INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The ability of cells to commit suicide is critical for development, tissue homeostasis, and protection of the organism against pathogens, including viruses. Cells kill themselves in a controlled process known as programmed cell death or apoptosis (reviewed in reference 34). Generally, movement of phosphatidylserine from the inner to the outer surface of the cell membrane, activation of cysteine proteases (caspases), and modifications of key regulatory proteins, such as P53 or members of the Bcl-2 family, are early apoptotic events. Degradation of nuclear DNA, leakage of DNA to the cytoplasm due to disruption of the nuclear envelope, cytoplasmic blebbing, and shrinkage of the cell are late apoptotic markers.

The Bcl-2 family contains about 15 protein members in mammalian cells (reviewed in references 1 and 13). The family includes prosurvival proteins (Bcl-2 and Bcl-X_L) and proapoptotic proteins that induce apoptosis either directly (Bax and Bid) or by heterodimerization and titration of the function of prosurvival members (Bax, Bad, and Bcl-X_S). A well-established role for prosurvival members is to maintain mitochondrial integrity, and Bcl-2 prevents release of cytochrome c to the cytoplasm. Cytochrome c facilitates a change in the structure of the adapter molecule Apaf-1 to allow pro-caspase-9 recruitment and activation (1, 13). Apart from heterodimerization with proapoptotic proteins, transcriptional control of the gene, phosphorylation, proteolysis, or induction of conformational changes may inactivate Bcl-2 (reviewed in reference 8).

Caspases exist as latent zymogens and are activated by cleavage of their N-terminal prodomains. They are classified as apoptotic initiators (caspase-9), apoptotic executioners (caspase -1, -3, -4, -6, and -7), and cytokine processors (caspase-1). Once activated by apoptotic signals, initiators cleave and activate apoptotic executioners that target pro- and antiapoptotic structural and homeostatic proteins to systematically dismantle the cell (reviewed in reference 44).

In the case of virus-infected cells, induction of early cell death would severely limit virus replication, and many viruses have evolved strategies to avoid or delay apoptosis. In addition, some viruses actively induce apoptosis during the late stages of infection to facilitate progeny spread (reviewed in references in 20 and 42). Herpes simplex virus type 1 (HSV-1) is a DNA virus and a ubiquitous human pathogen (reviewed in reference 6). During the lytic cycle, viral gene expression can be divided into three temporal stages. Transcription of the five immediate-early (IE) genes is initiated by the virion tegument protein VP16 (Vmw65) in the absence of de novo protein synthesis. IE proteins Vmw175 (ICP4), Vmw63 (ICP27), Vmw110 (ICP0), and Vmw68 act to orchestrate the expression of early and late genes. ICP27 and ICP4 are essential proteins, and elimination of their respective genes blocks the viral replication cycle at early stages of infection in tissue culture. ICP4 transactivates viral gene expression through DNA binding (2), and ICP27 is a multifunctional protein involved in the export, 3' processing, and poly(A) usage of viral RNAs (24, 25, 37). Early gene products are detectable by 4 to 5 h postinfection and are mostly enzymes involved in DNA synthesis and replication. Late genes are efficiently expressed after 6 to 7 h postinfection and mostly encode structural proteins.

Wild-type (wt) HSV-1 suppresses apoptotic DNA fragmentation and cell death (19), and early events during wt HSV-1 infection are required for protection against apoptosis (4). Infection of cells prior to treatment with a variety of apoptotic stimuli protects cells from apoptosis, and the effect has some cell type dependency (9). The host cell apoptotic mechanism is activated during HSV-1 infection, but the virus has evolved mechanisms to suppress it (9). Certain HSV-1 proteins were recently proposed to have antiapoptotic functions: viral protein kinase US3 and glycoprotein J (17), IE proteins ICP4 (21), and ICP27 (3), and late protein ₁34.5 (14). With the exception of ₁34.5, which blocks interferon-induced protein synthesis shutoff by stimulating dephosphorylation of eukaryotic initiation factor 2, the antiapoptotic functions of these proteins are poorly understood. Significantly, loss of ICP4 was linked to mitochondrial dysfunction and DNA fragmentation (10), and loss of functional ICP27 was associated with activation of caspase-3 and DNA damage (4).

We report results from studies that were performed to investigate in detail the antiapoptotic role of wt HSV-1 and its ICP4 and ICP27 proteins. We used the replication-defective mutant viruses 27lacZ, which lacks ICP27 (39), and tsk, which expresses an inactive form of ICP4 at the nonpermissive temperature of 38.5°C (33), to dissect the apoptotic events triggered by the cell during mutant virus invasion and to identify the mechanisms used in wt virus infection to subvert these events. Two approaches were used. First, we examined the cellular apoptotic pathways triggered during infection of cells with the mutant viruses and compared these to the pathways triggered in mock- and wt HSV-1-infected cells. Second, we examined how preinfection of cells with wt HSV-1 rescues cells from apoptosis induced by cisplatin, a DNA-damaging compound that induces stress kinase-related apoptosis (47), and compared this to results with mock- and mutant virus-infected cells treated with the drug. Cisplatin was used as an apoptotic stimulant since, during the course of our studies, we noted many similarities between cisplatin-, tsk-, and 27lacZ-induced cell death. To control for any cell type-specific responses, as virus-induced apoptosis has been reported to have some cell type dependency (9), we performed experiments in BHK cells, which are commonly used for HSV-1 studies, and in human HeLa cells; results were similar for both cell lines.

Apoptotic pathways activated by cisplatin, 27lacZ, and tsk involved decreased Bcl-2 protein levels, activation of caspases, cytoplasmic cytochrome c release, DNA degradation, and cell death, and importantly, since its overexpression was protective, decreased Bcl-2 levels appeared to be a key element in the apoptotic process. Downregulation of Bcl-2 levels during infection with the mutant viruses involved three mechanisms: (i) decreased bcl-2 RNA levels, (ii) caspase-dependent degradation of Bcl-2, and (iii) decreased half-life of Bcl-2 protein. wt HSV-1 subverted all three of these mechanisms. wt HSV-1 also protected against cisplatin-induced apoptosis by stabilizing bcl-2 RNA and protein levels, and we propose that ICP4 and ICP27 play an antiapoptotic role in this stabilization of bcl-2 RNA.

In a previous study, we observed activation of the stress kinases Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinase (p38MAPK) by VP16 at 6 h postinfection of cells with wt HSV-1, 27lacZ, and tsk viruses (46). Activation of the stress kinases has been related to cell death (47) and prevention of apoptosis (35) in a cell type- and stimulus-dependent manner. Moreover, activation of cellular signalling pathways by HSV-1 has been proposed to facilitate virus replication (31, 46), and therefore a further aim of the present study was to investigate the role of p38MAPK activation during HSV-1 infection. We observed that inhibition of p38MAPK activity during infection of cells with 27lacZ and tsk decreased the levels of apoptosis triggered by the viruses and resulted in an elevated half-life of Bcl-2 protein. Furthermore, we showed that activation of p38MAPK during wt HSV-1 infection increased the viral yield and enhanced expression of specific viral genes, without resulting in p38MAPK-dependent destabilization of Bcl-2. We suggest that activation of p38MAPK during HSV-1 infection is exploited to promote viral replication and that wt, but not 27lacZ and tsk, virus prevents the activated p38MAPK from destabilizing Bcl-2 protein.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and viruses. Baby hamster kidney (BHK) and HeLa cells were grown as described previously (46). tsk virus expresses an inactive form of ICP4 at the nonpermissive temperature of 38.5°C (33). In the 27lacZ virus, ICP27 is inactivated by insertion of a lacZ cassette (39).

Proteins, plasmids, and antibodies. Purified Jun and ATF-2 proteins were from Insight Biotechnology, Wembley, United Kingdom. Plasmid SEK-AL, coding for the Ala-220 Leu-224 dominant-negative MKK4 (SKK1) mutant, was a gift from J. R. Woodgett (47). A plasmid coding for full-length human Bcl-2 cloned as a 5'-KpNI-NotI-3' fragment in pCEP4 and expressed under control of the cytomegalovirus promoter was provided by J. Pietenpol. Plasmids pIE1CAT and pIE3CAT contain IE110 and IE175 gene promoters, respectively, upstream of a chloromphenicol acetyltransferase (CAT) reporter gene (7, 40). For expression of ICP27, pCMV63 was made by inserting an EcoRI/BamHI fragment with the full open reading frame into the expression vector pCMV-10 (41). For expression of ICP4, plasmid p175 was provided by R. Everett (32). Polyclonal antibodies against JunD; p38MAPK; procaspase-1, -3, -4, -6, and -7; P53; FasL; Bad; and Bax and monoclonal antibody against Mdm-2 were from Insight Biotechnology. Anti-procaspase-3 also recognizes the cleaved caspase-3 form. Anti-Bcl-2 (C-2) monoclonal antibody (Insight Biotechnology) was raised against amino acids 1 to 205 of the human protein. Polyclonal anti-poly (ADP-ribose)-polymerase (anti-PARP) antibody was from Boehringer Mannheim, Lewes, United Kingdom, and anti-cytochrome c antibody (clone 7H8.2C12) was from PharMingen, San Diego, Calif. Anti-ICP27 antiserum H1113 was from the Goodwin Institute for Cancer Research, Plantation, Fla. Anti-ICP0 (monoclonal) and anti-ICP4 (polyclonal) antibodies were provided by R. Everett. Monoclonal antibodies against gC and UL42 were from A. McLean, and anti-R1 and R2 polyclonal antibodies are described in reference 5.

Virus infection, transfection, and treatment with apoptotic stimuli and inhibitors. Cell monolayers were infected with wt HSV-1 (strain 17+) or 27lacZ at multiplicity of infection of 10 PFU per cell and grown at 37°C in 5% CO₂. In experiments with the tsk virus, infected and mock-infected cells were grown at 38.5°C. BHK cells were grown in 60-mm-diameter dishes and transiently transfected using Lipofectamine (Gibco BRL, Paisley, United Kingdom) as instructed by the manufacturer. For induction of apoptosis, cells were incubated with 30 µg of daily made cisplatin [cis-platinum(II) diamminedichloride] (Sigma) per ml prior to harvesting. To protect cells from cisplatin-induced apoptosis, cells were infected for 5 h with wt HSV-1 prior to addition of the drug. To block JNK, cells were transfected with plasmid coding for a dominant-negative SKK1 (MKK4) mutant and treated with apoptotic stimuli or infected with wt HSV-1 at 30 h posttransfection. To block p38MAPK, cells were preincubated for 1 h with 30 µM SB203580 (Calbiochem), a specific inhibitor, prior to addition of virus and throughout infection. For inhibition of caspases, 25 µM Z-VAD-FMK (Calbiochem) was added to cells 1 h prior to and 6 h after treatment with cisplatin and/or virus.

DNA fragmentation and viability assays. For detection of cytoplasmic DNA, 10⁶ cells per sample were analyzed with a Cell Death Detection ELISA^PLUS kit (Boehringer Mannheim). For viability assays, 10⁶ cells per sample were analyzed by trypan blue exclusion. During infection, cells gradually detached from the monolayers. Viability was determined by pooling floating and adherent cells. Experiments were repeated at least three times.

Immunoprecipitation from total cell extracts. Cells (5 × 10⁶) were lysed in 400 µl of ice-cold radioimmunoprecipitation assay buffer (1× phosphate-buffered saline, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 50 mM phenylmethylsulfate, 100 mM sodium orthovanadate) by passage through a 21-gauge needle. After centrifugation at 900 × g at 4°C for 5 min, extracts were incubated with 2 µg of the desired antibody and 20 µl of protein G (Insight Biotechnology) under rotation at 4°C overnight.

Immunoblotting procedure. Total cell extracts were prepared as described previously (46). Antibodies for caspase-1, -3, -4, -6, and -7 and Bcl-2 were used at dilutions of 1:100. Antibodies against JunD, PARP, Bax, Bad, P53, Mdm-2, and FasL were used at 1:300. Antibodies for ICP27, ICP4, R1, R2, gC, and UL42 were used at 1:1,000. Anti-ICP0 was used at 1:5,000. One hundred micrograms of total protein per sample was used for immunoblotting, with the exception of caspases, PARP, and Bcl-2, for which 300 µg of protein per sample was analyzed.

Localization of cytochrome c. Cells (5 × 10⁶ per sample) were resuspended on ice in 150 µl of buffer A (20 mM HEPES, 10 mM KCl, 1.5 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 250 mM sucrose, 50 mM phenylmethylsulfonyl fluoride) and passed through a 26-gauge needle. Lysates were centrifuged for 10 min at 750 × g at 4°C; supernatants were transferred to new tubes and centrifuged at 10,000 × g at 4°C for 20 min. The supernatants from the second centrifugation represent the cytosolic fractions, and 100-µg portions of extracts were immunoblotted with anti-cytochrome c antibody (2 µg/ml).

Immunocomplex kinase assay. For the immunocomplex kinase assay, Bcl-2 or p38MAPK proteins were immunoprecipitated. The matrix was washed extensively (20 mM HEPES [pH 7.6], 50 mM NaCl, 2.5 mM MgCl₂, 0.1 mM EDTA) and resuspended in 30 µl of kinase buffer (20 mM HEPES [pH 7.6], 20 mM MgCl₂, 20 mM -glycerophosphate, 0.1 mM NaVO₃, 2 mM dithiothreitol, 0.1 µg of okadaic acid per ml, 0.125 mM [-³²P]ATP) with addition of 1 µg of purified Jun (specific substrate for JNK) or ATF-2 (substrate for both JNK and p38MAPK). Phosphorylation was allowed to proceed for 30 min at 25°C, and samples were electrophoresed by SDS-polyacrylamide gel electrophoresis and analyzed by autoradiography.

[³⁵S]methionine pulse-chase analysis of Bcl-2 degradation rate. Cells (5 × 10⁶) were infected and, at 7 h postinfection, were washed and cultured in L-methionine-free medium (Life Technologies, Paisley, United Kingdom) for 1 h. Cells were pulsed with 100 µCi of ³⁵S-labeled L-methionine per ml for 2 h, washed twice, and cultured in medium containing unlabeled L-methionine for 0 to 4 h. Bcl-2 protein was immunoprecipitated, and SDS-polyacrylamide gels were analyzed with a phosphorimager.

Quantification of bands. All quantifications were performed using the Phosphorimager Quantity One program (Bio-Rad, Hemel Hempstead, United Kingdom).

Measurement of viral yields and virus release assay. BHK cells (5 × 10⁶) were infected at 10 PFU per cell, and viral yields were measured as described previously (31). Measurements at each time point were performed in triplicate, and experiments were repeated three times. For the virus release assay, cells and medium were separated by centrifugation (300 × g for 5 min; Sorvall H6000A rotor), and the percent virus release was calculated by dividing the amount of infectious virus released into the medium by the total viral yield.

Northern blot analysis. Total RNA was purified using Trizol (Life Technologies) according to the manufacturer's protocol. [-³²P]ATP (Amersham Pharmacia Biotech) was used for labeling bcl-2 and -actin DNA probes. The Bcl-2 probe was derived by cleavage of the bcl-2 expression plasmid with KpnI-NotI.

CAT enzyme assay. HeLa cells (5 × 10⁶) were transfected with 3 µg of CAT reporter plasmid, and liquid scintillation counting assay was performed using ³H-labeled chloramphenicol and n-butyryl-coenzyme A (Promega), following the manufacturer's protocol. Each sample was analyzed in triplicate, and experiments were repeated three times.

Analysis of results. Quantified data is presented in the figures; unless otherwise stated each numerical value is the mean from three independent experiments performed in either duplicate or triplicate. The maximum difference between the average value and the separate values is expressed as percent standard deviation, and this is indicated on the graphs by the error bars.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Cisplatin and 27lacZ and tsk mutant viruses induce DNA degradation, activation of caspases, cytoplasmic release of cytochrome c, decrease in Bcl-2 protein levels, and cell death, and wt HSV-1 protects cells from cisplatin-induced apoptosis without inhibiting activation of caspases. To investigate the antiapoptotic role of ICP27 and ICP4 viral proteins, BHK and HeLa cells were infected with mutant viruses 27lacZ (at 37°C) and tsk (at the nonpermissive temperature of 38.5°C) and tested for markers of programmed cell death, including DNA degradation, caspase activation, cytoplasmic cytochrome c release, and alterations in the levels of key regulatory cellular proteins. In addition, the ability of wt HSV-1 and mutant viruses to protect against apoptosis induced by cisplatin was also examined. For protection against cisplatin-induced apoptosis, cells were preinfected with wt or mutant viruses for 5 h prior to addition of the drug, as this was the minimum time required for wt HSV-1 to confer full protection against cisplatin-induced DNA damage and apoptosis (not shown). Results were always compared to those for the relevant mock-infected controls, i.e., cells incubated at 37°C for cisplatin, wt HSV-1, and 27lacZ treatments or cells incubated at 38.5°C for tsk infections. The data presented in Fig. 1 are from BHK cells. Experiments with HeLa cells produced similar results.

View larger version (51K): [in this window] [in a new window]   FIG. 1.   Apoptosis triggered by cisplatin (30 µg/ml) and tsk (38.5°C) and 27lacZ viruses includes DNA degradation, activation of caspase-3 and -6, cytoplasmic release of cytochrome c, and decrease in cell viability. Cells (106) were mock infected or treated with these stimuli for 16 h, unless otherwise indicated. When both cisplatin and virus were added, cells were infected 5 h prior to addition of the drug. (A) Detection of cytoplasmic DNA by ELISA at 6 to 24 h posttreatment. Each value represents the mean ± standard deviation from three experiments performed with duplicate samples. (B) Left panel, cleavage of inactive pro-caspase-3 to active caspase-3 was detected by immunoblotting of total cell extracts. Right panel, Immunoblotting for pro-caspase-6. (C) Total cell extracts were immunoblotted for the intact (112-kDa) and cleaved (85-kDa) forms of PARP. (D) Cytosolic extracts were immunoblotted for cytochrome c. (E) Cell viability was measured by trypan blue exclusion in mock-infected cells and in cells treated with cisplatin and/ or mutant viruses.

View larger version (47K): [in this window] [in a new window]   FIG. 2.   Apoptosis triggered by cisplatin (30 µg/ml) and tsk (38.5°C) and 27lacZ viruses targets bcl-2 protein levels. Cells (106) were mock infected or treated with these stimuli for 16 h, unless otherwise indicated. When both cisplatin and virus were added, cells were infected 5 h prior to addition of the drug. (A) Total cell extracts were immunoblotted for Bcl-2 (upper panels) and JunD (lower panels). (B) Quantification of Bcl-2 levels presented in panel A. (C) Immunoblotting analysis of total cell extracts for Bcl-2 (upper panel) and JunD (lower panel) at several time points postinfection with tsk.

A decreased level of Bcl-2 protein is a key parameter during cell death induced by cisplatin, 27lacZ, and tsk, and Bcl-2 protects against apoptosis induced by cisplatin, tsk, and 27lacZ by acting upstream of cytoplasmic cytochrome c release and DNA degradation. To investigate the role of decreased Bcl-2 levels during programmed cell death triggered by cisplatin and 27lacZ and tsk viruses, we tested whether overexpression of Bcl-2 protein rescued cells from apoptosis induced by these stimuli. The data presented in Fig. 3 are from BHK cells, and experiments with HeLa cells produced similar results. Cells were transfected with 6 µg of bcl-2 expression plasmid or empty vector as a control. At 30 h posttransfection cells were treated with 30 µg of cisplatin per ml or infected with 27lacZ, tsk, or wt HSV-1 for an additional 16 h and were tested for DNA damage, cytoplasmic cytochrome c release, and decrease in cell viability (Fig. 3).

View larger version (45K): [in this window] [in a new window]   FIG. 3.   Overexpression of Bcl-2 protects cells from DNA damage, cytoplasmic cytochrome c release, and cell death induced by cisplatin (30 µg/ml) and tsk (38.5°C) and 27lacZ viruses. (A) Cells (5 × 106) were transfected with 6 µg of bcl-2 plasmid or empty vector for 30 h, and apoptosis was induced with the stimuli for additional 16 h. Detection of cytoplasmic DNA by ELISA in cells mock infected or treated with wt HSV-1, tsk, 27lacZ, or cisplatin in the presence of bcl-2 plasmid or empty vector is shown. (B) Cells (5 × 106) were transfected with 0 to 6 µg of bcl-2 plasmid, and the amount of transfected DNA was normalized to 6 µg with empty vector. At 30 h posttransfection, cells were mock infected, treated with cisplatin, or infected with tsk or 27lacZ for 16 h. Upper panel, cytosolic preparations were immunoblotted for cytochrome c. Lower panel, Total cell extracts were immunostained for Bcl-2. (C) Cell viability was measured by trypan blue exclusion in cells transfected with Bcl-2 or vector as for panel A and infected for 12 to 24 h with tsk (left panel) and 27lacZ (right panel).

Possible mechanisms for decreased Bcl-2 levels during infection with the mutant viruses. (i) Decreased bcl-2 RNA levels during infections with tsk and 27lacZ viruses and a role for ICP4 and ICP27 in maintaining bcl-2 RNA levels. Regulation of bcl-2 expression is a well-established mechanism to control Bcl-2 protein levels (8). We examined whether downregulation of Bcl-2 protein during treatment of cells with cisplatin, tsk, and 27lacZ is caused by a decrease in bcl-2 RNA. The data presented in Fig. 4 are from HeLa cells, and experiments with BHK cells produced similar results.

View larger version (36K): [in this window] [in a new window]   FIG. 4.   Wt HSV-1, but not tsk and 27lacZ, maintains bcl-2 RNA levels during infection and treatment of cells with cisplatin, and transfection of ICP4 and ICP27 increases bcl-2 RNA levels during treatment of cells with cisplatin. (A) Cells (5 × 106) were mock-infected; infected with wt HSV-1, tsk (38.5°C), or 27lacZ, or treated with cisplatin (30 µg/ml) for 16 h. When both cisplatin and virus were added, cells were infected 5 h prior to addition of the drug. Total RNA was analyzed by Northern blotting using bcl-2 (upper panels) and -actin (lower panels) DNA probes and visualized with a phosphorimager. (B) Quantitative analysis of Bcl-2 levels as shown in panel A. (C) Cells (5 × 106) were cotransfected with 3 µg of ICP4- and 3 µg of ICP27-coding plasmids or transfected with 6 µg of empty vector for 30 h and then incubated in the absence or presence of cisplatin (30 µg/ml) for 16 h. Total RNA was analyzed by Northern blotting using bcl-2 (upper panel) and -actin (lower panel) DNA probes and visualized with a phosphorimager. (D) Quantitative analysis of Bcl-2 levels as shown in panel C.

(ii) Infections with tsk and 27lacZ viruses cause caspase-dependent degradation of Bcl-2, and caspases act upstream of Bcl-2. Regulation of Bcl-2 levels by caspase-dependent degradation was previously reported (8). To investigate whether activated caspases act upstream of Bcl-2 downregulation and whether caspase-dependent degradation of Bcl-2 occurs during apoptosis induced by the mutant viruses, we examined the effects, on the apoptotic pathway and on Bcl-2 levels, of blocking caspase activation using the inhibitor Z-VAD-FMK. Z-VAD-FMK is an inhibitor of caspase-1, -3, -4, and -7, and since no activation of caspase-1, -4, and -7 was observed during apoptosis induced by the mutant viruses, the effects observed using Z-VAD-FMK are most probably caspase-3 dependent. However, we cannot exclude the possibility that other caspases were inhibited. The data presented in Fig. 5 are from BHK cells. Experiments with HeLa cells produced similar results.

View larger version (41K): [in this window] [in a new window]   FIG. 5.   Inhibition of caspases during infection with 27lacZ and tsk increases Bcl-2 levels and protects from cytoplasmic release of cytochrome c and DNA degradation. (A) Cells (106) were mock infected or infected with tsk (38.5°C) or 27lacZ in the presence or absence of Z-VAD-FMK. When Z-VAD-FMK was included, cells were treated with 25 µM Z-VAD-FMK 1 h prior to and 6 h after infection. Cells were harvested at 16 h postinfection, and extracts were immunoblotted for Bcl-2 (upper panel) and JunD (lower panel). (B) Quantification of Bcl-2 levels as shown in panel A. (C) Cells (106) were mock infected or infected with tsk for 16 h the presence or absence of Z-VAD-FMK, as described for panel A. Cytosolic fractions were immunoblotted for cytochrome c. (D) Cells (106) were mock infected or treated with cisplatin (30 µg/ml), tsk, 27lacZ, or wt HSV-1 in the presence or absence of Z-VAD-FMK, as described for panel A. Cytoplasmic DNA was detected by ELISA. Each value represents mean ± standard deviation from four independent experiments performed in duplicate.

(iii) Infections with tsk and 27lacZ viruses cause a p38MAPK-dependent decrease in the half-life of Bcl-2 protein. Destabilization of Bcl-2 via phosphorylation has been identified as a further mechanism of Bcl-2 regulation following specific stimuli (8). In addition, the p38MAPK stress kinase is reported to phosphorylate Bcl-2 in vitro (23), and p38MAPK is activated during infection of cells with 27lacZ and tsk viruses (46). We therefore examined whether activation of p38MAPK was linked to destabilization of Bcl-2 protein and whether Bcl-2 interacted with a p38MAPK-associated kinase during infections with the mutant viruses. The data presented in Fig. 6 are from BHK cells, and experiments with HeLa cells produced similar results. Cisplatin treatment failed to activate p38MAPK (not shown) and was not used in these experiments.

View larger version (36K): [in this window] [in a new window]   FIG. 6.   Infection of cells with 27lacZ but not with the wt virus induces a p38MAPK-dependent decrease in Bcl-2 half-life, and p38MAPK kinase activity coimmunoprecipitates with Bcl-2 in cells infected with tsk and 27lacZ. (A) Left panel [35S]methionine pulse-chase analysis of Bcl-2 degradation rates during infection with wt HSV-1. Cells (5 × 106) were infected with wt HSV-1 and at 7 h post-infection cells were labeled with [35S]methionine and either lysed (0 h) or cultured in medium containing unlabeled methionine prior to immunoprecipitation for Bcl-2. The gel was analyzed by autoradiography. Right panel, [35S]methionine pulse-chase analysis of Bcl-2 degradation rates during infection with 27lacZ. Cells (5 × 106) were infected with 27lacZ, and at 7 h postinfection cells were analyzed as described for panel A. (B) Quantification of Bcl-2 levels from [35S]methionine pulse-chase analysis during infection with wt HSV-1 and 27lacZ as described for panel A. (C) Left panel, [35S]methionine pulse-chase analysis of Bcl-2 degradation rates during infection with 27lacZ in the presence of DMSO. Cells (5 × 106) were treated with 30 µM DMSO for 1 h and infected with 27lacZ in the continuous presence of DMSO. At 7 h postinfection cells were labeled with [35S]methionine, and Bcl-2 levels were analysed as described for panel A. Right panel, [35S]methionine pulse-chase analysis of Bcl-2 degradation rates during infection with 27lacZ in the presence of SB203580. Cells (5 × 106) were treated with 30 µM SB203580 for 1 h and infected with 27lacZ in the continuous presence of SB203580. At 7 h postinfection cells were labeled with [35S]methionine, and Bcl-2 levels were analyzed as described for panel A. Preimmune serum (pre-imm.) was used at 0 h postlabeling. (D) Quantification of Bcl-2 levels from [35S]methionine pulse-chase analysis during infection with 27lacZ in the presence of DMSO or SB203580, as described for panel C. (E) Cells (106) were pretreated with 30 µM SB203580 or DMSO for 1 h and mock infected or infected with tsk in the continuous presence of SB203580 or DMSO. In the upper panel, Bcl-2 was immunoprecipitated from total cell extracts and associated p38MAPK kinase activity was detected in immunocomplex kinase assays, using 1 µg of ATF-2 protein as the substrate. In the lower panel, total cell extracts from duplicate preparations similar to the ones analyzed in panel E, upper panel, were immunoblotted for Bcl-2. (F) Cells (106) were pretreated with 30 µM SB203580 or DMSO for 1 h, and mock infected or infected with 27lacZ, and analyzed as described for panel E. Upper panel, immunocomplex kinase assay for p38MAPK activity. Lower panel, immunoblotting of duplicate preparations for Bcl-2.

Inhibition of p38MAPK increases Bcl-2 levels and rescues cells from DNA damage during infection with the mutant viruses. To determine whether p38MAPK-dependent destabilization of Bcl-2 had any effect on the levels of apoptosis induced by the mutant viruses, we examined the results of inhibiting p38MAPK activity on downregulation of Bcl-2 levels and induction of DNA damage (Fig. 7). The data presented are from BHK cells. Experiments with HeLa cells produced similar results.

View larger version (28K): [in this window] [in a new window]   FIG. 7.   Inhibition of p38MAPK during infection with 27lacZ and tsk increases Bcl-2 levels and protects cells from DNA degradation, and activation of p38MAPK is independent of caspases. (A) Cells (106) were pretreated with 30 µM SB203580 or DMSO for 1 h and mock infected or infected with 27lacZ, tsk (38.5°C), or wt HSV-1. At 16 h postinfection, immunoblotting analysis of total cell extracts for Bcl-2 (upper panel) and JunD (lower panel) was performed. (B) Cells (106) were pretreated with 30 µM SB203580 or DMSO for 1 h and mock infected or infected with 27lacZ, tsk (38.5°C), or wt HSV-1 in the continuous presence of SB203580 or DMSO. At 16 h postinfection, cytoplasmic DNA was detected by ELISA. (C) Cells (106) were mock infected or infected with tsk in the presence or absence of Z-VAD-FMK. When Z-VAD-FMK was included, cells were treated with 25 µM Z-VAD-FMK 1 h prior to and 6 h after infection. At 8 h postinfection, p38MAPK was immunoprecipitated and immunocomplex kinase assays were performed using 1 µg of ATF-2 as the substrate. (D) Cells (106) were pretreated with 30 µM SB203580 or DMSO for 1 h and mock infected or infected with tsk (38.5°C) in the continuous presence of SB203580 or DMSO. At 16 h postinfection, total cell extracts were immunoblotted for caspase-3.

Activation of p38MAPK is independent of caspases. Both p38MAPK and caspases act upstream of Bcl-2 during apoptosis induced by the mutant viruses. We tested whether caspase activation and stimulation of p38MAPK were interdependent events.

Inhibition of p38MAPK kinase during wt HSV-1 infection decreases viral yield. Our results suggest that activation of the stress kinase p38MAPK during infection with the mutant viruses is part of an antiviral apoptotic response. However, we wished to determine whether activation of p38MAPK and of the closely related JNK stress kinases by wt HSV-1 (46) also serves some beneficial purpose for the virus. To test this hypothesis, viral yields were measured at 0-32 h postinfection of control BHK cells with wt HSV-1. BHK cells were preincubated for 1 h with either 30 µM DMSO (control) or SB203580 (p38MAPK inhibitor) and then infected with 10 PFU of wt HSV-1 per cell in the continuous presence of DMSO or SB203580. The yields at 18, 24, 28, and 32 h postinfection are given in Table 1. The titer of the virus increased in both experiments between 18 and 32 h, reaching a maximum by 28 h postinfection. At 32 h postinfection, the virus titer dropped slightly, presumably because most of the cells were lysed and some degradation of the virus occurred. Inhibition of p38MAPK resulted in decreases of 33% ± 3% in the yield of wt HSV-1 virus at 24 h postinfection, 58% ± 10% at 28 h postinfection, and 55% ± 6% at 32 h postinfection compared to that for the control wt HSV-1-infected cells treated with DMSO, and all of these differences were statistically significant as confirmed by the ² test (P < 0.05).

Activation of p38MAPK during wt HSV-1 infection enhances expression of specific viral gene promoters. Decreased viral yields in the presence of the SB203580 inhibitor might have resulted from inhibition of the release of assembled viral particles. Alternatively, inhibition of p38MAPK might reduce expression of all or specific viral proteins during infection with wt HSV-1. To test the first possibility, virus release assays were performed at 18, 24, 28, and 32 h postinfection of cells with wt HSV-1 in the presence of SB203580 or DMSO. No difference in virus release was observed at any of these time points (not shown).

View larger version (41K): [in this window] [in a new window]   FIG. 8.   Inhibition of p38MAPK decreases levels of viral proteins and transcriptional activity of specific promoters. (A) Cells (106) were incubated with 30 µM SB203580 or DMSO for 1 h and infected with wt HSV-1 for 6 to 12 h in the continuous presence of SB203580 or DMSO. Immunoblotting of total cell extracts was performed for viral proteins ICP0 (upper panel) and ICP4 (lower panel) at the time points indicated. (B) Cells (106) were treated with 30 µM SB203580 or DMSO as described for panel A and infected with wt HSV-1. At 9 h postinfection, total cell extracts were analyzed for viral proteins ICP27 (upper panel) and UL42 (lower panel) by immunoblotting. (C) Cells (5 × 106) were transfected with 3 µg of pIE1CAT or pIE3CAT. At 30 h posttransfection, cells were pretreated for 1 h with DMSO or SB203580 and infected with wt HSV-1 in the continuous presence of DMSO or SB203580. CAT activity was measured at 16 h postinfection.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we examined the apoptotic events triggered during infection of BHK and HeLa cells with the mutant viruses tsk and 27lacZ, which lack functional ICP4 and ICP27 proteins, respectively. Experiments with both cell lines produced similar data. Our results demonstrated that these mutant viruses induce an apoptotic pathway involving activation of caspases-3 and-6, activation of the p38MAPK stress kinase, decreases in Bcl-2 RNA and protein levels, release of cytochrome c to the cytoplasm, degradation of DNA, and cell death. Bcl-2 acts upstream of cytochrome c and DNA degradation. The caspases and p38MAPK act independently, and both act upstream of Bcl-2 protein.

We also demonstrated that the decrease in Bcl-2 levels is a key element in the apoptotic process and that downregulation of Bcl-2 protein during infection with the mutant viruses was mediated by three distinct mechanisms: (i) decreases in bcl-2 RNA levels, (ii) caspase-related degradation of Bcl-2, and (iii) p38MAPK-dependent decreases in the Bcl-2 half-life. wt HSV-1 evades apoptosis during infection by subverting all three of these mechanisms that target Bcl-2. Thus, activation of caspases and a decrease in bcl-2 RNA during infection of cells are prevented, and although p38MAPK is activated during wt HSV-1 infection (46), this does not alter the Bcl-2 half-life. Indeed, induction of p38MAPK activity during infection of cells with the wt virus increased viral yields by enhancing transcriptional activation of specific viral gene promoters and accumulation of their respective proteins. Cisplatin induces an apoptotic pathway that has a number of similarities to mutant virus-induced cell death, with decreased Bcl-2 RNA and protein levels, activation of caspase-3 and -6, cytoplasmic cytochrome c release, and DNA degradation. wt HSV-1 protected cells from cisplatin-induced apoptosis and stabilized bcl-2 RNA and protein levels without preventing activation of caspases.

Both tsk and 27lacZ infection failed to prevent cisplatin-induced apoptosis, suggesting a protective role for ICP4 and ICP27 during wt virus infection. Transfection of cells with both proteins was necessary to maintain Bcl-2 RNA and partially prevent cisplatin-induced apoptosis. However, we cannot rule out completely the involvement of other viral proteins. In addition, tsk and 27lacZ trigger activation of caspases and destabilization of Bcl-2 by activated p38MAPK during infection. Whether ICP4 and ICP27 play an antiapoptotic role in these two mechanisms is under investigation. It is possible that other viral early or late proteins that are not expressed in the ICP4 and ICP27 viral mutants play a role in evading caspase activation and p38MAPK-dependent destabilization of Bcl-2 during wt HSV-1 infections.

We provide evidence for mechanisms used by wt HSV-1 to circumvent and exploit the apoptotic host cell response to ensure and facilitate viral propagation that occur as early as 6 h postinfection. These are summarized schematically in Fig. 9. The only previously known apoptotic mechanism against HSV-1 involved host cell protein synthesis shutoff and the late viral protein ₁34.5 (14).

View larger version (22K): [in this window] [in a new window]   FIG. 9.   Schematic model of the mechanisms used by wt HSV-1 to block Bcl-2-mediated apoptosis. Apoptotic stimuli (for example, infection of cells with tsk or 27lacZ mutant viruses) target the levels of the pro-survival protein Bcl-2 by three mechanisms (route I, decrease of Bcl-2 RNA; route II, caspase-dependent decrease of Bcl-2 protein levels; route III, p38MAPK-dependent destabilization of Bcl-2 protein). wt HSV-1 blocks all of these mechanisms to prevent apoptosis. In addition, wt HSV-1 benefits from activation of p38MAPK by enhanced expression of specific viral genes and improved viral growth (route IV).

Infection of cells with 27lacZ or tsk mutant viruses triggers an apoptotic response that targets the key regulatory protein Bcl-2 through three cellular pathways (Fig. 9). In the first pathway (pathway I), the bcl-2 RNA levels (step I₁) and the Bcl-2 protein levels (step I₂) are decreased. wt-HSV-1 blocks step I₁ by maintaining cellular bcl-2 RNA levels, and viral proteins ICP4 and ICP27 are indispensable for that function. Numerous studies have indicated transcriptional regulation of Bcl-2, for example, by survival factors (28). Viruses commonly exploit the antiapoptotic function of Bcl-2 by encoding proteins that mimic its effect (20, 22) but only Epstein-Barr virus latent membrane protein 1 has been reported to stimulate cellular Bcl-2 expression (15).

The mechanism used by ICP4 and ICP27 to maintain bcl-2 RNA is under investigation. ICP4 acts as a transcription factor to transactivate viral genes (2). ICP27 regulates transcription of HSV-1 genes (36), modulates 3' RNA processing and poly(A) usage (24, 25), and is involved in the export of HSV-1 intronless RNAs (37). Furthermore, ICP27 interacts with ICP4 (30) and with host cell factors, including casein kinase 2 and the multifunctional heterogeneous nuclear ribonucleoprotein K, which is linked to transcription, RNA processing, transport, and translation (43). It is therefore possible that ICP27 and ICP4 regulate bcl-2 RNA levels through multiple mechanisms.

Second, the apoptotic mechanisms induced by the mutant viruses involved a caspase-dependent decrease in Bcl-2 levels (Fig. 9, pathway II). Infection of cells with 27lacZ or tsk viruses resulted in activation of caspase-3 and -6 (step II₁). Inhibition of caspase activity by Z-VAD-FMK during infection with the mutant viruses partially protected against decreased Bcl-2 levels (step II₂) and subsequent cytoplasmic cytochrome c release, degradation of nuclear DNA, and cell death, thus suggesting caspase-dependent degradation of Bcl-2 protein. Z-VAD-FMK is an inhibitor of caspase-1, -3, -4, and -7 but not of caspase-6. Since caspase-1, -4, and -7 were not activated during apoptosis triggered by cisplatin and the mutant viruses, it is probable that the Z-VAD-FMK effects are caspase-3 dependent. wt HSV-1 inhibits this apoptotic pathway by blocking the caspase-dependent decrease in Bcl-2 levels (step II₂).

Inactivation of the antiapoptotic function of Bcl-2 by caspase-3 cleavage at amino acids 31 and 34 has been reported (12). Cleavage removes the N-terminal BH4 region known to be essential for the death-protective activity of Bcl-2. In our study, we failed to observe characteristic products of Bcl-2 cleavage at these specific residues. It is possible that after specific cleavage of Bcl-2 by caspases, rapid degradation of the protein by noncaspase proteases occurs during infection of cells with the mutant viruses.

Our results are in agreement with previous observations showing that inhibition of caspases failed to protect against apoptosis induced by a virus that lacked functional ICP4 and Us3 genes (9) and that wt HSV-1 protects from apoptosis triggered by osmotic shock and heat without preventing caspase-3 activation (10). Z-VAD-FMK rescues Bcl-2 levels during infection with 27lacZ and tsk viruses by approximately 20 to 30%. This level of protection, although highly reproducible, was considerably lower than that obtained by transfection of cells with bcl-2. This implies that the caspase-dependent decrease in Bcl-2 levels plays a supplementary role in induction of apoptosis during infections with the mutant viruses and may explain why the wt virus does not need to block caspase activation to protect against cisplatin-induced apoptosis. Additional mechanisms could be used by wt HSV-1 to compensate for the caspase-dependent decrease in Bcl-2, for example, by increased translation efficiency of the stabilized bcl-2 RNA.

Activation of caspases during infection with tsk and 27lacZ viruses may result from cytoplasmic release of cytochrome c induced by decreased bcl-2 RNA levels and destabilization of Bcl-2 protein by p38MAPK (see below). Cleavage of Bcl-2 by caspases may induce additional release of cytochrome c and further promote caspase activation as part of a positive feedback loop (18). Alternatively, our results favor activation of caspases through a cytochrome c-independent mechanism, since stabilised expression of Bcl-2 decreased cytosolic cytochrome c but not activation of caspase-3 and wt HSV-1 abolished cytochrome c release caused by cisplatin but not caspase-3-dependent PARP cleavage. Furthermore, stimulation of stress kinases in our model (Fig. 9) is independent of caspase activation, in contrast to other apoptotic systems reported previously (29).

The third apoptotic pathway induced during infection with the mutant viruses resulted in a p38MAPK-dependent destabilization of Bcl-2 protein (Fig. 9, pathway III). The p38MAPK stress kinase is activated during infection with tsk and 27lacZ mutant viruses (46) (step III₁). Activated p38MAPK coimmunoprecipitated with Bcl-2, and stimulation of p38MAPK resulted in reduction of the Bcl-2 half-life by approximately 3.5 fold and of Bcl-2 levels by 40% (step III₂). The remaining decrease in Bcl-2 protein levels is probably a consequence of the decreased Bcl-2 RNA and of the caspase-related degradation of the protein. Inhibition of p38MAPK destabilization of Bcl-2 using the SB203580 compound (step III₂) reduces DNA damage induced by the mutant viruses by approximately 40% and increased cell survival by approximately 30%. The fact that inhibition of p38MAPK did not completely restore Bcl-2 stability to the levels observed in mock-infected cells or in cells infected with the wt virus suggests that additional pathways contribute to the decreased Bcl-2 half-life. The JNK pathway is also activated by infection with HSV-1 mutants (46), and whether activated JNK acts through a mechanism similar to that for p38MAPK is under investigation. Phosphorylation of Bcl-2 is essential for the antiapoptotic role of the protein (16) and inactivation of its antiproliferative function (38). The ability of p38MAPK to phosphorylate Bcl-2 was demonstrated in vitro (23), and a recent report suggested inactivation of Bcl-2 via phosphorylation by the closely related ASK1/JNK pathway, activated at the G₂/M phase of the cell cycle (45). Our results propose a role for p38MAPK in destabilizing Bcl-2 and blocking its antiapoptotic functions during mutant HSV-1 infections. p38MAPK activity coimmunoprecipitates with Bcl-2 during infection of cells with the mutant viruses, suggesting that Bcl-2 interacts with a p38MAPK-related kinase. This interaction may be direct or could be mediated by a third component. It is also possible that p38MAPK activation results in subsequent stimulation of downstream proteins, which interfere with Bcl-2 stability.

Although p38MAPK is also activated during infection of cells with wt HSV-1 (46), no significant p38MAPK activity coimmunoprecipitated with Bcl-2, and the rate of degradation of the protein was the same as in mock-infected cells. Possibly, activation of p38MAPK during infection of cells with HSV-1 resulted in stimulation of two different populations, one involved in apoptotic processes and another that targets transcriptional factors. If the wt virus inhibits activation of the p38MAPK target proteins that lead to destabilization of Bcl-2 and allows only the p38MAPK-dependent activation of transcription factors to occur, that may explain the ability of wt HSV-1 to evade the p38MAPK-dependent decrease in Bcl-2 half-life while simultaneously using activation of p38MAPK to promote viral replication. If Bcl-2 is a direct target of p38MAPK, wt HSV-1 may interfere, for example, with the cellular localization of p38MAPK during infection, thus affecting the availability of the p38MAPK substrates. If interaction of Bcl-2 with p38MAPK is mediated by a third factor, this factor may be absent or inhibited during wt HSV-1 infections.

The apoptotic pathway downstream of Bcl-2 includes mitochondrial dysfunction (release of cytochrome c to the cytoplasm), degradation of cellular DNA, and cell death. This is in accordance with observations for mutant virus d120 (lacking functional ICP4), which is reported to induce cytochrome c release and depolarization of the inner mitochondrial membrane (10). Recently published data indicated that mutant d120 induced caspase-dependent apoptosis in HEp-2 cells and that overexpression of Bcl-2 blocked d120-induced apoptosis without inhibiting caspase-3 activity (11). These data strongly support our findings regarding the central role of bcl-2 in mutant HSV-1-induced apoptosis. In addition, our observations provide evidence for cellular mechanisms that target Bcl-2 during infection and for strategies employed by the wt virus to subvert programmed cell death.

Activation of p38MAPK during infection with wt HSV-1 enhanced viral growth by up to 60% (Fig. 9, pathway IV). Although this twofold increase is modest, it could be very important in vivo for the balance between viral growth and a cellular antiviral response or, at a low multiplicity of infection, when lower numbers of viral particles per cell are present. Furthermore, it is possible that in the presence of the p38MAPK inhibitor, other cellular pathways partially compensate for the loss of p38MAPK function and that a multiplicity of infection of 10 in our experiments masks certain aspects of the virus-host cell interactions. Thus, the contribution of p38MAPK in viral growth could be more important than twofold. A recent study reported that activation of the other stress kinase, JNK, also enhanced HSV-1 replication (26).

Inhibition of p38MAPK during wt HSV-1 infection resulted in decreased levels of specific viral proteins. Although this decrease is again modest, the effects of the p38MAPK inhibitor on HSV-1 protein levels are specific for certain viral proteins and highly reproducible. Activation of transcription of the respective viral gene promoters, presumably through activation of transcription factors downstream of p38MAPK, appears to be a mechanism for p38MAPK-dependent increases in viral protein levels. More direct measures of transcription activity, for example nuclear run-on assays, will be very important for verifying our results from the CAT assays. It is also possible that additional mechanisms, for example, increased stability of viral mRNAs, participate in p38MAPK-dependent upregulation of viral protein levels. Unfortunately, little is known about regulation of HSV-1 gene promoters by cellular transcription factors. Transcription factor AP-1 is a common target of JNK and is activated during HSV-1 infection (46). Infection with HSV-1 also induces translocation of NF-B to the nucleus, and inhibition of this process reduces viral yield (31). Furthermore, HSV-1 stimulates accumulation and redistribution of E2F (27).

In conclusion, we propose that cellular kinase cascades and regulatory factors are manipulated in the infected cell and that these activities are important for the efficiency of viral proliferation.