Stimulation by Transforming Growth Factor-α of DNA Synthesis and Proliferation of Adult Rat Hepatocytes in Primary Cultures: Modulation by α- and β-Adrenoceptor Agonists

Experimental Procedures

Animals.

Male Wistar rats (200–250 g) were obtained from Saitama Experimental Animal Co. (Saitama, Japan). They were allowed to adapt to a humidity- and temperature-controlled room for at least 3 days before the experiment was started. They were fed on a standard diet and tap water ad libitum. The study reported here has been carried out according to the Josai University guidelines for ethical animal care and the Guide for Care and Use of Laboratory Animals as adopted and promulgated by the U.S. National Institutes of Health.

Hepatocyte Isolation and Culture.

Rats were anesthetized by an i.p. injection of sodium pentobarbital (45 mg/kg). Hepatocytes were isolated from normal liver by a two-step in situ collagenase perfusion technique to facilitate disaggregation of the adult rat liver (Seglen, 1975). The viability of hepatocytes as assessed by trypan blue exclusion was >97%. Unless otherwise indicated, isolated hepatocytes were plated onto collagen-coated plastic culture dishes (Sumitomo Bakelite Co., Tokyo, Japan) at a density of 3.3 × 10⁴cells/cm² (3 × 10⁵ cells/35-mm dish), and allowed to attach for 3 h on collagen-coated dishes in Williams’ medium E containing 5% fetal bovine serum, 0.1 nM dexamethasone, and 100 ng/ml aprotinin in 5% CO₂ in air. The medium was then changed by aspiration, and the cells were cultured in serum- and dexamethasone-free Williams’ medium E supplemented with various concentrations of TGF-α. Where appropriate, the following agents were added: α₁, α₂, and β₂adrenoceptor agonists or antagonists, cAMP-elevating agents, phorbol 12-myristate 13-acetate (PMA), and specific inhibitors of signal transducers.

Measurement of DNA Synthesis.

Hepatocyte DNA synthesis was assessed by measuring the incorporation of [³H]thymidine into acid-precipitable materials as described in Morley and Kingdon (1972). Briefly, after an initial attachment period of 3 h, hepatocytes were washed twice with serum-free Williams’ medium E and cultured in a medium containing TGF-α (0.1–10 ng/ml) for an additional 4 and 21 h. The cells were pulsed at 2 h and 19 h after TGF-α stimulation for 2 h with [³H]thymidine (1.0 μCi/well). Incorporation of [³H]thymidine into DNA was then determined. Hepatocyte protein content was measured by a modified Lowry procedure with BSA as a standard (Lee and Paxman, 1972). The results were expressed as disintegrations per minute per milligram of protein per hour.

Counting Nuclei.

The number of nuclei was counted instead of the cell number according to the previously described procedure ofNakamura et al. (1983a) with minor modifications. Briefly, the primary cultured hepatocytes were washed twice with 2 ml of Dulbecco’s PBS (pH 7.4). Then, the cells were lysed by incubation with 0.25 ml of 0.1 M citric acid containing 0.1% Triton X-100 for 30 min at 37^°C. An equal volume of the nucleus suspension was mixed with 0.3% trypan blue in Dulbecco’s PBS (pH 7.4), and the number of nuclei was counted in a hemocytometer. This procedure was performed because the hepatocytes had firmly attached to the collagen-coated plates and were not dispersed sufficiently by 0.02% EDTA-0.05% trypsin treatment.

Materials.

The following reagents were obtained from Sigma Chemical Co. (St. Louis, MO): genistein, aphidicolin, metaproterenol hemisulfate, phenylephrine hydrochloride, glucagon, forskolin, dobutamine hydrochloride, d-sphingosine, dexamethasone, aprotinin, and recombinant human EGF. Wortmannin was obtained from R&D Systems, Inc. (Minneapolis, MN). PMA, 8-bromo cAMP (8-br-cAMP), pertussis toxin, and rapamycin were purchased from Research Biochemicals Inc. (Natick, MA). Recombinant human TGF-α was obtained from Pepro Teck, Inc. (Rocky Hill, NJ). 1-[6-[[17β-3-Methoxyestra-1,3,5(10)-trien-17-yl]amino] hexyl]-1H-pyrrol-2,5-dione (U-73122), and 1-[6-[[17β-3-me thoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-2,5-pyrrolidine-dione (U-73343), andN-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride (H-89) were obtained from BIOMOL Research Laboratories, Inc. (Plymouth Meeting, PA). 5-Bromo-6-(2-imidazolin-2-yl-amino)quinoxaline (UK-14304) was a gift from Pfizer Central Research (Sandwich, UK). Monoclonal antibody against epidermal growth factor receptor (Ab-1) and 2′-amino-3′-methoxyflavone (PD98059) were obtained from Calbiochem-Behring Corp. (La Jolla, CA). Williams’ medium E and newborn calf serum were purchased from Flow Laboratories, Inc. (Ayrshire, Scotland). Collagenase (type II) was obtained from Worthington Biochemical Corp. (Freehold, NJ). [methyl-³H]Thymidine (20 Ci/mmol) was obtained from DuPont-New England Nuclear (Boston, MA). All other reagents were of analytical grade.

Statistical Analysis.

Group comparisons were made with ANOVA for unpaired data followed by post hoc analysis with Dunnett’s multiple comparison test. P values <.05 were regarded as statistically significant.

Results

Time Course Associated with Stimulation of Hepatocyte DNA Synthesis and Proliferation Induced by TGF-α and Their Modulation by α and β Adrenoceptor Agonists.

We studied the time course of DNA synthesis and the number of nuclei (proliferation) of hepatocytes in response to 1.0 ng/ml TGF-α at a cell density of 3.3 × 10⁴ cells/cm². In addition, we examined the effects of α and β adrenoceptor agonists on the TGF-α-induced hepatocyte DNA synthesis and proliferation at each time point. A significant increase in the DNA synthesis occurred 2 h after culture of hepatocyte with 1 ng/ml TGF-α, 2 h after culture with TGF-α and α₂ adrenoceptor agonist, 10⁻⁶ M UK-14304 (Ogihara, 1995), and 2 h after culture with TGF-α and α₁ adrenoceptor agonist, 10⁻⁷ M phenylephrine (Fig. 1A). The number of nuclei induced by 1 ng/ml TGF-α was significantly increased at ∼2.5 h after TGF-α addition, reached a plateau at 4 h, and sustained for an additional 17 h (Fig. 1B). As to the adrenoceptor-mediated modulation, it was found that the effects of 1 ng/ml TGF-α on hepatocyte proliferation were potentiated by 10⁻⁷ M phenylephrine or 10⁻⁶ M UK-14304. The hepatocyte DNA synthesis preceded the increase in the nucleus number. In contrast, hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α was found to be almost completely inhibited by the β₂adrenoceptor agonist 10⁻⁷ M metaproterenol. Phenylephrine, UK-14304, and metaproterenol alone did not significantly affect hepatocyte DNA synthesis and proliferation at any concentrations used in the present study.

Time Course Associated with Effects of Dexamethasone Pretreatment on TGF-α-Induced Hepatocyte DNA Synthesis and Proliferation.

To investigate the mechanism by which TGF-α rapidly stimulates hepatocyte DNA synthesis and proliferation, we examined the time course for effects of dexamethasone pretreatment on TGF-α-induced hepatocyte mitogenesis. Figure 2 shows that the time for the initiation of hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α is delayed, depending on increasing concentrations of dexamethasone. The maximum stimulation of hepatocyte DNA synthesis and proliferation was observed with 10⁻¹⁰and 10⁻⁹ M dexamethasone pretreatment. These results indicate that dexamethasone in culture is a major regulator of hepatocyte DNA synthesis and proliferation.

Dose-Dependent Effects of TGF-α on Hepatocyte DNA Synthesis and Proliferation.

We examined the dose-dependent effects of TGF-α on hepatocyte DNA synthesis and proliferation. Hepatocytes were cultured with various concentrations of TGF-α for 4 h at a cell density of 3.3 × 10⁴ cells/cm², followed by measurement of DNA synthesis and the number of nuclei. As shown in Fig. 3, TGF-α produced a dose-dependent increase in hepatocyte DNA synthesis and proliferation, and was significant at concentrations of 0.1 ng/ml and greater. A maximum increase in DNA synthesis was observed with 1 to 2 ng/ml TGF-α. The 50% effective concentration values (EC₅₀) for DNA synthesis and proliferation occurred at 0.36 mg/ml and 0.45 ng/ml, respectively.

Effects of an Anti-EGF Receptor Monoclonal Antibody on TGF-α- or EGF-Induced Hepatocyte DNA Synthesis and Proliferation.

Because TGF-α is reported to interact with the EGF receptor and is suggested to mediate all of its biological effects through the EGF receptor (Lee et al., 1995), we subsequently examined the effects of monoclonal anti-EGF receptor antibody (1–100 ng/ml) on the EGF- or TGF-α-induced hepatocyte DNA synthesis and proliferation at 4 and 21 h of culture. This antibody is known to inhibit EGF binding to its receptor (Kawamoto et al., 1983). Figure4 shows that the proliferative effects of 20 ng/ml EGF on hepatocyte DNA synthesis and proliferation were completely blocked by a monoclonal antibody against EGF receptor (>25 ng/ml). The IC₅₀ values for the hepatocyte DNA synthesis at 4 and 21 h were 27 ng/ml and 39 ng/ml, respectively. IC₅₀ values for the number of nuclei at 4 and 21 h were 28 ng/ml and 36 ng/ml, respectively. In contrast, the proliferative effects of 1 ng/ml TGF-α were not affected significantly by treatment of hepatocytes with various concentrations of monoclonal antibody against EGF receptor (1–100 ng/ml).

Influence of Cell Density on TGF-α-Stimulated Hepatocyte DNA Synthesis and Proliferation: Modulation by α₁, α₂, and β₂ Adrenoceptor Agonists.

To determine whether or not the proliferative effects of TGF-α are affected by initial plating densities, we investigated the density-dependence of hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α at 4 h of culture. Figure5A shows that hepatocyte DNA synthesis induced by 1 ng/ml TGF-α was significantly increased relative to the increasing initial plating densities (1.0–3.3 × 10⁴cells/cm²). It reached a maximum value at cell densities of 3.3 × 10⁴ cells/cm² (Fig. 5A) and did not decrease even at higher cell densities (5.0–10 × 10⁴cells/cm²). In general, there was a good correlation between the ability of 1 ng/ml TGF-α to stimulate hepatocyte DNA synthesis and the increase in the number of nuclei at various initial plating densities (Fig. 5). In addition, the effect of 1 ng/ml TGF-α on hepatocyte DNA synthesis and proliferation was potentiated by 10⁻⁷ M phenylephrine and 10⁻⁶ M UK-14304, whereas it was completely abolished by 10⁻⁷ M metaproterenol at the various cell densities tested. (Fig. 5). Each adrenoceptor agonist alone did not significantly influence hepatocyte DNA synthesis and proliferation at the various cell densities tested in the present study (data not shown).

Dose-Dependent Effects of Phenylephrine, UK-14304, and Metaproterenol on TGF-α-Induced Hepatocyte DNA Synthesis and Proliferation.

We next examined the dose-dependent effects of the α and β adrenoceptor agonists on the TGF-α-induced stimulation of hepatocyte DNA synthesis and proliferation during 4 h of culture. As shown in Fig. 6A, the hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α were potentiated dose-dependently by UK-14304 treatment (EC₅₀ = 5 × 10⁻⁷ M), whereas TGF-α effects were dose-dependently inhibited by metaproterenol (IC₅₀ = 1.5 × 10⁻⁸ M). The α₂ or β₂ adrenoceptor agonists alone did not significantly influence hepatocyte DNA synthesis and proliferation at the various concentrations tested (10⁻¹⁰–10⁻⁸ M). As shown in Fig. 6B, a specific α₁ adrenoceptor agonist, phenylephrine, dose-dependently potentiated the TGF-α-induced hepatocyte DNA synthesis and proliferation at 4 h of culture, but phenylephrine on its own had no significant effect on hepatocyte DNA synthesis and proliferation at the various concentrations tested (10⁻⁹–10⁻⁶ M). The phenylephrine effects occurred with an EC₅₀ value of 7 × 10⁻⁸M. Each effect of the α₁, α₂, and β₂ adrenoceptor agonists on the TGF-α-induced hepatocyte DNA synthesis was closely correlated with the alteration of hepatocyte proliferation as assessed by increases in the number of nuclei.

Effects of Specific Adrenoceptor Antagonists and H-89 on α₂ and β₂ Adrenoceptor Agonist-Induced Hepatocyte DNA Synthesis and Proliferation in the Presence of TGF-α.

To confirm the effect of α₂ adrenoceptor mediation of UK-14304 and β₂ adrenoceptor mediation of metaproterenol on hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α, we used specific antagonists of α₂ or β₂ adrenoceptors. In addition, to characterize the involvement of the cAMP/protein kinase A system in adrenoceptor-induced hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α, we investigated the effects of other cAMP-elevating agents and the specific protein kinase A inhibitor H-89 (Zusick et al., 1994) on these responses. As summarized in Table1, the UK-14304-induced stimulation of hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α was abolished by the specific α₂ adrenoceptor antagonist yohimbine (10⁻⁷ M), but not by the specific α₁ adrenoceptor antagonist prazosin (10⁻⁶M), confirming α₂ adrenoceptor mediation of the UK-14304 effect. However, metaproterenol-induced inhibition of hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α was reversed by the specific β₂ adrenoceptor antagonist butoxamine (10⁻⁷ M), but not by the specific β₁ adrenoceptor antagonist metoprolol (10⁻⁶M), confirming the β₂ adrenoceptor mediation of the metaproterenol effect. These results suggest that metaproterenol acts through β₂ adrenoceptors by increasing intracellular cAMP levels. If this is indeed the case, then other cAMP-elevating agents also will inhibit the TGF-α-induced DNA synthesis and proliferation in primary cultured hepatocytes. Consistent with this hypothesis, cell membrane-permeable cAMP analog 8-br-cAMP (10⁻⁷ M) also abolished the TGF-α-induced hepatocyte DNA synthesis and proliferation at 4 and 21 h of culture. H-89 (10⁻⁷ M) reversed the inhibitory effects of both 10⁻⁷ M metaproterenol and 10⁻⁷ M 8-br-cAMP on hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α at 4 and 21 h of culture. H-89 alone did affect the TGF-α-induced hepatocyte mitogenesis. In addition, the metaproterenol inhibition of the TGF-α-induced hepatocyte DNA synthesis and proliferation was significantly reversed by the α₂ adrenoceptor agonist UK-14304, suggesting that inhibition of adenylate cyclase activity via inhibitory G protein (G_i) is operational in cultured hepatocytes. The potentiating effects of UK-14304 (10⁻⁵ M) on TGF-α-induced hepatocyte DNA synthesis and proliferation were blocked by metaproterenol (10⁻⁷ M), but not by 8-br-cAMP (10⁻⁷ M), suggesting that UK-14304 acts at an α₂ adrenoceptor level. In addition, other cAMP-elevating agents, such as 10⁻⁷ M glucagon and 10⁻⁷ M forskolin, which stimulate adenylate cyclase activity by a different mechanism, also completely inhibited the hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α at 4 and 21 h of culture. Furthermore, in contrast to the case of metaproterenol, the β₁ adrenoceptor agonist dobutamine did not influence the TGF-α-induced hepatocyte DNA and proliferation at concentrations up to 10⁻⁶ M. None of these cAMP-elevating agents on its own affected hepatocyte mitogenesis.

Effects of Specific α Adrenoceptor Antagonist and Sphingosine on Phenylephrine- and PMA-Induced Hepatocyte DNA Synthesis and Proliferation in the Presence of TGF-α.

To confirm the α₁ adrenoceptor mediation of phenylephrine on hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α, we examined the effect of specific α₁ and α₂adrenoceptor antagonists on the hepatic mitogenesis at 4 and 21 h of culture. In addition, to characterize the involvement of the phospholipase C/protein kinase C system in hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α, we investigated the effect of the specific phospholipase C inhibitor U-73122 (Thompson et al., 1991) and protein kinase C inhibitor sphingosine (Merrill et al., 1989) on these responses. As shown in Table2, the ability of phenylephrine to potentiate hepatocyte DNA synthesis and proliferation was almost completely blocked by prazosin (10⁻⁷ M), but not by yohimbine (10⁻⁶ M). These results suggest that phenylephrine acts through α₁ adrenoceptors by activating phospholipase C and the subsequent increase in diacylglycerol and/or intracellular calcium levels. If this is indeed the case, a cell membrane-permeable analog of diacylglycerol, PMA (Castagana et al., 1982), also should stimulate hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α. As expected, although PMA alone had no significant effect on hepatocyte DNA synthesis and proliferation (data not shown), hepatocyte DNA synthesis and proliferation in the presence of 1 ng/ml TGF-α were potentiated by PMA (10⁻⁷M) after 4 and 21 h of culture. The potentiating effects of PMA was reversed by coincubation with the protein kinase C inhibitor sphingosine (5 × 10⁻⁶ M) for 4 and 21 h. Sphingosine (5 × 10⁻⁶ M) alone had no effect on hepatocyte DNA synthesis and proliferation during 4 and 21 h of culture (data not shown), but significantly reduced the TGF-α-induced hepatic mitogenesis during 4 and 21 h of culture. Similarly, to determine the possible involvement of Ca²⁺ mobilization in the phenylephrine-stimulated hepatocyte DNA synthesis and proliferation, cells were treated with the calcium ionophore ionomycin for 4 and 21 h in the presence of TGF-α (1 ng/ml). Potentiation of both hepatocyte DNA synthesis and proliferation was observed at a dose of 10⁻⁶ M ionomycin at 4 and 21 h of culture. Furthermore, U-73122 (10⁻⁵ M), a phospholipase C inhibitor, was found to significantly attenuate the 1 ng/ml TGF-α action on hepatocyte DNA synthesis and proliferation during 4 and 21 h of culture. U-73343 (10⁻⁵ M), a close structural analog of U-73122, which is known to have no such inhibitory action on phospholipase C, did not significantly affect the TGF-α-induced hepatocyte DNA synthesis and proliferation during early (4 h) and late (21 h) phases of culture (Table 3).

Effects of Specific Inhibitors of Signal-Transducing Elements on Hepatocyte DNA Synthesis and Proliferation Induced by TGF-α.

We investigated whether or not the mitogenic responses of hepatocytes to TGF-α (1 ng/ml) were mediated by such signal transducers as receptor tyrosine kinase, G_i, PI3K, MAP kinase kinase, and p70 S6K by corresponding specific inhibitors of the signal transducers (Table3). The TGF-α effects on hepatocyte DNA synthesis and proliferation were not affected significantly by a specific G_i protein inhibitor pertussis toxin (100 ng/ml) (Katada and Ui, 1982). In contrast, significant inhibitory effect of genistein (10⁻⁶M) (Akiyama et al., 1987), a specific inhibitor of receptor tyrosine kinase and wortmannin (5 × 10⁻⁷ M) (Baggiolini et al., 1987), a specific inhibitor of PI3K, on the TGF-α-induced hepatocyte DNA synthesis and proliferation during the early and late phases of culture were observed, suggesting that receptor tyrosine kinase and PI3K are involved in the hepatocyte proliferation induced by TGF-α. Furthermore, PD98059 (5 × 10⁻⁵ M) (Alessi et al., 1995), a specific inhibitor of MAP kinase kinase, also produced a significant attenuation of the TGF-α-induced hepatocyte DNA synthesis and proliferation, suggesting that MAP kinase kinase is an indispensable component of hepatocyte DNA synthesis and proliferation. Additionally, rapamycin (10 ng/ml) (Price et al., 1992), a specific inhibitor of p70 S6K inhibitor, almost completely blocked the hepatocyte DNA synthesis and proliferation induced by 1 ng/ml TGF-α. None of the specific inhibitors of signal transducers, namely, genistein (5 × 10⁻⁶ M), wortmannin (5 × 10⁻⁷ M), PD98059 (5 × 10⁻⁵ M), and rapamycin (10 ng/ml), on its own had a significant effect on hepatocyte DNA synthesis and proliferation.

Discussion

In this report, we demonstrated that TGF-α rapidly and strongly increases hepatocyte DNA synthesis and proliferation with a greater potency than EGF (Mead et al., 1989; Kimura and Ogihara, 1997a). The mechanism by which TGF-α rapidly stimulated hepatocyte DNA synthesis and proliferation may be the low concentration of dexamethasone in the culture medium (Fig. 2). The addition of the low concentration of dexamethasone (0.1 nM) to culture explains why the results obtained in our short-term studies were different from those of previous studies with longer-term culture. Moreover, in contrast to EGF, TGF-α is able to stimulate hepatocyte proliferation even at high initial plating densities. A similar phenomenon has been observed with other growth factors, such as insulin and PDGF (Kimura and Ogihara, 1997b, 1998a). The cell density-independent mechanism of TGF-α action remains to be established, but the involvement of a cell membrane modifier and/or autocrine secretion of some stimulatory factors have been suggested (Nakamura et al., 1983a,b).

The effects of TGF-α in a variety of in vitro assays with cultured cells have been reported to be essentially identical with those observed by EGF and to show little qualitative difference. EGF and TGF-α are ligands for the EGF receptor and act as mitogens for a variety of tissues (Winkler et al., 1989; Derynck, 1992;Strömblad and Andersson, 1993; Lee et al., 1995). The EGF receptor may contain two distinct binding sites, one for EGF and the other for TGF-α. As other investigators have suggested, these two sites could be very far apart in the extracellular domain of receptors (Lyons and Moses, 1990). Ab-1 monoclonal antibody is known to inhibit EGF binding to its receptor and is a competitive antagonist of the EGF-stimulated growth of hepatocytes (Kawamoto et al., 1983). With the specific antibody against the EGF receptor (Fig. 4), we showed that the proliferative effects of extracellular application of TGF-α and EGF may be largely mediated by binding to their own binding sites in the extracellular domain. In our assay system, TGF-α and EGF exert biologically different responses in terms of cell-density dependence, modulation by α and β adrenoceptor agonist, and involvement of PI3K as a signal transducer (Kimura and Ogihara, 1997a,b, 1998a,b). Therefore, one mechanism, namely, that a qualitatively different interaction of EGF and TGF-α with the receptor could lead to a difference in the balance between the effects of the activated receptor on a distinct signaling pathway is conceivable. However, more information is needed on possible differences in the signal transduction produced by EGF and TGF-α.

Adrenergic regulation is now thought to be involved in the hepatic regenerative process in vivo (Refsnes et al., 1992; Diehl and Rai, 1996; Michalopoulos and DeFrances, 1997). In a previous report, we demonstrated that proliferative effects of several growth factors (e.g., EGF, insulin, HGF, and PDGF) on hepatocyte proliferation were modulated differently by α and β adrenoceptor agonists (Kimura and Ogihara, 1997a,b, 1998a,b). Therefore, we used the same culture medium to examine the effect of TGF-α on DNA synthesis and proliferation of adult rat hepatocytes in primary culture. The result obtained showed that TGF-α effects were potentiated by specific α₁ and α₂ adrenoceptor agonists, whereas the TGF-α-induced hepatocyte mitogenesis was almost completely inhibited by a β₂ adrenoceptor agonist (Fig. 6A; Table 1). The present results suggest that potentiation of the TGF-α action by α₁ and α₂ adrenoceptor agonists in the liver are likely the first stimulator of hepatocytes to enter G₁ and subsequently promote their transit through the cell cycle. This α₁ adrenoceptor action is probably mediated by the phospholipase C/diacylglycerol/protein kinase C system and/or intracellular calcium mobilization (Berridge, 1993), given that PMA and ionomycin mimic the action of the α₁ adrenoceptor agonist. The effect of the α₂ adrenoceptor agonist may be mediated by a reduction in intracellualr cAMP via G_i protein. By contrast, the inhibitory modulation of the TGF-α action by the β₂ adrenoceptor agonist and other cAMP-elevating agents may be associated with an increment of intracellular cAMP. Consistent with this finding the addition of 8-br-cAMP mimics the effect of a β₂adrenoceptor agonist (Table 1). The inability of the β₁ adrenoceptor agonist dobutamine to inhibit the TGF-α-induced hepatocyte mitogenesis may be a result of the very low expression of β₁ adrenoceptors.

For the effects of TGF-α, metaproterenol and cAMP-elevating agents markedly decreased the mitotic effect, whereas they enhanced the response to HGF as described previously (Kimura and Ogihara, 1997c). The molecular mechanism of these contrasting results is unknown. However, TGF-α and HGF signal transduction pathways regulated by different cell surface receptors are known to use an MAP kinase (Boylan and Gruppuso, 1994; Gines et al., 1995). There are important reports showing that the effects of the cAMP/protein kinase A system on the MAP kinase pathway depend on both cell type and the type of tyrosine kinase receptor (Frödin et al., 1994; Calleja et al., 1997). Therefore, the differences in the cAMP effects on the MAP kinase cascade are likely to be a consequence of qualitative differences between TGF-α and HGF signaling. The cross-talk between the TGF-α- or HGF-signaling pathway and the cAMP/protein kinase A pathway remains to be explored.

Growth factors exert their growth-regulating effects on hepatocytes through intracellular signal transduction to bring about changes in nuclear DNA synthesis and proliferation (Sun and Tonks, 1994). Specific inhibitors of the intracellular signaling cascade are also useful probes with which to characterize target proteins involved in the activation of DNA synthesis and proliferation induced by the TGF-α in primary cultures of adult rat hepatocytes. As summarized in Tables 2and 3, hepatocyte DNA synthesis and proliferation induced by TGF-α was significantly blocked by genistein, U-73122, sphingosine, wortmannin, and rapamycin, suggesting that hepatocyte mitogenesis was stimulated through receptors that are associated with tyrosine kinase, and also was mediated via a phospholipase C, protein kinase C, PI3K, and p70 S6K. MAP kinase is commonly activated by a large number of extracellular stimuli and is hypothesized to play a key role in various intracellular signal transduction pathways (Davis, 1993; Gines et al., 1995). The use of PD98059 to specifically block the activation of MAP kinase kinase (Alessi et al., 1995) demonstrated that this signal transducer can be involved in hepatocyte DNA synthesis and proliferation induced by TGF-α (Table 3). Although, all of the above-mentnioned signal-transducing elements play a critical role in stimulating hepatocyte DNA synthesis and proliferation, the cascades of sequential phosphorylation events have not been definitively established yet. More research into the mechanisms of growth regulation by TGF-α in primary cultured adult rat hepatocytes is needed. In contrast, treatment with a G_i protein inhibitor, pertussis toxin, revealed that the TGF-α effects on hepatocyte DNA synthesis and proliferation were not affected significantly, suggesting that TGF-α receptors are not linked to Gi protein. This is in sharp contrast to the case of insulin-like growth factor II, the action of which is very sensitive to pertussis toxin (Kimura and Ogihara, 1998b).

In conclusion, this report represents several lines of new evidence supporting the important and distinct roles of TGF-α in vitro. Studies with specific inhibitors of signal-transducing elements demonstrate that among the elements that link signals of cell surface receptors to the nucleus, the proliferative action of TGF-α is mediated, at least, by tyrosine kinase, phospholipase C, PI3K, protein kinase C, MAP kinase kinase, and p70 S6K. However, the complete sequence of events in which TGF-α increases hepatocyte replication remains to be determined. These data suggest a role for endogenous TGF-α in the initiation and maintenance of regenerative hepatic growth in vivo. In addition, our critical observation is relevant for the modulation by specific adrenoceptor agonists of the TGF-α-induced hepatocyte DNA synthesis and proliferation. We propose that α₁ and α₂ adrenoceptor stimulation enhanced, and β₂ adrenoceptor stimulation decreased the TGF-α-induced hepatocyte DNA synthesis and proliferation, which in turn may contribute to the modulation of hepatic regeneration in vivo.