биохимия атеросклероза

Native LDL is not chemotactic for human monocytes. MM-LDL may be active in stimulating the endothelial cells to recruit the monocytes [4]. The monocytes migrate to the area in increasing numbers, and differentiate into tissue macrophages. Ox-LDL prevents both basal and stimulated migration of macrophages from the intima [17, 18].

In Vivo Studies

Initially the evidence for the Ox-LDL hypothesis of atherosclerosis only showed that oxidation occurred in atherosclerotic lesions. Studies demonstrated the presence of oxidized lipoproteins in atherosclerotic plaques [4]. These studies showed that the lipid and cholesterol peroxides could be correlated with disease severity.

This initial evidence began in 1971 when human peripheral lymph was found to contain an apoprotein of higher density than that found in plasma [19]. Raymond and Reynolds [20] isolated LDL from rabbit interstitial inflammatory fluid that had properties similar to Ox-LDL. Hoff et al. [21] then isolated apo B lipoproteins from the atherosclerotic aorta that contained peroxides, and Yla-Herttuala et al. [22] showed that isolated lipoprotein had many properties of Ox-LDL and was also chemotactic to monocytes.

Since then autoantibodies to Ox-LDL (aOx-LDL) have been isolated from humans and rabbits. Haberland et al. [23] and Palinski et al. [24] used these antibodies to identify modified LDL in atherosclerotic lesions of WHHL rabbits. Also, immunohistochemical techniques detected material in the macrophage-rich areas of the lesions that were shown to cross-react with these antibodies. Prospective studies have shown aOx-LDL can predict myocardial infarction (MI) [25]. aOx-LDL is elevated in established atherosclerosis and aOx-LDL are present in healthy individuals. However, there was no evidence to show that oxidation was the cause of disease not simply due to the presence of concentrated lipid. What is also unclear is the role of Ox-LDL in different stages and in the development of atherosclerosis. Finally, while it is believed that Ox-LDL is atherogenic, it is unclear whether aOx-LDL is also atherogenic [26].

Some of the most compelling direct evidence for this hypothesis comes from the studies of apo–/– mice, which are a good model of oxidative stress. These mice develop atherosclerosis in a manner similar to human atherosclerosis, and in these mice F2-isoprostanes (a lipid peroxidation product that is used as a marker of oxidative stress) are elevated in the plasma, tissue, and urine [27, 28].

The evidence to date is quite convincing that oxidation has a role in the development of atherosclerosis, although the evidence is mostly indirect. In summary, it is clear that some form of modification of the LDL is required for foam cell formation and that the most likely modification is through oxidation of the lipids in the LDL molecule. There is significant data on the

524 Lesley MacDonald-Wicks and Manohar Garg

Ox-LDL role in the formation of foam cells via the scavenger pathway in macrophages, and that the Ox-LDL exists in vivo and that it is present in aortic lesions. It is also clear that there are autoantibodies generated to OxLDL and that titers of aOx-LDL are generally correlated with the extent of atherosclerosis.

However, the evidence cannot provide answers at this stage for whether oxidation is an initiating factor for the development of atherosclerosis or whether the oxidation is caused by the inflammatory nature of atherosclerosis, and is therefore subsequent to the development of atherosclerosis. What is also still unclear is the exact location of the oxidation events.

Antioxidants

The definition of antioxidant is given by Halliwell [29] as a substance present in low concentrations when compared to the oxidizable substrate, which significantly delays or prevents oxidation of that substrate.

If oxidation is the cause of atherosclerosis then antioxidants must also be able to delay or prevent the progression of the atheroma development. In support of this idea studies of human populations have shown an association between the intake of antioxidants and the low prevalence of cardiovascular disease (CVD) [6, 30, 31, 32]. Similarly, most but not all animal studies have also been successful in showing antioxidant interventions have slowed the development of atherosclerosis [4, 6] with reductions of up to 60% in atherosclerotic lesions [33].

The most often studied antioxidants include probucol, vitamin E, β-carotene, and vitamin C [34, 35]. Several studies have demonstrated inhibition of atherosclerosis by probucol in animals [36–39]. Successful inhibition of atherosclerotic development has been shown in rabbits [37] and monkeys [38]. The treatment of LDL receptor-deficient rabbits with the hypolipidemic agent and antioxidant, probucol, had an antiatherogenic effect independent of the traditionally recognized cholesterol lowering effect of probucol. Presumably this added effect was due to the protection of the LDL from oxidation [39]. Studies of probucol compared to lovastatin, a drug known to reduce cholesterol to a similar degree to probucol, demonstrated a greater reduction in atherosclerosis from probucol [40]. Similarly, Mao et al. [41] demonstrated that a probucol analog inhibited LDL oxidation and atherosclerosis in the absence of any cholesterol lowering effects, suggesting the effect was mediated by an antioxidant mechanism. In a murine model of atherosclerosis, probucol promoted atherosclerosis in the aortic root, but inhibited disease progression at distal sites [42].

Subsequent tests in humans appear to have equivocal results, and are controversial. One study by Walldius et al. [43] showed no improvements in hypercholesterolemic subjects, while another study by Sawayama et al. [44] showed significant reduction in the progression of atherosclerosis in the carotid artery as assessed by intima-to-media thickness. It is unlikely that antioxidants

would have no role in the prevention of atherosclerosis as the basic etiology of atherosclerosis in humans and in animals is similar, therefore it would be difficult to understand why antioxidants work so effectively in animals but not in humans [45].

From the vitamin E studies it is clear that vitamin E alone is insufficient to prevent LDL oxidation, and it requires the synergistic effect of a number of lipid and aqueous soluble antioxidants to protect LDL from oxidation, termed coantioxidants. Studies have clearly shown that the lag time to oxidation ex vivo can be extended by the addition of vitamin E, and that only when the vitamin E is fully expended to form the tocopheryl radical can oxidation continue [35]. At best vitamin E can only delay the onset of peroxidation by 15–20 min, a small amount of time compared to an antioxidant, such as probucol, which can delay oxidation of the LDL fats by up to 20 h [45]. In vivo supplementation of 100 IU to 1600 IU of vitamin E lead to reduced ex vivo copper-induced oxidation of LDL in a dosedependent manner. Although the oxidizability of LDL was reduced there was no reduction in plasma LDL autoantibodies, indicating the aOx-LDL may not be a sensitive in vivo marker of the susceptibility of LDL to oxidation [46]. Wen et al. [46] speculated that this may be due to the long half-life of aOx-LDL or, that once initiated, there is a continuous immunological response to Ox-LDL contained in the atherosclerotic lesions. Dietary antioxidants appear to reduce in vivo markers of inflammatory chemokines and cytokines, as well as reducing the levels of autoantibodies to oxidatively modified proteins. This may well indicate a long-term attenuation of oxidation. Antioxidants also appear to ameliorate the proliferation of smooth muscle cells in culture whereas oxidants appear to promote proliferation. Also human studies appear to show that intimal–medial thickness is reduced upon antioxidant supplementation [33].

In support of these and many other animal, cell culture and ex vivo studies, epidemiological studies have shown an inverse relationship between vitamin E intake and ischemic heart disease. The positive results from animal studies and the presumed safety of naturally occurring antioxidant lead to large, prospective cohort studies. In particular two of these studies reported an association between antioxidant intake, serum concentration of antioxidants, and improved cardiovascular outcomes [47, 48]. As these were observational studies, no cause and effect relationship could be concluded. These favorable results were from the Health Professionals Follow-up Study [31, 32] and the Nurses Health Study [30]. These two studies could be considered to be primary prevention studies as individuals with existing disease were excluded from the analysis. Contrary to these findings, the Rotterdam Study [47], which focused on the elderly found a reduced risk of MI with β-carotene intake, but none with vitamin E or vitamin C. This may be due to the low rates of supplementation in the Rotterdam Study in comparison to the Health Profession Follow-up Study where the highest quintiles of vitamin E intake were achieved largely through supplementation.

526 Lesley MacDonald-Wicks and Manohar Garg

The concern is that, although antioxidant defenses are comprehensive, there is still evidence of lipid peroxidation products in the plasma, although there is no indication where the oxidation occurred.

The Evidence against

The evidence for the oxidation hypothesis is mounting though it is important to note that there is evidence that contradicts these findings and does not support the oxidative hypothesis under discussion. Williams and Fischer [6] argue that the following evidence indicates caution when interpreting the clinical role of Ox-LDL in CVD:

—Some of the animal studies failed to show improvement from antioxidant supplementation, this was shown even though there was reduction of LDL oxidation in vitro or in the vessel wall [3, 48].

—It appears that the oxidized material does collect in the vessel wall, but only after nonoxidized material accumulates initially, this questions the requirement of oxidation to initiate the atherosclerotic process [49]. Also, high-density lipoprotein (HDL) retards the progress of atheroma development but does so without altering oxidative processes.

—In some studies it has been shown that the retardation of the scavenger receptor increases the development of the atheroma process [50] and the overexpression of this receptor decreases the development of the atheroma

[51].This is contrary to the expectations of the oxidative hypothesis.

—Studies that have genetically manipulated the two enzymes believed to be involved with the oxidation of LDL, myeloperoxidase, and 15-lipoxyge- nase, in an effort to ameliorate their oxidative effect on LDL, have instead shown a protective role of these enzymes in some animal models of atherosclerosis. This is contrary to their believed role in the oxidative modification of LDL and the subsequent chain of events that leads to lesion development and atherosclerosis [6].

—Lastly, the correlation between autoantibodies to Ox-LDL and CVD in humans have not been consistent [3] indicating that there may be another explanation for the correlation when it does occur. Williams and Fischer

[6]state that from their studies hyperlipidemia and other factors correlate with the antibody concentration, but vascular disease does not.

Perhaps the most damning evidence against the hypothesis is the lack of efficacy of the myriad of antioxidant clinical trials that have sought to show that antioxidant supplementation halted or prevented the development of atherosclerosis. However you term the criticism of the antioxidant trials they did not achieve a clinical outcome and, for antioxidant supplementation to be effective, a clinical outcome that is measurable is required [3,6,52]. As atherosclerosis is essentially an inflammatory illness, any antioxidant used must perhaps also have an anti-inflammatory effect.

Antioxidant Trials

The body of evidence that was mounting for the effectiveness of antioxidants in atherosclerosis in animals, and the support provided by the epidemiological evidence in favor of antioxidant protection of LDL from oxidation has been followed up with large randomized clinical trials of antioxidant supplementation. These trials were largely secondary prevention trials where the patients recruited had known vascular disease, active tobacco use, asbestos exposure, or documented previous malignancies.

The results of the large clinical antioxidant supplementation trials appeared to disprove the role of antioxidants in preventing or halting CHD.

The conflicting evidence from the animal, in vitro, and clinical trial appears to have been resolved when a meta-analysis of these key studies was undertaken by Vivekananthan et al. [52], involving 138,113 persons enrolled in the β-carotene trials and 81,788 enrolled in the vitamin E trials using all cause mortality as the clinical end point in common. This analysis showed no clinical benefit from the supplementation of antioxidants in these randomized placebo-controlled clinical trials. In the vitamin E analysis the authors divided the trials into primary prevention and secondary prevention trials and this also showed no clinical benefit from the supplementation of antioxidants. This lack of clinical benefit was seen even when the antioxidant supplement was shown to increase antioxidant plasma concentration, to decrease oxidation lag time ex vivo, and to lower the levels of circulating Ox-LDL [53].

This is indeed damning evidence against the oxidation hypothesis of atherosclerosis. However, some authors argue that it is not as damming as it may seem. According to Parthasarathy et al. [33] the failure is more in the clinical trial study design, specifically in the inappropriateness of the outcome measures chosen, than in the oxidation hypothesis itself.

It is important to note that the decision to focus on vitamin E and β-carotene may not have been the appropriate route to take. There is very little evidence to show that either of these naturally occurring antioxidants is the most efficacious in protecting LDL from oxidation [45]. In particular, the stage of atherosclerotic development where plaque rupture occurs, although oxidation could indeed have a role in the lead up to the rupture, prevention of oxidation at this stage would only have limited effect and to achieve that limited effect large amounts of chain-breaking antioxidants would be required.

As most of the clinical trials have been secondary prevention trials, with established coronary artery disease (CAD) beyond fatty streak lesion formation, relating these patients to the oxidized hypothesis may be beyond the scope of the hypothesis. Indeed in the meta-analysis only one of the trials studied was a primary prevention trial of vitamin E alone, therefore the negative/neutral results were driven by the secondary prevention trials perhaps biasing the results [52].

528 Lesley MacDonald-Wicks and Manohar Garg

So, the end points of these trials, such as plaque rupture or MI do not relate to oxidation, and are not an appropriate end point to measure the OxLDL hypothesis. Markers of oxidative stress (autoantibodies to oxidatively modified proteins) and surrogate markers of arterial and tissue response (VCAM-1 and MCP-1) can be seen to be reduced by antioxidants—perhaps these are more appropriate markers to test the Ox-LDL hypothesis [33]. Again the authors [4, 52] of the meta-analysis clearly state that none of the large clinical trials use oxidative stress end points in their studies. The authors speculate that the vitamins could have been incorrectly dosed, that is given to patients with low oxidative stress, or given too late in the course of CVD to be efficacious in inhibiting LDL oxidation. The authors [52] state in their article that, although they report no benefit from antioxidant vitamins, this does not disprove the Ox-LDL hypothesis, a position supported by Steinberg and Witztum [45]. It is clear from the evidence to date that, in order to prove or disprove the Ox-LDL hypothesis in relation to atherosclerotic development, it is necessary to have biomarkers that would accurately assess the level of oxidative stress the patient is experiencing and also measure the efficacy of the antioxidant used. Until such biomarkers are tried and tested it is not possible to design an effective clinical trial to test the oxidative hypothesis of atherosclerosis.

In some of the antioxidant clinical trials there were small increases in antioxidant potential in the plasma reported due to the supplementation. The studies often relied on “natural”’ antioxidants, such as vitamin E or β-carotene, instead of pharmacologically devised, high-powered antioxidants where a clinically significant increase in antioxidant potential might have been able to be achieved. One study by Meagher et al. [69] indicated that despite very high doses of vitamin E, there was still evidence of lipid peroxidation products in plasma indicating that perhaps vitamin E is not as powerful a lipid protecting antioxidant as often purported.

There is building evidence that the studies performed to date have targeted the wrong micronutrients. Perhaps vitamin E and carotenoids do not have a protective role, but rather are markers for other nutrients also present in the foods. Folate may be of interest due to its role in the reduction of homocysteine concentrations, a risk factor of CHD. Although the mechanism whereby homocysteine is a risk factor of CHD is not clearly elucidated, it is believed to be related to modulation of nitric oxide (NO) production in the endothelium. Recently it has been shown that folate could be operating independently of homocysteine concentrations. It appears that the addition of folate to ex vivo LDL increases lag time of Cu2+ OxLDL [54]. This effect cannot be attributable to NO production or the indirect effect on homocysteine concentrations, but rather to the scavenging ability of folate itself.

Of particular interest are the phenolic compounds found in foods. One explanation for the French paradox (i.e., a seemingly high intake of dietary

fat, in particular, saturated fat, combined with low population rates of heart disease) is the phenolic substances present in red wine [55]. Also of particular interest are the polyphenols present in all plants, but not in vitamin supplements. Some studies have shown an inverse relationship between flavonoid intake and CVD [56, 57]. In particular, flavonoids can prevent the oxidation of LDL [58]. Flavonoids and other polyphenolics do have antioxidant behavior, but interestingly they also have other mechanisms whereby they could influence the development and progression of CVD. Flavonoids can inhibit platelet aggregation and adhesion, inhibit enzymes in lipid metabolism, and inhibit endothelium-dependent vasodilation. These molecules are also investigated for their role in modulating immune cell function and act in a anti-inflammatory manner [59]. Many of these studies are not well controlled, thus interpretation of the data must be done with care.

Oxidative Stress: Friend or Foe?

To say that oxidative stress is ubiquitously problematic is to ignore the many important cellular and protective functions that oxidation performs. This can be highlighted by the role of the oxidative burst in the immune system. Recently Williams and Fisher [6] reported a significantly important role for ROS and RNS in the posttranslational destruction of apolipoprotein-B regulation, and as such would have a role in the normal regulation of this potentially atherogenic protein. It is also clear that ROS are required for the downstream phosphorylation of tyrosine, and insulin signaling in hepatocytes indicating that oxidative stress is a normal and necessary function of insulin signaling in hepatocytes [6]. It should be noted that in the pancreas ROS are responsible for the destruction of pancreatic islet cells and due to this the decreased production of insulin. This is not a desirable result. It is clear that in some circumstances increased antioxidant presence would be desirable and lead to the prevention of chronic illness, and equally in other location or circumstance this would not be the case [6].

Susceptibility of LDL to Oxidation

Esterbauer and colleagues [60] have shown that in lipid oxidation within the LDL particle the rate of diene generation is initially slow and the induction time (lag period) to formation may vary but, once this period has been passed, the rate of diene generation reaches a maximum and continues for some time. This leads to questions about what aspects of the LDL particle could influence the lag period and also the maximum rate of peroxidation (or the slope of the curve).

530 Lesley MacDonald-Wicks and Manohar Garg

Type of Fat in the Diet

The Ox-LDL hypothesis holds oxidation to be the key process in the initiation and proliferation of atheromas in the intima of the arterial wall. This is in direct contradiction of the evidence that dietary intake of polyunsaturated fats, which are incorporated into the LDL molecule and are highly oxidizable, are protective of heart disease. Similarly, exercise and moderate alcohol consumption are also considered to be protective of heart disease, and themselves are initiators of oxidative stress.

Many studies from the 1990s indicated that the type of fatty acid in the LDL molecule could radically affect susceptibility to oxidation. From these studies it is clear that the presence of monounsaturated fatty acids (MUFA) in LDL particles decreased the presence of oxidation products, indicating the oleic acid in particular is resistant to oxidation in comparison to linoleic acid. In a study by Bonanome et al. [61] the ratio of oleic acid and linoleic acid in the LDL particle was inversely related to the peroxidation rate. The presence of oleic acid increased the induction time somewhat, but the main effect was to reduce the slope of the curve during the peroxidation period [62]. This indicates that when the antioxidant present in the particle is exhausted, the particle with greater oleic acid concentration resist oxidation in comparison with a high polyunsaturated fatty acid (PUFA) intake. The fatty acid composition of the lipoprotein will affect the stability of these lipoproteins independent of antioxidant concentration [61]. In fact in a study by Dimitriadis et al. [63] showed that the LDL from a diabetic population was more susceptible to oxidation than the control equivalent. This study showed differences in the fatty acid composition of the LDL with increased free/esterified cholesterol ratio in diabetic patients.

Studies by Williams and Fisher [6] indicate that n−3 fatty acids, which are highly unsaturated and therefore susceptible to oxidation, inhibit Apo-B secretion in the liver through stimulating its degradation by the postendoplasmic reticulum presecretory proteolysis (PERPP) process. This process is dependent on the oxidation of PUFAs and is potentially protective of atherosclerotic development as the Apo-B moiety is very atherogenic. Similarly, saturated fatty acids, such as myristic acid a 14 carbon saturated fatty acid, are not susceptible to oxidation, have been found to stimulate secretion of Apo-B containing lipoproteins and to protect them from the PREPP process [64].

Presence of Antioxidants

Most of the animal experiments looking at the role of antioxidant in protecting the LDL from oxidation were done using probucol. Probucol is a hypolipidemic agent with impressive reducing potential. Due to the success of probucol, it was decided to proceed with large clinical trials on antioxidants, and in fact the decision was made that all antioxidants, whether soluble in

lipid or aqueous solutions, could be grouped together due to their antioxidant actions being similar [45].

What is clear is that the addition of antioxidants lengthens the time to induction of LDL oxidation in vitro, that is the lag period, but does not affect the maximum slope of the curve during the peroxidation period [62]. Vitamin E was the most frequently used antioxidant in the clinical trials, however the evidence for a role for vitamin E was at best unclear. From animal trials the evidence is equivocal, there was some protection from atherosclerosis in the mouse model [27], but the rabbit models have not been as effective [65]. In ex vivo measurement of LDL oxidation there is only brief increase in lag time (15–20 min), whereas probucol has much greater peroxidation delaying ability (up to 20 h or more). A study by Jialal and Grundy [66] did show that the propagation of LDL was reduced by vitamin E supplementation, and in populations with increased oxidative stress such as cigarette smokers [67] and hemodialysis [68] vitamin E has been effective in protecting LDL from oxidation. This is not the case in a healthy population indicating that there is only benefit from taking vitamin E in situations where there is increased oxidative stress [69], or perhaps in a population that is deficient in vitamin E initially.

Vitamin E has been shown to be reduced by coantioxidants, such as vitamin C and uric acid [60], but in situations where the coantioxidants are lacking vitamin E has been shown to be a prooxidant [3]. Extensive research into the dual ability of vitamin E to be both proand antioxidant has lead to the conclusion that in cases of strong oxidative challenge vitamin E acts as an antioxidant with little concomitant lipid peroxide formation. When in mildly oxidizing conditions, vitamin E is slowly consumed and there is a build-up of lipid peroxides in the LDL molecule but there is no conversion to secondary lipid oxidation products (such as aldehydes) [3]. It is these secondary lipid peroxidation products that are highly atherogenic. So, in vivo, the vitamin E radical replaces the lipid peroxide radical, and is the promoter of autooxidation, but at a slower rate than incurred if the lipid peroxide radical was the initiator. Vitamin E is very reactive with free radicals, and the presence of vitamin E determines the rate of oxidation, the more vitamin E the more oxidation that occurs, and vitamin E acts as a prooxidant [3]. It is likely that the in vivo environment reflects the milder oxidation condition. This leads to the conclusion that vitamin E is not necessarily the most efficacious antioxidant to study the protection of LDL from oxidation or to test the LDL oxidation hypothesis of atherosclerosis [45].

β-Carotene is often used as a biomarker of fruit and vegetable intake, and fruit and vegetable intake is inversely related to CHD in many epidemiological studies [70]. It is not unreasonable to assume that β-carotene is protective of CHD itself. β-Carotene is not an effective chain-breaking antioxidant in comparison to vitamin E, it is an effective singlet oxygen quencher [3], so using β-carotene in trials to prevent LDL oxidation is not useful. It is particularly

532 Lesley MacDonald-Wicks and Manohar Garg

important that β-carotene should not be used to prove or disprove the OxLDL hypothesis of atherosclerosis [45].

The large clinical trials show there is no evidence for using vitamin E and β-carotene supplements in established CHD. This indicates that more potent antioxidants are required to test the role of Ox-LDL in atherosclerosis, Parthasarathy et al. [33] suggests antioxidants that are not naturally occurring, such as probucol, may be better suited to testing this hypothesis.

There is some evidence that folate protects against CHD other than through its potential for lowering homocysteine levels. Ex vivo testing of the antioxidant potential of folate suggest there may be a role for folate as an in vivo antioxidant [55]. Although it appears that there is little reducing power of the polyphenolics compared to other more traditional antioxidants in vivo [3], there have been some successful studies showing their efficacy in protection from development of atherosclerosis. Some studies suggest that the flavonoids present in onions, apples, and tea provide some protection against CHD [58], although the results are not consistent across all studies.

Perhaps the protective effect of the standard antioxidants, such as vitamin E and flavonoids, does not rely solely on antioxidant ability, this is particularly true as vitamin E has been isolated in atherosclerotic lesions [71]. Vitamin E has many antiatherosclerotic properties; it is an antithrombotic, it inhibits smooth muscle cell proliferation [72], and decreases monocyte adhesion to endothelium, decreases platelet adhesion and proliferation, stabilizes membranes, and affect intracellular calcium mobilization [9, 73]. It has also been shown to modulate superoxide formation and leakage from the mitochondria by preventing electron leakage, directly or indirectly impacting on the superoxide-generating systems, and by scavenging the superoxide that is generated [74]. Flavonoids are anti-inflammatory molecules and this is the role they may have in protecting against atherosclerosis, an established inflammatory condition.

Methods of Detecting Oxidized LDL

Indirect Methods

Initially the oxidation of LDL was measured by the ex vivo method, whereby oxidation of the unsaturated lipids in human LDL molecule was catalyzed by copper and the change in absorption at 234 nm measured. The change in absorption was due to the production of conjugated dienes, which absorb at a different frequency. Similarly the presence of thiobarbituric acid-reacting substance (TBARS) formation can be measured in the presence of copper ions. This is a severe form of oxidation that results in an LDL molecule that is capable of forming foam cells. In this case the LDL is completely devoid of antioxidant potential. This is contrary to the findings that LDL in fatty lesions in humans contain adequate concentrations of antioxidants, includ-