positive heterotropic allosteric regulators of dihydropyridine

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 17, Issue of June 15, PP. 10787-10795,1991 Printed in U. S. A.

Positive Heterotropic Allosteric Regulators of Dihydropyridine Binding Increase the Ca2+ Affinity of the L-type Ca2+ Channel STEREOSELECTIVE REVERSAL BY T H E NOVEL Ca2+ ANTAGONIST BM 20.1140*

(Received for publication, December 24, 1990)

Robert Staudinger, Hans-Gunther Knaus, and Hartmut GlossmannS From the Institute for Biochemical Pharmacology, University of Innsbruck, Peter-Mayr-Strasse I, A-6020 Innsbruck, Austria

The al-subunit of the voltage-dependent L-type Ca2+ channel has distinct, allosterically coupled binding domains for drugs from different chemical classes (dihydropyridines, benzothiazepines, phenylalkylamines, diphenylbutylpiperidines). (-)-BM 20.1140 (ethyl- 2,2-di-phenyl-4-(1-pyrrolidino)-5-(2-picolyl)-oxyval- erate) is a novel Ca2+ channel blocker which potently stimulates dihydropyridine binding ( K O . 5 = 2.98 nM) to brain membranes. This property is shared by (+)-cis- diltiazem, (+)-tetrandrine, fostedil and trans-diclofurime, but (-)-BM 20.1140 does not bind in a competitive manner to the sites labeled by (+)-~is-[~H]diltiazern. (+)-cis-Diltiazem and (-)-BM 20.1 140 have differential effects on the rate constants of dihydropyridine binding. (+)-BM 20.1140 reverses the stimulation of the positive allosteric regulators (pA2 value for reversal of (-)-BM 20.1140 stimulation = 7.4, slope 0.72). The underlying molecular mechanism of the potentia- tion of dihydropyridine binding has been clarified. The

for free Ca2+ to stabilize a high affinity binding domain for dihydropyridines on purified L-type channels from rabbit skeletal muscle is 300 nM. (+)-Tetran- dine (10 WM) increases the affinity %fold for free Ca2+ = 30.1 nM) and (+)-BM 20.114 (10 WM) inhibits the affinity increase ( K O . 5 for free Ca2+ = 251 nM). Similar results were obtained with membrane-bound Ca2+-channels from brain tissue which have higher affinity for free ca2+ ( K o . 5 for free ca2+ = 132 nM) and for dihydropyridines compared with skeletal muscle. I t is postulated that the dihydropyridine and Ca2+- binding sites are interdependent on the al-subunit, that the different positive heterotropic allosteric regulators (by their differential effects on Ca2+ rate constants) optimize coordination for Ca2+ in the channel pore and, in turn, increase affinity for the dihydropyridines.

Voltage-dependent L-type Ca2+-channels have several distinct allosterically communicating receptor domains for drugs (1-4) which are located on the al-subunit as shown by photolabeling (5-8). Rapid Ca2+ flux through and ion selectivity of the al-subunit are best explained by a multiple (at least two) ion-binding site model where ion-ion repulsion is the key feature to overcome the strong cation binding within the pore (9, 10). Whereas electrophysiological studies predict two func-

* This work was supported by Fonds zur Forderung der wissen- schaftlichen Forschung S-4501. 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.

$ T o whom correspondence and reprint requests should be ad- dressed.

tionally (but not necessarily structurally) symmetrical divalent cation-binding sites (9), most likely formed by noncon- tiguous sequences of the al-subunit, one internal putative EF hand in all known al-subunit sequences is predicted by the Tufty-Kretsinger test (11). This putative Ca2+-binding domain is within an evolutionary conserved area of a1 (12) and close to the phenylalkylamine receptor site as recently shown by photolabeling and site-directed antibodies (13). Whereas Ca2+ (and other divalent cations) inhibit the binding of radiolabeled phenylalkylamines to membrane-bound L-type Ca2+ channels with values in the high micromolar to millimolar range (see Refs. 3 and 4 for reviews), high affinity dihydropyridine labeling of L-type channels in brain, cardiac, or smooth muscle membranes is absolutely dependent on divalent cations (see 3, 4, 14-16). In skeletal muscle microsomes this dependence only becomes apparent in the presence of detergents (17) or after treatment with A 23187 (18). Evidence from electrophysiological and binding studies suggests that the dihydropyridine receptors and their coupled Ca2+-binding domains are on the extracellular face of the channel al- subunit (18-20). Dihydropyridine receptors, in contrast to the phenylalkylamine sites, also display the characteristic of tissue- (but not species) specific dissociation constants, typically %”-fold lower in brain, compared to skeletal muscle (3, 4).

Several drugs, chemically unrelated, stimulate equilibrium binding of radiolabeled dihydropyridines, often in a tissue- specific and temperature-dependent manner (21-27). Kinetic studies and saturation analysis have revealed that decreases of KO, of the dissociation rate ( k l ) and/or increased B,,, values are consequences of the binding of these allosteric regulators (23, 28, 29). (+)-cis-Diltiazem, a benzothiazepine, was the first representative of these positive heterotropic allosteric regulators, and its site was later shown to be recip- rocally allosterically coupled to the dihydropyridine receptor (3, 30-32). The idea that occupation of the benzothiazepine- selective site was necessary to exert the positive heterotropic allosteric action on dihydropyridine labeling was supported by the finding that the potent stimulator (+)-tetrandrine is a competitive ligand for (+)-cis- [3H]diltiazem-labeled sites in cardiac sarcolemma (26). On the other hand, truns-diclofurime (241, fostedil (22), and MDL 12330 (27), which also stimulate, may not bind directly to the benzothiazepine site as shown by kinetic analysis (3). Furthermore, dihydropyridine labeling of purified Ca2+ channels is (stereoselectively) enhanced by phenylalkylamines (33) which bind at the intracellular face of the channel (13) whereas (+)-cis-diltiazem acts from the opposite side (19).

Taken together these findings strongly argue against the hypothesis that binding to the benzothiazepine site is a pre- requisite to up-regulate dihydropyridine labeling.

Our interest in the molecular basis of positive heterotropic

10787

10788 Ca2+-binding Sites of the Ca'+ Channel

C H 2 I

FIG. 1. Structural formula of BM 20.1140. The assymmetric carbon atom is indicated by the asterisk.

allosteric mechanisms was raised when we found that the (-)- enantiomer of the novel Ca2+ channel blocker BM 20.1140 was a potent stimulator of dihydropyridine binding whereas its optical antipode was able to reverse the positive heterotropic allosteric effect to a major extent. We will report below that (-)-BM 20.1140 as well as other compounds, shown previously to be positive allosteric regulators, share a common mechanism at the molecular level.

EXPERIMENTAL PROCEDURES

Materials

(+)-[:'H]PN200-11O1 (75-85 Ci/mmol), (-)-['H]desmethoxyverapamil (84 Ci/mmol), and (+)-cis-['Hldiltiazem (143 Ci/mmol) were from Amersham, Vienna, Austria. (+)-BM 20.1140 (Fig. 1) and its optically pure enantiomers were kindly donated by Dr. Sponer, Bo- ehringer Mannheim GmbH, Germany. The calcium indicators, fura- 2 and Br-BAPTA were from Molecular Probes (Eugene, OR). The sources of the other unlabeled drugs and of most chemicals are given elsewhere (6, 29, 34).

Binding Studies with Membrane-bound and Purified Ca2+ Channels

Membrane Preparation-Partially purified T-tubule membranes from rabbit hind limb skeletal muscle and guinea pig cerebral cortex membranes were prepared and stored as described (29).

General Assay Conditions-All binding studies were performed in 60 mM Tris/HCl, 0.1 mM phenylmethylsulfonyl fluoride, pH 7.4 (buffer A), in a final assay volume of 0.5-1 ml. All drugs added were diluted in dimethyl sulfoxide (35) and directly added to the assay. The final dimethyl sulfoxide concentration never exceeded 1% (v/v). Receptor-bound radioligand was separated from unbound ligand by filtration with a medium (29) as recently described (34). All data reported in this paper refer to specifically bound radioligand (total binding minus radioligand bound in the presence of the nonspecific binding definition).

1,4-Dihydropyridine Binding-(+)-['H]PN200-110 (0.052-0.072 nM) was incubated in the absence or presence of the indicated concentrations of the allosteric modulators with 0.05-0.08 mg/ml guinea pig cerebral cortex membrane protein for 60 min at 37 "C; saturation analysis was performed over a range of radioligand concentration given in the legends to the figures. 0.12-0.960 nM (+)-["HI PN200-110 were incubated in the absence and presence of allosteric modulators with 0.013-0.018 mg/ml rabbit skeletal muscle microsomal protein a t 37 "C for 45 min. The definition of nonspecific binding was 1 p~ (+)-PN200-110 for guinea pig cerebral cortex membrane experiments and 3 p~ (+)-PN200-110 for rabbit skeletal muscle microsomes, respectively.

Phenylalkylamine Binding-0.39-0.43 nM (-)-["H]desmethoxyverapamil was incubated with 0.140-0.153 mg/ml of guinea pig cerebral cortex membrane protein in the absence or presence of allosteric regulators in buffer A or 0.52-0.83 nM (-)-[~'H]desmetboxyverapamil was incubated with 0.031-0.043 mg/ml of skeletal muscle T-tubule membrane protein as above. Nonspecific binding was defined by 3

' The abbreviations used are: PN200-110 (or isradipine), isopropyl 4-(2,1,3-benzoxadiazol-4-yl)-1,4-dihydro-2,6-dimethyl-5-methoxy- carbonyl-pyridine-3-carboxylate; BAPTA, 1,2-bis(o-aminophenoxy)- ethane-N,N,N',N'-tetraacetic acid; EGTA, [ethylenebis(oxy- ethylenenitriloj] tetraacetic acid; MOPS, 3-(N-morpholinoj- propanesulfonic acid.

p M (t)-desmethoxyverapamil for both membranes. Saturation analysis was performed as given in the legends to the figures.

(+)-ci.-fHlDiltiazem Binding--0.86-1.08 nM (+)-cis-['Hldilti- azem was incubated with 0.18-0.28 mg/ml of skeletal muscle T-tubule membrane protein in the absence and presence of allosteric regulators in buffer A for 60 min a t 30 "C. For (+)-ci~-[~H]diltiazem binding at 2 "C 0.056-1.32 nM (+)-cis-['Hldiltiazem was incubated with 0.124- 0.213 mg/ml rabbit skeletal muscle microsomal protein for 720 min as above. Nonspecific binding was defined by 10 p~ (+)-cis-diltiazem.

Kinetic Studies-For association kinetics of the Ca" channel- linked dihydropyridine receptors in guinea pig cerebral cortex membranes, allosteric regulators were added at the indicated concentrations in buffer A for 10 min a t 37 "C and the association reaction started by prewarmed radioligand. The reaction was terminated a t the given times by filtration (see above). The data were normalized employing the linearized, integrated second-order rate equation:

LO - Rdl * In[(& - x)/(& - x)] = k+, : t + A (1)

RO and Lo are the total concentrations of receptor and radioligand, x is the receptor-ligand complex at the indicated times, t , respectively and A is

[ l /(Lo - Rd] : ln(Lo/Ro)

Dissociation experiments were performed as previously published (29, 34). Data were transformed according to the following equation:

ln[B,/B,] = -k-, : t , (2)

where B, is the specifically bound radioligand at the indicated time after initiation of the dissociation by unlabeled ligand, and E, is the concentration of radioligand specifically bound a t equilibrium.

Purified Ca2+ Channels-L-type Ca2+ channels from skeletal muscle microsomes were purified up to the wheat germ agglutinin-seph- arose step (33) and stored without loss of binding activity for up to 2 months in liquid nitrogen.

Determination for Free Ca2+ and Ca'+ Dependence of Dihydropyridine Labeling

Incubations were done in buffer B (10 mM MOPS, 100 mM KCI, titrated with KOH a t 37 or 22 "C to pH 7.1, with 2.5 mM EGTA, 2.5 mM Ca'+-EGTA, or 2.5 mM Ca2+-EGTA plus additional amounts of CaC12). Cerebral cortex membranes from the guinea pig were pre- treated at 37 "C with EDTA and subsequently washed (at 2 "C) with Ca"-free (Chelex-100 treated-) buffer as described (29). Under these conditions >95% of the dihydropyridine binding is lost whereas phenylalkylamine binding is fully retained. The functional absence of the dihydropyridine receptor was also assessed by investigating the negative heterotropic allosteric regulation of (-)-["Hldesmethoxyver- apamil binding with unlabeled (+)-PN200-110 as reported (29). The determination of free Ca2+ and the assay conditions followed the protocol given by Ebata et al. (18) with some minor modifications. In brief, Ca2+-depleted cerebral cortex membranes were preincubated in buffer B with and without the positive heterotropic allosteric regulators and in the absence or presence of (+)-BM 20.1140 a t 37 "C for 15 min, followed by addition of the radioligand (+)-['H]PN200-110. The final concentration of EGTA in the assay for nominally -lo-' M free Ca'+ was 2.5 mM. Other free Ca'+ concentrations were prepared from mixing 2.5 mM Ca'+-EGTA buffers with 2.5 mM EGTA plus additional amounts of CaCL

Buffer B was adjusted to pH 7.1 at 22 "C and partially purified Ca'+ channels from skeletal muscle were preincubated as above for 15 min a t 22 "C, followed by addition of the radioligand. The definition of nonspecific binding was 10 p M (+)-PN200-110. For both types of assay, the final incubation time with radioligand was 60 min. For purified Ca'+ channels carrier proteins were added prior to filtration as described (29). The true concentration of free Ca'+ was measured in all buffers with the constituents of the respective binding assay, using the calcium-sensitive dye fura-2 (36) and 100 p M bromo- BAPTA (37), respectively. For fura-2 measurements, samples were excited at 340 and 380 nM, with emission a t 510 nM in a Perkin- Elmer-Cetus LS50 spectrofluorimeter. The 340:380 ratio was used to calculate free Ca'+. At free Ca'+ levels >1 p~ bromo-BAPTA was employed by measuring the absorbance change at 263 nM.

Determination of Ca'+ Antagonistic Activity

To determine Ca2+ antagonism of the BM 20.1140 enantiomers in a functional assay, the method described in detail elsewhere (38) was

Ca2+-binding Sites of the Ca2+ Channel 10789

employed. In brief, strips of taenia from the caecum of male guinea pigs were set up in 30-ml organ baths in K+-Tyrode (NaCl, 97 mM; KC1, 40 mM; NaHC03, 11.9 mM; NaH2P04, 0.4 mM; glucose, 5.5 mM), gassed with 95% 02, 5% C02 a t 35 "C. After complete relaxation of the strips (60 min), cumulative concentration-response curves for Ca'+ (0.03-1 mM) were obtained and the maximum isotonically re- corded tension taken as 100%. The enantiomers of BM 20.1140 (in the absence or presence of Bay K 8644) were preincubated a t different concentrations for 30 min and the Ca2+ concentration response curves repeated. EC,, values for Ca2+ in the absence or presence of drugs were calculated after log-logit transformation of the data and plotted according to (39) to obtain the pA2 value.

Protein Determination and Data Analysis Protein of the particulate membrane preparations was determined

according to Lowry et al. (40) and of partially purified skeletal muscle Ca'+ channels according to Bradford et al. (41) with bovine serum albumin as standard. Data are given as means f S.D. for a t least three independent experiments. Binding inhibition or stimulation curves were parameter-optimized using the general dose-response equation according to DeLean et al. (42). ICso and ECso values are the drug concentrations causing 50% inhibition or 50% stimulation, respectively. Inhibition is defined as 100 X (Bo-B)/B,, where B is specifically bound radiolabel in the presence and Bo in the absence of added drugs, respectively. Stimulation is defined as 100 X (BIB,). The dissociation constants ( K D ) and the maximal density of binding sites (BmaX) were obtained by linear regression analysis after Scat- chard transformation of the equilibrium saturation binding data.

RESULTS

(-)-BM 20.1140 Is a Potent Ca2+ Antagonis-(-)-BM 20.1140 inhibited Ca2+-induced contractions in taenia strips from guinea pig caecum (Fig. 2) with a pA2 value of 8.64 (ICbo value = 2.29 nM). In the presence of 1 pM Bay K 8644 (a L- type Ca2+ channel-activating dihydropyridine, see Ref. 43), the slope of the regression line was not significantly altered whereas the PA, value was decreased to 7.88 (ICbo value = 13.18 nM). The (+)-enantiomer was 22 times less active (PA, = 7.29; ICbo value = 51.28 nM), however, 1 p~ Bay K 8644 could not reverse the inhibitory effect of (+)-BM 20.1140. Thus, (-)-BM 20.1140 fulfilled criteria previously established for a subgroup of Ca2+ channel blockers which includes (+)- cis-diltiazem, verapamil, and antagonistic, dihydropyridines whereas (+)-BM 20.1140 behaved similar to another subgroup which included diphenylbutylpiperidines (pimozide) or flu- narizine (38).

(-)-BM 20.1 140 Stimulates Dihydropyridine Binding in

BM 20.1140 [ lOgao. MI

FIG. 2. Arunlakshana and Schild (39) plot of the antagonism of Ca'+-induced contractions by (-)-BM 20.1140 in the absence (0, PA, 8.64 f 0.2, slope 1.19) and in the presence of 1 PM Bay K 8644 (0, pA, 7.88 k 0.3, slope 1.24) and by (+)-BM 20.1140 in the absence (e, pA, 7.29 k 0.2, slope 0.95) and in the presence of 1 PM Bay K 8644 (0, PA, 7.14, slope 1.22), respectively. The results depicted are the means f S.E. of five to seven experiments.

Cerebral Cortex Membranes and (+)-BM 20.1140 Reverses the Stimulation-In Fig. 3A, it is shown that (-)-BM 20.1140 is a potent stimulator of dihydropyridine binding to guinea pig cerebral cortex membranes (mean ECb0 value = 5.02 f 1.2 nM, nH = 0.68 f 0.16). The maximum stimulation at 0.3-1 p~ was 169.4 f 18.8% ( n = 5). The (+)-enantiomer was without effect up to 1 p~ but was slightly inhibitory at 10 ptM (15 f 3.2%, n = 5). The mechanism of (-)-BM 20.1140 stimulation was first assessed by equilibrium saturation analysis and found to be mainly due to an increase in affinity (Fig. 3B) very similar to the effects of (+)-cis-diltiazem (23). We have calculated, using the equations derived by Ehlert (44), a * KA, the dissociation constant of an allosteric stimulator for the complex of dihydropyridine and receptor as well as a, the cooperativity factor. a was 0.238 k 0.129 ( n = 5) and a * KA = 2.98 f 1.05 nM. As a * KO is the dissociation constant of the radiolabeled dihydropyridine for binding to the complex of (-)-BM 20.1140 with the al-subunit, the predicted apparent KD of the experiments shown in Fig. 3B is 29 PM. KA, the calculated dissociation constant of (-)-BM 20.1140 for al- subunits was 12.5 nM. Kinetic analysis (Fig. 3, C and D) revealed that (-)-BM 20.1140 (1 p ~ ) accelerated the association reaction about 2.7-fold and the complex decay by 60% whereas (+)-cis-diltiazem (10 p ~ ) decreased the K+, by 30% and the K-I 2.8-fold. The dissociation constants (KD, apparent KD) derived from the kinetic constants were as follows. Con- trol: 66 PM, (+)-cis-diltiazem (10 p ~ ) present: 33 pM, (-)-BM 20.1140 (1 p~ present): 39 pM. The (+)-enantiomer of BM 20.1140 (at 10 p ~ ) accelerated the dissociation rate 3.8-fold and increased the K,, 2.3-fold, the kinetically derived apparent KD being 108 PM. The apparent equilibrium saturation KD with 0.3 p~ ( f ) -BM 20.1140 present was 43.8 PM, the control KD in the shown example 119 PM (see Fig. 3B). The reported average Ku is 75 PM (4). We have also investigated the racemate (f)-BM 20.1140. It stimulated dihydropyridine labeling with an ECao value of 21.6 f 6.5 nM, the maximum stimulation reached 161 f 10.1% (n = 3). At 1 p~ (+)-BM 20.1140 had similar effects on the kinetics as (-)-BM 20.1140 (not shown). Despite the apparently lacking action of (+)- BM 20.1140 on equilibrium binding (up to 3-5 p ~ ) it induced pronounced alterations of the dihydropyridine kinetics (Fig. 3, C and D). These findings prompted us to investigate the effects of the (-)-enantiomer in the presence of increasing concentrations of the (+)-enantiomer. An example of such an experiment is shown in Fig. 3F and indicated that the (+)- enantiomer shifted the stimulation curve of the (-)-enantiomer to the right. We have calculated apparent a * Ka values for (-)-BM 20.1140 dihydropyridine stimulation curves in the presence of increasing concentrations of the (+)-enantiomer. Replotting the data according to Ref. 44 yielded a slope of 0.72 and a PA, value of 7.4 (40 nM, Fig. 4). Similar experiments and calculations were performed with (+)-cis-diltiazem (Fig. 3E) as a positive heterotropic allosteric regulator. a was 0.25 in the absence of (+)-BM 20.1140 (KA for (+)-cis-diltiazem: 550 nM, reported IC6,, value at 37 "c = 137 nM, see Ref. 32) and increased to 0.56 at 5000 nM (+)-BM 20.1140 (KA = 4359 nM). The limited amount of data obtained did not allow replots as in Fig. 4. We estimated the KO, value of (+)-BM 20.1140 to reverse (+)-cis-diltiazem stimulation from these experiments: > 190 nM < 570 nM. As shown in Table I, the ICbo value of (+)-BM 20.1140 to reverse the stimulation by 1 pM (+)-cis-diltiazem was 196 nM. Clearly, (+)-BM 20.1140 is not a simple competitive antagonist for the sites on which (-)-BM 20.1140 exerts its allosteric regulation of the dihydropyridine receptors. As will be shown below (-)-BM 20.1140 is not a simple competitive ligand for benzothiazepine recep-

10790 Ca2+-binding Sites of the Ca2+ Channel

-10 -9 -8 -7 -6 -5 BH 20.1140 flOgi0. MI TIME [minuteal

-

0 .o

-0.5 - \

Y

g l . 5

m-1 .o

C r(

-2.0

-2.5 0 10 20 30 40 50

TIME [minUt66] -10 -9 -8 -7 -6

(-1 -BH 20.1140 [lop,,. HI FIG. 3. Effects of the BM 20.1140 enantiomers on guinea pig cerebral cortex membranes Caz+-

antagonist labeling. A, 0.072 nM (+)-[3H]PN200-110 were incubated with 0.080 mg/ml guinea pig cerebral cortex membrane protein in the presence of increasing concentrations of BM 20.1140. The concentrations given in parenthesis refer to specifically bound ligand at equilibrium. (-)-BM201140 (0, 13.2 PM) and ( f ) -BM 20.1140 (0, 13.6 PM) stimulated (+)-[3H]PN200-110 binding up to 172.5% f 16.1 and 161.1% f 10.1, respectively, with ECso values of 3.78 f 1.63 nM and 21.6 t 6.5 nM, respectively. (+)-BM 20.1140 (0, 13.2 pM) had no effect on (+)-[3H] PN200-110 binding up to 3 PM. B, saturation isotherm (Scatchard transformation) of equilibrium saturation data. KD and B,,, were calculated by linear regression analysis. 0.114 mg/ml guinea pig cerebral cortex membrane protein were labeled with 9.63-808.8 p~ (+)-[3H]PN200-110. Control (0): r = 0.98, KO = 119.1 p ~ , B,,, = 28.88 pM corresponding to 251.6 fmol/mg of protein; 300 nM ( f ) -BM 20.1140 present (+): r = 0.99, KD = 43.81 pM, BmaX = 35.35 PM corresponding to 307.1 fmol/mg of protein. C, linear transformation of the association data according to the second-order rate equation (see "Experimental Procedures") to determine k+l. Y axis intercepts were normalized (taking into account the stimulation of (+)-[3H]PN200-110 binding by (-)-BM 20.1140 and (+)-cis- diltiazem) to control association. Association was carried out in the absence (0) and presence of 1 PM (-)-BM 20.1140 (O), 10 PM (+)-BM 20.1140 (0), and 10 /IM (+)-cis-diltiazem (+). The slopes of the lines, as estimates of the association rate constants were: 0, control, 25.5 PM, 0.00037 PM" X rnin", r = 0.974; 0, (-)-BM 20.1140, 37.01 pM, 0.00099 pM" X rnin", r = 0.984; 0, (+)-BM 20.1140, 26.41 pM, 0.000865 pM" X rnin", r = 0.981; +, (+)-cis-diltiazem, 34.52 PM, 0.000266 PM" X rnin", r = 0.997. Specifically bound ligand at equilibrium ( B e ) is given in PM. D, guinea pig cerebral cortex membranes (0.078 mg/ml of protein) were labeled with 181.7 pM (+)- [3H]PN200-110 in the absence (O), 21.17 p~ or presence of 1 PM (-)-BM 20.1140 (O), 19.62 pM, 10 PM (+)-BM 20.1140 (0), 26.4 PM or 10 /IM (+)-cis-diltiazem (+), 27.1 PM, for 45 min at 37 "C. At the times indicated dissociation was initiated by 1 p~ (k)PN200-110. The following rate constants were obtained 0, K-l = 0.0234 rnin", r = 0.99; 0, K-, = 0.0367 rnin", r = 0.99; 0, = 0.106 rnin", r = 0.99; +, = 0.0088 rnin", r = 0.99. Specifically bound ligand at equilibrium (Be) is given in PM. E, guinea pig cerebral cortex membranes (0.129 mg/ml) were incubated with 61 p~ (+)-[3H]PN200-110 and the indicated concentrations of (+)-cis-diltiazem for 60 min at 37 "C, in the absence or presence of different concentrations of (+)-BM 20.1140. (+), control, 10.8 pM, ECso value 202 f 2.6 nM, n~ 1.61 f 0.41, maximal stimulation 153.7%; 0, 50 nM (+)"'VI 20.1140, 11.4 pM, E C ~ O value 220 f 4.6 nM, n~ 1.78 It 0.46, maximal stimulation 156.7%; A, 500 nM (+)-BM 20.1140, 14.2 pM, EC50 value 593 f 22 nM, n~ 1.15 t 0.68, maximal stimulation 142.7%; ., 5000 nM (+)-BM 20.1140, 13.8 pM, ECso value 1807 3z 162 nM, n~ 1.81 t 0.68, maximal stimulation 126.3%. F, stimulation by (-)-BM 20.1140 in the presence of different concentrations of (+)- BM 20.1140. (+)-[3H]PN200-110 (68.0 PM) was incubated with 0.054 mg/ml cerebral cortex membranes for 45 min at 37 "C. 0, control, 11.93 PM, EC,, = 4.51 f 0.2 nM, maximal stimulation 182.4 f 3.1%, n~ = 0.57 f 0.14; +, 10 nM (+)-BM 20.1140, 12.4 PM, EC,, = 8.8 t 1.4 nM, maximal stimulation 174 * 4.6%, nH = 0.72 t 0.11; A, 300 nM (+)-BM 20.1140, 11.6 p ~ , EC,, = 20.9 f 3.4 nM, maximal stimulation 171.4 f 6.1%, n~ = 0.62 f 0.19. Radioligand concentration was 68 pM, membrane protein 0.054 mg/ml.

tors, although it shares the stimulatory action of (+)-cis- cortex membranes. (-)-BM 20.1140 value = 5.8 nM), diltiazem on dihydropyridine labeling. (+)-BM 20.1140 (IC50 value = 51.2 nM), and the racemic

(-)-BM 20.1 140 Is an Allosteric Ligand for Dihydropyridine, compound (ICso value = 8.7 nM) were potent but only partial Phenylalkylamine, and Benzothiazepine Receptors-We next inhibitors of the phenylalkylamine sites of cerebral cortex L- investigated the effects of BM 20.1140 on three well-charac- type Ca2+ channels. For the three compounds the maximal terized drug receptor domains of a,-subunits in skeletal mus- average inhibition was 82.6 rfr 3.2% (not shown). The inter- cle microsomes and on phenylalkylamine receptors in cerebral action with the benzothiazepine-selective domain was studied


1.0 - * I

p: 0.5 0

- -

-8 -7 -6 (+) -BM 20.1140 [log,,. MI

FIG. 4. Arunlakshana and Schild (39) analysis of reversal of (-)-BM 20.1140 stimulation by (+)-BM 20.1140. ECso Val- ues and maximal stimulation were employed to calculate (Y, the cooperativity factor, and KA, the dissociation constant of (-)-BM 20.1140, according to Ehlert (44). First, the Y' value was derived, which is the fractional binding ( Y/Yo) of radioligand in the absence ( Y ) and in the presence (Yo) of infinitely large concentrations of a positive allosteric regulator, respectively. Employing the mean reported Kr, for (+)-[3H]PN200-110 (4), a was calculated from equation 8 from Ref. 44, and KA from equation 10. Calculated a : KA values obtained for different concentrations of (+)-BM 20.1140 were con- verted into dose-ratios (DR) where dose-ratios stands for [ a : KA] n p p / [ ~ : KA], where (Y * Ka is the mean dissociation constant of (-)- BM 20.1140 for the complex of (+)-[3H]PN200-110with a a,-subunits (2.98 nM, see "Results") and [ a * &Iapp the value in the presence of the indicated concentrations of (+)-BM 20.1140. The pA2 value was 7.4, the slope 0.72, r = 0.97.

TABLE I (+)-BM20.1140 reversal of stimulation of (+)-['H]PN200-110

binding induced by various positive heterotropic allosteric regulators In these experiments we have tested (in two different membranes)

various stimulators of dihydropyridine binding (at concentrations given in parenthesis) and repeated these measurements in the presence of increasing concentrations of (+)-BM 20.1140. Results from two independent (+)-BM 20.1140 inhibition experiments (each performed in duplicates) were pooled and computer-fitted to the general dose-response equation (42). In the presence of saturating concentrations of (+)-BM 20.1140 (10 p ~ ) control binding levels were achieved in all cases at 10 p ~ , except for (-)-BM 20.1140. For experiments with skeletal muscle microsomes, a mean protein concentration of 0.011 mg/ml, for cerebral cortex membranes of 0.066 mg/ml was used; the ligand concentrations, incubation volumes, and times were as described under "Experimental Procedures."

Compound/tissue % Stimulation" IC,, valueb nHc

nM

Skeletal muscle microsomes Fostedil (1 p ~ ) 208 * 19 152 f 54 0.79 * 0.12 (+)-Tetrandrine (1 p ~ ) 157 t 23 164 f 21 0.66 * 0.51 (+)-cis-Diltiazem (1 p M ) 148 * 16 255 f 89 1.11 * 0.63

Cerebral cortex membranes Fostedil (1 p M ) 176 f 24 229 * 42 1.44 f 0.32 (+)-Tetrandrine (1 p ~ ) 172 f 12 245 f 11 1.09 f 0.42 (+)-cis-Diltiazem (1 p ~ ) 145 * 13 196 f 94 >1.5 (-)-BM 20.1140 (0.3 p ~ ) 161 f 5 19 f 15d >1.5

100% = control binding in the absence of positive heterotropic allosteric regulators.

bThe IC,, value is the concentration of (+)-BM 20.1140 which causes 50% of reversal of stimulation induced by the allosteric regulators.

nH = apparent Hill slope of the (+)-BM 20.1140 inhibition curve. Only partial reversal to 132 f 6% of control binding at 10 p~

(+)-BM 20.1140.

in skeletal muscle microsomes (Fig. 5A) at two different temperatures. At 2 "C (+)-cis-diltiazem inhibited with an IC50 value of 99.9 f 18.2 nM (nH = 0.9 f 0.12), (-)-BM 20.1140 with an IC50 value of 21.6 f 16.4 nM (nH = 1.1 k 0.11, (+I- BM 20.1140 with an IC5, value of 139 f 37 nM (nH = 1.13 f 0.29) and (+)-BM 20.1140 with an ICs0 value of 63.8 f 33.2 nM (nH = 1.15 f 0.12) (n = 3, not shown). At 30 "c (+)-cis- diltiazem inhibited with an IC50 value of 84 nM (nH = 1.1). (-)-BM 20.1140 was an extremely potent inhibitor (IC50 value: 2.16 f 1.04 nM; nH = 1.11 k 0.1; n = 3), (+)-BM 20.1140 had an ICs0 value of 66.5 f 20.5 nM (nH = 1.11 f 0.1; n = 3) and (k)-BM 20.1140 inhibited with an ICso value of 5.56 2.60 nM (nH = 1.48 f 0.13; n = 5). The steep Hill slope for the racemate was not expected from the inhibition data of the two enantiomers. It was, however, confirmed in five separate experiments. Dissociation kinetics of the benzothiazepine receptors at 2 "C (Fig. 5B) revealed that (-)-BM 20.1140 slowed and (+)-BM 20.1140 accelerated dissociation. (+)-BM 20.1140 (1 FM) accelerated the decay of the (+)-cis- [3H]diltiazem-al-subunit complex to the same extent as (-)- BM 20.1140. These data prove that BM 20.1140 is not a simple competitive ligand for the benzothiazepine receptor.

(-)-BM 20.1140 was also a stimulator of dihydropyridine binding in skeletal muscle microsomes (Fig. 5C), but in contrast to cerebral cortex membranes the (+)-enantiomer was more inhibitory at concentrations >0.1 p ~ . Heparin was recently discovered to be a L-type Ca2+ channel activator which allosterically inhibits the phenylalkylamine, benzothiazepine, and dihydropyridine receptors with high affinity at the level of the al-subunit (34). A unique feature of the interaction of heparin with membrane-bound L-Type Ca2+ channels in mammalian skeletal muscle microsomes is the increase of its inhibition constant for the dihydropyridine receptors when the benzothiazepine site is liganded with (+)- cis-diltiazem. Fig. 5 0 shows that (+)-BM 20.1140 prevents this characteristic phenomen.

In rabbit skeletal muscle membranes, (-)-[3H]desmethoxyverapamil labeling was completely inhibited by (-)-BM 20.1140 (IC50 value: 1.94 f 0.24 nM), (+)"VI 20.1140 (IC50 value 102.6 f 32 nM) and (k)-BM 20.1140 (IC50 value: 7.6 f 1.2 nM). The apparent mean Hill slopes were not significantly different from unity. An example is shown in Fig. 5E. Analysis of the mechanism of inhibition yielded a clearly noncompet- itive type of interaction of (-)-BM 20.1140 with the phenylalkylamine sites (decrease of B,,,, increase in apparent KO, see Fig. 5F) . Taken together, BM 20.1140 does not bind in a simple competitive manner to any of the three well-defined receptor sites for dihydropyridines, phenylalkylamines, or benzothiazepines.

(+)-BM 20.1140 Inhibits the Affinity Increase of the al- Subunit for Free Ca2+ Induced by Positive Heterotropic Regu- lators-In Fig. 3, E and F, we have shown that (+)-BM 20.1140 can reverse the dihydropyridine stimulation by (+)- cis-diltiazem and (-)-BM 20.1140 in cerebral cortex membranes. In order to investigate if the reversal was tissue independent, the effects of increasing concentrations of the (+)-enantiomer at fixed concentrations of various allosteric stimulators were measured in skeletal muscle and cerebral cortex membranes. The results (Table I) demonstrate that reversal occurred in both membranes and for stimulators from different chemical classes. This prompted us to look for a more basic mechanism of action, shared by the positive heterotropic allosteric regulators but counteracted by (+)-BM 20.1140. As explained elsewhere in detail, dihydropyridine binding to the al-subunit is interdependent with Ca2+ binding (see Ref. 3 for a review) and the data are best described by a


2 4

150

6 I- t

0 .o

- m' -0.5 \

Y Y m

5 -1.0

-1.5 0 15 30 45 60 75 90

TIME [minuteel - 2 - 1 0 1 2 3

heparin t log,o ug/mll

FIG. 5. Effects of the BM 20.1140 enantiomers on rabbit skeletal muscle Ca2+-antagonist labeling. A, 1.27 nM (+)-~is-[~H]diltiazem were incubated with 0.174 mg/ml rabbit skeletal muscle T-tubule membrane protein for 60 min at 30 "C. ICs0 values and slopes ( n ~ values) are calculated as described under "Experimental Procedures" and the best fit curves drawn with the following parameters. (-)-BM 20.1140 (O), 14.1 PM, ICs0 1.10

17.7 PM, IC, 46.77 f 6.21, n~ 1.06 f 0.1. The concentrations given in parenthesis refer to specifically bound ligand at equilibrium. The data points are means from duplicate determinations. B, dissociation kinetics of (+)-ci~-[~H] diltiazem for rabbit skeletal muscle Ca2+ channels: 0.23 mg/ml rabbit skeletal muscle microsomal protein was labeled with 1.01 nM (+)-~is-[~H]diltiazem at 2 "C for 720 min. Dissociation (time zero) was initiated by addition of 10 p M (+)-cis-diltiazem (0), control, k-l = 0.0102, r = 0.98); 1 p M (-)-BM 20.1140 (e), k-1 = 0.0074, r = 0.99; 10 p~ (+)-BM 20.1140 (0), k-, = 0.0146, r = 0.99. C, 0.161 nM (+)-[3H]PN200-110 was incubated with 0.013 mg/ ml skeletal muscle microsomes for 45 min at 37 "C in the presence of increasing concentrations of BM 20.1140. At equilibrium the specifically bound ligand was 21.2 PM. (-)-BM201140 (0) and (f)-BM 20.1140 (0) stimulated (+)- [3H]PN200-110 binding up to 148.2 f 3.3% and 136.2 f 4.2, respectively, with ECw values of 5.44 f 0.4 nM and 8.87 f 2.8 nM, respectively. (+)-BM 20.1140 (0) inhibited (+)-[3H]PN200-l10 binding with an IC60 value > 10 pM. D, rabbit skeletal muscle microsomal protein (0.013 mg/ml) were incubated with 0.26 nM (+)-[3H]PN200-l10 in the presence of increasing concentrations of heparin in the absence (O), 86.3 pM and presence (O), 97.5 pM of 3 p~ (+)-cis-diltiazem or (0) 3 p~ (+)-cis-diltiazem and 1 p M (+)-BM 20.1140, respectively. The following binding for the heparin inhibition data were calculated 0, IC50 2.21 f 0.12 pg/ml, n~ 1.73 f 0.11; 0, IC60 8.21 f 0.67 pg/ ml, n~ 0.84 f 0.09; 0, ICso 3.12 f 0.14 pg/ml, n~ 1.07 f 0.06. E, 0.63 nM (-)-[3H]desmethoxyverapamil was incubated with 0.046 mg/ml rabbit skeletal muscle microsomal protein for 60 min at 22 "C. The following binding parameters were obtained for the shown experiments: (-)-BM 20.1140 (O), 39.4 pM; ICs0 2.29 f 0.11 nM, nH 1.18

6.4 nM, n~ 1.19 f 0.12. F, saturation isotherms (Scatchard transformation) for phenylalkylamine receptors. Rabbit skeletal muscle microsomes (0.032 mg of protein/ml) were incubated with (-)-[3H]desmethoxyverapamil (0.29- 17.49 nM). Control conditions (O), KD 2.58 nM, Bmax 191.1 pM, r = 0.99; addition of 1 nM (-)-BM 20.1140 (e), KO 3.67 nM, Emax 173.8 pM, r = 0.98; addition of 3 nM (- )"VI 20.1140 (A), KD 4.41 nM, Emax 135.1 pM, r = 0.99.

f 0.1 nM, n~ 1.15 f 0.14; (f)-BM 20.1140 (a), 11.2 pM, ICs0 4.15 f 0.32 nM, n~ 1.53 f 0.1; (+)"VI 20.1140 (0)

f 0.17; (f)-BM 20.1140 (0), 42.6 pM; IC60 8.7 f 1.2 nM, n~ 1.32 f 0.1; (+)"VI 20.1140 (o), 45.0 pM, ICs0 74.13 f

ternary complex model (Ca2+, al-subunit, dihydropyridine). It was recently reported that the K0.5 value for free Ca2+ to restore dihydropyridine binding to cardiac sarcolemma microsomes was 407 and 269 nM for partially purified Ca2+ channels from rabbit skeletal muscle (18). The Hill coefficients were 0.67 and 0.56, respectively. We can confirm these data for purified skeletal muscle Ca2+ channels (Fig. 6A). The KO, value in our hands was 301 nM, the Hill coefficient 0.70. With (+)-tetrandrine (10 p ~ ) present, the K0.5 value decreased to 30.1 nM and the Hill coefficient rose to 1.37. Simultaneous addition of 10 p~ (+)-BM 20.1140 and (+)tetrandrine (10 p M ) increased the value to 251 nM. To exclude that these effects were unique for skeletal muscle, we performed similar

experiments with L-type Ca2+ channel dihydropyridine labeling in cerebral cortex membranes (Fig. 6B). The values for free Ca2+ was 132 nM (nH = 1.02), and in the presence of 1 p~ (+)-tetrandrine the affinity for free Ca2+ increased 4- fold value 27.0 nM). Again, as in skeletal muscle, (+)- BM 20.1140 (1 p ~ ) reversed the affinity increase = 70.8 nM). Similar experiments were performed with (+)-cis-diltiazem (Fig. 6C) and (-)-BM 20.1140 (Fig. 6D) as structurally different positive heterotropic allosteric regulators. 0.3 p M (-)-BM 20.1140 and 1 p~ (+)-cis-diltiazem decreased the KO.& value for Ca2+ in cerebral cortex membranes to 60.1 and 67.9 nM, respectively. Again, (+)-BM 20.1140 reversed the affinity increase. (+)-BM 20.1140, when added alone, decreased the


CaP+ free [log,o, M I

-8 -7 -6 Caz* f ree [log,o, HI

0 1 -

" -8 -7 -6 Cas* f ree [ log,o. M I

Ca2* free log,^. H ]

FIG. 6. Reversal of (+)-tetrandrine, (+)-cis-diltiazem, and (-)-BM 20.1140 induced Ca2+ affinity increase by (+)-BM 20.1140. A, partially purified skeletal muscle dihydropyridine receptors (0.0098 mg of protein/ml) were incubated with 2.1 nM (+)-[3H]PN200-110 and increasing concentration of EGTA/CaC12 in buffer B for 60 min at 22 'C. For details see "Experimental Procedures." The following EC50 and n~ values were obtained by computer fitting: 0, control, ECso value = 300.1 & 14 nM, n~ 0.699 & 0.11; 0, 10 p M (+)-tetrandrine, ECso value = 30.1 & 3.4 nM, n~ 1.374 & 0.3; 0, 10 p M (+)-tetrandine and 10 ~ L M (+)-BM 20.1140, ECso value = 251.3 & 3.1 nM, nH 0.992 * 0.15. B, EDTA-treated guinea pig cerebral cortex membranes (0.129 mg of protein/ml) were incubated with 0.156 nM (+)-[3H]PN200-110 and increasing concentration of EGTA/CaC12 in buffer B for 60 min at 37 "C. The following EC,, and n~ values were obtained by computer fitting: 0, control, EC50 value = 104.4 * 1.8 nM, n~ 0.844 & 0.86; 0, 1 p M (+)-tetrandrine, EC50 value = 27.0 & 0.3 nM, IZH 1.58 * 0.45; 0, 1 p M (+)- tetrandrine and 1 p~ (+)-BM 20.1140, EC50 value = 70.8 & 2.3 nM, n~ 1.53 f 0.15. C, EDTA-treated guinea pig cerebral cortex membranes (0.072 mg of protein/ml) were incubated with 0.197 nM (+)-[3H]PN200-110 and increasing concentration of EGTA/CaC12 in buffer B for 60 min at 37 "C. The following EC50 and nH values were obtained by computer fitting: 0, control, EC50 value = 159.5 & 1.6 nM, n~ 1.20 & 0.17; 0, 1 p~ (+)-cis-diltiazem, ECso value = 67.9 & 0.7 nM, n~ 1.62 & 0.21; 0, 1 p M (+)-cis-diltiazem and 1 p~ (+)-BM 20.1140, EC50 value = 141.2 & 4.3 nM, nH 1.33 * 0.11. D, conditions as described for C. The following EC5, and n~ values were obtained by computer fitting: 0, control, EC50 value = 159.5 & 1.6 nM, n~ 1.20 & 0.17; 0, 300 nM (-)-BM 20.1140, EC5, value = 60.1 & 0.4 nM, n~ 1.51 & 0.19; +, 10 p M (+)-BM 20.1140, EC50 value = 297.3 & 5.4 nM, TLH 1.45 * 0.40; 0, 300 nM (-)-BM 20.1140 and 1 p~ (+)-BM 20.1140, E C ~ O value = 191.5 * 4.3 nM, n~ 1.30 & 0.43. Data shown are normalized for residual radioligand binding (= <5% of control binding) a t Ca2' free nominally <1 nM (=O% of control binding) and maximal binding at 10 (skeletal muscle) or 3 p~ (cerebral cortex) free Ca2+ (=loo%).

for Ca2+ in cerebral cortex membranes (Fig. 6D). Thus, although only representatives from the chemically heteroge- nous positive allosteric regulators were investigated, the Ca2+ affinity increase as well as the (+)-BM 20.1140 reversal is likely to be tissue independent and reflecting a general mechanism.

DISCUSSION

(-)-BM 20.1 140 Is a Potent ea2+ Antagonist and a Positive Heterotropic Allosteric Regulator of 1,4-Dihydropyridine Bind- ing-Our functional experiments with smooth muscle show that (-)-BM 20.1140 is a potent, stereoselective Ca2+ antagonist (45) with an eudismic ratio of 22. Surprisingly, the (+)- but not the (-)-enantiomer completely blocked the effects of the Ca2+ channel activator Bay K 8644. Such a behavior was previously described for highly lipophilic drugs which inhibit smooth muscle contraction presumably via an intracellular mechanism (38). Clearly, further studies are required to define the activities of both enantiomers on whole cell Ca2+ currents or at the single channel level.

Both enantiomers are potent allosteric regulators for the three distinct, well-defined drug receptor domains of the cy1-

subunit of the L-type calcium channel (1, 3, 4). (-)-BM 20.1140 was a positive heterotropic allosteric regulator of dihydropyridine binding, and (+)-BM 20.1140 was able to reverse the effects of (-)-BM 20.1140. We suggest that the two optical enantiomers of BM 20.1140 bind to different sites as has recently been shown for the enantiomers of DPI 201- 106 which act on Na+ channels (46). The difference between the two BM 20.1140 enantiomers resides in their differential effects on receptor dissociation in cerebral cortex membranes whereas association is accelerated to a similar extent. Raising the temperature will increase k+l but simultaneously k 1 .

Thus, an induced conformational change which mimics ele- vated temperature can be ruled out, and we are not aware of a drug which resembles BM 20.1140 with respect to dihydropyridine receptor association kinetics. (+)-BM 20.1140 also reversed the stimulation of dihydropyridine binding induced by fostedil, trans-diclofurime, (+)-cis-diltiazem, and (+)-te-


trandrine, without being a simple competitive ligand for the benzothiazepine ((+)-cis-diltiazem-selective) site as is, e.g. (+)tetrandrine (26).

Positive Allosteric Regulators Have Different Effects on Li- gand Kinetics but Share a Common Mechanism-The above findings prompted us to look for a common mechanism through which these positive allosteric regulators act. Previ- ously we suggested that (+)-cis-diltiazem increased the affinity of the Ca'+ channel for Ca2+ (47), based on kinetics (47) and ion requirement experiments (3). These earlier studies (14, 15), including our own (47), reported only data for total (added) cations. However, recently the dependence of dihydropyridine labeling of L-type Ca2+ channels on the free Ca2+ concentration was reported for the first time (18). Saturation analysis (performed in different tissues) revealed that only the density of high affinity sites for dihydropyridines but not their affinity is increased as a function of (total or) free Ca2+ (14-18). values therefore reflect a Ca2+ concentration where the fraction of a,-subunits that are in a high affinity state for dihydropyridines is half-maximal. The coupled equilibrium equations for a ternary complex of Ca'+, Lul-subunit, and dihydropyridine (3) predict that the affinity for Ca'+ must be extremely low for al-subunits that are not liganded by the dihydropyridine. Conversely, antagonistic dihydropyridines induce a high affinity state of cwl-subunits that in the ternary complex model have high affinity for Caz+. This view is in agreement with others (18, 48). Interestingly, the >95% con- version of a,-subunits in cerebral cortex membranes into a state of very low affinity for dihydropyridines left phenylalkylamine binding completely unaffected, even at nominal free Ca2+ concentrations of 1 nM. Ca2+ KO values of some calcium- binding proteins may be as low as 1 or 10 nM (18), but we regard it as unlikely that the phenylalkylamine-binding domain is directly coupled to Ca2+ coordination sites of such a high affinity. It is more likely that sidedness problems (as discussed in Refs. 18, 49) or Ca'+ in a completely protected (locked) conformation (3) may be responsible for the apparent lack of dependence of phenylalkylamine labeling on free Ca'+. Indeed, and in contrast to membrane-bound Ca'+ channels in cerebral cortex, phenylalkylamine labeling of purified rabbit skeletal muscle Ca2+ channels is highly dependent' on free Ca'+ as it is on the presence of dihydropyridines that are functionally channel blockers (33).

Having confirmed the results from the previous report with respect to the free Ca2+ requirement of high affinity dihydropyridine binding (18), we tested the hypothesis that the positive heterotropic allosteric regulators increase the affinity of the a,-subunit for free Ca'+. We demonstrated a very signifi- cant decrease in half-maximal Ca2+ requirement for high affinity dihydropyridine labeling by (+)-tetrandrine, (+)-cis- diltiazem, and (-)-BM 20.1140 and furthermore, inhibition of the decrease by (+)-BM 20.1140. Dihydropyridine receptor kinetics is affected by the chemically diverse positive allosteric regulators in two different patterns. (+)-cis-Diltiazem mainly slowed the off kinetics and (-)-BM 20.1140 mainly accelerated association, the result being stimulation of equilibrium dihydropyridine binding. Despite the different effects on kinetic constants for dihydropyridine binding, both (+)-cis- diltiazem and (-)-BM 20.1140 increased the affinity for free Ca2+. As for dihydropyridine receptor binding, the changes in

values for free Ca'+ in the ternary complex model (3) must reflect alterations in the kinetic constants of Ca2+ binding. Although we have not measured Ca2+-binding kinetics, we can formulate the following hypothesis (see also Fig. 7): Ca2+- and dihydropyridine-binding sites are interdependent

H. G. Knaus, unpublished observations.

[m prototype: ( - ) -MI 20.1140

/ \

K< (1 ,4-dihydropyridine receptor -

K-\a,T \-/

prototype: (+)-cis-diltiazem

FIG. 7. Positive heterotropic allosteric regulators of dihydropyridine-binding-postulated mechanism. The interdependent Ca2+- and dihydropyridine-binding sites of the al-subunit are symbolized by ouerlapping circles. (-)-BM 20.1140 stimulated dihydropyridine binding mainly by increasing the association rate of the dihydropyridine and is the prototype for a class of drugs (class I). (+)-cis-Diltiazem is a representative for another class of allosteric regulators (class ZZ) which predominantly reduces the dissociation rate constant for dihydropyridines. Arrows point to the kinetic constant which is preferentially altered. The model postulates that the increase in affinity of the al-subunit for Ca2+ is the mechanism through which the allosteric regulators (albeit by altering dihydropyridine kinetics differently) indirectly increase affinity for dihydropyridines. Furthermore, we assume that the action of the allosteric regulators on Caz+ association and dissociation rate constants mimics the measured effects on dihydropyridine-binding kinetics, as would be predicted from the coupled equilibria of a ternary complex model (3).

in such a manner that changes in the affinity constants for Ca2+, induced by positive allosteric regulators, are revealed by parallel changes of dihydropyridine receptor kinetics. In other words, we postulate that (+)-cis-diltiazem mainly slows dissociation of Ca'+ whereas (-)-BM 20.1140 preferentially fa- cilitates association of Ca'+ with the channel (see Fig. 7).

We have not integrated negative heterotropic allosteric regulators in our model. However, (+)-BM 20.1140, which is a (weak) negative heterotropic allosteric regulator of dihydropyridine binding, decreased the affinity for free Ca'+.

Structural Aspects-Coordination of Ca2+ ions in proteins has been studied at the crystallographic level. These struc- tyres show seven oxygen atoms at an average distance of 2.4 A from the Ca'+ ion, with either all ligands from atoms on the protein or with one, two, three, or even six water ligands in a pentagonal bipyramidal geometry (50). The exact location of the dihydropyridine-binding domain as well as that of the interdependent Ca2+-binding sites in the al-subunit primary amino acid sequence is not yet known. However, it is reason- able to assume that both domains are formed by noncontig- uous (separate) segments of the pore-forming polypeptide orientated to (or accessible from) the extracellular side (18- 20). This, of course, would exclude that the internal EF hand motif (11) is part of a high affinity dihydropyridine antagonist-binding domain. The positive allosteric regulators induce a conformational change to optimize the Ca2+ coordination of the sites that are linked to the dihydropyridine-binding domain. According to our hypothesis, binding of allosteric regulators alters Ca'+ kinetics, and it is tempting to speculate that the different classes of allosteric regulators induce changes in different parts of Ca'+ coordination sites. These conformational changes could affect entire helices or only 1 amino acid residue within a helix.

Ca2+-binding Sites

Target size analysis data of the dihydropyridine receptor in different tissues (summarized in Ref. 4) demonstrate a large decrease (180-110 kDa) of radiation-sensitive mass, induced by (+)-, but not by (-)-cis-diltiazem. This finding argues for large conformational changes of the al-subunit, as is also suggested by recent observations with heparin (34). (+)-BM 20.1140 blocked the characteristic affinity change for heparin when the membrane-bound Ca2+ channel in skeletal muscle microsomes was liganded with (+)-cis-diltiazem (34), but it remains to be tested if it can prevent the 70-kDa decrease in radiation-sensitive mass of the ul-subunit induced by this benzothiazepine diastereomer (4).

The al-subunit of the L-type Ca2+ channel is an extraordi- nary complex structure with multiple drug and divalent cation-binding sites. The enantiomers of BM 20.1140 have proved useful tools to refine the complex, interdependent interactions between dihydropyridines, other drug receptor sites, and Ca'+. The demonstration that increases in Ca2+ affinity by (+)-cis-diltiazem and (-)-BM 20.1140 (but not (+)-BM 20.1140) are due to different effects on kinetics of Ca2+-dependent dihydropyridine binding underlies the indis- putable utility of pure enantiomers. On the other hand we realize how little is currently known at the microscopic level of conformational changes within the al-subunit. I t is hoped that elucidation of the topology of the dihydropyridine-binding domain may give further clues and lead to refinement of the model.

Acknowledgments-We would like to thank Dr. Sponer for a gift of BM 20.1140 and its enantiomers, the Bayer AG (Wuppertal) for dihydropyridines, the Knoll AG (Ludwigshafen) for phenylalkylamines, and Goedecke AG (Freiburg) for benzothiazepines. We also thank David Ferry for very helpful discussions and suggestions in the final draft of the manuscript.

REFERENCES 1. Janis, R. A., Silver, P. J., and Triggle, D. J. (1987) Adu. Drug.

2. Hosey, M. M., and Lazdunski, M. (1988) J. Membr. Biol. 104,

3. Glossmann, H., and Striessnig, J. (1988) Vitam. Horm. 44, 155- 328

4. Glossmann, H., and Striessnig, J. (1990) Reu. Physiol. Biochem. Pharrnacol. 114, 1-105

5. Leung, A. T., Imagawa, T., and Campbell, K. P. (1987) J. Biol. Chem. 262, 7943-7946

6. Vaghy, P. L., Striessnig, J., Miwa, K., Knaus, H.-G., Itagaki, K., McKenna, E., Glossmann, H., and Schwartz, A. (1987) J. Biol. Chem. 262, 14337-14342

7. Naito, K., McKenna, E., Schwartz, A,, and Vaghy, P. L. (1989) J. Biol. Chem. 264, 21211-21214

8. Striessnig, J., Scheffauer, F., Mitterdorfer, J., Schirmer, M., and Glossmann, H. (1990) J. Biol. Chem. 265, 363-370

9. Tsien, R. W., Hess, P., McCleskey, E. W., and Rosenberg, R. L. (1987) Annu. Reu. Biophys. Biophys. Chem. 16, 265-290

10. Friel, D. D., and Tsien, R. W. (1989) Proc. Natl. Acad. Sei. U. S. A. 86,5207-5211

11. Babitch, J. (1990) Nature 346, 321-322 12. Grabner, M., Friedrich, K., Knaus, H. G., Striessnig, J., Schef-

fauer, F., Staudinger, R., Koch, W. J., Schwartz, A,, and Gloss- mann, H. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 727-731

13. Striessnig, J., Glossmann, H., and Catterall, W. A. (1990) Proc. Natl. Acad. Sci. U. S . A. 87, 9108-9112

Res. 16, 309-591

81-105

of the Ca2' Channel 10795

14. Gould, R. J., Murphy, K. M. M., and Snyder, S. H. (1983) Mol. Pharmacol. 25, 235-241

15. Luchowski, E. M., Yousif, F., Triggle, D. J., Maurer, S. C., Sarmiento, J . G., and Janis, R. A. (1984) J. Pharmacol. Exp. Ther. 230, 607-613

16. Gould, R. J., Murphy, K. M. M., and Snyder, S. H. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 3656-3660

17. Glossmann, H., and Ferry, D. R. (1983) Naunyn-Schmiedeberg's Arch. Pharmacol. 323, 279-291

18. Ebata, H., Mills, J. S., Nemcek, K., and Johnson, J. D. (1990) J. Bid. Chem. 265, 177-182

19. Affolter, H., and Coronado, R. (1986) Biophys. J . 4 9 , 767-771 20. Kass, R. S., Arena, J. P., and Chin, S. (1989) Am. J. Cardiol. 64,

21. Ferry, D. R., and Glossmann, H. (1983) Br. J . Pharmacol. 78, 81 (abstr.)

22. Linn, T., Ferry, D. R., and Glossmann, H. (1983) Naunyn- Schmiedebergk Arch. Pharmacol. 324, R45 (abstr.)

23. Ferry, D. R., and Glossmann, H. (1982) Naunyn-Schmiedeberg's Arch. Pharmacol. 321, 80-83

24. Mir, A. K., and Spedding, M. (1987) J. Cardiouasc. Pharrnacol. 9,469-477

25. De Pover, A., Matlib, M. A., Lee, S. W., Dub6, G. P., Grupp, I. L., Grupp, G., and Schwartz, A. (1982) Biochem. Biophys. Res. Commun. 108, 110-117

26. King, V. F., Garcia, M. L., Himmel, D., Reuben, J. P., Lam, Y. T., Pan, J., Han, G., and Kaczorowski, G. J. (1988) J . Biol. Chem. 263, 2238-2244

27. Lee, H. R., Jaros, J . A., Roeske, W. R., Wiech, N. L., Ursillo, R., and Yamamura, H. I. (1985) J. Pharmacol. Exp. Ther. 233,

28. Ferry, D. R., and Glossmann, H. (1982) FEBS Lett. 148, 331-

29. Glossmann, H., and Ferry, D. R. (1985) Methods Enzymol. 109,

30. Glossmann, H., Linn, T., Rombusch, M., and Ferry, D. R. (1983) FEBS Lett. 160, 226-232

31. Garcia, M. L., King, V. F., Siegl, P. K. S., Reuben, J. P., and Kaczorowski, G. J. (1986) J. Biol. Chem. 261, 8146-8157

32. Schoemaker, H., and Langer, S. Z. (1985) Eur. J . Pharmacol. 11 1, 273-277

33. Striessnig, J., Goll, A., Moosburger, K., and Glossmann, H. (1986) FEBS Lett. 197, 204-210

34. Knaus, H.-G., Scheffauer, R., Romanin, C., Schindler, H.-G., and Glossmann, H. (1990) J. Biol. Chem. 265, 11156-11166

35. Boer, R., Grassegger, A., Schudt, C., and Glossmann, H. (1989) Eur. J . Pharmacol. 172, 131-145

36. Grynkiewicz, G., Poenie, M., and Tsien, R. Y . (1985) J. Biol. Chem. 260,3440-3450

37. Tsien, R. Y. (1980) Biochem. J. 19, 2396-2404 38. Spedding, M., and Berg, C. (1984) Naunyn-Schmiedeberg's Arch.

39. Arunlakshana, O., and Schild, H. 0. (1959) Br. J . Pharmacol. 14,

40. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, F. J.

41. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 42. DeLean, A., Munson, P. J., and Rodbard, D. (1978) Am. J .

43. Schramm, M., Thomas, G., Towart, R., and Franckowiak, G.

44. Ehlert, F. J. (1988) Mol. Pharmacol. 33, 187-194 45. Fleckenstein, A. (1977) Annu. Reu. Pharrnacol. Tor. 17, 149-166 46. Wang, G., Dugas, M., Armah, I. B., and Honerjaeger, P. (1990)

47. Glossmann, H., Ferry, D. R., Goll, A., Striessnig, J., and Zernig,

48. Johnson, J. D. (1984) Biophys. J . 45, 134-136 49. Schilling, W. P. (1988) Biochim. Biophys. Acta 943, 220-230 50. Strynadka, N. C. J., and James, M. N. G. (1989) Annu. Reu.

35-42

611-616

337

513-550

Pharmacol. 328, 69-75

48-57

(1951) J. Biol. Chem. 193, 265-275

Physiol. 4, E97-El02

(1983) Nature 303, 535-537

Mol. Pharmacol. 37, 17-24

G. (1985) Arzneim. Forsch. 35, 1917-1935

Biochem. 58,951-998

positive heterotropic allosteric regulators of dihydropyridine

Documents