Alcohol. Vol. 2, pp. 157-161, 1985. ©Ankho InternationalInc. Printed in the U.S.A.
0741-8329/85 $3.00 + .00
Effect of Ethanol on Fatty Acid Metabolism in Cultured Hepatocytes: Dependency on Incubation Time and Fatty Acid Concentration N. G R U N N E T , *
F. J E N S E N , *
J. K O N D R U P t
A N D J. D I C H *
*Department o f Biochemistry A, The P a n u m Institute, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark tDepartment o f Medicine A, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen O, Denmark
GRUNNET, N., F. JENSEN, J. KONDRUP AND J. DICH. Effect of ethanol on fatty acid metabolism in cultured hepatocytes: Dependency on incubation time and fatty acid concentration. ALCOHOL 2(1) 157-161, 1985.--In a previous report it was shown that ethanol increases the rate of accumulation of triacylglycerol by 90% in hepatocytes in primary culture. This represents the f'trst known suitable model for in vitro studies of the ethanol-induced fatty liver. The biochemical alterations causing this accumulation of triacylglycerol remain to be elucidated, however. In the present report it is shown that (1) the effect of ethanol exhibits a time lag of 6-9 hours (2) the increment in the content of triacylglycerolcaused by ethanol is increased by increased concentrations of fatty acids (3) the fatty acid uptake is not affected by ethanol (4) fatty acid synthesis is inhibited 20% by ethanol (5) the contents of diacyiglycerol and phospholipids are not affected by ethanol (6) addition of ethanol increases the cytosolic and mitochondriai redox levels. It is concluded that ethanol is likely to exert its effect on the accumulation of triacylglycerolby redistributing fatty acids between oxidation and triacylglycerol synthesis and/or between storage and secretion of triacylglycerol. Ethanol
IN a previous report  it was shown that ethanol increases the rate o f accumulation of triacyiglycerol by about 90% in hepatocytes in primary culture. In short-term studies with perfused liver, or freshly isolated hepatocytes, no, or only insignificant, effects o f ethanol on the synthesis of triacylglycerol have been demonstrated [5,14]. In order to discover a possible time lag of the effect of ethanol, a timestudy of the effect of ethanol on the accumulation of triacylglycerol was undertaken. In addition, the effect of ethanol was investigated at different fatty acid concentrations. In order to further elucidate the effect of ethanol, we also measured fatty acid uptake, fatty acid synthesis, the content of phospholipid and diacyiglycerol and the production of ketone bodies. The effect of ethanol on lactate/pyrur a t e and 3-hydroxybutyrate/acetoacetate ratios are also reported. Possible artifacts in measuring the secretion of VLDL-triacylglyceroi were examined; rates of secretion of VLDL-triacylglyceroi are presented.
amethasone (10 -e M), and incubated at 37°C with air/CO2 (19:l). On the next morning (day 2, zero time), the medium was replaced with 3 ml o f serum- and hormone-free medium containing oleate (complexed to fatty acid free albumin in a molar ratio of 3.5) and ethanol (50 mM) as indicated. Hates containing ethanol were placed in a separate air-tight chamber together with a beaker containing 500 ml of 50 mM ethanol. At the time-points indicated, the medium was sucked off, and a sample was taken for determination of free fatty acids  and VLDL-triacyiglycerol. Krebs-Henseleit buffer was added to the plates, the cells were disrupted by sonication (40 W, l0 sec), and aliquots were taken for estimation of D N A  and total acylglycerol . F o r the separate determination o f di- and triacylglycerol and phospholipids, an aliquot o f the cell homogenate was extracted with chloroform:methanol , phospholipids were determined as total phosphorus in the lipid extract , and mono-, di-, and triacylglycerol were quantitated separately after t.l.c. [l l]. VLDL-triacylglycerol was isolated by centrifugation o f the medium which had a density of 1.006 g/ml for 187,000 g for 2 l hr. The floating VLDL-fraction was sucked off and subjected to the extraction and t.l.c, procedure also used for the cells. Lactate, pyruvate, acetoacetate and 3-hydroxybutyrate
METHOD Primary hepatocyte cultures were prepared as described earlier . In short, hepatocytes prepared from 16 hr starved female Wistar rats were plated on collagen-coated Petridishes with a modified HI-WO/BA medium  supplemented with antibiotics, 10% horse serum, insulin (10 -~ M) and r e x -
G R U N N E T , J E N S E N , K O N D R U P A N D DICH
Z £3 "6 E ::L
~.'.'- ~-.-~>--__..~v:r_~ 0
FIG. 1. Accumulation of triacylglyceroi in hepatocytes in primary culture. Hepatocytes were incubated with or without 50 mM ethanol and oleate at the concentrations indicated. Cells corresponding to 22.3+-i.9 #,g DNA (mean-S.E.M., n=7) were incubated in 3 ml medium. The symbols represent the meanxS.E.M. The initial concentrations of oleate were: ©, • 0.2 raM; A, & 0.5 raM; [2, • i .0 raM; ~, • 1.7 mM (only 3 experiments were performed with 1.7 raM). (a): Without ethanol; (b): with ethanol.
were measured enzymatically  on neutralized, citric acid (final concentration 7%) precipitated incubations. Materials were obtained as reported previously . Results are mean+-S.E.M, and statistical significance was calculated by the t-test for paired data. RESULTS
Triacylglycerol Accumtdation, Fatty Acid Uptake and Synthesis Figure I a and b shows the accumulation of triacylglycerol as a function of time at four different concentrations of oleate in the absence (a) and presence (b) of ethanol. The amount of triacylglycerol that accumulated was dependent on the initial concentration of oleate, in the absence as well as in the presence of ethanol. It is also apparent that ethanol increased the accumulation of triacylglycerol at the later time-points. The time-dependency of the effect of ethanol is more clearly shown in Fig. 2. The difference in the content of triacyiglycerol caused by the addition of ethanol is shown as a function of time for two different concentrations of oleate. With 1 mM oleate, 6 hours were required before a statistically significant effect of ethanol was seen: with 0.2 mM oleate 9 hours were required. The time-dependency of the ethanol effect is a real increase in the effect of ethanol with time: expressed as percent of control hepatocytes, the content of triacylglycerol with ethanol present (1 mM oleate) was 108+-6% at 3 hours as compared to 124_+4% at 12 hours, respectively (mean= S.E.M., n=7). The time lag observed for the effect of ethanol may explain that a 45 min peffusion study failed to reveal any effect of ethanol on the hepatic content, or synthesis, of triacylglycerol . The effect of ethanol was also dependent on the concentration of oleate (Fig. 3); this figure also suggests that some
triacylglycerol would accumulate even in the absence of added oleate but this remains to be experimentally verified. In the peffused liver, ethanol caused diacylglycerol to accumulate . In the present study there was no accumulation of diacylglycerol in the presence of ethanol: In three experiments the mean content of diacylglyceroi in the absence and presence of ethanol was 3.8 and 3.7 #,mol/mg D N A , respectively, after 12 hours incubation with I mM oleate initially. Figure 4a and b shows the time-course of the concentration of oleate in the medium at different oleate concentrations, in the absence or presence of ethanol. The uptake of oleate, as a function o f the initial oleate concentration, is shown in Fig. 5. The figure also shows that ethanol had no effect on the rate of fatty acid uptake. Within the first three hours the uptake of oleate at l mM was 39/~mol/mg DNA: when assuming a D N A content of 2.5 mg/g wet wt.  this uptake can be recalculated to 0.55 p.mol/min per g wet wt. which is identical to the uptake found in perfused liver [ l l ] . At later time-points the uptake of oleate decreased. probably because of the decline in the concentration of oleate. The average initial content of phospholipid in the cells was 24.6/~mol/mg D N A ; after 12 hours incubation the average content was 22.2 g.mol/mg DNA, the decrease was not statistically significant and there was no difference between control and ethanol treated cells. The rate of fatty acid synthesis corresponded to 6.8 nmol C~/min per mg DNA as measured by means of incorporation of tritiated water into lipids ; ethanol inhibited the synthesis by 20% (n=2). Since there was no increase in the uptake of fatty acids or the synthesis of fatty acids or decrease in the content of phospholipids it seems safe to conclude that ethanol must exert its effect by redistributing fatty acids between fatty
FATTY LIVER IN CULTURED
~ 0 -I
6 Time (hours)
FIG. 2. Ethanol-induced increase in the content of triacylglycerol: time dependency. Same experiments as depicted in Fig. 1. The ethanol-induced increase in the content of triacylglycerol was calculated for each time-point in each experiment. The mean values_+S.E.M, are depicted in the figure (n---7). Asterisks indicate that the increase was statistically different from zero (p<0.05, or less).
FIG. 3. Ethanol-induced increase in the content of triacylglycerol: dependency on the concentration of oleate. Same experiments as depicted in Fig. 1. The ethanol-induced increase in the content of triacylglycerol at 12 hours was calculated for each initial concentration of oleate in each experiment. The mean values-*g.E.M, are depicted in the f~gure (n=7).
6 Time (hours)
FIG. 4. Concentration of oleate in the medium. Same experiments as depicted in Fig. 1. The concentration of free fatty acids was measured in the medium. Mean values-+S.E.M, are depicted in the figure (n=7). The initial concentrations of oleate were as depicted in Fig. 1 (only three experiments were performed with 1.7 mM oleate). (a): Without ethanol; (b): with ethanol.
G R U N N E T , J E N S E N , K O N D R U P AND DICH Uptake
(p,mol/12 h per mg DNA)
FIG. 5. Uptake of el©ate. Same experiments as depicted in Fig. 1. The uptake of oleate was calculated from the difference in the concentrations measured at time-point zero and at 12 hours. The mean values±S.E.M, are depicted in the figure. ©, without ethanol present; O, with ethanol present.
acid oxidation and triacylglycerol synthesis and/or between storage and export of triacylglycerol. These possibilities are presently being investigated. In the following, some methodological aspects of measuring disposal of fatty acids as VLDL-triacylglycerol will be presented.
VLDL-Triacylglycerol Secretion The secretion of triacylglycerol in VLDL panicles was measured by isolating a d < 1.006 fraction of the medium. The possibifity that damaged cells might release intracellular cytoplasmic lipid droplets, which would float with the VLDL particles, was investigated by incubating the hepatocyte cultures with I0/~M cycloheximide to inhibit protein synthesis; the cells were also incubated with 1 mM [l-"C]oleate and after 24 hours incubation the radioactivity in total triacylglycerol and in the triacylglycerol of the VLDL fraction were determined. In control cells lipid radioactivity in the V L D L fraction accounted for 20% of the total triacylglycerol radioactivity in the cells and medium. With cycloheximide present, this figure declined to 1.5~ (average of 2 different cell preparationsk Thus at least 93c~ of the radioactivity in the VLD-triacylglycerol fraction was secreted in conjunction with protein. These results indicate that the contribution from damaged cells was negligible. The content of triacylglycerol, or the incorporation of [I-"C]oleate into triacylglycerol, in the cells plus medium was not affected. In order to discover a possible degradation of VLDLtriacylglycerol in the system, an aliquot of a previously isolated 14C-labelled VLDL fraction was incubated with cell cultures for 24 hours. The added VLDL-triacylglycerol was recovered by 83% which is similar to the recovery in the ultracentrifugation procedure itself (average of 2 experiments). This indicates that VLDL-triacylglycerol is not catabolized in the system. During day 2, cultures incubated with 1 mM oleate sec-
reted VLDL-triacylgiycerol at a rate of 5.2/~mol/24 hr per mg DNA (average of 3 experiments); during the following 24 hours (day 3) the rate rose to 14.8/~mol/24 hr per mg DNA. These results may suggest that accumulation of triacylglycerol induces an increase in the rate of secretion of triacylglycerol. The figure for day 2 can be recalculated to 9 nmol/min per g wet wt. which is 50% higher than that previously found in perfused liver in this laboratory [l 1]. This result indicates a well-preserved structure of the secretory pathway in the cultured hepatocytes. Intracellular triacylglycerol is divided into at least two pools, one pool associated with the endoplasmic reticulum and another pool associated with cytoplasmic lipid droplets [1 l, 13]. It is also clear from experiments with hepatocytes in culture, that more than one pool of triacylglycerol exists. After 5 hours incubation with 1 mM [1-14C]oleate the relative specific radioactivity of VLDL-triacylglycerol and intracellular triacylglycerol were 0.45+_0.02 and 0.30-+0.05, respectively (mean+_S.E.M.; n=4; p<0.025). This result indicates that triacylglycerol is secreted from an intracellular pool with a relative specific radioactivity which is higher than the average specific radioactivity of total triacylglycerol; this pool could be the one associated with the endoplasmic reticulum . The remaining triacylglycerol would presumably be associated with the cytoplasmic lipid droplets; several different approaches, including homogenization and ultrasonic disruption, have proven unsuccessful in trying to separate these two pools in cultured hepatocytes.
Ketone Body Release The release of ketone bodies was followed for various lengths of time for a total of 6 hours. Within the first 1.5 hours the release occurred at a rate equivalent to 1.9+_0.3 (n=4) p.mol oleate/hr per mg DNA (with 1 mM oleate). This can be recalculated to 79 nmoi Ct~/min per g wet wt. which is almost identical to the value found previously in freshly isolated hepatocytes in this laboratory . In the 3--6 hours period the release of ketone bodies had declined to 0.6-+0.2 tn=4) p.mol/hr per mg DNA. It is not likely, however, that this 70% decrease in ketone body production can be fully explained by the 25% decrease in the concentration of oleate (Figs. 4 and 5) in the 6 hr period. Ketone body formation was unaffected by ethanol.
Redox State of Hepatocytes in Culture One prevailing concept about the effect of ethanol on hepatic fatty acid metabolism is that an increased cytosolic NADH/NAD ~ ratio causes a more reduced state of the mitochondria which leads to a decreased rate of fl-oxidation and tricarboxylic acid cycle activity, causing fatty acids to accumulate as triacylglycerol. It was therefore investigated whether addition of ethanol to the hepatocytes in culture would reduce the cytosolic and mitochondrial compartments and whether these changes could be associated, in a timedependent manner, with the effect on the accumulation of triacylglycerol. In the absence of ethanol the lactatetpyruvate ratio at one and a half, three and six hours after the beginning of the incubation was 5.4 (n=2), 5.2 (n=2) and 6.4-+0.3 (n--4), respectively. With ethanol present the lactate/pyruvate ratios were 24.1 -+2.3 (n = 3), 28.4_+2.9 (n = 3) and 37.7-+0.9 (n=4), respectively. The cytosol thus became more reduced in the presence of ethanol. The 3-hydroxybutyrate/acet~cetate ratio, in the absence of ethanol, at one and a half, three and six hours were 3.1-+0.7
FATTY LIVER IN C U L T U R E D HEPATOCYTES
(n=4), 2.9__.0.8 (n--4) and 3.6_+0.9 (n=4), respectively. With ethanol present, the ratios were 4.5__.1.0 (n=4), 4.7_+1.0 (n=4) and 4.3-+ 1.1 (n=4), respectively. Addition of ethanol did cause a moderate reduction of the mitochondrial compartment, but the reduction was not more pronounced at 6 hours when the effect on the accumulation of triacylglycerol was statistically significant (Fig. 2). It seems that the accumulation of triacylglycerol cannot be explained by reduction of the mitochondrial NAD-pool. The lactate/pyruvate and 3-hydroxybutyrate/acetoacetate ratios and the changes caused by ethanol in these ratios, are very similar to values reported for fed rats in vivo [6,15]. DISCUSSION The biochemical integrity of hepatocytes in primary cultures has previously been discussed by us . It may now be added that the hepatocytes exhibit rates of fatty acid uptake, fatty acid synthesis, ketone body release and VLDLtriacylglycerol secretion which are almost identical to the rates measured previously in perfused liver or freshly isolated hepatocytes. In addition, the cultured hepatocytes exhibit redox levels of the cytosolic and mitochondrial compartments that are similar to those measured in vivo. The effect of ethanol on the accumulation of triacylglycerol in hepatocytes in primary culture is in contrast to experiments employing short-term studies with per-
fused liver or freshly isolated cells in which no effect of ethanol, or only small effects, are discernible [3, 5, 14]. The present paper suggests that this discrepancy is due to a time lag of the effect of ethanol of 6-9 hours. Preliminary experiments suggest that this time lag is due to the fact that secretion of VLDL-triacylglycerol is inhibited by ethanol (3040%) only after 6-9 hours exposure to ethanol. In the previous paper  it was concluded that alterations in the concentration of glycerol 3-phosphate could not explain the increased rate of accumulation of triacylglycerol. In the present paper it is further concluded that the fatty acids accumulating as triacylglycerol could not stem from increased fatty acid uptake or synthesis. In addition the content of phospholipids and diacylglycerol was unchanged. This means that ethanol must exert its effect by redistributing fatty acids between oxidation and triacylglycerol synthesis and/or between storage and secretion of triacylglycerol. This redistribution is probably not mediated via an increased redox level of the mitochondria.
ACKNOWLEDGEMENTS We thank Marie Nord and Ida Cohrt T~nnesen for their excellent technical assistance. The work was f'mancially supported by grants from King Christian X's Fund, Ebba Celinders Fund, The Danish Medical Research Council, grant No. 12-4406and NOVO's Fund.
REFERENCES i. Bergmeyer, H. U. Methoden der Enzymatischen Analyse. Weinheim: Verlag Chemic, 1970. 2. Dashti, N., W. J. McConathy and J. A. Ontko. Production of apolipoproteins E and A-I by rat hepatocytes in primary culture. Biochem Biophys Acta 618: 347-358, 1980. 3. Dich, J., B. Bro, N. Grunnet, F. Jensen and J. Kondrup. Accumulation of triacylglycerol in cultured rat hepatocytes is increased by ethanol and by insulin and dexamethasone. Biochem J 212: 617-623, 1983. 4. Fiszer-Szafarz, B., D. Szafarz and A. G. de Murilio. A general fast and sensitive micromethod for DNA determination: Application to rat and mouse liver, rat hepatoma, human leukocytes, chicken fibroblasts, and yeast cells. Anal Biochern 110: 165-170, 1981. 5. Grunnet, N. and J. Kondrup. Effect of ethanol, noradrenalin and 3',5'-cyclic AMP on oxidation of fatty acids and lipolysis in isolated hepatocytes. Pharmacol Biochem Behav 18: Suppl 1, 245-250, 1983. 6. Grunnet, N. and H. I. D. Thieden. The effect of ethanol concentration upon in vivo metabolite levels of rat liver. Life Sci 11: 983--993, 1972.
7. Ho, R. J. and H. C. Meng. A simple and ultrasensitive method for determination of free fatty acids by radiochemical assay. Anal Biochem 31: 426-436, 1969. 8. Jungas, R. L. Fatty acid synthesis in adipose tissue incubated in tritiated water. Biochemistry. 7: 3708-3716, 1968. 9. Kates, M. Techniques of Lipodology. Amsterdam: NorthHolland Publishing, 1972. I0. Kissane, J. M. and E. Robins. The fluorometric measurement of deoxyribonucleic acid in animal tissues with special reference to the central nervous system. J Biol Chem 233: 184-188, 1958. 1i. Kondrup, J. Metabolism of palmitate in perfused rat liver. Biochem J 184: 63-71, 1979. 12. Kondrup, J., B. Bro, J. Dich, N. Grunnet and H. I. D. Thieden. Fractionation and characterization of rat hepatocytes isolated from ethanol-induced fatty liver. Lab Invest 43: 182-190, 1980. 13. Kondrup, J., S. E. Damgaard and P. Fieron. Metabolism of palmitate in perfused liver. Biochem J 184: 73-81, 1979. 14. Kondrup, J., F. Lundquist and S. E. Damgaard. Metabolism of palmitate in perfused liver. Biochem J 184: 83-88, 1979. 15. Veech, R. L., R. Guynn and D. Veioso. The time course of the effect of ethanol on the redox and phosphorylation states of rat liver. Biochem J 127: 387-397, 1972.