Saumitra Das¹, Michael Ott², Akemi Yamane¹, Arun Venkatesan¹, Sanjeev Gupta² and Asim Dasgupta¹

Department of Microbiology, Molecular Genetics and Immunolog, UCLA School of Medicine, 10833 Le Conte Avenue, Los Angeles, CA 90095-1747¹ , ²Department of Medicine, Albert Einstein, College of Medicine of Yeshiva University, Bronx, New York 10461-1602

Received 9/18/98 Accepted 9/23/98

3. TRANSLATION OF VIRAL mRNA IN YEAST

3.1. Poliovirus RNA is not translated in yeast

The discovery that the yeast Saccharomyces cerevisiae contains a potent inhibitor of IRES translation occurred through scientific serendipity. While trying to express the infectious poliovirus cDNA in the yeast, it was observed that the yeast Saccharomyces cerevisiae was incapable of translating poliovirus RNA (12). A poliovirus cDNA construct, pBM POLIO (cloned into pBM258 vector, which contained the sequences for replication in both E. coli and yeast) under the control of the GAL10 promoter, and a control plasmid (pBM258) were transformed into S. cerevisiae. Transformed cells were grown in galactose to induce expression of poliovirus RNA from the cDNA (from GAL10 promoter) by the yeast RNA polymerase. Although full-length viral RNA was synthesized in cells transformed with pBMPOLIO, no viral protein could be detected in the cell.

The inability of yeast cells to translate poliovirus and P2CAT RNA in vivo was further recapitulated in vitro and appeared to be due to one or more translational inhibitors present in yeast cells. In fact, addition of small amounts of yeast cell lysates to the HeLa cell translation reaction prevented translation of RNAs containing the PV5´ -UTR (p2CAT RNA) whereas cap-dependent translation of CAT RNA (devoid of the PV 5´ -UTR sequences) remained unaffected. These observations suggested that the translational inhibitory effects are mediated through PV 5´ -UTR and the inhibitor was capable of acting in trans. Furthermore, addition of excess HeLa lysate was capable of rescuing the inhibition whereas addition of increasing amounts of template RNA to the translation reaction did not. Thus, the inhibitor appeared to interact with a component of the translational machinery rather than directly bind to the input viral RNA as an antisense RNA. In an attempt to purify the transacting factor (the inhibitor), the yeast cell lysates were fractionated on a DEAE sephacel column and the inhibitory activity was eluted with 1M KCl containing buffer. The partially purified inhibitor was further characterized and shown to be heat stable and resistant to phenol extraction, proteinase K digestion and DNase treatment, but sensitive to RNase digestion. These results suggested that the inhibitor is most probably an RNA molecule (12).

A later publication from this laboratory reported further purification of the DEAE sephacel purified fraction of the yeast inhibitor (15). Total RNA obtained from the 1M KCl fraction of the DEAE sephacel was treated with DNase and proteinase K followed by phenol extraction and alcohol precipitation. The purified total RNA was end labeled and resolved by 20% PAGE / 8M urea gel electrophoresis. Each RNA band was eluted from gel slices and tested for the inhibitory activity against PV-IRES driven translation. A single band associated with the inhibitory activity was identified and sequenced. On the basis of the RNA sequence of the small RNA molecule (60nt.) a synthetic clone was prepared (figure 1). RNA derived from the synthetic clone (pSDIR) by in vitro transcription (I-RNA) was shown to block the translation of poliovirus RNA both in vitro and in vivo (15). When HeLa monolayer cells were cotransfected with PV RNA and I-RNA, poliovirus specific proteins were not detected by immunoprecepitation in transfected cells. Also, host protein synthesis was restored in cells cotransfected with PVRNA and I-RNA compared to complete shut off seen in cells transfected with PV RNA alone. Furthermore, the in vivo effect of I-RNA was reversed by addition of equimolar amounts of antisense I-RNA in the cotransfection experiment (15). These findings strengthened the idea that the purified yeast inhibitor RNA specifically blocked viral RNA translation without adversely affecting host cell protein synthesis.

Figure 1. Sequence, cloning and deletion analysis of IRNA. The nucleotide sequence (A), cloning (B), and in vitro transcription of IRNA from the clone pSDIR (C) are shown. Panel D shows IRNA deletion mutant constructs. The nucleotide positions are indicated for each mutant. The numbers in parentheses indicate percentage of translation inhibition by mutant IRNAs compared with wild type IRNA (100%) , calculated by averaging results from three independent experiments. Reprinted with permission from J. Virology, Vol. 68, p.7200-7211, 1994, and Vol. 10, p. 1624-1632, 1996, ASM.

3.2. I-RNA specifically inhibits internal initiation of translation

To examine if the cloned and purified yeast inhibitor RNA (I-RNA) preferentially inhibits internal initiation of translation, its effect on translation from a bicistronic messenger was determined. For this purpose, bicistronic constructs containing CAT and luciferase (Luc) genes flanked by the IRES sequences of different picornaviruses were used. Translation of Luc is initiated internally and that of CAT is initiated in a cap- dependent manner. In the presence of I-RNA, significant inhibition of Luciferase synthesis was observed, whereas synthesis of CAT was almost completely unaffected in each case (13, 15, and Das & Dasgupta, unpublished observation). Monocistronic constructs having IRES sequences upstream of a reporter gene were tested for inhibition by I-RNA. As expected, the 5´ -UTR of poliovirus (p2CAT)- and the Bip mRNA(Bip)-mediated translation of the reporter genes was significantly inhibited by the addition of I-RNA. In contrast, addition of equivalent amounts of I-RNA did not inhibit cap dependent translation of CAT, Luciferase or the yeast a36 mRNA. Interestingly, EMCV-IRES mediated translation (pCITE) of the reporter gene was not inhibited by I-RNA (12).

3.3. Minimum sequences required for I-RNA activity

To determine I-RNA sequences required for inhibition of poliovirus IRES-mediated translation, a nested set of 15 nt long deletions were generated. The effect of these truncated RNAs on in vitro translation programmed by P2 CAT RNA containing poliovirus 5´ UTR was determined (13). The deletion analysis suggested that the minimum sequence required to inhibit PV IRES-mediated translation resides between nucleotides 30-45 (figure 1). This notion was supported by two observations. First, a deletion mutant (I-3 RNA) which contained the entire I-RNA sequence except nucleotides 31-45, was totally inactive in inhibiting viral IRES-mediated translation. Secondly, a truncated I-RNA (nt 30-45, I-9 RNA) retained considerable amount of translation-inhibitory activity (figure 1). However, a 25 nt. long truncated RNA (I-7 RNA) containing the I-9 RNA sequence appeared to be more active particularly in vivo (see figure 1). The shorter I-9 RNA was only 50% as active as I-RNA in vivo. The structure(s) of I-RNA or its truncated derivatives may be important in IRES-mediated translation-inhibition. The fact that addition of an extra ten nucleotides to the 3´-end of I-7 RNA (nt. 26-50) significantly reduces its (1-6 RNA, nt. 26-60) translation-inhibitory activity may be indicative of alteration of the structure of this RNA . Similarly, addition of another 5 nucleotides to the 5´ -end of I-6 RNA drastically reduces its (I-5, nt. 20-60) ability to inhibit translation (figure 1 ).

3.4. Inhibition of HCV IRES-mediated translation by IRNA in vivo and in vitro

To test the possibility that IRNA might interfere with the HCV IRES-mediated translation, human hepatocellular carcinoma cells (Huh-7) were transiently cotransfected with 3 plasmids: a reporter gene expressing luciferase programmed by the HCV-IRES element (pCD HCV-luc), SV40/b-gal plasmid to measure transfection efficiency, and the plasmid expressing IRNA (pCDIR.Ribo. D T7). All transfections were performed in triplicates and contained equal amounts of the luciferase reporter and b -gal plasmids. Increasing concentrations of the pCDIR.Ribo. D T7 plasmid were used in various reactions and the total amount of DNA in each reaction was kept constant by addition of an appropriate amount of a non-specific DNA (pCDNA3). Following transfection, luciferase activity was measured in cell-free extracts. At the lowest concentration of the IRNA plasmid, inhibition of luciferase activity from the pCD HCV-luc plasmid was approximately 50% compared to the control (14). However, at the highest concentration, 90% of luciferase activity was inhibited. Translation of luciferase from a control plasmid (pCDNA3-luc) without having the HCV-IRES was not significantly inhibited by IRNA. Expression of either ribozyme alone or a nonspecific RNA having similar length as IRNA, did not interfere with luciferase expression (14). These results suggested that HCV-IRES-mediated translation was specifically inhibited by IRNA in hepatoma cells, whereas cap-dependent translation of luciferase from the control plasmid lacking the HCV IRES element was not significantly affected by IRNA (14).

To confirm the results obtained in vivo, the effect of IRNA on HCV-IRES-mediated translation was determined in vitro. A bicistronic construct consisting of the HCV IRES flanked by CAT and luciferase genes was used in this experiment. While synthesis of CAT is mediated by cap-dependent translation, the downstream luciferase synthesis occurs by HCV IRES-mediated translation. Translation was measured by quantitating radioactivity incorporated into luciferase (Luc) and CAT polypeptides at 0, 0.5, 1, 2 and 4 m g of IRNA (figure 2). The specific inhibition of luciferase synthesis was normalized by determining the ratio of luciferase to CAT at each IRNA concentration. At one microgram IRNA, luciferase synthesis was inhibited to 20% of the control, whereas at 2 m g IRNA, specific inhibition of HCV IRES-mediated luciferase synthesis was 76% compared to the control. Although both luciferase and CAT synthesis were inhibited by IRNA in vitro, luciferase synthesis was affected much more than CAT at higher concentrations of IRNA. The inhibition of cap-dependent translation by IRNA could be due to its interaction with general RNA-binding proteins which have been implicated in facilitating cap-dependent translation. In addition, IRNA’s interaction with La may affect AUG start site selection during translation initiation, as suggested by recent studies from different laboratories (40, 61).

Figure 2. IRNA selectively inhibits HCV IRES-mediated translation in vitro. (A) a bicistronic construct containing the HCV IRES flanked by CAT and Luciferase (LUC) genes was transcribed by T7 RNA polymerase, and the bicistronic mRNA was translated in vitro in HeLa lysate in the absence (lane 1) and presence of 0.5, 1, 2 and 4 m g of IRNA (lanes 2, 3, 4 and 5, respectively). (B) The results in Figure 2A were quantitated and the ratios of LUC to CAT are presented as percentage of LUC/CAT translation against various concentration of IRNA. Reprinted with permission from J. Virology, Vol. 72, p. 5638-5647, 1998, ASM.

3.5. Constitutive expression of I-RNA in eukaryotic cells

To determine the long term effect of expression of IRNA in hepatoma cells, cell lines constitutively expressing IRNA were generated using a pCDNA based vector (14). The cell line (pCDIR) made initially contained the T7 sequences at the 5´ -end of the IRNA gene. Another cell line was made in which the hepatitis delta ribozyme sequence was added at the 3´-end of IRNA sequence for generation of the exact 3´-end (pCDIR-Ribo). A third cell line was prepared using the pCDIR-Ribo construct lacking the T7 promoter sequences (pCDIR-Ribo-D T7) (14). The control cells and cell lines expressing IRNA were cotransfected with HCV-IRES-Luc and b -gal DNA. Cell-free extracts were used to measure both luciferase and b -gal activities. Luciferase expression was plotted after normalizing with respect to b -gal activity. Approximately 60% inhibition of HCV-IRES-mediated translation was observed in the pCDIR cells compared to the control. The cells expressing IRNA with hepatitis delta antigen showed 65% inhibition of HCV-IRES-mediated translation. In cell lines expressing IRNA-Ribo without the T7 promoter sequence, almost 81% inhibition of luciferase expression was observed. These results clearly show that IRNA interferes with expression of luciferase programmed by HCV IRES. A titration of the reporter construct (HCV-IRES-Luc) in the cell line pCDIR-Ribo- T7 consistently showed 80-85% inhibition of HCV IRES-mediated translation of luciferase (figure 3). No significant inhibition of cap-dependent translation from the pCDNA-Luc construct was observed with cell lines expressing IRNA (14). That constitutive expression of IRNA in hepatoma cells is not detrimental to these cells is supported by continued viability of these cell lines for the last 6-8 months. Moreover, overall cellular transcription and translation were not significantly altered in the cell lines compared to the control cells (14).

Figure 3. HCV-luciferase reporter dose response and quantitation of IRNA and Luciferase mRNA in the hepatoma cell line expressing IRNA constitutively. (A) one, 2 and 3 m g of the pCD HCV-luc reporter plasmid were transfected into Huh-7 hepatoma cells (dotted bar) or the IRNA-expressing hepatoma cell line (pCDIR.Ribo.D T7) (white bars). A control b -gal plasmid was also cotransfected to normalize transfection efficiency. The ratio of luciferase (x10³ light units [LU]) to b -gal activities were plotted against various concentrations of the reporter plasmid. (B) IRNA expression level in the cell line was detected by RT-PCR; (lane 1) no IRNA; (lanes 2-4), 1, 2.5 and 5 ng of purified IRNA. (Lanes 5 to 7), 2, 1.5 and 1 m g of total RNA from the cell line expressing IRNA. (Lane 8) contained 2 m g of total RNA from control hepatoma cells. (C) Luciferase mRNA in control (lanes 4) versus IRNA-expressing cells (lane 5)) was determine by RT-PCR. Lanes 1-3 show no RNA, and 1 and 10 ng of luciferase mRNA control, respectively. Reprinted with permission from J. Virology, Vol. 72, p. 5638-5647, 1998, ASM.

3.6. Cells expressing I-RNA are refractory to PV and PV/HCV chimeric virus infection

To determine the effect of IRNA on HCV IRES-mediated translation during virus infection, the pCDIR.Ribo.D T7 cells were infected with a chimeric poliovirus (PV/HCV 701) in which PV IRES is replaced by the HCV IRES. PV/HCV 701 contained the 5´ -cloverleaf structure of PV, followed by HCV IRES (nt. 9-332) plus 123 amino acids of HCV core protein followed by the entire poliovirus ORF plus the 3´-UTR and poly (A) (35). Translation of viral proteins in cells infected with PV/HCV chimeric virus is mediated by the HCV IRES element. Huh-7 control cells and the hepatoma-IRNA cells (pCDIR.Ribo.D T7) were infected with polio and PV/HCV chimeric viruses. Following infection, cell free extracts were prepared from infected and mock-infected cells which were then used to further infect HeLa monolayer cells. Plaques characteristic of wt PV (panel A) and PV/HCV 701 (panel B) were apparent in HeLa cells infected with cell-free extract from control hepatoma cells (figure 4). Evidently viral replication was drastically affected in the cell line expressing IRNA (Hepatoma-IRNA) with either virus (figure 4). In a parallel experiment, the virus titers were measured using the serial dilution method and the results demonstrated more than 100 fold decrease in virus yield in IRNA-expressing hepatoma cells compared to the control cells. While the control hepatoma cells showed extensive damage after infection, the cells expressing IRNA were significantly protected from the cytopathic effect of the chimeric virus (figure 4). Thus, hepatoma cells constitutively expressing IRNA were significantly resistant to both PV and PV/HCV chimeric virus under the conditions used for infection (14).

Figure 4. Hepatoma cells constitutively expressing IRNA prevent PV and HCV-PV chimera infection. Huh-7 control cells or the IRNA-expressing hepatoma cell line (pCDIR-Ribo. D T7) were infected with either poliovirus (PV) or HCV-PV chimera (B and C). After 72 h, cells were stained for the observation of cytopathic effect (C) or cell extracts were made to further infect HeLa monolayer cells for plaque assay (A and B). (D) Average virus titers obtained from 3 independent plaque assays are shown. Reprinted with permission from J. Virology, Vol. 72, p.5638-5647, 1998, ASM.

To rule out the possibility that the cloned hepatoma cell line is simply less permissive to support viral infectious life cycle, its ability to support replication of another chimeric poliovirus was examined. For this purpose a chimeric PV [PV1 (ENPO)] containing, the EMCV IRES was used (14). We had previously shown that EMCV IRES-mediated in vitro translation was not inhibited by IRNA (12). Both the PV/HCV and PV/EMCV chimeric viruses were used to infect the Huh-7 control, hepatoma-IRNA and hepatoma cells expressing only the ribozyme. As can be seen in table 1, the PV/HCV chimeric virus titer was reduced approximately 100 fold in hepatoma-IRNA cells compared to Huh-7 control cells. In contrast, the PV/EMCV virus titer was not significantly reduced in hepatoma-IRNA cells compared to Huh-7 cells. This is consistent with our previous finding that EMCV IRES-mediated translation is not inhibited by IRNA in vitro (12). Also, the hepatoma-ribozyme cell line was as active in supporting PV/HCV (or PV/EMCV) replication as the control Huh-7 cells (14). These results suggest that the cloned hepatoma cell line expressing IRNA is not simply less permissive to support virus replication in general . These results also suggest that viral RNA synthesis or protein processing is not inhibited by IRNA.