Vitamin E Oxidation in Human Atherosclerotic Lesions
Oxidation of low-density lipoproteins (LDL) is a key process in atherogenesis, and vitamin E (α-tocopherol, TOH) has received attention for its potential to attenuate the disease. Despite this, the type and extent of TOH oxidation and its relationship to lipid oxidation in the vessel wall where lesions develop remain unknown. Therefore, we measured oxidized lipids, TOH, and its oxidation products, α-tocopherylquinone (TQ), 2,3- and 5,6-epoxy-α-tocopherylquinones by gas chromatography–mass spectrometry analysis in human lesions representing different stages of atherosclerosis. We also oxidized LDL in vitro to establish “footprints” of TOH oxidation product for different oxidants. The in vitro studies demonstrated that tocopherylquinone epoxides are the major products when LDL is exposed to the one-electron (ie, radical) oxidants, peroxyl radicals, and copper ions, whereas TQ preferentially accumulates with the two-electron (nonradical) oxidants, hypochlorite, and peroxynitrite. In human lesions, the relative extent of TOH oxidation was maximal early in the disease where it exceeded lipid oxidation. Independent of the disease stage, TQ was always the major oxidation product with all products together representing <20% of the total TOH present, and the oxidation product profile mirroring that formed during LDL oxidation by activated monocytes in the presence of nitrite. In contrast, oxidized lipid increased with increasing disease severity. These results suggest that two-electron oxidants are primarily responsible for TOH oxidation in the artery wall, and that the extent of TOH oxidation is limited yet substantial lipid oxidation takes place. This study may have important implications regarding antioxidant supplements aimed at preventing LDL oxidation and hence atherogenesis.
The oxidation of low-density lipoproteins (LDL) in the artery wall is considered to be a key contributor to atherogenesis.1 In support of this, a wide variety of markers of oxidized lipid and protein have been detected in human atherosclerotic lesions, including F2-isoprostanes,2 cholesteryl ester hydro(pero)xides (CE-O(O)H),3,4⇓ and chloro- and nitrotyrosine.5,6⇓ The exact identity of the oxidant(s) responsible for LDL oxidation in vivo remains undetermined although several oxidants have been implicated, including transition metals, 15-lipoxygenase, myeloperoxidase-generated hypochlorous acid (HOCl), and reactive nitrogen species such as peroxynitrite (ONOO−).7
Antioxidants are potential antiatherogenic agents because they can inhibit LDL oxidation. Most attention has focused on vitamin E (α-tocopherol, TOH), as it is the major lipid-soluble antioxidant in LDL, yet the outcome of interventions with TOH supplements has been overall disappointing.8 Also, in vitro mechanistic studies have revealed a complex role for TOH during lipoprotein lipid oxidation. Thus, in the absence of compounds capable of reducing the α-tocopheroxyl radical (TO·) and exporting the radical from LDL, the vitamin can promote lipid oxidation by the process of tocopherol-mediated peroxidation (TMP).9–11⇓⇓
Despite some evidence,12 there is surprisingly little direct evidence establishing that TOH attenuates lipid oxidation in the artery wall. One approach to resolve this issue is to examine the extent and type of TOH oxidation and how this relates to the extent of lipid oxidation in lesions. Because TOH is one of the most reactive components of LDL, such measurements should also provide a sensitive gauge for the extent of LDL oxidation in the artery wall. Furthermore, by measuring the distribution of specific TOH oxidation products, important information can be obtained on the nature of the oxidant causing LDL oxidation in the artery wall. At present, information on TOH oxidation in human atherosclerotic lesions is limited to the measurement of α-tocopherylquinone (TQ) in carotid plaque.3,4⇓
We therefore measured TOH and its oxidation products—TQ, 5,6-epoxy-α-tocopherylquinone (TQE1), and 2,3-epoxy-α-tocopherylquinone (TQE2) (Figure 1) in human atherosclerotic lesions representing different stages of the disease. We also compared the profile of TOH oxidation products in these lesions with that obtained for LDL exposed to various oxidants in vitro. We show that TQ is the major oxidation product in both lesions and LDL oxidized by the two-electron oxidants HOCl and ONOO−. This, together with the observation that the total concentration of TOH oxidation products in lesions is almost 10-fold lower than that of nonoxidized (parent) TOH and CE-O(O)H, suggests that overall TOH oxidation is limited and vitamin E fails to prevent LDL lipid oxidation in the artery wall.
Materials and Methods
Isolation of LDL
Human Lesion Material
All human lesion material was obtained with approval of the local ethics committee. Plaque specimens (approximately 1 gram wet weight) were obtained from patients undergoing carotid endarterectomy. All samples were typical of advanced lesions4 and were homogenized in sodium carbonate buffer (pH 11) according to established methods.4
Human aortic specimens (from men or women 20 to 60 years old; 10 to 50 grams) were obtained postmortem (<48 hours) and washed immediately with ice-cold, argon-flushed phosphate-buffered saline (PBS) containing 300 μmol/L ethylenediaminetetraacetic acid, 100 μmol/L diethylenetriaminepentaacetic acid, 10 mmol/L 3-amino-1,2,4-triazole, 20 μmol/L butylated hydroxytoluene, and 2 mmol/L benzamidine hydrochloride. Specimens were stored at −80°C for no longer than 6 months before processing. Aortic homogenates were prepared in sodium carbonate buffer as described above for plaque samples using ≈1 gram of intima resected from an appropriate area of aorta depending on the lesion type. Specimens were classified macroscopically into four distinct groups as defined by Stary et al15: no/initial lesion (I–II), fatty streak lesions (II–III), fibrofatty lesions (IV–V), and complex/ulcerated lesions (V–VI). A small piece of each specimen was fixed and examined for histology using hematoxylin and eosin staining. There was 80% to 90% agreement between macroscopic and histological classification of the lesions.
In Vitro LDL Oxidation
Oxidation of LDL was performed in PBS under air at 37°C for the indicated period of time and using the oxidant specified.16 Human peripheral blood monocytes were isolated by elutriation, and monocyte-mediated LDL oxidation was performed in Hanks’ balanced salt solution containing 100 μmol/L diethylenetriaminepentaacetic acid, as described previously by others.17
HPLC Sample Workup and Analysis
Aliquots of LDL (100 μL) or tissue homogenate (100 to 400 μL) were extracted and analyzed for cholesterol (FC), cholesteryl esters (CE), TOH, and CE-O(O)H by reversed-phase HPLC with UV, electrochemical, and postcolumn chemiluminescence detection, respectively, as described.14
Gas Chromatography–Mass Spectrometry (GC-MS) Sample Workup and Analysis
Analysis of TOH and its oxidation products—TQ, TQE1, and TQE2, was achieved using a stable-isotope dilution GC-MS method,18 with an Agilent 5973N GC-MS instrument. Run parameters were as described previously18 except that injections were carried out in a split-splitless inlet operating at 250°C in the pulsed-splitless mode with a pulse pressure of 23 psi for 0.8 minutes and a splitless period of 1 minute
Lesion samples were spiked with deuterated TOH, TQ, and TQEs to assess the extent of artifactual TOH oxidation and the recovery and stability of oxidation products during tissue homogenization and sample extraction. These experiments showed that the extent of artifactual TOH oxidation was <5% and that >85% of all species were recovered and were stable in the homogenate (see the online data supplement for a detailed description).
Statistical Analyses of Data
Statistical analyses were performed by one-way ANOVA with a Newman-Keuls posttest. A probability value of <0.05 was considered significant.
An expanded Materials and Methods section can be found in the online data supplement available at http://www.circresaha.org.
In Vitro LDL Oxidation
We first established the profiles of TOH oxidation products in LDL exposed in vitro to different chemically defined oxidants (Figure 2). Because lipoproteins isolated from human lesions still contain TOH,3,4⇓ we used “mild” oxidation conditions and determined both the extent of lipid oxidation and the profile of TOH oxidation products while vitamin E remained present.
We first used a typical radical oxidant. Incubation of LDL under conditions where aqueous peroxyl radicals were generated at a constant rate led to the steady increase of all three TOH oxidation products (Figure 2A). Overall, the oxidation products measured accounted for 90% of the TOH consumed. TQE2 was the major product formed followed by TQE1 and TQ, accounting for 55%, 20%, and 15% of the TOH consumed, respectively. After 6 hours of oxidation, when all eight molecules of TOH present initially in LDL were consumed, 60 molecules of CE-O(O)H were formed per LDL particle (Figure 2A). CE-O(O)H are the single major lipid oxidation products detected in in vitro–oxidized LDL19 and in atherosclerotic lesions.3 Thus, lipid oxidation proceeded in a chain reaction despite the presence of TOH, an observation that can be explained readily by TMP.9
We used copper ions (Cu2+) as an additional radical oxidant. Cu2+ efficiently catalyzes LDL lipid oxidation and causes the generation of lipid alkoxyl and peroxyl radicals.20 Similar to the situation with aqueous peroxyl radicals, large quantities of CE-O(O)H accumulated under these conditions, and TQE2 was again the major TOH oxidation product (Figure 2B). However, all TOH oxidation products together could account for only ≈25% of the vitamin consumed, although the oxidation profile suggests that this relatively low yield was not the result of degradation of the known TOH oxidation products. These results show that although Cu2+ generates presently unidentified TOH oxidation products, the exposure of LDL to radical oxidants is generally characterized by the preferential accumulation of TQE2, and the total extent of lipid oxidation exceeds that of TOH oxidation ≈10- to 40-fold.
We next exposed LDL to two-electron oxidants. For this we first used sydnonimine-1, a chemical that generates a steady flux of ONOO−. As can be seen, this led to the stoichiometric conversion of TOH to TQ (Figure 2C). LDL exposed to reagent HOCl also generated TQ as the principal TOH oxidation product and with only traces of TQEs formed (Figure 2D). Eighty percent of TOH consumed was accounted for by TQ alone. For both two-electron oxidants, the extent of lipid oxidation exceeded that of TOH oxidation by only ≈2-fold (Figures 2C and 2D).
Together, these results show that the TOH oxidation profiles induced by one- and two-electron oxidants markedly differ in that the former predominantly yield TQEs whereas the latter almost exclusively yield TQ. In addition, the two types of oxidation pathways differ in that radical oxidants give rise to substantially higher ratios of lipid to TOH oxidation products compared with two-electron oxidants.
Cell-Mediated LDL Oxidation
Activated leukocytes generate superoxide anion that dismutates to H2O2 and is utilized by myeloperoxidase to produce HOCl. In the presence of nitrite (NO2−), several additional reactive nitrogen species are formed.21–23⇓⇓ The various species have been implicated in LDL oxidation in the artery wall, as relevant markers such as 3-chlorotyrosine, p-hydroxyphenylacetaldehyde, and 3-nitrotyrosine are elevated in atherosclerotic lesions.5,6,24⇓⇓ We therefore measured TOH and lipid oxidation in LDL exposed to activated human monocytes in the presence of nitrite (500 μmol/L) and chloride ions (142 mmol/L). Under these conditions, TQ and TQEs accounted for 75% and 9% of the TOH consumed, respectively, and the extent of lipid oxidation was around six molecules of CE-O(O)H per LDL formed after 6 hours of oxidation (Figure 3A). Thus, the ratio of lipid to TOH oxidation products was approximately two. Experiments performed with 50 μmol/L nitrite yielded the same distribution of oxidation products, although the overall extent of oxidation was less (data not shown). We next examined the amounts of TOH oxidation (Figure 3B) and lipid oxidation products (Figure 3C) formed after 4 hours under varying oxidizing conditions. The complete system was comprised of the same conditions as for Figure 3A. TOH oxidation was significantly less in the absence of NO2− and was negligible for nonactivated cells (minus phorbol myristate acetate [PMA]). Where oxidation was observed, TQ always remained the major TOH oxidation product, except in the presence of catalase that suppressed TQ formation completely. The absence of nitrite decreased the production of TQ but not that of TQEs (Figure 3B). Lipid peroxidation was only partially nitrite-dependent, whereas catalase and butylated hydroxytoluene significantly suppressed such oxidation (Figure 3C). Together, these results indicate that in the presence of nitrite, activated human monocytes cause LDL TOH and lipid oxidation similar to that seen with chemically defined two-electron oxidants.
Next we measured the amounts of FC, CE, CE-O(O)H, TOH, and its oxidation products in homogenates of human aortic tissue exhibiting varying degrees of disease. FC and the sum of the two major CE in human LDL, cholesteryl linoleate and cholesteryl arachidonate, were higher in advanced than in early lesions when expressed per milligram of total tissue protein (Table 1), indicating lipid loading in the diseased vessel wall. TOH increased commensurately (per mg protein), consistent with the fact that it is delivered to the vessel wall along with other lipids in LDL. Therefore, when normalized to CE or FC, the concentration of TOH remained unaltered independent of lesion severity, except for a 62% decrease in the TOH/CE value in fatty streaks (Table 1). Compared with native LDL isolated from plasma, the TOH-to-FC ratio in lesions was lower (26.3±8.3 versus 7 to 12 mmol TOH/mol FC for LDL4 and lesions, respectively). However, the TOH-to-CE ratio was comparable in lesions and plasma LDL (16.2±4.5 versus 11 to 30 mmol TOH/mol CE for LDL4 and lesions, respectively). Considering that CE are more readily oxidized than FC and are the major class of lipid in LDL, these data indicate that on average, lesion lipids are not depleted of TOH.
Absolute amounts of protein-standardized oxidized lipid and TOH increased with increasing disease severity (Figure 4). At all stages, TQ was the major TOH oxidation product, present at 15- to 20-fold higher concentration than TQE1 or TQE2. In early lesions (type I–II), there was ≈10 times more TQ than CE-O(O)H (Table 2). The content of oxidized lipids increased markedly in more advanced lesions, where it was ≈7-fold higher than the total amount of oxidized TOH (Table 2).
The proportion of CE oxidized to CE-O(O)H also increased significantly in advanced lesions, with up to ≈2% of total CE oxidized (Figure 4C). In contrast, the proportion of TOH oxidized to TQ was highest in early lesions where it accounted for ≈17%, but did not increase thereafter (Figure 4C). TQEs were not prevalent (<1%) across lesion groups, with only a slight increase in the advanced (type V–VI) lesions (Figure 4C).
Finally, we measured TOH and lipid oxidation products in homogenates of human carotid plaque specimens. Overall, the results obtained were similar to those of type V–VI aortic lesions, with comparable absolute concentrations of FC, CE, and TOH (Table 1). TQ was the major (11% of total TOH) and TQE1 and TQE2 the minor oxidation products (Figure 5), and the extent of lipid oxidation exceeded that of TOH oxidation by ≈7-fold (Table 2).
The present study shows that TOH remains essentially intact in the human vessel wall as atherosclerotic lesions develop. Only a fraction of the vitamin is oxidized, and TQ is the single major oxidation product. Maximal, albeit limited, TOH oxidation is observed at the earliest stage of atherogenesis, at a time when the extent of lipid oxidation is minimal and overall less than that of the vitamin. The preferential formation of TQ and the low ratio of lipid to TOH oxidation closely resemble the situation when LDL is oxidized in vitro by reagent two-electron oxidants or those produced by activated human monocytes. Therefore, our findings indicate that oxidative events taking place in the artery wall during early atherogenesis are initiated primarily by two-electron oxidants.
Our observation that TOH oxidation exceeds lipid oxidation during early atherogenesis seems to suggest that TOH effectively protects arterial lipids from oxidation. Indeed, TOH is a highly effective scavenger of one-electron (ie, radical) oxidants,25 although for LDL lipids this is not necessarily associated with effective protection against lipid oxidation.9,11⇓ Also, the vitamin fails to protect LDL from two-electron oxidants26 that give rise to TQ in preference over TQEs, as our in vitro studies with isolated LDL and chemically defined oxidants (Figure 2) clearly establish, and as is seen in human lesions. In contrast, radicals generally cause more substantial oxidation of LDL lipid than TOH and preferentially give rise to TQEs. The observed low extent of lipid oxidation and high TQ to TQE ratio in early lesions are therefore best explained with TOH engaging in two-electron oxidation reactions rather than acting as radical-scavenging antioxidant protecting LDL lipids from oxidation.
The fact that two-electron oxidants give rise primarily to TQ rather than TQEs is explained readily on the basis of the underlying chemistry. Thus, two-electron oxidants convert TOH to the intermediate tocopheroxylium cation (Figure 1)27 that hydrolyzes rapidly to TQ, rather than giving rise to the radical intermediate TO·. Although TO· is a precursor to both TQ and TQEs, formation of the latter also requires the presence of peroxyl radicals (Figure 1). The most likely source of peroxyl radicals is LDL lipids undergoing lipid peroxidation, which itself is a radical chain reaction25 that is effectively initiated by one- but not two-electron oxidants.28 Hence, it is the lack of coexisting TO· and peroxyl radicals that explains the low concentration of TQEs when LDL is exposed to two-electron oxidants.
There are two obvious candidates for the two-electron oxidants implicated in the TOH and lipid oxidation seen in early atherosclerosis, ONOO− and HOCl. Although our results do not allow us to distinguish between different types of two-electron oxidants, the concomitant formation of nitric oxide and superoxide, eg, by vascular cells,29 is a potential source of ONOO−, and there is chemical6 and immunohistochemical30 evidence for its involvement in LDL oxidation in the artery wall. Alternatively, or in addition, HOCl or related oxidant(s) produced by activated phagocytes may be involved. Indeed, there is growing evidence for this.31 We showed earlier that HOCl-modified proteins, including apolipoprotein B–containing lipoproteins accumulate in human lesions.32 Consistent with this, Heinecke and coworkers showed preferential accumulation of lipid and protein products derived from oxidation with HOCl or myeloperoxidase/H2O2/Cl− in human lesion LDL.5,24⇓ Our in vitro studies indicate that activated human monocytes oxidize LDL predominantly via processes mediated by two-electron oxidants (Figure 3). Formation of TQ was largely nitrite-dependent (Figure 3B), suggesting a significant involvement of reactive nitrogen species such as nitryl chloride and ONOO−.33,34⇓ This conclusion seems at odds with an earlier report by Hazen et al23 that suggested the primary involvement of the one-electron oxidant nitrogen dioxide, although a participation of nitryl chloride and ONOO− was not excluded. The present findings therefore indicate that further studies are warranted to more clearly establish the major oxidants produced by activated human monocytes in the presence of NO2−.
The ratio of oxidized lipid to oxidized TOH increased steadily from ≈0.1 to 7 as the severity of the lesions increased (Table 2). Thus, lipid oxidation becomes more prevalent as lesions develop. However, even at the most advanced stage, the ratio of oxidized lipid to TOH still more closely resembled that seen with two- rather than one-electron oxidants, suggesting that, overall, radical oxidants play a less important role in LDL oxidation in the vessel wall than has been generally assumed.
As indicated, the concentrations of CE-O(O)H detected in fibrofatty and advanced lesions were higher than the total amount of TOH oxidation products. This seems surprising considering that TOH is ≈104-fold more reactive toward peroxyl radicals than CE, and lesions contain only 50- to 100-fold more CE than TOH. Thus, on purely kinetic grounds, one would expect at least 102-fold more oxidation of TOH than lipid. The observed ratio of CE-O(O)H to TOH oxidation products of around seven in advanced lesions is however fully consistent with TMP.9,11⇓ This model predicts that TO· can react more frequently with LDL lipids (to generate CE-O(O)H) than it reacts with other radicals, such as lipid peroxyl radicals, to inhibit lipid peroxidation. If TO· were to predominantly react with lipid peroxyl radicals, we would expect TQEs to be formed in high yields (Figure 1), whereas we detected only trace levels of these TOH oxidation products at all stages of the disease.
The above explanations imply that TOH does not effectively inhibit lipid oxidation in the vessel wall, and that TMP operates in vivo. Indeed, there is additional support for the idea that TMP contributes to lipid oxidation during atherogenesis. Thus, the distribution of configurational isomers of CE-O(O)H suggests that in human lesions most of the oxidized lipids are formed in the presence of TOH.11 Also, in animal models of atherosclerosis, compounds that inhibit TMP (ie, co-antioxidants)35–37⇓⇓ significantly decrease the accumulation of oxidized lipids in the vessel wall. Therefore, our present observations suggest that it is the lack of availability of co-antioxidants rather than depletion of vitamin E that explains why lipids become increasingly oxidized as lesions develop. The fact that advanced human atherosclerotic lesions contain ascorbate,3 a powerful co-antioxidant,10 suggests that vitamin C may be localized within cells, whereas lipoprotein lipid oxidation may occur outside cells.
There are several limitations to our study. Given the ethical implications involved, it was not possible to obtain and analyze the postmortem material within a substantially shorter period of time. This leaves open the possibility for postmortem artifacts, the most likely of which would be artifactual oxidation of TOH and lipids. Two lines of evidence argue against this. First, the overall extent of oxidation of TOH, the most readily oxidizable substance, is low, and normal levels of nonoxidized vitamin E remain present. Second, if significant artifactual oxidation had occurred, we would have expected more oxidized TOH and lipid to be present in the most advanced aortic lesions than in the comparably diseased carotid material that we obtained freshly (see however Figure 5). Another limitation to our study is the fact that we examined snapshot views of the disease process only, so that we cannot rule out the possibility that some of the TOH and/or lipid oxidation products measured had undergone metabolism. To address this issue, we spiked lesion homogenates with labeled TQ or TQEs before incubation at 37°C for up to 18 hours. Only marginal degradation of the labeled compounds was observed (see experimental procedures and supporting data in the online supplement available at http://www.circresaha.org). Also, topical application of TQEs to mice followed by analysis of the epidermis showed that TQEs are stable in vivo,38 and to date, there are no metabolic processes known to degrade TQEs. This argues against a preferential metabolism of TQEs over TQ that, if occurring, would limit the validity of our conclusion that oxidative events in the artery wall are primarily initiated by two-electron oxidants. TQ can be reduced to TOH in vitro,39 although this requires harsh chemical conditions unlikely to exist in vivo. Finally, our results would be confounded if TOH oxidation products other than those measured were formed in vivo. Indeed, our in vitro LDL oxidation experiments with Cu2+ show that this is feasible. However, there is no good evidence to suggest that Cu2+ participates in LDL oxidation in the vessel wall, and if it were involved, we would have expected more TQE2 than TQ to accumulate (Figure 2B). Because this was not the case and the TOH oxidation products detected accounted for most of the TOH lost in the other in vitro LDL oxidation conditions, substantial formation of unknown oxidation products of vitamin E seems unlikely. This is particularly so, as there is no evidence that a substantial amount of TOH becomes oxidized in the vessel wall as lesions develop.
In conclusion, we show that in human atherosclerotic lesions, most of the endogenous vitamin E remains intact and that the small extent of oxidation of the vitamin takes place early in atherogenesis and is primarily caused by two-electron oxidants. Lipid oxidation becomes more prevalent at later developmental stages, although this process does not appear to be inhibited by vitamin E. This together with the coexistence of TOH and oxidized lipid suggests that vitamin E supplements are not likely to be beneficial in preventing LDL lipid oxidation in the vessel wall. This view is supported by the overall disappointing outcome of human intervention studies aimed at lowering coronary heart disease with vitamin E supplements. Our findings suggest that co-antioxidants and/or antioxidants that scavenge two-electron oxidants should instead be considered as alternative antiatherogenic supplements.
This work was supported by the National Heart Foundation, the National Health and Medical Research Council of Australia, and the US National Cancer Institute. The authors thank Henkel Fine Chemicals for the generous gift of deuterated tocopherol acetates. We also thank the staff of the Department of Vascular Surgery, Royal Prince Alfred Hospital, NSW, for carotid endarterectomy specimens, and the staff of the NSW Institute of Forensic Medicine for human aorta specimens. Members of the Biochemistry Group, The Heart Research Institute, are thanked for many helpful discussions.
Original received September 14, 2001; revision received December 17, 2001; accepted December 18, 2001.
- ↵Suarna C, Dean RT, May J, Stocker R. Human atherosclerotic plaque contains both oxidized lipids and relatively large amounts of α-tocopherol and ascorbate. Arterioscler Thromb Vasc Biol. 1995; 15: 1616–1624.
- ↵Niu X, Zammit V, Upston JM, Dean RT, Stocker R. Co-existence of oxidized lipids and α-tocopherol in all lipoprotein fractions isolated from advanced human atherosclerotic plaques. Arterioscler Thromb Vasc Biol. 1999; 19: 1708–1718.
- ↵Leeuwenburgh C, Hardy MM, Hazen SL, Wagner P, Oh-ishi S, Steinbrecher UP, Heinecke JW. Reactive nitrogen intermediates promote low density lipoprotein oxidation in human atherosclerotic intima. J Biol Chem. 1997; 272: 1433–1436.
- ↵Bowry VW, Mohr D, Cleary J, Stocker R. Prevention of tocopherol-mediated peroxidation of ubiquinol-10-free human low density lipoprotein. J Biol Chem. 1995; 270: 5756–5763.
- ↵Upston JM, Terentis AC, Stocker R. Tocopherol-mediated peroxidation (TMP) of lipoproteins: implications for vitamin E as a potential antiatherogenic supplement. FASEB J. 1999; 13: 977–994.
- ↵Stary HC, Chandler AB, Dinsmore RE, Fuster V, Glagov S, Insull W Jr, Rosenfeld ME, Schwartz CJ, Wagner WD, Wissler RW. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. Arterioscler Thromb Vasc Biol. 1995; 15: 1512–1531.
- ↵O’Leary VJ, Darley-Usmar VM, Russell LJ, Stone D. Pro-oxidant effects of lipoxygenase-derived peroxides on the copper-initiated oxidation of low-density lipoprotein. Biochem J. 1992; 282: 631–634.
- ↵van der Vliet A, Eiserich JP, Halliwell B, Cross CE. Formation of reactive nitrogen species during peroxidase-catalyzed oxidation of nitrite: a potential additional mechanism of nitric oxide-dependent toxicity. J Biol Chem. 1997; 272: 7617–7625.
- ↵Hazen SL, Zhang R, Shen Z, Wu W, Podrez EA, MacPherson JC, Schmitt D, Mitra SN, Mukhopadhyay C, Chen Y, Cohen PA, Hoff HF, Abu-Soud HM. Formation of nitric oxide-derived oxidants by myeloperoxidase in monocytes: pathways for monocyte-mediated protein nitration and lipid peroxidation in vivo. Circ Res. 1999; 85: 950–958.
- ↵Hazen SL, Gaut JP, Crowley JR, Hsu FF, Heinecke JW. Elevated levels of protein-bound p-hydroxyphenylacetaldehyde, an amino acid-derived aldehyde generated by myeloperoxidase, are present in human fatty streaks, intermediate lesions and advanced atherosclerotic lesions. Biochem J. 2000; 352: 693–699.
- ↵Hazell LJ, Davies MJ, Stocker R. Secondary radicals derived from chloramines of apolipoprotein B-100 contribute to HOCl-induced lipid peroxidation of low-density lipoproteins. Biochem J. 1999; 339: 489–495.
- ↵Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res. 2000; 86: 494–501.
- ↵Carr AC, Zhu BZ, Frei B. Potential antiatherogenic mechanisms of ascorbate (vitamin C) and α-tocopherol (vitamin E). Circ Res. 2000; 87: 349–354.
- ↵Eiserich JP, Cross CE, Jones AD, Halliwell B, van der Vliet A. Formation of nitrating and chlorinating species by reaction of nitrite with hypochlorous acid: a novel mechanism for nitric oxide-mediated protein modification. J Biol Chem. 1996; 271: 19199–19208.
- ↵Witting PK, Pettersson K, Östlund-Lindqvist A-M, Westerlund C, Westin Eriksson A, Stocker R. Inhibition by a co-antioxidant of aortic lipoprotein lipid peroxidation and atherosclerosis in apolipoprotein E and low density lipoprotein receptor gene double knockout mice. FASEB J. 1999; 13: 667–675.
- ↵Thomas SR, Leichtweis SB, Pettersson K, Croft KD, Mori TA, Brown AJ, Stocker R. Dietary co-supplementation with vitamin E and coenzyme Q10 inhibits atherosclerosis in apolipoprotein E gene knockout mice. Arterioscler Thromb Vasc Biol. 2001; 21: 585–593.
- ↵Liebler DC, Burr JA. Effects of UV light and tumor promoters on endogenous vitamin E status in mouse skin. Carcinogenesis. 2000; 21: 221–225.
- ↵Wijesundara MBJ, Berger S. The redox pair vitamin E and vitamin C, a 13C-NMR study. Liebigs Ann Chem. 1994; 1239–1241.