SGM Journals
Research Article

Classical swine fever virus infection protects aortic endothelial cells from pIpC-mediated apoptosis

H. L. Johns, E. Bensaude, S. A. La Rocca, J. Seago, B. Charleston, F. Steinbach, T. W. Drew, H. Crooke, H. Everett

  • 1Virology Department, Veterinary Laboratories Agency, New Haw, Addlestone, Surrey KT15 3NB, UK
  • 2Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Surrey GU24 0NF, UK
  • 3Institute for Animal Health, Compton Laboratory, Compton, Newbury, Berkshire RG20 7NN, UK
  • Correspondence
    H. Everett
    h.everett{at}vla.defra.gsi.gov.uk

    • Journal of General Virology 2010; 91(4):1038–1046 · https://doi.org/10.1099/vir.0.016576-0

    View at publisher PubMed

    Abstract

    Classical swine fever virus (CSFV) causes severe disease in pigs associated with leukopenia, haemorrhage and fever. We show that CSFV infection protects endothelial cells from apoptosis induced by the dsRNA mimic, pIpC, but not from other apoptotic stimuli, FasL or staurosporine. CSFV infection inhibits pIpC-induced caspase activation, mitochondrial membrane potential loss and cytochrome c release as well as the pro-apoptotic effects of truncated Bid (tBid) overexpression. The CSFV proteins Npro and Erns both contribute to CSFV inhibition of apoptosis. We conclude that CSFV infection can inhibit apoptotic signalling at multiple levels, including at the caspase-8 and the mitochondrial checkpoints. By supporting viral replication, endothelial cells may promote CSFV pathogenesis.

    • ↵†Present address: Defra, Area 5C, 9 Millbank, 17 Smith Square, London SW1P 3JR, UK.

    • ↵‡These authors contributed equally to this work.

    INTRODUCTION

    Classical swine fever virus (CSFV), a pestivirus within the family Flaviviridae, is related in terms of its sequence and genome organization to other members of the flavivirus group that give rise to human diseases such as hepatitis C and dengue fever. CSFV is a positive-sense RNA virus, and is the causative agent of classical swine fever (CSF), a notifiable disease of pigs (Le Potier et al., 2006). Disease outbreaks are associated with considerable economic loss and adverse effects on animal welfare (Paton & Greiser-Wilke, 2003). In its most virulent form, CSFV infection causes severe immunopathological signs, including leukopenia, haemorrhage and fever, accompanied by high morbidity and mortality. However, within an infected population, some animals are found to recover or develop chronic disease (Le Potier et al., 2006). These observations suggest a complex interplay between the virus and the host immune system.

    CSFV infection is accompanied by depression of cellular immune defences (Ganges et al., 2008) and particularly innate responses mediated by interferon (Bensaude et al., 2004; La Rocca et al., 2005; Ruggli et al., 2005; Seago et al., 2007). In vivo, elevated levels of cell death by apoptosis are thought to account for the characteristic depletion of circulating leukocytes (Sanchez-Cordon et al., 2002; Summerfield et al., 2000, 2001). However, it has been found that apoptotic cells rarely contain detectable viral antigen, indicating that apoptotic cells are generally not CSFV-infected (Choi et al., 2004; Sato et al., 2000; Summerfield et al., 2001). This finding, together with the observation that acutely infected animals develop high viral loads, suggest a possible role for cell populations that are resistant to virus-induced apoptosis. These may include endothelial cells.

    Apoptosis is a controlled cell death programme employed by multicellular organisms to eliminate damaged, aberrant or infected cells (Benedict et al., 2002). A characteristic of apoptotic cell death is the activation of the caspase family of proteases. Caspases that function near the apex of cell death cascades are designated initiators and include caspases-8 and -9, whereas those involved in the terminal stages of cell death, such as caspases-3 and -7, are termed effector caspases (Cohen, 1997). Bcl-2 family proteins are also key regulators of cell death pathways and can be anti-apoptotic, such as Bcl-2, Bcl-XL and Mcl-1, or pro-apoptotic, such as Bax, Bak and Bid (Youle & Strasser, 2008).

    Cell death can be triggered by signals categorized as extrinsic if extracellular, or intrinsic if originating from within the cell. The prototypal cell death pathway is that of the Fas receptor and its extrinsic cell death ligand, FasL (Nagata, 1999). FasL induces oligomerization of the Fas receptor and formation of the death-inducing signalling complex (DISC) that results in activation of caspase-8. Caspase-8 cleaves Bid to produce active, truncated Bid (tBid), which then recruits pro-apoptotic Bcl-2 proteins to the mitochondria. Mitochondria undergo specific changes associated with propagation of the cell death signal. These changes include loss of inner-membrane potential (ΔΨm) and release of proteins from the intermembrane space, including cytochrome c. Cytochrome c forms a complex with caspase-9 that in turn activates caspases-3 and -7, thereby committing a cell to apoptosis. Intrinsic cell death processes frequently activate the cascade at the level of mitochondria (Rodriguez & Lazebnik, 1999).

    The effectiveness of apoptosis as an antiviral mechanism is illustrated by the diversity of countermeasures that viruses have developed to neutralize this host response (Benedict et al., 2002). RNA viruses potentially trigger apoptosis by activating cellular sensors designed to detect extraneous RNA derived from viral genomes or replicative intermediates. A well-described mechanism of cellular recognition of dsRNA is that mediated by PKR, which results in the inhibition of translation followed by apoptosis due to lack of protein expression (Garcia et al., 2007). Studies have also suggested a direct role for PKR in inducing apoptosis through caspase-8 and -9 (Iordanov et al., 2005a, b). Other dsRNA sensors implicated in inducing apoptosis include Toll-like receptor (TLR) 3 in endosomes (Salaun et al., 2006) and cytoplasmic RNA helicases such as RIG-I (Rintahaka et al., 2008). All of these are intracellular RNA sensors; however, a role for extracellular RNA as a biologically relevant molecule for signalling to these sensors, perhaps through scavenger receptors, has been suggested (Dinger et al., 2008). In addition, production of extracellular RNA is a known consequence of RNA virus infection (Majde et al., 1998).

    Previously, it was found that CSFV infection of primary porcine endothelial cells could counteract the pro-apoptotic effects of the dsRNA mimic pIpC (Bensaude et al., 2004). However, little was known concerning the mechanistic aspects of this process. It has been suggested that dsRNA such as viral replicative intermediates or pIpC could be degraded by CSFV proteins, one candidate being Erns, which has been shown, in the related pestivirus BVDV, to prevent pIpC-mediated activation of the interferon-stimulated Mx protein (Iqbal et al., 2004; Magkouras et al., 2008; Matzener et al., 2009). Interestingly, it was also reported that expression of the viral protein Npro alone is sufficient to block pIpC-mediated cell death (Ruggli et al., 2003, 2005). We therefore investigated, in more detail, the mechanism of apoptotic protection afforded by CSFV infection as well as the roles played by two viral proteins, Npro and Erns.

    RESULTS

    CSFV infection prevents pIpC-induced apoptosis of AOC cells

    We first investigated whether CSFV infection could circumvent dsRNA-induced apoptotic signalling in a previously characterized porcine aortic endothelial cell line (AOC) (Carrillo et al., 2002). AOC were infected with CSFV strain Alfort 187, and 24 h post-infection they were treated with increasing concentrations of pIpC for 4 h. Activity of the effector caspases, caspases-3 and -7, measured using a luminescent caspase activity assay, increased proportionately with pIpC concentration and reached a plateau at and above 100 μg ml−1 in uninfected cells. However, no increase in caspase-3/7 activity was seen in CSFV-infected cells (Fig. 1⇓). These results confirm that, like primary porcine endothelial cells, the AOC cell line is susceptible to dsRNA-induced apoptosis and that CSFV infection protects against this.

    Figure image not available in archive
    Fig. 1.

    Uninfected or 24 h CSFV-infected AOC were treated with increasing concentrations of pIpC for 4 h. Caspase-3/7 activity was measured using a luminescence assay and is expressed as mean arbitrary relative light units (RLU) ± sd for triplicate samples.

    CSFV infection inhibits the mitochondrial apoptotic pathway

    The release of cytochrome c from mitochondria is a critical factor in the activation of effector caspases. We therefore investigated whether apoptotic regulatory steps such as mitochondrial membrane potential (ΔΨm) loss and cytochrome c release were inhibited in CSFV-infected cells. CSFV-infected or uninfected AOC were treated with 200 μg pIpC ml−1 for 1, 3 or 5 h. To investigate whether the inhibition of apoptosis was specific to the dsRNA-induced pathway, we also treated cells with the apoptosis inducer staurosporine and the extrinsic death ligand FasL. Standard treatments of incubation with 2.5 μM staurosporine for 4.5 h, or with 10 ng ml−1 human membrane-bound FasL for 22 h were used. Cells were labelled with 20 nM of the mitochondrion-specific, potentiometric dye tetramethylrhodamine methyl ester (TMRM) during the final 20 min. Flow-cytometric analysis revealed that loss of ΔΨm, reflected by a reduction in the percentage of TMRM-positive cells, occurred following pIpC addition to uninfected but not to CSFV-infected cells (Fig. 2a⇓). In contrast, ΔΨm loss occurred in both infected and uninfected cells following FasL or staurosporine treatment (Fig. 2b⇓).

    Figure image not available in archive
    Fig. 2.

    (a, b) Effects of different apoptotic inducers on uninfected and CSFV-infected AOC, assessed by monitoring the loss of mitochondrial inner membrane potential (ΔΨm) using the potentiometric dye TMRM. The percentage of cells with a ΔΨm loss relative to controls is expressed as the mean±sd. (a) CSFV infection was found to block ΔΨm loss due to pIpC treatment, whereas (b) both infected and uninfected cells lost ΔΨm following FasL or staurosporine treatment. (c, d) Apoptosis-dependent redistribution of cytochrome c in uninfected or CSFV-infected AOC, monitored by immunoblot detection of cytochrome c in the cytosolic fraction of cells following treatment with (c) pIpC, or with (d) FasL, staurosporine (STS) or DMSO as a solvent control. CSFV infection was identified by the presence of the viral Npro protein, and equivalent protein loading was assessed by using an anti-γ-tubulin antibody.

    In parallel, the release of cytochrome c from mitochondria into the cytosol was monitored by immunoblotting of cytosolic fractions, following differential digitonin lysis. The appearance of cytochrome c in the cytosolic fraction was clearly detected in uninfected but not CSFV-infected cells treated with pIpC (Fig. 2c⇑). In contrast, infected cells were not protected against cytochrome c release following FasL or staurosporine treatment (Fig. 2d⇑).

    CSFV infection partially inhibits tBid-mediated apoptosis

    To investigate whether the anti-apoptotic effect of CSFV is exerted at the mitochondria or further upstream in the pathway, we looked at whether CSFV could inhibit the pro-apoptotic effects of truncated Bid (tBid) overexpression. Bid is cleaved by caspase-8 to give tBid, a pro-apoptotic protein that causes Bax oligomerization and insertion into the mitochondrial membrane, resulting in (ΔΨm) loss and cytochrome c release. CSFV-infected and uninfected AOC were transfected with a plasmid expressing human tBid fused to EGFP (Fig. 3a⇓), or EGFP alone (Fig. 3b⇓). Eight hours post-transfection, cells were fixed and labelled for CSFV E2 (red). Cell nuclei were visualized using DAPI (blue). Expression of EGFP-tBid (green) induced apoptosis in uninfected AOC, as demonstrated by nuclear chromatin condensation, whereas in CSFV-infected cells, the nuclei were intact (Fig. 3a⇓). Expression of EGFP alone had no effect on the nuclear morphology of uninfected or CSFV-infected cells (Fig. 3b⇓).

    Figure image not available in archive
    Fig. 3.

    (a, b) Uninfected or CSFV-infected AOC were transfected with a plasmid encoding pro-apoptotic tBid fused with EGFP (a) or EGFP alone (b) (green). Eight hours post-transfection, cells were fixed and labelled for CSFV E2 protein as a marker of virus infection (red). Nuclei were visualized using DAPI (blue). Scale bars, 10 μm. (c) Uninfected or CSFV-infected PK15 cells were mock-transfected, or transfected with EGFP alone or with EGFP-tBid. Caspase-3/7 activity was measured using a luminescence assay and is expressed as arbitrary mean relative light units (RLU) ± sd for triplicate samples. CSFV infection was able to partially inhibit caspase-3/7 activation due to tBid overexpression. *Statistically significant using Student's t-test (P<0.05).

    As the transfection efficiency of AOC is routinely less than 10 % as assessed by confocal microscopy, the effect of tBid expression was quantified using a porcine kidney cell line (PK15) that had a transfection efficiency of 30 % in these experiments. Caspase-3/7 activity was measured in CSFV-infected and uninfected PK15 cells expressing human tBid fused to EGFP. Caspase-3/7 activity was measured 6 h post-transfection (Fig. 3c⇑). Control transfections with no DNA or a plasmid encoding EGFP only were included to show the basal level of caspase-3/7 activity, and that the presence of the EGFP did not significantly increase apoptosis. Expression of tBid led to an increase in caspase-3/7 activity in the uninfected cells. The CSFV-infected cells showed a 48 % reduction in caspase-3/7 activity induced by tBid compared to uninfected cells. Transfection efficiencies for CSFV-infected and uninfected cells were equivalent. This result suggests that CSFV can partially protect cells from an intrinsic apoptotic trigger acting directly on mitochondria.

    CSFV infection inhibits caspase activation following pIpC treatment

    We next investigated which caspases are affected by CSFV infection. The activities of initiator caspases-8 and -9 and effector caspases-3 and -7 were measured at intervals after treatment of infected and uninfected AOC with 200 μg pIpC ml−1 (Fig. 4⇓). Following treatment, activity of all caspases monitored increased in uninfected cells but remained low in CSFV-infected cells. To investigate whether the CSFV inhibition of caspases, like the effect on mitochondria, is limited to the dsRNA-induced pathway, two other apoptotic stimuli were also used. CSFV-infected and uninfected cells were treated with FasL for 22 h or with staurosporine for 4.5 h. Comparable increases in caspase activity were observed in CSFV-infected and uninfected cells following both treatments, suggesting that CSFV infection specifically prevents caspase activation induced by dsRNA.

    Figure image not available in archive
    Fig. 4.

    Uninfected or CSFV-infected AOC were treated with 200 μg pIpC ml−1 over 5 h, with 10 ng ml−1 FasL for 22 h, or with with 2 μM staurosporine (STS) for 4.5 h. Activities of caspases-3/7, -9 and -8 were monitored with specific luminescence assays and are expressed as relative increase compared to time 0 (caspases) or untreated cells (FasL, STS), ± sd for triplicate samples.

    A parental CSFV infectious clone virus inhibits caspase-3 activation, unlike infectious clone viruses with an Npro deletion or Erns mutation (C-H346Δ)

    It is known that the protein Npro is involved in CSFV-mediated cell survival in response to pIpC (Ruggli et al., 2003). In order to investigate whether Npro is acting via a caspase-dependent pathway induced by pIpC, an Npro-deleted virus (EP#96/2) derived from an infectious clone plasmid and its parental strain (EP#98/2) were obtained (G. Meyers, FLI, Tübingen, Germany). As Npro has a role in the inhibition of type I interferon production (La Rocca et al., 2005; Ruggli et al., 2005; Seago et al., 2007), the Npro-deleted virus is more easily propagated in SK6 cells, as they are deficient in type I interferon production (Ruggli et al., 2003). To investigate a possible role of Erns in inhibition of apoptosis, we used a further infectious clone virus, TF230/1, derived from the same parental CSFV strain as the Npro mutant. This Erns mutant (C-H346Δ) has an amino acid deletion (His346) and is impaired in its RNase activity (von Freyburg et al., 2004). The presence of this mutation was confirmed by sequencing. SK6 cells were uninfected, or infected with parental CSFV, Npro-deleted CSFV or Erns C-H346Δ CSFV; 96 h after the initial infection, cells were treated with pIpC for 0, 1, 2 or 3 h before cell lysates were analysed by immunoblotting. Fig. 5(a)⇓ shows that activation of caspase-3 could be detected in uninfected cells following 3 h of pIpC treatment. Only low levels of active caspase-3 could be detected in cells infected with the parental CSFV, whereas marked caspase-3 activation was evident in cells infected with the Npro-deleted CSFV, indicating that this virus is impaired in its ability to inhibit caspase-3 activation in response to pIpC. Interestingly, following pIpC treatment, levels of E2 viral coat protein remained constant in cells infected with the parental CSFV but diminished rapidly in cells infected with Npro-deleted virus. This occurred despite a single flask of cells being infected with the respective viruses prior to preparation of individual cultures for the different treatments. Immunoblot analysis of untreated cells (t=0) also showed equivalent levels of E2, and lysates contained equivalent amounts of CSFV RNA based on qPCR (data not shown). These results suggest that, in the absence of Npro, viral E2 is depleted over time in response to pIpC. Fig. 5(b)⇓ shows a separate experiment comparing uninfected SK6 cells, parental CSFV-infected cells and Erns C-H346Δ CSFV-infected cells under the same pIpC treatment as Fig. 5(a)⇓. Active caspase-3 was produced earlier in this experiment, becoming detectable at 2 h then increasing at 3 h. However, as in Fig. 5(a)⇓, the parental CSFV-infected cells did not respond to pIpC. The cells infected with Erns C-H346Δ CSFV had higher levels of active caspase-3, although again less than uninfected cells. This suggests that functional Erns is required for inhibition of caspase-3 activation by pIpC. Interestingly, no depletion of E2 was observed in SK6 cells infected with the Erns C-H346Δ mutant virus, although caspase-3 levels were comparable with those seen in the Npro-deleted virus-infected cells.

    Figure image not available in archive
    Fig. 5.

    (a) Apoptosis-dependent activation of caspase-3 following treatment with pIpC, monitored in lysates of uninfected SK6 cells or cells infected with parental or Npro-deleted (ΔNpro) CSFV infectious clones. Immunoblot detection of CSFV E2 protein identified virus infection and the absence of Npro in Npro-deleted virus-infected cells was verified. Equivalent protein loading was assessed using an anti-γ-tubulin antibody. Using an antibody directed against active caspase-3, CSFV infection was found to inhibit caspase-3 activation following pIpC treatment, but this did not occur in uninfected and Npro-deleted CSFV-infected cells. (b) Apoptosis-dependent activation of caspase-3 following treatment with pIpC was monitored in lysates of uninfected SK6 cells or cells infected with parental or Erns C-H346Δ CSFV infectious clones. Detection of E2 and Npro protein identified virus infection. Equivalent protein loading was assessed using an anti-γ-tubulin antibody. Using an antibody directed against active caspase-3, parental CSFV infection was found to inhibit caspase-3 activation following pIpC treatment, but this did not occur in uninfected and Erns C-H346Δ CSFV-infected cells.

    DISCUSSION

    Endothelial cells have long been known to be highly sensitive to immune-activating effects of extracellular dsRNA (Doukas et al., 1994). In the context of CSFV infection, with associated haemorrhage and inflammation, it could be expected that cell lysis would result in the release of double-stranded viral replicative intermediates, which are potentially important signalling molecules (Dinger et al., 2008). This study demonstrates that CSFV infection of a porcine endothelial cell line exerts a protective effect against dsRNA-mediated cell death. We found that pIpC treatment of AOC activates caspase-8, as previously reported in human cells (Kaiser et al., 2004; Takahashi et al., 2006), and that CSFV infection inhibits the activation of this key initiator caspase as well as the downstream mitochondrial pathway involving caspase-9 activation, loss of mitochondrial membrane potential, release of cytochrome c and caspase-3/7 activation. CSFV infection does not appear to target caspase activation directly, as cell death triggered by two alternative pro-apoptotic stimuli, Fas ligand and staurosporine, was not inhibited in CSFV-infected cells. These results suggest that the inhibition of apoptosis by CSFV in response to dsRNA is at a stage upstream of caspase-8. It could be that CSFV is able to inhibit signalling by dsRNA sensors such as endosomal TLR3, or cytoplasmic RIG-I. One example of such a mechanism can be observed during HCV infection, where the TLR3 adaptor protein Trif is cleaved by viral protease NS3/4A (Li et al., 2005).

    We also investigated the effect of CSFV on mitochondria-initiated apoptosis. Infection with CSFV clearly reduced the pro-apoptotic effects of tBid overexpression, which directly targets the mitochondria. This suggests that there is an additional inhibition target at the mitochondria as well as upstream of caspase-8. Many viral proteins have been shown to inhibit apoptosis at the level of mitochondria, often through binding to pro- or anti-apoptotic proteins (Galluzzi et al., 2008).

    We used an Npro-deleted virus in order to examine in more detail the observation that Npro supports the survival of CSFV-infected cells (Ruggli et al., 2005). Our data also indicate a role for Npro in the inhibition of pIpC-dependent apoptosis. We have also shown that, in cells infected with the Npro-deleted virus, viral E2 itself is depleted in response to dsRNA treatment and this could represent an additional antiviral mechanism. Two major functions have been assigned to Npro, one being its protease activity (Rumenapf et al., 1998) and the other its ability to target IRF3 for proteasomal degradation (Bauhofer et al., 2007; La Rocca et al., 2005; Seago et al., 2007). It remains to be seen which of these functions, if either, is required for apoptosis inhibition by CSFV. Active IRF3 induces expression of a number of pro-apoptotic genes (Andersen et al., 2007), and this may be inhibited in CSFV-infected cells. The observation that apoptosis can be induced in SK6 cells, reported to have a defective interferon response (Ruggli et al., 2003), implies that the two pathways may be independent in these cells. It has also been suggested that inhibition of apoptosis by pestiviruses could be due to Erns, a viral protein with RNase activity that is most effective against ssRNA substrates at acidic pH. Erns has been found to block interferon-inducible Mx protein activation following dsRNA treatment (Iqbal et al., 2004; Magkouras et al., 2008). Our results, using the CSFV Erns C-H346Δ infectious clone virus, suggest that the RNase activity of Erns also contributes to the effective inhibition of caspase-3 activation in response to pIpC. However, the E2 depletion seen with the Npro mutant virus is not evident in cells infected with the Erns mutant virus. This argues against a general caspase-3-dependent protein cleavage event and highlights the need for further investigation into the mechanism of pIpC-induced apoptosis inhibition by each of the viral proteins. It is reasonable to suggest that active Npro and Erns each contribute to inhibition of apoptosis by CSFV, acting through different mechanisms. Our readout for apoptosis in these experiments was detection of active caspase-3, a marker for irreversible commitment to cell death. This is at the end of the apoptotic cascade and therefore most likely downstream of the step in the signalling pathway(s) targeted.

    We conclude that CSFV infection is able to inhibit pIpC-mediated apoptosis at multiple levels, including the caspase-8 and mitochondrial checkpoints. This multi-tiered protective effect employed by CSFV has also been observed with other viruses (Benedict et al., 2002). The presence of an apoptosis-resistant, infected cell population, such as endothelial cells, is consistent with the high viral loads detected in acutely infected animals. Endothelial cells are important immune modulators (Bensaude et al., 2004; Campos et al., 2004) and so infection of these cells is likely to play a key role in the immunosuppressant and haemorrhagic effects of CSFV.

    METHODS

    Virus and cells.

    CSFV strain Alfort 187 was kindly provided by the EU reference laboratory, Hannover, Germany. Parental infectious clone EP#98/2 derived from CSFV strain Alfort Tübingen, the corresponding Npro deleted infectious clone EP#96/2 and Erns C-H346Δ clone TF230/1 were a kind gift from Dr G. Meyers (FLI, Tübingen, Germany). Porcine aortic endothelial cells (AOC) (Carrillo et al., 2002) were cultured in RPMI with glutamax (Gibco) plus 10 % fetal calf serum (FCS), 50 IU penicillin ml−1, 50 μg streptomycin ml−1 and 0.25 μg fungizone ml−1 (Fisher). Porcine kidney cells (PK15), were cultured in MEM with glutamax plus 10 % FCS, 50 IU penicillin ml−1, 50 μg streptomycin ml−1 and 25 U nystatin ml−1. A porcine kidney cell line defective in the production of interferon (SK6) (Ruggli et al., 2003) was cultured in DMEM with glutamax (Gibco) plus 10 % FCS, 50 IU penicillin ml−1, 50 μg streptomycin ml−1 and 25 U nystatin ml−1.

    Virus infection.

    Virus was adsorbed to AOC at an m.o.i. of between 0.1 and 0.5 for 2 h at 37 °C and cells were incubated in complete medium for 24 h. Cells were then trypsinized, seeded into separate wells and treated with inducers of apoptosis beginning 24 h later, by which time the infection rate exceeded 98 % as verified by immunofluorescence.

    Caspase assays.

    CSFV-infected or uninfected AOC (5×105 cells) were incubated in fresh medium containing 200 μg polyinosinic-polycytidylic acid ml−1 (pIpC, Sigma) for up to 5 h, containing 2.5 μM staurosporine (Sigma) for 4.5 h, or containing 10 ng ml−1 membrane-bound FasL (Upstate) of human origin, which required an incubation time of 22 h to produce a measureable effect. Control cells were left untreated but in the case of staurosporine treatment, were treated with DMSO solvent. Caspase activity was measured using Caspase-Glo assay kits specific for caspase-3/7, caspase-8 or caspase-9 (Promega).

    Measurement of mitochondrial membrane potential and cytochrome c release.

    Apoptosis was induced as described for the caspase assays. To measure mitochondrial membrane potential, cells were labelled with 20 nM of the mitochondrion-specific potentiometric dye tetramethylrhodamine methyl ester (TMRM, Molecular Probes) during the final 20 min. Fluorescence of 20 000 cells per sample was analysed using a FASCalibur flow cytometer (Becton Dickinson). To measure cytochrome c release, cells were pelleted and subjected to differential digitonin lysis (Heibein et al., 1999). Immunoblot analysis of cytosolic fractions was conducted using a monoclonal antibody against cytochrome c (BD Pharmingen) and a rabbit polyclonal antibody, DS14 (La Rocca et al., 2005), to detect viral Npro as a marker of CSFV infection. Equivalent protein loading was assessed using an anti-γ-tubulin antibody (Sigma, clone GTU-88).

    tBid overexpression.

    For confocal microscopy, CSFV-infected or uninfected AOC were seeded on 13 mm borosilicate coverslips (VWR). After 24 h, cells were transiently transfected with a plasmid expressing human truncated Bid (tBid) fused to EGFP (a kind gift from Dr P. Powell), or EGFP alone, using Lipofectamine LTX (Invitrogen). Eight hours post-transfection, cells were fixed with 4 % paraformaldehyde and permeabilized with 0.1 % Triton. CSFV was labelled using the monoclonal antibody WH303 directed against the E2 viral coat protein (Edwards & Sands, 1990) and an Alexa 594-conjugated goat anti-mouse secondary antibody (Molecular Probes). Images were captured using a Leica SP2 confocal microscope.

    For quantification of apoptosis, caspase-3/7 activity was measured in CSFV-infected and uninfected PK15 cells transiently transfected with a plasmid expressing tBid fused to EGFP, or EGFP alone, using Lipofectamine 2000 (Invitrogen). Transfection efficiency was measured based on EGFP fluorescence using a Bioanalyser (Agilent). Caspase-3/7 activity was measured 6 h post-transfection using the Caspase-Glo kit (Promega).

    Analysis of CSFV infectious clones.

    Uninfected SK6 cells, or cells infected with Alfort/Tübingen-based infectious clone viruses EP#98/2 (parental), EP#96/2 (Npro deleted) and TF230/1 (Erns C-H346Δ) (von Freyburg et al., 2004), were cultured for 24 h, trypsinized and seeded into separate wells. After 48 h, cells were treated with 200 μg pIpC ml−1 (Sigma) for 0, 1, 2 or 3 h. Cells were pelleted and lysed using RIPA buffer (Upstate) before separation by SDS-PAGE and immunoblot detection with a monoclonal antibody to active caspase-3 (Cell Signaling). Viral Npro was labelled using the DS14 polyclonal rabbit polyclonal antibody and the anti-E2 monoclonal antibody WH303 was used as a marker of CSFV infection. Equivalent protein loading was assessed using an anti-tubulin antibody (Sigma, clone GTU-88).

    Acknowledgments

    This study was funded by Project SE0785 from Defra. The authors wish to thank Dr Penny Powell for supplying the EGFP-tBid construct and Dr Gregor Meyers for providing the infectious clone viruses (EP#96/2, EP#98/2 and TF230/1).

    References

    1. Andersen, J., VanScoy, S., Cheng, T. F., Gomez, D. & Reich, N. C. (2007). IRF-3-dependent and augmented target genes during viral infection. Genes Immun 9, 168–175.
    2. Bauhofer, O., Summerfield, A., Sakoda, Y., Tratschin, J. D., Hofmann, M. A. & Ruggli, N. (2007). Classical swine fever virus Npro interacts with interferon regulatory factor 3 and induces its proteasomal degradation. J Virol 81, 3087–3096.
    3. Benedict, C. A., Norris, P. S. & Ware, C. F. (2002). To kill or be killed: viral evasion of apoptosis. Nat Immunol 3, 1013–1018.
    4. Bensaude, E., Turner, J. L., Wakeley, P. R., Sweetman, D. A., Pardieu, C., Drew, T. W., Wileman, T. & Powell, P. P. (2004). Classical swine fever virus induces proinflammatory cytokines and tissue factor expression and inhibits apoptosis and interferon synthesis during the establishment of long-term infection of porcine vascular endothelial cells. J Gen Virol 85, 1029–1037.
    5. Campos, E., Revilla, C., Chamorro, S., Alvarez, B., Ezquerra, A., Dominguez, J. & Alonso, F. (2004). In vitro effect of classical swine fever virus on a porcine aortic endothelial cell line. Vet Res 35, 625–633.
    6. Carrillo, A., Chamorro, S., Rodriguez-Gago, M., Alvarez, B., Molina, M. J., Rodriguez-Barbosa, J. I., Sanchez, A., Ramirez, P., Munoz, A. & other authors (2002). Isolation and characterization of immortalized porcine aortic endothelial cell lines. Vet Immunol Immunopathol 89, 91–98.
    7. Choi, C., Hwang, K. K. & Chae, C. (2004). Classical swine fever virus induces tumor necrosis factor-α and lymphocyte apoptosis. Arch Virol 149, 875–889.
    8. Cohen, G. M. (1997). Caspases: the executioners of apoptosis. Biochem J 326, 1–16.
    9. Dinger, M. E., Mercer, T. R. & Mattick, J. S. (2008). RNAs as extracellular signaling molecules. J Mol Endocrinol 40, 151–159.
    10. Doukas, J., Cutler, A. H. & Mordes, J. P. (1994). Polyinosinic : polycytidylic acid is a potent activator of endothelial cells. Am J Pathol 145, 137–147.
    11. Edwards, S. & Sands, J. J. (1990). Antigenic comparisons of hog cholera virus isolates from Europe, America and Asia using monoclonal antibodies. Dtsch Tierarztl Wochenschr 97, 79–81.
    12. Galluzzi, L., Brenner, C., Morselli, E., Touat, Z. & Kroemer, G. (2008). Viral control of mitochondrial apoptosis. PLoS Pathog 4, e1000018
    13. Ganges, L., Nunez, J. I., Sobrino, F., Borrego, B., Fernandez-Borges, N., Frias-Lepoureau, M. T. & Rodriguez, F. (2008). Recent advances in the development of recombinant vaccines against classical swine fever virus: cellular responses also play a role in protection. Vet J 177, 169–177.
    14. Garcia, M. A., Meurs, E. F. & Esteban, M. (2007). The dsRNA protein kinase PKR: virus and cell control. Biochimie 89, 799–811.
    15. Heibein, J. A., Barry, M., Motyka, B. & Bleackley, R. C. (1999). Granzyme B-induced loss of mitochondrial inner membrane potential (ΔΨm) and cytochrome c release are caspase independent. J Immunol 163, 4683–4693.
    16. Iordanov, M. S., Kirsch, J. D., Ryabinina, O. P., Wong, J., Spitz, P. N., Korcheva, V. B., Thorburn, A. & Magun, B. E. (2005a). Recruitment of TRADD, FADD, and caspase 8 to double-stranded RNA-triggered death inducing signaling complexes (dsRNA-DISCs). Apoptosis 10, 167–176.
    17. Iordanov, M. S., Ryabinina, O. P., Schneider, P. & Magun, B. E. (2005b). Two mechanisms of caspase 9 processing in double-stranded RNA- and virus-triggered apoptosis. Apoptosis 10, 153–166.
    18. Iqbal, M., Poole, E., Goodbourn, S. & McCauley, J. W. (2004). Role for bovine viral diarrhea virus Erns glycoprotein in the control of activation of beta interferon by double-stranded RNA. J Virol 78, 136–145.
    19. Kaiser, W. J., Kaufman, J. L. & Offermann, M. K. (2004). IFN-α sensitizes human umbilical vein endothelial cells to apoptosis induced by double-stranded RNA. J Immunol 172, 1699–1710.
    20. La Rocca, S. A., Herbert, R. J., Crooke, H., Drew, T. W., Wileman, T. E. & Powell, P. P. (2005). Loss of interferon regulatory factor 3 in cells infected with classical swine fever virus involves the N-terminal protease, Npro. J Virol 79, 7239–7247.
    21. Le Potier, M., Mesplede, A. & Vannier, P. (2006). Classical swine fever and other pestiviruses. In Diseases of Swine, 9th edn, pp. 309–322. Edited by B. E. Straw, J. J. Zimmerman, S. D'Allaire & D. J. Taylor. Ames, Iowa: Blackwell.
    22. Li, K., Foy, E., Ferreon, J. C., Nakamura, M., Ferreon, A. C. M., Ikeda, M., Ray, S. C., Gale, M. & Lemon, S. M. (2005). Immune evasion by hepatitis C virus NS3/4A protease-mediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc Natl Acad Sci U S A 102, 2992–2997.
    23. Magkouras, I., Matzener, P., Rumenapf, T., Peterhans, E. & Schweizer, M. (2008). RNase-dependent inhibition of extracellular, but not intracellular, dsRNA-induced interferon synthesis by Erns of pestiviruses. J Gen Virol 89, 2501–2506.
    24. Majde, J. A., Guha-Thakurta, N., Chen, Z., Bredow, S. & Krueger, J. M. (1998). Spontaneous release of stable viral double-stranded RNA into the extracellular medium by influenza virus-infected MDCK epithelial cells: implications for the viral acute phase response. Arch Virol 143, 2371–2380.
    25. Matzener, P., Magkouras, I., Rumenapf, T., Peterhans, E. & Schweizer, M. (2009). The viral RNase Erns prevents IFN type-I triggering by pestiviral single- and double-stranded RNAs. Virus Res 140, 15–23.
    26. Nagata, S. (1999). FAS ligand-induced apoptosis. Annu Rev Genet 33, 29–55.
    27. Paton, D. J. & Greiser-Wilke, I. (2003). Classical swine fever – an update. Res Vet Sci 75, 169–178.
    28. Rintahaka, J., Wiik, D., Kovanen, P. E., Alenius, H. & Matikainen, S. (2008). Cytosolic antiviral RNA recognition pathway activates caspases 1 and 3. J Immunol 180, 1749–1757.
    29. Rodriguez, J. & Lazebnik, Y. (1999). Caspase-9 and APAF-1 form an active holoenzyme. Genes Dev 13, 3179–3184.
    30. Ruggli, N., Tratschin, J. D., Schweizer, M., McCullough, K. C., Hofmann, M. A. & Summerfield, A. (2003). Classical swine fever virus interferes with cellular antiviral defense: evidence for a novel function of Npro. J Virol 77, 7645–7654.
    31. Ruggli, N., Bird, B. H., Liu, L., Bauhofer, O., Tratschin, J. D. & Hofmann, M. A. (2005). Npro of classical swine fever virus is an antagonist of double-stranded RNA-mediated apoptosis and IFN-α/β induction. Virology 340, 265–276.
    32. Rumenapf, T., Stark, R., Heimann, M. & Thiel, H.-J. (1998). N-terminal protease of pestiviruses: identification of putative catalytic residues by site-directed mutagenesis. J Virol 72, 2544–2547.
    33. Salaun, B., Coste, I., Rissoan, M. C., Lebecque, S. J. & Renno, T. (2006). TLR3 can directly trigger apoptosis in human cancer cells. J Immunol 176, 4894–4901.
    34. Sanchez-Cordon, P. J., Romanini, S., Salguero, F. J., Nunez, A., Bautista, M. J., Jover, A. & Gomez-Villamos, J. C. (2002). Apoptosis of thymocytes related to cytokine expression in experimental classical swine fever. J Comp Pathol 127, 239–248.
    35. Sato, M., Mikami, O., Kobayashi, M. & Nakajima, Y. (2000). Apoptosis in the lymphatic organs of piglets inoculated with classical swine fever virus. Vet Microbiol 75, 1–9.
    36. Seago, J., Hilton, L., Reid, E., Doceul, V., Jeyatheesan, J., Moganeradj, K., McCauley, J., Charleston, B. & Goodbourn, S. (2007). The Npro product of classical swine fever virus and bovine viral diarrhea virus uses a conserved mechanism to target interferon regulatory factor-3. J Gen Virol 88, 3002–3006.
    37. Summerfield, A., Knoetig, S. M., Tschudin, R. & McCullough, K. C. (2000). Pathogenesis of granulocytopenia and bone marrow atrophy during classical swine fever involves apoptosis and necrosis of uninfected cells. Virology 272, 50–60.
    38. Summerfield, A., Zingle, K., Inumaru, S. & McCullough, K. C. (2001). Induction of apoptosis in bone marrow neutrophil-lineage cells by classical swine fever virus. J Gen Virol 82, 1309–1318.
    39. Takahashi, K., Kawai, T., Kumar, H., Sato, S., Yonehara, S. & Akira, S. (2006). Roles of caspase-8 and caspase-10 in innate immune responses to double-stranded RNA. J Immunol 176, 4520–4524.
    40. von Freyburg, M., Ege, A., Saalmuller, A. & Meyers, G. (2004). Comparison of the effects of RNase-negative and wild-type classical swine fever virus on peripheral blood cells of infected pigs. J Gen Virol 85, 1899–1908.
    41. Youle, R. J. & Strasser, A. (2008). The BCL-2 protein family: opposing activities that mediate cell death. Nat Rev Mol Cell Biol 9, 47–59.