Interferon α-induced apoptosis on rat preneoplastic liver is mediated by hepatocytic transforming growth factor β1

Authors

  • María de Luján Alvarez,

    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    Search for more papers by this author
  • María Teresa Ronco,

    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    Search for more papers by this author
  • J. Elena Ochoa,

    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    Search for more papers by this author
  • Juan A. Monti,

    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    Search for more papers by this author
  • Cristina E. Carnovale,

    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    Search for more papers by this author
  • Gerardo B. Pisani,

    1. Area de Morfología, Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, Rosario, Argentina
    Search for more papers by this author
  • María Cristina Lugano,

    1. Area de Morfología, Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, Rosario, Argentina
    Search for more papers by this author
  • María Cristina Carrillo

    Corresponding author
    1. Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
    • Instituto de Fisiología Experimental (IFISE), Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, Suipacha 570, 2000-Rosario, Argentina
    Search for more papers by this author
    • fax: +54 341-4399473


Abstract

In previous work we showed that interferon alfa-2b (IFN-α2b) increases apoptosis on rat hepatic preneoplastic foci. The aim of this study was to determine if transforming growth factor β1 (TGF-β1) was involved in the programmed cell death on the foci. Animals were divided into 6 groups: subjected to a 2-phase model (diethylnitrosamine plus 2-acetylaminofluorene) of preneoplasia development (group 1); treated with IFN-α2b during the 2 phases (group 2); treated with IFN-α2b during initiation with diethylnitrosamine (group 3); treated with IFN-α2b during 2-acetylaminofluorene administration (group 4); subjected only to an initiation stage (group 5); and treated with IFN-α2b during the initiation period (group 6). Serum TGF-β1 levels were increased in IFN-α2b–treated rats. Immunohistochemical studies showed that IFN-α2b significantly increased the quantity of TGF-β1–positive hepatocytes in groups 2 to 4. Phosphorylated-Smads-2/3 (p-Smads-2/3) proteins in liver nuclear extracts were significantly elevated. To determine the source of TGF-β1, isolated hepatocytes, Kupffer cells, and peritoneal macrophages from animals in groups 1 and 5 were cultured with or without IFN-α2b. IFN-α2b stimulus induced several-fold increases of TGF-β1 secretion from hepatocytes. Neither Kupffer cells nor peritoneal macrophages secreted detectable TGF-β1 levels when they were treated with IFN-α2b. IFN-α2b–stimulated cultured hepatocytes from preneoplastic livers showed enhanced apoptosis, measured by fluorescence microscopy and caspase-3 activity. They presented higher nuclear accumulation of p-Smads-2/3, indicating increased TGF-β1 signaling. When anti–TGF-β1 was added to the culture media, TGF-β1 activation and apoptosis induced by IFN-α2b were blocked. In conclusion, IFN-α2b–induced production of TGF-β1 by hepatocytes from preneoplastic liver is involved in the apoptotic elimination of altered hepatic foci. (HEPATOLOGY 2004;40:394–402.)

Induction of apoptosis through immune defense mechanisms or other homeostatic controls is an important process for the elimination of nascent tumor cells.1 The elucidation of mechanisms of apoptosis is of great importance, not only for understanding oncogenesis, but for developing new cancer therapy also. Rat liver provides a useful model to study physiological growth regulation and deregulated growth in carcinogenesis.2, 3

In a previous work,4 we showed that administration of interferon alfa-2b (IFN-α2b) decreased both number and volume percentage of the placental form of rat glutathione S-transferase (rGST-P)-positive preneoplastic foci through a mechanism of programmed cell death, or apoptosis. In this connection, we examined whether p53, Bax, Bcl-2 and Bcl-xL were involved in IFN-α2b–mediated apoptosis. Proapoptotic protein Bax levels were found increased in IFN-α2b–treated animals, in parallel with increases in p53. In addition, there were decreases of the antiapoptotic proteins Bcl-2 and Bcl-xL. We concluded that high apoptosis in preneoplastic IFN-α2b–treated hepatocytes, in vivo, appeared to largely depend on dysregulation of the Bcl-2 protein family and p53 overexpression.4

Transforming growth factor β1 (TGF-β1) belongs to a family of multifunctional cytokines that regulate a variety of biological responses such as proliferation, differentiation, apoptosis, and development.5 TGF-β1 binds to a heterometric complex of type I and type II receptors. Binding of the ligand to the type II receptor results in the recruitment and activation of the type I receptor that phosphorylates Smads-2 and -3. Once phosphorylated, Smads-2/3 form a complex with Smad-4 and translocate to the nucleus, where they positively or negatively regulate transcription of TGF-β target genes through cooperative interactions with DNA, transcription factors, coactivators, and corepresors.6, 7

TGF-β1 is an important physiological mediator for apoptosis in both normal and neoplastic liver.8 There are several lines of evidence: TGF-β11 induces apoptosis in primary hepatocyte cultures derived from both adult and fetal rat liver9–11; intravenous administration of TGF-β1 induces apoptosis in both normal and regressing liver9, 12 and also in hepatocytes from preneoplastic foci.13 In addition, several hepatoma cell lines are sensitive toward programmed cell death induction by TGF-β1.14, 15 Although the mechanisms by which TGF-β1 induces apoptosis in hepatocytes have not been fully elucidated, it is well established that production of reactive oxygen intermediates and activation of caspase-family proteases are involved.16 It has been shown that TGF-β1 activates caspase-3 and -7 processing in rat adult hepatocytes.17 Studies with hepatoma cell lines have confirmed that at least caspases -2, -3, -8 and -9 could be activated in TGF-β1–induced apoptosis14 and that caspase activation is mediated by the release of cytochrome c.18 TGF-β1 mediates radical oxygen species production that contributes to the down-regulation of Bcl-xL, the release of cytochrome c, and caspase activation.11 However, the involvement of Bcl-2, Bax, and Bcl-xL proteins in TGF-β1–induced apoptosis has not been completely established. TGF-β1 decreased the antiapoptotic protein Bcl-xL in diverse hepatoma cell lines,14, 15 whereas in other hepatoma cells no changes in Bax or Bcl-xL were observed.18 On the other hand, overexpression of Bcl-2 blocked induction of apoptosis by TGF-β1 in human hepatoma cells.19

In the current study, our first aim was to determine if the soluble mediator TGF-β1 was responsible for the observed IFN-α2b-induced apoptosis of preneoplastic hepatocytes.

Macrophages (Mϕs) play an important role by producing soluble cytokines that can modulate cell function. Well-established sources of TGF-β1 are Kupffer cells and peritoneal Mϕs. Hepatocyte apoptosis in severe acute pancreatitis occurs via TGF-β1 derived from peritoneal Mϕs.20 Kupffer cell-secreted TGF-β1 plays a pivotal role in the pathogenesis of alcoholic and fibrotic liver diseases.21 In spite of earlier concepts, evidence suggests that not only mesenchymal cells but also hepatocytes are able to produce and secrete TGF-β1.22

Thus, the second aim of this study was to clarify the source of TGF-β1 in IFN-α2b–treated rats.

Abbreviations

IFN-α2b, interferon alfa-2b; rGST-P, placental form of rat glutathione S-transferase; TGF-β1, transforming growth factor β1; Mϕs, macrophages; DEN, diethylnitrosamine; 2-AAF, 2-acetylaminofluorene; Ac-DEVD-pNA, N-acetyl-Asp-Glu-Val-Asp-p-nitroanilide; ELISA, enzyme-linked immunoassay; p-Smads-2/3, phosphorylated Smads-2/3; IFN, interferon.

Materials and Methods

Chemicals

Diethylnitrosamine (DEN), 2-acetylaminofluorene (2-AAF), and collagenase were obtained from Sigma Chemical Co. (St Louis, MO). IFN-α2b was kindly provided by BioSidus Laboratory (Buenos Aires, Argentina). Anti–p-Smad-2/3, anti–TGF-β1, anti–caspase-3, and secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti–rGST-P antibody was provided by Immunotech (Marseilles, France). The caspase-3 substrate N-acetyl-Asp-Glu-Val-Asp-p-nitroanilide (Ac-DEVD-pNA) was purchased from Alexis Co. (San Diego, CA). All other chemicals were of the highest grade commercially available.

Animals and Treatment

Adult male Wistar rats (330-380 g) were maintained in a room at constant temperature with a 12-hour light-dark cycle and free access to food (FormulabDiet 5008, Brentwood, MO) and water. All the experimental protocols were performed according to the NIH “Guide for the Care and Use of Laboratory Animals” (Publication no. 25-28, revised 1996). The animals were divided into 6 groups of 8 to 10 rats each.

Animals in groups 1 to 4 were subjected to a 2-phase model of rat hepatocarcinogenesis, as previously described.4 The initiation stage was induced by the administration of 2 intraperitoneal necrogenic doses of DEN (150 mg/kg body weight) 2 weeks apart. One week after the last injection of DEN, the rats received 20 mg/kg body weight of 2-AAF by gavage for 4 consecutive days per week for 3 weeks. Groups 2 to 4 also received IFN-α2b, 6.5 × 105 U/kg body weight, administered intraperitoneally three times per week. The dose used was comparable to that used for therapeutic purposes.4 Group 2 received IFN-α2b during the entire treatment; group 3 received IFN-α2b only during the initiation phase with DEN and group 4 received it only during the 2-AAF phase. Animals of groups 1 to 4 were killed at the end of the sixth week. Groups 5 and 6 underwent only the DEN stage. Group 6 also received IFN-α2b, three times per week. Both groups of rats were killed at the end of the third week.

All the animals were killed between 10 AM and 11 AM by cardiac puncture after administration of pentobarbital anesthesia (50 mg/kg body weight).

In Vivo Studies

TGF-β1 Assay.

Serum TGF-β1 levels were determined with an enzyme-linked immunoassay (ELISA) kit for human TGF-β1 (R&D Systems, Minneapolis, MN) that cross-reacts with the rat protein.

Immunohistochemical Study for TGF-β1.

Serially sectioned liver slices were examined by immunohistochemical staining with anti–rGST-P and anti–TGF-β1 by the peroxidase-antiperoxidase method.23 At least 30 random fields per tissue section were evaluated and scored. TGF-β1–positive cells were expressed per 1,000 hepatocytes.

Preparation of Nuclear Extracts and Western Blot of Phosphorylated–Smads-2/3 (p-Smads-2/3) proteins.

Liver tissue homogenates were prepared by homogenization in 300 mmol/L sucrose with protease inhibitors. Nuclei were sedimented by centrifugation at 1,000g.24 Pellets were washed and resuspended in radioimmunoprecipitation assay (RIPA) buffer containing 20 mmol phosphate-buffered solution (pH 8), 1% Triton, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 5 mmol EDTA, 200 mmol NaCl, and protease inhibitors. Pellets were incubated on ice for 1 hour, and centrifuged (8,000g, 15 minutes, 4°C). Proteins were quantified according to Lowry et al.25 Extracts were subjected to 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinyl difluoride membranes (PerkinElmer Life Sciences, Boston, MA). After blocking, blots were incubated overnight at 4°C with anti–p-Smad-2/3 antibody (1:500). Finally, they were incubated with peroxidase-conjugated secondary antibody (1:5,000), and bands were detected by enhanced chemiluminescence detection system (Amersham Pharmacia Biotech, Little Chalfont, Buckinghamshire, UK). Autoradiographs were obtained by exposing the membranes to Kodak XAR film (Rochester, NY), and the bands were quantitated by densitometry (Shimadzu CS-9000, Tokyo, Japan).

In Vitro Studies

Culture and IFN-α2b Treatment of Peritoneal Mϕs, Kupffer Cells, and Hepatocytes.

Animals of groups 1 and 5 were used in the in vitro studies. Peritoneal Mϕs were obtained aseptically from the peritoneal cavities. Cells were centrifuged at 500g for 15 minutes, resuspended in culture medium, and seeded. By nonspecific esterase staining, at least 85% of the cells in the peritoneal exudates appeared to be Mϕs. Hepatocytes and Kupffer cells were isolated by collagenase perfusion and mechanical disruption as described previously.26 A hepatocyte-enriched population from the entire liver containing both preneoplastic hepatocytes and hepatocytes from the surrounding nonpreneoplastic liver tissue was obtained by centrifugation twice at 40g for 2 minutes; centrifugation of the supernatants at 200g yielded an enriched population of adherent cells.27

Cells were seeded at similar densities and cultured at 37°C in a humidified atmosphere of 5% CO2 in RPMI medium with L-glutamine supplemented with 10% fetal calf serum, penicillin (100 U/mL) and streptomycin (100 μg/mL). Two hours later, cells were treated with fresh medium containing 105 U/mL IFN-α2b or left untreated.

The viability of all cell suspensions used, checked by trypan blue exclusion, was between 85% and 95%.

TGF-β1 Assay.

After 24 hours of culture, the hepatocyte-, peritoneal Mϕ-, and Kupffer cell-conditioned media were collected and TGF-β1 was quantitated using an ELISA.

Neutralizing Antibody Treatment.

Isolated hepatocytes (10 × 106/dish) from groups 1 and 5 were cultured for 24 hours with 105 U/mL IFN-α2b and 0.04 μg/mL anti–TGF-β1 at the same time.

Preparation of Cytosolic and Nuclear Extracts.

Hepatocytes were scraped off and centrifuged, resuspended in 300 mmol/L sucrose, disrupted by sonication, and incubated at 4°C for 15 minutes. The nuclei were sedimented and nuclear extracts were obtained as previously described.24 The postnuclear homogenate was further centrifuged at 45,000g for 1 hour and the obtained supernatant was used as the cytosolic extract.

Microscopic Determination of Apoptosis.

The relative numbers of apoptotic and necrotic cells after 24 hours of culture were determined by fluorescence microscopy after costaining with acridine orange and ethidium bromide.28, 29 Cells were considered apoptotic if they were stained with acridine orange and had nuclear and cytoplasmic condensation, nuclear fragmentation, membrane blebbing, and apoptotic body formation. The percentage of apoptotic hepatocytes was determined by examining more than 400 cells per dish. Necrosis was determined by the presence of ethidium bromide staining, which is only taken by nonviable necrotic cells, staining them red.

Analysis of Caspase-3 Activity.

Caspase-3 activity was evaluated by measuring the proteolytic cleavage of chromogenic substrate for caspase-3–like proteases Ac-DEVD-pNA, as described previously.30

Statistical Analysis.

Results were expressed as mean ± SE. Significance in differences was tested by 1-way ANOVA, followed by Tukey test. Differences were considered significant when the P value was less than .05.

Results

In Vivo Studies Analysis

Serum TGF-β1 Levels.

Changes in serum TGF-β1 levels are shown in Fig. 1. Serum TGF-β1 levels in the rats subjected to the 2-phase protocol and treated with IFN-α2b (groups 2 to 4) were increased compared with the values for control group 1 (P < .05). The animals from group 6 also showed increased serum TGF-β1 levels with respect to group 5 (P < .05).

Figure 1.

Serum TGF-β1 levels (ng/mL) of rats from the six experimental groups subjected to the experimental protocol. Values are expressed as mean ± SE. Serum TGF-β1 levels were significantly increased in IFN-α2b–treated animals compared with control group. *P < .05 versus control group 1. **P < .05 versus control group 5.

TGF-β1 Expression in Preneoplastic Liver.

To investigate whether the enzyme-altered foci or the surrounding tissue expressed TGF-β1 protein during DEN/2-AAF-induced carcinogenesis, serial sections of liver tissue were immunostained with anti–rGST-P and anti–TGF-β1 antibodies. Immunological signals for TGF-β1 were not found along the space of Disse. Interestingly, hepatocytes from the foci (rGST-P–positive) and hepatocytes from the surrounding tissue (rGST-P–negative) were strongly positive with anti–TGF-β1 antibody (Figs. 2A and B). These results differed from other reported,31 in which the areas of hepatocytes showing positive staining with the rGST-P and TGF-β1 antibodies were almost identical. Different experimental models of hepatocarcinogenesis used, however, may explain such discordance.

Figure 2.

Liver slices showing rGST-P–positive preneoplastic foci and TGF-β1–positive hepatocytes. A and C show rGST-P-positive enzyme-altered foci (magnification ×200). Liver tissue was obtained from rats subjected to (A) the 2-phase protocol and (C) 2-phase protocol plus IFN-α2b. (B and D) Serial sections of A and C respectively, showing TGF-β1-positive hepatocytes inside the foci (thick arrows) and in the surrounding tissue (thin arrows).

The same diffuse pattern of TGF-β1 intracellular expression was observed in liver tissues from rats that received IFN-α2b (Figs. 2C and D), though the quantity of TGF-β1–positive hepatocytes was significantly increased in IFN-α2b–treated rats, as shown in Fig. 3. IFN-α2b treatment on groups 2 to 4 increased the number of TGF-β1–positive hepatocytes 2- to 2.5-fold above that seen in group 1.

Figure 3.

Effect of IFN-α2b on the number of TGF-β1–positive hepatocytes. The results were expressed as positive cells scored per 1,000 hepatocytes. Data represent mean ± SE. IFN-α2b treatment on groups 2 to 4 significantly increased the quantity of TGF-β1–positive hepatocytes. *P < .05 versus group 1.

Because there were single initiated hepatocytes and very small foci in groups 5 and 6, it was not possible to determine if the TGF-β1–positive hepatocytes were inside/outside the foci.

Normal, untreated rat liver did not show positive immunohistochemical staining for TGF-β1 (data not shown).

Nuclear Content of pSmads-2/3 in Liver Tissue.

To demonstrate that the increase in TGF-β1 production corresponded with an increase in TGF-β1 signaling in hepatocytes, the nuclear content of p-Smads-2/3 was analyzed. Figure 4 represents changes in Smad phosphorylation and nuclear translocation. Western blot analysis showed that administration of IFN-α2b in animals subjected to the 2-phase protocol increased the nuclear content of p-Smads-2/3 (90% in group 2, 65% in group 3, and 82% in group 4 with respect to group 1; P < .05). In agreement with these results, immunohistochemical detection of hepatocytes showing p-Smads-2/3 nuclear staining was observed in liver tissues from animals of groups 2 to 4 (data not shown). However, in the animals that underwent only the initiation phase with DEN, IFN-α2b treatment did not increase the nuclear content of p-Smads-2/3 (group 6 vs. group 5).

Figure 4.

Western blot analysis of p-Smads-2/3 expression in nuclear extracts of liver tissue. Densitometric analysis was performed in at least 3 animals in each group. Data are expressed in percent values as mean ± SE, with control groups 1 and 5 considered 100%. Groups 2 to 4 showed increased p-Smads-2/3 proteins levels compared with control group (*P < .05 vs. group 1).

In Vitro Studies Analysis

Secretion of TGF-β1 to the Culture Media.

Peritoneal Mϕs, Kupffer cells, and hepatocytes from groups 1 and 5 were cultured in the presence of IFN-α2b. Neither peritoneal Mϕs nor Kupffer cells secreted detectable levels of TGF-β1 when they were stimulated with IFN-α2b. However, as can be seen in Fig. 5, hepatocytes produced and secreted TGF-β1 when they were stimulated with IFN-α2b. IFN-α2b presence in the culture media of hepatocytes from group 1 induced a 13-fold increase of TGF-β1 secretion, whereas TGF-β1 concentration in the culture media of hepatocytes from group 5 showed approximately 5.5-fold higher levels when cells were exposed to IFN-α2b. Because we used hepatocytes from the whole preneoplastic liver, we could not discriminate the TGF-β1 secreted by hepatocytes from the foci or from the surrounding tissue. Nevertheless, the higher secretion of TGF-β1 in group 1 was coincident with the higher proportion of liver occupied by preneoplastic foci in this group.4

Figure 5.

Release of TGF-β1 (ng/mL) from cultured hepatocytes without (−) and with (+) IFN-α2b addition. Data are expressed as mean ± SE. TGF-β1 secretion by hepatocytes was significantly increased by IFN-α2b addition in groups 1 and 5. *P < .05 versus unstimulated hepatocytes from each corresponding group.

The results of immunocytochemical studies were consistent with the prior observations that cultured hepatocytes were positive when they were immunostained with anti–TGF-β1, whereas peritoneal Mϕs and Kupffer cells were completely negative (data not shown).

Nuclear Content of p-Smads-2/3 in Cultured Hepatocytes.

To evaluate if increased TGF-β1 secretion correlates with increased TGF-β1 signaling in hepatocytes, nuclear content of p-Smads-2/3 was analyzed. Fig. 6 shows a significant increase in nuclear p-Smads-2/3 of IFN-α2b–stimulated hepatocytes from groups 1 and 5 (60% in group 1, 25% in group 5 vs. unstimulated cells; P < .05). These results indicate that TGF-β1 activation in IFN-α2b–treated hepatocytes has increased. When anti–TGF-β1 was added to the culture media, the nuclear translocation of p-Smads-2/3 induced by IFN-α2b was completely blocked.

Figure 6.

Western blot analysis of p-Smads-2/3 nuclear content in cultured hepatocytes from experimental groups 1 and 5. Densitometric analysis was performed in at least 3 animals in each group. Data are expressed in percent values as mean ± SE. The values of unstimulated hepatocytes (IFN-α2b[−]/anti–TGF-β1[−]) were arbitrarily considered 100%. IFN-α2b treatment increased p-Smads-2/3 protein levels compared with untreated cells. The addition of neutralizing anti-TGF-β1 blocked the increase in TGF-β1 signaling. *P < .05 versus unstimulated hepatocytes from each corresponding group.

Morphological Determination of Apoptosis in Cultured Hepatocytes.

The percentage of apoptotic hepatocytes in IFN-α2b–treated cultures from groups 1 and 5 increased 2-fold over unstimulated cultures (Fig. 7), indicating that IFN-α2b stimulus induced programmed cell death in hepatocytes. Furthermore, experiments were performed with anti–TGF-β1 to demonstrate if the apoptotic effect observed in IFN-α2b–treated cells is mediated by the production of TGF-β1 from hepatocytes themselves. If the secreted TGF-β1 functions as a positive regulator of apoptosis in these cells from preneoplastic liver, antibody-induced blockage of the endogenous TGF-β1 may indirectly inhibit apoptosis. As shown in Fig. 7, the addition of IFN-α2b and anti–TGF-β1 to cultured hepatocytes caused percentages of apoptotic hepatocytes similar to those in unstimulated cultures.

Figure 7.

Percentage of apoptotic cells in cultured hepatocytes. Hepatocytes from preneoplastic liver of groups 1 and 5 were untreated (−) or treated (+) with IFN-α2b alone, and IFN-α2b combined with anti–TGF-β1 (IFN-α2b[+]/anti–TGF-β1[+]). The percentage of apoptotic cells was determined by fluorescent microscopy as described in Materials and Methods. Results are from 4 independent experiments with duplicate dishes for each point. IFN-α2b augmented the percentage of apoptotic cells. The addition of anti–TGF-β1 blocked the effect of IFN-α2b. *P < .05 versus unstimulated hepatocytes from each corresponding group.

Finally, there was no increase in necrotic cells in either experimental group. The percentage of necrotic cells remained below 1% in control and treated cells.

Caspase-3 Activation in Cultured Hepatocytes.

IFN-α2b–treated hepatocytes showed a significant increase in caspase-3–like activity (30% and 40% in cultured hepatocytes from groups 1 and 5, respectively, compared with untreated cells). When anti–TGF-β1 was added to the culture media, no increase in caspase-3 activity could be detected (Fig. 8). These results were substantiated by measurements of procaspase-3 levels by Western blot. Accordingly, caspase-3 activation was demonstrated by the decrease in procaspase-3 levels after IFN-α2b treatment (Fig. 8).

Figure 8.

Caspase-3–like activity in cultured hepatocytes of groups 1 and 5. Cells were untreated (−) or treated only with IFN-α2b, and IFN-α2b combined with anti–TGF-β1 (IFN-α2b +/anti–TGF-β1 +). Caspase-3–like activity was determined 24 hours after treatment. Data are expressed in percent values as mean ± SE, with caspase-3 activity in control groups 1 and 5 considered 100%. Results are from 3 independent experiments with duplicate dishes for each point. IFN-α2b–treated hepatocytes showed increased caspase-3 activity. The addition of anti–TGF-β1 blocked the IFN-α2b-induced increase in the protease activity. Western blot expression of procaspase-3 was studied in all experimental groups. *P < .05 versus each corresponding control group.

Discussion

The growth of initiated hepatocytes leads to the development of liver foci and persistent nodules. These lesions exhibit a high rate of cell proliferation and cell death by apoptosis, with prevalence of cell production over cell loss; this allows their slow progression to more malignant stages.2, 32 Preneoplastic liver cells are highly susceptible to some apoptosis-inducing treatments, which can inhibit their evolution to more malignant stages.32

Given its antiproliferative, proapoptotic role in the liver, TGF-β1 could be expected to act as a tumor suppressor. However, various types of neoplastic liver cells respond quite differently to TGF-β1. Whereas some human and rat hepatoma cell lines are sensitive to TGF-β1,14, 15, 33 resistance has been reported for other hepatoma cells.33, 34 In addition, TGF-β1 overexpression seems to be a hallmark of human liver cancer.35 Thus, the relationship between TGF-β1 and cancer is complex: TGF-β1 may stimulate malignant progression itself; conversely, it can have tumor suppressor activity.36 The escape of certain hepatoma cells from TGF-β1–induced apoptosis seems to be an important and essential step in malignant progression.37, 38 Moreover, it has been suggested that TGF-β1 overexpression is a rather late event in human hepatocarcinogenesis.35 These data indicate that loss of TGF-β1 responsiveness is not an initiating or strongly predisposing event, but rather a late event in carcinogenesis.36, 39 Therefore, it is of interest to study if liver preneoplasia at an early stage is still sensitive toward TGF-β1 actions.

In a previous study,4 we showed that IFN-α2b administration decreased both number and volume of preneoplastic foci. These reductions were explained by a greater programmed cell death in the foci than in surrounding hepatocytes. Because the changes of pro- and antiapoptotic proteins induced by IFN-α2b were similar to those attributed to TGF-β1,14, 15, 19 we studied the possibility that TGF-β1 could be involved in the programmed cell death induced by IFN-α2b. Thus, the present work determines for the first time that endogenous TGF-β1 is implicated in the increased apoptosis on the foci of IFN-α2b–treated rats.

We observed that serum TGF-β1 levels in the animals treated with IFN-α2b were significantly increased. In accordance with this, immunohistochemical studies showed that IFN-α2b treatment significantly augmented the quantity of TGF-β1–positive hepatocytes in DEN/2-AAF–treated livers. At first sight, these findings seem to indicate that administration of IFN-α2b increases serum TGF-β1 production and the number of TGF-β1–positive hepatocytes. Although the mechanisms by which IFN-α2b treatment induced TGF-β1 in the preneoplastic livers were not completely explored, we observed, using Western blot analysis, that preneoplastic livers expressed higher levels of IFN-α receptors than control livers (data not shown). In addition, IFN-α2b administration in animals subjected to the preneoplastic protocol induced elevated levels of phosphorylated STAT1, indicating activation of the IFN-α pathway (data not shown).

Recent investigations have reported that the induction of apoptosis by endogenous TGF-β1 does not require an overall increase in its hepatic concentration.39 In view of the fact that TGF-β1 hepatic content may not reflect the induction of apoptosis by this cytokine, we determined the nuclear content of p-Smads-2/3. We observed high levels of p-Smads-2/3 proteins in the nuclear extracts of IFN-α2b–treated groups 2 to 4. These results corresponded with the increased number of TGF-β1–positive hepatocytes, indicating increased TGF-β1 activation in those experimental groups. However, an increase of 100% in p-Smads-2/3 reflects less than a 1.5-fold increase of TGF-β1. It has been shown that IFN-γ inhibits TGF-β–induced phosphorylation of Smads-2/3 through the induction of antagonistic Smad-7.40 Although the cross-talk between TGF-β and IFN-γ has been extensively described, there are no reported studies about cross-talk between TGF-β and IFN-α. Nevertheless, we cannot completely rule out the possibility that IFN-α2b could be raising inhibitory Smad-7 levels, a fact that could potentially mask a higher increase in p-Smads-2/3.

In animals subjected to an initiation stage, neither the percentage of TGF-β1–positive hepatocytes nor the accumulation of p-Smads-2/3 were increased after IFN-α2b treatment. Keeping in mind that the number of preneoplastic hepatocytes in the 2-stage model is 15-fold higher than in DEN treatment,4 we hypothesized that hepatocytes from the foci are more sensitive to IFN-α2b actions than hepatocytes from the surrounding tissue. Thus, it is likely that mainly preneoplastic hepatocytes produce TGF-β1 and have increased TGF-β1 signaling. Further work will be required to confirm this hypothesis. Specifically, hepatocytes from preneoplastic foci should be isolated and their TGF-β1 production and signaling compared with the values of the hepatocytes from the surrounding tissue. Nevertheless, in our model less than 3% of the liver is occupied by preneoplastic foci.4 This could yield a very low number of preneoplastic cells, since separation methods based on differences in metabolic and/or surface properties41, 42 only obtained subpopulations of preneoplastic hepatocytes due to heterogeneity in these properties.

Nonparenchymal cells, including Kupffer cells and peritoneal Mϕs, are the main source of hepatic TGF-β1.20, 21 Hepatocytes, however, may synthesize TGF-β1in vitro22 as well as during hepatocarcinogenesis.35 In the present study, neither peritoneal Mϕs nor Kupffer cells from DEN/2-AAF– and DEN-treated rats secreted detectable levels of TGF-β1 when they were stimulated with IFN-α2b. Conversely, hepatocytes from normal, untreated liver did not secrete TGF-β1 in the absence or presence of IFN-α2b (data not shown). Nevertheless, hepatocytes from preneoplastic livers produced and secreted detectable levels of TGF-β1 when they were cultured without IFN-α2b stimulus, and IFN-α2b presence in the culture media induced several-fold increases of TGF-β1 production. Previous studies clearly showed that IFN-α treatment induced significant apoptosis in primary human hepatocytes, in contrast to primary rat and mouse hepatocytes, which responded poorly to IFN-α. The differential response to IFN-α stimulation may be caused by different expression patterns of receptors in human, rat, and mouse hepatocytes.43 When we studied IFN-α2b–stimulated hepatocytes from control normal rats, TGF-β1 concentration in the culture media was not detectable. In addition, basal apoptosis in cultured cells from control normal animals was not potentiated by IFN-α2b: the percentage of apoptosis in unstimulated normal hepatocytes was 4.7 ± 0.3%; in IFN-α2b–treated normal hepatocytes it was 4.2 ± 0.2%. These data are consistent with the poor response of rat hepatocytes to IFN-α previously reported.43 In contrast to control rat hepatocytes, hepatocytes from preneoplastic livers responded very well to IFN-α2b, as shown by elevated levels of secreted TGF-β1 and increased apoptosis. Further studies are required to determine the reason for these differences, for example, if preneoplastic rat hepatocytes express different patterns of IFN-α receptors than normal control rat hepatocytes.

As stated, in vitro studies with isolated hepatocytes have allowed us to demonstrate that IFN-α2b induces apoptosis in hepatocytes from preneoplastic liver, measured by fluorescence microscopy and caspase-3 activity. These cells also had higher nuclear accumulation of p-Smads-2/3, indicating increased TGF-β1 activation. Although in vitro production of TGF-β1 by hepatocytes from group 5 in response to IFN-α2b was much lower than that from group 1, secreted TGF-β1 levels seemed to be enough to activate TGF-β1 signaling and to induce apoptosis because both groups showed the same increase in apoptotic cell death after IFN-α2b treatment was the same in both groups. When anti–TGF-β1 was added to the culture media, TGF-β1 activation and apoptosis induced by IFN-α2b were blocked, reaching the same levels as unstimulated cells. Therefore, the apoptotic effect of IFN-α2b is mediated by the production of TGF-β1 from hepatocytes.

Taken together, these data clearly show that TGF-β1, which is produced and secreted by hepatocytes from preneoplastic liver under IFN-α2b treatment, stimulates hepatocytes apoptotic cell death in an autocrine/paracrine fashion. This postulated mode of action is in agreement with data published previously.7, 39, 44 The reduction of preneoplastic foci by endogenous TGF-β1 early in the carcinogenesis process would likewise protect against tumor formation.

In summary, hepatocellular carcinomas are a leading cause of cancer incidence and mortality worldwide. Since long-term IFN-α treatment has been shown to prevent hepatocarcinoma in humans, the present study has relevant clinical implications. In an etiologically heterogeneous tumor entity such as hepatocellular carcinoma, a detailed knowledge of the pivotal pathways for preneoplastic lesions appears crucial to critically redefine therapeutic treatments.45 Importantly, delineating the components and targets of IFN-α proapoptotic pathways should enhance our understanding of their significance in hepatocarcinoma treatment.

Ancillary