Minimally Oxidized Low-Density Lipoprotein Increases Expression of Scavenger Receptor A, CD36, and Macrosialin in Resident Mouse Peritoneal Macrophages

Substantial evidence demonstrates that OxLDL can contribute to atherogenesis.¹²³⁴ Many studies indicate that macrophages in the artery wall take up OxLDL via scavenger receptors, leading to cholesterol ester accumulation and resulting in the formation of foam cells, the hallmark of the arterial fatty streak, which is recognized as the earliest atherosclerotic lesion.¹²³⁴

There is a broad and continuous spectrum of oxidized LDLs from minimally oxidized LDL (MM-LDL) to fully or highly oxidized LDL (OxLDL).⁵ These forms differ not only structurally but also functionally. MM-LDL is defined in part by the fact that it is still a ligand for the native LDL receptor, whereas OxLDL is not; MM-LDL is not a ligand for scavenger receptors, whereas OxLDL is.⁶ The extent of oxidation of MM-LDL, measured in terms of TBARS, is generally less than 10 nmol/mg LDL protein; for OxLDL, the value is usually over 30 or 40. The biological properties of MM-LDL can be quite different from those of OxLDL.⁶ For example, MM-LDL can induce expression of macrophage colony–stimulating factor,⁷ monocyte chemoattractant protein 1,⁸ and tissue factor⁹ by endothelial cells, but OxLDL cannot. Conversely, the surface expression of E-selectin, vascular cell adhesion molecule-1, and intercellular adhesion molecule 1 is not induced by treatment of endothelial cells with MM-LDL. Lysophosphatidylcholine, which is a major component of OxLDL, can induce vascular cell adhesion molecule-1 in endothelial cells.¹⁰ Moreover, the lipids in OxLDL are cytotoxic, whereas the lipids in MM-LDL are not.

It is now clear that the normal cells of the artery wall—endothelial cells and smooth muscle cells—have the ability to oxidatively modify LDL.¹³ Thus, MM-LDL may be generated before circulating monocytes have been recruited into the subendothelial space. Endothelial cells stimulated by MM-LDL could then produce adhesion molecules, chemokines, and cytokines that recruit monocytes and convert them to macrophages, accelerating lesion progression. Subsequently, MM-LDL is further oxidized, acquiring a wide array of biological properties, including the ability to bind to scavenger receptors.¹²³

Scavenger receptors are characterized as multiligand receptors, found mostly on macrophages, and play a variety of roles in host defenses.¹¹¹² Scavenger receptors that recognize OxLDL include SRA,¹³ CD36,¹⁴¹⁵ macrosialin/CD68,¹⁶¹⁷¹⁸ SR-B1,¹⁹²⁰ and its human homologue.²¹ The present studies were undertaken to test whether oxidized forms of LDL—MM-LDL or OxLDL—affect expression of their own receptors.

Methods

Lipoproteins

Native human LDL (density 1.019 to 1.063) was isolated from EDTA-plasma by sequential ultracentrifugation²² and stored at 4°C in 0.3 mmol/L EDTA. LPDS was isolated as the fraction with density >1.21. LDL was radioiodinated (specific activity, 150 to 300 cpm/ng LDL protein) using carrier-free Na ¹²⁵I and Iodogen.²³

Before oxidation, both ¹²⁵I-labeled and unlabeled native LDL were extensively dialyzed against EDTA-free PBS at 4°C in the dark to remove EDTA. OxLDL was prepared by incubating LDL (100 μg/mL) in Ham’s F-10 medium with 5 μmol/L copper for 24 hours at 37°C. MM-LDL was prepared by incubating with 2 μmol/L copper in PBS for 4 hours at 37°C. EDTA (100 μmol/L) was added immediately and the preparation was stored at 4°C under N₂ in the dark. Acetylation of LDL was carried out as described.²⁴

Lipid Extraction and Albumin Treatment of MM-LDL

One milliliter of DMEM containing 500 μg of MM-LDL was extracted by adding 2 mL of ice-cold methanol followed by 2 mL of chloroform.²⁵ The mixture was vortexed, put on ice for 15 minutes, and centrifuged at 1800g for 10 minutes at 4°C. The liquid supernatant (both phases) was transferred into new tubes, leaving behind the protein pellet. Chloroform (1 mL) and ice-cold distilled water (1 mL) were added, and the tubes were centrifuged at 1800g for 5 minutes at 4°C. The bottom layer was aspirated, dried under N₂, and redissolved in ethanol. In some experiments, MM-LDL was incubated for 24 hours with 10 mg/mL fatty acid–free BSA to reduce its content of polar oxidation products and then reisolated at d<1.21.

Metabolism of ¹²⁵I-OxLDL and ¹²⁵I-AcLDL

Resident macrophages were harvested by peritoneal lavage of female Swiss-Webster mice (2 to 3 months of age, weight 25 to 35 g) with ice-cold PBS. Cells were plated in 24-well plates at a cell density of 1×10⁶ cells per well in RPMI 1640 medium containing 10% heat-inactivated fetal calf serum. After 6 hours, nonadherent cells were removed by washing three times with RPMI medium without fetal calf serum. Great care was taken not to contaminate the incubation medium with lipopolysaccharide, and its concentration never exceeded 0.05 endotoxin units/mL.

Macrophages were incubated in DMEM, containing 2% LPDS with or without MM-LDL or OxLDL at concentrations of 25 or 50 μg/mL, for 24 to 48 hours. The macrophages were then washed three times with DMEM and then incubated in DMEM with 5 μg/mL ¹²⁵I-OxLDL or ¹²⁵I-AcLDL for 5 hours at 37°C. The media were harvested and assayed for determination of trichloroacetic acid–soluble noniodide degradation products.²⁶ Appropriate control incubations were conducted in wells with no cells, and the no-cell control values were subtracted from experimental values. The cells were washed three times with PBS, dissolved in 0.2N NaOH, and assayed for determination of cell protein and cell-associated ¹²⁵I radioactivity. Nonspecific cell association and degradation were defined as the values measured in the presence of 200 μg/mL (40-fold excess) unlabeled ligand.

Binding of ¹²⁵I-OxLDL and ¹²⁵I-AcLDL was determined by incubating macrophages with the radiolabeled ligands at concentrations ranging from 0 to 30 μg/mL in DMEM at 4°C for 2 hours. Cells were then washed once with ice-cold PBS containing 0.2% BSA and incubated in that medium for another 15 minutes. After three additional washes with BSA-free PBS, cells were solubilized with 0.2N NaOH, and cell protein and ¹²⁵I radioactivity were determined. Nonspecific binding was defined as the binding of the radiolabeled ligand in the presence of a 10-fold or 20-fold excess of unlabeled OxLDL or AcLDL, respectively.

RT-PCR Amplification

Total RNA was isolated from 5×10⁷ resident mouse peritoneal macrophages using the acid guanidinium thiocyanate–phenol–chloroform extraction procedure according to Chomczynski and Sacchi.²⁷ RNA was reverse transcribed using Superscript II (Life Technologies, Inc). The transcribed cDNA was then used for PCR amplification to estimate the expression of SRA, CD36, macrosialin, and GAPDH. Two specific primers matching the published sequences^14282930 were used to identify and amplify SRA (5′-ATGACAGAGAATCAGAGG-3′), (5′-CCCTCTGTCTCCCTTTTC-3′); CD36 (5′-CAGCCCAATGGAGCCATC-3′), (5′-CAGCGTAGATAGACCTGC-3′); macrosialin (5′-ATGCGGCTCCCTGTGTGTC-3′), (5′-TCAGAGGGGCTGGTAGGTTG-3′); and GAPDH (5′-TGCCATTTGCAGTGGCAAAGTGG-3′), (5′-TTGTCATGGATGACCTTGGCCAGG-3′). The amplified transcripts were analyzed by gel electrophoresis, and the signal intensity of the bands with the expected sizes (855 bp for SRA, 487 bp for CD36, 981 bp for macrosialin, and 439 bp for GAPDH) were quantified by densitometric scanning and compared with the internal standard of GAPDH, amplified under identical conditions.

Western Blot Analysis

Guinea pig antiserum against the C-terminal cytoplasmic tail of macrosialin (DS4) and against the internal portion of mouse CD36 was prepared as described.¹⁶³¹ Guinea pig antiserum against mouse SRA was a kind gift from Christopher K. Glass, Cellular and Molecular Medicine, University of California San Diego. Western blots were scanned densitometrically. Over the range of 0 to 30 μg of macrophage protein loaded (SRA), there was a near linear relation between optical density and amount of protein.

Macrophages preincubated for 48 hours with MM-LDL or OxLDL at 50 μg/mL or with medium alone (DMEM/2% LPDS) were harvested in ice-cold extraction buffer (50 mmol/L Tris-HCl, 150 mmol/L NaCl, 50 μmol/L PMSF, 5 mmol/L benzamidine, 50 U/mL aprotinin, 14.5 μmol/L pepstatin A, 0.1 mmol/L leupeptin, 1.9 μmol/L d-phenylalanyl-l-prolyl-l-arginine chloromethyl ketone), and the cells were washed twice with extraction buffer. The cells were resuspended in extraction buffer containing 40 mmol/L octylglucoside, mixed gently by pipetting and vortexing, and further incubated on ice for 2 hours. The insoluble material was removed by centrifugation at 11 200g for 15 minutes at 4°C, and the protein content in the supernatant was estimated using a BCA protein assay kit (Pierce).

Detergent extracts (20 μg protein) were electrophoresed on 8% SDS-polyacrylamide gels, and the proteins were electrotransferred to nitrocellulose membranes. Western blot analysis was carried out by first blocking the nitrocellulose strips with Super Block (Pierce) for 1.5 hours. The blots were rinsed with PBS/0.1% Tween 20, incubated with 1:100 dilution of primary antibody in PBS/0.1% Tween 20 for 3 hours, and then washed three times with the same buffer. The bound antibodies were detected by using ¹²⁵I-labeled goat anti–guinea pig IgG. The nitrocellulose strips were exposed to Kodak X-Omat MR film with an intensifier for 1 day at −70°C. Autoradiograms were assessed by densitometric scanning using a flatbed scanner attached to an IBM computer running ImageQuant (Molecular Dynamics).

Other Analytical Procedures

Cell protein and LDL protein were measured by the method of Lowry et al³² using BSA as a standard. Measurement of LDL oxidation was determined by fluorometric TBARS assay³³ and by electrophoretic mobility on agarose gel.³⁴ LDL samples were applied at 1 μg per lane on agarose gel and electrophoresed at 100 V for 30 minutes. After drying, the gels were stained with Oil Red O to visualize lipoprotein bands. Reactive amino groups in protein were estimated with TNBS using valine as the standard.³⁵

Statistics

Data on cell association and degradation assay were expressed as mean±SD. Results were analyzed by ANOVA with Fisher’s PLSD test. A value of P<.05 was considered significant.

Results

Evaluation of the Extent of Oxidation of MM-LDL and OxLDL

Table 1 shows values for TBARS, reactive lysine ε-amino groups (TNBS) and relative electrophoretic mobility for MM-LDL and OxLDL. As expected, the degree of oxidation-induced changes was much less for MM-LDL than for OxLDL. MM-LDL was still recognized by the receptor for native LDL. In direct ligand-binding experiments, the specific cell association of native ¹²⁵I-LDL with fibroblasts was completely inhibited by either unlabeled MM-LDL or native LDL, and the concentration-dependency curves were indistinguishable (Fig 1A). In contrast, the binding of ¹²⁵I-OxLDL to macrophages was unaffected by the presence of unlabeled MM-LDL or unlabeled native LDL, indicating that MM-LDL was not a ligand for the scavenger receptors (Fig 1B). Similar results were obtained using ¹²⁵I-AcLDL as ligand. The extent of oxidation of MM-LDL induced by brief incubation with copper, as used in the present study, was similar to that of MM-LDL as studied by other investigators at the University of California, Los Angeles.^47893637

Effect of Preincubation With Modified Lipoproteins on Scavenger Receptor Function

As shown in Table 2, preincubation with either MM-LDL or OxLDL for 24 to 48 hours increased cell association of ^l25I-OxLDL and ¹²⁵I-AcLDL by mouse peritoneal resident macrophages. However, the effect of MM-LDL was clearly greater than that of OxLDL. After a 48-hour preincubation with 50 μg/mL MM-LDL, cell association of the ligands was almost twice the control values. OxLDL under the same conditions increased cell association by only 50%. For both MM-LDL and OxLDL, the effects were greater at 50 μg/mL than at 25 μg/mL and greater at 48 hours than at 24 hours. Most of the subsequent studies were then done using a preincubation of 48 hours with 50 μg/mL OxLDL or MM-LDL.

Fig 2 compares the effects of MM-LDL and OxLDL preincubation using as references native LDL and AcLDL. MM-LDL at 50 μg/mL during the preincubation more than doubled the rate of ¹²⁵I-OxLDL and ¹²⁵I-acetyl LDL degradation during the subsequent incubations. Again, the effects of MM-LDL were greater than those of OxLDL both on cell association and degradation of either ligand, ¹²⁵I-OxLDL or ¹²⁵I-AcLDL. Neither native LDL nor AcLDL had any significant effects.

The Role of Oxidized Lipids in the Induction of Scavenger Receptor Activity

Because many or most of the biological effects of MM-LDL appear to be attributable to oxidized lipid components, we evaluated the role of the oxidized lipids in MM-LDL in the induction of scavenger receptor activity. As shown in Table 3, the lipids extracted from MM-LDL and added in ethanol (at final concentrations corresponding to those of intact MM-LDL studied in the same experiments) during the 48-hour preincubation significantly increased cell association and degradation of OxLDL and AcLDL. However, the effects were significantly smaller than that of the intact MM-LDL. For example, at 50 μg/mL, the lipid extracts increased OxLDL cell association by 34%, while the intact MM-LDL increased it by 73%. The corresponding changes in OxLDL degradation were +28% versus +61%.

Albumin is able to remove free fatty acids and certain other polar lipids from LDL. As shown in Table 3, albumin treatment of MM-LDL significantly reduced its effects on metabolism of ¹²⁵I-AcLDL (P<.01). With ¹²⁵I-OxLDL as ligand, the trend was in the same direction, but the effects did not reach statistical significance. These data suggest that at least a portion of the biological effect of MM-LDL is due to its lipid moiety. Because the extracted lipids reintroduced into the aqueous medium may not be as effective as they are when presented within a lipoprotein and because the stability of the extracted lipids is not known, these experiments probably give a minimum estimate of the effectiveness of the lipid moieties.

Characterization of the Specific Binding of ¹²⁵I-OxLDL and ¹²⁵I-AcLDL

The specific binding of OxLDL and AcLDL was measured as a function of concentration both in the untreated macrophages and in macrophages previously incubated for 48 hours in the presence of MM-LDL or OxLDL at 50 μg/mL. As shown in Table 4, both MM-LDL and OxLDL increased the B_max for both ligands without affecting the K_d. The effect of MM-LDL was somewhat greater than that of OxLDL. The results are consistent with an increase in the expression of receptors without a change in affinity.

Receptors Induced by Treatment With MM-LDL or OxLDL

The levels of mRNA for SRA, macrosialin, and CD36 were measured in control macrophages and in macrophages treated with MM-LDL or OxLDL. As shown in Fig 3, treatment with MM-LDL for 48 hours increased mRNA levels of SRA, macrosialin, and CD36 twofold to threefold. The effect of OxLDL was less than that of MM-LDL in the case of SRA, but greater than that of MM-LDL in the case of macrosialin. OxLDL had almost no effect on CD36 mRNA levels.

We then carried out a time course study of the response to MM-LDL, as shown in Fig 4. The increase in mRNA for SRA and for macrosialin was evident at 6 hours and became maximal at 48 hours. In contrast, the increase in CD36 mRNA did not begin until somewhere between 24 and 48 hours and became maximal at 48 hours. Native LDL, used as a control, had almost no effect on mRNA levels for all three receptors.

Fig 5 shows Western blots and their densitometric quantification for these three receptors. The changes in protein paralleled the changes in mRNA; ie, MM-LDL had a greater effect than OxLDL in the case of both SRA and CD36, but OxLDL increased macrosialin protein levels to a greater extent than did MM-LDL.

Discussion

These studies show that the binding and uptake of OxLDL by resident mouse peritoneal macrophages are under regulation by OxLDL itself. The effects of MM-LDL were distinctly larger than the effects of LDL oxidized for 24 hours (OxLDL). Prior incubation with MM-LDL for 48 hours almost doubled the cell association and degradation of both ¹²⁵I-OxLDL and ¹²⁵I-AcLDL, whereas equal concentrations of OxLDL increased metabolism of the substrates by only about 50%.

These effects on metabolism of OxLDL and AcLDL were associated with an increase in the total binding of the substrates, with no significant change in binding affinity. The increase in binding and uptake of modified LDL after 48 hours of incubation with MM-LDL was accompanied by comparable increases in receptor mRNA and receptor protein (SRA, macrosialin, and CD36). The effects of preincubation with OxLDL on mRNA and protein expression varied among the receptors measured. OxLDL had only a small effect on CD36, in contrast to the twofold effect of MM-LDL; OxLDL had a smaller effect also on SRA but a greater effect on the expression of macrosialin mRNA and protein than did MM-LDL.

Recently, Han et al³⁸ reported that both native and modified LDL increase the functional expression of CD36 in J774 cells, a murine macrophage cell line. They observed very rapid induction of CD36 mRNA, a maximum being reached 2 hours after treatment with AcLDL and 4 hours after native LDL or OxLDL. Increased expression of CD36 mRNA persisted for 24 hours with each treatment. These results differ in a number of ways from those reported here for normal resident mouse peritoneal macrophages. We found no effect of native LDL nor AcLDL; the magnitude of the induction was much less; and in the case of CD36, we saw no increase until 24 to 48 hours. Some or all of these differences may reflect altered regulatory patterns in the tumor cell line, but further studies are needed.

The exact mechanisms responsible remain to be established. However, it seems fair to conclude that much or most of the effect is exercised at the transcriptional level. The fact that MM-LDL has a greater effect than does OxLDL suggests that the effects are not due to interaction of the ligands with scavenger receptors. MM-LDL is not a ligand for scavenger receptors; at this early stage of oxidation it is still recognized by the native LDL receptor. A number of the biological effects of MM-LDL can be ascribed to oxidized lipids,⁶ and the signaling mechanism mediated by nuclear factor κB is centrally involved in at least some of them.³⁹ Watson et al⁴⁰ have recently identified two specific oxidized phospholipid molecules that can mimic some of the biological effects of MM-LDL. The present studies show that lipids extracted from MM-LDL and added to the medium in ethanol can partially mimic the effects of the intact MM-LDL. The fact that the effects were only about half those of intact MM-LDL could reflect the inefficient presentation of lipids added in this way; it remains a possibility that all of the effects are actually attributable to oxidized lipids. The facts that preincubation with native LDL had no effect at all on subsequent metabolism of OxLDL and that MM-LDL is a ligand for the native LDL receptor and not for scavenger receptors strengthen the possibility that the effects are all due to oxidized lipid moieties, rather than interactions with known receptors for LDL or modified LDL. Interactions with unidentified receptors are not ruled out.

What are the implications of these findings for the atherogenic process? LDL can be oxidized by the normal cells of the arterial wall, ie, endothelial cells and smooth muscle cells. Once they have been recruited, monocyte/macrophages can also participate in oxidation of LDL. MM-LDL can increase adhesion of monocytes to endothelial cells, partly by stimulating the release of monocyte chemoattractant protein 1 from endothelial cells,⁴ and it can speed the differentiation of the newly recruited monocytes to macrophages by stimulating the release of macrophage colony–stimulating factor.⁷ The present studies show that it can also upregulate the expression of scavenger receptors. Thus, when oxidation of LDL in the intima has proceeded to the point at which it has become a ligand for scavenger receptors, it will find the scavenger receptors already expressed at a higher level and able to take up the OxLDL more avidly. As has been pointed out previously,⁴¹ this has both its good side and bad side. On the good side, it allows for more rapid removal of OxLDL from the intima and therefore some “protection” from the cytotoxic effects and other damaging effects of OxLDL. On the bad side, it increases the rate of formation of foam cells, which may in the long run be even more deleterious. Only recently has a partial answer to this question been published. Suzuki et al⁴² have shown that atherosclerosis in mice targeted for SRA proceeds less rapidly than it does in wild-type mice. In other words, it appears that the presence of SRA contributes to progression of the atherosclerotic lesion.

The expression of scavenger receptors and their functional activity in macrophages are under complex regulation. A number of cytokines and other inflammatory agents have been identified as possible mediators involved in regulating expression of scavenger receptors either directly or indirectly. For example, SRA activity is inhibited by lipopolysaccharide,⁴³ but probably as a result of the increased production of tumor necrosis factor. Conversely, growth and maturation factors leading to macrophage activation appear to directly affect the expression of scavenger receptors.^4445464748

The exact mechanisms resulting in the changes of scavenger receptor expression reported here are not completely defined but are exercised at least in part by the lipid moiety of MM-LDL. Further studies will be needed to resolve this issue and to identify the receptors that play major roles in the metabolism of modified LDL in the developing arterial lesions.

Selected Abbreviations and Acronyms

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes