Macrophages Can Decrease the Level of Cholesteryl Ester Hydroperoxides in Low Density Lipoprotein*

Murine and human macrophages rapidly decreased the level of cholesteryl ester hydroperoxides in low density lipoprotein (LDL) when cultured in media non-permissive for LDL oxidation. This process was proportional to cell number but could not be attributed to the net lipoprotein uptake. Macrophage-mediated loss of lipid hydroperoxides in LDL appears to be metal ion-independent. Degradation of cholesteryl linoleate hydroperoxides was accompanied by accumulation of the corresponding hydroxide as the major product and cholesteryl keto-octadecadienoate as a minor product, although taken together these products could not completely account for the hydroperoxide consumption. Cell-conditioned medium possessed a similar capacity to remove lipid hydroperoxides as seen with cellular monolayers, suggesting that the activity is not an integral component of the cell but is secreted from it. The activity of cell-conditioned medium to lower the level of LDL lipid hydroperoxides is associated with its high molecular weight fraction and is modulated by the availability of free thiol groups. Cell-mediated loss of LDL cholesteryl ester hydroperoxides is facilitated by the presence of α-tocopherol in the lipoprotein. Together with our earlier reports on the ability of macrophages to remove peroxides rapidly from oxidized amino acids, peptides, and proteins as well as to clear selectively cholesterol 7-β-hydroperoxide, results presented in this paper provide evidence of a potential protective activity of the cell against further LDL oxidation by removing reactive peroxide groups in the lipoprotein.

Oxidation of low density lipoprotein (LDL) 1 in the arterial intima is believed to play a key role in atherogenesis (1,2). It is likely that the cells present in the arterial wall and developing lesions participate in the oxidative process (1,3). The mechanisms by which these cells promote lipoprotein oxidation are of great interest and have been investigated in numerous studies (4), as they could be potential targets for therapeutic or preventative intervention in the disease process. In addition to pro-oxidant action of cells toward LDL, it has been demonstrated that under certain conditions the cells could have potential antioxidant activity toward the lipoprotein. For example, in some situations macrophages could inhibit metalcatalyzed lipid peroxidation (5,6) by sequestering metals from the culture medium. Endothelial cells have also been shown to prevent formation of lipid hydroperoxides in LDL (7); this activity switched to a pro-oxidant action of the cells upon increasing the initial concentrations of hydroperoxides in LDL or metal ions in media. A potential protective role has been demonstrated for human hepatic cells, which selectively take up and detoxify cholesteryl ester hydroperoxides from high density lipoprotein particles (8). We have previously reported the capacity of murine macrophages to remove peroxides rapidly from oxidized amino acids, peptides, and proteins (9). In the case of the oxidized proteins at least, this process was likely to be extracellular, as it was not accompanied by significant net uptake of the protein molecules. In the present study, we demonstrated the ability of cells to extracellularly detoxify cholesteryl ester hydroperoxides in LDL in a culture medium that did not support cell-mediated LDL oxidation, and we studied possible mechanisms of this process. We also tested different cell types for their ability to reduce levels of lipid hydroperoxides in lipoprotein and regulation of this process. 6-week-old Quackenbush Swiss strain mice, after CO 2 asphyxiation, by peritoneal lavage with ice-cold DMEM, containing 0.38% (w/v) sodium citrate, penicillin G (100 units/ml), and streptomycin (100 g/ml). The isolated cells were immediately plated in 12-well (22-mm diameter) cell culture plates (Falcon) at 4 ϫ 10 6 cells per well or at the indicated density for cell number dependence experiments. The cells were incubated at 37°C for 2 h and then washed three times with PBS to remove non-adherent cells. The cells were further incubated in DMEM containing 10% FCS for 18 h. J774 A.1 mouse monocyte-macrophage cells (J774, obtained from the American Type Culture Collection, batch F-10089) were grown in DMEM containing 10% (v/v) FCS in 160-cm 2 cell culture flasks (Becton Dickinson). Cells were subcultured at 1:5 dilution every 6 -7 days. The day before an experiment the cells were plated in 6-well (35-mm diameter) cell culture plates (Falcon) at 2.5 ϫ 10 6 cells per well.
Human monocytes were isolated from white cell concentrates using centrifugal elutriation as described previously (11). Purified monocytes (Ͼ90% pure as judged by nonspecific esterase staining) were adhered in 22-mm cell culture wells (Falcon) at 4 ϫ 10 6 cells per well and cultured in RPMI 1640 containing 10% (v/v) HS for 7 days to allow them to differentiate to macrophages.
The cells were cultured in a humidified incubator at 37°C in 5% CO 2 in air. All tissue culture media were supplemented with 20 mM glutamine, 100 units/ml penicillin G, and 100 g/ml streptomycin. Cell viability was assessed by the trypan blue exclusion test. Prior to experiments cells were washed twice with warm PBS.
Preparation of Cell-conditioned Medium, Separation of High and Low Molecular Weight Fractions of Cell-conditioned Medium-1-ml portions of DMEM were added to cells and incubated for 2.5 h at 37°C in 5% CO 2 in air. The medium pooled from several cell culture wells was spun (1,000 ϫ g, 5 min, 5°C) to remove any detached cells, and the supernatant was used as the cell-conditioned medium.
In some experiments 2 ml of cell-conditioned medium was spun in a Centricon-10 concentration unit (Amicon) at 4,900 ϫ g and 4°C for 45 min, and then the high and low molecular weight fractions (retentate and filtrate, respectively) were reconstituted to the original volume with DMEM. Separation of high and low molecular weight components using filtration through a PD-10 column is described in detail in the legend to Fig. 3.
Preparation of LDL-LDL (density range 1.02-1.05) was prepared from plasma from fasted normolipidemic healthy volunteers as described in detail (12) by density gradient ultracentrifugation (Beckman L8-M ultracentrifuge) using a vertical (VTi50) rotor for 2.5 h at 242,000 ϫ g (mean) and 10°C. A second centrifugation at a density of 1.063 g/ml with Ti70 rotor (242,000 ϫ g (mean), 22 h, 10°C) removed traces of contaminating albumin. The LDL was subsequently dialyzed for 18 h at 4°C against 4 ϫ 1 liter of deoxygenated PBS containing chloramphenicol (0.1 g/liter) and EDTA (1.0 g/liter), filter-sterilized (0. 45-m), and stored in the dark at 4°C until use (within 7 days). Immediately prior to oxidation or incubation with cells, LDL was passed through two consecutive PD-10 columns equilibrated with Chelex 100-treated PBS to remove EDTA. LDL was further sterilized (0.45-m filter) for cellular experiments.
Mild Oxidation of LDL, in Vitro Preparation of Mildly Oxidized ␣-TOH-depleted LDL-An aliquot of LDL (in the range of 2-4 mg of LDL protein/ml) was incubated with 1 mM AAPH at 37°C for approximately 1 h. Under these conditions of low radical flux, oxidation of LDL resulted in formation of small amounts of lipid hydroperoxides with a less than 10% loss of ␣-tocopherol (␣-TOH) and no changes in protein and lipid composition (13,14). AAPH was removed by two sequential PD-10 column filtrations. The mildly oxidized LDL used in experiments in this study contained 28.5 Ϯ 12.5 nmol CEOOH per mg of LDL protein (mean Ϯ S.D., n ϭ 25) with TOH content being 8.2 Ϯ 3.9 nmol/mg (mean Ϯ S.D., n ϭ 25).
Oxidized ␣-TOH-depleted LDL was prepared as described in detail (13). LDL (ϳ2-4 mg/ml) was oxidized with 50 mM AAPH at 37°C for 20 -30 min. AAPH was removed by two sequential gel filtrations using the PD-10 columns. The resulting LDL contained undetectable levels of ␣-TOH and cholesteryl ester hydroperoxides at levels comparable to that reached in mildly oxidized TOH-containing LDL (prepared as above). Oxidized ␣-TOH-depleted LDL was without other significant detectable changes in lipid composition (13).
In Vitro Replenishment of TOH-depleted LDL with ␣-TOH-TOHdepleted LDL prepared as above was replenished with ␣-TOH as described (13). In detail, an aliquot of TOH-depleted LDL was incubated with lipoprotein-deficient serum (LPDS) (2:1, v/v) supplemented with ␣-TOH (dissolved in Me 2 SO (15), final concentration of Me 2 SO Ͻ1%) at 37°C for 4 h. LPDS was prepared from plasma from fasted normolipi-demic healthy volunteers (16) and contained no detectable levels of ␣-TOH. The final concentration of ␣-TOH in the incubation mixture was seven times that of the corresponding native unoxidized LDL. Control mildly oxidized and control TOH-depleted LDL were incubated with LPDS in the presence of Me 2 SO (Ͻ1%, v/v). After incubation the density of the mixture was adjusted with KBr to 1.21, and LDL was then reisolated using a quick LDL isolation method (17) with a 3.5-h centrifugation. The LDL was passed through two consecutive PD-10 columns, filter-sterilized, and used immediately.
Incubation of LDL with Cells-1-ml portions of DMEM (or 1.5-ml for J774 cultures) containing 100 g of LDL protein were incubated in cell culture wells at 37°C in 5% CO 2 in air with or without cells. In experiments using cell-conditioned medium, 1 ml of this was mixed with 100 g of LDL and incubated in cell-free wells under the above conditions.
Lipid Extraction-At the times indicated, the LDL-containing media were removed from the cell culture wells and spun in an Eppendorf centrifuge at 16,000 ϫ g for 2 min at 4°C to remove any detached cells. 0.5 ml of the supernatant (or cell-free medium) was extracted with 1 ml of cold methanol and 5 ml of hexane and then centrifuged (1,000 ϫ g, 5 min, 10°C). 4 ml of the hexane phase were withdrawn, evaporated under vacuum, and redissolved in 200 l of isopropyl alcohol. Samples were sealed in glass vials and stored at Ϫ80°C until HPLC analysis (within 7 days). Recovery of ␣-TOH and lipid in the extracts was Ն99%; ␣-TOH, cholesteryl esters (CE) and their hydroperoxides (CEOOH) were stable under this condition for at least 6 weeks (data not shown).
In experiments where cellular lipid analysis was carried out, the cells were washed three times with warm PBS, and the cell lipids were extracted twice with 1 ml of hexane/isopropyl alcohol (3:2, v/v) (12). The cells were then lysed by incubation in 0.5 ml of cold 0.2 M NaOH for 15 min at 4°C, and the lysates were stored at Ϫ20°C for protein assay (within 7 days).
HPLC Analysis-Lipid extracts from cells or LDL-containing media were analyzed using reverse phase-high performance liquid chromatography (RP-HPLC) with a Supelco ODS column (25 ϫ 0.46 cm, 5 M particle size with a 2-cm Pelliguard guard column) as described (18) using a mobile phase of ethanol/methanol/isopropyl alcohol (19.5:6:1, v/v/v) containing 5 mM lithium perchlorate. Analysis was performed using electrochemical or fluorescent detection (for ␣-TOH, with Ϸ0.1 pmol and pmol detection limit, respectively), UV 210 (for free cholesterol (FC) and cholesteryl esters (CE) detection), or UV 234 (for oxidized products of CE) detection. Cholesteryl ester hydroperoxides (CEOOH) were analyzed by post-column chemiluminescence (CL) detection (detection limit Ϸ1-5 pmol). In some experiments, HPLC analysis was performed as in Kritharides,et al. (19) with separation of CEOOH and cholesterol ester hydroxides (CEOH) peaks. In brief, the lipids were separated on an ODS column using acetonitrile/isopropyl alcohol/water (44:54:2, v/v/v) as a mobile phase with UV 210 detection for unoxidized FC and CE, UV 234 detection for CEOOH and CEOH, and UV 279 for cholesteryl keto-octadecadienoate (CEϭO). Analysis of tocopherylquinone was performed by HPLC with electrochemical detection as described (20); the response factor was 34,187 area units per pmol of tocopherylquinone. 2 Quantitation of FC, individual CEs, and ␣-TOH was performed using calibration curves for each of the commercially available compounds. CEOOH were quantified using a cholesteryl linoleate hydroperoxide standard prepared as described previously (17,21). The concentration of CEOH was calculated using a cholesteryl linoleate hydroxide standard prepared by reduction of corresponding hydroperoxide standard with NaBH 4 .
Preparation of the  (19) with UV 234 and on-line radiometric detection (Radiomatic Flo-One Beta detector, Packard Instrument Co.) using Ultima-FLQ M scintillation mixture at 1.5 ml/min. The activity eluting between 12.6 and 15 min (i.e. retention time of Ch18:2-OOH standard) was collected, dried under argon, and redissolved in isopropyl alcohol. The purified [ 3 H]Ch18:2-OOH (specific radioactivity 1.5 Ci/ mmol) was quantified using RP-HPLC with post-column CL detection as described (18) with Ch18:2-OOH as a standard.
To prepare unilamellar liposomes aliquots of stock solutions of DL-␣phosphatidylcholine dimyristoyl (DMPC), FC, ␣-TOH, Ch18:2-OOH, and [ 3 H]Ch18:2-OOH were mixed, the organic solvent was evaporated under vacuum, and the lipids were resuspended in PBS. 1 ml of the lipid emulsion contained 0.2 mg of DMPC, 120 nmol of FC, 20 nmol of ␣-TOH, 30 nmol of Ch18:2-OOH, and 1.5 Ci of [ 3 H]Ch18:2-OOH; the concentrations of ␣-TOH and Ch18:2-OOH in the liposome preparation were chosen to achieve levels similar to those in the experiments with LDL. The mixture was mixed vigorously, and the emulsion was sonicated in three 1-min cycles on ice (Branson Sonifier 450). The resultant unilamellar liposomes were sterilized (0.45-m filter) and used within 1 h.
In the experiments with cell-conditioned medium an aliquot of liposomes was mixed with the cell-conditioned medium from MPM (1:10 dilution) and incubated for 2.5 h at 37°C. The lipids were then extracted as above and analyzed using RP-HPLC with UV 234 and radiometric detection and isopropyl alcohol/acetonitrile/water mobile phase (see above). The 3 H-labeled tracers were identified by comparing their retention times with those of known, unlabeled standards and of purified [ 3 H]Ch18:2-OOH. To assess the total radioactivity in each sample, 50-l aliquots of methanol and hexane layers of the lipid extracts were mixed with Ultima Gold HFP scintillation mixture and quantified on Liquid Scintillation Analyzer Tri-carb 2100TR (Packard Instrument Co.). More than 99.2% of the total radioactivity in each analyzed sample was detected in the hexane layer of the lipid extract, and further analysis of the methanol layer was not performed. To measure the concentrations of ␣-TOH, FC, and Ch18:2ϭO in a sample, an aliquot of each lipid extract was also subjected to a separate RP-HPLC run using identical chromatographic conditions with fluorescent (for TOH), UV 210 (for FC), and UV 279 (for Ch18:2ϭO) detection.
Measurement of Copper Release by Macrophages-MPM (7 ϫ 10 6 cells per 35-mm well in 6-well plate) were washed three times with PBS (37°C) and subsequently incubated at 37°C in 1.5 ml of HBSS containing 125 M BCS and 100 M ascorbate for 3 h. This had no effect on cell viability (22). Parallel control incubations were performed in the absence of cells. The culture supernatants were removed and centrifuged to remove any cells, and their absorbance was measured at 482 nm. In some experiments the cells were incubated for 3 h in 1.5 ml of HBSS, lacking either ascorbate or BCS, or both of them. After the supernatants were removed and spun, these reagents were added to make the medium complete, i.e. containing 100 M ascorbate and 125 M BCS. The concentration of copper was calculated using an extinction coefficient ⑀ ϭ 12.15 mM Ϫ1 cm Ϫ1 for the BCS⅐Cu(I) complex (22). Three separate wells were used for each condition in three independent experiments.
Protein Assay-The protein content of LDL samples and cell lysates was measured using the bicinchoninic acid method (Sigma) with BSA as a standard. BSA standards were prepared in water or in 0.2 M NaOH for the LDL preparations or cell extracts, respectively. The samples were incubated for 60 min at 60°C, and the absorbance at 562 nm was measured.

Time and Cell Number Dependence of LDL Cholesteryl Ester
Hydroperoxides Degradation by Mouse Peritoneal Macrophages-Incubation of MPM with LDL (100 g/ml) containing modest levels of CEOOH (28.5 Ϯ 12.5 nmol/mg LDL protein, mean Ϯ S.D. from 25 independent LDL preparations) in DMEM, the medium non-permissive for LDL oxidation, resulted in a time-dependent loss of CEOOH in LDL (Fig. 1A). In the same medium the level of CEOOH remained stable in LDL incubated in cell-free wells (Fig. 1A). The levels of free cholesterol and cholesteryl esters in the medium remained unchanged during incubation of cells with LDL under these conditions (not shown). This indicates that the observed loss of CEOOH in LDL was not likely to be due to the whole LDL particle uptake by cells or to continuing lipid peroxidation. When LDL was incubated with cells plated at different densities, the decrease in CEOOH concentration occurred at different rates, with higher rate corresponding to the higher cell number condition (not shown). This resulted in the different amounts of CEOOH remaining in LDL at the end of incubation (Fig. 1B). Parallel to reduction of CEOOH level, incubation of MPM with LDL in DMEM resulted in a decrease in the concentration of TOH in LDL (Fig. 1C), whereas no significant loss of TOH was observed in cell-free conditions (Fig. 1C). Loss of TOH in LDL also occurred in cell number- (Fig. 1D) and timedependent manner. No detectable amounts of tocopherylquinone were formed in LDL by the end of incubation with cells (not shown). Stoichiometry between changes in CEOOH and TOH levels was not consistent between experiments and varied from 1:1 to 6.7:1 in 16 independent experiments with the mean 3.5 Ϯ 1.6. Loss of CEOOH in LDL during incubation with cells ( Fig. 1 and 2A) was accompanied by formation of the corresponding hydroxides (CEOH), but this was not stoichiometric (Fig. 2C). The amount of CEOH formed by the end of incubation of LDL with cells was on average 42.5 Ϯ 18.5% of loss of CEOOH (mean Ϯ S.D. from 16 independent experiments), with the lowest and highest values being, respectively, 7.5 and 72%. Analysis of cholesteryl keto-octadecadienoate (CEϭO), another product of hydroperoxide metabolism, in media samples demonstrated formation of this product in the presence of cells but not in cell-free controls (Fig. 2D). However, the amount of CEϭO accumulated in LDL during its incubation with cells was severalfold less than amount of CEOOH lost in LDL (Fig.  2, A and D). Statistical analysis of data from six experiments in which this lipid was analyzed demonstrated that average accumulation of CEϭO was 9.8 Ϯ 6.1% of loss of CEOOH in LDL. In a separate experiment we studied stability of the CEOH in LDL in the presence of cells to examine the possibility that CEOH undergo further cell-mediated metabolism. Mildly oxidized LDL prepared as described under "Materials and Methods" was treated with NaBH 4 to reduce chemically all hydroperoxides to the corresponding hydroxides. After removal of non-reacted NaBH 4 by gel filtration of LDL through two sequential PD-10 columns, the LDL was mixed with DMEM (100 g/ml) and incubated with 4 ϫ 10 6 MPM for 2. each run in triplicate), suggesting that CEOH in LDL was stable in the presence of cells. These results indicate that cell-mediated loss of CEOOH in LDL seen in DMEM could not be completely explained by its stoichiometric reduction to CEOH or CEϭO.
To study whether the observed loss of TOH in LDL in the presence of cells could be explained by a transfer of TOH from LDL to cells, we analyzed both cellular lipid extracts and LDL lipid extracts before and after incubation of LDL with cells. The basal level of TOH in MPM plated at 4 ϫ 10 6 was 3.26 Ϯ 0.26 pmol/well (mean Ϯ S.D. for triplicate cultures). After 2 h incubation with LDL cellular TOH levels increased to 4.86 Ϯ 0.63 pmol/well, and the concentration of TOH in LDL decreased from 8.1 Ϯ 0.7 nmol/well to 4.1 Ϯ 0.1 nmol/well (mean Ϯ S.D.). The fact that TOH loss in LDL exceeds by 250-fold the increase in cellular TOH content during their incubation, together with our earlier data on the stability of TOH content in macrophages incubated in the absence of LDL (12), indicates that loss of TOH in LDL under our experimental conditions could not be accounted for by its selective transfer to cells.
Incubation of MPM with LDL in the different metal ion-free medium, HBSS, also resulted in reduction in CEOOH level (not shown). The rate of CEOOH disappearance in LDL in HBSS was 1.9 Ϯ 0.2 times lower than in DMEM (mean Ϯ range of two separate experiments). Similarly to DMEM, levels of FC or CE in LDL incubated with cells in HBSS remained unchanged (not shown). We demonstrated that the different capacity of cells to reduce levels of CEOOH in the two different media could not be explained by the presence of phenol red in DMEM (264.9 mg/ liter) and its absence in HBSS (not shown), and the reason for different cellular activity to clear LDL CEOOH in the two media was not further investigated in this study.
To test whether the observed clearance of LDL lipid hydroperoxides by cells was specific for mouse peritoneal macrophages, we performed similar experiments with human monocyte-derived macrophages (hMDM) and murine macrophagelike J774 cells. As shown in Table I, incubation of 4 ϫ 10 6 hMDM or 2.5 ϫ 10 6 J774 cells with LDL (100 g/ml) in DMEM resulted in approximately 50% loss of CEOOH in LDL after 2.5 h, with no changes in hydroperoxides seen in cell-free incubations. Levels of FC and CE in LDL remained unchanged in the presence or absence of cells. Peritoneal macrophages from C57BL/6J mice were also able to decrease the level of CEOOH in LDL (data not shown).
Incubation of Cell-conditioned Medium from Macrophages with LDL Results in Reduction of the CEOOH Content of LDL-To investigate whether the observed activity of cells to reduce level of CEOOH in LDL is an integral cellular activity or is secreted from cells, we studied changes in lipid composition in LDL during its incubation in cell-conditioned medium. Table  II demonstrates that cell-conditioned medium possessed a similar capacity to remove lipid hydroperoxides in LDL to that of cells. Loss of CEOOH in LDL incubated in cell-conditioned medium was accompanied by increase in the level of corresponding hydroxides, and CEϭO was formed in trace amounts by the end of incubation (Table II). Similar to the cell-containing condition, incubation of LDL with cell-conditioned medium resulted in loss of TOH in LDL (Table II) without formation of detectable amount of tocopherylquinone (not shown). No changes were observed in the levels of FC and CE under this condition (not shown). In cell-free control incubations lipid composition of LDL remained unchanged (Table II). Analysis of stoichiometry between the loss of CEOOH and the formation of CEOH in LDL incubated with cell-conditioned medium in 25 independent experiments revealed that the amount of CEOH formed was on average 71.4 Ϯ 30.1% of the amount of CEOOH degraded (mean Ϯ S.D.) (the experiment in Table II being one with the lower value). Dilution of cell-conditioned medium with Chelex-treated PBS led to a decrease in its efficiency to decrease the content of CEOOH in LDL; 2.4 times less CEOOH (10.7% of the initial level of CEOOH in LDL) were degraded in LDL incubated for 2 h in 1 to 1 diluted cell-conditioned medium compared with undiluted cell-conditioned medium (25.5% loss of the initial CEOOH level). We determined that the activity of cell-conditioned medium to clear CEOOH in LDL was stable for at least 3 h, if the medium was kept at ϩ4°C (not shown).
We next studied the kinetics of LDL CEOOH loss in cellconditioned medium. In contrast to the almost linear cell-mediated loss of CEOOH (Fig. 1A), disappearance of CEOOH in LDL in cell-conditioned medium had a biphasic character; the first fast stage normally finishing by 5-10 min after the start of incubation followed by a slow stage of almost linear CEOOH degradation. During the fast stage almost 64.4 Ϯ 15.1% (mean Ϯ S.D. of six independent experiments) of the total loss of CEOOH in LDL occurred, with the remaining Ϸ35.6% being degraded over more than 2 h. The kinetics of TOH loss in LDL incubated in cell-conditioned medium had a similar pattern to that of CEOOH loss in this condition; 59.6 Ϯ 16.9% of total loss of LDL TOH was observed within the first 5-10 min.
Similarly to the cell-conditioned medium from MPM, loss of   (8) with the reisolation of LDL as described (17). Preliminary experiments on the incorporation of non-radiolabeled Ch18:2-OOH into LDL demonstrated that a significant proportion of Ch18:2-OOH in liposomes was converted into the corresponding hydroxide during this procedure, so that the ratio of Their further characterization was not performed in this study, but they may include various positional and regio-isomers of oxidation products of Ch18:2 and/or adduct(s) with TOH. We expect that additional information on the products of Ch18:2-OOH degradation mediated by cells or cell-conditioned medium could be obtained with the use of cholesteryl-[ 14 C]linoleate. However, this compound is not commercially available, and we were unable to pursue the investigation in this direction further.
Investigation of the Involvement of Metal Ions in Cell-mediated Loss of CEOOH and TOH in LDL-In the presence of traces of transition metals, decomposition of CEOOH could occur via a Fenton-type reaction. To study the possible involvement of the metals that might derive from cells or be present in trace amounts in the culture medium, we performed a series of experiments with metal chelators. These were as follows: BCS, a high affinity chelator for Cu(I), the iron-selective chelator desferal (DF), EDTA, and DETAPAC which efficiently bind both iron and copper ions. Addition of any of these chelators to cell culture medium did not have significant effect on loss of CEOOH in LDL during its incubation with MPM (Table III). Cellular viability (assessed by a trypan blue test) was not affected by the presence of the indicated amounts of a chelator in medium (not shown). Loss of CEOOH in LDL incubated with cell-conditioned medium also was not significantly affected by the presence of any of the metal chelators used (Table III). We also measured free copper in the medium before and after incubation with cells using the BCS-ascorbate assay. No copper was detected in cell culture medium by this method, which has a detection limit of approximately 0.4 nM copper (I). No significant effect of the chelators was observed on loss of CEOOH mediated by J774 cells or cell-conditioned medium from these cells (data not shown).
In another series of experiments, we compared the degree of  CEOOH degradation by cells in DMEM that was pretreated with Chelex-100 resin (3 g/1 liter, for 18 h) to remove any contaminating transition metals versus untreated DMEM. When 4 ϫ 10 6 MPM were incubated with 100 g of LDL in Chelex-treated DMEM, the loss of CEOOH observed after 2 h was 99.7 Ϯ 25.6% (mean Ϯ S.D. of four independent experiments) of that seen during incubation of LDL with cells in not Chelex-treated DMEM. This indicates that there was no inhibitory effect of preincubation of medium with Chelex on cellmediated CEOOH degradation.
We also investigated whether cell-derived heme in hematoproteins or heme-containing enzymes is responsible for cellmediated loss of CEOOH in LDL. Addition of 200 M of the heme-ligand, sodium cyanide (NaCN), to cell-conditioned medium from J774 cells did not have a significant effect on CEOOH degradation (Table III). Neither did cyanide affect loss of LDL CEOOH in the presence of the cells (Table III).
These results suggest that the loss of LDL lipid hydroperoxides observed in the presence of cells or in cell-conditioned medium is unlikely to be a free metal ion-or heme-catalyzed process.

LDL CEOOH Loss in Cell-conditioned Medium Is Inhibited by Heat Treatment of the Medium and Is Associated with Its
High Molecular Weight Components-We investigated the component(s) of cell-conditioned medium responsible for loss of CEOOH in LDL. First, sensitivity of this component(s) to heat treatment was studied. Incubation of an aliquot of cell-conditioned medium at 60°C for 30 min prior to mixing with LDL (100 g/ml) resulted in a 65.5% (mean from two experiments) reduction in LDL lipid hydroperoxide clearance during 2 h incubation in comparison with control, not heat-treated, cellconditioned medium. Boiling of cell-conditioned medium at 100°C for 5 min led to loss of 88% of hydroperoxide-reducing activity of cell-conditioned medium (mean from two experiments). These results suggest involvement of a heat-labile component, possibly protein, secreted by cells in the clearance of CEOOH from LDL.
To investigate further the possible involvement of a protein in degradation of CEOOH in LDL, we subjected cell-conditioned medium to filtration through a Centricon-10 unit, which allows separation of components with molecular mass above and below 10,000 Da. Results presented in Table IV demonstrate that the high molecular weight (HMW) fraction of cellconditioned medium retained the capacity to reduce levels of CEOOH and TOH in LDL, whereas the low molecular weight (LMW) fraction virtually lacked activity to degrade CEOOH and had 2.4 times lower activity to reduce level of TOH as compared with original cell-conditioned medium. Although cell-derived HMW components had a high activity to clear LDL lipid hydroperoxides, 100% recovery of the initial activity was not reached. Combining HMW and LMW fractions of cell-conditioned medium did not result in increased CEOOH-degrading capacity compared with HMW fraction alone (not shown). This could be possibly due to the loss of some activity during filtration or absorptive losses in the Centricon unit.
Proteins and LMW components of cell-conditioned medium were also separated by gel filtration through a PD-10 column. Fig. 3 demonstrates that the first 3-ml fraction (denoted as I), eluted from the column after loading 3 ml of cell-conditioned medium and equivalent to the excluded components with mass Ͼ25 kDa, had the highest activity to reduce level of CEOOH in LDL as compared with two other consecutively eluted 3-ml fractions (II and III). The first fraction also had a high activity to degrade LDL TOH, similar to that in the original cellconditioned medium, whereas this activity in two other fractions was accordingly 4.3 and 4.8 times lower than in the fraction I.
The treatment of the cell-conditioned medium with trypsin (10 g/ml) for 15 min at 37°C prior to mixing it with LDL did not result in the inhibition of the ability of cell-conditioned medium to reduce the level of CEOOH in two independent experiments (data not shown). We also investigated the pH dependence of the activity of cell-conditioned medium to decrease the level of CEOOH in LDL. Aliquots of freshly collected cell-conditioned medium with pH 7.8 were adjusted to pH 5, 6, or 7 with the citrate/phosphate buffer or to pH 9 with the Tris buffer and mixed with LDL (100 g/ml). The loss of LDL CEOOH observed after 2.5 h incubation did not vary significantly between the different conditions in two independent experiments (data not shown), indicating that there was no clear pH optimum for the activity of cell-conditioned medium to reduce the level of CEOOH in LDL in the pH range between 5 and 9, consistent with the involvement of a non-enzymatic factor in this process.
Thus, the association of the activity of cell-conditioned medium to decrease the LDL levels of hydroperoxide and TOH with its high molecular weight fraction (greater than 10,000 Da) together with the loss of this activity after heat treatment suggest involvement of a protein in this reaction. The absence of the effect of trypsin on this activity possibly indicates that the treatment with trypsin under the used conditions leaves the reactive center on the protein intact.
Importance of Thiols for Cell-mediated CEOOH Degradation in LDL-Among cellular metabolites, thiols have been implicated in peroxide detoxification; they are essential cosubstrates  3. Activity of cell-conditioned medium to reduce levels of CEOOH (A) and TOH (B) is more likely associated with its high molecular weight fraction. Three-ml aliquot of cell-conditioned medium (obtained from incubation of DMEM with 4 ϫ 10 6 MPM for 2.5 h) was filtered through PD-10 column equilibrated with DMEM, and three consecutively eluted fractions (I, II, and III, 3 ml each) were mixed with LDL (100 g/ml) and incubated for 2.5 h. Levels of CEOOH and TOH in LDL before incubation as well as after 2.5 h incubation in cell-free DMEM (CF) and in cell-conditioned medium (c/c) are also presented. Values are expressed as nmol/mg LDL protein and are means Ϯ S.D. for triplicate incubations from a single experiment.
for enzyme-catalyzed detoxification of various peroxides (23) and can directly reduce various biological molecules (24), including hydroperoxides (9,25). Decomposition of peroxides could also occur via thiol-driven Fenton reactions in the presence of traces of transition metals, although the latter seem to be unlikely under the conditions of this study (as shown above). It has been demonstrated that cells release sulfhydryl compounds into the medium and can reduce extracellular disulfides such as protein disulfides or cystine (22, 26 -28). Cells can utilize extracellular cystine to produce cysteine (26,27,29) which upon release into the medium can generate sulfhydryl groups on proteins via sulfhydryl-disulfide exchange reactions (26,28); this implicates cystine as an intermediate in protein disulfide reduction.
To investigate the relevance of thiols in the process of cellmediated CEOOH clearance, we compared the efficiency of CEOOH degradation in cell-conditioned medium before and after treatment with N-ethylmaleimide (NEM). Addition of 100 M NEM, an agent commonly used to block free thiol groups, to cell-conditioned medium resulted in approximately 65% inhibition in CEOOH degradation, confirming thiol importance (Fig. 4A). No effect of NEM was observed on the level of CEOOH (Fig. 4A) and other LDL lipids (not shown) in cell-free incubations in the original DMEM. Similar inhibitory action of NEM on loss of CEOOH in LDL was observed in cell-conditioned medium from human monocyte-derived macrophages (Fig. 4B). Pretreatment of mildly oxidized LDL (1.9 mg/ml) with 4 mM NEM for 1 h on ice followed by removal of NEM by gel filtration through a PD-10 column did not affect the ability of cells to decrease the level of CEOOH in this LDL compared with untreated LDL (not shown). This suggests that thiol groups on apoB in LDL do not play a significant role in cellmediated degradation of LDL CEOOH.
We then tested whether the presence of cystine, a low molecular weight disulfide, in DMEM has an effect on cell-mediated CEOOH degradation, particularly because of its capacity to act as an intermediate in protein disulfide reduction. MPM incubated with LDL in DMEM containing cystine degrade approximately 3.4 times more CEOOH during 2.5 h than cells incubated in the cystine-deprived medium (Fig. 5A). The level of CEOOH in LDL remained unchanged in LDL incubated without cells regardless of the presence of cystine in the medium (Fig. 5A). Cell-mediated loss of TOH in LDL was also accelerated by the presence of cystine in the medium (Fig. 5B), with no changes in TOH concentration detected in cell-free conditions (Fig. 5B). The efficiency of cell-conditioned medium to reduce the level of CEOOH in LDL also depended on the presence of cystine during conditioning: 52.5 Ϯ 0.8% of LDL CEOOH were lost during 2.5 h incubation of LDL (100 g/ml) in cystine-containing cell-conditioned medium versus 17.5 Ϯ 1.06% in cystine-deprived cell-conditioned media, which were prepared by preincubation of 4 ϫ 10 6 cells with, respectively, cystine containing or deprived DMEM for 2.5 h. These results indicate that the presence of cystine in the culture medium facilitates degradation of CEOOH in LDL by cells or cell-conditioned medium.
Previous reports have shown that several cell types, including macrophages, can utilize cystine from the medium to produce extracellular low molecular weight thiols (22,26,27), predominantly cysteine, and release of low amounts of glutathione (GSH) has also been suggested (27). We tested the potential involvement of such thiols in degradation of CEOOH in LDL by incubating LDL in cell-free DMEM in the presence of reduced glutathione or cysteine in the range of concentrations reported for thiols exported by macrophages (22,27). The level of CEOOH in LDL remained unchanged after its incubation in DMEM containing 5, 10, or even 20 M of either thiol for 2.5 h, suggesting that at these concentrations the low molecular weight thiols are not directly responsible for cell-mediated degradation of LDL CEOOH. This agrees with the results presented above that low molecular weight components of cellconditioned medium are inefficient in mediating loss of CEOOH in LDL. However, when 10 M reduced glutathione was added to the cell-conditioned medium incubated with LDL, the loss of CEOOH increased by an additional 11.6%. From these results we can suggest that the stimulatory effect of cystine on CEOOH degradation by cells or cell-conditioned medium was probably due to its reported capacity to enhance generation of extracellular protein thiols (26,28), which have been shown to be more resistant to autoxidation than free low molecular weight thiols (26,28). The stimulatory effect of GSH on LDL CEOOH degradation seen in cell-conditioned medium could be explained by the same mechanism. Since the high molecular weight fraction of the cell-conditioned medium has the greatest ability to decrease levels of CEOOH, it is most likely that protein sulfhydryl groups are important for cellmediated CEOOH clearance.
Thus, the inhibitory effect of NEM on clearance of LDL CEOOH in cell-conditioned medium demonstrates the involvement of thiols in the extracellular cell-mediated detoxification of LDL lipid hydroperoxides. Low molecular weight thiols generated by cells are unable to mediate directly extracellular loss of CEOOH in LDL, perhaps due to their very low levels, and the stimulatory effect of cystine in cell culture medium on CEOOH degradation by cells or cell-conditioned medium may be explained by its enhancement of secreted protein thiol content in the system.
The Involvement of TOH in LDL in Cell-mediated CEOOH Degradation-As been shown earlier (Figs. 1-3 and 5 and Table IV), one of the features of cell-mediated loss of CEOOH in LDL is the accompanying decrease of its TOH level. Factors that affected CEOOH degradation also influenced the loss of TOH. To determine the relationship between the loss of TOH and CEOOH, we studied the stability of CEOOH in LDL that was depleted of TOH (denoted as (ϪTOH)LDL) but that contained CEOOH at levels comparable to that in preoxidized TOH-containing LDL (denoted as (ϩTOH)LDL). The ability of MPM to degrade CEOOH was significantly impaired in LDL depleted of TOH. After 2 h incubation of LDL with MPM plated at a density between 1 ϫ 10 6 and 4 ϫ 10 6 , the level of CEOOH remained unchanged in (ϪTOH)LDL, whereas cell number-dependent loss of hydroperoxides was observed in (ϩTOH)LDL with all hydroperoxides being lost in incubations with 3 ϫ 10 6 and 4 ϫ 10 6 MPM (Fig. 6). No significant CEOOH degradation in either LDL was observed in cell-free conditions (not shown). It should be noted that in some experiments a small loss of CEOOH in (ϪTOH)LDL during its incubation with 4 ϫ 10 6 MPM was observed; however, degradation of CEOOH in (ϪTOH)LDL was on average 31.5 Ϯ 34.6% (mean Ϯ S.D.) that in (ϩTOH)LDL in six independent experiments.
To test whether decreased CEOOH clearance in (ϪTOH)LDL was due to the absence of TOH, rather than some other unidentified changes in LDL occurring during its preparation, we performed a series of experiments in which (ϪTOH)LDL was replenished with TOH (denoted as (Ϫ/ϩTOH)LDL) using the procedure described under "Materials and Methods." We demonstrated that incubation of (ϪTOH)LDL or (ϩTOH)LDL with LPDS at 37°C for 4 h, which is a feature of the replenishment procedure, did not affect the level of subsequently reisolated LDL CEOOH and kinetics of their degradation by cells (not shown). Replenishment of (ϪTOH)LDL with TOH restored its capacity to decrease the level of endogenous CEOOH in the presence of cells, influencing both the kinetics of the reaction (the loss of half of the initial amount of CEOOH in (ϩTOH)LDL, (ϪTOH)LDL, and (Ϫ/ϩTOH)LDL was achieved correspondingly after 1.5, 3, and 1.5 h of incubation in a single experiment) and the total loss of CEOOH detected after 3 h of incubation (Table V). These results indicate that TOH in LDL is an important component in the process of cell-mediated clearance of LDL CEOOH. DISCUSSION It is generally accepted that among events contributing to the development of atherosclerosis oxidation of LDL plays an important role (1,2). The cells present in the arterial intima and developing lesions (macrophages, monocytes, smooth muscle cells, endothelial cells, and leukocytes) can oxidatively modify LDL in vitro (1,3), although the precise mechanism(s) of cell-mediated LDL oxidation are, as yet, incompletely understood. Despite the numerous studies aimed to investigate the pro-oxidant and therefore potential proatherogenic action of cells of the arterial wall toward LDL, little attention has been attracted to the possible antioxidant role of the cells. One approach to investigate preventative action of cells in relation to lipoprotein oxidation has been to study the effect of antioxidant status of cells on their ability to oxidize LDL (12,30). However, as we have recently demonstrated, increased ␣-TOH content of macrophages does not influence kinetics of lipoprotein oxidation by these cells in vitro (12). Reports in the literature indicate that under certain conditions macrophages and endothelial cells could inhibit lipid oxidation in LDL (5-7), one of the suggested mechanisms being sequestration of metals from the culture medium (5). It has been demonstrated that human hepatic cells could play a potentially protective role by selectively removing and detoxifying cholesteryl ester hydroperoxides from high density lipoprotein (8). We have recently reported that murine macrophages could rapidly remove peroxides from oxidized amino acids, peptides, and proteins (9), the process which, at least in the case of proteins, is likely to be predominantly extracellular. In this paper we demonstrated for the first time the ability of human and murine macrophages to reduce level of cholesteryl ester hydroperoxides in LDL, thus potentially decreasing further deleterious oxidative modifications to the lipoprotein.
We demonstrated that incubation of mouse peritoneal macrophages with LDL containing modest levels of CEOOH (up to 40 nmol/mg LDL protein), but otherwise without significant changes in its protein and lipid composition, in the medium not permissive for LDL oxidation (DMEM or HBSS) resulted in substantial loss of lipoprotein CEOOH within 2-3 h. This process was time-dependent, and loss of CEOOH increased with increase of number of cells in culture. A reduction of CEOOH level in LDL was also seen in the presence of murine J774 macrophages and human monocyte-derived macrophages, indicating that this activity was not specific for mouse peritoneal macrophages (from both QS and C57BL/6J mice). It is noteworthy that the levels of oxidized cholesteryl linoleate (in mmol/mol of Ch18:2), the most abundant cholesteryl ester in LDL, used in this study were comparable with those determined in LDL fractions from advanced human atherosclerotic plaques (31), the difference being that in mildly oxidized LDL used here the major oxidized product of Ch18:2 was its hydroperoxide derivative, whereas in advanced plaques more than two-thirds of the oxidized lipid is present in the form of the hydroxide (31).
In the course of investigating of mechanisms underlying  cell-mediated loss of CEOOH in LDL, we demonstrated that transition metal ions were not likely to play an important role in this process. Although very low amounts (0.2 M) of Fe(III) are present in DMEM, according to the manufacturer's specification, and cells potentially are able to extracellularly reduce oxidized transition metal (6,22), this reaction does not seem to contribute to LDL lipid hydroperoxide decomposition by cells. This is based on our finding that loss of CEOOH was not affected by the pretreatment of culture medium with Chelex resin. The lack of effect of high affinity metal chelators (DF, BCS, EDTA, and DETAPAC), both in the presence of cells and in cell-conditioned medium, indicates that degradation of CEOOH was unlikely to be mediated by cell-derived free or protein-bound metal ions. Moreover, it is well established that in the presence of low concentrations of transition metals cells rather promote LDL oxidation (4,22,32), and increased levels of preformed lipid hydroperoxides in LDL usually facilitate oxidative processes in the lipoprotein (7,(33)(34)(35)(36). The oxidation of LDL mediated by cells is characterized by loss of cholesteryl esters and accumulation of significant amounts of CEOOH at early stages, followed by formation of oxysterols, e.g. 7-ketocholesterol, at more advanced stages of LDL oxidation (19,37). No such changes were observed during incubation of LDL with cells in this study, supporting our conclusion of metal ionindependent degradation of CEOOH mediated by cells in DMEM or HBSS. Involvement of cell-derived heme in cellmediated loss of CEOOH was also excluded as heme-chelator sodium cyanide failed to affect the reaction. The ability of cell-conditioned medium to reduce the level of CEOOH in LDL similarly to cells allows us to exclude several potential mechanisms in playing the major role in macrophagemediated clearance of lipid hydroperoxides in LDL. The latter could not be attributed to the cellular uptake of whole lipoprotein particles, which is also supported by the fact that levels of free cholesterol and cholesteryl esters in the medium remained unchanged, or to the endocytosis/retroendocytosis process. It is also unlikely to be due to selective uptake of LDL CEOOH by macrophages, as has been demonstrated for hepatic cells in relation to HDL CEOOH (8), since results with cell-conditioned medium indicate that the activity of cells to reduce the concentration of CEOOH in LDL is not an integral component of the cell but is secreted from it. However, the possibility that the activities predominantly responsible for CEOOH clearance by cells or by cell-conditioned medium are different cannot be excluded. Supporting this is the different efficiency of conversion of CEOOH to CEOH in these two systems, i.e. 42.5 versus 71.4% on average. The loss of the activity to decrease the level of CEOOH in LDL after heat treatment of cell-conditioned medium and its association with high molecular weight component(s) indicate that the secreted factor mediating CEOOH degradation in LDL is likely to be a protein. Among known cell-derived proteins able to detoxify lipid hydroperoxides, a representative of selenium-dependent glutathione peroxidase family, phospholipid hydroperoxide glutathione peroxidase, has been shown to reduce cholesterol ester hydroperoxides in oxidized LDL in vitro (38,39). However, to our knowledge there have been no reports on the existence of a secreted form of this enzyme. Moreover, reduction of CEOOH by phospholipid hydroperoxide glutathione peroxidase resulted in formation of equimolar amounts of the corresponding hydroxides and did not affect the TOH level in LDL (39), whereas cell-mediated degradation of CEOOH in LDL reported in our study was accompanied by the consumption of TOH in LDL, and CEOH, although being the major product of CEOOH degradation, could not account for all CEOOH loss. This argues against the involvement of a peroxidase as a sole player in cell-mediated loss of LDL lipid hydroperoxides. The enzyme thioredoxin reductase has been demonstrated to be able to reduce directly some lipid hydroperoxides to the corresponding alcohol (40). However, because of the absence of equimolar accumulation of CEOH during cell-dependent CEOOH loss, this enzyme was also unlikely to be solely responsible for the CEOOH clearance by macrophages. Moreover, the absence of a clear pH optimum for this activity is consistent with the involvement of a nonenzymatic factor in this process. It has been recently reported that apoB mediates reduction of lipid hydroperoxides, including cholesteryl ester hydroperoxides (41). Although a role of apoB-100 in the decrease of the level of CEOOH mediated by cells could not be excluded, our results from experiments with liposomes incubated with cell-conditioned medium, where the loss of Ch18:2-OOH was observed in the absence of apoB, show that it is not required for CEOOH degradation and suggest that some other cell-derived protein plays a role in this process. This activity was not due to contamination of LDL preparations with plasma glutathione peroxidase, as incubation of LDL with LPDS for 4 h did not result in a decrease in level of LDL CEOOH, in agreement with early reports on lack of activity of plasma glutathione peroxidase toward complex lipid hydroperoxides, including CEOOH (42). Neither was this activity due to the possible contamination of LDL with human serum albumin which as been reported recently is able to reduce phospholipid hydroperoxides (43). In addition to the absence of 1 to 1 stoichiometry between changes in the levels of CEOOH and CEOH and stability of CEOOH in LDL during its incubation with LPDS, this is supported by our result that cysteine or glutathione added to the cell-free incubation of LDL at the concentrations reported to be secreted by macrophages does not promote CEOOH loss.
Cell-mediated loss of cholesteryl ester hydroperoxides in LDL could hypothetically result from extracellular hydrolysis of cholesteryl esters with release of oxidized free fatty acids and cholesterol. As the masses of unoxidized cholesteryl esters and cholesterol in LDL during its incubation with cells remained unchanged in this study, this process does not seem to contribute to CEOOH degradation. Moreover, hormone-sensitive lipase expressed in mouse peritoneal macrophages and murine J774 macrophages (44 -46) and responsible for hydrolysis of CE in these cells is an intracellular enzyme, and we are unaware of any reports on its secretion by cells. Cholesteryl ester hydrolase synthesized in human monocyte-derived macrophages has been reported to be secreted by the cells, but it is inactive in the absence of bile salts (46). Preliminary results from our laboratory indicate that cell-conditioned media from MPM or J774 cells lack neutral cholesteryl ester hydrolase activity. 3 However, the strongest argument against hydrolysis of oxidized cholesteryl esters as the mechanism for cell-mediated CEOOH loss in LDL derives from the experiments with the radiolabeled Ch18:2-OOH. The absence of the detectable amounts of [ 3 H]cholesterol during the incubation of cell-conditioned medium with [ 3 H]Ch18:2-OOH-containing liposomes excludes the hydrolysis of the oxidized cholesteryl ester under these conditions.
The inhibition of the capacity of cell-conditioned medium to reduce levels of CEOOH in LDL by NEM suggests that this process depends on the availability of free thiol groups. As has been demonstrated earlier, macrophages secrete low and high molecular weight thiols (22), and thiol production significantly depends on the presence of cystine in the culture medium (27,22). LMW thiols are not directly responsible for loss of CEOOH as was demonstrated by the inactivity of the LMW fraction of 3 A. Brown, personal communication.
cell-conditioned medium and the lack of effect of added GSH or cysteine at relevant concentrations (22,27) on CEOOH levels in cell-free medium. However, they may contribute to the process by providing reducing equivalents to a protein (26,28). The latter could also be mediated by thioredoxin, which could act either extracellularly (47) or in association with the cell surface (48); however, this was not investigated in the present study. Protein-associated sulfhydryl groups are more resistant to autoxidation than LMW thiols (26,28) and therefore may be present at levels necessary for CEOOH degrading activity. Cystine in the culture medium probably helps to maintain the required level of protein-bound thiols in the medium, thus explaining its stimulatory effect on CEOOH degradation by cells or cell-conditioned medium. This could be also the mechanism by which GSH at concentrations not able to reduce directly CEOOH in cell-free medium stimulates its loss when added to cell-conditioned medium. It has been demonstrated that protein-associated thiols could determine folding of proteins and their overall structure (49). They could also play a key role in activation/inactivation of enzymes via rapid chemical transformation of the mercapto group. Thiols could also directly participate in the redox reactions with or without free radical formation.
␣-Tocopherol in LDL is an important component in reaction of CEOOH degradation mediated by cells. This is based on the following results: (i) reduction of CEOOH level in LDL was accompanied by the decrease in concentration of TOH in lipoprotein, which is not associated with its transfer to cells; (ii) factors that affected CEOOH degradation similarly influenced loss of TOH; and most important (iii) CEOOH degradation by cells was significantly inhibited in TOH-depleted LDL and restoration of TOH into the LDL also restored the activity. Although the mechanism of cellular detoxification of CEOOH has not yet been completely defined, we propose that LDL TOH facilitates reaction of CEOOH with cell-derived factors via, preferentially, some non-free radical mechanism, e.g. it could act as a reductant. We cannot, however, totally exclude formation of radical species in our complex system, as they, for example, could contribute to generation of small amounts of cholesteryl keto-octadienoate in LDL seen in the presence of cells. Interestingly, ␣-TOH has been reported to enhance the peroxidase activity of hemoglobin on phospholipid hydroperoxide in a mechanism hypothetically involving an alkoxyl radical intermediate (50).
In conclusion, the results presented in this paper provide evidence of the ability of macrophages to extracellularly detoxify endogenous LDL lipid hydroperoxides. This activity is metal ion-independent and appears to be associated with a cell-derived protein. Loss of cholesteryl ester hydroperoxides is greatly facilitated by the presence of redox active components, such as TOH in LDL and thiols, apparently of high molecular weight nature, in the medium. Independent of its precise mechanism, cell-mediated detoxification of hydroperoxides of cholesteryl ester and possibly other lipids in LDL presents a possible antioxidant step by removing reactive peroxide groups and eliminating potentially toxic hydroperoxides. As oxidation of LDL has been implicated in the development of atherosclerosis and cholesteryl ester hydroperoxides and hydroxides have been identified in atheroma and complicated lesions (31,51,52), the ability of macrophages to detoxify these molecules could be considered as potentially antiatherogenic.