Abstract
This study examined the effects of quinidine, quinine, and the quaternary quinidine derivative, quinidinium, on the conductance and activity of purified cardiac sarcoplasmic reticulum calcium release channels/ryanodine receptors (RyR) incorporated into planar lipid bilayers. Quinidine (50–500 μM) reduced the single-channel open probability in a voltage- and concentration-dependent manner. Reduction of channel activity was evident only at positive holding potentials where current flow is from the cytoplasmic to luminal side of the channel and when the drug was present only on the cytoplasmic face of the channel. A more pronounced effect was the appearance of a subconductance state at positive potentials. Single channel recordings and dose-response experiments revealed that at least two quinidine molecules were involved in reduction of the RyR activity. The permanently charged quinidinium compound produced nearly identical effects as quinidine when present only on cytoplasmic side of the channel, suggesting the positive-charged form of quinidine is responsible for the effects on the channel. There was no stereospecificity in the effects of quinidine because the levoisomer, 100 μM quinine, produced a similar subconductance activity of the channel. Ryanodine modification of the channel prevented subconductance activity. These findings suggest that the quinidine-induced subconductance activity may be the result of a partial occlusion of the channel pore interfering with ion conduction. Modification of the channel by ryanodine alters quinidine binding to the channel through a conformational change in protein structure.
The sarcoplasmic reticulum (SR) Ca2+ release channels/ryanodine receptors (RyR) play a vital role in the initiation of cellular contraction (for a review, see Meissner, 1994). In cardiac tissues, efflux of calcium from the SR is triggered by an influx of extracellular Ca2+ via the L-type Ca2+ channels through a process referred to as calcium-induced calcium release (Fabiato, 1983). Incorporation of purified RyR (Lai et al., 1988) into planar lipid bilayers has allowed for the study of the biophysical and pharmacological properties of the RyR. Agents that have been shown to have a direct influence on the RyR activity also affect release of Ca2+ from the SR and excitation-contraction coupling (for a review, see Zucchi and Ronca-Testoni, 1997).
The large molecular mass of the RyR (∼2000 kDa) has allowed for the visualization of the channel protein (Lai et al., 1988). These studies have revealed that the channel is a 4-fold symmetrical protein, measuring 27 × 27 × 14 nm with a 2-nm diameter pore. Detailed structural information on the channel pore has also been provided using permeant and impermeant organic cations andbis-quaternary ammonium blocking cations (Tinker and Williams, 1993a). These authors suggested the channel pore selectivity filter has a radius of ∼3.5Å located 90% into the voltage field from the cytoplasmic face with the entire electrical potential of the channel spanning 10.4Å. Block of RyR by large quaternary ammonium derivatives indicates that the pore may be the site for a number of positively charged agents to affect excitation-contraction coupling. We and other groups have recently demonstrated that local anesthetics can directly block single-channel activity of the RyR (Tinker and Williams, 1993b; Tsushima et al., 1996). Such agents have been shown to inhibit RyR and to possess negative inotropic activity (Bianchi and Bolton, 1967; Chamberlain et al., 1984).
Quinidine, an antiarrhythmic agent, is known to reduce contraction and alter cellular excitability in cardiac tissues (Parmley and Braunwald, 1967). It has been demonstrated that this agent interacts with multiple sarcolemmal voltage-gated ion channels (Lee et al., 1981; Hiraoka et al., 1986; Salata and Wasserstrom, 1987). The interaction of quinidine and other local anesthetic/antiarrhythmic agents with voltage-gated ion channels has been proposed to be the result of binding of these compounds within the inner vestibule of the channel pore (Hille, 1977). Mutagenesis studies have demonstrated critical intracellular residues important for quinidine binding to voltage-gated K+ and Na+ channels (Snyders et al., 1992; Ragsdale et al., 1995). This would imply that the cytoplasmic milieu is readily accessible to quinidine. The negative inotropic effects of quinidine have been, in part, associated with the inhibition of sarcolemmal L-type Ca2+ channels (Hiraoka et al., 1986; Salata and Wasserstrom, 1987); however, it is possible that blockade of RyR may also be involved in the depression of muscle contractility. Earlier studies have demonstrated [3H]quinidine binding to cardiac SR membranes (Besch and Watanabe, 1977) and alterations in cardiac SR Ca2+ handling by quinidine (Fuchs et al., 1968;Besch and Watanabe, 1977). Therefore, we examined the effects of quinidine on single-channel activity of RyR. The present study demonstrates that quinidine alters the conductance and activity of these channels in a voltage- and concentration-dependent manner. A more prominent effect is the appearance of a subconductance state. Further analysis revealed that ryanodine modification of the channel has a profound influence on the blocking properties of quinidine.
Experimental Procedures
Materials.
Quinidine and quinine were obtained from Sigma (St. Louis, MO). Quinidinium (N-methyl quinidine) was a gift from Dr. D. J. Synders (Vanderbilt University, Nashville, TN). Ryanodine was purchased from Calbiochem (San Diego, CA) and phospholipids were purchased from Avanti Polar Lipids, Inc. (Birmingham, AL). All other chemicals were analytical grade and were purchased from Sigma.
Isolation of Junctional SR Membrane Vesicles.
Canine cardiac junctional SR membrane vesicles were isolated as described previously (Tsushima et al., 1996). In brief, mongrel dogs were cared for according the Guide for the Care and Use of Laboratory Animals as adopted by the U.S. National Institutes of Health. Animals were anesthetized with sodium pentobarbital (65 mg/kg; iv), and ventricular tissue was excised and homogenized in 300 mM sucrose, 20 mM PIPES, and 0.5 mM EDTA (pH 7.4), in the presence of the protease inhibitors 100 nM aprotinin, 1 mM benzamidine, 1 mM iodoacetamide, 1 μM leupeptin, 1 μM pepstatin A, and 1 mM phenylmethylsulfonyl fluoride. These protease inhibitors were present in all solutions during the isolation of junctional SR membranes and purification of the RyR. The homogenate was centrifuged at 2800gmax for 20 min, and the subsequent supernatant was centrifuged at 120,000gmax to yield a mixed membrane preparation. This membrane preparation was fractionated on a 20 to 40% discontinuous sucrose-density gradient. Junctional SR membranes were recovered at the 30 to 40% interface, diluted with 1.5 volumes of 400 mM KCl, and pelleted at 120,000gmax. The pellet was resuspended in 300 mM sucrose and 5 mM HEPES/Tris solution (pH 7.2). Membranes vesicles were snap frozen and stored in liquid nitrogen.
Purification and Reconstitution of RyR.
RyR was purified and reconstituted into unilamellar liposomes as previously described (Tsushima et al., 1996). Junctional SR membranes were solubilized with the zwitterionic detergent Chaps followed by sucrose-density centrifugation. Chaps detergent was removed by dialysis. Purified channels were stabilized in proteoliposomes with phosphatidylcholine and stored in liquid nitrogen.
Planar Lipid Bilayer Measurements.
Single-channel activity was recorded from purified RyR incorporated into planar lipid bilayers containing phosphatidylethanolamine and phosphatidylserine (1:1; 40 mg/ml phospholipid in decane). Lipid bilayers were formed across a 200-μm hole in a Delrin partition that separates the cisand trans chambers of the bilayer apparatus. Single-channel activity was recorded under a symmetrical KCl solution (250 mM KCl, 0.15 mM CaCl2, 0.1 mM EGTA, and 10 mM HEPES/Tris, pH 7.4; free [Ca2+] = 50 μM). Channels were incorporated such that the cytoplasmic surface of the channel faced thecis chamber. Potentials given are those experienced at thecis (cytoplasmic) side relative to the trans(luminal) side, such that current flowing at positive holding potentials corresponds to current flow from cis totrans. Single-channel activity was measured using an Axopatch 200 amplifier and was stored directly into a 386 or 486 PC using pClamp software (Axon Instruments, Inc., Foster City, CA). Data were digitized at 10 kHz and filtered at 2 to 5 kHz. Quinidine was present in both the cis and trans chambers to prevent any asymmetrical surface potentials that may arise (Tinker et al., 1992).
Channel-state probabilities were determined from all points amplitude histograms fit with a Gaussian function. Kinetics of quinidine block of ryanodine-modified channels were determined as described previously (Tsushima et al., 1996). In brief, lifetime histograms were compiled using half-amplitude threshold analysis (Colquhoun and Sigworth, 1983). The cursor was manually set to the open, subconductance, or blocked state to derive open, substate, and blocked lifetimes, respectively, as performed by Tinker and Williams (1993b). Lifetime histograms were fit using pSTAT software (Axon Instruments), which uses Marquardt-Levenberg algorithms for estimates of the time constants. Statistical analysis was performed using a two-tailed unpaired t test. AP value less than 0.05 was deemed statistically significant. All values are presented as the mean ± S.E.M.
Results
Previous work from our laboratory has demonstrated that purified RyR displays a similar sensitivity to the physiologic (Ca2+, Mg2+, and ATP) and pharmacologic modulators (caffeine, ruthenium red, and ryanodine) as the native channel (Tsushima et al., 1996). The purified channel undergoes the characteristic 40% reduction in single-channel current amplitude with a prolongation in the open probability in the presence of ryanodine. We have demonstrated the pharmacological modulation of RyR channel activity by cocaine and dihydropyridine agonists (Tsushima et al., 1996; Sagawa et al., 2001). In the present study, we have further characterized the properties of RyR by examining the effects of quinidine and its quaternary derivative, quinidinium, on single-channel conductance and gating.
Effects of Quinidine on RyR.
The effects of quinidine on the purified RyR are illustrated in Figs. 1and 2. Quinidine reduced the open probability of the channel in a concentration-dependent manner but only at positive holding potentials (Figs. 1 and 2). The reduction in channel activity was induced only when quinidine was present on thecis (cytoplasmic) side of the channel. When quinidine was present on the trans (luminal) side, no block was observed (data not shown), suggesting the absence of a binding site on the luminal side of the channel. A more pronounced effect was the appearance of a subconductance state (substate). This substate was not the result of a second smaller conducting channel in the bilayer because we never observed this type of channel activity in the absence of quinidine. Previous studies examining the properties of the purified RyR have reported a common occurrence of subconductance activity (Liu et al., 1989). However, we observed little or no incidence of subconductance activity when the channel was solubilized in less Chaps detergent (0.5 versus 1.5%) for a shorter period of time (1 versus 2 h) compared with the studies mentioned above. Any channel displaying such activity under control recording conditions was discarded. Quinidine-induced subconductance activity was enhanced at more positive holding potentials. At negative potentials, we observed neither a change in channel activity nor the appearance of a substate (Figs. 1 and 2). Under symmetrical recording conditions (250 mM KCl), the channel displayed a linear current-voltage relationship with a slope conductance of 721 ± 4 pS (n = 8) (Fig.3). The quinidine-induced subconductance state had a slope conductance of 204 ± 9 pS (n = 8) (28% of control) at positive holding potentials, whereas the open-state conductance was not affected by quinidine (719 ± 5 pS,n = 8) (Fig. 3). This unidirectional block is similar to our previous observations with cocaine block of this channel (Tsushima et al., 1996).
Concentration Dependence of Quinidine Block.
The quinidine-induced subconductance activity was more apparent with increasing concentrations of quinidine (Fig. 1). The amplitude of the substate did not change with the progressive increase in quinidine levels. Initial inspection of the concentration dependence of quinidine block revealed that drug block did not seem to behave in a simple bimolecular process where occupation of the substate level would be enhanced with more quinidine as described by Scheme 1.
The ability of more than one quinidine molecule to induce channel block is demonstrated further by the concentration dependence of channel block (Fig. 5). Plotting the open channel probability (Popen) as a function of quinidine concentration reveals an IC50 value of 74.4 ± 12.7 μM (n = 6) and a Hill coefficient of 1.81 ± 0.64. The Hill coefficient of approximately 2 suggests that up to two molecules of quinidine may be involved to elicit full channel closure.
Voltage Dependence of Quinidine Block.
We also examined the voltage dependence of quinidine block (Fig.6). In the presence of 100 μM quinidine, increasing holding potential resulted in the steady decline in the open probability. In contrast, the probability of block was fairly constant from +40 to +55 mV but increased at potentials +60 mV and higher. More interestingly, substate probabilities demonstrated a biphasic response to quinidine showing a steady increase from 0.149 ± 0.031 (n = 8) at +40 mV to 0.293 ± 0.042 at +60 mV, but then declining to 0.185 ± 0.046 at +80 mV. The results suggest that at holding potentials between +40 and +55 mV, the decrease in the open probability is a consequence of quinidine-induced substate activity. With further depolarization, channel activity shifts from substate to block of the channel.
Quinidinium Block of RyR.
Under our experimental conditions, >90% of quinidine is in the protonated state; however, we cannot be certain from the above results whether the charged or uncharged form of the drug is responsible for the blocking behavior. To investigate this further, we examined the effects of the quaternary quinidine derivative, quinidinium (N-methyl quinidine), on RyR activity. The effects of quinidinium on channel activity are illustrated in Fig. 7. When present only on the trans face of the channel, 100 μM quinidinium did not elicit any subconductance activity or block at positive and negative holding potentials. Subsequent addition of quinidinium to the cytoplasmic side elicited substate gating at positive potentials (Fig.7).
We further examined the quinidinium block on channel-state probabilities of the RyR. Interestingly, quinidinium had nearly identical effects to quinidine on open, substate, and block probabilities (Fig. 4B). Quinidinium elicited a large reduction in the open probability of the channel from 0.868 ± 0.096 to 0.385 ± 0.047 at 100 μM (n = 4). As observed with quinidine, there was a biphasic effect on substate probabilities peaking at 150 to 200 μM (0.378 ± 0.067) and gradually decreasing at higher concentrations (0.236 ± 0.040 at 500 μM), whereas block probabilities steadily increased with quinidinium concentration. Plot of probability of unblocked channels as a function of quinidinium concentration revealed an IC50value of 117.7 ± 15.0 μM (n = 4) and a Hill coefficient of 1.35 ± 0.07 (Fig. 5). Although there is a slight reduction in the blocking affinity of quinidinium compared with quinidine (117.7 versus 74.4 μM), this difference did not reach statistical significance (P = 0.07). The similarity in blocking behavior compared with quinidine is not too surprising based on the near identical structural configuration of the two agents. The addition of the methyl group does not seem to alter dramatically the binding of quinidine to the channel pore. These results show that it is the charged-form of quinidine that is responsible for the blocking properties of the agent.
Effects of Quinine on RyR.
We examined the stereospecificity of quinidine block using the levoisomer of quinidine, quinine. Quinine (100 μM) reduced the single-channel open probability of the channel with concomitant production of a subconductance state (Fig.8). As observed with quinidine, block of the channel was not observed at negative holding potentials or when the drug was present on the luminal side of the channel. As a result of the similarity in the phenotypic changes to channel gating at similar drug concentration, we did not study further the kinetics of quinine block. Therefore, these data suggest that there is a lack of stereospecificity of quinidine block of RyR.
Ryanodine Modification on Quinidine Block.
Ryanodine induces an allosteric modification of RyR leading to changes in single-channel gating and conductance, and increases the sensitivity of the channel to Ca2+ (Du et al., 2001; Masumiya et al., 2001) due to marked structural changes in the configuration of the channel protein (Lindsay et al., 1994; Tu et al., 1994). Studies have determined the structural components of the ryanodine molecule responsible for altering channel gating and conductance (Tinker et al., 1996; Welch et al., 1997). In the presence of 1 μM ryanodine, there is a pronounced prolongation in the mean open time and a ∼40% reduction in single-channel amplitude, which differs from the quinidine-induced substate (Tsushima et al., 1996). Such changes to the channel by both ryanodine and quinidine may alter the interaction of these agents with the channel when present together. Therefore, we examined the interaction of quinidine on RyR modified with 1 μM ryanodine. In ryanodine-modified channels, quinidine reduced the open channel probability but did not induce subconductance activity (Fig.9). As with unmodified channels, quinidine had no effect on ryanodine-modified channels at negative holding potentials (data not shown). Kinetic analysis of the on- and off-rates of quinidine binding to the ryanodine-modified channel could be measured because open and closed lifetime histograms were best described by monoexponential functions. The kinetic values for quinidine on- and off-rates, kon andkoff, respectively, were derived from the time constants of the dwell-time histograms by the following equations.
Discussion
Quinidine block of RyR elicits a distinct subconductance state that seems to be precluded in channels previously modified by ryanodine. Further application of quinidine results in complete channel block due to an additional quinidine molecule interacting with the channel. Narrowing of the channel vestibule on the cytoplasmic face (Lindsay et al., 1994) or alterations in the cationic binding site in the pore due to conformational changes in RyR by ryanodine (Tanna et al., 2001) may explain the lack of substate activity in the presence of quinidine. This blockade differs from that we have observed previously with cocaine where a single blocking molecule binds within the pore to completely occlude ion flow (Tsushima et al., 1996).
Interpretations of the Quinidine-Induced Substate.
Many ion channels have inherent substate activity (for review, see Fox, 1987). Recording of purified RyR has demonstrated a frequent appearance of subconductance states, resulting from the presence of multiple conductance pathways (Liu et al., 1989). It is also possible that solubilization of the channel removes a regulatory protein involved in coordinating the gating of the homotetrameric protein. The FK506-binding protein, a cis-trans peptidy-prolyl isomerase has been recently shown to be associated with the RyR (Jayaraman et al., 1992) and, furthermore, to regulate the gating of expressed RyR (Brillantes et al., 1994). Interactions of this protein by FK506 or mice deficient in the FK506-binding protein result in the induction of subconductance activity of RyR (Ahern et al., 1994; Brillantes et al., 1994; Shou et al., 1998). With our purification conditions, we have a low frequency of substate activity, suggesting that the harsher solubilization technique used by the previous studies may account for the higher probability of substate activity (Lai et al., 1988). Our findings demonstrating quinidine-induced subconductance activity cannot be explained by an alteration in channel gating, wherein each individual subunit operates independently of one another or as a result of the loss of the FK506-binding protein. In such an instance, up to three subconductance states should be detected, each being a fraction of the open state. Only one quinidine-induced substate was observed having a conductance of 28% of the open state.
We initially envisioned the cationic quinidine molecule interacting within the RyR conduction pathway as observed with quinidine block of other channels where it occludes the channel pore (Snyders et al., 1992; Ragsdale et al., 1995). A similar blocking behavior has been described for tetraalkylammonium and QX314 block of sheep RyR (Tinker et al., 1992; Tinker and Williams, 1993b). These authors suggest the complex blocking interaction of these cationic compounds with the RyR channel is the result of an electrostatic barrier caused by the agent within the conduction pathway. Blockade of this nature was accurately modeled using multiple blocking molecules interacting within the channel pore (Tinker et al., 1992; Tinker and Williams, 1993b). Quinidine block of RyR involves the interaction of possibly two molecules interacting with the pore (Fig. 5). We and other groups have demonstrated the presence of two binding sites for cationic blockers within the RyR pore (Tinker et al., 1992; Tinker and Williams, 1993a,1993b; Tsushima et al., 1996). The data suggest the presence of a high-affinity hydrophobic cationic site located ∼90% into the voltage drop of the pore from the cytoplasmic side, which binds large tetraalkylammonium compounds, QX314, cocaine, and a low-affinity site further out in the pore (50%), which interacts with tetramethylammonium. We speculate one possible mechanism for quinidine block is a single quinidine molecule in the pore binding to the high-affinity site causing partial occlusion of the conduction pathway. A second quinidine can bind to the low-affinity site further out resulting in complete occlusion of the conduction pathway. This notion is similar to that proposed for the subconductance activities induced by large tetraalkylammonium compounds and QX314 as a result of the presence of more than one blocking molecule within the pore.
An alternative explanation for the subconductance activity and/or full closure of the RyR induced by quinidine is an allosteric conformational change in the structure of the channel due to binding of the drug outside the pore. Recent studies using cationic and neutral derivatives of ryanodine (21-amino-9α-hydroxyryanodine and ryanolol) have demonstrated a voltage dependence of subconductance activity on RyR (Tanna et al., 1998, 2000). It was initially speculated that translocation of the cationic ryanoid compound may account for the voltage dependence of substate activity (Tanna et al., 1998). Both the association and dissociation rates for ryanoid binding interactions with the channel were sensitive to the transmembrane voltage, suggesting that one possible mechanism for this effect was due to the translocation of the charged ryanoid into the electric field (Tanna et al., 1998). However, these authors suggested that voltage-dependent conformational changes leading to alterations in RyR affinity could also explain their findings. Subsequent analyses demonstrated the neutral ryanoid derivative, ryanodol, displayed qualitatively similar effects on the channel as the cationic ryanoid. These changes in channel gating were associated with voltage-dependent changes in the conformational state of the channel induced by the ryanoid compound, which were independent of the ryanoid translocating across the voltage field in the conduction pathway (Tanna et al., 2000). Quinidine (present study), and the ryanoid agents modify RyR gating when present only on the cytoplasmic side of the channel, and substate activity occurred only at positive holding potentials (Tanna et al., 1998,2000). The similarities in changes in RyR gating between these agents prevents us from excluding the possibility that one or both quinidine molecules induce their effects through allosteric binding outside the ion conducting pathway. Further studies are required to delineate the exact mechanism of quinidine block of RyR.
Implications of Changes in Quinidine Block by Ryanodine Modification.
Analysis of the ryanodine-modified RyR revealed that the channel undergoes a number of structural transformations (Lindsay et al., 1994; Tu et al., 1994). Alterations included a widening of the selectivity filter located deeper within the channel, a decreased density of the negative charges lining the pore and a narrowing of the outer vestibular region on the cytoplasmic face of the channel. Profound alterations such as these make it reasonable to speculate that channel block could be affected to some degree. Quinidine block was markedly altered after modification of the channel by ryanodine. Ryanodine modification of RyR resulted in the absence of a subconductance state in the presence of quinidine, while still allowing for up to two quinidine molecules to interact with the channel as observed in ryanodine-unmodified channels. The reduction in drug sensitivity after ryanodine modification (550 versus 75 μM) is similar to our previous observations in which the affinity for cocaine binding was reduced in ryanodine-modified channels (Tsushima et al., 1996). These effects could be the result of ryanodine-induced structural changes in RyR eliciting a change in the affinity of quinidine binding or could be due to the presence of only the lower affinity site remaining while the high-affinity site is occupied by ryanodine or not accessible. Similar findings with tetraalkylammonium and QX314 block of the ryanodine-modified RyR were observed, which were suggested to be the result of structural reorganizations in the conduction pathway with a relocation of the cationic drug site (Tinker and Williams, 1993c; Tanna et al., 2001).
In summary, we have demonstrated that quinidine elicits a unidirectional block of the RyR in a voltage- and concentration-dependent manner. Quinidine block of RyR would not seem to contribute to the negative inotropic effects of quinidine in cardiac tissue. Channel block was only observed at positive potentials where current flow is from the cytoplasmic to luminal side of the channel, opposite to direction of SR Ca2+ flux through these channels during excitation-contraction coupling. However, interaction of quinidine with the channel provides us with information on the structural framework of the channel pore. The substate blocking property of quinidine lacked any stereospecificity as judged by the similar effects of quinine on channel activity. Modification of the channel by ryanodine precludes substate activity, suggesting major alterations in the conduction pathway by ryanodine.
Acknowledgments
We thank Dr. Marlene Hosey (Department of Molecular and Cellular Pharmacology) for the use of her laboratory for the isolation, purification, and characterization of the purified RyR. We also thank Dr. Dirk Snyders (Vanderbilt University) for supplying us with the quinidinium compound.
Footnotes
-
↵1 Present address: Department of Medicine, University of Toronto, Toronto, Ontario, Canada, M5S 1A8.
-
This work was supported by grants from the National Heart, Lung and Blood Institute (HL 30724) and the American Heart Association of Metropolitan Chicago (to J.A.W.). R.G.T. was supported by a Medical Research Council of Canada fellowship during the tenure of this work. This work was presented in part elsewhere [Tsushima RG, Kelly JE, and Wasserstrom JA (1995) Quinidine-induced subconductance activity in purified cardiac SR Ca2+ release channels. Biophys J 68:A374].
- Abbreviations:
- SR
- sarcoplasmic reticulum
- RyR
- ryanodine receptor
- pS
- picoSiemen
- PIPES
- piperazine-N,N′-bis(2-ethanesulfonic acid
- PMSF
- phenylmethylsulfonyl fluoride
- Chaps
- 3-[3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate
- Received September 6, 2001.
- Accepted February 7, 2002.
- The American Society for Pharmacology and Experimental Therapeutics