INTRODUCTION

Top Abstract Introduction Conclusions References

REACTIVE OXYGEN SPECIES (ROS), such as superoxide radical anion (O₂), singlet oxygen (¹O₂), hydrogen peroxide (H₂O₂), hydroxyl radical (· OH), and hypochlorous acid (HOCl), are produced as by-products of oxidative metabolism, in which energy activation and electron reduction are involved. Their production is enhanced during inflammation, aging, radiation exposure, endotoxic shock, and ischemia-reperfusion of heart, intestine, liver, kidney, and brain. They have been implicated in various cell dysfunctions (91, 126). This can be indicated by the protection provided following treatments with free radical-scavenging enzymes (9, 54, 94, 104). The mechanisms of ROS action at the cellular level are not well understood. It is obvious, therefore, that the understanding of these mechanisms is important for developing therapeutic strategies at cellular sites of dysfunction. In particular, the role of cell membranes in compartmentation and transmembrane signal transduction renders the changes in their properties the early events that are associated with cell dysfunction. This review examines the interaction of ROS with membrane phospholipids and proteins that constitute ion transport pathways, i.e., ion channels, pumps, exchangers, and cotransporters of both internal and surface membranes in general and in muscles in particular.

	PRODUCTION, IDENTIFICATION, AND PATHOLOGIES OF ROS

The metabolic pathways that are known to produce ROS include 1) the xanthine (X)/xanthine oxidase (XO) system, 2) the cyclooxygenase pathway of the arachidonic acid metabolic system, 3) the electron transport system of mitochondria, 4) the activated neutrophil system, and 5) the amyloid  protein system. The significance of the contribution of each of these ROS sources is not well understood.

The superoxide radical anion O₂ is produced by the reduction of O₂ using an electron that can be supplied by superoxide-generating NADPH oxidase as follows

(1)

In aqueous solution, the production of H₂O₂ is as follows

(2)

The Fe³⁺-induced catalysis of · OH production is shown in the Fenton reaction (61) as follows

(3)

(4)

(5)

Experimentally, different ROS-generating and/or ROS-identifying systems have been used to examine the ROS-induced modifications of ion transport pathways (see Tables 1-9). These include 1) H₂O₂; 2) tert-butyl hydroperoxide (t-BHP), a substrate of glutathione peroxidase; 3) t-butoxy (RO ·) or t-butylperoxy (ROO ·) radical-generating systems; 4) hypoxanthine (HX)/XO, a source for O₂, H₂O₂, and · OH production; 5) dihydroxyfumaric acid (DHF); 6) cumene hydroperoxide or purine/XO; 7) photooxidizing rose bengal, a source for ¹O₂; 8) ionizing -irradiation, diethylenetriaminepentaacetic acid, and catalase/XO; 9) HX/XO, FeCl₃, and ADP; 10) phorbol myristate acetate activation of H₂O₂ production in neutrophils; and 11) the free radical scavenger N-acetyl-L-cysteine. Pharmacological identification and dissection of the combined ROS effects are achieved by examining the modulatory effects of specific scavengers for H₂O₂, O₂, · OH, and ¹O₂ such as catalase, superoxide dismutase (SOD), desferrioxamine, and histidine, respectively. It is thought that the effects of the non-free-radical H₂O₂ are caused by producing more highly reactive oxygen species such as the free radicals O₂ and · OH. In particular, · OH reacts rapidly with many substances, e.g., DNA, lipid, and carbohydrates.

A flowchart of the major processes of ROS underlying pathologies is shown in Fig. 1. The pathologies that have been attributed to ROS-induced cell dysfunction include 1) cardiac stunning and arrhythmia (see Refs. 35 and 61); 2) skeletal muscle injury (see Refs. 130 and 151); 3) neurological conditions (see Refs. 91 and 126), e.g., neuronal damage in Parkinson's disease (see Ref. 27); 4) neurotoxicity (107); 5) Alzheimer's disease (see Refs. 6 and 171); 6) diabetes (see Ref. 123), apoptosis of T lymphocytes (see Ref. 37), and gastric mucosal injury (see Ref. 160); and 7) hypertension (156). Some of these effects can be suppressed by free radical scavengers (9, 54, 94, 104).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Flow chart of major processes of reactive oxygen species (ROS) pathologies.

In skeletal muscle, exercise increases the rate of ROS production (30, 130, 158). The enhancement of ROS production due to the increase in activity of mitochondrial electron carriers, low catalase concentrations, the sudden changes in oxygen supply and consumption, and the presence of high levels of myoglobin acting as a catalyst for the formation of oxidants is thought to cause skeletal muscle injury (see Ref. 130). The increase of free radicals in skeletal muscle and liver cells during exhaustive exercise is associated with a decrease in mitochondrial respiratory control, loss of sarcoplasmic reticulum (SR)/endoplasmic reticulum (ER) integrity, and increased levels of peroxidation products and lipid peroxidation. These effects are similar to those observed in vitamin E-deficient animals (30).

	ROS INTERACTION WITH ION TRANSPORT PATHWAYS

The interaction of ROS with ion transport pathways in muscles can be deduced indirectly from changes in their membrane properties. Cosentino et al. (29) demonstrated the role of O₂ in the mediation of endothelium-dependent contraction. It has also been demonstrated that H₂O₂ potentiates twitch tension in cardiac (84, 136) and skeletal muscles (124, 136), and this induced tension can be decreased by catalase, a specific enzyme that hydrates H₂O₂ (136). The effects are usually characterized by amplification of tension and tension oscillation, followed by spontaneous contractions (84, 124). This effect of H₂O₂ is not mediated via end effects on the myofilaments (111, 124). This suggests that the signal transduction pathways are affected by ROS. Early studies revealed that the effects of ROS on membrane properties could be deduced from electrophysiological parameters of the membrane. These include changes in membrane current and potential, ionic gradients, action potential duration and amplitude, afterdepolarization, and spontaneous activity and loss of excitability (see Refs. 40, 166, 167).

The effects of ROS-generating systems on membrane potential are now well established. It has been demonstrated that X/XO as a ROS-generating system caused membrane depolarization and a decrease in the action potential amplitude and maximum rate of rise of action potentials in guinea pig ventricular myocardium (127). Delayed afterdepolarization and early afterdepolarization induced by t-BHP, DHF, and X/XO in guinea pig papillary muscle and canine ventricular myocytes have also been demonstrated (4, 5, 122). ROS-induced membrane depolarization has been attributed to inhibition of a Na⁺ current (11) or an inward K⁺ current (121), activation of an inwardly directed nonselective cation current (115, 152), and increase in a Ca²⁺ current that is associated with changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i). Similarly, the oscillation in [Ca²⁺]_i has been implicated in arrhythmogenic afterdepolarization (113).

ROS-induced shortening of the action potential duration has been attributed to a possible increase in a delayed rectifying K⁺ current and decrease in activation of ATP-sensitive K⁺ (K_ATP) channels and Ca²⁺ currents (121, 146). Exogenous ROS-induced changes in the electromechanical function and metabolism in isolated rabbit and guinea pig ventricles shortened the duration of the action potential, indicating a decrease in the Ca²⁺ current and time-dependent outward current (50). More recently, Tokube et al. (169) reported biphasic changes in the action potential duration, with initial lengthening of the action potential due to a rapid decrease in whole cell K⁺ currents and subsequent shortening due to a decrease of whole cell Ca²⁺ current and increase in the single ATP-sensitive time-dependent outward K⁺ current.

In cardiac, smooth, and skeletal muscles the deleterious effects of ROS, produced by leaked electrons from the electron transport system of the mitochondria, are due to their interaction with various ion transport proteins underlying transmembrane signal transduction (Fig. 2). Figure 2 indicates that an important feature of ROS interaction with ion transport proteins is the modification in Ca²⁺ homeostasis that ultimately causes muscle pathologies.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Generalized scheme showing ROS-modulated ion transport mechanisms that control Ca2+ homeostasis (Ca2+ pool) in muscles. These transport mechanisms include 1) ion channels: Ca2+ channels, including voltage-sensitive L-type Ca2+ currents (V Ca2+), ligand Ca2+ channels (R Ca2+), dihydropyridine receptor (DHPR) voltage sensor, ryanodine receptor (RyR) Ca2+-release channels, and D-myo-inositol 1,4,5-trisphosphate receptor (IP3-R) Ca2+-release channels; K+ channels, such as ATP-sensitive K+ channels; and Cl channels, such as the small Cl (SCl) channel; 2) ion pumps: such as sarcoplasmic reticulum (SR) and sarcolemmal (S) ATP Ca2+ pumps and Na+-K+-ATPase (Na+ pump); and 3) ion exchangers: Na+/Ca2+ exchanger. Excited cell membrane (sarcolemma of skeletal muscle) or specific receptor (R; in sarcolemma of cardiac and smooth muscles) communicate with the Ca2+ sender (SR in skeletal and smooth muscles and sarcolemma in cardiac muscle) by means of the T tubule (T-T) or by second messengers, such as cAMP (cardiac muscle) or IP3 (smooth muscle). M, mitochondria; MF, myofilaments; R, receptor  (cardiac muscle) and 1 (smooth muscles).

Ion Channels

Ca²⁺ channels. L-TYPE VOLTAGE-SENSITIVE CA²⁺ CURRENTS. L-type voltage-sensitive Ca²⁺ channels play an important role in Ca²⁺ homeostasis in ventricular myocytes. Hence numerous studies have been conducted to examine the effects of ROS on these channels and to determine their contribution to the alterations in Ca²⁺ homeostasis under adverse conditions (Table 1). It appears that the data for the effects of ROS on current peak, amount of current, and kinetics of L-type Ca²⁺ channels in ventricular myocytes are conflicting. It has also been reported that H₂O₂ has no influence on L-type Ca²⁺ current (100, 101). In contrast to the finding that ROS-induced reduction in peak current was associated with an increase in mean current due to slowing of the inactivation (26), Tokube et al. (169) reported a decrease in the current peak with no changes in the activation time course of this current. On the other hand, Cerbai et al. (20), Matsuura and Shattock (115), and Moghadam and Winlow (119) reported a decrease in L-type Ca²⁺ current. This decrease has been attributed to Ca²⁺-induced channel inactivation (see Ref. 115). The H₂O₂-induced decrease in the inward Ca²⁺ current in cultured Lymnaea neurons is dose dependent (119). There are data suggesting that overload due to Ca²⁺ influx through the voltage-gated Ca²⁺ channel can be ruled out, since free radicals and H₂O₂ inhibit the voltage-sensitive L-type Ca²⁺ current (48, 50, 51, 121). The inhibitory effects of HX/XO and DHF as ROS-generating systems were reversed with SOD and catalase, suggesting that both O₂ and H₂O₂ are effective (Table 1), whereas the effects of the cumene/XO ROS-generating system were irreversible (48). Internal oxidative agents used on ion channels also show that 4,4'-dithiodipyridine [DTDP; a lipophilic sulfhydryl (SH)-oxidizing agent] and thimerosal {[(o-carboxyphenyl)thio]ethyl mercury sodium salt, a hydrophilic SH-oxidizing agent} inhibit the activity of cloned rabbit smooth muscle L-type Ca²⁺ channels (23).

K⁺ channels. CA²⁺-ACTIVATED K⁺ CHANNELS. The role of ROS in modulating ion channels has also been inferred from the use of ion channel blockers together with ROS-generating systems. The K⁺ channel blocker quinidine hydrochloride reduced Ca²⁺-dependent chemiluminescence products, indicative of oxygen radical production, in human eosinophils (143). They postulated that production of oxygen free radicals by the membrane-bound NADPH oxidase may be mediated by Ca²⁺-activated K⁺ (K_Ca) channels in human eosinophils and that this mechanism may underlie the role of eosinophils in the pathogenesis of allergic diseases. Relaxation evoked by nonneurogenic electrical field stimulation, via generation of free radicals, also modified Ca²⁺-dependent channels (1, 81, 186). In contrast to the H₂O₂-induced reversible inhibition [with DTT and reduced glutathione (GSH)] of K_Ca channels in the plasma membrane of bovine aortic endothelial cells (18), the large K_Ca channel in skeletal muscle from mouse is insensitive to as high as 50 mM H₂O₂ concentration (179). Differences in H₂O₂-induced modification in channel activity could be attributed to difference in tissue types. For example, it has been found that reducing agents decrease the activity of K_Ca channels in pulmonary, but not in ear, arterial smooth muscle cells of rabbit (128). The significance of H₂O₂ inhibition of K_Ca channels derives from the fact that disruption of Ca²⁺ homeostasis is mediated via depolarization of the membrane potential (see Table 3) (18).

Na⁺ channels. Voltage-gated Na⁺ channels play an important role in cell excitability and conductance. They vary in their pharmacology and gating properties. There are few data on the effects of ROS on whole cell Na⁺ currents and no data on the unitary currents of single Na⁺ channels (Table 6). Indirect findings support the view that this channel is not very sensitive to ROS. For example, potential measurements of skeletal muscle fibers revealed that neither the resting potential nor the action potential was affected in fibers treated with 1.5 mM H₂O₂ for 30 min. This finding was taken to mean that the Na⁺ channel was not blocked (124). In experiments where the effect of SH-oxidizing agents on Na⁺ was examined, it was shown that DTDP induced no changes in expressed human cardiac Na⁺ current (23), whereas other reports suggest that SH-oxidizing agents do induce changes in this channel. It has been reported that Na⁺ channel inactivation is inhibited by some oxidants such as chloramine-T, halazone, and HOCl (177). Other oxidants such as H₂O₂ produce a shift in h, a kinetic parameter of the inactivation process, without modifying channel activation (133), whereas systems generating t-BHP (a substrate of glutathione peroxidase), t-butoxy (RO ·), or t-butylperoxy (ROO ·) radicals caused a slow inactivation and progressive decrease in Na⁺ current (11). This effect of t-BHP was selective to Na⁺ channels, since Ca²⁺ and K⁺ currents were unchanged. It is for this reason that such changes in the Na⁺ current have been proposed as mechanisms for H₂O₂-induced alterations in the electrical and contractile behavior of isolated cardiomyocytes (8).

View this table: [in this window] [in a new window]   Table 6.   Effects of ROS on Na+ and Cl channel types

Cl channels. The effects of ROS on Cl channels have not been widely examined (Table 6). However, there is evidence that Cl channels are sensitive to ROS. It has been previously shown that the Cl channel in the surface membrane of bovine trachea (134) and the voltage-dependent anion-selective channel in the mitochondrial outer membrane (190) are regulated by an oxidation-reduction mechanism. Pharmacological and biophysical studies also point to the presence of an O₂-sensing mechanism (GSH-GSSH) on the small Cl (SCl) channel in the SR of skeletal muscle (95, 97). Recently, it has been reported that the human skeletal muscle Cl current (hClC-1), as expressed in Xenopus oocytes and in human embryonic cells, is also modulated by SH-reactive compounds (105). They suggested that the mechanism of Zn²⁺-induced voltage-independent nonreversible current inhibition, like that of Cd²⁺, Hg²⁺, and other SH-reactive compounds, was mediated via binding to cysteine, histidine, or acidic side chains present near the extracellular side of the membrane. It has also been found that the ATP-sensitive SCl channel is modified by DTDP and H₂O₂ (95, 96, and unpublished observations).

Other channels. Ca²⁺ homeostasis could also be modified via nonselective ion channels. ROS have also been reported to activate a nonselective cation whole cell current in guinea pig ventricular myocytes (78). A Ca²⁺-activated nonselective cation channel has also been found to be modulated by SH reagents (92). However, ROS modulation of this channel has yet to be demonstrated. Recently, Koliwad et al. (93) reported that GSSH mediated cation channel activation in calf vascular endothelial cells during oxidant stress. The dependency, selectivity, and contribution of this channel to Ca²⁺ transport are not known.

Ion Pumps

SR and sarcolemmal Ca²⁺ pumps. These pumps are important in muscle relaxation. For example, in smooth muscles the relaxation is achieved by lowering [Ca²⁺]_cyt via Ca²⁺ efflux at the plasmalemma (Ca²⁺-ATPase and Na⁺/Ca²⁺exchange) as well as Ca²⁺ uptake via Ca²⁺-ATPase in SR (159). The two types of Ca²⁺ pumps in the plasmalemma and SR are structurally and immunologically distinct and are regulated differentially (56). Pharmacological probes also confirm these distinctions. For example, thapsigargin, a plant-derived sesquiterpene, inhibits the uptake pathway (109, 168) by binding to a specific site of various Ca²⁺-ATPase isoforms of SR and ER but not to sarcolemmal Ca²⁺-ATPase (109). The effects of ROS on Ca²⁺ pumps have been determined from modifications in 1) vasoconstrictor peptide ANG II-induced contractions of artery rings by cyclopiazonic acid, an SR Ca²⁺ pump inhibitor, 2) Ca²⁺ transients, 3) acylphosphate levels of the 115-kDa sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA2b) pump protein and 140- and 115-kDa pump proteins in the plasma membrane, 4) ATP-dependent azide-insensitive oxalate-stimulated Ca²⁺ uptake, 5) phosphate-stimulated plasma membrane Ca²⁺-ATPase, and 6) Ca²⁺-stimulated ATPase and ATP-dependent Ca²⁺ accumulation (Table 7). Both sarcolemmal and SR Ca²⁺ pumps in cardiac and smooth muscles are affected by ROS. Favero et al. (45) found that H₂O₂ (1-80 mM) had no effect on the ATP hydrolysis of the SR Ca²⁺-ATPase in skeletal muscle. By contrast, ROS induced depression in the heart sarcolemmal Ca²⁺-ATPase (86) and inhibition in Ca²⁺-ATPase in the SR (55-57, 108, 140). The plasmalemmal Ca²⁺ pump is less sensitive to ROS than the SR Ca²⁺ pump. H₂O₂ and O₂ uncouple the hydrolytic reaction of the plasmalemmal Ca²⁺ pump and inhibit the hydrolytic reaction of the SR Ca²⁺ pump (55, 56, 102). Grover et al. (57) studied interaction of ROS with the Ca²⁺ pump in the SR from pig coronary artery smooth muscle. It was found that 250 µM (K_0.5 = 74 µM) H₂O₂ inhibited ANG II- and cyclopiazonic acid-induced contractions and inhibited the increase in [Ca²⁺]_i as well as the Ca²⁺-dependent acylphosphate levels of the 115-kDa SERCA2b pump protein. They proposed that H₂O₂-induced damage to the SR Ca²⁺ pump diminishes the SR Ca²⁺ pool and decreases the smooth muscle response to ANG II. Superoxide has similar effects (57). Suzuki and Ford, (164) reported that, in SR of bovine aortic smooth muscle, exogenous HX (0.1-100 µM; average 5 µM) + XO (10 U/ml) induced concentration-dependent inhibition of Ca²⁺-ATPase. By comparison, 0.01-10 mM H₂O₂ (50% inhibition at 1 mM) inhibited the Ca²⁺-ATPase at 100 µM. H₂O₂ also, in the presence or absence of 100 µM FeSO₄, significantly depressed the SH content of L-cysteine. As far as the effect of ROS moieties are concerned, O₂ is the effective moiety, and not H₂O₂, since the inhibition of the Ca²⁺-ATPase is blocked by 100 U/ml SOD but not by 20 mM mannitol or 100 µM desferrioxamine. The free radical · OH produced irreversible inhibition of the Ca²⁺-ATPase activity and SH concentration, whereas H₂O₂ had no effect on SH concentrations (164). However, other studies have implicated ¹O₂ (102) and both O₂ and H₂O₂ (56, 57, 86, 87), whereas Rowe et al. (140) implicated H₂O₂ and · OH while the effectiveness of O₂ was not ruled out.

Na⁺-K⁺-ATPase (Na⁺ pump). Na⁺-K⁺-ATPase is a membrane-bound enzyme composed of two subunits: an -catalytic subunit with a relative molecular weight (M_r) of 90,000-110,000 and a -subunit with an M_r of 40,000-60,000. The Na⁺ pump is important for maintaining coronary tone. The pump transports three Na⁺ and two K⁺ per one ATP hydrolyzed against their concentration gradients, generating internal negative charges. The intracellular/extracellular concentrations in muscle cells are typically 5-115/130 mM Na⁺ and 130/5 mM K⁺ (see Ref. 42). There are several reaction steps in the function of the pump. The hydrolytic and transport reactions of the pump can be uncoupled. For example, treatment of inside-out erythrocyte vesicles with trypsin or chymotrypsin uncouples the transport of Na⁺ from the Na⁺-K⁺-ATPase (62). K⁺-activated ouabain-sensitive p-nitrophenylposphatase is associated with the hydrolytic step of the Na⁺ pump. Conflicting effects of ROS on Na⁺ pumps have been reported (see Ref. 62). Some of these differences in ROS effects are due to the methods used for ROS generation, in which different individual ROS may not act via the same mechanism (Table 8). These differences could also arise from resolution limitations of the flux experiments, which can be overwhelmed by the back flux of ROS-enhanced activated K⁺ channels.

H⁺ pump. The H⁺ pump is important for preventing a drastic intracellular acidification and for charge balance and membrane polarization. Oxidant stress-induced pH changes in peritoneal macrophages have been attributed to modifications in the plasmalemmal H⁺-ATPase (see Ref. 14).

Adenine nucleotide translocator, phosphate carrier, and uncoupling proteins. These proteins are present in mitochondria. The phosphate carriers catalyze the electroneutral exchange of phosphate for hydroxyl ion. The adenine nucleotide carrier binds and transports adenine nucleotides, whereas uncoupling proteins bind purine nucleotides but transport H⁺, OH, or Cl (see Ref. 137). The effects of ROS on these protein transporters are not known. However, it is very likely that they are affected by ROS. First, it is known that in skeletal muscle and liver cells free radicals increase during exhaustive exercise and this increase is associated with a decrease in mitochondrial respiratory control (30). Second, ROS have deleterious effects on mitochondrial metabolism (see Ref. 50) and are linked to a leakage of electrons from mitochondria (see Ref. 91). Third, the presence of the SH groups on the cysteine residues and the SH-induced modification in permeation of phosphate, Cl, and H⁺ (137) also suggest that ROS may modify these transporters.

Ion Exchangers

Na⁺/Ca²⁺ exchanger. The Na⁺/Ca²⁺ exchanger couples the transport of three Na⁺ to that of a single Ca²⁺ in the opposite direction in two consecutive, yet separate steps (see Ref. 28). The Na⁺/Ca²⁺ exchanger, together with Ca²⁺-ATPase of the ER/SR, regulates Ca²⁺ levels that underlie muscle contractility behavior under both normal and ischemic conditions (see Ref. 13). In cardiac muscle the Na⁺/Ca²⁺ exchanger contributes to force development, in particular, under glycosidic conditions (see Ref. 131). In smooth muscle, relaxation is achieved partially via a decrease in [Ca²⁺]_cyt efflux at the plasmalemma by means of the Na⁺/Ca²⁺ exchanger (159). There is evidence suggesting that this exchanger is a tetramer linked by disulfide bonds, and thus it is susceptible to modification by oxidizing and reducing agents as well as ROS (see Refs. 19, 88, 129). However, both decreases (24, 34, 88) and increases in Na⁺-dependent Ca²⁺ uptake (Na⁺/Ca²⁺ exchanger) (49, 135, 154) in both isolated and intact sarcolemmal vesicles have been reported (see Table 9). The exchanger is also inhibited by the oxidizing agent HOCl (46, 88), the SH-alkylating agent diamide (2, 24, 129), and SH-reducing agents GSH and DTT (131). There is also evidence that SH-alkylating diamide stimulates Ca²⁺/Na⁺ exchange (129). The nature of the inhibition or stimulation is not clear. The stimulation of the Na⁺/Ca²⁺ exchanger has been attributed to the increase in affinity to Ca²⁺, i.e., a decrease in K_m for Ca²⁺ (135, 154) with no changes in voltage dependency (154). The pathophysiological implication is that such stimulation of Na⁺/Ca²⁺ exchange by ROS may moderate the myocardial response to ischemia-reperfusion injury (154). DiPolo and Beauge, (33) proposed that the inhibition is due to a reduction in the affinity of the exchanger to Ca²⁺. The conflicting effects of ROS on the Na⁺/Ca²⁺ exchanger may be partially due to the use of a different ROS-generating system and different parameters to deduce the exchanger activity (see Table 9). For example, Kato and Kako (88) found that HOCl induced inhibition whereas H₂O₂ induced stimulation of the Na⁺/Ca²⁺ exchanger. H₂O₂-induced Cl current is used as an indicator of enhancement in the Na⁺/Ca²⁺ exchanger (149). However, both Coetzee et al. (24) and Goldhaber (49) obtained conflicting data, despite the fact that both used Ni²⁺ sensitivity of a membrane current as a marker for the Na⁺/Ca²⁺ exchanger. The conflicting effects of ROS on this exchanger are not due to the ROS-generating system, since it has been found that H₂O₂ and X/XO similarly enhanced this exchanger in ventricular myocytes, causing Ca²⁺ overload and triggering arrhythmia during reperfusion (49). The conflicting effects may be due to differences in the exchanger mode during which the effects of ROS were examined. There is evidence suggesting that Ca²⁺ and Na⁺ are translocated in separate consecutive steps (see Ref. 89) and that the Na⁺/Ca²⁺ exchanger may operate in reverse, i.e., efflux of Na⁺ and influx of Ca²⁺ during ischemia-reperfusion when cytoplasmic concentration of Na⁺ is increased (54). Furthermore, the exchanger is modulated by Ca²⁺ and/or ATP, which affect the exchange distribution between the active state and either of two inactive states (see Refs. 66 and 67). Other regulatory mechanisms may be involved, such as changes in lipid composition (106) and H₂O₂ production via insulin-NADPH oxidase interaction (88).

Na⁺/H⁺ exchange. The isoforms of the Na⁺/H⁺ exchanger are present in various epithelial and muscle cells. They play important physiological roles, such as regulation of intracellular pH, cell volume, and reabsorption of NaCl and NaHCO₃. There is little information on the effects of ROS on these exchangers in epithelial cells. ROS have been implicated in the increased activity of the cellular Na⁺/H⁺ exchanger that is activated by phosphorylation in vascular myocytes from hypertensive rats (156). It has also been reported that exposure of human neutrophils to 100 nM N-formyl-methionyl-leucyl-phenylalanine activated the amiloride-sensitive Na⁺/H⁺ exchanger, leading to an increase in intracellular pH from 7.22 to 7.8 (157). The ROS-generating system X/XO inhibited this transport system in isolated myocytes of rat heart and in sealed sarcolemmal vesicles of bovine heart (184), and the inhibition was reversed with catalase and SOD and, therefore, indicated that H₂O₂ and O₂ were the effective moieties. The effect of ROS on the Cl/HCO₃ exchanger is unknown.

Ion Cotransporters

Cation-Cl transporters. The electroneutral transporters have important physiological roles, such as regulatory volume decrease and transepithelial salt transport (see Ref. 120). Their activity depends on the presence of all the transported ions. However, they differ pharmacologically with respect to the identity and stoichiometry of the transported ions. There is little direct or indirect information on the effect of ROS on these transporters.

K⁺-CL COTRANSPORT.

NA⁺-K⁺-CL COTRANSPORT.

Other cotransporters. NA⁺-P_I COTRANSPORT. H₂O₂ and O₂ inhibited the Na⁺-P_i transport system in isolated myocytes of rat heart and in sealed sarcolemmal vesicles of bovine heart (184). Furthermore, this ROS-induced inhibition was reversed with catalase and SOD. The effect of ROS on Na⁺-HCO₃ and K⁺-HCO₃ cotransporters is unknown.

	THE PRIMARY TRANSPORT PATHWAY AS A TARGET FOR ROS

ROS-induced changes in membrane properties are considered early events in response to oxidative stress. However, the molecular mechanism(s) for ROS action on ion transport pathways is not known. Hypothetically, the effects of ROS can be caused via direct effects on ion transport proteins. Ion channels that have been thought to be a prime ROS target include a 106-kDa Ca²⁺-release channel (162, 185), DHPR and RyR Ca²⁺-release channels (125), and K⁺ channels (94, 104). It has been reported that the direct effect of H₂O₂ on K_ATP channels in skeletal muscle is mediated via oxidation of the channel protein (179). It has to be noted that the concentration of H₂O₂ used by Weik and Neumcke (179) greatly exceeded those reported in studies where the effect was thought to be indirect (50, 123). Tokube et al. (169) suggested that ROS directly affected the K_ATP channel by binding to the ATP-binding site, causing a decrease in the sensitivity of the channel to ATP in the range of 0.2-2 mM, without affecting ADP or glibenclamide binding sites. Indirect effects of ROS on ion transport pathways are mediated via membrane phospholipids. There are several examples where changes in ion transport have been attributed primarily to changes in membrane phospholipids. It has been argued that ROS caused peroxidation of membrane phospholipids and that this led to changes in the K_ATP channel (63) and the Ca²⁺-Mg²⁺-ATPase (see Table 7).

The data in Tables 1-9 show that the concentrations of ROS-induced changes are different for ion channels, pumps, and exchangers. It appears that the inhibitory concentrations for ion pumps are less than those required for ion channel inhibition. Thus ion pumps are more sensitive to ROS than ion channels. However, it is not known which of the ion pumps is the primary target. Attempts have been made to determine the primary ion pathway that is affected by ROS from the IC₅₀ of individual ROS. The data in Tables 7 and 8 show that Ca²⁺ uptake is more sensitive to H₂O₂ than ouabain-sensitive Rb⁺ uptake. However, Rb⁺ uptake is more sensitive to O₂ than Ca²⁺ uptake (42). These findings suggest that the primary ion pump that is affected by ROS depends not only on the type of pump but also on the individual ROS.

	MECHANISMS OF ROS-INDUCED MODIFICATIONS

Figure 3 shows the possible molecular targets underlying ROS effects on ion transport mechanisms. These molecular targets include 1) membrane phospholipids, 2) membrane proteins, 3) regulators of ion transport mechanisms, or 4) a combination of these targets.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Molecular targets underlying ROS-induced malfunction of transport mechanisms.

Oxidation of SH Groups

ROS characteristics. ROS are capable of reaching SH groups embedded in the membrane. For example, H₂O₂ can readily cross the cell membrane and be converted to · OH via the Fenton reaction, with consequent oxidizing of the SH groups. This SH oxidation produces intermolecular cross-links that underlie ROS-induced protein oligomers (70, 80, 88). The physical changes in the structure of the channel and pump proteins modify the function of the transporting proteins and/or the availability of regulatory sites on these proteins. It has been proposed that H₂O₂ modifies the redox state of the channel protein in such a way that oxidation of the cysteines involved in the "ball" and "chain" mechanism that gate the channel occurs (173). Jamme et al. (80) suggested that, during exposure to ROS-generating systems, the Na⁺-K⁺-ATPase forms cross-links without the isoforms of the 90-kDa -catalytic subunit and thus modifies the affinity and accessibility to the regulatory sites on this pump. Physical changes that modify ion transport mechanisms could also be brought about via ROS-induced changes in the properties of the phospholipids. The affinity or the accessibility of the ATP and ouabain binding sites could be modified by alteration in membrane integrity and fluidity during ROS-induced lipid peroxidation.

ROS mimic SH-oxidizing agents. It has been found that the SH-oxidizing agents H₂O₂ or DTNB prevent Ag⁺ contractions and Ag⁺ inhibition of E-C coupling in single skeletal muscle fibers from R. temporaria or R. catesbeiana and that these effects were reversible with the SH-reducing agents (125). The RyR Ca²⁺-release channels can also be enhanced by ROS and SH-oxidizing agents that induce Ca²⁺ release (Table 1 and Refs. 76, 145, 170, 188).

SH-reducing agents reverse ROS action. ROS-induced changes that have been reported to be reversed with SH-reducing agents, e.g., DTT, include 1) H₂O₂-induced increase in P_o of RyR in both cardiac and skeletal muscle (46, 124), 2) H₂O₂-induced decline in activity of Ca²⁺-activated K⁺ channels (18), 3) H₂O₂-induced depression in the Ca²⁺ pump (86, 87), 4) UV-C-generated · OH and peroxyl (ROO ·)- or H₂O₂-induced inhibition of Na⁺-K⁺-ATPase (80, 85), and 5) H₂O₂-induced inhibition of the Na⁺/Ca²⁺ exchanger (88). Cysteine block of ROS-induced inhibition of the SR Ca²⁺ pump also suggests the involvement of SH groups (164). In addition, ROS-induced mechanical dysfunction, due to impairment of Ca²⁺-ATPase, is prevented by SH-reducing DTT (38, 39, 46). It is assumed that SH-modifying agents act on ROS-induced disulfide by dissociating the H₂O₂-induced disulfide-linked RyR protein complex (45). However, the possibility that DTT has its own effect cannot be ruled out. Cai and Sauvé (18) suggested that oxidation of K_Ca channels by H₂O₂ forms disulfide bonds that differ from those induced by SH oxidation with DTNB and thimerosal.

Localization of the SH groups for ROS action. Localization of the SH groups on which ROS action occurs is achieved by using SH-modifying agents that differ in their pharmacological properties. The studies in which the poorly membrane-permeable thimerosal and the charged DTNB oxidizing agents were used suggest that H₂O₂ inhibits K_Ca channels by interacting with SH groups that are localized on the cytoplasmic side of the channel (see Ref. 18). On the other hand, rose bengal, a ROS-generating system that reverses the blocking effect of ryanodine (see Ref. 185), has an action that suggests a competition between ROS and ryanodine on a binding site that contains some SH groups. N-(7-dimethylamino-4-methyl-3-coumarinyl)maleimide labeling of cysteine indicates that this binding site (its oxidized SH groups keep the RyR in the active state) is embedded in the membrane away from the cytoplasmic side of the membrane (124). It is not known whether such differences in the localization of SH groups, cytoplasmic vs. internal, may account for differences in the proposed mechanisms of ROS action. The differences in the sensitivity of ion transport pathways to ROS-generating systems may be due to preferential binding of ROS to the SH groups of amino acids in the transport proteins (61). There is also evidence that indicates the presence of different sites underlying ROS-induced modifications in ion transport pathways. The opposite effects of H₂O₂ (inhibition) and DTDP (activation) on the gating of the SCl channel suggest that these oxidizing agents have different binding sites on the channel protein. Krippeit-Drews et al. (100, 101) reported that DTNB inhibited both Ca²⁺ and K_ATP currents, whereas H₂O₂ had no effect on the Ca²⁺ current while it enhanced the K_ATP current. These data point to the presence of another mechanism, other than SH oxidization, that may also be responsible for modulating ion channels. The presence of such different mechanisms may explain the opposite effects of H₂O₂ (12) and of ¹O₂ and O₂ radicals (69) observed on the RyR Ca²⁺-release channel. There is also evidence that the SH group modulating ATP-sensitive channels may be close to the ATP binding site. ATP inhibition of the K⁺ channel prevents the irreversible inhibitor NEM from reaching critical SH groups (179). Similarly, ATP inhibition of the SCl channel prevents the oxidizing agent DTDP from activating the channel (unpublished observations).

Other SH-modulated transport proteins. Some of the ion channels that are modulated by SH reagents have also been modulated by ROS in accordance with the SH hypothesis. One would expect that all ion channels and pumps that are modulated by SH-reducing and SH-oxidizing agents would also be modulated by ROS and the oxidation-reduction state in vivo. However, it should not be assumed that ROS would act in a manner similar to SH-oxidizing agents. As indicated above, there is evidence, contrary to such similarities, pointing to different mechanisms of actions. Some of the ion channels that are modulated by SH reagents, and not yet examined for ROS effects, include fast transient K⁺ (I_K(A)) channels (141), diphtheria toxin channels (116), and reduced human skeletal macroscopic Cl current (hClC-1) (105).

Changes in Ca²⁺ Homeostasis

Lipid Peroxidation