Raloxifene inhibits hepatitis C virus infection and replication.

Postmenopausal women with chronic hepatitis C exhibited a poor response to interferon (IFN) therapy compared to premenopausal women. Osteoporosis is the typical complication that occurs in postmenopausal women. Recently, it was reported that an osteoporotic reagent, vitamin D3, exhibited anti-hepatitis C virus (HCV) activity. Therefore, we investigated whether or not another osteoporotic reagent, raloxifene, would exhibit anti-HCV activity in cell culture systems. Here, we demonstrated that raloxifene inhibited HCV RNA replication in genotype 1b and infection in genotype 2a. Raloxifene enhanced the anti-HCV activity of IFN-α. These results suggest a link between the molecular biology of osteoporosis and the HCV life cycle.

Introduction

Hepatitis C virus (HCV) belongs to the Flaviviridae family and contains a positive single-stranded RNA genome of 9.6 kb. The HCV genome encodes a single polyprotein precursor of approximately 3000 amino acid residues, which is cleaved by the host and viral proteases into at least 10 proteins in the following order: Core, envelope 1 (E1), E2, p7, nonstructural 2 (NS2), NS3, NS4A, NS4B, NS5A, and NS5B [1-3].

The virological study and screening of antiviral reagents for HCV was difficult until the replicon system was developed [4-7]. In 2005, an infectious HCV production system was developed using genotype 2a HCV JFH-1 and hepatoma-derived HuH-7 cells, and the HCV life cycle was reproduced in a cell culture system [8]. We previously developed genome-length HCV reporter assay systems using HuH-7-derived OR6 cells [4]. In OR6 cells, the genotype 1b HCV-O with renilla luciferase (RL) replicates robustly. We also developed an HCV JFH-1 reporter infection assay system [9].

HCV infection frequently causes chronic hepatitis (CH) and leads to serious liver cirrhosis and hepatocellular carcinoma. Therefore, HCV infection is a major health problem worldwide. The elimination of HCV by antiviral reagents seems to be the most efficient therapy for preventing the fatal state of the disease. Pegylated-interferon (PEG-IFN) with ribavirin (RBV) is the current standard therapy for CH–C, but its sustained virological response (SVR) rate has remained 40–50%. Recently, a protease inhibitor, telaprevir, improved the SVR rate by up to 60–70% in combination with PEG-IFN/RBV [10]. The response to PEG-IFN/RBV therapy depends on host factors as well as viral factors. Among the host factors, age and gender are known to be associated with the outcome of IFN/RBV therapy [11,12]. Postmenopausal women with CH–C exhibited a poor response to IFN therapy compared to premenopausal women [11]. The decrease in estrogen may affect the response to IFN therapy. Dyslipidemia and osteoporosis are the typical complications in postmenopausal women. We and other groups reported that statins, which are dyslipidemia reagents, inhibited HCV proliferation in vitro and in vivo [13-17]. Recently it was reported that vitamin D3, an osteoporotic reagent, exhibited anti-HCV activity in vitro and in vivo [18-21]. It was also reported that 17β-estradiol inhibited the production of infectious HCV [22]. Taken together, these reports suggest an association between hepatitis C and complications due to the decrease of estrogen.

Raloxifene and tamoxifen are synthetic selective estrogen receptor modulators (SERMs) and are used for breast cancer and osteoporosis, respectively, in clinical settings. The responses of SERMs are mediated by estrogen receptors (ERs), either ERα or ERβ. SERMs exhibit agonistic actions in some tissues but antagonistic actions in others. Both raloxifene and tamoxifen are antagonists in breast and agonists in bone. However, only tamoxifen, and not raloxifene, exhibited agonistic activity in the uterus. It was reported that tamoxifen inhibited HCV RNA replication [23]. However, tamoxifen's agonist action leads to uterine cancer. Raloxifene belongs to an antiosteoporotic reagent and offers the advantage of safety without uterine cancer. Therefore, we decided to investigate whether or not raloxifene would exhibit anti-HCV activity in our developed cell culture systems.

Materials and methods

Reagents and antibodies

Raloxifene was purchased from LKT Laboratories, Inc. (St. Paul, MN). IFN-α and tamoxifen were purchased from Sigma–Aldrich (St. Louis, MO). Pitavastatin (PTV) was purchased from Kowa Company (Nagoya, Japan). The antibodies used in this study were those specific to HCV Core (CP11, Institute of Immunology, Tokyo, Japan), NS3 (Novocastra Laboratories, Newcastle, UK), and β-actin (Sigma).

Cell culture and HCV RNAs

HuH-7 cells were cultured in Dulbecco's modified Eagle's medium (Gibco-BRL, Invitrogen Life Technology, Carlsbad, CA) supplemented with 10% fetal bovine serum, penicillin, and streptomycin. HuH-7-derived OR6 and sOR cells were genome-length and subgenome HCV (O strain of genotype 1b) RNA harboring cells, respectively and cultured in the above medium supplemented with G418 (0.3 mg/ml; Geneticin, Invitrogen) [4]. HCVs replicating in OR6 and sOR cells contain RL and neomycin phosphotransferase (NPT) genes after 5′-untranslated region (UTR). HuH-7-derived RSc cells are cured cells, in which HCV RNA was eliminated by IFN-α; they are used for HCV JFH-1 infection [9]. RSc cells are also used for subgenomic JFH-1 RNA (JRN/35B) replication. JRN/35B contains RL and NPT genes after 5′-UTR.

RL assay

For the RL assay, 1.5×104 OR6 were plated onto 24-well plates in triplicate and cultured for 24 h. The cells were treated with each reagent for 72 h. Then the cells were harvested with Renilla lysis reagent (Promega, Madison, WI) and subjected to RL assay according to the manufacturer's protocol.

WST-1 cell proliferation assay

The cells (2×103 cells) were plated onto a 96-well plate in triplicate at 24 h before treatment with each reagent. At 72 h after treatment, the cells were subjected to a WST-1 cell proliferation assay (Takara Bio, Otsu, Japan) according to the manufacturer's protocol.

Western blot analysis

For Western blot analysis, 4×104 cells were plated onto 6-well plates, cultured for 24 h, and then treated with reagent(s) for 72 h and 120 h. Preparation of the cell lysates, sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and immunoblotting were then performed as previously described [24]. Immunocomplexes on the membranes were detected by enhanced chemiluminescence assay (Renaissance; Perkin Elmer Life Science, Wellesley, MA).

HCV infection

RSc cells (1.5×104 cells) were plated onto a 24-well plate 24 h before infection. To evaluate the effect of the treatment prior to infection, the cells were first treated with raloxifene for 24 h, then inoculated with reporter JFH-1 (JR/C5B/BX-2) supernatant at a multiplicity of infection (MOI) of 0.2, cultured for 48 h, and subjected to RL assay as described previously [9]. The JR/C5B/BX-2 contains the RL gene in the first cistron following the encephalomyocarditis virus-internal ribosomal entry site (EMCV-IRES) gene and the open reading frame (ORF) of JFH-1 in the second cistron. To evaluate the effect of the treatment after infection, the cells were inoculated with reporter JFH-1 supernatant at MOI of 0.2, cultured for 72 h, and subjected to RL assay.

Results

Raloxifene inhibited HCV RNA replication

The HCV RNA that replicated in HuH-7-derived OR6 cells was a genome-length HCV with RL, NPT, and EMCV-IRES in the first cistron and the ORF of HCV (O strain of genotype 1b) in the second cistron [4]. OR6 cells could not produce infectious HCV. Therefore, we can monitor the replication step in the HCV life cycle using OR6 cells. Raloxifene inhibited HCV RNA replication in a dose-dependent manner, and its 50% effective concentration (EC50) was 1 μM (Fig. 1A). Raloxifene did not exhibit cytotoxicity to OR6 cells until 2.5 μM (Fig. 1B). Raloxifene also inhibited intracellular Core and NS3 production in a dose- and time-dependent manner (Fig. 1C). The intensities of Core and NS3 in OR6 cells treated with 2.5 μM of raloxifene decreased to almost the level of cells treated with 10 IU/ml of IFN-α at 120 h after treatment. We also examined anti-HCV activity of raloxifene using subgenomic HCV replicon harboring sOR cells. Raloxifene exhibited weak anti HCV activity to sOR cells as compared with OR6 cells (Supplementary Figs. 1A and 1B). These results suggest that raloxifene exhibits anti-HCV activity and decreased the expression levels of HCV proteins more slowly compared to IFN-α.

Fig. 1

Raloxifene inhibited HCV RNA replication. (A) Anti-HCV activity of raloxifene in OR6 cells. OR6 cells were treated with raloxifene (0, 0.625, 1.25, and 2.5 μM) for 72 h. Relative RL activity (relative light unit: RLU) for HCV RNA replication is expressed as a percentage of control. Each bar represents the average with standard deviations of triplicate data points. (B) Effect of raloxifene on OR6 cell viability. Cell viability at 72 h after raloxifene treatment (0.15, 0.31, 0.625, 1.25, 2.5, 5, and 10 μM) was determined using WST-1 cell proliferation assay and is expressed as a percentage of control. (C) Raloxifene inhibited HCV proteins. OR6 cells were treated with IFN-α (10 IU/ml) or raloxifene (0, 0.625, 1.25, and 2.5 μM). After 72 or 120 h treatment, the production of Core and that of NS3 were analyzed by immunoblotting using anti-Core and anti-NS3 antibodies, respectively. OR6c cells were cured cells in which HCV RNA was eliminated using IFN-α, and were used as a negative control. β-actin was used as a control for the amount of protein loaded per lane.

Raloxifene enhanced anti-HCV activity of IFN-α

We investigated the anti-HCV activity of raloxifene in combination with a representative anti-HCV reagent, IFN-α. HCV RNA replication decreased in a dose-dependent manner after co-treatment with IFN-α and raloxifene (Fig. 2A). The results were almost similar to the expected effect of raloxifene in combination with IFN-α calculated from the anti-HCV activity of each reagent (Fig. 2B). These results indicate that the anti-HCV activity of raloxifene and IFN-α exhibited additive effect. We also examined the anti-HCV activity of previously reported SERM, tamoxifen. Tamoxifen also exhibited additive anti-HCV activity on HCV RNA replication in combination with IFN-α (Supplementary Figs. 2A–C). These results indicate that raloxifene as well as tamoxifen enhanced the anti-HCV activity of IFN-α. As both raloxifene and IFN-α are clinically used reagents, raloxifene seemed to be a candidate reagent as an add-on treatment to IFN-α in patients with CH–C.

Fig. 2

Raloxifene enhanced the anti-HCV activity of IFN-α. (A) Anti-HCV activity of raloxifene in combination with IFN-α. OR6 cells were co-treated with raloxifene (0, 0.625, 1.25, and 2.5 μM) and IFN-α (0, 0.63, 1.25, 2.5, 5, 10 IU/ml). Relative RL activity is shown as a percentage of control. Each bar represents the average with standard deviations of triplicate data points. (B) Expected anti-HCV activity was calculated based on the results when the cells were treated with only raloxifene or IFN-α.

Raloxifene antagonized anti-HCV activity of statin

We previously reported that statins exhibited anti-HCV activity using the OR6 assay system [14]. Statin is the first-choice reagent for dyslipidemia. As dyslipidemia and osteoporosis are major complications in postmenopausal women, we invested the effect of raloxifene on the anti-HCV activity of PTV. Raloxifene did not enhance the anti-HCV activity of PTV (Fig. 3A). Fig. 3B exhibits the expected anti-HCV activity of co-treatment with raloxifene and PTV calculated from the anti-HCV effect of either raloxifen or PTV alone. Raloxifene exhibited an antagonistic effect on PTV's anti-HCV activity. Raloxifene's antagonistic effect on PTV increased dose-dependently. The co-treatment with raloxifene (2.5 μM) and PTV (0.25, 0.5, and 1 μM) resulted in lower anti-HCV activity than did treatment with raloxifene alone (2.5 μM). These results suggest that we should be careful in the administration of statins with raloxifene to postmenopausal woman with CH–C.

Fig. 3

Statin antagonized the anti-HCV activity of raloxifene. (A) OR6 cells were co-treated with raloxifene (0, 0.625, 1.25, and 2.5 μM) and PTV (0, 0.25, 0.5, and 1 μM). Relative RL activity was shown as a percentage of control. Each bar represents the average with standard deviations of triplicate data points. (B) Expected anti-HCV activity was calculated based on the results when the cells were treated with only raloxifene or PTV.

Raloxifene inhibited infection of genotype 2a HCV

To further investigate the anti-HCV activity of raloxifene, we examined whether or not raloxifene could inhibit HCV infection. For this purpose, we used our recently developed JFH-1 reporter infection assay system [9]. HuH-7-derived RSc's are highly HCV-permissive cell lines. Raloxifene was pretreated at 24 h before HCV infection. The cells were inoculated with HCV JFH-1 virion with RL (JR/C5B/BX-2), and the infection was monitored with RL activity at 48 h after infection. As shown in Fig. 4A, raloxifene inhibited HCV infection in RSc cells in a dose-dependent manner. Next we examined the effect of raloxifene after HCV infection. RSc cells were inoculated with HCV JFH-1 virion with RL. After HCV infection, the cells were treated with raloxifene for 72 h and raloxifene's inhibitory effect on post-infection was assessed using the RL assay. Raloxifene inhibited HCV proliferation in a dose-dependent manner when it was added to the cells after infection in RSc cells, although inhibitory effect of raloxifene on JFH-1 HCV RNA replication seemed to be weak compared to the genotype 1b HCV-O RNA replication (Fig. 4B). Raloxifene did not exhibit cytotoxicity to RSc cells until 2.5 μM (Fig. 4C). We found that raloxifene could not inhibit subgenomic JFH-1 HCV (JRN/35B) RNA replication (Fig. 4D). We further examined the inhibitory action of raloxifene around infection step. RSc cells were treated for short time with raloxifene around infection step: for 1, 4, and 4 h before, during, and after inoculation, respectively (Fig. 4E). Raloxifene inhibited JFH-1 infection, when it was treated during inoculation but not just before or after inoculation. In case of genotype 2a JFH-1, raloxifene's anti-HCV activity is mainly due to the inhibition of infection. These results indicate that raloxifene inhibits JFH-1 infection but not its RNA replication.

Fig. 4

Raloxifene inhibited genotype 2a HCV infection. (A) Raloxifene inhibited HCV JFH-1 infection. RSc cells were treated with raloxifene (0, 0.31, 0.625, 1.25, 2.5, 5, and 10 μM) 24 h before infection. HCV JFH-1 reporter virion was used as an inoculum after removal of raloxifene. The cells were then infected with reporter JFH-1 virion and cultured for 48 h. The inhibition of HCV infection was assessed by relative RL activity and expressed as a percentage of control. (B) Raloxifene inhibited HCV JFH-1 proliferation after infection. RSc cells were inoculated with HCV JFH-1 reporter virion and cultured for 24 h. Then the cells were treated with raloxifene (0, 0.31, 0.625, 1.25, 2.5, 5, and 10 μM) for 48 h. The inhibitory effect on HCV proliferation after infection was assessed by relative RL activity and expressed as a percentage of control. Each bar represents the average with standard deviations of triplicate data points. (C) Effect of raloxifene on RSc cells viability. Cell viability at 72 h after raloxifene treatment (0.15, 0.31, 0.625, 1.25, 2.5, 5, and 10 μM) was determined using WST-1 cell proliferation assay and is expressed as a percentage of control. (D) Subgenomic JFH-1 RNA (JRN/35B) replicating RSc cells were treated with raloxifene (0, 0.625, 1.25, and 2.5 μM) for 72 h. Relative RL activity for HCV RNA replication is expressed as a percentage of control. Each bar represents the average with standard deviations of triplicate data points. (E) Raloxifene (0, 0.625, 1.25, and 2.5 μM) was treated for 1, 4, and 4 h before, during, and after JFH-1 inoculation to RSc cells at MOI of 0.2, respectively. The cells were then cultured for 72 h. The inhibition of HCV infection was assessed by relative RL activity and expressed as a percentage of control.

Discussion

In this study, we demonstrated that raloxifene, an osteoporotic reagent, inhibited the replication of genotypes 1b HCV RNA replication and inhibited genotype 2a HCV JFH-1 infection. Raloxifene additively enhanced the anti-HCV activity of IFN-α. On the other hand, raloxifene exhibited an antagonistic effect on statins.

PEG-IFN/RBV therapy led to a 40–50% SVR rate among patients with CH–C. Telaprevir with PEG-IFN/RBV increases the effect of PEG-IFN/RBV therapy by 10–20%. However, the major complication of anemia in PEG-IFN/RBV therapy increased when telaprevir was added. Considering that PEG-IFN/RBV-based therapy is less effective on postmenopausal women, an alternative therapy with minimal side effects is needed. Add-on therapy for postmenopausal women may be a candidate for improving the SVR in these patients. We focused on the reagents, which compensate for the lack of estrogen function. Dyslipidemia and osteoporosis are the major complications in postmenopausal women, and these complications are attributable to the decrease in estrogen. Statins are clinically used reagents for dyslipidemia; they inhibit HCV RNA replication in vivo as well as in vitro [13-17]. Therefore, we investigated whether or not raloxifene exhibits anti-HCV activity using genotype 1b HCV RNA replication and genotype 2a infection systems. In the HCV life cycle, raloxifene inhibited genotype 2a HCV infection and genotypes 1b HCV RNA replication. Raloxifene may be a potential reagent with different anti-HCV mechanisms in the HCV life cycle. Further study is needed to clarify these underlying mechanisms.

Recently it was reported that vitamin D3, an osteoporotic reagent, inhibited HCV production in cell culture systems [20,21]. Furthermore, it was reported that vitamin D3 was associated with the effect of therapy for patients with CH–C [18,19]. Statins inhibited HCV RNA replication by suppressing geranylgeranyl pyrophosphate (GGPP) production [14]. Another osteoporotic reagent, bisphosphonate, may possess anti-HCV activity, because it also inhibited the biosynthesis of GGPP in the mevalonate pathway by inhibiting farnesyl pyrophosphate synthetase. Taken together, these findings indicate it is likely that the HCV life cycle is associated with osteoporosis.

Raloxifene and tamoxifen are SERMs for osteoporosis and breast cancer, respectively. Tamoxifen is used for estrogen receptor-positive breast cancer, and it inhibits HCV RNA replication in cell culture [23]. Tamoxifen's anti-HCV activity is associated with ERα. In our study, raloxifene inhibited HCV infection as well as replication. To clarify the multi-potential effects of raloxifene, further study is needed. The incidence of side effects including uterine cancer is lower in raloxifene therapy than in tamoxifen therapy [25]. This is another advantage of raloxifene in clinical use for patients with CH–C.

As for the precise role of ERα or ERβ on the HCV life cycle, we could not reach a clear conclusion because microarray analysis revealed an absence of expression for both ERα and ERβ in OR6 cells (data not shown). Hayashida et al. [22] reported that the most potent physiological estrogen, 17-β-estradiol, inhibited infectious HCV production using HuH-7.5 cells, and that ERα-selective agonist inhibited infectious HCV production whereas ERβ-selective agonist did not. Watashi et al. [23] reported that RNA interference-mediated knockdown of ERα reduced HCV RNA replication. In our study, the anti-HCV activity of raloxifene in infection and replication did not seem attributable to ERα or ERβ. It is not clear why our HuH-7-derived OR6 cells did not express ERα or ERβ. HuH-7 cells were developed in 1982 at Okayama University and distributed worldwide [26]. Recently, Bensadoun et al. [27] reported that the genetic background of the IL28B genotype of HuH-7 cells differed among different laboratories. This may be a consequence of the polyploidal nature of hepatoma cells. A similar mechanism might cause the different expression levels of ERα and ERβ. Another ER, GPR30 [28], was expressed in OR6 cells (data not shown; from microarray analysis). GPR30 may be the responsible host factor for anti-HCV activity in OR6 cells. Further study is needed to clarify this issue.

In conclusion, we found that raloxifene inhibited HCV RNA replication in genotype 1b and infection in genotype 2a. Raloxifene additively enhanced the anti-HCV activity of IFN-α. The antagonistic effects of statins and raloxifene will yield information on the clinical use of these reagents. Our results, as well as the reports of vitamin D3's anti-HCV activity, will open new fields of treatment for both osteoporosis and HCV infection.