An efficient trans complementation system for in vivo replication of defective poliovirus mutants

ABSTRACT The picornavirus genome encodes a large, single polyprotein that is processed by viral proteases to form an active replication complex. The replication complex is formed with the viral genome, host proteins, and viral proteins that are produced/translated directly from each of the viral genomes (viral proteins provided in cis). Efficient complementation in vivo of replication complex formation by viral proteins provided in trans, thus exogenous or ectopically expressed viral proteins, remains to be demonstrated. Here, we report an efficient trans complementation system for the replication of defective poliovirus (PV) mutants by a viral polyprotein precursor in HEK293 cells. Viral 3AB in the polyprotein, but not 2BC, was processed exclusively in cis. Replication of a defective PV replicon mutant, with a disrupted cleavage site for viral 3Cpro protease between 3Cpro and 3Dpol (3C/D[A/G] mutant) could be rescued by a viral polyprotein provided in trans. Only a defect of 3Dpol activity of the replicon could be rescued in trans; inactivating mutations in 2CATPase/hel, 3B, and 3Cpro of the replicon completely abrogated the trans-rescued replication. An intact N-terminus of the 3Cpro domain of the 3CDpro provided in trans was essential for the trans-active function. By using this trans complementation system, a high-titer defective PV pseudovirus (PVpv) (>107 infectious units per mL) could be produced with the defective mutants, whose replication was completely dependent on trans complementation. This work reveals potential roles of exogenous viral proteins in PV replication and offers insights into protein/protein interaction during picornavirus infection. IMPORTANCE Viral polyprotein processing is an elaborately controlled step by viral proteases encoded in the polyprotein; fully processed proteins and processing intermediates need to be correctly produced for replication, which can be detrimentally affected even by a small modification of the polyprotein. Purified/isolated viral proteins can retain their enzymatic activities required for viral replication, such as protease, helicase, polymerase, etc. However, when these proteins of picornavirus are exogenously provided (provided in trans) to the viral replication complex with a defective viral genome, replication is generally not rescued/complemented, suggesting the importance of viral proteins endogenously provided (provided in cis) to the replication complex. In this study, I discovered that only the viral polymerase activity of poliovirus (PV) (the typical member of picornavirus family) could be efficiently rescued by exogenously expressed viral proteins. The current study reveals potential roles for exogenous viral proteins in viral replication and offers insights into interactions during picornavirus infection.

Similar to the completely processed viral proteins, processing intermediates derived from P2 or P3 (i.e., 2BC, 3AB, 3CD pro , etc.) are also produced from the polyprotein.Other processing intermediates that span these precursors (e.g., 2C3AB, 2ABC3AB) are produced but only in trace amounts (18,19).Both processing intermediates and fully processed proteins have critical roles in replication, such as remodeling of the endoplas mic reticulum by 2BC and 3A (20), stimulation of 3D pol activity by 3AB (21)(22)(23), efficient cleavage of P1 (4), switching of the viral genome from translation to RNA replication (24,25), and stimulation of uridylylation of 3B by 3CD pro (26).Disruption of each of the processing intermediate was lethal for infectivity (27), indicating the importance of the processing intermediates in replication.
RNA replication of PV predominantly depends on viral proteins provided in cis (28)(29)(30).Processing is critically controlled by cis cleavage (i.e., cleavage of the polyprotein by 2A pro /3C pro /3CD pro /3ABC pro , which are coded in the target polyprotein itself, thus authentic self-cleavage) and by trans cleavage (i.e., cleavage of polyprotein by the 2A pro / 3C pro /3CD pro /3ABC pro , which are coded in polyproteins other than the target polyprotein).Disruptions of the polyprotein downstream of the 2B (27) or introduction of mutations in the 2B2CP3 region without affecting the protease activity (31)(32)(33)(34)(35) interfered with cis cleavage, resulting in aberrant processing and lethality of virus, underscoring the importance of cis cleavage in picornavirus replication and of an intact 2B2CP3 precursor.Recently, involvement of a host PI4KB/OSBP family I (OSBP and OSBP2/ORP4) pathway (16,36) in processing of 3AB was discovered (37)(38)(39).The pathway is essential for the development of a viral replication organelle (RO) (40), formation of the replication complex and synthesis of viral plus-strand RNA (41)(42)(43), and enhancement of viral growth and infectivity (44).PI4KB and OSBP family I are the target of anti-EV drug candidates (36,(45)(46)(47)(48)(49)(50), suggesting that polyprotein processing is a promising target for antiviral development.Enigmatically, processing of 3AB occurs inefficiently especially in the early phase of replication (51); only 4% of 3AB is processed to provide 3A and 3B for RNA synthesis.Interestingly, resistant PV mutants against PI4KB/OSBP inhibitors have mutations in the 3A region, which enhance the processing of 3AB (37,40), suggesting that cleavage of 3AB or the final products (3A and 3B) could be a target of exogenous intervention to control infection.
Here, I have developed an efficient trans-rescue system for in vivo replication of defective PV replicons targeting polyprotein processing.I established cell lines that could conditionally express an entire precursor protein and analyzed the cis and trans roles of viral proteins.I found that cleavage of 3AB occurs exclusively in cis in the polyprotein.Among the defective PV replicon mutants examined, only a mutant (3C/D[A/G]) that has a disrupted cleavage site between 3C pro and 3D pol showed an efficient replication rescued in trans and produced pseudovirus at a high titer, comparable to that of the wild-type (WT) replicon.I identified an intact 3CD pro as the minimal protein required for the trans rescue.

Doxycycline (DOX)-inducible expression of PV non-structural proteins
To analyze the potential role of PV proteins provided in trans in replication, I generated a HEK293 cell line (Tet-AG-PV-2B2CP3[WT]) that could express a polyprotein of PV non-structural proteins (2BC3ABCD) in the presence of DOX (1 mg/L) as a form of an N-terminally Azami green (AG)-fused protein, which allowed a high expression level of protein (52) (Fig. 1).Expression of the AG-fused PV polyprotein caused rounded morphology of the cells, similar to the cytopathic effect observed in PV-infected cells.Localization of AG, which is cleaved from the polyprotein by 3C pro , was in the nucleus and cytoplasm (Fig. 1A).In the presence of a reversible 3C pro inhibitor GC376 (53), a normal morphology of the cells was retained even after protein expression.In the presence of GC376, AG, which remained attached to the polyprotein, showed a dot-like localization in the cytoplasm, in contrast to that of free AG.Processing of the polyprotein was partially inhibited in the presence of GC376 or rupintrivir (also known as AG7088, an irreversible 3C pro inhibitor) (54), and the polyprotein (161 kDa) remained as an intact precursor (Fig. 1A).In the absence of, or at lower concentrations of, the 3C pro inhibitors, processing intermediates (P3/3ABCD, 3CD pro ) and a fully processed viral protein (3D pol ) were detected by an anti-3D pol antibody similar to those present in PV-infected cells (Fig. 1B).These results suggested that Tet-AG-PV-2B2CP3(WT) cells could express a PV polyprotein precursor, which was subsequently processed in the absence of 3C pro inhibitors.

trans rescue of replication of defective PV replicons
Next, I attempted to trans rescue defective PV replicon mutants in Tet-AG-PV-2B2CP3(WT) cells (Fig. 2A).Tet-AG-PV-2B2CP3(WT) cells were first treated with DOX and GC376 to induce expression of the unprocessed polyprotein.The cells were then washed to remove DOX and GC376, and were transfected with RNA transcripts of the defective PV replicon mutants.Replication was monitored by firefly luciferase or mCherry reporters encoded in the replicons.PV replicon mutants that have disrupted cleavage sites (aa substitution of the conserved Q/G to A/G in the cleavage site) for 3C pro between the viral proteins (37,55,56), to inhibit the production of fully processed viral proteins (i.e., 2A pro , 2B, 2C ATPase/hel , 3A, 3B, 3C pro , and 3D pol ), or that lack each of the viral genes were examined (Fig. 2B).
In Tet-AG-PV-2B2CP3(WT) cells without DOX treatment, only the WT and Δ2A mutant could replicate (Fig. 3A), consistent with a previous report of a viable 2A pro deletion mutant (57).Other mutants produced only basal levels of signals in the cells, which are derived from initial protein synthesis from transfected RNA transcripts and could not be suppressed in the presence of a PV replication inhibitor guanidine hydrochloride (GuHCl) (a 2C ATPase/hel inhibitor), suggesting no replication occurred.In Tet-AG-PV-2B2CP3(WT) cells treated with DOX and GC376 before RNA transfection, interestingly, a mutant (3C/D[A/G]) could replicate as well as the WT and Δ2A mutant.Replication of 2B/C(A/G) and Δ2B mutants could be detected, albeit at low levels (two-to fourfold increase compared to that in GuHCl-treated cells) (Fig. 3B).This suggested that each viral protein (i.e., 2B, 2C ATPase/hel , 3A, 3B, and 3C pro ) is essential for replication and cannot be trans complemented.In addition, replication of a defective PV replicon, which could express 3CD pro but not 3C pro and 3D pol , could be efficiently rescued in trans by the precursor protein.

cis role of viral proteins in the replication of a PV 3C/D(A/G) mutant
Next, to determine the roles of viral proteins provided in cis in PV replication, mutations that inactivate activities of viral proteins were introduced into the 3C/D(A/G) mutant (Fig. 4A): 2C-K153A aa substitution that disrupts ATPase activity of 2C ATPase/hel (6), 3B-Y3F aa substitution that inhibits uridylylation of the 3B protein (58), 3C-C147A aa substitution that inactivates the protease activity of 3C pro /3CD pro by disruption of the catalytic triad (59), and 3D-D328N/D329N aa substitutions that inactivate the polymerase activity of 3D pol (60).In Tet-AG-PV-2B2CP3(WT) cells treated with DOX and GC376 before RNA transfection, no replication was observed for mutants with inactivated 2C ATPase/hel , 3B, or 3C pro /3CD pro (Fig. 4B).In contrast, a mutant with inactivated 3D pol (3C/D[A/G]-3D-D328N/ D329N) replicated as well as the WT and parental 3C/D(A/G) mutant.This suggested that the polymerase activity of the 3C/D(A/G) mutant can be rescued in trans and that cis activities of viral proteins (2C ATPase/hel , 3B, and 3C pro /3CD pro ) are essential for replication.

Identification of minimal viral proteins required for trans rescue of the replication of defective PV replicons
Because the polymerase activity seemed to be the target of the trans rescue of the 3C/D(A/G) mutant, I generated a HEK293 cell line (Tet-AG-PV-3CD[WT]) that expresses PV 3CD pro as a form of an N-terminally AG-fused protein in the presence of DOX, which could be self-processed into 3D pol with an intact N-terminus, which is essential for the polymer ase activity (61) (Fig. 5A).Fortuitously, a cell line (Tet-AG-PV-3CD[Δ4-5 aa]) that expresses a 3CD pro variant (deletion of aa 4 and 5 of 3C pro , possibly derived from mutations in the oligo DNAs used for the cloning) was also produced.Both cell lines expressed AG-3CD pro and a processing intermediate 3CD pro and 3D pol in the presence of DOX (Fig. 5B).Surpris ingly, replication of the 3C/D(A/G) mutant was rescued only in Tet-AG-PV-3CD(WT) cells but not in Tet-AG-PV-3CD(Δ4-5 aa) cells (Fig. 5C), suggesting that an intact 3CD pro is essential and sufficient for the trans rescue.Expression of N-terminally AG-fused 3D pol , which could be processed by 3CD pro of the 3C/D(A/G) mutant to give an intact 3D pol , could only partially rescue the replication of this mutant in trans (Fig. 6).The WT replicon replicated to similar levels in both cell lines, suggesting that the effect was specific to the trans-rescued replication.To further analyze the specificity of PV 3CD pro provided in trans, I attempted to rescue the replication of enterovirus 71 (EV-A71), which belongs to the Enterovirus A species, thus another species in EV.A mutation that disrupts the 3C pro cleavage site between 3C pro and 3D pol was introduced in an EV-A71 replicon (EV-A71-3C/D[A/G] mutant), and then replication in cells was analyzed (Fig. 7).In contrast to the PV replicon mutant, replication of the EV-A71-3C/D(A/G) mutant was not rescued in trans by a PV polyprotein or 3CD pro .These results suggested that the intact 3C pro domain of 3CD pro and viral species-specific interaction of 3CD pro are essential for the trans-active function.

trans rescue of replication of defective PV replicons by 3CD pro
To clarify the role of 3C pro in the trans-active function of 3CD pro , I introduced aa substitu tions into the 3C pro domain of 3CD pro , focusing on those involved in the binding to viral RNA and phospholipids (4,24,26,62).I introduced 3C-R13N, 3C-K82N, and 3C-R84S aa substitutions into 3CD pro , which abolish the binding to viral RNA and phospholipids (24,26,(62)(63)(64).I also introduced a mutation to disrupt the 3C pro cleavage site between AG and the 3C pro region to analyze the effect of potential steric hindrance for the interaction with target molecules around the N-terminus of 3CD pro on the trans rescue.I generated HEK293 cell lines that could express these 3CD pro variants and analyzed the trans-active function for PV1(Fluc) pv (3C/D[A/G]) infection (Fig. 8A).Unexpectedly, the 3CD pro variants (3C-R13N, 3C-K82N, and 3C-R84S) efficiently rescued the infection in trans, higher than the 3CD(Δ4-5 aa) variant.In contrast, the 3CD pro variants with uncleava ble AG could not substantially rescue the infection in trans irrespective of the deletion in the N-terminal region of 3C pro .This suggested that the integrity of the N-terminus of 3C pro , but not the binding activity to viral RNA or phospholipids, is essential for the transactive function of 3CD pro .
To analyze the effect of N-terminal modification of 3CD pro on the trans activity, I generated HEK293 cell lines that could simultaneously express 3CD pro (WT) or 3CD(Δ4-5 aa) variant with a PV polyprotein (2BC3ABCD), which lacks 3C pro protease activity with a 3C-C147A aa substitution (59) (Tet-AG-PV-2B2CP3[3C-C147A]+AG-PC-3CD[WT] cells and Tet-AG-PV-2B2CP3[3C-C147A]+AG-PC-3CD[Δ4-5 aa] cells) (Fig. 8B).Due to the lack of 3C pro activity, the polyprotein remained intact without producing processing intermedi ates and could not rescue the replication of the 3C/D(A/G) mutant in trans (Fig. 9).Simultaneous expression of this polyprotein with 3CD pro (WT) or 3CD pro (Δ4-5 aa) variant showed similar profiles of the precursor and processing intermediates AG-2BC, 2BC, 2C, and 3AB; interestingly, no 3A but only 3AB was observed in these cells, suggesting that 3AB is the target of cis cleavage (Fig. 8B).Both 3A and 3AB were observed in Tet-AG-PV-2B2CP3(WT) cells or in PV pv -infected cells.The 3AB protein was detected in the most diluted lysates of infected cells, at levels lower than those in the Tet-AG-PV-2B2CP3(3C-C147A)+AG-PC-3CD(WT) cells and Tet-AG-PV-2B2CP3(3C-C147A)+AG-PC-3CD(Δ4-5 aa) cells, confirming the absence of 3A in these cells.These results suggested that 3CD pro (WT) and 3CD pro (Δ4-5 aa) have similar trans cleavage activities and that processing of 3AB occurs exclusively in cis in the context of the polyprotein.

Production of PV pseudovirus (PV pv ) with defective PV replicons
To substantiate the observed high replication level of the defective PV replicons, I attempted to produce PV pv with the defective PV replicons and PV capsid proteins (65) (Fig. 10).Replication-competent PV WT replicon could produce PV pv to a titer (10 7 to 10 8 infectious units [IU] per mL) comparable to that of PV virus (65,66).
To produce PV pv , a plasmid expression vector for type 1 PV capsid proteins was transfected into Tet-AG-PV-3CD(WT) cells in the presence of DOX and GC376 (Fig. 10A).were harvested to determine the titer of PV pv .Collected PV pv was inoculated into Tet-AG-PV-3CD(WT) cells or Tet-AG-PV-3CD(Δ4-5 aa) cells (a control) pre-treated with DOX (or no DOX treatment as a control) before the infection, and then the signals of reporters (fluorescence of mCherry or luciferase activity) in the infected cells were analyzed (Fig. 10B).I observed fluorescence of mCherry in the cells infected with PV1(mCherry) pv (WT), irrespective of the DOX treatment and the cell types.In contrast, substantial replication of PV1(mCherry) pv (3C/D[A/G]) was observed only in Tet-AG-PV-3CD(WT) cells after pretreatment with DOX.The observed titer of PV1(mCherry) pv (3C/D[A/G]) was about 10 7 IU/mL, thus similar to that of PV1(mCherry) pv (WT).Replication in Tet-AG-PV-3CD(Δ4-5 aa) cells was significantly suppressed; about 1/100-fold or 1/100,000-fold lower than that of the WT replicon, in the presence or absence of DOX, respectively (Fig. 11A), supporting the weak trans-active function of 3CD pro (Δ4-5 aa).Infectivity of PV1(Fluc) pv (3C/D[A/G]) showed similar cell-type specificity and dependency on DOX treatment to that of PV1(mCherry) pv (3C/D[A/G]).To further confirm the presence of PV pv in the preparations, I performed neutralization tests for PV1(mCherry) pv (3C/D[A/G]) with anti-PV antisera (Fig. 11B).PV1(mCherry) pv (3C/D[A/G]) was incubated with type-specific anti-PV antibodies (i.e., anti-PV1, PV2, PV3 standard antisera) and then inoculated into Tet-AG-PV-3CD(WT) cells, pre-treated with DOX.Infection of PV1(mCherry) pv (3C/D[A/G]) was inhibited by pre-incubation with anti-PV1 antiserum, but not with anti-PV2 or PV3 antisera, suggest ing that the type 1 PV antigenicity is retained on PV1(mCherry) pv (3C/D[A/G]) similar to PV pv produced with the WT replicon (66,67).These results suggested that replication of defective PV replicons could be efficiently rescued in trans and could allow production of a high titer PV pv .

DISCUSSION
The importance of viral proteins provided in cis for replication of PV was initially sugges ted from analysis of defective interfering (DI) particles (28), which have in-frame dele tions in the capsid-coding region (P1 region) of the genome and retain an intact nonstructural protein coding region (P2P3 region) (29).These results suggested that the functions of viral proteins encoded in the P2P3 region could not be complemented by exogenous viral proteins, thus in trans.The cis and trans roles of the viral proteins in replication were intensively studied in the 1990s by trans complementation (or trans rescue) of replication of defective PV mutants (30-33, 37, 68-73).Main conclusions drawn from these studies include (i) trans rescue of defective PV mutants is inefficient, and (ii) a large intact precursor of the non-structural proteins is required for replication.Besides viral proteins, conserved RNA structures, encoded in the P2P3 region, were identified, including the CRE, RNase L ciRNA, α, and β (26,(74)(75)(76)(77)(78).The CRE is required for replication in cis as the template for uridylylation of 3B (26,79), confirming the cis role of the P2P3 region.These properties/roles of non-structural proteins and RNA structures in viral replication are generally conserved in picornavirus (80)(81)(82)(83)(84).
To elucidate the role of processing in viral replication, I established cell lines that could conditionally express a large precursor of PV non-structural proteins (2BC3ABCD) and performed trans rescue of defective PV replicons in cells instead of using helper virus/replicon (Fig. 1).One major advantage of using these cell lines is the high expres sion level of the precursor protein in the presence of inhibitors against viral proteins and controllable expression for a short period to avoid cytotoxicity caused by viral proteases 3C pro /3CD pro (85,86).Processing of the 2BC3ABCD precursor gave both final and inter mediate products, similar to those observed in PV-infected cells (27) (Fig. 1 and 8).I found that 2BC3ABCD could produce 2BC, 2B, 2C ATPase/hel , and 3AB (thus, also 3CD pro ) by 3CD pro provided in trans, but the processing of 3AB could occur only in cis (Fig. 8).In the trans rescue using the cell lines, replication was not detected for most of the mutants exam ined, except for a mutant with a disrupted cleavage site between 3C pro and 3D pol (3C/D[A/G] mutant), which showed a comparable level of replication to that of the WT replicon (Fig. 2 and 3).This allowed further analysis of the cis role of viral proteins in this mutant and revealed that only the activity of 3D pol , but not those of 3B, 2C ATPase/hel , or 3C pro , could be trans rescued (Fig. 4).PV 3CD pro lacks RNA polymerase activity (87,88); thus, the 3C/D(A/G) mutant, which could express only 3CD pro but not 3C pro nor 3D pol , was predicted to be deficient in polymerase activity irrespective of the introduction of inactivating mutations for 3D pol activity (Fig. 4).Previous reports suggested that 3D pol activity could be rescued in trans but with low efficiencies, similar to the rescue of 2C ATPase/hel activity (30,68,69).The high expression level of the viral proteins in the cell lines might have improved the efficiency of the rescue and provided all-or-none replication.trans complementation of a PV mutant (EG mutant) with a partially defective cleavage site between 3B and 3C pro has been reported ( 89), but I could not detect significant replica tion or trans-rescued replication of the corresponding mutant (3B/C[A/G] mutant, Fig. 2  and 3).This might suggest that partial or inefficient cis cleavage between 3B and 3C pro is sufficient for the trans-rescued replication.Studies on foot-and-mouth disease virus (FMDV), which belongs to the genus Aphthovirus in the family Picornaviridae (90,91), suggested that the defects in 3B uridylylation and 3D pol activity could be rescued in trans (34,35).FMDV is unique in coding multiple copies of 3B (three tandem copies of 3B).In addition, the recently discovered mosavirus has two copies of 3B (92), which could not be stably maintained in the PV genome (93).FMDV does not depend on host PI4KB/OSBP family I pathway for replication (94) in contrast to EVs, thus the mechanism and/or the role of uridylylation of 3B might be different from those in PV replication, in terms of trans rescue.Collectively, these results suggest a conserved trans role of 3D pol activity in picornavirus replication.
To provide trans 3D pol activity, expression of 3CD pro , but not of 3D pol , was essential and sufficient (Fig. 5 and 6); a large precursor was not necessarily essential in trans rescue.Unexpectedly, substitution of aa residues of 3C pro , which are involved in binding to viral RNA elements or phospholipids (4,24,26,62), did not abrogate the trans-active function.This is in contrast to requirements for 3CD pro in an in vitro trans complementation system; the interaction with RNA was essential for RNA synthesis and virus maturation (95,96).This may suggest a critical difference between the pre-initiation complexes formed in vitro and in vivo, which could be rescued in trans by the 3CD pol activity, possibly via a different pathway.I found that the N-terminal region of 3C pro is essential for in vivo trans rescue (Fig. 8).Primary structures of the N-terminal region of 3C pro , which protrudes outside in the tertiary structure, are conserved within the viral species (53) (Fig. 12).PV non-structural proteins did not rescue a defective EV-A71 replicon in trans, suggesting that the interaction is dependent on the viral species (Fig. 7).I propose a model in which the 3CD pro provided in trans provides 3D pol activity to the viral pre-initiation complex, formed by the defective replicons, via virus species-specific protein-protein interactions (80) (Fig. 13).
Based on the observations of DI particles, the capsid-coding region is dispensable for the replication of PV; PV replicons coding exogenous reporter genes in place of the capsid-coding region have been developed (24,97).Subsequently, generation of PV pv that encapsidates engineered replicon with PV capsid proteins provided in trans has been established by using helper PV (98,99), recombinant vaccinia virus systems (100,101), or virus-free capsid-protein expression plasmid vector (65).PV pv could serve as a safe alternative to live PV in biological tests, because no infectious virus is produced in the infected cells (65,102).A limitation of these methods is that only replication-competent replicon RNA could be encapsidated (98,103).In the present study, I partially solve this limitation to produce PV pv with a defective PV replicon after trans-rescued replication (Fig. 10 and 11).This new generation of PV pv may be useful for biological tests with enhanced safety.
The limitations of this study include the detection limits of trans-rescued replication; defective replicons give signals of reporters (firefly luciferase or mCherry) derived from initial protein synthesis before replication (about 1/10 2 or 1/10 4 of the signals at plateau, in the RNA-transfected cells or in PV pv -infected cells, respectively).Therefore, inefficient initial replication resulting only in quasi infection (70) might be missed in this study.In processing of a polyprotein (2BC3ABCD), only the processing of 3AB could occur in cis.However, a requirement for 2A pro , which can be deleted from the genome but requires a host factor SETD3 (104,105), in cis processing remains to be elucidated.
Collectively, this work reveals potential roles of exogenous viral proteins in PV replication and offers insights into protein/protein interactions during picornavirus infection.This will aid in elucidating the mechanism of multiple PV infection, including intra-species recombination that can reduce the effectiveness of novel PV vaccines toward eradication (106,107).Our findings might be useful for the development of effective antivirals targeting the polyprotein processing.

Cells
RD cells (human rhabdomyosarcoma cells) and HEK293 cells (human embryonic kidney cells) were cultured as monolayers in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal calf serum (FCS).RD cells were used for virus titration.HEK293 cells were used for expression of PV non-structural proteins and for production of PV pv .

General methods for molecular cloning
Escherichia coli strain XL10gold (Stratagene) was used for the preparation of plasmids.Ligation of DNA fragments was performed using an In-Fusion HD Cloning Kit (Clontech).PCR was performed using KOD Plus DNA polymerase (Toyobo).DNA sequencing was performed using a BigDye Terminator v3.1 cycle sequencing ready reaction kit (Applied Biosystems) and then analyzed with a 3500xL genetic analyzer (Applied Biosystems).

Lentivirus expression vectors for PV non-structural proteins pTet-AG-PV-2B2CP3(WT)
A cDNA fragment of a polyprotein of PV non-structural proteins (2BC3ABCD) was obtained by PCR with pPV-Fluc mc (a plasmid encoding cDNA of a PV replicon) (108) as the template and following primer set 1.This DNA fragment was inserted into a DNA fragment of a lentivirus vector plasmid with a TRE3G promoter, which was obtained by PCR with pLJM1-TRE3G-His-AG-FLAG-PreScission-OSBP(406-807) (52) as the template and using primer set 2.
Primer set 5
Primer set 25

Statistical analysis
Results of experiments are shown as means with standard deviations.Presented data are representative of at least two independent experiments with two or three biological replicates.Values of P < 0.05 by one-tailed t-test were considered to indicate a significant difference and were indicated by asterisks (*P < 0.05, **P < 0.01, ***P < 0.001).

FIG 2
FIG 2 Experimental design of trans rescue of replication of defective PV replicons.(A) Schematic view of the trans-rescue experiment.PV non-structural proteins were expressed in the presence of DOX (1 mg/mL) and GC376 (100 µM) at 37°C for 17 h.Then, RNA transcripts of each PV replicon that has firefly luciferase or mCherry reporter were transfected into cells in the absence of DOX and GC376.The luciferase signals or fluorescence of mCherry in the cells was analyzed at 7 h p.t. of the RNA transcripts.(B) Schematic view of PV replicon mutants.Firefly luciferase or mCherry was used as reporters for replication.Introduced amino acid substitutions in the cleavage sites by 3C pro or deletions of each viral gene are shown.

FIG 3
FIG 3 trans rescue of replication of defective PV replicons in Tet-AG-PV-2B2CP3(WT) cells.Without DOX treatment (A) or after DOX and GC376 treatment (B), the cells were transfected with RNA transcripts of each replicon that has firefly luciferase reporter, in the presence or absence of GuHCl (a 2C inhibitor).The luciferase signals measured at 7h p.t. of the RNA transcripts are shown.

FIG 4
FIG 4 cis roles of viral proteins in trans-rescued replication of defective PV replicons.(A) Schematic view of defective PV replicon mutants with a disrupted cleavage site between the 3C and 3D regions.Firefly luciferase was used as reporter of the replication.(B) trans rescue of replication of defective PV replicon mutants in Tet-AG-PV-2B2CP3(WT) cells.After DOX and GC376 treatment for 17 h, the cells were transfected with RNA transcripts of each replicon that has firefly luciferase reporter, in the presence or absence of GuHCl.The luciferase signals measured at 7 h p.t. of the RNA transcripts are shown.

FIG 5 (
FIG 5 (Continued) were detected by an anti-3D antibody.(C) trans rescue of replication of PV 3C/D(A/G) mutant in Tet-AG-3CD(WT) and Tet-AG-3CD(Δ4-5 aa) cells.After the DOX and GC376 treatment for 17 h, the cells were transfected with RNA transcripts of each replicon (WT and 3C/D[A/G] mutant) that has firefly luciferase reporter, in the presence or absence of GuHCl.The luciferase signals measured at 7 h p.t. of the RNA transcripts are shown.

FIG 6
FIG 6 trans rescue of replication of a defective PV replicon by PV 3D pol protein.(Top left) Generation of DOX-inducible HEK293 cell lines [Tet-AG-3D(WT)] that express PV 3D pol protein (WT).Fluorescent microscope images of the cells for AG after DOX treatment (for 17 h) are shown.(Bottom left) SDS-PAGE analysis of the expressed proteins in the cells.The cells were treated with or without DOX for 17 h.Fluorescence of AG on the gel is shown.(Right) trans rescue of replication of a defective PV replicon in Tet-AG-3D(WT) cells.After the DOX treatment, the cells were transfected with RNA transcripts of each replicon (WT or 3C/D[A/G] mutant) that has firefly luciferase reporter in the presence or absence of GuHCl.The luciferase signals measured at 7 h pt RNA are shown.

FIG 7
FIG 7 trans rescue of replication of a defective EV-A71 or PV replicon.After the DOX and GC376 treatment of indicated cells, the cells were transfected with RNA transcripts of each replicon that has firefly luciferase reporter in the presence or absence of GuHCl (a 2C inhibitor).The luciferase signals measured at 7 h pt RNA are shown.

FIG 10 (FullFIG 11
FIG 10 (Continued)Infectivity of PV pv produced with a defective PV replicon with mCherry reporter.Tet-AG-3CD(WT) or Tet-AG-3CD(Δ4-5 aa) cells were treated with or without DOX (1 mg/mL) at 37°C for 5 h.The cells were infected with 10 µL of PV pv solution and then incubated at 37°C for 17h.Fluorescent microscope images of the cells for mCherry and the observed titers of PV pv (IU/100 µL) are shown.

FIG 11 (
FIG11 (Continued)    anti-PV1, PV2, or PV3 antisera (26 U for each type) at 4°C for 7 h, and then added to the DOX-treated Tet-AG-3CD(WT) cells.The number of the cells positive for mCherry fluorescence was counted at 17 h post-infection (p.i.) PV1(mCherry) pv infection in the absence of antisera was taken 100%.

FIG 12
FIG 12 Comparison of primary and tertiary structures of 3C pro .(A) Sequence of aa 1-20 of 3C pro of enterovirus.Difference of aa from that of PV1(Mahoney) is highlighted in red.The aa 1-5 of PV1(Mahoney) is highlighted in green.(B) Tertiary structure of PV 3C pro (PDB: 4dcd).The aa 1-5 is highlighted in green.

FIG 13 A
FIG13  A model of PV replication by 3CD pro provided in trans.The 3CD pro provided in trans extends 3D pol activity to the pre-initiation complex via virus species-specific protein-protein interactions.