Comparative effect of ursodeoxycholic acid and calcium antagonists on the binding, uptake and degradation of LDL in isolated hamster hepatocytes

Comparative effect of ursodeoxycholic acid and calcium antagonists on the binding, uptake and degradation of LDL in isolated hamster hepatocytes

Biochi~ic~a ELSEVIER et Biophysica A~ta Biochimica et BiophysicaActa 1301 (1996) 230-236 Comparative effect of ursodeoxycholic acid and calcium anta...

741KB Sizes 0 Downloads 5 Views

Biochi~ic~a ELSEVIER

et Biophysica A~ta Biochimica et BiophysicaActa 1301 (1996) 230-236

Comparative effect of ursodeoxycholic acid and calcium antagonists on the binding, uptake and degradation of LDL in isolated hamster hepatocytes Bernard Bouscarel *, Susan Ceryak, Hans Fromm Division of Gastroenterology and Nutrition, Department of Medicine, The George Washington University Medical Center, 2300 l Street, N. W., #523 Ross Hall, Washington, DC 20037, USA

Received 18 December 1995; accepted 16 February 1996

Abstract

We have shown that ursodeoxycholic acid (UDCA) stimulates low density lipoprotein (LDL) metabolism (Biochem. J. 280 (1991) 589), as well as calcium mobilization (Am. J. Physiol. 264 (1993) G243) in isolated hepatocytes, Therefore, the effect of UDCA and that of different calcium antagonists on hepatic LDL metabolism was compared. Isolated hamster hepatocytes were incubated at 37°C for 60 rain in the presence of 12SI-labelled hamster LDL, increasing concentrations (25-100 /xM) of verapamil, nifedipine, and diltiazem, respectively, and with or without 700 /zM ursodeoxycholic acid (UDCA). At concentrations up to 100 /zM, neither verapamil nor nifedipine significantly affected cell associated LDL, but both agents decreased LDL degradation in a dose-dependent manner, with almost total inhibition with 100/xM of either agent. In contrast, 25 /xM diltiazem stimulated LDL binding and uptake, with a maximum increase of 15-20% of control, while 50 and 100 /xM diltiazem stimulated LDL degradation by 50 and 100%, respectively. UDCA increased native LDL binding and uptake by 20%, and degradation by 50%. None of the agents tested had any effect on the binding, uptake and degradation of methylated LDL. The increased hepatic LDL uptake induced by UDCA was not altered in the presence of calcium antagonists, while the increased degradation of LDL by UDCA was abolished by the addition of 50 /xM of either verapamil or nifedipine. However, 100 /xM diltiazem and 700 /zM UDCA stimulated LDL degradation without any additive effect. These studies show that different calcium antagonists have differential effects on hepatic LDL metabolism. The similarities between the effect of diltiazem and UDCA on LDL metabolism and the absence of any additive effect, suggest that these two agents have a similar mechanism of action, which may involve the integration of both agents into the plasma membrane lipid bilayer. Keywords: Verapamil; Nifedipine; Diltiazem; Ursodeoxycholicacid; Calcium antagonist; LDL metabolism; (Hamster hepatocyte)

1. Introduction

Ursodeoxycholic acid (UDCA), a bile acid used to dissolve cholesterol gallstones in man [1-4], has been reported to modulate the hepatic uptake of LDL [5,6]. Chronic feeding of U D C A to Golden Syrian hamsters increases receptor-dependent hepatic LDL uptake, in spite of a marked suppression of bile acid synthesis [7]. In addition, we showed that U D C A stimulates the binding, uptake and degradation of L D L in isolated hamster hepatocytes, and that the effect of U D C A is specific to the LDL

Abbreviations: LDL, low density lipoproteins; UDCA, ursodeoxycholic acid, i.e. 3a,7/3-dihydroxy-5fl cholan-24-oic acid; HMG-CoA, 3-hydroxy-3-methylglutaryl-CoA. * Corresponding author. Fax: + 1 (202) 9943435. 0005-2760/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0005-2760(96)00043-4

receptor [5,6]. We hypothesized that U D C A makes accessible a pool of LDL receptors present at the level of the hepatic plasma membrane [5]. However, the mechanism of the modulating action of this bile acid on LDL metabolism is unclear. From our studies [5,6] we concluded that this unique property of U D C A may be related to its distinct physicochemical characteristics or its structural conformation, since none of the other bile acids tested was able to significantly affect hepatic LDL metabolism. However, ketoconazole, a nonsteroidal agent, has been shown to increase the binding of LDL to the same extent as does U D C A [5]. Calcium antagonists, also called calcium channel blockers, are classified according to their chemical structure [8]. Although at least six different classes of calcium antagonists have been reported, the three major classes of cal-

B. BouscareIet al. / Biochimica et BiophysicaActa 1301 (1996)230-236

cium antagonists include the papaverine derivative, verapamil, the dihydropyridine derivative, nifedipine, and the benzothiazepine derivative, diltiazem. Calcium antagonists are mostly known for their anti-hypertensive actions (for review, see [9-11]). More recently, calcium antagonists have also been shown to play a role in the treatment of arteriosclerosis (for review see [12,13]). Initially, verapamil has been shown to prevent the formation of atherosclerotic plaques in rabbits fed a lipid-rich diet [14], probably through regulation of cellular LDL metabolism [15]. Increased LDL receptor synthesis by nifedipine has also been reported in human skin fibroblasts [16]. Since calcium antagonists have been shown to affect LDL metabolism in cells, such as fibroblasts [16,17], and since the liver is the major site for their metabolism [9], the purpose of the present study was to compare the effect of UDCA with that of agents from the major classes of calcium antagonists on LDL metabolism in isolated hamster hepatocytes. This investigation provides information on the specificity, as well as on the mechanism of the modulating action of certain non-steroidal agents on the receptor-dependent LDL metabolism in isolated hamster hepatocytes.

2. Materials and methods 2.1. Materials

~25I-labelled sodium iodide (specific activity 16-20 mCi//xg) was purchased from Amersham, Arlington Heights, IL. Verapamil, nifedipine, diltiazem and gelatin were purchased from Sigma (St. Louis, MO). UDCA, supplied by Tokyo Tanabe (Tokyo, Japan), was 98-99% pure as judged by gas-liquid chromatography. All other chemicals used were of the highest analytical or HPLC grade available from commercial sources.

231

wt per 1 ml Krebs-Henseleit bicarbonate buffer containing 118 mM NaC1, 5 mM KC1, 2.5 mM CaC12, 1.2 mM KH2PO 4, 1.2 mM MgSO 4 and 25 mM NaHCO 3, pH 7.4. Results from previous studies [5,6,22] have shown that the effect of bile acids on hepatic LDL metabolism was not mediated through collagenase activation. Furthermore, it has previously been reported that the hepatocellular binding and uptake of bile acids [20], as well as the biological effect [21], was similar when the cells were incubated in either a gelatin- or albumin-containing buffer. Therefore, unless otherwise indicated, the cells were incubated in Krebs-Henseleit buffer containing 1.5% gelatin. Prior to each experiment, the hepatocytes were incubated for 20-30 min at 37°C, under constant agitation and gassing (O2/CO 2, 95%:5%), to allow the cells to reach steady state. 2.4. LDL preparation and labeling

Blood was collected from human subjects and from hamsters into EDTA-containing vacuum collection tubes. The plasma was separated by centrifugation at 300 × g for 40 min at 4°C. Plasma lipoproteins were separated by density gradient ultracentrifugation, as previously described [5]. For some experiments, the human LDL fraction was reductively methylated according to the method of Weisgraber et al [23]. Human methylated LDL and hamster LDL particles were labeled with 125I to a specific activity of 80-200 c p m / n g protein, using the iodine monochloride technique modified for lipoproteins [24], and were dialyzed overnight against 0.15 M NaC1, 0.001 M NaN 3 and 0.01 M Tris-HC1 (pH 7.4). 125I-LDL were filtered through a 0.45 /zm cellulose-nitrate membrane filter (Millipore, Bedford, MA) and used within two days of preparation. The 125I radioactivity of the LDL particle was mostly trichloroacetic acid precipitable ( 9 6 _ 3%), while the lipid-associated radioactivity represented 3 + 2% of the total.

2.2. Rationale .[or the model 2.5. Uptake and degradation of 12~I-LDL

As previously reported [5], the freshly isolated hepatocyte model was selected since the effect of UDCA on the hepatic LDL metabolism has been linked in vivo to the cellular uptake of the bile acid [7]. However, although active transport is expressed in isolated hepatocytes, this effect disappears with time in culture [18]. Furthermore, although slower than in perfused liver [19], the binding, uptake and degradation of LDL occurs in isolated hepatocytes. 2.3. Isolation and incubation of hepatocytes

Male Golden Syrian hamsters (100-130 g body weight), fed a 0.027% cholesterol rodent chow diet, were used. Hepatocytes were isolated by perfusion of the liver with collagenase, as previously described [5,20,21]. The cells were suspended at a final concentration of 40-50 mg wet

Since previously, it has been reported that the hepatocellular uptake of LDL reaches steady state after 30 rain [5], cell-associated LDL was determined after 60 rain incubation at 37°C. Thus, to measure cell-associated LDL, isolated hepatocytes were incubated in bicarbonate buffer containing 20 to 30 /~g/ml ~25I-hamster LDL in the presence and absence of an excess of human LDL (1-2 mg/ml), with and without the addition of the respective agent tested. Native human LDL was used in order to determine the non-specific, cell-associated LDL. In addition, in certain experiments, methylated ~25I-human LDL was used instead of hamster LDL to study the effect of the agents on the uptake and degradation of LDL by the receptor-independent pathway. At the indicated periods of time, duplicate aliquots of the cell suspension were diluted in ice-cold bicarbonate buffer and centrifuged at 500 × g

232

B. Bouscarelet aL/ Biochimica et BiophysicaActa 1301 (1996)230-236

for 2 min. The cells were then washed and centrifuged 4-times [5]. Finally, the cells were resuspended in 0.2 N NaOH, and the radioactivity was measured in a Beckman model 4000 gamma counter (Beckman Instruments, Palo Alto, CA). LDL degradation was assessed by quantitating the radioactivity present in monoiodotyrosine residues in the incubation media as previously described [5] according to the method of Goldstein et al. [25].

~)~

2.6. UDCA binding to native and methylated LDL

°:21

The respective binding of 700/xM of 14C-labeled UDCA to native and methylated LDL was studied according to the method previously described by Ceryak et al. [26]. Briefly, aliquots of either native LDL or methylated LDL (40 + 8 /zg/ml) were transferred into nitrocellulose dialysis bags (Spectra Por, M r cutoff, 6000-8000) which were suspended in vials containing 0.05 /xCi [24-14 C]UDCA, and incubated at 37°C for 6 h. Duplicate aliquots of the dialysand and dialysate solutions were removed for determination of 14C radioactivity by scintillation counting. The UDCA specifically bound to the respective lipoproteins was calculated by subtracting the radioactivity in the dialysate (free bile salt) from the radioactivity in the dialysis bag (total bile salt).

t2" I

'

• • .e,. , .

UDCA 700uM VERAPAMIL NIFEDIPINE DILTIAZEM

~t

1t.5

9.5

I 0

I 20

I

40

I 80

I

60

[AGENT]

I t00

uM.

Fig. 1. Dose-dependent effect of different calcium antagonists and of UDCA on binding and uptake of 123I-hamsterLDL in isolated hamster hepatocytes. Cells were isolated from Golden Syrian hamsters and incubated in a Krebs-Henseleitbicarbonate buffer containing 1.5% gelatin at 37°C with 125I-hamsterLDL (25-30 /zg/ml) and with 700 /zM UDCA or increasing concentrations (10-100 /zM) of verapamil, nifedipine and diltiazem. After 60 min incubation, 2 aliquots were centrifugedat 500 × g, and washed 4-times with ice-coldbicarbonatebuffer. The radioactivityin the pellet was then counted in order to determine the amountof cell-associated l:5I-hamsterLDL, which was expressedas ng/mg wet wt cells in the 60 min period. Data points are mean+ S.E. of values of experiments done in triplicate. ( *) Significantlydifferent from control, P < 0.05.

3. Results

LDL in isolated hamster hepatocytes. After 60 min incubation with 25-30 / z g / m l hamster LDL, both verapamil and nifedipine were able to significantly increase the uptake of LDL by 10-15% only at a respective concentration of 100 /zM. Under the same conditions, diltiazem maximally stimulated LDL uptake by 15-20% at concentrations greater than or equal to 25 /xM ( P < 0.05). The effect of diltiazem on cell-associated LDL was similar to that observed with 7 0 0 / z M UDCA (Fig. 1). In the present study, a UDCA concentration of 700 ~ M was selected, since we have previously reported that the UDCA-induced increase in hepatic LDL binding and uptake is a function of the bile acid concentration, with a maximum effect observed at UDCA concentrations > 700 /xM [5]. The effect of the different calcium antagonists on cellassociated LDL after 60 min incubation, was also measured in the presence of 700 /zM UDCA (Fig. 2). While 100 /xM of verapamil, nifedipine and diltiazem significantly increased cellular LDL binding and uptake, respectively, none of these agents had an additive effect when tested in the presence of UDCA. The increase in cell-associated LDL of around 20% induced by the different calcium antagonists in the presence of UDCA was similar to that induced by UDCA alone (Fig. 2).

3.1. Effect of calcium antagonists and UDCA on cell-associated LDL

3.2. Effect of calcium antagonists and UDCA on cellular LDL degradation

The results reported in Fig. 1 show the dose-dependent effect of different calcium antagonists on cell-associated

The dose-dependent effect of the calcium antagonists on the degradation of LDL was studied in order to determine

2.7. Bile acid stock solution and protein assay

UDCA stock solutions were prepared by dissolving the sodium salt of the bile acid in normal saline (pH 7.5) with a final concentration of 100 mM. In all the experiments, the effect of the bile acid was compared to that of normal saline as a control. Protein was measured using the Pierce protein assay kit (Pierce, Rockford, IL) with bovine serum albumin as standard. However, in the experiments performed at 37°C, the cells were incubated in buffer containing 1.5% gelatin to protect the cells against proteolytic degradation. Therefore, since it is impossible to completely remove the gelatin in these experiments, the results were expressed per mg of wet weight of cells (mg cell, 1 mg of wet weight of cells was equal to 0.14-t-0.02 mg of protein, determined by measuring the protein content before and the cell wet weight after incubating the hepatocytes in Krebs-Henseleit bicarbonate buffer containing 1.5% gelatin). If not otherwise indicated, all experiments were performed in duplicate, and the results expressed as the mean + standard error (S.E.).

B. Bouscarel et al. / Biochimica et Biophysica Acta 1301 (1996) 230-236

233

12.5"



CONTROL t2.0-

W
**

5.5..5"

~

*

G

UDCA z

*.

UDCA700uM

• ,t

VERAPAMIL NIFEDIPINE

.

5-

O

*

,-I i11 u

u m

3

E

m

H'~. 121 C~

2

as J

m

.J o _J

1

t0.0. 0

*

*

9.5. -1

I

I

o

I

20

9.0.

CTL

VERAP

VERAP

50uM

IOOuM IOOuM

NIFEO

Fig. 2. Effect of different calcium antagonists on cell-associated J25Ihamster LDL in the presence and absence of UDCA. For experimental details see legend of Fig. 1 and Section 2. Isolated hamster hepatocytes were incubated at 37°C for 60 min with 125I-hamster LDL (29_+3 /xg/ml) and 50 /xM verapamil, or 100 /zM of either verapamil, nifedipine or diltiazem, in the presence or absence of 700 ~ M UDCA. Nonspecific cell associated LDL was determined under the same condition and in the presence of an excess of human LDL ( 1 - 2 m g / m l ) . The results represent the difference between the total and the non-specific cell-associated LDL and were expressed as n g / m g wet wt cells in the 60 min period. Data points are mean _+S.E. of values of experiments done in triplicate. ( * ) Significantly different from control, P < 0.05. ( * * ) Significantly different from experiment performed in absence of UDCA, P < 0.05.

whether these agents either have the same mechanism of action as that of UDCA, or decrease the cellular LDL degradation as previously reported in other models, including fibroblasts [27,28] and HepG2 cells [28]. The results reported in Fig. 3 indicate that verapamil and nifedipine inhibited the hepatocellular LDL degradation in a dose-dependent manner. Verapamil had a significant inhibitory effect of more than 50% at concentrations greater than 25 /xM, with an almost complete inhibitory effect at 100/zM. Nifedipine had an inhibitory effect on LDL catabolism similar to that of verapamil. On the other hand, diltiazem stimulated the hepatocellular LDL degradation with a maximum effect of 75 and 100% compared to control, at concentrations of 50 and 100 /xM, respectively. Diltiazem (100 /xM) stimulated the degradation of LDL to the same extent as that observed with 700 /zM UDCA. The more than 50% stimulation of LDL degradation by 700 /xM UDCA was abolished in the presence of either verapamil or nifedipine, while it remained unaffected by diltiazem (Fig. 4). The possibility that the degradation of LDL was affected by collagenase, which could have remained with the isolated cells, and been subsequently released into the incubation buffer, was investigated. In this study, incubation medium was collected following exposure to isolated hepatocytes for 40 rain at 37°C, and was then incubated, in the absence of cells, with 20 /zg/ml J25I-LDL in the

I

60 [AGENT]

DILT

100uM

I

40

GO

I

100

uH.

Fig. 3. Dose-dependent effect of different calcium antagonists and of UDCA on the degradation of I25I-hamster LDL in isolated hamster hepatocytes. For experimental details see legend of Fig. 1 and Section 2. Isolated hamster hepatocytes were incubated in a Krebs-Henseleit bicarbonate buffer containing 1.5% gelatin at 37°C with 2 9 + 3 /.Lg/ml of lZSI-hamster LDL without (control, CTL) or with the addition of increasing concentrations (25-100 /xM) of verapamil, nifedipine and diltiazem, respectively, or 7 0 0 / x M UDCA. After 60 min incubation, the radioactivity present as free iodotyrosine residues in the supernatant, following protein precipitation with 12.5% trichloracetic acid, was determined, and represented the amount of J25I-hamster LDL degraded in the 60 min incubation period. Data points are mean + S.E. of values of experiments done in triplicate. ( * ) Significantly different from control, P < 0.05.

presence of either 0.9% NaC1 (control), 700/xM UDCA or 100 IzM diltiazem for an additional 60 min. In a parallel experiment, the same concentration of ~25I-LDL was incut0-

~CONTROL

~u~A z

Hr4 I---~

,¢ m o u n" m (.9 E tld~

o mc

6

* A ~

o .J

2

o

--~ CTL

VERAP 50uM

VERAP IOOuM

NIFED IOOuM

DILT IOOUM

Fig. 4. Effect of UDCA in the presence and absence of different calcium antagonists on the degradation of 125I-hamster LDL in isolated hamster hepatocytes. For experimental details see legend of Fig. 1, Fig. 3 and Section 2. Data points are mean + S.E. of values of experiments done in triplicate. ( * ) Significantly different from control, P < 0.05. ( * * ) Significantly different form experiment performed in absence of UDCA, P < 0.05.

234

B. Bouscarel et al. / Biochimica et Biophysica Acta 1301 (1996) 230-236

Table 1 Effect of different calcium antagonists and of UDCA on cellullar association (binding and uptake) and degradation of methylated 125I-humanLDL in isolated hamster hepatocytes Treatment Methylated 25I-human LDL

Control UDCA Verapamil Nifedipine Diltiazem

cell-associated (ng/mg of cells)

degraded (ng/mg of cells)

5.20 _+0.355 6.08 _+0.503 5.51 _+0.578 5.36_+0.211 5.07 _+0.234

0.29 _+0.056 0.17 _+0.033 0.19 _+0.089 0.17 _+0.030 0.32 _+0.008

Cells were isolated from Golden Syrian hamsters and incubated in a Krebs-Henseleit bicarbonate buffer containing 1.5% gelatin at 37°C with methylated 125I-humanLDL (20_+5 /xg/ml) and with 0.9% NaC1 (control), 700 /zM UDCA or 100 /xM of either verapamil, nifedipine or diltiazem. After 60 min incubation, 2 aliquots of the respective cell suspension were centrifuged at 500 X g, and washed 4 times with ice-cold bicarbonate buffer. The radioactivity in the pellet was then counted in order to determine the amount of cell-associated 125I-LDL,which was expressed as ng/mg wet wt cells in the 60 min period. The radioactivity present as free iodotyrosineresidues in the supernatant, following protein precipitation with 12.5% trichloracetic acid, was determined as described in Section 2, and represented the amount of t25I-LDLdegraded in the 60 min incubation period. Data points are mean _+S.E. of values of triplicate experiments.

bated with fresh incubation buffer in order to determine the basal L D L degradation. At the end of this period, L D L degradation was determined as described in Section 2 and expressed as percentage of the initial dose of LDL. The LDL degraded represented 0.92 +__0.12% of the dose under control conditions, while it was 1.00 + 0.09% and 0.91 + 0.14% in the presence of U D C A and diltiazem, respectively. The basal degradation of LDL was 0.65 + 0.17%. Therefore, the collagenase-induced LDL degradation was approximately 0.3% of the dose, and was not affected by the presence of either U D C A or diltiazem. 3.3. Effect o f calcium antagonists and UDCA on cell associated and degradation o f methylated LDL

To determine the possible effect of the calcium antagonists and U D C A on LDL receptor-independent LDL metabolism, isolated hepatocytes were incubated for 60 min with 20 ___5 / x g / m l of methylated human L D L and in the presence of either 0.9% NaC1 (control), 700 /~M UDCA, 100 /xM verapamil, 100 /xM nifedipine or 100 /xM diltiazem. In contrast to the effect of these agents on either or both specific cell-association and degradation of native LDL, none of the agents studied had an effect on either the cellular uptake or degradation of methylated LDL (Table 1). Since bile acids have been shown to bind to lipoproteins [26], the binding of 700 /xM U D C A to 40 +_ 8 / z g / m l of either native LDL or methylated L D L was investigated, in order to explore the possibility that the observed differ-

ences in the effect of U D C A on the metabolism of native vs. methylated LDL was due to differences in binding of U D C A to these two particles. After 6 h incubation at 37°C, the specific binding of U D C A was 0.52 +__0.1 n m o l / / z g LDL and 0.83 ± 0.24 n m o l / / x g methylated LDL, respectively.

4. Discussion This study is the first to compare the short term effect of U D C A on hepatic LDL metabolism with that of different calcium antagonists. The results indicate a differential effect of the different classes of calcium antagonists on LDL metabolism. Only at concentrations of around 100 p~M, did verapamil and nifedipine, respectively, significantly stimulate hepatic LDL binding and uptake. Previously, Stein et al [15] have observed that, after a 3 hour incubation with 50 /xM verapamil, the uptake of L D L by both endothelial cells and human skin fibroblasts was increased by around 6%, while after 24 h incubation this increment was between 170% and 300%, respectively. Furthermore, verapamil has also been shown to increase the binding and uptake of LDL in other cell types, including HepG2 cells [28] and aortic cells [29]. Although not generally accepted [30], the mechanism of action of verapamil has been associated with the inhibition of either LDL degradation or cholesterol ester hydrolysis, or both [31,32]. Results from the present study suggest that, in the isolated hepatocyte model, the observed short-term increase in LDL uptake induced by verapamil and nifedipine is principally due to the inhibition of the degradation of LDL. This will, in turn, impede any possible cholesterolinduced down-regulation of the LDL receptor synthesis. However, the possibility of stimulation of the synthesis of the LDL receptor during prolonged exposure of the cells to these calcium antagonists cannot be ruled out. Furthermore, different calcium antagonists have differential effects on LDL metabolism. These differences may be related to the cell type, as well as to the conditions of incubation, i.e., short-term vs long-term exposure of the cells to the calcium antagonist. In our study, the results clearly show that verapamil and nifedipine have similar modulating effects on hepatic LDL metabolism. On the other hand, it is notable that diltiazem stimulated both the uptake and the degradation of L D L by isolated hepatocytes. Diltiazem has been reported to stimulate LDL degradation in HepG2 cells at concentrations as low as 10 /zM, while having an inhibitory effect at 50 and 100 /zM [28]. Diltiazem also has a biphasic effect on the degradation of LDL by fibroblasts from normal human subjects, while this calcium antagonist stimulates LDL degradation in fibroblasts isolated from type IIa heterozygous hypercholesterolemic patients [28]. Previous studies from this laboratory [5,6,22] indicate that the increase of cell-associated LDL by U D C A was not

B. Bouscarel et al. / Biochimica et Biophysica Acta 1301 (1996)230-236

related to a detergent or toxic effect of this bile acid on the plasma membrane. These results are supported by studies from other laboratories showing that, up to very high concentrations, UDCA is not only non-toxic, but shows evidence for preventing the cytotoxicity induced by other bile acids [33,34]. Furthermore, it has been suggested that the use of collagenase in the hepatocyte isolation technique may alter the LDL receptor [35], however, incubation of liver plasma membranes with high concentrations of collagenase (0.4 m g / m l ) does not affect the number of LDL receptors present at the membrane level [22], suggesting a collagenase-independent effect. In addition, the present study indicates that the effect of residual collagenase on LDL degradation is negligible, and that neither UDCA nor diltiazem modulates this effect. Although it has previously been reported by us that UDCA affects LDL uptake only through the LDL receptor-dependent pathway [5], it was not determined if this was also the case as far as LDL degradation was concerned. The present study clearly shows that none of the agents tested, including UDCA, verapamil, nifedipine and diltiazem affect the LDL receptor-independent pathway. However, the observed difference in the UDCA-induced increased uptake and degradation of native LDL compared to methylated LDL could be due to a difference in binding of the respective LDL particles to the bile acid. The present study indicates that the binding of U D C A to LDL is similar to what was previously reported for UDCA [5] and for its 7a-OH epimer, CDCA [26], and that there was no difference in the binding of UDCA to both native and methylated LDL. Therefore, the UDCA-induced increased degradation of native LDL compared to methylated LDL is not due to differences in binding of these two particles to UDCA. One of the possible mechanisms of action of UDCA and diltiazem to increase LDL metabolism may be associated with their integration into the plasma membrane lipid bilayer. This interpretation of the findings is supported, at least for UDCA, by the results showing that, while the effect of U D C A on hepatic LDL uptake is not altered by the presence of either verapamil or nifedipine, its effect on the degradation of LDL is completely abolished by the two calcium antagonists. This site of action is also suggested by our previous studies showing that the UDCA-induced increase in hepatic LDL metabolism occurs in the absence of both LDL receptor processing and cycling [5], as well as without any modification of the membrane lipid composition [22]. Furthermore, Giildiituna et al [33] have shown, by electron paramagnetic resonance spectroscopy, that the steroid nucleus of UDCA is localized in the apolar core of the membrane bilayer and could, therefore, increase membrane stabilization. Similarly, while verapamil, nifedipine and diltiazem are amphiphilic molecules, their incorporation into the membrane bilayer is different [36], probably due to their specific degree of hydrophobicity [30,32]. The nonpolar por-

235

tion of the calcium antagonists will insert itself into the hydrophobic part of the membrane. In contrast, the polar portion of the calcium antagonists is characterized by a unique association with specific membrane components, such as phospholipid [36,37]. This will result in the disruption of the membrane structure, in the case of verapamil [38], or will allow the molecule to reach the intracellular face of the membrane bilayer, in the case of diltiazem [39], which may contribute to membrane stabilization. Furthermore, the degree of hydrophobicity of papaverine and dihydropyridine derivatives is more comparable than that of benzothiazepine derivatives [36,37]. This supports the observed similar effects of verapamil and nifedipine, on hepatic LDL metabolism, in contrast to the effect of diltiazem. The regulation of cellular LDL metabolism through direct effects on the plasma membrane has also been suggested for different fatty acids [40], imidazole derivatives [41] and bile acids [5,6]. Another possible mechanism of action of certain calcium antagonists on LDL metabolism involves modulation of cellular calcium homeostasis [28]. Calcium antagonists have been shown to affect the influx of calcium from the extracellular space, as well as the release of calcium from intracellular calcium stores by binding to specific voltagesensitive calcium channels [42,43]. However, verapamil, nifedipine and diltiazem have different pharmacological activity and bind to different receptors or different calcium channels [44,45]. Furthermore, increased cytosolic calcium has been linked to the phosphorylation and inhibition of HMG-CoA reductase activity in rat hepatocytes [45]. However, in the present study, the period of time in which the uptake of LDL is studied is insufficient to allow for the subsequent transcriptional regulation of LDL uptake by this mechanism. In addition, this mechanism would not account for the decrease in LDL degradation observed with verapamil and nifedipine. In conclusion, the results of the present study indicate that verapamil and nifedipine modulate hepatocellular LDL metabolism by a mechanism distinct from that of diltiazem and UDCA. The short-term effect of verapamil and nifedipine is to stimulate LDL uptake through inhibition of the degradation of LDL. On the other hand, diltiazem and UDCA stimulate the degradation of LDL probably by increasing LDL uptake. The possible action of the latter two agents, as supported by previous studies [5,22] may be to associate with certain regions of the plasma membrane bilayer, and to make LDL receptors, present in a cryptic pool, accessible for binding.

Acknowledgements This work was supported in part by a grant from the American Heart Association (Nation's Capital Affiliate Inc.). Dr. S. Ceryak was a recipient of an American Liver Foundation Student Research Fellowship and is a fellow of

236

B. Bouscarel et al. / Biochimica et Biophysica Acta 1301 (1996) 230-236

the American Heart Association (Nation's Capital Affiliate Inc.). A portion of this work was presented at the annual meeting of the American Society for Clinical Investigation. Baltimore, MD in May, 1994.

References [1] Tint, G.S., Salen, G., Colalillo, A., Graber, D., Verga, D., Speck, J. and Shefer, S. (1982) Ann. Int. Med. 97, 351-356. [2] Fromm, H., Roat, J.W., Gonzalez, V., Sarva, R.P. and Farivar, S. (1983) Gastroenterology 85, 1257-1264. [3] Fromm, H. (1984) Gastroenterology 87, 229-233. [4] Leuschner, U., Leuschner, M., Sieratzki, J., Kurtz, W. and Hubner, K. (1985) Dig. Dis. Sci. 30, 642-649. [5] Bouscarel, B., Fromm, H., Ceryak, S. and Cassidy, M. (1991) Biochem. J. 280, 589-598. [6] Bouscarel, B., Ceryak, S. and Fromm, H. (1993) in Bile acids and the Hepatobiliary System (XII International Bile acid Meeting), pp. 234-242, Kluwer Academic, Dordrecht. [7] Malavolti, M., Fromm, H., Ceryak, S. and Roberts, I.M. (1987) J. Lip. Res. 28, 1281-1295. [8] Vanhoutte, P.M. (1987) Am. J. Cardiol. 59, 3A-8A. [9] Ram, C. VS. (1985) Am. J. Med. Sci. 290, 118-133. [10] Triggle, D.J. (1990) Stroke 21, IV49-1V58. [11] Hansson, L. (1991)Am. Heart J. 122, 308-311. [12] Henry, P.D. (1985) Circulation 72, 456-459. [13] Parmley, W.W. (1987) Am. J. Med. 82, 3-8. [14] Rouleau, J.-L., Parmley, W.W., Stevens, J., Wikman-Coffelt, J., Sievers, R., Mahley, R.W. and Havel, R.J. (1983) J. Am. Coll. Cardiol. 1, 1453-1460. [15] Stein, O., Leitersdorf, E. and Stein, Y. (1985) Arteriosclerosis 5, 35 -44. [16] Filipovic, I. and Buddecke, E. (1986) Biochem. Biophys. Res. Commun. 136, 845-850. [17] Block, L.H. Matthys, H., Emmons, L.R., Perrnchoud, A., Erne, P. and Roth, M. (1991) Proc. Natl. Acad. Sci. USA 88, 9041-9045. [18] Follmann, W., Petzinger, E., and Kinne, R.K.H. (1990) Am. J. Physiol. 258, C700-C712. [19] Griglio. S. (1986) in Research in Isolated and Cultured Hepatocytes (Guillouzo, A. and Guguen-Guillouzo, C., eds.), John Libbey Eurotext Ltd./INSERM, Montrouge, France. [20] Bouscarel, B., Nussbanm, R., Dubner, H. and Fromm, H. (1995) Hepatology 21, 145-154. [21] Bouscarel, B., Gettys, T., Fromm, H. and Dubner, H. (1995) Am. J. Physiol. 268, G300-G310.

[22] Bouscarel, B., Ceryak, S. and Fromm, H. (1995) Lipids 30, 607-617. [23] Weisgraber, K.H., Innerarity, T.L. and Mahley, R.W. (1978) J. Biol. Chem. 254, 6876-6879. [24] Bilheimer, D.W., Eisenber, S. and Levy, R.I. (1972) Biochim. Biophys. Acta 260, 212-221. [25] Goldstein, J.L., Basu, S.K. and Brown, M.S. (1983) Methods Enzymol. 98, 241-260. [26] Ceryak, S., Bouscarel, B. and Fromm, H. (1993) J. Lipid Res. 34, 1661-1674. [27] Ranganathan, S., Harmony, J.A.K. and Jackson, R.L. (1982) Biochem. Biophys. Res. Commun. 107, 217-224. [28] Paoletti, R., Bernini, F., Fumagalli, R. Allorio, M. and Corsini, A. (1988) Ann. N.Y. Acad. Sci. 522, 390-398. [29] Moil, S., Ito, H. and Yamamoto, K. (1988) J. Exp. Med. 154, 329-333. [30] Bernini. F., Catapano, A.L., Corsini, A., Fumagalli, R., and Paoletti, R. (1989) Am. J. Cardiol. 64, 129-134. [31] Stein, O. and Stein, Y. (1987) Arteriosclerosis 7, 5878-5884. [32] Etingin, O.R. and Hajjar, D.P. (1990) Circ. Res. 66, 185-190. [33] Giildiituna, S., Zimmer, G., Imhof, M., Bhatti, S., You, T. and Leuschner, U. (1993) Gastroenterology 104, 1736-1744. [34] Galle, P.R., Theilmann, L., Raedsch, R., Otto, G. and Stiehl, A. (1990) Hepatology 12, 486-491. [35] Havel, R.J. (1986) Methods Enzymol. 129, 591-612. [36] Mason, R.P., Moisey, D.M. and Shajenko, L. (1992) Mol. Pharmacol. 41,315-321. [37] Scheufler, E., Vogelgesang, R., Wilffert, B., Pegram, B.L., Hunter, J.B., Wermelskirchen, D. and Peters, T. (1990) J. Pharmacol. Exp. Ther. 252, 333-338. [38] Retzinger, G.S. and Cohen, L. (1992) J. Pharmacol. Sci. 81, 49-53. [39] Kimball, S.D., Floyd, D.M., Das, J., Hunt, J.T., Krapcho, J., Rovnyak, G., Duff, K.J., Lee, V.G., Moquin, R.V., Turk, C.F., Hedberg, S.A., Moreland, S., Bilttain, R.J., McMullen, D.M., Normandin, D.E. and Cucinotta, G.G. (1992) J. Med. Chem. 35, 780793. [40] Bihain, B.E., Deckelbaum, R.J., Yen, F.T., Gleeson, A.M., Carpenter, Y.A. and Witte, L.D. (1989) J. Biol. Chem. 264, 17316-17321. [41] Kempen, H.J., Van Son, K., Cohen, L.H., Griffioen, M., Verboom, H. and Havekes, L. (1987) Biochem. Pharmacol. 36, 1245-1249. [42] Rosales, C. and Brown, E.J. (1992) J. Biol. Chem. 267, 1443-1448. [43] Kato, H., Hayashi, T., Koshino, Y., Kutsumi, Y., Nakai, T. and Miyabo, S. (1992) Biochim. Biophys. Res. Commun. 188, 934-941. [44] Snyder, S.H. and Reynolds, l.J. (1985) N. Engl. J. Med. 313, 995 -998. [45] Gaggelli, E., Maccotta, A. and Valensin, G. (1992) J. Pharmacol. Sci. 81,367-370.