Sensitivity of a Ribavirin Resistant Mutant of Hepatitis C Virus to Other Antiviral Drugs

Background While ribavirin mono-therapy regimens have minimal effect on patients with chronic hepatitis C virus (HCV) infections, they can be efficacious when combined with interferon. Clinical studies show that interferon-free combination therapies containing ribavirin are also efficacious, suggesting that an interferon-free therapy could be adopted in the near future. However, generation of drug resistant mutants and cross resistance to other drugs could impair the efficacy of the treatment. Therefore, understanding the mechanism of HCV resistance to ribavirin and cross resistance to other antiviral drugs could be of major importance. Methods We tested the ability of a J6/JFH1 derived HCV ribavirin resistant mutant to grow in tissue cultured Huh7D cells in the presence of the mutagen 5-Fluorouracil and the nucleoside analog 2′-C-Methylcytidine. Virus replication was assessed by detecting HCV antigens by immunofluorescence and by titrating virus present in the supernatants. Recovered viruses were amplified by RT-PCR and sequenced. Results The sensitivity of HCV-RR relative to parental J6/JFH1 to the tested drugs varied. HCV-RR was more resistant than J6/JFH1 to 5-Fluorouracil but was not more resistant than J6/JFH1 to 2′-C-Methylcytidine. Growth of HCV-RR in 5-Fluorouracil allowed the selection of an HCV-RR derived mutant resistant to 5-Fluorouracil (HCV-5FU). HCV-5FU grows to moderate levels in the presence of high concentrations of 5-Fluorouracil and to parental levels in the absence of the drug. Sequence of its genome shows that HCV-5FU accumulated multiple synonymous and non-synonymous mutations. Conclusions These results indicate that determinants of resistance to ribavirin could also confer resistance to other anti-HCV drugs, shedding light toward understanding the mechanism of action of ribavirin and highlighting the importance of combination drug selection for HCV treatment. The results also show that it is possible to select a 5-Fluorouracil HCV resistant mutant that replicates to levels similar to parental virus when grown in the absence of 5-Fluorouracil.


Introduction
Hepatitis C virus (HCV) is an enveloped, positive strand RNA virus member of the genus Hepacivirus of the Flaviviridae family. The HCV genome consists of an RNA molecule of approximately 9.6 kb in size containing a single open reading frame flanked by structured 59 and 39 un-translated regions. An internal ribosome entry site directs the translation of a polyprotein precursor that is cleaved by cellular and viral proteases into 10 proteins (core, E1, E2, p7, NS2, NS3, NS4a, NS4b, NS5a, and NS5B) (reviewed in reference [1]). Human HCV infection causes chronic liver disease, cirrhosis, and is associated with hepatocellular carcinoma [2]. It is estimated that 180 million people worldwide are infected with HCV [3] and given the chronic nature of the infection it is expected that the number of patients with hepatocellular carcinoma will increase in the coming years.
The standard therapy for the treatment of chronically HCV infected patients consists of a combination of pegylated interferon alpha and ribavirin [4]. Recently, two protease inhibitors were approved by the FDA and are being used in the clinic [5,6]. Given the side effects associated with injections of interferon, an interferon-free regimen for the treatment of HCV infections is highly desirable. Recent studies have shown that ribavirin in combination with other antiviral drugs, without interferon, can be efficacious [7,8], suggesting that an interferon-free therapy containing ribavirin could be adopted in the near future. However, generation of drug resistant mutants and cross resistance to different drugs could impair the efficacy of the treatment. In addition, the anti HCV mechanism of action of ribavirin is not completely elucidated. Several mechanisms of action of ribavirin against HCV were proposed including a direct effect against the HCV RNA dependent RNA polymerase (NS5b); induction of misincorporation of nucleotides leading to lethal mutagenesis; depletion of intracellular guanosine triphosphate pools; alteration in the cytokine balance from a Th2 profile to a Th1 profile; and up-regulation of genes involved in interferon signaling [9,10].
In order to study the cross-resistance of HCV to ribavirin and other antiviral drugs that could have a mechanism of action similar to that of ribavirin, we tested the ability of a J6/JFH1 [11] HCV derived ribavirin resistant mutant, HCV-RR [12], to grow in the presence of the pyrimidine analog 5-Fluorouracil and the nucleoside analog 29-C-Methylcytidine in Huh7D cells (a Huh7 cell derivative more permissive to HCV replication) [13]. 5-Fluorouracil is broadly used in the clinic to treat cancer [14] including HCV associated hepatocellular carcinoma [15]. 5-Fluorouracil displays mutagenic activity leading to viral extinction in different RNA viruses including LCMV, VSV, EMCV, and FMDV when grown in tissue cultured cells [16] and a similar lethal mutagenic effect has also been observed for ribavirin on several viruses including poliovirus [17], coxsackievirus B3 [18], FMDV [19], West Nile virus [20], GB virus B [21], and Hantaan virus [22]. It has been shown that ribavirin also has a mutagenic effect on HCV, increasing its mutation rate in cultured cells [23][24][25][26] and in vivo [25,27]. Deep sequencing has recently revealed that ribavirin exerts mutagenic activity in chronic HCV infected patients by facilitating G to A and C to U nucleotide transitions [28]. 29-C-Methylcytidine, the active component of the experimental anti-HCV pro-drug valopicitabine [29] [30], has been tested in HCV clinical trials and shown to be a potent HCV inhibitor in patients [31] [32][33][34] and chimpanzees [35]. 29-C-Methylcytidine inhibited HCV RNA replication in the replicon assay and inhibited the HCV RNA polymerase in vitro in cell-free biochemical assays [36]. It has also been shown that ribavirin antagonizes the in vitro anti-HCV activity of 29-C-Methylcytidine [37], suggesting an interaction between the two drug pathways.
In this study we show that an HCV mutant resistant to ribavirin is more resistant than parental J6/JFH1 to 5-Fluorouracil, but is not more resistant than parental J6/JFH1 to 29-C-Methylcytidine. These results indicate that ribavirin resistant viruses could have elevated resistance to other inhibitors, highlighting the importance of combination drug selection for HCV treatment, and shedding light toward the understanding of the mechanism of action of ribavirin and HCV resistance to this drug.
In addition, the growth of HCV-RR in 5-Fluorouracil allowed us to select an HCV mutant resistant to 5-Fluorouracil that can replicate in vitro to moderate levels in the presence of concentrations as high as 3 m/ml of 5-Fluorouracil and to parental levels in the absence of drug. The 5-Fluorouracil resistant virus accumulated multiple mutations distributed throughout the HCV genome.

Growth of HCV in the Presence of 5-Fluorouracil
In order to test the sensitivity of an HCV ribavirin resistant mutant to 5-Fluorouracil, parental J6/JFH1 [11] and J6/JFH1 derived HCV-RR2 [12] were grown in zero, 0.5, 1, 2, and 5 mg/ ml of 5-Fluorouracil by serially passaging the viruses every 7 days in naïve Huh7D cells [13] as described in the materials and methods section. Virus growth was assessed by immunfluorscence (not shown) and by titration of virus present in the supernatants from each passage ( Figure 1). Both viruses grew similarly in medium containing no 5-Fluorouracil. HCV-RR2 grew to higher titers than J6/JFH1 in medium containing 0.5, 1, and 2 mg/ml of 5-Fluorouracil. Neither virus survived a concentration of 5 mg/ml of 5-Fluorouracil. The experiment was repeated several times and similar results were obtained. The result indicates that HCV-RR2 is more resistant to 5-Fluorouracil than parental J6/JFH1.

Growth of HCV in the Presence of 29-C-Methylcytidine
In order to test sensitivity of an HCV ribavirin resistant mutant to 29-C-Methylcytidine, parental J6/JFH1 and HCV-RR2 were grown in zero, 0.31, 0.62, 1.25, and 2.5 mM 29-C-Methylcytidine as described above for 5-Fluorouracil. Virus growth was assessed by immunofluorescence (not shown) and by titration of virus present in the supernatants from each passage ( Figure 2). Both viruses grew similarly in medium containing 0, 0.31, 0.62, and 1.25 mM of 29-C-Methylcytidine. At a concentration of 2.5 mM of 29-C-Methylcytidine, parental J6/JFH1 grew to titers of more than 10 2 ffu/ml and 10 3 ffu/ml by passage 3 and 4, while HCV-RR2 was extinguished after passage 2. This result indicates that HCV-RR2 is more sensitive to 29-C-Methylcytidine than parental J6/ JFH1 after passaging in medium containing 2.5 mM of 29-C-Methylcytidine.
We note variations in the growth of the viruses among experiments. For example, J6/JFH1 and HCV-RR2 grew to a lesser extent in 1.25 mM 29-C-Methylcytidine in the experiment shown in Figure S1 when compared to the growth attained in the experiment shown in Figure 2. Similarly, J6/JFH1 and HCV-RR2 grew to a lesser extent in 2 mg/ml 5-Fluorouracil in the experiment shown in Figure 3 (see below) when compared to the growth attained in the experiment shown in Figure 1. We don't know at this time the nature of these variations.

Virus Recovered from 5-Fluorouracil Treated Cells is Resistant to 5-Fluorouracil
The enhanced growth of HCV-RR2 in 5-Fluorouracil ( Figure 1) prompted us to obtain a 5-Fluorouracil HCV resistant virus. To that end, we subjected HCV-RR2 to six passages of 7 days each in Huh7D cells treated with 2.5 mg/ml of 5-Fluorouracil followed by a) one passage of 7 days in medium containing no 5-Fluorouracil to obtain HCV-5FU-1 or b) two passages (of 3 and 7 days) in medium containing no 5-Fluorouracil to obtain HCV-5FU-2. These last passages in medium containing no drug were performed in order to increase the titer of the viruses. Viruses were titrated to 2.2610 3 ffu/ml and 2.5610 3 ffu/ml respectively. In order to test whether the HCV-5FU obtained viruses were truly resistant to 5-Fluorouracil, J6/JFH1, HCV-RR2, HCV-5FU-1 and HCV-5FU-2 were grown in 0, 1.5, 2 and 2.5 mg/ml 5-Fluorouracil by serially passaging the viruses every 7 days in naïve Huh7D cells as described for the experiment shown in Figure 1. Virus growth was assessed by immunofluorescence (not shown) and by titration of the supernatants from each passage ( Figure 3). All viruses grew in medium containing 0 or 1.5 mg/ml 5-Fluorouracil. HCV-RR2, HCV-5FU-1 and HCV-5FU-2 grew in 2 mg/ml 5-Fluorouracil, while J6/JFH1 was extinct by passage 4. Only HCV-5FU-1 and HCV-5FU-2 grew in 2.5 mg/ml 5-Fluorouracil. This result indicates that viruses recovered from 5-Fluorouracil treated cells were more resistant to 5-Fluorouracil than parental HCV-RR2.

Kinetics of Virus Growth in Different Concentrations of 29-C-Methylcytidine for 9
We also studied the kinetics of the growth of J6/JFH1 and HCV-RR2 in concentrations of 0, 0.625, 1.25, 2.5, 5, and 10 mM of 29-C-Methylcytidine for one week as described in the material and methods section. Similar growth curves were observed for J6/ FH1 and HCV-RR2 ( Figure 6). The result indicates that both viruses have a similar sensitivity to 29-C-Methylcytidine by passage one. At later passages, J6/JFH1 seems to have a slight advantage over HCV-RR2 (Figure 2 and Figure S1).
We analyzed each of the ten positions where nucleotide substitutions leading to amino-acid changes were found in HCV-5FU-1-P6 by sequencing viruses recovered after passage 1, 2, 3, and 4 of HCV-RR2 in 2.5 mg/ml 5-Fluorouracil, and by sequencing HCV-5FU-1 ( Table 2). All the mutations found in HCV-5FU-1-P6 were already found in HCV-5FU-1, with the exception of T2080T/A and Y2293Y/H. Some of the mutations found in HCV-5FU-1 (and HCV-5FU-1-P6) were present at early passages of HCV-RR2 in 5-Fluorouracil, as L133F in the core protein, M405R in E2, and P2268L and G2397R in NS5a. Mutation T2715A in NS5b appeared after passage 3.

Discussion
In this study we compared the growth of an HCV ribavirin resistant mutant (HCV-RR2) to the growth of its parental J6/ JFH1 virus in the presence of two antiviral drugs: the mutagenic pyrimidine 5-Fluorouracil and the nucleoside analog 29-C-Methylcytidine. We show that HCV resistant to ribavirin is more resistant to 5-Fluorouracil but is not more resistant to 29-C-Methylcytidine (Figures 1 and 2) than it parental virus J6/JFH1. By passaging HCV-RR2 in 5-Fluorouracil we selected HCV-RR2 resistant to 5-Fluorouracil (HCV-5FU). We confirmed the 5-Fluorouracil resistant phenotype of HCV-5FU viruses by infecting naïve cells and showing that they can grow even in a concentration of 3 mg/ml 5-Fluorouracil (Figures 3, 4, and 5). HCV-5FU resistant to 5-Fluorouracil acquired synonymous and non-synonymous mutations that were distributed all along the genome ( Table 1).
The mechanism of action of ribavirin against HCV in vitro and in vivo and the mechanism of resistance to ribavirin by HCV ribavirin resistant mutants has not been completely elucidated. The differential sensitivity observed for HCV-RR2 to 5-Fluoro-uracil and 29-C-Methylcytidine when compared to parental J6/ JFH1 virus indicates that mechanisms and/or viral RNA sequences implicated in the antiviral activity of these drugs could be involved in mechanism and/or viral RNA sequences of the antiviral activity of ribavirin. This is supported by the fact that 5-Fluorouracil is a pyrimidine analog and 29-C-Methylcytidine and ribavirin are both nucleoside analogs. Given the nature of these drugs, it is tempting to speculate that ribavirin acts on HCV at the RNA replication level, as it has been shown for 29-C-Methylcytidine [36]. We previously showed that an HCV ribavirin resistant mutant has mutations in different positions of its genome including the RNA dependent RNA polymerase [12], and others found that the RNA dependent RNA polymerase from HCV can use ribavirin triphosphate as a nucleotide substrate. Once ribavirin monophosphate has been incorporated in the nascent chain, it can reduce or even block RNA elongation [38,39]. As noted above, mutagenic activity of ribavirin on HCV has been observed in vivo and in vitro [23][24][25][26] [25,27] [28].
We found that HCV-RR2 was more resistant to the mutagen 5-Fluorouracil than its parental J6/JFH1. This indicates that determinants conferring resistance to ribavirin also confer resistance to 5-Fluorouracil and could confer resistance to other antiviral drugs. Cross resistance in HCV has been observed by other investigators. As an example, in the replicon system, an HCV mutant resistant to 29-C-Methylcytidine showed crossresistance to the nucleoside analog 29-C-Methyladenosine but not to the nucleoside analog 49-Azidocytidine (R1479), interferon a-2a, or to non-nucleoside HCV polymerase inhibitors [36]. These observations in cultured cells could be also relevant in vivo. Therefore, careful considerations should be made in the clinic when selecting combined or sequential drug treatments. Given the observed efficacy of the inclusion of ribavirin in interferon free regimens in the treatment of chronic HCV, cross resistance of HCV to ribavirin and to other antiviral drugs could be of major importance.
In FMDV, a single point mutation (M296I) confers resistance to ribavirin [40]. This mutant was as sensitive as wild type FMDV to 5-fluorouracil when administered in combination with guanidine hydrochloride, indicating that mutation M296I did not confer a significant cross-resistance to 5-fluorouracil [41]. These and our results indicate that different viruses evolve to generate phenotypicaly different mutants to escape the action of antiviral drugs. Of note, the concentrations of ribavirin (5 mM) and 5-Fluorouracil (200 or 500 mg/ml) for the treatment of BHK-21 cells used by Perales and colleagues are between 1 and 3 orders of magnitude higher than those used in our study.
We isolated an HCV mutant resistant to 5-Fluorouracil (HCV-5FU). To our knowledge, no 5-Fluorouracil resistant viruses, including HCV, have been previously reported. When compared to its parental HCV-RR2, HCV-5FU acquired 33 mutations in the coding region, 10 of which encoded amino-acid changes. Of note, all the 38 mutations acquired by HCV-RR2 when compared to its parental J6/JFH1 [12] were maintained in HCV-5FU even after several passages for selection (see above, results, and Table 1). We don't know which mutation/s confer resistance to 5-Fluorouracil. Six of the ten non-synonymous mutations acquired by HCV-5FU are located in non-structural proteins NS5a and NS5b (Table 1). Mutation A8483G encodes a change from threonine to alanine at amino-acid 273 of NS5b, which, according to its crystal structure, is located in the finger domain [42,43]. HCV-RR2 acquired mutation G7710A which encodes a mutation at amino-acid 15 of NS5b, also located in the finger domain [12]. Mutations in the finger domain have been identified in ribavirin resistant mutants of poliovirus [44,45]. Five other non-synonymous mutations observed in HCV-5FU are located in positions corresponding to domains I (one mutation), II (three mutations) and III (one mutation) of NS5a [46]. Although the role of NS5a has not been completely elucidated, it is known that NS5a is essential for HCV replication, interacts with other HCV and cellular proteins forming multiprotein replication complexes [47][48][49] and has been associated with sensitivity to interferon [50]. Therefore, NS5a mutations carried by HCV-5FU may alter its interaction with other viral proteins as NS5b, critical for viral replication.

Conclusion
In this report, we show that a ribavirin resistant mutant of HCV has differential sensitivity to other antiviral drugs when compared to its parental virus. This suggests that mutations that are responsible for HCV resistance to ribavirin can be involved in the sensitivity to other drugs, implying that common antiviral mechanisms and common mechanisms of defense could be used by and against different drugs. This could be clinically important for drug selection, since an interferon-free regimen containing ribavirin for the treatment of HCV infections seems currently plausible. We isolated a mutant resistant to the potent mutagen 5-Fluorouracil. This mutant when grown without 5-Fluorouracil can replicate to parental levels. Analysis of the mutations responsible for the 5-Fluorouracil resistance phenotype may aid in understanding the mechanism of action of 5-Fluorouracil and other antivirals against HCV.

Antibodies
Monoclonal antibody 6G7 directed to the HCV core protein was kindly provided by Henry H. Hsu and Harry B. Greenberg (Stanford University, Palo Alto Veterans Administration Medical Center, Palo Alto, CA) [51].
Infection of Huh7D Cells with J6/JFH1, HCV-RR, and HCV-5FU Viruses and Treatment with 5-Fluorouracil or 29-C-Methylcytidine Huh7D cells grown in 48-well plates were mock infected or infected with the indicated viruses at a moi of 0.01. At 5 to 7 hours post infection, medium was replaced with 500 ml of medium containing the indicated amount of 5-Fluorouracil (Sigma) or 29-C-Methylcytidine (USBiological). At 7 days post infection, 200 ml of the supernatants were used to inoculate naïve Huh7D cells that were seeded the day before, and at 5 to 7 hours post infection medium was replaced with 500 ml of medium containing the corresponding concentration of corresponding drug. The rest of the supernatants were stored at 270uC. This procedure was repeated for the indicated number of passages. HCV antigen was detected in the remaining monolayers by immunofluorescence and HCV titers were obtained from the supernatants as described below.
Growth of HCV in 5-Fluorouracil or 29-C-Methylcytidine for One Week

Titration of Viruses
Monolayers of Huh7D cells grown in 96 well plates were infected with 100 ml of 10-fold serial dilutions of the corresponding virus (in quadruplicates). At three days post infection, viral antigen was detected by immunofluorescence as described below. Foci were counted and titers were expressed as the mean number of foci of each of the four replicates +/2 the standard deviation.

Sequencing of Ribavirin Resistant Viruses
Viral RNA was extracted from virus stocks using Trizol reagent as recommended by the manufacturer (Invitrogen). cDNA was synthesized using SuperScript III reverse transcriptase and random primers (Invitrogen). PCR amplification of the HCV genome was performed using the Expand High Fidelity PCR system (Roche) as recommended by the manufacturer and the following sets of primers: 2a40+ (59-atgaatcactcccctgtgag-39) and 2a1260-(59-gagcaattgcagtcttggac-39), 2a1101+ (59-TCACGCAGGGCTTGCGGACG-39) and 2a2690-(59-CCTTGATGTACCAAGCAGCC-39), 2a2431+ (59-CCAAAACATCGTGGACGTAC-39) and 2a3980-(59-AAGTGGGAGACCTTGTAACA-39), 2a3721+ (59-  PCR products were run in agarose gels and purified using geneelute agarose gel columns (Sigma) or the QIAquick PCR purification Kit (Qiagen) and sequenced using the BigDye terminator v3.1 cycle-sequencing kit (Applied Biosystems) and the 31306l Genetic analyzer (Applied Biosystems). In addition to the oligos used for PCR, we used the following oligos for  Figure S1 Growth of HCV in the presence of different drugs. J6/JFH1 and HCV-RR2 viruses were serially passaged in Huh7D cells in medium containing the indicated concentration of the indicated drugs. At each passage HCV titers were obtained as described in the text. Titers are expressed as the mean number of foci of each of four replicates. Error bars represent the standard deviation. (TIFF) Table 2. Nucleotide and deduced amino acid changes in early passages of HCV-RR2 in 5-Fluorouracil at positions found mutated in HCV-5FU-1-P6.