INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

OUR PREVIOUS WORK in endothelium-removed bovine coronary arteries suggests that the initial response to posthypoxic reoxygenation is mediated by a mechanism shown in Fig. 1 involving an increase in NADH oxidase-derived superoxide anion (O₂ ·), which produces a transient relaxation as a result of an increase in H₂O₂ metabolism by catalase stimulating the cytosolic or soluble form of guanylate cyclase (sGC) (17). Evidence for this mechanism included observations that relaxation to reoxygenation, but not to hypoxia, was significantly attenuated by a scavenger of O₂ ·, nitroblue tetrazolium, which prevents the formation of H₂O₂; by diphenyliodonium, a flavoprotein inhibitor of NADH oxidase activity; by methanol, an agent cometabolized by catalase in a manner that prevents catalase from stimulating sGC (4, 5); and by inhibition of sGC stimulation (LY-83583). Our previous studies demonstrated that the metabolism of high picomolar to low nanomolar levels of H₂O₂ by catalase stimulates the purified form of bovine lung sGC and that this mechanism appears to participate in the relaxation of endothelium-removed bovine pulmonary arteries to micromolar levels of H₂O₂ (4, 5, 10). The requirement for micromolar levels of H₂O₂ has been suggested to originate from the high intracellular levels of peroxide metabolism by the GSH peroxidase, which is thought to result in a marked lowering of H₂O₂ concentrations in the intracellular environment (29). It has also been demonstrated that endothelium-derived H₂O₂ elicits a similar cGMP-mediated relaxation of coronary arteries (11). Because a scavenger of O₂ · that promotes H₂O₂ formation (Tiron) did not inhibit the relaxation to reoxygenation, O₂ · did not appear to directly influence the signaling mechanism mediating this response (17). Our previous studies on bovine pulmonary (19, 24) and coronary (31) arteries, human placental vessels (23), and rat skeletal muscle arterioles (8) provide evidence that these vascular segments relax to physiological levels of lactate through a mechanism that appears to involve the stimulation of sGC by H₂O₂. Our studies in bovine pulmonary arteries (19, 24) suggest that the mechanism for this response involves an increase in the production of O₂ · by NADH oxidase, presumably as a result of the increased availability of NADH derived from the lactate dehydrogenase reaction, because the other metabolic product of this reaction pyruvate does elicit a similar response and the relaxation to lactate is inhibited by many of the probes that attenuated the coronary response to reoxygenation. Thus there is evidence that H₂O₂ metabolism by catalase and cGMP participate in certain vascular responses elicited by reoxygenation and lactate.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Model showing interactions of probes for GSH peroxidase (GSH Px) with hypothesized mechanism of relaxation to both endogenous and exogenous H2O2 involving stimulation of soluble guanylate cyclase activity as a result of alterations in peroxide metabolism by catalase. ATZ, 3-amino-1,2,4-triazole; SOD, superoxide dismutase; MS, mercaptosuccinic acid.

The influence of alterations in the function of GSH peroxidase, a key pathway of H₂O₂ metabolism, and the possibility that it interacts with signaling mechanisms as a result of modulating the redox status of GSH need to be considered in studying vascular responses mediated by H₂O₂. For example, there is evidence that increases in thiol oxidation have the ability to promote vascular relaxation through the opening of potassium channels (25, 28). Studies on signaling mechanisms in a variety of cell types have identified the potential for thiol redox-mediated regulation through processes that include the control of intracellular calcium storage, membrane potential, tyrosine phosphatase activity, the activity of sGC, key aspects of energy metabolism, and contractile protein function (3, 13, 27, 30). Interestingly, S-thiolation has been reported to reversibly inhibit the activity of sGC (2). It seems that the redox status of GSH appears to be a cellular control mechanism that potentially regulates systems through its influence on thiol redox as a result of modifications, which include S-thiolation and thiol oxidation (3, 13, 27, 30). Because the metabolism of peroxides by the GSH peroxidase reaction is probably the most important system through which H₂O₂ can influence the redox status of GSH (3, 6), agents that modulate the activity of this enzyme could be helpful in identifying the role of alterations in thiol redox in responses that are elicited by H₂O₂. As predicted by the model in Fig. 1, based on the organization of processes that cells use to metabolize peroxide (6), agents that influence the activity of GSH peroxidase should have opposite effects on responses mediated through mechanisms involving the metabolism of H₂O₂ by catalase compared with responses elicited through the oxidation of GSH. For example, the inhibition of GSH peroxidase should increase responses activated by the metabolism of H₂O₂ by catalase and inhibit responses elicited by the oxidation of GSH. Thus the purpose of this study was to investigate whether modulation of GSH peroxidase activity by pretreatment of coronary arteries with agents that either utilize intracellular GSH to catalyze (mimic) the glutathione peroxidase reaction [100 µM Ebselen (20)] or inhibit the activity of endogenous GSH peroxidase [10 mM mercaptosuccinic acid (MS) (7)] causes selective alterations in relaxation responses to posthypoxic reoxygenation, lactate, and exogenous H₂O₂, which have been previously hypothesized to be mediated through changes in H₂O₂ metabolism by catalase. In some experiments, 3-amino-1,2,4-triazole was used as an agent that inactivates catalase and prevents the stimulation of sGC by H₂O₂ (4). The sites of action of the probes employed in this study to alter the metabolism of peroxide are also included in Fig. 1.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Materials. MS and diamide were purchased from Sigma (St. Louis, MO). Ebselen was purchased from Cayman Chemical (Ann Arbor, MI). Other chemicals were obtained from sources previously described (4, 17, 18).

Measurement of changes in force in bovine coronary arteries. Isolated endothelium-removed arterial rings ~4 mm in diameter and length were prepared from the left anterior descending artery and circumflex coronary artery of bovine calf hearts obtained immediately after slaughter and studied by adaptation of previously described methods (17). Briefly, arterial rings were mounted on wire hooks attached to force displacement transducers (model FT-03, Grass) for measurements of changes in isometric force on a polygraph (model 7, Grass). Arteries were incubated for 1 h at an optimal passive tension of 5 g in individually thermostated (37°C) 10-ml baths (Metro Scientific). These studies were conducted in Krebs bicarbonate buffer (pH 7.4) gassed with 21% O₂5% CO₂ (balance N₂). The Krebs bicarbonate buffer contained the following (in mM): 118 NaCl, 4.7 KCl, 1.5 CaCl₂, 25 NaHCO₃, 1.1 MgSO₄, 1.2 KH₂PO₄, and 5.6 glucose. After a 1-h equilibration period, the vessels were depolarized with Krebs bicarbonate containing KCl in place of NaCl. This treatment produces maximal contraction and enhances the reproducibility of subsequent contractions. The arteries were then reequilibrated with Krebs solution for 30 min before the experiments were conducted. In all experiments, the functional removal of the endothelium by gently rubbing the lumen of the vessel was confirmed by examining the effect of 10  ⁸-10  ⁶ M acetylcholine on arteries precontracted with 10  ⁷ M serotonin. Endothelium-removed arteries did not show relaxation to acetylcholine and usually contracted to the 10  ⁶ M dose. After a 30-min equilibration period, experiments were conducted. In these experiments, arteries were typically precontracted to ~60% of maximal force (~7 g) with 20-30 mM KCl. Once a steady-state level of force was observed, the tissue was either exposed to a vasoactive agent or to hypoxia (PO₂  8-10 Torr) produced by use of a gas mixture of 95% N₂5% CO₂. After a 10-min exposure to hypoxia, arteries were reoxygenated with 21% O₂. The study of PO₂-elicited responses was designed to minimize the influence of O₂ gradients in the vessel wall on metabolism by examining a change in PO₂ from 150 Torr, which is well above the PO₂ range where mitochondrial metabolism would be altered, to 8-10 Torr, which should elicit a hypoxic metabolic response in the vessel wall. In mechanistic probing experiments, coronary arterial rings were preincubated for 30 min with either 10 mM MS or 100 µM Ebselen. Further studies examining the effects of the catalase inhibitor 3-amino-1,2,4-triazole employed a 30-min pretreatment with 50 mM followed by examination of the relaxation to H₂O₂ in the absence of 3-amino-1,2,4-triazole. None of these probes significantly altered the level of force generation to KCl.

Chemiluminescence detection of H₂O₂ release from vascular tissue. Employing methods that we previously described (18), we prepared endothelium-removed bovine coronary arterial rings as indicated in Measurement of changes in force in bovine coronary arteries for organ bath tone studies, but the rings were placed in plastic scintillation minivials containing 10 µM luminol plus 1 µM horseradish peroxidase (HRP) and other additions in a final volume of 1 ml of air-equilibrated Krebs solution buffered with 10 mM HEPES-NaOH (pH 7.4). The chemiluminescence elicited by H₂O₂ in the presence of luminol-HRP was measured in a liquid scintillation counter (Mark V, TmAnalytic, Elk Grove Village, IL) with a single active photomultiplier tube positioned in out-of-coincidence mode. All manipulations were performed in the darkroom with minimum lighting. Samples were initially at 37°C, but the temperatures subsequently equilibrated with the ambient temperature of ~34°C. After 5 min of dark adaptation, vials containing all components, with the exception of arterial rings (blanks), were counted once for 0.1 min over the next 10 min. This procedure was repeated twice after ~400 mg of an endothelium-removed arterial ring were placed in each vial. Blanks were then subtracted from the average of the relatively constant levels of chemiluminescence produced under each condition by the arteries to obtain the data reported as counts per minute per gram of tissue. The inhibition of chemiluminescence by the addition of 1 µM catalase was used to determine the relative amounts of H₂O₂ released from coronary arteries.

Statistical analysis. An ANOVA statistical analysis employing a post hoc Scheffé's test was used for all studies on vascular contractility. All chemiluminescence data were analyzed by Student's t-tests employing a Bonferroni correction for multiple comparisons. The acceptable level of significance was P < 0.05. The number of experimental determinations (n) in all cases is equal to the animals from which an arterial ring was employed for a treatment or a control group in contractile and tissue chemiluminescence studies. Values are means ± SE.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of treatment of isolated endothelium-removed bovine coronary arteries with probes for GSH peroxidase on force generation and relaxation to hypoxia and reoxygenation. As shown in Fig. 2, Ebselen and MS did not significantly alter the contractile response to increasing concentrations of KCl. In endothelium-removed coronary arteries precontracted to ~7 g of force with 20-30 mM K⁺, 100 µM Ebselen caused a 47% inhibition of the initial transient posthypoxic reoxygenation-elicited relaxation without significantly altering the relaxation to hypoxia (see Fig. 3A). In contrast, 10 mM MS caused a 121% enhancement of the initial transient posthypoxic reoxygenation-elicited relaxation and without significantly altering the relaxation to hypoxia (see Fig. 3B). The relaxation to posthypoxic reoxygenation of 53.1 ± 7.0% (n = 15) in the presence of MS was also significantly (P < 0.05) inhibited by 0.3 mM nitroblue tetrazolium (36.3 ± 5.8% relaxation = 43% inhibition; n = 15) and 10 µM diphenyliodonium (18.4 ± 7.0% relaxation = 65% inhibition; n = 11), probes previously shown (17) to attenuate the reoxygenation response by 42 and 74%, respectively, in the absence of MS. Because the contractile response to KCl appeared to show a trend toward a suppression of force generation in the presence of MS, the effects of this probe were examined on the response to 1 nM-3 µM serotonin, a contractile agent that appears to be more sensitive to the relaxant effects of H₂O₂ (4). The contraction to serotonin was inhibited by 41-82% in the presence of MS (P < 0.05; n = 16). For example, the increase in force of 7.23 ± 1.04 g elicited by 1 µM serotonin was reduced to 3.75 ± 0.65 g in the presence of MS. The catalase-inhibitable chemiluminescence observed to be derived from endothelium-removed bovine coronary arteries in the presence of luminol-HRP was used to examine the effects of inhibiting GSH peroxidase on the release of H₂O₂. In the absence of the MS inhibitor, catalase did not inhibit the near-background levels of chemiluminescence [control: 2.5 ± 0.7 × 10⁵ counts/min (cpm)/mg and plus catalase = 2.9 ± 1.2 × 10⁵ cpm/mg], whereas MS significantly increased chemiluminescence to 17.7 ± 3.4 × 10⁵ cpm/mg (P < 0.05; n = 14) and catalase reduced (P < 0.05) chemiluminescence to 5.1 ± 0.5 × 10⁵ cpm/mg. Thus MS causes a detectable release of H₂O₂ from endothelium-removed bovine coronary arteries, consistent with it impairing the metabolism of endogenously generated H₂O₂ by GSH peroxidase.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of 100 µM Ebselen (A; n = 10-12 rings) and 10 mM MS (B; n = 15 rings) on the generation of force by isolated endothelium-removed bovine coronary arteries elicited on exposure to increasing cumulative concentrations of KCl. There was no significant change between control and pretreated vessels.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effect of 100 µM Ebselen (A; n = 27 rings) and 10 mM MS (B; n = 15 rings) on the relaxation responses of isolated bovine coronary arteries precontracted with 20-30 mM KCl to hypoxia and posthypoxic reoxygenation. There was no significant effect of these probes on the relaxation to hypoxia.

Effect of probes for GSH peroxidase on relaxation of isolated endothelium-removed coronary arteries to lactate and H₂O₂. As shown in Fig. 4A, the relaxation of endothelium-removed bovine coronary arteries to 1-10 mM lactate was significantly attenuated in the presence of 100 µM Ebselen. In contrast, the data in Fig. 4B indicate that the relaxation to 1-10 mM lactate was significantly enhanced in the presence of 10 mM MS. The data shown in Fig. 5A demonstrate that the 100 µM Ebselen significantly inhibits the relaxation of endothelium-removed bovine coronary arteries to 1 µM-1 mM H₂O₂. In contrast, the relaxation to 1 µM-1 mM H₂O₂ was significantly enhanced by the presence of 10 mM MS at the 1 µM to 100 µM doses (see Fig. 5B). Pretreatment of endothelium-removed coronary arteries with catalase inhibitor 3-amino-1,2,4-triazole caused a marked attenuation of relaxation to H₂O₂ both in the absence and presence of MS (see Table 1). Thus Ebselen and MS alter the response to increases in H₂O₂ derived from both endogenous (lactate) and exogenous sources in a similar manner.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of 100 µM Ebselen (A; n = 30-38 rings) and 10 mM MS (B; n = 28-30 rings) on relaxation responses of isolated bovine coronary arteries precontracted with 20-30 mM KCl to lactate.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Effect of 100 µM Ebselen (A; n = 17-21 rings) and 10 mM MS (B; n = 15-17 rings) on relaxation responses of isolated bovine coronary arteries precontracted with 20-30 mM KCl to H2O2.

Absence of effect of probes for GSH peroxidase and catalase on relaxation to a nitric oxide donor. As shown by the data in Fig. 6, A and B, Ebselen and MS did not have a statistically significant effect on relaxation to the nitric oxide donor 10  ⁹-10  ⁵ M S-nitroso-N-acetylpenicillamine (SNAP). Inactivation of catalase with 3-amino-1,2,4-triazole pretreatment did not alter the response to SNAP (e.g., see Table 1).

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Absence of an effect of 100 µM Ebselen (A; n = 11 rings) and 10 mM MS (B; n = 11 rings) on relaxation responses of isolated bovine coronary arteries precontracted with 20-30 mM KCl to the nitric oxide donor S-nitroso-N-acetylpenicillamine (SNAP).

Effects of inhibition of catalase and the stimulation of sGC on coronary artery relaxation to an oxidant of GSH. Exposure of endothelium-removed bovine coronary arteries precontracted with KCl to the glutathione oxidant diamide (3) caused a 56.5 ± 8.2% relaxation (n = 10), and this response was not altered by inactivation of catalase with 3-amino-1,2,4-triazole pretreatment (see Table 1) or by inhibition of the stimulation of sGC with 10 µM LY-83583 (51.6 ± 7.3% relaxation; n = 9). Thus the response to diamide appeared to function through a mechanism that was different from the response to H₂O₂.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

This study shows that the modulation of GSH peroxidase activity has effects on relaxation of endothelium-removed bovine coronary arteries to posthypoxic reoxygenation, lactate, and exogenous H₂O₂, which were consistent with the hypothesized influence of this system in controlling the stimulation of sGC by peroxide metabolism through catalase. Inhibition of GSH peroxidase with MS enhanced, and increasing GSH peroxidase activity with Ebselen attenuated, each of the responses that were thought to be mediated by H₂O₂. The attenuating effect of an inhibitor of catalase 3-amino-1,2,4-triazole on relaxation to H₂O₂ and on the potentiating effect of MS on this response was used to confirm the role of catalase in mediating these responses. The model in Fig. 1 illustrates the influence of probes for GSH peroxidase on the hypothesized mechanism of relaxation to both endogenous and exogenous H₂O₂ involving the stimulation of sGC activity as a result of alterations in peroxide metabolism by catalase.

Studies on how mammalian cells metabolize peroxides have resulted in the identification of GSH peroxidases, heme peroxidases, and catalase as the primary systems involved (6). The GSH peroxidase enzymes that metabolize H₂O₂ are usually selenium-containing proteins, which have the common property of linking the oxidation of GSH to the reduction of peroxides (14). Because the measurement of the formation of oxidized GSH has typically been used to characterize the various GSH peroxidase activities present in tissue extracts, most tissues have been examined for the presence of enzyme activities through which peroxides can influence the redox status of GSH. Because of the redox status of GSH potentially contributing to the control of certain cellular signaling systems (3, 13, 27, 30), peroxide metabolism by GSH peroxidases is likely to be an essential component of the processes through which H₂O₂ influences systems controlled by GSH redox. In the present study, the attenuating effect of increasing tissue GSH peroxidase activity with Ebselen and the enhancing effect of inhibition of GSH peroxidases with MS on responses thought to be mediated by H₂O₂ are actions that are the opposite of what would be expected from a signaling mechanism linked to the oxidation of GSH. Whereas there is evidence that alterations in GSH redox activate vascular relaxation through the opening of potassium channels (25, 28), the actions of the probes for GSH peroxidase suggest that this mechanism does not contribute to the H₂O₂-mediated responses examined in the present study. Alternatively, cGMP has been demonstrated to regulate the opening of calcium-regulated potassium channels (1), and channels of this type have been demonstrated to participate in vasodilator responses to H₂O₂ (26). The activation of additional GSH redox-regulated mechanisms might either require exposure to increased levels of peroxide for their expression or processes most sensitive to alterations in GSH redox may not influence force generation under the conditions examined in the present study. However, the oxidant of GSH diamide elicited a relaxation response that was not altered by inhibition of catalase or the stimulation of sGC, suggesting that thiol oxidation has the potential to activate an additional cGMP-independent mechanism of relaxation, which could potentially function through a process involving modification of thiols. Thus signaling mechanisms other than alterations in GSH redox need to be considered in establishing the primary processes through which H₂O₂ elicits relaxation of endothelium-removed bovine coronary arteries.

Heme peroxidases and catalases are enzymes that could participate in linking peroxide metabolism to signaling mechanisms that control vascular responses. Cyclooxygenase seems to be only heme peroxidase potentially present in the vessel wall that is known to be involved in responses elicited by H₂O₂. Whereas cyclooxygenase-dependent mechanisms have been observed to be the dominant process contributing to the response of certain vascular segments to H₂O₂ (15, 16, 22, 32), the endothelium appears to be the origin of prostaglandin-mediated responses in isolated bovine coronary arteries (17, 21). In addition, H₂O₂-associated vascular relaxation to neutralized lactate does not seem to be linked to the generation of prostaglandin mediators (8, 23).

Our previous studies on the response of endothelium-removed bovine coronary arteries to posthypoxic reoxygenation, lactate, and H₂O₂ have provided evidence consistent with a role for the stimulation of sGC by H₂O₂ in these responses (17, 31). The metabolism of peroxide by catalase appears to be a key mediator of sGC stimulation by H₂O₂ in vascular tissue (4, 5), and the attenuation of relaxation to H₂O₂ observed in the present study by 3-amino-1,2,4-triazole supports the hypothesized role of the metabolism of peroxide by catalase in the response to H₂O₂. On the basis of the model in Fig. 1, inhibition of GSH peroxidase with MS would be expected to increase the metabolism of H₂O₂ by catalase and to enhance relaxation responses mediated by this process. In contrast the increase in GSH peroxidase-like metabolizing activity catalyzed by Ebselen would be expected to attenuate H₂O₂ metabolism by catalase and relaxation responses dependent on this process. The actions of MS and Ebselen appear to show selectivity for responses involving H₂O₂, because relaxation responses to the nitric oxide donor SNAP and to hypoxia were not altered by these probes. Because MS and Ebselen acted on the responses thought to be mediated by H₂O₂ that were consistent with their predicted effects on peroxide metabolism by catalase, the actions of these probes support a role for the mechanism proposed in Fig. 1. In this mechanism, the key role of GSH peroxidase in vascular responses involving H₂O₂ in endothelium-removed bovine coronary arteries is its modulating effect on the metabolism of peroxide by catalase.

The modulation of GSH peroxidase activity is likely to have effects on peroxide-associated responses that are specific to the vascular segments or models in which they are examined. Whereas vessel segments such as the porcine coronary artery show cGMP-associated relaxation to H₂O₂ (11), other mechanisms may dominate the response elicited by alterations in peroxide metabolism. For instance, the cGMP-associated relaxation to H₂O₂ is very sensitive to inhibition by O₂ · (9) and nitric oxide (18), and the relaxation may not be readily expressed in certain vascular segments. Alternatively, other mechanisms may be more sensitive to the effects of peroxides and they may dominate the response that is observed. The generation of vasoactive prostaglandins (15, 16, 22, 32) and stimulation of nitric oxide release (12) are examples of responses that have been demonstrated in other vascular preparations to be more sensitive to the actions of H₂O₂ than the responses in these segments, which appear to be similar to those examined in the present study. Thus it is likely that alterations in the function of GSH peroxidase will influence the response of different vascular segments or animal models in a manner that is determined by the function of the signaling mechanisms present.

Our previous studies on endothelium-removed bovine coronary arteries suggested that H₂O₂ originating from a O₂ · -producing NADH oxidase was an important contributor to the transient relaxation elicited in response to posthypoxic reoxygenation (17). The results of the present work provide evidence that the metabolism of H₂O₂ by GSH peroxidase and the processes controlled by the alterations in GSH redox caused by use of this metabolic pathway do not appear to be important participants in relaxation responses of bovine coronary arteries thought to be mediated by H₂O₂. However, as predicted by the model in Fig. 1, the function of GSH peroxidase seems to have a major influence on the expression of responses that are thought to be linked to the stimulation of cGMP-mediated relaxation as a result of the activation of sGC by increases in H₂O₂ metabolism by catalase. Because the expression of H₂O₂ metabolism by GSH peroxidase is a process that is potentially susceptible to metabolic and pathophysiological regulation, it is likely that the function of this peroxide-metabolizing pathway has a major influence in the expression of different vascular signaling mechanisms that are linked to the regulatory effects of the various metabolizing systems for peroxides. The present study documents the importance of the function of GSH peroxidase on the expression of responses thought to be mediated by the stimulation of sGC through H₂O₂ metabolism by catalase.