RNase-dependent inhibition of extracellular, but not intracellular, dsRNA-induced interferon synthesis by Erns of pestiviruses

Abstract

Recombinant pestivirus envelope glycoprotein E^rns has been shown to interfere with dsRNA-induced interferon (IFN-α/β) synthesis. This study demonstrated that authentic, enzymically active E^rns produced in mammalian cells prevented a dsRNA-induced IFN response when present in the supernatant of bovine cells. Strikingly, IFN synthesis of cells expressing E^rns was eliminated after extracellular addition, but not transfection, of dsRNA. Importantly, the same applied to cells infected with bovine viral diarrhea virus (BVDV) expressing E^rns but lacking the N-terminal protease N^pro. Free E^rns concentrations circulating in the blood of animals persistently infected with BVDV were determined to be approximately 50 ng ml^–1, i.e. at a similar order of magnitude as that displaying an effect on dsRNA-induced IFN expression in vitro. Whilst N^pro blocks interferon regulatory factor-3-dependent IFN induction in infected cells, E^rns may prevent constant IFN induction in uninfected cells by dsRNA that could originate from pestivirus-infected cells. This probably contributes to the survival of persistently BVDV-infected animals and maintains viral persistence in the host population.

We express our gratitude to Munir Iqbal and John W. McCauley (Division of Molecular Biology, Institute for Animal Health, Compton, UK) for providing recombinant, insect cell-derived E^rns, Martin Beer (Institute of Diagnostic Virology, Federal Research Institute for Animal Health, Greifswald-Insel Riems, Germany) for providing the BVDV strain Ncp7 lacking N^pro, and to O. Haller (Institute for Medical Microbiology and Hygiene, University of Freiburg, Germany) for supplying the Mx antibody. This work was supported by the Swiss National Science Foundation, grant no. 3100A0-109597 (to E. P.) and 3200-068305 (to M. S.). T. R. is funded by the Sonderforschungsbereich 535 der Deutschen Forschungsgemeinschaft (DFG). We thank R. Parham for critical reading of the manuscript.

Footnotes

^,†, Philippe Mätzener¹ †These authors contributed equally to this work.

Bovine viral diarrhea virus (BVDV), a member of the family Flaviviridae and a major ubiquitous cattle pathogen, causes two fundamentally different types of infection. Animals infected post-natally are transiently infected, mostly without showing obvious disease signs such as diarrhoea and coughing (Corapi et al., 1990; Meyers & Thiel, 1996; Ridpath et al., 2000; Thiel et al., 1996). When infected in utero as a result of maternal infection, fetuses may develop and be born normally. They remain infected for life and display a highly specific immunotolerance towards the infecting viral strain (Brownlie, 1990). Immunotolerance of persistently infected (PI) animals is explained by the early time point of infection between approximately 40 and 120 days of intrauterine development. In contrast to the adaptive immune response, innate immune defence mechanisms are operative even in the early stages of intrauterine development when the fetus is invaded by BVDV, and there is broad but indirect evidence that evasion of the host's interferon (IFN) defence is a crucial prerequisite for the establishment and maintenance of persistent infection (Adler et al., 1997; Charleston et al., 2001; Schweizer & Peterhans, 2001).

Over the past few years, it has become apparent that virtually all viruses express proteins that target the host's IFN defence mechanisms. Collectively, these proteins target either the induction pathways or the mechanism of IFN action (for a review, see Randall & Goodbourn, 2008). Induction of IFN is initiated by extracellular or intracellular pattern recognition receptors that sense molecular patterns that indicate viral infection (Beutler et al., 2007). In the case of positive- and negative-sense RNA viruses, dsRNA, a by-product of viral RNA replication, and 5'-triphosphorylated RNA, respectively, are believed to be the most important IFN inducers (Pichlmair & Sousa, 2007). Once the extracellular or intracellular pattern recognition receptors are activated, a signal pathway is initiated that results in IFN-α/β transcription and release of IFN from infected cells. IFN then binds to the type I receptor, with the ensuing signal cascade leading to the formation of over 100 IFN-stimulated proteins that mediate the antiviral effect (Der et al., 1998).

BVDV encodes two gene products that have been implicated in IFN evasion. The non-structural protein N^pro, a protease encoded at the 5' end of the 12.5 kb viral genome, has been shown to target the transcription factor interferon regulatory factor (IRF)-3 for proteasomal degradation (Hilton et al., 2006; Seago et al., 2007). The second protein, E^rns, is a highly N-glycosylated endoRNase that forms disulfide-linked homodimers. E^rns is an essential structural component of pestivirus particles but is also secreted from infected cells (Rümenapf et al., 1993; Schneider et al., 1993). Using E^rns produced in insect cells, it has been shown that it interacts with extracellular dsRNA, thereby targeting a major viral IFN-inducing signal (Iqbal et al., 2004). Insect cell-produced proteins significantly differ in the state of glycosylation and also functionally from their counterparts synthesized in mammalian cells (Kost et al., 2005). Therefore, we studied the mechanism and site of action of E^rns expressed in bovine cells and E^rns in the context of intact virus.

E^rns was expressed in bovine Madin–Darby bovine kidney (MDBK) Tet-On cells using a tetracycline-inducible expression plasmid as described previously (Krey et al., 2005). Briefly, wild-type (wt) and mutant E^rns cDNA was obtained from BVDV strain Ncp7 or classical swine fever virus (CSFV) strain Alfort by RT-PCR. H30F and H30R mutations in BVDV and CSFV E^rns, respectively, were introduced using a QuikChange mutagenesis kit (Stratagene). MDBK Tet-On cell lines stably transfected with E^rns were selected using G418 and puromycin. The resulting MDBK Tet-On/E^rns cell clones were monitored for inducible E^rns expression by immunoperoxidase staining with a monoclonal antibody (mAb 50F4-10) to E^rns at 48 h after the addition of 1 µg doxycycline ml^–1. MDBK-expressed BVDV Ncp7 E^rns was identical in apparent molecular mass to that secreted by BVDV-infected cells. Furthermore, MDBK Tet-On/E^rns cells allowed the propagation of BVDV strain Ncp7 with a deletion of the entire E^rns gene (Ncp7-ΔE^rns), suggesting its functional authenticity.

Initially, we compared the dose–response curves for inactivation of IFN induction by insect cell-grown E^rns (kindly provided by M. Iqbal, Institute for Animal Health, UK) with that expressed by MDBK cells. We found that insect cell-expressed recombinant E^rns of BVDV strain Pe515 (Iqbal et al., 2004) was able to prevent IFN induction in bovine turbinate (BT) cells (isolated as described by Schweizer & Peterhans, 2001) stimulated with the synthetic dsRNA poly(IC). As shown by the absence of Mx, a widely used sensitive marker for the activity of IFN (von Wussow et al., 1990) that can also be used in bovine cells (Schweizer & Peterhans, 2001; Schweizer et al., 2006), E^rns inhibited the induction of Mx in a dose-dependent fashion, as analysed by Western blotting. Complete inhibition was achieved at concentrations above 250 ng ml^–1 when the cells were stimulated with 1 µg poly(IC) ml^–1 (Fig. 1a). Next, the effect of insect cell-expressed E^rns was compared with that harvested from bovine cells. The latter was concentrated by ultracentrifugation of the cell-culture supernatant using a 10 kDa cut-off membrane (Sartorius Vivaspin). The concentration of E^rns was determined with a commercially available ELISA (Idexx Laboratories), taking purified insect cell-expressed E^rns as the standard. The RNase activity of E^rns was determined as described previously (Iqbal et al., 2004) using poly(U) as the substrate and was found to be similar to that of its counterpart expressed in the baculovirus system in insect cells (Fig. 1d). The bovine cell-expressed protein inhibited the effect of poly(IC) in a concentration range similar to that of insect cell-expressed E^rns (Fig. 1a and b). The supernatant of MDBK Tet-On cells, which constitutively express the reverse Tet-responsive transcriptional activator alone, served as a negative control and failed to inhibit the effect of poly(IC) (Fig. 1b). To assess the role of the RNase activity of E^rns, we used supernatants from MDBK Tet-On cells stably transfected with an E^rns plasmid containing a mutation of the catalytic residue His-30 to Phe (E^rns-H30F), resulting in a complete loss of enzymic activity (Fig. 1d). RNase-inactive E^rns failed to inhibit the effect of poly(IC), even when tested at high concentrations (9 µg ml^–1) in cells challenged with low amounts (1 µg ml^–1) of poly(IC) (Fig. 1c). Similarly, control supernatants of MDBK Tet-On cells expressing no E^rns that were concentrated in parallel to E^rns-H30F were inactive against poly(IC) and did not induce Mx when tested in the absence of poly(IC) (Fig. 1c).

(23K):

Fig. 1. Extracellular E^rns inhibits Mx induction in BT cells induced by poly(IC). Different concentrations of recombinant (rec) E^rns expressed in insect cells (a) or isolated from the supernatant of MDBK Tet-On/wt E^rns cells (b) and from MDBK Tet-On/E^rns H30F cells (c) were added with or without poly(IC) to BT cells for 18 h. Cytosolic protein extracts were analysed by Western blotting using an anti-Mx mAb (exposure time 4–16 min) and an anti-β-actin mAb (exposure time 10–30 s) as loading control. Cell-culture medium (MEM, a–c) and the supernatant of the MDBK Tet-On cells, used at equal (eq.) volumes as used for the E^rns samples (b and c), were used as negative controls. One representative experiment of three is shown. (d) Different concentrations of wt E^rns and RNase-inactive E^rns (E^rns-H30F) isolated from the supernatant of the MDBK Tet-On cell lines were assayed for RNase activity. The absorbance (A₂₆₀) at the lowest concentration of wt E^rns exhibiting maximal activity was measured for all samples at the same dilution. As positive controls, 0.2 µg insect cell-derived E^rns (rec E^rns) and 0.6 U RNase One were used, whilst the supernatant of MDBK Tet-On cells (Tet-On) and RNase assay buffer (Buffer) served as negative controls. The results shown are the means±SD of two independent experiments (for RNase One, n=6).

To test whether the concentrations that are active at inhibiting the effect of poly(IC) in the bioassays in vitro were present in vivo, we quantified free E^rns in the serum of BVDV PI animals after removing virus particles using ultrafiltration with a 300 kDa membrane (Sartorius Vivaspin). Quantification was carried out with an ELISA using insect cell-expressed E^rns as a standard as described above, and the absence of infectious virus was monitored by virus isolation. We found approximately 50 ng free E^rns ml^–1 in the blood of these animals (50.3±20.5 ng ml^–1, n=3), which was approximately in the same order of magnitude as the concentration active in the in vitro experiments. Notably, E^rns in the serum of PI animals as well as that produced by MDBK cells had a similar specific RNase activity to the recombinant insect cell-grown protein (not shown).

Regardless of its biological origin, enzymically active E^rns added to the medium was capable of interfering with the effect of dsRNA on Mx induction. To study the possible site of action, we used cells expressing E^rns or cells that were infected with the non-cytopathic BVDV strain Ncp7 and compared the effect of the addition of poly(IC) to the supernatant with that of lipofected poly(IC). MDBK Tet-On cells containing or not wt or RNase^– E^rns plasmids were stimulated for 48 h with doxycycline before removing the supernatant and adding 100 µg poly(IC) ml^–1 to the supernatant or lipofecting with 1 µg poly(IC) ml^–1 (Schweizer & Peterhans, 2001). At 24 h post-treatment, whole-cell extracts were prepared and assayed for Mx and IRF-3 by Western blotting. All cells expressed Mx in response to transfected poly(IC) and to recombinant bovine (rbo) IFN-α added to the supernatant as a positive control (Fig. 2). Mx expression was only eliminated when poly(IC) was added to the supernatant of cells expressing wt but not RNase^– E^rns of BVDV strain Ncp7 (Fig. 2a, b) or of CSFV strain Alfort (not shown). Remarkably, inhibition of Mx expression was observed even when poly(IC) was added to cells in fresh medium after removing the E^rns-containing medium and washing the cells. Importantly, both IFN-α and lipofected poly(IC) induced Mx as well as IRF-3 above the background level observed in mock-treated or mock-lipofected cultures. Cells infected with BVDV strain Ncp7 (m.o.i. of 3), in contrast to the other treatments, were clearly negative for both IRF-3 and Mx expression (Fig. 2). The results obtained in cells expressing RNase^– E^rns, as well as those of MDBK Tet-On cells, again indicated that the RNase activity of E^rns was essential for the elimination of IFN induction by dsRNA. Moreover, the fact that virus infection prevented Mx induction and at the same time led to the complete disappearance of IRF-3 suggested that the effect of intact virus may be independent of E^rns. Previous work has suggested that the effect on IRF-3 of pestiviruses may be mediated through N^pro, which targets IRF-3 for proteasomal degradation (Bauhofer et al., 2007; Chen et al., 2007; Hilton et al., 2006; Seago et al., 2007).

(45K):

Fig. 2. E^rns expressed as a single protein in MDBK cells inhibits Mx synthesis induced by extracellular but not intracellular poly(IC). MDBK Tet-On cells expressing wt E^rns (a) or an RNase-inactive mutant E^rns (b), or MDBK Tet-On cells as a control (c), were stimulated with doxycycline for 48 h. Thereafter, the supernatant was replaced with fresh medium containing either 100 µg poly(IC) ml^–1 (pIC-SN) or 30 ng recombinant bovine (rbo)IFN-α ml^–1, or 1 µg poly(IC) ml^–1 was transfected using lipofectin (pIC-Lipo). Lipofectin (Lipo) and medium alone (Mock) were used as controls. Alternatively, MDBK cells were infected with BVDV strain Ncp7 at an m.o.i. of 3. Whole-cell extracts were prepared 24 h later, and the expression of Mx (exposure time 1 min) and IRF-3 (exposure time 3 min) proteins was analysed by Western blotting. Parallel staining for β-actin (exposure time 5 s) served as a control for protein loading of individual lanes. One representative experiment of three is shown.

To assess the functions of the two proteins in the context of intact virus, rather than being expressed as single proteins, we made use of viral mutants lacking either N^pro or the RNase activity of E^rns (H30F). As shown previously (Schweizer & Peterhans, 2001), Mx (and hence IFN) production induced by extracellular and intracellular dsRNA was completely inhibited in cells infected with wt BVDV (Fig. 3a and b), provided all cells were infected, as analysed by immunoperoxidase staining (not shown). In cells infected with the BVDV strain lacking N^pro (Ncp7-ΔN^pro), Mx synthesis stimulated by up to 100 µg poly(IC) ml^–1 added extracellularly in fresh medium after removing the cell culture supernatant was also completely blocked, whereas no inhibition was observed upon transfection of dsRNA (Fig. 3c). Conversely, mutant viruses with RNase^– E^rns but still encoding N^pro were able to inhibit intracellularly as well as extracellularly applied dsRNA-induced responses and replicated to titres similar to those of the parental wt strain (not shown). This confirmed that the RNase activity of E^rns is not essential for viral growth in vitro. Mx synthesis induced by rboIFN-α was not reduced in cells infected with either strain, indicating that signalling through the IFN I receptor was not influenced by viral infection (Fig. 3) (Schweizer et al., 2006). It is worth noting that infection of BT cells with BVDV Ncp7-ΔN^pro induced Mx expression in the absence of stimulation with poly(IC) or rboIFN-α as early as approximately 2 days post-infection (not shown), as has also been reported for CSFV and BVDV (Gil et al., 2006; Ruggli et al., 2003). Thus, the inhibition of extracellularly applied dsRNA by Ncp7-ΔN^pro could not be demonstrated in BT cells. However, due to a lower replication efficiency in MDBK cells, it was only at about 5 days post-infection that we observed Mx expression by the mutant virus in the absence of any other stimulus (not shown), which enabled us to analyse the effect of poly(IC) at earlier time points (Fig. 3).

(28K):

Fig. 3. BVDV lacking N^pro but expressing E^rns is able to block Mx synthesis induced by extracellularly but not intracellularly applied dsRNA. MDBK cells were mock infected (a) or infected with the Ncp7 (Ncp7-wt) BVDV strain (b) or a mutant lacking N^pro (Ncp7-ΔN^pro) (c) and incubated for 55–65 h at 37 °C until all cells were infected. Thereafter, the supernatant was removed and 10 ng rboIFN-α ml^–1 or poly(IC) at concentrations of 0–100 µg ml^–1 were added in fresh medium. For the intracellular application of dsRNA, poly(IC) was transfected at 1 µg ml^–1 (Tr-1) with lipofectin. Lipofectin alone (Lipo) was used as a negative control. Cytosolic extracts were prepared 24 h later, and the expression of Mx protein (exposure time 3 min) and β-actin (exposure time 10 s) was analysed by Western blotting. One representative experiment of three is shown.

In conclusion, our experiments confirm and extend the previous demonstration that E^rns acts by targeting extracellular dsRNA, a major viral signal triggering IFN synthesis, as shown by E^rns expression in bovine cells but also using intact virus. This provides compelling evidence that N^pro and E^rns are non-redundant IFN antagonistic proteins of pestiviruses. In PI animals, E^rns reaches a concentration in the blood that implies that a biological function in viral persistence is highly likely. This view is compatible with the observation that both E^rns and N^pro, the latter inhibiting IFN production in infected cells, seem to be essential for the establishment of persistent infection in utero (Meyers et al., 2007). As not all cells are infected in PI animals, free E^rns may bind and degrade BVDV dsRNA that might emerge from infected cells, as has been described for influenza virus-infected MDCK cells (Majde et al., 1998), thereby preventing continued induction of IFN synthesis in uninfected cells. As systemic IFN production is well known to lead to fever and other disease signs (Pichler, 2006), free E^rns may contribute to maintaining PI animals as efficient virus replicators, which are essential for the persistence of BVDV in the cattle population.

References

Adler, B., Adler, H., Pfister, H., Jungi, T. W. & Peterhans, E. (1997). Macrophages infected with cytopathic bovine viral diarrhea virus release a factor(s) capable of priming uninfected macrophages for activation-induced apoptosis. J Virol 71, 3255–3258.[Abstract/Free Full Text]

Bauhofer, O., Summerfield, A., Sakoda, Y., Tratschin, J.-D., Hofmann, M. A. & Ruggli, N. (2007). Classical swine fever virus N^pro interacts with interferon regulatory factor 3 and induces its proteasomal degradation. J Virol 81, 3087–3096.[Abstract/Free Full Text]

Beutler, B., Eidenschenk, C., Crozat, K., Imler, J. L., Takeuchi, O., Hoffmann, J. A. & Akira, S. (2007). Genetic analysis of resistance to viral infection. Nat Rev Immunol 7, 753–766.[CrossRef][Medline]

Brownlie, J. (1990). Pathogenesis of mucosal disease and molecular aspects of bovine virus diarrhoea virus. Vet Microbiol 23, 371–382.[CrossRef][Medline]

Charleston, B., Fray, M. D., Baigent, S., Carr, B. V. & Morrison, W. I. (2001). Establishment of persistent infection with non-cytopathic bovine viral diarrhoea virus in cattle is associated with a failure to induce type I interferon. J Gen Virol 82, 1893–1897.[Abstract/Free Full Text]

Chen, Z. H., Rijnbrand, R., Jangra, R. K., Devaraj, S. G., Qu, L., Ma, Y. H., Lemon, S. M. & Li, K. (2007). Ubiquitination and proteasomal degradation of interferon regulatory factor-3 induced by N^pro from a cytopathic bovine viral diarrhea virus. Virology 366, 277–292.[CrossRef][Medline]

Corapi, W. V., Elliott, R. D., French, T. W., Arthur, D. G., Bezek, D. M. & Dubovi, E. J. (1990). Thrombocytopenia and hemorrhages in veal calves infected with bovine viral diarrhea virus. J Am Vet Med Assoc 196, 590–596.[Medline]

Der, S. D., Zhou, A. M., Williams, B. R. G. & Silverman, R. H. (1998). Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc Natl Acad Sci U S A 95, 15623–15628.[Abstract/Free Full Text]

Gil, L. H. V. G., Ansari, I. H., Vassilev, V., Liang, D. L., Lai, V. C. H., Zhong, W. D., Hong, Z., Dubovi, E. J. & Donis, R. O. (2006). The amino-terminal domain of bovine viral diarrhea virus N^pro protein is necessary for alpha/beta interferon antagonism. J Virol 80, 900–911.[Abstract/Free Full Text]

Hilton, L., Moganeradj, K., Zhang, G., Chen, Y. H., Randall, R. E., McCauley, J. W. & Goodbourn, S. (2006). The NPro product of bovine viral diarrhea virus inhibits DNA binding by interferon regulatory factor 3 and targets it for proteasomal degradation. J Virol 80, 11723–11732.[Abstract/Free Full Text]

Iqbal, M., Poole, E., Goodbourn, S. & McCauley, J. W. (2004). Role for bovine viral diarrhea virus E^rns glycoprotein in the control of activation of beta interferon by double-stranded RNA. J Virol 78, 136–145.[Abstract/Free Full Text]

Kost, T. A., Condreay, J. P. & Jarvis, D. L. (2005). Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nat Biotechnol 23, 567–575.[CrossRef][Medline]

Krey, T., Thiel, H. J. & Rümenapf, T. (2005). Acid-resistant bovine pestivirus requires activation for pH-triggered fusion during entry. J Virol 79, 4191–4200.[Abstract/Free Full Text]

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.[CrossRef][Medline]

Meyers, G. & Thiel, H.-J. (1996). Molecular characterization of pestiviruses. In Advances in Virus Research, pp. 53–118. Edited by K. Maramorosch, F. A. Murphy & A. J. Shatkin. San Diego: Academic Press.

Meyers, G., Ege, A., Fetzer, C., Von Freyburg, M., Elbers, K., Carr, V., Prentice, H., Charleston, B. & Schürmann, E. M. (2007). Bovine viral diarrhea virus: prevention of persistent fetal infection by a combination of two mutations affecting E^rns RNase and N^pro protease. J Virol 81, 3327–3338.[Abstract/Free Full Text]

Pichler, W. J. (2006). Adverse side-effects to biological agents. Allergy 61, 912–920.[CrossRef][Medline]

Pichlmair, A. & Sousa, C. R. E. (2007). Innate recognition of viruses. Immunity 27, 370–383.[CrossRef][Medline]

Randall, R. E. & Goodbourn, S. (2008). Interferons and viruses: an interplay between induction, signalling, antiviral responses and virus countermeasures. J Gen Virol 89, 1–47.[Abstract/Free Full Text]

Ridpath, J. F., Neill, J. D., Frey, M. & Landgraf, J. G. (2000). Phylogenetic, antigenic and clinical characterization of type 2 BVDV from North America. Vet Microbiol 77, 145–155.[CrossRef][Medline]

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 N^pro. J Virol 77, 7645–7654.[Abstract/Free Full Text]

Rümenapf, T., Unger, G., Strauss, J. H. & Thiel, H.-J. (1993). Processing of the envelope glycoproteins of pestiviruses. J Virol 67, 3288–3294.[Abstract/Free Full Text]

Schneider, R., Unger, G., Stark, R., Schneider-Scherzer, E. & Thiel, H. J. (1993). Identification of a structural glycoprotein of an RNA virus as a ribonuclease. Science 261, 1169–1171.[Abstract/Free Full Text]

Schweizer, M. & Peterhans, E. (2001). Noncytopathic bovine viral diarrhea virus inhibits double-stranded RNA-induced apoptosis and interferon synthesis. J Virol 75, 4692–4698.[Abstract/Free Full Text]

Schweizer, M., Mätzener, P., Pfaffen, G., Stalder, H. P. & Peterhans, E. (2006). "Self" and "nonself" manipulation of interferon defense during persistent infection: bovine viral diarrhea virus resists alpha/beta interferon without blocking antiviral activity against unrelated viruses replicating in its host cells. J Virol 80, 6926–6935.[Abstract/Free Full Text]

Seago, J., Hilton, L., Reid, E., Doceul, V., Jeyatheesan, J., Moganeradj, K., McCauley, J., Charleston, B. & Goodbourn, S. (2007). The N^pro 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.[Abstract/Free Full Text]

Thiel, H.-J., Plagemann, P. G. W. & Moennig, V. (1996). Pestiviruses. In Fields Virology, 3rd edn, pp. 1059–1073. Edited by B. N. Fields, D. M. Knipe & P. M. Howley. Philadelphia & New York: Lippincott–Raven Publishers.

von Wussow, P., Jakschies, D., Hochkeppel, H.-K., Fibich, C., Penner, L. & Deicher, H. (1990). The human intracellular Mx-homologous protein is specifically induced by type I interferons. Eur J Immunol 20, 2015–2019.[Medline]

Received 29 April 2008; accepted 16 June 2008.