Acetylsalicylic acid inhibits hepatitis C virus RNA and protein expression through cyclooxygenase 2 signaling pathways

Authors

  • Karina Trujillo-Murillo,

    1. Department of Biochemistry, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
  • Ana Rosa Rincón-Sánchez,

    1. University Center of Health Sciences, University of Guadalajara, Jalisco, Mexico
    Search for more papers by this author
  • Herminia Martínez-Rodríguez,

    1. Department of Biochemistry, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
  • Francisco Bosques-Padilla,

    1. Department of Gastroenterology, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
  • Javier Ramos-Jiménez,

    1. Department of Infectology, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
  • Hugo A. Barrera-Saldaña,

    1. Department of Biochemistry, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
  • Marcos Rojkind,

    1. Department of Biochemistry and Molecular Biology, George Washington, University Medical Center, Washington, DC
    Search for more papers by this author
  • Ana María Rivas-Estilla

    Corresponding author
    1. Department of Biochemistry, School of Medicine and University Hospital, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    • Department of Biochemistry, School of Medicine, Autonomous University of Nuevo Leon, Monterrey, Nuevo Leon, Mexico
    Search for more papers by this author
    • fax: (52) 81 8333-7747


  • Potential conflict of interest: Nothing to report.

Abstract

It has been reported that salicylates (sodium salicylate and aspirin) inhibit the replication of flaviviruses, such as Japanese encephalitis virus and dengue virus. Therefore, we considered it important to test whether acetylsalicylic acid (ASA) had anti–hepatitis C virus (HCV) activity. To this end, we examined the effects of ASA on viral replication and protein expression, using an HCV subgenomic replicon cell culture system. We incubated Huh7 replicon cells with 2-8 mM ASA for different times and measured HCV-RNA and protein levels by northern blot, real-time polymerase chain reaction, and western analysis, respectively. We found that ASA had a suppressive effect on HCV-RNA and protein levels (nearly 58%). ASA-dependent inhibition of HCV expression was not mediated by the 5′-internal ribosome entry site or 3′-untranslated regions, as determined by transfection assays using bicistronic constructs containing these regulatory regions. However, we found that HCV-induced cyclooxygenase 2 (COX-2) messenger RNA and protein levels and activity and these effects were down-regulated by ASA, possibly by a nuclear factor kappa B–independent mechanism. We also observed that the ASA-dependent inhibition of viral replication was due in part to inhibition of COX-2 and activation of p38 and mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2 (MEK1/2) mitogen-activated protein kinases (MAPKs). Inhibition of these kinases by SB203580 and U0126, respectively, and by short interfering RNA silencing of p38 and MEK1 MAPK prevented the antiviral effect of ASA. Taken together, our findings suggest that the anti-HCV effect of ASA in the Huh7 replicon cells is due to its inhibitory effect on COX-2 expression, which is mediated in part by the activation of MEK1/2/p38 MAPK. Conclusion: These findings suggest the possibility that ASA could be an excellent adjuvant in the treatment of chronic HCV infection. (HEPATOLOGY 2008.)

Infection with hepatitis C virus (HCV) is a common cause of liver disease, affecting 3% of the world's population.1 In most cases, the immune response is unable to clear the virus, and this results in its persistence. Patients with persistent infection are at high risk to develop chronic liver diseases, cirrhosis, and hepatocellular carcinoma.2 Current therapies of chronic hepatitis C have been restricted mainly to a combination of pegylated interferon α and ribavirin.3, 4 However, because the virus is not eliminated in approximately one-half of treated patients and because the treatment is expensive, alternative antiviral drugs that efficiently block virus replication are needed.5 Studies of HCV replication have been greatly facilitated by the development of an HCV replicon system that is capable of self-amplification in cultured hepatoma cells.6, 7 Although this system mimics only some aspects of HCV replication, it provides a valuable model for investigating viral expression and for characterizing those factors that regulate HCV expression, thus providing a unique tool for screening anti-HCV compounds.7

Recently, it has been reported that sodium salicylate and acetylsalicylic acid (ASA) inhibit the replication of flaviviruses, such as Japanese encephalitis virus (JEV) and dengue virus (DENV),8, 9 but the mechanisms have not been clearly established. Liao et al.8 reported that the mechanism by which salicylates suppress flavivirus infection may involve p38–mitogen-activated protein kinase (MAPK) activity, being independent of blocking nuclear factor kappa B (NF-κB).8 In addition, Mazur et al.9 suggested that ASA efficiently blocks influenza virus replication in vitro and in vivo by a mechanism involving expression of proapoptotic factors, inhibition of caspase activation, and blocking of the nuclear export of viral ribonucleoproteins. Salicylates are widely employed as nonsteroidal anti-inflammatory drugs (NSAIDs), which inhibit the cyclooxygenase (COX) activity, resulting in decreased synthesis of prostaglandins, leukotrienes, and thromboxane precursors.10 There are 3 known isoforms of the enzyme: COX-1, COX-2, and COX-3.11 Recently, the activation of COX-2 has been implicated in HCV replication12; however, the mechanism remains poorly defined. In this work, we investigated the effects of ASA on HCV-RNA and protein expression and analyzed possible mechanisms involved. Furthermore, the effects of inhibitors of pathways known to mediate regulation of the inducible COX-2 enzyme were analyzed. Our findings suggest that ASA exhibits anti-HCV properties in Huh7 replicon cells because of its inhibitory effect on COX-2 expression, which is mediated in part by the activation of mitogen-activated protein kinase/extracellular signal-regulated kinase kinase 1/2 (MEK1/2)/p38 MAPKs. These findings suggest the possibility that ASA could be an excellent adjuvant in the treatment of chronic HCV infection.

Abbreviations

ASA, acetylsalicylic acid; bp, base pair; cDNA, complementary DNA; COX, cyclooxygenase; DENV, dengue virus; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HCV, hepatitis C virus; IRES, internal ribosome entry site; JEV, Japanese encephalitis virus; JNK, c-Jun N-terminal kinase; MAb, monoclonal antibody; MAPK, mitogen-activated protein kinase; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; mRNA, messenger RNA; NF-κB, nuclear factor kappa B; NPT II, neomycin phosphotransferase II; NS5A, nonstructural protein 5A; NSAID, nonsteroidal anti-inflammatory drug; PCR, polymerase chain reaction; PGE2, prostaglandin E2; RLU, relative luciferase unit; RT-PCR, reverse-transcription polymerase chain reaction; siRNA, short interfering RNA; UTR, untranslated region.

Materials and Methods

Cell Culture and Aspirin Treatment.

The original wild-type pFKI389-NS3-3′ replicon DNA from genotype 1b and the generation of Huh7 HCV replicon cells have been described.6 Huh7 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 2% heat-inactivated fetal bovine serum, 1% nonessential amino acids, 100 U of penicillin G, and 100 μg of streptomycin/mL at 37°C in a humidified atmosphere with 5% CO2. Cells containing the HCV replicon were maintained in the culture in the presence of 500 μg of G418/mL. For ASA treatment, Huh7 HCV replicon cells were incubated with an increasing concentration of aspirin (0-8 mM) for up to 72 hours. There were no cytotoxic effects of aspirin at concentrations of 4 mM or less on HCV replicon–containing cells, as shown by the Alamar blue reduction assay (Fig. 1B). Hepatic enzyme activities [alanine aminotransferase (ALT) and aspartate aminotransferase (AST)] were measured in supernatant culture media with Vitros ALT/AST-DT slides (Ortho-Clinical Diagnostics, Rochester, NY). After each experiment, total cellular RNA and proteins were isolated from cell cultures and subjected to real-time reverse-transcription polymerase chain reaction (RT-PCR) and immunoblot analysis, respectively. ASA was purchased from Sigma (St. Louis, MO), and the chemical purity was ≥99%.

Figure 1.

(A) Effect of ASA on HCV-RNA levels in Huh7 HCV replicon cells. Huh7 HCV replicon cells (2 × 105 cells) plated in 6-well plates were incubated with 2, 4, or 8 mM ASA or without ASA at different time points (24, 48, and 72 hours). At the end of incubation, HCV-RNA levels were quantified by real-time RT-PCR using TaqMan probes (Applied Biosystems; *P < 0.05). (B) Assessing cell viability. The Huh7 HCV replicon cells were treated with ASA at indicated concentrations for 24, 48, and 72 hours. The cell viability was then determined in triplicate by the Alamar blue assay and compared with untreated cells (100% viability). (C,D) Evaluation of ALT and AST levels in parental and HCV replicon cells treated with 4 mM ASA. As a positive control of cellular damage, we treated the cells with 20 mM CCl4 for 6 hours, and then enzyme levels were quantified.

COX-2 and HCV Plasmid Constructs.

The COX-2 reporter plasmid P2-1900-Luc (−1796, +104), with two copies of upstream NF-κB binding sites, and COX-2 P2-431-NF-κB mut-Luc, with mutated NF-κB binding sites, were kindly provided by M. A. Íñiguez (Centro de Biología Molecular, Madrid, Spain).13 The bicistronic constructs, containing either the hepatitis C virus internal ribosome entry site (HCV-IRES) alone or the HCV-IRES and 3′-untranslated region (3′-UTR) under control of the T7-RNA polymerase and having the luciferase gene downstream, as well as the plasmid pFKI389-NS3-3′ wild type, have been described previously.14

Transient Transfection Assays.

Huh7 parental cells (2 × 105) were infected with recombinant vaccinia virus containing the T7-RNA polymerase gene15 for 1 hour, and this was followed by transfection with DNA containing the gene of interest in the expression vector under the control of T7 promoter. Cells were incubated in serum-free medium at 37°C for 6 hours, and this was followed by the addition of completed medium and incubation for an additional 48 hours; then, protein extraction was performed.

Luciferase Assays.

Transiently transfected cells (1 × 104 cells) were lysed in reporter lysis buffer according to the manufacturer's recommendation (Promega, Madison, WI), and cellular lysates were analyzed for luciferase expression with a luminometer. All transfections included Renilla expression vector (pRL-CMV) to serve as an internal control.

RNA Extraction.

Total RNA was extracted from Huh7 HCV replicon cells with Trizol (Life Technologies) according to the manufacturer's specification. RNA precipitates were then washed once in 75% alcohol and resuspended in 30 μL of RNase-free water.

RT-PCR for HCV-RNA Quantification.

Total RNA extracted was subjected to reverse transcription with a high-capacity complementary DNA (cDNA) archive kit (Applied-Biosystems, Foster City, CA) according to the manufacturer's specification. cDNAs (200 ng) were subjected to real-time polymerase chain reaction (PCR) for HCV and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) messenger RNA (mRNA) quantification. Amplifications were conducted in triplicate with the following primers: HCV forward (+75-93 nt), 5′-GCGTCTAGC CATGGCGTTA-3′; HCV reverse (+138-157 nt), 5′-GGTTCCGCAGACCACTATGG-3′; and the TaqMan probe (+94-110 nt), 5′-FAM-CTGCACGACACT CATAC-NFQ-3′. Thermal cycling conditions were as follows: initial setup at 50°C for 2 minutes and then 95°C for 10 minutes, followed by 40 cycles of 95°C for 15s and 60°C for 60s. For each PCR reaction, 12.5 μL of TaqMan PCR master mix, 1.25 μL of 20X assay mix, and 11.25 μL of cDNA were added. Fluorescence was monitored at the annealing step, and the amplification plots were generated. GAPDH-RNA expression was used to normalize the RNA concentration. For GAPDH-RNA quantification, we used a GAPDH (20×) assay (Applied Biosystems) according to the manufacturer's specification.

RT-PCR for COX-2–RNA Semiquantification.

Cells were treated with 4 mM ASA for the time points indicated and then harvested. Total cellular RNA was extracted and subjected to reverse transcription. cDNA (2 μg) was amplified by PCR for 28 cycles, each one consisting of 2 minutes at 95°C followed by cycles of 95°C for 1 minute, 60°C for 1 minute, and 72°C for 1 minute, with a final incubation at 72°C for 3 minutes. A set of primers, upper (sense; 5′-CCATCACCATCTTCCAGGAGCG-3′) and lower (antisense; 5′-AAGGCCATGCCAGT GAGCTTC-3′), were used to amplify a fragment of 313 base pairs (bp) of the COX-2 gene. RT-PCR for GAPDH mRNA (yielding a 483-bp fragment) was performed in parallel to show an equal amount of total RNA.

Northern Blot Analysis.

Huh7 cells (5 × 105) were treated with 4 mM ASA for the time points indicated in the figures, and then total cellular RNA was isolated according to the Chomczynski-Sacchi protocol.16 Ten micrograms of total RNA was denatured and subjected to northern analysis with a32P-labeled cDNA consisting of either the entire viral subgenomic sequence of pFKI389-NS3-3′ or ribosomal 18S DNA.17 The results of hybridization were visualized by autoradiography.

Protein Extraction and Immunoblot Assay.

Cell lysates were prepared and proteins were extracted as described previously.14 Equal amounts of protein were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes. Membranes were incubated with one of the following antibodies: mouse anti-HCV nonstructural protein 5A (NS5A) monoclonal antibody (MAb; 1:1000; Biodesign-International, Saco ME), anti–neomycin phosphotransferase II (anti–NPT-II; 1:1000; Cortex Biochem, San Leandro, CA), anti-luciferase (1:1000; Serotec, Raleigh, CA), anti–COX-2 MAb (1:1000; Cayman-Chemical, Ann Arbor, MI), anti-MEK1 (1:400; CST, Danvers, MA), anti-p38 (1:200; CST), or anti-actin MAb (1:1000; MP-Biomedicals, Aurora, OH). After being washed, membranes were incubated with horseradish peroxidase–conjugated goat anti-mouse immunoglobulin G or goat anti-rabbit immunoglobulin G. Detection was performed with the enhanced chemiluminescence detection system (Amersham, Freiburg, Germany). All the western blots were stripped and reprobed to detect NS5A, NPT-II, luciferase, COX-2, and actin proteins in the same blot.

Intracellular Prostaglandin E2 (PGE2) Measurements.

Huh7 HCV replicon cells (1 × 104) were treated with 4 mM ASA at different time points, and then cultured cells were harvested and cell membranes were hydrolyzed to release intracellular PGE2. PGE2 levels were then assayed with the PGE2 enzyme-linked immunosorbent assay system (Amersham) according to the manufacturer's protocol. The minimum detectable level in this assay was 2.5 pg/well.

RNA Interference Assay.

A prevalidated heterogeneous mixture of 21- to 22-bp short interfering RNAs (siRNAs) and prevalidated single-sequence siRNA that specifically inhibit p38 and MEK1 MAPK expression, respectively, were used according to the manufacturer's specification (CST). HCV replicon cells (1.5 × 105) were seeded as described before, and they were 30%-50% confluent at the time of transfection and incubated in the presence or absence of 4 mM ASA 30 minutes before transfection. The siRNAs were transfected at final concentrations of 100 and 20 nM for MEK1 and pool p38-MAPK, respectively, with the siPORT lipid transfection agent (Ambion, Austin, TX). As negative controls, we used silencer negative control siRNA (Ambion) and untransfected cells containing the siPORT lipid agent alone. Cells were incubated for 72 hours and then were harvested. Protein and total cellular RNA were extracted and analyzed by western blot and real-time RT-PCR quantification, respectively.

Statistical Analysis.

All variables were tested in triplicate, and experiments were repeated at least 3 times. One-way analysis of variance was used to test for differences in means, and the t test was used for comparisons. The differences were considered significant if P < 0.05.

Results

ASA Treatment Down-Regulates Subgenomic HCV- RNA Expression in a Dose-Dependent Manner.

To evaluate the effect of ASA on HCV-RNA expression in HCV replicon–containing cells, we incubated these cells with ASA at different concentrations (2-8 mM) and at 3 different time points (24, 48, and 72 hours), and then total cellular RNA was extracted and subjected to real-time RT-PCR for HCV-RNA quantification. ASA inhibited HCV-RNA levels in a concentration-dependent fashion compared with untreated cells, showing a higher effect at a concentration of 8 mM. In addition, this effect was time-dependent because we observed lower HCV-RNA expression in ASA-treated cells at 72 hours post-treatment (Fig. 1A). It has been reported that salicylates in concentrations of 1-5 mM are required for patients treated for chronic inflammatory diseases and that concentrations higher than 6.5 mM are too toxic for clinical use.8 To determine if the antiviral effects of ASA were due merely to the cytotoxic effect on treated cells or may reflect an effect of the differentiation or proliferation of the replicon cells, because cell proliferation is known to be tightly linked to HCV-RNA replication activity, total cell count and viability determinations were performed. Figure 1B shows that no significant difference in cell number and viability were detectable among untreated (100% viability) and treated cells with ASA concentrations of 2 and 4 mM (98% and 92%, respectively). Contrary to that, cells treated with 8 mM ASA showed lower cell survival (58%). On the basis of this finding, we selected 4 mM ASA for all subsequent experiments. To further confirm that this amount of ASA does not induce cellular damage, we quantified AST and ALT activities in the culture supernatants. As a positive control of cellular damage, we treated cells with 20 mM CCl4. Figure 1C suggests that 4 mM ASA leads to a light increase in AST activity of both replicon and parental cells at all 3 time points (comparing ASA-treated and nontreated cells); however, this effect does not reflect real cell damage. In contrast, Fig. 1D shows that 24, 48, and 72 hours after ASA treatment, the two cell types do not present significantly different levels of ALT enzyme compared with the basal levels shown by the same cell types in the absence of ASA. Together, these results confirm that the 4 mM ASA treatment did not appear to induce cytotoxicity, morphological changes, or apoptosis, as is reflected in the CCl4-treated cells.

We next explored whether the effect of ASA on HCV-RNA levels is time-dependent. Replicon cells were incubated with 4 mM ASA and then harvested at 3 different time points. Northern blot assay showed that viral RNA was down-regulated in a time-dependent fashion, reaching the lower value at 72 hours (Fig. 2A, lane 4). These effects were not associated with significant changes in the expression of cellular 18S ribosomal RNA. This result was further confirmed by real-time PCR assay, in which the maximum inhibition of the HCV-RNA level was also observed at 72 hours post-treatment (32% of inhibition at 48 hours and 58% of inhibition at 72 hours; **P < 0.01; Fig. 2B). These results reveal that 4 mM ASA possesses an antiviral effect in this replicon cell culture system.

Figure 2.

Effect of ASA on Huh7 HCV replicon cells at different time points. (A) Huh7 HCV replicon cells (5 × 105 cells) were incubated with (lanes 2-4) or without (lane 1) ASA (4 mM) for 24, 48, and 72 hours. Total cellular RNA was isolated from the cell cultures and subjected to northern blot assay to detect viral (top panel) and ribosomal 18S (bottom panel) RNA. The reactive bands were quantified with Phoretix 1D version 2003.02 software, and the ratio of HCV to S18 is indicated. (B) Huh7 HCV replicon cells (2 × 105 cells), plated in 6-well plates, were incubated with 4 mM ASA or without ASA for 24, 48, and 72 hours. Upon ASA treatment, HCV-RNA levels were quantified by real-time RT-PCR. RNA viral levels were normalized on the basis of the ratio of HCV to GAPDH mRNA, which was amplified in the same plate by the real-time RT-PCR. Data are expressed as HCV-RNA levels with respect (fold) to a nontreated control, which is defined as 1.0. The data shown are the mean ± standard deviation of triplicate cultures, and the experiment was repeated 3 times (**P < 0.01).

Aspirin Decreases HCV Protein Expression in Huh7 HCV Replicon Cells.

To determine whether ASA can influence the synthesis of viral proteins, NS5A and NPT-II protein levels were determined by western analysis in replicon cells treated with 4 mM ASA. We observed that ASA-treated cells expressed lower levels of NS5A and NPT-II proteins in a time-dependent fashion, as shown by the lower ratio of NS5A and NPT-II proteins at 72 hours compared with that of the control (nontreated cells; Fig. 3). This inhibitory effect of ASA was not likely due to its cytotoxic effect on replicon cells because total protein synthesis did not differ significantly among cells treated and not treated with the same amount of ASA (data not shown). These data suggest that ASA treatment may either diminish the translational rate of viral proteins or decrease viral protein stability in addition to the negative effect on HCV-RNA levels.

Figure 3.

HCV NS5A protein levels in Huh7 HCV replicon cells treated with ASA. Huh7 HCV replicon cells (5 × 105 cells) were incubated in the presence (lanes 5-7) or absence (lanes 1-4) of 4 mM ASA at different time points (24, 48, and 72 hours). Cell lysates were prepared, and equal amounts of protein extracts (40 μg) were subjected to immunoblot analysis to detect NS5A and NPT-II (top panel) and actin levels (bottom panel). The ratios of NS5A to actin and NPT-II to actin proteins from the immunoblot detection were quantified with Phoretix 1D version 2003.02 software.

ASA-Mediated Down-Regulation of HCV Expression Is Not Mediated by 5′-IRES and 3′-UTR Regions.

Regarding the aforementioned results, to test whether ASA could impair viral RNA translation and involve the HCV-IRES and/or 3′-viral UTR, we performed the following experiment. Bicistronic constructs containing the HCV-IRES or HCV-IRES and 3′-UTR downstream of the luciferase gene14 were transiently transfected into Huh7 parental cells and analyzed for their responses to 4 mM ASA treatment after 48 hours. We observed only a slightly nonsignificant change in luciferase activity upon treatment with 4 mM ASA (Figs. 4A,B), which suggested that the inhibitory effects of ASA shown in Figs. 1-3 are not due to inhibition of translation via the HCV-IRES and/or 3′-UTR elements.

Figure 4.

The effect of ASA on HCV replication does not involve HCV-IRES or the 3′-viral UTR regions. Huh7 parental cells (1 × 104 cells) were infected with recombinant vaccinia virus containing T7 RNA polymerase gene, and this was followed by transient transfection with 0.1 μg of (A) plasmid pHCV-IRES or (B) pHCV-IRES plus 3′-UTR downstream of the luciferase gene. Cells were incubated in serum-free medium at 37°C for 6 hours, and this was followed by the addition of completed medium and incubation with 4 mM ASA for an additional 48 hours. After this time, cells were harvested, and luciferase assays were performed. The values represent the averages of 3 separate experiments, each performed in triplicate. Data are expressed as relative luciferase units (RLUs) ± the standard deviation to the control (transfected cells without ASA treatment), which is defined as 1.0. All transfections included pRL-CMV to serve as an internal control.

HCV-Induced COX-2 mRNA and Protein Levels and Activity Are Down-Regulated by ASA in Huh7 HCV Replicon Cells.

The effectiveness of NSAIDs has been attributed to their ability to inhibit prostaglandin production by blocking COX-1/2 activity. Previous studies have reported that production of COX-2 and PGE2 modulates replication of human cytomegalovirus, gammaherpesvirus, and hepatitis B virus.18–20 To investigate the possible mechanism(s) responsible for ASA-mediated down-regulation of HCV-RNA expression, we examined COX-2 activity, mRNA, and protein levels in ASA-treated HCV replicon cells. In this study, we found that 4 mM ASA treatment down-regulated COX-2 protein (Fig. 5A) and mRNA levels (Fig. 5B) in replicon cells at 24, 48, and 72 hours, and this suggests that COX-2 gene expression may be involved in the ASA-mediated down-regulation of HCV expression.

Figure 5.

Effect of ASA on COX-2 expression and activity. (A) Huh7 HCV replicon cells (5 × 105 cells) were treated with ASA as described before (Fig. 3). Cell lysates were prepared, and equal amounts of protein extracts (40 μg) were subjected to immunoblot analysis to detect COX-2 (top panel) and actin levels (bottom panel). (B) Huh7 parental and Huh7 HCV replicon cells (2 × 105 cells), plated in 6-well plates, were incubated with or without 4 mM ASA for 24, 48, and 72 hours. At the end of incubation, total cellular RNA was extracted at each time point and analyzed by semiquantitative RT-PCR. PCR was performed for 28 cycles. RT-PCR for GAPDH mRNA was performed in parallel to show an equal amount of total RNA in each sample. The ratio of COX-2 to GAPDH mRNA from RT-PCR detection was quantified with Phoretix 1D version 2003.02 software. Data are expressed as relative band intensities to the control (cells without treatment), which is defined as 1.0. (C) Huh7 HCV replicon cells (1 × 104 cells) were incubated in the presence or absence of ASA (4 mM) and incubated at different time points (24, 48, and 72 hours). After the ASA treatment, the intracellular PGE2 levels were then assayed with the Biotrak PGE2 enzyme immunoassay system (Amersham Biosciences) according to the manufacturer's protocol.

We further evaluated COX-2 activity by monitoring PGE2levels. As we expected, in ASA-treated replicon cells, PGE2 levels were significantly decreased upon treatment in comparison with untreated control cells (Fig. 5C). These results suggest that the antiviral effect of ASA could be mediated by the blocking of COX-2 and/or its further product.

Overexpression of COX-2 Induces Viral Protein Expression.

We characterized the involvement of COX-2 in the modulation of HCV expression using transient cotransfection assays. Huh7 parental cells were transiently transfected with increasing amounts of the plasmid pcDNA-COX-2, which expresses COX-2 protein,21 and cotransfected with the same amount of plasmid expressing the HCV subgenomic DNA (pFKI389-NS3-3/wild type; 1 μg; Fig. 6).14 We observed that NS5A protein levels increased in parallel to the increase in COX-2 protein levels. Therefore, these findings suggest that an overexpression of COX-2 protein induces HCV-NS protein expression in HCV replicon cells.

Figure 6.

Overexpression of COX-2 induces viral protein expression. Huh7 parental cells (1 × 104 cells) were infected with recombinant vaccinia virus, transiently transfected with increasing amounts (0.1-0.5 μg) of the plasmid pcDNA-COX-2, which expresses COX-2 protein, and cotransfected with 0.5 μg of plasmid expressing the HCV subgenomic DNA (pFKI389-NS3-3/wild type). Cells were incubated in serum-free medium at 37°C for 6 hours, and this was followed by the addition of completed medium and incubation for an additional 48 hours. After this time, cells were harvested, and protein extraction was performed. Equal amounts of protein extracts (40 μg) were subjected to immunoblot analysis to detect NS5A and COX-2 (top panel) and actin levels (bottom panel).

ASA Regulates COX-2 Independently of Activation of NF-κB Binding Sites.

To further investigate the effect of ASA on the regulation of COX-2 expression in HCV replicon cells, we performed reporter assays. Huh7 parental cells were transiently transfected with the plasmid COX-2-P2-1900-Luc or pCOX-2-P2-431-NF-κB-mut and then stimulated with 4 mM ASA for 48 hours. No significant changes post-treatment were observed in luciferase expression directed by both plasmids (Fig. 7, lanes 2 and 5) in comparison with untreated cells (Fig. 7, lanes 1 and 4). To discriminate between COX-1 and COX-2 inhibition by ASA, we also treated transfected cells with the selective COX-2 inhibitor, NS398, at the concentration of 100 μM, and we observed the same effect obtained with ASA treatment (Fig. 7, lanes 3 and 6).

Figure 7.

ASA does not down-regulate COX-2 transcriptional activity. Huh7 parental cells (1 × 104 cells) were infected with recombinant vaccinia virus and transiently transfected with 0.05 μg of plasmid COX-2-P2-1900-Luc (luciferase gene expression controlled by the COX-2 promoter/enhancer with two copies of NF-κB binding sites upstream) and pCOX-2-P2-431-NF-κB-mut (luciferase gene expression controlled by the COX-2 promoter with deleted sites for NF-κB binding). Cells were incubated in serum-free medium at 37°C for 6 hours, and this was followed by the addition of completed medium and incubation with 4 mM ASA or 100 μM NS398 for an additional 48 hours. After this time, cells were harvested, and luciferase assays were performed. The values represent the averages of 3 separate experiments, each performed in triplicate. Data are expressed as relative luciferase units (RLUs) ± the standard deviation to the control (transfected cells without treatment), which is defined as 1.0.

Inhibition of p38 and MEK1/2 but Not c-Jun N-Terminal Kinase (JNK) Activation Reverses the Effect of ASA.

Recent evidence indicates that salicylates are able to activate MAPK.22 In addition, the expression of COX-2 is highly regulated by a number of MAPKs and NF-κB. Therefore, we determined whether the antiviral action of ASA was through the modulation of MAPK activity. We found that both of the MAPK inhibitors, SB203580 and U0126, which are specific inhibitors for p38 and MEK1/2 MAPK, respectively, partially reversed the negative effect of 4 mM ASA on HCV protein expression (Fig. 8A,B; compare lane 5 with lanes 6-8). In addition, we achieved specific inhibition of p38 and MEK1/2 MAPKs using RNA interference technology in Huh7 HCV replicon cells. When we introduced siRNA to down-regulate the expression of p38 and MEK1/2 MAPK, HCV-RNA levels were increased compared with those of control cells (untreated cells and cells transfected with nonsense control siRNA; Fig. 9A). However, when we down-regulated the expression of p38 and MEK1/2 MAPKs in cells treated with 4 mM ASA, the HCV-RNA levels were lower than those observed in cells treated only with the siRNAs but higher than those in cells treated only with ASA (Fig. 9A). We performed immunoblot analysis to confirm down-regulation of p38 and MEK1 proteins and to detect NS5A levels (Fig. 9B). We observed that p38 and MEK1/2 MAPK inhibition partially reversed the negative effect of 4 mM ASA on NS5A protein expression, as demonstrated previously by the specific inhibitors. Contrary to these results, when replicon cells were incubated with SP600125, a potent JNK inhibitor; we observed a dramatic inhibition of NS5A protein levels (Fig. 8C). Furthermore, when we treated cells with both 4 mM ASA and SP600125, we observed an additive effect between both compounds that inhibited viral protein levels (Fig. 8C, lanes 6-8). Together, these results suggest that the activity of MAPKs may play a role in the modulation of HCV subgenomic replication by ASA in cultured cells.

Figure 8.

ASA reduces HCV expression, and this effect is modulated by p38 and MEK1/2 but not JNK. Huh7 HCV replicon cells (5 × 105 cells) were incubated in the presence or absence of 4 mM ASA and with or without (A) SB203580, (B) U0126, or (C) SP600125 inhibitors for 48 hours. Cell lysates were prepared, and equal amounts of protein extracts (40 μg) were subjected to immunoblot analysis to detect NS5A and NPT-II (top panel) and actin levels (bottom panel).

Figure 9.

Effect of p38 and MEK1 MAPK siRNA on HCV expression. Huh7 HCV replicon cells were incubated in the presence or absence of 4 mM ASA 30 minutes before transfection. Cells were transfected with 20 nM pool p38 MAPK siRNA (Cell Signaling, part number 6386), 100 nM MEK1 siRNA (Cell Signaling, part number 6426), silencer negative control siRNA (Ambion, part number AM4611), or siPORT lipid agent alone and incubated for 72 hours. After this time, cells were harvested, and protein and total cellular RNA were extracted. (A) HCV-RNA levels were quantified by real-time RT-PCR and normalized on the basis of the ratio of HCV to GAPDH mRNA, which was amplified in the same plate by the real-time RT-PCR. Data are expressed as HCV-RNA levels with respect (fold) to cells transfected with silencer negative control siRNA, which is defined as 1.0. (B) Equal amounts of protein extracts (40 μg) were subjected to immunoblot analysis to detect NS5A, p38 (Cell Signaling, part number 9212), MEK1 (Cell Signaling, part number 9124; top panel) and actin levels (bottom panel). The experiment demonstrates silencing of p38 MAPK and MEK1 expression. The data shown are the mean ± standard deviation of triplicate cultures, and the experiment was repeated 3 times.

Discussion

It has been reported that salicylates induce antiviral activity against some flaviviruses8 and exert an influence on the replication of other viruses, such as the influenza virus, human cytomegalovirus, and varicella zoster virus.23, 24 Despite extensive studies of NSAIDs, little is known about their antiviral effect. In this study, we found that ASA suppressed HCV expression in a hepatoma cell line containing HCV subgenomic replicon.

To investigate the possible mechanism(s) involved, we examined COX-2 mRNA and protein levels and activity. Previously, we reported that COX-2 protein expression was already induced in Huh7 replicon cells, whereas it was almost undetectable in parental cells.25 In addition, in this study we found that ASA induced a time-dependent decrease in COX-2 mRNA and protein levels independently of the activation of NF-κB binding sites in HCV replicon cells. These findings suggest that COX-2 gene expression may be involved in the ASA-mediated down-regulation of HCV expression. Some contradictory results regarding the role of COX-2 in viral replication have been reported.26, 27 Recently, the activation of COX-2 mediated by Ca2+ signaling and reactive oxygen species has been implicated in HCV replication.12 Waris et al.12 reported increasing HCV-RNA levels in FCA4-HCV replicon cells incubated with COX-2 inhibitors, suggesting that COX-2 activity down-regulates HCV-RNA levels. Contrary to that, when we inhibited COX-1/2 activity by ASA treatment in HCV replicon cells, ASA significantly decreased COX-2 activity (Fig. 5C) and down-regulated HCV-RNA expression (Fig. 2A,B). There are some possible explanations for the divergent results. First, different stable replicon cell lines were used in our study in comparison with Waris et al.'s study (FCA4 cells are a Huh7 cell line stably expressing a HCV subgenomic replicon with a single adaptive mutation, so it is very likely that both replicons are genetically different). Second, in Waris et al.'s study, FCA4 cells were treated for 24 hours with two different COX-2 inhibitors (celecoxib and NS-398); we used ASA as a COX-1/2 inhibitor, and cells were treated for up to 72 hours. Third, in Waris et al.'s study, they reported a single dose of NSAID, in contrast to our work, in which we did a dose-dependent study. These results provide new clues to explain the mechanism(s) involved in the ASA-mediated anti-HCV effect; however, the mechanism still remains poorly defined, and additional experiments need to be performed to elucidate it.

Regarding modulation of HCV expression by COX-2, as we mentioned before, ASA down-regulated COX-2 mRNA levels and HCV protein expression. These findings suggest that COX-2 activity may be involved in the ASA-mediated down-regulation of HCV expression. Some reports show that NSAIDs contributed to controlling hepatic lipid peroxidation and hence oxidative stress promoted by different agents.28 ASA inhibits COX activity, in addition to scavenging a variety of free radicals, and acts also as an anti-inflammatory agent because it decreases steady-state COX-2 mRNA levels and catalytic activity. It is possible that additional immediate-early genes are also directly affected by ASA or COX-2 inhibitors and that the loss of their products might also affect downstream events. The decreased viral RNA levels by ASA treatment may be partially explained by these mechanisms.

We show in this communication that overexpression of COX-2 (Fig. 6) induces viral protein expression. In addition, it has been reported that inhibition of COX activity reduces rotavirus infection29; contrary to that, Chen et al.30 demonstrated that the antiviral action of salicylate is independent of the inhibition of COX activity because the doses used to suppress JEV amplification are higher than those required to inhibit prostaglandin synthesis. In the same study, MAPK inhibitors reversed the antiviral effect of salicylate. Taken together, these results show that salicylate suppresses JEV propagation in vitro through the modulation of the MAPK signaling pathway as we observed for HCV.30

Accumulated evidence indicates the antiviral effect of NF-κB inhibition in cytomegalovirus, DENV, respiratory syncytial virus, and human immunodeficiency virus 1 infection.23, 31, 32 However, in this study, the inactivation of NF-κB binding sites in the COX-2 promoter region failed to modify COX-2 promoter activity. Therefore, most likely the inactivation of NF-κB is not involved in the observed COX-mediated antiviral effect of ASA.

It has been reported that the activation of MAPKs can be modulated by salicylates.22 Some reports have demonstrated that the antiflavivirus effect of salicylates is partially reversed by blocking p38 and MEK1/2 activation.8, 30 Our results using MAPK inhibitors and introducing siRNAs to down-regulate the expression of p38 and MEK1/2 MAPK demonstrated that inhibition of p38 and MEK1/2 but not JNK is involved in the antiviral effect of ASA (Figs. 8 and 9). In addition, it has been reported recently that JNK inhibitor SP600125 reduces COX-2 expression by attenuating mRNA stability in activated murine J774 macrophages.33 Then, we can hypothesize that the severe down-regulation of NS5A protein in cells treated with both ASA and SP600125 (Fig. 8C) was due to an additive effect between ASA and JNK inhibitor reducing COX-2 activity in treated cells. Together, these results provide new clues to explain the mechanism(s) involved in the antiviral effect of ASA against HCV.

With respect to our results indicating that ASA-mediated inhibition of HCV expression is not mediated by 5′-IRES and 3′-UTR regions, further experiments should be performed in order to determine whether ASA treatment affects the stability of viral proteins at a posttranslational level.

Whether our results are relevant to designing therapeutic strategies aimed at improving the efficacy of anti-HCV drugs in HCV-infected patients is not clear yet because the Huh7 cell line may not reflect the complete scenario observed in infected hepatocytes. Indeed, contradictory results have been reported on the association of interferon-α with NSAIDs in treated patients, but they used different pharmacological agents and times of evaluation.34–37 Some reports have indicated the failure of NSAIDs and interferon combination therapy in improving interferon-resistant chronic hepatitis C,34, 38 whereas other authors, using ketoprofen plus interferon, have reported an improved virological response in chronic hepatitis C patients.35 Giambartolomei et al.39, 40 reported that indomethacin potentiates the interferon-α signaling pathway by increasing signal transducer and activator of transcription 1 phosphorylation in vitro, improving cellular response. It must be noted that the aforementioned studies used NSAIDs with different biochemical activities and did not include comparable groups of HCV patients. Indeed, we showed that, at least in vitro, ASA decreased the HCV viral expression. If these results are confirmed, ASA could be considered an excellent adjuvant in the treatment of chronic HCV infection.

Ancillary

Advertisement