Jvurnal of Virological Merho&. 30 ( 1990) 79--X8 79
Elsevier
VIRMJZT
01068
Use of monoclonal antibodies in an ELISA for the diagnosis of bovine leukaemia virus infection J. Ban, E. Gieciova,
0. Orlik and C. Altaner
Department of Molecubr Virology. Cancer Research Institute. Slovak Academy of Sciewes, Bratislava. Czechoslovakia (Accepted
19 June
1990)
Summary An ELISA diagnostic test for detection of bovine leukaemia virus (BLV) infected animals was developed. The test is based on the use of a mixture of monoclonal antibodies (MAbs) against envelope glycoprotein and against viral structural protein ~24. The sensitivity and specificity of the test were found to be dependent on the relative proportions of MAbs of the appropriate epitope specificity. Polystyrene microtitre plates, wells or sticks were firstly coated with a mixture of purified MAbs and then non-purified viral antigens were adsorbed from tissue culture fluid obtained from BLV-producing cells. The optimal conditions for adsorption of MAbs and viral antigens as well as for the ELISA procedure were established. The test is more sensitive and cheaper (no need for virus antigen purification) than the routinely used ELISA using purified virus antigens. The assay is highly specific, rapid, practical and could be easily automated. It is suitable for the detection of BLV-antibodies in blo~“serum or milk in the large-scale screening programs for BLV-infected animals. ELISA; Monoclonal ant&&dies; Bovine leukemia virus; Diagnostic test
Introduction Bovine leukaemia virus is the causative (for review, see Altaner, 1983; Ghysdael virus infection in cattle leads to leukosis cytosis, occurrence of antibodies directed _-. Correspondence IO: C. Academy
of Sciences.
016%8510/90/SOR.S(Kc)
agent of enzootic leukaemia of cattle et al., 1985; Bumy et al., 1987). The characterized by persistent lymphoagainst viral structural proteins, and
Altancr, Department of Molecular Virology, Cancer Research ul. csl. armady 2l., 812 32 Bratislava. Crechoslovakia.
1090 Elsevier Science Publishers B.V. (Biomedical
Division)
Institute,
Slovak
80
by the appearance of lymphoid turnouts after a long induction period in some infected animals. The viral infection is spread by contact between animals and therefore in most countries eradication of BLV-infected animals or whole herds is commonly practised. The diagnosis of BLV infection is based on the detection of viral antibodies by different methods. Since the eradication of infected cattle is the usual procedure, diagnosis must be made with methods of high sensitivity which certify the detection of early infected animals in which the level of antibodies is low. The antibody response in BLV-infected cattle is directed against viral glycoproteins at first but antibodies against other structural proteins have been found as well. From all different methods used for diagnostic purposes, the enzyme-linked immunosorbent assay (ELISA) (Engwall and Perlmann, 1971) is probably the most practical, being sensitive enough, simple, and easy to perform on large scale. The purity of BLV antigens used for the immunoassay is the most important factor which determines the quality of ELISA in terms of distinguishing clearly between seropositive and seronegative animals. The purification of BLV and its antigenic components to be appropriate for ELISA is rather difficult, mainly in terms of scale-up. Recently we produced several monoclonal antibodies against different epitopes of various proteins of BLV. In this report we describe the development of ELISA for detection of BLV infection using a mixture of monoclonal antibodies against gp5 1, gp30 and p24 of BLV. The procedure removes the need to purify antigens and is highly sensitive and practical for serodiagnosis of bovine leukaemia virus infection.
Materials
and Methods
Cells and virus
A cell clone of mouse myeloma cells Sp2/0 (Schulman et al., 1978) was used for hybridoma production. For bovine leukaemia virus production, the cells of highly virus-productive cell clone have been used (Altaner et al., 1987). The cell clone was originally derived from lamb kidney cell line FLK permanently infected with BLV (Van der Maaten and Miller, 1976). The BLV-producing rat cell line was recently established (Altanerova et al., 1989). Virus was purified by sucrose gradient centrifugation from tissue culture fluid harvested from roller cultures. Tissue culture medium containing 1% of fetal calf serum harvested from 3-day-old roller cultures at confluence was filtered through 0.45 p Millipore membrane and used as a non-purified BLV preparation. The virus-containing medium was disintegrated either by ultrasound (twice for 10 seconds at the highest output of 100 W MSE instrument), or by the addition of Tween-20 to be in final concentration of 0.5%.
81
Preparation of monoclonal antibodies Hybridoma cells producing monoclonal antibodies against BLV proteins were prepared by the usual hybridoma technology. Several modifications, described in detail recently (Orlik and Altaner 1988), were used to simplify the procedure which increased the yield of antibody producing cells. The detailed characterization of the whole panel of anti-BLV MAbs which had been prepared will be described elsewhere (Gieciova et al., 1990). The epitope specificity of monoclonal antibodies was determined by competition assays with MAbs of known epitope specificity’(Bruck et al., 1982). Production of monoclonal antibodies For mass production of MAbs, about 10 x 10” hybridoma cells were injected intraperitoneally in Pristane or incomplete Freund’s adjuvant-primed BALB/c mice, and ascitic fluid was collected from several mice 10-14 days later. The antibody concentration was determined by radioimmunoassay using polystyrene sticks coated with purified disrupted virus (Altaner et al., 1982). Affinity purified anti-mouse IgG labeled with “‘1 was used as second antibody in this assay. Pur@cation of monoclonal antibody The MAbs from ascitic fluid were purified by the chromatography on Protein A-Sepharose or Protein G-Sepharose gel columns (Pharmacia LKB, Uppsala) according to the manufacturer’s recommended procedure. From different methods tested these approaches were found to be the most suitable. ELBA test A mixture of monoclonal antibodies in PBS at pH 7.2 was adsorbed overnight at 4°C on the bottom of a 96-well microtitre plate or on the round part of polystyrene stick. The unbound antibody was discarded and the wells/sticks were washed once with 0.02 M Tris-HCl buffer, pH 7.5, containing 0.05% Tween-20 (washing buffer). The polystyrene surface was saturated with 3% Tween-20 in TEN buffer for 2 h at 37OC and washed three times with the washing buffer. The BLV antigens present in tissue culture fluid were bound either at 4OC overnight or at 37OC for 1 h. The coated solid phase was washed three times with washing buffer. The samples to be tested were diluted in 100 ~1 of reaction buffer (TEN containing 0.05% Tween-20) and incubated at 37OC for 1 h. The blood serum samples were diluted usually 25 times, milk was tested undiluted. The wells/sticks were then washed three times with washing buffer containing 0.05% Tween-20. The appropriate dilution of swine antibovine IgG conjugated with peroxidase (SWAB-Px) in TEN buffer containing 0.05% Tween-20 was added to wells and kept at 37°C for 1 h. Bound peroxidase was revealed by adding 100 ~1 of freshly prepared chromogenic substrate (3 mg of 3-amino-9-ethylcarbazole dissolved in
82
TABLE 1 List of anti-BLV MAbs used in development Monoclonal
antibody
BLVgpS I-6A 12/Al BLVgpS I-4E9/32A6 BLVgp5 I-25El l/2 BLVgp5 I-4E9/32Cl BLVgp30-94C 11 BLVp24-X 16C6A9 BLVp24-25E1 l/8 BLVp24-X48B3/C6
of the assay
Anti-
Isotype
Virus neutralization (titre)
Epitope specificity
gP51 gP5 1 gP5 1 gP5 1 gp30 ~24 ~24 ~24
IgGl IgG2b IgG 1 IgG 1 IgGl IgG 1 IgG 1 IgGl
5000 3000 3000 2000 0 0 0 0
G F,H H H new new 4’G9 new
2 ml of dimethylformamide) is added dropwise to 8 ml of 5 mM acetate buffer pH 4.8, or orthophenylendiamine (1 .O mg/ml in 0.05 M citrate-phosphate buffer, pH 4.5) in the presence of 0.01% hydrogen peroxide.
Results In ELISA, reaction takes place on a solid phase (usually polystyrene) to which components are sequentially exposed and successively bound. To detect antiviral antibody, the antigen may simply be adsorbed to the solid phase, or it may be selectively bound by specific antibody previously adsorbed to the solid phase. The latter process (known as double antibody sandwich procedure (Clark et al., 1986)) usually avoids the need to purify viral antigen before adsorption. The optimal conditions of the main steps for this type of assay by means of anti-BLV monoclonal antibody were investigated. Choice of MAhs To cover the spectrum of possible antiviral antibodies usually present in BLVinfected cattle, the following antibodies were chosen: MAbs directed against different epitopes of envelope glycoprotein gp51, MAb against transmembrane protein gp30 and MAbs directed against main viral structural protein ~24. Some properties of the MAbs used for the assay are presented in Table 1. The main feature which determines the quality and practical use of ELISA for detection of anti-BLV antibody in serum and milk is the level of non-specific signal detected with normal serum. Therefore the ratio of ELISA values of antibody positive to negative is a criterion for evaluation of the assay. Adsorption
of MAhs to solid phase
Tissue culture fluid harvested from hybridoma cells was not satisfactory for adsorption to the solid phase, since it gave low values. MAbs in the form of ascitic
83
fluid despite high end-point titers (in range of 10-40000 in RIA) were found also unsuitable for adsorption because of high background (data not shown). Therefore purified MAbs from ascitic fluids were used for adsorption to the solid phase only. Of the different purification procedures tested, chromatography on protein A-Sepharose, or Protein G-Sepharose gave the best results in terms of yield and quality expressed as a ratio of ELISA values. The pH of the buffer used for adsorption of MAbs was. found to be critical. Carbonate-bicarbonate buffer pH 9.6 commonly used for antigen adsorption was deleterious for the antigen-binding capacity of MAbs first adsorbed to the solid phase. The optimal pH values for adsorption were found to be in the range between 6.0 to 7.5. At increasing pH values, the amount of MAbS, bound to solid phase decreased. The MAbs dissolved in phosphate buffered saline (PBS) pH 7.2 were found optimal for the adsorption to solid phase. The MAbs were adsorbed to solid, polystyrene phase either at 37OC for 1 h or at 4OC overnight. Both techniques gave the same results. We prefer the adsorption at 4OC overnight. The lowest IgG concentration needed to give optimal results varied according to used MAbs. For MAb anti-gp51 and MAb anti-p24 the lowest amount of IgG giving still optimal result was found to be in range of 2-10 pg/ml or 8-10 &ml, respectively.
Adsorption of antigen There is no need to purify virus antigens for this type of assay. The sucrose gradient purified BLV did not perform better in comparison with the virus present in the tissue culture fluid from the virus producing cells. As-a source of BLV antigens the tissue culture fluid either disintegrated by ultrasound or by addition of Tween-20 from different virus-producing cells was used. No differences were found between the source of viral antigens produced either by ovine or bat cell lines. The virus produced from rat cell line was found to be less effective probably because of low titer and/or host-induced modification of gp51. The BLV-containing medium from ovine or bat cells was equally acceptable. The tissue culture fluid could be diluted at least 8 times without significant change in the efficiency (Table 2). Undiluted, BLV-c0ntainin.g tissue culture medium can be used repeatedly for antigen adsorption (data not shown). The virus antigen adsorption performed at 37OC for 1 h, or at 4’C overnight gave the same result.
Selection of propel” MAbs In order to establish the best combination of MAbs for the assay, several MAbs of different epitope specificity were tested. The positive and negative sera were titrated with different MAbs and with their cocktail (Fig. 1). The epitope specificity of MAbs was found to be important in their usability for the assay. This was true for MAbs directed against glycoprotein gp51 and MAbs against p24 protein as well. Some MAbs both against gp51 and p24 were found to be unsuitable for assay because of their low sensitivity. On the other hand, the
84
TABLE 2 Binding efficiency of viral antigens from different sources to monoclonal antibody-coated
solid phase
Dilution
Antigen Medium from BLV-nroducine. ovine cells FLK ’ v Medium from BLV-producing ovine cells; highly virus producing cell clone Bat cell line producing BLV Rat cell line producing BLV Medium from uninfected cells
None
2x
4x
6x
8X
20.09
14.0
13.8
10.1
8.0
21.0 13.0 6.0 1.0
16.0 12.0 4.0 NTb
12.5 10.5 3.0 NT
9.0 8.0 2.5 NT
8.0 7.5 2.0 NT
“Ratio of absorbancy values obtained in ELISA test with serum containing BLV antibodies/normal serum. bNT, not tested.
1.5
1 SERUM
3 DlLUTlON
(LOG,&
Fig. 1. Selection of different anti-BLV monoclonal antibodies for the use in the ELISA test. Different anti-gp51 MAbs, anti-p24 MAbs and a cocktail of MAbs were adsorbed to solid phase. The surface was coated with virus antigens. Negative and positive sera were titrated by ELISA procedure. The values represent ratio of absorbancies obtained with serum containing BLV antibodies/normal serum. (ratio P/N). A-A, BLV p24-Xl6C6A9: O-O, BLVgp5 I-6A12/A7: A-A, BLVp24X48B3/C6; l -e. BLVp24-25El l/8; o--o, cocktail BLVgp5 I-6A 12/A7+BLVp24-X48B3/C6.
others (for example BLV- gp5 1/6A12 and BLV-p24/X48) performed much better in the assay. The best results were obtained using a mixture of MAbs directed against epitope G of gp5 1 and undetermined epitope of ~24. The addition of gp30 specific MAb to the cocktail did not improve the assay (data not shown).
85 TABLE 3 Comparison of diagnostic reliability of classical and monoclonal antibody-based ELISA assays No. of sera
ELISA Classical
Monoclonal anti&
I
antibody-based anti-p24
anti-gp5l
+ + +
+ +
+ + +
+ -
+
57 7 4 3 2 6 74
+ + + f -
+ + -
Relative interception
89.5%
100%
+ anti-p24
-
+, Positive finding; f, dubious finding; -, negative finding.
Comparison
of MAb-based
ELISA with purified viral antigen-based
ELISA
In.order to determine the diagnostic value of the assay, a group of bovine sera and corresponding milk samples were tested with purified viral antigen-based ELISA (classical ELISA) and MAb-based ELISA. The data are summarized in Table 3. The relative sensitivity of this MAb-based ELISA increased about 10%. The test recognized animals infected at an early stage when the antibodies were either at low tines or were directed against viral glycoprotein only. The higher specificity of developed test contributed to the lower incidence of dubious cases which have been observed by classical ELISA. The data obtained with samples of serum and corresponding milk were similar.
Discussion There is a great need for a reliable diagnostic test which can detect reproducibly animals which have been infected with BLV. The large number of animals which need to be tested for BLV infection requires a simple, sensitive, reliable and inexpensive method. Those techniques which require more or less purified viral antigens, such as agar immunodiffusion, RIA and ELISA, the last probably fulfills most of the requirements. The sensitivity and reproducibility of the test depend on the quality of the viral antigen used for the assay. To prepare suitable viral antigens sufficiently pure, in large quantities, represents the main problem in preparation of diagnostic sets for screening purposes. Therefore the use of monoclonal antibodies against BLV antigens substantially solved the problem. A variant of the ELISA technique, involving a monoclonal anti-gp5 1 antibody and a competition ELISA involving two monoclonal anti-gp5 1 antibodies have both been developed for detection of BLV antibodies (Portetelle et al., 1983, 1989). The intention of our design of a monoclonal antibody-based ELISA for de-
86
tection of BLV-infected cattle was to include naturally occurring antibodies in infected animals as far as possible. At first we therefore chose a mixture of MAbs directed against envelope glycoproteins gp5 1, gp30 and main structural protein p24 for adsorption virus antigens from medium of virus-producing cells. It was important to choose the correct combination of epitope specificity. The presence of anti-gp30 MAb in .the cocktail did not improv.e the relative performance of the ELISA, therefore for the final technique for routine use the anti-gp30 MAb was not included. The low levels of anti-gp30 antibodies in bovine sera from infected cattle in comparison with the levels of the anti-gp51 and anti-p24 antibodies were the likely reason for this observation. The advantages of monoclonal antibody-based ELISA test for detection of BLV-infected cattle are obvious: 1. There is no need for purification of the virus antigens. 2. The ELISA procedure is standard because of the homogeneity of monoclonal antibodies. 3. The diagnostic tools (plates, sticks) can be prepared in advance and stored for the later use. 4. The procedure for serological detection is very simple, can be performed in a very short time (a few hours) and it can be done with many samples at once. It can also be easily automated. 5. The sensitivity of the assay allows detection of anti-BLV antibodies not only during early infection but also in samples of milk, as well as pooled samples of sera and milk. 6. The determination of antigp5 1 antibodies and anti-p24 antibodies in serum samples can be done separately which allows the monitoring of progression of natural infection of cattle. 7. The diagnostic kits can be produced at low cost and in unlimited amounts because of low amounts of MAbs and unpurified’virus antigens are needed for production. In order to eradicate bovine leukemia virus infection in cattle effectively it is necessary to detect all infected animals. Therefore a highly sensitive method for diagnosis must be used. The MAbs-based ELISA test is certainly the right choice at present.
Acknowledgements We would like to thank Mrs M. Zatkova, Ms. B. Puterova, Mrs H. Chorvatova and Mr M. Cebecauer for excellent technical assistance. We are indebted to Dr D. Portetelle for anti-gp51 monoclonal antibody of known epitope specificity, to Dr V. Janik for providing us samples of biological material from infected cattle and to Dr Val Jones for manuscript correction.
References Altaner, C. (1983) Bovine leukosis virus (BLV) (in Czech). Biol. Listy 48, 120-133. Altaner, C., Zajac, V. and Ban, J. (1982) A simple .and inexpensive method for detection of BLV infected cattle based on modified ELISA principle. Zbl. Vet. Med. B 29, 583-590. Altaner, C., Ban, J.: Zajac, V., Kettmann, R. and Bumy, A. (1985) Isolation and characterization of cell clones producing various amounts of bovine leukosis virus. Folia Biol. 31, 107-l 16. Altanerova, V., Portetelle, D., Kettmann, R. and Altaner, C. (1989) Infection of rats with bovine
87
leukaemia virus: establishment of a virus-producing rat cell line. J. Gen. Viral. 70, 1929-1932. Bruck, C., Mathot, S., Portetelle, D., Berte, C., Franssen, J.D., Herion, P. and Bumy, A. (1982) Monoclonal antibodies define eight independent antigenic regions on the bovine leukemia virus (BLV) envelope glycoprotein gp51. Virology 122, 342-352. Bumy, A., Cleuter, Y., Kettmann, R., Mammerickx, M., Marbaix, G., Portetelle, D., Van Den Broeke, A., Wiliems, L. and Thomas, R. (1987) Bovine leukemia virus: facts and hypotheses derived from the study of an infectious cancer. Cancer Surveys 6, 139-159. Clark, M.F., Lister, R.M. and Bar-Joseph M. (1986)ELISA technique. Methods Enzymol. 118,742-766. Engwall, E. and Perlmann, P. (1986) Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of IgG. Immunochemistry 8, 871-874. Ghysdael, J., Bruck, C., Kettmann, R. and Bumy, A. (1985) Bovine leukemia virus. Curr. Top. Microbial. Immunol. 112, l-19. Gieciova, E., Ban, J., Altanerova, V., Orlik, O., Portetelle, D. and Altaner, C. (1990) Characterization of monoclonal antibodies directed against bovine leukemia virus proteins (submitted for publication). Orlik, 0. and Altaner, C. (1988) Modifications of hybridoma technology which improve the yield of monoclonal antibody producing cells. J. Immunol. Methods 115, 55-59. Portetelle, D., Bruck, C., Mammerickx, M. and Bumy, A. (1983) Use of monoclonal antibody in an ELISA test for the detection of antibodies to bovine leukaemia virus. J. Virol. Methods 6, 19-29. Portetelle, D., Mammerickx, M. and Bumy, A. (1989) Use of two monoclonal antibodies in an ELISA test for the detection of antibodies to bovine leukaemia virus envelope protein gp51. J. Viral. Methods 23, 21 l-222. Schulman, C.D., Wilde, CD. and Kohler, G. (1978) A better cell line for making of hybridomas secreting specific antibodies. Nature 276, 269-270. Van der Maaten, M.J. and Miller, J.M. (1976) Replication of bovine leukemia virus in monolayer cell cultures. Bibl. Haematol. 43, 360-362.
Elsevier
VIRMJZT
01068
Use of monoclonal antibodies in an ELISA for the diagnosis of bovine leukaemia virus infection J. Ban, E. Gieciova,
0. Orlik and C. Altaner
Department of Molecubr Virology. Cancer Research Institute. Slovak Academy of Sciewes, Bratislava. Czechoslovakia (Accepted
19 June
1990)
Summary An ELISA diagnostic test for detection of bovine leukaemia virus (BLV) infected animals was developed. The test is based on the use of a mixture of monoclonal antibodies (MAbs) against envelope glycoprotein and against viral structural protein ~24. The sensitivity and specificity of the test were found to be dependent on the relative proportions of MAbs of the appropriate epitope specificity. Polystyrene microtitre plates, wells or sticks were firstly coated with a mixture of purified MAbs and then non-purified viral antigens were adsorbed from tissue culture fluid obtained from BLV-producing cells. The optimal conditions for adsorption of MAbs and viral antigens as well as for the ELISA procedure were established. The test is more sensitive and cheaper (no need for virus antigen purification) than the routinely used ELISA using purified virus antigens. The assay is highly specific, rapid, practical and could be easily automated. It is suitable for the detection of BLV-antibodies in blo~“serum or milk in the large-scale screening programs for BLV-infected animals. ELISA; Monoclonal ant&&dies; Bovine leukemia virus; Diagnostic test
Introduction Bovine leukaemia virus is the causative (for review, see Altaner, 1983; Ghysdael virus infection in cattle leads to leukosis cytosis, occurrence of antibodies directed _-. Correspondence IO: C. Academy
of Sciences.
016%8510/90/SOR.S(Kc)
agent of enzootic leukaemia of cattle et al., 1985; Bumy et al., 1987). The characterized by persistent lymphoagainst viral structural proteins, and
Altancr, Department of Molecular Virology, Cancer Research ul. csl. armady 2l., 812 32 Bratislava. Crechoslovakia.
1090 Elsevier Science Publishers B.V. (Biomedical
Division)
Institute,
Slovak
80
by the appearance of lymphoid turnouts after a long induction period in some infected animals. The viral infection is spread by contact between animals and therefore in most countries eradication of BLV-infected animals or whole herds is commonly practised. The diagnosis of BLV infection is based on the detection of viral antibodies by different methods. Since the eradication of infected cattle is the usual procedure, diagnosis must be made with methods of high sensitivity which certify the detection of early infected animals in which the level of antibodies is low. The antibody response in BLV-infected cattle is directed against viral glycoproteins at first but antibodies against other structural proteins have been found as well. From all different methods used for diagnostic purposes, the enzyme-linked immunosorbent assay (ELISA) (Engwall and Perlmann, 1971) is probably the most practical, being sensitive enough, simple, and easy to perform on large scale. The purity of BLV antigens used for the immunoassay is the most important factor which determines the quality of ELISA in terms of distinguishing clearly between seropositive and seronegative animals. The purification of BLV and its antigenic components to be appropriate for ELISA is rather difficult, mainly in terms of scale-up. Recently we produced several monoclonal antibodies against different epitopes of various proteins of BLV. In this report we describe the development of ELISA for detection of BLV infection using a mixture of monoclonal antibodies against gp5 1, gp30 and p24 of BLV. The procedure removes the need to purify antigens and is highly sensitive and practical for serodiagnosis of bovine leukaemia virus infection.
Materials
and Methods
Cells and virus
A cell clone of mouse myeloma cells Sp2/0 (Schulman et al., 1978) was used for hybridoma production. For bovine leukaemia virus production, the cells of highly virus-productive cell clone have been used (Altaner et al., 1987). The cell clone was originally derived from lamb kidney cell line FLK permanently infected with BLV (Van der Maaten and Miller, 1976). The BLV-producing rat cell line was recently established (Altanerova et al., 1989). Virus was purified by sucrose gradient centrifugation from tissue culture fluid harvested from roller cultures. Tissue culture medium containing 1% of fetal calf serum harvested from 3-day-old roller cultures at confluence was filtered through 0.45 p Millipore membrane and used as a non-purified BLV preparation. The virus-containing medium was disintegrated either by ultrasound (twice for 10 seconds at the highest output of 100 W MSE instrument), or by the addition of Tween-20 to be in final concentration of 0.5%.
81
Preparation of monoclonal antibodies Hybridoma cells producing monoclonal antibodies against BLV proteins were prepared by the usual hybridoma technology. Several modifications, described in detail recently (Orlik and Altaner 1988), were used to simplify the procedure which increased the yield of antibody producing cells. The detailed characterization of the whole panel of anti-BLV MAbs which had been prepared will be described elsewhere (Gieciova et al., 1990). The epitope specificity of monoclonal antibodies was determined by competition assays with MAbs of known epitope specificity’(Bruck et al., 1982). Production of monoclonal antibodies For mass production of MAbs, about 10 x 10” hybridoma cells were injected intraperitoneally in Pristane or incomplete Freund’s adjuvant-primed BALB/c mice, and ascitic fluid was collected from several mice 10-14 days later. The antibody concentration was determined by radioimmunoassay using polystyrene sticks coated with purified disrupted virus (Altaner et al., 1982). Affinity purified anti-mouse IgG labeled with “‘1 was used as second antibody in this assay. Pur@cation of monoclonal antibody The MAbs from ascitic fluid were purified by the chromatography on Protein A-Sepharose or Protein G-Sepharose gel columns (Pharmacia LKB, Uppsala) according to the manufacturer’s recommended procedure. From different methods tested these approaches were found to be the most suitable. ELBA test A mixture of monoclonal antibodies in PBS at pH 7.2 was adsorbed overnight at 4°C on the bottom of a 96-well microtitre plate or on the round part of polystyrene stick. The unbound antibody was discarded and the wells/sticks were washed once with 0.02 M Tris-HCl buffer, pH 7.5, containing 0.05% Tween-20 (washing buffer). The polystyrene surface was saturated with 3% Tween-20 in TEN buffer for 2 h at 37OC and washed three times with the washing buffer. The BLV antigens present in tissue culture fluid were bound either at 4OC overnight or at 37OC for 1 h. The coated solid phase was washed three times with washing buffer. The samples to be tested were diluted in 100 ~1 of reaction buffer (TEN containing 0.05% Tween-20) and incubated at 37OC for 1 h. The blood serum samples were diluted usually 25 times, milk was tested undiluted. The wells/sticks were then washed three times with washing buffer containing 0.05% Tween-20. The appropriate dilution of swine antibovine IgG conjugated with peroxidase (SWAB-Px) in TEN buffer containing 0.05% Tween-20 was added to wells and kept at 37°C for 1 h. Bound peroxidase was revealed by adding 100 ~1 of freshly prepared chromogenic substrate (3 mg of 3-amino-9-ethylcarbazole dissolved in
82
TABLE 1 List of anti-BLV MAbs used in development Monoclonal
antibody
BLVgpS I-6A 12/Al BLVgpS I-4E9/32A6 BLVgp5 I-25El l/2 BLVgp5 I-4E9/32Cl BLVgp30-94C 11 BLVp24-X 16C6A9 BLVp24-25E1 l/8 BLVp24-X48B3/C6
of the assay
Anti-
Isotype
Virus neutralization (titre)
Epitope specificity
gP51 gP5 1 gP5 1 gP5 1 gp30 ~24 ~24 ~24
IgGl IgG2b IgG 1 IgG 1 IgGl IgG 1 IgG 1 IgGl
5000 3000 3000 2000 0 0 0 0
G F,H H H new new 4’G9 new
2 ml of dimethylformamide) is added dropwise to 8 ml of 5 mM acetate buffer pH 4.8, or orthophenylendiamine (1 .O mg/ml in 0.05 M citrate-phosphate buffer, pH 4.5) in the presence of 0.01% hydrogen peroxide.
Results In ELISA, reaction takes place on a solid phase (usually polystyrene) to which components are sequentially exposed and successively bound. To detect antiviral antibody, the antigen may simply be adsorbed to the solid phase, or it may be selectively bound by specific antibody previously adsorbed to the solid phase. The latter process (known as double antibody sandwich procedure (Clark et al., 1986)) usually avoids the need to purify viral antigen before adsorption. The optimal conditions of the main steps for this type of assay by means of anti-BLV monoclonal antibody were investigated. Choice of MAhs To cover the spectrum of possible antiviral antibodies usually present in BLVinfected cattle, the following antibodies were chosen: MAbs directed against different epitopes of envelope glycoprotein gp51, MAb against transmembrane protein gp30 and MAbs directed against main viral structural protein ~24. Some properties of the MAbs used for the assay are presented in Table 1. The main feature which determines the quality and practical use of ELISA for detection of anti-BLV antibody in serum and milk is the level of non-specific signal detected with normal serum. Therefore the ratio of ELISA values of antibody positive to negative is a criterion for evaluation of the assay. Adsorption
of MAhs to solid phase
Tissue culture fluid harvested from hybridoma cells was not satisfactory for adsorption to the solid phase, since it gave low values. MAbs in the form of ascitic
83
fluid despite high end-point titers (in range of 10-40000 in RIA) were found also unsuitable for adsorption because of high background (data not shown). Therefore purified MAbs from ascitic fluids were used for adsorption to the solid phase only. Of the different purification procedures tested, chromatography on protein A-Sepharose, or Protein G-Sepharose gave the best results in terms of yield and quality expressed as a ratio of ELISA values. The pH of the buffer used for adsorption of MAbs was. found to be critical. Carbonate-bicarbonate buffer pH 9.6 commonly used for antigen adsorption was deleterious for the antigen-binding capacity of MAbs first adsorbed to the solid phase. The optimal pH values for adsorption were found to be in the range between 6.0 to 7.5. At increasing pH values, the amount of MAbS, bound to solid phase decreased. The MAbs dissolved in phosphate buffered saline (PBS) pH 7.2 were found optimal for the adsorption to solid phase. The MAbs were adsorbed to solid, polystyrene phase either at 37OC for 1 h or at 4OC overnight. Both techniques gave the same results. We prefer the adsorption at 4OC overnight. The lowest IgG concentration needed to give optimal results varied according to used MAbs. For MAb anti-gp51 and MAb anti-p24 the lowest amount of IgG giving still optimal result was found to be in range of 2-10 pg/ml or 8-10 &ml, respectively.
Adsorption of antigen There is no need to purify virus antigens for this type of assay. The sucrose gradient purified BLV did not perform better in comparison with the virus present in the tissue culture fluid from the virus producing cells. As-a source of BLV antigens the tissue culture fluid either disintegrated by ultrasound or by addition of Tween-20 from different virus-producing cells was used. No differences were found between the source of viral antigens produced either by ovine or bat cell lines. The virus produced from rat cell line was found to be less effective probably because of low titer and/or host-induced modification of gp51. The BLV-containing medium from ovine or bat cells was equally acceptable. The tissue culture fluid could be diluted at least 8 times without significant change in the efficiency (Table 2). Undiluted, BLV-c0ntainin.g tissue culture medium can be used repeatedly for antigen adsorption (data not shown). The virus antigen adsorption performed at 37OC for 1 h, or at 4’C overnight gave the same result.
Selection of propel” MAbs In order to establish the best combination of MAbs for the assay, several MAbs of different epitope specificity were tested. The positive and negative sera were titrated with different MAbs and with their cocktail (Fig. 1). The epitope specificity of MAbs was found to be important in their usability for the assay. This was true for MAbs directed against glycoprotein gp51 and MAbs against p24 protein as well. Some MAbs both against gp51 and p24 were found to be unsuitable for assay because of their low sensitivity. On the other hand, the
84
TABLE 2 Binding efficiency of viral antigens from different sources to monoclonal antibody-coated
solid phase
Dilution
Antigen Medium from BLV-nroducine. ovine cells FLK ’ v Medium from BLV-producing ovine cells; highly virus producing cell clone Bat cell line producing BLV Rat cell line producing BLV Medium from uninfected cells
None
2x
4x
6x
8X
20.09
14.0
13.8
10.1
8.0
21.0 13.0 6.0 1.0
16.0 12.0 4.0 NTb
12.5 10.5 3.0 NT
9.0 8.0 2.5 NT
8.0 7.5 2.0 NT
“Ratio of absorbancy values obtained in ELISA test with serum containing BLV antibodies/normal serum. bNT, not tested.
1.5
1 SERUM
3 DlLUTlON
(LOG,&
Fig. 1. Selection of different anti-BLV monoclonal antibodies for the use in the ELISA test. Different anti-gp51 MAbs, anti-p24 MAbs and a cocktail of MAbs were adsorbed to solid phase. The surface was coated with virus antigens. Negative and positive sera were titrated by ELISA procedure. The values represent ratio of absorbancies obtained with serum containing BLV antibodies/normal serum. (ratio P/N). A-A, BLV p24-Xl6C6A9: O-O, BLVgp5 I-6A12/A7: A-A, BLVp24X48B3/C6; l -e. BLVp24-25El l/8; o--o, cocktail BLVgp5 I-6A 12/A7+BLVp24-X48B3/C6.
others (for example BLV- gp5 1/6A12 and BLV-p24/X48) performed much better in the assay. The best results were obtained using a mixture of MAbs directed against epitope G of gp5 1 and undetermined epitope of ~24. The addition of gp30 specific MAb to the cocktail did not improve the assay (data not shown).
85 TABLE 3 Comparison of diagnostic reliability of classical and monoclonal antibody-based ELISA assays No. of sera
ELISA Classical
Monoclonal anti&
I
antibody-based anti-p24
anti-gp5l
+ + +
+ +
+ + +
+ -
+
57 7 4 3 2 6 74
+ + + f -
+ + -
Relative interception
89.5%
100%
+ anti-p24
-
+, Positive finding; f, dubious finding; -, negative finding.
Comparison
of MAb-based
ELISA with purified viral antigen-based
ELISA
In.order to determine the diagnostic value of the assay, a group of bovine sera and corresponding milk samples were tested with purified viral antigen-based ELISA (classical ELISA) and MAb-based ELISA. The data are summarized in Table 3. The relative sensitivity of this MAb-based ELISA increased about 10%. The test recognized animals infected at an early stage when the antibodies were either at low tines or were directed against viral glycoprotein only. The higher specificity of developed test contributed to the lower incidence of dubious cases which have been observed by classical ELISA. The data obtained with samples of serum and corresponding milk were similar.
Discussion There is a great need for a reliable diagnostic test which can detect reproducibly animals which have been infected with BLV. The large number of animals which need to be tested for BLV infection requires a simple, sensitive, reliable and inexpensive method. Those techniques which require more or less purified viral antigens, such as agar immunodiffusion, RIA and ELISA, the last probably fulfills most of the requirements. The sensitivity and reproducibility of the test depend on the quality of the viral antigen used for the assay. To prepare suitable viral antigens sufficiently pure, in large quantities, represents the main problem in preparation of diagnostic sets for screening purposes. Therefore the use of monoclonal antibodies against BLV antigens substantially solved the problem. A variant of the ELISA technique, involving a monoclonal anti-gp5 1 antibody and a competition ELISA involving two monoclonal anti-gp5 1 antibodies have both been developed for detection of BLV antibodies (Portetelle et al., 1983, 1989). The intention of our design of a monoclonal antibody-based ELISA for de-
86
tection of BLV-infected cattle was to include naturally occurring antibodies in infected animals as far as possible. At first we therefore chose a mixture of MAbs directed against envelope glycoproteins gp5 1, gp30 and main structural protein p24 for adsorption virus antigens from medium of virus-producing cells. It was important to choose the correct combination of epitope specificity. The presence of anti-gp30 MAb in .the cocktail did not improv.e the relative performance of the ELISA, therefore for the final technique for routine use the anti-gp30 MAb was not included. The low levels of anti-gp30 antibodies in bovine sera from infected cattle in comparison with the levels of the anti-gp51 and anti-p24 antibodies were the likely reason for this observation. The advantages of monoclonal antibody-based ELISA test for detection of BLV-infected cattle are obvious: 1. There is no need for purification of the virus antigens. 2. The ELISA procedure is standard because of the homogeneity of monoclonal antibodies. 3. The diagnostic tools (plates, sticks) can be prepared in advance and stored for the later use. 4. The procedure for serological detection is very simple, can be performed in a very short time (a few hours) and it can be done with many samples at once. It can also be easily automated. 5. The sensitivity of the assay allows detection of anti-BLV antibodies not only during early infection but also in samples of milk, as well as pooled samples of sera and milk. 6. The determination of antigp5 1 antibodies and anti-p24 antibodies in serum samples can be done separately which allows the monitoring of progression of natural infection of cattle. 7. The diagnostic kits can be produced at low cost and in unlimited amounts because of low amounts of MAbs and unpurified’virus antigens are needed for production. In order to eradicate bovine leukemia virus infection in cattle effectively it is necessary to detect all infected animals. Therefore a highly sensitive method for diagnosis must be used. The MAbs-based ELISA test is certainly the right choice at present.
Acknowledgements We would like to thank Mrs M. Zatkova, Ms. B. Puterova, Mrs H. Chorvatova and Mr M. Cebecauer for excellent technical assistance. We are indebted to Dr D. Portetelle for anti-gp51 monoclonal antibody of known epitope specificity, to Dr V. Janik for providing us samples of biological material from infected cattle and to Dr Val Jones for manuscript correction.
References Altaner, C. (1983) Bovine leukosis virus (BLV) (in Czech). Biol. Listy 48, 120-133. Altaner, C., Zajac, V. and Ban, J. (1982) A simple .and inexpensive method for detection of BLV infected cattle based on modified ELISA principle. Zbl. Vet. Med. B 29, 583-590. Altaner, C., Ban, J.: Zajac, V., Kettmann, R. and Bumy, A. (1985) Isolation and characterization of cell clones producing various amounts of bovine leukosis virus. Folia Biol. 31, 107-l 16. Altanerova, V., Portetelle, D., Kettmann, R. and Altaner, C. (1989) Infection of rats with bovine
87
leukaemia virus: establishment of a virus-producing rat cell line. J. Gen. Viral. 70, 1929-1932. Bruck, C., Mathot, S., Portetelle, D., Berte, C., Franssen, J.D., Herion, P. and Bumy, A. (1982) Monoclonal antibodies define eight independent antigenic regions on the bovine leukemia virus (BLV) envelope glycoprotein gp51. Virology 122, 342-352. Bumy, A., Cleuter, Y., Kettmann, R., Mammerickx, M., Marbaix, G., Portetelle, D., Van Den Broeke, A., Wiliems, L. and Thomas, R. (1987) Bovine leukemia virus: facts and hypotheses derived from the study of an infectious cancer. Cancer Surveys 6, 139-159. Clark, M.F., Lister, R.M. and Bar-Joseph M. (1986)ELISA technique. Methods Enzymol. 118,742-766. Engwall, E. and Perlmann, P. (1986) Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of IgG. Immunochemistry 8, 871-874. Ghysdael, J., Bruck, C., Kettmann, R. and Bumy, A. (1985) Bovine leukemia virus. Curr. Top. Microbial. Immunol. 112, l-19. Gieciova, E., Ban, J., Altanerova, V., Orlik, O., Portetelle, D. and Altaner, C. (1990) Characterization of monoclonal antibodies directed against bovine leukemia virus proteins (submitted for publication). Orlik, 0. and Altaner, C. (1988) Modifications of hybridoma technology which improve the yield of monoclonal antibody producing cells. J. Immunol. Methods 115, 55-59. Portetelle, D., Bruck, C., Mammerickx, M. and Bumy, A. (1983) Use of monoclonal antibody in an ELISA test for the detection of antibodies to bovine leukaemia virus. J. Virol. Methods 6, 19-29. Portetelle, D., Mammerickx, M. and Bumy, A. (1989) Use of two monoclonal antibodies in an ELISA test for the detection of antibodies to bovine leukaemia virus envelope protein gp51. J. Viral. Methods 23, 21 l-222. Schulman, C.D., Wilde, CD. and Kohler, G. (1978) A better cell line for making of hybridomas secreting specific antibodies. Nature 276, 269-270. Van der Maaten, M.J. and Miller, J.M. (1976) Replication of bovine leukemia virus in monolayer cell cultures. Bibl. Haematol. 43, 360-362.
