Increasing Type 1 Poliovirus Capsid Stability by Thermal Selection

Abstract

Poliomyelitis is a highly infectious disease caused by poliovirus (PV). It can result in paralysis and may be fatal. Integrated global immunization programs using live-attenuated oral (OPV) and/or inactivated (IPV) PV vaccines have systematically reduced its spread and paved the way for eradication. Immunization will continue posteradication to ensure against reintroduction of the disease, but there are biosafety concerns for both OPV and IPV. They could be addressed by the production and use of virus-free virus-like particle (VLP) vaccines that mimic the "empty" capsids (ECs) normally produced in viral infection. Although ECs are antigenically indistinguishable from mature virus particles, they are less stable and readily convert into an alternative conformation unsuitable for vaccine purposes. Stabilized ECs, expressed recombinantly as VLPs, could be ideal candidate vaccines for a polio-free world. However, although genome-free PV ECs have been expressed as VLPs in a variety of systems, their inherent antigenic instability has proved a barrier to further development. In this study, we selected thermally stable ECs of type 1 PV (PV-1). The ECs are antigenically stable at temperatures above the conversion temperature of wild-type (wt) virions. We have identified mutations on the capsid surface and in internal networks that are responsible for EC stability. With reference to the capsid structure, we speculate on the roles of these residues in capsid stability and postulate that such stabilized VLPs could be used as novel vaccines.

Importance: Poliomyelitis is a highly infectious disease caused by PV and is on the verge of eradication. There are biosafety concerns about reintroduction of the disease from current vaccines that require live virus for production. Recombinantly expressed virus-like particles (VLPs) could address these inherent problems. However, the genome-free capsids (ECs) of wt PV are unstable and readily change antigenicity to a form not suitable as a vaccine. Here, we demonstrate that the ECs of type 1 PV can be stabilized by selecting heat-resistant viruses. Our data show that some capsid mutations stabilize the ECs and could be applied as candidates to synthesize stable VLPs as future genome-free poliovirus vaccines.

Keywords: VLP; heat stable; poliovirus.

FIG 1

Poliovirus genome organization and morphogenesis. (A) Genomic structure of PV-1 showing structural protein genes in blue (VP1), green (VP2), red (VP3), and yellow (VP4). (B) Ribbon model of a PV protomer showing relative positions of the capsid proteins. (C) Cartoon model of the PV capsid, with a pentamer outlined. Five protomers assemble into a pentamer, 60 of which form a procapsid. Virion maturation involves cleavage of VP0 into VP2 and VP4. The diagrams are not to scale.

FIG 2

Evolution of heat-resistant PV through repeat cycles of selection and passage. (A) Wt PV-1 was incubated at a range of temperatures for 30 min and cooled to 4°C. The surviving pool of viruses was evaluated by plaque assay on HeLa cells. (B) Wt PV-1 was incubated at 51°C for 30 min, which resulted in 99.99% loss of infectivity, and immediately cooled to 4°C. The surviving pool of viruses was passaged at 37°C. After each cycle of passage, virus titers (pre- and postheating) were determined by plaque assays. The selection cycles were repeated until the pre- and postheating virus titers were approximately equal (n = 3 ± standard deviation [SD]; *, P < 0.05; **, P < 0.001; ****, P < 0.00001). (C) After 10 cycles of thermal selection at 51°C and passage at 37°C, thermal pressure was increased to 53°C with 12 successive passages at 37°C. The pre- and postheating titers were statistically different from passage 0 until passages 9 and 11 (n = 3 ± SD; *, P < 0.05; **, P < 0.001; ***, P < 0.0001). (D) After selection at 53°C, thermal selection pressure was subsequently increased to 57°C with 10 successive passages. The pre- and postheating titers were statistically different from passage 0 until passage 10 (n = 3 ± SD; *, P < 0.05; **, P < 0.001; ***, P < 0.0001; ****, P < 0.00001). Three titrations of the same selected pool were analyzed at each temperature.

FIG 3

Thermal-resistance profile of heat-selected virus pools. (A) Pools of PV-1 selected at 51°C (VS51), 53°C (VS53), and 57°C (VS57) were incubated at a range of temperatures between 37°C and 60°C for 30 min and immediately cooled to 4°C. Titers were determined by plaque assays on HeLa cells. The data represent titers at each temperature (n = 3 ± SD; P < 0.0001). Wt PV-1 and thermally selected purified virus samples were examined by differential scanning fluorometric assays (PaSTRy) using SYTO9 nucleic acid-binding dye and SYPRO orange protein-binding dye as described by Walter et al. (36). (B) Relative fluorescence of SYTO9 (n = 3 ± SD; P < 0.0001). (C) Relative fluorescence of SYPRO orange (n = 3 ± SD; P < 0.001). The error bars in panels B and C were omitted for clarity. AU, arbitrary units.

FIG 4

Generation and purification of virions and ECs. HeLa cells were infected with the wt, VS51, VS53, and VS57. Virus capsid proteins were radiolabeled with [³⁵S]Cys/Met at 2.75 h postinfection and incubated at 37°C. For EC preparation, 2 mM GuHCl was added at 3.15 h postinfection. The harvested particles were purified by ultracentrifugation. (A) Scintillation counts of fractions of 15 to 45% sucrose density gradients of the wt, VS51, VS53, and VS57. The gradients were fractionated into 300-μl aliquots, and 3% of each fraction was counted by scintillation. (B) Autoradiograph of proteins from virion and EC sucrose density gradient peak fractions of the wt, VS51, VS53, and VS57. Samples from virion and EC peaks were boiled in SDS loading buffer. Proteins were separated by SDS-PAGE, and [³⁵S]Cys/Met-radiolabeled virus bands were detected by autoradiography. (C) Electron micrographs of virion and EC peaks of the wt. Virion and EC peaks were harvested and dialyzed into 10 mM HEPES, 100 mM NaCl, 1 mM EDTA, 1 mM DTT, pH 6.4, and visualized by negative-staining transmission electron microscopy.

FIG 5

Antigenic-stability profile of heat-selected virus pools. Pools of thermally selected PV-1 at 51°C (VS51), 53°C (VS53), and 57°C (VS57) were radiolabeled with [³⁵S]Cys/Met and purified by differential and sucrose gradient centrifugation. Virion and EC peaks were identified by scintillation counting. Virion and EC fractions were harvested and incubated at a range of temperatures for 30 min and then cooled to 4°C. Heat-treated N-specific particles were immunoprecipitated using MAb 234. (A) Normalized scintillation counts of heat-treated virions (n = 3 ± SD; P < 0.05). (B) Normalized scintillation counts of heat-treated EC particles (n = 3 ± SD). It was not possible to compare values to the wt, as they convert completely to the H/C form at 10°C. (C) The temperatures at which 50% of the N/D antigenicity was converted to the H/C form were estimated for virion particles of selected viruses (VS51, VS53, and VS57) compared to the wt (*, P < 0.05; **, P < 0.001; ***, P < 0.0001; ns, not significant).

FIG 6

Positions of common mutations. (A) Ribbon model of a PV protomer. (B) Cartoon model of PV capsid, with VP1 (blue), VP2 (green), VP3 (red), and VP4 (yellow). The positions of VP1 mutations (1, V1087A; 2, I1194V; 3, A1026T; and 8, S1097P), a VP3 mutation (5, C3175A), and VP4 mutations (4, F4046L; 6, R4034S; and 7, D4045V) are indicated. Most mutations occurred at residues buried within the capsid. The diagrams are not drawn to scale.

FIG 7

Characterization of pocket mutations V1087A and I1194V. Two common mutations that occurred in all mutants were reintroduced by site-directed mutagenesis into an infectious clone of wt PV-1 individually and in combination (PV51). Viral RNA was generated by T7 transcription. A total of 5 × 10⁷ L cells were transfected with 5 μg RNA and incubated at 37°C. Virus proteins were radiolabeled with [³⁵S]Cys/Met and purified by differential and sucrose gradient centrifugation. (A) Virus titers were determined by plaque assays on HeLa cells (n = 3 + SD; *, P < 0.05). (B) Virions were immunoprecipitated with N/D-specific (MAb 234) and H/C-specific (MAb 1588) monoclonal antibodies and counted by scintillation. The sum of the radioactivity in immunoprecipitated particles from 30% of virus peak fractions is presented and compared to that of the wt (n = 3 + SD; *, P < 0.05; **, P < 0.001; ***, P < 0.0001). (C and D) Transfected cell lysates were blotted against anti-3CD (C) and polyclonal P1 antibody (SH-16) (D). The numbers on the left are in kilodaltons.

FIG 8

Infectivity of SDM-constructed mutants. Mutant viruses were constructed individually and in combinations by SDM, with and without I1194V. PV51δ (V1087A), PV53δ (A1026T, V1087A, F4046L), and PV57δ (V1087A, S1097P, C3175A, R4034S, D4045V) were generated and compared with constructs in which I1194V was present, i.e., PV51 (V1087A, I1194V), PV53 (A1026T, V1087A, F4046L, I1194V), and PV57 (V1087A, S1097P, C3175A, R4034S, D4045V, I1194V). A total of 1 × 10⁷ mouse L cells were transfected with 5 μg of in vitro-transcribed RNA. Posttransfection, the supernatant was harvested and the cells were freeze-thawed. Samples were heat treated at a range of temperatures, and virus titers were determined by plaque assays using HeLa cells. (A) Normalized thermal-inactivation assays for SDM mutants with I1194V (n = 3 ± SD; P < 0.00001). (B) Normalized thermal-inactivation assays for SDM mutants without I1194V (n = 3 ± SD; P < 0.00001). (C) Input virus titers (n = 3 ± SD; *, P < 0.05; **, P < 0.001). (D) 50% inhibitory concentration [IC₅₀] temperatures for thermal-inactivation curves of mutants with and without I1194V (n = 3 ± SD; **, P < 0.001; ***, P < 0.0001).

FIG 9

Thermal-stability profiles of virus mutants. Mutations identified by thermal selection were introduced by SDM into an infectious clone of wt PV-1, individually and in combinations with and without the I1194V mutation. The wt was included for comparison. Transcribed RNAs were transfected into L cells to recover infectious virus, and ECs were accumulated by GuHCl treatment at 3.15 h posttransfection. The harvested particles were purified through 15 to 45% sucrose density gradients. The sucrose was removed by dialysis, and virions and ECs were dialyzed out of the sucrose and incubated at a range of temperatures (from 37°C to 60°C) for 30 min and cooled to 4°C for 5 min before titration by plaque assays using HeLa cells. (A) Immunoprecipitation assay for heat-treated virions. (B) Immunoprecipitation assay for heat-treated ECs using N/D-specific (MAb 234) monoclonal antibodies. (C) Immunoprecipitation assay for heat-treated BRP vaccine controls, using N/D-specific (MAb 234) monoclonal antibodies. Samples were immunoblotted with anti-VP1 Millipore MAb 8650. Shown are normalized averages of semiquantitative densitometric plots of VP1 band intensities using ImageJ (n = 3 ± SD). (D) Temperatures at which 50% native (N/D) antigenicity was lost. n = 3 + SD. For virus, the data are compared to the wt, and for EC, to BRP (**, P < 0.001; ****, P < 0.00001).