SR-BI-directed HDL-cholesteryl ester hydrolysis

We have examined the metabolic fate of HDL cholesteryl ester (CE) delivered to cells expressing scavenger receptor class B type I (SR-BI). Comparison of SR-BI with a related class B scavenger receptor, CD36, showed a greater uptake and a more rapid and extensive hydrolysis of HDL-CE when delivered by SR-BI. In addition, hydrolysis of HDL-CE delivered by both receptors was via a neutral CE hydrolase. These data indicate that SR-BI, but not CD36, can efﬁciently direct HDL-CE to a neutral CE hydrolytic pathway. In contrast, LDL-CE was delivered and hydrolyzed equally well by SR-BI and CD36. Hydrolysis of LDL-CE delivered by SR-BI was via a neutral CE hydrolase but that delivered by CD36 occurred via an acidic CE hydrolase, indicating that SR-BI and CD36 deliver LDL-CE to different metabolic pathways. Comparison of inhibitor sensitivities in Y1-BS1 adrenal, Fu5AH hepatoma, and transfected cells suggests that hydrolysis of HDL-CE delivered by SR-BI occurs via cell type-speciﬁc neutral CE hydrolases. Further-more, HDL-CE hydrolytic activity was recovered in a membrane fraction of Y1-BS1 cells. These ﬁndings suggest that SR-BI efﬁciently delivers HDL-CE RHC-80267. Y1-BS1 (cid:9) ACTH or Fu5AH were incubated in the presence or absence of 400 (cid:2) M UBP or 10 (cid:2) M RHC-80267 in serum-free medium for 30 min prior to the addition of 6 mM [ 3 H]DFP and then incubated for 1 h at 37 (cid:3) C. Protein lysates were made as described above, and the proteins were separated on 4–15% gradient gels. The gels were impregnated with ﬂuor using EN 3 HANCE, dried and exposed to ﬁlm for 7 days at (cid:6) 80 (cid:3) C. The experiment repeated three times with the same results.

Scavenger receptor class B type I (SR-BI) is the first physiologically relevant HDL receptor to be identified and is one of the major participants in reverse cholesterol transport (RCT). RCT is the process whereby cholesterol is transported from peripheral tissues via plasma HDL to the liver for bile acid synthesis and secretion or to endocrine tissues for steroid production (1). A variety of studies indicate that SR-BI mediates the selective uptake (SU) of HDL cholesteryl ester (CE), the process whereby the HDL core CE is taken into the cell without degradation of the whole particle and its apolipoproteins. The SU pathway is the major route for delivery of HDL-CE to the liver and steroidogenic tissues of rodents in vivo and in vitro (2)(3)(4)(5)(6)(7). In addition, SR-BI stimulates the bidirectional flux of free cholesterol (FC) between cultured cells and lipoproteins (8)(9)(10)(11)(12), an activity that may be responsible for the rapid hepatic clearance of FC from plasma HDL and its resultant secretion into bile (13,14). SR-BI's role in the RCT pathway may be one of the contributing factors to how HDL protects against the development of atherosclerosis.
Although numerous studies have been conducted on the mechanism of SR-BI-mediated HDL-CE SU, little is known about the fate of the HDL-CE once incorporated into the cell. It is well known that LDL-CE internalized via the LDL receptor pathway is hydrolyzed by a lysosomal acid cholesteryl ester hydrolase (ACEH) (15)(16)(17). In contrast, a limited number of studies indicate that HDL-CE internalized via SU is hydrolyzed extra-lysosomally (18) by a neutral cholesteryl ester hydrolase (NCEH) (19,20). To complicate matters, total NCEH activity varies greatly among different tissues, and there appear to be multiple NCEH activities in some tissues, as suggested by studies with inhibitors and a limited amount of biochemical characterization. In general, tissues with an active cholesterol metabolism, such as liver, adrenal gland, ovary, testis, macrophage, and the gastrointestinal tract have the highest CEH activity. Interestingly, these are the tissues that express SR-BI at the highest levels, particularly adrenal and liver cells (21,22), raising the possibility that a specific NCEH in these cells may be responsible for hydrolysis of HDL-CE delivered by SR-BI. Identifying the enzyme(s) responsible for the hydrolysis of HDL-CE in cells, which nat-332 Journal of Lipid Research Volume 44, 2003 urally express SR-BI, is essential for understanding the complex metabolism of HDL-CE and the role played by SR-BI in RCT.
The present study investigates the fates of HDL and LDL-CE when presented to the cell by either SR-BI or CD36 receptors. Our results indicate that HDL-CE taken up by SR-BI is rapidly directed to a membrane-associated NCEH, whereas HDL-CE delivered by CD36 is slowly and inefficiently hydrolyzed. Differential inhibitor sensitivities in various cell types suggest that efficient SR-BIdirected HDL-CE hydrolysis occurs via cell type-specific NCEH(s).

Immunoblot analysis
Transiently transfected cells expressing SR-BI (in 35 mm wells) were washed twice with PBS (pH 7.4) and lysed with 300 l NP-40 cell lysis buffer (24,25) containing 1 g/ml pepstatin, 0.2 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, and 10 g/ml aprotinin. Protein concentrations were determined by the Lowry method (26). Immunoblots with antibodies directed to SR-BI and CD36 confirmed their expression (data not shown).

CE hydrolysis assays
Transiently transfected COS-7 cells (in 22 or 35 mm wells) were preincubated for 4 h (unless otherwise indicated) at 37 Њ C in the presence of an acyl CoA:cholesterol acyltransferase (ACAT) inhibitor, CP113,818 (gift from Pfizer, Groton, CT), at a final concentration of 2 g/ml in serum free medium. The cells were washed twice and [ 3 H]CE-hHDL 3 particles were added at a concentration of 10 g protein/ml in serum-free medium containing 2 g/ml ACAT inhibitor. After incubation at 37 Њ C for 4 h (unless otherwise indicated), the medium was removed and the monolayers were washed three times with PBS and allowed to dry. The lipids were extracted from the cell monolayers with isopropanol. One tenth of the extracted lipids was counted for total cell associated radioactivity. The remainder of the lipids was dried down under N 2 and resuspended in 100 l chloroform-methanol (1:1, v/v) with FC and [ 3 H]CE as unlabeled standards. The samples were run on activated silica gel ITLC plates in a TLC tank with 95% petroleum ether-5% diethyl ether-1% acetic acid as the solvent system. The plates were air dried, and spots corresponding to CE and FC were visualized with iodine, cut from the strip, and analyzed by liquid scintillation counting to determine the fraction of [

[ 3 H]diisopropylfluorophosphate labeling of cellular proteins
Y1-BS1 or Fu5AH cells were plated in 35 mm wells. The following day the Y1-BS1 cells were incubated in the presence or absence of ACTH (100 nM) overnight in growth medium. On day 3, the confluent cells were incubated in the presence or absence of 400 M diethylumbelliferyl phosphate (UBP) or 10 M RHC-80267 in serum-free medium for 30 min prior to the addition of 6 mM [ 3 H]diisopropylfluorophosphate (DFP) (Amersham Pharmacia Biotech) and then incubated for 1 h at 37 Њ C. Protein lysates were made, as described above, and the proteins were separated on preformed (Biorad Laboratories, Inc., Hercules, CA) 4-15% gradient gels. The gels were impregnated with fluor using EN 3 HANCE (NEN Life Science Products, Inc., Boston, MA), dried, and exposed to film for 7 days at Ϫ 80 Њ C.

Membrane purification
Four of eight flasks of Y1-BS1 cells were treated with ACTH (100 nM) overnight in growth medium. The cell monolayers were washed with serum-free medium, placed on ice, and scraped into 10 ml 10 mM Tris, pH 7.5/2.0 mM MgCl 2 . Cells were broken by N 2 cavitation after holding at 500 psi for 20 min, and unbroken cells removed by centrifugation for 10 min at 1,000 g at 4 Њ C. The supernatant was centrifuged for 30 min at 100,000 g in a Ti75 rotor at 4 Њ C to pellet the microsomal membrane fraction. Pellets were washed by suspension in 3 ml 10 mM Tris, pH 7.5/2.0 mM MgCl 2 and centrifuged again for 30 min at 100,000 g in a Ti75 rotor at 4 Њ C. Microsomal membrane pellets were suspended in 10 mM Tris, pH 7.5/2.0 mM MgCl 2 and divided into three equal aliquots. One aliquot was heated for 20 min at 65 Њ C, one was preincubated for 15 min with 1,000 M UBP, and the last aliquot was used as the untreated control. The samples were incubated for 1.5 h at 37 Њ C with [ 3 H]CE-hHDL 3 at a final concentration of 10 g/ml. Samples were then washed twice at 100,000 g as described above to remove excess labeled HDL particles. Pellets were resuspended in 10 mM Tris pH 7.5-2.0 mM MgCl 2 and aliquots were removed for protein determination by Lowry assay and for lipid extraction. One tenth of the extracted lipids was counted for total radioactivity and the remainder was analyzed by ITLC to determine the fraction of [

SR-BI efficiently directs HDL-CE for hydrolysis but CD36 does not
Previous studies from several laboratories have indicated that SR-BI and CD36, a closely related class B scavenger receptor, bind HDL with high affinity (21,23,(30)(31)(32). However, SR-BI mediates the SU of nonhydrolyzable HDL cholesteryl oleyl ether (HDL-COE) (33) or 1,1'-dioctadecyl-3,3,3 Ј ,3 Ј-tetramethylindocarbocyaine perchlorate (21) with a much higher efficiency than CD36, although most of the HDL-COE delivered by CD36 is also via the SU pathway as opposed to an endocytic pathway (disposition of HDL-COE with CD36: 91% SU vs. 0.1% endocytic uptake, with the remaining 8.9% COE associated with intact HDL particles within the cell or on the cell surface, unpublished data). This observation raised the question of whether there might be a difference in the hydrolysis of HDL-CE delivered to the cell via SR-BI versus CD36. In order to address this question, COS-7 cells were transiently transfected to express SR-BI or CD36 and then compared for their ability to hydrolyze HDL [ 3 H]CE. As previously demonstrated for HDL-COE, SR-BI delivered much more HDL-CE to COS-7 cells than CD36 (Fig. 1A). Notably, however, 75% of the HDL-CE delivered by SR-BI was hydrolyzed, whereas less than 20% of the HDL-CE delivered by CD36 was hydrolyzed (Fig. 1B). As seen in previous studies in transiently transfected COS-7 cells (23,33), when double-labeled 125 I-dilactitol tyramine-[ 3 H]CE-hHDL 3 particles were used in these hydrolysis experiments, similar levels of SR-BI and CD36 were expressed on the cell surface as determined by the amount of HDL binding (data not shown). Despite similar amounts of cell surface HDL binding with these two receptors, HDL-CE is not effi-ciently transferred from the HDL particle to the cell membrane by the CD36 receptor, nor is it efficiently hydrolyzed. Therefore, we believe that HDL-CE delivered to COS-7 cells via SR-BI is efficiently directed toward hydrolysis but HDL-CE delivered by CD36 is not.
Previous experiments revealed that high-efficiency SU of HDL-COE requires the extracellular domain (ECD) of SR-BI but not the transmembrane domains or the cytoplasmic C-terminal tail (23,32). This was also the case for a spectrum of other SR-BI-stimulated activities, including bidirectional FC flux, membrane cholesterol accumulation, and cholesterol oxidase sensitivity of plasma membrane FC (33). The transmembrane and cytoplasmic domains of CD36 may be substituted for those in SR-BI with no loss of these activities, whereas the converse substitutions did not confer these activities on the ECD of CD36. In order to address whether SR-BI's ECD also directs the metabolism of the HDL-CE, hydrolysis experiments were performed using SR-BI-CD36 chimeric receptors (Fig.  1C) (33). As with the other SR-BI-mediated activities, the data showed that the ECD, but not its transmembrane domains or the cytoplasmic tails, is responsible for highefficiency HDL-CE hydrolysis (Fig. 1D). With chimeric receptors containing the ECD of SR-BI (SR-BI, SR-CDT, CD-SR-CD, and SR-BI-D1), 80% of the delivered HDL-CE was hydrolyzed, whereas with chimeric receptors containing CD36s ECD (CD36, CD-SRT, SR-CD-SR) only 30% of the delivered HDL-CE was hydrolyzed. Substitution of the SR-BI C-terminal tail for that of CD36 (SR-CDT) gave a small increase in HDL-CE hydrolysis by CD36, but deletion of the C-terminal tail from SR-BI (SR-BI-D1) had no effect (Fig. 1D). Additionally, as compared with SR-BI, no reduction of hydrolysis activity was observed with chimeric receptors containing the ECD of SR-BI when cell incubations were carried out at 22ЊC, at which only 30% of the SR-BI-delivered HDL-CE was hydrolyzed (data not shown). Thus, as previously established with other SR-BI-dependent activities, efficient HDL-CE hydrolysis requires the ECD, but not the other domains of SR-BI.

SR-BI directs HDL-CE more efficiently for hydrolysis than LDL-CE
Previous work indicated that SR-BI was able to mediate selective COE uptake from LDL (34) and stimulate the esterification of cellular cholesterol in response to LDL (12). Also, the fractional transfer of lipoprotein COE by SR-BI was found to be ‫-7ف‬fold greater from the core of an HDL than from an LDL particle (34). Similar to HDL-COE, most of the LDL-COE delivered to the cell by SR-BI or CD36 was delivered in the SU pathway (SR-BI: 81% SU vs. 1.0% endocytic uptake vs. 18% associated with intact LDL; CD36: 83% SU vs. 2.0% endocytic uptake vs. 15% associated with intact LDL; data not shown). These data raised the question: Is LDL-CE hydrolyzed as efficiently as HDL-CE? To address this question, we measured the hydrolysis of LDL [ 3 H]CE delivered to COS-7 cells by SR-BI versus CD36. Note that the contribution of the LDL receptor in these experiments was minimized by growing the cells in serum-containing medium to down-regulate the LDL receptor and by subtracting the background uptake and hydrolysis with the vector transfected cells. The results revealed that, in contrast to HDL-CE, LDL-CE was delivered to the cell ( Fig. 2A) and hydrolyzed (Fig. 2B) approximately as well by both SR-BI and CD36 receptors.
Furthermore, the extent of LDL-CE hydrolysis was about half that of HDL-CE delivered by SR-BI. Similar amounts of CE by mass were delivered by SR-BI to the cells from HDL versus LDL (note: in Figs. 1, 2 the specific activity of HDL-CE was ‫5ف‬ times that of LDL-CE, see Materials and Methods). Therefore, the difference in the degree of hydrolysis was not due to saturation of the hydrolase by an excess of CE delivery from the LDL core. In other words, even when SR-BI delivers approximately the same amount of CE from HDL as LDL, the % CE hydrolysis is higher for HDL-CE than for LDL-CE. Hence, SR-BI directs HDL-CE to hydrolysis more efficiently than LDL-CE.

HDL-CE and LDL-CE delivered by SR-BI are hydrolyzed by a NCEH
LDL-CE internalized via the LDL receptor pathway is hydrolyzed by a lysosomal ACEH (15)(16)(17). In contrast, several studies have suggested that HDL-CE internalized via SU is hydrolyzed extralysosomally (18) by an NCEH (19,20). In order to address the question of whether SR-BI directs lipoprotein CE to an ACEH or an NCEH for hydrolysis, we performed HDL-CE and LDL-CE hydrolysis assays in the presence or absence of either chloroquine or UBP. Chloroquine inhibits ACEH localized to lysosomes or other low pH compartments, whereas UBP is an inhibitor of NCEH activity in hepatocytes, MA-10 Leydig cells, and macrophages (19,(35)(36)(37). As previously observed, with CD36 only 15% of the delivered HDL-CE was hydrolyzed in a 4 h period compared with 75% with SR-BI in the COS-7 cells (Fig. 3A). Furthermore, the HDL-CE that was hydrolyzed after delivery by CD36 was poorly inhibited by UBP or chloroquine. With SR-BI-delivered HDL-CE, where hydrolysis was high, there was no inhibition by chloroquine and only 13% inhibition of hydrolysis by UBP (Fig. 3A). In multiple experiments, the inhibition by UBP of hydrolysis of HDL-CE delivered by SR-BI was 21% (Table 1). These data indicate that most of the HDL-CE loaded into the plasma membrane by SR-BI is not acted on by an ACEH. The partial sensitivity to UBP for SR-BI delivered HDL-CE hydrolysis suggests that a fraction of the CE hydrolysis in COS-7 cells is due to a NCEH activity, although this activity in COS-7 cells appears significantly less sensitive to inhibition compared with hepatocytes (19).
In order to address the question of whether SR-BI delivered LDL-CE is hydrolyzed in the lysosomal pathway, we performed LDL-CE hydrolysis assays in the presence or absence of either chloroquine or UBP. In contrast to HDL-CE, we observed significant inhibition of CE hydrolysis by chloroquine but no inhibition by UBP when LDL-CE was delivered to COS-7 cells by CD36 (Fig. 3B). These data suggest that most of the CD36-delivered LDL-CE is hydrolyzed by an ACEH. Thus, the chloroquine sensitivity of LDL-CE hydrolysis with CD36 suggests different pathways for LDL-CE (chloroquine sensitive) and HDL-CE (chloroquine insensitive) hydrolysis with this receptor. In contrast, when LDL-CE was delivered by SR-BI, we observed partial inhibition of LDL-CE hydrolysis by UBP ‫)%03ف(‬ and no inhibition by chloroquine (Fig. 3B). These data suggest that neither LDL-CE nor HDL-CE delivered via SR-BI is hydrolyzed by an ACEH. As noted with HDL-CE, the partial sensitivity to UBP suggests that a fraction of the LDL-CE hydrolysis occurs via an NCEH, although this inhibitor does not efficiently block CE hydrolysis in COS-7 cells.

SR-BI-delivered HDL-CE is hydrolyzed by cell type-specific NCEHs
The fact that the NCEH activity we observed in the COS-7 cells was not inhibited efficiently by UBP suggested to us that different NCEHs may be responsible for HDL-CE hydrolysis in different cell types. Since SR-BI is not normally expressed in COS-7 cells, we asked whether CE hydrolysis is similar in cell types where SR-BI is expressed endogenously. To this end, several cell types expressing SR-BI, either naturally (Fu5AH and Y1-BS1) or after transient (COS-7) or stable transfection (WI38), were incubated in the presence or absence of inhibitors. As shown in Table 1, with ACTH-treated Y1-BS1 adrenal cells, in which HDL-CE uptake is almost entirely due to SR-BI (25), UBP inhibited HDL-CE hydrolysis by 51% but chloroquine had no effect. The lack of inhibition by chloroquine is in agreement with early studies by the laboratories of Pittman and Gwynne (18,38). The high percent inhibition by UBP confirms the role of a NCEH in the hydrolysis of HDL-CE in Y1-BS1 cells. Fu5AH cells, which also naturally express SR-BI, have an intermediate level of UBP sensitivity and a low level of chloroquine sensitivity, whereas SR-BI-expressing COS-7 and WI-38 cells exhibit low UBP and chloroquine sensitivities. The differential inhibitor sensitivities suggest that SR-BI-directed HDL-CE hydrolysis occurs via different NCEHs in various cell types. Note also that HDL-CE hydrolysis in differentiated THP-1 macrophages was substantially inhibited by both UBP and chloroquine. This is of interest because previous studies showed that macrophages process HDL by both SR-BImediated SU and by endocytic uptake (39,40), and possibly other pathways.
To further address the question of cell type-specific NCEHs, ACTH-stimulated Y1-BS1 and Fu5AH cells were incubated with a panel of NCEH inhibitors and assayed for HDL-CE hydrolysis as described above. The following were used: 1) RHC-80267, which selectively inhibits hormone-sensitive lipase hydrolysis of diacylglycerides and triglycerides in adipocytes (41); 2) masoprocol, a lipoxygenase inhibitor that inhibits the phosphorylation of hormone sensitive lipase, thereby inhibiting its activation (42); 3) papaverine HCl (Pap. HCl), a phosphodiesterase inhibitor that blocks the cellular redistribution of activated hormone sensitive lipase (43); 4) DFP and phenylmethylsulfonyl flouride (PMSF), which are active site-specific inhibitors of serine hydrolases; and 5) deoxyglucose (DG) and NaN 3 , which inhibit the translocation of hormone sensitive lipase to lipid droplets in 3T3-L1 adipocytes by abrogating cellular energy (44). The two cell types exhib-ited different sensitivities to each of the inhibitors implicating cell type-specific NCEHs in HDL-CE hydrolysis (Table 2).
In order to address the concern that the differences in inhibitor sensitivities between the Y1-BS1 and Fu5AH cells could be due to differences in cell permeability, which could cause differential inhibitor accessibility to cellular proteins, we incubated ACTH-unstimulated and -stimu-  Note: The amount of HDL-CE and LDL-CE uptake per experiment varied up to 2-fold among multiple experiments due to differences in receptor expression levels; however, the percentage of hydrolysis varied usually less than 5% from one experiment to another. Values designated asterisk differ from respective control value, P Ͻ 0.001. Values designated by a dagger differ from that designated by a double dagger, P Ͻ 0.005.
ACTH, adrenocorticotropic hormone; CE, cholesteryl ester; PMA, phorbol 12-myristate 13-acetate; SR-BI, scavenger receptor class B type I; UBP, diethylumbelliferyl phosphate. Cells were preincubated in the presence or absence of 400 M UBP or 50 M chloroquine and an ACAT inhibitor at a final concentration of 2 g/ml for 2 h at 37ЊC then incubated for 2 h at 37ЊC with 10 g/ml [ 3 H-CE]hHDL 3 in the presence or absence of UBP or chloroquine and the ACAT inhibitor. After incubation, the cells were processed to determine the percentage of the total number of counts that were each cell line and, as expected, a different pattern of labeled proteins between the two cell types was observed (Fig. 4). Labeling of most (UBP) or some (RHC-80267) of the proteins in both cell types was inhibited by these agents (Fig. 4). PMSF also inhibited protein labeling by [ 3 H]DFP (data not shown). These data support the idea that DFP, UBP, and RHC-80267 are fully accessible to the cellular hydrolases in the intact Y1-BS1 and Fu5AH cells.
The high lipid solubility of these agents is also supportive of this point. Nevertheless, we cannot completely rule out the possibility that cell type differences in inhibitor access to NCEHs could contribute to the quantitative differences among cell types. The same approach could not be used with papaverine and masoprocol since these inhibitors act indirectly to inhibit lipase activity.

Hydrolysis of HDL-CE occurs in a crude microsomal membrane fraction
Preliminary data suggested that SR-BI-mediated HDL-CE SU and subsequent CE hydrolysis are coupled and consequently may be occurring in a membrane versus a cytoplasmic compartment. Time courses at 37ЊC in COS-7 cells expressing SR-BI revealed that HDL-CE SU and hydrolysis occurred concomitantly (data not shown). In an attempt to dissociate HDL-CE uptake and hydrolysis and to address the compartmentalization of HDL-CE hy-  NaN 3 ), and an ACAT inhibitor (2 g/ml) for 2 h at 37ЊC then incubated for 2 h at 37ЊC with 10 g/ml [ 3 H-CE]hHDL 3 in the presence or absence of inhibitors. After incubation, the cells were processed to determine the percentage of the total number of counts that were [ 3 H]FC. Values are the mean Ϯ SD of six to nine replicates each. drolysis in a cell type that endogenously expresses SR-BI, hydrolysis assays were carried out with Y1-BS1 cells at various temperatures using a 1.5 h incubation period, and for various times at 22ЊC. The results show that HDL-CE uptake and CE hydrolysis are temporally coupled (Fig.  5A) and coincident over a wide temperature range (Fig.  5B). HDL-CE hydrolysis appears to occur rapidly following CE uptake. Note, however, that inhibitors of HDL-CE hydrolysis appear relatively ineffective in blocking HDL-CE SU. For example, the data show substantial inhibition of HDL-CE hydrolysis by DFP (Fig. 5C) over a broad concentration range with little inhibition of HDL-CE uptake. With UBP (Fig. 5D), substantial inhibition of CE hydrolysis occurs with no effect on HDL-CE uptake up to 100 M and partial inhibition of uptake above that. These data indicate that while HDL-CE uptake and hydrolysis are closely coupled, CE hydrolysis can be substantially inhibited with no effect on HDL-CE uptake. This led us to conclude that HDL-CE SU is not initially dependent on its subsequent hydrolysis. In addition, HDL-CE hydrolysis might be occurring in a membrane compartment since its hydrolysis is so closely coupled to its uptake.
To address whether SR-BI-directed HDL-CE hydrolytic activity was membrane associated, a crude microsomal membrane fraction prepared from ACTH-treated or -untreated Y1-BS1 cells was incubated with [ 3 H]CE HDL for 1.5 h at 37ЊC followed by washing of the membrane fraction. As shown in Fig. 6, the microsomal membrane fraction contained an HDL-CE hydrolytic activity that was sensitive both to heat and inhibition by UBP. The HDL-CE hydrolytic activity in the ACTH-treated Y1-BS1 microsomal membranes showed a modestly but consistently higher UBP inhibition than membranes from Y1-BS1 cells that were not treated with ACTH (62.5 Ϯ 2.8% vs. 50.8 Ϯ 0.4%, P Ͻ 0.006). A similar modest difference was seen with intact Y1-BS1 cells treated or not with ACTH (51 Ϯ 5% vs. 35 Ϯ 6%, P Ͻ 0.05) ( Table 1). We were unable to reliably estimate [ 3 H]CE HDL hydrolytic activity in the supernatant fractions of these preparations due to the background radioactivity from the labeled HDL.
Interestingly, when [ 3 H]CE was introduced directly into the Y1-BS1 membrane sample via a small volume of dimethylsulfoxide, significant CE hydrolysis also occurred (35%), but the hydrolysis was not sensitive to inhibition by UBP (data not shown). This result suggests the presence of multiple NCEH enzymes in the microsomal membrane preparation. As judged by UBP inhibition, NCEHs, responsible for HDL-CE hydrolysis in Y1-BS1 membranes, were not active in the hydrolysis of [ 3 H]CE delivered without the aid of the HDL particle. Therefore, SR-BI-directed HDL-CE hydrolysis occurs in a microsomal membrane fraction by an NCEH that is sensitive to the means of substrate presentation.

DISCUSSION
Although numerous studies have been conducted on the mechanism of SR-BI-mediated lipoprotein COE SU, little is known about the fate of lipoprotein CE once incorporated into the cell. The aim of this study was to investigate the fate of lipoprotein CE when presented to the cell by either SR-BI or CD36 scavenger receptors. As previously demonstrated for HDL-COE, SR-BI delivered more HDL-CE to COS-7 cells compared with CD36. In addition, the HDL-CE delivered to the cells by SR-BI was more efficiently hydrolyzed than the HDL-CE delivered by CD36. These data suggest that SR-BI not only delivers HDL-CE to the plasma membrane, but that it somehow directs the CE in such a way that it is efficiently hydrolyzed, a property not shared by CD36. These results complement previous data showing that delivery of HDL FC by SR-BI results in a 3-5-fold greater fractional esterification of newly influxed FC compared with CD36 (33,45). Furthermore, it has been shown that although CD36 can bind HDL with high affinity, it does not play a direct role in hepatic HDL metabolism, as judged by short-term expression studies using adenovirus vectors (46). Together these data support the idea that, unlike CD36, SR-BI delivers HDL-FC and CE into a metabolically active membrane pool. Whether this metabolically active pool reflects the localization of SR-BI in a physically distinct membrane domain compared with CD36 is unclear.
The data presented here indicate that 70-80% of the HDL-CE that is delivered to the cell by SR-BI is rapidly hydrolyzed. Interestingly, we have not seen a greater extent of hydrolysis of HDL-CE even with incubations up to 4 h. This may indicate that a fraction of the HDL-CE delivered to the cell by SU may be transferred to other cellular sites without hydrolysis. This possibility is supported by a recent study of SR-BI-expressing CHO cells that showed trafficking of CE to internal membranes in a complex with annexin II, cyclophilin 40, caveolin, and cyclophilin A (47). These findings suggest that there are two pathways for trafficking of SR-BI-delivered HDL-CE: one involving rapid hydrolysis and another involving transport to internal membranes prior to further metabolism.
Our results with the chimeric receptors revealed that, like its other lipid transport functions (23,32,33), SR-BI's ECD, but not its transmembrane domains or cytoplasmic tails, is responsible for efficient HDL-CE hydrolysis. It is unlikely that SR-BI itself has hydrolytic activity since UBP inhibits CE hydrolysis in a cell-specific manner. In addition, analysis of the SR-BI amino acid sequence failed to identify the core sequence motif that is common to all lipid hydrolases. More likely, SR-BI delivers the CE into the membrane or alters membrane organization in such a way as to facilitate CE presentation to the hydrolase. Alternatively, SR-BI may recruit an NCEH to the membrane domain into which CE is delivered. Although the mechanism of the efficient HDL-CE hydrolysis with SR-BI is unclear, it is likely that the rapid CE hydrolysis facilitates FC trafficking and utilization within the cell.
In contrast to HDL-CE, LDL-CE was delivered to the cell and hydrolyzed approximately as well by both SR-BI and CD36 receptors. Furthermore, the extent of LDL-CE hydrolysis was about half that of HDL-CE delivered by SR-BI. In this case, similar amounts of CE mass were delivered to the cells by HDL versus LDL. Therefore, the difference in the degree of hydrolysis was not due to saturation of hydrolytic activity. In other words, even when SR-BI delivers approximately the same amount of CE from HDL as LDL, the percent of CE hydrolysis is higher for HDL-CE than LDL-CE. Hence, SR-BI directs HDL-CE more efficiently than LDL-CE for hydrolysis. It has been hypothesized that there are different binding sites on SR-BI for HDL versus LDL (48). LDL, therefore, may bind to an SR-BI site that delivers the lipoprotein CE less efficiently and does not direct the CE efficiently for further metabolism. However, HDL may bind to a site on SR-BI that delivers the CE much more efficiently and allows it to be directed for metabolism.
Hydrolysis of HDL-CE delivered by SR-BI was not inhibited by chloroquine in COS-7, WI38, Fu5AH, or Y1-BS1 cells, indicating that hydrolysis is not via an ACEH. The lack of inhibition by chloroquine was previously noted in rat adrenal cells (18,38) and appears to be a general feature of the SR-BI-mediated delivery process in other cell types as well. In contrast, UBP inhibited hydrolysis of SR-BI-delivered HDL-CE in all cell types tested, although with variable efficiencies (Table 1). Up to 75% inhibition by UBP was seen in Y1-BS1 cells (Fig. 5D). Recent studies with the hormone-sensitive lipase (HSL)-deficient mouse Fig. 6. Hydrolysis of HDL-CE in a membrane fraction of Y1-BS1 cells. Crude microsomal membrane fractions prepared from Y1-BS1 cells with or without ACTH treatment for 24 h were isolated by centrifugation at 100,000 g. The membranes were incubated with 10 g/ml [ 3 H]CE-hHDL 3 for 1.5 h at 37ЊC in the presence or absence of UBP or heating to 65ЊC and reisolated at 100,000 g. A Bligh Dyer lipid extraction was performed on the membranes, after which the lipids were processed to determine the percentage of the total number of counts that were [ 3 H]FC. Values represent the mean Ϯ SD of four experiments. * Differs from respective control, P Ͻ 0.001. by guest, on July 19, 2018 www.jlr.org Downloaded from