ABSTRACT
Although combination therapy with interferon and ribavirin has improved the treatment for chronic hepatitis C virus (HCV) infection, the detailed anti-HCV effect of ribavirin in clinical concentrations remains uncertain. To detect the anti-HCV effect of ribavirin in lower concentrations, a sensitive and accurate assay system was developed using the reporter replicon system with an HCV genotype 2a subgenomic replicon (clone JFH-1) that exhibits robust replication in various cell lines. This reporter replicon was generated by introducing the luciferase reporter gene (instead of the neomycin resistance gene) into the subgenomic JFH-1 replicon. To assess the replication of this reporter replicon, luciferase activity was measured serially up to day 3 after transient transfection of Huh7 cells. The luciferase activity increased exponentially over the time course of the experiment. After adjustment for transfection efficiency and transfected cell viability, the impacts of interferon and ribavirin were determined. The administration of interferon and ribavirin resulted in dose-dependent suppression of replicon RNA replications. The 50% inhibitory concentration of interferon and ribavirin was 1.80 IU/ml and 3.70 μg/ml, respectively. In clinical concentrations, replications were reduced to 0.09% and 53.74% by interferon (100 IU/ml) and ribavirin (3 μg/ml), respectively. Combination use of ribavirin and interferon enhanced the anti-HCV effect of interferon by 1.46- to 1.62-fold. In conclusion, we developed an accurate and sensitive replicon system, and the antivirus effect of interferon and ribavirin was easily detected within their clinical concentrations by this replicon system. This system will provide a powerful tool for screening new antiviral compounds against HCV.
Hepatitis C virus (HCV) is a major public health problem, infecting an estimated 170 million people worldwide. HCV causes chronic liver diseases, including cirrhosis and hepatocellular carcinoma, because most patients fail to clear the virus and the persistent infection that follows (1, 11, 20). Current therapy for HCV-related chronic hepatitis is based on the use of interferon (IFN). However, virus clearance rates are limited to approximately 10 to 20% of cases treated with IFN only (9, 23, 26). Combination therapy with IFN and ribavirin improves the HCV clearance rate, although the molecular mechanism responsible for this improvement is not yet fully understood (23, 25, 26). However, some direct antiviral mechanisms of ribavirin have been proposed (19). One possible mechanism is the direct inhibition of HCV RNA-dependent RNA polymerase, and another possibility is the RNA mutagen effect that drives a rapidly mutating RNA virus over the threshold to “error catastrophe.” The detection of these direct anti-HCV effects has been hampered by the lack of an appropriate sensitive system for evaluating HCV replication.
Although HCV belongs to the Flaviviridae family and has a genome structure similar to those of the other flaviviruses (3, 27), efficient cell culture systems and small animal infection models for HCV have not yet been established. This disadvantage not only hampers the understanding of the life cycle of this virus but also prevents the development of adequate antiviral compounds against HCV infection. As an important step toward overcoming this disadvantage, a subgenomic HCV RNA replicon system has been developed and enabled the assessment of HCV replication in cultured cells (22). Although this represents a powerful tool in the study of HCV replication mechanisms and the search for potential antiviral agents, functional replicons have previously been reported only for genotype 1, and efficient replications of these replicons have been accomplished only in limited human hepatocyte-derived cell lines and with some adaptive mutations. To overcome these limitations, we developed an HCV genotype 2a subgenomic replicon system using a clone isolated from a patient with fulminant hepatitis (14, 15). This replicon system provides higher colony formation efficiency and robust replication not only in hepatocyte-derived cell lines but also in non-hepatocyte-derived cell lines, and adaptive mutations are not necessary for replication (6, 16). Recently, the culture cell-generated HCV particles of this clone have been demonstrated to be infectious for both Huh7 cells and a chimpanzee (21, 30, 32). This is the only clone which can produce infectious particles in Huh7 cells, and the replication of this clone in Huh7 cells is closely related to producing infectious particles. In the present study, we used the robust replicable subgenomic replicon of this clone to develop a sensitive and accurate assay system for anti-HCV effects, and we detected the suppression effect of both IFN and ribavirin in clinical concentrations.
MATERIALS AND METHODS
Cell culture system.Huh7 cells were cultured at 37°C in 5% CO2. Cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, as previously described (15).
IFN and ribavirin.Recombinant human alpha 2a IFN was obtained from Nippon Roche (Roferon-A; Tokyo, Japan). Ribavirin was purchased from Sigma-Aldrich (St. Louis, MO).
Reporter replicon constructs and RNA synthesis.The reporter replicon construct pSGR-JFH1/Luc was developed by rearrangement with pSGR-JFH1 (DDBJ/EMBL/GenBank accession number AB114136 ) that was constructed with the HCV genotype 2a clone JFH-1, which was isolated from a patient with fulminant hepatitis (14, 15). A DNA fragment encoding firefly luciferase was fused with the T7 promoter sequence and 5′ untranslated region of HCV clone JFH-1 by PCR and digested with EcoRI and PmeI (these restriction enzyme recognition sequences were artificially introduced in the primer site) and replaced the neomycin resistance gene of pSGR-JFH1 (Fig. 1). The construct of replication-deficient reporter replicon pSGR-JFH1/Luc-GND was also developed by introducing a point mutation at the GDD motif of RNA-dependent RNA polymerase to abolish this enzyme activity (Fig. 1).
Structures of the subgenomic and reporter replicon constructs pSGR-JFH1 (top), pSGR-JFH1/Luc (middle), and pSGR-JFH1/Luc-GND (bottom). Open reading frames (thick boxes) are flanked by untranslated regions (thin boxes). EcoRI, XbaI, MluI, and PmeI indicate positions of the respective restriction sites. In pSGR-JFH1/Luc and pSGR-JFH1/Luc-GND, the neomycin resistance gene was replaced by the luciferase reporter gene. GDD is the motif of HCV NS5B, RNA-dependent RNA polymerase. pSGR-JFH1/Luc-GND was constructed as a negative control by a point mutation altering GDD to GND. UTR, untranslated region; EMCV IRES, encephalomyocarditis virus internal ribosome entry site.
The XbaI-digested pSGR-JFH1/Luc and pSGR-JFH1/Luc-GND were purified and used as templates for RNA synthesis. The subgenomic reporter replicon RNAs were synthesized in vitro using the MEGAscript T7 kit (Ambion, Austin, TX). Synthesized RNA was treated with DNase I followed by acid phenol extraction to remove any remaining template DNA.
RNA transfection.The RNAs transcribed from pSGR-JFH1/Luc and pSGR-JFH1/Luc-GND were transfected into Huh7 cells by electroporation as follows. Trypsinized cells were washed with Opti-MEM I reduced-serum medium (Invitrogen, Carlsbad, CA), and 2.0 × 106 cells were resuspended in 400 μl of Cytomix buffer. Three micrograms of synthesized replicon RNA was mixed with the cell suspension. These cells were transferred to an electroporation cuvette (Precision Universal Cuvettes; Thermo Hybrid, Middlesex, United Kingdom) and pulsed at 260 V and 950 μF with the Gene Pulser II apparatus (Bio-Rad, Hercules, CA). Transfected cells were immediately transferred to 10 ml of culture medium and seeded into 12-well culture plates. Four hours after transfection, cells in a portion of the plates were harvested as a control for transfection efficacy, and a portion of the cells in the remaining plates received IFN or ribavirin in various doses. After administration of these agents, cells were harvested serially at 28 (day 1), 52 (day 2), and 76 (day 3) h after transfection.
MTS and luciferase assay.To adjust the number of viable cells, a 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt (MTS) assay was performed with the CellTiter 96 Aqueous One Solution cell proliferation assay (Promega), according to the manufacturer's instructions. Then, in the same well, luciferase activities were quantified with LUMAT LB9507 (EG&G Berthold, Bad Wildbad, Germany) and the luciferase assay system (Promega). Briefly, cells were lysed with 150 μl of cell culture lysis reagent (Promega), centrifuged, and mixed with luciferase assay reagent. Assays were performed at least in triplicate, and the results were expressed as luciferase activity relative to the luciferase activity at 4 h after transfection.
Northern blot analysis.Isolated RNA (4 μg) was separated in a 1% agarose gel containing formaldehyde, transferred to a positively charged nylon membrane (Hybond-N+; Amersham Pharmacia, Buckinghamshire, United Kingdom), and immobilized with a Stratalinker UV cross-linker (Stratagene, La Jolla, CA). Hybridization was performed with an [α-32P]dCTP-labeled DNA probe using Rapid-Hyb buffer (Amersham Pharmacia). The DNA probe was synthesized from the nonstructural (NS) 3-NS5b region of JFH-1 cDNA using the Megaprime DNA labeling system (Amersham Pharmacia). A DNA probe for β-actin was also synthesized as a control.
Reverse transcription-PCR and sequencing analysis.The cDNAs of the reporter replicon were synthesized from total RNA that was isolated from replicon RNA-transfected Huh7 cells with a primer in the 3′X region. A part of the reporter replicon cDNA fragment was amplified by nested PCR with DNA polymerase (TaKaRa LA Taq; Takara Bio Inc., Shiga, Japan) and primers as follows: 6764S-IH, 5′-AAGCCGTTTTTCCGGGATGAGGTCTCGTTC-3′, and 9382R-IH, 5′-GAGTAATGAGCGGGGTCGGGCGCGCGACAC-3′, for first-round PCR and 8717S-IH, 5′-GGTGATCCCCCCAGACCGGAATAT GACCTG-3′, and 9367R-IH, 5′-CACAGCGTGTCGCGCGCCCGACCCCGCTCA-3′, for second-round PCR. These primers were designed to amplify the approximately 650-bp cDNA fragment in the NS5b region. This fragment contains the amino acid position that was identified as being associated with resistance to ribavirin by Young et al. (31). Amplified fragments were cloned into the pCR-TOPO vector (Invitrogen Corp., Carlsbad, CA), and at least 22 isolated clones were sequenced with the ABI 3100 automatic DNA sequencer (Applied Biosystems Japan, Tokyo, Japan) to determine the population of reporter replicons in Huh7 cells.
Computer analysis.To calculate the genetic distances between isolated clones, sequences were aligned by use of Clustal W software (version 1.8; DDBJ), and the numbers of nucleotide substitutions per site were determined with MEGA software (version 2.1) (17).
Statistical analysis.The Student t test was used to analyze data. P values less than 0.05 were considered statistically significant.
RESULTS
Monitoring the reporter replicon replication with Northern blot analysis and luciferase assays.To determine the transient replication ability of the SGR-JFH1/Luc reporter replicon, Northern blot analysis was performed with total cellular RNA extracted from SGR-JFH1/Luc replicon RNA- and SGR-JFH1/Luc-GND replicon RNA-transfected cells. The correct size of reporter replicon RNA was detected only in SGR-JFH1/Luc replicon RNA-transfected cells (Fig. 2). Signal intensity peaked on day 2 and decreased on day 3. The luciferase activity in lysates of transfected cells was monitored at four time points: day 0 (4 h), day 1 (28 h), day 2 (52 h), and day 3 (76 h). The relative luciferase activity was calculated by adjusting the luciferase activity to be a multiple of the luciferase activity 4 h after transfection. In the case of the SGR-JFH1/Luc replicon, the relative luciferase activity increased exponentially over the time course of the experiment. However, in the case of the replication-deficient replicon, SGR-JFH1/Luc-GND, the relative luciferase activity showed no increases (Fig. 3).
Detection of reporter replicon RNA by Northern blot analysis. Total RNA from replicon RNA-transfected cells was analyzed by Northern blotting with DNA probes for the NS3-NS5b region of JFH-1 cDNA and β-actin genes. The 108 and 107 copies of in vitro-synthesized RNA were mixed with cellular RNA from untransfected Huh7 cells and used as positive controls. Arrows indicate target positions of reporter replicon RNA and β-actin. The Huh7 lanes contain cellular RNA from untransfected Huh7 cells as negative control.
Exponential replication of reporter replicon in Huh7 cells. Luciferase activity at days 1, 2, and 3 after RNA transfection is presented as multiples of the luciferase activity 4 h after transfection. Experiments were performed at least in triplicate. Data are presented as means and standard deviation bars.
Anti-HCV effects of IFN and ribavirin.To detect the anti-HCV effect of IFN, IFN was added to the culture medium at various doses 4 h after transfection. The luciferase activity was serially monitored every 24 h for 3 days. To adjust for the transfection efficiency, the relative luciferase activity was calculated as a multiple of the luciferase activity 4 h after transfection. To exclude the cytotoxic effects caused by the added agents and the variations in cell seeding, the number of viable cells in each well was normalized by MTS assay.
The administration of IFN at various doses resulted in a dose-dependent suppression of reporter replicon replication (Fig. 4A). When the same experiment was conducted with ribavirin, reporter replicon replication was also suppressed in a dose-dependent manner; but the suppression was substantially weaker than that mediated by IFN (Fig. 4B).
Dose-dependent suppression of reporter replicon replication by IFN and ribavirin. Transiently transfected Huh7 cells were treated with various doses of IFN and ribavirin after 4 h of transfection. Experiments were performed at least in triplicate. Data are presented as means and standard deviation bars. Rbv, ribavirin.
To assess the linear dose dependency of the antiviral effects of both agents, the percentages of relative luciferase activity at day 2 were plotted for each concentration. Both IFN and ribavirin showed linearly correlated dose dependency, and R2 was 0.987 and 0.976, respectively (Fig. 5). The 50% inhibitory concentration of IFN and ribavirin was 1.80 IU/ml and 3.70 μg/ml, respectively. Next, we compared the antiviral effects of these two agents in clinical concentrations. In a previous report, clinical concentrations of IFN and ribavirin in serum were found to be 40.2 to 116.0 IU/ml and 2.2 to 4.3 μg/ml, respectively (29). We found that the reporter replicon replication was suppressed to 0.09% by 100 IU/ml of IFN and to 53.74% by 3 μg/ml of ribavirin (Fig. 5). Thus, the antiviral effect of IFN was much greater than that of ribavirin in clinical concentrations.
Log dose-inhibition curve with IFN and ribavirin. Suppression of reporter replicon replication was calculated by comparison with control (without IFN and ribavirin) and presented as the percentage of replication at various concentrations of IFN and ribavirin. Reported clinical concentrations of IFN and ribavirin in serum are indicated. Rbv, ribavirin.
To elucidate the effect of IFN and ribavirin combined, these agents were administered simultaneously and the relative luciferase activity was measured 2 days after transfection. IFN was administered in three concentrations, 3, 10, and 30 IU/ml, and ribavirin was administered in two concentrations, 3 and 10 μg/ml. The addition of 3 μg/ml of ribavirin to various concentrations of IFN suppressed the relative luciferase activity by 54 to 66% (Table 1). Likewise, the addition of 10 μg/ml of ribavirin suppressed the relative luciferase activity by 10 to 20% (Table 1). The suppression of reporter replicon replication by IFN was presented as a linear regression (R2 = 0.995; Fig. 6). The additional administration of ribavirin in two concentrations, 3 and 10 μg/ml, shifted the dose-dependent inhibition curves to the left with conserved linear regression (R2 = 0.997 and 0.983, respectively; Fig. 6). The additional effect of ribavirin added to IFN was calculated to be 1.46- to 1.62-fold with 3 μg/ml of ribavirin and 3.94- to 6.14-fold with 10 μg/ml of ribavirin.
Log dose-inhibition curve with combination use of IFN and ribavirin. Suppression of reporter replicon replication after combination use of IFN and ribavirin is shown. Rbv, ribavirin.
Effect of IFN and ribavirin on HCV replication at day 2 after treatmenta
Effect of IFN and ribavirin on the mutation induction of the reporter replicon.The mutagen effect of ribavirin has been previously described (4, 5, 13, 28). To assess the mutagen effect of ribavirin in this system, sequences of replicating reporter replicon RNAs in Huh7 cells were determined after treatment with IFN (10 IU/ml), ribavirin (3 μg/ml), or both. These sequences were then compared with an untreated control. Total RNA was isolated 2 days after administration, and cDNA fragments were amplified by reverse transcription-PCR using primers covering the NS5B region. These amplified fragments were inserted into a cloning vector, and 22 to 24 clones in each treated well were sequenced. In the untreated control, the mutation induction rate was (2.9 ± 0.5) × 10−3/site in nucleotides and (4.2 ± 1.1) × 10−3 in amino acids. By administration of IFN, ribavirin, or both, the mutation induction rates were not statistically different from the untreated control (Table 2). To evaluate the complexity of quasispecies, mean genetic divergences between all possible isolated clone pairs were compared; there was no significant difference between the untreated control and treatment groups. Thus, the mutagen effect of ribavirin was not detected with this experimental system (Table 2). In addition, conserved amino acid mutations that indicate adaptation by use of IFN and ribavirin alone or in combination were not observed in the part of NS5b that was investigated (data not shown).
Numbers of mutations and genetic divergence among replicon populations
DISCUSSION
The development of an HCV replicon system has enabled the study of mechanisms for HCV replication and anti-HCV effects. Using this replicon system, the anti-HCV effects of IFN and ribavirin have been evaluated (13, 18, 28, 33). However, a number of these previous studies could not observe the anti-HCV effect of ribavirin in lower concentrations. In this report, we were able to identify the anti-HCV effect of ribavirin in clinical concentrations, because our replicon system has several advantages over the system used in previous reports. First, the HCV genotype 2a clone used in this system had potent replication activity in Huh7 cells (15). Previously reported replicons showed no exponential increment of replicon titer in the time course using normal Huh7 cells. By using a clone that efficiently replicates, the reporter replicon had an exponential increase in luciferase activity over time (Fig. 3). Additionally, this reporter replicon could replicate in a G418-free environment, although some of the previous replicons needed G418 selection during preparation or assessment for antiviral activities. G418 selection may alter the cellular characteristics of anti-HCV status or modify the sensitivity to anti-HCV agents. Thus, the robust replication of this reporter replicon may be essential to detect the anti-HCV effect with higher sensitivity and accuracy. Second, in this reporter system, we did not use the established replicon-hosting cells. Instead, we used the transient-transfection method. Replicon-hosting cells were selected to be sufficiently permissive for replicon replication (2). These cells, known as permissive cells, were expected to have disruptions in their antiviral systems, such as IFN signaling pathways. Thus, the study of established replicon-hosting cells to detect antiviral activities may lead to false conclusions. Besides, cell-derived ribavirin resistance was identified in Huh7 cells recently (7). Long-term cultivation with ribavirin may select the cells with these characteristics. To overcome this disadvantage, we used the transient-transfection assay for the reporter replicon system and used normal Huh7 cells. Third, we used a luciferase assay to quantify the reporter replicon replication in this study. Some of the previous data regarding anti-HCV effects with replicon systems were determined by colony-forming efficiency or replicon titer that was quantified by real-time detection PCR. Drawbacks of the assay for colony-forming efficiency are that it is affected by the condition of transfected cells and it consumes a lot of time. A disadvantage of real-time detection PCR is that it might be affected by degraded RNA fragments. We introduced the luciferase gene (instead of a neomycin resistance gene) into the replicon construct (Fig. 1); this allowed us to estimate the replicon RNA replication by measuring the luciferase activity. This reorganized replicon construct not only improved the accuracy and sensitivity of this system but also made replicon replication easy to measure. Finally, in this study, the luciferase activity data were adjusted by the luciferase activity 4 h after transfection and by the viable cell count (determined by MTS assay). Thus, we obtained more accurate and reliable data by avoiding the variations caused by transfection efficiency and cell seeding or the cytotoxic effects of anti-HCV agents and could detect the intracellular anti-HCV effects of IFN and ribavirin.
In some clinical studies, ribavirin monotherapy did not improve the clearance rate of HCV or reduce the viral load (8, 10, 12). This clinical observation may appear to conflict with our data that ribavirin has an antiviral effect against HCV. In our study, ribavirin in clinical concentrations certainly suppressed the replicon replication, but the suppression ratio was around 50%. Although ribavirin suppressed the HCV replication by half in vivo, it may be difficult to detect this suppression in the viral titer of circulating blood. However, combined with IFN, ribavirin enhanced the IFN effect by 1.46- to 1.62-fold; this boost may be crucial to clear the virus and may improve the efficacy of IFN therapy.
Several previous reports have described increases in mutation frequencies induced by ribavirin (4, 13, 18, 28). However, this mutagen effect of ribavirin was not detected in this study. Neither the observed mutation rate nor the mean genetic divergences between isolated clones treated with ribavirin alone or combined with IFN differed from those of the control or clones treated with IFN alone (Table 2). This discrepancy with previous data may be caused by differences in the ribavirin concentration and duration of administration. We used 3 μg/ml of ribavirin as a clinical concentration; this concentration is lower than that used in the previous reports. Moreover, the observation period in this study was only 48 h. This short duration was long enough to detect the antiviral effect of ribavirin but may be too short to detect the mutagen effect. Thus, the observed antiviral effect of ribavirin in this study does not appear to be the result of accumulated mutations. However, the mutagen effect of ribavirin administered for a longer time may not be negligible in clinical studies. Another possible explanation for the lack of ribavirin-induced mutations in this study is the characteristics of the HCV clone used in this system. The clone used in this system is JFH-1, which was isolated from a fulminant hepatitis patient (14). JFH-1 has exhibited efficient replication without adaptive mutations in various cell lines (6, 15, 16). According to ribavirin resistance-related mutations in NS5a reported by Pfeiffer et al., JFH-1 has Glu instead of Gly (amino acid 404 in NS5a) and Thr instead of Glu (amino acid 442 in NS5a), although the association of these mutations with ribavirin resistance is still obscure (25a). This clone may be more resistant against the mutagen effect of ribavirin than previously reported clones. Thus, it may be necessary to form the basis for resistance genotyping or phenotyping of patient HCV isolates using this new replicon system. The observed anti-HCV effect of ribavirin in this study cannot be attributed to the error-prone characteristics but may be the direct replication inhibition that has been reported for other viruses. Recently, a new model of HCV dynamics has been proposed (24). This model was based on the assumption that ribavirin reduces the infectious virion production and could explain the synergic effect of ribavirin and interferon. Unfortunately, our system cannot assess the infectivity of HCV, because it uses subgenomic replicons. Thus, a new system will be necessary to assess the HCV virion production and infectivity with the JFH-1 clone in order to verify this hypothesis.
In summary, we developed an accurate and sensitive replicon system using a luciferase reporter gene and JFH-1 HCV cDNA. The anti-hepatitis C virus effect of IFN and ribavirin was easily detected within their clinical concentrations by this replicon system. This system will provide a powerful tool for screening the newer antiviral compounds against HCV.
ACKNOWLEDGMENTS
This work was partially supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science and from the Ministry of Health, Labor and Welfare of Japan; by a grant from Toray Industries, Inc.; by the Program for Promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation (NIBIO); and by the Research on Health Sciences Focusing on Drug Innovation from the Japan Health Sciences Foundation.
FOOTNOTES
- Received 8 July 2005.
- Returned for modification 17 August 2005.
- Accepted 31 August 2005.
- Copyright © 2005 American Society for Microbiology
