Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (3)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Hufnagl, K.
Right arrow Articles by Grasl-Kraupp, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hufnagl, K.
Right arrow Articles by Grasl-Kraupp, B.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Carcinogenesis, Vol. 22, No. 8, 1247-1256, August 2001
© 2001 Oxford University Press

CARCINOGENESIS

Role of transforming growth factor {alpha} and prostaglandins in preferential growth of preneoplastic rat hepatocytes

Karin Hufnagl, Wolfram Parzefall, Brigitte Marian, Monika Käfer, Krystyna Bukowska, Rolf Schulte-Hermann and Bettina Grasl-Kraupp,1

Institut für Krebsforschung, University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
The role of transforming growth factor {alpha} (TGF{alpha}) and prostaglandins (PGs) in the preferential growth of preneoplastic liver cells was studied. Rats received the genotoxic hepatocarcinogen N-nitrosomorpholine (NNM); placental glutathione S-transferase (GSTp) was used as a marker to identify preneoplastic foci. Preneoplastic foci expressing TGF{alpha} (TGF{alpha}+) grew more rapidly than TGF{alpha} negative (TGF{alpha}–) ones. Almost all tumours studied were positive for TGF{alpha}. The key enzymes of prostaglandin synthesis, cyclooxygenase I (Cox-1) and II (Cox-2), were present in all unaltered and preneoplastic cells and tended to decrease in the later stages of hepatocarcinogenesis. Immunostaining revealed that cultures of hepatocytes, isolated from NNM-treated livers by collagenase perfusion, contained 1–2% GSTp-positive (GSTp+) and 9% TGF{alpha}+ hepatocytes; 0.6% of the cells were GSTp+/TGF{alpha}+. Cox-1 and Cox-2 were present in all cells. DNA replication was almost exclusively associated with expression of TGF{alpha}. GSTp+ hepatocytes showed a 3- to 4-fold higher probability of TGF{alpha} expression and of DNA synthesis than GSTp-negative (GSTp–) cells. PGE2 or PGF2{alpha} increased expression of TGF{alpha} and DNA replication in GSTp– cells but not in GSTp+ cells. PGA2 and PGJ2 decreased DNA synthesis in TGF{alpha}+ cells without an obvious effect on the intracellular levels of TGF{alpha}. The Cox-2 inhibitor SC236 suppressed DNA replication preferentially in GSTp+ cells; this inhibition was reversed by PGE2/F2{alpha}. Indomethacin had no effect. These results suggest the following conclusions. (i) Growth regulation of preneoplastic GSTp+ cells in culture exhibits distinct differences from GSTp– cells and elevated expression of TGF{alpha} contributes to their growth advantage. (ii) TGF{alpha} renders preneoplastic hepatocytes sensitive to suppression of DNA synthesis by PGA2/J2. (iii) SC236, a Cox-2 inhibitor, may have preventive value in hepatocarcinogenesis.

Abbreviations: BSA, bovine serum albumin; Cox-1, cyclooxygenase type I; Cox-2, cyclooxygenase type II; DMSO, dimethylsulphoxide; EGF, epidermal growth factor; GSTp, placental glutathione S-transferase; GSTp–, hepatocytes negative for placental glutathione S-transferase; GSTp+, hepatocytes positive for placental glutathione S-transferase; LI, labelling index; NNM, N-nitrosomorpholine; PB, phenobarbital; PG, prostaglandin; PPAR{gamma}, peroxisome proliferator activated receptor {gamma}; TBS, Tris-buffered saline; TGF{alpha}, transforming growth factor {alpha}; TGF{alpha}–, negative for transforming growth factor {alpha}; TGF{alpha}+, positive for transforming growth factor {alpha}.


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Liver cancer is among the eight leading causes of cancer death world wide; therapeutic possibilites are limited and prognosis is usually poor (1). Chronic cytotoxic and inflammatory events favour the development of this disease, indicating that mediators of inflammation, such as prostaglandins (PGs), may be causally involved in the pathogenesis (1–5). In many malignant tumours, including some liver tumours, increased levels of PGs, most notably PGE2 and PGF2{alpha}, have been detected (3,6). PGs stimulate tumour growth and they presumably act in most stages of carcinogenesis (3,4,7). Interest in PGs has increased even further after the observation that growth of adenomatous polyps in the colon can be delayed or prevented by treatment with non-steroidal anti-inflammatory drugs, inhibitors of PG synthesis (8,9). These findings have raised the question of whether decreased levels of PGs may be of prophylactic benefit for liver cancer.

Rat liver offers an excellent tool for mechanistic studies on all stages of hepatocarcinogenesis. When applying placental glutathione S-transferase (GSTp) as marker, putatively initiated GSTp-positive (GSTp+) single cells are detectable a few days after treatment with various genotoxic carcinogens, such as N-nitrosomorpholine (NNM) (10,11). A portion of these cells develops to preneoplastic GSTp+ foci, some of which evolve into adenomas and carcinomas (12–14). From the 2-cell stage onwards GSTp+ foci exhibit elevated rates of cell replication and cell death by apoptosis; this elevated cell turnover increased even further in hepatocellular adenomas and carcinomas (13,15). Tumour promoters, such as phenobarbital (PB) and the progestin cyproterone acetate, or increased food intake led to preferential inhibition of apoptosis of preneoplastic liver cells, thereby accelerating the selective growth of the lesion (16–18). This indicates enhanced sensitivity of preneoplastic cells to endogenous tumour promoters. The reasons for the enhanced sensitivity of liver preneoplasia and the identity of endogeneous tumour promoters are still known only partially (19,20). Recently preneoplastic liver cells at a very early stage of development have become available for investigation in vitro (21). This model, in combination with studies in vivo, has been used here to investigate the mechanisms underlying the growth advantage of preneoplastic hepatocytes.

PGs are important growth regulators in the liver of rats and humans, e.g. PGE2 and PGF2{alpha} are stimulatory for hepatocyte proliferation, while inhibitory effects have been attributed to PGA2, PGJ2 and PGD2 (22–25). The significance of PGs for hepatocarcinogenesis in rats has become evident in experiments with inhibitors of the key enzymes of PG synthesis, cyclooxygenase type I (Cox-1) and cyclooxygenase type II (Cox-2). Both Cox-1 and Cox-2 are inhibited by indomethacin, acetylsalicylic acid and other `classical' non-steroidal anti-inflammatory drugs, while a new generation of anti-inflammatory drugs, including SC236, has been developed to selectively affect Cox-2 (26–28). Non-steroidal anti-inflammatory drugs reduce proliferation of cultured hepatoma cells, inhibit the progression from G1 to S phase and may also induce apoptosis (29). Several studies documented a chemopreventive effect of non-steroidal anti-inflammatory drugs in rat hepatocarcinogenesis. Acetylsalicylic acid reduced the number and size of GSTp+ preneoplastic foci in the liver of rats fed a choline-deficient diet (30,31). In the post-initiation phase acetylsalicylic acid also antagonized growth promotion of GSTp+ foci by PB or acetylaminofluorene (32–34).

The mechanisms by which non-steroidal anti-inflammatory drugs inhibit hepatocarcinogenesis remain to be clarified. Evidence is increasing that in the liver PGs interfere with the network of other growth factors, such as epidermal growth factor (EGF) and transforming growth factor {alpha} (TGF{alpha}) (23,25,35,36). EGF and TGF{alpha} are functionally and structurally related peptides (37). They are produced by several tissues in demand for cell replication and are secreted outside the cell, where they may bind to and activate the EGF receptor, which transfers the signal via phosphorylation cascades to the nucleus. During hepatocarcinogenesis in experimental animals and in humans TGF{alpha} is synthesized by (pre)malignant cells and seems to stimulate tumour growth by an autocrine loop (38–41). Transgenic mice overexpressing TGF{alpha} in the liver develop hepatocellular carcinoma with high incidence and multiplicity (42–44). We have recently shown that TGF{alpha} is already overexpressed in the first stages of hepatocarcinogenesis and is associated with enhanced DNA synthesis (45).

PGs have been found to regulate the effects of exogenous EGF and TGF{alpha} on DNA synthesis in cultured hepatocytes (22). During regenerative liver growth PGs are produced locally (46,47). No data are available on the roles PGs and TGF{alpha} may play in the proliferation advantage of early prestages of liver cancer. Therefore, the following questions were investigated: (i) is intracellular TGF{alpha} involved in the enhanced DNA replication of preneoplastic cells; (ii) do PGE2/F2{alpha} and PGA2/J2 modify TGF{alpha} expression and DNA synthesis of unaltered and preneoplastic hepatocytes; (iii) do Cox inhibitors reduce DNA synthesis and TGF{alpha} expression in unaltered and preneoplastic hepatocytes? To tackle these questions unaltered and preneoplastic GSTp+ cells were studied in rat liver in vivo and in our newly developed cell culture model (21). The results suggest that: (i) TGF{alpha} expression plays a key role in the enhanced growth of preneoplastic GSTp+ hepatocytes; (ii) GSTp+ cells overexpressing TGF{alpha} do not require additional stimulation of DNA synthesis by PGE2/F2{alpha} but are the target of suppression of DNA synthesis by PGA2/J2 and by the Cox-2 inhibitor SC236. This supports the concept that Cox-2 inhibition may be of benefit for prevention of liver tumours.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Animals and treatment
Male SPF Wistar rats, ~3 weeks old, were obtained from the Forschungsinstitut für Versuchstierzucht und Versuchstierhaltung (Himberg, Austria). Animals were kept under standardized conditions (macrolon cages, 20 ± 3°C room temperature, 40–70% relative humidity, inverted light–dark cycle with lights on from 10 p.m. to 10 a.m.) and were fed Altromin 1321N (Altromin, Lage, Germany). Three weeks before treatment animals were adapted to rhythmic feeding (from 9 a.m. to 2 p.m.). This procedure synchronizes DNA synthesis in the liver to a single peak per day at ~8 p.m. (details in ref. 18). After adaptation animals were treated with NNM (Sigma, St Louis, MO); immediately before application NNM was dissolved in phosphate-buffered saline, pH 7.4, and was given as a single dose of 250 mg NNM/10 ml solution/kg body wt by gavage between 8 and 9 p.m. (18).

Immediately after treatment the daily food consumption decreased from 15.9 ± 1.0 to 3.6 ± 0.5 g at day 3 and gradually recovered to 15.9 ± 1.1 g on the following days. On days 4, 5 and 6 after NNM treatment PB (Fluka AG, Buchs, Switzerland) was dissolved in tap water and administered by gavage as a dose of 50 mg/10 ml/kg body wt at the end of the daily feeding period at 3 p.m. From day 7 onwards, when the animals had regained their normal food intake, PB was mixed with the powder diet (Altromin 1321N) and PB concentrations were adjusted every 14 days to provide a daily dose of 50 mg/kg body wt. Controls were fed the basal diet ad libitum, since PB treatment did not alter food intake or body weight (not shown). Animals were killed by decapitation under CO2 narcosis. Time points of killing were 8, 14, 20, 51, 91 and 266 days post-NNM treatment. The experiment was performed according to Austrian guidelines for animal care and protection.

Histology
Specimens of liver tissue were fixed in paraformaldehyde and were processed as described (15). Four serial sections, 2 µm thick, were cut; one of the sections was stained for TGF{alpha}, the second for GSTp, the third for Cox-1 and the fourth for Cox-2.

Immunostaining for GSTp, TGF{alpha}, Cox-1 and Cox-2
The primary antibodies used were: rabbit polyclonal IgGs raised against the Yp subunit of the rat enzyme (Biotrin International, Dublin, Eire); mouse monoclonal IgG raised against recombinant mature TGF{alpha} (originally clone 213-4.4; Oncogene Science, Uniondale, NY); goat polyclonal IgG raised against the C-terminus of rat Cox-1 (C-20; Santa Cruz Biotechnology, Santa Cruz, CA); mouse monoclonal IgG raised against purified sheep Cox-1(Cayman Chemical, Ann Arbor, MI); goat polyclonal IgG raised against the N-terminus of rat Cox-2 (N-20; Santa Cruz); mouse monoclonal IgG raised against C-terminal amino acids 368–604 of rat Cox-2 (clone 33; Transduction Laboratories, Lexington, KY). The following schedule was used: hydrogen peroxide to block endogenous peroxidases (3%) for 20 min at room temperature; 2.5% bovine serum albumin (BSA) in Tris-buffered saline (TBS) (0.05 M Tris, 0.3 M NaCl, pH 7.6) for 30 min at room temperature; incubation overnight at 4°C with primary antisera diluted in 2.5% BSA/TBS (anti-Yp, 1:5000; anti-TGF{alpha}, 1:50; mouse anti-Cox-1 and mouse anti-Cox-2, 1:50; goat anti-Cox-1 and goat anti-Cox-2, 1:100); rinsing with TBS; incubation for 90 min at room temperature with secondary antisera diluted in BSA/TBS [biotinylated goat anti-rabbit IgG, 1:600 (Dakopatts, Glostrup, Denmark); biotinylated rabbit anti-mouse IgG, 1:600 (Dakopatts); biotinylated mouse anti-goat IgG, 1:200 (Sigma)]; rinsing with TBS; addition of streptavidin–horseradish peroxidase conjugate (1:300 in TBS; Dakopatts) for 45 min at room temperature; colour development with diaminobenzidine. The specificity of immunohistochemistry was confirmed by omitting the primary antibodies.

Quantitative evaluation of GSTp+ foci
GSTp+ single cells and GSTp+ multicellular foci were identified with an anti-GSTp stain. The number of all foci consisting of at least five GSTp+ cells per cross-section were registered; the number of GSTp+ cells per focus cross-section was counted.

Primary hepatocyte cultures
Male Wistar rats were treated with a single dose of NNM (250 mg/kg body wt) as described above. Twenty-one days later livers were perfused with collagenase according to the technique of Seglen (48), with modifications as described (21,49). Before perfusion the processus papilliformus caudatus of the liver was tied and cut off for histology (see above for details).

Treatment of primary hepatocyte cultures
Petri dishes (3.5 cm diameter; NUNC, Roskilde, Denmark) were coated with diluted collagen (1:10 in distilled water). Cells were seeded at a densitiy of 30 000 viable cells/cm2 in 2 ml of William's E2 medium plus 10% foetal calf serum and were incubated at 37°C in an incubator with 5% CO2 in air at 98% relative humidity. After an attachment period of 1 h the monolayers were rinsed with medium and were left in 1.5 ml of serum-free Williams E2 medium for 2–3 h to recover from the isolation stress.

PGE2, PGF2{alpha}, PGA2 and PGJ2 were obtained from Biomol (Plymouth Meeting, PA). PGE2, PGF2{alpha} and PGJ2 were dissolved in 100% ethanol (100 mM stock for PGE2 and PGF2{alpha}; 50 mM stock for PGJ2). PGA2 was obtained as a solution in 100% acetone; the acetone was evaporated under liquid nitrogen and the remnant was dissolved quickly in 100% ethanol to produce a stock of 100 mM. Immediately before treatment stock solutions were further diluted in medium and aliquots were added to achieve the final concentrations. Indomethacin (Sigma) was diluted in 100% ethanol to give a 50 mM stock. In all treated and control groups the final concentration of ethanol was 0.1% in the medium. SC236 (4-[5-(4-chlorophenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide), generously supplied by Searle (Seattle, WA), was dissolved in dimethylsulphoxide (DMSO) to give a 20 mg/ml stock; 1 µl of a stock of recombinant human TGF{alpha} (10 µg in 10 µl of 10 mM acetic acid; UBI, Lake Placid, NY) was added per ml medium; tyrphostin A25 (synonym tyrphostin AG82; Calbiochem, La Jolla, CA) was prepared as a stock solution of 10 mg in 1 ml of DMSO and was added to the medium to a final concentration of 20 µg/ml; the concentration of DMSO in the medium of all groups, including the controls, was 0.2%. Treatment commenced 4 h after plating (time point 0 in the experimental protocol). Treatment with PGs and indomethacin was renewed with a medium change after 20 h. SC236 was applied only once, 4 h after plating, and the medium change 20 h later was omitted.

Double immunostaining for GSTp and TGF{alpha}
Cells in culture were fixed for 90 min at room temperature with 4% buffered formalin according to Lillie and were then kept in distilled water at 4°C until immunostaining. Then the following schedule was used: hydrogen peroxide to block endogenous peroxidases (3%) for 20 min at room temperature; primary antibodies diluted in 2.5% BSA in TBS (rabbit anti-GSTp, 1: 5000; mouse-anti-TGF{alpha}, 1:50) applied overnight at 4°C; rinsing with TBS; secondary antibodies diluted in 2.5% BSA/TBS [biotinylated goat-anti-mouse IgG, 1:600 (Dakopatts); alkaline phosphatase-labelled goat anti-rabbit] applied for 90 min at room temperature; rinsing with TBS; incubation with streptavidin–horseradish peroxidase conjugate (1:300 in TBS; Dakopatts) for 45 min at room temperature; colour development with diaminobenzidine (Sigma) and nitro-blue tetrazolium salt/5-bromo-4-chloro-3-indolylphosphate. The specificity of immunohistochemistry was confirmed by omitting both primary antibodies or anti-TGF{alpha} and anti-Yp only.

Double immunostaining for GST-P and Cox-1/Cox-2
The protocol exactly followed the procedure for double immunostaining for GSTp and TGF{alpha} except for the following modifications: anti-GSTp (anti-Yp, 1:500) and anti-Cox-1 or anti-Cox-2 (both 1:100; Santa Cruz Biotechnology) were applied together; alkaline phosphatase–mouse anti-rabbit antibody conjugate (1:100; Sigma) was applied, followed by biotinylated rabbit anti-goat antibody (1:200; Dakopatts) and horseradish–streptavidin conjugate; colour development as described above. The specificity of immunohistochemistry was confirmed by omitting both primary antibodies or anti-GSTp or anti Cox-1/anti Cox-2 only.

Determination of DNA replication and apoptosis in cultures
Newly synthesized DNA was labelled with [3H]thymidine (sp. act. 60– 80 Ci/mmol; ARC, St Louis, MO), which was added at 0.5 µCi/ml medium 24 h before harvesting.

Immunohistochemically stained plates (see above) were coated with 1% gelatin, 0.05% chromalaun, air dried and were then dipped into Ilford K5 photoemulsion (Ilford, Dreieich, Germany). Time of exposure was on average 24 h. Immediately after autoradiography plates were stained for 5 min with 8 µg/ml Hoechst 33258 (Riedel de Haen, Seelze, Germany) in phosphate-buffered saline, followed by two washes with distilled water. The plates were dried at room temperature and mounted in Kayser's glycerin gelatin (Merck, Darmstadt, Germany).

For determination of DNA synthesis a total of 1000 GSTp-negative (GSTp–) and TGF{alpha}-negative (TGF{alpha}–), 200–300 GSTp+ and 300–400 TGF{alpha}-positive (TGF{alpha}+) hepatocytes were evaluated per plate. Labelling index (LI) was calculated as percentage labelled hepatocyte nuclei of total number of hepatocyte nuclei counted. For evaluation of apoptosis at least 1000 hepatocyte nuclei per plate were scored for apoptotic morphology (condensed and fragmented nuclei) under a fluorescence microscope as described (50). The percentage of apoptotic bodies was calculated (apoptosis index).

Statistics
Routinely between four and six equally treated hepatocyte cultures per rat were run and harvested in parallel. If not stated otherwise, the means ± SD (vertical lines) of three cultures from at least three different donor rats are given. The significance of differences of means was tested either by Wilcoxon's test, the paired t-test or the Kruskal–Wallis test.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Expression of TGF{alpha} in unaltered liver and in GSTp+ preneoplastic foci
In immunostained sections of rat liver TGF{alpha} protein was found in all bile epithelial cells and in hepatocytes close to the central venules, as in previous reports (41,51). PB treatment did not alter the distribution or intensity of TGF{alpha} expression (not shown).

Treatment with the initiating carcinogen NNM induced GSTp+ preneoplastic foci in rat liver. A sub-group of animals received the potent tumour promoter PB from day 4 post-NNM onwards. As a result, the size of the foci expanded dramatically (data not shown), as described previously (16,31,40).

At day 91 post-NNM 25% of the GSTp+ foci in control or PB-treated rats contained TGF{alpha}+ cells, which occurred at a frequency of ~10% within the lesion. TGF{alpha}+ foci were significantly larger than TGF{alpha}– foci without and with tumour promotion by PB (Figure 1Go).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 1. TGF{alpha}+ foci are significantly larger than TGF{alpha}– foci. The average sizes of TGF{alpha}– and TGF{alpha}+ foci have been determined by the average number of GSTp+ cells per focus cross-section. At least four animals per time point and group were evaluated. Means ± SD are given. Statistics by paired t-test: aP < 0.05 for TGF{alpha}– versus TGF{alpha}+ foci.

 
Expression of Cox-1 and Cox-2 in unaltered liver and in GSTp+ preneoplastic foci
Cox-1 and Cox-2 were detected in the cytoplasm of all unaltered hepatocytes (Figure 2AGo). Some staining for both enzymes could be seen occasionally in fibroblasts and in hepatic stellate, Kupffer and vascular smooth muscle cells. In controls the intensity of hepatocellular expression was homogeneous within the liver lobule, while PB treatment induced both enzymes strongly in hepatocytes close to periportal areas (Figure 2BGo). Immunostains with two different antibodies against Cox-1 and against Cox-2 were identical.



View larger version (178K):
[in this window]
[in a new window]
  		Fig. 2. Expression of GSTp, TGF{alpha}, Cox-1 and Cox-2 in unaltered and (pre)neoplastic hepatocytes. Day 20 post-NNM treatment: expression of Cox-1 in unaltered liver of a control (A) and of a PB-treated rat (B); large arrows indicate periportal areas; the small arrow indicates the central venule. Day 91 post-NNM: a GSTp+ focus (C) in a PB-treated rat expressing Cox-1 (D) and Cox-2 (E); arrows indicate the border of the focus. Day 260 post-NNM: a hepatocellular carcinoma in a PB-treated rat expressing TGF{alpha} (F), Cox-1 (G) and Cox-2 (H). Nuclei were counterstained with hematoxylin. Magnifications: (A and B) x100; (C–H) x200.

 
At day 91 post-NNM all of the GSTp+ foci studied in controls expressed Cox-1 and Cox-2 with the same intensity as the surrounding tissue (in three animals 33 foci were evaluated with a total of 1232 GSTp+ cells). In GSTp+ foci of the PB-treated group both enzymes appeared induced to the level in periportal regions even when the foci were located near to central venules (in three animals 47 foci were evaluated with a total of 2440 GSTp+ cells) (Figure 2D and EGo). This suggests similar induction of Cox-1 and Cox-2 in unaltered periportal and preneoplastic cells.

Expression of GSTp, TGF{alpha}, Cox-1 and Cox-2 in hepatocellular carcinoma
After NNM and 260 days of PB treatment liver tumours had developed. Fifteen of 16 GSTp+ hepatocellular carcinomas studied (93%) contained TGF{alpha}+ cells at a frequency of nearly 100% (Figure 2FGo); 63 and 50% of the tumours expressed Cox-1 and Cox-2 at an extent similar to the perivenous tissue, respectively; the remaining tumours exhibited reduced levels (Figure 2G and HGo).

Expression of GSTp, TGF{alpha}, Cox-1 and Cox-2 in hepatocytes in primary culture
On days 20–22 post-NNM livers were perfused with collagenase to isolate and cultivate hepatocytes. To evaluate the reliability of the in vitro culture system we determined the percentages of GSTp+ and TGF{alpha}+ hepatocytes in tissue sections of a liver lobe left intact at perfusion and after 48 h in culture (see Materials and methods): the mean frequencies of GSTp+ hepatocytes were 1.92 ± 0.81% in intact liver and 1.63 ± 0.67% in culture; the frequencies of TGF{alpha}+ cells were 14.90 ± 1.9% in intact liver and 12.65 ± 0.67% in culture. This suggests that no selective loss or enrichment of GSTp+ and TGF{alpha}+ cells occurred during isolation and cultivation of the cells. Cultures were composed of all four possible categories of hepatocytes, which were present at the following percentages: GSTp–/ TGF{alpha}–, 89.9 ± 0.71%; GSTp–/ TGF{alpha}+, 8.36 ± 1.04%; GSTp+/ TGF{alpha}–, 1.05 ± 0.42%; GSTp+/ TGF{alpha}+, 0.59 ± 0.23%. Data were obtained from five animals (Figure 3A and BGo).



View larger version (119K):
[in this window]
[in a new window]
  		Fig. 3. Expression of GSTp, TGF{alpha} and Cox-1 in hepatocytes isolated from NNM-treated livers and kept in primary culture. Untreated cultures at 24 h: (A) a TGF{alpha}+ hepatocyte; (B) all hepatocytes are positive for Cox-1; the single GST+ hepatocyte is indicated by an arrow. Magnification: x400.

 
Immunocytochemical stains of primary cultures with two different antibodies against Cox-1 and against Cox-2 revealed that all hepatocytes expressed both isoenzymes in the cytoplasm (Figure 3BGo).

DNA synthesis of hepatocytes in culture with or without expression of TGF{alpha}
After 48 h in culture 63.7 ± 9.98% of all TGF{alpha}+ hepatocytes showed replicative DNA synthesis, which contrasts with only 1.54 ± 0.82% of TGF{alpha}– cells replicating DNA (n = 5, P < 0.0001, Wilcoxon test). Thus, the presence of TGF{alpha} is strongly associated with replicative DNA synthesis.

Addition of mature TGF{alpha} to cultures exerted no additional effect on the already high DNA synthesis of TGF{alpha}+ cells (Table IGo). In contrast, mature TGF{alpha} dramatically increased DNA synthesis in TGF{alpha}– hepatocytes, which was blocked by the EGF receptor tyrosine kinase inhibitor tyrphostin A25, according to classical concepts of growth signal transduction. Tyrphostin A25 exerted some effect on the limited DNA replication of TGF{alpha}– cells, which is indirect evidence for TGF{alpha} secretion into the medium (Table IGo). In conclusion, under our conditions hepatocytes expressing TGF{alpha} have a selective growth advantage and appear insensitive to additional growth stimulation by TGF{alpha} treatment.


View this table:
[in this window]
[in a new window]
  		Table I. Induction of DNA synthesis by TGF{alpha} selectively in TGF{alpha}– cells: blockade of TGF{alpha} effects by tyrphostin A25
 
Growth advantage of cultured GSTp+ hepatocytes due to expression of TGF{alpha}
In the intact liver rates of DNA replication are significantly higher in GSTp+ than in GSTp– hepatocytes (13). This difference persists in our ex vivo cell culture system, indicating an inherent growth advantage of the putatively initiated cell population (Figure 4Go), as has been shown before (21). We asked whether expression of TGF{alpha} is involved in the enhanced propensity of GSTp+ cells for replication. While only 8.5 ± 1.2% of GSTp– cells were positive for TGF{alpha}, 35.9 ± 3.8% of GSTp+ hepatocytes expressed TGF{alpha}, i.e. an ~4-fold higher frequency. In conclusion, the inherent growth advantage of GSTp+ cells is associated with an enhanced probability of synthesizing TGF{alpha} (Figure 4Go).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4. GSTp+ hepatocytes in culture express TGF{alpha} and synthesize DNA [LI (%)] more frequently than unaltered GSTp– cells. A total of 20 000 GSTp– cells and 6000 GSTp+ cells were evaluated. Means ± SD are given from 10 independent experiments at 48 h. Statistics by Wilcoxon test: aP < 0.01 for GSTp– versus GSTp+ cells.

 
Effects of PGE2 and PGF2{alpha} on TGF{alpha} expression and on DNA synthesis in GSTp– and GSTp+ hepatocytes in culture
PGE2 and PGF2{alpha} had no effect on cultured hepatocytes after 24 h treatment (not shown). At 48 h in the GSTp– population there was a dose-dependent increase in the percentage of cells expressing TGF{alpha} and of cells replicating DNA; the intracellular intensity of TGF{alpha} expression before and after PGE2/F2{alpha} treatment appeared identical, implying that only the number of TGF{alpha}+ cells was altered. In contrast, GSTp+ cells remained largely unaffected by PGE2/F2{alpha} (Figure 5Go). Figure 6Go shows that the increase in DNA synthesis occurred in GSTp– cells that had become TGF{alpha}+. This may indicate that PGE2 and PGF2{alpha} stimulate DNA replication by increasing the pool of hepatocytes expressing TGF{alpha}, i.e. upstream of TGF{alpha}. Accordingly, PGE2 and PGF2{alpha} do not affect GSTp+ cells, which are inherently rich in TGF{alpha}.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. PGE2 and PGF2{alpha} induce DNA synthesis and TGF{alpha} expression in GSTp– hepatocytes. For absolute LI (%) and absolute percentage of TGF{alpha}+ cells in the GSTp– and GSTp+ populations in controls at 48 h see Figure 4Go. Means ± SD are given from four independent experiments. Statistics by Kruskal–Wallis test: aP < 0.05 for GSTp– versus GSTp+ cells and for the relative increase in DNA synthesis or in the expression of TGF{alpha} in GSTp– cells.

 


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 6. Coincident induction of TGF{alpha} expression and DNA synthesis in GSTp– hepatocytes by PGE2 and PGF2{alpha}. Left ordinate, percentage of TGF{alpha}+ cells in the GSTp– cell population; right ordinate, LI (%) in TGF{alpha}+ cells in the GSTp– cell population. Means ± SD are given from four independent experiments at 48 h. Statistics by Kruskal–Wallis test: aP < 0.01 for the relative increases in DNA synthesis and in the expression of TGF{alpha} in GSTp– cells.

 
Effect of PGA2 and PGJ2 on expression of TGF{alpha} and on DNA synthesis in GSTp– and GSTp+ hepatocytes in culture
Cultured GSTp– and GSTp+ cells were treated with the growth inhibiting PGs PGA2 und PGJ2. The frequency of TGF{alpha}+ hepatocytes remained unchanged in both cell populations at all concentrations and time points studied (data not shown). At 24 h PGA2 and PGJ2 suppressed DNA synthesis selectively in GSTp– and GSTp+ cells that were TGF{alpha}+ (Figure 7Go); this effect had disappeared at the 48 h time point (not shown). In conclusion, PGA2/J2 appear to antagonize growth stimulation by endogenous TGF{alpha} without affecting its expression. This suggests that in the growth regulatory pathway PGA2/J2 may act downstream of TGF{alpha}.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 7. PGA2 and PGJ2 suppress DNA synthesis in TGF{alpha}+ hepatocytes. Absolute LI (%) of controls in GSTp–/TGF{alpha}– cells 0.27 ± 0.23%, in GSTp–/TGF{alpha}+ cells 16.54 ± 6.0%, in GSTp+/TGF{alpha}– cells 0.87 ± 0.17% and in GSTp+/TGF{alpha}+ cells 15.3 ± 2.9%. Means ± SD are given from four independent experiments at 24 h. Statistics by Kruskal–Wallis test: aP < 0.05 for TGF{alpha}– versus TGF{alpha}+ cells and for the relative decrease in DNA synthesis of TGF{alpha}+ cells; bP < 0.05 for the relative decrease in DNA synthesis of TGF{alpha}+ cells; cP < 0.05 for TGF{alpha}– versus TGF{alpha}+ cells.

 
Effect of the Cox inhibitors indomethacin and SC236 on expression of TGF{alpha} and on DNA synthesis in GSTp– and GSTp+ hepatocytes in culture
Indomethacin showed no effect on DNA replication or TGF{alpha} expression at any time point studied, even at concentrations that were close to the limit of solubility in 100% ethanol (not shown); a higher final concentration than 0.1% ethanol in the medium is toxic to hepatocytes (52).

SC236 suppressed DNA synthesis selectively in GSTp+ cells in a dose-dependent manner at the 24 h time point (Figure 8Go). This effect was so pronounced that it decreased replication of GSTp+ cells to the low level of GSTp– hepatocytes. After 48 h GSTp– cells were also somewhat affected (Figure 9Go). The expression of TGF{alpha} was not altered by SC236 (not shown).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 8. SC236 inhibits DNA synthesis in cultured GSTp+ hepatocytes. GSTp+ cells. Means ± SD are given from six independent experiments each at 24 h. Statistics by Kruskal–Wallis test: aP < 0.05 for the relative decrease in DNA synthesis of GSTp+ cells.

 


View larger version (31K):
[in this window]
[in a new window]
  		Fig. 9. Suppression of DNA synthesis in cultured hepatocytes by SC236 is abrogated by PGE2 and PGF2{alpha}. Open bars represent GSTp– cells, solid bars GSTp+ cells. For absolute LI (%) of controls see Figure 4Go. Means ± SD are given from three independent experiments at 48 h. Statistics by ANOVA analysis followed by a multiple Range test: aP < 0.05 for combined treatment versus treatment with PGE2 or PGF2{alpha} alone; bP < 0.05 for combined treatment versus treatment with SC236.

 
To study whether the growth inhibitory effect of SC236 may be due to depletion of endogenous PGs, SC236 was applied together with PGE2 or PGF2{alpha} (Figure 9Go). At 48 h, when growth stimulation by PGs was evident, suppression of DNA synthesis by SC236 was abrogated completely by addition of PGE2 or PGF2{alpha} at 50 (Figure 9Go) or 100 µM (not shown).

Frequency of apoptosis in cultured hepatocytes
Treatment with 50 µM PGA2 increased the frequency of apoptotic hepatocytes at least 5-fold after 24 and 48 h culture, reaching significance at the later time point (controls, 0.3 ± 0.07%; PGA2, 1.62 ± 1.13%; P < 0.05, Wilcoxon's test). PGE2, PGF2{alpha}, PGJ2, indomethacin and SC236 did not significantly alter the rate of apoptosis in primary hepatocyte cultures (not shown).


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
The present paper describes TGF{alpha} in early and later stages of hepatocarcinogenesis liver tumours as a key regulatory growth factor that modifies the stimulatory or inhibitory effects of PGs on DNA synthesis. A basal level of PGs appears necessary for growth of liver preneoplasias, since Cox-2 inhibition suppressed DNA replication of the preneoplastic cells. Several implications of these findings are discussed below.

TGF {alpha} as a growth factor for hepatocytes
In the present study DNA synthesis occurred almost only in cultured hepatocytes expressing TGF{alpha} (Figures 4 and 6GoGo). This coincidence of DNA replication and TGF{alpha} expression was further induced by PGE2 and PGF2{alpha} (Figure 5Go). Likewise, treatment of cultures with cyproterone acetate and stimulation of regenerative or hyperplastic liver growth in intact animals increased the frequency of hepatocytes expressing TGF{alpha} and synthesizing DNA; induction of TGF{alpha} preceded induction of DNA replication and occurred in G1 of the cell cycle (submitted for publication). This is evidence that several different growth stimuli, including PGE2 and PGF2{alpha}, may act by increasing intracellular TGF{alpha} in G1.

TGF{alpha} as a key determinant of growth of liver (pre)neoplasia
Preneoplastic GSTp+ cells show rates of DNA replication and of apoptosis ~3-fold higher than unaltered hepatocytes, which is evident in intact liver and in culture (Figure 4Go; 13,21). This finding suggests that preneoplastic hepatocytes bear an inherent defect in growth regulation that persists outside the body for at least 48 h in culture and, therefore, appears independent of intercellular contacts, endocrine or paracrine factors. The present work shows that considerably more GSTp+ cells express TGF{alpha} than do GSTp– cells (Figure 4Go). This may explain the intrinsic growth advantage of GSTp+ cells and supports their preneoplastic state.

In rats the expression of TGF{alpha} increases further in the course of hepatocarcinogenesis, reaching peak levels in carcinoma (Figures 1 and 2GoGo; 39,41). Human hepatocellular carcinoma and the pre-stages also show pronounced up-regulation of TGF{alpha} (38,40). This illustrates the similarity between rat and human hepatocarcinogenesis and supports the importance of TGF{alpha} for the development of liver tumours.

Possible interactions of TGF{alpha} with PGE2/F2{alpha}- and PGA2/J2-induced signal transduction pathways
PGE2/F2{alpha} stimulate DNA synthesis via preferential induction of TGF{alpha} expression in TGF{alpha}-poor GSTp– cells (Figure 5Go); without growth stimulation these cells usually do not cycle (19–21). On the other hand, PGA2/J2 suppress DNA replication in TGF{alpha}-rich hepatocytes that are mostly GSTp+ (Figure 7Go); these cells often synthesize DNA and therefore go through G1. These findings are in accord with the assumed molecular modes of action of PGE2/F2{alpha} and PGA2/J2. PGE2/F2{alpha} stimulate growth via indirect activation of MAPK (36,53). The TGF{alpha}–EGF receptor-induced signal transduction cascade also includes MAPK (37). Thus PGE2/F2{alpha} appear to act upstream of TGF{alpha} and to induce TGF{alpha} and TGF{alpha}-evoked transduction pathways in cells lacking this growth factor. PGA2 and PGJ2 are considered ligands for the transcription factor peroxisome proliferator-activated receptor {gamma} (PPAR{gamma}), which is expressed in hepatocytes (54). PGA2/J2 binding to a RAR/RXR–PPAR{gamma} complex is accompanied by G1 arrest of the cell cycle (55). In the present study expression of TGF{alpha} remained unaffected by PGA2/J2. From these findings it may be deduced that in the growth regulatory pathway PGA2/J2 exert their effects downstream of TGF{alpha}, i.e. after TGF{alpha} had driven the cell into G1. Thus the presence or absence of TGF{alpha} seems to determine whether growth stimulating or growth inhibiting PGs become effective (Figure 10Go).



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 10. Hypothetical mechanisms of interactions of TGF{alpha} with PGE2/F2{alpha} and PGA2/J2 in growth regulation of unaltered and preneoplastic hepatocytes. PGE2 and PGF2{alpha} appear to act upstream of TGF{alpha} in the growth regulatory pathways via induction of TGF{alpha}, which pushes unaltered cells into the cycle. TGF{alpha}+ cells have a high probability of being outside G0 and being targets for G1 arrest by PGA2/J2. Expression of TGF{alpha} is unaffected by PGA2/J2. This suggests that in the growth regulatory pathway PGA2 and PGJ2 act downstream of TGF{alpha}. Preneoplastic hepatocytes overexpressing TGF{alpha} do not require additional DNA synthesis induction by PGE2/F2{alpha}, but they are targets for suppression of DNA synthesis by PGA2/J2.

 
Growth suppression of preneoplastic liver cells by SC236
The Cox-2 inhibitor SC236 preferentially suppressed DNA synthesis in preneoplastic hepatocytes (Figure 8Go). Since this effect was abrogated by co-treatment with PGE2 or PGF2{alpha}, the growth inhibitory effect of SC236 may be caused by depletion of growth stimulatory endogenous PGs (Figure 9Go). This suggests that a basal intracellular level of PGs is necessary for a cell to undergo DNA synthesis and that therefore proliferating hepatocytes, i.e. mainly GSTp+ cells, are preferentially affected by depletion of PGs. This assumption is also substantiated by findings with the Cox inhibitor acetylsalicylic acid, which reduced the number and size of GSTp+ liver preneoplasias in rats treated with various tumour-promoting compounds (30–34). However, it cannot be excluded that other unknown mechanisms contribute to the antagonistic effects of PGE2/F2{alpha} and SC236 on replicative DNA synthesis.

Unlike SC236, the Cox inhibitor indomethacin failed to suppress DNA synthesis in cultured hepatocytes. These findings may be due to different effective concentrations of the two compounds, e.g. 5 µM indomethacin but only 0.005 µM SC236 are sufficient to halve the activity of purified Cox-2, while 0.1 µM indomethacin and 17 µM SC236 are required for a similar inhibition of Cox-1 (56; product information provided by Searle). Thus traces of SC236 may be sufficient and the effects of SC236 on preneoplastic hepatocytes in the present study may be attributed to a reduced activity of Cox-2. Indomethacin generally requires higher concentrations for inhibition of Cox-1 and Cox-2 and appears to abrogate PG biosynthesis only when pre-stimulated by exogenous TGF{alpha} or hepatocyte growth factor (35,36,57). Furthermore, other effects of SC236 may contribute to DNA synthesis inhibition in preneoplastic hepatocytes; the structurally diverse non-steroidal anti-inflammatory drugs differ considerably in their side-effects, as has been shown by extremely variable inhibition of the activity of GSH transferases and I{kappa}B kinaseß by this class of compounds (58,59).

Implications for human liver cancer
Rodent and human hepatocellular carcinomas are rich in TGF{alpha} (Figure 2Go; 38,40). Anticipating analogies with preneoplastic GSTp+ cells we can put forward the following hypotheses. (i) Human liver tumours may require a basal level of PGs to proliferate; reduced PG levels due to Cox inhibition may affect growth of the tumour considerably. This was also confirmed by an anti-proliferative effect of the Cox-2 inhibitor sulindac on human hepatocellular carcinoma cell lines (60). (ii) Tumour cells expressing TGF{alpha} may not require elevated activities of Cox enzymes, leading to high intracellular concentrations of both PGE2/F2{alpha} and PGA2/J2. PGE2/F2{alpha} may not stimulate additional growth of tumours with already accelerated cell replication, rather, the tumour cells would be in danger of growth inhibition by PGA2/J2. These assumptions are consistent with the observation that liver tumours do not overexpress the two Cox isoenzymes and even tend to exhibit decreased expression, as shown in the present and in previous studies on rats and humans (61,62). It remains to be elucidated whether tumours with elevated levels of PGE2 and PGF2{alpha} are poor in TGF{alpha}.

This study demonstrates differential effects of TGF{alpha}, PGs and Cox inhibitors on normal and preneoplastic hepatocytes. For tumour prevention and therapy it seems promising to focus future studies on these differences.


   Notes
 
1 To whom correspondence should be addressedEmail: bettina.grasl-kraupp{at}univie.ac.at Back


   Acknowledgments
 
This study was supported by the Herzfeldersche Familienstiftung.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Okuda,K., Kojiro,M. and Okuda,H. (1993) Neoplasms of the liver. In Schiff,L. and Schiff,E.R. (eds) Diseases of the Liver. Lippincott, Philadelphia, PA, pp. 1236–1296.
  2. Lupulescu,A. (1978) Enhancement of carcinogenesis by prostaglandins. Nature, 272, 634–636.[Medline]
  3. IARC (1997) IARC Handbooks of Cancer Prevention: Non-steroidal Anti-inflammatory Drugs, Vol. 1. IARC, Lyon.
  4. DuBois,R.N., Abramson,S.B., Crofford,L., Gupta,R.A., Simon,L.S., Van de Putte,L.B. and Lipsky,P.E. (1998) Cyclooxygenase in biology and disease. FASEB J., 12, 1063–1073.[Abstract/Free Full Text]
  5. Enomoto,N., Ikejima,K., Yamashina,S., Enomoto,A., Nishiura,T., Nishimura,T., Brenner,D.A., Schemmer,P., Bradford,B.U., Rivera,C.A., Zhong,Z. and Thurman,R.G. (2000) Kupffer cell-derived prostaglandin E2 is involved in alcohol-induced fat accumulation in rat liver. Am. J. Physiol., 279, G100–G106.[Abstract/Free Full Text]
  6. Ikeda,T., Tozuka,S., Hasumara,Y. and Takeuchi,J. (1988) Prostaglandin-E-producing hepatocellular carcinoma with hypercalcemia. Cancer, 61, 1813–1814.[ISI][Medline]
  7. Gupta,C., Banks,M. and Shinozuka,H. (1989) Elevated levels of prostaglandin E2 in the liver of rats fed a choline deficient diet: possible involvement in liver tumor promotion. Cancer Lett., 46, 129–135.[ISI][Medline]
  8. Thun,M.J., Namboodiri,M.M. and Heath,C.W.Jr (1991) Aspirin use and reduced risk of fatal colon cancer. N. Engl. J. Med., 325, 1593–1596.[Abstract]
  9. Elder,D.J.E. and Paraskeva,C. (1998) Cox-2 inhibitors for colorectal cancer. Nature Med., 4, 392–393.[ISI][Medline]
  10. Moore,M.A., Nakagawa,K., Satoh,K., Ishikawa,T. and Sato,K. (1987) Single GST-P positive liver cells—putative initiated hepatocytes. Carcinogenesis, 8, 483–486.[Abstract/Free Full Text]
  11. Dragan,Y.P., Peterson,J. and Pitot,H.C. (1993) Comparison of hepatocyte phenotypes at the glutathione transferase and albumin loci in Sprague–Dawley and Nagase analbumineic rats and F1 progeny after initiation and promotion. Carcinogenesis, 14, 1313–1319.[Abstract/Free Full Text]
  12. Stinchcombe,S., Buchmann,A., Bock,K.W. and Schwarz,M. (1995) Inhibition of apoptosis during 2,3,7,8-tetrachlorodibenzo-p-dioxin-mediated tumour promotion in rat liver. Carcinogenesis, 16, 1271–1275.[Abstract/Free Full Text]
  13. Grasl-Kraupp,B., Luebeck,G., Wagner,A., Löw-Baselli,A., de Gunst,M., Waldhör,T., Moolgavkar,S. and Schulte-Hermann,R. (2000) Quantitative analysis of tumor initiation in rat liver: role of cell replication and cell death. Carcinogenesis, 21, 1411–1421.[Abstract/Free Full Text]
  14. Sato,K. (1989) Glutathione transferases as markers for preneoplasia and neoplasia. In Vande Woude,G.F. and Klein,G. (eds) Advances in Cancer Research. Academic Press, San Diego, CA, pp. 205–255.
  15. Grasl-Kraupp,B., Ruttkay-Nedecky,B., Müllauer,L., Taper,H., Huber,W., Bursch,W. and Schulte-Hermann,R. (1997) Inherent increase of apoptosis in liver tumors: implications for carcinogenesis and tumor regression. Hepatology, 25, 906–911.[ISI][Medline]
  16. Schulte-Hermann,R., Timmermann-Trosiener,I., Barthel,G. and Bursch,W. (1990) DNA synthesis, apoptosis and phenotypic expression as determinants of growth of altered foci in rat liver during phenobarbital promotion. Cancer Res., 50, 5127–5135.[Abstract/Free Full Text]
  17. Schulte-Hermann,R., Bursch,W. and Grasl-Kraupp,B. (1995) Active cell death (apoptosis) in liver biology and disease. In Boyer,J.L. and Ockner,R.K. (eds) Progress in Liver Diseases. Saunders, Philadelphia, PA, pp. 1–35.
  18. Grasl-Kraupp,B., Bursch,W., Ruttkay-Nedecky,B., Wagner,A., Lauer,B. and Schulte-Hermann,R. (1994) Food restriction eliminates preneoplastic cells through apoptosis and antagonizes carcinogenesis in rat liver. Proc. Natl Acad. Sci. USA, 91, 9995–9999.[Abstract/Free Full Text]
  19. Michalopoulos,G.K. and DeFrances,M.C. (1997) Liver regeneration. Science, 276, 60–66.[Abstract/Free Full Text]
  20. Fausto,N. (2000) Liver regeneration. J. Hepatol., 32, 19–31.[ISI][Medline]
  21. Löw-Baselli,A., Hufnagl,K., Parzefall,W., Schulte-Hermann,R. and Grasl-Kraupp,B. (2000) Initiated rat hepatocytes in primary culture: a novel tool to study alterations in growth control during the first stage of carcinogenesis. Carcinogenesis, 21, 79–86.[Abstract/Free Full Text]
  22. Skouteris,G.G., Ord,M.G. and Stocken,L.A. (1988) Regulation of the proliferation of primary rat hepatocytes by eicosanoids. J. Cell. Physiol., 135, 516–520.[ISI][Medline]
  23. Refsnes,M., Thoresen,G.H., Dajani,O.F. and Christoffersen,T. (1994) Stimulation of hepatocyte DNA synthesis by prostaglandin E2 and prostaglandin F2{alpha}. Additivity with the effect of norepinephrine and synergism with epidermal growth factor. J. Cell. Physiol., 159, 35–40.[ISI][Medline]
  24. Khan,S.H. and Sorof,S. (1990) Preferential binding of growth inhibitory prostaglandins by the target protein of a carcinogen. Proc. Natl Acad. Sci. USA, 87, 9401–9405.[Abstract/Free Full Text]
  25. Hashimoto,N., Watanabe,T., Ykeda,Y., Yamada,A., Taniguchi,S., Mitsui,H. and Kurokawa,P. (1997) Prostaglandins induce proliferation of rat hepatocytes through a prostaglandin E2 receptor EP3 subtype. Am. J. Physiol., 272, G597–G604.[Abstract/Free Full Text]
  26. Meade,M.A., Smith,W.L. and DeWitt,D.L. (1993) Differential inhibition of prostaglandin endoperoxide synthase (cyclooxygenase) by aspirin and other non-steroidal anti-inflammatory drugs. J. Biol. Chem., 268, 6610–6614.[Abstract/Free Full Text]
  27. O'Neill,G.P., Mancini,J.A., Kargman,S., Yergey,J., Kwan,M.Y., Falgueyret,J.P., Abramovitz,M., Kennedy,B.P., Quellet,M., Cromlish,W., Culp,S., Evans,J.F., Ford-Hutchinson,A.W. and Vickers,P.J. (1994) Overexpression of human prostaglandin G/H synthase-1 and -2 by recombinant vaccinia virus: inhibition by nonsteroidal anti-inflammatory drugs and biosynthesis of 15-hydroxyeicosatetraenoic acid. Mol. Pharmacol., 45, 245–254.[Abstract]
  28. Gierse,J.K., McDonald,J.J., Hauser,S.D., Rangwala,S.H., Koboldt,C.M. and Seibert,K. (1996) A single amino acid difference between cyclooxygenase-1 (COX-1) and -2 (COX-2) reverses the selectivity of COX-2 specific inhibitors. J. Biol. Chem., 271, 15810–15814.[Abstract/Free Full Text]
  29. Bayer,B.M. and Beaven,M.A. (1979) Evidence that indomethacin reversibly inhibits cell growth in the G1 phase of the cell cycle. Biochem. Pharmacol., 28, 441–443.[ISI][Medline]
  30. Denda,A., Tang,Q., Endoh,T., Tsujiuchi,T., Horiguchi,K., Noguchi,O., Mizumoto,Y., Nakae,D. and Konishi,Y. (1994) Prevention by acetylsalicylic acid of liver cirrhosis and carcinogenesis as well as generation of 8-hydroxydeoxyguanosine and thiobarbituric acid-reactive substances caused by a choline-deficient, L-amino acid-defined diet in rats. Carcinogenesis, 15, 1279–1283.[Abstract/Free Full Text]
  31. Endoh,T., Tang,Q., Denda,A., Noguchi,O., Kobayashi,E., Tamura,K., Horiguchi,K., Ogasawara,H., Tsujiuchi,T., Nakae,D., Sugimura,M. and Konishi,Y. (1996) Inhibition by acetylsalicylic acid, a cyclooxygenase inhibitor, and p-bromophenacylbromide, a phospholipase A2 inhibitor, of both cirrhosis and enzyme-altered nodules caused by a choline-deficient, L-amino-acid-defined diet in rats. Carcinogenesis, 17, 467–475.[Abstract/Free Full Text]
  32. Denda,A., Ura,H., Tsujiuchi,T., Tsutsumi,M., Eimoto,H., Takashima,Y., Kitazawa,S., Kinugasa,T. and Konishi,Y. (1989) Possible involvement of arachidonic acid metabolism in phenobarbital promotion of hepatocarcinogenesis. Carcinogenesis, 10, 1929–1935.[Abstract/Free Full Text]
  33. Tang,Q., Denda,A., Tsujiuchi,T., Tsutsumi,M., Toshihiro,A., Murata,Y., Maruyama,H. and Konishi,Y. (1993) Inhibitory effects of inhibitors of arachidonic acid metabolism on the evolution of rat liver preneoplastic foci into nodules and hepatocellular carcinomas with or without phenobarbital exposure. Jpn. J. Cancer Res., 84, 120–127.[ISI][Medline]
  34. Tanaka,T., Kojima,T., Okumura,A., Sugie,S. and Mori,H. (1993) Inhibitory effect of the non-steroidal anti-inflammatory drugs, indomethacin and piroxicam, on 2-acetylaminofluorene-induced hepatocarcinogenesis in male ACI/N rats. Cancer Lett., 68, 111–118.[ISI][Medline]
  35. Skouteris,G.G. and McMenamin,M. (1992) Transforming growth factor-{alpha}-induced DNA synthesis and c-myc expression in primary rat hepatocyte cultures is modulated by indomethacin. Biochem. J., 281, 729–733.
  36. Adachi,T., Nakashima,S., Saji,S., Nakamura,T. and Nozawa,Y. (1995) Roles of prostaglandin production and mitogen-activated protein kinase activation in hepatocyte growth factor-mediated rat hepatocyte proliferation. Hepatology, 21, 1668–1674.[ISI][Medline]
  37. Kumar,V., Bustin,S.A. and McKay,I.A. (1995) Transforming growth factor alpha. Cell Biol. Int., 19, 373–388.[ISI][Medline]
  38. Yeh,Y.C., Tsai,J.F., Chuang,L.Y., Yeh,H.W., Tsai,J.H., Florine,D.L. and Tam,J.P. (1987) Elevation of transforming growth factor alpha and its relationship to the epidermal growth factor and alpha-fetoprotein levels in patients with hepatocellular carcinoma. Cancer Res., 47, 896–901.[Abstract/Free Full Text]
  39. Kaufmann,W.K., Zhang,Y. and Kaufmann,D.G. (1992) Association between expression of transforming growth factor-alpha and progression of hepatocellular foci to neoplasms. Carcinogenesis, 13, 1481–1483.[Abstract/Free Full Text]
  40. Schaff,Z., Hsia,C.C., Sarosi,I. and Tabor,E. (1994) Overexpression of transforming growth factor-alpha in hepatocellular carcinoma and focal nodular hyperplasia from European patients. Hum. Pathol., 25, 644–651.[ISI][Medline]
  41. Miller,R.T., Cattley,R.C., Marsman,D.S., Lyght,O. and Popp,J.A. (1995) TGF{alpha} is differently expressed in liver foci induced by diethylnitrosamine initiation and peroxisome proliferator promotion. Carcinogenesis, 16, 77–82.[Abstract/Free Full Text]
  42. Lee,G.H., Merlino,G. and Fausto,N. (1992) Development of liver tumors in transforming growth factor {alpha} transgenic mice. Cancer Res., 52, 5162–5170.[Abstract/Free Full Text]
  43. Sandgren,E.P., Luetteke,N.C., Qiu,T.H., Palmiter,R.D., Brinster,R.L. and Lee,D.C. (1993) Transforming growth factor alpha dramatically enhances oncogene-induced carcinogenesis in transgenic mouse pancreas and liver. Mol. Cell. Biol., 13, 320–330.[Abstract/Free Full Text]
  44. Santoni–Rugiu,E., Nagy,P., Jensen,M.R., Factor,V.M. and Thorgeirsson,S.S. (1996) Evolution of neoplastic development in the liver of transgenic mice co-expressing c-myc and transforming growth factor {alpha}. Am. J. Pathol., 149, 407–428.[Abstract]
  45. Schulte-Hermann,R., Bursch,W., Löw-Baselli,A., Wagner,A. and Grasl-Kraupp,B. (1997) Apoptosis in the liver and its role in hepatocarcinogenesis. Cell. Biol. Toxicol., 13, 339–348.[ISI][Medline]
  46. Little,P., Skouteris,G.G., Ord,M.G., Stocken,L.A. (1988) Serum from partially hepatectomized rats induces primary hepatocytes to enter S phase: a role for prostaglandins? J. Cell Sci., 91, 549–553.[Abstract/Free Full Text]
  47. Tsujii,H., Okamoto,Y., Kikuchi,E., Matsumoto,M. and Nakano,H. (1993) Prostaglandin E2 and rat liver regeneration. Gastroenterology, 105, 495–499.[ISI][Medline]
  48. Seglen,O.P. (1976) Preparation of isolated rat liver cells. In Prescott,D.M. (ed.) Methods in Cell Biology. Academic Press, New York, NY, pp. 29–81.
  49. Parzefall,W., Monschau,P. and Schulte-Hermann,R. (1989) Induction by cyproterone acetate of DNA synthesis and mitosis in primary cultures of adult rat hepatocytes in serum free medium. Arch. Toxicol.,