Mechanism of resistance of hepatitis C virus replicons to structurally distinct cyclophilin inhibitors

Abstract

The current standard of care for hepatitis C virus (HCV) infection, pegylated alpha interferon in combination with ribavirin, has a limited response rate and adverse side effects. Drugs targeting viral proteins are in clinical development, but they suffer from the development of high viral resistance. The inhibition of cellular proteins that are essential for viral amplification is thought to have a higher barrier to the emergence of resistance. Three cyclophilin inhibitors, the cyclosporine analogs DEBIO-025, SCY635, and NIM811, have shown promising results for the treatment of HCV infection in early clinical trials. In this study, we investigated the frequency and mechanism of resistance to cyclosporine (CsA), NIM811, and a structurally unrelated cyclophilin inhibitor, SFA-1, in replicon-containing Huh7 cells. Cross-resistance between all clones was observed. NIM811-resistant clones were selected only after obtaining initial resistance to either CsA or SFA-1. The time required to select resistance against cyclophilin inhibitors was significantly longer than that required for resistance selection against viral protein inhibitors, and the achievable resistance level was substantially lower. Resistance to cyclophilin inhibitors was mediated by amino acid substitutions in NS3, NS5A, and NS5B, with NS5A mutations conferring the majority of resistance. Mutation D320E in NS5A mediated most of the resistance conferred by NS5A. Taken together, the results indicate that there is a very low frequency and level of resistance to cyclophilin-binding drugs mediated by amino acid substitutions in three viral proteins. The interaction of cyclophilin with NS5A seems to be the most critical, since the NS5A mutations have the largest impact on resistance.

FIG. 1.

Susceptibility of wild-type, CsA-resistant, and CsA/NIM-resistant cells to CsA and NIM811. Wild-type (DMSOr), CsAr, and CsA/NIMr clones were treated for 48 h with CsA (A) or NIM811 (B). Viral RNA was quantified by qRT-PCR and normalized to total cellular RNA. The susceptibility curves of a representative experiment (A and B) show the level of viral RNA after drug treatment as the percentage of DMSO control cells. The standard deviation of six replicas is included. The experiment has been repeated several times, and the same trend has been observed. The EC₅₀ for CsA was 0.66, 3.39, and >10 μM in DMSOr, CsAr, and CsA/NIMr replicon-containing cells, respectively. The EC₅₀ for NIM811 was 0.21, 0.83, and 3.5 μM in DMSOr, CsAr, and CsA/NIMr replicon-containing cells, respectively.

FIG. 2.

Susceptibility of wild-type, SFA-resistant, and SFA/NIM-resistant cells to SFA-1 and NIM811. Wild-type (DMSOr), SFA-1r, and SFA/NIMr clones were treated for 48 h with SFA-1 (A) or NIM811 (B). Viral RNA was quantified by qRT-PCR and normalized to total cellular RNA. The susceptibility curves of a representative experiment (A and B) show the level of viral RNA after drug treatment as the percentage of DMSO control cells. The standard deviations from six replicas are included. The experiment was repeated several times, and the same trend was observed. The EC₅₀ for SFA-1 was 0.25, 4.55, and >10 μM in DMSOr, SFA-1r, and SFA/NIMr replicon-containing cells, respectively. The EC₅₀ for NIM811 was 0.37, 1.26, and 4.26 μM in DMSOr, SFA-1r, and SFA/NIMr replicon-containing cells, respectively.

FIG. 3.

Resistant replicon-containing cell lines contain several mutations. RNA of the resistant cell lines and of control cells grown for the same time period with DMSO, the solvent of the drugs, was sequenced. Shown are the mutations in the replicon that were detected only in the resistant cells.

FIG. 4.

Contribution of NS5A and NS5B mutations to resistance against NIM811. The NS5A gene of the CsA/NIMr clone was engineered into the wild-type replicon, and the NS5B mutation was added by site-directed mutagenesis. NIM811 susceptibility curves were established. The graph shows the results of two experiments with six replicas each, with standard deviations indicated by error bars. The EC₅₀ for NIM811 was 0.23, 0.94, and 1.14 μM in WT, WT + resistant NS5A, and WT + resistant NS5A & NS5B replicon-containing cells, respectively.

FIG. 5.

Contribution of the NS5A mutations D320E and R356Q to resistance against NIM811. The mutations were introduced separately or together into the wild-type replicon (A), and the mutations were reversed in the CsA/NIMr clone (B). Susceptibility to NIM811 was analyzed, and the EC₅₀s are shown in comparison to those of the wild-type and mutated NS5A gene containing all six mutations. The graphs are the results from two experiments, with standard deviations indicated by error bars.

FIG. 6.

Mutations of the CsA/NIMr replicon do not affect the replication rate. Equal numbers of clone A, DMSOr, and CsA/NIMr cells were seeded, and the HCV RNA copy number/ng total RNA was analyzed by qRT-PCR after 48 h of growth. The number of experimental repeats and standard deviations are included for each cell type (A). RNA of the indicated replicons (B) was transfected into Huh-cure cells, and the luciferase signal was analyzed after 24, 48, and 72 h. The graphs show the percent increase of the luciferase signal relative to the 24-h value. A repeat of this experiment showed the same trend.