Hepatitis C Virus Infection Suppresses GLUT2 Gene Expression via Downregulation of Hepatocyte Nuclear Factor 1α

Abstract

Hepatitis C virus (HCV) infection causes not only intrahepatic diseases but also extrahepatic manifestations, including type 2 diabetes. We previously reported that HCV replication suppresses cellular glucose uptake by downregulation of cell surface expression of glucose transporter 2 (GLUT2) (D. Kasai et al., J. Hepatol. 50:883–894, 2009). GLUT2 mRNA levels were decreased in both HCV RNA replicon cells and HCV J6/JFH1-infected cells. To elucidate molecular mechanisms of HCV-induced suppression of GLUT2 gene expression, we analyzed transcriptional regulation of the GLUT2 promoter using a series of GLUT2 promoter-luciferase reporter plasmids. HCV-induced suppression of GLUT2 promoter activity was abrogated when the hepatocyte nuclear factor 1α (HNF-1α)-binding motif was deleted from the GLUT2 promoter. HNF-1α mRNA levels were significantly reduced in HCV J6/JFH1-infected cells. Furthermore, HCV infection remarkably decreased HNF-1α protein levels. We assessed the effects of proteasome inhibitor or lysosomal protease inhibitors on the HCV-induced reduction of HNF-1α protein levels. Treatment of HCV-infected cells with a lysosomal protease inhibitor, but not with a proteasome inhibitor, restored HNF-1α protein levels, suggesting that HCV infection promotes lysosomal degradation of HNF-1α protein. Overexpression of NS5A protein enhanced lysosomal degradation of HNF-1α protein and suppressed GLUT2 promoter activity. Immunoprecipitation analyses revealed that the region from amino acids 1 to 126 of the NS5A domain I physically interacts with HNF-1α protein. Taken together, our results suggest that HCV infection suppresses GLUT2 gene expression via downregulation of HNF-1α expression at transcriptional and posttranslational levels. HCV-induced downregulation of HNF-1α expression may play a crucial role in glucose metabolic disorders caused by HCV.

INTRODUCTION

Hepatitis C virus (HCV) is the main cause of chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma. HCV is a single-stranded, positive-sense RNA virus that is classified into the Flaviviridae family, Hepacivirus genus (21). More than 170 million people worldwide are chronically infected with HCV. The 9.6-kb HCV genome encodes a polyprotein of approximately 3,010 amino acids (aa). The polyprotein is cleaved co- and posttranslationally into at least 10 proteins by viral proteases and cellular signalases: the structural proteins core, E1, E2, and p7 and the nonstructural proteins NS2, NS3, NS4A, NS4B, NS5A, and NS5B (21).

Persistent HCV infection causes not only intrahepatic diseases but also extrahepatic manifestations, such as type 2 diabetes. Clinical and experimental data suggest that HCV infection is an additional risk factor for the development of diabetes (26, 29, 30). HCV-related glucose metabolic changes and insulin resistance have significant clinical consequences, such as accelerated fibrogenesis, reduced virological response to alpha interferon (IFN-α)-based therapy, and increased incidence of hepatocellular carcinoma (29). Therefore, the molecular mechanism of HCV-related diabetes needs to be clarified.

We have sought to identify a novel mechanism of HCV-induced diabetes. We previously demonstrated that HCV suppresses hepatocytic glucose uptake through downregulation of cell surface expression of glucose transporter 2 (GLUT2) in a human hepatoma cell line (19). The uptake of glucose into cells is conducted by facilitative glucose carriers, i.e., glucose transporters (GLUTs). GLUTs are integral membrane proteins that contain 12 membrane-spanning helices. To date, a total of 14 isoforms have been identified in the GLUT family (24). GLUT2 is expressed in the liver, pancreatic β-cells, hypothalamic glial cells, retina, and enterocytes. Glucose is transported into hepatocytes by GLUT2 (34). We previously reported that GLUT2 expression was reduced in hepatocytes obtained from HCV-infected patients (19). We also demonstrated that GLUT2 mRNA levels were lower in HCV replicon cells and in HCV J6/JFH1-infected cells than in the control cells. GLUT2 promoter activity was suppressed in HCV-replicating cells. However, the molecular mechanism of HCV-induced suppression of GLUT2 gene expression remains to be elucidated.

In the present study, we aimed to clarify molecular mechanisms of HCV-induced suppression of GLUT2 gene expression. We analyzed transcriptional regulation of the GLUT2 promoter in HCV replicon cells. We demonstrate that HCV infection downregulates hepatocyte nuclear factor 1α (HNF-1α) expression at both transcriptional and posttranslational levels, resulting in suppression of GLUT2 promoter. We propose that HCV-induced downregulation of HNF-1α may play a crucial role in glucose metabolic disorders caused by HCV.

MATERIALS AND METHODS

Cell culture.

The human hepatoma cell line Huh-7.5 (4) was kindly provided by Charles M. Rice (The Rockefeller University, New York, NY). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (high glucose) with l-glutamine (Wako, Osaka, Japan) supplemented with 50 IU/ml penicillin, 50 μg/ml streptomycin (Gibco, NY), 10% heat-inactivated fetal bovine serum (Biowest, France), and 0.1 mM nonessential amino acids (Invitrogen, NY) at 37°C in a 5% CO₂ incubator. Cells were transfected with plasmid DNA using FuGENE 6 transfection reagents (Promega, Madison, WI).

Huh-7.5 cells stably harboring an HCV-1b subgenomic RNA replicon (SGR) were prepared as described previously (18), using pFK5B/2884Gly (a kind gift from R. Bartenschlager, University of Heidelberg, Heidelberg, Germany). The SGR cells express the genomic region from NS3 to NS5B of the HCV Con1 strain (19) (Fig. 1). Cells harboring a full-genome HCV-1b RNA replicon (FGR) derived from Con1 (27) or pON/C-5B (17, 19) (a kind gift from N. Kato, Okayama University, Okayama, Japan) were also used. The FGR cells express all of the HCV proteins (the region ranging from the core protein to NS5B).

Fig 1.

The HCV genome, chimeric HCV J6/JFH1, and the HCV RNA replicons. Schematic diagrams of the HCV genome, the chimeric HCV J6/JFH1 genome, SGR, and FGR are shown. IRES, internal ribosome entry site; EMCV, encephalomyocarditis virus; Neo, neomycin resistance gene.

The pFL-J6/JFH1 plasmid that encodes the entire viral genome of a chimeric strain of HCV-2a, J6/JFH1 (23), was kindly provided by Charles M. Rice. The HCV genome RNA was synthesized in vitro using pFL-J6/JFH1 as a template and was transfected into Huh-7.5 cells by electroporation (6, 9, 23, 37). The virus produced in the culture supernatant was used for infection experiments (6).

Cells were treated with 1,000 IU/ml of IFN-α (Sigma, St. Louis, MO) for 10 days to eliminate HCV replication (19).

Luciferase reporter assay.

We constructed the human GLUT2 promoter-luciferase reporter plasmid by cloning a 1.6-kb genomic fragment that encompasses the human GLUT2 promoter region from −1291 to +308, yielding pGLUT2(−1291/+308)-Luc (2, 19), into the pGL4 vector plasmid (Promega). The pGLUT2(−1291/+308)-Luc construct contains a 1,291-bp fragment of the human GLUT2 promoter upstream of the minimal promoter and the coding sequence of the Photinus pyralis (firefly) luciferase. We also used seven different GLUT2 promoter-luciferase reporter plasmids, i.e., pGLUT2(−1193/+308)-Luc, pGLUT2 (−1155/+308)-Luc, pGLUT2(−1100/+308)-Luc, pGLUT2(−1030/+308)-Luc, pGLUT2(−206/+308)-Luc, pGLUT2(+29/+308)-Luc, and pGLUT2(+126/+308)-Luc, which lack the binding sequence of the CCAAT/enhancer binding site (C/EBP), cyclic AMP (cAMP) response element (CRE), AP-1 binding site, HNF-1α binding site, CAAT box, TATA-like motif, and transcriptional initiation, respectively (Fig. 2A). The reporter plasmid pRL-CMV-Renilla (where CMV is cytomegalovirus) (Promega) was used as an internal control. Cells were transfected with each pGLUT2-Luc construct together with pRL-CMV-Renilla. At 48 h after transfection, samples were harvested and assayed for luciferase activity. The luciferase assays were performed using a dual-luciferase reporter assay system (Promega). Luciferase activity was measured by a Lumat LB 9501 instrument (Berthold Technologies GmbH & Co., Bad Wildbad, Germany). Firefly luciferase activity was normalized to Renilla luciferase activity for each sample. The number of relative light units (RLU) of the SGR cells or FGR cells transfected with each reporter plasmid is expressed as a ratio of the number of Huh-7.5 cells transfected with each reporter plasmid.

Fig 2.

HNF-1α-binding site is important for HCV-induced suppression of GLUT2 promoter. (A) A series of constructs in which genomic GLUT2 promoter DNA fragments were fused to a promoterless firefly luciferase gene of the pGL4 vector were generated with the 3′ end always terminating at bases +308 from transcriptional start site. The 5′ ends began at bases −1291, −1193, −1155, −1100, −1030, −206, +29, and +126 The regions that represent potential binding sites for transcription factors are shown, including a CCAAT/enhancer binding site (C/EBP), cAMP response element (CRE), AP-1 binding site, HNF-1α binding site, CAAT box, and TATA-like motif. The nucleotide at the beginning of the construct is indicated. (B) Huh-7.5 cells, SGR cells, and FGR cells (2.5 × 10⁵ cells/six-well plate) were transfected with each GLUT2 plasmid (0.5 μg) together with pRL-CMV-Renilla (25 ng). pRL-CMV-Renilla was used as an internal control. At 48 h posttransfection, cells were harvested and assayed for luciferase activities using a dual-luciferase reporter assay system. RLU is expressed as a ratio of the Huh-7.5 cells transfected with each reporter plasmid.

Expression plasmids.

Expression plasmids for core protein, p7, NS2, NS3, NS4A, NS4B, NS5A, and NS5B were described previously (9, 10, 18). To express E1 and E2 (E1/E2), the cDNA fragment of nucleotides (nt) 825 to 2676 derived from the HCV Con1 strain was amplified by PCR using the plasmid pFKI389neo/core-3′/Con1 (a kind gift from R. Bartenschlager) as a template. Specific primers used for PCR were as follows: sense primer, 5′-CCAGTGTGGTGAATTCACCATGGTGAACTATGCAACAGGGAA-3′; antisense primer, 5′-CGAAGGGCCCTCTAGAGATGTACCAGGCAGCACAGA-3′. To express NS3 and NS4A (NS3/4A), the cDNA fragment of nt 3420 to 5474 derived from the HCV Con1 strain was amplified by PCR. Specific primers were as follows: sense primer, 5′-CCAGTGTGGTGAATTCACCATGGCGCCTATTACGGCCTACTC-3′; antisense primer, 5′-CGAAGGGCCCTCTAGAGCACTCTTCCATCTCATCGAA-3′. These amplified PCR products were purified, and each of them was inserted into the EcoRI-XbaI site of pEF1/myc-His A (Invitrogen) using an In-Fusion HD-Cloning kit (Clontech, Mountain View, CA). To express a series of NS5A deletion mutants as hemagglutinin (HA)-tagged proteins, each fragment was amplified by PCR and cloned into the NotI site of pCAG-HA. pEF1A-NS5A (Con1)-myc-His was used as a template (18). The primer sequences used in this study are available from the authors upon request. The sequences of the inserts were extensively verified by sequencing (Operon biotechnology, Tokyo, Japan). The plasmids pEF1A-NS5A(1–126)-myc-His, consisting of residues 1 to 126 in NS5A, and pEF1A-NS5A(1–147)-myc-His were described previously (18).

Antibodies.

The mouse monoclonal antibodies (MAbs) used in this study were anti-FLAG (M2) MAb (F-3165; Sigma), anti-NS5A MAb (MAB8694; Millipore), anti-core protein MAb (2H9) (37), and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) MAb (MAB374; Millipore). Polyclonal antibodies (PAbs) used in this study were anti-HNF-1α rabbit PAb (sc-8986; Santa Cruz Biotechnology), anti-HNF-1α goat PAb (sc-6548; Santa Cruz Biotechnology), anti-NS5B goat PAb (sc-17532; Santa Cruz Biotechnology), anti-NS3 rabbit PAb (described elsewhere), and anti-actin goat PAb (C-11; Santa Cruz Biotechnology). Horseradish peroxidase (HRP)-conjugated anti-mouse IgG antibody (Cell signaling), HRP-conjugated donkey anti-goat IgG (Santa Cruz Biotechnology), and HRP-conjugated anti-rabbit IgG (Cell signaling) were used as secondary antibodies.

Real-time quantitative reverse transcription-PCR (RT-PCR).

Total cellular RNA was isolated using RNAiso reagent (TaKaRa Bio, Kyoto, Japan), and cDNA was generated using a QuantiTect Reverse Transcription system (Qiagen, Valencia, CA). Real-time quantitative PCR was performed using SYBR Premix Ex Taq (TaKaRa Bio) with SYBR green chemistry on an ABI Prism 7000 system (Applied Biosystems, Foster, CA), as described previously (11, 19). The β-glucronidase (GUS) gene was used as an internal control. The primers used for real-time PCR are as follows: for HNF-1α (NM_000545), 5′-AGCTACCAACCAAGAAGGGGC-3′ (nt 601 to 621) and 5′-TGACGAGGTTGGAGCCCAGCC-3′ (nt 801 to 781); HNF-1β (NM_000458), 5′-GTTACATGCAGCAACACAACA-3′ (nt 600 to 620) and 5′-TCATATTTCCAGAACTCTGGA-3′ (nt 801 to 782); GUS (NM_000181), 5′-ATCAAAAACGCAGAAAATACG-3′ (nt 1797 to 1817) and 5′-ACGCAGGTGGTATCAGTCTTG-3′ (nt 2034 to 2014).

Immunoblot analysis.

Immunoblot analysis was performed essentially as described previously (9, 33). The cell lysates were separated by 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride membrane (Millipore Corp., Billerica, MA). The membranes were incubated with primary antibody, followed by incubation with peroxidase-conjugated secondary antibody. The positive bands were visualized using ECL Western blotting detection reagents (GE Healthcare, Buckinghamshire, United Kingdom). To detect endogenous HNF-1α protein, ECL Plus Western blotting detection reagents were used (GE Healthcare).

Immunoprecipitation.

Cultured cells were lysed with a buffer containing 150 mM NaCl, 20 mM Tris-HCl (pH 7.4), 0.1% SDS, 1% NP-40, and Complete protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN). The lysate was centrifuged at 12,000 × g for 20 min at 4°C, and the supernatant was immunoprecipitated with appropriate antibodies. Immunoprecipitation was performed as described previously (10). Briefly, the cell lysates were immunoprecipitated with control IgG and Dynabeads protein A (Invitrogen) and incubated with appropriate antibodies at 4°C overnight. After being washed with the washing buffer (0.1 M Na-phosphate buffer, pH 7.4) five times, the immunoprecipitates were analyzed by immunoblotting.

Statistical analysis.

Results were expressed as means ± standard errors of the means (SEM). Statistical significance was evaluated by analysis of variance (ANOVA), and statistical significance was defined as a P value of <0.05.

RESULTS

HNF-1α-binding site is important for HCV-induced suppression of GLUT2 promoter.

To gain an insight into potential regulatory sequences involved in HCV-induced suppression of GLUT2 gene transcription, a 1.6-kb genomic fragment that encompasses the human GLUT2 promoter (−1291 to +308) and a series of deletion mutants were analyzed (Fig. 2A). The ability of the upstream region of the GLUT2 gene to function as a promoter was assessed by its capacity to drive the expression of a luciferase reporter gene. GLUT2 promoter activity was assessed by measuring luciferase activity of the cell extracts derived from transiently transfected Huh-7.5 cells, SGR cells, and FGR cells. As shown in Fig. 2B, a deletion of the promoter sequence to −1100 [pGLUT2(−1100/+308)-Luc [ΔAP-1]] showed lower luciferase activities in HCV replicon cells than in the control cells. Successive removal of nucleotides from −1100 to −1030 completely or almost completely abolished the suppression of the luciferase activity in both FGR and SGR cells, suggesting that the HNF-1α-binding site is important for HCV-induced suppression of GLUT2 promoter.

HCV infection reduces HNF-1α mRNA levels.

It is worth noting that HNF-1α is known to play a crucial role in diabetes. Mutations in the HNF-1α gene have been reported to cause a monogenic form of diabetes mellitus with autosomal dominant inheritance, termed maturity onset diabetes of the young 3 (MODY3) (25, 40). Cha et al. (7) reported that HNF-1α functions as a transcriptional transactivator in human GLUT2 gene expression in a human hepatoma cell line. These findings motivated us to further investigate a role of HNF-1α in HCV-induced glucose metabolic disorders in a human hepatoma cell line. To determine whether HCV infection suppresses HNF-1α mRNA expression, we quantified mRNA levels of HNF-1α and HNF-1β in HCV J6/JFH1-infected cells and in mock-infected cells by real-time RT-PCR. HNF-1α mRNA levels were significantly reduced in HCV J6/JFH1-infected cells from 5 days postinfection (dpi) to 14 dpi (Fig. 3A). On the other hand, HNF-1β mRNA levels remained unchanged until 14 dpi (Fig. 3B). These results suggest that HCV infection specifically downregulates HNF-1α mRNA expression.

Fig 3.

Quantitative RT-PCR analysis of mRNA for HNF-1α and HNF-1β in HCV J6/JFH1-infected cells. Huh-7.5 cells (2.5 × 10⁵ cells/six-well plate) were infected with HCV J6/JFH1 at a multiplicity of infection of 2. Cells were cultured and harvested at the indicated times. Total RNA was extracted, and the levels of HNF-1α mRNA and HNF-1β mRNA were determined by quantitative RT-PCR. Mock-infected cells served as negative controls. **, P < 0.01, compared with mock-infected cells.

HCV infection reduces HNF-1α protein levels.

To determine whether HCV infection reduces HNF-1α protein levels, endogenous HNF-1α protein levels were examined by immunoblot analysis. The HNF-1α protein level was much lower in J6/JFH1-infected cells than in the mock-infected control (Fig. 4A, upper panel, lane 2). To determine whether HCV infection is specifically involved in reduction of HNF-1α protein, we eliminated HCV by treatment of the cells with IFN-α (Fig. 4B, lower panel, compare lane 2 with lane 4). Upon elimination of HCV, the HNF-1α protein expression level recovered to the level of the mock-infected control (Fig. 4B, upper panel, compare lane 2 with lane 4). These results suggest that HCV infection specifically reduces HNF-1α protein levels.

Fig 4.

HCV infection induces lysosomal degradation of HNF-1α protein. (A) HCV infection decreased the levels of HNF-1α protein in Huh-7.5 cells. Huh-7.5 cells (2.5 × 10⁵ cells/six-well plate) were infected with HCV J6/JFH1 at a multiplicity of infection of 2. Cells were cultured and harvested at 5 days postinfection. Cells were analyzed by immunoblotting with anti-HNF-1α, anti-NS3, and anti-actin antibodies. The level of actin served as a loading control. The relative levels of protein expression were quantitated by densitometry and are indicated below the respective lanes. (B) HCV-induced downregulation of HNF-1α protein was restored by treatment of the cells with IFN-α. Huh-7.5 cells were plated at 2.5 × 10⁵ cells/six-well plate and cultured for 12 h. The cells were infected with HCV J6/JFH1 at a multiplicity of infection of 2 and cultured for 5 days. The cells were replated at 2.5 × 10⁵ cells/six-well plate and cultured in complete DMEM with or without 1,000 IU/ml IFN-α for 10 days to eliminate HCV. The cells cultured in DMEM without IFN-α served as negative controls. (C) HCV-induced reduction of HNF-1α protein was restored by treatment of the cells with lysosomal protease inhibitor. Huh-7.5 cells were plated at 2.0 × 10⁵ cells/six-well plate and cultured for 12 h. At 5 days postinfection, proteasome inhibitor (30 μM clasto-lactacystin β-lactone) or lysosomal protease inhibitors (40 μM E-64d and 20 μM pepstatin A) were administered to the cells. Cells were cultured for 12 h, harvested, and analyzed by immunoblotting as indicated. The level of GAPDH served as a loading control. DMSO, dimethyl sulfoxide; PepA, pepstatin A.

HCV-induced reduction of HNF-1α protein is restored by treatment of the cells with a lysosomal protease inhibitor.

As shown in Fig. 3A, HNF-1α mRNA levels in HCV J6/JFH1-infected cells decreased slowly at day 5 postinfection. One possible explanation is that suppression of HNF-1α mRNA is an indirect effect caused by HCV infection. The degree of the reduction of the HNF-1α protein was larger than that of HNF-1α mRNA (Fig. 4A), suggesting the involvement of protein degradation in reduction of HNF-1α protein levels. To determine whether protein degradation is involved in HCV-induced reduction of HNF-1α protein, we assessed the role of proteasome or lysosome proteases in the reduction of HNF-1α protein. We treated the cells with a proteasome inhibitor, clasto-lactacystin β-lactone, or lysosome protease inhibitors E-64d and pepstatin A. Clasto-lactacystin β-lactone had no effect on the levels of HNF-1α protein (Fig. 4C, upper panel, lane 6). This result suggests that proteasome is not involved in the reduction of HNF-1α protein. E-64d is a cysteine protease inhibitor, and pepstatin A is an aspartic protease inhibitor. Pepstatin A, but not E-64d, restored the levels of HNF-1α protein (Fig. 4C, upper panel, lanes 10 and 8). These results suggest that a lysosomal protease, such as an aspartic protease, is involved in HCV-induced reduction of HNF-1α protein.

Overexpression of NS5A protein suppresses GLUT2 promoter activity.

To determine which HCV protein is involved in the suppression of GLUT2 promoter, we examined the effects of transient expression of HCV proteins on GLUT2 promoter activity. Huh-7.5 cells were cotransfected with each HCV protein expression plasmid together with the GLUT2 promoter-luciferase plasmid. The pRL-CMV-Renilla plasmid was cotransfected as an internal control. At 48 h posttransfection, cells were harvested and assayed for luciferase activity. As shown in Fig. 5A, overexpression of the NS5A expression plasmid significantly reduced GLUT2 promoter activity. On the other hand, other HCV protein expression plasmids failed to suppress GLUT2 promoter activity (Fig. 5A, left and right panels). These results suggest that NS5A protein is involved in the suppression of GLUT2 promoter activity.

Fig 5.

HCV NS5A protein is involved in suppression of GLUT2 promoter activity and lysosomal degradation of HNF-1α protein. (A) Huh-7.5 cells were plated at 1 × 10⁵ cells/12-well plate. After cells were cultured for 12 h, cells were cotransfected with each HCV protein plasmid (0.5 μg), the human GLUT2 promoter reporter plasmid (0.5 μg), and pRL-CMV-Renilla (25 ng). pRL-CMV-Renilla was used as an internal control. At 48 h posttransfection, cells were harvested. Luciferase assays were performed by using a dual-luciferase reporter assay system. (B) Huh-7.5 cells were plated at 4 × 10⁵ cells/six-well plate and cultured for 12 h. Cells were transfected with increasing amounts of either NS5A plasmid or NS5B plasmid as indicated. At 48 h posttransfection, cells were harvested. Whole-cell lysates were analyzed by immunoblotting with anti-HNF-1α, anti-NS5A, and anti-NS5B antibodies. The level of GAPDH served as a loading control. (C) Huh-7.5 cells (2.5 × 10⁵ cells/six-well plate) were transfected with pEF1A-NS5A-myc-His₆. At 2 days posttransfection, proteasome inhibitor (30 μM clasto-lactacystin β-lactone) or lysosomal enzyme inhibitors (40 μM E-64d and 20 μM pepstatin A) were administered to the cells. Cells were cultured for 12 h and harvested, and the levels of endogenous HNF-1α protein were analyzed by immunoblotting with anti-HNF-1α goat PAb. The level of GAPDH served as a loading control. (D) Huh-7.5 cells (1.0 × 10⁵ cells/12-well plate) were transfected with the human GLUT2 promoter reporter plasmid (0.5 μg) and pRL-CMV-Renilla (25 ng). The plasmid pEF1A/myc-His (0.5 μg) was cotransfected to the control cells. Cells were transfected with the plasmid pEF1A-NS5A-myc-His (0.5 μg) together with either empty plasmid pCMV4 (10 ng) or pCMV-HNF-1α (10 ng). At 48 h posttransfection, cells were harvested. Luciferase assays were performed by using a dual-luciferase reporter assay system. *, P < 0.05, compared with control. (E) Huh-7.5 cells (1.2 × 10⁶ cells /10 cm-dish) were infected with HCV J6/JFH1 at a multiplicity of infection of 2 and cultured for 5 days. At day 5 postinfection, cells were plated at 1.0 × 10⁵ cells/12-well plate and cultured for 12 h. Mock-infected cells were plated similarly. Cells were transfected with the human GLUT2 promoter reporter plasmid (0.5 μg) and pRL-CMV-Renilla (25 ng) together with either empty plasmid pCMV4 or pCMV-HNF-1α, cultured for 48 h, and harvested. Luciferase assays were performed by using a dual-luciferase reporter assay system. *, P < 0.05, compared with control.

Overexpression of NS5A protein reduces the levels of endogenous HNF-1α protein.

To investigate a role of NS5A in the suppression of the GLUT2 promoter, we examined the effect of NS5A protein on the levels of endogenous HNF-1α protein. Huh-7.5 cells were transfected with increasing amounts of either an NS5A expression plasmid or NS5B expression plasmid. At 48 h posttransfection, cells were harvested, and the levels of endogenous HNF-1α protein were analyzed by immunoblot analysis. To detect endogenous HNF-1α protein, highly sensitive Western blotting detection reagents (ECL Plus Western blotting detection reagents) were used. Overexpression of NS5A (Fig. 5B, left panel) but not NS5B (Fig. 5B, right panel) significantly reduced endogenous HNF-1α protein. These results suggest that NS5A protein specifically reduces endogenous HNF-1α protein levels.

To determine if NS5A-dependent reduction of HNF-1α protein is due to lysosomal degradation, we treated the cells with lysosome protease inhibitors. Pepstatin A, but not E-64d, recovered the levels of HNF-1α protein (Fig. 5C, middle panel, lanes 5 and 6), which is consistent with the results found in HCV-infected cells. These results suggest that NS5A is responsible for HCV-induced lysosomal degradation of HNF-1α protein. Taken together, our results suggest that HCV infection suppresses GLUT2 promoter activity via NS5A-dependent lysosomal degradation of HNF-1α protein.

To verify a role of HNF-1α in the HCV-induced suppression of GLUT2 promoter activity, we examined the effects of ectopic expression of HNF-1α on GLUT2 promoter activity in NS5A-transfected cells as well as in HCV J6/JFH1-infected cells. As shown in Fig. 5D, overexpression of NS5A decreased GLUT2 promoter activity, and ectopic expression of HNF-1α restored GLUT2 promoter activity (Fig. 5D). Moreover, HCV J6/JFH1 infection significantly decreased GLUT2 promoter activity, and ectopic expression of HNF-1α restored GLUT2 promoter activity (Fig. 5E). These results are consistent with the notion that HNF-1α protein is a key regulator for HCV-induced suppression of GLUT2 promoter activity.

NS5A protein interacts with HNF-1α protein in Huh-7.5 cells and in FGR Con1 cells.

It was previously reported that in vitro translated HNF-1 protein was pulled down with glutathione S-transferase (GST)-NS5A protein (32). To determine whether NS5A physically interacts with HNF-1α protein in cultured cells, Huh-7.5 cells were cotransfected with each FLAG-tagged NS5A expression plasmid together with the HNF-1α expression plasmid. Immunoprecipitation analysis revealed that HNF-1α protein was coimmunoprecipitated with FLAG-NS5A protein using anti-FLAG MAb (Fig. 6A, third blot, lane 8). No band was detected using control IgG for immunoprecipitation (Fig. 6A, third blot, lane 7). Conversely, immunoprecipitation analysis revealed that NS5A protein was coimmunoprecipitated with HNF-1α protein using anti-HNF-1α rabbit PAb (Fig. 6B, fourth blot, lane 8). Moreover, NS5A protein was coimmunoprecipitated with endogenous HNF-1α protein (Fig. 6B, fourth blot, lane 6), suggesting that NS5A protein indeed interacts with HNF-1α protein.

Fig 6.

NS5A protein interacts with HNF-1α protein. (A) Huh-7.5 cells were plated at 1.2 × 10⁶ cells/10-cm dish and cultured for 12 h. Cells were transfected with plasmids as indicated. At 48 h after transfection, cells were harvested. Cell lysates were immunoprecipitated with either anti-FLAG mouse MAb (lanes 2, 4, 6, and 8) or control IgG (lanes 1, 3, 5, and 7), and bound proteins were immunoblotted with anti-HNF-1α rabbit PAb (third blot) or anti-NS5A mouse MAb (fourth blot). Protein expression of HNF-1α or FLAG-NS5A was confirmed using the same cell lysates by immunoblotting with either anti-HNF-1α rabbit PAb (first blot) or anti-NS5A mouse MAb (second blot). (B) Cell lysates were immunoprecipitated with either anti-HNF-1α rabbit PAb (lanes 2, 4, 6, and 8) or control IgG (lanes 1, 3, 5, and 7), and bound proteins were immunoblotted with either anti-HNF-1α rabbit PAb (third blot) and anti-NS5A mouse MAb (fourth blot). (C) Full-genome replicon Con1 (RCYM1) cells were plated at 1.2 × 10⁶ cells/10-cm plate and transfected with or without pCMV-HNF-1α plasmid and cultured for 48 h. Cells were harvested and assayed for immunoprecipitation with anti-HNF-1α rabbit PAb (lanes 2 and 4) or control IgG (lanes 1 and 3). Bound proteins were immunoblotted with anti-HNF-1α goat PAb (third blot) or anti-NS5A mouse MAb (fourth blot). Input samples were immunoblotted with either anti-HNF-1α PAb (first blot) or anti-NS5A MAb (second blot). IP, immunoprecipitation; IB, immunoblotting.

To confirm that HCV NS5A protein can interact with HNF-1α protein in HCV-replicating cells, we performed immunoprecipitation analysis using FGR Con1 (RCYM1) cells. NS5A protein was coimmunoprecipitated with endogenous HNF-1α protein (Fig. 6C, fourth blot, lane 2). Transfection of HNF-1α protein increased the level of coimmunoprecipitated NS5A protein (Fig. 6C, fourth blot, lane 4), suggesting that HCV NS5A protein indeed interacts with HNF-1α protein in HCV-replicating cells.

HNF-1α binds domain I of NS5A protein.

To map the HNF-1α-binding site on NS5A protein, coimmunoprecipitation analyses were performed. By use of a panel of NS5A deletion mutants (Fig. 7A), FLAG-HNF-1α protein was found to coimmunoprecipitate with all of the HA-NS5A proteins except HA-NS5A consisting of aa 357 to 447 [HA-NS5A(357–447), HA-NS5A(250–447), or HA-NS5A(214–447) (Fig. 7B, lower left panel). These results suggest that domain I of NS5A consisting of aa 1 to 213 is important for HNF-1α binding. FLAG-HNF-1α protein was also found to coimmunoprecipitate with NS5A(1–126)-myc-His₆ and NS5A(1–147)-myc-His₆. These data led to the conclusion that the HNF-1α-binding domain of NS5A protein was aa 1 to 126.

Fig 7.

Mapping of the HNF-1α-binding domain for NS5A protein. (A) Schematic representation of the hepatitis C virus NS5A protein. NS5A consists of three domains (domains I, II, and III) with domains separated by low-complexity sequences (LCS I and II). The position of the amino-terminal amphipathic helix membrane anchor is shown (labeled helix). The NS5A deletion mutants (a to j) contain the NS5A amino acids indicated to the left. Each NS5A deletion mutant contains either HA tag in the N terminus (a to h) or myc-His₆ tag in the C terminus (i and j). The gray region of each represents the HA tag sequence. The lattice region of each represents the myc-His₆ tag (i and j). Closed boxes represent proteins that are bound specifically to HNF-1α protein, and open boxes represent those that are not bound. (B) Huh-7.5 cells were transfected with each NS5A mutant plasmid together with a FLAG-HNF-1α expression plasmid. At 48 h posttransfection, cells were harvested, and cell lysates were immunoprecipitated with anti-FLAG beads. Input samples and immunoprecipitated samples were immunoblotted with anti-HA MAb (two left panels, top), anti-c-myc MAb (two right panels, top), or anti-HNF-1α PAb (all panels, bottom).

DISCUSSION

In this study, we aimed to clarify molecular mechanisms of HCV-induced suppression of GLUT2 gene expression. The reporter assays of the human GLUT2 promoter suggest that the HNF-1α-binding site is crucial for HCV-induced suppression of GLUT2 promoter activity (Fig. 2). HCV infection significantly reduced the levels of HNF-1α mRNA (Fig. 3A). Moreover, HCV infection remarkably decreased HNF-1α protein levels (Fig. 4A). Our results suggest that HCV infection suppresses GLUT2 gene expression via NS5A-mediated lysosomal degradation of HNF-1α protein (Fig. 5). Immunoprecipitation analyses revealed that NS5A protein physically interacts with HNF-1α protein (Fig. 6) and that domain I of NS5A is important for HNF-1α binding (Fig. 7). Taken together, our results suggest that HCV infection suppresses GLUT2 transcription via downregulation of HNF-1α expression at both transcriptional and translational levels (Fig. 8).

Fig 8.

A proposed mechanism of the HCV-induced suppression of GLUT2 via downregulation of HNF-1α. HCV infection downregulates HNF-1α at transcriptional and posttranslational levels, resulting in suppression of GLUT2 gene transcription. HCV NS5A protein physically interacts with HNF-1α protein and enhances lysosomal degradation of HNF-1α protein.

We demonstrated that HNF-1α protein levels were greatly reduced compared to the reduced levels of HNF-1α mRNA. We demonstrated that pepstatin A, but not E64-d, restored the levels of HNF-1α protein, suggesting that an aspartic protease is involved in the degradation of HNF-1α protein. Pepstatin A is widely used for investigation of autophagy and lysosomal degradation. Further studies are needed to elucidate how HCV induces lysosomal degradation of HNF-1α protein and how HNF-1α protein is selectively downregulated by HCV infection. Our data suggest that the HCV NS5A protein is responsible for the HCV-induced degradation of HNF-1α protein. Using a panel of NS5A deletion mutants, we demonstrated that domain I of NS5A is important for association with HNF-1α protein. NS5A domain I is relatively conserved among HCV genotypes compared to domains II and III, suggesting that NS5A–HNF-1α interaction is common to all the HCV genotypes. Domain I coordinates a single zinc atom per protein molecule and is essential for HCV RNA replication (35). The crystal structure of NS5A domain I revealed the presence of a zinc coordination motif and a C-terminal disulfide bond (36). NS5A domain I was found to bind many host proteins, RNA, and membranes (16). It is possible that physical interaction between NS5A protein and HNF-1α protein is important for selective degradation of HNF-1α protein. One possible mechanism is that NS5A protein may recruit HNF-1α protein to the lysosome. Further study is necessary to test this possibility.

We observed that deletion of the GLUT2 transcriptional start site enhances expression of the GLUT2 reporter in FGR cells (Fig. 2B). Cha et al. (7) previously reported that deletion down to nucleotide +73 of the GLUT2 promoter resulted in a marked increase and that further deletion to nucleotide +188 caused a drastic decrease in luciferase activity, indicating the presence of negative- and positive-regulator elements in the 5′ untranslated region. The role of these elements in HCV-infected cells remains to be elucidated.

We demonstrated that HCV J6/JFH1 infection reduced the HNF-1α mRNA level and HNF-1α protein level. Our results contradict an earlier report (32) demonstrating that expression of HNF-1 mRNA was increased in subgenomic replicon Huh.8 cells (3). We observed downregulation of HNF-1α mRNA and HNF-1α protein in SGR cells as well as in FGR cells (data not shown). We also demonstrated that the ectopic expression of NS5A protein decreased the endogenous HNF-1α protein level. The reasons for these discrepancies remain to be elucidated.

We along with other groups previously reported that HCV NS5A protein is involved in mitochondrial reactive oxygen species (ROS) production (11, 13, 38). Mitochondrial ROS generation is known to induce the autophagy pathway (22) and lysosomal membrane permeabilization (8). Therefore, it is necessary to determine whether NS5A-induced ROS production enhances autophagic degradation or lysosomal membrane permeabilization. Several groups have reported that autophagy vesicles accumulate in HCV-infected cells and that autophagy proteins can function as proviral factors required for HCV replication (14). Autophagy degrades macromolecules and organelles. Based on the means by which cargo is delivered to the lysosomes, three different autophagy pathways are described: macroautophagy, microautophagy, and chaperone-mediated autophagy (CMA). At first, autophagy was considered a nonselective bulk degradation process. CMA, however, results in specific degradation of the cytosolic proteins in a molecule-by-molecule fashion. Most known substrates for CMA contain a peptide sequence biochemically related to KFERQ (12). Although the typical KFERQ peptide motif is not found in HNF-1α protein, it is possible that KFERQ-like sequences can be generated by posttranslational modifications. It is also possible that HNF-1α protein possesses other degradation motifs. The molecular mechanism underlying NS5A-dependent lysosomal degradation of HNF-1α protein needs to be elucidated.

HNF-1α is a homeodomain-containing transcription factor, which is expressed in the liver, pancreatic β cells, and other tissues (1). Intriguingly, HNF-1α is known to play a crucial role in diabetes. Heterozygous germ line mutations in the gene encoding HNF-1α are responsible for an autosomal dominant form of non-insulin-dependent diabetes, MODY3 (40). Mutations in the HNF-1α gene disrupt GLUT2 function as a glucose sensor in pancreatic β cells, resulting in severe insulin secretory defects (39). It is unclear whether HNF-1α mutations in the liver affect glucose homeostasis in MODY3 patients. Two strains of HNF-1α-deficient mice have been reported. The mice of the first strain, created using standard methods for making knockout mice, are born normally, but most die postnatally around the weaning period after a progressive wasting syndrome (31). Mice of the second strain, created using the Cre-loxP recombination method, had a normal life span (20). The knockout mice of the second strain were dwarfed, diabetic, and infertile. Moreover, the knockout mice had enlarged livers and exhibited progressive liver damage.

HNF-1α was also identified as a tumor suppressor gene involved in human liver tumorigenesis since biallelic inactivating mutations of the HNF-1α gene were found in 50% of hepatocellular adenomas and, in rare cases, of well-differentiated hepatocellular carcinomas developed in the absence of cirrhosis (5). Moreover, HNF-1α has been shown to regulate a large number of genes related to glucose, fatty acid, bile acid, cholesterol, and lipoprotein metabolisms as well as inflammation (1). Therefore, it is possible that HCV-induced downregulation of HNF-1α may play a crucial role in metabolic disorders as well as tumorigenesis.

To determine which HCV protein is involved in the suppression of the GLUT2 promoter, we examined the effects of transient expression of HCV proteins on GLUT2 promoter activity. Overexpression of NS5A suppressed GLUT2 promoter activity, whereas overexpression of p7 enhanced GLUT2 promoter activity (Fig. 5A). SGR cells express NS5A protein but lack p7 protein. FGR cells express both NS5A protein and p7 protein. However, GLUT2 promoter activity was suppressed in both SGR and FGR cells (Fig. 2B). This discrepancy between transient expression system and replicon cells may result from the differences in trafficking of p7 because it is a complex process potentially regulated by both the cleavage from its upstream signal peptides and targeting signals within the protein sequence (15).

We previously reported that HCV infection promotes hepatic gluconeogenesis in HCV J6/JFH1-infected Huh-7.5 cells (11). HCV infection transcriptionally upregulates the genes for phosphoenolpyruvate carboxykinase (PEPCK) and glucose 6-phosphatase (G6Pase), the rate-limiting enzymes for hepatic gluconeogenesis. We demonstrated that gene expression of PEPCK and G6Pase was regulated by the transcription factor forkhead box O1 (FoxO1) in HCV-infected cells. Phosphorylation of the FoxO1 at Ser319 was markedly diminished in HCV-infected cells, resulting in increased nuclear accumulation of FoxO1. HCV NS5A protein was directly linked with FoxO1-dependent increased gluconeogenesis. HCV-induced downregulation of GLUT2 expression and upregulation of gluconeogenesis may cooperatively contribute to development of type 2 diabetes in HCV-infected patients at least to some extent. HCV-induced downregulation of GLUT2 expression and upregulation of gluconeogenesis may result in high concentrations of glucose in HCV-infected hepatocytes. As suggested in a recent study, low glucose concentrations in the hepatocytes inhibit HCV replication (28). Therefore, high glucose levels in the hepatocytes may confer an advantage in efficient replication of HCV.

In conclusion, we provided evidence suggesting that HCV infection downregulates HNF-1α expression at both transcriptional and posttranslational levels. HCV-induced downregulation of HNF-1α may play a crucial role in glucose metabolic disorders caused by HCV infection. Strategies aimed at HCV-induced downregulation of HNF-1α protein may lead to the development of new therapeutic agents for HCV-induced diabetes.