Induction of Apoptosis by Paramyxovirus Simian Virus 5 Lacking a Small Hydrophobic Gene

Abstract

Simian virus 5 (SV5) is a member of the paramyxovirus family, which includes emerging viruses such as Hendra virus and Nipah virus as well as many important human and animal pathogens that have been known for years. SV5 encodes eight known viral proteins, including a small hydrophobic integral membrane protein (SH) of 44 amino acids. SV5 without the SH gene (rSV5ΔSH) is viable, and growth of rSV5ΔSH in tissue culture cells and viral protein and mRNA production in rSV5ΔSH-infected cells are indistinguishable from those of the wild-type SV5 virus. However, rSV5ΔSH causes increased cytopathic effect (CPE) and apoptosis in MDBK cells and is attenuated in vivo, suggesting the SH protein plays an important role in SV5 pathogenesis. How rSV5ΔSH induces apoptosis in infected cells has been examined in this report. Tumor necrosis factor alpha (TNF-α), a proinflammatory cytokine, was detected in culture media of rSV5ΔSH-infected cells. Apoptosis induced by rSV5ΔSH was inhibited by neutralizing antibodies against TNF-α and TNF-α receptor 1 (TNF-R1), suggesting that TNF-α played an essential role in rSV5ΔSH-induced apoptosis in a TNF-R1-dependent manner. Examination of important proteins in the TNF-α signaling pathway showed that p65, a major NF-κB subunit whose activation can lead to transcription of TNF-α, was first translocated to the nucleus and was capable of binding to DNA and then was targeted for degradation in rSV5ΔSH-infected cells while expression levels of TNF-R1 remained relatively constant. Thus, rSV5ΔSH induced cell death by activating TNF-α expression, possibly through activation of the NF-κB subunit p65 and then targeting p65 for degradation, leading to apoptosis.

Apoptosis, or programmed cell death, is an important physiological process for host defense against viral infection (16). Viruses can activate a variety of cellular signaling pathways that lead to apoptosis. For example, cytokines such as tumor necrosis factor alpha (TNF-α) and interferons (IFNs) produced in response to viral infections can activate pathways leading to apoptosis (10, 26, 48). Apoptosis provides an opportunity for infected host organisms to clear viral infection by sacrificing virus-infected cells. However, many viruses have developed strategies to counteract apoptosis to prolong infections in their hosts (43). For example, cowpox virus encodes a viral protein, CrmA, that blocks apoptosis by inhibiting caspase-1 and caspase-3 (49, 53). Herpes simplex virus 1 can both induce and block apoptosis at multiple steps during infection and protects cells from exogenous apoptotic stimuli (20).

Apoptosis plays an important role in paramyxovirus pathogenesis. Many members of the Paramyxoviridae have been found to cause apoptosis. For example, Sendai virus causes apoptosis through activation of caspase-3 and caspase-8 by a mechanism that requires IFN regulatory factor 3 (5, 27). Measles virus induces apoptosis in the cells it infects, and apoptosis caused by measles virus infection is thought to facilitate virus release from infected cells (15, 19). Some members of the Paramyxoviridae have also been found to inhibit apoptosis. For example, mumps virus can inhibit hexadecylphosphocholine-induced apoptosis of human promonocytic cells U937 (22). Respiratory syncytial (RS) virus inhibits TNF-α-induced apoptosis in human respiratory epithelial cells and mononuclear cells (13, 30). Interestingly, RS virus also induces apoptosis in human respiratory epithelial cell A549 (4, 39). Mechanisms of activation and inhibition of apoptotic pathways by paramyxoviruses are not well understood.

TNF-α is a proinflammatory cytokine that can be induced by a variety of stimuli, including viral infection (3), and plays important roles in the control of virus infection (26). For example, TNF-α exerts strong anti-influenza virus activity which is even more potent than that exerted by IFNs (45). TNF-α expression has been detected in cells infected by many paramyxoviruses, such as Newcastle disease virus (NDV) (36) and Sendai virus (1, 2). Increased expression levels of TNF-α were detected both in the culture media of Sendai virus-infected cells and in Sendai virus-infected animal models (50). Pathogenicity caused by Sendai virus was reduced in vivo by treating the animals with neutralizing antibody against TNF-α (28, 50). How Sendai virus causes increased expression of TNF-α and how TNF-α exerts its cytotoxic effect on infected cells are not well understood.

Simian virus 5 (SV5) is a member of the Rubulavirus genus of the family Paramyxoviridae, which includes emerging viruses such as Hendra virus and Nipah virus as well as many important human and animal pathogens that have been known for years, such as mumps virus; human parainfluenza virus types 2, 3, and 4; NDV; Sendai virus; measles virus; canine distemper virus; rinderpest virus; and respiratory syncytial (RS) virus (33). Unlike most paramyxoviruses, SV5 can infect many cells with minimum cytopathic effect (CPE) and without inducing apoptosis (24). For example, SV5 can grow in MDBK cells productively up to 40 days without severe CPE (7). The ability of SV5 to grow productively without inducing apoptosis suggests that SV5 may have mechanisms to circumvent apoptosis. SV5 encodes eight known viral proteins, including a small hydrophobic (SH) integral membrane protein (33). The SH protein is a 44-residue membrane protein oriented with its N terminus in the cytoplasm. The SH gene is dispensable for virus replication in tissue culture cells. A recombinant SV5 without the SH gene (rSV5ΔSH) is viable and indistinguishable from the wild-type SV5 virus in single-step growth curve, plaque size, and viral protein and mRNA synthesis (23). Interestingly, rSV5ΔSH causes increased CPE and apoptosis in many cells. Coinfection of rSV5ΔSH with SV5 blocks CPE, and the blockage is dose dependent, i.e., the more wild-type SV5 that is used the less CPE that is induced by rSV5ΔSH, suggesting that the SH protein can inhibit apoptosis induced by virus infection (24). In this work the mechanism of rSV5ΔSH-induced apoptosis was investigated.

MATERIALS AND METHODS

Viruses and cells.

Generation of rSV5ΔSH was described previously (23). SV5 and rSV5ΔSH were grown in MDBK cells and were harvested 5 to 7 days postinfection (dpi) as described previously (41). Virus titers were determined by plaque assay with BHK 21F cells (41). To infect cells, cell monolayers were washed with phosphate-buffered saline (PBS) and then were incubated with viruses in DMEM-1% bovine serum albumin (BSA) at a multiplicity of infection (MOI) of 5 for 1 to 2 h at 37°C. The monolayers were washed and incubated with DMEM with 2% fetal calf serum (FCS) at 37°C with 5% CO₂.

L929 cells were maintained in DMEM containing 10% FCS. BHK 21F cells were maintained in DMEM containing 10% tryptose phosphate broth and 10% FCS. Virus-infected cells were grown in DMEM containing 2% FCS.

Purification of fragmented DNA, caspase assay, propidium iodide (PI) staining, Annexin-V staining, and terminal deoxynucleotidyltransferase-mediated dUTP-FITC nick end labeling (TUNEL) assay.

Fragmented DNAs were purified as described previously (52). Confluent L929 cells in 6-cm-diameter plates were infected with SV5 or rSV5ΔSH at an MOI of 5. At 24 and 48 h postinfection (hpi), similar numbers of L929 cells in 60-mm-diameter dishes were washed twice with PBS without Mg²⁺ or Ca²⁺ (PBS-) and then were incubated in 0.5 ml of TTE buffer (0.2% Triton X-100, 10 mM Tris, 15 mM EDTA, pH 8.0) at room temperature for 15 min. Cell lysates were harvested into microtubes and were subjected to centrifugation at 14,000 rpm (Microfuge 18; Beckman Coulter) for 20 min. Supernatants were digested with 100 μg of RNase A/ml at 37°C for 1 h. Samples were purified with phenol-chloroform extraction, precipitated, and washed with 70% ethanol. Pellets were air dried and redissolved in 10 μl of Tris-EDTA. Electrophoresis was performed on 2% agarose gels with size markers.

For caspase-3 assays, triplicates of infected cells (∼2 × 10⁶ to 10 × 10⁶ cells/ml) were lysed with cell lysis buffer containing 10 mM Tris-HCl, 10 mM NaH₂PO₄-NaHPO₄ (pH 7.5), 130 mM NaCl, 1% Triton X-100, and 10 mM sodium pyrophosphate. Protein concentrations of lysates were determined by using a bicinchoninic acid protein assay kit (Pierce, Rockford, Ill.). For each reaction mixture in a 96-well microtiter plate, 30 μg of cell lysate was added to a solution of 200 μl of reaction buffer (20 mM HEPES [pH 7.5], 20% glycerol, 4 mM dithiothreitol [DTT]) and 20 μM caspase-3 substrate AFC-DEVD (BD Biosciences Pharmingen). For caspase-2 assays, assay buffer containing 100 mM HEPES (pH 7.5), 10 mM DTT, and 100 μM caspase-2 substrate AFC-LDESD was used. Samples in the plates were mixed for 30 s and were incubated for 1 h at 37°C. The plates were then read on a microtiter plate reader with an excitation wavelength of 380 nm and an emission wavelength of 430 to 460 nm.

For PI staining, confluent L929 cells in 6-cm-diameter plates were infected with SV5 or rSV5ΔSH at an MOI of 5. Cells in monolayers were trypsinized and combined with floating cells in the media at different time points. The harvested cells were then centrifuged at 250 × g for 8 min at 4°C and were washed with PBS between each step. The cells were fixed with 0.25% formaldehyde for 2 h at 4°C. The fixed cells were resuspended in 0.5 ml of 50% DMEM-50% FCS and permeabilized by adding 1.5 ml of 70% ethanol at 4°C for at least 2 h and up to 3 days. To monitor expression of viral proteins the permeabilized cells were incubated with 0.5 ml of anti-V/P monoclonal antibody P-k at 1:500 in PBS-1% BSA at 4°C for 1 h and then with 0.5 ml of fluorescein isothiocyanate (FITC)-labeled anti-mouse secondary antibody (Organon-Teknika Corp., Charlotte, N.C.) at 1:1,000 in PBS-1% BSA for 1 h at 4°C. For PI staining the cells were incubated with 500 μl of a 50-μg/ml concentration of PI (Sigma-Aldrich) for 1 h at 4°C. The cells were then analyzed by using a flow cytometer (EPICS XL; Beckman-Coulter). Single cells were selected on FL2-W (cell width) versus FL2-A (DNA content) plots. Infected cells were selected on FL2-A (DNA content) versus FL1-H (V or P expression plots).

For TUNEL assay the cells were treated the same as for the PI staining before secondary antibody was added. Phycoerythrin-labeled (BD Bioscience no. 550589) instead of fluorescein isothiocyanate (FITC)-labeled secondary antibody was used because it has a different fluorescence spectrum from that of the TUNEL reagent. The cells were then incubated with 25 μl of TUNEL reaction mixture (in situ Cell Death Detection Kit, Fluorescein; Boehringer-Mannheim) for 2 to 3 h in the dark at 37°C. The cells were analyzed by flow cytometry.

For Annexin-V binding the cells were incubated with FITC-labeled Annexin-V (Annexin-V-FLUOS) for 15 min at room temperature according to the manufacturer's protocol (Roche Diagnostics Corp., Mannheim, Germany). The fluorescence of 10,000 cells was examined by using a flow cytometer.

UV inactivation of SV5.

Confluent L929 cells in 60-mm-diameter plates were infected with SV5 or rSV5ΔSH at an MOI of 5. The infected cells were incubated in 5 ml of 2% FCS-DMEM for 2 days. Covers of the plates were removed, and the plates were placed inside a Fisher Hamilton Biological Safety cabinet Class II and were UV treated for 30 min. The media were then filtered through a 0.22-μm-pore-size filter to remove cell debris. The effectiveness of the UV treatment to inactivate SV5 was confirmed by plaque assay. No plaques were formed in BHK 21F cells incubated with the UV-irradiated media.

Antibody treatment of infected cells.

Confluent L929 cells in 6-well plates were infected with rSV5ΔSH at an MOI of 5 as described before and were incubated in 1 ml of DMEM-2% FCS with neutralizing antibody against TNF-α (BD Pharmingen) at 1, 4, or 20 μg/ml or with control antibody (BD Pharmingen) that is the same subtype antibody as the neutralizing antibody at 20 μg/ml. At 2 dpi the cells were photographed by using a light microscope equipped with a digital camera. To quantify the effect of neutralizing antibody, the cells were collected and TUNEL assays were carried out as described before. To examine TNF-R1's involvement in rSV5ΔSH-induced apoptosis, L929 cells in 12-well plates were infected and incubated with anti-TNF-R1 (BD Pharmingen) or control antibody (BD Pharmingen) that is the same subtype as the neutralizing antibody at 75 μg/ml. At 2 dpi the cells were photographed. To quantify the effect of anti-TNF-R1, L929 cells in 96-well plates were infected and incubated in 50 μl of DMEM-2% FCS with antibody at 75 μg/ml for 2 days. The cells attaching to the plate were washed with PBS-, trypsinized, and combined with the cells in the media to obtain entire cell populations. The cells were centrifuged at 3,000 rpm (Microfuge 18; Beckman Coulter) and were resuspended in a 0.08% trypan blue solution made in DMEM. The cells were placed on a hemacytometer and were examined under microscope. Blue cells (dead cells) and transparent cells (live cells) were counted.

Enzyme-linked immunosorbent assay (ELISA) of TNF-α in media of infected cells.

L929 cells were infected and media were collected at different time points postinfection. The amounts of TNF-α were measured by using a murine TNF-α detection kit purchased from Amersham Pharmacia (Piscataway, N.J.) according to the manufacture's instructions. Fifty microliters of media from infected cells or standards in duplicate and 50 μl of biotinylated antibody against TNF-α were added to strips prelabeled with antibody against TNF-α. The strips were incubated at room temperature for 2 h. After being washed three times with wash buffer provided by the manufacturer, 100 μl of streptavidin-horseradish peroxidase conjugate was added to the strips, and the strips were incubated at room temperature for 30 min. The strips were then washed three times, and 100 μl of 3,3′,5,5′-tetramethylbenzidine substrate solution was added to each well. The strips were incubated in the dark at room temperature for 30 min, and 100 μl of stop solution was added to each well. Optical density at 450 nm was measured within 30 min. The amounts of TNF-α were calculated by using standard curves generated from known concentrations of TNF-α provided by the manufacture.

Detection of NF-κB p65 by using fluorescent microscopy.

L929 cells on glass coverslips were mock infected or were infected with rSV5 or rSV5ΔSH. At 1 dpi cells were washed with PBS and then were fixed in 0.5% formaldehyde for 15 min at room temperature. In all further steps, 0.1% Saponin was included in the PBS to permeabilize cells. The cells were washed with PBS-Saponin solution and were incubated for 30 min in a 1:100 dilution of monoclonal antibody specific for the p65 subunit of NF-κB factor. Cells were washed three times with PBS-Saponin, and FITC-labeled goat anti-mouse antibody (Jackson Laboratory, Bar Harbor, Maine) was added to the cells. The cells were incubated for 30 min and then were washed three times in PBS-Saponin. Fluorescence was examined and photographed by using an Olympus BX-60 fluorescent microscope.

EMSA.

To examine DNA binding of NF-κB in mock-, SV5-, and rSV5ΔSH-infected cells an electromobility shift assay (EMSA) was used as described previously (34). The infected L929 cells were trypsinized and washed in PBS at 1 and 2 dpi. The cells were counted, and 10⁶ cells were resuspended in 1 ml PBS- and were pelleted in an Eppendorf tube. The cells were resuspended in 400 μl of cold buffer (10 mM HEPES [pH 7.9], 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride) for 15 min on ice. Then 25 μl of 10% Nonidet NP-40 was added. The tubes were vortexed and then were centrifuged at 10,000 rpm (Microfuge 18; Beckman Coulter) for 30 s. The supernatant was discarded, and 50 μl of cold buffer (20 mM HEPES [pH 7.9], 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride) was added. The tubes were then rocked at 4°C for 15 min. This was followed by centrifugation at 10,000 rpm for 5 min. The supernatant was collected into new Eppendorf tubes and was stored at −70°C.

Oligomers containing NF-κB binding sites (5′-AGCTAAGGGACTTTCCGCTGGGGACTTTCCAGG-3′ and 5′-AGCTCCTGGAAAGTCCCCAGCGGAAAGTCCCTT-3′) were radiolabeled by using Klenow fragment in an end-filling reaction mixture. NF-κB oligomers, deoxynucleoside triphosphates (G, A, and T), [α-³²P]dCTP, reaction buffer, Klenow fragment, and water were incubated together at room temperature for an hour, and then the radiolabeled oligomers were purified by using a Sephadex column. To examine binding of nuclear extract from infected cells (described above) to the labeled NF-κB probe, 3 to 5 μl of lysate (mock, SV5, or rSV5ΔSH), NF-κB probe (about 20,000 cpm), 2 μl of 10× EMSA buffer (250 μl of 1 M HEPES [pH 7.5], 600 μl of 1 M KCl, 90.5 μl of glycerol, 2.0 μl of 0.5 M EDTA, 7.5 μl of 1 M DTT, 50 μl of 1 M MgCl₂), 2 μl of poly(dI-dC), and water up to 20 μl were incubated together at room temperature for 15 to 30 min. As controls, unlabeled NF-κB oligomers and nonspecific competitor NFAT were added to nuclear lysates from mock-, SV5-, or rSV5ΔSH-infected cells at over 50-fold excess. For supershift, up to 20 μg of anti-p65, anti-p50, anti-c-Rel, or anti-C/EBP-β (Santa Cruz Biotechnology, Inc., Santa Cruz, Calif.) was added to the DNA probe and nuclear lysates mixture. Mixtures of lysates, labeled oligomers, and antibodies were resolved on a 6% polyacrylamide gel at approximately 150 V. The gels were dried and radioactivities were detected by using a Storm System PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, Calif.).

Immunoblotting.

L929 cells in 6-cm-diameter plates were infected with rSV5ΔSH or rSV5 at 1 and 2 dpi. Cells were lysed in 0.5 ml of protein lysis buffer (2% sodium dodecyl sulfate [SDS], 62.5 mM Tris-HCl [pH 6.8], 2% DTT), and lysates were sonicated briefly to shear DNA. Up to 80 μl of the lysate was subjected to SDS-polyacrylamide gel electrophoresis (PAGE) by using a 10% gel. Polypeptides were transferred to polyvinylidene difluoride membrane by using a wet gel transfer apparatus. The membrane was first incubated with primary antibodies against TNF-R1, TRADD, Daxx, JNK2, p65 (Santa Cruz Biotechnology, Inc.), or actin (Sigma) and then were incubated with a mixture of anti-mouse and anti-rabbit secondary antibodies conjugated to horseradish peroxidase. The proteins on the membrane were detected by using the ECL+ kit (Amersham Pharmacia), and chemiluminescence was detected by using a Storm System PhosphorImager (Molecular Dynamics, Inc.).

RESULTS

rSV5ΔSH induced apoptosis in L929 cells.

Previously, studies of apoptosis induced by rSV5ΔSH infection were carried out mainly with MDBK cells, a bovine cell line. L929 cells, a mouse cell line, were used in this work because reagents such as mouse cytokines and antibodies against mouse proteins are readily available and because L929 cells are transfectable, allowing mechanistic studies on apoptosis induction that would be difficult with MDBK cells. In addition, a small-animal model based on mice lacking the stat1 gene has been established to study SV5 pathogenesis in vivo (24). Understanding apoptosis induced by SV5 in mouse cell lines will provide a foundation for future studies with the mouse model.

L929 cells were mock infected or were infected with SV5 or rSV5ΔSH virus at an MOI of 5 PFU/cell. CPE was observed in rSV5ΔSH-infected cells at 2 dpi but not in SV5-infected or mock-infected cells, consistent with previous observations (24) (Fig. 1A). To test whether the CPE is due to induction of apoptosis, the presence of fragmented chromosomal DNA in SV5-infected cells was examined. L929 cells were mock infected or were infected with SV5 or rSV5ΔSH and were harvested at 1 and 2 dpi. Fragmented DNAs were purified as described in Materials and Methods and were resolved in 2% agarose gels (Fig. 1B). Increasing amounts of fragmented DNA were detected in rSV5ΔSH-infected cells but not in mock- or SV5-infected cells, suggesting that rSV5ΔSH induced apoptosis in L929 cells. Caspases are key players in regulating apoptotic pathways. To investigate involvement of caspases in the CPE, activities of caspase-2 and caspase-3 in infected cells were examined. Consistent with reports of increased activities of caspase-2 and caspase-3 in rSV5ΔSH-infected MDBK cells (24), caspase-2 and caspase-3 were both activated in rSV5ΔSH-infected L929 cells (Fig. 1C).

FIG. 1.

rSV5ΔSH induced apoptosis in L929 cells. (A) rSV5ΔSH induced CPE in L929 cells. L929 cells were infected at 5 MOI and were photographed at 2 dpi. (B) Increased DNA laddering in rSV5ΔSH-infected cells. L929 cells were infected at 5 MOI and were collected at the times indicated. Preparations of fragmented DNAs were carried out as described in Materials and Methods. Mk, mock infection; Wt, wild-type SV5 infection; ΔSH, rSV5ΔSH infection. (C) Increased caspase-2 and caspase-3 activities in rSV5ΔSH-infected L929 cells. Infected L929 cells were collected at the times indicated, and caspase assays were performed as described in Materials and Methods. The results are representative of four different experiments. (D) TUNEL assay. (E) Annexin-V staining. (F) PI staining. The left-hand graphs of panels D, E, and F are results from analyzing cells 48 hpi by using flow cytometry. “A” represents the sub-G₀-G₁ cell population that is considered apoptotic. The right-hand graphs of panels D, E, and F are summaries of the results from analyzing cells at time the points indicated. Samples are triplicates. Errors are standard errors of means.

To further confirm that this CPE was due to induction of apoptosis in the rSV5ΔSH-infected cells and to quantify the apoptosis, three different measurements of apoptosis were performed: TUNEL assay, Annexin-V staining, and PI staining. L929 cells were mock infected or were infected with rSV5 or rSV5ΔSH at 5 MOI, and the infected cells were collected at different time points postinfection. The cells were subjected to TUNEL assay that directly measures the presence of nicked chromosomal DNA, a hallmark of apoptotic cells. Percentages of TUNEL-positive cells, i.e., apoptotic cells, were quantified by using a flow cytometer as described previously (24). As shown in Fig. 1D, at 36 hpi about 25% of cells infected by rSV5ΔSH were apoptotic. By 48 hpi the percentage of apoptotic cells increased to about 80%, while apoptotic cell populations in the mock or SV5 infections remained at background levels through 72 hpi. To confirm the results obtained by using TUNEL assay, the cells were infected and examined for the presence of phosphatidylserine (PS) on the outer leaflets of cell membranes, an indicator of cells undergoing apoptosis. The infected cells were harvested and incubated with FITC-labeled Annexin-V protein that binds to PS. As a result, the cells with PS on the outer leaflets of cell membranes were stained positive. Annexin-V staining was detected on cells infected by rSV5ΔSH but not on cells infected by wild-type SV5 at 48 hpi (Fig. 1E). The ability of rSV5ΔSH to induce apoptosis in L929 was further confirmed by PI staining, which detects DNA content profiles of cells. When cells undergo apoptosis, host chromosomal DNAs start to fragment. These fragmented DNAs are lost during the PI staining in which lipid membranes of cells were extracted. As a result, the DNA content profile of apoptotic cells is the same as that of sub G₀-G₁ cells. As shown in Fig. 1F, increased populations of sub G₀-G₁ cells, i.e., apoptotic cells, were detected only in rSV5ΔSH-infected cells starting at 48 hpi.

UV-irradiated culture media harvested from virus-infected cells caused CPE.

To investigate the mechanism of rSV5ΔSH-induced apoptosis, the ability of culture media from the infected cells to cause CPE was examined. L929 cells were infected with SV5 or rSV5ΔSH at 5 MOI for 2 days. The culture media from the infected cells were collected, UV-irradiated to inactivate SV5 virus, and filtered through 0.22-μm-pore-size filters to remove cell debris. Complete inactivation of virus by UV irradiation was confirmed by plaque assay. The media were then added to fresh L929 cells. After 20 h of incubation CPE was observed in the cells incubated in UV-treated media from rSV5ΔSH-infected cells; CPE was also observed in the cells incubated in UV-treated media from SV5-infected cells, albeit at a lesser degree. Very little CPE was observed in the cells incubated in UV-treated media from mock infection (Fig. 2A). This result shows that components secreted from virus-infected cells can cause CPE in L929 cells, suggesting the possibility that proteins, such as cytokines induced by virus infection, may be involved in apoptosis induced by rSV5ΔSH infection. The difference in severity of CPE caused by incubation with the UV-treated media suggests that more of the factor responsible for the CPE was produced in rSV5ΔSH-infected cells.

FIG. 2.

TNF-α was produced from virus-infected cells. (A) Media from rSV5ΔSH-infected cells can cause CPE. L929 cells in 6-cm-diameter plates were infected at 5 MOI. At 2 dpi, media in the plates were UV-irradiated and the UV-treated media were used to replace regular growth media for L929 cells in 6-well plates. After 20 h of incubation the cells were photographed. (B) TNF-α was produced from rSV5ΔSH-infected cells. L929 cells were infected and media samples were taken at different time points after infection. Amounts of TNF-α were measured by using ELISA. Samples are triplicates, and error bars are standard errors of the means. (C) TNF-α was produced from SV5-infected cells. The UV-treated media obtained as described for panel A were mixed with anti-TNF-α (20 μg/ml), control antibody (20 μg/ml), or no antibody in 4 wells of fresh L929 cells in a 48-well plate. The cells were harvested and counted by using trypan blue. The percentages of dead cells are shown. Error bars are standard errors of the means.

Increased levels of TNF-α were produced from virus-infected cells.

To identify factors released into the culture media of infected L929 cells that were responsible for inducing CPE, cytokine production in infected cells was examined. It was shown before that caspase-2 was activated in rSV5ΔSH-infected L929 cells (Fig. 1C). Caspase-2 can be activated by many stimuli, including TNF-α. To investigate the involvement of TNF-α in rSV5ΔSH-induced apoptosis, production of TNF-α by infected cells was measured by using ELISA. An increase of TNF-α was detected in rSV5ΔSH-infected cells compared with that of rSV5 infected cells starting 48 hpi at a range of 100 to 250 pg/ml (Fig. 2B), a range that is sufficient to induce CPE in L929 cells (data not shown).

To investigate whether TNF-α was involved in the CPE in the L929 cells that were incubated with the UV-treated media from SV5-infected cells (Fig. 2A), neutralizing antibody against TNF-α was added to the UV-treated media. Percentages of dead cells were examined by using a trypan blue dye exclusion assay (Fig. 2C). CPE caused by the UV-treated media from SV5-infected cells was inhibited by anti-TNF-α but not by control antibody, suggesting that a lower level of TNF-α that was not detectable in the ELISA (Fig. 2B) was released into the culture media of SV5-infected cells.

Inhibition of rSV5ΔSH-induced apoptosis by neutralizing antibody against TNF-α and TNF-R1.

The release of TNF-α into the medium of rSV5ΔSH-infected cells suggests that TNF-α could be involved in rSV5ΔSH-induced apoptosis. To study whether TNF-α indeed induced apoptosis in rSV5ΔSH-infected cells, L929 cells were infected with rSV5ΔSH and were incubated with neutralizing antibody against TNF-α. rSV5ΔSH infection caused severe CPE at 2 dpi as expected (Fig. 3A). The severity was progressively reduced with increasing amounts of anti-TNF-α in the medium. At 20 μg/ml CPE was entirely inhibited by the neutralizing antibody, while no inhibition was observed in the cells incubated with a 20-μg/ml concentration of the control antibody. Furthermore, the ability of TNF-α neutralizing antibody to inhibit apoptosis induced by rSV5ΔSH was tested and quantified by TUNEL assay by using flow cytometry. As shown in Fig. 3B, the apoptosis induced by rSV5ΔSH infection was inhibited when the infected cells were incubated with the neutralizing antibody against TNF-α but not when the cells were incubated with control antibody. Importantly, replication of the virus was not inhibited by the antibody treatment (data not shown).

FIG. 3.

Anti-TNF-α and anti-TNF-R1 inhibited rSV5ΔSH-induced apoptosis. L929 cells in 6-well plates were infected with 5 MOI of rSV5ΔSH and were incubated with no antibody, control antibody (20 μg/ml), or neutralizing antibody at 1, 4, or 20 μg/ml for 40 h. (A) Inhibition of CPE by the neutralizing antibody against TNF-α. (B) Inhibition of apoptosis by the neutralizing antibody against TNF-α. TUNEL assays were carried out on the cells infected with rSV5ΔSH and treated with antibodies. Samples are triplicates, and error bars are standard errors of the means. IgG, immunoglobulin G. (C) Inhibition of CPE by the neutralizing antibody against TNF-R1. L929 cells in 12-well plates were infected with rSV5ΔSH at 5 MOI and were incubated with nothing, control antibody (75 μg/ml), or neutralizing antibody (75 μg/ml) for 40 h. Percentages of dead cells were measured by using trypan blue dye exclusion assay. Results are averages of eight replicate samples, and error bars are standard errors of the means.

TNF-α can bind two receptors, TNF-R1 and TNF-R2. TNF-R1 is a 55-kDa membrane protein with a death domain whose activation by TNF-α leads to apoptosis. TNF-R2 is a 75-kDa glycoprotein whose activation by TNF-α leads to regulation of proliferation activity of cells mostly of the immune system. Although TNF-R2 lacks a death domain, it is implicated in triggering the apoptotic pathway (9). To investigate which TNF-α signaling pathway is involved in apoptosis induced by rSV5ΔSH, contributions of TNF-R1 and TNF-R2 in inducing apoptosis in rSV5ΔSH-infected cells were determined. L929 cells infected with rSV5ΔSH were incubated with neutralizing antibody that specifically inhibits TNF-R1 but not TNF-R2 activity for 2 days (46). Less CPE were observed in the cells incubated with anti-TNF-R1 than in the cells incubated with control antibody or no antibody at all, and the result was quantified by using a trypan blue dye exclusion assay. Only anti-TNF-R1 reduced dead cell populations in rSV5ΔSH-infected cells (Fig. 3C), indicating that the secreted TNF-α induced by rSV5ΔSH infection caused the apoptosis of rSV5ΔSH-infected cells through the TNF-R1 signaling pathway. Treating the infected cells with neutralizing antibody against TNF-R2 had no effect on rSV5ΔSH-induced CPE (data not shown).

Translocation of NF-κB p65 into nuclei and DNA-binding by NF-κB p65.

To further study activation of apoptosis in rSV5ΔSH-infected cells, important players in the TNF-α signaling pathway were examined. It is known that TNF-α can activate NF-κB and that the activated NF-κB can further up-regulate expression of TNF-α. Activation of NF-κB results in its translocation to the nucleus. To examine activation of NF-κB through its localization, L929 cells growing on glass coverslips were mock infected or were infected with SV5 or rSV5ΔSH. The cells were then fixed, permeabilized, incubated with antibody against p65 (a key subunit of NF-κB factor), and examined under a fluorescence microscope. As shown in Fig. 4A, only in rSV5ΔSH-infected L929 cells did p65 exhibit nucleus localization, suggesting activation of NF-κB in rSV5ΔSH-infected cells. Activation of p65 was further examined by EMSA. Nucleus lysates from the same numbers of infected cells were harvested at 1 and 2 dpi, incubated with ³²P-labeled NF-κB probes, and then resolved in 6% polyacrylamide gels. At 1 dpi the NF-κB probe was shifted by the nucleus lysate from rSV5ΔSH-infected cells (Fig. 4B), indicating likely activation of NF-κB factor in rSV5ΔSH-infected cells at 1 dpi. A lower-level shift of NF-κB probe by nucleus lysate from SV5-infected L929 cells was also observed, suggesting a lower level of NF-κB activation in SV5-infected cells. Interestingly, at 2 dpi very few NF-κB probes were shifted by the nucleus lysates from all infected cells (Fig. 4B). To investigate the identity of the shifted product, anti-p65, anti-p50, anti-c-Rel, or anti-C/EBPβ was added to the EMSA mixture and resolved in a 6% gel. Antibody against p65 further shifted the DNA-nuclear lysate complex while the other antibodies did not (Fig. 4C), suggesting the shifted complex is mainly made of p65 and DNA probe.

FIG. 4.

Activation of NF-κB in rSV5ΔSH-infected L929 cells. (A) Localization of NF-κB (p65) in rSV5ΔSH-infected cells. L929 cells were mock infected (Mk) or were infected with SV5 (Wt) or rSV5ΔSH (ΔSH), and immunofluorescence of p65 was carried out as described in Materials and Methods. (B) EMSA. L929 cells were infected and nuclei were lysed 1 and 2 dpi. The nucleus lysates were inoculated with radiolabeled NF-κB probes plus appropriate control DNA oligomers as described in Materials and Methods. (C) Supershift. Nuclear lysates collected at 1 dpi were incubated with DNA probes in the same conditions as those described for panel B plus antibodies against p65, p50, c-Rel, and C/EBP-β.

Degradation of p65.

To investigate the lack of shifting of the NF-κB probes at 2 dpi, expression levels of NF-κB subunit p65 and other important proteins in the TNF-α signaling pathway were examined. L929 cells were infected and harvested at 1 and 2 dpi. The total cell lysates were resolved in SDS-10% PAGE and were subjected to immunoblotting with antibodies against TNF-R1, TRADD, Daxx, and NF-κB (p65). Actin was used as a control for protein loading. While relative amounts of TNF-R1 and TRADD to actin changed little in mock-, SV5-, or rSV5ΔSH-infected cells over time (Fig. 5B), p65, a major NF-κB subunit, was not detected at 2 dpi in rSV5ΔSH-infected cells, in contrast to SV5-infected and mock-infected cells, in which p65 was readily detected (Fig. 5B). Interestingly, the amount of Daxx which is normally associated with induction of apoptosis (37) increased in rSV5ΔSH-infected cells (Fig. 5C). To investigate whether p65 degradation is a result of cell death, the amounts of p65 in L929 cells treated with TNF-α were examined. L929 cells were incubated with TNF-α at 20 ng/ml for 20 h, which led to severe CPE (data not shown). Amounts of p65 in total cell lysates were examined by using immunoblotting. As shown in Fig. 6A, the amount of p65 in the TNF-α-treated cells did not decrease. It has been shown earlier that the culture media from SV5 virus-infected cells induced CPE in L929 cells (Fig. 2A and C). To investigate what caused degradation of p65 in rSV5ΔSH-infected cells, the ability of culture media from virus-infected cells to induce p65 degradation was examined. Amounts of p65 in the cells incubated with the UV-irradiated media were examined and remained constant in the cells (Fig. 6B). To further examine the relationship between cell death and degradation of p65, amounts of p65 in JNK1^−/− cells, which resist rSV5ΔSH-induced apoptosis (Fig. 6C), were examined. JNK1^−/− cells were mock infected or were infected with SV5 or rSV5ΔSH and harvested at 1, 2, and 3 dpi. Amounts of p65 along with JNK2 and actin were compared by using immunoblotting. At 3 dpi p65 disappeared from rSV5ΔSH-infected JNK1^−/− cells, indicating that cell death is not a determining factor for the disappearance of p65 in rSV5ΔSH-infected cells.

FIG. 5.

Degradation of p65 in rSV5ΔSH-infected cells. Infected L929 cells were collected at 1 and 2 dpi. Lysates were subjected to SDS-PAGE on 10% gels and were immunoblotted by using antibodies against TNF-R1, Daxx, TRADD, p65, and actin as described in Materials and Methods. (A) Anti-p65. Anti-p65 is a monoclonal antibody generated from residues 1 to 286 of p65 (SC no. 8008; Santa Cruz Biotechnology). (B) Anti-TNF-R1 and anti-TRADD. (C) Anti-Daxx. Mk, mock infection; WT, SV5 infection; ΔSH, rSV5ΔSH infection.

FIG. 6.

Expression of p65. (A) Expression of p65 in TNF-α-treated cells. L929 cells were incubated with TNF-α (20 ng/ml) for 20 to 24 h, and all cells (including floating cells) were collected. The cell lysates were subjected to immunoblotting with anti-p65 and anti-actin. (B) Expression of p65 in the cells incubated with UV-irradiated media. The UV-irradiated media were obtained in the same way as that described in the legend to Fig. 2A. The cells incubated with the media were collected at 20 to 24 h after incubation and were subjected to immunoblotting with anti-p65 and anti-actin. (C) Infection of JNK1^−/− cells with SV5. JNK1^−/− cells in 6-cm-diameter plates were mock infected (Mk) or were infected with SV5 (Wt) or rSV5ΔSH (ΔSH) at 5 MOI and were photographed at 5 dpi. (D) Expression of p65 in virus-infected JNK1^−/− cells. The cells were mock infected or were infected with SV5 or rSV5ΔSH at 5 MOI and were collected at 1, 2, and 3 dpi. The cell lysates were subjected to immunoblotting with anti-p65, JNK2, and actin.

DISCUSSION

In this work we found that rSV5ΔSH induced apoptosis in L929 cells through TNF-α, suggesting that the SH protein of SV5 is required to block TNF-α induced apoptosis. Many viruses encode SH membrane proteins with diverse functions. Some of them play important roles in inducing CPE and apoptosis in infected cells. Poliovirus expresses 3A, an 87-amino-acid membrane protein residing in the endoplasmic reticulum, is a potent inhibitor of protein translocation from endoplasmic reticulum to Golgi, and is required for inducing CPE in poliovirus-infected cells (12). The 6K protein of Sindbis virus contributes to its ability to induce apoptosis (29). Other members of the Paramyxoviridae that contain SH genes are the closely related Rubulavirus (the mumps virus [14]) and Pneumovirus (RS virus [8]) genera. SH proteins of mumps virus and RS virus are type II integral membrane proteins of 57 and 64 amino acids, respectively (40). Functions of the SH proteins of mumps virus and RS virus are not clear. There are no obvious sequence homologies among SH proteins. Interestingly, both mumps virus and RS virus can block apoptosis, suggesting that they may also encode antiapoptosis proteins (22, 30). Like SV5 without SH, a mutant mumps virus which has a defect in expressing SH protein grew normally in tissue culture cells, suggesting that the SH protein is not essential for mumps virus growth in tissue culture cells (47). A recombinant RS virus without the SH gene grows efficiently in cell culture and is attenuated in the upper respiratory tract of the mouse when the virus was administered intranasally (6). The efficient growth in tissue culture cells and attenuation in animal models of RS virus without SH are reminiscent of the phenotypes of rSV5ΔSH. Intriguingly, RS virus is also reported to block TNF-α-induced cell death (13, 30). We hypothesize that the SH proteins from mumps virus and RS virus may be functional counterparts of the SH protein of SV5. The SH protein of SV5, a type II membrane protein with a short cytoplasmic tail, may define a new class of antiapoptosis proteins.

Cytotoxicity of TNF-α in L929 is well known. In fact, L929 is so sensitive to TNF-α-induced CPE that L929 cells are routinely used for detecting the presence of TNF-α in a biological assay (18). However, there are conflicting reports as to the nature of CPE. There are reports showing that TNF-α induces apoptosis in L929 cells, and there are reports showing that TNF-α only induces necrosis in L929 cells (17, 21, 32, 38, 44). The conflicting results may be due to the detection method used for apoptosis and different culture conditions used for L929 cells in different reports. In our experiments, without addressing whether TNF-α alone induces apoptosis or necrosis in L929 cells, we found that TNF-α played an essential role in rSV5ΔSH-induced apoptosis of L929 cells by using DNA laddering, caspase assay, TUNEL assay, and Annexin-V and PI staining (Fig. 1B to F). Interestingly, rSV5ΔSH infection caused degradation of the p65 subunit of NF-κB while TNF-α alone or media from the virus-infected cells did not cause p65 degradation even though they both induced cell death, suggesting that there may be different pathways activated by rSV5ΔSH and by TNF-α leading to cell death.

Many paramyxoviruses induce expression of cytokines, such as TNF-α, in infected cells. However, little is known about the involvement of viral proteins. SV5, a virus that causes little CPE in many cells, stimulated TNF-α production only very weakly (Fig. 2C). Deletion of SH from SV5 results in a virus that induced apoptosis through production of TNF-α. Other paramyxoviruses may use the same signaling pathway to induce apoptosis. For example, TNF-α is produced in NDV infection, and the infection causes apoptosis. Inducing apoptosis by TNF-α may be a common mechanism used by hosts to eliminate paramyxovirus infection. Not surprisingly, rSV5ΔSH is attenuated in vivo (24), likely due to clearance of apoptotic cells by the host immune system.

Translocation of the p65 subunit of NF-κB to the nucleus and increased DNA binding by p65 in rSV5ΔSH-infected cells were somewhat puzzling (Fig. 4A). NF-κB is known to activate transcription of TNF-α; thus, possible activation of NF-κB is consistent with increased TNF-α expression in rSV5ΔSH-infected cells. However, NF-κB can also activate expression of antiapoptotic genes which in turn inhibit apoptosis. However, apoptosis was induced by rSV5ΔSH at 2 dpi. The disappearance of p65 provides an explanation for rSV5ΔSH to induce cell death. Mechanisms of p65 degradation in rSV5ΔSH-infected cells is not known. It is unlikely that p65 degradation is a result of cell death, because cell death alone does not induce p65 degradation and p65 is degraded in the absence of CPE in rSV5ΔSH-infected JNK1^−/− cells. Treating cells with TNF-α led to severe CPE but not to degradation of p65 (Fig. 6). In addition, we found that rSV5ΔSH infection caused p65 degradation in JNK1^−/− cells without inducing cell death (Fig. B and C). These results suggest that p65 degradation is not a consequence of cell death; rather, it is a result of rSV5ΔSH infection. Degradation of p65 has also been reported for cells infected with African swine fever virus (ASFV), a large icosahedra double-stranded DNA virus (51). Infection with ASFV leads to disappearance of p65 and apoptosis in endothelial cells. It has been reported that caspase activated by deprivation of growth factors can cleave p65, and cleavage of p65 is a step required for, rather than a consequence of, apoptosis (35). A mutant p65 which is resistant to caspase cleavage and is an inhibitor of caspase can block the apoptosis induced by the growth factor starvation, suggesting that p65 plays an active role in inhibiting apoptosis and that degradation of p65 is required for the apoptosis to proceed (35). Increased caspase-2 and -3 activities were detected in rSV5ΔSH-infected L929 cells (Fig. 1C), suggesting it is possible that rSV5ΔSH induced p65 degradation through a caspase-dependent mechanism. However, we tried unsuccessfully to use caspase inhibitors to block p65 degradation because of high cytotoxicity of the caspase inhibitors in the L929 cells (data not shown).

No TNF-α was detected from the media of SV5-infected L929 cells by using ELISA (Fig. 2B), while CPE was caused by the UV-treated media from SV5-infected cells (Fig. 2A). The CPE observed in the L929 cells treated with UV-inactivated media from SV5-infected cells was blocked by neutralizing antibody against TNF-α, suggesting that low levels of TNF-α were produced in SV5-infected cells (Fig. 2C). The ELISA was not sensitive enough to detect the small amounts of TNF-α that were produced in SV5-infected cells.

Previously it was shown that caspase-2 and -3 played an essential role in rSV5ΔSH-induced apoptosis in MDBK cells (24). It has been reported that signal transducer and activator of transcription 1 (STAT1), a key regulator for IFN signaling, is essential for caspase-2 and caspase-3 transcription (31). However, SV5 encodes an anti-IFN protein, the V protein that blocks both IFN signaling by targeting STAT1 for degradation (11) and IFN-β production (25, 42). It has been shown that STAT1 was degraded in rSV5ΔSH-infected MDBK cells while rSV5ΔSH induced apoptosis, suggesting that IFN is not likely to be a critical factor in rSV5ΔSH-induced apoptosis. In this work we found that TNF-α played an essential role in rSV5ΔSH-induced apoptosis. Consistent with the results, neutralizing antibodies against IFN did not block CPE induced by rSV5ΔSH in L929 cells (data not shown).

It is not known how increased levels of TNF-α were produced in rSV5ΔSH-infected cells. We hypothesize that the increased expression of TNF-α in rSV5ΔSH-infected cells is a two-step process. The first step is initial induction of TNF-α that is triggered by wild-type SV5 or rSV5ΔSH infection, possibly through activation of NF-κB; the second step is amplification of TNF-α expression through TNF-α itself. The fact that small amounts of TNF-α were produced in the media from SV5-infected cells (Fig. 2C) supports the hypothesis that small amounts of TNF-α are produced in response to virus infection. Because TNF-α is an autocrine, more TNF-α is produced in rSV5ΔSH-infected cells. However, in cells infected by wild-type virus, further production of TNF-α by TNF-α may be blocked by SH. The facts that coinfection of wild-type SV5 with rSV5ΔSH blocked CPE induced by rSV5ΔSH and that the blockage is dosage dependent (24) support the hypothesis.