Highly Purified Scavenger Receptor Class B, Type I Reconstituted into Phosphatidylcholine/Cholesterol Liposomes Mediates High Affinity High Density Lipoprotein Binding and Selective Lipid Uptake*

The murine class B, type I scavenger receptor mSR-BI is a high and low density lipoprotein (HDL and LDL) receptor that mediates selective uptake of cholesteryl esters. Here we describe a reconstituted phospholipid/ cholesterol liposome assay of the binding and selective uptake activities of SR-BI derived from detergent-solu-bilized cells. The assay, employing lysates from epitope-tagged receptor (mSR-BI-t1)-expressing mammalian and insect cells, recapitulated many features of SR-BI activity in intact cells, including high affinity and saturable 125 I-HDL binding, selective lipid uptake from [ 3 H]cholesteryl ether-labeled HDL, and poor inhibition of HDL receptor activity by LDL. The novel properties of a mutated receptor (Q402R/Q418R, normal LDL binding but loss of most HDL binding) were reproduced in the assay, as was the ability of the SR-BI homologue CD36 to bind HDL but not mediate efficient lipid uptake. In this assay, essentially homogeneously pure mSR-BI-t1, prepared by single-step immunoaffinity chromatography, mediated high affinity HDL binding and efficient selective lipid uptake from HDL. Thus, SR-BI-mediated HDL binding and selective lipid uptake are intrinsic properties of the receptor that do not require the intervention of other proteins or specific cellular structures or compartments.

The murine class B, type I scavenger receptor mSR-BI is a high and low density lipoprotein (HDL and LDL) receptor that mediates selective uptake of cholesteryl esters. Here we describe a reconstituted phospholipid/ cholesterol liposome assay of the binding and selective uptake activities of SR-BI derived from detergent-solubilized cells. The assay, employing lysates from epitopetagged receptor (mSR-BI-t1)-expressing mammalian and insect cells, recapitulated many features of SR-BI activity in intact cells, including high affinity and saturable 125 I-HDL binding, selective lipid uptake from [ 3 H]cholesteryl ether-labeled HDL, and poor inhibition of HDL receptor activity by LDL. The novel properties of a mutated receptor (Q402R/Q418R, normal LDL binding but loss of most HDL binding) were reproduced in the assay, as was the ability of the SR-BI homologue CD36 to bind HDL but not mediate efficient lipid uptake. In this assay, essentially homogeneously pure mSR-BI-t1, prepared by single-step immunoaffinity chromatography, mediated high affinity HDL binding and efficient selective lipid uptake from HDL. Thus, SR-BI-mediated HDL binding and selective lipid uptake are intrinsic properties of the receptor that do not require the intervention of other proteins or specific cellular structures or compartments.
The LDL 1 receptor pathway for the delivery of lipoprotein cholesterol to cells involves clathrin-coated pit-mediated endocytosis and subsequent lysosomal degradation of the entire LDL particle (1). Almost 20 years ago a strikingly different mechanism for the cellular uptake of lipoprotein cholesterol, called selective cholesterol uptake, was identified during the analysis of HDL metabolism in vivo (2,3). Selective cholesterol uptake from HDL and other lipoproteins does not involve endocytosis and subsequent degradation of the entire lipoprotein particle (2, 3; reviewed in Refs. 4 and 5). In the case of HDL, the lipoprotein binds to the cell surface and transfers its cholesteryl esters to the cell, and then the lipid-depleted HDL particle dissociates from the cell and can re-enter the circulation.
The HDL receptor SR-BI (scavenger receptor, class B, type I) was the first cell surface receptor to be shown to mediate physiologically relevant selective lipid uptake (6)(7)(8)reviewed in Ref. 5). In vivo studies have established that SR-BI critically influences HDL structure and metabolism and apparently plays an important role in the transport of cholesterol from peripheral tissues to the liver for recycling or biliary excretion (5,(7)(8)(9)(10)(11)(12). This probably accounts for the ability of SR-BI to protect against atherosclerosis in murine models (9,(13)(14)(15). Recent studies have established that expression of SR-BI in mice is normally required for red blood cell development (16) and female fertility (9,17) and can prevent the development of myocardial infarctions, cardiac dysfunction, and premature death in apoE-deficient mice (18).
In vitro studies have shown that SR-B can bind and mediate lipid uptake from LDL as well as HDL (6, 19 -22). Strikingly, HDL competes efficiently for LDL binding, whereas LDL is a poor competitor of HDL binding (6,23). SR-BI also can facilitate the efflux of unesterified cholesterol from cultured cells (24,25), although the physiologic significance of this is not certain. Several studies support the proposal (26) that SR-BImediated transport of lipids between cells and lipoproteins involves two sequential steps: 1) productive lipoprotein binding and 2) binding-dependent, yet distinct, SR-BI-mediated lipid transfer (25)(26)(27)(28)(29). It has not yet been determined if SR-BImediated selective lipid uptake occurs only at the cell surface, or in some intracellular compartment followed by retroendocytosis (secretion) of the lipid-depleted lipoprotein, or both (30 -32; reviewed in Refs. 4

and 5).
A particularly important question regarding the mechanism of SR-BI activity has been: does SR-BI require the participation of one or more other proteins to mediate either ligand binding, lipid transport, or both, or are these activities autonomous properties of SR-BI (independent of other proteins)? This question has arisen, in part, because of the multiple and complex activities of SR-BI (5). A direct approach for studying the autonomous properties of SR-BI and its mechanism of action is to examine in an in vitro system the activity of the receptor purified away from other proteins. Here we describe an in vitro reconstituted liposome assay for SR-BI-mediated ligand binding and selective lipid uptake that reflects many of the characteristics of SR-BI activity in intact cells. This assay, effective with total cell lysates as well as highly purified protein, was used to show that, in the absence of other proteins, SR-BI can bind HDL and LDL and mediate efficient selective cholesteryl ether uptake from HDL.

Materials
Human high density lipoprotein (HDL), human low density lipoprotein (LDL), 125 I-labeled HDL, 125 I-labeled LDL, 1,1Ј-dioctadecyl-3,3,3Ј,3Ј-tetramethylindocarbocyanime perchlorate (DiI)-labeled HDL (DiI-HDL), and human lipoprotein-deficient serum were prepared as previously described (6,19). The phCD36 expression vector (33) was a generous gift from B. Seed (Massachusetts General Hospital). The KKB-1 antibody was a generous gift from Karen Kozarsky (25). The 1D4 antibody was obtained from the ATCC. The mixture of complete protease inhibitors was purchased from Roche Molecular Biochemicals. All other reagents were purchased from standard suppliers or obtained as indicated below. Cell culture supplies were purchased from Invitrogen, Irvine Scientific, or JRH Biosciences. The peptide TETSQVAPA was prepared in the biopolymer laboratory at Massachusetts Institute of Technology and was a gift from the G. Khorana laboratory.

Construction of Expression Vectors
Mammalian Expression Vectors-mSR-BI cDNA was amplified from pmSR-BI (6) by PCR using the primers BL5, 5Ј-GACACTGGTACCGA-TATCACGCGGACATGGGCGGCAGCTCCAG-3Ј, and BL3, 5Ј-CTGTC-ACTCGAGGTCGACTTAGGCAGGCGCCACTTGGCTGGTCTCTGTT-AGCTTGGCTTCTTGCAGCACCGTG-3Ј that resulted in a cDNA encoding a full-length mSR-BI protein that contained a 9-amino acid (TETSQVAPA) C-terminal extension representing the C terminus of bovine rhodopsin as an epitope tag. The protein product of this cDNA is designated "mSR-BI-t1Ј." The PCR product was treated with restriction endonucleases KpnI and XhoI, and the KpnI/XhoI fragment was cloned into the mammalian expression vector pcDNA3(ϩ) (Invitrogen) that includes an internal neomycin resistance marker for selection in mammalian cells. The ligation product was used to transform Escherichia coli cells, and the plasmid DNAs of selected clones were isolated and sequenced. One of these with the expected sequence was designated pmSR-BI-t1. The mSR-BI-t1 cDNA was then reconstructed into an expression plasmid (36) using an SalI and EcoRV restriction fragment from pmSR-BI-t1. This plasmid was designated pACmSR-BI-t1 and was used for the generation of the stable cell line HEK[mSR-BI-t1] (see below).

Mammalian Cell Culture and Transfection
The mammalian cell lines HEK293S (38) (gift from P. Reeves and G. Khorana) and COS M6 were maintained attached to the substratum in medium A (Dulbecco's modified Eagle's medium/Ham's F-12 medium (1:1) supplemented with 50 units/ml penicillin, 50 g/ml streptomycin, 2 mM glutamine (Invitrogen), and 10% fetal bovine serum (JRH Biosciences)) in a humidified 95% air/5% CO 2 incubator. For suspension culture, HEK293S cells were grown with constant gentle stirring at 20 -40 rpm in spinner bottles in medium B (HBGro medium (Irvine Scientific) supplemented with 50 units/ml penicillin, 50 g/ml streptomycin, 2 mM glutamine, and 10% fetal bovine serum) at 37°C in a humidified incubator in 5% CO 2 /95% air. For suspension growth, cells were inoculated at 1.2 ϫ 10 5 cells/ml in 500 ml of medium B and incubated for 5-7 days without changing the medium. Cells were harvested either after reaching confluence on dishes by treatment with trypsin/EDTA (JRH Biosciences) or after 6 days of growth in suspension.
For transient transfections, 1.5 ϫ 10 6 COS M6 cells were seeded in 10-cm dishes in medium A without antibiotics and grown at 37°C overnight. The cells were treated with 10 g per plate of DNA (pmSR-BI-t1 or the "empty" vector pcDNA3 without an expression cassette insert as a control) using the LipofectAMINE (Invitrogen) method according to the manufacturer's recommendations. The cells were grown in medium A for an additional 2 days, and then receptor activity assays were performed. Briefly, 24 h post-transfection, the cells were washed with PBS, harvested with trypsin/EDTA, and plated in 1 ml/well of medium A at 150,000 cells/well in 24-well plates. The receptor activity of the cells was analyzed 48 h post-transfection.
A stable cell line expressing high levels of mSR-BI-t1, HEK[mSR-BI-t1] (clone 7), was established as follows: HEK293S cells were transfected with the pACmSR-BI-t1 vector using the LipofectAMINE method and maintained for 2 weeks in medium A supplemented with 0.8 mg/ml G418. Individual colonies were isolated by screening for their abilities to take up the fluorescent dye DiI from DiI-HDL (10 g of protein/ml, 2 h at 37°C) with an inverted fluorescence microscope and by flow cytometry as previously described (26).

Insect (sf) Cell Culture and Baculovirus Expression
The Sf21 cells (gift from K. Cha, E. Getmanova, and G. Khorana) were maintained in SFM II medium (Invitrogen) in suspension at 27°C in a humidified incubator and passaged when the viable cell count reached 2 ϫ 10 6 cells/ml. The recombinant bacmid DNAs described above, encoding mSR-BI-t1, the 402 R/ 418 R mutant form of mSR-BI-t1 and CD36, were used to transfect Sf21 cells in T25 flasks using the Bac-to-Bac expression system (Invitrogen) according to the manufacturer's recommendations. After growth of the transfected Sf21 cells for 5 days at 27°C in Grace's medium, the media were collected as the recombinant baculovirus particle stocks and used for infection of Sf21 cells (procedures carried out according to the manufacturer's recommendations). The infected cells were grown in Grace's medium for 72-168 h and harvested by centrifugation. High level expression of the proteins was verified by SDS-PAGE, and immunoblotting analysis of cell extracts, prepared as described below, using polyclonal anti-SR-BI antibody KKB-1 or anti-CD36 antibody (BD PharMingen). Cells also were infected with an "empty" vector to permit generation of receptorfree cell extract controls. Cell lysates were prepared (see below) and used for immunoblot analysis and liposome reconstitution studies.

Determination of Receptor Cell Surface Expression and Function in Intact Mammalian Cells
Cell Surface Expression-Transiently transfected COS cells were incubated with polyclonal antisera (KKB-1) against mSR-BI, washed, incubated with fluorescein-conjugated goat anti-rabbit IgG (Amersham Biosciences), washed again, and harvested with trypsin, and receptor cell surface expression was determined by flow cytometry as previously described (23).
Cell Function Analysis-The 125 I-HDL binding assay was performed as previously described (6). The [ 3 H]CE-HDL cell association assay was similar to the 125 I-HDL binding assay, except for the following changes: the NaOH cell lysis step was preceded by the addition of 1 ml of isopropanol at room temperature for 30 min to extract the incorporated

Preparation of Total Cell Lysates
HEK[mSR-BI-t1] and untransfected HEK293S cells were plated at 1-2 ϫ 10 5 cells/plate in 10-cm plates and grown in 15 ml of medium A for 3-4 days. When the cells were confluent, the plates were washed twice with PBS, and the cells were harvested by scraping with a rubber policeman and concentrated into a pellet by centrifugation at 4°C at 1,460 ϫ g for 20 min in an SS34 rotor in a Sorvall RC-5B centrifuge (DuPont Instruments). The cell pellets were dissolved by adding 1 ml of lysis buffer A (1.5% (w/v) octyl glucoside and protease inhibitors (1ϫ) in PBS) per 0.1 g (wet weight) of pellet and incubating at 4°C for 30 min with gentle mixing. The lysates were clarified by centrifugation at 11,951 ϫ g using an SS34 rotor for 20 min at 4°C, and the supernatants were collected and used as total cell lysates. Protein concentration was determined by the DC protein assay (Bio-Rad) and was typically 1-2 mg/ml. The same procedure was used to prepare lysates from Sf21 cells.

Immunoaffinity Purification of mSR-BI-t1
mSR-BI-t1 was purified by immunoaffinity chromatography using a modification of the procedure of Reeves et al. (38). Briefly, 1.5 g (wet weight) of HEK[mSR-BI-t1] cells were incubated with 13 ml of lysis buffer A at 4°C for 30 min. The solution was clarified by centrifugation (4°C, 11,951 ϫ g, 20 min), and filtration (0.45 M Durapore membrane (Millipore)). The subsequent purification steps were carried out at room temperature. Five milliliters of the extract were loaded onto a 0.3-ml anti-C-terminal 1D4 antibody-Sepharose column, which was pre-equilibrated with lysis buffer A. After washing with 50 column-volumes of lysis buffer A, and 10 column-volumes of column buffer A (10 mM Tris (pH 6.0), 1.5% octyl glucoside, and protease inhibitors), bound proteins were eluted with 2 ml of elution buffer (column buffer A containing 100 M of the peptide epitope (TETSQVAPA), 0.3-ml fractions). Samples (10 l) were fractionated by SDS-PAGE and analyzed by either silver staining or immunoblotting with the anti-SR-BI antibody KKB-1. The purified protein was stored in the elution buffer at Ϫ20°C.

Liposome Preparation
The preparation of multilamellar vesicles was carried out according to Schneider et al. (39) with minor modifications. Briefly, egg yolk L-␣-phosphatidylcholine (Avanti) and cholesterol (Sigma Chemical Co.) (molar ratio of 5:1) in ether were dried and resuspended at 2 mg/ml in 50 mM Tris buffer (pH 6.0). Then 0.5 ml of the suspension and 2.4 ml of buffer B (50 mM Tris, pH 6.0, 150 mM NaCl, 2 mM CaCl 2 ) containing either 100 g of detergent-solubilized cell lysate or 2 g of detergentsolubilized purified mSR-BI-t1 were mixed, precipitated by 0.6 volume of ice-cold acetone, and the precipitate was recovered by centrifugation (4°C at 30,596 ϫ g, 20 min). The liposomes were resuspended in 300 l of buffer C (20 mM Tris, pH 8.0, 1 mM CaCl 2 , 20 mM NaCl, and protease inhibitors).

Filter Binding Assay
Liposomes (8 l) were diluted with 12 l of assay buffer (20 mM Tris, pH 8.0, 2 mM CaCl 2 , 80 mg/ml fat-free bovine serum albumin (Sigma), and protease inhibitors (1ϫ)) and the indicated amounts of 125 I-HDL or [ 3 H]CE-HDL (usually 10 g of protein/ml) in the presence (single incubations) or absence (duplicate incubations) of a 40-fold excess of unlabeled HDL. The mixture was incubated at 37°C for 2 h unless otherwise noted, and the liposomes with bound 125 I-HDL were isolated by filtration (39) using 0.45 M nitrocellulose membranes and a multifilter filtration manifold (Millipore, Milford, MA). Briefly, the filters in the manifold were moistened with wash buffer (20 mM Tris, pH 8.0, 50 mM NaCl, 20 M CaCl 2 , and 1 mg/ml fatty acid-free bovine serum albumin), the assay mixture was filtered at room temperature and then the filters were washed three times with wash buffer. The amounts of 125 I bound were measured with a gamma counter (LKB-Wallac, Finland). To measure 3 H, the filters were added to 4 ml of Hydrofluor scintillation fluid (National Diagnostics) and radioactivity was measured using a liquid scintillation analyzer (Packard Instrument Co., Meriden, CT).

RESULTS
The two goals of this study were 1) to develop an in vitro reconstituted liposome assay of SR-BI-mediated ligand binding and lipid uptake and 2) to use the assay to determine if highly purified SR-BI required the cooperation of other proteins to mediate HDL binding and selective lipid uptake or if it could do so independently of other proteins. To address the later goal, we developed a modification of the method of Reeves et al. (38), developed for the analysis of rhodopsin, to isolate essentially homogeneously pure and functional mSR-BI containing an exogenous epitope tag.
Comparison of Binding and Uptake Activities of Cells Transfected with Plasmid DNA of Wild-type and Tagged mSR-BI-To facilitate isolation of pure mSR-BI protein for reconstitution into liposomes, we slightly modified the approach described by Reeves, Thurmond, and Khorana (38) for the generation and purification of recombinant bovine rhodopsin from transfected mammalian cells. Reeves et al. (38) expressed bovine rhodopsin in HEK293S cells and purified the detergentsolubilized protein using monoclonal antibody affinity purification and epitope peptide elution. The anti-rhodopsin antibody 1D4 recognizes the C-terminal 9-amino acid peptide from rhodopsin (40). For the synthesis and purification of mSR-BI, we constructed a mammalian cell expression vector, pmSR-BI-t1, that encodes a chimeric protein (mSR-BI-t1) containing the full-length mSR-BI and, at its C terminus, the C-terminal epitope-tag TETSQVAPA from bovine rhodopsin. To determine if epitope tagging interfered with the activity of the receptor, we transiently transfected COS M6 cells with expression vectors encoding either wild-type mSR-BI (COS[mSR-BI]), epitope-tagged mSR-BI-t1 (COS[mSR-BI-t1]), or an empty vector (pcDNA) control (COS[control]) and measured the two best defined activities of SR-BI (5) Fig. 1 shows that the specific 125 I-HDL binding (panel A) or [ 3 H]cholesteryl ether uptake (panel C) values for mSR-BI and mSR-BI-t1 were similar. The relative levels of surface expression of the wild-type and tagged receptors were determined using an anti-SR-BI antibody (KKB-1) and quantitative flow cytometry (25). There was a somewhat higher level of expression of the tagged receptor (1.2-fold). When the 125 I-HDL binding and [ 3 H]cholesteryl ether uptake values for the wildtype receptor were normalized to account for this difference, the binding and lipid uptake curves for the two receptors were virtually identical (Fig. 1, B and D). The ratios of the maximal [ 3 H]CE-HDL uptake/ 125 I-HDL binding, expressed as equivalent ng of HDL protein/mg of cell protein, were 47 and 44, respectively, clearly showing that both receptors mediated selective lipid uptake. Therefore, the epitope tagging at the C terminus of mSR-BI did not appear to alter the key activities of mSR-BI in transfected COS cells, and this tagged mSR-BI should be useful for the isolation of purified SR-BI (see below) for studying its activity in a reconstituted system.
Reconstituted Liposome Assay for the Cell-free Analysis of SR-BI Function-We have developed a reconstituted liposome filter binding assay for measuring the binding of 125 I-HDL and the uptake of [ 3 H]cholesteryl ether from [ 3 H]CE-HDL mediated by SR-BI derived from detergent-solubilized cultured cells based on the LDL receptor binding assay of Schneider et al. (39) (see "Experimental Procedures"). In brief, phosphatidylcholine/ cholesterol vesicles were prepared using detergent-solubilized whole cell lysates or detergent-solubilized purified receptor protein. After incubation with radiolabeled lipoprotein (standard conditions: 5-10 g of protein/ml of lipoprotein for 2 h at 37°C, variations indicated below), the liposome-associated radioactivity was determined by ultrafiltration and counting. Standard controls included: preparation of liposomes either with lysates from cells that do not express recombinant receptor or without added cell lysate or purified receptor; and incubation with a 40-fold excess of unlabeled lipoprotein to compete for the specific binding of the ligands to the receptor. The nonspecific background binding of 125 I-HDL varied from 19 to 50% of the total binding and appeared to depend critically on the quality of the preparation, e.g. extent of radiolytic decomposition of the HDL, which depends on the age and specific activity of the preparations and can occur rapidly (41). The absolute values for binding and lipid uptake constants (apparent K d and B max (binding maximum) and U max (uptake maximum)) also varied somewhat between different receptor preparations and depended on the quality of the radiolabeled lipoprotein (41).
Validation of the Reconstituted Liposome Assay-The reconstituted liposome assay was initially validated using cell lysates from insect cells (Sf21) expressing high levels of tagged receptor (not shown) due to infection with a baculovirus encoding mSR-BI-t1. Previous studies have shown that insect cells can express on their surfaces functional mammalian SR-BI (42,43). Fig. 2A shows the ligand concentration dependence of the binding of 125 I-HDL (squares) and the uptake of [ 3 H]cholesteryl ether from [ 3 H]CE-HDL (circles) by mSR-BI-t1-containing insect cell lysate-reconstituted liposomes (open symbols) and control liposomes prepared with lysates from cells infected with baculovirus without the insertion (control, filled triangles). Specific 125 I-HDL binding to the mSR-BI-t1-containing liposomes (open squares) was high affinity (apparent K d ϳ 21 g of protein/ml, similar to that seen in mSR-BI-expressing mammalian cells (6,41)) and saturable (B max ϳ 15 ng of protein/assay), whereas there was very little specific binding to the mSR-BI-t1-negative control liposomes (filled triangles). Similar results were obtained for the uptake of [ 3 H]cholesteryl ether from [ 3 H]CE-HDL (open circles, apparent K d ϳ 11 g of protein/ml, U max ϳ 106 ng of protein/assay); the maximal value for lipid uptake was 7-fold greater than that for binding, indicating selective uptake.  (6). There was little binding or lipid uptake by the receptor-negative control liposomes (solid triangles). Fig. 4 shows the temperature dependence of 125 I-HDL binding and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL (10 g of protein/ml) by mSR-BI-t1-containing insect cell lysate-reconstituted liposomes. The value for 125 I-HDL binding at 0°C was somewhat lower (79%) than that at 37°C (100%), whereas there was a more substantial reduction in [ 3 H]cholesteryl ether uptake at 0°C (25%) compared with that at 37°C (100%). It has previously been noted that SR-BI-mediated binding of HDL (23,34) and selective uptake (34) are lower at 4°C than at 37°C and that the temperature sensitivity of selective uptake is greater than that of HDL binding (34). 2 To determine if the liposome assay could be used for recombinant SR-BI produced by mammalian cells, we transfected HEK293S cells with a mammalian expression vector for mSR-BI-t1 (pACmSR-BI-t1) and isolated a clone expressing high levels of the receptor, HEK[mSR-BI-t1] (clone 7). Detergentsolubilized lysates of these cells were reconstituted into liposomes, and the binding of 125 I-HDL and the uptake of [ 3 H]cholesteryl ether from [ 3 H]CE-HDL were measured. Fig. 5A shows results similar to those seen using extracts from the insect cells, i.e. high affinity HDL binding (apparent K d 8.5 g of protein/ml, B max 12.8 ng of HDL protein/assay) and selective lipid uptake (apparent K d 29.4 g of protein/ml, U max 220.9 ng of HDL protein/assay).
The specificity of the receptor's activity in mSR-BI-t1-containing mammalian cell lysate-reconstituted liposomes was examined by comparing the abilities of unlabeled HDL and LDL to inhibit [ 1. 125 I-HDL binding (A and B)

and [ 3 H]CE uptake from [ 3 H]CE-HDL (C and D) by COS cells expressing
wild-type mSR-BI or the C-terminal epitope-tagged mSR-BI-t1. COS M6 cells were transiently transfected with expression vectors for mSR-BI, mSR-BI-t1 (mSR-BI with a rhodopsin epitope tag incorporated as a C-terminal extension), or the control "empty" vector pcDNA. Two days later, the cells were incubated for 2 h at 37°C with the indicated concentrations of either 125 I-HDL or [ 3 H]CE-HDL in the presence (single determinations) or absence (duplicate determinations) of a 40fold excess of unlabeled HDL. Specific 125 I-HDL binding (A) and [ 3 H]CE uptake (C) were determined as described under "Experimental Procedures." The relative levels of cell surface expression of mSR-BI and mSR-BI-t1 were determined using the polyclonal anti-mSR-BI antibody KKB-1 by flow cytometry as described under "Experimental Procedures." The values for the binding and lipid uptake by the mSR-BI-expressing cells were corrected to account for the 1.2-fold difference in surface expression relative to that of the mSR-BI-t1-expressing cells and are shown in C and D.
high affinity, LDL is a poor inhibitor of HDL binding to SR-BI (6,25). Fig. 5B shows that, as is the case for intact cells, HDL was an effective inhibitor of mSR-BI-t1-mediated lipid uptake activity from [ 3 H]CE-HDL, whereas LDL was less effective. The [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL (10 g of protein/ml) by mSR-BI-t1-containing mammalian cell lysatereconstituted liposomes was inhibited by the polyclonal anti-SR-BI antibody KKB-1 to the same extent as by excess unlabeled HDL (Fig. 5C), whereas uptake was not inhibited by control pre-immune antibody. Control untransfected HEK293S cell lysate-reconstituted liposomes exhibited essentially no [ 3 H]cholesteryl ether uptake activity.
Although the above data support the validity of the liposome assay, we further tested the assay by determining the receptor activities of liposomes reconstituted with insect cell lysates containing in place of mSR-BI-t1 either a mutant form of mSR-BI-t1 or human CD36, another class B scavenger receptor (5,19). We have isolated a set of mutant mSR-BIs, which exhibit altered ligand-binding properties when expressed in transfected mammalian cells (23,25). One of these has a double substitution of arginines for the glutamines at positions 402 and 418 (designated " 402 R/ 418 R"). This 402 R/ 418 R mutant is as effective as wild-type mSR-BI in functioning as an LDL receptor in transfected mammalian cells, mediating high affinity LDL binding, uptake of metabolically active cholesterol from LDL, and efflux of cholesterol to LDL; however, it has lost most of the corresponding HDL receptor activity exhibited by the wild-type receptor (23). Lysates from insect cells expressing mSR-BI-t1, the 402 R/ 418 R mutant form of mSR-BI-t1, or no recombinant protein (control) were reconstituted into liposomes, and the abilities of the liposomes to bind 125 I-HDL or 125 I-LDL were determined. Fig. 6B shows that, as expected, the specific binding of 125 I-LDL (5 g of protein/ml) to the 402 R/ 418 R mutant (light gray bar) was similar to that of mSR-BI-t1 (open bar) and substantially greater than that of the receptor-free control (dark filled bar). In contrast, Fig. 6A shows that the binding of 125 I-HDL (5 g of protein/ml) to the 402 R/ 418 R mutant was much lower than that of mSR-BI-t1 and was not significantly different from that of the receptor-free control. Because the binding specificities of the whole insect lysatereconstituted liposomes reflected the specificities of the corresponding intact mammalian cells, it seems likely that the mechanism of SR-BI-mediated lipoprotein binding in the lysate-reconstituted liposomes is similar to that in intact mammalian cells.
CD36 is a class B scavenger receptor that is structurally similar to SR-BI and shares a number of ligand-binding activities (  HDL to cells (26,27). Fig. 2, B and C show that CD36-containing insect cell lysate-reconstituted liposomes bound 125 I-HDL (filled squares) at a level comparable to that of the corresponding mSR-BI-t1-containing liposomes ( Fig. 2A, open squares) 2B, filled circles, U max ϳ 3.5 ng of protein/assay) compared with that of mSR-BI-t1-containing liposomes ( Fig. 2A, open circles, U max ϳ 106.3 ng of protein/assay). Thus, the lipid uptake activities of mSR-BI-t1 and CD36 in whole insect lysate-reconstituted liposomes reflected their activities when expressed on the surfaces of intact mammalian cells.
Taken together the data in Figs. [2][3][4][5][6] show that the whole cell lysate/liposome assay recapitulated many key features of cellular SR-BI-mediated HDL receptor activity: 1) 125 I-HDL binding and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL were high affinity and saturable (6); 2) binding reached a steady state more rapidly than lipid uptake (6); 3) LDL was a poor inhibitor of HDL binding and lipid uptake (6, 25); 4) lipid transfer was specifically inhibited by an anti-SR-BI blocking antibody (25); 5) lipid uptake was substantially more temperature-sensitive than binding (34) 2 ; and 6) lipid transfer occurred via selective uptake (6). Furthermore, in this assay the activities of a mutant form of SR-BI ( 402 R/ 418 R) and the homologue CD36 also recapitulated those in intact cells. Therefore, the reconstituted liposome system provides a valid assay for the HDL binding and lipid uptake activities of SR-BI in detergent-solubilized, whole cell lysates.
One-step Immunoaffinity Purification of mSR-BI-t1-To examine the function of SR-BI using the reconstituted liposome assay in the absence of other proteins, we isolated highly purified mSR-BI-t1 from HEK[mSR-BI-t1] cells using a modified version of the rhodopsin purification of Reeves et al. (38). The receptor was purified from octyl glucoside-solubilized cells by immunoaffinity chromatography using an antibody to its Cterminal rhodopsin peptide epitope tag. Results of a typical purification are shown in Fig. 7, in which specimens obtained throughout the procedure were fractionated by SDS-PAGE and visualized by silver staining of the gel. Both the starting cell lysate (lane 1) and the initial column flow-through (material not retained by the column, lane 2) were highly complex protein mixtures. After washing the column so that no additional protein was detected by silver staining (lane 3), we eluted bound material with the rhodopsin C-terminal peptide and recovered virtually homogeneously pure mSR-BI-t1 (lanes 4 -8). The electrophoretic mobility of the bulk of the purified material corresponded to 82 kDa, as expected from previous studies (6). Immunoblotting of a replicate gel with anti-mSR-BI KKB-1 antibody established that the major protein band detected by silver staining, as well as the very low abundance minor bands (e.g. see lane 5), was either mSR-BI-t1 or minor proteolytic or aggregated forms of mSR-BI-t1 (not shown). Immunoblotting also revealed that a very small amount of mSR-BI-t1 was present in the column flow-through and wash fractions (corresponding to lanes 2 and 3, not shown). We estimate from quantitative immunoblotting that the overall recovery of purified mSR-BI-t1 from the lysate was ϳ80%. The yield of mSR-BI-t1 was 100 -150 g/liter of suspension cell culture.
Activity of Purified mSR-BI-t1 Reconstituted into Liposomes-With the availability of pure mSR-BI-t1 and a fully validated in vitro reconstituted liposome assay, we were able to address the main question of this study: could mSR-BI, independently of any other protein, mediate HDL binding and selective lipid uptake? Fig. 8 shows the results of an experiment in which we measured as a function of ligand concentration 125

I-HDL binding to (open squares) and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL by (open circles) liposomes
reconstituted with the immunoaffinity-purified receptor. Fig. 8 (open squares) shows that specific 125 I-HDL binding was of high affinity (apparent K d of 11.9 g of protein/ml) and saturable (B max ϳ 16.6 ng of protein/assay). The maximal binding, corrected for the amount of protein incorporated in the liposomes, was 311 ng of protein/g of liposome protein, a value 65-fold higher than the corresponding value for binding of 125 I-HDL to liposomes reconstituted with HEK[mSR-BI-t1] whole cell lysate (B max of 4.8 ng of protein/g of liposome protein). Fig. 8 (open circles)  ether uptake from [ 3 H]CE-HDL was also high affinity and saturable. For [ 3 H]cholesteryl ether uptake, the apparent K d was 13.8 g of protein/ml and the U max was 134 ng of protein/ assay or 2500 ng of protein/g of liposome protein. The relative amount of lipid uptake was substantially greater than that of binding, with a ratio of the maximal [ 3 H]CE-HDL uptake/ 125 I-HDL binding of 8, clearly showing that the pure receptor mediated selective lipid uptake from HDL. The HDL binding and lipid uptake activities of the mSR-BI-t1-containing liposomes increased linearly with the amount of mSR-BI-t1 incorporated into the liposomes (data not shown).
We conducted three additional experiments to determine if the selective lipid uptake activity of the purified receptor reconstituted into liposomes exhibited characteristics similar to those of SR-BI in intact cells. First, we compared the abilities of unlabeled HDL and LDL to inhibit 125 I-HDL binding and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL (10 g of protein/ml). Fig. 9 shows that excess unlabeled HDL (open squares and circles) effectively inhibited 125 I-HDL binding (panel A) and [ 3 H]cholesteryl ether uptake (panel B), whereas unlabeled LDL (filled squares and circles) did not. These results were similar to those observed using liposomes reconstituted with mSR-BI-t1-containing mammalian cell lysates (Fig. 5B) and mSR-BI expressed in intact mammalian cells (6,25). We did note that the extent of LDL competition for 125 I-HDL binding and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL varied somewhat from experiment to experiment (maximum percent inhibition of 8 -20% for binding and 0 -25% for uptake), perhaps reflecting the effects that small changes in the state of the HDL (e.g. oxidation) can have on lipoprotein binding affinities (41). Second, we tested the anti-SR-BI antibody KKB-1-specific inhibition of [ 3 H]CE-HDL uptake. Fig. 10 shows that the KKB-1 antibody inhibited the [ 3 H]cholesteryl ether uptake by mSR-BI-t1-reconstituted liposomes, but the control antibody from preimmune serum did not. These results were similar to those obtained with the mSR-BI-t1 mammalian cell lysate-reconstituted liposomes (Fig. 5C) and mSR-BI transfected intact mammalian cells (25). Third, we examined the temperature dependence of 125 I-HDL binding to (Fig. 11A)  In the same experiment, the background values for uptake by control (untransfected HEK293S cell-derived) liposomes in the absence of competitor or the presence of 400 g of protein/ml of either HDL or LDL were 7.7, 7.1, or 6.7 ng of protein/assay, respectively (not shown). C, the mSR-BI-t1-containing (open bars) or control (filled bars) liposomes were incubated in triplicate at 37°C for 2 h with 10 g of protein/ml of [ 3 H]CE-HDL in the absence (None) or presence of either 400 g of protein/ml of unlabeled HDL, 50 g/ml blocking antibody KKB-1, or 50 g/ml control antibody from preimmune serum. The liposomes were isolated and washed by filtration, and the amounts of [ 3 H]cholesteryl ether uptake were determined as described under "Experimental Procedures." The 100% of control value for lipid uptake in the absence of inhibitors was 19.1 ng of protein/assay. Error bars represent the range of variations in the triplicate determinations. uptake was substantially reduced at the lower temperature (35.7 versus 14.6 ng of protein/assay at 37°C and 0°C, respectively, 59.1% reduction). These findings were consistent with those obtained with mSR-BI-t1-containing cell lysatereconstituted liposomes (Fig. 4) and intact cells (34). 2 Thus, SR-BI needed no additional protein co-factors to allow it to mediate HDL binding and selective cholesteryl ether uptake. DISCUSSION We have developed an in vitro reconstituted liposome assay to measure the ligand binding and lipid transport activities of detergent-solubilized forms of the HDL receptor SR-BI. An epitope-tagged form of the recombinant receptor (mSR-BI-t1) in detergent-solubilized whole insect or mammalian cell lysates, or mSR-BI-t1 purified by immunoaffinity chromatog-raphy essentially to homogeneity, can be assayed using this liposome system. The liposome assay recapitulated many features of the HDL receptor activity of SR-BI expressed in intact cells (5): 1) 125 I-HDL binding and [ 3 H]cholesteryl ether uptake from [ 3 H]CE-HDL were high affinity and saturable, 2) binding reached a steady state more rapidly than lipid uptake, 3) LDL, although a tight binding ligand of the receptor, was a poor inhibitor of HDL binding and lipid uptake from HDL, 4) lipid transfer was specifically inhibited by an anti-SR-BI blocking antibody, 5) lipid uptake was substantially more temperature-sensitive than HDL binding, and 6) lipid transfer occurred via selective uptake. Furthermore, the novel binding properties of a doubly mutated receptor (Q402R and Q418R, normal LDL binding with the loss of  HEK[mSR-BI-t1] cells were grown in suspension culture and lysed, and the lysates were subjected to immunoaffinity chromatography using the 1D4 monoclonal anti-C-terminal epitope tag antibody. The bound protein was eluted from the column with the peptide epitope as described under "Experimental Procedures." Samples of the cell lysate, column flow-through (unbound material), column wash, and peptide-eluted fractions (10 l of each fraction) were fractionated by 10% SDS-PAGE, and the proteins in the gel were visualized using a Bio-Rad silverstaining kit. The mobilities of molecular weight standards are indicated on the left.
HDL binding (23)) were reproduced in the liposome assay, as was the ability of the SR-BI homologue CD36 to bind HDL but not mediate efficient lipid uptake (26,27,42). It should be possible to use this assay to explore in detail many features of the mechanism underlying the complex ligand binding and lipid transport activities of SR-BI. It should be noted that, prior to and after the discovery of SR-BI and its role as an HDL receptor for selective lipid uptake, several groups re-ported that intact, not solubilized, membranes isolated from adipocyte, steroidogenic, or hepatic tissues or cells could mediate selective lipid uptake (32,(45)(46)(47)(48)(49).
The first question regarding the mechanism of SR-BI activity that we addressed using this assay was: does SR-BI require the cooperation of other proteins to mediate HDL binding and selective lipid uptake, or are these activities autonomous properties of the receptor (independent of other proteins)? Several observations raised the possibility that the complex activities mediated by SR-BI might require the intervention or collaboration of other proteins. For example, SR-BI can be found in specialized membrane microdomains, including caveolar-like domains under certain conditions in some cultured cells (50) and microvillar channels in mammalian steroidogenic cells in vivo (51,52). Indeed, SR-BI expression in cultured cells can induce the formation of microvillar channel-like structures (43) and plays a role in the formation and/or stability of microvillar channels in steroidogenic cells in vivo (53,54). In addition, SR-BI has been shown to bind, via its C-terminal cytoplasmic tail, to a multiple PDZ domain-containing scaffold protein called CLAMP (55). The only unequivocal way to determine if other proteins are essential for key SR-BI activities was to examine the activity of the receptor purified away from other proteins. The liposome assay together with the highly efficient immunoaffinity purification of an epitope tagged form of the receptor permitted an unequivocal answer to this question. Essentially homogeneously pure mSR-BI-t1 incorporated into phosphatidylcholine/cholesterol liposomes did mediate high affinity and saturable binding of 125   capacity to function as an HDL receptor for selective lipid uptake in liposomes without the required intervention of other proteins or cellular structures or compartments. Therefore, these results suggest that SR-BI itself is primarily responsible for the lipid transfer step during SR-BI-mediated selective lipid uptake in vivo.
Acknowledgments-We thank M. Penman and S. Xu for help with lipoprotein preparations and advice regarding assays; T. Nieland and W. Evans for assistance; and X. Gu, B. Trigatti, W. Schneider, A. Rigotti, K. Cha, E. Getmanova, C. Bruel, S. Bell, F. Yang, G. Paradis, and R. Cook for advice and generously providing reagents. We are especially grateful to P. Reeves and G. Khorana who very generously provided advice, access to equipment, and assistance in adapting to

FIG. 10. Blocking antibody (KKB-1) inhibition of [ 3 H]CE-HDL uptake by liposomes reconstituted with immunoaffinity-purified mSR-BI-t1.
Immunoaffinity-purified mSR-BI-t1 isolated from HEK[mSR-BI-t1] cells was reconstituted into liposomes as described under "Experimental Procedures." Control liposomes without added receptor were prepared in parallel. The liposomes were incubated in triplicate at 37°C for 2 h with 10 g of protein/ml of [ 3 H]CE-HDL in the absence (Ϫ) or presence (ϩ) of unlabeled HDL (400 g of protein/ml), KKB-1 antibody (50 g/ml), or preimmune IgG (50 g/ml), isolated, and washed by filtration, and the amounts of [ 3 H]cholesteryl ether uptake were determined as described under "Experimental Procedures." The 100% of control value for lipid uptake by the receptor-containing liposomes in the absence of inhibitors was 39.9 ng of protein/assay. Error bars represent the range of variations in triplicate determinations.