Role of γ-glutamyl transpeptidase in redox regulation of K⁺ channel remodeling in postmyocardial infarction rat hearts

Abstract

γ-Glutamyl transpeptidase (γ-GT) is a key enzyme in GSH metabolism that regulates intracellular GSH levels in response to extracellular GSH (GSH_o). The objective of this study was to identify the role of γ-GT in reversing pathogenic K⁺ channel remodeling in the diseased heart. Chronic ventricular dysfunction was induced in rats by myocardial infarction (MI), and studies were done after 6–8 wk. Biochemical assays of tissue extracts from post-MI hearts revealed significant increases in γ-GT activity in left ventricle (47%) and septum (28%) compared with sham hearts, which paralleled increases in protein abundance and mRNA. Voltage-clamp studies of isolated left ventricular myocytes from post-MI hearts showed that downregulation of transient outward K⁺ current (I_to) was reversed after 4–5 h by 10 mmol/l GSH_o or N-acetylcysteine (NAC_o), and that the effect of GSH_o but not NAC_o was blocked by the γ-GT inhibitors, acivicin or S-hexyl-GSH. Inhibition of γ-glutamylcysteine synthetase by buthionine sulfoximine did not prevent upregulation of I_to by GSH_o, suggesting that intracellular synthesis of GSH was not directly involved. However, pretreatment of post-MI myocytes with an SOD mimetic [manganese (III) tetrapyridylporphyrin] and catalase completely blocked recovery of I_to by GSH_o. Confocal microscopy using the fluorogenic dye 2′,7′-dichlorodihydrofluorescein diacetate confirmed that GSH_o increased reactive oxygen species (ROS) generation by post-MI myocytes and to a lesser extent in myocytes from sham hearts. Furthermore, GSH_o-mediated upregulation of I_to was blocked by inhibitors of tyrosine kinase (genistein, lavendustin A, and AG1024) and thioredoxin reductase (auranofin and 13-cis-retinoic acid). These data suggest that GSH_o elicits γ-GT- and ROS-dependent transactivation of tyrosine kinase signaling that upregulates K⁺ channel activity or expression via redox-mediated mechanisms. The signaling events stimulated by γ-GT catalysis of GSH_o may be a therapeutic target to reverse pathogenic electrical remodeling of the failing heart.

chronic myocardial infarction (MI)-induced ventricular dysfunction elicits a pathogenic process of electrical remodeling that is characterized at the myocyte level by downregulation of K⁺ channel expression. This change in electrophysiological phenotype is proposed to contribute to arrhythmogenic abnormalities in repolarization (5) and cellular alterations in Ca²⁺ handling that impact contractile function (45). Recent studies from our laboratory suggest that redox-mediated mechanisms underlie K⁺ channel remodeling by oxidative stress in the rat heart with chronic MI (36, 37). It is proposed that redox control of K⁺ channels in heart involves endogenous thiol oxidoreductase systems, particularly 1) the thioredoxin (Trx) system, which operates with thioredoxin reductase (TrxR) and NADPH to catalyze the reduction of protein disulfides (20, 30, 32, 40, 47), and 2) the glutaredoxin (Grx) system, which uses reduced glutathione (GSH), glutathione reductase (GR) and NADPH to reduce protein-mixed disulfides (12, 13, 30). These oxidoreductase systems act as thiol repair mechanisms to maintain cell proteins in a reduced state, thus providing cellular protection and control of the physiological function of proteins against oxidative damage. The potential for redox mechanisms to control K⁺ channel activity was recently demonstrated in studies from our laboratory showing that the decreased density of K⁺ channels carrying the transient outward current (I_to) in rats with chronic MI (36, 37) or diabetes (46) is normalized by treatment with reduced extracellular GSH (GSH_o) or N-acetylcysteine (NAC_o), a putative GSH precursor. These data suggest that I_to channel upregulation by GSH_o involves oxidoreductase systems that may represent a novel approach to reverse electrical remodeling (25–28). However, the mechanisms by which GSH_o normalizes the density of I_to from post-MI myocytes are not clear.

It is well established that the ubiquitous tripeptide GSH (Glu-Cys-Gly) plays an important role in the modulation of protein thiols in physiological and pathophysiological states (13, 16). Intracellular GSH biosynthesis is controlled by the γ-glutamyl cycle, which is important for maintaining GSH homeostasis and normal redox status (10, 17, 18). γ-Glutamyl transpeptidase (γ-GT), the only enzyme of the cycle located on the outer surface of the plasma membrane, plays a key role in GSH homeostasis by breaking down GSH_o and providing cysteine, the rate-limiting substrate for intracellular de novo synthesis of GSH (17, 31, 41, 44, 49). However, a pro-oxidant effect of GSH_o catabolism by γ-GT has also been described, which has the capability of transactivating intracellular signaling pathways (31). Although serum γ-GT is mainly seen as a metabolic indicator of liver function (23, 44), it was also observed that high γ-GT levels are associated with an increased risk of MI and cardiac death (11, 38). Nevertheless, the functional impact of membrane-bound γ-GT in regulating cardiomyocyte function remains to be clarified. Therefore, the purpose of this study was to identify the role of γ-GT in reversing pathogenic K⁺ channel remodeling in the post-MI heart in response to GSH_o. Our data show that γ-GT activity, protein abundance, and mRNA expression are increased in post-MI left ventricle, and that upregulation of K⁺ channels by GSH_o is mediated by γ-GT and by the Trx system. Moreover, the electrophysiological effect of GSH_o involves reactive oxygen species (ROS)-dependent transactivation of tyrosine kinase signaling that regulates K⁺ channel activity or expression.

METHODS

Post-MI model and isolation of ventricular myocytes.

All animal procedures were approved by University of Nebraska Medical Center Institutional Animal Care and Use Committee, and conducted according to the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health (NIH Publication No. 85-23, Revised 1996).

A chronic post-MI model of ventricular dysfunction was used in the present investigation, as described previously (26, 36, 37, 48). Briefly, male Sprague-Dawley rats (180–200 g) were intubated and artificially ventilated under methohexital sodium (Brevital) anesthesia at 50 mg/kg ip. A left thoracotomy was performed, and the left coronary artery was ligated by a suture positioned between the pulmonary artery outflow tract and the left atrium. The thorax was closed and the rats were allowed to recover for 6–8 wk before experimentation. This ligation protocol produces infarcts of 30–40% of the left ventricular free wall and physiological signs of heart failure after several weeks (26, 36, 37). Sham-operated animals that served as controls underwent the same surgical procedure but were not subjected to coronary artery ligation.

Six to eight weeks after MI or sham operation, rats were given an overdose of pentobarbital sodium (100 mg/kg ip), and the hearts were excised to obtain tissue samples or to isolate ventricular myocytes. For the latter, myocytes were dissociated from Langendorff-perfused hearts by a collagenase digestion procedure described previously (26, 36, 37). Dispersed myocytes from surviving regions of the left ventricle (LV) and septum (SP) were suspended in DMEM and stored in an incubator at 37°C until used, usually within 6 h of isolation. In some experiments, right ventricular (RV) myocytes were used for voltage-clamp recordings. For the study of ionic currents (see below), aliquots of myocytes were transferred to a cell chamber on the stage of an inverted microscope and superfused with a solution containing (in mmol/l) 138 NaCl, 4.0 KCl, 1.2 MgCl₂, 1.8 CaCl₂, 18 glucose, 5 HEPES, and 0.5 CdCl₂ (to block Ca²⁺ channels), pH 7.4. Unless stated otherwise, all chemical reagents used in these studies were purchased from Sigma-Aldrich (St. Louis, MO).

Recording techniques.

Ionic currents were recorded using the whole cell configuration of the patch-clamp technique. Briefly, borosilicate glass capillaries were pulled (Sutter Instruments, Novato, CA) to an internal tip diameter of 1 to 2 μm and filled with a pipette solution containing (in mmol/l) 135 KCl, 3 MgCl₂, 10 HEPES, 3 Na₂-ATP, 10 EGTA, and 0.5 Na-GTP, pH 7.2. Filled pipettes with a resistance of 2–4 MΩ were coupled to a patch-clamp amplifier (Axopatch 200B, Molecular Devices, Sunnyvale, CA). After correction of the liquid junction potential and creation of a GΩ seal, the membrane within the pipette was ruptured, and at least 5 min were allowed for the contents of the pipette and cytoplasm to equilibrate. A computer program (pClamp, Molecular Devices) controlled command potentials and acquired current signals that were filtered at 2 kHz. Currents were sampled at 4 kHz by a 12-bit resolution analog-to-digital converter and stored on the hard disk of a computer. All electrophysiological experiments were done at room temperature (22–24°C).

The I_to was evoked by 500-ms depolarizing pulses to test potentials between −40 and +60 mV (0.2 Hz). The holding potential was set at −80 mV, and a 100-ms prepulse was applied to −60 mV to inactivate the fast Na⁺ current. For each test pulse, I_to amplitude was measured as the difference between the peak outward current and the steady-state current level at the end of the depolarizing pulse. All electrophysiological data were normalized as current densities by dividing measured current amplitude by whole cell capacitance.

Enzyme assays.

γ-GT activity was measured by standard spectrophotometric techniques (41). Briefly, tissue samples were homogenized in ice-cold Tris buffer (0.1 mol/l, pH 8.0, with 2 mmol/l EDTA) and centrifuged at 4°C (6,000 g) for 1 h. Purified membrane proteins were harvested, and the protein concentration was assayed by a kit (Pierce, Rockford, IL). Five microliters (10 μg) protein was added to a cuvette containing reaction buffer of the following composition: 0.6 ml Tris·HCl (0.1 mol/l, pH 8.0), 0.2 ml l-γ-glutamyl-p-nitroanilide (5 mmol/l, pH 8.0), and 0.2 ml glycylglycine (0.1 mmol/l, pH 8.0). The solution was allowed to reach 37°C in a spectrophotometer equipped with a temperature-controlled cuvette holder. The reaction was initiated by adding 5 μl purified membrane protein, and the rate of release of p-nitroaniline was recorded at 410 nm. Activity was expressed in milliunits per milligram protein and defined as the amount of enzyme catalyzing the reduction of 1 micromole p-nitroaniline per minute.

Western blot analysis.

Proteins from tissue samples were immunoblotted for γ-GT following standard protocols. Briefly, samples were prepared by homogenizing tissues in TE buffer (10 mmol/l Tris-glycerol-buffer, 1 μg/ml aprotinin, and 2 mmol/l EDTA, pH 7.0) with 1% Triton X-100. The protein concentration was assayed by a kit (Pierce), and equal amounts of protein from all samples were resolved in 10% polyacrylamide gel, transferred to polyvinylidene difluoride membrane, and blocked with 5% milk in Tris-buffered saline, pH 7.0, containing 0.05% Tween 20. The membranes were probed with primary antibodies against γ-GT (1:1,000 dilutions; sc-20639, Santa Cruz Biotechnology, Santa Cruz, CA) and secondary antibody of goat anti-rabbit IgG (1:5,000 dilutions, Pierce Chemical, Rockford, IL). The protein signals were detected by enhanced chemiluminescence reagent (Pierce Chemical) and analyzed using UVP BioImaging Systems (Upland, CA). The levels of target proteins were normalized with original total purified membrane proteins as a loading control.

Real-time PCR.

For real-time PCR, total RNA was extracted from tissue samples using an RNAeasy mini kit (Qiagen, Valencia, CA) following the manufacturer's protocol. cDNA was generated from 1 μg total RNA using random hexamer primers (TaqMan reverse transcription reagents), and 1 μl of the resulting cDNA was PCR amplified for 45 cycles using TaqMan gene expression assays for γ-GT and GAPDH. The primer sets were as follows: γ-GT, forward 5′-ACCACTCACCCAACCGCCTAC-3′, reverse, 5′-ATCCGAACCTTGCCGTCCTT-3′; GAPDH, forward 5′-ACCCCCAATGTATCCGTTGT-3′, reverse 5′-TACTCCTTGGAGGCCATGT-3′ (49). SYBR green qPCR master mix was purchased from Applied Biosystems (Foster City, CA). Real-time PCR was performed on duplicate samples using an Applied Biosystems 7500 Real-Time PCR System. For quantification, the target gene expression relative to GAPDH was determined, and data were expressed relative to the sham-operated control samples using the equation: R = 2^−ΔΔC, where C_t is the number of cycles needed to achieve a preset threshold value of fluorescence.

Confocal microscopy.

Cardiac myocytes isolated from sham or post-MI rats were loaded with the H₂O₂-sensitive fluorogenic dye 2′,7′-dichlorodihydrofluorescein diacetate (DCFH, 10 μmol/l, Invitrogen) for 30 min. DCFH fluorescence was visualized by confocal microscopy (Zeiss LSM 510) using an excitation wavelength of 543 nm and a rhodamine emission filter. Baseline detector and laser settings were kept constant across all samples within individual experiments, and sham and post-MI samples were processed in parallel. When adjustments in baseline fluorescence were necessary, the confocal detector gain was changed by the same amount for control and experimental samples. Cells were then stimulated with GSH (10 mmol/l) for 20 min, and fluorescence images were recaptured.

Statistical analysis.

All results are expressed as means ± SE. Statistical comparisons of two groups were made using a Student's t-test, whereas comparisons of more than two groups were made by ANOVA. When a significant difference between groups was indicated by the initial analysis, individual paired comparisons were made using a Student-Newman-Keuls t-test. Differences were considered significant at P < 0.05.

RESULTS

Role of γ-GT in I_to response to GSH_o.

We and others have shown in myocytes from post-MI rats that K⁺ channel remodeling is characterized by markedly decreased I_to density (5, 26, 33, 36, 37, 48) and downregulation of K_v4.2, K_v4.3, and K⁺ channel interacting protein-2 (KChIP2) protein expression (5, 26). In this experimental model we also found that I_to density can be normalized by treatment with GSH_o, which implicates redox modulation of electrophysiological phenotype (36). In the present study, we further explored the mechanisms by which GSH_o reverses K⁺ channel remodeling using an experimental protocol in which myocytes from post-MI and sham hearts were pretreated for 4–5 h with 10 mmol/l GSH_o. We determined that this was the minimal concentration of GSH_o that normalized I_to density by 5 h, based on concentration-response curves (0.1, 1, 3, 10, and 30 mmol/l) which yielded an EC₅₀ of 1.7 ± 0.3 mmol/l. Moreover, we found that the electrophysiological effect of GSH_o in myocytes from post-MI hearts was blocked by pretreatment with one of two specific inhibitors of γ-GT: acivicin (10 μmol/l) or S-hexyl-GSH (100 μmol/l). This is illustrated in Fig. 1A, which compares raw current traces from an untreated post-MI myocyte, one treated with 10 mmol/l GSH_o, and another treated with acivicin plus GSH^o. The mean current-voltage relations from several myocytes in each group are summarized in Fig. 1B, which show that acivicin completely blocked normalization of I_to density by GSH_o. A similar effect was observed with S-hexyl-GSH (Fig. 1C), whereas neither blocker alone significantly altered maximum I_to density (+60 mV) in myocytes treated for 4–5 h (Fig. 1, B and C). Moreover, the putative GSH precursor N-acetylcysteine (NAC_o; 10 mmol/l) also upregulated I_to density to normal levels in post-MI myocytes (36), but in contrast to GSH_o, acivicin did not block I_to upregulation by NAC_o (Fig. 1C). These data suggest that γ-GT plays a key role in rescuing K⁺ channel function in post-MI myocytes treated with GSH_o.

Fig. 1.

Effect of γ-glutamyl transpeptidase (γ-GT) inhibitors on outward K⁺ current (I_to) response to extracellular GSH (GSH_o). A: superimposed current traces recorded by voltage-clamp depolarizations from −40 to +60 mV are shown for myocytes from postmyocardial infarction (post-MI) hearts treated for 4 h with 10 mmol/l GSH, or pretreated for 1 h with the γ-GT inhibitor acivicin (10 μmol/l) with GSH for an additional 4 h. B: mean current-voltage relations for I_to of post-MI myocytes untreated and treated with GSH or acivicin + GSH are shown. V_m, membrane potential. C: maximum I_to density at +60 mV is compared in indicated experimental groups. S-hexyl-GSH (100 μmol/l), a second γ-GT inhibitor, was tested as with acivicin. N-acetylcysteine (NAC) was used at 10 mmol/l. *P < 0.05 compared with untreated group.

γ-GT in post-MI heart.

To further characterize the status of cardiac γ-GT after MI, tissue samples from LV, SP, and RV were assayed for γ-GT activity (41). Figure 2A shows that, compared with sham control (open bars), γ-GT activity in post-MI hearts was significantly increased by 47% and 28% in LV and SP, respectively, but decreased by 25% in RV. As a result, cardiac γ-GT activity was uniform across all regions of the post-MI hearts, which was different from sham hearts that showed a clear LV-to-RV gradient. To explore the molecular basis for changes in cardiac γ-GT activity in greater detail, Western blotting and real-time PCR analyses were done in tissue extracts. Figure 2B compares immunoblots of γ-GT protein in samples from LV, SP, and RV of post-MI and sham hearts. In agreement with measured activities (Fig. 2A), the abundance of γ-GT protein in LV (72% increase) and SP (40% increase) of post-MI hearts was significantly greater than sham control, whereas γ-GT protein was less in RV samples (23% decrease). Moreover, post-MI changes in mRNA expression of γ-GT measured by real-time PCR paralleled those observed in protein abundance. Thus, γ-GT mRNA levels were significantly increased in LV and SP from post-MI hearts while transcripts in RV from post-MI hearts were slightly but not significantly less than sham control (Fig. 2C).

Fig. 2.

Region-specific changes in γ-GT and I_to response to GSH_o. A: basal activity of γ-GT in tissue samples from left ventricle (LV), septum (SP,) and right ventricle (RV) in sham and post-MI rats. B: Western blot analysis of γ-GT protein levels in sham and post-MI rats. Purified membrane protein (25 μg) was loaded in each lane. C: mRNA expression of γ-GT relative to sham was measured by real-time PCR. In A–C, *P < 0.05 compared with sham and ^#P < 0.05 compared with sham LV. D: electrophysiological response of RV myocytes from post-MI hearts treated with 10 mmol/l GSH_o for 4–5 h. *P < 0.05 compared with sham and untreated post-MI myocytes. Numbers in parentheses in A–C represent number of hearts sampled. In D and all other figures, numbers in parentheses represent number of myocytes studied.

Since post-MI changes in activity and expression of γ-GT in RV were opposite to those observed in the LV and SP, we also examined the electrophysiological response of isolated RV myocytes to GSH_o. Figure 2D shows that GSH_o significantly increased I_to density in RV myocytes from post-MI hearts but not to control levels. Hence, maximum I_to density in GSH_o-treated cells (21.7 ± 2.7 pA/pF) was only 65% of sham controls (33.6 ± 3.0 pA/pF; P < 0.05). By contrast, GSH_o increased maximum I_to density in LV and SP myocytes from post-MI hearts to control levels (post-MI, 27.4 ± 2.8 pA/pF; sham, 28.3 ± 2.8 pA/pF; P > 0.05).

Signaling mechanisms involved in K⁺ channel response to GSH_o.

It has previously been reported that depleted levels of intracellular GSH and abnormal cell function can be reversed by supraphysiological concentrations of GSH_o (25, 36, 46). It is proposed that this effect is mediated by γ-GT, which degrades GSH_o to its constitutive amino acids that are then taken up by cells and resynthesized to intracellular GSH by γ-glutamylcysteine synthetase (18, 31). Since intracellular GSH levels are depleted in myocytes from post-MI hearts (36), studies were conducted to determine whether upregulation of I_to density by GSH_o was mediated by de novo intracellular GSH synthesis. Thus, myocytes from post-MI hearts were pretreated with the γ-glutamylcysteine synthetase inhibitor buthionine sulfoximine (BSO) for 18 h after which GSH_o (10 mmol/l) was added for an additional 4–5 h. We previously reported that BSO did not alter basal I_to density in post-MI myocytes (36), and as shown in Fig. 3A from the present study, BSO (10 and 100 μmol/l) also did not significantly affect upregulation of I_to by GSH_o, suggesting that intracellular de novo synthesis of GSH was not involved in the electrophysiological response. In related experiments, we tested the effects of two GSH precursors: l-2-oxothiazolidine-4-carboxylate (OTC), which supplies cysteine after enzymatic degradation by 5-oxo-l-prolinase, and glutathione monoethyl ester (GME), which enters the cell and releases GSH after esterase hydrolysis (3). However, neither OTC (10 mmol/l) nor GME (10 mmol/l) increased I_to density in myocytes from post-MI hearts when treated for 4–5 h (Fig. 3B).

Fig. 3.

Effect of buthionine sulfoximine (BSO) and GSH precursors on I_to density. A: myocytes from post-MI hearts were pretreated for 18 h with BSO (10, 100 μmol/l) to inhibit γ-glutamylcysteine synthetase and intracellular GSH synthesis, before addition of 10 mmol/l GSH_o for an additional 4–5 h. Time-matched data from untreated myocytes are also shown. *P < 0.05 compared with BSO + GSH_o-treated groups. B: myocytes from post-MI hearts were treated for 4–5 h with one of two GSH precursors: l-2-oxothiazolidine-4-carboxylate (OTC; 10 mmol/l) or glutathione monoethyl ester (GME; 10 mmol/l).

Recent studies have shown that γ-GT catalysis of GSH_o can produce ROS in amounts sufficient to influence signaling cascades (31), and thus we hypothesized that ROS generated from GSH_o breakdown underlies its electrophysiological effects. In support of this hypothesis, Fig. 4A shows that upregulation of I_to density by GSH_o was inhibited by pretreating myocytes for 30 min with manganese (III) tetrapyridylporphyrin (MnTPyP, 100 μmol/l), a membrane-permeable superoxide scavenger and SOD mimetic. A similar response was observed with catalase (400 U/ml) while neither compound alone affected basal I_to density (MnTPyP-treated, 17.7 ± 1.5 pA/pF, n = 6; catalase-treated, 15.9 ± 1.6 pA/pF, n = 4, untreated, 15.6 ± 2.4 pA/pF, n = 15; P > 0.05). When MnTPyP and catalase were combined, I_to density in the presence of GSH_o was closer to that measured in the untreated group (MnTPyP + catalase + GSH_o, 16.7 ± 1.5 pA/pF, n = 14; untreated, 15.6 ± 2.4 pA/pF, n = 15; P > 0.05) when compared with each antioxidant tested individually. We also examined cellular H₂O₂ production by GSH_o using confocal microscopy and the fluorescent probe DCFH. As shown in Fig. 4B (top), isolated myocytes from post-MI hearts were generally larger than in sham controls, which is consistent with compensatory hypertrophy in this experimental model (36). Subsequent panels in Fig. 4B show confocal images of the same myocytes after loading with DCFH (10 mmol/l) for 30 min. To compare the basal levels of H₂O₂ in sham versus post-MI myocytes, fluorescence images were captured (baseline) using identical confocal laser settings. These data show that cells isolated from post-MI rats exhibited a marked increase in baseline DCFH fluorescence, indicating an increase in H₂O₂ levels. To be able to detect GSH_o-induced changes in fluorescence in post-MI cells, the baseline fluorescence intensity was manually adjusted (adjusted baseline) by decreasing the confocal detector gain by 25%. After making this adjustment in sham and post-MI myocytes, the cells were exposed to GSH_o (10 mmol/l, 20 min) and confocal images were captured using the adjusted confocal settings. The bottom panels in Fig. 4B show that GSH_o elicited a greater increase in fluorescence in the post-MI myocyte compared with sham control and that this effect was blocked by the γ-GT inhibitor acivicin. Similar findings were observed in three other experiments and suggest that the increased activity of γ-GT in post-MI hearts results in greater ROS production in the presence of GSH_o.

Fig. 4.

Role of GSH_o-mediated H₂O₂ generation. A: maximum I_to density at +60 mV of myocytes from post-MI hearts treated with GSH_o is shown in the presence of 100 μmol/l manganese (III) tetrapyridylporphyrin (MnTPyP) or 400 U/ml catalase. *P < 0.05 compared with untreated group, ^#P < 0.05 compared with catalase + GSH group. Mean data from untreated and GSH_o-treated myocytes were replotted from Fig. 1C. B: GSH_o increased H₂O₂ levels in myocytes from post-MI hearts and to a lesser extent in sham cells. Representative confocal microscopy images showing 2′,7′-dichlorodihydrofluorescein diacetate (DCFH) fluorescence in ventricular myocytes isolated from sham or post-MI rats. Right-hand images were taken from a post-MI myocyte treated with 10 μmol/l acivicin. Fluorescence images were captured before (adjusted baseline) and 20 min after GSH_o treatment. All images were collected using the same magnification. Magnification bar equals 20 μm.

It is possible that receptor-mediated transduction events and exogenous factors forming ROS share common signaling mechanisms that control intracellular kinase activity. In this regard, it has been shown that insulin signaling in adipocytes involves the generation of H₂O₂ which inhibits tyrosine phosphatase (29). Moreover, it is known that activation of several kinase pathways, such as tyrosine kinase, can modulate K⁺ channels in heart (14, 28). Thus, to determine whether K⁺ channel current normalization by GSH_o was mediated by a phosphorylation mechanism, the GSH_o protocol was repeated after pretreating myocytes from post-MI hearts for 30 min with one of two pan-specific inhibitors of tyrosine kinase: genistein (1 μmol/l) or lavendustin A (10 μmol/l). We also examined the effect of AG1024 (100 nmol/l), a specific inhibitor of insulin receptor tyrosine kinase (9). As shown in Fig. 5, both genistein and lavendustin A blocked the effect of GSH_o to upregulate I_to density in myocytes from post-MI hearts. Control experiments also showed that neither blocker alone significantly altered basal I_to density (genistein-treated, 15.2 ± 1.8 pA/pF, n = 5; lavendustin A-treated, 14.9 ± 1.5 pA/pF, n = 5, untreated, 15.6 ± 2.4 pA/pF, n = 15; P > 0.05). In addition, the effect of GSH_o was blocked by the receptor tyrosine kinase inhibitor AG1024, which implicates this signaling cascade in the regulation of I_to channels. As with genistein and lavendustin A, AG1024 alone had no significant effect on basal I_to in post-MI myocytes (15.4 ± 2.3 pA/pF, n = 7; P > 0.05).

Fig. 5.

Upregulation of I_to density by GSH_o mediated by tyrosine kinase. Myocytes from post-MI hearts were treated for 4 h with GSH_o following incubation for 30 min with pan-specific inhibitors of tyrosine kinase, genistein (1 μmol/l), or lavendustin A (Laven A; 10 μmol/l), or a specific tyrosine kinase inhibitor for insulin receptor AG1024 (100 nmol/l). *P < 0.05 compared with untreated group. Mean data from untreated and GSH_o-treated myocytes were replotted from Fig. 1C.

Although the electrophysiological effect of GSH_o on I_to density implies direct involvement of GSH redox state, we recently reported that the Trx system plays an essential role in controlling K⁺ channel remodeling in diabetic rats (27) and rats with chronic MI (26). Thus, to determine whether the Trx system mediated I_to normalization by GSH_o in post-MI hearts, we examined the effects of two structurally different inhibitor compounds with relatively high selectivity profiles for TrxR: auranofin (AF, 10 nmol/l) and 13-cis-retinoic acid (RA, 1 μmol/l). Figure 6A illustrates that both compounds blocked upregulation of I_to density in post-MI rat myocytes by GSH_o. Neither blocker alone altered maximum I_to density in these cells when treated for 4 h (AF-treated, 16.8 ± 1.7 pA/pF, n = 5; RA-treated, 16.4 ± 2.8 pA/pF, n = 6; untreated, 15.6 ± 2.4 pA/pF, n = 15; P > 0.05) nor in sham control myocytes treated for the same duration (AF-treated, 29.7 ± 2.2 pA/pF, n = 5; RA-treated, 28.4 ± 2.7 pA/pF, n = 5; untreated, 30.1 ± 2.8 pA/pF, n = 11, P > 0.05). Moreover, the effect of GSH_o was blocked by inhibitors of transcription and translation (Fig. 6B). Specifically, in myocytes pretreated with actinomycin D (100 μmol/l) or cycloheximide (100 μmol/l) for 1 h, GSH_o failed to upregulate I_to density. These data confirm the electrophysiological role of the Trx system in controlling K⁺ channel expression and suggest a functional interaction between GSH_o catabolism and intracellular oxidoreductase activity.

Fig. 6.

Role of thioredoxin reductase (TrxR) and protein synthesis in the I_to response to GSH_o. A: mean current-voltage relations for myocytes from post-MI hearts pretreated for 30 min with 10 nmol/l auranofin (AF) or 1 μmol/l 13-cis-retinoic acid (RA) before addition of 10 mmol/l GSH_o for 4 h. *P < 0.05 compared with untreated group. B: mean current-voltage relations for post-MI myocytes pretreated for 1 h with 100 μmol/l cycloheximide or 100 μmol/l actinomycin D before adding 10 mmol/l GSH_o for 4 h. *P < 0.05 compared with untreated group. Mean data from untreated and GSH_o-treated myocytes in A and B were replotted from Fig. 1C.

DISCUSSION

Electrophysiological effects of GSH_o.

Protein thiols in the mammalian heart are mostly maintained in the reduced state under physiological conditions and protected from oxidation due to a high intracellular GSH concentration and the presence of thiol oxidoreductase systems, such as the Trx and Grx systems (10, 12, 13, 20, 30, 32, 40, 47). Under pathophysiological conditions, however, intracellular GSH levels are often depleted, which significantly affects cell function, including ion channel activity (7, 43). As a result, strategies to increase intracellular GSH levels have often been shown to restore cell function, and one such strategy is to supplement cells with exogenous GSH as we have done in the present study. Indeed, we and others have shown that treating ventricular myocytes from diabetic (39, 46) or post-MI (36) rats with GSH_o normalizes the density of I_to, an effect that requires several hours to develop. GSH depletion in canine and human atrium has also been correlated with decreased density of the L-type Ca²⁺ current (I_Ca), and this electrophysiological phenotype is reversed in vitro by GSH_o or NAC_o (7). Finally, hypoxia has been shown to modulate the transient and persistent Na⁺ current in guinea pig ventricular myocytes by a mechanism that is reversed by GSH_o (43). However, while experimental studies have demonstrated significant effects of GSH_o on ion channels, millimolar concentrations are usually necessary, which are significantly greater than what is measured in vivo. For example, GSH concentration in human plasma is ∼5 μmol/l, whereas in rat plasma it is near 15 μmol/l (1, 23). The reason for the need of supraphysiological GSH_o concentrations in in vitro studies is not clear but it may involve oxidation of a large fraction of GSH_o or lack of catalytic factors required to metabolize GSH_o (31). By accounting for these variables it is likely that the functional effects of GSH_o can be achieved with more physiological levels of this reductant.

While it is generally accepted that intracellular GSH can be influenced by GSH_o, mammalian cells do not take up intact GSH in significant amounts. Rather, it is proposed that GSH_o is first degraded by γ-GT to its constitutive amino acids (Gly, Cys, Glu), which are then taken up by cells where GSH is resynthesized in the cytoplasm by γ-glutamylcysteine synthetase (10, 18, 31). Indeed, we found in this study that the effect of GSH_o to upregulate I_to density in post-MI myocytes was mediated by γ-GT, whereas the effect of NAC_o was not (Fig. 1). However, the underlying mechanism for I_to upregulation did not involve de novo synthesis of intracellular GSH, since BSO did not block the effect of GSH_o nor did GSH precursors mimic the effects of GSH_o (Fig. 3). These latter data bring into question the mechanism for the effect of NAC_o, which is usually considered to supply intracellular cysteine after metabolism by N-deacetylase (3). Although not a main focus of our current experiments, our data suggest that NAC may have direct reducing effects on cellular proteins that are independent of intracellular GSH synthesis.

Our biochemical and molecular analyses indicated that γ-GT was significantly upregulated in the post-MI heart, although in a regionally dependent manner (Fig. 2). Increased expression of γ-GT in many cell types is elicited by factors such as inflammation and oxidative stress, which play a role in the etiology of several diseases, including cardiovascular disease (4, 17, 18, 41, 44). It is proposed that stress-induced upregulation of γ-GT activity is a cellular compensatory mechanism protecting proteins from irreversible oxidative damage, although this mechanism has not been well characterized in cardiac myocytes. In our study done 6–8 wk post-MI, total γ-GT activity (Fig. 2A) and protein abundance (Fig. 2B) in LV and SP were significantly greater than sham controls, which is consistent with (oxidative) stress-induced upregulation. In the RV, however, there was a significant decrease in γ-GT activity and protein abundance post-MI (Fig. 2B). The reason for downregulation of γ-GT in the RV is unclear, particularly since GSH depletion in this region occurs 8 wk after MI in the LV (19). It is also unclear as to the functional significance of the gradient in γ-GT activity in sham hearts (Fig. 2A, open bars). This may reflect regional differences in the sources of cysteine for GSH synthesis. Nevertheless, increased γ-GT activity in the LV and SP post-MI had a profound electrophysiological effect, albeit independent of intracellular GSH synthesis.

γ-GT-mediated ROS generation.

Our data suggest a unique mechanism for the cellular effect of GSH_o that is mediated by the generation of ROS. Several studies have postulated that the catabolism of GSH_o by γ-GT leads to the formation of superoxide anion (O₂^•−), which is facilitated by the presence of iron (31). The generated O₂^•− may directly attack cell components or be converted to the more stable H₂O₂ by spontaneous dismutation or by SOD (6). In the case of γ-GT-mediated H₂O₂ generation, it is possible that the extracellular isoform of SOD (ecSOD) was involved (6), although this mechanism requires further investigation. Nevertheless, we found that the effect of GSH_o to upregulate I_to density was blocked by the SOD mimetic MnTPyP and by extracellular catalase (Fig. 4A). Moreover, the significant blockade of the GSH_o response by catalase alone suggests that H₂O₂ was generated extracellularly and diffused into the cell. Thus, our confocal experiments showed a marked increase in intracellular DCFH fluorescence in myocytes from post-MI hearts exposed to GSH_o (Fig. 4B). Since γ-GT activity was increased post-MI in the LV and SP (Fig. 2), it follows that DCFH fluorescence in the presence of GSH_o was greater in myocytes from post-MI hearts compared with sham controls.

The functional effects of γ-GT-mediated ROS generation are not well understood and may involve redox-sensitive modulation of receptors, kinases, phosphatases, or transcription factors (31). Any of these mechanisms may have contributed to the electrophysiological effects of GSH_o in our studies, but our data identify tyrosine kinase signaling as playing a major role. In particular, blockade of the GSH_o effect in post-MI myocytes by AG1024 (Fig. 5) suggests that kinase signaling activated by the insulin receptor or related receptors plays a role in controlling the expression of K⁺ channels. In support of this hypothesis, we have previously shown that insulin upregulates I_to density in ventricular myocytes from diabetic rats (46) or rats treated with inhibitors of oxidoreductase systems (25). Whether transactivation of receptor tyrosine kinase signaling by GSH_o-mediated ROS involves redox modulation of receptors or phosphatases requires further investigation. Nevertheless, our data suggest that ROS generated by extracellular degradation of GSH_o have a significant functional effect on signal transduction in cardiac myocytes.

Functional role of the Trx system.

Previous work in diabetic and post-MI rats implicate the Trx system as a key regulator of Kv channel expression in heart, particularly the pore-forming K_v4.2 and K_v4.3 α-subunits (26). Our current experiments are in agreement with these previous studies because we found that GSH_o-mediated increase in I_to density was blocked by TrxR inhibitors (Fig. 6A). Thus, on the basis of the results of all our experiments, we propose a model, illustrated in Fig. 7, to explain the relationship of GSH_o catabolism, tyrosine kinase signaling, and oxidoreductase activity in increasing Kv channel expression post-MI. In this model, regulation of channel expression by the Trx system is postulated to occur mainly at the level of transcription, although posttranslational modifications of channel protein cannot be ruled out. Our experiments with GSH_o are consistent with a transcriptional mechanism since the increase in I_to density by GSH_o was blocked by actinomycin D (Fig. 6B) and required >4 h to fully develop (36). By comparison, the increase in ROS generation by GSH_o occurred in minutes (Fig. 4B), suggesting that stimulation of kinase signaling was rapid while subsequent transcriptional and translational events leading to increased Kv channel expression occurred with delay. Although we hypothesize that the Trx system is downstream of (receptor) tyrosine kinase signaling, it is possible that Trx controls Kv channel expression by modulating signaling intermediates. For example, Trx controls the redox state of protein tyrosine phosphatases that regulate the phosphorylation of receptor tyrosine kinases (15, 29). Trx also directly inhibits the phosphatase activity of phosphatase and tensin homolog (PTEN), which leads to activation of the phosphatidylinositol 3-kinase/Akt pathway that can impact transcription and translation (2). The signaling intermediate Ras is also regulated by Trx which may play a role in controlling the cardiac hypertrophic response to α-adrenergic agonists or growth factors (2, 22). Thus, it remains to be determined whether redox control of Kv channel expression in heart is directly or indirectly regulated by the Trx system.

Fig. 7.

Proposed mechanism of GSH_o-induced regulation of Kv channel expression in post-MI myocytes. In this model, GSH_o degradation by γ-GT generates O₂^•− and H₂O₂, which oxidize thiol side chains of receptors, kinases, or phosphatases to activate tyrosine kinase signaling. The stimulation of kinase signaling increases the reducing activity of the Trx system, which upregulates Kv channel activity/expression and I_to density. Although not directly involved in Kv channel regulation, γ-GT breaks down GSH_o to its constitutive amino acids (Glu, Cys, Gly), which are taken up and resynthesized to intracellular GSH by γ-glutamylcysteine synthetase (γ-GCS). Sites of blocking agents are also shown. Grx, glutaredoxin; GST, glutathione S-transferase; GPx, glutathione peroxidase.

Our current knowledge of the transcriptional control of ion channel remodeling in the heart suggests several candidate mechanisms that may be regulated by the Trx system, and hence sensitive to GSH_o-stimulated tyrosine kinase signaling. For example, MI-induced downregulation of K_v4.2, K_v4.3, K_v2.1, and K_v1.5 channel expression in mouse heart is mediated by activation of the Ca²⁺-dependent phosphatase calcineurin and the transcription factor nuclear factor of activated T cells c3 (NFATc3; 34). This proposed mechanism may involve increased β-adrenergic signaling which elevates intracellular Ca²⁺ concentration to activate the calcineurin/NFATc3 pathway (34, 35). Second, in neonatal rat ventricular myocytes, the transcription factor FOG2 suppresses the expression of K_v4.2 channels normally regulated by the transcription factor GATA4 (21). Third, transcription of the accessory protein KChIP2, which controls the abundance of K_v4 channels in the cell membrane, is decreased by the Iroquois transcription factor Irx5 acting with the repressor protein m-Bop (8). Finally, steady-state levels of Kv channel transcripts may be decreased by enhancing the degradation of mRNA, as has been shown for the effects of angiotensin II on K_v4.3 mRNA (50). Thus, several possibilities exist to explain how redox reactions controlled by Trx may regulate expression of Kv channels in the ventricle.

In summary, γ-GT activity is increased in the LV and SP of post-MI hearts, which correlates with increased protein abundance and mRNA expression. GSH_o elicits γ-GT- and ROS-dependent transactivation of tyrosine kinase signaling that functionally upregulates Kv channel activity/expression in post-MI myocytes via a redox mechanism involving the Trx system (see Fig. 7).

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grant HL-66446 (to G. J. Rozanski) and National Institutes of Health CoBRE Grant, Nebraska Redox Biology Center 5P20RR017675 (to M. C. Zimmerman).