Immunology Letters, 28 (1991) 65-72 Elsevier IMLET 01556
The production of human monoclonal anti-MMTV antibodies by in vitro immunization with anti-idiotypic antibodies M. B l a n k ~, I. N. S m o r o d i n s k y 2, I. Keydar 2, S. C h a i t c h i k 1 a n d Y. S h o e n f e l d 1 IResearch Unit, Department of Medicine "B', Chaim Sheba Medical Center, TeI-Hashomer, Israel; and 2Department of Microbiology, George Wise Faculty of Life Sciences, Tel-Aviv University, TeI-Aviv, Israel (Received 12 December 1990; accepted 15 December 1990)
1. Summary We have previously reported the production of a series of human monoclonal antibodies reacting with mouse mammary tumor viral antigens as well as the human counterpart H u M T V antigen. Against two of these antibodies (Bll and 4.6/6), antiidiotypic antibodies were generated, which appeared to be the internal image of viral antigen. Vaccination with the anti-idiotypic antibodies induced in mice humoral and cellular immunity against both MMTV and HuMTV. The current study describes a novel method to produce human anti-MMTV antibodies derived monoclonal antibodies. Normal peripheral blood lymphocytes (PBLs) were immunized in vitro in the presence of IL-2, with rabbit anti-Bll antibodies bound to silica beads. Following in vitro immunization, the lymphocytes were fused with the UC-792-6 human lymphoblastoid cell line. Four human hybridomas were found to secrete human monoclonal antibodies which bound to MMTV and HuMTV. Three of the secreted antibodies were of the IgG class and one of the IgM class. The specificity of binding was confirmed by direct and competitive ELISA assays and by radioimmunoprecipitation. Our study demonstrates the ability to generate hu-
Key words: Breast cancer; Hybridoma; HuMTV; MMTV; Antiidiotype; In vitro immunisation
Correspondence to: Y. Shoenfeld, M.D., Head, Dept. of Medicine "B", Chaim Sheba Medical Center, Tel-Hashomer 52621, Israel. 0165-2478 / 91 / $ 3.50 © 1991 Elsevier Science Publishers BN.
man monoclonal anti-viral antibodies by in vitro immunization, employing "internal image" antiidiotypic antibodies. The method can be used to generate human antibodies, especially against pathogenic agents or ill-defined antigens. 2. Introduction Several studies have suggested that a relationship exists between mouse mammary tumor virus (MMTV) and human breast cancer [1, 2]. Support for this notion comes from the ability to generate human monoclonal antibodies (MoAbs) with lymphocytes derived from either patients with breast cancer [3, 4] or laboratory workers exposed to MMTV and H u M T V [5] for prolonged periods [6, 7]. The human MoAbs reacting with MMTV or with its human counterpart H u M T V antigens [5] were established from a fusion in which the following sources of lymphocytes were fused: reactive but non-infiltrated lymph node [3], peripheral blood [7] and pleural effusion [4]. Analysis of the various human anti-MMTV monoclonal antibodies led to the generation of antiidiotypic antibodies [8] that were found to be the internal image of the MMTV viral antigen. Immunization of mice with these anti-idiotypic antibodies induced a humoral and cellular state of immunity against both MMTV and H u M T V [8]. Preliminary studies of in vitro immunization of PBLs with H u M T V [9], as well as the above studies, provided the impetus to try to generate human anti-MMTV antibody following in vitro immunization of normal PBLs with internal image anti-idiotypic antibodies. 65
In vitro immunization of normal PBLs with the rabbit anti-idiotypic antibody to Bll (R-anti-Bll-Id) [3] and their fusion with a human lymphoblastoid cell line, yielded human hybridomas secreting monoclonal anti-MMTV antibodies. This method might be useful in the preparation of antibodies to any viral antigen. 3. Materials and Methods
3.1. Antibodies Antibodies used were: Bll. Details of the production of the human MoAb Bll have been reported elsewhere [3]. Bll antibody reacts with the gp52 and gp36 envelope glycoproteins of the MMTV, as well as with several core proteins [3], but did not react with other retroviruses tested. The specificity of the binding was confirmed by radioimmunoprecipitation and competition assays. BR2. A human IgM MoAb prepared from PBLs of a breast cancer patient that was found to bind MMTV and HuMTV. DIN. A human IgG MoAb derived from PBLs of a patient with drug induced lupus disease. 3.2. Anti-Bll antibody Details of the preparation of the rabbit anti-Bll anti-idiotypic antibody have been reported previously [8]. Briefly, 2-month-old rabbits were subcutaneously immunized with 50 #g of purified Bll MoAb per injection in complete Freund's adjuvant (CFA) followed by immunization with the MoAb in incomplete Freund's adjuvant at intervals of 3 weeks. Sera were obtained and precipitated with 40% saturated ammonium sulphate (SAS), dialyzed against phosphate-buffered saline (PBS) and extensively absorbed on a Sepharose column conjugated to pooled human immunoglobulin, and later on, on a fetal calf serum (FCS) Sepharose column [8]. The flow-through fluid was further affinity purified on a column consisting of polyacrylhydrazide (PAH) agarose to which Bll antibodies were conjugated. Elution of the anti-Bll antibody was performed with 0.1 N acetic acid in 0.9% NaC1 (pH 3.3). The eluates were titrated to pH 5.0 with phosphate buffer (0.1 M, pH 8.0) and dialyzed against PBS. The specificity of the binding was confirmed by ELISA assay [8]. 66
3.3. Coupling of anti-Bll antibody to silica beads Fumed SiO 2 beads (Sigma, size 0.007/~) were suspended in water (10 mg/ml) and sonicated. Then, 5 mg/ml of anti-B11 affinity purified antibody was added and incubated for 18 h at 4 °C with gentle shaking. The silica beads with the attached protein were collected by centrifugation (10000 x g for 5 rain) and washed repeatedly with distilled water. The supernatant (after the first centrifugation) was tested for the presence of anti-Bll antibodies by ELISA. 3.4. In vitro immunization Peripheral blood lymphocytes (PBLs) were isolated from healthy donor by Ficoll-hypaque gradient separation. 107 PBLs were incubated during the first 24 h in 1 ml RPMI-1640 medium containing 0.5 mg anti-Bll antibodies attached to 1 mg of SiO 2 beads/l%0 inactivated (45 min, 5 6 ° C ) a u t o l o gous serum/2 mM glutamine/2 mM non-essential amino acids/2 mM sodium pyruvate/50 units/ml insulin/5 × 105 mercaptoethanol/20 mM glutathione/20 mM each of the following: adenosine, guanosine, uridine and cytidine/3 mM thymidine/30 mM hypoxanthine and 100 units/ml penicillin/streptomycin. After 24 h, the volume was increased to a final volume of 20 ml with the above medium containing 3% of PHA-induced MLR supernatant. The cultures were incubated for 4 more days at 37 °C in an atmosphere of 5% CO2 in air. 3.5. Filter immunoplaque assay In order to confirm the efficiency of the in vitro immunization, the filter immunoplaque assay was performed prior to hybridization. The plaque assay is based on enzyme immunoassay principles as previously described [10]. Briefly, nitrocellulose membranes (type HATF, pore size 0.45 m) were coated with 10 mg/0.2 ml of R-anti-Bll-Id, HuMTV, MMTV, hepatitis B surface antigen or normal rabbit serum. The antigen-coated nitrocellulose membranes were washed with PBS and put into Petri dishes (3 cm, Nunc). Antibody-producing cells (1 x 106 cells per dish in 1 ml enriched RPMI-1640 medium: 10%0FCS/2 mM glutamine/2 mM sodium pyruvate/2 mM essential amino acids/100 u/ml
penicillin/streptomycin) were added in triplicate for 6 h in a humidified incubator with 5% CO2. The nitrocellulose papers were then washed with PBS contain.ing 0.5% Tween 20 and the spots indicating cells that secrete antibodies could be visualized using horseradish peroxidase conjugated to rabbit antihuman immunoglobulins (Sigma) and 4 chloro-1naphthol as substrate (Sigma). Blue plaques appeared within 3 min and the enzymatic reaction was terminated after 15 - 30 min by washing the nitrocellulose membrane with distilled water. The filter membranes were dried and the blue plaques were counted in the microscope by using 25 × magnification. The total immunoglobulin concentration was determined by ELISA employing a titration curve, analyzing supernatants with identical numbers of cells incubated at the conditions without the nitrocellulose. 3.6. Cell fusion A single cell suspension of lymphocytes, immunized in vitro with rabbit-anti-Bll antibodies, was fused with the UC-792-6 human lymphoblastoid cell line according to the method described previously by us [3, 7, 11-13]. After the fusion, the cells in HAT medium (hypoxanthine, aminopterin, thymidine) were seeded into either 0.2 or 2.0 ml wells of tissue culture plates (Nunc), in density of 2× 105 lymphoblastoid cells per ml. The HAT medium was replaced every 5 days. Four weeks later when hybrids were observed microscopically, feeding was continued with the same medium without aminopterin. All clones were subjected twice to limiting dilution procedure in regular medium. Subsequent cultures were grown in vertical 75-mi flasks, and in roller bottles.
3.7. Purification of hybridoma antibodies Culture fluids from each hybridoma cell line were collected, the immunoglobulins were precipitated in 50% saturated ammonium sulphate, redissolved in, and dialyzed against PBS. The antibodies were then affinity-purified on goat anti-human IgG or IgM Sepharose columns (Sigma, St. Louis, MO). The immunoglobulins bound to the column were eluted with 5 M MgC12 dialyzed against three changes of TBS (10 mM Tris/0.15 M NaC1, pH 7.3), and con-
centrated by vacuum dialysis. The immunoglobulins in the eluates were quantitated by the ELISA technique as previously reported [12]. 3.8. Antigen For HuMTV, soluble proteins were derived from the media of T47D cell line (human breast carcinoma) as previously described [5, 14]. The culture media were precipitated with 65% saturated ammonium sulphate. The precipitated proteins were dissolved in, and dialyzed against PBS, pH 7.4. MMTV was prepared as detailed elsewhere [4]. 3.9. Direct b&ding of human monoclonal antibod-
ies to R-an ti-Bll M M T V and H u M T V antigens The monoclonal antibodies produced following the in vitro immunization (MIVs) and the serum of the individuals whose lymphocytes were used for in vitro sensitization were tested for their binding capacity to R-anti-Bll Id, MMTV and HuMTV antigens by the ELISA method. Polystyrene microtiter plates with 96 flat-bottom wells (Nunc) were coated with 10 #g/ml of R-anti-Bll-Id, HuMTV or MMTV in 0.05 M bicarbonate pH 9.6 and incubated overnight at 4 °C. Following three washes with PBS and PBS containing 1%0 Tween-20 (T-PBS), the plates were blocked with 0.5% gelatine for 2 h at room temperature. Then the MIV MoAbs and the control monoclonal and polyclonal antibodies, were added to the plates at different dilutions for 2 h at room temperature. Following three additional washes with PBS and T-PBS, goat anti-human IgG or IgM conjugated to alkaline phosphatase (Sigma) diluted 1:1000 was added to the wells and incubated for 2 h at room temperature. Bound alkaline phosphatase conjugate was detected by adding substrate (pnitrophenyl phosphate-disodium). Absorbance was determined at 405 nm. As positive control we used rabbit anti-MMTV antibody and enzyme linked goat anti-rabbit IgG as a secondary antibody. 3.10. Inhibition of binding of anti-Bll antibodies to
MoAb Different concentrations (0.1-10 #g) of MoAbs MIV (A1, A5, A7, B2, E3, E5), BR2 and Bll in 25 ml 67
PBS were incubated with 25 ml of 1:500 dilutions of rabbit anti-Bll affinity purified antibodies for 5 h at room temperature. Fifty/~1 of the final competitortreated preparations was then added to flatbottomed polystyrene microtiter plates (Nunc) previously coated with 50/A of 10/zg//xl Bll and blocked with 0.5o70 gelatine. Following overnight incubation at 4°C, alkaline phosphatase goat antirabbit IgG (Sigma) diluted 1:1000, was added for 2 h at room temperature. Bound alkaline-phosphatase conjugate was detected by adding the appropriate substrate. Absorbance was read at 405 nm.
3.11. Iodination of proteins MMTV was labelled with 125I using the Iodogen method [15]. MMTV (100 #g) was iodinated with 1 mCi 125I(Amersham, U.K.). Free 125Iwas removed by centrifugation of iodinated protein through a Sephadex G-25 coarse 1-ml column. 3.12. Radioimmunoprecipitation of M M T V The human MoAbs (MIVs), established by the in vitro immunization with the rabbit-anti-Bll were characterized for their ability to precipitate iodinated MMTV. As positive controls, rabbit-anti-MMTV and rabbit anti-gp52 antibody were employed. As negative controls, we used autologous donor sera, IgG secreted by the UC-792-6 lymphoblastoid cell line and other MIV MoAbs which do not bind to MMTV by ELISA assay. 125I-labelled MMTV (2.5×106 cpm/assay) was immunoprecipitated as follows: the iodinated protein was preabsorbed with either normal human sera or rabbit serum, followed by incubation with formalin-fixed Staphylococcus aureus Cowan-1 strain and centrifugation at 10000 rev./min. The pellets were discarded and to the preabsorbed antigen, 100 t~g/assay of human MoAbs or rabbit anti-MMTV or anti-gp52 antibodies were added and incubated for one hour at room temperature (23 °C) followed by 15 min incubation with Staphylococcus aureus at 37 °C. The mixtures were centrifuged through 1 ml cushions of 1 M sucrose in TNE buffer pH 7.4 (0.01 M TrisHC1/0.15 M NaC1/1 mM EDTA) containing 0.5% Nonidet P-40 (NP-40) and 0.5% deoxycholate. The pellets were washed three times with TNE buffer pH 7.4 containing 0.5% NP-40, and finally with TNE 68
buffer. The washed pellets were counted in a gamma scintillation counter, then resuspended in 40 ml of sample buffer (0.125 M Tris-HC1 pH 6.8/10% glycerol/l% SDS/1.5% /3-mercaptoethanol/0.1% bromophenol blue), boiled for 2 min, and subjected to SDS-polyacrylamide gel electrophoresis (SDSPAGE) [16].
3.13. Electrophoresis in a gradient polyacrylamide gel (SDS-PAGE)
SDS-
Electrophoresis was performed using gradient acrylamide gels 4 - 1 5 % . The gels underwent autoradiography of X-ray films (Agfa) with the aid of intensifying screens. 4. Results
In vitro immunized PBLs secreted antibodies that bound specifically to anti-B11, MMTV and H u M T V as can be seen from the immunoplaque assay (Table 1). The PBLs immunized in vitro with anti-Bll antibodies when incubated with the above mentioned antigens resulted in 3 0 0 - 500 plaques per 106 PBLs. Insignificant plaques were seen with normal rabbit serum and no plaques were detected with hepatitis B antigen. The immunoplaque-forming assay confirmed the success of the in vitro immunization with the anti-B11 antibodies and allowed hybridization to be continued with the appropriate lymphoblastoid cell line in order to produce new hybridomas. The hybridization of the in vitro immunized PBLs with UC-792-6 lymphoblastoid cell and following
TABLE 1 Antigen-specific plaque assay. The assay determines the specificity of the in vitro immunization of normal donor PBLs with rabbit-anti-Bl 1 antibodies, using the filter immuno-plaque assay. The results are given as the mean number of triplicate _+ SD of plaque forming cells (PFC)/106 viable cells. Antigen on filter
Number of PFC/10 6 cells
Total Ig/10 6 cells
Anti-B11 MMTV HuMTV Hepatitis B surface antigen Rabbit normal serum
313 _+ 11 626 +_ 15 477 _+ 9 0 7_+ 1
0.6 0.9 0.8 0 0.1
repeated limiting dilutions yielded 6 viable new hybridomas secreting MoAbs of the IgG and IgM isotypes, designated MIV: A1, A5, A7, B2, E3 and E5. Five hybridomas, MIVs A1, A5, B2, E3 and E5, secreted IgG at a concentration of 0.3-0.5 #g/ml (per 106 cells). One hybridoma, MIV A7, secreted IgM. Four of the MIV MoAbs, A1, A7, E3 and E5, bound to the rabbit anti-Bll antibodies in a dosedependent manner, as shown in Fig. 1 (E3 not shown). Competitive experiments supported these 1.5
data; MoAbs MIVs (A1, A7, E3, E5) competed with the human MoAb Bll for the binding of anti-Bll (Fig. 2). In this competitive ELISA assay, human MoAb Bll was used as the solid phase. Similar results were achieved when Bll was employed as competitor and the MIV (A1, A7, E3 and E5) antibodies as solid phase (results not shown). Since anti-Bll antibodies served as the internal 1.5
A
E
~_\
C
E
-.-E3
t.0
0 1.0
.-e-- ~AI
-.o- A 5 --~-- A 7 --o-- B 2
d
~ Bll
d O c" t~
8 g
..Q
~
Q5
O
0.5
..Q
The production of human monoclonal anti-MMTV antibodies by in vitro immunization with anti-idiotypic antibodies M. B l a n k ~, I. N. S m o r o d i n s k y 2, I. Keydar 2, S. C h a i t c h i k 1 a n d Y. S h o e n f e l d 1 IResearch Unit, Department of Medicine "B', Chaim Sheba Medical Center, TeI-Hashomer, Israel; and 2Department of Microbiology, George Wise Faculty of Life Sciences, Tel-Aviv University, TeI-Aviv, Israel (Received 12 December 1990; accepted 15 December 1990)
1. Summary We have previously reported the production of a series of human monoclonal antibodies reacting with mouse mammary tumor viral antigens as well as the human counterpart H u M T V antigen. Against two of these antibodies (Bll and 4.6/6), antiidiotypic antibodies were generated, which appeared to be the internal image of viral antigen. Vaccination with the anti-idiotypic antibodies induced in mice humoral and cellular immunity against both MMTV and HuMTV. The current study describes a novel method to produce human anti-MMTV antibodies derived monoclonal antibodies. Normal peripheral blood lymphocytes (PBLs) were immunized in vitro in the presence of IL-2, with rabbit anti-Bll antibodies bound to silica beads. Following in vitro immunization, the lymphocytes were fused with the UC-792-6 human lymphoblastoid cell line. Four human hybridomas were found to secrete human monoclonal antibodies which bound to MMTV and HuMTV. Three of the secreted antibodies were of the IgG class and one of the IgM class. The specificity of binding was confirmed by direct and competitive ELISA assays and by radioimmunoprecipitation. Our study demonstrates the ability to generate hu-
Key words: Breast cancer; Hybridoma; HuMTV; MMTV; Antiidiotype; In vitro immunisation
Correspondence to: Y. Shoenfeld, M.D., Head, Dept. of Medicine "B", Chaim Sheba Medical Center, Tel-Hashomer 52621, Israel. 0165-2478 / 91 / $ 3.50 © 1991 Elsevier Science Publishers BN.
man monoclonal anti-viral antibodies by in vitro immunization, employing "internal image" antiidiotypic antibodies. The method can be used to generate human antibodies, especially against pathogenic agents or ill-defined antigens. 2. Introduction Several studies have suggested that a relationship exists between mouse mammary tumor virus (MMTV) and human breast cancer [1, 2]. Support for this notion comes from the ability to generate human monoclonal antibodies (MoAbs) with lymphocytes derived from either patients with breast cancer [3, 4] or laboratory workers exposed to MMTV and H u M T V [5] for prolonged periods [6, 7]. The human MoAbs reacting with MMTV or with its human counterpart H u M T V antigens [5] were established from a fusion in which the following sources of lymphocytes were fused: reactive but non-infiltrated lymph node [3], peripheral blood [7] and pleural effusion [4]. Analysis of the various human anti-MMTV monoclonal antibodies led to the generation of antiidiotypic antibodies [8] that were found to be the internal image of the MMTV viral antigen. Immunization of mice with these anti-idiotypic antibodies induced a humoral and cellular state of immunity against both MMTV and H u M T V [8]. Preliminary studies of in vitro immunization of PBLs with H u M T V [9], as well as the above studies, provided the impetus to try to generate human anti-MMTV antibody following in vitro immunization of normal PBLs with internal image anti-idiotypic antibodies. 65
In vitro immunization of normal PBLs with the rabbit anti-idiotypic antibody to Bll (R-anti-Bll-Id) [3] and their fusion with a human lymphoblastoid cell line, yielded human hybridomas secreting monoclonal anti-MMTV antibodies. This method might be useful in the preparation of antibodies to any viral antigen. 3. Materials and Methods
3.1. Antibodies Antibodies used were: Bll. Details of the production of the human MoAb Bll have been reported elsewhere [3]. Bll antibody reacts with the gp52 and gp36 envelope glycoproteins of the MMTV, as well as with several core proteins [3], but did not react with other retroviruses tested. The specificity of the binding was confirmed by radioimmunoprecipitation and competition assays. BR2. A human IgM MoAb prepared from PBLs of a breast cancer patient that was found to bind MMTV and HuMTV. DIN. A human IgG MoAb derived from PBLs of a patient with drug induced lupus disease. 3.2. Anti-Bll antibody Details of the preparation of the rabbit anti-Bll anti-idiotypic antibody have been reported previously [8]. Briefly, 2-month-old rabbits were subcutaneously immunized with 50 #g of purified Bll MoAb per injection in complete Freund's adjuvant (CFA) followed by immunization with the MoAb in incomplete Freund's adjuvant at intervals of 3 weeks. Sera were obtained and precipitated with 40% saturated ammonium sulphate (SAS), dialyzed against phosphate-buffered saline (PBS) and extensively absorbed on a Sepharose column conjugated to pooled human immunoglobulin, and later on, on a fetal calf serum (FCS) Sepharose column [8]. The flow-through fluid was further affinity purified on a column consisting of polyacrylhydrazide (PAH) agarose to which Bll antibodies were conjugated. Elution of the anti-Bll antibody was performed with 0.1 N acetic acid in 0.9% NaC1 (pH 3.3). The eluates were titrated to pH 5.0 with phosphate buffer (0.1 M, pH 8.0) and dialyzed against PBS. The specificity of the binding was confirmed by ELISA assay [8]. 66
3.3. Coupling of anti-Bll antibody to silica beads Fumed SiO 2 beads (Sigma, size 0.007/~) were suspended in water (10 mg/ml) and sonicated. Then, 5 mg/ml of anti-B11 affinity purified antibody was added and incubated for 18 h at 4 °C with gentle shaking. The silica beads with the attached protein were collected by centrifugation (10000 x g for 5 rain) and washed repeatedly with distilled water. The supernatant (after the first centrifugation) was tested for the presence of anti-Bll antibodies by ELISA. 3.4. In vitro immunization Peripheral blood lymphocytes (PBLs) were isolated from healthy donor by Ficoll-hypaque gradient separation. 107 PBLs were incubated during the first 24 h in 1 ml RPMI-1640 medium containing 0.5 mg anti-Bll antibodies attached to 1 mg of SiO 2 beads/l%0 inactivated (45 min, 5 6 ° C ) a u t o l o gous serum/2 mM glutamine/2 mM non-essential amino acids/2 mM sodium pyruvate/50 units/ml insulin/5 × 105 mercaptoethanol/20 mM glutathione/20 mM each of the following: adenosine, guanosine, uridine and cytidine/3 mM thymidine/30 mM hypoxanthine and 100 units/ml penicillin/streptomycin. After 24 h, the volume was increased to a final volume of 20 ml with the above medium containing 3% of PHA-induced MLR supernatant. The cultures were incubated for 4 more days at 37 °C in an atmosphere of 5% CO2 in air. 3.5. Filter immunoplaque assay In order to confirm the efficiency of the in vitro immunization, the filter immunoplaque assay was performed prior to hybridization. The plaque assay is based on enzyme immunoassay principles as previously described [10]. Briefly, nitrocellulose membranes (type HATF, pore size 0.45 m) were coated with 10 mg/0.2 ml of R-anti-Bll-Id, HuMTV, MMTV, hepatitis B surface antigen or normal rabbit serum. The antigen-coated nitrocellulose membranes were washed with PBS and put into Petri dishes (3 cm, Nunc). Antibody-producing cells (1 x 106 cells per dish in 1 ml enriched RPMI-1640 medium: 10%0FCS/2 mM glutamine/2 mM sodium pyruvate/2 mM essential amino acids/100 u/ml
penicillin/streptomycin) were added in triplicate for 6 h in a humidified incubator with 5% CO2. The nitrocellulose papers were then washed with PBS contain.ing 0.5% Tween 20 and the spots indicating cells that secrete antibodies could be visualized using horseradish peroxidase conjugated to rabbit antihuman immunoglobulins (Sigma) and 4 chloro-1naphthol as substrate (Sigma). Blue plaques appeared within 3 min and the enzymatic reaction was terminated after 15 - 30 min by washing the nitrocellulose membrane with distilled water. The filter membranes were dried and the blue plaques were counted in the microscope by using 25 × magnification. The total immunoglobulin concentration was determined by ELISA employing a titration curve, analyzing supernatants with identical numbers of cells incubated at the conditions without the nitrocellulose. 3.6. Cell fusion A single cell suspension of lymphocytes, immunized in vitro with rabbit-anti-Bll antibodies, was fused with the UC-792-6 human lymphoblastoid cell line according to the method described previously by us [3, 7, 11-13]. After the fusion, the cells in HAT medium (hypoxanthine, aminopterin, thymidine) were seeded into either 0.2 or 2.0 ml wells of tissue culture plates (Nunc), in density of 2× 105 lymphoblastoid cells per ml. The HAT medium was replaced every 5 days. Four weeks later when hybrids were observed microscopically, feeding was continued with the same medium without aminopterin. All clones were subjected twice to limiting dilution procedure in regular medium. Subsequent cultures were grown in vertical 75-mi flasks, and in roller bottles.
3.7. Purification of hybridoma antibodies Culture fluids from each hybridoma cell line were collected, the immunoglobulins were precipitated in 50% saturated ammonium sulphate, redissolved in, and dialyzed against PBS. The antibodies were then affinity-purified on goat anti-human IgG or IgM Sepharose columns (Sigma, St. Louis, MO). The immunoglobulins bound to the column were eluted with 5 M MgC12 dialyzed against three changes of TBS (10 mM Tris/0.15 M NaC1, pH 7.3), and con-
centrated by vacuum dialysis. The immunoglobulins in the eluates were quantitated by the ELISA technique as previously reported [12]. 3.8. Antigen For HuMTV, soluble proteins were derived from the media of T47D cell line (human breast carcinoma) as previously described [5, 14]. The culture media were precipitated with 65% saturated ammonium sulphate. The precipitated proteins were dissolved in, and dialyzed against PBS, pH 7.4. MMTV was prepared as detailed elsewhere [4]. 3.9. Direct b&ding of human monoclonal antibod-
ies to R-an ti-Bll M M T V and H u M T V antigens The monoclonal antibodies produced following the in vitro immunization (MIVs) and the serum of the individuals whose lymphocytes were used for in vitro sensitization were tested for their binding capacity to R-anti-Bll Id, MMTV and HuMTV antigens by the ELISA method. Polystyrene microtiter plates with 96 flat-bottom wells (Nunc) were coated with 10 #g/ml of R-anti-Bll-Id, HuMTV or MMTV in 0.05 M bicarbonate pH 9.6 and incubated overnight at 4 °C. Following three washes with PBS and PBS containing 1%0 Tween-20 (T-PBS), the plates were blocked with 0.5% gelatine for 2 h at room temperature. Then the MIV MoAbs and the control monoclonal and polyclonal antibodies, were added to the plates at different dilutions for 2 h at room temperature. Following three additional washes with PBS and T-PBS, goat anti-human IgG or IgM conjugated to alkaline phosphatase (Sigma) diluted 1:1000 was added to the wells and incubated for 2 h at room temperature. Bound alkaline phosphatase conjugate was detected by adding substrate (pnitrophenyl phosphate-disodium). Absorbance was determined at 405 nm. As positive control we used rabbit anti-MMTV antibody and enzyme linked goat anti-rabbit IgG as a secondary antibody. 3.10. Inhibition of binding of anti-Bll antibodies to
MoAb Different concentrations (0.1-10 #g) of MoAbs MIV (A1, A5, A7, B2, E3, E5), BR2 and Bll in 25 ml 67
PBS were incubated with 25 ml of 1:500 dilutions of rabbit anti-Bll affinity purified antibodies for 5 h at room temperature. Fifty/~1 of the final competitortreated preparations was then added to flatbottomed polystyrene microtiter plates (Nunc) previously coated with 50/A of 10/zg//xl Bll and blocked with 0.5o70 gelatine. Following overnight incubation at 4°C, alkaline phosphatase goat antirabbit IgG (Sigma) diluted 1:1000, was added for 2 h at room temperature. Bound alkaline-phosphatase conjugate was detected by adding the appropriate substrate. Absorbance was read at 405 nm.
3.11. Iodination of proteins MMTV was labelled with 125I using the Iodogen method [15]. MMTV (100 #g) was iodinated with 1 mCi 125I(Amersham, U.K.). Free 125Iwas removed by centrifugation of iodinated protein through a Sephadex G-25 coarse 1-ml column. 3.12. Radioimmunoprecipitation of M M T V The human MoAbs (MIVs), established by the in vitro immunization with the rabbit-anti-Bll were characterized for their ability to precipitate iodinated MMTV. As positive controls, rabbit-anti-MMTV and rabbit anti-gp52 antibody were employed. As negative controls, we used autologous donor sera, IgG secreted by the UC-792-6 lymphoblastoid cell line and other MIV MoAbs which do not bind to MMTV by ELISA assay. 125I-labelled MMTV (2.5×106 cpm/assay) was immunoprecipitated as follows: the iodinated protein was preabsorbed with either normal human sera or rabbit serum, followed by incubation with formalin-fixed Staphylococcus aureus Cowan-1 strain and centrifugation at 10000 rev./min. The pellets were discarded and to the preabsorbed antigen, 100 t~g/assay of human MoAbs or rabbit anti-MMTV or anti-gp52 antibodies were added and incubated for one hour at room temperature (23 °C) followed by 15 min incubation with Staphylococcus aureus at 37 °C. The mixtures were centrifuged through 1 ml cushions of 1 M sucrose in TNE buffer pH 7.4 (0.01 M TrisHC1/0.15 M NaC1/1 mM EDTA) containing 0.5% Nonidet P-40 (NP-40) and 0.5% deoxycholate. The pellets were washed three times with TNE buffer pH 7.4 containing 0.5% NP-40, and finally with TNE 68
buffer. The washed pellets were counted in a gamma scintillation counter, then resuspended in 40 ml of sample buffer (0.125 M Tris-HC1 pH 6.8/10% glycerol/l% SDS/1.5% /3-mercaptoethanol/0.1% bromophenol blue), boiled for 2 min, and subjected to SDS-polyacrylamide gel electrophoresis (SDSPAGE) [16].
3.13. Electrophoresis in a gradient polyacrylamide gel (SDS-PAGE)
SDS-
Electrophoresis was performed using gradient acrylamide gels 4 - 1 5 % . The gels underwent autoradiography of X-ray films (Agfa) with the aid of intensifying screens. 4. Results
In vitro immunized PBLs secreted antibodies that bound specifically to anti-B11, MMTV and H u M T V as can be seen from the immunoplaque assay (Table 1). The PBLs immunized in vitro with anti-Bll antibodies when incubated with the above mentioned antigens resulted in 3 0 0 - 500 plaques per 106 PBLs. Insignificant plaques were seen with normal rabbit serum and no plaques were detected with hepatitis B antigen. The immunoplaque-forming assay confirmed the success of the in vitro immunization with the anti-B11 antibodies and allowed hybridization to be continued with the appropriate lymphoblastoid cell line in order to produce new hybridomas. The hybridization of the in vitro immunized PBLs with UC-792-6 lymphoblastoid cell and following
TABLE 1 Antigen-specific plaque assay. The assay determines the specificity of the in vitro immunization of normal donor PBLs with rabbit-anti-Bl 1 antibodies, using the filter immuno-plaque assay. The results are given as the mean number of triplicate _+ SD of plaque forming cells (PFC)/106 viable cells. Antigen on filter
Number of PFC/10 6 cells
Total Ig/10 6 cells
Anti-B11 MMTV HuMTV Hepatitis B surface antigen Rabbit normal serum
313 _+ 11 626 +_ 15 477 _+ 9 0 7_+ 1
0.6 0.9 0.8 0 0.1
repeated limiting dilutions yielded 6 viable new hybridomas secreting MoAbs of the IgG and IgM isotypes, designated MIV: A1, A5, A7, B2, E3 and E5. Five hybridomas, MIVs A1, A5, B2, E3 and E5, secreted IgG at a concentration of 0.3-0.5 #g/ml (per 106 cells). One hybridoma, MIV A7, secreted IgM. Four of the MIV MoAbs, A1, A7, E3 and E5, bound to the rabbit anti-Bll antibodies in a dosedependent manner, as shown in Fig. 1 (E3 not shown). Competitive experiments supported these 1.5
data; MoAbs MIVs (A1, A7, E3, E5) competed with the human MoAb Bll for the binding of anti-Bll (Fig. 2). In this competitive ELISA assay, human MoAb Bll was used as the solid phase. Similar results were achieved when Bll was employed as competitor and the MIV (A1, A7, E3 and E5) antibodies as solid phase (results not shown). Since anti-Bll antibodies served as the internal 1.5
A
E
~_\
C
E
-.-E3
t.0
0 1.0
.-e-- ~AI
-.o- A 5 --~-- A 7 --o-- B 2
d
~ Bll
d O c" t~
8 g
..Q
~
Q5
O
0.5
..Q
