Transforming Growth Factor β Acts as a Regulatory Molecule for Lipogenic Pathways among Hepatitis C Virus Genotype-Specific Infections

Hepatic steatosis is a frequent complication associated with chronic hepatitis C virus (HCV) infection and is a key prognostic indicator for progression to fibrosis and cirrhosis. Several mechanisms are proposed for the development of steatosis, especially with HCV genotype 3a. Our observations suggest that transforming growth factor β (TGF-β) and peroxisome proliferator-activated receptor alpha (PPARα)-associated mechanistic pathways in hepatocytes infected with HCV genotype 2a and 3a differ from those in cells infected with genotype 1a. The results suggest that a targeted therapeutic approach for enhanced PPARα and lipolysis may reduce HCV genotype-associated lipid metabolic disorder in liver disease.

ABSTRACT

Hepatitis C virus (HCV) infection promotes metabolic disorders, and the severity of lipogenic disease depends upon the infecting virus genotype. Here, we have examined HCV genotype 1-, 2-, or 3-specific regulation of lipid metabolism, involving transforming growth factor β (TGF-β)-regulated phospho-Akt (p-Akt) and peroxisome proliferator-activated receptor alpha (PPARα) axes. Since HCV core protein is one of the key players in metabolic regulation, we also examined its contribution in lipid metabolic pathways. The expression of regulatory molecules, TGF-β1/2, phospho-Akt (Ser473), PPARα, sterol regulatory element-binding protein 1 (SREBP-1), fatty acid synthase (FASN), hormone-sensitive lipase (HSL), and acyl dehydrogenases was analyzed in virus-infected hepatocytes. Interestingly, HCV genotype 3a exhibited much higher activation of TGF-β and p-Akt, with a concurrent decrease in PPARα expression and fatty acid oxidation. A significant and similar decrease in HSL, unlike in HCV genotype 1a, was observed with both genotypes 2a and 3a. Similar observations were made from ectopic expression of the core genomic region from each genotype. The key role of TGF-β was further verified using specific small interfering RNA (siRNA). Together, our results highlight a significant difference in TGF-β-induced activity for the HCV genotype 2a- or 3a-induced lipogenic pathway, exhibiting higher triglyceride synthesis and a decreased lipolytic mechanism. These results may help in therapeutic modalities for early treatment of HCV genotype-associated lipid metabolic disorders.

IMPORTANCE Hepatic steatosis is a frequent complication associated with chronic hepatitis C virus (HCV) infection and is a key prognostic indicator for progression to fibrosis and cirrhosis. Several mechanisms are proposed for the development of steatosis, especially with HCV genotype 3a. Our observations suggest that transforming growth factor β (TGF-β) and peroxisome proliferator-activated receptor alpha (PPARα)-associated mechanistic pathways in hepatocytes infected with HCV genotype 2a and 3a differ from those in cells infected with genotype 1a. The results suggest that a targeted therapeutic approach for enhanced PPARα and lipolysis may reduce HCV genotype-associated lipid metabolic disorder in liver disease.

INTRODUCTION

Hepatitis C virus (HCV) is a major cause of chronic liver disease, including metabolic disorders, steatosis, fibrosis/cirrhosis, and hepatocellular carcinoma (HCC) (1). HCV is classified into multiple genotypes. HCV genotype 1 accounts for approximately 42.6%, genotype 2 accounts for 9.1%, and genotype 3 for 30.1% of total HCV cases of infections worldwide (2). Viral factors play a central role in the rate of HCV genotype-associated steatosis progression. We have previously reported that HCV infection or ectopic expression of virus core protein alone promotes metabolic disorders and fibrosis progression (3). Steatosis involves the deposition of triglycerides in the liver of patients with chronic HCV infection (4).

Hepatic steatosis accelerates liver damage and enhances disease progression (5). A high percentage of HCV (∼51%) patients have hepatic steatosis, which is more prevalent in association with genotype 3a infection (6). We have previously shown that HCV infection leads to insulin resistance in human hepatocytes and regulates fork-head box transcription factor O1 (FoxO1) activity contributing to lipid accumulation (7). HCV core protein is relatively conserved among the genotypes and has multiple functions, including the modulation of intracellular signal transduction pathways for metabolic disorders (7, 8). The degree and the mechanisms by which HCV genotypes alter cellular metabolism may vary, although the reason for these differences remains unknown.

We and others have reported the growth of HCV genotypes 1a, 2a, and 3a in human cells of hepatocyte origin, and their replication rates vary (9–11). Several cell lines, including the HepG2 cell line of human hepatocyte origin are used for transfection of HCV core in studying the role of this protein alone. The relationship between HCV core protein expression and cellular lipid metabolism was suggested in an earlier study (12). HCV growth in cell culture and core protein expression in vitro allow for mechanistic analyses of lipid metabolic disorders induced by distinct virus genotypes.

In this study, we focused on understanding the mechanisms of differentially altering lipid metabolism associated with individual HCV genotypes and the involvement of core protein in these processes. Lipid metabolic regulation was studied by analyzing several molecules. The peroxisome activated receptor alpha (PPARα) belongs to a nuclear receptor protein superfamily, and fatty acid metabolism and ketogenesis are the most conserved PPARα-regulated biological processes (13). Xenobiotic metabolism and apolipoprotein synthesis pathways are specially controlled by PPARα agonism in human hepatocytes. Peroxisomal proliferation genes are induced upon activation of PPARα. Deficiency of PPARα leads to exaggerated lipid accumulation in the liver (14). Our observations suggested that PPARα-associated mechanistic pathways, especially decreased fatty acid oxidation, hormone-sensitive lipase (HSL) activation, and lipolysis regulated by HCV genotype 2a or 3a infection promote lipogenic disorders and differ significantly from HCV genotype 1a infection or respective core protein ectopic expression in hepatocytes.

RESULTS

HCV genotype 2a or 3a strongly promotes TGF-β/Akt/PPARα signaling axis.

HCV shows limited growth in cell lines of hepatocyte origin. HCV genotype 1a (clone H77), 2a, and 3a were grown in immortalized human hepatocytes (IHH) for our study. Transforming growth factor β (TGF-β) signaling in hepatocytes contributes to lipid accumulation and promotes the development of steatohepatitis (15). In vivo functions of the three isoforms, TGF-β1, TGF-β2, and TGF-β3, are known to have striking differences, with unique biological importance and functional nonredundancy. Both TGF-β1 and TGF-β2 play roles in immune tolerance and induce fibrosis, while TGF-β3 counteracts tissue fibrosis and has an opposite effect of TGF-β1 in wound healing (16). We have previously shown TGF-β1 or β2 is significantly upregulated in HCV-infected hepatocytes (17). We investigated TGF-β1 and β2 expression, phospho-Akt (p-Akt) (Ser473) activation, and PPARα status as the important regulatory molecules. The levels of biologically active TGF-β1 and β2 monomers were significantly increased in cells infected with HCV genotype 2a (∼12-fold and ∼2-fold, respectively) and 3a (∼28-fold and ∼8-fold, respectively) compared to those in cells infected with genotype 1a (Fig. 1A and B). Our quantitative PCR (qPCR) results for TGF-β1 and β2 followed the same patterns (see Fig. S1A and B in the supplemental material). p-Akt (S473) was elevated to a higher level in hepatocytes infected with cell culture-grown HCV genotype 2a or 3a than in those infected with genotype 1a (Fig. 1C). Free fatty acid (FFA) accumulation activates a potent nuclear PPAR, which is a master regulator of fatty acid metabolism (18). A difference in PPARα expression was observed between HCV genotype infected cells. HCV genotype 1a enhanced PPARα expression (∼2.2-fold), while HCV genotype 2a or 3a infection suppressed PPARα compared to that in mock-treated cells (Fig. 1D). Identical patterns were also observed in PPARα transcripts (Fig. S1C).

FIG 1.

HCV genotypes differentially activate TGF-β and p-Akt and express PPARα expression in infected cells. (A to D) Cell culture-grown HCV genotype 1a-, 2a-, or 3a-infected immortalized human hepatocytes (IHH) were analyzed for expression of TGF-β1, TGF-β2, phosphorylated Akt (S473), and PPARα by Western blotting and normalized with the level of actin expression in each lane. Results from densitometric scanning of the blot are shown on the right. Statistical significance was examined using the two-tailed Student’s t test. *, P < 0.05; **, P < 0.01.

HCV genotype 2a or 3a promotes SREBP-1/FASN expression.

Lipogenesis is tightly regulated by sterol regulatory element-binding protein (SREBP), a transcription factor involved in the regulation of lipogenesis associated genes. SREBPs are synthesized as precursor proteins bound to the endoplasmic reticulum (ER) membrane and undergo proteolytic cleavage after stimulation in the Golgi to release the transcriptionally active N-terminal domain. The mature SREBP then translocate to the nucleus and binds to sterol regulatory elements within its promoter and target genes (19). Activation of Akt signaling modulates SREBP-1 expression. The nuclear translocation of the Akt molecule results in accumulation and enhances the expression of several SREBP-1 target genes (20). To understand the mechanism for activation of SREBP-1 during HCV infection, we examined SREBP-1 status by Western blot analysis using a mouse monoclonal antibody raised against SREBP-1 (amino acids 301 to 407) of human origin. Infection of IHH with HCV genotype 2a (∼2.5-fold) or 3a (∼1.6-fold) increased the precursor SREBP-1 protein band (∼125 kDa) compared to that from infection with genotype 1a (Fig. 2A). However, we could not convincingly detect the mature SREBP-1 band (∼68 kDa). This could be due to the nature of the antibody used or the level of matured protein in our Western blot analysis. Thus, our results demonstrated that HCV genotype 2a or 3a modestly enhances precursor SREBP-1, unlike genotype 1a, and may generate the mature activated protein for translocation into the nucleus for downstream transcriptional activity.

FIG 2.

HCV genotypes 2a and 3a enhance SREBP-1 and FASN expression. (A) Cell culture-grown HCV genotype 1a-, 2a-, or 3a-infected cell lysates were analyzed for expression of SREBP-1 by Western blotting and normalized with actin. (B) Subcellular localization of SREBP-1 (green) and HCV NS5A protein (red) in HCV genotype 3a-infected IHH were analyzed by confocal microscopy. Mock-infected IHH were used as control for comparison. Cells were fixed, permeabilized for staining with DAPI or respective antibodies, and are shown at a magnification of ×60. (C) HCV genotype 1a-, 2a-, or 3a-infected cell lysates were analyzed for expression of FASN by Western blotting and normalized with actin. Densitometric scanning results of the protein bands are shown on the right. Statistical significance was examined using the two-tailed Student’s t test. *, P < 0.05; **, P < 0.01.

To determine whether the enhanced precursor form of SREBP-1 correlates with its activation, we performed immunofluorescence analysis to examine translocation of the mature protein in its active form to the cell nucleus. This activation plays a crucial role for enhanced nuclear localization of SREBP-1, thereby increasing its ability to activate the transcription of target genes in the nucleus. To understand the effect of HCV genotype 3a infection on nuclear localization of SREBP-1, IHH were infected with HCV genotype 3a, and after 3 days, cells were stained for the detection of endogenous SREBP-1 and HCV NS5A protein by immunofluorescence (Fig. 2B). Mock-infected cells displayed relatively low SREBP-1 primarily outside the nucleus compared to that in virus-infected cells. Together, our results suggested that HCV genotype 3a infection of hepatocytes enhances SREBP-1 expression and promotes localization into the nucleus in its active form.

Fatty acid synthase (FASN) is one of the primary enzymes associated with triglyceride synthesis, and its expression is directly regulated by SREBP-1. We examined FASN protein status by Western blot analysis. A modest increase in expression of FASN (∼1.4-fold) was observed in cells infected with cell culture-grown HCV genotype 2a or 3a (Fig. 2C). From qPCR results, we also observed that mRNA expression of FASN was increased by genotype 2a or 3a infection (Fig. S1D). Thus, HCV genotype 2a or 3a promotes the lipogenesis-related signaling cascade more efficiently than HCV genotype 1a. Together, our results indicate that infectious HCV genotype 2a or 3a plays a role in promoting SREBP-1 synthesis and activating nuclear localization, leading to a more efficient de novo lipogenic response compared to that with genotype 1a.

HCV genotype 2a or 3a exhibits reduced fatty acid oxidation and lipolysis-associated proteins.

Medium-chain acyl coenzyme A (acyl-CoA) dehydrogenase (MCAD) and short-chain acyl-CoA dehydrogenase (SCAD) are involved in catalyzing the first step in the β-oxidation of fatty acids (21). Here, we analyzed the expression of acyl-CoA dehydrogenases, MCAD and SCAD, in cells infected with each HCV genotype. Fatty acid accumulation seen in HCV genotype 1a-infected cells may trigger MCAD accumulation via activation of PPARα (22). Indeed, levels of MCAD and SCAD were observed to be enhanced in cells infected with HCV 1a. In contrast, the levels of MCAD and SCAD were decreased in cells infected with HCV genotype 2a or 3a, unlike with HCV 1a (Fig. 3A and B). We also found that the MCAD mRNA expression was increased with HCV genotype 1a but not with 3a infection (Fig. S1E). The results indicated that the functional inhibition of acyl-CoA dehydrogenase activity is common to HCV genotypes 2a and 3a, in contrast to genotype 1a, which agrees with the differential regulation of PPARα (shown in Fig. 1C).

FIG 3.

HCV genotype 1a infection enhances lipolytic enzymes expression. IHH were infected with HCV genotype 1a, 2a, or 3a. (A to C) Infected cells were analyzed for expression of MCAD, SCAD, and phosphorylated HSL (S563) by Western blot and normalized with actin level. Densitometric scanning results of the protein bands are shown on the right of each panel. Statistical significance was analyzed using the two-tailed Student’s t test. *P < 0.05, **P < 0.01.

Protein kinase A (PKA) increases the hydrolytic activity of HSL by the phosphorylation of a single site identified at Ser563. Akt activation has been observed to lead to HSL inactivation (23), and increased activation of HSL (S563) has been associated with an acceleration of lipolysis in a murine model. The regulation of HSL is dependent on PPARα activation (24). In agreement with our data, HCV genotype 1a infection enhanced p-HSL (S563) expression, while infection with genotype 2a and 3a decreased HSL activation (Fig. 3C). These results suggest that the lipolytic pathway may be inhibited by HCV 2a and 3a infection due to negative regulation of PPARα.

Ectopic expression of HCV core protein from genotype 2a or 3a strongly promotes TGF-β and Akt activation as seen in cell culture-grown virus infection.

Studies have reported that HCV core protein modulates TGF-β signaling (25, 26). We examined how core proteins from different HCV genotypes may modulate TGF-β expression. The HepG2 cell line was chosen for transfection with the HCV core gene to study the role of this virus protein alone upon TGF-β and Akt activation. The results from our Western blot analysis suggested that the biologically active monomeric TGF-β2 protein (∼12 kDa) was significantly elevated in HepG2 cells transfected with HCV core from genotype 2a (∼1.8-fold) or 3a (∼2.6-fold) compared to that from genotype 1a (Fig. 4A). To examine the importance of the biologically active TGF-β molecule in our study, we performed a TGF-β1/2/3 knockdown experiment using small interfering RNA (siRNA) transfection. siRNA treatment downregulated TGF-β2 precursor as well as active monomer expression in HCV core-transfected HepG2 cells (Fig. 4B). It has been established that TGF-β signaling directly activates Akt (27), and our result indicated that HCV core from genotype 2a or 3a promoted significant Akt activation (Fig. 4C). Furthermore, to examine whether knockdown of TGF-β affects Akt activation, we performed Western blot analysis for p-Akt in HCV genotype core 1a- and 3a-transfected cells. Our result showed that knockdown of TGF-β suppressed p-Akt activation (Fig. 4D). Furthermore, a reciprocal experiment using the addition of external TGF-β in IHH or HepG2 cells resulted in elevation of p-Akt, irrespective of the presence of HCV or HCV core protein (Fig. 4E). Taken together, our results suggested that TGF-β plays an important role in Akt activation. In addition, higher SREBP-1 expression was observed in genotype 2a or 3a core-transfected cells (Fig. 5A). Furthermore, a modest increase in FASN expression in HepG2 cells expressing HCV core from genotype 3a compared to that from 1a was detected by enzyme-linked immunosorbent assay (ELISA) (Fig. 5B). Therefore, HCV genotypes differentially regulate TGF-β activation to control Akt-mediated regulation of the lipogenic pathway.

FIG 4.

Differential expression of TGF-β and p-Akt regulation by core protein from HCV genotypes. (A) TGF-β2 expression was examined by Western blotting of HepG2 cells transfected with HCV core from genotypes 1a, 2a, and 3a and compared with mock-transfected control. (B) The role of core protein from genotype 1a and 3a upon TGF-β2 expression was separately verified using specific siRNA in transfected cells. Expression level of actin in each lane is shown as a loading control for comparison with expression of other proteins. (C and D) Densitometric scanning results of the protein bands are shown at the bottom of each panel. Phosphorylated Akt (S473) level was analyzed similarly. (E) Effect of exogenous TGF-β on Akt activation (S473) in naive IHH and HepG2 cells was also analyzed. Statistical significance using the two-tailed Student’s t test is shown. *, P < 0.05; **, P < 0.01.

FIG 5.

SREBP-1 and FASN are elevated in HepG2 cells transfected with HCV genotype 2a and 3a core proteins. (A) SREBP-1 expression was examined in HepG2 cells transfected with HCV core from genotypes 1a, 2a, and 3a and compared with mock-transfected control by Western blotting. Expression level of actin is shown as a loading control for comparison of SREBP-1 protein expression. Densitometric scanning results of the protein bands are shown at the bottom. (B) FASN level in core transfected cell lysates were separately analyzed by ELISA. Statistical significance using the two-tailed Student's t test is shown. *, P < 0.05; **, P < 0.01.

HCV core from genotype 1a does not suppress PPARα, unlike genotype 2a or 3a.

Core protein expression from HCV 1a did not significantly change or suppress PPARα, while core from HCV genotype 2a or 3a suppressed PPARα expression (Fig. 6A). TGF-β controls PPARα expression (28, 29). To understand whether PPARα is regulated primarily by active TGF-β generated from HCV core from genotype 1a and 3a transfection, we performed Western blot analysis for PPARα expression in TGF-β (isoforms 1, 2, and 3) knockdown cells. Introduction of TGF-β siRNA enhanced PPARα, suggesting its regulatory role on PPARα expression (Fig. 6B). These results suggested that PPARα expression is modulated via a TGF-β-mediated mechanism regardless of the presence of HCV core. Furthermore, addition of exogenous TGF-β in IHH or HepG2 cells resulted in inhibition of PPARα expression to different extents in both the cell lines (Fig. 6C) in the absence of HCV core protein.

FIG 6.

Differential expression of lipolytic pathway-related key molecules by HCV core protein from distinct genotypes. (A) PPARα expression was examined by Western blot in HCV core from genotype 1a-, 2a-, and 3a-transfected HepG2 cells and compared with mock-transfected controls. Expression level of actin in each lane is shown as a loading control for comparison with expression of other proteins. Densitometric scanning results of the protein bands are shown at the bottom. (B) The role of core protein from genotype 1a and 3a on TGF-β expression and its consequence on PPARα were separately verified using specific siRNAs in transfected cells. (C) Effect of exogenous TGF-β on PPARα expression in naive IHH and HepG2 cells was also analyzed. (D) The roles of HCV core protein from genotypes 1a, 2a, and 3a on phosphorylation and activation of HSL (S563) in transfected HepG2 cells were compared with that in mock-transfected controls by Western blotting. (E) HCV core transfected cells were analyzed for total triglyceride level by ELISA and compared with mock-transfected control cells. Statistical significance was analyzed using the two-tailed Student's t test. *, P < 0.05; **, P < 0.01.

Lipogenesis regulation by HCV core protein.

The regulation of HSL is dependent on PPARα activation (24). Transfection with HCV core from different genotypes displayed a strong difference in p-HSL (S563) status. HCV core from genotype 1a did not significantly alter the activation of HSL (S563), while core from genotype 2a or 3a significantly suppressed p-HSL (S563) (Fig. 6D). HCV genotype 3a infection is known to exert a stronger effect leading to fatty liver generation. We examined triglyceride synthesis in HCV core-transfected HepG2 cells by ELISA. Plasmid DNAs encoding the core genomic region from HCV genotypes 1a, 2a, and 3a were separately transfected into HepG2 cells and triglyceride synthesis was measured. An approximately 2.5-fold higher triglyceride level was observed in genotype 2a- or 3a-transfected cells than in genotype 1a-transfected or control cells (Fig. 6E). The expression levels of core protein were qualitatively measured from cell lysates by ELISA (Ortho Diagnostics kit; Ortho Clinical Diagnostics, Japan). This commercially available kit was used, since HCV core is very much conserved among the genotypes and monoclonal antibodies to conserved epitopes are included in this assay (30). Similar levels of core protein expression for genotypes 1a and 2a (∼18,000 fmol/liter) were observed, while core from genotype 3a was expressed at a considerably lower level (∼3,000 fmol/liter). These results were consistent from three transfection experiments.

DISCUSSION

We have investigated the mechanisms for hepatic lipid metabolic regulation associated with HCV genotype 1a, 2a, or 3a infection of cells of hepatocyte origin. Our novel findings include stronger TGF-β activation in cells infected with HCV genotypes 2a and 3a than with HCV genotype 1a infection (Fig. 7). Infection with HCV genotype 2a or 3a induced a significant difference in the magnitude of lipogenic pathways (TGF-β, PPARα, and decreased lipolysis) compared to that with HCV genotype 1a. HCV genotype 2a or 3a infection also decreased fatty acid oxidation compared to that with genotype 1a. HCV proteins are known to have distinct cell regulatory properties. HCV NS3 and NS5A proteins are known to bind with TGF-β receptor and modulate TGF-β signaling but do not modulate TGF-β expression (31, 32). Here, we observed HCV core from genotype 2a or 3a increased Akt and strongly enhanced SREBP-1, unlike core from 1a. A difference in HCV replication capacity or the level of core protein expression is unlikely to be the reason for the observed genotype-specific qualitative effect. This is rationalized by the fact that HCV genotype 1a displayed an opposite functional regulation even with a relatively weak virus genome replication in cell culture. An increase in SREBP-1/FASN expression in HCV genotype 2a- or 3a-infected cells and core-expressing cells, coupled to a loss of PPARα, led to an alteration of triglyceride accumulation. The serum triglyceride level also positively associates with HCV genotype 3 infection compared to that in steatosis patients infected with the other genotype (33). Lipogenesis is primarily controlled by SREBP-1 at the transcriptional level (34). SREBP-1 is one of the key regulators of fatty acid synthesis and acts as a vital building block for endogenous biosynthesis of triglyceride through FASN (35). HCV genotype 2a- or 3a-infected hepatocytes exhibited a modest increase in expression of SREBP-1, which was mirrored in genotype-specific core-transfected hepatocytes. SREBP-1 controls both cholesterol and fatty acid synthesis by directly influencing multiple downstream enzymes, including FASN. FASN plays a key role in the biosynthesis of fatty acids and is upregulated during HCV genotype 1a infection (36). An earlier study reported that HCV core from genotype 3a enhanced SREBP-1 function via a phosphoinositide 3-kinase (PI3K)/Akt-associated pathway (37). However, cells infected with HCV genotype 1a, 2a, or 3a exhibited increased p-Akt expression compared to mock-infected control cells to various degrees. SREBP-1 promotes the expression of lipogenic genes (35) through a PI3K/Akt/mTOR-dependent mechanism (38).

FIG 7.

Schematic diagram showing mechanism of HCV-associated lipid metabolic regulation. HCV genotype 1a significantly differs from genotypes 2a and 3a in the regulation of TGF-β and PPARα axes for dampening lipid catabolism.

It is unlikely that core protein expression level alone will have an impact on the expression of TGF-β, p-Akt, and PPARα expression and fatty acid oxidation, as the expression level in genotype 3a was significantly lower (∼6-fold), although the associations with TGF-β were different. Only HCV core protein can regulate TGF-β expression (25). We speculate that sequence variations in the core protein (26), between 71 and 165, among HCV genotypes may have a role in TGF-β regulation. A low concentration of core protein from HCV genotype 3a appears to be optimal, with a saturation effect for TGF-β upregulation. Increased core protein expression beyond its threshold may not impart a difference in the regulatory effect.

On the other hand, we have observed that HCV core from genotype 2a or 3a increased Akt and strongly enhanced SREBP-1, unlike core from genotype 1a. A difference in HCV replication capacity or the level of core protein expression alone is unlikely to be the reason for the observed genotype-specific qualitative effect. This is further rationalized by the fact that HCV genotype 1a displayed an opposite functional regulation even with a relatively weak virus genome replication in cell culture. An increase in SREBP-1/FASN expression in HCV genotype 2a- or 3a-infected cells and core-expressing cells, coupled to a loss of PPARα, led to alteration of triglyceride accumulation.

HCV core protein induces changes in lipid metabolism as well as the formation and redistribution of lipid droplets (39–42). HCV core protein contains RNA binding domains for suppressing the transcriptional activity of PPARα (43) and may be a primary cause in HCV-mediated PPARα downregulation. Studies using HCV core (with matching genotype 1a sequence) transgenic mice showed PPARα activation (44). The core protein serves as a coactivator and nuclear stabilizer of PPARα and may transactivate PPARα through extracellular signal-regulated kinase 1 or 2 (ERK1/2) activation and p38 mitogen-activated protein kinase (MAPK) phosphorylation (45). Interestingly, HCV core protein is highly conserved among different genotypes, except in minor regions, but sequence variations in core protein may cause significant differences in pathogenesis from distinct HCV genotypes (46). HCV core contributes to a transcriptional regulatory role, and specific sequence variation may alter these activities, for example, with PPARα. HCV infection modulates PPARα-mediated pathways affecting hepatic metabolism. PPARα mRNA and protein levels are significantly decreased in hepatitis C-associated steatohepatitis compared with that in nonsteatotic livers (47). Chronic HCV genotype 3a infection is associated with an alteration of PPARα in liver biopsy specimens of steatosis patients (48). PPARα in human liver is regulated by miR-21 and miR-27 (49). HCV genotype 2a infection induces miR-21 and miR-27 and negatively regulates PPARα protein status (50, 51). Interestingly, a significant downregulation of miR-21 in HCV genotype 4a-infected patients was reported (52). Therefore, HCV genotype variations may differentially regulate PPARα expression by microRNAs (miRNAs).

PPARα activation is associated with increased levels of steatosis, inflammation, and fibrosis in preclinical models of nonalcoholic fatty liver disease (NAFLD) (14). Inhibition of PPARα leads to decreased triglyceride breakdown and accumulation of fatty acids. Our study suggests that HCV genotype 2a or 3a decreased PPARα in vitro (43, 53), unlike HCV genotype 1a. Core protein expression from HCV genotypes 2a and 3a in HepG2 cells also markedly decreased the PPARα level. The mRNA expression of PPARα in liver biopsy specimens from HCV genotype 3a-infected patients was reported to be decreased significantly (53). These results indicated that the inhibition of PPARα by HCV 2a or 3a contributes to the buildup of triglycerides in hepatocytes compared to that by HCV genotype 1a. The MCAD and SCAD proteins are members of the acyl-CoA dehydrogenase family and catalyze the first step in the β-oxidation of straight-chain fatty acids. Many PPARα target genes are involved in fatty acid metabolism in tissues with high oxidative rates, such as in the liver. PPARα regulates mitochondrial β-oxidation by directly controlling MCAD, long-chain acyl-CoA dehydrogenase (LCAD), and very-long-chain acyl-CoA dehydrogenase (VLCAD) expression (54). Here, we have observed a significant reduction in PPARα, MCAD, and SCAD synthesis associated with HCV genotypes 2a and 3a. These results indicate that the inhibition of MCAD and SCAD may contribute to decreased free fatty acid oxidation in HCV-infected hepatocytes.

Intracellular lipases and their lipolytic products may also play important roles in fatty liver disease. Hormone-sensitive lipase (HSL) is a potential target for the treatment of lipid disorders. HSL is a key enzyme in the mobilization of fatty acids from stored triacylglycerols. Lipolysis is associated with phosphorylation of HSL and translocation to lipid droplets. HSL acts to further remove fatty acids from the triglyceride backbone (55). We have observed a significant reduction in HSL activation by HCV genotype 2a and 3a infection, or expression of their core proteins, but not from HCV genotype 1a. The results indicated that the inhibition of p-HSL may contribute to decreased lipolysis.

Biologically active TGF-β modulates lipid metabolism, initiating the development of steatohepatitis (15, 56, 57). Several studies have shown that HCV core protein contributes to the activation of TGF-β signaling (25, 26). TGF-β impairs the hepatocyte nuclear factor-4α (HNF4α) transcriptional regulator, which in turn enhances PPARα promoter activity (28, 29). This agrees with our observation of PPARα mRNA suppression in HCV genotype 2a- and 3a-infected cells, where TGF-β expression was higher. On the other hand, TGF-β signaling directly or indirectly activates Akt (27). Our study established that silencing of TGF-β resulted in upregulation of PPARα and downregulation of activated Akt. Indeed, our results suggested that core protein sequence variation within HCV genotypes provide a reason for differential expression of active TGF-β molecules. Our study also revealed a mechanism for HCV genotype 2a- or 3a-associated lipid accumulation by stronger activation of TGF-β. Thus, our observations may help in determining therapeutic modalities for treatment of HCV genotype-associated lipid metabolic disorders.

MATERIALS AND METHODS

Cell culture grown virus.

HCV shows limited growth in selected cell lines of hepatocyte origin. HCV genotype 1a (clone H77) was grown in IHH, as reported previously (9), while IHH was used for HCV genotype 2a growth (7). HCV genotype 3a full-length recombinant cDNA (DBN3acc) was kindly provided by Jens Bukh (University of Copenhagen, Denmark). HCV genotype 3a was grown in IHH. Cell culture supernatant was filtered through a 0.45-μm cellulose acetate membrane (Nalgene, NY), aliquoted as virus stock, rapidly frozen on dry ice, and stored at −70°C for single use. HCV RNA in virus stock was quantified (IU/ml) by real-time PCR (Prism 7500 real-time thermocycler; Applied Biosystems, Foster City, CA) with the use of analyte-specific reagents (Abbott, Chicago, IL) in the Pathology Clinical Laboratory of Saint Louis University. Hepatocytes were used in all experiments associated with HCV infection and verified by immunofluorescence for virus core protein detection using specific monoclonal antibody (C7-50; Santa Cruz Biotechnology). All the experiments were performed using a similar multiplicity of infection (MOI).

Cells and transfections.

HepG2 was chosen as a different cell line of human hepatocyte origin from other cells supporting infectious HCV growth and used for transfections to study the function of core protein alone. Cells were maintained in Dulbecco’s modified Eagle medium (DMEM), supplemented with 10% fetal bovine serum and penicillin-streptomycin at 37°C in a 5% CO₂ atmosphere. Cells were transfected with HCV core genomic region from genotype 1a under the control of a cytomegalovirus (CMV) early promoter in pcDNA3 expression vector (pcDNA3-core 1a). HCV core genomic region from genotype 2a or 3a was cloned into plasmid DNA ligated in the pIRES vector (pIRES-core 2a or 3a, respectively) and used in transfections utilizing Lipofectamine 3000 (Thermo Fisher). The pIRES-core plasmids were kindly provided by Francesco Negro (University Hospital of Geneva, Switzerland). Parental HepG2 cells transfected with empty vector DNA were used in parallel as a negative control for comparison. TGF-β (all three isoforms-specific siRNA was introduced into cells by a similar transfection method using Lipofectamine 3000.

qPCR analysis.

Total RNA was isolated from HCV-infected cells using TRIzol reagent (Invitrogen). cDNA was synthesized using a Superscript III reverse transcriptase kit (Invitrogen) with a random hexamer according to the manufacturer’s protocol. qPCR was performed for quantitation of gene expression using SYBR green PCR master mix and detected by real-time PCR (Applied Biosystems). For detection, TGF-β1 (Sigma identifier [ID] H1_TGFB1), TGF-β2 (Sigma ID H1_TGFB2), PPARα (Bio-Rad ID qHsaCID0011001), FASN (Sigma ID H1_FASN), and MCAD (Bio-Rad ID qHsaCID0010107) were used. GAPDH (forward primer 5′-CATGTTCGTCATGGGTGTGAACCA-3′ and reverse primer 5′-AGTGATGGCATGGACTGTGGTCAT-3′) was used as endogenous control. The relative gene expression was analyzed by using the threshold cycle (2^ΔΔCT) formula (ΔΔC_T = ΔC_T of the sample – ΔC_T of the untreated control).

Western blot analysis.

Proteins from HCV-infected or core-transfected cells were separated and transferred onto a nitrocellulose membrane (Bio-Rad Laboratories). The blot was blocked with 5% skim milk and incubated with commercially available antibodies to phospho-Akt (S473), Akt, phospho-FoxO1, and phospho-HSL (Cell Signaling Technology) or to SREBP-1, FASN, PPARα, MCAD, TGF-β1, TGF-β2, and SCAD (Santa Cruz Biotechnology), followed with a secondary antibody conjugated to horseradish peroxidase (HRP) (Bio-Rad Laboratories). The membrane was reprobed with enzyme-conjugated antibody to actin (Santa Cruz Biotechnology) as an internal control. The protein bands were detected with Super Signal West Pico ECL reagents (Pierce). The bands were scanned densitometrically using Image J software (NIH) and compared with those for the internal control.

FASN and triglyceride and core measurements.

FASN level was detected from infected or control sonicated cell lysates by a total fatty acid synthase sandwich ELISA kit (PathScan; Cell Signaling Technology), and absorbance was measured at 450 nm. Triglyceride levels were detected in mock- or core-transfected cell lysates using a triglyceride colorimetric assay kit (Cayman Chemical) and measured at 530 nm. HCV core antigen (HCVcAg) was determined in transfected cell lysates by ELISA (Ortho Clinical Diagnostics, Tokyo, Japan). The assay was performed according to the supplier's protocol. For this, HepG2 cells were transiently transfected with pIRES-core 2a or 3a or pcDNA-core 1a. Cell lysates were prepared after 48 h using a cell lysis buffer. The assay uses a capture ELISA (enzyme‐linked immunosorbent assay) format, wherein the solid phase is coated with two monoclonal antibodies (c11-3 and c11-7) with specificity to different regions of the core antigen. Conjugate-containing horseradish peroxidase-labeled monoclonal antibodies against HCVcAg (c11-10 and c11-14) were added to each well of the plate for detection of bound HCV core protein. The sensitivity of the HCVcAg assay is 93.4, 100, and 100 for genotypes 1, 2, and 3, respectively, in comparison with that of the reverse transcription-PCR (RT-PCR) assay (30).

Immunofluorescence.

HCV core plasmid DNA-transfected HepG2 cells were incubated for 72 h and fixed with 10% formalin. Cells were incubated with a purified rabbit anti-NS5a antibody (58) or SREBP-1 antibody (Cell Signaling) overnight at 4°C. Alexa flour 488 (green) for SREBP-1 and 568 (red) for NS5A were used to visualize stained virus protein or cellular protein, respectively. Cells were viewed and imaged using a confocal microscope (Olympus).

Statistical analysis.

Data were analyzed by Student's t tests with a two-tailed distribution. A P value of <0.05 was considered statistically significant. The results are presented as the mean ± standard deviation from at least three independent experiments.