About AVT
Browse Articles
Customer Services

Original article

In vitro selection of resistance to sofosbuvir in HCV replicons of genotype-1 to -6

Simin Xu1,, Brian Doehle1,*,, Sonal Rajyaguru1, Bin Han1, Ona Barauskas1, Joy Feng1, Jason Perry1, Hadas Dvory-Sobol1, Evguenia S Svarovskaia1, Michael D Miller1, Hongmei Mo1

1Gilead Sciences, Foster City, CA, USA

*Corresponding author e-mail: Brian.Doehle@gilead.com

Authors contributed equally to this work

Citation: Antiviral Therapy 2017; 22:587-597
doi: 10.3851/IMP3149

Date accepted: 03 February 2017
Date published online: 01 March 2017

Copyright (c) 2017 International Medical Press, all rights reserved.


Background: Sofosbuvir is a nucleoside analogue inhibitor of the HCV NS5B polymerase approved for treatment of HCV-infected patients in combination with ribavirin or with other antivirals. It has activity against all genotypes of HCV. Resistance to sofosbuvir in genotype-1 and -2 HCV is conferred by the S282T substitution in NS5B.

Methods: To begin to define the correlates of resistance to sofosbuvir in other genotypes, we performed selection experiments in cell culture using cell lines containing subgenomic replicons derived from genotypes-1b, -2a, -3a and -4a, or chimeric replicons in a genotype-1b background but encoding genotype-2b, -5a and -6a NS5B polymerase.

Results: In every case, S282T was selected following passage in the presence of increasing concentrations of sofosbuvir for 10 to 15 weeks. When introduced as a site-directed mutant, S282T conferred reductions in sofosbuvir susceptibility of between 2.4 and 19.4-fold. Other substitutions observed during the selections had relatively less impact on susceptibility, such as N237S in genotype-6a (2.5-fold). Replication capacity was affected by the introduction of S282T in all genotypes to variable extents (3.2% to 22% of wild type).

Conclusions: These results confirm that S282T is the primary sofosbuvir resistance-associated substitution and that replication capacity is reduced when it is present in all genotypes of HCV.


HCV infects approximately 185 million individuals globally and is associated with significant morbidity and mortality [1]. HCV is classified into 7 confirmed genotypes and is further divided into 67 subtypes [2], which vary in prevalence based on geographic and/or epidemiological factors [3,4]. HCV genotype can influence the outcome of some antiviral treatments [5,6].

The pan-genotypic HCV NS5B nucleotide inhibitor (NI) sofosbuvir is a chain-terminator that inhibits replication of HCV in vitro and in vivo [79]. Sofosbuvir has displayed high efficacy in patients infected with HCV and has been studied extensively in pre-clinical studies and Phase I, II and III clinical trials [1013]. Sofosbuvir was approved in 2013 for treatment of infection in combination with ribavirin (RBV) for treatment of genotype-2 and -3 HCV, and with pegylated interferon (PEG-IFN) and RBV for genotype-1 and -4 to -6 HCV. IFN-free treatment is an ongoing objective, and by combining 2 or more DAAs targeting different HCV proteins, high rates of sustained virological response can be achieved, for example, the recently approved combination of the NS5A inhibitor ledipasvir with sofosbuvir for genotype-1 [14,15]. There is a need for advances in the treatment of genotype-2 to -6 infection which remains a significant worldwide disease burden.

Treatment with DAAs has the potential for selection of drug-resistant HCV [16]. The S282T substitution in NS5B was first described as the major resistance-associated substitution for other investigational NIs [1720]. Selection experiments using cell lines harbouring subgenomic replicons of genotypes-1 (both 1a and 1b subtypes) and -2a grown in the presence of elevated concentrations of sofosbuvir (1 to 3 µM) identified the S282T as the primary resistance substitution [21]. In genotype-2a, T179A, M289L and I293L were also selected and found to contribute to reduce sofosbuvir susceptibility in combination with S282T. In addition, the combination of S96T and N142T has been observed following in vitro selection with R1479 [22], and a combination of L159F and L320F was observed in one patient who had a partial response during treatment with mericitabine [23] and one patient with viral breakthrough during treatment with mericitabine, danoprevir and RBV [24]. As previously described, in the Phase II ELECTRON clinical trial, S282T was detected in one patient infected with genotype-2b HCV who was treated with sofosbuvir as monotherapy but did not achieve SVR [13].

Relatively little is known about correlates of sofosbuvir resistance in HCV of genotypes-3, -4, -5 or -6. To broaden the knowledge base about sofosbuvir resistance, we performed in vitro selection experiments using a panel of sub-genomic HCV replicons that include NS5B sequences for these genotypes.



Sofosbuvir and the NS5A inhibitors ledipasvir and velpatisvir were synthesized by Gilead Sciences, Inc. (Foster City, CA, USA). RBV was purchased from Sigma–Aldrich (Atlanta, GA, USA).

Cell culture

All replicon cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) with GlutaMAX™ (GIBCO-BRL; Life Technologies Ltd, Madison, WI, USA), supplemented with 10% fetal bovine serum (HyClone, Chicago, IL, USA), 100 units/ml penicillin, 100 μg/ml streptomycin, (Gibco Life Technologies, Chicago, IL, USA), 0.1 mM non-essential amino acids (Gibco Life Technologies), and 0.5 mg/ml of Geneticin (Invitrogen, Madison, WI, USA). Cells were passaged every 5 to 7 days before reaching 100% confluence. The permissive cell line 1C, derived from Huh7 Lunet [25], were cultured under the same conditions but without G418.

HCV subgenomic replicons and cell lines

Generation of subgenomic replicon constructs, and cell lines stably expressing them, were described previously for HCV genotype-1a (H77) [25,26], -1b (Con1) [25,27], -2a (JFH-1) [28], -3a (S52) [29] and -4a (ED43) [30], and chimeric genotype-1b replicons carrying consensus NS5B coding sequences from genotype-2b, -5a and -6a [31]. All these replicon constructs carry a Renilla luciferase reporter gene. Chimeric replicons were used for genotype-2b, -5a and -6a because subgenomic replicons for these genotypes were not available at the initiation of these studies.

Resistance selection

In vitro resistance selection was conducted by passaging cells in the presence of 0.5 mg/ml G418 and gradually increasing concentrations of sofosbuvir. The G418 concentration was doubled to 1 mg/ml after day 63 and day 66 of selection for genotype-3a and -4a replicon cells, respectively. All other replicon cells were cultured in the complete medium containing 0.02% DMSO and 0.5 mg/ml G418, and were passaged in parallel with DMSO control treatment. Cells were split at 90% to 95% confluence. Cell culture medium containing sofosbuvir and G418 was refreshed every 2–3 days. The starting concentration of sofosbuvir treatment was 200 nM for all replicon cells, which is 2- to 3-fold higher than the 50% effective concentration (EC50) for genotype-1b, -2a, -3a and -4a replicons and 10- to 20-fold higher than the EC50 for chimeric genotype-2b, -5a and -6a replicons. Cells from each passage were collected and stored for phenotypic and genotypic analysis. Emergence of mutations was monitored by deep sequencing of NS5B (see Deep sequencing analysis for HCV NS5B). After the sofosbuvir concentration reached the micromolar range, additional passages were performed maintaining the highest concentration for some genotypes.

Deep sequencing analysis for HCV NS5B

Deep sequencing analysis was performed by Wuxi Apptec (Shanghai, China) using HCV RNA extracted from cell pellets, using RNeasy Mini Kit according to the manufacturer’s instructions (Qiagen, Carol Stream, IL, USA). cDNA was synthesized using High-Capacity cDNA Reverse Transcription Kit (Life Technologies, Chicago, IL, USA) according to the manufacturer’s instructions. The PCR products were processed for deep sequencing using either Life Technologies Ion Torrent PGM or Illumina Miseq (Genome Center, WuXi AppTec, Shanghai Shi, China). In validation experiments, both methods were found to provide equivalent results using a 1% mutation detection threshold. The data QC criteria in this project is data yield ≥6.0 Mbases for 3,000× coverage per sample. Deep sequencing data analysis was conducted by Gilead using in-house developed software.

Phenotypic analysis of sofosbuvir-selected replicon cell lines

Cell lines at various stages of selection were subsequently thawed and cultured for 3 days in the presence of both 0.5 mg/ml G418 and 200 nM sofosbuvir (except the passage containing >95% S282T, which was cultured in the selection concentration of sofosbuvir). Susceptibility to NS5A inhibitors GS-5885, GS-5816, sofosbuvir and RBV was determined on the selected passages of cells containing the genotype-1b, -2a, -3a and -4a subgenomic replicons and chimeric genotype-1b replicons carrying genotype-2b, -5a or -6a NS5B sequences. 4,000 replicon cells in a total volume of 100 µl of DMEM culture media (without G418) were added to 96-well plates. 3-fold serial dilutions of HCV inhibitors were prepared in 100% DMSO. Compounds were then diluted 1:100 in DMEM culture media. 100 μl of the compound/media mixture was then added to each respective well of the replicon cell-containing plates. Assay plates were cultured for 3 days at 37°C. Luciferase assays were performed using the Promega Renilla Luciferase Assay kit according to the manufacturer’s instructions (Promega, Madison, WI, USA). Data were collected from two independent experiments except where otherwise indicated. EC50 values were calculated using non-linear regression analysis in Prism 6 (GraphPad, La Jolla, CA, USA).

Transient transfection susceptibility assays

Substitutions in NS5B at conserved sites that were enriched during the selection process were selected for phenotypic susceptibility to a panel of anti-HCV compounds. Replicon variants containing site-directed mutations were generated in the wild-type Renilla luciferase reporter replicon from the corresponding genotype using the QuikChange® II Site-Directed Mutagenesis (SDM) Kit from Agilent (Santa Clara, CA, USA). Sequencing analysis was used to confirm the SDM.

Transient transfection susceptibility assays were performed to determine the susceptibility of SDMs to sofosbuvir and RBV. Replicon plasmids were linearized for run-off T7 transcription to generate replicon RNA transcripts using the RiboMax kit from Promega. 10 μg of in-vitro-transcribed RNA was transfected into 1C cells by electroporation as previously described [32]. Cells were electroporated, incubated at room temperature for 10 min, and resuspended. Threefold serial dilutions of HCV inhibitors were prepared in culture medium and added to each respective well of the replicon cell-containing plate, in duplicate. Assay plates were then cultured for 3 days at 37°C. Luciferase activity was measured at 4 and 72 h post-transfection using the Promega Renilla Luciferase Assay kit according to the manufacturer’s instructions (Promega, Madison, WI, USA). Replication capacity (RC) of a given SDM was defined as the luciferase activity of the SDM at 72 h in the absence of drug divided by that at 4 h as a percentage of the same ratio of the parental wild-type replicon. EC50 values were calculated using non-linear regression analysis in Prism.

Cloning, expression and purification of NS5B polymerase

The sequences coding for NS5B polymerase (wild-type or S282T SDM) from genotype-1b (Con-1 strain), -2a (JFH-1 strain), -3a (S52 strain) and -4a (ED43 strain) were PCR amplified from a plasmid encoding the I389luc-ubi-neo/NS3-3′/ET replicon and pJFH1, respectively. The 3′-PCR primers were designed to result in a construct excluding the C-terminal 21 amino acids of full-length NS5B and the addition of a C-terminal hexahistidine tag. The resulting PCR fragments were cloned into the pET21a, pET30a or pET24 protein expression vectors (Invitrogen), respectively.

BL21DE3 bacteria were transformed with the NS5B polymerase expression vectors and inoculated into a fermentation vessel (Sartorius BBI System Inc., Bethlehem, PA, USA), containing 18 l of fresh 2YT medium supplemented with 100 µg/ml ampicillin at 37°C. When cell densities reached an optical density (OD600) of 1, induction was initiated by the addition of 0.5 mM IPTG at 25°C (final concentration). After a 3-h induction, cells were harvested by centrifugation and stored as frozen pellets at -80°C prior to protein purification. Cell pellets were thawed and resuspended at 10 ml/g cells in lysis buffer containing 50 mM Tris-HCl (pH 7.6), 300 mM NaCl, 10% glycerol, 3 mM β-mercaptoethanol (BME) and 2 mM PMSF. The cell suspension was homogenized, filtered through cheesecloth and passed three times through a microfluidizer (Microfluidics, model 110Y) at 18,000 psi. The resulting lysate were centrifuged at 15,500 rpm for 45 min and the lysate supernatant was loaded onto a Ni-HiTrap column (GE healthcare, Piscataway, NJ, USA) pre-equilibrated with five column volumes of the lysis buffer. Proteins were eluted with a buffer containing 50 mM Tris-HCl (pH 7.6), 300 mM NaCl, 10% glycerol, and a 25–500 mM imidazole-HCl gradient. Fractions containing NS5B protein were collected, pooled and dialyzed overnight at 4°C into 50 mM Tris-HCl (pH 7.0), 50 mM NaCl, 10% glycerol, 2 mM DTT. The sample was then diluted with SP Buffer A (1:4) containing 50 mM Tris-HCl (pH 7.0), 10% glycerol and 2 mM DTT, and loaded onto an SP cation exchange column (GE healthcare) equilibrated in SP Buffer A. Recombinant NS5B protein was eluted with 50 mM Tris-HCl (pH 7.0), 2 mM DTT, 10% glycerol and a 0 to 1 M NaCl gradient. The enzyme was stored at -80°C in 50 mM Tris-HCl (pH 7.0), 600 mM NaCl, 10% glycerol and 2 mM DTT.

NS5B inhibition assays

Threefold serial dilutions of NTP analogues (starting at 20 or 200 μM) were prepared in duplicate in 96-well plates. A reaction mixture containing 50 mM Tris-HCl (pH 7.5), 10 mM KCl, 1 mM DTT, 0.1 mg/ml BSA, 0.2 unit/µl RNasin Plus RNase Inhibitor (Promega, Madison, WI, USA), 4 ng/µl sshRNA [33], 5 mM MgCl2 and 70–150 nM HCV NS5B was pre-incubated with NTP analogues for 5 min at room temperature. The reaction was initiated by the addition of a mixture containing 2.5 µM ATP, 2.5 µM CTP, 2.5 µM UTP, 1.25 µM genotypeP, 0.06 µCi/µl of α-33 P-genotypeP (3,000 mCi/mol). Reactions were allowed to proceed for 90 min at 30°C. 10 µl of the reaction mixture was spotted on DE81 anion exchange paper (Whatman, Maidstone, UK), which was then washed with 3× Na2HPO4 (125 mM, pH 9), 1× water and 1× ethanol. The filter paper was air-dried and exposed to the phosphor imager screen and the amount of synthesized RNA was quantified using Typhoon Trio Imager and Image Quant TL Software (GE Healthcare, Piscataway, NJ, USA).


In vitro resistance selection

Sofosbuvir resistance selections were performed using stably transfected cell lines harbouring genotype-1b, -2a, -3a and -4a subgenomic replicons, and chimeric 1b replicon encoding genotype-2b, -5a and -6a NS5B polymerase. All replicon cells were passaged in the presence of increasing concentrations of sofosbuvir and at concentration of 0.5–1 mg/ml G418 (Figure 1). In genotype-1b, -2a, -3a and -4a replicons, S282T was observed when the replicon cells were treated with sofosbuvir concentrations of 10 to 30× the wild-type EC50 (Table 1). Detection of the S282T substitution was associated with a decrease in susceptibility to sofosbuvir, with the presence of >98% S282T associated with 10- to 24-fold decreased susceptibility to sofosbuvir for all four genotypes. Although S282T was the only substitution selected in genotype-3a, several substitutions in addition to S282T were observed in genotype-1b, -2a and genotype-4a replicons. The S549N substitution was detected (>15%) in genotype-1b and -2a at passage levels where the percentage of S282T was increasing; in genotype-2a, it was not detected in later passages (Additional file 1). Multiple substitutions including V67A, E237G, R304K, A324V, K544N and C575G were observed in the genotype-4a replicon during the selection (Additional file 2). In nearly all cases, these additional substitutions were present as mixtures in early passages, with the K544N substitution representing the majority virus population (>85%; Table 1) at one passage prior to S282T reaching >99%. This substitution was also observed at >40% of the viral population in the DMSO control passaged cell (Additional file 2). Substitutions that were observed to be present at high proportions in combination with S282T are likely to be linked with S282T on the same genome. Additional phenotypic analysis from clonal phenotyping was undertaken, however the complex mixtures that included S282T failed to replicate (data not shown). Further investigation is needed to understand the relative contribution of these minor variants.

Figure 1.
Figure 1. Selection of sofosbuvir-resistant replicon cells

Cell lines harbouring subgenomic replicons of (A) the indicated genotype or (B) chimeric genotype-1b replicons containing NS5B from the indicated genotype were passaged in increasing concentrations of sofosbuvir for varying lengths of time as indicated.

Table 1.  Genotype and phenotype of SOF-selected HCV genotype-1b, -2a, -3a and -4a replicon cells and chimeric genotype-2b, -5a and -6a NS5B replicons cells
Table 1. Genotype and phenotype of SOF-selected HCV genotype-1b, -2a, -3a and -4a replicon cells and chimeric genotype-2b, -5a and -6a NS5B replicons cells

a Concentrations of sofosbuvir (SOF) at which replicon cells were cultured in and when the substitution(s) were identified. b Changes compared to wild-type (parental) replicon sequence. c Fold change (FC) in comparison to wild-type (parental replicon sequence). d Data were generated in n=2 independent experiments, except these values which are n=1. e Chimeric genotype-1b replicons containing the NS5B coding sequences from the indicated genotype. EC50, 50% effective concentration; LDV, ledipasvir; ND, not determined; RBV, ribavirin.

Cell lines harbouring chimeric genotype-1b replicons encoding the NS5B region from genotype-2b, -5a or -6a were similarly cultured in the presence of increasing concentrations of sofosbuvir. Again, the appearance of S282T coincided with decreases in sofosbuvir EC50 compared to the corresponding wild-type replicons. Deep sequencing analysis revealed that the levels of S282T reached 99% by day 81 in genotype-2b and by day 89 for genotype-6a chimeric replicon cells. In genotype-5a NS5B chimeric replicon cells, S282T levels reached 98% by day 70. Detection of the S282T substitution at levels of 98% or higher was associated with decreases in sofosbuvir susceptibility of between 27- and 37-fold. Additional substitutions were observed at levels >15% in genotype-2b and -6a chimeric replicons, including N237S, E375D, R498K and T580I (Table 1). Only E375D reached >85% of the viral population in the genotype-6a chimeric replicon, and resolved to a dominant mutant from a lower level mixture observed prior to the detection of S282T. This substitution was also observed in the DMSO control passaged cells as a dominant variant (data not shown). As has been reported in previous studies [21], the M289L substitution was observed in conjunction with S282T in the genotype-2a replicon cells (Table 1). It was also detected at low levels in genotype-5a (1.3%) and -6a (1.3%; not shown in Table 1 due to frequency <15%).

The susceptibility of representative passages of sofosbuvir-selected replicon cells to the NS5A inhibitors ledipasvir and velpatisvir, and RBV, was also tested. Passages of selected replicon cells with <1% S282T detected and 98–99% S282T detected were chosen, as well as the parental cells not exposed to sofosbuvir (Table 1). Susceptibility to LDV, velpatisvir and RBV remained largely unchanged regardless of the presence of S282T (FC EC50 0.4- to 1.9-fold). These data indicate that the NS5B S282T substitution does not confer cross-resistance to NS5A inhibitors or RBV.

Phenotypic analysis of individual sofosbuvir-selected amino acid substitutions

To assess the impact of the various substitutions detected by deep sequencing of sofosbuvir-selected replicon cells on sofosbuvir susceptibility and RC, they were introduced individually via SDM into the wild-type replicon of the genotype in which they were observed, and tested using a transient transfection phenotypic assay. Across all genotypes, replicons containing S282T consistently exhibited low RC (3.2 to 22%) compared to wild type and decreased sofosbuvir susceptibility (2.4- to 19.4-fold; Table 2). Other mutants showed typically no strong effect on RC and only small or no changes in sofosbuvir susceptibility (0.4- to 2.5-fold; Table 2). In the genotype-2a replicon, the M289L substitution, when present in isolation, had a minor change in sofosbuvir susceptibility (FC EC50 2.0-fold), and a small increase when added to S282T (FC EC50 3.4-fold) compared to S282T alone (2.4-fold). M289L appeared to partially restore the reduced RC imparted by S282T (22% versus 11%). To explore if this held across genotypes, M289L and the S282T+M289L double mutant were also made in genotype-5a replicons, despite M289L having only been observed at low levels during selection. Similar to genotype-2a, M289L alone showed a minor 2.3-fold change in sofosbuvir susceptibility in genotype-5a, however, the combination of the two mutants lowered the change in sofosbuvir susceptibility (10.9- from 19.4-fold) while still showing severely impaired replication. R498K, observed in genotype-2b passages during sofosbuvir selection, did not exhibit any phenotypic shift when present as single substitution (FC EC50 1.3-fold). When present as a double mutant (R498K+S282T), the fold shift in sofosbuvir was similar to S282T alone (FC EC50 14-fold versus 16.2-fold), indicating that this substitution does not confer a significant reduction in susceptibility to sofosbuvir. In genotype-6a chimeric replicons the N237S substitution displayed a small increase in sofosbuvir FC (2.5-fold) and reduced RC (6.9% compared to genotype-6a WT). For all other individual substitutions tested, no significant change in sofosbuvir susceptibility was observed (FC EC50 0.4 to 1.6-fold) indicating that these substitutions do not alter sofosbuvir activity in vitro.

Table 2.  Phenotypic analyses of NS5B substitutions observed during SOF selection
Table 2. Phenotypic analyses of NS5B substitutions observed during SOF selection

a Mean ±sd from duplicate experiments. b Not observed in selections but major polymorph in genotype-4 HCV. Wild-type (WT) replicon sofosbuvir (SOF) 50% effective concentrations (EC50) were as follows: genotype-1b: 110 nM; genotype-2a: 50 nM; genotype-2b: 15 nM; genotype-3a: 50 nM; genotype-4a: 40 nM; genotype-5a: 15 nM; genotype-6a: 14 nM. Bold type indicated values over twofold. FC, fold change.

Biochemical analysis of recombinant NS5B encoding the S282T substitution

As the NS5B S282T substitution was found to be the primary sofosbuvir resistance substitution in vitro, the effect of S282T on the activity of the active triphosphate form of sofosbuvir, GS-461203, was assessed in vitro using a biochemical assay. 50% inhibitory concentration (IC50) values derived from recombinant NS5B polymerase from genotypes-1b, -2a, -3a and -4a encoding S282T were compared to respective wild-type controls. The presence of S282T resulted in significant changes in GS-461203 activity with fold changes in IC50 ranging from 8.4- to 25-fold (Table 3). These data further confirm that S282T can confer reduced susceptibility to sofosbuvir in vitro and are consistent with the replicon data.

Table 3.  Effect of the S282T on GS-461203 susceptibility of recombinant NS5B
Table 3. Effect of the S282T on GS-461203 susceptibility of recombinant NS5B

a Mean ±sd values from at least three independent experiments. IC50, 50% inhibitory concentration.

Three-dimensional mapping of sofosbuvir-selected substitutions

The locations of the mutations observed during in vitro selection with sofosbuvir, mapped to the 3-D structure of the genotype-1b NS5B polymerase, are shown in Figure 2. Residues 51, 282, 286, 289 are the closest to the active site (Figure 2A), however, S282T is the only residue in clear contact with the substrate. NS5B residue 51 may be in contact with the gamma-phosphate while 286 and 289 form a cluster of residues on the back side of the active site along with residue 293 which was not modified in any of the sofosbuvir selections. The side chains of these residues do not come in direct contact with the substrate, and may be in contact with the template.

Figure 2.
Figure 2. Structural mapping of NS5B substitutions

The locations of selected mutations observed during in vitro selection with sofosbuvir (SOF), were mapped to the 3-D structure of the genotype-1b NS5B polymerase. (A) Residues within the active site, (B) residues residing within the palm domain and (C) residues found within the thumb domain. C-terminal residues 544, 549, 575 and 580 are not shown as they fall outside of the modelled structure.

Sofosbuvir-selected substitutions at other residues were more distal to the active site (Figures 2B and 2C). Residue 237 is in the fingers region, 421 is in the thumb and 445 (not figured) is on the beta-loop which plays a role in initiation and probably rests above the primer/template during elongation. Residue 549 is in the C-terminal and may also play a role in initiation. Since these models do not extend beyond residue 543, the positions of S549 genotype-1b and -2a and T580 in genotype-6a are only approximate.


Previous resistance selection studies in HCV genotype-1a, -1b and -2a reported S282T to be the primary substitution selected by and conferring reduced susceptibility to sofosbuvir [21] and the biochemical mechanism of the reduction in sofosbuvir sensitivity of this mutation has started to emerge [34]. The in vitro selection studies described here confirm these observations and further demonstrate that S282T is also the primary substitution selected by sofosbuvir in HCV replicons of genotype-2b, -3a, -4a, -5a and -6a. Consistently, S282T has been observed at virological failure in 2 patients (1 genotype-1a and 1 genotype-2b) treated with sofosbuvir in Phase II clinical trials [13,35].

S282T was selected following passage in the presence of increasing concentrations of sofosbuvir in every genotype tested, and conferred reductions in sofosbuvir susceptibility of between 2.4 and 19.4-fold along with impaired RC in all cases. The additional substitutions observed by deep sequencing were not individually associated with significant reductions in susceptibility to sofosbuvir, with the possible exception of N237S in genotype-6a (2.5-fold change in EC50) and M289L in genotype-2a and -5a (2.0- and 2.3-fold change in EC50, respectively). Furthermore, none of the amino acid substitutions selected by sofosbuvir in stable replicon cell lines, including S282T, conferred cross-resistance to NS5A inhibitors or RBV in any genotype.

The selection procedure used in this study was carefully optimized to allow for selection of sofosbuvir-resistant replicons. Compared to procedures typically used for other direct-acting antivirals, a higher number of cells and slower increases in the drug concentration were used to maximize the potential for resistance development. When selection experiments were conducted using fewer cells or more rapid escalations in drug concentration, no resistant colonies were detected. These observations are consistent with a low prevalence of pre-existing S282T variants and the relatively small increases in EC50 that this substitution imparts. A low prevalence of S282T in the absence of drug selection pressure is consistent with the reduced RC and high genetic barrier associated with S282T. Since all possible codons for serine (TCN or AGY) are only one nucleotide change away from encoding threonine (ACN), and since serine is the wild-type amino acid in all genotypes, it is unlikely that any differences in genetic barrier to S282T exist between genotypes. This would be the case even if the propensity for a G to C mutation was different from that of a T to A mutation, because genotype-1 to -6 HCV nearly all contain AGY at position 282. If these properties also apply to HCV in vivo, it may help explain why detection of S282T in virus from patients treated with sofosbuvir but not achieving sustained virological response is so rare.

A limitation of this study, also applicable to nearly all in vitro selection experiments using laboratory strains of a virus, is that it is not possible to mimic the in vivo viral sequence diversity within and between patients. Therefore, additional pathways to resistance to sofosbuvir might exist, as well as additional drug-selected mutations that impact RC of sofosbuvir-resistant viruses, may be discovered as it is used more widely in the clinic. Notably, the L159F and V321A mutations in NS5B have been reported to emerge in genotype-3 HCV-infected, sofosbuvir-exposed patients, but were not observed in vitro, and are not associated with significant reductions in sofosbuvir susceptibility [36].

The 289L substitution was observed in sofosbuvir-resistant replicon cell lines in three different genotype backgrounds (2a, 5a and 6a), but not in others tested so far. The propensity for this substitution to be selected may be affected by the nature of the wild-type amino acid at this position, which is methionine in genotype-2 and -5, but cysteine in genotype-1, and phenylalanine in genotype-3 and -4. In genotype-6, methionine is predominant in subtypes 6a, g, h, m, u and v, while leucine is the naturally occurring variant in other subtypes. In the combined analysis of more than 1,600 patients treated with sofosbuvir in the Phase II and III clinical trials, there was no significant association between having a 289I or L substitution at baseline with treatment outcome, and no subject that relapsed after treatment in these studies was observed to have a 289 substitution emerge [36]. Similarly, substitutions at position 237 were observed in both genotype-4a and genotype-6a replicons. Position 237 naturally encodes for a glutamic acid in all genotypes tested here, with the exception of genotype-6a, where the consensus amino acid is an asparagine. In genotype-4a E237G was observed during selection and showed a 1.3-fold change in sensitivity to sofosbuvir, and N237S developed in genotype-6a which showed a modest 2.5-fold change in sensitivity to sofosbuvir. E237G has also been observed in vivo in two genotype-1a subjects, and at low levels (1.3%) in a single genotype-1b subject all of whom relapsed after receiving a sofosbuvir-containing regimen [37]. None of these three subjects’ NS5B isolates showed phenotypic resistance to sofosbuvir (<2.5-fold change), and the SDM of this mutation in genotype-1 reference strains showed no fold shift (data not shown). A more detailed understanding of these substitutions and any impact on sofosbuvir sensitivity will require gathering of considerable amounts of additional in vivo data. Given that S282T develops so rarely in the clinic [36,38], the probability of observing these additional mutations in combination with S282T is also rare, and indeed none of these additional mutations have been observed in vivo (with or without S282T). Moreover, the majority of mutations tested in combination with S282T showed no obvious benefit on replication or sofosbuvir sensitivity. Further characterization of these combinations will be evaluated if they are observed in the clinic.

Taken together, these data demonstrate the S282T substitution in NS5B is the primary substitution conferring resistance to sofosbuvir across multiple genotypes. Other substitutions observed during the selections had no or minimal impact on susceptibility, and were not consistently observed in greater than 15% of the replicon population across genotypes.


This work was funded by Gilead Sciences Inc. Manuscript preparation was assisted by Neil Parkin, Data First Consulting, Inc. (Belmont, CA, USA), and thanks to Charlotte Hedskog (Gilead Sciences, Inc., Foster City, CA, USA) for critical review and editing.

Disclosure statement

All authors were employees of Gilead Sciences at the time of the research or during preparation of the manuscript.

Additional files

Additional file 1: A table of NS5B mutations observed during sofosbuvir selection in genotype-2a replicon cells can be found at https://www.intmedpress.com/uploads/documents/3967_Xu_Addfile_1.pdf

Additional file 2: A table of NS5B mutations observed during sofosbuvir selection in genotype-4a replicon cells can be found at https://www.intmedpress.com/uploads/documents/3967_Xu_Addfile_2.pdf


1. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology 2013; 57: 1333-1342. Medline doi:10.1002/hep.26141
2. Smith DB, Bukh J, Kuiken C, et al. Expanded classification of hepatitis C virus into 7 genotypes and 67 subtypes: updated criteria and genotype assignment web resource. Hepatology 2014; 59: 318-327. Medline doi:10.1002/hep.26744
3. Simmonds P. Genetic diversity and evolution of hepatitis C virus--15 years on. J Gen Virol 2004; 85: 3173-3188. Medline doi:10.1099/vir.0.80401-0
4. Simmonds P, Bukh J, Combet C, et al. Consensus proposals for a unified system of nomenclature of hepatitis C virus genotypes. Hepatology 2005; 42: 962-973. Medline doi:10.1002/hep.20819
5. Pawlotsky JM. Hepatitis C virus: standard-of-care treatment. Adv Pharmacol 2013; 67: 169-215. Medline doi:10.1016/B978-0-12-405880-4.00005-6
6. Shah N, Pierce T, Kowdley KV. Review of direct-acting antiviral agents for the treatment of chronic hepatitis C. Expert Opin Investig Drugs 2013; 22: 1107-1121. Medline doi:10.1517/13543784.2013.806482
7. Sofia MJ, Bao D, Chang W, et al. Discovery of a beta-d-2′-deoxy-2′-alpha-fluoro-2′-beta-C-methyluridine nucleotide prodrug (PSI-7977) for the treatment of hepatitis C virus. J Med Chem 2010; 53: 7202-7218. Medline doi:10.1021/jm100863x
8. Pockros PJ. Nucleoside/nucleotide analogue polymerase inhibitors in development. Clin Liver Dis 2013; 17: 105-110. Medline doi:10.1016/j.cld.2012.09.007
9. Gentile I, Borgia F, Buonomo AR, Castaldo G, Borgia G. A novel promising therapeutic option against hepatitis C virus: an oral nucleotide NS5B polymerase inhibitor sofosbuvir. Curr Med Chem 2013; 20: 3733-3742. Medline doi:10.2174/09298673113209990178
10. Jacobson IM, Gordon SC, Kowdley KV, et al. Sofosbuvir for hepatitis C genotype 2 or 3 in patients without treatment options. N Engl J Med 2013; 368: 1867-1877. Medline doi:10.1056/NEJMoa1214854
11. Lawitz E, Mangia A, Wyles D, et al. Sofosbuvir for previously untreated chronic hepatitis C infection. N Engl J Med 2013; 368: 1878-1887. Medline doi:10.1056/NEJMoa1214853
12. Lawitz E, Lalezari JP, Hassanein T, et al. Sofosbuvir in combination with peginterferon alfa-2a and ribavirin for non-cirrhotic, treatment-naive patients with genotypes 1, 2, and 3 hepatitis C infection: a randomised, double-blind, Phase 2 trial. Lancet Infect Dis 2013; 13: 401-408. Medline doi:10.1016/S1473-3099(13)70033-1
13. Gane EJ, Stedman CA, Hyland RH, et al. Nucleotide polymerase inhibitor sofosbuvir plus ribavirin for hepatitis C. N Engl J Med 2013; 368: 34-44. Medline doi:10.1056/NEJMoa1208953
14. Afdhal N, Zeuzem S, Kwo P, et al. Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection. N Engl J Med 2014; 370: 1889-1898. Medline doi:10.1056/NEJMoa1402454
15. Afdhal N, Reddy KR, Nelson DR, et al. Ledipasvir and sofosbuvir for previously treated HCV genotype 1 infection. N Engl J Med 2014; 370: 1483-1493. Medline doi:10.1056/NEJMoa1316366
16. Najera I. Resistance to HCV nucleoside analogue inhibitors of hepatitis C virus RNA-dependent RNA polymerase. Curr Opin Virol 2013; 3: 508-513. Medline doi:10.1016/j.coviro.2013.08.011
17. Ludmerer SW, Graham DJ, Boots E, et al. Replication fitness and NS5B drug sensitivity of diverse hepatitis C virus isolates characterized by using a transient replication assay. Antimicrob Agents Chemother 2005; 49: 2059-2069. Medline doi:10.1128/AAC.49.5.2059-2069.2005
18. Migliaccio G, Tomassini JE, Carroll SS, et al. Characterization of resistance to non-obligate chain-terminating ribonucleoside analogs that inhibit hepatitis C virus replication in vitro. J Biol Chem 2003; 278: 49164-49170. Medline doi:10.1074/jbc.M305041200
19. Stuyver LJ, McBrayer TR, Tharnish PM, et al. Inhibition of hepatitis C replicon RNA synthesis by beta-D-2′-deoxy-2′-fluoro-2′-C-methylcytidine: a specific inhibitor of hepatitis C virus replication. Antivir Chem Chemother 2006; 17: 79-87. Medline doi:10.1177/095632020601700203
20. Dutartre H, Bussetta C, Boretto J, Canard B. General catalytic deficiency of hepatitis C virus RNA polymerase with an S282T mutation and mutually exclusive resistance towards 2′-modified nucleotide analogues. Antimicrob Agents Chemother 2006; 50: 4161-4169. Medline doi:10.1128/AAC.00433-06
21. Lam AM, Espiritu C, Bansal S, et al. Genotype and subtype profiling of PSI-7977 as a nucleotide inhibitor of hepatitis C virus. Antimicrob Agents Chemother 2012; 56: 3359-3368. Medline doi:10.1128/AAC.00054-12
22. Le Pogam S, Jiang WR, Leveque V, et al. In vitro selected Con1 subgenomic replicons resistant to 2′-C-methyl-cytidine or to R1479 show lack of cross resistance. Virology 2006; 351: 349-359. Medline doi:10.1016/j.virol.2006.03.045
23. Tong X, Le Pogam S, Li L, et al. In vivo emergence of a novel mutant L159F/L320F in the NS5B polymerase confers low-level resistance to the HCV polymerase inhibitors mericitabine and sofosbuvir. J Infect Dis 2014; 209: 668-675. Medline doi:10.1093/infdis/jit562
24. Tong X, Li L, Haines K, Najera I. Identification of the NS5B S282T resistant variant and two novel amino acid substitutions that affect replication capacity in hepatitis C virus-infected patients treated with mericitabine and danoprevir. Antimicrob Agents Chemother 2014; 58: 3105-3114. Medline doi:10.1128/AAC.02672-13
25. Robinson M, Yang H, Sun SC, et al. Novel hepatitis C virus reporter replicon cell lines enable efficient antiviral screening against genotype 1a. Antimicrob Agents Chemother 2010; 54: 3099-3106. Medline doi:10.1128/AAC.00289-10
26. Blight KJ, McKeating JA, Marcotrigiano J, Rice CM. Efficient replication of hepatitis C virus genotype 1a RNAs in cell culture. J Virol 2003; 77: 3181-3190. Medline doi:10.1128/JVI.77.5.3181-3190.2003
27. Lohmann V, Korner F, Koch J-O, Herian U, Theilmann L, Bartenschlager R. Replication of subgenomic hepatitis C virus RNAs in a hepatoma cell line. Science 1999; 285: 110-113. Medline doi:10.1126/science.285.5424.110
28. Paulson MS, Yang H, Shih IH, et al. Comparison of HCV NS3 protease and NS5B polymerase inhibitor activity in 1a, 1b and 2a replicons and 2a infectious virus. Antiviral Res 2009; 83: 135-142. Medline doi:10.1016/j.antiviral.2009.04.003
29. Yu M, Corsa AC, Xu S, et al. In vitro efficacy of approved and experimental antivirals against novel genotype 3 hepatitis C virus subgenomic replicons. Antiviral Res 2013; 100: 439-445. Medline doi:10.1016/j.antiviral.2013.08.018
30. Peng B, Yu M, Xu S, et al. Development of robust hepatitis C virus genotype 4 subgenomic replicons. Gastroenterology 2013; 144: 59-61. Medline doi:10.1053/j.gastro.2012.09.033
31. Wong KA, Xu S, Martin R, Miller MD, Mo H. Tegobuvir (GS-9190) potency against HCV chimeric replicons derived from consensus NS5B sequences from genotypes 2b, 3a, 4a, 5a, and 6a. Virology 2012; 429: 57-62. Medline doi:10.1016/j.virol.2012.03.025
32. Shih IH, Vliegen I, Peng B, et al. Mechanistic characterization of GS-9190 (tegobuvir), a novel non-nucleoside inhibitor of hepatitis C virus NS5B polymerase. Antimicrob Agents Chemother 2011; 55: 4196-4203. Medline doi:10.1128/AAC.00307-11
33. Hung M, Gibbs CS, Tsiang M. Biochemical characterization of rhinovirus RNA-dependent RNA polymerase. Antiviral Res 2002; 56: 99-114. Medline doi:10.1016/S0166-3542(02)00101-8
34. Kulkarni AS, Damha MJ, Schinazi RF, et al. A complex network of interactions between S282 and G283 of hepatitis C virus nonstructural protein 5B and the template strand affects susceptibility to sofosbuvir and ribavirin. Antimicrob Agents Chemother 2016; 60: 2018-2027. Medline doi:10.1128/AAC.02436-15
35. Lawitz E, Poordad FF, Pang PS, et al. Sofosbuvir and ledipasvir fixed-dose combination with and without ribavirin in treatment-naive and previously treated patients with genotype 1 hepatitis C virus infection (LONESTAR): an open-label, randomised, Phase 2 trial. Lancet 2014; 383: 515-523. Medline doi:10.1016/S0140-6736(13)62121-2
36. Svarovskaia ES, Dvory-Sobol H, Parkin N, et al. Infrequent development of resistance in genotype 1-6 hepatitis C virus-infected subjects treated with sofosbuvir in Phase 2 and 3 clinical trials. Clin Infect Dis 2014; 59: 1666-1674. Medline doi:10.1093/cid/ciu697
37. Sarrazin C, Dvory-Sobol H, Svarovskaia ES, et al. Prevalence of resistance-associated substitutions in HCV NS5A, NS5B, or NS3 and outcomes of treatment with ledipasvir and sofosbuvir. Gastroenterology 2016; 151: 501-512. Medline doi:10.1053/j.gastro.2016.06.002
38. Gane E, Abergel A, Metivier S, et al. The emergence of the NS5B resistant associated variant S282T after sofosbuvir-based treatment. The Liver Meeting, AASLD. 13–17 November 2015, San Francisco, CA, USA.

Copyright © 2021 Nucleus Holdings Ltd. Part of Nucleus Global.
Design and Technology by Nucleus Global
Company registration No. 321 0712 (England & Wales). Registered Office address: Admiral House 76-78 Old Street, London EC1V 9AZ.