Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

L-type calcium channels are heteromultimeric proteins consisting of ₁, ₂, ₂/, and  subunits (Hofmann et al., 1994; Catterall, 1996). The ₁ subunit contains the binding sites for the three major classes of calcium channel modulators [phenylalkylamines, dihydropyridines (DHPs), and benzothiazepines; Striessnig et al., 1991]. When expressed in heterologous expression systems, this subunit is sufficient to encode channels with most of the biophysical and pharmacological properties of native channels (Mikami et al., 1989; Welling et al., 1993). Thus, studies of the molecular site(s) and mechanisms of action of calcium channel blockers have focused on chimeric and mutational analysis of the ₁ subunit (Varadi et al., 1995). In the case of DHPs, the most potent and selective calcium channel blocker (Kass, 1982), these studies have been restricted only to the neutral forms of these important compounds.

Initial photoaffinity labeling and antibody mapping experiments localized the receptor site for neutral DHPs to peptides that form domains III and IV of the ₁ subunit (Striessnig et al., 1991; Regulla et al., 1991). Mutational studies subsequently have revealed roles of shorter amino acid sequences and individual amino acid residues in the determination of high-affinity block and/or binding of neutral DHPs. In all cases, studies were carried out with the neutral drug isradipine. Several groups have now identified amino acids in IIIS5, IIIS6, and IVS6 transmembrane segments whose replacement by corresponding amino acids of DHP-insensitive ₁ subunits decreases DHP binding and/or inhibition (Tang et al., 1993; Grabner et al., 1996; Peterson et al., 1996; Schuster et al., 1996; Ito et al., 1997). Alanine-scanning mutagenesis has confirmed the importance of these residues and additionally revealed several amino acids conserved between DHP-sensitive and DHP-insensitive channels that contribute to neutral DHP activity (Peterson et al., 1997). An inverse approach similarly has been used to reconstruct a high-affinity DHP interaction site in DHP-insensitive _1A subunits (Grabner et al., 1996; Ito et al., 1997).

These findings support a proposed domain-interface model in which neutral DHPs act as allosteric effectors. After binding to the interface between domains III and IV of the ₁ subunit, the drugs alter domain interactions and subsequently influence channel gating (Hockerman et al., 1997b). In this model, estimation of the physical relationship between the key DHP binding residues and the outer surface of the calcium channel pore was based in part on data obtained with charged DHP derivatives and native L-type calcium channels (Bangalore et al., 1994). However, previous studies have shown that addition of a charged group to a DHP compound markedly alters the functional modulation of channel activity by these drugs (Bangalore et al., 1994).

These functional changes in channel modulation may be a consequence of restricted access to a common receptor binding domain for charged, but not neutral, forms of the drug as predicted within the framework of the modulated receptor hypothesis (Hille, 1977; Hondeghem and Katzung, 1977). However, it is also possible that addition of a charged group to a DHP molecule may restrict access of the DHP moiety to some or all of the residues that have been shown to constitute the neutral drug binding domain and/or underlie neutral drug-channel interactions. This restriction to a subset of key residues, could confer distinct functional consequences upon the drug-bound channel.

To distinguish between these possibilities, we have systematically investigated the sensitivity of an ionized DHP compound to mutation of individual amino acids of segments IIIS5, IIIS6, and IVS6 of 1c which have previously been shown to markedly reduce channel modulation and/or binding by the neutral compound isradipine. To select key residues that might affect charged DHP interactions we relied on previous site-directed and alanine scanning mutagenesis studies of both 1S and 1c subunits that have revealed important residues underlying neutral drug interactions. We mutated a tyrosine (Tyr1174) and two isoleucine (Ile1175, Ile1178) residues arranged in a YIXXI segment as well as a methionine (Met1183) in IIIS6. We also tested the effects of mutation of a tyrosine (Tyr1485), methionine (Met1486), and one of two isoleucines (Ile1493) in the YM (X)₅I sequence of IVS6 previously shown to be key to neutral DHP interactions (Peterson et al., 1996; Schuster et al., 1996; Bodi et al., 1997; Peterson et al., 1997). Finally, because the recent mutation of a threonine residue in IIIS5 (Thr1061) potently reduced neutral DHP inhibition of electrophysiologically-assayed channel activity as well as radioligand binding to expressed protein (Grabner et al., 1996; Mitterdorfer et al., 1996; Ito et al., 1997), we tested the effects of mutating this residue on charged DHP interactions.

We used custom synthesized DHP derivatives in which either a neutral or permanently charged head group was separated by a fixed length (10) hydrocarbon spacer chain from a common DHP moiety (Bangalore et al., 1994). We studied channel block by the neutral and charged forms of the test drug as a function of channel construct and membrane potential. Our results suggest that the presence of a permanently charged head group restricts the access of the attached DHP moiety to a subset of interaction residues on the _1c subunit in a voltage-dependent manner. Furthermore, these restricted interactions confer distinct functional properties upon the charged DHP molecules.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals

Neutral (DHPn) and charged (DHPch) dihydropyridines were custom synthesized DHP analogs, in which the active moiety was linked to either a neutral (CH₂CH₃) or charged [-⁺N(CH₃)₃] head group by an alkyl spacer chain containing ten methylene (CH₂) groups (Bangalore et al., 1994). This length of spacer chain was shown to optimize current blocking ability of both DHPn and DHPch in native cardiac L-type calcium channels (Bangalore et al., 1994). For structures of these compounds see Baindur et al. (1993) and Fig. 1A. All other chemicals were purchased from Sigma Chemical Corp. (St. Louis, MO).

View larger version (34K): [in this window] [in a new window]   Fig. 1.   A, structure of neutral and permanently charged DHPs used in the experiments. B, approximate position of amino acids exchanged in individual constructs. Only the third and fourth repeats are shown. Exchanged amino acids are highlighted by black letters in white circles.

Chimeras: Construction and Nomenclature. Chimeric channels were constructed by replacement of amino acids of IIIS5, IIIS6, or IVS6 segments of the rabbit _1cb cDNA (Biel et al., 1990) by the corresponding sequence of the _1E cDNA (Schneider et al., 1994). For details on mutational techniques see Schuster et al. (1996). The exchanged amino acids in individual chimeras were as follows (numbering according to the _1cb sequence): in the IIIS5 segment, Thr1061 was replaced by Tyr yielding what we refer to as the EC14 channel. Three amino acids were exchanged in IIIS6 (Ile1175Phe, Ile1178Phe, Met1183Val) to make the EC20 construct. The conserved Tyr1174 in IIIS6 was replaced by Ala to make the EC19 channel construct. Three amino acids (Tyr1485Ile, Met1486Phe, Ile1493Leu) were exchanged in segment IVS6 to make the chimera EC30 (identical with Ch30 in Schuster et al., 1996). The relative positions of amino acids that were mutated are displayed according to the _1cb sequence in Fig. 1B. In the text that follows we refer to the  subunit as _1c for simplicity because all amino acids that we mutated are common to both _1cb and _1ca. splice variants (Biel et al., 1990).

Transfection Procedures. For transient expression in HEK 293 cells, the full-length cDNAs of the chimeric constructs and the wild-type _1cb (WT) were cloned into the pcDNA 3 vector (Invitrogen). HEK 293 cells were transfected with either WT or chimeric plasmids together with the cDNA plasmids encoding the 2a and 2/ subunits by lipofection with Lipofectamine (Gibco BRL, Life Technologies, Grand Island, NY) at a DNA mass ratio of 1:1:1 as described previously (Bangalore et al., 1996).

Whole Cell Recording

In all experiments, K⁺ channel currents were suppressed by substituting Cs⁺ for K⁺ in both external and internal solutions. External sodium was replaced by N-methyl-D-glucamine to rule out the possibility that endogenous Na⁺ channel activity might contribute to our recordings. Lowering external Cl^- concentration minimized chloride currents.

Ionic currents were recorded using whole-cell patch clamp recording conditions with Ba²⁺ as the charge carrier. The extracellular solution contained (in mM): N-methyl-D-glucamine, 125; BaCl₂, 20; CsCl, 5; MgCl₂, 1; HEPES, 10; glucose, 5; pH 7.4 (HCl). The intracellular solution contained (in mM): CsCl, 60; CaCl₂, 1; EGTA, 11; MgCl₂, 1; K₂ATP, 5; HEPES, 10; aspartic acid, 50; pH 7.4 (CsOH). Under these conditions, we can expect that the measured current was carried by Ba²⁺ primarily through expressed calcium channels and consequently is referred to as I_Ba in the text.

Drugs were applied by a local solution changer and reached the cell membrane within 1 s. The effects of the DHPs were tested with 20 ms long voltage-clamp steps to +10 or +20 mV (peak of current-voltage relation for each individual cell) from HPs of 80 mV or 40 mV. Pulse frequency was 0.2 Hz. For each test configuration, drug effects were measured after attaining steady-state block, within 2 to 3 min after drug application.

Curve fitting and statistical analysis were carried out using the Origin 5.0 software package (Microcal Inc., Northampton, MA). The significance of observed differences was evaluated by nonpaired Student's t test. A probability of 5% or less was considered to be significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Mutations in IVS6 and IIIS6: Marked Effects on DHPn but not DHPch

IVS6 Mutations (EC30). We first focused on previously described mutations in IVS6 (EC30, see Materials and Methods and Fig. 1B; Peterson et al., 1996; Schuster et al., 1996) and tested for distinct effects of the mutations on channel block by DHPch and DHPn. In describing these experiments we introduce the strategy that we later used to study each channel construct. Currents were measured in response to pulses applied once every 5 s from HPs that were set at either 40 or 80 mV to provide insight into the role of membrane potential in channel block. In all experiments shown, Ba²⁺ was the charge carrier and we thus describe measured currents as I_Ba.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Differential effects of IVS6 (EC30) mutations on channel block by DHPn and DHPch. Concentration-dependent inhibition of expressed channel activity by DHPn (A, C) and DHPch (B, D) was recorded in WT (A, B) and EC30 mutant (C, D). (), concentration dependence measured from an HP of 80 mV; (), concentration dependence measured from an HP of 40 mV. The solid lines represent the fits of experimental data to the Hill equation (see Materials and Methods). Insets, family of barium currents recorded in control conditions and in presence of 10, 100, and 1000 nM (in order of descending amplitudes) of corresponding DHP. (), currents recorded from an HP of 80 mV; (), currents recorded from an HP of 40 mV. Scale bars in each panel mark 5 ms (horizontal) and 500 pA (vertical).

IIIS6 Mutations (EC20). We next focused on residues in domain IIIS6 as this region of _1c has also shown to be key to neutral DHP modulation of the channel (Grabner et al., 1996; Peterson et al., 1996, 1997; Bodi et al., 1997; Ito et al., 1997). We again compared the effects of these mutations on DHPn and DHPch block of channel activity measured from 40 and 80 mV. As is the case for the EC30 mutations of IVS6, we find little effect of EC20 mutations (defined in Materials and Methods and Fig. 1B) on DHPch block, but significant reduction of DHPn block from both HPs. Figure 3A illustrates the effects of both drugs (100 nM) on WT and EC20 currents recorded from 40 mV. Concentration-response data for EC20 are summarized in Figs. 3B and 5. From these experiments it is clear that mutation of IIIS6 sites that are not conserved between _1E and _1c subunits have much more pronounced effects on neutral compared with charged drug activity in a manner that is very similar to EC30 mutation in IVS6. Figure 3C summarizes experiments designed to test the effects of mutation of Tyr1174 (EC19) that is conserved between _1E and _1c type channels but has been shown to contribute to neutral DHP binding and channel modulation (Peterson et al., 1996, 1997; Bodi et al., 1997; Ito et al., 1997). Again, this mutation significantly reduced DHPn block of channel activity with a more pronounced effect on channel activity measured from 40 mV. However, channel block by DHPch is not inhibited by the EC19 mutation. In fact, the EC19 causes a statistically significant enhancement of I_Ba block for current measured from a 40 mV HP.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Influence of nonconserved IIIS6 amino acid residues (EC20) and conserved Tyr1174 (EC19) on channel block by DHPch and DHPn. A, current traces recorded in response to 40-ms voltage pulses (Vm = +20 mV) applied from a 40 mV HP in cells transfected with WT and EC20 cDNA as marked. Currents are shown in the absence () and presence () of 100 nM DHPn (left) and 100 nM DHPch (right). Complete dose-response curves for EC20 measured from 80 mV HP () or 40 mV HP () are shown in (B) for DHPn (left) and DHPch (right). The EC20 mutation has little effect on the sensitivity of expressed currents to DHPch but not to DHPn. C, currents measured for cells transfected with WT (solid columns) or EC19 (open columns) channels as described in Materials and Methods. Left, compares the effect of 100 nM DHPn on WT and EC19 channel measured from HPs of 80 mV (left) and 40 mV (right). Blocks represent the IBa amplitude measured in the presence of 100 nM DHPn in percentage of IBa amplitude measured under control conditions. *p < .05; ***p < .001. Significance of the difference between block of IBa through WT channel and IBa through EC19 channel was tested using unpaired Student's t test. Right, shows the same comparison as left for 100 nM DHPch. *p < .05 tested by unpaired Student's t test. Inset to (C) shows pairs of IBa through EC19 channel measured in the absence () and in the presence () of 100 nM DHPn (first two pairs) or DHPch (last two pairs). For each drug, left pair of traces was measured at the HP of 80 mV and right pair at the HP of 40 mV. Scale bars represent 10 ms (horizontal) and 100 pA (vertical), except for DHPch HP 80mV traces, where vertical bar represents 200 pA.

Mutation of IIIS5 (EC14): Selective Inhibition of DHPch Block of Currents Recorded from 80 mV

We next studied the effects of a single amino acid mutation (Thr1061) in segment IIIS5 of _1c (EC14) that has been shown to have marked affects on neutral DHP interactions with the channel (Grabner et al., 1996; Mitterdorfer et al., 1996; He et al., 1997; Ito et al., 1997). We tested the effects of the EC14 mutation on channel block by both DHPn and DHPch, and again, measured currents from 40 and 80 mV HPs. The results of these experiments are summarized in Fig. 4.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Mutation of IIIS5 Thr1061: preferential inhibition of DHPch block from 80 mV versus 40 mV. A, columns represent mean ± S.E.M. of relative amplitude of IBa in presence of 3 µM DHPn or DHPch, measured at the HPs of 80 mV or 40 mV, as noted in graph. Current was activated by 20-ms-long pulses from both HPs (as marked) to the peak of the current-voltage relationship with frequency of .2 Hz. 3 to 4 min were allowed to reach the new equilibrium amplitude. Inset shows family of currents recorded from EC14 channel from the HP 40 mV under control conditions (top) and in presence of 1 µM DHPch (bottom). Comparison of block of WT and EC14 channels by DHPn (left) and DHPch (right) is shown for currents recorded from 40 (B) and 80 mV (C). In both (B) and (C), the open symbols (WT) are the data from Fig. 2 replotted for comparison. Filled symbols represent the inhibition of the EC14 channel. The dose-response relationship for the inhibition of the EC14 channel by DHPch was fitted with the Hill equation (Materials and Methods) and the solid line in (B) represents a fit of experimental data from six cells to the Hill equation resulting in IC50 of 161 ± 43 nM and the Hill coefficient of .59 ± 0.09.

The Thr1061Tyr mutation reduces the sensitivity of expressed EC14 channels to DHPn, regardless of the HP from which channel activity is evoked. Fig. 4A illustrates the sensitivity of EC14 channels to block by 3 µM DHPn and 3 µM DHPch. EC14 channel activity, recorded from either HP (80 mV or 40 mV) is very insensitive to block by DHPn. DHPn (3 µM) blocks less than 20% of I_Ba through EC14 channels, in contrast with the complete block of WT channels at this concentration (B). As is the case for other previously investigated neutral DHP compounds, the EC14 mutation markedly reduces DHPn modulation of channel activity, confirming the importance of this residue in DHPn interactions.

In contrast with mutations of IVS6 and IIIS6, the EC14 mutation also has a pronounced effect on DHPch inhibition of I_Ba, but this effect is most evident for channel activity measured from 80 mV. As is the case for DHPn, 3 µM DHPch blocked less than 20% of I_Ba recorded from 80 mV HPs, despite virtually complete inhibition of WT channels recorded under the same voltage conditions (Fig. 4C). This result provides strong evidence that this site of interaction is common between DHPn and DHPch molecules when channel activity is evoked from hyperpolarized HPs.

However, as shown in Fig. 4, the situation is different when the HP is more depolarized. Channel activity evoked from a 40 mV HP is reduced by approximately 80% for EC14 channels, roughly the same percent reduction recorded for WT channels under the same recording conditions. In fact, for this depolarized HP, the concentration-response relationship for DHPch block of I_Ba through EC14 channels is the same as that for block of I_Ba through WT channels (Fig. 4B).

Summary of DHPch Interactions Sites: Evidence for Voltage-Dependent Changes in DHPch Interactions

Figure 5 summarizes the influence of channel construct on the estimated IC₅₀ of I_Ba inhibition by DHPn (left column) and DHPch (right column). Results are shown for experiments in which channel activity was evoked from 40 (top) and 80 mV (bottom). The figure shows that there is a clear difference between the contributions of specific residues to the interactions of DHPn and DHPch with expressed channels. Comparison between columns indicates differences in the effects of specific mutations on neutral and charged forms of the drug as well as, or in addition to differences caused by membrane potential (rows). The results clearly show that the presence of a charged head group changes the relative importance of residues in IVS6 and IIIS6 in regulating the activity of DHPch. Furthermore, changing the HP from 40 to 80 mV markedly increases the relative importance of Thr1061 (mutated in EC14) in determining DHPch channel inhibition.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Summary of the influence of mutations of 1C on inhibition by DHPn and DHPch. Shown in the figure are estimated IC50 values obtained from best fits to concentration-response data for inhibition of IBa by DHPn (left) and DHPch (right). As described in text, due to limitations in concentration responses, particularly to drug-insensitive channel constructs, these data only estimate the effective IC50 values. The data are presented for currents recorded from 40 (top) and 80 mV (bottom) HPs.

Distinct Functional Properties of DHPch

To provide a distinct profile of charged drug properties, we first compared drug-induced acceleration of apparent inactivation of current during depolarizing voltage pulses because such changes can reflect preferential interactions with either open and/or inactivated channels (Lee and Tsien, 1983). In the presence of DHPn, an extra block develops within 20 ms during pulse application for WT channels. This was also the case for all chimeric constructs in which channel block was decreased but still detectable, i.e., in EC19, EC20, and EC30 channels (see Figs. 2 and 3). It is possible, however, that pulses 20 ms in duration might have been too short to reveal the effect of DHPch on the time course of current decay. Thus, to test further the possible distinction in the mode of action of the two compounds, we also investigated current inhibition by both charged and neutral forms of our test DHPs measured during 100-ms-long depolarizing pulses (Fig. 6). For these experiments we chose DHP concentrations (both DHPch and DHPn) that inhibited 60 to 80% of the peak current. To illustrate the time course of additional inhibition that developed during depolarizing pulses, we formed the ratio of current traces recorded in the presence and absence of fixed drug concentrations (Fig. 6). The time course of additional inhibition in the presence of DHPn revealed by this analysis was well described as a single exponential process with average time constants of "extra block" of 19.2 ± 3.7 ms (n = 8). In contrast, DHPch did not affect the time course of currents measured during depolarizing pulses (Fig. 6B).

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Acceleration of the time course of IBa decay in WT channels induced by DHPn but not DHPch. Left: IBa was measured in control conditions and in the presence of 1µM DHPn (A) or 300 nM DHPch (B) during the application of 100-ms pulses to the peak of the current-voltage relationship (+20 mV). The record in the presence of drug was taken after new steady-state block had been achieved (after 20 pulses at a frequency of .1 Hz). The time courses (right) were obtained by dividing each current trace sample measured during the depolarizing pulse in the presence of the drug by the corresponding sample measured under drug-free conditions. Points preceding the peak of IBa were omitted. The solid line (A) represents the fit of experimental data to a single exponential with the time constant of 17.4 ms.

We next tested for functional properties of DHPch that distinguish this drug from DHPn and from verapamil, a phenylalkylamine that is also charged at physiological pH (Peterson et al., 1996) and blocks calcium channels in a use-dependent manner (Ehara and Kaufmann, 1978; Nawrath and Wegener, 1997). Figure 7 compares the effects of repetitive pulsing on the development of I_Ba block (use-dependent block) by verapamil and DHPch. In this experiment, control records were first recorded from a 80 mV HP to assay stability of currents in the absence of drug (). Then, with the HP maintained at 80 mV, test drugs were applied during a 2-min pulse-free period. Pulses were then resumed to evoke channel activity while drug application was maintained. In the presence of DHPch, current measured in response to the first pulse was markedly reduced, and there was little additional change in current amplitude in response to subsequent voltage pulses (Fig. 7A). In contrast, when a similar experiment was carried out with verapamil (Fig. 7B) there was little block of the first current measured after the pulse-free period, but a marked increase in channel block that occurred on a pulse-by-pulse basis after resumption of test pulse activity. DHPch exhibits marked tonic block with little use-dependent block. Verapamil is characterized by little tonic block, but pronounced use-dependent block. Clearly the actions of DHPch are distinct both from neutral DHPs and from charged phenylalkylamines.

View larger version (11K): [in this window] [in a new window]   Fig. 7.   Distinction in the voltage dependence of WT channel block by DHPch and verapamil. IBa was measured in cells transfected with WT (1C) cDNA. Current was measured in response to pulses (+10 mV, 40 ms) applied every 2 s from 80 mV in the absence (open symbols) of drug. Test drugs were then applied, and cells were held in the continued presence of drug for a 2-min pulse-free interval. Pulses were then resumed (+10 mV, 40 ms, .5 Hz) in the maintained presence of each drug. Shown are peak inward currents measured under these conditions for DHPch (300 nM; A) and verapamil (10 µM; B). An arrow indicates tonic block for each case.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study we used mutational analysis of the L-type calcium channel _1c subunit in combination with neutral and permanently charged forms of a custom synthesized DHP compound to test for distinction in sites of action between neutral and charged DHPs. Our results clearly show that residues in IVS6 and IIIS6 of the _1c subunit that are crucial for neutral DHP modulation of expressed channel activity only slightly influence block by DHPch. This can clearly be seen in the data summarized in Fig. 5 where mutations in IVS6 and IIIS6 increase the estimated IC₅₀ values for DHPn approximately 10- (80 mV) to 30-fold (40 mV), but only 2-fold for DHPch (40 and 80 mV).

The voltage dependence of L-type calcium channel modulation by neutral DHP compounds has been interpreted within the general framework of the modulated receptor hypothesis, which has been very successful in predicting differences in the activity of charged and neutral Class Ib local anesthetics. The central concept of this hypothesis is that there is a common channel-associated receptor for neutral and charged drug forms, but that hydrophobic (neutral) and hydrophilic (charged) drugs reach this single receptor via distinct hydrophilic and hydrophobic pathways. Furthermore, binding rates and equilibria for interactions with this common receptor depend on the state of the channel (Hille, 1977; Hondeghem and Katzung, 1977). The pronounced voltage dependence of the current inhibition by neutral DHPs is consistent with high-affinity binding of the drug to inactivated channels and low-affinity binding to channels in the resting state (Sanguinetti and Kass, 1984; Bean, 1984). In contrast, the weak voltage dependence of current inhibition by DHPch and lack of any effect on the kinetics of current decay during depolarizing pulses suggests that DHPch with different voltage-dependent states of the channel. In combination with the mutagenesis results, these data suggest that DHPn and DHPch interact differently with the amino acid residues that constitute the receptor binding domain for neutral drugs (Hockerman et al., 1997b). Thus, the addition of a charged head group may not only alter access of the DHP moiety to the receptor binding domain, but also change the interaction of the drug molecule with the amino acid residues on the channel protein that constitute the binding domain itself.

Perhaps most interesting in our study is the finding that the IIIS5 mutation Thr1061Tyr (EC14) has pronounced effects on both DHPn and DHPch test drugs. In the case of DHPn, mutation of this residue markedly reduces drug potency for both 40 and 80 mV HPs (see Fig. 5) similar to the effects of this mutation on the neutral drug DHP (+)isradipine (Mitterdorfer et al., 1996; Ito et al., 1997). However, in the case of DHPch, the mutation preferentially affects channel inhibition for currents evoked from a negative (80 mV) HP (Fig. 5). This property distinguishes Thr1061 from all other residues investigated in this study, which, when mutated, had only minor effects on DHPch block. But the relative importance of Thr1061 to DHPch inhibition of current depends critically on membrane potential. The simplest interpretation of this result is that the change in membrane potential from 80 mV to 40 mV alters the conformation of the expressed channel protein in a manner that reduces the relative importance of Thr1061 in determining inhibitory actions of DHPch. This could occur if Thr1061 contributed directly or via allosteric coupling to DHPch block of current recorded from a 80 mV but not from 40 mV HP. In this case, mutation of Thr1061 would be expected to alter DHPch inhibition of current evoked from 80 mV but not from 40 mV, consistent with our experimental data. It is likely then that, depolarization to 40 mV causes a conformational change of the channel protein that restricts access to these residues through a physical restriction imposed upon the DHP moiety of the DHPch by the presence of the charged head group. This is likely to be the case because the neutral drug, DHPn, interacts with these residues at both HPs. Because L-type calcium channels do not open at 40 mV, this conformational change is likely to accompany either a closed state transition or a transition of the channel into the inactivated state. Gating charge movement has been reported to occur in the absence of channel openings over the same voltage range (Bangalore et al., 1996). It is attractive to speculate that the same changes in channel conformation underlie this charge movement and voltage-dependent change in DHPch block of EC14 channels.

Previous studies have shown that DHPch and other similar molecules with varying hydrocarbon spacer chains separating the charged head group from the DHP moiety inhibit channel activity of native L-type channels as well as binding of [³H]-(+)-PN200-100 (Baindur et al., 1993; Bangalore et al., 1994) in a spacer chain-dependent manner. Our present results are consistent with these previous observations and suggest that the voltage-dependent limitation on the interaction of DHPch with Thr1061 may result from structural restrictions imposed on the DHPch molecule by the charged head group. As suggested previously (Bangalore et al., 1994), this may be caused simply by exclusion of the hydrophilic charged head group from the membrane lipid bilayer.

On the other hand, it may also be possible that the positively charged head group of DHPch interacts in an electrostatic manner with specific negative charges on the _1c subunit and that this interaction contributes to the physical restriction of the charged drugs. Each pore region (P-region) of the L-type calcium channel _1c subunit contains a negatively charged glutamate residue that contributes both to calcium channel selectivity (Sather et al., 1994) and to allosteric modulation of DHP binding (Mitterdorfer et al., 1995; Peterson and Catterall, 1995). These residues may interact with DHPch and thereby restrict its access to residues key to the DHP binding domain in the channel pore. Mutation of P-region glutamates has recently been shown to modify block of expressed currents by verapamil, a charged phenylalkylamine derivative (Hockerman et al., 1997a). Thus, it will be of interest to test directly for a role of the P-region glutamates in the action of charged DHPs, such as DHPch, particularly because the modulatory activity of DHPch is distinct from that of verapamil (see Fig. 7). It is entirely possible that other, as yet unidentified, residues may affect DHPch interaction in an allosteric manner. Similar allosteric interactions have been reported to be the case for residues in IS6 and IIIS2 of human _1c that regulate channel sensitivity channel to (+)-isradipine (Soldatov et al., 1995; Welling et al., 1997; Zuhlke et al., 1998).