Covalent Binding of Oxidized Cholesteryl Esters to Protein IMPLICATIONS FOR OXIDATIVE MODIFICATION OF LOW DENSITY LIPOPROTEIN AND ATHEROSCLEROSIS*

It has been proposed that plasma low density lipopro-teins (LDL) undergo oxidative modification before they can produce foam cells in atherosclerosis. The oxidation of LDL generates a variety of reactive aldehydic products, which covalently bind to the LDL apolipoprotein B-100 (apoB). In the present study, to investigate the mechanisms contrib-uting to the modification of LDL, we analyzed oxidized cholesteryl esters generated during the autoxidation of LDL and characterized their covalent binding to the lysine residues of LDL apoB. In addition, we raised a monoclonal antibody specific to a lysine-bound oxidized cholesteryl ester and determined its production in human atherosclerotic lesions. The peroxidation of LDL with Cu 2 (cid:1) produced 9-oxononanoylcholesterol (9-ONC) and 5-oxovaleroylcho-lesterol as the major oxidized cholesteryl esters. We observed that the levels of 9-ONC and 5-oxovaleroylcholes-terol peaked at 12 h and significantly decreased thereafter. The reduction of the core aldehyde molecule due probably modification agarose gel electrophoresis/immunoblot analysis of -oxidized formation immunoreactive not osclerosis. The participation of reactive aldehydes in atherosclerosis is also suggested by a series of immunohistochemical analyses of atherosclerotic lesions from the human aorta using various antibodies against short-chain aldehyde-amino acid adducts, as 4-hydroxy-2-nonenal-histidine (20), MDA-ly-sine (21, 22), and acrolein-lysine (23), in which the immunoreactivities are commonly associated with cells, primarily macrophages. These studies provided evidence for the presence of oxidized LDL (or at least of antigens closely related to it) in arterial lesions at significant concentrations and suggested that reactive aldehydes may play a role in the formation of arterial foam cells.

It has been proposed that plasma low density lipoproteins (LDL) undergo oxidative modification before they can produce foam cells in atherosclerosis. The oxidation of LDL generates a variety of reactive aldehydic products, which covalently bind to the LDL apolipoprotein B-100 (apoB). In the present study, to investigate the mechanisms contributing to the modification of LDL, we analyzed oxidized cholesteryl esters generated during the autoxidation of LDL and characterized their covalent binding to the lysine residues of LDL apoB. In addition, we raised a monoclonal antibody specific to a lysine-bound oxidized cholesteryl ester and determined its production in human atherosclerotic lesions. The peroxidation of LDL with Cu 2؉ produced 9-oxononanoylcholesterol (9-ONC) and 5-oxovaleroylcholesterol as the major oxidized cholesteryl esters. We observed that the levels of 9-ONC and 5-oxovaleroylcholesterol peaked at 12 h and significantly decreased thereafter. The reduction of the core aldehyde levels was accompanied by (i) the formation of free 7-ketocholesterol and 7-ketocholesteryl ester core aldehydes and (ii) an increase in the amounts of apoB-bound cholesterol and 7-ketocholesterol, suggesting that the cholesteryl ester core aldehydes were further converted to their 7-ketocholesterol-and apoBbound derivatives. To detect the protein-bound 9-ONC, we raised the monoclonal antibody 2A81, directed against 9-ONC-modified protein, and found that it extensively recognized protein-bound cholesteryl ester core aldehydes. Agarose gel electrophoresis followed by immunoblot analysis of the oxidized LDL clearly demonstrated the formation of antigenic structures. Furthermore, immunohistochemical analysis of the atherosclerotic lesions from the human aorta showed that immunoreactive materials with mAb 2A81 were indeed present in the lesions, in which the intense immunoreactivity was mainly located in the macrophage-derived foam cells and the thickening neointima of the arterial walls. The results of this study suggest that the binding of cholesteryl ester core aldehydes to LDL might represent the process common to the oxidative modification of lipoproteins.
Atherosclerosis is a complex vascular disorder that can lead to more serious conditions such as myocardial infarction. The events in atherosclerosis are monocyte migration from the blood stream, its differentiation into macrophages in situ, the uptake of low density lipoprotein (LDL) 1 by the macrophage scavenger receptor, transformation of the lipid-laden macrophage into foam cells, smooth muscle cell proliferation and transformation into foam cells, and thus the accumulation of foam cells leading to fatty streaks and subsequent plaque formation. Various lines of evidence indicate that an important part of the pathogenesis of atherosclerosis is the oxidative modification of plasma LDL (1,2). It has been proposed that LDL undergoes oxidative modification before it can give rise to foam cells, the key component of the progression of atherosclerosis, and that the oxidation of LDL in vivo can be reproduced by the in vitro incubation of LDL with cultured cells, such as endothelial cells, smooth muscle cells, and macrophages, or by autoxidation catalyzed by cupric ion in the absence of cells (1). During the oxidation of LDL, the LDL molecule undergoes a large number of structural changes that alter its metabolism (1). Although the detailed mechanism for the modification of LDL has not yet been established, it is generally accepted that the primary generation of lipid hydroperoxides initiates a reaction cascade leading to rapid propagation and to amplification of the number of reactive oxygen species formed; this ultimately leads to extensive fragmentation of the fatty acid chains (3) and conversion of the LDL to a more atherogenic form (4).
Lipid peroxidation leads to the formation of a broad array of different products with diverse and powerful biological activities. Among them are a variety of different aldehydes (5). The primary products of lipid peroxidation, lipid hydroperoxides (6), can undergo carbon-carbon bond cleavage via alkoxyl radicals in the presence of transition metals, giving rise to the formation of short-chain, unesterified aldehydes (6, 7) or a second class of aldehydes still esterified to the parent lipid. These esterified aldehydes are commonly termed core aldehydes (8). Although these compounds have received less attention, the formation of phospholipid and cholesteryl ester core aldehydes during the oxidation of LDL has been demonstrated (9,10). Due to their structural similarity to the platelet-activating factor and their rather hydrophilic nature, phospholipid core aldehydes display a number of platelet-activating factor-related biological activities. More recently, Podrez et al. (11) characterized a structurally conserved family of oxidized phosphatidylcholines that serve as novel high affinity ligands for cells stably transfected with CD36, mediating the recognition of multiple oxidized forms of LDL, and demonstrated their formation during the oxidation of LDL by multiple distinct pathways and their participation in the CD36-mediated recognition of different forms of oxidized LDL. On the other hand, with regard to oxidized cholesteryl esters, Kamido et al. (10) isolated and identified 9-oxononanoylcholesterol (9-ONC) and 5-oxovaleroylcholesterol (5-OVC) as the oxidation products of cholesteryl linoleate and cholesteryl arachidonate, respectively, in oxidized LDL. The oxidation of high density lipoprotein resulted in the formation of primarily 5-OVC and 9-ONC and, although to a lesser extent, the corresponding 7-ketocholesterol derivatives (12). In line with the potential pathophysiological role of these compounds, the cholesteryl ester and 7-ketocholesteryl ester core aldehydes of varying chain lengths were identified in human atheromas (13,14). In contrast to the phospholipid core aldehydes, however, much less is known about the formation, metabolism, and biological properties of cholesteryl ester core aldehydes.
In the present study, to investigate the mechanisms contributing to the modification of LDL, we analyzed the oxidized cholesteryl esters generated during the autoxidation of LDL and characterized their covalent binding to lysine residues of LDL apoB. In addition, based on the fact that the cholesteryl ester-and 7-ketocholesteryl ester core aldehydes of varying chain length were identified in human atheromas (13,14), we raised a monoclonal antibody directed to a lysine-bound oxidized cholesteryl ester and determined its production in human atherosclerotic lesions.
Autoxidation of Cholesteryl Esters-The autoxidation of the cholesteryl esters was performed by incubating the emulsified cholesteryl ester liposome (20 mM) with Fe 2ϩ (10 M) and ascorbate (1 mM) in 50 mM sodium phosphate buffer (pH 7.4).
In Vitro Oxidation of LDL-LDL (1.019 -1.063 g/ml) was prepared from the plasma of healthy humans by sequential ultracentrifugation and then extensively dialyzed against phosphate-buffered saline (10 mM sodium phosphate buffer, pH 7.4, containing 150 mM NaCl) containing 0.01% EDTA at 4°C. The LDL used for the oxidative modification by Cu 2ϩ was dialyzed against a 1000-fold volume of phosphatebuffered saline at 4°C. The oxidation of LDL was performed by incubating 0.5 mg of LDL with CuSO 4 (5 M) in 1 ml of 50 mM sodium phosphate buffer (pH 7.4) at 37°C. Informed consent was obtained under the study approved by our institutional review board. The study was performed in accordance with the Helsinki Declaration of 1975 as revised in 1983.
Lipid Peroxidation Assay-The degree of lipid peroxidation of the cholesteryl ester and LDL was determined by measurement of 2-thiobarbituric acid-reactive substances (TBARS). The reaction mixture (0.1 ml) was treated with 0.5 ml of 2.8% (w/v) trichloroacetic acid and 0.5 ml of 1% 2-thiobarbituric acid in 0.05 N NaOH and then boiled for 20 min. After cooling, the sample was centrifuged (11,000 ϫ g, 3 min), and the absorbance of the supernatant solution was measured at 532 nm. Malondialdehyde bis(dimethyl acetal), which yields malondialdehyde (MDA) by acid treatment, was used as a standard.
Core Aldehyde Analysis-The cholesteryl ester core aldehydes in the reaction mixtures (100 l) were derivatized with an equal volume of 0.1% DNPH in 2 N HCl at room temperature for 30 min. The formed DNPH derivatives were extracted with chloroform, evaporated, and dissolved in ethanol. 10 l of the aliquot was injected and analyzed on a Develosil ODS-MG-5 column (4.6 ϫ 250 mm) with a linear gradient elution from isopropyl alcohol/methanol (2:8) to isopropyl alcohol/methanol (8:2) for 45 min at a flow rate of 0.8 ml/min. The elution profiles were monitored by absorbance at 370 nm. For the LC-MS measurement with a Jasco PlatformII-LC instrument, 0.5% ammonium hydroxide was added to the mobile phases.
Cholesteryl Ester Analysis-Lipids in LDL samples (100 l) were extracted with an equal volume of chloroform/methanol (1:1), and the 25-l chloroform layer was directly injected on a Develosil ODS-MG-5 column (4.6 ϫ 250 mm) with a linear gradient elution from isopropyl alcohol/methanol (2:8) to isopropyl alcohol/methanol (8:2) for 45 min at a flow rate of 0.8 ml/min and monitored at 210 nm. Quantification was performed by peak area comparison with external standards.
Analysis of ApoB-bound Cholesteryl Esters-The isolation of apoB was performed as previously described (15). Briefly, 1 l of 1 N HCl and 100 l of chloroform/methanol (1:1) were added to 100 l of LDL samples, mixed, and centrifuged at 7,000 rpm for 10 min. The insoluble apoB was collected and washed twice each with ether, H 2 O, and acetone. The isolated apoB was solubilized in 10 mM NaOH containing 2% octyl glucoside. The cholesteryl esters associated with the apoB were saponified by incubation of the protein with 1 N KOH in methanol for 1 h at 37°C. After incubation, 6 N HCl was added until an acidic pH was reached, and chloroform was added to extract the lipids. The chloroform phase was evaporated to dryness and dissolved in 50 l of ethanol. 10 l of the aliquot was injected on a Develosil ODS-MG-5 column (4.6 ϫ 250 mm) in a solvent system of isopropyl alcohol/methanol (2:8) at a flow rate of 0.8 ml/min with UV detection at 210 nm. Quantification was performed by peak area comparison with external standards.
Reaction of N ␣ -Benzoyl-glycyl-lysine with 9-ONC-9-ONC (0 -10 mM in methanol) was mixed with an equal volume of 2 mM N ␣ -benzoylglycyl-lysine in phosphate buffer (pH 7.4), sonicated for 1 min, and incubated for 4 h at room temperature. After the incubations, the reaction mixtures were treated with 50 mM NaCNBH 3 for 1 h at room temperature. The 9-ONC-N ␣ -benzoyl-glycyl-lysine adducts were extracted with an equal volume of chloroform, evaporated, dissolved in methanol, and analyzed by LC-MS equipped with a Develosil ODS-HG-5 column (4.6 ϫ 250 mm) using a linear gradient from 90% aqueous acetonitrile to 100% acetonitrile for 30 min with UV detection (240 nm). The elution profiles were monitored by absorbance at 240 nm.
Modification of BSA by 9-ONC-250 l of 9-ONC dissolved in methanol was added to the BSA solution in 750 l of 50 mM sodium phosphate buffer (pH 7.2) and incubated at 37°C for 24 h. The final concentrations of BSA and 9-ONC were 0.5 mg/ml and 0 -5 mM, respectively.
Amino Acid Analysis-An aliquot (0.1 ml) of the protein samples incubated in the absence or presence of 9-ONC was treated with 10 mM EDTA (10 l), 1 N NaOH (10 l), and 100 mM NaBH 4 (10 l). After incubation for 1 h at 37°C, the mixture was treated with 10% trichloroacetic acid. After centrifugation at 10,000 ϫ g for 3 min, the proteins were hydrolyzed in vacuo with 6 N HCl for 24 h at 105°C. The hydrolysates were then concentrated and dissolved in 50 mM sodium phosphate buffer (pH 7.4). The amino acid analysis was performed using a JEOL JLC-500 amino acid analyzer equipped with a JEOL LC30-DK20 data analyzing system.
Antibody Preparation-The antigen was prepared by incubating KLH (1.0 mg/ml) with 5 mM 9-ONC in 10 ml of 0.1 M sodium phosphate buffer (pH 7.4) at 37°C for 24 h. The 9-ONC-treated KLH was further treated with NaCNBH 3 to reductively aminate the 9-ONC-lysine Schiff base adducts. Female BALB/c mice were immunized three times with the 9-ONC-modified KLH. Spleen cells from the immunized mice were fused with P3/U1 murine myeloma cells and cultured in hypoxanthine/ aminopterin/thymidine selection medium. The culture supernatants of the hybridoma were screened using an enzyme-linked immunosorbent assay (ELISA), employing pairs of wells of microtiter plates on which were absorbed 9-ONC-treated BSA as the antigen (0.5 g of protein/well).
ELISA-A 100-l aliquot of the antigen solution was added to each well of a 96-well microtiter plate and incubated for 20 h at 4°C. The antigen solution was then removed, and the plate was washed with Tris-buffered saline (TBS) containing 10% Tween 20 (TBS/Tween). Each well was incubated with 200 l of 1% BSA in TBS/Tween for 30 min at 37°C to block the unsaturated plastic surface. The plate was then washed once with TBS/Tween. A 100-l aliquot of the antibody was added to each well and incubated for 1 h at 37°C. After discarding the supernatants and washing three times with TBS/Tween, a 100 l of a 5 ϫ 10 3 dilution of goat anti-mouse IgM conjugated to horseradish peroxidase in TBS/Tween was added. After incubation for 1 h at 37°C, the supernatant was discarded, and the plates were washed three times with TBS/Tween. The enzyme-linked antibody bound to the well was revealed by adding 100 l/well of 1,2-phenylenediamine in 0.1 M citrate/ phosphate buffer (pH 5.0) containing 0.003% hydrogen peroxide. The reaction was terminated by the addition of 2 M sulfuric acid, and the absorbance at 492 nm was read on a micro-ELISA plate reader.
In a competitive ELISA, a competitor was incubated with antibody for 20 h at 4°C to yield competitor/antibody mixtures containing antibody at 0.8 g/ml and variable concentrations of the competitor. A 100-l aliquot of competitor/antibody mixtures was added to each well and incubated for 1 h at 37°C. After discarding the supernatants and washing three times with TBS/Tween, the second antibody was added, and the enzyme-linked antibody bound to the well was revealed as described above. Results were expressed as the ratio B:B o , where B represents absorbance in the presence of competitor minus background absorbance (no antibody, and B o is absorbance in the absence of competitor minus background absorbance. Agarose Gel Electrophoresis/Immunoblot Analysis-Agarose gel electrophoresis of LDL was performed with the Helena TITAN GEL High Resolution Protein System (Helena Laboratories, Saitama, Japan). The samples were run on two separate gels. One gel was used for staining with Fat Red 7B; the other was transblotted to nitrocellulose membranes, incubated with Block Ace (40 mg/ml) for blocking, washed, and treated with the primary antibody (mAb 2A81). This procedure was followed by the addition of horseradish peroxidase conjugated to a goat anti-mouse IgM and ECL reagents (Amersham Biosciences). The bands were visualized by exposure of the membranes to autoradiography film.
Immunohistochemistry-This investigation was carried out on aortic wall samples obtained at autopsy from five patients with generalized arteriosclerosis without diabetes mellitus. Each autopsy was performed at Tokyo Women's Medical University after the patients' family members granted informed consent according to the established guideline. The postmortem time before starting autopsy varied from 1 to 3 h. No significant difference of morphological changes in the aorta was seen among these cases with different postmortem time. Each sample was prepared for 10% formalin-fixed, paraffin-embedded materials and for frozen materials embedded in the optimum cutting temperature compound (Sakura Finetechnical, Tokyo, Japan) at Ϫ80°C. Multiple 6-mthick sections were cut from these paraffin-embedded and frozen materials and used for histopathological and immunohistochemical examinations. Paraffin-embedded sections were deparaffinized in xylene and ethanol, rehydrated in distilled water, quenched for 10 min with 3% hydrogen peroxide for inhibiting endogenous peroxidase activity, rinsed in phosphate-buffered saline, pH 7.6, and pretreated for 30 min at room temperature with 3% nonimmune animal serum, from the same species as those producing secondary antibodies, in phosphatebuffered saline for blocking nonspecific antibody binding. Prior to immunostaining for CD68, paraffin sections were treated for 30 min at 37°C with 0.1% trypsin in phosphate-buffered saline for antigen retrieval. Frozen sections were dried, postfixed or not in 10% formalin, rehydrated, processed with the same subsequent procedures as those for paraffin sections, and pretreated with the endogenous avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA). For staining of frozen sections, 1 mM EDTA was added to each solvent for removal of transition metals. These prepared sections were then incubated overnight at 4°C with mAb 2A81 at a concentration of 0.5 g/ml and the mouse monoclonal anti-CD68 IgG antibody (KP-1; Dako, Glostrup, Den-mark) at a dilution of 1:500. CD68 was used as a specific marker of macrophages. Immunoreaction was visualized by the avidin-biotin-immunoperoxidase complex method using Vectastain ABC kit (Vector). The chromogen was 3,3Ј-diaminobenzidine tetrahydrochloride, and the counterstain was hematoxylin. Sections from which the primary antibodies were omitted served as negative reaction controls. An immunoabsorption test was performed on sections incubated with mAb 2A81 pretreated with 50 g/ml 9-ONC-conjugated BSA. Immunohistochemical localization of the immunoreaction product deposits was verified by light microscopy on consecutive sections with hematoxylin-eosin or immunostained for CD68.

RESULTS
Transient Accumulation of Cholesteryl Ester Core Aldehydes during Peroxidation of Cholesteryl Esters and LDL-We analyzed the aldehydic molecules generated during the peroxidation of cholesteryl linoleate and cholesteryl arachidonate. As shown in Fig. 1A, when cholesteryl linoleate was treated with 10 M Fe 2ϩ and 1 mM ascorbate in 50 mM sodium phosphate buffer (pH 7.4), the TBARS values reached a maximum at 24 h and decreased thereafter. The LC-MS analysis following the DNPH derivatization demonstrated that, in addition to the short-chain aldehydes such as acetaldehyde, hexanal, and 4-hydroxy-2-nonenal (Fig. 1B), the metal-catalyzed peroxidation of cholesteryl linoleate mainly produced 9-ONC (Fig. 1C). Upon autoxidation of cholesteryl arachidonate, 5-OVC was detected as the major product (Fig. 1D). These core aldehydes were estimated to account for a minimum of 1-2% of the consumed cholesteryl esters.
Subsequently, we characterized the formation of these core aldehydes during the peroxidation of LDL. To efficiently induce LDL peroxidation, Cu 2ϩ , instead of Fe 2ϩ , was used as the catalyst. Both Cu 2ϩ and Fe 2ϩ can be used to initiate LDL oxidation in vitro, since the resulting oxidized LDL has similar biological activities to that oxidized in vivo. However, Cu 2ϩ is more frequently used for the in vitro LDL oxidation, because Cu 2ϩ alone can effectively oxidize LDL, whereas a chelating agent, such as citrate, is required for the Fe 2ϩ -initiated LDL oxidation (16). As shown in Fig. 2A, the TBARS values reached a maximum between 12 and 24 h. In addition, the cholesteryl esters contained in the LDL were eventually consumed, and all of them disappeared within 24 h of incubation (Fig. 2B). In parallel to the extent of the LDL oxidation, there was a progressive increase in the number of cholesteryl ester core aldehydes, 9-ONC and 5-OVC (Fig. 2C). The decrease in the cholesteryl ester concentrations closely reflected the formation of these core aldehydes. It was noted that the amounts of both core aldehydes peaked (150 nmol of 9-ONC and 25 nmol of 5-OVC per mg of LDL) at 12 h and continuously decreased thereafter (Fig. 2D). This result suggests that these cholesteryl ester core aldehydes may be the intermediates that undergo further reactions during the LDL peroxidation.
Oxidative Conversion of Cholesteryl Ester Core Aldehydes to Their 7-Ketocholesterol Derivatives-It was anticipated that the decrease in the cholesteryl ester core aldehyde concentrations during LDL peroxidation might be due to their conversion to the corresponding oxidized products, such as the 7-ketocholesterol ester core aldehydes. To assess this presumption, we examined the formation of 7-ketocholesterol and its ester core aldehydes by LC-MS following DNPH derivatization. Because the peak for the DNPH derivative of 7-ketocholesterol overlapped with the peaks for the DNPH derivatives of 5-oxovaleroyl-7-ketocholesterol and 9-oxononanoyl-7-ketocholesterol (Fig. 3A), we determined these products together as the total amounts of 7-ketocholesterol and its ester core aldehydes. As shown in Fig. 3B, in contrast to the formation of TBARS and the cholesteryl ester core aldehydes (Fig. 2), the yields of 7-ketocholesterol and its ester core aldehydes were relatively low during the early incubation period of 8 h; however, accompanied by the decrease in the 9-ONC and 5-OVC concentrations after 12 h, the formation of 7-ketocholesterol and its ester core aldehydes was dramatically enhanced. These data suggest that the consumption of the cholesteryl ester core aldehydes closely reflects, at least in part, the oxidation of the cholesterol moieties to their corresponding 7-keto derivatives.
Covalent Binding of Cholesteryl and 7-Ketocholesteryl Ester Core Aldehydes to LDL ApoB-An alternative mechanism for the consumption of the cholesteryl ester core aldehydes may include their covalent binding to the apoB of LDL. During the oxidation of LDL, there is a progressive decrease in the amino groups. It has been generally assumed that this reflected the conjugation of short-chain aldehydes (e.g. malondialdehyde or 4-hydroxy-2-nonenol) with the lysine ⑀-amino groups (17). We wanted to learn to what extent, if at all, the bound cholesteryl ester core aldehydes contributed to the blocking of the lysine ⑀-amino groups. To characterize the putative bound cholesteryl esters, we prepared the oxidized LDL, extracted lipids, and then treated the apolipoprotein with dilute alkali. If the linkage occurs between a fragmented side-chain fatty acid aldehyde of the oxidized cholesteryl esters and a lysine residue of apoB, saponification should result in the release of cholesterol (or oxidized cholesterol). In fact, we determined the release of primarily cholesterol (Fig. 4A) and 7-ketocholesterol (Fig. 4B). Furthermore, the amount of the released cholesterol increased right after the beginning of the incubation and reached a plateau at 24 h, whereas the release of 7-ketocholesterol was observed at and after 8 h of incubation (Fig. 4C). These data establish the fact that the cholesteryl ester core aldehydes could be covalently linked to apoB.
9-ONC-Lysine Schiff Base Adduct-9-ONC was presumed to be covalently bound to the lysine residues of the protein via a Schiff base linkage. Indeed, upon incubation of BSA with 9-ONC, the loss of the lysine residues was quite significant (Fig. 5A). In an attempt to detect the 9-ONC-lysine Schiff base adduct in protein by the amino acid analysis, the authentic 9-ONC-lysine adduct was prepared by incubating N ␣ -benzoylglycyl-lysine with 9-ONC, reductively aminating it with NaC-NBH 3 , and then characterizing it by LC-MS and amino acid analyses. As shown in Fig. 5B, the LC-MS analysis of the major product eluted at 21 min showed a pseudomolecular ion peak at m/z 832.1 (M ϩ H) ϩ , corresponding to the reduced form of the 9-ONC-lysine adduct. Upon amino acid analysis of the reaction mixtures, we detected a new peak, which might have originated from the 9-ONC-lysine adduct (data not shown). However, it was found that the peak of 9-ONC-lysine completely overlapped that of histidine in the amino acid analysis.
Preparation of Monoclonal Antibody against 9-ONC-Lysine Schiff Base Adduct-As an alternative approach for the detection of the 9-ONC-lysine Schiff base adduct in the protein, we attempted to raise the monoclonal antibody specific to the 9-ONC-modified protein. Mice were immunized with the 9-ONC-modified KLH, and, during the preparation of the monoclonal antibodies, hybridomas were selected by the reactivities of the culture supernatant to the 9-ONC-modified BSA. We finally obtained one clone (2A81), which showed the most distinctive recognition of the 9-ONC-modified BSA. The isotype on the monoclonal antibody was IgM with a light chain. Upon incubation with BSA, 9-ONC generated immunoreactive materials with mAb 2A81 in time-and dose-dependent manners (Figs. 6, A and B). Despite an extensive screening of the hybridomas, which produce monoclonal antibodies specific to the 9-ONC-modified BSA, it is still conceivable that the antibody recognizes epitopes originating from other lipid peroxidation products. Hence, we examined the immunoreactivity of the antibody to the aldehyde-treated BSA by direct ELISA and found that, among the aldehydes tested, 9-ONC was the only source of immunoreactive materials generated in the protein (Fig. 6C). In addition, the monoclonal antibody was found to recognize the protein-bound 9-ONC reductively aminated with NaCNBH 3 , indicating that the Schiff base double bond might not be essential for the antibody recognition.
Formation of Antigenic Structures in the Protein Exposed to the Iron/Ascorbate-mediated Oxidation of Cholesteryl Esters and in Oxidized LDL-To examine whether the peroxidation of cholesteryl esters could generate antigenic structures in protein, we incubated the cholesteryl esters (cholesteryl linoleate, cholesteryl arachidonate, and cholesteryl oleate) with an iron/ ascorbate-mediated free radical generating system in the presence of BSA and examined their formations in the modified proteins by ELISA. As shown in Fig. 7A, the antibody showed a significant immunoreactivity with the protein exposed not only to the oxidized cholesteryl linoleate but also to the oxidized cholesteryl arachidonate. In addition, the competitive ELISA analysis showed that the antibody binding to the 9-ONC-modified protein was not inhibited by free 9-ONC and 7-ketocholesterol but significantly inhibited by the protein-and peptide-bound 9-ONC (Fig. 7, B and C). These data suggest that the antibody extensively recognizes the protein-bound cholesteryl ester core aldehydes, including 9-ONC-lysine and 5-OVC-lysine.
To determine whether antigenic structures recognized by mAb 2A81 are formed in oxidized LDL, LDL treated with 5 M Cu 2ϩ was subjected to an agarose gel electrophoresis followed by immunoblot analysis with mAb 2A81. As shown in Fig. 7D, the native form of the LDL appeared as a single protein band that was readily visualized by Fat Red 7B staining; however, the LDL incubated with 5 M Cu 2ϩ exhibited enhanced anodic mobility compared with the native LDL, indicating an in-  6. Specificity of mAb 2A81. A, dose-dependent formation of immunoreactive materials with mAb 2A81 in the 9-ONC-treated BSA. BSA (0.5 mg/ml) was incubated with 9-ONC (0 -5 mM) in 50 mM sodium phosphate buffer (pH 7.4) for 24 h at 37°C. B, time-dependent formation of immunoreactive materials with mAb 2A81 in the 9-ONC-treated BSA. BSA (0.5 mg/ml) was incubated with 5 mM 9-ONC in 50 mM sodium phosphate buffer (pH 7.4) at 37°C. C, immunoreactivity of mAb 2A81 to the aldehyde-treated BSA. Affinity of mAb 2A81 was determined by direct ELISA using aldehyde-treated BSA as the absorbed antigen. A coating antigen was prepared by incubating BSA (0.5 mg/ml) with 5 mM aldehyde in 1 ml of 50 mM sodium phosphate buffer (pH 7.4) for 2 h at 37°C. creased negative charge of the molecule due probably to the modification of the ⑀-amino group of the lysine residues (upper panel). An agarose gel electrophoresis/immunoblot analysis of the Cu 2ϩ -oxidized LDL using mAb 2A81 revealed the formation of immunoreactive materials, which were not detected in the native LDL (lower panel).
Localization of Antigenic Structures in Human Atherosclerotic Lesions-The recent findings that the level of 9-ONC increases in the plasma of patients with atherosclerosis (13,14) suggested that the 9-ONC-lysine Schiff base adduct was most likely accumulated in an atherosclerotic lesion. Hence, atherosclerotic aortas were immunohistochemically examined for protein-bound cholesteryl ester core aldehydes using mAb 2A81. As shown in Fig. 8, the macrophages appearing in the atherosclerotic plaques were identified in hematoxylin-eosin-stained (A) or CD68-immunostained sections (B). The cytoplasm in most of the foamy or spindle macrophages was immunoreactive with mAb 2A81 (Fig. 8C). No immunoreaction product deposits were detectable in a section incubated with mAb 2A81 preabsorbed with the specific antigen (Fig. 8D). There was no significant difference in the localization of the antigenic structures between the paraffin-embedded and frozen sections (Supplementary Figs. 1 and 2). In addition, neither formalin postfixation nor EDTA addition had any effect on the immunohistochemical results (Supplementary Figs. 1 and 2). These data suggest that the generation of antigenic structures recognized by the antibody during the dehydration, embedding, rehydration, and staining processes is unlikely and that the observed immunoreactivities reflect the endogenous products generated in vivo. Thus, the detection of protein-bound cholesteryl ester core aldehydes in the atherosclerotic plaques supports the no-tion that the reaction between oxidized cholesteryl esters and primary amines might represent a process common to the LDL modification during aging and its related diseases. DISCUSSION The formation of lipid peroxidation products bound to proteins in vascular lesions, such as the atherosclerotic lesion, is a phenomenon common in most, if not all, types of vascular damage associated with oxidative stress. The possibility that aldehydic molecules generated from lipid peroxidation play a role in the pathogenesis of atherosclerosis is suggested by the facts that (i) the level of reactive aldehydes increases in plasma in relation to extensive aortic atherosclerosis, (ii) high concentrations of aldehydes can be generated during the oxidation of LDL phospholipids, (iii) the structural and functional changes associated with the in vitro oxidation of LDL can also be produced by the direct interaction of LDL with aldehydes, and (iv) the reaction of aldehydes with a critical number of lysine residues of the LDL apoB produces internalization by the scavenger receptor of human monocyte-macrophages and subsequent intracellular accumulation of the lipoprotein-derived cholesteryl ester. On the other hand, it has been shown that atherosclerotic lesions of varying severity from the human aorta contain material recognized by antibodies raised against protein adducts with aldehydes. Haberland et al. (18) and Palinski et al. (19) demonstrated that monoclonal antibodies directed against MDA-modified LDL bind to epitopes in atherosclerotic lesions. They proposed the hypothesis that the formation of MDA and modification of the lysine residues of LDL apoB occur in vivo as a prerequisite to the formation of arterial foam cells and contribute to the development of ather- osclerosis. The participation of reactive aldehydes in atherosclerosis is also suggested by a series of immunohistochemical analyses of atherosclerotic lesions from the human aorta using various antibodies against short-chain aldehyde-amino acid adducts, such as 4-hydroxy-2-nonenal-histidine (20), MDA-lysine (21,22), and acrolein-lysine (23), in which the immunoreactivities are commonly associated with cells, primarily macrophages. These studies provided evidence for the presence of oxidized LDL (or at least of antigens closely related to it) in arterial lesions at significant concentrations and suggested that reactive aldehydes may play a role in the formation of arterial foam cells.
It is well established that unesterified aldehydes formed during LDL oxidation can modify lysine or histidine residues of apolipoproteins by Schiff base or Michael adduct formation (24 -26). LDL modified in this way is no longer recognized by the LDL receptor but internalized via hepatic or macrophage scavenger receptor-mediated pathways (27). However, the ability of peroxidized lipids to covalently react with proteins is not limited only to low molecular weight short-chain aldehydes. It was demonstrated that, after oxidation, radiolabeled phospholipids that were incorporated into the LDL became associated with apoB (28,29). Itabe et al. (30) also provided evidence that oxidized phospholipids form complexes with lysine residues on proteins due to the presence of 9-oxononanoylphosphatidylcholine. In addition, the Schiff base adducts of phospholipid core aldehydes with lysine residues of myoglobin (31) and thyroglobin (32) have been identified. Furthermore, the binding of phospholipids to apoB during LDL oxidation was directly shown by measuring the phosphorus incorporated into the protein (15). These findings strongly suggested that the formation of lipid-protein complexes could occur via interactions between the carbonyl groups of the phospholipid core aldehydes and lysine residues on proteins. Meanwhile, the possibility that oxidized cholesteryl esters might bind to the protein moiety during LDL oxidation has received little attention. As far as we know, there is only one study, in which upon incubation of the purified 9-ONC with a lysine derivative, the formation of the 9-ONC-lysine Schiff base adduct was ascertained by LC-MS analysis (13).
In the present study, to investigate the mechanisms contributing to the modification of LDL, we analyzed the oxidized cholesteryl esters generated during the autoxidation of LDL and characterized their covalent binding to the lysine residues of LDL apoB. We observed that the peroxidation of LDL with Cu 2ϩ produced cholesteryl ester core aldehydes, such as 9-ONC and 5-OVC, as the major oxidized cholesteryl esters. The levels of 9-ONC and 5-OVC reached peaks at 12 h and significantly decreased thereafter (Fig. 2D). In addition, the reduction of the core aldehyde levels was accompanied by the formation of the free 7-ketocholesterol and 7-ketocholesteryl ester core aldehydes (Fig. 3) and by the increase in the amounts of the apoBbound cholesterol and 7-ketocholesterol (Fig. 4), suggesting that the cholesteryl ester core aldehydes were further converted to their 7-ketocholesterol and apoB-bound derivatives. On average, there were about 97 mol of cholesterol and 7-ketocholesterol attached per mol of apoB after 48 h of the coppercatalyzed oxidation. The overall decrease in the number of lysine residues averaged 250 mol/mol of apoB. Thus, 38.4% of the conjugated lysine amino groups appeared to be attributable to the cholesteryl ester core aldehydes. The remainder is presumably attributable to conjugation with phospholipid core aldehydes and with nonester fragments from polyunsaturated fatty acids, including MDA and 4-hydroxy-2-nonenal, and the many other aldehydic fragments that have been demonstrated (17).
To further ascertain the formation of 9-ONC covalently bound to protein in vivo, we obtained the murine monoclonal antibody, mAb 2A81, which clearly distinguished the 9-ONCmodified protein from the native protein (Fig. 6). Characterization of the antibody revealed that the monoclonal antibody was directed against protein-bound cholesteryl ester core aldehydes (Figs. 6 and 7). The in vivo detection of antigenic structures using mAb 2A81 was attempted in the tissue samples from the patients with atherosclerosis, which is considered to be a form of chronic inflammation resulting from the interaction between the modified lipoproteins, monocyte-derived macrophages, T cells, and the normal cellular elements of the arterial wall. We confirmed that atheromatous lesions indeed contained proteinbound cholesterols, colocalizing mainly with foamy macrophages (Fig. 8). It is known from the in vitro studies that all of the major cell types within the atherosclerotic lesions are capable of promoting the oxidation of LDL (1,2). Therefore, the observed cell-associated staining patterns are probably attributable to the cellular oxidation of LDL by endothelial cells, macrophages, and smooth muscle cells. The resulting oxidized LDL may be taken up by cells and may be the ultimate source of the lipids that accumulate in the atherosclerotic lesions. In addition, the intracellular granular staining observed in the atherosclerotic lesions represents the presence of proteinbound cholesterols that had already been taken up by the macrophages and are present within the cell in cytoplasmic FIG. 8. Immunohistochemical detection of antigenic structures in atherosclerotic aorta. In the neointima, foamy macrophages (arrows), which were identified in sections stained with hematoxylin-eosin (A) and immunostained with anti-CD68 antibody (B), were immunoreactive with mAb 2A81 (C). For a competitive experiment, mAb 2A81 preincubated with the specific antigen was used (D). A-D, magnification was ϫ100.
organelles (33). This leads to speculation that the modification by cholesteryl ester core aldehydes renders proteins relatively resistant to intracellular proteolytic degradation, resulting in the marked accumulation of epitopes in the macrophages. On the other hand, it has also been shown that core aldehydes produced from oxidized cholesteryl ester are less efficiently hydrolyzed by an acidic macrophage cholesteryl ester hydrolase than the native cholesteryl ester and that the site-specific release of core aldehydes in lysosomes could lead to the irreversible intralysosomal deposition of the lipid-protein complexes (13). This is compatible with lipid engorgement in lysosomes (34) and partial lysosomal enzyme inactivation after uptake of the oxidized LDL (35). Also, the predominant lysosomal localization of the epitopes specific for oxidized proteins (33), the lysosomal accumulation of ceroid (36), and the partially degraded apoB from oxidized LDL (37) indicate that the lysosome is the site where potential pathophysiological consequences of the core aldehydes could be manifested. In line with the suggestion of a relatively slow turnover, cholesteryl ester core aldehydes of different chain lengths have been found in the lipid extracts of atherosclerotic plaques (13,14).
In summary, we analyzed the oxidized cholesteryl esters, such as 9-ONC and 5-OVC, generated during the peroxidation of LDL, and found that these core aldehydes were converted to their 7-ketocholesterol and apoB-bound derivatives. Moreover, we raised the monoclonal antibody directed to the proteinbound 9-ONC and obtained new murine monoclonal antibodies, mAb 2A81, that clearly distinguished the 9-ONC-modified protein from the native protein. The observation that the monoclonal antibody did not cross-react with proteins that had been treated with the reactive aldehydes but rather with the oxidized cholesteryl esters suggested that the antibody extensively recognized the protein-bound cholesteryl ester core aldehydes. Using mAb 2A81, we showed that immunoreactive materials were indeed formed not only in the oxidized LDL but also in the atherosclerotic lesions of the human aorta. These data suggested that the reaction between oxidized cholesteryl esters and primary amines of the proteins might represent a process common to the formation of carbonyl-modified proteins during aging and its related diseases.