Minimally oxidized LDL inhibits macrophage selective cholesteryl ester uptake and native LDL-induced foam cell formation.

Scavenger receptor-mediated uptake of oxidized LDL (oxLDL) is thought to be the major mechanism of foam cell generation in atherosclerotic lesions. Recent data has indicated that native LDL is also capable of contributing to foam cell formation via low-affinity receptor-independent LDL particle pinocytosis and selective cholesteryl ester (CE) uptake. In the current investigation, Cu2+-induced LDL oxidation was found to inhibit macrophage selective CE uptake. Impairment of selective CE uptake was significant with LDL oxidized for as little as 30 min and correlated with oxidative fragmentation of apoB. In contrast, LDL aggregation, LDL CE oxidation, and the enhancement of scavenger receptor-mediated LDL particle uptake required at least 3 h of oxidation. Selective CE uptake did not require expression of the LDL receptor (LDL-R) and was inhibited similarly by LDL oxidation in LDL-R−/− versus WT macrophages. Inhibition of selective uptake was also observed when cells were pretreated or cotreated with minimally oxidized LDL, indicating a direct inhibitory effect of this oxLDL on macrophages. Consistent with the effect on LDL CE uptake, minimal LDL oxidation almost completely prevented LDL-induced foam cell formation. These data demonstrate a novel inhibitory effect of mildly oxidized LDL that may reduce foam cell formation in atherosclerosis.

the Institutional Animal Care and Use Committee. Animals were maintained in a pathogen-free facility with a 12 h light/dark cycle and free access to food and water. Lipoprotein isolation, characterization, radiolabeling, and modifi cation LDL ( = 1.019-1.063 g/ml) was isolated from human plasma by density gradient ultracentrifugation ( 19 ). Protein concentrations (mg/ml) were determined by the method of Lowry et al. ( 20 ), and TC, FC, and CE content (nmol/mg LDL protein) using cholesterol E-kits (Wako). LDL was analyzed by SDS-PAGE and nondenaturing gradient gel electrophoresis. LDL was 125 iodinated by the iodine monochloride method ( 21 ). To measure LDL CE uptake, [ 3 H]CE was incorporated into 125 I-LDL using the CETP method as previously described ( 22 ). After labeling, 125 I-and 3 H-specifi c activities were calculated by dividing the radioactivity (dpm/µL) by the LDL CE content (nmol CE/µL) and was between 3,000-8,000 and 100-1,000 dpm/nmol LDL CE for 125 I and 3 H, respectively. LDL (1 mg/ml) was oxidized with 20 µM CuSO 4 at 37°C for 0.5-24 h as described ( 23 ) except that chloramphenicol and Chelex-100 resin were not included in the preoxidation dialysis buffer. Oxidation was terminated by the addition of EDTA (10 mM fi nal concentration) followed by dialysis into 0.3 mM EDTA/saline. oxLDL was analyzed by agarose gel electrophoresis (1% gel) to determine relative electrophoretic mobility (REM) and LDL aggregation. LDL aggregates were defi ned as the proportion of 125 I in each lane that remained trapped within in the well. SDS-PAGE was used to analyze apoB fragmentation, quantifi ed as the percent of 125 I in each lane of the developed gel present as intact apoB. oxLDL lipids were analyzed by TLC with 3

Bone marrow-derived macrophage isolation and culture
Murine bone marrow-derived macrophages (BMMs) were obtained from C57/BL6 mice (WT) or from LDL-R-defi cient (LDL-R Ϫ / Ϫ , purchased from Jackson Laboratories) mice and cultured by standard procedures ( 24 ). Isolated cells were suspended in medium A [RPMI 1640 (Gibco, 11875) plus 15% (v/v) L-cellconditioned medium (LCM), 10% (v/v) heat-inactivated FBS, 50 IU/ml penicillin G, and 50 µg/ml streptomycin] and plated in 24-well plates (with or without glass cover slips for microscopy and cholesterol content determination, respectively) or in 48well plates (for radiolabeled LDL uptake experiments). BMMs atherosclerosis ( 12 ) as well as disappointing results with antioxidants in human clinical trials have generated much controversy about the role of oxidized LDL in the disease ( 9 ).
More recently, native LDL has been implicated in foam cell formation ( 13 ). Many early studies demonstrated that native LDL induces little cholesterol accumulation in cultured macrophages compared with acetylated LDL ( 6 ) or oxLDL ( 14 ). However, these investigations were carried out at relatively low LDL concentrations (5-100 µg/ml LDL). At high concentrations ( ‫ف‬ 1000 µg/ml LDL), nonsaturable mechanisms including fl uid-phase pinocytosis contribute signifi cantly to native LDL uptake and result in foam cell formation ( 13 ). These concentrations are comparable to levels found in the serum and atherosclerotic lesions of patients with cardiovascular disease (often >200 mg/dl of LDL-cholesterol) ( 15 ).
In a recent study, we characterized the pathway of cholesterol delivery to macrophages during LDL-induced foam cell formation ( 16 ). Surprisingly, the majority of the cholesterol was acquired via selective uptake of cholesteryl ester (CE), with smaller amounts acquired by pinocytic uptake of whole LDL particles. Selective CE uptake has been previously reported in other cells and tissues and is the major mode of uptake of cholesterol from HDL as mediated by scavenger receptor class B, type I (SR-BI ) ( 17 ). The mechanism of SR-BI-mediated selective CE uptake is poorly understood but is thought to involve transfer of CE from lipoprotein particles to the cell plasma membrane followed by internalization and hydrolysis by neutral cholesteryl esterase(s) ( 18 ). In macrophages, however, selective CE uptake is independent of SR-BI or the related scavenger receptor CD36 and LDL-derived CE acquired by this pathway appears to be hydrolyzed in lysosomes, indicating a pathway that is distinct from the established SR-BI-mediated pathway ( 16 ). The linear relationship of selective CE uptake to LDL concentration does not support the involvement of any high-affi nity receptor binding in the selective uptake pathway and further studies are needed to defi ne the mechanism of macrophage selective CE uptake.
One question that arose during the initial characterization of the macrophage selective CE uptake pathway was whether LDL modifi cation would alter the rate of selective CE uptake. LDL acetylation resulted in increased LDL particle uptake and unchanged selective CE uptake, while LDL aggregation resulted in increased LDL particle uptake and completely eliminated selective CE uptake ( 16 ). Altered selective CE uptake may impact the processing of LDL-cholesterol and the development of foam cells within atherosclerotic lesions. Therefore, the current study was undertaken to investigate the effect of LDL oxidation, a physiologically relevant LDL modifi cation, on macrophage selective CE uptake.

Animals and reagents
C57/BL6 mice were housed in the Veterans Affairs Medical Center (Lexington, KY) and all experiments were approved by of oxidation. Oxidized LDL exhibited saturable uptake with respect to LDL concentration ( Fig. 2A ), consistent with the known involvement of the high-affi nity scavenger receptors SRA and CD36 ( 7 ). Total [ 3 H]CE uptake did not follow the same pattern ( Fig. 1A ,  I-LDL uptake (( 3 H)-( 125 I)) is defi ned as selective CE uptake, and indicates CE acquired by cells independent of LDL particle uptake. Selective CE uptake was progressively impaired by LDL oxidation and was reduced by >85% in 24 hr oxLDL ( Fig. 1A , gray circles). As we previously reported ( 16 ) and in contrast to 125 I-oxLDL uptake, selective CE uptake was nonsaturable with respect to LDL concentration ( Fig. 2B ). This is the fi rst report of reduced selective CE uptake from oxidized LDL. To investigate the mechanism by which this occurs, two hypotheses were considered: 1 ) LDL oxidation alters the chemical composition and/or structure of LDL particles, thereby preventing cell-surface binding or selective CE transfer, and 2 ) reactive products generated during oxidation inhibit the cell-mediated selective uptake pathway.
First, the relationship between LDL oxidation and selective CE uptake was examined by correlating reductions in selective CE uptake to several established modifi cations that occur during LDL oxidation: increased relative electrophoretic mobility (REM), LDL aggregation, apoB fragmentation, and LDL CE oxidation ( CE oxidation was observed after a lag phase of 2 h and reached >85% by 24 h. Taken together, the inhibition of selective uptake during LDL oxidation correlated well with apoB fragmentation and to a lesser degree with increased electrophoretic mobility but did not correlate with CE oxidation or LDL aggregation.
While the extent of CE oxidation did not correlate with reduced selective uptake, this modifi cation is especially relevant to CE uptake because the CE oxidation products may themselves have altered selective uptake capacity were cultured in medium A for 7 days before being used for experiments, with culture media changed on days 3, 5, 6, and 7.

Determination of CE and FC uptake
LDL CE uptake was determined by treating BMMs with radiolabeled LDL or oxidized LDL, as indicated in the fi gure legends, in medium B [RPMI 1640 plus 15% (v/v) LCM, 0.5% (w/v) fatty acid-free BSA (Sigma), 10 mM HEPES, pH 7.4, 50 IU/ml penicillin G, and 50 µg/ml streptomycin]. After the treatment period, 125 I-LDL uptake, [ 3 H]CE uptake and selective [ 3 H]CE uptake were quantifi ed as previously described ( 16 ). Briefl y, cells were lysed by the addition of 0.1 M NaOH, collected, and counted for cell-associated 125 I-LDL. Cell lysates were also used to determine cell protein content (mg/well) by BCA. Culture medium was precipitated with trichloroacetic acid, extracted with chloroform, and the aqueous phase counted to determine degraded 125 I-LDL. Lipid was extracted from cell lysates using Dole reagent (isopropanolheptane-1M sulfuric acid, 40:10:1) and counted for 3 H by scintillation counting. Cellular 3 H (dpm/mg cell protein) was used to quantify CE uptake (nmol CE/mg cell protein). 125 I uptake (dpm/mg cell protein, the sum of cell-associated and degraded 125 I) was used to calculate CE equivalents (nmol/mg cell protein) assuming whole-particle uptake with concomitant CE delivery, as previously validated ( 25 ). Selective CE uptake was calculated as the difference between total [ 3 H]CE uptake and the CE uptake equivalents contributed by the uptake of whole-particle 125 I-LDL and represents CE uptake that cannot be accounted for by wholeparticle uptake.

Cellular FC and CE content determinations and microscopy
For cellular lipid determination, lipid was extracted from cells with hexane-isopropanol (3:2), dried, and resuspended in aqueous solution with 1% Triton X-100-H 2 O. TC and FC content (nmol cholesterol/mg cell protein) were determined using cholesterol E-kits. CE content was calculated as the difference between TC and FC content. Cell protein was determined using a BCA protein assay kit (Thermo Scientifi c). For microscopy, BMMs were treated as indicated in the fi gure legends then fi xed for 5 min in freshly prepared 1% paraformaldehyde and stained with Oil Red O. Images were obtained from at least fi ve randomly chosen fi elds for each condition.

Statistics and data analysis
Results are expressed as the mean ± SEM as indicated in the fi gure legends. Where not visible, error bars are contained within the symbols for results with n у 2. Statistical analysis was carried out using unpaired, two-tailed t -tests. Best-fi t curves were generated using GraphPad Prism4 software.

RESULTS
We previously demonstrated that selective CE uptake is a major pathway of cholesterol delivery to macrophages from native LDL ( 16 ). However, the effect of LDL oxidation on this selective CE uptake pathway is not known. In the current investigation, [ ( Fig. 3A ), consistent with previous reports ( 23 ). However, selective CE uptake was impaired with each of these labeled ligands, regardless of the extent of CE tracer oxidation ( Fig. 3B ). In addition, when a [ 3 H]CEt tracer was incorporated after LDL oxidation, to ensure that none of the tracer was oxidized (confi rmed by TLC before and after the experiment), selective CE uptake was reduced similarly compared with the previous experiments (supplementary Fig. II). These results provide strong evidence that LDL oxidation reduces selective uptake of intact CE and that this effect (hypothesis 1 above). However, in the experiments described above ( Figs. 1 and 2 ), the rate of uptake of intact CE versus oxidized CE could not be determined because the uptake assay relied on quantifi cation of total cellular 3 H that would represent both [ 3 H]CL and oxidized derivatives. Therefore, the effect of CE oxidation on selective CE uptake was further studied by incorporating three different radiolabeled CEs into parallel LDL preparations:   CE uptake was found to be dose-dependent and was observed at concentrations of 2 hr oxLDL as low as 25 µg/ml (supplementary Fig. III). While it was possible that the labeled [ 3 H]CL/ 125 I-LDL could have been oxidized by exposure to oxLDL and cells during the treatment period, this is thought to be unlikely because of the high concentration of antioxidants and absence of Cu 2+ in RPMI medium. This was confi rmed by collecting the treatment medium from cells at the end of the experiment and quantifying the extent of [ 3 H]CL/ 125 I-LDL oxidation as described for Fig. 1 . There was no signifi cant [ 3 H]CL oxidation by TLC (data not shown); however, there was a small but significant amount of apoB fragmentation as determined by SDS-PAGE ( Fig. 5B ). Media containing 4 hr oxLDL or 24 hr oxLDL were associated with a ‫ف‬ 10% and 18.5% reduction, respectively, in intact apoB of the coincubated [ 3 H] CL/ 125 I-LDL. However, this corresponds to a lesser degree of LDL oxidation than that observed with LDL oxidized by Cu 2+ for 30 min ( Fig. 1C ), and therefore is not suffi cient to explain the reduction in selective uptake observed ( Fig. 5A ). These data suggest that oxLDL inhibits selective CE uptake from coincubated native LDL, possibly via cellmediated effects (hypothesis 2 above).
The contribution of a direct inhibitory effect of oxLDL on cells was further investigated by determining the initial rates of LDL uptake and selective uptake in cells pretreated with or without oxLDL ( I-oxLDL. Compared with LDL, pretreatment with oxLDL had a modest inhibitory effect on subsequent 125 I-oxLDL uptake but no effect on 125 I-LDL uptake ( Fig. 6A ). In contrast, pretreatment with oxLDL is not the result of reduced selective uptake of oxidized CE derivatives.
In contrast to the relationship to CE oxidation, reduced selective uptake did correlate well with apoB fragmentation ( Fig. 1 ). Chemical or oxidative apoB modifi cation is known to alter binding of LDL to the LDL receptor (LDL-R) ( 27,28 ). Therefore, the possible contribution of LDL-R to selective CE uptake from LDL and oxLDL was investigated using BMMs from LDL-R Ϫ / Ϫ mice ( Fig. 4 ). 125 I-LDL uptake was signifi cantly lower in LDL-R Ϫ / Ϫ BMMs compared with WT, with uptake of LDL, 1 hr-oxLDL, and 4 h-oxLDL reduced by 48%, 34%, and 24%, respectively ( Fig. 4A ). In contrast, selective CE uptake from LDL was at most 14% lower in LDL-R Ϫ / Ϫ BMMs ( Fig. 4B ). LDL oxidation impaired selective CE uptake similarly in LDL-R Ϫ / Ϫ compared with WT BMMs. Therefore, LDL-R likely does not contribute to selective CE uptake and reduced binding to this receptor does not explain loss of selective uptake with LDL oxidation.
As stated in hypothesis ( 2 ) above, it is possible that bioactive products generated during LDL oxidation inhibit the cellular pathway of selective uptake. In fact, oxidized lipids including oxidized CE are known to exert many activities in macrophages including induction of pinocytosis ( 29 ). To test this, BMMs were incubated simultaneously with [ 3 H]CL/ 125 I-LDL and unlabeled oxLDL to determine if the presence of oxLDL would inhibit uptake of the coincubated unoxidized [ 3 H]CL/ 125 I-LDL ( Fig. 5 ). Interestingly, oxLDL inhibited selective CE uptake by as much as 94%, depending on the extent of oxidation of the oxLDL used in coincubation ( Fig. 5A ). In contrast, only 24 hr ox-LDL caused a signifi cant reduction in 125 I-LDL uptake and caused only ‫ف‬ 50% inhibition ( Fig. 5A ). In an independent experiment, the inhibitory effect of oxLDL on selective   oxLDL compared with native LDL is consistent with previous work demonstrating lysosomal trapping and reduced effl ux of CE derived from extensively oxidized LDL ( 30,31 ). In contrast, CE derived from selective CE uptake is effi ciently hydrolyzed and subject to effl ux from macrophages ( 16 ).
LDL-induced foam cell generation was assessed by Oil Red O staining of neutral lipid droplets ( Fig. 8 ). Cells treated for 24 h without lipoproteins did not contain Oil Red O-positive lipid droplets (data not shown). 8 hr ox-LDL or 24 hr oxLDL induced foam cell generation at either 100 µg/ml or 500 µg/ml. In contrast, native LDL treatment resulted in visible neutral lipid droplet deposition only at the higher concentration (500 µg/ml), consistent with our previous observation of lipid loading by native LDL through a low-affi nity and nonsaturable process ( 16 ). Interestingly, 1.5 h oxLDL induced much less lipid deposition compared with native LDL. The morphology of the lipid-loaded cells was examined at higher magnifi cation and revealed distinct differences in the type and distribution of the Oil Red O-stained droplets ( Fig. 8B ). The lipid droplets formed during treatment with native LDL were less numerous but much larger than those formed during treatment with oxidized LDL and appeared localized toward the cell periphery. In contrast, the droplets induced by oxLDL were widely spread throughout the resulted in pronounced inhibition of selective CE uptake from both LDL as well as oxLDL ( Fig. 6B ). These results indicate that oxLDL exerts a strong effect on selective CE uptake by modulating the macrophage pathway responsible for such uptake.
The effect of LDL oxidation on selective uptake would be predicted to infl uence cholesterol accumulation and foam cell formation. However, the cholesterol content of cells is infl uenced by cholesterol effl ux during treatments and does not always relate quantitatively to measured CE uptake ( 16 ). Therefore, net cholesterol accumulation was examined by measuring cellular cholesterol content after a 24 h loading period with native or oxLDL ( Fig. 7 ). BMMs treated without lipoproteins contained ‫ف‬ 117 nmol free cholesterol (FC) per mg cell protein and did not contain a signifi cant amount of CE (calculated as the difference between total cholesterol and FC content). LDL treatment resulted in signifi cant concentration-dependent cholesterol accumulation, with FC and CE increasing 1.7 and 4.7-fold, respectively, in cells treated with 500 µg/ml LDL. In contrast, 1.5 hr oxLDL induced only minimal changes in FC (<30%) and did not signifi cantly alter CE content. As expected, 8 h oxLDL or 24 hr oxLDL treatment resulted in pronounced accumulation of both FC and CE, which increased 5.8-and 26.7-fold, respectively. The greater degree of cholesterol accumulation induced by    cytoplasm. Therefore, cells loaded with native LDL were easily distinguished from those treated with oxLDL.

DISCUSSION
We recently demonstrated that high-capacity selective CE uptake is a major pathway of cholesterol delivery to macrophages during native LDL-induced foam cell formation ( 16 ). The current results extend the previous study by showing that LDL oxidation results in progressive inhibition of this pathway, in contrast to the well-known enhancement of LDL particle uptake that occurs with LDL oxidation due to recognition by macrophage scavenger receptors. The mechanism is unknown, but our studies suggest that this occurs via inhibition of macrophage uptake activity, resulting in reduced selective uptake from both oxLDL as well as native LDL. The impact of altered CE uptake at each stage of LDL oxidation was refl ected by signifi cant differences in macrophage foam cell formation. Native LDL cholesterol was acquired largely via lowaffi nity or nonsaturable selective CE uptake, leading to the formation of foam cells containing large lipid droplets. Minimally oxidized LDL was associated with reduced selective CE uptake and induced signifi cantly less lipid accumulation. Extensively oxidized LDL was taken up with high affi nity as whole particles and resulted in profuse deposition of small lipid droplets, similar to previous reports of scavenger receptor-mediated uptake ( 32 ).
LDL oxidation involves a complex series of chemical and physical changes; therefore, in this study, special attention was given to the degree of oxidation and its relationship to the CE uptake. A signifi cant reduction in selective CE uptake was observed with as little as 30 min of oxidation while altered macrophage LDL particle uptake, increased LDL electrical charge, LDL aggregation, and LDL CE oxidation required at least 3 h of oxidation. These results indicate that the impairment of selective uptake is not associated with the modifi cations that promote scavenger receptor-mediated uptake of extensively oxidized LDL. Instead, the impairment of selective CE uptake correlated closely with apoB fragmentation, a well-known feature of apoB that occurs early in the oxidation process ( 33 ). Oxidative peptide fragmentation is not fully understood; however, it is thought to involve direct scission of peptide bonds by free radical-containing phospholipid oxidation products. Phospholipid and protein forms the shell through which CE must traverse in order to be transferred to cells by selective uptake, so it is conceivable that derangement of this surface layer in minimally oxidized LDL could impact the rate of selective CE uptake.
We also considered the possibility that oxidation or fragmentation of apoB may alter binding of LDL to LDL-R or another cell-surface binding site. The current investigation provides evidence that LDL-R is not involved in macrophage selective CE uptake from LDL. Interestingly, oxLDL and oxidized phospholipids were previously shown to inhibit selective uptake of HDL CE in SR-BI expressing cells, with a variable effect on HDL binding depending on cell type ( 34,35 ). The current study is consistent with a similar effect of oxidized phospholipids or other bioactive lipids in macrophages. However, we recently demonstrated that macrophage selective CE uptake from LDL is not dependent on SR-BI ( 16 ). It is possible that the SR-BIindependent selective uptake pathway in macrophages shares a similar oxLDL-sensitive step in common with the SR-BI-mediated pathway in other cell types. This is the fi rst report of an oxLDL (minimally oxidized LDL) inducing less macrophage lipid accumulation than native LDL, a fi nding that was apparent only at a high LDL concentration and was likely overlooked in previous investigations due to the lower LDL concentrations used. Several conclusions may be inferred from this observation. First, this provides additional evidence that oxidation is not involved in native LDL-induced lipid loading, because mild oxidation of LDL that may occur during LDL isolation and during treatment of cells would be expected to decrease the observed loading. Second, while lipid loading was almost completely eliminated by minimal oxidation of LDL, LDL particle uptake including pinocytosis was not decreased, suggesting an essential role for selective CE uptake in LDL-induced foam cell formation. Previous studies established that macrophage LDL pinocytosis is required for foam cell formation ( 13 ) but did not rule out the possibility that another pathway also contributed to cholesterol uptake. In fact, selective CE uptake was reduced by minimal LDL oxidation, suggesting that the selective uptake pathway is also required. This hypothesis is supported by the large proportion of cholesterol supplied by the selective uptake pathway and by previous studies of the metabolism of CE following selective uptake ( 16 ).
Finally, reduced foam cell formation with minimally oxidized LDL suggests that LDL oxidation that occurs in atherosclerotic lesions may reduce cellular lipid uptake and accumulation in some circumstances in which the extent of oxidation is limited. We propose that this effect is relevant to atherosclerosis because analysis of oxidized lipids in human lesions suggests that LDL is minimally oxidized (<2.5% and <1% of cholesteryl linoleate and cholesterol oxidized, respectively) ( 36 ). As stated above, in the current study, a signifi cant inhibitory effect of oxLDL on macrophages was observed that persisted even after the oxLDL was removed. This result is important because it indicates that the presence of oxLDL in lesions and other tissue sites may infl uence not only its own metabolism but also that of native LDL. It is not known whether this effect is mediated by intact oxLDL particles or by one or more chemical constituents such as oxidized lipid products. The isolation and identifi cation of such products would be of great interest because they may serve as pharmacologic agents for reducing foam cell formation in the treatment of atherosclerosis.
In summary, this study demonstrates that oxLDL inhibits selective LDL CE uptake, resulting in reduced foam cell generation in macrophages treated with minimally oxidized LDL. Further studies are needed to elucidate the mechanism of this phenomenon and to determine the signifi cance of these pathways in atherosclerosis.