Human Plasma Cholesteryl Ester Transfer Protein Enhances the ...

THEJOURNALOF BlOLOClCkL CHEMtSTRY 0 1987 by The American Society of Biological Chemists. Inc.

Voi. 262, No. 8,ksue of March 15,pp. 392-3*7,1987 Prmted m U.S.A.

Human Plasma Cholesteryl Ester Transfer Protein Enhances the Transfer of Cholesteryl Ester from HighDensity Lipoproteins into Cultured HepG2 Cells* (Received for publication, July 28, 1986)

Esther Granot,Ira Tabas, andAlan R. Tall From the Division of ~ ~ t r ~ ~ t e rCollege o ~ gofy P, ~ y s ~ cand i a Surgeons, ~ C o ~ u U m n~~ v~e r s i New ~ , York, New York IO032

The role of human plasma cholesterylester transfer protein (CETP) in the cellular uptake of high density lipoprotein (HDL) cholesteryl ester (CE) was studied in a liver tumor cell line (HepGZ). When HepGZ cells were incubated with [3H]cholesteryl ester-labeled HDLBin the presence of increasing concentrations of CETP there was a progressive increase in cell-associated radioactivityto levels that were2.8 times control. The CETP-dependent uptake of HDL-CE was found to be saturated by increasingconcentrations of both CETP and HDL. The CETP-dependent uptake of CE radioactivity increased continuously during an 18-h incubation. In contrast to the effect on cholesteryl ester, CETP failed to enhance HDL protein cellassociation or degradation. Enhanced uptakeof HDL cholesteryl ester wasshown for thed > 1.21 g/ml fraction of human plasma, partially purified CETP, and CETP purified to homogeneity, but not for the d 1.21 g/ml fraction of rat plasma which lacks cholesteryl ester transfer activity. HDL cholesteryl ester entering the cell under theinfluence of CETP was largely degraded to free cholesterol by a process inhibitable by chloroquine. CETP enhanced uptake of HDL r3H]CE in cultured smooth muscle cells and to a lesser extent in fibroblasts but did not significantly influence uptake in endothelial cells or 5774 macrophages, These experiments show that, in addition to its known role in enhancing the exchange ofCE between lipoproteins, plasma CETP can facilitate in thevitro selective transfer of CE from HDL into certain cells.

In several species, including humans, plasma cholesteryl esters are synthesized within high density lipoproteins (HDL)’ as a result of the activity of the enzyme lecithincholesterol acyltransferase (1). HDL-cholesteryl esters m a y be transferrred to less dense triglyceride-rich lipoproteins (chylomicrons and very low density lipoproteins) by a cholesteryl ester transfer protein ( C E T P ) (2,3). Since the remnants of triglyceride-rich lipoproteins are taken up by specific receptors in the liver @), the CETP potentiallyprovides a mechanism for the transfer of cholesteryl ester from plasma to liver. In contrast to this indirect route of catabolism, HDLcholesteryl esters are also thought to be taken up directly by ~~~

~

* This work was supported by National Institutes of Health Grant HL22682. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 LJ.S.C. Section 1734 solelyto indicate this fact. * The abbreviations used are: HDL, high density Iipoproteins; LDL, low density lipoproteins; CE, cholesteryl ester; CETP, cholesteryl ester transfer protein; DMEM, Dulbecco’s modified Eagle’smedium; BSA. bovine serum albumin.

certain tissues in a process which is selective for the cholesteryl esters. Thus, a dispropo~~onate uptake of HDL cholesteryl ester compared to apoprotein A-I has been shown i n rat

ovary, adrenal, and liver, and also in hepatocyte cultures (5, 6). In the present investigation we have examined whether cholesteryl ester transfer protein can play a role in the direct transfer of

HDL-CE into model liver cells, thereby providing

an additional potential pathway by which CETP may influence HDL catabolism. A human liver tumor cell line (HepG2) served as a model for studying the effect of cholesteryl ester transfer protein on cellular HDL cholesteryl ester uptake. EXPERIMENTAL PROCEDURES

~a€eria~ ~ e l ~ - H u m a nhepatoma cell line (HepG2) cells were kindly provided by Drs. Knowles, Howe, and Aden of the Wistar Institute. The 5774 macrophage-like cell line was obtained from Jay Unkeless (Rockefeller University). Porcine aortic endothelial cells (sixth passage) and rabbit aortic smooth muscle cells (third passage) were kindly provided by Dr. Ken Pomerantz, Columbia University. Fibroblasts were from neonatal human foreskin (sixth-eighth passage). All cells were frozen in liquid nitrogen and thawed rapidly prior to use. For each experiment the cells were plated in 16 X 35-mm plastic Petri dishes in Dulbecco’s modified Eagle’s medium ( D ~ E M con) taining 10% heat-inactivated fetal bovine serum, 100 unitslml penicillin, 100 pg/ml streptomycin, 292 pg/ml L-glutamine. Plates were incubated at 37 “C in an atmosphere containing 8% Cop, 92% air. HepG2 cells were routinely split 1:6 every 4-5 days. At the time experiments were performed the cells appeared confluent. L i ~ p r o ~ i nand s Cholesteryl EsterTransfer Prote~n-~uman HDL, (1.125-1.21 g/mlf containing radiolabeled cholesteryl esters was prepared as described previously (7). The HDL contained 99% of radioactivity in cholesteryl esters and 1%in cholesterol. Compositional analysis showed that thelabeled HDL contained 55% protein, 20% phospholipid, 18% cholesteryl ester, 4% cholesterol, and 3% triglyceride and was therefore similar in composition to HDLB.Also, the radiolabeled preparation co-eluted with HDL, on a 100-cm column of6% agarose, indicating a similar particle size. In some instances cold HDL, wasadded to theradiolabeled HDL to achieve the desired specific activity, 2660 cpm/pg ofCE. Human HDLB was radiolabeled with lzSI by the iodine monochloride method of MacFarlane (8).Approximately 4% of the radiolabel was associated with lipids. Human plasma cholesteryl ester transfer proteinwas routinely purified about 500-fold from pooled blood-bank plasma through the carboxymethyl cellulose step (9). In selected experiments this CETP preparation was further purified to homogeneity by incubation with a synthetic lipid emulsion containing egg phosphatidylcholine, triolein, and oleic acid. The mixture was subjected to chromato~aphy on a Sepharose 4B column, and the active CETP was then obtained from the emulsion following delipidation with ethanol/ether. Based on activity, the homogeneous CETP was purified 55,000-fold relative to theplasma d > 1.21 g/ml fraction.‘ Human and ratlipoprotein-poor plasma fractions were isolated by preparative ultracentrifugation of human and ratplasma at d 1.21 g/ Hesler, C., Swenson, T. L., and Tall, A.R. (1987) J. Biol. Chem. 262,2275-2282.

3432

Lipid Facilitated

intoTransfer

Cells

3483

ml for 72 h in a Beckman Ti-50.2 rotor at 45,000 rpm at 10 "C. [,HI Cholesteryl ether-labeled HDL was prepared by incubating a sonicated emulsion of 20%egg phosphatidylcholine, 65% cholesteryl linoleate, 15% triolein, and [3H]cholesterylhexadecyl ether (100 pCi) in human plasma (37 "C, 18 h) followed by isolation of the d 1.1251.21 g/ml fraction by sequential preparative ultracentrifugation. Dulbecco's modified Eagle's medium, penicillin (10,000 units/ml), streptomycin (10,000 pg/ml), glutamine (200 mM), and trypsin were from Gibco. Fetal bovine serum was obtained from MA Bioproducts. Bovine serum albumin (essentially fatty acid-free) and chloroquine were obtained from Sigma. NaIz5I (in NaOH) (508 mCi/ml), was purchased from ICN Biochemicals Inc. 3H-Labeled cholesterol (23.7 Ci/mmol) and 3H-labeled cholesteryl hexadecyl ether (46.8 Ci/mmol) were from New England Nuclear.

RESULTS

T o assess the role of CETP in the uptakeof HDL cholesteryl ester by HepG2 cells, cells were incubated in medium containing [3H]CE-labeled HDL in presence the of increasing concentrations of partially purified CETP (Fig. 1A). With increasing CETP mass, there was a progressive increase in the cellular uptake of HDL-CE up to2.8 times that of HDLCE uptake without CETP (Fig. l A , closed circles). To see whether CETP caused a similar increase in cellular uptake of HDL protein,we incubated lZ5I-HDLwith HepG2 cells inthe presence of increasing concentrations of CETP. Cell uptake of HDL protein was calculated as the amount of '"I-HDL that remained associated with the cells plus the amount of Methods lZ5I-HDLdegraded during the incubationperiod as calculated Cell Association Assays-At the time of the experiments the cells from the amount of trichloroacetic acid-soluble radioactivity were incubated at 37 "C with 1 ml of DMEM containing 0.1% BSA in themedium, i.e. cell association and degradation.As shown (essentially fatty acid-free) and lz5I-HDL3or [,H]CE-labeled HDL3. in Fig. lA (open circles),the presence of CETP did not result Incubations were performed with or without added cholesteryl ester in a significant enhancementof cellular HDL protein uptake. transferprotein. Cholesteryl estertransfer protein was dialyzed In six different experiments similarresponses were obtained: against DMEM before being added to theincubation medium. At the mean cellular HDL-CE uptake was 1.1 f 0.16% of the total end of the incubation the dishes were placed on ice and the medium wasremoved. The monolayers were then washed as described by added radioactivity without CETP and increasedprogressively to 2.9 f 0.36% in the presence of CETP (210 pg/ml). Tabas and Tall (11):three rapid (< 1 min) washes in ice-cold 0.05 M Tris-HCL, pH 7.4, containing 0.15 M NaCl and 0.2% BSA (TBS-A), Uptake of HDL protein was 1.1 f 0.09% without CETP and three 10-min washes in TBS-A, and finally three rapid washes in 1.1% k 0.05% in the presenceof CETP. Thus, in theabsence TBS. The monolayer was then dissolved in 1 ml of 0.1% sodium of CETP, uptakeof HDL protein and HDL-CE were in similar

dodecyl sulfate, an aliquot removed for protein assay (Lowry method (12)), and the radioactivity associated with the cells determined. The average protein contents per dish were: HepG2 cells, 0.65 mg; 5774 macrophages, 0.8 mg; endothelial cells, 0.22 mg; smooth muscle cells, 0.27mg; and fibroblasts, 0.26mg. To ensure that cholesteryl ester transfer activity did not diminish during the incubation period, CE transfer activity was measured in the medium at the end of the 18-h incubations. The activity in medium of incubations performed in the presence of CETP was compared to thatin medium from incubations performed without CETP. For all cell systems studied we demonstrated atleast constant or slightly increased cholesteryl ester transfer activity in the medium throughout the incubation period. In experiments in which the amount of cell membrane-bound HDL-CE was determined, the cells were treated, at theend of the 18-h incubation, with trypsin at varying concentrations between 0.05 and 1.0% for 5 min. At these trypsin concentrations cell membranes remained intact, as demonstrated by a negative trypan blue stain at termination of the 5-min incubations. To determine the amount of Iz5I-labeledHDL bound to the cell surface following the nine sequential washes (see above) monolayers were treated with 0.05% trypsin and incubated for 5 min a t 37 "C. After 5 min 1 ml of DMEM with 10% fetal bovine serum was added to inhibit further proteolysis. The cell suspension was then centrifuged (1000 X g ) for 10 min at 4 "C. The radioactivity in the supernatant was the trypsin-releasable value. The cell pellet was washed with TBS-A and respun at 1000 X g for 10 min. The supernatantwas discarded and the pellet counted (trypsin-resistant value). To determine whether the Iz5I-HDLtrypsin-resistant radioactivity represents cellular uptake of Iz5I-HDLby the cells, displacement experiments were carried out in which excess unlabeled HDL (1 mg of protein/ ml) was added to the cells at the end of the 18-h incubation. Cells were washed rapidly three times with TBS and incubated for 3 h at 37 "C with excess unlabeled HDL. Degradation of '"I-HDL was determined in the medium, at the end of the 18-h incubations, after precipitation of protein with trichloroacetic acid (13). Measurement of Radiolabeled Free Cholesterol and Cholesteryl Esters in Cells-Following incubations with [3H]CE-labeled HDL3 the medium was removed and the cells were washed as described above. After the last wash, cells were scraped with a rubber policeman, suspended in TBS, and spun at 1000 rpm for 10 min, and then the supernant was discarded. Cellular lipids were extracted by the Folch method (14), and the extracts were applied to thin layer chromatography plates and analyzed in a solvent system of hexane/ether/acetic acid (70:301). Free [3H]cholesteroland [3H]cholesteryl ester, identified using reference standards, were scraped off the plates, and their radioactivity was determined in a liquid scintillation counter.

I

A

140

70

210

350

PROTEIN MASS ( p g )

0.1 0.2

0.6

1.0

CETP MASS ( p g )

FIG. 1. A, effect of CETP on uptake of HDL, cholesteryl ester and protein by HepG2 cells. Cells were incubated for 18 h at 37 "C with 1 mlof DMEM, 0.1%BSA containing [3H]cholesteryl ester-labeled HDL, or 1251-HDL,plus the amount of partially purified CETP indicated on the X axis. The [,H]CE-labeled HDL specific activity was 2660 cpm/pg of CE and the "'1-HDL specific activity was 100 cpm/ng of protein. The HDL concentration was 30 pg of CE/ml in both experiments. Cholesteryl ester (closed circles)and protein (open circles) celluptake were expressed as percent of control (i.e. Y - H D L or radioactive cholesteryl ester cell uptake in the absence of CETP). Mean control Iz5I-HDLcell association and degradation value was 1.6 pg of protein/ml of cell protein and control cholesteryl ester cell uptake was 0.55 pg of CE/mg of cell protein. B , effect of homogenous CETP on uptake of HDL, cholesteryl ester by HepG2 cells. Cells were incubated for 18 h at 37 "C in 1 mlof DMEM, 0.1% BSA containing [3H]cholesterylester-labeled HDL, (specific activity 2660 cpm/pg of CE, 30 pg of CE/ml) in the presence of increasing concentrations of purified, homogenous cholesteryl ester transfer protein. 1.0 pg of this CETPpreparation is equivalent in activity to approximately 350pgof partially purified CETP. Radioactive cholesteryl ester cell uptake was expressed as percent of control (i.e. cellular cholesteryl ester uptake in the absence of CETP). Each point represents the mean of two parallel experiments. Results varied by not more than 5%.

3484

Facilitated Lipid intoTransfer

proportion to the original HDL. However, in the presence of CETP, there was a selective enhancement of cellular HDL cholesteryl ester uptake compared to HDL protein uptake. Further experiments were conducted to determine whether the stimulated uptake of HDL-CE was a specific property of CETP. CETPwas purified to homogeneity as described under "Methods." The effect of CETP on cellular uptake of HDLCE was also shown by the purified homogenous CETP with enhanced cellular HDL-CE uptake at increasing CETP concentrations (Fig. 1B). Based on specific activity (CE transferred/mg of protein in an HDL-LDL exchange assay) the homogenous CETP was purified approximately 350-fold compared to the partially purified CM52 preparation. The fact that a similar increase in specific activity was found for the cellular uptake of HDL-CE (cf. Fig. 1, A and B ) suggests that the effect of the partially purified fraction on cellular uptake of CE can be entirely accounted for by its content of CETP. This was confirmed in an experiment where immunoprecipitation of the partially purified fraction with CETP-specific IgG2 completely abolished its ability to stimulate cellular uptake of HDL-CE, whereas non-immune IgG did not. To prove further that CETP activity was responsible for the selective uptake of HDL-CE, we took advantage of the fact that human d > 1.21 g/ml lipoprotein-poor plasma shows CE transfer activity, whereas rat plasma is devoid of such activity. Therefore, we compared the effect of human and rat lipoprotein-poor plasma on the cellular uptake of HDL cholesteryl ester (Fig. 2). Cellular [3H]CE-labeled HDL uptake increased as afunction of increasing concentrations of human lipoprotein-poor plasma (closed circles). By contrast, the rat d > 1.21 g/ml plasma fraction did notenhanceHDL-CE uptake by HepG2 cells (open circles). To characterize further the effect of CETP on the uptake of HDL cholesteryl ester, we studied the time course of the cellular uptake of HDL-CE (Fig. 3). The effect of CETP on HDL-CE uptake was clearly apparent after 6 h of incubation and increased throughout the length of incubations. The effect of increasing HDL-CE mass on the CETP enhancement of cellular HDL-CE uptake was also examined. Incubations were performed with increasing concentrations of [3H]CE-labeled HDL but constant CETP concentrations (Fig. 4). With increasing HDL mass there was an increase in both the basal and CETP-stimulateduptake of HDL cholesteryl ester. However, the increment in CE uptake specifically due to CETP (i.e. the difference of the two curves in Fig. 4) reached a maximum value at a HDL concentration of45 gg/ml, indi-

1

PROTEIN MASS IN dzl-21 p/ml FRACTION Imp)

FIG. 2. Effect of human versus rat lipoprotein-poor plasma on uptake of HDLs cholesteryl ester byHepG2 cells. Cells were incubated for 18 h a t 37 "C with 1ml ofDMEM, 0.1% BSA containing [3H]cholesterylester-labeled HDL3 (specific activity 2660 cpm/pg of CE, 30 pgof CE/ml) in the presence of human d > 1.21 g/ml lipoprotein-poor plasma (closed circles) or rat d > 1.21 g/ml lipoprotein-poor plasma (open circles). Cellular HDL3 cholesteryl ester uptake was expressed as counts/min per mgof cell protein X Results represent the mean S.E. of four experiments.

Cells

TIME ( h )

FIG. 3. Time course of the uptake of HDLs cholesteryl ester by HepG2 cells, with and without CETP. Cells were incubated for 18 h a t 37 "C with 1 ml of DMEM, 0.1% BSA containing [3H] cholesteryl ester-labeled HDL (specific activity 2660 cpm/pg of CE, 30 pg of CE/ml). Cells were incubated in the absence of CETP (open circles) or in the presence of 210 pg/ml CETP (closed circles), for 2, 4,6,8, and 18 h. Cellular uptake of HDL-cholesteryl ester is expressed Results represent the as counts/min per mgof cell protein X mean of six experiments.

w

0

V

,

I

30

60

90

HDL-CE MASS ( p p / m l )

FIG. 4. Effect of HDLs cholesteryl ester mass on the uptake of HDLs cholesteryl ester by HepG2 cells. Cells were incubated for 18h a t 37 "C with 1ml of DMEM, 0.1%BSA containing increasing concentrations of [3H]cholesterylester-labeled HDL3 (specific activity 2660 cpm/pg of CE) alone (open circles) or in the presence of 210 pg/ml CETP (closed circles). Cellular uptake ofHDL3 cholesteryl Each ester was expressed as counts/min per mgof cell protein X point representsthe mean of two parallel experiments. Results varied by not more than 5%.

cating that, under these conditions, the CETP-specific uptake was saturated at relatively low concentrations of HDL. Further experiments were conducted to see whether facilitated CE transfercould be observed at physiological levels of HDL and transfer activity. In the earlier experiments (Fig. 2), itwas noted that theeffect of the human d > 1.21 fraction reached a maximum value at 7.2 mg of protein, at which point the ratio of HDL-CE/d > 1.21 protein approximates the physiological value. In a further experiment the concentrations of both HDL andd > 1.21 fraction were increased, using a fixed, physiological ratio of HDL/d > 1.21 fraction (Fig. 5). There was a continuous increase in both the basal and facilitated transfer of HDL-CE into the cells, with the -fold increase mediated by the d > 1.21 fraction approximately constant. Thus, theenhanced transfer of HDL-CE was observed with physiological concentrations of HDL and d > 1.21 protein in the medium. HepG2 cells have the capacity to synthesize and secrete apoB-containing lipoproteins (15). Therefore the observed effect of CETP on cellular HDL-CE uptake might actually occur through CETP-mediated transfer of CE from HDL to apoB-containing lipoproteins and subsequent cellular uptake of these lipoproteins. Sodium dodecyl sulfate-gel analysis of the apoproteins isolated from the d < 1.063 fraction of 4 ml of 24-h conditioned medium showed the presence of several micrograms of apoB-100; by contrast, the d 1.063-1.210 g/ml

Facilitated Lipid intoTransfer

Cells

3485

TABLEI Cholesteryl ester transfer activity (countslmin transferred) in HepG2 medium CE transfer activity was assayed in conditioned and unconditioned medium at the end of 18-h incubations performed in the presence of [3H]CE-labeledHDL with (+CETP) or without (-CETP) added CETP. CETP activity was assayed by adding exogenous LDL (0.1 mg of protein) to the medium, followed by incubation at 37 "C for 16 h, precipitation of apoB-containinglipoproteins with heparin/MnC12, then determination of radioactivity remaining in the supernatant. The value obtained by subtracting the measured radioactivity (counts/min) in the supernatant in each incubation condition studied from the radioactivity (counts/min) in the supernatant of unconditioned media without CETP constituted CE transfer activity. Unconditionedmedium without added CETP had no transfer activity. Results are the mean of two parallel experiments andvaried by less than 5%. Cells No cells 52 I 0 42 0 8 I56 HDL-CE MASS (gg/ml)

-CETP

+CETP

406

1906 1597

0

-fold increase dueto CETP. These results suggest that cellular cholesterol stores may influence HDL-CE uptake, but do not FIG. 5. HDL-CE uptake by HepG2 cells at a physiological indicate a specific role for the LDL receptor in the CETPratio of HDLa:d> 1.21 fraction. Cells were incubated for 20 h at mediated component of uptake. 37 "C with 1 ml of DMEM, 0.1% BSA containing the indicated Recentexperiments from thislaboratory3indicatethat concentrations of [3H]cholesteryl ester-labeledHDL, (specific activcholesteryl estertransferactivityaccumulatesin a timeity 5826 cpm/pg of CE) alone (closed circles) or together with human d > 1.21 g/ml lipoprotein-poor plasma (open circles). The protein dependent fashion in the medium of cultured HepG2 cells, concentration of d > 1.21 fraction used at each HDL3 concentration suggesting synthesis of CETP by these cells. To compare the was such that the mass ratio of d > 1.21 fraction protein:HDL,-CE amount of CEtransferactivityaddedto cellswith that was 173. Both the [3H]HDL3 and d > 1.21 fractions were dialyzed secreted by the cells duringan18-hincubation,theCE extensively against DMEM before they wereadded to the cells. transfer activity was measured in cell-conditioned or unconCellular HDL3-CE uptake was expressed as counts/min per mg of of exogenous cell protein X lo+. Results represent the mean of triplicate experi- ditioned medium with and without addition CETP (TableI). The resultsshow 20-30% higher activity for ments +- S.E. the cell-conditioned medium with or without added CETP, fraction did not contain detectable apoB (not illustrated). Weindicating accumulationof a small amountof endogenous CE transfer activity in themedium. It ispossible that thisendogshowed that, when HepG2 cells are incubated in medium enous CE transfer activity contributes to the uptake of HDL containing both HDL and CETP, the CETP can mediate the cholesteryl ester; however, it is a relatively small amount of transfer of cholesteryl ester from HDL to thed < 1.063 g/ml exogenous CETP. fraction, which contains the apoB-containing lipoproteins. In activity compared to that added with Further experimentswere performed to elucidate the mechincubations performed without CETP, 0.7 k 0.05% of the total radiolabeled CE in the incubation (480,000 cpm/well) anism of cellular HDL-CE uptake. As noted above, CETP transferred from HDL to thed < 1.063 g/ml fraction and, in had no effect on total cellular uptake of lz5J-HDL. Furtherthe presence of CETP, the transferred radiolabeled CE was more, CETP was also found to have no effect on lZ5I-HDL cell association (0.13% of total added radioactivity in the increased to 1.8 & 0.03% of total radioactivity. In order to presence of CETP as compared to 0.12% without CETP) or evaluate the possibility that the CETP-stimulated cellular on the fraction of the cell-associated radioactivity that was uptake of HDL-CE occurs via the uptake of these apoBcontaining lipoproteins, an aliquot of the d < 1.063 g/ml displaceable by cold HDL (40.8% in the presence of CETP and 40.3% in control experiments without CETP). Of the fraction (containing5,000 cpm) was isolated from the medium cell-associated protein radioactivity a major fraction (about after an 18 h incubation in the presence of CETP and re55%) was trypsin-releasable. By contrast, in another experiincubated for 18 h with new HepG2 cells. There was no ment it was found that only 6-9% of CETP-stimulated [3H] detectable cellular uptake of radioactivity (ie. less than 0.1% cholesterylesteruptake was trypsin-releasable,indicating of incubated counts/min). Thus, although CETP did enhance internalization of HDL-CE. transfer of CE to d < 1.063 g/ml lipoproteins in the medium, Sincethe major fraction of cell-associated HDL-CEis cellular uptake of CE present in d < 1.063 g/ml lipoproteins internalized by the cell, the intracellular fate of cholesteryl is unlikely to account for the CETP-stimulated uptake of ester was studied. Following cellular uptake cell lipids were HDL cholesteryl ester.T o evaluate further a possible role of extracted and separated by thin layer chromatography. In the the LDL receptor in the CETP-mediated uptake of HDL-CE, presence of CETP the amount of cell-associated radioactivity HepG2 cells were grown in medium containing LDL or lipo- of both free cholesterol and CE was augmented 2-3-fold. Since protein-deficient serum for24 h, themedia was removed, and more than 99% of the label in HDL was in cholesteryl ester thenthe cells were incubatedwithHDL k CETP.LDL and since CETP does not promote transfer of unesterified preincubation resulted in an 82%decrease in subsequent lZ5I- cholesterol, these results suggest hydrolysis of CE following LDL degradation by the cells. For cells preincubated in lipo- entry into thecell. This was confirmedby the addition of 25 protein-deficient serum the percent uptake of the HDL-CE FM chloroquine, which inhibitedthedegradation of HDL was 1.1% (HDL)and 2.2% (HDL + CETP),whereasthe cholesteryl ester t o free cholesterol (Fig. 6). These results corresponding values were 0.6 and 1.1% for cells preincubated indicate that HDL cholesteryl esters entering thecell under with LDL. Thus, the preincubation with LDL decreased the absolute values of HDL-CE uptake but did not influence the J. Simmond, T. Swenson, and A. R. Tall, unpublished data.

3486

Facilitated Lipid Transfer intoCells

t

associated [3H]cholesteryl ether was 1.40-fold higher in the presence of CETP as compared to control. In a simultaneous experiment thecell-associated [3H]cholesteryl ester increased 1.49-fold, at 18 h, in the presence of CETP. Thus, CETP caused stimulation of both cholesteryl ether and cholesteryl ester uptake. Therelatively small effect observed was due to the lower specific activity of thepartially purified CETP 500 preparation used in this experiment. In order to determine whether CETP enhancement of HDL cholesteryl ester uptake is specific for HepG2 cells we evalul o 0 l -CETP ated the effect of CETP on HDL-CE uptake by other cell CETP +CETP -CETP types (Fig. 7). CETP enhanced uptake of HDL-CE by smooth -CHLOROQUINE +CHLOROQUINE muscle cells; HDL-CE uptake increased to levels that were FIG. 6. Effect of chloroquine on the degradation of HDLB 2.6 times the control. CETPalso enhanced HDL-CE uptake cholesteryl ester. Cells were incubated for 18 h a t 37 “C in1 ml of DMEM, 0.1% BSA containing [3H]cholesteryl ester-labeled HDL, by fibroblasts but to a lesser extent than in smooth muscle total counts 240,000 cpm/well (specific activity 2660 cpm/rg of CE) cells. In endothelial cells and 5774 macrophages there was no cellular HDL-CE uptakewith increasing alone or plus 25 pM chloroquine. For both sets of conditions incuba- significant change in CETP concentrations, indicating that CETPdoes not facilitions were performed in the absence or presence of CETP (210 pl/ ml). At the end of the incubation period the cells were washed (as tate uptake of HDL-CE in thesecell types. I5O0

T

h

described under “Methods”) and lipids extracted by the Folch method (14). Free cholesterol and cholesteryl ester were separated by thin layer chromatography. Results represent the mean S.E. of six experiments. Openbars, free cholesterol; shadedbars, cholesteryl ester.

*

TABLE I1 The effect of chloroquine and acyl-CoA:cholesterol acyltransferase inhibitor on the degradation of HDL3-cholesteryl ester Cells were incubated for 18 h at 37 “C in 1 mlof DMEM, 0.1% BSA containing [3H]CE-labeled HDL, total radioactivity 240,000 cpm/well (specific activity 2,660 cpmlpg of CE) alone or plus 25 p~ chloroquine andan acyl-CoA:cholesterol acyltransferase inhibitor (58-035) (5 wg/ml). For both sets of conditions incubations were performed in the absence or presence of CETP (210 pg/ml). At the end of the incubation period cells were washed, lipids extracted, and free cholesterol and cholesteryl ester separated by thin layer chromatography as described under “Methods.” The results are themean of two parallel experiments and varied by not more than 5%. Cell-associated radioactivity Cholesteryl Free cholesterol ester cpm

-CETP +CETP [3H]CE-HDL+ chloroquine + acyl-CoA:cholesterolacyltransferase inhibitor

543 1238

-CETP 674 +CETP 1529775

436 702

DISCUSSION

The human plasma CETP has been well documented to facilitateboththe exchange andnettransfer of CEand triglyceride between plasma lipoproteins. Steinet al. (16) have recently shown that CETP remove can cholesteryl esters from lipoproteinsboundto cell surface locationsandcan also remove CE from intracellular locations but only following treatment that permeabilizes the cells. We present evidence for afunction of CETP that has not been previously described. Our observations indicate that CETP can promote the transfer of cholesteryl ester from HDL into intact HepGZ and smooth muscle cells. The radiolabeled cholesteryl esters are internalized andundergo lysosomal degradation. Based on the kinetics of lipid exchange between the lipoproteins, two models for CETP-dependent CE transfer have been suggested. A ping-pong model proposes that CETP acts as a carrier of CE between donor and acceptor lipoproteins (17). Alternatively, CETP may enhance theexchange of lipids during formation of a ternary collision complex consisting of donor and acceptor lipoprotein and CETP (18).Either model could potentially explain the effect ofCETP oncellular HDLCE uptake. However, the failure of CETP to promote cell association or degradation of HDLprotein suggests that

599 V w

304

the influence of CETP are susceptible tolysosomal degradation. To evaluate further the possibility that the accumulating CE reflected re-esterification of [3H]cholesterol by acylCoA:cholesterol acyltransferase,the combinedeffects of chloroquine and 58-035, a specific acyl-CoA:cholesterol acyltransferaseinhibitor, were studied(Table 11). Despitethe w u presence of an acyl-CoA:cholesterol acyltransferase inhibitor, cellular cholesterylesterradioactivity was markedlyin70 140 210 350 creased. Therefore, the accumulating cholesteryl ester was PROTEIN MASS I p 9 ) directly derived from HDL and had not been derived from reFIG. 7. Effect of CETP on cellular uptake of HDLB cholesesterification of [3H]cholesterolby acyl-CoAcholesterol acyl- teryl ester. Smooth muscle cells ( S M , A), endothelial cells (EC, O), transferase. 5774 macrophages (MO, O), and fibroblasts (FB, A) were incubated To verify further that CETP-mediated enhancement of for 18 h a t 37 “C with 1 ml of DMEM, 0.1% BSA containing [3H] HDL-CE uptakewas indeed an effect on neutral lipid transfer,cholesteryl ester-labeled (specific activity 2660 cpm/pg of CE, 30 pg we studied HDL cholesteryl ether uptake by HepG2 cells of CE/ml), in the presence of increasing concentrations of CETP. Cellular uptake of HDL,-cholesteryl ester was expressed as percent underthe influence of CETP (cholesteryl ethersarenot of control (i.e. cholesteryl ester uptake in the absence of CETP). susceptible to degradationby hydrolases, and are taken up by Mean control cellular uptake was: smooth muscle cells, 0.4 pg of CE/ cells asintact molecules). HepGZ cells were incubatedin mg of cell protein; endothelial cells, 1.1 pg of CE/mg of cell protein; DMEM, 0.1% BSA containing [3H]CE-labeled HDL aloneor 5744 macrophages, 0.59 pg of CE/mg of cell protein; fibroblasts, 0.31 in the presence of CETP. Following an 18-h incubation cell- pg of CE/mg of cell protein.

Facilitated Lipid Transfer into Cells CETP is not acting by enhancing binding or fusion of HDL with the cell surface. The saturationof CETP-dependent CE uptake by increasing donor (HDL) concentration is typical of the kinetics of carrier-mediated lipid transfer (17). If the transfer into thecell is carrier-mediated, then the findings of the presentstudy imply the existence of a cell surface binding site that canbe recognized by CETP. CETP binds readily to phospholipid surfaces in emulsion, lipoproteins, and vesicles, especially in the presence of an increased negative charge (7, 19, 20), suggesting that CETPmight also bind to thelipids of the plasma membrane. Another possibility is that CETPbinds to a cell surface receptor. However, a specific role of the LDL receptor inmediating the CETP-dependent uptake seems unlikely, since down-regulation of the LDL receptor did not alter the-fold stimulation of uptake due to CETP.An analogy to the present results is suggested by the previous descriptions of lipoprotein lipase-enhanced cellular uptake of CE from liposomes or lipoproteins (21). Both the lipase and CETP molecules may have bindingsites for neutral lipid which allow them toact as carriers of CE. CETP increased the uptake of HDL-CE but did not affect the cell association or degradation of HDL protein. Thus, CETP does not enhance binding or internalization of whole HDL particles. These findings are reminiscent of previous studies in which it had been shown that, in selected tissues, there is a disproportionate uptake of HDL-CE compared to HDL protein (5,6,22). HDL-CEuptake into rat adrenal cells was not changed during metabolic inhibition of sucrose pinocytotic processes, implying that thedisproportionate uptake of HDL-CEin excess of HDL apoA-I probably does not involve whole particle uptake through receptor-mediated endocytosis (23). Although these studies have mostly been performed in the rat, which lacks cholesteryl ester transferactivity, selective uptake of HDL-CE has also been shown in perfused rabbit liver ( 2 2 ) , which in other studies has been shown to accumu1at.echolesteryl ester transfer activity (24). The CETP promotes neutral lipid transfer between lipoproteins by facilitating acholesteryl ester-triglyceride heteroexchange process (25). The phenomenon of CETP-dependent cellular uptake of HDL-CE might also involve lipid exchange. Because the amount of CE taken up by HepG2 cells is in the range of 1.5-1.6 ,ug ofCE/mg of cell protein, it was not possible to document net mass changes. Currently, there is a paucity of data to show that, upon entry into cells, HDL-derived CE affects cellular metabolism. However, we have shown that internalized HDL-CE undergoes hydrolysis via a lysosomal pathway. The observed increase in cellular free cholesterol suggests that HDL-derived CE has the potential to influence cellular cholesterol metabolism. HDL-CE uptake does promote cellular prostanoidsynthesis,in part by transfer of cholesteryl arachidonate from HDL to cellular lipid pools containing arachidonate (10). The CETP has been found to increase markedly the HDL-induced prost,anoid release by smooth muscle cells partly as a result of increased incorporation of HDL-CE-derived arachidonate into prostanoids.* These experiments clearly show that fatty acid derived from CETP-induced CE entry can influence cellular metabolism and strongly imply that thefacilitated entry of HDL-CE into cells is not a simple lipid exchange process. An intriguing aspect of the CETP-dependent uptake of "

...____

K. Pomerantz, A. R. Tall, P. Cannon, and E. Granot, unpublished data.

3487

HDL-CE is the cellular specificity. We have observed CETP enhancement of HDL-CE uptake in HepG2 cells, a human tumor cell line, in rabbit smooth muscle cells, and toa lesser extent in human fibroblasts. No effect of CETP on HDL-CE uptake was noted in 5774 macrophages or porcine endothelial cells. Specificity is not related to HDL binding, as endothelial cells have been shown to bind IWI-HDL3to a higher degree than eithersmooth muscle cells or fibroblasts (11) and, similarly, we have also observed that, in the absence of CETP, HDL-CE cell association per mg of cell protein was highest for endothelial cells. CETP augmentation of HDL-CE uptake could be related to cell surface characteristics that enable the cell membrane to bind CETP or to the existence of intracellular lipid pools that provide lipid for CETP-mediated exchange processes. The physiological significance of the CETP-dependentuptake of HDL-CE is unknown. As theseexperiments were conducted in a model cell system, definite conclusions relating to in vivo conditions cannot be made. However, the CETP did enhance cellular HDL-CE uptake at a physiological ratio of HDL-CE to human d > 1.21 g/ml fraction (Fig. 2) and at physiological concentrations of HDL and d > 1.21 fraction (Fig. 5). Since the d > 1.21 fraction shows similar CE transfer activity to whole p l a ~ m athis , ~ result indicates that theeffect was observed at physiological levels of CE transfer activity and HDL. In those species that possess plasma cholesteryl ester transfer activity, the CETP-mediated selective uptake of HDL-CE might constitute a pathway for direct incorporation of HDL-CE, exclusive of other HDL components, into the liver or other tissues. REFERENCES 1. Glomset, J. A., and Norum, K.R. (1973) Adv. Lipid Res. 1,l-65 2. Nichols, A. V., and Smith, L. (1965) J. Lipid Res. 6,206-210 3. Pattnaik, N. M., Montes. A,. Hughes. L. B.. and Zilversmit. D. B. (1978) Biochim. Biophys. Acta 530,458-438 4. Mahley, R. W., Hui, D. Y., Innerarity, T. L., and Weisgraher, K. H. (1981) J. Clin. Invest. 6 8 , 1197-1206 5. Glass. C.. Pittman. R. C., Weinstein, D. B., and Steinberg, D. (1983) Proc. Natl. Acad. Sci. U . S. A. 80,Ei435-5439 6. Glass, C., Pittman, R. C., Given, M., and Steinberg, D. (1985) J. Biol. Chem. 260,744-750 7. Tall, A: R., Sammett, D., and Granot, E. (1986) J. Clin. Invest. 77,11631172 8. MacFarlane, A. S. (1958) Nature 1 8 2 , 53-54 9. Tall, A. R., Ahreu, E., and Shuman, J. (1983) J. Biol. Chem. 2 5 8 , 217491nn

10. Poziiantz, K. B., Fleisher, L. N., Tall, A. R., and Cannon P. J. (1985) J. Lipid Res. 26,1269-1276 11. Tabas, I., and Tall, A. R. (1984) J. BioZ. Chem. 259, 13897-13905 12. Lowry, 0. H.,Rosehrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Bid. Chem. 193,265-275 13. Bierman, E. L., Stein, O., and Stein, Y.(1974) Circ. Res. 3 5 , 136-150 14. Folch, J., Lees, M., and Sloane Stanley, G. H. (1957) J. BioL Chem. 2 2 6 , 497-509 15. Rash, J. M., Rothblat,, 6. H., and Sparks, C. E. (1981) Biochim. Biophys. Acta 666,294-298 16. Stein, 0.. Halperin, G., and Stein, Y. (1986) Arteriosclerosis 6, 70-78 17. Barter, P. J., and Jones, M. E. (1980) J. Lipid Res. 2 1 , 238-249 18. Ihm, J., Quinn, D. M., Busch, S. J., Chataing, B., and Harmony, J. A. K. (1982) J. Lipid Res. 2 3 , 1328-1341 19. Pattnaik, N. M., and Zilversmit, D. B. (1979) J. Bid. Chem. 2 5 4 , 27822786 20. Sammett, D., and Tall, A. R. (1985) J. Biol. Chem. 260,6687-6697 21. Chajek-Saul, T., Friedman, G., Halperin, G., Stein, 0.. and Stein, Y.(1981) Biochim. Biophys. Acta 666,147-155 22. Mackinnon, M., Savage, J., Wishart, R., and Barter, P. (1986) J. BioZ. Chem. 2 6 1 , 2548-2552 23. Pittman, R. C., Knecht, T. P., and Taylor, A. C. (1985) Circulation 72,111 377 24. Parscav, L. 13..and Fielding, P. E. (1984) J. Lipid Res. 2 5 , 721-728 25. Morton, R. E., and Zilversmit, D. B. (1983) J. Biol. Chem. 258, 1175111757 ~-

-

'A. R. Tall, E. Granot, C. Hesler, R. Brocia, I. Tabas, and K. J. Williams (1987)J. Clin. Inuest., in press.