Selection and Characterization of Rupintrivir-Resistant Norwalk Virus Replicon Cells In Vitro

Abstract

Human norovirus (HuNoV) is a major cause of nonbacterial gastroenteritis worldwide, yet despite its impact on society, vaccines and antivirals are currently lacking. A HuNoV replicon system has been widely applied to the evaluation of antiviral compounds and has thus accelerated the process of drug discovery against HuNoV infection. Rupintrivir, an irreversible inhibitor of the human rhinovirus 3C protease, has been reported to inhibit the replication of the Norwalk virus replicon via the inhibition of the norovirus protease. Here we report, for the first time, the generation of rupintrivir-resistant human Norwalk virus replicon cells in vitro Sequence analysis revealed that these replicon cells contained amino acid substitutions of alanine 105 to valine (A105V) and isoleucine 109 to valine (I109V) in the viral protease NS6. The application of a cell-based fluorescence resonance energy transfer (FRET) assay for protease activity demonstrated that these substitutions were involved in the enhanced resistance to rupintrivir. Furthermore, we validated the effect of these mutations using reverse genetics in murine norovirus (MNV), demonstrating that a recombinant MNV strain with a single I109V substitution in the protease also showed reduced susceptibility to rupintrivir. In summary, using a combination of different approaches, we have demonstrated that, under the correct conditions, mutations in the norovirus protease that lead to the generation of resistant mutants can rapidly occur.

Keywords: antiviral agents; drug resistance mechanisms; noroviruses; protease inhibitors; proteases.

FIG 1

Isolation of rupintrivir-resistant HGT-NV replicon cells. (A) Dose-response curve of effect of rupintrivir on HuNoV replicon RNA levels in HGT-NV cells. The level of replicon RNA in HTG-NV cells treated with DMSO or rupintrivir was measured by quantitative RT-PCR. Inhibition was plotted as a percentage relative to that in DMSO-treated cells. Error bars represent means ± standard deviations from three independent experiments. (B) Reduction in the GI replicon RNA level over time in HGT-NV cells treated with rupintrivir in the absence of G418. The levels of HuNoV replicon RNA relative to those observed in control DMSO-treated cells were plotted. Error bars represent means ± standard deviations from three biological replicates. Dashed line, detection limit for replicon RNA. (C) Colony formation assays with HGT-NV cells treated with DMSO or rupintrivir for 12 days. Colonies were stained and photographed on day 7 after treatment with G418 (1.5 mg/ml) in the absence of rupintrivir. (D) Schematic overview of the procedure used for the repetitive cultivation of HGT-NV cells in the presence of increasing concentrations of rupintrivir. HGT-NV cells were maintained as subconfluent cultures in the presence of DMSO or rupintrivir and G418 and passaged every 2 to 3 days for 45 days.

FIG 2

Amino acid sequence alignment and structural comparison of norovirus proteases. (A) Amino acid sequence alignment of the GI (GenBank accession number M87661), GII (GenBank accession number DQ658413), and GV (GenBank accession number DQ285629) proteases used in this study. Identical amino acids (in black) are highlighted. Black arrowheads, positions of the amino acids where substitutions were observed in rupintrivir-resistant replicon cells. (B) Crystal structures of the GI Southampton norovirus protease in complex with a peptidyl inhibitor (gray) (PDB accession number 2IPH) (19) and the GV MNV protease (PDB accession number 4ASH) (20). Residues A105 and I109 in the GI and GV proteases are indicated in red. The crystal structures of poliovirus (PV) protease in complex with GC376 (blue) (PDB accession number 4DCD) (24) and human rhinovirus 2 (HRV2) protease in complex with rupintrivir (cyan) (PBD accession number 1CQQ) (26) are also shown. Residue G128 in the PV protease and residues N165, E3, and A103 in the HRV2 protease, involved in rupintrivir resistance, are indicated in red (23, 25). The catalytic residues C139 (GI), A139 mutated from cysteine (GV), C147 (PV), and C147 (HRV2) are indicated in orange. (C) The crystal structure of the GI, GV, or PV protease was aligned to the HRV2 protease crystal structure in complex with rupintrivir and shown as the hydrophobic surface of each protease with rupintrivir (cyan). Residues A105 and I109 in the GI and GV protease and residue G128 in the PV protease are indicated in red.

FIG 3

Inhibitory effects of rupintrivir on the mutant proteases in a cell-based FRET assay. (A) The FRET signal from a sensor (YFP/CFP ratio) was lost upon cotransfection with WT or mutant protease but not with the inactive (H30A) protease and the empty vector (EV). The GI, GII, and GV mutant proteases showed cleavage activity comparable to that of the wild type (WT). (B) Western blot analysis of HEK293T cells cotransfected with the GI, GII, or GV NS1/2-NS3 FRET sensor along with either the WT or mutant proteases. Similar cleavage products corresponding to CFP and YFP were observed in the WT and mutant proteases in all genogroups. MM, molecular mass. (C) Dose-response inhibition curves showing the effect of rupintrivir on either the WT or mutant proteases of GI, GII, and GV genotypes in the cell-based FRET assay. (D) IC₅₀s of rupintrivir along with the standard deviations of the means from three independent experiments. Asterisks indicate statistically significant differences, as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Abbreviation: n.s., not significant.

FIG 4

Impact of rupintrivir resistance-conferring mutations on MNV polyprotein processing in vitro. In vitro translation reactions were programmed with equal quantities of in vitro-transcribed MNV full-length capped RNA containing either the WT or mutant proteases in the presence and absence of 80 μM rupintrivir. The [³⁵S]methionine-labeled products were subsequently analyzed by SDS-PAGE and phosphorimaging analysis. The percentage of fully processed NS3, determined on the basis of the expected molecular mass of the proteins, was calculated and expressed relative to the total amount of viral protein produced.

FIG 5

Impact of rupintrivir resistance-conferring mutations on the recovery of recombinant MNV. (A and B) Virus titers obtained at 24 h after transfection of BSR-T7 cells with MNV cDNA (A) or in vitro-transcribed and capped MNV RNA (B). The virus titers in BV-2 cells were determined by the TCID₅₀ method. Error bars represent means ± standard deviations from three independent experiments. Dotted lines, detection limit of the assay. (C) The increase in the RNA level (the RNA level at 12 h postinoculation minus the RNA level at 1 h postinoculation) was determined by qRT-PCR after three passages with BV-2 cells. The supernatant of BSR-T7 cells transfected with MNV cDNA was inoculated onto BV-2 cells. At 48 h postinoculation, the supernatant of BV-2 cells was next inoculated onto BV-2 cells. This passage was repeated two times. After three passages, the supernatant was inoculated onto BV-2 cells and the increase in the RNA level was measured. No increase in the amount of viral RNA of the A105V and A105V/I109V recombinant viruses was observed.

FIG 6

The I109V substitution in the MNV protease results in reduced susceptibility to rupintrivir during viral replication in cell culture. (A) The growth kinetics of the WT and I109V recombinant MNV were assessed following infection of BV-2 cells at MOIs of 0.1, 1, and 10 TCID₅₀/cell. Viral infectivity levels were determined by measuring the release of infectious virus into the culture supernatant at various times postinfection by determination of the TCID₅₀. (B) The inhibitory effects of rupintrivir and 2CMC against either the WT or the recombinant MNV I109V mutant were assessed following infection of BV-2 cells at MOIs of 0.1 and 1 TCID₅₀/cell. A significant difference in viral titers following treatment with 10 and 20 μM and 10, 20, and 40 μM rupintrivir was observed between the WT and I109V recombinant viruses at MOIs of 0.1 and 1, respectively (*, P < 0.05; **, P < 0.01; ***, P < 0.001). In contrast, there were no significant differences in virus titers following treatment with 2CMC.