Oxidation and Antioxidation of Human Low-Density Lipoprotein and

accumulation of cholesteryl ester hydro(pero)xides, the loss of lysine (Lys) and tryptophan (Trp) residues, and the change in relative electrophoretic...
0 downloads 0 Views 166KB Size
484

Chem. Res. Toxicol. 1998, 11, 484-494

Oxidation and Antioxidation of Human Low-Density Lipoprotein and Plasma Exposed to 3-Morpholinosydnonimine and Reagent Peroxynitrite Shane R. Thomas,* Michael J. Davies,† and Roland Stocker The Biochemistry and EPR Groups, The Heart Research Institute, 145 Missenden Road, Camperdown, Sydney, NSW 2050, Australia Received September 23, 1997

As peroxynitrite is implicated as an oxidant for low-density lipoprotein (LDL) in atherogenesis, we investigated this process using reagent peroxynitrite (ONOO-) and 3-morpholinosydnonimine (SIN-1, which produces peroxynitrite via generation of NO• and O2•-). LDL oxidation was assessed by the consumption of ubiquinol-10 (CoQ10H2) and R-tocopherol (R-TOH), the accumulation of cholesteryl ester hydro(pero)xides, the loss of lysine (Lys) and tryptophan (Trp) residues, and the change in relative electrophoretic mobility. Exposure to ONOO- or SIN-1 resulted in rapid ( urate > ascorbate. With SIN-1, these antioxidants inhibited Trp consumption, while only the co-antioxidants ascorbate and 3-HAA prevented R-TOH consumption and lipid peroxidation. Exposure of human plasma to SIN-1 resulted in the loss of ascorbate followed by loss of CoQ10H2 and bilirubin. Lipid peroxidation was inhibited during this period, though proceeded as a radical-chain process after depletion of these antioxidants and in the presence of R-TOH and urate. Bicarbonate at physiological concentrations decreased ONOO--induced lipid and protein oxidation, whereas it enhanced SIN-1-induced lipid peroxidation, Trp consumption, and R-tocopheroxyl radical formation in LDL. These results indicate an important role for tocopherol-mediated peroxidation and co-antioxidation in peroxynitrite-induced lipoprotein lipid peroxidation, especially when peroxynitrite is formed time-dependently by SIN-1. The studies also highlight differences between ONOO-- and SIN1-induced LDL oxidation with regards to the effects of bicarbonate, ascorbate, and urate.

Introduction Oxidative modification of low-density lipoprotein (LDL)1 is implicated as an important early event in atherogenesis (1). As such, great interest has focused on LDL antioxidants as potential antiatherogenic agents. Indeed, some (2, 3), though not all (4), synthetic lipid-soluble antioxidants slow the progression of atherosclerosis in animal models. Human LDL contains several endogenous antioxidants, including R-tocopherol (R-TOH) and ubiquinol-10 (CoQ10H2) (5). R-TOH, biologically and chemically the most active form of vitamin E (6), is the most abundant * Corresponding author. Tel: +61(2) 9550 3560. Fax: +61(2) 9550 3302. E-mail: [email protected]. † The EPR Group. 1 Abbreviations: AAPH, 2,2′-azobis(2-amidinopropane) hydrochloride; CEO(O)H, cholesteryl ester hydro(pero)xides; CoQ10H2, ubiquinol10; EPR, electron paramagnetic resonance; 3-HAA, 3-hydroxyanthranilate; LDL, low-density lipoprotein; Lys, lysine; •OH, hydroxyl radical; ν, chain length; REM, relative electrophoretic mobility; SIN-1, 3-morpholinosydnonimine; TMP, tocopherol-mediated peroxidation; R-TOH, R-tocopherol; R-TO•, R-tocopheroxyl radical; Trp, tryptophan.

lipid-soluble antioxidant in LDL extracts (7) and as such has received the most interest with respect to research into “antioxidation” of LDL lipids. The role of R-TOH during radical-initiated LDL lipid peroxidation has been controversial, with both antioxidant and pro-oxidant activities being described. A recent study (8) has helped clarify this issue and provided further experimental support for the model of tocopherol-mediated peroxidation (TMP), which predicts that during radical-initiated LDL lipid peroxidation, the R-tocopheroxyl radical (R-TO•) formed acts as a radical-chain-transfer agent to promote the peroxidation of LDL’s lipids (9-11). Thus, in the absence of co-antioxidants (see below), the reactivity of the radical oxidant and the frequency with which radicals react with LDL determine whether R-TOH acts as an antioxidant or pro-oxidant for LDL’s lipids (8). For example, when LDL is exposed to Cu2+ at a ratio of 1016 Cu2+/LDL (7), R-TOH exhibits antioxidant activity; lipid peroxidation is inhibited during the phase of consumption of R-TOH (2 h), R-TOH exhibits pro-oxidant activity; lipid peroxidation proceeds via a free-radical-chain process in the presence of the vitamin, is accelerated by enriching and markedly suppressed by decreasing the R-TOH content of LDL, and is faster in the presence of R-TOH than immediately after its complete consumption (8, 9, 13, 14). Furthermore, the less reactive the oxidant the greater the range of oxidant concentration over which R-TOH exhibits pro-oxidant activity for a given LDL concentration (8). A wide variety of one-electron oxidants or conditions that result in one-electron reactions have been demonstrated to initiate oxidation of LDL’s lipids by TMP, including peroxyl radicals, Cu2+, Ham’s F-10 medium in the absence and presence of human monocyte-derived macrophages, hydroxyl radicals (•OH), 15-lipoxygenase, and horseradish peroxidase/H2O2 (8, 9, 13-15). In contrast, the two-electron oxidant HOCl is a poor TMP agent; it induces only minor LDL lipid peroxidation though substantial apoB-100 modification (16). TMP is prevented by co-antioxidants, defined as compounds capable of efficiently scavenging the R-TO• in oxidizing LDL and exporting the radical into the aqueous phase (17). Physiological co-antioxidants include ascorbate (9), CoQ10H2 (13), and the tryptophan (Trp) metabolite 3-hydroxyanthranilate (3-HAA) (18). In contrast, urate, reduced glutathione, and carotenoids are not coantioxidants. Peroxynitrite2 [oxoperoxonitrate(1-)] is a potent oxidant formed from the near-diffusion-limited reaction of nitrogen monoxide (NO•) with superoxide anion radical (O2•-) (19). Peroxynitrite is capable of reacting with carbohydrates (20), nucleic acids (21), lipids (22), proteins (23), and antioxidants (24). Peroxynitrite may be produced by endothelial and phagocytic cells, capable of simultaneously generating O2•- and NO• (25, 26). Interestingly, under limited arginine supply cellular nitric oxide synthase can produce both O2•- and NO• and hence give rise to peroxynitrite (27). There is evidence for peroxynitrite production in human renal allograft rejection (28), rheumatoid arthritis (29), and atherosclerosis (30, 31), although the use of detection of 3-nitrotyrosine as evidence for in vivo peroxynitrite production remains ambiguous as this amino acid modification can also be generated by HOCl plus nitrite (32). At neutral pH, the peroxynitrite anion (ONOO-) is in equilibrium with its conjugated acid, peroxynitrous acid (HOONO, pKa 6.8), which itself is highly unstable and readily isomerizes to nitrate (33). During decomposition, HOONO can either directly oxidize substrates by secondorder reactions or be converted to an activated trans intermediate (HOONO*) that oxidizes substrates via reactions that are first-order in peroxynitrite and zerothorder in substrate (indirect oxidation mechanism) (34). Peroxynitrite is capable of mediating both two- and oneelectron oxidations (35). The one-electron oxidations have been suggested to be due to the formation of •OH and nitrogen dioxide from the decomposition of HOONO (36, 37). However, both thermodynamic and kinetic 2 Peroxynitrite refers to the anion, its conjugate acid, and the activated trans intermediate of peroxynitrous acid. In this manuscript reagent peroxynitrite is abbreviated as ONOO-, HOONO is used where the conjugated acid is addressed specifically, and HOONO* refers to the activated trans intermediate.

Chem. Res. Toxicol., Vol. 11, No. 5, 1998 485

considerations (33) and experimental evidence (35, 38) suggest that homolysis is unlikely to occur and HOONO* is responsible for the one-electron oxidations. In vitro studies commonly use 3-morpholinosydnonimine (SIN-1) as a peroxynitrite generator. SIN-1 timedependently releases equimolar amounts of O2•- and NO• (39). A number of studies have established that reagent peroxynitrite (ONOO-) and SIN-1 oxidize LDL’s lipids and protein moieties (40-45). However the mechanism(s) of these oxidation reactions has not been investigated in detail. Therefore, we examined the potential role of TMP and co-antioxidation in LDL and plasma lipid peroxidation induced by peroxynitrite and investigated the action of bicarbonate and physiological aqueous antioxidants on such oxidations. Furthermore, we compared LDL oxidation induced by bolus versus continuously formed peroxynitrite.

Materials and Methods Materials. 2,2′-Azobis(2-amidinopropane) hydrochloride (AAPH) was obtained from Polysciences (Warrington, PA). Ascorbate, urate, and bilirubin were purchased from Sigma, R-TOH was from Henkel (Sydney, Australia), and 3-hydroxyanthranilate (3-HAA) was from Aldrich. Sodium bicarbonate was obtained from BDH. SIN-1 was a generous gift from Casella, Germany. Coenzyme Q capsules (50 mg of ubiquinone10/capsule) were from Blackmores Ltd. (Sydney, Australia). ONOO- was synthesized as described previously (46) and its concentration determined spectrophotometrically in alkaline solution using 302 ) 1.67 × 103 M-1 cm-1. Nanopure water (MODULAB) was used for all buffers, which were subsequently treated with Chelex-100 (Bio-Rad, Richmond, CA) to remove contaminating, redox-active transition metals. All organic solvents used (HPLC quality) were from either Mallinckrodt or Merck. Preparation of Plasma and LDL. Blood was obtained from normolipidemic male or female donors (n ) 10, age 2640 years) and drawn into heparinized vacutainers (Becton Dickinson). Plasma and LDL were prepared as described previously (47). Plasma was used immediately, whereas LDL was stored at 4 °C for e5 days before use. No specific care was taken to preserve the small quantities of CoQ10H2 in LDL, unless CoQ10H2 data were required specifically (see below). Prior to oxidation, LDL was passed through a PD-10 gel-filtration column (Pharmacia), equilibrated with 50 or 200 mM phosphate buffer (pH 7.4). In Vitro Enrichment of LDL with r-TOH. Freshly prepared human plasma was treated with R-TOH (dissolved in DMSO) at a final concentration of 0.5 mM or an equal volume of DMSO [final concentration in each case < 0.3% (v/v)] and incubated at 37 °C for 4-6 h (48). LDL was then isolated from the plasma as described above. FIVE Patient (in Vivo) r-TOH Supplementation Regime. Plasma was obtained from a patient (male, 29 years) with familial isolated vitamin E deficiency syndrome (FIVE), from the Department of Pediatrics, University of Hamburg, Germany. This deficiency is characterized by a very low level of plasma R-TOH unless the patient receives daily supplements of R-TOH. Initially the patient was without R-TOH supplements for 5 consecutive days, after which time the first blood sample was taken, and this provided the R-TOH-deficient sample. Following this, the patient received supplementation with vitamin E for the ensuing 5 days. Blood samples were taken on each day thus providing R-TOH-supplemented plasma samples with increasing amounts of R-TOH with increasing time of supplementation. Plasma from all samples was prepared immediately, supplemented with sucrose (0.6%, w/v), frozen, and shipped on dry ice to Sydney. The plasma was kept at -80 °C until used. LDL was prepared from these various plasma samples as described above.

486 Chem. Res. Toxicol., Vol. 11, No. 5, 1998 In Vivo Enrichment of LDL with CoQ10H2. Healthy normolipidemic human subjects (n ) 3) were used for in vivo coenzyme Q supplementation. Initially a blood sample was removed, and plasma was isolated as described above and then stored under argon for 5-6 days at 4 °C. This represented the control, unsupplemented sample. CoQ10H2 was stable in the plasma under these storage conditions (not shown). Following collection of a control blood sample, the subjects were supplemented with coenzyme Q (150 mg/day) for the ensuing 5 days. On day 6 a second blood sample was taken and the plasma isolated. This constituted the CoQ10H2-supplemented sample. LDL was isolated from both control and supplemented plasma samples as described above using argon-flushed chelexed buffer and then used immediately for oxidation experiments. Oxidation of LDL and Plasma. Oxidation of isolated LDL (0.5 mg of protein/mL or 1 µM LDL) with ONOO- or SIN-1 was carried out under air at 25 °C in 200 or 50 mM phosphate buffer (pH 7.4), respectively. ONOO- was added as a single bolus, and samples were removed after 15 min. Preliminary experiments with ONOO- showed that oxidation reactions measured were complete within the first minute, and there were no differences in the extent of oxidation up to 15 min. All experiments involving bicarbonate were performed in 200 mM phosphate buffer (pH 7.4). For oxidation of plasma, SIN-1 was added at a final concentration of 5 mM and the plasma incubated under air for up to 24 h at 25 °C. At 25 °C, SIN-1 decomposes linearly over ≈7 h (data not shown), as assessed by the rate of formation of SIN-1C (39). Determination of Lipophilic and Aqueous Antioxidants and Neutral and Oxidized Lipids. Aliquots (50-100 µL) of the plasma or LDL reaction mixtures were withdrawn at the indicated times, extracted with 1 mL of acidified methanol (0.1%, v/v, acetic acid) and 5 mL of hexane, mixed, and centrifuged at 4 °C at 1000g for 5 min (47). For LDL experiments, extracts were stored at -20 °C for no longer than 48 h prior to analyses. The resulting hexane layer (4 mL) was removed, dried under vacuum, and resuspended immediately in 2-propanol (200 µL) for analyses of the lipid-soluble components. The levels of R-TOH and CoQ10H2, neutral lipids (mainly free cholesterol and cholesteryl esters), and cholesteryl ester hydroperoxides were determined by reversed-phase HPLC with electrochemical, UV (210 nm), and postcolumn chemiluminescence detection, respectively (47). In addition, the levels of oxidized cholesteryl esters were also quantified by UV (234 nm) which measures both cholesteryl ester hydroperoxides and hydroxides [CE-O(O)H]. Lipid-soluble components were standardized internally against free cholesterol (47). For LDL experiments, the concentrations of R-TOH and CE-O(O)H are expressed as the number of molecules per LDL particle. For ascorbate and urate analyses, the aqueous methanol phase was removed, filtered (0.2 µm), and analyzed immediately by HPLC with electrochemical detection (49). For bilirubin analysis, the filtered aqueous methanol phase was analyzed by HPLC at 440 nm as described previously (50). All compounds were quantitated by area comparisons with reagent standards. Determination of Relative Electrophoretic Mobility and Trp and Lys Residues. LDL’s relative electrophoretic mobility (REM) was determined as described (16). Trp and lysine (Lys) residues of LDL’s apoB-100 were measured by fluorescence (16). Lys determination was not possible with SIN1-containing samples due to interference of the sydnonimine with the assay (not shown). Electron Paramagnetic (EPR) Studies. EPR spectra of reaction mixtures contained in standard aqueous solution cells were recorded at room temperature using a Bruker EMX spectrometer equipped with 100-kHz modulation. Hyperfine coupling constants were measured directly from the experimental scans. Typical spectrometer conditions were gain 3 × 106, modulation amplitude 0.4 mT, time constant 81 ms, scan time 42 s, center field 347.8 mT, field scan 10.0 mT, microwave power 31 mW, and frequency 9.76 GHz, with (typically) 20 spectral scans averaged. Signal intensities (which are proportional to

Thomas et al. absolute radical concentrations) were compared by measurement of peak-to-peak line heights from spectra recorded under identical conditions. In some cases sequential spectra were recorded from reaction mixtures starting from 90 s after the initiation of the reaction for up to 45 min.

Results Treatment of LDL with ONOO- or SIN-1 resulted in oxidative alterations to both the lipoprotein’s lipid and protein components. Oxidation by ONOO- was instantaneous ( R-TOH g lycopene > β-carotene, with the majority (75-90%) of the R-TOH lost converted to R-tocopherylquinone (data not shown). Both oxidants induced accumulation of CE-O(O)H and phosphatidylcholine hydroperoxides at a molar ratio of ≈3-5:1; they also induced consumption of apoB-100’s Trp and Lys residues and an increase in LDL’s REM. REM values of LDL treated with 1 mM ONOO- (after 15-min incubation) and SIN-1 (after 7-h incubation) were ≈1.5 and 1.6, respectively. With ONOO-, it appeared that a greater proportion of the oxidant reacted with protein rather than lipid. Thus with 200 µM ONOO-, ≈70 µM oxidant could be accounted for by the loss of Trp plus Lys residues, whereas only ≈18 µM lipophilic substrates (antioxidants and lipid hydroperoxides) were accounted for. In comparison, peroxynitrite formed continuously by SIN-1 induced ≈10 times more lipid peroxidation than equimolar ONOO-, although a comparable extent of protein oxidation (as assessed by loss of Trp residues and increase in REM) was noted with both oxidants. Role of r-TOH during ONOO--Induced LDL Oxidation. To examine the role of R-TOH in peroxynitriteinduced LDL oxidation, we initially compared the oxidizability of in vitro R-TOH-enriched LDL versus native, nonenriched LDL. When LDL (1 µM) with varying R-TOH content was exposed to 100 µM, the R-TOH-enriched LDL accumulated less CE-O(O)H than the corresponding native LDL, and this antioxidant activity of the vitamin coincided with R-TOH depletion in the corresponding native LDL samples. While the yield of CE-O(O)H was low over the entire concentration range of ONOO- examined, it was highest at low ONOOconcentrations where R-TOH was not depleted. Similar results regarding lipid peroxidation were obtained for FIVE patient LDL exposed to ONOO- (not shown), indicating that the results obtained with in vitro manipulated LDL were not artifactual. In contrast to lipid peroxidation, there was no difference in the extent of ONOO-induced consumption of Trp or Lys residues between native and in vitro R-TOH-enriched LDL (Figure 1C). To verify the pro-oxidant activity of R-TOH, more concentrated LDL (5 µM) was exposed to increasing concentrations of ONOO- (Figure 2). This increased the ONOO- concentration range for which R-TOH was pro-oxidant: more CE-O(O)H accumulated in the R-TOHenriched than native LDL up to 400 µM ONOO-. Despite this, the yield of CE-O(O)H formed decreased when

Peroxynitrite and Lipoprotein Oxidation

Chem. Res. Toxicol., Vol. 11, No. 5, 1998 487

Figure 2. Lipid oxidation of concentrated solutions of native and in vitro R-TOH-enriched LDL induced by ONOO-. In vitro R-TOH-enriched (9) or native (0) LDL (2.5 mg of protein/mL, 5 µM LDL) was exposed to increasing concentrations of ONOOadded as a bolus. After 15 min, LDL aliquots were removed and analyzed for R-TOH (A) and CE-O(O)H (B). Results are the mean of two independent experiments.

Figure 1. Oxidation of native and in vitro R-TOH-enriched LDL induced by ONOO-. In vitro R-TOH-enriched (circles) or native (squares) LDL (0.5 mg of protein/mL, 1 µM LDL) was exposed to increasing concentrations of ONOO- added as a bolus. After 15 min, LDL aliquots were removed and analyzed for CE-O(O)H (A), R-TOH (B), Trp (closed symbols, C) and Lys (open symbols, C). Results are the mean ( SEM of four independent experiments.

compared to the experiments using less concentrated LDL (cf. Figure 1A and Figure 2B). Together, the results indicate that depending on the oxidant-to-LDL ratio employed, R-TOH can exhibit both pro-oxidant and antioxidant activity during ONOO--induced lipid peroxidation. Role of r-TOH during SIN-1-Induced LDL Oxidation. Previous studies have demonstrated that generation of both O2•- and NO• from SIN-1 is required for this compound to efficiently oxidize LDL, implying peroxynitrite as the damaging oxidant (41-43). We now show that with urate > ascorbate (Figure 6). With SIN-1 as the oxidant, all antioxidants inhibited Trp consumption (Figure 7A). However, whereas 3-HAA and ascorbate prevented both R-TOH consumption and lipid peroxidation, urate increased LDL lipid peroxidation, although it attenuated R-TOH consumption (i.e., urate increased the ν of lipid peroxidation) (Figure 7B). This “urate paradox” provides further support that SIN-1induced LDL lipid peroxidation proceeds via TMP (for a detailed discussion of the “urate paradox”, see ref 9). Effect of Bicarbonate on Peroxynitrite- and SIN1-Induced LDL Oxidation. Human plasma contains ≈25 mM bicarbonate in equilibria with CO2 (≈1.5 mM), with which peroxynitrite reacts rapidly (51). Thus, peroxynitrite-mediated oxidations in vivo may involve the peroxynitrite-CO2 adduct (i.e., nitrosoperoxycarbonate anion, ONOOCO2-) which can give rise to further reactive intermediates. Therefore, we examined the effect of bicarbonate on peroxynitrite-induced LDL oxidation. Addition of 25 mM bicarbonate inhibited LDL oxidation induced by ONOO-, as indicated by decreased CE-O(O)H

Thomas et al.

Figure 4. Oxidation of in vitro R-TOH-enriched and native LDL-induced by SIN-1. In vitro R-TOH-enriched (circles) or native (squares) LDL (0.5 mg of protein/mL) was exposed to 40 µM (A; closed symbols in C) or 400 µM (B; open symbols in C) SIN-1. At the indicated times, LDL aliquots were removed and analyzed for CE-O(O)H (closed symbols in A and B), R-TOH (open symbols in A and B), and Trp (C). Results are the mean ( SEM of four independent experiments.

accumulation, decreased consumption of R-TOH, Lys, and Trp, and lower REM (Figure 8). Bicarbonate also inhibited CE-O(O)H accumulation and consumption of R-TOH, Lys, and Trp in R-TOH-enriched LDL (not shown). In sharp contrast to ONOO-, the presence of 25 mM bicarbonate enhanced the rate of SIN-1-induced Trp and R-TOH consumption and formation of CE-O(O)H, independent of the SIN-1 concentration (Figure 9). Although bicarbonate caused a small increase in the pH of the 200 mM buffer (