Ribavirin-Resistant Mutants of Human Enterovirus 71 Express a High Replication Fidelity Phenotype during Growth in Cell Culture

Abstract

It has been shown in animal models that ribavirin-resistant poliovirus with a G64S mutation in its 3D polymerase has high replication fidelity coupled with attenuated virulence. Here, we describe the effects of mutagenesis in the human enterovirus 71 (HEV71) 3D polymerase on ribavirin resistance and replication fidelity. Seven substitutions were introduced at amino acid position 3D-G64 of a HEV71 full-length infectious cDNA clone (26M). Viable clone-derived virus populations were rescued from the G64N, G64R, and G64T mutant cDNA clones. The clone-derived G64R and G64T mutant virus populations were resistant to growth inhibition in the presence of 1,600 μM ribavirin, whereas the growth of parental 26M and the G64N mutant viruses were inhibited in the presence of 800 μM ribavirin. Nucleotide sequencing of the 2C and 3D coding regions revealed that the rate of random mutagenesis after 13 passages in the presence of 400 μM ribavirin was nearly 10 times higher in the 26M genome than in the mutant G64R virus genome. Furthermore, random mutations acquired in the 2C coding regions of 26M and G64N conferred resistance to growth inhibition in the presence of 0.5 mM guanidine, whereas the G64R and G64T mutant virus populations remained susceptible to growth inhibition by 0.5 mM guanidine. Interestingly, a S264L mutation identified in the 3D coding region of 26M after ribavirin selection was also associated with both ribavirin-resistant and high replication fidelity phenotypes. These findings are consistent with the hypothesis that the 3D-G64R, 3D-G64T, and 3D-S264L mutations confer resistance upon HEV71 to the antiviral mutagen ribavirin, coupled with a high replication fidelity phenotype during growth in cell culture.

INTRODUCTION

Human enterovirus 71 (HEV71) belongs to the genus Enterovirus within the family Picornaviridae. HEV71 is composed of a single-stranded positive-sense RNA of approximately 7,500 nucleotides surrounded by an icosahedral capsid assembled from 60 copies of each of the four structural proteins, VP1 to VP4. Since its discovery in 1969 (1), HEV71 has been identified as the cause of epidemics of the mild febrile exanthema known as hand, foot, and mouth disease (HFMD). Although most HEV71 infections result in self-limiting conditions, such as HFMD, herpangina, or aseptic meningitis, some cases are associated with severe neurological complications, such as encephalitis and poliomyelitis-like paralysis (2, 3). In the Asia-Pacific region, HEV71 has recently emerged as a significant cause of acute neurological disease during HFMD epidemics, in particular, a syndrome of acute brainstem encephalitis associated with a high mortality rate (3).

The RNA genome of HEV71 encodes a single polyprotein that is proteolytically cleaved in infected cells to produce four structural (P1, VP1 to VP4) and seven nonstructural (P2, 2A to 2C; P3, 3A to 3D) viral proteins. The nonstructural protein 3D is an RNA-dependent RNA polymerase (RDRP) responsible for viral genome replication (4, 5). The HEV71 3D polymerase is a 52-kDa polypeptide that conforms to the “right-hand” consensus polymerase motif, consisting of “palm,” “fingers,” and “thumb” domains (6, 7). Based on data derived from poliovirus, it has been suggested that picornaviral RDRPs have evolved primarily by their selection for high-speed replication, which has resulted in a concomitant reduction in replication fidelity (8). Indeed, the lack of a proofreading function of viral RDRPs has been shown to result in high rates of genome mutation that are thought to contribute to viral RNA genome diversity and adaptability (8).

Ribavirin (1-β-d-ribofuranosyl-1,2,4-triazole-3-carboxamide) is a mutagenic guanosine analogue. The antiviral activity of ribavirin is thought to be due to its ability to increase the nucleotide incorporation error rate during both DNA and RNA viral replication (9–11). Furthermore, the selection of ribavirin resistance in poliovirus has been achieved by serial passages of the virus in cell culture in the presence of increasing concentrations of ribavirin (10). Of particular interest, there appears to be a link between ribavirin resistance and the fidelity of polioviral RNA replication. In particular, ribavirin-resistant strains with a G64S mutation in the 3D polymerase-coding region have been shown to express increased replication fidelity (10, 12, 13). Indeed, the key function of residue 64 (3D-G64) as a replication fidelity checkpoint in the poliovirus 3D polymerase has been identified through detailed genetic and phenotypic analyses of ribavirin-susceptible and -resistant strains (10, 14).

A recent study reported on the replication fidelity properties of wild-type and mutant 3D polymerases of poliovirus (15). This was achieved by observing for the selection of mutations that confer resistance to the growth-inhibitory effect of guanidine on RNA viruses. The resistance of poliovirus to growth inhibition by guanidine is known to be associated with point mutations in coding region 2C (16). The frequency of nucleotide misincorporations within 2C during viral replication can be estimated by measuring the percentages of cloned virus populations that develop resistance to the growth-inhibiting effects of guanidine (10). The guanidine resistance assay provides a useful phenotypic marker for assessing mutation frequencies within coding region 2C that are associated with viral RNA synthesis by wild-type and mutant 3D polymerases (13).

In this study, we aimed to investigate the effects of altering the amino acid residue at position G64 in the HEV71 3D polymerase on ribavirin resistance and on replication fidelity. We selected three viable 3D-G64 variant virus populations using site-directed mutagenesis of a full-length infectious cDNA clone of HEV71 (17). Clone-derived viruses (CDVs) containing the parental 3D-G64 sequence or variant 3D-G64N, 3D-G64R, or 3D-G64T sequences were generated. In addition, a CDV population containing the ribavirin passage-associated mutation, 3D-S264L, was also prepared. We then examined ribavirin susceptibility and replication fidelity of the parental and mutant 3D polymerase virus populations using both genetic and phenotypic studies.

MATERIALS AND METHODS

Virus, cell lines, and media.

HEV71 strain 26M/AUS/4/99 subgenotype B3 (called 26M in this study) was isolated in western Australia in 1999 (17, 18) during an outbreak of HFMD. Previously, 26M was fully sequenced and used to generate an infectious cDNA clone (17). Human rhabdomyosarcoma (RD; ATCC CCL-136), African green monkey kidney (Vero; ATCC CCL-81), and simian virus 40-transfected African green monkey kidney (COS-7; CRL-1651) cell lines were maintained at 37°C in 5% CO₂. Dulbecco's modified Eagle's medium (DMEM) (HyClone) served as the maintenance medium and was supplemented with 2% fetal bovine serum (FBS) (Bovogen Biologicals) and 2 mM l-glutamine (Sigma). The growth medium was DMEM supplemented with 5% FBS and 2 mM l-glutamine.

Cloning.

The plasmid pcDNA3-3D, which contained the full 26M 3D coding region, was constructed for use in this study. G64A, G64E, G64L, G64N, G64P, G64S, G64T, and S264L mutations were introduced into pcDNA3-3D using site-directed mutagenesis with the QuikChange Lightning mutagenesis kit (Stratagene); the primers used in the mutagenesis assays are shown in Table S1 in the supplemental material. The 3D gene fragments carrying the G64A, G64E, G64L, G64N, G64P, G64S, G64T, and S264L mutations were cloned into the 26M infectious cDNA clone by digestion with MluI and EcoRV to produce HEV71-3D-G64A, HEV71-3D-G64E, HEV71-3D-G64L, HEV71-3D-G64N, HEV71-3D-G64P, HEV71-3D-G64S, HEV71-3D-G64T, and HEV71-3D-S264L, respectively.

The plasmid pUC18-2C, which contained the full 26M 2C coding region, was also constructed for use in this study. The M175K, K185R, K298N, E272K, and R293K mutations were introduced into pUC18-2C using site-directed mutagenesis with the QuikChange Lightning mutagenesis kit (Stratagene); the primers used in the mutagenesis assays are shown in Table S1 in the supplemental material. The 2C gene fragments carrying the M175K, K185R, K298N, E272K, and R293K mutations were cloned into the 26M infectious cDNA clone by digestion with ClaI and EcoRV to produce HEV71-2C-M175K, HEV71-2C-K185R, HEV71-2C-K298N, HEV71-2C-E272K, and HEV71-2C-R293K, respectively.

Transfection.

Recombinant full-length HEV71 plasmids were transfected into COS-7 cells using Lipofectamine 2000 (Invitrogen) to rescue clone-derived viruses (CDVs). Briefly, 1.6 μg of either HEV71-3D-G64A, HEV71-3D-G64E, HEV71-3D-G64L, HEV71-3D-G64N, HEV71-3D-G64P, HEV71-3D-G64S, HEV71-3D-G64T, HEV71-3D-S264L, HEV71-2C-M175K, HEV71-2C-K185R, HEV71-2C-E272K, HEV71-2C-R293K, or HEV71-2C-K298N and 1.6 μg pCMVT7-Pol expressing the T7 RNA polymerase (Pol) were diluted into a final volume of 100 μl containing 4 μl Lipofectamine 2000 in 96 μl Opti-MEM (Invitrogen). The two solutions were incubated for 5 min and then combined and incubated for an additional 20 min at room temperature. COS-7 cells in 12-well tissue culture plates (CellStar) were covered with 400 μl of the Opti-MEM solution. After 4 h of incubation at 37°C, 600 μl of growth medium was added to the plates, and the transfected cells were incubated for 72 h. The transfected cells were then subjected to freeze-thawing (3×), and the supernatants were clarified by centrifugation at 6,500 × g for 5 min; 200 μl of transfected COS-7 cell lysate was used to infect RD cells in 6-well plates. CDVs were passaged on RD cells to obtain working stocks. HEV71-infected cell culture supernatants were subjected to viral RNA extraction using the QIAamp viral RNA minikit (Qiagen), and viral cDNA synthesis was performed using Superscript III (Invitrogen) and the 3′ untranslated region reverse (3′ UTR-R) primer (see Table S1 in the supplemental material). The viral cDNA was amplified by PCR followed by nucleotide sequencing of the 3D coding regions to confirm the presence of the 3D-G64A, 3D-G64E, 3D-G64L, 3D-G64N, 3D-G64P, 3D-G64S, 3D-G64T, and 3D-S264L mutations and in the 2C coding region to confirm the presence of the M175K, K185R, E272K, R293K, and K298N mutations. The resulting CDV populations were named G64A, G64E, G64L, G64N, G64S, G64T, S264L, M175K, K185R, E272K, R293K, and K298N, respectively (the full-length 3D-G64P cDNA clone did not give rise to a viable clone-derived virus population) (Table 1).

Table 1.

Selection and stability of seven alternative amino acid residues at position 3D-G64 in HEV71^a

Amino acid replacement at position 3D-G64	Amino acid substitution in the CDV	Viability of the CDV	Stability of the genotype in RD cells
G64A (GGG→GCC)	W (TGG)	Viable	Further A→W/P mutation at passage 1
		Nonviable	W/P at passage 2
G64S (GGG→TCC)	G (GGG)	Viable	Reverted to parental sequence at passage 2
G64L (GGG→CTG)	G (GGG)	Viable	Reverted to parental sequence at passage 2
G64T (GGG→ACC)	T (ACC)	Viable	Stable at passage 5
G64N (GGG→AAC)	N (AAC)	Viable	Stable at passage 5
G64P (GGG→CCC)	—	Nonviable	—
G64E (GGG→GAG)	R (AGG)	Viable	Further E→R mutation at passage 2
			R stable to passage 10

^a–––, not applicable.

Characterization of HEV71 ribavirin-resistant mutants.

In order to assess the susceptibility of each viable CDV variant to growth inhibition in the presence of ribavirin, the growth of the G64N, G64R, G64T, and S264L CDV populations were compared to that of 26M in an RD cell culture for 24 h in the presence of either 0 μM, 800 μM, or 1,600 μM ribavirin. Briefly, RD cells in 12-well plates (CellStar) were covered with DMEM containing 0 μM, 800 μM, or 1,600 μM ribavirin for 1 h at 37°C. The ribavirin-pretreated RD cells were then washed with phosphate-buffered saline (PBS). Cells were then infected (multiplicity of infection [MOI], 0.1) with parental virus or one of the variants for 1 h at 37°C and washed with PBS, and maintenance medium containing either 0 μM, 800 μM, or 1,600 μM ribavirin was added. The infected cell cultures were then incubated at 37°C in 5% CO₂ for 24 h. At the conclusion of the experiment, virus was released from the cells by freeze-thawing (3×), and the lysate was clarified by centrifugation at 6,500 × g for 5 min. The titer of the lysate was determined using a 50% tissue culture infective dose (TCID₅₀) assay on Vero cells following the Reed-Muench method (19).

Plaque purification.

Twelve-well tissue culture trays (CellStar) were seeded with Vero cells at a density of 2.2 × 10⁵ cells/well and were grown overnight at 37°C. Ten-fold serial dilutions of virus were inoculated onto the Vero cell monolayers at a volume of 200 μl per well. After incubation for 30 min at 37°C, the inoculum was removed, and the cells were washed with PBS. Cells were then overlaid with 1 ml maintenance medium containing 0.5% immunodiffusion-grade agarose (ICN) and were incubated at 37°C in 5% CO₂. After 4 days of incubation, an additional 0.5 ml of maintenance medium containing 0.5% immunodiffusion-grade agarose was added to each well, and the cultures were incubated for an additional 3 days. Isolated plaques were selected using a pipette tip to draw up the cells that were showing cytopathic effects (CPEs) within the plaque. This protocol was repeated three times to select plaque-purified virus populations.

Characterization of the replication fidelity of plaque-purified populations of HEV71 ribavirin-resistant mutants.

The sensitivities of the plaque-purified populations of G64N, G64R, G64T, and S264L to guanidine were compared to the plaque-purified 26M population as measured by growth in the presence of either 0 mM or 0.5 mM guanidine for 48 h. Briefly, RD cells were infected (MOI, 0.1) in 6-well plates (CellStar) for 1 h at 37°C. Cells were washed with PBS, and maintenance medium containing either 0 mM or 0.5 mM guanidine was added. Cells were incubated at 37°C in 5% CO₂ for 48 h. Virus was released from cells by freeze-thawing (3×), and the lysate was clarified by centrifugation at 6,500 × g for 5 min. The titer of the lysate was determined using a TCID₅₀ assay, as described above (19).

RNA extraction, cDNA synthesis, and nucleotide sequence analysis.

Viral RNA was extracted from the parental virus (26M) and from each of the 3D-G64 polymerase mutants, after 1 or 13 serial passages in RD cells in the presence of the mutagen ribavirin (400 μM), using the QIAamp viral RNA minikit (Qiagen). Viral RNA was reverse transcribed using Superscript III (Invitrogen) and the 3′ UTR-R primer (see Table S1 in the supplemental material). Viral cDNA was amplified by PCR and gel purified using a MinElute gel extraction kit (Qiagen). Nucleotide sequencing of the 2C and 3D coding regions was performed by the Australian Genome Research Facility using the primers listed in Table S1 in the supplemental material.

Single-step growth kinetics.

RD cell monolayers (1 × 10⁵ cells per well) that were grown overnight in 48-well tissue culture plates (CellStar) were infected (MOI, 10) for 1 h at 37°C. Cells were then washed with PBS (3×) and overlaid with 300 μl of maintenance medium per well. Virus was collected at four-hour intervals for 24 h by freeze-thawing (3×) and was clarified by centrifugation at 6,500 × g for 5 min. The first time point for virus collection (0 h) was immediately after the addition of the maintenance medium. The yields were quantified using a TCID₅₀ assay, as described above (19).

Characterization of HEV71 guanidine-resistant mutants.

The sensitivities of the M175K, K185R, K298N, E272K, and R293K CDVs to guanidine were compared to that of the parental virus as measured by growth in the presence of either 0 mM or 4 mM guanidine for 72 h. Briefly, RD cells were infected (at an MOI of 0.1) in 6-well plates (CellStar) for 1 h at 37°C. Cells were washed with PBS, and maintenance medium containing either 0 mM or 4 mM guanidine was added. Cells were incubated at 37°C in 5% CO₂ for 72 h. Virus was released from the cells by freeze-thawing (3×), and the lysate was clarified by centrifugation at 6,500 × g for 5 min. The titer of the lysate was determined using a TCID₅₀ assay, as described above (19).

Statistical methods.

All statistical analyses were performed using one-way analysis of variance (ANOVA) and Tukey's multiple comparison tests or the two-tailed Student t test. A P value of <0.05 was considered statistically significant. The data were analyzed using the GraphPad Prism online software package version 5.04.

Nucleotide sequence accession numbers.

Sequences newly determined in this study have been deposited in GenBank under accession numbers KC118542 (26M1) and KC118543 (26M2).

RESULTS

Engineering of HEV71 strains with mutations at position 3D-G64.

It has been shown that a single mutation (G64S) in the 3D polymerase of poliovirus is able to confer ribavirin resistance, which is also associated with high-fidelity 3D polymerase activity (10, 15, 20). Genetic evidence indicates that 3D-G64 is allied with a key fidelity checkpoint of the poliovirus 3D polymerase (14).

In order to investigate the replication fidelity properties of the HEV71 3D polymerase, we examined the effect of incorporating several alternative amino acid residues at amino acid position 64, which is also a glycine (G) residue in the parental 26M 3D polymerase (17). The mutations 3D-G64A, 3D-G64E, 3D-G64L, 3D-G64N, 3D-G64P, 3D-G64S, and 3D-G64T were introduced into the 26M infectious cDNA clone by site-directed mutagenesis. The CDVs were passaged five times in RD cells to assess the stabilities of the introduced mutations. Of the seven amino acid residues introduced at position 3D-G64, one substitution, 3D-G64P, did not produce a viable virus population (Table 1). By contrast, incorporation of the other residues at position 64 in the 3D polymerase resulted in the rescue of viable virus populations. However, the 3D-G64A mutation changed at the first RD passage to encode tryptophan (W), which was included in a mixed codon that included proline (P); the W/P mixed codon virus population failed to generate an observable cytopathic effect (CPE) after two passages in RD cells and was not examined further. Interestingly, the 3D-G64E mutation changed to encode arginine (R) during RD cell passage. Furthermore, the 3D-G64R mutation remained stably present through 10 additional passages in RD cells. The 3D-G64S and 3D-G64L mutations reverted to the parental (G) sequence after a single passage in RD cells.

Thus, only three of the 3D-G64 mutations—3D-G64N, 3D-G64R, and 3D-G64T—resulted in viable CDV populations that remained stable during subsequent passages in RD cells (Table 1). Given these findings, we decided to study the ribavirin resistance phenotype of the G64N, G64R, and G64T CDVs to determine if any of these mutations resulted in high-fidelity 3D polymerase variants.

Characterization of the ribavirin resistance phenotype of the HEV71 3D-G64 mutants in RD cell culture.

In order to examine the susceptibility of each of the viable 3D variant CDV populations to ribavirin, the G64N, G64R, and G64T variants and the parental virus were grown in RD cells for 24 h in the presence or absence of 800 μM or 1,600 μM ribavirin. The results of this experiment are shown in Fig. 1. The G64R and G64T variants were both capable of growth in the presence of 1,600 μM ribavirin (G64R, 7.9 × 10⁴ TCID₅₀/ml; G64T, 3.9 × 10⁴ TCID₅₀/ml), with titers approximately 10-fold below those in the absence of ribavirin (G64R, 7.9 × 10⁵ TCID₅₀/ml; G64T, 3.1 × 10⁵ TCID₅₀/ml), whereas the growth of G64N and parental virus was inhibited in the presence of 800 μM ribavirin (parental virus, 5.4 × 10⁴ TCID₅₀/ml; G64N, 3.9 × 10⁴ TCID₅₀/ml) by approximately 40- to 50-fold below that in the absence of ribavirin (parental virus, 2.5 × 10⁶ TCID₅₀/ml; G64N, 1.3 × 10⁶ TCID₅₀/ml). These experiments confirmed that the G64R and G64T mutations in the 3D polymerase of HEV71 can confer resistance to the growth-inhibiting effect of ribavirin, as has been observed with poliovirus (10, 15). Furthermore, the ribavirin resistance phenotype in HEV71 is amino acid residue specific and is different than that of poliovirus.

Fig 1.

Characterization of HEV71 ribavirin-resistant mutants in RD cells. (A) Titers of G64N, G64R, G64T, and 26M after 24 h of growth in RD cells (MOI, 0.1) in the presence or absence of 800 μM or 1,600 μM ribavirin. The resulting viral titers were determined by TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 standard error of the mean (SEM). (B) The ribavirin susceptibility ratio was calculated as the viral titer in the absence of ribavirin divided by the titer in the presence of ribavirin, based on the final titers shown in panel A. The experiment was performed in triplicate, and each error bar indicates 1 SEM.

Single-step growth kinetics of the 3D-G64 mutants.

Single-step growth kinetics studies were performed in order to determine if the higher titer growth of the G64R and G64T polymerase variants in RD cells in the presence of ribavirin compared to that of the parental virus was due either wholly or partly to increased replicative capacity. RD cells were infected (MOI, 10), and virus growth was measured over a 24-h time course. The growth kinetics in the RD cells of G64N, G64R, and G64T mutants were compared to those of the parental virus (Fig. 2).

Fig 2.

Comparison of the single-step growth properties of G64N, G64R, G64T, and HEV71-26M in RD cells. RD cell monolayers in 48-well tissue culture trays were infected with G64N, G64R, G64T, or 26M (MOI, 10). The viruses were harvested at four-hour intervals for 24 h, and yields were quantified using a TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 SEM. The differences between the growth curves were assessed using one-way ANOVA and Tukey's multiple-comparison tests. The differences in the single-step growth titers of the G64R and 26M populations were not found to be significant (P > 0.05). By contrast, a significant difference in growth was observed between the parental 26M and G64T populations (P < 0.05). A significant difference in growth was also observed between the parental 26M and G64N populations (P < 0.05). (Data were analyzed using the GraphPad Prism online software package, version 5.04.

In the absence of ribavirin pressure, G64R mutant and parental virus populations grew efficiently in RD cells with no statistically significant differences (P > 0.05) observed between their single-step growth kinetics (Fig. 2). By contrast, both G64T and G64N replicated poorly in RD cells, with maximum titers approximately 10- to 30-fold below that of the parental virus (P < 0.05). Thus, the 3D-G64R mutation does not significantly impact the growth of the G64R CDV population in RD cells, whereas the 3D-G64N and 3D-G64T mutations significantly impact the growth of the G64N and G64T mutant CDV populations in RD cells.

Genotypic and phenotypic analyses of cloned populations of parental virus and the 3D-G64N, 3D-G64R, and 3D-G64T mutants after passages in the presence of ribavirin.

Previous studies have shown that ribavirin resistance in poliovirus is associated with the selection of high replication fidelity 3D polymerase mutants (10). Poliovirus 3D polymerase replication fidelity variants have been used increasingly to study the role of genome diversity in viral fitness and virulence (10, 15). In order to investigate the replication fidelity of the HEV71 3D polymerase, we examined the effects of serial passage of the parental and mutant CDVs in the presence of ribavirin on mutagenesis within the viral genome.

Three plaque-purified virus populations from independently derived stocks of G64N, G64R, G64T, and 26M were passaged (13×) in RD cells in the presence of 400 μM ribavirin. The plaque-purified populations are designated 26M1 to 26M3, G64N1 to G64N3, G64R1 to G64R3, and G64T1 to G64T3, respectively. Following the passages, the nucleotide sequences of the 2C and 3D coding regions of the three plaque-purified populations of G64N, G64R, and G64T and the parental virus were compared at RD cell passages 1 and 13.

Nucleotide sequences of the 2C and 3D coding regions of parental virus and the 3D-G64N, 3D-G64R, and 3D-G64T mutants after passage in the presence of ribavirin.

In the nucleotide sequences of the 2C and 3D coding regions of the 26M1 to 26M3, G64N1 to G64N3, G64R1 to G64R3, and G64T1 to G64T3 CDV populations at RD cell passages 1 and 13, no differences from that of the parental virus (17) were observed in these coding regions in any of the plaque-purified virus populations at RD passage 1 (Table 2).

Table 2.

Nucleotide and deduced amino acid sequence mutations identified in the 2C and 3D coding regions of parental and 3D-G64 mutant virus populations after 1 or 13 passages in RD cells in the presence of the mutagen ribavirin (400 μM)

Population	Nucleotide			Amino acid			BLOSUM62a (similarity score)
	Positionb	Amino acid sequence mutation at:		Position	Amino acid sequence mutation at:
		Passage 1	Passage 13		Passage 1	Passage 13
26M	4604	T	A	2C-175	M	K	−1
	4633 (2)	A	C	2C-185	K	R	2
	4634 (2)	A	G	2C-185	K	R	2
	4974	G	T	2C-298	K	N	0
	4995	G	A	2C-305	T	T	4
	6048 (2)	A	G	3D-36	G	G	6
	6130	G	C	3D-64	G	R	−2
	6500	T	C	3D-187	L	L	4
	6532	T	C	3D-198	F	L	0
	6730–6731	TC	CT	3D-264	S	L	−2
G64N	4638 (2)	T	C	2C-186	D	D	6
	4894	G	A	2C-272	E	K	1
	4896	G	A	2C-272	E	K	1
	4958	G	A	2C-293	R	K	2
	6048	A	G	3D-36	G	G	6
	6071	A	G	3D-44	H	R	0
	6074	G	C	3D-45	S	T	1
	6108	G	C	3D-56	Q	H	0
	6121–6122	AA	CC	3D-61	K	P	−1
	6172 (2)	G	C	3D-78	A	P	−1
	6429 (2)	C	A	3D-163	R	R	5
G64T	6448 (2)	G	A	3D-169	K	K	5
	6451	G	T	3D-170	K	N	0
G64R	6981	G	C	3D-347	T	T	4
S264L	6225	G	C	3D-95	M	I	1

^aBLOSUM62 is an amino acid substitution matrix for pairwise protein sequence alignments used to analyze the similarities between amino acid residues. BLOSUM62 matrices with high scores are associated with closely related residues. In contrast, low scores are associated with distantly related residues (39).

^bValues in this column in parentheses represent the number of repetitions for each mutation observed out of the total number of nucleotides sequenced.

At RD passage 13, nucleotide changes were observed in the 3D coding regions of the 26M1, 26M2, and 26M3; G64N1, G64N2, and G64N3; G64T1 and G64T2; and G64R1 populations (Table 2). Fewer mutations were present in the 3D coding regions of G64R and G64T than in those of 26M and G64N at RD passage 13. Interestingly, the plaque-purified 26M1 population had acquired a 3D-G64R mutation at RD cell passage 13, indicating that the selection of ribavirin resistance had occurred during passage in the presence of 400 μM ribavirin.

Five nucleotide changes were detected in coding region 2C of the parental virus populations, and four changes were detected in the G64N populations after 13 passages in the presence of 400 μM ribavirin (Table 2). In contrast, no mutations were identified in the 2C coding regions of the G64R and G64T mutant populations after 13 passages in the presence of ribavirin.

Susceptibility of parental virus and 3D-G64N, 3D-G64R, and 3D-G64T mutants to ribavirin after passage in the presence of ribavirin.

We next examined the susceptibility of each of the plaque-purified virus populations to ribavirin. The 26M1 to 26M3, G64N1 to G64N3, G64T1 to G64T2, and G64R1 CDVs at RD passage levels 1 and 13 were grown in RD cells for 24 h in the presence or absence of 800 μM and 1,600 μM ribavirin, respectively. As expected, the G64R1, G64T1, and G64T2 populations were resistant to growth in the presence of 1,600 μM ribavirin at RD passages 1 and 13 (Fig. 3). At RD passage 13, the 26M1 and 26M2 populations had acquired the ability to grow in the presence of 1,600 μM ribavirin, which was associated with single amino acid changes in coding region 3D at positions 64 (G→R) and 264 (S→L), respectively. By contrast, growth of the 26M3 and G64N1 to G64N3 populations was inhibited in the presence of 800 μM ribavirin at RD passages 1 and 13 (Fig. 3). These findings confirm the role of the 3D-G64R mutation in the ribavirin-resistant phenotype of HEV71 and also identify a new 3D mutation (S264L) that confers ribavirin resistance upon HEV71. By contrast, the new mutations identified in the 3D coding regions of the 26M3, G64N1 to G64N3, G64T1 to G64T2, and G64R1 populations at RD passage 13 were not associated with ribavirin resistance.

Fig 3.

Characterization of the ribavirin-resistant phenotype of plaque-purified populations of parental virus and the 3D-G64 polymerase mutants. (A) Titers of 26M1-3 (passages RD1 and RD13), G64N1-3 (passages RD1 and RD13), G64R1 (passages RD1 and RD13), and G64T1-2 (passages RD1 and RD13) (MOI, 0.1) after 24 h of growth in RD cells in the presence or absence of 800 μM or 1,600 μM ribavirin. The resulting viral titers were determined using a TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 SEM. (B) The ribavirin susceptibility ratio was calculated as the viral titer in the absence of ribavirin divided by the titer in the presence of ribavirin, based on the final titers shown in panel A. The experiment was performed in triplicate, and each error bar indicates 1 SEM.

Susceptibility of parental virus and 3D-G64N, 3D-G64R, and 3D-G64T mutants to guanidine after passage in the presence of ribavirin.

In the poliovirus model, the rate of random incorporation of guanidine resistance mutations, located in gene 2C (16, 21), has been used to estimate the replication fidelity of ribavirin-resistant mutants (10, 13). We recently showed that resistance to guanidine in HEV71 is associated with several amino acid substitutions in coding region 2C (22). Thus, we developed a guanidine resistance assay to assess the replication fidelity of the parental virus and the G64N, G64R, and G64T mutant virus populations.

The sensitivities of seven plaque-purified populations of the parental and mutant viruses (26M1 to 26M7), G64N1 to G64N7, G64R1 to G64R7, and G64T1 to G64T7) to guanidine were compared after growth in RD cells for 48 h in the presence or absence of 0.5 mM guanidine, and the results are shown in Fig. 4. The plaque-purified parental virus populations were capable of higher growth rates in the presence of 0.5 mM guanidine (10- to 100-fold) than were the plaque-purified G64R1 to G64R7 and G64T1 to G64T7 populations (Fig. 4). Although there were variations between the plaque-purified populations of G64N1 to G64N7, on average, we observed no statistically significant differences in the development of guanidine resistance between the plaque-purified G64N populations and the parental virus. These data provide further evidence to support the hypothesis that G64R and G64T have higher replication fidelity during growth in RD cells than 26M and G64N, preventing the random incorporation of mutations in coding region 2C that confer a guanidine-resistant phenotype.

Fig 4.

Characterization of the guanidine-resistant phenotype in seven plaque-purified populations of parental virus and the 3D-G64 polymerase mutants. Seven plaque-purified populations of parental 26M1 to 26M7 (A), G64N1 to G64N7 (B), G64T1 to G64T7 (C), and G64R1 to G64R7 (D) were cultured for 48 h in RD cells (MOI, 0.1) in the presence or absence of 0.5 mM guanidine (Gua). The resulting viral titers were determined using a TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 SEM. The guanidine resistance ratios of G64N1 to G64N7 (E), G64T1 to G64T7 (F), and G64R1 to G64R7 (G) were calculated as the viral titer in the presence of guanidine divided by the titer in the absence of guanidine, based on the final titers shown in panels A and B, A and C, and A and D, respectively. The experiment was performed in triplicate, and each error bar indicates 1 SEM.

In order to examine more fully the functional roles of these mutations on the guanidine resistance and cell culture growth phenotypes of HEV71, we introduced the 2C-M175K, 2C-K185R, 2C-E272K, 2C-R293K, and 2C-K298N mutations identified in the plaque-purified 26M and G64N populations after 13 passages in the presence of 400 μM ribavirin (see Table 2) into the 26M infectious clone. The sensitivities of the M175K, K185R, E272K, R293K, and K298N CDVs to guanidine were compared to that of the 26M CDV population as measured by virus growth after 72 h in the presence of 0 mM or 4 mM guanidine (Fig. 5). In the presence of 4 mM guanidine, the M175K, K185R, E272K, R293K, and K298N populations grew efficiently in RD cells, with significantly (P < 0.05) higher growth phenotypes than the parental virus (Fig. 5). In contrast, no mutations were found in the 2C coding regions of any of the plaque-purified G64R and G64T mutant populations at RD passage 1 or 13 (Table 2). It is interesting to note that the M175K, K185R, E272K, R293K, and K298N mutations in coding region 2C have not been shown to confer guanidine resistance upon HEV71 (1) or upon other picornaviruses (16, 21–24).

Fig 5.

Guanidine resistance profiles of clone-derived HEV71 2C variants. (A) Titers of 26M, M175K, K185R, K298N, E272K, and R293K after 72 h of growth in RD cells in the presence or absence of 4 mM guanidine. (B) The guanidine resistance (GuR) frequency was calculated as the titer in the presence of guanidine divided by the titer in the absence of guanidine, based on the final titers shown in Fig. 4A. The experiment was performed in duplicate, and each error bar indicates 1 SEM. The differences in titers were assessed using one-way ANOVA and Tukey's multiple-comparison tests. Statistically significant differences in growth titers were identified between M175K, K185R, E272K, R293K, K298N, and 26M (P < 0.05). Statistical analyses were performed using the GraphPad Prism online software package, version 5.04.

Taken together, these data provide further confirmation of increased 3D polymerase replication fidelity in the G64R and G64T mutant CDV populations, resulting in a reduced likelihood of the incorporation of random mutations into coding region 2C that confer guanidine resistance during serial passages in the presence of ribavirin.

Phenotypic characterization of the 3D-S264L mutant.

The newly identified 3D-S264L mutation in the 3D coding region of 26M2 was also associated with ribavirin resistance (Fig. 3). In order to confirm this observation, we incorporated the S264L mutation into the 3D coding region of the 26M infectious clone and were able to rescue a viable CDV population (S264L). The S264L CDV population was able to grow in RD cells in the presence of 1,600 μM ribavirin, with titers approximately 10-fold lower than those in the absence of ribavirin (Fig. 6), confirming the role of the 3D-S264L mutation in the ribavirin-resistant phenotype of HEV71. Furthermore, we examined the guanidine sensitivity phenotype of seven plaque-purified S264L CDV populations after 13 passages in the presence of 400 μM ribavirin. The S264L1 to S264L7 CDV populations remained susceptible to guanidine at RD passage 13 (Fig. 7), suggesting that the S264L mutation confers a high-fidelity phenotype upon the HEV71 3D polymerase. Furthermore, no mutations were identified in the 2C coding regions of three randomly selected plaque-purified S264L CDV populations after 13 passages in the presence of ribavirin. A single nucleotide change was identified in coding region 3D of one plaque-purified S264L CDV population at RD passage 13 (Table 2).

Fig 6.

Characterization of the ribavirin resistance phenotype of 3D-S264L in RD cells. (A) Titers of S264L and 26M after 24 h of growth in RD cells (MOI, 0.1) in the presence or absence of 800 μM or 1,600 μM ribavirin. The resulting viral titers were determined using a TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 SEM. (B) The ribavirin susceptibility ratio was calculated as the viral titer in the absence of ribavirin divided by titer in the presence of ribavirin, based on the final titers shown in panel A. The experiment was performed in triplicate, and each error bar indicates 1 SEM.

Fig 7.

Characterization of the guanidine resistance phenotype of S264L in RD cells. (A) Seven individually plaque-purified populations of S264L (S264L1 to S264L7) were cultured for 48 h in RD cells (MOI, 0.1) in the presence or absence of 0.5 mM guanidine. The resulting viral titers were determined using a TCID₅₀ assay on Vero cells (19). The experiment was performed in triplicate, and each error bar indicates 1 SEM. (B) The guanidine resistance ratios of S264L1 to S264L7 and parental 26M1 to 26M7 were calculated as the viral titer in the presence of guanidine divided by the titer in the absence of guanidine, based on the final titers shown in Fig. 4A and panel A, respectively. The experiment was performed in triplicate, and each error bar indicates 1 SEM.

Estimation of genome mutation frequencies of the parental and G64N, G64R, G64T, and S264L viruses after passage in the presence of ribavirin.

We estimated the mutation frequencies in the G64N, G64R, G64T, S264L, and parental virus genomes after 13 passages in RD cells in the presence of 400 μM ribavirin. The mutation frequencies were estimated using nucleotide sequencing of the 2C and 3D coding regions of three randomly selected plaque-purified populations for each CDV (a total of 8,099 nucleotides sequenced for each CDV). We estimated that parental 26M had acquired an average of 12.81 mutations per genome (7,412 nucleotides) at RD passage 13. Interestingly, the G64N mutant genome was also highly diverse, having acquired an average of 13.72 mutations per genome (P > 0.05). By contrast, the G64R, G64T, and S264L populations had acquired averages of only 0.91 (P < 0.05), 2.74 (P < 0.05), and 0.91 (P < 0.05) mutations per genome, respectively (Table 3).

Table 3.

Mutation frequencies in parental and 3D-G64 mutant virus populations after 13 passages in RD cells in the presence of the mutagen ribavirin (400 μM)

Virus	Total no. of mutationsa	No. of mutations per genomeb	Statistical analyses (P value)c
26M	14	12.81
G64N	15	13.72	>0.05
G64R	1	0.91	<0.05
G64T	3	2.74	<0.05
S264L	1	0.91	<0.05

^aNumber of mutations observed (over the total number of nucleotides sequenced [8,099 nucleotides for each virus population]). The 2C and 3D coding regions of three plaque-purified populations of the parental and mutant viruses were sequenced.

^bThe mutation frequency of each 3D-G64 polymerase mutant is represented as the average number of changes from the parental 26M full-genome consensus sequence (the denominator is 7,412 nucleotides) (17).

^cThe differences between the numbers of mutations compared to the parental 26M population were assessed using the Student t test. Statistical significance was set at a P value of <0.05. The data were analyzed using the GraphPad Prism online software package, version 5.04.

DISCUSSION

Ribavirin is a nucleotide analogue that has antiviral activities against a variety of RNA (9, 12) and DNA (9) viruses, including HEV71 (11). Ribavirin has been shown to diminish viral protein translation and RNA replication via three activities: as a mutagen by incorporation into nascent viral RNA molecules (8–10, 12), by the inhibition of IMP dehydrogenase (IMPDH), which results in a decrease in intracellular GTP levels (25–27), and by immunomodulatory effects, including a switch in the T-helper cell phenotype from type 2 to type 1 (28–32). It has been suggested that during virus replication in newly synthesized genomes, a decrease in cellular GTP levels may increase the frequency of ribavirin incorporation in newly synthesized genomes during virus replication through competitive inhibition (9, 10, 12, 13). However, the mutagenic effect of ribavirin has been demonstrated to be the primary antiviral mechanism of ribavirin against RNA viruses, which results from the incorporation of ribavirin into viral RNA, leading to a decrease in the specific infectivity of viral RNA and, ultimately, to lethal mutagenesis (9).

The 3D coding region of the HEV71 genome is predicted to encode a 52-kDa polypeptide that functions as an RDRP. Genetic and biochemical data indicate that the 3D polymerase is involved in the synthesis of both plus- and minus-strand viral RNAs when using the coding region 2B peptide VPg as a primer (5, 33–35). The structure of the 3D polymerase conforms to a right-hand motif, with palm, fingers, and thumb subdomains (4, 7). The HEV71 3D polymerase contains six four-amino-acid sequence conserved motifs designated A through F. Although motifs A through D, which are positioned in the palm domain, are clearly characterized in all classes of polymerases, motif E in the palm domain and motif F in the fingers domain are unique in both their 3D polymerases and reverse transcriptases (7).

The active site of the 3D polymerase is situated on a three-stranded β-sheet scaffold located in the palm region, in which four conserved aspartate residues are thought to facilitate catalysis (6). The residues, Arg-174, Asp-233, Asp-238, Ser-288, Gly-289, Thr-293, Asn-297, Gly-327, Asp-328, and Asp-329, have been demonstrated to play important roles in active site interactions of the poliovirus 3D polymerase (4, 6). In the poliovirus 3D polymerase, the amino acid position 3D-G64 maps onto the “index” motif of the fingers domain, adjacent to the nucleotide-binding site (4, 36). Moreover, the residues Gly-1, Gly-64, Ala-239, and Leu-241 formed a tetrahedral hydrogen bond network with the buried N-terminal glycine residue in a pocket in the fingers domain (6, 37). Thus, it is likely that the interaction of G64 with the buried N terminus of 3D polymerase serves to position Asp-238 in the active site and results in a direct connection between G64 and nucleoside triphosphate (NTP)-driven active closure (6, 36). It has been shown that Asp-238 is an essential 3D polymerase active-site residue that interacts with the 2′-OH group of incoming nucleotides (36, 38). Notably, the introduction of bulky serine, valine, alanine, threonine, or leucine residues at position 3D-G64 in the poliovirus 3D polymerase has been shown to increase replication fidelity (10, 15), possibly by reducing the space available within the active site of the enzyme for base mispair formation (38).

To the best of our knowledge, this study provides the first demonstration of the selection of ribavirin-resistant variants of HEV71 that express high replication fidelity. We introduced seven mutations into the 3D polymerase coding region of our 26M infectious cDNA clone using site-directed mutagenesis in order to produce the G64A, G64E, G64L, G64N, G64P, G64S, and G64T clone-derived virus populations. Previous studies with poliovirus have determined that five alternative amino acid residues to the parental 3D-G64 (A, L, S, T, and V) generated viable and genetically stable virus populations (15, 20). It is interesting to note that the 3D-G64S mutation, which is genetically stable in poliovirus and produces a viable virus population (10, 15, 20), did not produce a genetically stable HEV71 population in this study and reverted to the parental (G) sequence at its first passage in RD cells. However, it is possible that the G64S mutation may have been retained in the HEV71 genome if further RD cell passages had been made in the presence of ribavirin.

The 3D-G64N and 3D-G64T mutations produced viable HEV71 populations that were found to be genetically stable through five passages in RD cells. By contrast, the 3D-G64P mutation did not produce a viable HEV71 CDV population. The 3D-G64E mutation was genetically unstable and changed to arginine (R) after the first passage in RD cells; the 3D-G64R mutation was found to produce a viable virus population that was genetically stable through 10 more passages in RD cells. The 3D-G64A, 3D-G64E, and 3D-G64L mutations were genetically unstable in HEV71 and reverted to the parental (G) sequence after a single passage in RD cells. Furthermore, the 3D-G64A mutation changed to encode tryptophan, which was occupied by a W/P mixed codon, during the early RD passages. It is interesting to note that the G64W mutant appeared to allow the nonviable G64P mutant to propagate in RD cell culture in a mixed population.

Given that the 3D-G64N, 3D-G64R, and 3D-G64T mutations yielded viable and genetically stable CDV populations, we chose to investigate the ribavirin resistance and replication fidelity phenotypes of these three 3D polymerase mutant CDVs. Growth of the parental 26M population in RD cell culture was strongly inhibited (50-fold) in the presence of 800 μM ribavirin. By contrast, the G64R and G64T mutant virus populations were capable of growth in cell culture in the presence of 1,600 μM ribavirin, with titers approximately 10-fold below those in the absence of ribavirin, demonstrating that these mutations are directly responsible for the resistance of HEV71 to growth inhibition in the presence of ribavirin. Interestingly, growth of the G64N mutant virus population in cell culture was inhibited in the presence of 1,600 μM ribavirin, with titers approximately 40-fold below that in the absence of ribavirin, which was similar to that of the parental virus. In addition, the single-step growth kinetics in RD cells of the G64R and parental 26M populations (in the absence of ribavirin) did not differ significantly. However, the single-step growth kinetics of the G64T and G64N mutant viruses in the absence of ribavirin were reduced approximately 10- to 30-fold (P < 0.05) compared to that in the parental virus. These findings indicate that the high replication fidelity phenotype of G64R did not exert a significant impact on virus growth in cell culture. By contrast, the 3D-G64N and 3D-G64T mutations exerted a significantly adverse effect on the growth phenotype of HEV71. The mechanism for the poor cell culture growth phenotype of these two mutant viruses is unclear.

A possible mechanism for the resistance of the 3D-G64R and 3D-G64T virus populations to ribavirin is that increased 3D polymerase replication fidelity reduces the error frequency that is induced by the presence of ribavirin during replication. Several point mutations were observed in the 3D coding regions of 26M, G64N, G64R, and G64T virus populations after 13 passages in RD cells in the presence of 400 μM ribavirin. The high-fidelity CDV populations G64R and G64T acquired significantly fewer mutations in the 3D coding region than did the parental 26M and G64N populations.

The acquisition of a guanidine-resistant phenotype during viral passage in the presence of 0.5 mM guanidine, a potent inhibitor of enterovirus replication (16, 21), was used to evaluate the replication fidelity in G64 polymerase mutants in this study. Guanidine resistance in picornaviruses is associated with the development of point mutations in coding region 2C (16, 21, 23, 24), including HEV71 (22). After 13 passages in RD cells in the presence of 400 μM ribavirin, we showed that the G64R and G64T virus populations grew poorly in the presence of 0.5 mM guanidine relative to growth of the G64N and parental virus populations. Furthermore, 26M and the G64N variant had acquired several point mutations in coding region 2C, whereas the 2C coding regions of the G64R and G64T variants were unchanged. Incorporation of the observed 2C mutations (M175K, K185R, E272K, R293K, and K298N) into the 26M infectious clone confirmed the role of these mutations in conferring guanidine resistance.

It is of interest that one of the parental virus populations, 26M1, had acquired the 3D-G64R mutation by RD passage 13, providing the first demonstration of the selection of the 3D-G64R mutation under ribavirin pressure; this provides further confirmation that 3D-G64R is a determinant of ribavirin resistance in HEV71. The 26M2 population had also acquired ribavirin resistance by RD passage 13 under ribavirin selection (400 μM), which was associated with the presence of a 3D-S264L mutation. The ability of the S264L CDV population to grow in RD cells in the presence of 1,600 μM ribavirin confirmed that this mutation also confers ribavirin resistance upon HEV71. Furthermore, after 13 passages in RD cells under ribavirin selection pressure, we showed that seven plaque-purified S264L CDV populations grew poorly in the presence of 0.5 mM guanidine relative to growth of the ribavirin-passaged parental virus populations. Nucleotide sequencing showed that the ribavirin-passaged S264L CDV populations had retained the parental virus sequence (17) in coding region 2C.

From these data, we conclude that the selection of ribavirin resistance had occurred during the passage of parental virus in the presence of ribavirin, in addition to ribavirin-induced random mutagenesis, which is indicated by the generation of guanidine resistance in the absence of guanidine selection pressure. Furthermore, it is possible that several of the amino acid mutations observed in the 2C and 3D coding regions may have been selected during prolonged passages in RD cell culture.

Our data show clearly that the 3D-G64R, 3D-G64T, and 3D-S264L mutations confer high replication fidelity upon HEV71, resulting in an average of 10-fold (G64R, S264L) and 3-fold (G64T) fewer mutations being introduced into the HEV71 genome under the pressure of ribavirin-induced mutagenesis relative to the number of mutations introduced in the parental virus. Taken together, our data support the hypothesis that three point mutations in the 3D polymerase coding region, G64→R, G64→T, and S264→L, are sufficient to confer a high replication fidelity phenotype upon HEV71.

In conclusion, this study provides the first evidence of the genetic control of HEV71 3D polymerase replication fidelity during viral replication in cell culture. This phenomenon has been well characterized in poliovirus (10, 15, 20) and in coxsackievirus B3 (37). The key function of the 3D-G64 residue as a copying fidelity checkpoint during viral RNA replication has also been well established (15). We have demonstrated that the mutations 3D-G64R, 3D-G64T, and 3D-S264L confer ribavirin resistance upon HEV71 by increasing the replication fidelity of the 3D polymerase. Knowledge of the replication fidelity phenotypes of the 3D-G64R, 3D-G64T, and 3D-S264L mutations may be of use in the preparation of candidate live-attenuated HEV71 vaccines, as the enhanced replication fidelity due to these mutations may help to improve the stability and safety of live-attenuated HEV71 vaccines.