|Home | About | Journals | Submit | Contact Us | Français|
Several mammalian viruses have been shown to induce a cellular DNA damage response during replication, and in some cases, this response is required for optimal virus replication. However, nothing is known about whether a DNA damage response is stimulated by DNA viruses in invertebrates. Cell cycle arrest and apoptosis are two of the downstream effects of the DNA damage response, and both are stimulated by baculovirus infection, suggesting a possible relationship between baculoviruses and the DNA damage response. In the study described in this report, we found that replication of the baculovirus Autographa californica M nucleopolyhedrovirus (AcMNPV) in the cell line Sf9, derived from the lepidopteran insect Spodoptera frugiperda, stimulated a DNA damage response, as indicated by an increased abundance of the S. frugiperda P53 protein (SfP53) and phosphorylation of the histone variant protein H2AX. Stimulation of the DNA damage response was dependent on viral DNA replication. Inhibition of the DNA damage response prevented both the increase in SfP53 accumulation and H2AX phosphorylation and also caused a 10- to 100-fold reduction in virus production, along with decreased viral DNA replication and late gene expression. However, silencing of Sfp53 expression by RNA interference did not significantly affect AcMNPV replication or induction of apoptosis by a mutant of AcMNPV lacking the antiapoptotic gene p35, indicating that these processes are not dependent on SfP53 in Sf9 cells.
Eukaryotic cells constantly monitor the integrity of their genomes and rapidly respond to the presence of damaged DNA by activating DNA damage response pathways, which are initiated largely through the activation of two members of the phosphatidylinositol 3-kinase superfamily, ataxia telangiectasia mutated (ATM) and ATM and rad-3 related (ATR) (for a recent review, see reference 50). ATM is activated mainly in response to double-strand breaks in DNA, while ATR responds mainly to single-strand breaks and stalled replication forks. However, there is overlap in the downstream substrates of these kinases. ATM and ATR phosphorylate numerous effector proteins which function in cell cycle checkpoints, DNA repair, and stimulation of apoptosis. One of the many important substrates of ATM and ATR is the transcription factor P53, which regulates the expression of numerous genes that function in all three of these processes. Studies in Drosophila melanogaster have demonstrated that DNA damage responses, including the roles of ATM, ATR, and P53, share many similarities in insects and mammals (53).
There is increasing evidence that the ability to manipulate the DNA damage response and associated downstream pathways is crucial for the replication of many viruses. Several mammalian viruses stimulate DNA damage response pathways as a consequence of infection, and these viruses have in turn evolved mechanisms to manipulate these pathways for their own benefit by exploiting or actively inhibiting different parts of the pathways (reviewed in reference 7). For example, herpesviruses, papillomaviruses, and HIV-1 have been shown to activate ATM, and ATM signaling is important for virus replication (12, 21, 24, 34). However, one of the downstream targets of ATM, P53, is actively inhibited by many mammalian DNA viruses because of its involvement in inhibiting cell cycle progression and stimulating apoptosis in response to DNA damage. At this time, nothing is known about the roles of the DNA damage response or P53 in relation to replication of invertebrate DNA viruses.
Insect genomes examined to date contain single orthologs of p53 and do not contain recognizable orthologs of the related genes p63 and p73, which appear to have evolved specifically in the vertebrate lineage (46). The D. melanogaster p53 gene (Dmp53) is required for rapid apoptosis in embryos following treatment with ionizing radiation (5, 18, 40, 52). However, in both mammals and D. melanogaster, p53-independent pathways are also triggered by DNA damage (31, 58). Dmp53 does not appear to be required for cell cycle arrest following DNA damage, although it has been reported to be required for arrest following metabolic stress (30). P53 activation in mammals involves phosphorylation and stabilization of P53 protein, which is normally kept at low levels by ubiquitin-mediated degradation via the E3 ubiquitin ligase MDM2. As in mammals, activation of DmP53 appears to involve phosphorylation (6), but the question of whether and how P53 protein stability is regulated in insects is somewhat unclear, since there are no detectable orthologs of MDM2 in insects, and DmP53 levels have been reported to not increase following DNA damage in embryos (6). However, DmP53 stability was recently reported to be influenced by an E2 ubiquitin-conjugating enzyme, dRad6 (8).
Baculoviruses such as Autographa californica M nucleopolyhedrovirus (AcMNPV) have large (134 kbp in the case of AcMNPV) circular double-stranded DNA genomes and replicate in the nucleus of susceptible insect cells. Replication of viral DNA is required for expression of late and very late viral genes, which is initiated soon after viral DNA replication begins. The genome of AcMNPV has been shown to encode six proteins that are required for viral DNA replication in transient assays, and several additional viral proteins may be required for production of infectious genomes in infected cells (19, 25, 28, 39, 55). The mechanism of baculovirus DNA replication is not well understood. Although it was first suggested to be a rolling-circle mechanism, more recent data suggest that it may also involve a recombination-dependent mechanism, similar to that in herpesviruses (38). AcMNPV-infected cells contain short-lived fragments of viral DNA which are presumed to be replication intermediates and would mimic both double- and single-strand breaks (33). In addition, it has been known for many years that baculovirus replication induces high rates of recombination in cells, and all baculoviruses encode homologs of alkaline nuclease, a homolog of bacteriophage lambda red α exonuclease that is required for production of infectious viral DNA (45).
AcMNPV has been shown to cause cell cycle arrest at G2/M and S, depending on the cell cycle stage of the cell upon infection (4, 17). In addition to being required for late and very late gene expression, viral DNA replication also appears to be the main trigger for AcMNPV-induced apoptosis and shutoff of cellular protein synthesis, which are both also stimulated shortly after initiation of viral DNA replication (11, 20, 49). Although apoptosis is initiated by AcMNPV infection, the virus blocks apoptosis at a downstream point by expressing P35, an inhibitor of effector caspases (reviewed in reference 9).
A protein conserved in all known baculoviruses, Ac92 (also known as P33), was shown to bind to human P53 and enhance apoptosis stimulated by overexpression of human P53 (44), raising the question of whether baculoviruses modulate P53 function. Ac92 has sulfhydryl oxidase activity, is associated with viral envelopes, and is required for formation of infectious virus (27, 36, 59). However, the relevance of the interaction of Ac92 with human P53 is not currently understood.
A lepidopteran ortholog of p53, Sfp53, was recently identified in the fall armyworm Spodoptera frugiperda, a host of AcMNPV (16). SfP53 shares low but significant sequence similarity with DmP53, including conserved amino acids involved in DNA binding. Overexpression of SfP53 was shown to stimulate apoptosis, but unlike DmP53, levels of endogenous SfP53 protein were increased when cells of Sf9, a S. frugiperda cell line, were treated with the DNA-damaging agents UV or camptothecin (CPT) (16). This increased protein abundance is similar to what is seen for P53 in mammalian cells following DNA damage, but whether it is due to increased transcription or altered protein stability is not known.
It has been proposed that baculovirus infection may stimulate DNA damage response pathways because silencing AcMNPV genes involved in DNA replication prevents virus-induced apoptosis and shutoff of host protein synthesis, which are common features of DNA damage responses (49). However, there is no direct evidence supporting this hypothesis. In this report, we present evidence indicating that infection of Sf9 cells with AcMNPV induces a DNA damage response. Induction of this response was dependent on viral DNA replication. In addition, we found that inhibition of ATM and ATR signaling caused decreased viral DNA replication and late gene expression. Finally, although infection resulted in increased accumulation of SfP53 protein, we found that silencing Sfp53 expression did not affect either virus replication or virus-induced apoptosis, indicating that while AcMNPV stimulates a DNA damage response and this response is required for optimal replication, SfP53 does not appear to be required for AcMNPV replication or AcMNPV-induced apoptosis in Sf9 cells.
Sf9 cells (clonal isolate 9 derived from the S. frugiperda cell line IPLB-SF21-AE) were purchased from Invitrogen and maintained at 27°C in TC-100 medium (Invitrogen) supplemented with 10% fetal bovine serum (Atlanta Biologicals), penicillin G (60 μg/ml), streptomycin sulfate (200 μg/ml), and amphotericin B (0.5 μg/ml). The AcMNPV strain used as a representative of wild type was vAcWT-PG (59), which was constructed using the bacmid bMON14272 by insertion of polyhedrin (polh) and enhanced green fluorescent protein (egfp) genes. The p35 mutant virus vAcP35KO-PG was constructed by inserting polh and egfp genes (as described for vAcWT-PG ) into the bacmid vAcP35-KO (26). Bacmids were maintained in Escherichia coli strain DH10B. To produce virus, bacmid DNA was purified and transfected into Sf9 cells, and budded virus was harvested and passaged twice before titers in Sf9 cells were determined by 50% tissue culture infective dose endpoint dilution assays as previously described (43), using Sf9 cells for vAcWT-PG and BTI-TN-5B1-4 (Hi5) cells for vAcP35KO-PG. Infections were performed using a multiplicity of infection (MOI) of 10 PFU/ml unless otherwise indicated.
When indicated, cells were treated with 5 mM caffeine (Sigma), 5 μg/ml aphidicolin (Sigma), 15 to 25 μM KU-55933 (Selleck Chemicals), or 10 μM CPT (Sigma). Aphidicolin, KU-55933, and CPT were dissolved in dimethyl sulfoxide, while caffeine was dissolved in water. When combined with virus infection, chemicals were added to culture medium 1 h prior to addition of virus and maintained throughout infection.
Full-length open reading frame sequences were used as templates for PCR, with the T7 promoter sequence TAATACGACTCACTATAGGG added at each 5′ terminus. The resulting PCR products were used to synthesize double-stranded RNA (dsRNA) by in vitro transcription using an AmpliScribe T7 high-yield transcription kit (Epicentre) according to the protocol of the manufacturer. Sf9 monolayers (1.0 × 106 cells) were washed with unsupplemented Grace's insect medium (Invitrogen) lacking serum before transfection with a dsRNA-liposome mixture (equivalent to Lipofectin) as previously described (32), incubated for 5 h, washed twice with Grace's medium, and incubated with TC-100 medium supplemented with 10% fetal bovine serum (FBS). Time zero was set at the end of the 5-h incubation period. dsRNA at 150 and 100 μg/well was used to silence Sfp53 and p143, respectively.
Cells were collected and washed with phosphate-buffered saline, lysed in an equal volume of 2× Laemmli buffer, and incubated at 100°C for 5 min. Cell lysates were subjected to SDS-PAGE, and proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). The membranes were incubated with the following primary antibodies (diluted 1:3,000, unless indicated otherwise): mouse monoclonal anti-GP64 (eBioscience), rabbit polyclonal anti-VP39 (23), rabbit polyclonal anti-GP41 (42), rabbit polyclonal anti-VP1054 (41), rabbit polyclonal anti-SfP53 (16) (diluted 1:5,000), and mouse monoclonal anti-LEF-3 (diluted 1:2,000; obtained from Linda Guarino). They were then washed and incubated with horseradish peroxidase-conjugated secondary antibodies (Sigma). Signal development was conducted as described previously (16). All immunoblotting experiments were done multiple times using samples from at least three biological replicates, and representative results are presented.
Sf9 monolayers (1.0 × 106 cells) were infected at the indicated MOI. When RNAi was performed, infection was performed 24 h after dsRNA transfection. Cells were incubated with virus for 1 h on a rocking shaker, washed twice with 10% FBS-supplemented TC-100 medium, and cultured in 2 ml fresh medium in the presence of caffeine. The supernatants were collected at different time points, and titers were determined as described above.
Total RNA was isolated from cell samples using TRIzol (Invitrogen) and extracted with chloroform. The samples were centrifuged, and the RNA was precipitated with isopropanol. RNA pellets were washed with 75% ethanol, air dried, and resuspended in 30 to 50 μl RNase-free water. For each sample, 1.2 μg RNA was digested by DNase I, and then cDNA was synthesized using ImProm-II reverse transcriptase (RT; Promega) and oligo(dT) as a primer. Quantitative reverse transcriptase PCR (qRT-PCR) was performed using ABsolute QPCR SYBR green fluorescein mix (Thermo Scientific) with a Bio-Rad iCycler apparatus according to the protocol of the manufacturer. The real-time PCR data for viral genes were normalized to the data for β-actin as 2−ΔCT (where CT is the threshold cycle) (48). The primers used for PCR were as follows: vp39F, AACTTTTTGCGCAACGACTTTAT; vp39R, AGACGGCTATTCCTCCACCTG; polhF, ACCCGGCAAGAACCAAAAACTCACT; polhR, TCCAAGTTTCCCTGTAGAACTCTTTTCCTT; Sfp53F, GGGACATGGTGAAGGACGAGACG; Sfp53R, CAGCATAGTTTTGAGGCCGCATAGTG; actinF, AAGGCTAACCGTGAGAAGATGAC; and actinR, GATTGGGACAGTGTGGGAGAC.
To detect viral DNA replication, cells were infected with vAcWT-PG (MOI = 10) or vAcP35KO-PG (MOI = 1), treated with caffeine as described above, and harvested at various times postinfection, and total DNA was isolated with a Wizard genomic DNA purification kit (Promega). Using 50 ng of total DNA as template, quantitative PCR (qPCR) was performed using primers (gp41F, CGTAGTGGTAGTAATCGCCGC; gp41R, AGTCGAGTCGCGTCGCTTT) to amplify a 100-bp region of the AcMNPV gp41 gene (56). Preparation of standard DNA samples and analysis of real-time PCR results were done as previously described (59).
Caspase activity was determined using the mammalian caspase-3 substrate N-acetyl-Asp-Glu-Val-Asp-(7-amino-4-trifluoromethylcoumarin) (DEVD-AFC; MP Biomedicals) as described previously (16). Cell viability was determined by photographing four random fields of view at ×400 magnification at the indicated times and counting the number of cells with normal (nonapoptotic) morphology relative to untreated cells (set at 100%). To induce DNA damage, cells were treated with UV (254 nm) by placing tissue culture plates on a transilluminator for 8 min, or CPT was added at a final concentration of 10 μM.
Histones were isolated from Sf9 cells using an acid extraction procedure based on a previously published protocol (51). Cells were lysed in hypotonic buffer (10 mM Tris-Cl, pH 8.0, 1 mM KCl, 1.5 mM MgCl2, and 1 mM dithiothreitol) containing phenylmethylsulfonyl fluoride (1 mM) and 1 tablet protease inhibitor cocktail (Roche) added to 10 ml lysis buffer just before use. Cells were lysed for 30 min on a rotator, and intact nuclei were pelleted by spinning at 10,000 × g for 10 min at 4°C. Nuclei were resuspended in 0.4 N HCl, incubated on a rotator overnight at 4°C, and centrifuged at 16,000 × g for 10 min at 4°C, and the supernatant was incubated with trichloroacetic acid (final concentration, 33%) on ice for 30 min and centrifuged at 16,000 × g for 10 min at 4°C. Precipitated histones were washed twice with ice-cold acetone (followed each time by centrifugation at 16,000 × g for 5 min at 4°C), air dried for 20 min, and resuspended in water. The resulting histone fraction was incubated with Laemmli buffer, heated at 100°C for 5 min, and analyzed by SDS-PAGE. Proteins were transferred to a PVDF membrane using CAPS buffer (10 mM CAPS [Sigma], pH 11, 20% methanol). PVDF membranes containing transferred protein were blocked with 5% bovine serum albumin for 1 h, incubated with anti-phospho-histone H2AX polyclonal antibody (catalog number PA-40278; Pierce) diluted 1:5,000 overnight at 4°C, washed, and incubated with peroxidase-labeled goat anti-rabbit secondary antibody (Abcam) at a 1:5,000-fold dilution for 1.5 h. Signal development was performed as described previously (16).
Since many viruses inactivate P53 and AcMNPV encodes Ac92, which interacts with human P53, we examined the levels of SfP53 protein during AcMNPV infection by immunoblot analysis using an antiserum raised against recombinant SfP53. Interestingly, the levels of SfP53 protein increased in AcMNPV-infected Sf9 cells beginning at between 6 and 12 h postinfection (hpi) and continued to increase through 48 hpi (Fig. 1A), similar to what was observed following treatment with the DNA-damaging agent UV or CPT (Fig. 2B) (16). However, qRT-PCR analysis revealed that Sfp53 transcript levels did not increase relative to actin during infection, and both Spf53 and actin transcripts decreased in abundance beginning at 12 hpi, as indicated by increasing CT values (Fig. 1B). Treatment with DNA-damaging agents also did not significantly alter the level of the Sfp53 transcript relative to that of the actin transcript, with ΔCT values being less than 0.5 (Fig. 1C). This decrease in transcript levels for Sfp53 and actin during infection is consistent with what has been observed for many cellular transcripts following baculovirus infection, due to shutoff of host transcription that is dependent on viral DNA replication. Indeed, the decrease in Sfp53 and actin transcript levels during infection was blocked by aphidicolin, which blocks viral DNA replication (Fig. 1B). These results indicate that the mechanism responsible for increased SfP53 protein abundance following infection or DNA damage does not involve increased Sfp53 transcription or mRNA stability and may instead involve increased protein stability, similar to the increased stability of mammalian P53 that is observed following DNA damage.
Caffeine is widely used in studying DNA damage responses since it is an effective inhibitor of ATM and ATR (3). Importantly, the increase in SfP53 protein accumulation resulting from either AcMNPV infection or DNA-damaging agents was inhibited by treatment with 5 mM caffeine, a concentration that is commonly used to inhibit ATM and ATR (Fig. 2A and B). Caffeine also prevented the decrease in Sfp53 and actin transcript levels during infection, similar to aphidicolin (Fig. 1B), and prevented formation of viral occlusion bodies (data not shown). While caffeine was highly effective at inhibiting the response to either AcMNPV infection or UV, it was somewhat less effective at inhibiting the response to CPT, which apparently is a very strong signal at the concentration used (Fig. 2B and Fig. 3A). To verify the caffeine results, we treated infected cells with the specific ATM inhibitor KU-55933 (Fig. 2C). KU-55933 is a small-molecule inhibitor that specifically inhibits human ATM at 10 to 15 μM (15). It was necessary to increase the concentration of KU-55933 to 20 μM to see an inhibitory effect on SfP53 accumulation in Sf9 cells (Fig. 2C), possibly due to dissimilarities between human and lepidopteran ATM. Treatment with 20 or 25 μM KU-55933 also inhibited accumulation of the viral protein GP64, which is expressed mainly at late times (Fig. 2C).
To further verify that a DNA damage response was induced by AcMNPV, we examined phosphorylation of the histone protein H2AX. H2AX is a component of chromatin that is closely related to histone H2A but contains a slightly longer C terminus that includes a conserved SQ phosphorylation motif (60). Following DNA damage, H2AX is phosphorylated at this conserved serine by activated ATM and ATR, and this is widely used as a reliable marker for initiation of the DNA damage response in both mammals and Drosophila (29, 60). S. frugiperda encodes an ortholog of H2AX that contains the conserved C-terminal SQ phosphorylation motif (unpublished results). Immunoblotting with an antibody specific for the phosphorylated form of H2AX indicated that H2AX phosphorylation occurred in AcMNPV-infected Sf9 cells (Fig. 3A). As a positive control, cells were treated with CPT, which also induced H2AX phosphorylation (Fig. 3A). As expected, treatment with caffeine decreased H2AX phosphorylation stimulated either by AcMNPV infection or by CPT (Fig. 3A). Together, these results confirm that AcMNPV infection induces a bona fide DNA damage response in Sf9 cells.
The timing of increased accumulation of SfP53 and H2AX phosphorylation strongly suggested that the DNA damage response may be triggered by viral DNA replication, which begins at about 6 hpi in AcMNPV-infected Sf9 cells. To test whether viral DNA replication was involved in stimulating the DNA damage response, we sought a method to prevent viral DNA replication. Treatment with aphidicolin was not an option because it blocks elongation, not initiation, resulting in stalled replication forks which are themselves a trigger of the DNA damage response (22). We instead used RNAi to silence the expression of a viral gene that is required for initiation of viral DNA replication, the helicase gene (p143). While p143 mutants exhibit a complete lack of DNA replication (14), it has been reported previously that RNAi-mediated silencing of p143 reduced virus replication but did not block it completely, presumably due to incomplete silencing (49). We observed similar effects, as expression of two viral structural proteins, GP64 and VP39 (Fig. 3B), and formation of occlusion bodies (data not shown) were reduced but not completely eliminated by p143 dsRNA treatment compared to the levels of expression and formation for cells treated with control dsRNA. Treatment of infected cells with p143 dsRNA also reduced SfP53 accumulation compared to that in cells treated with negative-control dsRNA derived from the bacterial chloramphenicol acetyltransferase (cat) gene (Fig. 3B). This result, along with the timing of DNA damage induction, strongly suggests that viral DNA replication is necessary for triggering the DNA damage response.
In order to determine whether the DNA damage response plays a role in AcMNPV replication, cells were treated with caffeine or KU-55933 during infection and the effects on viral replication and gene expression were quantified. Treatment with either drug caused a delay in release of budded virus of about 12 h (Fig. 4). In addition, the final accumulation of budded virus in treated cells was reduced by approximately 10- to 100-fold (Fig. 4), indicating that the DNA damage response is important for full levels of viral replication. To further define the effect of caffeine on replication, the expression of selected viral genes and viral DNA replication were examined. Caffeine reduced the accumulation of transcripts expressed from the late viral structural genes gp41 and vp39 and the very late gene polh (Fig. 5A to C), as well as accumulation of late or very late viral proteins GP64, GP41, VP1054, VP39, and polyhedrin (Fig. 5D). To test whether caffeine affected expression of early genes, we examined levels of the early protein LEF-3. Levels of LEF-3 were reduced somewhat by caffeine; however, this reduction was similar to that seen when viral DNA replication was inhibited by silencing p143 (Fig. 5D). This reduction is likely due to a requirement for DNA replication (and the resulting increase in genome copy number) to achieve full levels of early gene expression. Treatment with caffeine also delayed and reduced the accumulation of viral DNA (Fig. 6A). Infected cells treated with caffeine had a 10-fold lower viral genome copy number than infected cells at 12 and 24 hpi, but this difference was reduced to 2-fold by 72 hpi. Thus, we conclude that caffeine does not have a significant direct effect on early gene expression. Instead, these results indicate that optimal viral DNA replication and late gene expression require the activity of ATM and/or ATR.
Infection of S. frugiperda cells with AcMNPV mutants that lack p35 causes high levels of caspase activity and apoptotic cell death (10), and triggering of apoptosis is dependent on viral DNA replication (49). Since the DNA damage response is also capable of triggering apoptosis, it was of interest to determine whether apoptosis induced by AcMNPV infection requires the DNA damage response. Caffeine treatment reduced the DNA replication of a mutant virus in which the p35 gene was deleted, vAcP35KO-PG (Fig. 6B), similar to its effect on vAcWT-PG. Caspase activity and apoptosis were also greatly reduced in cells infected with vAcP35KO-PG when the cells were treated with caffeine (Fig. 6C and D), consistent with the effect of caffeine on viral DNA replication. Thus, on the basis of these results, it is not possible to determine whether the DNA damage response is directly involved in apoptosis after it is stimulated by AcMNPV or whether it is required only for high levels of viral DNA replication and, thus, for stimulation of apoptosis by AcMNPV.
To look further into the relationship between the DNA damage response and virus replication and virus-induced apoptosis, we examined the role of SfP53 in virus infection. P53 is a downstream target of the ATM kinase, and P53 is required for rapid apoptosis induced by DNA damage in Drosophila embryos, although P53-independent apoptosis can also be induced by DNA damage (31, 58). Therefore, we examined whether SfP53 was required for AcMNPV replication or virus-stimulated apoptosis by silencing expression of Sfp53 using RNAi. Transfection of cells with Sfp53 dsRNA drastically reduced SfP53 protein levels for up to 72 h posttransfection, although some SfP53 protein was observed at between 72 and 96 h (Fig. 7A). To test the effect of silencing Sfp53 on virus replication, cells were transfected with dsRNA at 24 h prior to infection with vAcWT-PG. In cells transfected with control cat dsRNA, SfP53 protein levels increased during infection (Fig. 7B), similar to the finding in untransfected cells (Fig. 1A), while infected cells treated with Sfp53 dsRNA contained relatively low levels of SfP53, indicating successful gene silencing during infection (Fig. 7B). However, the reduced expression of Sfp53 did not have a significant effect on the timing or the final levels of virus replication at either a low or a high MOI (Fig. 7C and D).
To determine whether SfP53 plays a role in induction of apoptosis by AcMNPV, we silenced Sfp53 expression in Sf9 cells and then infected the cells with vAcP35KO-PG. Sfp53 expression was silenced during vAcP35KO-PG infection, as shown by immunoblotting (Fig. 8A). However, apoptosis still proceeded normally in vAcP35KO-PG-infected cells with silenced Sfp53 expression (Fig. 8B and C). In fact, caspase activity was consistently higher in cells treated with Sfp53 dsRNA than cells treated with control dsRNA (Fig. 8B). Similar results were observed in cells treated with UV or CPT; silencing Sfp53 expression had no effect on apoptosis stimulated by these DNA-damaging agents and resulted in higher caspase activity than in control cells (Fig. 9).
It is becoming increasingly clear that DNA damage response signaling plays an important role in the replication of many viruses. This study represents the first report of a virus stimulating the DNA damage response in invertebrate cells. Our results demonstrate that replication of the baculovirus AcMNPV stimulates signaling through DNA damage response pathways, similar to what has been reported for several DNA viruses that infect vertebrates. Stimulation of the DNA damage response is dependent on initiation of viral DNA replication; however, efficient levels of viral DNA replication in turn depend on the DNA damage response (Fig. 10). It is likely that intermediates formed during viral DNA replication (33) are interpreted by the cell as damaged DNA and that this is the trigger for the DNA damage response, although at this time we cannot rule out the possibility that a late gene product is responsible. One or more unknown cellular factors activated by the DNA damage response are presumably utilized by the virus to achieve full levels of DNA replication and, thus, late and very late gene expression, which depends on DNA replication. Although the identities of these cellular factors remain to be determined, they may include enzymes involved in DNA repair or recombination.
Using caffeine to inhibit DNA damage response signaling caused decreased viral DNA replication and late gene expression and also reduced virus-induced apoptosis. A similar dependence on DNA damage response signaling for optimal virus replication has been reported for several mammalian viruses, including the herpesviruses human cytomegalovirus (HCMV) (12) and herpes simplex virus type 1 (HSV-1) (24). Both of these herpesviruses stimulate the DNA damage response via immediate early gene expression (12, 24), which differs from our results with AcMNPV, where stimulation of the DNA damage response appears to be dependent on viral DNA replication. However, with both HCMV and HSV-1, abrogation of the DNA damage response inhibits viral DNA replication, similar to what we have observed with AcMNPV. In the case of both HSV-1 and HCMV, inhibition of the DNA damage response reduces the formation of mature viral replication centers, perhaps by inhibiting localization of DNA repair enzymes that are necessary for viral replication to the viral centers (7, 12, 24). It has been noted that DNA replication of baculovirus is similar in many ways to that of herpesviruses, as both appear to be dependent on recombination (45), and baculoviruses also form nuclear replication foci (37). Therefore, it will be interesting to determine whether a similar effect on formation of replication centers is observed in AcMNPV-infected cells that are treated with inhibitors of the DNA damage response.
The ability of caffeine to inhibit virus-induced apoptosis is probably due to the fact that caffeine inhibited viral DNA replication of the vAcP35KO-PG mutant, which is the main trigger for virus-induced apoptosis (49). On the basis of these data, we cannot determine whether the DNA damage response is also involved in apoptosis downstream in the pathway, after induction by viral DNA replication (Fig. 10). One of the main mechanisms by which the DNA damage response triggers apoptosis is through activation of P53, but silencing Sfp53 expression had no effect on apoptosis stimulated by either virus infection or treatment with UV or CPT. In Drosophila, DmP53 upregulates expression of the proapoptotic gene reaper after DNA damage (5). The Reaper protein functions as an inhibitor of apoptosis (IAP) antagonist and can trigger degradation of the protective IAP protein DIAP1 (54). S. frugiperda cells also express an IAP protein, SfIAP, that is required for cell viability (35). It has been shown that AcMNPV infection causes a decline in the levels of SfIAP protein (57). The decrease in SfIAP is also triggered by viral DNA replication and is partially due to proteosome-mediated degradation (57), but exactly how viral DNA replication triggers SfIAP degradation is not clear. Because of their interdependent nature, it is difficult to tease apart the processes of baculovirus DNA replication, the DNA damage response, shutoff of host gene expression, and apoptosis (Fig. 10). Thus, the exact role of the DNA damage response in AcMNPV-induced apoptosis is still unclear.
Silencing Sfp53 expression had no discernible effect on virus replication. This was not unexpected, since many viruses encode proteins that inactivate P53. P53 is only one of the downstream targets of ATM, and it is likely that other arms of the DNA damage response are responsible for boosting virus replication. However, although P53-independent apoptosis can occur in response to DNA damage (31, 58), it was somewhat surprising that silencing Sfp53 did not affect apoptosis stimulated by infection or by UV or CPT treatment. In fact, caspase activity was consistently higher in cells where Sfp53 was silenced before treatment. This higher caspase activity could potentially be explained by the ability of SfP53 to stimulate expression of one or more antiapoptotic genes. Similar results have been reported in mammalian cells lacking p53 and are thought to be due to the ability of P53 to stimulate expression of proteins involved in cell cycle arrest and DNA repair, such as the cyclin-dependent kinase inhibitor p21 (13), although p21 does not appear to be a transcriptional target of P53 in Drosophila embryos (1, 6). Regardless, the ability of apoptosis to occur normally, even though levels of SfP53 were drastically reduced by RNAi, suggests that SfP53 is not required for apoptosis stimulated by infection or the DNA-damaging agent UV or CPT in Sf9 cells. It is possible that SfP53 is normally involved in apoptosis triggered by these stimuli, but when SfP53 levels are reduced, other pathways are sufficient for apoptosis. It is also possible that the Sf9 cell line may have undergone genetic alterations that make it less dependent on SfP53 for apoptosis than normal for S. frugiperda cells. While we cannot rule out the possibility that silencing of Sfp53 did not reduce SfP53 levels enough to affect its function, we consider this unlikely since the kinetics of apoptosis were unaffected by the drastically reduced SfP53 levels (data not shown).
At this time, it is not clear whether the AcMNPV protein Ac92, which binds to human P53 (44), plays any role in regulating the function of SfP53 during AcMNPV infection. Although Ac92 was shown to enhance apoptosis stimulated by overexpression of human P53, this was done using a protein overexpression approach, the results of which can be difficult to interpret. Infection with AcMNPV stabilizes P53, as does infection with simian virus 40 (SV40). Interestingly, interaction with SV40 large T antigen stabilizes P53 but inhibits P53 function (2, 47). Although we have observed that Ac92 interacts with SfP53 (unpublished results), the question of whether Ac92 regulates SfP53 function needs to be further investigated.
Our results indicate that the increased abundance of SfP53 that is observed after triggering of the DNA damage response in Sf9 cells is not due to increased Sfp53 transcript levels. Instead, the accumulation of SfP53 is likely due to increased protein stability, as it is in mammalian cells, although this remains to be directly demonstrated. Although DmP53 has been reported to not increase in abundance following irradiation of embryos (6), an E2 ubiquitin-conjugating enzyme, dRad6, has been reported to affect the stability of DmP53 in Drosophila cell lines (8). The difference between our results and those reported in Drosophila may be due to using embryos versus established cell lines or to a species-specific difference.
In conclusion, we have shown that AcMNPV replication stimulates DNA damage response signaling in Sf9 cells, as demonstrated by overaccumulation of SfP53, H2AX phosphorylation, and the ability of inhibitors of ATM and ATR to affect these processes. In turn, optimal replication of AcMNPV depends on yet-to-be-determined factors stimulated by the DNA damage response. Although induction of apoptosis by AcMNPV depends on viral DNA replication, the downstream effector SfP53 is not required for AcMNPV-induced apoptosis in Sf9 cells. Further investigation will continue to unravel the role of the DNA damage response in AcMNPV replication and viral stimulation of apoptosis.
We thank Julie Olszewski, Lois Miller (deceased), and Linda Guarino for providing antibodies.
This research was supported by the Kansas Agricultural Experiment Station.
†This is contribution no. 12-020-J from the Kansas Agricultural Experiment Station.
Published ahead of print on 14 September 2011.