Hepatitis C Virus Core Protein Activates Nuclear Factor k B-dependent Signaling through Tumor Necrosis Factor Receptor-associated Factor*

Hepatitis C virus (HCV) core protein, a viral nucleo-capsid, has been shown to affect various intracellular events including the nuclear factor k B (NF- k B) signaling supposedly associated with inflammatory response, cell proliferation, and apoptosis. In order to elucidate the effect of HCV core protein on the NF- k B signaling in HeLa and HepG2 cells, a reporter assay was utilized. HCV core protein significantly activated NF- k B signaling in a dose-dependent manner not only in HeLa and HepG2 cells transiently transfected with core protein expression plasmid, but also in HeLa cells induced to express core protein under the control of doxycycline. HCV core protein increased the DNA binding affinity of NF- k B in the electrophoretic mobility shift assay. Acetyl salicylic acid, an IKK b -specific inhibitor, and dominant negative form of IKK b significantly blocked NF- k B activation by HCV core protein, suggesting HCV core protein activates the NF- k B pathway mainly through IKK b . Moreover, the dominant negative forms of TRAF2/6 significantly blocked activation of the pathway by HCV core protein, HCV core protein mimics proinflammatory cytokine activation of the NF- k B pathway through TRAF2/6. In fact, HCV core protein activated b NF- k B pathway. this of HCV core may play an GAGTCAAGCGGAAGCTGGGATGGTC-3 9 , and then cloned into pCXN2. For construction of pCXN2-HAcore151, the region of HCV core aa 1–151 was amplified with F1 and R453 (an antisense primer, nt 438–453), 5 9 -CCGCTCGAGTCACAGGGCCCTGGCAGCG-3 9 , and then cloned into pCXN2-HA. Similarly, pCXN2-HAcore92–191, which encodes N-terminal 91-aa deleted core protein, was constructed by using F276 (a sense primer, nt 276–292), 5 9 -CCGCTCGAGACCATGGGGT-GGGCAGGATGGCT-3 9 and R573. To generate a tetracycline-regulated HCV core expression plasmid (pTRE2-core), full-length HCV core cDNA was prepared by digesting pCXN2-core with Xho I and then subcloned into the Sal I site of pTRE2 (CLONTECH). This construct allowed for the expression of HCV core protein under the transcriptional control of a tetracycline-controlled transactivator-dependent promoter. A series of core expression plasmids were sequenced using a cycle DNA sequencing system (PE Applied Biosystems, Foster City, CA), as described previously (19), to confirm the integration of core genes. luciferase assay dual sea pansy luminometer LB9507, Bad Wildbad, were triplicate. luciferase luciferase activity normalized

Hepatitis C virus (HCV), 1 a member of the Flaviviridae family, is one of the major causes of chronic hepatitis, which can result in cirrhosis and finally hepatocellular carcinoma (1). Its genome consists of a linear, positive-strand RNA molecule of ϳ9,500 nucleotides (nt) encoding a single polyprotein precursor of ϳ3,000 amino acids (aa) that is processed into three to four structural proteins at the amino terminus (Core, E1, and E2/p7) and six nonstructural proteins at the carboxyl terminus (nonstructural proteins 2, 3, 4a, 4b, 5a, and 5b) (2,3). The genomic region of the putative core protein encodes 191 aa and has an apparent molecular mass of 21 kDa (2). The core protein, relatively conserved among all identified HCV isolates (4), may be the fundamental unit for the encapsidation of genomic RNA to help in virus morphogenesis. In addition, previous studies suggested that HCV core protein has various biological properties, one of which is its effect on the nuclear factor B (NF-B) pathway (5)(6)(7)(8)(9).
NF-B belongs to a highly conserved Rel-related protein family, which includes RelA (p65), RelB, c-Rel, NF-B1 (p105/ p50), and NF-B2 (p100/p52). Of these, the p50/p65 heterodimer, commonly referred to as NF-B, is the most abundant and ubiquitous. One of the most intensively studied signals to the NF-B is induced by tumor necrosis factor (TNF), a proinflammatory cytokine associated with inflammation, immune response, and apoptosis. Currently, this signal transduction pathway is understood as follows (10 -15); when TNF binds and activates the TNF receptor 1 (TNFR1), TNFR-associated death domain (TRADD), TNFR-associated factor 2 (TRAF2), and receptor interacting protein (RIP) form a complex with TNFR1. Subsequently NF-B inducing kinase (NIK) and/or mitogen-activated protein kinase/extracellular signalregulated kinase kinase 1 (MEKK1) are activated. Activated NIK and/or MEKK1 phosphorylate and activate both IB kinase (IKK) ␣ and IKK␤. Activated IKK␣/␤ phosphorylate IB␣, which associates with and sequesters NF-B in the cytoplasm. Phosphorylated IB␣ is ubiquitinated and degraded, and then NF-B translocates into the nucleus and binds to DNA to initiate the transcription of various genes associated with inflammation, the immune response, cell growth, and survival.
There have, however, been conflicting reports up until now about the effect of HCV core protein on this NF-B pathway. Recently, it was shown that HCV core protein activated the NF-B pathway (7)(8)(9). On the contrary, it was previously shown that core protein suppressed TNF-induced NF-B activation (5,6). At present, the mechanism for core protein's effect on the NF-B pathway remains unclear; therefore, we focused our attention on the effect HCV core protein has on the NF-B pathway and tried to determine how core protein affects NF-B signaling.

EXPERIMENTAL PROCEDURES
Cell Lines-Human cervical carcinoma cells (HeLa), human hepatoma cells (HepG2), and monkey kidney cells (COS-7) were obtained from the RIKEN cell bank (Tsukuba Science City, Japan). HeLa Tet-Off cells, which constitutively express the tetracycline-controlled transactivator, were purchased from CLONTECH (Palo Alto, CA). Cells were grown in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) * This work was supported in part by the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Pharmaceutical Safety and Research of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
HCV Core Protein-expressing Plasmids-Type 1b HCV core region (nt 1-575 and aa 1-191 of the prototype HCV type 1b, HCV-J; Ref. 16) was amplified by reverse transcription-polymerase chain reaction (PCR) using the HCV RNA extracted from the sera of a patient with chronic hepatitis C as a template. Reverse transcription-PCR was performed as previously described (17) using the following primers having an XhoI restriction site (underlined) (the nucleotide positions in HCV-J are shown in parentheses): F1 (a sense primer, nt 1-20), 5Ј-CCGCTC-GAGACCATGAGCACGAATCCTAAACC-3Ј, and R573 (an antisense primer, nt 555-573), 5Ј-CCGCTCGAGTCAAGCGGAAGCTGGGATG-GTC-3Ј. The amplified product was digested with XhoI and then cloned into the XhoI site of pCXN2 (kindly provided by J. Miyazaki, Osaka University, Osaka, Japan), a mammalian expression vector having a ␤-actin promoter and cytomegalovirus enhancer (18), to generate pCXN2-core.
To generate a tetracycline-regulated HCV core expression plasmid (pTRE2-core), full-length HCV core cDNA was prepared by digesting pCXN2-core with XhoI and then subcloned into the SalI site of pTRE2 (CLONTECH). This construct allowed for the expression of HCV core protein under the transcriptional control of a tetracycline-controlled transactivator-dependent promoter.
A series of core expression plasmids were sequenced using a cycle DNA sequencing system (PE Applied Biosystems, Foster City, CA), as described previously (19), to confirm the integration of core genes. Expression of core proteins was examined by the ECL-plus Western blotting detection system (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom), using extracts of COS-7 cells transfected with core expression plasmids and mouse anti-HCV core antigen IgG fraction (Austral Biologicals, San Ramon, CA) or anti-HA polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA).
NF-B Pathway Reporter Plasmids-To examine the effect of HCV core protein on the NF-B pathway, NF-B-inducible reporter plasmid (pNF-B-Luc) containing the Photinus pyralis (firefly) luciferase reporter gene driven by a basic promoter element (TATA box) plus five repeats of B cis-enhancer element (TGGGGACTTTCCGC) (Stratagene, La Jolla, CA) was utilized. pFC-MEKK (Stratagene), which expresses constitutively active MEKK1 (amino acids 360 -672) driven by a cytomegalovirus promoter, was used as a positive control plasmid to activate the NF-B pathway. pRL-TK (Promega, Madison, WI), a control plasmid that expresses Renilla reniformis (sea pansy) luciferase driven by the herpes simplex virus thymidine kinase promoter, was used to monitor the efficiency of transfection.
Reporter Plasmids Having Interleukin (IL)-1␤ and TNF␣ Promoters-In addition to the reporter plasmid having synthetic B cis enhancer element, luciferase reporter plasmids having IL-1␤ or TNF␣ promoter containing NF-B binding sites, were constructed. A fragment of 691 base pairs (positions Ϫ585 to ϩ106) of the TNF␣ gene was amplified by PCR using a primer set of 5Ј-ggggtaccgcttgtcctgctacccc-3Ј and 5Ј-cccaagcttgtcaggggatgtggcgt-3Ј. The amplified fragment was digested with KpnI and HindIII, and then cloned into pGL3 basic vector (Promega) (pTNF␣-Luc). Similarly, a fragment of 1125 base pairs (Ϫ1110 to ϩ15) of IL-1␤ gene was amplified by PCR using primer set of 5Ј-ggggtacccctgtagtcccagctg-3Ј and 5Ј-ctagctagctcgaagaggtttggtatct-3Ј. Those fragments were digested with KpnI and NheI, and then cloned into pGL3 basic vector (pIL-1␤-Luc).
All cloned plasmids were purified using the Endfree plasmid kit (Qiagen, Hilden, Germany). Nucleotide sequencing of constructed plasmids was performed using an autosequencer (PE Applied Biosystems) and the dye termination method as described previously (19) to confirm gene expression.
Construction of HeLa Cells Induced to Express Core Protein-HeLa cells induced to express HCV core protein were generated with use of a tetracycline-regulated gene expression system (Tet-Off gene expression system, CLONTECH). HeLa Tet-Off cells were cotransfected with pTRE2-core and pTK-Hyg (CLONTECH), a selection vector that confers hygromycin resistance, followed by selection in culture medium containing 200 g/ml hygromycin (CLONTECH) and 1 g/ml doxycycline (Sigma). Hygromycin-resistant clones, termed HeTOC, were examined for expression of the core protein upon withdrawal of doxycycline by Western blotting using mouse anti-HCV core antigen IgG fraction, as described previously . Positive clones were expanded and rescreened by Western blotting of cells grown in the presence and absence of doxycycline.
Transfection-We used the Effectene transfection reagent (Qiagen) for all transfection experiments. Approximately 4 ϫ 10 5 HeLa cells were plated into the well of a six-well tissue culture plate (Iwaki Glass, Chiba, Japan) 24 h before transfection. To examine the effect of HCV core protein on the NF-B pathway, HeLa cells were transfected with a total of 0.4 g of plasmids consisting of 0.19 g of pNF-B-Luc, 0.01 g of pRL-TK, and 0.2 g of pCXN2 or pCXN2-core. As a positive control, pFC-MEKK was added to the transfection complexes containing pCXN2, or human TNF␣ (Strathmann Biotech GmbH, Hamburg, Germany) was added to the medium of transfected HeLa cells at a concentration of 20 ng/ml 6 h before harvest. The effect of HCV core protein on the NF-B pathway in HeLa cells was examined using HepG2 cells with the same protocol as that used for HeLa cells. To examine the dose-dependent effect of core protein on the NF-B pathway, HeLa cells were transfected with a total of 1.2 g of plasmids consisting of 0.38 g of pNF-B-Luc, 0.02 g of pRL-TK, 0 -0.8 g of pCXN2-core, and 0 -0.8 g of pCXN2 adjusted to total 1.2 g.
In addition to the HeLa cells, which express core protein transiently, HeTOC cells, which can be induced to express core protein under the control of doxycycline, were used to examine the effect of HCV core protein on the NF-B pathway. Approximately 4 ϫ 10 5 HeTOC cells were plated into the well of a six-well tissue culture plate containing 1 g/ml doxycycline 24 h before transfection. Cells were transfected with 0.39 g of pNF-B-Luc and 0.01 g of pRL-TK and cultured in medium with or without doxycycline.
To elucidate how HCV core protein affects the NF-B pathway, dominant negative forms of IKK␣/␤, TRAF2/6, and TAK1, and an IKK␤specific inhibitor were used. To initially examine the effect of the dominant negative form of components of the NF-B pathway, HeLa cells were transfected with a total of 1.2 g of plasmids consisting of 0.38 g of pNF-B-Luc, 0.02 g of pRL-TK, 0.4 g of pCXN2-core, and 0.4 g of one of the following pIKK␣ (K44A) , pIKK␤ (K44A) , pTRAF2-(87-501), or pTRAF6-(289 -522). Similarly, HeLa cells were transfected with a total of 1.6 g of plasmids consisting of 0.38 g of pNF-B-Luc, 0.02 g of pRL-TK, 0.4 g of pCXN2-core, 0.4 g of pCMV-TAK1(K63W), and 0.4 g of pCMV-TAB1. Second, we added 5 mM acetyl salicylic acid (Sigma), which inhibits cyclooxygenase (24) and IKK␤ (25), to the medium of HeLa cells transfected with 0.19 g of pNF-B-Luc, 0.01 g of pRL-TK, and 0.2 g of pCXN2 or pCXN2-core. Acetyl salicylic acid was added 1 h after the transfection of plasmids. Instead of acetyl salicylic acid, 25 M indomethacin (Sigma), a nonsteroidal anti-inflammatory drug, which inhibits cyclooxygenase but not IKK␤ (24,25), was added to the medium as a control.
Luciferase Assay-The entire cell lysate was collected 36 h after transfection. A luciferase assay was performed with the PikkaGene dual sea pansy system (Toyo Ink, Tokyo, Japan) using a luminometer (Lumat LB9507, EG&G Berthold, Bad Wildbad, Germany). Assays were conducted at least in triplicate. Firefly luciferase activity and sea pansy luciferase activity were measured as a relative light unit. Firefly luciferase activity was then normalized for transfection efficiency based on sea pansy luciferase activity.
Indirect Immunofluorescence Staining of HCV Core Proteins-Indi-rect immunofluorescence staining was performed as previously described (26); COS-7 cells were transfected with pCXN2-core, pCXN2-core173, or pCXN2-HAcore151. Forty-eight hours after transfection, COS-7 cells were washed with phosphate-buffered saline (PBS) and subsequently fixed with 3.7% formaldehyde in PBS for 1 h at room temperature. After washing with PBS, cells were then permeabilized with 0.1% Tween 20 in PBS for 1 h. The cells were then blocked with PBS containing 2% normal rabbit serum for 1 h and incubated with mouse anti-core antigen IgG fraction (1:500) for 1 h at room temperature. After washing three times with PBS, cells were incubated with fluorescein isothiocyanate-conjugated rabbit anti-mouse IgG antibody (1:40) (Dako, Carpinteria, CA) for 1 h at room temperature. Cells were then washed with ice-cold PBS, coated with fluorescent mounting medium (Dako), covered with glass, and observed by microscope using an epifluorescent attachment (AX80, Olympus, Tokyo, Japan). Electrophoretic Mobility Shift Assay-Approximately 2 ϫ 10 6 He-TOC-22 cells were plated into a 10-cm dish (Iwaki Glass) and cultured in medium with or without 1 g/ml doxycycline. Forty-eight hours later, the cells were harvested and their nuclear extracts were prepared according to mini-nuclear extraction methods (27). The concentration of the nuclear extracts was determined by a Micro BCA protein assay reagent kit (Pierce) and was adjusted to give equal concentrations. Electrophoretic mobility shift assay (EMSA) was performed using a Gel shift assay system (Promega) according to the manufacturer's protocol. Briefly, a synthetic double-stranded oligonucleotide having a B site (5Ј-AGTTGAGGGGACTTTCCCAGGC-3Ј) was end-labeled with [ 32 P]ATP using T4 polynucleotide kinase and incubated with 10 g of nuclear extracts for 20 min at room temperature. For competition, an unlabeled competitor oligonucleotide or an unlabeled noncompetitor oligonucleotide (5Ј-GATCGAACTGACCGCCCGCGGCCCGT-3Ј) was added to the reaction mixture in 100-fold osmolar excess over the labeled probe to examine the binding specificity. Reaction mixtures were loaded onto a 4% polyacrylamide gel and then separated by electrophoresis in 0.5ϫ Tris borate/EDTA electrophoresis buffer (0.045 M Tris borate and 0.001 M EDTA).
Statistics-Data were expressed as means Ϯ S.D. Statistical analyses were performed using the t test. A p value of less than 0.05 was considered statistically significant.

Detection of Transiently Expressed HCV Core Proteins by
Western Blotting-Expression of full-length and three deleted core proteins was examined in the soluble protein cell extracts of transiently transfected COS-7 cells by Western blotting (Fig.  1, A and B). Full-length (pCXN2-core) and aa 1-173 core (pCXN2-core173) were detected by mouse anti-HCV core antigen IgG fraction (Fig. 1A). HA-tagged core aa 1-151 (pCXN2-HAcore151) and aa 92-191 (pCXN2-HAcore92-191) were detected by anti-HA polyclonal antibody (Fig. 1B). The size of each core protein was consistent with the expected size.
Doxycycline-regulated HCV Core Protein Expression in He-TOC Cells-Expression of HCV core protein was detected by Western blotting in HeTOC cells, a clone termed HeTOC-22, which can be induced to express core protein under the control of doxycycline, 48 h after the removal of doxycycline. The induction ratio was greater than 1,000 when the relative amount of expressed HCV core protein with or without doxycycline was analyzed using a LAS-1000 image analyzer (photo film from Fuji, Tokyo, Japan) (data not shown).
Activation of the NF-B Pathway by the HCV Core Protein-In HeLa cells, HCV core protein (0.2 g of pCXN2-core) significantly activated the NF-B pathway at a value 6.2 Ϯ 3.4 (mean Ϯ S.D.) times higher than that of the control, whereas TNF-␣ (20 ng/ml) activated the pathway at a value 8.6 Ϯ 2.0 times higher ( Fig. 2A). Activation of the NF-B pathway increased in relation to the amount of plasmid utilized for pCNX2-core (Fig. 3). In HepG2 cells, HCV core protein activated the pathway at a value 4.3 Ϯ 0.9 times higher than the control, whereas TNF-␣ activated the pathway at a value 4.5 Ϯ 2.5 ( Fig. 2A).
HeTOC cells, which can be induced to express core protein under the control of doxycycline, were used to examine activation of the NF-B pathway. Core-expressing HeTOC-22 cells cultured in a medium without doxycycline showed a pathway activation value 5.2 Ϯ 1.5 times higher than that found with HeTOC-22 cells cultured in a medium with doxycycline. Addition of TNF␣ to core-expressing HeTOC-22 cells did not affect significantly the activation of NF-B pathway by core protein (Fig. 2B).
Core Protein Enhances NF-B-DNA Binding Activity-To examine whether core protein enhances NF-B-DNA binding, EMSA was performed using the nuclear extracts of HeTOC-22 cells, which were induced to express core protein. As shown in Fig. 4, NF-B-DNA binding activity was enhanced in He- TOC-22 cells expressing core protein (about 2.9 times) as compared with that in HeTOC-22 cells without expressing core protein under the existence of doxycycline. This NF-B-DNA binding activity observed in these assays was ablated by an excess of unlabeled competitor but not by an excess of unlabeled noncompetitor (Fig. 4). Addition of an antibody directed against p65 or p50 (Santa Cruz Biotechnology) generated a supershifted band, suggesting that this NF-B-DNA complex was containing p65 and p50 (data not shown).
Mapping the Region of HCV Core Protein Responsible for Activation of the NF-B Pathway-To determine the region of HCV core protein responsible for activation of the NF-B pathway, we constructed three plasmids, which expressed deletion mutant forms of core protein: pCXN2-core173, pCXN2-HA-core151, and pCXN2-HAcore92-191. HeLa cells were transfected with full-length or each deletion mutant form of HCV core expressing plasmids in combination with pNF-B-Luc (Fig. 5). None of the deletion mutant forms of core protein activated the NF-B pathway, whereas full-length HCV core protein did activate the pathway. These results suggest that both the N terminus (aa 1-91) and C terminus (aa 174 to 191) of the core protein may play an important role in activation of the NF-B pathway.
We then examined subcellular localization of full-length and two deletion mutants of HCV core protein by indirect immunofluorescence assay. As shown in Fig. 6, full-length core protein (aa 1-191) was located diffusely in the cytoplasm, in contrast to the perinuclear localization of the C-terminal 18 aa-deleted core protein (aa 1-173) and the nuclear localization of the C-terminal 40 aa-deleted core protein (aa 1-151).
HCV Core Protein Activates the NF-B Pathway through IKK␤-To examine whether activation of the NF-B pathway by HCV core protein is transduced through IKK␣ or IKK␤, HeLa cells were cotransfected with pCXN2/pCXN2-core, pNF-B-Luc, and pIKK␣ (K44A) /pIKK␤ (K44A) . Expression of IKK␤ (K44A) , catalytically inactive IKK␤, significantly reduced HCV core induced NF-B activation to about one-tenth, whereas expression of IKK␣ (K44A) , catalytically inactive IKK␣, reduced the activation to about two-fifths (Fig. 7). To confirm the participation of IKK␤ in activation of the NF-B pathway by HCV core protein, we added 5 mM acetyl salicylic acid, an IKK␤-specific inhibitor (28), to the HeLa cells transfected with pCXN2/pCXN2-core and pNF-B-Luc. Activation of the pathway by core protein was significantly inhibited by acetyl salicylic acid but not by indomethacin, a cyclooxygenase inhibitor (Fig. 8). These results suggest that HCV core protein activates the NF-B pathway through IKK, especially IKK␤.
Dominant Negative Forms of TRAF2/6 Reduced the Activation of the NF-B Pathway by HCV Core Protein-To examine whether activation of the NF-B pathway by HCV core protein was transduced through TRAF2/6, HeLa cells were cotransfected with pCXN2/pCXN2-core, pNF-B-Luc, and pTRAF2-(87-501)/pTRAF6-(289 -522). Expression of the dominant negative form of TRAF2 (aa 87-501) significantly reduced coreinduced NF-B activation to about two-fifths in HeLa cells (Fig.  9). Expression of the dominant negative form of TRAF6 (aa 289 -522) also reduced core-induced NF-B activation (Fig. 9). These results suggest that HCV core protein activates the HeLa cells were transfected with pCXN2, pCXN2-core, pCXN2-core173, pCXN2-HAcore151, pCXN2-HAcore92-191, and reporter plasmids. Luciferase activities were measured and expressed as described in the legend for Fig. 2. Any deletion of mutant forms of core protein did not activate the NF-B pathway, whereas full-length HCV core protein activated the pathway.
NF-B pathway through TRAF2/6. Catalytically Inactive TAK1 Has No Effect on Activation of the NF-B Pathway by HCV Core Protein-The kinase TAK1 was shown to act upstream of NIK in the IL-1-activated signaling pathway and associate with TRAF6 during IL-1 signaling (23). To examine whether activation of the NF-B by HCV core protein was transduced through TAK1, HeLa cells were cotransfected with pCXN2/pCXN2-core, pNF-B-Luc, pCMV-TAK1, and pCMV-TAB1. Although expression of catalytically inactive TAK-1 and its activator, TAB1, efficiently suppressed IL-1-induced activation of the NF-B pathway, they had no effect against core-induced activation of the pathway (Fig. 10), suggesting that HCV core protein activates the NF-B pathway independent of TAK1.
The Effect of HCV Core Protein on IL-1␤ or TNF␣ Promoter-HCV core protein activated the IL-1␤ promoter 2.5 Ϯ 0.6 times higher than the control. However, core protein did not have significant influence on the TNF␣ promoter. This activation of IL-1␤ promoter was cancelled by catalytically inactive IKK␤, or dominant negative TRAF2/6 ( Fig. 11), suggesting HCV core protein activates IL-1␤ promoter mainly through NF-B signaling.

DISCUSSION
In this study, HCV core protein clearly activated the NF-B pathway not only in a dose-dependent manner in mammalian cells transiently transfected with core protein expression plasmid, but also in HeLa cells inducibly expressing core protein. In fact, activation of NF-B signaling by transfection of 0.4 g of pCXN2-core corresponded to activation by 20 ng/ml TNF␣ ( Fig.  2; Ref. [7][8][9]. This concentration of TNF␣, which is widely used to study responses to TNF, is enough to induce cytolysis in murine fibrosarcoma L929 cells (29). However, there are contradicting data regarding the effect core protein has on the NF-B pathway; two studies demonstrated that core protein enhanced NF-B signaling in cells stably expressing the core protein when using an EMSA (8) or in cells transiently expressing the core protein when using a reporter assay (7,9), while other studies showed that HCV core protein suppressed TNFinduced NF-B activation in cells stably expressing HCV core protein with use of an EMSA (5, 6). These contradicting results may be due to differences in the type of cells stably or transiently expressing core protein and the method for detecting NF-B activation. To solve this problem, we adopted a tightly regulated, high level core protein expression system responsive to doxycycline (30). This system allowed us to overcome the problem of clonal selection of permanent transfectants that differ in characteristics from parental cell lines, as well as the problem of low transfection efficiency in the transient transfection assay. Thus, we could analyze the effect of HCV core protein on NF-B signaling in the same cells with or without doxycycline by use of both a reporter assay and an EMSA.
This study also demonstrated that HCV core protein activates the NF-B pathway through IKK␣/␤ (Fig. 7). Core protein may predominantly modulate the activity of one of the two FIG. 6. Subcellular localization of full-length and two deletion mutants of the core proteins by indirect immunofluorescence assay. COS-7 cells were transfected with three types of core protein expression plasmid, pCXN2-core, pCXN2-core173, and pCXN2-HA-core151. After 48 h, cells were fixed and incubated with mouse anti-core antigen IgG fraction and then stained with fluorescein isothiocyanateconjugated rabbit anti-mouse IgG antibody. Full-length core protein (aa  was located diffusely in the cytoplasm (A), whereas the Cterminal 18 aa-deleted core protein (aa 1-173) was located in the perinuclear region (B) and the C-terminal 40 aa-deleted core protein (aa 1-151) was found mainly in the nucleus (C). Acetyl salicylic acid, an IKK␤-specific inhibitor, was added to the medium of HeLa cells transfected with pCXN2/pCXN2core and pNF-kB-Luc at a concentration of 5 mM. Luciferase activities were measured and expressed as described in the legend for Fig. 2. Activation of the pathway by HCV core protein was significantly inhibited by acetyl salicylic acid but not indomethacin, a cyclooxygenase inhibitor. IB kinases, IKK␤. The dominant negative mutant of IKK␤ completely abrogated activation of the NF-B pathway by HCV core protein, in contrast to the minor role played by the comparable mutant IKK␣. Moreover, acetyl salicylic acid, an IKK␤specific inhibitor (25), significantly blocked the action of HCV core protein on NF-B. Recently, it was shown that only IKK␤ phosphorylation contributes to IKK activation by proinflammatory cytokines or by cotransfected NIK and MEKK1 (31). This result is consistent with the genetic analysis of IKK function; whereas disruption of the IKK␣ locus has no effect on IKK activation and IB degradation in response to proinflammatory stimuli, disruption of the IKK␤ locus results in a major defect in both events (32,33). In fact, the dominant negative mutant of IKK␤ blocked activation of the pathway by HCV core protein and TNF more efficiently than IKK␣ (Fig. 7). These results suggest that HCV core protein mimics proinflammatory cytokine activation of the NF-B pathway.
Furthermore, the dominant negative forms of TRAF2/6 significantly blocked activation of the NF-B pathway by HCV core protein. TRAF proteins are known to function as signal transducers for distinct receptor families. TRAF2 is thought to be a common mediator of TNFR and CD40 signaling (21), whereas TRAF6 is thought to be a signal transducer for IL-1 (22). Because CD40 is not expressed and CD40 signaling is not activated by signal-activating anti-CD40 antibody (34) (monoclonal antibody 89, Immunotech, Marseille, France) in HeLa and HepG2 cells (data not shown), TNFR or IL-1 signaling is the temporary candidate for the target of HCV core protein.
Although TRAF6 is thought be involved in IL-1 signaling and not in TNFR signaling, previous (35) and present studies have shown that the dominant negative form of TRAF6 suppresses the activation of TNF␣ activation of NF-B (Fig. 9). On the other hand, the dominant negative form of TRAF2 suppresses only TNFR signaling and not IL-1 signaling. Moreover, catalytically inactive TAK1, which links TRAF6 to the NIK-IKK cascade in the IL-1 signaling pathway, has no effect against the core-induced activation of NF-B. These data may imply that HCV core protein mimics proinflammatory cytokine activation of the NF-B pathway, especially TNFR signaling through TRAF2/6.
Recently, HCV core protein was shown to interact with the cytoplasmic tail of lymphotoxin-␤ receptor (36,37), a member of the TNFR family, and also with the cytoplasmic domain of TNFR1 (6), where TRADD and RIP interact to produce TNFinduced NF-B activation (11,12). There may be the possibility that HCV core protein activates the NF-B pathway through interaction with TNFR1. Our finding that the N-terminal 91-aa deleted core protein did not activate the NF-B pathway may support this idea because the N-terminal 115 aa of HCV core protein is important for the interaction with TNFR1 (6). However, we could not detect an in vivo interaction between core protein and TNFR1, TRADD, and TRAF2 by coimmunoprecipitation assay (data not shown). Deletion analysis of the core protein showed that the C-terminal 18 aa, a highly hydrophobic region (38), is also important for NF-B activation. It was suggested that this region is responsible for the association of HCV core protein with intracellular membranes, and the Cterminal deleted core protein translocates into the nucleus (39,40). As shown in Fig. 6, the C-terminal deleted core proteins are located in the perinuclear region or the nucleus, whereas full-length core protein was diffusely located in the cytoplasm. These data may support our finding that core protein activates NF-B in the cytoplasm through TRAF2/6.
The NF-B pathway is known to be activated by oncogenic viral proteins such as X protein of the hepatitis B virus, Tax of human T-cell leukemia virus type 1, and latent membrane protein 1 (LMP-1) of the Epstein-Barr virus (41). Hepatitis B virus X protein interacts directly with IB␣ to probably prevent the reassociation of IB␣ with NF-B (42). Tax was shown to interact with components of the IKK complex, such as MEKK1 (28), IKK (43), and the NF-B essential modulator (44), thereby activating the NF-B pathway. Tax has been shown to also associate with IB␣, p105, p100, RelA, and c-Rel (41). Thus, more than one molecule may be the target of HCV core protein to activate the NF-B pathway as well as Tax.
The NF-B pathway plays an important role in cellular response to proinflammatory cytokines such as TNF-␣ and IL-1 and induces an inflammatory response by the up-regulation of many cytokines, including IL-1, -2, -6, -8, and -12, and TNF-␣ (10). In this study, HCV core protein is shown to mimic proinflammatory cytokine activation of the NF-B pathway, especially TNFR signaling., and actually activates IL-1␤ promoter mainly through NF-B signaling pathway. Recently we have shown that HCV core protein activates also IL-8 promoter FIG. 10. Catalytically inactive TAK1 had no effect on activation of the NF-B pathway by HCV core protein. HeLa cells were transfected with pCXN2/pCXN2-core, pNF-B-Luc, pCMV-TAK1, and pCMV-TAB1. Luciferase activities were measured and expressed as described in the legend for Fig. 2. Catalytically inactive TAK-1 and its activator, TAB1, had no effect on core-induced activation of the pathway. IL-1 was added 6 h before harvest to function as an inducer of the pathway (positive control). through NF-B pathway (9).Therefore, it is quite conceivable that HCV core protein could induce an inflammatory response and cause "hepatitis." In fact, the eradication of HCV by interferon leads to the rapid resolution of acute and chronic hepatitis (52,53). Moreover, serum or intrahepatic expression of IL-1␤, 2, 6, 8 and TNF-␣ are elevated from 2 to 10 times higher in patients with active chronic hepatitis C than those of a control group, and reduced after eradication of the virus by interferon treatment (54 -58). Although the host immune response caused by cytotoxic T lymphocytes is believed to play a pivotal role in the pathogenesis of C-viral hepatitis (50), our findings suggest that HCV core protein directly induces hepatitis through inflammatory cytokine production. Therefore, blockage of the activation of the NF-B pathway may become an attractive option for the treatment of chronic hepatitis C in the future.