Proc. Nat. Acad. Sci. USA Vol. 68, No. 5, pp.

1047-1050, May 1971

Binding Sites for Cholinergic Ligands in a Particulate Fraction of Electrophorus Electroplax

M. E. ELDEFRAWI, A. T. ELDEFRAWI, AND R. D. O'BRIEN
Section of Neurobiology and Behavior, Cornell University, Ithaca, N.Y. 14850

Communicated by Thomas Eisner, February 22, 1971

ABSTRACT The binding of four radioactive cholinergic ligands to a particulate fraction of Electrophorus electricus electroplax was investigated by equilibrium dialysis and found to be reversible. A single site (concentration, 0.020.03 nmol/g electroplax) that bound muscarone and nicotine was found to be on a phospholipoprotein; it is suggested that this is the acetylcholine receptor. Four binding sites for decamethonium were found; the highestaffinity site (0.03 nmol/g) may be the same one that binds muscarone and nicotine, both because of the chemical nature of the binding macromolecule and because decamethonium competes with muscarone and nicotine in binding. Dimethylcurare bound to three sites. The loweraffinity sites appeared to be on different macromolecules. When binding of the four cholinergic ligands (at 0.01 AM) to 28 arbitrarily selected proteins was determined, it was concluded that muscarone would be the most specific for the acetylcholine receptor, in that it bound to none of the 28 proteins tested; nicotine bound to one, dimethylcurare to two, and decamethonium to ten of the proteins. Nicotine or muscarone appear to be the ligands of choice in studying binding to the acetylcholine receptor; decamethonium appears to have inadequate specificity.

A major problem in the attempted isolation of acetylcholine (ACh) receptor macromolecules is how to identify them in vitro. Acting on the hypothesis that in vitro, ACh receptor retains its affinity for cholinergic ligands and binds them reversibly and competitively, we searched for such macromolecules in different excitable tissues. We found them in a particulate fraction of Torpedo electroplax and a 100,000 X g, 1 hr supernatant of housefly heads (1-5). They bound muscarone, nicotine, decamethonium, and dimethyl-dtubocurarine (dimethylcurare); in the housefly head and rat brain, atropine was bound also (3, 6). When binding was measured by equilibrium dialysis over a restricted range of ligand concentration (0.1-1 ,AM), only a single binding site was detected for each ligand in the above cases; but when the concentration was extended (0.001-100 AM), several sites were revealed in Torpedo electroplax (7). On the basis of concentrations of binding sites, reversibility of binding, and competition by other cholinergic ligands, it was suggested that either one or more of the sites formed part of the acetylcholine receptor macromolecule. The binding macromolecules were phospholipoproteins of nicotinic nature in Torpedo, and proteins of nicotinic and muscarinic nature in houseflies (3, 5). Further supporting evidence that one or more of the sites was on the ACh receptor was that in the housefly extract, toxic but not nontoxic nicotine analogs were found to block binding of muscarone and nicotine (4). Also, in electroplaxes,

there was excellent correlation between the high concentrations of organophosphates needed to reversibly block the response of the ACh receptor in vivo in Electrophorus (8), and to block the in vitro binding of muscarone and acetylcholine to Torpedo (9). With the majority of our extensive binding data coming from Torpedo electroplax (1, 2, 4, 5, 7, 9), but the in vivo information coming from Electrophorus electroplax, we found it important to do in vitro binding studies on the latter. Direct comparisons would thus be feasible between biochemical and physiological studies in the same species, as well as between binding studies in electroplaxes of the two species. Comparison of such results, obtained over a wide range of ligand concentration, with several ligands and in the absence of a detergent, could also be made with those found for decamethonium on the same tissue (10, 11). The presence of several binding sites for each ligand should also be sought, and those with properties closest to those of the ACh receptor determined. Such a study should increase our knowledge of the characteristics of ACh receptor macromolecules in vitro and help in their final identification before attempts at purification are made.
METHODS A 5-kg, 155-cm long eel (Electrophorus electricus) yielded 2 kg of electroplax tissue from the three organs. The tissue was cut into 30-g pieces and stored at - 16C. Each piece was homogenized in 100 ml of distilled water (2C) in a Sorvall Omnimixer, and the homogenate was centrifuged for 20 min at 45,000 X g. The supernate was discarded when no significant binding of cholinergic ligands to it could be detected. The pellet was rehomogenized in 20 ml of water for 1 min, sonicated for 2 min in a Branson Sonifier at half maximum capacity, then filtered through cheesecloth. The volume was adjusted to 30 ml to give a final concentration of 1 g of electroplax per ml. Protein analysis (12) showed that this preparation contained 10 mg of protein per ml. The sources and specific activities of [3H ]muscarone,

Abbreviation: ACh, acetylcholine.

[3Hjnicotine, [3H]decamethonium, and [14C]dimethylcurare have previously been reported (3, 5). Equilibrium dialysis (for 16 hr at 4C) was performed for each 1 ml of electroplax preparation in 100 volumes of modified Krebs-Ringer solution (3) (pH 7.4, ionic strength 0.2), containing the radioactive ligand. After dialysis, the difference between radioactivity in samples of electroplax over equal volumes of solution, represented amount of ligand bound. Details of the procedure, sampling and counting, sources and treatment with enzymes, reversibility measurements, and statistical analysis have all been previously described (5, 7).

1047

1048

Biochemistry: Eldefrawi et al.

Proc. Nat. Acad. Sci. USA 68 (1971)

a
-J

b
N

-J

x
C.I)
0

x
(V)
0

.1

-J

x
CI)
0

x
C.)

.5

.5 B

gram

FIG. 1. Scatchard plots of the binding of ligands to a particulate preparation of eel electroplex (B, amount bound in nanomoles per of fresh tissue; L, molar concentration of ligand). a, muscarone; b, nicotine; c, dimethylcurare; d, decamethonium. The insert of id shows data at low concentrations of ligand. Each point represents an average of two experiments, three samples each. RESULTS

In contrast to the total homogenate of Torpedo electroplax, that of the eel did not significantly bind muscarone at 1 1AM when the tissue concentration was 25% (w/v). No binding was observed in the 100,000 X g supernate and very little binding was associated with the pellet. However, when tissue concentration in the reconstituted pellet was increased to 1 g of original electroplax per ml (10 mg of protein per ml) significant binding was observed. Binding was then studied over a wide ligand concentration (0.001-100 AM); muscarone and nicotine were found to bind significantly only below 1 MM, but decamethonium and dimethylcurare were bound over the entire ligand concentration studied. Scatchard plots (Fig. 1) and the iterative computer analysis of the binding data (Table 1) showed that muscarone and nicotine bound to single sites, but several binding sites were revealed for decamethonium and dimethylcurare. The Scatchard curve for dimethylcurare (Fig. ic) had an early rising portion at low ligand concentration, similar to that exhibited in positive cooperativity (13). The concentrations of binding sites for muscarone and nicotine and of the first two sites for decamethonium (B,, B2) and the second binding site (B2) for dimethylcurare were similar (Table 1). Reversibility of binding of the four ligands was studied by a second dialysis of the particulate fraction against 100 volumes of ligand-free Krebs-Ringer solution. The amount bound

at 0.1 uM of ligand was reduced to within 10% of that bound at 0.001 pM of ligand as calculated from Lineweaver-Burk plots. This indicated that binding (at the 0.1 AM concentration) was totally reversible.

TABLE 1. Binding constants and concentration of binding sites (B) in a particulate preparation of Electrophorus electroplax for four cholinergic ligands
B

Ligand Muscarone Nicotine Decamethonium

Dimethylcuraret

Constant, K (M) 5.5 X 10-8 6.3 X 10-8 K, = 2.5 X 10-9 K2 =5.5 X 10-8 K3= 2.5 X 10-6 K4= 1 X 10-4 K2 = 8 X 10-8 K3= 4 X 10-5

(nmol/g tissue)
0.021 0.033 B, = 0.03 B2 = 0.02 B3 = 0.25 B4= 4.0

% error*
11.8

7.1

17.0
6.5

B2 = 0.05 B3 = 14.0

Data derived from iterative treatment of Scatchard plot

percent error. t For dimethylcurare, it is assumed that there exists also a K, site, which cooperates positively with one other site, and whose K, cannot be estimated (see Fig. ic and text).

(Fig. 1). * Root-mean-square

Proc. Nat. Acad. Sci. USA 68

(1971)
x
W
0

Binding of Ligands to Electroplax

1049

TABLE 2. Effect of enzyme treatment on binding

% Reduction in binding PhosphoChymolipase Trypsin C trypsin
71 63 58 52 40 45 21

.02L
'o

Ligand (pM) Muscarone (0.01) Nicotine (0.01) Decamethonium (0. 001) (0.01) (0. 1) (1.0) Dimethylcurare (0.01) (. 1)

(6)*

71 62 38 44 37 43 42 36

50 23 25

0 0)

N.

.01

00)*
(10)* (-14)*t ( 2)* (-13)*t

_0

.0
E
c

Controls showed significant binding of decamethonium to the three enzymes, and of dimethylcurare to trypsin and phospholipase C. Values in the table were corrected accordingly. * Values in parentheses are not significant as judged by the t test. t Negative values indicate greater binding after enzyme treatment.

.05 .5 .1 106 x CLI FIG. 2. Binding to a particulate preparation of eel electroplax of nicotine alone (0), and in the presence of 0.01 ,&M curare
(A),
or 0.005j

t'.01

I.0

M decamethonium (0). L, nicotine concentration

(AM).
To determine the nature of the binding macromolecules, and whether the binding sites are on the same or different molecules, we investigated whether the binding is sensitive to the action of several hydrolases. Binding of muscarone, nicotine, and decamethonium at 0.001 ,&M was reduced by treatment with trypsin, chymotrypsin, and phospholipase C (Table 2), which suggests that these binding sites are on phospholipoproteins. However, binding of decamethonium at 0.01, 0.1, or 1 ;&M or dimethylcurare at 0.01 AM was affected only by trypsin and chymotrypsin. Binding of dimethylcurare at 0.1 1M was reduced by treatment only with chymotrypsin. If the different cholinergic ligands studied bind to the same site, then they should compete with one another. Indeed, binding of nicotine was reduced by the presence of decamethonium (0.005 uM) or d-tubocurarine (curare) (0.01 ;&M)
(Fig. 2), and binding of muscarone was reduced in the presence of nicotine, curare, or decamethonium. Because the previous unsuccessful attempts to isolate ACh receptor macromolecules indicated that cholinergic ligands bind to macromolecules not related to the receptor (14), and especially in view of the many binding sites revealed for decamethonium and dimethylcurare in this eel preparation, it was important to obtain a general idea of the specificity of these ligands. Several proteins were selected arbitrarily and binding was determined (Table 3) at the low ligand concentration of 0.01 /AM, so as to detect only very specific binding. Surprisingly, muscarone proved to be the most specific for ACh receptor-it did not bind to any of the proteins listed in Table 3; nicotine bound only to butyrylcholinesterase; dimethylcurare bound to butyrylcholinesterase and

TABLE 3. Binding of radioactive ligands (at 0.01

AIM) to proteins
Dimethylcurare
1.7 1.0 0 0 0
0 0 0 0 0

Proteins
Enzymes

Muscarone
0 0 0 0 0

Ligand (nmol bound/g protein) Nicotine Decamethonium

1.7 0 0 0 0
0 0 0 0 0

Butyrylcholinesterase (EC 3.1.1.8) PhospholipaseC (EC3.1.4.3) (EC 3.1.4.4) Phospholipase D (EC 3.4.4.4) Trypsin (EC 3.4.4.5) Chymotrypsin Nonenzymic proteins Casein a
Albumin

2.6 1.8 1.8 1.4 1.4

1.5 1. 5 1.1 1.9 1.4

Globulins a
RNA Acetone powder of pig kidney

0 0 0 0 0

No significant (P < 0.01) binding of the four ligands was found for ribonuclease (EC 2.7.7.16), lipase (EC 3.1.1.3), acetylcholinesterase (EC 3.1.1.7), deoxyribonuclease (EC 3.1.4.5), hyaluronidase (EC 3.2.1.e), lysozyme (EC 3.2.1.17), peptidase (EC 3.4.1.1), pepsin (EC 3.4.4.1), papain (EC 3.4.4.10), collagenase (EC 3.4.4.19), insulin, globulinsft, globulins y, ,6-lipoproteins, mucin, DNA, or acetone powders of pig liver or pig brain.

1050

Biochemistry: Eldefrawi et al.

Proc. Nat. Acad. Sci. USA 68

(1971)

to phospholipase C; and decamethonium was quite nonspecific, binding to 10 of the 28 proteins tested.
DISCUSSION

The present study demonstrates the presence, in this particulate preparation of eel electroplax, of macromolecules that reversibly bind four cholinergic ligands with high affinity. The similarity in the binding (Table 1) of the two relatively specific agonists, muscarone and nicotine, is remarkable. They reversibly bind to single sites of similar concentration and compete in binding. We suggest that binding macromolecules, which are phospholipoproteins in both cases (Table 2), are ACh receptors with a common binding site for the two agonists. Similarities in the binding of muscarone and nicotine were also found in a particulate preparation of Torpedo electroplax, but there were two sites for binding each of these ligands (7); the evidence strongly suggested that both sites are on the ACh receptor, which perhaps exists in two conformations with different affinities to ligands, as has been suggested from physiological studies (15). We cannot exclude the possibility that the ACh receptor in our eel preparation has another binding site, with different affinity for nicotine and muscarone, which we were unable to detect either because of large differences between the two affinities or because of the predominance of one conformation. Equilibrium dialysis has its limitations, a lower one determined by the radioactivity that can be detected, and an upper one where the radioactivity of the amount bound is so small so as to be masked by the standard deviation of that of the bath samples. The similarity of the highest affinity site (B1) of decamethonium to that of muscarone and nicotine, in its concentration (Table 1) and sensitivity to hydrolases (Table 2), as well as the competition of 0.005 ,.M decamethonium with nicotine binding (Fig. 2), leads to the suggestion that there may be one site on the ACh receptor that binds these three ligands. By virtue of their insensitivity to phospholipase C (Table 2), the three other sites that bind decamethonium and the two that bind dimethylcurare seem to be on different molecules. Such a possibility is likely, especially in view of their low specificity (Table 3). Yet, as found in Torpedo electroplax (7), a paradoxical situation is presented by the antagonist dimethylcurare, which competes with nicotine and muscarone for binding. An explanation is that competition may be allosteric between two topographically distinct sites on the macromolecule, and the highest affinity sites that bind dimethylcurare may be on the protein portion of the ACh receptor. For that matter, the second site (B2) that binds decamethonium, which has similar concentration and sensitivity to enzyme treatment, may also be binding at this allosteric site. Physiological data have led to many suggestions (16) that the ACh receptor carries more than one binding site for cholinergic ligands. The concentration of high-affinity sites is very similar for the three agonists (0.02-0.03 nmol/g of electroplax), and similar to the corresponding concentration (7) for Torpedo electroplax (0.06-0.08 nmol/g of electroplax). For Electrophorus, these values correspond to 2-3 nmol/g of protein; they are compatible with the value of 26 nmol/g of protein of a partially purified membrane preparation from Electrophorus, as estimated by its ability to bind a-bungarotoxin (11). Changeux et al. (10) calculated dissociation constants for six ligands; only decamethonium was used both in our study

and theirs. Their dissociation constant for "receptor in solution" (i.e., after treatment of "microsacs" with deoxycholate and using the supernatant after centrifugation at only 28,000 X g) is 0.9 AM for decamethonium. This is closest to our K3 value for decamethonium (2.5 ,uM), and presumably this is the site which these authors were studying. It appears from our data that this site does not bind muscarone or nicotine. As we found with Torpedo electroplax (7), our constants for the in vitro dissociation of cholinergic ligands from what we believe to be the ACh receptor are much lower than physiologically determined values (10, 17). Differences in the same direction and of similar magnitude have been found between in vivo and in vitro dissociation constants for Electrophorus acetylcholinesterase (18). An interesting question is whether disruption of the membrane structure favors conformations of the ACh receptor that have higher affinities for ligands. Our data suggest that for Electrophorus electroplax, muscarone or nicotine are the preferred ligands with which to study the ACh receptor as purification proceeds. Decamethonium is the least desirable ligand for this purpose.
We are grateful to Prof. H. Howland for providing a linear computer program for the Scatchard plot, and to Miss L. P. Gilmour for technical assistance. Financial support is gratefully acknowledged from U.S. Public Health Service research grants NS 09144 and GM 07804 and Training Grant ES 98.
1. O'Brien, R. D., and L. P. Gilmour, Proc. Nat. Acad. Sci. USA, 63, 496 (1969). 2. O'Brien, R. D., L. P. Gilmour, and M. E. Eldefrawi, Proc. Nat. Acad. Sci. USA, 65, 438 (1970). 3. Eldefrawi, A. T., and R. D. O'Brien, J. Neurochem., 17, 1287 (1970). 4. Eldefrawi, M. E., A. T. Eldefrawi, and R. D. O'Brien, J. Agr. Food Chem., 18, 1113 (1970). 5. Eldefrawi, M. E., A. T. Eldefrawi, and R. D. O'Brien, Mol. Pharmacol., 7, 104 (1971). 6. O'Brien, R. D., M. E. Eldefrawi, A. T. Eldefrawi, and J. T. Farrow, in Cholinergic Ligand Interactions, ed. D. J.Triggle, E. A. Barnard, and J. F. Moran (Academic Press, N.Y., (1971), p. 49. 7. Eldefrawi, M. E., A. T. Eldefrawi, L. P. Gilmour, and R. D. O'Brien, Mol. Pharmacol. (submitted). 8. Bartels, E., and D. Nachmansohn, Arch. Biochem. Biophys., 133, 1 (1969). 9. Eldefrawi, M. E., A. G. Britten, and R. D. O'Brien, Pestic. Biochem. Physiol., 1, in press. 10. Changeux, J. P., M. Kasai, M. Huchet, and J. C. Meunier, C. R. Acad. Sci., Paris, 270, 2864D (1970). 11. Changeux, J. P., M. Kasai, and C. Y. Lee, Proc. Nat. Acad. Sci. USA, 67, 1241 (1970). 12. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem., 193, 265 (1951). 13. Cook, R. A., and D. E. Koshland, Biochemistry, 9, 3337

(1970).
14. Ehrenpreis, S., Nature, 201, 887 (1964); Chagas, C., Ann. N.Y. Acad. Sci., 81, 345 (1959); Trams, E. G., Biochim. Biophys. Acta, 79, 521 (1964). 15. Rang, H. P., and J. M. Ritter, Mol. Pharmacol., 6, 357 (1970); Katz, B., and S. Thesleff, J. Physiol., 138, 63 (1957). 16. Clark, A. J., J. Physiol., 61, 123 (1926); Ariens, E. J., and A. 1M. Simonis, Ann. N.Y. Acad. Sci., 144, 842 (1967); Flacke, W., and T. S. Yeoh, Brit. J. Pharmacol. Chemother., 33, 154 (1968); Rang, H. P., and J. M. Ritter, Mol. Pharmacol., 5, 394 (1969); Podleski, T., J. C. Meunier, and J. P. Changeux, Proc. Nat. Acad. Sci. USA, 63, 1239 (1969). 17. Higman, H. B., T. Podleski, and E. Bartels, Biochim. Bio-

phys. Acta, 75, 187 (1963). 18. Webb, G. D., and R. L. Johnson, Biochem. Pharmacol., 18, 2153 (1969).