THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 265, No. 10, Issue of April 5, pp. 5546-5553, 1990 Printed in U.S.A.

The Efflux of Lysosomal Cholesterol from Cells*

(Received for publication, September 25, 1989)

William J. Johnson+, George K. Chacko, Michael C. Phillips, and George H. Rothblat From the Department of Physiology and Biochemistry, The Medical College of Pennsylvania, Philadelphia, Pennsylvania 19129

To gain insight into the transport of sterol from lysosomes to the plasma membrane, we studied the efflux of lysosomal free cholesterol from intact Fu5AH rat hepatoma cells to high density lipoprotein (HDL) and other extracellular acceptors that promote sterol desorption from the plasma membrane. The procedures involved pulsing cells at 15 “C with low density lipo- protein that had been reconstituted with [3H]choles- teryl oleate and then incubating the cells at 37 “C! in the presence of a sterol acceptor, while monitoring both the hydrolysis of [‘Hlcholesteryl oleate in lysosomes and the efflux of the resulting [3H]free cholesterol to the acceptor. After warming cells to 37 “C, rapid hy- drolysis of [3H]cholesteryl oleate began after lo-20 min, and the lysosomally generated [3H]free choles- terol became available for efflux after an additional delay of 40-50 min. The kinetics of hydrolysis and the delay between hydrolysis and efflux were unchanged over a wide range of HDL3 concentrations (10-1000 pg of protein/ml), and with acceptors that do not inter- act with HDL-specific cell surface binding sites (phos- pholipid vesicles, dimethyl suberimidate cross-linked HDL). In addition, the delivery of lysosomal cholesterol to the plasma membrane was unaffected when cellular cholesterol content was elevated 2.6-fold above the normal control level, or when the activity of cellular acyl-coenzyme A/cholesterol acyltransferase (ACAT) was stimulated with exogenous oleic acid. We conclude that in the Fu5AH cell, a maximum of 40-50 min is required for the transport of cholesterol from lyso- somes to the plasma membrane and that this transport is not regulated in response to either specific extracel- lular acceptors or the content of sterol in cells. The lack of effect of increased ACAT activity implies that the pathway for this transport does not involve passage of sterol through the rough endoplasmic reticulum, the subcellular location of ACAT.

Mammalian cells acquire sterols in the following three ways: by cle nouo synthesis in the endoplasmic reticulum, by the degradation of low density lipoprotein (LDL)’ and other

* This research was supported by National Institutes of Healt,h Program Project Grant HL22633 and Grant ROl-HL37550. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “uduer- tisement” in accordance with 18 USC. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed. 1 The abbreviations used are: LDL, low density lipoprotein; FC,

free (unesterified) cholesterol; r[“H-COILDL, reconstituted [3H]cho- lestervl oleate-labeled LDL: HDL. high densitv lioourotein: BSA, bovini serum albumin; DMS, dimeihylsuberimidate; TLC, thin layer chromatography; CoA, coenzyme A; MEM, Eagle’s minimum essen- tial medium; HEPES, N-2-hydroxyethylpiperazine-N’-2-ethylsul- fonic acid; ACAT, acyl-CoA:cholesterol acyltransferase; RER, rough endoplasmic reticulum.

sterol-rich lipoproteins in lysosomes (Spady and Dietschy, 1989), and by the influx of free (unesterified) sterol from lipoproteins into the plasma membrane (Johnson et al., 1988). Sterols are utilized mostly in the plasma membrane (Lange et al., 1989) where they have a major influence on the fluidity of the lipid bilayer (Phillips, 1972). The metabolism of sterols occurs in the endoplasmic reticulum (the site of bile acid synthesis in hepatocytes and of sterol esteriflcation in a variety of cells) and in the mitochondrion (the site of steroid- ogenesis in steroidogenic cells) (Carey, 1982; Suckling and Stange, 1985; Lambeth et al., 1987). Sterols also function in the endoplasmic reticulum as feedback regulators of choles- terol biosynthesis (Davis and Poznansky, 1987). These obser- vations imply that there is a complex system for the transport of sterol between compartments in the cell and that the control of this transport may be important in the maintenance of membrane structure and the control of sterol metabolism.

Most studies on the intracellular transport of sterol have focused on the movement of newly synthesized cholesterol from the endoplasmic reticulum to the plasma membrane. The major features of this process are that it occurs rapidly (half-times range from 10 min to l-2 h, depending on cell type), appears to be mediated by lipid-rich vesicles and re- quires metabolic energy and a temperature above 15 “C (Daw- idowicz, 1987; DeGrella and Simoni, 1982; Kaplan and Si- moni, 1985a; Lange and Mattheis, 1984; Lange and Steck, 1985). Another potentially important aspect of intracellular sterol transport is the movement of sterol from lysosomes to the cell surface. Until recently, this process has received very little attention. Interest has increased with the discovery that in fibroblasts from individuals with the inherited Niemann- Pick C disorder, LDL is taken up and degraded normally, but the resulting free cholesterol is retained for an abnormally long time in lysosomes. This results in sluggish regulation of sterol metabolism and delayed availability of the cholesterol for desorption from the plasma membrane (Blanchette- Mackie et al., 1988; Liscum and Faust, 1987; Liscum et al., 1989; Pentchev et al., 1987).

Most non-hepatic cells have little or no ability to oxidize sterol. In this situation, sterol homeostasis is maintained by a balance of sterol synthesis, LDL uptake, and sterol influx on the one hand and the efflux of sterol to plasma lipoproteins on the other. The efflux of cellular sterol is a component of reverse cholesterol transport, the process by which cholesterol is removed from cells and transported to the liver for elimi- nation from the body (reviewed by Rothblat et al., 1986). This process appears to be dependent on high density lipoprotein (HDL), which in tissue culture is much more efficient than other lipoproteins at promoting sterol efflux (reviewed by Phillips et al., 1987). The participation of HDL in reverse cholesterol transport is thought to explain the well-estab- lished negative correlation between plasma HDL levels and atherosclerosis, a disease characterized by the accumulation of cholesterol and cholesteryl ester in the walls of arteries

5546

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

Efflux of Lysosomal Sterol

(Gordon et al., 1986). The mechanism of sterol efflux involves the passive desorption of sterol from the plasma membrane into the extracellular aqueous phase, followed by rapid ab- sorption by an HDL particle. The same mechanism allows efflux to a variety of phospholipid-containing acceptors. In all essential respects, this process is not different from the simple diffusion of sterol between phospholipid vesicles sus- pended in water (Phillips et al., 1987). This suggests that the efflux of sterol from the plasma membrane is not mediated in any way, and therefore is not subject to any strict regulation.

Recently, Oram, Bierman, and colleagues (Slotte et al., 1987; Aviram et al., 1989) have reported that in fibroblasts, endothelial cells, and macrophages the specific binding of HDL to the cell surface stimulates the translocation of newly synthesized sterol from the endoplasmic reticulum to the plasma membrane. This observation implies that some as- pects of the intracellular trafficking of sterol may be regulated by signals from the cell’s environment. It also suggests a unique role for HDL in reverse cholesterol transport, since the enhancement of sterol translocation to the cell surface might enhance the ability of HDL to remove cholesterol from the internal pools of sterol that are involved in the formation of cholesteryl ester.

In the present work, we studied the efflux of lysosomal cholesterol from intact Fu5AH rat hepatoma cells to extra- cellular sterol acceptors, with the intent of gaining insight into the transport of sterol from lysosomes to the plasma membrane. The methods involved the selective labeling of lysosomal cholesterol using LDL containing [3H]cholesteryl oleate, followed by careful monitoring of the kinetics of hy- drolysis of the labeled ester in lysosomes and the kinetics of efflux of the resulting free [3H]cholesterol in the presence of HDL and other extracellular acceptors. For this approach to yield reliable information on the intracellular transport of lysosomal sterol, an important condition was that the move- ment of sterol from lysosomes to the extracellular acceptor not be rate limited by slow desorption from the plasma membrane. This requirement argued against the use of fibro- blasts and a variety of other commonly studied cells, in which the minimum half-time for efflux of plasma membrane cho- lesterol is about 10 h (Phillips et al., 1987) and thus much longer than known half-times for intracellular transport proc- esses. A much better choice appeared to be the Fu5AH cell, in which the minimum half-time for efflux of plasma mem- brane cholesterol is about 1 h (Rothblat and Phillips, 1982; Johnson et al., 1986). In experiments with this cell, we dem- onstrated a delay of 40-50 min between the generation of labeled free cholesterol in lysosomes and the availability of the sterol for efflux to extracellular acceptors, suggesting 40- 50 min as an upper limit for the lysosome-to-plasma mem- brane transport time. The delay was independent of the type of acceptor in the medium, the concentration of the acceptor, and the level of sterol mass in the cells, suggesting the operation of a constitutive, rather than regulated, transport process.

EXPERIMENTAL PROCEDURES

Materials-N-Ethylmaleimide, thrombin, oleic anhydride, oleic acid, bovine serum albumin (BSA), cholesterol (FC), cholesteryl oleate, and cholesteryl methyl ether were from Sigma. Henarin- Sepharose was from Pharmadia LKB Biotechnology-Inc. Dimethyl suberimidate (DMS) was from Pierce Chemical Co. Svlon CT (silv- lation reagent for the treatment of test tubes used tb reconstitute LDL) was from Supelco, Inc. (Bellefonte, PA). Potato starch, sol- vents, and other unlabeled reagents were from Fisher. Plates for thin layer chromatography (TLC) were from Analabs, Inc. (North Haven, CT) and Gelman Sciences (Ann Arbor, MI). ]4-Y?]Cholesterol ([4- “C]FC) and [1,2-“H]FC were from Du Pont-New England Nuclear.

[‘C]FC was purified by preparative TLC (on Silica Gel G with a mobile phase of diethylether) before use. [1,2-3H]cholesteryl oleate was prepared from the labeled free sterol by derivatization with oleic anhydride in anhydrous benzene (Lentz et al., 1984), followed by isolation of the labeled ester by preparative TLC (on Silica Gel G with benzene/hexane, 60:40 (v/v) as the mobile phase). Eagle’s min- imum essential medium (MEM, from GIBCO) was supplemented with 50 pg/ml of gentamicin (Tri Bio Laboratories, State College, PA) and buffered to pH 7.4 with either 25 mM NaHC03 or 14 mM HEPES. Bovine calf and fetal bovine sera were also from GIBCO.

Lipoprotein-deficient plasma and all lipoproteins were isolated by sequential ultracentrifugation (Hatch and Lees, 1968) from fresh human plasma that had been treated with 5 mM N-ethylmaleimide to irreversibly inhibit 1ecithin:cholesterol acyltransferase (Johnson et nl., 1986). Density intervals for different lipoprotein and plasma fractions were as follows: LDL, 1.019-1.066 g/ml, HDLs, 1.125-1.21 g/ml, and lipoprotein-deficient plasma, >1.21 g/ml. Before use, HDLa was chromatographed twice on heparin-Sepharose (Quarfordt et al., 1978) to remove any particles containing apolipoproteins B or E. DMS was used to covalently cross-link theproteins in HDL3, accord- ing to the procedure of Chacko et al. (1988). The reaction conditions were chosen to cross-link proteins within individual HDL particles, without cross-linking particles to each other. Control HDL for this modification was treated identically except that the exposure to DMS was omitted. After all chromatographic and modification procedures, HDL, was dialyzed against MEM-HEPES, sterilized by passage through a 0.45~pm pore diameter filter, and stored at 4 “C until use. Human lipoprotein-deficient plasma was dialyzed against 0.15 M NaCl, clotted with thrombin to produce lipoprotein-deficient serum, which was separated from the clot by centrifugation, and then steri- lized by filtration (0.45 am). Lipoprotein-deficient serum was stored at -15 “C until use. When needed in native form, LDL was dialyzed against 0.15 M NaCl, filter sterilized (0.45 Fm), and stored at 4 “C.

For reconstitution, LDL was dialyzed and treated essentially as described by Krieger et al. (1978), with the following differences. 1) Shell freezing of the LDL-potato starch mixture was done with dry ice-acetone, rather than liquid nitrogen. 2) [1,2-3H]cholesteryl oleate (1 mCi/G mg) was used, instead of cholesteryl linoleate. 3) Elution of the reconstituted LDL ($H-CO]LDL) from potato starch was done with 10 mM Tris-HCI buffer, pH 8.4, rather than 10 mM Tricine buffer. The final step in the reconstitution of LDL was the removal of residual starch by passage of the r[“H-CO]LDL solution through a 0.8-pm pore diameter filter. Typically, the r[“H-CO]LDL filtrate contained 1 mg/ml protein, 1.2 mg (200 PCi) of [3H]cholesteryl oleate/ ml, and no detectable free cholesterol bv mass. The 3H label in rLDL had normal /3 migration on agarose gel electrophoresis, and >97% of the label migrated as steryl ester when the lipids were extracted and analyzed by TLC. The proteins in r[“H-CO]LDL were identical to those in native LDL, as judged by polyacrylamide gel electrophoresis in the presence of mercaptoethanol and sodium dodecyl sulfate, followed by staining with Coomassie Blue.

Delipidized serum protein was prepared from bovine calf serum (Arbogast et al., 1976, as modified by Capriotti and Laposata, 1987), dissolved in MEM-HEPES, filter sterilized (0.45 pm), and stored at -15 “C! until use. Small unilamellar vesicles of egg phosphatidylcho- line were prepared by sonication in 0.15 M NaCl (Barenholz et al., 1977). After centrifugation to remove titanium particles, the vesicle suspension was dialyzed against MEM-HEPES, filter sterilized (0.45 pm), and stored at 4 “C. Cholesterol-rich liposomes (FC/phosphati- dylcholine 2 2 mol/mol) and control liposomes (FC = 0) were pre- pared with egg phosphatidylcholine in 0.15 M NaCl as described (Arbogast ef al., 1976) and filter sterilized (0.45 pm) before use.

Incubation Methods-All media were based on MEM and were buffered to pH 7.4 with either bicarbonate or HEPES. Bicarbonate- buffered media were used in an atmosphere of 5% Coo/air; HEPES- buffered media were used in air. The incubation temperature was 37 “C, and the cells were in monolayer culture, unless otherwise indicated. Cultures of Fu5AH cells were grown to confluence in T75 flasks containing 15 ml of 5% bovine calf serum. Cells were usuallv plated 4 days before experiments, by dispersing the cells from one or two confluent T75 flasks with trvnsin (0.5 ma/ml). susnendina the cells in fresh 5% calf serum medium, and then’distributing the-cells into 22-mm wells of 12-well tissue culture plates at a ratio of 200- 300 wells/ T75 flask (1 ml of medium/well). The cells were allowed to attach and grow for 2 days. The partially confluent monolayers were rinsed once with room temperature MEM-HEPES, and then, to upregulate LDL receptors and simultaneously to label the cells with [“C]FC, they were incubated for 2 days with 1 ml/well of medium

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

5548 Efflux of Lysosomal Sterol

containing 10 mg of protein/ml of lipoprotein-deficient serum and 0.025-0.05 &i/ml of [“C]FC (55 Ci/mol) dispersed with 0.1% ethanol and 5 pg/ml of egg phosphatidylcholine. After this incubation (on the day of an experiment), the cells were rinsed three times with room temperature MEM-HEPES and then incubated 1 h at 37 “C with medium containing 5 mg/ml of delipidized serum protein, to remove any loosely associated [“‘CIFC. Afterward, the cells were rinsed once at room temperature and twice at 15 “C with MEM-HEPES. To begin the labeling of lysosomes, the cells were incubated 5 h at 15 “C with 0.5 ml/well of medium containing 0.2% (w/v) bovine serum albumin and 5-10 pg of protein/ml of r[3H-CO]LDL. At this temperature, there was substantial uptake of r[“H-CO]LDL (approximately 500 ng of LDL protein/mg cell protein) but very little hydrolysis of its steryl ester (see “Results”). After 5 h, the r[3H-CO]LDL medium was removed, and the cells were rinsed five times at 4 “C with MEM- HEPES. Hydrolysis of [3H]cholesteryl oleate and the efflux of FC then were initiated by quickly aspirating the last rinse and applying 0.5 ml/well of prewarmed (37 “C) medium containing either HDL or some other acceptor of interest. The incubations were continued for various times (up to 8 h) at 37 “C in humidified air.

When additional chemical treatments were intended for the efflux period (e.g. exposure to oleic acid), an additional half-hour incubation at 4 “C in the presence of the desired chemical agent(s) (but in the absence of any efflux-promoting acceptor) was inserted between the rinses that followed incubation with $H-CO]LDL and the beginning of the 37 “C efflux period. In the experiments involving enrichment of cells with cholesterol, the preparation of cells differed from the above procedure as follows. 1) Plating was done 6 days before the experiment. 2) Beginning 3 days before the experiment, the cells were incubated for 2 days in medium containing 10 mg of protein/ml of lipoprotein-deficient serum. 3) Beginning 1 day before the experi- ment, cells were enriched with cholesterol by incubation for 1 day in medium containing 50 ag of protein/ml of native LDL, 100 Kg of FC/ ml of cholesterol-rich liposomes (FC/phosphatidylcholine 2 2 mol/ mol), 1 rg/ml of the Sandoz compound 58-035 (an inhibitor of acyl- CoA:cholesterol acyltransferase) dispersed with 0.5% (v/v) dimethyl sulfoxide, and 0.05 &i/ml of [14C]FC (55 Ci/mol) dispersed with 1% fetal bovine serum and 0.1% ethanol. Control (unenriched) cells were incubated identically, except that the medium lacked LDL and the liposomes contained only phosphatidylcholine. 4) After incubation with one of these media, cells were incubated in delipidized serum protein and labeled with r[3H-CO]LDL as described above, except that exposure to compound 58-035 (dispersed with dimethyl sulfox- ide) was maintained until the termination of efflux.

The efflux of cholesterol was terminated by recovering the medium from each well and chilling it to 4 “C. The efflux media were spun in a refrigerated (4 “C) centrifuge for 15 min at 3000 rpm (2200 X g) to sediment any suspended cells, and then aliquots of the media were taken for the determination of ‘H and i4C by liquid scintillation counting, and for the extraction of lipids, which were analyzed later by a combination of TLC and liquid scintillation counting to deter- mine the distribution of isotopic tracers between free and esterified sterol. Immediately after removal of the efflux media, the cells were rinsed at 4 “C, once with phosphate-buffered saline (Johnson et al., 1986) supplemented with 0.2% BSA, and twice with phosphate- buffered saline, and then cell lipids were extracted overnight at room temperature with 2 ml/well of isopropyl alcohol. Aliquots of the cell extract were taken for determination of total 3H and i4C content and the distribution of the tracers between free and esterified sterol.

Analytical Methods-Extractions, chemical assays, TLC, liquid scintillation counting, and gas-liquid chromatography were as de- scribed previously (Johnson et al., 1986), except that the internal standard for the determination of cholesterol by gas-liquid chroma- tography was cholesteryl methyl ether, rather than coprostanol. All values are the means f 1 S.D. of replicate (23) determinations.

RESULTS

In all experiments, the FuSAH cells were prelabeled for 1 or 2 days with [i4C]FC that had been dispersed with either serum or phosphatidylcholine. With either form of dispersal, [14C]FC would have entered cells mostly by diffusing into the plasma membrane. To a greater or lesser extent, the tracer then would have equilibrated with the other pools of cell cholesterol. We have assumed complete equilibration, and therefore define i4C as tracing “whole-cell” cholesterol. The results of Lange et al. (1989) suggest that nearly all (80-90%)

of the free cholesterol in mammalian cells is located in the plasma membrane, which implies that in the present studies, at least 80-90% of the [14C]FC in cells would have been in the plasma membrane at the end of the prelabeling period. Thus, the efflux of this tracer was representative essentially of the efflux of plasma membrane cholesterol. It was impor- tant in the present studies to have a relatively simple measure of the efflux of plasma membrane cholesterol because this provided one of the means of knowing whether a change in the release of lysosomally generated [3H]FC was due to a change in the rate of intracellular sterol transport or to a change in the rate of sterol desorption from the plasma membrane.

Cholesterol Efflux in the Presence of HDL3-In the presence of HDL3 at a concentration of 1 mg of protein/ml, the initial efflux of whole-cell [i4C]cholesterol was rapid (.& = 108 min), and the kinetics of release over an 8-h period suggested simple equilibration of the tracer between a single medium compart- ment and single cellular compartment (Fig. L4). These results are largely consistent with our previous studies on the efflux of cholesterol from the Fu5AH cell (Johnson et al., 1986; Karlin et al., 1987), although in the earlier studies, the tlh for efflux of whole-cell cholesterol was approximately 1 h under the present conditions. The longer tlh in the present studies may have resulted from differences in the prelabeling condi- tions.

When cells that had been prelabeled with [‘4C]cholesterol were then incubated with 10 pg of protein/ml of r[3H-CO] LDL for 5 h at 15 “C, they absorbed approximately 600 ng of LDL cholesteryl oleate/mg cell protein (equivalent to 500 ng of LDL protein/mg cell protein) (Fig. 1 legend). This result agrees well with data of Brown and Goldstein (1976) on the

14 -A.[‘4C]FC (sxchonge label) tot.2

0 -*p------o%

A-----4 medium

7 B.[%]FC (lysosomol) total

1 J

incubation time, min

FIG. 1. The efflux of whole-cell [‘“Cl and lysosomal [‘HI cholesterol from Fu5AH rat hepatoma cells in the presence of HDL3. The preparation of cells is described under “Materials and Methods.” To label cells with [3H]cholesterol, the concentration of r[3H-CO]LDL was 10 rg protein/ml. The graphs show the efflux of labeled FC into the incubation medium (A), the retention of labeled FC in cells (O), and the total recovery of labeled FC in cells + medium (O), versus time of incubation at 37 “C in the presence of human HDL3 (1000 pg protein/ml) and 0.2% (w/v) BSA. All values are the means f SD. of replicate (23) determinations. In most cases, the error bar is smaller than the symbol, and therefore cannot be seen. At f = 0, the level of cell protein was 227 C 8 &well, 97.5 + 0.1% of total cell [‘4C]cholesterol was unesterified, and the total uptake of rLDL was 7946 + 358 3H cpm/well (508 ng of LDL protein/mg cell protein). During the 480-min efflux period, the total release of un- hydrolyzed [3H]cholesteryl oleate into the medium (as determined by liquid scintillation counting and TLC analysis of medium lipids) was 160 + 99 cpm/well (2% of the initial uptake of r[3H-CO]LDL).

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

Efflux of Lysosomal Sterol 5549

uptake of native human LDL by fibroblasts at low tempera- ture. If reasonable assumptions are made concerning the difference between binding at 4 “C! and uptake at 15 “C, the result also is in good agreement with data of Friedman et al. (1987) on the binding of canine LDL to the Fu5AH cell. At 15 “C, 97% of the 3H-labeled cholesterol in cells was esterified at the end of the 5-h uptake period, implying that very little rLDL had been delivered to lysosomes. This material probably was located mostly in prelysosomal endocytic compartments (Brown and Goldstein, 1976). Upon warming to 37 “C in the presence of 1 mg/ml of HDL3 (Fig. lB), there was a lo-20 min delay, and then rapid hydrolysis of [3H]cholesteryl oleate began. The maximum extent of hydrolysis was achieved in 4 h, with approximately 75% conversion of the ester to FC (over several experiments, the maximum extent of hydrolysis ranged from 68 to 90%). Half-maximal hydrolysis was achieved in about 60 min. In the first 60 min of incubation, the efflux of [3H]FC was negligible; then starting at 60 min, there was rapid movement of the lysosomally generated sterol into the medium. Thus, there was an approximate 40-min delay between the beginning of rapid generation of [3H]FC and the beginning of release of the sterol to the medium. Maximal net release of [3H]FC was achieved in 8 h, at which time the distributions of [i4C]FC and [3H]FC between cells and medium were identical. During the 8-h efflux period, there was 2% release of intact [3H]cholesteryl oleate from the cells (data not shown). Presumably, this was due to a small amount of loosely bound r[3H-CO]LDL that was not removed during the rinses that followed incubation at 15 “C. The omission of HDL from the medium had no effect on the hydrolysis of [3H]cholesteryl oleate but reduced the efflux of [‘*C]FC and [3H]FC to negligible levels (data not shown).

In Fig. 2, data from the initial 2 h of the experiment described in Fig. 1 are replotted using an enlarged time scale. This graph emphasizes the 40-min delay between the gener- ation of [“H]FC in lysosomes and the desorption of the sterol into the medium. This result implies that it takes a maximum of 40 min for cholesterol to be transported from the lysosome to the plasma membrane in the Fu5AH cell.

The dependence of cholesterol efflux on HDL3 concentra- tion is shown in Fig. 3. Three concentrations (10, 100, and 1000 pg of protein/ml) were compared. The data at the highest concentration are identical to those in Fig. 1. The results suggest only two effects as HDL, concentration was lowered:

FIG. 2. Enlarged view of the efflux of whole-cell and lyso- somal cholesterol in the presence of HDL$. The data are from Fig. 1, with the time scale enlarged to emphasize the 40-min delay between the beginning of hydrolysis of [3H]cholesteryl oleate (t = 20 min) and the beginning of efflux of [3H]FC (t = 60 min).

incubation time. min

efflux of cholesterol from Fu5AH cells with various concentrations of HDL in the extracellular medium was measured as described in Fig. 1. Each efflux medium con- tained 0.2% BSA and HDLa at a concentration of 10,100, or 1000 pg of protein/ml. The graphs show efflux of labeled FC into media (A, A) at the three HDL concentrations and the total recovery of labeled FC in cells + medium for incubations with [HDL] = 1000 +g/ml (Cl). The hydrolysis of [3H]cholesteryl oleate and the total recovery of labeled FC were the same at all three concentrations of HDL and in the complete absence of HDL (data not shown). A, whole-cell [‘“Cl FC. B, lysosomal [‘H]FC.

1) the efflux of whole-cell cholesterol was reduced (Fig. 3A), and 2) the efflux of lysosomal cholesterol after 60 min was reduced, in parallel with the reduction of whole-cell choles- terol (Fig. 3B). Other features of the data (the kinetics of hydrolysis of [3H]cholesteryl oleate, and the 40-min delay between [“Hlcholesteryl oleate hydrolysis and [“H]FC efflux) were not affected by changes in the HDLB concentration (Fig. 3B). The similarity between changes in the efflux of whole- cell and lysosomal cholesterol was very striking. At 10 and 100 wg of HDL3/ml, the initial rate of release of whole-cell cholesterol was 3 and 21%, respectively, of its value at 1000 fig/ml (Fig. 3A). After 120 min of incubation, the correspond- ing values for the accumulation of [“H]FC in media were 4 and 22% (Fig. 3B). These results suggest that lysosomal cholesterol was transported to the cell surface at the same rate, regardless of HDL concentration, and that differences in the efflux of the lysosomal sterol were due completely to differences in the rate of desorption of the sterol from the plasma membrane.

Dependence of Cholesterol Efflux on Acceptor Properties- HDL binds specifically and with high affinity to the Fu5AH cell (Gottlieb and Marsh, 1987; Karlin et al., 1987). Data from Oram and colleagues (Slotte et al., 1987; Aviram et al., 1989) suggest that in several types of cells, the specific binding of HDL stimulates the transport of newly synthesized sterol to the plasma membrane. To determine whether the intracellular transport of lysosomal sterol is dependent on HDL binding, we examined the efflux of lysosomal cholesterol in the pres- ence of HDLB that had been pretreated with DMS to inhibit its specific interaction with the cell surface. The validity of this approach was established in previous studies, which showed that DMS produced a 90% reduction in the affinity of HDL for binding sites on fibroblasts and rat liver mem- branes (Chacko et al., 1988), without affecting the ability of the lipoprotein to promote sterol efflux from the plasma membrane (Johnson et al., 1988). In the present experiment (Fig. 4), untreated (control) and DMS-treated HDL, were used at a concentration of 1 mg of protein/ml. The inhibition

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

5550 Efflux of Lysosomal Sterol

0 100 200 300 4w 500

incubation time, min

FIG. 4. Effect of treating HDL, with DMS on the efflux of whole-cell and lysosomal cholesterol. Conditions were similar to those described in Fig. 1. HDLa was treated with DMS as described by Chacko et al. (1988). Each efflux medium contained 0.2% BSA and 1000 pg of protein/ml of either DMS-HDL, (open symbols) or unmodified HDL3 (closed symbols) from the same donor. A, whole- cell [14C]FC. B, lysosomal [3H]FC.

of HDL binding with DMS did not affect the efflux of whole- cell cholesterol (Fig. 4A), the hydrolysis of LDL cholesteryl oleate in lysosomes, or the efflux of lysosomally generated free cholesterol (Fig. 4B). Thus, none of these processes appear to be regulated in the Fu5AH cell by the specific binding of HDL to the cell surface.

To further explore the issue of acceptor specificity, HDL, was compared with a completely nonspecific acceptor, unila- mellar vesicles prepared from egg phosphatidylcholine. Pre- liminary experiments suggested that the efflux of plasma membrane cholesterol was 20 times more rapid with HDL, than with vesicles, when the two acceptors were present at equal phospholipid concentrations. To compensate for this difference, the efflux of lysosomal cholesterol was examined at vesicle and HDLB concentrations of 1000 and 50 pg of phospholipid/ml, respectively. The efflux media were not supplemented with BSA, to avoid any unintentional addition of apolipoproteins, which sometimes contaminate BSA (Fain- aru et al., 1981) and might interact with specific cell surface receptors. The initial efflux of whole-cell cholesterol with the two acceptors was identical (Fig. 5A). The net release of [‘“Cl FC continued for a longer period with the vesicle-containing medium, probably due to its greater capacity for solubilizing sterol. The hydrolysis of LDL-derived [3H]cholesteryl oleate was identical with the two acceptors (data not shown). In the first 4 h of incubation, the efflux of lysosomally generated free cholesterol was identical (Fig. 5B). After 4 h, there was greater release to the medium-containing vesicles, resembling the difference in efflux of whole-cell cholesterol noted above. Thus, the vesicles were effective in the removal of lysosomal cholesterol from cells. This removal differed from that pro- moted by HDLs in ways that could be attributed completely to differences in the desorption of sterol from the plasma membrane. The results of this experiment and those in Fig. 4 suggest that the transport of lysosomal cholesterol to the cell surface in the Fu5AH cells does not depend in any way on the presence of specific acceptors in the incubation medium.

Effect of Increased Cell Sterol on Cholesterol Efflux-The enrichment of cells with cholesterol is known to induce changes in several aspects of cellular sterol metabolism. To test for possible effects of enrichment on the intracellular

0 100 200 300 400

incubation time. min

0

FIG. 5. Efflux of whole-cell cholesterol and lysosomal cho- lesterol, comparing HDL, (A) and egg phosphatidylcholine (PC) small unilamellar vesicles (SUP’) (A) as acceptors. Con- ditions were similar to those described in Fig. 1, except the efflux media did not contain BSA, and the concentrations of HDLZ and vesicles were 50 and 1000 fig of phospholipid/ml, respectively. The concentration of HDL,, in terms ofprotein, was 121 fig/ml. Hydrolysis of LDL-derived [3H]cholesteryl oleate was the same with both accep- tors (data not shown). A, whole-cell [14C]FC in media. B, lysosomal [3H]FC in media.

transport of lysosomal sterol, Fu5AH cells were incubated in a cholesterol-rich medium that also contained the Sandoz compound 58-035 (to inhibit sterol esterification) and [‘“Cl FC (to label whole-cell cholesterol). Control (unenriched) cells were prepared under similar conditions that did not produce enrichment (for details see “Materials and Methods” and Fig. 6 legend). While maintaining the exposure to compound 58 035, the enriched and control cells were incubated with r[3H- CO]LDL at 15 “C (to label lysosomes with [3H]FC) and then with 1 mg/ml of HDL, at 37 “C (to promote efflux). At the beginning of the efflux period, the enriched cells contained 2.6 times as much FC as control cells (20 and 53 pg of FC/mg of cell protein, respectively). During incubation of control cells with HDL3, the hydrolysis of [“Hlcholesteryl oleate (Fig. 6B) and the efflux of whole-cell and lysosomal FC (Fig. 6, A and C) were essentially identical to earlier results (e.g. Fig. 1). The delay between the production of [3H]FC in lysosomes and its availability for efflux was about 50 min in the present experiment (Fig. 6, B and C). The enrichment of cells with cholesterol had the following effects. 1) The fractional efflux of whole-cell cholesterol was reduced by 28% (Fig. 6A). 2) The uptake of rLDL was reduced 38% and the total produc- tion of [3H]FC was reduced 46%, although the rate of hydrol- ysis of [3H]cholesteryl oleate during the initial 40 min of incubation was unaffected by enrichment (Fig. 6B and leg- end). 3) The delay between the beginning of rapid [3H]cho- lesteryl oleate hydrolysis and rapid [3H]FC efflux was not affected by the enrichment, although at incubation times greater than 90 min, the fractional release of lysosomal [3H] FC was reduced by about 20% (Fig. 6, B and C). Thus, the delay between the generation of [3H]FC in lysosomes and its availability for efflux was the same in control and enriched cells. The fractional efflux of lysosomal cholesterol from enriched cells was somewhat less than from control cells. A similar reduction was apparent in the fractional efflux of whole-cell cholesterol. Results similar to these were also ob- tained at a much lower HDLB concentration (132 rg of pro- tein/ml) and with vesicle acceptors at 1 mg/ml (data not

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

Efflux of Lysosomal Sterol 5551

14 B

I

--!J--- 4 E 12 ,_/o--

5 10

incubation time, Ini”

FIG. 6. Effect of cellular cholesterol enrichment on the ef- flux of whole-cell and lysosomal cholesterol. The incubation conditions and the preparation of control (open symbols) and choles- terol-enriched (closed symbols) cells are described under “Materials and Methods.” Cells were exposed to compound 58-035 (a specific inhibitor of acyl-CoA:cbolesterol acyltransferase) beginning with the enrichment (or control) incubation and extending through the efflux incubation. Cell parameters at the beginning of the efflux period were as follows: 7) for control cells, protein = 0.33 f 0.02 mg/well, FC = 20 f 3 pg/mg protein, total [‘“Clcholesterol = 15,236 + 228 cpm/well (free/total = 0.998 + O.OOl), and total [sH]cholesterol = 20,990 -C 994 cpm well (free/total = 0.044 + 0.004). 2) For enriched cells, protein = 0.33 + 0.01 mg/well, FC = 53 f 3 rg/mg protein, total [‘“Cl cholesterol = 8,652 + 139 cpm/well (free/total = 0.997 + O.OOl), and total [3H]cholesterol = 12,948 f 304 cpm/weIl (free/total = 0.063 _t 0.008). The efflux medium contained 1 mg of protein/ml of HDLJ, 0.2% BSA, and 1 @g/ml of compound 58-035. A, the efflux of [‘“Cl FC, expressed as fraction of [i4C]FC initially present in cells. B, the hydrolysis of LDL-derived [“Hlcholesteryl oleate. The reduced pro- duction of [3H]FC in enriched cells was due mostly to reduced LDL uptake, but was also due partially to a lower efficiency of hydrolysis in enriched cells (final total [3H]FC/initial 3H in cells = 0.68 and 0.60 in control and enriched cells, respectively). C, the efflux of [3H]FC, expressed as fraction of the total [3H]FC present in cells + medium at t = 480 min of incubation (14,350 f 767 and 7,785 + 50 cpm/well with control and enriched cells, respectively). FCh, free cholesterol.

shown). The results of this experiment suggest that the en- richment of the Fu5AH cells with cholesterol has little or no influence on the transport of sterol from lysosomes to the plasma membrane. The enrichment produces a small reduc- tion in the fractional release of lysosomal cholesterol, but this seems to be due to sluggish desorption of sterol from the plasma membrane.

Effect of Increased ACAT Activity on Cholesterol Efflux- The only known metabolism of cholesterol in the Fu5AH cell is esterification to fatty acid, catalyzed by the enzyme acyl- CoA:cholesterol acyltransferase (ACAT). In other cells, ACAT is located in the rough endoplasmic reticulum, RER (reviewed by Suckling and Stange, 1985). In Fu5AH and other cells, ACAT activity can be increased by providing oleic acid in the incubation medium (McCloskey et al., 1988). The activity can be inhibited specifically with the Sandoz com-

pound 58035 (Ross et al., 1983). Recognizing that LDL- derived cholesterol becomes available for esterification in cells (Tabas et al., 1985) and that a variety of components destined for the plasma membrane pass through the RER (Pfeffer and Rothman, 1987; Kobayashi and Pagano, 1989), we postulated that a step in the transport of lysosomal cholesterol to the cell surface might be delivery to the RER. To test this hy- pothesis, whole-cell and lysosomal cholesterol were labeled in Fu5AH cells by the usual procedures, and then the cells were incubated at 37 “C in the presence of BSA-solubilized oleic acid (to stimulate ACAT) and HDL, (to promote sterol ef- flux). In control incubations, conditions were identical, but the stimulation of ACAT was prevented with compound 58- 035. Assuming passage of lysosomal sterol through the RER and a uniform distribution of ACAT within this compartment, the stimulation of ACAT was expected to reduce the delivery of lysosomal cholesterol to the plasma membrane, an effect that would be apparent as reduced efflux.

The stimulation of ACAT was assessed by plotting the quantity of esterified [Wlcholesterol in cells uersus time of incubation at 37 “C (Fig. 7A). Initially, 3.8% of the [‘“Cl cholesterol in cells was esterified. In the absence of compound 58-035 (solid symbols and line), a significant stimulation of esterification was apparent within 60 min. Nearly maximal esterification was achieved after 4 h, at which time the level

2

0 4

d- ,A’

4. ” C

J 0 100 200 300 400 500

incubation time. min

FIG. 7. Influence of ACAT on the efflux of whole-cell and lysosomal cholesterol. Conditions were similar to those described in Fig. 1, except that the efflux media contained 1000 pg of protein/ ml of HDLa, 1% (w/v) BSA, 100 pg/ml oleic acid (Na’ salt), 0.5% (v/ v) dimethyl sulfoxide (used to disperse compound 58-035), and where indicated (dashed lines, open symbols) 1 pg/ml of the Sandoz com- pound 58-035. In addition, between the rinses that followed labeling with r[sH-CO]LDL and the efflux period, cells were incubated for 30 min at 4 “C in medium containing 1% BSA, 100 fig/ml of oleic acid, 0.5% dimethyl sulfoxide, and, if efflux was to occur in the presence of compound 58-035, 1 pg/ml of this inhibitor. The graphs show labeled cholesterol in different locations and forms versus incubation time at 37 “C in the presence of HDL+ A, quantity of esterified [‘“Cl cholesterol in cells. B, [i4C]FC in media. C, lysosomal 13H]FC in medium (A, A) and in cells + medium (U, n ).

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

5552 Efflux of Lysosomal Sterol

of esterified [Wlcholesterol was 14% of the total [Wlcholes- terol initially present in the cells. The change in [“‘Clester content was equivalent to the esterification of 1.7 pg of cholesterol/mg of cell protein. The stimulation of ACAT was prevented with compound 58-035 (Fig. 7A, dashed line).

The stimulation had no perceptible effect on the efflux of whole-cell [‘YJ]FC within the first 2 h of incubation (Fig. 7B), suggesting that the [W]FC for esterification was not being drawn directly from the plasma membrane but must have been drawn from intracellular pools not in rapid equilibrium with the plasma membrane pool. The stimulation of ACAT seemed to produce a small decrease in the net generation of [3H]FC from LDL-derived steryl ester, but in the first 2 h of incubation, there was no perceptible effect on the efflux of the lysosomally generated sterol (Fig. 7C). After 2-4 h at 37 “C, the stimulation of ACAT produced slight reduction (6- 8%) in the net release of [3H]FC to HDL. This effect paral- leled a small reduction in net hydrolysis of [3H]cholesteryl oleate in ACAT-stimulated cells and was not noticeable until long after both the initiation of efflux of the lysosomal sterol and the stimulation of esterification of [Y!]FC. The total production of [3H]FC in lysosomes was about 0.6 pg/mg cell protein, which was far less than the quantity of [W]FC esterified during the efflux period. Thus, the lack of esterifi- cation of the lysosomal sterol cannot be attributed to insuf- ficient ACAT activity in the cells. These results show that lysosomal cholesterol is not preferentially esterified by the Fu5AH cell under conditions of increased ACAT activity, but rather is much more likely to be transported to the cell surface and made available for efflux. This finding suggests that little, if any, of the lysosomal sterol passes through the RER during transport to the cell surface.

DISCUSSION

Kinetics of Sterol Transport from Lysosomes to the Cell Surface-The present data show that in the F&AH cell 40- 50 min elapses between the generation of free cholesterol in lysosomes and the availability of this sterol for desorption from the plasma membrane. This time is independent of the type of sterol acceptor in the medium, the concentration of the acceptor, the level of sterol in the cell, and the level of ACAT activity in the cell. In the 40-50-min interval, several steps may need to occur in sequence for the delivery of lysosomal sterol to the cell surface. These are transfer of sterol from the surface of the partially degraded LDL to the lysosomal membrane, translocation across the lysosomal membrane, transport to the plasma membrane, and translo- cation across the plasma membrane. For this reason, the value of 40-50 min is an upper limit for the lysosome-to-plasma membrane transport time. Among the above steps, bilayer translocations may be unnecessary (for instance, if the trans- port mechanism involves vesicle fission/fusion events); if necessary, the translocations probably are very rapid and therefore not rate-limiting (Dawidowicz and Backer, 1981). The potential rate-limiting steps are movement of sterol from LDL to the lysosomal membrane and the transport of the sterol within the cytoplasm. The relative importance of these two steps cannot be determined using the present results. This determination probably will require subcellular fraction- ation studies in which lysosomal and plasma membrane sterol are assayed separately. In the discussion that follows, we have assumed that the 40-50-min value applies to the movement of cholesterol through the cytoplasm.

Studies on the transport of various membrane-associated materials from intracellular locations to the plasma mem- brane have suggested half-times for transport as follows: 10

min to 1 or 2 h for the transport of newly synthesized cholesterol (Degrella and Simoni, 1982; Lange and Mattheis, 1984), 30-150 min for the constitutive transport of secreted and integral membrane proteins from the RER to the cell surface (reviewed by Pfeffer and Rothman, 1987), 20-30 min for the transport of newly synthesized sphingomyelin and glucosylceramide (Lipsky and Pagano, 1985), and approxi- mately 2 min for the transport of newly synthesized phospha- tidylcholine and phosphatidylethanolamine (Kaplan and Si- moni, 1985b; Sleight and Pagano, 1983). In addition, Liscum et al. (1989) recently presented data which suggest a delay of l-2 h between the lysosomal generation of free cholesterol and its availability for efflux from normal human fibroblasts. These results and those obtained in the present study suggest that most plasma membrane constituents, with the exception of phosphatidylcholine and phosphatidylethanolamine, are moved from intracellular locations to the cell surface at sim- ilar rates. This similarity may imply partial equivalency in the mechanisms of transport.

Mechanism and Regulation of Intracellular Sterol Moue- ment-There are three conceivable mechanisms for the move- ment of lysosomal sterol to the cell surface: 1) unmediated diffusion in the aqueous cytosol, 2) mediation by a cytosolic protein carrier, and 3) transport in vesicles, either directly from lysosomes, or via another organellar system (e.g. the endoplasmic reticulum or the Golgi apparatus). The present results provide no evidence either for or against the first two possibilities. The kinetics of movement of lysosomal choles- terol are consistent with a possible relationship to the system of vesicle traffic and membrane flow that carries newly syn- thesized proteins from the endoplasmic reticulum to the Golgi apparatus and then to the cell surface (reviewed by Pfeffer and Rothman, 1987). Increased ACAT activity has very little effect on the efflux of lysosomal cholesterol (Fig. 7), which argues against passage of this sterol through the RER. Assum- ing some degree of cotransport with newly synthesized pro- tein, the lack of effect of ACAT on efflux implies passage of lysosomally generated sterol through the Golgi apparatus. Direct evidence for the delivery of lysosomal sterol to the Golgi apparatus was presented recently by Blanchette-Mackie et al. (1988), who found that during the degradation of LDL in fibroblasts, the Golgi apparatus became enriched in sterol, as assessed by staining with filipin. Evidence for vesicle- mediated transport of sterol in cells has been presented by Kaplan and Simoni (1985a) and Lange and Steck (1985), who found that newly synthesized sterol is associated with lipid- rich vesicles while in transit between the endoplasmic retic- ulum and the plasma membrane. These vesicles do not cor- respond in properties to any of the well-established intracel- lular particulate fractions, implying that they may be part of a transport system that is distinct from that for newly syn- thesized proteins. It remains to be seen whether similar lipid- rich vesicles participate in the intracellular transport of ly- sosomal sterol.

Oram and colleagues (Slotte et al., 1987; Aviram et al., 1989) have reported that in a variety of non-hepatic cells, the specific binding of HDL to the cell surface stimulates the movement of newly synthesized sterol to the plasma mem- brane. This result suggests that some aspects of intracellular sterol transport may be regulated by signals from the cell’s environment. In contrast, the present results show that lyso- somal sterol in the Fu5AH cell is moved rapidly to the cell surface under a variety of conditions, which include the ex- posure of cells to acceptors other than HDL. Thus, the trans- port of lysosomal sterol in this cell appears to be a constitutive (i.e. unregulated) process. The contrast between the present

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

Efflux of Lysosomal Sterol 5553

results and those of Oram and colleagues may have resulted from differences between hepatic and non-hepatic cells, or from differences between the mechanisms of lysosomal and biosynthetic cholesterol transport. It is reasonable to specu- late that hepatocytes may possess a unique mechanism for the transport of lysosomal cholesterol, in view of the polarized organization of these cells and their ability to secrete lysoso- ma1 contents into the bile canaliculus (Renaud et al., 1989). To resolve these possibilities, it will be useful to examine the transport of lysosomal sterol in both hepatic and non-hepatic cells and to perform direct comparisons of lysosomal and biosynthetic sterol transport in a few representative cell types.

Functional Significance of Rapid Intracellular Sterol Trans- port-During incubation of Fu5AH cells with a high concen- tration of HDL, the efflux of lysosomal sterol was limited to some extent by both the rate of sterol transport to the plasma membrane (transport time = 40-50 min) and the rate of sterol desorption from the cell surface ( tl12 2 110 min in the present study). As noted in the Introduction, the tnh for efflux of plasma membrane cholesterol is unusually short in the Fu5AH cell. With most other mammalian cells, plasma membrane cholesterol desorbs with a minimum tlj2 of about 10 h (Phillips et al., 1987). Assuming rapid intracellular transport of lyso- somal sterol in all cells, the slow desorption of plasma mem- brane cholesterol from most cells implies that typically this desorption step is absolutely rate limiting for the movement of cholesterol from lysosomes to HDL. Thus, in most cells, factors other than the rate of lysosome-to-plasma membrane sterol transport probably control the movement of lysosomal sterol out of cells to HDL. In general terms, the important controlling factors probably are those that determine the distribution of sterol between the plasma membrane and sites of esterification and that directly influence the rate of sterol desorption from the plasma membrane.

Acknowledgments-We are grateful to Timothy M. Sullivan for excellent technical assistance. We are also grateful to Dr. Jerry Faust for advice on the reconstitution of LDL and to Dr. John Heider of Sandoz Corporation for the gift of compound 58-035.

REFERENCES

Arbogast, L. Y., Rothblat, G. H., Leslie, M. H., and Cooper, R. A. (1976) Proc. Natl. Acad. Sci. U. S. A. 73. 3680-3684

Aviram,’ M., Bierman, E. L., and Oram, J. 6. (1989) J. Lipid Res. 30, 65-76

Barenholz, Y., Gibbes, D., Litman, B. J., Goll, J., Thompson, T. E., and Carlson, F. D. (1977) Biochemistv 16, 2806-2810

Blanchette-Mackie, E. J., Dwyer, N. K, Amende, L. M., Kruth, H. S.. Butler. J. D.. Sokol. J.. Comlv, M. E.. Vanier, M. T., Aurmst. J. T.; Brady, R. O:, and @en&hev,P. G. (1688) Prdc. Natl. Acid. &i. U. S. A. 85, 8022-8026

Brown, M. S., and Goldstein, J. L. (1976) Cell 9, 663-674 Capriotti, A. M., and Laposata, M. (1987) J. Tissue Culture Methods

10,219-221 Carey, M. C. (1982) in The Liuer: Biology and Pathobiology (Arias, I.

M., Popper, H., Schachter, D., and Shafritz, D. A., eds) pp. 429- 465, Raven Press, New York

Chacko, G. K., Mahlberg, F. H., and Johnson, W. J. (1988) J. Lipid Res. 29, 319-324

Davis, P. J., and Poznansky, M. J. (1987) Proc. Natl. Acad. Sci. U. S. A. 84.118-121

Dawidowicz, E. A. (1987) Annu. Reu. Biochem. 56, 43-61 Dawidowicz, E. A., and Backer, J. M. (1981) Biochim. Biophys. Acta

644, 373-375 DeGrella, R. F., and Simoni, R. D. (1982) J. Biol. C&m. 257,14256-

14262 Fainaru, M., Schaeffer, Z., and Deckelbaum, R. (1981) Prep. Biochem.

11,273-290 Friedman, G., Wernette-Hammond, M. E., Hui, D. Y., Mahley, R.

W., and Innerarity, T. L. (1987) J. Lipid Res. 28, 1482-1492 Gordon, D. J., Knoke, J., Probstfield, J. L., Superko, R., and Tyroler,

A. (1986) Circulation 74, 1217-1225 Gottlieb, B. A., and Marsh, J. B. (1987) Atherosclerosis 67, 251-259 Hatch, F. T., and Lees, R. S. (1968) Adu. Lipid Res. 6, l-68 Johnson, W. J., Bamberger, M. J., Latta, R. A., Rapp, P. E., Phillips,

M. C., and Rothblat, G. H. (1986) J. Biol. Chem. 261, 5766-5776 Johnson, W. J., Mablderg, F. H., Chacko, G. K., Phillips, M. C., and

Rothblat. G. H. (1988) J. Biol. Chem. 263.14099-14106 Kaplan, M: R., and Simbni, R. D. (1985a) J. ‘Cell Biol. 101, 446-453 Kaplan, M. R., and Simoni, R. D. (1985b) J. Cell Biol. 101, 441-445 Karlin. J. B.. Johnson. W. J.. Benedict. C. R.. Chacko. G. K.. PhilliDs.

M. c., and Rothblat, G. H: (1987) J. Biol. dhem. 282,12557-12564 Kobayashi, T., and Pagano, R. E. (1989) J. Biol. Chem. 264, 5966-

5973 Krieger, M., Brown, M. S., Faust, J. R., and Goldstein, J. L. (1978)

J. Biol. Chem. 253,4093-4101 Lambeth, J. D., Xu, X. X., and Glover, M. (1987) J. Biol. Chem. 262,

9181-9188 Lange, Y., and Matthies, H. J. G. (1984) J. Biol. Chem. 259, 14624-

14630 Lange, Y., and Steck, T. L. (1985) J. Biol. Chem. 260, 15592-15597 Lange, Y., Swaisgood, M. H., Ramos, B. V., and Steck, T. L. (1989)

J. Biol. Chem. 264,3786-3793 Lentz, B. R., Barenholz, Y., and Thompson, T. E. (1984) Chem. Phys.

Lipids 15,216-221 Lipsky, N. G., and Pagano, R. E. (1985) J. Cell Biol. 100, 27-34 Liscum, L., and Faust, J. R. (1987) J. Biol. Chem. 262, 17002-17008 Liscum, L., Ruggiero, R. M., and Faust, J. R. (1989) J. Cell Biol. 108,

1625-1636 McCloskey, H. M., Glick, J. M., Ross, A. C., and Rothblat, G. H.

(1988) Biochim. Biophys. Acta 963,456-467 Pentcbev, P. G., Comly, M. E., Kruth, H. S., Tokoro, T., Butler, J.,

Sokol, J., Filling-Katz, M., Quirk, J. M., Marshall, D. C., Patel, S., Vanier, M. T., and Brady, R. 0. (1987) FASEB J. 1, 40-45

Pfeffer, S. R., and Rothman, J. E. (1987) Annu. Rev. Biochem. 56, 829-852

Phillips, M. C. (1972) in Progress in Surface and Membrane Science, Vol. 5 (Danielli, J. F., Rosenberg, M. D., and Cadenhead, D. A. eds) pp. 139-222, Academic Press, New York

Phillips, M. C., Johnson, W. J., and Rothblat, G. H. (1987) Biochim. Biophys. Acta 906, 223-276

Quarfordt, S. H., Jain, R. S., Jakoi, L., Robinson, S., and Shelburne, F. (1978) Biochem. Biophys. Res. Commun. 83, 786-793

Renaud, G., Hamilton, R. L., and Havel, R. J. (1989) Hepatology 9, 380-392

Ross, A. C., Go, K. J., Heider, J. G., and Rothblat, G. H. (1984) J. Biol. Chem. 259,815-819

Rothblat, G. H., and Phillips, M. C. (1982) J. Biol. Chem. 257,4775- 4782

Rothblat, G. H., Bamberger, M., and Phillips, M. C. (1986) Methods Enzymol. 129,628-644

Sleight, R. G., and Pagano, R. E. (1983) J. Biol. Chem. 258, 9050- 9058

Slotte, J. P., Oram, J. F., and Bierman, E. L. (1987) J. Biol. Chem. 262,12904-12907

Spady, D. K., and Dietschy, J. M. (1989) J. Lipid Res. 30,559-569 Suckling, K. E., and Stange, E. F. (1985) J. Lipid Res. 26, 647-671 Tabas, I., Weiland, D. A., and Tall, A. R. (1985) Proc. Natl. Acad.

Sci. U. S. A. 82,416-420

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from

W J Johnson, G K Chacko, M C Phillips and G H RothblatThe efflux of lysosomal cholesterol from cells.

1990, 265:5546-5553.J. Biol. Chem. 

  http://www.jbc.org/content/265/10/5546Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/10/5546.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on Decem

ber 26, 2019http://w

ww

.jbc.org/D

ownloaded from