Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Many studies have shown that the generation of reactive oxygen species (ROS) is increased in the reperfused ischemic myocardium and that this increase is related to the progression of myocardial damage (Garlick et al., 1987; Kuzuya et al., 1990). However, more recent studies have demonstrated that ROS generation is not limited to reperfusion but also occurs during ischemia despite a low level of O₂ (Becker et al., 1999). Moreover, ROS implication has not only been shown in acute ischemia/reperfusion but also in the failing heart (Ide et al., 2000) including idiopathic dilated cardiomyopathy (Bäumer et al., 2000). Thus, the role of ROS in heart dysfunction still remains a field of growing interest.

Electrophysiological effects of ROS generally consist of a reduction in action potential amplitude and an increase in action potential duration (APD) followed by a reduction (Tarr and Valenzeno, 1989; Barrington, 1994; Satoh and Matsui, 1997), although either only a reduction (Goldhaber et al., 1989; Hayashi et al., 1989; Coetzee et al., 1990) or only an increase in APD (Barrington, 1994) have also been reported. The ionic mechanisms underlying AP modifications could involve changes in I_Na, I_Ca, and I_K (Goldhaber et al., 1989; Cerbai et al., 1991; Goldhaber and Liu, 1994; Satoh and Matsui, 1997; Ward and Giles, 1997) and other membrane currents such as I_K1 (Matsuura and Shattock, 1991).

Outward repolarizing K⁺ currents are important determinants of APD. Effects of ROS on voltage-gated K⁺ channels have been studied both in single cardiac myocytes isolated from animals and in transfected cell lines. Rose Bengal-generated ROS were shown to suppress I_K in frog atrial cells (Tarr and Valenzeno, 1991). In guinea pig ventricular myocytes, dihydroxyfumarate-generated ROS reduced I_K (Cerbai et al., 1991), while a direct application of 100 µM H₂O₂ appeared to slightly increase I_K at highly (+70 mV) depolarizing pulses (Satoh and Matsui, 1997). However, in these experiments, no attempt was made to differentiate between effects of ROS on I_Kr and I_Ks, the two components of the delayed rectifier K⁺ current (Sanguinetti and Jurkiewicz, 1990). H₂O₂ has been shown to inhibit several noncardiac K⁺ channels expressed in Xenopus oocytes (de Miera and Rudy, 1992), but no effect was detected on other K⁺ channels including hIsK (Duprat et al., 1995). Taglialatela et al. (1997) recently reported that a generator of ROS (FeSO₄/ascorbic acid) was able to increase HERG expressed in Xenopus oocytes due to a shift in channel inactivation without any change in activation properties. However, preliminary results from our laboratory (Bérubé and Daleau, 2000) showed that H₂O₂ was consistently able to change activation and deactivation kinetics of HERG expressed in a mammalian cell line.

The aim of the present study was to better characterize effects of hydrogen peroxide, an important producer of oxygen free radicals, which diffuse easily through hydrophobic membranes (Yu, 1994), on the gating properties of HERG expressed in a Chinese hamster ovary (CHO) cell line. We show that H₂O₂ at concentrations of 30 µM, 100 µM, and 1 mM accelerates HERG activation and deactivation with a different time course. Intracellular application of the antioxidant-scavenging enzyme catalase blunted these effects, while SOD selectively prevented the acceleration of HERG activation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Transfection and Cell Culture (Bérubé et al., 1999). CHO cells were maintained in Iscove's modified Dulbecco's medium (Life Technologies, Burlington, ON, Canada) supplemented with 10% fetal bovine serum (Life Technologies), 1% penicillin-streptomycin (Life Technologies), and 1% L-glutamine (Life Technologies) and incubated in a 5% CO₂ humid atmosphere incubator. HERG cDNA in pcDNA3 expression vector (kindly provided by Dr. G. Robertson, University of Wisconsin, Madison, WI) was cotransfected with the surface marker protein CD8 (EBO-pcD-CD8) in a 1:1 ratio to allow assessment of the transfection efficiency (5-10%) and identification of cells for electrophysiological study (Margolskee et al., 1993). Transfections were made by calcium phosphate precipitation using a mixture containing 10 µg of HERG/pcDNA3 and 10 µg of EBO in 500 µl of 250 mM CaCl₂. To increase the efficiency of the cotransfection, a 10% glycerol solution (glycerol shock) was applied for 3 min after 6 h of contact with the transfection mixture. After the glycerol shock, cells were washed five to six times with phosphate-buffered saline, briefly trypsinized, washed with Iscove's modified Dulbecco's medium, and plated on 35-mm Petri dishes for use within the next 36 h. For electrophysiological studies, cells were incubated with anti-CD8-coated beads for 5 min (Dynabeads M-450 CD8) prepared according to the manufacturer (Dynal, Oslo, Norway). After incubation, cells were washed twice with the bath solution to eliminate unbound beads. No morphological changes due to the cotransfection were observed in the CHO cells. The resting membrane potential of the cells was hyperpolarized by about 30 mV in HERG-transfected cells. HERG channel activation/inactivation I-V relationships were not modified by the presence of the CD8 antigen.

Whole-Cell Voltage-Clamp Recordings. Recordings were performed at 20-22°C with CHO cells in 35-mm Petri dishes mounted on the stage of an inverted microscope (IX50; Olympus, Tokyo, Japan). Currents were recorded in the whole-cell configuration of the patch-clamp technique using Axopatch 200A amplifier (Axon Instruments, Burlingame, CA). Voltage-clamp was controlled by the pCLAMP software package (version 6.02, Axon Instruments). Patch pipettes used were heat-polished to obtain a tip resistance of about 3 M when filled with intracellular solution containing the following (in mM): 130 KCl, 1 MgCl₂, 5 EGTA, 5 MgATP, 5 HEPES (pH = 7.2). Cells were superfused with a control external solution containing the following (in mM): 137 NaCl, 4 KCl, 1.8 CaCl₂, 1 MgCl₂, 10 glucose, 10 HEPES (pH = 7.4). Hydrogen peroxide (30 µM, 100 µM, 1 mM) was superfused for 10 min following a 5- to 7-min control period. Series resistance was compensated >80% to improve whole-cell voltage-clamp measurements. Currents were filtered at 1 kHz using a four-pole Bessel filter (3 dB/octave) and sampled at 2 kHz.

Chemicals. H₂O₂, catalase, and SOD from bovine liver were purchased from Sigma Chemical Co. (St. Louis, MO).

Statistical Analysis. For each recording, the time course of activation and deactivation were fitted using pCLAMP software package (ClampFit version 6.02, Axon Instruments). An iterative simplex method was used with sum of squares minimization to ensure the reliability of the fitting procedure. Best fits were obtained using a simple exponential fitting for activation and a biexponential fitting for deactivation and inactivation. Fast _deact and slow _deact correspond, respectively, to the rapid and slow time constants of deactivation. Paired t tests were used to evaluate the statistical significance. P < 0.05 was considered to indicate significance. Group data are expressed as mean ± S.E.M.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Figure 1 presents typical HERG potassium currents elicited from a holding potential of 80 mV by five subsequent depolarizing voltage pulses from 40 to +40 mV (20-mV steps), followed by repolarization to 60 mV. Panel A shows the recordings obtained under control conditions, while panel B presents currents recorded after 10-min superfusion with an external solution containing 30 µM H₂O₂. The time-dependent activation currents show inward rectification properties, which have been previously described (Sanguinetti et al., 1995). No significant effects on the isochronal (following a 1250-ms activation pulse) activated peak tail current amplitude was observed following 10-min exposure to all H₂O₂ concentrations tested. For example, following a +40-mV depolarizing pulse, tail current amplitude (elicited upon repolarization to 60 mV) changed from 460 ± 55 to 374 ± 27 pA (P = 0.18; n = 6) after 1 mM H₂O₂ superfusion. In the experiments presented in Fig. 1, although voltage steps were applied every 10 mV between 40 and +40 mV, only half of the current traces are presented for the sake of clarity. Tail currents measured in control and under H₂O₂ superfusion saturated for V_tests  +30 mV.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Effects of extracellular application of 30 µM H2O2 on HERG currents. Current traces elicited by depolarizing voltage pulses (1.25 s) between 40 and +40 mV (20-mV steps) from a holding potential of 80 mV in control conditions (A) and after 10-min superfusion with H2O2 (B).

The voltage at which rectification occurred (i.e., about +5 mV) was negatively but not significantly shifted by H₂O₂ application (Fig. 2A). The isochronal I-V curves obtained from deactivating currents (Fig. 2B) were fitted to a Boltzmann function, and the voltage at which the current was half-activated (V1/2) was changed from 1.9 ± 4.3 to 13.7 ± 6.0 mV by exposure to 30 µM H₂O₂ (P < 0.05; n = 4). The slope of the I-V relationship was not significantly affected in the presence of 30 µM H₂O₂ (8.9 ± 0.7 versus 10.5 ± 0.9; P = 0.31; n = 4).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effects of extracellular application of 30 µM H2O2 on HERG activation and rectification. Amplitude normalized to the maximum activating (A) and tail (B) currents are plotted versus test potential in controls () and in the presence of H2O2 ().

Extracellular application of H₂O₂ significantly effected activation and deactivation time constants (Table 1). For example, for a 0-mV depolarizing pulse, the time constant of activation (_act) decreased by 26%. Biexponential curve fitting of tail currents obtained upon repolarization from 0 to 60 mV showed a decrease in fast _deact by 47%, while slow _deact remained unchanged. Similar effects were observed for all H₂O₂ concentrations tested. Figure 3 shows effects of 30 µM H₂O₂ on normalized activating (panel A) and deactivating (panel B) currents.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Effects of extracellular application of 30 µM H2O2 on HERG activation and deactivation time constants. A, activating currents obtained during a 0-mV depolarizing pulse in control and following external application of 30 µM H2O2 are normalized to the maximum activation amplitude. B, deactivating currents obtained upon repolarization to 60 mV following a 1.25-s depolarizing pulse to 0 mV. Tail currents are normalized to the maximum deactivation amplitude. Curve fittings of activation (monoexponential) and deactivation (biexponential) are superimposed.

Changes in HERG activation were also assessed with three H₂O₂ concentrations (i.e., 30, 100, and 1000 µM) using the "envelope of tails" protocol with depolarizing pulses of various durations to 0, +20, and +40 mV. It appears clearly in Fig. 4 that envelope of tails current amplitudes elicited following a 0-mV voltage pulse (for which HERG inactivation is removed) increase more rapidly in the presence of H₂O₂ (30, 100, and 1000 µM), corresponding to an acceleration of HERG activation. This effect was also observed for envelope of tails protocols elicited by depolarizing pulses to +20 and +40 mV for all concentrations tested (data not shown). Activation time constant derived from the envelope of tails protocol was reduced from 870 ± 53 to 375 ± 52 ms (P = 0.003; n = 4) for a 0-mV voltage pulse and from 293 ± 46 to 220 ± 45 ms (P < 0.05; n = 4) for a +20-mV voltage pulse.

View larger version (36K): [in this window] [in a new window]   Fig. 4.   H2O2-induced changes in HERG activation assessed by an envelope of tails protocol. Envelope of tails protocols were performed to better characterize H2O2-induced changes in HERG current activation. Cells were stimulated from 80 to 0 mV for 50- to 1950-ms duration (100-ms increments), and tail currents were obtained upon a 2.5-s repolarization step to 60 mV. For clarity, only the peak deactivating current amplitudes () are presented for each pulse duration in the presence of H2O2. Panels A, B, and C present control recordings with superimposed peak tail current amplitudes from different cells obtained after a 10-min superfusion period for 30 µM, 100 µM, and 1 mM H2O2, respectively.

The time course of the changes in HERG kinetics induced by 1 mM H₂O₂ is presented in Fig. 5. In four experiments where HERG was activated by 0-mV pulses every 10 s, reduction in fast _deact was delayed compared with the reduction in _act. The mean lag time between the two phenomena was 278 ± 130 s. Similar lag times were obtained in experiments using 100 µM H₂O₂.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Time course of 1 mM H2O2 effects on HERG activation and fast deactivation time constants. A, upper panel, presents HERG activating and deactivating currents (Vtest = 0 mV) recorded before and after 400-s H2O2 superfusion. A, lower panel, presents currents recorded before and after 700-s H2O2. Curve fittings of activating and tail currents for each trace are superimposed to emphasize changes of kinetics. B, time course of H2O2 effects on normalized act () and fast deact (). Dashed lines indicate the delay between the onset of act and fast deact changes.

We assessed effects of H₂O₂ on HERG properties in the presence of intracellular catalase (1000 U/ml). As shown in Fig. 6A, no effect was detected following the application of 1 mM H₂O₂ when this reactive oxygen species scavenger was present. Intracellular application of the oxygen free radicals scavenging enzyme SOD (1000 U/ml) blunted H₂O₂-induced activation changes without affecting effects on deactivation. In the absence of H₂O₂, intracellular application of catalase or SOD did not affect HERG currents during 8- to 10-min control periods. While external application of catalase blunted the effects of H₂O₂ on HERG, superfusion of SOD did not affect H₂O₂-induced HERG modulation.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   Effects of H2O2 on HERG in the presence of intracellular catalase or SOD. Normalized HERG activating and tail currents are shown in control and following 10-min superfusion of 1 mM H2O2 in the presence of 1000 U/ml intracellular catalase (A). Intracellular application of SOD (1000 U/ml) inhibited acceleration of HERG activation induced by H2O2 but did not affect changes in deactivation (B). Activating and deactivating currents were normalized to their respective peak amplitudes.

Hydrogen peroxide (1 mM) did not affect re-inactivation elicited upon return to 0 mV following a brief hyperpolarization to 110 mV (protocol from Yang et al., 1997). Biexponential fitting of the fast inactivation showed no change in fast _inact (from 4.9 ± 1.0 to 3.4 ± 1.4 ms; P = 0.29, n = 5) and in slow _inact (from 7.6 ± 0.3 to 7.0 ± 1.2 ms; P = 0.7, n = 5). Application of 100 µM H₂O₂ did not cause any change in the voltage dependence of HERG fast inactivation (Fig. 7; n = 4).

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Effects of H2O2 on HERG channel inactivation. Steady-state inactivation curves of HERG potassium channels in control conditions and upon 100 µM H2O2 exposure are shown in the left panel (A). The data were fitted by a Boltzmann function. V1/2 and slope of the conductance (k) were not different in control and H2O2 conditions. The values for V1/2 and k were 60 ± 9 mV (n = 4) and 26 ± 1 (n = 4), respectively, in control conditions and 59 ± 12 mV and 23 ± 3, respectively, in the presence of H2O2. Measurements were made using the voltage-clamp protocol shown at the top right of panel B.

Effects of extracellular application of 30 µM H₂O₂ was also assessed on HERG currents elicited by a ventricular AP wave form voltage-clamp (350 ms) previously recorded in a guinea pig ventricular myocyte using the current-clamp mode (Fig. 8). In these experiments, the temperature was set to 37°C to better appreciate the physiological effect of H₂O₂ on HERG. Current showed a gradual increase in amplitude during the AP plateau, followed by a reduction in current amplitude during the repolarization phase. After a 10-min superfusion with H₂O₂, the amplitude of the current increased more rapidly during the AP plateau, while the deactivation occurred faster during the early diastole.

View larger version (13K): [in this window] [in a new window]   Fig. 8.   Effect of extracellular application of 30 µM H2O2 on HERG current during an action potential clamp. An action potential wave form derived from previous recording of a guinea pig ventricular myocyte (dashed line) was used to voltage-clamp CHO cells overexpressing the HERG channel. Ionic currents recorded in one cell superfused with the control external solution and in the presence of H2O2 are presented. In these experiments, the temperature was set to 37°C to better appreciate the physiological effect of H2O2 on HERG.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Results from this study show that hydrogen peroxide modulates the activation and deactivation kinetics of HERG, a channel that produces a current with characteristics similar to the major cardiac repolarizing current I_Kr. External application of H₂O₂ induced an acceleration of activation and deactivation of HERG and a negative shift of about 10 mV in V1/2 of the activation I-V curve (which reflects the acceleration of the activating current) at all H₂O₂ concentrations tested (30 µM, 100 µM, and 1 mM). These effects were not concentration-dependent in the range tested. The lower H₂O₂ concentrations tested in this study (30 and 100 µM) likely have physiological and pathological significance (Cavarocchi et al., 1986; Drummond et al., 2000). A faster HERG activation allows a greater amount of potassium ions to flow through the channel during the first 150 ms of the plateau phase of the action potential (where I_Kr is predominant). This effect can be involved in the APD reduction observed in different cardiac and cardiomyocyte preparations during superfusion with H₂O₂ (Goldhaber et al., 1989; Coetzee et al., 1990; Goldhaber and Liu, 1994). The reduction in deactivation time constant is expected to reduce the potassium current at the end of the action potential and during the first part of the diastole, which, synergistically with reduced pH during ischemia/reperfusion (Bérubé et al., 1999), could preclude the normal protective effect of the I_Kr slow deactivation against premature beats.

Effects of H₂O₂ superfusion were prevented by intracellular application of catalase (reactive oxygen species scavenger enzyme). This indicates that H₂O₂ diffuses into the intracellular medium before acting on HERG. SOD prevented H₂O₂-induced acceleration of HERG activation but not its effects on deactivation. This indicates that the oxygen free radicals superoxide anion might mediate H₂O₂ effects on HERG activation. Changes in deactivation properties that were still observed in the presence of SOD suggest that H₂O₂ affects HERG through a superoxide-independent mechanism such as the oxidation of glutathione catalyzed by glutathione peroxidase, which generates oxidized glutathione and subsequent S-thiolation of proteins (Ziegler, 1985).

H₂O₂-modulated HERG activation and deactivation kinetics were not concomitant. The reduction in _act preceded the reduction in fast _deact by about 4 min, suggesting that these effects were not produced by the same intracellular pathway or at the same site on HERG protein.

Recently, Taglialatela et al. (1997) reported that a generator of ROS (FeSO₄/ascorbic acid) was able to increase the current amplitude of HERG expressed in Xenopus oocytes due to a shift in channel inactivation without any change in activation properties. We have not observed any shift in HERG inactivation I-V curve in transfected CHO cells. The reasons for the discrepancy between their findings and ours are not clear. The difference could be explained by the dissimilar procedures used for free radicals generation, different relative importance of oxygen free radicals scavenging systems, or differences in expression systems. Disparity in ionic channel properties has already been demonstrated depending on the expression system used (i.e., oocyte versus mammalian cell line) (Baroudi et al., 2000). The effect of an externally applied compound could vary between expression systems due, for example, to the presence of oocyte yolk and the diffusion barrier of the vitelline membrane as suggested by Kiehn et al. (1996).

H₂O₂ can directly oxidize channel proteins. Recently, the N terminus of HERG was crystallized, and its three-dimensional structure corresponds to a PAS (Per, Arnt, Sim) domain. This domain is found in a wide variety of proteins both in prokaryotes and eukaryotes (Morais Cabral et al., 1998). The function of PAS domains in those proteins range from protein-protein interaction to sensitivity to redox state (Morais Cabral et al., 1998). Another target for H₂O₂ oxidative stress may be -SH groups. Cysteine residues are present both on intra- (19 residues) and extracellular (two residues) sides of the channel (three additional cysteine residues are located within membrane-spanning segments) (Splawski et al., 1998) and are putative effectors of the oxidative stress induced by hydrogen peroxide.

Other mechanisms by which H₂O₂ can modulate channel proteins include induction of membrane phospholipids peroxidation (Janero et al., 1991) and activation of intracellular signal cascades. Hydrogen peroxide is known to directly increase protein kinase C activity in ventricular myocytes (Ward and Moffat, 1995); protein kinase C is a kinase that was shown to modulate HERG channel gating (Barros et al., 1998). Furthermore, 1 mM H₂O₂ was shown to decrease the -adrenoceptor-linked signal transduction pathway in the heart by changing the functions of G_s proteins and the catalytic subunit of adenylyl cyclase (Persad et al., 1998), a mechanism by which HERG may be regulated (Kiehn et al., 1998; Thomas et al., 1999). Subfamilies of mitogen-activated protein kinases such as extracellular signal-regulated kinases are also activated by H₂O₂ in cardiac myocytes (Aikawa et al., 1997; Sabri et al., 1998), but effects of phosphorylation cascades related to the activation of such kinases on HERG are still unexplored. Further investigations should be undertaken to separate the direct effect of hydrogen peroxide on HERG channel from those that may arise from various signal transduction pathways. The changes in HERG current kinetics reported in this study are of pathological relevance and could be involved in ischemia/reperfusion-induced arrhythmias.