Vascular non-protein thiols: prooxidants or antioxidants in atherogenesis?

Abstract-cell-mediated lipoprotein oxidation may be due to generation of non-protein thiols (NP-SH) from cystine with formation of oxidizing species. However, NP-SH, especially GSH, may also exert antioxidant effects in vitro and in vivo. To further investigate whether vascular NP-SH could be prooxidants or antioxidants in atherosclerosis, we have correlated the aortic content of NP-SH with that of lipoperoxides in 10 rabbits fed on a fat-enriched atherogenic diet for 9 weeks. As compared to 7 control rabbits, aortic NP-SH and lipoperoxides were significantly increased in the fat-fed animals. The levels of NP-SH were strongly and inversely correlated with those of lipid peroxidation in the atherosclerotic aorta (r(s) -0.92, P < 0.0001 for thiobarbituric acid reactive substances, and r(s) -0.80, P < 0.01 for fluorescent damage products of lipid peroxidation). A similar trend was evident also in the control rabbits (r(s) -0.60 for both indices of lipid peroxidation). Thus, the present data suggest that vascular NP-SH exert significant antioxidant-antilipoperoxidative effects in vivo especially in fat diet-related atherogenic conditions.

Antioxidant properties of ticlopidine on human low density lipoprotein oxidation.

We found that ticlopidine, at therapeutically relevant concentrations (2.5-10 microM), but not aspirin nor salicylate, significantly counteracted copper-driven human LDL oxidation. Ticlopidine, at 5 and 10 microM, was also antioxidant on peroxyl radical-induced LDL oxidation; yet it was ineffectual on thiol and ascorbate oxidation mediated by peroxyl radicals themselves, suggesting that drug antioxidant capacity is somehow related to the lipoprotein nature of the oxidizable substrate, but not to radical scavenging. The drug could not indeed react with the stable free radical 1,1-diphenyl-2-pycrylhydrazyl, nor had apparent metal complexing-inactivating activity. Thus, ticlopidine has antioxidant effects on LDL oxidation, which, together with its anti-platelet activity, could confer peculiar antiatherogenic properties to the drug in vivo.

Glutathione-related antioxidant defenses in human atherosclerotic plaques.

BACKGROUND: Oxidative stress, resulting from an antioxidant/prooxidant imbalance, seems to be crucial in atherogenesis. Recent evidence has emerged, however, of a surprisingly high content of low-molecular-weight antioxidants in human atherosclerotic plaques, although other antioxidant systems have not been investigated in these lesions. METHODS AND RESULTS: We studied glutathione-related antioxidant defenses (which play a key role in tissue antioxidant protection) in carotid atherosclerotic plaques of 13 patients subjected to endarterectomy and in normal internal mammary arteries of 13 patients undergoing coronary artery bypass surgery. Selenium-dependent glutathione peroxidase activity was undetectable in the plaques of 7 patients; the other 6 patients with plaques showed a mean enzymatic activity approximately 3.5-fold lower than that of mammary arteries. Glutathione reductase activity was also markedly lower in the plaques than in the arteries. Glutathione transferase instead had comparable activity in the two tissues. Remarkably, 5 of the 7 patients with an undetectable selenium-dependent glutathione peroxidase activity but none of the 6 with a detectable one were characterized by multivascular atherosclerotic involvement (3 patients) or stenosis of the contralateral carotid artery (2 patients). CONCLUSIONS: A weak glutathione-related enzymatic antioxidant shield is present in human atherosclerotic lesions. Although the cause of this phenomenon remains to be determined, the present data suggest that a specific antioxidant/prooxidant imbalance operative in the vascular wall may be involved in atherogenic processes in humans.

Antioxidant activity of allopurinol on copper-catalysed human lipoprotein oxidation.

We found that allopurinol, at therapeutically relevant concentrations (9-58 microM), significantly counteracted copper-catalysed human non-HDL lipoprotein oxidation, as assessed by thiobarbituric acid reactant content and kinetics of conjugated diene formation. Oxypurinol was ineffectual. Both drugs had no activity on metal-independent, peroxyl radical-induced lipoprotein oxidation. Specific fluorescence-quenching experiments revealed that only allopurinol could interact with copper antagonizing metal binding to lipoproteins. Thus, therapeutic allopurinol concentrations can inhibit copper-catalysed lipoprotein oxidation through metal complexation, suggesting some antioxidant-antiatherogenic activity of the drug in vivo.

Captopril has no significant scavenging antioxidant activity in human plasma in vitro or in vivo.

1. Captopril has been reported to possess hydroxyl radical (OH.) and hypochlorous acid (HOCl) scavenging effects, which could contribute to its therapeutic activity in the clinical setting. 2. The objective of the present study was to determine whether therapeutically achievable captopril concentrations could augment antioxidant properties of human plasma and protect it against OH.- and HOCl-driven oxidant injury in vitro. Possible drug influences on systemic oxidative stress status in vivo were also investigated in subjects taking 50 mg captopril orally by measuring plasma and red blood cell peroxidation, as well as plasma protein thiols. 3. The results show that captopril is incapable of enhancing antioxidant properties of human plasma, of protecting it against specific oxidative attack and of decreasing systemic oxidant load in vivo. 4. The present data, therefore, do not support the contention of a beneficial action of captopril through systemic antiradical-antioxidant effects in human beings.

Impaired glutathione biosynthesis in the ischemic-reperfused rabbit myocardium.

Non-protein thiols (NP-SH) and the activities of the glutathione status-regulating enzymes gamma-glutamylcysteine synthetase (G-GCS), gamma-glutamyl transpeptidase (G-GT) and glutathione reductase (GR) were assessed in perfused rabbit hearts subjected to severe (60 min) or mild (7 min) total ischemia and 30 min reperfusion. Severe ischemia significantly decreased NP-SH, which were further depressed on reperfusion together with a significant decline in G-GCS activity; G-GT and GR activities were unchanged. Specific analytes were unaffected by mild ischemia-reperfusion. Thus, impaired enzymatic biosynthesis of GSH is operative in the reperfused rabbit myocardium after 60 min ischemia. This phenomenon may favour myocardial GSH depression and oxidative reperfusion injury after severe ischemia.

Antioxidant properties of omeprazole.

Potential antioxidant properties of therapeutically achievable concentrations of the protonated, active form of omeprazole (OM) were investigated in vitro at specific acidic pH values to mimic intragastric conditions in the clinical setting. We found that OM is a powerful scavenger of hypochlorous acid (HOCl) even at a drug concentration of 10 microM at pH 5.3 or 3.5. This effect is also evident in the presence of the physiological HOCl scavenger ascorbate. Moreover, 10 and 50 microM OM inhibit significantly both iron- and copper-driven oxidant damage at pH 5.3 and 3.5, respectively. Since oxidative stress is involved the gastric injury of peptic ulcer and gastritis, it may be hypothesized that some therapeutical effects of OM could also be related to its antioxidant properties.

Plasma copper and lipid peroxidation in cigarette smokers.

Plasma levels of copper and lipid peroxidation were evaluated in 14 smokers as compared to 14 nonsmokers. Plasma copper concentrations were higher in smokers than in nonsmokers (122.5 +/- 19.15 vs. 101.5 +/- 16.2 micrograms/dl, P < .01). Plasma lipoperoxidation, evaluated as fluorescent damage products of lipid peroxidation (FDPL), also was higher in smokers than in nonsmokers (20.35 +/- 2.6 vs. 17.1 +/- 2.95 units of relative fluorescence/ml, P < .01). A significant and positive correlation between the number of cigarettes smoked, expressed as pack years, and the levels of either FDPL (r = .61, P < .025) or copper (r = .55, P < .05) was found. Moreover, a significant and positive relationship between copper and FDPL values was observed in smokers (r = .64; P < .025), but not in nonsmokers. These data indicate that cigarette smoke-related plasma oxidant load may be partly due to enhanced levels of the prooxidant metal cooper, potentially suggesting the supplementation of specific antioxidants (e.g., zinc) to counteract cigarette smoke-induced oxidative stress in smokers.

Aminophylline: could it act as an antioxidant in vivo?

Potential antioxidant properties of aminophylline and theophylline were investigated. We have found that these drugs, though ineffective against superoxide anion and hydrogen peroxide, are scavengers of hydroxyl radical (OH.). Indeed, second-order rate constants (k) of aminophylline and theophylline with OH. are about 1.9 x 10(10) mol-1 s-1 and 4.5 x 10(9) mol-1 s-1, respectively. Ethylenediamine, which is present in the aminophylline molecule, significantly contributes to this marked OH. scavenging activity, since it is characterized by a high k value, i.e. about 8 x 10(9) mol-1 s-1. However, when using therapeutically relevant concentrations of the methylxanthines (9 and 13 micrograms mL-1), significant antioxidant effects against OH.-induced oxidant injury are evident only with aminophylline. Although all three substances can apparently bind and inactivate iron, only aminophylline is effective at 9 and 13 micrograms mL-1; also this action is favoured by ethylenediamine. Moreover, therapeutic concentrations of aminophylline, but not of theophylline, are capable of antagonizing hypochlorous acid (HOCl); this effect is entirely due to the presence of ethylenediamine. Oxidant species, such as OH. and HOCl, have been implicated in the pathophysiology of asthma; it could be hypothesized, therefore, that some therapeutic effects of aminophylline may be related to its antioxidant properties, which are partly or totally attributable to ethylenediamine, depending on the chemical identity of the prooxidant antagonized (e.g. iron/OH. or HOCl). Aminophylline antioxidant capacity should be taken into account when investigating the lung epithelial lining fluid antioxidant capacity and oxidative stress indices in humans.

The prooxidant properties of captopril.

The thiol drug captopril has been reported to possess reducing and transition metal-binding properties, which could result in specific changes in iron and copper prooxidant capacity. Thus, the effects of captopril on iron- and copper-induced oxidative injury were evaluated using deoxyribose as the oxidizable substrate in the presence of physiological phosphate concentrations but in the absence of the non-physiological chelator EDTA. In an iron(III)/H2O2/ascorbate oxidant system, captopril enhanced deoxyribose oxidation only when it was pre-mixed with iron, whereas it did not influence sugar degradation when not pre-mixed with the metal or when ascorbate was omitted. The physiological thiol GSH acted in a similar manner, whereas the SH-lacking angiotensin-converting enzyme inhibitor ramiprilat did not influence iron-induced deoxyribose oxidation, indicating that the thiol group is crucial in favouring enhanced iron reactivity due to 'malignant' chelation. Further specific experiments designed to evaluate possible thiol-dependent iron(III) reduction failed to demonstrate ferric to ferrous reduction by either captopril or reduced glutathione (GSH). When iron(III) was replaced by copper(II) to induce deoxyribose oxidation, captopril was prooxidant both in the presence and absence of ascorbate, and when pre-mixed or not with copper. On the other hand, GSH was prooxidant up to 2:1 molar ratio with respect to copper but markedly inhibited copper-dependent sugar oxidation beginning at molar ratio of 4:1. Ramiprilat did not significantly influence copper-induced deoxyribose oxidation. Moreover, unlike the experiments performed with iron, captopril, as well as GSH, readily reduced copper(II) to copper(I). Hence, captopril can act as prooxidant in the presence of iron or copper. In the former case, only 'malignant' iron chelation by the drug is involved in oxidant injury, whereas in the latter both copper chelation and reduction are operative, although specific chelating mechanisms are crucial in enhancing copper-induced oxidant injury. Captopril, therefore, cannot be considered simply as an 'antioxidant drug', and its catalytic transition metal-related prooxidant capacity should be taken into account in experimental and clinical investigations.

Cephalosporins are scavengers of hypochlorous acid.

Potential scavenging properties of cephalosporins (i.e. cefamandole, cefotaxime and ceftriaxone) towards hypochlorous acid (HOCl) as well as the antibacterial activity of control and HOCl-reacted antibiotics were investigated. We found that these drugs, at therapeutically relevant concentrations, are indeed scavengers of HOCl, with ceftriaxone showing the highest anti-HOCl capacity. However, the efficiency of cephalosporins in protecting biological molecules is also related to the chemical identity of such molecules. Indeed, the polyenoic compound beta-carotene is much better protected that the thiol compound GSH against HOCl attack. Moreover, the drugs do not appear to form chloramine derivatives as a result of their reaction with HOCl, and they inhibit taurine-chloramine formation. After HOCl challenge, the antibacterial activity of cefamandole, cefotaxime and ceftriaxone (tested against the standard strain Escherichia coli ATCC 25922) is approx. 8-, 5- and 4-fold lower, respectively, than that of the HOCl-unreacted antibiotics. The depression of the antibacterial activity of cephalosporins appears inversely related to their HOCl scavenging capacity, suggesting that the drug antioxidant groups may protect the beta-lactam ring against HOCl attack. In conclusion, physiological biomolecules are protected by cephalosporins against HOCl-driven oxidative injury with varying efficiency, this antioxidant defence being a consequence of a direct drug scavenging capacity towards HOCl. The interaction of cephalosporins with HOCl, however, results in a depression of their antibacterial activity.

Cigarette smoke, ferritin, and lipid peroxidation.

Gas and tar phases of commercially available filter cigarettes were tested for ferritin-iron-releasing effects and polyunsaturated-fatty-acid oxidant capacity in vitro. A vacuum pump-dependent apparatus with Cambridge filters was used to separate gas and tar; the former was directly smoked into reaction mixtures, while the latter was extracted from Cambridge filters in aqueous medium and freshly used at 40 to 80% final concentrations. Both phases induced ferritin iron release, which was not antagonized by superoxide dismutase (SOD). In specific experiments, we have also shown that gas and tar extracts could cross an organic (i.e., chloroform)-phospholipid layer before mobilizing ferritin iron. Once delocalized from ferritin, iron could trigger lipid peroxidation; however, a marked prooxidant effect (inhibited by 20 microM deferoxamine mesylate and significantly decreased by 40 microM vitamin E) was observed only with gas, whereas tar extracts showed antioxidant effects. Accordingly, tar extracts could also antagonize lipid peroxidation driven by non-chelated iron or by peroxyl radicals. In the absence of ferritin, gas-induced lipid peroxidation was very low, and tar extracts were apparently ineffective. Thus, the intrinsic lipoperoxidative capacity of cigarette smoke is low and is due to gas; however, when smoke interacts with ferritin, a marked iron-driven peroxidation becomes manifest essentially with gas, tar components acting as antioxidants. The present data suggest that cigarette-smoke-mediated iron mobilization from ferritin may represent a specific prooxidant mechanism related to cigarette smoking in vivo.

Ambroxol is a scavenger of hypochlorous acid and monochloramine.

Experimental studies have suggested that ambroxol, a drug clinically used to enhance pulmonary surfactant production, may also exert some anti-inflammatory effects, though specific mechanisms are not yet fully understood. Thus, potential scavenging properties of ambroxol towards the most toxic neutrophil-arising prooxidants hypochlorous acid and monochloramine were investigated. We have found that the drug at 25, 50 and 100 mumol/l exerted a significant, concentration-related antagonizing effect on both chlorine species, with a maximal specific activity detected against hypochlorous acid. These new pharmacological properties of ambroxol may play a role in vivo in disease entities characterized by white blood cell activation and uncontrolled oxidant generation, such as inflammatory/ischemic conditions.

Transmural distribution of antioxidant defences and lipid peroxidation in the rabbit left ventricular myocardium.

The left ventricular subendocardial and subepicardial layers of six perfused rabbit hearts were tested for enzymatic and non-enzymatic antioxidant defences and for lipid peroxidation. The subendocardium showed significantly lower catalase activity and contents of non-protein thiol compounds and vitamin E associated with a higher degree of lipid peroxidation. The activities of Cu,Zn- and Mn-superoxide dismutases, glutathione reductase, gamma-glutamylcysteine synthetase and gamma-glutamyl transpeptidase showed no significant transmural differences, and Se-independent glutathione peroxidase activity was not detectable in either layer. Comparable results were observed in another group of six unperfused rabbit hearts. In five H2O2-perfused rabbit hearts, lipid peroxidation was higher, and myocardial creatine phosphokinase activity lower, in the subendocardium than in the subepicardium. In this group, only the subendocardium had significantly higher lipid peroxidation levels than the control hearts. Thus, a lower antioxidant capacity and a greater oxidative stress are present in the rabbit subendocardium. These findings could provide insight into the problem of subendocardial vulnerability to free radical-mediated processes, such as occurs in ischaemia-reperfusion injury.

H2-receptor antagonists are scavengers of oxygen radicals.

Potential oxygen radical scavenging properties of the H2-receptor antagonists cimetidine, ranitidine and famotidine were investigated. These drugs, although ineffective against superoxide anion and hydrogen peroxide, can scavenge hydroxyl radical (OH.) with a very high rate constant, which is about tenfold higher than that of the specific scavenger mannitol for famotidine (1.7 x 10(10) mol-1 s-1) and cimetidine (1.6 x 10(10) mol-1 s-1), ranitidine displaying a rate constant of 7.5 x 10(9) mol-1 s-1. These OH. savenging effects are significant beginning from 10, 28 and 100 mumol l-1 concentration for famotidine, cimetidine and ranitidine, respectively, thus suggesting that the drugs may effectively act as OH. scavengers in vivo especially in the gastric lumen. Only cimetidine can apparently bind and inactivate iron, which further emphasizes its antioxidant capacity. Moreover, all drugs, even at 10 mumol l-1 concentration, show powerful scavenging effects on hypochlorous acid and monochloramine, which are cytotoxic oxidants arising from inflammatory cells, such as neutrophils. These data suggest that some therapeutical effects of H2-receptor antagonists in peptic ulcer may also be related to their antiradical-antioxidant capacity, and that these drugs could potentially be used in other disease entities characterized by free radical-mediated oxidative stress in vivo.

Blood cardioplegia reduces oxidant burden in the ischemic and reperfused human myocardium.

In 20 patients receiving cold crystalloid cardioplegia (n = 10) or cold blood cardioplegia (n = 10) during elective coronary artery bypass grafting, the atrial myocardium was tested for glutathione-related antioxidant defenses and lipid peroxidation. In both groups, ischemia and reperfusion induced a significant increase in lipid peroxidation values (p < 0.05) that was associated with a depression of nonprotein thiol compound levels (p < 0.05). Compared with the cold crystalloid cardioplegia-treated patients, the cold blood cardioplegia-treated patients showed a lower lipid peroxidation (p < 0.05) and higher values of nonprotein thiol compounds (p < 0.05). Moreover, a significant ischemia and reperfusion-dependent activation of glutathione transferase was observed only in the cold crystalloid cardioplegia-treated patients. Selenium-dependent glutathione peroxidase and glutathione reductase activities did not change after release of the aortic cross-clamp and did not differ between the two groups. The highest postoperative plasma level of the myocardial-specific isoenzyme of creatine kinase was significantly more elevated in the cold crystalloid cardioplegia patients. Overall, these tissue biochemical features indicate a lower oxidant burden in the myocardium of cold blood cardioplegia-treated patients, a finding suggesting superior protection for the ischemic and reperfused human myocardium also through antioxidant-type mechanisms, apparently medicated by the antioxidant capacity of erythrocytes and specific plasma molecules.

Hypochlorous acid-induced zinc release from thiolate bonds: a potential protective mechanism towards biomolecules oxidant damage during inflammation.

It has been proposed that metalloprotein zinc mobilization mediated by hypochlorous acid (HOCl) may induce cell injury (see H. Fliss and M. Ménard (1991), Archives of Biochemistry and Biophysics, 287, 175-179). In the present paper, we have demonstrated using a dimercaptopropanol-zinc complex that, once released from thiolate bonds by HOCl, zinc can exert a significant antioxidant effect on both linolenic acid and deoxyribose oxidation induced by iron. In these experimental conditions, however, the antagonism towards deoxyribose oxidation is notably less than that towards linolenic acid peroxidation, thus suggesting a more specific inhibitory effect of zinc on iron-mediated oxidant damage when polyunsaturated fatty acids represent the oxidizable substrate. The antioxidant effects of zinc are strictly related to the "free" form; indeed, the dimercaptopropanol-zinc complex per se is stimulatory even on biomolecules oxidant damage, apparently as a result of the prooxidant prooxidant interaction of the thiol compound with iron. In light of these results, it may be proposed that the zinc released from thiolate bonds by HOCl could specifically limit tissue oxidative burden in pathological conditions involving neutrophil accumulation and activation, such as inflammation and ischemia-reperfusion.

Hypoxia-induced coronary flow changes in the perfused rat heart: effects of high L-carnitine concentrations.

1. Hemodynamic effects of physiological (0.04 and 0.07 mM) and high (8, 15 and 25 mM) L-carnitine (LC) concentrations were tested on the normally oxygenated and hypoxic perfused rat heart. 2. No effect was detected on aerobic hearts, whereas a dose-dependent rise in coronary flow (CF) during both the hyperemic and constrictive phases was observed in hypoxic hearts with 8, 15 and 25 mM LC. This action was apparently unrelated to a resting tension (RT) improvement, which was observed only with 25 mM LC. 3. When 11 mM glucose was replaced by 11 mM mannitol in the perfusion buffer, LC effects on the hyperemic phase were abolished; however, 25 mM LC resulted in CF-values still significantly higher than those detected without the drug, though RT was similar in these glucose-free groups. 4. It may be concluded that LC is ineffective on the perfused rat heart in aerobic conditions, whereas high LC concentrations can enhance CF only during hypoxia, these effects being independent of heart function improvements and partly unrelated to glucose presence.

Heparin: does it act as an antioxidant in vivo?

Previous studies have shown that heparin antagonizes oxygen radical-mediated injury to endothelial cells, suggesting an antioxidant role of the drug. In the present investigation, the hypothesis that heparin exerts direct antioxidant effects was tested in several experimental models. We have found that 1, 3, 5, 10, 20, 40 and 80 U/mL of heparin do not scavenge superoxide anion, hydrogen peroxide, hydroxyl radical or the stable free radical 1,1-diphenyl-2-pycrylhydrazyl. Moreover, the drug is ineffective towards iron-driven linolenic acid peroxidation, autooxidation of brain homogenate and linolenic acid peroxidation mediated by human internal mammary artery homogenate. Specific studies on the potential iron-binding-inactivating capacity of heparin prove the drug to be totally ineffective. Finally, the loss of protein sulphydryls from human plasma induced by hypoxanthine-xanthine oxidase-generated oxygen radicals is not prevented by heparin. In conclusion, heparin, even at concentrations far higher than those usually used therapeutically, has no direct antioxidant properties. Thus, other mechanisms not strictly antioxidant-type must be involved in heparin-mediated cell protection against toxic oxygen metabolites.