Acetaldehyde Selectively Stimulates Collagen Production in Cultured Rat Liver Fat-storing Cells but not in Hepatocytes HAN MOSHACE,ALESWDRO CASINIAND CHARLESS. LIEBER Alcohol Research and Treatment Center and Section of Liver Disease and Nutrition, Bronx Vetemns Administrcrtion Medical Center, Bronx, NY 10468 and Mount Sinai School of Medicine (CUNY), New York, NY 10029

Hepatocytee and fat-storing cells have been implicated in the production of collagen, under both normal and pathological conditions. In thin study, short-term primary cultures of rat hepatocyh-t, maintained in a serum-free,h o r m o d y defined medium without dexamethasone and cultured on a flbronectin-collagemtype IV eubetratum, were wed. Primary and p1and 2 cultures of fat-storing cells maintained on t h e culture plastic were also studied. Hepatocytes produced significant amounts of collagen type m, but formation of collagen type I was not detectable. Laminin and collagen type IV production were very low. Hepatocytes maintained their ability to metabolizeethanol (at levels comparable to those observed at 2 hr)for at lea&48 hr after plating and this metabolism was inhibited 86% to 95% by Pmeth3lpyraaole (1 mmoUL). Neither ethanol (50 mmoUL) nor acetaldehyde (175 pmoUL, initial concentration) had any effect on the production of collagen type m or laminin. Fat-doring cells (95% to 100%deemin-positive) produced significant amounts of both type I and type III collagen. Production of collagen type IV and laminin was very low. M e t a b o h of ethanol by these cultures was not detected. Addition of ethanol had no effect on collagen or 7 ' ' production in fat-storing cella In contrast,acetaldehyde significantlyincreased the production of collagen type 1, but did not alter the production of collagen type III, IV or laminin. Incorporation of %-proline into total protein was not affected by addition of ethanol or acetaldehyde to fat-doring cells or hepatocytee. Jkpomwe of fat-storing cells to ethanol or acetaldehyde did not change .Hcollwn degradhg activity in the media. We conclude that fat-storing cells are likely effector cella in the

Received November 8, 1989; accepted March 27, 1990. WBB supported by Department of Health and Human Services granta No. DKS2810 and AA03508, the Department of Veterans Affairs, the Alcoholic Beverage M e d i d Research Foundation and a NATO-fellowship provided by the Netherlands organization for Scientific Reeearch (to H.M.). A preliminary account of this work wae presented at the annual meeting of the American Asaoeiation for the Study of Liver Diaeaees, Chiesgo. Illinois, October 31, 1989 (HEPATO~OCU 1989;10:6301. Address reprint requeata to: Charlea S. Lieber. M.D. (ISUG), Alcohol Reeearch and Treatment Center, Bronx VA Medical Center, 130 West Kingsbriw Road, Bronx. New York 10468.

This resear&

31/1/a%888

increased production of collagen type I in alcoholic liver fibrosis. (HEPATOLOGY 1990;12:511-518.)

The deposition of collagen in the Disse space and around the hepatocytes represents an important morphological feature of alcohol-induced liver injury, but the mechanisms responsible for this increased hepatic collagen deposition remain unknown. Alcohol-induced liver fibrosis may occur in the absence of polymorphonuclear inflammation, both in baboons (1,2)and in man (3).These findings, therefore, suggest that mechanisms other than such inflammation can induce the hepatic fibrosis. Fat-storing cells (FSC, also called Ito cells or perisinusoidal lipocytes), the principal cells residing in the Disse space of the liver, have been considered to play an important role in the development of alcohol-induced liver fibrosis (4, 5). FSCs can transform into activated transitional cells in experimental (6, 7) and human ( 8 ) alcoholic liver injury. Furthermore, it has been demonstrated that FSCs can synthesize and release M e r e n t types of collagen, either in uiuo or in uitm (9-12). Therefore activated FSCs have been incriminated in the alcohol-induced perisinusoidal fibrosis. The lack of any stimulatory effect of ethanol on collagen production in cultured human liver, skin, fetal (13) and lung fibroblasts (14) and baboon liver myofibroblasts (15)suggests that ethanol itself does not directly induce hepatic fibrogenesis. On the other hand, it has been reported that acetaldehyde and lactate stimulate collagen production in baboon liver myofibroblasts (16) and in human skin and hepatic fibroblasts (13). Acetaldehyde itself also increases collagen gene transcription in cultured human fibroblasts (16). The effects of ethanol and its metabolites on proteoglycan synthesis and proliferation of cultured rat liver FSCs have been evaluated (17), but thus far no stimulation, by ethanol or acetaldehyde, of the synthesis of different types of collagen by normal FSC has been reported. Also, no definitive data exist about the possible involvement of hepatocytes in the ethanol-induced liver fibrogenesis. Conflictingresults have been reported with regard to the liver cell types responsible for the increased collagen deposition and the synthesis of different col-

511

512

MOSHAGE. CASINI AND LIEBER

lagen types. In particular, it has not been clarified as yet whether hepatocytes are involved (9, 18-20) or not (10,11, 21,22)in liver collagen synthesis. We report here (with the use of an ELISA method) that primary cultures of adult rat hepatocytes produce only collagen type 111, whereas collagen type I, 111, IV and laminin are detectable in FSC cultures. We also describe the effects of ethanol and acetaldehyde on the synthesis of different types of collagen and laminin in cultured rat liver FSCs and in primary culture of adult rat hepatocytes. MATERIALS AND METHODS Materials. Williams medium E, Dulbecco's modified Eagle's

HEPATOLOGY

resuspended in HBSS and FSCs were obtained by centrifugation over a 11.4%(wt/vol)Nycodenz gradient (25) for 17 rnin at 1,400g.After centrifugation, FSCs were collected from the interface, washed with HBSS and resuspended in Dulbecco's modified Eagle's medium (Grand Island Biological Co.) containing 20% FCS, 2 mmol/L L-glutamine and antibiotics, at a concentration of 2 x lo-' cells/ml. Viability of the cells, evaluated by the trypan blue exclusion test, was > 90%. Cells were then seeded on 35 mm plastic tissue culture dishes at a cells/lO cm2 and incubated at starting density of 1 x 37" C in a 5% C0,-air humidified atmosphere. The medium was changed daily. Subcultures were obtained by trypsinization using a 0.025% trypsin solution containing 0.01% EDTA (Grand Island Biological Co.). FSCs were identified by their typical phase-contrast light microscopic appearance and the constant positive immunofluorescence staining for desmin that was maintained even after the first and second passages in vztro (26). Indirect Immunofluorescence. For indirect immunofluorescence studies, small aliquots of cells were cultured on Lab-Tek chamber slides (Thomas Scientific, Swedesboro, NJ). Slides were air-dried for 2 hr, fixed in cold acetone/chloroform (1: 1, vol/vol) for 20 min and then incubated with monoclonal antidesmin antibody (1: 10 diluted in PBS) for 30 min at room temperature in a humidified chamber. Slides were then washed three times, for 5 min each time, in PBS and incubated for a further 30 min at room temperature with FITCconjugated antimouse IgG-F(ab')2, diluted 1:20 in PBS containing 10% normal rat serum. Controls were carried out by omitting the first antibody and including instead mouse total IgG at the same dilution as used for the monoclonal antibody. After mounting coverslipswith 90% glycerol in PBS, the slides were examined with a Zeiss fluorescence microscope (Carl Zeiss Inc., Thornwood, NY) equipped with epiillumination. Hepatocyte Cultures. Hepatocytes were isolated from young female Sprague-Dawley rats (175 to 200 gm body weight) by a two-step perfusion method using collagenase as described originally by Seglen (27). The liver was first perfused through the portal vein with Ca' '/Mg' '-free HBSS, pH 7.4, maintained at 37" C (10 min, flow rate = 20 ml/min), followed by the same buffer containing Ca ' (5.7 mmoUL) and collagenase (150 U/ml; 10 min, flow rate = 8 mumin), using the cannulated superior vena cava as the outflow port. The liver was then removed and placed in HBSS. Hepatocytes were released by gentle teasing of the softened liver and filtered through a 60 mesh sterile nylon gauze. Cells were then washed three times with HBSS at 50 g for 5 min and the supernatant, containing sinusoidal cells, was discarded. The final cell pellet was resuspended in serum-free Williams E medium containing antibiotics and the following supplements (28): epidermal growth factor (25 ng/ml), glucagon (10 nmoVL), thyroxine (10 pmoVL), insulin (5 pg/ml), transferrin (5 pg/ml), sodium selenite (5 ng/ml), nicotinamide (10 mmol/LJ, CuSO, (0.1 pmol/L), growth hormone (10mU/ml), prolactin (20 mU/ml) and L-glutamine (2 mmol/L). Cell viability, assessed by the trypan blue exclusion test, exceeded 90%. Hepatocytes were plated on 35 mm plastic culture dishes coated with a defined matrix composed of collagen type IV (50 pg/dish) and fibronectin (15 pg/dish). Cells were allowed to attach for 2 to 4 hr at 37" C in a humidified 5%COJair atmosphere; unattached cells were then removed and fresh medium was added.

medium, heat-inactivated fetal calf serum (FCS), L-glutamine, HBSS, penicillin/streptomycin, fungizone and sodium bicarbonate were purchased from Grand Island Biological Co., (Grand Island, NY).Epidermal growth factor, insulin, transferrin, sodium selenite, L-thyroxine, nicotinamide, glucagon, growth hormone, prolactin, sodium ascorbate, paminoproprionitrile (P-APN),N-ethylmaleimide,PMSF, collagenase type I, protease type XIV (pronase E), collagen type I11 and IV (human placenta), collagen type I (rat tail),fibronectin (bovine plasma), monoclonal antidesmin antibody (clone DEU-101, FITC-conjugated anti-mouse IgG F(ab'12 and alkaline phosphatase-conjugated goat antirabbit IgG were from Sigma Chemical Co. (St. Louis, MO). Type-specific rabbit polyclonal antibodies against rat collagen type I, I11 and IV were obtained from the Pasteur Institute (Lyon, France) (23). Rabbit antimouse laminin polyclonal antibody was kindly provided by Dr.D. Schuppan, Department of Gastroenterology, Klinikum Steglitz, Free University of Berlin, West Berlin, FRG. Nycodenz was from Accurate Chemical & Scientific Co. (Westbury, NY). Tissue culture plastic dishes were from the Falcon Division of Becton Dickinson & Co. (Oxnard, CAI. Immunoassay plates (96 wells, flat bottom) were purchased from Biorad Laboratories (Richmond, CAI. ~ 4 23, , 4, 5-3H]proline (specific activity = 100 Ci/mmol) was from ICN Radiochemicals, Irvine, CA. [ l , 2,-"C]-ethanol (specific activity = 1.08 mCi/ml or 62.8 pCi/mmol) was from Research Products International Corp. (Mount Prospect, IL). N-[propionate-2, 3-3H]-propionylatedrat type I collagen (specific activity = 100 Ci/mmol) was from DuPont NEN Medical Products (Boston, MA). Acetaldehyde was obtained from Eastman Kodak Co. (Rochester, NY).Ethanol and 4-methylpyrazole were from Aldrich Chemical Co. (Milwaukee,WI). Preparation of Fat-Storing Cells. FSCs were isolated from old female Sprague-Dawley rats (500 to 700 gm body weight) that had free access to water and Purina Chow diet. These animal studies were carried out in accordance with the guidelines of our Institutional Animal Committee. Nonparenchymal liver cells were isolated by the pronase-collagenase method of b o o k , Seffelaar and DeLeeuw (24) with some minor modifications. After the animals were anesthetized with pentobarbital, the abdomen was opened and the portal vein was canndated; the liver was perfused with Ca + * /Mg ' -free HBSS, pH 7.4, for 10 min at a flow rate of 10 mVmin to wash the blood out. The liver was then perfused with HBSS containing 0.1% pronase E and 0.05% collagenase type I (both from Sigma Chemical Co.) for another 10 min at the same flow Cell Culture Zncubatiom with Ethanol and Acetaldehyde. rate. After perfusion, the liver was removed, cut into small pieces and incubated in HBSS containing 0.05% collagenase For ethanol and acetaldehyde incubations, airtight 35 mm and 0.02% pronase at 37" C for 20 min with shaking. After culture dishes were used. Hepatocytes (primary cultures) and passing through a filter (mesh size 200 pm), cells were washed, FSCs (primary, passage 1and 2 cultures) were incubated with +

Vol. 12, No. 3, 1990

ACETALDEHYDE STIMULATES FAT-STORING CELL COLLAGEN PRODUCTION

513

FIG.1. Phase contrast micrograph ( A ) of FSCs, cultured on tissue culture plastic, 96 hr after d n g . Lipid droplets are visible in the perinuclear zone of the cells (original magnification x 100). Immunofluorescence micrograph (B) of FSCs (passage 1). Desmin staining was performed as described in "Materials and Methods" (original magnification x 200).Phase contrast micrograph (C) of hepatocytes cultured in hormonally defined medium on a defined matrix composed of collagen type IV (50 &dish) and fibronectin (15 & l i s h ) 72 hr after seeding (original magnification x 100).

50 mmol/L ethanol or 175 Fmol/L acetaldehyde (initial concentrations) in serum-free medium containing 50 &ml ascorbic acid and 100 pg/ml P-APN for 24 hr and 48 hr at 37" C. Ascorbic acid and P-APN were added to increase the collagen proline hydroxylation and secretion and to prevent collagen cross-linking. Experiments with hepatocytes started 24 hr after cell seeding. Experiments with FSCs started when the cultures reached confluency. In the incubations of hepatocytes with acetaldehyde, 4-methylpyrazole and cyanamide (finalconcentrations of 2 mmolL and 100 FmoVL respectively) were added to the medium to prevent metabolism of acetaldehyde. As controls, hepatocytes and FSCs were cultured in the same media without ethanol or acetaldehyde. Each experiment (control,ethanol and acetaldehyde) was performed with cells of the same isolate and from the same passage. The media were changed after the first 24 hr and the experiments continued using fresh culture medium and fresh additives for another 24 hr. The concentrations of ethanol and acetaldehyde were monitored by head space gas chromatography as described before (15). Supernatants were collected and stored at - 20" C. Cell layers were scraped off, sonicated and stored at - 20" C. To determine the amount of collagen already present in the

cells at the beginning of each experiment, the cell layers of two dishes per experiment were harvested at t = 0 and stored at - 20" C. The cell layer collagen content at t = 0 was subtracted from the total cell collagen content. ColZugen Assay. Collagen was determined in culture media and cell layers by an ELISA method adapted from Rennard et al. (29),using the direct method. The alkaline phosphatase method was found to be slightly more sensitive than the horseradish peroxidase method. Standards and samples were assayed in triplicate. ELISA immunoassay plates were coated overnight at 4" C with 150 ~1 of the appropriate dilutions of medium and cell layer samples in 0.1 m o m carbonatehicarbonate buffer, pH 9.6. After coating, the plates were incubated for 1.5hr at 37" C with 200 Fl/well of PBS containing 5% drymilk (PBS-Blotto),to block uncoated sites; 100 @well type-specific collagen antibodies diluted in PBSBlotto; 100 clliwell goat antirabbit IgG-alkaline phosphatase conjugated complex, diluted in PBS-Blotto (5%) and finally with 100 &well 10% diethanolamine buffer, pH 9.8, containing 50 pmol/L Mg' + and 1 mg/ml p-nitrophenylphosphate as substrate for alkaline phosphatase. Between all the steps, plates were washed five times with saline containing 0.05% Tween-

514

MOSHAGE, CASINI AND LIEBER

HEPATOLOGY

TABLE 1. %-Ethanol metabolism in cultured rat hepatocytee Time (hr)

SemDle

2 24 48

634 ? 230 615 ? 147 546 f 174

Sample

Inhibition

+ 4-MP

(%)

57 2 2 77 a 14 45 f 2

89.5 f 3.5 91.5 f 2.8 91.5 ? 2.5

Data (mean values f S.E.M.) are expressed ae picomole of ethanol per microgram of DNA per minute. Results were obtained according to the following calculations:

where: net dpm = dpm sample - dprn background. Time indicated is time after the attachment period (2 to 3 hr after seeding). 4-MP = 4methylpyrazole. TABLE 2. Collagen and laminin production in FSC cultures

Type1 Type I11 Typem Laminin

948 390 0.9 8

350-2,210 265-517 0-3.1 1-27

880 365 0.4 7

223-1,320 214-504 0-1.2 1-37

Data are expremed ae nanograms per microgrm of cellular DNA and represent the amount of collagens and laminin detected in the culture media.Collagen type I, 111, IV and laminin in the cell layers were legs than 10% of total recovered collagen and laminin. Determination~in primary, passage 1 and 2 cultures were averaged. Data are mean and range of five experiments.

20. After development of the color at room temperature, the optical density was determined at 405 nm. The detection limits for collagen type I, 111, IV and laminin were 2, 2.5, 0.25 and 0.25 d w e l l , respectively. Variability of results for triplicate samples was less than 8%. No cross-activity was deteded between collagen and nonspecific collagen antibodies, even when a 20-fold higher amount of antigen was used than with the highest specific standard. Results were expressed as nanograms of mllagen per microgram of cellular DNA. Total PtPtein S y n t b b . The incorporation of tritiated proline into protein was used to measure total protein synthesis in the cell cultures. ~ - [ 2 ,3, 4, 5-'H]-proline (10 pCi/ml; final concentration 'H-proline: 0.1 pmoVL; final concentration cold proline: 0.25 mmoUL) was added to cultures of hepatocytes and FSCs in the last 6 hr of the total 48 hr incubation period. Media and cells were then harvested separately and stored at -20" C. The total incorporation of 'H-proline into protein was determined by trichloroacetic acid precipitation as previously described (30). The free intracellular proline pool was also determined in the cultures. Cells were harvested, submitted to 10% trichloroacetic precipitation and proline was measured in the resulting supernatants by a gas chromatographic-mass spectrometric procedure (31). Separation of amino acids were performed with a Hewlett-Packard 5890A gas chromatograph (Hewlett-Packard Co., Analytical Products, Palo Alto, CAI on a 25 m by 0.2 mm fused silica HP-1 capillary column (Hewlett-Packard Co.) (0.5 pm film thickness) with a spitless injection. Gas chromatographic-maas spectrometric measurement was made with a Hewlett Packard 5970A mass selective detector (Hewlett-Packard Co.).

Ethanol Metabolbm. Ethanol metabolism in the cell cultures was determined as described elsewhere (32) with some modification for its use in monolayer cultures. Hepatocytes and FSCs were cultured in multiwell plates in 200 pl serum-free medium containing "C-ethanol (final concentration = 50 mmoVL; corresponding to 3.14 pCi/ml medium) for 2 hr at 37" C. To assess the contribution of alcohol dehydrogenaseto ethanol metabolism, 4-methylpyrazole(final concentration = 1 mmom) was added to some wells. Background was represented by the amount of radioactivity deteded in the cell layers harvested at t = 0. Hepatocyte experiments were carried out 2 hr, 24 hr and 48 hr after the attachment period (2 to 3 hr after cell seeding). FSCs were tested for their ability to metabolize ethanol at the same time as the collagen production studies. AU assays were performed in quadruplicate. Results were expressed as picomole of ethanol per microgram of DNA per minute. DNA Determination. The amount of DNA in the cells, harvested at the end of the experiments, was determined by a fluorimetric assay as previously described (33).DNA content was determined for cell culture conditions at the end of all the experiments. thZ&genase h n a y . Collagenase activity in cell culture media was determined as described by Hu, Crombie and Franzblau (34) with some minor modifications: 0.1 pCi 'H-labeled collagen and 50 pg cold rat t ail collagen type I (Sigma Chemical Co.) were incubated with 100 p1 of media in a reaction mixture containing 50 mmoVL Tris-HC1 (pH 7.41, 0.2 moVL NaC1, 5 mmoVL CaCl,, 0.2 mmoVL PMSF and 5 mmoW N-ethylmaleimide in a final volume of 600 111for 1h r at room temperature with constant shaking. To determine the total enzyme activity, aminophenylmercuric acetate (final concentration = 1mmoVL) was added to the reaction mixture to activate any latent enzyme that might be present (35). The same reaction mixture in which samples were replaced by 100 p1 test buffer represented the blank. As positive control, 15 pg of collagenase type I (Sigma Chemical Co.) were added. After the intact collagen was precipitated by centrifugation at 10,000 g for 10 min, aliquots of the supernatants were counted in a liquid scintillation counter. Statistics. Data were expressed as mean S.E.M. Statistical analysis was performed by ANOVA (36).

*

RESULTS Characterization of Cultured Cells. Primary cultures of FSCs reached confluency after 14.5 & 1.9 days. Passage 1 and 2 subcultures were confluent after 8.8 f 0.9 days. When examined by phase-contrast microscopy, they had the typical stellate shape and many

515

ACETALDEHYDE STIMULATES FAT-STORING CELL COLLAGEN PRODUCTION

V O ~12, . NO. 3,1990

0E T W L

1In

I ACETALDEHYDE

T

2 COLLAGEN I

I

DAY

I

DAY

COCLAGEN

2

IU

;1 I

DAY

2

LAMININ

FIG.2. Effects of ethanol and acetaldehyde on collagen and laminin production by FSCs. Data are mean 5 S.E.M. of five experiments. Data ~ DNA and e x p d as percentage of control. The effect of acetaldehyde on type I collagen production waa were calculated as n g / cellular significant (p c 0.01).Day 1 and 2 refers to the first and second 24-hr incubation periods of cultures with ethanol or acetaldehyde.

TABLE3. Collagen and laminin production in hepatocyte culturea uhr

M-

N.D. 58.4 0.5 0.4

Type1 Type I11 TypeN Laminin

48hr

Rrppe

Mean

Ranpe

-

N.D. 58.3 0.4 0.5

42-83 0.3-1 0.4-0.6

30-113 0.3-1 0.4-0.5

Data (expressedas nanograms per micrograms of cellular DNA) are from h e p a b y t e culture supematanta.Collagen and laminin in the cell monolayera were less than 10% of total recovered collagen and laminin. Type I collagen waa not detectable (N.D.). Laminin and collagen type N accumulation in the culture media waa very low. Data are mean and range of five experiments.

lipid droplets were visible in the perinuclear zone (Fig. 1A). Desmin positivity was 93% to 95% in primary FSC cultures and 100% after the first and second passage (Fig. 1B). Hepatocytes were easily recognizable by phase-contrast microscopy (Fig. 1C). Desmin staining was also performed in hepatocyte cultures to check for the possible presence of FSCs in the monolayer, which could be responsible for the collagen synthesis (21). Primary cultures of hepatocytes contained less than 2% desmin-positive cells at the end of the 48 hr incubation time with ethanol or acetaldehyde (72 hr after seeding on plastic dishes). Ethanol and Acetaldeh* in CeU culturn Medium. To avoid evaporation of ethanol, and particularly evaporation of acetaldehyde, culture dishes were tightly capped during the incubation. After 24 hr, the concentration of ethanol was 40 2 2.2 mmoVL in FSC cultures and 37 f 2.5 mmoVL in hepatocyte cultures. The concentration of acetaldehyde in FSC culture media was 65.3 & 0.73 pmoVL after 2 hr incubation and 42 5 9 pmoUL after 24 hr. Acetaldehyde concentration in hepatocyte cultures was 56.2 2 3.4 pmoVL after 2 hr incubation and 44 7.5 pmoVL after 24 hr. Medium with 50 mmoVL ethanol contained no more than 5 pmoVL of acetaldehyde and medium with 175

*

pmoVL acetaldehyde contained no more than 50 pmoVL ethanol. Hejuatotyte and FSC Ethanol Metabolism. The rate of ethanol metabolism in hepatocyte cultures is reported in Table 1.Hepatocytes maintained their ability to metabolize ethanol 48 hr after plating at levels comparable to those observed at 2 hr. This metabolism was markedly decreased (86%to 90%) by 4-methylpyrazole (1mmoVL). Cultured FSCs showed no evidence of ethanol metabolism. Over the 2 hr period for the determination of ethanol metabolism, ethanol concentration dropped from 50 mmoVL to 48 1 mmoVL (96% f 2% of the initial value). Hepcrtoeyte and FSC Corkrgen Productto * w Primary as well as passage 1 and 2 FSC cultures produced SigNficant amounts of both type I and type I11 collagen; production of collagen type IV and laminin was much lower (Table 2). Primary cultures of hepatmytea synthesized e m c a n t amounts only of collagen type 111, whereas collagen type I was undetectable; laminin and c o l l w n type IV production was very low (Table 3). Collagens and laminin in the cell monolayers of both FSC and hepatocyte cultures were less than 10%of the total recovered collagens or laminin. Addition of ethanol had no effect on collagen or

516

HEPATOLOGY

MOSHAGE, CASINI AND LIEBER

TABLE 4. Total protein synthesis in hepatocyte and FSC cultures Hepatocytes

FSCs

Variables

Media

Cells

Media

Cells

Control Ethanol Acetaldehyde

100 94 2 9 95 2 11

100 97 -t 10 108 5 10

100 82 2 20 1 1 2 2 12

100 110 2 13 97 2 16

Data are calculated as disintegrations per minute per microgram cellular DNA and expressed as percentage of control. Mean value -t S.E.M. of five experiments. Control values (100%) for 3H-proline incorporation into protein were 6,325 2 427 and 1,874 2 205 d p d k g cellular DNA in hepatocyte and FSC cultures, respectively. Ethanol = 50 mmoVL; acetaldehyde = 175 FmoVL.

TABLE 5. Effect of ethanol and acetaldehyde on the intracellular free-proline pool in hepatocyte and FSC cultures Variables

Hepatocytes

FSCs

Control Ethanol Acetaldehyde

2.16 0.14 2.23 2 0.49 2.20 5 0.18

0.63 2 0.05 0.65 5 0.14 0.58 2 0.11

Data are expressed as nanomoles per microgram of cellular DNA. Mean value 5 S.E.M. of four experiments.

laminin production in FSC cultures. In contrast, acetaldehyde significantly(p c 0.01)enhanced the production of collagen type I but did not modify the FSC production of collagen type I11 and laminin (Fig. 2). Collagen type IV values were too close to the detection limit of the ELISA assay to be quantitated. The effect of acetaldehyde on collagen type I was evident after 24 and 48 hr in primary cultures and in passage 1and 2 cultures. Neither ethanol nor acetaldehyde significantly affected the production of collagen type I11 in the primary cultures of hepatocytes: 85% & 8% and 80% 2 21% of controls 24 and 48 hr (after ethanol incubation, respectively; 102% 5%and 91% & 3% of controls 24 and 48 hr after acetaldehyde incubation, respectively (mean 5 S.E.M.). Type IV collagen and laminin values in the media of hepatocyte cultures were too low to be quantitated. No significant differences were found between primary and passaged cultures in the absolute amount of collagen produced or the extent of stimulation by acetaldehyde. DNA content of hepatocyte and FSC cultures was not significantly affected by ethanol or acetaldehyde (data not shown). Total Protein Synthesis. As shown in Table 4, the incorporation of 'H-proline into total protein was not affected by exposure of FSCs or hepatocytes to ethanol or acetaldehyde. Ethanol and acetaldehyde did not significantly modify the free proline pool either in FSC or in hepatocyte cultures (Table 5). Cblhgen &grading Activity. Ethanol and acetaldehyde did not sigdicantly modify the 3H-collagen

*

TABLE 6. Collagenolytic activity in FSC media Variables

24 br

48 hr

Control Ethanol Acetaldehyde

100 102 4 98 2 9

100 100 2 2 99 5 5

*

*

Data are expressed as percentage of control. Mean value S.E.M. of five experiments. The collagenolytic activity in the controls, calculated as reported in Table 1, was: 45.1 2 0.7 and 42.5 1.2 fmoVpg DNA/hr at 24 and 48 hr, respectively.

*

degrading activity in FSC and hepatocyte culture media after 24 and 48 hr, compared with controls (Table 6).

DISCUSSION This study demonstrates that acetaldehyde doubles the production of collagen type I in cultured rat liver FSCs. This effect was evident both after 24 and 48 h r of incubation. FSCs also synthesized significant amounts of collagen type I11 and lesser amounts of collagen type IV and laminin, but only type I collagen production was enhanced by acetaldehyde, whereas total protein synthesis was not modified. In the same experimental model, ethanol itself did not significantly affect the production of collagens and laminin. Because the collagenolytic activity was not decreased in the cell media after acetaldehyde exposure, the enhancement of collagen accumulation appears to have been caused by an active production by FSCs. These in uitro findings confirm previous in uivo studies that reported that hepatic fibrosis is associated with transformation of lipocytes to transitional cells surrounded by collagen fibers and characterized by a depletion of lipid droplets and a hypertrophy of the rough endoplasmic reticulum, both in baboons (7)and in man (6). At variance with our findings, Shiratori et al. (37) reported that acetaldehyde increases collagen production in cultured FSCs isolated from carbon tetrachloride-treated rats but not in FSCs of normal animals. However, in the study of Shiratori et al., FSCs were exposed to acetaldehyde 18 hr after isolation, whereas in our experiments, collagen production in FSCs was determined when the cultures had reached confluency (14.5 & 1.9 days after isolation). It is not unlikely that during this period, a partial shift toward a transitional cell-like phenotype occurs, accompanied by a gradual increase in collagen production, as has recently been demonstrated by Geerts et al. (38). This could also explain why no significant difference was observed in the absolute amount of collagen produced and the extent of stimulation by acetaldehyde between primary cultures of FSCs and passage 1 or 2 cultures. It should also be pointed out that in all our experiments FSCs maintained desmin-positivity, both in primary and in passage 1or 2 cultures, suggesting a FSC phenotype. Savolainen and coworkers (15)reported that acetaldehyde and lactate, but not ethanol, stimulate total collagen synthesis of cultured baboon liver myofibro-

V O ~12, . NO. 3, 1990

ACETALDEHYDE STIMULATES FAT-STORING CELL COLLAGEN PRODUCTION

517

Acknow1edgments:We thank Drs. N. Funaki and blasts. Data presented here extend these findings to Y. Tanaka for assistance in the free proline assay and Dr. FSCs, which can transform into myofibroblast-like transitional cells in alcohol-induced liver fibrosis (6, 7). D. Schuppan for the gift of the antilaminin antibody. Furthermore, other in uitro experimental models have REFERENCES shown that acetaldehyde stimulates the synthesis of 1 Lieber CS, DeCarli LM. An experimental model of alcohol feeding collagen by fibroblasts (13, 16). and liver injury in the baboon. J Med Primatol 1974;3:153-163. The ethanol concentration was easily maintained at Popper H, Lieber CS. Histogenesis of alcoholic fibrosis and constant levels in FSC culture media because ethanol 2 cirrhosis in baboon. Am J Pathol 1980;98:695-716. itself was not significantly metabolized by these cells. By 3 Nakano M, Worner TM, Lieber CS. Perivenular fibrosis in contrast, acetaldehyde levels rapidly decreased in media, alcoholic liver injury: ultrastructure and histologic progression. Gastroenterology 1982;83:777-785. but never dropped below about 40 p.mol/L after 24 hr of Y,Hashamura Y,Takeuchi J . The role of fat-storing cells incubation. The replacement with new medium con- 4 Minato in Disse space in fibrogenesis in alcoholic liver disease. HEPATOLOGY taining fresh acetaldehyde provoked an increase of 1983;3:559-566. collagen type I synthesis even during the second 24 hr of 5 French SW.Mivamoto K. Wone K. Jui L. Briere L. Role of the Ito cell in liver p k e n c h y m i fibroiis’in rats fed alcohol and a high incubation. The ethanol concentration in the medium fat-low protein diet. Am J Pathol 1988;132:73-85. corresponds to the blood level of baboons fed ethanol Mak KM, Lieber CS. Lipocytes and transitional cells in alcoholic chronically (39).The acetaldehyde concentrations main- 6 liver disease: a morphometric study. HEPAMLCGY 1988;8:1027tained in the cultures were either the same or higher 1033. than the hepatic venous blood levels observed in the 7 Mak KM, Leo MA, Lieber CS. Alcoholic liver injury in baboons: transformation of lipocytes to transitional cells. Gastroenterology baboons given alcohol (40); however, it is known that 1984;87:188-200. acetaldehyde concentrations in the liver are higher than 8. Horn T, Jumge J, Christoffersen P. Early alcoholic liver injury: those observed in the blood (41) and, therefore, concenactivation of lipocytes in acinar zone 3 and correlation to degree of trations used in the cell culture media can be expected to collagen formation in the Disse space. J Hepatol 1986;3:333-340. be close to the intracellular hepatic acetaldehyde levels. 9 Clement B, Grimaud JA, Campion JP, Deugnier Y,Guillouzo A. Cell types involved in collagen and fibronectin production in Collagens I and I11 are the most abundant types in the and fibrotic human liver. HEPATOLCGY 19866:225-234. liver; together they account for about 80% of total liver 10 normal De Leeuw AM, McCarthy SP, Geerts A, b o o k DL. Purified rat collagen (42,43).Collagen type I11 is the most abundant liver fat-storing cells in culture divide and contain collagen. collagen type in the liver lobule and it is in close 1984;4:392-403. HEPATOLCGY association with processes of Ito cells in the space of 11 Friedman SL, Roll FJ, Boyles J , Bissel DM. Hepatic lipocytes: the principal collagen-producing cells of normal rat liver. Proc Natl Disse (44). Furthermore, an increased accumulation of Acad Sci USA 1985;82:8681-8685. type I11 collagen in the perivenular zone and in perisi- 12 Kawase T, Shiratori Y,Sugimoto T. Collagen production by rat nusoidal areas has been described in the early stages of liver fat-storing cells in primary cultures. Exp Cell Biol 198634: 183-189. alcohol-induced liver fibrosis (45).Therefore, one might have expected this type of collagen to be enhanced by 13 Holt K, Bennett M, Chojkier M. Acetaldehyde stimulates collagen and noncollagen protein production by human fibroblasts. HEPAacetaldehyde in our in uitro model. On the other hand, TOLOCY 1984;4:843-848. it is well known that collagen type I11 predominates only 14 Thanassi NM, Rokowski RJ, Sheehy J , Hart B, Absher M, in the early stages of hepatic fibrosis, whereas type I Cutroneo KR. Non-selective decrease of collagen synthesis by cultured fetal lung fibroblasts after non-lethal doses of ethanol. becomes predominant as the deposition of collagen Biochem Pharmacol 1980;29:2417-2424. proceeds in the liver parenchyma; indeed, the type I/type 15 Savolainen ER, Leo MA, Timpl R, Lieber CS. Acetaldehyde and I11 ratio, which is 1 in normal liver, increases to above 2 lactate stimulate collagen synthesis of cultured baboon liver in the cirrhotic liver (46, 47). Moreover, an increased myofibroblasts. Gastroenterology 1984;87:777-787. expression of the specific mRNA for procollagen type I 16 Brenner DA, Chojkier M. Acetaldehyde increases collagen gene transcription in cultured human fibroblasts. J Biol Chem 1987; has been demonstrated in the baboon model of alcohol262:17690-17695. induced liver fibrosis, even in the absence of cirrhosis 17 Gressner AM, Althaus M. Effect of ethanol, acetaldehyde and ~~~~

(48).

In our in uitro model, hepatocytes produced collagen, but this synthesis was limited to type 111, was not affected by ethanol or acetaldehyde exposure and was much less extensive than that of FSCs. Our data are in agreement with other reports (10, 11, 21, 22) that stressed the role of nonparenchymal cells, particularly FSCs, as the principal collagen-producing cells and demonstrated that type I collagen is the most abundant type (11). In conclusion, the results obtained in this study suggest that acetaldehyde (a metabolite of ethanol), but not ethanol itself, plays an important role in the development of alcohol-induced liver fibrosis and that FSCs are the target of the acetaldehyde-mediated increased collagen synthesis.

~

~

lactate on proteoglycan synthesis and proliferation of cultured rat liver fat-storing cells. Gastroenterology 1988;94:797-807. 18 Chojkier M. Hepatocyte collagen production in vivo in normal rats. J Clin Invest 1986;78:333-339. 19 Chojkier M, Lyche K, Filip M. Increased production of collagen in vivo by hepatocytes and nonparenchymal cells with carbon 1988;8:808tetrachloride-induced hepatic fibrosis. HEPAMLOGY 814. 20 Chojkier M, Brenner DA, Leffert HL. Vasopressin inhibits type I

collagen and albumin gene expression in primary cultures of adult rat hepatocytes. J Biol Chem 1989;264:9583-9591. 21 Maher J J , Bissel DM, Friedman SL, Roll FJ. Collagen measured in primary cultures of normal rat hepatocytes derives from lipocytes within the monolayer. J Clin Invest 1988;82:450-459. 22 Milani S, Herbst H, Schuppan D, Hahn EG, Stein H. In situ hybridization for procollagen types I, 111 and N mRNA in normal and fibrotic rat liver: evidence for predominant expression in 1989;10:84-92. nonparenchymal liver cells. HEPATOUJCY 23 Hartman DJ,Ronziere MC, Grimaud JA, Herbage D, Ville GB.

518

MOSHAGE, CASINI AND LIEBER

Comparativestudy for collagen iodination techniques. In: Hofer R, Bergmann H, ed. Radioaktive Isotope in Klinik und Forschung. vol. 14. Munich: Egermann Publisher, 1980:533-540. 24. h o o k DL, Seffelaar AM, De Leeuw AM. Fat-storing cells of the rat liver. Exp Cell Res 1982;139:468-471. 25. Schafer S,Zerbe 0,Greesner AM. The synthesis of proteoglycans in fat-storing cells of rat liver. ~ A T C I L O G Y 1987;7:680-687. 26. Takaee S,Leo MA, Nouchi T, Lieber CS. Desmin distinguishes fat-storing cells from myofibroblasts, smooth muscle cells and fibroblasts in the rat. J Hepatol 1988,6267-276. 27. !3eglen PO. Preparation of isolated rat liver cells. In: Prescott DM, ed. Methods of cellular biology. New York Academic Press, 197629-83. 28. Reid LM, Narita M, Fujita M, Murray Z,Liverpool C, Rosenberg L. Matrix and hormonal regulation of differentiation in liver cultures. In: Guillowo A, Guguen-GuillouzoC, eds. Research in isolated and cultured hepatocytes. London: John Libbey & Co. La., 1986:225-258. 29. Rennard SI, Berg R, Martin GR, Foidart JM, Gehron-Robey P. Enzyme-linked immunoassay (ELISA) for connective tissue components. Anal Biochem 1980;104:205-214. 30. Moshage HJ, Janssen JAM, Franssen JH,Hafkenscheid JCM, Yap SH. Study of the molecular mechanism of decreased liver synthesis of albumin in inflammation. J Clin Invest 1987;79:16351641. 31. Chaves Das Neves HJ, Vasconcelos AMP. Capillary gas chromatography of amino acids, including asparaghe and glutamine: sensitive gas chromatographic-mass spectrometric and selected ion monitoring gas chromatographic-mass spectrometric detection of the N,O(S)-tert-butyldimethylsilylderivatives. J Chromatop 1987;392:249-258. 32. Takagi T, Alderman J, Gellert J, Lieber CS. Assessment of the role of non-ADH ethanol oxidation in vivo and in hepatocytes from deermice. Biochem Pharmacol 1986;35:3601-3606. 33. KapuscinekiJ, Skouylas B. Simple and rapid fluorimetric method for DNA microassay. Anal Biochem 1977;83:252-257. 34. Hu CL. Crombie G, Franzblau C. A new assay for collagenolytic -~ activi~.Anal Biocdem 1978;88:638-643.

HEPATOLOGY

35. Seller A, Cartwright E, Murphy G, Reynolds JJ. Evidence that latent collagenases are enzyme-inhibitor complexes. Biochem J 1977;163:303-307. 36. Snedecor GW, Cochran WG. In: Statistical methods, 7th ed. Ames, IA: Iowa State University Preas, 1980:215. 37. Shiratori Y, Ichida T, Kawase T, Wise E. Effect of acetaldehydeon collagen synthesis by fat-storing cells isolated from rats treated with carbontetrachloride. Liver 1986;6246-251. 38. Geerts A, Vriisen R, Rauterberg J, Burt A, Schellinck P, Wisse E. In vitro differentiation of fat-storing cells parallels marked increase of collagen synthesis and secretion. J Hepatol1989;95968. 39. Jauhonen P, Baraona E, Miyakawa H, Lieber CS. Mechanism for selective perivenular hepatotoxicity of ethanol. Alcoholism Clin Exp Res 1982;6:350-357. 40. Baraona E, DiPadova C, Tabasco J, Lieber CS. Transport of acetaldehyde in red blood cells.Alcohol Alcoholism 1987;(suppl. 1):203-206. 41. Baraona E, Matsuda Y, Pikkarainen P, Finkelman F, Lieber CS. Effects of ethanol on hepatic protein secretion and microtubules: possible mediation by acetaldehyde. In: Galanter M, ed. Currents in alcoholism. vol. 8. New York: Grune & Stratton, 1981:421-434. 42. Seyer JM, Hutcheson ET, Kana AH. Collagen polymorphism in normal and cirrhotic human liver. J Clin Invest 1977;59:241-247. 43. Rojkind M, GiambroneMA, Biempica L. Collagen typesin normal and cirrhotic liver. Gastroenterology 1979;76:710-719. 44. Oikawa K. A scanning electron microscopic study of the liver of mice. Tohoku J Exp Med 1979;129:373-387. 45. Sat0 S, Leo MA, Lieber CS. Ultrastructural localization of type I11 procollagen in baboon liver. Am J Path01 1986;122:212-217. 46. Rojkind M. Liver fibrosis. In: Gitnick G, ed. Current hepatology. Vol. 6. Chicago: Year Book Medical Publishers, 1986:lll-129. 47. Rojkind M. Liver cirrhosis. In: Gitnick G, ed. Current hepatology. Vol. 5. Chicago: Year Book Medical Publishers, 1985:465-491. 48. Zern MA, Leo MA, Giambrone MA, Lieber CS. Increased type I procollagen mRNA levels and in vitro protein synthesis in the baboon model of chronic alcoholic liver disease.Gastroenterology 1985:89: 1123-1131.