Scavenger Receptor, Class B, Type I-dependent Stimulation of Cholesterol Esterification by High Density Lipoproteins, Low Density Lipoproteins, and Nonlipoprotein Cholesterol*

Scavenger receptor, class B, type I (SR-BI) is a cell surface glycoprotein that mediates selective uptake and efflux of sterols from high density lipoproteins (HDL) to cells. A Chinese hamster ovary cell line that is deficient in functional LDL receptors, but has high expression levels of recombinant SR-BI (ldlA7-SR-BI), was used to examine the effect of SR-BI on the trafficking of sterols between lipoproteins and cells. To monitor the fate of sterols transported by SR-BI into cells, we measured the incorporation of [14C]oleate into cholesterol esters by acyl-CoA:cholesteryl acyltransferase in the endoplasmic reticulum. We show that incubation of ldlA7-SRBI cells with either LDL or HDL resulted in an equally dramatic increase in the formation of [14C]oleate-labeled cholesterol esters. The lipoprotein-stimulated, SR-BI-dependent increase in cholesterol esterification was inhibited by chloroquine. The uptake of sterols and their incorporation into cholesterol esters by SR-BI from LDL was largely a selective process. The addition of free cholesterol to ldlA7-SRBI cells also stimulated cholesterol ester formation in a chloroquine-sensitive fashion. We also show that SR-BI mediates the efflux of endogenously synthesized sterols from the cell membrane. From these studies we conclude that, in the absence of the LDL receptor, overexpression of SR-BI can mediate significant transport of sterols between lipoproteins and the endoplasmic reticulum of cells.

Scavenger receptor, class B, type I (SR-BI) 1 is an ϳ82-kDa, palmitylated, cell surface glycoprotein that binds high density and low density lipoproteins (HDL and LDL) with high affinity (1)(2)(3). Expression of SR-BI in cultured cells is associated with the selective transfer of lipid, but not apoproteins from HDL to cells (1). SR-BI is expressed at high levels in the liver and in the major steroid-producing regions of the adrenal gland (zona fasiculata and zona reticularis), ovary (corpus luteum), and testes (interstitial cells) of rodents (4 -6). Levels of SR-BI are regulated by adrenocorticotrophic hormone (5-7) and human chorionic gonadotropin (4), which stimulate synthesis of corticosteroids by the adrenal gland and testosterone by the testes, respectively. This tissue pattern of expression and regulation of SR-BI in rodents is consistent with SR-BI being the functionally well characterized HDL receptor that mediates uptake of cholesterol esters in steroidogenic tissues and the liver (8 -11). The phenotype of mice lacking functional SR-BI (SR-BI Ϫ/Ϫ mice) confirms the central role of this receptor in HDL metabolism of rodents; SR-BI Ϫ/Ϫ mice accumulate large, cholesterol ester-rich HDL particles in their plasma and have a greater than 2-fold increase in plasma HDL-cholesterol levels (12).
The mechanism by which SR-BI delivers cholesterol esters to cells is unknown, but it appears different from the well characterized LDL receptor endocytic pathway (13). LDL binds to LDL receptors, which are clustered in coated pits on the cell surface. The lipoprotein particles, together with the receptors, are taken up and delivered to endosomes, where the receptors and ligands dissociate, and the receptors are recycled to the cell surface. LDL particles are then transported to lysosomes, where the protein components of the lipoproteins are degraded and the cholesterol esters are hydrolyzed by acid lipase. Free cholesterol released from the particle is transported across the lysosomal membrane, a process that requires NPC1, the gene defective in the lysosomal storage disease Nieman-Pick, type C (14). The lipoprotein-derived cholesterol that enters the endoplasmic reticulum (ER) is re-esterified by acyl-CoA:cholesteryl acyltransferase (ACAT) and stored as cholesterol esters in lipid droplets within the cell (15).
In contrast to the LDL receptor, SR-BI is concentrated in caveolae, which are cholesterol-and sphingomyelin-rich microdomains of the cell surface membrane (3,16). HDL binds to SR-BI with high affinity, and this binding is associated with the uptake of cholesterol esters from the lipoprotein without the concomitant degradation of apolipoproteins, by a mechanism that remains poorly understood (1). The fate of the lipids that are transported by SR-BI into cells has not been clearly delineated. Specifically, it has not been shown whether lipoprotein-derived lipids transported by SR-BI are substrates for ACAT.
In this study we have examined the ligand-binding characteristics of SR-BI and determined if sterols internalized by SR-BI are transported from the cell surface to the ER for esterification by ACAT (15). We show that SR-BI expression is associated with a dramatic increase in lipoprotein-stimulated cholesterol esterification. We also show that LDL, which binds SR-BI with high affinity (2,17), is as effective as HDL in stimulating cholesterol esterification in an SR-BI-dependent manner in cells lacking a functional LDL receptor. The SR-BImediated uptake of lipids from LDL is largely selective and is * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  not associated with significant degradation of the protein components of the lipoprotein particles. We also show that SR-BI expression allows free, nonlipoprotein-associated cholesterol to enter cells and to stimulate cholesterol esterification. Thus, SR-BI-mediated cholesterol uptake is not limited to cholesterol contained in lipoproteins. The SR-BI-mediated stimulation of cholesterol esterification by nonlipoprotein cholesterol as well as by LDL and HDL required the maintenance of acidic compartments in the cell; esterification is inhibited by the addition of chloroquine, whether the cholesterol is provided as free cholesterol or as part of a lipoprotein particle. Finally, we demonstrate that SR-BI mediates the efflux of endogenously synthesized sterols to LDL as well as to HDL.

EXPERIMENTAL PROCEDURES
Cell Lines and Tissue Culture-The mouse SR-BI cDNA (kindly provided by Monty Krieger, Massachusetts Institute of Technology, Cambridge, MA) was subjected to restriction digestion using XhoI and HindIII and subcloned into pcDNA3 (Invitrogen, Carlsbad, CA). Chinese hamster ovary (CHO) cells that lack LDL receptor activity (ldlA7 cells) (18) were stably transfected with pcDNA3-SRBI (ldlA7-SRBI) using the MBS mammalian transfection kit (Stratagene, La Jolla, CA). Clones were screened for expression of SR-BI employing immunoblot analysis using polyclonal rabbit antipeptide antibodies directed against the last 14 amino acids of the COOH terminus of murine SR-BI (IgG-SRBI). A subclone of the cell line with the highest expression of immunodetectable SR-BI (ldlA7-SRBI) and two other independently isolated subclones (ldlA7-SRBIa and ldlA7-SRBIb) were selected for further study. Immunofluorescence studies were performed by incubating the cells with either IgG-SRBI or an affinity-purified rabbit anti-human caveolin-1 polyclonal antibody (Transduction Laboratory, Lexington, KY) and then with fluorescein-conjugated goat anti-rabbit IgG (Zymed Laboratories Inc., San Francisco, CA) (4).
The ldlA7-SRBI cells were maintained in 100-mm dishes in medium A (1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) and Ham's F-12 medium with 100 units/ml penicillin, 100 g/ml streptomycin sulfate), supplemented with 5% (v/v) fetal calf serum and 0.25 mg/ml Geneticin ® (G418 sulfate, Life Technologies, Inc.). The ldlA7 and ldlA7-SRBI cells were plated at a cell density of 70,000 cells/well in six-well plates in medium A with 5% (v/v) fetal calf serum. On day 2, the cells were washed twice with phosphate-buffered saline (PBS) and refed with medium A containing 5% (v/v) newborn calf lipoprotein-deficient serum (NCLPPS).
Lipoproteins-Human HDL (density (d) ϭ 1.09 -1.21 g/liter) and LDL (d ϭ 1.04 -1.063 g/liter) was prepared by zonal centrifugation (21). The lipoproteins were delipidated and 50 g was subjected to 4 -15% SDS-polyacrylamide gradient gel electrophoresis and stained using Coomassie Blue or silver stain reagents; the lipoproteins used for the experiments described in this paper contained no detectable apolipoprotein E (data not shown). The total cholesterol (cholesterol/HP system pack, Boehringer Mannheim) and protein mass (22) of the lipoproteins were measured; the mass ratios of cholesterol to protein were ϳ1:4 for HDL, and 1:1.3 for LDL. The lipoproteins were iodinated using the iodine monochloride method (23), and the specific activity ranged from 75 to 100 cpm/ng protein. All lipoproteins were dialyzed extensively against 150 mM NaCl and 0.24 mM sodium EDTA, pH 7.4, before being used.
Cholesterol Esterification-To measure SR-BI-mediated stimulation of cholesterol esterification, the cells were grown as described previously except that 10 M sodium compactin (Sigma) and 100 M sodium mevalonate (Sigma) were added to the medium on day 2. On day 3, the cells were washed three times with PBS and incubated in 2 ml of DMEM containing 2 mg/ml fatty-acid-free bovine serum albumin (BSA) prior to addition of varying concentrations of HDL, LDL, or cholesterol. After a 5-h incubation with lipoproteins, the cells were pulsed with 0.2 mM [ 14 C]oleate (ϳ10,000 dpm/nmol) (NEN Life Science Products) for 2 h. Then the cells were washed three times with buffer A (50 mM Tris, 0.9% NaCl, and 0.2% BSA, pH 7.4), and once with buffer B (50 mM Tris and 0.9% NaCl, pH 7.4). The lipids were extracted twice from the cell monolayer and analyzed by thin-layer chromatography (TLC) (Silica Gel G, Machete-Nagel, Dueren, Germany) (23). The cells were dissolved in 1 ml of 0.1 N NaOH, and the protein was quantitated using the Lowry method (22).
Degradation of 125 I-LDL-The ldlA7 and ldlA7-SRBI cells were incubated with 125 I-LDL in medium A supplemented with 2 mg/ml fattyacid-free BSA for 5 h and pulsed with 0.2 mM [ 14 C]oleate for 2 h. The medium was removed, and the proteins were precipitated using trichloroacetic acid. The supernatant was extracted with chloroform and oxidized with hydrogen peroxide. An aliquot was counted to determine the amount of acid-soluble material formed by the cells (which is mostly [ 125 I]iodotyrosine) (23). The degradation activity, representing the celldependent proteolysis, is expressed in micrograms of 125 I-labeled, acidsoluble protein formed per mg of total cell protein. The cell monolayers were washed, and lipids were extracted and analyzed by TLC as described above.
Cholesterol Efflux Studies-Cells were maintained as described above except that on day 2 the medium was removed and the cells were washed twice with PBS and refed with medium A containing 5% NCLPPS serum.  (24,25). After 24 h the cells were washed three times with buffer A and once with buffer B prior to incubation in DMEM with 2 mg/ml fatty acid-free BSA and either HDL or LDL in the concentrations indicated. The medium was removed and the cells were washed twice with PBS prior to being lysed by adding 1 ml of 0.1 N NaOH. Aliquots of the medium and [ 3 H]cholesterol-labeled cells were analyzed using ␤-scintillation counting. For the [ 14 C]acetate and [ 3 H]mevalonate-labeled cells, aliquots of the medium and cells were analyzed exactly as described (24,25).

RESULTS
To study the role of SR-BI independently of the LDL receptor, we expressed recombinant mouse SR-BI (1) in CHO cell lines devoid of functional LDL receptors (ldlA7 cells) (18). Immunofluorescence studies were performed using IgG-SRBI and an anti-caveolin antibody. A similar linear punctate pattern of staining was seen on the cell surface of the SR-BI-expressing cells using both antibodies (data not shown), which is consistent with SR-BI being located in caveolae, as has been described previously (3). Immunoblot analyses of membrane fractions from the parental cell line (CHO-7) (lane 1), ldlA7 cells (lane 2), and subclones of stably transformed ldlA7 cells expressing variable amounts of SR-BI (lanes 3-5) were performed using IgG-SRBI ( Fig. 1, top panel). Low levels of SR-BI were present in the CHO-7 and ldlA7 cells. The levels of immunodetectable SR-BI were up to 75-fold higher in the ldlA7-SRBI cells (lane 3) than in the ldlA7 cells. Two other ldlA7-SRBI subclones that expressed lower amounts of SR-BI are shown in lanes 4 and 5.
A doublet was apparent in one of the subclones (lane 5), but was not seen in the other SR-BI-expressing cell lines. The slightly smaller immunoreactive band in the ldlA7-SRBIb cells (lane 5) likely represents SR-BI protein that has not been fully glycosylated. As expected, immunoreactive LDL receptor was present in the wild-type CHO-7 cells (lane 1, second panel); no LDL receptor was detected in any of the other cell lines even after a prolonged exposure of the blot to film. As a control, all of the cell lines expressed similar amounts of LRP (20), another cell surface receptor. Increased levels of expression of SR-BI were not associated with any changes in the steady state levels of caveolin-1, a structural protein of caveolae (bottom panel).
Although it was shown previously that SR-BI mediates selective transport of lipids from HDL into cells (1), the intracellular fate of the lipoprotein-derived lipids has not been defined.
To determine if SR-BI expression increases the delivery of sterols from lipoproteins to the ER for cholesterol esterification, ldlA7 and ldlA7-SRBI cells were incubated with increasing amounts of human HDL or LDL for 5 h, followed by a 2-h incubation with [ 14 C]oleate. The total mass of cholesterol converted to cholesterol ester within the 2 h interval was measured by thin layer chromatography (TLC) as described under "Experimental Procedures." Almost no increase in esterification was seen in the ldlA7 cells after the addition of either LDL or HDL (Fig. 2, panel A). In contrast, addition of LDL or HDL to the ldlA7-SRBI cells resulted in a dramatic increase in esterification (panel B). For any given amount of lipoprotein mass added to the medium, the amount of cholesterol esters formed was higher with LDL than with HDL. When the same data was replotted as a function of the amount of total cholesterol added to the cells (panel C), a similar response curve was seen with LDL and HDL. Uptake of [ 14 C]oleate (1300 versus 1200 pmol/mg) and incorporation into triglycerides (34,500 versus 36,000 pmol/mg) and phospholipids (59,000 versus 54,000 pmol/mg) were not significantly different between ldlA7 and ldlA7-SRBI with both LDL and HDL. Thus, in the absence of the LDL receptor, SR-BI expression is associated with an increase in lipoprotein-stimulated cholesterol esterification, and LDL is as effective as HDL as a sterol donor.
Data for this experiment, as well as all the other experiments described in this paper were generated using the ldlA7-SRBI cells. To ensure that the observations made were not a clonespecific artifact, every experiment was repeated in the ldlA7-SR-BIa and ldlA7-SRBIb cells; similar results were obtained, and the magnitude of the changes was proportional to the amount of SR-BI protein expressed (Fig. 1).
A time course of the lipoprotein-stimulated incorporation of [ 14 C]oleate into cholesterol esters in the ldlA7-SRBI cells is shown in Fig. 3. Cells were incubated with 100 g of protein/ml of either HDL or LDL for the indicated time periods, after which the cells were labeled with [ 14 C]oleate for an additional 2 h. Cells incubated with [ 14 C]oleate in the absence of lipoproteins had a low level of [ 14 C]oleate incorporation (ϳ50 pmol/2 h/mg), which did not change with time (data not shown). A small increase in cholesterol esterification was seen in the ldlA7 cells incubated with either LDL or HDL (panel A), which is likely due to the low levels of SR-BI present in these cells (see Fig. 1). In the ldlA7-SRBI cells, a linear increase in cholesterol esterification began 1 h after the addition of either LDL or HDL (panel B).
To determine the effect of SR-BI expression on the amount of cellular cholesterol, we measured the level of total cholesterol in the ldlA7 and ldlA7-SRBI cells after incubation with media alone or the addition of either LDL or HDL (100 g/ml). The level of total cellular cholesterol was 1.4-fold higher in the ldlA7-SRBI cells than the ldlA7 cells with no addition of lipoproteins (13.5 versus 9.5 g of cholesterol/mg). No significant change in the level of cholesterol was seen in the ldlA7 cells after the addition of LDL and HDL. In the ldlA7-SRBI cells there was a ϳ1.5and 1.8-fold increase in total cellular cholesterol content after addition of HDL and LDL, respectively.
The pathway by which SR-BI expression stimulates lipoprotein-associated cholesterol esterification was probed using chloroquine, a weak base that accumulates in acidic compartments and neutralizes their pH (26). Cells were preincubated for 15 min with chloroquine prior to the addition of 100 g of protein/ml HDL or LDL for 5 h. The cells were then pulsed with [ 14 C]oleate for 2 h. A dose-dependent decrease in lipoproteinassociated cholesterol esterification was seen with the addition of increasing amounts of chloroquine to the cells (Fig. 4). Lipoprotein-stimulated cholesterol esterification was completely blocked at 50 M chloroquine (panel B), which is similar to the level at which this drug inhibits uptake of LDL by the LDL receptor (27). To ensure that chloroquine did not inhibit ACAT activity, the cells were incubated with 25-hydroxycholesterol for 5 h and then pulsed with [ 14 C]oleate for 2 h in the presence or absence of 50 M chloroquine; the amount of cholesterol esters formed was nearly identical in the two sets of cells (panel C). These data are consistent with SR-BI-dependent stimulated cholesterol esterification requiring the maintenance of some acidic compartment(s) within the cell.
Previously, Acton et al. (1) showed that uptake of cholesterol from HDL via SR-BI was not coupled to degradation of the protein component. To determine if SR-BI mediates selective uptake of lipid from LDL as well as HDL, we repeated the esterification experiments using 125 I-labeled LDL (Fig. 5). In CHO-7 cells, LDL-stimulated esterification was completely inhibited by 50 M chloroquine (Fig. 5, panel A). These cells, which are the parental cell line for the ldlA7 cells, express both the LDL receptor and SR-BI (Fig. 1). In contrast, no esterification of cholesterol was seen in the LDL receptor-deficient ldlA7 cells in the presence or absence of chloroquine (data not shown). In the ldlA7-SRBI cells, which lack LDL receptors but have supraphysiologic levels of SR-BI, the level of LDL-stimulated cholesterol esterification was high (panel B) and markedly reduced with the addition of chloroquine.
The degradation of the protein component of the LDL was measured in the same experiments. In CHO-7 cells, degrada-  CHO-7, ldlA7, and ldlA7 cells (ldlA7-SRBI, ldlA7-SRBIa, and ldlA7-SRBIb) that stably express recombinant murine SR-BI. Fifty micrograms of solubilized membrane proteins from each cell line were size-fractionated on a 6.5% SDS-polyacrylamide gel (12.5% for caveolin) and transferred to nitrocellulose. Immunoblot analysis was performed using a rabbit polyclonal antibody to mouse SR-BI (IgG-SRBI), bovine LDL receptor (19), mouse LRP (20), and human caveolin-1. Subsequently, the filters were incubated with a horseradish peroxidase-coupled donkey anti-rabbit secondary antibody. The filters were developed using the enhanced chemiluminescence (ECL) system and exposed to Kodak DuPont NEF 496 film for 5 min.
tion of 125 I-lipoprotein was reduced by addition of chloroquine (panel C), but there still was an unexpectedly high level of residual degradation in these cells. To confirm that our lipoprotein particles were not denatured, parallel experiments were performed in cultured human fibroblasts and CHO-KI cells, the parental cell line of the CHO-7 cells. A complete inhibition of 125 I-LDL degradation was seen with addition of 50 M chloroquine (data not shown). Therefore, the relative resistance of the CHO-7 cells to LDL degradation by chloroquine was cell-specific, and not due to an artifact associated with lipoprotein labeling. CHO-7 cells, unlike CHO-K1 cells, have been adapted to grow in NCLPPS (28), which may have selected for an alternative, yet-to-be-defined lipid uptake mechanism.
In contrast to the parental cell line (CHO-7), the ldlA7 cells did not degrade 125 I-LDL in the absence or presence of chloroquine (data not shown). Low levels of 125 I-LDL degradation were seen in the ldlA7-SRBI cells (panel D), which was completely blocked by chloroquine. The ratio of cholesteryl [ 14 C]oleate formation to LDL degradation in the ldlA7-SRBI cells (without chloroquine) was 2.9, whereas in the CHO-7 cells it was 0.9. The same experiments were performed using HDL, and no degradation was seen in either the ldlA7 or ldlA7-SRBI cells (data not shown). Taken together, these data are consistent with SR-BI mediating selective uptake of sterols from LDL as well as HDL particles.
To determine if sterol uptake by SR-BI requires the lipid to be presented as part of a lipoprotein particle, free cholesterol (solubilized in ethanol) was added to cells and the incorporation of [ 14 C]oleate into cholesterol esters was measured (Fig. 6). Addition of ethanol alone had no effect on cholesterol esterification (data not shown). Cholesterol produced a modest increase in cholesterol esterification in the ldlA7 cells (Fig. 6, panel A) and CHO-7 cells (data not shown). In contrast, a marked, concentration-dependent increase in cholesterol esterification was seen in ldlA7-SRBI cells; the cholesterol-induced esterification was blocked by the addition of 50 mM chloroquine (panel B). To demonstrate that cholesterol-stimulated esterification was attributable to the expression of SR-BI, we tested two other subclones of ldlA7-SRBI cells, ldlA7-SRBIa and ldlA7-SRBIb (see Fig. 1). The amount of cholesterol-stimulated esterification seen in these cells were proportion to the levels of immunodetectable SR-BI (data not shown). The time course of cholesterol-stimulated esterification in the ldlA7-SRBI cells was delayed compared with that seen with HDL and LDL. The first detectable increase in esterification occurred between 2 and 3 h, and the rate of accumulation of cholesteryl [ 14 C]oleate increased in a linear fashion over the ensuing 5 h (data not shown). The finding that SR-BI is able to increase the delivery of cholesterol from the medium to the ER demonstrates that cholesterol need not be presented as part of a lipoprotein or vesicle for recognition by SR-BI. (Fig. 7, panel A). In the ldlA7 cells, HDL was somewhat better than LDL as an acceptor of cholesterol from the cells (panel A). SR-BI expression was associated with a marked increase in efflux of sterols from cells to either HDL or LDL. Up to 80% of the radioactivity was transferred to the medium after 5 h (panel B). Greater than 99% of the radioactivity in the medium was precipitable with trichloracetic acid, suggesting that the cholesterol was contained within lipoproteins (data not shown). The medium was also analyzed by TLC, and all of the radioactivity migrated the same distance from the origin as the cholesterol standard (data not shown).

It has been shown previously that expression of recombinant SR-BI in CHO cells is associated with an increase in cholesterol efflux from [ 3 H]cholesterol-labeled cells to HDL (29, 30). To determine if SR-BI can donate cholesterol to LDL as well as to HDL, we labeled ldlA7-SRBI cells with [ 3 H]cholesterol for 24 h and then incubated the cells with increasing amounts of LDL or HDL for 5 h
The time course of transfer of labeled cholesterol to either LDL or HDL (100 g/ml) is shown in panels C and D of Fig. 7. A linear increase in cholesterol efflux to LDL and HDL was seen in the ldlA7 cells (panel C). After 24 h, ϳ20% of the cholesterol label had been transferred to lipoproteins. In the ldlA7-SRBI cells, there was a more rapid increase in [ 3 H]cho- FIG. 2. Stimulation of cholesteryl-[ 14 C]oleate formation by HDL (closed symbols) or LDL (open symbols) in ldlA7 (panel A) and  ldlA7-SRBI cells (panels B and C). The ldlA7 cells, which are LDL receptor-negative CHO cells, and ldlA7-SRBI cells, which express recombinant mouse SR-BI, were plated in six-well dishes (70,000 cells/well) in 2 ml of medium A containing 5% (v/v) fetal calf serum (day 0). The cells were washed twice on day 2 with PBS and refed with 2 ml of medium A containing 5% (v/v) NCLPPS, 10 M compactin, and 100 M mevalonate. On day 3, the cells were washed three times with PBS and fed with 2 ml of DMEM containing 2 mg/ml fatty acid-free BSA and the indicated concentrations of either HDL or LDL. After 5 h, 0.2 mM [ 14 C]oleate (ϳ10,000 dpm/nmol) was added to the medium. Two h later, the cells were washed and the cholesterol ester levels were analyzed by TLC as described under "Experimental Procedures." Data points represent the mean of triplicate values. The experiment was repeated three times with similar results. Fig. 2  lesterol efflux that reached a plateau at ϳ12 h (panel D). The efflux was temperature-dependent; no cholesterol efflux occurred at 4°C (data not shown).

FIG. 3. Time course of lipoprotein-mediated [ 14 C]oleate incorporation into cholesterol esters in ldlA7 (panel A) and ldlA7-SRBI (panel B) cells. Cells were grown as described in
In the experiment shown in Fig. 7, the cells were incubated in medium containing [ 3 H]cholesterol dissolved in ethanol. Under these conditions most of the [ 3 H]cholesterol can be expected to be located in the cell membrane (31). To determine if SR-BI can mediate the efflux of endogenously synthesized cholesterol, we labeled cells for 24 h with [ 14 C]acetate, a precursor of cholesterol in the biosynthetic pathway, and then added increasing concentrations of either LDL or HDL for 5 h (Fig. 8).
In the ldlA7-SRBI cells, the amount of labeled cholesterol in the lipoproteins rose in proportion to the concentration of lipoproteins added to the medium (panel B). When the same data were expressed according to the amount of total cholesterol added to the cells, HDL appeared to be a slightly better acceptor than LDL under these conditions (data not shown). The efflux of [ 14 C]lanosterol, an intermediate in the cholesterol biosynthetic pathway, was also measured by TLC in the lipoproteins after incubation with ldlA7 cells (panel C) and ldlA7-SRBI cells (panel D). The SR-BI-associated efflux of lanosterol paralleled that of cholesterol. HDL was a better acceptor of lanosterol than LDL. No radiolabeled squalene was found in the medium and only very low levels of squalene were detected in the cells (ϳ20 dpm) (data not shown). DISCUSSION The major new findings in this paper are that SR-BI mediates uptake of lipids from LDL, as well as HDL, and that this uptake is associated with an increase in cholesterol esterification. The cholesterol content of the lipoproteins added to cells, rather than their apoprotein mass, correlated directly with the amount of cholesterol esters formed by the ldlA7-SRBI cells. Free cholesterol (or cholesterol aggregates) is as effective as lipoproteins at stimulating cholesterol esterification in the SR-BI-expressing cells (Fig. 6). Thus, cholesterol need not be associated with lipoproteins for SR-BI to mediate its uptake by cells. The SR-BI-dependent stimulation of cholesterol esterification was inhibited by chloroquine, a weak base that neutralizes the pH of acidic compartments. Finally, we show that SR-BI can mediate efflux of endogenously-synthesized sterols to lipoproteins and efflux of cholesterol from the plasma mem- FIG. 4. Dose-response effects of chloroquine administration on lipoprotein-stimulated cholesterol esterification by SR-BI. Cells were grown as described in Fig. 2. On day 3, the ldlA7 (panel A) and ldlA7-SRBI cells (panel B) were incubated in 2 ml of DMEM containing 2 mg/ml fatty acid-free BSA and increasing concentrations of chloroquine. After 15 min, 100 g of protein/ml of either HDL or LDL was added. In panel C, the cells were not treated with compactin and 25-hydroxycholesterol was added to ldlA7-SRBI in the presence or absence of chloroquine. After 5 h, the cells were pulsed for 2 h with [ 14 C]oleate. The amount of cholesterol [ 14 C]oleate formation was determined as described under "Experimental Procedures." Each data point represents the mean of triplicate measurements. The experiment was repeated three times with similar results.

FIG. 5. Cholesterol esterification and LDL degradation in CHO-7 (panels A and C) ldlA7-SRBI cells (panels B and D) incubated in the absence (closed circles) or presence (open circles) of chloroquine (C).
Cell monolayers were grown as described in Fig. 2 and incubated for 5 h in DMEM containing 2 mg/ml fatty acid-free BSA and the indicated concentrations of 125 I-LDL (75 cpm/ng) in the absence (ϪC) or presence (ϩC) of 50 M chloroquine. [ 14 C]Oleate (0.2 mM) was added to the cells for 2 h prior to collection of the medium and cells. The medium was subjected to trichloroacetic acid precipitation, and the acid-soluble counts were determined. The cells were extracted with hexane:isopropanol (3:2), and the lipids were subjected to TLC. The levels of cholesterol esters were determined as described under "Experimental Procedures." Each point represents the mean of triplicate measurements, and this experiment was repeated five times with similar results. brane to LDL, as well as HDL (29,30).
SR-BI binds native LDL, as well as HDL, with high affinity (2,17). Our studies provide evidence that SR-BI not only binds LDL but also mediates the transport of sterols from the lipoprotein to cells. SR-BI expression was associated with a dramatic increase in LDL-stimulated cholesterol esterification and this process was not accompanied by significant degradation of the lipoprotein. These results are consistent with prior in vivo studies which suggest that the same receptor that mediates selective uptake of cholesterol esters from HDL in steroidogenic tissues also takes up sterols selectively from LDL (32).
Previously Reaven and colleaugues (33) showed that apoliprotein E-free HDL stimulated cholesterol esterification in ovarian granulosa cells, which we now know express high levels of SR-BI (4). The results of our experiments are consistent with SR-BI being responsible for the delivery of HDL-derived cholesterol for esterification. We show that, in the absence of a functional LDL receptor, LDL is as effective as HDL at stimulating cholesterol ester formation in cells expressing SR-BI. Fielding and Fielding (31) have shown that some LDL-associated free cholesterol is taken up by a selective mechanism that is independent of the LDL receptor, and this uptake may be mediated by SR-BI. However, this uptake mechanism is unlikely to have physiological significance in delivering LDL from the plasma, as is evidenced by the very high plasma levels of LDL in humans without any functional LDL receptor (homozygous familial hypercholesterolemia). On the other hand, these individuals do not have clinically significant deficiencies of steroid hormones, which may be explained by the action of SR-BI. SR-BI may thus serve as a back-up receptor that en-sures sufficient delivery of lipids to steroidogenic tissues to support reproduction and the stress response.
The LDL and HDL particles used in these experiments share no apolipoproteins in common, suggesting that a shared lipid, rather than protein, component is recognized by the receptor. Moreover, the fact that cholesterol alone was able to stimulate cholesterol esterification in an SR-BI-dependent manner, suggests that lipids, and not proteins, are the natural ligands for this receptor. These results are consistent with prior in vivo studies demonstrating that selective uptake of lipoprotein sterols by tissues does not require the presence of specific apolipoproteins (34). In this regard, Xu et al. (35) showed that SR-BI also binds some lipid-free proteins, including apoA-I, apoA-II, and apoC-III. The ability of SR-BI to bind both lipids and proteins is a feature that this receptor shares with a close relative in the class B scavenger receptor family, CD-36. CD-36 binds numerous proteins (thrombospondin, type 1 collagen) as well as lipids (free fatty acids) (for review, see Ref. 36). Both SR-BI and CD-36 bind acetylated and oxidized LDL as well as anionic phospholipids (2,37). It will be of interest to determine the structural features of these two proteins that allow them to interact with such a diverse range of ligands.
SR-BI expression was associated with a marked increase in cholesterol esterification when free cholesterol was added to the medium (Fig. 6). Previously, it has been shown that the addition of low concentrations of cholesterol (Ͻ5 g/ml) to fibroblasts does not result in an appreciable increase in cholesterol esterification by the cells (38). Only addition of higher levels of cholesterol (Ͼ5 g/ml) stimulate cholesterol esterification in fibroblasts (38). Cholesterol forms aggregates at concentrations greater than 5 g/ml, which may alter the pathway by which the sterol enters cells (i.e. via endocytosis rather than

FIG. 7. Effect of SR-BI expression on cholesterol efflux from [ 3 H]cholesterol-labeled ldlA7 (panel A) and ldlA7-SRBI cells (panel B) to HDL (closed symbols) and LDL (closed symbols).
Cells were grown as described under "Experimental Procedures" until day 2 when they were washed and refed with medium A containing 5% NCLPPS and 5 Ci/well [ 3 H]cholesterol. On day 3, the cells were washed prior to the addition of either HDL or LDL. Cells were allowed to incubated with the indicated concentration of HDL or LDL for 5 h (panels A and B) or with 100 g of protein/ml of either lipoprotein for the indicated times (panels C and D). Medium was collected, and the cells were washed twice with PBS prior to lysis. The radioactivity was quantitated in the medium and cells. At the zero time point, the ldlA7 cells and ldlA7-SRBI cells had 330,000 and 630,000 dpm/mg cellular protein, respectively. TLC analysis showed that all the label was in cholesterol. This experiment was repeated five times with similar results. A and B) and lanosterol (panels C and D) from ldlA7 cells labeled with [ 14 C]acetate. Cells were grown for 2 days in medium A with 5% fetal calf serum, then washed twice with PBS and refed with medium A containing 5% NCLPPS and 5 Ci/well [ 14 C]acetate. On day 3, cells were washed and refed with DMEM plus 2 mg/ml fatty acid-free BSA and the indicated amount of HDL (closed symbols) or LDL (open symbols). After 5 h, the medium was collected and the cells were washed and lysed. The ldlA7 and ldlA7-SRBI cells had 955,000 and 705,000 dpm/mg, respectively at the zero time point. The lipids were extracted from the cells and medium and analyzed by TLC as described under "Experimental Procedures." Each point represents the mean of triplicate determinations. The experiment was repeated twice with similar results. by molecular transfer), as has been proposed previously (38). No increase in cholesterol esterification was seen when we incubated the SR-BI-expressing cells with 2 or 5 g/ml cholesterol (data not shown). Addition of 10 g/ml cholesterol to the medium resulted in a 2-fold increase in esterification in both the ldl7A and ldlA7-SRBI cells (Fig. 6). It was only after addition of high concentrations of cholesterol, which are invariably associated with aggregate formation, that we saw a significant increase in cholesteryl [ 14 C]oleate formation in the SR-BI-expressing cells. The physiological relevance of this observation is not clear. Perhaps SR-BI mediates sterol uptake from cholesterol aggregates/crystals that are present in necrotic tissues and atherosclerotic plaques.

FIG. 8. The effect of SR-BI expression on the efflux of endogenously synthesized cholesterol (panels
The SR-BI-dependent stimulation of cholesterol esterification by free cholesterol, as well as by lipoprotein-derived cholesterol, was inhibited by chloroquine. At this time we can only speculate as to the intracellular pathway(s) lipids traverse after interacting with SR-BI on the cell membrane. SR-BI resides in caveolae, and sterols may enter the cell via this route. Caveolin, which is the major protein in caveolae, binds cholesterol (39) and can be stimulated to move between the cell membrane, the ER, and the Golgi apparatus (40). Caveolae trafficking from the cell membrane, a process referred to as potocytosis, requires acidification of the vesicles, which would expected to be disrupted by chloroquine (41). Alternatively, SR-BI may facilitate the entry of sterols via endocytic vesicles (33,42,43). Endosomes are known to have a high cholesterol content (44) and uptake into this compartment is blocked by chloroquine. Cholesterol esters internalized by the selective uptake pathway accumulate in the cytoplasm, prior to being hydrolyzed by a cholesterol esterase (45). Sparrow and Pittman (45) showed that chloroquine does not block hydrolysis of these cholesterol esters. They also showed that in Wolman's syndrome, in which individuals have no functioning acid lipase, the HDL-derived cholesterol esters are hydrolyzed normally, suggesting they do not enter the cell via the classic endocytic pathway (44). Thus, if lipoprotein-associated cholesterol esters are the major source of sterols for esterification in the SR-BIexpressing cells, the cholesterol esters must pass through an acidic compartment en route to the ER after being hydrolyzed.
We do not know if the lipoprotein-stimulated increase in cholesterol esterification in SRBI-expressing cells is due to the uptake of cholesterol or cholesterol esters (or both) from lipoprotein particles. In preliminary experiments, we have shown that radiolabeled cholesterol is incorporated into cholesterol esters in an SR-BI-dependent fashion (data not shown). Experiments are in progress to determine the relative contribution of cholesterol and cholesterol esters to the lipoproteinstimulated SRBI-dependent cholesterol esterification in the ldlA7-SRBI cells.
SR-BI not only mediates the influx of lipids from lipoproteins but also promotes efflux of cholesterol residing in the cell membrane (29,30). The transport of newly synthesized cholesterol, as well as membrane cholesterol, to lipoproteins is enhanced by SR-BI expression. It is likely that the directionality of the transfer of lipids between membranes and lipoproteins is governed by the relative concentrations of lipid in both compartments, as well as the relative partition coefficient of lipids into the membrane and lipoprotein.