Norovirus Polymerase Fidelity Contributes to Viral Transmission In Vivo

Abstract

Intrahost genetic diversity and replication error rates are intricately linked to RNA virus pathogenesis, with alterations in viral polymerase fidelity typically leading to attenuation during infections in vivo. We have previously shown that norovirus intrahost genetic diversity also influences viral pathogenesis using the murine norovirus model, as increasing viral mutation frequency using a mutagenic nucleoside resulted in clearance of a persistent infection in mice. Given the role of replication fidelity and genetic diversity in pathogenesis, we have now investigated whether polymerase fidelity can also impact virus transmission between susceptible hosts. We have identified a high-fidelity norovirus RNA-dependent RNA polymerase mutant (I391L) which displays delayed replication kinetics in vivo but not in cell culture. The I391L polymerase mutant also exhibited lower transmission rates between susceptible hosts than the wild-type virus and, most notably, another replication defective mutant that has wild-type levels of polymerase fidelity. These results provide the first experimental evidence that norovirus polymerase fidelity contributes to virus transmission between hosts and that maintaining diversity is important for the establishment of infection. This work supports the hypothesis that the reduced polymerase fidelity of the pandemic GII.4 human norovirus isolates may contribute to their global dominance. IMPORTANCE Virus replication fidelity and hence the intrahost genetic diversity of viral populations are known to be intricately linked to viral pathogenesis and tropism as well as to immune and antiviral escape during infection. In this study, we investigated whether changes in replication fidelity can impact the ability of a virus to transmit between susceptible hosts by the use of a mouse model for norovirus. We show that a variant encoding a high-fidelity polymerase is transmitted less efficiently between mice than the wild-type strain. This constitutes the first experimental demonstration that the polymerase fidelity of viruses can impact transmission of infection in their natural hosts. These results provide further insight into potential reasons for the global emergence of pandemic human noroviruses that display alterations in the replication fidelity of their polymerases compared to nonpandemic strains.

Keywords: RNA polymerases; noroviruses; polymerase fidelity; quasispecies; virus transmission.

FIG 1

Substitution I391L in the MNV NS7 RNA polymerase leads to restricted genetic diversity in viral populations. (A to C) Sensitivity of the WT strain and mutants I391L and S313T to different mutagenic compounds: 5-fluorouracil (A), favipiravir (B), and ribavirin (C). Sensitivity to each mutagen is expressed as the change in virus titer relative to the value for infections in the absence of mutagens. Each value is the average of results of three independent biological replicates (n = 3), with the error bars representing standard errors of the means (SEM). An apparent absence of SEM bars indicates that the bar is smaller than the respective symbol for these values. Sensitivity values for each MNV NS7 mutant were compared with WT values. Hence, those mutants with titer change curves above the WT curve were less sensitive to drug treatment, whereas those with curves below the WT curve showed greater sensitivity. Statistical analysis of relative levels of sensitivity in each NS7 mutant compared to the WT was performed by two-way analysis of variance (ANOVA) (*, P < 0.05; **, P < 0.01; ***, P < 0.001). (D) Mutation frequency analysis of MNV populations obtained after 8 serial passages in RAW264.7 cells in the absence of mutagenic treatment. Viral RNA was extracted and amplified by RT followed by PCR as explained in Materials and Methods. Amplified cDNA molecules were ligated to PCR Blunt vector, and positive colonies containing MNV insertions were sequenced. Statistical significance in the genetic diversity of different mutant viruses (S313T and I391L) with respect to that of the WT was determined by a Mann-Whitney test, which was used to compare the ranked scores of the numbers of mutations found in individual clones grouped by population, as further explained in Materials and Methods (ns, not significant; **, P < 0.05). Repeated mutations were taken into consideration only once for the analysis, as previously described (10). These mutation frequencies correspond to the identification of 34, 23, and 10 unique mutations during the sequence analysis of 127,870, 71,844, and 117,540 nucleotides (nt) in molecular clones from WT, S313T, and I391L populations, respectively. The total numbers of clones analyzed (n) in this study were 60, 34, and 56 for the WT, S313T, and I391L populations, respectively.

FIG 2

The MNV NS7 RNA polymerase mutant I391L shows WT-like replication kinetics during infection in cell culture. (A) Virus yield after reverse genetics recovery of full-length cDNA constructs containing either wild-type (WT) MNV or the polymerase mutants S313T and I391L. The values shown are the averages of results of at least three independent virus titer determinations for up to 4 independent cDNA transfections. Error bars represent SEM, with the statistical significance determined by one-way ANOVA (ns, not significant; ***, P < 0.001). (B to D) Replication kinetics of WT and mutant MNVs in RAW264.7 cells infected at the following different MOI’s: 0.01 TCID₅₀/cell (B), 5 TCID₅₀/cell (C), and 10 TCID₅₀/cell (D). The values represent the averages of results of three independent biological replicates (n = 3), with the error bars representing SEM. Data were analyzed by two-way ANOVA (*, P < 0.05). The dashed line in panels A to C represents the limit of detection.

FIG 3

MNV NS7 RNA polymerase mutant I391L shows delayed replication kinetics in mice. (A and B) Virus titers (A) and viral RNA levels (B) shed in the feces of animals inoculated with 3 × 10² TCID₅₀ MNV (n = 5). Mice were inoculated with recombinant viruses generated following recovery of MNV NS7 variants in BHK-21 cells by reverse genetics followed by 4 serial infections in RAW264.7 cells, as previously described (39). The presence of the desired mutations in the viral genome was confirmed by sequencing. Data were analyzed by two-way ANOVA (*, P < 0.05; **, P < 0.01; ***, P < 0.001). gRNA eq, genomic RNA equivalents. (C) Animals were infected as described for panel A, but the mice were inoculated with a dose of 3 × 10³ TCID₅₀ MNV per mouse (n = 5). Data represent results from a two-way ANOVA test (***, P < 0.001). (D) Viral RNA levels detected in various tissues collected from mice inoculated with 3 × 10² TCID₅₀ of WT MNV or the NS7 polymerase mutants at 48 h postinfection. MLN, mesenteric lymph node (two-way ANOVA test results were nonsignificant). In all panels, the dashed line represents the detection limit. The apparent absence of SEM bars for some values in panels B and C indicates that the bar is smaller than the respective symbol.

FIG 4

MNV NS7 RNA polymerase mutant I391L is less sensitive to favipiravir treatment in vivo. Six animals were infected with 3 × 10² TCID₅₀ MNV (n = 6). After 7 days, persistently infected mice were subjected to favipiravir treatment twice a day over a period of 4 days. Mouse feces samples were collected, and the titers (TCID₅₀ per gram) were determined as previously described (10). The virus titer values represented are relative to the average virus titer at the day 0 time point, collected immediately before the beginning of treatment. Absolute virus titer values at day 0 were 5.80 and 5.51 log₁₀ TCID₅₀/g stool for the WT and the I391L mutant, respectively. Error bars represent SEM from a two-way ANOVA test (ns, not significant; *, P < 0.05; **, P < 0.01).

FIG 5

MNV NS7 RNA polymerase mutant I391L shows delayed transmission in susceptible hosts. Virus transmission from infected to uninfected mice was determined as detailed in Materials and Methods. In brief, five naive mice in each cage were housed with one mouse inoculated by oral gavage with 3 × 10² TCID₅₀ MNV-3. Every mutant was tested in 4 independent cages (n = 20). MNV transmission was determined by qPCR detection of MNV in the feces of naive mice along a series of time points. (A and B) Graphs showing the percentages of noninfected animals along a series of time points (the log rank test was used for statistical analysis). (A) WT and S313T showed no significant (ns) difference in their transmission kinetics between mice. (B) The WT strain was transmitted significantly faster than the I391L mutant (**; P < 0.01). S313T also showed significantly increased transmission compared to I391L (not shown in the figure; P < 0.05). Animals directly infected by oral-gavage inoculation are not represented. (C) Viral RNA levels found in the feces of noninoculated animals after 4 days in contact with infected mice (n = 20). The dashed line represents the limit of detection, and the numbers above the groups represent the numbers of animals shedding viral RNA at levels below the detection limit. Animals directly infected by oral-gavage inoculation are not represented. Data represent results of a two-way ANOVA test (**, 0.001 < P < 0.01).