Phenobarbital Suppresses Growth and Accelerates Restoration of Differentiation Markers of Primary Culture Rat Hepatocytes in the Chemically Defined Hepatocyte Growth Medium Containing Hepatocyte Growth Factor and Epidermal Growth Factor

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

241, 445– 457 (1998)

EX984085

Phenobarbital Suppresses Growth and Accelerates Restoration of Differentiation Markers of Primary Culture Rat Hepatocytes in the Chemically Defined Hepatocyte Growth Medium Containing Hepatocyte Growth Factor and Epidermal Growth Factor Masahiro Miyazaki,*,† Wendy M. Mars,* Dieter Runge,* Tae-Hyoung Kim,* William C. Bowen,* and George K. Michalopoulos*,1 *Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261; and †Department of Cell Biology, Institute of Molecular and Cellular Biology, Okayama University Medical School, Okayama 700, Japan

Phenobarbital (PB), a liver-tumor promoter, at a concentration of 3 mM dramatically inhibited the growth of adult rat hepatocytes in the chemically defined medium, HGM, with added hepatocyte growth factor (HGF) and epidermal growth factor (EGF). In concurrence with these findings, PB down-regulated expression of the HGF receptor (c-met) and suppressed production of the autocrine growth factor transforming growth factor-a (TGF-a). Furthermore, PB down-regulated expression of transcription factors associated with proliferation such as AP1 and NF-kB. In the presence of PB, hepatocytes remained morphologically differentiated and restoration of the expression of mature hepatocyte markers, such as albumin and cytochrome P450s (1A, 2B1/2, and 2E1), was accelerated after an initial phase of growth. Additionally, PB strongly suppressed expression of the mRNA for a-fetoprotein, a protein primarily expressed by fetal liver, and the accelerative effect of PB on restoration of mature hepatocyte markers showed a correlation with the up-regulation of the hepatocyte-enriched transcription factors HNF3 and HNF4. When the effects of PB on various extracellular matrix proteins were examined, the data indicated that PB specifically suppressed laminin and fibronectin production by hepatocytes, suggesting an important role for these proteins in growing hepatocyte cultures. © 1998 Academic Press Key Words: phenobarbital; hepatocyte; HGF; c-met; phenotypic changes; extracellular matrix.

INTRODUCTION

The liver has a high potential to regenerate. After partial hepatectomy, DNA synthesis is initiated in the remaining liver cells, followed by division of the cells and subsequent regeneration of the liver. The factors 1 To whom correspondence and reprint requests should be addressed. Fax: (412) 648-9846. E-mail: [email protected].

related to liver regeneration have been a topic of intense scrutiny for several decades. Recently, several studies have shown that hepatocyte growth factor (HGF), epidermal growth factor (EGF), and transforming growth factor-a (TGF-a) are primary mitogens for hepatocytes and stimulate DNA synthesis of these cells in chemically defined medium [1]. Levels of HGF or an HGF-like factor increase remarkably in the plasma and livers of mice and rats exposed to carbon tetrachloride or subjected to partial hepatectomy, prior to liver regeneration [2–5]. An elevation of the mRNA levels of this factor in the liver, spleen, and lung has also been reported [6–10]. Furthermore, in regenerating liver after partial hepatectomy, the increase in HGF mRNA in mesenchymal cells occurs in parallel with a rise in TGF-a and acidic fibroblast growth factor production [11, 12]. In fact, injection of HGF, TGF-a, or EGF into rats induces DNA synthesis in the in vivo hepatocytes directly or, especially, after pretreatment of the animals with manipulators such as collagenase perfusion or nutritional alterations [13–17]. Thus, these growth factors seem to play an important role in compensatory liver regeneration. In spite of the high capacity of hepatocytes to proliferate in vivo, under most culture conditions cells placed in primary culture enter into only a limited number of divisions under the influence of primary mitogens before they degenerate and die [18, 19]. This phenomenon has been shown to be associated with changes in gene expression as well as levels of specific transcription factors [20]. The limited capacity of hepatocytes to proliferate in primary culture has hindered studying the conditions that contribute to the growth potential and phenotypic transition of hepatocytes in regenerating liver. Recently, we documented conditions that allow hepatocytes to enter into sustained clonal growth, resulting in long-term expansion of the cell population [21]. The new, chemically defined hepatocyte growth medium

445

0014-4827/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

446

MIYAZAKI ET AL.

(HGM) allows such growth in the presence of the hepatocyte primary mitogens HGF, TGF-a and/or EGF. In HGM with growth factors, hepatocytes undergo multiple proliferative cycles, express altered levels of hepatocyte-associated transcription factors and lose characteristic gene expression markers, such as albumin and cytochrome P450 2B1/2 (CYP2B1/2), while expressing aspects of the bile duct epithelial phenotype (cytokeratins 14, 19). In the presence of specific matrix components and growth factors, these simplified cells can be manipulated either to revert to a mature hepatocyte phenotype or to form acinar structures composed of cells with mixed ductular and hepatocytic appearance. In the present study, we investigated the effects of phenobarbital (PB) on the growth and phenotypic conversion of proliferating hepatocytes stimulated by HGF and EGF in HGM. HGM with added growth factors allows long-term, serum-free culture of rat hepatocytes. PB is known to be a potent promoter of hepatocarcinogenesis [22–24]. Previously, we reported that PB at 3 mM in Eagle’s minimal essential medium (MEM) efficiently supports survival of functional hepatocytes from normal adult rats for a long period of time, i.e., at least 1.5 months in the presence of 20% bovine serum in primary culture [25]. Without PB, these cells die off after about 1 week in culture. This action of PB on primary cultured hepatocytes is concentration dependent, with a maximal effect observed at 3 mM both in the presence [25, 26] and in the absence of serum [27]. Furthermore, maintenance of the PB effect on hepatocytes is observed only when the chemical is continuously present in the medium [25–27]. It should be noted that PB is not unique in its actions, as a number of other barbituric acid derivatives show similar activities on primary cultured hepatocytes [26, 28]. Here, we document that PB at 3 mM remarkably suppresses the growth of normal adult rat hepatocytes cultured with HGF and EGF in HGM, efficiently accelerating a differentiated phenotype in the cells after a period of initial cell growth. MATERIALS AND METHODS Materials. Male Fischer 344 rats from Charles River were used for all experiments. Collagenase for hepatocyte isolation was purchased from Boehringer Mannheim (Germany); vitrogen (type I collagen) for coating culture vessels was from Celtrix Lab. (Palo Alto, CA); EGF was from Collaborative Research (Walthan, MA); [3H]thymidine (50 Ci/mmol, 1 mCi/ml) was from ICN Radiochemicals (Irvine, CA); and other general reagents were from Sigma Chemical Co. (St. Louis, MO). HGF was the D5 variant, the mature, two-chain form, and was kindly donated by Snow Brand Co. (Tochigi, Japan). Isolation and culture of hepatocytes. Rat hepatocytes were isolated by an adaptation of the calcium two-step collagenase perfusion technique as previously described from our laboratory [29]. Hepatocytes, which were suspended in an attachment medium (see below), were seeded onto collagen-coated 80-mm dishes (for RNA and protein

extraction) or 12-well plates (for tritiated thymidine uptake) and left to attach for 2 h. Sixty thousand hepatocytes per square centimeter were inoculated for all the experiments described. The attachment medium was changed to HGM (see below), with growth factors (HGF and EGF), 2 h after seeding and was replenished thereafter every 48 h, unless otherwise specified. A minimum of two animal perfusions were used to set up each type (RNA, protein, nuclei, etc.) of study. Results were reproducible within each set as well as from set to set. Coating of culture vessels with collagen. Vitrogen (type I collagen) was diluted to a concentration of 10% with sterile, doubledistilled water and used for coating. Dry coating of culture vessels with collagen was done as previously described [30]. Composition of the attachment medium and HGM. The attachment medium consisted of MEM containing nonessential amino acids (Gibco/BRL 41600, Gaithersburg, MD), insulin at 50 mg/ml, bovine serum albumin (BSA) at 500 mg/ml, and gentamycin at 100 mg/ml. As we described previously [21], the basal HGM consisted of DMEM (Gibco/BRL 11966) supplemented with 2.0 g/liter BSA, 2.25 g/liter glucose, 2.0 g/liter galactose, 0.1 g/liter ornithine, 0.030 g/liter proline, 0.305 g/liter nicotinamide, 0.544 mg/liter ZnCl2, 0.750 mg/ liter ZnSO4 z 7H2O, 0.200 mg/liter CuSO4 z 5H2O, 0.025 mg/liter MnSO4, 146.1 mg/liter glutamine, 1.0 g/liter ITS mixture (Boehringer Mannheim) [5.0 mg/liter rh-insulin, 5.0 mg/liter human transferrin (30% diferric iron saturated), 5.0 mg/liter selenium], and 1027 M dexamethasone. Penicillin and streptomycin were added to the basal HGM at 100 U/ml and 100 mg/ml, respectively. The growth factors HGF and EGF were added to the basal HGM at 40 and 20 ng/ml, respectively. PB treatment in culture. A stock solution of PB was prepared at a concentration of 0.3 M by dissolving the sodium salt in MEM, storing at 4°C, and using within the week. The PB stock solution was added to the primary hepatocyte cultures to obtain a final concentration of 3 mM when the medium was changed. Unless otherwise specified, treatment of hepatocytes with PB was started 2 h after the cells were plated, when the medium was switched to HGM. Measurement of [3H]thymidine uptake. For the measurement of DNA synthesis, primary hepatocyte cultures on 12-well plates were treated with 1 mCi per well of [3H]thymidine for 24 h. Each time point for each set of experiments was determined in triplicate. The incorporation of [3H]thymidine into DNA was measured as described previously [21]. The measurement of DNA content of cells was done by the fluorometric method of Brunk et al. [31] with calf thymus DNA as a standard. For these studies, [3H]thymidine uptake is expressed as radioactivity per microgram of DNA in culture. Feeding rats with PB and preparation of parenchymal and nonparenchymal liver cell fractions. Rats were fed with PB, which was dissolved at a concentration of 0.05% in drinking water. At the indicated times, the rats were subjected to the calcium two step collagenase liver perfusion as described above [29]. Liver cells suspended in Hanks’ balanced salt solution were centrifuged at 50g for 1 min. Supernatant was kept on ice for preparation of the nonparenchymal cell fraction. Cell pellet was resuspended in Hanks’ solution and centrifuged under the same conditions. The cell pellet thus obtained was designated as the parenchymal cell fraction. The two supernatants were combined and centrifuged at 50g for 1 min to remove parenchymal cells. Then the final supernatant was centrifuged at 150g for 3 min to obtain the nonparenchymal cells. The parenchymal and nonparenchymal cell fractions were immediately frozen in liquid nitrogen and kept at 280°C until used for extraction of total RNA. Analysis of gene expression by Northern blots. Total RNA was extracted from phosphate-buffered saline (PBS)-washed hepatocyte cultures, parenchymal and nonparenchymal liver cell fractions from PB-fed rats using RNAzol B (Tel-Test Laboratories, Houston, TX) and purified per the manufacturer’s guidelines. RNA concentration

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES and purity were determined by routine spectrophotometry. Samples of RNA (20 mg per lane) were size fractionated on 1% agarose gels and transferred to nylon membranes (Amersham, Arlington Heights, IL) by the capillary method. After cross-linking under UV light, membranes were prehybridized and then hybridized overnight with specific cDNAs (as indicated in the figures) that had been labeled with [a-32P]dCTP using an Amersham random primer kit. Membranes were subsequently washed under high stringency conditions and exposed to X-omat film (Kodak, NY). cDNA probes. The plasmids containing cDNA probes used to study gene expression were obtained from the following sources: plasmids pB1RSA59 (rat albumin) and pB1RAF39 (rat a-fetoprotein, AFP) [32] from Dr. Joseph Locker (University of Pittsburgh, Pittsburgh, PA) and plasmids HL-40 (human laminin B1, LAMB1), HL210 (human laminin B2, LAMC1), pFH154 (human fibronectin), HSP-pUC18 (human vitronectin), Hf677 (human type I collagen-a 1, COL1A1), Hf32 (human type I collagen-a2, COL1A2), Hf934 (human type III collagen-a1, COL3A1), pCIV-1-PE16 (mouse type IV procollagen-a1, COL4A1), HHCH75 (human ribosomal RNA, 28S), and HFBCC79 (human cytochrome P450 2E1, CYP2E1) from The American Type Culture Collection (Rockville, MD). Other cDNAs were obtained from the following sources: TGF-a (rat) from Dr. David Lee (University of North Carolina, Chapel Hill, NC), transforming growth factor-b1 (TGF-b1, human) [33], and cytochrome P450 2B1/2 (CYP2B1/2) from Dr. Steve Strom (University of Pittsburgh, Pittsburgh, PA). When necessary, the inserted cDNAs were isolated from the plasmids by restriction enzyme digestion followed by agarose gel electrophoresis and band excision from the gel with subsequent purification using the GeneClean II kit (BIO 101 Inc., La Jolla, CA) per the manufacturer’s guidelines. Hybridizations were performed using the following cDNA inserts: 0.31-kb HindIII fragment of rat albumin; 0.94-kb SstI–SalI fragment of rat AFP; 3.7-kb EcoRI fragment of human LAMB1; 3.6-kb EcoRI fragment of human LAMC1; 0.7-kb PvuII fragment of human fibronectin; 1.54-kb EcoRI fragment of human vitronectin; 1.5-kb EcoRI fragment of human COL1A1; 1.1-kb HindIII–PstI fragment of human COL1A2; 0.7- and 0.6-kb EcoRI–HindIII fragments of human COL3A1; 1.6-kb BamHI fragment of mouse COL4A1; 2.4-kb EcoRI fragment of human CYP2E1; 1.05-kb EcoRI fragment of human TGF-b1; and 0.7-kb EcoRI fragment of human ribosomal RNA 28S. Western blot analysis. Cells in culture were placed on ice and washed twice in PBS. Monolayers were lysed by addition of 1 ml per 80-mm dish of a lysis buffer that consisted of 10 mM Tris–HCl (pH 8.0), 1.0% sodium dodecyl sulfate (SDS), and protease inhibitors [10 mg/ml leupeptin, 100 mM phenylmethylsulfonyl fluoride (PMSF), 10 mg/ml pepstatin, 5 mM Na2EDTA, 10 mg/ml Glu-Gly-Arg-chloromethylketone, 5 mg/ml trans-epoxysuccinyl-L-leucylamido-(4-guanidino)butane (E-64), 10 mg/ml aprotinin, 50 mM phenanthroline]. The cell lysates were stored at 280°C until use. Protein content was assayed using a bicinchoninic acid assay [34]. Equivalent amounts of protein from each sample were placed in buffer, as described by Laemmli [35], with b-mercaptoethanol and heated to 95°C for 5 min before loading onto 10% polyacrylamide gels prepared in a vertical slab gel unit (SE280; Hoefer, San Francisco, CA). After electrophoresis, proteins were electrophoretically transferred (TE22; Hoefer) to Immobilon-P membrane (Millipore, Bedford, MA). Transfer was performed overnight at 250 mA in a buffer containing 50 mM Tris–HCl, 95 mM glycine, and 0.005% SDS. Membranes were reversibly stained with 0.2% Ponceau S to assess transfer efficiency before blocking with 5% powdered milk in blotto base [20 mM Tris–HCl (pH 7.5), 150 mM NaCl, 0.1% Tween 20]. Primary antibodies were applied at a concentration of 2 mg/ml in blocking buffer for a minimum of 2 h at room temperature. The membranes were then washed well with 1% powdered milk in blotto base, and secondary antibody, conjugated to horseradish peroxidase, was applied at a concentration of 0.4 mg/ml in blocking buffer. After 1 h,

447

filters were washed extensively with blotto base and the bands were visualized using the ECL detection system (Amersham). Antibodies. The primary antibodies used for Western blot analysis were as follows: goat anti-rat CYP1A and goat anti-rat CYP2E1 (Gentest Corp., Woburn, MA); goat anti-rat fibronectin (Calbiochem 341653, La Jolla, CA); goat anti-rat laminin (Gibco/BRL 12116-018), and rabbit anti-mouse MET (Santa Cruz sc162). The secondary antibodies conjugated to horseradish peroxidase used for this analysis were as follows: goat anti-rabbit IgG (Sigma A6154) and rabbit anti-goat IgG (Sigma A4174). Nuclear extract preparation. Cells were washed twice in 40 mM Tris/HCl (pH 7.6), 14 mM NaCl, 1 mM Na2EDTA, and harvested in the same buffer (2 ml/plate). The cells of three to five plates were pooled and centrifuged at 80g. Next the cells were resuspended in 2 ml of hypotonic buffer [10 mM Hepes (pH 7.9), 10 mM NaH2PO4, 1.5 mM MgCl2, 1 mM dithiothreitol, 0.5 mM spermidine, 1 mM NaF, 1% nonfat dry milk, 1 mM PMSF, and 1 mg/ml each leupeptin, aprotinin, antipain, E-64, and pepstatin A], incubated at 4°C for 10 min and homogenized in a Dounce homogenizer (60 –100 strokes). Homogenates were examined under the light microscope to control breakage of the cells. The released nuclei were pelleted by centrifugation (5 min, 800g). Nuclear proteins were extracted in 50 –100 ml of hypertonic buffer [30 mM Hepes (pH 7.9), 25% glycerol, 450 mM NaCl, 12 mM MgCl2, 6 mM dithiothreitol, 0.3 mM Na2EDTA, 1 mM PMSF, and 1 mg/ml each leupeptin, aprotinin, antipain, E-64, and pepstatin A] for 45 min at 4°C. After centrifugation at 30,000g for 30 min the supernatant was recovered and dialyzed for 2 h against the same solution, but containing 150 mM NaCl. Protein concentration was determined according to the method of Bradford [36]. Bandshift assays. Single-stranded oligonucleotides, described in our previous study [21], were annealed by incubation at 80°C for 10 min, followed by slow cooling to room temperature. Labeling was performed by incubating 1 mg of the double-stranded oligonucleotides with 2 U Klenow DNA polymerase I (Boehringer Mannheim), 50 mCi of [32P]dCTP, 4 mM dATP, 4 mM dTTP, 4 mM dGTP in 13 reaction buffer of the Multiprime DNA labeling systems kit (Amersham) at room temperature overnight. The reaction was stopped by adding blue juice [0.03% (w/v) bromophenol blue, 5% (w/v) glycerol in 1 mM Tris–HCl (pH 7.6), 0.2 mM Na2EDTA]. The unincorporated nucleotides were removed by spinning through an S200 microspin column (Pharmacia). Specific activities were about 1.5 3 107 cpm/mg. The binding reaction (final volume 20 ml) contained 2.5 (NF1, HNF3, HNF4) or 5 mg (AP1, NF-kB) nuclear extract protein in 15 mM Hepes (pH 7.9), 75 mM NaCl, 6 mM MgCl2, 1.15 mM Na2EDTA, 5 mM Tris–HCl (pH 7.6), 12.5% glycerol, 3 mM dithiothreitol, 0.5 mM PMSF, and 0.5 mg/ml each E-64, pepstatin A, leupeptin, antipain, and aprotinin. After a 15-min incubation at 4°C, the labeled probe (40,000 cpm/assay) was added and the incubation was continued for 20 min at room temperature. The mixture was electrophoresed through a 6% polyacrylamide gel in TBE running buffer (90 mM Tris– borate, 2 mM Na2EDTA) for 2.5 h. Gels were dried and exposed to X-omat film (Kodak, Rochester, NY).

RESULTS

Suppressive Effect of PB on Growth of Hepatocytes Stimulated by HGF and EGF in the HGM In our previous study [21], a high BRdU nuclear labeling index (about 80%) was observed using similar culture conditions as described in the present study. Furthermore, the results documented that the proliferating cells were not a minor contaminant in the population but, rather, they derived directly from the mature parenchymal hepatocytes. In the present

448

MIYAZAKI ET AL.

Suppressive Effect of PB on Expression of c-met in Hepatocytes To elucidate the mechanisms by which PB suppresses hepatocyte growth, expression of the HGF receptor was examined by Northern and Western blot analyses in primary cultures, with or without PB. The expression of the proto-oncogene c-met, the HGF receptor gene, steadily increased by day 9 in the absence of PB, whereas it was suppressed in the presence of PB (Fig. 2A). The MET protein was also present at a much lower concentration in the PB-treated cultures than in the untreated cultures (Fig. 2B). Furthermore, when PB was added to cultures after 6 days without treatment, the presence of MET protein was strongly decreased (Fig. 2B).

FIG. 1. (A) Incorporation of [3H]thymidine into DNA in primary hepatocyte cultures at different times after seeding. Cells were cultured in HGM containing HGF and EGF in the presence or absence of PB. Thymidine values are expressed per microgram of DNA. (B) Amount of DNA per plate at different times in culture in the presence or absence of PB as indicated. Each point represents the mean of three wells. Vertical bars indicate SD.

study, cells were plated at a much higher density than previously. DNA synthesis, expressed as tritiated thymidine uptake per microgram of DNA in culture, sharply increased from day 2, peaked around day 4 (showing a 22-fold increase compared with day 1), and was rapidly slowed by day 6 (Fig. 1A). DNA content in the hepatocyte cultures showed a gradual increase from day 4 which reached a plateau by day 10, with a 5-fold increase when compared with day 1 (Fig. 1B). When 3 mM PB was added, DNA synthesis increased from day 2 with a peak at day 3 (Fig. 1A). The peak, however, was about 60% of the peak observed in the absence of PB which occurred on day 4. DNA content in the PB-treated cultures also increased slightly, beginning at day 4 and reaching a plateau at day 10 (Fig. 1B). There was, however, a remarkable difference in the DNA content between the PB-treated and -untreated hepatocyte cultures; at day 10, DNA in the PB-treated cultures was less than 50% of that in the untreated cultures. Thus, PB remarkably suppressed DNA synthesis and proliferation of hepatocytes in HGM with added growth factors.

FIG. 2. Northern and Western blot analyses of expression of c-met (HGF receptor) in primary hepatocyte cultures in HGM containing HGF and EGF at different times after seeding. (A) RNA samples were derived from PB-treated cultures (lanes 3, 5, 7, 9, 11, and 13) or untreated cultures (lanes 1, 2, 4, 6, 8, 10, and 12). The intensity of 28S RNA was used as an internal control to compare the amount of loaded RNA between lanes. The numbers on the left ordinate refer to the size (kilobases) of each specific message. The exposure times of autoradiography were as follows: c-met, 24 h; 28S, 30 min. (B) Protein samples were derived from cultures treated with PB (lanes 3, 5, 7, and 10, continuously after seeding; lanes 8 and 11, from day 6 onward) or untreated cultures (lanes 1, 2, 4, 6, and 9). The number on the left ordinate refers to the size (kilodaltons) of MET protein.

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES

449

FIG. 3. Phase-contrast micrographs of primary hepatocyte cultures in HGM containing HGF and EGF in the absence (A, B, and C) or presence of PB (D, E, and F). A and D, day 1; B and E, day 2; C and F, day 5. Bar, 100 mm.

Effect of PB on Phenotypic Changes of Hepatocytes during Proliferation In the absence of PB, the morphology of the hepatocytes varied over the days following culture in HGM. From a normal hepatocyte morphology at day 1 (Fig. 3A), the hepatocytes acquired long projections assum-

ing the typical phenotype described previously as due to the ‘‘scattering’’ effect of HGF on the cells (Fig. 3B) [37]. With time, the proliferating hepatocytes lost most of their cytoplasmic granules, the projections diminished, and the cells began to grow as monolayer patches (Fig. 3C). These patches merged as the cells

450

MIYAZAKI ET AL.

FIG. 4. Phase-contrast micrographs of primary hepatocyte cultures in HGM containing HGF and EGF in the absence (A and B) or presence of PB (C and D, continuously after seeding; E and F, from day 6). A, C, and E: day 9; B, D, and F: day 13. Bar, 100 mm.

continued to grow, forming a continuous monolayer (Figs. 4A and 4B). In confluent cultures (around day 9) there was focal overgrowth of fibroblast-like cells. (In less dense primary hepatocyte cultures cultured in HGM, confluence occurs around day 15 and contaminating fibroblast-like cells are seen at nearly the same time point [21]). Although the nature of these cells is

not clear, it has been shown that cells similar to these are derived mostly from hepatic Ito cells [38]. Interestingly, in these areas, the cytoplasm of the hepatocytes appeared more granular. In the presence of PB, hepatocytes still acquired long projections when the cells were cultured in HGM (Figs. 3D and 3E). The cells, however, barely lost their cyto-

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES

451

plasmic granules and retained most of their morphological characteristics, during and after the period of cell proliferation (Figs. 3F, 4C, and 4D). Although the cells diminished their projections and grew as monolayer patches, fibroblast-like cells rarely grew in the presence of PB (Figs. 4C and 4D). Furthermore, when PB was added to cultures after 6 days without treatment, it efficiently induced morphological ‘‘redifferentiation’’ of hepatocytes and severely inhibited fibroblast-like cells from growing (Figs. 4E and 4F). Proliferating Hepatocytes Lose Hepatic Differentiation Markers, but PB Accelerates Restoration of Those Markers We examined the effects of PB on expression of CYP2B1/2 mRNA which is known to be induced in the liver by this chemical [39]. The expression of CYP2B1/2 mRNA markedly decreased with hepatocyte proliferation in HGM independently of the presence or absence of PB; however, although the message reappeared much earlier in the PB-treated cultures than in the untreated cultures (Fig. 5A), it was not constitutively present. In fact, the results suggested that CYP2B1/2 was behaving more like a hepatocyte differentiation marker rather than a PB marker. To confirm this, we next examined expression of other CYPs, such as CYP2E1 (ethanol inducible) and CYP1A (aromatic hydrocarbon inducible) in the presence or absence of PB. The expression of CYP2E1 mRNA was decreased throughout the culturing period in the absence of PB (Fig. 5A). In the presence of PB, the message was also diminished but reappeared again after day 5 in culture (Fig. 5A). Western blot analysis of CYP2E1 protein confirmed these findings as the protein was greater in the PB-treated cultures than in the untreated cultures (Fig. 5B). In addition, there was a striking increase in CYP1A protein in the presence of PB (Fig. 5B). To further confirm the induction of a more differentiated phenotype, cells were examined for expression of albumin. Freshly isolated hepatocytes from adult rats were abundant in albumin mRNA; however, in the absence of PB the message dramatically decreased as the hepatocytes proliferated (Fig. 5C). Albumin mRNA became enhanced again by day 9 when proliferation of

FIG. 5. Northern and Western blot analyses of liver proteins in primary hepatocyte cultures in HGM containing HGF and EGF in the presence or absence of PB. (A) RNA samples were derived from PB-treated cultures (lanes 3, 5, 7, 9, 11, and 13) or untreated cultures (lanes 1, 2, 4, 6, 8, 10, and 12). The intensity of 28S RNA was used as an internal control to compare the amount of loaded RNA between lanes. The numbers on the left ordinate refer to the size (kilobases)

of each specific message. The exposure times of autoradiography were as follows: CYP2B1/2, 5 days; CYP2E1, 5 days; 28S, 30 min. (B) Protein samples were derived from cultures treated with PB (lanes 2 and 4, continuously after seeding; lane 5, from day 6) or untreated cultures (lanes 1 and 3). The numbers on the left ordinate refer to the size (kilodaltons) of each specific protein. (C) RNA samples were derived from cultures treated with PB (lanes 3, 6, and 9, continuously after seeding; lanes 5 and 8, from day 6) or untreated cultures (lanes 1, 2, 4, and 7). The exposure times of autoradiography were as follows: albumin, 4 h; AFP, 2 days; 28S, 1 h.

452

MIYAZAKI ET AL.

the hepatocytes had almost ceased and was maintained at an elevated level. In the presence of PB, albumin mRNA was also initially diminished in cultures; however, the message reappeared much earlier in the PB-treated cultures than in the untreated cultures (Fig. 5C). AFP is a serum a-globulin which is primarily synthesized by fetal liver [40]. Still, the AFP message can be detected at low levels in adult hepatocyte cultures [41]. In the present study, we detected a signal for the expression of AFP mRNA by day 9 in cultures lacking PB which diminished with time in culture (Fig. 5C). In cultures with added PB, the expression of AFP mRNA was suppressed by the presence of PB, even when the PB was added to the cultures 6 days after initiation of cell growth. Thus, in addition to up-regulating markers associated with a mature phenotype, PB inhibits the expression of a marker that is primarily expressed by fetal hepatocytes. PB Suppresses Expression of Cell Proliferation Associated Transcription Factors (AP1, NF-kB) but Enhances Hepatocyte-Enriched Transcription Factors (HNF3, HNF4) in Primary Hepatocyte Cultures To confirm the phenotypic and growth-related changes in hepatocytes stimulated by HGF and EGF in the presence or absence of PB, transcription factors were analyzed in nuclear extracts prepared from cultures at different time points. Bandshift assays were performed to assess hepatocyte-enriched transcription factors (HNF3 and HNF4), general factors associated with cell proliferation (AP1 and NF-kB) and a family of housekeeping factors (NF1). Scanning densitometry using NF1 as an internal control, showed that levels of AP1 and NF-kB were generally lower in the presence of PB (Fig. 6A), reflecting the suppression of proliferation by this chemical. The exceptions were day 3 where AP1 was slightly higher and day 4 where NF-kB was higher. On the other hand, PB-treated cultures generally expressed more hepatocyte-enriched transcription factors such as HNF3 and HNF4, when compared with untreated cultures (Fig. 6B). PB Severely Suppresses Expression of Extracellular Matrix Proteins in Primary Hepatocyte Cultures Extracellular matrix is thought to be not only a physical scaffold, but also a modulator of biologic processes including cell attachment, migration, differentiation, repair, and development [42]. As PB affected the hepatocyte phenotype in HGM, we examined its effects on the expression of extracellular matrix proteins. LAMC1 and fibronectin mRNAs dramatically increased until day 9 when proliferation of hepatocytes had almost ceased and then gradually decreased (Fig.

FIG. 6. Bandshift assay of transcription factor expression in primary hepatocyte cultures in HGM medium containing HGF and EGF in the presence or absence of PB. Nuclear proteins were extracted from PB-treated (1) or untreated cultures (2) at the indicated times: (A), days 1, 2, 3, and 4; (B), days 3 and 9.

7A). The protein quantities of fibronectin and a laminin chain the size of LAMC1 also increased throughout the culture period of 13 days, as shown by Western blot analyses (Fig. 7B). On the other hand, in the presence of PB the levels of LAMC1 and fibronectin mRNAs also increased until day 9 but, at a considerably lower level than in untreated cultures (Fig. 7A). Furthermore, the protein levels of the laminin chain and fibronectin were also much less in the presence of PB (Fig. 7B). These suppressive effects of PB were also clearly observed when the chemical was added to untreated cultures at day 6 (Figs. 7B and 7C); however, as expected, the suppression was not as intense due to the relatively long half-life of fibronectin protein [43, 44].

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES

453

FIG. 7. Northern and Western blot analyses of extracellular matrix protein expression in primary hepatocyte cultures in HGM containing HGF and EGF in the presence or absence of PB. (A) RNA samples were derived from PB-treated cultures (lanes 3, 5, 7, 9, 11, and 13) or untreated cultures (lanes 1, 2, 4, 6, 8, 10, and 12). The intensity of 28S RNA was used as an internal control to compare the amount of loaded RNA between lanes. The numbers on the left ordinate refer to the size (kilobases) of each specific message. The exposure times of autoradiography were as follows: LAMC1, 18 h; fibronectin, 18 h; vitronectin, 18 h; COL1A1, 18 h; COL1A2, 18 h; COL3A1, 18 h; COL4A1, 18 h; 28S, 30 min. (B) Protein samples were derived from cultures treated with PB (lanes 3, 5, 7, and 10, continuously after seeding; lanes 8 and 11, from day 6) or untreated cultures (lanes 1, 2, 4, 6, and 9). The numbers on the left ordinate refer to the size (kilodaltons) of each specific protein. (C) RNA samples were derived from cultures treated with PB (lanes 3, 6, and 9, continuously after seeding; lanes 5 and 8, from day 6) or untreated cultures (lanes 1, 2, 4, and 7). The exposure times of autoradiography were as follows: fibronectin, 18 h; COL1A1, 18 h; COL4A1, 18 h; 28S, 1 h.

Significant amounts of mRNAs for COL1A1, COL1A2, and COL3A1 could not be detected for the first 7 days in the primary hepatocyte cultures; however, the transcripts of these genes began to appear around day 9 and remarkably increased thereafter until day 13 (Fig. 7A). COL4A1 mRNA appeared around day 7, increased until day 11, and then decreased rapidly (Fig. 7A). None of these collagen mRNAs was detectable in the cultures in the presence of PB (Fig. 7A). Furthermore, these suppressive effects of PB on collagen gene expression were also clearly observed when the chemical was added to untreated cultures at day 6 (Fig. 7C).

Vitronectin mRNA was also abundant in freshly isolated hepatocytes; however, the expression level of this mRNA decreased following normal stimulation of hepatocytes and was dramatically suppressed by day 13 (Fig. 7A). In the presence of PB, the expression level of vitronectin mRNA decreased over time in culture until day 9 and thereafter increased to some extent. Thus, in the presence of PB, hepatocytes maintained expression of vitronectin mRNA after prolonged growth in primary culture. To determine if our in vitro results with the expression of the extracellular matrix protein genes resulted from the PB suppression of hepatocyte culture-derived

454

MIYAZAKI ET AL.

mal cells at days 4 and 7. Thus, PB transiently affected the expression of all these extracellular matrix components. Expression of the mRNA for TGF-b1, a potent mitoinhibitor of hepatocytes, steadily increased in the nonparenchymal liver cell fractions in rats fed with PB, which is consistent with results reported by Jirtle et al. [45]. This growth inhibitor has been reported to upregulate the gene expression of extracellular matrix components [46] and may account for the increases which occurred over time with these genes. DISCUSSION

FIG. 8. Northern blot analysis of expression of extracellular matrix protein genes and TGF-b gene in the liver of rats fed with PB. Rats were fed with 0.05% PB in drinking water. At the indicated times, total RNA was extracted from the parenchymal (PC) and nonparenchymal cell fractions (NPC) prepared from the PB-fed rat liver. The intensity of 28S RNA was used as an internal control to compare the amount of loaded RNA between lanes. The numbers on the left ordinate refer to the size (kilobases) of each specific message. The exposure times of autoradiography were as follows: fibronectin, 17 h; vitronectin, 17 h; COL3A1, 21 h; COL4A1, 17 h; TGF-b1, 4 days; 28S, 1.5 h.

subpopulations such as the fibroblast-like cells that appear late in the cultures, we fed rats with 0.05% PB in drinking water, prepared parenchymal and nonparenchymal liver cell fractions, and then extracted total RNAs from them for Northern blot analysis (Fig. 8). Fibronectin mRNA was present only in the parenchymal liver cell fraction when feeding of rats with 0.05% PB began. By 1 day after feeding, the fibronectin mRNA dramatically decreased and then increased again by day 4. At day 11, message transiently appeared in the nonparenchymal cells as well. COL3A1 and COL4A1 mRNAs were abundant in the nonparenchymal liver cell fractions at day 0. Expression of both the mRNAs decreased at day 1. COL3A1 mRNA increased again gradually from around day 4 while the COL4A1 mRNA increased sharply around day 4 and maintained the higher level until day 11. Vitronectin mRNA was detected both in the parenchymal and in the nonparenchymal liver cell fractions at day 0 but the amount was much more abundant in the former than in the latter. The vitronectin message steadily increased in the parenchymal liver cell fractions until day 11 and was slightly decreased in the nonparenchy-

In the present study, we once again demonstrated that hepatocytes from adult rats could be induced to enter into sustained clonal growth in the chemically defined HGM with added HGF and EGF, even when plated at high density. We also documented that the addition of PB at 3 mM dramatically suppressed the growth of hepatocytes and accelerated restoration of the expression of mature hepatocyte markers. Furthermore, we showed that PB affected the expression of various extracellular matrix proteins in this system. Subsequent to growth stimulation, there were steady increases in the quantities of mRNAs and proteins for laminin and fibronectin until the growth phase ceased. When hepatocyte growth was suppressed by adding PB, the expression of laminin and fibronectin was also remarkably suppressed, indicating a close relationship between these matrix proteins and hepatocyte growth. PB is known as a potent promoter for hepatocarcinogenesis in vivo [22–24] although the reasons are unclear. Usually, it is given to the hepatocarcinogeninitiated rats at a concentration of 0.05% (about 2 mM) in drinking water. Therefore, the actual PB concentration to which the initiated hepatocytes are exposed in the liver is likely to be lower than 2 mM. Administration of 0.05% PB causes transient stimulation of DNA synthesis in normal rat liver [47]. Similarly, growth of primary hepatocytes in vitro is effectively enhanced in the presence of mitogen(s) when PB is added at concentrations less than 2 mM [48 –50]. PB is also known to be an inducer of CYP2B1/2 in vivo [39]. The present study shows that PB can also induce CYP2B1/2 in primary culture hepatocytes, however, there may be different mechanisms at work since we also found an unexpected increase of CYPs 1A and 2E1. To our knowledge, this is the first report of CYPs 1A and 2E1 induction using PB and probably reflects the differentiation state of the cells, rather than a direct effect of PB. Finally, no in vivo studies have previously been reported for the effects of PB on liver matrices. The two most notable changes detected by our in vivo studies were seen with fibronectin (in parenchymal cells) and collagen (in nonparenchymal cells). These messages were transiently decreased at day 1, with a return to

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES

normal or higher level by day 4. Fibronectin expression was also seen in nonparenchymal cells at day 11 although it is possible this is an artifact caused by contamination with parenchymal cells. The significance of these findings is unclear but suggests that matrix effects are important for tumor promotion. The mechanisms by which PB enhances or inhibits hepatocyte growth are not yet clear. PB up-regulates or maintains the number of EGF receptors when added at low concentrations to EGF-stimulated hepatocytes in primary culture [49]. On the contrary, high-dose PB remarkably reduced EGF binding to primary hepatocytes from adult rats at concentrations greater than 3 mM [51–53]. In the present study, PB at 3 mM strongly suppressed the expression of the mRNA for c-met (the HGF receptor gene) in primary cultures of adult rat hepatocytes stimulated with HGF and EGF. The quantity of the HGF receptor protein was also much less in the PB-treated hepatocyte cultures than in the untreated ones. Previously, we documented that there was a steady increase in mRNA for TGF-a, one of the primary mitogens for hepatocytes, when the cells were stimulated by HGF and EGF in the HGM in primary culture [21]. TGF-a utilizes the same receptor as EGF and is considered to be an autocrine growth factor. In the present study, mRNA for TGF-a also increased until day 11. Interestingly, the expression of TGF-a was remarkably suppressed in the presence of PB (data not shown). Combined, these findings suggest that the mitoinhibitory effect of PB on hepatocytes at concentrations greater than 3 mM is at least partly due to down-regulation of the growth factor receptors and suppression of the production of autocrine growth factor(s). Bandshift assays correlate this data as PB generally down-regulated the expression of the proliferation-related transcription factors AP1 and NF-kB. In the presence of 3 mM PB, hepatocytes barely lost their cytoplasmic granules and retained most of their morphological characteristics, even during the cell proliferation phase. This may be due to prevention by PB of hepatocytes from morphologic degeneration by stabilizing their cytoplasmic membranes, as indicated in our previous studies [54]. Besides the morphological maintenance effects, PB apparently accelerated restoration of the expression of mature hepatocyte markers, such as albumin and CYPs 1A, 2B1/2, and 2E1, in proliferating hepatocytes. In concurrence, bandshift assays showed that PB efficiently enhanced the expression of the hepatocyte-enriched transcription factors HNF3 and HNF4. Furthermore, the expression of mRNA for AFP, a typical fetal hepatocyte marker, was suppressed in the presence of PB. Although the mechanisms by which PB exerts these effects on hepatocytes still remains to be resolved, the system described here is useful for studying hepatocyte growth and differentiation.

455

We acknowledge the excellent technical assistance of Ms. Linda Shab and the personnel of the Pathology Photography Laboratory. We also appreciate Ms. Valerie Tyner for her excellent technical assistance. This work was supported by National Cancer Institute and National Institute of Health Grants CA43632, CA35373, and CA31241.

REFERENCES 1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Michalopoulos, G. K. (1990). Liver regeneration: Molecular mechanisms of growth control. FASEB J. 4, 176 –187. Gohda, E., Hayashi, Y., Kawaida, A., Tsubouchi, H., and Yamamoto, I. (1990). Hepatotrophic growth factor in blood of mice treated with carbon tetrachloride. Life Sci. 46, 1801–1808. Asami, O., Ihara, I., Shimidzu, N., Shimizu, S., Tomita, Y., Ichihara, A., and Nakamura, T. (1991). Purification and characterization of hepatocyte growth factor from injured liver of carbon tetrachloride-treated rats. J. Biochem. 109, 8 –13. Lindroos, P. M., Zarnegar, R., and Michalopoulos, G. K. (1991). Hepatocyte growth factor (hepatopoietin A) rapidly increases in plasma before DNA synthesis and liver regeneration stimulated by partial hepatectomy and carbon tetrachloride administration. Hepatology 13, 743–749. Gohda, E., Matsunaga, T., Kataoka, H., Kasada, C., and Yamamoto, I. (1992). Purification and characterization of a mouse hepatocyte growth-stimulating factor from the liver of carbon tetrachloride-treated mice. J. Pharmacobio-Dyn. 15, 131–137. Nakamura, T., Nishizawa, T., Hagiya, M., Seki, T., Shimonishi, M., Sugimura, A., Tashiro, K., and Shimizu, S. (1989). Molecular cloning and expression of human hepatocyte growth factor. Nature 342, 440 – 443. Okajima, A., Miyazawa, K., and Kitamura, N. (1990). Primary structure of rat hepatocyte growth factor and induction of its mRNA during liver regeneration following hepatic injury. Eur. J. Biochem. 193, 375–381. Selden, C., Jones, M., Wade, D., and Hodgson, H. (1990). Hepatotropin mRNA expression in human foetal liver development and in liver regeneration. FEBS Lett. 270, 81– 84. Zarnegar, R., DeFrances, M. C., Kost, D. P., Lindroos, P., and Michalopoulos, G. K. (1991). Expression of hepatocyte growth factor mRNA in regenerating rat liver after partial hepatectomy. Biochem. Biophys. Res. Commun. 177, 559 –565. Yanagita, K., Nagaike, M., Ishibashi, H., Niho, Y., Matsumoto, K., and Nakamura, T. (1992). Lung may have an endocrine function producing hepatocyte growth factor in response to injury of distal organs. Biochem. Biophys. Res. Commun. 182, 802– 809. Kan, M., Huang, J. S., Mansson, P. E., Yasumitsu, H., Carr, B., and McKeehan, W. L. (1989). Heparin-binding growth factor type I (acidic fibroblast growth factor): A potential biphasic autocrine and paracrine regulator of hepatocyte regeneration. Proc. Natl. Acad. Sci. USA 86, 7432–7436. Mead, J. E., and Fausto, N. (1989). Transforming growth factor TGFa may be a physiological regulator of liver regeneration by means of an autocrine mechanism. Proc. Natl. Acad. Sci. USA 86, 1558 –1562. Bucher, N. L., Patel, U., and Cohen, S. (1977). Hormonal factors concerned with liver regeneration. Ciba Found Symp. 55, 95– 107. Ishiki, Y., Ohnishi, H., Muto, Y., Matsumoto, K., and Nakamura, T. (1992). Direct evidence that hepatocyte growth factor is a hepatotrophic factor for liver regeneration and has a potent antihepatitis effect in vivo. Hepatology 16, 1227–1235.

456 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

MIYAZAKI ET AL. Liu, M. L., Mars, W. M., Zarnegar, R., and Michalopoulos, G. K. (1994). Collagenase pretreatment and the mitogenic effects of hepatocyte growth factor and transforming growth factor-a in adult rat liver. Hepatology 19, 1521–1527. Webber, E. M., Godowski, P. J., and Fausto, N. (1994). In vivo response of hepatocytes to growth factors requires an initial priming stimulus. Hepatology 20, 489 – 497. Roos, F., Ryan, A. M., Chamow, S. M., Bennett, G. L., and Schwall, R. H. (1995). Induction of liver growth in normal mice by infusion of hepatocyte growth factor/scatter factor. Am. J. Physiol. 268, 380 –386. Michalopoulos, G. K., Ciancilli, H. D., Novotny, A. R., Kligerman, A. D., Strom, S. C., and Jirtle, R. L. (1982). Liver regeneration studies with rat hepatocytes in primary culture. Cancer Res. 42, 4673– 4682. Gohda, E., Yamasaki, T., Tsubouchi, H., Kurobe, M., Sakiyama, O., Aoki, H., Niidani, N., Shin, S., Hayashi, K., Hashimoto, S., Daikuhara, Y., and Yamamoto, I. (1990). Biological and immunological properties of human hepatocyte growth factor from plasma of patients with fulminant hepatic failure. Biochim. Biophys. Acta 1053, 21–26. Rana, B., Mischoulon, D., Xie, Y., Bucher, N. L., and Farmer, S. R. (1994). Cell-extracellular matrix interactions can regulate the switch between growth and differentiation in rat hepatocytes: Reciprocal expression of C/EBPa and immediate-early growth response transcription factors. Mol. Cell. Biol. 14, 5858 –5869. Block, G. D., Locker, J., Bowen, W. C., Petersen, B. E., Katyal, S., Strom, S. C, Riley, T., Howard, T. A., and Michalopoulos, G. K. (1996). Population expansion, clonal growth, and specific differentiation patterns in primary cultures of hepatocytes induced by HGF/SF, EGF and TGFa in a chemically defined (HGM) medium. J. Cell Biol. 132, 1133–1149. Peraino, C., Fry, R. J. M., and Staffeldt, E. (1971). Reduction and enhancement by phenobarbital of hepatocarcinogenesis induced in the rat by 2-acetylaminofluorene. Cancer Res. 31, 1506 –1512. Weisburger, J. H., Madison, R.M., Ward, J. M., Viguera, C., and Weisburger, E. K. (1975). Modification of diethylnitrosamine liver carcinogenesis with phenobarbital but not with immunosuppression. J. Natl. Cancer Inst. 54, 1185–1188. Kitagawa, T., and Sugano, H. (1978). Enhancing effect of phenobarbital on the development of enzyme-altered islands and hepatocellular carcinomas initiated by 39-methyl-4-(dimethylamino)azobenzene or diethylnitrosamine. Gann 69, 679 – 687. Miyazaki, M., Handa, Y., Oda, M., Yabe, T., Miyano, K., and Sato, J. (1985). Long-term survival of functional hepatocytes from adult rat in the presence of phenobarbital in primary culture. Exp. Cell Res. 159, 176 –190. Miyazaki, M., Handa, Y., Suzuki, Y., and Sato, J. (1987). Effect of various barbituric acid derivatives on survival of functional hepatocytes from adult rats in primary culture. Res. Exp. Med. 187, 105–117. Miyazaki, M., Suzuki, Y., Oda, M., Kawai, A., Bai, L., and Sato, J. (1989). Improved maintenance of adult rat hepatocytes in a new serum-free medium in the presence or absence of barbiturates. In Vitro Cell. Dev. Biol. 25, 839 – 848. Miyazaki, M., Handa, Y., Suzuki, Y., and Sato, J. (1987). Synthesis of barbituric acid derivatives and their effect on survival of functional hepatocytes from adult rats in primary culture. In Vitro Cell. Dev. Biol. 23, 2–9. Kost, D. P., and Michalopoulos, G. K. (1991). Effect of 2% dimethyl sulfoxide on the mitogenic properties of epidermal

30.

31.

32.

33.

34.

35.

36.

37.

38.

39. 40.

41.

42.

43.

44.

45.

46.

growth factor and hepatocyte growth factor in primary hepatocyte culture. J. Cell. Physiol. 147, 274 –280. Michalopoulos, G. K., and Pitot, H. C. (1975). Primary culture of parenchymal liver cells on collagen membranes. Morphological and biochemical observations. Exp. Cell Res. 94, 70 –78. Brunk, C. F., Jones, K. C., and James, T. W. (1979). Assay for nanogram quantities of DNA in cellular homogenates. Anal. Biochem. 92, 497–500. Schulz, W. A., Crawford, N., and Locker, J. (1988). Albumin and a-fetoprotein gene expression and DNA methylation in rat hepatoma cell lines. Exp. Cell Res. 174, 433– 447. Derynck, R., Jarrett, J. A., Chen, E. Y., Eaton, D. H., Bell, J. R., Assoian, R. K., Roberts, A. B., Sporn, M. B., and Goeddel, D. V. (1985). Human transforming growth factor-b complementary DNA sequence and expression in normal and transformed cells. Nature 316, 701–705. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M. B., Olson, J., and Klenk, D. C. (1985). Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76 – 85. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 – 685. Bradford, M. M. (1976). A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248 –254. Michalopoulos, G. M., Bowen, W., Nussler, A. K., Becich, M. J., and Howard, T. A. (1993). Comparative analysis of mitogenic and morphogenic effects of HGF/SF and EGF on rat and human hepatocytes maintained in collagen gels. J. Cell. Physiol. 156, 443– 452. Maher, J. J., Bissell, D. M., Friedman, S. L., and Roll, F. J. (1988). Collagen measured in primary cultures of normal rat hepatocytes derives from lipocytes within the monolayer. J. Clin. Invest. 82, 450 – 459. Waxman, D. J., and Azaroff, L. (1992). Phenobarbital induction of P-450 gene expression. Biochem. J. 281, 577–592. Gitlin, D., and Boesman, M. (1967). Sites of serum a-fetoprotein synthesis in the human and in the rat. J. Clin. Invest. 46, 1010 –1016. Tateno, C., and Yoshizato, K. (1996). Long-term cultivation of adult rat hepatocytes that undergo multiple cell divisions and express normal parenchymal phenotypes. Am. J. Pathol. 148, 383–392. Martinez-Hernandez, A., and Amenta, P. S. (1995). The extracellular matrix in hepatic regeneration. FASEB J. 9, 1401– 1410. Polin, R. A., Yoder, M. C., Douglas, S. D., McNelis, W., Nissim, I., and Yudkoff, M. (1989). Fibronectin turnover in the premature neonate measured with [15N]glycine. Am. J. Clin. Nutr. 49, 314 –319. Chou, L., Firth, J. D., Uitto, V. J., and Brunette, D. M. (1995). Substratum surface topography alters cell shape and regulates fibronectin mRNA level, mRNA stability, secretion and assembly in human fibroblasts. J. Cell. Sci. 108, 1563–1573. Jirtle, R. L., Hankins, G. R., Reisenbichler, H., and Boyer, I. J. (1994). Regulation of mannose 6-phosphate/insulin-like growth factor-a receptors and transforming growth factor beta during liver tumor promotion with phenobarbital. Carcinogenesis 15, 1473–1478. Jakowlew, S. B., Mead, J. E., Danielpour, D., Wu, J., Roberts, A. B., and Fausto, N. (1991). Transforming growth factor-b (TGF-b) isoforms in rat liver regeneration: Messenger RNA expression and activation of latent TGF-b. Cell Regul. 2, 535–548.

PHENOBARBITAL, HEPATOCYTE GROWTH, AND PHENOTYPIC CHANGES 47.

Peraino, C., Fry, R. J. M, Staffeldt, E., and Christopher, J. P. (1975). Comparative enhancing effects of phenobarbital, amobarbital, diphenylhydantoin, and dichlorodiphenyltrichloroethane on 2-acetylaminofluorene-induced hepatic tumorigenesis in the rat. Cancer Res. 35, 2884 –2890. 48. Edwards, A. M., and Lucas, C. M. (1985). Phenobarbital and some other liver tumor promoters stimulate DNA synthesis in cultured rat hepatocytes. Biochem. Biophys. Res. Commun. 131, 103–108. 49. Sawada, N., Staecker, J. L., and Pitot, H. C. (1987). Effects of tumor-promoting agents 12-O-tetradecanoylphorbol-13-acetate and phenobarbital on DNA synthesis of rat hepatocytes in primary culture. Cancer Res. 47, 5665–5671. 50. Yusof, Y. A. M., and Edwards, A. M. (1990). Stimulation of DNA synthesis in primary rat hepatocyte cultures by liver tumor promoters: Interactions with other growth factors. Carcinogenesis 11, 761–770. Received December 15, 1997 Revised version received March 13, 1998

51.

457

Manjeshwar, S., Rao, P. M., Rajalakshmi, S., and Sarma, D. S. R. (1992). Inhibition of DNA synthesis by phenobarbital in primary cultures of hepatocytes from normal rat liver and from hepatic nodules. Carcinogenesis 13, 2287–2291. 52. Meyer, S. A., and Jirtle, R. L. (1989). Phenobarbital decreases hepatocyte EGF receptor expression independent of protein kinase C activation. Biochem. Biophys. Res. Commun. 158, 652– 659. 53. Meyer, S. A., Gibbs, T. A., and Jirtle, R. L. (1989). Independent mechanisms for tumor promoters phenobarbital and 12-O-tetradecanoylphorbol-13-acetate in reduction of epidermal growth factor binding by rat hepatocytes. Cancer Res. 49, 5907–5912. 54. Miyazaki, M., Utsumi, K., and Sato, J. (1989). Mechanisms responsible for long-term survival of adult rat hepatocytes in the presence of phenobarbital in primary culture. Exp. Cell Res. 182, 415– 424.