Studies on nicotinic acetylcholine receptors in mammalian brain. Interaction of solubilized protein with cholinergic ligands.

Binding of alpha-bungarotoxin, labeled with 125I, has been studied in detergent extracts and affinity purified acetylcholine receptor from rat cerebral cortex. Binding to detergent extracts is saturable and appears to be due to one class of binding sites present at a level of 0.27 pmol/mg of protein. The association constant is 2 X 10(7) liters mol-1 . Competition with cholinergic ligands indicates that toxin binding to both detergent solubilized and affinity purified receptor retains its nicotinic nature. Values for the ligand concentrations required to produce 50% inhibition of extent and rate of toxin binding are presented.

In recent years, rapid progress has been made in elucidating the nature of the nicotinic acetylcholine receptor' from a wide range of sources (see Ref. 1 for a collection of pertinent papers). This advance has been made possible by the introduction by Changeux (2) of the use of a-neurotoxins isolated from the venom of various elapid snakes. These neurotoxins bind specifically and with high affinity to the nAChR (3, 4). They, especially a-bungarotoxin, obtained from Bungurus multicinctus, the Formosan krait, and cobroxin, from various Naja species, have been used extensively to identify and quantitate the nAChR (5-8).
The techniques for the assay, solubilization, and purification of the nAChR were developed for the electric organ of electric eels and other fish, a tissue that is highly enriched in the nAChR. These techniques were then successfully adapted to study the nAChR of the neuromuscular junction in the intact tissue and after its removal and culture (1). Study of the nAChR in cerebral cortex has proven to be much more difficult because the predominant form of the AChR is muscarinic and not nicotinic (9). Farrow and O'Brien (10) found a high affinity binding site for atropine, a muscarinic ligand, at a concentration of 89 pmol/g of tissue in rat cerebral cortex; Yamamura and Snyder (II), using 3-quinuclidinyl benzylate, a potent central muscarinic antagonist, found a density of muscarinic sites of 65 pmol/g of rat brain. In 1972, Moore and Loy (12) demonstrated the presence of an a-Btx binding component in hog brain. The following year Salvaterra and Moore (13) determined the number of a-Btx binding sites as 3.4 pmol/g of rat cerebral cortex. These values were confirmed and extended to a number of brain regions and subcellular fractions in a study by Salvaterra et al. (14). These levels are very low as compared to the muscarinic AChR in the same preparation, to the 1 to 2 nmol/g of tissue in Electrophorus electricus (81, or to 0.5 to 1 nmol/g of tissue in Torpedo californica (5). Eterovic and Bennett (15) reported on the binding of tritiated a-Btx to crude mitochondrial fractions from rat cerebral cortex and the inhibition of binding by a variety of cholinergic drugs. Moore and Brady (16) and McQuarrie et al. (17) studied the binding of a-Btx, labeled with lZ51, to crude membrane preparations of rat cerebral cortex. The latter study, as well as a recent one by Schmidt (18), reported on the inhibition of a-Btx binding to such preparations by a number of cholinergic and noncholinergic drugs. Salvaterra and Mahler (191,Lowy et al. (20), and Moore and Brady (16) have described the solubilization of the nAChR from rat cerebral cortex by detergents, i.e. 1% Triton X-100 and 1% Emulphogene; however, these studies do not deal with the inhibition of toxin binding by cholinergic drugs. This paper reports on some lZ51-a-Btx binding properties of detergent extracts of a crude membrane preparation from rat cerebral cortex. We also report on the inhibition of rate of lZ51a-Btx binding by some cholinergic drugs to the solubilized AChR purified according to Salvaterra and Mahler (19 and Naja naja siamensis venom were purchased as lyophilized powders from the Miami Serpentarium. iz51 was purchased as carrier-free NalZ51 in 0.1 N NaOH from New England Nuclear.
All other chemicals and reagents were of reagent grade or hotter and purchased from commercial sources.

RESULTS
A binding isotherm for 'Y-a-Btx to the Triton X-100 extract containing the AChR receptor in solubilized form is presented in Fig. 1  crude membrane preparations. The apparent association constant K',, as calculated from the concentration at which 50% saturation occurs, is found to be 2 x lo7 1 mall'. In Fig. 2 are presented the extent of inhibition at equilibrium of cu-Btx binding in Triton extracts by several acetylcholine agonists and antagonists. The nicotinic antagonist, dtubocurarine, the acetylcholine agonist, carbamylcholine, and acetylcholine itself are all more effective inhibitors of lzSI-(u-Btx binding than are the muscarinic ligands, atropine and oxotremorine. In Fig. 3 we present the inhibition of the rate of binding of lz51-oc-Btx in similar Triton X-100 extracts by a series of cholinergic drugs. The most effective inhibitor is d-tubocurarine, followed closely by carbamylcholine and acetylcholine. The nicotinic antagonist, gallamine (flaxedil), and the muscarinic antagonist, atropine, are much less effective, while hexamethonium and the muscarinic agonist, scopolamine, are essentially without effect. Unlabeled a-Btx inhibits the rate of binding of 50% at a concentration somewhat less than 2.4 X It9 M.
The data of Fig. 4    d-Tubocurarine 10-6 5 x 10-1 1.9 x 10-e 10-S 2 x 10-S Acetylcholine 3 x 10-S Carbamylcholine lo-4 9 x 10-S 10-S 10-S Gallamine 3.5 x 10-G Hexamethonium 9 x 10-d Atropine 10-S 8 x 10-a 1.6 x 10m3 10-a 9 x 10-a Oxotremorine 2.0 x 10-a Protein concentration (mglml) Labeled a-Btx concentration CM) 3-4.5 5.8 1 1 6.2 x 10m9 5 x 10-g 7 x 10-10 3 x 10-S 3 x 10-S methonium, which is an effective antagonist on both ganglion cells and the electroplax, is without effect on cerebral cortex or on muscle end plates (35). Very recently, Moore and Brady (36) reported on the solubilization of the AChR from rat cerebral cortex by the detergent Emulphogene BC 720 and on some of the binding characteristics of these detergent extracts. The inhibition of the extent of iz51-a-Btx binding by d-tubocurarine and hexamethonium was very similar to that reported here, while carbamylcholine was lo-fold less effective. The relative effectiveness of the ligands is altered upon solubilization of the AChR. The ability of gallamine, a nicotinic acetylcholine antagonist, to inhibit a-Btx binding is reduced 2 orders of magnitude upon solubilization and purification of AChR, while the effectiveness of carbamylcholine is enhanced by a factor of 10 under these conditions. Upon solubilization of the AChR from the neuromuscular junction, Dolly and Barnard (37) found a slight increase in the effectiveness of inhibition of the rate of binding of ["HIa-Btx by various cholinergic agents, except for carbamylcholine, which exhibited a lower affinity for the solubilized receptor. Colquhoun and Rang (38) found a slight, uniform increase in affinity for all ligands tested under equilibrium conditions. Schmidt and Raftery (25) found little change in the effectiveness of the various cholinergic drugs upon solubilization of the AChR from the electric organ of E. electricus.
When studying competitive inhibition of binding of an essentially irreversible ligand such as a-Btx, the inhibition of the rate of its association by the competitive ligands provides a much more sensitive measure of their relative affinities than do measurements of displacement close to equilibrium (5). The data presented in Table I clearly demonstrate this phenomenon. The concentration at which 50% inhibition occurs with tubocurarine, acetylcholine, and carbamylcholine is much lower when determined by measuring their effect on rate rather than the extent of iZ51-a-Btx binding to the solubilized receptor. After its purification similar effects are seen for tubocurarine and acetylcholine, while carbamylcholine appears somewhat less effective in reducing the rate of toxin binding than it is under equilibrium conditions. Similar results were found by Changeux and co-workers (28, 39) on the AChR from E. electricus and by Raftery and co-workers (25, 40) on the AChR from Torpedo californica.
After extensive purification by affinity chromatography the solubilized AChR retains its affinities for all cholinergic ligands with the exception of carbamylcholine, the affinity of which reverts to a value resembling that found with the membrane-bound AChR. Dolly and Barnard (37) reported no 3.3 x 10m7 3.9 x 10m7 5.6 x lo-' 1 x 10-1 Acetylcholine 1 x 10-G 7.7 x 1O-6 2.5 x 1O-6 Carbamylchohne 0.9 x 10m5 1.9 x 1O-6 1.6 x lOma 4.5 x 10m5 Gallamine 1.3 x 10-T 2 x 10-1 Hexamethonium 6.2 x 1O--5 3.3 x 10m5 2 x 10-6 Atropine change between solubilized and affinity purified AChR from muscle, although they raised the a-Btx concentration lo-fold to determine the rate of binding to the purified receptor. Comparable studies on the AChR from the electric organ have not been reported in the literature.
In Table III are presented data taken from the literature for the values of the concentration of cholinergic ligands required for 50% inhibition of the rate of binding of labeled oc-neurotoxin to purified AChR. From a comparison of these values with the corresponding ones summarized in Table I, we conclude that the AChR found in cerebral cortex is similar but not identical with that found in the electric organ of electric fish.
Maelicke et al. (34) have recently also completed an exhaustive analysis of the binding of Naja naja siamensis o-neurotoxin to AChR from the electric organ of E. electricus after its purification by affinity chromatography and the inhibition of that binding by various cholinergic ligands. The use of an aneurotoxin with a reversible, instead of an essentially irreversible, binding mode, such as oc-Btx, enabled the authors to develop relationships between the amount of labeled toxin bound to the AChR, and Ki, the inhibitor dissociation constant of the cholinergic ligand, determined under equilibrium conditions. These relationships were derived from basic thermodynamic arguments with a minimum of assumptions; Ki values calculated from these expressions represent the most accurate determinations of these values since they are based on firmer theoretical grounds than assuming Ki to be equal to the concentration at which 50% inhibition occurs. The values reported by Maelicke et al. (34) were generally at least 10 times lower than previously published values, obtained by a variety of techniques. Although it is clear that this method for the determination of inhibitor dissociation constants is the most accurate, it is not readily applicable to brain AChR due to the much lower concentration of AChR in this tissue as compared to that found in the electric organ. The preparation of suffkient purified AChR to carry out the evaluation of Ki values according to that method would represent an enormous effort in terms of time and cost. The data presented in this report are more or less self-consistent and do allow one to detect changes in binding properties during the various stages of purification. Since they were determined under conditions similar to many of those previously reported for AChR derived from a variety of tissues, they should still prove useful for comparison between the AChR in brain and other vertebrate tissues.

Achnowledgment
-We express our appreciation to Ms. Rene Foders for excellent technical assistance.