behaviour of phospholipase modified-hdl towards cultured hepatocytes. ii. increased cell cholesterol...

Eltr chtmitu et Biophyst('a A(':::. 1081 ( 1991 ) 211 - 2 1 9

' 1991 Elsevier Science Publishers B.V_ (Biomedical Divisi(~n) 0005-2760/91/$03 5~ A DONIS 0005276{)9100078S

BBALIP 53544

Behaviour of phospho!ipase modified-HDL towards cultured hepatocytes. II. Increased cell cholesterol storage and bile

acid synthesis

Xav ie r Col le t , C l a u d e Vieu, H u g u e s C h a p a n d B e r t r a n d - P i e r r e Per re t

lnseroz Unil~ 326. Pho.~pholipides memhranairc,.~', mgnahsarton celh:latre et hpoprott;tne& tl~p:tal Pu.rpatl. TouhJ~t~e ('cdex ( l"rance)

( Received 24 J,,llv l'4gll;

211

Key ~Jrds: HDL: Phospholipase: ('latdesterol: Bile acid- (Ral hcpatocyte)

Human total HDL (hydrated density. 1.070-1.210), HDL., ( !.070-1.125), HDL~ (i.125-1.210) or t tDL separated by heparin affinity chromatography were treated with or without purified phospholipase A z from Crotal~ a~'amanteus. Control and treated HDL were reisolated and were then incubated with cultured hepatocytes. !. Ma.~s measurements evidenced a time-dependent cholesterol enrichment in hepatocytes cultured in the absence of lipoproteins. Addition of H D L 2 still enhanced by 25% the cell cholesterol content and down-regulated endogenous sterol synthesis in similar proportions. Conversely, I-IDL 3 slightly decreased the amount of free cholesterol in hepatocytes ( - 12%). 2. Incubations with phospholipase A 2-treated H D L resulted in a 35%-50% increase of both the cellular cholesterol esterification and the cholester~lester accumulation, when compared to cells cultured it~ the pre~nce of control-ItDL. This effect was observed with HDL2, HDL3 and combil~ing the data with all subfractions. 3. Cultured hepatocytes .secreted cholic and O-muricholic acids as major bile acids and HDL~ showed a tendency Io stimulate their secretion. Phospholip~tse treatment of t tDL again induced an increased production by hepatucytes of those two bile acids. Thtt~, wherea~s HDLz and HDL 3 display different hehaviours with respect to cell cholesterol content, neosynthesis and bile acid .,~cretion, their modifications by phnspholipases always orientate the cell sterol m,;tabolism in the same direction: increased cholesterylester accumulation and bile acid production.

Introduction

Hi~h-density lipoproteins are a pivot in the process of reverse cholesterol transport. In the past recent years, numerous studies have described the uptal.e of HDL lipids and apoproteins by liver cells [1-4], and there has been several attempts to characterize HDL binding sites [5-7]. Besides specific interactions between HDL and cells, it has been proposed that hepatic lipase, a lipolytic enzyme present in liver capillaries and also detected in adrenals and ovaries [8], would favour the delivery of HDL cholesterol to liver and to steroidogenic tissues [9]. This effect, at first suggested by in vivo studios on the

Abbreviations: HDL: high-density lipoprotein: PLA,-HDL. phospht~- lipase A.,-treated HDL; F.C., free cholesterol; GLC. gas-liquid chromatography.

Correspondence: B. Ferret, INSERM Unitt~ 326, Phospholipides membranaires, signalisation cellutaire et lipoprot~ines, Hfpital Put- pan, 31059 Toulouse Cedex. France.

immunological inhibition of the enzyme [10,11], would be mediated by the phospholipase A-i activity of hepatic lipase to,wards large, sterol-rich, HDL particles [12,13].

Indeed, in a first report 114], we observed that phospholipase-modified HDL delivered more radiolabelled free cholesterol, cholesteryl ether and apoproteins to cultured hepatocytes, compared to control HDL. Thi,~ was associated with an enhanced binding capacity of lipolysed HDL onto cells. Furthermore, although more cholesterol was transferred from (control) HDL~ than from HDL 3, the phospholipase-induccd modifications during further incubations with cells were recorded with both subfractions.

Nevertheless, whether those phospholipase-stimulated transfers of HDL components result in an increased utilization of HDL-cholesterol, or whether they merely reflect an accelerated transit of the particles through the cells, is so far unknown. The reported observations of reversible interactions between HDL and hepatocytes [3], and of a retroendocytotic pathway in hepatoma cells [15] could support either hypothesis. Finally, if many HDL subfractions can interact with hepatocytes, involv-

212

ing apoproteins A-I A-II and A-IV as ligands [5- 7.16,17], the final sterol level in cells may depend on the particle size or on their lipid content.

Hepatocytes can utilize .exogenous cholesterol for bile acid synthesis [18], and can store excess sterol as cholesterylesters, components of nascent lipoproteins such as VLDL [19]. In this paper, we have follcwed the effects of native and phospholipase-modified HDL subfractions on several aspects of cholesterol homeostasy in cultured hepatocytes: cholesterol accumulation, esterification and neosynthesis and production of bile acids.

Materials and Methods

[4-]4C]Cholesterol and [4-14Clglycocholic acid were purchased from Amersham-France (Paris Los Ulis. France) and [2-~4C]acetic acid fr.gm CEA (Gif sur Yvette, France).

The following compounds were obtained from Sigma (St Louis. MO, U.S.A.): type IS collagenase, human insulin; cholestyramine; fatty acid free bovine serum albumin (BSA): lithocholic acid; taurodeoxycholic acid; giycochenodeoxycholic acid; glycocholic acid; hyocholic acid and ursodeoxycholic acid. fl-Muricholic acid was a generous gift from Dr. JC Montet. Mar';eille, France [201. Heparin Sepharose CL6B was from Pharmacia- LKB (Uppsala, Sweden). Sep-Pack Clx columns came from Waters Associates (Milford, MA, U.S.A.)

Col/c,dture Male Wistar rats (200 g) fed ad libitum, were treated

for 48 h by cholestyramine as included in the diet (4%, ~ /v) . Hepatocytes were prepared following a recycling perfusion of the liver in presence of collagenase for t0 rain at 37°C, as described by Durrington et ai. [21], The monolayer cultures of hepatocytes were prepared and incubated as previously described [14].

l.s'olatior~ of plasma lipoproteins HDL were isolated from normolipidemic healthy wo-

men by uhracentrifugation. Total HDL were isolated between the densities of 1.07 and 1.21 and were washed at their lower density limit. Purity of the HDL preparations was ascertained by electrophoresis on 2-3% polyacrylamide, under non-denaturing conditions (Sebia-lssy los Moulineaux, France). Their chemical composition, expressed as percent weight, was the following: protein 51.6 +_ 1.7% phosphotipid 26.0% + 1.3go; esterified cholesterol 19.1% + 1.1%; free cholesterol 2.3% + 0.2% and triacylglycerol 1.4% + 0.2%. Total HDL were either subfractionated as HDL, (1.070-1.125) and HDL 3 (1.125-.1.210), or were separated by heparin affinity chromatography [14,22]. All subfractions were checked for Ihc presence of apo E by 0.1% (w/v) S D S / l l % (w/v) polyacrylamide gel electrophoresis, followed by

Coomassie blue staining or by immunoblotting with a specific monoclonal antibody (see below). Apoprotein E was also quantified using a specific radioimmunoassay [231.

As previously described [14]. HDL~ and the heparin-Sepharose unretained fraction displayed similar hydrated densities and chemical characteristics (cholesterol : protein = 0,62 + 0.07 p, mol /mg) and will be re- ferred to as HDL 3. As well, HDL1 and the retained fraction will be assimilated (cholesterol:protein, 0.98 + 0.11 t tmol/mg). The lipoproteins were ..tialysed against Tris-NaCl buffer, 0.01 M/0.135 M (pH 7.4). containing NaN 3 (0.01%. w/v) and EDTA (0.25 mM) and were stored in a dark place under nitrogen.

Apoprotein A-I accounted for 82% + 4% of HDL 2 protein, and 71% +_. 3% in HDL.~. Apoprotein E represented only 1% of total apoprotein in HDL2 and trace amounts were occasionally detected in HDL 3 by im- munoblot assays.

Labelling of HDL with unesterified [14Clcholesterol was performed through exchange, by incubating the labelled sterol, spotted onto ac id /e thanol washed Whatman paper No. 1, with the lipoprotein preparations (0.6 p, Ci/p, mol HDL cholesterol). Incubations were for 240 min at 37°C, in the Tris-NaCl buffer containing EDT,t, and NaN 3. as above. Labelled HDL were recovered by filtration. Their specific radioactivities ranged between 3, l0 s and 1.7,10 ~' dpm// tmol free cholesterol and no '4C-labelling of HDL esterified cholesterol was delected.

Phospholipase A, treatment of HDL subfractions HDL subfractions (0.6 mM total cholesterol) were

incubated in the presence of bovine serum albumin (1%, w/v) , 3 mM CaCI_, and with or without purified phospholipase A2 from Crotalus adamanteus (200 mU/ml ) , during 60 min at 37 °C [14,24]. The reaction was stopped by addition of 0.01 M EDTA (pH 7.4) and HDL were re-isolated by ultracentrifugation at d : 1.21. They were then dialysed against 0.135 M NaCI,/0.015 M Hopes (pH 7.4) and were further sterilised by filtering through 0.22 p,m Millex GS filters. The average phosphatidyl- choline hydrolysis was 64.5% + 3.2%. The properties of modified lipoproteins have been described elsewhere [24t.

locorporation of [ t4C]acetate into celhdar sterois Incorporation of [14C]acetate into cell sterols was

followed according to Azhar et al. [25]. Hepatocytes were incubated for 6 and 24 h without or with lipopro- reins (0.13 mM cholesterol) and the culture supernatants were removed and stored at - 2 0 ° C , till analysis. The cells were then washed five times with Hanks" medium, the first two washes containing 0.2% (w/v) albumin. Hepatocytes were then incubated for 2 h at 37°C with [~4Clacetate, 1 FCi/ml, in 4 ml DMEM,

supplemented with albumin (1%. w/v) and glucose (5 raM). At the end of the labelling period, the medium was removed and the cell monolayer was washed twice again in the presence of albumin and then five times without albumin but in the pre~ence of sodium acetate (2 raM), so as to get rid of excess, non incorporated, labelled acetate. CeUs were then taken off in 2 ml Hanks" medium containing 0.05 M NaOH and were sonicated 4 x 15 s. at 0-4 ° C. Duplicate aliquots were taken for protein measurements and lipids were extracted. The chloroform phase was washed three times with methanoi/H_~O (1/1, v/v), to remove traces of free in4 C]acetate. Control extractions including [ 3 H]cholesterol and [*4C]acetate indicated a recovery of 95~ for the foxmer, with a very low contamination by [l":C]acetate (<0.2%). Labelled lipids were then separated by thin-layer chromatography on silica gel G60 plates, using a first migration (up to 17 cm) with chloroform/methanol/acetic acid (98:2:1 , v/v) and a second one, until the top of the plate, with hexane/ diethylether/acetic acid (96 : 4:0.2, v/v). The different labelled sterols were detected by autoradi~gaaphy and radioactivity was measured after scraping the differents spots, identified by comparison with reference standards.

Measurements of bile acids

1. Extraction of conjugated bile acMs. The extraction procedure was realised according to de Mark et al. [26], using reverse phase chromatography on Sep-Pack C I,~ columns. The latter were first washed with 10 ml methanol and then with 10 ml distilled water. Culture supernatants collected after incubation with hepatocytes, were centrifuged at 1000 × g for 10 min to get rid of cell debris, and were then diluted with an equivalent volume of 0.1 M NaOH. This solution was then applied onto the Sep-Pack column at a flow rate of 1 ml/min. Different solvent mixtures were then passed through: 5 mi 0.1 M NaOH (which removed 98% of HDL cholesterol, as assessed by running HDL samples radio- labeiied on free and esterified cholesterol): then 20 ml water and finally 10 ml methanol/H.,O (2:3, v/v). Conjugated bile acids were elated with methanol /H,O (2 : 1, v/v) and were left to dry. The bile acids were then dissolved in 4 ml water. A second more rapid purifica- tion step was performed, using a new Sep-Pack column prepared as above. The bile acid mixture was loaded and the column was washed with 10 mi H,O, followed by 10 ml hexane and then 10 ml H,O. The final elation was carried out with 20 ml methanol/H_~O (3: 1, v/v). After evaporation, bile acids were again dissolved into water (1.8 ml), before extraction according to the Bligh and Dyer procedure [27], as modified by addiiioa of 1.4 M NH4OH (0.04 ml/ml). The upper phase was evaporated and dissolved in 1 ml methanol. Two sam-

213

pies containing [14C]glycoeholic acid as a tracer allowed to estimate the extraction yield in each series (n = 6}: between 85% and 95%. Some parallel experiments were performed to follow the incorporation of HDL-free [laClcholesterol into conjugated bile acids. Distribution of radioactivity in tauro- or gly,o-conjugates was measured by thin-layer chromatography in chloroform/ isopropanol/acetic ac id /H 20 (30 : 30:4 : 1, v/v).

2. AIkaline hydrolysis of conjugated bile acids. Hydroi- d'sis was carried out according to Youssef et ai. [281. The dried extracts were dissolved in 5 M NaOH and were incubated in sealed tubes for 3 h at I15°C. After cooling, the extract was washed three *Ames with 4 ml petroleum ether to remorse any coJltanlinant cholesterol. The extract was then acidified with 2.5 M HCI (until pH i). Free bile acids were then extracted three times with diethyl ether and were then put to dryness.

3. Bile acid derit'ati:ation, Methyl ester derivatives were obtained by incubation of free bile acids with methanol/acetyl chloride (20: I, v/v), heated at 50°C for 1 h. To prepare acetate derivatives, the extract was again dissol'-ed in 0.5 ml acetyi chloride and heated at 50°C for 5 rain. Derivatized bile acids were dissolved in 100 ~tl ethyl acetate,

4. Gas-liquid chromatography (GLC). Derivatized bile acids were :malysed on an Intersmat (Paris, France) IGC 120 DLF gas chromatograph, equipped with a solventless injector flame ionisation detector and a Hewlett-Packard fused silica capillary column (5 m × 0.31 mm i.d.), coated with SP 2100 (cross-linked methyl silicone). The oven temperature was programmed from 230°C to 260°C at a rate of 4 C ° / m i n and the carrier gas was hydrogen (0.05 bar). Retention times and areas were measured with a Spectra Physics (Paris. France) SP 4270 integrator.

Bile acid quantitation was reaIised using an internal standard of ursodeoxycholic acid (1,25 /tg), which was added to each culture supernatant before loading onto Sep-pack columns.

et nalrtWal methodv Total and free cholesterol in HDL were assayed with

the cholesterol esterase/cholesterol oxidase technique [29] using commercial kits (Boehringer-Mannheim, F.R.G.) and aqueous cholesterol standards (Sigma). This method was adapted to cell lipid extrat,~s with the following modifications: dried lipid extracts and cholesterol standards were first dissolved in 0.05 ml methanol before addition of enzymatic reagents. Con- trol determinations performed by gas-liquid chromatography showed the excellent correlation between both methods. Lipids were extracted according to Bligh and Dyer [271, unless otherwise stated. Total phospholipid was measured as the lipid phosphorus content [30]. Phospholipid classes were separated by monodimen- sional thin-layer chromatography (TLC) on ,~ilica gel G

214

60 plates (Merck, D a r m s t a d t , F .R .G , ) , using the so lvent :~ystem of Skipski et al. [31], W h e n cells were i ncuba t ed in p resence o f H D L l abe l l ed wi th unes t e r i f i ed [14C]cholesterol, s epa ra t i on of the cell r ad ioac t ive s te ro ls was pe r fo rmed by TLC, using p e t r o l e u m e t h e r / d i e t h y l e t h e r / a c e t i c acid (165 : 35 : 2, v / v ) , as a solvent .

Prote in was d e t e r m i n e d acco rd ing to Lowry et al.

[32], using BSA as a s t anda rd . Rad ioac t iv i ty was measu red us ing a l iquid sc in t i l la -

t ion coun te r ( P a c k a r d 4530, Zur ich , Swi tze r land) wi th

au toma t i c quench ing cor rec t ion . A p o l i p o p r o t e i n s A-I and B were d e t e r m i n e d by im-

munoe lec t rod i f fu s ion us ing a commerc i a l kit (Sebia , Issy les Moul ineaux , F r ance ) [33]. A p o p r o t e i n E in

H D L subf rac t ions was measu red by c o m p e t i t i v e rad io -

immunoas say , us ing V L D L a p o E as the i m m o b i l i z e d ant igen and the 6C5 monoc iona l an t i body , k ind ly p ro - v ided by Dr. Ross Milne , Mon t r ea l , C a n a d a . [23].

Statistics Resul ts are means __ S.E. S ta t i s t ica l c o m p a r i s o n s were

pe r fo rmed using S tuden t ' s t - test for p a i r e d samples .

Resul ts

I. Effects of HDL subfractions on the cholesterol content of cultured hepatocytes. Modulation by HDL phospholipol)'sis

Dur ing i ncuba t ions of cu l tu red h e p a t o c y t e s in the absence of l ipopro te ins , the ce l lu la r cho les te ro l con t en t i n c r e a s e d w i th t i m e , f r o m 34.50 4- 2 .74 n m o l c h o l e s t e r o l / r a g cell p ro t e in at 0 t ime to 46.44 + 3,55 a f te r 24 h ( P < 0.02) (Fig . 1). Th is rise mos t ly con- ce rned the a m o u n t of es ter i f ied choles tero l , which in- c reased f rom 3.05_+ 1.92 n m o l / m g at 0 t ime up to

8.45 _+ 1.05 and 14.40 + 4.45 n m o l / m g at 6 and 24 h,

701 - - 6 0

01o 0

m c

~ 3 0

0 [ . . . . . . 0 6 h

(a)"

24 h Time

Fig. 1. Effects of different HDL suhfractions on the cell cholesterol content. Hepatocytes were incubated for various times in the presence of HDL subfractions at a concentration of 0.164+0.012 mM cholesterol. HDL 3 (El). HDL~ (11). or without HDL (I-1). After washing the cells, protein and cholesterol were measured. Data represent means (4- S.E.) from six experiments. Statistical comparisons between HDL~ and HDL2 at 6 and 24 h la), between controls at 0 h and at 24 h lb). and between HDL3 and controls at 24 h (c): *P10 .05 :

** P <0.02.

respec t ive ly ( P < 0.02). O n the o t h e r hand , the c o n t e n t in free cho les te ro l d i d not change d u r i n g i n c u b a t i o n s ( f rom 30.7 + 3.1 to 32.2 4- 2.7 n m o l / m g be tween 0 a n d

24 h). T h e ef fec ts o f H D L sub f r ac t i ons on this p rocess were

inves t iga ted , l a the p resence of H D L 3 , a d d e d at an ave rage c o n c e n t r a t i o n o f 2 6 5 _ 30 # g p r o t e i n / m l , equ iva len t to 174 n m o l / m l H D L choles te ro l , the cell s teroi e n r i c h m e n t wi th t ime was less p r o n o u n c e d (Fig . 1). A t 24 h, there was a s ign i f i can t r educ t i on in cell

cho les te ro l , m o s t l y ref lec ted in the free cho les te ro l con-

tent , when c o m p a r e d to con t ro l cells i n c u b a t e d wi thou t

TABLE I

Cholesterol content of hepatocvws following 24-h im'ubatimts with H DL suh/ractions. Effects of phospholipol.rsLs

Hepatocytes were incubated with total HDL, HDL~ or HDL~ 10.166+0.02 mM chol.) treated or not by PLA~, After washing, the cells were scraped off "~,iih a rubber policeman and free cholesterol, total cholesterol and proteins were measured. Results are means _+ S.E. Statistical ct~mparison between control HDL and PLAa-HDL: * P < 0.05: ** P < 0.01.

IIDL PLase .,,ubfractions treatment

Cell cholesterol content (nmo[ cholesterol/rag cell protein)

total cholesterol free cholesterol cholesleryl ester

Total HDL - 44.4 + 3.2 35.5 4- 2.7 8.7 4-1.9 ( n = 61 + 53.1 + 5.8 43.1 + 5.1 9.9 4- 1 . 7

HIlL, - 57.1 +8.2 38.9__+5.3 17.8_+ 5.6 (n = 5) + 63.9+ 5.9 40.1 ±4.6 24.2± 5.8 *

HDL~ - 42.1 _+2.6 28.3_+2.2 13.5_+4A ( , = 8) + 47.7+4.4 28.0+ 2A 19,3_+5.3 *

All suhfraction~,, - 46.8 ± 3.1 34.0 + 2.2 ! 2.9 +_ 2.5 combined + 53.7±3.5 ** 35.9+2.6 17.6+_3.1 ** ( n = t9)

215

tipoproteios (28.3 + 2.1 versus 32.1 + 2.7 n m o i / m g respectively. P < 0.05, not shown).

Conversely, cells incubated wit'l HDL 2 at 169 __. 21 gg protein/ml , i.e., at comparable cholesterol con- centrations (159 nmol /ml) , increased by 23% their content in total chol 'sterol (57.1 + 8.1 with HDL~ versus 46.44 + 3.55 n m o l / m g in control), both as free and esterified cholesterol, although non-significantly if compared to 24 h controls. However. the differences between the cholesterol content in presence of HDL z or H D L 3 were significant at both incubation times.

Addition of unfractionated H D L resulted in a com- bination of those two opposite effects and thus induced almost no change in the cell cholesterol content (44.4 _ 3.2 with total HDL, versus 46.44 + 3.55 n m o l / m g in 24 h control cells). (Table I)

Whatever be the H D L subfractions addressed, the effect of a prior phospholipase treatment was unequiv- ocal and led to a further cellular enrichment in cholesterol ( + 15%, on average). This tendency was observed with any H D L subfraction (Table I) and considering all the data combined, this rise in cell total cholesterol in the presence of l ipolysed-HDL was found significant ( P < 0 . 0 1 ) compared to incubations with controI-HDL. This increase mostly concerned the content in esterified cholesterol ( P < 0.05 for PLAa-HDL: , PLAz-HDL3 and P < 0,01, all data combined).

IL Effects of native and iipolysed HDL subfractions on roe cholesterol synthesis

The increase in cellular cholesterol with time suggests an active neosynthesis of sterols in cultured hepatocytes. The latter was estimated by incorporation of [taC]acetate into sterols and the effects on this process of HDL2, HDL3 and their phospholipase-treated coun- terparts were investigated. Without addition of lipoprotein, acetate incorporation into cell sterols was maximal after 6 h culture and remained still at a high level after 24 h (+50% and 30% over 0 time respectively, not shown). Incuba t ion with HDL.~ (0.13 mM as cholesterol), had limited effects on the further incorporation of [taC]acetate, except for a slight and non-significant increase compared to control cultures incubated without lipoprotein (Table II). Conversely, HDL,_, present at a comparable cholesterol concentration, significantly reduced cell cholesterol synthesis ( - 2 1 % , Table il). These variations, although of low amplitude, corroborate the observations on the cell cholesterol content during incubations with HDL. For instance, HDL2, which provide hepatocytes with cholesterol, down-regulates its endogenous synthesis. Despite the very different effects of H D L , versus HDL.~. acetate incorporation in presence of phospholipolysed HDL subfracfions was 15% lower than with native particles (Table II). This is again concordant with the data on cell cholesterol content, comparing the cultures

1

2 4

5

8

l

4 5

il ' t ' 7

Fig. 2. Separation of bile acids produced by cultured hepatocytes by gas-liquid chromatography (GLC). A GLC recording of a standard mixture (A) including the following bile acids: (i) lithocholic acid: 12) deoxvchotic acid: (3) chenodexychotic acid: 14) cholic acid: (5) urso- dcoxycholic acid {internal standard): (6) hyocholic acid: (7) ,8- muricholic acid: (N.I.) non-identified. (B) Recording from a culture super'Latant previously purified by reverse-phase chromatography.

Cells were incubated for 6 h in the presence of HDL 10.13 raM).

in presence of control or l ipolysed-HDk, as observed above with both subfractions tested.

i l l Esterification of HDL-derited chole.~terol. Compari- son between natit,e and lipolysed HDL

Since the differences in cell cholesterol content between cultures exposed to either lipolysed or control HDL were found mostly in esterified cholesterol, we wondered whether incubations with phospholipase- modified HDL wottld stimulate intracellu!ar sterot esterification. To follow the fate of cell free cholesterol, in presence of control or treated HDL, hepatocytes were incubated with HDL, labelled with unesterified [14C]cholesterok prior to any tzeatment. The distribu-

216

TABLE I I

Rehttire mcorp,~ration of [t4Cjacetate m cell chofe~wrol in presence of control or PLA ,-treated HDL subfractitms

Cultured rat hepatocytes were incubated for 6 or 24 h with or without HDL subfractions (0.130_+0.006 mM as cholesterol), treated or nol by PLA z. After washing, cells were further incubated during 121:) rain with [14('l-acetate, I .uCi per ml. The radio, activity incorporated into free + esterified cholesterol in cells was then determined by TLC. The incorporation of [~'~C] acetate following p~ior incubations with HDL ~s expressed relative to control cells, preincuhated in absence of lipopmtein. When significantly different from comrol cells: *P < 0.1)5: '~ " P < O.Ol. Taken from four determinations.

H D L PLA _, Relative incorporation of subfraclion treatment [t'*CJacetate into cell stcrols

(ratio over incorporation in control cells)

HI)L: - 0.79+_0.06 * HI)l., + 0.67+_0.03 * * tIDI.~ - 1.14_+0.10 H DLa + 0.97 _+ 0.08

14 tion of C-radioact ivi ty a m o n g cell sterols was then determined. Table i l l depicts such exper iments using H D L 3 or unfrac t ionated H D L . Unde r these condi t ions , there is no major change in the cell cholesterol content (see above), so that the labelling of cell sterols results mainly f rom a bidirectional exchange.

'4C-radioactivi ty gradual ly appeared in hepatocytes as a function of time (not shown). Af ter 24 h, it represented, per mg cell protein, 11.5% of the initial free [14C]cholesterol in H D L (Table III). Part of the cell radioactivity was found as [14C]cholesteryl esters, pre- sumably thrc, ugh the action of A C A T . since the initial label was only in free cholesterol.

Phospholipa:;e-treated H D L delivered 1.25-fold more [;4C]cholestero', to hepatocytes, with an increase in bo th the ceil-associated free cholesterol ( + 2 5 % NS) and esterified cholesterol ~ + 5 0 %. P < 0.05). However . the specil'ic cadioactivities for cell free [14 C]cholesterol were

not much different after incubat ion with either control or lipolysed H D L . This implies that a direct compar i son between the amoun t s of ~4C-radioactivity in cholesteryl esters in the two systems, reflects differences in the mass amoun t s of ester[fled cholesterol derived from the cell labelled free cholesterol. Thus, it was est imated that 1.5-fold more cholesteryl esters are generated intra- cellularly in presence of l ipo lysed-HDL, than with unmodif ied particles. When hepatocytes were challenged with H D L 2, labelled likewise with free [~4C]cholesterol and present at similar concentra t ions , the radioact ivi ty recovered as cell free and ester[fled cholesterol was twice higher than with HDI_ 3 and was again s t imulated by phosphol ipases (not shown).

1 V. Influence o f native and lipolysed H D L on the bile acid secretion by hepatotytes

In these experiments , hepatocytes or iginated f rom choles tyramine- t rea ted rats in order to s t imulate the bile acid synthesis. Those cells were incubated in the presence of H D L subfract ions, at an average cholesterol concen t ra t ion o f 0,123 m M cholesterol and the bile acid p roduc t ion was measured in the cul ture media by gas- ' [quid ch roma tog raphy , fol lowing selective extract ion and derivat izat ion. Two major metabol i tes were identified: cholic acid and f l-muricholic acid (Fig. 2). Thei r secretion into cul ture med ium was maximal at 6 h and levelled off for longer incuba t ion per iods (not shown). In absence o f H D L , after 6 h incubat ion, the mean p roduc t ion o f f l -murichol ic acid was 0,41 _+ 0.08 n m o l / m g cell pro te in and that o f cholic acid, 0,55 4- 0.19 n m o l / m g (Fig. 3). Add i t ion o f H D L 2 increased their respective secret ions up to 0.71 4-0.23 and 0.97 5= 0.31 n m o l / m g , a l though the differences over cont ro l cells incuba ted wi thout l ipoprotc ins were not significant, due to in terexper iment variat ions. (Fig. 3). The effects o f H D L 3 or of unf rac t iona ted H D L on the p roduc t ion o f those bile acids by hepa tocy tes were minimal.

Pre t rea tment of H D L with phosphol ipase A 2 had a

TABLE IIl

Cell uptake and esterificalton of It DL free [ I~C]choleslero!

Rat cultured hepatocytes were incubated for 24 h with HDL (either total HDL or HDLD. radiolabelled with free [~4Clcholesterol (FC) and were treated with or without phospholipase A 2, The average HDL concentration was 166 #M total cholesterol, corresponding to 29/.tM FC. The specific radioactivity ranged between 3.10 s and 1.7. l0 ~' dpm per p.mol HDL-FC. The radioactivity recovered in cells and associated with both free and esterified cholesterol, was then determined, as well as the specific radioactivity of the former. Results are the means+ S.E from seven experiments: four using total tlDL and three with HDL 3. Statistical comparisons between PLA2-HDL and (control) HDL: *P < 0,05.

. . . . Cell uptake of HDL free [~4C]chnlesterol recovered as:

t,,ml [~'~C]Chol. frdTCaC]Chol. • ester[fled [14ClChol. [percent of initial percent of initial specific (percent of initial HDL [~C]FC HDL [~aCIFC radioactivity HDL [t4C]FC (~i/rag cell protein)) (%/mg cell protein) (dpm/nmol) (%/mg cell protein))

HDL 11.5_+2.7 10.3+2.4 510+150 1.2±0.4 PLA,-HDL 14.7_+4.2 * 12.9+2.6 575-+ 178 1.9+0.5 *

217

~.4-

A 1.2"

g ._m " 1 .O - re

.~= ~ 0.6'

m E ~ 0.4 N

~ 0.2-

Cholic acitl f~ muricholic acid

Fig. 3. Effects of different HDL subfractions on the bile acid production. Hepatocytes ,.,.'ere incubated for 6 h in the presence o! HDL subfractions at a concentra'don of O. ! 30 -+ 0.005 mM cholesterol H DL {'-:-), HDL 2 (IlL or withoul HDL 112). Analysis of bile acids was performed by GLC after extraction and alkaline hydrofysi~. Results

are the means ( + S.E.) from five experiments_

st imulatory effect on the further bile acid synthesis by cultured hepatocytes. In spite o f variations between experiments, this tendency was observed with H D L 2 . H D L 3 or unfract ionated H D L (Table IV). Combin ing the data o!. tained with all H D L subfractions, the product ions of /3-muricholic and cholic acids were increased by 70% ( P < 0.05) and 80% ( P < 0.02), respectively, in the presence o f l ipolysed-HDL, compared to incubat ions with unmodif ied particles.

TABLE IV

Bile acid production hv cuhured hepatotTtes m the presem'e of tlDL sub#actions. Effects o] HDL phospholipo/vsi.~

Hepatocytes were incubated during 6 h with H DL subfractions (0.123 _.+ 0.006 mM cholesterol), treated or not by phospholipase A,. Quanti- fication of the two major bile acids secreted into the medium was performed by GLC. Statistical comparison between control HDL and PLA:-HDL: = P < 0.05, * * P < 0.02.

HDL PLase Bile acid synthesis subfractlons treatment (nmol/mg cell protein per 6 hi

cholic acid B-muricholic acid

Total H DL - 0.63 5:0.10 0.'~ _, O. 10 (n=5) + 1.98+0.20 ** 1.22_+~}A5 *

HDL z - 0.07_+0.31 0.71 +_(}.23 (n =4l + 1.08_+0.31 0.7t~_+ 1).11

HDL., - 0.55+0.13 0.20+_0.08 (n = 4l + 0.77+0.15 0.61 _+0.22

All subfractions - 0.71 _+0.12 I).48 _+0.11 combined + 1.27 +0.19 * * 02,1 _+0.12 * ~n=13)

Discu:¢:don

The present s tudy was designed to compare the effects of phospholipase-modif ied versus untreated H D L on the cholesterol metabolism of cultured hepatocytes.

It was observed that. whereas apo E free HDL.~ had limited effects on the cell sterol content and bile acid product ion. H D L : clearly induced a cholesterol accu- mulat ion irt hepatocytes and p romoted the secretion of cholic and B-muricholic acids. However. a pretreatment of HIDL with phospholipases induced in all cases the sam~ enhancemen t of cholesterol metabol ism in hepatocytes.

Human H D L is a heterogeneous group of l ipoprotein particles, which may differ, among other factors, by their apoprote in composit ion. The presence or absence of apo A-II . associated to apo A-I on the same lipopro- rein particle [34]. or else. the contr ibut ion of apo E to certain H D L subclasses 135] are possible determinants of H D L heterogeneity. As regards the binding of H D L to hepatic cells, both apo A-! and apo A-l l have been proposed as possible iigands [5,7,16,17]. On the other hand. apo E,conta in ing H D L may be taken up through the L D L - a p o B / E receptor [361, or by a receptor specific for apo E, such as the L R P (LDL-recep tor Related Proteinl, recently characterized in Hep G_,-cells and human liver membranes [37]. In human plasma, H D L are also distr ibuted between two main subfractions. according to their size. density and lipid content . In this study, only H D L , showed a positive effect on the cell cholesterol content down regulating its endogenous synthesis and promot ing , to some extent, bile aO.d production. Conversely. H D L 3 had little effects, except for a 12% decrease in the cellular free cholesterol. In con- nection with the present observations, we recently reported [141 that, at similar protein concentrat ions, H DL2 deliver 3 -4 - fo ld more free cholesterol and cholesteryl ether than HDL~. These differences might be partly explained by the presence o f apo E in our H D L : preparations. However, its contr ibut ion did not exceed !% of H D L , protein and all the incubat ions between cc..~ and H D L wcrc performed in absence of calcium, usually a necessary factor for the interactions through the LDL-recep to r [36]. On the other hand, H D L , con- tain twice more free and esterified cholesterol than HDL~ and the ratio o f free cholesterol to phospholipid is also doubled in the former [131. It is thus possible that, the high lipid content of H D L , may favour both a positive diffusion ~radient for free cholesterol between the H D L , surface and hepatic membranes [38] and a selective transfer of cholesterylesters, as evidenced in

several reports [2,1.!]. Hence, a l though H D L , and H D L a can transfer cholesterol to hepatocytes, the final sterol balance in cells depends on the nature o f the subfractions addressed. This is concordant with the general view that small-sized H D L (the predominant HDL3)

218

would be best acceptors for cellular cholesterol [39] subsequently returned to the liver by large, sterol-ricb HDL_,. If hepatic lipase is to play a role in this process [9]. it may be reminded that HDL~ [12] or, more generally, cholesterol-rich HDL I13] are preferential substrates when considering its phospholipase A-I activity.

Phospholipase-treatment of H D L subfractlons induced always the same modifications, as regards their influence on cell sterol metabolism: a 40% stimulation of cholesterol esterification and cholesteryl ester accumulation and an increased bile acid production. Consid- ering that the stimulations were of low amplitude (between +15~_ and +100%) and that some parameters showed large variations between experiments, some changes reached statistical significance, only when the data on all HDL subfractions were taken into account. This appeared legitimate, since the same tendencies were observed following phospholipase treatment of any subclass.

Compared to incubations performed with control- HDL, hepatocytes increase their content in esterified cholesterol by 35% on average. As well the formation of radiolabelled chotesterylesters from HDL-derived free [14Clcholesterol was stimulated in similar proportions. This close agreement suggests that the accumulation of esterified sterol in presence of modified H D L may well be explained by a higher influx of free cholesterol from lipolysed HDL, coupled to a stimulated esterification by the cell acyl cholesterol acyl transferase (ACAT) activity.

The uptake of HDL-radiolabelled cholesterylethers, non hydrolysable analogs and tracers of H D L esterified cholesterol, has been described with various cultured cells, including hepatocytes 12,14]. We recently reported that phospholipases enhance by 2-6-fold this process [14]. When such experiments were repeated using total HDL labelled with radioactive cholesteryl esters, the latter were also taken up by cultured hepatocytes, a process again stimulated by HDL-phospholipid hydrolysis. However, almost 80% of the transferred radioactive cholesteryl esters were found hydrolysed within the cells, after 6 or 24 h culture with either control or lipo;ysed HDL. As a consequence, the cellular radioactivity associated with non-hydrolysed cholesteryl esters represented only between 1.9 and 3.0 nmol derived from control and lipolysed HDL, respectively (X. Coltet, personal communication).

These data strengthen the conclusion that interaction with l ipolysed-HDL would, finally, provide the cells with more free cholesterol, part of it being secondarily reacylated through the action of ACAT. Another part of the transferred sterols would be chanelled through the synthesis of bile acids.

Cultured hepatocytes isolated from rat or chicken embryos constitute a reliable model to study the synthe-

sis of bile acids and its regulations [18.40]. Using labelled lipoproteins, it was reported that HDL free and esteri- fled cholesterol are potential substrates for conversion to bile acids [18]. However, the quantitative contribu- tion of either substrate is difficult to estimate, since the transformation concerns only a small proportion of the cellular cholesterol, and since the specific radioactivities of the free and esterified sterols w;thin cells are continu- ously changing with incubation t,mes [18]. Thus. in the present study, only mass determinations were performed. In agreement with others [18,41], we found cholic and /3-muricholic acids to be the r~ajor metabo- litcs in cultured rat hepatocytes and a stimutatory effect of apo E-containing H D L 2 on their secretion. Further- more, we observed that phospholipase-modified H D L was more efficient in sustaining this secretion of bile acids. Thus, after previous studies on cultured granulosa cells [42] hepatocytes constitute another type of cells, which enhance their production of steroid metabolites upon interaction with lipolysed HDL.

In conclusion, phospholipid hydrolysis in H D L pro- motes a net influx of H D L cholesterol to hepatocytes, as evidenced by the cellular accumulation of esterified cholesterol and the secretion of bile acids. Although this effect was observed with different H D L subfractions, we hypothesize that, under physiological conditions, it would mostly concern the large sterol-rich HDL. In- deed, these particles appear as efficient cholesterol donors to hepatocytes and are prone to phospholipid degradation by hepatic lipase.

Acknowledgements

Thanks are due to Mrs J.A. Puchaes and Mrs Y. Jonquiere for secretarial assistance.

References

10'Ma|ley. J.P.. Sottys. P.A. and Portman. O.N. (1981) J. Lipid Res. 22, 1214-1220.

2 Glass. C,, Pillman, R.C.. Civen, M. and Steinberg. D. (1985) J. Biol. Chem. 260. 744-750.

3 Bachorik, P.S.. Franklin, F.A., Virgil. D.G. and Kwlterovieh, P.O., Jr. (1985)Arteriosclerosis 5, 142-152.

4 Dashti. N., Wolfbaver, G. and Alaupovic, P. (19851 Bitv,:him. Biophys. Acta 833, i00-110.

5 Hoeg. J.M., Demosky. S.J.. Edge. S.B., Gregg, R.E., Osborne, P.C. and Brewer, H.B., Jr. (19851 Atherosclerosis 5, 228-237.

6 Monaco, L., Bond, H.M., Howell, K.E. and Corlese, R. (1987) EMBO J. 6. 3253-3260.

7 Graham. L.D. and ()ram, J.F. (19871 J. Biol. Chem. 262, 7439- 7442.

8 Doolinle. M.H., Wong, H., Davis, R.C. and Schotz, M.C. (1987)J. Lipid Res. 28. 1326-1334.

9 Jansen, H. and Hiilsmann, W.C. ~i985) Eur. J. Clin. Invest. 15. 95-99.

10 Jansen, H. 119851 Biochem. Biophys. Res. Commun. 131. 574-578. 11 Daggy, B.P. and Bensadoun, A. (1986) Biochim. Biophcs. Acta

877. 252-261.

t2 Shirai, W., Barnhart, R.L. and Jackson, R.L. 11981) Bicchem. Biophys. Res. Commun. 100. 591-599.

13 Simard. G.. Perret. B.. Durand. S.. Co!let X, Chap. H and Do.uste-Blazy, L, (19891 Biochim. Biof, hys. Acla 1001. 225-239,

14 Collet, X.. Perret, B.. Simard, G., Vieu, C. and Douste-Blazy, L. (1990) Biochim. Biophys. Acta 1043, 301-3IO,

15 Delamartre, J.G., Sarphie, T.G.. Archibold, R.C. and Hornick. C.A. (19901J. Lipid Res. 31. 191-202.

16 Fidge, N.H. and Nestel, PJ. (19851J. Biol, Chem. 260. 35711-3575. 17 Dvorin, E., Gorder. N.L.. Benson, D.M. and Gotto. A.. Jr.(19861 J.

Biol. Chem. 261, 15714-15718. 18 Herscovitz, H., Ronen. I., Bilu, S. and Tietz, A. (19861 Bitx'him

Biophys. Acta 8-78, 426-434. 19 Khan, B.. Wilcox. H.G. and Heimberg. M. ( 19891 Bit~hem. J. 259,

807-816. 20 Montet, J.C., Parquet, M., Sacquet, E., Montet. A.M.. lnfante, R.

and Amic, J. (19871 Biochim. Lgiophys. Acta qlg. 1-10, 21 Durringlon. P.N., Newton, R.S.. Weinstein, D.B. and Steinberg,

D. (1982) J. Clin. Invest. 70, 63-73. 22 Weisgraber, K,H. and Mahley, R.W. (19801 J, Lipid Res. 21.

316-325. 23 Marcel. Y.L.. Weech. P.K., Nguyen, T.D., Milne, R.W. and Mc

Conathy, J. (1984) Eur. J. Biochem. 143. 467-476. 24 Petter, B.P., Chollet. F. Durand. S. Simard, G. Chap, H. and

Douste-Blazy, L (19871 Eur. J. Bit-chem. 162, 279-286. 25 Azhar. S. Chen. YD.I. and Reaven. GM. (19841 Biochemistry 23,

4533-4538. 26 De Mark, B.R., Everson. G.T., Klein. P.D., Showatter, R.B. and

Kern, F., Jr. (1982) J. Lipid Res. 23, 204-210. 27 Bligh. E.G. and Dyer. W.J. (19591 Can. J. Bitxhent. Physit~l. 37,

911-917.

219

28 Youssef, I.M., Ho, J. and Jeejeebhoy, K.N. (1978} Can. J. Bio- them, 56. 780-783.

29 Roschlau. P., Bernel. E. and Gruher. N. 11974} Z_ Klin_ Chem. Kiln, Biochem. 12, 403-417.

30 Bi.~ttcher, C.F.J.. Van Gent. C.M. and Pries, C. (19611 Anal. Chim. ~cta 24. 203-204.

3t Skipskt. V.P.. Pelerson. F.F. ;,nd Barclay. N. (19641 Bit~hem. J. 90. 374-379.

32 Lowry., O.H., Rosebrough, N.J.. Farr, A,L. and Randall, R.J. t |ggl} J. Biol. Chem. 193, 265-275.

33 Cur:3', MD_. Gustafson, A., Alaupovic, P. and Mc Conath'*', W.J, (19771 Clin. Chem. 24, 280-286.

34 Cheung, M.C. and Albers. J.J. (19841 J. Biol. Chem. 259, 12201- 12209.

35 Funke. H., Bo~les, J,, Neisgraber, KH., Ludwig. E_H., Hui, D.Y_ and Mahley, R.N. (1984) Atherosclerosis 5, 452-461.

36 Brown, M.S., Kovanen, P.T. and Goldstein, J.L. (19811 Science 212, 628-635.

37 Beistegel, U., Weber, W., lhrke, G,, Herz, J, and Stanley, K.K. (19891 Nature 341, 162-164.

38 Rolhblat, G,H., Argohast, L.Y. and Ray, E.K. (19781 J. Lipid. Res, 19, 350-358_

39 Rothblat, G.H. and Phillips. M.C. (19821 J. Biol, Chem., 257, 4775-4782.

40 Davis, R.A., Hyde, P.M., Kuan, J.C.W.. Malone-Me Neal, M. and Archambealt-Schewnayder, J. 0983) J. Biol. Chem. 258, 3661- 3667.

41 Ford, R.P., EIt,lham, K M. Sucklln~, K_ l::. :~9d Bo)'d, G.S. 1,19851 FEBS Lett. 179, 177-180.

42 Parinaud, J . Petter, B.. Ribbes, H.. Chap, H.. Pomonnier, G. and Douste-Bhuy. L. (19871J. Clin. Endt~rinol. Metab. 64,1147-1152.

behaviour of phospholipase modified-hdl towards cultured hepatocytes. ii. increased cell cholesterol...

Documents