K+ Currents in Human Coronary Artery Vascular Smooth Muscle Cells
Abstract K+ channels and their currents are important in vascular tone regulation and are potential therapeutic targets; however, K+ channels in human coronary artery vascular smooth muscle cells (VSMCs) have received little attention. We examined K+ currents in freshly isolated VSMCs from human coronary arteries (n=368 from 32 human hearts) with conventional patch-clamp or perforated-patch techniques with nystatin. We detected four different K+ currents: (1) the delayed rectifier K+ current, IK(dr); (2) the Ca2+-activated K+ current, IK(Ca); (3) the nonrectifying noninactivating outward ATP-dependent K+ current, IK(ATP); and (4) the spontaneous transient outward K+ current, IK(STOC). K+ channels underlying spontaneous transient outward currents probably represent a single clustered population of Ca2+-activated K+ channels functionally associated with Ca2+ release channels in the sarcoplasmic reticulum. Inwardly rectifying K+ currents were not observed. K+ currents were unevenly distributed in that they were not uniformly exhibited by all cells. The most prominent K+ currents were IK(Ca) (100%) and IK(dr) (46%). IK(STOC)s, which have not been previously described in humans, were present in 67% of VSMCs. IK(ATP) was small under physiological conditions; however, IK(ATP) increased markedly after cell stimulation with exogenous or endogenous coronary vasodilators. Thus, IK(ATP) may be particularly relevant in ischemia and could be of special importance as a therapeutic target. We conclude that human coronary VSMCs have unique K+ currents that differ sufficiently from those of other species, thus making the investigation of human material clinically relevant. The findings suggest potential avenues for further therapeutic research.
- K+ channels
- delayed rectifier current
- transient outward current
- K+ channel openers
- pituitary adenylate cyclase–activating peptide
Plasma membrane K+ channels are important in coronary blood flow regulation.1 Such a role was demonstrated by using K+ channel blockers to induce coronary vasospasm2 and K+ channel openers to relax arteries precontracted with serotonin.3 In coronary and other VSMCs, hyperpolarization after K+ channel activation causes voltage-dependent Ca2+ channels to deactivate, leading to relaxation.1 VSMC K+ channels are targets for vasoactive drugs and endogenous ligands.1 4 5 The coronary vasorelaxation of several vasodilators, including pinacidil, prostacyclin, and PACAPs, can be blocked by specific K+ channel inhibitors.3 6 7 Indirect evidence suggests that a glibenclamide-sensitive K+ channel is responsible for hypoxic dilatation of coronary arteries.8 Further, the basic defect in vasospastic angina pectoris may be impaired K+ conductance in coronary VSMCs.2
IK(Ca) and IK(dr) have been found in human mesenteric VSMCs but have not been described in human coronary arteries.9 Moreover, K+ channels are diverse, and their distribution varies in different vascular beds.10 In coronary arteries from other species, five different K+ currents have been described. A large-conductance IK(Ca) was found in rabbit, pig, and guinea pig.11 12 13 14 IK(dr) was identified in rabbit11 15 but not in canine14 coronary VSMCs. IK(STOC)s have been described in rabbit and guinea pig VSMCs but not in human VSMCs.12 16 Most recently, IK(ir)s have been described in VSMCs of intracardiac arterioles from guinea pigs.17 Finally, IK(ATP)s have been identified in porcine coronary VSMCs by several authors.18 19 IK(ATP)s are activated by synthetic vasodilators such as pinacidil3 4 20 and are targets for endogenous vasodilators.5 18 However, no electrophysiological information has been published from human coronary VSMCs.
Our aim was to identify and to characterize the basic electrophysiological properties and the outward K+ currents in freshly isolated single VSMCs from human coronary arteries. Whole-cell K+ currents were measured by conventional patch-clamp technique or by the perforated-patch method with nystatin. We provide evidence for four distinct K+ currents, namely, IK(Ca), IK(dr), IK(ATP), and IK(STOC). IK(STOC) and IK(ATP) have not been described in humans. IK(dr) in human coronary VSMCs is different from IK(dr) in human mesenteric arteries. Our results show that K+ channels are distributed heterogeneously, which may indicate heterogeneous cell populations within the coronary vessel wall.21 IK(Ca), IK(dr), and IK(STOC) are prominent in the majority of the cells. IK(ATP) is small under physiological conditions; however, IK(ATP) can be induced by cell stimulation with exogenous or endogenous K(ATP) channel agonists, such as pinacidil or PACAP-27. This finding may have unique clinical significance.
Materials and Methods
Coronary Artery Preparations
Human coronary arteries were obtained from the excised native hearts of 21 male and 10 female transplant patients with dilatative cardiomyopathy or coronary heart disease as well as from one male donor heart that could not be transplanted for technical reasons. The tissue was immediately placed in cold (8°C) Hanks’ solution (mmol/L: NaCl 119, KCl 4.7, KH2PO4 1.2, NaHCO3 25.0, MgSO4 1.2, glucose 11.1, EDTA 0.026, and CaCl2 2.5 at 5% CO2/95% O2) during transportation to the laboratory for further dissection. Left coronary arteries, right coronary arteries, or branches from left, right, or circumflex coronary arteries (diameter, ≈1.5 mm) were dissected and cleansed of adhering tissue and fat in the Hanks’ solution. All coronary arteries were stored in the Hanks’ solution at 4°C before use in electrophysiological measurements.
Isolation of VSMCs
The cells were isolated as previously described for rabbit basilar artery and porcine coronary artery with some modifications.3 22 The vessels were cut into small segments (≈4 to 8 mm in length) and placed in a Ca2+-free Hanks’ solution (mmol/L: NaCl 137, KCl 5.4, KH2PO4 0.44, NaH2PO4 0.42, MgCl2 2, CaCl2 0.14, EGTA 0.05, glucose 11.11, and HEPES 10, pH 7.4 with NaOH) for 2 to 4 minutes at room temperature (20°C to 24°C). After a longitudinal section was performed, the segments were washed twice in this solution. The segments were then placed in the Ca2+-free solution containing 3 mg/mL collagenase (type IA, Sigma Chemical Co), 10 mg/mL BSA (Sigma), and 1 mg/mL elastase (type IIA, Sigma) and were incubated for 30 to 50 minutes with gentle agitation at 36°C. After completion of the digestion, single cells were dispersed by gentle agitation in the Ca2+-free Hanks’ solution. The cells were stored in Hanks’ solution (Ca2+, 0.12 mmol/L) containing 1 mg/mL BSA or in a solution containing (mmol/L) NaCl 90, KH2PO4 1.2 , MgCl2 5, glucose 5, taurine 20, and HEPES 5 (pH 7.1 with NaOH) at 4°C. The isolation procedure produced high yields of relaxed coronary VSMCs ≈60 to 120 μm in length and 5 to 12 μm in diameter. The cells were studied between 2 and 12 hours after isolation.
K+ Current Recordings
Whole-cell K+ currents were measured according to the conventional patch-clamp method of Hamill et al23 (for details see References 24 and 25) or using the perforated-patch method with nystatin.26 Cells were held at −80 mV, and linear voltage-ramp pulses at 0.67 V/s from −100 to +100 mV or 300-millisecond (or 500-millisecond) depolarizing step pulses to different voltages were applied (stimulation frequency, 0.3 Hz). The external solution E1 contained (mmol/L) NaCl 140, CaCl2 1.8, MgCl2 1, KCl 5.4, CdCl2 0.1, glucose 10, and sodium HEPES 10 (pH 7.4). The patch pipette (resistance, 1 to 4 MΩ) was filled with solution I1 containing (mmol/L) potassium aspartate 80, KCl 30, NaCl 20, MgCl2 1, Mg-ATP 3, EGTA 10, and potassium HEPES 5 (pH 7.4). The Cs+-dialyzing pipette solution I2 contained (mmol/L) cesium aspartate 80, CsCl 40, TEA 10, MgCl2 1, Mg-ATP 3, EGTA 10, and cesium HEPES 5 (pH 7.4). Solutions were superfused through the chamber by gravity at a rate of 2 mL/min. Experiments were done at room temperature (20°C to 24°C). Nystatin (Sigma) was dissolved in dimethyl sulfoxide and diluted into the pipette solution to give a final concentration ranging from 50 to 100 μg/mL. Whole-cell access was achieved by nystatin within 10 to 20 minutes after seal formation. Whole-cell K+ currents were recorded using an Axopatch 200A or a List EPC-7 amplifier, filtered at 5 kHz using an eight-pole low-pass Bessel filter instrument (Frequency Devices), digitized at 10 kHz using a CED1401 interface (Cambridge Electronic Design Limited), and analyzed using CED Patch and Voltage Clamp Software Version 6.08. Series resistance and total cell capacitance were calculated from uncompensated capacitative transients, from 10-millisecond hyperpolarizing linear ramp pulses (10 mV), or by adjusting the Axopatch 200A amplifier series resistance and whole-cell capacitance controls to eliminate the resulting current transitions. Total cell capacitance of the cells was compensated only in some experiments. Series resistance in whole-cell recordings was <10 MΩ and was not corrected. Amplitude of currents in this configuration was always <1000 pA, resulting in a voltage error of <8 mV. Series resistance in perforated-patch recordings was <30 MΩ. Series resistance in perforated-patch recordings was corrected for when currents were >200 pA. The membrane input resistance of the cells was measured using small hyperpolarizing voltage pulses (10 mV, 10 milliseconds) from a holding potential of −80 mV. In this voltage range, the input resistance was large, and time-dependent outward currents were not activated. All data were corrected for a 10-mV liquid junction potential. Only acutely dispersed, spindle-shaped, relaxed cells were examined for K+ currents in the electrophysiological experiments. Their passive electrical properties and K+ channel currents are analyzed in the present study.
Pinacidil and iberiotoxin were obtained from RBI. DIDS, niflumic acid, glibenclamide, EGTA, 4-AP, HEPES, and A23187 were purchased from Sigma-Aldrich. PACAP-27 was from Peninsula. All salts were obtained from Merck. Stock (10 mmol/L) solutions of pinacidil, A23187, and glibenclamide were made using dimethyl sulfoxide as the solvent. BAPTA-AM (30 μmol/L), a Ca2+ chelator that diffuses through membranes (Calbiochem), was used to chelate intracellular Ca2+.
All values are given as mean±SEM. The term n represents the number of cells tested. The Wilcoxon rank sum test or the Mann-Whitney-Wilcoxon test was used to determine significant differences. A value of P<.05 was considered statistically significant.
Cell Appearance and Passive Membrane Properties
The enzyme treatment released ≈70% single, spindle-shaped, relaxed coronary VSMCs from the hearts of the patients with cardiomyopathy and from the normal donor heart. The isolation method released only 20% to 30% spindle-shaped relaxed cells from the atherosclerotic plaque-filled coronary arterial walls of the two hearts from patients with coronary heart disease. In 52 cells, the length ranged from 65 to 180 μm, with a mean of 120 μm. Almost all spindle-shaped cells remained relaxed even after 15- to 20-minute superfusion with a physiological salt solution containing 1.8 mmol/L CaCl2 (E1 without CdCl2), which indicated that they were Ca2+ tolerant. Serotonin and external K+ (50 mmol/L) cause direct constriction of human coronary arterial rings.3 Usually a high proportion of the cells contracted with 5 μmol/L serotonin or 50 mmol/L K+ by ≈40% in 20 seconds (using external solution E1 without Cd2+) and formed many blebs after contraction. Recovery from serotonin or K+ was generally slow (in the order of minutes) and not usually complete. However, it was possible to obtain multiple contractions to serotonin or K+ in the same cell. Our observations suggest that the isolated cells were viable and had the necessary receptors, membrane conductances, and intracellular contractile machinery necessary to initiate contraction. Only spindle-shaped, relaxed, freshly isolated cells were used in the following electrophysiological experiments.
The resting membrane potential of 21 cells from the hearts with cardiomyopathy ranged from −35 to −70 mV, with a mean of −53 mV (measured in E1 without Cd2+). Similar resting membrane potentials were measured in cells isolated from the nonfailing donor heart. In these cells, the resting membrane potential was −56 mV (n=8). These values are in the range of resting membrane potentials reported for other types of isolated VSMCs, including nonhuman coronary arteries.1 Using small hyperpolarizing pulses from a holding potential of −80 mV, we calculated the membrane input resistance and the total cell capacitance. Mean values for these parameters were 5.1±0.2 GΩ (n=178) and 32.1±1.2 pF (n=131), respectively, from the hearts with dilatative cardiomyopathy and 4.8±0.4 GΩ (n=26) and 29.3±1.7 pF (n=18), respectively, in cells from the nonfailing donor heart. Assuming a specific capacitance of 1 μF/cm2, the membrane surface area of a single cell was ≈3×10−5 cm2 in both cases. These data suggest that coronary VSMCs isolated from patients with cardiomyopathy are similar to cells isolated from the nonfailing donor heart, since resting membrane potential, input resistance, and calculated membrane surface area of the single cells were not different.
Voltage-dependent K+ currents were studied using 300- or 500-millisecond voltage steps to potentials ranging from −70 to +100 mV from a holding potential of −80 mV or using voltage ramps from −100 to +100 mV. The external solution contained 100 or 200 μmol/L CdCl2 to block interfering voltage-dependent inward Ca2+ currents. Recordings from a total of 368 single acutely dispersed VSMCs were examined for K+ currents. Currents were recorded using the standard whole-cell configuration, dialyzing the cells with the internal pipette solution (130 mmol/L K+-rich solution with 3 mmol/L ATP and 10 mmol/L EGTA), or using the perforated-patch configuration with nystatin. The latter configuration has the advantage that ionic currents can be examined without the necessity of internal dialysis of the cells, ie, without changing the intracellular milieu.
Voltage-Dependent K+ Outward Currents (K(dr) and K(Ca) Channels)
Fig 1A⇓ illustrates examples of membrane currents elicited during voltage steps or ramp depolarizations in a single coronary VSMC. Step depolarizations elicited outward currents at membrane potentials positive to −40 mV. The current increased as the test potential became more positive. Close examination of the current tracings revealed a sigmoid onset of the outward current with a correspondingly larger and faster onset with depolarization. The corresponding I-V relationship (I-V curve) of the net outward currents is plotted in Fig 1B⇓.
To verify that K+ was the major charge carrier of the net outward current, K+ was replaced by Cs+ and TEA in the whole-cell patch pipette solution. Under these conditions, no outward currents were observed, suggesting that outward currents were carried by K+ ions (n=4). The Cl− channel antagonists niflumic acid (20 μmol/L, n=5, 4±7% current inhibition at +40 mV) and DIDS (100 μmol/L, n=2, 3% and 5% inhibition at +40 mV) had no effect on the outward currents, suggesting that the contribution of Cl− channels was minimal in our recording conditions. Tail currents recorded at various repolarizing potentials after a prepulse step to +80 mV revealed an average tail current EK of −78±8 mV (n=4). This value is close to the calculated Nernstian EK (−83 mV). Our experiments indicate that K+ was the major charge carrier.
Two components of the outward current were distinguished by using both voltage-step and -ramp protocols. The first component, IK(dr), was activated near −40 mV and was small in amplitude, and the current noise was minimal. The second component, IK(Ca), was activated positive to −20 mV, was large in amplitude, and was extremely noisy. Both current components displayed no or little inactivation during 300- to 500-ms voltage steps (Fig 1A⇑). The I-V curve of the outward current was N-shaped and showed two maxima of slope conductances at +20 and ≥100 mV. This appearance may reflect the fact that the first small low-noise component of the outward current, IK(dr), is maximally activated (Fig 1A⇑ and 1C⇑). Further evidence that the two macroscopic currents, IK(dr) and IK(Ca), were carried by two distinct K+ channels was obtained from experiments that examined their sensitivity to the K+ channel blockers and intracellular Ca2+ concentration.
Fig 2⇓ shows the dose-dependent effects of TEA on the currents. TEA (0.1 to 5 mmol/L) decreased the amount of the IK(Ca) elicited during strong depolarizations with voltage steps or ramps (Fig 2A⇓ and 2B⇓) and reduced the current noise at these positive potentials. In contrast, another fraction of the currents elicited by depolarizations to potentials ≤+20 mV was relatively insensitive to TEA over this concentration range. However, higher doses of TEA (5 mmol/L) also inhibited IK(dr) at these more negative potentials. To quantify the differential sensitivities of IK(Ca) and IK(dr) to TEA, the inhibition of outward current was compared at depolarizations to +80 and +20 mV. Dose-response curves for the inhibition of the two components at +20 and +80 mV are illustrated in Fig 2C⇓. The percent current in the presence of TEA as fraction of its control value (I/Imax) at +80 mV is given as follows:
where [TEA] is the dose of TEA, and Ki1 is the apparent dissociation constant. The data points were well fitted by this equation, consistent with a 1:1 binding of TEA to a receptor with a Ki1 of 255 μmol/L at +80 mV.
The dose-response curve measured for outward currents at +20 mV was fitted by the following equation:
where A and B are constants determining the current fractions inhibited by TEA with Ki1 and Ki2, respectively; Ki1 is apparent dissociation constant obtained from Equation 1; and Ki2 is a second dissociation constant. The data points were well fitted by this equation, consistent with a voltage-independent 1:1 binding of TEA to the same receptor as described for TEA binding on large-conductance K(Ca) channels in arterial VSMCs27 and an additional 1:1 binding to a second receptor with a Ki2 of 21.2 mmol/L at +20 mV. Functionally, the equation indicates that TEA inhibits two different K+ channels or interacts with two binding sites with different affinities on a single-channel protein.
In other experiments, blockade of K+ outward currents by 4-AP was examined. At all concentrations tested, 4-AP appeared to preferentially inhibit the low-noise current, IK(dr), which was activated at negative potentials (Fig 3⇓). The current noise of IK(Ca) at positive potentials was little affected by 4-AP; however, a small decrease in current at +80 mV was observed. This decrease in total outward current was attributed to blockade of IK(dr), which may also contribute current at positive potentials. Dose-response curves for the inhibition of IK(dr) by 4-AP at +20 mV is shown in Fig 3C⇓. The points were well fitted by the following equation:
consistent with a 1:1 binding of 4-AP to a receptor with a Ki of 1.02 mmol/L at +20 mV. At positive potentials (+80 mV), the highest dose of 4-AP (5 to 10 mmol/L) produced only a 20% inhibition of the total current. These data suggest that one population of K+ channels, namely, K(dr), is blocked by 4-AP. The combination of 4-AP (5 mmol/L) and TEA (5 mmol/L) almost completely blocked the outward K+ currents at +20 and +80 mV (n=6).
Iberiotoxin is a selective high-affinity blocker of large-conductance K(Ca) channels. We used this compound to test the contribution of these channels to the macroscopic currents. Fig 4A⇓ shows an experiment with iberiotoxin. Iberiotoxin (100 and 300 nmol/L) administration was similar to low concentrations of TEA, in that the toxin selectively inhibited the large noisy component of the K+ outward current, IK(Ca) (100 nmol/L, n=12; 300 nmol/L, n=23). No difference was detected between the effect of 100 and 300 nmol/L iberiotoxin on the total outward K+ current; 300 nmol/L iberiotoxin did not further reduce the K+ current that had been reduced by 100 nmol/L iberiotoxin (n=12). Both concentrations of iberiotoxin had little effect on the smaller, less noisy component of the outward IK(dr).
Studies examining the sensitivity of the two components to the intracellular Ca2+ concentration were performed using the perforated-patch method with nystatin. Addition of the Ca2+ ionophore A23187 (10 μmol/L, n=12), which increases membrane permeability for Ca2+, induced an enhancement of the large noisy component of the outward current (Fig 4B⇑). This enhancement was voltage dependent and was blocked by iberiotoxin (300 nmol/L, n=3; Fig 4B⇑) or by low doses of TEA (0.5 to 1 mmol/L). In contrast, addition of the membrane-permeable Ca2+ chelator BAPTA-AM (30 μmol/L, n=15) reduced the large noisy component of the outward current from 360±45 to 78±27 pA at +80 mV. During both the voltage-step and -ramp protocols, there was far less noise than was present under the control condition. The less noisy component of outward current that was activated at more negative potentials was relatively insensitive to BAPTA-AM. At +20 mV, BAPTA-AM had no effect on the outward current (Fig 4C⇑). These data suggest that the large noisy component of outward current, IK(Ca), which was activated at more positive potentials, was Ca2+ sensitive. The second, small, low-noise component of outward current, IK(dr), was relatively insensitive to Ca2+. Since IK(Ca) is a large, noisy, voltage-dependent, Ca2+-sensitive current that is blocked by iberiotoxin or low doses of TEA, it most likely can be attributed to the opening of large-conductance K(Ca) channels. The small, low-noise component of current was also voltage dependent, displayed Ca2+ insensitivity, and was blocked by 4-AP. Because this current resembles delayed rectifier K+ currents in portal vein,28 cerebral artery,29 and renal artery cells,30 we referred to this current as IK(dr).
Inactivation kinetics of IK(dr) were studied in 10 single cells by a double-pulse protocol and in the presence of iberiotoxin (100 nmol/L) and Cd2+ (200 μmol/L) (to inhibit IK(Ca)) (Fig 4D⇑). The degree of inactivation was assessed by examining the peak outward current at test potentials of +20 mV (circles in Fig 4D⇑) after holding the membrane (preconditioning) potential at voltages between −80 and +80 mV for 10 seconds. The peak outward current should be proportional to the degree of inactivation that occurred during the preconditioning potential. Membrane depolarization increased inactivation, plotted in Fig 4A⇑ as a decrease in the availability of the current for activation. V0.5 was −26 mV and increased as much as e-fold per 12.1 mV (steepness factor k) depolarization, Imax was 73.1 pA, and a noninactivating component was 19.0 pA (26% of Imax). Similar results were obtained using test potentials of +80 mV.
Both IK(Ca) and IK(dr) were revealed by N-shaped I-V curves and were found in 136 (46%) of 293 cells isolated from coronary arteries of 31 patients with dilatative cardiomyopathy and in 18 (64%) of 28 cells isolated from coronary arteries of the nonfailing donor heart (see Fig 1⇑). In 155 of 321 cells (145 of 293 cells of patients with dilatative cardiomyopathy and 10 of 28 cells isolated from coronary arteries of the nonfailing donor heart), I-V curves were S-shaped, as predicted for an outward current carried by one current component (Fig 5⇓). The current was noisy, activated at potentials positive to +10 mV, showed one maximum of slope conductances at +90 mV (Fig 5A⇓ through 5D), and was blocked by low concentrations of TEA (Ki, 150 μmol/L) (Fig 6A⇓ and 6C⇓). The current was almost completely blocked by iberiotoxin (100 nmol/L, n=6; 300 nmol/L, n=12) (Fig 6E⇓), increased by A23187 (30 μmol/L, n=9) (Fig 6F⇓), but was not affected by 5 mmol/L 4-AP (Fig 6B⇓ and 6D⇓). These data indicate that cells with S-shaped I-V curves of outward current expressed only IK(Ca).
An analysis of I80 (outward current with dominant IK(Ca)) on the membrane capacitance revealed a positive correlation between I80, of cells with N-shaped I-V curves, and membrane capacitance (r=.79) (Fig 7A⇓). The correlation coefficient of .79 indicates that K(Ca) channels are expressed at a similar density in small and large cells with N-shaped I-V curves. I80 correlated with the membrane capacitance of cells with S-shaped I-V curves (r=.75), indicating that K(Ca) channels in these cells are also expressed at a similar density (Fig 7B⇓).
I80 of cells with N-shaped I-V curves was 277±14 pA (n=142), corresponding to a current density of 6.2 pA/pF (d1, estimated by linear regression; see Fig 7A⇑). I80 of cells with S-shaped I-V curves was 255±11 pA (n=141), corresponding to a current density of 4.8 pA/pF (d2, estimated by linear regression; see Fig 7B⇑). At +20 mV, I20 (outward current with significant IK(dr)) of cells with N-shaped I-V curves was 120±8 pA (n=142), corresponding to a current density of 1.9 pA/pF (d3, estimated by linear regression; see Fig 7A⇑). I20 of cells with S-shaped I-V curves was 32±3 pA (n=141), which is close to the leakage current (estimated by the calculated input resistance). Since the sum of d2 and d3 is close to d1 (plus the leakage current density), we suggest that K(Ca) channels are expressed in a similar density in both cell types (eg, in cells with N-shaped and S-shaped I-V curves), whereas K(dr) channels are expressed in a similar density in only one population (≈50%) of cells (with N-shaped I-V curves).
Pinacidil is a K(ATP) channel opener.4 20 To provide evidence for IK(ATP) in human coronary VSMCs, we examined the effect of this drug on K+ outward currents. In 16 cells, pinacidil (1 to 20 μmol/L) induced a large nonrectifying current. At +40 mV, 1 and 20 μmol/L pinacidil increased the outward current from 160±28 to 295±40 pA (n=9) and from 148±47 to 309±52 pA (n=4), respectively. As shown in Fig 8⇓, EK was shifted to −72±8 mV (n=16). This value is close to the calculated EK, suggesting that the pinacidil-activated current was carried by the movement of K+ through K+ channels. The pinacidil-induced current was linear over most of the voltage range and showed no threshold of activation, as expected if the channels show little voltage dependence. The pinacidil-induced current occurred within seconds, showed no inactivation, and was reversible after the drug was removed from the bath (not shown).
Glibenclamide (3 μmol/L) blocked the pinacidil-induced (1 μmol/L) current from 329±20 to 158±27 pA at +40 mV (Fig 8B⇑, n=10). The glibenclamide-sensitive current induced by pinacidil was linear over the voltage range, was not inactivating during 300- to 500-millisecond step pulses, and showed no inactivation using double-pulse protocols as described in Fig 8⇑ (data not shown). Glibenclamide had no effect on the outward curves of IK(Ca) and IK(dr) (at +40 mV, 210±25 and 218±21 pA before and after glibenclamide, respectively; n=6) (Fig 8A⇑). The voltage independence and glibenclamide sensitivity of the pinacidil-induced current suggest an activation of a current that is conducted by K(ATP) channels.
Further evidence that K(ATP) channels are present in human coronary VSMCs was obtained from experiments that examined the effect of a putative humoral activator of IK(ATP). We used PACAP-27, which relaxes human coronary arteries in a dose-dependent and glibenclamide-sensitive manner.7 An experiment with PACAP-27 is shown in Fig 8C⇑. Similar to pinacidil, PACAP-27 (100 nmol/L) induced a large nonrectifying current. In the presence of PACAP-27, the outward current was increased from 178±26 to 380±31 pA at +40 mV (n=8). EK was shifted to more negative potentials, suggesting that the current activated by PACAP-27 is carried by K+ ions. The effects of PACAP-27 were reversible upon washout of the hormone. Glibenclamide (3 μmol/L) blocked the PACAP-27–induced (100 nmol/L) current from 368±34 to 239±24 pA at +40 mV (Fig 8C⇑, n=5). The glibenclamide-sensitive current induced by PACAP-27 was almost linear over the voltage range, was not inactivating during 300- to 500-millisecond step pulses, and showed no inactivation using double-pulse protocols as described in Fig 4D⇑ (data not shown). These data indicate that IK(ATP) is small in physiological conditions but can be stimulated after application of exogenous or endogenous K(ATP) channel agonists, namely, pinacidil or PACAP-27, respectively. However, in addition to enhancing IK(ATP), pinacidil and PACAP-27 may also stimulate a glibenclamide-insensitive K+ channel current (IK(Ca) and/or IK(dr)).31 32
IK(to) (K(to) Channels)
In 4% of the cells studied (n=12/321), a voltage-dependent outward current was observed. This current was activated by depolarization and had many of the characteristics of the fast transient K+ current (“A”-type current) observed in neurons. Currents were activated at potentials positive to 0 mV. The currents were transient and were inactivated to steady state levels over the duration of the pulse (within 100 to 200 milliseconds). The decay of inactivation was adequately fitted by a single exponential using the following equation:
where A and B are constants, t is the time, and tau is the time constant of decay. Using this equation, at a membrane potential of +50 mV, tau was 65±11 milliseconds (n=6) and decreased with more positive pulse potentials. The cells with IK(to) were spindle-shaped, as were all cells used in this study. The cells had an input resistance of 5.2±1.8 GΩ, a capacitance of 29.7±8.4 pF, and a calculated membrane area of 3×10−5 cm2. Because of the low number of cells exhibiting transient outward currents, we did not perform further pharmacological characterization of IK(to).
IK(STOC)s were observed in ≈70% of coronary VSMCs studied with the perforated-patch technique. Cells from coronary arteries of patients with dilatative cardiomyopathy showed IK(STOC)s in 67%, whereas those from nonfailing donor hearts demonstrated IK(STOC)s in 70%. Membrane current recordings from a voltage-clamped cell at different voltages are shown in Fig 9⇓. At −40 mV, the IK(STOC)s had a duration of ≈100 milliseconds. The current increased to a peak of 55 pA in the largest transients in <30 milliseconds and then decayed more slowly to the basal current level. The amplitude of the largest IK(STOC)s decreased with hyperpolarization (16±7 pA at −60 mV, n=4; 48±16 pA at −30 mV, n=6; and 112±34 at 0 mV, n=5), as did the frequency of the IK(STOC) discharges (0.37±0.18 Hz at −60 mV, 1.4±0.6 Hz at −30 mV, and 2.3±0.8 at 0 mV; time of recording, 3 to 6 minutes). EK of IK(STOC)s was seen between −70 and −60 mV (interpolated), close to the calculated EK. Thus, the generated currents appeared to be carried by K+ ions.
The IK(STOC)s were blocked completely by 300 nmol/L iberiotoxin (n=12, Fig 9B⇑). IK(STOC)s were never observed when the cells were held under voltage clamp using standard patch-clamp recording in the whole-cell configuration with 10 mmol/L EGTA in the pipette solution, thereby increasing intracellular Ca2+ buffering (n=50). These data suggest that an elevated concentration of subsarcolemmal Ca2+ was required to stimulate large-conductance K(Ca) channels generating IK(STOC)s in human coronary VSMCs.
Step depolarizations elicited similar outward currents (IK(dr) and IK(Ca)) when the holding potential was −40 mV instead of −80 mV (see Fig 1⇑). In contrast, step hyperpolarizations from a holding potential of −40 mV to potentials negative to EK did not elicit inward currents (with exception of the leakage current) (n=21). Ba2+ (0.5 mmol/L) had no effect on currents elicited at potentials negative to −40 mV (n=3). Thus, the corresponding I-V relationship (I-V curve) of the net currents showed strong outward rectification, which suggests that IK(ir)s are not detectable in coronary artery VSMCs.
The present study is the first to characterize and identify K+ currents in freshly isolated single human coronary VSMCs. Despite similar microscopic appearance and passive electrical properties, the K+ channels displayed functional heterogeneity. The most prominent K+ currents in human coronary VSMCs were IK(Ca) (100%), IK(dr) (46%), and IK(STOC) (67%). IK(STOC)s have not been described in humans; these currents may represent K+ currents through a single clustered population of maximal K(Ca) channels that have been activated by local and transient Ca2+ release from the sarcoplasmic reticulum.33 34 IK(ir)s were not observed. IK(ATP) was small under physiological conditions but increased markedly when the cell was activated by stimulation with exogenous or endogenous coronary vasodilators. This observation suggests that the K(ATP) channel is closed under most circumstances. However, under pathological conditions, such as ischemia, this channel may be a useful therapeutic target. Although the majority of our data show that the basic properties of K+ channel types in human coronary VSMCs are similar to those measured in a vast number of other vascular preparations, our results demonstrate that K+ currents in coronary arteries have features distinct from other vascular beds in humans9 and coronary VSMCs of other species.11 13 14 35
IK(dr) and IK(Ca)
We found that the outward K+ current that activated upon depolarization could be divided into two components in our cells. One component was carried by large-conductance K(Ca) channels, and the other was carried by channels whose characteristics resemble those of delayed rectifier channels in other types of smooth muscle cells. We termed these components IK(Ca) and IK(dr), respectively. The current activated at potentials negative to +20 mV was mainly IK(dr) and was blocked by 4-AP and possibly also by high concentrations of TEA. The current activated at positive potentials consisted of both currents and was thus only partially blocked by 4-AP. IK(Ca) was blocked by low concentrations of TEA, as well as by iberiotoxin (100 and 300 nmol/L), and was sensitive to the internal Ca2+ concentration. Measurements of K+ channel unitary current amplitudes underlying TEA- and iberiotoxin-sensitive IK(Ca) directly from low-noise whole-cell current recordings revealed a large unitary conductance of ≈130 pS between −30 and +10 mV (M. Gollasch and R. Bychkov, unpublished data, 1995), as reported for large-conductance maxi Ca2+-activated K+ channels. The combination of 4-AP (5 mmol/L) and TEA (5 mmol/L) almost completely blocked the outward current.
Large-conductance K(Ca) channels have been described in almost all VSMC types studied so far, including coronary arteries of guinea pigs,12 dogs,14 and rabbits.11 13 Single-channel experiments revealed that there are at least three K(Ca) channel subtypes (KL, KS, and KM) in VSMCs.36 In human coronary arteries, the Ki value from inhibiting IK(Ca) with TEA (Ki at +80 mV, 150 μmol/L) was 6.3-fold lower than Ki observed after TEA in human mesenteric arteries (Ki at +80 mV, 850 μmol/L). Both currents were completely blocked by 100 nmol/L iberiotoxin. Although the basic properties of K(Ca) channels in human coronary arteries are not different from those described in arteries from other species,11 13 27 30 our results indicate that there are differences in TEA sensitivity between K(Ca) channels in human coronary arteries and in human mesenteric arteries. Since both IK(Ca) TEA-inhibition dose-response curves were well fitted by a Langmuir equation with a Hill coefficient of 1, both preparations may express only a single dominant population of different K(Ca) channels. This possibility may represent an important difference in rabbit coronary VSMC IK(Ca) compared with human IK(Ca) (Ki at +60 mV, between 0.3 and 1 mmol/L; Hill coefficient unequal to 1). Whether or not different K(Ca) channels in human coronary and mesenteric smooth muscle cells have different functions in the regulation of vascular tone and/or smooth muscle cell proliferation remains to be determined.
K(dr) channels have been described in noncoronary VSMCs9 28 29 30 and in coronary VSMCs from rabbits.11 13 15 In human coronary VSMCs, we found a 4-AP–sensitive component of outward current that had characteristics similar to the basic properties of IK(dr)s reported by others in smooth muscle cells from different preparations.13 15 28 29 30 The Ki of 4-AP in human coronary VSMCs was 1.02 mmol/L. This value is close to the sensitivity of IK(dr) to 4-AP in human mesenteric arteries (Ki at +20 mV, 1.04 mmol/L).9 However, marked differences were observed in the kinetic properties of IK(dr)s in both preparations. First, IK(dr) in human mesenteric VSMCs inactivated more rapidly than in human coronary VSMCs. In mesenteric arteries, IK(dr) inactivated by ≈50% within 300 milliseconds (at +80 mV). In human coronary arteries, on the other hand, this value was ≈10%. Second, in human mesenteric arteries, V0.5 was seen at −38.0 and −29.7 mV for transient and sustained current components, respectively. The currents increased as much as e-fold per 5.5-mV (steepness factor k) and 6.2-mV depolarization, respectively. In human coronary arteries, V0.5 was observed at −26.0 mV, and IK(dr) increased e-fold per 12.1 mV (k). These differences may indicate that diverse K(dr) channels are expressed in different human vascular beds. The inactivation parameters of human coronary IK(dr) are also different from those reported for K(dr) channels in many differing nonhuman vascular preparations, including coronary arteries.13 15 28 29 37 38
4-AP–sensitive K(dr) channels were not expressed in all VSMCs isolated from human coronary arteries. The reason for this finding is unclear. One possible explanation is the existence of heterogeneous coronary VSMCs. In agreement with this suggestion are morphological and biochemical studies demonstrating the heterogeneity of VSMCs in the arterial wall of pulmonary arteries.21 Alternatively, some cells may not express the channels because of different metabolic states induced by the cell isolation procedure or because of channel rundown after the isolation procedure. In this context, it should be noted that 4-AP–sensitive K(dr) channels were detected by Xu and Lee35 in canine coronary arteries but not by Buljubasic et al,14 who studied the same preparation.
We observed a voltage-dependent transient outward current, IK(to), in 4% of VSMCs. This current was activated at potentials positive to 0 mV and inactivated very rapidly over the duration of the pulse (tau, 65 milliseconds at +50 mV). The time constant of IK(to) decay decreased with more positive pulse potentials. The current had characteristics of the fast transient K+ current (“A”-type current) observed in neurons and with a 4-AP–sensitive voltage-dependent outward K+ current, Ifo (activation positive to −65 mV; tau, ≈65 milliseconds at +20 mV), in VSMCs of the portal vein.39 However, the time constant of Ifo decay increased as the amplitude of the voltage step increased. Therefore, we suggest that IK(to) has not been described in any VSMCs so far. Possibly, it is present only in human coronary VSMCs. The current’s function is unknown. Because of the low number of cells exhibiting transient outward currents, we could not perform further pharmacological and electrophysiological characterization of IK(to).
We found that pinacidil and PACAP-27 at concentrations that induce human coronary vasorelaxation3 7 led to activation of a K+ current that shared several properties with the IK(ATP) that was activated by pinacidil and other K+ channel openers in noncoronary vascular preparations.4 20 This K+ current showed little voltage sensitivity. For example, the current responded almost instantaneously to changes in voltage. Moreover, the current was sensitive to glibenclamide. The current also shifted the reversal potential of the entire transmembrane current, thereby determining resting membrane potential, to potentials near the EK.
The results of our studies may have therapeutic implications. For instance, cardiac and coronary K(ATP) channels apparently operate with very low activity under normal metabolic conditions. However, they are activated when the oxygen supply and, consequently, the intracellular high-energy phosphate values fall below critical levels.8 Thus, the opening of K(ATP) channels may be considered as an “emergency” response to prevent energy failure and to preserve the viability of the tissue during ischemic episodes. Recent studies suggest that pinacidil may have beneficial effects in the ischemic myocardium. Pinacidil and other K+ channel openers are viewed as exogenous “ischemic preconditioners,” which enable the heart to survive during limited periods of ischemia by opening K(ATP) channels. Blockade of K(ATP) channels with glibenclamide interrupts this process in several animal species, including rabbits, dogs, and pigs.40 In humans, glibenclamide at oral doses sufficient to treat type II diabetes mellitus prevented the beneficial effects of preconditioning.41 The cardiovascular mortality was threefold higher in diabetics treated with tolbutamide, another sulfonylurea that blocks K(ATP) channels, compared with treatment with insulin.42 Thus, opening of K(ATP) channels appears to be a necessary link in the chain of events leading to cardioprotection and “preconditioning” initiated by endogenous signals, such as factors from heart muscle (adenosine) or perivascular nerves (calcitonin gene–related peptide and PACAP-27) that regulate smooth muscle membrane potential. These factors could conceivably place the heart in a state of preconditioning by opening K(ATP) channels. The present data demonstrating that PACAP-27 and pinacidil activate coronary K(ATP) channels in humans support the view.
IK(STOC)s were observed in 67% of cells tested. IK(STOC)s with a similar duration were observed in VSMCs from guinea pig coronary artery, rabbit ear artery, and dog carotid artery; however, IK(STOC)s have not been described in humans.12 16 43 Even when internal Ca2+ was low-buffered with EGTA, IK(STOC)s were not observed in canine coronary artery, canine renal artery, or rabbit coronary artery.11 13 30 IK(STOC)s may reflect K+ currents through a single clustered population of iberiotoxin-sensitive maxi K(Ca) channels that have been activated by local and transient Ca2+ release from sarcoplasmic reticulum.16 33 Nelson et al34 proposed that IK(STOC)s control the diameter of small myogenic cerebral arteries. Although the physiological meaning of IK(STOC)s for the regulation of large epicardial arteries is presently unclear, the striking similarity of the present results and those obtained in other smooth muscle cells suggests that these phenomena may be common in VSMCs, including coronary arteries of humans.
I-V relationships recorded around the resting membrane potential in submucosal and cerebral arterioles and in isolated vascular cells of resistance-sized cerebral arteries show inward rectification; namely, the conductance is higher for inward than for outward currents.44 Our data indicate that inward rectifier K+ channels are not functional in VSMCs of large epicardial coronary arteries. In contrast to their putative role in resistance-sized coronary arteries,1 they do not underlie the resting K+ conductance and so do not play a major role in the maintenance of the resting potential of epicardial coronary arteries in humans.
In conclusion, the present study is the first characterization of K+ channel currents in human coronary VSMCs. Although the basic properties of the K+ channel types in human coronary arteries are similar to those measured in a vast number of other preparations, we found two K+ currents (IK(STOC) and K(to)) that have not been described in humans. We found K(dr), which has features distinct from the voltage-dependent IK(dr), in human mesenteric artery. We observed IK(ATP), which was largely quiescent until activated by vasodilators. The characterization was possible because of the isolation method, which used fresh human coronary arteries. The identified K+ channels may function as important mechanisms against coronary artery VSMC depolarization and, hence, coronary vasoconstriction. The K+ channels possibly serve to prevent vasospasm and are potential targets of pharmacological vasodilators. The electrophysiological and pharmacological profile of the currents we identified allows the investigation of different K+ channel types in intact coronary arteries and may foster the development of antianginal drugs selectively targeting K+ channels.
Selected Abbreviations and Acronyms
|d1, d2, d3||=||current density of 6.2, 4.8, and 1.9 pA/pF|
|EK||=||K+ reversal potential|
|I/Imax||=||percent current as fraction of its control value|
|I20, I80||=||peak outward current measured at +20 and +80 mV|
|IK(ATP)||=||ATP-dependent K+ channel current|
|IK(Ca)||=||Ca2+-activated K+ current|
|IK(dr)||=||delayed rectifier K+ current|
|IK(ir)||=||inward rectifier K+ current|
|IK(STOC)||=||spontaneous transient outward K+ current|
|IK(to)||=||transient outward K+ current|
|K(ATP) channel||=||ATP-dependent K+ channel|
|K(Ca) channel||=||Ca2+-activated K+ channel|
|K(dr) channel||=||delayed rectifier K+ channel|
|K(to) channel||=||transient outward K+ channel|
|PACAP||=||pituitary adenylate cyclase–activating peptide|
|VSMC||=||vascular smooth muscle cell|
This study was supported by the Deutsche Forschungsgemeinschaft (Ha 1388/2-3) and by the Bundesministerium für Forschung und Technologie (Dr Gollasch). We thank Prof R. Hetzer from the Deutsches Herzzentrum, Berlin, for supplying us with tissue from human hearts during orthotopic heart transplantations. We are grateful to Prof G. Baumann for support and to Drs L. Bruch and A. Kästner for helpful discussions.
- Received May 30, 1995.
- Accepted January 10, 1996.
- © 1996 American Heart Association, Inc.
Gollasch M, Bychkov R, Ried C, Behrendt F, Scholze S, Luft FC, Haller H. Pinacidil relaxes porcine and human coronary arteries by activating ATP-dependent potassium channels in smooth muscle cells. J Pharmacol Exp Ther. 1995;275:681-692.
Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT. Hyperpolarizing vasodilators activate ATP-sensitive K+ channels in arterial smooth muscle. Science. 1989;245:177-180.
Gollasch H, Hescheler J, Nelson MT. Arterial vasorelaxation induced by iloprost is prevented by blockers of ATP-dependent K+ channels. Naunyn Schmiedebergs Arch Pharmacol. 1991;344(suppl):R40. Abstract.
Kästner A, Bruch L, Will-Shahab L, Modersohn D, Baumann G. Pituitary adenylate cyclase activating peptides are endothelium-independent dilators of human and porcine coronary arteries. Agents Actions. 1995;45:283-290.
Daut J, Maier-Rudolph W, von Beckerath N, Mehrke G, Günther K, Goedel-Meinen L. Hypoxic dilation of coronary arteries is mediated by ATP-sensitive potassium channels. Science. 1990;247:1341-1344.
Smirnov SV, Aaronson PI. Ca2+-activated and voltage-gated K+ currents in smooth muscle cells isolated from mesenteric arteries. J Physiol (Lond). 1992;457:454-462.
Ganitkevich V, Isenberg G. Isolated guinea pig coronary smooth muscle cells: acetylcholine induces hyperpolarization due to sarcoplasmic reticulum calcium release activating potassium channels. Circ Res. 1990;67:525-528.
Leblanc N, Wan X, Leung PM. Physiological role of Ca2+-activated and voltage-dependent K+ currents in rabbit coronary myocytes. Am J Physiol. 1994;266:C1523-C1537.
Buljubasic N, Marijic J, Kampine JP, Bosnjak ZJ. Calcium-sensitive potassium current in isolated canine coronary smooth muscle cells. Can J Physiol Pharmacol. 1993;72:189-198.
Klieber HG, Daut J. A glibenclamide sensitive potassium conductance in terminal arterioles isolated from guinea pig heart. Cardiovasc Res. 1994;28:823-830.
Frid MG, Moiseeva EP, Stenmark KR. Multiple phenotypically distinct smooth muscle cell populations exist in the adult and developing bovine pulmonary arterial media in vivo. Circ Res. 1994;75:669-681.
Gollasch M, Hescheler J, Quayle J, Patlak J, Nelson MT. Single Ca-channel currents from arterial smooth muscle at physiological calcium concentrations. Am J Physiol. 1992;263:C948-C952.
Gollasch M, Haller H, Schultz G, Hescheler J. Thyrotropin-releasing hormone induces opposite effects on Ca2+ channel currents in pituitary cells by two pathways. Proc Natl Acad Sci U S A. 1991;88:10262-10266.
Gollasch M, Kleuss C, Hescheler J, Wittig B, Schultz G. Gi2 and protein kinase C are required for thyrotropin-releasing hormone-induced stimulation of voltage-dependent Ca2+ channels in pituitary GH3 cells. Proc Natl Acad Sci U S A. 1993;90:6265-6269.
Horn R, Marty A. Muscarinic activation of ionic currents measured by a new whole-cell recording method. J Gen Physiol. 1988;92:145-159.
Langton PD, Nelson MT, Huang Y, Standen NB. Block of calcium-activated potassium channels in mammalian arterial myocytes by tetraethylammonium ions. Am J Physiol. 1991;260:H927-H934.
Robertson BE, Nelson MT. Aminopyridine inhibition and voltage dependence of K+ currents in smooth muscle cells from cerebral arteries. Am J Physiol. 1994;267:C1589-C1597.
Gelband CH, Hume JR. Ionic currents in single smooth muscle cells of the canine renal artery. Circ Res. 1992;71:745-758.
Gelband CH, McCullough JR. Modulation of rabbit aortic Ca2+-activated K+ channels by pinacidil, cromakalim, and glibenclamide. Am J Physiol. 1993;264:C1119-C1127.
Nelson MT, Cheng H, Rubart M, Santana LF, Bonev AD, Knot HJ, Lederer WJ. Relaxation of arterial smooth muscle by calcium sparks. Science. 1995;270:633-637.
Parratt JR, Kane KA. KATP channels in ischaemic preconditioning. Cardiovasc Res. 1994;28:783-787.
Tomai F, Crea F, Gaspardone A, Versaci F, De Paulis R, Penta de Peppo A, Chiariello L, Gioffrè PA. Ischemic preconditioning during coronary angioplasty is prevented by glibenclamide, a selective ATP-sensitive K+ channel blocker. Circulation. 1994;90:700-705.
Prout TE, Knatterud GL, Meinert CL, Klimt CR. The UGDP controversy: clinical trials versus clinical implication. Diabetes. 1972;21:1035-1040.
Quayle JM, McCarron JG, Brayden JE, Nelson MT. Inward rectifier K+ currents in smooth muscle cells from rat resistance-sized cerebral arteries. Am J Physiol. 1993;265:C1363-C1370.