Cellular Protein Modification by Poliovirus: the Two Faces of Poly(rC)-Binding Protein

Plasmid design.(i) The synthetic oligonucleotide primers used were RP12(+) (5′ CATGCCATGGGATCGCATCACCATCACCATCAC 3′), RP13(−) (5′ GGTGGCATGGGGAGCAGCGACTACAAGGACGACGATGCAAGTAATAGGAATTCCGG 3′), RP18(+) (5′ GGCAATGCAAATCGATCATTTTCCCA 3′), RP19(−) (5′ TGGGAAAATGATCGATTTGCATTGCC 3′), RP20(+) (5′ GGCAATGCAAGCTAGCCATTTTCCCA), RP21(+) (5′ TGGGAAAATGGCTAGCTTGCATTGCC 3′), RP22(+) (5′ AATCGGGCGTGCTAGCGCCAAAATCA 3′), RP23(−) (5′ TGATTTTGGCGCTAGCACGCCCGATT 3′), RP24(+) (5′ TATACCATGGGCAGCAGCCATCAT 3′), RP25(−) (5′ ATCCTCGAGTTACTATTGACTCTG 3′), RP31(−) (5′ CCGGAATTCTTACTACTGTTGCATTGC 3′), pET22b-PCPBP2(+) (5′ GAT ATA CAT ATG GAC ACC GGT GTG ATT 3′), and pET22b-PCBP2(−) (5′ GTG CTC GAG GCT GCT CCC CAT GCC ACC CGT CTC 3′).

(ii) pET28a-PCBP2-ΔKH3.Synthetic oligonucleotide primers RP12(+) and RP31(−) were used to generate a PCR fragment containing nucleotides 121 to 930 of the pQE30-PCBP2 plasmid that was then cloned into pET28a (Novagen) by using the NcoI (RP12) and EcoRI (RP31) restriction sites. The resulting plasmid (pET28a-PCBP2-ΔKH3) encodes amino acids 1 to 253 of PCBP2 in pQE30-PCBP2, producing a protein that has an N-terminal hexahistidine tag (in italics in the primer sequence) and migrates at ∼26 kDa. Restriction sites are underlined in the primer sequences.

(iii) pET28a-PCBP2.Synthetic oligonucleotide primers RP12(+) and RP13(−) were used to generate a PCR fragment containing nucleotides 121 to 1266 of pQE30-PCBP2, which was cloned into pET28a by using NcoI and EcoRI (RP13). The FLAG tag sequence is in italics (above). The encoded protein has an N-terminal hexahistidine tag and a C-terminal FLAG tag.

(iv) pET28a-PCBP2 (Q253A, Q253I S254D, and Q306A).QuikChange mutagenesis (Stratagene) was used to generate the required mutations with the following synthetic oligonucleotide primers: Q253I S254D, RP18(+) and RP19(−); Q253A, RP20(+) and RP21(−); Q306A, RP22(+)and RP23(−). The template for mutagenesis was pET28a-PCBP2 (described above). The mutated proteins also have an N-terminal hexahistidine tag and a C-terminal FLAG tag.

(v) pET22b-PCBP2.Synthetic oligonucleotides were used to generate a PCR fragment with pQE30-PCBP2 as the template. The resulting PCR fragment was cloned into pET22b at the NdeI and XhoI restriction sites. pQE30-PCBP1 (10), pQE30-PCBP2-mKH1 (69), and pQE30-PCBP2-ΔKH2 (6) have been previously described. pET15b-3CD(μ10) and pET15b-3C (C147A) were also described previously (52, 53).

(vi) pET15-3C (C147A).Synthetic oligonucleotides RP24(+) and RP(25−) were used for PCR mutagenesis to clone the C147A mutation into 3C. pET15b-3CD (C147A) was used as the template (53).

Purification of recombinant PCBPs.Purification of hexahistidine-tagged PCBP1 (from the pQE30-PCBP1 plasmid), PCBP2 (from the pQE30-PCBP2 plasmid), and mKH1 (from the pQE30-mKH1 plasmid) was carried out as previously described (53). Hexahistdine-tagged PCBP2-FLAG wild-type (WT) and mutated proteins (from pET28a-PCBP2) and PCBP2-ΔKH3 (from pET28a-PCBP2-ΔKH3) were expressed in BL21DE3(Rosetta) cells. Cells were grown at 37°C to an A₆₀₀ of 0.2 and then induced with isopropyl-β-d-thiogalactopyranoside (IPTG; 1 mM final concentration) for 3 h at 25°C. Cells were lysed in buffer containing lysozyme (25 mM Tris-HCl [pH 8.0], 2.0 × 10⁵ U/ml chicken egg white lysozyme σ) by incubation on ice for 30 min, followed by sonication. The soluble fraction following lysis was precipitated with ammonium sulfate (20%, wt/vol). The resulting pellet fraction was dialyzed for 4 h or overnight in I60 buffer (20 mM Tris-HCl [pH 7.9], 250 mM NaCl, 60 mM imidazole, 10% glycerol, 0.5% NP-40) and subjected to Ni²⁺ ion-based affinity chromatography (GE Healthcare). PCBP2-containing fractions were pooled and further purified by size exclusion chromatography on an S200 Superdex column (GE Healthcare). The resulting proteins were ∼90% pure as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by Coomassie blue R250 staining. Purification of 3C (WT and C147A mutant) and 3CD (WT and C147A mutant) was carried out essentially as described by Parsley and coworkers (52).

In vitro cleavage of recombinant PCBPs.Recombinant purified proteins PCBP1 and PCBP2 (WT and Q253A, Q253I S254D, and Q306A mutant proteins) and the 3C (WT and C147A mutant) and 3CD (WT and C147A mutant) proteinases were incubated at a 4:1 molar ratio of substrate to proteinase in cleavage buffer (20 mM HEPES [pH 7.4], 1 mM dithiothreitol [DTT], 150 mm potassium acetate [KOAc]) for 6 h at 30°C. The cleavage products were analyzed by SDS-PAGE and Western blotting with an anti-His (Abcam) or anti-FLAG primary monoclonal antibody (Stratagene), followed by an alkaline phosphatase-conjugated anti-mouse secondary antibody.

Mass spectrometry analysis of cleavage products.Cleavage reaction mixtures containing recombinant PCBP2 and recombinant 3C were resolved by SDS-PAGE and then blotted onto a polyvinylidene difluoride membrane via Western blotting methods. The isolated protein products were subjected to mass spectrometry analysis, followed by N-terminal sequencing by the Protein Facility of the Iowa State University Office of Biotechnology.

RNA electrophoretic mobility shift analysis.RNA mobility shift assays were performed as previously described (69). Recombinant purified poliovirus 3CD (C147A) and/or recombinant purified PCBP2, PCBP2-ΔKH3, PCBP2-mKH1, or PCBP2-ΔKH2 were incubated in RNA-binding buffer (5 mM HEPES-KOH [pH 7.4], 25 mM KCl, 2.5 mM MgCl₂, 20 mM dithiothreitol, 3.8% [vol/vol] glycerol, 1 mg/ml Escherichia coli tRNA σ, 8 U of RNasin [Promega], 0.5 mg/ml bovine serum albumin [New England BioLabs]) for 10 min at 30°C. Proteins were then incubated with a radiolabeled poliovirus stem-loop I or stem-loop IV RNA probe at a final concentration of 5 nM in a 10-μl reaction volume and incubated for 10 min at 30°C. Following incubation, 2.5 μl of 50% glycerol was added and the resulting complexes were resolved at 4°C on native 4% polyacrylamide gels.

HeLa cell S10 preparation and poly(rC) depletion of cytoplasmic extracts.Preparation of HeLa cell S10 cytoplasmic extracts was done essentially as previously described (69). RNA affinity chromatography-mediated depletion of the extracts was also carried out as previously described (69), with the exception that ribosomal salt wash was not added to the extracts prior to depletion.

In vitro translation-RNA replication reaction mixtures.In vitro translation and RNA replication reaction mixtures were prepared with HeLa cell S10 extracts by using the RibPVE2A(MluI) RNA as previously described (69), except that the reaction mixtures did not contain ribosomal salt wash and total RNA from the replication reaction mixtures was purified with an RNAqueous spin column (Ambion) as previously described (18).

In vitro translation and luciferase assays.Translation assays were carried out as previously described (68), with minor modifications. Briefly, 20-μl reaction mixtures contained 50% HeLa cell S10 cytoplasmic extract; 50 fmol 5′PV-Luc (luciferase) reporter RNA (68); 1 mM ATP; 250 μM each GTP, CTP, and UTP; 30 mM creatine phosphate; 400 μg/ml creatine kinase (Boehringer Mannheim/Roche); and 15% (by volume) ribosomal salt wash buffer containing PCBP2 (WT), PCBP2-ΔKH3, or buffer alone. The reaction mixtures were incubated at 30°C for 90 min, and reactions were terminated with passive lysis buffer (Promega), followed by incubation on ice. The luciferase activity in 10 μl of sample was measured by adding 100 μl of reconstituted luciferase assay substrate (Promega). Relative light units were measured for 10 s with a Berthold luminometer.

Endogenous PCBP is modified during poliovirus infection of cultured HeLa cells.The three proteinases (2A, 3C, and 3CD) encoded in the genomes of different picornaviruses modify several cellular proteins during infection. As shown in Table 1, many of these modifications disrupt key cellular processes that would likely compete with virus replication. Since PCBP2 provides dual functions during translation and replication of the poliovirus genome, it seemed plausible that modification of this protein could initiate a switch in functionality. To determine the fate of PCBP during a poliovirus infection, HeLa cells were mock infected or infected with poliovirus at a multiplicity of infection of 30. Cytoplasmic extracts were harvested from these cultures over a 7-h time course and analyzed by SDS-PAGE and Western blotting with anti-PCBP polyclonal antiserum (Fig. 1). During poliovirus infection, the native PCBP species (apparent molecular mass of ∼40 kDa) was observed to disappear with the concomitant appearance of a faster-migrating species of PCBP with an apparent molecular mass of ∼26 kDa (Fig. 1, lanes 15 to 18). Furthermore, this faster-migrating species first detected at 3 h postinfection accumulates during the course of infection. This species is not observed in the mock-infected samples. The abundance of full-length PCBP also remained constant in the mock-infected samples. These results support the hypothesis that PCBP is modified during poliovirus infection of HeLa cells. It should be noted that the recombinant PCBP2 shown in lane 9 of Fig. 1 has an amino-terminal hexahistidine tag, thus accounting for its slower electrophoretic mobility than the authentic forms of PCBP from HeLa cells. In addition, the doublet observed migrating at a molecular mass of ∼40 kDa represents two of the four isoforms of PCBP in the cell. The specific isoform that is modified or the source of modification is not clear from these experiments. Therefore, inspection of the amino acid sequence of the PCBP isoforms for potential proteinase cleavage sites was a necessary next step in identifying the cleavage products.

• Open in new tab
• Download powerpoint

FIG. 1.

PCBP cleavage products accumulate during poliovirus infection of HeLa cells. Crude extracts from mock-infected (lanes 1 to 8) or poliovirus-infected (lanes 11 to 18) HeLa cells were examined over a 7-h time course by Western blot analysis. The sample from each time point represents a crude extract from approximately 5 × 10⁵ HeLa cells. For size comparison, 0.5 μg of histidine-tagged recombinant PCBP2 (rPCBP2; lane 9) and a molecular weight (MW) marker (lane 10) were included. Western blot analysis was carried out with rabbit polyclonal anti-PCBP serum as the primary antibody and an alkaline phosphatase-conjugated goat anti-rabbit (Sigma) secondary antibody. The electrophoretic mobilities of endogenous (uncleaved) PCBP isoforms and the putative cleavage products of PCBP are indicated by the arrows on the right.

3C/3CD consensus recognition sites are present in the amino acid sequences of PCBP isoforms.An amino acid sequence alignment of all of the mouse and human sequences for the four isoforms of PCBP is shown in Fig. 2. The poliovirus proteinases 3C and 3CD specifically recognize cleavage sites with a glutamine (Q) at the P1 position and a glycine (G) at the P1′ position of the scissile bond, although amino acids such as serine (S) or alanine (A) in the P1′ position are also cleaved by picornavirus 3C enzymes. An alanine or other amino acid with a small, aliphatic side chain is preferred at the P4 position of the sequence (23). Two such consensus 3C/3CD proteinase cleavage sites were identified in a PCBP amino acid sequence alignment and are highlighted in Fig. 2. The first potential cleavage site, referred to as the pre-KH3 cleavage site, is in the linker between the KH2 and KH3 domains at position Q253 S254 (PCBP2 numbering), and the second putative site is at Q306 G307 in the KH3 domain and is referred to as the KH3 cleavage site. The pre-KH3 cleavage site has an alanine at the P4 position, which makes it an ideal 3C/3CD cleavage site. The P4 amino acid at the KH3 cleavage site is an isoleucine, and its presence would be predicted to reduce the rate of proteolytic cleavage compared to the presence of alanine in the same position (51). Therefore, it was our hypothesis that PCBP2 is cleaved primarily at the pre-KH3 cleavage site. A consensus cleavage site for viral proteinase 2A was not found in the PCBP amino acid sequences. Several other Q/G amino acid pairs were observed in the PCBP1 and PCBP2 sequences; however, these sites did not have preferred amino acids at the P4 position of the scissile bond (data not shown). Since the only strong consensus 3C/3CD cleavage sites mapped to the linker region between KH2 and KH3 (pre-KH3 cleavage site) and the KH3 domain (KH3 cleavage site), only the sequences corresponding to these regions of interest are shown in Fig. 2. However, the alignments and potential cleavage site analyses were carried out for the complete amino acid sequences of all four PCBPs.

• Open in new tab
• Download powerpoint

FIG. 2.

Alignment of the human and mouse PCBP KH3 domain sequences showing putative 3C/3CD cleavage sites. The sequence alignment shown corresponds to the C-terminal region of the PCBPs (linker between KH2 and KH3 and the KH3 domain). The KH3 domain is in the white box. The positions of the putative scissile bonds for the two consensus 3C/3CD proteinase cleavage sites (pre-KH3 and KH3 cleavage site, respectively) are indicated, and the amino acids forming the cleavage site are boxed in gray. The critical fourth residue (P4 position) upstream of the putative cleavage site is indicated above or below the alignment. To the left of the amino acid sequences are the amino acid numbers for each protein. Several nonconsensus 3C/3CD cleavage sites have also been identified elsewhere in the PCBPs but are not indicated here. The PDB accession numbers for the human sequences are as follows: PCBP1, NM_006196; PCBP2, NM_005016; PCBP3, NM_020528; PCBP4, NM_033008. The PDB accession numbers for the mouse sequences are as follows: PCBP1, NM_011865; PCBP2, NM_011042; PCBP3, NM_021568; PCBP4, NM_021567.

Recombinant PCBP2 is cleaved in vitro by recombinant 3C and 3CD proteinases.We predicted that PCBP could be directly cleaved by the viral proteinases 3C and 3CD. To test this prediction, recombinant PCBP1, PCBP2, 3C, and 3CD were expressed in E. coli. All four proteins had an N-terminal hexahistidine tag for purification and detection purposes. PCBP1 or PCBP2 was incubated with WT 3C or 3CD for 6 h at 30°C, and the cleavage products were analyzed by SDS-PAGE, followed by Western blot analysis with an anti-PCBP polyclonal serum. The recombinant 3C and 3CD proteinases with a C147A mutation in the active site (which inactivates the proteinase) were also included in the assay as negative controls. As shown in Fig. 3A, both PCBP1 and PCBP2 were cleaved by the 3C or 3CD proteinase but not by the inactive enzymes. Furthermore, the data indicate equivalent levels of cleavage by either proteinase under the conditions tested. We also analyzed the kinetics of in vitro proteolysis of PCBP2 by 3C or 3CD. Cleavage products of PCBP2 were detected within the first 30 min of incubation; no apparent differences in the rate of cleavage of PCBP2 by 3C or 3CD were observed (data not shown). It should be noted that we included recombinant PCBP1 as a substrate in the assay shown in Fig. 3A only to show that the 3C/3CD cleavage is site specific rather than isoform specific. Because PCBP1 has not been definitively shown to be required for both translation and RNA replication of poliovirus RNA, we have focused the remainder of this study on PCBP2.

• Open in new tab
• Download powerpoint

FIG. 3.

Recombinant PCBP1 and PCBP2 are proteolytically cleaved in vitro by 3C and 3CD. (A) Purified preparations of recombinant PCBP1 or PCBP2 were incubated with recombinant 3C, 3CD, or inactive forms of both proteinases (C147A mutant forms). The substrate and proteinase at a concentration of 20 μM each were incubated in cleavage buffer (20 mM HEPES [pH 7.4], 1 mM DTT, 1 mM EDTA, 0.15 M KOAc) at 30°C for 6 h. The cleavage reaction mixtures were analyzed by Western blotting with anti-PCBP rabbit polyclonal serum. (Top) Lane 1, recombinant PCBP1 (rPCBP1) incubated alone; lane 2, rPCBP1 incubated with 3C; lane 3, rPCBP1 incubated with 3C (C147A); lane 4, rPCBP1 incubated with 3CD; lane 5, rPCBP1 incubated with 3CD (C147A). (Bottom) Lane 1, rPCBP2 incubated alone; lane 2, rPCBP2 incubated with 3C; lane 3, rPCBP2 incubated with 3C (C147A); lane 4, rPCBP2 incubated with 3CD; lane 5, rPCBP2 incubated with 3CD (C147A). The relative electrophoretic mobilities of uncleaved and cleaved PCBP are indicated by arrows on the left, and the molecular masses of two marker polypeptides are indicated on the right. (B) Two cleavage fragments are observed upon the cleavage of recombinant PCBP2 by 3C. Increasing amounts of recombinant PCBP2 were incubated with recombinant 3C in cleavage buffer (20 mM HEPES [pH 7.4], 1 mM DTT, 150 mm KOAc) for 6 h at 30°C. The cleavage products were analyzed by Western blotting. PCBP2 has a hexahistidine tag at the N terminus and a FLAG epitope tag at the C terminus. 3C has a hexahistidine tag at the N terminus. (Top) An anti-histidine monoclonal antibody was used as the primary antibody. (Bottom) An anti-FLAG monoclonal antibody was used as the primary antibody. Lane 1, PCBP2 alone; lane 2, 3C alone; lanes 3 to 5, 50, 75, and 150 pmol of PCBP2 incubated with 50 pmol of 3C proteinase. Two cleavage products of PCBP2 are shown in lanes 4 and 5, an N-terminal fragment (∼26 kDa) detected with the anti-histidine antibody and a C-terminal fragment (∼14 kDa) detected with the anti-FLAG antibody.

To determine whether PCBP cleavage by 3C/3CD occurred at the pre-KH3 cleavage site, the KH3 cleavage site, or both, a dual tagged form of PCBP2 with a hexahistidine tag at the N terminus and a FLAG tag at the C terminus of the protein was expressed in E. coli. The purified protein was incubated in an in vitro cleavage reaction mixture with 3C proteinase (Fig. 3B), and the reaction mixtures were analyzed by SDS-PAGE, followed by Western blot analysis with either anti-His or anti-FLAG monoclonal antibodies. On the basis of the two possible cleavage sites, the expected molecular masses of the cleavage fragments are shown in Table 2. As shown in Fig. 3B, lane 5, a ∼26-kDa cleavage fragment corresponding to the N-terminal portion of PCBP2 was observed with the anti-His antibody. A ∼14-kDa fragment was observed with the anti-FLAG antibody. Both cleavage fragments were detected only in the reaction mixtures that contained active 3C proteinase and PCBP2, indicating that these fragments were the result of authentic proteolytic cleavage. Mass spectrometry and N-terminal sequence analysis of the cleaved fragments demonstrated that the cleavage occurred at the pre-KH3 cleavage site (data not shown). The corresponding molecular masses determined for each of the cleavage fragments (28 kDa and 12 kDa) are indicated in Table 2.

Mapping of the 3C and 3CD proteinase cleavage sites in PCBP2.To confirm the data obtained by mass spectrometry analyses, the pre-KH3 and KH3 cleavage sites were mutated to generate three mutated versions of PCBP2. The pre-KH3 cleavage site was mutated to Q253I S254D, where both the P1 and P1′ positions of the scissile bond were changed, and to Q253A, where only the P1 position was changed. At the KH3 cleavage site, a Q306A mutation was generated at the P1 position. By using the cleavage assay conditions described for the experiment shown in Fig. 3, the three recombinant forms of mutated PCBP2 were incubated with 3C or 3C (C147A). The resulting cleavage products were analyzed by immunoprecipitation of the cleavage reaction mixtures with anti-His monoclonal antibody, followed by Western blot analysis with anti-PCBP polyclonal serum (Fig. 4). Immunoprecipitation was used in these assays because of impurities in some of the preparations of recombinant proteins that comigrated with the cleavage products and produced a high background level in Western blot assays (data not shown). The results showed that the WT PCBP2 protein was cleaved by proteinase 3C (Fig. 4.), while the double mutation at the pre-KH3 cleavage site (Q253I S254D) rendered PCBP2 uncleavable by the purified proteinase. PCBP2 with a single mutation (Q253A) at the pre-KH3 cleavage site showed significantly reduced cleavage by 3C. These data suggest that both the P1 and P1′ positions may be important for proteinase recognition of this cleavage site. PCPB2 with a mutation at the KH3 cleavage site (Q306A) was cleaved by 3C, albeit to a lower extent than the WT protein. This site may also be used for cleavage by 3C/3CD but is not the major site cleaved within PCBP2.

• Open in new tab
• Download powerpoint

FIG. 4.

Cleavage site mutations prevent 3C/3CD-mediated cleavage of PCBP2. Two mutations were constructed at the pre-KH3 cleavage site (Q253I S254D and Q253A), and one mutation was generated at the KH3 cleavage site in PCBP2. Recombinant WT and mutant proteins were incubated with the 3C, 3C (C147A), 3CD, or 3CD (C147A) proteinase under the cleavage assays conditions described in the legend to Fig. 3. Cleavage reaction mixtures were analyzed by immunoprecipitation with an anti-histidine monoclonal antibody, followed by Western blotting with anti-PCBP polyclonal serum. 3C (C147A) corresponds to the C147A mutation of the proteinase. Uncleaved and cleaved PCBP are indicated.

Differential binding of PCBP2-ΔKH3 to stem-loop structures in the 5′ NCR of viral RNA.To test the biological function(s) of a truncated version of PCBP2 that would be the equivalent of a product of proteolytic cleavage at the pre-KH3 site, we generated a recombinant protein lacking the sequences downstream of this site, including the entire KH3 domain. This protein, called PCBP2-ΔKH3, was purified from E. coli and used in electrophoretic mobility shift assays with either stem-loop I (also known as cloverleaf) or stem-loop IV in the 5′ NCR of poliovirus RNA. On the basis of previous studies that analyzed the RNA-binding properties of mutated polypeptides with charge-to-alanine substitutions in the KH3 domain of PCBP2, as well as the functional consequences of such mutations for viral translation and RNA replication (69), we predicted that PCBP2-ΔKH3 would be unable to bind efficiently to stem-loop IV RNA but might retain its ability to form an RNP complex with stem-loop I. Results from mobility shift experiments with recombinant PCBP2 (WT) or PCBP2-ΔKH3 protein and stem-loop IV RNA are displayed in Fig. 5A. Only the WT version of PCBP2 formed an RNP complex with the stem-loop IV RNA (represented by a slower-migrating species in lanes 6 and 7, Fig. 5A). Similar concentrations of PCBP2-ΔKH3 failed to shift the stem-loop IV RNA probe (lanes 12 and 13, Fig. 5A). Higher concentrations of PCPB2-ΔKH3 were also tested with similar results (data not shown).

• Open in new tab
• Download powerpoint

FIG. 5.

Differential RNA-binding capacities of PCBP2 and PCBP2-ΔKH3. (A) RNA mobility shift assays performed as described in reference 69 and Materials and Methods. Recombinant purified PCBP2 or PCBP2-ΔKH3 was incubated with a radiolabeled poliovirus stem-loop IV RNA probe. Following incubation, the resulting complexes were resolved on a native 4% polyacrylamide gel. The electrophoretic mobility of the RNP complex between stem-loop IV RNA and PCBP2 is indicated on the right (S-L IV/PCBP2), as is that of the free stem-loop IV probe (Free S-L IV). (B) RNA mobility shift assay with recombinant purified PCBP2, PCBP2-ΔKH3, mKH1-2, and ΔKH2 in the absence or presence of poliovirus 3CD (C147A) to analyze the abilities of the different forms of PCBP2 to form a ternary complex with 3CD and stem-loop I RNA (also known as cloverleaf RNA). The radiolabeled probe corresponded to poliovirus stem-loop I RNA, and the conditions of incubation and electrophoretic separation were as described for panel A. The protein concentrations in each reaction mixture are indicated above the autoradiograph of the gel. Complexes formed with stem-loop I RNA and PCBP2 or 3CD alone, as well as the ternary complex formed with stem-loop I RNA, PCBP2 or PCBP2-ΔKH3, and 3CD are indicated. As described in the text, mKH1-2 and ΔKH2 are mutated forms of PCBP2 that are unable to bind to stem-loop I or stem-loop IV in the poliovirus 5′ NCR.

To analyze the binding of PCBP2-ΔKH3 to stem-loop I RNA, we assayed for the formation of a ternary complex that is formed between the RNA and two binding partners, PCBP2 and poliovirus protein 3CD. The formation of this complex has previously been shown to be essential for poliovirus RNA synthesis (28, 53). As shown in Fig. 5B, lanes 5 and 6, a slow-migrating, heterogeneous complex was formed following the addition of WT PCBP2 and recombinant 3CD to labeled stem-loop I RNA. This complex was not observed when the individual proteins were used in binding assays (Fig. 5B, lanes 2 and 3). As shown in Fig. 5B, lanes 9 and 10, PCBP2-ΔKH3 was capable of forming a ternary complex with stem-loop I RNA and viral protein 3CD. Although this complex was more diffuse than the one detected with WT PCBP2, the formation of a ternary complex with stem-loop I RNA by PCBP2-ΔKH3 suggests that the proteolytic cleavage of PCBP2 may not disrupt the ability of the protein to generate a functional RNA replication complex. It should be noted that PCBP2-ΔKH3 appears to have a lower affinity for stem-loop I than does the WT protein (Fig. 5B). At a concentration of 500 nM PCBP2-ΔKH3, ternary complex formation was ∼45% of the WT level (compare lanes 5 and 9 in Fig. 5B), while at 1,000 nM PCBP2-ΔKH3, ternary complex formation was ∼65% of the WT level (compare lanes 6 and 10 in Fig. 5B). The lower affinity of PCBP2-ΔKH3 may reflect the loss of the KH3 domain-mediated stabilization of PCBP2 binding to stem-loop I that we have previously noted (69), or it may be due to overall changes in protein folding mediated by the loss of a major structural domain of PCBP2. Mutated PCPB2 proteins mKH1-2 and ΔKH2 were included as negative controls (Fig. 5B, lanes 11 to 14), since these proteins have previously been shown to be defective in RNA binding (6, 69).

Cleaved PCBP2 does not function in poliovirus translation in vitro.To determine if the cleaved form of PCBP2 was still active in poliovirus translation, in vitro translation assays were carried out with a dicistronic viral RNA, RibPVE2A(MluI), as previously described (48, 69). As shown in Fig. 6A, the P1 structural polyprotein of poliovirus is translated with the type I IRES of poliovirus while the P2 and P3 nonstructural precursors are translated via the type II IRES element of encephalomyocarditis virus (EMCV; which does not require PCBP2 to function). Therefore, when RNA from this construct was used for in vitro translation reactions in PCBP-depleted HeLa cell cytoplasmic extracts, the nonstructural proteins were translated without the addition of exogenous PCBP2. However, translation of the structural proteins (through the type I IRES) required the addition of recombinant PCBP2 to the depleted extracts (69). This in vitro translation assay was used to determine the ability of PCBP2-ΔKH3 to rescue translation in PCBP-depleted extracts compared to the WT protein. Figure 6B shows the results of translating RibPVE2A(MluI) RNA in nondepleted (lane 1), mock-depleted (lane 2), or PCBP-depleted HeLa cell cytoplasmic extracts (lanes 5 to 16). Recombinant PCBP2 (WT) or recombinant PCBP2-ΔKH3 was added to the PCBP-depleted extracts to rescue the translation of the P1 polyprotein. The results show that the WT protein could rescue translation in PCBP-depleted extracts in a dose-dependent manner, as indicated by the synthesis of the P1 structural precursor and its cleavage products VP0, VP1, and VP3. However, PCBP2-ΔKH3 could not rescue translation to any significant levels (Fig. 6B, lanes 10 to 14), suggesting that the 3C/3CD proteinase-mediated cleavage of PCBP2 impairs the function of this protein in poliovirus translation. A mutated form of PCBP2, mKH1-2, was used as a negative control in the translation reaction mixtures. This mutated PCBP2 is defective in binding stem-loop IV of the poliovirus IRES and therefore cannot rescue poliovirus type I IRES-driven translation (69).

• Open in new tab
• Download powerpoint

FIG. 6.

Rescue of viral RNA translation in PCBP-depleted HeLa cell S10 extracts. (A) Genetic organization and processing map of dicistronic poliovirus RNA used for in vitro translation and RNA replication assays. The values in parentheses are the molecular masses (in kilodaltons) of precursor polypeptides and proteolytic cleavage products produced following the translation of this RNA. (B) In vitro translation reaction mixtures prepared with either nondepleted (lane 1), mock-depleted (lane 2), or PCBP-depleted (lanes 4 to 17) HeLa cell cytoplasmic S10 extracts with the dicistronic replicon RNA, RibPVE2A(MluI), as previously described (69). The depleted extracts were supplemented with recombinant PCBP2 (WT), ΔKH3, or mKH1-2 at the indicated concentrations. The mutated PCBP2 protein mKH1-2 was used as a negative control since it is unable to function in translation (see reference 68). Synthesis of the structural polyprotein (P1) and its cleavage products (VP0, VP1, and VP3) was used as an indicator of differential translation between the ΔKH3- and PCBP2 (WT)-supplemented reaction mixtures.

To quantify the reduced levels of translation produced by the addition of PCBP2-ΔKH3 to in vitro translation assay mixtures, we used a luciferase reporter gene fused to the poliovirus 5′ NCR (5′PV-Luc) in mock-depleted or PCBP-depleted HeLa cell cytoplasmic extracts (Fig. 7). These in vitro translation assay mixtures were supplemented by the addition of recombinant PCBP2 (WT) or PCBP2-ΔKH3 protein. As shown in Fig. 7, the maximum level of translation stimulated by PCBP2-ΔKH3 was less than 20% of that stimulated by WT PCBP2. When recombinant PCBP2-ΔKH3 was added to nondepleted HeLa cell S10 extracts, the translation of 5′PV-Luc could not be inhibited (data not shown). Thus, PCBP2-ΔKH3 appears to be unable to incorporate itself into functional translation initiation complexes and act as a dominant negative inhibitor of translation.

• Open in new tab
• Download powerpoint

FIG. 7.

Rescue of poliovirus translation in PCBP-depleted HeLa cell cytoplasmic S10 extracts. The translation of a luciferase reporter gene driven by the poliovirus IRES was analyzed in HeLa cell cytoplasmic S10 extracts that were either mock depleted, depleted of PCBP, or depleted of PCBP and reconstituted with recombinant WT PCBP or ΔKH3 protein. The translation reaction mixtures were prepared in triplicate and incubated at 30°C for 1.5 h. Luciferase assays were carried out with 10 μl of the translation reaction mixture as previously described (68). RLU, relative light units.

Cleaved PCBP2 is active in poliovirus RNA replication.Although PCBP2-ΔKH3 was unable to function in translation because of its inability to bind to stem-loop IV in the 5′ NCR of viral RNA, it still binds to stem-loop I and forms a ternary complex with 3CD. Therefore, we predicted that this protein should retain its ability to function in viral RNA replication. To test this prediction, we made use of the dicistronic RNA construct (noted above) because it allows the uncoupling of poliovirus translation in vitro from RNA replication via translation of the nonstructural proteins from the EMCV IRES (69). We determined if the amino-terminal cleavage fragment of PCBP2 (PCBP2-ΔKH3) was able to rescue RibPVE2A(MluI) RNA replication in PCBP-depleted HeLa cell cytoplasmic extracts. The results shown in Fig. 8 demonstrated that both PCBP2 (WT) and PCBP2-ΔKH3 could rescue RNA replication in a dose-dependent manner. The level of rescue by PCBP2-ΔKH3 was slightly lower than that of the WT protein (compare lane 5 to lanes 10, Fig. 8), perhaps reflecting its lower affinity for stem-loop I upon loss of the KH3 domain. As mentioned earlier, even though the KH1 domain is the primary determinant for RNA binding, the KH3 domain also makes a contribution to the binding of stem-loop I (69). It should be noted that at the highest concentration of PCBP2-ΔKH3 used in this assay (500 nM), the mutated protein reproducibly appears to cause an inhibition of RNA synthesis. This may suggest an altered state of multimerization at high concentrations of the truncated protein or perturbation of the binding equilibrium between the ternary complex and the viral RNA-dependent RNA polymerase required for initiation of replication. Importantly, the ability of PCBP2-ΔKH3 to rescue the replication of the dicistronic RNA provides direct evidence that the KH3 domain is dispensable for RNA replication. These data are also in agreement with previous studies that showed the primary function of the KH3 domain to be translation initiation, perhaps by recruiting ribosomes to the viral RNA through its interaction with SRp20 (5).

• Open in new tab
• Download powerpoint

FIG. 8.

Rescue of viral RNA replication in PCBP-depleted HeLa cell S10 extracts. In parallel experiments to the in vitro translation reactions shown in Fig. 6B, in vitro RNA replication reaction mixtures were prepared with either nondepleted (lane 1) or PCBP-depleted (lanes 3 to 16) HeLa cell cytoplasmic S10 extracts with the dicistronic replicon RNA RibPVE2A(MluI) as the template. The depleted extracts were supplemented with recombinant PCBP2 (WT), ΔKH3, or mKH1-2 at the indicated concentrations. As a negative control, 2 mM guanidine hydrochloride (GuHCl) was added to an S10 extract sample to inhibit RNA replication (lane 2). Replicated poliovirus (PV) RNA is indicated on the left.

Model of poliovirus template selection during translation and replication of the genome.On the basis of the data derived from this study, as well as previously reported observations by other investigators, a model of PCPB2-mediated template selection is proposed in Fig. 9. Following the initial rounds of translation and RNA replication during the early phases of the poliovirus infectious cycle (Fig. 9A), the positive-strand viral genome is used as the template for translation, leading to high-level synthesis of viral proteins, including the 3C/3CD proteinases. These proteins mediate the cleavage of endogenous PCBP1 and PCBP2 in the cytoplasm of the infected cell (Fig. 9B). This cleavage releases the KH3 domain, and as a result, SRp20, from the stem-loop IV/PCBP2/SRp20 complex, rendering it inactive in translation initiation. The remaining cleavage fragment (PCBP2-ΔKH3) loses its affinity for stem-loop IV. Because of a lack of new initiation events, the viral RNA is eventually cleared of translating ribosomes and becomes available to function as the template in negative-strand RNA synthesis. The cleaved form of PCBP2 (PCBP2-ΔKH3) is still able to bind stem-loop I, where it forms a ternary complex with 3CD. This ternary complex may interact with PABP bound to the 3′ end of the viral genome to circularize the template, thereby allowing RNA replication to proceed.

• Open in new tab
• Download powerpoint

FIG. 9.

Model of template selection during poliovirus translation and viral RNA replication. Depicted is poliovirus positive-strand RNA (black line) under conditions that favor cap-independent translation (A) or those that favor negative-strand RNA synthesis (B). Under conditions that favor translation early during the infectious cycle (A), the RNA is bound by ribosomes (designated 80S on the gray spheres) continuously initiating and elongating nascent viral polypeptide chains (depicted as orange curly lines). Initiation of translation is facilitated by the binding of cellular protein PCBP2 to the poliovirus IRES in the 5′ NCR of positive-strand RNA (secondary structures with thickened black lines) and the possible bridging of RNA to the ribosome by the cellular protein SRp20. The presence of multiple ribosomes on viral RNA may preclude the viral RNA polymerase (3D) from elongating any newly initiated negative-strand RNAs. Under conditions that favor negative-strand RNA synthesis later during the infectious cycle (B), PCBP2 has been cleaved by the accumulated 3CD (or 3C) proteinase polypeptides (as indicated by the orange arrow). This inhibits the binding of ribosomes to the IRES, thereby reducing the levels of translation initiation complex formation and releasing SRp20 from PCBP2. The cleaved form of PCBP2 (PCBP2-ΔKH3) can still participate in ternary complex formation with the 5′ stem-loop I structure and 3CD. This may facilitate the interaction of the 5′ RNP complex with the 3′ poly(A) tract via a bridging interaction with PABP. This latter interaction would lead to initiation of negative-strand RNA synthesis on templates that are now cleared of translating ribosomes. The gray arrows indicate the direction of ribosomes traversing the RNA during translation, while the red arrows indicate the direction of the 3D RNA polymerase during negative-strand RNA synthesis. The blue lines represent nascent negative-strand RNAs with a VPg (small solid red circle) at their 5′ ends. The UUUUUU sequence represents the 5′ oligo(U) tract that is templated by the 3′ poly(A) tract on genomic positive-strand RNAs. The separation of the two processes is shown for illustration only; it is likely that the viral replication cycle is a dynamic process with both activities occurring simultaneously (but not on the same template RNA).

In our model, PCBP2 plays a major role in both the translation and the replication of the poliovirus genome, and the observations in this study have allowed a PCBP2-mediated mechanism to be proposed for template selection. However, a mechanism that includes other cellular proteins such as PTB may also need to be considered in light of the significant intracellular protein modifications and vesicle rearrangements observed during picornavirus infections. Furthermore, the proposed model does not assume complete cleavage of PCBP2 or complete down regulation of translation, as this has not been observed during poliovirus infections. Rather, it proposes a shift in the balance of activities between two processes (translation and RNA replication) within localized environments surrounding specific RNA templates.

It should be noted that the proposed mechanism of template selection is most likely not used by picornaviruses like EMCV or foot-and-mouth disease virus. These viruses with type II IRES elements do not appear to use PCBP2 during translation of the viral genome, and to date there have been no reports of a PCPB2 requirement for cardiovirus or aphthovirus RNA replication. It will be interesting to determine the alternative mechanisms that drive template selection for these and other positive-strand RNA viruses. Finally, PCBP2 has been identified in RNP complexes in other viral systems such as hepatitis C virus and human papillomavirus type 16. Although an interaction of PCBP2 with the hepatitis C virus IRES has been observed, the functional significance of this binding is unknown (64). For human papillomavirus, PCBP2 is used as a translational regulator of late gene expression (16). Collectively, these data suggest that PCBP2 is a multifunctional cellular protein that falls victim to molecular scavenging by several different viruses. However, its ability to control virus replication via multiple mechanistic roles is only now beginning to be understood.