Inhibition of Recombinant Ca2+ Channels by Benzothiazepines and Phenylalkylamines: Class-Specific Pharmacology and Underlying Molecular Determinants

  1. Dongming Cai,
  2. Jennifer G. Mulle and
  3. David T. Yue
  1. Program in Molecular and Cellular System Physiology, Department of Biomedical Engineering, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205
     
    Next Section

    Abstract

    To understand the molecular basis of state-dependent pharmacological blockade of voltage-gated Ca2+ channels, we systematically characterized phenylalkylamine and benzothiazepine inhibition of three molecular classes of Ca2+ channels (α1C, α1A, and α1E) expressed from cDNA clones transfected into HEK 293 cells. State-dependent blockade figures importantly in the therapeutically desirable property of use-dependent drug action. Verapamil (a phenylalkylamine) and diltiazem (a benzothiazepine) were imperfectly selective, so differences in the state dependence of inhibition could be compared among the various channels. We found only quantitative differences in pharmacological profile of verapamil: half-maximal inhibitory concentrations spanned a 2-fold range (70 μm for α1A, 100 μm for α1E, and 110 μm for α1C), and inhibition was state dependent in all channels. In contrast, diltiazem produced only state-dependent block of α1C channels; α1A and α1Echannels demonstrated state-independent block despite similar half-maximal inhibitory concentrations (60 μm for α1C, 220 μm for α1E, and 270 μm for α1A). To explore the molecular basis for the sharp distinction in state-dependent inhibition by diltiazem, we constructed chimeric channels from α1C and α1A and localized the structural determinants for state dependence to repeats III and IV of α1C, which have been found to contain the structures required for benzothiazepine binding. We then constructed a mutant α1C construct by changing three amino acids in IVS6 (Y1490I, A1494S, I1497M) that have been implicated as key coordinating sites for avid benzothiazepine binding. Although these mutations increased the half-maximal inhibitory concentration of diltiazem inhibition by ∼10-fold, the state-dependent nature of inhibition was spared. This result points to the existence of physically distinct elements controlling drug binding and access to the binding site, thereby favoring a “guarded-receptor” rather than a “modulated-receptor” mechanism of drug inhibition.

    Pharmaceutical agents that modulate voltage-gated Ca2+ channels have found widespread application as basic research tools and therapeutic drugs. From the biophysical perspective, L-type Ca2+ channel modulation by the most established generation of organic Ca2+ channel blockers (dihydropyridines, phenylalkylamines, and benzothiazepines) provides rich experimental models of allosteric modification of gating and/or physical blockade of the conduction pathway. In physiological experiments, these organic compounds, together with neuroactive peptides that block specific molecular classes of Ca2+channels (1), have been used to delineate the contributions of different channel types to specific biological functions. In the clinical setting, members of each of the three classes of organic compounds act on L-type channels to treat disorders as varied as angina, hypertension, and cardiac arrhythmia (2-4). Moreover, recombinant neuropeptides that inhibit neuronal N-type channels hold promise for the treatment of chronic intractable pain (5). Finally, mutations in neuronal P/Q-type channels have been linked to a inherited forms of migraine and ataxia (6), raising the possibility that as-yet-undiscovered compounds that select for P/Q channels may provide novel therapies for the more general forms of these neurological disorders. For all these reasons, there is enormous interest in understanding the molecular basis of Ca2+ channel inhibition and discovering new compounds with high specificity for selected molecular classes of Ca2+ channels.

    The cloning and expression of several classes of Ca2+channels provide powerful tools for addressing just these issues. First, mutant and chimeric Ca2+ channel analysis on the main α1 subunits have already revealed several key amino acid residues that can account for much of the class-specific difference in binding affinity of dihydropyridines (7-9), phenylalkylamines (9-11), and benzothiazepines (12, 13). Second, the ability to express a homogeneous population of a selected type of channel, in the virtual absence of contaminating currents, provides an ideal opportunity to assess the relative effect of an agent on a particular class of Ca2+ channel. The coexistence of multiple channel types in neurons and other native cells may complicate quantitative attribution of drug inhibition to particular channel types.

    Despite the rapid advances and molecular approaches outlined above, there is still little understanding of the molecular mechanism of Ca2+ channel inhibition. All recombinant channel studies to date have focused mainly on simple differences in the potency of inhibition of different recombinant Ca2+ channels by various organic blockers. No systematic comparison has been made of the characteristic phenotype of block. In particular, there is little information on the relative state dependence of block, a feature that is critical to the therapeutically useful property of “use dependence” (16, 17). Use-dependent blockers preferentially inhibit channels during unusually high electrical activity, as found in a cardiac arrhythmia or seizure focus. Differences in state-dependent block among the various recombinant channels could point the way to the structural determinants of use dependence. These molecular elements may be distinct from the structures that specify simple binding affinity according to the guarded-receptor hypothesis (18).

    Here, we report the systematic characterization of phenylalkylamine and benzothiazepine inhibition of three molecular classes of Ca2+ channels expressed in HEK 293 cells from cDNA clones encoding α1A, α1E, and α1Csubunits. The first two α1 clones correspond to neuronal P/Q-type and R-type channels (19), and the latter clone corresponds to cardiac L-type channels. We chose to study these two classes of organic compounds because they proved to be imperfectly selective, so differences in the state dependence of block could be compared among the various classes of Ca2+ channels. Dihydropyridines were almost perfectly selective for L-type channels (7; but see 14, 15), so no such comparison can be made.

    Previous SectionNext Section

    Materials and Methods

    Expression of recombinant voltage-gated Ca2+channels.

    cDNAs encoding calcium channel α1, β2a (20), and α2 (21) subunits were subcloned into mammalian expression plasmids. α1C (22), α1E (23), and β2a were subcloned into cytomegalovirus-promotor expression plasmid pGW1H (British Biotechnologies, Oxford, UK); α1A (24, 25), as well as chimeric and/or mutant α1 subunits (α1CCAA, α1AACC, and α1C-ISM, as described below), was subcloned into CMV-promotor expression plasmid pcDNA3 (InVitrogen Corporation, San Diego); and α2 was subcloned in the constitutively active, metallothionein-promotor expression plasmid pZEM229R (Zymogenetics, Seattle, WA). Low-passage-number (< 20) HEK 293 cells, obtained from Dr. Jeremy Nathans (26), were transiently transfected with plasmids containing α1 and β2a subunits (10 μg each/10-cm plate) using a calcium-phosphate precipitation procedure (27). With chimeric channels, the α2 subunit and pAdVAntage vector (Promega, Madison, WI) were sometimes cotransfected (10 and 5 μg/10-cm plate, respectively) to enhance expression, as noted. Cotransfection of the α2 subunit had no detectable effect on the pharmacological profile of diltiazem (not shown). pAdVAntage encodes the transcription-enhancing genes VAI and VAIIfrom the adenovirus genome (28). Recombinant currents were observed in > 30% of transfected cells for the majority of constructs; transfection with a β-galactosidase reporter gene resulted in a > 50% overall expression rate.

    Mock-transfected cells were cotransfected with β2a and α2 subunits. No high-voltage-activated channels were detected (n = 25 cells in three rounds of transfection). In mock-transfected cells, we occasionally (∼10–20% of cells) observed endogenous, low-voltage-activated Ca2+currents of small amplitude. Although endogenous currents of such small amplitude would contribute negligibly to most of our results, such cells were nevertheless rejected. Therefore, all data reflect the expression of recombinant Ca2+ channels.

    Construction of mutant and chimeric Ca2+ channel α1 subunits.

    Chimeric Ca2+ channels were generated in which repeats I and II were interchanged between α1C and α1A, yielding α1CCAAand α1AACC (see Fig. 8, top). To construct α1CCAA, we amplified a region spanning repeats III and IV of α1A (nucleotides 2024–6639, with nucleotide 1 at the start codon, here and throughout), using PCR catalyzed byPfu DNA polymerase (Stratagene, La Jolla, CA). The PCR product was initially blunt-end-ligated into pCR-Script using a commercial cloning kit (Stratagene). PCR primers contained flankingEcoRI and XbaI restriction endonuclease sites, enabling transfer of the PCR product into unique sites on the α1C construct. The resulting chimeric channel contained amino acids 1–746 from α1C, followed by 675-2213 from α1A. To construct α1AACC, we PCR-amplified a region spanning repeats III and IV of α1C (nucleotides 3198–5193). PCR primers contained flanking AccI andXbaI restriction sites, permitting ready transfer of the PCR product into these sites on the α1A construct. The resulting construct contained amino acids 1–1335 of α1A, followed by 1067–1732 of α1C and a premature stop codon. The premature stop codon was designed so as to delete the last 439 amino acids of the α1C carboxyl tail because it has been previously demonstrated that such truncation significantly enhances expression of current without an appreciable change in pharmacological profile (29).

    Figure 8
    View larger version:
    Figure 8

    Diltiazem inhibition of chimeric Ca2+ channels, derived from α1C and α1A. Left, data for α1-CCAA.Top left, chimera diagrammed (dark portions are derived from α1A).Right, results for α1-AACC. Top right, diagram. A and B (top), exemplar current traces (○) before and (•) during exposure to 200 μm diltiazem. A and B (bottom), diary plots of peak current (○) before, (•) during, and (cross-centered circle) after exposure to 200 μm diltiazem. Arrows, traces shown above. C and D, diary plots of r100, the fraction of peak current remaining at the end of 100-msec depolarization. These demonstrate that diltiazem produced little change in current decay of (C) α1-CCAA, in contrast to obvious acceleration of current decay of (D) α1-AACC. C, For α1-CCAA, the average change inr100 (drug − control) was small (−0.07 ± 0.04; three cells) and lacked statistical significance. D, In contrast, for α1-AACC, the change inr100 was large (−0.26 ± 0.06; four cells) and significant (p < 0.05). Test depolarizations were delivered every 15 sec from a holding potential of −80 mV, with 30 mm Ba2+ as charge carrier. Both chimeric α1 constructs were coexpressed with β2 and α2 subunits.

    We generated a mutant α1C construct in which three amino acids in the IVS6 region were changed (Y1490I, A1494S, I1497M, numbering relative to α1C) to those found in analogous positions of α1A, yielding the mutant construct α1C-ISM. To facilitate mutagenesis, two silent mutation were introduced into the standard α1C construct (C4974T and G4977A), yielding a unique SfuI restriction site at nucleotide 4973. In the resulting α1CSfuI+ construct, the IVS6 region was now bracketed by unique EcoRV andSfuI restriction sites, which are separated by only 626 base pairs. To generate α1C-ISM, we used PCR mutagenesis by overlap extension (30). The outside primers contained EcoRV and SfuI restriction sites, permitting transfer of the PCR product (containing the desired mutations) into these sites on α1CSfuI+. Portions of chimeric and mutant channel constructs derived from PCR were verified in their entirety with the use of the fluorescent dideoxy terminator method of thermocycle sequencing on an automated DNA sequencer (Applied Biosystems Division 373a; Perkin-Elmer Cetus, Norwalk, CT)

    Electrophysiology.

    Whole-cell recordings were conducted at room temperature 1–3 days after transfection. The external solution contained 150 mmN-methyl-d-glucamine aspartate, 10 mm glucose, 10 mm HEPES, 10 mm4-aminopyridine, and 0.1 mm EGTA, pH 7.3–7.4 with 1mN-methyl-d-glucamine aspartate; 2–30 mm CaCl2 or BaCl2 was added as charge carrier. The internal solution contained 135 mmCs-methanesulfonate, 5 mm CsCl, 10 mm EGTA, 10 mm HEPES, 1 mm MgCl2, and 4 mm MgATP, pH 7.2–7.3 with CsOH.

    L-type Ca2+ channel antagonists, verapamil (Calbiochem, La Jolla, CA), and diltiazem (Sigma Chemical, St. Louis, MO), were dissolved in distilled water to make stock solutions (10 mmor 1 m) that were stored at −20°. Aliquots were diluted to the external solution to obtain the final desired concentrations.

    For convenience, both Ca2+ and Ba2+ were used as charge carriers. Although an earlier study reported that the potency of Ca2+ channel blockers was modulated by the extracellular concentration and species of charge carrier (31), we found no evidence for such modulatory effects on α1C and α1Achannels with the use of verapamil and diltiazem. We found that neither concentration (Ca2+, 5 versus 30 mm) nor charge species (5 mm Ca2+ versus 5 mmBa2+, or 30 mm Ca2+ versus 30 mm Ba2+) significantly altered the extent of blockade by verapamil or diltiazem. Similar results were reported in the study of dihydropyridines (32, 33).

    Transfected HEK 293 cells were grown onto coverslips, which facilitated easy transfer to a recording chamber just before electrophysiological recording. Fresh external solution continuously perfused the chamber at a flow rate of 1–2 ml/min. The bath was grounded by a 0.5m KCl agar bridge attached to a Ag-AgCl wire. Whole-cell current records were obtained by standard patch-clamp techniques. Series resistance was typically < 5 MΩ and compensated > 60%. Leak and capacity transients were assessed between each test depolarization by a P/8 protocol and subtracted from test currents in all subsequent data analysis. Currents were filtered at 2 kHz (−3 dB, four-pole Bessel), sampled at 10 kHz, and stored digitally for data analysis.

    Statistical analysis.

    All pooled data are reported as mean ± standard error. Statistical significance was assessed by two-sided, paired t test, with p < 0.05 taken as the minimal level of significance. N.S. denotes a comparison with no statistical significance (p > 0.05).

    Previous SectionNext Section

    Results

    Verapamil and diltiazem inhibition of multiple classes of Ca2+ channels.

    Although verapamil was developed as a blocker of L-type Ca2+ channels, Fig. 1illustrates that the identical concentration of this compound (50 μm) produced significant block of all three classes of channels tested. The pharmacological profiles of α1C, α1A, and α1E are presented in three columns, as labeled. The exemplar currents at the top, taken before and during application of verapamil (Fig. 1, A–C), explicitly demonstrate the inhibition of each type of channel. The diary plots of peak current shown below indicate the rapid inhibition of current on exposure to verapamil (•), which is readily reversed on wash with control solution (▵), except in the case of α1C. Slow or incomplete reversal of block was characteristic of α1C.

    Figure 1
    View larger version:
    Figure 1

    Verapamil inhibition of multiple classes of Ca2+ channels. Top (A–C), exemplar current traces from (A) α1C, (B) α1A, and (C) α1E, all coexpressed with β2, in the (○) absence and (•) presence of 50 μm verapamil. The charge carrier was 5 mm Ca2+, and currents were evoked every 15 sec by 100-msec depolarizing pulses from a holding potential of −80 mV. Bottom (A–C), diary plots of peak current (○) before, (•) during, and (▵) after exposure to 50 μm verapamil. Arrows, traces shown above. D–F, Diary plots of r100, the fraction of peak current remaining at the end of 100-msec depolarization. Plots quantify the enhancement of current decay during verapamil inhibition. Results for α1C, α1A, and α1Eare representative of those observed in a total of seven, four, and five cells, respectively.

    Apart from inhibition of peak current, a notable effect of verapamil was to accelerate the decay of current during maintained step depolarization. This effect is apparent in the exemplar traces (top, A–C), as well as in the diary plots of the fraction of peak current remaining at the end of 100-msec depolarizing steps (Fig. 1, D–F, r100). This acceleration of the decay of current is consistent with a state-dependent blocking mechanism in which verapamil preferentially inhibits the channel when it resides in or near the open and/or inactivated state. Such preferential inhibition of open/inactivated states could underlie the substantial use-dependent component of block observed in all channels with our 1/15 Hz stimulation frequency (data not shown). Overall, there was no fundamental difference in the action of verapamil on the three classes of channels.

    In contrast, diltiazem produced fundamentally different effects on α1C compared with α1E or α1A channels. Fig. 2 summarizes the pharmacological profile of diltiazem, following the same format as in Fig. 1. Although peak currents were readily inhibited by 100 μm diltiazem for all three channels (Fig. 2, A–C), the acceleration of current decay during maintained depolarizing steps was present in only α1C. The latter finding is demonstrated by the exemplar records (top, A–C), as well as in the diary plots of r100 (D–F). The flatr100 plots for α1A (Fig. 2E) and α1E (Fig. 2F) quantitatively demonstrate the lack of change in current waveform with diltiazem exposure. These results suggested that diltiazem might selectively demonstrate preferential open/inactivated-state block of α1C but not α1A and α1E channels. This idea was consistent with the absence of appreciable use-dependent block with α1A and α1E channels (data not shown). If confirmed, such a clearcut distinction in fundamental action could permit structural analysis of the basis of state-dependent block by diltiazem.

    Figure 2
    View larger version:
    Figure 2

    Diltiazem inhibition of multiple classes of Ca2+ channels. The format is identical to that in Fig. 1, except that channels were inhibited by 100 μm diltiazem. Results for (A) α1C, (B) α1A, and (C) α1E are representative of those observed in a total of four, four, and seven cells, respectively.

    Dose-response curves for verapamil and diltiazem block.

    As another potential test for state-dependent block in a given drug/channel combination, we characterized the dose dependence of blockade by verapamil and diltiazem for all three types of channels. If a drug binds with different affinity to channels in distinct conformations, then the dose-response curve for drug blockade might be described by multiple binding isotherms corresponding to various channel states. On the other hand, if a drug binds with the same affinity, regardless of channel state, then the dose-response relation should describe a single Langmuir function.

    Fig. 3 shows the dose-response of verapamil and diltiazem blockade for the three types of Ca2+ channels. In each case of verapamil blockade, fits of relations with single binding isotherms (not shown) consistently underestimated the degree of block in the 1–10 μm range, with differences between data and theoretical predictions (“residuals”) ranging between 0.1 and 0.2. In contrast, fits with functions composed of two binding isotherms (solid gray curves) yielded no such consistent deviations in residuals, as if channels were roughly distributed among two groups of states: one with a high affinity for verapamil and the other with low affinity. These results fit nicely with the consistent acceleration of current decay produced by verapamil on all three classes of channels (Fig. 1). Therefore, both dose-response (Fig. 3A) andr100 data (Fig. 1, D–F) were consistent with state-dependent blockade of all three channels by verapamil.

    Figure 3
    View larger version:
    Figure 3

    Concentration dependence of verapamil and diltiazem inhibition. Ordinate, peak current during steady state exposure to the blockers, normalized by the peak current before drug exposure, defined as Idrug/Ictrl.Points, mean ± standard error obtained from three to five cells. We applied only one concentration of drug to each cell, and the extent of drug block was measured as soon as peak currents showed no visible change from one pulse to the next. We thereby minimized possible distortion that might have occurred with a cumulative dose-response protocol, in which variable rundown could complicate interpretation. A, In each case of verapamil blockade, data were fit by a dual binding-site relation: Idrug/Ictrl = 1 − f1*[drug]/([drug] +Kd1) - f2*[drug]/([drug] +Kd2), where f1 + f2 = 1, and f1 and f2 represent the fraction of channels in high and low affinity binding conformations, respectively; andKd1 andKd2 represent dissociation constants for high and low affinity states, respectively. Dose-response relations for channels with high and low affinity states would be described by this two-isotherm relation as long as equilibration of drug binding is rapid relative to exchange between high and low affinity states. Solid gray curve, fit of this relation. Dashed and dotted curves, contribution of individual binding isotherms. A computer-based, iterative, nonlinear fitting algorithm was used. For α1C,Kd1 = 0.29 μm,Kd2 = 220 μm, f1 = 0.19, and f2 = 0.81. For α1A,Kd1 = 1.57 μm,Kd2 = 124 μm, f1 = 0.21, and f2 = 0.79. For α1E,Kd1 = 0.13 μm,Kd2 = 123 μm, f1 = 0.09, and f2 = 0.91. When fits were performed with a one binding isotherm, there was a consistent deviation in residuals (theoretical prediction-data ∼0.1 to 0.2) in the 1–10 μm range for all three channels. We observed no such consistent deviation in residuals with fits of the two-binding-isotherm function. For all data, currents were elicited by depolarizations to +10 mV, delivered every 15 sec from a holding potential of −80 mV, with 2 mmCa2+ as charge carrier. B, Concentration dependence of diltiazem inhibition. The format is identical to that in A, except that single-binding isotherms were used to fit the data for α1A and α1E(Idrug/Ictrl =Kd1/([drug] +Kd1)). The fit parameters were as follows. For α1C,Kd1 = 4.48 μm,Kd2 = 129 μm, f1 = 0.29, and f2 = 0.71. For α1A,Kd1 = 270 μm. For α1E,Kd1 = 220 μm. In the case of α1C, the fit with a single-binding isotherm function consistently showed a residual theoretical prediction-data) averaging −0.15 at 10 μm diltiazem. No such deviation in residuals was observed with two-binding-isotherm fit shown (top,solid gray curve). For α1A and α1E, fits of a single binding isotherm (solid gray curves, middle and bottom, respectively) showed no consistent deviation of residuals.

    The form of dose-response relations for diltiazem inhibition also agreed with our initial assignment of state-dependent block to α1C (Fig. 2D), and state-independent block to α1A (Fig. 2E) and α1E (Fig. 2F). The α1C data were well fit by a relation with two binding isotherms (Fig. 3B, top), which is consistent with the channel being distributed between higher and lower affinity states. Fits of relations with one binding isotherm (not shown) yielded consistent deviations in residuals, as observed above with verapamil. In contrast, the α1A and α1E dose-response data were well fit by a single binding term (Fig. 3B, middleand bottom), as if diltiazem had an equal affinity for all states.

    Holding-potential dependence of drug inhibition.

    Another critical test for state-dependent block in certain drug/channel combinations is to determine whether the degree of drug inhibition is dependent on the holding potential between voltage pulses (32, 34). So far, all of our experiments were conducted with a fixed holding potential of −80 mV. At different holding potentials, the distribution of channels among various states was likely to be different. Consequently, if a drug bound with different affinity to different states, then the degree of inhibition would be affected by holding potential. Alternatively, if the drug bound with equal affinity to all states, no such holding potential dependence would be observed.

    Fig. 4 shows the effects of holding potential on verapamil inhibition. The schematic (top) represents the holding potentials used in the various phases of the protocol. Brief test pulses were delivered every 15 sec from the indicated holding potentials. Because all measurements were taken over a short time period, we minimized the effects of drift and rundown. Before the application of verapamil, the holding potential was initially set at a depolarized level (−60 mV for α1C and α1A; −80 mV for α1E), where the channel populated deep resting, shallow resting, and inactivated states. After stabilization at this potential (○), we set the holding potential to a hyperpolarized value (−100 mV for α1C and α1A; −110 mV for α1E) while still in the absence of verapamil. After equilibration (▵), inactivation was minimal, and most channels were in a deep resting state. The control currents obtained at these two potentials served to normalize currents inhibited by verapamil. While maintaining the hyperpolarized holding potential, verapamil was then applied, and the steady state level of inhibition was allowed to develop (▴). The holding potential was then adjusted to the initial depolarized level and the steady level of inhibition allowed to develop with verapamil still present (•). Verapamil was then removed, and the current allowed to recover at the depolarized (cross-centered circle) and hyperpolarized (cross-centered triangle) holding potentials. The degree of inhibition at the two holding potentials was obtained by normalizing inhibited currents by their corresponding control currents (hyperpolarized inhibition = ▴ ÷ ▵; depolarized inhibition = • ÷ ○).

    Figure 4
    View larger version:
    Figure 4

    Effect of holding potential on verapamil inhibition. Top, diagram of holding potentials used in the protocol described below. A–C, Diary plots of peak current measured during various phases of protocol. D–F, Extent of verapamil inhibition at different holding potentials for (D) α1C, (E) α1A, and (F) α1E.Ordinate, fraction of control current remaining in the presence of verapamil. Statistical significance was fromn cells as indicated. Brief test pulses to +10 mV were delivered every 15 sec from the indicated holding potentials, with 5 mm Ca2+ as charge carrier. Before the application of verapamil, we set the holding potential to a depolarized level, where the channel populated deep resting, shallow resting, and inactivated states. After stabilization at this potential (○), we changed the holding potential to a hyperpolarized value while still in the absence of verapamil. After equilibration (▵), inactivation was minimized, and most channels populated deep resting states. The control currents obtained at these two potentials served to normalize currents inhibited by verapamil. While maintaining the hyperpolarized holding potential, verapamil was applied, and the steady state level of inhibition was allowed to develop (▴). The holding potential was then adjusted to the initial depolarized level, and the steady level of inhibition was allowed to develop with verapamil still present (•). Verapamil was then removed, and the current was allowed to recover at the depolarized (cross-centered circle) and hyperpolarized (cross-centered triangle) holding potentials. The degree of inhibition at the two holding potentials was obtained by normalizing inhibited currents by their corresponding control currents (hyperpolarized inhibition = ▴ ÷ ▵; depolarized inhibition = • ÷ ○). We chose verapamil concentrations of (B) 30 μm for α1A, (A) 50 μm for α1C, and (C) 100 μm for α1E, at which comparable degrees of inhibition were produced at holding potentials of −110 to −100 mV.  

    Results of this experiment for each of three classes of channels are reported in separate columns in Fig. 4, as labeled. At the top of each column (A–C) is a diary plot of peak current. Shown below is the degree of inhibition at hyperpolarized and depolarized holding potentials. For each channel type, inhibition was clearly greater at the depolarized holding potential, which is consistent with state-dependent block of all channels by verapamil.

    Fig. 5 summarizes results for the same holding potential experiment with diltiazem, using the identical format. Here, only α1C showed greater inhibition at the depolarized holding potential (Fig. 5A); α1A and α1E were inhibited to the same extent (Figs. 5, B and C), regardless of holding potential. These results support the view that only α1Cshowed state-dependent block by diltiazem.

    Figure 5
    View larger version:
    Figure 5

    Effect of holding potential on diltiazem inhibition. The format and protocol are identical to those in Fig. 4, except that 100 μm diltiazem was used for drug inhibition throughout.

    Unequivocal support for the results of these two-point holding potential experiments came with quantitative shape analysis of complete steady state inactivation (h∞) curves, obtained in the presence and absence of drug. Following the same rationale as the streamlined holding potential protocol above, state-dependent blockade should not only depress theh∞ curve but also shift its position after renormalization. State-independent block should depress theh∞ curve without a shift in position. Fig.6 shows the results of such an experiment, in whichh∞ curves were obtained with 20-sec prepulses to various voltages. Data for the three classes of channels are presented in separate columns, as labeled. At the top are shownh∞ curves obtained in the absence (open symbol) and presence (filled symbol) of verapamil. In each case, there was a substantial depression of h∞ curves by verapamil. At the bottom are shown twoh∞ curves after normalization to facilitate shape comparison. Clearly, there were changes in shape with drug inhibition of all channels, which is consistent with state-dependent inhibition throughout. Fig. 7 summarizes identical experiments, now with diltiazem as the blocker. After normalization ofh∞ curves (bottom, A–C), only α1C gave evidence of state-dependent block.

    Figure 6
    View larger version:
    Figure 6

    Effect of verapamil on steady state inactivation (h∞) curves. Data for three classes of channels are presented in separate columns as labeled.Top, h∞ curves obtained in the (open) absence and (filled) presence of 50 μm verapamil. Ordinate, peak test currents after normalization by the peak test current obtained following a prepulse to −110 mV in the absence of drug.Bottom, two h∞ curves, after normalization to facilitate shape comparison. All channel types show shifts along the voltage axis. Protocol, every 60 sec, test depolarizations to +10 mV were delivered after a 20-sec prepulse to voltages from −110 mV to 0 mV. The holding potential was −110 mV throughout. The charge carrier was 5 mmCa2+. Separate groups of cells were averaged for (open ) control and (filled ) drug data. Each group contained four or five cells.

    Figure 7
    View larger version:
    Figure 7

    Effect of diltiazem inhibition on steady state inactivation (h∞) curves. Data for the three classes of channels are presented in separate columns as labeled. The format is identical to that in Fig. 6, except that 100 μm diltiazem was used for drug inhibition. After normalization to facilitate shape comparison (bottom row), shifts along the voltage axis are apparent for only α1C; the normalized h∞relations for α1A and α1E essentially superimpose, arguing for state-independent block. Control and drug data were averaged from different groups of cells. The control (open) data are replotted from Fig. 6. The drug (filled) data are averaged from four or five cells in each group.

    Structural determinants of state-dependent block by benzothiazepines.

    Taken together, the r100(Figs. 1 and 2), dose-response (Fig. 3), and holding-potential (Figs.4, 5, 6, 7) experiments provided good evidence that a benzothiazepine produced state-dependent block of only α1C channels, not α1A or α1E channels. This clear distinction in a fundamental pharmacological property furnished us a novel opportunity to apply chimeric and mutant Ca2+ channel analysis to gain insight into the structural determinants of state-dependent block.

    According to the guarded-receptor hypothesis of drug block, the drug-binding site and guard elements that control drug access to the binding site are physically distinct structures (18). Critical amino acids for diltiazem binding have been localized to IVS6 (12), with potential ancillary contribution from IIIS6 (13). We therefore wondered whether structures specifying properties of state dependence were located on repeats III and IV near the binding site or could be entirely distinct structures residing on repeats I and II. Accordingly, we examined diltiazem blockade of chimeric Ca2+ channels in which repeats I and II were interchanged between α1C and α1A. The resulting chimeric channels are schematized in Fig. 8 (top). The profile of diltiazem blockade for the two constructs appear in separate columns, as labeled. Both channel constructs were readily inhibited by 200 μmdiltiazem, as demonstrated in Fig. 8, A and B, by the exemplar current records (top) and the diary plots of peak current (bottom). Interestingly, the recovery from inhibition was rapid in α1CCAA but slow in α1AACC, suggesting that the characteristic slow recovery of α1C(Fig. 2A) was mediated by structures in repeats III and IV near the binding site. The most important new findings came with the effects of diltiazem on the decay rate of currents. Close inspection of the exemplar traces shows that diltiazem hardly affected the waveform of α1CCAA (Fig. 8A, top) but significantly accelerated the decay of current of α1AACC (Fig. 8B,top). The r100 diaries, shown below in Fig. 8, C and D, confirm these impressions, which argue for the predominance of repeats III and IV in mediating the state-dependent phenotype.

    Such predominance of repeats III and IV gave reason to wonder whether the very amino acids that are critical for higher affinity diltiazem binding in α1C (12) are also essential to the state-dependent phenotype. A tyrosine, an alanine, and an isoleucine located in IVS6 of α1C (Y1490, A1494, and I1497 in the rabbit cardiac α1C used here) have been proposed to be crucial to diltiazem inhibition; when the corresponding amino acids in α1A are mutated to these amino acids, higher affinity diltiazem blockade is conferred to the α1A backbone (12). Here, we performed the converse experiment and mutated the critical amino acids in α1C to their α1Acounterparts. We then asked whether these amino acids alone, which are a critical part of the binding pocket, are essential to the state-dependent phenotype.

    Fig. 9 shows the effects of diltiazem on the resulting construct, α1C-ISM. The exemplar currents and diary plot (Fig. 9A) demonstrate the ready block of peak current by diltiazem, although the half-maximal inhibitory concentration was increased by ∼10-fold (not shown). More telling was the effect of diltiazem on the decay of current in the exemplar records. In addition to the substantial inhibition of current, diltiazem still produced a marked acceleration of the current decay. This impression was confirmed by the obvious decline in the diary plot of r100 during diltiazem application (Fig. 9B, bottom). These results suggested that diltiazem still preferentially inhibited channels in the open and/or inactivated states.

    Figure 9
    View larger version:
    Figure 9

    Diltiazem inhibition of α1C-ISM, an α1C construct in which three point mutations in IVS6 have been made to remove the high affinity coordinating sites for diltiazem (Y1490I, A1494S, I1497M). Top, location of the point mutations is diagrammed. A (top), exemplar current traces (○) before and (•) during exposure to 200 μm diltiazem. A (bottom), diary plots of peak current (○) before, (•) during, and (▴) after exposure to 200 μm diltiazem. Arrows, traces shown above. B, Diary plot of r100, the fraction of peak current remaining at the end of 100-msec depolarization, corresponding to the same cell and protocol as in A. The depression of r100 with diltiazem demonstrates obvious speeding of current decay with drug application. For the exemplar experiment (A and B), 5 mmBa2+ was the charge carrier. There was no difference in the change of r100 upon drug application with Ba2+ or Ca2+ serving as charge carrier. Overall, the change in r100 was large and reproducible, averaging −0.16 ± 0.02 (p< 0.01 for seven cells: three with 5 mmBa2+ and four with 5 mm Ca2+as charge carrier). C, Diary plot of peak current measured during various phases of holding potential (VH) protocol diagrammed (top). Protocol was identical to that in Fig.5. D, Extent of diltiazem inhibition at different holding potentials, derived from diary plots as in C. Statistical significance was assessed from four cells, with 5 mm Ca2+ as charge carrier. α1C-ISM was coexpressed with β2and α2 subunits throughout.

    To bolster the idea that diltiazem block of α1C-ISM was state dependent, we examined the holding-potential dependence of diltiazem blockade of α1C-ISM. Fig. 9, C and D, summarizes the results of a two-point holding potential experiment that is identical in every regard to that reported for α1C in Fig. 5, A and D. The data clearly argue that there was still an appreciable effect of holding potential on diltiazem blockade.

    The experiments shown in Fig. 9 consistently demonstrate that there must be other distinct structural elements in repeats III and IV, different from the amino acids most critical to diltiazem binding (Y1490, A1494, I1497), that are essential for state-dependent blockade. This result has important structural implications on which we elaborate in the Discussion.

    Previous SectionNext Section

    Discussion

    We report on the first systematic characterization of phenylalkylamine and benzothiazepine inhibition of three molecular classes of Ca2+ channels (α1A, α1E, and α1C), expressed individually from recombinant clones. Although phenylalkylamines and benzothiazepines were developed as blockers of L-type (α1C) channels, we find that verapamil and diltiazem inhibit all three classes of channels tested, with half-maximal inhibitory concentrations that differ by less than several-fold (Fig. 3). There are striking qualitative differences in the state dependence of inhibition observed in different drug/channel combinations. Although verapamil shows evidence of state-dependent inhibition in all three channels, diltiazem produces state-dependent block of only α1C channels (not α1A and α1E channels). The determination of state dependence is based on mutually consistent results from current decay (Figs. 1 and 2), dose-response (Fig. 3), and holding-potential (Figs. 4, 5, 6, 7) experiments. This is the first instance in which an organic Ca2+ channel blocker demonstrates similar overall potency in different channels but a fundamental difference in the state-dependent character of blockade. Verapamil, D888, and mibefradil all show signs of state-dependent blockade of several molecular classes of Ca2+ channels (9-11, 35).

    Relation to previous studies.

    For channel/drug combinations that manifest state-dependent blockade, the extent of blockade will vary considerably depending on the holding potential (Figs. 6 and 7) and stimulation rate. Therefore, in comparing the half-maximal inhibitory concentrations specified in Fig. 3 with those in the literature (12), it is worth emphasizing that the data in these figures were obtained with a uniform holding potential and stimulation rate (−80 mV and 1/15 Hz, respectively). In instances in which there was considerable state-dependent blockade, the half-blocking concentration would certainly have been lower with depolarization of the holding potential and/or increase in the stimulation rate.

    Our finding of selective state dependence of diltiazem blockade with α1C (but not with α1A and α1E) contrasts somewhat with another study (12) in which recombinant α1A currents, expressed in Xenopus laevis oocytes, showed evidence of state-dependent diltiazem blockade. The difference may arise from the use of different α1A clones [rat brain in our case (24) versus rabbit brain in the other study (36)] or different β subunits [rat brain β2 in our case (20) versus rabbit skeletal β1a (37)]. In the former case, it may be that a restricted number of differences between the two α1Aclones account for the difference in state dependence, but it is interesting that our finding of state-independent diltiazem block extends to a construct as different as α1E. Differences in pharmacological profile related to alternative splicing (38) looms as another possibility with enormous therapeutic potential. Further work is clearly necessary to characterize the structural determinants of the state-dependent phenotype.

    Implications for drug discovery.

    The finding that diltiazem blocks only α1C in a state-dependent manner gives reason to expect that there exist organic compounds that would demonstrate selective state-dependent blockade of certain neuronal Ca2+channels. Our results with diltiazem argue that the general functional requirements of Ca2+ channels have not imposed common structural themes that require organic compounds to demonstrate similar state-dependent blocking properties across channels, as in the case of verapamil. In fact, eliprodil, an organic drug initially developed as an N-methyl-d-aspartate channel antagonist, seems to demonstrate selective blockade of N-type (α1B) and P/Q-type (α1A) channels while sparing L-type (α1C) and R-type (α1E) channels (39). There is great interest in discovering organic compounds with selectivity for neuronal channels because they may prove to be easier to obtain and deliver than neuropeptides, which despite their proven selectivity for particular types of neuronal channels, may have difficulty crossing the blood-brain barrier to reach targets in the central nervous system (35).

    Molecular basis of state-dependent diltiazem blockade.

    Chimeric Ca2+ channel analysis between α1Cand α1A localized most of the structural determinants for state-dependent diltiazem blockade to repeats III and IV of α1C, which likely contain the elements required for benzothiazepine binding (12, 13). Analysis of α1C-ISM, a construct in which the predominant structural determinants of higher-affinity diltiazem inhibition of α1C have been changed to their α1A counterparts, revealed that state-dependent diltiazem inhibition was spared, despite a ∼10-fold increase in half-maximal inhibitory concentration relative to α1C. These results suggest that although the elements responsible for state dependence (“guards”) are largely contained within repeats III and IV, they may be structurally distinct from the most critical part of the actual binding region in IVS6. The existence of physically distinct elements controlling drug binding and access to the receptor fits naturally with a guarded-receptor (18) rather than a modulated receptor mechanism of drug inhibition (16, 17). Such a physical dissociation of elements controlling binding and access has also been demonstrated in voltage-gated Na+ channels (40). It will be interesting to see whether such a physical distinction turns out to be a general design theme for state-dependent block of all voltage-gated channels.

    Previous SectionNext Section

    Acknowledgments

    We are grateful to Dr. Shao-kui Wei (Johns Hopkins University School of Medicine) for his collaboration on certain experiments, to Yan Wang for constructing α1CSfuI+, and to Dr. Paul Fuchs, David Brody, Lisa Jones, and Parag Patil for discussion and comments. We acknowledge the following individuals for generously providing Ca2+ channel clones: Terry P. Snutch (University of British Columbia, Vancouver, British Columbia, Canada) (α1A, α1E, α2), Edward Perez-Reyes (Loyola University, Chicago, IL) (α1C, β2a), and the late Chris Wei (Georgia Medical College) (α1C).

    Previous SectionNext Section

    Footnotes

    • Send reprint requests to: David T. Yue, M.D., Ph.D., Program in Molecular and Cellular System Physiology, Department of Biomedical Engineering, Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205. E-mail: dyue{at}bme.jhu.edu

    • This work was supported by Specialized Center of Research in Sudden Cardiac Death Grant NIH P50-HL52307 (D.T.Y.) and Postdoctoral Training Fellowship NIH 5-T32-HL07581 (D.M.C.).

    • Abbreviations:
      HEK
      human embryonic kidney
      PCR
      polymerase chain reaction
      EGTA
      ethylene glycol bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid
      HEPES
      4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
      • Received December 11, 1996.
      • Accepted February 13, 1997.
    Previous Section
     

    References

    1. ↵
      1. Olivera B. M.,
      2. Miljanich G. P.,
      3. Ramachandran J.,
      4. Adams M. E.
      (1994) Calcium channel diversity and neurotransmitter release: the omega-conotoxins and omega-agatoxins. Annu. Rev. Biochem 63:823–867.
      CrossRefMedlineWeb of Science
    2. ↵
      1. Fleckenstein A.
      (1983) History of calcium antagonist. Circ Res. 52:I3–I16.
    3. ↵
      1. Partridge L. D.,
      2. Leach J. K.
      1. Janis R. A.,
      2. Triggle D. J.
      (1991) Drugs action on calcium channels. in Calcium Channels: Their Properties, Functions, Regulation, and Clinic Relevance, eds Partridge L. D., Leach J. K. (CRC Press, Boca Raton, FL), pp 195–249.
    4. ↵
      1. McDonald T. F.,
      2. Pelzer S.,
      3. Trautwein W.,
      4. Pelzer D. J.
      (1994) Regulation and modulation of calcium channels in cardiac, skeletal, and smooth muscle cells. Physiol. Rev. 74:365–507.
      FREE Full Text
    5. ↵
      1. Xiao W. H.,
      2. Bennett G. J.
      (1995) Synthetic ω-conopeptides applied to the site of nerve injury suppress neuropathic pains in rats. J. Pharmacol. Exp. Ther. 274:666–672.
      Abstract/FREE Full Text
    6. ↵
      1. Ophoff R. A.,
      2. Terwindt G. M.,
      3. Vergouwe M. N.,
      4. van Eijk R.,
      5. Oefner P. J.,
      6. Hoffman S. M. G.,
      7. Lmerdin J. E.,
      8. Mohrenweiser H. W.,
      9. Bulman D. E.,
      10. Ferrari M.,
      11. Haan J.,
      12. Lindhout D.,
      13. van Ommen G.-J. B.,
      14. Hofker M. H.,
      15. Ferrari M. D.,
      16. Frants R. R.
      (1996) Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene CACNL1A4. Cell 87:543–552.
      CrossRefMedlineWeb of Science
    7. ↵
      1. Grabner M.,
      2. Wang Z.,
      3. Hering S.,
      4. Striessnig J.,
      5. Glossmann H.
      (1996) Transfer of 1,4-dihydropyridine sensitivity from L-type to class A (BI) calcium channels. Neuron 16:207–218.
      CrossRefMedlineWeb of Science
    8. ↵
      1. Peterson B. Z.,
      2. Tanada T. N.,
      3. Catterall W. A.
      (1996) Molecular determinants of high affinity dihydropyridine binding in L-type calcium channels. J. Biol. Chem. 271:5293–5296.
      Abstract/FREE Full Text
    9. ↵
      1. Schuster A.,
      2. Lacinova L.,
      3. Klugbauer N.,
      4. Ito H.,
      5. Birnbaumer L.,
      6. Hofmann F.
      (1996) The IVS6 segment of the L-type calcium channel is critical for the action of dihydropyridines and phenylalkylamines. EMBO J. 15:2365–2370.
      MedlineWeb of Science
    10. ↵
      1. Hockerman G. H.,
      2. Johnson B. D.,
      3. Scheuer T.,
      4. Catterall W. A.
      (1995) Molecular determinants of high affinity phenylalkylamine block of L-type calcium channels. J. Biol. Chem. 270:22119–22122.
      Abstract/FREE Full Text
    11. ↵
      1. Doring F.,
      2. Degtiar V. E.,
      3. Grabner M.,
      4. Striessnig J.,
      5. Hering S.,
      6. Glossmann H.
      (1996) Transfer of L-type calcium channel IVS6 segment increases phenylalkylamine sensitivity of α1A. J. Biol. Chem. 271:11745–11749.
      Abstract/FREE Full Text
    12. ↵
      1. Hering S.,
      2. Aczel S.,
      3. Grabner M.,
      4. Doring F.,
      5. Berjukow S.,
      6. Mitterdorfer J.,
      7. Sinnegger M. J.,
      8. Striessnig J.,
      9. Degtiar V. E.,
      10. Wang Z.,
      11. Glossmann H.
      (1996) Transfer of high sensitivity for benzothiazepine from L-type to class A calcium channels. J. Biol. Chem. 271:24471–24475.
      Abstract/FREE Full Text
    13. ↵
      1. Kraus R.,
      2. Reichl B.,
      3. Kimball S. D.,
      4. Grabner M.,
      5. Murphy B. J.,
      6. Catterall W. A.,
      7. Striessnig J.
      (1996) Identification of benz(othi)azepine binding regions within L-type calcium channel α1 subunits. J. Biol. Chem. 271:20113–20118.
      Abstract/FREE Full Text
      1. Diochot S.,
      2. Richard S.,
      3. Baldy-Moulinier M.,
      4. Nargeot J.,
      5. Valmier J.
      (1995) Dihydropyridines, phenylalkylamines and benzothiazepines block N-, P/Q- and R-type calcium currents. Pflueg. Arch. Eur. J. Physiol. 431:10–19.
      CrossRefMedlineWeb of Science
      1. Ishibashi H.,
      2. Yatani A.,
      3. Akaike N.
      (1995) Block of P-type Ca2+ channels in freshly dissociated rat cerebellar Purkinje neurons by diltiazem and verapamil. Brain Res. 695:88–91.
      CrossRefMedlineWeb of Science
    16. ↵
      1. Hille B.
      (1977) Local anesthetics: hydrophilic and hydrophobic pathways for the drug-receptor reaction. J. Gen. Physiol. 69:497–515.
      Abstract/FREE Full Text
    17. ↵
      1. Hondeghem L. M.,
      2. Katzung B. G.
      (1977) Time- and voltage-dependent interactions of antiarrhythmic drugs with cardiac sodium channels. Biochim. Biophys. Acta 472:373–398.
      Medline
    18. ↵
      1. Starmer C. F.,
      2. Grant A. O.,
      3. Strauss H. C.
      (1984) Mechanisms of use-dependent block of sodium channels in excitable membranes by local anesthetics. Biophys. J. 46:15–27.
      MedlineWeb of Science
    19. ↵
      1. Randall A.,
      2. Tsien R. W.
      (1995) Pharmacological dissection of multiple types of Ca2+ channel currents in rat cerebellar granule neurons. J. Neurosci. 15:2995–3012.
      Abstract
    20. ↵
      1. Perez-Reyes E.,
      2. Castellano A.,
      3. Kim H. S.,
      4. Bertrand P.,
      5. Baggstrom E.,
      6. Lacerda A. E.,
      7. Wei X. Y.,
      8. Birnbaumer L.
      (1992) Cloning and expression of a cardiac/brain β subunit of the L-type calcium channel. J. Biol. Chem. 267:1792–1797.
      Abstract/FREE Full Text
    21. ↵
      1. Tomlinson W. J.,
      2. Stea A.,
      3. Bourinet E.,
      4. Charnet P.,
      5. Nargeot J.,
      6. Snutch T. P.
      (1993) Functional properties of a neuronal class C L-type calcium channel. Neuropharmacology 32:1117–1126.
      CrossRefMedlineWeb of Science
    22. ↵
      1. Wei X. Y.,
      2. Perez-Reyes E.,
      3. Lacerda A. E.,
      4. Schuster G.,
      5. Brown A. M.,
      6. Birnbaumer L.
      (1991) Heterologous regulation of the cardiac Ca2+ channel α1 subunit by skeletal muscle β and γ subunits: implications for the structure of cardiac L-type Ca2+ channels. J. Biol. Chem. 266:21943–7.
      Abstract/FREE Full Text
    23. ↵
      1. Soong T. W.,
      2. Stea A.,
      3. Hodson C. D.,
      4. Dubel S. J.,
      5. Vincent S. R.,
      6. Snutch T. P.
      (1993) Structure and functional expression of a member of the low voltage-activated calcium channel family. Science (Washington D. C.) 260:1133–1136.
      Abstract/FREE Full Text
    24. ↵
      1. Starr T. V.,
      2. Prystay W.,
      3. Snutch T. P.
      (1991) Primary structure of a calcium channel that is highly expressed in the rat cerebellum. Proc. Natl. Acad. Sci. USA 88:5621–5625.
      Abstract/FREE Full Text
    25. ↵
      1. Stea A.,
      2. Tomlinson W. J.,
      3. Soong T. W.,
      4. Bourinet E.,
      5. Dubel S. J.,
      6. Vincent S. R.,
      7. Snutch T. P.
      (1994) Localization and functional properties of a rat brain α1A calcium channel reflect similarities to neuronal Q- and P-type channels. Proc. Natl. Acad. Sci. USA 91:10576–10580.
      Abstract/FREE Full Text
    26. ↵
      1. Gorman C. M.,
      2. Gies D. R.,
      3. McCray G.
      (1990) Transient production of proteins using an adenovirus transformed cell line. DNA Prot. Eng. Techniques 2:3–10.
    27. ↵
      1. Dhallan R. S.,
      2. Yau K. W.,
      3. Schrader K. A.,
      4. Reed R. R.
      (1990) Primary structure and functional expression of a cyclic nucleotide-activated channel from olfactory neurons. Nature (Lond.) 347:184–187.
      CrossRefMedline
    28. ↵
      1. Kaufman R. J.
      (1985) Identification of the components necessary for adenovirus translational control and their utilization in cDNA expression vectors. Proc. Natl. Acad. Sci. USA 82:689–693.
      Abstract/FREE Full Text
    29. ↵
      1. Wei X.,
      2. Neely A.,
      3. Lacerda A. E.,
      4. Olcese R.,
      5. Stefani E.,
      6. Perez-Reyes E.,
      7. Birnbaumer L.
      (1994) Modification of Ca2+ channel activity by deletions at the carboxyl terminus of the cardiac α1 subunit. J. Biol. Chem. 269:1635–1640.
      Abstract/FREE Full Text
    30. ↵
      1. Erlich H. A.
      1. Higuchi R.
      (1989) Using PCR to engineer DNA. in PCR Technology, ed Erlich H. A. (Stockton Press, Basingstoke, Hants, UK), pp 61–70.
    31. ↵
      1. Lee K. S.,
      2. Tsien R. W.
      (1983) Mechanism of calcium channel blockade by verapamil, D600, diltiazem and nitrendipine in single dialysed heart cells. Nature (Lond.) 302:790–794.
      CrossRefMedline
    32. ↵
      1. Bean B. P.
      (1984) Nitrendipine block of cardiac calcium channels: high-affinity binding to the inactivated state. Proc. Natl. Acad. Sci. USA 81:6388–6392.
      Abstract/FREE Full Text
    33. ↵
      1. Kass R. S.,
      2. Krafte D. S.
      (1987) Negative surface charge density near heart calcium channels: relevance to block calcium channels. J. Gen. Physiol. 89:629–644.
      Abstract/FREE Full Text
    34. ↵
      1. McDonald T. F.,
      2. Pelzer D.,
      3. Trautwein W.
      (1984) Cat ventricular muscle treated with D600: effects on calcium and potassium currents. J. Physiol. (Lond) 352:203–216.
      Abstract/FREE Full Text
    35. ↵
      1. Bezprozvanny I.,
      2. Tsien R. W.
      (1995) Voltage-dependent blockade of diverse types of voltage-gated Ca2+ channels expressed in Xenopus oocytes by the Ca2+ channel antagonist mibefradil (Ro 40–5967). Mol. Pharmacol. 48:540–549.
      Abstract
    36. ↵
      1. Mori Y.,
      2. Friedrich T.,
      3. Kim M.-S.,
      4. Mikami A.,
      5. Nakai K.,
      6. Ruth P.,
      7. Rosse E.,
      8. Hoffmann F.,
      9. Flockerzi V.,
      10. Furuichi T.,
      11. Mikoshiba K.,
      12. Imoto K.,
      13. Tanabe T.,
      14. Numa S.
      (1991) Primary structure and functional expression from complementary DNA of brain calcium channel. Nature (Lond.) 350:398–402.
      CrossRefMedline
    37. ↵
      1. Ruth P.,
      2. Rohrkasten A.,
      3. Miel M.,
      4. Bosse E.,
      5. Regulla S.,
      6. Meyer H. E.,
      7. Flockerzi V.,
      8. Hoffmann F.
      (1989) Primary structure of the β subunit of the DHP-sensitive calcium channel from skeletal muscle. Science (Washington D.C.) 245:1115–1118.
      Abstract/FREE Full Text
    38. ↵
      1. Sakurai T.,
      2. Westenbroek R. E.,
      3. Rettig J.,
      4. Hell J.,
      5. Catterall W. A.
      (1996) Biochemical properties and subcellular distribution of the BI and rBA isoforms of α1A subunits of brain calcium channels. J. Cell Biol. 134:511–528.
      Abstract/FREE Full Text
    39. ↵
      1. Avenet P.,
      2. Biton B.,
      3. Depoortere H.,
      4. Scatton B.
      (1996) Neuroprotective compound eliprodil (SL 82.0715) blocks N-, P- but not L-type Ca2+ channels in rat cultured cerebellar granule cells. Soc. Neurosci. Abstr. 22:689.5.
    40. ↵
      1. Ragsdale D.,
      2. McPhee J. C.,
      3. Scheuer T.,
      4. Catterall W. A.
      (1994) Molecular determinants of state-dependent block of Na+ channels by local anesthetics. Science (Washington D. C.) 265:1724–1728.
      Abstract/FREE Full Text
    « Previous | Next Article »Table of Contents

    This Article

    1. Molecular Pharmacology May 1, 1997 vol. 51 no. 5 872-881
    1. AbstractFree
    2. » Full Text
    3. Full Text (PDF)

    Classifications

    Responses

    1. Submit a response
    2. No responses published

    Navigate This Article

    Current Issue

    1. November 2009, 76 (5)
    1. Current Issue
    1. Alert me to new issues of Molecular Pharmacology