Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

High voltage-activated Ca²⁺ channels in excitable cells such as myocytes, smooth muscle cells, and neurons play important roles, including contraction of myocytes, electrical excitement in neurons, and modulation of hormone and neurotransmitter release (Tsien et al., 1991). High voltage-activated Ca²⁺ channels are pharmacologically classified into at least five different subclasses (L-, N-, P-, Q-, and R-type), the characteristics of which are determined by the pore-forming ₁ subunit. The subunits _1C, _1D, and _1S form L-type Ca²⁺ channels and bind dihydropyridines (DHPs), phenylalkylamines, and benzothiazepines with high affinity, whereas the subunits _1B, _1A, and _1E form N-, P/Q-, and R-type Ca²⁺ channels, respectively, which show low affinities for these drugs (Hockerman et al., 1997; Hering et al., 1998; Striessnig et al., 1998). Because nifedipine, the prototype of the DHPs, exclusively blocked muscular L-type Ca²⁺ channels (Fleckenstein, 1983), DHPs had been considered as selective blockers for L-type channels.

Recent studies have shown, however, that two DHPs, amlodipine (Furukawa et al., 1997) and cilnidipine (Fujii et al., 1997; Uneyama et al., 1997), blocked N-type Ca²⁺ channels as well. These findings indicate that DHPs are no longer considered L-type specific blockers, and suggest that some DHPs may block other subtypes of Ca²⁺ channels, such as P/Q-, and R-types. DHPs are widely used clinically in the treatment of hypertension, angina pectoris, and cerebrovascular diseases. However, pharmacological profiles of the effects of DHP on each Ca²⁺ channel subtype are not understood well enough for them to be used with confidence with these drugs.

Non-L-type Ca²⁺ channels are diversely distributed in peripheral and central nervous cells (Tsien et al., 1991). However, native neuronal cells and cell lines possess multiple subtypes of Ca²⁺ channels in a single cell, which hampers quantitative comparison of effects of a given drug on a single subtype of Ca²⁺ channel. To address these issues at the molecular level, a single subclass of the Ca²⁺ channel ₁ subunit (_{1A, 1B}, _1C, or _1E) was coexpressed with the same auxiliary ₂ and  subunits in Xenopus oocytes, and 10 DHP derivatives used clinically were examined for their channel-blocking effects.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Methods for in vitro transcription of cRNAs specific to the Ca²⁺ channel ₁, ₂, and _1a subunits and procedures for functional expression of Ca²⁺ channels in Xenopus oocytes were described previously (Furukawa et al., 1998). After removal of the follicular cell layer, Xenopus oocytes were injected with 0.3 µg/µl ₁ [_1C (Mikami et al., 1989), _1B (Fujita et al., 1993), _1A (Mori et al., 1991), or _1E (Niidome et al., 1992)] cRNA in combination with 0.2 µg/µl ₂ (Mikami et al., 1989) cRNA and 0.1 µg/µl _1a (Mori et al., 1991) cRNA. In some experiments, the cRNA for the _2b or ₃ subunit (Hullin et al., 1992) was used instead of that for _1a subunit.

The oocytes were cultured for 2 to 4 days and then subjected to electrophysiological measurement. The oocytes were placed in a small chamber perfused with extracellular solution (10 mM Ba²⁺, 90 mM Na⁺, 2 mM K⁺, 5 mM HEPES, and 0.3 mM niflumic acid, pH 7.5, with methanesulfonic acid), and Ba²⁺ currents through expressed channels were measured by the two-microelectrode voltage-clamp method with a GeneClamp 500 amplifier (Axon Instruments, Foster City, CA). The experimental chamber was 0.5 ml in volume, and it was perfused continuously (2.0-3.0 ml/min) with the extracellular solution. Commercial software (pClamp version 6.0; Axon Instruments) was used for generating voltage pulses, acquiring data, and analyzing the currents. Typically, oocytes were clamped at a holding potential of 100, 80, or 60 mV and depolarized to +10 mV for 200 ms every 15 s. Microelectrodes were filled with 3 M KCl, and those showing a resistance of 0.5 to 1.2 M were used.

The drug effects were evaluated after a 5-min perfusion of bath solution containing a DHP derivative. In experiments to obtain concentration-response relationships, the concentrations of the DHPs were changed successively. Each experiment was finished within 20 min to avoid possible run down of Ba²⁺ currents. Because no detectable current change was observed during exposure to the vehicle for DHP [0.2% dimethyl sulfoxide (DMSO)] as reported previously (Furukawa et al., 1997), the concentration of DMSO in the bath solution was maintained at 0.2% throughout the experiments. DHPs were dissolved into DMSO just before each experiment and added to the bath solution to make the final concentration. The following DHPs were generous gifts from pharmaceutical companies: amlodipine from Sumitomo Pharmaceuticals Co., Ltd. (Tokyo, Japan); benidipine from Kyowa-Hakko Pharmaceuticals Co., Ltd. (Tokyo, Japan); barnidipine from Yamanouchi Seiyaku Co., Ltd. (Tokyo, Japan); cilnidipine from Ajinomoto Co., Ltd. (Tokyo, Japan); felodipine from Hoechst-Marion-Roussel (Tokyo, Japan); and nilvadipine from Fujisawa Pharmaceuticals Co., Ltd. (Tokyo, Japan). Other drugs were purchased from Sigma Chemical Co. (St. Louis, MO) unless otherwise noted. Statistical data were represented by the mean ± S.E.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

As reported previously (Furukawa et al., 1998), injection of cRNA specific for Ca²⁺ channel ₁ subunit (_1B, _1A, or _1C) in combination with cRNAs for the Ca²⁺ channel ₂ and _1a subunits resulted in functional expressions of Ca²⁺ channels that possessed the native characteristics of -conotoxin GVIA-sensitive N-type, -agatoxin IVA-sensitive P/Q-type, and nifedipine-sensitive L-type channels, respectively. In addition, N-type (_1B21a) and P/Q-type (_1A21a) channels were not blocked by 10 µM nifedipine (Furukawa et al., 1998). In oocytes injected with cRNAs for Ca²⁺ channel _1E, ₂, and _1a subunits, inward currents through _1E21a channels were not blocked by any of these toxins or nifedipine (n = 5), which is a characteristic feature of the R-type Ca²⁺ channel (Birnbaumer et al., 1994).

To conduct an effective screening of the selectivities of the 10 DHP derivatives (nifedipine, nilvadipine, barnidipine, nimodipine, nitrendipine, amlodipine, nicardipine, benidipine, felodipine, and cilnidipine) in blocking each subtype of Ca²⁺ channel (Table 1; Fig. 1), we first examined the effects of 10 µM DHPs on the channel subtypes at a holding potential of 80 mV (Fig. 2). All of the DHPs examined inhibited L-type (_1C21a) Ca²⁺ channels by 30 to 65%, whereas none of them blocked R-type (_1E21a) channels more than 10%. In contrast to R-type channels, N-type (_1B21a) and P/Q-type (_1A21a) channels were appreciably blocked by 5 DHPs among the 10 tested (barnidipine, amlodipine, nicardipine, benidipine, and cilnidipine). The blocking action of these DHPs on N- and P/Q-type channels was not related to the blocking potency for L-type channels. Moreover, each DHP shared a blocking action on both N- and P/Q-type channels. The remaining DHPs (nifedipine, nilvadipine, nimodipine, nitrendipine, and felodipine) scarcely blocked the two channel subtypes. In the case of amlodipine, N-type channels were more profoundly inhibited than were L- and P/Q-type channels. Thus, barnidipine, amlodipine, nicardipine, benidipine, and cilnidipine were not categorized as L-type-selective DHPs, and their effects on N- and P/Q-type channels were investigated in more detail.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Structures of the 10 DHPs tested.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Comparison of the potencies of various DHPs in blocking four subtypes of Ca2+ channels: L-, N-, P/Q-, and R-types. The oocytes were clamped at a holding potential of 80 mV and were depolarized to +10 mV for 200 ms every 15 s. Blocking of peak inward currents after 5-min perfusion of each DHP (10 µM) was measured. The DHPs are ordered according to the blocking potencies observed for L-type Ca2+ channels.

The Ca²⁺ channel  subunit is known to modulate channel kinetics and DHP sensitivity, and we used the skeletal muscle form of the  subunit. To get insight into the modulation of DHP effect by -subunit subclass, we compared the effect of amlodipine on _1B21a, _1B22b, and _1B23 channels. The block of Ba²⁺ current at a holding potential of 80 mV by 10 µM amlodipine were 78.4 ± 7.2 (n = 12) for the _1B21a channel, 69.6 ± 10.2 for the _1B22b channel (n = 7), and 71.2 ± 13.2 for the _1B23 channel (n = 8), respectively. The blocking actions of amlodipine on N-type Ca²⁺ channels with different -subunit subclasses were not substantially different.

The blockade of L-type Ca²⁺ channels by amlodipine is known to develop slowly (Kass and Arena, 1989). Under the present experimental conditions (at a holding potential of 80 mV), the time constant of blocking by 10 µM amlodipine was 42.8 ± 7.8 s (n = 8) for an N-type channel or 45.8 ± 5.5 s (n = 6) for a P/Q-type channel. At a holding potential of 100 mV, the time constant of blocking for an N-type Ca²⁺ channel was 1 min (Furukawa et al., 1997). Considering these time constants of blocking, we determined that perfusion of amlodipine for 5 min was enough to reach steady-state blocking. Blockades of N- and P/Q-type channels by other DHPs developed much faster, and a steady-state inhibition was reached within 1 min (n = 5-12).

Figure 3 shows the effects of the aforementioned five DHPs on the current-voltage (I-V) relationships of N-type Ca²⁺ channels. These DHPs blocked Ba²⁺ currents through the N-type (_1B21a) channels at each membrane potential without shifting peak I-V relationships. Similar results were obtained for P/Q-type (_1A21a) Ca²⁺ channels (Fig. 4).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Effects of five DHPs on I-V relationships of N-type Ca2+ channels. Membrane currents were elicited by step depolarizations of 200-ms duration from holding potential of 80 to +50 or +70 mV every 10 mV. Membrane currents in response to the step pulse to +10 mV before (Control) and after 5-min perfusion of barnidipine (A), amlodipine (B), nicardipine (C), benidipine (D), or cilnidipine (E) are presented with current traces (left) and I-V curves of peak currents (right). Note that these DHPs reduced the membrane currents at each test pulse potential.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Effects of five DHPs on I-V relationships of P/Q-type Ca2+ channels. The same protocols as in Fig. 3 were used. Membrane currents before (Control) and after 5-min perfusion of barnidipine (A), amlodipine (B), nicardipine (C), benidipine (D), or cilnidipine (E) are presented with current traces (left) and I-V curves of peak currents (right). Note that these DHPs also reduced the membrane currents at each test pulse potential, as observed in the case of N-type channels (Fig. 2).

In the next step, we investigated the concentration- and voltage-dependent effects of these five DHPs on L-, N-, and P/Q-type channels. The oocytes expressing L-type Ca²⁺ channel were clamped at holding potentials of 80 and 60 mV, and those expressing N- or P/Q-type Ca²⁺ channels were clamped at holding potentials of 100, 80, and 60 mV. We then measured the blockade of Ba²⁺ currents by DHPs at various concentrations. Figures 5 through 9 summarize the effects of DHP derivatives on these three subtypes of Ca²⁺ channels. To describe the concentration-response relationships in a quantitative way, we performed a least-squares fit to data as follows:
where IC₅₀ is the concentration at half-blockade, [D] is the drug concentration, and n_H is the Hill coefficient. All five DHPs showed a concentration-dependent blocking of both N- and P/Q-type Ca²⁺ channels. In the case of N-type channels (Figs. 5-9, left), the blockade by these DHPs was more potentiated by depolarization of the holding potential, and the voltage-dependence of blocking was not prominent for benidipine, although it was still significant (Fig. 6, left). In contrast to N-type channels, P/Q-type channels were not blocked by benidipine or cilnidipine in a voltage-dependent manner (Figs. 6 and 9, right). Moreover, blockades of P/Q-type channels by amlodipine, nicardipine, and barnidipine were less voltage-dependent than were those of N-type channels (Figs. 5, 7, and 8, right). Amlodipine, barnidipine, and nicardipine had simple concentration- and voltage-dependent blocking actions on L-type Ca²⁺ channels. However, the effect of benidipine and cilnidipine was variable and complex. The effect of benidipine on an L-type Ca²⁺ channel was variable as shown by the S.E. bars in Fig. 6. Depolarizing the holding potential decreased the block at drug concentrations of 1 to 10 µM, and benidipine did not block the Ba²⁺ current completely even at a very high concentration (30 µM) and at a holding potential of 60 mV. Moreover, cilnidipine at concentrations of 1 to 10 µM enhanced the Ba²⁺ current in some experiments. The Ba²⁺ current was blocked again at higher concentrations. Benidipine and cilnidipine are known to have both inhibitory and stimulatory effects on L-type Ca²⁺ channels (Terada et al., 1987; Oike et al., 1990), which may contribute to these complex concentration-response relationships.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves for barnidipine at holding potentials of 100 (), 80 (), and 60 mV (). The IC50 and nH for an L-type Ca2+ channel are 3.1 ± 0.8 µM and 0.5 ± 0.1 at 80 mV (n = 7), and 1.2 ± 0.3 µM and 0.7 ± 0.1 at 60 mV (n = 6), respectively. Those for an N-type Ca2+ channel are 1370 ± 499 µM and 0.4 ± 0.1 at 100 mV (n = 6), 74.9 ± 29.2 µM and 0.4 ± 0.2 at 80 mV (n = 6), and 7.1 ± 2.6 µM and 0.6 ± 0.1 at 60 mV (n = 6), respectively. Those for a P/Q-type Ca2+ channel are 213 ± 54 µM and 0.8 ± 0.2 at 100 mV (n = 6), 40.3 ± 8.9 µM and 0.9 ± 0.2 at 80 mV (n = 6), and 13.1 ± 1.8 µM and 1.1 ± 0.3 at 60 mV (n = 6), respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Concentration-response curves for amlodipine at holding potentials of 100 (), 80 (), and 60 mV (). The IC50 and nH for an L-type Ca2+ channel are 4.2 ± 3.1 µM and 0.7 ± 0.1 at 80 mV (n = 12) and 1.2 ± 0.6 µM and 1.0 ± 0.2 at 60 mV (n = 11), respectively. Those for an N-type Ca2+ channel are 7.9 ± 2.6 µM and 0.7 ± 0.2 at 100 mV (n = 14), 1.9 ± 0.8 µM and 0.7 ± 0.1 at 80 mV (n = 12), and 0.14 ± 0.05 µM and 0.5 ± 0.1 at 60 mV (n = 8), respectively. Those for a P/Q-type Ca2+ channel are 11.5 ± 2.4 µM and 1.0 ± 0.2 at 100 mV (n = 7), 7.3 ± 1.4 µM and 0.8 ± 0.1 at 80 mV (n = 7), and 3.0 ± 0.9 µM and 0.8 ± 0.2 at 60 mV (n = 6), respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Concentration-response curves for nicardipine at holding potentials of 100 (), 80 (), and 60 mV (). The IC50 and nH values for an L-type Ca2+ channel are 24.1 ±1.9 µM and 0.4 ± 0.1 at 80 mV (n = 8) and 9.6 ± 0.6 µM and 0.7 ± 0.1 at 60 mV (n = 7), respectively. Those for an N-type Ca2+ channel are 7590 ± 1320 µM and 0.4 ± 0.1 at 100 mV (n = 6), 201 ± 49 µM and 0.6 ± 0.1 at 80 mV (n = 6), and 59.9 ± 10.2 µM and 0.6 ± 0.1 at 60 mV (n = 14), respectively. Those for a P/Q-type Ca2+ channel are 85.0 ± 12.5 µM and 0.9 ± 0.2 at 100 mV (n = 14), 39.6 ± 6.8 µM and 1.1 ± 0.2 at 80 mV (n = 14), and 21.1 ± 1.0 µM and 1.1 ± 0.3 at 60 mV (n = 14), respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Concentration-response curves for benidipine at holding potentials of 100 (), 80 (), and 60 mV (). The IC50 and nH for an L-type Ca2+ channel are 13.8 ± 4.5 µM and 0.3 ± 0.1 at 80 mV (n = 13) and 4.9 ± 1.5 µM and 0.5 ± 0.1 at 60 mV (n = 10), respectively. Those for an N-type Ca2+ channel are 72.0 ± 23.6 µM and 0.7 ± 0.1 at 100 mV (n = 6), 29.5 ± 5.8 µM and 0.7 ± 0.2 at 80 mV (n = 14), and 35.3 ± 7.1 µM and 0.4 ± 0.1 at 60 mV (n = 6), respectively. Those for a P/Q-type Ca2+ channel are 43.8 ± 17.9 µM and 0.8 ± 0.2 at 100 mV (n = 7), 50.4 ± 16.4 µM and 0.7 ± 0.1 at 80 mV (n = 7), and 33.2 ± 6.2 µM and 0.8 ± 0.2 at 60 mV (n = 14), respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 9.   Concentration-response curves for cilnidipine at holding potentials of 100 (), 80 (), and 60 mV (). The IC50 and nH for an L-type Ca2+ channel are 5.3 ± 2.1 µM and 0.2 ± 0.1 at 80 mV (n = 10) and 12.7 ± 5.8 µM and 0.9 ± 0.4 at 60 mV (n = 9), respectively. Those for an N-type Ca2+ channel are 39.4 ± 7.9 µM and 0.8 ± 0.2 at 100 mV (n = 6), 18.8 ± 3.2 µM and 0.7 ± 0.2 at 80 mV (n = 6), and 4.2 ± 1.7 µM and 0.5 ± 0.1 at 60 mV (n = 6), respectively. Those for a P/Q-type Ca2+ channel are 22.6 ± 7.3 µM and 0.5 ± 0.1 at 100 mV (n = 6), 58.5 ± 9.0 µM and 0.4 ± 0.1 at 80 mV (n = 6), and 20.8 ± 7.4 µM and 0.4 ± 0.1 at 60 mV (n = 6), respectively.

To compare the selectivity of blocking action of DHPs on L-, N-, and P/Q-type Ca²⁺ channels, the IC₅₀ values for these three Ca²⁺ channels are summarized in Fig. 10. Because voltage dependence of the blocking action on a channel subtype was different in each DHP, we could not simply compare the selectivities of DHPs in blocking Ca²⁺ channel subtypes. At the holding potential of 80 mV, the rank order of the IC₅₀ values for an N-type Ca²⁺ channel was amlodipine < cilnidipine < benidipine < barnidipine < nicardipine. The ranking was not changed except the order of benidipine and nicardipine at the holding potential of 60 mV. The IC₅₀ values for amlodipine effect on a P/Q-type Ca²⁺ channel were smaller compared with the other DHP compounds. The IC₅₀ for nicardipine effect for a P/Q type Ca²⁺ channel was considerably smaller than that for an N-type Ca²⁺ channel. The other three DHPs (benidipine, cilnidipine, and barnidipine) showed similar IC₅₀ values for both N- and P/Q-type Ca²⁺ channels. The IC₅₀ values were not correlated with molecular weight, solubility, or acid dissociation constant (pK_a) value for DHPs (Table 1). The IC₅₀ values for the effects of these five DHPs on an L-type Ca²⁺ channel stayed in a narrow range compared with those for N- and P/Q-type Ca²⁺ channels at the holding potentials of 80 and 60 mV. Thus, amlodipine blocked all three Ca²⁺ channel subtypes with comparable potency, and nicardipine had lower affinity to N-type Ca²⁺ channels compared with the other four DHPs (benidipine, cilnidipine, barnidipine and amlodipine). Benidipine, cilnidipine, and barnidipine showed similar selectivities for N- and P/Q-type Ca²⁺ channels.

View larger version (25K): [in this window] [in a new window]   Fig. 10.   Comparison of IC50 values for the effect of five DHP derivatives on N-, P/Q-, and L-type Ca2+ channels. IC50 values for three Ca2+ channel subtypes are plotted against the same semilogarithmic concentration axis.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, five DHP derivatives (amlodipine, benidipine, cilnidipine, nicardipine, and barnidipine) differentially blocked L-, N-, and P/Q-type Ca²⁺ channels that were expressed functionally in Xenopus oocytes, an in vivo expression system. These findings indicate that some DHPs are not selective antagonists for L-type channels. Although our data covered only narrow and hyperpolarized membrane potentials, amlodipine showed high affinity for both N- and P/Q-type channels, and nicardipine preferentially antagonized the P/Q-type channel. Moreover, blockades by benidipine and cilnidipine were voltage-dependent in N-type channels but not in P/Q-type channels. However, further systematic studies are necessary to clarify the channel subtype selectivity of the DHPs because the voltage-dependent effects of the five DHPs on L-type Ca²⁺ channels were different and complex. These findings are consistent with the observations of blocking of N-type channels by cilnidipine in native tissues, dorsal root ganglion neurons (Fujii et al., 1997), and sympathetic neurons (Uneyama et al., 1997). The other five DHPs (nifedipine, nilvadipine, nimodipine, nitrendipine, and felodipine) were strongly selective for L-type channels. Thus, we were able to categorize five DHPs (nifedipine, nilvadipine, nimodipine, nitrendipine, and felodipine) as L-type-selective DHPs and the other five DHPs (amlodipine, benidipine, cilnidipine, nicardipine, and barnidipine) as L-type-nonselective DHPs. These categorizations indicate the importance of determining which DHPs block which subtypes of Ca²⁺ channels for the therapeutic application of these drugs.

In this study, we investigated the DHP effect on the Ca²⁺ channels with the same combination of auxiliary Ca²⁺ channel subunits. Although the ₂ and  subunit combination is known to modulate the DHP action on L-type Ca²⁺ channels (Mitterdorfer et al., 1994; Wei et al., 1995; Welling et al., 1995), we showed that the effect of amlodipine on N-type Ca²⁺ channels was not substantially modulated by exchanging the  subunit subclass (_1a, _2b, or ₃). Therefore, these differential blockades of channel subtypes by DHPs are most possibly accounted for by the structural differences of the ₁ subunits (Hockerman et al., 1997; Hering et al., 1998; Striessnig et al., 1998). Nine amino acid residues in the segments IIIS5, IIIS6, and IVS6 of the _1C or _1S subunit of L-type Ca²⁺ channels have been shown to interact directly with DHPs (Hockerman et al., 1997; Hering et al., 1998; Striessnig et al., 1998). Four of these nine amino acid residues are conserved in the ₁ subunits of non-L-type Ca²⁺ channels, including N-, P/Q-, and R-type channels, and the incorporation of these amino acid sequences made P/Q-type (Sinnegger et al., 1997) and R-type Ca²⁺ channels (Ito et al., 1997) sensitive to DHPs. Taken together with our results that five DHPs (amlodipine, benidipine, cilnidipine, nicardipine, and barnidipine) were able to block N- and P/Q-type channels, these suggest that N- and P/Q-type channels are endowed with a basic structure designed for interaction with the five DHPs. However, R-type channels were not appreciably blocked by any of the DHPs examined. Therefore, amino acid residues other than these conserved four residues may play a critical role in DHP selectivity between these non-L-type channels. The five L-type-nonselective DHPs had larger molecular weights compared with the other DHPs used in this study. Nevertheless, the blocking potency of DHPs on non-L-type channels was not correlated with the structural characteristics of the DHPs, including the degree of ionization (Kass and Arena, 1989) and partition coefficients (Table 1). Further studies using additional DHPs and site-directed mutagenesis on the _1B and _1A subunits will be necessary to determine the direct interaction between specific amino acid residues on the ₁ subunits and specific DHPs.

This is the first report that P/Q-type Ca²⁺ channels, as well as N-type channels, were blocked by DHPs with a voltage dependence. DHPs are known to show a voltage- and state-dependent action on L-type Ca²⁺ channels (Hockerman et al., 1997; Striessnig et al., 1998). DHP antagonists bind to the inactivated state of the Ca²⁺ channel with the highest affinity, resulting in a blockade that is steeply voltage dependent. Cilnidipine and benidipine showed a weak voltage dependence in blocking N- and P/Q-type channels. In contrast, both of these drugs have been shown to block L-type Ca²⁺ channels with a sharp voltage dependence (Oike et al., 1990; Yamamoto et al., 1990). This discrepancy may be accounted for in part by the differences between the inactivation mechanisms of neuronal Ca²⁺ channels and those of L-type Ca²⁺ channels (Patil et al., 1998), in which N-, P/Q-, and R-type channels, but not L-type channels, inactivate profoundly during a train of action potential.

The difference in selectivity of DHPs in blocking Ca²⁺ channel subtypes provides a clue for understanding the different clinical effects of DHPs. Reviewing the reports about neurohormonal modulation by DHPs, we could find possible relationships between the subtype selectivity and the clinical features of the various DHPs. Reflex tachycardia (Ram and Featherston, 1988) is a common adverse effect of most of the L-type selective DHPs such as nifedipine (Scholz, 1997), nitrendipine (Warltier et al., 1984), and nimodipine (Duncker et al., 1988). On the other hand, the L-type nonselective DHPs such as amlodipine (Burges et al., 1989), barnidipine (van Zwieten, 1998), benidipine (Fuji et al., 1988), cilnidipine (Minami et al., 1998b), and nicardipine (Chen et al., 1995) have been shown not to increase heart rate significantly. However, this DHP effect on heart rate may be closely related to the pharmacokinetics of DHPs because felodipine, one of the L-type selective DHPs, does not increased heart rate (Bicchi et al., 1998).

Three DHPsamlodipine (Abernethy et al., 1988), barnidipine (Argenziano et al., 1998), and cilnidipine (Minami et al., 1998b)are also known to decrease the serum catecholamine level. In addition, it has been reported that amlodipine and cilnidipine reduce sympathetic tones as measured by heart rate variability (Hamada et al., 1998; Minami et al., 1998a,b). To our knowledge, these kinds of sympatholytic reactions have not been reported for L-type-selective DHPs. The blocking action of these DHPs on N-type channels may be related to the characteristics of autonomic modulation. This is consistent with the fact that analgesic effects have been shown in -conotoxin (Miljanich and Ramachandran, 1995), a peptide blocker of N-type channels, as well as amlodipine (Dogrul et al., 1997), which blocked N-type channels most profoundly among the L-type-nonselective DHPs. In addition, the L-type-selective antagonists, such as nifedipine, verapamil, and diltiazem, are considerably less potent in inducing antinociception (Miljanich and Ramachandran, 1995).

P/Q-type channel malfunction recently has been implicated in the pathophysiology of certain forms of migraine, ataxia, and epilepsy (Doyle and Stubbs, 1998; Ophoff et al., 1998). Although the in vivo effects of DHP action on P/Q-type Ca²⁺ channels are totally unknown, these Ca²⁺ channels represent therapeutic targets for novel drugs. For this reason, the major challenge for Ca²⁺ channel pharmacologists is to develop selective nonpeptide modulators of certain non-L-type channels. The findings in this study would provide a new insight into the classification of DHPs based on the selectivity in Ca²⁺ channel subtypes, as well as helpful information for determining selective DHPs for every non-L-type channel.