Macrophage plasma membrane chondroitin sulfate proteoglycan binds oxidized low-density lipoprotein

Atherosclerosis 149 (2000) 5 – 17 www.elsevier.com/locate/atherosclerosis

Macrophage plasma membrane chondroitin sulfate proteoglycan binds oxidized low-density lipoprotein Marielle Kaplan, Michael Aviram * The Lipid Research Laboratory, The Bruce Rappaport Faculty of Medicine, Technion, The Rappaport Family Institute for Research in the Medical Sciences and Rambam Medical Center, Haifa, Israel Received 31 December 1998; received in revised form 8 June 1999; accepted 13 July 1999

Abstract Lipoprotein interactions with macrophage proteoglycans (PGs) is believed to play an important role in the cellular uptake of lipoproteins and in macrophage cholesterol accumulation. Recently, we have shown the participation of macrophage plasma membrane glycosaminoglycans (GAGs) in the cellular uptake of oxidized LDL (Ox-LDL). The aim of the present study was to identify the specific cell surface proteoglycans involved in this interaction. J-774 A.1 macrophage-like cell line plasma membrane proteoglycans were isolated by anion exchange chromatography from cells that were prelabeled with [35S]sodium sulfate. Using Sepharose 6B chromatography, cell surface major proteoglycans were identified as chondroitin sulfate (CS) proteoglycans (77%) and heparan sulfate (HS) proteoglycans (23%). Binding rates of these 35S-labeled proteoglycans to Ox-LDL and to native LDL were analyzed by their ability to bind lipoproteins coupled to a CnBr-activated Sepharose CL-4B chromatography. Of the total labeled cell surface proteoglycans added to the column, 57% were bound to the Sepharose-coupled Ox-LDL, whereas 73% of the cell surface proteoglycans were bound to the Sepharose-coupled native LDL. Binding of the plasma membrane macrophage 35 S-labeled proteoglycans to Ox-LDL was inhibited by adding increasing concentrations of non-labeled chondroitin sulfate, or by pretreatment of the 35S-labeled proteoglycans fraction with chondroitinase ABC. In contrast, neither the addition of non-labeled heparan sulfate, nor pretreatment of the labeled proteoglycans fraction with heparinase III, had any significant effect on proteoglycan binding to Ox-LDL. These findings were further supported by using mutant cells characterized by specific glycosaminoglycan deficiencies. Ox-LDL binding and degradation by mutant 745 CHO cells which are characterized by a deficiency in both heparan sulfate and chondroitin sulfate, was decreased by 28 and 27% respectively, compared to the binding of Ox-LDL to the wild-type CHO cells. Ox-LDL binding and degradation by mutant 677 CHO cells, which lack heparan sulfate but have increased levels of chondroitin sulfate, however, was found to be increased by 29 and 19%, respectively, compared to Ox-LDL binding to the wild-type CHO cells. Finally, analysis of the cell surface proteoglycans in macrophages that were subjected to oxidative stress, by their preincubation with angiotensin II, exhibited a 51 – 59% increase in their cell surface proteoglycan content, with a major effect on chondroitin sulfate proteoglycans. The present study thus demonstrated that Ox-LDL can specifically bind to macrophage surface chondroitin sulfate proteoglycans, and the macrophage content of this proteoglycan is increased under oxidative stress. The interaction between macrophage chondroitin sulfate proteoglycans and Ox-LDL can contribute to enhanced uptake of Ox-LDL with the formation of cholesterol-loaded foam cells, and accelerated atherosclerosis. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Chondroitin sulfate; Heparan sulfate; Lipid peroxides; Macrophages; Oxidized LDL; Proteoglycans

1. Introduction

* Corresponding author. Lipid Research Laboratory, Rambam Medical Center, Haifa 31096, Israel. Tel.: + 972-4-8542970; fax: +972-4-8542130. E-mail address: [email protected] (M. Aviram)

Oxidation of LDL appears to play a significant role in atherogenesis [1–4]. Unlike the cellular uptake of native LDL via the LDL receptor, which is regulated by cellular cholesterol content, Ox-LDL uptake via the macrophage scavenger receptors, was shown to induce cellular cholesterol accumulation [5–7]. Recently, using

0021-9150/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 9 9 ) 0 0 2 8 7 - 7

6

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

specific glycosaminoglycans hydrolyzing enzymes, we have shown that macrophage plasma membrane glycosaminoglycans (GAGs) also contribute to the cellular uptake of Ox-LDL [8], but the exact proteoglycans required for the specific interaction with Ox-LDL were not characterized. Proteoglycans (PGs) consist of a core protein to which glycosaminoglycan chains of repeating disaccharide units are linked, and these polyanionic glycosaminoglycan chains dominate the physical properties of proteoglycans [9]. Proteoglycans were shown to be involved in atherogenesis, either as cell surface lipoprotein receptors [10 – 12] or as a component of the extracellular matrix of the atherosclerotic lesions [13 – 15]. Cell surface proteoglycans include syndecan that contains three to five chondroitin sulfate and heparan sulfate glycosaminoglycan chains [16], perlecan with a large core protein and three heparan sulfate chains [17], and some other proteoglycans containing only chondroitin sulfate chains [18]. Although heparan sulfate proteoglycan is a ubiquitous cell surface PG of most cell types, some monocyte, as well as macrophage cell lines, such as M1 and P388D1, and human peritoneal macrophages, did not exhibit heparan sulfate proteoglycan on their cell surfaces and were shown to produce exclusively chondroitin sulfate proteoglycans [19 – 21]. Moreover, in post-mortem samples from humans, it has been shown that chondroitin sulfate increases preferentially during fatty streak formation [22]. Arterial wall proteoglycans and especially chondroitin sulfate proteoglycans have been known to interact specifically with apolipoprotein B100-containing lipoproteins, mainly with LDL, Lp(a) and, to a lesser extent with VLDL, but not with HDL [23 – 25]. Lipoprotein–proteoglycan interaction leads to lipoprotein entrapment, modification, and uptake by macrophages in the intimal extracellular space thereby promoting foam cell formation [26 – 28]. Native LDL binds proteoglycans via ionic interactions between the positively charged lysine and arginine residues of the LDL apolipoprotein B-100, and the negatively charged sulfate and carboxyl groups of the GAG chains in proteoglycans [29,30]. Oxidative stress has been shown to affect not only the LDL lipids, but also cellular lipids including those in arterial wall macrophages which can undergo lipid peroxidation under oxidative stress [31,32]. We have previously shown that angiotensin II can induce oxidative damage to macrophages [32]. In addition, lipid peroxidized macrophages that were obtained following cells incubation with angiotensin II exhibited an increased ability to oxidize LDL, and to take up Ox-LDL [32]. These lipid-peroxidized macrophages also contain elevated levels of proteoglycans [33]. Murine J-774 A.1 macrophages, unlike other macrophage cell lines, are considered to be a good

model to study atherogenesis, since they are able to take up modified lipoproteins at enhanced rate and to accumulate large amounts of cholesterol ester, leading to foam cell formation [34]. Thus, these cells represent an appropriate model for foam cell formation, which is a prominent feature of the early atherosclerotic plaque. The purpose of the present study was to identify J-774 A.1 macrophage cell surface proteoglycan(s) which specifically bind Ox-LDL, and can thus contribute to its enhanced uptake by macrophages, in comparison to native LDL. In addition, since lipid-peroxidized macrophages exhibit an increased Ox-LDL cellular uptake, cell surface proteoglycans from macrophages that were submitted to oxidative stress, were also analyzed.

2. Methods

2.1. Materials Sodium [35S]sulfate and carrier-free Na [125I] were purchased from Du Pont, New England Nuclear (Boston, MA). J-774 A.1 murine macrophage-like cell line, as well as the Chinese hamster ovary cells (CHO), were purchased from the American Type Culture Collection (ATCC, Rockville, MD). Mutant CHO cells, including mutants number 677 and 745, were kindly provided by Dr J. Esko, University of California, San Diego, USA. Dulbecco’s modified Eagle’s medium (DMEM), fetal calf serum (FCS), phosphate-buffered saline (PBS), bovine serum albumin (BSA) and trypsin were purchased from Biological Industries (Beit Haemek, Israel). Heparinase III, chondroitinase ABC, chondroitinase AC, chondroitin sulfate A, heparan monosulfate and angiotensin II were purchased from Sigma (St Louis, MO), DEAE–Sephacel, Sepharose 6B, and CnBr-activated Sepharose CL-4B were from Pharmacia (LKB, Sweden). Dimethylmethylene Blue was purchased from Aldrich (Milwaukee, WI).

2.2. Lipoproteins Low-density lipoprotein (LDL) was prepared from human plasma (drawn into 1 mM of Na2 EDTA) from fasted normolipidemic volunteers. LDL (d=1.019– 1.063 g/ml) was prepared by discontinuous density gradient ultracentrifugation, as described previously [35]. The lipoprotein was washed at d= 1.063 g/ml and dialyzed against 150 mM NaCl, 1 mM Na2 EDTA, pH 7.4. LDL was then sterilized by filtration and was used within 2 weeks. The protein content of the lipoproteins was determined with the folin phenol reagent [36]. LDL was radioiodinated using the iodine monochloride method [37] modified for lipoproteins. Oxidized LDL was prepared by an overnight dialysis of LDL or

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17 125

I-labeled LDL (1 mg of lipoprotein protein/ml) against PBS to remove any residual EDTA, followed by incubation with 10 mM CuSO4 for 18 h at 37°C. Oxidation was terminated by refrigeration and the addition of 0.1 mM Na2 EDTA. The degree of LDL oxidation was determined by analysis of malondialdehyde (MDA) equivalents using the thiobarbituric acid reactive substance (TBARS) assay [38], and it ranged between 18 and 25 nmol of MDA equivalents/mg lipoprotein protein compared to 0.5 – 1.0 nmol of MDA equivalents/mg lipoprotein protein in native LDL. Electrophoresis of lipoproteins was performed on 1% agarose using a Hydragel-Lipo kit, (Sebia, France). Changes in apolipoprotein B-100 fragmentation and size after LDL oxidation were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis [39].

2.3. Cell culture and labeling J-774 A.1 cells were plated at 2.5 ×105 cells/16-mm dish in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS), 100 U of penicillin/ml, 100 mg streptomycin/ml, and 2mM glutamine. The cells were fed every 3 days and used for experiments within 7 days of plating. After cell washing with PBS, the macrophages were radiolabeled with 30 mCi/ml of [35S]sodium sulfate for 24 h at 37°C in DMEM supplemented with 10% FCS.

2.4. Cell surface proteoglycans isolation and purification Trypsin treatment was used to isolate cell surface proteoglycans. J-774 A.1 macrophages were washed with PBS, and incubated with 0.05% trypsin and 0.02% EDTA in PBS for 10 min at 37°C. The cells were then scrapped into trypsin and further incubated for 5 min at 37°C. To stop the trypsin action, the cell suspensions were removed to an ice bath, and fresh medium with FCS was added to the suspensions. The cells were separated from the trypsin by centrifugation for 5 min at 450×g. The trypsin fluids (supernatant) contained the ectodomains of cell surface proteoglycans. For proteoglycan purification, samples were dialyzed in a buffer consisting of 7 mol/l urea and 0.1% Triton X-100 in 0.05 mol/l Tris, pH 7.2, and applied onto a DEAE– Sephacel column (8 ml). The column was washed with five bed volumes of the above buffer with 0.15 mol/l NaCl added, followed by an additional five bed volumes of this buffer without NaCl. A continuous gradient (48 ml) of 0.0–0.8 mol/l of NaCl in the same buffer was used to elute the proteoglycans from the column [40]. Fractions (1 ml) were collected and measured for their radioactivity content.

7

2.5. Cell surface glycosaminoglycan characterization To prepare glycosaminoglycan chains, aliquots (2 ml) of purified cell surface proteoglycans were treated with 200 ml of 10 N NaOH for 18 h at 26°C with constant shaking, and then neutralized with 10 N HCl. For glycosaminoglycan purification, core proteins were removed by adjusting the sample concentration to 7 mol/l urea, followed by their loading on a DEAE minicolumn (1 ml) that was washed with three bed volumes of 7 mol/l urea, 0.1% Triton X-100, and 0.2 mol/l NaCl in 0.05 mol/l Tris, pH 7.2. Glycosaminoglycans were eluted with a solution of 7 mol/l urea, 0.1% Triton X-100, and 1 mol/l NaCl in 0.05 mol/l Tris, pH 7.2. Urea was removed by dialysis against 10 mmol/l Tris and 0.1% Triton X-100, pH 7.2. Preincubations of the 35 S-labeled glycosaminoglycans with 0.1 U/ml of either heparinase III or chondroitinase ABC for 18 h at 37°C were used in order to hydrolyze chondroitin sulfate and heparan sulfate, respectively. Non-hydrolyzed and hydrolyzed glycosaminoglycans were chromatographed on a Sepharose 6B column using 0.2 mol/l NaCl as eluant [40].

2.6. Glycosaminoglycan determination Macrophage glycosaminoglycan content was analyzed using the 1,9-dimethylmethylene blue (DMMB) spectrophotometric assay for sulfated glycosaminoglycans [41]. Briefly, 2.5 ml of ice-cold DMMB working solution (46 mM DMMB, 40 mM glycine, 40 mM NaCl in 5% ethanol adjusted to pH 3.0) were added to 100 ml of cell sonicate or plasma membrane fraction on ice. The absorbence at 525 nm was then immediately measured. Chondroitin sulfate A was used as a standard and was included within each series of assays.

2.7. Binding of macrophage membrane proteoglycans to Ox-LDL–Sepharose The lipoproteins (Ox-LDL or native LDL) were coupled to CnBr-activated Sepharose CL-4B as described previously for native LDL coupling [42]. Briefly, OxLDL or native LDL (3 mg) was coupled to 1 g CnBractivated Sepharose CL-4B in coupling buffer (0.1 M NaHCO3 containing 0.5 M NaCl pH 8.3) for 18 h at 4°C. The amount of lipoproteins bound to the column was determined by the measurement of the lipoprotein protein levels in the fraction that was first eluted from the column, and were found to be 0.95 9 0.05 mg of protein for Ox-LDL and 0.80 9 0.04 mg of protein for LDL (n= 3). Since the initial amount of Ox-LDL and native LDL added to the column was 3 mg, almost identical amounts of Ox-LDL and native LDL were able to bind to the CnBr-activated Sepharose column. The lipoprotein levels that bound to the Sepharose

8

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

column were in excess thus, lipoprotein concentration was not a limiting step in the interaction between PGs and lipoproteins. Elution of the Ox-LDL or LDL from a blocked CnBr–Sepharose control column illustrated only minor unspecific binding. Prior to use, the lipoprotein-Sepharose column was washed three times with one volume of binding buffer (30 mM CaCl2, in 10 mM Tris –HCl, pH 7.2) containing 0.1% bovine serum albumin (BSA), followed by 3 washes with one volume of binding buffer without BSA. Equal amounts of purified cell surface 35S-labeled proteoglycans were diluted in 1 ml of binding buffer and incubated with 4 ml of either Ox-LDL or native LDL coupled to Sepharose for 15 min at 25°C. Samples were centrifuged at 250×g for 1 min, and the supernatant removed. Washes (1 ml × 3) with binding buffer containing 0, 0.01, 0.05, 0.1, 0.25, 0.5 and 0.8 M of NaCl were used to release the bound PG from the lipoprotein coupled to Sepharose. Following each wash, samples were centrifuged at 250× g for 1 min and the supernatant removed. Aliquots of supernatants were then analyzed for their radioactivity content. Bound proteoglycans were determined as the total amount of 35S-labeled proteoglycans that required greater than 0.05 M NaCl to be released from the Ox-LDL or from the LDL substituted Sepharose column.

2.8. Cellular lipid peroxidation To analyze the cell surface PGs in macrophages that were submitted to oxidative stress, J-774 A.1 macrophages were incubated with increasing concentrations (10 − 8 –10 − 6 M) of angiotensin II [32] together with the 35S-radiolabeled sodium sulfate for 24 h at 37°C. The level of cellular lipid peroxidation was determined by lipid peroxide determination in a sonicated cells preparation [43].

2.9. Macrophage plasma membrane preparation In order to analyze the GAG levels in plasma membrane of cells that were submitted to oxidative stress, cell plasma membrane fraction was isolated from control and from lipid-peroxidized macrophages. Cells (2× 106/35-mm dish) were washed with cold PBS (× 3), harvested, and suspended in 2 ml of 250 mM sucrose containing 5 mM Tris – HCl buffer, pH 7.4. The cells were then sonicated for 20 s at 20 W (× 2), homogenized in a Teflon/glass homogenizer (15 strokes), and centrifuged at 500×g for 10 min. The supernatant was then centrifuged at 10 000× g for 45 min to precipitate the lysosomal-rich fraction. The remaining supernatant was centrifuged at 100 000× g for 60 min to precipitate the plasma membrane fraction [44]. The plasma membrane fraction was then resuspended in saline and analyzed for its GAG content, as

described below, as well as for the incorporation of the 35 S label in cell surface proteoglycans.

2.10. Lipoprotein binding to chondroitin sulfate LDL that was previously dialyzed against PBS was oxidized using increasing concentrations (2.5–20 mM) of CuSO4. The degree of oxidation was determined by analysis of the thiobarbituric acid reactive substance (TBARS) assay and expressed as malondialdehyde (MDA) equivalents. The lipoprotein preparations were analyzed for their capacity to bind chondroitin sulfate. LDL (200 mg of lipoprotein protein/ml) was incubated with chondroitin sulfate (CS, 100 mg/ml) for 30 min at room temperature. The lipoprotein was then precipitated with a commercial kit for HDL cholesterol reagent (phosphotungstic acid/MgCl2; Sigma, St. Louis, MO), that precipitated all the LDL present in the samples followed by a 10-min centrifugation at 2000× g [45]. After discarding the supernatant, the precipitated LDL was dissolved in 0.1 N NaOH, and analyzed for its glycosaminoglycan (GAG) content using the 1,9-dimethylmethylene blue (DMMB) spectrophotometric assay for sulfated glycosaminoglycans [41]. The absorbance at 525nm was then immediately analyzed. Chondroitin sulfate was used as a standard and it was included in each series of assays. Similar preparations of LDL, with no chondroitin sulfate added, were used as control. GAG content obtained in the control was subtracted from the GAG content in LDL preparations that were incubated with chondroitin sulfate (CS).

2.11. CHO cell mutants Chinese hamster ovary (CHO) cells were cultured in Ham’s F-12 medium containing 10% fetal bovine serum. Mutant 745 is defective in xylosyltransferase and therefore lacks both heparan sulfate and chondroitin sulfate. The mutant 677 is characterized by a deficiency in heparan sulfate and an excess of chondroitin sulfate [45].

2.12. Cellular metabolism of lipoproteins 2.12.1. Lipoprotein cellular degradation Cell-mediated degradation of Ox-LDL or native LDL was measured following incubation of CHO cells (1×106/16-mm dish) with 10 mg of lipoprotein protein/ ml of [125I]Ox-LDL or [125I]LDL in serum-free DMEM containing 0.2% bovine serum albumin (BSA) for 4 h at 37°C. Cell-mediated hydrolysis of the lipoprotein protein was assayed by determination of the trichloroacetic acid (TCA) soluble, chloroform insoluble radioactivity in the incubation medium [46]. Degradation of the lipoproteins in the absence of cells was minimal and was always subtracted from the total lipoprotein degradation.

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

9

2.12.2. Lipoprotein cellular binding Binding of Ox-LDL to CHO cells was studied by incubation of 10 mg of lipoprotein protein/ml of 125I-labeled Ox-LDL with the cells for 4 h at 4°C. After extensive washing with PBS ( ×4), cells were solubilized by incubation in 0.1 N NaOH for 1 h at room temperature, and the level of bound radiolabeled lipoproteins was then counted in a g counter [47]. 2.13. Statistical analyses The Student t-test was used in order to analyze the significance of the results. Results are given as mean 9 SD.

3. Results

3.1. Macrophage plasma membrane proteoglycan characterization Cell surface proteoglycans that were isolated from [35S]sodium sulfate radiolabeled J-774 A.1 macrophages by trypsin treatment, were further purified by DEAE– Sephacel ion exchange chromatography, using a linear gradient from 0–1 mol/l NaCl for elution (Fig. 1). As seen in Fig. 1, the 35S-labeled proteoglycans from J-774 A.1 macrophages were eluted with 0.2 – 0.5 M NaCl. In

Fig. 1. Elution profile of macrophage cell surface proteoglycans on DEAE – Sephacel. J-774 A.1 macrophages were radiolabeled for 24 h at 37°C, with 30 mCi/ml of [35S]sodium sulfate. Proteoglycans were isolated from the cells as described in Section 2. The labeled proteoglycans were then chromatographed on a DEAE–Sephacel column (8 ml). Proteoglycans were eluted with a linear gradient (48 ml) of 0.0–0.8 mol/l NaCl in urea/Tris buffer. Fractions (1 ml) were collected and analyzed for their radioactivity content. A representative profile of five different experiments is shown.

Fig. 2. Elution profiles of 35S-labeled glycosaminoglycans isolated from macrophages cell surface after chromatography on Sepharose 6B. Glycosaminoglycans were prepared by base treatment of the cell surface proteoglycans as described under Section 2. Glycosaminoglycans were purified by additional chromatography on DEAE–Sephacel and preincubated in the absence (control, A) or presence of 0.1 U/ml of either heparinase III (B) or of chondroitinase ABC (C) for 18 h at 37°C. Non-hydrolyzed and hydrolyzed glycosaminoglycans were then chromatographed on a Sepharose 6B column using 0.2 mol/l NaCl as the eluant. Fractions (1 ml) were collected and analyzed for their radioactivity content. One representative profile of five different experiments is shown.

order to find out whether the isolated 35S-labeled cell surface material represents PGs molecules or free GAGs, a control experiment including incubation of the fraction with or without NaOH (to release free GAGs from PGs) and its further elution on a Sepharose 6B column, was performed. Indeed, size separation chromatography of the 35S-labeled cell-associated material, before and after alkali treatment, exhibits a complete shift after alkali treatment, indicating that almost all the 35S-labeled cell surface fractions consisted of PGs and not of free GAGs (data not shown). To identify specific glycosaminoglycans, the fractions obtained after DEAE–Sephacel chromatography were incubated in the absence (Fig. 2A) or presence of 0.1 U/ml of either heparinase III (Fig. 2B) or chondroitinase ABC (Fig. 2C) for 18 h at 37°C, prior to their separation on Sepharose 6B column. The elution profi-

10

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

les of isolated [35S]glycosaminoglycan chains, isolated from the cell surface PGs, on Sepharose 6B as shown in Fig. 2A, illustrated a similar elution profile when glycosaminoglycans were measured by their 35S radioactivity, or by the dimethylmethylene blue assay for glycosaminoglycans mass. As seen in Fig. 2A, the control non-hydrolyzed fraction exhibits two major peaks (I and II). Following treatment with heparinase III (Fig. 2B), peak II disappeared, while peak I remained unchanged. Following treatment with chondroitinase ABC (Fig. 2C), peak number I disappeared, while peak II remained unchanged. These results suggest that peak I is chondroitin sulfate, whereas peak II is heparan sulfate. From both 35S radioactivity in glycosaminoglycans and the glycosaminoglycan content analyses, chondroitin sulfate was found to represent 77 – 81% of the total macrophages cell membrane glycosaminoglycans, whereas heparan sulfate represented 19 – 23% of the total glycosaminoglycans isolated from macrophage cell surfaces (Fig. 2). Treatment of the control fraction with chondroitinase AC, which hydrolyzes only chondroitin sulfate, resulted in the same elution profile as treatment with chondroitinase ABC (data not shown), suggesting that dermatan sulfate is not present in these macrophage cell surface proteoglycans.

3.2. Interaction of macrophage plasma membrane proteoglycans with Ox-LDL The potential of macrophage membranal proteogly-

Fig. 3. Macrophage membranal proteoglycans binding to Ox-LDL or to native LDL. Cell membrane proteoglycans were diluted in ‘binding buffer’ (0.1 M NaHCO3, 0.5 M NaCl, pH 8.3), and incubated with either Sepharose-coupled Ox-LDL or native LDL for 15 min at 25°C. Samples were centrifuged at 250×g for 1 min and the supernatant was removed. Lipoprotein-bound proteoglycans were released from Ox-LDL or from native LDL–Sepharose by washes (1 ml × 3) with ‘binding buffer’ containing 0, 0.01, 0.05, 0.1, 0.25, 0.5, and 0.8 mol/l NaCl. For each sequential wash, samples were centrifuges at 250 × g for 1 min, and the resulting supernatants were analyzed for their radioactivity content. Results represent mean 9 SD of four different experiments.

cans to interact with Ox-LDL, in comparison to native LDL, was assessed by their ability to bind Ox-LDL or native LDL that were pre-coupled to a Sepharose column (Fig. 3), followed by release of the PG bound fraction using an increasing salt gradient. Of the total loaded cell membrane proteoglycans, 57% were bound to the Sepharose-coupled Ox LDL (as measured by a total of 1620 cpm/ml released by 0.05–0.8 M of NaCl out of 2850 cpm/ml loaded on the Sepharose-coupled Ox-LDL column). Most of this bound fraction was released by low (0.10–0.25 M) NaCl concentrations (Fig. 3). Unbound 35S-labeled PGs (the remaining 43% counts) were eluted from the column with buffer prior to the introduction of the NaCl gradient. In contrast, about 73% of the loaded cell surface proteoglycans were bound to the Sepharose-coupled native LDL (as measured by a total of 2360 cpm/ml released by the 0.05–0.8 M of NaCl gradient out of 3240 cpm/ml that were loaded on the Sepharose-coupled LDL column). Most of this bound fraction was released by higher (0.25–0.50 M) NaCl concentrations than that required for the elution of Ox-LDL (Fig. 3). The unbound 35 S-labeled PG fraction (the remaining 27% counts) was eluted from the column with elution buffer before introduction of the NaCl gradient. Radiolabeled proteoglycans did not bind at all to control Sepharose which was prepared in the absence of lipoproteins (data not shown). The specificity of macrophage proteoglycans binding to Ox-LDL was next demonstrated by competition studies. Binding of the macrophage membranal 35S-labeled proteoglycans to Sepharose-coupled Ox-LDL was competitively inhibited by preincubation of the lipoprotein with increasing concentrations (10– 50 mg/ml) of non-labeled chondroitin sulfate (Fig. 4). The competition by chondroitin sulfate was concentration dependent, and in the presence of 50 mg/ml of chondroitin sulfate the binding of radiolabeled proteoglycans to Ox-LDL was inhibited by 73%, indicating that chondroitin sulfate specifically interacts with OxLDL (Fig. 4). Preincubation of the Sepharose-coupled Ox-LDL with increasing concentrations (10–50 mg/ml) of non-labeled heparan sulfate however, did not significantly inhibit the binding of proteoglycans to Ox-LDL, suggesting that heparan sulfate is not involved in the binding of macrophage membranal proteoglycans to Ox-LDL (Fig. 4). To further analyze the nature of the proteoglycans involved in the interaction between macrophage membranal proteoglycans and Ox-LDL, the isolated macrophage membrane 35S-labeled proteoglycans were pretreated with 0.1 U/ml of either chondroitinase ABC or heparinase III for 18 h at 37°C, prior to their incubation with Sepharose-coupled Ox-LDL (Fig. 5). As seen in Fig. 5, the 35S-labeled proteoglycan fraction treated with chondroitinase ABC lost most of its capac-

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

Fig. 4. Glycosaminoglycan competition for Ox-LDL binding to macrophage membranal proteoglycans to Ox-LDL. Sepharose-coupled Ox-LDL was preincubated with the indicated amount of competitor (heparan sulfate or chondroitin sulfate) for 5 min at 25°C, in ‘binding buffer’ (0.1 M NaHCO3, 0.5 M NaCl, pH 8.3). Samples were then incubated with 1 ml of 35S-labeled macrophage membranal proteoglycans for 15 min at 25°C, centrifuged for 1 min at 250 × g and the supernatant was removed. Ox-LDL-bound proteoglycans were released by washes (1 ml ×3) with ‘binding buffer’ containing 0.8 mol/l NaCl and aliquots of the supernatants were analyzed for their radioactivity content. Results represent mean 9SD of four different experiments.

ity to bind Ox-LDL, compared to the capacity of the non-hydrolyzed fraction, since only 13% of the chondroitinase-treated 35S-labeled proteoglycan fraction still binds to Ox-LDL (as measured by a total of 534 cpm/ml released by 0.05 – 0.8 M of NaCl out of 4229 cpm/ml that were loaded on the Sepharose-coupled Ox-LDL column), compared to a 57% binding for the non-treated fraction (see Fig. 3). In contrast, the heparinase-treated 35S-labeled proteoglycans fraction retained almost the same capacity to bind Ox-LDL as the non-treated fraction (Fig. 5), as 45% of the heparinasetreated 35S-labeled proteoglycan fraction binds OxLDL (as measured by a total of 1934 cpm/ml released by 0.05–0.8 M of NaCl out of 4301 cpm/ml that were loaded on the Sepharose-coupled Ox-LDL column), compared to 57% binding for non-hydrolyzed proteoglycans (see Fig. 3A). These results indicate that the chondroitin sulfate glycosaminoglycan chains are mostly involved in the interaction between the macrophage membranal proteoglycans and Ox-LDL. To further evaluate the role of macrophage membranal chondroitin sulfate proteoglycans in the binding of Ox-LDL to the cells, mutant Chinese hamster ovary

11

(CHO) cells characterized by glycosaminoglycan deficiencies were used. Ox-LDL was incubated with the wild-type CHO cells, with mutant 745 CHO cells (which lack both heparan sulfate and chondroitin sulfate), or with the mutant 677 CHO cells (which lack heparan sulfate but have increased levels of chondroitin sulfate), for 4 h either at 4°C for determination of the lipoprotein cellular binding (Fig. 6A), or at 37°C for determination of the lipoprotein cellular degradation (Fig. 6B). At 10 mg of Ox-LDL protein/ml, Ox-LDL binding and degradation to the mutant 745 CHO cells were decreased by 28 and 27%, respectively, compared to Ox-LDL binding to the wild-type CHO cells, (Fig. 6A). Ox-LDL binding and degradation to the mutant 677 CHO cells was increased by 29 and 19%, respectively, compared to Ox-LDL binding to the wild-type CHO cells (Fig. 6A). These results further indicate that cell surface chondroitin sulfate, but not heparan sulfate, binds Ox-LDL. For comparison, native LDL degradation by CHO mutant 677, 745 and the wild type was analyzed by incubation of 125I-labeled LDL (10 mg lipoprotein protein/ml) with either the wild-type CHO cells, with mutant 745 CHO cells, or with mutant 677 CHO cells, for 4 h at 37°C, followed by determination of the lipoprotein cellular degradation. LDL degradation by mutant 677 CHO cells and by mutant 745 CHO cells was decreased by 23 and 39%, respectively, in comparison to the wild type cells. A decrease from

Fig. 5. The effect of treatment of macrophage membranal proteoglycans with chondroitinase ABC or with heparinase III on their binding to Ox-LDL. Purified labeled macrophage membranal proteoglycans were treated with 0.1 U/ml of either chondroitinase ABC or with heparinase III for 18 h at 37°C. Samples were then diluted in ‘binding buffer’ and incubated with 1 ml of Sepharose-coupled Ox-LDL for 15 min at 25°C, centrifuged at 250 × g for 1 min, and the supernatant was removed. Lipoprotein-bound proteoglycans were released by washes (1 ml × 3) with ‘binding buffer’ containing 0, 0.01, 0.05, 0.1, 0.25, 0.5 and 0.8 M NaCl. Following each wash, samples were centrifuged at 250 ×g for 1 min, and aliquots of supernatants were analyzed for their radioactivity content. Results represent mean 9SD of four different experiments.

12

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

negative charge (data not shown). As seen in Fig. 7, increasing LDL oxidation rates, as measured by MDA equivalents levels, induced a substantial increase (up to 2.5-fold) in the ability of Ox-LDL to bind chondroitin sulfate up to 22 nmol of MDA/mg lipoprotein protein. Above this level of LDL oxidation, however, the ability of Ox-LDL to bind chondroitin sulfate was decreased and reached levels that were even lower than those obtained with native LDL (Fig. 7).

3.4. Analysis of cell surface proteoglycans in lipid-peroxidized macrophages

Fig. 6. Cellular binding and degradation of Ox-LDL to glycosaminoglycan-deficient Chinese hamster ovary cells (CHO). 125I-labeled OxLDL (10 mg of protein/ml) was incubated with either the wild-type CHO cells, or with mutant 745 CHO cells (characterized by a deficiency in heparan sulfate and chondroitin sulfate), or with mutant 677 CHO cells (which lack heparan sulfate but contain increased levels of chondroitin sulfate) for 4 h at 4°C for binding determination (A) and at 37°C for degradation determination (B). Results represent mean 9 SD of four different experiments. *PB0.01 versus control cells).

To analyze cell surface proteoglycans in macrophages that were submitted to oxidative stress, J-774 A.1 macrophages were incubated in the absence (control cells) or presence of increasing concentrations of angiotensin II (10 − 8 –10 − 6 M) together with [35S]sodium sulfate for 25 h at 37°C. Then, cellular glycosaminoglycan content, as well as 35S-labeled proteoglycans, were determined in the isolated cell plasma membranes. In parallel, the extent of cellular lipid peroxidation was also determined. Incubation of the cells with angiotensin II induced a dose-dependent increment in cellular lipid peroxidation compared to control cells. Macrophage lipid peroxide content increased by up to 3-fold following incubation with 10 − 6 M of angiotensin II, in comparison to control cells (Table 1). Both the 35 S-labeled proteoglycan and glycosaminoglycan con-

0.319 0.04 mg of lipoprotein protein/mg cell protein in wil-type CHO cells, to 0.2490.01 and 0.199 0.01 mg of lipoprotein protein/mg cell protein in the mutant 677 CHO cells and in mutant 745 CHO cells, respectively, was found. These results indicate that HS and CS could be involved in the binding of native LDL, unlike the binding of only CS to Ox-LDL.

3.3. The effect of the le6el of CuSO4 -induced LDL oxidation on its interaction with chondroitin sulfate To analyze whether the extent of LDL oxidation affects the ability of the lipoprotein to bind chondroitin sulfate, LDL oxidation was induced with increasing concentrations of CuSO4, prior to the analysis of its binding to chondroitin sulfate. Oxidation of LDL with increasing concentrations of CuSO4 (2.5 – 20 mM) led to an increase in their MDA content, as well as their

Fig. 7. The effect of the extent of LDL oxidation on its binding to chondroitin sulfate LDL (200 mg of lipoprotein protein/ml) which was oxidized by using increasing concentrations (2.5 – 20 mM) of CuSO4 for 6 h at 37°C, was incubated with chondroitin sulfate (CS, 100 mg/ml) for 30 min at room temperature. Ox-LDL was then precipitated, and lipoprotein-associated GAGs content was analyzed in the precipitate using the DMMB assay, as described in Section 2. Results are presented as mean 9S.D. The degree of LDL oxidation was determined by the thiobarbituric acid reactive substance (TBARS) assay.

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

13

Table 1 Cell membrane content of glycosaminoglycans and proteoglycans in angiotensin II lipid-peroxidized macrophagesa

Control cells + Angiotensin II (10−8 M) + Angiotensin II (10−7 M) + Angiotensin II (10−6 M)

Cell membrane 35S-labeled PGs (cpm/mg cell membrane protein)

Cellular lipid peroxides (nmol/mg cell protein)

Cell membrane glycosaminoglycans (mg/mg cell membrane protein)

11.5 9 1.1 14.7 9 2.6

48.3 9 5.4 47.594.0

731 728

23.8 92.2*

63.8 93.8*

814

33.6 91.82*

72.8 95.1*

1165

a

J-774 A.1 macrophages were incubated in the absence or presence of increasing concentrations of angiotensin II (10−8–10−6 M), together with [ S]sodium sulfate for 25 h at 37°C. Then, the glycosaminoglycans content, as well as the incorporated 35S-labeled proteoglycans (PGs) were determined in the isolated cell membranes. Cellular lipid peroxidation was determined in the cells by peroxide assay. Results are given as mean9 S.D. * PB0.01 versus control cells. 35

tent in the macrophage cell membranes increased by up to 59 and 51%, respectively, in cells that were lipid peroxidized with 10 − 6 M angiotensin, compared to control cells (Table 1). Cell surface 35S-labeled proteoglycans from macrophages that were preincubated with 10 − 6 M angiotensin II were isolated by DEAE – Sephacel chromatography (data not shown). Following base treatment and release of the 35S-labeled glycosaminoglycan fraction (using DEAE – Sephacel chromatography), the 35S-labeled glycosaminoglycans were pretreated with 0.1 U/ml of either heparinase III or chondroitinase ABC for 18 h at 37°C, prior to their elution on a Sepharose 6B column. As seen in Fig. 8, lipid peroxidized cells exhibited an increase by 56% (pB0.01) in the chondroitin sulfate fraction, whereas the heparan sulfate fraction remained similar, as compared to control cells. These results thus demonstrate that lipid peroxidized cells are characterized by a significant increase in cell surface proteoglycan content, especially chondroitin sulfate proteoglycans.

and hence became more atherogenic, also exhibited increased levels of cell membrane chondroitin sulfate proteoglycans. Heparan sulfate and chondroitin sulfate proteoglycans have been previously shown to be major cell surface proteoglycans [48–50], but their distribution differs among different cell types. In the peri-cellular compartment of pigeon macrophages, 40% of the proteoglycans were found to be chondroitin sulfate proteoglycans, and the remainder heparan sulfate

4. Discussion The present study demonstrated that Ox-LDL can specifically bind to macrophage surface chondroitin sulfate proteoglycans. Using 35S labeling of the cell surface proteoglycans, chondroitin sulfate was identified as the major glycosaminoglycan in the plasma membrane of J-774 A.1 macrophage-like cell line, whereas heparan sulfate represented only a minor fraction of the glycosaminoglycans in the macrophage cell surface. Using affinity chromatography techniques, and cells with specific glycosaminoglycan deficiencies, it was clearly demonstrated that cell surface chondroitin sulfate proteoglycans bind Ox-LDL. In addition, macrophages that were submitted to oxidative stress,

Fig. 8. Cell membrane distribution of chondroitin sulfate and heparan sulfate in lipid-peroxidized macrophages, compared to control cells. Following incubation of J-774 A.1 macrophages in the absence (control) or presence of 10 − 6 M of angiotensin II (lipid-peroxidized cells), 35S-labeled glycosaminoglycan fraction was isolated from cell surface 35S-labeled proteoglycans as described in Section 2. The radioactivity associated with either chondroitin sulfate or heparan sulfate was then analyzed based on the susceptibility of 35S-labeled glycosaminoglycan fraction to hydrolysis by incubation with 0.1 U/ml of either chondroitinase ABC or heparinase III for 18 h at 37°C, respectively. Results represent mean 9SD of four different experiments. *PB 0.01 versus control cells.

14

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

proteoglycans [42]. In contrast, the macrophage-like cell lines P338D1, U-937 and M1 produced no detectable heparan sulfate proteoglycans, but mainly chondroitin sulfate proteoglycans [19,20,51]. In addition, structural changes in cell surface chondroitin sulfate proteoglycans were apparent when THP-1 monocytes differentiated into macrophages [40]. In another study, when analyzing the glycosaminoglycan content of human aortas, the total glycosaminoglycan levels, and especially the chondroitin sulfate content, exhibited a marked increase with age [52]. A recent study has shown that cell surface J-774 A.1 macrophages contain mainly HS and a small amount of CS, and that cell surface HSPGs but not CSPGs were able to bind Ox-LDL [48], as opposed to our study which shows chondroitin sulfate as the major macrophage surface proteoglycan. The discrepancy with our study could reside in the fact that, in the study of Halvorsen et al. [48], most of the 35S-labeled material released from the cell surfaces with trypsin were GAGs and not PGs, as opposed to the results in our study where all of the 35 S-material released from the cell surfaces with trypsin were PGs. Indeed, our study analysis of the 35S-labeled fraction released by trypsin treatment from the macrophages cell surfaces, on a CL-6B Sepharose column, before and after alkali treatment, exhibits a complete shift after alkali treatment, indicating that these fractions originally consisted of PGs and not of free GAGs. In addition, it is well documented that all GAGs (except hyaluronan) are synthesized by assembly onto a core polypeptide [53], therefore all GAGs that were isolated from the cell surface of the macrophages originated from PGs. Native LDL has been previously shown to bind proteoglycans [23,30,54], and the interaction between LDL and proteoglycans has been previously shown to be mediated by negatively charged glycosaminoglycans [29,30]. Moreover, the principal proteoglycan-binding site (amino acids 3362 – 3364) in LDL was identified using a single-point mutation in apolipoprotein-B100 [55,56], and was shown to differ from the LDL receptor binding site. In the present study we showed that 57% of the total cell membrane proteoglycans isolated from J774 A.1 macrophages bind to oxidized LDL, whereas more than 70% of these proteoglycans bind to native LDL. Thus, in spite of the LDL oxidative modification, and its increased negative charge, Ox-LDL was still able to substantially bind cell membrane proteoglycans. Since both oxidized LDL and proteoglycans share a negative charge, the ability of Ox-LDL to bind chondroitin sulfate may be related to the exposition of novel sites and to conformational changes that take place during LDL oxidation [57]. Local positive charges in the lipoprotein particle can also still be present after LDL oxidation [57]. Moreover, proteolysis of apoB-100 has been shown to increase the binding of LDL to

proteoglycans, suggesting the exposure of buried heparin binding sites [58]. Studies of interactions between PGs and Ox-LDL, as well as the nature of the PGs which are involved in this interaction, has led to contradictory reports in the literature. LDL binding to proteoglycans can induce molecular inter-conversion, including structural modifications of the PGs, as well as of the lipoprotein [59,60]. Ox-LDL was also shown to bind proteoglycans, with alteration of its polyanionic glycosaminoglycan binding sites, possibly by intervention of reactive oxygen species [61,62]. These modifications can result in an altered length or molecular weight of the glycosaminoglycan chains, or in modification of the core protein structure, which may then exhibit different affinities for lipoproteins [63]. A recent study has shown that oxidation of LDL particles decreases their ability to bind to aortic proteoglycans [64]. In this study, affinity chromatography of proteoglycans and oxidized LDL yielded two lipoprotein fractions with the same content of lipid peroxides, but with different binding capabilities to proteoglycans. In addition, this study [64] analyzes Ox-LDL binding to proteoglycans isolated from the intima of human aortas, whereas our study investigated Ox-LDL binding specifically to proteoglycans from the macrophage cell membrane. In the present study, the extent of LDL oxidation was found to affect the ability of Ox-LDL to bind chondroitin sulfate. Mild oxidation of LDL increased its binding to chondroitin sulfate, whereas heavy oxidation reduced LDL binding to CS to levels lower than that of native LDL binding to CS. Similar results were previously reported, showing that mild oxidation of LDL or of VLDL increased their ability to bind to heparan sulfate and lipoprotein [65]. Recently, chondroitin sulfate proteoglycans (CSPGs) secreted by human monocyte-derived macrophages were shown to bind to mildly oxidized LDL but not to extensively oxidized LDL [66], illustrating that the level of lipoprotein oxidation, as well as the method of LDL oxidation, could be of major importance in determining factors which are involved in the interaction between PGs and Ox-LDL. Moreover, when studying binding of lipoproteins to the subendothelial matrix [57] or to the macrophage-derived matrix [67], Ox-LDL was found to bind proteoglycans from the extracellular matrix with a 4-fold greater affinity than native LDL, and this interaction was dependent on the extent of LDL oxidation [57]. In light of these results we can conclude that the increased affinity of Ox-LDL to CS is related to LDL modifications, which take place during early stages of oxidation. Chondroitin sulfate PGs, as opposed to HSPGs, have been shown in the present study to specifically bind Ox-LDL. These results could be related to the struc-

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

tural differences in the chondroitin sulfate and heparan sulfate glycosaminoglycan chains. Glycosaminoglycans are macromolecules which are built of repeated units of disaccharides, that include glucosamine and uronic acid. The chondroitin sulfate sequence is characterized by glucosamine units, which are exclusively N-acetylgalactosamine and a uronic acid fraction containing no iduronic acid but exclusively a glucuronic acid. In addition chondroitin sulfate has been characterized by a lower sulfatation degree than heparan sulfate and this phenomenon is also associated with a lower negative charge density [68]. Finally, we questioned whether under atherogenic conditions, such as cellular lipid peroxidation, the cells can express more chondroitin sulfate and, hence, take up more Ox-LDL and form cholesterol-loaded macrophages. We have previously shown that angiotensin II can induce oxidative damage to macrophages [32]. Administration of angiotensin II to mice has been also shown to induce a 60% increase in the cellular glycosaminoglycan content of their peritoneal macrophages [33]. This increase in macrophage glycosaminoglycans was shown in the present study to be located in the cell membrane and could be related to chondroitin sulfate proteoglycans. Thus, increased OxLDL uptake by oxidized macrophages [32] may be related to increased synthesis of chondroitin sulfate proteoglycans which is induced by oxidative stress. In conclusion, the present study shows that Ox-LDL can specifically bind to cell surface chondroitin sulfate proteoglycans, and thus contribute to enhanced uptake of Ox-LDL by macrophages. Lipid peroxidation of macrophages under oxidative stress can further increase the cellular uptake of Ox-LDL, probably via stimulation of cell surface chondroitin sulfate proteoglycans synthesis. If elevation in macrophage plasma membrane chondroitin sulfate proteoglycans is indeed associated with increased atherogenesis, secondary to the enhanced macrophage uptake of Ox-LDL and foam cell formation, then intervention means to reduce macrophage chondroitin sulfate content should be searched for.

[2] [3] [4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Acknowledgements The authors would like to thank Dr Jeffrey D. Esko, for the CHO cells. This study was supported by a grant from the Israel Science Foundation and by a grant from the Rappaport Family Institute for Research in Medical Sciences, Haifa, Israel.

[16]

[17]

[18]

[19]

References [1] Steinberg D, Pathasarathy S, Carew TE, Khoo JC, Witztum JL.

15

Beyond cholesterol. Modifications of low density lipoprotein that increase its atherogenicity. New Engl J Med 1989;320:915– 24. Witztum JL. The oxidation hypothesis of atherosclerosis. Lancet 1994;344:793 – 5. Aviram M. Modified forms of low density lipoprotein and atherogenesis. Atherosclerosis 1993;98:1 – 9. Aviram M. Interaction of oxidized low density lipoprotein with macrophages in atherosclerosis and the antiatherogenicity of antioxidants. Eur J Clin Chem Clin Biochem 1996;34:599–608. Brown MS, Goldstein JL. Lipoprotein metabolism in the macrophage: implications for cholesterol deposition in atherosclerosis. Annu Rev Biochem 1983;52:223 – 61. Brown MS, Basu SP, Falck JR, Ho YR, Goldstein JL. The scavenger pathway for lipoprotein degradation: specificity of the binding site that mediates the uptake of negatively charged LDL by macrophages. J Supramol Struct 1980;13:67 – 81. Henriksen T, Mahoney EM, Steinberg D. Enhanced macrophage degradation of low density lipoprotein previously incubated with cultured endothelial cells: recognition by receptors for acetylated low density lipoprotein. Proc Natl Acad Sci USA 1981;78:6499– 503. Kaplan M, Williams JK, Mandel H, Aviram M. Role of macrophage glycosaminoglycans in the cellular catabolism of oxidized LDL by macrophages. Arterioscler Thromb Vasc Biol 1998;18:542 – 53. Jackson RL, Busch SJ, Cardin AD. Glycosaminoglycans: molecular properties, protein interactions and role in physiological processes. Physiol Rev 1991;71:481 – 539. Williams KJ, Fless GM, Petrie KA, Bnyder PL, Brocia RW, Swenson TL. Mechanisms by which lipoprotein lipase alters cellular metabolism of LP(a) LDL and nascent lipoproteins. J Biol Chem 1992;267:13284– 92. Mulder M, Lombardi P, Jansen H, Van Berkel TJC, Frants RR, Havekes LM. LDL receptor internalizes LDL and VLDL that are bound to heparan sulfate proteoglycans via lipoprotein lipase. J Biol Chem 1993;268:9369– 75. Eisenberg S, Shayek E, Olivecrona T, Vlodavsky I. Lipoprotein lipase enhances binding of lipoproteins to heparan sulfate on cell surfaces and extracellular matrix. J Clin Invest 1992;90:2013– 21. Hurt E, Camejo G. Effect of arterial proteoglycans on the interaction of LDL with human monocyte-derived macrophages. Atherosclerosis 1987;67:115 – 26. Vijayagopal P, Srinivasan SR, Radhakrishanamurthy B, Berenson GS. Lipoprotein-proteoglycan complexes from atherosclerotic lesions promote cholesteryl ester accumulation in human monocytes macrophages. Atheroscler Thromb 1992;12:237– 49. Wight N. The extracellular matrix and atherosclerosis. Curr Opin Lipidol 1995;6:326 – 34. Rapraeger A, Jalkanen M, Endo E, Koda J, Bernfield M. The cell surface proteoglycans from mouse mammary epithelial cells bears chondroitin sulfate and heparan sulfate proteoglycans. J Biol Chem 1985;260:11046– 52. Williams KJ, Fuki IV. Cell-surface heparan sulfate proteoglycans: dynamic molecules mediating ligand catabolism. Curr Opin Lipidol 1997;8:253 – 62. Hedman K, Christner J, Julkunen I, Vaheri A. Chondroitin sulfate at the plasma membranes of cultured fibroblasts. J Cell Biol 1993;97:1288 – 93. McQuillan DJ, Yanagishita M, Hascall VC, Bickel M. Proteoglycan biosynthesis in murine monocytic leukemic (M1) cells before and after differentiation. J Biol Chem 1989;264:13245– 51.

16

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

[20] Christner JE. Biosynthesis of chondroitin sulfate proteoglycan by P388D1 macrophage-like cell line. Arteriosclerosis 1988;8:535 – 43. [21] Uhlin-Hansen L, Wik T, Kjellen L, Berg E, Forsdahl F, Kolset SO. Proteoglycan metabolism in normal and inflammatory human macrophages. Blood 1993;82:2880–9. [22] Wagner WD, Salisbury BG, Rowe HA. A proposed structure of chondroitin sulfate proteoglycans of human and normal adjacent atherosclerotic plaque. Arteriosclerosis 1988;6:407–17. [23] Bihari-Varga M, Gruber E, Rotheneder M, Zechner R, Kostner GM. Interaction of lipoprotein Lp(a) and low density lipoprotein with glycosaminoglycans from human aorta. Arteriosclerosis 1988;8:851 – 7. [24] Ji ZS, Brecht WJ, Miranda RD, Hussain MM, Innerarity TL, Mahley RW. Role of heparan sulfate proteoglycans in the binding and uptake of apolipoprotein E-enriched remnant lipoproteins by cultured cells. J Biol Chem 1993;268:10160–7. [25] Ji ZS, Faio S, Lee YL, Mahley RW. Secretion-capture role for apolipoprotein E in remnant lipoprotein metabolism involving cell surface heparan sulfate proteoglycans. J Biol Chem 1994;269:2764– 72. [26] Williams KJ, Tabas I. The response to retention hypothesis. Arterioscler Thromb 1995;15:551–61. [27] Hurt Camejo E, Camejo G, Rosengreen B, Lopez F, Ahlstrom C, Fager G, Bondjers G. Effect of arterial proteoglycans and glycosaminoglycans on low density lipoprotein oxidation and its uptake of human macrophages and arterial smooth muscle cells. Arterioscler Thromb 1992;12:569–83. [28] Hurt E, Bondjers G, Camejo G. Interaction of LDL with human arterial proteoglycans stimulates its uptake by human monocytederived macrophages. J Lipid Res 1990;31:443–54. [29] Cardin AD, Weintraub HJ. Molecular modeling of protein-glycosaminoglycan interactions. Arteriosclerosis 1989;9:21–32. [30] Camejo G, Olofson SO, Lopez F, Carlsson F, Carlsson P, Bondjers G. Identification of Apo B-100 segments mediating the interaction of low density lipoproteins with arterial proteoglycans. Arteriosclerosis 1988;8:368–77. [31] Fuhrman B, Oiknine J, Aviram M. Iron induces lipid peroxidation in cultured macrophages increases their ability to oxidatively modify LDL and affects their secretory properties. Atherosclerosis 1994;111:65–78. [32] Fuhrman B, Oiknine J, Keidar S, Ben-Yaish L, Kaplan M, Aviram M. Increased uptake of low density lipoprotein (LDL) by oxidized macrophages is the result of an initial enhanced LDL receptor activity and a further progressive oxidation of LDL. Free Radic Biol Med 1997;23:34–46. [33] Keidar S, Attias J. Angiotensin II injection into mice increases the uptake of oxidized LDL by their macrophages via a proteoglycan-mediated pathway. Biochem Biophys Res Commun 1997;239:63 – 7. [34] Tabas I, Weiland DD, Tall AR. Unmodified low density lipoprotein causes cholesteryl ester accumulation in J-774 macrophages. Proc Natl Acad Sci USA 1985;82:416–20. [35] Aviram M. Plasma lipoprotein separation by discontinuous density gradient ultracentrifugation in hyperlipidemic patients. Biochem Med 1983;30:111–8. [36] Lowry OH, Rosenbrough NJ, Farr JL, Randall RI. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193:265 – 75. [37] Bilheimer DW, Eisenberg S, Levy RI. The metabolism of very low density lipoprotein, I: preliminary in vitro and in vivo observations. Biochim Biophys Acta 1972;260:212–21. [38] Buege JA, Aust SD. Microsomal lipid peroxidation. Methods Enzymol 1978;52:302–10. [39] Swaney JB, Kuehl KS. Separation of apolipoproteins by an acrylamide gradient sodium dodecyl sulfate gel electrophoresis system. Biochim Biophys Acta 1976;446:561–5.

[40] Edwards IJ, Xu H, Obunike JC, Goldberg IJ, Wagner WD. Differentiated macrophages synthesize a heparan sulfate proteoglycan and an oversulfated chondroitin sulfate proteoglycan that bind lipoprotein lipase. Arterioscler Thromb Vasc Biol 1995;15:400 – 9. [41] Van de Lest CH, Versteeg EM, Veerkamp JH, van Kuppevelt TH. A spectrophotometric method for the determination of heparan sulfate. Biochim Biophys Acta 1994;1201:305–11. [42] Owens RT, Wagner WD. Proteoglycans produced by cholesterol-enriched macrophages bind plasma low density lipoprotein. Atherosclerosis 1991;91:229 – 40. [43] El-Saadani M, Esterbauer M, El-Sayed M, Goher M, Nassar AY, Jurgens J. A spectrophotometric assay for lipid peroxides in serum lipoproteins using a commercially available reagent. J Lipid Res 1989;30:627 – 30. [44] Aviram M, Bar M, Maor I. Macrophage localization of unesterified cholesterol affect cellular cholesterol regulation and its cytotoxicity: atherogenicity of excess plasma membrane cholesterol. Bull Mol Biol Med 1995;20:49 – 54. [45] Assman G, Schriewer H, Schmitz G, Hagele EO. Quantification of high-density-lipoprotein cholesterol by precipitation with phosphotungstic acid-MgCl2. Clin Chem 1983;29:2026. [46] Esko JD, Stewart TE, Taylor WH. Animal cell mutants defective in glycosaminoglycans biosynthesis. Proc Natl Acad Sci USA 1985;82:3197 – 201. [47] Aviram M, Dankner G, Brook JG. Platelet secretory products increase LDL oxidation enhance its uptake by macrophages and reduce its fluidity. Arteriosclerosis 1990;10:559 – 63. [48] Halvorsen B, Aas UK, Kulseth MA, Drevon CA, Christiansen EN, Kolset SO. Proteoglycans in macrophages: characterization and possible role in the cellular uptake of lipoproteins. Biochem J 1998;331:743 – 52. [49] Uhlin-Hansen L, Wik T, Kjellen L, Berg E, Forsdahl F, Koslet S. Proteoglycan metabolism in normal and inflammatory human macrophages. Blood 1993;82:2880 – 9. [50] Williams KJ, Fuki I. Cell-surface heparan sulfate proteoglycans: dynamic molecules mediating ligand catabolism. Curr Opin Lipidol 1997;8:253 – 62. [51] Kolset SO, Ivhet I, Overtn A, Nilsson K. Differentiation-associated changes in the expression of chondroitin sulfate proteoglycan in induced U-937 cells. Cancer Res 1988;48:6103 – 8. [52] Tovar AMF, Cesar DCF, Leta GC, Mourao PAS. Age related changes in populations of aortic glycosaminoglycans. Arterioscler Thromb Vasc Biol 1998;18:604 – 14. [53] Murray RK. Glycoproteins and proteoglycans. In: Murray RK, Graner H, Mayes PA, Rodel J, editors. Harper’s Biochemistry. Appleton and Lange, 1990:591 – 608. [54] Hurt Camejo E, Olsson U, Wiklund O, Bondjers G, Camejo G. Cellular consequences of the association of apoB lipoproteins with proteoglycans. Arterioscler Thromb Vasc Biol 1997;17:1011 – 7. [55] Boren J, Olin K, Lee I, Chait A, Wight TN, Innerarity TL. Identification of the principal proteoglycan-binding site in LDL. A single point mutation in apo-B100 severely affects proteoglycan interaction without affecting LDL. J Clin Invest 1998;101:2658– 64. [56] Weisgraber KH, Rall SC. Human apolipoprotein B-100 heparinbinding sites. J Biol Chem 1987;262:11097– 111103. [57] Saxena U, Klein MG, Vanni TM, Goldberg IJ. Lipoprotein lipase increases low density lipoprotein retention by subendothelial cell matrix. J Clin Invest 1992;89:373 – 80. [58] Paanaven K, Kovanen PT. Proteolysis and fusion of low density lipoprotein particles independently strengthen their binding to exocytosed mast cells granules. J Biol Chem 1994;269:2023–31. [59] Camejo G, Rosengren B, Olson U, Lopez F, Olofson SO, Westerlund C, Bondjers G. Molecular basis of the association of arterial proteoglycans with low density lipoproteins: its effect on

M. Kaplan, M. A6iram / Atherosclerosis 149 (2000) 5–17

[60]

[61]

[62]

[63]

[64]

the structure of the lipoprotein particle. Eur Heart J 1990;11:164 – 73. Srinivasan SR, Radhakrishnamurthy B, Vijayagopal P, Berenson GS. Proteoglycans, lipoproteins and atherosclerosis. Adv Exp Med Biol 1991;285:373–81. Chang MY, Lees AM, Lees RS. Low density lipoprotein modification and arterial wall accumulation in a rabbit model of atherosclerosis. Biochemistry 1993;32:8518–24. Siegel G, Malmstem M, Klussendorf D, Leonhardt W, Wood WG. Physicochemical interaction between plasma lipoproteins and endothelial cell membrane and vascular matrix proteoglycans. Abstract of the 11th International Symposium on Atherosclerosis, Paris. Atherosclerosis 1997;134(1):8. Bihari-Varga M, Haynal I. Proteoglycan-lipoprotein interaction. Abstract of the 11th International Symposium on Atherosclerosis, Paris. Atherosclerosis 1997;134(1):8. Oorni K, Pentikainene MO, Annila A, Kovanen PT. Oxidation of low density lipoprotein particles decreases their ability to bind

.

[65]

[66]

[67]

[68]

17

to human aortic proteoglycans: dependence on oxidative modification of the lysine residues. J Biol Chem 1997;272:21303– 11. Makoveichuk E, Lookene A, Olivecrona G. Mild oxidation of lipoproteins increases their affinity for surfaces covered by heparan sulfate and lipoprotein lipase. Biochem Biophys Res Commun 1998;252:703 – 10. Chang M, Olin KL, Tsoi C, Wight TN, Chait A. Human monocyte derived macrophages secrete two forms of proteoglycan-macrophage colony stimulating factor that differ in their ability to bind low density lipoprotein. J Biol Chem 1998;273:15985– 92. Kaplan M, Aviram M. Oxidized LDL binding to a macrophagesecreted extracellular matrix. Biochem Biophys Res Commun 1997;237:271 – 6. Sambandan T, Baker JR, Christner JE, Ekborg SL. Specificity of the low density lipoprotein-glycosaminoglycan interaction. Arterioscler Thromb 1991;11:561 – 8.