- Email: [email protected]
Modulation of liver-specific cellular response to ethanol in vitro in hep G2 cells
Toxicology
in Vitro 12 (1998) 111-122
Modulation of Liver-specific Cellular Response to Ethanol In Vitro in Hep G2 Cells * R. G. CAMERON’, M. G. NEUMAN2t, N. H. SHEAR2, S. BELLENTANI” and C. TIRIBELL13 ‘Department Toronto,
G. KATZ’,
of Pathology, The Toronto Hospital, ‘Division Clinical Pharmacology, Sunnybrook HSC. Ontario, Canada and 3Centro Studi Fegato Modena-Trieste, Department of B.B.C.M, University of Trieste, Trieste, Italy
(Accepted 6 May 1997) Abstract-The aim of this study was to investigate in vitro in a human hepatoblastoma cell line, Hep G2, the effect of ethanol (EtOH) toxicity. The ultrastructural changes were assessed by performing quantitative light and transmission electron microscopy. The second objective of this study was to define further EtOH-induced biochemical changes associated with mitochondrial function. In comparison with controls, after exposure to 80 mM EtOH cells showed: a threefold increase in length of mitochondria; proliferation, vesiculation and dilatation of smooth endoplasmic reticulum, and twofold incrcascs m the size of lipid droplets and in their number/cell. Exposure of cells to two doses of EtOH augmented the ultrastructural alterations observed after a single dose. Cytoviability, assessed by metabolism of methylxanthine dye decreased significantly by (P c 0.0001) to 68% of the control after one dose and was further reduced after the second dose of EtOH (P < 0.001). Succinate dehydrogenase activity in cells treated for 24 hr with 80 mM EtOH was decreased to about 80% of control values after one 24-hr treatment with 80 mM EtOH and to about 55% of control values after two 24-hr exposures. This in vitro model of ethanol-induced cytotoxicity in Hep G2 cells is able to reproduce essential ultrastructural features of alcohol-related hepatitis, in humans, including steatosis and dose-dependent hepatocytotoxicity. The present work represents the first morphometric study to measure changes produced by EtOH exposure in human-derived liver cells. c 1998 Elsevier Science Ltd. All rights reserved Abbreviations: ADH = alcohol dehydrogenase; ANOVA = analysis of variance; EM = ultrastructural examination; EtOH = ethanol; FCS = foetal calf serum; LM = light microscopy; MEM = minimum essential medium; MFO = mixed function oxidase; PBS = phosphate buffered saline; SDH = succinate dehydrogenase (E.C. 1.3.99.1); TEM = transmission electron microscopy; XTT = methyl-xanthine tetrazolium salt (sodium 3,3’-[(phenylamino) carbonyll-3,4-tetrazolium-bis(4-methoxy-6-nitro) benzene sulfonit acid hydrate).
INTRODUCTION
In humans, ethanol-induced liver disease in the acute stage (or acute superimposed on chronic or chronic) has a characteristic morphology (Scheuer, 1982) including macrovesicular steatosis and acute fatty liver hepatitis with Mallory bodies (Biava, 1964; Edmondson, 1980; Schaffner et al., 1963). Ultrastructural examination (EM) reveals pleomorphic mitochondria with giant mitochondria (Baraona et al., 1975; lseri and Gottlieb, 1971) macrovesicular and microvesicular steatosis was performed at *This work Pharmacology, Sunnybrook Health Department of Pathology, The Toronto, Ontario, Canada. TAuthor for correspondence. 0887-2333/98/$19.00+0.00 PII: SO887-2333(97)00095-7
0
Division Clinical Science Center and Toronto Hospital,
(Bruguerd et al., 1977; Iseri et al., 1964; Lischner et al., 1971; Stewart and Dincsoy, 1982) Mallory bodies of different types (Denk et al., 1981; French, 1981; Smuckler, 1968; Yokoo et al., 1972) and proliferation of smooth endoplasmic reticulum (Ishii et al., 1973; Meldolesi, 1967; Rubin and Lieber, 1968). The understanding of the metabolic, structural and functional changes which occur in acute or acute superimposed on chronic exposure is very important. Therefore, to have a reproducible in virro model and to be able to identify morphopathological processes associated with the ethanolinduced changes might provide information that will be useful in the clinical setting. Human-derived hepatoblastoma cells, Hep G2, display many of the phenotypic and genotypic features of normal liver cells and carry no viral infection (Sassa et al., 1987). This cell line preserves
1998 Elsevier Science Ltd. All rights reserved.
Printed
in Great
Britain
most of the functions of normal human liver. foi example carrying out cytochrome P-450-dependent activities oxidase mixed-function (MFO) (Darlington el (I/.. 1987: Dawson cf (I/., 19X5: Limbosch, 1983). In these cells irl vitro glucuronic acid conjugation also responds to inducing agents such as: phenobarbital. I ,2-benz[n]anthracene and 3-methylcholanthrene (Diamond c’t (I/.. 1980: Doodstdr c’t u/.. 1988; Neuman and Tiribelli. 1995: Shear ct al., 1995). We showed (Neuman cf u/.. 1993 and 1995) that an irz Virgo model for acute ethanol-induced toxicity reproduces essential morphological features including steatosis, pleomorphism and large mitochondria and dilatation of smooth endoplasmic reticulum. The aim of the present study was to quantify by morphometry ethanol-induced damage observed at the ultrastructural level and to correlate it with the cytoviability assessment.
Cells were exposed to r-MEM (control) or to EtOH (80 mM) for 24 hr. A second set of cells was incubated with two consecutive doses of 80 mM EtOH (24 hr incubation for each dose). Controls were exposed to (r-MEM for the same period. For each treatment six flasks were used.
EtOH concentration was monitored in flasks containing 80 mM EtOH incubated for 24 hr in the presence or absence of cells (either Hep G2 or SCC13). To detect the EtOH we used a standard gas chromatographic procedure (Jaen, 1971) using a Varian Gas Chromatograph 6000 Vista Series (Varian Instrument Group. Walnut Creek Division, Walnut Creek. CA, USA) with flash vaporization ports and glass insert injection. For preparation we used a 6ft column with 0.25 in. O.D. The parameters were set as follows: helium carrier gas (Oz and HZ0 by high capacity gas purifier) 30 mlimin.
MATERIALS AND METHODS
Ethanol (EtOH) was purchased from Alcohol Ltd (Toronto. Ontario, Canada). From Sigma Chemical Company (St Louis, MO. USA) wre tetrazolium salt obtained the methylxanthine (XTT: sodium [3,3‘-[(phenylamino)carbonyl]-3.4tetrazolim]-bis(4-methoxy-6-nitro) benzene sulfonic acid hydrate), haematoxylin-eosin stain for light microscopy and all the chemicals used for the buffers in cell fractionation as well for succinate dehydrogenase (SDH) activity. r-Minimum essential medium (MEM), calcium chloride. HEPES buffer and phosphate buffered saline (PBS) were from Gibco (Burlington. Ontario). Trypsin was from Difco (Detroit. Ml, USA). Araldite for embedding was purchased from (Cargilles Laboratories Inc.. Cedar Grove, NJ, USA). All stains for electron microscopy were purchased from Electron Microscopy Sciences (Fort Washington, PA, USA).
The human hepatoma cells Hep G2 obtained from Wistar Institute (Philadelphia. PA, USA) were seeded in 75 Falcon flasks and in microplates (I x 10h cells/ml). In some experiments human squamous carcinoma cell line (SSC13) obtained from Harvard Medical School (Boston, MA, USA) was used. The cell counts were monitored using H Couiter counter (Coulter Electronics Inc., Hialeah. FL, USA). Liver long-term cultures were grown in r-MEM supplemented with 10% (v/v) heat-inactivated foetal calf serum (FCS). FCS was not added when the cells were exposed to the treatment. All components were filtered-sterilized and the entire procedure was conducted under aseptic conditions.
Cytotoxicity was assayed by the XTT test (see below). The study was carried out simultaneously in 96-well microtitre trays. Cultures of Hep G2 and SSC13 ceils were set up as described previously (Neuman <‘t rd., 1993). The ceils were seeded at a density of IO6 per well. The growth medium was rMEM containing 10% FCS (as described above) for the first 2 or 3 days (until 70% confluence was reached in each well). When EtOH was added all the dilutions and cultures were performed in xMEM without FCS. In each plate six wells were used for control cells and for each of the treatments. A total of five plates (30 wells/treatment) were used in the XTT measurements.
The Hep G2 cells treated in different combinations as described previously were prepared for light and transmission electron microscopy (TEM) using a standard procedure as outlined below. The medium was removed from the flasks and the cells were washed twice with PBS. To each flask 5 ml I % trypsin solution was added and 2 min later the cells were washed with PBS and then resuspended in r-MEM. Cell suspensions were centrifuged at 5Og for IO min. Pellets were immediately fixed in 2.5 (v/v) glutaraldehyde for at least 24 hr. Blocks of cells were separated, postfixed in 1% (v/v) osmium tetroxide, dehydrated with a graded series of acetone concentrations and embedded in Araidite. Sections (I pm thick) were viewed by light microscopy. For light microscopy studies an Olympus microscope equipped with Leco 2005, Image Processing Analysis and System (LECO Instruments, Toronto, Ontario, Canada) was used. For electron microscopy representative blocks were
Hepatic changes after ethanol exposure selected, subjected to ultrathin sectioning and stained with uranyl acetate and lead citrate. Electron micrographs were taken with a transmission electron microscope JEOL 1200 E x II (JOEL Institute Inc., MA, USA). For EM five different grids/flask were used in each experiment. On each grid 200400 intact cells were examined. An average of 9000 (300 cells/grid x five grids/flask x six flasks/treatment) cells were analysed for each one of treatments. The electron micrographs presented in this work are the most representative for each group of treatments, showing features that can be observed in 60-70% of the cells. Morphometric analysis
Only intact cells with nuclei were assessed for both LM and TEM. The system used for LM morphometry was a modulator high-performance image processing and analysis system, extended with a high-resolution camera which gave a true colour image processing. For each block five slides were studied and six cells/slide were measured. The morphological dimensions (particle sizing) were implemented by a combination of hardware and software to ensure an optimized performance of Microsoft” Visual BasicTM. Magnification for the EM for morphometry was calibrated at x2500 for all the pictures used in quantification of number and size of lipid vesicles and size of mitochondria. The photographs were taken randomly. We were able to fix the boundaries of mitochondria and lipid droplets and to measure their length and axial ratio (short over long diameter) For each cell numerical density (the number of lipid droplets/per cell cytoplasm) was quantified. XTT assay
XTT was used to assay the cytotoxicity. The principle for this test is that viable intact cells take up the dye which is transformed to a red product, which can be quantified spectrophotometrically at an endpoint mode using two wavelengths: absorbance at 560 nm and 690 nm reference (Roehm et al., 1991; Stevens and Olsen, 1993). Cells were grown directly in the 96-well plate. To each well, 100 ~1 XTT solution (1 mg/ml in cr-MEM) were added. The cells were incubated for 1 hr protected from light. At the end of the period the colour was read. Cell viability quantified by metabolism of XTT is expressed as a percentage of controls (control cells viability = 100%). Each determination was carried out in six wells per 96-well plate. For each experiment we used five plates. The results are expressed as the mean &-SD of 30 wells/treatment. Succinate dehydrogenase (SDH)
activity
Enzyme activity was measured The cells from six flasks/treatment
intracellularly. were scraped.
113
The pellet was than homogenized in sucrose 0.01 mol/litre K-HEPES (pH 6.8) using a PotterElvehjem glass homogenizer with a glass pestle (Thomas Scientific, Swedessboro, NJ, USA). The homogenate was than fractionated by rate zonal centrifugation (Wilson et al., 1978) using a refrigerated centrifuge RC 2B (Sorvall; DuPont Co, Wilmington, DE, USA), using a rotor (SZ-14, Sorvall). After discarding the 3000 rpm fraction, the rotor was accelerated to 11,000 rpm to obtain the mitochondrial fraction which was used to measure SDH activity. SDH activity was measured at 37°C by measuring reduction of 2,6-dichlorophenol-indophenol spectrophotometrically at 600 nm (Sottocasa et al., 1967). A solution containing potassium phosphate buffer (50 mM/litre), pH 7.5, ethylenediaminetetraacetic acid (EDTA) (1 mM/litre), 2,6-dichlorophenol-indophenol (1.66 mg/dl), potassium cyanide (1 mM/litre) sodium succinate (20 mM/htre) and phenazine methosulfate (0.33 mg/ml) was used. The enzyme was prepared in potassium phosphate buffer (0.01 M/litre) and added to the solution. 180 ~1 to above described solution was added to 200 /*l mitochondrial fraction. Results are presented as nmol/mg proteimmin. Protein content was determined by the method of Bradford (1976). Bovine serum albumin was used as a standard. For XTT and SDH measurements a Maxline Microplate Reader was connected to a computer using SOFT MAX software 2.3 for WindowsTM (Molecular Devices Corporation, Menlo Park, CA, USA). Data analvsis
Morphometric changes are expressed as mean + SD. Statistical analysis was performed with the Kruskall-Wallis test (Mattews and Farell, 1985). For SDH and ‘XTT measurements the data were analysed using one-way analysis of variance (ANOVA) between the groups with a NeumanKeuls multiple comparison between groups. P < 0.05 was considered significant.
RESULTS
Ethanol metabolism
After 24 hr of exposure, the concentration of EtOH in the absence of Hep G2 cells decreased to 74.8 f 0.3 mM; in the presence of SSC13 cells it from 8OmM to 73.4 f 0.4 mM decreased (mean _t SD). When Hep G2 cells were incubated with 80 mM EtOH, at the end of 24-hr period the EtOH concentration was 61.32 f 2.4 mM. Transmission electron microscopy
Hep G2 cells that were not exposed to EtOH had normal organelles: abundant mitochondria, rough and smooth endoplasmic reticulum and occasional
114
R. G. Cameron
small lipid vesicles (Plate I). Bile canaliculus-like structures were observed at the confluence of the cells. With the addition of 80 IIIM EtOH an enlargement of some cells was seen, mitochondria were swollen, some of the mitochondrial cristae had disappeared and endoplasmic reticulum was enlarged (Plate 2). Lipid accumulation was seen. Lipid vacuoles were seen as rounded inclusions with a smooth homogeneous surface of low electron density having a sharp delimitation. Some of the cells were enlarged (Plate 3) presenting variations in size, shape and number of mitochondria, or mitochondria with rarefied matrix. Between the individual fatty droplets there was a heterogeneity of size and electron density (Plates 3 and 4) a phenomenon that could be observed also by LM (macrovesicular and microvesicular steatosis). The endoplasmic reticulum showed vesiculation. dilatation and ballooning. A second dose of 80 mM EtOH increased the number of lipid droplets and the mitochondrial damage. Some cells preserved their normal nucleus but there were ‘giant’ mitochondria with fewer cristae. other cells were apoptotic with irregular nuclei and dark condensed cytoplasm (Plate 5). The number of lipid droplets increased in comparison with that found after I day of EtOH exposure. Several cells showed vesiculation and dilatation of endoplasmic reticulum.
c’f cd.
secutive periods of 24 hr, cytoviability was significantly further reduced (to 52 _t 4.5%; P < 0.05) in comparison with exposure for only one 24-hr period. The findings confirm our observation (Neuman et cd.. 1993) that EtOH produced a dosedependent cell damage in Hep G2 cells which increased with the number of doses. In contrast. SSC 13 cells treated with an equimolar dose of ethanol for the same period of time did not show a significant decrease in viability (10%). SDH activity (nmol/mg protein/min) in mitochondria of treated cells was significantly decreased compared with the control cells (94 + 4.5 ~1. 120 + 4.0: P < 0.001). The second treatment further reduced the SDH activity (64 +4.5 I’. 116 k 2.5: P < 0.0001).
DISCUSSION
We previously reported that Hep G2 cells represent a good model for the study of ethanolinduced liver damage using a specific mitochondrial marker MTT (Neuman ct al., 1993) and XTT (Neuman et cd., 1995). The XTT response to the increment in the dose of ethanol, was similar with that obtained by MTT (Neuman et cd., 1995). Cells exposed to 20 or 40 mM EtOH did not show reduced viability. The cytotoxicity increased with the dose, beginning with 60 mM and with the number of 80m~ EtOH treatments (one or two conThe image processing of slides using the held secutive doses). In studies (Konrad and Reed, 1977) measurements of intact cells revealed significant using 2-hr exposure at high EtOH doses of 50. differences (P < 0.05) between the area of cells 174 mM, rat postnatal liver cells showed no toxicity. exposed to EtOH (6025 + 345 inn’) and the cells In 1974, Walker et ul. obtained cytotoxicity in norexposed only to plain media (4425 k 525 pm’. In mal human liver cells after exposure to 69-l 74 nlM the intact cells examined in the morphometric stuEtOH. In our model. the cytotoxic effect was dies the nuclear area was not changed significantly obtained only after 16 hr of exposure of Hep G2 (860 k 122 pm2 in control cells V. 780 k 135 pm’ in cells to 80 tnM EtOH. suggesting that the toxic cells exposed to 80 mM EtOH). The ratio of the cell effect is related to a metabolic rather than to a surface/nuclear surface in control cells was 5.145 direct cause (Neuman rt al.. 1993). This conclusion whereas in treated cells the ratio was 7.724. is also supported by the presence of alcohol dehydrogenase (ADH) activity in Hep G2 treated cells EM morphomrtr~~ studies (0.025 + 0.002 IUjmg DNA). In HeP G2 cells, There was a threefold increase in length (pm) of ADH activity is much lower than the activity found mitochondria in the cells exposed to EtOH in comin normal human liver primary cultures exposed to an equimolar dose of EtOH (0.4 _t 0.05 IUI parison with the controls (1.2 +- 0.6 L’. 0.4 k 0.2; mgDNA); Neuman et ul., 1995. P < 0.001) with occasional very long ‘giant’ mitoAt the ultrastructural level. human Hep G2 cells chondria (3 pm). EM studies also revealed proliferrespond to ethanol exposure with a dose-dependent ation, vesiculation and dilatation of smooth steatosis. pleomorphism of mitochondria and with endoplasmic reticulum in the EtOH-treated cells. In comparison with the controls, EtOH-exposed cells enlargement, dilatation and ballooning of endoplasshowed an increase in diameter (pm) of lipid dromic reticulum. In particular, the increase in size of plets (1 .O + 0.4 r. 0.4 _t 0.2 pm; P < 0.0001) and a mitochondria is a morphological pattern very simigreater number of lipid droplets/cell (4.0 & 1.0 11. lar to that observed in patients with alcohol-related 2.0 f 0.5 pm; P < 0.001). liver disease (Iseri and Gottlieb, 1971). The enlargement of mitochondria in the Hep G2 model is related to the dose and length of exposure to the toxicant (Neuman et al.. 1993) Cells exposed to 80 ITIM EtOH showed a significantly decreased viability (68.0% k 2; P < 0.001). In humans ethanol-induced liver damage can be When cells were treated with EtOH for two condefined morphologically as a progressive disease of
Plate I Transmission electron micrograph of control (untereated) Hep G2 cells. Cells were plated in 75 Falcc ,n flasks at a density of 106cells/ml and were grown in medium supplemented with FCS. At 70% in medium without FCS. for 24 hr. After removing the medium the confl uence, cells were incubated cells were prepared for TEM as described in Materials and Methods. The EM shows: normal looking cells with normal central located nucleus (flN), numerous normal mitochondria (\m). One of the cell displ ays vesiculation of the cisternae of the smooth endoplasmic reticulum (\er). Several small lipid (\tj) between two adjacent cells can be observed. drop lets are also present @id). Tight junctions Magnification x 5055.
II5
clcctron micr-ograph 01‘ Hcp (;2 cell \ cuposed to 80 ~F\I ttOH l’or 24 hl-. Cell?. 7. 7 ranwiwon plated 11175 balcon Ha\k\ ill ‘1 dcn\il! 01‘ ltPwll\ ml and w-xc gron n 1n medium \upplemt :nted f-C. At 70”. contluencc. cell\ wrc ~nculxrtcd in mcd~um containing X0 mh~ ethanol without IFCS. : of 11xcells wre 01’ normal G/c. R~IIC others ;IIC slightly enlarged. The nucleus (flN) looks norThe cells show considerable enlargement of many mitochondria (JM). although some normal size :hondria can bc observed (/ml. and prtrmincnt lipid droplets (did). The cisternac of endopla smic Ilum chow vesiculation (/er) and dilatation (‘\ER). Between three cells a ‘bile canaliculus likebtructurc’ (c) can be seen. Magnification x 4825.
Plate 3. Transmission electron micrograph of treated Hep G2 cells (80 mM EtOH for 24 hr) at higher magnification showing a large cell with normal nucleus (fl), some normal looking mitochondria (\m), many large and giant mitochondria (MJ), up to 3 pm in length, and multiple lipid droplets (JLD). A tight junction can be observed between two adjiacent cells (/tj). The upper cell has dilatation of endoplasmic reticulum (\er). A ‘bile canaliculus looking-structure’ (-) is formed at the upper corner between three cells. Magnification x 6080.
II7
Plate 3. TI-anw~ission electron micrograph \ho\lms Hcp G? cells. treated with dox of X0 nl.vt EtOH fol 24 hr. The upper cell has a normal nucleus (flN) and Intact plawamembrane (flpm).The large cell in 1he 111 liddle ih viable. has an Intact plasma mcmbranc. and mitochondria ( -m) that look normal. Note the la rgr number of big lipid droplets (tiLD) of greater clcctron density than is observed in the previous plate. Hecausc of the lipids the nucleus @iis pushed to the basal extremity of the cell. but it is a viabk . nucleus containing a nucleolus (n). Vcsiculation (\er), dilatation (/er) and ballooning (ER) ol rndoplasmic reticulum can bc wcn. Mngutication x 5050.
I IS
Plate 5. Transmission electron micrograph of a cell treated with 80 mM EtOH for two consecutive 24-hr periods. When 70% confluent, cells were washed with PBS and medium containing 80mM EtOH was added (time 0). After 24-hr incubation, followed by aspiration of the medium, a second equimolar dose of EtOH was added for a second period of 24 hr. The cells formed a monolayer that has not been disrupted by the trypsinization. There is mitochondrial pleomorphism with normal size mitochondria (Jm) and large mitochondria (cm). Some of the cells contain enlarged endoplasmic reticulum with vesiculation (/er) and dilatation. A ‘bile canaliculus like-structure’(r) can be observed between three cells. At the left upper corner the cells contain a normal looking nucleus (N). The cell in the centre has a smaller, shrunken nucleus (n). There are two apoptotic cells (A) containing preserved organelles (one cell is in the upper centre and the other is at the lower left corner). Apoptotic bodies (tab) can be observed. These cells show characteristic features of apoptotic cells including elongation, controtion and shrinking. They have intact borders retracted from the surrounding cells. Some vesiculation of ER and fat can be seen. The integrity of organelles and of the plasma membrane remains intact and the cells show dense chromatin aggregated against the nuclear envelope. The cytoplasm is condensed and dark. Magnification x 4425.
II9
Hepatic changes after ethanol exposure
the hepatic parenchyma that includes steatosis, alcoholic hepatitis and cirrhosis (Chedid et al., 1993; Danan, 1993; Lishcher et al., 1971). Also of interest are image analysis results that shows significant liver-cell enlargement after ethanol exposure. Hepatocytes enlargement in chronic exposure to alcohol was observed by Baraona et al. (1975) and Israel et al. (1982). We observed in Hep G2 cells exposed to ethanol the co-existance of normal size and appearance cells (Plate 2) with enlarged, functional cells (Plate 4) or shrunken, non-functional apoptotic cells (Plate 5). The dose-dependent damage, as observed either by reduction in cytoviability using XTT assay or by morphological changes, support the conclusion that the in vitro model described may be predictive. The potential usefulness of this model is increased by its reproducibility and the opportunity it presents for quantifying cell damage by morphometric analysis. In Hep G2 cells, the cytotoxic effects of ethanol, including steatosis and mitochondrial or endoplasmic reticulum changes, suggest more direct causes of alcohol-induced liver damage. The changes in cell organelles are paralleled by changes in SDH activity, suggesting that one of the consequences of EtOH-induced damage is affecting mitochondria. We have been able to show that the morphological changes in cultured Hep G2 cells exposed to EtOH are very similar to those observed in liver biopsies from humans with alcoholism (Neuman et al., 1993). These changes (e.g. steatosis and endoplasmic reticulum dilatation) can occur after exposure to ethanol for only a few days in culture. Experimental animal models for ethanol-induced liver damage that can reproduce features encountered in alcohol-induced liver injury in humans do already exist (Iimuro et al., 1996; Kalant et al., 1975; Lieber et al., 1989; Lindros et al., 1996; Rubin and Lieber, 1974). However, existing in vivo models of ethanol-induced liver damage have required prolonged periods (weeks) of ethanol exposure in order to induce changes compatible with EtOH-liver disease and are expensive (Rubin and Lieber, 1974; Tsukamoto et al., 1984). As the scientific world has become more aware of ethical issues about the use of laboratory animals, the need to develop appropriate corresponding in vitro models has been recognized. For the in vitro study of alcohol-related liver disease the model system should present a lesion that reproduces all the biochemical and morphological abnormalities seen in humans with the disease. In the present study we have focused on the morphometric differences between Hep G2 cells exposed and to EtOH and unexposed control cultures. Caution should be used in the extrapolation of data obtained in cultured cells to the much more complicated processes occurring in vivo in human hepatocytes, which preserve high cytochrome P-450 activities. However, the data described here indicate
121
that Hep G2 cells could be useful in assessing the mechanism(s) of the cytotoxic effects of ethanol. The results of our parallel study in which we used Hep G2 cells to evaluate the effects of therapeutic bile acids in modifying alcohol-induced toxicity in vitro suggest that this model has potential for the study of factors that may prevent or reverse alcohol-related hepatotoxicity (Neuman et al., 1995).
Acknowledgemenrs-The
authors thank Izabella Malkiewicz (Department of Clinical Pharmacology, Sunnybrook HSC) and Lois Lanes, Julia Hwong and Richard Leung (Departments of Pathology, Hospital for Sick Children and The Toronto Hospital) for the skilful technical assistance with EM preparation and to Fisher Co. for kindly providing us with the molecular device reader. The authors also acknowledge the following support: S. Bellentani-Fondo Studi Fegato (FSF), Trieste, Italy; R. G. Cameron-Dept of Pathology, Toronto Hospital Research Foundation; M. G. Neuman-personal award from University of Toronto Drug Safety Research Group; N. H. Shear--career scientist Ontario Ministry of Health, Canada; C. Tiribelli-MURST 40% (Fegato Cirrosi Epatica e Epatiti Virali), Roma and FSF, Trieste, Italy. REFERENCES Baraona E., Leo M. A., Borowsky S. A. and Lieber C. S. (1975) Alcoholic hepatomegaly: accumulation of protein in the liver. Science 190, 794795. Biava C. (1964) Mallory alcoholic hyalin: a heterofore unique lesion of hepatocellular ergastoplasm. Labbratory Investigation 13,.301-320.
Bradford M. (1976) A ranid and sensitive method for the quantification of’microgram quantities of protein utilizing the principle of protein dye binding. Analytical Biochemistry Ii, 248-236.
Bruauera M.. Bertran A.. Bombi J. A. and Rodes J. (1977) &ant mitochondria in hepatocytes. A diagnostic first for alcoholic liver disease. Gastroenrerology 73, 13831387.
Chedid A., Mendenhall C. L., Moritz T. E., French S. W., Chen T. S., Morgan T. R., Roselle G. A., Nemchausky B. A., Tamburo C. H. and Schiff E. R. (1993) Cellmediated hepatic injury in alcoholic liver disease. Gastroenterology 105, 294-296.
Danan G. (1993) Definitions et criteres d’imputation des atteintes hepatiques aigues medicamenteuses. Conclusions d’une reunion internationale de consensus. Gastroenterology and Clinical Biology 17, H 18-H2 1. Darlington G. J., Kelly J. H. and Buffone G. J. (1987) Growth and hepatospectrum gene expression of human hepatoma cells in a defined medium. In Vitro Cell Development and Biology 23, 349-354.
Dawson J. R., Adams D. J. and Wolf C. R. (1985) Induction of drug metabolizing enzymes in human liver cell line Hep G2. FEBS Letters 183, 219-222. Denk H., Franke W. W., Dragisics B. and Zeiler I. (1981) Pathology of cytoskeleton of liver cells: demonstration of Mallory bodies (alcoholic hyalin) in murine and human hepatocytes by immunofluorescence microscopy, using antibodies to cytokeratin polypeptides from hepatocvtes. HeDatolonv 1, 9-20. Diamond L.,. Kruszewski F., Aden D. P., Knowles B. B. and Baird W. M. (1980) Metabolic activation of benzo(a) pyrene by a human hepatoma cell line. Carcinogenesis 1, 871-875.
172
R. C. Cam
Doodstdr H., Duthie S. J., Burke M. D., Melvin W. T. and Grant M. H. (1988) The influence of culture medium composition on drug metabolising enzyme activities of the human liver derived Hep G2 cell line. F‘EBS Lerter.~ 1-2, IS- 18. Edmondson H. A. (1980) Pathology of alcoholism. Americuw
Journul
01 Clinicul
Pathologic
74, 725 742.
French S. W. (1981) The Mallory body: structure. composition and pathogenesis. Hepatolog~~ 1, 76-83 limuro Y., lkejima K.. Rose M. L., Bradford B. U. and Thurman R. G. (1996) Nimodipine. a dihydropyridinctype calcium channel blocker, prevents alcoholic hepatitis caused by chronic intragastrrc ethanol exposure rn the rat. Hc~prrto/o,ej~ 24, 39 I 397. lseri 0. A. and Gottlieb L. S. (1971) Alcoholic hyalin and megamitochondria as separated and distinct entities in liver disease associated with alcoholism. G‘crsrroenterolog~ 60, 1027~1035. Iseri 0. A., Lieber C. S. and Gottlieb L. S. (1964) The ultrastructure of fatty liver induced by prolonged ethanol ingestion. ilnwricrm Journul of Pathology. 48, 535 555. Ishii H.. Joly J. and Lieber C. S. (1973) Effect of ethanol on the amount and enzyme activities of hepatic rough and smooth microsomal membrane. Biwhimk et Bioph.wica Acttr 291, 41 l-420. Israel Y., Britton R. and Orrego H. (1982) Liver- cell enlargement induced by chronic alcohol consumption: studies on its cause and consequences. Clink/ Biochemi.Ytr~~ 15, I899 I92 Jaen N. C. (1971) Direct blood Injectton method ]‘<)I-g;ta chromatographic determination of alcohols and other volatile compounds. Clirricnl Cl7cwi.cty 17, 82 X5. Kalant H.. Orrego H.. Khanna J. M. and Phrlhps J. M. (1975) Experimental alcohol-induced hepatic necrosis: suppression by propylthiouracil. Procc&irrg.r o/’ r/re Nutiontrl Acuden~~~ of Scietucc of the Y.S. A 12, 1137~1141. Konrad K. D. and Reed G. (1977) Organ culture of portnatal rat liver: effect of ethanol on lipid metabolism Esl,erimental
and Moieculur
Putholog~~ 26, 425 433.
286. Lindros K. O., Badger T., Ronis M.. Ingelman-Sundbel-p M. and Koivusalo M. (1996) Phenyl-lsocyanate. a new dietary liver aldehyde dehydrogenase inhibitor. .Iorrrrrtr/ Phurmacolo~q>~ (tnd 79-83.
E.vperimenttrl
Theruputic~t
cell injuries in the human Hep G2 cell line. 109, 5555563. Neuman M. G.. Koren G. and Tiribelli C. (1993) In vitro assessment of the ethanol-induced hepatotoxicity on Hep G2 cell line. Biochemical and Biophysical Research Communicuiions
2, 932-942.
Neuman M. G. and Tiribelli able in preclinical stage?
C. (1995) Drugs are predictModels and techniques. In
Liver tmd Drup ‘94. Progrers 1, pp. 339. Bo-Art, Brdtisiava.
in Hepato-Pharmacoll~g~,
Phillips M. J., Poucell S.. Pattersson J. and Valencia P. ( 1987) Tiw Liver. An Arlus of L~/fru.structurul Puthology. Raven Press. New York. Roehm N. W., Rogers G. H.. Hatfield S. M. and Glasebrook A. L. (1991) An improved calorimetric assay for cell proliferation and viability utilizing the tetrazolium salt XTT. Journal of Immunologicul Methotlv 142, 251-265.
Rubin E. and Lieber C. S. (1968) Alcohol induced injury in non-alcoholic volunteers. Nen England of’ Medicine
hepatic Journal
278, 869-876.
Rubin E. and Lieber C. S. (1974) Fatty liver, aicohoiic hepatitrs and cirrhosis produced by alcohol in primates. ,Yru, England Journul of‘ Medicine 290, 128-I 35. Sassa S., Sugita O., Galbraith R. A. and Kappas A. (1987) Drug metabolism by the human hepatoma cell, Hep G2. Biochemiccd 143, 52-57.
und
Biophysiccd
Research
Communicution.c
Schaffner F., Loebel A.. Weiner H. A. and Barka T. (1963) Hepatocellular cytoplasmic changes in acute alcoholic hepatitis. Journal of rhr Americun Mrdirul Associution 183, 343-349.
Scheuer P. J. (1982) The morphology of alcoholic liver disease. British Medicul Bulletin 38, 63-65. Shear N. H., Malkiewicz I., Klein D., Randor S., Koren G. and Neuman M. G. (1995) Acetaminophen-induced toxicity to keratinocytes and to hepatocytes in vitro is diminished by Silymarin. Skin Pharmacology 8, 45-56. Smuckler E. A. (1968) The ultrastructure of human alcoholic hyaline. .Amcricttn Journul of PathoLogl, 49, 790-~ 797.
Lieber C. S.. Baraona E.. Hernandez-Munoz R. and Kubota S. (1989) Impaired oxygen utilization 4 new mechanism for the hepatotoxicity of ethanol in subhuman primates. Journul of’ C/iniu/ Inwvtigtrtio~~ 83, 16X2-1690. Limbosch S. (1983) Benzo (a)pyrene and aldrin metabohzing activities in cultured human and rat hepatoma cell line. Journtrl of iIre Nrrtionul Canwr Ir~stitute 71, 2X 1
of
induced
Gasiroenterolog~~
275.
Lischner M. W.. Alexander _I. F. and Galambos J. T. (1971) Natural history of alcoholic hepatitis. I. The acute disease. Americctn Journul of’ Digestiw Di.trrrs~.~ 16, 48 I 494. Mattews. D. E. and Farewell. V. (19X5) L:sirq wu/ U~der.r/unding Mediical .Stu/i.rfic.s pp. I 12 I I6 and I3 129. S. Karger, AG. Basel. Meldolesi J. (1967) On the significance of the hypertrophy of the smooth endoplasmic reticulum in liver cells after administration of drugs. Biochemical Pharmucolog~ 16, 125-131.
Neuman M. G.. Cameron R. G., Shear N. H., Bellentam S. and Tiribelli C. (1995) Effect of Tauroursodeoxycholate and ursodeoxycholic acid on ethanol-
Sottocasa G. L.. Kuitenstierna B., Ernster L. and Bergstrand L. (1967) An electron transport system associated with the outer membrane of liver mitochondria: a biochemical and morphological study. Journul qf’ Cell Biology
32, 415-438.
Stevens M. G. and Olsen S. C. (1993) Comparative analysis of using MTT and XTT in calorimetric assays for quantitating bovine neutrophil bacterial activity. Journul
of
Immunologicul
Methods
157, 225-233.
Stewart R. V. and Dincsoy H. P. (1982) The significance of giant mitochondria in liver biopsies as observed by light microscopy. American Journal of Clinicul Purholog~~ 78, 293-298.
Tsukamoto H., Reiderberger R. D., French Largman C. (1984) Long-term cannulation blood sampling and intragastric infusion American
Journal
of Phvsiologv
S. W. and model for in the rat.
241, R595-R599.
Walker F., Elmslie N., Fraser- R. A., Snape P. E. and Watt B. C. (1974) Cvtotoxic effect of alcohol on liver cells and tibroblast rn*vitro. Scottish Medic,ul Journcd 19, IL- 127. Wilson M. A., Cascarano J., Wooten W. I. and Pickett C. B. (1978) Quantitative isolation of liver mitochondria by zonal centrifugation. Analytical Biochemistry 85, 225-264.
Yokoo H., Minick 0. T.. Batti F. and Kent Morphological variants of alcoholic hyalin. Journal
of Pathology,
69. 25-32.
G. (1972) Americrm
in Vitro 12 (1998) 111-122
Modulation of Liver-specific Cellular Response to Ethanol In Vitro in Hep G2 Cells * R. G. CAMERON’, M. G. NEUMAN2t, N. H. SHEAR2, S. BELLENTANI” and C. TIRIBELL13 ‘Department Toronto,
G. KATZ’,
of Pathology, The Toronto Hospital, ‘Division Clinical Pharmacology, Sunnybrook HSC. Ontario, Canada and 3Centro Studi Fegato Modena-Trieste, Department of B.B.C.M, University of Trieste, Trieste, Italy
(Accepted 6 May 1997) Abstract-The aim of this study was to investigate in vitro in a human hepatoblastoma cell line, Hep G2, the effect of ethanol (EtOH) toxicity. The ultrastructural changes were assessed by performing quantitative light and transmission electron microscopy. The second objective of this study was to define further EtOH-induced biochemical changes associated with mitochondrial function. In comparison with controls, after exposure to 80 mM EtOH cells showed: a threefold increase in length of mitochondria; proliferation, vesiculation and dilatation of smooth endoplasmic reticulum, and twofold incrcascs m the size of lipid droplets and in their number/cell. Exposure of cells to two doses of EtOH augmented the ultrastructural alterations observed after a single dose. Cytoviability, assessed by metabolism of methylxanthine dye decreased significantly by (P c 0.0001) to 68% of the control after one dose and was further reduced after the second dose of EtOH (P < 0.001). Succinate dehydrogenase activity in cells treated for 24 hr with 80 mM EtOH was decreased to about 80% of control values after one 24-hr treatment with 80 mM EtOH and to about 55% of control values after two 24-hr exposures. This in vitro model of ethanol-induced cytotoxicity in Hep G2 cells is able to reproduce essential ultrastructural features of alcohol-related hepatitis, in humans, including steatosis and dose-dependent hepatocytotoxicity. The present work represents the first morphometric study to measure changes produced by EtOH exposure in human-derived liver cells. c 1998 Elsevier Science Ltd. All rights reserved Abbreviations: ADH = alcohol dehydrogenase; ANOVA = analysis of variance; EM = ultrastructural examination; EtOH = ethanol; FCS = foetal calf serum; LM = light microscopy; MEM = minimum essential medium; MFO = mixed function oxidase; PBS = phosphate buffered saline; SDH = succinate dehydrogenase (E.C. 1.3.99.1); TEM = transmission electron microscopy; XTT = methyl-xanthine tetrazolium salt (sodium 3,3’-[(phenylamino) carbonyll-3,4-tetrazolium-bis(4-methoxy-6-nitro) benzene sulfonit acid hydrate).
INTRODUCTION
In humans, ethanol-induced liver disease in the acute stage (or acute superimposed on chronic or chronic) has a characteristic morphology (Scheuer, 1982) including macrovesicular steatosis and acute fatty liver hepatitis with Mallory bodies (Biava, 1964; Edmondson, 1980; Schaffner et al., 1963). Ultrastructural examination (EM) reveals pleomorphic mitochondria with giant mitochondria (Baraona et al., 1975; lseri and Gottlieb, 1971) macrovesicular and microvesicular steatosis was performed at *This work Pharmacology, Sunnybrook Health Department of Pathology, The Toronto, Ontario, Canada. TAuthor for correspondence. 0887-2333/98/$19.00+0.00 PII: SO887-2333(97)00095-7
0
Division Clinical Science Center and Toronto Hospital,
(Bruguerd et al., 1977; Iseri et al., 1964; Lischner et al., 1971; Stewart and Dincsoy, 1982) Mallory bodies of different types (Denk et al., 1981; French, 1981; Smuckler, 1968; Yokoo et al., 1972) and proliferation of smooth endoplasmic reticulum (Ishii et al., 1973; Meldolesi, 1967; Rubin and Lieber, 1968). The understanding of the metabolic, structural and functional changes which occur in acute or acute superimposed on chronic exposure is very important. Therefore, to have a reproducible in virro model and to be able to identify morphopathological processes associated with the ethanolinduced changes might provide information that will be useful in the clinical setting. Human-derived hepatoblastoma cells, Hep G2, display many of the phenotypic and genotypic features of normal liver cells and carry no viral infection (Sassa et al., 1987). This cell line preserves
1998 Elsevier Science Ltd. All rights reserved.
Printed
in Great
Britain
most of the functions of normal human liver. foi example carrying out cytochrome P-450-dependent activities oxidase mixed-function (MFO) (Darlington el (I/.. 1987: Dawson cf (I/., 19X5: Limbosch, 1983). In these cells irl vitro glucuronic acid conjugation also responds to inducing agents such as: phenobarbital. I ,2-benz[n]anthracene and 3-methylcholanthrene (Diamond c’t (I/.. 1980: Doodstdr c’t u/.. 1988; Neuman and Tiribelli. 1995: Shear ct al., 1995). We showed (Neuman cf u/.. 1993 and 1995) that an irz Virgo model for acute ethanol-induced toxicity reproduces essential morphological features including steatosis, pleomorphism and large mitochondria and dilatation of smooth endoplasmic reticulum. The aim of the present study was to quantify by morphometry ethanol-induced damage observed at the ultrastructural level and to correlate it with the cytoviability assessment.
Cells were exposed to r-MEM (control) or to EtOH (80 mM) for 24 hr. A second set of cells was incubated with two consecutive doses of 80 mM EtOH (24 hr incubation for each dose). Controls were exposed to (r-MEM for the same period. For each treatment six flasks were used.
EtOH concentration was monitored in flasks containing 80 mM EtOH incubated for 24 hr in the presence or absence of cells (either Hep G2 or SCC13). To detect the EtOH we used a standard gas chromatographic procedure (Jaen, 1971) using a Varian Gas Chromatograph 6000 Vista Series (Varian Instrument Group. Walnut Creek Division, Walnut Creek. CA, USA) with flash vaporization ports and glass insert injection. For preparation we used a 6ft column with 0.25 in. O.D. The parameters were set as follows: helium carrier gas (Oz and HZ0 by high capacity gas purifier) 30 mlimin.
MATERIALS AND METHODS
Ethanol (EtOH) was purchased from Alcohol Ltd (Toronto. Ontario, Canada). From Sigma Chemical Company (St Louis, MO. USA) wre tetrazolium salt obtained the methylxanthine (XTT: sodium [3,3‘-[(phenylamino)carbonyl]-3.4tetrazolim]-bis(4-methoxy-6-nitro) benzene sulfonic acid hydrate), haematoxylin-eosin stain for light microscopy and all the chemicals used for the buffers in cell fractionation as well for succinate dehydrogenase (SDH) activity. r-Minimum essential medium (MEM), calcium chloride. HEPES buffer and phosphate buffered saline (PBS) were from Gibco (Burlington. Ontario). Trypsin was from Difco (Detroit. Ml, USA). Araldite for embedding was purchased from (Cargilles Laboratories Inc.. Cedar Grove, NJ, USA). All stains for electron microscopy were purchased from Electron Microscopy Sciences (Fort Washington, PA, USA).
The human hepatoma cells Hep G2 obtained from Wistar Institute (Philadelphia. PA, USA) were seeded in 75 Falcon flasks and in microplates (I x 10h cells/ml). In some experiments human squamous carcinoma cell line (SSC13) obtained from Harvard Medical School (Boston, MA, USA) was used. The cell counts were monitored using H Couiter counter (Coulter Electronics Inc., Hialeah. FL, USA). Liver long-term cultures were grown in r-MEM supplemented with 10% (v/v) heat-inactivated foetal calf serum (FCS). FCS was not added when the cells were exposed to the treatment. All components were filtered-sterilized and the entire procedure was conducted under aseptic conditions.
Cytotoxicity was assayed by the XTT test (see below). The study was carried out simultaneously in 96-well microtitre trays. Cultures of Hep G2 and SSC13 ceils were set up as described previously (Neuman <‘t rd., 1993). The ceils were seeded at a density of IO6 per well. The growth medium was rMEM containing 10% FCS (as described above) for the first 2 or 3 days (until 70% confluence was reached in each well). When EtOH was added all the dilutions and cultures were performed in xMEM without FCS. In each plate six wells were used for control cells and for each of the treatments. A total of five plates (30 wells/treatment) were used in the XTT measurements.
The Hep G2 cells treated in different combinations as described previously were prepared for light and transmission electron microscopy (TEM) using a standard procedure as outlined below. The medium was removed from the flasks and the cells were washed twice with PBS. To each flask 5 ml I % trypsin solution was added and 2 min later the cells were washed with PBS and then resuspended in r-MEM. Cell suspensions were centrifuged at 5Og for IO min. Pellets were immediately fixed in 2.5 (v/v) glutaraldehyde for at least 24 hr. Blocks of cells were separated, postfixed in 1% (v/v) osmium tetroxide, dehydrated with a graded series of acetone concentrations and embedded in Araidite. Sections (I pm thick) were viewed by light microscopy. For light microscopy studies an Olympus microscope equipped with Leco 2005, Image Processing Analysis and System (LECO Instruments, Toronto, Ontario, Canada) was used. For electron microscopy representative blocks were
Hepatic changes after ethanol exposure selected, subjected to ultrathin sectioning and stained with uranyl acetate and lead citrate. Electron micrographs were taken with a transmission electron microscope JEOL 1200 E x II (JOEL Institute Inc., MA, USA). For EM five different grids/flask were used in each experiment. On each grid 200400 intact cells were examined. An average of 9000 (300 cells/grid x five grids/flask x six flasks/treatment) cells were analysed for each one of treatments. The electron micrographs presented in this work are the most representative for each group of treatments, showing features that can be observed in 60-70% of the cells. Morphometric analysis
Only intact cells with nuclei were assessed for both LM and TEM. The system used for LM morphometry was a modulator high-performance image processing and analysis system, extended with a high-resolution camera which gave a true colour image processing. For each block five slides were studied and six cells/slide were measured. The morphological dimensions (particle sizing) were implemented by a combination of hardware and software to ensure an optimized performance of Microsoft” Visual BasicTM. Magnification for the EM for morphometry was calibrated at x2500 for all the pictures used in quantification of number and size of lipid vesicles and size of mitochondria. The photographs were taken randomly. We were able to fix the boundaries of mitochondria and lipid droplets and to measure their length and axial ratio (short over long diameter) For each cell numerical density (the number of lipid droplets/per cell cytoplasm) was quantified. XTT assay
XTT was used to assay the cytotoxicity. The principle for this test is that viable intact cells take up the dye which is transformed to a red product, which can be quantified spectrophotometrically at an endpoint mode using two wavelengths: absorbance at 560 nm and 690 nm reference (Roehm et al., 1991; Stevens and Olsen, 1993). Cells were grown directly in the 96-well plate. To each well, 100 ~1 XTT solution (1 mg/ml in cr-MEM) were added. The cells were incubated for 1 hr protected from light. At the end of the period the colour was read. Cell viability quantified by metabolism of XTT is expressed as a percentage of controls (control cells viability = 100%). Each determination was carried out in six wells per 96-well plate. For each experiment we used five plates. The results are expressed as the mean &-SD of 30 wells/treatment. Succinate dehydrogenase (SDH)
activity
Enzyme activity was measured The cells from six flasks/treatment
intracellularly. were scraped.
113
The pellet was than homogenized in sucrose 0.01 mol/litre K-HEPES (pH 6.8) using a PotterElvehjem glass homogenizer with a glass pestle (Thomas Scientific, Swedessboro, NJ, USA). The homogenate was than fractionated by rate zonal centrifugation (Wilson et al., 1978) using a refrigerated centrifuge RC 2B (Sorvall; DuPont Co, Wilmington, DE, USA), using a rotor (SZ-14, Sorvall). After discarding the 3000 rpm fraction, the rotor was accelerated to 11,000 rpm to obtain the mitochondrial fraction which was used to measure SDH activity. SDH activity was measured at 37°C by measuring reduction of 2,6-dichlorophenol-indophenol spectrophotometrically at 600 nm (Sottocasa et al., 1967). A solution containing potassium phosphate buffer (50 mM/litre), pH 7.5, ethylenediaminetetraacetic acid (EDTA) (1 mM/litre), 2,6-dichlorophenol-indophenol (1.66 mg/dl), potassium cyanide (1 mM/litre) sodium succinate (20 mM/htre) and phenazine methosulfate (0.33 mg/ml) was used. The enzyme was prepared in potassium phosphate buffer (0.01 M/litre) and added to the solution. 180 ~1 to above described solution was added to 200 /*l mitochondrial fraction. Results are presented as nmol/mg proteimmin. Protein content was determined by the method of Bradford (1976). Bovine serum albumin was used as a standard. For XTT and SDH measurements a Maxline Microplate Reader was connected to a computer using SOFT MAX software 2.3 for WindowsTM (Molecular Devices Corporation, Menlo Park, CA, USA). Data analvsis
Morphometric changes are expressed as mean + SD. Statistical analysis was performed with the Kruskall-Wallis test (Mattews and Farell, 1985). For SDH and ‘XTT measurements the data were analysed using one-way analysis of variance (ANOVA) between the groups with a NeumanKeuls multiple comparison between groups. P < 0.05 was considered significant.
RESULTS
Ethanol metabolism
After 24 hr of exposure, the concentration of EtOH in the absence of Hep G2 cells decreased to 74.8 f 0.3 mM; in the presence of SSC13 cells it from 8OmM to 73.4 f 0.4 mM decreased (mean _t SD). When Hep G2 cells were incubated with 80 mM EtOH, at the end of 24-hr period the EtOH concentration was 61.32 f 2.4 mM. Transmission electron microscopy
Hep G2 cells that were not exposed to EtOH had normal organelles: abundant mitochondria, rough and smooth endoplasmic reticulum and occasional
114
R. G. Cameron
small lipid vesicles (Plate I). Bile canaliculus-like structures were observed at the confluence of the cells. With the addition of 80 IIIM EtOH an enlargement of some cells was seen, mitochondria were swollen, some of the mitochondrial cristae had disappeared and endoplasmic reticulum was enlarged (Plate 2). Lipid accumulation was seen. Lipid vacuoles were seen as rounded inclusions with a smooth homogeneous surface of low electron density having a sharp delimitation. Some of the cells were enlarged (Plate 3) presenting variations in size, shape and number of mitochondria, or mitochondria with rarefied matrix. Between the individual fatty droplets there was a heterogeneity of size and electron density (Plates 3 and 4) a phenomenon that could be observed also by LM (macrovesicular and microvesicular steatosis). The endoplasmic reticulum showed vesiculation. dilatation and ballooning. A second dose of 80 mM EtOH increased the number of lipid droplets and the mitochondrial damage. Some cells preserved their normal nucleus but there were ‘giant’ mitochondria with fewer cristae. other cells were apoptotic with irregular nuclei and dark condensed cytoplasm (Plate 5). The number of lipid droplets increased in comparison with that found after I day of EtOH exposure. Several cells showed vesiculation and dilatation of endoplasmic reticulum.
c’f cd.
secutive periods of 24 hr, cytoviability was significantly further reduced (to 52 _t 4.5%; P < 0.05) in comparison with exposure for only one 24-hr period. The findings confirm our observation (Neuman et cd.. 1993) that EtOH produced a dosedependent cell damage in Hep G2 cells which increased with the number of doses. In contrast. SSC 13 cells treated with an equimolar dose of ethanol for the same period of time did not show a significant decrease in viability (10%). SDH activity (nmol/mg protein/min) in mitochondria of treated cells was significantly decreased compared with the control cells (94 + 4.5 ~1. 120 + 4.0: P < 0.001). The second treatment further reduced the SDH activity (64 +4.5 I’. 116 k 2.5: P < 0.0001).
DISCUSSION
We previously reported that Hep G2 cells represent a good model for the study of ethanolinduced liver damage using a specific mitochondrial marker MTT (Neuman ct al., 1993) and XTT (Neuman et cd., 1995). The XTT response to the increment in the dose of ethanol, was similar with that obtained by MTT (Neuman et cd., 1995). Cells exposed to 20 or 40 mM EtOH did not show reduced viability. The cytotoxicity increased with the dose, beginning with 60 mM and with the number of 80m~ EtOH treatments (one or two conThe image processing of slides using the held secutive doses). In studies (Konrad and Reed, 1977) measurements of intact cells revealed significant using 2-hr exposure at high EtOH doses of 50. differences (P < 0.05) between the area of cells 174 mM, rat postnatal liver cells showed no toxicity. exposed to EtOH (6025 + 345 inn’) and the cells In 1974, Walker et ul. obtained cytotoxicity in norexposed only to plain media (4425 k 525 pm’. In mal human liver cells after exposure to 69-l 74 nlM the intact cells examined in the morphometric stuEtOH. In our model. the cytotoxic effect was dies the nuclear area was not changed significantly obtained only after 16 hr of exposure of Hep G2 (860 k 122 pm2 in control cells V. 780 k 135 pm’ in cells to 80 tnM EtOH. suggesting that the toxic cells exposed to 80 mM EtOH). The ratio of the cell effect is related to a metabolic rather than to a surface/nuclear surface in control cells was 5.145 direct cause (Neuman rt al.. 1993). This conclusion whereas in treated cells the ratio was 7.724. is also supported by the presence of alcohol dehydrogenase (ADH) activity in Hep G2 treated cells EM morphomrtr~~ studies (0.025 + 0.002 IUjmg DNA). In HeP G2 cells, There was a threefold increase in length (pm) of ADH activity is much lower than the activity found mitochondria in the cells exposed to EtOH in comin normal human liver primary cultures exposed to an equimolar dose of EtOH (0.4 _t 0.05 IUI parison with the controls (1.2 +- 0.6 L’. 0.4 k 0.2; mgDNA); Neuman et ul., 1995. P < 0.001) with occasional very long ‘giant’ mitoAt the ultrastructural level. human Hep G2 cells chondria (3 pm). EM studies also revealed proliferrespond to ethanol exposure with a dose-dependent ation, vesiculation and dilatation of smooth steatosis. pleomorphism of mitochondria and with endoplasmic reticulum in the EtOH-treated cells. In comparison with the controls, EtOH-exposed cells enlargement, dilatation and ballooning of endoplasshowed an increase in diameter (pm) of lipid dromic reticulum. In particular, the increase in size of plets (1 .O + 0.4 r. 0.4 _t 0.2 pm; P < 0.0001) and a mitochondria is a morphological pattern very simigreater number of lipid droplets/cell (4.0 & 1.0 11. lar to that observed in patients with alcohol-related 2.0 f 0.5 pm; P < 0.001). liver disease (Iseri and Gottlieb, 1971). The enlargement of mitochondria in the Hep G2 model is related to the dose and length of exposure to the toxicant (Neuman et al.. 1993) Cells exposed to 80 ITIM EtOH showed a significantly decreased viability (68.0% k 2; P < 0.001). In humans ethanol-induced liver damage can be When cells were treated with EtOH for two condefined morphologically as a progressive disease of
Plate I Transmission electron micrograph of control (untereated) Hep G2 cells. Cells were plated in 75 Falcc ,n flasks at a density of 106cells/ml and were grown in medium supplemented with FCS. At 70% in medium without FCS. for 24 hr. After removing the medium the confl uence, cells were incubated cells were prepared for TEM as described in Materials and Methods. The EM shows: normal looking cells with normal central located nucleus (flN), numerous normal mitochondria (\m). One of the cell displ ays vesiculation of the cisternae of the smooth endoplasmic reticulum (\er). Several small lipid (\tj) between two adjacent cells can be observed. drop lets are also present @id). Tight junctions Magnification x 5055.
II5
clcctron micr-ograph 01‘ Hcp (;2 cell \ cuposed to 80 ~F\I ttOH l’or 24 hl-. Cell?. 7. 7 ranwiwon plated 11175 balcon Ha\k\ ill ‘1 dcn\il! 01‘ ltPwll\ ml and w-xc gron n 1n medium \upplemt :nted f-C. At 70”. contluencc. cell\ wrc ~nculxrtcd in mcd~um containing X0 mh~ ethanol without IFCS. : of 11xcells wre 01’ normal G/c. R~IIC others ;IIC slightly enlarged. The nucleus (flN) looks norThe cells show considerable enlargement of many mitochondria (JM). although some normal size :hondria can bc observed (/ml. and prtrmincnt lipid droplets (did). The cisternac of endopla smic Ilum chow vesiculation (/er) and dilatation (‘\ER). Between three cells a ‘bile canaliculus likebtructurc’ (c) can be seen. Magnification x 4825.
Plate 3. Transmission electron micrograph of treated Hep G2 cells (80 mM EtOH for 24 hr) at higher magnification showing a large cell with normal nucleus (fl), some normal looking mitochondria (\m), many large and giant mitochondria (MJ), up to 3 pm in length, and multiple lipid droplets (JLD). A tight junction can be observed between two adjiacent cells (/tj). The upper cell has dilatation of endoplasmic reticulum (\er). A ‘bile canaliculus looking-structure’ (-) is formed at the upper corner between three cells. Magnification x 6080.
II7
Plate 3. TI-anw~ission electron micrograph \ho\lms Hcp G? cells. treated with dox of X0 nl.vt EtOH fol 24 hr. The upper cell has a normal nucleus (flN) and Intact plawamembrane (flpm).The large cell in 1he 111 liddle ih viable. has an Intact plasma mcmbranc. and mitochondria ( -m) that look normal. Note the la rgr number of big lipid droplets (tiLD) of greater clcctron density than is observed in the previous plate. Hecausc of the lipids the nucleus @iis pushed to the basal extremity of the cell. but it is a viabk . nucleus containing a nucleolus (n). Vcsiculation (\er), dilatation (/er) and ballooning (ER) ol rndoplasmic reticulum can bc wcn. Mngutication x 5050.
I IS
Plate 5. Transmission electron micrograph of a cell treated with 80 mM EtOH for two consecutive 24-hr periods. When 70% confluent, cells were washed with PBS and medium containing 80mM EtOH was added (time 0). After 24-hr incubation, followed by aspiration of the medium, a second equimolar dose of EtOH was added for a second period of 24 hr. The cells formed a monolayer that has not been disrupted by the trypsinization. There is mitochondrial pleomorphism with normal size mitochondria (Jm) and large mitochondria (cm). Some of the cells contain enlarged endoplasmic reticulum with vesiculation (/er) and dilatation. A ‘bile canaliculus like-structure’(r) can be observed between three cells. At the left upper corner the cells contain a normal looking nucleus (N). The cell in the centre has a smaller, shrunken nucleus (n). There are two apoptotic cells (A) containing preserved organelles (one cell is in the upper centre and the other is at the lower left corner). Apoptotic bodies (tab) can be observed. These cells show characteristic features of apoptotic cells including elongation, controtion and shrinking. They have intact borders retracted from the surrounding cells. Some vesiculation of ER and fat can be seen. The integrity of organelles and of the plasma membrane remains intact and the cells show dense chromatin aggregated against the nuclear envelope. The cytoplasm is condensed and dark. Magnification x 4425.
II9
Hepatic changes after ethanol exposure
the hepatic parenchyma that includes steatosis, alcoholic hepatitis and cirrhosis (Chedid et al., 1993; Danan, 1993; Lishcher et al., 1971). Also of interest are image analysis results that shows significant liver-cell enlargement after ethanol exposure. Hepatocytes enlargement in chronic exposure to alcohol was observed by Baraona et al. (1975) and Israel et al. (1982). We observed in Hep G2 cells exposed to ethanol the co-existance of normal size and appearance cells (Plate 2) with enlarged, functional cells (Plate 4) or shrunken, non-functional apoptotic cells (Plate 5). The dose-dependent damage, as observed either by reduction in cytoviability using XTT assay or by morphological changes, support the conclusion that the in vitro model described may be predictive. The potential usefulness of this model is increased by its reproducibility and the opportunity it presents for quantifying cell damage by morphometric analysis. In Hep G2 cells, the cytotoxic effects of ethanol, including steatosis and mitochondrial or endoplasmic reticulum changes, suggest more direct causes of alcohol-induced liver damage. The changes in cell organelles are paralleled by changes in SDH activity, suggesting that one of the consequences of EtOH-induced damage is affecting mitochondria. We have been able to show that the morphological changes in cultured Hep G2 cells exposed to EtOH are very similar to those observed in liver biopsies from humans with alcoholism (Neuman et al., 1993). These changes (e.g. steatosis and endoplasmic reticulum dilatation) can occur after exposure to ethanol for only a few days in culture. Experimental animal models for ethanol-induced liver damage that can reproduce features encountered in alcohol-induced liver injury in humans do already exist (Iimuro et al., 1996; Kalant et al., 1975; Lieber et al., 1989; Lindros et al., 1996; Rubin and Lieber, 1974). However, existing in vivo models of ethanol-induced liver damage have required prolonged periods (weeks) of ethanol exposure in order to induce changes compatible with EtOH-liver disease and are expensive (Rubin and Lieber, 1974; Tsukamoto et al., 1984). As the scientific world has become more aware of ethical issues about the use of laboratory animals, the need to develop appropriate corresponding in vitro models has been recognized. For the in vitro study of alcohol-related liver disease the model system should present a lesion that reproduces all the biochemical and morphological abnormalities seen in humans with the disease. In the present study we have focused on the morphometric differences between Hep G2 cells exposed and to EtOH and unexposed control cultures. Caution should be used in the extrapolation of data obtained in cultured cells to the much more complicated processes occurring in vivo in human hepatocytes, which preserve high cytochrome P-450 activities. However, the data described here indicate
121
that Hep G2 cells could be useful in assessing the mechanism(s) of the cytotoxic effects of ethanol. The results of our parallel study in which we used Hep G2 cells to evaluate the effects of therapeutic bile acids in modifying alcohol-induced toxicity in vitro suggest that this model has potential for the study of factors that may prevent or reverse alcohol-related hepatotoxicity (Neuman et al., 1995).
Acknowledgemenrs-The
authors thank Izabella Malkiewicz (Department of Clinical Pharmacology, Sunnybrook HSC) and Lois Lanes, Julia Hwong and Richard Leung (Departments of Pathology, Hospital for Sick Children and The Toronto Hospital) for the skilful technical assistance with EM preparation and to Fisher Co. for kindly providing us with the molecular device reader. The authors also acknowledge the following support: S. Bellentani-Fondo Studi Fegato (FSF), Trieste, Italy; R. G. Cameron-Dept of Pathology, Toronto Hospital Research Foundation; M. G. Neuman-personal award from University of Toronto Drug Safety Research Group; N. H. Shear--career scientist Ontario Ministry of Health, Canada; C. Tiribelli-MURST 40% (Fegato Cirrosi Epatica e Epatiti Virali), Roma and FSF, Trieste, Italy. REFERENCES Baraona E., Leo M. A., Borowsky S. A. and Lieber C. S. (1975) Alcoholic hepatomegaly: accumulation of protein in the liver. Science 190, 794795. Biava C. (1964) Mallory alcoholic hyalin: a heterofore unique lesion of hepatocellular ergastoplasm. Labbratory Investigation 13,.301-320.
Bradford M. (1976) A ranid and sensitive method for the quantification of’microgram quantities of protein utilizing the principle of protein dye binding. Analytical Biochemistry Ii, 248-236.
Bruauera M.. Bertran A.. Bombi J. A. and Rodes J. (1977) &ant mitochondria in hepatocytes. A diagnostic first for alcoholic liver disease. Gastroenrerology 73, 13831387.
Chedid A., Mendenhall C. L., Moritz T. E., French S. W., Chen T. S., Morgan T. R., Roselle G. A., Nemchausky B. A., Tamburo C. H. and Schiff E. R. (1993) Cellmediated hepatic injury in alcoholic liver disease. Gastroenterology 105, 294-296.
Danan G. (1993) Definitions et criteres d’imputation des atteintes hepatiques aigues medicamenteuses. Conclusions d’une reunion internationale de consensus. Gastroenterology and Clinical Biology 17, H 18-H2 1. Darlington G. J., Kelly J. H. and Buffone G. J. (1987) Growth and hepatospectrum gene expression of human hepatoma cells in a defined medium. In Vitro Cell Development and Biology 23, 349-354.
Dawson J. R., Adams D. J. and Wolf C. R. (1985) Induction of drug metabolizing enzymes in human liver cell line Hep G2. FEBS Letters 183, 219-222. Denk H., Franke W. W., Dragisics B. and Zeiler I. (1981) Pathology of cytoskeleton of liver cells: demonstration of Mallory bodies (alcoholic hyalin) in murine and human hepatocytes by immunofluorescence microscopy, using antibodies to cytokeratin polypeptides from hepatocvtes. HeDatolonv 1, 9-20. Diamond L.,. Kruszewski F., Aden D. P., Knowles B. B. and Baird W. M. (1980) Metabolic activation of benzo(a) pyrene by a human hepatoma cell line. Carcinogenesis 1, 871-875.
172
R. C. Cam
Doodstdr H., Duthie S. J., Burke M. D., Melvin W. T. and Grant M. H. (1988) The influence of culture medium composition on drug metabolising enzyme activities of the human liver derived Hep G2 cell line. F‘EBS Lerter.~ 1-2, IS- 18. Edmondson H. A. (1980) Pathology of alcoholism. Americuw
Journul
01 Clinicul
Pathologic
74, 725 742.
French S. W. (1981) The Mallory body: structure. composition and pathogenesis. Hepatolog~~ 1, 76-83 limuro Y., lkejima K.. Rose M. L., Bradford B. U. and Thurman R. G. (1996) Nimodipine. a dihydropyridinctype calcium channel blocker, prevents alcoholic hepatitis caused by chronic intragastrrc ethanol exposure rn the rat. Hc~prrto/o,ej~ 24, 39 I 397. lseri 0. A. and Gottlieb L. S. (1971) Alcoholic hyalin and megamitochondria as separated and distinct entities in liver disease associated with alcoholism. G‘crsrroenterolog~ 60, 1027~1035. Iseri 0. A., Lieber C. S. and Gottlieb L. S. (1964) The ultrastructure of fatty liver induced by prolonged ethanol ingestion. ilnwricrm Journul of Pathology. 48, 535 555. Ishii H.. Joly J. and Lieber C. S. (1973) Effect of ethanol on the amount and enzyme activities of hepatic rough and smooth microsomal membrane. Biwhimk et Bioph.wica Acttr 291, 41 l-420. Israel Y., Britton R. and Orrego H. (1982) Liver- cell enlargement induced by chronic alcohol consumption: studies on its cause and consequences. Clink/ Biochemi.Ytr~~ 15, I899 I92 Jaen N. C. (1971) Direct blood Injectton method ]‘<)I-g;ta chromatographic determination of alcohols and other volatile compounds. Clirricnl Cl7cwi.cty 17, 82 X5. Kalant H.. Orrego H.. Khanna J. M. and Phrlhps J. M. (1975) Experimental alcohol-induced hepatic necrosis: suppression by propylthiouracil. Procc&irrg.r o/’ r/re Nutiontrl Acuden~~~ of Scietucc of the Y.S. A 12, 1137~1141. Konrad K. D. and Reed G. (1977) Organ culture of portnatal rat liver: effect of ethanol on lipid metabolism Esl,erimental
and Moieculur
Putholog~~ 26, 425 433.
286. Lindros K. O., Badger T., Ronis M.. Ingelman-Sundbel-p M. and Koivusalo M. (1996) Phenyl-lsocyanate. a new dietary liver aldehyde dehydrogenase inhibitor. .Iorrrrrtr/ Phurmacolo~q>~ (tnd 79-83.
E.vperimenttrl
Theruputic~t
cell injuries in the human Hep G2 cell line. 109, 5555563. Neuman M. G.. Koren G. and Tiribelli C. (1993) In vitro assessment of the ethanol-induced hepatotoxicity on Hep G2 cell line. Biochemical and Biophysical Research Communicuiions
2, 932-942.
Neuman M. G. and Tiribelli able in preclinical stage?
C. (1995) Drugs are predictModels and techniques. In
Liver tmd Drup ‘94. Progrers 1, pp. 339. Bo-Art, Brdtisiava.
in Hepato-Pharmacoll~g~,
Phillips M. J., Poucell S.. Pattersson J. and Valencia P. ( 1987) Tiw Liver. An Arlus of L~/fru.structurul Puthology. Raven Press. New York. Roehm N. W., Rogers G. H.. Hatfield S. M. and Glasebrook A. L. (1991) An improved calorimetric assay for cell proliferation and viability utilizing the tetrazolium salt XTT. Journal of Immunologicul Methotlv 142, 251-265.
Rubin E. and Lieber C. S. (1968) Alcohol induced injury in non-alcoholic volunteers. Nen England of’ Medicine
hepatic Journal
278, 869-876.
Rubin E. and Lieber C. S. (1974) Fatty liver, aicohoiic hepatitrs and cirrhosis produced by alcohol in primates. ,Yru, England Journul of‘ Medicine 290, 128-I 35. Sassa S., Sugita O., Galbraith R. A. and Kappas A. (1987) Drug metabolism by the human hepatoma cell, Hep G2. Biochemiccd 143, 52-57.
und
Biophysiccd
Research
Communicution.c
Schaffner F., Loebel A.. Weiner H. A. and Barka T. (1963) Hepatocellular cytoplasmic changes in acute alcoholic hepatitis. Journal of rhr Americun Mrdirul Associution 183, 343-349.
Scheuer P. J. (1982) The morphology of alcoholic liver disease. British Medicul Bulletin 38, 63-65. Shear N. H., Malkiewicz I., Klein D., Randor S., Koren G. and Neuman M. G. (1995) Acetaminophen-induced toxicity to keratinocytes and to hepatocytes in vitro is diminished by Silymarin. Skin Pharmacology 8, 45-56. Smuckler E. A. (1968) The ultrastructure of human alcoholic hyaline. .Amcricttn Journul of PathoLogl, 49, 790-~ 797.
Lieber C. S.. Baraona E.. Hernandez-Munoz R. and Kubota S. (1989) Impaired oxygen utilization 4 new mechanism for the hepatotoxicity of ethanol in subhuman primates. Journul of’ C/iniu/ Inwvtigtrtio~~ 83, 16X2-1690. Limbosch S. (1983) Benzo (a)pyrene and aldrin metabohzing activities in cultured human and rat hepatoma cell line. Journtrl of iIre Nrrtionul Canwr Ir~stitute 71, 2X 1
of
induced
Gasiroenterolog~~
275.
Lischner M. W.. Alexander _I. F. and Galambos J. T. (1971) Natural history of alcoholic hepatitis. I. The acute disease. Americctn Journul of’ Digestiw Di.trrrs~.~ 16, 48 I 494. Mattews. D. E. and Farewell. V. (19X5) L:sirq wu/ U~der.r/unding Mediical .Stu/i.rfic.s pp. I 12 I I6 and I3 129. S. Karger, AG. Basel. Meldolesi J. (1967) On the significance of the hypertrophy of the smooth endoplasmic reticulum in liver cells after administration of drugs. Biochemical Pharmucolog~ 16, 125-131.
Neuman M. G.. Cameron R. G., Shear N. H., Bellentam S. and Tiribelli C. (1995) Effect of Tauroursodeoxycholate and ursodeoxycholic acid on ethanol-
Sottocasa G. L.. Kuitenstierna B., Ernster L. and Bergstrand L. (1967) An electron transport system associated with the outer membrane of liver mitochondria: a biochemical and morphological study. Journul qf’ Cell Biology
32, 415-438.
Stevens M. G. and Olsen S. C. (1993) Comparative analysis of using MTT and XTT in calorimetric assays for quantitating bovine neutrophil bacterial activity. Journul
of
Immunologicul
Methods
157, 225-233.
Stewart R. V. and Dincsoy H. P. (1982) The significance of giant mitochondria in liver biopsies as observed by light microscopy. American Journal of Clinicul Purholog~~ 78, 293-298.
Tsukamoto H., Reiderberger R. D., French Largman C. (1984) Long-term cannulation blood sampling and intragastric infusion American
Journal
of Phvsiologv
S. W. and model for in the rat.
241, R595-R599.
Walker F., Elmslie N., Fraser- R. A., Snape P. E. and Watt B. C. (1974) Cvtotoxic effect of alcohol on liver cells and tibroblast rn*vitro. Scottish Medic,ul Journcd 19, IL- 127. Wilson M. A., Cascarano J., Wooten W. I. and Pickett C. B. (1978) Quantitative isolation of liver mitochondria by zonal centrifugation. Analytical Biochemistry 85, 225-264.
Yokoo H., Minick 0. T.. Batti F. and Kent Morphological variants of alcoholic hyalin. Journal
of Pathology,
69. 25-32.
G. (1972) Americrm