Enterovirus A71 Containing Codon-Deoptimized VP1 and High-Fidelity Polymerase as Next-Generation Vaccine Candidate

Abstract

Enterovirus A71 (EV-A71) is a major pathogen that causes hand-foot-and-mouth disease (HFMD), which occasionally results in severe neurological complications. In this study, we developed four EV-A71 (rgEV-A71) strains by reverse genetics procedures as possible vaccine candidates. The four rgEV-A71 viruses contained various codon-deoptimized VP1 capsid proteins (VP1-CD) and showed replication rates and antigenicity similar to that of the wild-type virus, while a fifth virus, rg4643C4VP-CD, was unable to form plaques but was still able to be examined by median tissue culture infectious dose (TCID₅₀) titers, which were similar to those of the others, indicating the effect of CD on plaque formation. However, the genome stability showed that there were some mutations which appeared during just one passage of the VP1-CD viruses. Thus, we further constructed VP1-CD rgEV-A71 containing high-fidelity determinants in 3D polymerase (CD-HF), and the number of mutations in CD-HF rgEV-A71 was shown to have decreased. The CD-HF viruses showed less virulence than the parental strain in a mouse infection model. After 14 days postimmunization, antibody titers had increased in mice infected with CD-HF viruses. The mouse antisera showed similar neutralizing antibody titers against various CD-HF viruses and different genotypes of EV-A71. The study demonstrates the proof of concept that VP1 codon deoptimization combined with high-fidelity 3D polymerase decreased EV-A71 mutations and virulence in mice but retained their antigenicity, indicating it is a good candidate for next-generation EV-A71 vaccine development.IMPORTANCE EV-A71 can cause severe neurological diseases with fatality in infants and young children, but there are still no effective drugs to date. Here, we developed a novel vaccine strategy with the combination of CD and HF substitutions to generate the genetically stable reverse genetics virus. We found that CD combined with HF polymerase decreased the virulence but maintained the antigenicity of the virus. This work demonstrated the simultaneous introduction of CD genome sequences and HF substitutions as a potential new strategy to develop attenuated vaccine seed virus. Our work provides insight into the development of a low-virulence candidate vaccine virus through a series of genetic editing of virus sequences while maintaining its antigenicity and genome stability, which will provide an additional strategy for next-generation vaccine development of EV-A71.

Keywords: codon deoptimization; enterovirus A71; high fidelity; reverse genetics virus; vaccine.

FIG 1

Construction of VP1 codon-deoptimized EV-A71. (a) 4643/TW/98 (4643Y12) infectious clone was used as the backbone. The capsid region of genotype C4 was synthesized and swapped into 4643 (genotype C2) and designated EV-A71 4643C4VP (parental clone). Different scales of codon-deoptimized VP1 fragments were introduced into 4643C4VP to generate VP1 codon-deoptimized EV-A71. The white area is the wild-type region (4643), the light gray area is the capsid region from genotype C4, and the black area is the codon-deoptimized region. (b) Using the codon-deoptimized viruses as the template, 3D-G64R and 3D-L123F were introduced into the viral genome through site-directed mutagenesis. The two stars indicate the high-fidelity determinants of 3D-G64R and 3D-L123F.

FIG 2

Production of reverse genetics viruses in RD cells. RD cells were transfected with RNA derived from the infectious clones with various EV-A71 cDNAs and examined under an inverted light and fluorescence microscope after immunofluorescent staining the viruses with Mab979. Evans blue was used as the counterstain. All recombinant viruses were successfully produced in the transfected RD cells. (a) Native 4643, rg4643, and rg4643C4VP. (b) rg4643C4VP, rg4643C4VP-CD, rg4643C4VP-CD-C, and rg4643C4VP-CD-N. (c) rg4643C4VP-HF, rg4643C4VP-CD-HF, rg4643C4VP-CDC-HF, and rg4643C4VP-CDN-HF. rg, reverse genetics virus; CD, VP1 codon deoptimized; CD-C, VP1 C-terminal codon deoptimized; CD-N, VP1 N-terminal codon deoptimized; HF, high-fidelity viruses with G64R and L123F substitutions in 3D polymerase.

FIG 3

Growth kinetics of native 4643 and reverse genetics EV-A71. Growth kinetics were analyzed in RD and Vero cells at virus MOIs of 0.001 and 10. Virus growth was quantified by plaque assay. The closed triangle (blue) represents native 4643, the closed circle (red) represents reverse genetics 4643, and the closed square (green) represents rg4643C4VP, which contains capsid proteins of the C4 genotype.

FIG 4

Codon usage of EV-A71 4643/TW/98. (a and b) Codon usage of 4643/TW/98 polyprotein (a) and codon-deoptimized VP1 (b). The numbers represent the amounts of time the codon was used in the gene.

FIG 5

Growth kinetics of codon-deoptimized rgEV-A71 in vitro. Growth kinetics were analyzed in RD cells at virus MOIs of 0.001 (a), 0.01 (b), and 10 (c). Samples were collected after 1, 2, 3, and 4 days (a and b) or every 2 h postinfection (c) (x axis) and were quantified by plaque assay (y axis). The closed circle (blue) represents rg4643, the closed square (red) represents rg4643C4VP, the closed triangle (orange) represents rg4643C4VP-CDC, and the inverted triangle (green) represents rg4643C4VP-CDN. The experiment was performed in duplicates, and each error bar indicates 1 SD. The statistical differences between the growth kinetics were analyzed using two-way ANOVA (*, P < 0.05; **, P < 0.01; ***, P < 0.001). No statistical difference was seen. (d) Plaque morphology of each virus. (e) Viral growth kinetics at an MOI of 10. Samples were collected every 2 h postinfection (x axis) and were quantified by TCID₅₀ (y axis). The purple circle, orange square, green triangle, yellow inverted triangle, and pink diamond represent rg4643C4VP, rg4643C4VP-HF, rg4643C4VP-CD-HF, rg4643C4VP-CDC-HF, and rg4643C4VP-CDN-HF, respectively. The experiment was performed in triplicates, and each error bar indicates 1 SD.

FIG 6

Growth kinetics of high-fidelity codon-deoptimized rgEV-A71 in vitro. (a) Growth kinetics of viruses were analyzed in RD cells at an MOI of 0.001. Samples were collected 1, 2, 3, and 4 days postinfection (x axis) and were quantified by plaque assay (y axis). The closed circle represents rg4643, the closed square represents rg4643C4VP, the closed triangle represents rg4643C4VP-CDC, the inverted triangle represents rg4643C4VP-CDN, the open square represents rg4643C4VP-HF, the open triangle represents rg4643C4VP-CDC-HF, and the open inverted triangle represents rg4643C4VP-CDN-HF. The experiment was performed in duplicates, and each error bar indicates 1 SD. The statistical differences between the growth kinetics were analyzed using two-way ANOVA, and no significant difference was identified (*, P < 0.05; **, P < 0.01; ***, P < 0.001). (b) Plaque morphology of CD-HF viruses.

FIG 7

Neutralizing antibody titers of sera generated after immunization with various CD-HF viruses. BALB/c mice were immunized with CD-HF viruses. The sera were collected 14, 28, 45, and 60 days postimmunization. The data were assessed using two-way ANOVA, and no significant difference of neutralizing antibody generated from different CD-HF viruses was observed at any time point (P > 0.05). Mean values and one standard deviation are shown.

FIG 8

Body weight change of the C57BL/6 mice infected with codon-deoptimized high-fidelity rgEV-A71 viruses. Three-day-old C57BL/6 mice were infected with 100 μl of 10⁶-TCID₅₀ codon-deoptimized high-fidelity viruses via the intraperitoneal (i.p.) route, and rg4643C4VP was also included as a parental control. The open circle represents the mock infection group, the closed square represents rg4643C4VP, the closed triangle represents rg4643C4VP-HF, the inverted triangle represents rg4643C4VP-CD-HF, the diamond represents rg4643C4VP-CDC-HF, and the closed circle represents rg4643C4VP-CDN-HF. The data are shown as mean values and one standard deviation.

FIG 9

Body weight change, clinical score, and survival rate of the ICR mice infected with various codon-deoptimized high-fidelity rgEV-A71 viruses. Three-day-old ICR mice were infected with 100 μl of 10⁶-TCID₅₀ codon-deoptimized high-fidelity rgEV-A71 viruses via the intraperitoneal (i.p.) route. The virulent strain rgEV-A71-MP4 was also included as a control. The data are shown as body weight (a), clinical score (b), and survival rate (c). The close circle represents the mock infection group, the closed square represents rgEV-A71-MP4, the closed triangle represents rg4643C4VP-HF, the inverted triangle represents rg4643C4VP-CD-HF, the diamond represents rg4643C4VP-CDC-HF, and the open circle represents rg4643C4VP-CDN-HF. The data are shown as mean values and one standard deviation.

FIG 10

Virus load of viruses collected from organs and tissues of mice infected with rgEV-A71 and CD-HF viruses. (a to c) Viruses were collected from the intestine, brain, spinal cord, and muscle of mice infected with rEV-A71-MP4, r4643C4VP-HF, r4643C4VP-CD-HF, r4643C4VP-CDN-HF, and r4643C4VP-CDC-HF, represented as red, yellow, green, blue, and black, respectively. Each point represents a single sample. The organs and tissues were extracted 2 days postinfection (a), 3 days postinfection (b), and 6 days postinfection (c). Viral titers were calculated using plaque assay (PFU/mg) (y axis). The data are shown as mean values and one standard deviation.

FIG 11

Codon usage of human genome sequence. The codon usage of humans (Homo sapiens) was analyzed using the Kazusa database. The numbers represent the number of times each codon was used.