Viable Polioviruses That Encode 2A Proteins with ... - Journal of Virology

JOURNAL OF VIROLOGY, Feb. 2010, p. 1477–1488 0022-538X/10/$12.00 doi:10.1128/JVI.01578-09

Vol. 84, No. 3

Viable Polioviruses That Encode 2A Proteins with Fluorescent Protein Tags䌤 Natalya L. Teterina, Eric A. Levenson,† and Ellie Ehrenfeld* Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892 Received 30 July 2009/Accepted 12 November 2009

The 2A proteins of the Picornaviridae enterovirus genus are small cysteine proteinases that catalyze essential cleavages in the viral polyprotein in cis and in several cellular proteins in trans. In addition, 2A has been implicated in the process of viral RNA replication, independent of its protease functions. We have generated viable polioviruses that encode 2A proteins containing fluorescent protein tag insertions at either of two sites in the 2A protein structure. Viruses containing an insertion of Discosoma sp. red fluorescent protein (DsRed) after residue 144 of 2A, near the C terminus, produced plaques only slightly smaller than wild-type (wt) virus. The polyprotein harboring the 2A-DsRed fusion protein was efficiently and accurately cleaved; fluorescent 2A proteinase retained protease activity in trans and supported translation and replication of viral RNA, both in vitro and in infected cells. Intracellular membrane reorganization to support viral RNA synthesis was indistinguishable from that induced by wt virus. Infected cells exhibited strong red fluorescence from expression of the 2A-DsRed fusion protein, and the progeny virus was stable for three to four passages, after which deletions within the DsRed coding sequence began to accumulate. Confocal microscopic imaging and analysis revealed a portion of 2A-DsRed in punctate foci concentrated in the perinuclear region that colocalized with replication protein 2C. The majority of 2A, however, was associated with an extensive structural matrix throughout the cytoplasm and was not released from infected cells permeabilized with digitonin. The picornavirus genome is a single-stranded, positive-sense RNA strand that is translated into a single polyprotein of ⬃250 kDa which is cleaved during translation to generate an Nterminal capsid protein precursor (P1) and a P2-3 precursor to the nonstructural proteins. Although genetic and some biochemical studies conducted over many years have implicated all the nonstructural proteins as having some role(s) in viral RNA synthesis, the precise biochemical functions of most of these proteins remain only vaguely defined. Picornaviral 2A protein sequences, defined by their position C-terminal to the capsid protein precursor in the viral polyprotein, are highly variable in size, sequence, and function among the different genera within the family Picornaviridae. In the enterovirus genus, the 2A protein is a small cysteine proteinase, homologous to trypsin-like small serine proteases, with a Cys residue substituting for the usual serine to form a catalytic triad in the active site of the protease. The 2A proteins of cardio- and aphthoviruses have no sequence similarity to the 2A proteins of enteroviruses or to any other known proteinases. While smaller in size, the aphthovirus 2A proteins are homologous to the C-terminal ends of 2A proteins from cardioviruses. Their coding sequence causes ribosomes to “skip” formation of a peptide bond at the junction of the 2A and downstream sequences, leading to production of two proteins from a single open reading frame (36). Other picornaviruses exhibit no identifiable activities in the 2A region. During translation of viral RNA, poliovirus (PV) 2Apro cat-

alyzes the cleavage of its own N terminus in cis, thereby releasing the capsid proteins in the P1 region from the nascent nonstructural proteins in the P2-3 region. A second 2A cleavage site in the polyprotein resides in the N-terminal portion of 3D. Cleavage of 3CD at this site generates two products, 3C⬘ and 3D⬘; however, removal of the 3CD scissile bond by mutagenesis had no effect on poliovirus replication in HeLa cells (19). In addition to cleavage of the viral polyprotein, 2Apro catalyzes specific cleavages in a small number of host cell proteins whose activities affect virus replication in different ways. Cleavage of the translational initiation factor eIF4G prevents translation of capped mRNAs, generating a C-terminal cleavage product of eIF4G that stimulates utilization of the poliovirus internal ribosome entry site (IRES) (10, 16, 21). Poly(A) binding protein (PABP) is also cleaved by 2Apro although this cleavage is preferentially targeted to PABP not associated with polysomes. 3Cpro is responsible for polysome-associated PABP cleavage, which appears to be responsible for synergistic inhibition of cap-dependent translation (9, 18, 35). 2Apro was found to increase the stability of viral RNA and to stimulate and prolong translation in vitro, independent of the RNA stabilizing effect (17). All of these 2A-induced events serve to enhance viral protein synthesis. In addition, poliovirus infection or 2A expression in HeLa cells causes alterations in the nuclear pore structure resulting from cleavage of specific components of the nuclear pore complex (8, 15, 33). These alterations produce bidirectional increases in permeability of the nuclear envelope, which permit redistribution of cellular proteins that normally reside in the nucleus to the cytoplasm, where they are available for viral translation or other replication reactions. Several genetic studies have implicated a direct role for 2A in viral RNA replication. Deletion of a C-terminal negatively

* Corresponding author. Mailing address: Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, NIH, 50 South Drive, Room 6120, Bethesda, MD 20892. Phone: (301) 5941654. Fax: (301) 435-6021. E-mail: [email protected]. † Present address: Department of Chemistry and Biochemistry, University of Delaware, Newark, DE 19716. 䌤 Published ahead of print on 25 November 2009. 1477

1478

TETERINA ET AL.

J. VIROL.

FIG. 1. (A) Schematic diagram of PV genomes with identified transprimer insertions in protein 2A coding sequences. Amino acids encoded by insertions are shown in bold; wild-type PV amino acids are underlined. Numbers indicated the amino acids between which the insertions are located. (B) Location of insertion sites on three-dimensional structure of protein 2A from coxsackie virus B4. (Protein Data Bank code 1Z8R) (5). NTR, nontranslated region.

charged cluster of amino acids present in poliovirus 2A is lethal to viral RNA replication without significantly affecting 2A protease function, and growth could be rescued by providing wildtype 2Apro in trans (20). Similarly, deletion of sequences coding for the first 45 amino acids of 2A greatly reduced replication of a subgenomic replicon in transfected cells; replication of this replicon was also rescued in trans when 2A was provided by a helper RNA (12). In another study using a dicistronic viral RNA with an encephalomyocarditis virus (EMCV) IRES inserted between the P1 and P2 coding regions so that the cis cleavage function of 2A was not required to generate viral proteins, deletion of 2Apro coding sequences rendered the transcripts incapable of replication (27). Insertion of an IRES between the 2A and 2B genes generated a small-plaque virus, indicating that neither the intact P2 polypeptide nor the 2AB cleavage fragment of P2 is required for virus growth. Thus, although little insight has been provided into the possible function(s) of 2A in the mechanism of viral RNA replication, several studies have suggested that 2A contributes some essential role to this process. In an effort to develop new approaches to studying the potential role of polio 2A in viral RNA replication, we generated viable polioviruses that encode 2A proteins tagged with a fluorescent protein (FP), inserted at either of two different positions in 2A. We report here the characterization and properties of these viruses, and we describe for the first time the intracellular location and distribution of protein 2A during the course of virus infection in HeLa cells. MATERIALS AND METHODS Construction of plasmids and DNA manipulation. pPV-2A50-tp and pPV2A144-tp plasmids carrying the full-length genome of PV with insertions of 15 nucleotides (nt) after nt 3535 or 3817 that encoded in-frame insertions of 5 amino acids (aa) after aa 50 or 144 of PV protein 2A, respectively (Fig. 1A), were constructed using standard cloning techniques. The insertion sites were selected from a random library of viable transposon insertion mutants of PV (to be published elsewhere). Discosoma sp. red fluorescent protein (DsRed) or Aequorea coerulescens GFP (AcGFP) coding sequences were produced by PCR using pDsRed-Monomer-N1 or pAcGFP-N1 (Clontech) as a template. Plasmid

pPVM-2A50-DsRed contained insertion of the entire DsRed coding sequence after nt 3538 (the last nucleotide in the codon for aa 51 in protein 2A) of the PV genome. This plasmid also encodes two extra amino acids preceding DsRed and one extra amino acid on the C terminus of DsRed, introduced during cloning of this insert, followed by duplication of PV 2A aa 50 and 51 (Fig. 2). Plasmids pPVM-2A144-DsRed and pPVM-2A144-GFP encoded the PV genome with the insertion of DsRed or GFP coding sequences in the PmeI site of the initial 15-nt insertion contained in pPV-2A144-tp from the transprimer, with an additional 9 nt coding for three additional spacer amino acids on each side. A schematic representation of all recombinant PV genomes is shown in Fig. 2. Sequences of all the recombinant genomes will be provided upon request. The pXpA-RenR replicon has been described previously (6). pXpA-Ren-2A144-DsRed and pXpA-Ren-2A144-GFP were produced by conventional subcloning from pPVM2A144-DsRed and pPVM-2A144-GFP, respectively. RNA transcription and transfection. Plasmids were linearized at the EcoRI restriction site located downstream of the PV poly(A) sequence prior to transcription with T7 RNA polymerase in vitro (39). HeLa cell monolayers in six-well plates were transfected with serial dilutions of RNA transcripts using a TransItmRNA transfection kit (Mirus Bio), according to the manufacturer’s protocol. Briefly, cell culture medium was replaced by 1 ml of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% fetal bovine serum (FBS). Serial dilutions of 2.5 ␮g of RNA transcript in 20 ␮l of Opti-MEM were individually mixed with 250 ␮l of Opti-MEM, 4 ␮l of RNA Boost (Mirus), and 6 ␮l of TransIt RNA (Mirus). The RNAs were applied to HeLa cells and incubated at 37°C for 2 to 3 h. The medium was replaced with DMEM containing 0.4% agarose. Plates were incubated at 37°C for the times indicated in the figure legends, and individual plaques were isolated or plates were stained with crystal violet. Analysis of protein synthesis. HeLa cells in 35-mm plates (⬃5 ⫻ 105 cells/well) were infected at a multiplicity of infection (MOI) of 20 PFU/cell. At 2.5, 4, and 5.5 h postinfection, cells were washed once with DMEM without serum and then incubated for 30 min at 37°C in 1 ml of DMEM supplemented with 0.1% FBS and 20 ␮Ci of [35S]methionine-cysteine (EasyTag; Perkin Elmer). After the labeling step, cells were washed once with cold phosphate-buffered saline (PBS), lysed with 300 ␮l of lysis buffer containing 100 mM Tris-HCl (pH 7.4), 0.5% Triton X-100, and protease inhibitors, and nuclei were pelleted from the cell lysates by centrifugation at 13,000 ⫻ g for 5 min at 4°C. Protein concentration was measured (Bradford reaction), and equal amounts of protein were resolved by SDS–12% polyacrylamide gel electrophoresis. Gels were dried and analyzed by autoradiography, or proteins were transferred to nylon membranes for immunoblot analysis using specified antibodies. Anti-eIF4G rabbit serum was the kind gift of R. Lloyd. In vitro translation/replication assays. Assays were performed as described previously (40). Renilla luciferase replicon assay. Assays were performed essentially as described previously (7). HeLa cells grown in 96-well plates were transfected with poliovirus replicon RNAs (0.8 ng/well) transcribed from pXpA-RenR, pXpA-

VOL. 84, 2010

FLUORESCENT 2A PROTEIN IN REPLICATING POLIOVIRUS

1479

FIG. 2. Schematic representation of PV genomes and plaques produced after transfection of HeLa cells. Underlined amino acids are part of PV genome; bold amino acids show DsRed or GFP sequences; exogenous sequences unrelated to DsRed or GFP are unmarked. Insertions of fluorescent protein coding sequences are indicated by black boxes. Plaques were stained 48 h after transfection with wild-type PV, PV-2A144DsRed, and PV-2A144-GFP RNAs and 72 h after transfection with PV-2A50-DsRed RNA.

Ren-2A144-DsRed, or pXpA-Ren-2A144-GFP. The transfection mix contained 60 ␮M EnduRen Renilla luciferase substrate (Promega). Plates were incubated at 37°C in a Molecular Device MV microplate reader, and light emission from luciferase activity was measured every hour. One-step growth curve. HeLa cells in 35-mm plates were infected with viruses at an MOI of 10 PFU/cell in a total volume of 300 ␮l and incubated for 30 min at room temperature for virus attachment. Excess virus was removed, cells were covered with 1.5 ml of DMEM supplemented with 5% FBS, and plates were incubated at 37°C. At various times cells were scraped from the plates into DMEM and subjected to three freeze-thaw cycles, and the crude virus stocks were clarified by centrifugation at 2,000 ⫻ g for 5 min. Virus titers were determined by plaque assay on HeLa cell monolayers. Viral RNA isolation and reverse transcription (RT)-PCR. Total RNAs were isolated from 140 ␮l of crude virus stocks using a QIAamp Viral RNA kit (Qiagen) according to the manufacturer’s instructions and eluted in 60 ␮l of elution buffer. First-strand cDNA synthesis was performed using 3 ␮l of isolated total RNA, 10 pmol of V3 oligo(dT) primer, and a MonstreScript 1st-Strand cDNA Synthesis Kit (Epicentre Biotechnologis). The reaction mixtures (20 ␮l) were incubated for 5 min at 42°C and 40 min at 60°C according to the manufacturer’s protocol, after which the enzyme was inactivated by incubation at 95°C

for 5 min. Two microliters of RT reaction mixture was used as a template for PCR. Reaction mixtures contained a 200 ␮M concentration of each deoxynucleoside triphosphate (dNTP), 200 nM (each) upstream and downstream primers, PCR buffer, and 1 U of high-fidelity DNA polymerase (Phusion HighFidelity PCR Kit; Finnzymes). The reaction mixtures were heated for 2 min at 95°C and cycled 30 times for 30 s at 95°C, 30 s at 62°C, and 90 s at 72°C. Primer pair 1 consisted of primer fpDsRed⫹ (5⬘-GGACAACACCGAGGACGTCATC AAG) and primer rpDsRed⫺ (5⬘-CTGGGAGCCGGAGTGGCGGGCC) specific for the 5⬘ and 3⬘ ends of the DsRed coding sequence, respectively. Primer pair 2 consisted of primer 2942⫹ (5⬘-TACCAAATTATGTACGTACCACC AGG) and primer rp4657⫺ (5⬘-CAGGTCGTCCATAATCACCACTCCCTGT TG). PCR products were analyzed on 1% agarose gels. Digitonin treatment. Digitonin treatment was performed essentially as previously described (7). Briefly, HeLa cells were grown on coverslips in 24-well plates. Cells were transfected using Fugene 6 reagent (Roche) with plasmids expressing the membrane-binding marker Sec61-yellow fluorescent protein (YFP) or free DsRed protein for 24 h or infected with PV-2A144-DsRed at an MOI of 10 PFU/cell for 6 h. Cells were washed once with KHM buffer (110 mM K-acetate, 2 mM MgCl2, 20 mM HEPES-KOH, pH 7.4), treated for 1 min with

1480

TETERINA ET AL.

50 ␮M digitonin in KHM buffer, washed with KHM buffer, and fixed with 4% formaldehyde in PBS and processed for fluorescence microscopy. Microscopy. For confocal microscopy, HeLa cells were seeded in 24-well plates on glass coverslips and infected at an MOI of 10 PFU/cell with recombinant viruses. At the times indicated in the figure legends, cells were fixed with 4% paraformaldehyde in PBS for 20 min, permeabilized with 0.2% Triton X-100 in PBS for 10 min, and stained with mouse monoclonal antibodies against PV protein 2C (generous gift from K. Bienz and D. Egger) and goat anti-mouse secondary antibodies labeled with Alexa Fluor 488. Nuclei were labeled with Hoechst 33342 (Invitrogen) at a final concentration of 1 ␮g/ml during the last wash. Images were collected on a Leica SP5 confocal microscope (Leica Microsystems, Exton, PA) using a 63⫻ oil immersion objective with a numerical aperture of 1.4 at zoom 4. Fluorochromes were excited using a 405-nm laser for 4⬘,6⬘-diamidino-2-phenylindole (DAPI), an argon laser at 488 nm for Alexa Fluor 488, and a 561-nm laser for DsRed. Detector slits were configured to minimize any cross talk between the channels (PMT1, 430 nm to 480 nm [DAPI]; PMT2, 492 nm to 550 nm [Alexa Fluor 488]; PMT3, 571 nm to 700 nm [DsRed]). Differential interference contrast (DIC) images were collected simultaneously with the fluorescence images using the transmitted light detector. Images were processed using Leica LAS-AF software (version 1.9.1) and Adobe Photoshop CS4 (Adobe systems). Raw data were deconvolved to increase resolution using Huygens Essential software (version 3.3; Scientific Volume Imaging BV, Hilversum, the Netherlands). Sequential z-sections of stained cells were acquired for three-dimensional (3D) reconstruction and isosurface modeling of representative cells with Imaris software (version 6.2.1; Bitplane AG, Zurich, Switzerland). In order to determine more quantitatively the extent of colocalization of 2ADsRed and 2C proteins, quantitative colocalization analyses were performed in Imaris. For colocalization analyses, channel masking and the semiautomatic threshold technique of the Imaris colocalization module were used to calculate colocalization statistics and Pearson’s correlation coefficients in the colocalized 3D volume as well as percentages of channel volume colocalization. Analysis of fluorescent proteins in cells infected with different passages of PV-2A144-DsRed and distribution of fluorescent proteins after digitonin treatment was performed on a Leica SP5 confocal microscope. EM. HeLa cells were seeded onto Thermanox plastic coverslips in 12-well plates and infected with wild-type or recombinant viruses at an MOI of 10 PFU/cell. Four hours after infection with wild-type virus or 6 h after infection with recombinant viruses, cells were fixed in 2.5% glutaraldehyde solution in 0.1 M sodium cacodylate buffer. Cells were processed for electron microscopy (EM) in the NIAID Rocky Mountain Lab Electron Microscopy facility.

RESULTS PV protein 2A tolerates small insertions at two locations. A random library of transposon-generated 15-nucleotide insertions in the P2 coding region of the poliovirus polyprotein produced two different viable genomes encoding five amino acid insertions within the 2A sequence (Fig. 1A). These were identified by their abilities to induce plaques after transfection into HeLa cells. A detailed description of the library and identification of insertion sites will be presented elsewhere. The five amino acid residues had been inserted either between 2A residues 50 and 51 or 144 and 145. Figure 1B shows the predicted locations of the two tolerated insertion sites mapped on a structural model of the 2A protein from closely related coxsackievirus B4 resolved in solution by nuclear magnetic resonance (NMR) spectroscopy (5). The insertion at residue 50 of 2A (2A50) is located in a loop region that separates two domains of the protein. The 2A144 insertion is located near the C-terminal end of the protein in a region with apparently little stable structure. Viruses carrying either of these insertions in 2A formed plaques on HeLa cell monolayers with nearly wildtype size and morphology (data not shown). Poliovirus tolerates insertion of fluorescent protein at 2A insertion sites. Since little is known about the intracellular localization of the PV 2A protein in infected cells, we constructed recombinant PV genomes with FP coding sequences

J. VIROL. TABLE 1. Infectivity of in vitro synthesized viral RNAs and comparative plaque sizes RNA transcript

Infectivity (PFU/ ␮g of RNA)

Plaque size (mm)a

PVMc PV-2A50-DsRed PV-2A144-DsRed PV-2A144-GFP

3.2 ⫻ 106 1.2 ⫻ 105 4.4 ⫻ 106 1.6 ⫻ 106

2–4 ⬍0.5b 1–2 0.5–1.5

a b c

Plaques were stained after 48 h of incubation under agarose. Plaques were stained after 72 h of incubation under agarose. PVM, PV strain Mahoney.

inserted into the sites that we had identified as tolerant of the small insertions. Either DsRed or GFP coding sequences were inserted (Fig. 2); both of these proteins (25 to 30 kDa) are larger than the viral 2A protein into which they were inserted. Nevertheless, when HeLa cells were transfected with RNA transcripts from these constructs, all RNAs encoding 2A-FP generated viable viruses and displayed infectivities similar to the infectivity of wild-type RNA (Table 1), suggesting that no adaptive mutations were required for production of viruses carrying these large insertions. Examination of the plaques generated after transfection with 2A-DsRed RNAs (Fig. 2) showed that transcripts encoding DsRed insertion after aa 50 produced virus with a minute-plaque phenotype, indicating severe inhibition of virus growth produced by FP insertion at this position; these viruses were not used for further experiments. RNA transcripts encoding a DsRed insertion after aa 144, however, produced clear virus plaques that were smaller than wild-type (Fig. 2 and Table 1). A comparison of DsRed and GFP insertions at position 144 showed that GFP had a more debilitating effect on virus growth than insertion of DsRed, despite the similar sizes of these two FP insertions (Fig. 2). Plaques formed by PV-2A144-GFP were somewhat smaller than those formed by PV-2A144-DsRed. These preliminary analyses showed that the C-terminal insertion site after amino acid residue 144 in 2A was more tolerant for the insertion of FP than the internal site after residue 50, and DsRed was accepted better than GFP in this position although insertions of both FPs imposed some negative effects on virus growth. In addition, PV-2A144-DsRed formed plaques of similar size at 32o and 37o, whereas those formed by wild-type virus were significantly larger at 37o. Recombinant PV-2A144-FP viruses were isolated from individual plaques produced after transfection and used to infect fresh HeLa cell monolayers to produce passage 1 virus stocks. As PV-2A144-DsRed virus had better growth properties, we selected this virus for further analysis. Genetic stability of PV-2A144-DsRed. HeLa cells were infected with PV-2A144-DsRed virus from passage 1 to generate passage 2, which was subsequently serially passaged up to passage 8 (see Materials and Methods). Expression of DsRed was analyzed after each passage by fluorescence microscopy, and the percentage of fluorescent cells was determined by manual counting of individual cells in two fields for each passage. Cells infected with passage 2 PV-2A144-DsRed virus exhibited strong DsRed-specific fluorescence in more than 90% of all cells. Infections with passage 6 virus produced only 50 to 65% of cells positive for DsRed fluorescence, and the intensity of

VOL. 84, 2010


the fluorescent signal was significantly decreased (Fig. 3A). By the eighth passage, only 20 to 35% of cells showed detectable fluorescence (data not shown). The right-hand panels in Fig. 3A show that all cells were infected and produced viral proteins at both passages, as indicated by staining of cells with antibodies to viral protein 2C. We interpreted these results to reflect a gradual, continuous loss of all or part of the DsRed coding sequences during replication of the viral genome. We used RT-PCR to analyze the genomes extracted from virus stocks harvested after various numbers of passages. When RT-PCR was performed with primers annealing to the ends of the DsRed coding sequence such that products would be generated exclusively from viral RNAs that retained the insert, a fragment of the size expected for the intact parental RNA transcript (673 bp) was formed from RNAs isolated from all passages, albeit in decreasing amounts from the later-passage virus stocks (Fig. 3B). This confirmed that at least part of the virus stock after eight passages of the original PV-2A144DsRed had retained the complete DsRed coding sequence and was consistent with the appearance of some red fluorescent cells in cultures infected with this virus stock. When the RTPCR was performed using primers annealing to PV sequences upstream and downstream of the DsRed coding sequence such that all viral RNAs would generate products, fragments shorter than those that contain a complete insert (2,421 bp) could be detected in passage four and in greater amounts with increasing passages (Fig. 3C). These results suggest that during replication of the viral genome, deletion events occurred that conferred some selective advantage to the virus. Sequence analysis of non-plaque-purified, passaged virus stocks revealed that deletions of various lengths generally occurred at the 5⬘ end of the insertion, with ⬃150 nt from the 3⬘ end of the insert retained in the emerging viruses. However, virus stocks passaged fewer than three to four times demonstrated the presence of inserted sequences and expression of 2A-DsRed fusion protein in over 80% of cells. This level of stability was deemed sufficient for use in studies of virus replication and warranted further characterization of the fluorescent virus growth properties. Single-cycle growth curve of fluorescent virus. Despite the presence of the rather long insertion in its 2A protein, PV2A144-DsRed manifested a specific infectivity similar to the parental wild-type virus (Table 1) and retained a relatively stable genotype over multiple passages. Nevertheless, plaques were smaller than those of wild-type virus, and further investigations of the growth properties of PV-2A144-DsRed were deemed necessary. The reduced plaque size of the 2A144-FP insertion mutants suggested that FP insertions caused some impairment of virus growth. To characterize the growth kinetics of chimeric virus, we performed a single-cycle growth analysis by infecting HeLa cell monolayers at a high MOI with either wild-type or PV-2A144-DsRed virus. Cells were harvested at various times after infection, and viral titers in cell lysates were determined by plaque assay (Fig. 4). The singlecycle growth curve shows that the PV-2A144-DsRed virus accumulated more slowly than wild-type virus, and maximum virus accumulation was approximately 1 log10 lower than for wild-type virus. These data are consistent with the smaller plaque phenotype we observed for this virus (Fig. 2).

1481

FIG. 3. Genetic stability of PV-2A144-DsRed. Virus was passaged in HeLa cells, as described in Materials and Methods, to produce passages 2 to 8. (A) HeLa cells were cultured in 24-well plates and were infected with virus from passage 2 or passage 6 at an MOI of 10 PFU/cell. At 8 h postinfection, cells were fixed and examined by fluorescence microscopy. The merged images of 2A-DsRed (red) and light-field microscopy are shown in the left panels, and nuclei (blue) and 2C protein (green) are shown in the right panels. (B and C) RT-PCR analysis of viral RNAs isolated from passages 2 to 8 of PV-2A144-DsRed using oligonucleotide primers mapping within inserted DsRed sequence (B) or oligonucleotide primers mapping in PV sequence flanking the inserted sequence (C). M, DNA molecular weight markers; T, RT-PCR of pPV-2A144-DsRed RNA transcript. Products were resolved on a 1.0% agarose gel stained with ethidium bromide.

1482

TETERINA ET AL.

J. VIROL.

FIG. 4. Virus replication in HeLa cells. Cells were infected with wild-type or PV-2A144-DsRed virus (passage 1) at an MOI of 10 PFU/cell. At the indicated times, cells were collected as described in Materials and Methods, and virus titers were determined by plaque assay.

The 2A-DsRed fusion protein is properly processed, manifests protease activity in cis and in trans, and supports translation of viral RNA. The presence of a large insertion in the viral polyprotein could significantly alter the polyprotein folding so as to affect presentation of individual sites needed for subsequent protein processing. In addition, insertion of sequences within the 2A coding sequence might directly affect its proteolytic activity. To test for defects in polyprotein processing that might be caused by the DsRed protein insertions in 2A, we monitored the pattern of viral proteins in HeLa cells infected with wild-type or mutant virus under single-step growth conditions. Cells were labeled with [35S]methionine for 60 min prior to extraction at 3, 4.5, or 6 h postinfection, and proteins were analyzed by SDS-PAGE and autoradiography. Figure 5A shows that, as expected for wild-type virus, maximum viral protein synthesis was observed between 2 and 3 h postinfection, and complete shutoff of cellular protein synthesis was evident by that time. The DsRed 2A144 mutant displayed somewhat delayed synthesis of viral proteins and inhibition of cellular protein synthesis. Maximum synthesis of virus-specific proteins was detected between 3.5 and 4.5 h postinfection. These kinetics of protein synthesis were consistent with the single-step growth curve (Fig. 4) and relative plaque sizes (Fig. 2). As expected, a band representing the fully processed wildtype 2A protein was not observed in the profile of viral proteins (Fig. 5A). Western blot analyses confirmed that the predicted 2A fusion proteins with molecular masses of ⬃42 kDa were present, detected by antibodies against PV protein 2A as well as against DsRed monomer protein (data not shown). No other significant differences in the viral protein profiles were observed. Western blot analysis with anti-2C antibodies demonstrated production of 2BC and 2C proteins, as seen in wildtype-infected cells, and failed to detect any higher-molecularweight precursors containing 2C sequences that might indicate abnormal processing (data not shown). Thus, the insertion of DsRed protein in 2A did not impair the overall folding and processing of the polyprotein. Importantly, correctly processed VP1 protein was detected in cells infected with mutant virus, demonstrating retention of the proteolytic activity of 2A required for this cleavage.

FIG. 5. Effects of DsRed insertion on polyprotein processing and 2A protease activity. (A) Protein synthesis in cells infected with wildtype PV or PV-2A144-DsRed virus. HeLa cells were mock infected (lane 1) or infected with wild-type PV (lanes 2 to 4) or PV-2A144DsRed (lanes 5 to 7) at an MOI of 10. Proteins were metabolically labeled with [35S]methionine-cysteine at 3 (lanes 2 and 5), 4.5 (lanes 3 and 6), or 6 (lanes 4 and 7) h postinfection and analyzed by SDS– 12.5% PAGE. (B) Cleavage of eIF4G in HeLa cell extracts by 2A-FP protein. RNAs coding for wild-type or mutant 2A protein were translated in HeLa S10 extracts, and proteins were analyzed by immunoblotting with anti-eIF4G antibodies. The mobilities of marker proteins are indicated. hpi, hours postinfection.

We also examined the ability of 2A harboring an FP insertion to mediate cleavage of the cellular protein eIF4G. The profile of labeled proteins synthesized in cells infected with mutant virus (Fig. 5A) indicated efficient inhibition of host protein synthesis, albeit with similar delay compared with wildtype virus as observed for production of viral proteins. This observation suggested that the 2A144-DsRed fusion protein functioned in cleavage of the cellular eIF4G translation initiation factor, which is largely responsible for inhibition of cellular translation. Direct demonstration of eIF4G cleavage by the 2A-DsRed fusion proteins was provided by immunoblot analysis of eIF4G in extracts after translation of PV RNAs in vitro (Fig. 5B). The integrity of eIF4G was maintained during at least 3 h of incubation in mock reactions (Fig. 5B, Mock); in

VOL. 84, 2010


1483

FIG. 6. Replication of PV-2A144-DsRed mutant RNA. (A and B) S10 extracts from HeLa cells were programmed for translation/replication of indicated full-length RNA transcripts. Lanes 1, no RNA control. (A) Viral proteins were synthesized and processed in reaction mixtures containing [35S]methionine, and labeled viral proteins were analyzed on 12.5% SDS-polyacrylamide gels. The arrows indicate the position of 2A-DsRed protein. (B) RNA synthesis was measured in reaction mixtures containing preinitiation replication complexes isolated after translation of the indicated RNAs in the presence of 2 mM guanidine hydrochloride (GuHCl) and then resuspended in the absence of GuHCl. (C) Kinetics of replication of PV replicons with insertion of DsRed in protein 2A. HeLa cells were transfected with PV Renilla luciferase replicon RNA transcripts in 96-well plates and incubated in the presence of cell-permeable Renilla luciferase substrate. Light readings were taken every hour with a Molecular Devices MV reader. Error bars indicate standard deviation from the mean of 32 replicate samples. RLU, relative light units; wt, wild type.

contrast, reactions programmed with PV-2A144-DsRed RNA contained only cleavage products of eIF4G, similar to those reactions programmed with wild-type PV RNA. Thus, insertion of FP in 2A144 did not affect the trans-proteolytic activity of 2A proteinase. Replication of 2A144-DsRed insertion mutant RNAs. Individual assays described above revealed no detectable effects of the DsRed insertion in 2A on translation, protein processing, or proteolytic activities of the recombinant 2A proteins. Protein 2A also has been implicated in an undefined function during replication of viral RNA, which involved C-terminal residues 140 to 144 (20, 22). This finding prompted us to examine whether insertion of DsRed at position 144 of 2A specifically affected viral RNA replication. We first examined synthesis of RNA in vitro in translation/replication extracts prepared from uninfected HeLa cells in the absence of more stringent requirements for virus production in cells. Figure 6A shows analysis of the protein produced in in vitro translation reactions performed in the presence of [35S]methionine-cysteine from the corresponding RNA replication reactions. The comparable amounts of labeled viral proteins present in each lane indicate that the overall translation levels were approximately equal for wild-type and mutant RNAs. As expected, based on the analysis of the viral proteins produced in infected cells (Fig. 5A), no processing defects were observed, and the fusion protein 2A-DsRed is clearly visible among the trans-

lation products of these RNAs. When these reactions were analyzed for synthesis of viral RNA in vitro, we found that the levels of RNA synthesized in reactions programmed with RNAs encoding 2A144-DsRed were similar to those produced by wild-type RNA, albeit slightly decreased (Fig. 6B). To directly measure the effect of the 2A144-DsRed insertion on viral RNA replication efficiencies in HeLa cells, we engineered this FP insertion in the background of a PV replicon RNA containing the Renilla luciferase gene in place of the structural protein coding region. HeLa cells were transfected with the replicon RNAs, and Renilla luciferase activity was monitored in intact cells as a measure of replicon RNA synthesis. Figure 6C shows that the replicon encoding the DsRed protein insertion in 2Apro replicated with similar efficiency as the wild-type replicon. This result demonstrates that 2A144DsRed fusion proteins support viral RNA replication almost as well as wild-type 2A. Thus, the slower production of infectious virus particles determined by plaque assay (Fig. 2) and singlestep growth curves (Fig. 4) is likely caused by assembly and packaging of the insertion mutant RNAs, whose lengths are increased over wild-type by ⬃10%. Replication complexes formed by PV-2A144-DsRed. EM analysis of the cells infected with PV-2A144-DsRed virus was performed to examine the membrane alterations and morphology of replication complexes in infected cells (Fig. 7). As is typical for wild-type PV infections, remodeling of intracellular

1484

TETERINA ET AL.

J. VIROL.

membranes to generate clusters of membrane vesicles was observed in cells infected with virus harboring the 2A-DsRed fusion. The overall cellular ultrastructure (Fig. 7A), as well as the morphology of the vesicles (Fig. 7B), was indistinguishable from that observed in control cells infected with wild-type PV (Fig. 7C). Similar alterations of membrane structures were also observed in cells infected with PV-2A144-GFP (data not shown). These results demonstrate that insertions of DsRed at the C terminus of 2A did not affect the reorganization of membranes in infected cells or the proliferation of vesicle clusters that represent the sites of viral RNA replication characteristic of PV infection. Visualization of 2A-FP in infected cells. Live-cell imaging of PV-2A144-DsRed-infected HeLa cells revealed some red cells detectable by 3.5 to 5 h postinfection (not shown). By 8 h postinfection, almost all cells developed a very strong fluorescent signal. The delay in visualization of 2A-DsRed expression by fluorescence microscopy compared to detection of radiolabeled proteins by SDS-PAGE (Fig. 5A) or immunoblot analysis of infected cell extracts (data not shown) apparently represents the time required for maturation of the DsRed chromophore. To characterize the distribution and localization of 2ADsRed in infected cells, we analyzed the subcellular localization of 2A-DsRed during the course of infection. HeLa cells were infected with PV-2A144-DsRed at an MOI of 10 PFU/ cell and fixed at 3.5, 5, 6, and 8 h postinfection. Figure 8a shows that at early times after infection, the DsRed fluorescence appeared predominantly as a punctate pattern in the cytoplasm. As infection progressed and cytopathic changes developed (Fig. 8b and c), increased DsRed fluorescence became distributed extensively throughout the cytoplasm, with some accumulation along the cell periphery near the plasma membrane (Fig. 8d and Fig. 9a). To determine the relationship between the locations of 2ADsRed and the PV replication complexes, HeLa cells were infected with PV-2A144-DsRed virus at an MOI of 10 PFU/ cell and fixed at 8 h postinfection. Cells were labeled with monoclonal antibodies that react with protein 2C, a standard marker for membrane-bound PV replication sites, and examined by immunofluorescence microscopy. Images of infected cells at different times postinfection were obtained by confocal microscopy and analyzed by deconvolution and reconstitution software (see Materials and Methods). Side views of the threedimensional volume of representative cells produced by reconstitution are shown in the bottom row of Fig. 9. Panels a and d show that the 2A-DsRed protein was restricted to the cytoplasm and appeared as multiple punctate foci near the nucleus as well as being extensively distributed throughout the cytoplasm, also shown in Fig. 3. PV 2C protein was localized exclusively in the perinuclear area in a punctate pattern (Fig. 9b and e), as has been described previously (11). As apparent from the merged images

FIG. 7. Electron micrographs of HeLa cells infected with PV2A144-DsRed or wild-type PV. Cells were fixed at 5.5 h postinfection for PV-2A144-DsRed (A and B) and at 4 h postinfection for PV (C). Scale bar, 500 nm. nuc, nucleus.

VOL. 84, 2010


FIG. 8. Intracellular distribution of PV 2A-DsRed during infection. HeLa cells were infected with PV-2A144-DsRed at an MOI of 10 PFU/cell. Cells were fixed with 4% paraformaldehyde at 3.5, 5, 6, and 8 h postinfection, and nuclei were stained with Hoechst 3334 as described in Materials and Methods. The images were acquired on a Leica SP5 confocal microscope using a 63⫻ (1.4 numerical aperture) oil immersion objective.

shown in Fig. 9c and f, a portion of the 2A-DsRed protein was found together with 2C protein: at 8 h postinfection, approximately 20% of the 2A-DsRed fluorescence colocalized with the 2C protein (Fig. 9c and f). However, a larger fraction of 2A-

1485

DsRed was distributed throughout the cytoplasm, apparently associated with an intracellular network outside of the perinuclear replication complexes. These data suggested that, at least later in infection, the PV-2A144-DsRed-infected cells contained two types of subcellular complexes containing 2ADsRed: one appeared to be associated with membrane-bound, perinuclear PV replication complexes, whereas the other was organized into a structural matrix throughout the cytoplasm, not colocalized with replication complexes. A similar appearance of GFP was observed in cells infected with PV-2A144GFP (data not shown), indicating that the observed staining was not an artifact resulting from the insertion of specific DsRed sequences into 2A protein. To further characterize the distribution and localization of 2A-DsRed in infected cells, we treated cells with digitonin prior to fixation to selectively permeabilize the plasma membrane containing high concentrations of cholesterol, thereby allowing soluble cytoplasmic components to be washed out and released from the infected cells (34). Intracellular membrane organelles and other insoluble subcellular structural components are unaffected by digitonin treatment. Figure 10 shows the results of the digitonin permeabilization assay. As controls, cells were transfected with plasmids expressing the endoplasmic reticulum (ER) protein Sec61 fused to yellow fluorescent protein (Sec61-YFP) or soluble DsRed protein (Fig. 10A). Cells expressing soluble DsRed protein displayed diffuse fluorescence throughout the cytoplasm as well as an accumulation of DsRed in nuclei. In cells expressing Sec61-YFP protein, fluorescence was observed in an ER-like network in the cytoplasm. Treatment of cells expressing free DsRed proteins with digitonin resulted in complete loss of cytoplasmic fluorescence, with only nuclei retaining a fluorescent signal, whereas in cells expressing Sec61-YFP, the ER staining remained intact after digitonin treatment. As shown in the previous experiment, cells

FIG. 9. Localization of 2A-DsRed and 2C proteins in infected cells. HeLa cells were infected at an MOI of 10 PFU/cell. At 8 h postinfection cells were fixed with 4% paraformaldehyde, permeabilized with Triton X-100, and stained with PV-2C protein-specific antibodies (green) as described in Materials and Methods. The top row (a to c) shows the xy-axis view of the reconstructed three-dimensional volume of the cell. The bottom row (d to f) shows the z-stack side view of the reconstructed volumes of the two cells at the bottom of the fields shown in the top row. The red channel represents 2A-DsRed, and green represents protein 2C. Merged green and red channels together show colocalization in yellow.

1486

TETERINA ET AL.

FIG. 10. Protein 2A is associated with cellular structures in HeLa cells. (A) HeLa cells were transfected with plasmids expressing free DsRed protein or ER-targeted Sec61-YFP. After 24 h cells were treated with digitonin, fixed, and examined by fluorescence microscopy. (B) HeLa cells were infected with PV-2A144-DsRed at an MOI of 10 PFU/cell. At 6 h postinfection cells were treated with digitonin and fixed as described for panel A.

infected with PV-2A144-DsRed exhibited a punctate fluorescence concentrated predominantly in the perinuclear region with the appearance of multiple small foci (Fig. 10B). This fluorescence distribution was clearly different from the diffuse distribution of free DsRed and indicated that 2A-DsRed protein was associated with some structures in the infected cell. This pattern of fluorescence was retained after digitonin permeabilization of the cellular membrane, confirming that 2ADsRed protein had some stable interactions with intracellular membranes. DISCUSSION The use of fluorescent proteins as protein tags for cellular or viral proteins, in combination with advanced imaging techniques, has proven very useful for visualization of dynamic processes in living cells. In the present study we describe PV genomes that allow direct visualization of PV protein 2A in living cells. Central to this work was the identification of sites in this protein that tolerated insertion of fluorescent tags. Most commonly, fusion of tags to proteins of interest is performed at one or the other terminus of the polypeptide chain. However, as is the case for many positive-strand RNA viruses, expression

J. VIROL.

of the picornavirus genome produces a large polyprotein that is subsequently cleaved by viral proteases to generate individual gene products. This gene expression strategy constrains the addition of tags to the ends of proteins since the presence of the inserted tag sequence usually interferes with the subsequent cleavage of the precursor. An alternate approach to attempting to incorporate individual, tagged proteins expressed from separate expression vectors into functional PV replication complexes would not likely succeed since complementation has generally proven to be extremely inefficient (13, 32, 41). To identify sites in protein 2A that would accept insertion of fluorescent proteins and still generate viable virus, we used a modification of the transposon-based insertion mutagenesis method that was previously used for developing hepatitis C virus (HCV) replicons with GFP insertions in an NS5A sequence (28) and Sindbis viruses with GFP insertions in nonstructural proteins nsP2 and nsP3 (4, 14). This approach generated a pool of random insertions within the PV nonstructural proteins, which was subsequently used to select those sites with insertions that would yield viable viruses when introduced into an infectious RNA genome. After transfection of the in vitro synthesized RNA into HeLa cells and selection of plaqueforming viruses, two permissive sites were identified in PV protein 2A, in the predicted unstructured loop after aa 50 and in the unstructured region near the C terminus of the protein. Insertion of DsRed or GFP into the unstructured loop region severely impaired virus growth; however, DsRed protein insertion into the C-terminal site of protein 2A affected virus growth to a significantly lesser extent, indicating considerable flexibility of the 2A protein in that region. The similarly sized insertion of GFP in the same site of 2A protein caused a more debilitating effect on virus growth. Thus, even sites capable of tolerating insertions of long polypeptides display preferences for specific amino acid sequences, which have a significant influence on protein function and virus growth, so selection of the most useful tags remains an empirical process for each site. The preferred insertion site was located directly upstream of a C-terminal sequence (EEAMEQ) in the 2A protein that was previously implicated as being essential for viral RNA replication although its precise function was not identified (20). Our data show that the large FP insertion just upstream of this sequence had no significant adverse effect on viral RNA replication. Thus, it appears that the presence of the EEAMEQ sequence at the C-terminal end of 2A is sufficient to express the unidentified function of 2A in replication, and the protein sequence upstream has a much lesser effect. Several attempts have been made previously to insert foreign genes into the PV genome, usually with the aim of using PV as a vector for the delivery of foreign antigens. One approach was to replace the neutralizing antigenic sites located on the surface of the PV capsid with antigenic sites from PV of different types (24, 31) or other more distantly related picornaviruses (2, 37). Another approach attempted to develop more versatile vectors expressing larger foreign protein sequences. The foreign genes were either inserted as separate open reading frames in bicistronic genomes (1, 23) or inserted at the amino terminus of the polyprotein or at the junction between the P1 and P2 coding regions and separated by an introduced 3C proteinase cleavage site (3, 25, 26, 30, 38, 42). In all cases both the size and the nature of the inserted sequence

VOL. 84, 2010


affected the growth and stability of recombinant viruses. For example, although insertion of 155 aa of hepatitis B virus core protein was reported to be stable during at least six passages of recombinant virus (42), other studies demonstrated a maximum size of inserted foreign sequence in the range of 100 aa (25, 26). Most of these recombinant viruses were characterized as temperature sensitive. Insertions of larger size were found to be severely restricting for viral replication and were genetically unstable. Thus, when GFP coding sequences were inserted in the PV genome at the beginning of the polyprotein open reading frame, a significant delay and decrease in virus yield after transfection occurred, and a major part of the GFP coding sequence was lost after six passages (30). Although we also observed viral genomes with deletions of a significant portion of the inserted sequence upon passaging, insertion of DsRed protein in the C-terminal sequence of 2A proteinase was stable in the majority of viral progeny at least until passage 4, which readily permits live imaging of infected cells. Despite the large size of the DsRed protein insertion (235 aa) which increased the size of the 2A protein by ⬃160%, placement of DsRed sequences in the C-terminal site of 2A had no adverse effect on polyprotein processing. This indicates that the proteolytic activity of the 2A proteinase responsible for the cleavage between its own N terminus and VP1 was not affected. Similarly, there appears to be no effect on the availability or presentation of the cleavage site between 2A and 2B, which is subsequently cleaved by the 3C proteinase. There are a number of other proteolytic cleavages catalyzed by the 2A proteinase on cellular protein substrates that are likely important or essential for efficient virus infection; here, we examined only one such cleavage and showed that insertion of the DsRed tag near the C terminus of the 2A protein did not reduce its ability to cleave eIF4G. Since the PV 2A protein has been implicated in an antagonistic effect on interferon (IFN) signaling (29) and since we have not determined whether the 2ADsRed fusion protein retains this activity, it will be important to examine this effect in a transgenic mouse model. In cells infected with PV-2A144-DsRed the intracellular distribution of fluorescence was punctate, with a majority of fluorescence distributed throughout the cytoplasm. A fraction of 2A-DsRed protein (⬃20 to 25%) colocalized with viral protein 2C, a known marker of PV replication complexes, lending credence to its proposed role in viral RNA replication. Importantly, a large fraction of 2A-DsRed protein was not associated with 2C-containing membrane-bound replication complexes and apparently formed another type of complex in the cytoplasm, whose possible function(s) at this time remains unknown. ACKNOWLEDGMENTS This research was supported by the Intramural Research Program of the NIH, NIAID. We thank Juraj Kabat, Lily Koo, and Owen M. Swartz of the Biological Imaging Facility, NIAID, NIH, for assistance with confocal microscopy and Juraj Kabat for help with image processing and analysis. Elizabeth Fischer, RML Electron Microscopy Unit, NIAID, generated excellent electron microscopic images. We are grateful to Rick Lloyd for the gift of anti-eIF4G serum, K. Bienz and D. Egger for monoclonal anti-2C antibodies, and George Belov for pXpA-RenR plasmid. We thank Don Weaver for preparation of HeLa S10 extracts and Eric Belk for technical assistance.

1487

REFERENCES 1. Alexander, L., H.-H. Lu, and E. Wimmer. 1994. Poliovirus containing picornavirus type 1 and/or type 2 internal ribosomal entry site elements: genetic hybrids and the expression of a foreign gene. Proc. Natl. Acad. Sci. U. S. A. 91:1406–1410. 2. Altmeyer, R., A. D. Murdin, J. J. Harber, and E. Wimmer. 1991. Construction and characterization of a poliovirus/rhinovirus antigenic hybrid. Virology 184:636–644. 3. Andino, R., D. Silvera, S. D. Suggett, P. L. Achacoso, C. J. Miller, D. Baltimore, and M. B. Feinberg. 1994. Engineering poliovirus as a vaccine vector for the expression of diverse antigens. Science 265:1448–1451. 4. Atasheva, S., R. Gorchakov, R. English, I. Frolov, and E. Frolova. 2007. Development of Sindbis viruses encoding nsP2/GFP chimeric proteins and their application for studying nsP2 functioning. J. Virol. 81:5046–5057. 5. Baxter, N. J., A. Roetzer, H. D. Liebig, S. E. Sedelnikova, A. M. Hounslow, T. Skern, and J. P. Waltho. 2006. Structure and dynamics of coxsackievirus B4 2A proteinase, an enyzme involved in the etiology of heart disease. J. Virol. 80:1451–1462. 6. Belov, G. A., N. Altan-Bonnet, G. Kovtunovych, C. L. Jackson, J. LippincottSchwartz, and E. Ehrenfeld. 2007. Hijacking components of the cellular secretory pathway for replication of poliovirus RNA. J. Virol. 81:558–567. 7. Belov, G. A., C. Habbersett, D. Franco, and E. Ehrenfeld. 2007. Activation of cellular Arf GTPases by poliovirus protein 3CD correlates with virus replication. J. Virol. 81:9259–9267. 8. Belov, G. A., P. V. Lidsky, O. V. Mikitas, D. Egger, K. A. Lukyanov, K. Bienz, and V. I. Agol. 2004. Bidirectional increase in permeability of nuclear envelope upon poliovirus infection and accompanying alterations of nuclear pores. J. Virol. 78:10166–10177. 9. Bonderoff, J. M., J. L. Larey, and R. E. Lloyd. 2008. Cleavage of poly(A)binding protein by poliovirus 3C proteinase inhibits viral internal ribosome entry site-mediated translation. J. Virol. 82:9389–9399. 10. Borman, A. M., R. Kirchweger, E. Ziegler, R. E. Rhoads, T. Skern, and K. M. Kean. 1997. elF4G and its proteolytic cleavage products: effect on initiation of protein synthesis from capped, uncapped, and IRES-containing mRNAs. RNA 3:186–196. 11. Cho, M. W., N. Teterina, D. Egger, K. Bienz, and E. Ehrenfeld. 1994. Membrane rearrangement and vesicle induction by recombinant poliovirus 2C and 2BC in human cells. Virology 202:129–145. 12. Collis, P. S., B. J. O’Donnell, D. J. Barton, J. A. Rogers, and J. B. Flanegan. 1992. Replication of poliovirus RNA and subgenomic RNA transcripts in transfected cells. J. Virol. 66:6480–6488. 13. Egger, D., N. Teterina, E. Ehrenfeld, and K. Bienz. 2000. Formation of the poliovirus replication complex requires coupled viral translation, vesicle production, and viral RNA synthesis. J. Virol. 74:6570–6580. 14. Frolova, E., R. Gorchakov, N. Garmashova, S. Atasheva, L. A. Vergara, and I. Frolov. 2006. Formation of nsP3-specific protein complexes during Sindbis virus replication. J. Virol. 80:4122–4134. 15. Graham, K. L., K. E. Gustin, C. Rivera, N. M. Kuyumcu-Martinez, S. S. Choe, R. E. Lloyd, P. Sarnow, and P. J. Utz. 2004. Proteolytic cleavage of the catalytic subunit of DNA-dependent protein kinase during poliovirus infection. J. Virol. 78:6313–6321. 16. Hambidge, S. J., and P. Sarnow. 1992. Translational enhancement of the poliovirus 5⬘ noncoding region mediated by virus-encoded polypeptide 2A. Proc. Natl. Acad. Sci. U. S. A. 89:10272–10276. 17. Jurgens, C. K., D. J. Barton, N. Sharma, B. J. Morasco, S. A. Ogram, and J. B. Flanegan. 2006. 2Apro is a multifunctional protein that regulates the stability, translation and replication of poliovirus RNA. Virology 345:346– 357. 18. Kuyumcu-Martinez, N. M., M. Joachims, and R. E. Lloyd. 2002. Efficient cleavage of ribosome-associated poly(A)-binding protein by enterovirus 3C protease. J. Virol. 76:2062–2074. 19. Lee, C. K., and E. Wimmer. 1988. Proteolytic processing of poliovirus polyprotein: elimination of 2Apro-mediated, alternative cleavage of polypeptide 3CD by in vitro mutagenesis. Virology 166:405–414. 20. Li, X., H. H. Lu, S. Mueller, and E. Wimmer. 2001. The C-terminal residues of poliovirus proteinase 2A(pro) are critical for viral RNA replication but not for cis- or trans-proteolytic cleavage. J. Gen. Virol. 82:397–408. 21. Liebig, H.-D., E. Ziegler, R. Yan, K. Hartmuth, H. Klump, H. Kowalski, D. Blaas, W. Sommergruber, L. Frasel, B. Lamphear, R. Rhoads, E. Kuechler, and T. Skern. 1993. Purification of two picornaviral 2A proteinases: interaction with eIF-4g and influence on in vitro translation. Biochemistry 32: 7581–7588. 22. Lu, H.-H., X. Li, A. Cuconati, and E. Wimmer. 1995. Analysis of picornavirus 2Apro proteins: separation of proteinase from translation and replication functions. J. Virol. 69:7445–7452. 23. Lu, H. H., L. Alexander, and E. Wimmer. 1995. Construction and genetic analysis of dicistronic polioviruses containing open reading frames for epitopes of human immunodeficiency virus type 1 gp120. J. Virol. 69:4797– 4806. 24. Martin, A., C. Wychowski, T. Couderc, R. Crainic, J. Hogle, and M. Girard.

1488

25.

26.

27.

28.

29.

30.

31.

32. 33.

TETERINA ET AL.

1988. Engineering a poliovirus type 2 antigenic site on a type 1 capsid results in a chimaeric virus which is neurovirulent for mice. EMBO J. 7:2839–2847. Mattion, N. M., P. A. Reilly, E. Camposano, S. L. Wu, S. J. DiMichele, S. T. Ishizaka, S. E. Fantini, J. C. Crowley, and C. Weeks-Levy. 1995. Characterization of recombinant polioviruses expressing regions of rotavirus VP4, hepatitis B surface antigen, and herpes simplex virus type 2 glycoprotein D. J. Virol. 69:5132–5137. Mattion, N. M., P. A. Reilly, S. J. DiMichele, J. C. Crowley, and C. WeeksLevy. 1994. Attenuated poliovirus strain as a live vector: expression of regions of rotavirus outer capsid protein VP7 by using recombinant Sabin 3 viruses. J. Virol. 68:3925–3933. Molla, A., A. V. Paul, M. Schmid, S. K. Jang, and E. Wimmer. 1993. Studies on dicistronic polioviruses implicate viral proteinase 2Apro in RNA replication. Virology 196:739–747. Moradpour, D., M. J. Evans, R. Gosert, Z. Yuan, H. E. Blum, S. P. Goff, B. D. Lindenbach, and C. M. Rice. 2004. Insertion of green fluorescent protein into nonstructural protein 5A allows direct visualization of functional hepatitis C virus replication complexes. J. Virol. 78:7400–7409. Morrison, J. M., and V. R. Racaniello. 2009. Proteinase 2Apro is essential for enterovirus replication in type I interferon-treated cells. J. Virol. 83:4412– 4422. Mueller, S., and E. Wimmer. 1998. Expression of foreign proteins by poliovirus polyprotein fusion: analysis of genetic stability reveals rapid deletions and formation of cardioviruslike open reading frames. J. Virol. 72:20–31. Murray, M. G., R. J. Kuhn, M. Arita, N. Kawamura, A. Nomoto, and E. Wimmer. 1988. Poliovirus type 1/type 3 antigenic hybrid virus constructed in vitro elicits type 1 and type 3 neutralizing antibodies in rabbits and monkeys. Proc. Natl. Acad. Sci. U. S. A. 85:3203–3207. Novak, J. E., and K. Kirkegaard. 1994. Coupling between genome translation and replication in RNA virus. Genes Dev. 8:1726–1737. Park, N., P. Katikaneni, T. Skern, and K. E. Gustin. 2008. Differential targeting of nuclear pore complex proteins in poliovirus-infected cells. J. Virol. 82:1647–1655.

J. VIROL. 34. Plutner, H., H. W. Davidson, J. Saraste, and W. E. Balch. 1992. Morphological analysis of protein transport from the ER to Golgi membranes in digitonin-permeabilized cells: role of the P58 containing compartment. J. Cell Biol. 119:1097–1116. 35. Rivera, C. I., and R. E. Lloyd. 2008. Modulation of enteroviral proteinase cleavage of poly(A)-binding protein (PABP) by conformation and PABPassociated factors. Virology 375:59–72. 36. Ryan, M. D., G. Luke, L. E. Hughes, V. M. Cowton, E. ten Dam, X. Li, M. L. L. Donnelly, A. Menrotra, and D. Gani. 2002. The aphtho- and cardiovirus “Primary” 2A/2B polyprotein “Cleavage,” p. 213–223. In B. L. Semler and E. Wimmer (ed.), Molecular biology of picornaviruses. ASM Press, Washington, DC. 37. Sverdlov, E. D., S. A. Tsarev, S. V. Markova, V. M. Rostapshov, T. L. Azhikina, I. P. Chernov, A. E. Gorbalenya, M. S. Kolesnikova, L. I. Romanova, N. L. Teterina, E. A. Tolskaya, and V. I. Agol. 1989. Insertion of short hepatitis virus A amino-acid sequences into poliovirus antigenic determinants results in viable progeny. FEBS Lett. 257:354–356. 38. Tang, S., R. van Rij, D. Silvera, and R. Andino. 1997. Toward a poliovirusbased simian immunodeficiency virus vaccine: correlation between genetic stability and immunogenicity. J. Virol. 71:7841–7850. 39. Teterina, N. L., A. E. Gorbalenya, D. Egger, K. Bienz, M. S. Rinaudo, and E. Ehrenfeld. 2006. Testing the modularity of the N-terminal amphipathic helix conserved in picornavirus 2C proteins and hepatitis C NS5A protein. Virology 344:453–467. 40. Teterina, N. L., M. S. Rinaudo, and E. Ehrenfeld. 2003. Strand-specific RNA synthesis defects in a poliovirus with a mutation in protein 3A. J. Virol. 77:12679–12691. 41. Teterina, N. L., W. D. Zhou, M. W. Cho, and E. Ehrenfeld. 1995. Inefficient complementation activity of poliovirus 2C and 3D proteins for rescue of lethal mutations. J. Virol. 69:4245–4254. 42. Yim, T. J., S. Tang, and R. Andino. 1996. Poliovirus recombinants expressing hepatitis B virus antigens elicited a humoral immune response in susceptible mice. Virology 218:61–70.