Thermal stabilization of enterovirus A 71 and production of antigenically stabilized empty capsids

Abstract

Enterovirus A71 (EVA71) infection can result in paralysis and may be fatal. In common with other picornaviruses, empty capsids are produced alongside infectious virions during the viral lifecycle. These empty capsids are antigenically indistinguishable from infectious virus, but at moderate temperatures they are converted to an expanded conformation. In the closely related poliovirus, native and expanded antigenic forms of particle have different long-term protective efficacies when used as vaccines. The native form provides long-lived protective immunity, while expanded capsids fail to generate immunological protection. Whether this is true for EVA71 remains to be determined. Here, we selected an antigenically stable EVA71 virus population using successive rounds of heating and passage and characterized the antigenic conversion of both virions and empty capsids. The mutations identified within the heated passaged virus were dispersed across the capsid, including at key sites associated with particle expansion. The data presented here indicate that the mutant sequence may be a useful resource to address the importance of antigenic conformation in EVA71 vaccines.

Keywords: EVA71; empty capsids; infantile paralysis; stabilization.

Fig. 1.

Selection of thermally resistant virus. (a) Schematic representation of the experimental process used to select thermally resistant EVA71. (b) To determine a suitable incubation temperature for thermal stressing while retaining the ability to recover infectious virus, a series of temperatures were used. Virus stocks at 10⁶ p.f.u. ml⁻¹ were incubated at the indicated temperature (x-axis) for 30 min and samples were titred by plaque assay, n=1. (c) To generate a virus population resistant to incubation at the selected temperature (52.5 °C), cycles of heating and recovery were carried out. Heated virus samples were titred by plaque assay and compared to an unheated virus control population, n=3 in duplicate, graphed mean±sem.

Fig. 2.

Location of mutations in the stabilized virus population. (a) A single protomer highlighted within the context of the full EVA71 capsid (left; PDB: 3VBS, [30]), with icosahedral axes indicated (5, fivefold; 3, threefold; 2, twofold). The protomer is shown in an expanded view (right) with the sites of mutations indicated. VP1, cyan; VP2, green; VP3, magenta; VP4, yellow. (b–e) Expanded views showing the sites of mutation indicated in (a), with selected side chains shown. Adjacent protomers are indicated by pale colours. In some cases, segments of the backbone were removed or made partially transparent to improve clarity.

Fig. 3.

Verification of thermal resistance. (a) Characterization of the thermal resistance of WT and stabilized virus populations after incubation at the temperatures indicated along the x-axis. (b) Comparison of the thermal resistance of stabilized virus and recombinantly produced stabilized virus after incubation at the temperatures indicated along the x-axis (UT, untreated). n=3 each in duplicate, graphed mean±sem.

Fig. 4.

Antigenic form of WT and stabilized ECs and virions. (a) Gradient-purified WT and recombinant-stabilized viral samples were assessed for the presence of VP0/VP2 by Western blot using mAb 979, n=3 (representative blots shown). (b, c) Fractions were assessed by sandwich ELISA for (b) expanded particles using mAb 979, (c) and native particles using a 16-2-2D scFv. For clarity, fractions are aligned across all panels with wild-type samples on the left and recombinant stabilized samples on the right. For each, n=3 in duplicate, graphed mean±sem, dotted line indicates the background of the assay, wells containing no antigen.

Fig. 5.

Thermal stability of WT and stabilized ECs and virions. Samples of virus and EC were heated to a range of temperatures to induce antigenic conversion. Graphed HAg-reactive (a) EC and (b) virus, and NAg-reactive (c) EC and (d) virus. HAg assays use mAb 979 and NAg assays use the 16-2-2D scFv in the detection phase of a sandwich ELISA. Graphed mean±sem normalized to 2× blank well OD, n=3 each in duplicate. WT, wild-type; rs, recombinant stabilized.