- Email: [email protected]
Monoclonal rat anti-Torpedo electroplax nicotinic acetylcholine receptor antibodies: Immunochemical characterization
Journal of Immunological Methods, 74 (1984) 129-138 Elsevier
129
JIM 03252
Monoclonal Rat Anti-Torpedo Electroplax Nicotinic Acetylcholine Receptor Antibodies" Immunochemical Characterization * Ronald J. Lukas Division of Neurobiology, Barrow Neurological Institute, 350 West Thomas Road, Phoenix, A Z 85013, U.S.A.
(Received 14 June 1984, accepted 18 July 1984)
The interaction of Torpedo californica nicotinic acetylcholine receptor (nAcChoR) with three rat monoclonal antibodies (mcab) directed against nAcChoR (Gomez et al., 1979) was studied by use of four different radioimmunoassay protocols. Each mcab reacts poorly with formalin-fixed Staphylococcus aureus or S. aureus Protein A, which requires that an additional incubation with second antibody (goat anti-rat immunoglobulin G) is included in each radioimmunoassay paradigm. One mcab inhibits 125I-labeled a-bungarotoxin binding to nAcChoR, recognizes a subset of solubilized nAcChoR-toxin complexes, and shows higher titer against nAcChoR in the absence of toxin. Thus, it appears to be directed against nAcChoR antigenic determinants that at least partially overlap with toxin binding sites. Two other mcab react with nAcChoR in a toxin-independent manner. Receptor-mcab dissociation constants are less than 10 nM, according to each assay paradigm. Estimates of antibody titer or concentrations of antibody in stock hybridoma supernatants vary according to the assay used. This predictable result is attributed to differences in design and sensitivity of assay protocols. The data provide a basis for further utilization of monospecific antibodies as probes in characterization of nAcChoR structure and function.
Key words: acetylcholine receptor - nicotinic receptor - monoclonal antibodies
* Supported by National Institutes of Health Grant NS-16821, by a grant from the Epilepsy Foundation of America, by funds from Epi-Hab of Arizona, Inc., and by the Men's and Women's Boards of the Barrow Neurological Foundation. Abbreviations: CNS, central nervous system; AcChoR, acetylcholine receptor; nAcChoR, nicotinic acetylcholine receptor; IgG, immunoglobulin G; mcab, monoclonal antibody; Bgt, a-bungarotoxin; I-Bgt, 125I-labeled monoiodinated a-bungarotoxin; memb nAcChoR, nicotinic acetylcholine receptor-rich membranes from Torpedo californica electric organ; sol nAcChoR, detergent-solubilized nAcChoR from Torpedo; staph a, formalin-fixed Staphylococcus aureus cells; GAR, goat anti-rat IgG antiserum; I-Protein A, 125I-labeled iodinated Protein A; TNP7.4-BSA, 100 mM NaCl, 50 mM NaEHPO4, pH 7.4 supplemented with 1% Triton X-100 and 0.1% bovine serum albumin; NP7.4-BSA, 100 mM NaC1, 50 mM Na2HPO4, pH 7.4, supplemented with 0.1% bovine serum albumin. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
130 Introduction
Our ultimate understanding of molecular mechanisms of central nervous system (CNS) acetylcholine receptor (AcChoR) function is dependent upon identification of suitable receptor-specific probes. Small cholinergic ligands and atropinemimetic and curaremimetic neurotoxins have been used to identify putative CNS AcChoR sites, but there remains uncertainty regarding their functional relevance and potency (Schmidt et al., 1980; Morley and Kemp, 1981). In a previous manuscript (Lukas, 1984b), interaction between nicotinic AcChoR (nAcChoR) from Torpedo californica electric organ and antibodies in polyclonal antisera raised against nAcChoR from Electrophorus" electricus electric organ was described. While antisera-antigen interactions are well-behaved, and provide information on antibody and nAcChoR characteristics, the heterogeneity of immunoglobulin G (IgG) species in the antisera, and their limited availability, restricts rigorous application of these antibodies as receptor-specific ligands. Mono-specific anti-nAcChoR antibodies, obtained in large quantities by using the hybridoma technique (K6hler and Milstein, 1975), may prove more useful as molecular probes in characterization of receptor structure and function (Mochly-Rosen et al., 1979: Lennon et al., 1980; Tzartos and Lindstrom, 1980; Yelton and Scharff, 1981). Hybridoma supernatants containing monoclonal antibodies (mcab) directed against Torpedo nAcChoR have been obtained (Gomez et al., 1979; Richman et al., 1980). With a view to using these mcab as probes for authentic nAcChoR in the mammalian CNS, characterization of mcab-Torpedo nAcChoR interactions has been undertaken.
Materials and Methods
Preparation of a-bungarotoxin (Bgt), 125I-labeled Bgt (I-Bgt), and nAcChoR-rich membranes from Torpedo (memb nAcChoR) is as described previously (Lukas, 1984a). Procedures for solubilization of Torpedo nAcChoR (sol nAcChoR), preparation of toxin- and radiotoxin-labeled memb and sol nAcChoR, iodination of Protein A, and treatment of formalin-fixed Staphylococcus aureus cells (staph a), also appear elsewhere (Lukas, 1984b). Media from three hybridoma clonal lines, derived from fusion of spleen cells of Torpedo nAcChoR-innoculated female Lewis rats with cells of the murine myeloma line P3-X63-Ag8, were the generous gift of Dr. David Richman of the University of Chicago. Clones 56I, 60D and 73G each produce anti-Torpedo nAcChoR mcab, which are designated in this manuscript as 56, 60, and 73, respectively. Dr. Richman also supplied supernatants from the parent myeloma line, which are designated as P3. Goat anti-rat IgG antiserum (GAR; Cappel 0113-0911) and culture supernatants are fractionated into 1-2 ml aliquots, and stored at - 2 0 ° C or - 8 0 ° C until use. Experimental protocols for four different radioimmunoassay schemes, described in detail previously (Lukas, 1984b), are modified for assay of mcab activity by insertion of a second antibody incubation step between mcab-antigen reactions and
131
staph a or 125I-labeled iodinated Protein A (I-Protein A) addition. Toxin binding inhibition assays are also as previously described (Lukas, 1984b).
Results
Immunoprecipitation assay Mcab-antigen complexes were not precipitated by staph a in the absence of GAR. Experiments were conducted routinely to establish the staph a and GAR requirements for quantitative immunoprecipitation of mcab-nAcChoR complexes under the assay conditions used below. The antigen dependence for immunoprecipitation of sol nAcChoR-I-Bgt with mcab is shown in Fig. 1. In each case, saturation of antibody capacity is achieved at high concentrations of antigen. About 6.8, 3.3, and 6.5 fmol sol nAcChoR-I-Bgt are precipitated by the given dilution of 56, 60, and 73, respectively, which provides estimates of the concentration of mcab in stock hybridoma supernatants given in Table I. Also provided in Table I are values for the apparent dissociation constant for mcab-sol nAcChoR interaction. Antibody titration curves obtained in the presence of 1 nM sol nAcChoR-I-Bgt indicate quantitative precipitation of antigen by mcab 56 and 73 at low dilution (Fig. 2). By contrast, only about 30% of available antigen is precipitated at saturation by mcab 60. Other experiments (data not shown) demonstrate that once 30% of available sol nAcChoR-I-Bgt have been removed from reaction mixtures by immunoprecipitation with mcab 60, no further precipitation of antigen occurs on addition of excess 60, and that precipitable and non-precipitable antigens are fully complexed with I-Bgt throughout the course of the experiment. Data obtained from 800
i
i
20
40
x "o
D.
Q" 400
t
I
['=Sl-~-Bgt : nAChR]
I
60 (n_M)
Fig. 1. Antigen titration, immunoprecipitation. In a 30-/.tl reaction volume are mixed sol nAcChoR-I-Bgt (100 c p m / f m o l ) at the final concentration indicated on the abscissa and 56 ( × ), 60 (O), or 73 (©), at final dilutions of 1/2430, for 2 h. 3 0 / t l of a 1 / 3 0 0 dilution of G A R is added for 1 h, followed by 10 #1 of staph a for 1 h, and the reaction is quenched as described (Lukas, 1984b). nAcChoR-I-Bgt specifically immunoprecipitated (relative to non-specific sedimentation of nAcChoR-I-Bgt in the presence of rat normal serum at the same dilution) is plotted on the ordinate. Less than 100 cpm nAcChoR-I-Bgt is precipitated in control samples.
132
=/
-,o
0 e~
,q. tt'5
0
t--
,-d ,-J
¢.
e-q
"U~
e-
t,.., 0
o.
0
i.
?
<
~.
.r,~
© ,..d
.< [..,
.<
' r , ~"= -
o
133 Fig. 2 yields values for antigen-recognition efficiency and apparent affinity (Lukas, 1984b), which are summarized in Table I. Immunoprecipitation assays with memb n A c C h o R - I - B g t were found to yield higher apparent antibody titers than experiments with sol n A c C h o R - I - B g t . In addition, quantitative precipitation of m e m b nAcChoR-1-Bgt was effected by mcab 60, as well as by mcab 56 and 73. Characteristic values for radioimmunoassay parameters obtained by use of this protocol are summarized in Table I. Competition r a d i o i m m u n o a s s a y results indicated that 50% inhibition of nAcC h o R - I - B g t precipitation by 2.8 nM 56, 1.4 nM 60, or 2.6 nM 73, occurs in the presence of 2, 4, or 2 nM sol n A c C h o R , respectively.
Solid phase, Protein A coat radioimrnunoassay Poly-(vinylchloride) microtiter wells, coated with Protein A as described (Lukas, 1984b), were used to immobilize G A R for subsequent reaction with mcab and radiolabeled antigen. The requirement for second antibody was found to be similar to the requirement for interaction of mcab and I-Protein A (see below). A n t i b o d y titration curves indicated that the mcab binding capacity of immobilized Protein A - G A R complexes (6 fmol immobilized IgG per well) saturates on addition of mcab at 1 / 3 0 dilution. At that dilution, wells treated with 56 and 73 specifically b o u n d 1000-1200 fmol m e m b n A c C h o R - I - B g t or 12 fmol sol n A c C h o R I-Bgt at antigen excess, while wells treated with mcab 60 showed little m e m b n A c C h o R - I - B g t binding capacity. However, apparent dissociation constants were 3 - 1 0 times higher than in analogous immunoprecipitation assays.
Solid phase, antigen coat radioimmunoassay The n A c C h o R binding capacity of poly-(vinylchloride) microtiter wells has been ,
,
(¢ //
,
A
'
B'
X o
0 0
X
/
0¢
o
"5.
/•0
/
o X
c
"
I
/
-° .
,
~//
c~
1/10 1S 1/2 antibody added (dilution)
z~
1/1
-3 '
"
;-2"
Jt
-^- 5a-
antibody added [log dln)
Fig. 2. Antibody titration, immunoprecipitation. In a 30-#1 reaction volume are mixed 10 ~tl of sol nAcChoR-I-Bgt ( - 180 cpm/fmol) at a concentration of 3 nM and 20 #1 of 56 ( x ), 60 (e), 73 (©) or P3 (zx) at the dilutions indicated on the abscissa. Following a 1-h incubation, 20 Izl of a 1/60 dilution of GAR and 10 #1 of staph a are added. After a 1-h incubation, samples are processed as described, and nAcChoR-I-Bgt precipitated is plotted on the ordinate.
134
described previously (Lukas, 1984b). Routine determinations of the second antibody dependence for I-Protein A binding to immobilized antigen-mcab complexes indicated that addition of 0.1-0.3/tl G A R is adequate to provide maximal I-Protein A binding. I-Protein A titration curves showed that maximum specific radiolabel binding occurs at about 25 nM I-Protein A, and that the apparent dissociation constant for I-Protein A binding to immobilized a n t i g e n - m c a b - G A R complexes is about 3 4 nM. Non-specific binding levels are considerable, and vary linearly with I-Protein A, according to the equation: cpm bound = 10 [I-Protein A (nM)]. Illustrated in Fig. 3 is an antibody titration curve for I-Protein A binding to immobilized a n t i g e n - m c a b - G A R complexes (60 fmol immobilized memb nAcChoR). Specific binding through mcab 56 or 73 saturates at about 300 nl applied mcab, and about 3.3 fmol I-Protein A is bound. Less than 2 fmol I-Protein A is bound to wells incubated with up to 30 ~tl mcab 60, while non-specific binding in the presence of P3 myeloma line supernatant remains at background levels, independent of the amount of added control material. If memb nAcChoR (100 nM) is added to the fluid phase along with mcab, no significant binding of I-Protein A is detected (data not shown). While I-Protein A binding capacity of wells treated with mcab 56 or 73 was found to be unaffected by addition of Bgt or cholinergic ligand, mcab 60-nAcChoR interaction was 50% inhibited in the presence of 10/~M Bgt and 75% inhibited in the presence of 500/~M carbamylcholine (data not shown).
Membrane pellet radioimmunoassay I-Protein A titration curves for binding through G A R to mcab complexed with memb nAcChoR were found to saturate at radioligand concentrations of about 60 nM, and the reaction was characterized by an apparent dissociation constant of 15 nM. 20
16
.o
8
9" -
4 IK
o. 3
Zl
A
o13 antibody added
3'0 (/~1)
Fig. 3. Antibody titration: solid phase, antigen coat. Microtiter wells coated with 30/tl of n A c C h o R are subsequently treated with a 30-/~I sample containing the indicated a m o u n t s (e), 73 (O), or P3 (zx). Wells are then sequentially exposed to 30/~l of a 1/100 dilution of #1 of 3 nM I-Protein A (380 c p m / f m o l ) , prior to assay of I-Protein A bound (plotted on
20 nM m e m b of 56 ( × ) , 60 GAR, and 30 the ordinate).
135
The second antibody dependence for precipitation of I-Protein A with antigen-mcab complexes indicated that about 2 ffl G A R is sufficient to react with memb nAcChoR-associated mcab, and provide a substrate for interaction with l-Protein A. Antigen titration curves for interaction with mcab at a final dilution of 1/120 are shown in Fig. 4. At high concentrations of added nAcChoR, about 20, 40 or 20 fmol I-Protein A are specifically bound to antigen-mcab complexes for incubation with 1 ffl 56, 60, or 73, respectively. Based on the tentative assumption that one mol I-Protein A is bound per mol immune complex, these results yield approximate concentrations of nAcChoR-ab in stock hybridoma supernatants as shown in Table I. Also shown (Table I) are apparent dissociation constants (with respect to nAcChoR concentration) estimated from the data in Fig. 4. Fig. 5 illustrates antibody titration curves for interaction of memb nAcChoR (200 fmol added) with mcab. Levels of I-Protein A bound through second antibody to membrane-associated P3 remain essentially at background, independent of the amount of control supernatant added. Specific binding through 73 approaches saturation at relatively low levels of added supernatant, while saturation is not achieved for 56 or 60, even on addition of 20 #1 of the stock hybridoma supernatant. Nevertheless, if one assumes that 2 mol I-Protein A is bound per mol memb nAcChoR, antigen-recognition efficiencies and apparent affinities may be derived from the amount of antibody required to yield 50% maximal interaction with memb nAcChoR (Table I). Toxin binding inhibition B i n d i n g of I-Bgt to m e m b n A c C h o R , i m m o b i l i z e d o n poly-(vinylchloride) microt i t e r wells, w a s i n h i b i t e d b y 30% w h e n s a m p l e s w e r e t r e a t e d w i t h 30 ~1 o f a 1 / 1 5 i
25
A
Q
2O
X //
,~ 1o
5
? 210
I 40
60
[nAChRlf (nM)
Fig. 4. Antigen titration, membrane pellet radioimmunoassay. In a 120-~1 reaction volume are mixed memb nAcChoR at the final concentrations indicated (abscissa), GAR at a 1/360 final dilution, and 56 ( x ) , 60 (O) or 73 (O) at a final dilution of 1/120. Samples are sedimented and resuspended with 40 #1 of 24 nM I-Protein A (610 cpm/fmol) prior to assay as described. Control sample I-Protein A binding is - 5000 cpm at all nAcChoR concentrations.
136
g.// •
S///
// •
5
/
_~R ~--'-~
antibody added
(/~1)
;
0
2'0
Fig. 5. Antibody titration, membrane pellet radioimmunoassay. Memb nAcChoR (40 /~1 or 5 nM) and 0-20/L1 (as indicated on the abscissa) of 56 ( x ) , 60 (e), 73 (O), or P3 (zx) are mixed in a 60-/d reaction volume for 1 h, prior to addition of 20 /L1 of a 1 / 6 dilution of GAR, and sample sedimentation. Subsequent reaction with 40 p,l of 53 n M l-Protein A (930 c p m / f m o l ) and treatment for l-Protein A binding determination is as described.
dilution of 60. Binding of radiotoxin (1 nM final concentration) to memb nAcChoR (1 nM final concentration) was blocked by 25% in the presence of mcab 60 at a final dilution of 1/15 and 50% at a 1/1.5 dilution of mcab. Neither mcab 56 nor mcab 73 blocked I-Bgt binding, even at very low dilutions. Similar results were obtained for memb nAcChoR in suspension, or for sol nAcChoR.
Discussion
Monoclonal antibodies directed against detergent-solubilized and purified nAcChoR from Torpedo californica (Gomez et al., 1979) interact with both sol and memb Torpedo nAcChoR, as judged by the results of four different radioimmunoassay paradigms. These antibodies have been assigned to heavy chain classes IgG2a and IgG2b (Richman et al., 1980). According to each of the four radioimmunoassay paradigms utilized, they interact very poorly with staph a or native- or zzsI-labeled iodinated-Protein A. Assays for mcab activity succeed only when a second antib o d y / g o a t anti-rat IgG antisera incubation step is included in each assay. Nevertheless, the interactions between mono-specific anti-nAcChoR antibodies and sol and memb nAcChoR are well-behaved, and yield useful data compatible with rigorous analysis of typical ligand-receptor interactions. Data obtained by analysis of solid phase, Protein A coat and immunoprecipitation radioimmunoassays antibody titration curves is not quantitatively comparable,
137 insofar as it refers to antibody binding capacities and affinities of staph a and microtiter well-immobilized Protein A. Nevertheless, antigen titration curves show lower apparent affinity of mcab for antigen in solid phase assay, as opposed to immunoprecipitation radioimmunoassay, which may reflect steric hindrance of antibody combining sites on microtiter wells, or some degree of loss of conformational specificity of poly-(vinylchloride)-bound immune complexes. Second antibody dependence and I-Protein A titration curves are similar for both solid phase, antigen coat and membrane pellet radioimmunoassays, given that higher concentrations of mcab are used in the membrane pellet paradigm. However, the quantity of bound I-Protein A at saturation in the antigen coat assay is only a small fraction of the quantity of immobilized nAcChoR, as measured in toxin binding assay. This result suggests that poly-(vinylchloride)-bound memb nAcChoR have lost antigenic specificity, or that antigenic sites in the solid phase are less accessible to antibody, second antibody, and I-Protein A than to I-Bgt. The consequent reduction in effective concentration of antigenic sites partially accounts for the observed saturation of sites by mcab at high dilution. In both solid phase, Protein A coat and immunoprecipitation assays, mcab 60 removes only 30% of sol nAcChoR-I-Bgt from the fluid phase of the reaction mixture. No further precipitation/immobilization of sol nAcChoR-I-Bgt in reaction mixture supernatants is effected on supplemental reaction with large quantities of mcab 60. A potentially trivial explanation of these results is discounted by the observation that all toxin binding sites are occupied throughout the course of the experiment. Thus, this result suggests that mcab 60 detects a subset of sol nAcChoRI-Bgt, which in turn might reflect a heterogeneity in solubilized receptor-toxin complexes. Quantitative precipitation of memb nAcChoR-I-Bgt is obtained in immunoprecipitation assays with mcab 60. Thus, heterogeneity of nAcChoR-I-Bgt recognition by mcab 60 is only manifest in detergent solution, or there is uniform distribution on Torpedo membranes of nAcChoR-I-Bgt classes that show heterogeneity in interaction with mcab 60. Other results are consistent with idiosyncratic interaction of mcab 60 with nAcChoR-I-Bgt. For example, the molar ratio of sol nAcChoR to antibody needed to provide 50% inhibition of sol nAcChoR-I-Bgt precipitation is 1 : 1 for mcab 56 and 73, but 3:1 for mcab 60. Thus, mcab 60s antigen binding capacity is approx. 3-fold higher for interaction with sol nAcChoR than with sol nAcChoR-I-Bgt. Moreover, measures of antigen-recognition efficiency taken from antibody titration curves are higher for mcab 60 than mcab 56 or 73 in the membrane pellet assay, whereas 56 and 73 show higher antigen-recognition efficiency than mcab 60 in the other assays. These results suggest that the affinity of mcab 60 for native nAcChoR is higher in the absence of Bgt, which is also suggested by the ability of antibody to partially block toxin binding to nAcChoR. Thus, the present results are consistent with observations regarding the sensitivity of mcab 60 interaction with nAcChoR to the presence of Bgt (as judged by a hemagglutinin assay), and with the interpretation that antigenic determinants recognized by mcab 60 partially overlap with the nAcChoR toxin binding site (see Gomez et al., 1979). There are quantitative differences in the estimates of the concentrations of
138
anti-AcChoR antibodies in hybridoma supernatant stock solutions, as gathered from antigen titration curves performed by use of immunoprecipitation and membrane pellet radioimmunoassays. These results are observed with polyclonal antisera as well (Lukas, 1984b), and predictably reflect non-equilibrium conditions in membrane pellet assays, and enhanced efficiency of memb nAcChoR-I-Bgt immunoprecipitation, relative to immunoprecipitation of sol nAcChoR-I-Bgt. Nevertheless, antigen titration curve-based estimates of antibody-antigen dissociation constants from membrane pellet assays are in agreement with similar estimates based on immunoprecipitation assays. In summary, characteristics of monoclonal anti-nAcChoR antibody interaction with Torpedo nAcChoR have been found to be sensitive to different aspects of antigen-antibody specificity a n d / o r immunoassay design. These observations suggest that a battery of immunoassays may be necessary to realize full characterization of antibody-antigen interactions.
Acknowledgements The author acknowledges the generous gifts of hybridoma and myeloma supernatants provided by Dr. David Richman, the excellent secretarial services of Deirdre Anne Janus, and the expert technical assistance of Mary Jane Cullen.
References Gomez, C.M., D.P. Richman, P.W. Berman, S.A. Burres, B.G.W. Arnason and F.W. Fitch, 1979, Biochem. Biophys. Res. Commun. 88, 575. KOhler, G. and C. Milstein, 1975, Nature (London) 256, 495. Lennon, V.A., M. Thompson and J. Chen, 1980, J. Biol. Chem. 255, 4395. Lukas, R.J., 1984a, Biochem. 23, 1152. Lukas, R.J., 1984b, J. Immunol. Methods, submitted. Mochly-Rosen, D., S. Fuchs and Z. Eshhar, 1979, FEBS Len. 106, 389. Morley, B.J. and G.E. Kemp, 1981, Brain Res. Rev. 3, 81. Richman, D.P., C.M. Gomez, P.W. Berman, S.A. Burres, F.W. Fitch and B.G.W. Aranson, 1980, Nature (London) 286, 738. Schmidt, J., S.P. Hunt and G. Polz-Tejera, 1980, in: Neurotransminers, Receptors and Drug Action, ed. W.B. Essman (Spectrum Publishing, New York) p. 1. Tzartos, S.J. and J.M. Lindstrom, 1980, Proc. Natl. Acad. Sci. U.S.A. 77, 755. Yelton, D.E. and M.D. Scharff, 1981, Ann. Rev. Biochem. 50, 657.
129
JIM 03252
Monoclonal Rat Anti-Torpedo Electroplax Nicotinic Acetylcholine Receptor Antibodies" Immunochemical Characterization * Ronald J. Lukas Division of Neurobiology, Barrow Neurological Institute, 350 West Thomas Road, Phoenix, A Z 85013, U.S.A.
(Received 14 June 1984, accepted 18 July 1984)
The interaction of Torpedo californica nicotinic acetylcholine receptor (nAcChoR) with three rat monoclonal antibodies (mcab) directed against nAcChoR (Gomez et al., 1979) was studied by use of four different radioimmunoassay protocols. Each mcab reacts poorly with formalin-fixed Staphylococcus aureus or S. aureus Protein A, which requires that an additional incubation with second antibody (goat anti-rat immunoglobulin G) is included in each radioimmunoassay paradigm. One mcab inhibits 125I-labeled a-bungarotoxin binding to nAcChoR, recognizes a subset of solubilized nAcChoR-toxin complexes, and shows higher titer against nAcChoR in the absence of toxin. Thus, it appears to be directed against nAcChoR antigenic determinants that at least partially overlap with toxin binding sites. Two other mcab react with nAcChoR in a toxin-independent manner. Receptor-mcab dissociation constants are less than 10 nM, according to each assay paradigm. Estimates of antibody titer or concentrations of antibody in stock hybridoma supernatants vary according to the assay used. This predictable result is attributed to differences in design and sensitivity of assay protocols. The data provide a basis for further utilization of monospecific antibodies as probes in characterization of nAcChoR structure and function.
Key words: acetylcholine receptor - nicotinic receptor - monoclonal antibodies
* Supported by National Institutes of Health Grant NS-16821, by a grant from the Epilepsy Foundation of America, by funds from Epi-Hab of Arizona, Inc., and by the Men's and Women's Boards of the Barrow Neurological Foundation. Abbreviations: CNS, central nervous system; AcChoR, acetylcholine receptor; nAcChoR, nicotinic acetylcholine receptor; IgG, immunoglobulin G; mcab, monoclonal antibody; Bgt, a-bungarotoxin; I-Bgt, 125I-labeled monoiodinated a-bungarotoxin; memb nAcChoR, nicotinic acetylcholine receptor-rich membranes from Torpedo californica electric organ; sol nAcChoR, detergent-solubilized nAcChoR from Torpedo; staph a, formalin-fixed Staphylococcus aureus cells; GAR, goat anti-rat IgG antiserum; I-Protein A, 125I-labeled iodinated Protein A; TNP7.4-BSA, 100 mM NaCl, 50 mM NaEHPO4, pH 7.4 supplemented with 1% Triton X-100 and 0.1% bovine serum albumin; NP7.4-BSA, 100 mM NaC1, 50 mM Na2HPO4, pH 7.4, supplemented with 0.1% bovine serum albumin. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
130 Introduction
Our ultimate understanding of molecular mechanisms of central nervous system (CNS) acetylcholine receptor (AcChoR) function is dependent upon identification of suitable receptor-specific probes. Small cholinergic ligands and atropinemimetic and curaremimetic neurotoxins have been used to identify putative CNS AcChoR sites, but there remains uncertainty regarding their functional relevance and potency (Schmidt et al., 1980; Morley and Kemp, 1981). In a previous manuscript (Lukas, 1984b), interaction between nicotinic AcChoR (nAcChoR) from Torpedo californica electric organ and antibodies in polyclonal antisera raised against nAcChoR from Electrophorus" electricus electric organ was described. While antisera-antigen interactions are well-behaved, and provide information on antibody and nAcChoR characteristics, the heterogeneity of immunoglobulin G (IgG) species in the antisera, and their limited availability, restricts rigorous application of these antibodies as receptor-specific ligands. Mono-specific anti-nAcChoR antibodies, obtained in large quantities by using the hybridoma technique (K6hler and Milstein, 1975), may prove more useful as molecular probes in characterization of receptor structure and function (Mochly-Rosen et al., 1979: Lennon et al., 1980; Tzartos and Lindstrom, 1980; Yelton and Scharff, 1981). Hybridoma supernatants containing monoclonal antibodies (mcab) directed against Torpedo nAcChoR have been obtained (Gomez et al., 1979; Richman et al., 1980). With a view to using these mcab as probes for authentic nAcChoR in the mammalian CNS, characterization of mcab-Torpedo nAcChoR interactions has been undertaken.
Materials and Methods
Preparation of a-bungarotoxin (Bgt), 125I-labeled Bgt (I-Bgt), and nAcChoR-rich membranes from Torpedo (memb nAcChoR) is as described previously (Lukas, 1984a). Procedures for solubilization of Torpedo nAcChoR (sol nAcChoR), preparation of toxin- and radiotoxin-labeled memb and sol nAcChoR, iodination of Protein A, and treatment of formalin-fixed Staphylococcus aureus cells (staph a), also appear elsewhere (Lukas, 1984b). Media from three hybridoma clonal lines, derived from fusion of spleen cells of Torpedo nAcChoR-innoculated female Lewis rats with cells of the murine myeloma line P3-X63-Ag8, were the generous gift of Dr. David Richman of the University of Chicago. Clones 56I, 60D and 73G each produce anti-Torpedo nAcChoR mcab, which are designated in this manuscript as 56, 60, and 73, respectively. Dr. Richman also supplied supernatants from the parent myeloma line, which are designated as P3. Goat anti-rat IgG antiserum (GAR; Cappel 0113-0911) and culture supernatants are fractionated into 1-2 ml aliquots, and stored at - 2 0 ° C or - 8 0 ° C until use. Experimental protocols for four different radioimmunoassay schemes, described in detail previously (Lukas, 1984b), are modified for assay of mcab activity by insertion of a second antibody incubation step between mcab-antigen reactions and
131
staph a or 125I-labeled iodinated Protein A (I-Protein A) addition. Toxin binding inhibition assays are also as previously described (Lukas, 1984b).
Results
Immunoprecipitation assay Mcab-antigen complexes were not precipitated by staph a in the absence of GAR. Experiments were conducted routinely to establish the staph a and GAR requirements for quantitative immunoprecipitation of mcab-nAcChoR complexes under the assay conditions used below. The antigen dependence for immunoprecipitation of sol nAcChoR-I-Bgt with mcab is shown in Fig. 1. In each case, saturation of antibody capacity is achieved at high concentrations of antigen. About 6.8, 3.3, and 6.5 fmol sol nAcChoR-I-Bgt are precipitated by the given dilution of 56, 60, and 73, respectively, which provides estimates of the concentration of mcab in stock hybridoma supernatants given in Table I. Also provided in Table I are values for the apparent dissociation constant for mcab-sol nAcChoR interaction. Antibody titration curves obtained in the presence of 1 nM sol nAcChoR-I-Bgt indicate quantitative precipitation of antigen by mcab 56 and 73 at low dilution (Fig. 2). By contrast, only about 30% of available antigen is precipitated at saturation by mcab 60. Other experiments (data not shown) demonstrate that once 30% of available sol nAcChoR-I-Bgt have been removed from reaction mixtures by immunoprecipitation with mcab 60, no further precipitation of antigen occurs on addition of excess 60, and that precipitable and non-precipitable antigens are fully complexed with I-Bgt throughout the course of the experiment. Data obtained from 800
i
i
20
40
x "o
D.
Q" 400
t
I
['=Sl-~-Bgt : nAChR]
I
60 (n_M)
Fig. 1. Antigen titration, immunoprecipitation. In a 30-/.tl reaction volume are mixed sol nAcChoR-I-Bgt (100 c p m / f m o l ) at the final concentration indicated on the abscissa and 56 ( × ), 60 (O), or 73 (©), at final dilutions of 1/2430, for 2 h. 3 0 / t l of a 1 / 3 0 0 dilution of G A R is added for 1 h, followed by 10 #1 of staph a for 1 h, and the reaction is quenched as described (Lukas, 1984b). nAcChoR-I-Bgt specifically immunoprecipitated (relative to non-specific sedimentation of nAcChoR-I-Bgt in the presence of rat normal serum at the same dilution) is plotted on the ordinate. Less than 100 cpm nAcChoR-I-Bgt is precipitated in control samples.
132
=/
-,o
0 e~
,q. tt'5
0
t--
,-d ,-J
¢.
e-q
"U~
e-
t,.., 0
o.
0
i.
?
<
~.
.r,~
© ,..d
.< [..,
.<
' r , ~"= -
o
133 Fig. 2 yields values for antigen-recognition efficiency and apparent affinity (Lukas, 1984b), which are summarized in Table I. Immunoprecipitation assays with memb n A c C h o R - I - B g t were found to yield higher apparent antibody titers than experiments with sol n A c C h o R - I - B g t . In addition, quantitative precipitation of m e m b nAcChoR-1-Bgt was effected by mcab 60, as well as by mcab 56 and 73. Characteristic values for radioimmunoassay parameters obtained by use of this protocol are summarized in Table I. Competition r a d i o i m m u n o a s s a y results indicated that 50% inhibition of nAcC h o R - I - B g t precipitation by 2.8 nM 56, 1.4 nM 60, or 2.6 nM 73, occurs in the presence of 2, 4, or 2 nM sol n A c C h o R , respectively.
Solid phase, Protein A coat radioimrnunoassay Poly-(vinylchloride) microtiter wells, coated with Protein A as described (Lukas, 1984b), were used to immobilize G A R for subsequent reaction with mcab and radiolabeled antigen. The requirement for second antibody was found to be similar to the requirement for interaction of mcab and I-Protein A (see below). A n t i b o d y titration curves indicated that the mcab binding capacity of immobilized Protein A - G A R complexes (6 fmol immobilized IgG per well) saturates on addition of mcab at 1 / 3 0 dilution. At that dilution, wells treated with 56 and 73 specifically b o u n d 1000-1200 fmol m e m b n A c C h o R - I - B g t or 12 fmol sol n A c C h o R I-Bgt at antigen excess, while wells treated with mcab 60 showed little m e m b n A c C h o R - I - B g t binding capacity. However, apparent dissociation constants were 3 - 1 0 times higher than in analogous immunoprecipitation assays.
Solid phase, antigen coat radioimmunoassay The n A c C h o R binding capacity of poly-(vinylchloride) microtiter wells has been ,
,
(¢ //
,
A
'
B'
X o
0 0
X
/
0¢
o
"5.
/•0
/
o X
c
"
I
/
-° .
,
~//
c~
1/10 1S 1/2 antibody added (dilution)
z~
1/1
-3 '
"
;-2"
Jt
-^- 5a-
antibody added [log dln)
Fig. 2. Antibody titration, immunoprecipitation. In a 30-#1 reaction volume are mixed 10 ~tl of sol nAcChoR-I-Bgt ( - 180 cpm/fmol) at a concentration of 3 nM and 20 #1 of 56 ( x ), 60 (e), 73 (©) or P3 (zx) at the dilutions indicated on the abscissa. Following a 1-h incubation, 20 Izl of a 1/60 dilution of GAR and 10 #1 of staph a are added. After a 1-h incubation, samples are processed as described, and nAcChoR-I-Bgt precipitated is plotted on the ordinate.
134
described previously (Lukas, 1984b). Routine determinations of the second antibody dependence for I-Protein A binding to immobilized antigen-mcab complexes indicated that addition of 0.1-0.3/tl G A R is adequate to provide maximal I-Protein A binding. I-Protein A titration curves showed that maximum specific radiolabel binding occurs at about 25 nM I-Protein A, and that the apparent dissociation constant for I-Protein A binding to immobilized a n t i g e n - m c a b - G A R complexes is about 3 4 nM. Non-specific binding levels are considerable, and vary linearly with I-Protein A, according to the equation: cpm bound = 10 [I-Protein A (nM)]. Illustrated in Fig. 3 is an antibody titration curve for I-Protein A binding to immobilized a n t i g e n - m c a b - G A R complexes (60 fmol immobilized memb nAcChoR). Specific binding through mcab 56 or 73 saturates at about 300 nl applied mcab, and about 3.3 fmol I-Protein A is bound. Less than 2 fmol I-Protein A is bound to wells incubated with up to 30 ~tl mcab 60, while non-specific binding in the presence of P3 myeloma line supernatant remains at background levels, independent of the amount of added control material. If memb nAcChoR (100 nM) is added to the fluid phase along with mcab, no significant binding of I-Protein A is detected (data not shown). While I-Protein A binding capacity of wells treated with mcab 56 or 73 was found to be unaffected by addition of Bgt or cholinergic ligand, mcab 60-nAcChoR interaction was 50% inhibited in the presence of 10/~M Bgt and 75% inhibited in the presence of 500/~M carbamylcholine (data not shown).
Membrane pellet radioimmunoassay I-Protein A titration curves for binding through G A R to mcab complexed with memb nAcChoR were found to saturate at radioligand concentrations of about 60 nM, and the reaction was characterized by an apparent dissociation constant of 15 nM. 20
16
.o
8
9" -
4 IK
o. 3
Zl
A
o13 antibody added
3'0 (/~1)
Fig. 3. Antibody titration: solid phase, antigen coat. Microtiter wells coated with 30/tl of n A c C h o R are subsequently treated with a 30-/~I sample containing the indicated a m o u n t s (e), 73 (O), or P3 (zx). Wells are then sequentially exposed to 30/~l of a 1/100 dilution of #1 of 3 nM I-Protein A (380 c p m / f m o l ) , prior to assay of I-Protein A bound (plotted on
20 nM m e m b of 56 ( × ) , 60 GAR, and 30 the ordinate).
135
The second antibody dependence for precipitation of I-Protein A with antigen-mcab complexes indicated that about 2 ffl G A R is sufficient to react with memb nAcChoR-associated mcab, and provide a substrate for interaction with l-Protein A. Antigen titration curves for interaction with mcab at a final dilution of 1/120 are shown in Fig. 4. At high concentrations of added nAcChoR, about 20, 40 or 20 fmol I-Protein A are specifically bound to antigen-mcab complexes for incubation with 1 ffl 56, 60, or 73, respectively. Based on the tentative assumption that one mol I-Protein A is bound per mol immune complex, these results yield approximate concentrations of nAcChoR-ab in stock hybridoma supernatants as shown in Table I. Also shown (Table I) are apparent dissociation constants (with respect to nAcChoR concentration) estimated from the data in Fig. 4. Fig. 5 illustrates antibody titration curves for interaction of memb nAcChoR (200 fmol added) with mcab. Levels of I-Protein A bound through second antibody to membrane-associated P3 remain essentially at background, independent of the amount of control supernatant added. Specific binding through 73 approaches saturation at relatively low levels of added supernatant, while saturation is not achieved for 56 or 60, even on addition of 20 #1 of the stock hybridoma supernatant. Nevertheless, if one assumes that 2 mol I-Protein A is bound per mol memb nAcChoR, antigen-recognition efficiencies and apparent affinities may be derived from the amount of antibody required to yield 50% maximal interaction with memb nAcChoR (Table I). Toxin binding inhibition B i n d i n g of I-Bgt to m e m b n A c C h o R , i m m o b i l i z e d o n poly-(vinylchloride) microt i t e r wells, w a s i n h i b i t e d b y 30% w h e n s a m p l e s w e r e t r e a t e d w i t h 30 ~1 o f a 1 / 1 5 i
25
A
Q
2O
X //
,~ 1o
5
? 210
I 40
60
[nAChRlf (nM)
Fig. 4. Antigen titration, membrane pellet radioimmunoassay. In a 120-~1 reaction volume are mixed memb nAcChoR at the final concentrations indicated (abscissa), GAR at a 1/360 final dilution, and 56 ( x ) , 60 (O) or 73 (O) at a final dilution of 1/120. Samples are sedimented and resuspended with 40 #1 of 24 nM I-Protein A (610 cpm/fmol) prior to assay as described. Control sample I-Protein A binding is - 5000 cpm at all nAcChoR concentrations.
136
g.// •
S///
// •
5
/
_~R ~--'-~
antibody added
(/~1)
;
0
2'0
Fig. 5. Antibody titration, membrane pellet radioimmunoassay. Memb nAcChoR (40 /~1 or 5 nM) and 0-20/L1 (as indicated on the abscissa) of 56 ( x ) , 60 (e), 73 (O), or P3 (zx) are mixed in a 60-/d reaction volume for 1 h, prior to addition of 20 /L1 of a 1 / 6 dilution of GAR, and sample sedimentation. Subsequent reaction with 40 p,l of 53 n M l-Protein A (930 c p m / f m o l ) and treatment for l-Protein A binding determination is as described.
dilution of 60. Binding of radiotoxin (1 nM final concentration) to memb nAcChoR (1 nM final concentration) was blocked by 25% in the presence of mcab 60 at a final dilution of 1/15 and 50% at a 1/1.5 dilution of mcab. Neither mcab 56 nor mcab 73 blocked I-Bgt binding, even at very low dilutions. Similar results were obtained for memb nAcChoR in suspension, or for sol nAcChoR.
Discussion
Monoclonal antibodies directed against detergent-solubilized and purified nAcChoR from Torpedo californica (Gomez et al., 1979) interact with both sol and memb Torpedo nAcChoR, as judged by the results of four different radioimmunoassay paradigms. These antibodies have been assigned to heavy chain classes IgG2a and IgG2b (Richman et al., 1980). According to each of the four radioimmunoassay paradigms utilized, they interact very poorly with staph a or native- or zzsI-labeled iodinated-Protein A. Assays for mcab activity succeed only when a second antib o d y / g o a t anti-rat IgG antisera incubation step is included in each assay. Nevertheless, the interactions between mono-specific anti-nAcChoR antibodies and sol and memb nAcChoR are well-behaved, and yield useful data compatible with rigorous analysis of typical ligand-receptor interactions. Data obtained by analysis of solid phase, Protein A coat and immunoprecipitation radioimmunoassays antibody titration curves is not quantitatively comparable,
137 insofar as it refers to antibody binding capacities and affinities of staph a and microtiter well-immobilized Protein A. Nevertheless, antigen titration curves show lower apparent affinity of mcab for antigen in solid phase assay, as opposed to immunoprecipitation radioimmunoassay, which may reflect steric hindrance of antibody combining sites on microtiter wells, or some degree of loss of conformational specificity of poly-(vinylchloride)-bound immune complexes. Second antibody dependence and I-Protein A titration curves are similar for both solid phase, antigen coat and membrane pellet radioimmunoassays, given that higher concentrations of mcab are used in the membrane pellet paradigm. However, the quantity of bound I-Protein A at saturation in the antigen coat assay is only a small fraction of the quantity of immobilized nAcChoR, as measured in toxin binding assay. This result suggests that poly-(vinylchloride)-bound memb nAcChoR have lost antigenic specificity, or that antigenic sites in the solid phase are less accessible to antibody, second antibody, and I-Protein A than to I-Bgt. The consequent reduction in effective concentration of antigenic sites partially accounts for the observed saturation of sites by mcab at high dilution. In both solid phase, Protein A coat and immunoprecipitation assays, mcab 60 removes only 30% of sol nAcChoR-I-Bgt from the fluid phase of the reaction mixture. No further precipitation/immobilization of sol nAcChoR-I-Bgt in reaction mixture supernatants is effected on supplemental reaction with large quantities of mcab 60. A potentially trivial explanation of these results is discounted by the observation that all toxin binding sites are occupied throughout the course of the experiment. Thus, this result suggests that mcab 60 detects a subset of sol nAcChoRI-Bgt, which in turn might reflect a heterogeneity in solubilized receptor-toxin complexes. Quantitative precipitation of memb nAcChoR-I-Bgt is obtained in immunoprecipitation assays with mcab 60. Thus, heterogeneity of nAcChoR-I-Bgt recognition by mcab 60 is only manifest in detergent solution, or there is uniform distribution on Torpedo membranes of nAcChoR-I-Bgt classes that show heterogeneity in interaction with mcab 60. Other results are consistent with idiosyncratic interaction of mcab 60 with nAcChoR-I-Bgt. For example, the molar ratio of sol nAcChoR to antibody needed to provide 50% inhibition of sol nAcChoR-I-Bgt precipitation is 1 : 1 for mcab 56 and 73, but 3:1 for mcab 60. Thus, mcab 60s antigen binding capacity is approx. 3-fold higher for interaction with sol nAcChoR than with sol nAcChoR-I-Bgt. Moreover, measures of antigen-recognition efficiency taken from antibody titration curves are higher for mcab 60 than mcab 56 or 73 in the membrane pellet assay, whereas 56 and 73 show higher antigen-recognition efficiency than mcab 60 in the other assays. These results suggest that the affinity of mcab 60 for native nAcChoR is higher in the absence of Bgt, which is also suggested by the ability of antibody to partially block toxin binding to nAcChoR. Thus, the present results are consistent with observations regarding the sensitivity of mcab 60 interaction with nAcChoR to the presence of Bgt (as judged by a hemagglutinin assay), and with the interpretation that antigenic determinants recognized by mcab 60 partially overlap with the nAcChoR toxin binding site (see Gomez et al., 1979). There are quantitative differences in the estimates of the concentrations of
138
anti-AcChoR antibodies in hybridoma supernatant stock solutions, as gathered from antigen titration curves performed by use of immunoprecipitation and membrane pellet radioimmunoassays. These results are observed with polyclonal antisera as well (Lukas, 1984b), and predictably reflect non-equilibrium conditions in membrane pellet assays, and enhanced efficiency of memb nAcChoR-I-Bgt immunoprecipitation, relative to immunoprecipitation of sol nAcChoR-I-Bgt. Nevertheless, antigen titration curve-based estimates of antibody-antigen dissociation constants from membrane pellet assays are in agreement with similar estimates based on immunoprecipitation assays. In summary, characteristics of monoclonal anti-nAcChoR antibody interaction with Torpedo nAcChoR have been found to be sensitive to different aspects of antigen-antibody specificity a n d / o r immunoassay design. These observations suggest that a battery of immunoassays may be necessary to realize full characterization of antibody-antigen interactions.
Acknowledgements The author acknowledges the generous gifts of hybridoma and myeloma supernatants provided by Dr. David Richman, the excellent secretarial services of Deirdre Anne Janus, and the expert technical assistance of Mary Jane Cullen.
References Gomez, C.M., D.P. Richman, P.W. Berman, S.A. Burres, B.G.W. Arnason and F.W. Fitch, 1979, Biochem. Biophys. Res. Commun. 88, 575. KOhler, G. and C. Milstein, 1975, Nature (London) 256, 495. Lennon, V.A., M. Thompson and J. Chen, 1980, J. Biol. Chem. 255, 4395. Lukas, R.J., 1984a, Biochem. 23, 1152. Lukas, R.J., 1984b, J. Immunol. Methods, submitted. Mochly-Rosen, D., S. Fuchs and Z. Eshhar, 1979, FEBS Len. 106, 389. Morley, B.J. and G.E. Kemp, 1981, Brain Res. Rev. 3, 81. Richman, D.P., C.M. Gomez, P.W. Berman, S.A. Burres, F.W. Fitch and B.G.W. Aranson, 1980, Nature (London) 286, 738. Schmidt, J., S.P. Hunt and G. Polz-Tejera, 1980, in: Neurotransminers, Receptors and Drug Action, ed. W.B. Essman (Spectrum Publishing, New York) p. 1. Tzartos, S.J. and J.M. Lindstrom, 1980, Proc. Natl. Acad. Sci. U.S.A. 77, 755. Yelton, D.E. and M.D. Scharff, 1981, Ann. Rev. Biochem. 50, 657.