Title: The molecular basis of viral oncolysis: usurpation of the Ras signaling pathway by reovirus
Abstract: Article15 June 1998free access The molecular basis of viral oncolysis: usurpation of the Ras signaling pathway by reovirus James E. Strong James E. Strong Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Matthew C. Coffey Matthew C. Coffey Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Damu Tang Damu Tang Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Pauline Sabinin Pauline Sabinin Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Patrick W.K. Lee Corresponding Author Patrick W.K. Lee Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author James E. Strong James E. Strong Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Matthew C. Coffey Matthew C. Coffey Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Damu Tang Damu Tang Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Pauline Sabinin Pauline Sabinin Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Patrick W.K. Lee Corresponding Author Patrick W.K. Lee Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 Search for more papers by this author Author Information James E. Strong1, Matthew C. Coffey1, Damu Tang1, Pauline Sabinin1 and Patrick W.K. Lee 1 1Department of Microbiology and Infectious Diseases, University of Calgary Health Sciences Centre, Calgary, Alberta, Canada, T2N 4N1 *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:3351-3362https://doi.org/10.1093/emboj/17.12.3351 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info NIH-3T3 cells, which are resistant to reovirus infection, became susceptible when transformed with activated Sos or Ras. Restriction of reovirus proliferation in untransformed NIH-3T3 cells was not at the level of viral gene transcription, but rather at the level of viral protein synthesis. An analysis of cell lysates revealed that a 65 kDa protein was phosphorylated in untransformed NIH-3T3 cells, but only after infection with reovirus. This protein was not phosphorylated in infected or uninfected transformed cells. The 65 kDa protein was determined to be the double-stranded RNA-activated protein kinase (PKR), whose phosphorylation leads to translation inhibition. Inhibition of PKR phosphorylation by 2-aminopurine, or deletion of the Pkr gene, led to drastic enhancement of reovirus protein synthesis in untransformed cells. The emerging picture is one in which early viral transcripts trigger PKR phosphorylation in untransformed cells, which in turn leads to inhibition of translation of viral genes; this phosphorylation event is blocked by an element(s) in the Ras pathway in the transformed cells, allowing viral protein synthesis to ensue. The usurpation of the Ras signaling pathway therefore constitutes the basis of reovirus oncolysis. Introduction Although the presence or absence of virus receptors on the cell surface remains a major determining factor of the susceptibility of a cell to virus infection, there is now increasing evidence that the intracellular environment plays an important role in dictating the outcome of viral invasion. In the case of the human reovirus, the receptor is the ubiquitous sialic acid (Gentsch and Pacitti, 1985, 1987; Paul et al., 1989; Choi et al., 1990), a fact accounting for the observation that reovirus binds to most mammalian cells. However, neither virus binding nor even internalization assures a productive outcome, suggesting that downstream events are required for reovirus infection. An interesting clue has come from earlier studies which showed that normal and transformed cells manifested differential sensitivity to reovirus infection. Hashiro et al. (1977) reported that certain virally and spontaneously transformed cell lines of murine origin were susceptible to reovirus infection, whereas normal human and subhuman primate cells, primary mouse cells, normal rat kidney cells and baby hamster kidney cells were not. Duncan et al. (1978) found that normal and SV40-transformed WI-38 cells exhibited different sensitivities to reovirus infection, with cytopathology observed only in the transformed cells and not in normal cells, which nonetheless produced virus for a sustained period. Collectively, these observations suggest that reovirus infection efficiency is somehow linked to the transformed state of the cell. However, the molecular basis of this correlation remains obscure. We recently reported that two mouse cell lines (NR6 and B82) expressing no epidermal growth factor receptors (EGFRs) were relatively resistant to reovirus infection, whereas the same cell lines transfected with the gene encoding EGFR manifested significantly higher susceptibility (Strong et al., 1993). This enhancement of infection efficiency requires a functional EGFR, since it was not observed in cells expressing a mutated (kinase-inactive) EGFR. Thus, the reovirus infection process is closely coupled to the EGFR-mediated cell signal transduction pathway. Furthermore, we found that reovirus is capable of binding directly to the N-terminal extracellular domain of EGFRs (Tang et al., 1993). Taken together, these observations suggest two alternative explanations for the augmentation of reovirus infection by functional EGFRs. The first possibility is that reovirus plays an active role by first binding to EGFRs, thereby activating the tyrosine kinase activity of the latter and triggering a cell signaling cascade which is required for subsequent steps of the infection process. This mechanism would be similar to that proposed for Salmonella typhimurium invasion of mammalian cells (Galan et al., 1992; Pace et al., 1993), in which binding of the bacteria to cell surface structures stimulates the EGFR, leading to a signaling cascade that promotes Salmonella invasion. The second possibility is that reovirus takes advantage of an already activated signal transduction pathway conferred by the presence of functional EGFR on the host cell. In this case, the binding of the virus to EGFRs would represent a fortuitous event that is unrelated to the ensuing infection. The latter possibility is favored because of the following considerations. First, a single infectious reovirus particle is sufficient to initiate the infection process. Second, reovirus recognizes cell surface sialic acid residues and is therefore capable of interacting with a variety of cell surface sialoglycoproteins, rather than with a single species such as the EGFR (Choi et al., 1990; Tang et al., 1993). Third, even if this interaction (between a reovirion and an EGFR) occurs and results in the triggering of a signal, it is doubtful that the signal from a single bound virus is strong enough to generate an intracellular environment that is now conducive to the subsequent infection process. More recently, we demonstrated that NIH-3T3 cells, which express a low number of EGFRs and do not respond mitogenically to EGF, are poorly infectible by reovirus. However, when transformed with the v-erbB oncogene, they become highly susceptible (Strong et al., 1996). This enhanced susceptibility is abrogated by treatment of the cells with genistein, an inhibitor of tyrosine protein kinases. Since v-erbB is essentially a truncated EGFR lacking the extracellular ligand-binding domain (residues 1–555) and possessing ligand-independent, constitutive tyrosine kinase activity (Fung et al., 1983; Ullrich et al., 1984; Miles and Robinson, 1985; Raines et al., 1985; Massoglia et al., 1990; Carter et al., 1995), our results strongly suggest that the mechanism of enhancement of infection efficiency conferred by EGFR and v-erbB is through the opportunistic utilization by the virus of an already activated signal transduction pathway. In the present study, we probe the mechanism of reovirus oncolysis using NIH-3T3 cells transformed with intermediates in the EGFR and Ras signaling pathway. We show that the block in reovirus-resistant NIH-3T3 cells is at the level of translation of viral transcripts, and that activated intermediates in the Ras signaling pathway such as Son of Sevenless (Sos) or Ras are capable of releasing this block. Inhibition of viral mRNA translation in parental NIH-3T3 cells is accompanied by the phosphorylation of a 65 kDa cellular protein identified as the double-stranded RNA (dsRNA)-activated protein kinase (PKR). Phosphorylated PKR is known to catalyze the phosphorylation of the α-subunit of eukaryotic initiation factor 2 (eIF-2α) on Ser51, leading to inhibition of the initiation of protein synthesis (for reviews, see Hershey, 1991; Redpath and Proud, 1994; Proud, 1995). PKR is not phosphorylated in the reovirus-infected Sos- or Ras-transformed cells. Inactivation of PKR by 2-aminopurine, or deletion of the Pkr gene in untransformed cells, results in enhanced translation of viral transcripts in these cells. Taken together, our results indicate that reovirus exploits the ability of an element(s) of the cellular Ras signaling pathway to down-regulate PKR. Transformed cells with an activated Ras signaling pathway are therefore particularly susceptible to reovirus infection. Results Activated intermediates in the Ras signaling pathway augment reovirus infection efficiency Previously, we showed that NIH-3T3 cells and their derivatives lacking EGFRs are poorly infectible by reovirus, whereas the same cells transformed with either EGFR or v-erbB are highly susceptible as determined by cytopathic effects, viral protein synthesis and virus output (Strong et al., 1993; Strong and Lee, 1996). To determine which downstream mediators of the EGFR signal transduction pathway may be involved in this capacity, we took advantage of the availability of a number of different NIH-3T3-derived cell lines transformed with constitutively activated oncogenes downstream of the EGFR and assayed for their relative susceptibility to reovirus infection. Of particular interest were intermediates in the Ras signaling pathway (reviewed by Barbacid, 1987; Cahill et al., 1996). To this end, NIH-3T3 parental cell lines and NIH-3T3 lines transfected with activated versions of sos (Aronheim et al., 1994) or ras (Mundschau and Faller, 1992) oncogenes were exposed to reovirus and compared in terms of their capacity to promote viral protein synthesis. Detection of viral proteins initially was carried out using indirect immunofluorescent microscopy (Figure 1A). On comparing the uninfected parental cell lines with the various transformed cell lines, it was apparent that the morphology of the cells was quite distinct upon transformation. Whereas the NIH-3T3 cells adopted a typically flattened, spread-out morphology with marked contact inhibition, the transformed cells often grew as spindle-shaped cells with much less contact inhibition. Upon challenge with reovirus, it became clear that the parental NIH-3T3 line was poorly infectible (<5%; this occurred regardless of the source of the parental NIH-3T3 line). The observation that a small proportion of cells were infectible was probably due to the fact that NIH-3T3 cells can sometimes undergo transformation spontaneously (Rubin et al., 1995). In contrast, cell lines transformed with Sos or Ras demonstrated relatively pronounced immunofluorescence by 48 h post-infection. Figure 1.Effect of activated Sos and activated Ras on host cell susceptibility to reovirus infection. (A) Immunofluorescence assay of viral proteins expressed in reovirus-infected NIH-3T3 cells, Sos-transformed cells (TNIH#5), H-Ras- and EJ-Ras-transformed cells. Cells were infected with reovirus at an estimated m.o.i. of 10 p.f.u. per cell. At 48 h post-infection, cells were fixed, processed and reacted with rabbit anti-reovirus type 3 antibody and then with FITC-conjugated goat anti-rabbit immunoglobulin G. The magnification for all panels is ×200. (B) Reovirus protein synthesis in mock-infected and reovirus-infected NIH-3T3 cells, Sos-transformed cells (TNIH#5), H-Ras- and EJ-Ras-transformed cells. Cells were labeled with [35S]methionine from 12 to 48 h post-infection. Lysates were then prepared and either analyzed directly by SDS–PAGE (left panel) or immunoprecipitated with the polyclonal anti-reovirus type 3 serum and then analyzed by SDS–PAGE (right panel). The positions of reovirus proteins are indicated on the right. Download figure Download PowerPoint To demonstrate further that viral protein synthesis was more efficient in the Sos- or Ras-transformed cell lines, cells were labeled continuously with [35S]methionine from 12 to 48 h post-infection and the proteins were analyzed by SDS–PAGE. The results (Figure 1B) show clearly that the levels of viral protein synthesis were significantly higher in the Sos- or Ras-transformed cells than in parental NIH-3T3 cells. The identities of the viral bands were confirmed by immunoprecipitation of the labeled proteins with polyclonal anti-reovirus antibodies. Since the uninfected NIH-3T3 cells and their transformed counterparts displayed comparable levels of cellular protein synthesis and doubling times (data not shown), the observed difference in the level of viral protein synthesis could not be due to intrinsic differences in growth rates or translation efficiencies for these cell lines. The long-term fate of infected NIH-3T3 cells was followed by passaging these cells for at least 4 weeks. They grew normally and appeared healthy, with no sign of lytic or persistent infection; no virus could be detected in the medium after this time (data not shown). Enhanced infectibility conferred by the activated oncogenes is not due to long-term transformation or the generalized transformed state of the cell To determine whether the differences in susceptibility may be the result of some long-term effects of transformation, or just the result of the presence of the activated oncogene itself, a cell line expressing a Zn-inducible cellular Harvey-ras (c-H-ras) gene was tested for its susceptibility to reovirus as before. These cells (called 2H1) were derived from the murine fibroblast cell line C3H 10T1/2 which we have shown previously to be poorly infectible by reovirus (unpublished data), and carry the c-H-ras gene under the control of the mouse metallothionein-I promoter (Trimble et al., 1986). Cells were either mock treated or pre-treated with 50 μM ZnSO4 18 h prior to infection, followed by indirect immunofluorescent analysis of these cells at 48 h post-infection or mock infection. The results (Figure 2A) demonstrate that uninduced cells were poorly infectible (<5%) whereas those induced for only 18 h were much more susceptible (>40%). Enhanced viral protein synthesis in the Zn-induced 2H1 cells was confirmed further by metabolic labeling of the cells with [35S]methionine followed by SDS–PAGE analysis of virus-specific proteins (Figure 2B). Based on these observations, we conclude that the augmentation of reovirus infection efficiency in the transformed cells is a direct result of the activated oncogene products, and not due to other factors that might have contributed to a stably transformed state of these established cell lines. Figure 2.Effect of transient induction of activated Ras or c-Myc on host cell susceptibility to reovirus infection. (A) Immunofluorescence assay of viral proteins expressed in reovirus-infected 2H1 cells (containing a Zn-inducible ras gene) in the presence or absence of zinc. 2H1 cells were pre-treated with 50 μM ZnSO4 (+Zn) or mock treated (−Zn) for 18 h prior to infection. Cells were then infected with reovirus at an estimated m.o.i. of 10 p.f.u. per cell or mock infected. At 48 h post-infection, cells were fixed, processed and reacted with rabbit anti-reovirus type 3 antibody, followed by FITC-conjugated goat anti-rabbit IgG. The magnification for all panels is ×200. (B) Reovirus protein synthesis in mock-infected and reovirus-infected 2H1 cells in the presence or absence of ZnSO4. Cells were labeled with [35S]methionine from 12 to 48 h post-infection. Lysates were then prepared and either analyzed directly by SDS–PAGE (left panel) or immunoprecipitated with the polyclonal anti-reovirus type 3 serum and then analyzed by SDS–PAGE (right panel). The positions of reovirus proteins are indicated on the right. (C) NIH-3T3 tet-myc cells (containing the human c-myc gene whose expression is repressed in the presence of 2 μg/ml tetracycline) grown in the presence (+) or absence (−) of tetracycline were infected with reovirus and labeled with [35S]methionine from 12 to 48 h post-infection. Lysates were then prepared and either analyzed directly by SDS–PAGE or immunoprecipitated with the polyclonal anti-reovirus serum and then analyzed by SDS–PAGE. Download figure Download PowerPoint To show further that susceptibility to reovirus infection is not a result of transformation per se (i.e. the transformed state of the host cell), we examined NIH-3T3 cells containing a tetracycline-controlled human c-myc gene (tet-myc cells) (Helbing et al., 1997). These cells normally are maintained in tetracycline (2 μg/ml) which represses the expression of c-myc. Removal of tetracycline under normal growth conditions (10% fetal bovine serum) leads to accumulation of the c-Myc protein and the cells display a transformed phenotype. We found that these cells were unable to support virus growth in either the presence or absence of tetracycline (Figure 2C), which again suggests that susceptibility to reovirus infection is not due to the general transformed state of the host cell, but rather requires specific transformation by elements of the Ras signaling pathway. A good indicator of activation of the Ras signaling pathway is the activation of the MAP kinases ERK1 and ERK2 (for a review, see Robinson and Cobb, 1997). In this regard, we have found that compared with untransformed cells, Ras-transformed cells have a significantly higher ERK1/2 activity (unpublished observation). Furthermore, an examination of a number of human cancer cell lines has revealed an excellent correlation between the level of ERK1/2 activity and susceptibility to reovirus infection (unpublished observation), although ERK1/2 itself does not appear to play any role in it (see below). Not surprisingly, mouse L cells and human HeLa cells, in which reovirus grows very well, both manifest high ERK1/2 activity (data not shown). Viral transcripts are generated, but are not translated, in reovirus-resistant NIH-3T3 cells To elucidate the role(s) of these oncogenes in reovirus infection, it was important first to identify the step at which reovirus infection is blocked in the non-susceptible NIH-3T3 cells. We have demonstrated previously that virus binding and virus internalization into non-susceptible cells are comparable with those observed for susceptible cells (Strong et al., 1993). It would therefore be of interest to determine whether early transcription of viral genes, a translation-independent process, proceeds normally in the non-susceptible cells. Accordingly, the relative amounts of reovirus s1 transcripts generated in NIH-3T3 cells and in the H-Ras-transformed cells during the first 12 h of infection were compared after amplification of these transcripts by quantitative PCR. The results, shown in Figure 3, demonstrate that the rates of accumulation of s1 transcripts in the two cell lines were similar, at least up to 12 h post-infection. Similar data (not shown) were obtained when other reovirus transcripts were compared. These experiments therefore demonstrate that infection block in the non-susceptible cells is not at the level of transcription of viral genes, but rather at the level of translation of these transcripts. At later times, the level of viral transcripts present in untransformed NIH-3T3 cells decreased significantly whereas transcripts in transformed cells continued to accumulate (data not shown). The inability of these transcripts to be translated in NIH-3T3 cells probably contributed to their degradation. As expected, the level of viral transcripts in infected L cells was at least comparable with that in infected Ras-transformed cells (data not shown). Figure 3.Reovirus s1 mRNA levels in infected NIH-3T3 cells and H-Ras cells. Cells were infected with reovirus at an estimated m.o.i.of 10 p.f.u. per cell. At various times post-infection, cells were harvested and RNA was extracted from them. Equal amounts of RNA from each sample were then subjected to RT–PCR, followed by selective amplification of reovirus s1 cDNA and the constitutively expressed GAPDH, which served as a PCR and gel loading control. The PCR products were separated on a 2% agarose gel and visualized with ethidium bromide under UV light. Download figure Download PowerPoint The finding that viral transcripts were generated in untransformed NIH-3T3 cells led to the question as to whether these cells should still be identified as ‘resistant’, ‘non-susceptible’ or ‘not infectible’. We have opted not to change these long-held designations since there is little or no infectious outcome in these cells (i.e. the infection is abortive). Thus, cells that are ‘resistant’ or ‘non-susceptible’ to reovirus infection could still harbor viral transcripts but these transcripts are not translated. A 65 kDa protein is phosphorylated in reovirus-treated NIH-3T3 cells, but not in reovirus-infected transformed cells The above observation that viral transcripts were generated, but were not translated in NIH-3T3 cells led us to entertain the possibility that PKR could be activated (phosphorylated) in these cells [e.g. by s1 mRNA transcripts which have been shown to be potent activators of PKR (Bischoff and Samuel, 1989)], which in turn leads to inhibition of translation of viral genes. The corollary of such a scenario would be that in the case of the transformed cells, this activation is prevented, allowing viral protein synthesis to ensue. To test the above hypothesis, NIH-3T3 cells and v-erbB- or Ras-transformed cells (designated THC-11 and H-ras, respectively) were treated with reovirus (or mock-treated), and at 48 h post-infection were subjected to in vitro kinase reactions, followed by SDS–PAGE for autoradiographic analysis. The results (Figure 4A) clearly show that there was a distinct phosphoprotein migrating at ∼65 kDa (the expected size of PKR) only in the NIH-3T3 cells and only after exposure to reovirus (lane 3). This protein was not labeled in the lysates of either the uninfected or infected transformed cell lines (lanes 4–7), although it was present in comparable amounts in all the cell lines tested (data not shown). Instead, a protein migrating at ∼100 kDa was found to be labeled in the transformed cell lines after viral infection (Figure 4A, lanes 5 and 7). This protein was absent in both the pre-infection and post-infection lysates of the NIH-3T3 cell line (lanes 2 and 3), and was not a reovirus protein since it did not react with an anti-reovirus serum that precipitated all reovirus proteins (data not shown). A similar 100 kDa protein was also found to be 32P-labeled in in vitro kinase reactions of post-infection lysates of the Sos- or Raf-1-transformed cell lines (data not shown). The identity of this 100 kDa protein presently is unknown. Figure 4.In vitro kinase reactions of lysates from uninfected and reovirus-infected cells. (A) In vitro kinase reactions of lysates from uninfected and reovirus-infected NIH-3T3, THC-11 and H-Ras cells. Cells were harvested at 48 h post-infection and lysates were prepared. After normalizing for total protein concentration, the lysates were incubated with [γ-32P]ATP for 30 min at 37°C, followed by SDS–PAGE analysis. Lane 1 represents the marker lane which shows the positions of reovirus type 3 structural proteins. (B) In vitro kinase reactions of lysates from uninfected and reovirus-infected 2H1 cells and Zn-induced 2H1 cells. The positions of molecular weight markers are indicated on the left. Download figure Download PowerPoint That intermediates in the Ras signaling pathway are responsible for the lack of phosphorylation of the 65 kDa protein was confirmed further by the use of the 2H1 cells which contain a Zn-inducible ras oncogene. As shown in Figure 4B, uninduced 2H1 cells [shown above (Figure 2A and B) to be relatively resistant to reovirus infection] were capable of activating the 65 kDa protein only after exposure to reovirus (compare lanes 1 and 2). However, 2H1 cells subjected to Zn induction of the H-ras oncogene showed significant impairment of the activation of this protein (compare lanes 2 and 4). This impairment coincided with the enhancement of viral synthesis in the presence of zinc as shown in Figure 2A and B. Our results therefore eliminated the possibility that phosphorylation of the 65 kDa protein is strictly an NIH-3T3-specific event, and clearly established the role of Ras in preventing (or reversing) this phosphorylation. It is interesting that the Zn-induced 2H1 cells did not produce the 100 kDa phosphoprotein seen in the infected, stably transformed H-Ras cells. The reason for this is unclear at present. Identification of the 65 kDa phosphoprotein as PKR To determine whether the 65 kDa protein was indeed PKR, a dsRNA-binding experiment was carried out in which poly(I)–poly(C)–agarose beads were added to the 32P-labeled lysates. After incubation for 30 min at 4°C, the beads were washed, and bound proteins were released and analyzed by SDS–PAGE. The results (Figure 5A) show that the 65 kDa phosphoprotein present in the post-infection NIH-3T3 cell lysates was capable of binding to dsRNA, a well-recognized characteristic of PKR. In contrast, the 100 kDa phosphoprotein detected in the infected H-Ras-transformed cell line did not bind to the poly(I)–poly(C)–agarose. That the 65 kDa phosphoprotein was indeed PKR was confirmed further by the demonstration that it was immunoprecipitable with a PKR-specific antibody (Figure 5B). Figure 5.Identification of the 65 kDa phosphoprotein as PKR. (A) Interaction of the 65 kDa phosphoprotein with poly(I)–poly(C). Cytoplasmic extracts were made from uninfected or reovirus-infected NIH-3T3 or H-Ras cells and subjected to in vitro kinase reaction in the presence of [γ-32P]ATP. The labeled extracts were then mixed with a 50% agarose–poly(I)–poly(C) slurry and incubated at 4°C for 1 h. The beads were then washed extensively and the adsorbed proteins were released and analyzed by SDS–PAGE. The leftmost lane represents the marker lane which shows the reovirus serotype 3 (ST3) structural proteins. (B) Immunoprecipitation of the 65 kDa phosphoprotein with anti-PKR antibody. In vitro kinase reactions of cytoplasmic extracts from uninfected or reovirus-infected NIH-3T3, THC-11 or H-Ras cells were subjected to immunoprecipitation with either pre-immune or anti-PKR serum, followed by SDS–PAGE analysis. Download figure Download PowerPoint Induction of PKR phosphorylation requires active viral transcription Since phosphorylation of PKR occurred only in non-susceptible cells, and only after the cells had been exposed to reovirus, it was of interest to determine whether active viral transcription is required for this event. To this end, reovirus was UV treated to inactivate its genome prior to addition to NIH-3T3 cells. Such treatment efficiently abolished viral gene transcription (as analyzed by PCR), and hence viral infectivity (data not shown). The cells were then incubated for 48 h and lysates were prepared and subjected to in vitro 32P labeling as before. The results are shown in Figure 6. Again, NIH-3T3 cells infected with untreated reovirus produced a prominent 65 kDa 32P-labeled band (PKR) not found in uninfected cells (compare lanes 2 and 3). However, cells exposed to the UV-inactivated reovirus behaved similarly to the uninfected controls, manifesting little PKR phosphorylation (lane 4). These results demonstrate that the induction of PKR phosphorylation requires the transcription of viral genes, and is not due to the presence of dsRNA in the input virus. This observation is compatible with the notion that PKR is phosphorylated upon direct binding with the viral transcripts. Figure 6.Comparison of reovirus and UV-inactivated reovirus in their ability to induce phosphorylation of the 65 kDa protein in NIH-3T3 cells. Lysates from uninfected NIH-3T3 cells (Un), and from NIH-3T3 cells exposed to reovirus (REO) or to UV-inactivated reovirus (UV) for 48 h were prepared and subjected to in vitro kinase reaction in the presence of [γ-32P]ATP, followed by SDS–PAGE analysis. Lane 1 represents the marker lane which shows the positions