Arsenic Trioxide Inhibits Hepatitis C Virus RNA Replication through Modulation of the Glutathione Redox System and Oxidative Stress

Abstract

Arsenic trioxide (ATO), a therapeutic reagent used for the treatment of acute promyelocytic leukemia, has recently been reported to increase human immunodeficiency virus type 1 infectivity. However, in this study, we have demonstrated that replication of genome-length hepatitis C virus (HCV) RNA (O strain of genotype 1b) was notably inhibited by ATO at submicromolar concentrations without cell toxicity. RNA replication of HCV-JFH1 (genotype 2a) and the release of core protein into the culture supernatants were also inhibited by ATO after the HCV infection. To clarify the mechanism of the anti-HCV activity of ATO, we examined whether or not PML is associated with this anti-HCV activity, since PML is known to be a target of ATO. Interestingly, we observed the cytoplasmic translocation of PML after treatment with ATO. However, ATO still inhibited the HCV RNA replication even in the PML knockdown cells, suggesting that PML is dispensable for the anti-HCV activity of ATO. In contrast, we found that N-acetyl-cysteine, an antioxidant and glutathione precursor, completely and partially eliminated the anti-HCV activity of ATO after 24 h and 72 h of treatment, respectively. In this context, it is worth noting that we found an elevation of intracellular superoxide anion radical, but not hydrogen peroxide, and the depletion of intracellular glutathione in the ATO-treated cells. Taken together, these findings suggest that ATO inhibits the HCV RNA replication through modulation of the glutathione redox system and oxidative stress.

Hepatitis C virus (HCV) is the causative agent of chronic hepatitis, which progresses to liver cirrhosis and hepatocellular carcinoma. HCV is an enveloped virus with a positive single-stranded 9.6-kb RNA genome, which encodes a large polyprotein precursor of approximately 3,000 amino acid residues. This polyprotein is cleaved by a combination of 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 (30).

Alpha interferon has been used as an effective anti-HCV reagent in clinical therapy for patients with chronic hepatitis C. The current combination treatment with pegylated alpha interferon and ribavirin, a nucleoside analogue, has been shown to improve the sustained virological response rate to more than 50% (15). However, the adverse effects of the combination therapy and the limited efficacy against genotype 1b warrant the development of new anti-HCV reagents.

Arsenic trioxide (ATO) (As₂O₃, arsenite) has been used as a therapeutic reagent in acute promyelocytic leukemia, which bears an oncogenic PML-retinoic acid receptor alpha fusion protein resulting from chromosomal translocation (51, 52, 68, 70). The ATO treatment induces complete remission through degradation of the aberrant PML-retinoic acid receptor α (70). The PML tumor suppressor protein is required for formation of the PML nuclear body (PML-NB), also known as nuclear dot 10 or the PML oncogenic domain, which is often disrupted by infection with DNA viruses, such as herpes simplex virus type 1, human cytomegalovirus, and Epstein-Barr virus (17). The treatment with ATO results in degradation of the PML protein and disruption of the PML-NB (70). Therefore, ATO has been become a useful probe for investigating the functions of the PML-NB, including cell growth, apoptosis, stress response, and viral infection. Indeed, ATO has been shown to increase retroviral infectivity, such as human immunodeficiency virus type 1 (HIV-1) and murine leukemia virus infectivity, but the mechanisms of this change are not well understood (5, 6, 32, 44, 47, 50, 57). In contrast, ATO was recently reported to inhibit the replication of HCV subgenomic replicon RNA (24). However, it also remains unclear how ATO inhibits the HCV RNA replication. In this study, using genome-length HCV RNA replication systems, we investigated the molecular mechanism(s) of the anti-HCV activity of ATO, and we provide evidence that ATO inhibits HCV RNA replication through modulation of the glutathione redox system and oxidative stress.

MATERIALS AND METHODS

Reagents.

ATO, N-acetyl-cysteine (NAC), ascorbic acid (vitamin C), and l-buthionine sulfoximine (BSO) were purchased from Sigma (St. Louis, MO). Arsenic pentoxide (APO) (As₂O₅, arsenate) was purchased from Wako (Osaka, Japan). Both ATO and APO were dissolved in 1 N NaOH at 0.1 M as a stock solution. An inducible nitric oxide synthase (iNOS) inhibitor, 1400W, was purchased from Calbiochem (Merck Biosciences, Darmstadt, Germany).

Cell culture.

293FT cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum. The following four HuH-7-derived cell lines or their parental HuH-7 cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum as described previously (25): O cells, harboring a replicative genome-length HCV-O RNA (O strain of genotype 1b) (25); OR6 cells, harboring the genome-length HCV-O RNA with luciferase as a reporter (25); sO cells, harboring the subgenomic replicon RNA of HCV-O (31); and RSc cured cells, which cell culture-generated HCV-JFH1 (JFH1 strain of genotype 2a) (58) could infect and effectively replicate in (2, 3). The O, OR6, and sO cells were maintained in the presence of G418 (300 μg/ml Geneticin; Invitrogen).

RNA interference.

Oligonucleotides with the following sense and antisense sequences were used for the cloning of short hairpin RNA (shRNA)-encoding sequences targeted to PML (56) in a lentiviral vector: 5′-GATCCCCAGATGCAGCTGTATCCAAGTTCAAGAGACTTGGATACAGCTGCATCTTTTTTGGAAA-3′ (sense) and 5′-AGCTTTTCCAAAAAAGATGCAGCTGTATCCAAGTCTCTTGAACTTGGATACAGCTGCATCTGGG-3′ (antisense). These oligonucleotides were annealed and subcloned into the BglII-HindIII site, downstream from an RNA polymerase III promoter of pSUPER (8), to generate pSUPER-PMLi. To construct pLV-PMLi, the BamHI-SalI fragments of pSUPER-PMLi were subcloned into the BamHI-SalI site of pRDI292, an HIV-1-derived self-inactivating lentiviral vector containing a puromycin resistance marker allowing for the selection of transduced cells (7). pLV-Chk2i was described previously (3).

Lentiviral vector production.

The vesicular stomatitis virus (VSV) G-pseudotyped HIV-1-based vector system has been described previously (42). The lentiviral vector particles were produced by transient transfection of the second-generation packaging construct pCMV-ΔR8.91 (1, 71) and the VSV G envelope-expressing plasmid pMDG2 as well as pRDI292 into 293FT cells with FuGene6 (Roche Diagnostics, Mannheim, Germany).

HCV infection experiments.

The supernatants was collected from cell culture-generated HCV-JFH1 (58)-infected RSc cells (2, 3) at 5 days postinfection and stored at −80°C after filtering through a 0.45-μm filter (Kurabo, Osaka, Japan) until use. For infection experiments with HCV-JFH1 virus, RSc cells (1 × 10⁵ cells/well) were plated onto six-well plates and cultured for 24 h. We then infected the cells with 50 μl (equivalent to a multiplicity of infection of 0.05 to 0.1) of inoculum. The culture supernatants were collected at 97 h postinfection, and the levels of the core protein were determined by enzyme-linked immunosorbent assay (Mitsubishi Kagaku Bio-Clinical Laboratories, Tokyo, Japan). Total RNA was isolated from the infected cellular lysates using an RNeasy minikit (Qiagen, Hilden, Germany) for quantitative reverse transcription-PCR (RT-PCR) analysis of intracellular HCV RNA. The level of intracellular HCV RNA in the RSc cells was >10⁸ copies/μg total RNA at 4 days postinfection.

Quantitative RT-PCR Analysis.

The quantitative RT-PCR analysis for HCV RNA was performed by real-time LightCycler PCR (Roche) as described previously (25). We used the following forward and reverse primer sets for the real-time LightCycler PCR: PML, 5′-GAGGAGTTCCAGTTTCTGCG-3′ (forward), 5′-GCGCCTGGCAGATGGGGCAC-3′ (reverse); β-actin, 5′-TGACGGGGTCACCCACACTG-3′ (forward), 5′-AAGCTGTAGCCGCGCTCGGT-3′ (reverse); HCV-O, 5′-AGAGCCATAGTGGTCTGCGG-3′ (forward), 5′-CTTTCGCGACCCAACACTAC-3′ (reverse); and HCV-JFH1, 5′-5′-AGAGCCATAGTGGTCTGCGG-3′ (forward), 5′-CTTTCGCAACCCAACGCTAC-3′ (reverse).

Western blot analysis.

Cells were lysed in buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 4 mM EDTA, 1% Nonidet P-40, 0.1% sodium dodecyl sulfate, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride. Supernatants from these lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by immunoblot analysis using anti-PML (A301-168A-1; Bethyl Laboratories, Montgomery, TX), anti-Chk2 (DCS-273; Medical & Biological Laboratories, MBL, Nagoya, Japan), anti-HCV core (CP-9 and CP-11; Institute of Immunology, Tokyo, Japan), anti-HCV NS5A (no. 8926; a generous gift from A Takamizawa, The Research Foundation for Microbial Diseases of Osaka University, Japan), anti-signal transducer and activator of transcription 3 (anti-STAT3) (BD Bioscience, San Jose, CA), anti-phospho-STAT3 (Tyr705) (Cell Signaling Technology, Danvers, MA) anti-poly(ADP-ribose) polymerase 1 (anti-PARP-1) (C-2-10; Calbiochem), or anti-β-actin antibody (Sigma).

MTT assay.

HuH-7 or O cells (5 × 10³ cells/well) were plated onto 96-well plates and cultured for 24 h. The cells were treated with ATO, APO, or NaOH for 24, 48, or 72 h and then subjected to the colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay according to the manufacturer's instructions (cell proliferation kit I; Roche). The absorbance was read using a microplate reader (model 2550; Bio-Rad Laboratories, Hercules, CA) at 550 nm with a reference wavelength of 690 nm.

RL assay.

OR6 cells (1.5 × 10⁴ cells/well) were plated onto 24-well plates and cultured for 24 h. The cells were treated with each reagent for 72 h and then subjected to the Renilla luciferase (RL) assay according to the manufacturer's instructions (Promega, Madison, WI). A Lumat LB9507 luminometer (Berthold, Bad Wildbad, Germany) was used to detect RL activity.

FL assay.

Plasmids were transfected into O cells (2 × 10⁴ cells/well in 24-well plates) using FuGene6 and cultured for 24 h. The cells were treated with or without 1 μM ATO for 24 h, and then firefly luciferase (FL) assays were performed according to the manufacturer's instructions (Promega).

Immunofluorescence and confocal microscopic analysis.

Cells were fixed in 3.6% formaldehyde in phosphate-buffered saline (PBS), permeabilized in 0.1% NP-40 in PBS at room temperature, and incubated with anti-PML antibody (PM001; MBL) at a 1:300 dilution in PBS containing 3% bovine serum albumin at 37°C for 30 min. They were then stained with fluorescein isothiocyanate-conjugated anti-rabbit antibody (Jackson ImmunoResearch, West Grove, PA) at a 1:300 dilution in PBS containing bovine serum albumin at 37°C for 30 min, followed by staining with 4′,6-diamidino-2-phenylindole (DAPI) at room temperature for 15 min. Following extensive washing in PBS, the cells were mounted on slides using a mounting medium of 90% glycerin-10% PBS with 0.01% p-phenylenediamine added to reduce fading. Samples were viewed under a confocal laser-scanning microscope (LSM510; Zeiss, Jena, Germany).

Measurement of intracellular O₂⁻ and H₂O₂ production.

The intracellular superoxide anion radical (O₂⁻) levels were measured with an oxidation-sensitive fluorescent probe, dihydroethidium (DHE) (Invitrogen Molecular Probes), that is highly selective for detection of O₂⁻ among reactive oxygen species (ROS). DHE is cell permeative and reacts with O₂⁻ to form ethidium, which in turn intercalates in DNA, thereby exhibiting a red fluorescence. The intracellular hydrogen peroxide (H₂O₂) levels were measured with another oxidation-sensitive fluorescent probe dye, 6-carboxy-2′,7′-dichlorodihydrofluorescein diacetate (carboxy-H₂DCFDA) (Invitrogen Molecular Probes). Carboxy-H₂DCFDA was intracellularly deacetylated with esterase and further oxidized with peroxidase to the fluorescent 2′,7′-dichlorodihydrofluorescein (DCF). The ATO- or BSO-treated O cells were washed with PBS and incubated with 5 μM DHE and 20 μM carboxy-H₂DCFDA in PBS at 37°C for 30 min. Cells were then washed twice with PBS. The DHE or DCF fluorescence intensity was measured using a FACSCalibur flow cytometer. For each sample, 10,000 events were collected. The O₂⁻ or H₂O₂ levels are indicated as mean fluorescence intensities, which were determined with the CellQuest software (BD Bioscience).

Detection of intracellular glutathione.

Intracellular glutathione levels were analyzed using CellTracker Green (5-chloromethylfluorescein diacetate [CMFDA]; Molecular Probes, Invitrogen). CMFDA is a membrane-permeative dye used to determine intracellular glutathione levels. Cytoplasmic esterase converts the nonfluorescent CMFDA to the fluorescent 5-chloromethylfluorescein (CMF), which can then react with glutathione. The excitation peak is at 492 nm, and the fluorescence emission peak is at 517 nm. O cells treated with 1 μM ATO for 72 h were washed with PBS and incubated with 5 μM CMFDA at 37°C for 30 min. The CMF fluorescence intensity was measured using a FACSCalibur flow cytometer. For each sample, 10,000 events were collected. The glutathione levels are given as the relative mean fluorescence intensities, which were determined with CellQuest software.

RESULTS

ATO inhibits HCV RNA replication.

First, we quantitatively examined the effect of ATO on the HCV RNA replication in HuH-7-derived O cells harboring a replicative genome-length HCV-O RNA (25). We found that submicromolar concentrations of ATO markedly inhibited genome-length HCV-O RNA replication in the O cells at 72 h after administration (Fig. 1A). The 50% effective concentration (EC₅₀) of ATO required for inhibition of genome-length HCV-O RNA replication was 0.19 μM (Fig. 1A). Consistent with this finding, the expression levels of the HCV core and NS5A proteins were also significantly decreased in the cell lysates of O cells treated with ATO for 72 h (Fig. 1B). In addition, ATO markedly inhibited the replication of the subgenomic replicon RNA (31), with an EC₅₀ of 0.48 μM at 72 h after the treatment (Fig. 1C). We next examined the effect of ATO on HCV reproduction by HCV-JFH1 infection (58). The results revealed that ATO significantly inhibited the intracellular RNA replication of HCV-JFH1, with an EC₅₀ of 0.27 μM, as well as the release of core protein into the culture supernatants in HuH-7-derived RSc cells at 97 h after inoculation of the HCV-JFH1 virus (Fig. 1D and E). Thus, we have demonstrated for the first time that ATO can inhibit the reproduction of HCV and particularly HCV RNA replication.

FIG. 1.

Inhibition of HCV RNA replication by ATO. (A) The level of genome-length HCV RNA in O cells after the treatment with ATO was monitored by real-time LightCycler PCR. Experiments were done in triplicate and, bars represent the mean percentage of HCV RNA. Error bars indicate standard deviations. (B) HCV core and NS5A protein expression levels in O cells after treatment with ATO. The results of Western blot analysis of cellular lysates with anti-HCV core, anti-HCV NS5A, or anti-β-actin antibody in O cells at 72 h after treatment with ATO at the indicated concentration are shown. (C) The level of subgenomic replicon RNA was monitored by real-time LightCycler PCR. Results from three independent experiments conducted as described for panel A are shown. (D) The level of intracellular genome-length HCV-JFH1 RNA was monitored by real-time LightCycler PCR. RSc cells were pretreated with the indicated concentration of ATO for 13 h, followed by inoculation of the HCV-JFH1 virus, and then the infected cells were further incubated with ATO for 97 h. Results from three independent experiments conducted as described for panel A are shown. (E) The levels of the core protein in the culture supernatants treated as described for panel D were determined by enzyme-linked immunosorbent assay. Experiments were done in triplicate, and bars represent the mean core protein levels.

Effect of APO on HCV replication.

Arsenic is known to exist in two oxidation states, As(III) in ATO and As(V) in APO. As ATO in the lower valence state has been reported to be more toxic than APO (48), we compared their anti-HCV activities using an OR6 assay system, which was recently developed as a luciferase reporter assay system for monitoring genome-length HCV RNA replication in HuH-7-derived OR6 cells (Fig. 2A) (25). The results showed that APO could not strongly suppress HCV replication at submicromolar concentrations, while ATO strongly inhibited it, with an EC₅₀ of 0.33 μM (Fig. 2B and C), indicating that ATO has unique anti-HCV activity. In this context, it is relevant that the expression level of HCV core protein was also remarkably decreased in the cell lysates of O cells treated with ATO, but not those treated with APO, for 72 h (Fig. 2D). Thus, APO seems to be a useful negative probe to clarify the mechanism of the anti-HCV activity of ATO.

FIG. 2.

Effect of APO on HCV replication. (A) Schematic representation of genome-length HCV RNA encoding the RL gene as a reporter (ORN/C-5B/KE RNA) replicated in OR6 cells. The RL is expressed as a fusion protein with neomycin phosphotransferase (Neo^R). The position of an adaptive mutation, K1609E in NS3, is indicated by an open triangle. (B) Effect of ATO on genome-length HCV RNA replication. At 72 h after treatment of OR6 cells with ATO at the indicated concentrations, the replication level of HCV RNA was monitored by the RL assay. The relative RL activity is shown. The results shown are means from three independent experiments. Error bars indicate standard deviations. (C) Effect of APO on genome-length HCV RNA replication. At 72 h after treatment of OR6 cells with APO at the indicated concentrations, the replication level of HCV RNA was monitored by the RL assay as described for panel B. (D) HCV core protein expression level in O cells after treatment with either ATO or APO. The results of Western blot analysis of cellular lysates with anti-HCV core or anti-β-actin antibody in O cells at 72 h after treatment with either 1 μM ATO or 1 μM APO are shown.

ATO does not affect cell growth at submicromolar concentrations.

ATO has been reported to induce apoptosis (11, 14, 20, 21, 26-28, 33, 48, 66). Therefore, such an ATO-induced apoptosis may be involved in the anti-HCV activity. To test this possibility, we examined the effect of ATO or APO at various concentrations on cell proliferation by colorimetric MTT assay. In this context, we demonstrated that ATO did not affect the cell proliferation of O cells or the parental HCV-negative HuH-7 cells at submicromolar concentrations (Fig. 3A and E). In contrast, 4 or 8 μM ATO significantly inhibited cell proliferation (Fig. 3B and F). Similarly, APO did not affect the cell proliferation at less than 2 μM (Fig. 3C and D). Consistent with the above results, ATO-treated O cells exhibited normal growth rates and cell viabilities, at least at 1 μM for 72 h (Fig. 3G). Furthermore, we did not observe the cleavage of PARP-1, which is known to be an important substrate for activated caspase 3, in O cells treated with 1 μM ATO at least until 72 h (Fig. 3H), indicating that 1 μM ATO did not induce apoptosis in O cells. Thus, we concluded that the anti-HCV activity was independent of ATO-induced apoptosis or cell toxicity, at least at submicromolar concentrations.

FIG. 3.

Effect of ATO on cell growth and viability. (A and B) MTT assay of O cell lysates at the indicated times after treatment with ATO at various concentrations. NaOH (10 μM) was used as the solvent for ATO. The results shown are means from three independent experiments. Error bars indicate standard deviations. (C and D) MTT assay of O cell lysates at the indicated times after treatment with APO at various concentrations. (E and F) MTT assay of HuH-7 cell lysates at the indicated times after treatment with ATO at various concentrations. (G) Growth curve and viability of O cells after treatment with either 10 μM NaOH (Con) or 1 μM ATO (As). (H) Western blot analysis of cellular lysates with anti-PARP-1 or anti-β-actin antibody in O cells at the indicated times after treatment with 1 μM ATO.

PML and Chk2 are dispensable for the anti-HCV activity of ATO.

Since PML is known to be a target of ATO (70), we first examined the subcellular localization of PML in O cells treated with either 1 μM ATO or 1 μM APO for 72 h by means of an anti-PML antibody (PM001; MBL) that can recognize most of the PML splicing variants and is useful for immunofluorescence analysis. The results showed that PML was localized predominantly in punctate nuclear speckles termed PML-NBs in control O cells (Fig. 4A). Interestingly, we noticed that some nuclear PML, but not all, disappeared and was translocated into discrete cytoplasmic bodies in the O cells treated with 1 μM ATO (Fig. 4A). We also observed cytoplasmic translocation of PML in the O cells treated with 1 μM APO for 72 h (Fig. 4A). Furthermore, we observed a similar cytoplasmic translocation of PML in the HCV-negative 293FT or HeLa cells after the treatment with ATO (data not shown). Thus, we concluded that the cytoplasmic translocation of PML after the treatment with ATO was not associated with anti-HCV activity. Next, Western blot analysis to compare PML expression in the lysates of O cells treated with 1 μM ATO or 1 μM APO for 72 h was performed using another anti-PML antibody, A301-168A-1 (a gift from Bethyl Laboratories), which can recognize the longest isoform, PML I, but not shorter PML isoforms such as PML VI and which has been proven useful for Western blot analysis. Consistent with the previous finding that ATO promotes PML degradation (70), the expression level of the PML I protein was lower in the ATO-treated O cells than in the APO-treated O cells (Fig. 4B), suggesting that PML degradation by ATO is associated with anti-HCV activity. To further examine whether PML is directly involved in the anti-HCV activity of ATO, we used lentiviral vector-mediated RNA interference to stably knock down PML in the O cells. To express an shRNA targeted to all PML isoforms (56), we used the VSV G-pseudotyped HIV-1-based vector system (1, 42, 71). We used the puromycin-resistant pooled cells at 10 days after the lentiviral transduction in this experiment. Immunofluorescence and Western blot analysis demonstrated a very effective knockdown of PML in the O cells (Fig. 4C and D). We quantitatively examined the level of HCV RNA in the PML knockdown O cells treated with or without either 1 μM ATO (Fig. 4E) or 1 μM APO (Fig. 4F) for 72 h. The results showed that the replication level of genome-length HCV-O RNA in the untreated PML knockdown cells was similar to that in control cells (Fig. 4E), suggesting that PML is dispensable in HCV RNA replication. Importantly, ATO effectively inhibited the HCV RNA replication in both the PML knockdown cells and control cells compared with that of the APO-treated cells (Fig. 4E and F). Thus, we concluded that PML was dispensable for the anti-HCV activity of ATO. Since the Chk2 checkpoint kinase has recently been implicated in ATO-induced apoptosis and in association with PML (27, 63, 64, 66), we examined the anti-HCV activity in the ATO-treated Chk2 knockdown O cells (3). As we previously described, Western blot analysis demonstrated very effective knockdown of Chk2 in O cells (Fig. 4G). Accordingly, we examined the level of HCV RNA in Chk2 knockdown cells treated with or without either 1 μM ATO (Fig. 4H) or 1 μM APO (Fig. 4I) for 72 h. Consistent with our recent finding that Chk2 is required for HCV RNA replication, the replication of genome-length HCV RNA in the untreated Chk2 knockdown cells was remarkably suppressed (Fig. 4H). However, ATO strongly inhibited the HCV RNA replication in the Chk2 knockdown cells compared with that in the APO-treated Chk2 knockdown cells (Fig. 4H and I), suggesting that Chk2 is not implicated in the anti-HCV activity of ATO.

FIG. 4.

PML and Chk2 are not required for the anti-HCV activity of ATO. (A) Subcellular localization of PML in O cells at 72 h after treatment with 10 μM NaOH (−), 1 μM ATO, or 1 μM APO. PML was detected by indirect immunofluorescence analysis with anti-PML antibody (PM001). DAPI staining of the nuclear DNA is also shown. (B) Induction of PML degradation by ATO but not by APO. The results of Western blot analysis of cellular lysates of O cells at 72 h after treatment with 10 μM NaOH (−), 1 μM ATO, or 1 μM APO with anti-PML (A301-168A-1) or anti-β-actin antibody are shown. (C) Stable knockdown of PML by shRNA-producing lentiviral vector in O cells. PML was detected by indirect immunofluorescence analysis with anti-PML antibody (PM001) in O cells expressing shRNA targeted to PML (PMLi) as well as in O cells transduced with a control lentiviral vector (Con). (D) Western blot analysis of cellular lysates with anti-PML (A301-168A-1) or anti-β-actin antibody in PML knockdown O cells (PMLi) as well as in control O cells (Con). (E and F) The level of genome-length HCV-O RNA was monitored by real-time LightCycler PCR in PML knockdown O cells (PMLi) as well as in control O cells (Con) after treatment with 10 μM NaOH (−), 1 μM ATO (+) (E), or 1 μM APO (+) (F) for 72 h. Results from three independent experiments conducted as described in the legend to Fig. 1A are shown. (G) Inhibition of Chk2 expression by shRNA-producing lentiviral vector. The results of Western blot analysis of cellular lysates with anti-Chk2 or anti-β-actin antibody in O cells expressing shRNA targeted to Chk2 (Chk2i) as well as in O cells transduced with a control lentiviral vector (Con) are shown. (H and I) The level of genome-length HCV-O RNA was monitored by real-time LightCycler PCR in Chk2 knockdown O cells (Chk2i) as well as in control O cells (Con) after treatment with 10 μM NaOH (−), 1 μM ATO (+) (H), or 1 μM APO (+) (I) for 72 h. Results from three independent experiments conducted as described in the legend to Fig. 1A are shown.

Effect of ATO on the stress-signaling pathways.

To date, the focus has been on PML and PML-retinoic acid receptor α as major targets of ATO (70). On the other hand, arsenic has been reported to modulate other cell-signaling pathways, especially stress-responsive transcription factors, such as nuclear factor κB (NF-κB), activator protein 1 (AP-1), and STAT3 (12, 37, 38, 62). Therefore, we examined the involvement of several stress-responsive pathways in the anti-HCV activity of ATO by luciferase-based reporter assays or Western blot analysis using an antibody which specifically recognizes STAT3 phosphorylated at tyrosine 705. Although it has been reported that ATO inhibited the NF-κB signaling pathway through a direct interaction with IKKβ at a high concentration (more than 10 μM) (29), neither 1 μM ATO nor 1 μM APO affected the endogenous NF-κB transcriptional activity in the present study (Fig. 5A and B). Conversely, ATO at least slightly stimulated mitogen-activated protein kinase kinase kinase (MEKK)-mediated NF-κB activation (Fig. 5A and B). Since NF-κB activation has been shown to stimulate HCV replication (60), the NF-κB pathway would seem not to be essential for the anti-HCV activity of ATO. Next, regarding the AP-1 signaling pathway, both ATO and APO are known to activate c-Jun N-terminal kinase (JNK) (45). Importantly, there was no stimulation of JNK activity at a dose below 30 μM (45). In fact, 50 μM ATO but not 50 μM APO strongly stimulates AP-1 activity by inhibiting a JNK phosphatase (10). Consistently, we found that both 1 μM ATO and 1 μM APO had a marginal effect on the AP-1 signaling pathway (Fig. 5C and D), suggesting that the AP-1 pathway is also not involved in the anti-HCV activity of ATO. Regarding the STAT3 signaling pathway, ATO has been reported to inhibit the phosphorylation of the STAT3 tyrosine at 705, leading to inactivation of the JAK-STAT signaling pathway (12, 62). In contrast, it has been reported that HCV constitutively phosphorylates and activates STAT3 (49, 59, 67). In this context, we observed constitutive tyrosine phosphorylation of STAT3 in untreated O cells (Fig. 5E). Furthermore, the marginal effect of 1 μM ATO on STAT3 phosphorylation and interleukin-6-mediated STAT3 activation was also observed (Fig. 5E and F). Taken together, these results at least suggest that the NF-κB, AP-1, and STAT3 pathways may not be associated with the anti-HCV activity of ATO at submicromolar concentrations.

FIG. 5.

Effect of ATO on the stress-signaling pathways. (A and B) Effect of ATO or APO on the NF-κB signaling pathway. O cells were transfected with 100 ng of reporter plasmid, pNF-κB-Luc, and/or 100 ng of pFC-MEKK (Stratagene, La Jolla, CA). Cells were treated with either 1 μM ATO (A) or 1 μM APO (B), and an FL assay was performed 24 h later. The results shown are means from three independent experiments. The relative FL activity is shown. (C and D) Effect of ATO or APO on the AP-1 signaling pathway. O cells were transfected with 100 ng of pAP-1-Luc and/or 100 ng of pFC-MEKK (Stratagene). Cells were treated with either 1 μM ATO (C) or 1 μM APO (D), and an FL assay was performed 24 h later as described for panels A and B. (E) Effect of ATO on the phosphorylation level of STAT3 at tyrosine 705. The results of Western blot analysis of cellular lysates with anti-phospho-STAT3 (Tyr705), anti-STAT3, or anti-β-actin antibody in O cells treated with either 1 μM ATO or 1 μM APO for 24 h are shown. (F) Effect of ATO on the STAT3 signaling pathway. O cells were transfected with 100 ng of STAT3 reporter APRE-Luc (41) (STAT3-Luc, a generous gift from T. Hirano, Osaka University, Japan). Cells were treated with 1 μM ATO for 19 h and then stimulated with 100 ng/ml of interleukin-6 for 5 h, and an FL assay was performed as described for panels A and B.

The anti-HCV activity of ATO is associated with the glutathione redox system and oxidative stress.

Finally, we focused on the involvement of the glutathione redox system and oxidative stress in the anti-HCV activity of ATO. For this, we analyzed the HCV replication level after combination treatment with ATO and antioxidants such as NAC and vitamin C using the OR6 assay system. When OR6 cells were treated with either 100 μM vitamin C or 10 mM NAC alone for 24 h or 72 h, the HCV replication was slightly enhanced (Fig. 6A and B), indicating that the antioxidant can activate HCV replication. Although the anti-HCV activity in the OR6 cells treated with 1 μM ATO and in combination with 100 μM vitamin C for 24 h was weakly reduced, 10 mM NAC completely and partially eliminated the anti-HCV activity of ATO after 24 h (Fig. 6A) and 72 h (Fig. 6B) of treatment, respectively, suggesting that oxidative stress and the glutathione redox system are associated with the anti-HCV activity of ATO. In contrast, the iNOS inhibitor 1400W did not suppress the HCV RNA replication or eliminate the anti-HCV activity of ATO, suggesting that NO is not involved in the anti-HCV activity of ATO (Fig. 6C). To further examine the involvement of oxidative stress in the anti-HCV activity of ATO, we examined ROS production in ATO-treated cells using two oxidative-sensitive fluorescent probes, DHE for detection of intracellular O₂⁻ and DCF for detection of intracellular H₂O₂. We found that 1 μM ATO could generate a significant level of intracellular O₂⁻ but not intracellular H₂O₂, while 2 μM BSO, an inhibitor of glutathione synthesis (14, 20, 33), could induce both O₂⁻ and H₂O₂ (Fig. 6D to H). Importantly, NAC diminished the ATO-dependent O₂⁻ induction (Fig. 6F). Since glutathione is a major antioxidant in cells and can clear away superoxide anion free radical, we also analyzed the changes of the intracellular glutathione level in ATO-treated O cells using CMF fluorescence, which can react with glutathione. As a result, we observed significant glutathione depletion in the cells treated with at least 1 μM ATO (Fig. 6I). To further confirm the involvement of glutathione in the anti-HCV activity of ATO, we examined the effect of cotreatment with ATO and BSO. When the OR6 cells were treated with 1 μM BSO alone, the HCV replication level was suppressed by about 30% compared with that of the control cells, and this occurred without cell toxicity (data not shown). However, consistent with previous reports in which ATO-induced apoptosis was enhanced by BSO (14, 20, 33), most of the cells died, possibly through apoptosis, when the OR6 cells were cotreated with 1 μM ATO and 1 μM BSO for 72 h (data not shown), suggesting that ATO and BSO synergistically generate ROS and deplete glutathione, resulting in induction of oxidative damage. Taken together, these results suggest that ATO may inhibit the HCV RNA replication by modulating the glutathione redox system and oxidative stress.

FIG. 6.

The anti-HCV activity of ATO is associated with the glutathione redox system and oxidative stress. (A and B) The anti-HCV activity of ATO is eliminated by treatment with the antioxidant NAC. OR6 cells were treated with 1 μM ATO alone and in combination with 100 μM vitamin C (VC), with or without 10 mM NAC, for 24 h (A) or 72 h (B). The replication level of HCV RNA was monitored by the RL assay. The relative RL activity is shown. The results shown are means from three independent experiments; error bars indicate standard deviations. The results of Western blot analysis of cellular lysates with anti-HCV core or anti-β-actin antibody in OR6 cells at 72 h after the treatment with 1 μM ATO alone and in combination with 100 μM VC, with or without 10 mM NAC, are also shown. (C) Effect of combination treatment with ATO and the iNOS inhibitor 1400W on HCV RNA replication. OR6 cells were treated with 1 μM ATO alone and in combination with 1400W at the indicated concentrations for 72 h. The replication level of HCV RNA was monitored by the RL assay as described for panels A and B. (D and E) Effect of ATO on production of a ROS, O₂⁻, in O cells. O cells were treated with 1 μM ATO (D) or 2 μM BSO (E) for 24 h. The intracellular O₂⁻ level was measured by flow cytometry using DHE as described in Materials and Methods. (F) Inhibition of ATO-dependent O₂⁻ induction by NAC. O cells were treated with either 1 μM ATO or 10 mM NAC alone and in combination with 10 mM NAC for 24 h. (G and H) Effect of ATO on production of a ROS, H₂O₂, in O cells. O cells were treated with 1 μM ATO (G) or 2 μM BSO (H) for 24 h. The intracellular H₂O₂ level was measured by flow cytometry using DCF as described in Materials and Methods. (I) Effect of ATO on the intracellular glutathione level in O cells. O cells were treated with 1 μM ATO for 72 h. The intracellular glutathione level was measured by flow cytometry using CellTracker Green CMFDA as described in Materials and Methods.

DISCUSSION

ATO has been reported to affect multiple biological functions, such as PML-NB formation, apoptosis, differentiation, stress response, and viral infection (38). Indeed, ATO has been shown to increase retroviral infectivity, including infectivity of HIV-1, HIV-2, feline immunodeficiency virus, simian immunodeficiency virus from rhesus macaques, and murine leukemia virus, although the mechanisms responsible for these changes are not well understood (5, 6, 32, 44, 47, 50, 57). PML, which is involved in host antiviral defenses, is required for the formation of the PML-NB, which is often disrupted or sequestered in the cytoplasm by infection with DNA or RNA viruses (17). The fact that ATO promotes the degradation of PML and alters the morphology or distribution of PML-NBs suggests that ATO enhances HIV-1 infection by antagonizing an antiviral activity associated with PML. In fact, HIV-1 infection has been reported to alter PML localization (57), although others have failed to confirm this finding (5). Furthermore, Berthoux et al. demonstrated that ATO stimulated retroviral reverse transcription (5). Moreover, ATO has been shown to have an inhibitory effect on host restriction factors, such as TRIM5a, Ref1, and Lv1, in a cell type-dependent manner (5, 6, 32, 44, 47, 50). In contrast, we have demonstrated that ATO strongly inhibited genome-length HCV RNA replication without cell toxicity (Fig. 1A and 2A). In addition, we observed the cytoplasmic translocation of PML in the HCV RNA-replicating O cells after the treatment with ATO (Fig. 4A). However, PML was dispensable for the anti-HCV activity of ATO as well as HCV RNA replication (Fig. 4E). In this regard, it is worth noting the recent report by Herzer et al. that the HCV core protein interacts with PML isoform IV and abrogates the PML function (22). Thus, PML may be involved in the HCV life cycle. In any case, the sensitivity to ATO and the cellular target of ATO seem to be different between HCV and HIV-1.

HCV infection has been shown to cause a state of chronic oxidative stress like that seen in chronic hepatitis C, which may contribute to fibrosis and carcinogenesis in the liver (16, 18, 40). In particular, HCV replication has been associated with the endoplasmic reticulum (ER), where HCV causes ER stress. Indeed, HCV NS5A and core, the ER-associated proteins, have been reported to trigger ER stress (4, 55). Therefore, HCV infection causes production of ROS and lowering of mitochondrial transmembrane potential through calcium signaling (4, 36). Among the HCV proteins, core, E1, NS3, and NS5A have been shown to be potent ROS inducers, and these HCV proteins also alter intracellular calcium levels and induce oxidative stress, thereby inducing DNA damage, and constitutively activate STAT3 and NF-κB, which are associated with HCV pathogenesis (19, 34, 36, 43, 49, 59, 60, 67). In fact, oxidative stress has been shown to trigger STAT3 tyrosine phosphorylation and nuclear translocation, which correlate with the activation of STAT3, leading to its DNA-binding activity (9). In contrast, ATO inhibited the STAT3 tyrosine phosphorylation through direct interaction with JAK kinase, thereby suppressing the transcriptional activity of STAT3 (12, 62). Importantly, STAT3 activation has been reported to be associated with HCV RNA replication (59, 69). The STAT3 Tyr705 dominant negative mutant has been shown to inhibit HCV RNA replication, suggesting that STAT3 positively regulates HCV replication (59). In contrast, others have reported that STAT3 induces anti-HCV activity (69). In this study, we analyzed the potential effect of ATO treatment on a set of stress-signaling events, including the NF-κB, AP-1, and STAT3 pathways, since ATO is known to modulate various signaling pathways. However, at 1 μM, which exerted an anti-HCV activity, the respective signaling pathways were not affected, arguing that the anti-HCV activity is independent of these pathways (Fig. 5). In this regard, these stress-signaling pathways have been reported to be constitutively activated in HCV core- or NS5A-expressing cells (19, 36, 49, 59, 60, 67). In addition, previous studies demonstrated that ATO modulates the NF-κB, AP-1, and STAT3 pathways at higher concentrations (NF-κB, >10 μM; AP-1, >30 μM; STAT3, >4 μM). Therefore, we may have only observed the marginal effect of ATO in this study (Fig. 5). On the other hand, the HCV core or NS3 protein as well as HCV infection induces NO, leading to induction of double-stranded DNA breaks and accumulation of mutations of cellular genes (35). However, the iNOS inhibitor 1400W could not suppress HCV RNA replication and the anti-HCV activity of ATO, indicating that NO is not associated with the anti-HCV activity or with HCV replication (Fig. 6C).

It has been indicated that oxidative damage plays an important role in the effect of ATO (38). ROS generated in response to ATO exposure lead to accumulation of intracellular H₂O₂. Glutathione peroxidase and catalase are key enzymes regulating the levels of ROS and protecting cells from ATO-induced damage (26). However, the gastrointestinal glutathione peroxidase was drastically downregulated in cells harboring HCV replicons, which are rendered more susceptible to oxidative stress (39). The glutathione redox system has been implicated in the cellular defense system (14, 20). Glutathione, a major antioxidant in cells, is a tripeptide synthesized from cysteine, glutamic acid, and glycine, and it can scavenge superoxide anion free radicals. ATO has been shown to bind to the sulfhydryl group of glutathione and deplete the intracellular glutathione, resulting in enhancement of the sensitivity to oxidative damage (20, 33). Conversely, the antioxidant NAC is readily taken up by cells and serves as a precursor to elevate intracellular glutathione (53). In fact, ATO-induced apoptosis has been shown to be inhibited by NAC (11, 14, 21, 28). In this study, we have demonstrated that the anti-HCV activity of ATO was completely eliminated by treatment with NAC for 24 h (Fig. 6A). In addition, we found that ATO increased intracellular O₂⁻ but not H₂O₂ and depleted the intracellular glutathione in HCV RNA-replicating cells (Fig. 6D to I). Importantly, NAC diminished the ATO-dependent O₂⁻ induction (Fig. 6F). This finding could strengthen the link between ATO-dependent oxidative stress and anti-HCV activity. Similarly, Wen et al. reported an increase in ROS and enhanced susceptibility to glutathione depletion in the HCV core- expressing HepG2 cells (61). Accordingly, ROS have been shown to significantly suppress RNA replication in HCV replicon-harboring cells treated with H₂O₂ (13). In addition, HCV replication has been shown to be inhibited by lipid peroxidation of arachidonate, and this peroxidation could be blocked by lipid-soluble antioxidants such as vitamin E (23). Conversely, several antioxidants, such as vitamin C, vitamin E, and NAC, enhanced HCV replication in the present study (Fig. 6A and B) (65). Thus, we suggest that ATO inhibited HCV RNA replication by modulating the glutathione redox system and oxidative stress. In contrast to the above findings with HCV, NAC has been shown to suppress HIV-1 replication by preventing the activation of HIV-1 long terminal repeat transcription by NF-κB, suggesting a correlation between a decrease in glutathione levels and activation of HIV-1 replication (46, 53, 54). In this context, ATO has shown opposite effects on HIV-1 and HCV replication, stimulating the former and inhibiting the latter. Considering all of these results together, ATO can be regarded as a useful, novel anti-HCV reagent. In addition, the host redox system may be critical for HCV replication and may represent a pivotal target for the clinical treatment of patients with chronic hepatitis C.