Eur. J. Biochem. 55, 505-515 (1975)

Interconversion between Different States of Affinity for Acetylcholine of the Cholinergic Receptor Protein from Torpedo marmorata Hiroyuki SUGIYAMA and Jean-Pierre CHANGEUX Neurobiologie Moleculaire, Institut Pasteur, Paris (Received November 26, 1974/February 11, 1975)

In receptor-rich membrane fragments from Torpedo, acetylcholine binds, in the presence of 70 pM Tetram, to a homogeneous population of high-affinity sites with Kd = (3.4 k 0.8) x M. Dissolution of these membrane fragments by sodium cholate causes a decrease of affinity associated with the appearance of medium-affinity (Kd FZ lo-’ M) and low-affinity (Kd 2 lop6M) sites. Dissolution by neutral detergents Triton X-100 or Emulphogene preserves the high affinity of the acetylcholine binding sites. In all the soluble states of the receptor protein, Ca2+ ions and local anaesthetics no longer enhance the affinity for acetylcholine. Elimination of sodium cholate by dilution leads to the reassociation of the receptor protein, the recovery of high-affinity sites and the control by Ca2 ions and local anaesthetics. Purification by affinity chromatography of the receptor protein in Triton X-100 is accompanied by a conversion of a majority of the acetylcholine sites into their state of low affinity. High-affinity sites can no longer be recovered by detergent dilution from these low-affinity ones. +

The cholinergic receptor protein from both Electrophorus [ l - 61 and Torpedo [7 - 111 electric organ has now been purified in quantities such that its structural analysis appears readily feasible [1,4,5,12- 141. The mechanism by which this membrane-bound protein controls the selective translocation of alkali cations remains, however, largely unknown. We considered that a first step in this analysis was to investigate in detail the binding properties of the cholinergic receptor protein from Torpedo marmorata in its membranebound, detergent-dissolved and purified states. Starting from the receptor-rich membrane preparation of Cohen et al. [15], the cholinergic receptor protein can be shown to exist under at least three classes of binding states. Interconversion between these states can be achieved by varying the concentration of detergent. A preliminary report of this work has been published [16]. ____

Enzyme. Acetylcholinesterase (EC 3.1.1.7). Trivial names. Tetram, O,O’-diethyl-S-(2-diethylamino)ethyl phosphorothiolate; Dimethisoquin, l-(2-dimethylaminoethoxy)-3n-butylisoquinoline hydrochloride; Triton X-100 is a condensate of ethylene oxide with an octylphenol and can be represented by the formula (CH,), - CCH, - C(CH,), - C6H,0- (CH,CH,O),H the mean value of n for this material being 10; Emulphogene BC-720, analogue of Triton, without the phenol ring. Eur. J. Biochem. 55 (1975)

MATERIALS AND METHODS Preparation of Membrane Fragments Receptor-rich membrane fragments were prepared from fresh electric tissue of Torpedo marmorata following the method previously described [17], which is a modification of the method of Cohen et al. [15]. The resulting membrane suspensions contained usually about 10 mg protein/ml in 0.8- 1.0 M sucrose, 0.02 % NaN,, and they were stored until use in an ice-bath under argon. The specific activity of the preparations used ranged from 1000 to 2000nmol of ~t-[~H]toxin binding sites per g protein. Dissolution of the Membrane Fragments by Detergents Membrane fragments were dissolved by anionic (sodium cholate) or neutral (Triton X-100 or Emulphogene BC-720) detergents. The dissolution procedure that we shall refer to as “standard” in the present study was as follows: one volume of membrane suspension (FZ 10 mg protein/ml in 0.8- 1.0 M sucrose) was added under continuous stirring into an equal volume of solution containing 25% (w/v)

506

Interconversion between Affinity States of Cholinergic Receptor

sodium cholate, 1 M Tris . HCI, pH 8.0, and 10 mM EDTA. After gentle stirring for 15 rnin at 4 "C, the mixture was centrifuged at 100000 x g for 1 h at 4 "C. The yields in the supernatant were 90- 95 % of the proteins and 75 - 85 of the cx-toxin binding sites. These yields did not change significantly when the mixture was diluted with Torpedo Ringer's solution to a final concentration of 3 % (w/v) sodium cholate and subsequently centrifuged. Other methods of dissolution were sometimes used. For instance, a 50% (w/v) solution of sodium cholate was added to the membrane suspension to a final concentration of 1 The mixture was stirred for 30 rnin at 4 'C and centrifuged. In the case of Emulphogene or TritonX-100, methods of dissolution were similar to those used with sodium cholate. More details are given in the legend of Table 2.

:,;

s;,.

Reassociation bj, Dr tergen t Dilution

Dissolved membrane fragments were reassociated by reducing the concentration of detergent simply by diluting the detergent extract with ice-cooled Torpedo Ringer's solution (250 mM NaCI, 5 mM KCI, 5 mM CaCI,, 4 m M MgCl,, 5 mM sodium phosphate buffer, pH 7.0). Binding of Revrrsihle Ligands

Binding of [3H]acetylcholine (specific activity : 290 mCi/mmol, The Radiochemical Centre, Amersham) was followed by two methods: ultracentrifugation and equilibrium dialysis. The former was used in the case of native and reassociated membrane fragments. The membrane suspension (or the detergent solution in the case of a reassociation experiment) was diluted into Torpedo Ringer's solution (usually 50 times or more), and preincubated with 70 pM Tetram, i.e. O,O-diethyl-S-(2-diethylamino)ethyl phosphorothiolate. for at least 30 min; then, [3H]acetylcholine was added to the mixture [18-201. After incubation for about 10 rnin at 4 "C, the mixture was centrifuged at 100000 x g for 120 min still at 4 T ; 100-pl aliquots of the mixture were counted in duplicate before and after the centrifugation in 10 ml of Bray's solution. It was checked that all the a-toxin sites sediment after centrifugation, especially in the case of a reassociation experiment. When the effect of a local anaesthetic was studied, the diluted mixture was preincubated for about 15 min at 4 ' C with the desired concentration of a local anaesthetic before the addition of [3H]acetylcholine. When the effect of CaZ+ ions was examined, Torpedo Ringer's solution depleted of Ca2+and Mg2+

ions was used ; then, the desired concentration of CaCI, was added. Binding of [3H]acetylcholine or [3H]decamethonium (400 mCi/mmol, The Radiochemical Centre, Amersham) to dissolved (and reassociated) membrane fragments or to the purified receptor protein was studied by the technique of equilibrium dialysis [21]. The solution was diluted with ice-cooled Torpedo Ringer's solution to the desired concentration of detergent and incubated with 70 pM Tetram for at least 30 min. Aliquots (300-400 pl) of the sample were dialyzed at 4 "C for 15 h against 15 ml of Torpedo Ringer's solution supplemented with the same concentration of detergent, 7 pM Tetram, and the desired concentration of [3H]acetylcholine or [3H]decamethonium. Radioactivity inside the bags and in the buffer solution was measured as above. Purification oj Acetylcholine Receptor Protein

Acetylcholine receptor protein was purified by affinity chromatography following the method of Olsen et al. [22] from membrane fragments dissolved by 2-5 % Triton X-100. The purified protein was concentrated by small DEAE-cellulose columns 1221. No further purification was performed by sucrosegradient centrifugation except for the exchange of Triton X-100 with sodium-cholate. Binding ~ f a - [ ~ H ] T o x i n

The number of a-toxin sites was always measured by the method of Olsen et al. [22], using "helper" fraction and ~ - [ ~ H ] t o xfrom i n Naja nigricollis prepared by the method of Menez et al. [23] (specific activity 10- 14 Ci/mmol). Exchange of Detergent

Sodium cholate was exchanged with Triton X-100 by centrifugation in a discontinuous sucrose gradient or by dissolution of reassociated pellet. In the former case, a membrane solution prepared by the standard procedure was diluted into water to a sodium cholate concentration of 5%. 2 ml of this solution were then layered on a sucrose gradient, composed of 1.5 ml each of 20 %and 5 % (w/v) sucrose in 5 ':(;Triton X-100, 2 mM sodium phosphate buffer, pH 7.0. The gradients were centrifuged in a Spinco SW65 rotor at 60000 rev./ rnin for 3.5 h. 65 % of the a-toxin sites initially present were collected between the two layers of sucrose and diluted 2-fold with Torpedo Ringer's solution containing 5 % Triton X-100. [3H]Acetylcholine binding was measured by equilibrium dialysis directly with this last solution. E:ur. J . Biochem. 5.5 (1975)

507

H. Sugiyama and J.-P. Changeux 50t A

I

Ii

I/

/*’

native membrane

/

0 1 / [ K h I free

iFM~’)

I i 0 t

a-toxin sites

\ \

0.1

0.2 [Achlbound

0.3

0.4

0.5

(WM)

Fig. 1. Consequences of dissolution by d r u m cholute on the binding of acetylcholine to the cholinergic receptor site. Receptor-rich membrane fragments were dissolved by the standard procedure and diluted to 3 ”/, sodium cholate with Ringer’s solution (see Methods). (A) Double-reciprocal plot. The ordinate was expressed as the ratio of the total concentration of receptor sites as measured with the ~-[ ~ H] t o x i n[AChR],,,,,, , to the concentration of bound acetylcholine, [AChIbound. The total number of sc-toxin sites was 1.4 pmollg and protein for the native receptor-rich membranes (@--O) 1.5 pmol/g protein after dissolution by sodium cholate (-0).

The concentration of proteins was 0.036 and 0.47 rng/ml for membranes and dissolved preparation, respectively. (B) Scatchard plot of the same data as (A). For the two curves, the maximal value of the ratio [ACh]b,,,,/[ACh]free was taken as one. After iiormalisation the concentration of toxin sites becomes 0.039 pM (arrow) for the native membranes and 0.83 pM for the dissolved preparation. For reasons of scale, the two points corresponding to the lowest and highest [ACh],,,,,, obtained in the presence of 3 % sodium cholate in this figure are not shown in (A)

Sodium cholate was also exchanged with Triton X-100 by another method. One volume of supernatant dissolved in sodium cholate was diluted 30 times with water, and centrifuged at 100000 x g for 120 min. The pellet was resuspended with approx. 0.5 vol. of 5 M NaCI, 20 mM phosphate buffer, pH 7.4. The suspension was added into the equal volume of 10 % Triton X-100, 10 mM EDTA, stirred at 4 “C for 15 min and centrifuged at 100000 x g for 120 min. The supernatant was used for equilibrium dialysis.

All acetylcholine bound by the membrane fragments was displaced by an excess (1 - 2 pM) of N . nigricollis a-toxin and was therefore [24] associated with the cholinergic receptor site [19,24]. On Fig. 1 are given the double-reciprocal and Scatchard plots of the binding data. With these particular preparations, the cooperative binding of acetylcholine [I91 was not detected and the data were fitted by an hyperbola. The single dissociation constant determined from these lo-* M (Table 1). The numplots was (3.4 k 0 . 8 ) ~ bers of acetylcholine and a-toxin binding sites were always close to each other [I91 (Table 1). Cohen et al. [20]have found that, with the receptorrich membrane fragments, CaZf ions (up to 5 mM) and local anaesthetics (up to 0.3 mM dimethisoquin) cause a significant increase of affinity of the highaffinity sites for acetylcholine. This important finding was confirmed under the present conditions of assay (Fig. 5s).

RESULTS CONSEQUENCES OF DISSOLUTION ON THE BINDING PROPERTIES OF THE CHOLINERGIC RECEPTOR PROTEIN

Acetylcholine Binding to Membrane Fragments

The binding of acetylcholine to receptor-rich membrane fragments was always measured by centrifugation in the presence of 70 pM Tetram, a powerful inhibitor of acetylcholinesterase (see Methods). With the membrane preparation used and under the present conditions of assay, 70 pM Tetram did not significantly modify the interaction of acetylcholine with the receptor site [18 - 201 (significant decrease of acetylcholine binding was observed only above 200 pM). Eur. J. Biochem. 55 (1975)

Dissolution by Detergents

Membrane fragments were dissolved in the presence of sodium cholate, Triton X-100 or Emulphogene BC-720 under the conditions given in Methods. With each solution, the total number of ~ [ ~ H I t o x sites in was routinely measured and the binding of [3H]acetylcholine followed by equilibrium dialysis in the presence

lnterconversion between Affinity States of Cholmergic Receptor

SO8

Table 1. Bindinl: qf’pH]ucrrylclioline ro the cholinergic receptor in its membrane-hound, sodium ckolare-dissolved and retrs.sociated stutt’.s Acetylcholine (ACh) binding was measured by centrifugation for membrane fragments and by equilibrium dialysis in the dissolved and reassociated states, this membrane preparation was different from that used in Fig. 1. Numbers in parentheses are the estimated difference between the number of r-toxin sites and the number of medium-affinity or high-affinity acetylcholine binding sites. Sensitivity refers to the increase of ACh binding by dimethisoquin (up to 0.3 mM) or C a Z +ions (up to 5 mM) and was examined by equilibrium dialysis (dissolved states) or centrifugation (membranes and reassociated state) ~

Receptor Stdtr‘

~~

~

Number of sites for ~

a-toxin

x ~

~~

K , (ACh)

ACh

nmolig protein

*

1990 200 1700 f 100

Reaswciated

1960 & 340

440 (1 300)

50

680 f 70 (1000)

Sodium cholate 1 ‘ I , ,

1700

local anaesthetic

CaZ ion

+

+

+

M

~~~~

Membrane fragments Detergent solution Sodium cholatc 3 ‘’

Sensitivity to

~~~~~

100

*

Sodium choldte 0 3‘’

790 (900)

Sodium cholntc 0 1 ’’ 30 1 z 30

acetylcholine binding sites with a high affinity for acetylcholine represented only 60 to 75 % of the total number of a-toxin sites. The rest of the acetylcholine sites had a dissociation constant larger than 3 x 10-7 M. Ca2+ ions (up to 3 mM) and the local anaesthetic dimethisoquin (up to 30 pM) still no longer increase the affinity for acetylcholine despite the fact that the acetylcholine binding sites are in their high-affinity state. REASSOCIATION BY DILUTION

Sensitivity to anaesthetic

Ca2+ ions

-

-

-

-

Yield of a-toxin sites by dissolution (conditions)

%

binding curve given by a reassociated sodium cholate extract. After dilution a significant fraction of the sites (approximately 50% of the a-toxin sites in the experiment of Table 1) binds acetylcholine with a dissociation constant close to 3 x l o p 8 M, i.e. in the same range as that of the membrane-bound receptor. The reassociation by dilution therefore leads to the recovery of the high-affinity state of the cholinergic receptor site. The recovery is however incomplete. About 50 % of the total number of a-toxin sites still bind acetylcholine with a dissociation constant larger than 10-7 M.

Recovery of High-Affinity Sites for Acetylcholine Binding from Crude Sodium Cholate Extracts

Dependence on Sodium Cholate Concentration

It has been shown that the dissolution by sodium cholate or deoxycholate of Electrophorus [25] and Torpedo [17] membrane fragments followed by the dialysis of the detergent leads to the reconstitution of closed microsacs. The method we have developed here to eliminate the detergent after dissolution is much simpler: the soluble membrane extract is diluted into detergent-free Torpedo Ringer’s solution. Under such conditions, as long as the concentration of detergent falls below a critical concentration, then all the a-toxin binding activity reassociates and sediments when centrifuged at 100000 x g for 120 min. Binding of [3H]acetylcholineto these aggregates was measured by centrifugation (see Methods). Fig. 2 shows the

In the experiment illustrated by Fig. 3, a crude extract of membrane fragments in 3 % sodium cholate was diluted into buffers containing different amounts of sodium cholate. The final concentration of protein in the diluted extracts was kept constant. Between 1 and 0 . 3 x (w/v) sodium cholate, the affinity for acetylcholine increased sharply (approx. 6-fold). An interconversion between “medium-affinity’’ and “high-affinity’’ states of the acetylcholine binding sites took place. Since the binding of acetylcholine was always measured in the presence of 70 pM Tetram, we considered the possibility that the concentration of sodium cholate primarily affects the binding of Tetram

Eur. J. Biochem. 55 (1975)

510

Interconversion between Affinity States of Cholinergic Receptor

.o

25

1

20

0.8

H

C= r

0.6

. V 4

I

w

2 0.4 a

I

J 5

.i"

reassociated (0.25% cholate)

P a - toxin

sitii-..

/

reassociated (0.25% cholate)

0.2

.v-

d - 10

0 0

10

20

60

30 [ACh]bOund (nM)

Fig. 2. Biridirip of uci~t~ii~holinr 10 the cholinergic receptor site after dissolution ti), wdiun7 cholate and reassociation by dilution. Native receptor-rich membrane fragments were dissolved by the standard procedure and diluted to the indicated concentrations of sodium cholate with Ringer's solution (see Methods). The data in 3"i, cholate are the came as in Fig. 1 . (A) Double-reciprocal plot. The heterogeneit) i)f thc population of acetylcholine binding sites in the presence 01' 3",, sodium cholate is more evident in this figure than in Fig. 1 ( A ) . Three points corresponding to the lowest [ACh],,,,,,, in the 3 sodium cholate curve from Fig. 1 (B) are not

0 0

I

1

I

1

2

3

Sodium cholate

6

w/v)

a - t o x i n sites

shown for reasons of scale. The total number of r-toxin sites were 1.5 pmol/g protein and 0.54 pmol:g protein for dissolved and reassociated membranes, respectively. The concentrations of protein was 0.47 mg/ml and 0.16 mg/ml, respectively. (13) Scatchard plot of the data of (A). Concentrations of r-toxin sites were normalized as in Fig. 1 (B), and after normalisation becomes 0.83 pM and 0.043 pM (arrow) for dissolved and reassociated membranes, respectively. For reasons of scale only, the first scven points on the curve of 3 % sodium cholate in Fig. 1 (N) arc shown in this figure

- 0

0

1 Sodium cholate

2

3

'0

wiv)

Fig. 3 . L'crricirion of rhc di,tsocrutiori c'onstunt Jor acetylcholine urid of thr number. of midiurn-uffinitq plus high-ujjznity sites us afunction of .sodium c~liolati,concmtration. Receptor-rich membranes were dissolved by the standard procedure (see Methods), then the solution was diluted into Torpedo Ringer's solution to the indicated concentration or sodium cholate. ( 0 ) The dissociation constant, Kd. and (0) number of sites for acetylcholine with high and medium affinity at each sodium cholate concentration

Fig. 4. Vurialion oj the umourit of f ' H l c r i c~t,rlc~lroiitie (0) und ( " H I decamethonium (e) bound to the cholinergic receptor- site us a function of' sodium chelate concentrution. Native membrane fragments were dissolved by the standard procedure (see Methods) and diluted to the indicated concentration of sodium cholate. The concentration of receptor sites was the same for all the points (0.07 p b l for [3H]acetylcholine and 0.55 pM for [3H]decamelhoniu~n)as well as the free concentration of [3H]acetylcholine (0.016 p M ) or f3H]dccamethonium (1.0 pM) only the concentration of sodium cholate was varied

rather than that of acetylcholine. Binding of decamethonium was therefore determined at various concentrations of detergent in the absence of Tetram. Fig. 4 shows that the amount of ligand bound increa-

ses for both decamethonium and acetylcholine and within the same range of sodium cholate concentration. The transition therefore primarily affects the binding of cholinergic ligands. tu-.J . Niochem. 5.5 (1975)

H. Sugiyama and J.-P. Changeux

51 1

Recovery of the Effects of Ca2' Ions and Local Anaesthetics

151 P

After dissolution Ca2 and local anaesthetics no longer increase the affinity of the receptor site for acetylcholine. Fig. 5 illustrates that reassociation by dilution leads to a recovery of this property. The variation of acetylcholine binding with the concentrations of Ca2+ and of a local anaesthetic (dimethisoquin) followed almost exactly the same pattern with the reassociated and with the membrane-bound receptor. +

Consequence of the Exchange of Detergent on the AJj'inityfor Acetylcholine

native membrane\:\ \

0

1 .o

10

100

loo0

[Dimethisoquin](pM)

Fig.5. Eflbcts of ( A ) Cu" ions und oj ( B ) a local anaesthetic jdirnethisoquin) on the binding of acetylcholine to cholinergic receptor site after reussociution by dilution. Receptor-rich membranes were dissolved by the standard procedure and diluted to 0.13% (wiv) sodium cholate for reassociation or 3.0% for dissolved receptor protein (AChR). In the case of native membrane and reassociated receptor, the binding of acetylcholine was measured by centrifugation with the total concentration of [3H]acetylcholine in the centrifuge tube 0.02 pM in each case. In the case of dissolved receptor (3.0 sodium cholate), the binding was measured by equilibrium dialysis in the presence of 0.2 pM [3H]acetylcholine in the outer solution. The concentration of @-toxin sites and protein were 0.086 pM and 0.2 mglml (A), and 0.033 pM and 0.02 mg/ml, 0.086 pM and 0.05 mg/ml, and 0.65 pM and 0.4 mg/ml for native membrane, reassociated receptor, and dissolved preparation, respectively (B)

Relevant to these observations, it is appropriate to mention that the critical micelle concentration of sodium cholate is about 1.8 % [ 3 8 ] . Since the value of this concentration depends upon parameters such as temperature, ionic strength, pH etc., the exact value for cholate under the conditions where acetylcholine binding was measured, although not known, might be smaller than 1.8% and, therefore, close to the concentration of cholate at which the transition takes place. Eur. J. Biochem. 55 (1975)

The receptor protein from membrane fragments dissolved with Triton X-100 exhibit a much higher affinity for acetylcholine than that which results from dissolution by sodium chokdte. Considering the possibility that the difference of affinity was primarily caused by the presence of a different detergents during the binding assay, membrane fragments were first dissolved with sodium cholate, then sodium cholate was exchanged with Triton X-100. This last step was done by two different procedures (see Methods) : (a) ultracentrifugation of the sodium cholate extract on a discontinuous sucrose gradient in 5 % Triton X-100; (b) reassociation of the sodium cholate extract by dilution and subsequent dissolution of the pellet by 5 %Triton X-100. Table 3 shows that in both cases, acetylcholine binds with a dissociation constant close to 2 x M to a population of sites corresponding to 20- 25 % of the total number of cc-toxin sites. The dissociation constant of acetylcholine for the other sites was M. larger than In other words, after exchange of sodium cholate by Triton X-100, the cholinergic receptor site exhibits the same affinity for acetylcholine as that found with the starting sodium cholate extract. Although it is likely that under these conditions the great majority of cholate is replaced by Triton X-100 [26-281, we cannot eliminate the possibility that trace amounts of cholate remain bound to the receptor protein in the presence of Triton X-100. Effect of Storage on the Recovery of High-Affinity Sites from a Sodium Cholate Extract

When a 3% sodium cholate extract is stored for 23 h in an ice bath, then reassociated by dilution, the yield of the recovery of the high-affinity sites does not significantly differ from that given by a fresh extract. On the other hand, this yield decreases by 60% after

512

Interconversion between Affinity States of Cholinergic Receptor

Table 3. Consequences ofthr exchange of sodium cholate by Triton X-100 on the binding of [3H]ucetylcholine (ACA) to thi. cholinergi:ic.rec'eptor proteinlrom a crude chohte extract Exchange of detergent

Final detergent

108x Kd (ACh)

ACh sites

( % a-toxin sites) 0

By centrifugation

Triton X-100 5 "0 Triton X-100 5

By re-dissolution

2s the rest 20 the rest

x

Table 4. Consequences OJ' .storage on the binding properties of the cholinergic recepior in u crude detergent extract The acetylcholine (ACh) binding was measured after the treatment as follows. ( I ) Dissolved by standard procedure, diluted to 3"/, sodium cholate with Ringer's solution, and (a) diluted immediately to 0.1 y;, sodium cholate or (b) kept for 23 h in ice-bath, then diluted to 0.1 y i sodium cholate. (11) Dissolved in 7.27; sodium cholate, 0.5 M Tris . HCl, pH 8.0,20 mM EDTA, centrifuged, and (c) diluted immediately to 0.5",, sodium cholate and kept for 72 h in icebath, then diluted to 0.1 2,sodium cholate or (d) kept for 72 h in ice-bath without any treatment. and then diluted to 0.1 "/, sodium cholate Storage conditions

I N o storage (a) Stored in 3 cholate for 23 h ( b )

I1 No storage (c) Stored in 7.2", cholate for 72 h (d)

ACh

High-affinity sites

nmol/g protein

%, &-toxinsites

720

so

670

45

-

50

-

20

a storage of 72 h in the presence of 7 sodium cholate (Table 4). The capacity to recover high-affinity sites can therefore be lost by storage. Consequences of Purification of the Properties of the Receptor Protein Reassociated by Dilution

Since the presence of a negatively charged detergent interferes with the chromatography on an affinity column with a cholinergic arm, all the purification procedures were done in the presence of Triton X-100. In the crude Triton X-100 extract before purification, the numbers of ~ - [ ~ H ] t o xsites i n and [3H]acetylcholine sites were identical. All these sites exhibit a high affinity for acetylcholine (& % 2 x lo-' M). Subsequent decrease of detergent concentration by dilution did not significantly modify these dissociation constants (Table 5 ) . After purification (Table 5), the binding curve of acetylcholine no longer follows an hyperbola. A heterogeneity of binding is again present. The majority

Yield of a-toxin sites by the treatment

M 20

65

2 100

17

20

2 100

of the sites bind acetylcholine with a low affinity ( z2 x M); the rest (10-30(5/, of the total number of a-toxin sites) exhibiting a high affinity ( % 3 x l o p 8 M). In one experiment, the precision was such that the total number of acetylcholine sites, with both low and high affinity, could be determined. It was found again close to the total number of a-[3H]toxin sites (Table 5). In addition, the yield of the purification was high enough to eliminate the eventuality that the purification selected a small fraction of sites present in the crude extract. In other words, the affinity of the receptor protein changes as a result of the purification. When centrifuged on a continuous sucrose gradient which contained 1 "/, sodium cholate (Fig. 6), the purified protein shows, like the crude extracts [29 - 301, two peaks of toxin binding activity. Acetylcholine binding was measured with these two distinct fractions in the presence of 1 "/, sodium cholate. N o significant difference in the ratios between high-affinity and low-affinity sites was noticed between these two fractions (Table 5). With these purified fractions the decrease of cholate concentration below 0.3 O 0 does not change the ratio between these two populations of site. In contrast with what was observed with the crude membrane extracts, with the purified protein dilution of the detergent does not yield a significant recovery of the high-affinity sites.

DISCUSSION Because of their high content of cholinergic receptor site, the purified membrane fragments from Torpedo of Cohen et al. [20] are particularly suitable preparations to study the binding properties of the cholinergic receptor protein and their control by the membrane environment. In agreement with the results of Weber and Changeux [19], we have found that acetylcholine binds to these membrane fragments in the presence of Tetram, with a high affinity and a single dissociaM). The K , appears tion constant (& = 3 x Eur. J. Biochem. S.5 (1975)

H. Sugiyama and J.-P. Changeux

513

Table 5. Binding of'acetylcholine ( A C h ) to the purified cholinergic receptor protein: consequences of detergent exchange and dilution Receptor protein

Yield

cr-Toxin sites

Detergent in ACh binding medium

ACh sites with affinity ~

high

(Kd = 3 x lo-@M) nmol/mg protein Crude extract After affinity column

nmol/g protein

(Kd z 2 x

100

870

Triton X-100

(0.2) (0.03)

900 900

n. d . b n. d . b

28

6900

Triton X-100

(1)

470

7400

22 fr. I"

2500

sodium cholate (1) (0.2)

610 670

2200 2500

2900

sodium cholate (1) (0.2)

600 660

2300 2300

fr. 11"

a

( %)

low M)

Fractions I and I1 of sucrose gradient centrifugation shown in Fig. 6. Not detected.

/"'

0

-5

50

100

150

1 / [ A C h ] t r ~(KM-')

Fig. 6. S u c r o . ~grudieiir ci.irri.ifirgorioii oJ rhe cholinergic receptor protein pur(fied by ujjinity chromutogruphy and binding of acetylcholine to twoJractions obtained by the centrijiugution. The cholinergic receptor protein was purified as described in Methods in the presence of 1 Triton X-100 (specific activity 4400 nmol r-toxin sites/g protein), and centrifuged on a linear ( 5 - 20 %) sucrose gradient containing 1"i, sodium cholate, 33 mM Tris HC1 buffer, pH 7.5, and 0.007 NaN,. Two fractions were pooled as indicated in the inserted figure. The recovery of a-toxin sites of the centrifugation was 71%. Each fraction was then diluted 5-times with Ringer's solution with or without 3 % sodium cholate, and [3H]acetylcholine binding was measured, as described in Methods, in Ringer's solution with 1.0% or 0.2% sodium cholate, respectively. The final concentration of a-toxin sites in the dialysis bags was 0.1 1 pM and 0.25 pM for fractions I (circles) and I1 (squares), respectively. Open symbols indicate the binding in l.OO/, sodium cholate, and closed symbols 0.2 sodium cholate