Lipid peroxidation and decomposition-Conflicting roles in plaque vulnerability and stability

Abstract

The low density lipoprotein (LDL) oxidation hypothesis has generated considerable interest in oxidative stress and how it might affect atherosclerosis. However, the failure of antioxidants, particularly vitamin E, to affect the progression of the disease in humans has convinced even staunch supporters of the hypothesis to take a step backwards and reconsider alternatives.

Preponderant evidence for the hypothesis came from animal antioxidant intervention studies. In this review we point out basic differences between animal and human atherosclerosis development and suggest that human disease starts where animal studies end. While initial oxidative steps in the generation of early fatty streak lesions might be common, the differences might be in the steps involved in the decomposition of peroxidized lipids into aldehydes and their further oxidation into carboxylic acids. We suggest that these steps may not be amenable to attenuation by antioxidants and antioxidants might actually counter the stabilization of plaque by preventing the formation of carboxylic acids which are anti-inflammatory in nature. The formation of such dicarboxylic acids may also be conducive to plaque stabilization by trapping calcium.

We suggest that agents that would prevent the decomposition of lipid peroxides and promote the formation and removal of lipid hydroxides, such as paraoxonase (PON 1) or apo A1/high density lipoprotein (HDL) might be more conducive to plaque regression.

Oxidation hypothesis of atherosclerosis

Atherosclerosis is the major manifestation of cardiovascular diseases. The role of lipoproteins in atherosclerosis development has been a topic of interest for over five decades. It is now established beyond doubt that high levels of low density lipoprotein (LDL) and low levels of high density lipoprotein (HDL) contribute significantly to the development and progression of cardiovascular diseases [1–5].

The biochemical processes that contributed to the formation of early atherosclerotic lesions, the fatty streak lesions, are still under debate. The LDL oxidation hypothesis was put forward in the eighties to explain the formation of fatty streak lesions and over 4000 publications and countless reviews have appeared on the topic to date [6–11] providing evidence for the presence of oxidative processes in human disease. The potential steps involved in the development of early atherosclerotic lesion as outlined by the hypothesis are listed in table 1.

Table 1.

Steps involved in fatty streak lesion formation:

1.	Increased plasma and intimal LDL.
2.	Oxidation of lipids associated with LDL
3.	Chemotactic recruitment of monocytes and their differentiation into macrophages.
4.	Modification of apolipoprotein B100 to a negatively charged form (Oxidatively modified LDL) by products derived from lipid peroxidation.
5.	Uptake of oxidatively modified LDL by macrophages.
6.	Retention of lipid laden macrophages in the intima.

The oxidized LDL

The hypothesis originated from a simple observation that in vitro incubation of macrophages with oxidized LDL and not with native LDL led to intracellular cholesterol ester accumulation [12–15]. The oxidation of LDL is a complex process during which both the protein and lipids undergo oxidative changes and form complex products. The peroxidized lipids decompose generating both free and core aldehydes and ketones that covalently modify ε- amino groups of lysine residues of the protein moiety [16–17]. The latter not only generates Schiff’s bases, thus modifying charges on the amino acids, but also results in both intra- and inter- molecular cross-links between proteolyzed apo B. The oxidative modification is not unique to LDL as other lipoproteins such as very low density lipoprotein (VLDL), beta very low density lipoprotein (β-VLDL), and even high density lipoprotein (HDL) have been suggested to undergo similar oxidative changes [18–22] accompanied by changes in their pro- or anti-atherosclerotic behavior. The take home message from these studies was that the modification of LDL might be the key determinant of lipid uptake by macrophages.

Potential mechanisms of oxidation

The mechanism(s) by which oxidation of LDL might occur in vivo or the extent of oxidation are subject to speculation and are predominantly derived from in vitro studies (Table 2). Practically every type of oxidative system has been suggested to contribute to the oxidation of LDL. In vivo evidence in support or against any of these mechanisms is ambiguous. For example, overexpression/knockout strategies involving specific enzymes often yielded conflicting results. Overexpression of arachidonate 12/15 lipoxygenase in rabbits protected against atherosclerosis [23], while overexpression and knock out of this enzyme in mice yielded results in favor of the involvement of this enzyme in oxidative processes [24, 25]. Similarly, myeloperoxidase (MPO) gene knockout in mice was expected to protect against atherosclerosis, while the results showed an enhanced atherosclerosis [26]. However, overexpression of human MPO in mice appeared to marginally increase atherosclerosis susceptibility [27]. These studies were often interpreted as evidence in favor of distinct human MPO protein; however, it is hard to understand how the MPO activity could be different from different sources although variations might exist in protein amino acid sequence and subunits unless the inference is that only in humans MPO plays a role and not in mice. Similarly, conflicting results have emerged from studies that have manipulated antioxidant enzymes such as superoxide dismutase [28].

Table 2.

Potential mechanisms for the oxidation of LDL

Mechanism	Ref.
1. Lipoxygenase reaction	23,24,25
2. Copper and ceruloplasmin mediated oxidation	28, 30
3. Iron mediated oxidation	235, 236
4. Peroxidase mediated oxidation including Myeloperoxidase and heme	26,27,35
5. Peroxynitrite mediated oxidation	29,30
6. Thiol-dependent oxidation	38,39
7. Xanthine Oxidase, NADPH Oxidase and other superoxide generators	28
8. AAPH or other means of radical generation including cytochromes	29, 30

The oxidation systems vastly differ in their properties and many of them require co-oxidants. MPO mediated oxidations require co-oxidants such as H₂O₂ or lipid peroxides and thus over expression or knockout of MPO alone would have little effect on the oxidative status. Some oxidants damage the proteins more readily than others. For example treatment of LDL with 2,2′-azobis(2-amidinopropane) dihydrochloride (AAPH), a radical generator or peroxynitrite results in more protein oxidation than lipid peroxidation [29, 30]. Similarly, peroxidase catalyzed oxidation generate very little aldehyde products as compared to metal catalyzed oxidations [31]. Lipoxygenase reactions which are exclusively intracellular, might require additional reactions or transfers before or after lipoxygenase reactions occur [32]. To our knowledge, no one has characterized oxidized LDL isolated from animals or humans to the extent that mechanistic insights could be derived, although there are numerous claims of the presence of an electronegative LDL (presumably minimally oxidized LDL) in human and animal plasma [33, 34].

Antioxidants and atherosclerosis-prooxidant nature is ignored

While all these reactions are oxidative in nature, they are not uniformly amenable to inhibition by traditional antioxidants. Vitamin E or simple phenols such as tyrosine or estradiol actually enhance peroxidase-mediated oxidation of LDL [35–37]. This has been taken as evidence for an intermediate pro-oxidant role for antioxidants. Similarly, many thiols may enhance or inhibit the oxidation of LDL depending on the peroxide/metal content of LDL [38–39]. The bottom line is we neither know of the oxidation system(s) that might be involved nor the effects of specific antioxidant chemicals on such reactions.

Antioxidants and animal models of atherosclerosis

One of the strongest evidence for the oxidation hypothesis was derived from animal studies that tested the effects of antioxidants on atherosclerosis. A variety of antioxidants have been tested using a number of different animal models. These include WHHL rabbits, cholesterol-fed rabbits, hypercholesterolemic hamsters and monkeys, and various mouse models of atherosclerosis. Animal models were developed to study the formation of early atherosclerotic lesion in isolated environment so that contribution of individual risk factors and associated biochemical events could be segregated and studied. The human atherosclerosis is a chronic disease associated with the ageing process. During the lifelong process spanning decades, numerous risk factors contribute to atherosclerosis. The necessity to study the disease in a short and meaningful period of time (combined with the human impatience and the pressure of the time constraints of the scientific grant funding process) has created numerous models that have shrunk the atherogenic process to a mere few months. All these models depend on the increases in plasma cholesterol levels beyond what one would see in a human being. These models were developed with the sole purpose of studying the contribution of cholesterol carrying lipoproteins or other associated processes to the formation of macrophage-rich fatty streak lesions and have performed remarkably well to this end. In fact, the end points of these studies are often lesion size/areas or lesion lipid parameters.

These animal models and the use of numerous antioxidants (Table 3) have suggested that antioxidants do attenuate the atherogenic process. However, problems related to the very high concentrations of antioxidants (sometimes even more than that of cholesterol in the diet itself!) and toxicity of many different types of antioxidants (BHT, DPPD) were often overlooked. There was a race to demonstrate that every chemical with known antioxidant property was anti-atherosclerotic.

Table 3.

Animal models and antioxidants in the prevention of atherosclerosis:

Model	Antioxidant(s) used
LDL Receptor Knockout mice	Vitamin E, beta-carotene, vitamin C, probucol, S-nitroso-N-acetylcysteine, nitric oxide-releasing aspirin, 2,3-dihydro-5-hydroxy-2, 2-dipentyl-4,6-di-tert-butylbenzofuran.
Apolipoprotein E knockout mice	N-acetylcysteine, vitamin E, probucol, N,N'-diphenyl 1,4-phenylenediamine (DPPD), calcium channel blockers- lacidipine, disodium ascorbyl phytostanyl phosphate, taurine, tocotrienol, catechins.
WHHL Rabbits	Probucol, vitamin E, and fluvastatin, an HMG-CoA (hydroxy-3-methylglutaryl coenzyme A), ubiquinone-10, N-acetylcysteine, astaxanthin, water-soluble vitamin E derivative, N,N'-diacetyl-L-cystine.
Cholesterol-fed NZW rabbits	Probucol, vitamin E, polyphenol, U74006F, BHT, N,N'-diphenyl 1,4-phenylenediamine (DPPD)
Cholesterol-fed hamsters	Isoflavones, vegetable extract (Oxxynea), alpha-lipoic acid (ALP), Vitamin E, Green tea, probucol, grape extracts.
Cholesterol-fed guinea pigs	Vitamin C, Vitamin E, Huajiao volatile oil
Monkeys	Vitamin E, probucol, various soy components.

Why did human trials using antioxidant fail?

Combined with epidemiological evidence and success in several animal trials in a number of species using a variety of antioxidants, the hypothesis appeared to be on solid ground. This euphoria of initial success led to clinical trials that were planned mostly to be the "first" to validate the hypothesis. These trials have been reviewed in many articles [40–42] and will not be discussed here. These ill-conceived clinical trials were performed without understanding the a) enzymes and factors involved in the oxidative process including those contributed by associated risk factors, b) nature of the oxidative process and associated changes in the lipid and protein moieties, c) whether other lipoproteins, particularly HDL was affected by these processes, d) whether antioxidants ameliorated any specific steps in the oxidative process, e) the compatibility of the oxidative process/antioxidant action with current therapy, f) the steps involved in the atherogenic process and the role(s) of oxidative processes, if any, on these steps, g) the distinction between the early fatty streak lesions and the advanced vulnerable lesions that lead to plaque rupture, thrombosis, and acute coronary events, and h) above all the distinction between the animal models and human atherosclerotic development. The latter is of great significance as the antioxidant trials completely ignored pathological findings and clinical outcomes.

Tocopherol-vitamin or antioxidant?

Obviously vitamin E was the chosen antioxidant as the expectation of the general public and scientists was that a natural antioxidant (such as vitamin E) would have less undesirable effects. Accordingly, a number of clinical trials were performed using vitamin E alone or combined with vitamin C or other antioxidants at pharmacological doses often disregarding the definition of the term “vitamin” that suggested a requirement at miniscule doses. Some of these trials were piggyback studies on cancer and other diseases. These trials with vitamin E have not been overwhelmingly supportive of the hypothesis. Understanding the sequence of events that led to the hypothesis, the molecular basis of antioxidant action, the potential interaction with other drugs, and the stage of the disease process to which oxidation was suggested to contribute, dosage, the transport mechanisms involved in the release and utilization of vitamin E in the body, its interaction(s) with cytochrome systems, and its role as a prooxidant in peroxidase-mediated oxidations, played little role in the design of the study. More importantly, little if any attention was paid to factors that would affect and control oxidation in vivo.

Oxidized LDL in humans

While the potential use of antioxidants in controlling the disease process has lost its supporters, more and more evidence supports a role for an oxidative process in atherosclerosis and associated risk factors. The prevalence of antibodies that recognize epitopes of Ox-LDL seem to correlate well with the disease process [43–48]. These antibodies also detect such LDL in the plasma of subjects who have higher risk associated with atherosclerosis [49–52] and seem to recognize oxidative tailored lipid epitope. Interestingly, such antibodies are not available commercially to study animal atherosclerosis and the human oxidized LDL kit does not appear to work with animal blood samples. This does not appear to be related to the specific antibodies used in the kit sounding yet another ominous bell of caution that the fatty streak lesions seen in animals or in children might not involve the generation of such epitopes.

Myeloperoxidase products in humans

Oxidative enzymes such as MPO and products derived from their actions seem to predict the vulnerable population and seem to correlate with the severity of the disease [53–57]. A variety of products, such as chloro and nitro tyrosines have been detected at increased levels in the plasma of subjects with established coronary artery disease as compared to control subjects. Considering that Vitamin E might actually enhance oxidation [35, 58–60], the clinical trials might have actually endorsed the hypothesis rather than shooting it down. Perhaps, when inhibitors of specific oxidases are available, additional trials might be warranted to induce the development of anti-atherosclerotic pharmacological agents.

Human atherosclerosis and animal models

The human form of atherosclerotic disease differs considerably from the disease models in animals. As mentioned before, the animal models are predominantly setup to develop and study the very early fatty streak lesions with macrophage foam cells. Such lesions start very early in life in humans (even in infants and children) below the age of 15 [61–68]. Maternal hypercholesterolemia seems to strongly influence the development of atherosclerotic lesions in the young. Again, interestingly, while plasma LDL levels positively correlate to the severity of coronary artery disease, infants and children have much lower LDL levels as compared to adults and the elderly population suggesting that the correlation between age and plasma LDL levels could be a statistical coincidence rather than one of clinical significance except when specific sub-fractions with increased atherogenic potency such as oxidized LDL or small dense LDL account for the increase. However, the presence of fatty streak atherosclerotic lesions alone in the absence of significant other risk factors might be clinically inconsequential especially considering lower plasma lipid levels and compounding risk factors in children as compared to adults. Also, quantitatively, adults may have greater lesion surface area as compared to children, in addition to differences in the type of lesions. Acquisition of additional risk factors might significantly affect the progression or regression of the early childhood lesions.

Progression of human disease-Calcification

The progression of human atherosclerotic lesions has been well studied by several pioneers [69–73]. They unanimously conclude that the macrophage-lipid-rich early fatty streak lesions progress to extra-cellular lipid rich raised complex lesions later in adult life with or without calcification [69–71]. Atherosclerotic calcification begins as early as the second decade of life, just after fatty streak formation. The lesions of younger adults have revealed small aggregates of crystalline calcium among the lipid particles of lipid cores. Calcium deposits are found more frequently and in greater amounts in elderly individuals and in more advanced lesions. In most advanced lesions, when calcification dominates, components such as lipid deposits and increased fibrous tissue may also be present. The extent of calcification also has been correlated with plaque burden. It has been demonstrated that coronary calcification is detected in the vast majority of patients with a first MI. Age-related calcium score is highly predictive for myocardial infarction event risk in prospective studies and the association between coronary calcium and CHD events usually remains significant after adjusting for other CHD risk factors. These study results support the concept that coronary calcium is associated with a relative profound independent increased risk of CHD events in women and men. However, there has been a great deal of debate about the role of calcification in plaque rupture [74–79].

Calcification has been noted as both intra- and extra- cellular deposits [78], the latter predominantly is associated with extracellular lipids. Many researchers believe that coronary arterial calcification may represent an attempt to protect the weakened atherosclerotic plaque prone to rupture [77, 79]. Calcified lesions and fibrotic lesions are much stiffer than cellular lesions and are unlikely to be associated with sites of plaque rupture [77, 80]. In fact, a recent study showed that less than 15% of the ruptured human lesions showed evidence of associated calcification [80]. It is speculated that the plaque rupture often occurs at the interface between a calcified and non-calcified atherosclerotic areas of the lesion. Thus, calcification could be seen as a potential stabilizing force that may increase the biochemical stability of the plaque by imparting rigidity while at the same time decreasing the plaque’s mechanical stability.

Calcified human lesions

Neither the mechanism(s) involved in atherosclerotic calcification nor the nature of the calcium deposits has been positively identified. Calcification is identified mostly by Von Kossa staining or by physical imaging techniques such as computerized tomography. While they give valuable information, they fail to indicate the chemical form in which calcium is deposited. It is generally believed that calcium phosphate (hydroxyapatite, Ca₃[−PO₄]₂₋ ×Ca[OH]₂), precipitates in diseased coronary arteries by a mechanism similar to that found in active bone formation and remodeling [77, 81–87]. Others believe that specific sulfated proteoglycans with high affinity for calcium are involved in the deposition of calcium in lesion areas [88–91]. Support for this comes from the variety of proteoglycans, such as decorin, that are synthesized and secreted by smooth muscle cells (cells that are proximally associated with calcification). The synthesis of these proteins is also regulated by factors that influence the atherosclerotic process thus, providing credibility to their involvement.

Inflammation and atherosclerosis

Yet another major difference between human and animal atherosclerosis is the contribution of inflammatory cytokines and matrix digesting enzymes (MMPs) to the advancement of the vulnerable plaque [92–98]. Most of these cytokines and MMPs are induced by oxidative stress [99–125] suggesting that ongoing oxidative processes might be involved in the generation of the vulnerable plaque. Inflammatory cytokines, matrix disruption, release of lytic lipids such as lyso phosphatidylcholine, aldehydes, release of reactive oxygen species, and proteolytic enzymes have been suggested to play major roles in plaque vulnerability and rupture.

Risk factors and oxidative stress

Well-known risk factors might contribute to such vulnerability by promoting the formation of inflammatory cytokines and the generation of oxygen radicals (Fig. 1). Diabetes, ageing, hypertension, smoking, physical inactivity and obesity, hyperlipidemia, particularly hyper cholesterolemia including the prevalence of small dense LDL, hyperhomocysteinemia, are recognized risk factors for atherosclerosis. It is interesting to note that most of the risk factors for atherosclerosis also are risk factors for heart failure suggesting a chronic perturbation of the immune system and endothelial activation. Innumerable studies have documented that every risk factor for atherosclerosis is associated with an increased oxidative stress. One would assume these genetic, environmental, and life style associated risk factors somehow would increase oxidative and inflammatory stress and promote the progression of early lesions to the vulnerable plaque leading to plaque rupture.

Figure 1.

a) Diabetes and oxidative stress

Oxidative stress has long been associated with diabetes and hypertension. Both clinical and experimental studies show the excessive reactive oxygen species generation in hyperglycemic conditions. The toxicity of diabetic serum has long been attributed to the presence of oxidized lipids. Hyperglycemic conditions are associated with high levels of oxidant production [126–128]. The ability of glucose to generate oxygen radicals and H₂O₂ non-enzymatically and enzymatically has been known for a long time [128–130]. In addition, hyperglycemia appears to induce lipid oxidative enzymes such as 12/15 lipoxygenases and NADPH oxidases [131, 132]. In addition, glucose also reacts with amino groups of proteins and lipids to generate advanced glycation end products (AGEs) [133, 134]. These products among other properties also induce oxygen radical production thus amplifying the production of H₂O₂ and LOOH [135, 136].

b) Hypertension and oxidative stress

Similarly, molecules associated with hypertension, for example nitric oxide and angiotensin II are intimately associated with oxidative stress [137–140]. While the former’s beneficial vaso dilatating action is cancelled by oxygen radicals generating a more virulent form of oxidant, peroxynitrite [141], angiotensin II potentiates such inactivation by generating superoxide radicals by way of enzymes that utilize NADPH oxidation [137].

c) Cigarette smoke and oxidation

The fundamental mechanism by which cigarette smoke is thought to contribute to atherosclerosis has been suggested to be mediated by causing oxidative stress [142–144]. Innumerable articles have alluded to the presence of free radicals in the gas/tar phase of cigarette smoke [142–144]. In addition, the ability of cigarette smoke to deplete plasma antioxidants and induce the oxidation of LDL in vitro and ex vivo has also been well documented [145–148].

d) Physical activity and oxidative stress

Recently there have been several exciting findings in relation to obesity/physical inactivity and oxidative stress. Leptin has been shown to induce oxidative stress as well as being induced by oxidative stress [149, 150]. While vigorous physical activity in early stages might deplete body’s antioxidants and promote oxidative stress, sustained and long term physical activity has been noted to raise antioxidant defense and reduce oxidative stress [151–158].

The propensity of small dense LDL to undergo increased oxidation has also been observed [159–161]. While small dense LDL may well represent one of the most atherogenic forms of LDL, it remains to be seen whether its oxidation specifically has any relevance in vivo.

In many of the above considerations, the underlying assumption is that the oxidative stress contributes to the formation of oxidized LDL and that the components of oxidized LDL could act to stimulate the production of inflammatory mediators that are hallmarks of subjects with advanced atherosclerotic lesions. There is also an underlying expectation that new lesions are continuously generated and oxidative stress is not only involved in their generation but somehow promote their conversion to advanced vulnerable plaque.

Oxidative stress and body’s antioxidant defense

The failure of antioxidants (at least vitamin E containing supplements) to benefit human subjects with advanced disease might suggest that either the hypothesis is flawed or some form of oxidative stress might actually be beneficial. Could oxidative stress be beneficial? Some of the arguments in favor of a “beneficial” oxidative stress are illustrated in Table 4. First, oxidative stress has been known to induce body’s defense mechanisms by triggering the synthesis of antioxidant enzymes. The induction of catalase [162], GSH [163] and GSH peroxidase [164], nitric oxide synthase [165], MnSOD [166,167] and heme oxygenase [168,169] by oxidants has been well described in literature. We described the ability of oxidized fatty acids to induce apo A1 (a major component of HDL) synthesis by liver and intestinal cells [170] which also could signify a beneficial effect. We also recently described the induction of apo AV gene expression and a decrease in plasma triglycerides by oxidized fatty acids [171]. Some of these effects could be attributed to PPAR ligand like actions of oxidized lipids [170, 172–174], a point that is conveniently ignored when considering the deleterious actions of oxidized LDL. Oxidation also has been suggested to play a key role in the generation and removal of senescent apoptotic cells [175–177] and perhaps involved in the detoxification of many cardiovascular drugs including statins [178].

Table 4.

Potential benefits of oxidative stress:

1.	Oxidative stress and oxidized lipids have been shown to induce antioxidant enzymes such as, nitric oxide synthase, catalase, MnSOD, GSH synthesis, and heme oxygenase.
2.	Oxidized lipids induce apo A1 gene expression and synthesis.
3.	Induce apo AV gene expression and decrease plasma triglyceride levels. Also act as PPAR ligands and induce fatty acid oxidation
4.	Might be involved in senescent cell clearance and cellular signaling mechanisms.
5.	Might be involved and essential for the detoxification of many cardiovascular drugs.
6.	Might actually be involved in the beneficial actions of exercise, long chain polyunsaturated fat, and estradiol.

Exercise, PUFA, and Estradiol

More importantly, the three well-known modalities that are deterrents of cardiovascular diseases, exercise, the consumption of polyunsaturated fatty acids (PUFA), and the estrogenic hormones are intimately associated with oxidative stress. Vigorous physical activity has been shown to deplete antioxidants and promote oxidant production [151–158]. Similarly, PUFA is more readily oxidized as compared to saturated or monounsaturated fatty acids [179–181], and estradiol has been shown to potentiate the actions of neutrophils and MPO action [36, 37]. Consistently and repeatedly, physical activity and PUFA have been shown not only to prevent the development of atherosclerotic lesions but also to promote regression [182–187]. While the precise mechanisms of action of these are yet unknown, oxidative stress-induced antioxidant defense has been suggested to be an important factor [157].

Antioxidant activities of cardiovascular drugs

However, such actions must be weighed together with the fact that many of the currently available cardiovascular drugs are antioxidants in nature (Table 5). These drugs include statins [187, 188], calcium channel blockers [189–193], angiotensin II receptor antagonists [194], ACE inhibitors [195–199], compounds such as dobutamine [200], nitric oxide donors [201–202], salicylate [203, 204] and many others. Regardless of their action as antioxidants in vivo, their chemical nature as antioxidants cannot be ignored. It is likely that the metabolic profiles of these compounds are more favorable than that of vitamin E.

Table 5.

Cardiovascular drugs that have antioxidant properties

1.	Nitric oxide donors.
2.	Calcium channel blockers
3.	Angiotensin type II receptor blockers.
4.	Dobutamine.
5.	Statins.
6.	Certain ACE inhibitors.
7.	Salicylate and related compounds.

Too much aldehydes from peroxidized lipids?

We considered the possibility that the oxidation hypothesis did not extend “far” enough to include the decomposition of peroxidized lipids. The oxidation hypothesis took into consideration that the aldehyde products generated by the decomposition of peroxidized lipids covalently modify the €-amino groups of lysine residues to generate the oxidatively modified LDL. Considering that even the simplest PUFA, linoleic acid, could generate several aldehyde products (e.g. 2-hydroxy hexanal, 4-hydroxy nonenal, oxo-nonanoic acid), one would end up with a glut of such products. Earlier studies by Haberland and coworkers have suggested that derivatization of as little as 15% of lysine residues of apoprotein B₁₀₀ are enough to impart macrophage scavenger receptor recognition [16, 17]. A mole of LDL particle is expected to have about 350 mols of free amino groups. About 15% modification of these (excluding other oxidative structural changes) would be equal to the modification of approximately 50 amino groups. Considering that about 1 mol equivalent of LDL would have about or more than 0.5 mol equivalent of PUFA and each PUFA is capable of generating at least two more aldehyde products, one would have 20 or more equivalent of aldehydes per modifiable lysine residues. In addition, the lysosomal proteolysis would release aldehydes bound lysine peptides as well as free aldehydes due to the instability of Schiff bases in such environments. Of course such arguments consider that all oxidizable PUFA will undergo oxidation once oxidation is initiated and there would be a glut of free and core aldehydes.

Pro-atherogenic effects of aldehydes

The pro-atherogenic effects of aldehyde products have received relatively little attention as compared to the oxidized lipids themselves. The initial burst of activities centered on the modification process per se and the results seemed to indicate that small molecular water-soluble aldehydes such as malondialdehyde (MDA) readily diffused from the lipoprotein particle [205]. Pioneering studies by Esterbauer and coworkers brought attention to more reactive and lipophilic aldehydes such as 4-hydroxy nonenal (4-HNE) in to focus [206–209]. Besides being capable of protein modification and crosslinking 4-HNE is now recognized to affect the atherogenic process by a number of different ways. Aldehydes such as 4-HNE might cause enzyme inactivation, apoptosis, modify proteins associated with cell signaling, and induce pro-inflammatory cytokines [206–217].

Aldehydes to carboxylic acids

While 4-HNE attracted considerable attention, the other half of the decomposition product of oxidized linoleic acid (oxo-valeric acid, oxo-nonanoic acid and other oxo-acids and core aldehydes) has attracted little attention [218]. The core aldehydes (which are esterified to lipids such as cholesterol or lysophospholipids) are seen only as markers of oxidation or as antigenic lipids or as substrates for enzymes such as platelet activating factor acetyl hydrolase (PAF-acetyl hydrolase) or phospholipases. One such product (oxo-nonanoic acid) that is derived from the oxidation of linoleic acid is of great biological significance as it gets readily oxidized to nonane dioic acid (azelaic acid). Azelaic acid is a lipophilic dicarboxylic acid as opposed to short chain dicarboxylic acids such as malonic acid, despite the reactive methylene group reactivity of the latter. Therefore, it would be expected to be formed and to accumulate in greater amounts in lipid-rich domains.

Azelaic acid, a lipid peroxidation-derived lipophilic dicarboxylic acid

Azelaic acid would have two important anti-atherogenic properties: one, as other dicarboxylic acids, it would have great affinity for calcium and precipitate calcium in lipid-rich domains [219–222]. This might account for the unusually high calcification associated with lipid core domains. Second, azelaic acid has been noted to have anti-inflammatory properties and has been noted to decrease cytokine production by macrophages and inflammatory cells [223–225]. It also has been noted to reduce oxidant production by leukocytes [223]. Thus, it has been added as an active ingredient in topical skin medications, particularly as anti-acne agent [224–226].

The oxidation of oxo-nonanoic acid to azelaic acid is a simple oxidation step and is inhibited by antioxidants [227] analogous to the oxidation of other aldehydes [228–230]. The oxidation of aldehydes into acids is catalyzed by both enzymatic (aldehydes oxidase, xanthine oxidase) as well as non-enzymatic oxidation.

Could antioxidants inhibit the conversion of aldehydes to carboxylic acids?

Thus, in human secondary prevention trials with antioxidants, there is a distinct possibility that the high levels of antioxidants could have prevented the generation of anti-atherogenic and anti-inflammatory dicarboxylic acids. Such dicarboxylic acids also might be involved in calcification process and the inhibition of their formation by antioxidants would be conducive to the formation of non-calcified vulnerable plaque (Fig. 2). Until the specific nature of calcium deposits are identified in human lesions, one could only speculate the role of such dicarboxylic acids in calcification. Either they can complex calcium and thus could represent calcified lesions, or they could destabilize the calcium phosphate/calcium extracellular proteoglycan complex thus acting against calcification. It remains to be seen whether agents that would prevent the decomposition of lipid peroxides into long chain aldehydes or enhance the conversion of aldehydes into dicarboxylic acids would have a better cardiovascular outcome. The carboxylic acids, including dicarboxylic acids are readily oxidized in the mitochondria and peroxisomes and thus could be readily eliminated. Molecules such as apo A1/HDL or enzymes such as paraoxonases or glutathione peroxidase might be inherently anti-atherogenic by not only promoting the conversion of hydroperoxides into hydroxides as well as by complexing with the aldehyde products but also by removing the hydroxides from the artery for elimination [231–234] as suggested by Fogelman and associates.

Figure 2.

Summary

In this review we point out the differences between animal and human atherosclerosis and suggest that human atherosclerosis might have progressed beyond the point of oxidation. We suggest that the end products of lipid peroxidation, carboxylic acids (and not aldehydes) might influence calcification and inflammation. Antioxidants might prevent the formation of innocuous carboxylic acids from toxic aldehydes.