This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Babenko, A. Articles by Vassort, G. Search for Related Content PubMed PubMed Citation Articles by Babenko, A. Articles by Vassort, G.

(Circulation Research. 1997;80:589-600.)
© 1997 American Heart Association, Inc.

Articles

Enhancement of the ATP-Sensitive K⁺ Current by Extracellular ATP in Rat Ventricular Myocytes

Involvement of Adenylyl Cyclase–Induced Subsarcolemmal ATP Depletion

Andrey Babenko, , Guy Vassort

From INSERM U-390, Physiopathologie Cardiovasculaire, Montpellier, France.

Correspondence to Dr Guy Vassort, INSERM U-390, Physiopathologie Cardiovasculaire, CHU Arnaud de Villeneuve, F-34295 Montpellier Cedex 05, France. E-mail vassort{at}u390.montp.inserm.fr

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Abstract ATP-sensitive K⁺ (K_ATP) channels are present at high density in membranes of cardiac cells, where they regulate cardiac function during metabolic impairment. The present study analyzes the effects of extracellular ATP (ATP_e), a P₂-purinergic agonist that can be released under various conditions in the myocardial cell bed, on K_ATP current (I_K-ATP) in rat ventricular myocytes. Under the whole-cell patch-clamp configuration at a physiological level of intracellular ATP, applying ATP_e in the micromolar range did not activate I_K-ATP. However, dialyzing the cell with a low-ATP (100 µmol/L) pipette solution elicited a slowly, quasilinearly increasing I_K-ATP that was markedly enhanced by applying ATP_e in the presence of a P₁-purinergic antagonist. The effect was reversible on washing out the agonist. The I_K-ATP enhancement was inhibited by cholera toxin treatment of the myocytes, suggesting that a G_s protein was involved to mediate the effect. Experiments on excised patches allowed us to exclude a membrane-delimited G protein–dependent pathway. Rather, the results suggested that ATP_e activates the adenylyl cyclase, since its inhibition by 2'-deoxyadenosine 3'-monophosphate and SQ-22536, which respectively interact with the purine and catalytic site of the cyclase, strongly reduced the ATP_e-induced I_K-ATP enhancement, whereas neither compound affected I_K-ATP in inside-out patches. Inhibition of cAMP-dependent protein kinase by protein kinase inhibitor peptide 5-24 did not alter the purinergic effect. The findings suggests that ATP_e triggers the activation of adenylyl cyclase, which causes a subsarcolemmal ATP depletion sufficient to enhance I_K-ATP as it develops during low-ATP dialysis of rat ventricular myocytes.

Key Words: ATP-sensitive K⁺ channel • purinergic stimulation • G protein • adenylyl cyclase • ATP depletion

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
ATP, a universal energy source, is necessary to fuel a variety of cellular mechanisms, including metabolic regulation of membrane excitability via modulation of ionic channel function. One type of K⁺ channel, initially discovered in cardiac cells by Noma¹ is specifically inhibited by ATP_i. The opening of these K_ATP channels is responsible for the long-reported action potential shortening in cardiomyocytes in anoxia.² ATP_i regulates the cardiac K_ATP channel in two different ways. First, ATP_i inhibits the channel by interacting with a high-affinity specific nucleotide-binding site. Second, in the presence of Mg²⁺ ions, ATP_i is required to maintain the channel in the operational state and reactivates the channel from rundown, presumably through a mechanism requiring ATP hydrolysis.^{3 4} The K_ATP channel has been shown to consist of a weak inward rectifier subunit, Kir 6.2, plus a member of the ATP-binding cassette superfamily, SUR in pancreatic beta cells.⁵ More recently, cardiac-like K_ATP channels were reconstituted by coexpression of Kir 6.2 and an isoform of SUR, designated SUR2.⁶

Modulation of the cardiac K_ATP channel has been the basis of numerous pharmacological studies, since these channels, abundant in the sarcolemma of mammalian cardiomyocytes, including human ventricular cells,⁷ may link directly the metabolic status of a cell to its membrane potential.¹ Sulfonylureas are the most selective inhibitors of these channels, and they are used for long-term treatment of non–insulin-dependent type II diabetes, whereas numerous synthetic K_ATP channel openers are currently suggested for their cardioprotective effects.⁴ Endogenous ligands may also regulate K_ATP channels. Adenosine and acetylcholine are expected stimulators of cardiac K_ATP channels, since it has been reported that the activation of A₁-adenosine and muscarinic receptors in the presence of GTP at the inner side of sarcolemma fragments excised from neonatal rat ventricular cells⁸ and adult guinea pig atrial⁹ and ventricular¹⁰ myocytes induces an increase in K_ATP channel activity at submillimolar ATP_i. According to these authors, such an activation occurs via a direct membrane-delimited pathway that involves a PTX-sensitive G protein. However, G^_i-1, G^_i-2, or G_o protein–mediated decrease in ATP_i sensitivity of K_ATP channel, a mechanism attributed to this direct pathway,¹¹ is not sufficient by itself to activate the K_ATP channel at physiological (millimolar) ATP_i.¹² Alternatively, second messenger–dependent pathways could be implicated. A stimulatory effect of acetylcholine on I_K-ATP in cat atrial cells has been reported to be mediated via PKC, ie, an indirect non–membrane-delimited signaling pathway.¹³ An indirect cAMP-dependent pathway has been proposed to account for the ß-adrenergic–induced increase of the pinacidil-activated I_K-ATP in canine ventricular myocytes.¹⁴ More recently, a stimulatory effect of isopro-terenol on I_K-ATP appearing during ATP_i depletion in cat ventricular myocytes was shown to be mediated by adenylyl cyclase activation without involvement of PKA.¹⁵

Although adenosine is the most considered purinergic agonist released during ischemia and other pathophysiological conditions, the ATP concentration in the extracellular space might also transiently increase.^{16 17} ATP_e affects heart function through multiple mechanisms, including modulation of different ion channels.¹⁸ ATP_e was first shown to activate a nonselective cationic conductance^{19 20} and to increase Ca²⁺ current.²¹ ATP_e also activates a Cl^- current²² and is reported to stimulate inwardly rectifying, G protein–gated, delayed rectifier– as well as muscarinic receptor–activated K⁺ channels in atrial cells.^{23 24 25}

In the present study, the effects of ATP_e were examined on I_K-ATP in isolated rat ventricular myocytes. Under whole-cell patch-clamp conditions, ATP_e, in the presence of an adenosine receptor antagonist, enhanced I_K-ATP during its slow development in cells dialyzed with a low ATP–containing solution. This ATP_e-induced modulation of I_K-ATP implied that a cholera toxin–sensitive G protein might be involved. Experiments in inside-out patch configuration demonstrated that the addition of GTP to the inner side of membrane fragments bathed in the low ATP–containing "intracellular" solution increases the P_o of K_ATP channels independent of the presence of ATP_e at the outer side of membrane. I_K-ATP enhancement was prevented by adenylyl cyclase inhibition; however, PKA activation was not necessary to mediate this effect of ATP_e. Thus, it is proposed that ATP_e-induced enhancement of I_K-ATP in rat ventricular myocytes is caused by subsarcolemmal ATP depletion subsequent to adenylyl cyclase activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cardiomyocyte Preparation
Adult rat ventricular cardiomyocytes were isolated as described previously.²⁶ Briefly, the heart was rapidly excised from pentobarbital (40 mg/kg)–anesthetized 200- to 250-g male Wistar rats and perfused according to the Langendorff method at 37°C, first with a nominally Ca²⁺-free buffer solution at 6 mL/min for 5 minutes and then for 50 minutes at 4 mL/min with the same solution supplemented with 1.3 mg/mL type A collagenase (Boehringer) and 20 µmol/L Ca²⁺. Ventricles were then separated from atria and gently dissociated by pipetting. Cells, filtered through nylon mesh and washed out from collagenase in the enzyme-free solution, were allowed to precipitate, and the supernatant was discarded. Cells from the pellet were resuspended and incubated at 37°C for 15 minutes. In the meantime, the Ca²⁺ concentration in the solution was increased step by step up to 0.3 mmol/L. Finally, cells were resuspended in the solution containing 1 mmol/L Ca²⁺ and 0.25% bovine serum albumin (Sigma Chemical Co). Cells from dissociations giving not <6x10⁶ rod-shaped cardiomyocytes were kept at low concentration (10³ cells/mL) for 8 to 10 hours at 37°C. Thirty microliters of cell suspension was dropped in a specially designed 0.7-mL perfused bath chamber mounted on a Diaphot 200 inverted Nikon microscope 10 minutes before the electrophysiological experiment, in which only a clearly striated myocyte with smooth surface was used.

Solutions and Reagents
The buffer solution for cell preparation had the following composition (mmol/L): NaCl 123, KCl 5.4, NaHCO₃ 5, NaH₂PO₄ 2, MgCl₂ 1.6, glucose 10, taurine 20, and HEPES 20 (pH 7.2 at 23°C adjusted with 1 mol/L NaOH). The standard external control solution ("extracellular" bath solution in whole-cell and outside-out configuration or pipette solution in cell-attached and inside-out configuration of the patch-clamp technique) contained (mmol/L) NaCl 140, KCl 5, MgCl₂ 1, CaCl₂ 1, and HEPES 10 (pH 7.4 adjusted with 1 mol/L NaOH). In some experiments, K⁺-rich external solution containing (mmol/L) KCl 145, MgCl₂ 1, CaCl₂ 1, and HEPES 10 (pH 7.4 adjusted with 1 mol/L KOH) was used. DPCPX (10 µmol/L), a potent adenosine receptor antagonist, was added to the external solution, unless otherwise noted. The standard internal solution ("intracellular" pipette solution in whole-cell and outside-out configurations or bathing solution for inside-out patches) containing (mmol/L) KCl 140, EGTA 5, HEPES 5, KOH 10, and MgCl₂ 1 (pH 7.2 adjusted with 1 mol/L KOH) was supplemented by 0.1 mmol/L Na₂ATP and 0.2 mmol/L GTP lithium salt, unless otherwise noted. Other nucleotides and different agents were added to either the external or the internal solution according to the experimental protocol from stock solutions just before use; the pH was carefully checked again after dilution of nucleotides. The calculated free-Ca²⁺ concentration in the internal solution was 0.1 nmol/L, and the estimated concentration of free Mg²⁺ was kept constant at 1 mmol/L by adjusting the concentration of MgCl₂ for the Mg²⁺-binding properties of nucleotides supplementing the solution.²⁷ The pipette solution used in the perforated-patch configuration contained (mmol/L) potassium aspartate 110, KCl 30, HEPES 10, and MgSO₄ 2 (pH 7.2 adjusted with 1 mol/L KOH). Such a composition was fitted to minimize Donnan's potential between this solution and the cytoplasm. Amphotericin B (1 mg) dissolved in 8 µL of DMSO was added to a small portion of the pipette solution to a final antibiotic concentration of 200 µg/mL. The solution was sonicated, light-protected, and stored at 4°C for no longer than 1 hour before use. All aqueous and DMSO-based stock solutions were stored at -20°C. DMSO, at a concentration 2-fold higher than that in any external and internal solution, did not affect the whole-cell or the single-channel current.

DPCPX (Research Biochemicals Inc), SQ-22536 (Biomol Research Laboratories, Inc), and PKI_5-24 (GIBCO-BRL, Life Technologies) were used. Other compounds were obtained from Sigma.

Different bathing solutions, including the control one, were applied to a cell using an RSC-100 rapid solution changer system (Biologic) at a flow rate of 150 to 200 µL/min and a tube switching time of 200 milliseconds. Electrophysiological experiments were performed at 23°C to 24°C.

Whole-Cell and Single-Channel Current Recordings
Whole-cell current and single-channel current were recorded using the patch-clamp technique.²⁸ Micropipettes were manufactured from borosilicate glass capillary tubes (GC120F-10, Clark Electromedical Instruments) using a P-80/PC Flaming-Brown programmable puller (Sutter Instrument Co). Pipettes with resistances of 1 to 2, 2 to 4, or 4 to 8 M when filled with the pipette solution were used for whole-cell current recordings in perforated-patch and conventional whole-cell modes or for single-channel current recordings in cell-attached and excised-patch configurations, respectively. The liquid junction potential between standard external and internal solutions was <2 mV and was not taken into account. The junction potential between the standard external solution and the pipette solution for perforated-patch recordings was 10 mV and was subtracted from the command potential. Electrophysiological criteria for the acceptance of a cell for an experiment were a membrane potential more negative than -70 mV (estimated in zero current-clamp mode) and a steady state outward current at -40-mV holding potential. Whole-cell currents were recorded using an RK-400 cell/patch-clamp amplifier (Biologic) at a holding potential of -40 mV and during repetitive slow (0.05-mV/ms) voltage-ramp stimuli from +50 to -100 mV (voltage from -40 to +50 mV and from -100 to -40 mV was changed at the voltage-ramp rate of 1 mV/ms). This voltage protocol allowed for essentially complete inactivation of Na⁺ current, and Ca²⁺ current was also not detectable at 0 mV during the ramp stimuli. Command potential generation as well as acquisition and on-line analysis of currents elicited during voltage-clamp stimuli were performed using a PCL-718 interface (Advantech Co) on a Pentium computer and BioQuest software (WC2.6 version, developed by Dr A. Alekseev, Mayo Clinic, Rochester, Minn). The whole-cell current was continuously monitored on an NIC-310 digital oscilloscope (Nicolet Instrument) and stored on magnetic tape using a DTR-1800 digital tape recorder (Biologic) for later analysis. The currents recorded in conventional whole-cell patch-clamp configuration were accepted at the start of the 10th minute after membrane patch rupture if series resistance, ranging from 8 to 16 MW, was quite stable (<10% increase during the experiment). The currents recorded in perforated whole-cell patch configuration were taken into account after 10 minutes of achieving stable access resistance ranging from 8 to 16 MW. Series resistances were compensated by 70% through the patch-clamp amplifier circuit. Whole-cell currents were normalized to cell capacitance (C_m), calculated as C_m=Q/V, where Q is the charge estimated from the capacitative transient current recorded during a hyperpolarizing 10-mV pulse (V) when micropipette capacitance was compensated. Series resistance was calculated as /C_m, where was estimated from the best fit of the capacitive transient current relaxation by exponential function. Leak compensation was not applied.

Under the designed ionic conditions and voltage protocol, ATP_e activates substantial inward currents through voltage-independent nonselective cationic²⁰ and Cl^- channels²² with reversal potential around 0 mV for both currents in the present ionic conditions. Thus, K⁺ current was estimated from the difference between currents recorded at 0 mV either during voltage-ramp stimuli applied at the time of interest or at the beginning of whole-cell current recording.

Single-channel currents were recorded using the integrator mode of the patch-clamp amplifier in cell-attached configuration at a holding potential of -70 mV in quasiphysiological ionic conditions (Na⁺-rich bathing and pipette solutions) or 0 mV with Na⁺-rich bathing and K⁺-rich pipette solutions and in excised-patch configurations at a holding potential of -40 mV in quasisymmetrical ionic conditions (K⁺-rich intracellular and extracellular solutions) or 0 mV in quasiphysiological ionic conditions (Na⁺-rich extracellular and K⁺-rich intracellular solution) at a five-pole Bessel filter cut-off frequency of 1 kHz and sampled at a step of digitization of 0.3 milliseconds on-line or during replay of records. Single-channel current-voltage relations were taken in the interactive mode using subtraction of the single-channel current traces recorded during the voltage-ramp stimuli applied when channels were predominantly in open or closed states. Single-channel currents were analyzed using BioQuest software (CHAN4.8 program). Slow fluctuations of the no-channel-open baseline current were removed by fitting of the baseline with a spline curve and subtraction of this fit from the signal. Quasi–steady state channel activity was characterized by NP_o. NP_o was calculated by dividing the mean patch current estimated over a 30- to 60-second test interval by the mean unitary current amplitude. Mean current amplitudes were calculated from the difference between peaks in a multiple Gaussian fit to all-points current amplitude histograms constructed from corresponding 30- to 60-second current record segments. The NP_o calculated over shorter (a few seconds) intervals was shown to fluctuate significantly. For presentation, long-time current records were digitized with a 6- or 60-millisecond step and plotted using SigmaPlot 5.0 software (Jandel Corp).

Data Analysis
Averaged values and error bars are expressed as mean±SD. Statistical significance was evaluated by paired and unpaired Student's t test, and differences with values of P<.05 were considered to be significant. Fitting of experimental data by theoretical functions was performed using a nonlinear Marquardt-Levenberg curve-fit algorithm using SigmaPlot 5.0 or BioQuest software.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
ATP_e Does Not Activate K_ATP Channels in Rat Ventricular Cells at Physiological Levels of ATP_i
In preliminary experiments on intact isolated rat ventricular cardiomyocytes, K_ATP channel activity was never observed in the cell-attached configuration with ATP-free (11 cells), 50 µmol/L ATP–containing (12 cells), or "two-layer" (ATP-free solution in the micropipette tip plus 50 µmol/L ATP–containing solution in remaining part of the pipette, 8 cells) pipette solution while bathing the cells in control external solution with or without 50 µmol/L ATP as a purinergic agonist. However, K_ATP channels were activated in the same membrane fragments after excising the patch to the inside-out configuration in intracellular ATP-free solution (19 of 23 cells). In 5 cells voltage-clamped at -40 mV in the conventional whole-cell configuration with 5 mmol/L ATP and 200 µmol/L GTP in the pipette solution, repetitive (ranging from seconds to a few minutes) applications of 50 µmol/L ATP within 1 hour after breaking the membrane induced inward currents as previously described.^{20 22} However, a detectable outward holding current was never activated under these conditions, although application of K⁺ channel openers (300 µmol/L cromakalim or 100 µmol/L pinacidil) or metabolic inhibition by 0.5 µmol/L carbonyl cyanide p-(trifluoro-methoxy)phenylhydrazone produced a gli-benclamide-sensitive outward K⁺ current with a magnitude of several nanoamperes. Similar observations were obtained on 4 cells in which whole-cell currents were recorded using the perforated-patch technique. The results of these preliminary studies (not illustrated) suggest that K_ATP channels in the operational state (and also in experimental conditions minimizing rapid desensitization and/or native intracellular milieu disturbance) cannot be activated by ATP_e in nondialyzed isolated rat ventricular cells or in cells dialyzed with a millimolar ATP–containing solution.

ATP_e Enhances I_K-ATP Developing During Cell Dialysis With a Low-ATP Solution
As described previously, depletion of ATP_i by dialysis of ventricular cardiomyocytes with ATP-free solution^{15 29} or treatment of cells voltage-clamped in the perforated-patch configuration with oxidative phosphorylation and glycolysis blockers (in our preliminary experiments) increased I_K-ATP; the increase soon gave way to rundown. However, dialyzing cells in the conventional whole-cell patch configuration with a low ATP (100 µmol/L)– and GTP (200 µmol/L)–containing solution elicited a slowly, quasilinearly increasing outward current that exhibited delayed rundown. As shown in Fig 1A, soon after breaking the patch, only a negligible steady state current occurred at a holding potential of -40 mV. Current traces recorded during voltage-ramp stimuli (Fig 1 inset) demonstrated that the initial holding current was predominantly due to I_K1, which remained relatively constant during at least 1 hour of recording in the presence of 10 µmol/L glibenclamide and was not significantly altered by ATP_e application (not shown, data from four cells). The external application of 50 µmol/L ATP during the early period of low-ATP cell dialysis induced only an inward current with an initial large surge, as previously described.^{20 22} Within 15 minutes, the basal current started to become more outward; then, ATP_e triggered an outward current that developed on top of the inward current and recovered on ATP_e washout. ADP negligibly increased the previously developed outward current, whereas AMP and adenosine were not effective. The subsequent applications of ATP_e as well as of its poorly hydrolyzable analogue, ATPS, similarly modulated the outward current. The late application of 10 µmol/L glibenclamide inhibited the current. As shown in the inset, the current developing during ATP_i depletion as well as the current facilitated by the purinergic stimulation had a reversal potential close to -85 mV, the equilibrium potential for K⁺ in the given ionic conditions at 23°C; the latter current demonstrated a quasiohmic current-voltage relation at potentials positive to the reversal potential. Therefore, this glibenclamide-sensitive current through K⁺ channels was identified as I_K-ATP. These results show that ATP_e in the presence of a P₁-purinoceptor antagonist mediates a substantial increase in I_K-ATP only when the current has already been partially activated during the ATP_i depletion. Thus, the modulatory effect of ATP_e may be defined as an enhancement of I_K-ATP.

View larger version (29K): [in this window] [in a new window]   Figure 1. Effects of purinergic agonists on the holding current in rat ventricular cardiomyocytes dialyzed with a 100 µmol/L ATP– and 200 µmol/L GTP–containing internal solution. A, Various purinergic agonists were applied at 50 µmol/L in the presence of DPCPX (10 µmol/L) on a cell bathed in the standard solution and held at -40 mV in the whole-cell patch-clamp configuration. Only ATP and ATPS enhanced the outward current developed during the low-ATP dialysis. The late application of glibenclamide (Glib, 10 µmol/L) inhibited IK-ATP. Voltage-ramp stimuli were applied repetitively to establish current-voltage relations of quasi–steady state current appearing under the different experimental conditions. The ramp-elicited current traces are displayed as sharp vertical deviations at this time base. Note that because of the low rate of digitization and compression during the current trace plotting, the length of the traces might not reflect full values of currents during voltage-ramp stimuli. Traces marked alphabetically are presented in the inset. Here and in the other figures, currents are normalized to cell membrane capacitance. The dashed line indicates zero current level, and the upward deviation of current trace corresponds to outwardly directed current. B, Algorithm of quantification of the ATPe-induced IK-ATP enhancement is shown. ATP-induced changes in holding current are caused by several currents that are superimposed at -40 mV (left). Taking advantage of the fact that the transient currents through both nonselective cationic and Cl- channels have a reversal potential at 0 mV under the designed experimental ionic conditions, IK-ATP was estimated from the difference between currents recorded at 0 mV during the voltage-ramp stimuli applied either at the time of interest or at the beginning of whole-cell current recording (middle). The time course of IK-ATP at 0 mV was constructed from a sequence of ramp current traces (right). The ATPe-induced increase in slope of quasilinearly rising IK-ATP (ie, "net slope," expressed in pA/pF per minute and estimated by the difference between the slopes during agonist application and basal IK-ATP development) was used to characterize the ATPe-induced enhancement of IK-ATP.

A true steady state level of I_K-ATP was not reached over 1 hour of whole-cell recordings under designed conditions of low-ATP dialysis; instead, I_K-ATP increased quasi-linearly. Thus, the ATP_e-induced changes in the rate of quasilinear increase in I_K-ATP; ie, the "slope" of I_K-ATP rather than the absolute values of the current were analyzed. An algorithm of the I_K-ATP slope calculation is illustrated in Fig 1B. First, I_K-ATP was estimated from the difference between currents recorded at 0 mV during voltage-ramp stimuli applied either at the time of interest or at the beginning of whole-cell current recording to eliminate contaminating steady state I_K1 and substantial ATP_e-induced inward currents. Second, the time course of I_K-ATP at 0 mV was constructed from a sequence of ramp current traces. Finally, the ATP_e-induced increase in slope of quasilinearly rising I_K-ATP (ie, "net slope") was estimated to quantitatively characterize the ATP_e-induced enhancement of I_K-ATP.

ATP_e Enhances I_K-ATP Dose-Dependently
To establish a dose dependence of the purinergic-induced enhancement of I_K-ATP, each cell had to be evaluated with a single application of ATP at a given concentration when a similar level of the outward current (2 to 4 pA/pF at -40 mV) was reached to a priori minimize cumulative phenomena, including desensitization, and dependence of the ATP_e-triggered effect on the previous I_K-ATP level (Fig 2). As shown in the inset, the best fit of experimental data was obtained by a pseudo-Hill equation: I=I_max [Cⁿ(Cⁿ+C₅₀ⁿ)^-1], where I and I_max are the actual and maximal increases in the quasilinear rise of I_K-ATP, respectively; C and C₅₀ are the actual and half-maximal stimulatory concentrations of ATP_e, respectively; and n is the pseudo-Hill coefficient, with C₅₀=1.7 µmol/L, n=1, and I_max=4.8 pA/pF per minute.

View larger version (25K): [in this window] [in a new window]   Figure 2. Dose dependence of ATPe-induced enhancement of IK-ATP. Representative current recordings are shown at -40-mV holding potential and during voltage-ramp stimuli obtained on cells evaluated with a single application of ATP at a given concentration in the presence of DPCPX (10 µmol/L) when a similar level of the outward current (2 to 4 pA/pF at -40 mV) was reached during cell dialysis with the low-ATP solution. Time after the membrane rupture is indicated on horizontal scales. Inset shows the best fit of experimental data by the pseudo-Hill equation: I=Imax[Cn(Cn+C50n)-1], where I and Imax are the actual and maximal increases in the quasilinear rise of IK-ATP, respectively; C and C50 are the actual and half-maximal stimulatory concentrations of ATPe, respectively; and n is the pseudo-Hill coefficient (C50=1.7 µmol/L, n=1, and Imax=4.8 pA/pF per minute). Error bars and numbers near the bars indicate SD and the number of cells, respectively.

Purinoceptor-Induced Enhancement of I_K-ATP Involves a G_s Protein
In the absence of selective agonists and antagonists to discriminate between the ionotropic P_2X (transmitter-gated cationic channels) and the P_2Y (G protein–coupled) purinoceptors, we first checked the involvement of any G protein in mediating the I_K-ATP enhancement by ATP_e by substituting 0.4 mmol/L GDPßS for 0.2 mmol/L GTP in the low ATP–containing pipette solution. As shown in Fig 3, this intervention substantially reduced the ATP_e-induced enhancement of I_K-ATP (1.45±0.82 versus 4.86±1.35 pA/pF per minute in control; P<.05 for paired and unpaired t tests on four cell pairs from four heart dissociations). To further check the type of G protein involved in mediating the effect of ATP_e, cells were preincubated for 6 to 7 hours at 37°C with either 0.5 µg/mL PTX or 5 µg/mL CTX before the electrophysiological experiments. Immobilization of G_i,o proteins in the inactive state by PTX treatment did not affect the purinergic enhancement of I_K-ATP (5.06±3.10 versus 4.78±3.56 pA/pF per minute in myocytes similarly maintained during 5 to 8 hours but without PTX; P>.05 for paired and unpaired t tests on five cell pairs from five heart dissociations). On the other hand, the ATP_e effect was strongly reduced in CTX-treated cardio-myocytes (1.04±0.42 versus 4.49±1.81 pA/pF per minute in cells similarly maintained but without CTX; P<.05 for paired t tests on five cell pairs from five different heart dissociations). This inhibitory effect of CTX treatment was attributed to specific immobilization of G_s proteins in the GTP-bound state rather than to some nonspecific cell damage, because Ca²⁺ current density was 3-fold higher than in nontreated myocytes from the same batch (data not shown). These results indicate that a G_s protein is involved in the signaling mechanism mediating the ATP_e-induced enhancement of I_K-ATP.

View larger version (27K): [in this window] [in a new window]   Figure 3. ATPe-induced IK-ATP enhancement implies a CTX-sensitive G protein. A, Representative whole-cell current recordings in cells dialyzed with control low-ATP GTP-containing solution (control) and with the same solution containing 0.4 mmol/L GDPßS instead of 0.2 mmol/L GTP and recordings in CTX- and PTX-treated myocytes illustrating the effects of ATPe on holding current. ATP was applied when a similar level of the outward current (2 to 4 pA/pF at -40 mV) was reached in each cell. B, Comparative analysis of IK-ATP enhancements induced in the above experimental conditions. *Significantly different from control group by paired and unpaired Student's t test (P<.05).

ATP-Induced Enhancement of I_K-ATP Is Not Attributable to a Membrane-Delimited GTP-Dependent Mechanism
To check the involvement of a membrane-delimited G protein–dependent signaling pathway mediating the enhancement of I_K-ATP, we examined the effect of GTPs applied to the cytoplasmic side of inside-out patches bathed in low-ATP solution on K_ATP channel activity when ATP as an extracellular ligand was present at the outer side of the membrane. In preliminary experiments on patches bathed in 100 µmol/L ATP–containing intracellular solution, without agonist in the pipette solution, it was observed that values of NP_o calculated over 30- to 60-second intervals might fluctuate spontaneously by up to 50% to 70%. The applications of 200 µmol/L GTP or 50 µmol/L GTPS were thus considered effective if they produced not less than a 2-fold increase in NP_o. In experiments on 45 patches with 50 µmol/L ATP in the pipette solution, four different situations were observed (Fig 4): (1) In four so-called "GTPS+, GTP+" patches, both GTP and GTPS triggered the NP_o increase (an observation that is usually attributed to an agonist-dependent G protein–induced modulation of K_ATP channel activity). (2) In 13 "GTPS+, GTP-" patches, GTP was not effective, whereas GTPS increased the channel activity, which indicates that in these membrane fragments, a G protein regulating the channel was not linked to the receptor to ATP_e. (3) In 13 "GTPS-, GTP-" patches, neither GTP nor GTPS stimulated the channel activity, whereas removal of ATP from the intracellular solution increased NP_o to a value not less than 50% of the initial one appearing immediately after the patch isolation in nucleotide-free solution (a result interpreted as a lack of G protein–dependent modulation of otherwise operational channels). (4) In 15 patches, NP_o demonstrated a >50% decrease within the first minute after the patch excision, or in the case of ineffective application of both GTP and GTPS, NP_o estimated over the first minute after removal of ATP from the bath was <50% of the initial level; the situation was classified as channel rundown (rundown patches). The probability of observations attributable to a purinoceptor-induced G protein–mediated modulation of K_ATP channel activity was low (13.3% of non-rundown patches in which NP_o was increased 8.05±3.99 times) (Fig 4B). However, a rapid desensitization might have occurred during the time of gigaseal formation in the presence of ATP in the pipette (combined with loss of mechanism of recovery from desensitization upon patch excision). In 12 experiments with a "two-layer" ATP-containing pipette solution expected to minimize such a desensitization phenomenon, GTP increased the channel activity in one patch of 12; ie, there was a similar low probability (12.5%, 7.1-fold increase in NP_o). Moreover, in eight outside-out patches with 100 µmol/L ATP– and 200 µmol/L GTP–containing intracellular pipette solution, the addition of 50 µmol/L ATP to the bath did not affect the channel activity (not shown). The increase in K_ATP channel activity on applying GTP might be caused by an agonist-independent GTP modulation of the channel as previously shown for another G protein–regulated K⁺ channel.³⁰ In another series of 37 patches without ATP in the pipette solution, the application of GTP increased K_ATP channel activity with a probability similar to that determined in the presence of added ATP_e (12.5%, 7.7±3.6-fold increase in NP_o). In a final series of experiments, we checked that the increase in K_ATP channel activity in the absence of added ATP_e was not due to stimulation by ATP carried through the patch membrane from the bath to the pipette solution. To prevent both the appearance of ATP or an effect of adenosine, its hydrolysis product, the pipette solution included apyrase (0.1 U/mL), DPCPX (10 µmol/L), and glucose (5 mmol/L, to maintain ectonucleotidase activity). The application of GTP and GTPS was as effective as in the previous experimental series (12.5%; NP_o was increased 6.90±3.25 times). Thus, the ATP_e-induced enhancement of I_K-ATP did not involve a direct G protein–coupling between purinoceptors and the K_ATP channel.

View larger version (32K): [in this window] [in a new window]   Figure 4. Stimulation of KATP channels seen on GTP application to the intracellular face of inside-out patches is not dependent on the presence of ATP at the outer side of membrane. A, Examples of single-channel current recordings in which applications of both GTPS and GTP significantly increased NPo (GTPS+, GTP+ patch). Dotted lines indicate recording interval over which NPo was calculated. The patch was held at -40 mV with standard intracellular bathing solution and K+-rich external solution in the pipette. Zero current level is indicated by arrows. Here and in the next figure, downward deflections correspond to inward current during channel openings, and small periodical upward deflections are the amplifier resettings due to use of integrator mode. B, Distribution of patches classified in relation to the effectiveness of GTPs to increase NPo. A patch was considered to be sensitive to GTP or GTPS when NPo was increased not less than 2-fold over basal value in the presence of 100 µmol/L ATP. Relative increases in NPo in subpopulation of the GTPS+, GTP+ patches were similar with ATP in the pipette solution, ATP in a two-layer pipette solution, and no ATP in the pipette solution and under conditions excluding uncontrolled P2- or/and P1-purinoceptor stimulation.

ATP-Induced Enhancement of I_K-ATP Is Dependent on Adenylyl Cyclase Activity
G_s proteins, whose activation is required to mediate the ATP_e-induced enhancement of I_K-ATP, are known to be linked to adenylyl cyclase. To examine the involvement of adenylyl cyclase in the signal pathway under these experimental conditions, we tested the effects of 2'd3'-AMP and SQ-22536, two structurally different adenylyl cyclase inhibitors, on the whole-cell I_K-ATP enhancement as well as on single K_ATP channel currents in inside-out patches (Fig 5). In myocytes dialyzed with the low-ATP intracellular solution added with either 200 µmol/L 2'd3'-AMP or 200 mmol/L SQ-22536 (in the latter case after the cells had been preincubated with 200 µmol/L for 20 minutes at 37°C), ATP_e-elicited I_K-ATP enhancement was significantly reduced (1.30±0.50 pA/pF per minute in five 2'd3'-AMP–dialyzed cells versus 4.09±1.74 pA/pF per minute in five control cells from the same batch; 1.37±0.82 pA/pF per minute in four SQ-22536–dialyzed myocytes versus 4.11±2.42 pA/pF per minute in four control cells from the same batch; P<.05 by paired and unpaired t test). However, at the same concentration, neither of the compounds affected K_ATP channel activity or the single-channel current amplitude in inside-out patches bathed in either nucleotide-free or low-ATP– and GTP-containing intracellular solution in the presence of ATP at the outer side of membrane. Thus, the data suggest that adenylyl cyclase activation plays an essential role in the ATP_e-induced enhancement of I_K-ATP.

View larger version (30K): [in this window] [in a new window]   Figure 5. Adenylyl cyclase is involved in mediating the ATPe-induced enhancement of IK-ATP. A, Effects of ATPe on holding current are shown in cells dialyzed with control low-ATP GTP-containing solution or the same solution supplemented by either 200 µmol/L 2'd3'-AMP or 200 µmol/L SQ-22536 (in the last case after pretreatment of cell with 200 µmol/L SQ-22536 at 37°C for 20 minutes). ATP was applied when a similar level of the outward current (2 to 4 pA/pF at -40 mV) was reached in each cell. B, Both adenylyl cyclase inhibitors reduced the enhancement of IK-ATP by ATP. *Significantly different from the control group by paired and unpaired Student's t test (P<.05). C, The two adenylyl cyclase inhibitors had no effect on KATP channel activity or the single-channel current amplitude when added to either nucleotide-free or low ATP– and GTP-containing intracellular solution in the presence of ATP at the outer side of membrane in inside-out configuration. The patch was held at -40 mV in symmetrical K+ conditions (standard intracellular bathing solution and K+-rich external solution in the pipette). Zero current level is indicated by dashed line. Similar results were obtained in two other inside-out patches in the presence of ATP in the pipette solution and in three patches without the agonist in the pipette solution. The illustrated recording represents a situation related to the GTPS-, GTP- patch. Openings with smaller single-channel current amplitude are due to contaminating activity of the strongly inwardly rectifying K+ channels.

Additional experiments were performed to compare the P₂-purinoceptor–mediated enhancement with the one mediated by isoproterenol (Fig 6). Isoproterenol, applied at 1 µmol/L under the same experimental conditions, was more effective than ATP (50 µmol/L) to enhance I_K-ATP (7.95±2.17 versus 4.49±1.81 pA/pF per minute; P<.05 by unpaired t test). The ß-adrenoceptor–mediated I_K-ATP enhancement was comparable to the one reported in cat ventricular cells.¹⁵ There was no significant additive effect when the two agonists were applied simultaneously (8.75±3.23 pA/pF per minute in the presence of both 1 µmol/L isoproterenol and 50 µmol/L ATP_e versus 7.95±2.17 pA/pF per minute in the presence of 1 µmol/L isoproterenol alone; P>.05 by unpaired t test). Furthermore, in all three groups, CTX similarly reduced the I_K-ATP enhancement (23.7±4.6%, 21.0±5.5%, and 24.6±5.8% in the presence of 50 µmol/L ATP_e, 1 µmol/L isoproterenol, and both 50 µmol/L ATP_e and 1 µmol/L isoproterenol, respectively; P>.05 by unpaired t test).

View larger version (14K): [in this window] [in a new window]   Figure 6. Comparison of the P2-purinoceptor– and ß-adrenoceptor–mediated IK-ATP enhancement. Iso indicates isoproterenol. Experimental conditions were as described in Fig 3. #Significantly different from ATP by unpaired Student's t test (P<.05). *Significantly different from control group by paired and unpaired Student's t test (P<.05).

PKA Does Not Mediate the ATP_e-Induced Enhancement of I_K-ATP
The possibility that a subsequent PKA phosphorylation of the K_ATP channel itself or some interacting channel regulatory protein after cAMP production by adenylyl cyclase might be responsible for I_K-ATP enhancement was then examined (Fig 7). Dialysis of cells with low ATP–containing solution supplemented with PKI_5-24 (10 µmol/L) did not alter the occurrence of I_K-ATP. Furthermore, although PKI_5-24 administration prevented 10 µmol/L forskolin–elicited increase in Ca²⁺ current observed in control cells from the same batch (when Ca²⁺ current was measured in cells dialyzed with the low-ATP–containing solution before I_K-ATP development and in cells dialyzed with the same solution added with 5 mmol/L ATP; data not shown), ATP_e was similarly effective at enhancing I_K-ATP in five pairs of cells dialyzed with low-ATP pipette solution with or without PKI_5-24 (3.86±2.20 and 4.24±1.68 pA/pF per minute in PKI_5-24-dialyzed and control cells, respectively). The results demonstrate that PKA is not necessary for mediating the I_K-ATP enhancement by ATP_e under these experimental conditions, assuming that PKA inhibition was achieved as was inferred from parallel experiments in which the forskolin-induced increase in Ca²⁺ current was prevented.

View larger version (28K): [in this window] [in a new window]   Figure 7. PKA activity is not a factor limiting the purinergic-induced enhancement of IK-ATP. A, Representative examples are shown of whole-cell current recordings in cells dialyzed with the control low ATP– and GTP-containing solution or with the same solution added with 10 µmol/L PKI5-24, illustrating the effects of ATPe applied when a similar level of the outward current (2 to 4 pA/pF at -40 mV) was reached in each cell. B, ATPe-induced enhancements of IK-ATP in cells treated or not by PKI5-24 are not different.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The present study has yielded several important findings: (1) ATP as a P₂-purinergic agonist enhances reversibly the I_K-ATP appearing under conditions of ATP_i depletion but not at physiological levels of ATP_i in rat ventricular cells. (2) This effect involves a CTX-sensitive G protein, although it is not directly mediated by G proteins. (3) The blocking of adenylyl cyclase inhibits the ATP_e-induced I_K-ATP enhancement; however, I_K-ATP modulation is not prevented by PKA inhibition. Altogether, these results suggest that ATP_e facilitates subsarcolemmal ATP depletion sufficient to enhance I_K-ATP by activating adenylyl cyclase.

These results were obtained using an experimental approach that allowed us to examine the agonist effect on currents through preactivated latent channels under conditions that preserved cell integrity. The whole-cell configuration of the patch-clamp technique was selected in most experiments to minimize disruption of native regulatory interaction of the K_ATP channel with the cytoskeleton³¹ or other neighboring proteins. Such a disorganization of natural membrane architecture might occur upon either isolation of membrane fragments to achieve excised-patch configurations³² or permeabilization of the sarcolemma to achieve an open-cell attached configuration.³³ Under our experimental conditions, I_K-ATP was activated during ATP_i depletion in cells dialyzed with submillimolar (100 µmol/L) ATP–containing glucose-free solution. This method was chosen because cell treatment with K⁺ channel openers or metabolic inhibitors activating K_ATP channels in intact cardiomyocytes is known to alter the channel sensitivity to nucleotides and sulfonylureas^{34 35 36 37} and because severe ATP_i depletion during cell dialysis with ATP- and glucose-free solution induced a nonlinear increase in I_K-ATP that rapidly gave way to rundown.³⁸ The chosen ATP_i concentration of 100 µmol/L was, on the one hand, low enough to induce a substantial I_K-ATP, and, on the other hand, high enough to maintain K_ATP channels in an operational state. Moreover, at this submillimolar ATP concentration, K_ATP channel activity could be maximally increased after a rightward shift of the curve describing the P_o-ATP_i concentration relation, an effect attributed to the G_i,o protein that mediates the P₁-purinoceptor stimulation.^{11 12}

ATP_e-Induced Enhancement of I_K-ATP Implies P₂-Purinoceptor Stimulation Coupled to a G_s Protein–Dependent Non–Membrane-Delimited Mechanism
Under the above experimental conditions, a slowly developing current and the ATP_e-enhanced current have all the characteristics of I_K-ATP, including voltage-independent gating, reversal potential close to the equilibrium potential for K⁺, pseudo-ohmic current-voltage relation, and inhibition by glibenclamide. The dose-dependent enhancement of I_K-ATP elicited by ATP_e shows a C₅₀ of 1.7 µmol/L, a value in the same range as most of the previously reported ATP_e effects on the same cells.¹⁸ The ATP_e stimulatory effect is attributable to the activation of a P₂-purinergic G protein–coupled receptor (a member of the seven transmembrane–spanning domain purinoceptor superfamily, probably a P_2Y subtype but not a P_2X member of the superfamily of ionotropic purinoceptors/transmitter-gated cationic channels) since the effect (1) is observed in the presence of a P₁-purinoceptor antagonist when AMP and adenosine were ineffective; (2) is mediated by various poorly hydrolyzable ATP analogues, including those that do not activatethe nonselective cationic current³⁹ ; and (3) involves a G protein, since the ATP_e-induced enhancement of I_K-ATP was inhibited in the presence of GDPßS. The G protein was sensitive to CTX but not to PTX, suggesting that a G_s rather than a G_i,o protein mediates the ATP_e effect. It has already been proposed that stimulation of purinergic receptors in rat ventricular cells implies a G_s protein to enhance Ca²⁺ current.²¹ On the other hand, inward rectifier K⁺ channels are known to be modulated by a membrane-delimited mechanism that generally involves a PTX-sensitive G protein; this has recently been extended to the ATP_e-induced modulation of the atrial muscarinic channel.²⁵ Our results do not support the contention that ATP_e increases activity of the K_ATP channel in rat ventricular myocytes, a member of the same two transmembrane–spanning domain K⁺ channel superfamily, through such a direct mechanism. In the present work, GTP applied to the cytoplasmic side of inside-out patches bathed in low-ATP solution increases channel activity similarly in the absence and presence of external ATP as well as under conditions excluding uncontrolled stimulation of P₂- or P₁-purinoceptors. The inhibitory effect of CTX treatment on the ATP_e-induced enhancement of I_K-ATP is unlikely to be due to the cross regulation of a G_i,o protein involving its phosphorylation by PKA,⁴⁰ since neither ADP ribosylation nor PKA inhibition alters the ATP_e effect. The previous observations of an unchanged K_ATP channel activity when applying a purified preactivated G_s subunit to the intracellular side of inside-out patches from rat ventricular cells⁸ are in line with a lack of direct G_s protein control of K_ATP channel activity.

Activation of the G_s protein generally triggers phosphorylation of various proteins by PKA as a consequence of adenylyl cyclase activation and cAMP production. Both Kir 6.2 and SUR2, the subunits reconstituting a cardiac-like K_ATP channel, contain several consensus PKA phosphorylation sites.^{5 6} At least one report involving canine myocytes described a cAMP-dependent pathway to account for the ß-adrenergic–induced increase in pinacidil-activated I_K-ATP in the presence of high (5 to 10 mmol/L) ATP_i.¹⁴ However, in cat myocytes dialyzed with nucleotide-free solution, mimicking the ß-adrenergic activation of adenylyl cyclase by adding 8-(4-chlorophenylthio)-cAMP did not induce I_K-ATP enhancement; similarly, the ß-adrenergic facilitation of I_K-ATP was not sensitive to PKA inhibition.¹⁵ The latter authors suggested that whatever their experimental conditions, PKA could not be an important contributor to ß-adrenergic stimulation of I_K-ATP, since Ca²⁺ current was not simultaneously increased by isoproterenol in myocytes internally dialyzed with solutions containing low (300 µmol/L) ATP even in the presence of glucose. Under our experimental conditions, dialyzing the cell with PKI_5-24 did not prevent the ATP_e-induced enhancement of I_K-ATP. Our results indicate that PKA-dependent phosphorylation of the K_ATP channel, if any, is not a limiting factor for its enhancement by ATP_e in these experiments. However, our results do not exclude the possibility that under more physiological conditions, PKA-dependent phosphorylation might alter K_ATP channel properties.

Purinergic Stimulation of Adenylyl Cyclase Induces I_K-ATP Enhancement by Depleting Subsarcolemmal ATP
That CTX prevents the purinergic-induced enhancement of I_K-ATP suggests the involvement of adenylyl cyclase pathway even though PKA-dependent phosphorylation is not mediating the effect. A similar observation was made with the ß-adrenergic–induced stimulation of I_K-ATP in cat ventricular cells, since the ß-adrenergic effect was shown to be prevented by cell dialysis with solution supplemented by 2'd3'-AMP, an adenylyl cyclase inhibitor.¹⁵ However, several recent reports demonstrate direct inhibition of K_ATP channels by various adenonucleotide derivatives (ADP-ribose⁴¹ and diaden-osine polyphosphates⁴² ). The activation of this cyclase to mediate the purinergic enhancement of I_K-ATP was supported by the observations that two adenylyl cyclase inhibitors, 2'd3'-AMP and SQ-22536, interacting respectively with the purine and the catalytic site of the cyclase, strongly and to a similar level reduce the ATP_e-induced I_K-ATP enhancement. Moreover, we checked that both compounds, used at the same concentration in both whole-cell and single-channel experiments, do not directly alter the channel function in inside-out patches. Whether ATP_e enhances the cAMP level in cardiac cells is still debatable. It is generally reported that ATP_e does not significantly affect the basal cAMP level but may increase or decrease the ß-adrenergic–mediated cAMP rise.^{21 43 44} However, in a detailed study of rat ventricular cells taken in part from the same dissociation batches as used in the present study, we found that the extracellular application of ATPS induces a significant 2-fold increase in the cAMP level (M. Pucéat and G. Vassort, unpublished data, 1996). Because of these observations and the fact that ATP is a substrate of the adenylyl cyclase, it can be proposed that purinergic stimulation by cyclase activation induces subsarcolemmal ATP_i depletion that will result in a relief of the ATP_i-dependent channel block and an increase in I_K-ATP. This proposal is supported by previous observations that in the cell-attached configuration, K_ATP channels can be activated in hypoxia without changes in the main intrinsic properties as tested after excising the same membrane patch.⁴⁵ A similar relief of ATP_i-dependent inhibition of cardiac I_K-ATP by localized ATP_i consumption has already been proposed after ß-adrenergic stimulation in cat ventricular cells, a situation that more obviously implies adenylyl cyclase activation.¹⁵ Under our experimental conditions, ATP_e was less effective than isoproterenol at enhancing I_K-ATP; the effects of both agonists were not additive and were blocked by CTX. These observations indicate that a common factor is involved, which among others could be a G_s protein. However, it was recently proposed that adenylyl cyclase, the component distal to receptor and G_s protein, limits agonist-mediated increases in the effector system activity.⁴⁶ Since several adenylyl cyclase isoforms are expressed in rat heart,⁴⁷ we cannot distinguish whether both agonists activate the same isoforms but to different extents or whether isoproterenol activates the same isoforms as ATP_e as well as some others, which would account for its larger effect. A relief of ATP_i-dependent modulation of cardiac I_K-ATP also occurs with activation of the Na⁺-K⁺ pump (Priebe et al³⁸ ). Priebe et al further demonstrated that the backward-running pump, being then a nonphysiological source of ATP_i, inhibits K_ATP channels. It has been reported that glycolytically generated ATP preferentially controls K_ATP channels,⁴⁸ and this has led to the idea that a functionally compartmentalized ATP_i (specifically a subsarcolemmal pool with a lower ATP concentration that the cytosolic bulk) controls channel activity. The above experiments were performed in the absence of glucose; nevertheless, I_K-ATP recovered its basal level on ATP_e removal. The subsarcolemmal ATP is thus controlled by the cytosolic ATP content. There should not be a significant diffusional barrier limiting access to the internal face of the membrane, in agreement with effective modulation of I_K-ATP by switching on/off the Na⁺-K⁺ pump,³⁸ even though in the latter work part of the effects could be attributed to ADP production or removal, since ADP antagonizes ATP_i inhibition of K_ATP channel.⁴⁹

Are Additional Mechanisms Involved in Mediating the ATP_e-Induced I_K-ATP Enhancement?
In the above experiments involving CTX treatment as well as 2'd3'-AMP or SQ-22536 administration, the purinergic-induced enhancement of I_K-ATP could be inhibited to 30%. This could be consequent to incomplete ADP ribosylation of the G_s protein by CTX or to partial inhibition of the adenylyl cyclase but to similar levels by the two types of compounds. Also consider that ATP_e has multiple effects,¹⁸ and several of them might enhance I_K-ATP despite our controlled experimental conditions. Intracellular acidification is known to increase this current.⁴⁹ However, under bicarbonate-free conditions and at room temperature, knowing the purinergic-induced Cl^-/HCO₃^- exchanger activation to be characterized by a very high temperature-dependent coefficient,²⁶ we would expect this pathway to be strongly limited. Purinergic stimulation is also expected to load the cell with Na⁺ and Ca²⁺ ions after activation of the nonselective cationic current and Ca²⁺ release by the sarcoplasmic reticulum.¹⁸ Na⁺ influx might activate the Na⁺-K⁺ pump and thus contribute to ATP_i depletion.³⁸ However, ATP_e-induced inward cationic currents inactivate faster than I_K-ATP increases, and contribution of the pump should be limited in our experiments by using a Na⁺-free internal solution. Similarly, a 30% remaining enhancement of I_K-ATP was observed after inhibition of either G_s protein– or the adenylyl cyclase–dependent step mediating the current enhancement by isoproterenol,¹⁵ a stimulation not known to induce Na⁺ influx but to induce Cl^--HCO₃^- exchanger activation and acidosis.⁵⁰ Ca²⁺_i has long been known to affect I_K-ATP.⁵¹ Lately, a Ca²⁺_i-induced actin cytoskeleton disassembly was shown to reduce K_ATP channel activity by facilitation of channel rundown³¹ and to increase channel activity as a result of a decrease in sensitivity to ATP_i.⁵² Moreover, an involvement of sarcolemmal Ca²⁺-ATPase–dependent ATP_i depletion could be speculated. However, under our experimental conditions, the free Ca²⁺ concentration in the pipette was buffered at subnanomolar levels by 5 mmol/L EGTA. Furthermore, a similar occurrence and ATP_e-induced enhancement of I_K-ATP was observed in three cells dialyzed with 10 mmol/L BAPTA, a Ca²⁺ buffer with faster kinetics, which should better prevent localized Ca²⁺ variations.⁵³ Purinergic stimulation of cardiomyocytes is also known to activate phospholipase C, which might not only produce inositol trisphosphate but also activate PKC by diacylglycerol. Several isoforms of this kinase, for which many consensus-phosphorylation sites are present on Kir 6.2 and SUR2,^{5 6} are dependent on Ca²⁺ ions, although it should be noted that Ca²⁺-independent PKC isoforms are predominantly translocated during purinergic stimulation of rat ventricular myocytes.⁵⁴ PKC has already been suggested to mediate the acetylcholine-induced stimulation of cardiac I_K-ATP.¹³ It is also reported that treatment with phorbol 12-myristate 13-acetate increases I_K-ATP induced by pinacidil or metabolic inhibition in rabbit ventricular cells, with the latter effect requiring concomitant adenosine receptor activation.⁵⁵ Furthermore, superfusion of rabbit and human ventricular myocytes with phorbol 12,13-didecanoate favors activation of I_K-ATP in rabbit cells dialyzed with submillimolar ATP–containing solutions, probably by reducing the channel sensitivity to ATP_i.⁵⁶ However, a mixture of purified constitutively active isoforms of PKC (, ß, , and ) was shown to reduce activity of K_ATP channels in the presence of substantially lower (50 µmol/L) ATP at the inner side of inside-out membrane fragments from rabbit ventricular cells.⁵⁷ Other, yet poorly investigated, mechanisms are possible, namely, the purinergic-triggered tyrosine kinase activation of phospholipase C,⁵⁸ which could contribute to ATP_e-induced I_K-ATP modulation in particular, since the phospholipase C substrate, phosphatidylinositol diphosphate, also regulates actin cytoskeleton assembly. However, we can conclude that at least in our experimental conditions, an adenylyl cyclase–dependent, but PKA-independent, signaling mechanism is the predominant pathway for the purinergic-induced enhancement of I_K-ATP.

Physiopathological Implications
The naturally occurring stimulation of K_ATP channels by ATP_e could be of major physiopathological importance in view of the reported effects of K⁺ channel openers.⁵⁹ Except at high concentrations, at which these drugs may accelerate automaticity and promote reentry, most recent studies have shown that K⁺ channel openers are effective in suppressing polymorphic ventricular tachyarrhythmias induced by early afterdepolarizations and triggered activity in vivo. During hypoxia and ischemia, a concomitant increase in intracellular ADP and acidosis will act to shift the apparent ATP_i sensitivity of the K⁺ channel.³ ATP that is released under these conditions may exert a cardioprotective effect by augmenting the already occurring cytoprotective action potential shortening in these cells. Because of the large K_ATP channel conductance and its high density, <1% of the channel population is sufficient to account for the degree of action potential shortening seen during hypoxia^{60 61 62} ; consequently, ATP_e has potential significant modulatory effects on action potential duration. A diverse situation might happen in the neighboring healthy cells, since the purinergic-induced enhancement would not occur at normal ATP_i levels. In these cells, a sudden ATP_e rise will not activate K_ATP channels but will depolarize the cells and trigger abnormal automaticity. This will add to intrinsic heterogeneity between epicardium and endocardium and be the basis of arrhythmias. Furthermore, K_ATP channels may be involved in the protective effect of ischemic preconditioning in relation to their expected activation by adenosine. Since ATP is released in ischemia, a role of P₂-purinergic stimulation in preconditioning can also be foreseen.

	Selected Abbreviations and Acronyms

 

2'd3'-AMP = 2'-deoxyadenosine 3'-monophosphate ATPe = extracellular ATP ATPi = intracellular ATP CTX = cholera toxin DMSO = dimethyl sulfoxide DPCPX = 8-cyclopentyl-1,3-dipropylxanthine IK-ATP = ATP-sensitive K+ current IK1 = strongly inwardly rectifying K+ channel current KATP channel = ATP-sensitive K+ channel NPo = Po, where N is the number of channels in the patch PKA = cAMP-dependent protein kinase PKC = protein kinase C PKI5-24 = PKA inhibitor peptide 5-24 Po = mean open probability PTX = pertussis toxin SUR = sulfonylurea receptor

	Acknowledgments

 
This study was supported in part by Ministère de l'Enseignement Supérieur, Fondation pour la Recherche Médicale (France), and Russian Foundation for Basic Research 95-04-12056 grants to Dr Andrey Babenko. The authors wish to thank Dr Michel Pucéat for helpful discussion.

Received August 23, 1996; accepted January 10, 1997.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Noma A. ATP-regulated K channels in cardiac muscle. Nature. 1983;305:147-148. [Medline] [Order article via Infotrieve]

2. Trautwein W, Gottstein U, Dudel J. Der Aktionstrom der Myokardfaser im Sauerstoffmangel. Pflugers Arch. 1954;260:40-60. [Medline] [Order article via Infotrieve]

3. Nichols CG, Lederer WJ. Adenosine triphosphate-sensitive potassium channels in the cardiovascular system. Am J Physiol. 1991;261:H1675-H1686. [Abstract/Free Full Text]

4. Terzic A, Jahangir A, Kurachi Y. Cardiac ATP-sensitive K⁺ channels: regulation by intracellular nucleotides and K⁺ channel-opening drugs. Am J Physiol. 1995;269:C525-C545. [Abstract/Free Full Text]

5. Inagaki N, Gonoi T, Clement JP IV, Namba N, Inazawa J, Gonzalez G, Aguilar-Bryan L, Seino S, Bryan J. Reconstitution of I_K(ATP): an inward rectifier subunit plus the sulfonylurea receptor. Science. 1995;270:1166-1170. [Abstract/Free Full Text]

6. Inagaki N, Gonoi T, Clement JP IV, Wang CZ, Aguilar-Bryan L, Bryan J, Seino S. A family of sulfonylurea receptors determines the pharmacological properties of ATP-sensitive K⁺ channels. Neuron. 1996;16:1011-1017. [Medline] [Order article via Infotrieve]

7. Babenko AP, Samoilov VO, Kazantseva ST, Shevchenko YL. ATP-sensitive K⁺-channels in the human adult ventricular cardiomyocyte membrane. FEBS Lett. 1992;313:148-150. [Medline] [Order article via Infotrieve]

8. Kirsch GE, Codina J, Birnbaumer L, Brown AM. Coupling of ATP-sensitive K⁺ channels to A₁ receptors by G proteins in rat ventricular myocytes. Am J Physiol. 1990;259:H820-H826. [Abstract/Free Full Text]

9. Ito H, Tung RT, Sugimoto T, Kobayashi I, Takahashi K, Katada T, Ui M, Kurachi Y. On the mechanism of G protein ß subunit activation of the muscarinic K⁺ channel in guinea pig atrial cell membrane: comparison with the ATP-sensitive K⁺ channel. J Gen Physiol. 1992;99:961-983. [Abstract/Free Full Text]

10. Kurachi Y, Tung RT, Ito H, Nakajima T. G protein activation of cardiac muscarinic K⁺ channels. Prog Neurobiol. 1992;39:229-246. [Medline] [Order article via Infotrieve]

11. Terzic A, Tung RT, Inanobe A, Katada T, Kurachi Y. G proteins activate ATP-sensitive K⁺ channels by antagonizing ATP-dependent gating. Neuron. 1994;12:885-893. [Medline] [Order article via Infotrieve]

12. Ito H, Vereecke J, Carmeliet E. Mode of regulation by G protein of the ATP-sensitive K⁺ channel in guinea pig ventricular cell membrane. J Physiol (Lond). 1994;478:101-108. [Abstract/Free Full Text]

13. Wang YG, Lipsius SL. ß-Adrenergic stimulation induces acetylcholine to activate ATP-sensitive K⁺ current in cat atrial myocytes. Circ Res. 1995;77:565-574. [Abstract/Free Full Text]

14. Tseng GN, Hoffman BF. Actions of pinacidil on membrane currents in canine ventricular myocytes and their modulation by intracellular ATP and cAMP. Pflugers Arch. 1990;415:414-424. [Medline] [Order article via Infotrieve]

15. Schackow TE, Ten Eick RE. Enhancement of ATP-sensitive potassium current in cat ventricular myocytes by ß-adrenoreceptor stimulation. J Physiol (Lond). 1994;474:131-14. [Abstract/Free Full Text]

16. Gordon JL. Extracellular ATP: effects, sources and fate. Biochem J. 1986;233:309-319. [Medline] [Order article via Infotrieve]

17. Vial C, Owen P, Opic II, Posel D. Significance of release of adenosine triphosphate and adenosine induced by hypoxia or adrenaline in perfused rat heart. J Mol Cell Cardiol. 1987;19:187-197. [Medline] [Order article via Infotrieve]

18. Vassort G, Pucéat M, Scamps F. Modulation of myocardial activity by extracellular ATP. Trends Cardiovasc Med. 1994;4:236-240.

19. Friel DD, Bean BP. Two ATP-activated conductances in bullfrog atrial cells. J Gen Physiol. 1988;91:1-27. [Abstract/Free Full Text]

20. Scamps F, Vassort G. Mechanism of extracellular ATP-induced depolarization in rat isolated ventricular cardiomyocytes. Pflugers Arch. 1990;417:309-316. [Medline] [Order article via Infotrieve]

21. Scamps F, Rubin V, Pucéat M, Tkachuk V, Vassort G. A G_s couples P₂-purinergic stimulation to cardiac Ca channels without cyclic AMP production. J Gen Physiol. 1992;100:675-701. [Abstract/Free Full Text]

22. Kaneda M, Fukui K, Doi K. Activation of chloride current by P₂-purinoceptors in rat ventricular myocytes. Br J Pharmacol. 1994;111:1355-1360. [Medline] [Order article via Infotrieve]

23. Friel DD, Bean BP. Dual control by ATP and acetylcholine of inwardly rectifying K⁺ channels in bovine atrial cells. Pflugers Arch. 1990;415:651-657. [Medline] [Order article via Infotrieve]

24. Fu C, Pleumsamran A, Oh U, Kim D. Different properties of the atrial G protein-gated K⁺ channels activated by extracellular ATP and adenosine. Am J Physiol. 1995;269:H1349-H1358. [Abstract/Free Full Text]

25. Matsuura H, Sakaguchi M, Tsuruhara Y, Ehara T. Activation of the muscarinic K⁺ channel by P₂-purinoceptors via pertussis toxin-sensitive G proteins in guinea-pig atrial cells. J Physiol (Lond). 1996;490:659-671. [Abstract/Free Full Text]

26. Pucéat M, Clément O, Vassort G. Extracellular MgATP activates the Cl^-/CHO₃^--exchanger in single rat cardiac cells. J Physiol (Lond). 1991;444:241-256. [Abstract/Free Full Text]

27. Fabiato A. Computer programs for calculating total from specified free or free from specified total ionic concentrations in aqueous solutions containing multiple metals and ligands. Methods Enzymol. 1988;157:378-417. [Medline] [Order article via Infotrieve]

28. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 1981;391:85-100. [Medline] [Order article via Infotrieve]

29. Noma A, Shibasaki T. Membrane current through adenosine-triphosphate-regulated potassium channels in guinea-pig ventricular cells. J Physiol (Lond). 1985;363:463-480. [Abstract/Free Full Text]

30. Okabe K, Yatani A, Brown AM. The nature and origin of spontaneous noise in G protein-gated ion channels. J Gen Physiol. 1991;97:1279-1294. [Abstract/Free Full Text]

31. Furukawa T, Yamane TI, Terai T, Katayama Y, Hiraoka L. Functional linkage of the cardiac ATP-sensitive K⁺ channel to the actin cytoskeleton. Pflugers Arch. 1996;431:504-512. [Medline] [Order article via Infotrieve]

32. Trube G, Hescheler J. Inward-rectifying channels in isolated patches of the heart cell membrane: ATP-dependence and comparison with cell-attached patches. Pflugers Arch. 1984;401:178-184. [Medline] [Order article via Infotrieve]

33. Kakei M, Noma A, Shibasaki T. Properties of adenosine-triphosphate-regulated potassium channels in guinea-pig ventricular cells. J Physiol (Lond). 1985;363:441-462. [Abstract/Free Full Text]

34. Thuringer D, Escande D. Apparent competition between ATP and the potassium channel opener RP 49356 on ATP-sensitive K⁺ channels of cardiac myocytes. Mol Pharmacol. 1989;36:897-902. [Abstract]

35. Nakayama K, Fan Z, Marumo F, Hiraoka M. Interrelation between pinacidil and intracellular ATP concentrations on activation of the ATP-sensitive K⁺ current in guinea pig ventricular myocytes. Circulation. 1990;82:2044-2051. [Abstract/Free Full Text]

36. Deutsch N, Weiss JN. ATP-sensitive K⁺ channel modification by metabolic inhibition in isolated guinea-pig ventricular myocytes. J Physiol (Lond). 1993;465:163-179. [Abstract/Free Full Text]

37. Krause E, Englert H, Gögelein H. Adenosine triphosphate-dependent K currents activated by metabolic inhibition in rat ventricular myocytes differ from those elicited by the channel opener rilmacalim. Pflugers Arch. 1995;429:625-635. [Medline] [Order article via Infotrieve]

38. Priebe L, Friedrich M, Benndorf K. Functional interaction between K_ATP channels and the Na⁺-K⁺ pump in metabolically inhibited heart cells of the guinea-pig. J Physiol (Lond). 1996;492:405-417. [Abstract/Free Full Text]

39. Babenko AP, Vassort G. P₂-purinoceptors-mediated enhancement of cardiac ATP-sensitive K⁺ channel (K_ATP) current. Prog Biophys Mol Biol. 1996;65:89. Abstract.

40. Levistre R, Berguerand M, Bereziat G, Masliah J. The cross-regulation of G_i-protein by cholera toxin involves a phosphorylation by protein kinase A. Biochem J. 1995;306:765-769.

41. Kwak YG, Park SK, Kim UH, Han MK, Eun JS, Cho KP, Chae SW. Intracellular ADP-ribose inhibits ATP-sensitive K⁺ channels in rat ventricular myocytes. Am J Physiol. 1996;40:C464-C468.

42. Jovanovic A, Terzic A. Diadenosine tetraphosphate-induced inhibition of ATP-sensitive K⁺ channels in patches excised from ventricular myocytes. Br J Pharmacol. 1996;117:233-235. [Medline] [Order article via Infotrieve]

43. Zheng JS, Christie A, DeYoung MB, Levy MN, Scarpa A. Synergism between cAMP and ATP in signal transduction in cardiac myocytes. Am J Physiol. 1992;262:C128-C135. [Abstract/Free Full Text]

44. Yamada M, Hamamori Y, Akita H, Yokoyama M. P₂-purinoceptor activation stimulates phosphoinositide hydrolysis and inhibits accumulation of cAMP in cultured ventricular myocytes. Circ Res. 1992;70:477-485. [Abstract/Free Full Text]

45. Babenko AP, Samoilov VO, Kazantseva ST. Potassium channels in the cardiomyocyte sarcolemma: initial opening under influence of hypoxia. Basic Appl Myol. 1992;2:317-324.

46. Post SR, Hilal-Dandan R, Urasawa K, Brunton LL, Insel PA. Quantification of signalling components and amplification in the ß-adrenergic-receptor-adenylate cyclase pathway in isolated adult rat ventricular myocytes. Biochem J. 1995;311:75-80.

47. Espinasse I, Iourgenko V, Defer N, Samson F, Hanoune J, Mercadier JJ. Type V, but not type VI, adenylyl cyclase mRNA accumulates in the rat heart during ontogenetic development: correlation with increased global adenylyl cyclase activity. J Mol Cell Cardiol. 1995;27:1789-1795. [Medline] [Order article via Infotrieve]

48. Weiss JN, Lamp ST. Glycolysis preferentially inhibits ATP-sensitive K⁺ channels in isolated guinea pig myocytes. Science. 1987;238:67-69. [Abstract/Free Full Text]

49. Lederer WJ, Nichols CG. Nucleotide modulation of the activity of rat heart ATP-sensitive K⁺ channels in isolated membrane patches. J Physiol (Lond). 1989;419:193-211. [Abstract/Free Full Text]

50. Désilets M, Pucéat M, Vassort G. Chloride dependence of pH modulation by ß-adrenergic agonist in rat cardiomyocytes. Circ Res. 1994;75:862-869. [Abstract/Free Full Text]

51. Findlay I. ATP-sensitive K channels in rat ventricular myocytes are blocked and inactivated by internal divalent cations. Pflugers Arch. 1987;410:313-320. [Medline] [Order article via Infotrieve]

52. Terzic A, Kurachi Y. Actin microfilament disrupters enhance K_ATP channel opening in patches from guinea-pig cardiomyocytes. J Physiol (Lond). 1996;492:395-404. [Abstract/Free Full Text]

53. Marty A, Neher E. Potassium channels in cultured bovine adrenal chromaffin cells. J Physiol (Lond). 1985;367:117-141. [Abstract/Free Full Text]

54. Pucéat M, Vassort G. Signalling by protein kinase C isoforms in the heart. Mol Cell Biochem. 1996;157:65-72. [Medline] [Order article via Infotrieve]

55. Liu Y, Gao WD, O'Rourke B, Marban E. Synergistic modulation of ATP-sensitive K⁺ currents by protein kinase C and adenosine: implications for ischemic preconditioning. Circ Res. 1996;78:443-454. [Abstract/Free Full Text]

56. Hu K, Duan D, Li G-R, Nattel S. Protein kinase C activates ATP-sensitive K⁺ current in human and rabbit ventricular myocytes. Circ Res. 1996;78:492-498. [Abstract/Free Full Text]

57. Light PE, Allen BG, Walsh MP, French RJ. Regulation of adenosine triphosphate-sensitive potassium channels from rabbit ventricular myocytes by protein kinase C and type 2A protein phosphatase. Biochemistry. 1995;34:7252-7257. [Medline] [Order article via Infotrieve]

58. Pucéat M, Vassort G. Purinergic stimulation of rat cardiomyocytes induces tyrosine phosphorylation and membrane association of phospholipase C: a major mechanism for InsP₃ generation. Biochem J. 1996;318:723-728.

59. Escande D, Cavero I. Potassium channel openers in the heart. In: Escande D, Standen N, eds. K⁺ Channels in Cardiovascular Medicine. Paris, France: Springer-Verlag; 1993:225-244.

60. Findlay I, Deroubaix E, Guiraudou P, Coraboeuf E. Effects of activation of ATP-sensitive K⁺ channels in mammalian ventricular myocytes. Am J Physiol. 1989;257:H1551-H1559. [Abstract/Free Full Text]

61. Nichols CG, Lederer WJ. The regulation of ATP-sensitive K⁺ channel activity in intact and permeabilized rat ventricular myocytes. J Physiol (Lond). 1990;423:91-110. [Abstract/Free Full Text]

62. Nichols CG, Ripoll C, Lederer WJ. ATP-sensitive potassium channel modulation of the guinea pig ventricular action potential and contraction. Circ Res. 1991;68:280-287.[Abstract/Free Full Text]

This article has been cited by other articles:

				R. Fischmeister, L. R.V. Castro, A. Abi-Gerges, F. Rochais, J. Jurevicius, J. Leroy, and G. Vandecasteele Compartmentation of Cyclic Nucleotide Signaling in the Heart: The Role of Cyclic Nucleotide Phosphodiesterases Circ. Res., October 13, 2006; 99(8): 816 - 828. [Abstract] [Full Text] [PDF]

				K. Shinmura, K. Tamaki, T. Sato, H. Ishida, and R. Bolli Prostacyclin attenuates oxidative damage of myocytes by opening mitochondrial ATP-sensitive K+ channels via the EP3 receptor Am J Physiol Heart Circ Physiol, May 1, 2005; 288(5): H2093 - H2101. [Abstract] [Full Text] [PDF]

				C.-Y. Xiao, A. Hara, K.-i. Yuhki, T. Fujino, H. Ma, Y. Okada, O. Takahata, T. Yamada, T. Murata, S. Narumiya, et al. Roles of Prostaglandin I2 and Thromboxane A2 in Cardiac Ischemia-Reperfusion Injury: A Study Using Mice Lacking Their Respective Receptors Circulation, October 30, 2001; 104(18): 2210 - 2215. [Abstract] [Full Text] [PDF]

				G. Vassort Adenosine 5'-Triphosphate: a P2-Purinergic Agonist in the Myocardium Physiol Rev, April 1, 2001; 81(2): 767 - 806. [Abstract] [Full Text] [PDF]

				A. Nishiyama, D. S. A. Majid, K. A. Taher, A. Miyatake, and L. G. Navar Relation Between Renal Interstitial ATP Concentrations and Autoregulation-Mediated Changes in Renal Vascular Resistance Circ. Res., March 31, 2000; 86(6): 656 - 662. [Abstract] [Full Text] [PDF]

				A. M. Gomez, B. G. Kerfant, and G. Vassort Microtubule Disruption Modulates Ca2+ Signaling in Rat Cardiac Myocytes Circ. Res., January 7, 2000; 86(1): 30 - 36. [Abstract] [Full Text] [PDF]

				E. Carmeliet Cardiac Ionic Currents and Acute Ischemia: From Channels to Arrhythmias Physiol Rev, July 1, 1999; 79(3): 917 - 1017. [Abstract] [Full Text] [PDF]

				A. P. Babenko, G. Gonzalez, L. Aguilar-Bryan, and J. Bryan Reconstituted Human Cardiac KATP Channels : Functional Identity With the Native Channels From the Sarcolemma of Human Ventricular Cells Circ. Res., November 30, 1998; 83(11): 1132 - 1143. [Abstract] [Full Text] [PDF]

				A. P. Babenko and J. Bryan A Conserved Inhibitory and Differential Stimulatory Action of Nucleotides on KIR6.0/SUR Complexes Is Essential for Excitation-Metabolism Coupling by KATP Channels J. Biol. Chem., December 21, 2001; 276(52): 49083 - 49092. [Abstract] [Full Text] [PDF]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Babenko, A. Articles by Vassort, G. Search for Related Content PubMed PubMed Citation Articles by Babenko, A. Articles by Vassort, G.