|Home | About | Journals | Submit | Contact Us | Français|
Oxidative stress is recognized as an essential mechanism of atherogenesis and plaque progression. However, the origin of increased free radical production has not yet been well described. Furthermore, therapy with antioxidants has not shown convincing results.
To consider questions concerning the impact of oxidative stress, and the effects and usefulness of antioxidants.
Atherosclerotic plaques were induced in rabbits by feeding them a cholesterol-rich diet (2%) for six weeks. Thereafter a normal diet was given up to 68 weeks. Body weight, food intake, plasma lipid concentration and antioxidative capacity were determined at various time intervals. Aortic plaque size, morphology and radical production were determined in groups of animals killed after six, 14, 21, 29, 40 and 74 weeks, and compared with values in untreated controls. Chemiluminescent methods were used to determine antioxidative capacity of plasma, generation of free radicals and redox reactivity of various antioxidants.
Antioxidative capacity, occurrence of modified low density lipoprotein and generation of free radicals indicated oxidative stress during plaque progression; however, they showed different correlations to cellular components of the plaques. Furthermore it was shown that some antioxidants have both anti- and pro-oxidative properties.
Oxidative stress during atherogenesis seems to correlate with different phases of plaque development and can be associated with different types of reactive species. Because plaque remodelling and stabilization may also be a phase of increased free radical generation, therapeutic antioxidants must exert specific and selective activity; in particular, whether their oxidized form acts pro-oxidatively must be determined.
Based on experimental and clinical evidence, the oxidative stress hypothesis of atherogenesis describes the importance of free radical generation and oxidative modification of low density lipoproteins (LDL) as essential mechanisms in plaque formation (1–3). Initial events within the arterial intima as well as final developments causing atherothrombosis can be related to effects exerted or modulated by reactive oxygen species (ROS) or modified LDL. However, antioxidant therapy has not showed convincing improvement in atherosclerotic patients (4,5). Here we consider some aspects of oxidative stress during atherogenesis and report results obtained in an animal model.
Oxidative stress is a problem of aerobic life and means the increased formation of oxygen species with reactive unpaired electrons (that is, superoxide anion, hydroxyl radicals) or hydrogen peroxide produced by incomplete reduction of oxygen. By spontaneous dismutation, and reactions with myeloperoxidase or nitric oxide, further species with high reactive potential such as singlet oxygen, hypochlorite or peroxynitrite, respectively, can also be formed. Furthermore, many other molecules such as unsaturated fatty acids and even proteins can form radical intermediates. The main sources of ROS within the cells are the mitochondrial respiratory chain, peroxisomes, various enzymatic systems such as NADPH oxidase and lipoxygenase, and metal-catalyzed reactions. The high reactivity of the intermediates implies more or less short survival times and fast reactions with other cellular components. Detoxification occurs by reaction with various antioxidants such as vitamins E and C, or with the cellular glutathione system (5). Other reactions are clearly related to pathophysiological and injuring mechanisms such as lipid peroxidation, which can occur by a self-propagating chain reaction, DNA modification or several types of protein derivatization. However, as discussed below, important intracellular second messenger functions may also be provided by ROS.
There is overwhelming evidence that atherosclerotic plaque is formed by inflammatory reactions with increased secretion of ROS, lipid peroxidation and other oxidative mechanisms (1,2,5,6). However, although the cell types involved in intimal thickening – that is, smooth muscle cells, endothelial cells and macrophages – were shown to be able to produce ROS in vitro, little is known about their contribution to plaque formation. Therefore, we studied the morphology of plaque development in relation to parameters of oxidative stress in hypercholesterolemia-induced atherosclerotic lesions in rabbits.
Male New Zealand rabbits received a cholesterol-rich diet (2% cholesterol) for six weeks and a normal diet for up to 68 weeks more. Groups of animals (n=10 until week 40, n=6 in the last group) were killed at six, 14, 21, 29, 40 and 74 weeks. As already described in more detail (7), blood plasma was analyzed for the following parameters: lipid fractions and total cholesterol content; lag time of LDL oxidation during incubation with copper ions (according to 3); and antioxidative capacity by measuring the plasma-induced time of suppression of chemiluminescence generated by spontaneous decay of azoamidinopropane hydrochloride (AAPH, according to 8). Cross-sectional slices of segments of aortic arch and thoracic aorta were used for determination of intimal plaque area (hematoxylin and eosin staining), and intimal content of macrophages, smooth muscle cells and oxidized LDL, which were detected by immunohistochemical and morphometrical methods using antibodies against ram 11, alpha-smooth muscle actin and oxidized phospholipids, respectively (9). Free radical secretion within a specimen of proximal thoracic aorta was measured by luminol-enhanced chemiluminescence and by incubation with the fluorescent TEMPO (Molecular Probes, the Netherlands), and following fluorimetric detection (excitation wave length 365 nm, emission above 480 nm) (10). To characterize anti-oxidative properties of various compounds, the spontaneous generation of free radicals from AAPH as measured by luminol-enhanced chemiluminescence was used. In this assay chemiluminescence is quenched by antioxidants and increased by pro-oxidants (11). Statistical analysis was done using JMP 3.1 (SAS Institute Inc, USA) by calculating analysis of variance and comparing the mean values by Tukey-Kramer test. Significance was set at P<0.05.
Table 1 summarizes the results of the determinations of blood parameters and aortic plaque areas. The cholesterol concentration rose steeply after the feeding time of six weeks to a maximum of about 70 mol/L and fell to normal values after 40 weeks. Intimal plaque area increased to a maximum of about 7.4 mm2 after 21 weeks and was not significantly reduced during the further regression time. Interestingly, the lag time of copper-induced oxidation was prolonged during hypercholesterolemia, indicating that the oxidizability of LDL was decreased during the plaque progression phase. On the other hand, the antioxidant capacity of plasma, which is influenced mainly by the presence of antioxidative vitamins, urate and protein, was transiently increased after six weeks but decreased through to week 40.
Cellular analysis of the aortic plaques showed that macrophages were the dominant cell type of initial lesions (weeks 6 and 14), but that they were overgrown by smooth muscle cells after 21 weeks. Despite the reduction of plasma cholesterol, smooth muscle cells remained within the plaques throughout the experiment, but they showed remodelling with accumulation beneath the subendothelium, thus forming a fibrous cap.
The determinations of the parameters of oxidative stress were correlated with the cellular composition of the plaques. Significant correlations are given in Table 2. Very strong correlations were found between oxidized LDL, as detected within the intima by an antibody recognizing oxidized phospholipids, and the number of macrophages, and between the fluorimetrically determined radical production (corresponding to superoxide anion and hydroxyl radicals) of aortic rings and the number of intimal smooth muscle cells. Negative correlations were found between antioxidative capacity of plasma and intimal plaque area. The very low values of basic radical production that were determined by chemiluminescence indicated a time course similar to that of the amount of macrophages; however, because of large variations, the correlation was not found to be significant.
The use of the luminol-enhanced AAPH method for characterization of antioxidants showed that substances can be ordered into three categories. First, antioxidants such as nor-dihydroguaretic acid or trolox quenched the constant rate of the indicator chemiluminescent reaction in a concentration-dependent manner, in such a way that the initial rate of chemiluminescence was obtained after recovery. Second, substances such as glutathione induced a similar lag time in chemiluminescence, but thereafter a transient increase above the initial rate of the measuring signal occurred. Finally, vitamin C or thiols such as homocysteine induced a very strong, long lasting, concentration-dependent increase in chemiluminescence after the lag time.
Experimental models of atherosclerosis that deliver plaques similar to those found in humans – that is, in which endothelium is not removed – are very restricted. We used the model of hypercholesterolemic rabbits, in which plaques can be induced in a relatively short time, yet are very similar to human lesions. The results presented here clearly show oxidative stress during atherosclerotic plaque development. It seems that under the conditions of the model used, various reactive species can be produced from the distinct vascular cells during different phases of plaque development. Probably as a consequence of the inflammatory reaction, plasma antioxidative capacity is decreased at least during the phase of plaque progression and stabilization. Because lag time of LDL oxidation is increased simultaneously, these results mean that the usual assumption that atherogenesis is enforced by increased oxidizability of LDL is not generally valid. Therefore, one can assume that oxidation of LDL within the vessel wall during plaque formation depends not only on the lipoprotein-associated antioxidants but also on the complete antioxidative systems present in interstitial fluid and blood plasma. Therefore, these parameters should also be determined to completely describe the oxidative situation in atherogenesis.
In the plaque, the occurrence of oxidized LDL was closely related to that of macrophages. Because their contribution to inflammatory reaction and oxidative stress is well known (12), it is suggested that these cells are responsible for LDL oxidation and, concomitantly, induction of lipid deposits in the vessel wall. The method used did not discriminate between extracellular and intracellular modified LDL but because it was decreased even during further plaque progression, macrophage-dependent removal seems to be possible.
ROS production was measured by two independent methods with different specificity for the various oxygen radicals. The results showed that the signal indicated by TEMPO-fluorescence, which is selective for superoxide and hydroxyl radicals, correlated with intimal content of smooth muscle cells, whereas the unspecific luminol-enhanced chemiluminescence paralleled the number of macrophages. This suggests that both cell types should be regarded as a source of ROS and that during plaque development various reactive species are involved in different pathophysiological mechanisms. The increased TEMPO-fluorescence, which paralleled the increase in smooth muscle cells, can be related to the fact that these cells when present in the intima undergo a proliferative modulation and are metabolically activated. In cell culture, it was shown that proliferation of smooth muscle cells was accompanied by increased secretion of hydrogen peroxide (13), which can be formed from superoxide anions by the dismutase reaction.
As already mentioned, many steps finally leading to atherothrombosis can be associated with oxidative stress or the effects of modified LDL (1–3,5,14). However, recent findings support the view that important physiological effects, too, have to be considered for ROS. It was shown that many hormones, growth factors, etc, use NADPH oxidase-derived superoxide anion as an intracellular second messenger. In addition to effects on intracellular calcium regulation, inhibitory effects on protein phosphatases were described, thus enhancing protein phosphorylation by the kinase systems (reviewed in 15). These signalling mechanisms are even active in regulating gene expression (16). Further important aspects of reactive species are their contribution to the immune defence function and their role in smooth muscle cell proliferation, which is involved in the formation of the fibrous cap responsible for plaque stabilization.
This means that the functions of reactive metabolites are ambivalently related to the pathophysiology of atherosclerosis and that not only beneficial effects may be expected when antioxidants are used in therapy.
Further problems with using antioxidants arise from their chemical properties. Depending on their redox potential, antioxidants are oxidized by reactions with corresponding oxidants. The generated oxidized form must be eliminated or reduced again by the cellular metabolism, otherwise the anti-oxidant would become a pro-oxidant. One important part of the cellular antioxidative protection system is provided by a chain of redox partners involving vitamins E and C, as well as glutathione, which is finally reduced enzymatically by NADPH provided by the pentose phosphate cycle. Only by the complete reaction sequence is detoxification of reactive species achieved. Antioxidants used in therapy have to fit into these redox cycles depending on their solubility and their reactivity. In isolated systems as represented by LDL particles in vitro it was shown that under certain conditions vitamin E does not act as an antioxidant but mediates oxidative chemical reactions leading to LDL modification (17).
To characterize the specificity and the reaction mechanism of antioxidants, we developed various test assays (11). In principle, various reactive species (superoxide anions, hydroxyl radicals and other free radicals) were generated and measured by corresponding methods. The quenching effects of the antioxidants were determined. We showed that thiol compounds including homocysteine and N-acetylcysteine behave as antioxidants; however, in the AAPH reaction system, some of them (such as homocysteine) form strong pro-oxidative intermediates. The nature of the intermediary compounds remains to be established, but it is conceivable that similar reaction products are also generated in vivo, which might be responsible for oxidative damages induced by, for example, homocysteine.
Oxidative mechanisms – that is, production of ROS and free radicals as well as the formation of products of lipid peroxidation, protein modification, etc – seem to be most important during atherogenesis. However, it is evident that there are different transient phases of oxidative stress probably arising from different intimal cellular sources that may be important in different phases of plaque development. Problems with antioxidative therapy may arise from unknown or unspecific interferences with the oxidative mechanisms or unclear sites of action. Another problem may arise when the oxidized antioxidant is pro-oxidative and incompletely reduced by the cellular metabolism.