Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The nicotinic acetylcholine receptor (nAChR) is the paradigm of the neurotransmitter-gated ion channel superfamily. The nAChR is a pentamer of homologous subunits with composition ₂ or ₂ in embryonic or adult muscle, respectively. The nAChR exists in different functional states: resting, active, and desensitized. The transitions between these states are affected by agonists and competitive antagonists acting at the acetylcholine (ACh) binding sites and by a broad class of noncompetitive inhibitors (NCIs). These inhibitors decrease the probability of channel opening by different general mechanisms: steric blockade of the ion pore, allosteric inhibition of the protein, or enhancement of desensitization.

The acridine derivative quinacrine has been shown to act as an NCI of the nAChR. However, its mechanism of action is still a matter of controversy. Both open-channel blockade and allosteric inhibition of ion flux have been reported (Adams and Feltz, 1980; Valenzuela et al., 1992; Tamamizu et al., 1995; Arias, 1997). In addition, different locations for the quinacrine binding site on the nAChR have been proposed (see review in Arias, 1998). Voltage clamp experiments suggested that quinacrine binding site is located within the ion pore. However, fluorescence quenching and energy transfer studies showed that quinacrine binding sites exist at the lipid-protein interface (Valenzuela et al., 1992; Arias et al., 1993). Arginine209 and Proline211, located at the N-terminal of the M1 transmembrane domain, have been specifically photolabeled by quinacrine azide (Cox et al., 1985; DiPaola et al., 1990). Measurements of currents evoked by ACh on Xenopus laevis oocytes expressing mutant Torpedo californica nAChRs showed that mutations at R209, P211, and Y213 of Torpedo nAChR change the sensitivity to quinacrine (Tamamizu et al., 1995). The M1 domain seems closely associated with both the ion conducting pathway and the lipid bilayer, as revealed by substituted cysteine accessibility and labeling by hydrophobic reagents (Blanton and Cohen, 1994; Akabas and Karlin, 1995). Applying the substituted-cysteine accessibility method to the M1 domain, Akabas and Karlin (1995) have shown that some residues of the N-terminal third of M1 contribute to the lining of the pore. Residue P211 is accessible to the hydrophilic reagent only in the absence of ACh.

Here we investigate the mechanistic bases for the noncompetitive action of quinacrine. We describe for the first time the kinetic changes in nAChR at the single channel level and examine the kinetic properties of macroscopic currents. In addition, we compare the effects of quinacrine between adult and embryonic nAChRs. We conclude that inhibition of nAChR by quinacrine is coincident with an increase in the desensitization rate together with stabilization of a desensitized state, with no significant increase in agonist binding affinity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of nAChR. Mouse cDNAs were subcloned into the cytomegalovirus-based expression vector pRBG4 (Bouzat et al., 1994). HEK293 cells were transfected with , , , and  cDNA subunits using calcium phosphate precipitation at a subunit ratio of 2:1:1:1, respectively, essentially as described previously (Bouzat et al., 1994, 1998). For transfections, cells at 40 to 50% confluence were incubated for 8 to 12 h at 37°C with the calcium phosphate precipitate containing the cDNAs in DMEM plus 10% fetal bovine serum. Cells were used for patch clamp recordings 1 or 2 days after transfection.

Patch-Clamp Recordings. Recordings were obtained in the cell-attached and outside-out patch configurations at 20°C (Hamill et al., 1981). For cell-attached patch recordings, patch pipettes were pulled from 7052 capillary tubes (Garner Glass, Claremont, CA) and coated with Sylgard (Dow Corning, Midland MI). The pipette resistances ranged from 5 to 7 M. The bath and pipette solutions contained 142 mM KCl, 5.4 mM NaCl, 1.8 mM CaCl₂, 1.7 mM MgCl₂ and 10 mM HEPES, pH 7.4. ACh at final concentrations of 1, 10, or 60 µM and quinacrine at final concentrations of 1, 10, 30, or 60 µM were added to the pipette solution. Channels were typically recorded at a membrane potential of 70 mV. For studying the voltage dependence of the effect of quinacrine on nAChR, channels were also recorded at a membrane potential of 40 and 90 mV and +70 mV. Single-channel currents were recorded using an Axopatch 200 B patch-clamp amplifier (Axon Instruments, Inc., CA) and digitized at 94 kHz with a PCM adapter (VR-10B; InstruTECH, Port Washington, NY). Data were transferred to a computer using the program Acquire (Bruxton Corporation, Seattle, WA) and detected by the half-amplitude threshold criterion using the program TAC 3.0 (Bruxton Corporation) at a final bandwidth of 10 kHz. Open- and closed-time histograms were plotted using a logarithmic abscissa and a square root ordinate and fitted to the sum of exponential functions by maximum likelihood using the program TACFit (Bruxton Corporation). Clusters of openings corresponding to a single channel were identified as a series of closely spaced events preceded and followed by closed intervals greater than a specified duration (_crit); this duration was taken as the point of intersection of the predominant closed-time component and the succeeding one in the closed-time histogram. For nAChRs activated by 60 µM ACh, the major intermediate component associated to dwell times within clusters typically varied between 0.5 and 1 ms and _crit between 5 and 8 ms. Cluster duration histograms were constructed and fitted by using the program TACfit setting the burst resolution to the calculated _crit. Open probability within clusters (P_open) was experimentally determined at each ACh concentration by calculating the mean fraction of time the channel is open within a cluster.

Ligand Binding Measurements. Binding of ACh was measured by competition against the initial rate of [¹²⁵I]-bungarotoxin ([¹²⁵I]-BTX) binding as described previously (Sine and Taylor, 1979; Sine et al., 1994) and compared with binding in the presence of quinacrine or proadifen. HEK cells expressing adult nAChRs were resuspended in high potassium Ringer's solution in the absence or presence of quinacrine or proadifen and divided into aliquots for ligand binding measurements. Cells were first incubated for 30 min with different concentrations of ACh; [¹²⁵I]-BTX was subsequently added to a final concentration of 5 nM, and the cells were incubated for an additional 20 min to allow occupancy of no more than 50% of the binding sites by -BTX (Sine et al., 1994). The total number of binding sites was determined by incubating cells with 5 nM [¹²⁵I]-BTX for 2 h in the absence of ACh. Binding was terminated by the addition of potassium Ringer's solution containing 20 mM carbamylcholine. Nonspecific binding was determined in the presence of 20 mM carbamylcholine. Rates of -BTX binding in the absence and presence of competing ligand were calculated from binding measured at 20 and 120 min (Fu and Sine, 1996). These rates are related to ligand occupancy by k_obs = k_max (1  Y), where k_obs is the rate of toxin binding in the presence of a specified concentration of competing ligand, k_max is the rate in the absence of competing ligand, and Y is the occupancy function for the competing ligand, given by the Hill equation. Fractional occupancy by ACh was fitted by the Hill equation: 1  fractional occupancy = [1 / (1 + ([ACh] / K_d)^n_H], where K_d is the apparent dissociation constant and n_H is the Hill coefficient.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Quinacrine on Single-Channel Currents

Quinacrine Reduces the Frequency of nAChR Opening Events. Adult nAChR channels were recorded 1 and 8 min after seal formation from cell-attached patches using a pipette in which the tip was filled with 1 µM ACh alone and the shaft was filled with 1 µM ACh and 10 µM quinacrine. Under these conditions, rapid sealing allowed us to record channel activity in the absence and presence of quinacrine on the same patch. During the first minute, the channel activity is similar to that observed in the absence of quinacrine, with opening events typical of adult muscle nAChRs (Bouzat et al., 1994, 1998). After a few minutes, a dramatic decrease in channel activity is observed because of the diffusion of quinacrine to the tip of the pipette (Fig. 1a). The frequency of opening events decreases more than 90% at 10 µM quinacrine (Fig. 1b). The decrease in the frequency of channel opening is seen in the closed-time histogram as a displacement of the predominant closed component to longer durations (13.3 ms and 76.3 ms for t = 1 min and t = 8 min, respectively; Fig. 1c). Open-time distributions of nAChRs recorded in the presence of 10 µM quinacrine are similar to control histograms (Fig. 1c, 1.07 ms and 870 µs for t = 1 min and t = 8 min, respectively). Quinacrine does not affect the channel amplitude; the amplitudes were 5.2 ± 0.2 pA and 5.1 ± 0.2 pA for t = 0 and t = 8 min, respectively (n = 5).

View larger version (36K): [in this window] [in a new window]   Fig. 1.   Effects of quinacrine on nAChR activity. a, effects of quinacrine on the frequency of opening events. nAChR channels were recorded using a pipette containing 1 µM ACh in the tip and 1 µM ACh plus 10 µM quinacrine in the shaft. Channels were recorded immediately after seal formation (t = 1; top) and 8 min later (t = 8, bottom). Membrane potential: 70 mV. Inward currents are plotted upwards. Filter 10 kHz. b, decrease of the frequency of opening events in the presence of quinacrine. f is the ratio between the frequency of opening events (events per second) recorded 8 min (t = 8) and 1 min (t = 1) after seal formation. n = 4 at the specified quinacrine concentration. Data are shown as mean ± S.D. c, open- and closed-time histograms in the absence and presence of quinacrine. 0 µM, histograms corresponding to nAChR channels recorded during the first minute after rapid sealing (1 µM ACh). 10 µM, histograms corresponding to channels recorded during minute 8 after seal formation (1 µM ACh plus 10 µM quinacrine in the shaft of the pipette).

Quinacrine Reduces the Duration of Clusters of Single-Channel Currents. To evaluate the influence of quinacrine on channel activation, clear clusters of events corresponding to a single channel were activated with 60 µM ACh. Each activation period (cluster) begins with the transition of a single receptor from the desensitized to the activatable state and terminates by returning to the desensitized state (Fig. 2, left). In these experiments, control and quinacrine recordings were performed on separate cell-attached patches.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Clusters of single-channel currents in the presence of quinacrine. Left, channel traces recorded in the absence and presence of 10 and 30 µM quinacrine in the pipette solution. ACh, 60 µM. Membrane potential: 70 mV. Filter, 10 kHz. Right, closed-, cluster-, and open-duration histograms corresponding to nAChRs recorded at the given quinacrine concentrations.

Dependence of Mean Cluster Duration, Openings per Cluster, Mean Open Time, and Mean Closed Time on Quinacrine Concentration. The dependence of channel kinetics on the concentration of quinacrine is shown in Fig. 3. The main effect of quinacrine is a dose-dependent reduction in the duration of clusters (Fig. 3a); this decrease is about 95% at 30 µM quinacrine. The decrease in the number of openings per cluster parallels that of cluster duration (Fig. 3b), indicating that the decrease in cluster duration is caused mainly by a decrease in the number of openings. The mean open time shows a slight concentration-dependent decrease that is statistically significant in the presence of quinacrine concentrations higher than 30 µM (p < 0.001) (Fig. 3c). The duration of the intermediate component of the closed-time histogram, corresponding to closings within clusters, remains stable up to 10 µM quinacrine but increases 3-fold at 60 µM (Fig. 3d). The relative area of this component decreases as a function of quinacrine concentration (Fig. 3d).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Dependence of mean cluster duration (a), openings per cluster (b), mean open time (c), and mean closed time (d) on quinacrine concentration. Clusters of openings corresponding to a single channel were identified as a series of closely spaced events preceded and followed by closed intervals greater than a specified duration; this duration was taken as the point of intersection of the predominant closed time component and the succeeding one in the closed time histogram. The mean open time was obtained from the corresponding open-time histograms. The closed time and its relative area corresponds to the intermediate closed component of the closed time histogram. Data are shown as mean ± S.D. for three to eight experiments for each condition.

Effects of Quinacrine on nAChR as a Function of Membrane Potential. To investigate whether the action of quinacrine on the nAChR is voltage-dependent, we recorded nAChR channels activated by 60 µM ACh in the absence and presence of 10 µM quinacrine at different membrane potentials. The mean open time decreases exponentially when the membrane is depolarized for both control and quinacrine-treated channels (Fig. 4). The mean open times change e-fold per 89 and 94 mV for control and treated nAChRs, respectively. Thus the magnitude of such decrease is similar in both cases, suggesting that quinacrine does not change the intrinsic voltage dependence. The dominant effect of quinacrine is a reduction in cluster duration. Figure 4 shows that in the presence of quinacrine, the relationship between cluster duration and membrane potential also parallels that of the control channels.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effects of quinacrine on nAChR as a function of membrane potential. nAChRs activated by 60 µM ACh were recorded in the absence (open symbols) or presence of 10 µM quinacrine (solid symbols) at different membrane potentials. Dependence of the mean open time (circles and solid line) and cluster duration (squares and dotted line) on the membrane potential. The voltage dependence of the mean open-time constant is fitted by a simple exponential decay. Data were obtained from the corresponding histograms and are expressed as the mean ± S.D. of 4 to 10 different patches.

Dependence of Cluster Duration as a Function of ACh Concentration. Mean cluster duration decreases with the ACh concentration, showing values of 224 ± 10 (n = 3), 120 ± 50 (n = 3), and 60 ± 10 ms (n = 10) at 10, 30, and 60 µM ACh. The decrease in cluster duration by 10 µM quinacrine does not change with ACh concentration. This decrease is of 76 ± 7% (n = 3), 88 ± 4% (n = 3), and 85 ± 5% (n = 10) at 10, 30, and 60 µM ACh, respectively.

Effects of Quinacrine on Macroscopic nAChR Currents.

Quinacrine Increases Current Decay Due to Desensitization. To determine the overall consequences of quinacrine on nAChR activation, we studied the effect of the drug on outside-out patches rapidly perfused with 300 µM ACh. Figure 5a shows ensemble currents obtained from a single patch exposed to brief applications of ACh alone (control) and together with different concentrations of quinacrine. In control data, the current reaches the peak after 0.1 to 1 ms and then decays with a time constant (_d) of 44 ± 16 ms because of desensitization. When quinacrine is present in both the ACh-free and ACh-containing solutions, a concentration-dependent increase in the decay rate is observed (Fig. 5a).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of quinacrine on the decay of macroscopic currents activated by 300 µM ACh. a, ensembled mean currents obtained from rapid perfusion of 300 µM ACh alone or with different concentrations of quinacrine onto an outside-out patch during 200 ms. Membrane potential of 50 mV. Each trace represents the average of 12 to 20 applications of agonist. Curves from right to the left correspond to: control (three curves), 1, 6, 10, and 30 µM quinacrine. The calculated decay time constants (d) are 26.3, 26.9, and 29.1 ms for control curves and 8.3, 2.5, 1.7, and 0.86 for 1, 6, 10, and 30 µM quinacrine, respectively. b, dependence of d on quinacrine concentration. Each point represents the mean ± S.D. of 3 to 13 measurements.

Quinacrine Decreases the Rate of Recovery from Desensitization. The rate of recovery from desensitization was studied by using a two-pulse protocol (Dilger and Liu, 1992). After a 300-ms application of 300 µM ACh, about 90% of the channels are desensitized (Fig. 6a). The degree of recovery increases with the interval between ACh applications and more than 80% recovery is reached within 435 ms (Fig. 6a top). Recovery is dramatically slower in the continuous presence of 1 µM quinacrine (Fig. 6a, bottom).

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Recovery from desensitization in the presence of quinacrine. a, superimposed pair of current responses to 300 µM ACh obtained using a double agonist application protocol with a variable interpulse time interval (15-435 ms). The first application lasts 300 ms and the second one is applied during 30 ms. Experiments were performed in the absence (top traces) and continuous presence of 1 µM quinacrine (bottom traces). Membrane potential, 50 mV. b, time course of recovery from desensitization. The plot show the relation of the peak current ratio with the interval between the end of the first and the beginning of the second ACh application (interval duration) for control (), 1 (), 3 (), and 6 µM () quinacrine. The peak current ratio was calculated as the ratio of the peak current induced by the second pulse with respect to the first peak current.

Effects of Quinacrine Application Protocol on Macroscopic Current Desensitization. In the macroscopic current experiments described thus far, patches were equilibrated with quinacrine before application of the ACh + quinacrine solution (+/+ protocol). When quinacrine is omitted from the preincubation solution (/+ protocol), the effect of quinacrine on desensitization is less pronounced (Fig. 7). In this example, the control value of _d is 41.3 ms and the value of _d in the constant presence of 3 µM quinacrine is 4.7 ms (+/+ protocol). However, when 3 µM quinacrine is applied simultaneously with ACh, _d is 14.9 ms (/+ protocol). Although quinacrine has no obvious effect on closed channels, it seems that previous incubation of the patch with quinacrine is necessary for its full action. We also looked at the effect of preincubation of the patch with quinacrine followed by application of ACh alone (+/ protocol, Fig. 7). The effect of quinacrine remains (_d = 6.3 ms) even though the drug is quickly removed from the aqueous solution around the patch during the +/ protocol.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Effects of quinacrine application protocol on macroscopic current desensitization. Superimposed current responses to 300 µM ACh and 3 µM quinacrine applied using different protocols. a, from right to left, curves correspond to control conditions (/ protocol), simultaneous ACh/quinacrine application without preincubation with quinacrine (/+ protocol), and simultaneous ACh/quinacrine application with previous incubation with quinacrine (+/+ protocol). b, from right to left, curves correspond to control conditions (/ protocol), ACh application after the preincubation with quinacrine (+/ protocol) and simultaneous ACh/quinacrine application with previous incubation with quinacrine (+/+ protocol).

Inhibition of Embryonic-Type nAChR by Quinacrine. To determine quinacrine inhibition of embryonic nAChRs, we measured macroscopic currents activated by 300 µM ACh on BC3H-1 cells in the absence and presence of different quinacrine concentrations. As described for adult nAChRs, all current decays were fitted by a single exponential function. In the absence of quinacrine, embryonic nAChR desensitization rate is similar to that of adult nAChR (Table 1). At high quinacrine concentrations, decay time constants differ slightly from those of -containing nAChRs (Table 1). Thus,  or  subunits are not the main subunits involved in the quinacrine binding site.

View this table: [in this window] [in a new window]   TABLE 1 Influence of quinacrine concentration and subunit composition on current decay The table shows the decay time constants (d) of currents obtained from outside-out patches exposed to 300 µM ACh containing adult (2) and embryonic (2) nAChRs. Values are expressed as the mean ± S.D. Number of experiments for each condition are shown in parentheses. Membrane potential, 50 mV.

Effects of Quinacrine on Equilibrium Binding of ACh. To determine whether quinacrine introduces changes in equilibrium agonist binding, we compared the inhibition of -BTX binding by ACh in the absence and presence of 10 µM quinacrine. As shown in Fig. 8, similar profiles were obtained in both cases. The apparent affinity constants (K_d) and Hill coefficients (n_H) are shown in Table 2. The calculated values for control conditions are in good agreement with those reported previously (Sine et al. 1994; Prince and Sine, 1999). Thus, although 10 µM quinacrine dramatically enhances desensitization as measured in electrophysiology experiments, no changes in the equilibrium binding of ACh are observed. Increasing quinacrine concentration from 10 to 60 µM leads to a modest shift of the curve that is statistically insignificant (Table 2). Previous work established that some noncompetitive inhibitors convert the nAChR to a state in which the affinity coincides with that of the desensitized state (Sine and Taylor, 1982). We therefore measured ACh binding in the presence of a saturating concentration of the noncompetitive inhibitor proadifen (Prince and Sine, 1998). As expected, proadifen shifts the binding curve to lower ACh concentrations and decreases the Hill coefficient to unity (Fig. 8, Table 2; Sine and Claudio, 1991; Prince and Sine, 1998). When cells are preincubated with 60 µM proadifen plus 10 µM quinacrine there is no additional shift in the binding curve (Fig. 8, Table 2).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Effects of quinacrine on equilibrium binding of ACh. ACh binding was measured by competition against the initial rate of -bungarotoxin binding in the absence () or presence () of 10 µM quinacrine, 60 µM quinacrine (), 60 µM proadifen (), and 60 µM proadifen plus 10 µM quinacrine (). The curves are fits to the Hill equation. Each point represents the mean ± S.D. of three to eight different experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Here we identify the mechanistic bases for the noncompetitive action of quinacrine: increase in the rate of desensitization together with stabilization of a desensitized state. At the single-channel level, the main effect of quinacrine is a profound decrease in the frequency of openings. Analysis of single channels activated by high concentrations of ACh permits a better understanding of the mechanistic bases of quinacrine action. Under these conditions, single-channel openings are clustered into unambiguous activation periods (Sakmann et al., 1980). A cluster starts when one nAChR recovers from desensitization and continues with the receptor undergoing cycles of agonist association/dissociation and channel gating. Auerbach and Akk (1998) demonstrated that the value of (_cP_o)¹ (where _c is the mean cluster duration and P_o is the probability of being open within a cluster) is a direct measure of the rate constant of desensitization. Under control conditions with 60 µM ACh, we find P_o = 0.7 and _c = 61 ± 10 ms, so that (_cP_o)¹ = 24 ± 5 s¹. This agrees with desensitization rates calculated from the decay of macroscopic currents (Dilger and Liu, 1992; Franke et al., 1993). The duration of clusters decreases in the presence of quinacrine, mainly because of a decrease in the number of successive openings. The early termination of clusters suggests that quinacrine increases the desensitization rate. Between 1 and 10 µM quinacrine, P_o is similar to control values (0.7), but there is a slight decrease at 30 µM quinacrine (P_o = 0.55). The desensitization rate, estimated by the product (_cP_o)¹, is 122 ± 50, 280 ± 80, and 1190 ± 300 s¹ in the presence of 1, 10, and 30 µM quinacrine, respectively.

Conclusions obtained from single-channel recordings are in good agreement with those deduced from macroscopic currents activated by 300 µM ACh. The decay time constant (_d) for desensitization varied between 28 and 60 ms. At 300 µM ACh Popen > 0.9 and the fraction of current remaining after desensitization is less than 1% of the peak current. Thus, we can assume that 1/_d equals the rate of desensitization from the double liganded open state, being its value of 26 ± 8 s¹, and the reopening rate of desensitized nAChRs is very low (<0.3 s¹). The desensitization rate thus calculated is in good agreement with that estimated from our single-channel data. Quinacrine increases the rate of desensitization to 110 ± 30, 220 ± 40, 370 ± 30, 560 ± 80, and 1100 ± 100 s¹ at concentrations of 1, 3, 6, 10, and 30 µM, respectively. Thus, information obtained from both single channel and macroscopic current recordings indicates that quinacrine profoundly increases the rate of desensitization in a concentration-dependent manner.

In the presence of quinacrine, the extrapolated peak current is similar to that of control, suggesting that quinacrine has a strong preference for interacting with the open state of the nAChR. This is consistent with the idea that quinacrine accelerates desensitization, a process that proceeds mainly from the doubly liganded open state (Dilger and Liu, 1992; Auerbach and Akk, 1998). Experiments in which quinacrine azide is used to photolabel T. californica nAChR also indicate that the drug binds preferentially to the open state (Johnson and Ayres, 1996).

Open channel block was the mechanism proposed for the action of quinacrine at the frog endplate nAChR (Adams and Feltz, 1980). The open channel blocking mechanism predicts that the decay should be biexponential (Dilger et al., 1997; Forman, 1999). However, we found that a single exponential function is always adequate to describe the current decay in the presence of quinacrine. This is true even for macroscopic currents activated by 60 or 10 µM ACh, for which the desensitization decay is slower. We conclude that open channel block is unlikely to be the cause of inhibition by 30 µM quinacrine. We were unable to extend the concentration range because macroscopic currents in the presence of 60 µM quinacrine were almost undetectable. Single-channel recordings in the presence of 60 µM quinacrine show a very low number of openings, but analysis of clusters suggests a slight increase in the duration of closings within clusters (Fig. 3). Because of the few openings, the duration of openings and closings within clusters at high quinacrine concentrations is less well determined than at lower concentrations. The increase in the duration of closings within clusters, together with a decrease in the mean open time could be associated with an open-channel blockade. Thus, at quinacrine concentrations higher than 30 µM, both mechanisms, increased desensitization and open channel blockade, may occur. Multiple sites of action have been described for other noncompetitive inhibitors (Spitzmaul et al., 1999). We speculate that for quinacrine, the relative occupancy of at least two sites is concentration-dependent and may also vary with the type of nAChR.

By using a two-pulse protocol, we determined the rate of recovery from desensitization in the absence of ACh, which involves agonist dissociation and return to the closed, resting state (Dilger and Liu, 1992). Recovery is slower in the presence of quinacrine, as evidenced by the 6-fold decrease in recovery rate at 6 µM quinacrine. Therefore, in addition to increase the rate of the nAChR to reach the desensitized state, quinacrine makes this state more stable.

During continued exposure to agonist, nAChR undergoes a slow conversion to a high affinity state that correlates closely with the extent of desensitization. To gain an overall mechanistic picture of the action of quinacrine, we evaluated changes in equilibrium agonist binding. The apparent affinity of nAChRs previously equilibrated with agonist reflects a weighted average of the affinity of the agonist for the two forms of the receptor, low and high affinity, as well as the fraction of receptor in each state (Sine and Taylor, 1979). Because electrophysiological data show that quinacrine profoundly increases desensitization, we expected a shift of the curve to lower agonist concentration as observed with anesthetics (Sine and Taylor, 1982; Sine et al., 1995). Surprisingly, neither 10 µM nor 60 µM quinacrine affected the equilibrium between high and low affinity states. We also determined binding at equilibrium in the presence of proadifen. For the adult muscle nAChR, the limiting shift in affinity has been shown to occur at 30 to 60 µM proadifen (Prince and Sine, 1998). In contrast to quinacrine, 60 µM proadifen produced the expected decrease in K_d value.

An explanation for our results is that quinacrine stabilizes a conformational intermediate state of the nAChR; one that exhibits the permeability but not the agonist binding properties of desensitization. This intermediate could represent either a short-lived state that cannot be distinguished in the absence of quinacrine or a novel state induced by quinacrine.

Scheme 1 is a general model in which each state of channel activation and block is considered to have a desensitized counterpart (Dilger and Liu, 1992):
Scheme 1.

Our results from single channel-data suggest that d₊₄ in Scheme 1 is one of the main kinetic steps affected by quinacrine. However, equilibrium binding studies suggest that quinacrine stabilizes an intermediate conformation of the transition from resting to desensitized states. This conformational intermediate (A₂I) could be either in the A₂R* to A₂D* path or in a branched path in which both A₂I and A₂D are connected to A₂R*.

In accordance with our results, conformational intermediates between the resting and desensitized states of the nAChR have recently been described. Ryan et al. (2001), with the use of infrared difference spectroscopy, showed that domains of the nAChR interconvert between the resting and desensitized states independently of each other and that binding to the noncompetitor inhibitor binding site may lead to the formation of a conformation that is structural intermediate between both states. Further studies using more specific techniques (Ryan et al., 2001) will be required to confirm our explanation for quinacrine's action.

Changes in the solution application protocol show that the effect of quinacrine develops and recovers slowly (i.e., between 50 ms and 5 s). One interpretation is that quinacrine binds to the nAChR closed state and dissociates slowly from this state but does not exert its effect until the channel opens. Early work from Grunhagen and Changeux (1976) used quinacrine to monitor structural changes that take place upon binding of cholinergic ligands. Interestingly, a slow equilibration process of quinacrine-treated membranes after addition of carbamylcholine was observed. The kinetics of the slow phase were interpreted to represent the time course of transitions in the nAChR, which in turn affect the energy transfer to quinacrine. Alternatively, quinacrine, because of its hydrophobicity, may require >10 ms to equilibrate with the patch membrane. In this scenario, quinacrine would have access from the membrane to its inhibitory binding site on the nAChR. Similar explanations have been proposed for the slow kinetics of action of 3-(trifluoromethyl)-3-(m-iodophenyl)diazirine (Forman, 1999).

Arginine209 and Proline211 of the M1 transmembrane domain have been specifically photolabeled by quinacrine azide (Cox et al., 1985; DiPaola et al., 1990). The disposition and functional role of the M1 transmembrane domain remain uncertain. Our results reveal that M1 may contribute to the desensitization process.