- Email: [email protected]
In vitro neutralization by monoclonal antibodies of α-bungarotoxin binding to acetylcholine receptor
Toxiron Vol. 27, No. 12, pp . 1261268, 1989 . Printed in Great Britain.
0041-0101/89 53 .00+ .00 ~ 1989 Pergamon Press pk
IN VITRO NEUTRALIZATION BY MONOCLONAL ANTIBODIES
OF a-BUNGAROTOXIN BINDING TO ACETYLCHOLINE RECEPTOR* ANDREW
R.
PACHNER
arid NANCY RICALTON
Department of Neurology, Georgetown University Hospital, 3800 Reservoir Rd, Washington, DC 20007, U.S .A. (Acceptedfor publication 5 May 1989)
A. R . PACHNER and N . RICALTON . In vitro neutralization by monoclonal antibodies of a-bungarotoxin binding to acetylcholine receptor. Toxicon 27, 1263-1268, 1989 .-In order to develop monoclonal antibodies that would neutralize binding of a-bungarotoxin to acetylcholine receptor in vitro, mice were hyperimmunized with native toxin. Frequent small doses of toxin were used. Hybridoma supernatants were screened by ELISA and six monoclonal antibodies isolated and tested . The anti-a-bungarotoxin monoclonal antibodies consisted of IgM, IgGI or IgG2a antibodies. In an in vitro neutralization assay measuring the effect of the antibodies on the binding of iodinated «-bungarotoxin to BC3H1 and TE671 (mouse and human cell lines bearing acetylcholine receptor), three of the six monoclonal antibodies were able to neutralize toxin binding. These studies demonstrate the feasibility of using native toxin for the generation of hybridomas, and the potential of using in vitro neutralization assays to screen hybridomas for in vivo neutralization . INTRODUCTION
in the venoms of snakes from the families Elapidae (e .g. cobras, kraits, mambas) and Hydrophidae (sea snakes) which exert their toxicity by binding to the acetylcholine receptor (AChR) and blocking neuromuscular transmission (LEE, 1972). Both short-chain (60-62 amino acids) and long-chain (71-74 a.a.) toxins share areas that are highly conserved or invariant, which are felt to represent an area called the `toxic' loop (LENTZ et al., 1987). Antibodies to neurotoxins are important, thus, for at least two reasons: (i) to advance our understanding of toxins as ligands for cell receptors, and (ü) to work toward developing effective serotherapy of venomous snake bites (MENEZ, 1985) . Investigations into toxin-receptor interactions and serotherapy can be aided by the availability of monoclonal antitoxin antibodies and demonstration of their neutralizing ability, if present, in in vitro systems. We report the production of murine polyclonal and monoclonal antibodies to the purified neurotoxin of Bungaris multicinctus a-bungarotoxin NEUROTOxnvs are present
'Supported by a grant from the National Institute of Neurological and Communicative Disorders and Stroke (NS 23538) . 1263
1264
A. R. PACHNER and N. RICALTON
(a-BGT), produced by immunization with native toxin. These antibodies were characterized in in vitro AChR-binding assays to native AChR on the surface of human and mouse tumor lines. MATERIALS AND METHODS Mice and immunizations C57B1/6 female mice (Jackson Laboratories, Bar Harbor, MN) were immunized with purified a-BGT (Sigma, St Louis, MO). Prior to immunizations the a-BGT was tested for toxicity by i.v. injection, and in vivo potency verified by identity of the t.n,~ with the previously published ~n, m confirmed (L~, 1972). Initial immunization was with 21+g subcutaneously in complete Freund's adjuvant . Subsequent immunizations were at monthly intervals and consisted of 2 t+g subcutaneously or intraperitoneally in incomplete Freund's adjuvant. In one hybridization, from which the F hybridomas were obtained, there were five immunizations; the V hybridization was performed after seven hybridizations . The last immunization was three days prior to hybridization and consisted of 3 kg of a-BGT in phosphate-buffered saline i.v . One day prior to sacrifice of the mice, they were bled from the retro-orbital plexus, and a high titer of anti-a-BGT antibodies confirmed by ELISA. Production and cloning of anti-BGT hybrtdomas Monoclonal antibodies were produced by a modification of the method of KoHt .ex and MttsrFax (1975) . In brief, spleens from mice immunized and boosted with native a-BGT were dissected, dispersed into a single cell suspension, and red blood cells lysed with an ammonium chloride solution . After washing, the spleen cells were mixed with cells of the myeloma cell line P3X63-Ag8 (kindly supplied by Dr K. Bottomly), and polyethylene glycol 1450 (Kodak, Rochester, N.Y .) added . After a 1 min incubation, the cells were washed carefully, resuspended in HAT (hypoxanthine, aminopterin, thymidine) medium (Sigma, St Louis, MO), and aliquots placed into 24-well tissue culture plates . Supernatants from wells positive for growth were tested for anti-a-BGT antibody activity. Cells from antibody positive wells were then cloned on a bed of irradiated syngeneic splenic 'feeder' cells. Growth positive clones were then retested for antibody . Supernatants from clones from the same original well were tested for isotype, subclass and neutralizing activity; only clones different in these characteristics were saved. Ascites production CB6 (C57B1/6 x Balb/C) mice were initially immunosuppressed by sublethal irradiation . One half milliliter of pristane (2,6,10,14-tetramethylpentadecane) was injected i.p ., and 10 days later, 5 x 10° hybridoma cells were injected i.p . Over the ensuing weeks, satires wire drained by insertion of a 20 gauge needle into the abdomen. AChR-bearing cett lines The TE671 line was obtained as a gift from Dr Steven Sine, Yale University School of Medicine. Initially the cells were gown in DMEM with 10% fetal calf serum (Whittaker Bioproducts, Walkersville, MD) and were conditioned gradually over 2weeks for growth in 10% newborn calf serum (Flow, McLean, VA). Confluent cells were dispersed in 2.5% Viokase (A . H. Robins, Richmond, VA), washed three times in DMEM, and resuspended in culture medium at a 1 :5 dilution, and replated . Cells generally required feeding with fresh media twice a week . Similar methods were used for the cell line BC3H1, initially obtained as a gift from Dr Edward Hawrot, Yale University School of Medicine. In contrast to the TE671 line, however, the BC3H1 line required gowth in 20% newborn calf serum, and 4 days prior to use for receptor studies, the cells were `serum-deprived' by incubation in 0.5% serum. ELISA (enzyme-linked immunosorbent assay) for measurement of anti-a-BGT antibodies and determination of isotype The techniques for the ELISA have been previously described (P~cFUVex and Kernnx, 1983). In brief, a-BGT was coated onto polystyrene microtitration plates (Flow Laboratories, Maclean, VA) by overnight incubation at 4°C in 0.06 M carbonate buffer (pH 9.6). After washing we performed serial incubations with the test solution (serum or supernatant), goat anti-mouse immunoglobuline (or goat antimouse IgM, IgGI, IgG2a, IgG2b) conjugated with horseradish peroxidase (Cappel Laboratories, Cochranville, PE), and TMB substrate solution (Kierkegaard and Perry Laboratories, Gaithersburg, MD). Color was read on an ELISA reader (HioRad, Richmond, CA). Assay for interference with binding oja-BGT to AChR-bearing cell lines Confluent TE671 or BC3H 1 cells in 60 mm tissue culture dishes (Corning, Corning, NY) were washed with DMEM after which `depolarizing buffer' was added for at least 30 min. This buffer, used to clamp the membrane
Antibodies to a-bungarotoxin
1265
potential to 0 mV, consisted of 140mM KCI, 1 .8 mM CaC12, 1 .7 mM MgCI 25 mM HEPES [(2-hydroxymethyl)1-piperazine~thanesulfonic acid], 0.03 mg/ml bovine serum albumin, adjusted to pH 7.4 with 10-11 mM NaOH . The antibody preparations to be tested were pre-incubated with 1.25x10 -'curies of "'I-a-bungarotoxin (Amersham, Arlington Heights, IL, or New England Nuclear, Boston, MA) for 30 min and then added to the washed cells and allowed to incubate for 60 min. Cells were then washed with buffer three times and exposed to 1% Triton at 37°C overnight. The Triton extract was counted in a gamma counter for 1 min. Non-specific binding was determined in the presence of 10 mM carbamylcholine. Experiments in which non-specific binding was greater than 20% were discarded, and generally signalled overgrowth of cell cultures by variants lacking the AChR .
RESULTS
Anti-a-BGT antibody titer in mice hyperimmunized with a-BGT The anti-a-BGT binding activity of the serum from immunized mice was very high, as can be seen in Fig . 1 . The titer of antibody by ELISA was greater than 1 :30,000. Four immunizations were necessary before such titers were achieved. The isotype composition of the serum was a mixture of IgGI and IgG2a with no detectable IgG2b or IgM . Isotype of anti-a-BGT monoclonal antibodies The three neutralizing monoclonal anti-a-BGT antibodies were F2.2(IgG2a), F4.1(IgG2a) and V1 .2(IgG2a) . The three non-neutralizing anti-a-BGT monoclonal antibodies were F2.1(IgG 1), F6.1(IgM) and V 1 .1(IgG2a) . Although four of the six mcAbs were IgG2a, the isotype determinations of the anti-a-BGT polyclonal antiserum revealed both IgGI and IgG2a with somewhat less of the latter than the former (data not shown). Dilutions of supernatant of the anti-a-BGT hybridomas beyond 1 :100 to 1 :300 resulted in loss of detection of activity, while ascites preparations retained detectable anti-a-BGT antibody out to 1:10,000, indicating the much higher concentration of antibody in ascites relative to serum. There was considerable variability in the mcAbs with regard to ability to produce ascites with anti-a-BGT activity : the best ascites producer was F4.1, while ascites produced by injection of hybridomas F2.1 and F6.1 lacked anti-a-BGT activity . mo eoo soo d ~Oo d aoo soo 100 0 Dllutlon FIG. 1 . ELISA of ANTI a-BGT HYPERIMMUNE SERUM AND CONTROL HYPERIMMUNE ANTI-B$A sERUM FROM A MOUSE. a-BGT binding by the anti-a-BGT hyperimmune serum is detectable out to 1:30,000. Standard deviation of each point is less than or equal to 15% (n = 4) .
1266
A. R. PACHNER and N. RICALTON
~O
u
a°" â t s 0
m bulfsr
F2 .1
F2 .2
F4.1
FH.1
Antlbodlss
V1 .1
V1 .2
antl " BOT wrum
FIG. 2 . EFFECT OF VARIOUS MONOCLONAL AND POLYCLONAL ANTI-a-BGT AhTiBODIES ON THE BINDING of '=SI-BGT AChR oN BC3H1 CELLS.
ro
Binding of 100% represented iodinated a-BGT incubated with buffer alone without anti-a-BGT antibody; the CPM of this point was 6783 . Neither non-immune serum nor irrelevant monoclonal antibodies had any effect in this assay (data not shown) . Non-specific binding, as measured by 'z'I-a-BGT binding in the presence of d-tubocurarine was 8l7cpm, or 12% . Error bars represent standard deviations (n = 7) .
Neutralization of a-BGT in vitro The monoclonal antibodies were tested for their ability to prevent binding to the AChR present on the nicotinic AChR-bearing cell lines BC3H1 (mouse) and TE671 (human). Monoclonals F4.1, F2.2, and V1 .2 were potent in their ability to neutralize a-BGT in vitro, while the other monoclonals had no activity (Fig. 2). Antisera from a-BGT-hyperimmunized mice were also able to neutralize a-BGT in vitro. Results shown are those for the BC3H1 line; comparable results were obtained with the TE671 line . The difference in neutralizing capacity of the monoclonal antibodies was not due to differences in concentrations of the antibodies used in the neutralization studies since the supernatants used had similar anti-BGT titers by end-point dilution. The neutralizing capacity of the antibodies was concentration-dependent, as shown by the representative example of the ascites from F4.1 (Fig. 3). The other monoclonals had similar curves although plateau
FIG. 3. NEUTRALIZATION OF'ZSI-a-BGT HINDING TO AC11R ON BC3H1 CELLS AS A FUNCTION OF CONCENTRATION OF F4 .1 ASCITES. Various concentrations of F4 .1 ascites dilution in buffer were preincubated with iodinated a-BGT. Binding of 100% represented iodinated a-BGT incubated with buffer alone without anti-a-BGT antibody; the cpm of this point was 10,343 . Non-specific binding as measured by "°I-a-BGT binding in the presence of d-tubocurarine, was 9%, or 926 cpm . Error bars represent standard deviations (n=2).
Antibodies to a-bungarotoxin
126 7
concentrations were not reached because of the fact that the supernatants had a significantly lower antibody concentration than the ascites of F4.1 . Pre-incubation of the monoclonal antibodies with the iodinated a-BGT was not necessary for neutralization . The neutralization experiments were also performed so that the monoclonal antibodies were present in the petri dish when the iodinated bungarotoxin was added; for monoclonals F2.2, F4.1 and V1.2, binding was decreased to 10-20% of control by both methods. DISCUSSION
Our understanding of the molecular relationships between toxic components of venoms and their target tissues has increased dramatically in the last decade . The primary toxin of many venoms of the Elapidae and Hydrophidae are post-synaptic neuromuscular blockers that bind to the acetylcholine receptor (AChR) and block binding of acetylcholine (PopoT and CxANCSUx, 1984). The most thoroughly studied interaction between a toxin and the receptor to which it binds is that between a-BGT and the AChR. Alpha-BGT binds to the alpha subunit of the AChR and the area around residues 192 and 193 on this AChR subunit, which are dual cysteines, is the most likely candidate for the a-BGT binding site (PnTIUCx et al., 1987 ; WILSON et al., 1985). Of similar interest are studies on areas on the a-BGT molecule that seem to be involved in the binding to the AChR. Chemical modifications of a large number of different amino acids including arginines, lysines, tyrosines, tryptophan (LsN~rz and WILSON, 1988) have not resulted in losses of toxicity or affinity of more than a log-fold . The change in toxin structure leading to the most dramatic reduction in affinity has been disruption of the disulfide bonds (MAR~rIN et al., 1983), demonstrating the need for the rigidity in the molecule brought about by the five disulfide bonds. Another probable explanation for the lack of great effect of chemical modification studies is that multiple points of contact between a-BGT and AChR are probably necessary. Loop 2 of the toxin molecule, containing invariant residues Arg37 and Trp29, is the most likely candidate as the part of the a-BGT molecule binding to AChR, based on earlier work as well as more recent studies of interference of 'zsI-a-BGT binding to AChR by a-BGT and cobrotoxin synthetic peptides (Ll:rrrZ et al., 1987). It would be reasonable to hypothesize that neutralizing monoclonal antibodies bind at or near the toxic loop, while non-neutralizing anti-a-BGT monoclonal antibodies bind to parts of the molecule distant from this site . This hypothesis is the subject for further investigation. Monoclonal antibodies to a-BGT have previously been described (DANSi: et al., 1986), but these were obtained by immunization with denatured a-BGT, a process which may affect the generation of antibodies to the toxic portion of a-BGT. These antibodies and monoclonal antibodies to other post-synaptic toxins (BOULAIN et al., 1982; BOULAIN et al., 1985) have been shown to have neutralizing capacity when pre-incubated with the toxin and injected into experimental animals. However, it requires further study to determine whether any monoclonal antibodies can serve to protect experimental animals when injected into the whole animal prior to injection of the toxin. The in vitro system described above for testing neutralization of anti-toxin antibodies, utilizing native AChR on the surface of transformed cell lines, has significant advantages over testing in vivo . It does not require the use of animals, and is cheap and easily reproducible . Although it is unknown how well in vitro and in vivo testing correlate, it is
1268
A. R. PACHNER and N . RICALTON
difficult to conceive of a monoclonal antibody being unable to neutralize in vitro, but able to neutralize in vivo ; thus, in vitro studies may prove to be a useful initial screening assay for more extensive in vivo testing of anti-toxin monoclonal antibodies. This report is the first on monoclonal antibodies to a-BGT induced by immunization with native a-BGT. Three out of six monoclonal antibodies derived from hybridizations from these immunizations had neutralizing ability in vitro. This high proportion is possibly related to the use of native a-BGT in the immunization process. Studies are in progress to evaluate the ability of these antibodies to protect from a-BGT injection in vivo, and to correlate in vivo with in vitro neutralizing ability. REFERENCES Bour.wnv, J . C., Menez, A., CotroEtec, J ., FwtrnE, G ., Ltwcoeouros, P. and FROStwcmr, P. (1982) Neutralizing monoclonal antibody specific for Naja nigricollis toxin-alpha : preparation, characterization and localization of the antigenic binding site. Biochemistry 21, 2910-2915 . BOULAIN, J . C., FROMAGEOT, P. and M~z, A. (1985) Further evidence showing that neurotoxin-acetylcholine receptor dissociation is accelerated by monoclonal newotoxin-specific immunoglobulin . Mol . Immw. 22, 553556. Dwnse, J .-M., Toussw~xr, J .-L. and KEMPF, J . (1986) Neutralization of alpha-bungarotoxin by monoclonal antibodies. Toxicon 24, 141-151 . Kot~e, G. and Must~N, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 49598 . Lee, C . Y. (1972), Chemistry and pharmacology of polypeptide toxins in snake venoms . Ann. Rev. Pharmac. 12, 265-286 . Lerrrz, T . L . and W~tsort, P. T. (1988) Newotoxin-binding site on the acetylcholine receptor . Int . Rev. Neurobiol. 29, 117-160. Lexrz, T. L ., Hwwxor, E. and Wusox, P . T . (1987) Synthetic peptides corresponding to sequences of snake venom neurotoxins and rabies virus glycoprotein bind to the nicotinic acetylcholine receptor. Prot . Struc. Func. Genet . 2, 298-317 . Mwttrtx, B. M ., GnaHea, B . A. and MwaucKE, A. (1983) The sites of neurotoxicity in alpha~obratoxin . J. biol. Chem. 258, 8714-8722 . Mt:rrsz, A . (1985) Molecular immunology of snake toxins . Pharmac . Ther. 30, 91J113 . PwctnaeR, A . R. and Kwrrrox, F . S . (1983) The relation of clinical disease to antibody titre, proliferative response and newophysiology in marine experimental autoimmune myasthenia gravis . Clip. exp . Immun . 51, 543-550. Pwrxtcx, J ., Bout,rEx,1 ., GOLDMAN, D ., Gwxnxex, P. and HE~x~tnNrr, S . (1987) Molecular biology of nicotinic acetylcholine receptors. Ann. N. Y. Acad. Sci. 505, 194-207 . Poeor, J .-L . and Cttwxaetnt, J .-P. (1984) Nicotinic receptor of acetylcholine: structure of an oligomeric integral membrane protein . Physiol . Rev. 64, 1162-1239 . TREt~wu, O ., BOULAIN, J .-C ., Couneac, J ., FROMAGFAT, P. and MexEZ, A . (1986) A monoclonal antibody which recognized the functional site of snake neurotoxins and which neutralizes all short-chain variants . FEBS Lett. 208, 236-240 . W~tsotv, P. T ., LErrrz, T. L . and Hwwaor, E. (1985) Determination of the primary amino acid sequence specifying the alpha-bungarotoxin binding site on the alpha subunit of the acetylcholine receptor for Torpedo califoneica. Proc . natn . Acad. Sci . U .S .A . 82, 8790-8794 .
0041-0101/89 53 .00+ .00 ~ 1989 Pergamon Press pk
IN VITRO NEUTRALIZATION BY MONOCLONAL ANTIBODIES
OF a-BUNGAROTOXIN BINDING TO ACETYLCHOLINE RECEPTOR* ANDREW
R.
PACHNER
arid NANCY RICALTON
Department of Neurology, Georgetown University Hospital, 3800 Reservoir Rd, Washington, DC 20007, U.S .A. (Acceptedfor publication 5 May 1989)
A. R . PACHNER and N . RICALTON . In vitro neutralization by monoclonal antibodies of a-bungarotoxin binding to acetylcholine receptor. Toxicon 27, 1263-1268, 1989 .-In order to develop monoclonal antibodies that would neutralize binding of a-bungarotoxin to acetylcholine receptor in vitro, mice were hyperimmunized with native toxin. Frequent small doses of toxin were used. Hybridoma supernatants were screened by ELISA and six monoclonal antibodies isolated and tested . The anti-a-bungarotoxin monoclonal antibodies consisted of IgM, IgGI or IgG2a antibodies. In an in vitro neutralization assay measuring the effect of the antibodies on the binding of iodinated «-bungarotoxin to BC3H1 and TE671 (mouse and human cell lines bearing acetylcholine receptor), three of the six monoclonal antibodies were able to neutralize toxin binding. These studies demonstrate the feasibility of using native toxin for the generation of hybridomas, and the potential of using in vitro neutralization assays to screen hybridomas for in vivo neutralization . INTRODUCTION
in the venoms of snakes from the families Elapidae (e .g. cobras, kraits, mambas) and Hydrophidae (sea snakes) which exert their toxicity by binding to the acetylcholine receptor (AChR) and blocking neuromuscular transmission (LEE, 1972). Both short-chain (60-62 amino acids) and long-chain (71-74 a.a.) toxins share areas that are highly conserved or invariant, which are felt to represent an area called the `toxic' loop (LENTZ et al., 1987). Antibodies to neurotoxins are important, thus, for at least two reasons: (i) to advance our understanding of toxins as ligands for cell receptors, and (ü) to work toward developing effective serotherapy of venomous snake bites (MENEZ, 1985) . Investigations into toxin-receptor interactions and serotherapy can be aided by the availability of monoclonal antitoxin antibodies and demonstration of their neutralizing ability, if present, in in vitro systems. We report the production of murine polyclonal and monoclonal antibodies to the purified neurotoxin of Bungaris multicinctus a-bungarotoxin NEUROTOxnvs are present
'Supported by a grant from the National Institute of Neurological and Communicative Disorders and Stroke (NS 23538) . 1263
1264
A. R. PACHNER and N. RICALTON
(a-BGT), produced by immunization with native toxin. These antibodies were characterized in in vitro AChR-binding assays to native AChR on the surface of human and mouse tumor lines. MATERIALS AND METHODS Mice and immunizations C57B1/6 female mice (Jackson Laboratories, Bar Harbor, MN) were immunized with purified a-BGT (Sigma, St Louis, MO). Prior to immunizations the a-BGT was tested for toxicity by i.v. injection, and in vivo potency verified by identity of the t.n,~ with the previously published ~n, m confirmed (L~, 1972). Initial immunization was with 21+g subcutaneously in complete Freund's adjuvant . Subsequent immunizations were at monthly intervals and consisted of 2 t+g subcutaneously or intraperitoneally in incomplete Freund's adjuvant. In one hybridization, from which the F hybridomas were obtained, there were five immunizations; the V hybridization was performed after seven hybridizations . The last immunization was three days prior to hybridization and consisted of 3 kg of a-BGT in phosphate-buffered saline i.v . One day prior to sacrifice of the mice, they were bled from the retro-orbital plexus, and a high titer of anti-a-BGT antibodies confirmed by ELISA. Production and cloning of anti-BGT hybrtdomas Monoclonal antibodies were produced by a modification of the method of KoHt .ex and MttsrFax (1975) . In brief, spleens from mice immunized and boosted with native a-BGT were dissected, dispersed into a single cell suspension, and red blood cells lysed with an ammonium chloride solution . After washing, the spleen cells were mixed with cells of the myeloma cell line P3X63-Ag8 (kindly supplied by Dr K. Bottomly), and polyethylene glycol 1450 (Kodak, Rochester, N.Y .) added . After a 1 min incubation, the cells were washed carefully, resuspended in HAT (hypoxanthine, aminopterin, thymidine) medium (Sigma, St Louis, MO), and aliquots placed into 24-well tissue culture plates . Supernatants from wells positive for growth were tested for anti-a-BGT antibody activity. Cells from antibody positive wells were then cloned on a bed of irradiated syngeneic splenic 'feeder' cells. Growth positive clones were then retested for antibody . Supernatants from clones from the same original well were tested for isotype, subclass and neutralizing activity; only clones different in these characteristics were saved. Ascites production CB6 (C57B1/6 x Balb/C) mice were initially immunosuppressed by sublethal irradiation . One half milliliter of pristane (2,6,10,14-tetramethylpentadecane) was injected i.p ., and 10 days later, 5 x 10° hybridoma cells were injected i.p . Over the ensuing weeks, satires wire drained by insertion of a 20 gauge needle into the abdomen. AChR-bearing cett lines The TE671 line was obtained as a gift from Dr Steven Sine, Yale University School of Medicine. Initially the cells were gown in DMEM with 10% fetal calf serum (Whittaker Bioproducts, Walkersville, MD) and were conditioned gradually over 2weeks for growth in 10% newborn calf serum (Flow, McLean, VA). Confluent cells were dispersed in 2.5% Viokase (A . H. Robins, Richmond, VA), washed three times in DMEM, and resuspended in culture medium at a 1 :5 dilution, and replated . Cells generally required feeding with fresh media twice a week . Similar methods were used for the cell line BC3H1, initially obtained as a gift from Dr Edward Hawrot, Yale University School of Medicine. In contrast to the TE671 line, however, the BC3H1 line required gowth in 20% newborn calf serum, and 4 days prior to use for receptor studies, the cells were `serum-deprived' by incubation in 0.5% serum. ELISA (enzyme-linked immunosorbent assay) for measurement of anti-a-BGT antibodies and determination of isotype The techniques for the ELISA have been previously described (P~cFUVex and Kernnx, 1983). In brief, a-BGT was coated onto polystyrene microtitration plates (Flow Laboratories, Maclean, VA) by overnight incubation at 4°C in 0.06 M carbonate buffer (pH 9.6). After washing we performed serial incubations with the test solution (serum or supernatant), goat anti-mouse immunoglobuline (or goat antimouse IgM, IgGI, IgG2a, IgG2b) conjugated with horseradish peroxidase (Cappel Laboratories, Cochranville, PE), and TMB substrate solution (Kierkegaard and Perry Laboratories, Gaithersburg, MD). Color was read on an ELISA reader (HioRad, Richmond, CA). Assay for interference with binding oja-BGT to AChR-bearing cell lines Confluent TE671 or BC3H 1 cells in 60 mm tissue culture dishes (Corning, Corning, NY) were washed with DMEM after which `depolarizing buffer' was added for at least 30 min. This buffer, used to clamp the membrane
Antibodies to a-bungarotoxin
1265
potential to 0 mV, consisted of 140mM KCI, 1 .8 mM CaC12, 1 .7 mM MgCI 25 mM HEPES [(2-hydroxymethyl)1-piperazine~thanesulfonic acid], 0.03 mg/ml bovine serum albumin, adjusted to pH 7.4 with 10-11 mM NaOH . The antibody preparations to be tested were pre-incubated with 1.25x10 -'curies of "'I-a-bungarotoxin (Amersham, Arlington Heights, IL, or New England Nuclear, Boston, MA) for 30 min and then added to the washed cells and allowed to incubate for 60 min. Cells were then washed with buffer three times and exposed to 1% Triton at 37°C overnight. The Triton extract was counted in a gamma counter for 1 min. Non-specific binding was determined in the presence of 10 mM carbamylcholine. Experiments in which non-specific binding was greater than 20% were discarded, and generally signalled overgrowth of cell cultures by variants lacking the AChR .
RESULTS
Anti-a-BGT antibody titer in mice hyperimmunized with a-BGT The anti-a-BGT binding activity of the serum from immunized mice was very high, as can be seen in Fig . 1 . The titer of antibody by ELISA was greater than 1 :30,000. Four immunizations were necessary before such titers were achieved. The isotype composition of the serum was a mixture of IgGI and IgG2a with no detectable IgG2b or IgM . Isotype of anti-a-BGT monoclonal antibodies The three neutralizing monoclonal anti-a-BGT antibodies were F2.2(IgG2a), F4.1(IgG2a) and V1 .2(IgG2a) . The three non-neutralizing anti-a-BGT monoclonal antibodies were F2.1(IgG 1), F6.1(IgM) and V 1 .1(IgG2a) . Although four of the six mcAbs were IgG2a, the isotype determinations of the anti-a-BGT polyclonal antiserum revealed both IgGI and IgG2a with somewhat less of the latter than the former (data not shown). Dilutions of supernatant of the anti-a-BGT hybridomas beyond 1 :100 to 1 :300 resulted in loss of detection of activity, while ascites preparations retained detectable anti-a-BGT antibody out to 1:10,000, indicating the much higher concentration of antibody in ascites relative to serum. There was considerable variability in the mcAbs with regard to ability to produce ascites with anti-a-BGT activity : the best ascites producer was F4.1, while ascites produced by injection of hybridomas F2.1 and F6.1 lacked anti-a-BGT activity . mo eoo soo d ~Oo d aoo soo 100 0 Dllutlon FIG. 1 . ELISA of ANTI a-BGT HYPERIMMUNE SERUM AND CONTROL HYPERIMMUNE ANTI-B$A sERUM FROM A MOUSE. a-BGT binding by the anti-a-BGT hyperimmune serum is detectable out to 1:30,000. Standard deviation of each point is less than or equal to 15% (n = 4) .
1266
A. R. PACHNER and N. RICALTON
~O
u
a°" â t s 0
m bulfsr
F2 .1
F2 .2
F4.1
FH.1
Antlbodlss
V1 .1
V1 .2
antl " BOT wrum
FIG. 2 . EFFECT OF VARIOUS MONOCLONAL AND POLYCLONAL ANTI-a-BGT AhTiBODIES ON THE BINDING of '=SI-BGT AChR oN BC3H1 CELLS.
ro
Binding of 100% represented iodinated a-BGT incubated with buffer alone without anti-a-BGT antibody; the CPM of this point was 6783 . Neither non-immune serum nor irrelevant monoclonal antibodies had any effect in this assay (data not shown) . Non-specific binding, as measured by 'z'I-a-BGT binding in the presence of d-tubocurarine was 8l7cpm, or 12% . Error bars represent standard deviations (n = 7) .
Neutralization of a-BGT in vitro The monoclonal antibodies were tested for their ability to prevent binding to the AChR present on the nicotinic AChR-bearing cell lines BC3H1 (mouse) and TE671 (human). Monoclonals F4.1, F2.2, and V1 .2 were potent in their ability to neutralize a-BGT in vitro, while the other monoclonals had no activity (Fig. 2). Antisera from a-BGT-hyperimmunized mice were also able to neutralize a-BGT in vitro. Results shown are those for the BC3H1 line; comparable results were obtained with the TE671 line . The difference in neutralizing capacity of the monoclonal antibodies was not due to differences in concentrations of the antibodies used in the neutralization studies since the supernatants used had similar anti-BGT titers by end-point dilution. The neutralizing capacity of the antibodies was concentration-dependent, as shown by the representative example of the ascites from F4.1 (Fig. 3). The other monoclonals had similar curves although plateau
FIG. 3. NEUTRALIZATION OF'ZSI-a-BGT HINDING TO AC11R ON BC3H1 CELLS AS A FUNCTION OF CONCENTRATION OF F4 .1 ASCITES. Various concentrations of F4 .1 ascites dilution in buffer were preincubated with iodinated a-BGT. Binding of 100% represented iodinated a-BGT incubated with buffer alone without anti-a-BGT antibody; the cpm of this point was 10,343 . Non-specific binding as measured by "°I-a-BGT binding in the presence of d-tubocurarine, was 9%, or 926 cpm . Error bars represent standard deviations (n=2).
Antibodies to a-bungarotoxin
126 7
concentrations were not reached because of the fact that the supernatants had a significantly lower antibody concentration than the ascites of F4.1 . Pre-incubation of the monoclonal antibodies with the iodinated a-BGT was not necessary for neutralization . The neutralization experiments were also performed so that the monoclonal antibodies were present in the petri dish when the iodinated bungarotoxin was added; for monoclonals F2.2, F4.1 and V1.2, binding was decreased to 10-20% of control by both methods. DISCUSSION
Our understanding of the molecular relationships between toxic components of venoms and their target tissues has increased dramatically in the last decade . The primary toxin of many venoms of the Elapidae and Hydrophidae are post-synaptic neuromuscular blockers that bind to the acetylcholine receptor (AChR) and block binding of acetylcholine (PopoT and CxANCSUx, 1984). The most thoroughly studied interaction between a toxin and the receptor to which it binds is that between a-BGT and the AChR. Alpha-BGT binds to the alpha subunit of the AChR and the area around residues 192 and 193 on this AChR subunit, which are dual cysteines, is the most likely candidate for the a-BGT binding site (PnTIUCx et al., 1987 ; WILSON et al., 1985). Of similar interest are studies on areas on the a-BGT molecule that seem to be involved in the binding to the AChR. Chemical modifications of a large number of different amino acids including arginines, lysines, tyrosines, tryptophan (LsN~rz and WILSON, 1988) have not resulted in losses of toxicity or affinity of more than a log-fold . The change in toxin structure leading to the most dramatic reduction in affinity has been disruption of the disulfide bonds (MAR~rIN et al., 1983), demonstrating the need for the rigidity in the molecule brought about by the five disulfide bonds. Another probable explanation for the lack of great effect of chemical modification studies is that multiple points of contact between a-BGT and AChR are probably necessary. Loop 2 of the toxin molecule, containing invariant residues Arg37 and Trp29, is the most likely candidate as the part of the a-BGT molecule binding to AChR, based on earlier work as well as more recent studies of interference of 'zsI-a-BGT binding to AChR by a-BGT and cobrotoxin synthetic peptides (Ll:rrrZ et al., 1987). It would be reasonable to hypothesize that neutralizing monoclonal antibodies bind at or near the toxic loop, while non-neutralizing anti-a-BGT monoclonal antibodies bind to parts of the molecule distant from this site . This hypothesis is the subject for further investigation. Monoclonal antibodies to a-BGT have previously been described (DANSi: et al., 1986), but these were obtained by immunization with denatured a-BGT, a process which may affect the generation of antibodies to the toxic portion of a-BGT. These antibodies and monoclonal antibodies to other post-synaptic toxins (BOULAIN et al., 1982; BOULAIN et al., 1985) have been shown to have neutralizing capacity when pre-incubated with the toxin and injected into experimental animals. However, it requires further study to determine whether any monoclonal antibodies can serve to protect experimental animals when injected into the whole animal prior to injection of the toxin. The in vitro system described above for testing neutralization of anti-toxin antibodies, utilizing native AChR on the surface of transformed cell lines, has significant advantages over testing in vivo . It does not require the use of animals, and is cheap and easily reproducible . Although it is unknown how well in vitro and in vivo testing correlate, it is
1268
A. R. PACHNER and N . RICALTON
difficult to conceive of a monoclonal antibody being unable to neutralize in vitro, but able to neutralize in vivo ; thus, in vitro studies may prove to be a useful initial screening assay for more extensive in vivo testing of anti-toxin monoclonal antibodies. This report is the first on monoclonal antibodies to a-BGT induced by immunization with native a-BGT. Three out of six monoclonal antibodies derived from hybridizations from these immunizations had neutralizing ability in vitro. This high proportion is possibly related to the use of native a-BGT in the immunization process. Studies are in progress to evaluate the ability of these antibodies to protect from a-BGT injection in vivo, and to correlate in vivo with in vitro neutralizing ability. REFERENCES Bour.wnv, J . C., Menez, A., CotroEtec, J ., FwtrnE, G ., Ltwcoeouros, P. and FROStwcmr, P. (1982) Neutralizing monoclonal antibody specific for Naja nigricollis toxin-alpha : preparation, characterization and localization of the antigenic binding site. Biochemistry 21, 2910-2915 . BOULAIN, J . C., FROMAGEOT, P. and M~z, A. (1985) Further evidence showing that neurotoxin-acetylcholine receptor dissociation is accelerated by monoclonal newotoxin-specific immunoglobulin . Mol . Immw. 22, 553556. Dwnse, J .-M., Toussw~xr, J .-L. and KEMPF, J . (1986) Neutralization of alpha-bungarotoxin by monoclonal antibodies. Toxicon 24, 141-151 . Kot~e, G. and Must~N, C. (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 49598 . Lee, C . Y. (1972), Chemistry and pharmacology of polypeptide toxins in snake venoms . Ann. Rev. Pharmac. 12, 265-286 . Lerrrz, T . L . and W~tsort, P. T. (1988) Newotoxin-binding site on the acetylcholine receptor . Int . Rev. Neurobiol. 29, 117-160. Lexrz, T. L ., Hwwxor, E. and Wusox, P . T . (1987) Synthetic peptides corresponding to sequences of snake venom neurotoxins and rabies virus glycoprotein bind to the nicotinic acetylcholine receptor. Prot . Struc. Func. Genet . 2, 298-317 . Mwttrtx, B. M ., GnaHea, B . A. and MwaucKE, A. (1983) The sites of neurotoxicity in alpha~obratoxin . J. biol. Chem. 258, 8714-8722 . Mt:rrsz, A . (1985) Molecular immunology of snake toxins . Pharmac . Ther. 30, 91J113 . PwctnaeR, A . R. and Kwrrrox, F . S . (1983) The relation of clinical disease to antibody titre, proliferative response and newophysiology in marine experimental autoimmune myasthenia gravis . Clip. exp . Immun . 51, 543-550. Pwrxtcx, J ., Bout,rEx,1 ., GOLDMAN, D ., Gwxnxex, P. and HE~x~tnNrr, S . (1987) Molecular biology of nicotinic acetylcholine receptors. Ann. N. Y. Acad. Sci. 505, 194-207 . Poeor, J .-L . and Cttwxaetnt, J .-P. (1984) Nicotinic receptor of acetylcholine: structure of an oligomeric integral membrane protein . Physiol . Rev. 64, 1162-1239 . TREt~wu, O ., BOULAIN, J .-C ., Couneac, J ., FROMAGFAT, P. and MexEZ, A . (1986) A monoclonal antibody which recognized the functional site of snake neurotoxins and which neutralizes all short-chain variants . FEBS Lett. 208, 236-240 . W~tsotv, P. T ., LErrrz, T. L . and Hwwaor, E. (1985) Determination of the primary amino acid sequence specifying the alpha-bungarotoxin binding site on the alpha subunit of the acetylcholine receptor for Torpedo califoneica. Proc . natn . Acad. Sci . U .S .A . 82, 8790-8794 .