Proc. Natl. Acad. Sci. USAVol. 92, pp. 12008-12012, December 1995Pharmacology

Cocaine: Mechanism of inhibition of a muscle acetylcholinereceptor studied by a laser-pulse photolysis technique

(caged carbamoylcholine)

Li Niut, LEO G. ABOODt, AND GEORGE P. HEsst§tSection of Biochemistry, Molecular and Cell Biology, Division of Biological Sciences, 216 Biotechnology Building, Cornell University, Ithaca, NY 14853-2703;and tDepartment of Pharmacology, University of Rochester Medical Center, Rochester, NY 14627

Contributed by George P. Hess, June 30, 1995

ABSTRACT Effects of cocaine on the muscle nicotinicacetylcholine receptor were investigated by using a chemicalkinetic technique with a microsecond time resolution. Thismembrane-bound receptor regulates signal transmission be-tween nerve and muscle cells, initiates muscle contraction,and is inhibited by cocaine, an abused drug. The inhibitionmechanism is not well understood because of the lack ofchemical kinetic techniques with the appropriate (microsec-ond) time resolution. Such a technique, utilizing laser-pulsephotolysis, was recently developed; by using it the followingresults were obtained. (i) The apparent cocaine dissociationconstant of the closed-channel receptor form is '50 ,uM. Highcarbamoylcholine concentrations and, therefore, increasedconcentrations of the open-channel receptor form, decreasereceptor affinity for cocaine -6-fold. (ii) The rate of thereceptor reaction with cocaine is at least '30-fold slower thanthe channel-opening rate, resulting in a cocaine-induced de-crease in the concentration of open receptor channels withouta concomitant decrease in the channel-opening or -closingrates. (iii) The channel-closing rate increases -1.5-fold as thecocaine concentration is increased from 20 to 60 ,uM but thenremains constant as the concentration is increased further.The results are consistent with a mechanism in which cocainefirst binds rapidly to a regulatory site of the receptor, whichcan still form transmembrane channels. Subsequently, a slowstep (ti,2 70 ms) leads to a receptor form that cannot formtransmembrane channels, and acetylcholine receptor-mediated signal transmission is, therefore, blocked. Implica-tions for the search for therapeutic agents that alleviatecocaine poisoning are mentioned.

Cocaine is a commonly abused drug (1), with more than fivemillion addicts in the United States alone (2). Cocaine abusecan result in myocardial ischemia, seizures, hypothermia, andsudden death (3). The drug inhibits nicotinic acetylcholinereceptors (4-6) on central nervous system neurons and musclecells and muscarinic acetylcholine receptors on cardiac, vas-cular, and other smooth muscles (7, 8). Cocaine and its analogsantagonize nicotine-induced seizures in mice by inhibitingnicotinic ion-channel receptors (9). The mechanism by whichcocaine inhibits the muscle nicotinic acetylcholine receptor isthe subject of this report.The results of many electrophysiological and chemical la-

beling experiments led to a commonly accepted mechanism bywhich inhibitory organic cations, including procaine, atropine,scopolamine, and cocaine, enter the receptor channel after ithas opened and sterically block the transmembrane ion flowand, thereby, signal transmission between cells (10-15). De-viations from this simple channel-blocking mechanism havebeen noted in some electrophysiological experiments (16-20),and the binding of inhibitors to closed-channel forms of the

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicate this fact.

receptor has been suggested (4, 5, 16). A mechanism thatinvolves the binding of an inhibitor to a regulatory site on theclosed- and open-channel forms of the receptor (Fig. 1) can bedistinguished from the channel-blocking mechanism, provid-ing one can determine the effects of inhibitors on the rateconstants for both channel opening and closing (23). A chem-ical kinetic technique suitable for such measurements wasrecently developed (23-25) to obtain information about thechemical mechanism of neurotransmitter-gated receptors thatcan form open ion channels in the submillisecond region.The technique, employed here, utilizes photolabile inert

precursors of neurotransmitters that can be equilibrated withreceptors on the surface of single cells (23-25). A laser pulseliberates the neurotransmitter in the microsecond time region.The released neurotransmitter binds to the receptor, whichforms an open channel; the resulting current can be measuredby a whole-cell current-recording technique (26). The tech-nique we used, with a 100-,us time resolution, allows one to (i)separate sequential steps of the reaction along the time axisand (ii) investigate the reaction in whole cells over a wideconcentration range of neurotransmitter and other ligands.Consequently, it is possible to determine (i) the rate constantsfor both the opening (kop) and closing (kc1) of receptorchannels (25), (ii) the dissociation constant for the neurotrans-mitter-binding site that controls channel opening (25), and (iii)the effects of inhibitors on kop and kc1 separately and, therefore,the dissociation constants for the binding of inhibitors to theclosed- and open-channel receptor forms independently (23).Such measurements have been made with the acetylcholinereceptor in BC3H1 muscle cells (25), including the effects ofthe inhibitor procaine on channel-opening and -closing rates(23). In this report, we have used the same technique and cellline to study the effects of cocaine on the receptor. The resultsare inconsistent with the channel-blocking mechanism (10-15). A mechanism in which cocaine binds to a regulatory siteon the receptor and inhibits the opening of the receptorchannel is consistent with the results.

MATERIALS AND METHODSCell Culture and Whole-Cell Recording. The BC3H1 cell

line, which expresses muscle type nicotinic acetylcholine re-ceptors (27), was cultured as described (28). The equilibrationof these cells with various ligands using a cell-flow technique(29-31) and the determination of the whole-cell current (26)due to opening of receptor channels have been described indetail (29-31).

Laser-Pulse Photolysis. Caged carbamoylcholine [N-(a-carboxy)-2-nitrobenzyl]carbamoylcholine (24) in the presenceor absence of cocaine was equilibrated with a BC3H1 cell in thewhole-cell recording mode. A Candela UV 500 flash-lamp-pumped dye laser, with oxazine 720 perchlorate (Exciton) asthe laser dye, was used for photolysis at 343 nm. The time

§To whom reprint requests should be addressed.

12008

Dow

nloa

ded

by g

uest

on

Janu

ary

1, 2

021

Proc. Natl. Acad. Sci. USA 92 (1995) 12009

I+L+A ~K,koI + L + A - AL + AL2 open channelI kCI '2kcK K,1K K'

IA IAL2 k IAL2 open channel

k'bl k'fIA*Ln inhibited channel

FIG. 1. Minimum mechanisms for the activation and inhibition ofthe nicotinic acetylcholine receptor. The reaction is initiated when an

activating ligand (L) such as carbamoylcholine binds to the activenondesensitized receptor form (A). The reaction contains severalequivalent binding steps for L, characterized by the dissociationconstants Kl, and leads to receptor-ligand complexes (e.g., AL1 andAL2, where the subscript represents the number of ligand moleculesbound). Several transitions in protein conformation result. A trans-membrane channel (AL2) opens in the microsecond-millisecond timeregion, characterized by rate constants for channel opening (kop) andclosing (kc1). In the presence of L, the receptor is also rapidly (-200ms) (21) converted to inactive desensitized receptor forms with alteredligand-binding properties (22). The transitions to desensitized receptorforms are not shown. Ki and Kj are the inhibitor dissociation constantsfor the closed (A, AL1, and AL2) and open (AL2) forms of the receptorchannel, respectively. The rate constants for the openipgfkop and kop)and closing (kc, and kgl) of the open-channel forms AL2 and IAL2,respectively, are as indicated.k' and kl are overall rate constants forthe formation of the receptor-inhibitor complexes IA*L, that cannotform open channels (n = 0, 1, or 2).

constants for the rising phase of the whole-cell current were

fitted by u,sing the data analysis programs ORIGIN (MicroCalSoftware) and LOTUS. All other details of the procedure havebeen described (23, 25).

RESULTS AND DISCUSSION

In laser-pulse photolysis experiments (Fig. 2), an inactivephotolabile precursor of carbamoylcholine (caged carbamoyl-choline) (24) was equilibrated with a BC3H1 cell suspendedfrom a current-recording electrode (26). When carbamoylcho-line was released from the caged carbamoylcholine by a laserpulse, acetylcholine receptor channels opened and the currentflowing through them was recorded (Fig. 2). From these resultsthe effects of cocaine on both the rates of channel opening andclosing and the amplitude of the whole-cell currents, a measure

of the concentration of open receptor channels, can be deter-mined.

Effect of Cocaine on the Concentration of Open ReceptorChannels. The amplitude of the observed whole-cell currentdecreased with increasing cocaine concentration, indicatingthat cocaine inhibits the receptor (Fig. 2). To determine theeffect of cocaine on the concentration of open receptorchannels, we measured the ratio of the maximum currentamplitude in the absence,Im., and presence, Imax(10), of cocaineas a function of cocaine concentration [Fig. 3A and Appendix,Eq.I(i)]. The maximum current amplitude was measured withlaser-pulse photolysis and with a cell-flow technique (10-mstime resolution) that allows one to correct the observedcurrent for receptor desensitization (30). The photolysis andcell-flow techniques give similar results within experimentalerror (Fig. 3A). The solid lines in Fig. 3A represent experi-ments in which the cell was preincubated with cocaine for 200ms. This is sufficient time to inhibit the receptors at allconcentrations of cocaine used; preincubation with cocaine foras long as 1 min caused no further inhibition. Three carbamoyl-choline concentrations [20 AtM (Fig. 3A, line a), 100 txM (Fig.3A, line a), and 1 mM (Fig. 3A, line d)] were used to vary theconcentration of open receptor channels. For the mechanisms

2.0_

-

o 1

0

0

0= 1 r

o

0 2 4Time (ms)

6 8 10

FIG. 2. Laser-pulse photolysis of caged carbamoylcholine withBC3H1 cells at pH 7.4,22°C, and -90 mV. (Upper) Whole-cell currentsgenerated by 20 ALM released carbamoylcholine in the absence (curvea) and presence of 20,uM (curve b) or 120,uM (curve c) cocaine. Theobserved first-order rate constants and maximum amplitudes of thewhole-cell currents for curves a, b, and c are 510 ± 4, 730 ± 2, and 790± 7s-1 and 1470, 1094 and 450 pA, respectively. (Lower) Whole-cellcurrents generated by 100,uM released carbamoylcholine in theabsence (curve a) and presence of 20,uM (curve b) and 90,uM (curvec) cocaine. The observed first-order rate constants and maximumamplitudes of the whole-cell currents for curves a, b, and c are 1830± 10, 1840 ± 10, and 2700 ± 22 s-1 and 2354, 1542, and 825 pA,respectively. In all these experiments, the cells were preincubated withcocaine for 200 ms before the measurements were made. The currentspike at time zero is an instrument artifact.

of channel opening in Fig. 1, the slopes of the lines are inverselyproportional to the apparent inhibition constant for cocaine,KI(app). Eq. 1(i) (Appendix) indicates that KI(app) for the inhibitorand receptor forms A, AL, AL2, and AL2 (Fig. 1) is inverselyproportional to the fraction of receptors in any particular form.As the carbamoylcholine concentration is increased, the frac-tion of receptors in the open-channel form will increase. Ifcocaine bound only to the open channel (10-15), as thecarbamoylcholine concentration is increased, the receptoraffinity for cocaine would increase, corresponding to a de-crease in KI(app) for cocaine [Appendix, Eq. 1(i)] and an increasein the slope of theplots shown in Fig. 3A. However, theopposite results were obtained. In the experiments in whichFig. 3, lines a and d, were generated, the slope of line d,measured at a high carbamoylcholine concentration, is lessthan the slope of line a, measured at a low concentration. Thedecrease in slope obtained at high carbamoylcholine concen-trations corresponds to a 6-fold increase in KI(app). Moreover,the increase in KI(app) from 50 to 300,uM, in the range ofcarbamoylcholine concentration used (20,uM to 1 mM),corresponds closely to the decreasing fraction of receptors inthe closed-channel forms (30). These results indicate thatcocaine binds at least six times more strongly to the closed-channel forms of the receptor A, AL, and AL2 than to theopen-channel form AL2 (Fig. 1).

Effect of Preincubation of the Receptor with Cocaine. Theexperiments in Fig. 3 allow one to obtain an estimate of therate at which cocaine inhibits the acetylcholine receptor. Thedashed lines in Fig. 3A represent experiments in which thereceptor was preincubated with cocaine for only 70ms, in thepresence of either 20,uM or 100,uM carbamoylcholine. Thedecrease in the slopes of lines b and c in Fig. 3A indicates that

a3 ~~~~~b

C

a

5-b~- C

Pharmacology: Niu et al.

1 .c

Dow

nloa

ded

by g

uest

on

Janu

ary

1, 2

021

Proc. Natl. Acad. Sci. USA 92 (1995)

A

-10

E

7- xEj

500Cocaine (,AM)

B 1200-

c+-

@1 800-

-

Y 400-0

75

1000

a

b

6 6 5 0

Time (ms)

FIG. 3. Cell-flow and laser-pulse photolysis measurements ofmaximum current amplitudes made with BC3H1 cells at pH 7.4, 22°C,and -90 mV in the absence and presence of cocaine. (A) Ratio of themaximum current amplitudes in absence, Imax, and presence, Imax(IO),of cocaine. Each data point is the mean of at least three measurementsmade with each of three cells. The solid lines represent experimentsin which the receptor was preincubated with cocaine for 200 ms. Thesymbols represent measurements in the presence of 20 ,uM (a), 100,iM (A), and 1000 ,uM (o) carbamoylcholine, and the solid and opensymbols represent the measurements made by laser-pulse photolysisand cell-flow techniques, respectively. The data are plotted accordingto Eq. I(i) (Appendix); the solid lines are the best fits and yield theapparent KI value of 50 ± 10 ,uM for 20 ,uM and 100 ,uM carbamoyl-choline (line a) and of 300 ± 70 AM for 1 mM carbamoylcholine (lined). The dashed lines represent measurements in the presence of 20 ,uM(&, line b) and 100 ,lM (O, line c) carbamoylcholine after the receptorwas preincubated with cocaine for only -70 ms. (B) Whole-cellresponses evoked by 20 ,uM carbamoylcholine in the absence (part a)and presence of 60 ,uM cocaine (parts b and c). The receptors were

preincubated with cocaine in part b for 70 ms and in part c for 200 ms.The whole-cell current amplitudes, after correction for desensitizationoccurring during the rising phase of the current (30), are 1680 ± 20 pA,1250 ± 34 pA, 816 ± 31 pA, for parts a-c, respectively. All the currenttraces were recorded for 12 s but for clarity only the first 5 s are

displayed.

when the cell was preincubated with cocaine for only 70 ms(line b), less inhibition occurred than when it was preincu-bated for 200 ms (line c). Furthermore, less inhibitionoccurred in the presence of 100 ,tM carbamoylcholine (Fig.3A, line c) than in the presence of 20 ,uM carbamoylcholine(Fig. 3A, line b), even though in the former case theconcentration of open receptor channels is 10 times higher(30).The channel-blocking mechanism predicts that as the con-

centration of open receptor channels increases both the ratewith which the inhibitor reacts and the affinity of the receptorfor cocaine increase. The experiments in Fig. 3 indicate,however, that cocaine reacts more rapidly with the closed-channel form than with the open-channel form and thatcocaine binds with a 6-fold greater affinity to the closed-channel forms than to the open-channel form (Fig. 3A). Theseresults are inconsistent with a mechanism in which cocaine can

inhibit the receptor only by entering the receptor channel afterit has opened and blocking it, i.e., the channel-blockingmechanism (10-15). Whether the binding site associated withthe open-channel form, which binds cocaine 6 times lessstrongly than the closed-channel forms, is the same as ordifferent from the high-affinity inhibitor-binding site on theclosed-channel forms cannot be determined from these exper-iments.The effect of the duration of receptor exposure to cocaine

is examined in more detail in Fig. 3B. Solutions containing 20,uM carbamoylcholine in the absence (Fig. 3B, part a) orpresence of 60 AM cocaine (Fig. 3B, parts b and c) were usedin these cell-flow measurements. The current first increased asthe receptor channels opened and then decreased as thereceptors desensitized (became transiently inactive). The de-crease in the maximum current amplitude (in parts b and c)reflects receptor inhibition by cocaine. In part b, the receptorswere preincubated with cocaine for -70 ms and in part c thereceptors were preincubated with cocaine for 200 ms. Theresults indicate that in 70 ms the inhibition reaction was only-50% complete. The inhibition reaction is, therefore, slowcompared to channel opening (t112 < 2 ms). This accounts forthe observation (Figs. 2 and 3) that the current amplitudedecreases with increasing inhibitor concentration and predictsthat the decrease in current amplitude is not accompanied bya concomitant decrease in the rate of either channel openingor closing (Eq. II).

Effect of Cocaine on the Rates of Channel Opening andClosing. First-order plots of the rising phase (Fig. 2) of thecurrent measured in the absence or presence of cocaine areshown in Fig. 4A. The slope of each line is proportional to theobserved rate constant (kobs) for the rising phase of thecurrent, which follows a single exponential rate over 85% ofthe reaction in all the measurements. At 100 ,uM releasedcarbamoylcholine, where the observed rate constant reflectsmainly kop (23, 25), cocaine has no effect on kobs. Surprisingly,however, at 20 ,uM released carbamoylcholine, where kobsreflects kci (23, 25), k.bs increased 1.5-fold in the presence ofcocaine. The effects of cocaine concentration on the values ofkobs for the current rise (Fig. 2) at these concentrations ofreleased carbamoylcholine are summarized in Fig. 4B. Notonly does cocaine not decrease the rate of either channelopening or closing, it in fact increases the rate of channelclosing.The bimolecular rate constant for the interaction of another

cationic inhibitor, procaine, with the acetylcholine receptorhas been reported to be in the region of 107 M-l.S-1 (16), whichis typical for small molecule-protein interactions (32). Thetime required for cocaine to inhibit the receptor (Fig. 3), in thecocaine concentration range investigated, is too long to beconsistent with a bimolecular process. Diffusional access ofcocaine to its binding site can be slow, but it is not expected todepend on the concentration of carbamoylcholine as wasobserved in the experiments (Fig. 3A). That cocaine inhibitionis solely due to its effect on the cell membrane is unlikelybecause cocaine competes with the local anesthetic QX-222 forthe inhibition site (L.N. and G.P.H., unpublished results). Inthe absence of additional evidence, the mechanism in Fig. 1assumes a slow cocaine-induced interconversion of activereceptor-cocaine complexes (IAn) to inactive ones (IA*Ln).The increase in the channel-closing rate when cocaine waspresent (Fig. 4) is also consistent with the existence of an activecocaine-receptor complex. As the cocaine concentration wasincreased from 20 to 60 ,uM, the rate of channel closingincreased 1.5-fold. If this increase in channel-closing raterepresents a bimolecular reaction of the receptor with inhibitorleading to a closed channel, the channel-closing rate is ex-pected to increase as the concentration of inhibitor is in-creased. However, when the cocaine concentration was in-creased from 60 ,uM to 120 ,uM, the rate remained the same.

12010 Pharmacology: Niu et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

1, 2

021

Proc. Natl. Acad. Sci. USA 92 (1995) 12011

A 0O

x

E

- 1E

-i-J

-2

B 3

7U,

0

xn00

-y 1

0

0 2 4Time (ms)

0 100 200 400

Cocaine (pM)

FIG. 4. (A) First-order plot of the whole-cell current rise deter-mined by the laser-pulse photolysis technique as a function of time (seeEq. II). The current induced by 20 ,uM carbamoylcholine in theabsence (open symbols) or presence (solid symbols) of 60 ,iM cocaineis indicated by circles. The data shown were from a single cell; theywere collected at a sampling frequency of 50 kHz (for clarity every 25thpoint is shown). Control, k.bs = 500 + 10 s-1; 60 ,uM cocaine, kobs =900 ± 35 s-'. Currents induced by 100 ,uM carbamoylcholine in theabsence (open symbols) or presence (solid symbols) of 400 ,uM cocaineare indicated by squares. The data shown were from another singlecell; they were collected at a sampling frequency of 50 kHz (for clarityevery 5th point is shown). In the presence and absence of 400 ,uMcocaine, kobs = 2060 ± 125 s (B) Effects of cocaine on k.b. for thecurrent rise time determined by the laser-pulse photolysis technique.Each data point represents three measurements, each measurementmade with a different cell. The observed first-order rate constant wasmeasured from the rising phase of the whole-cell current generated by20 ,iM (cl) or 100 ,uM (0) released carbamoylcholine in the presenceor absence of cocaine.

The data can be accounted for by the rapid formation of anopen-channel-inhibitor complex, IAL2, followed by a conver-sion to an inactive receptor form IA*L2 that is slow comparedto channel closing (Fig. 1). The observed cocaine-inducedincrease in the channel-closing rate can be accounted for by aclosing rate associated with IAL2 that is larger than thatassociated with AL2 (Fig. 1). The observed cocaine-inducedincrease in the channel-closing rate predicts a decrease in thelifetime of the open channel. A cocaine-induced decrease inthe lifetime of the acetylcholine receptor has been observed insingle-channel current recordings from frog muscle (5) andBC3H1 cells (data not shown).

Mechanism of Inhibition by Cocaine. The conclusions de-rived from the present study are summarized in the form of aminimum mechanism in Fig. 1 and are as follows. (i) Cocainebinds to a regulatory site on the receptor and inhibits channelopening. Cocaine also binds to the open-channel form, butwith a 6-fold lower affinity than to the closed-channel forms.Support for this conclusion is drawn from the followingobservations. When the receptors are mainly in the open-channel form, inhibition requires an -6-fold higher concen-tration of cocaine than when they are mainly in the closed-channel forms (Fig. 3A). The rate of the inhibition reactiondecreases as the concentration of open channels increases (Fig.3A). (ii) In the presence of cocaine, the concentration of openchannels is reduced due to a slow process that leads to inactivereceptor forms, IA*L,, where n is 0, 1, or 2. This is indicatedby the results in Fig. 3 showing that inhibition of the receptorby cocaine is slow (tQ12 70 ms) compared to channel opening(tK12< 2 ms; Fig. 2). The slow step in receptor inactivation bycocaine accounts for the observed decrease in the number ofreceptor channels that can open in the presence of cocainewithout a concomitant decrease in the rate constants forchannel opening and closing (Figs. 2 and 4).The assumptions used to solve rate equations for the mech-

anism in Fig. 1 are given in the Appendix. When the observedrate constant for the reaction leading to receptor inhibition ismuch smaller than the value of kobs for channel opening, theintegrated rate equation for the current rise time in presenceof cocaine is

It = Imax(1 - exp kobs t).

kobs = koP(L + K) + kc,(K, 2 0

+ kI(KI+

[IIA]

[IIB]

The symbols are defined in the legend to Fig. 1, and it isassumed that kop is not affected by the binding of cocaine (Fig.4B). It and Ima, represent the current observed in photolysisexperiments at time t and when the current has reached itsmaximum value. Eqs. I and II account for the results shown inFig. 2, namely, decreasing current amplitudes and increasingclosing rate constants with increasing cocaine concentrationsbut no apparent effect of cocaine on the rate constant forchannel opening.The results described here, and those obtained previously (4,

23), indicate the existence of a regulatory (inhibitory) site onthe muscle nicotinic acetylcholine receptor that inhibits theopening of the ion channel and, therefore, signal transmissionbetween cells. The design of therapeutic agents for alleviatingthe toxic effects of inhibitors such as cocaine depends on aknowledge of the inhibition mechanism. The regulatory sitemechanism suggests the possibility of finding therapeuticagents that compete with cocaine for the regulatory site but donot inhibit channel opening and signal transmission. Thisstrategy may not work if cocaine binds inside receptor channelsonly after they open and sterically blocks them (10-15). Thepresent findings add to previous studies (for review, see ref. 33)demonstrating the usefulness of rapid chemical kinetic tech-niques developed to elucidate the underlying mechanisms ofneurotransmitter receptors in living cells (23-25, 33). Thesetechniques may be useful in answering many of the remainingquestions regarding the chemical mechanism by which neuro-transmitter receptors are activated, regulated, and inhibited bya large number of clinically interesting compounds (34), theaction of which, on the molecular level, is not well understood.

T 0D 0U +I I~~~~~~

TI

1'.

rP

Pharmacology: Niu et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

1, 2

021

Proc. Natl. Acad. Sci. USA 92 (1995)

APPENDIXThe relationship between the ratios of maximum currentamplitudes in absence, Imax, and presence, Imax(,o), of aninhibitor and inhibitor concentration, IO, has been derived (4):

1max 1 + Io[FA + FAL + FAL2 + FAL2]Imax(Io) KI KI'

Io

KI(app) [1(i)]

KI(app) is the apparent dissociation constant of inhibitor; FA,FAL, and FAL2 represent the fraction of the receptors in theclosed-channel forms; and FAL2 represents the fraction in theopen-channel forms. At low ligand concentrations, KI(app) isconsidered to reflect KI, and at high ligand concentrations, itreflects KJ. The maximum observed current amplitude inpresence of cocaine is given by:

Imax(I,) =

maxL[KIKC + Io(Kc + 1)]

(L+K1)2'D + L2[1 + IoKK )][KK+ I + 1)]

[I(ii

KI and KI have been defined above and KC = kb/lk (see Fig. 1).The following assumptions were made in deriving the equa-

tions per.taining to the reaction scheme in Fig. 1: (i) Theconcentrations of carbamoylcholine and cocaine are muchgreater than the concentration of receptor sites in the mem-brane. (ii) Receptor forms A, AL, AL2, IA, IAL, and IAL2(Fig. 1) are in rapid preequilibrium, and the formation of theopen receptor-channel forms (AL2 and IAL2) and of thereceptor-inhibitor complexes IA*L, that cannot form openchannels is rate limiting. (iii) Only a single inhibitor-bindingsite is present on the closed- and open-channel forms of thereceptor. (iv) A single equilibrium constant, Kc, accounts forthe interconversion from receptor forms with inhibitor boundthat can lead to open channels to inactive receptor forms thatcannot. Consequently, the observed slow formation of theinactive receptor forms reflects the transition from IAL, toIA*L, and IAL2 to IA*L2 (Fig. 1). This assumption is neitherrequired nor excluded by the experiments. (v) kop kop (Figs.1 and 4B). (vi) kop, kci, k1l >> k', k' (Fig. 1). With theseassumptions, the differential equations pertaining to the re-action scheme in Fig. 1 are readily solved. The integrated rateequation (Eq. II) is given in the text.

We are grateful to Lisa Miller for preparing the manuscript andSusan Coombs for editing it. This work was supported by a grant(NS08527) awarded to G.P.H. by the National Institutes of HealthInstitute for Neurological Diseases and Stroke.

1. Gawin, F. H. (1991) Science 251, 1580-1586.2. Redda, K. K. & Walker, C. A. (1993) in Cocaine, Marijuana,

Designer Drugs: Chemistry, Pharmacology, and Behavior, eds.Redda, K. K., Walker, C. A. & Barnett, G. (CRC, Boca Raton,FL), pp. 1-3.

3. Lathers, C. M., Tyau, L. S. Y., Spino, M. M. & Agarwal, I. (1988)J. Clin. Pharmacol. 28, 584-593.

4. Karpen, J. W. & Hess, G. P. (1986) Biochemistry 25, 1777-1785.5. Swanson, K. L. & Albuquerque, E. X. (1987) J. Pharmacol. Exp.

Ther. 243, 1202-1210.6. Heidmann, T. & Changeux, J.-P. (1978) Annu. Rev. Biochem. 47,

317-357.7. Sharkey, J., Ritz, M. C., Schenden, J. A., Hanson, R. C. & Kuhar,

M. J. (1988) J. Pharmacol. Exp. Ther. 246, 1048-1052.8. Flynn, D. D., Vaishnav, A. A. & Mash, D. C. (1992) Mol. Phar-

macol. 41, 736-742.9. Lerner-Marmarosh, N., Carroll, F. L. & Abood, L. G. (1995) Life

Sci. 56, 67-70.10. Adams, P. R. (1976) J. Physiol. (London) 260, 531-532.11. Neher, E. & Steinbach, J. H. (1978) J. Physiol. (London) 277,

153-176.12. Adler, M., Albuquerque, E. X. & Lebeda, F. J. (1978) Mol.

Pharmacol. 14, 514-529.13. Ogden, D. C. & Colquhoun, D. C. (1985) Proc. R. Soc. London

B 225, 329-355.14. Galzi, J.-L., Revah, F., Bessis, A. & Changeux, J.-P. (1991)Annu.

Rev. Pharmacol. 31, 37-72.15. Lester, H. A. (1992) Annu. Rev. Biophys. Biomol. Struct. 21,

267-292.16. Adams, P. R. (1977) J. Physiol. (London) 268, 291-318.17. Tiedt, T. N., Albuquerque, E. X., Bakry, N. M., Eldefrawi, M. E.

& Eldefrawi, A. T. (1979) Mol. Pharmacol. 16, 909-921.18. Neher, E. (1983) J. Physiol. (London) 339, 663-678.19. Gage, P. W. & Wachtel, R. E. (1984) J. Physiol. (London) 346,

331-339.20. Papke, R. L. & Oswald, R. E. (1989)J. Gen. Physiol. 93, 785-811.21. Hess, G. P., Cash, D. J. & Aoshima, H. (1979) Nature (London)

282, 329-331.22. Katz, B. & Thesleff, S. (1957) J. Physiol. (London) 138, 63-80.23. Niu, L. & Hess, G. P. (1993) Biochemistry 32, 3831-3835.24. Milburn, T., Matsubara, N., Billington, A. P., Udgaonkar, J. B.,

Walker, J. W., Carpenter, B. K., Webb, W. W., Marque, J., Denk,W., McCray, J. A. & Hess, G. P. (1989) Biochemistry 29, 49-55.

25. Matsubara, N., Billington, A. P. & Hess, G. P. (1992) Biochem-istry 31, 5507-5514.

26. Hamill, 0. P., Marty, A., Neher, E., Sakmann, B. & Sigworth,F. J. (1980) Pflugers Arch. 391, 85-100.

27. Schubert, D., Harris, A. J., Devine, E. E. & Heinemann, S. (1974)J. Cell Biol. 61, 398-402.

28. Sine, S. M. & Taylor, P. (1979) J. Biol. Chem. 254, 3315-3325.29. Krishtal, 0. A. & Pidoplichko, V. I. (1980) Neuroscience 5,

2325-2327.30. Udgaonkar, J. B. & Hess, G. P. (1987) Proc. Natl. Acad. Sci. USA

84, 8758-8762.31. Niu, L., Grewer, C. & Hess, G. P. (1995) in Techniques in Protein

Chemistry VII, ed. Marshak, D. R. (Academic, San Diego), inpress.

32. Hammes, G. G. (1982) Enzyme Catalysis and Regulation (Aca-demic, New York).

33. Hess, G. P. (1993) Biochemistry 32, 989-1000.34. Gilman, A. G., Rall, T. W., Nies, A. S. & Taylor, P., eds. (1992)

The Pharmacological Basis of Therapeutics (Pergamon, NewYork), 8th Ed.

12012 Pharmacology: Niu et al.

Ol

Dow

nloa

ded

by g

uest

on

Janu

ary

1, 2

021