Expression of class A scavenger receptor is enhanced by high glucose in vitro and under diabetic conditions in vivo: one mechanism for an increased rate of atherosclerosis in diabetes.

In the early stage of atherosclerosis, macrophages take up chemically modified low density lipoproteins (LDL) through the scavenger receptors, leading to foam cell formation in atherosclerotic lesions. To get insight into a role of the scavenger receptors in diabetes-enhanced atherosclerotic complications, the effects on class A scavenger receptor (SR-A) of high glucose exposure in vitro as well as the diabetic conditions in vivo were determined in the present study. The in vitro experiments demonstrated that high glucose exposure to human monocyte-derived macrophages led to an increased SR-A expression with a concomitant increase in the endocytic uptake of acetylated LDL and oxidized LDL. The endocytic process was significantly suppressed by an anti-SR-A neutralizing antibody. Stability analyses revealed a significant increased stability of SR-A at a mRNA level but not a protein level, indicating that high glucose-induced up-regulation of SR-A is due largely to increased stability of SR-A mRNA. High glucose-enhanced SR-A expression was prevented by protein kinase C and NAD(P)H oxidase inhibitors as well as antioxidants. High glucose-enhanced production of intracellular peroxides was visualized in these cells, which was attenuated by an antioxidant. The in vivo experiments demonstrated that peritoneal macrophages from streptozotocin-induced diabetic mice increased SR-A expression when compared with those from nondiabetic mice. Endocytic degradation of acetylated LDL and oxidized LDL were also increased with these macrophages but not with the corresponding macrophages from diabetic SR-A knock-out mice. These in vitro and in vivo results probably suggest that reactive oxygen species generated from a protein kinase C-dependent NAD(P)H oxidase pathway plays a role in the high glucose-induced up-regulation of SR-A, leading to the increased endocytic degradation of modified LDL for foam cell formation. This could be one mechanism for an increased rate of atherosclerosis in patients with diabetes.

The presence of a massive cluster of macrophage-derived foam cells in situ in the subendothelial spaces is one of the characteristic features in the early stages of atherosclerotic lesions (1). Foam cells produce various bioactive molecules, such as cytokines, growth factors, and proteases, which play an important role in the development and progression of atherosclerotic lesions (1). Macrophages take up chemically modified low density lipoproteins (LDLs), 1 such as oxidized LDL (Ox-LDL) and acetylated LDL (acetyl-LDL) through the scavenger receptors, and transform into foam cells in vitro (2)(3)(4). Several scavenger receptors so far identified include class A scavenger receptor (SR-A) (5), class B scavenger receptor (CD36) (6), class B scavenger receptor type-I (SR-BI) (7), and lectin-like oxidized LDL receptor-1 (LOX-1) (8). However, their contribution to endocytic uptake of modified LDL and subsequent foam cell formation has not been well established. The degradation capacity by peritoneal macrophages obtained from SR-A knockout mice was reduced to 30% for acetyl-LDL and to 50% for Ox-LDL when compared with wild-type macrophages (9). Nozaki et al. (10) using CD36-deficient human monocyte-derived macrophages demonstrated that CD36 accounts for 40% of Ox-LDL uptake. A more recent experiment demonstrated that endocytic uptake of acetyl-LDL and Ox-LDL by macrophages obtained from double knock-out mice of SR-A and CD36 was reduced to 10 -25% as compared with those from wild-type mice, and subsequent accumulation of cholesteryl esters (CEs) (foam cell formation) in macrophages from these double knockout mice was negligibly low (11), suggesting that both SR-A and CD36 could serve as major scavenger receptors for endocytic uptake of modified LDL.
Atherosclerosis-related disorders are serious vascular complications and the major cause of death in diabetic patients; type 2 diabetic patients possess a 2-3-fold higher risk of coronary artery disease (12,13). There is a strong positive correlation between glycated hemoglobin levels and mortality from cardiovascular disease, even in nondiabetic subjects (14), demonstrating that hyperglycemia is one of the risk factors for cardiovascular disease. Five possible links of diabetes to atherosclerosis have been proposed, which include (i) an increase in polyol pathway flux, (ii) an increase in oxidative stress, (iii) an increase in advanced glycation end products (AGE), (iv) activation of protein kinase C (PKC), and (v) an increase in hexosamin pathway flux (15,16). However, molecular mechanisms underlying the increased rate of atherosclerosis in diabetic patients are poorly understood. Griffin et al. (17) recently reported that high glucose induces up-regulation of CD36 in human monocyte-derived macrophages at a translational level, which might well explain a link between diabetes and atherosclerosis because CD36 is one of the scavenger receptors for endocytic uptake of modified lipoproteins and AGE (18). More recently, Li et al. reported that high glucose increases LOX-1 expression in human monocyte-derived macrophages, which enhances their capacity to adhere to endothelial cells (19) and to endocytose Ox-LDL (20), suggesting a role of LOX-1 in form cell formation in atherosclerotic lesions.
Previous studies have established an important role of SR-A in foam cell formation in the atherosclerotic lesions (9,21,22). However, no solid information has been available on its role in diabetes-enhanced atherosclerotic complications. The purpose of the present study was to examine the in vitro effect of high glucose exposure to human monocyte-derived macrophages on SR-A expression. The in vivo effect of diabetic conditions on SR-A expression was also examined in peritoneal macrophages obtained from streptozotocin (STZ)-induced diabetic SR-A knock-out and wild type mice. The results indicate that high glucose or diabetic conditions increase generation of reactive oxygen species (ROS) from PKC-dependent NAD(P)H oxidase and induce SR-A up-regulation, which might lead to an enhanced foam cell formation by monocyte-derived macrophages in the early atherosclerotic lesions. The SR-A up-regulation induced under diabetic conditions could be one mechanism for an increased rate of atherosclerosis in patients with diabetes.
Lipoproteins and Their Modifications-Human LDL (d ϭ 1.019 -1.063 g/ml) was isolated by sequential ultracentrifugation from human plasma of consenting normolipidemic subjects after overnight fasting. LDL was dialyzed against 0.15 mol/liter NaCl and 1 mmol/liter EDTA (pH 7.4). Acetyl-LDL was prepared by chemical modification of LDL with acetic anhydride as described previously (23). To prepare Ox-LDL, LDL was dialyzed against PBS to remove EDTA. LDL (0.1 mg/ml) was then incubated for 20 h at 37°C with 5 M CuSO 4 , followed by the addition of 1 mM EDTA and cooling (23). The levels of endotoxin associated with these lipoproteins were less than 1 pg/g of protein, which were measured by a commercially available kit (Toxicolor system; Seikagaku Corp., Tokyo, Japan) (24). Moreover, cell viability was not affected by endotoxin at concentrations less than 1.0 ng/ml in our experimental system. Acetyl-LDL and Ox-LDL was labeled with 125 I as described by McFarlane (25) to a specific radioactivity of 600 and 800 cpm/ng, respectively.
Monoclonal Anti-SR-A Antibodies-The monoclonal antibodies to SR-A were raised in SR-A knock-out mice by immunizing the recombinant human SR-A corresponding to amino acid sequences 131-451 (26). Among five antibodies obtained, two antibodies (E5 and C6) were used in the present study. E5 having a cross-reactivity to SR-A from human, monkey, cat, rabbit, rat, and mouse was used for immunoblot analyses of SR-A in human monocyte-derived macrophages. In contrast, the immunoreactivity of C6 is restricted to human SR-A, and our preliminary experiments have revealed that the endocytic degradation of 125 I-acetyl-LDL and 125 I-Ox-LDL by human monocytederived macrophages is effectively inhibited by this antibody, dem-onstrating that C6 is a first-time developed anti-SR-A-neutralizing antibody. Therefore, C6 was used as an anti-SR-A-neutralizing antibody in the present study. 2F8 purchased from Wako Pure Chemical Industries (Osaka, Japan) was used for immunoblot analyses for SR-A in mouse peritoneal macrophages.
Animals-The experimental protocol was approved by the Ethics Review Committee for animal experimentation of our institution. Mice lacking SR-A were established from C57BL/6 by targeted disruption of exon 4 of the SR-A gene, which is essential for the formation of functional trimeric receptors in A3-1 ES cells according to Suzuki et al. (9). These mice (5 weeks old) were maintained in a temperature-and humidity-regulated room (22 Ϯ 2°C, 55 Ϯ 2%) with controlled lightening (12 h light/dark cycle). They had free access to tap water and commercial regular chow for a week before receiving an STZ injection and throughout the experiments. After overnight fasting, mice were injected with STZ (100 mg/kg) into the peritoneal cavity in 50 mM citrate-buffered saline (pH 4.5) to induce diabetes under light anesthesia with ether. Control mice were injected with vehicle (citrate-buffered saline). Two weeks after injection of STZ, fasting blood glucose elevated to 350 Ϯ 45 mg/dl (19.8 mM) (n ϭ 30) from 90 Ϯ 10 mg/dl (5.0 mM) (n ϭ 25).
Cell Culture-Human peripheral mononuclear cells were isolated from blood of healthy volunteers by Ficoll density gradient centrifugation (Ficoll-Paque from Amersham Biosciences). Monocytes were purified according to a modified method of Connor et al. (27). Purified monocytes were suspended in RPMI 1640 at 2 ϫ 10 6 cells/ml and seeded onto 6-cm dishes (2 ϫ 10 6 ) or 10-cm dishes (1 ϫ 10 7 ) (BD Biosciences PRIMARIA, Tokyo, Japan). After incubation for 1 h for adherence, the medium was replaced by RPMI 1640 with indicated glucose or mannitol concentrations and supplemented with 10% pooled human serum, streptomycin (0.1 mg/ml), and penicillin G (100 units/ml). Cells were incubated for 7 days to differentiate into macrophages (the medium was replaced every 3 days). Differentiation of monocytes into macrophages (human monocyte-derived macrophages) was judged after 7 days of incubation by three categories, such as the adherence to the culture plates, the morphological features as mononuclear cells after Giemsa staining, and the capacity to take up carbon particles (28). Under these conditions, the cells contained Ͼ95% macrophages and were Ͼ98% viable as determined by trypan blue staining. These human monocytederived macrophages were used for each experiment. Thioglycolateelicited macrophages were collected from STZ-induced diabetic C57BL/6J mice or SR-A knock-out mice 4 days after intraperitoneal injection of 1.5 ml of thioglycolate (29). The collected peritoneal cells were suspended in RPMI 1640 medium supplemented with 10% heatinactivated fetal bovine serum, 0.1 mg/ml streptomycin, and 100 units/ml of penicillin and seeded onto 12-well plates (1 ϫ 10 5 cells). After a 90-min incubation, nonadherent cells were removed by triplicate washing with PBS. These cells were judged by the three categories described above and used for each experiment. All cellular experiments were performed at 37°C in a humidified atmosphere of 5% CO 2 in air.
Cellular Assay for Endocytic Uptake of 125 I-acetyl-LDL or 125 I-Ox-LDL-Human monocyte-derived macrophages or mouse peritoneal exudate macrophages (1 ϫ 10 6 cells) were seeded to a 22-mm plastic culture dish and incubated for 1 h. The monolayers thus formed were washed three times with 1.0 ml of PBS. Each well was incubated with the indicated concentrations of 125 I-acetyl-LDL or 125 I-Ox-LDL for 5 h in the absence or presence of an excess amount (200 g/ml) of the unlabeled acetyl-LDL or Ox-LDL. Endocytic degradation was determined by trichloroacetic acid-soluble radioactivity in the medium as described previously (30). Cells were solubilized with 1.0 ml of 0.1 N NaOH, and the cell-associated radioactivity was determined (30).
Assay for Foam Cell Formation (CE Accumulation)-Cell monolayers were incubated with acetyl-LDL for 16 h in the presence of 0.1 mM [ 3 H]oleate conjugated with bovine serum albumin, and cellular lipids were extracted for determination of radioactivity of cholesteryl-[ 3 H]oleate as described previously (31).
Immunoblot Analyses-Human monocyte-derived macrophages were solubilized with 1% Triton X-100, and was pretreated with boiling for 3 min in 2% SDS and 2-mercaptoethanol, followed by incubation for 24 h at 37°C with 3 units of N-glycosidase (Roche Applied Science). These samples were run on 4 -20% gradient SDS-polyacrylamide gels, followed by electrophoretic transfer to nitrocellulose membranes (Millipore Corp., Bedford, MA). The membranes were exposed to an anti-human SR-A antibody (E5) and visualized by horseradish peroxidase-conjugated anti-mouse IgG antibody with ECL Western blotting detection reagent (Amersham Biosciences) (26). This membrane was reblotted with an anti-␤-actin antibody (Sigma) as an internal control. The molecular size of SR-A detected by this immunoblot was ϳ50 kDa, SR-A expression was normalized by ␤-actin contents and expressed by SR-A protein/␤-actin in each figure. Mouse peritoneal macrophages were pretreated at room temperature with 2% SDS without 2-mercaptoethanol, run on 4 -20% gradient SDS-polyacrylamide gels, and transblotted, followed by reaction with the anti-mouse SR-A antibody (2F8). The membrane was also reblotted with an anti-␤-actin antibody (Sigma) as an internal control. The molecular size detected in this system was ϳ250 kDa, representing a trimer of monomeric SR-A (Fig.  8A). Amounts of SR-A expression were similarly corrected by ␤-actin expression. The peritoneal macrophages from SR-A knock-out mice were also subjected to parallel immunoblot analyses, the band corresponding to SR-A was not detected as expected (data not shown). The density of bands was measured with Image Gauge software in LAS-1000plus (FUJIFILM, Tokyo, Japan).
Northern Blot Analyses-SR-A cDNA inserted into pGEM-T was amplified and used for preparation of digoxygenin-labeled antisense RNA probes according to the manufacturer's instructions (digoxygenin Northern starter kit from Roche Diagnostics). Total RNA from macrophages (2 ϫ 10 6 cells, 10 cm in diameter; BD Biosciences) was extracted with TRIzol (Invitrogen). Ten g/lane of total RNA was fractionated by electrophoresis through a denaturing formaldehyde 2% agarose gel, transferred to Hybond-Nϩ nylon membrane by capillary transfer with 10ϫ SSC for 20 h, and then cross-linked by UV (FS 1500; Funakoshi, Tokyo, Japan). The membrane was hybridized with the heat-denatured RNA probe (100 ng/ml) and further incubated for 30 min with the anti-digoxygenin antibody conjugated with alkaline phosphatase. SR-A mRNAs were visualized using CDP-Star (Roche Applied Science) chemiluminescence substrate as described previously (27). Density of bands was measured with an Image Gauge software in LAS-1000plus (FUJIFILM).
Flow Cytometric Analyses of Peroxide Production-Human monocyte-derived macrophages cultured under various conditions for 5 days were washed with the culture medium and further incubated for 15 min at 37°C with 5 M CDCFHDA(AM) in a 6-cm dish. The cells were collected using cell scraper into microtubes with 5% fetal calf serum-PBS. After keeping on ice for 5 min, 6-carboxy-2,7-dichlorofluorescein (CDCFH) oxidation was measured using a flow cytometer (FACSVantage TM ; BD Biosciences). CDCFHDA(AM) is a nonpolar compound that readily diffuses into cells, where it is hydrolyzed by cytosolic enzymes to the nonfluorescent polar derivative CDCFH, which is trapped within the cell. CDCFH is rapidly oxidized to highly fluorescent product in the presence of intracellular hydroperoxides.
Determination of SR-A mRNA and Protein Stability-After 7 days' incubation of human monocyte-derived macrophages with 5.5 or 27.5 mM glucose, actinomycin D (3 g/ml) was added to block transcription, and total RNA was collected from the cells at various time points. For quantifying SR-A and glyceraldehyde-3-phosphate dehydrogenase transcript, the LightCycler System (Roche Applied Science) was used (32). PCRs were performed using SYBR Green I master mix and specific primer for human SR-A. To assess the specificity of the amplified PCR products, a melting curve analysis was performed after the final cycle. In the case of a stability assay for SR-A protein, cycloheximide (10 g/ml) was added, and whole cell lysates were harvested at various time points, followed by immunoblot analyses using anti-SR-A antibody (E5) to quantify the expression of SR-A protein as described above.
Statistical Analysis-All data were expressed as mean Ϯ S.D. Differences between groups were examined for statistical significance using Student's t test. A p value less than 0.05 denoted the presence of a statistically significant difference.

High Glucose Promotes the Expression of SR-A-Immunoblot
analysis showed that SR-A expression was increased during differentiation of human monocytes into macrophages; its expression was obvious at day 3 and reached a maximal level at day 7 (Fig. 1A). When the incubation time was fixed at day 7, SR-A expression was enhanced by glucose in a dose-dependent manner and reached a plateau level at 16.5 mM glucose, whereas mannitol as an osmolality control did not enhance the expression of SR-A (Fig. 1B). The expression of SR-A was increased by high glucose at 27.5 mM (Fig. 1B), whereas high glucose affected the expression of neither galectin-3, a differentiation marker for macrophages as well as one of AGEreceptors ( Fig. 2A), nor SR-BI (data not shown), indicating that a high glucose-induced increase in the expression was selective for SR-A. Northern blot analysis showed that high glucose induced an increase in the amount of mRNA of SR-A, type I and type II, suggesting that SR-A expression was enhanced at a mRNA level (Fig. 2B).
Effect of High Glucose on the Stability of SR-A Protein and mRNA-Cycloheximide was used to determine whether high glucose altered the translation of SR-A protein in human monocyte-derived macrophages. Although the initial SR-A protein level treated with high glucose was 2-fold higher than that treated with normal glucose, the decay rate in the protein level after cycloheximide addition was indistinguishable from each other (Fig. 2C), excluding the possibility that high glucose increased the stability of SR-A protein. Real time PCR analyses showed that the decay of SR-A mRNA after the actinomycin D addition was significantly slower in the high glucose-treated cells than that in control cells (half-life: 36 h versus 15 h, respectively) (Fig. 2D), suggesting that high glucose increased the stability of SR-A transcript. It seems likely therefore that the increased stability of SR-A mRNA may play an important role in high glucose-enhanced SR-A protein expression.
FIG. 1. Effect of glucose on SR-A expression in human monocyte-derived macrophages. A, human monocyte-derived macrophages were incubated with 5.5 mM glucose for the indicated times. The cells were harvested and subjected to immunoblot analyses with an anti-human SR-A antibody (E5) as described under "Experimental Procedures." B, the cells were incubated for 7 days with the indicated glucose concentrations (closed column) or 5.5 mM glucose with the indicated mannitol concentrations (hatched column). The cells were harvested and subjected to immunoblot analyses as described above. The transblotted membrane was reblotted with an anti-␤-actin antibody as an internal control. Bands of SR-A and ␤-actin were subjected to densitometric analyses. Amounts of SR-A were normalized by ␤-actin contents (lower panel) as described under "Experimental Procedures." Experiments were repeated three times with almost identical results.
Intracellular Signal Pathway in High Glucose-enhanced SR-A Expression-Several inhibitors were tested to elucidate the signal pathway(s) involved in the high glucose-enhanced SR-A expression. The SR-A expression enhanced by high glucose (27.5 mM) was effectively inhibited either by calphostin C as PKC inhibitor, wortmannin as phosphatidylinositol-3 kinase (PI3K) inhibitor, or NAC as antioxidant (Fig. 3A). Under the identical conditions, PD98059 as inhibitor for mitogen-activated protein kinase/extracellular signal-regulated kinase kinase suppressed the expression of SR-A, whereas SB203580 as p38 mitogen-activated protein kinase inhibitor had no effect (Fig. 3B). These findings indicate the involvement of PKC, PI3K, mitogen-activated protein kinase/extracellular signalregulated kinase kinase, and intracellular oxidation in high glucose-enhanced SR-A expression.
The role of ROS in particular was further pursued in the present study. High glucose-enhanced SR-A expression was significantly inhibited by apocynin as NAD(P)H oxidase inhibitor (Fig. 3C). Insulin, which was reported to enhance NAD(P)H oxidase activity (33), significantly increased the expression of SR-A under normal glucose, but not under high glucose conditions (Fig. 3C). High glucose-enhanced SR-A expression was also inhibited by superoxide dismutase mimetic, MnTBAP (Fig.  3D). Under the present conditions, all reagents tested, such as calphostin C, wortmannin, NAC, PD98059, SB203580, insulin, apocynin, and MnTBAP, had no effect on the expression of galectin-3 (data not shown). These results suggested that an intracellular signal system(s) related to ROS via NAD(P)H oxidase could be operative. Finally, the generation of ROS was visualized in cultured human monocyte-derived macrophages with flow cytometric analysis using CDCFHDA(AM). Intracellular ROS production was increased 1.5-fold by high glucose " Experiments were repeated three times with almost identical results. *, p Ͻ 0.05. C, the cells were incubated for 7 days with 5.5 mM glucose (normal glucose) or 27.5 mM glucose (high glucose). The cells were then treated with cycloheximide (10 g/ml) and harvested at different time points. The cells were subjected to immunoblot analyses for SR-A and amounts of SR-A expression was normalized by ␤-actin as described under "Experimental Procedures." Experiments were repeated three times with almost identical results. D, the cells were incubated for 7 days with 5.5 mM glucose (normal glucose) or 27.5 mM glucose (high glucose). The cells were then treated with actinomycin D (3 g/ml) and harvested at different time points. Total RNA was isolated, and SR-A mRNA was quantified by a real time reverse transcription-PCR. Results were normalized to the amount of glyceraldehyde-3-phosphate dehydrogenase measured in parallel. Values expressed as ratios to SR-A mRNA at the time of actinomycin D addition are the mean of four independent donors. Significant differences in amounts of SR-A mRNA between high glucose and normal glucose were observed 1, 6, and 24 h after actinomycin D addition as shown by asterisks (*, p Ͻ 0.05). The inset shows the bar graph of these three time points. Experiments were repeated three times with almost identical results.
incubation when compared with normal glucose (Fig. 4, A and  B), which was reduced to a basal level with coincubation with NAC ( Fig. 4A).
High Glucose Promotes Endocytic Uptake and Foam Cell Formation-We next determined effects of high glucose on SR-A functions such as endocytic uptake of acetyl-LDL and subsequent CE accumulation. Amounts of cell association of 125 I-acetyl-LDL with human monocyte-derived macrophage treated with high glucose were significantly higher than the cells treated with normal glucose, whereas mannitol medium exhibited a level similar to normal glucose (Fig. 5A). Amounts of 125 I-acetyl-LDL degraded were increased dose-dependently. However, the degradation capacity of the cells treated with high glucose was 2.5 times greater than that by the cells treated with normal glucose (Fig. 5B), indicating that the high glucose-induced increase in SR-A expression was accompanied by an increase in its function. This was also the case with CE accumulation that occurred after endocytic uptake of acetyl-LDL by these macrophages. When human macrophages treated with high glucose were incubated with 40 g/ml acetyl-LDL for 16 h, CE accumulation was 2.8 nmol/mg of protein, which was 1.5 times higher than that obtained from cells treated with normal glucose, whereas cells treated with man-nitol showed a CE accumulation capacity similar to that of cells treated with normal glucose (Fig. 6). CE accumulation of the cells treated either with high glucose, normal glucose, or mannitol was negligible when these cells were not incubated with acetyl-LDL (nonload) (Fig. 6). These results clarify that treatment of human monocytes with high glucose leads to up-regulation of SR-A functions such as endocytic uptake of acetyl-LDL and subsequent CE accumulation.
We next determined the contribution of SR-A to degradation of acetyl-LDL by human monocyte-derived macrophages by using a recently developed anti-SR-A-neutralizing antibody (26). Degradation of acetyl-LDL by human macrophages treated with normal glucose was inhibited by 74% by the antibody, whereas nonimmune mouse IgG had no effect (Fig. 7A). The inhibitory effect of this antibody was also prominent for degradation by high glucose-treated macrophages; amounts of 125 I-acetyl-LDL degraded by high glucose-treated cells were reduced by 80% by the antibody (Fig. 7A). Macrophages treated with high glucose also showed a significant increase in their endocytic degradation of Ox-LDL; amounts of Ox-LDL degraded by high glucose-and normal glucose-treated cells were reduced by 61 and 40%, respectively, by the antibody, whereas nonimmune mouse IgG had no effect (Fig. 7B). These results strongly suggest that SR-A serves as a major contributor to the endocytic uptake of acetyl-LDL and Ox-LDL by high glucosetreated human monocyte-derived macrophages.

Peritoneal Macrophages from Diabetic Mice Increase SR-A Expression, Acetyl-LDL Degradation, and CE Accumulation-
The high glucose-induced SR-A up-regulation is supported by the results of the above experiments in vitro. However, the most important issue is whether this phenomenon does occur in vivo. Peritoneal macrophages from STZ-induced diabetic mice were compared with those from nondiabetic mice in their SR-A expression, endocytic uptake of acetyl-LDL, and CE accumulation. Immunoblot analyses showed that the expression of SR-A in peritoneal macrophages obtained from diabetic mice was significantly higher than that from normal nondiabetic mice (Fig. 8A). Degradation of acetyl-LDL by the macrophages from diabetic mice was significantly higher than that from control mice (Fig. 8B). Moreover, acetyl-LDL-induced CE accumulation by these macrophages from diabetic mice was 2 times greater than that from nondiabetic mice (Fig. 8C). It is clear from these results that SR-A up-regulation does occur under diabetic conditions in vivo. Fig. 8 clearly demonstrated that diabetic conditions increased the SR-A expression of macrophages and their endocytic uptake and CE accumulation. To determine the contribution of SR-A to the diabetes-enhanced endocytic uptake of acetyl-LDL and Ox-LDL, peritoneal macrophages obtained from wild-type and SR-A knock-out mice were used. Degradation of acetyl-LDL by macrophages from diabetic wild-type mice was increased 1.5-fold when compared with that by macrophages from nondiabetic wild-type mice (a versus c in Fig. 9A, p Ͻ 0.01). In contrast, degradation for acetyl-LDL by peritoneal macrophages from diabetic SR-A knock-out mice was not significantly higher than that by macrophages from nondiabetic SR-A knock-out mice (b versus d in Fig. 9A). Furthermore, acetyl-LDL degradation by macrophages from diabetic SR-A knock-out mice was significantly lower (by 64%) than that by macrophages from diabetic wild-type mice (c versus d in Fig. 9A; p Ͻ 0.01). We also conducted the same experiment with Ox-LDL. Degradation of Ox-LDL by macrophages from diabetic wild-type mice was significantly higher than that from nondiabetic wild-type mice (a versus c in Fig.  9B; p Ͻ 0.04). Although Ox-LDL degradation by peritoneal macrophages from diabetic SR-A knock-out mice was slightly higher (but significant) than that by macrophages from nondiabetic SR-A knock-out mice (b versus d in Fig. 9B; p Ͻ 0.003), degradation of Ox-LDL by macrophages from diabetic SR-A knock-out mice was significantly lower than that by macrophages from diabetic wild-type mice (c versus d in Fig. 9B; p Ͻ 0.04). These results taken together support the notion that SR-A plays a crucial role in diabetes-enhanced endocytic uptake of modified LDL by macrophages. DISCUSSION The present study provided in vitro and in vivo evidence for the high glucose-induced SR-A up-regulation. The in vitro experiments with human monocyte-derived macrophages demonstrated the high glucose-enhanced expression of SR-A (Figs. 1  and 2), which was accompanied by a parallel increase in the endocytic degradation of acetyl-LDL and Ox-LDL (Fig. 5) that was effectively inhibited by an anti-SR-A-neutralizing antibody (Fig. 7), suggesting that SR-A is responsible for a high glucoseinduced increase in macrophage functions such as endocytic degradation of modified LDLs and subsequent intracellular CE accumulation (foam cell formation). The stability analysis clarified that glucose-induced up-regulation of SR-A is due mainly to increased stability of SR-A mRNA. This phenomenon was suppressed either by apocynin, a specific inhibitor for NAD(P)H oxidase, calphostin C, a PKC inhibitor, or NAC, a trapping reagent for ROS (Fig. 3) and fluorescence-activated cell sorting analyses visualized increased ROS production in high glucose-treated cells (Fig. 4), providing strong evidence for a crucial role of ROS generated from a PKC-dependent NAD(P)H oxidase pathway in high glucose-induced SR-A upregulation. The in vivo experiments disclosed the increased SR-A expression in peritoneal macrophages from STZ-induced diabetes mice with a concomitant increase in endocytic capacity for acetyl-LDL and subsequent CE accumulation (Fig. 8). However, the diabetes-enhanced degradation of acetyl-LDL did not occur to peritoneal macrophages from diabetic SR-A knock-out mice (Fig. 9). Therefore, the present study has clarified that the SR-A up-regulation induced by high glucose in vitro also occurs in diabetic conditions in vivo.

Role of SR-A in Diabetes-induced Up-regulation of Endocytic Uptake of Modified LDL by Peritoneal Macrophages-The results in
A recent study by Griffin et al. (17) demonstrated that CD36 expression in carotid artery and coronary artery was higher in diabetic patients than in nondiabetic patients in vivo and that the CD36 expression in human monocyte-derived macrophages in vitro was enhanced by high glucose at the translational level (17). Since it is known that CD36 recognizes Ox-LDL (7) and AGE-proteins as efficient ligands (34) and that LDL oxidation (35,36) and formation of AGE proteins (15) are enhanced in diabetic patients, high glucose-induced CD36 up-regulation might enhance the development of atherosclerotic complications in diabetes. Since the present study also clarified the up-regulation of SR-A in high glucose-treated macrophages in vitro and peritoneal macrophages in diabetic mice in vivo, it is interesting to speculate upon the contribution of CD36 and SR-A to high glucose-enhanced endocytic uptake by human monocyte-derived macrophages. Although no direct comparison between SR-A and CD36 is available, the present study has provided some information relevant to this question. The endocytic degradation of acetyl-LDL by human monocyte-derived macrophages treated either with high glucose or normal glucose was inhibited by 74 -80% by the anti-SR-A-neutralizing antibody (Fig. 7A). The inhibitory effect of the same antibody on the endocytic degradation of Ox-LDL was also significant: 40% for normal glucose-treated cells and 61% for high glucosetreated cells (Fig. 7B). Furthermore, The degradation capacity of acetyl-LDL and Ox-LDL by peritoneal macrophages from diabetic SR-A knock-out mice was reduced by 64 and 40%, respectively, when compared with that from diabetic wild-type mice (Fig. 9). These findings do not rule out the contribution of CD36, but probably suggest that SR-A is a main contributor to the increased endocytic capacity of macrophages in diabetes. Further studies are needed to determine the exact contribution of CD36 to diabetes-induced up-regulation of macrophages in vivo.
Li et al. recently reported that high glucose enhanced the LOX-1 expression in human aortic endothelial cells (20). A report from the same group (19) also revealed that high glucose exposure to human monocyte-derived macrophages led to the increased LOX-1 expression. High glucose-induced increase in the endocytic function, which was determined by a semiquantitative method using a fluorescence-labeled ligand (DiI-Ox-LDL) was significantly but partially inhibited by an anti-LOX- against SR-A and CD36 and the quantitative assay for foam cell formation. However, our preliminary experiment showed that the anti-LOX-1-neutralizing antibody (donated from Dr. Sawamura) did not show any significant inhibition for the endocytic degradation of 125 I-Ox-LDL by high glucose-and normal glucose-treated human monocyte-derived macrophages, under which the anti-SR-A antibody inhibited the degradation of Ox-LDL by 61 and 40%, respectively (Fig. 7B). 2 The present study clearly demonstrated that high glucose enhanced the expression of SR-A protein by stabilizing SR-A mRNA (Fig. 2D), leading to an increase in endocytic degradation of modified LDL. In contrast, high glucose-induced CD36 expression in human monocyte-derived macrophage is reported to be due mainly to an increased translation efficiency of CD36 mRNA dependent on ribosomal reinitiation that results in the increased expression of CD36 protein (17), indicating that the regulatory mechanism for the expression of SR-A apparently differs from that for CD36. Previous studies also demonstrated the increase in the stability of SR-A. Sjef et al. (37) reported that ROS stimulate the expression of SR-AI but not SR-AII in human monocyte cell line THP-1 in which hydroxyl radicals produced from superoxide anions and hydrogen peroxide in the presence of free iron contribute to an increased SR-A activity by stabilizing MSR-AI mRNA. Fitzgerald et al. (38) also reported that exposure to lipopolysaccharide causes J774A.1 and RAW264.7 mouse macrophage cell lines to increase SR-A transcript by 3-and 5-fold, respectively. Lipopolysaccharide induction does not increase SR-A gene transcription or affect alternative transcript splicing but mildly increases the stability of mature transcript. The increase in mRNA stability is also known to occur to natural resistance-associated macrophage protein-1 in macrophage cell line, RAW 264.7 in which the activation of mitogen-activated protein kinase increases the stability of natural resistance-associated macrophage protein-1 mRNA through an oxidant-mediated signaling pathway (39).
Effects of several inhibitors on high glucose-enhanced SR-A expression were examined to elucidate a signaling pathway from glucose to SR-A expression. High glucose-enhanced SR-A expression was suppressed by several inhibitors such as PI3K, PKC, NAD(P)H oxidase, and antioxidant (Fig. 3, A and C). Phosphorylated PI3K is known to increase the translocation of glucose transporter-4 from an intracellular compartment to the plasma membrane (40 -43). Since a PI3K inhibitor inhibited high glucose-enhanced SR-A expression (Fig. 3A), glucose transporter-4-mediated glucose uptake could modulate SR-A expression. Since high glucose-enhanced SR-A expression was also inhibited by antioxidant, we next measured the involvement of ROS generated from NAD(P)H oxidase, because high glucose-induced activation of PKC-dependent NAD(P)H oxidase was already shown by Inoguchi et al. (44). Inhibitors for PKC and NAD(P)H oxidase resulted in significant suppression of high glucose-enhanced SR-A expression (Fig. 3, A and C). However, insulin, which was known to enhance NAD(P)H oxidase activity, increased SR-A expression in normal glucose but not in high glucose (Fig. 3C). Therefore, it seems likely that ROS generated from PKC-dependent NAD(P)H oxidase pathway are involved in the high glucose-enhanced SR-A expression. This result is consistent with the previous report that ROS enhance the SR-A expression in human monocyte-derived macrophages (45). In addition, the effective suppression of high glucose-enhanced SR-A expression by extracellular signal-regulated kinase inhibitor (Fig. 3B) also indicates a role of extracellular signal-regulated kinase. Although an exact mechanism for an intracellular signaling pathway to the increase in the stability of SR-A mRNA is not known, a recent study has characterized the intracellular signaling pathway in high glucose-enhanced LOX-1 expression in human monocyte-derived macrophages, where high glucose induces cells to generate ROS, which stimulates NF-B and AP-1 through PKC and mitogen-activated protein kinase pathways (19). Since an NF-B inhibitor had no effect on our system, 2 the intracellular signaling pathway for high glucose-induced SR-A up-regulation could differ from that for high glucose-induced LOX-1 up-regulation, although sharing some part by both pathways is possible. It would be safer at present to propose the notion that high glucose induces these macrophages to stimulate NAD(P)H oxidase to generate ROS, which leads to the increased stability of SR-A mRNA through PKC, PI3K, and extracellular signalregulated kinase pathways. Further detailed studies are needed to clarify this issue.
AGE-modified proteins, which are formed under high glucose conditions, were reported to be involved in the development of diabetic vascular complications (18). Since incubation of AGE proteins with human monocyte-derived macrophages led to enhancement of SR-A expression at an mRNA level (46), it is possible that SR-A expression might be enhanced in the present study by AGE-proteins formed during 7 days' culture. However, bovine serum albumin that was incubated with 27.5 mM glucose for 7 days in a cell-free system did not enhance the expression of SR-A in human monocyte-macrophages. 2 This suggests that interaction of AGE proteins with these macrophages might not lead to the high glucose-induced SR-A upregulation in vitro, although a meaningful contribution of AGEproteins to SR-A up-regulation in vivo is not excluded.