Thermostability of the Foot-and-Mouth Disease Virus Capsid Is Modulated by Lethal and Viability-Restoring Compensatory Amino Acid Substitutions

Abstract

Infection by viruses depends on a balance between capsid stability and dynamics. This study investigated biologically and biotechnologically relevant aspects of the relationship in foot-and-mouth disease virus (FMDV) between capsid structure and thermostability and between thermostability and infectivity. In the FMDV capsid, a substantial number of amino acid side chains at the interfaces between pentameric subunits are charged at neutral pH. Here a mutational analysis revealed that the essential role for virus infection of most of the 8 tested charged groups is not related to substantial changes in capsid protein expression or processing or in capsid assembly or stability against a thermally induced dissociation into pentamers. However, the positively charged side chains of R2018 and H3141, located at the interpentamer interfaces close to the capsid 2-fold symmetry axes, were found to be critical both for virus infectivity and for keeping the capsid in a state of weak thermostability. A charge-restoring substitution (N2019H) that was repeatedly fixed during amplification of viral genomes carrying deleterious mutations reverted both the lethal and capsid-stabilizing effects of the substitution H3141A, leading to a double mutant virus with close to normal infectivity and thermolability. H3141A and other thermostabilizing substitutions had no detectable effect on capsid resistance to acid-induced dissociation into pentamers. The results suggest that FMDV infectivity requires limited local stability around the 2-fold axes at the interpentamer interfaces of the capsid. The implications for the mechanism of genome uncoating in FMDV and the development of thermostabilized vaccines against foot-and-mouth disease are discussed.IMPORTANCE This study provides novel insights into the little-known structural determinants of the balance between thermal stability and instability in the capsid of foot-and-mouth disease virus and into the relationship between capsid stability and virus infectivity. The results provide new guidelines for the development of thermostabilized empty capsid-based recombinant vaccines against foot-and-mouth disease, one of the economically most important animal diseases worldwide.

Keywords: capsid; foot-and-mouth disease virus; protein engineering; thermal stability; vaccine.

FIG 1

Structure of FMDV indicating residues at the interpentamer interfaces selected for analysis of capsid thermostability. (A) Scheme of the icosahedral FMDV capsid. A biological protomer consisting of VP1 (blue), VP2 (green), VP3 (red), and the internal polypeptide VP4 (not visible) is shown in color. Two neighboring pentamers are delineated in cyan or violet. The interface between these two pentamers is enclosed within a dashed rectangle. (B) Atomic structure of FMDV (C-S8c1) (37) in the same orientation shown in panel A. VP1, VP2, and VP3 are colored blue, green and red, respectively. (C). Wire frame atomic model of two neighboring pentamers corresponding to the color-contoured pentamers in panel A. The amino acid residues at the interpentamer interface whose role in capsid thermostability was analyzed here are represented using space-filling models and color coded as follows: yellow, D3195 and A2065; green, N2019; and red, other tested, charged residues. One of each pair of equivalent residues belonging to either pentamer has been labeled. (D, E) Close-up ribbon models of two regions around interpentamer interfaces. VP2 and VP3 are colored green and red, respectively. (D) The β-annulus at each capsid 3-fold axis (S3) joining 3 pentamers together. Residue E2011 is represented using a magenta space-filling model. (E) A region around the capsid 2-fold axis (S2) at the center of each interpentamer interface. At the right and left, two equivalent β-sheets (each formed by two β-strands from VP2 from one pentamer [green] and five β-strands from VP3 from the neighboring pentamer [red]) contribute to join the two pentamers together. At the center there are two equivalent α-helices (orange) from symmetry-related VP2 subunits across a capsid 2-fold axis (S2). Residues R2018, H3141, and N2019 are represented using magenta, yellow, and blue space-filling models, respectively.

FIG 2

Capsid protein expression and processing and capsid assembly. (A) Representative results of Western blot assays to compare the expression and processing of wt and mutant capsid protein VP3 (indicated by the corresponding labels). Lanes labeled M correspond to molecular weight markers. The positions of unprocessed polyprotein P1-2A and the processed product VP3 are indicated. (B) Percent yields of the mutant mature capsid protein VP3s relative to the yield of the wt control. For each mutant, the averaged value obtained from 2 independent experiments and the corresponding error bar (standard deviation) are indicated. (C) A representative example of the centrifugation analysis carried out to estimate relative capsid yields (in this case, the yield of mutant capsid E2011A is compared to that of the wt capsid, used as the internal control in the same experiment). The larger peaks correspond to assembled capsids; the minor peaks on the left correspond to free pentameric subunits. (D) Percent yields of different mutant assembled capsids relative to the yield of the wt control. For each mutant for which results are presented in panels B and D, the averaged value obtained from 2 independent experiments and the corresponding error bar (standard deviation) are indicated.

FIG 3

Electron micrographs of capsid preparations. (A) Capsids that had been extensively purified through method I. (B) Capsids that had been only partially purified. Arrows point to individual assembled capsids. (Insets) Enlarged images of single capsids to better assess size and shape. Scale bars are included.

FIG 4

Thermal stability of purified E2011A (A), R2018A (B), and H3141A (C) mutant empty capsids containing charged-to-neutral (Ala) substitutions at the interpentamer interfaces (filled circles) relative to that of the wt empty capsid control (open circles). The percentages of intact capsids remaining after different incubation times in a representative experiment at 42°C are indicated in each case. Data were fitted to exponential decay processes, and the dissociation rate constants were obtained. The gray curve was obtained using the control wt capsid. (D) Ratios between the rate constant obtained for each mutant (k_mut) and the rate constant obtained for the wt control (k_wt). Values are averages and standard deviations obtained from 2 (E2011A and R2018A) or 4 (H3141A) independent experiments.

FIG 5

Thermal stability of 5 additional partially purified mutant empty capsids containing charged-to-neutral (Ala) substitutions at the interpentamer interfaces relative to that of the wt empty capsid control. The percentages of intact capsids remaining after different incubation times in a representative experiment at 42°C are indicated in each case. For each mutant, the average value obtained from 2 independent experiments and the corresponding error bar (standard deviation) are indicated.

FIG 6

Thermal stability of purified mutant empty capsids containing the lethal charged-to-neutral substitution H3141A and/or the viability-restoring, charge-compensating substitution N2019H relative to that of the wt empty capsid control. (A) Percentage of intact capsids remaining after different incubation times at 42°C in a representative experiment. Open circles, wt capsids; filled triangles, mutant H3141A; filled inverted triangles, mutant N2019H; filled squares, mutant H3141A/N2019H. Data were fitted to exponential decay processes, and dissociation rate constants were obtained. The gray curve was obtained using the control wt capsids. (B) Ratios between the rate constant obtained for each mutant and the rate constant obtained for the wt control. Values are averages and standard deviations obtained from 3 (N2019H) or 4 (all other variants) independent experiments.

FIG 7

Thermal stability of purified mutant virions containing substitution N2019H alone (filled triangles) or together with the H3141A change (filled squares) relative to that of the wt virion control (open circles). The percentages of intact virions remaining after different incubation times at 42°C in a representative experiment are indicated. Data were fitted to exponential decay processes, and dissociation rate constants were obtained. The gray curve was obtained using the control wt capsids. The table indicates the ratios between the rate constant obtained for each mutant and the rate constant obtained for the wt control. Values are averages and standard deviations obtained from 2 or 3 independent experiments.

FIG 8

Thermal stability of purified mutant empty capsids containing virion-stabilizing, neutral-to-charged A2065H or charged-to-neutral D3195N substitutions relative to that of the wt empty capsid control. (A, B) The percentages of intact capsids remaining after different incubation times at 42°C in a representative experiment are indicated. Open circles, wt capsid; filled circles, A2065H mutant (A) or D3195A mutant (B). Data were fitted to exponential decay processes, and dissociation rate constants were obtained. The gray curve was obtained using the control wt capsids. (C) Ratios between the rate constant obtained for each mutant and the rate constant obtained for the wt control. Values are averages and standard deviations obtained from 2 independent experiments.

FIG 9

pH sensitivity to dissociation of mutant FMDV empty capsids containing the D3195N, A2065H, or H3141A substitutions relative to that of the wt empty capsid control. The percentages of intact capsids remaining after different incubation times in a representative experiment at pH 5.0 and 25°C are indicated in each case. For each mutant, the average value obtained from 2 independent experiments and the corresponding error bar (standard deviation) are indicated.