Alternatively Spliced IS6 Segments of the α1C Gene Determine the Tissue-Specific Dihydropyridine Sensitivity of Cardiac and Vascular Smooth Muscle L-Type Ca2+ Channels
Abstract Dihydropyridines (DHPs) block the vascular smooth muscle L-type Ca2+ channel at lower concentrations than the cardiac Ca2+ channel, although their α1 subunit, which binds the DHPs, is derived from the same gene. This α1C gene gives rise to several splice variants, among which the α1C-b variant is affected by lower concentrations of nisoldipine than the α1C-a variant. Functional expression of chimeras of α1C-a and α1C-b subunits demonstrated that the transmembrane segment IS6 is responsible for the different dihydropyridine sensitivity. Northern blot analysis showed that transcripts coding for the IS6 segment of the α1C-a subunit were expressed in heart but not in aorta, whereas the IS6 segment of the α1C-b subunit was expressed predominantly in vascular smooth muscle. In situ hybridization of rat heart sections confirmed this expression pattern of IS6 α1C-a and IS6 α1C-b in ventricular and smooth muscle myocytes, respectively. These results suggest that the different dihydropyridine sensitivities of cardiac and vascular L-type Ca2+ channels are caused at least partially by the tissue-specific expression of alternatively spliced IS6 segments of the α1C gene.
High-voltage–activated Ca2+ channels are the pathways for voltage-gated Ca2+ entry in heart, smooth muscle, and neurons. These Ca2+ channels are multimeric protein complexes composed of up to four different proteins, namely, the α1, α2/δ, β, and γ subunits.1 The α1 subunit is the principal subunit and contains the voltage-gated channel pore. Six distinct genes have been identified encoding homologous α1 subunits. The “cardiac” (α1C-a)2 and the “smooth muscle” (α1C-b)3 proteins are splice products of the class-C Ca2+ channel α1 gene. These α1C subunits are the major targets for all known Ca2+ channel blockers, such as DHPs, phenylalkylamines, and benzothiazepines, which are powerful drugs in the treatment of a wide variety of cardiovascular disorders, including hypertension, angina, and some forms of cardiac arrhythmias. Key to most of these applications is the high degree of specificity of these drugs for the vascular smooth muscle L-type Ca2+ channel.4
DHPs are the most selective and potent blockers of L-type Ca2+ channels. Their affinity for channel block depends on the membrane potential5 6 ; ie, the extent of channel block increases with a shift of the membrane potential from −80 mV to more positive values. Myocardial and vascular smooth muscle cells express the same L-type Ca2+ channel, which is derived from the α1C gene. The resting potential of most smooth muscle cells is more positive than that of myocardial cells. Therefore, it was suggested that the higher selectivity of DHPs for the vascular smooth muscle channel than for the cardiac channel is caused by the lower resting membrane potential of smooth muscle cells.4 7 However, recent reports8 9 10 showed that slight differences in the primary sequence of the α1C subunit are sufficient to affect significantly the DHP block of expressed channels. These findings indicated that the well-documented high sensitivity of the smooth muscle Ca2+ channel for DHPs might be caused not only by the lower membrane potential of smooth muscle but also by structural differences between the cardiac and vascular L-type Ca2+ channel.
The α1C-a and α1C-b channel differ only at four sites constituting ≈5% of their amino acid sequence, namely, the amino terminus, the transmembrane segments IS6 and IVS3, and an insert in the cytoplasmic loop connecting repeat I and repeat II, which is only present in the α1C-b subunit. Interestingly, photoaffinity labeling studies indicated that the DHPs interact with extracellular sites of the channel close to IS6, IIIS6, and IVS6.11 12 Recent studies using site-directed mutagenesis of the α1C, α1A, and α1E subunit confirmed the interaction sites at the IIIS5/S6 and IVS6 segment of the α1C subunit.13 14 15 16 17 18 In order to define the sequence critical for the distinct DHP sensitivity, all possible chimeras between the α1C-a and α1C-b subunit were constructed and functionally expressed, and their pharmacological properties were studied.
The results demonstrate that the alternatively spliced IS6 segments of the α1C gene are responsible for the different DHP sensitivity of smooth and cardiac muscle L-type Ca2+ channels and that the IS6 α1C-a and IS6 α1C-b segments are selectively expressed in cardiac and vascular smooth muscle, respectively.
Materials and Methods
Construction of Chimeras
For construction of carboxy-terminal deletion mutants of the α1C-a subunit (termed pcDNA3HK1733) and α1C-b subunit (pcDNA3LK1728), plasmids pcDNA3HK1 and pcDNA3LK119 were truncated at amino acids 1733 and 1728, respectively, by a PCR protocol.20 For details of construction, see Seisenberger et al.9 To construct chimeras pcDNA3HK2, pcDNA3LK2, pcDNA3LK3, pcDNA3LK4, pcDNA3LK5, pcDNA3LK6, pcDNA3LK7, pcDNA3LK8, pcDNA3LK10, and pcDNA3LK12, the 2510-bp Afl II (nucleotide 2674) of pcDNA3LK1728–Xba I (polylinker site of pcDNA3LK1728) fragment and the following cDNA fragments from pcDNA3HK1733 were used: Afl II (nucleotide 2689)–Xba I (polylinker site of pcDNA3HK1733) fragment, Cla I (nucleotide 257)–Afl II (nucleotide 2689) fragment, Cla I (nucleotide 257)–Dra II (nucleotide 1386) fragment, Dra II (nucleotide 1386)–Nar I (nucleo-tide 1608) fragment, and Mun I (polylinker site of pcDNA3HK1)–Mun I (nucleotide 498) fragment. For the construction of pcDNA3LK7 and pcDNA3LK8, the EcoRI (nucleotide −11)–EcoRI (nucleotide 2199) fragment of pcDNA3LK1 was ligated into pBluescript II (Stratagene) to obtain pBlueLK1EcoRI. The Cla I–Dra II fragment was removed and replaced by the corresponding fragment of pcDNA3HK1, yielding pcDNA3LK7. To obtain pcDNA3LK8, the Dra II–Nar I fragment of pcDNA3HK1 was inserted into pBlueLK1EcoRI, and the resulting fragment was reinserted in pcDNA3LK1728.
Transfection of HEK 293 Cells
HEK 293 cells were transiently transfected with the expression vectors for the various chimeras (0.5 μg) by a lipofection method using lipofectamine according to the manufacturer (GIBCO-BRL Life Technologies). After transfection, the cells were grown for 16 to 18 hours, washed with PBS, and grown for another 24 hours before beginning the electrophysiological experiments.
IBa was measured at room temperature using the whole-cell patch-clamp technique. The external solution contained (mmol/L) NaCl 82, TEA-Cl 20, BaCl2 30, CsCl 5, MgCl2 1, EGTA 0.1, glucose 10, and HEPES 5, pH 7.4 (NaOH). The pipette solution contained (mmol/L) CsCl 102, TEA-Cl 10, EGTA 10, MgCl2 1, ATP 3, and HEPES 5, pH 7.4. The investigated cell was superfused constantly by the external solution. The fast capacitive component was compensated after forming the seal; the slow component and the series resistance were compensated after breaking into the whole-cell configuration. The leak current was compensated by subtracting the adequate multiple of a linear component obtained by small depolarizing voltage steps from a negative HP. Data were sampled at 5 kHz and filtered at 1 to 3 kHz. Current inhibition by nisoldipine was measured during trains of 40-ms test pulses applied once every 5 s as described previously.6 8 9 Rested-state channel block (protocol A) was measured by applying 40-ms test pulses once every 5 or 30 s from −80 mV to +20 or +30 mV. The developed block is termed frequency-independent block. In protocol B, channel block was measured after changing the HP from −80 to −40 mV (onset of inactivated-state channel block), and recovery from inhibition was studied by returning the HP to −80 mV (recovery from inactivated-state channel block).6 8 The test pulse voltage was either +20 or +30 mV, according to the peak of the current-voltage relation under these experimental conditions. Usually, inactivated- and rested-state channel blocks were each measured in different experiments with a single drug concentration. Nisoldipine was diluted to the final concentrations from a stock solution of 20 mmol/L in ethanol into the external solution and applied with a rapid solution exchanger. Data plotting and statistical analysis was carried out using ORIGIN software (Microcal). Pooled data are given as mean±SEM.
RT-PCR and Northern Blot Analysis
Total RNA from rat heart and aorta was isolated by the guanidinium thiocyanate method, and poly(A+) RNA was separated by oligo(dT) cellulose chromatography (Stratagene). Poly(A+) RNA was reverse-transcribed using oligo(dT) primers. PCR amplification for 40 cycles was done with a primer pair flanking the IS6 segment (primer 1 [nucleotides 1343 to 1362] and primer 2 [nucleotides 1477 to 1497]) of α1C-a.21 Reaction products were characterized by restriction mapping with several enzymes, including BamHI and Hae III, cloned in a pUC19-derived vector, and sequenced on both strands. For Northern blotting, 10 μg poly(A+) RNA from rat heart and aorta was electrophoresed on a 1.2% agarose gel, and equal RNA loading was confirmed by ethidium bromide staining of the gel. Separated RNA was transferred to Hybond-N nylon membranes (Amersham) and hybridized under high stringency with random-primed labeled cDNA probes. The probes were the 154-bp fragments corresponding to the type-a and type-b IS6 segments of the rat α1C sequence. The blot was exposed to BAS-III image plates (Fuji) and Hyperfilm MP (Amersham). Signals on the image plate were quantified with a BAS-1500 phosphorimager (Fuji). For rehybridizations, the blot was stripped in 0.1× SSC and 0.1% SDS at 95°C for 30 minutes, and complete removal of the probe was checked by autoradiography.
In Situ Hybridization
Adult Sprague-Dawley rats (250 to 350 g) were anesthetized with sodium pentobarbital (60 mg/kg IP). The heart was removed, frozen in isopentane, and cut at 12-μm thickness in a cryostat. Sections were fixed with 4% paraformaldehyde in PBS, pH 7.4. Slices were pretreated with proteinase K, acetylated with 0.25% acetic anhydride, prehybridized for 2 hours at 42°C in hybridization buffer (10 mmol/L Tris [pH 8.0], 0.3 mol/L NaCl, 1 mmol/L EDTA, 1× Denhardt’s solution, 10% dextran, 50% deionized formamide, and 50 mmol/L dithiothreitol), and hybridized with 35S-UTP–labeled cRNA probes (5×106 cpm/mL hybridization buffer) for 16 hours at 55°C. Slides were washed in 2× SSC, incubated in RNase A, and washed then in 0.1× SSC at 70°C. Sections were dipped in autoradiography emulsion Kodak NTB-2, exposed for 6 weeks, counterstained with hematoxylin/eosin, and examined with dark-field illumination. cRNA probes were in vitro–transcribed from the two IS6 segment variants cloned in a pUC 19–derived vector. The α1C common probe is directed against nucleotides 2745 to 3022 in the loop region between IIS6 and IIIS1 of the α1C-a subunit.21
Electrophysiological Characterization of the Ca2+ Channel Chimeras
A previous analysis of the two splice variants of the α1C gene, the cardiac α1C-a and the smooth muscle α1C-b protein, showed that low concentrations of nisoldipine induced a rested-state channel block only in the smooth muscle α1C-b channel; conversely, recovery from inactivated-state block was significantly greater in the α1C-a than the α1C-b channel.8 The native cardiac L-type Ca2+ channel had the same properties as the expressed α1C-a channel, suggesting that the different DHP sensitivity of the cardiac and smooth muscle L-type Ca2+ channel may reside in structural differences. The two splice variants of the α1C gene differ at the four regions, A, B, C, and D (Fig 1A⇓). To identify the region responsible for the different nisoldipine sensitivity, all possible chimeras of α1C-a (named HK1) and α1C-b (named LK1) subunits were constructed (termed LK2 to LK8, LK10, LK12, and HK2) (Fig 1B⇓) and functionally expressed in the absence of auxiliary subunits. All channels were carboxyterminal-truncated, since the truncated α1C subunit yields higher IBa without affecting the basic channel kinetics.9 20 22
Initial electrophysiological characterization of these chimeras indicated that indeed the truncated α1C chimeras induced regular L-type IBa. Neither the current-voltage relations nor the steady state inactivation curves differed significantly from those obtained previously with the full-length or truncated α1C-b or α1C-a subunit.8 9 22 IBa started at membrane potentials positive to −30 mV, was maximal between +20 and +30 mV, and reversed at +75 mV. Half-maximal inactivation under steady state conditions occurred between −8 and −9 mV. The maximal current densities of the truncated channels LK1 and HK1 were 18±1.2 pA/pF (n=22) and 8±0.7 pA/pF (n=20), respectively (Fig 2⇓). Comparison of the current densities obtained with the different chimeras showed that LK2 (19 pA/pF), LK4 (25 pA/pF), LK8 (13 pA/pF), and LK12 (19 pA/pF) had current densities similar to LK1 (Fig 2⇓). LK5 (5 pA/pF), LK6 (7 pA/pF), HK2 (8 pA/pF), and HK1 (8 pA/pF) had significantly smaller current densities, and LK3 (30 pA/pF), LK7 (39 pA/pF), and LK10 (19 pA/pF) had significantly higher current densities than LK1. All constructs with the “cardiac” amino terminus (region A) had low IBa densities, whereas intermediate current densities correlated with the “lung” amino terminus. The highest current densities were obtained with the combination of the lung amino terminus (region A) and the cardiac IS6 segment (region B) (Fig 2⇓). Obviously, not only the carboxy terminus20 22 but also the amino terminus of the α1C proteins influences the current density of the expressed α1C channels.
Nisoldipine Sensitivity of the Chimeras
The sensitivity of the chimeras against the nisoldipine block was tested by the two protocols, which were used and tested in previous studies,8 9 to distinguish between smooth muscle and cardiac type of block. In protocol A, the rested-state channel block (protocol designed to measure also frequency-independent block) was tested by superfusion of cells expressing genuine or chimeric channels with 1 or 10 nmol/L nisoldipine at the resting membrane potential of −80 mV and a stimulation frequency of 0.2 (not shown) or 0.033 Hz (see Figs 3⇓ and 4⇓). Nisoldipine reduced IBa of the genuine LK1 channel at a concentration of only 1 nmol/L and an HP of −80 mV, even at the very low stimulation frequency of 0.033 Hz (Fig 3⇓). The current was blocked to the same amount at 0.2 Hz (not shown). Such a rested-state channel block was not observed with the cardiac HK1 channel (Fig 4⇓). Further analysis showed that 1 nmol/L nisoldipine exerted a rested-state channel block in chimeras LK2, LK5, LK6, LK8, and LK12, whereas IBa of LK3, LK4, LK7, LK10, and HK2 was hardly or not blocked at all (Fig 5⇓). Close inspection of the chimeras revealed that a rested-state channel block occurred in parallel with the presence of the IS6 segment of the α1C-b sequence (segment B in Fig 1A⇑ ) but not with any of the other three alternatively spliced regions. In protocol B, the presence and extent of the inactivated-state channel block and the recovery from this block were tested. As already shown in previous studies,8 9 a shift of the HP from −80 to −40 mV decreased IBa by ≈20% in both channels in the absence of nisoldipine. In agreement with these previous results, 1 nmol/L nisoldipine reduced IBa to 55% and 67% for the cardiac versus the smooth muscle channel, respectively, and 10 nmol/L nisoldipine reduced IBa close to zero in both channels (Figs 3⇓ and 4⇓). The time course of the onset of block was faster in the LK1 than in the HK1 cells in the presence of 1 or 10 nmol/L nisoldipine (compare Figs 3⇓ and 4⇓ for 10 nmol/L drug), but the differences for both onset or extent of block at −40 mV in the presence of 1 or 10 nmol/L nisoldipine were not statistically significant. In agreement with these electrophysiological results, the DHP isradipine bound in radioligand studies to both channels with the same Kd value (≈0.1 nmol/L).23
However, the α1C-a and the α1C-b channels and the chimeras differed significantly in their recovery from the inactivated-state channel block. Therefore, this criterion was used in further experiments to distinguish the different types of inhibition. In the presence of 1 nmol/L nisoldipine, the reversal of the membrane potential from −40 to −80 mV unblocked almost 70% of the current in the HK1-expressing cells but only 47% in the LK1-expressing cells. The reversal of the membrane potential in the presence of 10 nmol/L nisoldipine revealed the clearest difference with an unblock of almost 50% in the HK1-expressing cells but only a minimal amount of current in the LK1-expressing cells (Fig 6⇓). After reversal of the membrane potential from −40 to −80 mV in the presence of 10 nmol/L nisoldipine, only 15% to 22% of IBa was recovered in cells expressing LK1 and the chimeras LK2, LK5, LK6, LK8, and LK12, whereas 41% to 46% of IBa was recovered in cells expressing HK1 and the chimeras LK4, LK7, and HK2. The extent of recovery of the chimeric channel LK10 was with 32% less than that of the HK1 channel, but this value was still significantly higher than that of the LK1 channel. Inspection of all chimeric channels indicated again that chimeras that contained the IS6 segment of the α1C-b subunit had a poor recovery from voltage-dependent block, whereas those that contained the IS6 segment of the α1C-a subunit had a good recovery. The only exception was chimera LK3. The current of this channel recovered by 27%, which is a higher recovery than that of the LK1 channel but lower than that of the HK1 channel. Together, these results suggested that all chimeras retained the property of a voltage-dependent DHP block. However, similar to the in vivo situation,8 the cardiac α1C-a channel recovered better from the voltage-dependent nisoldipine block than did the smooth muscle α1C-b channel. This difference was caused by the primary sequence of the IS6 segment.
Tissue-Specific Expression of the IS6 Segments
The above results indicated that the use of alternative IS6 segments significantly affected the DHP sensitivity of the α1C subunit. It was of great importance to see whether the two IS6 segments were expressed selectively in cardiac and smooth muscle. To address this question, the 154-bp fragment encoding transmembrane segment IS6 was RT-PCR–amplified from rat heart and aorta poly(A+) RNA (Fig 7A⇓). The sequence of the amplified fragment from rat heart was 96.8% and 100% identical at the nucleotide level to the corresponding sequences cloned from rabbit heart2 and rat brain,21 respectively, demonstrating that it encodes the type-a variant of the IS6 segment. In contrast, the fragment amplified from rat aorta showed 98.7% and 99.4% identity at the nucleotide level to the type-b variant of IS6 originally cloned from rabbit lung3 and mouse brain,24 respectively. The PCR products from each tissue were digested with the restriction enzymes BamHI and Hae III. BamHI cuts only the IS6 segment of the α1C-a sequence, and Hae III cuts only the IS6 segment of the α1C-b sequence. This analysis demonstrated almost tissue-specific representation of the IS6 segment in heart and aorta RNA (Fig 7A⇓, lanes 1 to 4). To further characterize the relative abundance of the two IS6 variants, the amplicons were subcloned and analyzed by restriction digestion. Eighteen of 20 cardiac PCR clones contained the IS6 segment of the α1C-a sequence, and only 2 contained the IS6 segment of the smooth muscle α1C-b sequence. In contrast, 17 of 20 aorta PCR clones had the sequence of the smooth muscle type-b IS6 segment. These results supported the hypothesis that both IS6 splice variants were expressed in a tissue-specific manner, with IS6 α1C-a being present in heart and IS6 α1C-b being present in smooth muscle cells.
To provide an independent confirmation of the expression level of the two IS6 splice variants, Northern blot analysis of rat heart and aorta poly(A+) RNA was performed (Fig 7B⇑). A 1:1 mixture of the two IS6 segment–specific probes detecting the entire pool of IS6 segment splice variants hybridized with transcripts of 8.5 and 14 kb in both tissues (Fig 7B⇑, blot a). Obviously, heart poly(A+) RNA contained a higher proportion of α1C mRNA than did aorta poly(A+). The blot was stripped and rehybridized with the probe recognizing the IS6 segment of the α1C-a subunit. Transcripts of 8.5 and 14 kb were labeled exclusively in cardiac poly(A+) RNA (Fig 7B⇑, blot b). In contrast, rehybridizing with a probe specific for the IS6 segment of α1C-b detected transcripts of 8.5 and 14 kb in aorta poly(A+)RNA and, to a smaller extent, also in cardiac poly(A+) RNA tissues (Fig 7B⇑, blot c). These signals were quantified using a phosphorimager. Based on the finding that nearly 100% of IS6 segments detected in aorta poly(A+) RNA were of the smooth muscle type (Fig 7B⇑, blot b), comparison of the relative band intensities of blots a and c showed that heart poly(A+) RNA contained 79% of the cardiac-type and 21% of the smooth muscle–type IS6 segments.
In Situ Hybridization
Finally, the tissue-specific expression of the two α1C splice variants was tested by in situ hybridization on rat heart sections. Hybridization of a frontal section through the heart ventricles with an α1C common probe detecting both type a and type b of the α1C gene showed strongly scattered labeling of ventricular myocytes and coronary arteries (Fig 7C⇑, a). As a control, labeling of an adjacent section with the corresponding sense riboprobe showed no signal (Fig 7C⇑, b). A cRNA probe directed against IS6 α1C-a strongly labeled ventricular myocytes but yielded no signal over different coronary arteries (Fig 7C⇑, c). In contrast, the probe specific for the IS6 segment of the α1C-b subunit strongly labeled the wall of the coronary arteries, whereas only weak labeling of ventricular myocytes was detected (Fig 7C⇑, d). Similar results as described above were seen with in situ hybridization of the three probes on sections through heart atria. Only IS6 α1C-b was detected in sections through the aorta (not shown).
The results of this and two previous studies8 9 showed that the nisoldipine sensitivity of the a and b splice variants of the α1C gene is different. This difference has been observed (1) with the full-length α1C-a and α1C-b subunits stably expressed in Chinese hamster ovary cells,8 (2) with the carboxyterminal-truncated α1C-a and α1C-b subunits stably expressed in HEK 293 cells,9 and (3) with the transiently expressed, carboxyterminal-truncated α1C-a and α1C-b subunits (the present study). The different nisoldipine sensitivity of the two splice variants is due to the presence of different IS6 segments within the α1C-a and α1C-b proteins. PCR amplification, Northern analysis, and in situ hybridization showed that the a and b splice variants of the IS6 segment are expressed in a tissue-specific manner. The IS6 segment of the α1C-a subunit is exclusively expressed in cardiac myocytes, whereas the IS6 segment of the α1C-b subunit is expressed in vascular smooth muscle and, to a small extent, also in cardiac myocytes. This tissue-specific expression of the isoforms of the IS6 segment clearly demonstrates that the preferential action of DHPs on the vascular L-type Ca2+ channel is caused at least partially by the structural properties of the α1C subunit present in vascular smooth muscle cells.
The major difference between the a1C-a and a1C-b channels was that (1) low concentrations of nisoldipine alone caused an frequency-independent block of the α1C-b channel, and (2) recovery from the inactivated-state channel block was observed readily with only the cardiac α1C-a, but not the smooth muscle α1C-b, channel. These findings suggested that nisoldipine interacted not only with the inactivated but also with the resting or closed state of the α1C-b channel. Moreover, the occurrence of a rested-state channel block indicates that the DHP block of the α1C-b channel has a voltage-independent component that is not apparent with the cardiac α1C-a channel. The association of the rested-state channel block with the presence of the IS6b segment that is part of the smooth muscle Ca2+ channel is in excellent agreement with the physiology of this tissue. A change in vascular smooth muscle tone is obtained without large changes in the membrane potential.7 The smooth muscle membrane potential remains relatively constant for long time periods. Under such conditions, a rested-state channel block will be more efficient than a block depending on strong depolarization of the membrane potential. In contrast to the vascular smooth muscle channel, the cardiac channel containing the IS6a segment showed no rested-state channel block and an excellent recovery from inactivated-state channel block. This property is in agreement with the physiology and pharmacology of the native cardiac L-type Ca2+ channel showing a frequency-dependent increase in the DHP block. The increase in frequency shifts the equilibrium between the different channel states to the inactivated state.
In general, the results of the present study confirm previous findings that the DHP sensitivity of both channels is different. As observed also by others,10 the difference between both channels is less apparent at depolarized membrane potentials but significant at more hyperpolarized membrane potentials. In agreement with these functional results, the DHP isradipine binds to both channels with the same Kd value (≈0.1 nmol/L).23 In these binding studies, the channel is believed to be in its inactivated state.5 6 The amino acid residues that are required for the high-affinity interaction with DHPs have been identified in the IIIS5, IIIS6, and IVS6 segments.13 14 15 16 17 18 Both channels have the same sequence in these regions, explaining the identity of the two binding constants and the similarity in the extent of the nisoldipine block at a HP of −40 mV.
The IS6 segment has been implicated in the process of voltage-dependent inactivation of the channel.25 26 Furthermore, it has been reported12 that amino acid residues close to IS6 of the α1S subunit interact with DHPs. These previous findings lend further support to the hypothesis that the IS6 segment is involved in a “low-affinity” interaction with nisoldipine and other DHPs that results in a rested-state block of the channel. The rested-state channel block may require interaction with critical amino acid residues of the IIIS5, IIIS6, and IVS6 region of the α1C subunit. This low-affinity interaction is modulated by the use of the alternative IS6 segments. The results of the present study suggest that the high sensitivity of the smooth muscle Ca2+ channel for DHPs is caused not only by the more depolarized potentials of vascular smooth muscle but additionally by the splice variant type b of the IS6 segment, which is specifically expressed in this tissue.
Selected Abbreviations and Acronyms
|HEK||=||human embryonic kidney|
|PCR||=||polymerase chain reaction|
This study was supported by SET (Stiftung zur Förderung der Erforschung von Ersatz- und Ergänzungsmethoden zur Einschränkung von Tierversuchen), Deutsche Forschungsgemeinschaft, and Fond der Chemie. We thank Barbara Lehnert and Sabine Stief for excellent technical assistance and Eva Roller for the graphical work.
© 1997 American Heart Association, Inc.
- Received April 14, 1997.
- Accepted July 22, 1997.
- © 1997 American Heart Association, Inc.
Triggle DJ. Calcium-channel drugs: structure-function relationships and selectivity of action. J Cardiovasc Pharmacol. 1991;18:S1-S6.
Bean BP. Nitrendipine block of cardiac calcium channels: high-affinity binding to the inactivated state. Proc Natl Acad Sci U S A. 1984;81:6388-6392.
Sanguinetti MC, Kass RS. Voltage-dependent block of calcium channel current in calf cardiac Purkinje fiber by dihydropyridine calcium channel antagonists. Circ Res. 1984;55:336-348.
Nelson MT, Laher I, Worley J. Membrane potential regulates dihydropyridine inhibition of single calcium channels and contraction of rabbit mesenteric artery. In: Vanhoutte PM, Paoletti R, Govoni S, eds. Calcium Antagonists: Pharmacology and Clinical Research. New York, NY: New York Academy of Science; 1988:47-50.
Welling A, Kwan YW, Bosse E, Flockerzi V, Hofmann F, Kass RS. Subunit-dependent modulation of recombinant L-type calcium channels: molecular basis for dihydropyridine tissue selectivity. Circ Res. 1993;73:974-980.
Soldatov NM, Bouron A, Reuter H. Different voltage-dependent inhibition by dihydropyridines of human Ca2+ channel splice variants. J Biol Chem. 1995;270:10540-10543.
Striessnig J, Murphy BJ, Catterall WA. Dihydropyridine receptor of L-type Ca2+ channels: identification of binding domains for [3H](+)-PN200-110 and [3H]azidopine within the α1 subunit. Proc Natl Acad Sci U S A. 1991;88:10769-10773.
Peterson BZ, Tanada TN, Catterall WA. Molecular determinants of high affinity dihydropyridine binding in L-type calcium channels. J Biol Chem. 1996;271:5293-5296.
Mitterdorfer J, Wang Z, Sinnegger MJ, Hering S, Striessnig J, Grabner M, Glossmann H. Two amino acid residues in the IIIS5 segment of L-type calcium channels differentially contribute to 1,4-dihydropyridine sensitivity. J Biol Chem. 1996;271:30330-30335.
He M, Bodi I, Mikala G, Schwartz A. Motif III S5 of L-type calcium channels is involved in the dihydropyridine binding site: a combined radioligand binding and electrophysiological study. J Biol Chem. 1997;272:2629-2633.
Ito H, Klugbauer N, Hofmann F. Transfer of the high-affinity dihydropyridine sensitivity from L-type to non-L-type calcium channel. Mol. Pharmacol. In press.
Wei X, Neely A, Lacerda AE, Olcese R, Stefani E, Perez-Reyes E, Birnbaumer L. Modification of Ca2+ channel activity by deletions at the carboxyl terminus of the cardiac α1 subunit. J Biol Chem. 1994;269:1635-1640.
Hofmann F, Klugbauer N, Lacinová L, Seisenberger C, Schuster A, Welling A. Interaction of the L-type calcium channel with calcium channel blockers. In: Endoh M, Iijima T, Morad M, Scholz H, eds. Molecular and Cellular Mechanisms of Cardiovascular Regulation. New York, NY: Springer Verlag; 1996:231-242.
Hofmann F, Biel M, Bosse E, Flockerzi V, Ruth P, Welling A. Functional expression of cardiac and smooth muscle calcium channels. In: Brown AM, Catterall WA, Kaczorowski GJ, Spooner PS, Strauss HC, eds. Ion Channels in the Cardiovascular Systems: Function and Dysfunction. Armonk, NY: Futura Publishing Co; 1994:369-381.
Ma WJ, Holz RW, Uhler MD. Expression of a cDNA for a neuronal calcium channel α1 subunit enhances secretion from adrenal chromaffin cells. J Biol Chem. 1992;267:22728-22732.