Monoclonal Anti-idiotypic Antibodies to Opioid Receptors*

Two monoclonal anti-idiotypic antibodies (anti-Id-135 and anti-Id-14, both of the IgM class) which inter- act with the binding site of opioid receptors were generated. A monoclonal anti-@-endorphin antibody (3- E7) which displays binding characteristics for opioid ligands similar to opioid receptors served as the anti- gen (Gramsch, c., Meo, T., Riethmuller, G., and Herz, A., (1983) J, Neurochem. 40, 1220-1226; Meo, T., Gramsch, C., Inan, R., Hollt, V., Weber, E., Herz, A., and Riethmuller, G. (1983) Proc. Nutl. Acad. Sci. U.S.A. 80, 4048-4088) and the hybridomas obtained were screened for anti-idiotypic antibodies with Fab fragments of 3-E7. The anti-idiotypes were then screened for opioid binding to rat brain membrane receptors, yielding several positive clones two of which were more intensively studied. Both anti-idiotypic an- tibodies were about equally potent in displacing the p-and &opioid receptor ligands [3H]dihydromorphine, 1261-labeled &endorphin, [~-Ala~,~-Leu~-~H]enkepha-lin and [3H]naloxone from rat

Our findings demonstrate the successful generation of monoclonal antibodies interacting with membranebound and solubilized opioid receptors of the p-and 6type.
It has been shown in several systems that the anti-idiotypic antibodies (anti-Id)' provide a powerful tool for the characterization and functional modulation of membrane receptors (for review, see Ref. 3). This approach, using polyclonal antibodies, was recently employed to study opioid receptors (4)(5)(6)(7). The anti-idiotypic procedure may be advantageous over the classical method of raising anti-receptor antibodies which * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. requires a purified receptor preparation for immunization. Nevertheless, the conventional technique has been successfully applied to the production of anti-opioid receptor antibodies (8).
We first demonstrated that the anti-idiotypic route of raising antibodies is applicable to opioid receptors ( 5 , 6). The antigen used was the monoclonal anti-0-endorphin antibody 3-E7 which binds opioid peptides in a similar manner to opioid receptors (1,2). The anti-Ids that were generated behaved apparently like an internal image of the 3-E7 antibody, being able to interact with opioid receptors. However, in studying these polyclonal anti-Ids several difficulties are encountered first, due to a low titer, high concentrations of antisera are required to inhibit ligand binding to receptors or to initiate a biological response. Thus, additional processing is necessary for the concentration and purification of the limited amounts of antibodies. Second, the cyclic nature of the anti-idiotypic response may result in the disappearance of anti-idiotypic antibodies and appearance of a further set of antibodies, the anti-anti-isotypes, which neutralize the antiidiotypic antibodies as well as the ligand for the receptor (9).
Considering our successful anti-idiotypic approach to raising antibodies to opioid receptors and to avoid the drawbacks of polyclonal antibodies, we attempted to generate monoclonal anti-idiotypic anti-opioid receptor antibodies, whose unlimited availability would be of considerable advantage. Several monoclonal anti-idiotypic antibodies were obtained from BALB/c mice immunized with 3-E7; two of these were studied in detail and their characteristics are described below.

Immunization
The mouse monoclonal anti-P-endorphin antibody 3-E7 was raised as described (1). Protein A-purified 3-E7 (Fusion I) or its Fab fragment coupled to bovine thyroglobulin (Fusion 11) were used for immunization. BALB/c mice were injected intradermally at multiple sites with 100 pg of antigen emulsified in Freund's complete adjuvant at 4-week intervals. The immunization efficiency was controlled by periodical testing of mice sera for the development of anti-Fab-(3-E7) antibodies by ELISA (details see below). At the end of the immunization period, 100 pg of the immunogen was injected both intraperitoneally and intravenously on 3 consecutive days before the mice spleens were harvested 24 h after the last injection and used for cell fusion (10).

Fusion
The hybridization of the immunized spleen cells with mouse myeloma cells (X63-Ag8.653) were carried out as described by Kearney et al. (11). Cloning of hybrid cells was conducted by limiting dilution with mouse peritoneal feeder cells in microculture plates (Costar). Spleen cells (1 X 10') of a mouse showing the highest titer were mixed with X63-Ag8.653 myeloma cells (1.4 X lo7) and fused by use of polyethylene glycol 4000 (Fusion I) or polyethylene glycol 1500 (Fusion 11). The cells were distributed into the wells of flat-bottomed tissue culture plates and cultivated in Dulbecco's modified Eagle's medium supplemented with 15% fetal calf serum (Boehringer Mann-5853 heim). The medium contained hypoxanthine, aminopterin, and thymidine in a 7% CO,/air atmosphere, at 37 'C. Anti-F(ab'),-(3-E7) activity of spent media was tested by ELISA. Wells which showed high reactivity were identified and grown in serum-free medium to produce sufficient quantities of antibodies in the supernatant for anti-idiotype tests.

Anti-F(abO2-(3-E7) ELISA
The hybridoma culture medium from wells with growing cell lines was assayed for anti-Fab-(3-E7) antibodies by a double-antibody sandwich ELISA. Microtiter plates (Costar 2596) were prepared by coating each well with 150 pl of 250 pg/ml F(ab'),-(3-E7) fragments in PBS (16 h, 37 "C). After washing twice with PBS, 0.1% Tween 20, the wells were blocked with 200 p1 of 3% BSA in PBS (37 "C, 3 h) and again washed with PBS/Tween 20. 100 pl of hybridoma supernatant were transferred to the plates and incubated overnight at room temperature on a shaker. After intense washing with PBS/Tween 20, 100 pl of a 1:lOOO dilution of rabbit anti-mouse IgM(Fc)-peroxidase conjugate dissolved in PBS containing 10% horse serum were added and incubated for 1 h at 37 "C and for 1 h at room temperature. After washing the plates several times, 100 pl of substrate H20,/2,2'-azinodi-3-ethylhenzthiazoline sulfonic acid (13) were added to each well. The enzyme reaction was allowed to proceed for 30-60 min, stopped with 50 pl of 20 mM NaN3, and the absorbance was recorded at 414 nm (Immuno Reader, Intermed, Nunc).

Solid-phase Inhibition Assay
Microtiter plates were coated with @-endorphin (100 pl/well, 300 pmol/ml in PBS) for 16 h at 37 "C. Thereafter, the plates were blocked for 4 h at 37 "C with 3% BSA dissolved in PBS. Subsequently, the plates were washed three times with PBS, 0.1% Tween 20 and twice with distilled water. Samples or solutions of control IgM (MOPC 104E) in 100 p1 of radioimmunoassay buffer (0.02 M phosphate, pH 7.4, containing 0.15 M NaCl, 1% BSA, 0.05% Tween 20, 0.1% polylysine 4000) were incubated at 4 "C overnight together with Iz5I-labeled 3-E7 (20,000 cpm in radioimmunoassay buffer). The plates were then washed three times with PBS/Tween 20 and twice with distilled water. The individual wells were counted for radioactivity in a y-spectrometer.

Iodination of 3-E7
Iodinations of 3-E7 (IgG,.) were performed with chloramine T at pH 7.4 in 0.5 M phosphate buffer according to Bolton and Hunter (14). The molar ratios of reactants were 2:1.25:100100 for Na'"I/ IgG/chloramine T/metabisulfite. The antibody was exposed to chloramine T for 30 s at room temperature and the reaction was stopped by the addition of metabisulfite. The iodinated IgG was first separated on a Sephadex G-25 column (0.5 X 10 cm) and further on a protein A-Sepharose column consisting of 500 pl of gel. The adsorbed Iz5Ilabeled 3-E7 was eluted with citrate, pH 4.0, and immediately neutralized with 1 N NaOH.

Ascites Production
Hybridoma cells (5-10 X lo6) were injected intraperitoneally into pristane-primed nude mice (15). The ascitic fluids were centrifuged 1000 X g for 10 min after dilution in appropriate buffer to remove cells. The supernatant was applied to a protein A-Sepharose column (3-E7 isolation) or treated with ammonium sulfate (50% saturation) for precipitation and further purification of the anti-Ids (IgM class).

Purification of Anti-idiotypic Antibodies from Cell Culture Supernatants and from Ascites
The ammonium sulfate precipitate of IgM was dissolved in 0.025 M phosphate buffer, pH 6.7, and ultrafiltrated on an Amicon XM 300 membrane. Samples were applied to a fast protein liquid chromatog-raphy system, equipped with an anion exchange column (Mono Q HR 5/5, Pharmacia LKB Biotechnology Inc.) and eluted with a step gradient from 0.025 M phosphate, pH 6.7, 0.05% Lubrol PX to 0.3 M phosphate, pH 6.5,0.05% Lubrol PX, at a flow rate of 1.5 ml/min.

Preparation of Fab Fragments from 3-E7
Papain digestion of 3-E7 was carried out by a modified method of Mage (16) using 4 mg of papain/100 mg of IgG. After alkylation with iodoacetamide, the Fc fragments and undigested 3-E7 were removed from the reaction mixture by means of protein A-Sepharose. After ultrafiltration, the flow-through containing the Fab fragments was used for preparation of a thyroglobulin conjugate.
F(ab'), fragments were prepared by pepsin digestion as described (17) with minor modifications: 3-E7 was dissolved in 0.1 M citrate, pH 4.0; the optimal digestion time was determined to be 6 h at 37 "C. The generation of F(ab'), was controlled by hydroxylapatite chromatography (0.01 M NaH2P04, pH 6.8, to 50% 0.5 M NaHZPO,, pH 6.8). The mixture was ultrafiltrated on an Amicon YM 10 membrane and separated on a protein A-Sepharose column as described above. The purity of the fragments was controlled by SDS-PAGE under reducing and nonreducing conditions.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis
To monitor the proteins in the column fractions and to estimate their molecular weight and purity, samples were separated under reducing conditions on a 13% SDS-PAGE slab gel according to the method of Laemmli (18). Gels were stained with Coomassie Blue. Molecular weight standards (low molecular weight protein standards, Bio-Rad) consisted of phosphorylase b (molecular weight 92,500), BSA (66,200), ovalbumin (45,000), carbonic anhydrase (31,000), soybean trypsin inhibitor (21,500), and lysozyme (14,300).

Opioid Receptor Binding
Rat Brain Membranes-Neural membranes were prepared as previously described (19), and binding studies were conducted in a final volume of 200 pl. Each sample consisted of 100 pl of membrane preparation (20 pg of protein; Bio-Rad protein assay, Munich, Federal Republic of Germany) and 50 p1 of the tritiated tracer in Tris buffer (50 mM, pH 7.4). Unless otherwise stated, incubation was conducted at 25 "C for 30 min (maximal binding is observed within 20 min). Total binding was measured at 0.8-1.0 nM of the tritiated drugs, and nonspecific binding was obtained in the presence of 10 pM of the respective unlabeled substance. Thereafter, samples were submitted to Whatman GF/B glass fiber filters which were washed twice under reduced pressure with 4 ml of cold Tris buffer and then counted by liquid scintillation spectrometry. The binding studies for non-opioid ligands followed the same procedures. The total binding measured for the tritiated ligands was in the range of 0.4-0.7 pmol/O.l mg of protein. The specific binding of the compounds collectively amounted to 90%.
For 'Z51-labeled @-endorphin binding assays the method described by Toogood et al. (20) was slightly modified. Briefly, rat brain cortex (2 g) was homogenized in 100 ml of 20 mM Hepes buffer, pH 7.4, containing 3 mM MgSO,, and per ml, 100 pg of BSA and 50 pg of bacitracin. Binding was assayed with 2 pg of protein (cortex membranes) in a final volume of 100 pl. After incubation for 20 min at 25 "C, maximal binding was observed. The tubes were centrifuged for 4 min (12,000 X g ) , the supernatant was removed, and the pellet counted for radioactivity. The total binding amounted to 35 fmol, 10 pg of protein. The specific binding was 80-85%.
Effect of Anti-idiotypic Antibodies on the Binding of Opioids and Non-opioids-Antibodies were concentrated employing Amicon U1trafiltration tubes (Centricon, M, cutoff 30,000). To measure inhibition of opioid and non-opioid binding, the antibodies were incubated with the receptor material 15 min prior to exposure to the radiolabeled opioids (30 min). Reversihility of opioid binding by anti-Ids was tested by preincubation of the radiolabeled opioids with the receptor material (15 min) followed by an incubation (30 min) with the appropriate antibody concentrations. The subsequent procedure followed was that described in the previous section.

Immunoprecipitation
Rat brains (minus cerebellum) were prepared as described (21). The homogenate was vigorously shaken in the presence of 2% digitonin (30 min, 4 "C). Following ultracentrifugation (100,000 X g, 1 h, 4 "C) the clear supernatant was used for binding studies. Varying amounts of purified anti-Id-135 or anti-Id-14 were added to 700 pl of solubilized opioid receptor solution and incubated at 4 "C for 48 h. Subsequently, 300 pl of goat anti-mouse IgM, conjugated to Sepharose 4B, was added and the samples shaken at 37 "C for 2 h. After centrifugation at 12,000 X g for 2 min, the supernatant was removed and [3H]DIP binding to the supernatant was measured using the opioid receptor binding assay (see above). cAMPAssay in NG 108CC15 Hybrid Cells Neuroblastoma x glioma hybrid cells (NG 108CC15) were cultured in Dulbecco's modified Eagle's medium as described by Hamprecht (26). Cells were harvested at the state of confluency, and each assay consisted of 2 X lo4 cells. Cells were exposed to anti-Id-14 in Dulbec-CO'S modified Eagle's medium (37 "C, final volume 80 pl). After 1 h PGE, M) was added in the presence of a phosphodiesterase inhibitor (Ro 20-1724, 10" M), each in 10 pl of Dulbecco's modified Eagle's medium. The stable DADL was given to the incubation system 10 min prior to stimulation. The generation of CAMP was terminated by addition of 500 pl of 0.1 N HCl (95 "C), and measured by radioimmunoassay (25 Germany). 2,2'-azinodiethylbenzthiazoline sulfonic acid and fetal calf serum were obtained from Boehringer Mannheim; goat anti-mouse IgM, rabbit anti-IgG/Fc peroxidase conjugate, and rabbit anti-IgM/Fc peroxidase conjugate were from Nordic Immunological Laboratories (Tilburg, The Netherlands) and human @endorphin from Novabiochem (Laufelfingen, Switzerland). Naloxone was obtained from Du Pont de Nemours, and microtiter plates from Costar. The supernatants of stable cell lines that were positive in the solid-phase inhibition test were further screened in the opioid receptor assay to identify anti-idiotypic antibodies capable of interaction with the opioid receptor. The lZ5Ilabeled @-endorphin binding to rat brain membranes was inhibited to varying degrees by 10 of the concentrated hybridoma supernatants.

Fusion and Tests for the Interaction of
Two of the most ELISA positive IgM producing cell lines (1.55.153.135 (anti-Id-135) from Fusion I and 11.30.14 (anti-Id-14) from Fusion 11)) were selected for production either in culture (serum-free medium) or in nude mice (ascitic fluid) SO as to obtain sufficient material for purification and further characterization.
Purification of the Anti-idiotypic Antibodies-Attempts to purify both anti-Ids (of IgM class) by conventional gel chromatography (Sephacryl S-300) proved unsuccessful; the strong adsorption properties of the antibodies resulted in a poor yield. Including 0.05% Lubrol PX in the elution buffer and using a Mono Q anion exchange column, we obtained a good separation of the anti-Ids from protein contaminants originating in culture medium or ascitic fluid (Fig. 1).
Further analysis of the material eluting from the Mono Q column was performed using SDS-PAGE under reducing conditions (Fig. 2). IgM of the hybridoma line 1.55.153.135 was eluted by the 100% step of 0.3 M phosphate buffer with only low amounts of protein contaminations. Two main bands (M, 82,000 and 25,000) were separated as described for the heavy and light chain of IgM (22). Similar results were obtained for IgM of clone 11.30.14 (data not shown). The remainder of the study described here deals with the two fast protein liquid chromatography-purified antibodies.
Demonstration of Anti-idiotypic Actiuity-The anti-idiotypic activity of both antibodies was examined by testing their ability to inhibit the interaction of 3-E7 with its ligand @-endorphin. Fig. 3 shows that the binding of iodinated 3-E7 to @-endorphin (coated to a solid support) can be inhibited by increasing concentrations of both antibodies. The same degree of displacement was observed when the binding of lZ5I-labeled @-endorphin to immobilized F(ab'), fragments of 3-E7 was measured (data not shown). This demonstrates that both anti-Ids interact with the @-endorphin binding site of 3-E7. To obtain a 50% inhibition of binding, an antibody concentration of about 130 nM was necessary for both antibodies.
Interaction of Anti-Id-135 and Anti-Id-14 with the Binding of Opioid Ligand to Rat Brain Membranes-At the outset studies were conducted to examine the time course of interaction of the anti-Ids (50 nM) with opioid receptor binding a t brain membranes. The data revealed a rapid inhibition of lZ5Ilabeled @-endorphin binding, reaching maximal effects 20 min Proteins were eluted with a step gradient of 0.025 M phosphate buffer, pH 6.7, 0.05% Lubrol PX to 0.3 M phosphate, pH 6.5, 0.05% Lubrol PX (----). IgM was measured by the anti-Fab ELISA (0) (see "Experimental Procedures") and the inhibition of the binding of P-endorphin to rat brain membranes (0) determined by the receptor assay.  For detailed binding studies brain membranes were exposed to increasing concentrations of anti-Ids prior to incubation with a constant concentration of radiolabeled opioids. fashion. Half-maximal inhibitions were estimated to be 20 and 130 nM for anti-Id-14 and anti-Id-135, respectively; i.e. anti-Id-14 was 6 times more potent in competing for the binding site of the p-ligand as compared to anti-Id-135. More than 200 nM of each antibody was required to completely impede specific binding of the p-ligand. The ICso of DHM in the binding assay was 2 nM. Thus, the anti-Id-14 was (on a molar basis) 10-fold less effective than the alkaloid ligand. Anti-Id-14 was employed to further characterize the specificity of the interaction with opioid receptors (Fig. 5). The antibody did not affect or only moderately reduced the binding of K-receptor ligands (['HI-EKC and [3H]bremazocine) and of ["HIDIP, an opioid antagonist with high affinity for K-, 6-, and preceptors (23). In contrast, the antibody competed effectively with F -and &receptor ligands. ['H]Naloxone, a narcotic antagonist like DIP, displays a somewhat higher preference for preceptors, and its receptor binding appeared to be sensitive to the action of the antibody. The anti-Id-14 affected the binding of the opioid peptide '2sI-labeled P-endorphin somewhat less (ICso 50 nM) than compared to the alkaloid DHM.

Monoclonal Anti-idiotypic Antibodies to Opioid Receptors
The inhibition of binding activity was specific for opioid receptors: the binding of ["Hlflunitrazepam was not altered by anti-Id-14 even at the highest concentration employed, and similar results were obtained with ['H]scopolamine, a muscarinic antagonist (data for scopolamine not shown). Con- Each point represents the mean of at least two independent experiments. trol IgM MOPC 104E was completely inactive in the opioid binding assay and heat denaturation (95 "C, 5 min) totally abolished the receptor activity of both antibodies (data not shown).
Inhibition of opioid binding activity was observed regardless of whether the antibody was incubated with the membrane prior to, or following, the addition of radioligand, although the IC,, was 2-3 times higher in the latter case. Since the anti-Id may noncompetitively interact with opioids ( Fig. 6) this limited data available does not permit to draw any firm conclusion regarding the interpretation of this observed difference.
Mode of Interaction of Anti-Id with Opioid Receptors-To examine the opioid receptor interaction of the anti-Id-14, rat brain membranes were exposed to 50 nM of the antibody, a concentration known to cause 60-70% inhibition of P-endorphin. This was followed by increasing concentrations of '*,Ilabeled P-endorphin. As demonstrated in the Scatchard analysis in Fig. 6, the interference of anti-Id-14 with /'+endorphin binding occurred in a noncompetitive fashion. This is indi-cated by a decline of the Bmax value (total number of receptors) with an apparently unchanged affinity of the receptor for the ligand. Similar experiments conducted with [3H]DHM also revealed a noncompetitive interaction between the antibody and the opiate (data not given).
Immunoprecipitation of Opioid Receptors-The above demonstration of an interaction of the anti-idiotypic antibodies described here with the binding of opioid ligands to opioid receptors prompted attempts to precipitate these solubilized molecules. Fig. 7 shows the ability of both anti-Id-135 and anti-Id-14 to precipitate [3H]DIP binding sites from a solubilized rat brain membrane preparation. Anti-Id-135, which was less potent in competing for opioid binding in brain membrane preparations proved slightly less efficient than anti-Id-14 to immunoprecipitate soluble opioid receptors. No significant precipitation was obtained with control IgM (MOPC 104E).
Investigations into the Function of Anti-Id-14-To study whether the anti-Id behave like an opioid agonist or antagonist, the ability to interfere with cAMP production in NG 108CC15 cells was tested. Table I

DISCUSSION
It has been suggested that an anti-ligand antibody, suitable for the generation of anti-receptor anti-idiotypic antibodies, should share with the receptor a high degree of similarity in binding properties (3). This requirement is partially fulfilled by the monoclonal anti-@-endorphin antibody 3-E7; the antibody recognizes virtually all known endogenous opioid peptides, but does not bind any of the acetylated, iodinated, or N-terminal shortened peptides. Furthermore, opioids with alkaloid structure or opioid peptides bearing D-amino acids in position 2 are not recognized by the 3-E7 antibody (1). In principle, as shown in our previous study with polyclonal antibodies (4,5 ) and confirmed in the present study, the 3-E7 is useful in generating both xenogeneic and syngeneic antiidiotypic antibodies.
The reason for using a murine monoclonal antibody for the immunization of mice resides in the fact that no anti-isotypic or anti-allotypic antibodies were to be expected. The difficulties inherent in a syngeneic system (i.e. screening for anti" antibodies from the same species as the immunogen) was circumvented by the use of the Fab fragment of 3-E7 and measuring the binding of the anti-Id antibodies to solid phaseadsorbed Fab fragments with the aid of Fc-specific antibodies. The screening for anti-opioid receptor antibodies revealed that both fusions carried out here resulted in antibodies of the IgM class (molecular weight 980,000) which showed high adsorption as well as a relatively low solubility.
The following observation provides some indication as to the specificity of the interaction of the two anti-idiotypic antibodies described in this paper with opioid receptors: (i) the inhibitory activity of the two antibodies was strictly confined to opioid receptors since they did not show interference with the binding of benzodiazepine or muscarinic receptor ligands; (ii) the opioid receptor-binding activity of both anti-idiotypic antibodies was shown to depend upon their tertiary structure since heat denaturation completely abolished binding.
An interesting point of this study is that both antibodies displayed a broad spectrum of different IC,, values in inhibiting opioid ligand binding to p-or &opioid receptor types. It demonstrates the ability of the anti-Ids to discriminate between different opioid receptor types. Both p-and &ligands are inhibited with similar potencies, whereas the binding of [3H]EKC or [3H]bremazocine, which have preferential affinity for the K-receptor, is poorly affected by both antibodies. This signalizes that the receptor preference of the two antibodies is similar to that exhibited by enkephalin, an opioid peptide with affinity for p-and &receptors. Such a result was expected on the basis of the concept that the combining site of both antibodies represents the internal image of the N terminus of P-endorphin (identical to Met-enkephalin), which was employed to generate the idiotypic antibody 3-E7. In fact, although 3-E7 does not interact with non-peptide opioids, it recognizes the N-terminal pentapeptide sequence of opioids and, therefore, the resulting anti-idiotypic antibodies may be expected to carry the internal image of this sequence. Consequently, both of the anti-Ids are able to compete with opioid peptides as well as with alkaloids for the binding to the receptors. This finding is in line with Jernes (24) postulation that an anti-Id may reflect the internal image of the ligand.
A similar preference of binding to both p-and &opioid receptors was recently described by Glasel and Myers (7) for a polyclonal anti-opioid receptor antibody generated from an anti-morphine antibody. Since the idiotype was specific for the p-ligand morphine, these authors expected anti-idiotypic receptor antibodies to be directed to the p-receptor. However, they obtained antibodies which interacted not only with Fbut also with &receptors. K-Receptors were not affected by this antibody. Similarly, the anti-opioid receptor antibody described by Bidlack and Denton (8) does interfere with pand &receptors but fails to exhibit activity on K-receptors.
The ability of our anti-idiotypic antibodies to interact with solubilized receptors is a further indication of their specificity. This is confirmed by the fact that the two IgMs can be used to quantitatively immunoprecipitate all the opioid-binding activity in solution. In view of the high molecular weight of the IgMs some constraints of their accessibility to membranebound receptor proteins may be expected. Nevertheless, both of the antibodies were able to completely inhibit the binding of [3H]DHM to rat brain membranes in a way similar to that previously found by us for polyclonal anti-idiotypic antibodies (5); this is in contrast to findings with anti-opioid receptor antibodies generated by others (6)(7)(8), and suggests that the receptor binding sites are truly occupied.
The anti-Id failed to display opioid agonistic activity as is indicated by its inability to reduce adenylate cyclase activity in NG 108-15 hybrid cells. This contrasts the function of a polyclonal anti-Id opioid receptor antibody (IgG) in the identical test system ( 5 ) . On the other hand, the anti-Id-14 exerts opioid antagonistic activity as is indicated by its ability to reduce the action of DADL on CAMP production in NG cells.
Although antagonistic activities have been reported for antibodies to non-opioid receptors (27), information regarding anti-opioid receptor antibodies have been missing so far. Although the antagonistic activity demonstrated here was incomplete, the binding data given (Fig. 5) would suggest a complete blockade of the action of DADL in the NG hybrid cells.