Biochlmtca et Biophysi~ Acta, 1081 (19911293-3U0 ,~ 1991 ElsevierScience Publishers B.V (aiomedieal Division)0005-27611/9|/S03.50 ADONIS 0005276091000904
293
Effects of probucol on lipid metabolism and secretion in long-term cultures of adult rat hepatocytes Francisco M. De La Vega * and Tomfis Mendoza-Figueroa Deparlmenl of Pha~lacoloD" and To.~icolog~'. Centro de b~cestigacidn y de Estudlos At~.,~,zadosdel IPN. Me.,ico City. D.F (Mdxtco)
(n~eived 19 September 1990) Key ,.vords: Probucol; Hepatocyte primary,cullure: Lipid secretion: Cholesterol; Phospholipid:Triacylglycerol TO study the effects of prohucul on hepatic lipid metabolism, we used adult rat hepatocytes cultared on a feeder layer of 3'I"3 cells lethally treated with mitomycin C. These cultures synthesize 'and secrete for at least 2 weeks various liplds from [MC]aeetate and IMCluleate precursors. Treatment with 20 / z g / m l of probucul for 7 and 14 days decreased the secretion of various radhilaheled lipid species to the culture medium and prnduced an intracytoplasmic acemnulation of triacylglycerul droplets. The lipids whose secretion was most decreased were free and esteriiied cholesterol (50-70% reduction). Secretion of triacylglycesols and pbusphulipids was also reduced but to a lower extent. Intracytoplasmle trlac3~lglycerols accumulated and the actlvlty of glycerol phosphate dehydrogenase, a marker enzyme of glycerolipid synthesis, also increased (35-56%). The total incorporation of both radioactive precursors into free and estecir~d chulesteml and phusphulipids was reduced 20-60%. Our data show that 2-week treatment of 3T3-hepatoc~ le cultmes with pharmacological concentrations of probuenl reduces significantly lipid secretion and s u r e s t that at least part of the in vlvo hypullpldemle effect of probuenl could be attributed to a decrease in the secretion of lipids (i.e.. llpoprutelus) by helmtoeytes. Introduction Probucol is an effective drug to treat hypercholesterolemia and to prevent both atheroselerosis and cardiovascular disease [1]. It is the only drug that prevents the progression of atherosclerotic plaques [2] apparently through inhibition of lipid storage in macrophages, an effect probably related to its antioxidant properties [31. It is also one of the few drugs effective in the treatment of homozygous familial hypercholestesolemia [4,5]. It produces a mean reduction of 12% in
* Present address: Department of Genetics and Mol~ular Biology, Centto de lnvesligaci6n y de Esludios Avanzados del I.P.N., AP 14-740, M~'xi¢0D.F, 07000, Mdxico. Abbreviations: DMSO. dimeli:yt sulro~de; dpm, disint©gralionsper minute; GPD, t-glycerol-3-phosphate dehydrogenase: HDL. highdensity lipoprolein; HMG-Co,~., 3-hydroxy-3-raethyl glutapjl ¢oenzyme A; LDL, low-density lipoprotein; S~D., slandard deviation: TIC. lain-layer chromatography; VLDL. very-low-density lipopr~ teln. Trivial and systematic names: Probu¢ol, 4,4'-(isopmpylidenm dithio)bis(2,6-di-t-bulylphenol); Tris, 2-amino-2-hydroxymelaylpropane-l,3-diol. Correspondence: T. Mend0ga-Figueroa.Department of Pharmacology and Toxicology,Ceatro de lavestiga¢i6ny de Estudio~sAvanzado~del IPN, AP 14-740, Mexico City, 07000, D.F,, M~eo.
plasma cholesterol [6], and 8-15% and 9 25% reductions in LDL and H D L cholesterol, respectively [6,7]. Despite its importance, the mechanism of the hypolipidemic effect has not been unambiguously established 18,91. In the lasting state the liver accounts for most of the lipoprutein production and removal from plasma. Plasma lipid-lowering drugs ultimately work by decreasing lipoprotein production or increasing their intravascular catabolism and removal from the bloodstream [8]. Neither in human patients nor in experimental animals is clear if probuenl affects hepatic lipid metabolism and secretion, or if its effects are mainly intravascular [10]. Whereas Balasubramauiam et el. [111 and Mieninen [12] reported a reduction in cholesterol synthesis by the liver, Tawara et aL [13] showed that this effect does not occur in hypereholesterolemic mice treated with probucol. Miettinen [12] and Tawara et al. [13] reported an increase in bile salt production and secretion by the liver, but Kesiiniemi and Grundy [7] found no effect of probucol on bile acid or cholesterol excretion and Li et al. [14] found in fact a decrease in sterol balance. Hattori et ai. [15] suggested that probucol reduces the synthesis of tdacylglyeerul rich lipoproteins by the liver. On the other hand, results of Namszewicz et al. [16] and Steinberg [10] suggested that probuenl lowers LDL levels
294 by increasing the rate of their removal from the bloodstream and argued that effects of the drug on the liver or on the intravascular production of LDL do not contribute significantly to the hypolipidemic effect. However, Trezzi et al. [17] found no effect of probucol on the fractional catabolic rate of LDL in rabbits. Other suggested effects of probucol are an increase in the catabolism of H D L j to HDL~ [15], reduction in H D L size [5], and a possible interference on lipoprotein assembly [18]. These contradictory results may be partially attributed to the use of intact animals where the measurement of direct effects of the drug or, the hepatic lipid metabolism and secretion might be obscured by the contribution of other effects on the intravascular metabolism of lipoproteins. Primary cultures of adult rat hepatocytes have provided a useful model to study hepatic effects of several drugs and chemicals on lipid metabolism without the interference of other tissues (vascular, adipose, etc. [1922]). In this paper we report the long-term effects of pharmacological concentrations of probucol on hepatic lipid metabolism and secretion using as model system a primary culture of adult rat hepatocytes seeded on a feeder layer of 3T3 cells. The 3T3-hepatocyte cultures have been shown to retain for several weeks the hepatic characteristic morphology and ultrastructure and various liver-specific functions, such as albumin secretion, basal and inducible levels of cytochrome P-450, and synthesis and secretion of lipids to the culture medium [19-21]. The results obtained show that prnbucol treatment decreases the secretion of lipids to the culture medium, mainly free and esterified cholesterol and reduces the incorporation of labeled precursors into most cellular lipids. Concurrently with these effects there is an intracytoplasmic triacylglycerol accumulation. Materials and Methods
Materials Collagenase type IV, hydrocortisone, insulin, d-biotin, Oil red O, Percoll and dimethyl sulfoxide (DMSO) were purchased from Sigma Chemical (St. Louis, M e ) . Calf serum was obtained from HyClone Laboratories (Logan, UT). Aluminum plates coated with silica gel 60 (0.2 m m thickness) for thin-layer chromatography (TLC) were obtained from E. Merck (Darmstadt, F.R.G.). [l,2-14C]Acetic acid (55 m C i / m m o l ) and [1-14C]oleic acid (57 m C i / m m o l ) were bought from New England Nuclear (Boston, MA). Mytomicin C and probucol were kind gifts from Bristol-Myers (Denver, CO) and Laboratories Lepetit de Mrxico (Mexico City), respectively. All other reagents were of analytical grade and were purchased from Sigma Chemical and J.T. Baker (Mexico City).
Cell cultures Hepatocytes were isolated from male Wistar rats
(180-200 g) by the collageaase-perfusion method of Berry and Friend [23] modified as described earlier [191. Briefly, the liver was perfused with a salt solution containing 5.5 m M glucose. 137 mM NaCI. 50 m M KCI, 0.4 mM N a 2 H P O 4, 25 mM Tris-HCl and 0.01% phenol red (pH 7.4). for 3 rain at 37°C and at a flow rate of 15 m i / m i n and again for 15 min at a rate of 7 m l / m i a with the salt solution containing 100 u n i t s / m l bf collagenase iype IV at 37°C. Cells were dispersed in 40 ml of Eagie's medium modified by Dulbecco and ViSgt supplemented with 77o calf serum, 5 t*g/ml insulin and 0.l tiM d-biotin (basal medium). The cell suspension was filtered through a nylon mesh and allowed to stand for 5 min. The supernatant was discarded, the cells were resuspended in basal medium containing 2070 Pcrcoll [24] and then ~edimented for 3 rain at 300 × g in a Beckman TJ-6 centrifuge. The cell viability, assessed by Trypan blue exclusion, was higher than 9870. The hepatocytes were inoculated ( 4 . 5 . 1 0 s / 3 5 m m dish) on a feeder layer of 3T3 edls [25] that had been lethally treated with mitomycin C [26] and inoculated 1 day earlier at 4 . 1 0 ~ cells/35 m m dish [19]. The cells were placed in a 37°C humidified incubator gassed with 90~, air/10% CO 2. 30 rain after seeding, the dishes were rinsed with serum-free medium and refed with 1 ml of basal medium supplemented with 10 ~ g / m l of hydro-. cortisone (complete medium). Untreated cultures were maintained with complete medium, whereas treated cultures were changed 24 h after cell seeding to medium containing the indicated concentration of probucol previously dissolved in DMSO. The final concentration of D M S O in the culture medium of untreated and treated cell cultures was 1~. The presence of D M S O is not detrimental to hepateeytes [27], but favors the permanence of hepatic differentiated functions, as shown by Isom et al. [28,29]. Cultures of lethally treated 3T3 cells were maintained with complete medium containing 1% D M S O with or without probucol to assess the contribution of the feeder layer to the observed effects. Culture medium was changed daily. The concentration of probucol used throughout this study was widlin the range observed in the plasma of treated subjects [30]. Also, we selected the exposure tirn,~ of 1 and 2 weeks because pharmacological effects of probucol are seen in the rat after treatment for 1 week [11].
Cell morphology and ultrastructure The cultures were examined and photographed with an Olympus I M T inverted photomicroscope equipped with a PM-10AD photographic system. For transmission electron microscopy, cells were processed as described [20] and were observed with a ,leol-2000EX electron microscope at 80 kV.
Incorporation of [taC]acetate and [14C]oleate into cellular and secreted lipids At the indicated culture times. 0.2 / t C i / m l
of
295 [l~C]acetic acid were added to the cultures. After 24 h incubation, the medium was removed and the cells were washed, scraped and disrupted by sonication. For studies with [laC]oleic acid, the cultures were incubated for 2 h with 0.2 ~tCi/ml of the radioactive precursor. Then. the cultures were rinsed twice with 2 ml of serum-free medium and refed with 0.5 ml of fresh non-radioactive medium. After 24 h incubation the medium was remo~ed and the cells processed as before. Lipids from both culture media (secreted lipi.~) and cells (cellular lipids) were extracted as described by Chum and Knowles [311 and analyzed by TLC.
Thin-layer chrorr~atography analysi.s of #ipids ann triatylglycerol cell content The extracted lipids were mixed with rat serum lipids as carriers, evaporated under a nitrogen stream and dissolved in 30 ~tl of chloroform. Samples of 15/tl were analyzed by TLC using a mobile phase of heptane/ diisopropyl ether/acetic acid ( 6 0 : 4 0 : 4 , v / v ) [32]. The radioactive lipids that c~,"r.~grzted with purified standards of triacylglycerols, phospholipids, cholesterol and cholesterol esters, were determined by liquid scintillation in a Packard Tri-Carb 460 counter (Downers Grove, lL) using Aquasol scintillation cocktail INew England Nuclear, Boston, MA). Triaeylglycerol content in the cells was determined by staining the dishes with Oil red O, extracting the dye retained with 1 ml isopropanol and measuring its absorbance at 510 nm. This procedure stains specifically triacylglycerols without interference from phospholipids (Ramirez-Zacaralas, Castre-Muhozledo and KuriHarcuch, unpublished). Triacylglycerol concentration was determined from a standard calibration curve [20].
Enzyme activities Cell extracts used to determine enzyme activities
were prepared as described by Kuri-Harcuch and Green [33]. Malic enzyme (L-malate:NADP oxidoreductase (descarboxylating); EC 1.1.1A0) and L-glycerol-3-phosphate dehydrogeaase (EC 1.1.].8) activities were determined as described elsewhere [34--36]. Protein was determined by the method of Lowry et al. [37].
Data presentatwn and analysis Unless otherwise indicated, results are expressed as the mean _+ S.D. of determinations from three or four culture dishes in a typical experiment; similar results were obtained in three independent experiments. As protein content was very similar between dishes and it did not change during culture for 2 weeks, neither with treatment, all values are expressed per dish. Statistical comparison between groups was performed with the unpaired Studenfs t-test after a randomized analysis of variance revealed significance ( P < 0.05). Results
Alteration ef heparoQ.i ? morpholog)' and ultrastructure by probuvol Control cultures showed under both optical and electron microscope their typical morphology and ultrastructure during culture time, as reported before (Figs. la, c, 2a: [19]). Hepatocytes treated for 7 days with probucol accumulated small lipid droplets homogeneously distributed within the cytoplasm (Fig. l b ) which probably fused arising large droplets after 14 days of exposure {Fig. ld). Electron micrographs revealed that the lipid droplets were not surrounded by membranes (Fig. 2b). and that some times they were associated with large glycogen areas (Fig. 2b, c). Large irregular milcchondria, 3ome autophagic vesicles and abundant microfilament bundles dispersed throughout the cytoplasm were also seen in the treated cells (Fig. 2e, d). All other cellular organelles presented normal morphology.
Fig. l. Effect of probucol on hepatocylc morphology,HepatocyIes were cultured in complele medlum supplememed with 1% DM$O for 7 (a) or 14 days (c) or in complete medium supplemented with 1¢,l DMSO plu.~ 20 ,~g/mI of probucol for 7 (b) or 14 days (d). Phase contrast optics. Bar, 50 pm.
296
Fig. 2. Ultrastmcture of untreated and probucol-treated hepatocytes. Hepalocytes were cultured for 7 days and proe~sed for transmission elc~ctr~ microscopy as de~ribed in Methods (a) Unlreated hepatocyte: N, nucleus; n, nt,cleolus; m, mit~h~dria. (b-d) Hepatocytes treated with 20 pg/ml of probucol: L. lipid droplet; g, glycogenarea; a. autophagic vacuole~ m, mit~hondria; ern, extracellular matrix; arl,owhead,microfilamcnt bundle; arrow, glycogengranules. Bar, a-c, 1/lm; d. 5COnm.
Tria~TIg(~'cerol accumulation in hepato¢yles treated with probucol Triacylglycerol cell content increased in the cultures treated for 14 days in a dose dependent manner (Fig. 3). The increase was significant at 10 and 20 # g / m l of probucol. Triaeyiglycerol accumulation was also timedependent as was evident from the daily morphological observations. Effect of probucol on lipid synthesis and secretion To study the effect of probucol treatment on de novo lipid synthesis and secretion by the hepatocytes, we determined the incorporation of [t4C]acetate and [14C]oleate into cellular, secreted and total (cellular plus secreted) lipids The experiments with [14C}acetate were 24 h pulse labeling experiments whereas the experiments with [laC]oleate were 2 h pulse-24 It chase experiments. The long [14C]acetate pulse was necessary because short pulses resulted in little incorporation (data
not shown). Additionally, with this protocol a steady state of labeling might be obtained, so the amount of label incorporated into a particular lipid species could be an estimate of the amount of the lipid. Ptobucol reduced the incorporation of [14C]acetate into total lipids from 82.6 4- 7.9.103 to 43.5 4- 4 . 3 . 1 0 J d p m / d i s h (mean 5: S.D., n = 4; 47% reduction, P < 0.001) and from 92.6 4- 2 8 . 7 . 1 0 ~ to 57.3 5:30.2.103 d p m / d i s h ( P > 0.05, not significant), after 7 and 14 days of treatment, respectively. TLC analyses of [14C]acetate-labeled lipid species (Fig. 4) revealed that after 7 days of treatment the total incorporation o[ [14C]acetate into free and esterified cholesterol was strongly reduced (Fig. 4C, D). On the other hand, radiolabeled triacylglycerols was the only species whose cellular content increased though its total amount was unchanged (Fig. 4A). The cell content of all other lipids was reduced leading to a net reduction in total incorporation (Fig. 4 B - D ) . Remarkably, the secretion of free
297
45.
~"
*
25
15
0.01
0,1
1,0
Probu¢ol
10
100
(/~g/ml)
Fig. 3. Dose dependent cellular triacylglycerol accumulation reduced by probucol treatment. Hepat~ytes were cultured for 14 days in complete medium supplemented with 1% DMSO and the indicated concentrations of probueol, Culture dishes were fixed and stained with Oil red O. The stain was extracted with i~propanol and its absorbanee determined at 510 am. Triacylglycerol ~ntent (/~g/dish) was determined from a standard calibration cu~e. Poinls and bars represent mean± range of two culture dishes in a typical experimenl. Asterisks indicate significant difference wilh respect to control values ( P < 0.~5 ~.
a n d esterified cholesterol to c u l t u r e m e d i u m w a s red u c e d by 5 0 - 7 0 % (Fig. 4C, D). T h e secretion of labeled triglycerides a n d p h o s p h o l i p i d 3 w a s also r e d u c e d b u t to a lower e x t e n t (Fig. 4A, B). A f t e r 14 days e x p o s u r e , the secretion a n d total i n c o r p o r a t i o n i n t o all labeled lipids w a s n o t a b l y decreased (Fig. 4 E - H ) , b e i n g free a n d esterified cholesterol t h e m o s t decreased species (e.g.. 9 0 ~ a n d 75% r e d u c t i o n of secretion, respectively; Fig.
i"°!
4 G . H). T h e cellular c o n t e n t s of radiolabelcd p h o s p h o lipids a n d cholesterol esters were also reduced (Fig. a F , G ) , whereas t h o s e of cholesterol a n d triacylglyccrols were u n c h a n g e d at this t r e a t m e n t time (Fig. 4 E , H). W h e n we a n a l y z e d the i n c o r p o r a t i o n of [V~C]oleate i n t o the various h p i d species, the results were c o m p a r a ble to those o b t a i n e d with []4C]acetate (Fig. 5). T h e r e were s o m e v a r i a i i o n s that c o u l d be d u e to differences in radio!abeled p r e c u r s o r pools a n d in e x p e r i m e n t a l design. A l t h o u g h the secretion of labekxt cholesterol esters w a s decreased after 7 days t r e a t m e n t , their cell c o n t e n t increased w h e r e a s total i n c o r p o r a t i o n w a s u n c h a n g e d IFig. 5CI. However. as in e x p e r i m e n t s w i t h [laC]acetate the secretion of triacylglycerols decreased while their cellular c o n t e n t increased (Fig. 5 A L T h e i n c o r p o r a t i o n of [~aC]oleate i n t o total, cellular a n d secreted p h o s p h o lipids was reduced significantly IFig. 5B). After 14 days of e x p o s u r e the secretion of all species also decreased (Fig. 5 D - F L Nevertheles,~, at this t i m e the cell c o n t e n t of labeled triacylglycerols a n d cholesterol esters d i d n o t increased significantly (Fig. 5D. F). T h e i n c o r p o r a t i o n or [laC]oleate i n t o cellular a n d total p h o s p h o l i p i d s was similar to t h a t observed a f t e r 7 days of t r e a t m e n t IFig. 5E). T h e c o n t r i b u t i o n of the f~,~eder layers to these effects was insignificant, as studies w i t h cultures of 3T3 cells revealed. T h e i n c o r p o r a t i o n of labeled precursors i n t o lipids of 3T3 cells was I - 5 % of t h a t f o u n d in 3T3h e p a t o c y t e cultures, e x c e p t for p h o s p h o l i p i d s w h e r e the i n c o r p o r a t i o n r e a c h e d a b o u t 10%. F u r t h e r m o r e . we detected n o c h a n g e s in radiotabcled lipid synthesis a n d
!
,
Fig, 4. Eff~I of prohucol on [14Cla~tale incorporation into hepatocyte cellular and ~ecreted lipids. Hcpat~yt~ were cultured for "/IA D) or 14 days (E H) in medium with I% DMSO (unfilled bars) or medium with I% DMSO containing 20/~g/ml probucol (filled barsL Then, cultures wen incubated with 0.2 ~Ci/ml nf [t4C]acetate for 24 h. The incorporation of [b4Clacetate into medium (secteled) and cell ]ipids was measured after extraction and separation by TLC. Total incorporation was the sum of radioactivity in medium and calls. Bars represent values of mean ± S D ( n - 4) in a typical experiment. Asterisks indicatu significant difference with respect IO control values ( * P < 0.05: * * P < 0.01 ). Note that scales for panels C and D ale different.
298 T ~[lfl~
[
pfio~Ollpl~
[ Choeserat~ors
I
I
' nlil n,l,'i0 llll Fig. 5. Effect of probucol on [14C]oleate incorporation into hepatocyte cellular and secreted lipids. Hepato~tes were cultured for 7 (A-C) or 14 days (D-F) in medium with 1~ DM$O (unfilled bars) or medium with 1~ DMSO and 20 pg/ml probucol (filled bars). Then, cultures were incubated with 0.2 #Ci/ml of 114Cloleatefor 2 h. finsed. feted with fresh medium and incubated further for 24 h. The incorporation of [14C]oleate into medium (secreted) and cell lipids wa~ measured after extraction and separation by TLC. Total incorporation was the sum of radioactivity in medium and cells. Bars represent values of mean_+S.D.(n 4) in a typical experiment, Asterisks indicate significant difference with respect to control values ( * P < 0.05; • * P < 0.01). Note that scales for panels C and F are different. secretion by 3T3 cells due to probucol treatment (data not shown).
Effect of probucol on the activity of two lipogenic enzymes We determined the activities of malic enzyme and L-glycerol-3-phosphate dehydrogenase (GPD), markers of fatty acid and glycerolipid synthesis, respectively [38]. The exposure to probucoi increased G P D activity significantly from 77.6 + 18.6 to 121 + 16 (mean 4- S.D., P < 0.05, n = 3) and from 103.1 ++ 13.4 to 140.5 + 26.4 ( P < 0.05, n = 3) nmol N A D • mg protein train 1, at 7 and 14 days of treatment, respectively. On the other hand, the effect of probucol on malic enzyme activity was variable. After 7 days exposure there was a slight decrease, whereas at 14 days there was an increase, although not statistically significant (data not shown). Discussion Our results show that the main effect of probucol on cultured rat hepatocytes was a reduction in the secretion of lipids, as evidenced by the decrease in the amount of radiolabeled lipids released to the culture medium by the treated cultures (Figs. 4 and 5). The secretion of free and esterified cholesterol was strongly decreased, although triacylglycerol and phospholipid secretion was also diminished to a lesser extent (Figs. 4
and 5). At longer treatment times the effects "¢~e~e usually more potent. Our results are consistent with the in viva observed 12 20% reduction of plasma cholesterol [1,30] and the lower reduction of plasma triacylglycerols [1,15,30] produced by probucol, though tiic decrease in lipid secretion by hepatocytes was greater. This is not surprising since hepatecyte cultures may resemble more closely the fasting liver. Due to the limited supply of extracellular lipids there is an increased synthesis of lipids and hence the effect of probucol might be overemphasized. These results suggest that probucol could decrease the secretion of cholesterol-bearing lipoproteins ( H D L or V L D [ ) or their cholesterol content [4,15]. It is not clear which lipoprotein species could be affected by probucol since the secretion of triacylglycerols and phospholipids decreased also. Khan et al. [39] have shown that reduction in the intracellular pool of esterifled cholesterol, an effect of probucol on cultured hepatocytes (Fig. 4C, G), leads to a decrease in V L D L secretion. Further work is necessary to determine directly the effects of probuool on the secretion of H D L and V L D L lipids by hepatoeyte cultures. Simultaneously with the reduction in lipid secretion, we found a reduction in the incorporation of [~4C]aectate into total free and esterified cholesterol as well as other lipids (Fig. 4). Since this is a steady state measurement, this result suggest that probu¢ol decrease lipid synthesis, in accordance with the results of Miettinen [12] and Li et al. [14], or increase their catabolism, as suggested by Tawara et al. [131 for cholesterol. Tawara et al. [131 have shown that probucol increases the liver synthesis of bile acids only in cholesterol fed mice whereas in fasting mice there is an inhibition in the synthesis of cholesterol. Any of these effects could deplete the intracellular pool of cholesterol, and thus decrease its availability for lipoprotein symliesis. We found a decrease in the incorporation of [taC]acetate into cellular cholesterol and cholesterol esters (Fig. 4), the storage form of cholesterol, which could be compatible with this mechanism. However, with the present data it is not d e a r if probueol affects cholesterol synthesis or catabolism, or both. Balasubramaniam et al. fill showed that during the initial reduction of plasma cholesterol in the rat, probucol did not decreased the activities of hepatic H M G - C o A reductase or 7a-mono-oxygenase. Nevertheless, the available dat~ can not discard the possibility of a direct effect of probucol on lipoprotein assembly [16] or apoprotein synthesis [11], as others have suggested. Probuool decreased the cellular content of ll4Clacetate labeled cholesterol esters but increased that of [laC]oleate-labeled esters, although in both cases their secretion was markedly reduced (Figs. 4C vs. 5C). These results suggest that probucol affects the synthesis of cholesterol molecule without affecting its esterification. The accumulation of tdacylglycerols within liver cells
299 we observed (Figs. 1-31 can resuh from a decreased secretion of triacylglycerol-rich lipoproteins (e.g., V L D L ; Figs. 4A and 5A) o r / a n d increase¢ triacylglycerol synthesis. Although G P D activity is an indirect measure of glycerolipid synthesis [38], the increase in its activity by probucol treatment suggest that glycerolipid synthesis could be increased which is consistent with the increased content of cell tdacylglycerols seen in various instances (Figs. 3, 4A and 5A). An increase in fa,ty acid synthesis or an inhibition in their oxidation produced b y probucoI ean also lead to an increase in the cytoplasmic pool of triacylglycerols. However. the data of Schladt et al. [40] show that probucol has no effect on peroxisomal/~'-oxidation, and in our cultures there was no consistent effect of probucol on the incorporation of [taC]acetate into fatty acids (data nol shown) or on malic Enzyme activity (see Results). It remains to be investigated if probucol has any effect on mitochondriai a-oxidation. Further, it has been demonstrated that intracellular cholesterol esters are necessary for V L D L secretion and that a reduction of their pool by lovastatin decreases also triacylglycerol secretion and promotes its accumulation in the liver, even if triacylglycerol synthesis is not affected [391. Thus, since one of the main effects of probucol we have found is a depletion of cellular free and esterifiEd cholesterol, this would suggest that triacylglyeerol accumulation is mahfly the result of reduced V L D L secretion, although a contribution by other pathways should be considered. U n d e r the electron mieroseope we observed few signs of cell damage resulting from probucol exposure for 2 weeks. The most conspicuous morphologic alteration was the accumulation of intraeytoplasmic lipid droplets that are constituted mainly by triacylglyeerols (Fig. 3; [41]). W e o~served these droplets after 4 - 5 days of treatment. This latter effect has not been reported before in intact animals. In summary, the results obtained in this work show that long-term exposure of ctfi:ured hepatoeytes to pharmacological concent~,,ions of probucol decreases the secretion of various lipids, specially free and esterifled cholesterol. This effect could Contribute to the in rive observed reduction of plasma cholesterol and other lipids induced by probucol treatment. Further experiments are necessary to determine the aetailed mechanism by which probucol reduces lipid secretion by the hepatocytes. However, our work suggests that the 3T3hepatocyte culture is a useful model to unravel the role of the liver in the hypolipidemic effect of probucol and possibly other drugs.
Acknowledgements WE are indebted to Asceneirn Hernfindez and Ma. de Lourdes L 6 p e z for skillful technical assistance and photographic artwork and to Ignacio Cruz for drawing
artwork. W e also thank Dr. Walid Kuri-Harcuch for making helpful suggestions during the development of the research and to Drs. Ruben L6pez-Revilla and Gabriel Guarneros for the critical review of ~,be manuscript. F.M.V. was supported by a scholarship from the Consejo N a t i o n a l de Ciencia y Teenohigla (CONACYT>, Mrxico, during part of this work. This research was partially supported by C O N A C Y T grant P219CCOL882377. References 1 Kalams, Z.. Dacquisto. M. and Kornetl, G.S. (1971} Curt, Ther. Res. 12. 6q2 694. 2 Kita, T, Nagano, Y.. Yokode. M., Ishii, K., Kume, K,, Oo.shlma, A., Yoshida, H. and Kawai, C. (19871 Pr~. Natl. A~d. g~i, USA 84. 5928-5931. 3 Yamamoto, A., Takaichl. $., Hard. H.. Nishikawa, O., Yokoyama, S., Yamamura. T. and Yamagu¢hi. T. (i9861 Athero~clerosis 62, 209-217. 4 Yamamoto, A, Matsuzawa, Y., Kishino. B,, Hayashi. R., Hirobe, K. and Kikkawa. T. 11983)Atherosclerosis 48, 157-161. 5 Yamamoto, A., Matsu~awa. Y.. Yokoyama. $.. Funahashi, T. Yamarnura. T and Kasluno, B. (19861 Ant J. Cardiol. 57. 29H35H. 6 lllinworth, D.R. [19881Clin. Chem. 34/8(B1. BI23-B132. 7 Ke-sAnierni, Y.A. and Grundy, S.M. (19841 J. Lipid Res. 25. 780-79G. 8 Levy,.RI. (19881 Hospital Practice 23($upl II, 14-21. 9 Strandberg. T.E, Vanhanen, H. and Mieninen, T.A. [19881 Gen. Pharmacol. 19. 317-320. 10 Steinberg. D. (1986) Am J. Cardiol. 57,16H 21H. 11 Balasuhramarliam, S,. Baths. D,M, and Siroons. L.A. (19811 Clin, Sci. 61,615-619. 12 Mieltinen, T.A. (19721 Atherosclero~is 36, 16~-176 13 Taw~a. K.. Tomikawa, M. and Abiko, Y. (1986l Japan J. Pharmacoh 40, 123-133. 14 Li, J.R, Hol~sL R.J. and Kntlke. B.A. 11980) Atherosclerc~is 36. 559-565 15 Hallori, M.. Tsuda, K., Taminato, T., Nishi, S.. Fuiila, J.. Tsuji. K., Kur~e, T.. Koh. G.. Seinn, Y and Imufa, H !19871 Cu~. Ther. Reb, 42, 967-973. 16 Naru~wi~, M., Carew, T.E. Pittman, R.C.. Witztum. J.l_ and Steinberg, D. (19841J. Lipid R~. 25, 1206 1213. 17 Trezea, E. Reran, P., Berini, F.. FumagallL R. and Catapano. A.L. (19841 Alherosclerosis 52, 309-316, 18 Mc Lean, L.R. ~ d Hagaman, K.A. (19881 Bioehim. Blophys. Acla 959, 201 205, ~9 Kun-l-!~rcuch. W. and Mendoza-Figueroa, T, fl9891 Differentialion 41. 148-157. 20 Mendo~-Figueroa. "r., Hernfindez, A.. Lrpez, M.L. and KudHarcuch, W. (19881Toxi~logy 52. 273-286. 21 Mendo~-Figueroa, T., Hernfindez, A. and l~pez, M.L. (1989) J. TnxicoL Env. Health 26, 293-308~ 22 Forle. T.M. (1984) Annu~ Rev, Physiol. 46, 403 410. 23 Berry. M N~ and Friend. D.S. (1969) J. Ceil Biol. 43. 506-520. 24 Dater, C. F©hlmann. M. and Debey. P. (1982} Anal. Biochem. 122. 119-123. 25 Todaro. G.J, and Green. H. (1933l J. Cell Biol. 17. 299-313. 26 Macpherson. I. and Bryden, A. ~197I) Exp. Cell. Res. 69. 240-24I. 27 Santone. K.S., Krc~chel. D.M. and Powis. G. (19861 Biochem. Pharmacol. 35,1287-1292. 28 Isom. H., Secott, T, Georgoff, 1.. Woodworth. C. and Mummaw, J. (1985~ Prec. Natl. Acad. Scl. USA 82. 3252-3256.
300 29 Isora, H . Georgoff. I., Salditt-Georgieff, M. and Da~ell, JE, (]987) J. Cell Biol. 105. 2877-28185. 30 Daehet, C . Jacotot, B. and Bux:orf. J . C (1085) Alherosclerosis 58. 261-268. 31 Chain. B.E. and Knowleg, BR. (19761 J, Lipid Res. 17. 176-181. 32 Kates. M. (19721 in Laboratory Techniques in Biochemistry and Molecular Biology. Vok 3 (Work, T.S. and Work. E,, eds.), pp. 502-59. American Elsevier, New York, 33 Kuri-Ha~ueh. W. and Green. H. !1977) J. Biol. Chem. 252. 2]58-2160. 34 Wise, E M , and Ball, E.G, (19641 PrC¢, Natl. Acad. Sci. USA 52, 1255-1263. 35 Kosak. L.P. and 5¢nsen. J.T. (19741 J. Biol Chem. 249, 7775 7781.
36 Kuri-Hareueh, W., Wise. L.S. and Green. ft. (19781 Cell 4. 53-59. 37 Lowry. O.H.. Rosenbmg. N,J.. Farr. A.L. and Randal, R J, (19511 J. Biol. Chem, 193. 26-~ 275. 38 Newsholme, E,A. and Start, C. (19731 Regulation in Metabolism. pp. 293 328, John Wiley and Sons, Chiehester. 39 Khan. B.V., Fungwe. T , V . Wilcox, H.f5. and Heimberg. M. (1990~ Bioehim. Biophys. Acla 1044, 297 ~04 4O Sehlad~, I , Hartmama, R., Timras, C.. Strolin-Benedetti. M.. Dosterl, P.. Womer, W. and Oegeh. F. 0987) Biochem. Pharmacol 36, 345-351. 41 Mooney, R . A and Lane. M.D. (19817 J. Biol. Chem. 256. 1172411733.
293
Effects of probucol on lipid metabolism and secretion in long-term cultures of adult rat hepatocytes Francisco M. De La Vega * and Tomfis Mendoza-Figueroa Deparlmenl of Pha~lacoloD" and To.~icolog~'. Centro de b~cestigacidn y de Estudlos At~.,~,zadosdel IPN. Me.,ico City. D.F (Mdxtco)
(n~eived 19 September 1990) Key ,.vords: Probucol; Hepatocyte primary,cullure: Lipid secretion: Cholesterol; Phospholipid:Triacylglycerol TO study the effects of prohucul on hepatic lipid metabolism, we used adult rat hepatocytes cultared on a feeder layer of 3'I"3 cells lethally treated with mitomycin C. These cultures synthesize 'and secrete for at least 2 weeks various liplds from [MC]aeetate and IMCluleate precursors. Treatment with 20 / z g / m l of probucul for 7 and 14 days decreased the secretion of various radhilaheled lipid species to the culture medium and prnduced an intracytoplasmic acemnulation of triacylglycerul droplets. The lipids whose secretion was most decreased were free and esteriiied cholesterol (50-70% reduction). Secretion of triacylglycesols and pbusphulipids was also reduced but to a lower extent. Intracytoplasmle trlac3~lglycerols accumulated and the actlvlty of glycerol phosphate dehydrogenase, a marker enzyme of glycerolipid synthesis, also increased (35-56%). The total incorporation of both radioactive precursors into free and estecir~d chulesteml and phusphulipids was reduced 20-60%. Our data show that 2-week treatment of 3T3-hepatoc~ le cultmes with pharmacological concentrations of probuenl reduces significantly lipid secretion and s u r e s t that at least part of the in vlvo hypullpldemle effect of probuenl could be attributed to a decrease in the secretion of lipids (i.e.. llpoprutelus) by helmtoeytes. Introduction Probucol is an effective drug to treat hypercholesterolemia and to prevent both atheroselerosis and cardiovascular disease [1]. It is the only drug that prevents the progression of atherosclerotic plaques [2] apparently through inhibition of lipid storage in macrophages, an effect probably related to its antioxidant properties [31. It is also one of the few drugs effective in the treatment of homozygous familial hypercholestesolemia [4,5]. It produces a mean reduction of 12% in
* Present address: Department of Genetics and Mol~ular Biology, Centto de lnvesligaci6n y de Esludios Avanzados del I.P.N., AP 14-740, M~'xi¢0D.F, 07000, Mdxico. Abbreviations: DMSO. dimeli:yt sulro~de; dpm, disint©gralionsper minute; GPD, t-glycerol-3-phosphate dehydrogenase: HDL. highdensity lipoprolein; HMG-Co,~., 3-hydroxy-3-raethyl glutapjl ¢oenzyme A; LDL, low-density lipoprotein; S~D., slandard deviation: TIC. lain-layer chromatography; VLDL. very-low-density lipopr~ teln. Trivial and systematic names: Probu¢ol, 4,4'-(isopmpylidenm dithio)bis(2,6-di-t-bulylphenol); Tris, 2-amino-2-hydroxymelaylpropane-l,3-diol. Correspondence: T. Mend0ga-Figueroa.Department of Pharmacology and Toxicology,Ceatro de lavestiga¢i6ny de Estudio~sAvanzado~del IPN, AP 14-740, Mexico City, 07000, D.F,, M~eo.
plasma cholesterol [6], and 8-15% and 9 25% reductions in LDL and H D L cholesterol, respectively [6,7]. Despite its importance, the mechanism of the hypolipidemic effect has not been unambiguously established 18,91. In the lasting state the liver accounts for most of the lipoprutein production and removal from plasma. Plasma lipid-lowering drugs ultimately work by decreasing lipoprotein production or increasing their intravascular catabolism and removal from the bloodstream [8]. Neither in human patients nor in experimental animals is clear if probuenl affects hepatic lipid metabolism and secretion, or if its effects are mainly intravascular [10]. Whereas Balasubramauiam et el. [111 and Mieninen [12] reported a reduction in cholesterol synthesis by the liver, Tawara et aL [13] showed that this effect does not occur in hypereholesterolemic mice treated with probucol. Miettinen [12] and Tawara et al. [13] reported an increase in bile salt production and secretion by the liver, but Kesiiniemi and Grundy [7] found no effect of probucol on bile acid or cholesterol excretion and Li et al. [14] found in fact a decrease in sterol balance. Hattori et ai. [15] suggested that probucol reduces the synthesis of tdacylglyeerul rich lipoproteins by the liver. On the other hand, results of Namszewicz et al. [16] and Steinberg [10] suggested that probuenl lowers LDL levels
294 by increasing the rate of their removal from the bloodstream and argued that effects of the drug on the liver or on the intravascular production of LDL do not contribute significantly to the hypolipidemic effect. However, Trezzi et al. [17] found no effect of probucol on the fractional catabolic rate of LDL in rabbits. Other suggested effects of probucol are an increase in the catabolism of H D L j to HDL~ [15], reduction in H D L size [5], and a possible interference on lipoprotein assembly [18]. These contradictory results may be partially attributed to the use of intact animals where the measurement of direct effects of the drug or, the hepatic lipid metabolism and secretion might be obscured by the contribution of other effects on the intravascular metabolism of lipoproteins. Primary cultures of adult rat hepatocytes have provided a useful model to study hepatic effects of several drugs and chemicals on lipid metabolism without the interference of other tissues (vascular, adipose, etc. [1922]). In this paper we report the long-term effects of pharmacological concentrations of probucol on hepatic lipid metabolism and secretion using as model system a primary culture of adult rat hepatocytes seeded on a feeder layer of 3T3 cells. The 3T3-hepatocyte cultures have been shown to retain for several weeks the hepatic characteristic morphology and ultrastructure and various liver-specific functions, such as albumin secretion, basal and inducible levels of cytochrome P-450, and synthesis and secretion of lipids to the culture medium [19-21]. The results obtained show that prnbucol treatment decreases the secretion of lipids to the culture medium, mainly free and esterified cholesterol and reduces the incorporation of labeled precursors into most cellular lipids. Concurrently with these effects there is an intracytoplasmic triacylglycerol accumulation. Materials and Methods
Materials Collagenase type IV, hydrocortisone, insulin, d-biotin, Oil red O, Percoll and dimethyl sulfoxide (DMSO) were purchased from Sigma Chemical (St. Louis, M e ) . Calf serum was obtained from HyClone Laboratories (Logan, UT). Aluminum plates coated with silica gel 60 (0.2 m m thickness) for thin-layer chromatography (TLC) were obtained from E. Merck (Darmstadt, F.R.G.). [l,2-14C]Acetic acid (55 m C i / m m o l ) and [1-14C]oleic acid (57 m C i / m m o l ) were bought from New England Nuclear (Boston, MA). Mytomicin C and probucol were kind gifts from Bristol-Myers (Denver, CO) and Laboratories Lepetit de Mrxico (Mexico City), respectively. All other reagents were of analytical grade and were purchased from Sigma Chemical and J.T. Baker (Mexico City).
Cell cultures Hepatocytes were isolated from male Wistar rats
(180-200 g) by the collageaase-perfusion method of Berry and Friend [23] modified as described earlier [191. Briefly, the liver was perfused with a salt solution containing 5.5 m M glucose. 137 mM NaCI. 50 m M KCI, 0.4 mM N a 2 H P O 4, 25 mM Tris-HCl and 0.01% phenol red (pH 7.4). for 3 rain at 37°C and at a flow rate of 15 m i / m i n and again for 15 min at a rate of 7 m l / m i a with the salt solution containing 100 u n i t s / m l bf collagenase iype IV at 37°C. Cells were dispersed in 40 ml of Eagie's medium modified by Dulbecco and ViSgt supplemented with 77o calf serum, 5 t*g/ml insulin and 0.l tiM d-biotin (basal medium). The cell suspension was filtered through a nylon mesh and allowed to stand for 5 min. The supernatant was discarded, the cells were resuspended in basal medium containing 2070 Pcrcoll [24] and then ~edimented for 3 rain at 300 × g in a Beckman TJ-6 centrifuge. The cell viability, assessed by Trypan blue exclusion, was higher than 9870. The hepatocytes were inoculated ( 4 . 5 . 1 0 s / 3 5 m m dish) on a feeder layer of 3T3 edls [25] that had been lethally treated with mitomycin C [26] and inoculated 1 day earlier at 4 . 1 0 ~ cells/35 m m dish [19]. The cells were placed in a 37°C humidified incubator gassed with 90~, air/10% CO 2. 30 rain after seeding, the dishes were rinsed with serum-free medium and refed with 1 ml of basal medium supplemented with 10 ~ g / m l of hydro-. cortisone (complete medium). Untreated cultures were maintained with complete medium, whereas treated cultures were changed 24 h after cell seeding to medium containing the indicated concentration of probucol previously dissolved in DMSO. The final concentration of D M S O in the culture medium of untreated and treated cell cultures was 1~. The presence of D M S O is not detrimental to hepateeytes [27], but favors the permanence of hepatic differentiated functions, as shown by Isom et al. [28,29]. Cultures of lethally treated 3T3 cells were maintained with complete medium containing 1% D M S O with or without probucol to assess the contribution of the feeder layer to the observed effects. Culture medium was changed daily. The concentration of probucol used throughout this study was widlin the range observed in the plasma of treated subjects [30]. Also, we selected the exposure tirn,~ of 1 and 2 weeks because pharmacological effects of probucol are seen in the rat after treatment for 1 week [11].
Cell morphology and ultrastructure The cultures were examined and photographed with an Olympus I M T inverted photomicroscope equipped with a PM-10AD photographic system. For transmission electron microscopy, cells were processed as described [20] and were observed with a ,leol-2000EX electron microscope at 80 kV.
Incorporation of [taC]acetate and [14C]oleate into cellular and secreted lipids At the indicated culture times. 0.2 / t C i / m l
of
295 [l~C]acetic acid were added to the cultures. After 24 h incubation, the medium was removed and the cells were washed, scraped and disrupted by sonication. For studies with [laC]oleic acid, the cultures were incubated for 2 h with 0.2 ~tCi/ml of the radioactive precursor. Then. the cultures were rinsed twice with 2 ml of serum-free medium and refed with 0.5 ml of fresh non-radioactive medium. After 24 h incubation the medium was remo~ed and the cells processed as before. Lipids from both culture media (secreted lipi.~) and cells (cellular lipids) were extracted as described by Chum and Knowles [311 and analyzed by TLC.
Thin-layer chrorr~atography analysi.s of #ipids ann triatylglycerol cell content The extracted lipids were mixed with rat serum lipids as carriers, evaporated under a nitrogen stream and dissolved in 30 ~tl of chloroform. Samples of 15/tl were analyzed by TLC using a mobile phase of heptane/ diisopropyl ether/acetic acid ( 6 0 : 4 0 : 4 , v / v ) [32]. The radioactive lipids that c~,"r.~grzted with purified standards of triacylglycerols, phospholipids, cholesterol and cholesterol esters, were determined by liquid scintillation in a Packard Tri-Carb 460 counter (Downers Grove, lL) using Aquasol scintillation cocktail INew England Nuclear, Boston, MA). Triaeylglycerol content in the cells was determined by staining the dishes with Oil red O, extracting the dye retained with 1 ml isopropanol and measuring its absorbance at 510 nm. This procedure stains specifically triacylglycerols without interference from phospholipids (Ramirez-Zacaralas, Castre-Muhozledo and KuriHarcuch, unpublished). Triacylglycerol concentration was determined from a standard calibration curve [20].
Enzyme activities Cell extracts used to determine enzyme activities
were prepared as described by Kuri-Harcuch and Green [33]. Malic enzyme (L-malate:NADP oxidoreductase (descarboxylating); EC 1.1.1A0) and L-glycerol-3-phosphate dehydrogeaase (EC 1.1.].8) activities were determined as described elsewhere [34--36]. Protein was determined by the method of Lowry et al. [37].
Data presentatwn and analysis Unless otherwise indicated, results are expressed as the mean _+ S.D. of determinations from three or four culture dishes in a typical experiment; similar results were obtained in three independent experiments. As protein content was very similar between dishes and it did not change during culture for 2 weeks, neither with treatment, all values are expressed per dish. Statistical comparison between groups was performed with the unpaired Studenfs t-test after a randomized analysis of variance revealed significance ( P < 0.05). Results
Alteration ef heparoQ.i ? morpholog)' and ultrastructure by probuvol Control cultures showed under both optical and electron microscope their typical morphology and ultrastructure during culture time, as reported before (Figs. la, c, 2a: [19]). Hepatocytes treated for 7 days with probucol accumulated small lipid droplets homogeneously distributed within the cytoplasm (Fig. l b ) which probably fused arising large droplets after 14 days of exposure {Fig. ld). Electron micrographs revealed that the lipid droplets were not surrounded by membranes (Fig. 2b). and that some times they were associated with large glycogen areas (Fig. 2b, c). Large irregular milcchondria, 3ome autophagic vesicles and abundant microfilament bundles dispersed throughout the cytoplasm were also seen in the treated cells (Fig. 2e, d). All other cellular organelles presented normal morphology.
Fig. l. Effect of probucol on hepatocylc morphology,HepatocyIes were cultured in complele medlum supplememed with 1% DM$O for 7 (a) or 14 days (c) or in complete medium supplemented with 1¢,l DMSO plu.~ 20 ,~g/mI of probucol for 7 (b) or 14 days (d). Phase contrast optics. Bar, 50 pm.
296
Fig. 2. Ultrastmcture of untreated and probucol-treated hepatocytes. Hepalocytes were cultured for 7 days and proe~sed for transmission elc~ctr~ microscopy as de~ribed in Methods (a) Unlreated hepatocyte: N, nucleus; n, nt,cleolus; m, mit~h~dria. (b-d) Hepatocytes treated with 20 pg/ml of probucol: L. lipid droplet; g, glycogenarea; a. autophagic vacuole~ m, mit~hondria; ern, extracellular matrix; arl,owhead,microfilamcnt bundle; arrow, glycogengranules. Bar, a-c, 1/lm; d. 5COnm.
Tria~TIg(~'cerol accumulation in hepato¢yles treated with probucol Triacylglycerol cell content increased in the cultures treated for 14 days in a dose dependent manner (Fig. 3). The increase was significant at 10 and 20 # g / m l of probucol. Triaeyiglycerol accumulation was also timedependent as was evident from the daily morphological observations. Effect of probucol on lipid synthesis and secretion To study the effect of probucol treatment on de novo lipid synthesis and secretion by the hepatocytes, we determined the incorporation of [t4C]acetate and [14C]oleate into cellular, secreted and total (cellular plus secreted) lipids The experiments with [14C}acetate were 24 h pulse labeling experiments whereas the experiments with [laC]oleate were 2 h pulse-24 It chase experiments. The long [14C]acetate pulse was necessary because short pulses resulted in little incorporation (data
not shown). Additionally, with this protocol a steady state of labeling might be obtained, so the amount of label incorporated into a particular lipid species could be an estimate of the amount of the lipid. Ptobucol reduced the incorporation of [14C]acetate into total lipids from 82.6 4- 7.9.103 to 43.5 4- 4 . 3 . 1 0 J d p m / d i s h (mean 5: S.D., n = 4; 47% reduction, P < 0.001) and from 92.6 4- 2 8 . 7 . 1 0 ~ to 57.3 5:30.2.103 d p m / d i s h ( P > 0.05, not significant), after 7 and 14 days of treatment, respectively. TLC analyses of [14C]acetate-labeled lipid species (Fig. 4) revealed that after 7 days of treatment the total incorporation o[ [14C]acetate into free and esterified cholesterol was strongly reduced (Fig. 4C, D). On the other hand, radiolabeled triacylglycerols was the only species whose cellular content increased though its total amount was unchanged (Fig. 4A). The cell content of all other lipids was reduced leading to a net reduction in total incorporation (Fig. 4 B - D ) . Remarkably, the secretion of free
297
45.
~"
*
25
15
0.01
0,1
1,0
Probu¢ol
10
100
(/~g/ml)
Fig. 3. Dose dependent cellular triacylglycerol accumulation reduced by probucol treatment. Hepat~ytes were cultured for 14 days in complete medium supplemented with 1% DMSO and the indicated concentrations of probueol, Culture dishes were fixed and stained with Oil red O. The stain was extracted with i~propanol and its absorbanee determined at 510 am. Triacylglycerol ~ntent (/~g/dish) was determined from a standard calibration cu~e. Poinls and bars represent mean± range of two culture dishes in a typical experimenl. Asterisks indicate significant difference wilh respect to control values ( P < 0.~5 ~.
a n d esterified cholesterol to c u l t u r e m e d i u m w a s red u c e d by 5 0 - 7 0 % (Fig. 4C, D). T h e secretion of labeled triglycerides a n d p h o s p h o l i p i d 3 w a s also r e d u c e d b u t to a lower e x t e n t (Fig. 4A, B). A f t e r 14 days e x p o s u r e , the secretion a n d total i n c o r p o r a t i o n i n t o all labeled lipids w a s n o t a b l y decreased (Fig. 4 E - H ) , b e i n g free a n d esterified cholesterol t h e m o s t decreased species (e.g.. 9 0 ~ a n d 75% r e d u c t i o n of secretion, respectively; Fig.
i"°!
4 G . H). T h e cellular c o n t e n t s of radiolabelcd p h o s p h o lipids a n d cholesterol esters were also reduced (Fig. a F , G ) , whereas t h o s e of cholesterol a n d triacylglyccrols were u n c h a n g e d at this t r e a t m e n t time (Fig. 4 E , H). W h e n we a n a l y z e d the i n c o r p o r a t i o n of [V~C]oleate i n t o the various h p i d species, the results were c o m p a r a ble to those o b t a i n e d with []4C]acetate (Fig. 5). T h e r e were s o m e v a r i a i i o n s that c o u l d be d u e to differences in radio!abeled p r e c u r s o r pools a n d in e x p e r i m e n t a l design. A l t h o u g h the secretion of labekxt cholesterol esters w a s decreased after 7 days t r e a t m e n t , their cell c o n t e n t increased w h e r e a s total i n c o r p o r a t i o n w a s u n c h a n g e d IFig. 5CI. However. as in e x p e r i m e n t s w i t h [laC]acetate the secretion of triacylglycerols decreased while their cellular c o n t e n t increased (Fig. 5 A L T h e i n c o r p o r a t i o n of [~aC]oleate i n t o total, cellular a n d secreted p h o s p h o lipids was reduced significantly IFig. 5B). After 14 days of e x p o s u r e the secretion of all species also decreased (Fig. 5 D - F L Nevertheles,~, at this t i m e the cell c o n t e n t of labeled triacylglycerols a n d cholesterol esters d i d n o t increased significantly (Fig. 5D. F). T h e i n c o r p o r a t i o n or [laC]oleate i n t o cellular a n d total p h o s p h o l i p i d s was similar to t h a t observed a f t e r 7 days of t r e a t m e n t IFig. 5E). T h e c o n t r i b u t i o n of the f~,~eder layers to these effects was insignificant, as studies w i t h cultures of 3T3 cells revealed. T h e i n c o r p o r a t i o n of labeled precursors i n t o lipids of 3T3 cells was I - 5 % of t h a t f o u n d in 3T3h e p a t o c y t e cultures, e x c e p t for p h o s p h o l i p i d s w h e r e the i n c o r p o r a t i o n r e a c h e d a b o u t 10%. F u r t h e r m o r e . we detected n o c h a n g e s in radiotabcled lipid synthesis a n d
!
,
Fig, 4. Eff~I of prohucol on [14Cla~tale incorporation into hepatocyte cellular and ~ecreted lipids. Hcpat~yt~ were cultured for "/IA D) or 14 days (E H) in medium with I% DMSO (unfilled bars) or medium with I% DMSO containing 20/~g/ml probucol (filled barsL Then, cultures wen incubated with 0.2 ~Ci/ml nf [t4C]acetate for 24 h. The incorporation of [b4Clacetate into medium (secteled) and cell ]ipids was measured after extraction and separation by TLC. Total incorporation was the sum of radioactivity in medium and calls. Bars represent values of mean ± S D ( n - 4) in a typical experiment. Asterisks indicatu significant difference with respect IO control values ( * P < 0.05: * * P < 0.01 ). Note that scales for panels C and D ale different.
298 T ~[lfl~
[
pfio~Ollpl~
[ Choeserat~ors
I
I
' nlil n,l,'i0 llll Fig. 5. Effect of probucol on [14C]oleate incorporation into hepatocyte cellular and secreted lipids. Hepato~tes were cultured for 7 (A-C) or 14 days (D-F) in medium with 1~ DM$O (unfilled bars) or medium with 1~ DMSO and 20 pg/ml probucol (filled bars). Then, cultures were incubated with 0.2 #Ci/ml of 114Cloleatefor 2 h. finsed. feted with fresh medium and incubated further for 24 h. The incorporation of [14C]oleate into medium (secreted) and cell lipids wa~ measured after extraction and separation by TLC. Total incorporation was the sum of radioactivity in medium and cells. Bars represent values of mean_+S.D.(n 4) in a typical experiment, Asterisks indicate significant difference with respect to control values ( * P < 0.05; • * P < 0.01). Note that scales for panels C and F are different. secretion by 3T3 cells due to probucol treatment (data not shown).
Effect of probucol on the activity of two lipogenic enzymes We determined the activities of malic enzyme and L-glycerol-3-phosphate dehydrogenase (GPD), markers of fatty acid and glycerolipid synthesis, respectively [38]. The exposure to probucoi increased G P D activity significantly from 77.6 + 18.6 to 121 + 16 (mean 4- S.D., P < 0.05, n = 3) and from 103.1 ++ 13.4 to 140.5 + 26.4 ( P < 0.05, n = 3) nmol N A D • mg protein train 1, at 7 and 14 days of treatment, respectively. On the other hand, the effect of probucol on malic enzyme activity was variable. After 7 days exposure there was a slight decrease, whereas at 14 days there was an increase, although not statistically significant (data not shown). Discussion Our results show that the main effect of probucol on cultured rat hepatocytes was a reduction in the secretion of lipids, as evidenced by the decrease in the amount of radiolabeled lipids released to the culture medium by the treated cultures (Figs. 4 and 5). The secretion of free and esterified cholesterol was strongly decreased, although triacylglycerol and phospholipid secretion was also diminished to a lesser extent (Figs. 4
and 5). At longer treatment times the effects "¢~e~e usually more potent. Our results are consistent with the in viva observed 12 20% reduction of plasma cholesterol [1,30] and the lower reduction of plasma triacylglycerols [1,15,30] produced by probucol, though tiic decrease in lipid secretion by hepatocytes was greater. This is not surprising since hepatecyte cultures may resemble more closely the fasting liver. Due to the limited supply of extracellular lipids there is an increased synthesis of lipids and hence the effect of probucol might be overemphasized. These results suggest that probucol could decrease the secretion of cholesterol-bearing lipoproteins ( H D L or V L D [ ) or their cholesterol content [4,15]. It is not clear which lipoprotein species could be affected by probucol since the secretion of triacylglycerols and phospholipids decreased also. Khan et al. [39] have shown that reduction in the intracellular pool of esterifled cholesterol, an effect of probucol on cultured hepatocytes (Fig. 4C, G), leads to a decrease in V L D L secretion. Further work is necessary to determine directly the effects of probuool on the secretion of H D L and V L D L lipids by hepatoeyte cultures. Simultaneously with the reduction in lipid secretion, we found a reduction in the incorporation of [~4C]aectate into total free and esterified cholesterol as well as other lipids (Fig. 4). Since this is a steady state measurement, this result suggest that probu¢ol decrease lipid synthesis, in accordance with the results of Miettinen [12] and Li et al. [14], or increase their catabolism, as suggested by Tawara et al. [131 for cholesterol. Tawara et al. [131 have shown that probucol increases the liver synthesis of bile acids only in cholesterol fed mice whereas in fasting mice there is an inhibition in the synthesis of cholesterol. Any of these effects could deplete the intracellular pool of cholesterol, and thus decrease its availability for lipoprotein symliesis. We found a decrease in the incorporation of [taC]acetate into cellular cholesterol and cholesterol esters (Fig. 4), the storage form of cholesterol, which could be compatible with this mechanism. However, with the present data it is not d e a r if probueol affects cholesterol synthesis or catabolism, or both. Balasubramaniam et al. fill showed that during the initial reduction of plasma cholesterol in the rat, probucol did not decreased the activities of hepatic H M G - C o A reductase or 7a-mono-oxygenase. Nevertheless, the available dat~ can not discard the possibility of a direct effect of probucol on lipoprotein assembly [16] or apoprotein synthesis [11], as others have suggested. Probuool decreased the cellular content of ll4Clacetate labeled cholesterol esters but increased that of [laC]oleate-labeled esters, although in both cases their secretion was markedly reduced (Figs. 4C vs. 5C). These results suggest that probucol affects the synthesis of cholesterol molecule without affecting its esterification. The accumulation of tdacylglycerols within liver cells
299 we observed (Figs. 1-31 can resuh from a decreased secretion of triacylglycerol-rich lipoproteins (e.g., V L D L ; Figs. 4A and 5A) o r / a n d increase¢ triacylglycerol synthesis. Although G P D activity is an indirect measure of glycerolipid synthesis [38], the increase in its activity by probucol treatment suggest that glycerolipid synthesis could be increased which is consistent with the increased content of cell tdacylglycerols seen in various instances (Figs. 3, 4A and 5A). An increase in fa,ty acid synthesis or an inhibition in their oxidation produced b y probucoI ean also lead to an increase in the cytoplasmic pool of triacylglycerols. However. the data of Schladt et al. [40] show that probucol has no effect on peroxisomal/~'-oxidation, and in our cultures there was no consistent effect of probucol on the incorporation of [taC]acetate into fatty acids (data nol shown) or on malic Enzyme activity (see Results). It remains to be investigated if probucol has any effect on mitochondriai a-oxidation. Further, it has been demonstrated that intracellular cholesterol esters are necessary for V L D L secretion and that a reduction of their pool by lovastatin decreases also triacylglycerol secretion and promotes its accumulation in the liver, even if triacylglycerol synthesis is not affected [391. Thus, since one of the main effects of probucol we have found is a depletion of cellular free and esterifiEd cholesterol, this would suggest that triacylglyeerol accumulation is mahfly the result of reduced V L D L secretion, although a contribution by other pathways should be considered. U n d e r the electron mieroseope we observed few signs of cell damage resulting from probucol exposure for 2 weeks. The most conspicuous morphologic alteration was the accumulation of intraeytoplasmic lipid droplets that are constituted mainly by triacylglyeerols (Fig. 3; [41]). W e o~served these droplets after 4 - 5 days of treatment. This latter effect has not been reported before in intact animals. In summary, the results obtained in this work show that long-term exposure of ctfi:ured hepatoeytes to pharmacological concent~,,ions of probucol decreases the secretion of various lipids, specially free and esterifled cholesterol. This effect could Contribute to the in rive observed reduction of plasma cholesterol and other lipids induced by probucol treatment. Further experiments are necessary to determine the aetailed mechanism by which probucol reduces lipid secretion by the hepatocytes. However, our work suggests that the 3T3hepatocyte culture is a useful model to unravel the role of the liver in the hypolipidemic effect of probucol and possibly other drugs.
Acknowledgements WE are indebted to Asceneirn Hernfindez and Ma. de Lourdes L 6 p e z for skillful technical assistance and photographic artwork and to Ignacio Cruz for drawing
artwork. W e also thank Dr. Walid Kuri-Harcuch for making helpful suggestions during the development of the research and to Drs. Ruben L6pez-Revilla and Gabriel Guarneros for the critical review of ~,be manuscript. F.M.V. was supported by a scholarship from the Consejo N a t i o n a l de Ciencia y Teenohigla (CONACYT>, Mrxico, during part of this work. This research was partially supported by C O N A C Y T grant P219CCOL882377. References 1 Kalams, Z.. Dacquisto. M. and Kornetl, G.S. (1971} Curt, Ther. Res. 12. 6q2 694. 2 Kita, T, Nagano, Y.. Yokode. M., Ishii, K., Kume, K,, Oo.shlma, A., Yoshida, H. and Kawai, C. (19871 Pr~. Natl. A~d. g~i, USA 84. 5928-5931. 3 Yamamoto, A., Takaichl. $., Hard. H.. Nishikawa, O., Yokoyama, S., Yamamura. T. and Yamagu¢hi. T. (i9861 Athero~clerosis 62, 209-217. 4 Yamamoto, A, Matsuzawa, Y., Kishino. B,, Hayashi. R., Hirobe, K. and Kikkawa. T. 11983)Atherosclerosis 48, 157-161. 5 Yamamoto, A., Matsu~awa. Y.. Yokoyama. $.. Funahashi, T. Yamarnura. T and Kasluno, B. (19861 Ant J. Cardiol. 57. 29H35H. 6 lllinworth, D.R. [19881Clin. Chem. 34/8(B1. BI23-B132. 7 Ke-sAnierni, Y.A. and Grundy, S.M. (19841 J. Lipid Res. 25. 780-79G. 8 Levy,.RI. (19881 Hospital Practice 23($upl II, 14-21. 9 Strandberg. T.E, Vanhanen, H. and Mieninen, T.A. [19881 Gen. Pharmacol. 19. 317-320. 10 Steinberg. D. (1986) Am J. Cardiol. 57,16H 21H. 11 Balasuhramarliam, S,. Baths. D,M, and Siroons. L.A. (19811 Clin, Sci. 61,615-619. 12 Mieltinen, T.A. (19721 Atherosclero~is 36, 16~-176 13 Taw~a. K.. Tomikawa, M. and Abiko, Y. (1986l Japan J. Pharmacoh 40, 123-133. 14 Li, J.R, Hol~sL R.J. and Kntlke. B.A. 11980) Atherosclerc~is 36. 559-565 15 Hallori, M.. Tsuda, K., Taminato, T., Nishi, S.. Fuiila, J.. Tsuji. K., Kur~e, T.. Koh. G.. Seinn, Y and Imufa, H !19871 Cu~. Ther. Reb, 42, 967-973. 16 Naru~wi~, M., Carew, T.E. Pittman, R.C.. Witztum. J.l_ and Steinberg, D. (19841J. Lipid R~. 25, 1206 1213. 17 Trezea, E. Reran, P., Berini, F.. FumagallL R. and Catapano. A.L. (19841 Alherosclerosis 52, 309-316, 18 Mc Lean, L.R. ~ d Hagaman, K.A. (19881 Bioehim. Blophys. Acla 959, 201 205, ~9 Kun-l-!~rcuch. W. and Mendoza-Figueroa, T, fl9891 Differentialion 41. 148-157. 20 Mendo~-Figueroa. "r., Hernfindez, A.. Lrpez, M.L. and KudHarcuch, W. (19881Toxi~logy 52. 273-286. 21 Mendo~-Figueroa, T., Hernfindez, A. and l~pez, M.L. (1989) J. TnxicoL Env. Health 26, 293-308~ 22 Forle. T.M. (1984) Annu~ Rev, Physiol. 46, 403 410. 23 Berry. M N~ and Friend. D.S. (1969) J. Ceil Biol. 43. 506-520. 24 Dater, C. F©hlmann. M. and Debey. P. (1982} Anal. Biochem. 122. 119-123. 25 Todaro. G.J, and Green. H. (1933l J. Cell Biol. 17. 299-313. 26 Macpherson. I. and Bryden, A. ~197I) Exp. Cell. Res. 69. 240-24I. 27 Santone. K.S., Krc~chel. D.M. and Powis. G. (19861 Biochem. Pharmacol. 35,1287-1292. 28 Isom. H., Secott, T, Georgoff, 1.. Woodworth. C. and Mummaw, J. (1985~ Prec. Natl. Acad. Scl. USA 82. 3252-3256.
300 29 Isora, H . Georgoff. I., Salditt-Georgieff, M. and Da~ell, JE, (]987) J. Cell Biol. 105. 2877-28185. 30 Daehet, C . Jacotot, B. and Bux:orf. J . C (1085) Alherosclerosis 58. 261-268. 31 Chain. B.E. and Knowleg, BR. (19761 J, Lipid Res. 17. 176-181. 32 Kates. M. (19721 in Laboratory Techniques in Biochemistry and Molecular Biology. Vok 3 (Work, T.S. and Work. E,, eds.), pp. 502-59. American Elsevier, New York, 33 Kuri-Ha~ueh. W. and Green. H. !1977) J. Biol. Chem. 252. 2]58-2160. 34 Wise, E M , and Ball, E.G, (19641 PrC¢, Natl. Acad. Sci. USA 52, 1255-1263. 35 Kosak. L.P. and 5¢nsen. J.T. (19741 J. Biol Chem. 249, 7775 7781.
36 Kuri-Hareueh, W., Wise. L.S. and Green. ft. (19781 Cell 4. 53-59. 37 Lowry. O.H.. Rosenbmg. N,J.. Farr. A.L. and Randal, R J, (19511 J. Biol. Chem, 193. 26-~ 275. 38 Newsholme, E,A. and Start, C. (19731 Regulation in Metabolism. pp. 293 328, John Wiley and Sons, Chiehester. 39 Khan. B.V., Fungwe. T , V . Wilcox, H.f5. and Heimberg. M. (1990~ Bioehim. Biophys. Acla 1044, 297 ~04 4O Sehlad~, I , Hartmama, R., Timras, C.. Strolin-Benedetti. M.. Dosterl, P.. Womer, W. and Oegeh. F. 0987) Biochem. Pharmacol 36, 345-351. 41 Mooney, R . A and Lane. M.D. (19817 J. Biol. Chem. 256. 1172411733.
