Serine Residue in the IIIS5-S6 Linker of the L-type Ca2+ Channel α1C Subunit Is the Critical Determinant of the Action of Dihydropyridine Ca2+Channel Agonists*

The dihydropyridine (DHP)-binding site has been identified within L-type Ca2+ channel α1Csubunit. However, the molecular mechanism underlying modulation of Ca2+ channel gating by DHPs has not been clarified. To search for novel determinants of high affinity DHP binding, we introduced point mutations in the rat brain Ca2+ channel α1C subunit (rbCII or Cav1.2c) based on the comparison of amino acid sequences between rbCII and the ascidian L-type Ca2+ channel α1 subunit, which is insensitive to DHPs. The α1C mutants (S1115A, S1146A, and A1420S) and rbCII were transiently expressed in BHK6 cells with β1a and α2/δ subunits. The mutation did not affect the electrophysiological properties of the Ca2+channel, or the voltage- and concentration-dependent block of Ca2+ channel currents produced by diltiazem and verapamil. However, the S1115A channel was significantly less sensitive to DHP antagonists. Interestingly, in the S1115A channel, DHP agonists failed to enhance whole-cell Ca2+ channel currents and the prolongation of mean open time, as well as the increment of NP o. Responsiveness to the non-DHP agonist FPL-64176 was also markedly reduced in the S1115A channel. When S1115 was replaced by other amino acids (S1115D, S1115T, or S1115V), only S1115T was slightly sensitive to S-(−)-Bay K 8644. These results indicate that the hydroxyl group of Ser1115 in IIIS5-S6 linker of the L-type Ca2+ channel α1C subunit plays a critical role in DHP binding and in the action of DHP Ca2+ channel agonists.

type Ca 2ϩ channel currents showed three distinct modes of Ca 2ϩ channel gating: brief openings (mode 1), no openings due to channel unavailability (mode 0), and long-lasting openings and very brief closing (mode 2) (6). Bay K 8644 1 enhances Ca 2ϩ channel current by promoting mode 2, whereas nitrendipine inhibits the current by favoring mode 0 (6). However, the conformational change accompanied with channel gating and its modulation by Ca 2ϩ channel agonists and antagonists is largely unknown. Determination of amino acids that are critical for the specific interaction with Ca 2ϩ channel agonists and antagonists will help clarify the binding pockets and the molecular mechanism underlying modulation of Ca 2ϩ channel gating. DHP-binding sites have been determined by photoaffinity labeling, radioligand binding, chimera, and alaninescanning mutagenesis based on sequence differences between ␣ 1C (or ␣ 1S ) and the DHP-insensitive ␣ 1A subunit (7)(8)(9)(10)(11). However, other amino acids conserved in both DHP-sensitive and DHP-insensitive ␣ 1 subunits may also participate in the binding pocket for DHPs. Indeed, alanine-scanning mutagenesis of IIIS6 and IVS6 showed that some amino acids conserved in both ␣ 1C and ␣ 1A are required for the high affinity binding of Ca 2ϩ channel antagonists (12).
Recently, ascidian Ca 2ϩ channel (TuCa1) was cloned (13). TuCa1 belongs to the L-type Ca 2ϩ channel and has all amino acids that have been determined as critical sites for DHP binding so far. However, Ba 2ϩ currents through TuCa1 were poorly sensitive to the DHP antagonist, nitrendipine, and insensitive to the DHP agonist, S-(Ϫ)-Bay K 8644, when expressed in Xenopus oocytes. The jellyfish Ca 2ϩ channel (CyCa␣1) was also insensitive to DHPs (14). These reports imply that amino acid residues conserved in DHP-sensitive and DHP-insensitive Ca 2ϩ channel ␣ 1 subunits are also required for the high affinity binding pocket of DHPs.
Thus, in the present study, we searched for amino acids, which are conserved in ␣ 1A , ␣ 1B , ␣ 1C , ␣ 1D , ␣ 1E , and ␣ 1S but not in TuCa1 or CyCa␣1, to identify novel determinants of DHP binding (see Fig. 1). We identified Ser 1115 , located in the poreforming region between IIIS5 and IIIS6 of the L-type Ca 2ϩ channel ␣ 1C subunit, as a critical determinant of DHP binding and of the action of DHP agonists.

EXPERIMENTAL PROCEDURES
Point Mutations in rbCII-Single mutations were generated by introducing TuCa1-type amino acids into rat brain ␣ 1C subunit (rbCII, Ca v 1.2c), which was kindly supplied by Dr. T. P. Snutch (15). Single point mutations, S1115A, and S1146A, were introduced into IIIS5-S6 * This work was supported by a grant-in-aid from the Japanese Ministry of Education, Science, Sports, and Culture (11771426) and a grant-in-aid for research on Health Sciences Focusing on Drug Innovation from the Japan Health Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Whole-cell Patch-clamp Recording-The whole-cell L-type Ca 2ϩ channel currents were recorded with Ca 2ϩ (2 mM) as a charge carrier in bath solution containing (in mM): 137 NaCl, 5.4 KCl, 1 MgCl 2 , 10 glucose, 10 HEPES, 2 CaCl 2 (pH 7.4, adjusted with NaOH) at room temperature. The resistance of the heat-polished microelectrodes ranged between 2 and 4 M⍀ when filled with the internal solution composed of (in mM): 120 CsCl, 20 tetraethyl ammonium chloride, 14 EGTA, 5 MgATP, 5 disodium creatine phosphate, 0.2 GTP, 10 HEPES, 0.2 cAMP (pH 7.3, adjusted with CsOH). Whole-cell currents were measured using a patch/whole-cell clamp amplifier (Nihon Kohden, Tokyo, Japan) or Axopatch 1D (Axon Instruments, Inc., Foster City, CA) via an analog to digital converter (Digidata 1200, Axon Instruments, Inc.). Voltage-clamp protocols and data acquisitions were performed using pCLAMP6 software (Axon Instruments, Inc.). Transfected cells were identified by the expression of GFP. The percentage of cells expressing GFP was about 40 -60%, and Ca 2ϩ channels were expressed in about 60 -80% of the GFP-expressing cells. Because the inactivation kinetics of Ca 2ϩ channels are influenced by the current density, we used cells expressing Ca 2ϩ channel current density between 10 and 100 pA/pF.
Single-channel Recording-Cell-attached single-channel recordings were performed with a high K ϩ bath solution to cancel membrane potential (in mM): 5 KCl, 112 potassium aspartate, 5 NaCl, 3 MgCl 2 , 1 Mg-ATP, 2 EGTA, 10 glucose, 10 HEPES (pH 7.3, adjusted with KOH) at room temperature. The resistance of the heat-polished microelectrodes was between 2 and 9 M⍀ when filled with the internal solution composed of (in mM): 110 BaCl 2 , 10 HEPES (pH 7.4, adjusted with tetraethyl ammonium hydroxide). Single-channel currents were measured using EPC9 (HEKA , Germany). Voltage-clamp protocols and data acquisitions were performed by use of PULSE software (HEKA, Germany). Analysis of data was performed using TAC version 3.0 (Bruxton Corp., Seattle, WA), Igor Pro software, and Patch Analyst Pro software (MT Corp.).
Materials-Diltiazem (generous gift from Tanabe Seiyaku) and verapamil (purchased from Nacalai Tesque, Kyoto, Japan) were dissolved in distilled water and stored at 4°C as 1 mM stock solutions. Nitrendipine (purchased from Funakoshi Seiyaku, Tokyo, Japan), S-(Ϫ)-Bay K 8644 (purchased from Sigma Chemical Co. St. Louis, MO), R-(ϩ)-Bay K 8644 (purchased from Sigma), Bay Y 5959 (generous gift from Bayer AG, Wuppertal, Germany), (ϩ)-isradipine (generous gift from Sandoz AG, Basel, Switzerland), and FPL-64176 (purchased from Sigma) were dissolved in EtOH and stored at Ϫ20°C as 3 mM stock solutions. Drugs were dissolved in the external solution and applied by either perfusion (at 4 ml/min) or via concentration-clamp apparatus (Vibraspec, Inc., Philadelphia, PA) in the whole-cell patch-clamp recording. The concentration-clamp apparatus allowed us to exchange the extracellular solution within 50 ms (17). In single-channel recording experiments, each drug was directly added to the bath solution to make up the final concentration.
Statistics-Data are expressed as means Ϯ S.E. Statistical significance was assessed with the Student-Welch's t test for simple comparisons. Differences at p Ͻ 0.05 were considered to be significant.
Sensitivity of Mutant Ca 2ϩ Channels to Various Ca 2ϩ Channel Antagonists-First, we examined the effect of diltiazem. All mutants and rbCII were equivalently blocked by diltiazem (1 M) at holding potentials of Ϫ70 and Ϫ50 mV without any differences in the voltage dependence of the block (Fig. 3, A and  B). We also compared the block of Ca 2ϩ channel currents by verapamil (1 M, 10 M). Ca 2ϩ channel currents of S1115A were blocked to the same degree as that of rbCII (Fig. 3, C and  D). On the other hand, sensitivity of S1115A to DHP antagonists, such as nitrendipine, at a holding potential of Ϫ70 mV was significantly decreased compared with those of rbCII, S1146A, and A1420S (Fig. 4B). When holding potential was changed to Ϫ50 from Ϫ70 mV, nitrendipine block of S1115A FIG. 2. Electrophysiological properties of mutant Ca 2؉ channels were not different from those of rbCII Ca 2؉ channels. A, current-voltage relationships of S1115A (f, n ϭ 8), S1146A (ࡗ, n ϭ 14), A1420S (OE, n ϭ 11), and rbCII (E, n ϭ 14) are superimposed. Currents are normalized by peak current amplitude. There were no significant differences in peak Ca 2ϩ channel current density between rbCII and mutant Ca 2ϩ channels. B, steady-state inactivation curves of Ca 2ϩ channel currents. Steady-state inactivation curves (B-c) of S1115A (f, n ϭ 6), S1146A (ࡗ, n ϭ 12), A1420S (OE, n ϭ 8), and rbCII (E, n ϭ 7) were measured with voltage protocol shown in B-a. Representative current traces are shown in B-b (rbCII). Pulses were applied every 30 s. C, restitution curve of Ca 2ϩ channel currents. Recovery from the inactivation of S1115A (f, n ϭ 6), S1146A (ࡗ, n ϭ 5), A1420S (OE, n ϭ 5), and rbCII (E, n ϭ 7) were measured with voltage protocol shown in C-a. mV. Peak amplitudes of Ca 2ϩ channel currents were normalized to those of control measured at the respective holding potentials. There were no significant differences between rbCII and mutant Ca 2ϩ channels. C, representative current traces of rbCII (a) and S1115A (b) elicited by a 100-ms test pulse to 0 mV from a holding potential of Ϫ70 mV at 0.2 Hz. Open circles represent control Ca 2ϩ channel currents. Gray and black circles represent Ca 2ϩ channel currents blocked by verapamil at 1 and 10 M, respectively. D, concentration-dependent block by verapamil of Ca 2ϩ channel currents through rbCII and S1115A. The error bars show S.E. channel was enhanced but still significantly weaker than that of rbCII. The concentration-response curve for the nitrendipine block of Ca 2ϩ channel currents was shifted to the right by 57.5-fold in S1115A (IC 50 values: 7.8 ϫ 10 Ϫ7 M (rbCII) versus 4.5 ϫ 10 Ϫ5 M (S1115A), Fig. 4C). The affinity for (ϩ)-isradipine was also reduced by 24.1-fold (IC 50 values at Ϫ60 mV: 3.7 ϫ 10 Ϫ7 M (rbCII) versus 8.9 ϫ 10 Ϫ6 M (S1115A), (n ϭ 3), data not shown). In S1115A, R-(ϩ)-Bay K 8644, a stereoisomer of Ca 2ϩ channel agonist S-(Ϫ)-Bay K 8644 (18,19), again, produced significantly smaller block of Ca 2ϩ channel currents than in rbCII (Fig. 4, D and E). Thus, S1115A was significantly less sensitive to DHP antagonists.
Sensitivity of Mutant Ca 2ϩ Channels to Ca 2ϩ Channel Agonists-Ca 2ϩ channel currents of rbCII, S1146A, and A1420S were equally increased by S-(Ϫ)-Bay K 8644 (1 M). However, S1115A was not enhanced at all. This result was against our expectation that S-(Ϫ)-Bay K 8644 would produce smaller enhancement of Ca 2ϩ channel currents in S1115A than in rbCII as was the case with the Ca 2ϩ channel block produced by R-(ϩ)-Bay K 8644 (Fig. 5, A and B). Ca 2ϩ channel currents of S1115A were rather slightly decreased by S-(Ϫ)-Bay K 8644 (1 M) (98.6 Ϯ 3.45% of control) to the same degree as time control (92.9 Ϯ 2.15% of control at 90 s, data not shown). Another DHP Ca 2ϩ channel agonist, Bay Y 5959 (20), also did not enhance the Ca 2ϩ channel current of S1115A, but rather slightly decreased it in a concentration-dependent manner (Fig. 5, C and  D). These results demonstrate that S1115A is insensitive to DHP Ca 2ϩ channel agonists. Thus we verified whether S1115A lacks the sensitivity to only DHP Ca 2ϩ channel agonists or Ca 2ϩ channel agonists generally. A benzoylpyrrole-type Ca 2ϩ channel agonist, FPL-64176 binds to the Ca 2ϩ channel at its binding site distinct from that of DHPs (21,22). Ca 2ϩ channel currents of S1115A were slightly enhanced by FPL-64176 (1 M), unlike the results with S-(Ϫ)-Bay K 8644 or Bay Y 5959. However, the enhancement of Ca 2ϩ channel currents through S1115A were markedly smaller than that of rbCII (Fig. 5, E  and F). Because Ser 1115 is located only three amino acids away from Glu 1118 , which is the determinant of Ca 2ϩ selectivity of the ␣ 1C subunit, the sensitivity of S1115A for DHPs may be affected by charge carriers. Therefore, we tested whether the above results were affected when Ba 2ϩ was used as a charge carrier in place of Ca 2ϩ , and we obtained the same results as observed with Ca 2ϩ , such as insensitivity to S-(Ϫ)-Bay K 8644 (data not shown).
Importance of the Hydroxyl Group of Ser 1115 -We replaced Ser 1115 by other amino acid residues (Asp, Thr, and Val) to identify the functional group responsible for the action of Ca 2ϩ channel agonists. Replacement by valine (S1115V) or aspartic acid (S1115D) abolished the effect of S-(Ϫ)-Bay K 8644 (1 M). In contrast, in S1115T, S-(Ϫ)-Bay K 8644 produced slight but significant enhancement of Ca 2ϩ channel currents (Fig. 6). The Ca 2ϩ channel block produced by nitrendipine was also reduced in S1115T. The IC 50 value of S1115T was 1.9 ϫ 10 Ϫ5 M, which was higher than that of rbCII by a factor of 25 but lower than that of S1115A by 2.3-fold (Fig. 4C). These results indicate that the hydroxyl group of Ser 1115 forms part of the DHP binding pocket and plays a critical role in mediating the action of DHP Ca 2ϩ channel agonists.  (Fig. 7A)). Open-time histograms of rbCII and S1115A with and without S-(Ϫ)-Bay K 8644 are shown in Fig. 7B. The mean open time of rbCII was also increased by S-(Ϫ)-Bay K 8644 in rbCII but unchanged in S1115A (mean open time in ms: 1.60 Ϯ 1.93 3 3.26 Ϯ 3.23, 2.07 Ϯ 3.46 3 2.24 Ϯ 3.36, respectively). These results indicate that prolongation of Ca 2ϩ channel opening by Ca 2ϩ channel agonists is mostly absent in S1115A.

DISCUSSION
Electrophysiological Properties of Mutant Ca 2ϩ Channels-DHP Ca 2ϩ channel antagonists bind to Ca 2ϩ channels in the inactivated state with high affinity (23). It is possible that the affinity of Ca 2ϩ channel antagonists for Ca 2ϩ channels is influenced by their inactivation kinetics. However, we did not find any differences in the electrophysiological properties between rbCII and mutant ␣ 1C subunits, indicating that the mutation introduced into the ␣ 1C subunit did not affect the inactivation kinetics (Fig. 2). Therefore, any differences in the responsiveness to DHPs between rbCII and mutant Ca 2ϩ channels should result from the conformational change of the DHP binding pocket or the gating moiety linked to the DHP-binding site.

FIG. 4. Voltage-dependent block of Ca 2؉ channel currents by nitrendipine (0.1 M) and R-(؉)-Bay K 8644 (1 M).
A, representative current traces of rbCII (a), S1115A (b), S1146A (c), and A1420S (d) elicited by a 100-ms test pulse to 0 mV at 0.1 Hz. Open circles represent control Ca 2ϩ channel currents at a holding potential of Ϫ70 mV. Gray and black circles represent Ca 2ϩ channel currents blocked by nitrendipine (0.1 M) at holding potentials of Ϫ70 mV and Ϫ50 mV, respectively. B, inhibition of Ca 2ϩ channel currents by nitrendipine at holding potentials of Ϫ70 mV and Ϫ50 mV. *p Ͻ 0.05 versus S1146A, A1420S, and rbCII. C, concentration-response curve for nitrendipine of rbCII and S1115A. The IC 50 value for S1115A and S1115T are 57.5 and 25 times higher than that of rbCII, respectively. D, representative current traces of rbCII (a) and S1115A (b). Open and black circles represent control Ca 2ϩ channel currents and their block produced by R-(ϩ)-Bay K 8644 (1 M), respectively. E, inhibition by R-(ϩ)-Bay K 8644 of Ca 2ϩ channel current through rbCII and S1115A. The error bars show S.E. *p Ͻ 0.05 versus rbCII.
Pharmacological Properties of S1115A-IC 50 values of the Ca 2ϩ channel block produced by nitrendipine was higher in S1115A by 57.5-fold than in rbCII. Another DHP Ca 2ϩ channel antagonist, R-(ϩ)-Bay K 8644, also produced significantly smaller block in S1115A than in rbCII (Fig. 4). The mutation in IIIS6 (Y1152A) has been shown to reduce the affinity of (ϩ)isradipine by 25-fold (12). In the present study, the affinity of (ϩ)-isradipine for S1115A was also lower than that of rbCII by 24-fold. These results indicate that the replacement of Ser 1115 by Ala reduced the affinity for DHP antagonists. In contrast, the Ca 2ϩ channel currents of S1115A were inhibited by dilti-azem and verapamil in the same way as that of rbCII (Fig. 3). Therefore, Ser 1115 in IIIS5-S6 linker appears to play a critical role in the binding pocket specific to DHPs.
Unexpectedly, Ca 2ϩ channel currents of S1115A were insensitive to DHP Ca 2ϩ channel agonists and rather slightly decreased by S-(Ϫ)-Bay K 8644 (Fig. 5, A and B) in a way similar to that of time control (data not shown). R-(ϩ)-Bay K 8644, a DHP Ca 2ϩ channel antagonist, and S-(Ϫ)-Bay K 8644, a DHP Ca 2ϩ channel agonist, are enantiomers. However, effects of both chemical compounds were not linearly altered in S1115A: antagonistic effects of R-(ϩ)-Bay K 8644 were reduced, but agonistic effects of S-(Ϫ)-Bay K 8644 were completely eliminated. Even higher concentration (3 M) of S-(Ϫ)-Bay K 8644, a concentration more than 300 times higher than EC 50 value of S-(Ϫ)-Bay K 8644 (ϳ10 nM), did not enhance Ca 2ϩ channel currents of S1115A (data not shown). Thus the impairment of the action of both stereoisomers of DHPs were not simply explained by a decrease of affinity. Therefore, Ser 1115 appears to serve as an important site for agonistic action of DHP Ca 2ϩ channel agonists. Indeed, Ca 2ϩ channel currents of S1115A were also rather slightly inhibited by another DHP Ca 2ϩ channel agonist, Bay Y 5959, regardless of concentration (Fig. 5, C  and D). In contrast, the benzoylpyrrole (non-DHP) Ca 2ϩ channel agonist, FPL-64176, slightly enhanced Ca 2ϩ channel currents of S1115A, but the effect was significantly smaller than that of rbCII (Fig. 5, E and F). Considering that the bindings of S-(Ϫ)-Bay K 8644 and FPL-64176 to the Ca 2ϩ channel do not compete with each other, i.e. that the binding sites for both compounds are distinct, the two compounds may share the  Y 5959 (0.1, 1 M), and FPL-64176 (1, 10 M) on Ca 2؉ channel currents. A, representative current traces of rbCII (a), S1115A (b), S1146A (c), and A1420S (d) elicited by a 100-ms test pulse to 0 mV from a holding potential of Ϫ70 mV at 0.1 Hz. Open and black circles represent Ca 2ϩ channel currents measured before and after application of S-(Ϫ)-Bay K 8644 (1 M), respectively. B, relative increase of peak Ca 2ϩ channel currents by S-(Ϫ)-Bay K 8644 is summarized. *p Ͻ 0.05 versus S1146A, A1420S, or rbCII. C, representative current traces of rbCII (a) and S1115A (b). Open circles represent control Ca 2ϩ channel current. Gray and black circles represent Ca 2ϩ channel currents enhanced by Bay Y 5959 at 0.1 and 1 M, respectively. D, relative increase of peak Ca 2ϩ channel currents by Bay Y 5959. *p Ͻ 0.05 versus rbCII. E, representative current traces of rbCII (a) and S1115A (b). Open circles represent control Ca 2ϩ channel current. Gray and black circles represent Ca 2ϩ channel current enhanced by FPL-64176 at 1 and 10 M, respectively. F, relative increase of peak common site, such as Ser 1115 , for producing Ca 2ϩ channel agonist effects. These results indicate that Ser 1115 plays a critical role in the action of Ca 2ϩ channel agonists.
Binding of Ca 2ϩ to the Ca 2ϩ channel pore may contribute to stabilize high affinity DHP binding (5,24). However, the direct interaction between the Ca 2ϩ -binding site of the Ca 2ϩ channel pore with DHPs has not been shown. The critical role of Ser 1115 in DHP binding strongly supports this idea, because Ser 1115 is located at three amino acids away from the key Glu 1118 of the channel pore (25) (Fig. 8A). In addition, our findings are consistent with the previous studies: 1) DHP agonists and antagonists gain access to their binding sites from the extracellular side of the channel, 2) the DHP-binding site resides 11-14 Å from the extracellular surface of the cell membrane (26 -29), 3) photoreactive DHPs specifically label the connecting loop between IIIS5 and IIIS6 (30 -33), 4) in analysis of chimeric Ca 2ϩ channels, IIIS5-S6 linker is necessary for transferring DHP sensitivity from L-type Ca 2ϩ channel to P/Q-type (␣ 1A ) Ca 2ϩ channels (7). The previous studies have identified 14 amino acids, indicated as gray circles in Fig. 8A, as critical sites for DHP binding, which were basically determined by comparison between the sequences of the DHP-insensitive ␣ 1A subunit and the DHP-sensitive ␣ 1C (␣ 1S ) subunit. However, the alaninescanning mutagenesis performed in the whole area of the transmembrane region IIIS6 and IVS6 showed that the DHP binding pocket includes amino acids conserved in both ␣ 1A and ␣ 1C subunits. For this reason, the amino acids highly conserved in both ␣ 1A and ␣ 1C subunits may also serve as critical sites for DHP binding. Our finding of Ser 1115 in IIIS5-S6 linker as a novel determinant of DHP binding is the first step toward understanding the molecular mechanism underlying the conformational change of the pore region involved in the modulation of Ca 2ϩ channel gating by DHPs. Involvement of other domains such as motifs IS5-S6 and IIS5-S6 are under investigation.
The single-channel study clearly showed that, in S1115A, Ca 2ϩ channel agonists (or DHP agonist) failed to prolong Ca 2ϩ channel opening. However, unexpectedly, the open probability was slightly enhanced by S-(Ϫ)-Bay K 8644 (Fig. 7A), although the whole-cell Ca 2ϩ channel current of S1115A was not increased at all (Fig. 5A). These somewhat contradictory results may be explained as follows: 1) in whole-cell recording, natural run-down of the Ca 2ϩ channel current may have masked the slight enhancement of S1115A by S-(Ϫ)-Bay K 8644 and resulted in the slight decrease of the current, because S-(Ϫ)-Bay K 8644 reduced the Ca 2ϩ channel current to a degree similar to that of time control; 2) the difference of charge carrier (Ca 2ϩ and Ba 2ϩ ) may influence the effects of S-(Ϫ)-Bay K 8644, because Ser 1115 is located near a Ca 2ϩ -selective filter (Glu 1118 ). However, the latter possibility is unlikely, because effects of S-(Ϫ)-Bay K 8644 on whole-cell Ca 2ϩ channel currents were the same in rbCII and S1115A whether the charge carrier was Ca 2ϩ (2 mM) or Ba 2ϩ (2 mM) (data not shown).
Importance of the Hydroxyl Group of Ser 1115 -We replaced The upper lines represent voltage protocol. Holding potential was Ϫ70 mV, and test potential was Ϫ20 to 0 mV. Unitary currents through rbCII were strongly enhanced by Ca 2ϩ channel agonists but not in S1115A. Therefore, different voltages were used as test pulses in rbCII (Ϫ20 mV) and S1115A (Ϫ10, 0 mV) to optimize the recording condition. In each experiment, the same test pulse was used before and after drug application. B, mean open time distributions for rbCII and S1115A with and without S-(Ϫ)-Bay K 8644. Mean open time of rb-CII was increased (1.60 Ϯ 1.93 ms (n ϭ 75) 3 3.26 Ϯ 3.23 ms (n ϭ 623)), whereas it was unchanged in S1115A (2.07 Ϯ 3.46 ms (n ϭ 64) 3 2.24 Ϯ 3.36 ms (n ϭ 210)).
Ser 1115 by other amino acid residues (Asp, Thr, Val) to identify the important functional group of this serine. Only S1115T was slightly increased by S-(Ϫ)-Bay K 8644. The binding affinity of S1115T for DHP antagonist was significantly lower than that of rbCII but higher than that of S1115A. These results indicate that the hydroxyl group at Ser 1115 is important for DHP binding and the action of Ca 2ϩ channel agonists. The steric hindrance by the methyl group at ␤-carbon of threonine may be the reason why effects of DHPs were significantly reduced in S1115T.
The conformational dynamics of IIIS5-S6 linker and its role in the gating mechanism have not been clarified yet. When Gln 1043 in IIIS5 (10) and Met 1161 in IIIS6 (35) were replaced, agonistic effects of DHPs were selectively abolished, while antagonistic effects remained. Therefore, these two amino acids may be responsible for the conformational change of IIIS5-S6 caused by DHP Ca 2ϩ agonists. Based on our finding that Ser 1115 contributes to DHP binding and plays a critical role in producing the action of Ca 2ϩ channel agonists, Ser 1115 may be involved, in cooperation with Gln 1043 and Met 1161 , in the conformational change of IIIS5-S6 linker upon binding of DHP Ca 2ϩ channel agonists. Considering that electrophysiological properties of S1115A were identical to those of rbCII, Ser 1115 appears to be a critical amino acid that links the binding of DHP agonists and the modulation of Ca 2ϩ channel gating, which is not a result of modulation of Ca 2ϩ channel activation (36,37) nor inactivation (34,38) but rather may be the result of direct stabilization of the Ca 2ϩ channel pore in the open state. Our finding further implies that the conformational change of IIIS5-S6 linker may be a critical step of Ca 2ϩ channel gating and its modulation by Ca 2ϩ channel antagonists or agonists. Therefore, we propose a model that DHPs interact with the hydroxyl group of Ser 1115 , and, especially upon the binding of Ca 2ϩ channel agonists, such interaction stabilizes the Ca 2ϩ channel in the open state (Fig. 8B).
Summary-We identified a novel critical site (Ser 1115 in the IIIS5-S6 loop of the L-type Ca 2ϩ channel ␣ 1C subunit) for DHP binding. Ser 1115 turned out to be a critical determinant of action of DHP agonists. FIG. 8. Schematic drawing of DHP Ca 2؉ channel agonists and antagonists interaction domains. A, part of the putative pore region, IIIS5-IIIS6 and IVS5-IVS6, of Ca 2ϩ channel ␣ 1 subunit is shown. Amino acid residues shown in gray circles have been shown to participate in the interaction with dihydropyridines. Glutamic acids in IIIS5-IIIS5 linker and IVS5-S6 linker are part of the selective filter for permeation of Ca 2ϩ . Ser 1115 , a novel determinant of action of DHP agonist, is located in the linker between IIIS5 and IIIS6 (indicated by an arrow). B, hypothetical models for role of Ser 1115 in the interaction between DHPs and Ca 2ϩ channel pore. S indicates Ser 1115 . DHP agonists may interact with ϪOH of Ser 1115 and stabilize Ca 2ϩ channels in the open state.