Involvement of a Nonstructural Protein in Poliovirus Capsid Assembly

Abstract

Virus capsid proteins must perform a number of roles. These include self-assembly and maintaining stability under challenging environmental conditions, while retaining the conformational flexibility necessary to uncoat and deliver the viral genome into a host cell. Fulfilling these roles could place conflicting constraints on the innate abilities encoded within the protein sequences. In a previous study, we identified a number of mutations within the capsid-coding sequence of poliovirus (PV) that were established in the population during selection for greater thermostability by sequential treatment at progressively higher temperatures. Two mutations in the VP1 protein acquired at an early stage were maintained throughout this selection procedure. One of these mutations prevented virion assembly when introduced into a wild-type (wt) infectious clone. Here we show, by sequencing beyond the capsid-coding region of the heat-selected virions, that two mutations had arisen within the coding region of the 2A protease. Both mutations were maintained throughout the selection process. Introduction of these mutations into a wt infectious clone by site-directed mutagenesis considerably reduced replication. However, they permitted a low level of assembly of infectious virions containing the otherwise lethal mutation in VP1. The 2A^pro mutations were further shown to slow the kinetics of viral polyprotein processing, and we suggest that this delay improves the correct folding of the mutant capsid precursor protein to permit virion assembly.IMPORTANCE RNA viruses, including poliovirus, evolve rapidly due to the error-prone nature of the polymerase enzymes involved in genome replication. Fixation of advantageous mutations may require the acquisition of complementary mutations which can act in concert to achieve a favorable phenotype. This study highlights a compensatory role of a nonstructural regulatory protein, 2A^pro, for an otherwise lethal mutation of the structural VP1 protein to facilitate increased thermal resistance. Studying how viruses respond to selection pressures is important for understanding mechanisms which underpin emergence of resistance and could be applied to the future development of antiviral agents and vaccines.

Keywords: 2Apro; cleavage; evolution; poliovirus; selection.

FIG 1

Identification of thermally selected mutations. (A) Cartoon structure of the PV genome, showing the 5′ and 3′ UTRs flanking the open reading frame comprising the structural (P1) and nonstructural (P2 and P3) regions. (B) Evolution of PV-1 under thermal selection. Viral RNA was extracted from each passage of virions selected at 51°C (i.e., VS51), 53°C (i.e., VS53), and 57°C (i.e., VS57) (25). The entire genome of the evolving population at each passage was reverse transcribed and amplified by PCR. The virus pool was sequenced and aligned against the wt PV-1 sequence by ClustalOmega. A cartoon representation of selected mutations is shown. Solid black vertical lines represent wt sequences of the P1 and 2A regions. Nonsynonymous mutations are presented as colored shapes in VP4, VP2, VP3, VP1, and 2A^pro, as shown in the key inset. (C) Sequence comparisons of 2A^pro among enteroviruses. Alignment of a 60-residue region of 2A^pro of thermally selected viruses VS51, VS53, and VS57 against enterovirus A (coxsackievirus A1 [CVA-1]), enterovirus B (echovirus 1 [EV1]), enterovirus C (PV-1, PV-2, PV-3, and CVA-2), enterovirus D: human enterovirus 94 (EV-94); enterovirus E: bovine enterovirus (BEV-1); enterovirus F (BEV-2), enterovirus G (porcine enterovirus G1 [EV-G1]), enterovirus H (simian enterovirus SV4 [SV4]), rhinovirus A (human rhinovirus A [HRV-A]), rhinovirus B (HRV-B), and rhinovirus C (HRV-C). Thermally selected virions VS51, VS53, and VS57 are underlined. Enterovirus C members are annotated with a bracket. The 2A^pro consensus residues of wt PV-1 are shown in bold. Matching residues are shown as dots beneath the corresponding residues of wt PV-1. Variable residues are shown underneath the corresponding positions of the wt PV-1 residues. Asterisks on the consensus sequence indicate positions that correspond to residues I99 and G102, respectively. Sequences were aligned using the default alignment algorithms of CLC sequencing viewer version 6.

FIG 2

Effect of 2A^pro mutations on cis-cleavage activity at the P1/2A junction. (A) Cartoon of the construct used in the transcription/translation (T_NT) assay. (B) Autoradiographs of SDS-PAGE of T_NT samples. Following incubation at 30°C for 90 min, further incorporation of [³⁵S]Cys-Met was prevented by the addition of excess unlabeled Cys-Met and samples collected at 30-minute intervals. Samples were separated by SDS-PAGE and protein bands detected by autoradiography. Arrows correspond to the P1/2A^pro precursor and processed P1. Time points represent chase. (C and D) Normalized densitometry of the P1/2A^pro precursor (C) and cleaved P1 (D) over time. Graphs represent the intensity of the P1 band of phosphoscreen scans of autoradiograph. Scanned images were analyzed by ImageJ version 1.47t (n = 2; error bars show standard errors of the means [SEM]; **, P < 0.001 compared to wt).

FIG 3

Polyprotein processing of wt and 2A^pro mutant PV-1. (A) Both 2A mutations were introduced into an infectious clone of wt PV-1 (i.e., pt7Rbz). RNA transcripts of pt7Rbz or 2A-I99V/-G102R were used in HeLa cell-free reactions. Following incubation at 34°C for 2 h, excess Cys-Met was added and samples taken at various time points. Samples (lanes 1 to 12) were separated by 8% SDS-PAGE and radiolabeled proteins detected by autoradiography. To identify specific bands, pcDNA constructs of P1, P2, noncleavable P1-2A, and noncleavable P1-P2 were expressed in T_NT reactions in the presence of [³⁵S]Cys-Met and incubated for 3 h (lanes 13 to 17). (B and C) Band intensities were quantified from a phosphoscreen image and the levels of P1-P2 (B) and P1-2A (C) presented as normalized percent intensity over background phosphorescence. Scanned images were analyzed by ImageJ version 1.47t (n = 3; error bars show SEM; *, P < 0.05; **, P < 0.001 [compared to wt at each time point]).

FIG 4

Effect of 2A^pro mutations on PV-1 replicon replication. (A) HeLa cells were transfected with T7 RNA transcripts and replication monitored by GFP fluorescence over time using an IncuCyte Zoom. A replication-deficient mutant, 3D-GNN, was included as a control for input translation. (B) The data from panel A at 22 h posttransfection (total GFP-positive cells) were also plotted as a bar graph for clarity (n = 3; error bars show SEM; *, P < 0.05; **, P < 0.001; ***, P < 0.0001).

FIG 5

Effects of 2A^pro mutations on virus recovery. RNA transcripts were generated from infectious clones of the wt or VP1-I194V in the presence or absence of the 2A^pro mutations 2A-I99V/-G102R, transfected into mouse L cells or HeLa cells, and incubated at 37°C for 24 h. (A) Virus titers recovered from transfected mouse L cells (n = 3; error bars show SEM; ***, P < 0.0001) (B) Virus titers recovered from transfected HeLa cells. (n = 3; error bars show SEM; ***, P < 0.0001) (C) Virus samples recovered from HeLa cells were diluted in serum-free medium to equal starting titers, incubated at a range of temperatures between 37°C and 55°C for 30 min, cooled to 4°C, and titrated by plaque assays using HeLa cells (n = 2; error bars show standard deviations [SD]; *, P < 0.05 compared to wt).

FIG 6

Effects of reduced translation on an assembly-deficient VP1-I194V mutant. (A) Mouse L cells were treated with increasing concentrations of cycloheximide and assayed for toxicity by using MTS. The IC₅₀ was evaluated by using dose-dependent curves (n = 2; error bars show SD). (B) Mouse L cells were transfected with T7 RNA transcripts of the wt and treated with cycloheximide at increasing concentrations. Cell lysates were harvested using RIPA buffer. Supernatants and cell lysates were separated by SDS-PAGE and immunoblotted against anti-PV-1 VP1 MAb 8560. Representative results from two biological repeats are shown. (C) Mouse L cells were transfected with T7 RNA transcripts of the wt and VP1-I194V mutant and treated at two concentrations of cycloheximide. Supernatants were clarified by low-speed centrifugation. Virus titers of supernatants from cycloheximide-treated wt- and VP1-I194V mutant-transfected cells were determined by plaque assays using HeLa cells (n = 2; error bars show SEM; *, P < 0.05; ***, P < 0.0001 [compared to untreated wt]).