Hepatitis C viral proteins interact with Smad3 and differentially regulate TGF-β/Smad3-mediated transcriptional activation

Abstract

Transforming growth factor-β (TGF-β) is a pleiotropic cytokine implicated as a pathogenic mediator in various liver diseases. Enhanced TGF-β production and lack of TGF-β responses are often observed during hepatitis C virus (HCV) infection. In this study, we demonstrate that TGF-β-mediated transactivation is decreased in cells exogenously expressing the intact HCV polyprotein. Among 10 viral products of HCV, only core and nonstructural protein 3 (NS3) physically interact with the MH1 (Mad homology 1) region of the Smad3 and block TGF-β/Smad3-mediated transcriptional activation through interference with the DNA-binding ability of Smad3, not the nuclear translocation. However, the interactive domain of NS3 extends to the MH2 (Mad homology 2) region of Smad3 and a distinction is found between effects mediated, respectively, by these two viral proteins. HCV core, in the presence or absence of TGF-β, has a stronger suppressive effect on the DNA-binding and transactivation ability of Smad3 than NS3. Although HCV core, NS3, and the HCV subgenomic replicon all attenuate TGF-β/Smad3-mediated apoptosis, only HCV core represses TGF-β-induced G1 phase arrest through downregulation of the TGF-β-induced p21 promoter activation. Along with this, HCV core, rather than NS3, exhibits a significant inhibitory effect on the binding of Smad3/Sp1 complex to the proximal p21 promoter in response to TGF-β. In conclusion, HCV viral proteins interact with the TGF-β signaling mediator Smad3 and differentially impair TGF-β/Smad3-mediated transactivation and growth inhibition. This functional counteraction of TGF-β responses provides insights into possible mechanisms, whereby the HCV oncogenic proteins antagonize the host defenses during hepatocarcinogenesis.

Introduction

Chronic infection with hepatitis C virus (HCV) is a major factor in the development of liver cirrhosis and hepatocellular carcinoma (HCC) (Seeff, 2002). Despite the well-known course of disease through fibrosis to hepatocarcinogenesis (Lauer and Walker, 2001), the mechanisms underlying the disturbance of the host defense system in chronic HCV-infected patients remain obscure. HCV is a positive-strand RNA virus with a 9.5-kb viral genome encoding a large precursor protein, which is processed by both cellular and viral proteases into 10 mature structural and nonstructural viral proteins (Reed and Rice, 2000). Among the 10 viral proteins, core and nonstructural protein 3 (NS3) have received much attention, since both of them have a multifunctional nature including transforming and antiapoptotic activity. HCV core has been shown to act in trans on numerous cellular or viral promoters (Lai and Ware, 2000), many of which encode proteins involved in the regulation of cellular proliferation. Additionally, HCV core targets a wide spectrum of cellular factors (Tellinghuisen and Rice, 2002). These associations result in modulations that affect numerous biological activities. Furthermore, transgenic mice with constitutive expression of HCV core show liver steatosis and eventually develop HCC (Moriya et al., 1997; 1998).

HCV NS3 is a bifunctional protein. The N-terminal region of NS3 encodes a protease activity that is responsible for the proteolytic cleavage at the various boundaries between different NS proteins (De Francesco and Steinkuhler, 2000). The C-terminal region of NS3 encodes RNA helicase/NTPase activities that are essential for viral replication (Kwong et al., 2000). Previous studies also revealed that NS3 confers transformation and antiapoptotic ability on nontumorigenic fibroblast cells and these are largely dependent on its protease function (Sakamuro et al., 1995; Fujita et al., 1996; Zemel et al., 2001).

Transforming growth factor-β (TGF-β) is one of the best-characterized cytokines that affect cell growth, cell death, differentiation, and morphogenesis (Derynck et al., 2001). The principal downstream effect of the intracellular signal transmission by the TGF-β family is initiated via binding to plasma membrane serine/threonine kinase receptors and the activation of specific downstream cytoplasmic effectors, the Smad family of proteins (Massague, 2000). The receptor-activated Smads involved in TGF-β signaling are Smad2 or Smad3. Once phosphorylated, they form heteromeric complexes with the Smad4 and then translocate to the nucleus, where they act as the transcriptional regulators. Structural analysis of the Smad molecules reveals the presence of the two conserved domains, one is the N-terminal Mad homology 1 (MH1), which contains a nuclear localization signal and confers DNA-binding ability, and the other is the C-terminal Mad homology 2 (MH2), which is responsible for protein–protein interaction (Jayaraman and Massague, 2000). Smad proteins regulate target gene expression via binding to CAGA- or GTCT-like sequences (termed the Smad binding element, SBE) in a variety of promoters (Dennler et al., 1998; Zhang et al., 1998). For promoters lacking an SBE, Smad proteins are also able to confer effects through cooperation with other sequence-specific DNA-binding transcription factors, such as Sp1 (Pardali et al., 2000).

Physiologically, TGF-β is expressed in the nonparenchymal cells of the liver, but not in the hepatocytes from normal or regenerating liver (De Bleser et al., 1997). The microenvironment of the liver keeps hepatocyte function under control and thus may prevent the progression of transformed hepatocytes to HCC. As the triggering of apoptotic cell death by TGF-β is a well-recognized phenomenon in the differentiation and the maintenance of the liver (Oberhammer et al., 1992), its underlying mechanism has been extensively studied in a variety of cell types originating in the liver. From these studies, Smad/SBE-dependent transcriptional activation of proapoptotic genes has emerged as one of the important processes of this response in hepatic cells (Jang et al., 2002). Clinical studies also revealed that the level of TGF-β in the plasma is elevated in HCV-infected patients and is even further increased in patients with HCC (Ito et al., 1991; Matsuzaki et al., 2000; Ray et al., 2003). Apparently, the antiproliferation activities of TGF-β can no longer protect HCV-infected patients from hepatocarcinogenesis even with enhanced TGF-β production. It is likely that the disappearance of the antiproliferation activities of TGF-β could partly result from an effect of the HCV viral products.

Here, we show that HCV core and NS3 interact with Smad3 and repress the TGF-β/Smad3-mediated transactivation through interference with the SBE binding ability of Smad3. Although core and NS3 both repress TGF-β/Smad3-induced SBE-containing promoter activity (e.g. on the plasminogen activator inhibitor (PAI)-1 promoter), they display a differential repression of the TGF-β/Smad3-induced transcriptional activation of the p21WAF1/Cip1/Sid1 (p21) promoter, which lacks an SBE. This distinction also occurs in terms of their inhibitory potency on the binding of Smad3/Sp1 complex to the proximal p21 promoter. This is further influential in how NS3 and core are able to impair differentially the TGF-β-mediated growth-inhibitory response. These results provide the possible mechanisms whereby the HCV oncogenic proteins counteract the host defenses during hepatocarcinogenesis.

Results

HCV full-length polyprotein inhibits TGF-β-induced transcriptional activation

The TGF-β signal exerts its influence on cellular responses mainly through the modulation of the expression of numerous cellular factors. To investigate the viral protein-mediated effect on TGF-β pathway, TGF-β-induced transcriptional activity in the presence of the full-length HCV polyprotein was examined. HuH-7 cells were transiently cotransfected with the intact HCV polyprotein expression construct pSRα/HCV-FL (0.35 or 0.7 μg) and the TGF-β-responsive p3TP-Lux reporter (0.3 μg). This reporter construct contains three repeats of a 12-O-tetradecanoylphorbol-13-acetate (TPA) response element (TRE) and a promoter fragment (nucleotide positions –636 to –740) of the human PAI-1 and has been shown to be efficiently stimulated by TGF-β in a variety of cell lines (Wrana et al., 1992; Abe et al., 1994). As shown in Figure 1, the TGF-β induced about 5.8-fold transactivation of the p3TP-Lux reporter in HuH-7 cells, while in the presence of HCV polyprotein, a 1.3- to 1.6-fold decrease of TGF-β-induced transactivation occurred, depending on the expression level of HCV polyprotein as indicated by the level of core protein (bottom panel). In addition, the repression effect on TGF-β-mediated transcriptional activity was also observed in the context of the subgenomic replicon Ava.5 cells (data not shown). The subgenomic replicon Ava.5 cells contain all the nonstructural proteins of HCV including NS3, and its physiological condition bears a striking resemblance to a cell containing replicating HCV (Blight et al., 2000). Our results thus suggest that loss of TGF-β-induced transcriptional activation occurs with the full-length HCV polyprotein as well as the subgenomic replicon.

Figure 1
figure1

Effects of HCV polyprotein on TGF-β-induced transactivation. The full-length HCV polyprotein expression construct pSRα/HCV-FL (0.35 or 0.7 μg) together with the TGF-β-responsive 3TP-Lux reporter plasmid (0.3 μg) was cotransfected into HuH-7 cells. The total amount of plasmid DNA used for transfection was 1 μg each and adjusted by adding pSRα vector or salmon sperm DNA, as applicable. TGF-β treatment was carried out at day 3 post-transfection and luciferase activity was measured 24 h after incubating with or without 10 ng/ml TGF-β. The relative expression level of HCV core protein in the transfected cells were visualized by Western blot and was shown at the bottom. The results are represented as the mean±s.d. from at least three separate experiments and the number above the bar represents the activation fold obtained by comparing the luciferase activity in the presence or absence of TGF-β

HCV core and NS3 suppress the TGF-β-induced transcriptional activation

To examine which HCV viral protein is responsible for the suppression of TGF-β-mediated transcriptional activation, expression constructs (1 or 2 μg) for each of the HCV viral proteins and p3TP-Lux reporter (1 μg) were cotransfected into HuH-7 cells (Figure 2a). A dosage-dependent repression (maximum repression two- to threefold) on the TGF-β-induced p3TP-Lux activity was observed only in HuH-7 cells expressing the HA-tagged core, NS3, or NS3/4A. The effect of other HCV viral proteins, including E1, E2, p7, NS2, NS4B, NS5A, and NS5B, on the TGF-β response was relatively minor, if any. We also examined if the HuH-7 or HeLa stable cell lines expressing the HCV core (HuH-7/C190) or NS3 (HuH-7/NS3, HeLa/NS3 #38 and #4) displayed the same suppression effect on TGF-β-induced transcriptional activation. HuH-7/C190 and HuH-7/NS3 are pooled population of HCV viral protein-producing cells, while both HeLa/NS3 #38 and #4 are tetracycline-regulated NS3-producing clones (see Materials and methods). The expression of HCV viral proteins in these stable cell lines was confirmed by Western blotting (Figure 2b, left panel). As shown in Figure 2b (left panel), the NS3 expression of HeLa/NS3 #38 and #4 was conditionally repressed by the addition of doxycycline (a tetracycline analog) (compare lanes 8 and 10 with lanes 7 and 9). Similar reporter experiments were conducted in these cell lines and the results are shown in Figure 2b (right panel) and indicate about two- to 3.8-fold repression in the TGF-β-mediated transactivation in these core- or NS3-producing stable cell lines (lanes 4, 8, 14, and 18) as compared to the controls (lanes 2, 6, and 10). Notably, HeLa/NS3 #38 exhibited a stronger repressive effect than a second clone (HeLa/NS3 #4), which is consistent with the relative expression level of NS3 in these cells (left panel). Additionally, when the NS3 expression was turned off in the presence of doxycycline, both HeLa/NS3 clones could restore TGF-β-mediated p3TP-Lux reporter activation to the similar extent (8.8- to 9.1- fold activation) of the control cells (right panel, compare lanes 16 and 20 with lane 12). This, in combination with the dosage-dependent effects observed from transient transfection experiment (Figure 2a), suggests that the repression effects on TGF-β-mediated transactivation are elicited by these two viral proteins. Thus, core and NS3 are two crucial HCV viral products that are responsible for the inhibition of TGF-β-induced p3TP-Lux reporter activation.

Figure 2
figure2

HCV core and NS3 inhibit TGF-β-induced transcriptional activation. (a) 3TP-Lux reporter plasmid (1 μg) was transfected into HuH-7 cells together with 1 or 2 μg of vector or expression plasmids encoding HA-tagged HCV viral products as indicated. (b) HCV core (HuH-7/C190)-, NS3 (HuH-7/NS3, HeLa/NS3 # 38, or HeLa/NS3 #4)- producing cells and their control cells (HuH-7, HuH-7/Vector, or HeLa) were cotransfected with 3TP-Lux reporter (1 μg) as indicated. The HeLa/NS3 clones were cultured with (+) or without (−) the addition of 5 μg/ ml doxycycline (Dox; Clontech) for 24 h. The expression of HCV viral protein in these cell lines as detected by immunoblot is shown in the left panel. The immunoblots of input cell lysates with antibody against tubulin were also shown at the bottom panel. Luciferase activity was measured 24 h after incubation with or without 10 ng/ml TGF-β. The results are represented as the mean±s.d. from at least three separate experiments and the number above the bar represents the activation fold obtained by comparing the luciferase activity in the presence or absence of TGF-β

HCV core and NS3 differentially inhibit TGF-β/Smad3-mediated transcriptional activation

Since p3TP-Lux reporter contains AP-1 sites resided within TRE region, we first investigated the effect of these two viral proteins on TGF-responsive transactivation of pAP-1-Lux reporter. As shown in Figure 3a, a fourfold induction on pAP-1-Lux activity was observed in TGF-β-treated HuH-7 cells (compare lane 2 with lane 1). Interestingly, both HCV core and NS3 elicited a slight increase (maximum increase 1.4- to 1.8-fold) of TGF-β-induced transactivation on pAP-1-Lux reporter (Figure 3a, compare lanes 6 and 10 with lane 2), and this apparently is an opposite effect to that on p3TP-Lux reporter (see Figure 2). These findings imply that the involvement of AP-1 proteins on the suppression ability of HCV core and NS3 protein on TGF-β signaling is minor, if any.

Figure 3
figure3

HCV core and NS3 differentially affect TGF-β/Smad3-mediated transcriptional activation. (a) HCV core and NS3 activate TGF-β-induced transcriptional activation of pAP-1-Lux reporter activity. HuH-7 cells were cotransfected with increasing amount of HCV viral protein expression constructs, pcDNA3/HA-core (HA-core, lanes 3–6) or pcDNA3/FLAG-NS3 (FLAG-NS3, lanes 7–10) (1.0 or 2.0 μg), along with 1 μg of pAP-1-Lux reporter. Luciferase activity was measured 24 h after incubation with or without 10 ng/ml of TGF-β. (b) HCV core and NS3 inhibit TGF-β-induced transcriptional activation of CAGA box reporter activity. Increasing amount of HCV viral protein expression constructs, pcDNA3/HA-core (HA-core) and pEF/NS3 (NS3) (0.25, 1, or 2.0 μg), along with 1 μg of (CAGA)6TK-Lux reporter were cotransfected into HuH-7 cells. Luciferase activity was measured 24 h after incubation with or without 10 ng/ml of TGF-β. (c) Dosage-dependent repression of HCV viral proteins on the transcriptional activation ability of Smad3 complex. HuH-7 cells were transfected with (CAGA)6TK-Lux reporter (1 μg) along with a combination of Smads expression constructs (FLAG-Smad2/Myc-Smad4, Myc-Smad3/Myc-Smad4, or DN-Smad3/Myc-Smad4, 1 μg for each) and increasing amounts of HA-core or FLAG-NS3 expression plasmids (0.5 or 1.0 μg) as indicated. (d) HuH-7 cells were transfected with PAI-1 promoter-containing reporter p800-Lux (1 μg), together with increasing amounts of pcDNA3/HA-core (HA-core) or pEF/NS3 (NS3) (0.25, 1, or 2.0 μg). The total amount of DNA transfected was kept constant in each transfection experiment by adding the empty vector. Luciferase activity was measured 24 h after incubation with or without 10 ng/ml of TGF-β. (e) All experimental conditions were similar to panel d, except that the p21 promoter-containing reporter pWWP-Lux (1 μg) was used. The results are represented as the mean±s.d. from at least three separate experiments and the number above the bar represents the activation fold obtained by comparing the luciferase activity in the presence or absence of TGF-β

Next, we assessed the role of Smad proteins in HCV viral protein-mediated repression of TGF-β-induced transcriptional activation. The reporter activity of the SBE (termed CAGA box)-containing construct, (CAGA)6TK-Lux, in the presence or absence of viral proteins was examined. This CAGA box has been shown to interact specifically with Smad3 and Smad4 (Dennler et al., 1998; Jayaraman and Massague, 2000). As shown in Figure 3b, TGF-β treatment triggered a 4.6-fold enhancement in CAGA box reporter activity in HuH-7 cells transfected with empty vectors (vector). However, a significant dosage-dependent repression (maximum repression 2.6- to 4.7-fold) of CAGA box reporter activity was observed in the presence of increasing amounts of HCV core or NS3 (see Figure 3b, lanes 3–8), suggesting that these two viral proteins have a suppressive effect on transcriptional activation activity of Smad complexes. In addition, HuH-7 cells cotransfected with Smad3/Smad4 elicited a 20-fold increase in CAGA box reporter activity, which is much higher in level than those transfected with Smad2/Smad4 (3.8-fold) or dominant-negative (DN)-Smad3/Smad4 (1.86-fold) (Figure 3c), supporting the role of Smad3 in the induction of CAGA box activation. Interestingly, HuH-7 cells expressing either core or NS3 displayed a strong repression effect (maximum reduction two- to fourfold) on Smad3/Smad4-mediated but not Smad2/Smad4- or DN-Smad3/Smad4-mediated transactivation of (CAGA)6TK-Lux (Figure 3c). Taken together, these results indicate that HCV core and NS3 target TGF-β-induced, Smad3-dependent SBE activation. Also noted, HCV core displays a stronger suppressive effect (about twofold) on the CAGA box reporter as compared to NS3 (see Figure 3b and c).

Experiments were also carried out using two other reporters that contain the natural TGF-β-responsive promoter, the PAI-1 promoter alone (p800-Lux) and the CDK inhibitor p21 promoter (pWWP-Lux). The former is an SBE-dependent reporter and the latter is an SBE-independent one that is activated by Smad3/Sp1 complex through a critical Sp1 site (Datto et al., 1995; Zhang et al., 1998). As shown in Figure 3d and e, HCV core displayed about twofold suppression of both the TGF-β-induced transactivation of p800-Lux and pWWP-Lux. Surprisingly, NS3 had no effect on transactivation of pWWP-Lux in response to TGF-β (Figure 3e), but it did inhibit TGF-β-induced transactivation of p800-Lux to a similar extent as the core protein (Figure 3d). Thus, although core and NS3 both repress TGF-β/Smad3-induced SBE-containing promoter activity, they also display differential repression of TGF-β/Smad3-induced transcriptional activation of a promoter without SBE.

HCV core and NS3 specifically interact with Smad3

Previous studies have demonstrated that Smad proteins associate with viral proteins (Nishihara et al., 1999; Lee et al., 2001; 2002a; 2002b). Thus, it is likely that the repression of TGF-β-mediated transcriptional activity, as observed for HCV core- or NS3-producing cells, is through the physical interaction of these two HCV viral proteins with the Smad proteins. To test this, a GST pull-down assay was performed using 35S-labeled in vitro-translated HA-tagged core or FLAG-tagged NS3 and bacterially expressed GST/Smad proteins. As shown in Figure 4a and b, the GST fusion protein of Smad3 (lane 4), but not Smad2 (lane 3) or Smad4 (lane 5), interacted directly with epitope tagged HCV core or NS3. To further confirm the interaction of HCV core and Smad3 in vivo, a coimmunoprecipitation (co-IP) experiment was performed on HuH-7 cells transiently cotransfected with the expression constructs of HA-tagged HCV core and Myc-tagged Smad3. As shown in Figure 4c, the Myc-Smad3 was co-precipitated by the anti-HA antibody in the presence, but not in the absence, of HA-tagged HCV core (lanes 7 and 8), suggesting that HA-tagged HCV core forms a complex with Myc-tagged Smad3 in vivo. In the case of NS3, the Myc-Smad3 expression construct was transiently transfected into NS3-producing HeLa cells (HeLa/NS3 #38) and anti-Myc antibody was then used for the immunoprecipitation (IP). The HeLa/NS3 #38 cell line is a NS3 highly expressing cell clone (see Figure 2b). To facilitate the detection of complex formation between NS3 and the Smad molecules, we used this cell line (designated HeLa/NS3) for co-IP experiment. The results as shown in Figure 4e indicate that NS3 was co-precipitated by the anti-Myc antibody from the HeLa/NS3 cells in the presence, but not in the absence, of Myc-Smad3 (lanes 7 and 8), suggesting that NS3 interacts with Smad3 in vivo. In contrast, when the Myc-Smad4 expression construct was introduced into HA-tagged HCV core-transfected HuH-7 cells or HeLa/NS3 cells, no significant complex formation of Smad4 with these two viral proteins was found (Figure 4d and f, lane 8). Similarly, NS3 protein from Ava.5 cell lysates could only form immune complex with Myc-Smad3, but not Myc-Smad4 (Figure 4g). It should also be noted that the antibody-conjugated agarose resins used for the assay were unable to cause precipitation with any of the controls (Figure 4c and e, lane 9). This strongly suggests that there is specificity of complex formation between Smad3 and these two viral proteins.

Figure 4
figure4

HCV core and NS3 interact with Smad3 in vitro and in vivo. (a and b) An in vitro binding assay of GST/Smad fusion proteins and in vitro-translated HCV viral proteins. The GST- (lane 2) or GST/Smad protein- (lanes 3–5) bound resins were incubated either with the in vitro-translated [35S]Met-labeled HA-tagged HCV core or FLAG-tagged HCV NS3 (6 μl). After extensive washing, proteins bound on the resins were analysed by SDS–PAGE and autoradiography (bottom panel). Lane 1, input control (2 μl). Coomassie blue staining of the GST/Smad variants used in the binding assay are also shown above. (c and d) In vivo co-IP of HCV core with Smad3 and Smad4. The cells extracts were prepared from HuH-7 cells without transfection (lane 5) or transfected with Myc-tagged Smad constructs (lanes 7 and 8) or/and the HA-tagged HCV core expression construct (lanes 6 and 8) and subjected to IP using anti-HA antibody, followed by immunoblotting with antibodies against Smad3, Smad4 (top panel) or HCV core (bottom panel). (e and f) In vivo co-IP of HCV NS3 with Smads. HeLa/NS3 # 38 cell (HeLa/NS3) and its control HeLa cells (HeLa) were transfected with the Myc-tagged Smads constructs as indicated. IP was performed using anti-Myc antibody (lanes 5–8) and this was followed by immunoblotting with antibodies against HCV NS3 (top panel), Smad3 or Smad4 (bottom panel). The relative expression of HCV viral proteins and Myc-tagged Smads in cell lysates is shown in lanes 1–4 of the panels. The conjugated agarose resins used for the assay were shown as a blank control (c and e, lane 9). (g) HCV subgenomic replicon containing cell (Ava.5) and its control HuH-7 cells (HuH-7) were co-transfected with Myc-tagged Smads constructs as indicated. Immunoprecipitation was performed using anti-NS3 antibody (lanes 3 and 4) and this was followed by immunoblotting with antibodies against HCV NS3 (top panel) or anti-Myc antibody (bottom panel). The relative expression of NS3 protein and Myc-tagged Smads in cell lysates was shown in lanes 1 and 2 of the panels

To gain more insights into the interaction between Smad3 and these two viral products, we defined the interaction domain of Smad3 with the HCV viral proteins. A series of GST/Smad3 deletion mutants were constructed (see Materials and methods) and used for a series of GST pull-down assays of 35S-labeled in vitro-translated HA-tagged core or FLAG-tagged NS3 (Figure 5a and b). As shown in Figure 5c, the truncated mutants of Smad3 that contain either part of (residues 1–70) (lane 2) or the whole region (residues 1–145) (lane 3) of N-terminal MH1 domain retained the ability to interact with the core protein. However, the mutants of Smad3 that consist of linker region (residues 145–215) (lane 5) or MH2 region (residues 215–447) (lane 7) or both regions (residues 145–447) (lane 6) did not bind the core protein, indicating that the MH1 domain harbors the core protein interaction domain. Similarly, these GST/Smad3 deletion constructs were used for pull-down assays of in vitro-translated FLAG-tagged NS3 and results revealed that the interaction domain for NS3 resides in both the MH1 and the MH2 domains (see Figure 5d, lanes 3–5, 7, and 8), but not in the linker region (lane 6), of Smad3. Thus, the direct interacting regions of these two viral proteins with Smad3 are not identical but do overlap within the MH1 domain (see Figure 5a).

Figure 5
figure5

Domain mapping between Smad3 and HCV viral proteins. (a) Schematic representation of the functional domains of Smad3 and the GST/Smad3 variants used for in vitro binding analysis. The locations of MH1 domain (dark gray), linker region, and MH2 domain (light gray) are shown. The numbers mark the amino-acid positions. Also shown on the right panel is the result of the in vitro binding analysis carried out as part of this study. (b) Coomassie blue staining of purified GST/Smad3 variants was used for the binding analysis. (c) In vitro binding assay of GST/Smad3 variants and in vitro-translated HCV core. The GST- (lane 1) and GST/Smad3 variant- (lanes 2–8) bound resins were incubated with in vitro-translated [35S]Met-labeled HA-tagged core (6 μl). Lane 9, input control (2 μl). (d) In vitro binding assay of GST/Smad3 variants and in vitro-translated HCV NS3. All experimental conditions were essentially similar to panel c, except that an in vitro-translated [35S]Met-labeled FLAG-tagged NS3 was used for in vitro binding analysis

HCV core and NS3 do not block the nuclear translocation of Smad3 in response to TGF-β

The Smad-mediated transactivation in response to TGF-β requires phosphorylation and nuclear translocation of the Smads (Zhang et al., 1996). To understand the molecular basis of the repression of TGF-β-mediated transcriptional activation in HCV viral protein-producing cells, we examined whether these viral proteins altered nuclear translocation of the Smad proteins. As only Smad3 interacts with the HCV viral proteins (see Figure 4), it was only necessary to test the nuclear translocation of this signaling mediator. HCV core-producing HuH-7 cells (HuH-7/C190) and NS3-producing HeLa cells (HeLa/NS3) and their parental cells were transfected with the Myc-tagged Smad3 expression construct and the subcellular localization of this epitope-tagged Smad3 in these cell lines before and after TGF-β treatment was detected by immunoblotting. As shown in Figure 6a, following stimulation by 10 ng/ml of TGF-β for 30 min, the amount of Myc-Smad3 in nuclear fraction slightly increased in the HuH-7 control cells. The level of nuclear translocation of Myc-Smad3 in HuH-7/C190 cells was comparable to the control cells, indicating that HCV core did not perturb the nuclear translocation of Smad3 in response to the TGF-β signal. Similar results were obtained for the NS3-producing HeLa/NS3 (Figure 6b) and Ava.5 cells (data not shown). Quantification of the level of Myc-Smad3 in the nuclear (N) and cytosolic (C) fraction (displayed as an N/C ratio) also suggests that there is no viral protein effect on the nuclear translocation of Myc-Smad3 triggered by TGF-β (Figure 6c). This conclusion was further supported by the immunofluorescence staining experiments on Myc-tagged Smad3 expression construct-transfected cells. As shown in Figure 6d and e, while the nuclear translocation of Myc-Smad3 was detected after 30 min of TGF-β treatment, examination of the subcellular distribution of Myc-Smad3 by immunofluorescence staining revealed no significant difference in the percentage of cells displaying the Smad3 nuclear translocation (around 45%) among these TGF-β-treated viral protein-expressing cells and their control cells. Notably, in the absence of TGF-β treatment, both HCV core and NS3 were found in nuclear and cytoplasmic compartments. However, a predominantly nuclear localization of these two viral proteins was observed in the presence of TGF-β (see Figure 6d and e; core and NS3 in the left panel). An enhanced nuclear translocation of NS3 after TGF-β triggering was also noted (Figure 6b, compare lane 7 with lane 8). This implies that complex formation with Smad3 may facilitate the nuclear translocation of these viral proteins. All together, these results suggest that the inhibition of TGF-β-induced transactivation in viral protein-producing cells could not be attributed to a blockage of the nuclear translocation of Smad3.

Figure 6
figure6

HCV core and NS3 do not interfere with the nuclear translocation of Smad3. (a) Analysis of Smad3 nuclear translocation in HCV core-producing cells. HuH-7/C190 and control HuH-7 cells were transiently transfected with the Myc-tagged Smad3 expression plasmid (1 μg). The cytoplasmic fraction (Cyto) (80 μg) and the nuclear fraction (NE) (40 μg) were prepared from transfected cells after being treated with (+) or without (−) 10 ng/ml TGF-β for 30 min. These fractions were analysed by Western blotting using the specific antibodies as indicated. (b) Analysis of Smad3 nuclear translocation in HCV NS3-producing cells. All experimental conditions are similar to panel a, except that HeLa/NS3 #38 (HeLa/NS3) and control HeLa cells were used for the experiment. (c) The N/C ratio of nuclear Myc-Smad3 protein (N) to cytoplasmic Myc-Smad3 protein (c) in cells treated with or without 10 ng/ml of TGF-β. The amount of Myc-Smad3 was determined from the intensity of the specific immunoblot band for Myc-Smad3 using Densitometer™ SI. Data represent the mean±s.d. from three separate experiments. (d and e) Subcellular localization and nuclear translocation of Smad3 were analysed using indirect immunofluorescence. HuH-7/C190 (d), HeLa/NS3 #38 (e) and their control cells grown on coverslips were transfected with Myc-tagged Smad3 expression plasmid (1 μg). After treating with (bottom panel) or without (top panel) 10 ng/ml of TGF-β for 30 min, HCV core (d, green), NS3 (e, green), and Smad3 (d and e, red) were detected in the cells by immunofluorescence analysis. Nuclei were counterstained with DAPI (blue)

HCV core and NS3 differentially block the Smad3/DNA complex formation in vitro and in vivo

When it is considered that both the NS3 and core interact with the N-terminal MH1 region of Smad3, a DNA-binding domain for Smad3, it is likely that these two viral proteins may be able to modulate the DNA-binding activity of Smad3. A gel shift assay was performed using purified GST fusion proteins and a 32P-labeled oligonucleotide probe containing three Smad3/Smad4 binding elements (see Materials and methods). Since the autoinhibitory interaction between region MH1 and MH2 of Smad3 blocks the DNA-binding activity of MH1 domain (Hata et al., 1997; Jayaraman and Massague, 2000), we determined the effect of the HCV viral protein specifically on the DNA-binding activity of GST/Smad3(1–145), which contains the N-terminal 1–145 amino acids of Smad3 but lacks its linker and MH2 regions. The gel shift assay revealed that both GST/Smad3(1–145) and GST/Smad4 bound to this probe specifically, because the retarded bands could be competed out by an excess (50-fold) of unlabeled wild-type (WT) probe but not by a mutant probe (compare lanes 2, 9, and 10 in Figure 7a–d). The existence of more than one kind of Smad4/SBE complexes in this assay (Figure 7b and d) could be relevant to the fact that Smad4 is capable of forming homo-oligomeric complexes (Hata et al., 1997; Pardali et al., 2000). However, in the presence of increasing amounts of the GST/core (GST/C195) or GST/NS3 fusion protein but not GST protein, the DNA-binding activity of GST/Smad3(1–145) protein was diminished (see Figure 7a and c, lanes 5–8), whereas under the same conditions, the DNA-binding activity of GST/Smad4 protein was not affected by these two viral proteins (Figure 7b and d, lanes 5–8). This result is in accord with the earlier finding that these two viral proteins are associated with Smad3 but not Smad4 (see Figure 4a and b). This suggests that the association of Smad3 with core or NS3 may interfere in complex formation with specific SBE element in vitro.

Figure 7
figure7

HCV core and NS3 differentially inhibit Smad3/DNA complex formation. (a and b) The effect of HCV core on Smads/DNA complex formation. EMSAs were performed using 32P-labeled SBE probe. Purified GST/Smad3(1−145) (a) or GST/Smad4 (b) in the presence of GST or GST-C195 fusion protein was incubated with 32P-labeled SBE probe. Complexes were resolved using 6% native polyacrylamide gel electrophoresis. Competition experiments were carried out by incubating unlabeled oligonucleotides with MT (50-fold excess) or WT (50-fold excess) SBE. Lane 1, SBE probe. (c and d) The effect of NS3 on Smads/DNA complex formation. The effects of GST-NS3 protein on the DNA-binding ability of Smad3 (c) or Smad4 (d) were examined as described above. Lane 1, SBE probe. (e) EMSA was preformed with nuclear extracts (10 μg) prepared form TGF-β-treated or -untreated cells as indicated. All experimental conditions are similar to panels a–d, except 0.15 μg of specific anti-Smad3 antibody was used in the supershift assay (lane 8). (f) The 32P-labeled p21 probe was used in EMSA together with nuclear extracts (10 μg) prepared from TGF-β-treated cells as indicated. Specific antibody (0.15 μg) against Sp1 (lane 5), Smad3 (lane 6), or Sp3 (lane 7) was used in the supershift assay. Competition experiments were carried out by incubating an excess amount of unlabeled MT or WT p21 (−120/−60) oligonucleotides

A gel shift assay was also performed with the SBE probe and nuclear extracts prepared from both TGF-β-treated and -untreated cells. As shown in Figure 7e, at least six retarded bands of Smad/SBE complex were present in HuH-7 cells (lane 2) and four out of them were TGF-β inducible (lane 5). Moreover, these four retarded bands were supershifted in the presence of the Smad3-specific antibody (lane 8), but diminished in the presence of an excess of unlabeled WT probe but not by a mutant probe (compare lanes 5, 9, and 10 in Figure 7e). This suggests that these Smad/SBE complexes in HuH-7 nuclear extracts are Smad3 specific. It should also be noted that these four Smad3/SBE complexes were disrupted by both HCV core and NS3 after the TGF-β treatment (compare lanes 6 and 7 with lane 5). However, without TGF-β treatment, the inhibition of Smad3/SBE complexes formation was evident only in the presence of HCV core but not NS3 (compare lanes 3 and 4 with lane 2). Therefore, it appears that HCV core represses in vivo Smad3/SBE complex formation regardless of whether the TGF-β signal is activated, while the NS3 interferes with the in vivo Smad3/SBE complex formation only in the presence of TGF-β.

When it is considered that these two viral proteins display differential inhibitory effect on TGF-β induced p21 promoter activity (Figure 3e), it is pertinent to know whether these effects were resulted from the differential deregulation of Smad3/Sp1 binding to the p21 promoter. Similar experiments were preformed using a 32P-labeled oligonucleotide corresponding to the TGF-β-responsive element in the proximal p21 promoter (nucleotide positions −120 to −60, designated p21 probe) (Figure 7f). At least seven retarded bands of protein–DNA complex were revealed in nuclear extracts prepared from TGF-β-treated HuH-7 cells (lane 2). The two slowest migrating bands correspond to the Smad3/Sp1 complexes since they were supershifted in the presence of the Sp1- or the Smad3-specific antibody but not by the Sp3 antibody (compare lanes 5–7). Moreover, these two retarded bands were diminished in the presence of an excess amount of unlabeled WT probe but not by mutant probe harboring point mutations in the consensus Sp1 sequence (at nucleotide −83 to −77; see Materials and methods) (compare lanes 8 and 9 with lane 2), further supporting the idea that these complexes are Sp1- and Smad3-specific. Most interestingly, binding of the p21 probe to the protein complexes including Smad3/Sp1 was strongly repressed by HCV core (lane 3), while NS3 showed essentially no difference as compared to the control (compare lane 4 with lane 2). Therefore, HCV core, but not NS3, downregulates Smad3/Sp1–DNA complex formation on the proximal p21 promoter in response to TGF-β, which is in agreement with finding a differential regulation of pWWP-Lux reporter activity (see Figure 3e).

HCV core and NS3 have differential effects on the TGF-β-induced G1 phase arrest

TGF-β plays a key role in a variety of biological processes including cell cycle progression and apoptosis (Massague, 2000). As noted above, there are variations between NS3 and core in their transcriptional modulation on TGF-β target promoters (see Figure 3) and these may lead to differential effects on TGF-β-induced biological responses. To explore this idea, we examined whether HCV core or NS3 could affect the antiproliferative effect of TGF-β in vivo. To this end, the DNA content of HuH7/C190, HuH-7/NS3, and the control HuH-7 cells at various time points post-treatment with TGF-β were analysed by flow cytometry (see Materials and methods). As shown in Figure 8a, following stimulation with 10 ng/ml of TGF-β, the growth of the control HuH-7 cell lines was effectively inhibited and arrested in G1 phase. In contrast, HCV core prevented the cells from undergoing this growth arrest after exposure to TGF-β. Specifically, the percentage of HuH-7/C190 arrested at G1 phase was significantly lower than that for the control cells 24–48 h after TGF-β treatment (10–25% reduction) (Figure 8a). In the case of NS3, the TGF-β-induced G1 phase arrest observed in HuH-7/NS3 cells at 12–48 h showed essentially no difference when compared to the control cells (HuH-7/Vector, Figure 8b), a finding consistent with the lack of interference in TGF-β-induced transactivation of pWWP-Lux in HuH-7/NS3 cell lines (Figure 3e). In addition, we also examined whether the subgenomic replicon Ava.5 cells have any influence on TGF-β-induced G1 phase arrest. Prior to the TGF-β treatment, Ava.5 cells had a higher proportion of cells in G1 phase than the parental HuH-7 cells (60% for Ava.5 versus 50% for HuH-7 control in Figure 8c), suggesting the deregulation of the cell cycle in Ava.5 cells. Nevertheless, following stimulation with TGF-β, the growth of Ava.5 and the control HuH-7 cells were both effectively inhibited and arrested in G1 phase (20% enhancement in G1 phase for each). In this regard, there is essentially no difference between Ava.5 cells and the parental cells with respect to TGF-β-induced G1 phase arrest. Taken together, we found that HCV core, but not NS3 or the subgenomic replicon, has an ability to cause resistance to TGF-β-induced G1 phase arrest.

Figure 8
figure8

HCV NS3 and core exert differential regulation on the TGF-β-mediated G1 phase arrest. (a and b) HCV core (HuH-7/C190)-, NS3 (HuH-7/NS3)-producing cells and their control HuH-7 cells (HuH-7 or HuH-7/Vector) were stimulated with 10 ng/ml TGF-β for various durations (0–48 h) and then stained with propidium iodide. The DNA content was analysed by flow cytometry as described in ‘Materials and methods’. Results were expressed as the percentage of cells in G1 phase. Data represent the mean±s.d. from at least three separate experiments. (c) All experimental conditions were essentially similar to (a and b), except that HCV subgenomic replicon Ava.5 cells were used for TGF-β-induced G1 phase arrest analysis

Both HCV core and NS3 resist the TGF-β/Smad3-mediated apoptosis

As the HCV core and NS3 differentially regulated the TGF-β-induced G1 arrest of cell cycle (Figure 8a and b), we further examined whether they also confer different effects on TGF-β-induced apoptosis. Under serum-free conditions, the control HuH-7 cells did not undergo an induction of apoptosis unless they were treated with a high concentration of TGF-β (25–100 ng/ml; data not shown). We also investigated whether Smad3 is involved in TGF-β-induced apoptosis using a dual staining method (see Materials and methods). Our results, as shown in Figure 9a, indicate that overexpression of the Myc-Smad3 increased the sensitivity of HuH-7 to TGF-β-induced apoptosis, whereas overexpression of DN-Smad3 enhanced resistance of HuH-7 cells to this response. Specifically, the amount of TGF-β-induced apoptotic cells in Myc-Smad3 plasmid-transfected sample was twice as much as that in the control sample, while in the DN-Smad3 plasmid-transfected sample, the amount of apoptotic cells was only half of the control (Figure 9a, left and right panels). The proportion of apoptotic cells expressing the DN-Smad3 was much lower than those expressing the Myc-Smad3 in response to TGF-β (Figure 9a, middle panel). These findings suggest that Smad3-mediated pathways play a role in TGF-β-induced cell death of HuH-7 cells.

Figure 9
figure9figure9

HCV viral proteins downregulate TGF-β/Smad3-induced apoptosis. (a) The Smad3-mediated apoptosis was assessed by a dual staining analysis in HuH-7 cells transfected with 1 μg of vector alone, Myc-Smad3, or DN-Smad3 construct. After incubating in serum-free medium containing 100 ng/ml TGF-β for 24 h, the transfected cells were fixed and stained with Smad3 antibody for intracellular Myc-Smad3 or DN-Smad3, and then subjected to TUNEL assay for apoptotic cell determination as described in Materials and methods. The stained cells were analysed using flow cytometry. The overlaid TUNEL staining histogram is shown on the left, the 2-D plot plotted by the populations of Myc-Smad3- or DN-Smad3-expressing cells (y-axis) versus the apoptotic cells (x-axis) is shown in the middle, and the quantitative results for the proportion of TUNEL staining-positive cells are shown on the right. The relative expression levels of Myc-Smad3 and DN-Smad3 protein in the transfected cells were visualized by Western blot using antibody against Smad3 and is shown at the bottom. (b and c) HCV viral protein-producing HuH-7 cells and their control HuH-7 cells were incubated in serum-free medium containing 100 ng/ml of TGF-β for various durations (0–48 h). The treated cells were stained with propidium iodide and the DNA content was analysed by flow cytometry. Results are expressed as the percentage of cells in sub-Go phase, and DNA histograms are shown on the right. Data represent the mean±s.d. from at least three separate experiments. (d) All experimental conditions were essentially similar to (b and c), except that HCV subgenomic replicon Ava.5 cells were used for the TGF-β-induced cell death analysis

We next tested the level of TGF-β-induced cell death in HCV core- and NS3-producing HuH-7 cell lines by flow cytometry (Figure 9b and c). In the control HuH-7 cells, apoptotic cell death was induced by incubation with 100 ng/ml of TGF-β in serum-free medium at 24 h (Figure 9b and c, left panels). Their DNA histograms after treatment with TGF-β for 36 and 48 h showed a typical sub-G0 or shoulder shape that represents the apoptotic cells (Figure 9b and c, right panels). However, a marked resistance to TGF-β-induced apoptosis was noted in both of the two HCV viral protein-producing cell lines as compared to the control HuH-7 cells at 36 h (5–7.5% apoptotic cells for HuH-7/C190 and HuH-7/NS3 cells versus 25–27.5% for the control cells) (Figure 9b and c). After 48 h of TGF-β treatment, the HuH-7/C190 cells still retained a strong inhibition against TGF-β-induced apoptosis (7.5% apoptotic cells for HuH-7/C190 versus 35% for the control cells). This is not the case for HuH-7/NS3 cells, since in these 48 h-treated cells the apoptotic cells had reached 25% of the population (Figure 9c). In the case of the subgenomic replicon Ava.5 cells, after treating with 100 ng/ml of TGF-β in serum-free medium for 48 h, a marked resistance to apoptosis was also noted compared to the parental HuH-7 cells (Figure 9d, 13% for Ava.5 versus 32% for parental HuH-7 cell). All together, our results indicate that both core and NS3 have an ability to block TGF-β/Smad3-mediated apoptosis, and the HCV subgenomic replicon exerts influence on TGF-β-induced apoptosis but not on G1 phase arrest, which is consistent with the findings obtained for NS3-producing cells (Figures 8b and 9c).

Discussion

In this study, we have demonstrated the inhibitory effect of the HCV full-length polyprotein, the oncogenic core and NS3 viral proteins on the TGF-β-mediated transcriptional activation (see Figures 1 and 2). Resistance to TGF-β-mediated apoptosis was also observed in the core- or NS3-producing cell lines, however, only the core protein represses the TGF-β-induced G1 phase arrest (see Figures 8a, b and 9b, c). Interestingly, the HCV subgenomic replicon attenuated TGF-β-induced apoptosis but not G1 phase arrest (see Figures 8c and 9d), a property reflecting the characteristic effects of NS3 on TGF-β resistance. The suppressive effect of HCV viral proteins on the TGF-β-responsive reporter is directed to the specific Smad-binding CAGA box (SBE) in a Smad3-dependent manner (see Figure 3). Physical interaction between Smad3 and these two viral proteins was detected using both in vitro and in vivo approaches (see Figure 4). Additionally, fine mapping of the interaction domain revealed that the HCV core associates with the MH1 region of Smad3, while NS3 interacts with both the MH1 and the MH2 regions of the same Smad molecule (see Figure 5). This core or NS3 association with Smad3 was able to reduce the DNA-binding ability of Smad3 (see Figure 7) but did not lead to its nuclear exclusion in response to TGF-β (see Figure 6). Thus, the binding of these two proteins with Smad3 and the interference with its SBE binding ability is apparently a potential mechanism whereby HCV core and NS3 can perturb the TGF-β/Smad3-mediated transcriptional activation of target genes. Nevertheless, since in this study all the biological effects on TGF-β response were examined in different cell lines, the possibility of the involvement of unknown host factors in the HCV viral protein-mediated differential effects cannot be completely excluded out.

Smads, a small family of structurally related proteins, are central to most activities of the TGF-β family. Given that a slight difference in the structure of the Smads is sufficient to change their recognition specificity to the receptors or associated partners (Massague, 2000), it is likely that the binding specificity of HCV core and NS3 on Smad molecules is also dictated by one or more discrete structural elements in each Smad. Consistent with this, we found that Smad3, but not the structurally related Smad2 or Smad4, can associate with HCV core and NS3 (see Figure 4). This result, together with the finding that these two viral proteins selectively inhibit complex formation of Smad3/SBE but not Smad4/SBE (see Figure 7a–d), strongly supports the finding that these two viral proteins specifically target Smad3-mediated transcriptional activity (see Figure 3c).

In this work, we did observe several common features conferred on the TGF-β-induced responses by HCV core and NS3, but certain different effects elicited by these two viral proteins were also present. The MH2 domain of Smad3, which is responsible for interaction with various transcription factors, is an interaction domain for NS3 but not core (see Figure 5). This discrepancy may underlie the functional differences between Smad3 complexes formed with these two viral proteins. Along this line, HCV core acts as a stronger repressor of the CAGA box reporter (see Figure 3b and c) and has a marked blocking effect on in vivo Smad3/SBE interaction (see Figure 7e) than NS3. Additionally, HCV NS3 interferes with the in vivo Smad3/SBE complex formation only in the presence of TGF-β, while core represses in vivo Smad3/SBE complex formation regardless of whether the TGF-β signal is activated or not (see Figure 7e). Furthermore, we found that HCV core suppresses TGF-β-induced PAI-1 and p21 promoter activity (see Figure 3d and e). However, NS3 did not affect the TGF-β-induced p21 promoter activity (see Figure 3e). The diversity of the TGF-β/Smad3 targeted promoters leads to the requirement of different mechanisms for these viral proteins to repress the TGF-β/Smad3-induced transactivation. This is due to the fact that these promoters do not necessarily contain Smad binding elements. The PAI-1 promoter contains three SBEs and is regulated mainly by Smads in an SBE binding-dependent manner (Jayaraman and Massague, 2000). Conceivably, by suppressing the formation of TGF-β-inducible Smad3/SBE complex (see Figure 7e), both viral proteins could deregulate TGF-β-induced PAI-1 promoter activity (see Figure 3d). In contrast to the PAI-1 promoter, the p21 promoter contains a critical Sp1 binding sequence, but not an SBE within its defined TGF-β-responsive element. In this case, Smad3 activates p21 promoter by its ability to increase the binding of Sp1 protein to this specific Sp1 site (Zhang et al., 1998). In line with these, our results indicate that following TGF-β stimulation, the core protein, but not NS3, represses the binding of the Smad3/Sp1 complex to this specific Sp1 site (Figure 7f), which is consistent with the previous results showing that, with or without TGF-β treatment, the HCV core protein represses the p21 promoter via the Sp1 site (Ray et al., 1998; Dubordeau et al., 2002; Lee et al., 2002c). Thus, the core protein, but not NS3, may interfere with TGF-β-induced p21 promoter activity by disrupting Smad3/Sp1 complex formation or by blocking the DNA-binding affinity of this complex. It should also be noted that the influences of these two viral proteins on TGF-β-regulated target genes are closely correlated with their modulating abilities on the TGF-β-induced biological responses. Since TGF-β causes growth arrest at the G1 phase of the cell cycle, at least in part by upregulating the cyclin-dependent kinase inhibitor p21 (Datto et al., 1995), the observed resistance to TGF-β-induced growth arrest in HuH-7/C190 cells (Figure 8a) thus correlates with the finding of core-mediated repression of TGF-β-induced p21 promoter activation (see Figure 3e). Likewise, the lack of repression of TGF-β-induced p21 promoter activation by NS3 is also in accord with the fact that NS3 is unable to overcome TGF-β-induced growth arrest (see Figures 3e and 8b). However, when considering that the amounts of viral products in natural HCV-infected cells are likely different from the in vitro culturing experiment, these viral protein-mediated differential effects on TGF-β response as noted above may not be similar to the clinical status.

There are precedents for viral proteins that modulate the TGF-β signal pathway, suggesting that the TGF-β pathway is a favored target of viral proteins. Several viral oncogenic proteins including adenovirus E1A, HPV E6 and E7, HBV HBx, HTLV Tax, and EBV LMP1 influence the TGF-β-induced response through either binding with Smads or modulating other TGF-β signaling-related mechanisms. These viral proteins are similar in several ways to HCV core in their oncogenic potency, their alterations to signal transduction and their effect on transcriptional activity. Interestingly, viral proteins can be involved at various regulatory points on the TGF-β signaling pathway. In line with this, the common strategies used by viruses in order to interfere with TGF-β signaling include altering the promoter activity of TGF-β or its receptors genes. E6 or E1A seems to be a case (Kim et al., 1996; Dey et al., 1997). However, this mechanism does not fit the results for HCV core or NS3 because there was no significant difference in the expression level of TGF-β type II receptor in either HCV core- or NS3-producing cells as compared with the control cells (data not shown). An interaction between viral proteins and Smads has also been reported. Examples of this group include the E7, E1A, HBx and Tax (Nishihara et al., 1999; Lee et al., 2001; 2002a; 2002b), and HCV core and NS3 reported in this work (see Figure 4). These viral proteins impair TGF-β signaling through direct binding to Smad proteins and blockage of either the DNA-binding or protein-binding properties of Smads. In addition, cooperating with survival pathways is another possible strategy. Evidence supporting this notion has been reported in various previous studies and includes constitutive phosphorylation of ERK 1/2 and activation of NF-κB by LMP1 and the activation of PI3-kinase signaling by HBx, thereby causing a resistance to TGF-β-mediated growth inhibition (Shih et al., 2000; Prokova et al., 2001). This may also be the case for HCV core or NS3 because the repression of TGF-β signaling is not the only example of HCV viral proteins being involved in the regulation of signaling pathways (Tellinghuisen and Rice, 2002). Some of these signals play critical roles in controlling cell survival or in counteractive action following TGF-β stimulation. In HuH-7 cells, we found that TGF-β-induced cell death is modulated by Smad proteins, suggesting that Smad-mediated pathways do play a role in the apoptosis response (see Figure 9a). However, whether the HCV viral protein inhibits TGF-β induced apoptosis through deregulating Smad-dependent transactivation of proapoptotic gene still awaits further study. Nevertheless, it should be noted that the cooperation of viral proteins with survival signaling pathways and their interference with the intrinsic properties of Smads may contribute to some degree or other to counteract the TGF-β-induced antiproliferation response.

TGF-β is one of the most potent regulators of tumorigenesis. Recently, several studies have supported a biphasic role for TGF-β in human cancer (Derynck et al., 2001). In the initial stage of tumorigenesis, TGF-β acts as a tumor suppressor by inducing growth arrest or promoting apoptosis. Transformative cells often lose their TGF-β-mediated antiproliferation responses as a result of functional inactivation of TGF-β receptors or Smads (Derynck et al., 2001). In the later stages of various cancers, an increased production of TGF-β has been associated with invasiveness and metastasis due to its carcinogenic function, which includes stimulation of angiogenesis, production of cell-adhesion proteins and extracellular-matrix proteins. Interestingly, elevated levels of plasma TGF-β are often detected in HCV-infected patients who have either chronic liver disease or HCC, but it is rare in HBV-related or non-HCV-related patients (Ito et al., 1991; Matsuzaki et al., 2000; Ray et al., 2003). Presumably therefore, TGF-β is involved in HCV pathogenesis including the occurrence of cirrhosis and HCC development.

In this work, we demonstrate that core and NS3 are two crucial HCV viral products that are responsible for the inhibition of the TGF-β-induced responses, albeit a distinction between these two viral protein-mediated effects also exists. Interference with Smad3 DNA binding is likely to be one of the potential mechanisms that allow these two HCV viral proteins to inhibit the transcriptional activation of TGF-β/Smad3-mediated target genes containing SBEs. The ability to act against either TGF-β-induced growth arrest or apoptosis was also revealed for the core- or NS3-producing cells. Therefore, we would like to suggest that perturbations of sensitivity to TGF-β by the HCV oncoproteins, in combination with enhanced TGF-β production in HCV infection, may contribute to HCV-associated hepatocarcinogenesis.

Materials and methods

Plasmids

The HCV core expression construct pcDNA3/HA core that produces the HA-tagged full-length HCV core was constructed by insertion of a 593 bp PCR-generated EcoRI–XhoI fragment of the HCV core gene derived from plasmid pECE/HCVC-KF (Shih et al., 1993) into a EcoRI/XhoI-digested pcDNA3-HA vector (Invitrogen). A series of HA- or FLAG-tagged HCV viral protein expression plasmids was constructed by insertion of PCR-generated EcoRI–XhoI fragments containing the E1 (576 bp), E2 (1.1 kb), p7 (189 bp), NS2 (651 bp), NS3 (1.9 kb), NS3/4A (2.0 kb), NS4B (783 bp), NS5A (1.3 kb), or NS5B (1.8 kb) gene fragments derived from p90/HCVFLlongpU (kindly provided by Apath, USA) (Kolykhalov et al., 1997) into either the pcDNA3-HA or pcDNA3-FLAG vector using EcoRI/XhoI sites. The HCV NS3 expression construct pEF/NS3, which can direct the expression of the full-length NS3 protein (68 kDa), was obtained by insertion of a HindIII–XhoI (filled in) fragment derived from pGEM4/Flag-NS34A (provided by Dr L-H Hwang, National Taiwan University, Taiwan) into EcoRI/XbaI- (filled in) digested p/3EFpro vector (kindly provided by MD Tokushige K, Tokyo Women's Medical College, Japan). Reporter plasmid pAP-1-Lux containing seven copies of the AP-1 site in the regulatory region was purchased from Stratagene. The plasmid (CAGA)6TK-Lux reporter contains two copies of the CAGA box primers (5′-IndexTermCAGCCAGACAAAAAGCCAGACATTTAGCCAGACAGGTAC-3′) inserted at the KpnI site upstream of HSV1 thymidine kinase (TK) promoter (nucleotides −105 to +32 related to transcription start site) of a pGL2 basic plasmid (Promega) derivative that has the TK promoter fragment insertion at the HindIII/SmaI site. The pGST/NS3 construct that directs the expression of NS3 protein fused with the C-terminus of GST was generated by direct cloning of a HindIII–XhoI fragment derived from pEF/NS3 into HindIII/XhoI-digested pGEX-5X-3 (Promega). pET-21a/NS3-385, which expresses the His-tagged N-terminal 385 amino acids of the NS3 protein under IPTG induction, was constructed by excising a 0.7 kb HindIII–EcoRI fragment from pET-21a/NS3-3360 (kindly provided by Dr L-H Hwang, National Taiwan University, Taiwan). This was then ligated via a 10-mer HindIII linker through blunt end ligation. The plasmid pGST/HCVc195 is a derivative of pGEX-3KS and directs the synthesis of the full-length of the HCV core protein fused to the C-terminus of GST (Chen et al., 2003). The HCV full-length polyprotein expression plasmid pSRα/HCV-FL has been described previously (Chen et al., 2003). The TGF-β-responsive reporter constructs, p3TP-Lux (Wrana et al., 1992) and p800-Lux (Abe et al., 1994), were provided by Dr R-H Chen (National Taiwan University, Taiwan) and Dr C-K Chou (National Yang-Ming University, Taiwan), respectively. The plasmid pWWP-Lux, which covers the −2325/+8 region of p21 promoter, was obtained from Dr K-H Lan (Taipei Veterans General Hospital, Taiwan) (El-Deiry et al., 1993). The Smad protein expression constructs, pcDNA3/Flag-Smad2, pcDNA3/Myc-Smad3, pcDNA3/Myc-Smad4, and pRK/Smad3 ΔCF, were provided by Dr K Miyazono (The Cancer Institute of the Japanese Foundation for Cancer Research and Research for the Future Program, Japan) (Zhang et al., 1996; Nakao et al., 1997). The plasmids pGST/Smad2, pGST/Smad3, and pGST/Smad4 that direct the expression of the various Smad proteins fused with the C-terminus of GST were constructed by subcloning the 1.4 kb, 1.3 kb, or 1.6 kb EcoRI–XbaI fragments of the Smad region from the respective pcDNA3/Flag-Smad2, pcDNA3/Myc-Smad3, and pcDNA3/Myc-Smad4 constructs into EcoRI/XbaI-digested pGEX-5X-1 (Promega). The pGST/Smad3 serial deletion constructs pGST/Smad3(1–70), pGST/Smad3(1–145), pGST/Smad3(1–215), pGST/Smad3(1–447), pGST/Smad3(145–215), pGST/Smad3(145–447), and pGST/Smad3(215–447) were generated by insertion of a PCR-generated EcoRI–NotI fragment of the indicated Smad3 coding region into EcoRI/NotI-digested pGEX-5X-1 (Promega).

Cells and transfection

HCV core-producing human hepatoma cell line HuH-7/C190 and its parental cell line HuH-7 were cultured as described previously (Chen et al., 1997). To establish the full-length NS3-expressing stable cell lines, HuH-7 cells were transfected with a plasmid encoding the full-length NS3 (pEF/NS3) by electroporation. Geneticin (G418; Sigma) at 1 mg/ml was added to the culture medium at 48 h post-transfection, and the NS3-expressing cell line HuH-7/NS3 was then selected after 2 weeks. A control HuH-7/Vector cell line was also established after transfection with the p/3EFpro vector. To establish the NS3-expressing stable HeLa cell line HeLa/NS3, a BamHI–KpnI fragment (2 kb) of NS3-4A coding region was excised from pGEM4/Flag-NS34A, filled in, and cloned into the filled EcoRI site of the pTRE vector from the Tet-Off ™ gene expression system (Clontech). The resulting plasmid, pTRE/NS34A, was cotransfected with pTK-Hyg into Tet-Off control HeLa cells and the NS3 cell line was selected as specified by the manufacturer. Two of the stable transfectants, designed NS3 #4 and NS3 #38, expressed the HCV NS3 protein with different level (see Figure 2b) and these were chosen for this study. Ava.5 cells, the HuH-7-derived cell lines harboring the autonomously replicating HCV replicon covering the NS3 to NS5B region, were provided by CM Rice and Apath (St Louis, MO, USA) (Blight et al., 2000). The expression of core and NS3 in these cell lines was confirmed by Western blotting (see Figures 2b and 4g). The anti-HCV core antiserum has been described previously (Chen et al., 1997). Rabbit anti-NS3 antiserum was prepared using Ni2+ affinity column-purified His-NS3-385 protein as the antigen, which was expressed from the expression construct pET-21a/NS3-385. Ava.5 was maintained in medium plus 1 mg/ml of G418. Transfection was performed by the calcium phosphate co-precipitation method (Graham and van der Eb, 1973) or by using FuGENE 6 transfection reagent (Roche Molecular Biochemicals). After transfection with the indicated plasmids, cells were treated with 10 ng/ml TGF-β (R&D Systems) for 24 h and the luciferase assay was performed as described previously (You et al., 1999).

Analysis of Smad3 translocation

To analyse translocation of Smad3, cells transfected with Myc-Smad3 expression construct (pcDNA3/Myc-Smad3) were treated with 10 ng/ml of TGF-β for 30 min, and then harvested to allow the preparation of cytosolic and nuclear extracts. Subcellular fractionation was performed as described previously (You et al., 1999). Aliquots of the cell extracts were analysed by SDS–PAGE and immunoblotting with the indicated antibodies using the ECL system (Pierce). Anti-Smad3, anti-Smad4, and anti-Myc (9E10) antibodies were purchased from Upstate Biotechnology, and the control antibodies, anti-α-tubulin, and anti-B23 were purchased from Santa Cruz Biotechnology.

Immunofluorescence

The localization of Myc-tagged Smad3 in transfected cells was determined by immunofluorescence microscopy. Cells grown on coverslips were transfected with 1 μg of the Myc-Smad3 expression construct (pcDNA3/Myc-Smad3). At 24 h after transfection, the cells were treated with 10 ng/ml of TGF-β for 30 min, and then fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 1 h at room temperature. Cells were washed with PBS three times and permeabilized with 0.5% Triton X-100 in PBS for 15 min. The primary antibodies used for immunofluorescence were rabbit anti-HCV core antiserum (1 : 500), rabbit anti-NS3 antiserum (1 : 1000), and mouse anti-Myc (1 : 1000; Upstate Biotechnology). All secondary antibodies were conjugated with fluorescein isothiocyanate or rhodamine (The Jackson Laboratories) and used at a dilution of 1 : 1000. During the incubation period for the secondary antibodies, 4′,6-diamidino-2-phenylindole (DAPI) were used simultaneously for nuclei staining. Cells were visualized using a fluorescence microscope.

Expression and purification of GST fusion proteins and the in vitro pull-down assay

GST, GST/C195, GST/NS3, or GST/Smads fusion proteins were expressed in DH5α bacterial cells. Purification of the GST fusion proteins was performed as described previously (Shih et al., 1993). For the GST pull-down assay, 20 μl of glutathione–sepharose 4B beads (Pharmacia) bound to GST or GST/Smad fusion proteins (4 μg) were incubated with in vitro-translated 35S-labeled HCV core or NS3 mixtures at 4°C for overnight. The beads were washed four times with 1 ml of PBS. Proteins bound to the beads were eluted with sample buffer, separated by SDS—PAGE, and then detected by autoradiography.

In vivo co-IP

HuH-7 cells were cotransfected with the HA-tagged HCV core (pcDNA3/HA-core) or Myc-tagged Smad3/Smad-4 (pcDNA3/Myc-Smad3 or pcDNA3/Myc-Smad4) expression plasmids. Alternatively, HeLa/NS3 or its control cells were transfected with Myc-Smad3 or Myc-Smad4 expression construct, and Ava.5 cell or its control cells were cotransfected with Myc-Smad3 and Myc-Smad4 expression constructs. After 48 h, the cells were washed twice with ice-cold PBS and collected by centrifugation. Whole-cell extracts were prepared in PBS containing 0.5% NP-40 and 1 × protease inhibitor mixture (Complete™, Roche Molecular Biochemicals). The extracts were cleared by centrifugation and then the supernatants were incubated overnight at 4°C with anti-HA, anti-NS3 antibody, or anti-Myc-conjugated Protein G Sepharose™ (Amersham Pharmacia Biotech) in binding buffer (50 mM Tris-HCl (pH 8.0), 100 mM NaCl, 5 mM MgCl2, 20% glycerol). The beads were washed four times with binding buffer, and the bound proteins were separated by SDS–PAGE and analysed by Western blotting.

Electrophoretic mobility shift assays (EMSA)

EMSA were performed by incubating 50 fmol of 32P-end-labeled synthetic double-stranded oligonucleotide with either nuclear extracts (10 μg) or purified GST fusion protein in binding buffer (10 mM Tris-HCl (pH 7.5), 1 mM MgCl2, 0.5 mM EDTA, 50 mM NaCl, 0.5 mM DTT, 4% glycerol). The sequence of the oligonucleotide probe used had three SBE with the sequence 5′-IndexTermTCGAGAGCCAGACAAAA AGCCAGACATTTAGCCAGACAC-3′; the sequence for the mutant SBE probe is 5′-IndexTermTCGAGAGC T A C A T AAAAAGC T A C A T ATTTAGC T A C A T AC-3′. The sequence of the p21 probe corresponds to the −120/−60 region of the proximal p21 promoter; the same oligonucleotide harboring point mutations within the Sp1 consensus sequence in −83/−77 region (WT sequence, 835′-IndexTermGGTCCCGCC-3′77; mutant type, 835'-IndexTermGGTCGACCC-3′−77) was used as the mutant p21 probe. After incubation at room temperature for 45 min, the DNA–protein complex was separated from the free oligonucleotide on a 6% native polyacrylamide gel using buffer containing 0.25 × TBE (22.5 mM Tris-borate, 0.5 mM EDTA (pH 8.0)). After electrophoresis, the gel was dried and visualized with a PhosphorImager. Competition experiments were carried out by incubating unlabeled 50-fold excess of either the mutated (MT) or the WT probe. For supershift assays, 0.15 μg of specific anti-Sp1, -Sp3, or -Smad3 antibody (Upstate) was added to the reaction mixture prior to the addition of the radiolabeled oligonucleotide.

Flow cytometry assay for DNA content

For the analysis of the TGF-β-induced cell cycle arrest at G1 phase, cells were first treated with 10 ng/ml TGF-β for various times in complete medium, then collected by trypsinization, and resuspended in 2 ml ice-cold PBS. The suspensions were fixed with 5 ml ethanol drop by drop. After incubation at 4°C for at least 48 h, the cells were recovered by centrifugation and stained by incubating cells for 30 min at 37°C in 1 ml PBS containing 5 μg/ml of propidium iodide. Cells were then analysed by flow cytometry without further washings. Flow cytometry was carried out using a FACScan (Becton Dickinson) with argon laser excitation at 488 nm. The population of cells at each cell cycle phase was finally analysed by ModFit LT software (Becton Dickinson Immunocytometry Systems).

Determination of apoptotic cells

For flow cytometric detection of dual staining of the intracellular Smad3 variants and the apoptotic cells, cells were transfected with the Myc-Smad3 or DN-Smad3 expression construct (pRK/Smad3ΔCF). The transfected cells were treated with 100 ng/ml TGF-β for 24 h in serum-free medium, and then fixed by incubating in 2% paraformaldehyde/PBS solution at 25°C for 20 min. The fixed cells were then permeabilized with 0.1% Saponin/PBS (Sigma) at 25°C for 10 min. The Smad3 variants staining procedure is similar to that used in immunofluorescence experiment, except a 1 : 10 dilution of rabbit anti-Smad3 antibody and 10 μg of rhodamine-labeled goat anti-rabbit antibody were used. After this, a fluorescein-based terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) staining was employed using the In Situ Cell Death Detection Kit (Roche). The staining procedure was performed according to the instruction manual, and the stained cells were then subjected to flow cytometry.

References

  1. Abe M, Harpel JG, Metz CN, Nunes L, Loskutoff DJ and Rigkin DB . (1994). Anal. Biochem., 216, 276–284.

  2. Blight KJ, Kolykhalov AA and Rice CM . (2000). Science, 290, 1972–1974.

  3. Chen CM, You LR, Hwang LH and Lee YHW . (1997). J. Virol., 71, 9417–9426.

  4. Chen SY, Kao CF, Chen CM, Shih CM, Hsu MJ, Chao CH, Wang SH, You LR and Lee YHW . (2003). J. Biol. Chem., 278, 591–607.

  5. Datto MB, Li Y, Panus JF, Howe DJ, Xiong Y and Wang XF . (1995). Proc. Natl. Acad. Sci. USA, 92, 5545–5549.

  6. De Bleser PJ, Niki T, Rogiers V and Geerts A . (1997). J. Hepatol., 26, 886–893.

  7. De Francesco R and Steinkuhler C . (2000). Curr. Top. Microbiol. Immunol., 242, 149–169.

  8. Dennler S, Itoh S, Vivien D, Dijke PT, Huet S and Gauthier JM . (1998). EMBO J., 17, 3091–3100.

  9. Derynck R, Akhurst RJ and Balmain A . (2001). Nat. Genet., 29, 117–129.

  10. Dey A, Atcha IA and Bagchi S . (1997). Virology, 228, 190–199.

  11. Dubordeau M, Miyamura T, Matsuura Y, Alric L, Pipy B and Rousseau D . (2002). J. Hepatol., 37, 486–492.

  12. El-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein B . (1993). Cell, 75, 817–825.

  13. Fujita T, Ishido S, Muramatsu S, Itoh M and Hotta H . (1996). Biochem. Biophys. Res. Commun., 229, 825–831.

  14. Graham FL and van der Eb AJ . (1973). Virology, 52, 456–467.

  15. Hata A, Lo RS, Wotton D, Lagna G and Massague J . (1997). Nature, 388, 82–87.

  16. Ito N, Kawata S, Tamura S, Takaishi K, Shirai Y, Kiso S, Yabuuchi I, Matsuda Y, Nishioka M and Tarui S . (1991). Cancer Res., 51, 4080–4083.

  17. Jang CW, Chen CH, Chen CC, Chen JY, Su YH and Chen RH . (2002). Nat. Cell Biol., 4, 51–58.

  18. Jayaraman L and Massague J . (2000). J. Biol. Chem., 275, 40710–40717.

  19. Kim DH, Chang JH, Lee KH, Lee HY and Kim SJ . (1996). J. Biol. Chem., 272, 688–694.

  20. Kolykhalov AA, Agapov EV, Blight KJ, Mihalik K, Feinstone SM and Rice CM . (1997). Science, 277, 570–574.

  21. Kwong AD, Kim JL and Lin C . (2000). Curr. Top. Microbiol. Immunol., 242, 171–196.

  22. Lai MM and Ware CF . (2000). Curr. Top. Microbiol. Immunol., 242, 117–134.

  23. Lauer GM and Walker BD . (2001). N. Engl. J. Med., 345, 41–52.

  24. Lee DK, Kim BC, Brady JN, Jeang KT and Kim SJ . (2002a). J. Biol. Chem., 277, 33766–33775.

  25. Lee DK, Kim BC, Kim IY, Cho EA, Satterwhite DJ and Kim SJ . (2002b). J. Biol. Chem., 277, 38557–38564.

  26. Lee DK, Park SH, Yi Y, Choi SG, Lee C, Parks WT, Cho H, de Caestecker MP, Shaul Y, Roberts AB and Kim SJ . (2001). Genes Dev., 15, 455–466.

  27. Lee MN, Jung EY, Kwun HJ, Jun HK, Yu DY, Choi YH and Jang KL . (2002c). J. Gen. Virol., 83, 2145–2151.

  28. Massague J . (2000). Nat. Rev. Mol. Cell. Biol., 1, 169–178.

  29. Matsuzaki K, Date M, Furukawa F, Tahashi Y, Matsushita M, Sakitani K, Yamashiki N, Seki T, Saito H, Nishizawa M, Fujisawa J and Inoue K . (2000). Cancer Res., 60, 1394–1402.

  30. Moriya K, Fujie H, Shintani Y, Yotsuyanagi H, Tsutsumi T, Ishibashi K, Matsuura Y, Kimura S, Miyamura T and Koike K . (1998). Nat. Med., 4, 1065–1067.

  31. Moriya K, Yotsuyanagi H, Shintani Y, Fujie H, Ishibashi K, Matsuura Y, Miyamura T and Koike K . (1997). J. Gen. Virol., 78, 1527–1531.

  32. Nakao A, Imamura T, Souchelnytskyi S, Kawabata M, Ishisaki A, Oeda E, Tamaki K, Hanai J, Heldin CH, Miyazono K and ten Dijke P . (1997). EMBO J., 16, 5353–5362.

  33. Nishihara A, Hanai JI, Imamura T, Miyazono K and Kawabata M . (1999). J. Biol. Chem., 274, 28716–28723.

  34. Oberhammer FA, Pavelka M, Sharma S, Tiefenbacher R, Purchio AF, Bursch W and Schulte-Hermann R . (1992). Proc. Natl. Acad. Sci. USA, 89, 5408–5412.

  35. Pardali K, Kurisaki A, Moren A, ten Dijke P, Kardassis D and Moustakas A . (2000). J. Biol. Chem., 275, 29244–29256.

  36. Prokova V, Mosialos G and Karassis D . (2001). J. Biol. Chem., 277, 9342–9350.

  37. Ray RB, Steele R, Meyer K and Ray R . (1998). Gene, 208, 331–336.

  38. Ray S, Broor SL, Vaishnav Y, Sarkar C, Girish R, Dar L, Seth P and Broor S . (2003). J. Gastroenterol. Hepatol., 18, 393–403.

  39. Reed KE and Rice CM . (2000). Curr. Top. Microbiol. Immunol., 242, 55–84.

  40. Sakamuro D, Furukawa T and Takegami T . (1995). J. Virol., 69, 3893–3896.

  41. Seeff LB . (2002). Hepatology, 36, S35–S46.

  42. Shih CM, Lo SJ, Miyamura T, Chen SY and Lee YHW . (1993). J. Virol., 67, 5823–5832.

  43. Shih WL, Kuo ML, Chuang SE, Cheng AL and Doong SL . (2000). J. Biol. Chem., 275, 25858–25864.

  44. Tellinghuisen TL and Rice CM . (2002). Curr. Opin. Microbiol., 5, 419–427.

  45. Wrana JL, Attisano L, Carcamo J, Zentella A, Doody J, Laiho M, Wang XF and Massague J . (1992). Cell, 71, 1003–1014.

  46. You LR, Chen CM, Yeh TS, Tsai TY, Mai RT, Lin CH and Lee YHW . (1999). J. Virol., 73, 2841–2853.

  47. Zemel R, Gerechet S, Greif H, Bachmatove L, Birk Y, Golan-Goldhirsh A, Kunin M, Berdichevsky Y, Benhar I and Tur-Kaspa R . (2001). J. Viral Hepatitis., 8, 96–102.

  48. Zhang Y, Feng XH and Derynck R . (1998). Nature, 394, 909–913.

  49. Zhang Y, Feng XH, We RY and Derynck R . (1996). Nature, 383, 168–172.

Download references

Acknowledgements

We thank CM Rice and Apath (St Louis, MO, USA) for generously providing the Ava.5 cells and p90/HCVFLlongpU plasmid. We also thank L-H Hwang, R-H Chen, C-K Chou, K-H Lan, K Tokushige and K Miyazono for providing plasmids used in this study. We are grateful to R Kirby for critical reading and comments on this manuscript. This work was supported by the following grants to Y-HW Lee: NSC 89-2315-B-010-006-MH, NSC 89-2320-B-010-121, NSC 90-2320-B-010-077, NSC 91-2320-B-010-040, and NSC 92-2320-B-010-064 from National Science Council; and in part by grants NHRI-EX91-9002BL, NHRI-EX92-9002BL, and NHRI-EX93-9002BL from the National Health Research Institute.

Author information

Correspondence to Yan-Hwa Wu Lee.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • HCV
  • core protein
  • NS3 protein
  • TGF-β
  • Smad3
  • hepatocarcinogenesis

Further reading