200,000+ products from a single source!
sales@angenechem.com
Home > Synthesis, Rotamer Orientation, and Calcium Channel Modulation Activities of Alkyl and 2-Phenethyl 1,4-Dihydro-2,6-dimethyl-3-nitro-4-(3- or 6-substituted-2-pyridyl)-5-pyridinecarboxylates
Nadeem Iqbal, Murthy R. Akula, Dean Vo, Wandikayi C. Matowe, Carol-Anne McEwen,
Michael W. Wolowyk,‡ and Edward E. Knaus*
Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2N8
A group of racemic alkyl and 2-phenethyl 1,4-dihydro-2,6-dimethyl-3-nitro-4-(3- or 6-substituted-2-pyridyl)-5-pyridinecarboxylates (13a-q) was prepared using a modified Hantzsch reaction that involved the condensation of a 3- or 6-substituted-2-pyridinecarboxaldehyde (7a-j) with an alkyl or 2-phenethyl 3-aminocrotonate (11a-d) and nitroacetone (12). Nuclear Overhauser (NOE) studies indicated there is a significant rotamer fraction in solution where the pyridyl nitrogen is oriented above the 1,4-dihydropyridine ring, irrespective of whether a substituent is located at the 3- or 6-position. A potential Hbonding interaction between the pyridyl nitrogen free electron pair and the suitably positioned 1,4-dihydropyridine NH moiety may stablize this rotamer orientation. In vitro calcium channel antagonist and agonist activities were determined using guinea pig ileum longitudinal smooth muscle (GPILSM) and guinea pig left atrium (GPLA) assays, respectively. Compounds having an i-Pr ester substituent acted as dual cardioselective calcium channel agonists (GPLA)/smooth muscle-selective calcium channel antagonists (GPILSM), except for the C-4 3-nitro-2-pyridyl compound which exhibited an antagonist effect on both GPLA and GPILSM. In contrast, the compounds with a phenethyl ester group, which exhibited antagonist activity (IC50 ) 10-5-10-7 M range) on GPILSM, were devoid of cardiac agonist activity on GPLA.
Structure-activity relationships showing the effect of a substituent (Me, CF3, Cl, NO2, Ph) at the 3- or 6-position of a C-4 2-pyridyl moiety and a variety of ester substituents (Me, Et, i-Pr, PhCH2CH2-) upon calcium channel modulation are described. Compounds possessing a 3- or 6-substituted-2-pyridyl moiety, in conjuction with an i-Pr ester substituent, are novel 1,4-dihydropyridine calcium channel modulators that offer a new drug design approach directed to the treatment of congestive heart failure and may also be useful as probes to study the structure-function relationships of calcium channels.
The structure-activity relationships for Hantzsch type 1,4-dihydropyridines, with respect to calcium channel antagonist-agonist modulation, have presented a significant challenge.1-14 The calcium ion channel is an important drug design target since it possesses specific drug binding sites for both antagonist and agonist ligands that are modulated by the closed or open conformational state of the channel. Different states of the channel have different affinities and/or access for drugs, and drugs may exhibit both quantitative and qualitative differences in structure-activity relationships, including stereoselectivity between channel states.14 Accordingly, ion channels can be viewed as multiple drug binding receptors that typically have 4-8 discrete binding sites which may be individually linked to each other and to the gating and permeation machinery of the ion channel by complex allosteric interactions.15 1,4-Dihydropyridines of the nifedipine class [dialkyl 1,4-dihydro-2,6-dimethyl-4-(substituted-aryl)-3,5-pyridinedicarboxylates] are flexible molecules (1a), in which the C-4 aryl moiety and the C-3/C-5 ester substituents can rotate and the conformation of the 1,4-dihydropyridine ring can change (Figure 1). The 1,4-dihydropyridine ring exists in a flat-boat conformation with the C-4 aryl moiety in a pseudoaxial position and orthogonal to the plane of the 1,4-dihydropyridine ring.
In addition, two rotamers may exist if the C-4 aryl ring is substituted at its ortho or meta position (X * H). The X-substituent either could then be on the same side as the C-4H as in 1b [synperiplanar (sp) to the C-4H or distal to the 1,4-dihydropyridine ring] or, following rotation of the phenyl ring, could be oriented above the 1,4-dihydropyridine ring as in 1c [antiperiplanar (ap) to the C-4H or proximal to the 1,4-dihydropyridine ring].16 Although the rotation barrier about the C(4)-C(phenyl) central C-C single bond is small, molecular mechanics calculations for Hantzsch 1,4-dihydropyridines show the rotational barrier separating the distal (sp) and proximal (ap) rotamers is higher for a C-4 phenyl ring having an ortho-substituent relative to analogues having a meta- or para-substituent.17 Ab initio STO-3G calculations similarly showed that an o-phenyl substituent also favors the sp rotamer orientation, although the energy difference or rotational barrier between the sp and ap rotamers is not sufficiently large to exclude the ap rotamer from participation in binding to the calcium channel receptor.18 The observation that the fraction of the sp rotamer in solution showed a positive correlation with vasorelaxant activity and receptor binding affinity suggests the sp rotamer of nonrigid unsymmetrically substituted 4-phenyl-1,4-dihydropyridine calcium antagonists is the receptor-bound conformation.
A normal vs capsized DHP boat model was recently proposed to explain structural and conformational requirements for modulation of calcium channel action where a left-hand-side alkoxy (cis) carbonyl interaction is required for maximal DHP receptor affinity, and the effect on channel action is determined by the orientation of the 4-aryl moiety. Thus enantiomers having an uporiented pseudoaxial aryl group (normal DHP boat) elicit antagonist activity, while enantiomers having a down-oriented pseudoaxial aryl group (capsized DHP boat) exhibit agonist activity.20 However, the issue of antagonism or agonism is dependent not only on the structure and stereochemistry of the 1,4-DHP but also upon the state of the calcium channel due to its membrane potential.15 This latter phenomenon is clearly
illustrated by the potent Ca2+ agonist (S)-Bay K 8644 (see Figure 2) which acts as an agonist at polarized membrane potentials and as an antagonist at depolarized membrane levels, respectively.21
Although the interaction of calcium antagonists with Ca2+ has received little attention, a recent study employing nicardipine showed the predominant species to be a 2:1 drug: Ca2+ sandwich complex that involved coordination of Ca2+ to oxygen of the m-nitrophenyl substituent and the carbonyl moiety of the C-3 ester substituent. Based on this result, it was suggested that the Ca2+-bound form
of DHP drugs may constitute their biological active species in the nonpolar milieu of a lipid bilayer.22
A novel third-generation class of isomeric 1,4-dihydro-2,6-dimethyl-3-nitro-4-(pyridyl)-5-pyridinecarboxylates (2a-c; see Figure 2) with different calcium channel modulation activities was recently discovered.23 The 2-pyridyl isomer (()-2a acted as a dual cardioselective calcium channel agonist/smooth muscle-selective calcium channel antagonist. On the other hand, the 3-pyridyl [(()-2b] and 4-pyridyl [(()-2c] isomers acted as calcium channel agonists on both cardiac and smooth muscle. The (+)-2-pyridyl enantiomer (+)-2a exhibited agonist activity on both cardiac and smooth muscle, whereas the (-)-2-pyridyl enantiomer (-)-2a exhibited cardiac agonist and smooth muscle antagonist actions. It was therefore of interest to extend these structure-activity data by replacing the isopropyl moiety of 2a by a 2-phenethyl substituent and/or introducing a 3- or 6-substituent (CH3, CF3, Cl, NO2, C6H5) into the 2-pyridyl ring since this novel type of 1,4-DHP calcium channel modulator could provide a potentially
new approach directed toward the treatment of congestive heart failure (CHF) and for use as probes to study the structure-function relationships of calcium channels. We now report the synthesis of alkyl and 2-phenethyl 1,4-dihydro-2,6-dimethyl-3-nitro-4-(3- or 6-substituted-2-pyridyl)-5-pyridinecarboxylates (13a-q), their rotameric orientation, and their in vitro calcium channel modulating actions on guinea pig smooth muscle and left atrium.
Chemistry
The 3- and 6-substituted-2-pyridinecarboxaldehydes 7a-e, required for the modified Hantzsch 1,4-dihydropyridine reaction, were prepared using a four-step synthetic sequence as illustrated in Scheme 1. Thus, oxidation of the 3- or 6-substituted-2-methylpyridines 3b-e with H2O2 in AcOH yielded the corresponding N-oxide derivatives 4a-d in 55-89% yields. Reaction of 4a-d with acetic anhydride afforded the respective 2-(acetoxymethyl)-3(or 6)-substituted-pyridines 5a-d in 75-81% yields. Hydrolysis of the acetate derivatives 5a-d with either 1 N NaOH or K2CO3/MeOH afforded the corresponding 2-(hydroxymethyl)-3(or 6)-substitutedpyridines 6a-d in 51-86% yields. In contrast, 3-chloro-2-(hydroxymethyl)pyridine (6e) was prepared by oxidation of 3-chloro-2-methylpyridine (3f) with KMnO4 to
yield 3-chloro-2-pyridinecarboxylic acid (5e; 45%) which on treatment with ClCO2Et to form the mixed anhydride and reduction with NaBH4 gave 6e (67%). Oxidation of the 2-(hydroxymethyl)-3(or 6)-substituted-pyridines 6a-e using either dicyclohexylcarbodiimide (DCC) or MnO2 yielded the respective 3(or 6)-substituted-2-pyridinecarboxaldehydes 7a-e (36-52%).
The Et3N-catalyzed reaction of 2-phenylethanol (9)with diketene (8) afforded 2-phenethyl acetoacetate (10; 73%) which was elaborated to 2-phenethyl 3-aminocrotonate (11d; 90%) on treatment with NH3 in MeOH (see Scheme 2). The racemic alkyl or 2-phenethyl 1,4-dihydro-2,6-dimethyl-3-nitro-4-(3- or 6-substituted-2-pyridyl)-5-pyridinecarboxylates 13a-q were prepared by a modified Hantzsch reaction.
Accordingly, condensation of the respective aldehyde (7a-j) with the respective alkyl or 2-phenethyl 3-aminocrotonate (11a-d) and nitroacetone (12) afforded the title compounds (13a-q) in 24-71% yields as illustrated in Scheme 3 and summarized in Table 1.
Results and Discussion
The development of calcium channel modulators that are useful for treating CHF will be dependent upon the separation and/or elimination of their vasoconstrictor effect from their cardiostimulant-positive inotropic property.24 Differences in the molecular electrostatic potentials between agonist and antagonist structures, with respect to binding of the C-3 and C-5 DHP regions, have been observed, which may provide a mechanism by which the receptor differentiates between agonist and antagonist ligands. In this regard, calcium channel agonists have been shown to possess a strong negative potential in the region of the C-3 nitro substituent, while antagonists showed a positive potential in this region.25
In addition, the effect of aromatic substituents on the C-4 phenyl ring of 1,4-dihydropyridine agonists and antagonists is also different.26 These observations, in conjunction with the dual cardioselective agonist/smooth muscle-selective antagonist calcium channel-modulating effects exhibited by isopropyl 1,4-dihydro-2,6-dimethyl-3-nitro-4-(2-pyridyl)-5-pyridinecarboxylate (2a),23 prompted us to study analogues of 2a where the isopropyl substituent is replaced by other alkyls or a 2-phenethyl substituent and a substituent is introduced at the 3- or 6-position of the C-4 2-pyridyl moiety.
The nature (electronic properties, steric size) and position (3 or 6) of the substituent on the 2-pyridyl ring system was expected to be a determinant of the electronic charge distribution at the pyridyl ring carbons, the global conformation of the molecule due to nonbonded interactions between the C-3, C-4, and C-5 substituents, and the rotameric orientation and/or preference (sp/ap) of the substituted-2-pyridyl moiety.
These factors may provide an approach to optimize calcium channel binding, calcium channel modulation, and/or tissue specificity. It has been suggested that molar refraction (MR) values are a crude, but useful, measure of substituent bulk (size) which have been used in some quantitative structure-activity relationship (QSAR) studies.27 A series of 1,4-dihydropyridines 13a-q were prepared wherein the 3- or 6-substituent on the C-4 2-pyridyl moiety possesses a broad range of MR values (H, 1.03; CF3, 5.02; CH3, 5.65; Cl, 6.03; NO2, 7.36; C6H5-, 25.36).27 The van der Waals radi for H, Cl, and CH3 are 1.2, 1.8, and 2.0 Å, respectively. Examination of the 1H NMR spectra for 13a-q indicated that the 1,4-DHP H-4 proton appeared in the δ 5.32-6.49 range (CDCl3). The H-4 proton was always more shielded (higher field)
when a C-6 pyridyl R1-substituent was present (δ 5.32-5.66 range) as compared to the same C-3 pyridyl R2-substituent (δ 5.74-6.49 range). Low-temperature 1H NMR studies (300 MHz) of 13a-q at -50 °C in CDCl3 did not give rise to any dual resonances. Although this absence of dual resonances indicates the possibility that a single rotamer and/or preference for the C-4 2-pyridyl moiety exists in solution, it is equally plausible that the rotation barrier is too small to stop rotation at -50 °C or that there is a thermodynamic preference for one rotamer regardless of the magnitude of the energy barrier to rotation.
However, Goldmann and Geiger28 reported that the two o-methyl resonances for dimethyl 1,4-dihydro-2,6-dimethyl-4-(2,4,6-trimethylphenyl)-3,5- pyridinedicarboxylate appear as a broad singlet in
CDCl3 at 25 °C, coalescence occurred at -18 °C (∆G(-18 °C) ) 51.1 kJ/mol), and two separate resonances were observed at -50 °C. 1H NMR difference nuclear Overhauser enhancement (NOE) studies were
performed for 13g and the three pairs of isomers 13a and 13d, 13h and 13i, and 13p and 13q to acquire information pertaining to the rotamer orientation of the C-4 2-pyridyl moiety. The percent NOE enhancements (DMSO-d6, 22 °C) clearly show that a significant rotamer fraction is present in which the pyridyl nitrogen atom is oriented above the 1,4-DHP ring in all these cases irrespective of whether the substituent (H, Me, Ph) is located at the C-3 or C-6 position of the 2-pyridyl moiety (see Figure 3).
Similar NOE enhancement studies (CDCl3, 22 °C) were performed for 13a, 13h (see Figure 4) since the H-bonding potential between the compound 13 and bulk solvent differs significantly between CDCl3 and DMSO. In contrast to the NOE studies using DMSO-d6 as solvent, compounds 13a,h showed NOE enhancements from the 1,4-DHP NH proton to the C-6 Me substituent on the pyridyl ring of 1.6% and 3.0%, respectively.
However, a NOE from the NH hydrogen to the pyridyl H-3 hydrogen of 13a or 13h was not observed (CDCl3, 22 °C), which suggests a rotamer in which the pyridyl H-3 hydrogen is oriented above the 1,4-DHP moiety and the pyridyl nitrogen atom is sp to the 1,4-DHP H-4 hydrogen either is not present or is a less predominant rotamer in solution. Rovnyak et al. 29 have elegantly determined the fraction of ap rotamer by measurement of NOEs from the NH hydrogen to an o-phenyl hydrogen (interatomic distances of 3.2-3.5 Å) for Hantzsch 1,4-DHPs. Variable temperature 1H NMR spectroscopy is a useful method to study H-bonding interactions. Lowering temperature may stop NH exchange (gives rise to a sharp, or coupled, NH resonance) which can enhance H-bonding that results in a deshielding (downfield shift) of the NH proton. Conversely, increasing temperature may disrupt H-bonding resulting in a more rapid rate of NH exchange (gives rise to a broader resonance) that results in an upfield shift (shielding
effect) for the NH proton due to disruption of Hbonding.30 Accordingly, it was also observed (1H NMR spectra) that the chemical shift of the NH proton in 13a,h in CDCl3 was highly temperature dependent. For example, the NH proton for 13a appeared at δ 10.28 (sharp singlet, 10 °C), 9.66 (sharp singlet, 22 °C), and 7.67 (broad singlet, 61 °C). Similarly, 13h showed NH resonances at δ 10.13 (sharp singlet, 10 °C), 9.37 (sharp singlet, 22 °C), and 8.11 (broad singlet, 50 °C). All other resonances for 13a or 13h showed minor changes in chemical shift positions of less than δ 0.09, irrespective of temperature. These NH chemical shift dependence data indicate that the NH group must be H-bonded (sharp NH, more deshielded) at 10 and 22 °C and that H-bonding is disrupted (broad NH, more shielded) upon heating to 61 °C (13a) or 50 °C (13h).
It has been reported, using a group of dimethyl 1,4-dihydro-2,6-dimethyl-4-(2-halogenophenyl)-3,5-pyridinedicarboxylates, that the fraction of sp rotamer (fs) increased with increasing van der Waals radius of the halogen substituent (F, 0.69; Cl, 0.84; Br, 0.92; I, 0.95) in solution.19 The calcium channel antagonist nifedipine29 and agonist Bay K 86446 both exist as sp rotamers. In contrast, the 4-(3-nitrophenyl) moiety present in the agonist Bay K 8643 has the unexpected ap rotamer orientation (X-ray structure), which places the m-nitro substituent on the phenyl ring directly above the 1,4-DHP ring (see structures in Figure 2).8
The restricted 2-(trifluoromethyl)phenyl ring sp orientation of Bay K 8644 in the solid state is primarily due to steric contacts between the CF3 fluorine atoms and the carbonyl ester and nitro oxygen atoms which would occur in the ap orientation.6 Accordingly, the ap rotamer for Bay K 8643 may be stablized by an electronic attraction between the nitro group on the phenyl ring and the 1,4-DHP moiety. In methyl 1,4-dihydro-2,6-dimethyl-3-nitro-4-(aryl)-5-pyridinecarboxylates such as Bay K 8644 and Bay K 8643, the structure is stablized by a H-bond (2.94 ( 0.04 Å) between the NH and the 3-nitro oxygen atom which is cis to the C(2)dC(3) double bond.8 The amine group of Bay K 8644 is significantly more acidic, due to electron delocalization, and is more capable of forming a stronger H-bond than the related antagonists having ester substituents at the 3- and 5-positions of the 1,4-DHP ring.6 It was therefore of interest to investigate the conformation of 13a, which exhibited the best calcium channel agonist activity on guinea pig left atrium, to gain further information pertaining to the H-bonding effect observed for the NH hydrogen in CDCl3 at 22 °C and the rotamer orientation of the pyridyl ring system. Weaver et al.31 have shown that AM1 semiempircal molecular orbital conformational analyses of 1,4-DHP calcium channel modulators are best suited to the modeling of DHP geometry.31
Some interatomic distances for the AM1-minimized structure of 13a are shown in Figure 5. The distance between the amine and pyridyl nitrogen atoms (3.47 Å) is shorter than that between the amine nitrogen and nitro oxygen atom that is cis to the C(2)dC(3) double bond (4.18 Å) or the carbonyl oxygen atom (4.49 Å). These data suggest there may be a possibility for the amine NH hydrogen atom to H-bond to the pyridyl nitrogen free electon pair. A H-bond of this type would position the NH hydrogen closer to the pyridyl nitrogen free electron pair where the donor NH distance is less than 3.2 Å and the angle made by covalent bonds to the donor and acceptor atoms is less than 120°. Although the preferred geometry of the amine nitrogen is trigonal with the NH hydrogen projected away from the DHP
ring, the formation of a H-bond between the NH and pyridyl nitrogen free electron pair may compensate for the decrease in energy resulting from a change in the NH orientation. Accordingly, a H-bond between the pyridyl nitrogen free electron pair and the 1,4-DHP NH, which would place the pyridyl nitrogen atom above the 1,4-DHP ring, would remove the requirement for a highenergy rotational barrier that would be necessary to exclusively, or preferentially, favor this rotameric orientation. It is also possible that an electronic attraction between the pyridyl ring and the 1,4-DHP moiety would
increase the fraction of this rotamer population. The interatomic distance between the amine nitrogen and pyridyl CH3 of 4.81 Å is consistent with the NOE effect (1.6%) shown in Figure 4. For the purpose of comparison, the distance between the amine nitrogen and H-3 on the pyridyl ring, when 13a is minimized (AM1) with the pyridyl nitrogen sp to H-4, was 3.26 Å. The observation that no NOE effect was observed from NH to the pyridyl H-3 in 13a also suggests that 13a exists preferentially as the ap rotamer shown in Figure 4. The nitro group of 13a in the AM1-minimized structure is in the plane of the C(2)dC(3) bond to which it is attached since the C(2)dC(3)-N-O (cis-oxygen) and C(2)dC(3)-N-O (trans-oxygen) torsion angles are -2.9° and 176.2°, respectively. AM1 geometry optimizations for compounds 13a,d,g-i,p,q (see ap rotamer orientation shown in Figure 3) were performed for both the preferred ap rotamer (pyridyl N anti to H-4) and the less favored sp rotamer (pyridyl N syn to H-4). In all cases, the preferential ap rotamer orientation showed a thermodynamic preference, viz.: 13a (6-Me), 3.00 kcal; 13d (3-Me), 7.47 kcal; 13g (3-H), 3.54 kcal; 13h (6-Me), 3.59 kcal; 13i (3-Me), 7.92 kcal; 13p (6-phenyl), 2.96
kcal; and 13q (3-phenyl), 6.37 kcal. The observation that the energy differences between the two rotamer orientations were greater for 13i (3-Me) and 13q (3-phenyl) is attributed to the larger steric interactions between the 1,4-dihydropyridine C-3, C-4, and C-5 substituents. Thermodynamic preferences g 3 kcal are considered to be significant with respect to rotamer orientation. The results of these studies do not indicate whether the NH is H-bonded to the cis-oxygen atom of the 3-nitro substituent or the pyridyl nitrogen free electron pair. An X-ray structure for 13a will be required to distinguish between these two potential alternatives.
The in vitro calcium channel antagonist- and agonistmodulating activities of racemic compounds 13a-q were determined using guinea pig ileum longitudinal smooth muscle (GPILSM) and guinea pig left atrium (GPLA), respectively. The calcium channel antagonist activities of 13a-q determined as the concentration required to produce 50% inhibition of GPILSM contractility are presented in Table 1. Compounds 13a-q exhibited weaker antagonist activity (IC50 ) 10-5-10-7 M range) than the reference drug nifedipine (IC50 ) 1.40 × 10-8 M). The R3-ester substituent was a determinant of antagonist activity for compounds possessing a C-4 6-methyl-2-pyridyl moiety where the potency order was Me (13c) > PhCH2CH2 (13h) > Et (13b) and i-Pr (13a).
A similar activity profile was observed for compounds possessing a C-4 6-chloro-2-pyridyl moiety [PhCH2CH2- (13l) > i-Pr (13e)]. In contrast, for compounds having a C-4 3-nitro-2-pyridyl moiety, the i-Pr ester (13f) was more active than the PhCH2CH2- ester (13o). In the isopropyl ester group of compounds, the nature of the R1-substituent at the 6-position [Me (13a) ≈ Cl (13e)], or the R2-substituent at the 3-position [Me (13d) ≈ NO2 (13f)], of the 2-pyridyl moiety was not a determinant of antagonist activity. In the phenethyl ester group of compounds 13g-q, incorporation of a substituent (Me, CF3, Cl, NO2, Ph) at either the 3- or 6-position of the C-4 2-pyridyl moiety decreased antagonist activity (IC50) 1.51 × 10-6 to >5.96 × 10-5 M range) relative to the unsubstituted 2-pyridyl compound 13g (IC50 ) 6.39 ×
10-7 M). In this latter series of compounds 13h-q, the nature of the R1-substituent at the 6-position of the 2-pyridyl ring (13h,j,l,n,p) had a small effect on antagonist activity (IC50 ) 1.41 × 10-6 to 9.98 × 10-6 M range). In contrast, the effect of a R2-substituent at the 3-position of the 2-pyridyl ring (13i,k,m,o,q) produced a larger effect on antagonist activity (IC50 ) 3.35 × 10-6 to >5.96 × 10-5 M range) where the relative potency order was Ph g Cl and Me > NO2 > CF3. The effect which the position of the substituent on the C-4 2-pyridyl moiety (6-R1 versus 3-R2) had upon antagonist activity was variable where R2 > R1 [Me (13d) > Me (13a); Me (13i) > Me (13h); Ph (13q > Ph (13p)], but in other cases R1 > R2 [CF3 (13j) > CF3 (13k); Cl (13l) > Cl (13m); NO2 (13n) > NO2 (13o)]. These isomeric differences in antagonist activities for the ortho-likeR2-substituent that is sp to the 1,4-DHP H-4 versus the meta-like-R1-substituent that is ap to the 1,4-DHP H-4 (see Figure 2) could be due to a number of possibilities
which include differences in the drug-receptor interaction or preferential affinity or access to the R1-subunit binding site of the L-type calcium channel.
© 2019 Angene International Limited. All rights Reserved.