Lithocholic acid-cholesterol interactions in rat liver plasma membrane fractions

Biochimica

et Biophysics

Ada,

345

196 (1984) 345-353

Elsevier

BBA 51782

LITHOCHOLIC ACID-CHOLESTEROL MEMBRANE FRACTIONS I.M. YOUSEF,

M. LEWITTES,

B. TUCHWEBER,

Departments of Pediatrics and Nutrition, Hospital, Montreal, Quebec (Canada) (Received February (Revised manuscript

INTERACTIONS

IN RAT LIVER PLASMA

C.C. ROY and A. WEBER

University

of Montreal,

and the Pediatric

Research

Center,

SainteJustine

lOth, 1984) received June 4th, 1984)

Key words: Lithocholic acid; Cholesterol;

Cholestasis; Lipid-lipid

interaction; (Rat liver cell)

This study was designed to elucidate the steps involved in the incorporation of lithocholic acid and the increase in cholesterol in liver plasma membranes after lithocholic acid injection. In vitro, cholesterol incorporation or binding to liver plasma membrane fractions enriched in bile canalicular structures occurred only when cholesterol was added simultaneously with lithocholic acid. The addition of cholic acid did not prevent the incorporation or binding of lithocholic acid and of cholesterol. However, when cholic acid was incubated with membranes already containing lithocholic acid and cholesterol, the ratio of cholesterol to lithocholic acid increased from 2 to more than 3 via a reduction of lithocholic acid. The binding of lithocholic acid and cholesterol to membranes rose 5-fold in the presence of cytosolic proteins. By electron microscopy the canalicular membrane structures with a high cholesterol content exhibited few microvilli, and their lumen appeared to have collapsed. These data suggest that simultaneous interaction of lithocholic acid and cholesterol, and not prior incorporation or binding of lithocholic acid to the membrane, may be a prerequisite to cholesterol accumulation in the membrane.

Introduction Several investigators [l-15] have established that lithocholic acid injection induces intrahepatic cholestasis in a variety of animal species. Recent studies [7,10,12-151 aimed at understanding the pathogenesis of this type of cholestasis have been aided by the isolation of liver cell plasma membrane fractions enriched in bile canalicular structures (canalicular fractions) [16-181, the most obvious .morphologic site of lithocholate cholestasis [5,8-12,191. Lithocholic acid injection has been shown to be associated with its incorporation or binding to the canalicular fraction and with an increase in membrane cholesterol [13-151. Changes in membrane enzymes have been found to be secondary to membrane alterations and not in0005-2760/84/$03.00

0 1984 Elsevier Science Publishers

B.V.

volved in the mechanism of cholestasis [14]. Consequently, in the model of lithocholate-induced cholestasis, it was proposed that lithocholic acid is bound or incorporated in the canalicular fraction, causing a disruption of membrane organization, and an increase of cholesterol in the canalicular fraction [14]. Thus, lithocholic acid binding or incorporation in the canalicular fraction appeared to be a prerequisite for cholestasis. This suggestion was supported by the fact that cholic acid, infused with lithocholic acid, prevented the development of cholestasis. The protective effect was thought to be due to the formation of a complex of cholic acid with lithocholic acid, which prevented binding or incorporation of the latter in the canalicular fraction as well as the subsequent accumulation of cholesterol in the membrane [7,11,15].

346

Although these data [I3-15] supported the proposed mechanism of cholestasis, it should be noted that they were obtained from analysis of the canalicular fraction at various time intervals after infusion of the cholestatic bile acid. The objective of the present study was to examine in vitro the effect of lithocholic acid in cholesterol incorporation in liver cell plasma membranes. Materials and Methods A. Preparation of liver cell plasma membrane fractions Liver cell plasma membrane fractions consisting of contiguous areas of hepatocytic plasma membrane with attached canalicular structures or containing vesicles and sheets of membrane from the sinusoidal cell surfaces were isolated from 200-250 g male Wistar rats (High Oak Ranch, Montreal, Quebec), as described elsewhere 117,181. The purity of the fractions was determined by measuring glucose-6-phosphatase, S-nucleotidase, (Na+-K+)-adenosinetriphosphatase, Mg2+-adenosinetriphosphatase, leucine aminopeptidase, as outlined earlier [18], and adenylate cyclase, cytochrome oxidase and galactosyltransferase, as described by Beaufay et al. [20], and acid phosphatase, as suggested by Trouet [21]. The membranes, suspended in 0.075 M NaC1/0.075 M KC1 buffer (pH 7.4) at a concentration of I mg protein/ml, were always used on the day of preparation. For some studies, hepatic cytosol was added to the incubation mixtures. It was prepared by centrifugation of the post-mitochondrial supernatant at 124000 X g for 2 h 1221. B. Preparation of bile acid and lipid suspensions Several suspensions containing phosphatidylcholine and cholesterol, phosphatidylcholine and either lithocholic acid or cholic acid, phosphatidylcholine with cholesterol and either lithocholic acid or cholic acid, phosphatidylcholine with cholesterol and both lithocholic acid and cholic acid were prepared as follows. (I) Cholesterol. Dipalmitoylphosphatidylcholine and cholesterol (both from Sigma Chemical Company, St. Louis, MO) were purified by thin-layer chromatography and mixed with [1,2(n)-

(New England Nuclear, Mon”!real; spec. act. 50 Ci/ mmol) and L-a-di]PC]paim”itoylphosphatidylchoiine (New England Nuclear; spec. act. 80 mCi/mmol) at a final 2 : 1 ratic of cholesterol to phospholipid and at a concentration of 1000 nmol of cholesterol per ml and spec. act. of I . IO6 + 5% dpm/m.mol of cholesterol or phsspholipid. The lipids were dissolved in chloroform. The solvent was then evaporated under nitrogen after which the lipids were suspended in 0.075 M NaC1/0.075 M KC1 and sonicated for 10 rntin m 15-m% flasks kept in ice. Following sonication, the suspension was centrifuged at 21000 x g for 6e? min for possible sedimentation of undispersed after centrifugation, lipids. However, the cholesterol and phospholipids were measured in the suspension, and no sedimentation was found. (2) Lithocholic acid or cholic acid Li&ocho%ic acid or cholic acid suspensions were prepared in dipalmitoylphosphatidylcholine by dissolving equal molar amounts of lithocholic acid sodium salt (Calbiochem, San Diego, CA; 99% pure as checked by gas-liquid chromatography) or cholic acid sodium salt (Calbiochem; 94% pure) with phosphatidylcholine in methanol. [ccs&~~~lt4C]lithochelic acid (Amersham, Arlington Heights, IL; spec. act. 55 mCi/mmol) or [ca.&xyl-“C]cholic acid (New England Nuclear: spec. act. 50 mCi/mmol) or H-ci-di-[9,lO’H]palmj&oylphosphatidylcholine (Applied Science Laboratories Inc., State College, PA; spec. act. 10 mCi/ mmol) was also added. The mixtures were dried under nitrogen. Alicpots of 0.075 M NaCl and 0.075 M KCB were added and sonicated for 10 min. The suspensions were then centrifuged at 21000 x g for 60 ,min, after which no sedimentation was detectable. The final concentration of bile acids was 500 nmoI/ml and the spec, act. was 1 . 106 _ 5% dpm/mmol for each of the bile acids and for phosphatidylcholine. (3) Cholesterol with either lithocholic acid or cholic acid. Lithocholic acid sodium salt or cholic acid sodium salt was dissolved in methanol after which cholesterol and phosphatidylcholine were added. [‘4C]Lithocholic acid or [r4C]cholic acid and [3H]cholesterol were also added. The bile acids and lipids were suspended, as described previously, to give final concentrations of 500, 1000 and 500 nmoI/ml of lithocholic acid or cholic acid? 3 H]cholesterol

347

cholesterol and phosphatidylcholine, respectively. The spec. act. of the bile acids and cholesterol was 1. lo6 + 5% dpm/mmol. (4) Cholesterol with lithocholic acid and cholic acid. A suspension of lithocholic acid, cholic acid and cholesterol was prepared with phosphatidylcholine, as described above. In this suspension, only [i4C]lithocholic acid and [ 3H]cholesterol were used. The final concentrations were 500, 500, 1000 and 500 nmol/ml of lithocholic acid, cholic acid, cholesterol and phosphatidylcholine, respectively. The spec. act. of lithocholic acid and cholesterol was 1. lo6 + 5% dpm/mmol. C. Experimental conditions for incubation of plasma membranes with lipids (I) Effect of time of incubation. To determine the optimal time of incubation, aliquots of membrane suspensions containing 1 mg protein were mixed with lithocholic acid, cholic acid or cholesterol suspensions in a volume of 4 ml and incubated at 37°C for 5, 15 or 30 min. After incubation, the samples were centrifuged at 3000 x g for 30 min at 4°C and aliquots of the supernatants were counted in duplicate. The pellet was washed twice with 0.075 M NaC1/0.075 M KC1 buffer, then suspended in 1 ml of 0.075 M NaC1/0.075 M KCl, after which 0.5 ml was solubilized in 2 M NaOH solution and aliquots were counted in duplicate. The binding or incorporation of the bile acids or lipids was calculated from the radioactivity. In addition, cholesterol and phospholipids were measured in the rest of the membrane pellet. Control incubations were performed with suspensions of the membrane previously boiled for 15 min, to determine nonspecific binding or incorporation, which was subtracted from the total binding to obtain the true binding or incorporation value. Both incubation procedures produced the same results for binding or incorporation, calculated from radioactivity and chemical analysis of the membranes. (2) Effect of membrane protein concentration. Aliquots of membrane suspensions containing 0.2, 0.4, 0.6, 1 or 1.5 mg protein were adjusted to 3 ml with 0.075 M NaC1/0.075 M KCI. Lithocholic acid, cholic acid or cholesterol suspensions of 1 ml were added, and binding or incorporation of the various lipids was estimated as outlined earlier.

(3) Effect of cytosol. l-ml membrane suspensions (containing 1 mg protein) were diluted with various amounts of cytosol containing 0.25, 0.50, 1.0 or 2.0 mg protein. The volume was completed to 3 ml with NaCl/KCl buffer. Lithocholic acid, cholic acid or cholesterol suspensions of 1 ml were added, and incubation was prolonged for 5 min, after which the binding or incorporation was calculated as before. Control studies were performed by boiling the cytosol for 15 n-tin prior to its addition to the incubation mixture. D. Interaction of lithocholic acid, cholic acid and cholesterol in plasma membranes (I) Specificity of lithocholic acid binding to canalicular fractions. The specificity of lithocholic acid binding to the canalicular fraction was tested by incubating 1 mg protein of this membrane fraction with 50, 100, 200, 400 or 500 nmol of lithocholic acid sodium salt suspended in 500 nmol of phosphatidylcholine in the presence of 1 mg cytosolic protein for 5 min. The total incubation volume was 4 ml. The incubation mixture was then centrifuged at 3000 X g for 30 ruin and the resulting membranes were incubated in the presence of 500 nmol of [‘4C]lithocholic acid. After 5 min of incubation, the membranes were isolated, and the [‘4C]lithocholic acid bound to the membrane was determined as before. (2) Effect of lithocholic acid binding on cholesterol incorporation in canalicular fractions. Incubation of 1 mg membrane protein with lithocholic acid for 5 min was followed by removal of the supernatant and re-incubation of the membrane pellet with cholesterol for 5 more minutes. In other incubations, the membrane suspension was incubated with the lithocholic acid and cholesterol suspension in the presence or absence of 1 mg cytosolic protein. Binding of the bile acid and cholesterol was determined as above. (3) Effect of cholic acid. Two series of incubations were performed, the first with a mixture of cholic acid, lithocholic acid and cholesterol. In the second, the membranes were incubated initially with lithocholic acid and cholesterol for 5 min. After removal of the supernatant, a cholic acid suspension was added, and the incubation was continued for 5 more minutes. Portions of the pellets resulting from incubations following which

348

lithocholic acid and/or cholesterol were detected in the membrane were examined by electron microscopy. Results The liver cell plasma membrane fractions used in the present study were similar to those previously prepared in our laboratory [17,18]. Of the two types of fractions obtained, one appeared morphologically to consist of structures similar to the canalicular pole of the hepatocyte, as seen in Biver tissue in situ (canaliculus with microviili and portions of the lateral membranes with tight junctions). This fraction will be referred to as the canalicular fraction. The other fraction mainly contained sheets and vesicles of membranes with no bile canalicular or tight junction structures. Accordingly, it will be referred to as the sinusoidal fraction. The enrichment of (Na’-I[(.+)-ATPase activity noted in the canalicular fraction was in agreement with the recent study of Schenk and Eeffert [23], who located the enzyme, by monoclonal antibodies to rat (Na+-K+)-ATPase, in the canahcular side of the liver cell. The relatively high specific activity of Mg’+-ATPase, 5’-nucleotidase and leucine aminopeptidase in the canalicular fraction further supported the enrichment of the

canahcular fraction in the intact canalicuk ccmplexes. Moreover, the sinusoidal fraction was di!ferent from the canahcular fraction with regard :o the enrichment of adenylate cyclase activity. Va%ues for these enzymes and for lipid cornnosition have already been reported in an e&her paper i24]. The two liver cell plasma membrane fractions revealed a relatively low specific activiggi for ghcose6phosphatase (o.o2-o.Q5> and c)~Eochrome oxidase (0.03-0.06) with no detectable activities o.f galactosyltransferase or acid phosphatase, indiceting that they were not contaminated with rn~crosomes, mitochondria, Go:gi or cytoso2. Binding or incorporation of lii2hocMic a&cd, ch& acid, cho!esleroi and phospholipids to kxr celf plasma membrane Table I shows that lithocholic acid was bound in significantly higher amounts than chohc acid, cholesterol or phospholipids to the canahcular fraction in comparison to the sinusoidal fraction. The binding reached a maximum at 5 min and was hnear with protein concentrations (Table HI). The presence of cylosohc protein (up to 1 mgj in the incubation medium increased lithochohc acid binding 5-foId only in the canahcular fraction and did not affect the incorporation of other lipids or bile acids (Table III). When cytosoi was boiled for

TABLE I INFLUENCE

OF TIME ON LIPID BINDING TO LIVER CELL PLASMA MEMBRANE

FRACTIONS

IN ViTWO

The values are in nmol (meani SD. from eight experiments, for each of which the membranes were obtained from one hver). The incubation medium contained 1 mg of protein, 500 nmoi of lithochoiic acid or cholic acid or 1000 nmol of cholesterol. The lipids (lithochoiic acid, cholic acid and cholesterol) were suspended in 500 nmol of phospholipids. The nonspecific bindicg with both fractions of the liver cell plasma membranes was equal, and the values (nmol/mg protein) were as follows: lithochok acid 4.8 k 0.8 and 5.2 i 1.1; cholic acid 3.9 -I_0.8 and 4.7 i 0.9; cholesterol 3.4i 0.7 and 3.8 i 0.8; and phospholipids 3.6 F 1 .I and 2.9 t 0.9 for the canalicular fraction and sinusoidal fraction, respectively. Time of incubation (min)

Canalicular 5 I5 30 Sinusoidal 5 15 30

Lipids Lithocholic acid

Cholic acid

Cholesterol

Phospholipids

31.7i9.3 34.5 + 8.1 38.5 * 8.2

2.4* 0.2 3.8 + 0.4 2.4kO.2

2.9 f 0.4 4.3 & 0.4 2.7 zt 0.3

3.8 IO.5 2.8 + 0.4 4.1 kO.4

5.0*0.4 6.0 + 0.5 5.8 i 0.4

2.3 4 0.2 1.210.2 3.7 kO.4

2.3 50.2 4.4 & 0.2 3.0 * 0.3

3.850.2 2.4kO.3 2.1 * 0.2

fraction

fraction

349 TABLE

II

IN VITRO EFFECT OF MEMBRANE MEMBRANE FRACTIONS The values determined.

are in nmol

(mean& S.D. from

Membrane concentration (ng protein) Canalicular 200 400 600 1000

The lipid

BINDING

concentrations

TO LIVER

are the same

CELL

as in Table

Cholic acid

Cholesterol

Phospholipids

6.4+ 2.0 12.7+ 3.8 19.1* 5.8 31.7+ 9.3 41.2kl1.9

1.450.4 1.9kO.2 3.1 -f 0.3 2.4kO.2 6.OkO.7

2.1 IO.2 2.4 & 0.3 2.1 kO.3 2.9 !c 0.4 3.0t0.4

n.d. nd. 2.1* 0.2 3.8 + 0.5 4.lkO.4

1.OJrO.l 1.2kO.l

1.8iO.l 2.3 + 0.2

nd.

2.3kO.3 3.4kO.3 4.1* 0.4

2.4+0.3 4.1+ 0.4 5.2+0.6

PLASMA

I. n.d.,

not

fraction 2.0* 3.3+

0.1 0.3

4.7* 0.3 6.4+ 0.6 8.95 0.8

min

before

mixture,

it was added

its effect

the membrane

was diminished,

for six membrane the canalicular

acid binding

lithocholate

transferred

and a value (X f

lar

capacity

reduced

the

[Tables Table

with unbinding

fraction,

acid.

As phospholipids

the

fusion

was

ruled

to the membrane

with

the

were not

to the canalicu-

membrane

out

that

was due bilayer

I-III]. IV shows

that cholesterol

rated in the canalicular

of

amounts

possibility

acid binding

to liposome

to

n.d. 2.4i0.4 3.150.2 4.3 + 0.5

in significant

lithocholic

As shown in Fig. 1,

had a limited

this bile acid since pre-incubation

labeled

to

protein was obtained

incubations.

fraction

[i4C]lithocholic

to the incubation

on lithocholic

S.D.) of 36.7 J~I8.9 nmol/mg

TABLE

experiments).

ON LIPIDS

Lithocholic acid

600 1000 1500

bind

eight

CONCENTRATIONS

fraction

1500 Sinusoidal 200 400

15

PROTEIN

fraction

was incorpo-

only when it was

III

IN VITRO

EFFECT

OF CYTOSOL

ON LIPIDS

BINDING

TO LIVER

CELL PLASMA

MEMBRANE

FRACTIONS

The values are in nmol (mean + S.D. of eight experiments). The incubation medium was similar to that described in Table I with the exception of the presence of cytosolic protein. The incubation time was 5 min. Increasing the incubation time to 15 or 30 min did not result in a significant change in the binding of the various lipids. When cytosol was boiled for 15 min and allowed to reach room temperature before it was added to the incubation mixture at a concentration of 1 mg of protein, lithocholic acid binding was 36.7+8.9 nmol/mg of membrane protein (meanS.E. from eight experiments). The addition of cytosol did not affect nonspecific binding, and the values were not different from those obtained in Table I. Amount of cytosol (mg protein)

Lithocholic acid

Cholic acid

Cholesterol

Phospholipids

Canalicular 0.25 0.50 1.00 2.00

69.05 8.3 98.4+ 13.7 155.2 + 28.6 165.1 k25.3

3.8 + 0.4 3.8 zk0.5 4.3 + 0.5 4.8 + 0.4

2.4 k 0.2 2.3 i 0.5 4.8kO.3 4.2 CO.5

3.1 i 0.3 2.9 + 0.2 2.8 + 0.3 3.2+0.3

3.9kO.4

2.7*0.3 2.3 f 0.2 2.8 * 0.2

3.7 + 0.4 4.3 +0.4 3.8 5 0.3 3.7 &-0.4

Sinusoidal 0.25 0.50 1 .oo 2.00

fraction

fraction

4.9+ 4.7+ 4.9+ 5.32

0.6 0.5 0.6 0.4

4.7 +0.5 2.8 + 0.2 3.9+0.3

3.1+0.3

added with Bitbocholic ac!d as a csmplex. 7he ratio of cholesterol to litbocholic acid bound to :he canalicurar fraction averaged 2, and was not aEfected by cytosolic protein. Addition. of choHic acid at the same time as lithochoiic acid and chokslerci did not prevent their binding or incorpora~ioc. WoweverTe when cholic acid was incubated w;k membranes already containing lithochok acid and cholesterol, the radio of &olesteroB to lithcchoiic acid increased from 2.05 to 3.68 due kc a rechdsra of iithocholic acid binding (Table IV>. When cholesterol was enhanced in the membranes, tke ___I_ 20 10 5 50

LCA

TABLE

400

200

!OO

nmoledmg

-----.-._

Fig. 1. Specificity of lithocholic acid (LCA) binding to the canalicular fraction. The membranes were preincubated iwith various concentrations of uniabeled lithocho!ic acid and ihen incubated with a suspension containing 5OC nmol [ “Cll~t~ochohc acid. The values (means+ SD.) represent tOtei ii4CjIirhocholic acid binding - [ I4Cjlithocholic acid nonspecific hinding; n. 4 samples.

500

protein

1V

IN VITRO PRESENCE

BINDING OF LITHOCHOLIC ACID OF VARIOUS LIPID MIXTURES

AND

CHOLESTEROL

IN THE

CANALICULAR

FRACTION

IN THE

The values are in nmol (mean+S.D. from eight experiments). The amount of cytosol added to the incubation mixture was I mg protein. The addition of cholic acid before or with choiesterol did not result in binding of cholic acid or cholesterol to the membrane. Incubated

with

Incubated

without

cyt0s0i

cytosol

Cholesterol added after initial binding of lithocholate Lithocholate Clnolesterol Cholesterol/lithocholate

159.9 & 32.8 3.1 i 0.9 0.02+ 0.003

32.6 i 9.8 2.2 i 0.6 O.Q7 + 0.009

Cholesterol and lithocholate added simultaneously Lithocholate Cholesterol Cholesterol/lithochoIate

162.7 i 25.9 328.9 i 43.3 2.04 ?r 0.230

33.2 911.9 67.9 k15.1 2.05 I 0.247

Cholesterol, cholic acid and lithocholate added simultaneousiy Lithocholate Cholesterol Cholesterol/lithocholate

155.5 * 30.3 318.5 k46.9 2.05 & 0.238

31.3 + 8.1 65.2 kl7.5 2.09 + 0.236

Cholic acid added after initial binding of cholesterol and lithocholate Lithocholate Cholesterol Cholesterol/lithocholate

88.2 k22.9 322.6 i 52.7 3.68 f 0.384

17.7 i 7.3 61.3 iP8.0 3.49 f 0.367

351

Fig. 2. Liver cell plasma membrane fraction enriched in bile canaliculi incubated with lithocholic acid (A) and lithocholic acid + cholesterol (B). In A, the bile canalicular structure is normal and does not differ from non-incubated membranes. However, in 8, the canalicular structures show less microvilli and are partially collapsed. x 10.933

microvilli in the canalicular fraction were virtually absent and the canalicular lumens were partially collapsed. However, in other incubations when cholesterol content was not increased, the membranes were not altered morphologically (Fig. 2). Discussion The aim of the present study was to elucidate the steps involved in the accumulation of cholesterol in the canalicular fraction following lithocholic acid injection in vivo [l-4]. The experiments were performed in light of the fact that, in vivo, the effect of lithocholic acid injection is rapid: a reduction in bile flow (cholestasis) occurs within 45-60 min [l-4] and membrane changes are noted within 15 min [3]. Therefore, the membrane incubations were of a maximum duration of 30 min. The results listed in Table I depict maximal lithocholic acid binding at 5 min of incubation with a limited capacity of the membrane to bind this bile acid (Fig. 2). Binding was much higher in canalicular fraction than in the sinusoidal fraction (Tables I-III), indicating either the presence of a specific binding protein for lithocholate in the canalicular membrane or a specific arrangement in the lipid bilayer which allows the incorporation of this molecule. It has been shown that in mem-

branes isolated from control rats, the protein polypeptides in the canalicular fraction differ from those in the sinusoidal fraction [18]. Moreover, similar polypeptides have been obtained from canalicular fractions of lithocholate-treated animals and controls [3]. However, a freeze-fracture replica study of the liver after lithocholic acid injection has demonstrated that, in plasma membranes, these proteins seem to translocate to the side of the membrane adjacent to the tight junctions [7]. Therefore, it is reasonable to suggest that lithocholic acid may be incorporated in canalicular fraction lipids. The influence of cytosol on the elevated binding of lithocholic acid cannot be readily explained. Since the effect was linear with increasing concentrations of cytosolic proteins in the incubation medium, certain cytosolic factors may be involved in the binding or incorporation of lithocholic acid and cholesterol in the canalicular fraction. Indeed, the cytosol has been found to contain not only a lithocholic acid-binding protein [25] but also a cholesterol-binding protein [26]. Moreover, it has been reported that the presence of albumin increases cholesterol incorporation in red blood cell membranes [27]. Therefore, such binding proteins may be required for the lithocholic acid binding process in the canalicular fraction. Further support for this view has been provided by the ab-

352 sence

of the cytosoi effect when it was boiled before use. When cholic acid was added to membranes containing lithocholic acid and cholesterol, only bthocholic acid was extracted from the membranes, at least in the first 5 min. This could be interpreted as evidence that Sithocholic acid is bound to the surface of the membranes. Kakis and Yousef (131 suggested that the hydrophobic part of the molecule is inserted in the bi!ayer while the hydroxyl and carboxyl groups are attracted to the cell cytoplasm. The addition of chohc acid at the same time as lithochohc acid and cholesterol does not prevent lithocholic acid binding, which indicates that cholic acid protection in vivo may result from the formation of a lithochohc acid-cholic acid complex in the membrane itself. Choles!erol is not incorporated in the membrane when it is added alone or after hthocholic acid binding. This signifies that lithochoiate binding to the canalicular fraction is not a prerequisite for the entry of cholesterol in the bilayer, as postulated by Kakis and Yousef [13]. However, when cholesterol is added with hthocholic acid in a complex, it enters the membrane immediately at a constant cholesterol/lithocholic acid ratio of 2, suggesting an association of lithocholic acid and cholesterol prior to their incorporation in the membrane. Since the binding of lithochohc acid and cholesterol is increased dramatically in the presence of cytosol, there may be a common mechanism for the incorporation of both molecules In the membrane. The studies undertaken with cholic acid following the incorporation of the lithocholic acidcholesterol complex in the membrane and resulting mainly in the removal of lithocholic acid indicate that cholesterol is inserted in the Lipid bilayer of the membrane. The constant ratio of cholesterol to hthochohc acid in the membrane may be explained as follows: the lithocholic acid molecule measures 30-35 A in length. Assuming that it lies flat on the bilayer surface of the membrane, such an arrangement would create a gap in the membrane bilayer sufficient to allow the incorporation of two molecules of cholesterol each measuring 15 A diameter. The electron microscopic observation that incubation with lithocholic acid and its binding did not

induce any morpho%ogic changes in the me,m$raae and that the presence of cholesterol caused the isappearance of micrcvilli from most of the membrane preparations (Fig. 2B) indicates ihat the increase in cholesterol ehcits one of the early rmphologic alterations seen in tithochohc acid-in-duced cholestasis in vim It should be pointed out, however, that the disappearance of microvilli per se may not be a sign of intrahepatic chclestasis [2S]. The lamellar transformation of the inembrane, noted during lln viva experiments on hthocho!ic acid-induced cholestasis [5,7], was not observed in these studies, wbtich is not surprismg in view of the recent report that lamelIar transformation is not associated with an increase in Ihe cholesterol content of -membranes $29]. Hn conclusion, tne present investigation suggests an interaction between hthochohc acid and cholesterol proir to their accumulation an the canahcular fraction and this interaction may represent an important step in the development oi lithocholic acid-induced cholestssis in vivo.

This work was supported by the Medicai Research Council of Canada. The authors gratefui!y acknowledge the assistance of Ovid M. Da S&a of Better Communications Reg’d, Montreal, and Mrs. Diane Leblanc and Sylvie Tasse for typing the manuscript. References 1 Javitt, N.B. (1966) Nature 210, 1262-1263 2 Javitt, N.B. and Emerman, S. (1968) J. Ciin. Invest. 47, 1002-1014 3 Fisher, MM., Magnusson, R. and Miyai, K. (1971) Lab. Invest. 21, 88-91 4 King, J.E. and Schoenfield, i. (1971) J. Clin. Invest. 50, 2305-2312 5 Miyai, K., Price, V.M. and Fisher, MM. (1971) Lab. Bnvesr. 24, 292-302 6 Priestly, B.G., Cote, MC. and Plaa, G.L. (1971) Can. J. Physioi. Pharmacol. 49, 1078-1091 69: 7 Layden, T. and Boyer, J.L. (1975) Gastroenteroiogy A40,‘846 8 Layden, T.J., Schwarz, J. and Bayer, J.L. (1975) Gastroenterology 69, 724-738 AL. (1975) Lab. 9 Miyai, K., Mayr, W.W. and Rkhardson, Invest. 32, 527-535 10 Layden, T.J. and Boyer, J.E. (1977) Gastroenterology 73. 12.0-128

353

11 Miyai, K., Richardson, A., Mayr, W.W., Warner, M. and Javitt, N.B. (1977) Lab. Invest. 36, 249-258 12 Bonvicini, F., Gamier, A., Gandiol, D. and Borel, G.A. (1978) Lab. Invest. 38, 487-495 13 Kakis, G. and Yousef, I.M. (1978) Gastroenterology 75, 595-607 14 Kakis, G., Phillips, M.J. and Yousef, I.M. (1980) Lab. Invest. 43, 73-81 15 Kakis, G. and Yousef, I.M. (1980) Gastroenterology 78, 1402-1411 16 Song, C., Rubin, W., Rifkind, A.B. and Kappas, A. (1964) J. Cell. Biol. 4, 124-132 17 Fisher, M.M., Bloxam, D.L., Oda, M., Phillips, M.J. and Yousef, I.M. (1975) Proc. Sot. Exp. Biol. Med. 150,177-184 18 Yousef, I.M. and Murray, R.K. (1978) Can. J. Biochem. 56, 713-721 19 Boyer, J.L., Layden, T.J. and Hruban, Z. (1977) in Membrane Alteration as Basis of Liver Injury (Popper, H., Bianchi, L. and Reutter, W., eds.), pp. 353-369, MTP Press, Lancaster, U.K.

20 Beaufay, H., Amar-Costesec, A. and Feytmans, G. (1974) J. Cell. Biol. 61, 188-200 21 Trouet, A. (1974) Methods Enzymol. 31, 323-329 22 Fleischer, S. and Kervina, M. (1974) Methods Enzymol. 31, 6-40 23 Schenk, D.B. and Leffert, H.L. (1983) Proc. Natl. Acad. Sci. USA 80, 5281-5285 24 Yousef, I.M. and Tuchweber, B. (1984) Biochim. Biophys. Acta 796, 336-344 25 Strange, R.C., Nimmo, I.A. and Percy-Robb, I.W. (1977) in Bile Acid Metabolism in Health and Disease (Paumgartner, G. and Stiehl, A., eds.), pp. 125-132, MTP Press, Lancaster, U.K. 26 Erickson, S., Meyer, D.J. and Gould, R.G. (1978) J. Biol. Chem. 253,1817-1826 27 Hui, S.W., Stewart, C.M. and Carpentier, M.P. (1980) J. Cell. Biol. 85, 283-291 28 Niemchausky, B., Layden, T.J. and Boyer, J.L. (1977) Lab. Invest. 36, 259-267 29 Jung, W., Gebhardt, R. and Robenek, H. (1982) Eur. J. Cell. Biol. 29, 77-82