ABSTRACT
Internal ribosome entry site (IRES)-driven translation is a common strategy among positive-sense, single-stranded RNA viruses for bypassing the host cell requirement of a 5′ cap structure. In the current study, we identified the ribosomal protein L13 (RPL13) as a critical regulator of IRES-driven translation of foot-and-mouth disease virus (FMDV) but found that it is not essential for cellular global translation. RPL13 is also a determinant for translation and infection of Seneca Valley virus (SVV) and classical swine fever virus (CSFV), and this suggests that its function may also be conserved in unrelated IRES-containing viruses. We further showed that depletion of DEAD box helicase DDX3 disrupts binding of RPL13 to the FMDV IRES, whereas the reduction in RPL13 expression impairs the ability of DDX3 to promote IRES-driven translation directly. DDX3 cooperates with RPL13 to support the assembly of 80S ribosomes for optimal translation initiation of viral mRNA. Finally, we demonstrated that DDX3 affects the recruitment of the eukaryotic initiation factor eIF3 subunits e and j to the viral IRES. This work provides the first connection between DDX3 and eIF3e/j and recognition of the role of RPL13 in modulating viral IRES-dependent translation. This previously uncharacterized process may be involved in selective mRNA translation.
IMPORTANCE Accumulating evidence has unveiled the roles of ribosomal proteins (RPs) belonging to the large 60S subunit in regulating selective translation of specific mRNAs. The translation specificity of the large-subunit RPs in this process is thought provoking, given the role they play canonically in catalyzing peptide bond formation. Here, we have identified the ribosomal protein L13 (RPL13) as a critical regulator of IRES-driven translation during FMDV infection. Our study supports a model whereby the FMDV IRESs recruit helicase DDX3 recognizing RPL13 to facilitate IRES-driven translation, with the assistance of eIF3e and eIF3j. A better understanding of these specific interactions surrounding IRES-mediated translation initiation could have important implications for the selective translation of viral mRNA and thus for the development of effective prevention of viral infection.
INTRODUCTION
In eukaryotic cells, translation initiation occurs through two alternative mechanisms, a cap-dependent mechanism in the majority of mRNAs or a 5′-end-independent mechanism driven by internal ribosome entry site (IRES) elements in a few cases (1). IRESs are present in cellular genes involved in the stress response, cell cycle, and apoptosis; it is estimated that 5 to 10% of cellular mRNAs contain IRES elements (2). IRESs are categorized into four classes based on their length, nucleotide sequence, and secondary and tertiary structures, as well as their mode of action. Classes I and II require the assistance of cellular factors, eukaryotic initiation factors (eIFs), for efficient formation of ribosomal initiation complexes, while class III IRESs require only a subset of eIFs, and class IV can promote translation without any eIFs (3). Since its discovery, the process of IRES-mediated translation has been regarded as a critical step for viral infection, with important effects on virulence, tissue tropism, and pathogenicity (2, 4, 5). When dominant cap-dependent translation is inhibited, the single-stranded positive-sense RNA viruses of picornaviruses and flaviviruses can take advantage of IRES-driven translation to subvert the host translation machinery (5). Several canonical eIFs can influence viral translation initiation efficiency. For instance, enterovirus 71 (EV-A71) in the family Picornaviridae requires eIF2, eIF3, eIF4A, eIF4G, eIF4B, and eIF1A (6), and eIF3 and eIF5B are necessary to direct the synthesis of proteins of hepatitis C virus (HCV) in the family Flaviviridae (7).
DExD/H box helicases are vital for the recognition of RNA and metabolism and are critical for the stimulation of antiviral innate immunity; the well-known eIF4A and retinoic acid-inducible gene 1 (RIG-I) are representative members of the class. Asp-Glu-Ala-Asp (DEAD) box polypeptide 3 (DDX3) is known to play roles in various key aspects of RNA metabolism, including transcriptional regulation, splicing, mRNA export, ribosome biogenesis, and translational regulation (8–10). In addition, DDX3 is a component of the innate immune response (11–14). DDX3 may accomplish modulation of cellular mRNA translation by interacting with RNA and specific initiation factors such as eIF2 (15), eIF3 (16), eIF4E (17), eIF4G, and poly(A)-binding protein (PABP) (18), but it does not directly interact with eIF1A or eIF5 (19). These observations suggest that helicase DDX3 is an active component of the translation initiation machinery. Furthermore, DDX3 positively regulates viral translation of HCV (19) and EV-A71 (20) for efficient propagation. DDX3 is required for translation of viral transcripts of IRES-containing viruses, but given its great complexity, the mechanistic basis for its mode of action is not fully understood.
The eukaryotic ribosome consists of four ribosomal RNAs (28S, 18S, 5.8S, and 5S rRNAs) and 79 ribosomal proteins (RPs), which are primarily responsible for protein synthesis from mRNAs (21, 22). RPs may exert ribosome-independent activities that are implicated in tumorigenesis, immune signaling, and diseases, and they may regulate translation of cellular mRNAs as constituents of the ribosome (23); this suggests that the ribosome is capable of much greater control in key cellular processes than previously thought. Various viruses have in fact evolved to hijack specific RPs to achieve optimal viral protein synthesis; RPL22 (24) and RPLPs (25, 26), as well as RACK1 (27), RPS5 (28, 29), RPS6 (30), and RPS25 (31, 32), facilitate translation of viral transcripts of IRES-containing viruses. The relationship of RPs and DDX3 in IRES-driven translation of specific mRNAs, however, remains to be clarified.
Foot-and-mouth disease virus (FMDV) belongs to the genus Aphthovirus within the family Picornaviridae. The FMDV genome encodes a single, long polyprotein that is cotranslationally processed by virally encoded proteinases into four structural proteins, VP1 to VP4, and eight nonstructural proteins, 2A, 2B, 2C, 3A, 3B, 3Dpol, leader proteinase Lpro, and 3 Cpro. The IRES of FMDV is one of the strongest IRESs described to date (33). FMDV IRES function in mediating initiation has been well studied; canonical initiation factors (such as eIF3, 4A, 4B, and 4G) and IRES trans-acting factors (ITATs) that bind to the IRES are required to reconstitute the translation initiation process in vitro (34–36). In the current study, we found that DDX3 binds to FMDV IRES directly. RPL13 participates in IRES-driven translation in a DDX3-dependent manner, and a similar translational mechanism is also seen in Seneca Valley virus (SVV) in the family Picornaviridae and classical swine fever virus (CSFV) in the family Flaviviridae. We also found that the eIF3 subunits e and j are involved in association with DDX3 and the IRES. This previously uncharacterized process may be involved in the selective translation of viral mRNA.
RESULTS
RPL13 plays a supportive role in FMDV infection.By using the biotinylated RNA pulldown method, followed by liquid chromatography/electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS), we identified several RPs that bind both the 5′ untranslated region (5′UTR) and IRES, including RPL13, RPLP0, RPL8, RPL5, RPL19, RPL7, RPS6, RPS3a, RPS3, and RPS2 (Fig. 1A). Numerous research groups have demonstrated that these RPs may be utilized by diverse viruses for efficient viral infection (25, 30, 37–44). Notably, the candidate protein RPL13, a member of the L13E family, is conserved in a wide range of species, including plants, yeast, and animals (45, 46). RPL13 has been reported to be the product of a stress-inducible gene (47), and it was also found to act as a positive regulator for NF-κB signaling by modulating the translation of specific mRNA (48). The role of RPL13 in viral IRES-dependent translation, however, has until now been uncharacterized.
RPL13 is required for FMDV replication but not for cell viability. (A) Venn diagram showing the overlap between RPs identified as potentially interacting with the FMDV 5′UTR and IRES. Green and yellow ovals depict RPs bound to the FMDV 5′UTR and IRES, respectively. (B) BHK-21 cells transfected with a negative-control siRNA (siCtrl) or the indicated RP-targeting siRNA for 48 h were infected with FMDV (MOI, 0.5), and cells and supernatants were harvested at the indicated times. The virus titer was determined by TCID50 assay. (C) Cells were silenced with the indicated siRNAs for 72 h, and then dead cells were estimated using the LIVE/DEAD viability/cytotoxicity kit (column) and cell viability was assessed using the MTS assay (scatter). Cycloheximide (CHX), a protein synthesis inhibitor, was used as positive control. The data in panels B and C represent the means from three independent experiments, with error bars indicating standard deviation (SD) (*, P < 0.05; **, P < 0.01).
Intriguingly, a variety of RPs have been confirmed to act as translational regulator components of the ribosome to confer transcript-specific translational control. For example, large-subunit RPs (RPL22 and RPLPs) and small-subunit RPs (RACK1, RPS25, and RPS5) facilitate translation of viral transcripts of IRES-containing viruses (24–29, 31, 32). RPL40 promotes efficient translation for cap-dependent transcripts of vesicular stomatitis virus (VSV) in the family Rhabdoviridae (21, 49). Meanwhile, unlike RPS11, which definitely affects cell viability (50, 51), RACK1, RPS25, and RPL40 are not essential for global protein synthesis and cell proliferation. To investigate whether the RPs indicated above might play a role in FMDV infection, we used small interfering RNA (siRNA) to knock down RPs in BHK-21 cells and then infected the cells with FMDV. As shown in Fig. 1B, we found that the depletion of RPS11 and RPLP0 led to strong reductions in viral yield, but it caused detectable cell death with high cytotoxicity (Fig. 1C). In comparison, the depletion of RPL13, RACK1, RPS25, or RPS5 greatly depressed FMDV titers, but only the depletion of RPS5 led to an increase in cell death. Virus yields were slightly affected by the depletion of RPL40 and RPL22. The reduction in viral protein expression was consistent with the observed decrease in virus titer (data not shown). Importantly, we found that RPL13 silencing results in reduced production of viral particles by ∼5-, ∼14-, ∼22-, and ∼25-fold at 2, 4, 6, and 8 h postinfection (hpi), respectively, in BHK-21 cells, with no detectable cytotoxicity at the given doses. Overall, our data indicate that FMDV replication requires the ribosomal factor RPL13, which is otherwise not essential for cell viability.
DDX3 is identified as an RPL13-interacting protein.To explore the possible mechanisms of RPL13 involvement in FMDV replication, we combined immunoprecipitation (IP) with quantitative mass spectrometry (MS) to identify cellular proteins that interact with RPL13. As shown in Table S1 in the supplemental material, 50 proteins were found as potential RPL13-interacting proteins. Proteins involved in the eukaryotic translation machinery were identified, including ribosomal proteins RPL32, RPL4, RPL15, and RPL29 and subunits of eIF2 and eIF5. Of these, DDX3, with sequence coverage of 35%, was selected for further characterization. We first confirmed that RPL13 interacts with DDX3. As shown in Fig. 2A, RPL13 was coimmunoprecipitated with DDX3 regardless of the presence of viral infection; conversely, DDX3 was coprecipitated with RPL13 in BHK-21 cells (top). It is also likely that RPL13 interacted with DDX3 regardless of RNase A treatment (bottom), demonstrating that the interaction is RNA independent.
The N-terminal region of DDX3 is required for interaction with RPL13. (A) Top, cells were challenged with FMDV (MOI, 0.5) for the indicated times, and cell lysates were then immunoprecipitated with control IgG, DDX3, or RPL13 antibody. Bottom, equal amounts of lysates from mock-infected cells were subjected to immunoprecipitation with RPL13 antibody. The resulting coprecipitates were treated with 0.5 mg/ml RNase A at 37°C for 30 min. − RNase A and no treat., samples that were incubated in the absence of RNase A for 30 min at 37°C or at 4°C, respectively. Immunoblotting was then performed to detect RPL13 and DDX3. (B) Cells transfected with either control siRNA (siCtrl) or siRNA targeting DDX3 or RPL13 for 48 h were infected with FMDV (MOI, 0.5) for the indicated times. Cell lysates were then generated and subjected to Western blot analysis. (C) Docking results for RPL13 (red) with DDX3 (cyan for N domain and blue for C domain) using ZDOCK. The interacting residues are shown in stick model, with green residues for DDX-N and purple ones for DDX-C. (D) Top, schematic representation of recombinant constructs of FLAG-tagged WT DDX3 and mutants (DDX3△C and DDX3△N). Bottom, Western blot analysis of immunoprecipitations of cells cotransfected with FLAG-tagged WT DDX3, DDX3△C, or DDX3△N and GFP-tagged RPL13. (E and F) Recombinant proteins were mixed and passed over a Superdex G75 column. In panel E, the inset shows 10% SDS-PAGE analysis of markers (lane 1) and the eluted peak protein (lane 2). In panel F, the inset shows markers (lane 1), DDX3-N (lanes 2 and 3), RPL13 (lane 4), and eluted peak proteins (lanes 5 and 6).
The role of DDX3 in FMDV replication was then assessed. Following infection by FMDV, a significant decrease in the level of viral proteins was observed in DDX3-depleted cells at 4 h or 8 h postinfection (Fig. 2B). We also detected that the level of expression of DDX3 gradually increased in BHK-21 cells after FMDV infection. In contrast, the level of RPL13 did not undergo an obvious change with viral infection, indicating that its endogenous level was sufficient to support viral infection. Furthermore, the level of DDX3 is not affected when RPL13 is knocked down, and the level of RPL13 is not affected by depletion of DDX3. These results reveal that both RPL13 and DDX3 support viral infection, and the regulation of their expression is not reciprocal.
The N-terminal region of DDX3 is required for interaction with RPL13.DDX3 performs ATP-dependent helicase activity (at the N-terminal region, ranging from residue 133 to 406) and RNA-binding activity (at the C-terminal region, 407 to 584) (52, 53). To unravel the interaction between DDX3 and RPL13, RPL13 (PDB no. 6EK0) docking into DDX3 (PDB no. 5E7I) was performed using the ZDOCK algorithm in Discovery Studio 2018 (Accelrys, San Diego, CA, USA). This result indicated that RPL13 interacts mainly with the surrounding residues of N-terminal domain of DDX3 (Fig. 2C). Moreover, we used a FLAG-specific antibody to immunoprecipitate ectopically expressed FLAG-tagged DDX3 and variants DDX3ΔC and DDX3ΔN and used specific antibodies in immunoblotting to detect green fluorescent protein (GFP)-tagged RPL13 in the immune complex. The wild-type (WT) DDX3 and DDX3ΔC were apparently able to interact with RPL13 (Fig. 2D). This preliminary result and the docking results shown in Fig. 2C together suggest that the N-terminal region plays a role in DDX3 interaction with RPL13.
To determine whether RPL13 interacts with DDX3 directly, His or glutathione S-transferase (GST) fusion proteins were expressed and purified. The molecular weight (MW) of the N-terminal domain from DDX3 (denoted His-DDX3-N) was approximately 33 kDa and that of the C-terminal DDX3 (His-DDX3-C) was 22 kDa, while that of the GST fusion RPL13 (GST-RPL13) was 53 kDa. SDS-PAGE and chemical cross-linking analyses confirmed the formation of a homodimer of DDX3-C in the presence of cross-linking agent glutaraldehyde (54), while a monomer was obtained with both DDX3-N and RPL13 (data not shown). To address the interaction between recombinant proteins, DDX3-N and DDX3-C were mixed and passed over a Superdex G75 column; they formed a heterotrimer structure with a MW of 78 kDa, which is approximately the sum of one DDX3-N and two DDX3-C molecules (Fig. 2E). Furthermore, DDX3-N and RPL13 formed a heterodimer structure with a MW of 86 kDa (Fig. 2F), and no evidence supports interaction of DDX3-C and RPL13. Collectively, these results indicate the direct interaction between RPL13 and the N-terminal region of DDX3.
RPL13 and DDX3 for IRES-dependent translation and viral replication.To explore whether DDX3 and RPL13 are specific factors for FMDV IRES-driven translation, a bicistronic luciferase plasmid, psiCHECK-FMDV, containing a cap-dependent Renilla luciferase (RLuc) gene and an FMDV IRES-dependent firefly luciferase (FLuc) gene were constructed. The ratio of FLuc expression to RLuc expression (FLuc/RLuc ratio) shows the relative FMDV-IRES activity. First, a cap-dependent reporter plasmid (psiCHECK) or monocistronic mRNA (siCHECK) was transfected into cells, and the RLuc value, which shows cap-dependent activity, was measured; this value was unaffected by a decrease or increase of either DDX3 or RPL13 (Fig. 3A and B). Next, using transfection of bicistronic plasmid psiCHECK-FMDV (Fig. 3C) or in vitro-transcribed bicistronic reporter mRNA siCHECK-FMDV (Fig. 3D), we compared the effect of DDX3 and RPL13 on FMDV IRES activity to that of RACK1, which has been well studied as a regulator of IRES translation and facilitator of selective translation of transcripts of IRES-containing viruses (27). We found that IRES activity was strongly decreased when the expression of DDX3 or RPL13 was knocked down, to the same extent as in RACK1-depleted BHK-21 cells (Fig. 3C and D, left). The ectopic expression of RPL13 had no effect on IRES-dependent reporter activity, and ectopic expression of DDX3 stimulated the translation of the FMDV IRES (Fig. 3C and D, right). In addition, the decrease or increase in DDX3 or RPL13 expression did not affect the amount of bicistronic reporter RNA in the cells, and this demonstrates that DDX3 and RPL13 affect translation, rather than transcription and RNA stability (data not shown). Overall, our results indicate that both RPL13 and DDX3 are required specifically for IRES-dependent translation of FMDV and that they are not essential for cap-dependent translation.
Both RPL13 and DDX3 are required for FMDV IRES-driven translation and viral replication but not for cap-dependent translation. (A and B) Top, schematic diagrams of cap-dependent reporter plasmid psiCHECK (A) and mRNA siCHECK (B). Bottom, BHK-21 cells were transfected with either control siRNA (siCtrl) or siRNA targeting RACK1, DDX3, or RPL13 for 48 h or were transfected with FLAG vector, FLAG-DDX3, or FLAG-RPL13 for 24 h. The monocistronic construct psiCHECK (A) or mRNA siCHECK (B) was then transfected into the cells. At 24 h (A) or 6 h (B) posttransfection, RLuc activity was measured. (C) Top, schematic illustration of bicistronic FMDV IRES constructs (psiCHECK-FMDV). The translation of Renilla luciferase (RLuc) is cap dependent, whereas translation of firefly luciferase (FLuc) is mediated by the FMDV IRES. Bottom, cells transfected with either the indicated siRNA or FLAG-tagged plasmid were transfected with the bicistronic construct psiCHECK-FMDV. At 24 h posttransfection, the RLuc and FLuc activities normalized to mRNA levels were determined. RACK1 was used as a positive control for selective translation of IRES-containing virus. (D) Top, schematic diagram of capped bicistronic mRNAs (siCHECK-FMDV). Bottom, cells transfected with either the indicated siRNA or FLAG-tagged plasmid were transfected with siCHECK-FMDV mRNA. At 6 h posttransfection, the RLuc and FLuc activities were assayed. (E) BSR-T7 cells transfected with either the indicated siRNA or FLAG-tagged plasmid were transfected with the replicon rFMDV-EGFP. At 24 h posttransfection, the cells were subjected to fluorescence analysis. For panels A to E, results were normalized to the control (arbitrarily set to 100%) and are presented as mean ± SD from three independent experiments.
To determine the effects of RPL13 and DDX3 on FMDV replication, we used an FMDV subgenomic replicon with an enhanced GFP (EGFP) reporter gene, rFMDV-EGFP. In agreement with the results shown in Fig. 3C and D, the levels of specific fluorescence were strongly decreased in RPL13- or DDX3-depleted BSR-T7 cells. However, replicon activity was elevated when DDX3 was ectopically expressed and was unaffected when RPL13 was ectopically expressed (Fig. 3E). Specifically, the lower IRES-dependent translation efficiencies, together with the lower activity of the replicon in RPL13-depleted cells, indicates that RPL13 acts as a positive regulator of viral translation. Our results together indicate that RPL13 and DDX3 facilitate IRES-dependent translation, and this consequently promotes efficient replication of FMDV.
RPL13 and DDX3 associate with the FMDV IRES.To determine whether RPL13 and DDX3 associate with the IRES of the FMDV 5′UTR, BHK-21 cells were incubated with FMDV at a multiplicity of infection (MOI) of 1 for 5 h. The cells were lysed, and the RNA-protein complexes were immunoprecipitated by an antibody specific to RPL13 or DDX3 or by isotype anti-IgG. The total RNAs isolated from these immunoprecipitates were subjected to reverse transcription-PCR (RT-PCR) analysis for the presence of the FMDV IRES, coding region, and 3′UTR and mRNAs for RPS16 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (control). We identified the FMDV IRES, 3D, and 3′UTR in the samples immunoprecipitated with antibody specific to DDX3 or RPL13 (Fig. 4A, lanes 3, 4, 9, and 10). RT-PCR did not detect RPS16 and GAPDH in the precipitated samples (Fig. 4B, lanes 3, 4, 9, and 10). No RT-PCR band representing FMDV IRES, 3D, or 3′UTR, RPS16 mRNA, or GAPDH mRNA was detected with the isotype IgG antibody (lanes 5 and 11), without antibody (lanes 6 and 12), or with H2O as a template (lanes 7 and 13). These results confirm that DDX3 and RPL13 specially bind the FMDV genome in FMDV-infected cells.
RPL13 associates with the FMDV IRES in DDX3-dependent manner. (A and B) FMDV-infected BHK-21 cell lysates at 5 h postinfection (MOI, 1) were generated and subjected to immunoprecipitation assay with RPL13 or DDX3 antibody. Following washing and dissociation, RNA was extracted and subjected to the RT-PCR using primers that are specific to the FMDV IRES or 3D/3′UTR (A) or RPS16 or GAPDH (B). (C) Top, schematic representation of the FMDV genome. Bottom, cell lysates were incubated separately with the biotin-labeled 5′UTR, S fragment, cre, IRES, and 3′UTR of FMDV. Nonbiotinylated RNA probes were used as controls. After being pulled down by streptavidin beads, the protein complex was dissolved for Western blotting. (D to F) Cells were treated with either a control siRNA (siCtrl) or an siRNA targeting RPL13 (D) or DDX3 (E and F) for 48 h. Cell extracts were then harvested and subjected to biotinylated IRES pulldown assay followed by Western blot analysis.
To further assess the reciprocal interactions and to identify the functional elements in the FMDV genome required for these interactions, we then performed RNA-protein pulldown assay. RNA probes corresponding to the positions of the 5′UTR, S fragment, Cre, IRES, and 3′UTR were biotinylated and incubated with BHK-21 cell extracts (Fig. 4C, top). The RNA-protein complex was pulled down with streptavidin beads and analyzed by immunoblotting. The results revealed that either RPL13 or DDX3 is pulled down by the beads if biotinylated 5′UTR or IRES is used (Fig. 4C, bottom). Collectively, the results verify the association of RPL13 and DDX3 with the IRES of the FMDV 5′UTR.
RPL13 associates with the FMDV IRES in a DDX3-dependent manner.It is important to determine whether RPL13 or DDX3 binds to IRES elements directly. Our results showed that the depletion of RPL13 does not affect DDX3 binding to the biotinylated IRES (Fig. 4D), whereas depletion of DDX3 impairs the association between RPL13 and the IRES (Fig. 4E). This result revealed that DDX3 is the functional link between RPL13 and IRES-dependent translation. Under the same experimental conditions, we assessed other RPs belonging to the 60S ribosomal subunit (RPLP0, RPL22, and RPL11) and the 40S ribosomal subunit (RPS5, RPS6, and RACK1) for involvement in DDX3-mediated IRES translation. It should be noted that all of the RPs tested and biotinylated IRES showed an interaction; depletion of DDX3 did not disrupt the specific interactions (Fig. 4F). Thus, RPL13 may act as the unique RP interacting with DDX3. These results reinforce the conclusion that RPL13 interacts with the FMDV IRES in a DDX3-dependent manner.
C-terminal DDX3 is required for association with the IRES but is not sufficient for IRES-driven translation.DDX3 was proposed to promote translation of cellular and viral transcripts harboring complex 5′UTRs via its RNA helicase activity or to stimulate translation by supporting ribosomal subunit joining in an ATP-independent manner (18–20, 55). To address whether the ATP-binding domain of DDX3, the RNA-binding domain, or both are responsible for association with the viral IRES, an RNA pulldown assay was employed using lysates from BHK-21 cells transiently overexpressing WT DDX3 or variant DDX3ΔC or DDX3ΔN. We found that either DDX3 WT or DDX3ΔN binds to biotinylated IRES elements (Fig. 5A). Furthermore, biotinylated IRES and recombinant proteins of DDX3-N, DDX3-C, or RPL13 were mixed; the RNA-protein complex was pulled down with streptavidin beads and analyzed by immunoblotting. The results revealed specific interaction between the IRES and DDX3-C (Fig. 5B). Collectively, these results support that the C-terminal region of DDX3 is required for association with the viral IRES.
The C-terminal region of DDX3 is required for association with the FMDV IRES but is not sufficient for IRES-driven translation. (A) Cell lysates from BHK-21 cells transfected with wild-type DDX3 or the variant DDX3△C or DDX3△N were generated and subjected to biotinylated IRES pulldown assay. hnRNPK was set as a positive control for IRES binding. (B) Biotinylated IRESs were incubated with increasing concentrations of recombinant DDX3-N (lanes 2 to 4), DDX3-C (lanes 5 to 7), or RPL13 (lanes 8 to 10). The bound complexes were analyzed by immunoblotting, and the “input” was detected by Coomassie blue staining. Lane 1, IRES alone, incubated without added protein. (C) Cells transfected with 1 μg of vector encoding wild-type DDX3 or the indicated variants for 24 h were transfected with 0.3 μg of bicistronic plasmid psiCHECK-FMDV. The RLuc and FLuc activities were determined at 24 h posttransfection. (D) BHK-21 cells were transfected with siRNAs either control siRNA (siCtrl) or an siRNA targeting DDX3 for 48 h and then treated and assessed as described in for panel C. For panels C and D, results were normalized to the control (arbitrarily set to 100%) and are presented as mean ± SD from three independent experiments.
To reveal the role that C-terminal region of DDX3 plays in IRES-dependent translation, cells were transiently overexpressed with WT DDX3, DDX3ΔN, DDX3ΔC, or DDX3 variants, with subsequent transfection of cells with the bicistronic plasmid psiCHECK-FMDV. The results revealed an increasing IRES activity stimulated by the ectopic expression of WT DDX3 but not variant DDX3ΔN or DDX3ΔC. DDX3 variants (Q207A, K230E, E348Q, and S382L) that affect ATPase activity or unwinding activity (56) can promote IRES-dependent translation to an extent comparable to that with WT DDX3 (Fig. 5C). Furthermore, rescue experiments were performed using WT DDX3 and variants in DDX3-silenced cells. The results showed that the DDX3 Q207A, K230E, E348Q, and S382L variants partly restore IRES activity; DDX3ΔN and DDX3ΔC, however, had a negligible effect (Fig. 5D). It seems that C-terminal region of DDX3 is not sufficient for IRES activity. Our findings suggest that multiple active-site residues on DDX3 are involved in regulating IRES-driven translation and suggest that both N- and C-terminal regions are essential for function in DDX3.
RPL13 is a downstream interconnection for DDX3 stimulation of IRES activity and viral replication.There has as yet been no reported study of the way in which ribosomal proteins are related to DDX3 stimulation of IRES-driven translation. We therefore explored whether depletion of RPL13 influences the DDX3 stimulation of IRES activity and replication of FMDV. DDX3 inhibitors RK-33 (57) and Ketorolac salt (58), which selectively dock into its ATP-binding pocket and potently inhibit its RNA helicase activity, were used. The present study showed that RK-33 at 10 μM or Ketorolac salt at 500 μM potently depressed the levels of DDX3 and viral proteins (Fig. 6A), as well as IRES-dependent translation (Fig. 6B), but had a negligible effect on cap-dependent translation and cell viability (data not shown). Importantly, the depletion of RPL13 did not significantly further reduce IRES activity in cells treated with RK-33 (Fig. 6C) or Ketorolac salt (Fig. 6D). Moreover, ectopic expression of DDX3 promoted IRES-dependent translation, but this promotion effect disappeared when RPL13 was knocked down (Fig. 6E). These results indicate that RPL13 modulates DDX3-mediated FMDV IRES translation.
DDX3 cooperates with RPL13 to promote the IRES activity and replication of FMDV. (A) BHK-21 cells pretreated with indicated amounts of RK-33 (left) and Ketorolac salt (right) for 30 h were infected with FMDV (MOI, 0.5) for 8 h in the presence of the indicated inhibitor. The cell lysates were collected and subjected to Western blotting. (B) BHK-21 pretreated with indicated amounts of RK-33 and Ketorolac salt for 30 h were transfected with bicistronic plasmid psiCHECK-FMDV for 24 h in the presence of the indicated inhibitor. The cell lysates were then subjected to luciferase reporter assay. (C to E) Control and RPL13-depleted BHK-21 cells were transfected with psiCHECK-FMDV DNA in the presence of the indicated concentrations of RK-33 (C) and Ketorolac salt (D), or siRNA-treated cells were transfected with FLAG empty vector or increasing amounts of FLAG-DDX3 for 24 h and then transfected with psiCHECK-FMDV (E). The RLuc and FLuc activities were determined at 24 h posttransfection. (F) BSR-T7 cells with the same treatment as described for panels C to E were transfected with the rFMDV-EGFP. At 24 h posttransfection, the cells were subjected to fluorescence analysis. For panels B to F, results were normalized to the control (arbitrarily set to 100%) and are presented as mean ± SD from three independent experiments.
To further strengthen these findings, we assessed the effect of treatment with RK-33, Ketorolac salt, or ectopic expression of DDX3 on the activity of the replicon in RPL13-depleted BSR-T7 cells. Consistent with the findings for the IRES activity, there was no significant effect on the replication of FMDV (Fig. 6F). Collectively, these results prove that RPL13 is a downstream interconnection for DDX3 stimulation of IRES-driven translation and replication of FMDV.
Distribution of RPL13 and DDX3 on ribosome subunits.To address the distribution of RPL13 and DDX3 on the ribosome subunits, we carried out sucrose gradient fractionation of extracts from mock- or FMDV-infected BHK-21 cells to separate ribosomal subunits (40S and 60S), monosomes (80S), and progressively larger polysomes. As shown in Fig. 7A and B, the peaks in the polysome profiles were generated by measuring profiles of the optical density at 254 nm (OD254) in the gradient fractions and by immunoblot probing for the ribosomal proteins RPS6 (40S component) and RPL13 (60S component); both ribosomal proteins were detectable in the 80S as well as in the polysomal fractions. eIF2α, a canonical translation initiation factor, was found mostly at the top of the gradient, as well as partially sedimenting in the 40S and 60S fractions. Absorbance monitoring at 254 nm showed decreased polysome distribution profiles after FMDV infection, indicating that FMDV infection compromises global translation.
DDX3 cooperates with RPL13 to support the assembly of 80S ribosomes for optimal translation initiation of FMDV mRNA. (A and B) Top, extracts were generated from mock-infected (A) and FMDV-infected (MOI, 1) (B) BHK-21 cells (5 h postinfection), sedimented through 10 to 50% sucrose gradients, and fractionated. Bottom, the collected fractions were subjected to Western blotting to determine the cosedimentation of DDX3 and RPL13 with ribosomal subunits (40S and 60S), monosomes (80S), or polysomes. (C and D) BHK-21 cells transfected with either a control siRNA (siCtrl) or siRNA targeting DDX13 (top) or RPL3 (bottom) were infected with FMDV (MOI, 1). At 4 h postinfection, cell lysates were resolved and fractionated through sucrose gradients. The fractions were then subjected to RT-qPCR analysis for FMDV 3D (C) and GAPDH mRNA (D) levels. The transcript number is expressed as the percentage of total mRNA transcripts recovered and is plotted against fraction number. (E) Polysome profiles of mock-infected BHK-21cells transfected with a control siRNA (siCtrl) or siRNA targeting DDX3 (top) or RPL13 (bottom). (F) BHK-21 cells transfected with either a control siRNA (siCtrl) or siRNA targeting RPL13 (left) or RPL22 or RPL40 (right) were infected with FMDV (MOI, 1). At 4 h postinfection, cell lysates were generated and subjected to immunoprecipitation (IP) with DDX3 antibody followed by Western blot analysis with the indicated antibodies. (G) Lysates from control (siCtrl) and RPL13-depleted (siRPL13) cells infected with FMDV were fractionated through sucrose gradients to isolate ribosomal complexes followed by IP with DDX3 antibody. The “input” and the bound complexes were analyzed by Western blotting with the indicated antibodies.
Furthermore, probing for DDX3 revealed that most of the protein remained in the 40S gradient fractions and that a small quantity of the protein tailed into the 60S and 80S fractions in both mock-infected (Fig. 7A) and FMDV-infected (Fig. 7B) cells. DDX3 was undetected in the polysome fractions, indicating that it does not participate in the elongation phase of the translation process; this is in agreement with previous reports (19). However, RPL13 was present in the 60S, 80S, and polysome fractions, with no detection as part of the 40S fraction, with or without FMDV infection. These results suggest that the interaction of DDX3 and RPL13 for regulating the translation initiation takes place specifically on the ribosome.
DDX3 and RPL13 are specific for the translation of viral mRNA.To gain insight into the mechanism of translational control by DDX3 and RPL13, we compared the formation of ribosomal complexes on FMDV mRNA with that for the cellular transcript GAPDH. Cells depleted of DDX3 or RPL13 were infected with FMDV, and lysates were fractionated through sucrose gradients. Following fractionation, the association of ribosomal subunits with specific mRNA was determined by quantitative RT-PCR analysis. As shown in Fig. 7C, FMDV mRNA levels peaked at fractions 14 to 15 in control cells, and the distribution of FMDV mRNA shifted considerably toward lighter fractions when DDX3 (top) or RPL13 (bottom) was knocked down, peaking at fractions 9 to 10. This indicates that FMDV mRNA formed on average smaller polysomes after DDX3/RPL13 depletion. In contrast, the distribution of GAPDH mRNA was not affected by DDX3/RPL13 knockdown (Fig. 7D). Furthermore, consistent with previous reports (15, 16, 18, 19), polysome profiles in mock-infected cells revealed that canonical global translation is not compromised by DDX3 depletion (Fig. 7E, top) and is also not affected by RPL13 depletion (Fig. 7E, bottom). Given that the selective translation of monosomes versus polysomes has been reported (59), it is tempting to speculate that the increase in the polysome/80S ratio may be indicative of preferential translation of selective mRNAs. Taken together, these results demonstrate that DDX3 and RPL13 are required for translation initiation of FMDV mRNAs but not for canonical cellular transcripts.
The functional interaction of DDX3 with RPL13 on the ribosome.We next attempted to gain insight into the mechanisms of the cooperation of DDX3 and RPL13 in viral translation initiation. Here, we carried out immunoprecipitation (IP) with DDX3 antibody from cytoplasmic extracts of FMDV-infected BHK-21 cells and analyzed the precipitate for associated RPs by Western blotting. As shown in Fig. 7F, DDX3 specifically coprecipitated with these tested RPs, except for RPL40. Depletion of RPL13 disrupted the interactions of DDX3 with 60S subunit RPs (RPL22 or RPLP0) but not with 40S subunit RPs (RPS5, RPS6, RPS11, or RACK1) (left). In contrast, depletion of RPL22 or RPL40 did not affect the interactions of DDX3 with other RPs (right). These results demonstrate the role that RPL13 plays in DDX3-mediated translation initiation.
To further address the functional interaction of DDX3 with RPL13 on the ribosome, various ribosome subunits from lysates of FMDV-infected BHK-21 cells were isolated, and IP with DDX3 antibody was then performed. As shown in Fig. 7G, a small quantity of RPL13 coprecipitated with DDX3 in 60S ribosomal subunits (lane 5), but large amounts of RPL13 bound to DDX3 in 80S ribosomes (lane 7). Furthermore, knockdown of RPL13 decreased the amount of DDX3 embedded in the 80S ribosome and suppressed joining of the 60S subunit (RPL22) (lanes 6 and 8). Interactions of DDX3 with different components (eIF2a, eIF4A, and RPS6) of the 43S preinitiation complex, however, were not affected by depletion of RPL13 (lanes 1 to 4). Collectively, these results suggest that DDX3 recognition of RPL13 supports the assembly of 80S ribosomes for optimal translation initiation of viral mRNA.
DDX3 affects the ability of eIF3e and -3j to bind to FMDV IRES.FMDV shuts down cellular global and cap-dependent translation, allowing for translation of its own viral mRNA through the IRES to recruit ribosomes to specific viral mRNA and a number of eIFs, such as eIF2, eIF3, and eIF4 (60–64). In the current study, it was demonstrated that eIF4A, eIF4G, and eIF5B were cleaved in FMDV-infected BHK-21 cells (Fig. 8A), consistent with previous data (60, 65, 66). We also detected that two eIF3 subunits, eIF3a and eIF3d, underwent substantial reduction, and eIF3h was partially cleaved in the late stages of FMDV infection. In addition, eIF3e did not undergo proteolytic modification. Instead, there was a progressive upregulation of the subunit eIF3j during FMDV infection (Fig. 8A). These results reveal that FMDV can exert a variety of actions upon the eIFs in order to efficiently initiate viral protein synthesis.
DDX3 affects the ability of eIF3e and -3j to bind to the FMDV IRES. (A) Lysates of BHK-21 cells infected with FMDV (MOI, 0.5) were harvested at the indicated times and subjected to Western blotting with the indicated antibodies. (B and C) BHK-21 cells transfected with either a control siRNA (siCtrl) or siRNA targeting DDX3 (B) or RPL13 (C) for 48 h were challenged with FMDV for 0, 4, and 8 h. Cell lysates were harvested and analyzed by Western blotting. (D and E) Lysates of siRNA-transfected cells were collected and subjected to biotinylated IRES pulldown assay followed by Western blot analysis. (F) Cells transfected with a control siRNA (siCtrl) or the indicated eIF3 subunit-targeting siRNA for 48 h were infected with FMDV, and cells and supernatants were harvested for the indicated times. Virus titer was determined by TCID50 assay. (G) Cells were silenced with the indicated siRNAs for 72 h, dead cells were estimated using the LIVE/DEAD viability/cytotoxicity kit (column), and cell viability was measured using the MTS assay (scatter). The data in panels F and G represent the mean ± SD from three independent experiments (*, P < 0.05; **, P < 0.01).
Given the importance of eIFs in FMDV translation (34–36, 61), we asked whether several eIFs may specially participate in the observed effects of DDX3/RPL13 on viral translation initiation. We found that depletion of DDX3 or RPL13 did not affect the expression level of the eIFs indicated above (Fig. 8B and C). A biotin-RNA pulldown assay was further performed. Western blot analysis of the pulldown samples demonstrated that eIF2α, -3a, -3e, -3h, -3j, -4A, and -4G and PABP can bind to biotinylated IRES (Fig. 8D and E, lanes 3); the exceptions are eIF3d and -5B, and this suggests that they may not be required for FMDV IRES translation initiation. Notably, depletion of DDX3 greatly impaired the ability of eIF3e and -3j to bind to the IRES (Fig. 8D, lane 4); depletion of RPL13 only moderately affected the association of the IRES with eIF3e (Fig. 8E, lane 4). These results demonstrate that DDX3 affects the ability of eIF3e and -3j to bind to the FMDV IRES.
Furthermore, we found that the depletion of either eIF3e or -3j greatly depressed the viral yield (Fig. 8F) and level of viral protein (data not shown); depletion of eIF3j had no detectable cytotoxicity, and depletion of eIF3e minimally decreased cell viability (Fig. 8G). The depletion of eIF3a and -3d, on the other hand, reduced the production of viral particles to an extent comparable to that for eIF3e or -3j, but it caused detectable cell death with high cytotoxicity (Fig. 8F and G). Collectively, these results indicate that both eIF3e and -3j are involved in DDX3-mediated FMDV IRES translation.
RPL13 is required for replication of SVV and CFSV but not for replication of VSV.To determine whether the depletion of RPL13 affects the translation of other viral IRESs, IRES activity for SVV (type III-like IRES) and CSFV (type III IRES) (3, 5) was tested in RPL13-depleted cells. The results with bicistronic plasmid and reporter mRNA showed that IRES activity was strongly decreased to 50% to 60% (Fig. 9A and B). Furthermore, depletion of RPL13 greatly depressed the viral yields of SVV (Fig. 9C) and CFSV (Fig. 9D), by 15-fold at 12 hpi and by 6.3-fold at 48 hpi, respectively. The RNA levels of SVV and CFSV were also depressed (Fig. 9E and F). It seems that RPL13 is an essential determinant for translation of SVV and CFSV, indicating that its regulator function is conserved for IRES-containing picornaviruses and flaviviruses.
RPL13 is required for SVV and CSFV replication but not for VSV replication. (A and B) BHK-21 cells were treated with a control siRNA (siCtrl) or RPL13-targeting siRNA (siRPL13) for 48 h before transfection of the indicated bicistronic plasmid psiCHECK-SVV or psiCHECK-CSFV (A) or the corresponding capped bicistronic mRNAs (B). The RLuc and FLuc activities were determined at 24 h (A) or 6 h (B) posttransfection. Results were normalized to the control (arbitrarily set to 100%) and are presented as mean ± SD from three independent experiments. (C to F) Control and RPL13-depleted IBRS-2 cells or ST cells were infected with SVV (MOI, 0.5) or CSFV (MOI, 0.05), respectively. At the indicated times, the supernatants and cell lysates were collected, and virus yields were determined by TCID50 assay (C and D); total RNAs were extracted and subjected to RT-qPCR (E and F). (G to I) Control and RPL13-depleted Vero cells were challenged with VSV-EGFP (MOI, 0.5). At 8 h postinfection, cells were monitored by fluorescence microscopy (G), and the viral yields and viral RNA levels for the indicated time points were determined by TCID50 (H) and RT-qPCR (I), respectively. In Panels C to F, H, and I, data represent the mean ± SD from three independent experiments (**, P < 0.01).
Interferon-deficient Vero cells were transfected with a specific siRNA targeting RPL13 and then infected with a reporter VSV (VSV-EGFP) that expresses EGFP as a marker of infection, and the percentage of infected cells was determined by fluorescence microscopy. We observed that depletion of RPL13 did not affect viral replication (Fig. 9G), viral yields (Fig. 9H), or the RNA level of VSV (Fig. 9I). Amplification of VSV, a virus that depends solely on cap-dependent translation, was not hindered when RPL13 was depleted. This agrees with our findings that RPL13 does not affect cap-dependent translation or significantly hinder global translation rates.
These findings reveal a target for the development of broad antiviral interventions against viruses that use 5′ cap-independent mechanisms for the translation of their RNAs.
DISCUSSION
The structure of the 60S ribosomal subunit contains 42 proteins, of which 16 are present in all domains of life, 20 are present in both eukarya and archaea, and 6 are eukaryote specific (22, 67, 68). In eukaryotes, certain RPs belonging to the large 60S subunit, such as RPL38 and RPL35, regulate selective translation of specific mRNAs (69–71). With assistance from RPL38, cellular IRESs directly recruit ribosomes to initiate translation of Hox genes (69, 71). Like RPL38, RPL13 is not homologous to any other ribosomal proteins of known structure, which hence expands the range of protein folds found in ribosomes (22, 67). Long terminal extensions of eukaryotic/archaeal protein RPL13 mediate contacts between the tips of expansion segments, which serve as binding platforms for eukaryote-specific proteins or eukaryote-specific extensions of RPs (67). In the current study, dual-luciferase reporter, virus replicon, RNA-protein IP/pulldown, and polysome profiles showed that mRNA translation initiation and replication of FMDV depend specifically on RPL13 (Fig. 3, 4, 6, and 7). These results are beginning to provide evidence that RPL13 contributes to the translation efficiency of viral IRESs.
RPL13 plays an essential role in the progression of some gastrointestinal malignancies and may make cancer cells more resistant to apoptotic stimuli (72); the decrease of RPL13 produced a less-inhibitory response in the normal fibroblast cell lines (73). Furthermore, several lines of evidence indicate that global translation is unaffected by the depletion of RPL13: (i) the activities of reporter Renilla luciferase, which utilize cap-dependent translation, are not inhibited (Fig. 3A and B); (ii) there is no defect in global translation initiation based on polysome analysis (Fig. 7E, bottom); and (iii) amplification of VSV, which relies solely on cap-dependent translation for viral protein synthesis, is not inhibited (Fig. 9G to I). These lines of evidence suggest the preferential engagement of RPL13 in bolstering viral IRES-driven translation.
Furthermore, RPL13 directly interacts with DDX3 (Fig. 2) and associates with IRES elements of FMDV by binding to DDX3 (Fig. 4). Depletion of RPL13 reduces the efficacy of DDX3 improvement of viral IRES-driven translation and replication (Fig. 6C to F), supporting the conclusion that RPL13 exerts a previously unappreciated regulatory function. DDX3 reduces the 5′ RNA secondary structures and removes the proteins bound to the 5′ extremity, to facilitate 43S ribosome binding (18), but DDX3 depletion has no effect on the formation of the 48S complex (19). Moreover, DDX3 is present in newly assembling 80S ribosomes but does not contribute to translation elongation (19). The polysome profile with IP on the ribosome subunit revealed that DDX3 cooperates with RPL13 to assist the 60S subunit joining process to assemble functional 80S ribosomes for optimal translation initiation of FMDV mRNA (Fig. 7). Taken together, our results support a model in which the activity of IRESs is augmented by DDX3, which recruits RPL13, and both proteins participate in the formation of ribosomal initiation complexes. These results make important contributions to the understanding of how RPL13 is kinetically recognized and recruited by DDX3 during viral IRES-driven translation.
Viral IRESs may directly bind eIFs to either inhibit host protein synthesis or efficiently express viral proteins (1–5). Recent reports offer evidence that multisubunit eIF3 is a major partner of DDX3; they directly interact to mediate a conserved general function in promoting cap-dependent translation (16, 74). During canonical initiation, eIF3 stimulates ternary complex recruitment to the 40S ribosome as a component of the 43S ribosome (75), and it also prevents premature association of 43S complexes with 60S ribosomal subunits (76, 77). By binding to eIF3, HCV-like IRESs would reduce the competition with this factor for binding to the 40S subunit and impair formation of 43S, which, in turn, might aid the ability of these IRESs to compete with cellular mRNAs (78).
In mammals, eIF3 is composed of 13 nonidentical subunits, including the octameric core of subunits (a, c, e, f, h, k, l, and m) and peripheral subunits (b, d, g, i, and j) (79, 80). eIF3e induces or regulates eIF4E phosphorylation (81), and this alteration in phosphorylation level is related to the translation of specific mRNAs involved in cellular transformation, immune responses, and viral infection (2, 82, 83). Intriguingly, the subunit eIF3e in yeast is involved in translation of a selective set of RNAs (84). Another eIF3 subunit of eIF3j, which is located in the decoding center of the 40S subunit, facilitates the recruitment of mRNA to the 40S subunit (85). This subunit is required for high-affinity binding of eIF3 to 40S subunits in vitro and in vivo (86). Notably, several observations support a role for eIF3j in selective mRNA translation, and eIF3j is specifically required for HCV IRES-driven translation (27, 63, 87). In the current study, depletion of DDX3 was found to impair the eIF3e/j binding to the FMDV IRES (Fig. 8D). The current results raise the possibility that DDX3 and eIF3e/j act together in preferential translation of viral mRNAs.
Our study supports a model whereby the FMDV IRESs recruit DDX3 recognizing RPL13 for translation initiation. Future studies will be required to analyze the three-dimensional structure of the complex to clarify the functions in efficient translation of mRNAs encoding viral proteins.
MATERIALS ANS METHODS
Cells and viruses.BHK-21 (baby hamster kidney, ATCC CCL-10), IBRS-2 (porcine kidney, ATCC CRL-1835), Vero (African green monkey kidney, ATCC CCL-81), and ST (swine testicular, ATCC CRL-1746) cells were maintained in Dulbecco modified Eagle medium (DMEM) (Gibco, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco) and penicillin-streptomycin (100 U/ml and 100 μg/ml, respectively) (Gibco) at 37°C in 5% CO2. BSR-T7 cells (a kind gift from Karl-Klaus Conzelmann [88]), which stably express T7 RNA polymerase, were grown in DMEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 1 mg/ml G418 (Life Technologies, CA, USA).
FMDV serotype O strain O/BY/CHA/2010 (GenBank accession no. JN998085.1) was preserved by the OIE/National Foot-and-Mouth Disease Reference Laboratory (Lanzhou, China). Propagation of FMDV was carried out in BHK-21 cells, and viral titers were determined by a 50% tissue culture infective dose (TCID50) assay in BHK-21 cells. SVV (accession no. KY747511.1) propagated in IBRS-2 cells and CSFV (accession no. AY805221.1) propagated in ST cells were stored in our laboratory. A reporter vesicular stomatitis virus that expresses EGFP (VSV-EGFP) was generated as described previously (89) and amplified in Vero cells.
Antibodies and reagents.Anti-RPL13, anti-RPS5, anti-RPS11, anti-RPLP0, anti-RPL22, anti-RPL40, anti-RACK1, anti-DDX3, anti-RPS6, anti-hnRNPK, anti-eIF3a, anti-eIF3d, anti-eIF3e, anti-eIF3h, anti-eIF3j, anti-eIF4GI, and anti-PABP were purchased from Abcam (Cambridge, MA, USA). Anti-eIF4A, anti-eIF5B, anti-GFP, anti-FLAG, and anti-β-actin were purchased from Santa Cruz Biotechnology (CA, USA). Anti-eIF2α was purchased from Cell Signaling Technology (Danvers, MA, USA). Anti-RPL11, anti-His, and the secondary antibodies conjugating with horseradish peroxidase (HRP) and fluorescein isothiocyanate (FITC), were purchased from Sigma-Aldrich (St. Louis, MO, USA). Polyclonal pig antiserum against FMDV was prepared by our laboratory. The DDX3 inhibitors RK-33 and Ketorolac salt were purchased from Selleck Chemicals (Houston, TX, USA).
RNAi.For RNA interference (RNAi), small interfering RNAs (siRNAs) targeting candidate genes and negative-control (NC) siRNA were synthesized by Genepharma (Shanghai, China). The sequences of the siRNAs are as follows: hamster RACK1, 5′-GGUCCAGGAUGAGAGUCAU-3; hamster RPS5, 5′-CCGAUGAUGUGCAGAUCAA-3′; hamster RPS11, 5′-GCAAGACUGUGCGCUUCAA-3′; hamster RPS25, 5′-GCUGGUUUCUAAGCACAGA-3′; hamster RPLP0, 5′-CCGAGAAGACCUCUUUCUU-3′; hamster RPL13, 5′-GCCCUACAGUGAGAUACCA-3′; hamster RPL22, 5′-CCUGAAGAAGAACAAUCUU-3′; hamster RPL40, 5′-CCAAGAUAAGGAAGGCAUU-3′; hamster DDX3, 5′-GCAGUCGUGGACGUUCUAA-3′; hamster eIF3a, 5′-GGAACUGUGUGUGGAUCUU-3; hamster eIF3d, 5′-GCCGUUCAGCAAAGGAGAU-3; hamster eIF3e, 5′-GCAGGAUGUUAUUUGACUA-3; hamster eIF3h, 5′-GCCUGGUAGUAUUAAAGAU-3; hamster eIF3j, 5′-GGCAGAUUAUGGUGGAUAU-3; porcine RPL13, 5′-GGAAUGGCAUGAUCCUGAA-3′; and Chlorocebus sabaeus RPL13, 5′-GCCGGAAUGGCAUGAUCUU-3′.
Cells grown to 70% confluence were treated with siRNA using Lipofectamine RNAi MAX (Invitrogen, CA, USA) according to the manufacturer’s instructions.
Plasmid constructs.cDNAs of RPL13 and DDX3 were amplified from total mRNA of BHK-21 cells by RT-PCR and inserted into the EcoRI/XhoI sites of the pCMV-N-FLAG vector (Beyotime Biotechnology, Shanghai, China) to obtain FLAG-tagged DDX3/RPL13 or were cloned into the XhoI/EcoRI sites of pEGFP-N1 vector (Clontech, USA) to obtain a GFP-tagged construct. Site-directed mutagenesis of pCMV-FLAG-DDX3 was carried out with the QuikChange mutagenesis kit (Agilent Technologies). FLAG-tagged DDX3ΔN (with deletion of residues 133 to 406) or DDX3ΔC (with deletion of residues 407 to 584) was obtained by inserting the synthesized gene (Genewiz, Suzhou, China) into pCMV-N-FLAG vector. pET30-DDX3-N/DDX3-C/RPL13 were obtained by inserting the corresponding cDNAs into the XhoI/EcoRI sites of the pET30a vector (EMD Biosciences). pGST-RPL13 was obtained by subcloning the EcoRI/XhoI fragment from the pCMV-FLAG-RPL13 construct into the pGEX-6p vector (Amersham Biosciences). Bicistronic reporter plasmids were constructed as previously described (30). Briefly, the sequences of FMDV (IRES), SVV (IRES), or CSFV (IRES plus 69 bases of the coding region) fragments were cloned into the psiCHECK-2 vector (Promega, WI, USA). All DNA constructs were verified by sequencing.
Immunoprecipitation assay.BHK-21 cells were lysed with radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime Biotechnology) for 1 h on ice and then subjected to centrifugation at 15,000 × g for 20 min at 4°C. Supernatants were immunoprecipitated with the appropriate antibodies at 4°C overnight. The immune complexes were incubated with protein G-agarose beads (GE Healthcare, Chicago, IL, USA) for 2 h, washed five times with lysis buffer, and eluted in 1× SDS-PAGE sample buffer for Western blot analysis.
Gel filtration analysis.The purified proteins eluted from a His-Ni column or glutathione Sepharose 4B beads were loaded onto a Superdex G75 column in a solution buffer of 20 mM Tris-HCl, pH 8.0. The peak MW was estimated by comparing the substrate with the protein standards running on the same column. The analytical column was calibrated with a series of individual runs of molecular mass standard proteins as markers, including bovine serum albumin (68 kDa), egg white albumin (43 kDa), and ribose nucleotidase (13.7 kDa). The peak protein was collected and analyzed by 10% SDS-PAGE.
Bicistronic dual-reporter assay.Cells in which DDX3 or RPL13 was knocked down or overexpressed were transfected with the indicated bicistronic constructs. At 24 h posttransfection, cell extracts were generated using a passive buffer and examined for RLuc and FLuc activity using the Dual-Glo luciferase assay system (Promega). For the bicistronic mRNA reporter assay, the capped mRNA was obtained by incorporating a Ribo m7G cap analog (Promega) into the transcript during the RiboMAX transcription reaction and transfected into pretreated cells. After 6 h posttransfection, cells were harvested and subjected to dual-luciferase reporter assay.
Construction of the FMDV-EGFP replicon, rFMDV-EGFP.The subgenomic replicon rFMDV-EGFP was constructed by replacing the Lb and P1 genes of the full-length infectious clone of FMDV O/HN/CHA/93 (pOFS-K1234) with an EGFP reporter gene (90). The replicon linearized with NotI was transfected into BSR/T7 cells using Lipofectamine LTX (Invitrogen).
RNA immunoprecipitation and RT-PCR.Lysates from FMDV-infected BHK-21 cells for use in immunoprecipitation assays were collected at 5 h postinfection and preincubated with protein A-agarose (GE Healthcare) on ice for 1 h. Nonspecific complexes were pelleted by centrifugation at 1,000 × g at 4°C for 10 min. The supernatants were recovered, and 100-μl samples were each diluted with 450 μl of lysis buffer and then added to either 8 μl of the indicated antibody or 8 μl of buffer containing no antibody. This was followed by incubation on ice for 2 h. Prewashed protein A-agarose was added to each sample, which was then incubated on ice for 1 h. RNA-protein coimmunoprecipitation complexes were pelleted by centrifugation at 1,000 × g at 4°C for 5 min and washed three times with lysis buffer. Each pellet was resuspended in 400 μl of proteinase K buffer (100 mM Tris-HCl [pH 8.0], 12.5 mM EDTA, 150 mM NaCl, 1% SDS) and incubated with 100 μg predigested proteinase K for 30 min at 37°C. RNA was extracted with TRIzol reagent (Invitrogen) and subsequently RT-PCR amplified. RT-PCR was performed using a PrimeScript one-step RT-PCR kit (TaKaRa, Dalian, China) and primers specific to the FMDV IRES (5′-CACAGGTTCCCACAACCGACAC-3′ and 5′-CAGTGATAGTTAAGGAAAGGC-3′), primers specific to FMDV 3D and 3′UTR (5′-GTTGCTAGTGATTATGACTTGGAC-3′ and 5′-CTTACGGCGTCGCTCGCCTCAGAG-3′), primers specific to RPS16 RNA (5′-TCGCAGCCATGCCGTCCAAGGGT-3′ and 5′-TCATTAAGATGGGCTCATCGGT-3′), or primers specific to GAPDH RNA (5′-TCCATGCCATCACGGCCACCCAG-3′ and 5′-ACTCTTGAAGTCGCAGGAGACAAC-3′). PCR products were resolved in a 1% agarose gel prestained with Gel Red nucleic acid gel stain (Biotium, CA, USA).
In vitro transcription and synthesis of biotinylated RNA.Viral cDNAs corresponding to the 5′UTR, the S fragment, the cis-acting replication elements (cre), the IRES, and the 3′UTR of the FMDV genome were amplified from the cDNA of FMDV strain O/BY/CHA/2010 and cloned into the pcDNA3.1 vector (Invitrogen). Prior to in vitro transcription, these cDNAs were linearized by BamHI digestion. RNA transcripts were synthesized using the RiboMAX Large Scale RNA Production System-T7 (Promega), according to the manufacturer’s instructions.
Biotinylated RNAs were synthesized using the Pierce RNA 3′-end desthiobiotinylation kit (Pierce Biotechnology, Rockford, IL, USA), following the manufacturer’s instructions. Briefly, T4 RNA ligase was used to attach a single biotinylated cytidine bisphosphate nucleotide to the 3′ terminus of the RNA transcript.
Biotinylated RNA pulldown assay.Cell extracts were prepared as described above, and the protein concentration was determined using the Pierce bicinchoninic acid (BCA) protein assay kit (Pierce Biotechnology). The biotinylated RNAs (50 pmol) were incubated with prewashed streptavidin magnetic beads (50 μl) in an RNA capture buffer for 15 to 30 min at room temperature with agitation to allow binding. A reaction mixture containing 200 μg of cell extracts, 30 μl of 50% glycerol, RNA-bound beads, and protein-RNA binding buffer was made. The mixture (with a final volume of 100 μl) was incubated 60 min at 4°C with agitation, and then the protein-RNA complexes were washed three times with a wash buffer, after which 50 μl of elution buffer was added to the beads and the mixture incubated for 15 min at 37°C with agitation to dissociate the complexes from the beads. Ten microliters of 6× SDS–PAGE sample buffer was added to the samples. The samples were then boiled, subjected to 12% SDS-PAGE, and analyzed by Western blotting.
Polysome profile analysis.Cells were mock infected or infected with FMDV at an MOI of 1. At 4 h postinfection, the cells were incubated with 0.1 mg/ml cycloheximide (CHX) (Sigma-Aldrich) for 5 min at 37°C to arrest the ribosome. Subsequently, the cells were washed twice with ice-cold phosphate-buffered saline (PBS) containing CHX and collected by cell scraping. Cells were resuspended in polysomal extraction buffer (20 mM Tris-HCl [pH 7.5], 5 mM MgCl2, 100 mM KCl, 1% Triton X-100, 0.1 mg/ml CHX, 1× protease inhibitor cocktail [EDTA free], and 50 U/ml RNase inhibitor), vortexed briefly, and incubated on ice for 1 h. Cell lysates were centrifuged at 15,000 × g for 10 min at 4°C, and then supernatants were resolved on a linear 10 to 50% (wt/vol) sucrose gradient (composed of 20 mM Tris-HCl [pH 7.5], 5 mM MgCl2, and 100 mM KCl) by centrifugation at 35,000 rpm for 3 h in a Beckman SW41 Ti rotor at 4°C. Fractions were collected, and the absorbance at 254 nm was recorded using an Isco fractionator (Teledyne, USA). Total RNA was phenol-chloroform extracted, ethanol precipitated, and dissolved in distilled water (dH2O) for quantitative RT-PCR. Protein in fractions was analyzed by Western blotting. Where indicated, the separated ribosome subunits were desugared and subjected to IP with DDX3 antibody.
Cell viability assay.After the cells in 96-well plates reached 70% confluence, specific siRNAs were transfected into the cells. NC siRNA was used as a control. For 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay, 10 μl of the CellTiter 96 Aqueous One solution reagent (Promega) was directly added to the cells at 72 h posttransfection and incubated for 4 h, and then the absorbance at 490 nm was recorded. To detect the percentage of dead cells, cells were stained based on membrane integrity and intracellular esterase activity using the LIVE/DEAD viability/cytotoxicity kit (Invitrogen). After being washed twice in PBS, cells were incubated with staining reagent (2 mM calcein AM and 4 mM ethidium homodimer in PBS) at room temperature for 30 min. Images of stained cells were obtained with an Olympus IX73 inverted microscope.
ACKNOWLEDGMENTS
We are deeply grateful to Zhi-Fang Zhang of the Chinese Academy of Agricultural Sciences for kindly providing the reporter vesicular stomatitis virus that expresses EGFP (VSV-EGFP). We also thank Karl-Klaus Conzelmann (Max von Pettenkofer Institute & Gene Center, Germany) for generously providing the BSR-T7 cells.
This work was supported by the National Natural Science Foundation of China (31873023, 31672592, 31811540395, 31772739, 315725153, and 1811540395) and the National Key R&D Program of China (2017YFD0500900, 2017YFD0502300, and 2017YFD0502200).
FOOTNOTES
- Received 29 September 2019.
- Accepted 11 October 2019.
- Accepted manuscript posted online 16 October 2019.
Supplemental material is available online only.
- Copyright © 2020 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵