Phylogenetic analysis of classical swine fever virus in Taiwan

Summary. Two envelope glycoprotein (E^rns and E2) regions of the classical swine fever virus (CSFV) were amplified by RT-PCR and sequenced directly from 158 specimens collected between 1989 and 2003 in Taiwan. Phylogenetic analysis of the two regions revealed a similar tree topology and the E^rns region provided better discrimination than the E2 region. One hundred and fifteen isolates out of the 158 isolates were clustered within subgroup 2.1 (further classified as 2.1a and 2.1b) and 2.2, which were considered to be likely of the introduced strains, whereas the remaining 43 isolates were clustered within subgroup 3.4 and were considered to be of the endemic strains. The subgroup 2.1a viruses were first detected in 1994 and predominated from 1995 onwards. However, subgroup 3.4 viruses were prevalent in the early years, not being isolated after 1996. We have observed a dramatic switch in genotype from subgroup 3.4 to 2.1a. The subgroup 2.1a isolates are closely related to the Paderborn and Lao isolates, whereas 2.1b isolates have a close relationship to the Chinese Guangxi isolates. The phylogenetic tree of 27 CSFV sequences based on the complete envelope glycoprotein gene (E^rns–E2) displayed better resolution than that based on the complete open reading frame.     

Analysis of synonymous codon usage in classical swine fever virus

Using the complete genome sequences of 35 classical swine fever viruses (CSFV) representing all three genotypes and all three kinds of virulence, we analyzed synonymous codon usage and the relative dinucleotide abundance in CSFV. The general correlation between base composition and codon usage bias suggests that mutational pressure rather than natural selection is the main factor that determines the codon usage bias in CSFV. Furthermore, we observed that the relative abundance of dinucleotides in CSFV is independent of the overall base composition but is still the result of differential mutational pressure, which also shapes codon usage. In addition, other factors, such as the subgenotypes and aromaticity, also influence the codon usage variation among the genomes of CSFV. This study represents the most comprehensive analysis to date of CSFV codon usage patterns and provides a basic understanding of the mechanisms for codon usage bias. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Phylogenetic comparison of classical swine fever virus in China.

An N-terminal fragment of the E2 gene of classical swine fever (CSF) virus encoding major immunogenic sites was amplified by RT-PCR directly from 110 clinical specimens representing 109 epizootic sites during the last decade in China. Phylogenetic relationships between these viruses as well as 20 reference strains were determined by comparison of their nucleotide sequences. A phylogenetic tree showed that 103 of the 110 field viruses (93.6%) were clustered within group 2 and subdivided into three subgroups, while the remaining seven viruses (6.4%), along with two Chinese reference strains, Shimen and HCLV (attenuated vaccine strain), were clustered into subgroup 1.1 within group 1. However, none of the Chinese CSF viruses were members of subgroup 1.2 (represented by reference strain Brescia). This is the first report on the distribution of CSF virus genotypes in China. Results indicated that CSF viruses predominating in recent epizootics within China are genetically divergent from the reference strain Shimen and the vaccine strain HCLV.     

Propagation of classical swine fever virus in vitro circumventing heparan sulfate-adaptation

Amplification of natural virus isolates in permanent cell lines can result in adaptation, in particular enhanced binding to heparan sulfate (HS)-containing glycosaminoglycans present on most vertebrate cells. This has been reported for several viruses, including the pestivirus classical swine fever virus (CSFV), the causative agent of a highly contagious hemorrhagic disease in pigs. Propagation of CSFV in cell culture is essential in virus diagnostics and research. Adaptation of CSFV to HS-binding has been related to amino acid changes in the viral E(rns) glycoprotein, resulting in viruses with altered replication characteristics in vitro and in vivo. Consequently, a compound blocking the HS-containing structures on cell surfaces was employed to monitor conversion from HS-independency to HS-dependency. It was shown that the porcine PEDSV.15 cell line permitted propagation of CSFV within a limited number of passages without adaptation to HS-binding. The selection of HS-dependent CSFV mutants was also prevented by propagation of the virus in the presence of DSTP 27. The importance of these findings can be seen from the altered ratio of cell-associated to secreted virus upon acquisition of enhanced HS-binding affinity, a phenotype proposed previously to be related to virulence in the natural host.     

Mutation of E1 glycoprotein of classical swine fever virus affects viral virulence in swine

Transposon linker insertion mutagenesis of a full-length infectious clone (IC) (pBIC) of the pathogenic classical swine fever virus (CSFV) strain Brescia was used to identify genetic determinants of CSFV virulence and host range. Here, we characterize a virus mutant, RB-C22v, possessing a 19-residue insertion at the carboxyl terminus of E1 glycoprotein. Although RB-C22v exhibited normal growth characteristics in primary porcine macrophage cell cultures, the major target cell of CSFV in vivo, it was markedly attenuated in swine. All RB-C22v-infected pigs survived infection remaining clinically normal in contrast to the 100% mortality observed for BICv-infected animals. Comparative pathogenesis studies demonstrated a delay in RB-C22v spread to, and decreased replication in the tonsils, a 10(2) to 10(7) log10 reduction in virus titers in lymphoid tissues and blood, and an overall delay in generalization of infection relative to BICv. Notably, RB-C22v-infected animals were protected from clinical disease when challenged with pathogenic BICv at 3, 5, 7, and 21 days post-RB-C22v inoculation. Viremia, viral replication in tissues, and oronasal shedding were reduced in animals challenged at 7 and 21 DPI. Notably BICv-specific RNA was not detected in tonsils of challenged animals. These results indicate that a carboxyl-terminal domain of E1 glycoprotein affects virulence of CSFV in swine, and they demonstrate that mutation of this domain provides the basis for a rationally designed and efficacious live-attenuated CSF vaccine.     

Human MxA protein inhibits the replication of classical swine fever virus

Classical swine fever virus (CSFV) has a spherical enveloped particle with a single stranded RNA genome, the virus belonging to a pestivirus of the family Flaviviridae is the causative agent of an acute contagious disease classical swine fever (CSF). The interferon-induced MxA protein has been widely shown to inhibit the life cycle of certain RNA viruses as members of the Bunyaviridae family and others. Interestingly, it has been reported that expression of MxA in infected cells was blocked by CSFV and whether MxA has an inhibitory effect against CSFV remains unknown to date until present. Here, we report that CSFV replicated poorly in cells stably transfected with human MxA. The proliferation of progeny virus in both PK-15 cell lines and swine fetal fibroblasts (PEF) continuously expressing MxA was shown significantly inhibited as measured by virus titration, indirect immune fluorescence assay and real-time PCR.     

Evidence for an acid phosphatase in African swine fever virus

Summary An acid phosphatase activity has been detected in purified preparations of African swine fever virus. Purified viral cores obtained after Nonidet-P₄₀ and 2-mercaptoethanol treatment of the virus retained the activity as assayed with nitrophenyl phosphate as substrate at pH 5. Enzyme cytochemistry by electron microscopy showed that the acid phosphatase activity is localized mainly inside the core of the virion. The molecular weight and the isoelectric point of the virus acid phosphatase activity confirmed that it was distinct from the host cellular enzyme.     

Detection of classical swine fever virus in the ovaries of experimentally infected sows

Six sows were infected intranasally with a Korean isolate of classical swine fever virus (CSFV). The distribution of virus in ovarian tissues was then assessed for 21 days by in-situ hybridization and immunohistochemistry. CSFV was detected in the ovaries between 7 and 21 days post-inoculation (dpi) by both methods, but the labelling was particularly intense and widespread at 7 dpi. CSFV nucleic acid and antigen were located almost exclusively within the cytoplasm of cells shown by haematoxylin and eosin staining to be macrophages, which were numerous in atretic follicles. Small numbers of CSFV nucleic acid-positive cells with distinctly round morphology and oval nuclei, resembling monocytes, were also observed in the blood vessels of sows at 7 and 14 dpi. CSFV nucleic acid and antigen were not observed in primordial, primary or secondary follicles from infected sows at 7, 14 or 21 dpi. The results suggest that CSFV replicates in circulating peripheral monocytes and gains access to ovarian tissues from the bloodstream, and that this contributes to the distribution of CSFV in macrophages throughout the atretic follicles.     

Phylogenetic analysis of recent isolates of classical swine fever virus from Colombia

The ability to discriminate between different classical Swine fever virus (CSFV) isolates is a prerequisite for identifying the possible origin of an outbreak. To determine the relatedness between Colombian isolates from different geographical regions, genetic sequences of the glycoprotein E2 and the 5'UTR of CSFV were amplified by PCR, sequenced and compared with reference strains of different genetic grouping. The viruses originated from classical swine fever (CSF) outbreaks in Colombia during 1998-2002. All viruses characterized belonged to genogroup 1 and were members of the subgroup 1.1. The results indicate that the outbreaks from the year 2002 are caused by a strain related to the virus CSF/Santander, isolated in 1980, suggesting that the current CSF outbreaks are the consequence of a single strain that continues to circulate in the field. For the first time, an association between isolates from outbreaks in Colombia in the 1990s was established with a virus isolate from Brazil, indicating a possible origin of the virus causing the outbreak.     

Molecular chaperone Jiv promotes the RNA replication of classical swine fever virus

The nonstructural protein 2 (NS2) of classical swine fever virus (CSFV) is a self-splicing ribozyme wherein the precursor protein NS2-3 is cleaved, and the cleavage efficiency of NS2-3 is crucial to the replication of viral RNA. However, the proteolytic activity of NS2 autoprotease may be achieved through a cellular chaperone called J-domain protein interacting with viral protein (Jiv) or its fragment Jiv90, as evidence suggests that Jiv is required for the proper functioning of the NS2 protein of bovine viral diarrhea virus. Hence, the expression of Jiv may be correlated with the replication efficiency of CSFV RNA. We investigated the expression levels of Jiv and viral RNA in CSFV-infected cells and tissues using Real-time RT-PCR or Western blot analysis. The obtained results show that Jiv90 possibly plays an important role in the lifecycle of CSFV because the distribution of Jiv90 protein shows a positive correlation with the viral load of CSFV. Furthermore, the overexpression or knockdown of Jiv90 in swine cells can also significantly promote or decrease the viral load, respectively. The detection of Flow cytometry shows that the overexpression of Jiv90 prolongs the G1 phase of cell cycles but has no effect on apoptosis. These findings are likely to be of benefit in clarifying the pathogenesis of the CSFV.     

Pathogenicity and kinetics of virus propagation in swine infected with the cytopathogenic classical swine fever virus containing defective interfering particles

Summary.  To analyze the pathogenicity and in vivo kinetics of the cytopathogenic (cp) classical swine fever virus (CSFV) WB82 strain, which is composed of cp defective interfering (DI) particles and noncytopathogenic (noncp) helper virus (WB82/E⁺ strain), WB82 and WB82/E⁺ strains were administered separately to domestic pigs. After inoculation with either strain, all pigs showed typical symptoms of classical swine fever (CSF), such as leucopenia and high fever. There were few differences in clinical signs and survival times between each group. However, the appearance of some symptoms of CSF had a tendency to be delayed following infection with the WB82 strain, when compared with the WB82/E⁺ strain. Virus isolation and detection of subgenomic (sg) and full-length viral (flv) RNA by RT-PCR was carried out using sera, 10% homogenized organs and oral, nasal and rectal swabs. Both noncytopathogenic helper virus and cp DI particles were detected in samples from pigs infected with the WB82 strain, but only noncp phenotype virus was isolated from pigs infected with the WB82/E⁺ strain. Interestingly, the cp DI particles appeared six to seven days later than helper virus in sera from pigs infected with the WB82 strain. Although active cp DI particles could not be isolated from swabs, sg RNA as well as flv RNA was detected in swabs from animals infected with the WB82 strain. These results suggest that progeny cp DI particles are replicated from parent DI particles after noncp virus replication, and subsequently discharged from infected animals. Furthermore, propagation of DI particles or replication of sg RNA, following propagation of helper virus, appears to inhibit the appearance of CSF symptoms induced by virulent helper CSFV. Received January 14, 2002; accepted August 19, 2002     

Identification of an NTPase motif in classical swine fever virus NS4B protein

Classical swine fever (CSF) is a highly contagious and often fatal disease of swine caused by CSF virus (CSFV), a positive-sense single-stranded RNA virus within the Pestivirus genus of the Flaviviridae family. Here, we have identified conserved sequence elements observed in nucleotide-binding motifs (NBM) that hydrolyze NTPs within the CSFV non-structural (NS) protein NS4B. Expressed NS4B protein hydrolyzes both ATP and GTP. Substitutions of critical residues within the identified NS4B NBM Walker A and B motifs significantly impair the ATPase and GTPase activities of expressed proteins. Similar mutations introduced into the genetic backbone of a full-length cDNA copy of CSFV strain Brescia rendered no infectious viruses or viruses with impaired replication capabilities, suggesting that this NTPase activity is critical for the CSFV cycle. Recovered mutant viruses retained a virulent phenotype, as parental strain Brescia, in infected swine. These results have important implications for developing novel antiviral strategies against CSFV infection.     

Complete genome sequence of attenuated low-temperature Thiverval strain of classical swine fever virus

The Thiverval vaccine strain of classical swine fever virus (CSFV) was derived from virulent Alfort strain through the serial passages in cells at 29–30°C. In this study, we determined the complete genome sequence of this strain and found that its genome contains one open reading frame (ORF) that encodes a polyprotein with 3,898 amino acids. The 5′-UTR of Thiverval is 373 nt long with only one mutation at position 220. In contrast, the length of 3′-UTR is highly heterogeneous ranging from 233 to 259 bp. The heterogeneity of length of the 3′-UTR was due to an insertion of a variable length of T-rich sequence ranging from 6 to 32 nt. The insertion may change the structure and free energy of the 3′-UTR, resulting in a destabilization of the 3′-UTR. Sequence alignment of Thiverval and other CSFV strains showed 85.2–99.6% identities at the nucleotide level and 92.5–99.5% at the amino acid level. The phylogenetic tree analysis of the complete ORF, partial region of E2, and NS5B suggests that the CSFV Thiverval strain belongs to genetic group 1 and subgroup 1.1. The results from this study provide insight into the molecular mechanism of the attenuation of Thiverval vaccine strain.     

Evidence for the type of nucleic acid in African swine fever virus

Summary The growth of African swine fever virus (ASFV) in pig kidney cells (PK-15) was inhibited by 5-iodo- and 5-bromo-2-deoxyuridine (IUDR and BUDR), whereas the fluoro-compound was inactive in this respect. In another cell line of the same origin (PK-C) virus inhibition occurred with as little as 1g/ml. of BUDR. The inhibitory effect was reduced when the compound was added more than 6 hours after infection, i.e. about 4 hours preceding the first appearance of progeny virus and was completely reversed in the presence of excess thymidine. Cultures in which virus multiplication was completely inhibited showed no cytopathic effects. Actinomycin-D also suppressed the growth and cytopathogenicity of ASFV in PK-C cells, the minimal inhibitory concentration being 0.2g/ml.; the antibiotic-sensitive phase of virus multiplication was 2–6 hours after infection.The growth of pseudorabies virus was inhibited in parallel with that of ASFV but the multiplication of foot-and-mouth disease virus in PK-C cells was not affected by these compounds.When ASFV was grown in the presence of¹⁴C-thymidine, followed by centrifugal concentration, enzymatic treatment and fractionation in sucrose density gradients, a peak of radioactivity coinciding with that of viral infectivity was obtained. No such peak occurred when the virus was grown in the presence of¹⁴C-uridine.It is concluded that ASFV is a DNA virus and should no longer be considered as a possible myxovirus.     

First molecular identification and characterization of classical swine fever virus isolates from Nepal

Classical swine fever (CSF) is a major constraint to pig production worldwide, and in many developing countries, the epidemiological status is unknown. Here, for the first time, molecular identification and characterization of CSFV isolates from two recent outbreaks in Nepal are presented. Analysis of full-length E2-encoding sequences revealed that these isolates belonged to CSFV subgenotype 2.2 and had highest genetic similarity to isolates from India. Hence, for CSFV, Nepal and India should be regarded as one epidemiological unit. Both Nepalese isolates exhibited significant sequence differences, excluding a direct epidemiological connection and suggesting that CSFV is endemic in that country.     

Characterisation of the discrepancy between PCR and virus isolation in relation to classical swine fever virus detection

In order to confirm and characterise further the discrepancies observed between diagnostic RT-nPCR and virus isolation results for the detection of classical swine fever virus (CSFV), a test panel of three new RT-PCRs was designed, amplifying parts of the NS2, NS3 and NS5A regions. Screening of negative samples by virus isolation with the new panel not only confirmed the discrepancies previously observed but also indicated that these were not associated with a specific genomic region. However, none of the PCR-positive samples were positive on all the different PCRs and preferential amplification was not obtained even when a more sensitive real-time RT-PCR was used. Furthermore, the primer-dependent amplification, most likely caused by the presence of viral fragments, demonstrates the necessity of confirming a single positive PCR result, certainly in the presence of contradictory virus isolation results. The new PCR panel, in combination with sequencing, can be used as a tool to provide additional information on the nature of the viral RNA present in the sample.     

Removal of a N-linked glycosylation site of classical swine fever virus strain Brescia Erns glycoprotein affects virulence in swine

E(rns) glycoprotein, along with E(1) and E(2), is one of the three envelope glycoproteins of classical swine fever virus (CSFV). E(rns) is a heavily glycosylated protein involved in several functions, including virus attachment and entry to target cells, production of neutralizing antibodies, and virulence. The role of added glycans to CSFV strain Brescia E(rns) on virus virulence was assessed in swine. A panel of virus mutants was constructed and used to investigate whether the removal of each of seven putative glycosylation sites in the E(rns) glycoprotein would affect viral virulence in swine. Only N269A/Q substitution rendered attenuated viruses (N1v/N1Qv) that, unlike BICv and other mutants, produced a transient infection in swine characterized by mild symptoms and decreased virus shedding. Notably, N1v efficiently protected swine from challenge with virulent BICv at 3 and 21 days post-infection suggesting that glycosylation of E(rns) could be modified for development of CSF live-attenuated vaccines.