uptake of hdl unesterified and esterified cholesterol by human endothelial cells. modulation by hdl...

Biochimica et Biophysics Acta 958 (1988) 81-92 Elsevier

BBA 52717

Uptake of HDL unesterified and esterified cholesterol by human endothelial

cells. Modulation by HDL phospholipolysis and cell cholesterol content

Xavier Collet, Bertrand Perret, Fraqois Chollet, Fraqoise Hullin, Hugues Chap and Louis Douste-Blazy

INSERM Unite 101, Biochimie des Lipides, HGpital Purpan, Toulouse (France)

(Received 31 July 1987)

Key words: Cholesterol uptake; HDL; Phospholipase; Hepatic triacylglycerol lipase; (Human endothelial cell)

Human HDL (1.070-1.210), doubly labelled with 3H/‘4 C-labelled unesterified cholesterol and 3H-labelled esterified cholesterol were incubated for 1-5 h with monolayer cultures of human endothelial cells. HDL were preincubated for 60-120 min the presence of albumin and with/without purified phospholipase A, (control HDL, phospholipase A, HDL) before dilution in the cell culture medium. Average phosphatidylcholine (PC) degradation was 62.10% f 2.57% (range 45-80s). A purified lipase/phospholipase A, from guinea pig pancreas was used in some experiments (range of PC hydrolysis: 16-70s). (1) 3H/ l4 C-labelled unesterified cholesterol and 3H-labelled esterified cholesterol appeared in cells during O-5 h incubations.

Trypsin treatment allowed a simple adsorption of HDL onto the cell surface to be avoided, and most of the 3H-labelled esterified cholesterol transferred to cells was hydrolysed. Cell uptake of radioactive cholesterol increased as a function of HDL concentration but no saturation was achieved at the highest lipoprotein concentration used (200 pg cholesterol/ml). Flux of 3H/‘4 C-labelled unesterified cholesterol was related

to the cell cholesterol content, suggesting that it might partly represent an exchange process. The cell cholesterol content was slightly increased after 5 h incubation with HDL ( + 16%). (2) Pretreatment of HDL

with purified phospholipase A, doubled on average the amount of cell recovered 3H-labelled esterified cholesterol, while the flux of 3H/‘4C-labelled unesterified cholesterol was enhanced by 15-25%. Both transfer and cell hydrolysis of 3H-labelled esterified cholesterol were increased. A stimulation was also observed using purified lipase/phospholipase A,, provided that a threshold phospholipid degradation was achieved (between 27 and 45%). (3) Endothelial cells were conditioned in different media so as to modulate their charge in cholesterol. The uptake of 3H-labelled esterified cholesterol was found to be significantly higher in cholesterol-enriched cells compared to the sterol-depleted state. Finally, movements of ‘H-labelled esterified cholesterol from HDL to endothelial cells were essentially unaffected by cell density or by the presence of partially purified cholesterol ester transfer protein. The possible roles of the transfer of HDL esterified cholesterol to endothelial cells and its modulation by phospholipases are discussed.

Introduction

Abbreviations: PC. phosphatidylcholine; DTNB, 5,5- dithiobis(2nitrobenzoic acid).

The endothelial cell monolayer is the primary site of interaction between the vessel wall and

Correspondence: B. Perret, INSERM Unite 101, Biochimie des blood cells or plasma lipoproteins, and it exerts a Lipides. HBpital Purpan, 31059 Toulouse CCdex, France. protective role against thrombosis and athero-

0005-2760/88/$03.50 0 1988 Elsevier Science Publishers B.V. (Biomedical Division)

82

sclerosis developments. Growing endothelial cells can take up plasma cholesterol via the well regulated LDL-receptor pathway, but these receptors are progressively suppressed as cells organize and reach confluence [l-3]. Contact-inhibited endothelial cells may acquire esterified cholesterol from other sources such as chylomicrons, by a mechanism relatively independent of apolipoprotein uptake and degradation [4,5]. Lipoprotein lipase, which is physiologically bound onto the vascular endothelial surface [6], was reported to help the delivery of cholesteryl esters from chylomicrons or artificial liposomes towards smooth muscle cells, skin fibroblasts, mesenchymal heart and endothelial cells [7,8]. Endothelial cells, either growing or confluent, can also interact with circulating HDL, but low rates of uptake and degradation of the HDL apolipoprotein have been reported, following binding of the particles [9-111. However, little is still known about the disposal of HDL cholesterol by endothelial cells. This prompted us to set up this investigation with two components: processing of HDL free and esterified cholesterol.

In animal models, the liver is the major cata- bolic site for HDL radiolabelled esterified cholesterol, yet adrenals and ovaries also take it up with a high specific radioactivity [12,13]. HDL choIestero1 may be a precursor for bile acid synthesis by hepatocytes, as demonstrated by in vivo evidence or cell culture experiments in several species including humans 114-161. The liver pos- sesses the hepatic triacylglycerol lipase, a heparin-releasable lipolytic enzyme, which was also detected in rat ovaries and adrenals [17]. Since phospholipid in the large HDL subfractions (HDL,) is a preferential substrate for hepatic t~acylglycerol lipase [18], the hypothesis emerged that in situ phospholipolysis would trigger the delivery of HDL cholesterol to steroid-making cells, which was later verified at least for unesterified cholesterol, with rat hepatoma and human granulosa cells [19,20]. However, in liver, the highest degradative activities towards lipoprotein components were detected in non-parenchymal cells 121-231. In addition, hepatic triacylglycerol lipase has been located at the endothelial surface of liver sinusoid capillaries [24]. Thus, liver endothelial cells are in close contact both with the plasma compartment, and at the site of HDL

phospholipolysis. The present study was under- taken to find whether phospholipid degradation would modulate the delivery of HDL free and esterified cholesterol to endothelial cells, taken here as a model of non-parenchymal cells.

Materials and Meth~s

Materials All materials and media for cell cultures were

obtained from Seromed (Munich, F.R.G.). Highly purified phospholipase A 2 from Crotalus ada~ante~ (52 I.U./ml) was kindly provided by Professor R.F.A. Zwaal [25] and was kept at - 20°C in a 50 mM Tris-HCl buffer (pH 7.4) containing 50% (v/v) glycerol and 5 mM CaCl,. Purified lipase/phospholipase A, from guinea pig pancreas was isolated in our laboratory as described by Fauvel et al. [26]. Bovine serum albumin fraction V powder essentially fatty acid free and 5,5’-dithiobis(2-nitrobenzoic acid) (DTNB) were from Sigma (St Louis, MO, U.S.A.). [7(n)-3H]Cholesterol and [4-‘4C]cholesterol were obtained from Amersham-France (Paris Les Ulis, France).

Cell culture Endothelial cells were obtained from human

umbilic~ veins according to the method of Slater [27]. The cells were resuspended in medium 199 containing 20% (v/v) fetal calf serum, 100 U/ml penicillin, 50 pg/ml streptomycin and 2 mM L- glutamine, and were seeded in 25 cm2 flasks (Falcon, Becton Dickinson, France). After 24 h plating, the medium was changed for M-199 containing 20% (v/v) AB human serum. When cells had reached confluence (2 . lo6 cells/25 cm2 flask), they were subcultured in a ratio of 1: 2 by brief exposure to 0.05% trypsin/0.02% EDTA (w/v). The experiments reported here were carried out with cells derived from the same umbilical cord in the second-third passage at confluence. Morpho- logical and immunological criteria [28] were used to verify homogeneity of the endothelial cell cultures. All incubations were carried out at 37 * C in humidified atmosphere and 5% CO,.

Lipoprotein isolation Human plasma was obtained in the blood bank

83

(Centre Regional de Transfusion Sanguine, Tou- louse, France) from fasting healthy women who had cholesterol and triacylglycerol levels in the normal range. After removal of any chylomicron by a 30 min ultracentrifugation at 120000 x g,,, sodium azide (O.Ol%, w/v) and EDTA (O.Ol%, w/v) were added. HDL were isolated by sequen- tial ultracentrifugations at 120000 X g,, and at 5” C between the limit densities of 1.070 and 1.210, for 20 h and 44 h, respectively. HDL were washed at their two-limit densities. Each prepara- tion was extensively dialysed against 0.135 M NaCl/O.OIS M Hepes (pH 7.4). The purity of the fractions was ascertained by lipoprotein electrophoresis on 2-3% polyacrylamide gels (Sebia, Issy-les-Moulineaux, France), apolipoprotein electrophoresis on polyacrylamide (11% w/v containing 0.1% SDS) and by the absence of measured amounts of apolipoprotein B. Lipoproteins were stored at 4OC in darkness and under nitro- gen.

LabeIIing of lipoprotein with L31i]- and (‘“Cl- cholesterol

To label with [ 3H]cholesterol/cholesteryl ester, an albumin-stabilized emulsion of [7(n)- 3H]cholesterol was formed in 0.1 M phosphate buffer (pH 7.4) containing 2% bovine serum albumin (w/v) and 1.25 mM EDTA, and was then added to plasma (4 ~Ci/ml). Incubation was performed at 37°C for 12-15 h in the presence of 0.01% (w/v) sodium azide. [3H]Cholesterol- labelled HDL were isolated as above. After isolation, excess 3H-labelled unesterified cholesterol was removed by exchange with washed human erythrocytes diluted in 0.135 M NaCl/0.015 M Hepes buffer (pH 7.4) (25% hematoc~t). Two to three 4 h incubations at 37” C were performed and HDL were washed at d 1.21 following this step. More recently, an alternative method was used to label HDL with [3H]cholesteryl esters. After labelling total plasma, LDL was isolated at d 1.019-1.063. Freshly prepared HDL were incubated with 3H-labelled LDL and the plasma fraction of d > 1.21 g/ml in the presence of 0.4 mM dithionitrobenzoic acid (DTNB) to inhibit lecithin-cholesterol acyltransferase activity. After labelling, the 3H-labelled HDL was re-isolated at d 1.070-1.210. The percentage of [3H]cholesterol

in cholesteryl esters ranged between 70 and 80% with both methods.

To label HDL with 4-‘4C-labelled unesterified cholesterol, two methods were used. First, [4- “C]cholesterol (1 pCi/ml) was incubated with total plasma for 4 h at 37 * C in the presence of 0.4 mM DTNB. Lipoprotein fractions were isolated as above. Afterwards, pure HDL (1.070-1.210) were labelled by incubation with acid-ethanol washed Whatman No. 1 paper (1 cm’), onto which [‘4C]cholesterol was spotted (1.5 pCi/mg HDL cholesterol). Under this condition, i4 C-labelled HDL were washed at their limit densities after the labdling procedure. Around 90% of [‘4C]cholesterol was reisolated in the HDL density interval and 87-100% was in unesterified cholesterol.

The chemical composition of each HDL pre- paration was determined. For unlabelled HDL the average weight composition was as follows: protein, 48.6% (+ 1.2%); phospholipid, 23.0% (+1.4%); cholesteryl ester, 21.7% (il.l%); free cholesterol, 2.5% ( i 0.4%); triacylglycerol, 4.2% (+ 1.2%), means (i SE.) n = 12. Also, at phospholipid degradation by phospholipase A 2 as high as 62.10% (* 2.57) (see Results), no loss of apolipoprotein A-I or cholesterol could be detected in the reisolated HDL. 14C as well as 3H-labelled HDL, labelled by the cholesterol ester transfer reaction in the presence of 3H-labelled LDL, dif- fered little from their unlabelled homologues. HDL labelled with [3H]cholesterol/cholesteryl esters in total plasma with an active lecithin-cholesterol acyltransferase exhibited decreases in free cholesterol ( - 6%), phospholipid ( - 11.5%), and a concomitant enrichment in esterified cholesterol f + 14.7%). All parameters were calculated per mg apolipoprotein A-I, which accounted for 80.3% 1: 6% of the total protein measured in HDL, with no difference following 3H/‘4C labelling. SDS-polyacrylamide gel electrophoresis of HDL protein revealed almost exclusively the presence of apolipoprotein A-I and A-II. In overloaded gels however, a faint contamination by albumin was occa- sionally detected, as well as traces of an inter- mediate band possibly corresponding to apolipoprotein E. Mixed “C/3H-labelled HDL were analysed by equilibrium density sedimentation as in Ref. 29. Both labels gave a single peak and the average hydrated density of our HDL prepara-

84

tions was 1.146 f 0.005 (n = 4) with a spectrum

spaning between d 1.115 and 1.19.

Treatment of HDL with phospholipase A,

HDL doubly labelled with 3H and i4C were allowed to equilibrate for 30 min at 37 ’ C. Then, they were separated in equal parts and incubated with bovine serum albumin 1% (w/v), CaCl, 3 mM, and with or without phospholipase A, from C. adamanteus (200 mIU/ml). The lipoprotein concentration was adjusted to l-l.2 mM total cholesterol. After 60 min incubation at 37” C, HDL were sterilized by filtration through a Milex GV 0.22 pm filter unit (Millipore, Molshein,

France) and transferred to O-4’ C. Phospholipase A, from guinea pig pancreas (5-10 IU/ml) was used for 120 min in the same medium as phos-

pholipase A, except that CaCl, and bovine serum albumin were omitted. These two components were

added at the end of the incubation.

Experimental conditions Cells were used at confluence except when

stated in Results. On the day of the experiment, the medium was changed for M-199 medium containing 20% human serum (v/v), and preincubation was allowed for 4 h.

Cells were washed twice with Ca2+/Mg2+-free phosphate buffer saline and the medium was re- placed by serum-free M-199 containing antibio-

tics. Doubly labelled HDL (25-200 pg cholesterol) and bovine serum albumin 1% (w/v) present in

Hepes/NaCl were added. The final CaCl, concentration was 2 mM. Cells were incubated under standard conditions for l-5 h. At the end of the

incubation, the medium was withdrawn and stored at - 20 ’ C till analysis. The cell layer was washed three times with 0.2% bovine serum albumin (w/v)

in phosphate-buffered saline, and three times with phosphate-buffered saline alone at room temperature. The first wash contained between 4 and 6% of the 3H and i4C radioactivities, and the last one less than 0.1% of both labels. The recoveries of radioactive material in all fractions (supernatant, cells, washes) was about 90%. Cells were scraped off with a rubber policeman in methanol and were immediately submitted to lipid extraction.

When cellular proteins were to be measured, cells were recovered in the presence of 0.05 N

NaOH. Duplicate aliquots were taken for protein determination and the remaining was treated for

lipid extraction as above. In a series of experiments, cells were detached with 0.05% trypsin/0.02% EDTA (w/v) in Ca2+/Mg2+-free phosphate-buffered saline. 1 ml of this solution

was added to each flask and incubated at room temperature for 5 min under mild shaking. The cells were transferred to plastic tubes and following addition of 20% (v/v) fetal calf serum in phosphate-buffered saline to inhibit further pro- teolysis, the cells were separated by centrifugation at 1500 X g for 15 min. The supernatant was collected and is referred to as ‘trypsin-releasable’ material. The cell pellet was washed twice. Protein determination and lipid extraction were per-

formed as above. Aliquots of post-incubation supernatants were

taken for radioactivity measurements and 0.5-l

ml fractions were withdrawn for lipid analyses. Cell washes and trypsin-releasable media were

also assayed for 3H/‘4C radioactivities. Labelled cell lipids were counted after solvent evaporation.

Analytical method

Lipids were extracted according to Bligh and Dyer [30] after acidification with 0.012 ml formic acid/ml aqueous phase. When 0.05 N NaOH was used to remove cells no other addition was made. Control determination showed parallel recoveries of esterified cholesterol using either method and the extraction efficiency for radiolabelled cholesterol was close to 100%. Protein was determined by the procedure of Lowry et al. [31]. HDL total and unesterified cholesterol were measured with the cholesterol esterase/cholesterol oxidase technique [32] using commercial kits (Boehringer, Mannheim, F.R.G.) and aqueous

cholesterol standards (Sigma). Cell cholesterol was measured in the lipid extract by an adaptation [33] of the latter method. Cell or HDL phospholipid was estimated as the lipid phosphorus, according to Bottcher et al. [34]. Triacylglycerols were measured as described in Ref. 35. Phospholipid classes were separated by monodimensional thin-layer chromatography (TLC) on silica gel GeO plates (Merck, Darmstadt, F.R.G.), using the solvent system of Skipski et al. [36]. Individual spots were scraped off and analysed as above. 3H/‘4C-

85

labelled unesterified and esterified cholesterol from the supernatants and the cells were separated by TLC using petroleum ether/ diethyl oxide/ acetic acid (165 : 35 : 2 by vol.) as a solvent. The spots were scraped off and assayed for radioactivity. Radioactive counts were measured in a Packard Tri-Carb 4530 liquid scintillation apparatus with automatic quenching correction (Packard Instru- ment International, Zurich, Switzerland).

Apolipoproteins A-I and B were determined by immunoelectrodiffusion using a commercial kit (Sebia, Issy les Moulineaux, France) [37].

Cff~cuIfftions and statistics Transfer of unesterified cholesterol was calcu-

lated from the 14C radioactivity in cells and the 14C specific radioactivity of the free sterol in HDL. As well, the cell non-hydrolysed 3H-labelled esterifed cholesterol was determined from the specific radioactivity of HDL 3H-labelled cholesteryl ester. Intracellul~ hydrolysis of the transferred HDL esterified cholesterol was estimated according to the following formula:

Cell 13H]UC- (Cell [r4C]UC x 1s.R)

[3H]- and [‘4C]UC represent the respective radioactivities in cellular unesterified cholesterol after analysis by thin-layer chromatography, and 1s.R

4

represents the 3H/‘4C isotopic ratio of unesterified cholesterol in HDL. The latter was similar in pre- or post incubation HDL.

The results are given as the mean + SE. Statis- tical comparisons were performed using paired Student’s t-test.

Red ts

1. Kinetics of unesterified cholesterol exchange he- tween HDL and endothelial cells

Radioactive unesterified cholesterol gradually appeared in endothelial cells during O-5 h incubations with doubIy-labelled HDL (Fig. 1A). The amount of sterol taken up by cells was calculated from the i4C specific radioactivity in HDL. The process was dependent on the amount of HDL added and tended to level off at high lipoprotein concentrations (Fig. 1B).

The cell phospholipid content, taken as an indi- cator of the cell density, varied little between different flasks in the same culture: the variation factor was 15.2 f 2.6% on average. However, vari- ations were noted between different culture experiments (average phospholipid content 180.2 nmol t_ 18.11 nmol/flask, range 80-250 nmol), so that comparative data are normalized per pmol cell phospholipid.

The analysis of different expe~ments per-

t

” ;

3 5 25 50 100 200

TIME (hours) HDL_COtICE NTRAT ION

()JQ cholesterol /ml )

Fig. 1. Kinetics of exchange of HDL “C-labelled unesterified cholesterol (UC) by endothelial cells. A. Time kinetics over a O-5 h period. One experiment representative of four. Cells were incubated at 37’C with 4 ml of M-199 containing 1% bovine serum

albumin and [3H]/[‘4C]cholesterol-labe~led HDL, treated (0) or not (0) by purified phospholipase A,. The HDL concentration was

100 pg cholesterol/ml. Nanomoles of transferred unesterified cholesterol (UC) were calculated from the 14C specific radioactivity in HDL. B. Concentration kinetics. One experiment representative of two. Cells were incubated for 3 h in the same conditions as in A

with increasing concentrations (O-250 pg total cholesterol/ml) of [ 3H]/[‘4C]cholesterol-labelled HDL, treated (0) or not (0) by

purified phospholipase A,.

86

.

.

//

*=0.862

P<O.Ol - ----__ __

MOLE RATIO

Fig. 2. Relationship between the exchange of t4C-labelled unesterified cholesterol (UC) and the cell cholesterol content. Endothelial cells were incubated for 3 h with [~H]/[*4C]cholesteroi-labelied control HDL as in Fig. 1. HDL concentration was 83.7 + 4.4 lg chol~terol/~. All points are from individual experiments and are normahzed per pmof cell phospholipid. Cell cholesterol and phosphohpid were de-

termined as described in Materials and Methods.

Transfer of ‘H radioactivity to cells may represent fluxes of both HDL unesterified and esterified cholesterol. However, the use of HDL doubly labelled on free cholesterol enables the fate of the cholesteryl ester taken up by cells to be followed. An example of such a calculation is shown in Table I. After interaction with HDL, very low amounts of 3H esterified cholesterol were detected in the cell extracts. However, the isotopic 3H/‘4C ratio for unesterified cholesterol was higher in cells than in pre- or post-incubation HDL, indi- cating intracellular hydrolysis of the transferred 3H-Iabelled esterified cholesterol. The latter was calculated as described in Naterials and Methods. Total uptake of esterified cholesterol thus repre-

formed at 3 h incubation showed a positive correlation (r = 0.862, P < 0.01) between the flux of HDL-labelled unesterified cholesterol and the cellular cholesterol content expressed as the sterol/ phospholipid mole ratio (Fig. 2). This latter value ranged between 0.28 and 0.48 in our cultured cells.

Quantitatively, the measured fluxes of HDL unesterified cholesterol toward endothelial cells were weak. The contribution of cell-unesterified cholesterol to the total present in incubations ranged from 15 to 65% (mean: 37.2 f 4.8%), while the corresponding proportions of “C unesterified cholesterol recovered in cells after 180 min were only l.l-4.5% (mean 2.35 i 0.33%). Hydrolysis of HDL glycerophospholipid by phospholipase A 2 had little effect on the movements of HDL 14C- labelled free cholesterol (Fig. 1).

TABLE I

CALCULATIONS FOR THE CELL UPTAKE OF 3H/“4C-LABELLED UNESTERIFIED CHOLESTEROL (UC) AND OF ‘H-LABELLED ESTERIFIED CHOLESTEROL (EC)

Endothelial cells were incubated with [ 3H]/[‘4C]UC- and [ 3H]EC-Iabelled HDL. Radioactive specific activities were 330 dpm/nmol for [t4C]UC and 790 dpm/nmol for 13H]EC. sH/14C ratio in HDL- UC was 2.12 (i&06). One example of calculation. Transfer of UC was calculated from the cell 14C radioactivity and the [r4C]UC specific radioactivity in HDL. The cell non-hydrolysed [‘HJEC was determined from the [ 3 H]EC specific radioactivity in HDL. Hydrolysis of [ 3H]EC was estimated as described in Materials and Methods. UC, unesterified cholesterol; EC, esterified cholesterol.

Conditions Labelled UC

i3W’114Cl (dpm) (dpm)

ratio cell transfer (nmol)

[ sH]UC derived from [ 3HjEC hydrolysis

@pm) (nmol)

Labelled EC

‘H cell non hydrolysed

(dpm) (nmol)

HDL (0 time) HDL (5 b) Cell (5 h)

174 140 : 81246 (2.14) 766215

158862 : 77652 (2.05) 714430

6240: 1740 (3.58) 5.35 2552 3.20 1750 2.20

sented the sum of the cell-recovered and of the cell-hydrolysed 3H-labelled esterified cholesterol.

The transfer of 3H-labelled esterified cholesterol tended to a plateau after 3 h of incubation (Fig. 3A). The uptake was dependent on the lipoprotein concentration in the medium, but no clear saturation was observed up to 200 pg HDL cholesterol/ml (equivalent to 650 pg/ml HDL protein) (Fig. 3B).

Endothelial cells extracted 3H esterified cholesterol from phospholipolysed HDL more ef- ficiently than from control particles, at all incubation times and over all the concentration range tested (Fig. 3).

In some experiments, to assess the cell internalization of the transferred 3H-labelled esterified cholesterol, cells were taken off by a mild treatment with trypsin following HDL incubations. Some 20.8 + 1.8% of the 3H and 19.1 rt 2.1% of the i4C radioactivities (means k SE., n = 4) were so released by trypsin with no significant difference following incubation with either control or phospholipase-treated HDL. The 3H/*4C ratio in this trypsin supernatant was slightly above (X 1.17 + 0.05) that measured in cells but was clearly far below the same ratio in HDL (X 2.85 -t_ 0.40 that of cells, on average). Thus, the radiolabelled cholesterol released by trypsin would essentially originate from lysed cells or from membrane frag- ments rather than from HDL particles.

87

TABLE II

TRANSFERS OF HDL 3H/‘4C-LABELLED UNESTERI- FIED CHOLESTEROL (UC) AND 3H-LABELLED ESTERIFIED CHOLESTEROL (EC) TO ENDOTHELIAL CELLS

Cultured endothelial cells were incubated for 3 h with 3H/14C-fabelled HDL. After washing and scraping off the cells, the distribution of lipid radioactivity was determined. Results are from nine experiments and are normalized by pmol/cell phospholipid. The cell hydrolysis of [ 3H]EC was calculated as in Materials and Methods and is expressed as percent of the total [3H]EC taken up by cell (hydrolysed and non-hydrolysed). Statistical comparison between phosphoh- pase A, HDL and control HDL: * P < 0.05, * * P < 0.01; as., not significant; PL, phospholipid: UC, unesterified cholesterol; EC, esterified cholesterol.

HDL [3H]/[‘4C]UC transfer (nmol/p mol per cell PL)

Control Phospholipase HDL A, HDL

19.74 25.30 ** (_+ 3.66) (54.50)

HDL f3H]EC uptake

(nmol/~mol per cell PL) [ ‘H]EC hydrolysis

(%)

11.94 18.48 * ( _+ 3.55) (+6.18)

61.18 69.29 n.S. (* 8.54) (* 7.65)

In Table II is summarized the quantitation of the HDL cholesterol fluxes towards endothelial cells during 3 h incubations, normalized on a cell phospholipid basis. Large interexperimental varia-

-_ , .

r

25 50 100 200

HDL_CONCENTRATION

(,iio ch01~51croI /ml)

Fig 3. Kinetics of uptake of HDL 3H-labelled esterified cholesterol (EC) by endothelial cells. A. Time kinetics. B. Concentration kinetics. Legends as in Fig. 1. Total uptake of [3H]EC represents the sum of the cell recovered [3H]EC and the cell hydrolysed

[ ‘H]EC. Calculations are detailed in the text.

88

tions were noted for the uptake of 3H-labelled esterified cholesterol. Its value represented on average 60% that of the flux of 14C-labelled un-

esterified cholesterol. Most of the transferred 3H- labelled esterified cholesterol was found hydrolysed in the cells and this proportion increased

with time: from 56 + 7.5% at 1 h to 72.7 k 3.8 at 5 h. Pretreatment of HDL with phoshpolipase A,

proved always stimulatory for the cell uptake of

[ 3H]cholesteryl esters. This is better illustrated when every culture flask containing phospholipase AZ-treated HDL was compared to its paired coun- terpart incubated with control HDL (Fig. 4). The schedule of all our results shows a 100-120% stimulation for the uptake of 3H-labelled esterified cholesterol at 3-5 h, whereas the fluxes of unesterified cholesterol were only poorly affected

(* 10-25s). In most cases, about two thirds of HDL phos-

phatidylcholine (62.10 f 2.57%) were degraded by

the added phospholipase A,. In three experiments, hydrolysis was induced by phospholipase A, in some samples, while other ones received

Y

0 1 3 5

INCUBATION TIME (hours)

Fig. 4. Stimulatory effect of HDL phospholipolysis on the uptake of radiolabelled HDL cholesterol. Endothelial cells were incubated either with control or with phospholipase A,-

treated HDL. Fluxes of 3H/14C-labelled unesterified

cholesterol (0) and of 3H-labelled esterified cholesterol (0)

were calculated as in Table III. All the experimental points

performed with lipolysed samples were compared to their

paired counterparts incubated with control HDL (value of 1.0).

The stimulation factor is expressed for phospholipase A, HDL

(means and SE.). The average degree of phosphatidylcholine degradation was 62.10 * 2.57. Statistical significance: * P <

0.05, ** P < 0.05, * ** P < 0.001. NS, not significant.

phospholipase A, so as to obtain different degrees of phospholipolysis and to compare the two en-

zymes. Using phospholipase A,, no stimulation was observed for a moderate phospholipid degradation ( < 27%). With phospholipase A z induc-

ing a 45% PC breakdown, the stimulation factor ranged between 1.12 and 1.8 for the uptake of

3H-labelled esterified cholesterol. An increased effect (1.73-2.73) was noted at higher phospholipid hydrolyses (> 70%) using either phospholipase A, or phospholipase A, (not shown). Thus, the enhanced uptake induced by phospholipase treatment seems related to the degree of lipolysis rather than to the position of the acyl ester bond hydrolysed.

Since the cell cultures were heterogeneous in terms of cholesterol content (Fig. 2), we looked for

a possible modulation of the transfer of HDL cholesterol by the amount of cellular cholesterol.

As stated above (Fig. 2) a linear relationship was observed when plotting different experiments for the movements of HDL unesterified cholesterol. As regards the uptake of HDL esterified cholesterol, a significant correlation (r = 0.78; P < 0.05) was noted at 5 h incubation, whereas it was not significant at 3 h (not shown). However, since interexperimental heterogeneity may arise from several other factors, an attempt was made to modulate the cellular cholesterol within the same cultures. Three different preconditioning media were used in parallel: medium containing human serum (‘normal’), lipoprotein-deficient serum (‘cholesterol depletion’) or lipoprotein-defi-

cient serum plus a high amount of LDL (‘cholesterol enrichment’). The uptake of HDL 3H-labelled esterified cholesterol at 3 h was measured thereafter. The cholesterol/phospholipid ratio could thus be varied from 0.35 to 0.52 (Table III). Surprisingly, the uptake of 3H-labelled

esterified cholesterol was significantly enhanced (P < 0.01) when comparing the sterol-enriched to the cholesterol-depleted state.

Cell uptake of 3H-labelled esterified cholesterol was also compared at confluence and 48 h earlier when cells were actively dividing. Expressed on a cell phospholipid basis, the measured transfers were essentially similar between the two stages (not shown). Finally, in one experiment, a partially purified cholesteryl ester transfer protein

TABLE III

EFFECTS OF DIFFERENT PREINCUBATION CONDI-

TIONS ON THE CELL CHOLESTEROL CONTENT AND

ON HDL CHOLESTEROL TRANSFERS

Under condition 1, cultures were preincubated with 10% lipo-

protein-deficient serum (LPDS) for 48 h. Under condition 2,

cells were preincubated for 48 h with 20% (v/v) human serum.

Under condition 3, cells were preincubated with 10% (v/v)

LPDS for 24 h and then 24 h with LPDS plus LDL (100 pg

protein/ml). Thereafter, cultures were washed and fresh

medium was then added with labelled HDL (50 pg

cholesterol/ml) as described above. Cells were incubated for 3

h and the uptake of [ 3H]/[14C] cholesterol was calculated as in

Table II. Values of each experimental condition are the mean

( + S.E.) of three experiments. Statistical comparison between 1

vs. 2 (a), 2 vs. 3 (b), 1 vs. 3 (c). * P < 0.05, * * P i 0.01. ns.,

not significant; PL, phospholipid; UC unesterified cholesterol;

EC, esterified cholesterol.

Conditions Cholesterol/ Transfer of HDL Uptake of

phospholipid [3H]/[‘4C]UC HDL [ ‘H]EC (mole ratio) (nmol/pmol (nmol/~mol

per PL) per cell PL)

1. LPDS 0.35 a,” 5. 17.60 a. n.s. 17.30 a,“.s. ( * 0.05) c. * ( $2.47) ‘, ” ’ (&2.51)‘,**

2. Human 0.45 15.17 20.95

serum (kO.03) ( + 2.60) (+ 2.50)

3. LPDS 0.52 ‘. * 18.54 b.* * 26.22 b. KS.

+ LDL ( & 0.04) ( f 2.77) ( * 2.02)

isolated from rabbit plasma [38] was included in

HDL cell incubations (0.6 mg protein correspond-

ing to 80 I.U./flask cholesteryl ester transfer activity). The fluxes of free cholesterol were almost unaffected in the presence of the protein. On the other hand, the uptake of 3H-labelled esterified cholesterol from control HDL was 24% lower when cholesteryl ester transfer protein was included, and the inhibition was still more pronounced with lipolysed HDL (-45%). This inhibitory effect might result from binding of HDL particles to the enzymatic protein.

4. Phospholipid and cholesterol contents in cells, during incubation with HDL

To check whether HDL incubations would have an influence on endothelial cell lipids, cholesterol and phospholipid mass were measured and referred to the cellular protein. The cell cholesterol content displayed almost no change following 5 h incubations with no lipoprotein (from 81.1 & 7.0 to 83.4

89

k 11.9 nmol/mg cell protein), whereas a slight but significant enrichment was registered in the presence of HDL (95.1 + 12.6 nmol/mg protein, P < 0.05 compared to zero time, n = 3). In all cases, 80-85s of the sterol was in non-esterified form. The amount of cell phospholipid remained constant during incubation with control HDL (221.9 + 10.2 nmol/mg protein). No significant difference was noted when addressing phospholipase-treated compared to untreated HDL, al- though cholesterol and phospholipid values were on average 3-6s greater in the former case. The cell phospholipid distribution was comparable

whether phospholipase A, was included or not in

the medium, yet some accumulation of lysophosphatidylcholine was detected with the enzyme present (7.4 * 1.9% versus 4.4 & 1.4%, P <

0.05, n = 4). As there was no evidence of hydrolysis of the cell glycerophospholipid (not shown), these lyso compounds may have originated from the surface of the lipolysed particles.

Discussion

The present study describes the movements of radiolabelled cholesterol from HDL to human endothelial cells. Confluent monolayer cultures took up HDL esterified cholesterol, representing an amount of 2.5-5% their own cell cholesterol content, an actively hydrolysed the acquired sterol.

Pretreatment of HDL with purified phospholipases (A2 or A,) doubled the transfer of HDL esterified cholesterol and its further hydrolysis.

Fluxes of free cholesterol between HDL and cells were also registered and were proportional to the pool size of the acceptor compartment, suggesting that it may predominantly represent an exchange phenomenon. This latter process was less affected by lipoprotein lipolysis.

The interaction between HDL, labelled on their apolipoprotein moiety, and endothelium has been described with cells from umbilical cord vein [lo], aorta [9,11], or with fetal heart cells [lo]. In these cell models, as in skin fibroblasts [39-411, HDL- receptor activities differ in many respects from the well characterized LDL-receptor pathway [42]. For instance, HDL binding is difficult to saturate, displays a low affinity and leads to very little secondary internalization and degradation

[9,39,41]. Furthermore, the latter does not seem sensitive to lysosome blockers [9]. The specificity of apolipoprotein recognition is also questionable, since LDL or various apolipoprotein-liposome complexes may compete for the binding of HDL, to endothelial cells [9-111 or fibroblasts [40]. In both cell types, HDL receptors were shown to be up regulated after loading the cells with non-lipoprotein or LDL cholesterol [11,41,433. Finally, the association of HDL with endothelial cells is poorly affected by the cell density or the degree of cell organization, contrary to what is observed for the processing of LDL apolipoprotein [3]. In a recent report however, a HDL-binding protein of a 110 kDa molecular mass has been identified in several cell types, including bovine aorta endothelial cells

]441. The uptake of HDL cholesteryl ester described

here, which represent a selective delivery of molecules to cells, fits in with some features mentioned above for HDL apolipoprotein. Uptake showed no saturation at high lipoprotein concentrations, and was quantitatively comparable in confluent or non-confluent cell layers. This uptake could also be modulated by varying the cell cholesterol content, as reported earlier for the binding of HDL apolipoprotein [41,43].

However, in other respects, the processing of HDL cholesterol strongly differs from that described for apolipoprotein. Most of the transferred cholesteryl ester was found hydrolysed in cells (about 75% at 5 h) which suggests an internalization or a membrane-associated hydrolysis of the component. In addition, only 20% of the cell radioactivity could be released by trypsin, indicat- ing again that most of the transferred cholesterol was recovered inside the cells. Differences also appear when comparing the amount of sterol taken up and that reported for HDL apolipoprotein in several studies on endothelial cells [9-111. At high HDL levels (up to 200 pg HDL protein/ml), the cell-recovered apolipoprotein after 3-6 h of incubation ranged between 100 and 300 ng/mg cell protein, which would correspond to the transfer of 0.4-1.2 nmol esterified cholesterol per ymol phospholipid, if HDL particles were taken up as a unit. The values of 3 H-labelled esterified cholesterol transfer measured here, at comparable concentrations, were lo-times greater on average, which

suggests different cellular fates for the different components of HDL. A preferential delivery of HDL esterified cholesterol over that of apolipoprotein A-I has been described, but in other cell types as hepatocytes or adrenals, which can utilize HDL cholesterol for steroidogenesis

I451. Confluent endothelial cells may also derive and

further hydrolyse cholesteryl ester and triacylglycerol from chylomicrons, and interestingly this process is competed for by HDL but not by LDL [4].

Those observations raise the question of the mechanism by which confluent cells can acquire cholesterol from HDL or triacylglycerol-rich lipoprotein. Unesterified cholesterol may exchange or diffuse in a continuous lipid layer through binding of particles onto cell membranes. As regards apolar lipids, a receptor-mediated endocytosis and degradation of the whole particles does not account for most of the transferable cholesteryl esters from HDL or chylomicrons. Rather, HDL may specifi- cally bind to endothelial cells by its apolipoprotein part, a process possibly influenced by lipid-lipid interactions [II], and cholesteryl ester molecules would diffuse to cells during this re- versible fusion before HDL particles are released. Another hypothesis, supported by microscopic evidence on cholesterol-en~ched macrophages or endothelial cells 1461, involves an intracellular traffic for internalized HDL, which proceed through cytoplasmic vesicles and are further re- secreted out of cells, thus escaping the lysosomal degradation. However, in this transit, lipid hydrolysis could occur via cytoplasmic hydrolases (471.

HDL phospholipid degradation always caused an increased delivery of HDL cholesterol to endothelial cells. Stimulation of the free cholesterol flux was already reported using hepatoma cells [19], or granulosa cells [20], which can either ac- comodate extra cholesterol or convert it into secreted steroids [20]. By contrast, in the present study, phospholipolysis enhanced the uptake from HDL of esterified cholesterol molecules more than of free cholesterol. Phospholipid breakdown, by altering the particle surface might enhance the cell binding. Alternatively, HDL binding would be minimally affected but the secondary transfer of apolar lipids could be facilitated by the removal of

91

phospholipid from the particle outer shell. An acceleration of the intracellular traffic of HDL may also be evoked.

The lysophospholipids generated are well- known fusogenic compounds, but albumin was present with a giycerophospholipid-to-bovine serum albumin mole ratio close to 1.0, thus in excess to take up most of the lipolysis products [29]. Moreover, when albumin concentration was varied between 0.03 mM and 0.3 mM, a similar stimulation was observed with phospholipases (not shown). However, one cannot exclude the possibil- ity that trace amounts of lysophosphatidylcholine or free fatty acids may affect HDL binding.

Several reports demonstrated that addition of purified lipoprotein lipase could trigger the delivery of labelled cholesteryl ester from liposomes to cultured endothelial cells, even when the former contained no substrate for lipoprotein lipase [7,8]. This is in favour of a ligand interaction between lipid particles and the cell-bound lipase. However, such an anchoring role seems unlikely for our enzymes, since the stimulated transfers were observed with two phospholipases, which are secretory enzymes, free in solution and unable to hydrolyse cell phospholipids, and because a threshold phosphatidylcholine hydrolysis was necessary to observe the lipase-induced stimulatory effect.

The supplying of HDL cholesteryl ester may deserve several roles in endothelial cell functions. First, as a source of exogenous cholesterol in contact-inhibited cell layers when both LDL uptake and de novo cholesterol synthesis are suppressed [1,2]. Secondly, HDL have proved mito- genie in a dose-dependent manner towards several types of endothelial cells, yet the nature of this effect is still unclear [48]. A recent report pro- posed a parallelism between this effect and the phosphorylation of a membrane protein following interaction of endothelial cells with HDL apolipoprotein or other growth factors [49]. In other studies [50], the mitogenic effect was related to the induction by HDL of the cellular hydroxymethyl- glutaryl-CoA reductase activity, joined with a supply in exogenous cholesterol. Human and rat HDL can also promote the prostaglandin I, production by porcine, bovine or human endothelial cells [51]. A possible mechanism is that HDL

cholesteryl esters would provide the cells with arachidonic acid, secondarily metabolized in pros- tacyclin or reacylated in cell phospholipids [52].

Finally, as a model, the uptake and hydrolysis of HDL esterified cholesterol by endothelial cells and its stimulation by phospholipolysis may raise hypotheses on the possible interrelations between parenchymal and non-parenchymal cells in steroidogenic tissues. Reports on the utilization of HDL cholesterol have mostly focused on HDL-free sterol as a possible precursor for bile acid synthesis [14], or progesterone production [20]. As well, HDL phospholipoIysis was found to favour mainly the fluxes of unesterified cholesterol from HDL. However, several reports indicate that non- parenchymal cells rather than parenchymal are most active in the degradation of lipoprotein both in vivo [21] and in vitro [23] and, accordingly, these cells were found enriched in acid lipase and cholesteryl esterase activities 1221. On the other hand, hepatic t~acylglycerol lipase has been located at the luminal surface of the liver capillary endothelium [24]. Thus, one might contemplate a cooperative scheme, in which the endothelium- bound hepatic triacylglycerol lipase would specifi- cally interact with circulating, large (sterol-rich) HDL, promote HDL phospholipid breakdown and thus an increased flux of free cholesterol/ cholesteryl esters and the subsequent hydrolysis of the latter in endothelial cells. The free sterol and other amphipathic molecules such as some fatty acids and lysophospholipids generated during lipolysis might be transferred to the underlying parenchymal cells by lateral diffusion through an interfacial continuum as suggested by Scow and coworkers 1531 for the interaction between lipolysed chylomicrons and fat cells.

In conclusion, the experimental model used throughout the study strongly argues in favour of a major effect of phospholipolysis on the transfer of esterified cholesterol from HDL to cultured endothelial cells. Whether this reflects exactly what occurs in sinusoidal liver endothelial cells in vivo remains still questionable however, since the transfer of free cholesterol was found to be much more affected by phospholipase A treatment in other systems such as hepatoma [19] or granulosa cells [20]. Therefore, further studies using purified liver endothelial cells containing or not containing

92

hepatic triacylglycerol lipase adsorbed anto their surface will be necessary to better define the pre- cise role of the enzyme in the processing of HDL cholesterol by the liver.

References

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Vlodavsky, I., Fielding, P.E., Fielding, C.J. and Gospo-

darowicz, D. (1978) Proc. Natl. Acad. Sci. USA 75,356-366.

Fielding, P.E., Vlodavsky, I., Gospodarowicz, D. and Field-

ing, C.J. (1979) J. Biol. Chem. 254, 749-755.

Kenagy, R., Bierman, E.L., Schwartz, S. and Albers, J.J.

(1984) Arteriosclerosis 43, 365-375.

Fielding, C.J., Vlodavsky, I., Fielding, P.E. and Gospo-

darowicz, D.D. (1979) J. Biol. Chem. 254, 8861-8868.

Chajek-Shaul, T., Friedman, G., Halperin, G., Stein, 0. and

Stein, Y. (1981) B&him. Biophys. Acta 666, 147-1255.

Cheng, C.F., Oosta, G.M., Bensadoun, A. and Rosenberg,

R.D. (1981) J. Biol. Chem. 256, 12893-12898.

Friedman, G., Chajek-Sham, T., Stein, O., Olivecrona, T.

and Stein, Y. (1981) Biochim. Biophys. Acta 666, 156-164.

Stein, O., Halperin, G., Leitesdorf, E., Olivecrona, T. and

Stein, Y. (1984) B&him. Biophys. Acta 795, 47-59.

Tauber, J.P., Goldminz, D., Vlodavsky, I. and Gospo-

darowicz, D. (1981) Eur. J. B&hem. 119, 317-325.

Havekes, L., Schouten, D., Van Hinsbergh, V. and De Wit,

E. (1984) Biochem. Biophys. Res. Commun. 122, 786-790.

Tabas. I. and Tall, A.R. (1984) J. Biol. Chem. 259,

13897-13906.

Stein, Y., Dabach. Y., Hollander, G., Halperin, G. and

Stein, 0. (1983) Biochim. Biophys. Acta 752, 98-105.

Glass, C., Pittman, R.C., Weinstein, D.B. and Steinberg, D.

(1983) Proc. Natl. Acad. Sci. USA 80, 5435-5439.

Halloran, L.G., Schwartz, CC., Vlahcevic, Z.R., Nisman,

R.M. and Swell, L. (1987) Surgery 84, l-6.

Hylemon, P.B., Gurley, E.C., Kubaska, W.M., Whitehead,

T.R., Guzelian, P.S. and Vlahcevic, Z.R. (1985) J. Biol.

Chem. 260, 1015-1019.

Herscowitz, M., Roneu, I., Bilu, S. and Tietz, A. (1986)

B&him. Biophys. Acta 878, 426-434.

Jansen, H. and De Greef. W.J. (1981) Biochem. J. 196,

739-743.

Shirai, K., Barnhart, R.L. ad Jackson, R.L. (1981) B&hem.

Biophys. Res. Commun. 100, 591-599.

Bamberger, M., Lund-Katz, S., Phillips, M.C. and Roth-

blat, G.H. (1985) Biochemistry 24, 3693-3701.

Parinaud, J., Perret, B., Ribbes, H., Chap, H., Pontonnier. G. and Douste-Blazy? L. (1987) J. Clin. Endocr. Metab. 64,

409-417.

Van Berkel, T.J.C. and Van Tol. A. (1978) Biochim. Bio- phys. Acta 530, 299-304.

Van Berkei, T.J.C., Vaandrager, H., Kruut, J.K. and Foster, J.F. (1980) Biochim. Biophys. Acta 617, 446-457.

Wandel, M., Norum, K.R., Berg, T. and Ose, L. (1981) Stand. J. Gastroenterol. 16, 71-80.

Kuusi, T., Nikkila. EA., Virlanen, J. and Kinnunen, P.K.J.

(1975) B&hem. J. 181, 245-246.

Chap, H., Zwaal, R.F.A. and Van Deenen, L.L.M. (1977)

Biochim. Biophys. Acta 467, 1466164.

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

Fauvel, J., Bonnefis, M.J., Sarda, L., Chap, H., Thouvenot,

J.P. and Douste-Blazy, L. (1981) Biochim. Biophys. Acta 663, 446-457.

Slater, D.N. (1976) Atherosclerosis 25, 237-244

Jaffe, E.A., Nachman, R.L., Becker, C.G. and Minnick, C.R. (1973) J. Clin. Invest. 52, 2745-2756.

Perret, B.P., Chollet, F., Durand, S., Simard, G., Chap, H.

and Douste-Blazy, L. (1987) Eur. J. Biochem. 162,279-286.

Bligh, E.G. and Dyer, W.J. (1959) Can. J. Biochem. Phys- iol. 37, 911-918.

Lowry, OH., Rosebrough, N.J., Farr, A.L. and Randall,

R.J. (1951) J. Biol. Chem. 193, 265-275.

Roschlau, P., Bernet, E. and Gruber, N. (1974) Z. Klin.

Chem. Klin. Biochem. 12, 403-407.

De Hoof, J.L., Davidson, L.M. and Kritchevsky, D. (1978)

Clin. Chem. 24,433-435.

Bottcher, C.J.F., Van Gent, C.M. and Pries, C. (1961) Anal. Chim. Acta 24, 203-204.

Wahefeld, A.W. (1974) in Methoden der enzymodischen

Analyse (Bergmeyer, H.U., ed.), Tome II, p. 1978, Verlag-

Chemie, Weinheim.

Skipski, V.P., Peterson, F.F. and Barclay, M. (1964) Bio-

them. J. 90, 374-379.

Curry. M.D., Gustafson, A., Alaupovic, P. and McConathy,

W.J. (1977) Clin. Chem. 24, 280-286.

Ierides, M., Dousset, N., Potier, M., Manent, J., Carton, M.

and Douste-Blazy, L. (1985) FEBS Lett. 193, 59-62.

Fidge, N.H., Nestel, P.J. and Suzuki, N. (1983) B&him.

Biophys. Acta 753, 14-21.

Fidge, N.H. and Nestel, P.J. (1985) J. Biol. Chem. 260,

3570-3575.

Oram, J.F., Brinton, E.A. and Bierman, E.C. (1983) J. Clin. Invest. 72, 1611-1621.

Goldstein, J.L. and Brown. M.S. (1977) Annu. Rev. Bio-

them. 46, 897-930.

Tauber, J.D., Goldminz, D. and Gaspodarowicz, D. (1981)

Eur. J. Biochem. 199, 327-339.

Graham, L. and Oram, J.F. (1987) J. Biol. Chem. 262,

7439-7442.

Leitersorf, E., Israeli, A., Stein, O., Eisenberg, S. and Stein,

Y. (1986) B&him. Biophys. Acta 878, 320-329.

Schmitz, G., Assman, G., Robenek, H. and Brennhausen,

B. (1985) Proc. Natl. Acad. Sci. USA 82, 6305-6309.

Brecher, P., Kessler, M., Cliffard, C. and Chabanian, A.V.

(12973) Biochim. Biophys. Acta 316, 386-389.

Tauber, J.P., Cheng, J. and Gospodarowicz, D. (1980) J.

Clin. Invest. 66, 696-708.

Darbon, J.M., Tournier, J.F., Tauber, J.P. and Bayard. F.

:1986) J. Biol. Chem. 261, 8002-8008.

Cohen, DC., Massoglia, S.L. and Gospodarowicz, D. (1982) 1. Biol. Chem. 257, 9429-9437.

Fleisher, L., Tall, A., Witte, L. and Cannon, P. (1982) J. Biol. Chem. 257, 6653-6655.

Pommerantz, K.B., Fleisher, IN., Tall, A.R. and Cannon,

P.J. (1985) J. Lipid Res. 26, 1269-1276. Scow, R.O., Blanchette-Mackie, J. and Smith, L.C. (1975)

Xrc. Res. 39, 149-162.

uptake of hdl unesterified and esterified cholesterol by human endothelial cells. modulation by hdl...

Documents