INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

MOLECULAR OXYGEN OCCUPIES an essential role in many of the metabolic processes associated with aerobic existence. Hydrogen peroxide (H₂O₂), superoxide radical anion (O), singlet oxygen (¹O₂), and hydroxyl radical (·OH) are produced by intracellular metabolism from a variety of cytosolic enzyme systems. NADPH oxidase and nitric oxide synthase (NOS) family contributes to oxidative stress due to the generation of free radicals. Free radicals production is regulated by an enzymatic defensive system (superoxide dismutase, catalase, glutathione peroxidase) and a nonenzymatic system that includes pyruvate, ascorbate, carotenes, and glutathione. The equilibrium between free radical production and antioxidant defenses determines the degree of oxidative stress (13).

The levels of reactive oxygen species (ROS) are enhanced during inflammation, radiation exposure, endotoxic shock, and ischemia-reperfusion. The pathologies that have been attributed to ROS-induced cell dysfunction include skeletal muscle injury (27, 30) and myocardial damage during ischemia and reperfusion. In skeletal muscle, exercise increases the rate of ROS production. This increase is associated with increased levels of lipid peroxidation and peroxidation products, low catalase concentrations, and the presence of high levels of myoglobin acting as a catalyst for the formation of oxidants (8, 27). Kourie (19) reviews the effects of ROS when interacting with ion transport systems.

Calcium- and voltage-sensitive channels of large unitary conductance (K_V,Ca) are distributed in different cells and tissues, where they modulate many cellular processes (20, 21). Because cytosolic Ca²⁺ activates K_V,Ca channels, they play an important role in coupling chemical to electric signaling. K_V,Ca channels are present abundantly in virtually all types of smooth muscle cells, where they control the resting tone (1, 17, 24). K_V,Ca channels are also redox modulated (10, 38, 39). For instance, oxidizing agents such as H₂O₂ promote channel inhibition, and the reducing agent dithiothreitol (DTT) augments channel activity (10). The effect of H₂O₂ on the K_V,Ca channel was studied by DiChiara and Reinhart (10). They reported that hSlo currents were downmodulated by the oxidizing agent with a right shift of the probability of opening (P_o) vs. voltage curves and a decrease in the single-channel P_o. In the present study, we examined the mode of action of H₂O₂ in detail, with the aim of finding a mechanistic explanation for its deleterious effects on K_V,Ca channels. We found that 1) the targets of H₂O₂ action are located in the intracellular aspect of the K_V,Ca channel; and 2) the H₂O₂ effect on the K_V,Ca channel activity is mediated by ·OH. The general conclusion is that redox modulation most probably involves a disulfide/thiol exchange of thiol groups of some of the numerous cysteines present in the carboxy terminus of the human Slowpoke (hSlo) protein.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Planar lipid bilayers and single channel recordings. Lipid bilayers were made from an 8:1 mixture of 1-palmitoyl, 2-oleoyl phosphatidylethanolamine (POPE) and 1-palmitoyl, 2-oleoyl phosphatidylcholine (POPC) in decane (13 mg lipid/ml). This lipid solution was applied across a small hole (0.2-0.3 mm in diameter) made in the wall of a Delrin cup separating two chambers of 3.5 ml cis- and 0.35 ml trans-containing symmetric salt solutions of 150 mM KCl, 10 mM 3-[N-morpholino]propanesulfonic acid K⁺ salt , pH 7, [Ca²⁺]  5 µM. Bilayer formation was followed by measuring membrane capacitance. Bilayer capacitance was measured at the end of each experiment to determine possible changes in bilayer area and/or thickness. Rat skeletal muscle was used to prepare tubule T membrane vesicles containing K_V,Ca channels as previously described (22). Membrane vesicles were added very close to the bilayer. Because depolarizing voltages and cytoplasmic Ca²⁺ activate K_V,Ca channels, the internal side of the membrane was defined according to the voltage and Ca²⁺ dependence of the channel. H₂O₂ from a concentrated stock solution was added to the indicated concentrations. The solutions were stirred for 30 s, and single-channel current records (3-60 min) were obtained at a constant applied potential of +60 mV unless otherwise stated.

Data acquisition and analysis. The current across the bilayer was measured with a low-noise current-to-voltage converter (6) connected to the solution through agar bridges made with 1 M KCl. Continuous 3- to 60-min single-channel current records were taped on a video recorder. For analysis, the current was filtered at 400 Hz with an eight-pole Bessel low-pass active filter and digitized at 500 µs/point. The electrophysiological convention was used, in which the external side of the channel was defined as zero potential. The experiments were conducted at room temperature (22 ± 2°C).

Oocyte isolation and RNA injection. Ovarian lobes were surgically removed from adult female X. laevis (Nasco) and placed in 100-mm petri dishes containing OR-2 solution (in mM: 83 NaCl, 2.5 KCl, 1 MgCl₂, 5 HEPES; pH 7.6). To dissociate the oocytes, the lobes were incubated for 60 min at 18°C in OR-2 solution containing 1 mg/ml collagenase (GIBCO BRL). Dissociated oocytes were placed in ND-96 solution (100 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 5 mM HEPES, and 50 µg/ml gentamicin, pH 7.4) and were injected with 50 nl of a solution of human myometrial cRNA of the hSlo channel -subunit containing 100 ng/µl. Oocytes were kept at 18°C in an incubator and were used for the experiments 3-4 days after RNA injection.

Electrophysiology. Macroscopic currents were recorded in cell-attached macropatches and excised inside-out patches. Patch pipettes resistance were ~1 M. Bath and pipettes contained (in mM): 110 KMES, 10 HEPES, 5 HEDTA (the affinity of HEDTA for iron is about 100-fold less compared with EDTA), pH 7.0, and the indicated Ca²⁺ concentrations. The acquisition and basic analysis of the data were performed with pClamp 6.0 software (Axon Instruments) driving a 12-bit analog interface card (Labmaster DMA, Scientific Solutions).

Variance analysis. A series of current traces were recorded after pulsing to a positive voltage from the holding potential. The average basal variance at the holding voltage was subtracted from the variance obtained during the test pulse. The subtracted variance (²) was plotted vs. mean current [I(t)] and the data were fitted using (31)

(1)

where i is the single-channel current amplitude and N the number of channels; i was obtained from the initial slope; N was obtained from the nonlinear curve-fitting analysis done using Microsoft Excel. The maximum open probability (P) was obtained according to the relation: P = I_max/iN, where I_max is the maximum mean current measured in the experiment.

Reagents. POPE and POPC in chloroform were purchased from Avanti Polar Lipids (Birmingham, AL). The 5,5'-dithio-bis-(2-nitrobenzoic acid) (DTNB), DTT, and n-decane were purchased from Sigma (St. Louis, MO). The perhydrol 30% of hydrogen peroxide, chloroform, ethanol, and methanol were purchased from Merck Chemical.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of H₂O₂ addition on the K_V,Ca single channels. Fig. 1, A and B, shows single-channel current records in the absence (control) and presence of 23 mM H₂O₂ added to the external and internal side, respectively. This experiment shows that H₂O₂ does not affect the P_o when added to the external side; P_o value remains constant even after 30 min of H₂O₂ addition. On the other hand, when the internal side was exposed to the same [H₂O₂], P_o decreased from 0.621 ± 0.032 to 0.081 ± 0.022 (n = 5) after just 3 min of addition (Fig. 1, B and C). However, the unitary conductance remained constant during the time experiments (see Fig. 1B, insets a and b). At this [H₂O₂], channel P_o decreases to a very low value in a few seconds. It is surprising that H₂O₂, despite its large membrane permeability coefficient, is ineffective when applied to the external side. This is due to the fact that, in these experiments, H₂O₂ (23 mM) was added to an external compartment having a volume of 0.35 cm³, and the internal compartment had a volume of 3.3 cm³. Considering a H₂O₂ permeability coefficient of 10⁴ cms¹ and a bilayer area of 3.14 × 10⁴ cm², the maximum [H₂O₂] that can be reached in the internal compartment is only about 2 mM, and the time to reach this concentration is >1,000 h. Even if this volume is restricted by the unstirred layers (~100 µm in thickness), the time needed to reach a concentration >4 mM (the smallest [H₂O₂] tested; see Fig. 2B) would be several hours.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Hydrogen peroxide (H2O2) induces a decreased open probability (Po) value only from the intracellular side. A, B: single-channel recordings in the absence (control) and in the presence of 23 mM H2O2 in the extracellular and intracellular side, respectively. Records a and b (in B) are insets of square section. Transmembrane potential was +50 mV. Arrows indicate the closed state. Data from 4 experiments are summarized as means ± SD in bar graphs in C and D, respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   H2O2 decreases the Po of the Ca2+-sensitive voltage-dependent K+ (KV,Ca) channel. A: single-channel current records in the absence (control) and in the presence of 8 mM internal H2O2 after 1, 12, 25, and 50 min of addition. Arrows indicate the closed state. B: time course of the KV,Ca channel Po decreases after addition of 4 (), 8 (), and 23 mM H2O2 (). , Data taken in the absence of H2O2. Transmembrane potential was +60 mV. Data are presented as means ± SD for n = 5. Solid lines are fits to the data using Eq. 4. For 4 mM H2O2, n = 103 and  = 8.35 ± 0.13 min; for 8 mM H2O2, n = 103 and  = 2.15 ± 0.30 min. The data obtained using 23 mM H2O2 were fit to a single exponential with  = 0.62 ± 0.09 min.

View larger version (35K): [in this window] [in a new window]   Fig. 3.   Reversibility of the H2O2 effect on the KV,Ca channel. A: after a control period in the absence of the oxidizing agent (top), H2O2 was added to the internal side of the channel to a final concentration of 23 mM (middle). After 30 s in the presence of H2O2, Po decreased to 0.105; after several minutes in the presence of H2O2, the internal compartment was perfused with 10 volumes of a H2O2-free solution (bottom). B: experiment done as in A, except that reversibility was tested by perfusing the internal compartment with a H2O2-free solution containing 60 µM Ca2+. C: KV,Ca channel activity can be recovered if, after 18 mM H2O2 treatment, the internal compartment is perfused with a solution containing 200 µM of Ca2+. The [Ca2+] was 5 µM during the control period and after addition of H2O2. D, E: average from 4 different experiments under the conditions described in A and B (means ± SD), respectively. F: data from 5 experiments under the conditions described in C (means ± SD).

Desferrioxamine and cysteine protect K+ channel activity against ·OH. H₂O₂ in the presence of Fe(II) can generate ·OH via the Fenton reaction (28), where H₂O₂ is reduced according to the following scheme: Fe²⁺ + H₂O₂  ·OH + Fe³⁺ + OH. This reaction is a well-known source of ·OH, and there is evidence that the iron-mediated production of ·OH is an important source of lipid peroxidation and oxidation of amino acid residues in proteins (28).

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Desferrioxamine (DFA) and cysteine (Cys) protect against the deleterious effect of H2O2. A: after a control period of 3 min, 60 µM DFA was added to the internal side of the bilayer. DFA protects against the effect of H2O2, as can be observed in the third single-channel record. B: Cys (1 mM) added before H2O2 (23 mM) protects from the decrease in Po. C, D: average of 4 experiments (means ± SD) under the conditions described in A and B, respectively.

Effect of sulfhydryl groups reducing agents. Confirming previous reports (10, 38, 39), we found that channel activity in lipid bilayers was increased by exposure of the intracellular side to the sulfhydryl (SH) reducing agent DTT (2 mM; Fig. 5A). For this experiment, we chose channels with a low P_o at the calcium concentration used (~5 µM). Figure 5, A and C, shows that P_o increased 2.3-fold after we added 2 mM DTT (0.260 ± 0.096 to 0.608 ± 0.134, n = 4). After DTT addition, the increase in P_o took less than 30 s to reach a steady state and remained constant during usual recording times (15-30 min). Perfusion of the internal side with a DTT-free solution did not reverse the DTT effect. The effect of 23 mM H₂O₂ on channel activity was partially reversed by adding 2 mM DTT after perfusing the intracellular side with about ten times the volume of the internal compartment with a H₂O₂-free buffer (Fig. 5, B and D). P_o values were: control P_o, 0.896 ± 0.124; P_o in the presence of 23 mM H₂O₂, 0.054 ± 0.006; and P_o in the presence of 2 mM DTT, 0.511 ± 0.156. 

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Dithiothreitol (DTT) can partially recover the activity of H2O2-modified KV,Ca channels. A: channels having a Po of ~0.3 at 5 µM Ca2+ were chosen; 2 mM DTT added to the internal side of the bilayer increased their Po. B: the addition of 2 mM DTT partially recovers the activity of a H2O2-modified channel. C: the average increase in Po was from 0.260 to 0.608, n = 4. D: average recovery for 4 experiments was 55% of control values.

Modification of SH groups by DTNB. To determine if SH residues were involved in channel modulation by redox agents, we used DTNB. DTNB is a hydrophilic oxidative reagent that attacks specifically SH groups in proteins in a reaction that involves a thiol-disulfide exchange mechanism. Figure 6A shows the effect of 2 mM internal DTNB on channel activity. P_o decreased 17-fold (P_o = 0.052 ± 0.007) compared with control (P_o = 0.889 ± 0.103). Channel activity was not restored by withdrawal of DTNB from the internal side of the channel, suggesting a covalent modification. However, channel activity was almost fully restored by application of 2 mM DTT to the internal side (Fig. 6, A and B). On the average, in the presence of 2 mM internal DTT, P_o increased to 0.76 ± 0.13 of the initial control value. This observation strongly suggests that the observed inhibitory effect of DTNB is specifically related to the oxidation of SH groups.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   DTT recovers the activity of 5,5'-dithio-bis-(2-nitrobenzoic acid) (DTNB)-modified channels. A: currents recorded in control conditions (top trace), after 2 mM DTNB was added to the internal side of the channel (middle trace), and after removing DTNB by washing and adding 2 mM DTT (bottom trace). B: average of 4 different experiments. Control Po values decreased from 0.889 to 0.052 with DTNB. Addition of 2 mM DTT to the cytoplasmic side of the channel increased Po to 0.758. [Ca2+] was 5 µM.

Effect of H₂O₂ on the macroscopic current induced by hSlo channels. Figure 7 shows the effect of H₂O₂ on the macroscopic K_v,Ca currents expressed in X. laevis oocytes. The addition of H₂O₂ to the external side (in the pipette) caused a decrease of ~5% of the current after 3-5 min of seal formation. No further changes were observed thereafter (Fig. 7A). Incubation of an inside-out patch for 30 min in the presence of 8 mM H₂O₂ reduced the macroscopic current by 60% (Fig. 7B). After 1 h, the current decreased further to about 20% of the control value. Under these conditions, we obtained data by directly plotting the peak tail current amplitude at a constant postpulse potential (60 mV) and as a function of the test prepulse potential in symmetrical 110 mM K⁺ (Fig. 7C). Note that H₂O₂ addition at the internal side induces a gradual shift of the tail conductance vs. voltage curves, with a clear decrease in the maximum conductance.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   H2O2 inhibits KV,Ca channel steady-state currents when added to the intracellular side. A: macroscopic currents in the cell-attached configuration were recorded in the range 150 to 140 mV. H2O2 (8 mM) was added in the pipette. B: KV,Ca current recorded in an inside-out patch was recorded in the range 50 to 140 mV at 10-mV increments. H2O2 (8 mM) was added in the bath. Internal [Ca2+] = 56 nM. C: conductance (Gtail) vs. voltage curves for the traces shown in B. Data were obtained measuring the peak amplitude of the tail currents after repolarization to 60 mV in the range 40 mV to 250-300 mV. Each point is the average of 5 patches in the inside-out configuration. The experimental data were fitted to a Boltzmann equation of the form Gtail = Gmax/{1 + exp[(V1/2 - V)/k]} where V is the applied voltage, V1/2 is the half-activation voltage, and k the slope of the Gtail vs. voltage curve. Gmax was 3,500, 3,100, 2,750, 2,600, and 2,150 pS for the Gtail vs. voltage curves taken at 0, 5, 10, 20, and 30 min, respectively.

(2)

(3)

View larger version (34K): [in this window] [in a new window]   Fig. 8.   Variance analysis of hSlo channel at different times after H2O2 addition. A, D, and G: mean current traces obtained from 256 traces recorded with the patch technique from a holding potential of 0 mV to a test pulse potential 120 mV at 0, 10, and 30 min after addition of 8 mM H2O2 to the internal side, respectively. B, E, and H: time course of the variance. C, F, and I: variance (2) vs. mean current [I(t)] fitted to the function: 2 = iI(t)  I(t)2/N (solid line), where N is the number of channels and i the unitary current. At t = 0 min, maximum Po was 0.63, i = 37.0 ± 2.3 pA, and N was 155,700 ± 15; after 10 min of H2O2 exposure, Po = 0.60, i = 34.6 ± 2.7 pA, and N = 111,000 ± 8. After 30 min of H2O2 exposure, a clear maximum cannot be observed, which implies that Po  0.5. i was calculated from the slope of the 2 vs. I(t) curve. Internal [Ca2+] was 728 nM, and V is 120 mV.

Effect of calcium after treatment with H₂O₂ on hSlo channel current. At low [Ca²⁺], the addition of 18 mM H₂O₂ produces a decrease in macroscopic current after 20 min of exposure (Fig. 9A). This effect was reversed by increasing the internal [Ca²⁺] to 100 µM. This effect was similar to that observed in single-channel experiments (Fig. 3, C and F). H₂O₂ addition at the internal side induces a shift of the tail conductance vs. voltage curves, with a clear decrease in the maximum conductance. As can be observed in Fig. 9B, the effect of the oxidant can be fully reversed by perfusing the internal side of the channel with a H₂O₂-free solution containing 100 µM Ca²⁺. In the presence of high Ca²⁺ (100 µM), the addition of 18 mM H₂O₂ does not produce a decrease in the macroscopic current after 20 min of exposure (Fig. 10, A and B).

View larger version (22K): [in this window] [in a new window]   Fig. 9.   High [Ca2+] recovers the activity of H2O2-modified channels. A: macroscopic currents were recorded using the patch-clamp technique in the inside-out configuration. The condition is the same as that in Fig. 7, but internal [Ca2+] = 498 nM and [H2O2] = 18 mM. B: Gtail vs. voltage curves for the traces shown in A. Each point is the average of 5 patches in the inside-out configuration. Gmax was 15,500, 6,000, and 14,800 pS for the Gtail vs. voltage curves taken at 0, 20, and 30 min, respectively. [Ca2+] at t = 30 min was 100 µM.

View larger version (23K): [in this window] [in a new window]   Fig. 10.   High [Ca2+] supports activity of KV,Ca against the deleterious effect of H2O2. A: macroscopic current was recorded in the inside-out configuration under the same conditions described for Fig. 7. B: relative conductance (G/Gmax = Po) vs. voltage curves. In the control condition at t = 0, [Ca2+] was 100 µM (). , After 20 min of exposure to 18 mM H2O2 addition. , After 40 min of exposure to 18 mM H2O2. The fit curve was the same that Fig. 7.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The interest in H₂O₂ as a biologically active oxygen-derived intermediate is evident, because it is associated to a series of alterations and effects in many different types of cells. H₂O₂ is not by itself reactive enough to oxidize organic molecules in an aqueous environment. Nevertheless, H₂O₂ has the ability to generate highly reactive hydroxyl free radicals through its interaction with redox-active transitional metals (2). Hydroxyls result from the decomposition of H₂O₂ via the Fenton reaction and by interaction of superoxide with H₂O₂ through the Haber-Weiss reaction (41). The biological importance of H₂O₂ stems from its participation in the production of more reactive chemical species such as ·OH, and its role as a source of free radicals has been emphasized rather than its chemical reactivity. ·OH is considered one of the most potent oxidants encountered in biological systems. However, because of its extremely short half-life, it is effective only near the locus of its production. The diffusion capability of ·OH is restricted to only about two molecular diameters before it reacts with water (41). Highly reactive ·OH readily react with a variety of molecules, such as amino acids and lipids, by removing hydrogen or by addition to unsaturated bonds (28). The lag time and its dependence on [H₂O₂] can be explained using a simple model in which n number of successful and independent hits of the H₂O₂ with different amino acid residues are necessary to "kill" a channel. A successful hit is described by the irreversible reaction
where X_mod is the residue modified by the hydrogen peroxide and k₁ the second-order rate constant describing the its chemical modification. In this model, the time dependence of the probability of opening, P_o(t), is given by the expression

(4)

where  = 1/k₁[H₂O₂] is the time constant. The best fit to the data of Fig. 2B (4 and 8 H₂O₂ mM) was obtained with n = 103. This number is very large indeed and can be explained on the basis of the large number of cysteine residues (108) present in the carboxy terminal of the K_V,Ca channel. Below we show that cysteine residues are the main target of the oxidant agent, although it is likely that methionine or tryptophan residues are also oxidized.¹ Hovewer, other mechanisms are possible, and the interpretation of the n should be taken cautiously. The model proposed demands that the value of 1/ must be directly proportional to the [H₂O₂]. The inset of Fig. 2B shows that the 1/[H₂O₂] data are well described by a straight line with a second-order rate constant k₁ = 0.56 s¹M¹. The value obtained for k₁ indicates that the oxidation is extremely slow indeed, considering that the forward rate constant in a diffusion-limited reaction is of the order of 10⁸10⁹ s¹M¹. The very short half-life of the ·OH can explain the low forward rate constant, k₁, for the channel oxidizing reaction (Fig. 2B), obtained using the multiple-hit model because the effective [·OH] at the target sites in the protein would be low. We note here that the oxidizing reaction is a multistep process in which ·OH are formed via the Fenton reaction, reacting afterwards with, for example, an SH group at a diffusion-controlled rate to form sulfinic or sulfonic acid derivatives (see, e.g., 18, 36). Therefore, the effective [·OH] should be in the nanomolar range to explain the low reaction rate found in Fig. 2B.

In the present work we found that adding H₂O₂ to the cytoplasmic aspect of the K_V,Ca channel produces a decrease in its P_o (Fig. 2). This effect was not reversed by washing or after increasing [Ca²⁺] to 60 µM (Fig. 3B). Therefore, H₂O₂ could be involved in redox reactions with some vulnerable amino acid residues (28). The protective effect showed by desferrioxamine and cysteine before treatment with 23 mM H₂O₂ (Fig. 4) implies that the decrease in P_o is mediated by the ·OH generated by the Fenton reaction.

Of course, it is possible that the oxidant agent affects other components associated to the membrane or to the channel; for example, the target of the oxidizing agent could be an auxiliary -subunit or some membrane-bound enzyme able to promote channel phosphorylation. We think that the bilayer experiments argue against that possibility, because the skeletal muscle preparation does not contain -subunits and we are working in the absence of second messengers such as ATP or cAMP. In what follows, we assume that the primary target of the ·OH is the channel-forming protein.

The decrease in the K_V,Ca channel activity by addition of H₂O₂ to the intracellular side is highly dependent on the oxidant concentration. Low [H₂O₂] (8 mM) does not affect the P_o of K_V,Ca channels in the first 8 min of a reaction. Afterwards, and in a very short time span, P_o values change drastically. Based on the effect of desferrioxamine, the reduction in P_o can be attributed to the oxidizing action of the ·OH on free SH residues of cysteines associated with the opening of the K_V,Ca channel. The differences in the effect of the H₂O₂ when it is added to the intracellular or extracellular side imply different access to essential targets. In particular, oxidation of free SH residues of cysteines, present in greater proportion at the intracellular side (27 amino acid residues per subunit), could explain the observed difference. Twenty-four of these residues occur in transmembrane or intracellular domains and are largely concentrated in the carboxy terminus of the hSlo protein. We think that the external cysteines (C14, C141, and C277) do not play an important functional role in determining P_o, because their replacement by serine produces channels indistinguishable from the wild-type in terms of voltage-Ca²⁺ dependence and H₂O₂ sensitivity (data not shown).

To determine whether the P_o decrease of the K_V,Ca channel by ·OH could be attributed to the oxidation of free SH residues of cysteine, the effect of H₂O₂ was compared with that of DTNB. This is a hydrophilic agent specific for the oxidation of free SH groups in proteins. The presence of 2 mM DTNB decreased the P_o value to an extent similar to that observed with H₂O₂. This effect was reversed by addition of 2 mM DTT (Fig. 6A). On the other hand, the addition of 2 mM DTT to a channel previously treated with H₂O₂ only produces a partial recovery in the P_o (Fig. 5B). These results would indicate that K_V,Ca channel of rat skeletal muscle can be regulated by compounds that alter the redox states of sulfhydryl groups. Similar behavior was described for a smooth muscle K_V,Ca channel (38, 39) and a voltage-insensitive K(Ca²⁺) channel of intermediate conductance present in bovine aortic endothelial cells (5). Cai and Sauvé (5) show that the oxidative effects of H₂O₂ were observed at H₂O₂ concentrations ranging from 0.5 to 10 mM. The oxidative effect of H₂O₂ was similar to hydrophilic oxidative reagents such as DTNB. The difference observed between the K_V,Ca channel activity recovery by DTT in pretreated samples with H₂O₂ (Fig. 5C) (recovery 57%) and that with DTNB (Fig. 6A) (recovery 85%) indicates that the ·OH could have a more generalized oxidative effect than DTNB. Besides cysteines, other amino acids such as tryptophan and/or methionine can be the target of the ·OH (7, 23, 28).

We note here that the addition of H₂O₂ in presence of high calcium (Fig. 9) does not produce a macroscopic conductance decrease after 20 min of exposure. In this condition, high internal [Ca²⁺] would act as a protective Ca²⁺ agent for the different sensitive groups exposed in the carboxy-terminal region, particularly at calcium binding sites (29). These experimental facts could be important, due to the fact that the intracellular Ca²⁺ is modulating and potentially protecting groups that can be oxidized and are associated to the opening and closing of the K_V,Ca channel.

Oxidation effects depend on the type of K+ channels. The effect of ROS on the K_V,Ca channel differs from that reported for voltage-dependent K⁺ or the human ether-à-gogo-related gene (HERG) channels. For example, t-butyl hydroperoxide reversibly increases the activity of both Kv1.4 and Kv3.4. This effect was attributed to an attenuation or removal of the fast inactivation processes (11). Enhancement of ROS production induced by the perfusion with Fe²⁺ and ascorbic acid caused an increase in HERG outward K⁺ currents (34). On the other hand, a decrease in ROS levels achieved by perfusion with ROS scavengers inhibited the resting outward currents induced by HERG channels and prevented their increase induced by ROS. Rose bengal (generator of singlet oxygen, ¹O₂) produced a decrease of channel activity in the case of Shaker, Kv1.3, Kv1.4, Kv1.5, and Kv3.4 channels expressed in Xenopus oocytes. Duprat et al. (11) argued that these observations might be important in disease states. Kv1.4 and Kv1.5 are fast inactivating channels expressed in cardiac cells, and their inhibition by ROS can contribute to the major electrophysiological disorders that occur during reperfusion-induced arrhythmias after ischemia and during heart failure induced by chronic pressure overload (3). Evidence of a direct effect of H₂O₂ on ATP-sensitive K⁺ (K_ATP) channels was inferred from studies where ROS effects were examined on excised membrane patches. Ichinari et al. (16) observed a dose-dependent H₂O₂-induced increase in P_o of the K_ATP channel. H₂O₂-induced irreversible inhibition of the activity of K_ATP channel in skeletal muscle has been attributed to inhibition via oxidation of SH groups (40). On the other hand, sarcoplasmic reticulum Ca²⁺ release channel (ryanodine receptor) is differentially affected by different ROS. Singlet oxygen causes an irreversible damage of the ryanodine from cardiac muscle after a brief transient period of activation (15). However, 5 mM H₂O₂ activates the same channel even at 0.45 nM cytosolic calcium, a condition in which the channel is normally silent. The activation occurs abruptly after a lag period of a few minutes (4). The activating effect of H₂O₂ has also been found for the ryanodine receptor from skeletal muscle channel of the rabbit and frog (12, 26). The rabbit channel is somewhat more sensitive; it becomes activated at 0.1 mM, and in this preparation 1-3 mM H₂O₂ inhibit channel activity (12).