Viral Translation Is Coupled to Transcription in Sindbis Virus-Infected Cells

Abstract

During the late phase of Sindbis virus infection, the viral subgenomic mRNA is translated efficiently in BHK cells, whereas host protein synthesis is inhibited. However, transfection of in vitro-generated Sindbis virus subgenomic mRNA leads to efficient translation in uninfected BHK cells, whereas it is a poor substrate in infected cells. Therefore, the structure of the subgenomic mRNA itself is not sufficient to confer its translatability in infected cells. In this regard, translation of the subgenomic mRNA requires synthesis from the viral transcription machinery. The lack of translation of transfected viral mRNAs in infected cells is not due to their degradation nor is it a consequence of competition between viral transcripts and transfected mRNAs, because a replicon that cannot produce subgenomic mRNA also interferes with exogenous mRNA translation. Interestingly, subgenomic mRNA is translated more efficiently when it is transfected into uninfected cells than when it is transcribed from a transfected replicon. Finally, a similar behavior was observed for other RNA viruses, such as vesicular stomatitis virus and encephalomyocarditis virus. These findings support the notion that translation is coupled to transcription in cells infected with different animal viruses.

In cells infected with animal viruses, the regulation of translation plays a pivotal role in both viral replication and gene expression. This is the case for Sindbis virus (SV), a prototype member of the Alphavirus genus belonging to the Togaviridae family (36). The SV lytic cycle comprises two distinct translation stages. In the early phase, synthesis of viral nonstructural proteins (nsPs) takes place alongside host mRNA translation. However, a few hours later, cellular protein synthesis is strongly inhibited whereas viral structural proteins (sPs) are synthesized in abundance. Synthesis of the SV sPs is not necessary to trigger the inhibition of translation (8). Noncytopathic SV variants that do not fully block translation contain mutations in the nsp region and particularly in the nsp2 gene, suggesting that viral replication is implicated in the abrogation of host translation (1, 7, 11, 16, 30, 31). However, an SV variant that efficiently induces the shutoff of cellular mRNAs despite a low replication rate has been described previously (11). Replication of cytopathic SV induces a cellular stress response, leading to phosphorylation of the α subunit of eukaryotic initiation factor 2 (eIF2α) to block the translation machinery. In particular, protein kinase R is activated in SV-infected cells, thus promoting the phosphorylation of this initiation factor. However, protein synthesis is also inhibited in cells that lack the protein kinase R gene, so other inhibitory pathways may be activated during SV infection (17, 32, 39). It has been proposed that, during Semliki Forest virus infection, assembly of stress granules after eIF2α phosphorylation and their localized disassembly may lead to general translational arrest that only viral subgenomic mRNA can elude (23). The alteration of intracellular ionic concentrations has also been proposed as a mechanism that modulates translation in alphavirus-infected cells. Thus, subgenomic mRNA translation is not affected by elevated cytoplasmic sodium concentration, as occurs after alphavirus infection, whereas translation of host mRNAs is severely impaired under these conditions (5, 14, 15).

At least two sequences in SV subgenomic mRNA influence its translatability. In a recent article, we proposed that the 5′ untranslated region of the SV subgenomic mRNA provides independence of eIF4G to recruit this mRNA to the translation machinery (6). Moreover, the first 275 nucleotides (nt) downstream from the translation initiation codon that encode part of the C protein act as a translation-enhancing element for SV 26S mRNA (9, 10). This sequence may fold in an extensively base-paired structure thought to favor the translatability of this mRNA only in infected cells (10). As a result, 26S mRNA has a low requirement for eIF2α (39). The prevailing idea is, therefore, that the high translatability of 26S mRNA is derived from its structure. Infection of animal cells by a given virus provokes changes that favor viral mRNA translation but that would be detrimental for host protein synthesis. Contrary to this view, we now provide evidence that viral mRNAs transfected into SV-infected cells are poorly translated. The present findings suggest that SV translation is coupled to transcription of subgenomic mRNA.

MATERIALS AND METHODS

Cell line and viruses.

BHK-21 cells and vesicular stomatitis virus (VSV), encephalomyocarditis virus (EMCV), or SV were used to perform the experiments. SV virus stock was prepared from a pT7 SVwt infective cDNA clone (where wt is wild type) (33). Viral infection of BHK cells was carried out in Dulbecco's modified Eagle medium (DMEM) without serum for 40 min to permit virus attachment. Next, this medium was removed and infection continued in DMEM with 10% fetal calf serum.

Plasmids.

Plasmids were used as DNA templates for in vitro RNA transcription with T7 or Sp6 RNA polymerases. The transcription mixture always contained an m⁷G(5′)ppp(5′)G cap analog except when mRNAs containing a picornavirus internal ribosome entry site (IRES) were prepared. pT7 SVwt (33) was used as the parental plasmid for all of the constructs. The luciferase gene was derived from the plasmid pKS-Luc (38). pT7 rep C+Luc was obtained by the insertion of a double PCR product digested with AatII and ApaI in the same sites of pT7 SVwt. The double PCR product was made as follows. For the first PCRs, the primers 5′AatII and 3′joint C+luc (the sequences are shown below) were used, with pT7 SVwt as the DNA template, and primers 5′joint C+luc and 3′ApaI-luc were used, with pKs-Luc as the DNA template. A mixture of these products and the primers 5′AatII and 3′ApaI-luc were then used for the second PCR. pT7 Δnsps rep C+Luc was derived from pT7 rep C+Luc, which has deleted the sequence between the SmaI and HpaI restriction sites. pT7 C+Luc was made by inserting the SacI/ApaI-digested PCR product obtained using the oligonucleotides 5′SacI-T7prom and 3′ApaI-luc and pT7 rep C+Luc as the DNA template in the same sites of pT7 SVwt. pT7 rep C has been described previously (34). pT7 rep −26S was made by inserting the PCR product obtained with the oligonucleotides 5′HpaI and 3′ApaI-nsP4 and pT7 SVwt as the DNA template digested with HpaI/ApaI in the same sites of pT7 SVwt. Plasmids pToto1101/Luc (4), pTM1-Luc (2), and pT7 5′NCpolioLUC (38) were used to obtain luciferase RNA preceded by the SV genomic leader sequence, the EMCV IRES sequence, and the poliovirus leader sequence, respectively. pT7 SV(P₇₂₆G) was obtained by insertion of a double PCR product digested with ClaI and SpeI in the same sites of pT7 SVwt. First, oligonucleotides 5′SV-ClaI and 3′nsp2 P₇₂₆G or 5′nsp2 P₇₂₆G and 3′SV-SpeI, respectively, were used as primers and pT7 SVwt as the DNA template. Next, a mixture of the products obtained with 5′SV-ClaI and 3′SV-SpeI as primers were used for the second PCR.

Oligonucleotides.

The primer sequences are as follows: for 5′AatII, TTGTTCGACGTCAAGAAC; for 5′HpaI, TATGGCGTTAACCGGTCTG; for 5′joint C-luc, CAGAAGAGTGGTCCGCACATATGGAAGACGCCAAAAAC; for 3′joint C-luc, GTTTTTGGCGTCTTCCATATGTGCGGACCACTCTTCTG; for 5′SacI-T7prom, GCGGGCGAGCTCTAATACGACTCACTATAGATAGTCAGCATAGT; for 3′ApaI-luc, ACGCGCGGGCCCTTACAATTTGGACTTTCC; for 3′ApaI-nsP4, CAGGATTTATCCCGGGTCCATC; for 5′nsp2 P₇₂₆G, GCCTTAACGGAGGAGGCACC; for 3′nsp2 P₇₂₆G, GGTGCCTCCTCCGTTAAGGC; for 5′SV-ClaI, CATTGAATCGATATTACAG; and for 3′SV-SpeI, GTCCATACTAGTAATAGAG.

RNA transfection.

Subconfluent BHK cells, uninfected or infected according to each experiment, were harvested, washed with ice-cold phosphate-buffered saline, and resuspended at a density of approximately 2.5 × 10⁶ cells/ml in the same buffer. Then, 20 μg of in vitro-transcribed RNA synthesized from the different constructs was added to 0.4 ml of cells and the mixture was transferred to a 2-mm cuvette. Extraction of mRNAs from culture cells was carried out by use of QIAGEN Oligotex Direct mRNA. Electroporation was performed at room temperature by generating two consecutive 1.5-kV, 25-mF pulses with a Genepulser apparatus (Bio-Rad), as previously described (21). Transfection efficiency measured by in situ immunofluorescence was always higher than 90% of cells.

Analysis of protein synthesis.

BHK cells were seeded into 24-well Costar plates at a concentration of 10⁵ cells/well. At the times indicated for each experiment, the media were removed and proteins were labeled for 30 min with 0.2 ml DMEM without methionine-cysteine supplemented with 2 μl trans-labeled [³⁵S]Met-Cys (15 mCi/ml; Amersham) per well. The cells were then collected in the appropriate gel loading buffer and analyzed using 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The gels were subjected to fluorography by immersion in 1 M sodium salicylate and then dried and exposed to X-ray film.

Western blotting.

Goat antiluciferase antibodies (ABcam) and mouse monoclonal antiactin antibodies (Chemicon International) were used. The former were used to determine the luciferase production of each sample and the latter as controls to determine the amount of a housekeeping cell protein. Also, rabbit polyclonal antibodies against anti-SV capsid protein were used.

Analysis of mRNA by real-time RT-PCR.

Luciferase mRNA levels in transfected BHK cells were determined by real-time quantitative reverse transcription-PCR (RT-PCR) (6). For this purpose, total RNA was extracted from 2 × 10⁵ cells at the times indicated in each figure by use of an RNeasy commercial kit (QIAGEN) following the manufacturer's recommendations. The isolated RNA was resuspended in 30 μl of nuclease-free water, and 3 μl was subjected to analysis. Real-time quantitative RT-PCR was performed with a LightCycler thermal cycler system (Roche Diagnostics) and an RNA Master SYBR green I kit (Roche Diagnostics) as described by the manufacturer. The primers nSP2-forward (5′-GGAGGGGCTCCAGGCGGACATCG-3′), corresponding to nt 1652 to 1674 in the SV sequence, and nSP2-reverse (5′-GCTCCTCTTCTGTATTCTTGGCG-3′), corresponding to nt 2034 to 2012 in the SV sequence, were used to quantify the genomic RNA present in the cells that were transfected with the different Luc-containing replicons. The primers Luc-forward (5′-CCTTGATTGACAAGGATGGATGGC-3′), corresponding to nt 1142 to 1165 in the open reading frame of luciferase, and Luc-reverse (5′-CATCGTCGGGAAGACCTGCCACGCCC-3′), corresponding to nt 1339 to 1314 in the open reading frame of luciferase, were employed to quantify the total RNA. Subgenomic RNA was calculated, in each case, as the difference between total RNA and genomic RNA. The amount of exogenous mRNA C+Luc present inside the electroporated cells was determined by employing the oligonucleotides against the luciferase sequence. These primers were designed to amplify sequences of 250 to 300 nt to maximize the efficiency of the reaction. RT-PCR was carried out in 20 μl of LightCycler RNA Master SYBR green I solution containing 3 mM manganese acetate and a 1 μM concentration of each primer. RT was performed at 61°C for 20 min. After that, PCR amplification was initiated with incubation at 95°C for 2 min, followed by 45 cycles of 95°C for 5 s, 58°C for 12 s, and 72°C for 20 s. Data analysis was done using Roche Molecular Biochemicals LightCycler software (version 3.3). The specificity of the amplification reactions was confirmed by analyzing the corresponding melting curves.

Dot blot analysis of mRNA.

Biotinylated C+Luc RNA was obtained by in vitro transcription with the addition of biotin-21-UTP (0.5 mM) to the transcription mixture. For electroporation, 5 × 10⁶ BHK cells infected with SV for 3 h with 100 PFU/cell or uninfected cells were used with 40 μg of in vitro-synthesized RNA. The cells were then seeded in wells of an L-6 plate and harvested after 1 h. Total RNA was extracted from the cultures using an RNeasy kit (QIAGEN) and resuspended in 40 μl RNase-free water. A Bio-Dot SF Microfiltration apparatus was used to fix 5 and 20 μl of each sample on a nitrocellulose filter. The nitrocellulose filter was subsequently washed with 0.2 mM NaCl, 0.1 M Tris-HCl (pH 7.5), 0.05% Triton X-100, and 3% bovine serum albumin and then vacuum dried at 80°C for 1 h. For detection, streptavidin-alkaline phosphatase was conjugated according to an enhanced chemiluminescence method (Amersham).

Measurement of luciferase activity.

Cells were lysed in a buffer containing 0.5% Triton X-100, 25 mM glycylglycine (pH 7.8), and 1 mM dithiothreitol at different times postelectroporation. Luciferase activity was determined using a Monolith 2010 luminometer (Analytical Luminescence Laboratory), as described previously (38).

RESULTS

Low translatability of transfected subgenomic mRNA in SV-infected cells.

The subgenomic 26S mRNA from SV is efficiently translated in infected cells during the late phase of infection, whereas cellular translation is inhibited. Our initial aim was to analyze the translational behavior of the SV subgenomic mRNA under infection or noninfection conditions. To achieve this goal, two different RNAs (Fig. 1A) were synthesized by in vitro transcription from the corresponding plasmids and electroporated into SV-infected or mock-infected BHK cells. One of the RNAs is an SV replicon (rep C+Luc), while the other RNA, representing the subgenomic mRNA C+Luc, is unable to replicate. After electroporation of rep C+Luc, initial translation generates the SV nsPs involved in RNA replication and transcription. The replicative complexes produce minus-strand RNA, which in turn serves as the template for synthesis of genomic (rep C+Luc) and subgenomic mRNA C+Luc by transcription from an internal promoter. Translation of this subgenomic mRNA gives rise to SV capsid (C) and luciferase (Luc) proteins. As expected, host translation was strongly inhibited at 3 h postelectroporation (hpe) of BHK cells with rep C+Luc and only translation of subgenomic mRNA was detected (Fig. 1B). The chimeric protein C+Luc was synthesized and rapidly processed by the autoproteolytic activity of C to release itself and Luc. A progressive increase in Luc activity was apparent over time, consistent with continuous luciferase production (data not shown). This result indicates that the subgenomic mRNA C+Luc generated in BHK cells transfected with the replicon is efficiently translated.

FIG. 1.

Translation of SV subgenomic mRNA in infected BHK cells. (A) Schematic representation of the SV genome, the rep C+Luc replicon, and the subgenomic mRNA C+Luc. (B) About 10⁶ BHK cells were electroporated with 20 μg of rep C+Luc RNA, seeded in three wells of an L-24 plate, and labeled for 30 min with 15 μCi [³⁵S]Met-Cys at the times indicated. Next, SDS-PAGE was carried out, followed by fluorography and autoradiography. Mock cells were electroporated with 50 μl of transcription buffer. (C and D) Mock-infected and SV-infected BHK cells (100 PFU/cell) were electroporated with in vitro-synthesized mRNA C+Luc at different times after infection, as indicated in the text. (C) Half of the cells were seeded in L-24 wells for 1 h and then labeled for 30 min. (D) The remaining cells were recovered in 60 μl luciferase lysis buffer and 20 μl used to estimate Luc activity at 1 hpe, as described in Materials and Methods. (E) Translatability of different Luc-containing mRNAs. Different Luc-containing mRNAs made by in vitro transcription were electroporated into uninfected or SV-infected BHK cells (100 PFU/cell for 3 h). Luc activity was measured at the postelectroporation times indicated. Bars represent standard deviations from two independent experiments. RLU, relative luciferase units; SG.P., subgenomic promoter.

Translation of in vitro-synthesized mRNA C+Luc was then compared in SV-infected and uninfected BHK cells. To this end, this mRNA was electroporated into mock-infected or SV-infected BHK cells at different times postinfection and Luc activity was measured at 1 hpe (Fig. 1D). Protein synthesis was analyzed by radioactive labeling in parallel (Fig. 1C). SV proteins were detected from 2 hpe, indicating that viral translation continued after electroporation. Moreover, SV-induced shutoff of host protein synthesis was produced at 3 hpe. Notably, mRNA C+Luc was efficiently translated when it was transfected into uninfected BHK cells (Fig. 1D, “Mock” lane), while lower Luc activity was obtained after electroporation of SV-infected cells with that mRNA (Fig. 1D, lanes 3 and 4). The inhibition of mRNA C+Luc translation got progressively stronger over time in SV-infected cells.

To test whether the inhibition of mRNA C+Luc translation observed for SV-infected cells occurs with other types of mRNAs, protein synthesis directed by several mRNAs bearing different leader sequences followed by the Luc gene were analyzed. The different leader sequences tested were the SV genomic 49S mRNA, the genuine Luc leader sequence, the EMCV IRES, and the entire poliovirus 5′ untranslated region. SV-infected cells were electroporated with the different mRNAs at 3 h postinfection (hpi). Translation of all of these mRNAs was inhibited by more than 80% in SV-infected cells at 1 and 3 hpe compared to levels with mock-infected BHK cells (Fig. 1E). These results suggest that mRNAs electroporated into SV-infected cells are poorly engaged in the translation machinery, whereas subgenomic mRNAs transcribed from a replicon are translated.

Several possibilities can be put forward to account for the lack of mRNA C+Luc translatability in SV-infected cells. One of them is that this mRNA is poorly transfected into those cells. To examine this possibility, the amount of mRNA transfected into both SV-infected and uninfected BHK cells was analyzed by several methods. First, biotinylated mRNA C+Luc was transfected into cells and the amount present was measured at 1 hpe by dot blot assay (Fig. 2A). It was noteworthy that, at 1 hpe, more mRNA was recovered from SV-infected cells than from uninfected cells. In addition, the amount of transfected mRNA was estimated by quantitative RT-PCR. In this case, uninfected or SV-infected cells were electroporated with in vitro-synthesized mRNA C+Luc and total mRNA was extracted at different hpe to quantitate the remaining amount of mRNA C+Luc. In agreement with the dot blot data (Fig. 2A), the amount of mRNA C+Luc recovered from infected cells was 1.4 times larger than that recovered from uninfected cells after electroporation of mRNA C+Luc (Fig. 2B). These findings support the notion that transfection of exogenous mRNAs is not hampered in SV-infected cells.

FIG. 2.

Quantitation of subgenomic mRNA C+Luc in electroporated cells. (A) BHK cells (2 × 10⁶) were electroporated with 40 μg of biotinylated, in vitro-synthesized mRNA C+Luc. Total RNA was recovered at 1 hpe in 50 μl H₂O and analyzed by dot blot assay with conjugated streptavidin-peroxidase. (Lane descriptions refer to rows a and b, respectively.) Lane 1 (control), 0.1 and 0.025 μg from biotinylated, in vitro-synthesized mRNA C+Luc; lane 2, 20 and 5 μl of RNA extracted from mock-infected BHK cells; lane 3, 20 and 5 μl of RNA extracted from BHK electroporated cells; lane 4, 20 and 5 μl of RNA extracted from SV-infected and electroporated cells. (B) In vitro-generated mRNA C+Luc (20 μg) was electroporated into uninfected or SV-infected BHK cells (100 PFU/cell for 3 h) and seeded in three wells of an L-24 plate, and at 1, 3, or 5 hpe total RNA was extracted from the cultures to be used as the template to quantify the amount of mRNA C+Luc by real-time RT-PCR. On the y axis, the relative amounts of C+Luc obtained are indicated, with 100 corresponding to the amount extracted at 1 hpe from uninfected cells. Bars represent standard deviations from two independent experiments. (C) Translatability of the mRNAs recovered from electroporated cells. mRNA C+Luc (40 μg) was electroporated into uninfected or SV-infected BHK cells (100 PFU/cell for 3 h). Two and one-half hours later, total mRNA was extracted from the cultures and used to electroporate uninfected cells. Luc activity was measured at 90 min after electroporation. Bars represent standard deviations from three independent measures from the same experiment. RLU, relative luciferase units.

Another possible explanation for the low translatability of transfected mRNAs in SV-infected cells is that they are quickly degraded or inactivated. However, extensive mRNA degradation did not occur, since transfected mRNAs were equally stable in both mock and SV-infected cells, as determined by RT-PCR (Fig. 2B). To analyze whether the mRNAs electroporated into SV-infected cells were inactivated for translation, total mRNA from infected and uninfected cells electroporated with mRNA C+Luc was isolated at 2.5 hpe. This mRNA extracted from both types of electroporated cells was employed to further electroporate uninfected cells. The mRNA isolated from SV-infected cells produced 27% more Luc activity than mRNA extracted from uninfected cells at 90 min after transfection (Fig. 2C), in accordance with the larger amount of mRNA C+Luc detected by RT-PCR in infected cells. Therefore, the mRNA electroporated into SV-infected cells is neither degraded nor irreversibly inactivated for translation.

Alternatively, the nontranslatability of electroporated mRNAs in SV-infected cells could be due to differences between in vitro-synthesized mRNAs and those produced in infected cells. For instance, it is possible that in vitro-synthesized mRNAs possess a cap structure different from that of mRNAs transcribed in the infected cells. To analyze this possibility, total mRNA from cells electroporated with the replicon rep C+Luc was extracted at 7 hpe. Translation of this genuine mRNA was analyzed by the detection of Luc activity after electroporation of uninfected or infected cells. The RNA electroporated includes, apart from cellular mRNA, viral subgenomic mRNA C+Luc and rep C+Luc. This rep C+Luc can replicate and give rise to mRNA C+Luc upon transcription. To avoid Luc activity from newly synthesized mRNA C+Luc, Luc activity was measured at early times after electroporation (Fig. 3). As occurred with in vitro-synthesized C+Luc, Luc activity from genuine mRNA extracted from cells was strongly inhibited in SV-infected cells. As a control, Luc activity in cells electroporated with in vitro-generated rep C+Luc was determined. Background levels of Luc activity were obtained in uninfected BHK cells electroporated only with rep C+Luc until 2 hpe, suggesting that this is the time necessary to generate subgenomic mRNAs. SV-infected cells transfected with rep C+Luc synthesize Luc earlier than uninfected cells but progress more slowly. In this case, the replicative complexes are already formed as a result of SV infection before rep C+Luc electroporation but this RNA should compete for minus- and plus-strand RNA synthesis with SV RNAs.

FIG. 3.

Translatability of genuine mRNA C+Luc extracted from cells. BHK cells (2 × 10⁶) were electroporated with 40 μg of rep C+Luc, and at 7 hpe total mRNA was extracted to be used for electroporation of uninfected or SV-infected cells (100 PFU/cell for 3 h). As a control, 20 μg of in vitro-synthesized rep C+Luc was also electroporated. At the times indicated (mpe, minutes postelectroporation), Luc activity was measured. Bars represent standard deviations from two independent experiments. RLU, relative luciferase units.

Transfected replicons are competent for replication in SV-infected cells.

Taken together, the above data indicate that exogenous mRNAs are excluded from the protein-synthesizing machinery in SV-infected cells. However, replicons transfected into those cells could be used as templates for replication and transcription. This idea is reinforced by the use of ΔnsPs rep C+Luc (Fig. 4A), which lacks most of the coding sequence for SV nsPs. This SV-derived RNA requires the viral nsPs provided in trans by SV-infected cells for its replication and transcription. ΔnsPs rep C+Luc RNA was electroporated into mock- or SV-infected BHK cells and Luc activity measured at 1, 2, 3, and 5 hpe (Fig. 4B). Obviously, Luc activity was not observed in uninfected cells, since they lack SV nsPs. However, increasing Luc activity was detected throughout the time course upon transfection of ΔnsPs rep C+Luc in SV-infected cells. These data indicate that the RNA transfected into SV-infected cells is not destroyed and can be used as a template for replication and transcription. Moreover, the subgenomic mRNA C+Luc synthesized can be translated in SV-infected cells if it is transcribed in these cells. As a control, rep C+Luc, which can replicate autonomously, was also electroporated. Notably, Luc activity obtained from this replicon in infected cells was similar to that obtained from ΔnsPs rep C+Luc (Fig. 4B). Probably, synthesis of nsPs was inhibited from rep C+Luc in SV-infected cells, as occurred with the other exogenous mRNAs. However, rep C+Luc can serve as a template for minus-strand and next for plus-strand RNA synthesis in SV-infected cells in the same way as ΔnsPs rep C+Luc. These results indicate that electroporated RNAs are not sequestered and localize in places in SV-infected cells where replication, transcription, and translation can take place.

FIG. 4.

Electroporated mRNA acts as a template for transcription. (A) Schematic representation of ΔnsPs rep C+Luc. (B) Δnsps rep C+Luc or rep C+Luc (20 μg) was electroporated into uninfected or SV-infected BHK cells (100 PFU/cell for 3 h). Luc activity was measured at the times indicated. Bars represent standard deviations from two independent experiments. RLU, relative luciferase units.

Correlation with inhibition of host translation.

The experiments described above indicated that the inhibition of translation for exogenous mRNAs and that for host mRNAs are correlated in SV-infected cells (Fig. 1C and D). An SV variant unable to block host protein synthesis has been described previously; this viable SV variant contains a point mutation in nsP2 and is able to replicate and produce viral proteins in BHK cells without interfering with cellular translation (11, 16). This SV variant was constructed by introducing a mutation, P₇₂₆G, in the sequence for nsP2 of our SV clone. BHK cells were electroporated with RNA from control SV or from this variant, and protein synthesis was analyzed at 7 hpe. As expected, this variant did not inhibit cellular protein synthesis, whereas the capsid protein could be detected both by radioactive labeling and by Western blotting (Fig. 5). Cells were then electroporated a second time with in vitro-synthesized mRNA C+Luc, and Luc activity was measured 90 min later. In agreement with the experiments described above, Luc activity was inhibited by more than 90% in cells electroporated with SV RNA compared to levels for uninfected BHK cells, whereas no inhibition was detected for the SV(P₇₂₆G) variant. Therefore, the blockade of cellular protein synthesis is associated with the inhibition of translation of exogenous mRNAs in SV-infected cells.

FIG. 5.

Translatability of exogenous mRNA in BHK cells infected with an SV variant defective in the shutoff of host translation. About 2 × 10⁶ BHK cells were electroporated with transcription buffer or with 40 μg of SV or SV(P₇₂₆G) in vitro-generated mRNAs. Half of the electroporated cells were labeled with [³⁵S]Met-Cys for 30 min at 7 hpe and analyzed by SDS-PAGE (top) or by Western blotting using anticapsid antibodies (middle). The remaining cultures were electroporated with mRNA C+Luc, and Luc activity was measured at 90 min after electroporation. Relative values of Luc activity are shown (bottom), with 100 being the value obtained with BHK cells previously electroporated with transcription buffer.

Lack of competition between viral transcripts and transfected mRNAs.

Our next aim was to determine whether viral mRNAs derived from transcription and the exogenous mRNAs compete for the translation machinery. Two replicons were employed for this experiment: rep C, which produces large amounts of C protein, and a replicon with a deletion of the subgenomic promoter after the stop codon for the nsP4 protein (rep −26S) (Fig. 6A). This second replicon does not synthesize any sPs. Both replicons were equally efficient at shutting off cellular protein synthesis (Fig. 6B). BHK cells were electroporated with equivalent amounts of rep C, rep −26S, or transcription buffer as a control, and a second electroporation with mRNA C+Luc was carried out 4 h later. Luc activity and protein synthesis were measured 1 or 3 h after the second electroporation. Large amounts of C protein were indeed synthesized from rep C, whereas no viral proteins were detected from rep −26S (Fig. 6B). Notably, Luc activity in both types of replicon-expressing cells was strongly inhibited (∼90%) compared with activity in control cells (Fig. 6C). These findings indicate that viral transcription from the subgenomic promoter is not necessary to block the translation of exogenous mRNAs. Therefore, inhibition of the translation of transfected mRNAs in SV-infected cells does not seem to be a consequence of competition with the subgenomic mRNAs produced by the viral replication machinery.

FIG. 6.

Competition between endogenous viral transcripts and transfected mRNAs. (A) Schematic representation of the different replicons. (B and C) BHK cells (2 × 10⁶) were electroporated with transcription buffer or with 40 μg of in vitro-synthesized replicons. Four hours later, a second electroporation was carried out with 40 μg of in vitro-synthesized mRNA C+Luc. (B) Half of the cells were labeled with [³⁵S]Met-Cys for 30 min 2 or 4 h after the second electroporation and analyzed by SDS-PAGE. (C) The remaining cells were used to measure Luc activity. Bars represent standard deviations from two independent experiments. RLU, relative luciferase units.

To determine the translational efficiencies of subgenomic mRNA C+Luc in the different situations assayed above, BHK cells were electroporated with mRNA C+Luc or rep C+Luc and SV-infected cells with mRNA C+Luc, and the amount of Luc RNA was quantified by real-time RT-PCR and compared with the Luc activity obtained in each case. Remarkably, the highest translational efficiency corresponded to exogenous mRNA C+Luc electroporated into uninfected cells (Table 1). Although greater Luc activity was obtained from rep C+Luc-electroporated cells, the ratio of luciferase activity to the amount of each mRNA was lower than the ratio for translation of exogenous mRNA C+Luc in uninfected cells. mRNA C+Luc transfected into infected cells clearly exhibited the lowest translation ratio (about 60 times lower than in uninfected BHK cells).

TABLE 1.

Translatability in mock-infected or SV-infected BHK cells^a

RNA/BHK cells	Relative amt of mRNA (% [SD])	Luc activity (arbitrary units [SD])	Ratio of Luc activity/mRNA
C+Luc/mock infected	1.0 (±11)	74,124 (±16%)	74,124
C+Luc/SV infectedb	1.7 (±12)	2,245 (±10%)	1,290
rep C+Luc/mock infected	114.1 (±6)	4.4 × 106 (±13%)	39,454

^aBHK cells were electroporated with C+Luc or rep C+Luc RNA and SV-infected cells with C+Luc RNA. Luciferase activity and Luc RNA were measured and extracted at 1 hpe in C+Luc-electroporated cells. The same procedure was followed at 4 hpe for rep C+Luc-electroporated cells.

^bSV-infected cells were used at 3 hpi.

Translation of several mRNAs transfected into BHK cells infected with different animal viruses.

Finally, the impairment of translation of transfected mRNAs was investigated in BHK cells infected with different animal viruses. For this experiment, different mRNAs (mRNA C+Luc, L.luc-Luc, and L.EMC-Luc) were electroporated into uninfected or BHK cells infected with VSV, EMCV, or SV. Protein synthesis was determined at 4 hpi. The cells were then electroporated, and Luc activity was examined at 1 and 3 hpe (Fig. 7A and B). Compared with results for uninfected cells, an extensive blockade of Luc synthesis was observed for cells infected with the three different RNA viruses assayed. Notably, mRNA L.EMC-Luc was strongly inhibited in EMCV-infected cells, despite the fact that this mRNA contains the EMCV IRES. In conclusion, these data suggest that the suppression of exogenous viral mRNA translation occurs in cells infected with other RNA viruses, such as VSV and EMCV, which belong to families other than Togaviridae.

FIG. 7.

Transfection of several Luc mRNAs in BHK cells infected with different viruses. Cells (2 × 10⁶) mock infected or infected with SV (50 PFU/cell), VSV (40 PFU/cell), or EMCV (20 PFU/cell) for 4 h were electroporated with 40 μg of the indicated in vitro-synthesized mRNAs. (A) Half of the cells were labeled with [³⁵S]Met-Cys at 4 hpi and the proteins analyzed by SDS-PAGE. (B) The remaining cultures were electroporated and luciferase activity measured at 1 and 3 hpe. Bars represent standard deviations from two independent experiments. RLU, relative luciferase units.

DISCUSSION

SV-infected cells provide a paradigmatic model to analyze the regulation of translation in mammalian cells. At 3 to 4 hpi, these cells efficiently translate subgenomic viral mRNA, whereas cellular protein synthesis is abrogated. This could be explained by the induction of cellular modifications after viral infection, such that the viral 26S mRNA can be translated preferentially because viral mRNAs have evolved structures that optimize their translatability under these conditions. Particular structures present in viral mRNAs often promote interaction with the protein-synthesizing machinery, and so these mRNAs have a low requirement for some initiation factors (12, 18, 19). Hence, some of the initiation factors are inactivated in a number of viral infections (22, 26). For the subgenomic mRNA from alphaviruses, a low requirement for eIF4E (the cap binding protein), eIF4B (37), eIF4G (6), and eIF2α (17, 39) has been reported. In fact, eIF2α becomes phosphorylated in SV-infected cells.

According to our present findings, transfected subgenomic SV mRNA is poorly translated in infected cells. Therefore, this mRNA must meet two criteria in order to direct protein synthesis. First, it must have particular sequences that confer translation efficiency (9, 10, 35; our unpublished results). Second, this mRNA should be synthesized by the viral transcription machinery. The first 275 nt of the C sequence in the SV subgenomic mRNA form a tertiary structure that acts as a translation-enhancing element in infected cells (9, 10). However, the presence of special motifs or structures in the subgenomic SV mRNA does not in itself guarantee translatability when a cell is infected.

The finding that SV translation is coupled to viral transcription provides new insight into the inhibition of translation in the infected cells. Previous studies indicated that SV subgenomic mRNA had a sequence that enhanced translation in infected cells but exhibited the opposite effect in uninfected ones (9, 10, 35), suggesting that the cellular modifications arising from viral infection enhanced its translation. In contrast, our results indicate that the SV subgenomic mRNA is translated efficiently in uninfected cells. The results reported by Frolov and Schlesinger (10) on the translation-enhancing element present in SV subgenomic mRNA suggested that this element was detrimental in uninfected cells. However, these authors assayed only chimeric mRNAs encoding the truncated but not the entire C sequence with lacZ. They also employed an expression system that synthesized mRNAs in the nuclei of uninfected cells. A similar work with Semliki Forest virus concluded that the subgenomic mRNA had a translation-enhancing region that was functional only in virus-infected cells (35). In that study, all recombinant subgenomic mRNAs tested, including the entire C sequence fused to lacZ, were able to promote comparable levels of protein synthesis. Our present findings indicate that the SV subgenomic mRNA is translated at least twice as efficiently in uninfected BHK cells as in cells transfected with the corresponding replicon. Moreover, the transfected subgenomic mRNA is translated about 60 times more effectively in uninfected BHK cells than in SV-infected cells. This feature cannot be attributed to deficient electroporation of the infected cells because the amount of exogenous RNA present in these cells was larger than that present in uninfected cells. A general inhibition of translation due to the electroporation process did not occur, since active viral protein synthesis was evidenced by radioactive labeling. The exclusion of transfected mRNAs cannot be ascribed to competition by the viral mRNAs produced in the infected cells. This was illustrated by the fact that an SV replicon that does not synthesize subgenomic mRNA still impairs translation of transfected viral mRNA. Although the subgenomic mRNA has a lower translatability in an infected environment, its particular structure together with the large amounts present in infected cells produced the viral sPs in abundance sufficient to form new virus particles.

The finding that an SV variant defective to abolish host translation is also unable to inhibit translation of electroporated mRNAs suggests that similar mechanisms can operate in both circumstances. Perhaps viral replication recruits some components of the translation machinery to couple them to translation. The P₇₂₆G variant has a lower replicative capacity, which might imply less capacity to recruit translation components. Alternatively, the mutated nsP2 could be involved in the interaction with a putative translation component. Further studies to localize electroporated mRNAs and the translation components in infected cells could shed more light on the translation inhibition mechanism.

Other functions in animal viruses are certainly coupled. This is the case for picornaviruses (13) or ambisense viruses (28), where viral transcription necessitates continuous viral protein synthesis. In addition, coupling between synthesis of viral genomes and encapsidation has been described to occur during the morphogenesis of poliovirus (29), flavivirus (20), and bromovirus (3). Cytoplasmic replication of animal viruses could rely on coupling between transcription and translation, as occurs with bacterial mRNAs (24). Findings reported for SV may also be applicable to cells infected with other animal viruses, since IRES-containing mRNAs are also excluded from translation in cells infected with a picornavirus (EMCV) or a rhabdovirus (VSV). During the preparation of the manuscript, two studies describing how preferential translation of different viral mRNAs was dependent on viral transcription were published (25, 40). In one of these studies, it was found that coupling between translation and transcription of RNA2 may occur in cells infected with red clover necrotic mosaic virus (25). In this case, the uncapped viral mRNA is not translated in uninfected cells and directs protein synthesis only when RNA replication occurs. In the other study, it was discovered that preferential translation of VSV mRNAs is also conferred by transcription from the viral genome (40). At present, little about the molecular mechanism responsible for coupling between viral translation and transcription is known. Montgomery et al. reported that the nsp2 proteins from different alphaviruses interact with ribosomal protein S6 (27). According to the authors, the nsp2 protein, besides participating in replication, may also alter the ribosome contributing to the differential translation of host and viral mRNAs. Future work will determine whether coupling between translation and transcription is a common feature in different families of viruses and what molecular mechanisms are implicated in this coupling.