Introduction

Top Abstract Introduction Materials and Methods Results Discussion Appendix References

In recent years, there has been increased support for the idea that general anesthetics can act directly on membrane proteins, especially ion channels (Mihic et al., 1997; Franks and Lieb, 1998). The muscle-type nicotinic acetylcholine receptor (AChR) is the site of action of muscle relaxants and may play a role in relaxant effects of some anesthetics. In addition, this receptor is a useful prototype for studying the effects of anesthetics on ligand-gated channels (Barann et al., 1998; Raines and Zachariah, 1999, 2000).

Previously, we described how the volatile anesthetic isoflurane causes AChR channel activity to occur in bursts of briefer-than-normal openings (Dilger et al., 1992). We showed how this behavior was well described by a model in which isoflurane blocks both the open and closed states of the receptor pore (Dilger et al., 1993). We also found evidence for cooperative interactions between the channel-blocking local anesthetic QX-222 and diethyl ether, the parent compound of isoflurane (Dilger and Vidal, 1994). Forman and his colleagues (Forman et al., 1995; Forman, 1997) studied site-directed mutants of the AChR in which polar amino acids in the second membrane-spanning region of the receptor (M2) were replaced with large hydrophobic residues. These mutations affected the affinity of the receptor for isoflurane and some alcohols. This was interpreted as evidence for anesthetic binding within the pore of the AChR. However, the kinetic predictions of the model were not tested in this study. It was also observed that the mutations affected channel desensitization. The consequences of this finding for the interpretation of the data were not completely addressed.

In this article, we consider more conservative mutations at two sites within the pore of the channel. We studied wild-type (₂, WT) and two mutant receptors [₂S10'AT10'A (10') and ₂S6'AS6'A (6')] (Fig. 1). In the experiments conducted by Charnet et al. (1990), the binding of QX-222 was found to be affected by mutations at both the 10' and 6' regions of M2. Polar-to-nonpolar mutations at 10' resulted in tighter binding of QX-222. In contrast, polar-to-nonpolar mutations at 6' resulted in less tight binding of QX-222. The authors reasoned that the amphipathic QX-222 molecule binds with its charged nitrogen near the 6' level and its ring moiety near 10'. Our hypothesis is that the hydrophobic general anesthetic isoflurane binds near 10'. Thus, inhibition by isoflurane will be sensitive to mutations at 10' but not at 6'.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   The amino acid sequences of the M2 region of the four subunits of the embryonic muscle AChR.

In any study involving site-directed mutagenesis of proteins, it is essential to determine whether the mutation is simply causing a fairly localized change in the protein or a global change that affects multiple functions of the protein. The first part of this study concerns the behavior of WT and the two mutant receptors in the absence of isoflurane. We found that the mutations affect both channel gating and desensitization, and these changes are different for the two mutants. The second part of the study deals with the effects of isoflurane. We find that the 10' mutation increases the sensitivity of the ACh receptor to isoflurane, whereas the 6' mutation does not. These changes are probably not caused by the differential effects of the mutation on channel gating and desensitization. Finally, we found that the increased sensitivity of the 10' receptor to isoflurane is state-dependent; the mutation changes the affinity of the closed state but not the open state of the channel. Preliminary results have been presented in abstract form (Barann et al., 1996).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Appendix References

Mouse nicotinic acetylcholine receptors were expressed transiently in HEK-293 cells (American Type Culture Collection, Manassas, VA) using a calcium-phosphate transfection protocol (Sine, 1993). Cells were maintained in Dulbecco's modified Eagle's medium with 10% fetal calf serum. Solutions of 2× HEPES (275 mM NaCl, 1.5 mM Na₂HPO₄·7H₂O, 55 mM HEPES, pH 7.05) and calcium/DNA (123 mM CaCl₂ plus cDNA for each AChR subunit) were prepared. To transfect three 30-mm dishes, we used 7.5, 3.75, 3.75, and 3.75 µg of cDNA (subcloned into the pRGB4 expression vector) for the  (WT or mutant),  (WT or mutant), , and  (WT or mutant) subunits, respectively, and 1 µg of cDNA for CD8 (gift from Dr. Brian Seed), a T-cell antigen used as a marker. A precipitate was formed by adding 0.5 ml of the calcium/DNA solution dropwise into the 0.5 ml of the 2× HEPES buffer solution. After 30 min, the precipitate was distributed over each dish of cells. After 8 to 12 h, the transfection solution was replaced with Dulbecco's modified Eagle's medium plus 10% fetal calf serum. Experiments were performed on cells between 1 and 4 days after transfection.

We prepared a stock solution of saturated (15 mM; Firestone et al., 1986) isoflurane (1-chloro-2,2,2-trifluroethyl difluoromethyl ether, Forane; Anaquest, Madison, WI) in extracellular solution (ECS) by adding isoflurane to a glass bottle filled with ECS, closing the bottle with a cap providing a Teflon seal, and stirring the solution overnight with a Teflon-coated magnetic stirring bar. The stock solution was then diluted with ECS to obtain the desired concentration of isoflurane. The solution was quickly transferred to a solution reservoir, an intravenous drip bag that was made air-tight with a dialysis bag clamp.

To prepare cells for patch-clamp recording, the culture medium was replaced with an ECS containing 150 mM NaCl, 5.6 mM KCl, 1.8 mM CaCl₂, 1.0 mM MgCl₂, and 10 mM HEPES, pH 7.3. To help identify transfected cells, we added 5 to 10 µl of polysterene beads coated with a monoclonal antibody specific for the CD8 antigen (Dynabeads; Dynal, Lake Success, NY). After 15 min of equilibration with beads, ECS was flowed into the culture dish to remove loose beads. Cells with two or more beads attached (about 10%) were considered likely to express AChR as well as CD8 (Jurman et al., 1994). We found that patches from some of the cells with beads did not have AChR channel activity; perhaps these cells expressed such a low density of channels that most excised patches had no channels. We also used morphological clues to help determine which of those cells with beads also contained AChR channels. These clues included large, spread-out cells, cells with a granular appearance, and cells having areas that appeared as plateaus emerging from the surface.

Patch pipettes were filled with a solution consisting of 140 mM KCl, 5 mM EGTA, 5 mM MgCl₂, and 10 mM HEPES, pH 7.3, and had resistances of 3 to 6 M. An outside-out patch (Hamill et al., 1981) with a seal resistance of 10 G or greater was excised from a cell and moved into position at the outflow of a perfusion system. The perfusion system consisted of solution reservoirs, manual switching valves, a solenoid-driven pinch valve, and two Teflon tubes inserted into the culture dish (Liu and Dilger, 1991). One tube contained ECS without agonist (normal solution); the other contained ECS with a known concentration of ACh (test solution). In the resting position of the pinch valve, normal solution perfused the patch. The solenoid was then triggered to stop the flow of normal solution and start the flow of test solution. After the patch was exposed to the test solution for 0.1 to 0.25 s, the pinch valve was returned to its resting position for several seconds. In this way, we exposed the patch to a series of timed exposures to the agonist-containing solution while minimizing the desensitizing effects of prolonged exposure to ACh. The perfusion system allows a rapid (0.1-1 ms) exchange of the solution bathing the patch. During single-channel recording, we used a similar but slower perfusion system that was adequate for the relatively nondesensitizing conditions of 0.2 µM ACh.

The currents flowing during exposure of the patch to ACh were measured with a patch-clamp amplifier (EPC-7; List-Medical Instruments, Darmstadt, Germany), filtered at 3 kHz (3-db frequency, 8-pole Bessel filter; Frequency Devices, Haverhill, MA), digitized, and stored on the hard disk of a laboratory computer (PDP11; Digital Equipment Corporation, Maynard, MA). Data analysis was performed off-line as described previously (Dilger et al., 1997). Experiments were performed at room temperature (20-23°C).

For macroscopic currents, we recorded current responses (at 50 mV) during 200-ms applications (at 5-s intervals and sampled at 100-200 µs per point) of ECS containing 100 µM ACh (test solution), a concentration that activates about 90% of the AChR channels from BC₃H-1 cells (Dilger and Brett, 1990). We subsequently used this test solution to quantify the loss of channel activity. Both the normal and test perfusion solutions were then switched to isoflurane-containing solutions. Responses of the patch to 100 µM ACh applications during continuous exposure to isoflurane were recorded. The drug-free solutions were then re-introduced, and recovery currents were measured. Data were accepted when the initial and recovery currents changed by less than 10%. Some experiments were performed with other perfusion protocols. We used the following nomenclature: (/) = control (no drug in either normal or test solutions), (+/+) = equilibrium (drug in both normal and test solutions), and (/+) = onset (drug in test solution onlysimultaneous application of agonist and drug).

The ensemble mean current was calculated from 10 to 30 individual current traces. Mean currents were fit to single- or double-exponential functions to obtain peak and steady-state current values and a time constant of the decay caused by desensitization (Dilger and Liu, 1992). The 2-exponential fit was considered better if was at least 3-fold greater than . Relative peak-response amplitudes are calculated as the ratio of the peak current response to application of a given concentration of ACh to that obtained in response to application of 100 µM ACh. Fractional inhibition of the peak mean current, the maximum inward current obtained during perfusion of ACh, was calculated as the ratio of current in the presence of drug to current in the absence of drug.

Single-channel recordings were made while the patch was transiently exposed to ECS containing 0.2 µM ACh at a patch potential of 100 mV. We digitized data in 5-s segments at a rate of 50 µs per point. The patch was perfused with normal ECS for at least 10 s in between recordings. This differs from our previous protocol and was necessary because we found that mutant AChRs desensitize quickly and extensively when exposed to this low concentration of ACh. We obtained enough segments of data to provide 300 to 1000 single-channel events. Data recording was repeated with ECS containing isoflurane and then again with normal ECS (recovery). Data were accepted if, after analysis, channel kinetics during recovery were within 20% of the values obtained during the initial data-collection segments. Data analysis, including correction for missed events, was performed off-line as described previously (Dilger et al., 1997). The kinetic parameters of Scheme 2 (see Results) were determined by fitting the open duration-versus-concentration data to obtain values for  and f (eq. 3). These values were used to determine b/' from the number of openings per burst versus concentration data (eq. 5). Finally, b + ' was determined from the gap duration averaged for all concentrations (eq. 4). Thus, we calculated both b and ' values. Next, 90% confidence limits were obtained from the fits, and these confidence limits were propagated in quadrature, e.g.
Data are expressed as mean ± S.D. Statistical comparisons were made using the Student's t test. A level of p < 0.05 was considered to be significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Appendix References

Comparison of WT and Mutant AChR Currents and Channels. When an outside-out patch is rapidly perfused with 3 to 100 µM ACh, a macroscopic current develops as channels are opened by ACh within less than 1 ms and subsequently desensitizes within tens of ms in the continued presence of ACh. Figure 2 shows examples of macroscopic currents from WT, 10'-mutant, and 6'-mutant AChRs activated by rapid perfusion of 3, 10, and 100 µM ACh. In these examples, both of the mutant AChR currents desensitize more quickly and seem to be more sensitive to ACh than do WT. As was seen previously with the AChR in the clonal murine BC₃H-1 cell line (Dilger and Brett, 1990) and in embryonic receptors expressed in Xenopus oocytes (Bufler et al., 1993), there is a large patch-to-patch variability in the rate of fast desensitization for AChRs expressed in HEK-293 cells (compare currents in Fig. 2 with those in Fig. 4). Table 1 summarizes the data on the desensitization of the three types of receptor by 100 µM ACh. When the decay phase (desensitization) of the current is fit to a single exponential function, the desensitization time constant, , for the 10' mutant receptors is 60% faster than that for WT receptors (p < 0.03); the 26% lower value of  for 6' compared with that for WT receptors was not significant (p > 0.1). The time course of desensitization is sometimes better fit by a 2-exponential decay function. Using the criterion   3 , a 2-exponential fit was better in 15, 47, and 17% of the patches containing WT, 10', and 6' receptors, respectively. Table 1 compares the two time constants, _fast, _slow, and the ratios of the amplitudes of the two components, a_fast/a_slow, for WT and mutant receptors. The 10'-mutant receptors differ from WT in that both components of desensitization are faster. In addition, both components have approximately the same amplitude in 10' receptors, whereas the slow component dominates in WT receptors. Comparisons of these parameters between 6' mutant and WT show no significant differences (p > 0.5). The data for WT receptors are similar to published data for the native fetal mouse AChR in BC₃H1 cells (Dilger and Brett, 1990); these data are included in Table 1.

View larger version (8K): [in this window] [in a new window]   Fig. 2.   Macroscopic current recordings from WT, 6', and 10' AChR in response to rapid perfusion with 100, 10, and 3 µM ACh. The applied potential was 50 mV.

View this table: [in this window] [in a new window]   TABLE 1 Desensitization time constants for native, WT, and mutant AChRs is the time constant of a single-exponential fit to the decay in the macroscopic current response to application of 100 µM ACh at 50 mV. For those cases in which a double-exponential fit was superior (see Materials and Methods), the average values of fast, slow, and afast/aslow are given. Results are expressed as mean ± S.D. (number of patches). Data for receptors in BC3H1 cells are from Dilger and Brett, 1990.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   ACh concentration-response curves for WT and mutant AChR. Solid lines are drawn according to p = [ACh]nH/(EC + [ACh]nH; the best fit parameters are given in Table 2.

View this table: [in this window] [in a new window]   TABLE 2 Activation of native, WT, and mutant AChRs by ACh Concentration-response data (Fig. 3) were fit to the Hill equation (see Results). Results are presented as best fit parameter ±90% confidence limit. Data for receptors in BC3H1 cells are from Dilger and Brett, 1990, and consider a wider concentration range (0.3-1000 µM) than the other experiments.

Macroscopic Current Response to Perfusion of Decamethonium. The shift in ACh sensitivity induced by the mutations (Fig. 3) could arise from either a change in agonist affinity or agonist efficacy. Scheme 1 is the standard 4-state model for AChR channel activation.

View larger version (5K): [in this window] [in a new window]   Scheme 1.

View this table: [in this window] [in a new window]   TABLE 3 Activation of native, WT, and mutant AChRs by 100 µM decamethonium Results are expressed as mean ± S.D. (number of patches) in the presence of 100 µM decamethonium normalized to 100 µM ACh. Data for receptors in BC3H1 cells are from Liu and Dilger, 1993.

Single-Channel Experiments. The results of single-channel experiments with WT, 10', and 6' AChR are summarized in Table 4. Single channels are activated by transient exposure to 0.2 µM ACh. As observed with AChRs from BC₃H1 cells, the three types of expressed AChRs exhibit two-component open, burst, and closed duration histograms (see Fig. 7). A brief closed component with a time constant of 50 µs is expected to arise from repeated transitions of a single channel between A₂R and A₂R* before the agonist dissociates from its binding site (Zhang et al., 1995). This component is not fully resolved in our experiments, so the value of the observed fast time constant does not indicate the lifetime of the A₂R state. This also suggests that the observed open duration overestimates the actual value. The long closed time corresponds to the time between activation of separate channels in the patch and depends on the number of active channels in the patch and on the agonist concentration. For both WT and mutant receptors, the long closed time varied from patch-to-patch, between 30 and 700 ms. The origin of the two open and burst components is not certain, but a fraction of the brief duration openings is believed to originate from the opening of single-ligand receptors whereas, the long duration openings represent transitions to the more stable, double-liganded open state, A₂R* (Dilger et al., 1992). The single-channel current at 100 mV for both 10' and 6' AChRs was the same as for WT. This corresponds to a conductance of 40 pS.

View this table: [in this window] [in a new window]   TABLE 4 Single-channel properties of native, WT, and mutant AChRs Results are presented as mean ± S.D. for nine patches of WT, ten patches of 10', and seven patches of 6' (all expressed in HEK-293 cells). Data for receptors in BC3H1 cells are from Dilger et al., 1992, except for the brief long burst durations, which were not previously published (but are based on the same experiments with 20 patches) 0.2 µM ACh, 100 mV.

The Effects of Isoflurane on Macroscopic AChR Currents. Isoflurane inhibits macroscopic currents elicited by rapid perfusion of ACh (Dilger et al., 1993). Figure 4 presents examples of currents elicited by 100 µM ACh in controls and in the presence of 1 mM isoflurane for WT (Fig. 4A), 10' mutant (Fig. 4B), and 6' mutant (Fig. 4C) AChRs. Whereas 1 mM isoflurane produces the same inhibition of WT and 6' currents (46 and 47%, respectively), it produces more inhibition of the 10' channel current (88%). Inhibition curves for the three types of receptors are given in Fig. 5. These data were fit to a Hill equation (Fig. 5, solid lines):

(1)

where K_i is the drug concentration for 50% inhibition and n is the Hill coefficient.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Macroscopic current recordings from WT, 6', and 10' AChR. Activation by 100 µM ACh in the absence and presence of 1 mM isoflurane (equilibrium, +/+ protocol). Applied potential was 50 mV.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Isoflurane concentration-response curves for WT, 6', and 10' AChR. Solid lines are fits to the Hill equation (eq. 1); the broken line for the 10' AChR represents the fit to the two-site model (eq. 2). Fitting parameters are given in Table 3.

(2)

The Kinetics of Inhibition of Macroscopic Currents by Isoflurane. One way to test the idea that the apparent high affinity of 10' receptors for isoflurane is caused by an increase in binding to open channels is to examine the state dependence of block using different rapid application protocols (Dilger et al., 1993, 1997; Barann et al., 2000). For BC₃H1 cells, we found that simultaneous rapid perfusion of high concentrations of ACh and 1 mM isoflurane (/+ onset protocol) resulted in a macroscopic current that began close to the control level and decayed to close to the equilibrium level within a few hundred microseconds (Dilger et al., 1993). The situation for WT and 6' receptors is similar (Fig. 6 top, left and right, respectively). In contrast, the /+ perfusion protocol performed on the 10' receptors revealed comparatively less inhibition of the current (Fig. 6, top, center). For these figures, we used a concentration of isoflurane that produced nearly 50% inhibition of each mutant receptor (1.8 mM for WT, 1.6 mM for 6', 0.6 mM for 10').¹ We estimated the degree of inhibition of open channels by comparing the current after the fast decay process during a /+ protocol with the control peak current. The results of /+ protocol experiments over a range of isoflurane concentrations were fit to eq. 1 to determine the potency of isoflurane to inhibit open channels [K_i = 2.5 ± 0.8 mM for the 6' mutant (23 patches) and K_i = 1.6 ± 0.2 mM for the 10' mutant (19 patches)]. Thus, for the 10' mutant, isoflurane is much more effective when applied to closed channels (0.3 mM) then when applied to open channels (1.6 mM). There is also a marked difference between the mutants when the +/ protocol is used (Fig. 6, bottom panels). The WT and 6' mutant receptors begin to recover from inhibition by isoflurane within 1 ms, but the 10' receptors do not.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Macroscopic current responses to 1 mM ACh obtained under different perfusion protocols with equipotent concentrations of isoflurane. Left, WT and 1.8 mM isoflurane; center, 6' AChR and 1.6 mM isoflurane; right, 10' AChR and 0.6 mM isoflurane. The top panels show onset kinetics and compare the simultaneous perfusion of ACh and isoflurane (/+) with control (/ ) and equilibrium (+/+) application protocols. The bottom panels illustrate recovery kinetics and compare currents during rapid removal of isoflurane (+/) with control and equilibrium application protocols.

The Effects of Isoflurane on AChR Single-Channel Currents. If the difference in affinity for isoflurane between WT and 10' receptors is, indeed, caused by a difference in the binding of isoflurane within the closed but not open pore of the channel, then we expect there to be no differences in the behavior of single AChR channels in the presence of the drug. Single-channel recordings allow us to examine both the association and dissociation rates of drugs interacting with open channels.

View larger version (15K): [in this window] [in a new window]   Scheme 2.

(3)

(4)

(5)

View larger version (45K): [in this window] [in a new window]   Fig. 7.   Single-channel records (500-ms each) obtained from WT, 6', and 10' AChR in the absence and presence of 1 mM isoflurane. Applied potential = 100 mV, 0.2 µM ACh. Closed time distributions compiled from 30 to 60-s recording epochs are shown to the right of each record. The distributions are fit to two-exponential probability distribution functions, except for the distribution for 6' AChR in the presence of isoflurane, which required a three-exponential distribution function.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   Summary of the single-channel results for apparent open durations (A), gap durations (B), and openings per burst (C) for WT, 6', and 10' AChR. The inset to A shows a reciprocal plot for apparent open durations. Solid lines are the fits of the data to eqs. 3 to 5 as described in the text; the parameters are listed in Table 6. The broken line on the openings per burst versus [isoflurane] graph corresponds to the expected result if isoflurane were to bind to open 10' AChR channels 4-fold more tightly than to WT receptors.

View this table: [in this window] [in a new window]   TABLE 6 Parameters used in fitting the single-channel data The best fit values for ' and b for WT receptors are about 2-fold higher than those reported for the interaction of isoflurane with single channels in BC3H1 cells (Dilger et al., 1992). This difference may be the result of an incomplete correction for missed events in the analysis of the latter data.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Appendix References

We found no significant differences between WT ACh receptors expressed in HEK-293 cells and the native channel found in murine BC₃H1 cells either in the absence or presence of isoflurane. This suggests that the membrane and intracellular environment provided by the nonmuscle host cell line do not induce biophysical changes in receptor function.

Comparison of WT and Mutant AChR Currents and Channels. The mutant receptors studied here were chosen because they affect the interaction of QX-222 with the AChR channel (Charnet et al., 1990). Substitution of the small, nonpolar alanine residue for the polar serine and threonine residues at the 6' and 10' regions of M2 was considered to have an effect that is localized to the pore of the channel. However, our experiments show that these point mutations are not benign; they affect the global behavior of the channel. Other mutations in M2 have also been shown to affect the gating and/or binding behavior of the AChR (Filatov and White, 1995; Forman et al., 1995; Labarca et al., 1995; Chen and Auerbach, 1998; Grosman and Auerbach, 2000).

Interactions of Isoflurane with WT and Mutant AChR. Isoflurane inhibited ACh-activated macroscopic currents in the three channel types studied. The 10' mutant channel had a 4.5-fold greater affinity for isoflurane than did WT receptors; the affinity of 6' receptors was not different from that of WT (Fig. 5; Table 5). In all cases, the isoflurane concentration dependence was steep (n_H = 1.4-1.5). When the concentration dependence was fit by a two-site cooperative inhibition model, both sites on the 10' receptor exhibited greater affinity for isoflurane. These results agree with our hypothesis that polar-to-nonpolar mutations at the 10' level, but not at the 6' level, would increase the binding affinity for a hydrophobic drug such as isoflurane. However, the kinetic and single-channel data suggest that this simple interpretation is incomplete.

Comparisons with Published Data. In another study of the interaction of isoflurane and alcohols with 10' mutant receptors (Forman et al., 1995), it was also found that the binding affinity, as judged by inhibition of macroscopic currents, was enhanced by nonpolar substitutions at 10' on single and multiple subunits. Most of their experiments were done by use of the /+ protocol, so this suggests that the mutations changed the affinity of open channels for the drugs. The substitutions chosen for that study were larger residues (isoleucine, valine, and phenylalanine) than the alanine substitution used here. One could argue that large residues substantially modify the structure of the pore so as to create an anesthetic binding site where one did not exist in WT receptors. This is also a caveat for our study. However, the substitution of alanine (volume in protein interior = 90 Å³/residue; Harpaz et al., 1994) for serine (94 Å³/residue) has little effect on volume, and the substitution of alanine for threonine (120 Å³/residue) decreases volume by 30%. In contrast, the substitution of valine (139 Å³/residue), isoleucine (165 Å³/residue), or phenylalanine (193 Å³/residue) for the two -subunit serines increases the volume of the 10' region by 96, 150, and 210%, respectively. Forman et al. (1995) examined neither the state dependence nor the single-channel kinetics of the interaction between isoflurane and the mutant receptors.