Demonstration of the Specificity of Poliovirus Encapsidation Using a Novel Replicon Which Encodes Enzymatically Active Firefly Luciferase

VIROLOGY 243, 1±11 (1998) ARTICLE NO. VY989046

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Demonstration of the Specificity of Poliovirus Encapsidation Using a Novel Replicon which Encodes Enzymatically Active Firefly Luciferase

Donna C. Porter,1 David C. Ansardi, Jun Wang, Sylvia McPherson,* Zina Moldoveanu, and Casey D. Morrow

Department of Microbiology; and *Center for AIDS Research, University of Alabama at Birmingham, Birmingham, Alabama 35294 Received September 22, 1997; returned to author for revision November 11, 1997; accepted January 14, 1998

The specificity of poliovirus encapsidation has been studied using a novel chimeric genome in which the gene encoding firefly luciferase has been substituted for the VP2±VP3±VP1 genes of the poliovirus capsid (P1) gene. Transfection of RNA transcribed in vitro from this genome resulted in a VP4-luciferase fusion protein which retained luciferase enzyme activity. Since the detection of enzyme activity was dependent upon replication of the transfected RNA genome, we refer to these genomes as replicons. The replicon encoding luciferase was encapsidated upon transfection of the genomic RNA into cells previously infected with a recombinant vaccinia virus, VV-P1, which encodes the poliovirus type 1 capsid proteins (P1). Infection of cells with each serial passage, followed by analysis of luciferase enzyme activity, revealed that encapsidated replicons could be detected at the first passage with VV-P1. Amplification of the titer of encapsidated replicons occurred upon serial passage with VV-P1, as evidenced by the high expression levels of luciferase enzyme activity following infection. Serial passage of the luciferase replicons with poliovirus type 1, 2, or 3 resulted in the trans encapsidation into the type 1, 2, or 3 capsids, respectively. In contrast, serial passage with bovine enterovirus, Coxsackievirus A21 or B3, or enterovirus 70 did not result in trans encapsidation, even though co-infection of cells with the replicon and different enteroviruses resulted in high-level expression of luciferase. The results of this study highlight the specificity of poliovirus encapsidation and point to the use of encapsidated replicons encoding luciferase as a reagent for dissecting elements of replication and encapsidation. © 1998 Academic Press

INTRODUCTION the viral polymerase (Kuhn and Wimmer, 1987; Harris et al., 1990; Palmenberg, 1990). Picornaviruses are small, plus-strand RNA viruses The development of an infectious cDNA clone for po- which include the enteroviruses (poliovirus, bovine en- liovirus facilitated studies to delineate the role of viral terovirus, Coxsackieviruses, and enterovirus 70), rhino- proteins and regions of the genome in replication (Kita- viruses, cardioviruses, and apthoviruses (Koch and Koch, mura et al., 1981; Racaniello and Baltimore, 1981). Early 1985). Picornavirus virions contain a single-stranded studies from Kaplan and Racaniello demonstrated that RNA genome of approximately 7500 bp in length. Upon regions of the P1 gene could be deleted without com- infection of cells, the viral genome is translated into a promising the capacity of the RNA to undergo self-repli- single polyprotein which is processed by virally encoded cation (Kaplan and Racaniello, 1988). Follow-up studies proteases to generate the viral proteins (Kitamura et al., demonstrated that insertion of foreign genes into the P1 1981; Semler et al., 1981). This long single polyprotein is region resulted in expression as long as the translational functionally divided into three distinct regions: P1 en- reading frame was maintained between the foreign gene codes the capsid or structural proteins, while P2 and P3 and the poliovirus P2 and P3 regions. Previous studies encode the nonstructural or replication proteins. An ini- from this laboratory have described the construction of tial cleavage of the polyprotein between the P1 and P2 self-replicating genomes, termed replicons, which en- regions is catalyzed by the 2A protease (2Apro) and oc- code the nonstructural proteins of poliovirus down- curs cotranslationally at a tyrosine±glycine amino acid stream from a foreign gene substituted for the P1 region pair (Nicklin et al., 1986, 1987; Toyoda et al., 1987). Sub- (Porter et al., 1993, 1995, 1996, 1997; Ansardi et al., 1994; sequent cleavages of the P2 and P3 region are catalyzed Moldoveanu et al., 1995; Anderson et al., 1996a,b). The by the protease 3Cpro at glutamine±glycine amino acid transfection of the replicon RNA into cells resulted in pairs. Cleavage of the P1 protein occurs in trans by the replication and expression of the gene substituted for the 3CD protein, which is a fusion between 3Cpro and 3Dpol, P1 region. Although the details of the mechanism and specificity of poliovirus encapsidation are not known, early studies 1 To whom correspondence and reprint requests should be ad- identified poliovirus genomes which contained deletions dressed. Fax: (205) 934-1580. in the P1 gene that could be encapsidated if serially

FIG. 1. Diagram of the replicon genome which contains the firefly luciferase gene. (A) Depiction of plasmid pT7-IC, which contains the complete poliovirus infectious cDNA clone with specific restriction sites indicated. This plasmid contains the T7 RNA polymerase promoter with a unique SalI restriction site for linearization of DNA before in vitro transcription of RNA. (B) Construction of pT7-IC-Luc. The complete gene encoding firefly luciferase (nucleotides 105±1754) was amplified by PCR and cloned between unique XhoI and SnaBI restriction sites in the modified poliovirus cDNA. passaged in the presence of wild-type poliovirus (Cole et licons correlated with the infectious dose. Co-infection of al., 1971; Cole and Baltimore, 1973a,b,c; Cole, 1975). cells with this replicon and other picornaviruses was These RNAs acted as defective interfering genomes (DI). used to analyze replication, as well as to examine the Further analysis of DI genomes revealed that the major- specificity of RNA encapsidation. The results of these ity of the deletions occurred in the VP2±VP3±VP1 genes studies are discussed with respect to the use of this of the P1 region (Nomoto et al., 1975). Percy et al. (1992) novel replicon as a means of dissecting and character- constructed a replicon which contained the gene encod- izing features of poliovirus replication and encapsida- ing chloramphenicol acetyltransferase (CAT) substituted tion. for VP4, VP2, and part of the VP3 genes. Transfection of this RNA into cells followed by infection with wild-type RESULTS poliovirus resulted in the encapsidation of the replicon Construction of a replicon encoding luciferase encoding CAT. A detailed analysis of the encapsidation of this replicon, though, was complicated by the pres- Previous studies from this laboratory have character- ence of the wild-type helper virus in these preparations ized the expression of foreign proteins from poliovirus- which prevented the isolation and characterization of a based replicon genomes. In these studies, the complete homogeneous population of encapsidated replicons. We HIV-1 Pr55gag gene sequences were substituted for ei- have demonstrated that transfection of replicon RNAs ther the entire P1 coding region or for the VP2, VP3, and into cells previously infected with a recombinant vaccinia VP1 coding sequences (Porter et al., 1995). Upon serial virus which expresses P1 (VV-P1) resulted in the encap- passage of in vitro transcribed RNA from each replicon sidation of the RNA by the P1 protein provided in trans. with VV-P1, a difference in the levels of encapsidated Serial passage of the encapsidated replicon in the pres- RNA was detected. These results suggested that al- ence of VV-P1 resulted in the generation of stocks of the though the VP4 capsid sequences are not required for encapsidated replicons. encapsidation, the presence of the VP4 gene sequences Previous studies from this laboratory have described might affect the efficiency of encapsidation. Based on the construction and characterization of encapsidated these results, we constructed replicons encoding the replicons encoding foreign proteins. Although immuno- luciferase gene substituted for the VP2±VP3±VP1 capsid precipitation was used to detect the encapsidated rep- sequences, which maintains the VP4 capsid gene se- licons, this technique was not sensitive enough for quences positioned 5Ј to the luciferase gene sequences. detailed analysis of poliovirus replication and encapsi- In order to subclone the luciferase gene, the poliovirus dation. In the present study, we have generated encap- cDNA pT7-IC was modified to contain a unique XhoI sidated replicons which encode the firefly luciferase restriction site between the VP4 and VP2 capsid genes marker protein. We demonstrate that infection of cells (Fig. 1A). Since poliovirus is translated as a single with this replicon results in the production of enzymati- polyprotein which is subsequently processed to gener- cally active luciferase protein. The level of luciferase ate mature proteins, a unique SnaBI restriction site was detected from cells infected with the encapsidated rep- introduced at nucleotide 3359, followed by coding se- ENCAPSIDATED REPLICONS ENCODING LUCIFERASE 3 quences for 8 amino acids constituting a cleavage site for the 2A protease: Thr-Lys-Asp-Leu-Thr-Thr-Tyr-Gly (cleavage site underlined). To construct the replicon pT7- Luc, PCR was used to amplify a 1.7-kb gene encoding luciferase (nucleotides 105±1754) which was then sub- cloned into the modified poliovirus cDNA between unique XhoI and SnaBI restriction sites (Fig. 1B). To determine that RNA from the replicon genomes encoding luciferase would replicate in transfected cells, RNA was transcribed in vitro from pT7-IC-Luc, as previously described (Porter et al., 1995). The RNA was then transfected into cells which had been previously infected for 2 h with VV-P1 at an m.o.i. of 10. After 6 h, the infected/transfected cells were metabolically labeled, and radiolabeled proteins were immunoprecipitated using anti-3Dpol antibodies. Examination of protein expression resulted in the detection of the poliovirus 3CD pro- FIG. 2. Comparison of the levels of encapsidation after RNA trans- tein (data not shown). We have shown previously that fection and serial passage with VV-P1. Cells infected with VV-P1 at an detection of 3CD by metabolic radiolabeling is indicative m.o.i. of 10 were transfected with RNA transcribed in vitro from the of RNA replication (Porter et al., 1995). High levels of pT7-Luc replicon genome. Encapsidated replicons were harvested luciferase enzyme activity were also detected from the from the transfected cells and a portion was analyzed for luciferase activity as measured by light units (designated passage 0). The remain- transfected cells (data not shown). der was used for serial passages in cells which had been previously infected with VV-P1 at an m.o.i. of 10 (passages 1±4). A portion of the Encapsidation of the replicon encoding luciferase lysates from the each of the serial passages was used to infect cells for the analysis of luciferase activity as described under Materials and In previous studies, we have demonstrated that RNA Methods. from poliovirus-based replicon genomes could be encapsidated if transfected into cells previously infected metabolically labeled at 6 h postinfection (Fig. 3A). Upon with a recombinant vaccinia virus, VV-P1, which ex- incubation of the cell lysates with anti-3Dpol antibodies, a presses the P1 protein (Ansardi et al., 1994; Porter et al., 72-kDa protein, which is consistent with the molecular 1993, 1995, 1996, 1997). Serial passage of encapsidated mass of the poliovirus 3CD protein, was detected (Fig. replicons in VV-P1-infected cells resulted in the genera- 3A, lane 2). Previous studies from this laboratory have tion of stocks of the encapsidated replicon. In order to shown that upon immunoprecipitation of radiolabeled generate a stock of encapsidated replicons encoding proteins from poliovirus-infected cells with anti-3Dpol an- luciferase (referred to as vIC-Luc), RNA transcribed in tibody under similar labeling conditions, the 3CD protein vitro was transfected into cells infected with VV-P1; the is predominantly detected (Porter et al., 1995). We next replicon was subsequently serially passaged in the pres- wanted to characterize the luciferase protein expressed ence of VV-P1. An aliquot of each passage was used to from the replicon. After incubation of replicon-infected infect HeLa H1 cells and the extracts were analyzed for cell lysates with anti-luciferase antibody, a protein with a luciferase activity (Fig. 2). We found that the levels of molecular mass of approximately 70 kDa was detected, luciferase activity from the transfected cells (passage 0) which is consistent with the size of the VP4-luciferase dropped (from 4,119,175 light units to approximately fusion protein (Fig. 3A, lane 4). We did not detect lucif- 231,450 light units) after one passage with VV-P1. By erase proteins of lower molecular mass, which might passage 2, an increase in the luciferase levels was indicate deletions of the luciferase gene within the rep- detected from the infection of cells; by passages 3 and 4, licon after serial passage. Analysis of the replicon virion the levels of luciferase detected in cells infected with RNA by reverse transcription PCR indicated that the encapsidated replicons had dramatically increased. We predominant DNA product amplified was the same size have further passaged this stock of encapsidated repli- as the initial VP4-luciferase gene (data not shown). con for approximately 50 serial passages to derive a Since previous studies have suggested that the lucif- large scale stock suitable for further analysis. erase enzyme is unstable in vivo (Andino et al., 1993), we examined the stability of the luciferase fusion protein Characterization of the encapsidated replicon expressed in replicon-infected cells by pulse-chase encoding luciferase analysis and immunoprecipitation using anti-luciferase To examine protein expression from the vIC-Luc en- antibodies (Fig. 3B). For these studies, cells were in- capsidated replicons, cells were infected with titered fected with vIC-Luc at an m.o.i. of 10 and pulse-labeled encapsidated replicon virus stock at an m.o.i. of 10 and for 30 min at 5 h postinfection. After pulse labeling, 4 PORTER ET AL.

FIG. 3. Analysis of protein expression from cells infected with replicons which encode luciferase. (A) Cells were infected at an m.o.i. of 10 with vIC-Luc encapsidated replicons and metabolically labeled with Tran35S-label. Cell lysates were immunoprecipitated with the indicated antibodies. The immunoprecipitates were analyzed by sodium dodecyl sulfate±polyacrylamide gel electrophoresis followed by fluorography/autoradiography. Lanes 1 and 3, mock-infected cells; lanes 2 and 4, cells infected with vIC-Luc replicons. (B) Cells infected with vIC-Luc replicons were pulse-labeled for 30 min, followed by the addition of complete media for 0.5-, 1-, and 2-h chase periods. Cell lysates were incubated with anti-luciferase antibodies and the immunoreactive proteins were analyzed by sodium dodecyl sulfate±polyacrylamide gel electrophoresis followed by fluorography/autoradiography. Lane 1, mock-infected cells; lanes 1±5, cells infected with vIC-Luc replicons. Molecular mass standards (MW) and positions of relevant proteins are indicated; kDa, kilodalton; ab, antibody. complete medium was added for 0.5-, 1-, and 2-h chase erase antibodies at various time points postinfection periods. After the designated chase periods, infected (Fig. 4B). Consistent with the results from Fig. 4A, the cells were processed as described for immunoprecipi- luciferase fusion protein was first detected at 4 h postin- tation with the anti-luciferase antibodies. A protein with a fection (lane 3). The levels of luciferase fusion protein molecular mass consistent with that of the VP4-lucif- continued to increase until a maximum level was de- erase fusion protein (ϳ70 kDa) was detected from in- tected at 6 to 8 h postinfection (Fig. 4B, lanes 4 and 5). fected cells which were pulse-labeled. After incubation These results are in contrast with the results in Fig. 4A, of the pulse-labeled cells in complete medium, decreas- which shows the peak levels of luciferase enzyme activ- ing amounts of the luciferase protein were detected as ity at 12 h postinfection. These differences might be a the chase times increased from 0.5 to 2 h (Fig. 3B, lanes result of a limitation of the anti-luciferase antibody react 2±5); the amount of the luciferase fusion protein did with all of the luciferase present in the samples. In other decrease over time but was still detectable even after the words, the immunoprecipitations were not quantitative, 2-h chase period. especially when the cell lysates contained large We next characterized the kinetics of protein expres- amounts of the VP4-luciferase present at later times sion from cells infected with the replicons encoding postinfection. luciferase. For this experiment, equivalent amounts of To further characterize the encapsidated replicons cells were infected with vIC-Luc replicons at an m.o.i. of which encode luciferase, we analyzed the correlation 10. Cells were also infected with another replicon, vIC- between the input virus and the measured light units Gag, which served as a negative control (Porter et al., following a defined time postinfection. Using a known 1993). Infected cells were then harvested for standard quantity of the encapsidated replicon which encodes luciferase enzyme assays as described under Materials luciferase, we determined the relationship between the and Methods at the time points indicated (Fig. 4A). Lu- amount of virus (infectious units) and the measured light ciferase activity was first detected at 4 h postinfection units following infection (Fig. 5). We found a linear rela- (measured in light units) from cells infected with vIC-Luc. tionship between the input infectious units of replicons The amount of detectable light units increased steadily and the measured light units from cells infected with 0.01 and peaked at 12 h postinfection. Between 12 and 24 h to 1 infectious unit per cell (see inset graph). The limit of postinfection, the number of light units detected from the the level of sensitivity was at approximately 0.01 i.u. per cells infected with vIC-Luc declined steadily, which is cell. probably a result of increasing cytopathic effects on the cells from the replicon infection. To determine if the trans encapsidation of replicons encoding luciferase enzymatic activity detected from lysates from cells infected with vIC-Luc replicons correlated with luciferase In a previous study, we have found that the replicon fusion protein expression, equivalent numbers of cells encoding HIV-1 gag could be trans-encapsidated when were infected at an m.o.i. of 10, metabolically labeled, serially passaged with poliovirus type 1 (Porter et al., 1993). and harvested for immunoprecipitation with anti-lucif- To demonstrate the presence of the replicon within the ENCAPSIDATED REPLICONS ENCODING LUCIFERASE 5

after infection. From the examination of light units, we determined that the levels of luciferase activity in cells co-infected with vIC-Luc and type 1 poliovirus (Fig. 6A) or Coxsackievirus B3 (Fig. 6B) were reduced by approximately 30±50% from that of cells infected with replicons alone. In contrast, the expression of luciferase from the replicon in the presence of poliovirus types 2 and 3, bovine enterovirus, and Coxsackievirus A21 was enhanced at least 2-fold. Surprisingly, the level of luciferase activity in cells co-infected with replicons and enterovirus 70 was increased by approximately 10-fold over that of cells infected with replicon alone. To determine whether the replicon could be encapsidated into virions following serial passage in the presence of a designated picornavirus, the replicons were co-infected with the viruses at equivalent moi (Fig. 7). Following complete cell lysis, the lysate was harvested and used to reinfect cells. The next day, the infected cells were harvested; the cell lysate was designated as passage 1 (P1). This procedure was repeated for an additional serial passage to generate passage 2 (P2). Material from P1 and P2 was then used to infect cells which were assayed for luciferase activity. Passage of the replicon alone resulted in an approximate 30-fold drop in luciferase activity after the first passage (P1); by passage 2, the levels of luciferase activity from vIC-Luc replicons alone were barely detectable (approximately 100 light units (Figs. FIG. 4. Determination of luciferase enzyme activity from cells infected with encapsidated replicons. (A) Cells were infected with vIC- 7A and 7B)). This result is consistent with the fact that Luc replicons or Gag replicons (negative control) at an m.o.i. of 10. At the time points indicated (hours postinfection), the infected cells were harvested and analyzed for luciferase enzyme activity as described under Materials and Methods. (B) Cells were infected with vIC-Luc and metabolically labeled with Tran35S-label. At specific time points, cell lysates were immunoprecipitated with anti-luciferase antibodies. Im- munoreactive proteins were separated by sodium dodecyl sulfate± polyacrylamide gel electrophoresis followed by fluorography/autoradiography. Lane 1, mock-infected cells; lanes 2±7, cells infected with vIC-Luc and harvested at 2, 4, 6, 8, 12, and 24 h postinfection. Molecular mass standards (MW) are indicated and luciferase fusion proteins are specified by arrowheads; kDa, kilodalton. stocks of poliovirus, we immunoprecipitated proteins from co-infected cells with antibody specific for the HIV-1 gag protein. However, detection of the replicon genome in the presence of the wild-type genome was complicated due to competition for radiolabel by both the replicon and wild- type RNAs resulting in a loss of sensitivity. The use of the replicon encoding luciferase allows for a more sensitive assessment of replication in cells co- infected with different picornaviruses. For these studies, FIG. 5. Correlation between initial virus multiplicity of infection and cells were co-infected with equivalent numbers of repli- luciferase activity from cells infected with vIC-Luc. Equivalent amounts cons encoding luciferase and poliovirus types 1, 2, or 3 of cells were infected with 0.01, 0.1, 1, and 10 infectious units (i.u.) per (Sabin strains) (Fig. 6A) or related picornaviruses, bovine cell of vIC-Luc replicons. At 6 h postinfection, the cells were harvested for luciferase assays as described under Materials and Methods. enterovirus, Coxsackievirus A21, Coxsackievirus B3, and (Inset) Expanded graph depicting the input infectious units of replicons enterovirus 70 (Fig. 6B). The replication of the replicon and luciferase activity from infections with 0.01 to 1.0 i.u. per cell; the was monitored by determination of luciferase activity 6 h level of detection was approximately 0.01 i.u. per cell. 6 PORTER ET AL.

FIG. 6. Replication of the replicon encoding luciferase in cells co-infected with other picornaviruses. Cells were co-infected with vIC-Luc replicons and poliovirus type 1, 2, or 3 (A), or bovine enterovirus, enterovirus 70, Coxsackievirus A21, or Coxsackievirus B3 (B) at an equivalent m.o.i. of 10 for both the replicons and the different picornaviruses. At 6 h postinfection, the co-infected cells were processed for luciferase enzyme activity as described under Materials and Methods. the replicons do not encode their own capsid proteins picornaviruses, we passaged vIC-Luc in the presence of and therefore undergo only one round of infection. Coxsackie A21 or B3 viruses, enterovirus 70, or bovine Upon co-infection of cells with vIC-Luc and poliovirus enterovirus (Fig. 7B). After P2 of the replicons with the types 1, 2, or 3, luciferase activity was detected from each different enteroviruses, the luciferase activity was below serial passage (Fig. 7A). We observed a reduction of the levels detected from the passage of replicons alone. about 50% in the levels of luciferase activity from P1 to These results demonstrate that although the replicons P2; the levels of activity though were still significantly can replicate and express luciferase in cells co-infected higher than background levels, demonstrating that the with the different picornaviruses, a mechanism exists replicons were trans-encapsidated upon passage with which excludes the replicons from trans encapsidation types 1 and 2 poliovirus. In contrast, the vIC-Luc pas- with the different picornaviruses. saged in the presence of poliovirus type 3 remained at high levels for each serial passage. Although the lucif- DISCUSSION erase replicons are capable of competing with all three serotypes of poliovirus for capsid proteins, the replicons In this report, we have described the construction and then appear to compete more efficiently for type 3 polio- characterization of a replicon which encodes firefly lucif- virus capsids. erase protein. Transfection of cells with the replicon To determine if trans encapsidation of the poliovirus resulted in the expression of enzymatically active lucif- replicons would occur upon serial passage with different erase. The replicons were encapsidated in poliovirions ENCAPSIDATED REPLICONS ENCODING LUCIFERASE 7

FIG. 7. trans encapsidation of replicon genomes encoding luciferase upon passage in cells co-infected with other picornaviruses. Cells were co-infected with vIC-Luc replicons and poliovirus type 1, 2, or 3 (A) or bovine enterovirus, enterovirus 70, Coxsackievirus A21, and Coxsackievirus B3 (B) at an equivalent m.o.i. of 10. The replicons were passaged with the wild-type picornaviruses for two successive passages. Cells were infected with material from each passage and processed for luciferase enzyme activity as described under Materials and Methods; P1, passage 1; P2, passage 2. by cotransfecting into cells previously infected with VV- Coxsackieviruses A21 or B3, enterovirus 70, or bovine P1. Serial passage in the presence of VV-P1 resulted in enterovirus. the generation of stocks of the encapsidated replicon. The availability of an encapsidated replicon which ex- Expression of luciferase protein correlated with the initial presses enzymatically active luciferase provides an impor- starting amount of the encapsidated luciferase replicon. tant experimental tool for dissecting elements of poliovirus The expression of luciferase from the replicon was de- RNA replication and encapsidation. Previous studies have creased by co-infection of cells with type 1 poliovirus and reported on the construction of poliovirus replicons which Coxsackievirus B3; luciferase expression was increased encode luciferase (Andino et al., 1993). Consistent with our in the presence of types 2 and 3 poliovirus, Coxsackievi- results, upon transfection into cells, the luciferase protein rus A21, bovine enterovirus, and enterovirus 70. The expressed from the replicon was enzymatically active and luciferase replicon could be encapsidated into poliovirus the levels correlated with replication of the genome. One type 1, 2, and 3 virions, but not trans encapsidated into difference between this study and the present study was 8 PORTER ET AL. that the luciferase protein was expressed as a fusion pro- encapsidation. Therefore, the generation of encapsi- tein with the poliovirus VP4 protein. Interestingly, the ex- dated stocks of the replicon might then represent a pression of luciferase as a fusion protein did not drastically selection of replicon RNAs which are efficiently encap- affect the enzymatic activity of the protein. To our knowl- sidated. Alternatively, it is possible that transfected RNA edge, the expression of enzymatically active firefly lucif- is localized to a different cellular compartment than the erase as a fusion protein has not been reported. Previous RNA which enters the cell via infection. While this differ- studies using a replicon encoding luciferase protein which ence does not affect RNA replication, it could influence was not a fusion protein with the VP4 capsid sequences subsequent encapsidation. indicated that this enzyme was unstable in cells (Andino et Early studies aimed at delineating the complexities of al., 1993). The results of our pulse±chase experiment, poliovirus RNA encapsidation have described the use of though, demonstrate that the enzyme is stable for at least poliovirus DI genomes for investigating the specificity of 2 h (and possibly longer). It is possible that the VP4 protein encapsidation (Kajigaya et al., 1985). In these studies, it stabilizes or protects the luciferase enzyme from degrada- was shown that DI genomes of poliovirus could be trans- tion in the replicon-infected cells. Alternatively, since the encapsidated upon passage with poliovirus types 1, 2, VP4 protein retains the signal for myristylation, this could and 3, but not with Coxsackievirus B1. Percy et al. (1992) target the VP4-luciferase fusion protein to a different intra- have previously described chimeric poliovirus genomes cellular compartment than luciferase expressed without the containing the CAT gene which were encapsidated upon VP4 protein. Additional studies using this replicon will be transfection into wild-type poliovirus-infected cells. How- needed to resolve these possibilities. ever, the drawback to this approach for analysis of en- Our study is unique since this is the first demonstra- capsidation is that the amount of RNA encoding CAT in tion that replicons encoding luciferase protein can be the mixed virus stocks is dependent not only upon the encapsidated and, through serial passage with VV-P1, efficiency of the transfection but also on the capacity of grown in quantities sufficient for studies of RNA replica- the CAT RNA to amplify in the presence of wild-type tion and encapsidation. Indeed, we have passaged the virus. In contrast, encapsidated replicons encoding lucif- encapsidated replicons described in the present study erase allow for a more direct evaluation of RNA encap- for over 50 serial passages in the presence of VV-P1 and sidation by the poliovirus P1 capsid protein. In the found no deletion of the luciferase gene. Infection of cells present study, we found that the replicon encoding lucif- with this replicon results in the synthesis of enzymati- erase could be detected in each of the serial passages cally active luciferase which can be detected from an with poliovirus types 1, 2, or 3. We detected the highest infection initiated with as few as 0.01 i.u. in our culture levels of luciferase enzyme activity from the replicons system. Using the encapsidated replicons, we have de- passaged with type 3 poliovirus. The expression of lucif- termined that the kinetics of luciferase expression, as erase from replicons in cells infected with either polio- measured by an enzyme activity assay, parallels the virus type 1 or type 2 was lower than that in the presence infectious life cycle of poliovirus. These results extend of type 3 poliovirus, although still significantly above our previous reports on the construction and character- background. We do not think these differences are due to ization of replicons encoding foreign genes which relied interference with the replication of the replicon RNA in on immunoprecipitation of the foreign protein using spe- the co-infected cells. In fact, luciferase expression was cific antibodies. enhanced in cells co-infected with replicon and either We have used the luciferase enzyme assay to charac- type 2 or type 3 poliovirus. Our results, then, suggest that terize the trans encapsidation of the replicon in the the replicons compete more efficiently for type 3 capsids presence of VV-P1. We found that the amount of encap- than for type 1 or 2 capsids. If this is true, we would sidated replicon RNA in the first passage after the initial predict that a recombinant vaccinia virus encoding type transfection in the presence of VV-P1 was extremely low. 3 poliovirus capsid proteins should allow for higher en- However, after 4 serial passages, the levels of luciferase capsidation efficiency than a virus encoding type 1 po- activity (and thus the amounts of encapsidated replicon) liovirus capsids. Experiments are ongoing to test this had increased to significant levels, which were compa- possibility. rable to those obtained from cells transfected with rep- Previous studies have suggested the possibility of licon RNA. Since the cells are infected with VV-P1 at a trans encapsidation of poliovirus genomes into other multiplicity of infection such that every cell is infected, picornavirus virions (Trautman and Sutmoller, 1971). If the results of these studies suggest that the transfected this had occurred, it might have important consequences replicon RNA, although competent for replication, does for the pathogenesis of poliovirus, as well as for picor- not interact with the encapsidation machinery as effi- naviruses in general. For example, immunization of in- ciently as RNA from the infection of cells with the encap- fants that have a concurrent infection with another picor- sidated replicon. This could be due to subtle changes in navirus (e.g., Coxsackie virus) with the poliovirus vaccine the replicon RNA during replication such that only a might result in trans encapsidation of the poliovirus ge- subpopulation of the transfected RNA is competent for nome into a Coxsackie virion which would not be neu- ENCAPSIDATED REPLICONS ENCODING LUCIFERASE 9 tralized by anti-poliovirus antibodies. Using the replicon Construction of replicons containing the luciferase genomes encoding luciferase, we examined the speci- gene ficity of replicon RNA encapsidation upon passage with To generate the replicons containing the luciferase several different enteroviruses. We found that co-infection of cells with the luciferase replicon and Coxsack- gene, a modified version of the poliovirus cDNA, pT7- ievirus A21 or enterovirus 70 resulted in an enhancement IC, was used (Porter et al., 1995). The replicon ge- of luciferase activity. Even so, it was clear from analysis nome, pT7-ENV (825 aa)/VP4, contains the HIV-1 en- of the passage of the replicons with the different picor- velope gene substituted for the poliovirus VP2, VP3, naviruses that trans encapsidation of the replicon RNA and VP1 capsid genes (Anderson et al., 1996a,b). This did not occur. Thus, although a common encapsidation genome contains a unique XhoI restriction site at signal exists between the three serotypes of poliovirus, it nucleotide 949, preceded by the VP4 capsid gene. The is distinct from that for Coxsackievirus A21 and B3, en- genome also contains a unique SnaBI restriction site terovirus 70, and bovine enterovirus. Although the spe- at the 3Ј end of the P1 region at nucleotide 3359, cific encapsidation signal for poliovirus is at present followed by 8 amino acids encoding a cleavage site for unknown, one candidate which could be involved in the 2A protease (cleavage site underlined): Thr-Lys- encapsidation is the 5Ј genome-linked protein VPg. In Asp-Leu-Thr-Thr-Tyr-Gly. support of this possibility, a previous study has identified To generate the replicon cDNA pT7-Luc, the luciferase mutations within the VPg coding region that resulted in gene (nucleotides 105±1754) was amplified from the RNA which replicated but did not produce infectious plasmid pGEM-luc (Promega) using the following prim- virus, suggesting that genomes with mutant VPg proteins ers: 5Ј-CTC-GAG-GAA-GAC-GCC-AAA-AAC-ATA-AAG-3Ј were incapable of being encapsidated (Reuer et al., and 5Ј-GTT-AAC-CAA-TTT-GGA-CTT-TCC-GCC-3Ј. The 1990). Although VPg is covalently linked to both plus- and amplified DNA fragment contains a unique XhoI site at minus-strand RNA, previous studies have demonstrated the 5Ј end and a HpaI site at the 3Ј end. The fragment that poliovirus RNA encapsidation is highly specific for was digested with XhoI and HpaI and ligated into the plus-strand RNA, suggesting that encapsidation is also plasmid pT7-ENV (825 aa)/VP4 after digestion of the dependent upon signals specific for positive-sense RNA plasmid with XhoI and SnaBI. The resulting plasmid pT7- (Novak and Kirkegaard, 1991). Thus, the encapsidation Luc, contains the complete gene encoding firefly lucif- signal for poliovirus RNA is most likely a complex con- erase positioned between nucleotides 949 and 3359, sisting of both RNA and protein determinants. The de- substituted in-frame for the VP2, VP3, and VP1 capsid velopment of an encapsidated replicon encoding lucif- genes of poliovirus. erase, as described in this study, provides an important experimental tool with which to now dissect the complexities of poliovirus RNA encapsidation. Encapsidation of replicons Previous studies from this laboratory have described MATERIALS AND METHODS the encapsidation and serial passage of replicon ge- Tissue culture medium was purchased from Gibco/ nomes containing foreign gene sequences using the BRL Company. The Tran35S-label (methionine and cys- recombinant vaccinia virus VV-P1 to provide P1 in trans teine) and methionine- and cysteine-free Dulbecco's (Porter et al., 1993, 1995, 1996, 1997). To encapsidate the modified Eagle's medium (DMEM) were purchased from replicon RNA, HeLa T4 cells were infected with 5 PFU/ ICN Biochemicals. Synthetic DNA primers were pur- cell of VV-P1, which expresses the poliovirus capsid chased from Gibco/BRL. The luciferase assay kit was precursor protein P1 (Ansardi et al., 1991). At 2 h postin- purchased from Promega. fection, the cells were then transfected with RNA transcribed in vitro from the pT7-Luc replicon genome (Porter Tissue culture cells and viruses et al., 1995). The cultures were harvested and encapsi- Cell monolayers of HeLa T4, BSC-40, or HeLa H1 cells dated replicons were collected by ultracentrifugation (obtained from ATCC) were grown in DMEM supple- (Porter et al., 1995). mented with 10% fetal bovine serum (Porter et al., 1995, To generate virus stocks of encapsidated replicons 1996). The recombinant vaccinia virus, VV-P1, which ex- encoding the luciferase protein, HeLa H1 cells were presses the P1 capsid precursor protein in infected cells infected with 10±20 PFU/cell of VV-P1 for 2 h. The cells has been previously described (Ansardi et al., 1991; Mor- were then infected with lysate from the VV-P1-infected row et al., 1994, 1995, 1996a,b). Poliovirus types 1, 2, and cells transfected with replicon RNA. The cells were har- 3 (Sabin strains) were obtained from Wyeth-Lederle, Vac- vested at 24 h postinfection and the supernatants, des- cines. Coxsackie B3, Coxsackie A21, enterovirus 70, and ignated pass 1 (P1), were then stored at Ϫ70°C or used bovine enterovirus were obtained from ATCC; viruses immediately for additional serial passages in the pres- were grown and titered in HeLa H1 cells. ence of VV-P1 (Porter et al., 1995). 10 PORTER ET AL.

Metabolic labeling and immunoprecipitation of viral lysis buffer [25 mM Tris±phosphate, pH 7.8, 2 mM DTT proteins (dithiothreitol), 2 mM DCTA (1, 2 diaminocyclohexane- N,N,NЈ,NЈ-tetraacetic acid), 10% glycerol, and 1% Triton In order to metabolically label proteins from infected/ X-100] and incubated at room temperature for 10 min. In a transfected cells, the cultures were starved for methio- separate tube, 10 ␮l of the room-temperature cell lysate nine/cysteine, at the times indicated, by incubation in was mixed with 100 ␮l of the luciferase assay reagent (20 DMEM minus methionine/cysteine followed by the addi- mM tricine, 1.07 mM (MgCO ) Mg(OH) ⅐ 5HO, 2.67 mM tion of Tran35S-label. For pulse±chase analysis, complete 3 4 2 2 MgSO , 0.1 mM EDTA, 33.3 mM DTT, 270 ␮M coenzyme A, medium was added to infected cells after labeling with 4 470 ␮M luciferin, 530 ␮M ATP, pH 7.8), and 50 ␮lofthe Tran35S-label, and infections were harvested at the indi- mixture was assayed for light detection using a luminom- cated time points. Infected cells were processed for eter (Monolight 2010, Analytical Luminescence Laborato- radioimmunoprecipitation as previously described and ries). The data are presented as relative light units. incubated with the appropriate antibodies (Porter et al., 1995). The immunoprecipitates were collected after the Co-infection of cells with replicons and other addition of protein A±Sepharose and analyzed by SDS± picornaviruses polyacrylamide gel electrophoresis as previously described (Porter et al., 1995). Cells were co-infected with the encapsidated replicons encoding luciferase and designated picornaviruses Estimation of virus titers of encapsidated replicon at an equivalent m.o.i. of 10. Viruses were allowed to virus stocks adsorb for1hat37°C. The inoculum was then removed and replaced with complete medium. The infection was Encapsidated replicons do not encode their own cap- allowed to proceed for 6 h, at which time cells were sid proteins and thus cannot form plaques on cell mono- processed for luciferase assays. For the encapsidation layers. To determine the virus titers of encapsidated experiments, the co-infected cultures were incubated for replicons encoding luciferase, cells were infected with 24 h, at which time a complete cytopathic effect had serial dilutions of the encapsidated replicon virus stocks; occurred. Following three freeze±thaw cycles and re- similar amounts of cells were infected with specific moval of cell debris, one-half of the supernatant was amounts of type 1 Mahoney poliovirus (as determined by used to reinfect cells; the other half was stored at plaque assay). At 5 h postinfection, cells were metabol- Ϫ70°C. After a 1-h adsorption, the inoculum was re- ically labeled, and specific viral proteins were immuno- moved and the infection continued for an additional 24 h. precipitated with anti-3Dpol antibodies. After separation The cultures were processed as before and repassaged of proteins by SDS±PAGE, the gels were subjected to on cells for a total of two passages. After two serial phosphorimagery to quantitate the radioactivity in each passages, cells were infected for 6 h with material from sample. A standard curve was generated from the quan- each passage and processed for luciferase assays. titation of protein expressed from known amounts (plaque-forming units) of type 1 Mahoney poliovirus. The ACKNOWLEDGMENTS values obtained from the quantitation of protein expressed from the luciferase replicons were then plotted The CFAR Molecular Biology Core is acknowledged for the construction of the luciferase replicon (AI-27767). We thank Li Hua Feng for on the poliovirus standard curve to estimate the titer. In expert technical assistance in the preparation of virus stocks of en- previous studies, we have shown that the titer of the capsidated replicons and John Eldridge of Wyeth-Lederle Vaccines for encapsidated replicon virus stocks, referred to as infec- the Sabin strains of poliovirus. This work was supported by a grant from tious units (i.u.), correlated with plaque-forming units National Institutes of Allergy and Infectious Disease (AI 25005 and AI from wild-type poliovirus (Ansardi et al., 1994; Morrow et 28147) (C.D.M.). al., 1994, 1996a,b). REFERENCES

Luciferase assays Anderson, M. J., Moldoveanu, Z., Fletcher, T. M., III, McPherson, S., Porter, D. C., and Morrow, C. D. (1996a). Characterization of the For luciferase assays, equivalent amounts of cells were expression and immunogenicity of poliovirus replicons which en- infected with titered virus stocks of encapsidated replicons code simian immunodeficiency virus SIVmac239 Gag or envelope SU proteins. AIDS Res. Hum. Retroviruses 13, 53±62. with designated moi. At 5±7 h postinfection, the culture Anderson, M. J., Porter, D. C., Fultz, P. N., and Morrow, C. D. (1996b). medium was removed, and the infected cells were washed Poliovirus replicons that express the Gag or the envelope surface

2ϫ with PBS (phosphate-buffered saline). The cells were protein of simian immunodeficiency virus SIVsmm PBj14. Virology 219, then scraped from the tissue culture dishes, resuspended 140±149. in 1 ml PBS, and pelleted at 14,000g for 2 min at room Andino, R., Rieckhof, E., Achacoso, P. L., and Baltimore, D. (1993). Poliovirus RNA synthesis utilizes an RNP complex formed around the temperature. After discarding the supernatant, the cell pel- 5Ј-end of viral RNA. EMBO J. 12, 3587±3598. lets were used for assays immediately or stored at Ϫ70°C. Ansardi, D. C., Moldoveanu, Z., Porter, D. C., Walker, D. E., Conry, R. M., For assays, the cell pellets were resuspended in 100 ␮l1ϫ LoBuglio, A. F., McPherson, S., and Morrow, C. D. (1994). Character- ENCAPSIDATED REPLICONS ENCODING LUCIFERASE 11

ization of poliovirus replicons encoding carcinoembryonic antigen. Nicklin, M. J. H., KraÈusslich, H. G., Toyoda, H., Dunn, J. J., and Wimmer, Cancer Res. 54, 6359±6363. E. (1987). Poliovirus polypeptide precursors: expression in vitro and Ansardi, D. C., Porter, D. C., and Morrow, C. D. (1991). Coinfection with processing by exogenous 3C and 2A proteinases. Proc. Natl. Acad. recombinant vaccinia viruses expressing poliovirus P1 and P3 pro- Sci. USA 84, 4002±4006. teins results in polyprotein processing and formation of empty capsid Nicklin, M. J. H., Toyoda, H., Murray, M. G., and Wimmer, E. (1986). structures. J. Virol. 65, 2088±2092. Proteolytic processing in the replication of polio and related viruses. Cole, C. N. (1975). Defective interfering (DI) particles of poliovirus. Prog. BioTechnology 4, 33±42. Med. Virol. 20, 180±207. Nomoto, A., Jacobson, A., Lee, Y. F., Dunn, J., and Wimmer, E. (1975). Cole, C. N., and Baltimore, D. (1973a). Defective interfering particles of Defective interfering particles of poliovirus: mapping of deletion and poliovirus. II. Nature of the defect. J. Mol. Biol. 76, 325±361. evidence that the deletions in the genome of DI (1), (2) and (3) are Cole, C. N., and Baltimore, D. (1973b). Defective interfering particles of located in the same region. J. Mol. Biol. 128, 179±196. poliovirus. III. Interference and enrichment. J. Mol. Biol. 76, 345±361. Novak, J. E., and Kirkegaard, K. (1991). Improved method for detecting Cole, C. N., and Baltimore, D. (1973c). Defective interfering particles of poliovirus negative strands used to demonstrate specificity of posi- poliovirus. IV. Mechanisms of enrichment. J. Virol. 12, 1414±1426. tive-strand encapsidation and the ratio of positive to negative stand Cole, C. N., Smoler, D., Wimmer, E., and Baltimore, D. (1971). Defective in infected cells. J. Virol. 65, 3384±3387. interfering particles of poliovirus. I. Isolation and physical properties. Palmenberg, A. C. (1990). Proteolytic processing of picornaviral J. Virol. 7, 478±485. Harris, K. S., Hellen, C. U. T., and Wimmer, E. (1990). Proteolytic pro- polyprotein. Annu. Rev. Microbiol. 44, 603±623. cessing in the replication of picornaviruses. Semin. Virol. 1, 323±333. Percy, N., Barclay, W. S., Sullivan, M., and Almond, J. W. (1992). A Kajigaya, S., Arakawa, H., Kuge, S., Koi, T., Imura, N., and Nomoto, A. poliovirus replicon containing the chloramphenicol acetyl-trans- (1985). Isolation and characterization of defective-interfering parti- ferase gene can be used to study the replication and encapsidation cles of poliovirus Sabin 1 strain. Virology 142, 307±316. of poliovirus RNA. J. Virol. 66, 5040±5046. Kaplan, G., and Racaniello, V. R. (1988). Construction and characteriza- Porter, D. C., Ansardi, D. C., Choi, W. S., and Morrow, C. D. (1993). tion of poliovirus subgenomic replicons. J. Virol. 62, 1687±1696. Encapsidation of genetically engineered poliovirus mini-replicons Kitamura, N., Semler, B. L., Rothberg, P. G., Larsen, G. R., Adler, C. J., which express HIV-1 gag and pol proteins upon infection. J. Virol. 67, Dorner, A. J., Emini, E. A., and Hanecak, R. (1981). Primary structure, 3712±3719. gene organization and polypeptide expression of poliovirus RNA. Porter, D. C., Ansardi, D. C., and Morrow, C. D. (1995). Encapsidation of Nature (London) 291, 547±553. poliovirus replicons encoding the complete human immunodefi- Koch, F., and Koch, G. (1985). ``The Molecular Biology of Poliovirus.'' ciency virus type1 gag gene using a complementation system which Springer-Verlag, Vienna. provides the P1 capsid protein in trans. J. Virol. 69, 1548±1555. Kuhn, R. J., and Wimmer, E. (1987). The replication of picornaviruses. In Porter, D. C., Melsen, L. R., Compans, R. W., and Morrow, C. D. (1996). ``Molecular Biology of Positive Strand RNA Viruses'' (D. J. Rowlands, Release of virus-like particles from cells infected with poliovirus M. A. Maho, and B. W. J. Mahy, Eds.), pp. 17±51. Academic Press, replicons which express human immunodeficiency virus type 1 Gag. London. J. Virol. 70, 2643±2649. Moldoveanu, Z., Porter, D. C., Lu, A., McPherson, S., and Morrow, C. D. Porter, D. C., Wang, J., Moldoveanu, Z., McPherson, S., and Morrow, (1995). Immune responses induced by administration of encapsi- C. D. (1997). Immunization of mice with poliovirus replicons express- dated poliovirus replicons which express HIV-1 gag and envelope ing the C-fragment of tetanus toxin protects against lethal challenge proteins. Vaccine 13(11), 1013±1022. with tetanus toxin. Vaccine 15(3), 257±264. Morrow, C. D., Ansardi, D. C., and Porter, D. C. (1996a). Methods for the Racaniello, V., and Baltimore, D. (1981). Cloned poliovirus complemen- use of poliovirus vectors for gene delivery. In ``Gene Therapy Proto- tary DNA is infectious in mammalian cells. Science 214, 916±919. cols: Methods in Molecular Biology'' (P. Robbins, Ed.), pp. 103±116. Reuer, Q., Kuhn, R. J., and Wimmer, E. (1990). Characterization of Humana, Totowa, NJ. poliovirus clones containing lethal and nonlethal mutations in the Morrow, C. D., Moldoveanu, Z., Anderson, M. J., and Porter, D. C. (1996b). Poliovirus replicons as a vector for mucosal vaccines. In genome-linked protein VPg. J. Virol. 64, 2967±2975. ``Mucosal Vaccines'' (H. Kiyono, J. McGhee, and P. Ogra, Eds.), pp. Semler, B. L., Hanecak, R., Anderson, C. W., and Wimmer, E. (1981). 137±144. Academic Press. Cleavage sites in the polyopeptide precursors of poliovirus protein Morrow, C. D., Porter, D. C., and Ansardi, D. C. (1995). New methods to P2-X. Virol. 114, 589±594. study poliovirus assembly and encapsidation of genomic RNA. In Toyoda, J., Yang, C., Takeda, N., Nomoto, A., and Wimmer, E. (1987). ``Methods in Molecular Genetics, Vol. 7, Viral Gene Techniques'' (K. W. Analysis of RNA synthesis of type 1 poliovirus by using an in vitro Adolph, Ed.), Chap. 19, pp. 299±314. Academic Press. Orlando, FL. molecular genetic approach. J. Virol. 61, 2816±2822. Morrow, C. D., Porter, D. C., Ansardi, D. C., Moldoveanu, Z., and Fultz, P. N. Trautman, R., and Sutmoller, P. (1971). Detection and properties of a (1994). New approaches for mucosal vaccines for AIDS: Encapsidation genomic masked viral particle consisting of foot-and-mouth disease and serial passage of poliovirus replicons which express HIV-1 proteins virus nucleic acid in bovine enterovirus protein capsid. Virology 44, upon infection. AIDS Res. Hum. Retroviruses 10, S61±S66. 537±543.