Down-Regulation of p53 by Double-Stranded RNA Modu
http://www.100md.com
病菌学杂志 2005年第17期
Departments of Cancer Biology
Molecular Biology, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195
Department of Medicine, Yale University, New Haven, Connecticut 06516
Department of Obstetrics and Gynecology, Mazda Hospital, 2-15 Aosakiminami, Fuchu-cho, Aki-gun Hiroshima, 735-8585, Japan
ABSTRACT
p53 has been well characterized as a tumor suppressor gene, but its role in antiviral defense remains unclear. A recent report has demonstrated that p53 can be induced by interferons and is activated after vesicular stomatitis virus (VSV) infection. We observed that different nononcogenic viruses, including encephalomyocarditis virus (EMCV) and human parainfluenza virus type 3 (HPIV3), induced down-regulation of p53 in infected cells. Double-stranded RNA (dsRNA) and a mutant vaccinia virus lacking the dsRNA binding protein E3L can also induce this effect, indicating that dsRNA formed during viral infection is likely the trigger for down-regulation of p53. The mechanism of down-regulation of p53 by dsRNA relies on translation inhibition mediated by the PKR and RNase L pathways. In the absence of p53, the replication of both EMCV and HPIV3 was retarded, whereas, conversely, VSV replication was enhanced. Cell cycle analysis indicated that wild-type (WT) but not p53 knockout (KO) fibroblasts undergo an early-G1 arrest following dsRNA treatment. Moreover, in WT cells the onset of dsRNA-induced apoptosis begins after p53 levels are down-regulated, whereas p53 KO cells, which lack the early-G1 arrest, rapidly undergo apoptosis. Hence, our data suggest that the down-regulation of p53 facilitates apoptosis, thereby limiting viral replication.
INTRODUCTION
p53 was first characterized as the major cellular protein associated with the T antigen encoded by simian virus 40 (SV40), a small DNA virus (22, 24). Following its initial discovery, it became clear that p53 was very important in preventing aberrant cell growth and tumor development (10). Observations that the p53 gene is mutated in most human cancers and the exquisite susceptibility of p53 knockout (KO) mice to spontaneous appearance of cancer clearly established the importance of p53 as a tumor suppressor (14, 16). SV40 and other oncogenic viruses target p53, the major tumor suppressor protein in the cell, in order to induce cell proliferation, thus increasing the number of cells carrying their genomes (2, 40).
In addition to the SV40 T antigen, several proteins encoded by viruses with malignant potential have also been shown to interact with p53 (reviewed in reference 5). Due to the oncogenic potential of these viruses, viral proteins targeting p53 were generally thought to have a role in tumorigenesis. However, viruses without tumorigenic potential, such as vaccinia virus (VV), have also been shown to encode proteins that can target p53 (32). Moreover, p53 KO mice, in contrast to wild-type (WT) mice, are highly sensitive to vesicular stomatitis virus (VSV) infection (38). Since VSV is a nononcogenic small RNA virus, it is not clear how p53 is activated to promote apoptosis of the infected cells. Further evidence for a role for p53 in antiviral defense came from the observation that it can be induced by interferon (IFN), a classical antiviral cytokine (38).
The double-stranded RNA (dsRNA)-activated protein kinase (PKR) is a potent antiviral protein and has been shown to be essential for resistance against VSV and other viruses in vivo (2). PKR has been shown to phosphorylate p53 in vitro, whereas PKR KO cells showed impaired p53-mediated responses to doxorubicin (6, 7). However, it is not clear whether interaction between PKR and p53 has a role in antiviral defense.
While characterizing the cellular responses to viral infection, we observed that two nononcogenic viruses, encephalomyocarditis virus (EMCV) and human parainfluenza virus type 3 (HPIV3), induce down-regulation of p53 in infected cells. To address the importance of p53 in the cellular responses to these viruses, we performed virus yield experiments, which demonstrated that the absence of p53 had a detrimental effect on the growth of EMCV and HPIV3, in contrast to a positive effect on the replication of VSV. These observations suggest that the down-regulation of p53 in response to virus infection plays a role in cellular responses to certain viruses. Interestingly, we observed that while the WR strain of VV did not induce down-regulation of p53, a mutant VV deleted for the E3L gene (VVE3L) did. Because E3L codes for a dsRNA binding protein, we hypothesized that dsRNA produced during viral infection was the trigger to induce down-regulation of p53. Indeed, transfection of dsRNA induces down-regulation of p53 as efficiently as viral infection. We further demonstrate that the PKR and RNase L pathways activated by dsRNA are required to down-regulate p53 levels by inhibiting translation. Unexpectedly, p53 KO cells are more sensitive to dsRNA-induced apoptosis, suggesting that the down-regulation of p53 sensitizes cells to dsRNA-induced cell death in order to limit viral replication.
MATERIALS AND METHODS
Cells, viruses, and reagents. HT1080 cells, CV1 cells, and L929 cells were grown in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Primary cultures of mouse embryo fibroblasts (MEFs) were prepared as described elsewhere (45) and maintained in DMEM supplemented with 10% FBS from U.S. Biochemical Corp. (Cleveland, OH). Viral stocks for VSV strain Indiana, EMCV, and HPIV3 (kindly provided by Amiya Banerjee of the Cleveland Clinic Foundation) were prepared in VERO, L929, and CV1 cells, respectively. Sendai virus was purchased from Charles River Laboratories (North Franklin, CT). Antibodies against human p53, p21, total eukaryotic translation initiation factor 2 (eIF2), and murine PKR were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), -actin from Sigma (St. Louis, MO), phospho-eIF2 and phospho-p53 from Cell Signaling (Beverly, MA), mouse p53 from Novocastra Laboratories (Newcastle, United Kingdom), human Mdm2 from Oncogene Research Products (San Diego, CA), mouse IFN regulatory factor 3 (IRF-3) from Zymed Laboratories (South San Francisco, CA), and glyceraldehyde-3-phosphate dehydrogenase from Chemicon International (Temecula, CA). Antibodies against p56 were a gift from Ganes Sen (Cleveland Clinic Foundation), and antibodies against mouse RNase L were a gift from Robert Silverman. 2-Aminopurine (2-AP), doxorubicin, and MG132 were purchased from Sigma. CP31398 was provided as a gift from Pfizer. Z-VAD-FMK was purchased from Calbiochem. Poly(I:C) was purchased from Sigma Aldrich and poly(I) from Amersham Pharmacia Biotech (Piscataway, NJ). Fugene reagent was purchased from Roche Diagnostics (Indianapolis, IN).
Poly(I:C) transfections. Poly(I:C) was transfected using Fugene reagent (Roche) in all experiments according to protocols provided by the manufacturer. Briefly, 2 μg of poly(I:C) per 3 μl of Fugene was incubated in 100 μl of serum-free DMEM for 15 to 30 min before being added to the supernatants of cells containing 10% FBS.
Western blotting and IP. Western blotting and immunoprecipitation (IP) were performed as described elsewhere (34). Briefly, cells were lysed in 50 mM Tris buffer, pH 7.4, containing 150 mM of NaCl, 50 mM of NaF, 10 mM of -glycerophosphate, 1% Triton X-100, 0.1 mM of EDTA, 10% glycerol, and protease/phosphatase inhibitors. The samples were kept on ice for 10 min, vortexed, and centrifuged for 15 min at 14,000 rpm; the supernatant was collected in a new tube; and protein concentrations were determined using the protein assay kit from Bio-Rad (Hercules, CA). Thirty micrograms of total protein was separated on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE), transferred to Immobilon-PSQ membranes (Millipore Corp., Belford, MA), and probed with the indicated antibodies. For the IP, 0.5 to 1.0 mg of total protein was incubated with a polyclonal antibody against p53 (FL-393; Santa Cruz) overnight at 4°C. The immune complex was precipitated using protein G-Sepharose beads (Amersham Biosciences). The beads were washed five times with lysis buffer, and the immunoprecipitate was separated on SDS-PAGE before Western blotting with the appropriate antibodies.
Metabolic labeling. Metabolic labeling of p53 was performed as described elsewhere (28). Briefly, cells were grown in DMEM without methionine for 2 h, and [35S]methionine (Amersham) was added to the culture medium for 30 min. The medium was removed, the monolayer was washed three times with phosphate-buffered saline (PBS), and fresh complete medium was added with or without poly(I:C). The whole-cell extracts were collected at the indicated times after the addition of fresh medium. p53 from 1 mg of total protein from the lysates was immunoprecipitated with a rabbit polyclonal antibody (FL-393; Santa Cruz). The immunoprecipitate was separated on a 10% SDS-PAGE, and isotopic incorporation into p53 was determined using a phosphorimager.
To measure protein synthesis, cells were treated for the indicated times, the medium was removed, and the cells were first washed with PBS and then incubated with DMEM containing [35S]methionine for 15 min before the extracts were collected. Fifty micrograms of total protein was precipitated with trichloroacetic acid (TCA), and the amount of isotope incorporated was measured using an LS 6000 IC Beta Counter (Beckman Coulter, Fullerton, CA).
Virus yield. Virus yield experiments were performed as described elsewhere (45). Briefly, cells were infected for 1 h with the indicated virus in DMEM without serum, the inoculum was removed, and new medium containing 10% FBS was added. After the indicated times, the supernatants were collected for titration of progeny viruses. Virus titers in the supernatants were determined by the Reed and Muench end point calculation method or by plaque assay. Virus yields were determined in L929 cells or CV1 cells. For the plaque assay, 0.5% methylcellulose (Sigma) was added to the medium in order to restrict viral spread.
RNA extractions and Northern blotting. For Northern blotting, total RNA was extracted from 10-cm dishes using the TRIZOL reagent according to the protocols provided by the manufacturer (Invitrogen, Carlsbad, CA). The RNA was separated by electrophoresis on a formaldehyde agarose gel and transferred to a Hybond-N membrane (Amersham). Hybridizations were performed as described elsewhere (8) using a probe containing the coding region for human p53 (kindly provided by Inder Verma, Salk Institute).
Lentivirus vectors. Lentivirus vectors expressing short hairpin RNAs (shRNA) have been described previously (4). Pools of HT1080 cells containing the lentivirus vectors were selected and used for the experiments described. Lentivirus vectors expressing human p53 or green fluorescent protein (GFP) have been described previously (13). Stocks of recombinant lentiviruses were prepared using the 293T cell line transfected with the lentivirus plasmids along with packaging plasmids encoding viral structural proteins and the G-protein of VSV using LipofectAMINE reagent (Invitrogen) as recommended by the manufacturer. The virus-containing medium from 293T cells was collected 48 h later. All infections with lentiviruses were performed in a medium containing 8 μg/ml of Polybrene for 6 h, after which the medium was removed and regular medium added.
Cell cycle analysis. For cell cycle analysis, cells were treated for the indicated times and fixed in 70% ethanol for 30 min at 4°C. After fixation, cells were stained with propidium iodide, processed using the Cellular DNA Flow Cytometric Analysis kit (Roche), and analyzed in a FACScan (Becton Dickinson, San Jose, CA).
TUNEL. For the terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) assay, cells were treated for the indicated times and fixed with 1% paraformaldehyde in PBS followed by 70% ethanol for terminal deoxynucleotidyltransferase-mediated detection of broken DNA ends using the APO-BRDU kit (BD Biosciences, Palo Alto, CA). The percentage of TUNEL-positive cells detected by FACScan analysis for each condition is plotted.
Caspase 3/7 activity. Cells were grown in black-wall 96-well plates with transparent bottoms and treated as indicated. All treatments were performed in triplicate, and the results shown represent the averages of the three wells. Caspase activity was measured directly from the 96-well plate using the Apo-ONE Homogeneous Caspase-3/7 assay according to the protocols provided by the manufacturer (Promega, Madison, WI).
RESULTS
Viral infection induces down-regulation of p53. To characterize the role of p53 in antiviral defense, different viruses were used to infect HT1080 and HepG2 cells, both of which express wild-type p53. p53 levels were monitored at different times postinfection (p.i.) (Fig. 1A and B). In the initial screen, four negative-sense RNA viruses, the rhabdovirus VSV and three paramyxoviruses (respiratory syncytial virus [RSV], Sendai virus, and HPIV3), and one positive-sense RNA virus, the picornavirus EMCV, were tested. The WR strain of the poxvirus VV, a large DNA virus, and the VV mutant VVE3L, which codes for a dsRNA binding protein capable of blocking dsRNA-mediated activation of PKR and IRF-3, were also included (42). Unexpectedly, infection with EMCV, HPIV3, Sendai virus, or VVE3L induced down-regulation of p53. In contrast, infection with VSV, RSV, or VV did not affect p53 levels (Fig. 1A and B). The observation that the mutant VVE3L induced down-regulation of p53 while WT VV did not suggested that the trigger for the down-regulation of p53 was most likely the dsRNA produced during viral replication (Fig. 1B; compare lanes 7 and 8). Although VSV has been shown to induce up-regulation of p53 in MEFs (38), we could not detect changes in p53 levels after VSV infection in either HT1080 or HepG2 cells (Fig. 1A and B). It is clear from these experiments that the role of p53 in the response to viruses is dependent on the cell type and the type of virus.
Influence of p53 on viral replication. p53 is an important mediator of stress responses, and previous observations have implicated p53 in antiviral defense against VSV (38). Conversely, the results shown in Fig. 1 indicate that virus infection can also induce down-regulation of p53. Therefore, we decided to investigate the role of p53 in the context of infection with viruses that induced down-regulation of p53. Some viruses, such as Sendai virus, RSV, and the VVE3L mutant, have a restricted host range and cannot grow in HT1080 cells (data not shown). VSV and EMCV have a rapid replicative lytic cycle in HT1080 cells, while HPIV3 has a slower replicative cycle. Since p53 is involved in the defense against VSV in MEFs (38), we used this virus in comparison to EMCV and HPIV3. Stable pools of HT1080 cells expressing shRNA targeting p53 were generated where p53 expression was selectively silenced compared to shRNA-GFP controls (Fig. 2A). When GFP was expressed, specific silencing was seen in the pools expressing the shRNA-GFP (Fig. 2B). Importantly, these shRNA did not activate the IFN system, as indicated by the expression of the IFN- and dsRNA-induced protein 56 (p56) (Fig. 2C). It has been reported that the H1 promoter, the one used in our vectors, is less likely to induce the IFN response (3, 12, 26). Despite the absence of p56 expression in unstimulated cells, the pools expressing the shRNA-GFP or shRNA-p53 still responded to dsRNA treatment by expressing p56 (27). Interestingly, HT1080 cells expressing the shRNA-p53 showed less accumulation of p56 in response to dsRNA (Fig. 2C). The different pools were infected with VSV, HPIV3, or EMCV for the indicated times, the supernatant was collected, and virus yields were determined (Fig. 2D). VSV, which did not induce down-regulation of p53, showed a 60-fold increase in virus yield in cells expressing shRNA-p53 compared to control cells (Fig. 2D). This is consistent with observations that, in p53 KO MEFs, VSV shows increased virus yield (38). Conversely, we observed 2.7-fold and 22-fold decreases in virus yields for EMCV and HPIV3, respectively, in pools expressing shRNA-p53 compared to control cells (Fig. 2C). To determine if the 2.7-fold difference in the yield of EMCV was consistent and to rule out possible problems associated with cell lines and in vitro manipulation of gene expression, virus yields from primary MEFs derived from p53 KO mice were also determined. Consistently, we observed a 3.4-fold reduction in EMCV yields from p53 KO MEFs compared to those for primary MEFs from WT mice (Fig. 2F). To further determine whether the absence of p53 alone was the cause of the difference between WT and KO cells, the KO MEFs were transduced with a lentivirus expressing human p53 prior to infection with EMCV. By Western blotting, it was observed that human p53 was successfully expressed in the KO cells (Fig. 2E). The expression of human p53, but not GFP, in the p53 KO MEFs partially restored the EMCV yield obtained in a concentration-dependent manner compared to control WT cells (Fig. 2F). Therefore, the down-regulation of p53 likely has a physiological role in controlling viral replication.
p53 levels are down-regulated in a dose- and time-dependent manner after dsRNA treatment. We then decided to test whether dsRNA produced during viral replication was the trigger for the down-regulation of p53. HT1080 cells, which not only retain WT p53 but also manifest a robust response following exposure to dsRNA (23, 27, 30), were transfected with poly(I:C), a synthetic analogue of dsRNA. As with viral infection, after transfection of poly(I:C), p53 protein levels were down-regulated in a time-dependent manner (Fig. 3A). As an indication of dsRNA treatment, we also observed accumulation of p56 following poly(I:C) treatment (Fig. 3A). The down-regulation of p53 was temporally coincident with the appearance of cleaved poly(ADP-ribose) polymerase (PARP) an indication of caspase activation and apoptosis (Fig. 3A). This raised the question of whether the down-regulation of p53 was a side effect of dsRNA-induced apoptosis. To rule this out, HT1080 cells were treated with poly(I:C) in the presence of Z-VAD-FMK, a general caspase inhibitor. The results (Fig. 3B) show that the down-regulation of p53 was not prevented by Z-VAD-FMK, although PARP cleavage and apoptosis were inhibited.
dsRNA treatment does not change the half life of p53. p53 protein levels are controlled mainly by the rate of degradation rather than synthesis (1). Therefore, we decided to determine whether dsRNA treatment affected the rate of p53 degradation. The half-life of p53 was first measured by treating cells with cycloheximide to block protein synthesis in the presence or absence of poly(I:C) treatment. Surprisingly, no difference was detected in the rate of decay of p53 protein levels with or without dsRNA treatment (Fig. 3C and D). The same results were seen using another chemical inhibitor of protein synthesis, emetine (data not shown). Because chemical inhibitors may have secondary effects, the half-life of p53 was also determined by metabolic pulse-labeling of HT1080 cells. Following labeling, the cells were washed and treated with poly(I:C) for different times, p53 was immunoprecipitated from cell extracts, and the amount of 35S incorporated into p53 was measured (Fig. 3E). Confirming the results obtained using protein synthesis inhibitors, no differences in the half-life of p53 with or without poly(I:C) treatment were detected.
Inhibition of the proteasome can block down-regulation of p53 in response to dsRNA. dsRNA treatment did not interfere with the rate of p53 degradation by the proteasome. However, it was still important to determine if basal activity of the proteasome was required for the down-regulation of p53. Accordingly, HT1080 cells were treated with increasing concentrations of the proteasome inhibitor MG132, and p53 levels were measured. Interestingly, even very low concentrations of MG132 (1 μM) that could barely induce changes in p53 basal levels in untreated cells had a strong inhibitory effect on the down-regulation induced by dsRNA (compare lanes 4 and 5 in Fig. 3F). It should be noted that at the highest concentration of MG132, 5 μM, in combination with poly(I:C), extensive cytotoxicity was observed, which could explain the smaller amount of -actin detected in Fig. 3F, lane 6.
dsRNA treatment does not affect the interaction between p53 and Mdm2. Mdm2 is the major E3 ubiquitin ligase controlling p53 degradation by the proteasome (1). The interaction between p53 and Mdm2 is regulated by posttranslational modifications of both proteins. To verify if dsRNA treatment could regulate the interaction between p53 and Mdm2, we performed co-IP experiments with HT1080 cells. Likely due to low levels of p53 in poly(I:C)-treated cells, we could not detect any association with Mdm2 (compare levels in whole-cell lysates [WCL] in Fig. 3H, lane 12, to lane 13 in Fig. 3G). For that reason, we used MG132 to block the down-regulation of p53 induced by dsRNA and then immunoprecipitated p53 from cells treated with MG132 alone or with poly(I:C) for 0.5, 1.0, 2.5, 5.0, and 7.0 h. As expected, MG132 induced accumulation of p53 and Mdm2 in WCL (Fig. 3H). However, poly(I:C) treatment did not consistently affect the interaction between p53 and Mdm2 (compare lanes 3 and 4, 5 and 6, 7 and 8, 9 and 10, 11 and 12 in Fig. 3G). Although there was somewhat less interaction between p53 and Mdm2 in poly(I:C)-treated cells, we also observed lower levels of p53 in the immunoprecipitate and in the WCL (Fig. 3G and H). Consistently, dsRNA treatment did not affect the half-life of p53 (Fig. 3D and E) or the interaction with Mdm2 (Fig. 3G).
Activation of PKR and RNase L pathways by dsRNA induces down-regulation of p53. dsRNA treatment can activate PKR and RNaseL pathways, resulting in protein synthesis inhibition (41, 44). To determine if the protein synthesis inhibition after treatment correlated with down-regulation of p53, HT1080 cells were treated with poly(I:C), and protein synthesis was measured by incorporation of [35S]methionine into newly synthesized proteins (Fig. 4A and C). The level of protein synthesis inhibition was dependent on the dose and the time after treatment and could be correlated with phosphorylation of eIF2 (Fig. 4A, B, C, and D). eIF2 is an in vivo substrate for PKR that, when phosphorylated, leads to inhibition of translation initiation (41). 2-AP, an ATP analogue capable of inhibiting PKR-induced eIF2 phosphorylation, showed dose-dependent inhibition of eIF2 phosphorylation in HT1080 cells treated with poly(I:C) (Fig. 4E). At the same time, 2-AP blocked the down-regulation of p53 in a dose-dependent manner (Fig. 4E). Since it has been shown previously that 2-AP can also inhibit pathways other than PKR that are involved in phosphorylation and stabilization of p53 (17), it is likely that, in our experiment, inhibition of PKR by 2-AP stabilizes p53 (Fig. 4E). We also monitored phosphorylation of p53 on serine 392, since PKR was shown to phosphorylate p53 on that residue in vitro. However, we did not detect any significant changes in phosphorylation of serine 392 after dsRNA treatment (Fig. 3H). To determine whether RNase L was activated following dsRNA treatment, we isolated total RNA from cells treated with poly(I:C). After separating the RNA on a denaturing gel, we observed clear rRNA cleavage, an indication of RNase L activation (44) (Fig. 4F). The same RNA was transferred to a nylon membrane, and Northern blotting was performed using a probe for p53. Following the same pattern of rRNA degradation, p53 mRNA levels were also down-regulated after dsRNA treatment, implicating RNase L in the down-regulation of p53 (Fig. 4F). To further confirm the requirements for PKR and RNase L, primary MEFs derived from PKR or RNase L knockout mice were transfected with poly(I:C). In both PKR and RNase L primary single-KO MEFs, down-regulation of p53 in response to dsRNA was partially impaired compared to that in the WT controls (Fig. 4G). To further verify the involvement of both PKR and RNase L pathways, primary MEFs derived from PKR/RNase L double-knockout (DKO) mice were transfected with poly(I:C). In DKO cells the down-regulation of p53 and protein synthesis inhibition in response to dsRNA were abrogated, indicating cooperation between the two pathways (Fig. 4H and I). We also observed that dsRNA treatment of DKO cells induced up-regulation of p53 due to the lack of protein synthesis inhibition in response to dsRNA (Fig. 4H). This is likely due to induction of IFN-/, consistent with the observations of Takaoka et al. (38). IFN production in WT cells also results in the up-regulation of PKR and RNase L, proteins known to be IFN inducible (Fig. 4H) (36). There is a high background in Western blots measuring p53 levels using whole-cell extracts from primary MEFs due to the very low expression levels of p53 and the lack of highly specific antibodies. Although the abilities of different antibodies to detect p53 in primary MEFs differed, down-regulation of p53 was observed with all of them (data not shown). We also confirmed that the antibodies were recognizing p53 by using extracts obtained from p53 KO cells or WT cells treated with doxorubicin as positive controls to increase the levels of p53 (Fig. 5D and data not shown). Moreover, when human p53 was overexpressed in primary MEFs, its down-regulation after dsRNA treatment could still be detected (Fig. 2E). In order to clearly illustrate the down-regulation of p53 levels in primary MEFs, we performed quantitation using ImageQuant 1.2 software and normalized the numbers obtained to -actin (represented as the p53/-actin ratio compared to that for untreated controls in Fig. 4G and H).
p53 knockout cells show increased sensitivity to dsRNA-induced apoptosis. We have shown that viral infection leads to down-regulation of p53 through inhibition of translation induced by dsRNA produced during viral replication. Moreover, in the absence of p53, certain viruses exhibit decreased replication, indicating that the down-regulation of p53 observed during viral infection might be important for the antiviral response. Therefore, we decided to address the mechanism. dsRNA is a trigger for apoptosis, which is one mechanism that infected cells use to limit viral replication (11). We observed a temporal correlation between the down-regulation of p53 and apoptosis in dsRNA-treated cells (Fig. 3A). However, the down-regulation of p53 is not caused by, and occurs independently of, apoptosis (Fig. 3B). Because of the known proapoptotic role of p53, we determined the cellular responses to dsRNA in the absence of p53. In order to address the mechanism by which the absence of p53 could influence the response of cells to dsRNA or viral infection, primary cultures of MEFs derived from WT and p53 KO mice were transfected with poly(I:C). Cell cycle analysis was performed by flow cytometry determining the DNA content of WT and p53 KO cells at different times. Following dsRNA treatment, WT cells show an early-G1/S arrest (increased G1/G2 ratio) that is not seen in p53 KO cells, indicating that it is p53 dependent (Fig. 5A). Interestingly, in WT cells, the down-regulation of p53 (see Fig. 4G) occurs after the G1 arrest and precedes the outcome of apoptosis (sub-G1 cells in Fig. 5B). p53 KO cells, which have no apparent early-G1 arrest, rapidly exhibit an increase in sub-G1 apoptotic cells as early as 8 h after treatment (Fig. 5A and B). To further verify the differences in cell death, apoptosis was measured by TUNEL staining after 16 h of treatment. MEFs derived from p53 KO mice treated with dsRNA exhibited about fourfold more apoptosis than wild-type controls treated similarly (Fig. 5C). To determine if the p53 KO cells were more sensitive not only to dsRNA but to any apoptotic stimuli, cells were treated with a combination of tumor necrosis factor (TNF) or untransfected poly(I:C) with cycloheximide, agents known to induce apoptosis (9). The results (Fig. 5C) confirmed that even in the presence of cycloheximide, p53 KO MEFs were more sensitive to dsRNA- but not TNF-induced apoptosis than WT MEFs. If down-regulation of p53 increases apoptosis, then in PKR and RNase L KO cells, apoptosis should be inhibited due to impaired down-regulation of p53. Indeed, PKR and RNase L single-KO cells were more resistant to various apoptotic stimuli (9, 44) and showed some degree of resistance to dsRNA-induced caspase activation (Fig. 5E). Since PKR/RNase L DKO cells did not show any down-regulation of p53 after dsRNA treatment, they were tested for their sensitivity to dsRNA-induced apoptosis. Compared to WT or PKR/RNase L single-KO MEFs, primary DKO MEFs were highly resistant to apoptosis induced by dsRNA, as measured by caspase 3/7 activation (Fig. 5E).
DISCUSSION
Although p53 has a clear role in genotoxic stress responses and maintaining genomic stability, our results demonstrate that its role in antiviral defense is complex. Previous studies have shown that oncogenic viruses target p53 to promote cell growth and that viruses such as VSV induce p53-dependent apoptosis (5, 38). We show that in the absence of viral mechanisms that activate or neutralize p53, dsRNA generated during infection will induce down-regulation of p53 as a default response. This does not appear to be a general mechanism used by all viruses, even though they are all likely to produce dsRNA at some stage during their replicative cycle. We observed that for both HPIV3 and EMCV, which induced down-regulation of p53 in infected cells, viral replication was reduced in the absence of p53. Some other viruses do not induce cell death and replicate at lower rates, establishing persistent infection (43).
dsRNA treatment did not induce any posttranslational modification of p53, nor were there changes in the protein half-life or association of p53 with Mdm2 (Fig. 3C, D, E, and G). We cannot completely rule out other possible mechanisms, but protein synthesis inhibition caused by activation of the PKR and RNase L pathways by dsRNA is required (Fig. 4). Therefore, translation inhibition prevents new p53 protein synthesis, leading to down-regulation of its levels, while the turnover of p53 through the proteasome pathway is unaffected (Fig. 3F).
dsRNA produced during viral infection is a trigger for down-regulation of p53, which in turn inhibits viral replication. There was a temporal coincidence between the outcome of apoptosis and the down-regulation of p53 (Fig. 3A). Cells can limit viral replication by undergoing apoptosis, and dsRNA produced during infection is a trigger for apoptosis (21). Furthermore, due to the role of p53 in apoptosis and in antiviral defense against VSV, we expected the absence of p53 to have severe consequences. Surprisingly, down-regulation of p53 was not a consequence of apoptosis but rather seemed to sensitize cells to death (Fig. 3B and 5C). Accordingly, apoptosis in response to dsRNA is not triggered by p53 and can occur in the absence of p53 function (39). Apoptosis in response to dsRNA seems to require IRF-3 and PKR, although the pathways are not clearly defined (39, 41). We observed that p53 KO cells displayed greater sensitivity to apoptosis than WT cells, implying that p53 is protective against dsRNA-induced cell death (Fig. 4B and C). Raj et al. (29) have shown that p53 induces cell cycle arrest at G2 and protects cells from adeno-associated virus-mediated cell death. In WT cells, we observed an increase in the G1/G2 ratio at 2 h after dsRNA treatment, suggesting G1 arrest that is p53 dependent, since it is absent in p53 KO cells (Fig. 5A). However, 8 h after treatment, the G1/G2 ratios in WT and p53 KO cells are similar, indicating that at least at later time points, p53 does not play a significant role in controlling the cell cycle (Fig. 5A). In WT cells we observed that the sub-G1 apoptotic cell population increases at later time points (8 to 24 h), when p53 levels are significantly lower (Fig. 5B). However, in p53 KO cells, a dramatic increase in sub-G1 cells can be detected as early as 8 h after treatment (Fig. 5B). This difference was confirmed by TUNEL staining of cells at 16 h after treatment (Fig. 5C). Thus, our data indicate that cells containing WT p53 undergo cell cycle arrest at G1 after dsRNA treatment and delay apoptosis until p53 levels are low enough to allow progression through the cell cycle. In the absence of p53, or after p53 is down-regulated, cells are released from the G1 arrest and undergo apoptosis. Although the p53-dependent G1 arrest after dsRNA treatment is likely to delay apoptosis, it is not clear if that is the sole mechanism for the increased viral replication in the presence of p53. As observed in Fig. 2C, HT1080 cells where p53 expression was ablated showed impaired gene induction in response to dsRNA, as previously shown for p53 KO MEFs (19). Most dsRNA-induced genes have an antiviral function, implying that reduced levels in p53-null cells should facilitate viral replication. Instead, we observed decreased viral replication for EMCV and HPIV3. It remains possible that some dsRNA-induced genes might be required for delaying apoptosis, thereby explaining the increased sensitivity in the absence of p53.
These observations link the down-regulation of p53 to the limitation of viral replication by increasing dsRNA-induced apoptosis. This provides an explanation of how the down-regulation of p53 in response to infection with EMCV, HPIV3, and possibly other viruses restricts viral replication. The magnitude of the difference in virus yield in the presence and absence of p53 depended on the replicative cycle of the virus. HPIV3 has a slower replicative cycle than EMCV and is more likely to have its replication impaired by apoptosis of the infected cell. Nevertheless, the results with EMCV can also be considered significant considering its rapid replicative cycle, since apoptosis is necessary but not sufficient to limit viral replication. Accordingly, MEFs derived from NF-B p50 KO mice, which display a three- to fourfold increase in apoptosis after infection with EMCV, show a mild decrease in virus yield at the same time postinfection (33). Furthermore, in control cells, p53 levels are down-regulated, which might dampen differences in virus yield obtained in the presence or absence of p53. Interestingly, Yeung et al. (43) have shown that ablating PKR expression leads to persistent EMCV infection and constitutive expression of p53 in U937 cells. This provides additional evidence for the mechanism we propose, where, after EMCV infection, PKR promotes down-regulation of p53, thus facilitating apoptosis and preventing establishment of persistent infection.
The complexity of the p53 response to viral infection is also supported by studies with HCT116 colon cancer cells. The presence or absence of p53 had no effect on cell death induced by several RNA viruses in HCT116 cells (18), although in the absence of p53, HCT116 cells were more effectively killed by adeno-associated virus infection (29). The absence of p53 in HCT116 cells resulted in increased EMCV yield in our experiments (J. T. Marques and B. R. G. Williams, unpublished results). However, HCT116 cells are very resistant to EMCV infection as well as to dsRNA-induced apoptosis and cannot limit virus replication by undergoing apoptosis. Therefore, the mechanism described here is probably not pertinent in these cells (Marques and Williams, unpublished).
Our results and the observations of others raise concerns about the use of viruses, either as oncolytic viruses or as gene therapy vectors, in the treatment of cancer. Virus vectors, usually replication-deficient viruses, still activate a stress response within the infected cell, which is often stronger than the response activated by wild-type viruses (31). Although we did not directly test any of the deficient viruses currently used in clinical trials, we tested two viruses that, in our model, can infect the cell but do not generate a new progeny. These two viruses, Sendai virus and VVE3L, induced down-regulation of p53 to different degrees (data not shown). Several different oncolytic viruses that are currently being tested include reoviruses, paramyxoviruses (Newcastle disease virus), herpesviruses, adenoviruses, and VSV (15, 20, 25, 37). These viruses are likely to activate antiviral responses, and in the case of reoviruses, they induce protein synthesis inhibition through RNase L and PKR (35). Although p53 can have an antiapoptotic role by inducing cell cycle arrest (29; also our results), there is no doubt about the antitumor and proapoptotic role of p53 in response to radio- and chemotherapies. Our results suggest that both oncolytic and deficient viruses that are used to deliver suicide genes to tumor cells may also inadvertently cause down-regulation of p53. Hence, cancer treatment strategies relying on viral infection may impair subsequent radio- and chemotherapies.
ACKNOWLEDGMENTS
We thank Anthony Sadler, Mark Whitmore, Michelle Holko, and Patricia Stanhope-Baker for helpful discussions, Curt Horvath for helpful discussions and sharing unpublished data, Cathy Shemo for help with the cell cycle analysis and TUNEL assays, and Joe Didonato, Amiya Banerjee, Ganes Sen, and Kristi Peters for providing reagents.
This work was supported by NIH grants RO1 AI34039 and PO1 CA 62220.
REFERENCES
Ashcroft, M., and K. H. Vousden. 1999. Regulation of p53 stability. Oncogene 18:7637-7643.
Balachandran, S., P. C. Roberts, L. E. Brown, H. Truong, A. K. Pattnaik, D. R. Archer, and G. N. Barber. 2000. Essential role for the dsRNA-dependent protein kinase PKR in innate immunity to viral infection. Immunity 13:129-141.
Bridge, A. J., S. Pebernard, A. Ducraux, A. L. Nicoulaz, and R. Iggo. 2003. Induction of an interferon response by RNAi vectors in mammalian cells. Nat. Genet. 34:263-264.
Budanov, A. V., A. A. Sablina, E. Feinstein, E. V. Koonin, and P. M. Chumakov. 2004. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of bacterial AhpD. Science 304:596-600.
Collot-Teixeira, S., J. Bass, F. Denis, and S. Ranger-Rogez. 2004. Human tumor suppressor p53 and DNA viruses. Rev. Med. Virol. 14:301-319.
Cuddihy, A. R., S. Li, N. W. Tam, A. H. Wong, Y. Taya, N. Abraham, J. C. Bell, and A. E. Koromilas. 1999. Double-stranded-RNA-activated protein kinase PKR enhances transcriptional activation by tumor suppressor p53. Mol. Cell. Biol. 19:2475-2484.
Cuddihy, A. R., A. H. Wong, N. W. Tam, S. Li, and A. E. Koromilas. 1999. The double-stranded RNA activated protein kinase PKR physically associates with the tumor suppressor p53 protein and phosphorylates human p53 on serine 392 in vitro. Oncogene 18:2690-2702.
da Fonseca, F. G., R. L. Silva, J. T. Marques, P. C. Ferreira, and E. G. Kroon. 1999. The genome of cowpox virus contains a gene related to those encoding the epidermal growth factor, transforming growth factor alpha and vaccinia growth factor. Virus Genes 18:151-160.
Der, S. D., Y. L. Yang, C. Weissmann, and B. R. Williams. 1997. A double-stranded RNA-activated protein kinase-dependent pathway mediating stress-induced apoptosis. Proc. Natl. Acad. Sci. USA 94:3279-3283.
Donehower, L. A., M. Harvey, B. L. Slagle, M. J. McArthur, C. A. Montgomery, Jr., J. S. Butel, and A. Bradley. 1992. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356:215-221.
Everett, H., and G. McFadden. 1999. Apoptosis: an innate immune response to virus infection. Trends Microbiol. 7:160-165.
Fish, R. J., and E. K. Kruithof. 2004. Short-term cytotoxic effects and long-term instability of RNAi delivered using lentiviral vectors. BMC Mol. Biol. 5:9.
Gurova, K. V., J. E. Hill, O. V. Razorenova, P. M. Chumakov, and A. V. Gudkov. 2004. p53 pathway in renal cell carcinoma is repressed by a dominant mechanism. Cancer Res. 64:1951-1958.
Harvey, M., M. J. McArthur, C. A. Montgomery, Jr., J. S. Butel, A. Bradley, and L. A. Donehower. 1993. Spontaneous and carcinogen-induced tumorigenesis in p53-deficient mice. Nat. Genet. 5:225-229.
Hirasawa, K., S. G. Nishikawa, K. L. Norman, M. C. Coffey, B. G. Thompson, C. S. Yoon, D. M. Waisman, and P. W. Lee. 2003. Systemic reovirus therapy of metastatic cancer in immune-competent mice. Cancer Res. 63:348-353.
Hollstein, M., D. Sidransky, B. Vogelstein, and C. C. Harris. 1991. p53 mutations in human cancers. Science 253:49-53.
Huang, S., L. K. Qu, A. R. Cuddihy, R. Ragheb, Y. Taya, and A. E. Koromilas. 2003. Protein kinase inhibitor 2-aminopurine overrides multiple genotoxic stress-induced cellular pathways to promote cell survival. Oncogene 22:3721-3733.
Huang, S., L. K. Qu, and A. E. Koromilas. 2004. Induction of p53-dependent apoptosis in HCT116 tumor cells by RNA viruses and possible implications in virus-mediated oncolysis. Cell Cycle 3:1043-1045.
Hummer, B. T., X. L. Li, and B. A. Hassel. 2001. Role for p53 in gene induction by double-stranded RNA. J. Virol. 75:7774-7777.
Ikeda, K., T. Ichikawa, H. Wakimoto, J. S. Silver, T. S. Deisboeck, D. Finkelstein, G. R. Harsh IV, D. N. Louis, R. T. Bartus, F. H. Hochberg, and E. A. Chiocca. 1999. Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses. Nat. Med. 5:881-887.
Kibler, K. V., T. Shors, K. B. Perkins, C. C. Zeman, M. P. Banaszak, J. Biesterfeldt, J. O. Langland, and B. L. Jacobs. 1997. Double-stranded RNA is a trigger for apoptosis in vaccinia virus-infected cells. J. Virol. 71:1992-2003.
Lane, D. P., and L. V. Crawford. 1979. T antigen is bound to a host protein in SV40-transformed cells. Nature 278:261-263.
Leaman, D. W., A. Salvekar, R. Patel, G. C. Sen, and G. R. Stark. 1998. A mutant cell line defective in response to double-stranded RNA and in regulating basal expression of interferon-stimulated genes. Proc. Natl. Acad. Sci. USA 95:9442-9447.
Linzer, D. I., and A. J. Levine. 1979. Characterization of a 54K dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell 17:43-52.
Nemunaitis, J., I. Ganly, F. Khuri, J. Arseneau, J. Kuhn, T. McCarty, S. Landers, P. Maples, L. Romel, B. Randlev, T. Reid, S. Kaye, and D. Kirn. 2000. Selective replication and oncolysis in p53 mutant tumors with ONYX-015, an E1B-55kD gene-deleted adenovirus, in patients with advanced head and neck cancer: a phase II trial. Cancer Res. 60:6359-6366.
Pebernard, S., and R. D. Iggo. 2004. Determinants of interferon-stimulated gene induction by RNAi vectors. Differentiation 72:103-111.
Peters, K. L., H. L. Smith, G. R. Stark, and G. C. Sen. 2002. IRF-3-dependent, NF-B- and JNK-independent activation of the 561 and IFN- genes in response to double-stranded RNA. Proc. Natl. Acad. Sci. USA 99:6322-6327.
Qu, L., S. Huang, D. Baltzis, A. M. Rivas-Estilla, O. Pluquet, M. Hatzoglou, C. Koumenis, Y. Taya, A. Yoshimura, and A. E. Koromilas. 2004. Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3. Genes Dev. 18:261-277.
Raj, K., P. Ogston, and P. Beard. 2001. Virus-mediated killing of cells that lack p53 activity. Nature 412:914-917.
Rasheed, S., W. A. Nelson-Rees, E. M. Toth, P. Arnstein, and M. B. Gardner. 1974. Characterization of a newly derived human sarcoma cell line (HT-1080). Cancer 33:1027-1033.
Roth, M. D., Q. Cheng, A. Harui, S. K. Basak, K. Mitani, T. A. Low, and S. M. Kiertscher. 2002. Helper-dependent adenoviral vectors efficiently express transgenes in human dendritic cells but still stimulate antiviral immune responses. J. Immunol. 169:4651-4656.
Santos, C. R., F. M. Vega, S. Blanco, R. Barcia, and P. A. Lazo. 2004. The vaccinia virus B1R kinase induces p53 downregulation by an Mdm2-dependent mechanism. Virology 328:254-265.
Schwarz, E. M., C. Badorff, T. S. Hiura, R. Wessely, A. Badorff, I. M. Verma, and K. U. Knowlton. 1998. NF-B-mediated inhibition of apoptosis is required for encephalomyocarditis virus virulence: a mechanism of resistance in p50 knockout mice. J. Virol. 72:5654-5660.
Silva, A. M., M. Whitmore, Z. Xu, Z. Jiang, X. Li, and B. R. Williams. 2004. Protein kinase R (PKR) interacts with and activates mitogen-activated protein kinase kinase 6 (MKK6) in response to double-stranded RNA stimulation. J. Biol. Chem. 279:37670-37676.
Smith, J. A., S. C. Schmechel, B. R. Williams, R. H. Silverman, and L. A. Schiff. 2005. Involvement of the interferon-regulated antiviral proteins PKR and RNase L in reovirus-induced shutoff of cellular translation. J. Virol. 79:2240-2250.
Stark, G. R., I. M. Kerr, B. R. Williams, R. H. Silverman, and R. D. Schreiber. 1998. How cells respond to interferons. Annu. Rev. Biochem. 67:227-264.
Stojdl, D. F., B. D. Lichty, B. R. tenOever, J. M. Paterson, A. T. Power, S. Knowles, R. Marius, J. Reynard, L. Poliquin, H. Atkins, E. G. Brown, R. K. Durbin, J. E. Durbin, J. Hiscott, and J. C. Bell. 2003. VSV strains with defects in their ability to shutdown innate immunity are potent systemic anti-cancer agents. Cancer Cell 4:263-275.
Takaoka, A., S. Hayakawa, H. Yanai, D. Stoiber, H. Negishi, H. Kikuchi, S. Sasaki, K. Imai, T. Shibue, K. Honda, and T. Taniguchi. 2003. Integration of interferon-/ signalling to p53 responses in tumour suppression and antiviral defence. Nature 424:516-523.
Weaver, B. K., O. Ando, K. P. Kumar, and N. C. Reich. 2001. Apoptosis is promoted by the dsRNA-activated factor (DRAF1) during viral infection independent of the action of interferon or p53. FASEB J. 15:501-515.
Werness, B. A., A. J. Levine, and P. M. Howley. 1990. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 248:76-79.
Williams, B. R. 2001. Signal integration via PKR. Sci. STKE 2001:RE2.
Xiang, Y., R. C. Condit, S. Vijaysri, B. Jacobs, B. R. Williams, and R. H. Silverman. 2002. Blockade of interferon induction and action by the E3L double-stranded RNA binding proteins of vaccinia virus. J. Virol. 76:5251-5259.
Yeung, M. C., D. L. Chang, R. E. Camantigue, and A. S. Lau. 1999. Inhibitory role of the host apoptogenic gene PKR in the establishment of persistent infection by encephalomyocarditis virus in U937 cells. Proc. Natl. Acad. Sci. USA 96:11860-11865.
Zhou, A., J. Paranjape, T. L. Brown, H. Nie, S. Naik, B. Dong, A. Chang, B. Trapp, R. Fairchild, C. Colmenares, and R. H. Silverman. 1997. Interferon action and apoptosis are defective in mice devoid of 2',5'-oligoadenylate-dependent RNase L. EMBO J. 16:6355-6363.
Zhou, A., J. M. Paranjape, S. D. Der, B. R. Williams, and R. H. Silverman. 1999. Interferon action in triply deficient mice reveals the existence of alternative antiviral pathways. Virology 258:435-440.(Joao T. Marques, Dominiqu)
Molecular Biology, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195
Department of Medicine, Yale University, New Haven, Connecticut 06516
Department of Obstetrics and Gynecology, Mazda Hospital, 2-15 Aosakiminami, Fuchu-cho, Aki-gun Hiroshima, 735-8585, Japan
ABSTRACT
p53 has been well characterized as a tumor suppressor gene, but its role in antiviral defense remains unclear. A recent report has demonstrated that p53 can be induced by interferons and is activated after vesicular stomatitis virus (VSV) infection. We observed that different nononcogenic viruses, including encephalomyocarditis virus (EMCV) and human parainfluenza virus type 3 (HPIV3), induced down-regulation of p53 in infected cells. Double-stranded RNA (dsRNA) and a mutant vaccinia virus lacking the dsRNA binding protein E3L can also induce this effect, indicating that dsRNA formed during viral infection is likely the trigger for down-regulation of p53. The mechanism of down-regulation of p53 by dsRNA relies on translation inhibition mediated by the PKR and RNase L pathways. In the absence of p53, the replication of both EMCV and HPIV3 was retarded, whereas, conversely, VSV replication was enhanced. Cell cycle analysis indicated that wild-type (WT) but not p53 knockout (KO) fibroblasts undergo an early-G1 arrest following dsRNA treatment. Moreover, in WT cells the onset of dsRNA-induced apoptosis begins after p53 levels are down-regulated, whereas p53 KO cells, which lack the early-G1 arrest, rapidly undergo apoptosis. Hence, our data suggest that the down-regulation of p53 facilitates apoptosis, thereby limiting viral replication.
INTRODUCTION
p53 was first characterized as the major cellular protein associated with the T antigen encoded by simian virus 40 (SV40), a small DNA virus (22, 24). Following its initial discovery, it became clear that p53 was very important in preventing aberrant cell growth and tumor development (10). Observations that the p53 gene is mutated in most human cancers and the exquisite susceptibility of p53 knockout (KO) mice to spontaneous appearance of cancer clearly established the importance of p53 as a tumor suppressor (14, 16). SV40 and other oncogenic viruses target p53, the major tumor suppressor protein in the cell, in order to induce cell proliferation, thus increasing the number of cells carrying their genomes (2, 40).
In addition to the SV40 T antigen, several proteins encoded by viruses with malignant potential have also been shown to interact with p53 (reviewed in reference 5). Due to the oncogenic potential of these viruses, viral proteins targeting p53 were generally thought to have a role in tumorigenesis. However, viruses without tumorigenic potential, such as vaccinia virus (VV), have also been shown to encode proteins that can target p53 (32). Moreover, p53 KO mice, in contrast to wild-type (WT) mice, are highly sensitive to vesicular stomatitis virus (VSV) infection (38). Since VSV is a nononcogenic small RNA virus, it is not clear how p53 is activated to promote apoptosis of the infected cells. Further evidence for a role for p53 in antiviral defense came from the observation that it can be induced by interferon (IFN), a classical antiviral cytokine (38).
The double-stranded RNA (dsRNA)-activated protein kinase (PKR) is a potent antiviral protein and has been shown to be essential for resistance against VSV and other viruses in vivo (2). PKR has been shown to phosphorylate p53 in vitro, whereas PKR KO cells showed impaired p53-mediated responses to doxorubicin (6, 7). However, it is not clear whether interaction between PKR and p53 has a role in antiviral defense.
While characterizing the cellular responses to viral infection, we observed that two nononcogenic viruses, encephalomyocarditis virus (EMCV) and human parainfluenza virus type 3 (HPIV3), induce down-regulation of p53 in infected cells. To address the importance of p53 in the cellular responses to these viruses, we performed virus yield experiments, which demonstrated that the absence of p53 had a detrimental effect on the growth of EMCV and HPIV3, in contrast to a positive effect on the replication of VSV. These observations suggest that the down-regulation of p53 in response to virus infection plays a role in cellular responses to certain viruses. Interestingly, we observed that while the WR strain of VV did not induce down-regulation of p53, a mutant VV deleted for the E3L gene (VVE3L) did. Because E3L codes for a dsRNA binding protein, we hypothesized that dsRNA produced during viral infection was the trigger to induce down-regulation of p53. Indeed, transfection of dsRNA induces down-regulation of p53 as efficiently as viral infection. We further demonstrate that the PKR and RNase L pathways activated by dsRNA are required to down-regulate p53 levels by inhibiting translation. Unexpectedly, p53 KO cells are more sensitive to dsRNA-induced apoptosis, suggesting that the down-regulation of p53 sensitizes cells to dsRNA-induced cell death in order to limit viral replication.
MATERIALS AND METHODS
Cells, viruses, and reagents. HT1080 cells, CV1 cells, and L929 cells were grown in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS). Primary cultures of mouse embryo fibroblasts (MEFs) were prepared as described elsewhere (45) and maintained in DMEM supplemented with 10% FBS from U.S. Biochemical Corp. (Cleveland, OH). Viral stocks for VSV strain Indiana, EMCV, and HPIV3 (kindly provided by Amiya Banerjee of the Cleveland Clinic Foundation) were prepared in VERO, L929, and CV1 cells, respectively. Sendai virus was purchased from Charles River Laboratories (North Franklin, CT). Antibodies against human p53, p21, total eukaryotic translation initiation factor 2 (eIF2), and murine PKR were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), -actin from Sigma (St. Louis, MO), phospho-eIF2 and phospho-p53 from Cell Signaling (Beverly, MA), mouse p53 from Novocastra Laboratories (Newcastle, United Kingdom), human Mdm2 from Oncogene Research Products (San Diego, CA), mouse IFN regulatory factor 3 (IRF-3) from Zymed Laboratories (South San Francisco, CA), and glyceraldehyde-3-phosphate dehydrogenase from Chemicon International (Temecula, CA). Antibodies against p56 were a gift from Ganes Sen (Cleveland Clinic Foundation), and antibodies against mouse RNase L were a gift from Robert Silverman. 2-Aminopurine (2-AP), doxorubicin, and MG132 were purchased from Sigma. CP31398 was provided as a gift from Pfizer. Z-VAD-FMK was purchased from Calbiochem. Poly(I:C) was purchased from Sigma Aldrich and poly(I) from Amersham Pharmacia Biotech (Piscataway, NJ). Fugene reagent was purchased from Roche Diagnostics (Indianapolis, IN).
Poly(I:C) transfections. Poly(I:C) was transfected using Fugene reagent (Roche) in all experiments according to protocols provided by the manufacturer. Briefly, 2 μg of poly(I:C) per 3 μl of Fugene was incubated in 100 μl of serum-free DMEM for 15 to 30 min before being added to the supernatants of cells containing 10% FBS.
Western blotting and IP. Western blotting and immunoprecipitation (IP) were performed as described elsewhere (34). Briefly, cells were lysed in 50 mM Tris buffer, pH 7.4, containing 150 mM of NaCl, 50 mM of NaF, 10 mM of -glycerophosphate, 1% Triton X-100, 0.1 mM of EDTA, 10% glycerol, and protease/phosphatase inhibitors. The samples were kept on ice for 10 min, vortexed, and centrifuged for 15 min at 14,000 rpm; the supernatant was collected in a new tube; and protein concentrations were determined using the protein assay kit from Bio-Rad (Hercules, CA). Thirty micrograms of total protein was separated on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE), transferred to Immobilon-PSQ membranes (Millipore Corp., Belford, MA), and probed with the indicated antibodies. For the IP, 0.5 to 1.0 mg of total protein was incubated with a polyclonal antibody against p53 (FL-393; Santa Cruz) overnight at 4°C. The immune complex was precipitated using protein G-Sepharose beads (Amersham Biosciences). The beads were washed five times with lysis buffer, and the immunoprecipitate was separated on SDS-PAGE before Western blotting with the appropriate antibodies.
Metabolic labeling. Metabolic labeling of p53 was performed as described elsewhere (28). Briefly, cells were grown in DMEM without methionine for 2 h, and [35S]methionine (Amersham) was added to the culture medium for 30 min. The medium was removed, the monolayer was washed three times with phosphate-buffered saline (PBS), and fresh complete medium was added with or without poly(I:C). The whole-cell extracts were collected at the indicated times after the addition of fresh medium. p53 from 1 mg of total protein from the lysates was immunoprecipitated with a rabbit polyclonal antibody (FL-393; Santa Cruz). The immunoprecipitate was separated on a 10% SDS-PAGE, and isotopic incorporation into p53 was determined using a phosphorimager.
To measure protein synthesis, cells were treated for the indicated times, the medium was removed, and the cells were first washed with PBS and then incubated with DMEM containing [35S]methionine for 15 min before the extracts were collected. Fifty micrograms of total protein was precipitated with trichloroacetic acid (TCA), and the amount of isotope incorporated was measured using an LS 6000 IC Beta Counter (Beckman Coulter, Fullerton, CA).
Virus yield. Virus yield experiments were performed as described elsewhere (45). Briefly, cells were infected for 1 h with the indicated virus in DMEM without serum, the inoculum was removed, and new medium containing 10% FBS was added. After the indicated times, the supernatants were collected for titration of progeny viruses. Virus titers in the supernatants were determined by the Reed and Muench end point calculation method or by plaque assay. Virus yields were determined in L929 cells or CV1 cells. For the plaque assay, 0.5% methylcellulose (Sigma) was added to the medium in order to restrict viral spread.
RNA extractions and Northern blotting. For Northern blotting, total RNA was extracted from 10-cm dishes using the TRIZOL reagent according to the protocols provided by the manufacturer (Invitrogen, Carlsbad, CA). The RNA was separated by electrophoresis on a formaldehyde agarose gel and transferred to a Hybond-N membrane (Amersham). Hybridizations were performed as described elsewhere (8) using a probe containing the coding region for human p53 (kindly provided by Inder Verma, Salk Institute).
Lentivirus vectors. Lentivirus vectors expressing short hairpin RNAs (shRNA) have been described previously (4). Pools of HT1080 cells containing the lentivirus vectors were selected and used for the experiments described. Lentivirus vectors expressing human p53 or green fluorescent protein (GFP) have been described previously (13). Stocks of recombinant lentiviruses were prepared using the 293T cell line transfected with the lentivirus plasmids along with packaging plasmids encoding viral structural proteins and the G-protein of VSV using LipofectAMINE reagent (Invitrogen) as recommended by the manufacturer. The virus-containing medium from 293T cells was collected 48 h later. All infections with lentiviruses were performed in a medium containing 8 μg/ml of Polybrene for 6 h, after which the medium was removed and regular medium added.
Cell cycle analysis. For cell cycle analysis, cells were treated for the indicated times and fixed in 70% ethanol for 30 min at 4°C. After fixation, cells were stained with propidium iodide, processed using the Cellular DNA Flow Cytometric Analysis kit (Roche), and analyzed in a FACScan (Becton Dickinson, San Jose, CA).
TUNEL. For the terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) assay, cells were treated for the indicated times and fixed with 1% paraformaldehyde in PBS followed by 70% ethanol for terminal deoxynucleotidyltransferase-mediated detection of broken DNA ends using the APO-BRDU kit (BD Biosciences, Palo Alto, CA). The percentage of TUNEL-positive cells detected by FACScan analysis for each condition is plotted.
Caspase 3/7 activity. Cells were grown in black-wall 96-well plates with transparent bottoms and treated as indicated. All treatments were performed in triplicate, and the results shown represent the averages of the three wells. Caspase activity was measured directly from the 96-well plate using the Apo-ONE Homogeneous Caspase-3/7 assay according to the protocols provided by the manufacturer (Promega, Madison, WI).
RESULTS
Viral infection induces down-regulation of p53. To characterize the role of p53 in antiviral defense, different viruses were used to infect HT1080 and HepG2 cells, both of which express wild-type p53. p53 levels were monitored at different times postinfection (p.i.) (Fig. 1A and B). In the initial screen, four negative-sense RNA viruses, the rhabdovirus VSV and three paramyxoviruses (respiratory syncytial virus [RSV], Sendai virus, and HPIV3), and one positive-sense RNA virus, the picornavirus EMCV, were tested. The WR strain of the poxvirus VV, a large DNA virus, and the VV mutant VVE3L, which codes for a dsRNA binding protein capable of blocking dsRNA-mediated activation of PKR and IRF-3, were also included (42). Unexpectedly, infection with EMCV, HPIV3, Sendai virus, or VVE3L induced down-regulation of p53. In contrast, infection with VSV, RSV, or VV did not affect p53 levels (Fig. 1A and B). The observation that the mutant VVE3L induced down-regulation of p53 while WT VV did not suggested that the trigger for the down-regulation of p53 was most likely the dsRNA produced during viral replication (Fig. 1B; compare lanes 7 and 8). Although VSV has been shown to induce up-regulation of p53 in MEFs (38), we could not detect changes in p53 levels after VSV infection in either HT1080 or HepG2 cells (Fig. 1A and B). It is clear from these experiments that the role of p53 in the response to viruses is dependent on the cell type and the type of virus.
Influence of p53 on viral replication. p53 is an important mediator of stress responses, and previous observations have implicated p53 in antiviral defense against VSV (38). Conversely, the results shown in Fig. 1 indicate that virus infection can also induce down-regulation of p53. Therefore, we decided to investigate the role of p53 in the context of infection with viruses that induced down-regulation of p53. Some viruses, such as Sendai virus, RSV, and the VVE3L mutant, have a restricted host range and cannot grow in HT1080 cells (data not shown). VSV and EMCV have a rapid replicative lytic cycle in HT1080 cells, while HPIV3 has a slower replicative cycle. Since p53 is involved in the defense against VSV in MEFs (38), we used this virus in comparison to EMCV and HPIV3. Stable pools of HT1080 cells expressing shRNA targeting p53 were generated where p53 expression was selectively silenced compared to shRNA-GFP controls (Fig. 2A). When GFP was expressed, specific silencing was seen in the pools expressing the shRNA-GFP (Fig. 2B). Importantly, these shRNA did not activate the IFN system, as indicated by the expression of the IFN- and dsRNA-induced protein 56 (p56) (Fig. 2C). It has been reported that the H1 promoter, the one used in our vectors, is less likely to induce the IFN response (3, 12, 26). Despite the absence of p56 expression in unstimulated cells, the pools expressing the shRNA-GFP or shRNA-p53 still responded to dsRNA treatment by expressing p56 (27). Interestingly, HT1080 cells expressing the shRNA-p53 showed less accumulation of p56 in response to dsRNA (Fig. 2C). The different pools were infected with VSV, HPIV3, or EMCV for the indicated times, the supernatant was collected, and virus yields were determined (Fig. 2D). VSV, which did not induce down-regulation of p53, showed a 60-fold increase in virus yield in cells expressing shRNA-p53 compared to control cells (Fig. 2D). This is consistent with observations that, in p53 KO MEFs, VSV shows increased virus yield (38). Conversely, we observed 2.7-fold and 22-fold decreases in virus yields for EMCV and HPIV3, respectively, in pools expressing shRNA-p53 compared to control cells (Fig. 2C). To determine if the 2.7-fold difference in the yield of EMCV was consistent and to rule out possible problems associated with cell lines and in vitro manipulation of gene expression, virus yields from primary MEFs derived from p53 KO mice were also determined. Consistently, we observed a 3.4-fold reduction in EMCV yields from p53 KO MEFs compared to those for primary MEFs from WT mice (Fig. 2F). To further determine whether the absence of p53 alone was the cause of the difference between WT and KO cells, the KO MEFs were transduced with a lentivirus expressing human p53 prior to infection with EMCV. By Western blotting, it was observed that human p53 was successfully expressed in the KO cells (Fig. 2E). The expression of human p53, but not GFP, in the p53 KO MEFs partially restored the EMCV yield obtained in a concentration-dependent manner compared to control WT cells (Fig. 2F). Therefore, the down-regulation of p53 likely has a physiological role in controlling viral replication.
p53 levels are down-regulated in a dose- and time-dependent manner after dsRNA treatment. We then decided to test whether dsRNA produced during viral replication was the trigger for the down-regulation of p53. HT1080 cells, which not only retain WT p53 but also manifest a robust response following exposure to dsRNA (23, 27, 30), were transfected with poly(I:C), a synthetic analogue of dsRNA. As with viral infection, after transfection of poly(I:C), p53 protein levels were down-regulated in a time-dependent manner (Fig. 3A). As an indication of dsRNA treatment, we also observed accumulation of p56 following poly(I:C) treatment (Fig. 3A). The down-regulation of p53 was temporally coincident with the appearance of cleaved poly(ADP-ribose) polymerase (PARP) an indication of caspase activation and apoptosis (Fig. 3A). This raised the question of whether the down-regulation of p53 was a side effect of dsRNA-induced apoptosis. To rule this out, HT1080 cells were treated with poly(I:C) in the presence of Z-VAD-FMK, a general caspase inhibitor. The results (Fig. 3B) show that the down-regulation of p53 was not prevented by Z-VAD-FMK, although PARP cleavage and apoptosis were inhibited.
dsRNA treatment does not change the half life of p53. p53 protein levels are controlled mainly by the rate of degradation rather than synthesis (1). Therefore, we decided to determine whether dsRNA treatment affected the rate of p53 degradation. The half-life of p53 was first measured by treating cells with cycloheximide to block protein synthesis in the presence or absence of poly(I:C) treatment. Surprisingly, no difference was detected in the rate of decay of p53 protein levels with or without dsRNA treatment (Fig. 3C and D). The same results were seen using another chemical inhibitor of protein synthesis, emetine (data not shown). Because chemical inhibitors may have secondary effects, the half-life of p53 was also determined by metabolic pulse-labeling of HT1080 cells. Following labeling, the cells were washed and treated with poly(I:C) for different times, p53 was immunoprecipitated from cell extracts, and the amount of 35S incorporated into p53 was measured (Fig. 3E). Confirming the results obtained using protein synthesis inhibitors, no differences in the half-life of p53 with or without poly(I:C) treatment were detected.
Inhibition of the proteasome can block down-regulation of p53 in response to dsRNA. dsRNA treatment did not interfere with the rate of p53 degradation by the proteasome. However, it was still important to determine if basal activity of the proteasome was required for the down-regulation of p53. Accordingly, HT1080 cells were treated with increasing concentrations of the proteasome inhibitor MG132, and p53 levels were measured. Interestingly, even very low concentrations of MG132 (1 μM) that could barely induce changes in p53 basal levels in untreated cells had a strong inhibitory effect on the down-regulation induced by dsRNA (compare lanes 4 and 5 in Fig. 3F). It should be noted that at the highest concentration of MG132, 5 μM, in combination with poly(I:C), extensive cytotoxicity was observed, which could explain the smaller amount of -actin detected in Fig. 3F, lane 6.
dsRNA treatment does not affect the interaction between p53 and Mdm2. Mdm2 is the major E3 ubiquitin ligase controlling p53 degradation by the proteasome (1). The interaction between p53 and Mdm2 is regulated by posttranslational modifications of both proteins. To verify if dsRNA treatment could regulate the interaction between p53 and Mdm2, we performed co-IP experiments with HT1080 cells. Likely due to low levels of p53 in poly(I:C)-treated cells, we could not detect any association with Mdm2 (compare levels in whole-cell lysates [WCL] in Fig. 3H, lane 12, to lane 13 in Fig. 3G). For that reason, we used MG132 to block the down-regulation of p53 induced by dsRNA and then immunoprecipitated p53 from cells treated with MG132 alone or with poly(I:C) for 0.5, 1.0, 2.5, 5.0, and 7.0 h. As expected, MG132 induced accumulation of p53 and Mdm2 in WCL (Fig. 3H). However, poly(I:C) treatment did not consistently affect the interaction between p53 and Mdm2 (compare lanes 3 and 4, 5 and 6, 7 and 8, 9 and 10, 11 and 12 in Fig. 3G). Although there was somewhat less interaction between p53 and Mdm2 in poly(I:C)-treated cells, we also observed lower levels of p53 in the immunoprecipitate and in the WCL (Fig. 3G and H). Consistently, dsRNA treatment did not affect the half-life of p53 (Fig. 3D and E) or the interaction with Mdm2 (Fig. 3G).
Activation of PKR and RNase L pathways by dsRNA induces down-regulation of p53. dsRNA treatment can activate PKR and RNaseL pathways, resulting in protein synthesis inhibition (41, 44). To determine if the protein synthesis inhibition after treatment correlated with down-regulation of p53, HT1080 cells were treated with poly(I:C), and protein synthesis was measured by incorporation of [35S]methionine into newly synthesized proteins (Fig. 4A and C). The level of protein synthesis inhibition was dependent on the dose and the time after treatment and could be correlated with phosphorylation of eIF2 (Fig. 4A, B, C, and D). eIF2 is an in vivo substrate for PKR that, when phosphorylated, leads to inhibition of translation initiation (41). 2-AP, an ATP analogue capable of inhibiting PKR-induced eIF2 phosphorylation, showed dose-dependent inhibition of eIF2 phosphorylation in HT1080 cells treated with poly(I:C) (Fig. 4E). At the same time, 2-AP blocked the down-regulation of p53 in a dose-dependent manner (Fig. 4E). Since it has been shown previously that 2-AP can also inhibit pathways other than PKR that are involved in phosphorylation and stabilization of p53 (17), it is likely that, in our experiment, inhibition of PKR by 2-AP stabilizes p53 (Fig. 4E). We also monitored phosphorylation of p53 on serine 392, since PKR was shown to phosphorylate p53 on that residue in vitro. However, we did not detect any significant changes in phosphorylation of serine 392 after dsRNA treatment (Fig. 3H). To determine whether RNase L was activated following dsRNA treatment, we isolated total RNA from cells treated with poly(I:C). After separating the RNA on a denaturing gel, we observed clear rRNA cleavage, an indication of RNase L activation (44) (Fig. 4F). The same RNA was transferred to a nylon membrane, and Northern blotting was performed using a probe for p53. Following the same pattern of rRNA degradation, p53 mRNA levels were also down-regulated after dsRNA treatment, implicating RNase L in the down-regulation of p53 (Fig. 4F). To further confirm the requirements for PKR and RNase L, primary MEFs derived from PKR or RNase L knockout mice were transfected with poly(I:C). In both PKR and RNase L primary single-KO MEFs, down-regulation of p53 in response to dsRNA was partially impaired compared to that in the WT controls (Fig. 4G). To further verify the involvement of both PKR and RNase L pathways, primary MEFs derived from PKR/RNase L double-knockout (DKO) mice were transfected with poly(I:C). In DKO cells the down-regulation of p53 and protein synthesis inhibition in response to dsRNA were abrogated, indicating cooperation between the two pathways (Fig. 4H and I). We also observed that dsRNA treatment of DKO cells induced up-regulation of p53 due to the lack of protein synthesis inhibition in response to dsRNA (Fig. 4H). This is likely due to induction of IFN-/, consistent with the observations of Takaoka et al. (38). IFN production in WT cells also results in the up-regulation of PKR and RNase L, proteins known to be IFN inducible (Fig. 4H) (36). There is a high background in Western blots measuring p53 levels using whole-cell extracts from primary MEFs due to the very low expression levels of p53 and the lack of highly specific antibodies. Although the abilities of different antibodies to detect p53 in primary MEFs differed, down-regulation of p53 was observed with all of them (data not shown). We also confirmed that the antibodies were recognizing p53 by using extracts obtained from p53 KO cells or WT cells treated with doxorubicin as positive controls to increase the levels of p53 (Fig. 5D and data not shown). Moreover, when human p53 was overexpressed in primary MEFs, its down-regulation after dsRNA treatment could still be detected (Fig. 2E). In order to clearly illustrate the down-regulation of p53 levels in primary MEFs, we performed quantitation using ImageQuant 1.2 software and normalized the numbers obtained to -actin (represented as the p53/-actin ratio compared to that for untreated controls in Fig. 4G and H).
p53 knockout cells show increased sensitivity to dsRNA-induced apoptosis. We have shown that viral infection leads to down-regulation of p53 through inhibition of translation induced by dsRNA produced during viral replication. Moreover, in the absence of p53, certain viruses exhibit decreased replication, indicating that the down-regulation of p53 observed during viral infection might be important for the antiviral response. Therefore, we decided to address the mechanism. dsRNA is a trigger for apoptosis, which is one mechanism that infected cells use to limit viral replication (11). We observed a temporal correlation between the down-regulation of p53 and apoptosis in dsRNA-treated cells (Fig. 3A). However, the down-regulation of p53 is not caused by, and occurs independently of, apoptosis (Fig. 3B). Because of the known proapoptotic role of p53, we determined the cellular responses to dsRNA in the absence of p53. In order to address the mechanism by which the absence of p53 could influence the response of cells to dsRNA or viral infection, primary cultures of MEFs derived from WT and p53 KO mice were transfected with poly(I:C). Cell cycle analysis was performed by flow cytometry determining the DNA content of WT and p53 KO cells at different times. Following dsRNA treatment, WT cells show an early-G1/S arrest (increased G1/G2 ratio) that is not seen in p53 KO cells, indicating that it is p53 dependent (Fig. 5A). Interestingly, in WT cells, the down-regulation of p53 (see Fig. 4G) occurs after the G1 arrest and precedes the outcome of apoptosis (sub-G1 cells in Fig. 5B). p53 KO cells, which have no apparent early-G1 arrest, rapidly exhibit an increase in sub-G1 apoptotic cells as early as 8 h after treatment (Fig. 5A and B). To further verify the differences in cell death, apoptosis was measured by TUNEL staining after 16 h of treatment. MEFs derived from p53 KO mice treated with dsRNA exhibited about fourfold more apoptosis than wild-type controls treated similarly (Fig. 5C). To determine if the p53 KO cells were more sensitive not only to dsRNA but to any apoptotic stimuli, cells were treated with a combination of tumor necrosis factor (TNF) or untransfected poly(I:C) with cycloheximide, agents known to induce apoptosis (9). The results (Fig. 5C) confirmed that even in the presence of cycloheximide, p53 KO MEFs were more sensitive to dsRNA- but not TNF-induced apoptosis than WT MEFs. If down-regulation of p53 increases apoptosis, then in PKR and RNase L KO cells, apoptosis should be inhibited due to impaired down-regulation of p53. Indeed, PKR and RNase L single-KO cells were more resistant to various apoptotic stimuli (9, 44) and showed some degree of resistance to dsRNA-induced caspase activation (Fig. 5E). Since PKR/RNase L DKO cells did not show any down-regulation of p53 after dsRNA treatment, they were tested for their sensitivity to dsRNA-induced apoptosis. Compared to WT or PKR/RNase L single-KO MEFs, primary DKO MEFs were highly resistant to apoptosis induced by dsRNA, as measured by caspase 3/7 activation (Fig. 5E).
DISCUSSION
Although p53 has a clear role in genotoxic stress responses and maintaining genomic stability, our results demonstrate that its role in antiviral defense is complex. Previous studies have shown that oncogenic viruses target p53 to promote cell growth and that viruses such as VSV induce p53-dependent apoptosis (5, 38). We show that in the absence of viral mechanisms that activate or neutralize p53, dsRNA generated during infection will induce down-regulation of p53 as a default response. This does not appear to be a general mechanism used by all viruses, even though they are all likely to produce dsRNA at some stage during their replicative cycle. We observed that for both HPIV3 and EMCV, which induced down-regulation of p53 in infected cells, viral replication was reduced in the absence of p53. Some other viruses do not induce cell death and replicate at lower rates, establishing persistent infection (43).
dsRNA treatment did not induce any posttranslational modification of p53, nor were there changes in the protein half-life or association of p53 with Mdm2 (Fig. 3C, D, E, and G). We cannot completely rule out other possible mechanisms, but protein synthesis inhibition caused by activation of the PKR and RNase L pathways by dsRNA is required (Fig. 4). Therefore, translation inhibition prevents new p53 protein synthesis, leading to down-regulation of its levels, while the turnover of p53 through the proteasome pathway is unaffected (Fig. 3F).
dsRNA produced during viral infection is a trigger for down-regulation of p53, which in turn inhibits viral replication. There was a temporal coincidence between the outcome of apoptosis and the down-regulation of p53 (Fig. 3A). Cells can limit viral replication by undergoing apoptosis, and dsRNA produced during infection is a trigger for apoptosis (21). Furthermore, due to the role of p53 in apoptosis and in antiviral defense against VSV, we expected the absence of p53 to have severe consequences. Surprisingly, down-regulation of p53 was not a consequence of apoptosis but rather seemed to sensitize cells to death (Fig. 3B and 5C). Accordingly, apoptosis in response to dsRNA is not triggered by p53 and can occur in the absence of p53 function (39). Apoptosis in response to dsRNA seems to require IRF-3 and PKR, although the pathways are not clearly defined (39, 41). We observed that p53 KO cells displayed greater sensitivity to apoptosis than WT cells, implying that p53 is protective against dsRNA-induced cell death (Fig. 4B and C). Raj et al. (29) have shown that p53 induces cell cycle arrest at G2 and protects cells from adeno-associated virus-mediated cell death. In WT cells, we observed an increase in the G1/G2 ratio at 2 h after dsRNA treatment, suggesting G1 arrest that is p53 dependent, since it is absent in p53 KO cells (Fig. 5A). However, 8 h after treatment, the G1/G2 ratios in WT and p53 KO cells are similar, indicating that at least at later time points, p53 does not play a significant role in controlling the cell cycle (Fig. 5A). In WT cells we observed that the sub-G1 apoptotic cell population increases at later time points (8 to 24 h), when p53 levels are significantly lower (Fig. 5B). However, in p53 KO cells, a dramatic increase in sub-G1 cells can be detected as early as 8 h after treatment (Fig. 5B). This difference was confirmed by TUNEL staining of cells at 16 h after treatment (Fig. 5C). Thus, our data indicate that cells containing WT p53 undergo cell cycle arrest at G1 after dsRNA treatment and delay apoptosis until p53 levels are low enough to allow progression through the cell cycle. In the absence of p53, or after p53 is down-regulated, cells are released from the G1 arrest and undergo apoptosis. Although the p53-dependent G1 arrest after dsRNA treatment is likely to delay apoptosis, it is not clear if that is the sole mechanism for the increased viral replication in the presence of p53. As observed in Fig. 2C, HT1080 cells where p53 expression was ablated showed impaired gene induction in response to dsRNA, as previously shown for p53 KO MEFs (19). Most dsRNA-induced genes have an antiviral function, implying that reduced levels in p53-null cells should facilitate viral replication. Instead, we observed decreased viral replication for EMCV and HPIV3. It remains possible that some dsRNA-induced genes might be required for delaying apoptosis, thereby explaining the increased sensitivity in the absence of p53.
These observations link the down-regulation of p53 to the limitation of viral replication by increasing dsRNA-induced apoptosis. This provides an explanation of how the down-regulation of p53 in response to infection with EMCV, HPIV3, and possibly other viruses restricts viral replication. The magnitude of the difference in virus yield in the presence and absence of p53 depended on the replicative cycle of the virus. HPIV3 has a slower replicative cycle than EMCV and is more likely to have its replication impaired by apoptosis of the infected cell. Nevertheless, the results with EMCV can also be considered significant considering its rapid replicative cycle, since apoptosis is necessary but not sufficient to limit viral replication. Accordingly, MEFs derived from NF-B p50 KO mice, which display a three- to fourfold increase in apoptosis after infection with EMCV, show a mild decrease in virus yield at the same time postinfection (33). Furthermore, in control cells, p53 levels are down-regulated, which might dampen differences in virus yield obtained in the presence or absence of p53. Interestingly, Yeung et al. (43) have shown that ablating PKR expression leads to persistent EMCV infection and constitutive expression of p53 in U937 cells. This provides additional evidence for the mechanism we propose, where, after EMCV infection, PKR promotes down-regulation of p53, thus facilitating apoptosis and preventing establishment of persistent infection.
The complexity of the p53 response to viral infection is also supported by studies with HCT116 colon cancer cells. The presence or absence of p53 had no effect on cell death induced by several RNA viruses in HCT116 cells (18), although in the absence of p53, HCT116 cells were more effectively killed by adeno-associated virus infection (29). The absence of p53 in HCT116 cells resulted in increased EMCV yield in our experiments (J. T. Marques and B. R. G. Williams, unpublished results). However, HCT116 cells are very resistant to EMCV infection as well as to dsRNA-induced apoptosis and cannot limit virus replication by undergoing apoptosis. Therefore, the mechanism described here is probably not pertinent in these cells (Marques and Williams, unpublished).
Our results and the observations of others raise concerns about the use of viruses, either as oncolytic viruses or as gene therapy vectors, in the treatment of cancer. Virus vectors, usually replication-deficient viruses, still activate a stress response within the infected cell, which is often stronger than the response activated by wild-type viruses (31). Although we did not directly test any of the deficient viruses currently used in clinical trials, we tested two viruses that, in our model, can infect the cell but do not generate a new progeny. These two viruses, Sendai virus and VVE3L, induced down-regulation of p53 to different degrees (data not shown). Several different oncolytic viruses that are currently being tested include reoviruses, paramyxoviruses (Newcastle disease virus), herpesviruses, adenoviruses, and VSV (15, 20, 25, 37). These viruses are likely to activate antiviral responses, and in the case of reoviruses, they induce protein synthesis inhibition through RNase L and PKR (35). Although p53 can have an antiapoptotic role by inducing cell cycle arrest (29; also our results), there is no doubt about the antitumor and proapoptotic role of p53 in response to radio- and chemotherapies. Our results suggest that both oncolytic and deficient viruses that are used to deliver suicide genes to tumor cells may also inadvertently cause down-regulation of p53. Hence, cancer treatment strategies relying on viral infection may impair subsequent radio- and chemotherapies.
ACKNOWLEDGMENTS
We thank Anthony Sadler, Mark Whitmore, Michelle Holko, and Patricia Stanhope-Baker for helpful discussions, Curt Horvath for helpful discussions and sharing unpublished data, Cathy Shemo for help with the cell cycle analysis and TUNEL assays, and Joe Didonato, Amiya Banerjee, Ganes Sen, and Kristi Peters for providing reagents.
This work was supported by NIH grants RO1 AI34039 and PO1 CA 62220.
REFERENCES
Ashcroft, M., and K. H. Vousden. 1999. Regulation of p53 stability. Oncogene 18:7637-7643.
Balachandran, S., P. C. Roberts, L. E. Brown, H. Truong, A. K. Pattnaik, D. R. Archer, and G. N. Barber. 2000. Essential role for the dsRNA-dependent protein kinase PKR in innate immunity to viral infection. Immunity 13:129-141.
Bridge, A. J., S. Pebernard, A. Ducraux, A. L. Nicoulaz, and R. Iggo. 2003. Induction of an interferon response by RNAi vectors in mammalian cells. Nat. Genet. 34:263-264.
Budanov, A. V., A. A. Sablina, E. Feinstein, E. V. Koonin, and P. M. Chumakov. 2004. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of bacterial AhpD. Science 304:596-600.
Collot-Teixeira, S., J. Bass, F. Denis, and S. Ranger-Rogez. 2004. Human tumor suppressor p53 and DNA viruses. Rev. Med. Virol. 14:301-319.
Cuddihy, A. R., S. Li, N. W. Tam, A. H. Wong, Y. Taya, N. Abraham, J. C. Bell, and A. E. Koromilas. 1999. Double-stranded-RNA-activated protein kinase PKR enhances transcriptional activation by tumor suppressor p53. Mol. Cell. Biol. 19:2475-2484.
Cuddihy, A. R., A. H. Wong, N. W. Tam, S. Li, and A. E. Koromilas. 1999. The double-stranded RNA activated protein kinase PKR physically associates with the tumor suppressor p53 protein and phosphorylates human p53 on serine 392 in vitro. Oncogene 18:2690-2702.
da Fonseca, F. G., R. L. Silva, J. T. Marques, P. C. Ferreira, and E. G. Kroon. 1999. The genome of cowpox virus contains a gene related to those encoding the epidermal growth factor, transforming growth factor alpha and vaccinia growth factor. Virus Genes 18:151-160.
Der, S. D., Y. L. Yang, C. Weissmann, and B. R. Williams. 1997. A double-stranded RNA-activated protein kinase-dependent pathway mediating stress-induced apoptosis. Proc. Natl. Acad. Sci. USA 94:3279-3283.
Donehower, L. A., M. Harvey, B. L. Slagle, M. J. McArthur, C. A. Montgomery, Jr., J. S. Butel, and A. Bradley. 1992. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 356:215-221.
Everett, H., and G. McFadden. 1999. Apoptosis: an innate immune response to virus infection. Trends Microbiol. 7:160-165.
Fish, R. J., and E. K. Kruithof. 2004. Short-term cytotoxic effects and long-term instability of RNAi delivered using lentiviral vectors. BMC Mol. Biol. 5:9.
Gurova, K. V., J. E. Hill, O. V. Razorenova, P. M. Chumakov, and A. V. Gudkov. 2004. p53 pathway in renal cell carcinoma is repressed by a dominant mechanism. Cancer Res. 64:1951-1958.
Harvey, M., M. J. McArthur, C. A. Montgomery, Jr., J. S. Butel, A. Bradley, and L. A. Donehower. 1993. Spontaneous and carcinogen-induced tumorigenesis in p53-deficient mice. Nat. Genet. 5:225-229.
Hirasawa, K., S. G. Nishikawa, K. L. Norman, M. C. Coffey, B. G. Thompson, C. S. Yoon, D. M. Waisman, and P. W. Lee. 2003. Systemic reovirus therapy of metastatic cancer in immune-competent mice. Cancer Res. 63:348-353.
Hollstein, M., D. Sidransky, B. Vogelstein, and C. C. Harris. 1991. p53 mutations in human cancers. Science 253:49-53.
Huang, S., L. K. Qu, A. R. Cuddihy, R. Ragheb, Y. Taya, and A. E. Koromilas. 2003. Protein kinase inhibitor 2-aminopurine overrides multiple genotoxic stress-induced cellular pathways to promote cell survival. Oncogene 22:3721-3733.
Huang, S., L. K. Qu, and A. E. Koromilas. 2004. Induction of p53-dependent apoptosis in HCT116 tumor cells by RNA viruses and possible implications in virus-mediated oncolysis. Cell Cycle 3:1043-1045.
Hummer, B. T., X. L. Li, and B. A. Hassel. 2001. Role for p53 in gene induction by double-stranded RNA. J. Virol. 75:7774-7777.
Ikeda, K., T. Ichikawa, H. Wakimoto, J. S. Silver, T. S. Deisboeck, D. Finkelstein, G. R. Harsh IV, D. N. Louis, R. T. Bartus, F. H. Hochberg, and E. A. Chiocca. 1999. Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses. Nat. Med. 5:881-887.
Kibler, K. V., T. Shors, K. B. Perkins, C. C. Zeman, M. P. Banaszak, J. Biesterfeldt, J. O. Langland, and B. L. Jacobs. 1997. Double-stranded RNA is a trigger for apoptosis in vaccinia virus-infected cells. J. Virol. 71:1992-2003.
Lane, D. P., and L. V. Crawford. 1979. T antigen is bound to a host protein in SV40-transformed cells. Nature 278:261-263.
Leaman, D. W., A. Salvekar, R. Patel, G. C. Sen, and G. R. Stark. 1998. A mutant cell line defective in response to double-stranded RNA and in regulating basal expression of interferon-stimulated genes. Proc. Natl. Acad. Sci. USA 95:9442-9447.
Linzer, D. I., and A. J. Levine. 1979. Characterization of a 54K dalton cellular SV40 tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell 17:43-52.
Nemunaitis, J., I. Ganly, F. Khuri, J. Arseneau, J. Kuhn, T. McCarty, S. Landers, P. Maples, L. Romel, B. Randlev, T. Reid, S. Kaye, and D. Kirn. 2000. Selective replication and oncolysis in p53 mutant tumors with ONYX-015, an E1B-55kD gene-deleted adenovirus, in patients with advanced head and neck cancer: a phase II trial. Cancer Res. 60:6359-6366.
Pebernard, S., and R. D. Iggo. 2004. Determinants of interferon-stimulated gene induction by RNAi vectors. Differentiation 72:103-111.
Peters, K. L., H. L. Smith, G. R. Stark, and G. C. Sen. 2002. IRF-3-dependent, NF-B- and JNK-independent activation of the 561 and IFN- genes in response to double-stranded RNA. Proc. Natl. Acad. Sci. USA 99:6322-6327.
Qu, L., S. Huang, D. Baltzis, A. M. Rivas-Estilla, O. Pluquet, M. Hatzoglou, C. Koumenis, Y. Taya, A. Yoshimura, and A. E. Koromilas. 2004. Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3. Genes Dev. 18:261-277.
Raj, K., P. Ogston, and P. Beard. 2001. Virus-mediated killing of cells that lack p53 activity. Nature 412:914-917.
Rasheed, S., W. A. Nelson-Rees, E. M. Toth, P. Arnstein, and M. B. Gardner. 1974. Characterization of a newly derived human sarcoma cell line (HT-1080). Cancer 33:1027-1033.
Roth, M. D., Q. Cheng, A. Harui, S. K. Basak, K. Mitani, T. A. Low, and S. M. Kiertscher. 2002. Helper-dependent adenoviral vectors efficiently express transgenes in human dendritic cells but still stimulate antiviral immune responses. J. Immunol. 169:4651-4656.
Santos, C. R., F. M. Vega, S. Blanco, R. Barcia, and P. A. Lazo. 2004. The vaccinia virus B1R kinase induces p53 downregulation by an Mdm2-dependent mechanism. Virology 328:254-265.
Schwarz, E. M., C. Badorff, T. S. Hiura, R. Wessely, A. Badorff, I. M. Verma, and K. U. Knowlton. 1998. NF-B-mediated inhibition of apoptosis is required for encephalomyocarditis virus virulence: a mechanism of resistance in p50 knockout mice. J. Virol. 72:5654-5660.
Silva, A. M., M. Whitmore, Z. Xu, Z. Jiang, X. Li, and B. R. Williams. 2004. Protein kinase R (PKR) interacts with and activates mitogen-activated protein kinase kinase 6 (MKK6) in response to double-stranded RNA stimulation. J. Biol. Chem. 279:37670-37676.
Smith, J. A., S. C. Schmechel, B. R. Williams, R. H. Silverman, and L. A. Schiff. 2005. Involvement of the interferon-regulated antiviral proteins PKR and RNase L in reovirus-induced shutoff of cellular translation. J. Virol. 79:2240-2250.
Stark, G. R., I. M. Kerr, B. R. Williams, R. H. Silverman, and R. D. Schreiber. 1998. How cells respond to interferons. Annu. Rev. Biochem. 67:227-264.
Stojdl, D. F., B. D. Lichty, B. R. tenOever, J. M. Paterson, A. T. Power, S. Knowles, R. Marius, J. Reynard, L. Poliquin, H. Atkins, E. G. Brown, R. K. Durbin, J. E. Durbin, J. Hiscott, and J. C. Bell. 2003. VSV strains with defects in their ability to shutdown innate immunity are potent systemic anti-cancer agents. Cancer Cell 4:263-275.
Takaoka, A., S. Hayakawa, H. Yanai, D. Stoiber, H. Negishi, H. Kikuchi, S. Sasaki, K. Imai, T. Shibue, K. Honda, and T. Taniguchi. 2003. Integration of interferon-/ signalling to p53 responses in tumour suppression and antiviral defence. Nature 424:516-523.
Weaver, B. K., O. Ando, K. P. Kumar, and N. C. Reich. 2001. Apoptosis is promoted by the dsRNA-activated factor (DRAF1) during viral infection independent of the action of interferon or p53. FASEB J. 15:501-515.
Werness, B. A., A. J. Levine, and P. M. Howley. 1990. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 248:76-79.
Williams, B. R. 2001. Signal integration via PKR. Sci. STKE 2001:RE2.
Xiang, Y., R. C. Condit, S. Vijaysri, B. Jacobs, B. R. Williams, and R. H. Silverman. 2002. Blockade of interferon induction and action by the E3L double-stranded RNA binding proteins of vaccinia virus. J. Virol. 76:5251-5259.
Yeung, M. C., D. L. Chang, R. E. Camantigue, and A. S. Lau. 1999. Inhibitory role of the host apoptogenic gene PKR in the establishment of persistent infection by encephalomyocarditis virus in U937 cells. Proc. Natl. Acad. Sci. USA 96:11860-11865.
Zhou, A., J. Paranjape, T. L. Brown, H. Nie, S. Naik, B. Dong, A. Chang, B. Trapp, R. Fairchild, C. Colmenares, and R. H. Silverman. 1997. Interferon action and apoptosis are defective in mice devoid of 2',5'-oligoadenylate-dependent RNase L. EMBO J. 16:6355-6363.
Zhou, A., J. M. Paranjape, S. D. Der, B. R. Williams, and R. H. Silverman. 1999. Interferon action in triply deficient mice reveals the existence of alternative antiviral pathways. Virology 258:435-440.(Joao T. Marques, Dominiqu)