- Volume 82, Issue 6, 2001

A live attenuated porcine reproductive and respiratory syndrome (PRRS) vaccine virus has been shown to revert to virulence under field conditions. In order to identify genetic virulence determinants, ORF1 from the attenuated vaccine virus and three Danish vaccine-derived field isolates was sequenced and compared with the parental strain of the vaccine virus (VR2332). This revealed five mutations that had occurred independently in all three vaccine-derived field isolates, indicating strong parallel selective pressure on these positions in the vaccine virus when used in swine herds. Two of these parallel mutations were direct reversions to the parental VR2332 sequence and were situated in a papain-like cysteine protease domain and in the helicase domain. The remaining parallel mutations might be seen as second-site compensatory mutations for one or more of the mutations that accumulated in the vaccine virus sequence during cell-culture adaptation. Evaluation of the remaining mutations in the ORF1 sequence revealed stronger selective pressure for amino acid conservation during spread in pigs than during vaccine production. Furthermore, it was found that the selective pressure did not change during the time period studied. The implications of these findings for PRRS vaccine attenuation and reversion are discussed.

The coronavirus genome is a positive-strand RNA of extraordinary size and complexity. It is composed of approximately 30000 nucleotides and it is the largest known autonomously replicating RNA. It is also remarkable in that more than two-thirds of the genome is devoted to encoding proteins involved in the replication and transcription of viral RNA. Here, a reverse-genetic system is described for the generation of recombinant coronaviruses. This system is based upon the in vitro transcription of infectious RNA from a cDNA copy of the human coronavirus 229E genome that has been cloned and propagated in vaccinia virus. This system is expected to provide new insights into the molecular biology and pathogenesis of coronaviruses and to serve as a paradigm for the genetic analysis of large RNA virus genomes. It also provides a starting point for the development of a new class of eukaryotic, multi-gene RNA vectors that are able to express several proteins simultaneously.

Recombination events are known to occur in non-segmented RNA viruses like polioviruses or alphaviruses. Analysis of the subgenomic sequences of dengue virus type 1 (DENV-1) structural genes has recently allowed the identification of possible recombination breakpoints. Because DENV is a major human pathogen, this discovery might have important implications for virus pathogenicity, vaccine safety and efficiency, or diagnosis and, therefore, requires clear confirmation. We report the complete sequence determination of one Asian and two African strains of DENV-1 isolated from human patients. Rigorous sequence analysis provided strong evidence for the occurrence of intragenomic recombination events between DENV-1 strains belonging to different lineages. Singapore S275/90 strain appears to be the evolutionary product of a recombination event between viruses belonging to two distinct lineages: one lineage includes an African strain isolated in Abidjan (Ivory Coast) and the other includes isolates from Djibouti and Cambodia. The ‘Recombination Detection Program’, bootscanning and analysis of diversity plots provided congruent results concerning the existence of a two-switch recombination event and the localization of recombination breakpoints. Thus, the 5′ and 3′ genomic ends of the Singapore S275/90 strain were inherited from a Djibouti/Cambodia lineage ancestor and an internal fragment located in the envelope/NS1 region originated from an Abidjan lineage ancestor.

A full-length cDNA clone of the hepatitis C virus (HCV) genotype 1 prototype (subtype 1a) sequence was constructed. Synthetic RNA produced from the initial cDNA clone was not infectious following intrahepatic inoculation of a chimpanzee. A consensus clone was prepared by comparison with multiple full-length HCV sequences of genotypes 1, 2 and 3. A total of 11 non-consensus amino acid residues were altered by mutagenesis. Synthetic RNA from the repaired clone initiated a typical, acute-resolving HCV infection following intrahepatic inoculation of a chimpanzee. In addition, at least one of three chimeric cDNA clones constructed between the HCV-1 and H77 genotype 1a strains of HCV was infectious in a chimpanzee. This is the first example of an infectious chimeric HCV clone. An infectious cDNA clone of HCV-1 will be of particular value, since it is the prototype HCV sequence and many commonly used reagents are based on this sequence.

The non-structural 3 (NS3) protein is one of the most conserved proteins of hepatitis C virus, and T helper 1 (Th1)-like responses to NS3 in humans correlate with clearance of infection. Several studies have proposed that DNA-based immunizations are highly immunogenic and prime Th1-like responses, although few head-to-head comparisons with exogenous protein immunizations have been described. A full-length NS3/NS4A gene was cloned in eukaryotic vectors with expression directed to different subcellular compartments. Inbred mice were immunized twice in regenerating tibialis anterior (TA) muscles with either plasmid DNA or recombinant NS3 (rNS3). After two 100 μg DNA immunizations, specific antibody titres of up to 12960 were detected at week 5, dominated by IgG2a and IgG2b. NS3-specific CD4+ T cell responses in DNA-immunized mice peaked at day 13, as measured by proliferation and IL-2 and IFN-γ production. Mice immunized with 1–10 μg rNS3 without adjuvant developed antibody titres comparable to those of the DNA-immunized mice, but dominated instead by IgG1. CD4+ T cell responses in these mice showed peaks of IL-2 response at day 3 and IL-6 and IFN-γ responses at day 6. With adjuvant, rNS3 was around 10-fold more immunogenic with respect to speed and magnitude of the immune responses. Thus, immunization with rNS3 in adjuvant is superior to DNA immunization with respect to kinetics and quantity in priming specific antibodies and CD4+ T cells. However, as a DNA immunogen, NS3 elicits stronger Th1-like immune responses, whereas rNS3 primes a mixed Th1/Th2-like response regardless of the route, dose or adjuvant.

The pathogenesis of bone marrow atrophy during classical swine fever (CSF) was investigated in vitro by using CSF virus (CSFV) infection of bone marrow haematopoietic cells (BMHC). The monocytic lineage had the highest susceptibility to CSFV infection, whereas the more mature SWC8+ granulocytic cells were not directly susceptible to infection. However, myelomonocytic precursors were targets for CSFV infection and continued to differentiate into SWC8+ granulocytic cells, which remained infected. This explains the occurrence of infected peripheral blood granulocytes during CSF. The infection of BMHC resulted in increased apoptosis and necrosis, mainly within the granulocytic lineage. Caspases 3 and 9 were particularly activated, relating to the mitochondrial pathway of apoptosis. Interestingly, the majority of infected cells were non-apoptotic, the apoptotic cells being primarily non-infected. This indicated an indirect mechanism for induction of apoptosis, but no role could be identified for bone marrow stroma cells such as macrophages or fibroblastoid cells. Furthermore, soluble factors including cytokines and reactive oxygen species were not primarily responsible. In contrast, contact between infected and non-infected BMHC was critical for increasing apoptosis in the latter. Taken together, these results in vitro relate to and help to explain further the apoptosis of BMHC that occurs in vivo during CSF. This experimental system will also be particularly useful for the study of CSFV gene products involved in leukocyte apoptosis.

In this study, tick-borne encephalitis (TBE) viruses from Siberia and far-eastern Asia were characterized in order to determine virus subtype distribution. TBE viruses were isolated from ticks (Ixodes persulcatus) collected in the far-eastern (Khabarovsk and Vladivostok) and Siberian (Irkutsk) regions of Russia in 1999. Phylogenetic analysis showed that isolates formed distinct clusters of far-eastern and Siberian subtypes. There was also a minor difference in antigenicity between the Irkutsk isolates and other TBE virus strains, as demonstrated by the reactivity of monoclonal antibodies. Amino acid alignments of the E gene showed that the Irkutsk isolates had a single amino acid change at position 234 (Q or H); this amino acid position is considered to be a ‘signature’ of Siberian subtype TBE viruses. Strains isolated in Irkutsk also exhibited equivalent or somewhat higher virulence in mice compared with far-eastern TBE virus isolates. All viruses isolated in this study (i.e. far-east Asian and Siberian isolates) have 3′ non-coding regions (NCRs) of almost the same length, which contrasts with the various sizes of 3′NCRs of other TBE viruses strains reported previously. The data presented in this study show that the 3′NCR is uniform among TBE viruses isolated from Siberia and far-eastern Asia and that the 3′NCR is essential for TBE virus growth in tick and/or rodent host cells.

Vaccine-associated paralytic poliomyelitis is a serious concern while using the live attenuated oral polio vaccine for the eradication of poliomyelitis. The bonnet monkey model of poliovirus central nervous system (CNS) infection following experimental inoculation into the ulnar nerve allows the comparative study of wild-type and attenuated poliovirus invasiveness. Dosages ⩾104 TCID50 of Mahoney strain of poliovirus type 1 [PV1(M)] result in paralysis. In contrast, even with 107 TCID50 of Sabin attenuated strain of poliovirus type 1 (LSc/2ab), no paralysis occurs, but virus spreads into the CNS where viral RNA is found in spinal cord neurons. While wild-type PV1(M) viral RNA replicates in neurons (and possibly in glial cells) and in cells around vessel walls, which may be mononuclear or endothelial cells, attenuated viral RNA is detected only in neurons. Systemic viraemia and gastrointestinal virus shedding occurs only in PV1(M)-infected animals. While a systemic serologic response is detected in both groups of animals, cerebrospinal fluid antibodies are detected only in animals infected with PV1(M). Both the PV1(M) and LSc/2ab strains spread to the cervical spinal cord and then to the lumbar spinal cord following ulnar nerve inoculation. Neuronophagia and neuronal loss are only seen in PV1(M)-infected monkeys in whom clinical paralysis is observed. Infection with LSc/2ab does not result in neuronophagia, neuronal loss or clinical paralysis. Spread of attenuated poliovirus in spinal cord neurons without causing paralysis following inoculation into the ulnar nerve is an important finding.

The secondary structure of the 3′ untranslated region (3′UTR) of picornaviruses is thought to be important for the initiation of negative-strand RNA synthesis. In this study, genetic and biological analyses of the 3′ terminus of coxsackievirus B2 (CVB2), which differs from other enteroviruses due to the presence of five additional nucleotides prior to the poly(A) tail, is reported. The importance of this extension was investigated using a 3′UTR mutant lacking the five nucleotides prior to the poly(A) tail and containing two point mutations. The predicted secondary structure within the 3′UTR of this mutant was less energetically favourable compared with that of the wild-type (wt) genotype. This mutant clone was transfected into green monkey kidney cells in four parallel experiments and propagated for multiple passages, enabling the virus to establish a stable revertant genotype. Genetic analysis of the virus progeny from these different passages revealed two major types of revertant. Both types showed wt-like growth properties and more stable and wt-like predicted secondary structures than the parent mutant clone. The first type of revertant neutralized the introduced point mutation with a compensatory second-site mutation, whereas the second type of revertant partly compensated for the deletion of the five proximal nucleotides by the insertion of nucleotides that matched the wt sequence. Therefore, the extended 3′ end of CVB2 may be considered to be a stabilizing sequence for RNA secondary structure and an important feature for the virus.

The defence of a host against viral infections is strongly influenced by cytokines. We investigated the role of the B-cell stimulating cytokines IL-5 and IL-6, and the immuno-suppressive cytokine IL-10, during primary and secondary immune responses in mice against infection with Friend retrovirus (FV) (Murine leukaemia virus). IL-5−/− mice were comparable to C57BL/6 wild-type mice in their ability to control acute FV infection. In contrast, IL-6−/− and IL-10−/− mice showed significantly enhanced virus loads in spleen cells. However, this impaired control of acute FV replication did not alter the long-term control over persistent FV in IL-6−/− and IL-10−/− mice. Immunization with a live attenuated vaccine virus prior to challenge protected all three types of cytokine-deficient mice from high levels of spleen virus, despite the finding that the vaccinated IL-5- and IL-6-deficient mice had significantly reduced titres of virus-neutralizing IgG class antibodies. The results indicate that IL-6 and IL-10 contribute to primary immune responses against FV, but are dispensable during persistent infection and vaccine-primed secondary responses.

Peripheral blood leukocytes (PBLs) and tissue samples from 36 sheep were examined for jaagsiekte sheep retrovirus (JSRV) by hemi-nested PCR. Animals were classified according to the status of sheep pulmonary adenomatosis (SPA), which was confirmed by pathological examination, as follows: (i) sheep with classical SPA (cSPA, n=10), (ii) sheep with atypical SPA (aSPA, n=6), (iii) non-affected sheep from SPA-affected flocks (in-contact, n=10) and (iv) non-affected sheep from SPA-free flocks (control, n=10). JSRV proviral DNA was detected in the PBLs of 10/10 cSPA, 5/6 aSPA, 4/10 in-contact and 0/10 control sheep. Lung tumours and lymphoid organs were also found to be JSRV-positive. The number of positive PCR results was greater for sheep in the cSPA group than for those in the aSPA and in-contact groups. For the first time, it is concluded that JSRV can be detected in naturally infected sheep before the onset of clinical disease and even before the development of discernible tumours.

Two conflicting in vitro observations suggest that retroviral recombinations are temperature dependent. Ouhammouch & Brody (Nucleic Acids Research 20, 5443–5450, 1992) suggested that retroviral recombination rates should increase as temperature increases. However, Shimomaye & Salvato (Gene Analysis Techniques 6, 25–28, 1989) and Brooks et al. (Biotechniques 19, 806–812, 814–815, 1985) found that at low temperature the tightly folded structure of RNAs may hinder reverse transcription proceeding along the RNA template, which increases its chance of dissociating from the template; therefore, raising the reaction temperature was the simplest way to overcome template secondary structure and prevent premature termination of cDNA synthesis. In this report, two vectors based on murine leukaemia virus (MLV) were constructed. The first contained two mutated gfp genes in tandem positions. The upstream gfp gene encoded a mutation at its 3′ end, while the downstream gfp gene encoded a mutation at its 5′ end. The recombination that occurred between the two mutated gfp genes restored a functional gfp gene. The cells that contained the functional gfp gene were green when observed under a fluorescence microscope. The second MLV vector contained a functional gfp gene with two identical sequences flanking either end. A recombination that occurred between the two identical sequences resulted in deletion of the gfp gene. Cells containing the vector with the gfp deletion were colourless or clear when observed under the microscope. Using these two vectors, we have demonstrated that retroviral recombination is temperature dependent and the rate of recombination decreases as temperature is raised from 31 to 43 °C.

Adult immunocompetent mice inoculated with Ebola (EBO) or Marburg (MBG) virus do not become ill. A suckling-mouse-passaged variant of EBO Zaire ’76 (‘mouse-adapted EBO-Z’) causes rapidly lethal infection in adult mice after intraperitoneal (i.p.) inoculation, but does not cause apparent disease when inoculated subcutaneously (s.c.). A series of experiments showed that both forms of resistance to infection are mediated by the Type I interferon response. Mice lacking the cell-surface IFN-α/β receptor died within a week after inoculation of EBO-Z ’76, EBO Sudan, MBG Musoke or MBG Ravn, or after s.c. challenge with mouse-adapted EBO-Z. EBO Reston and EBO Ivory Coast did not cause illness, but immunized the mice against subsequent challenge with mouse-adapted EBO-Z. Normal adult mice treated with antibodies against murine IFN-α/β could also be lethally infected with i.p.-inoculated EBO-Z ’76 or EBO Sudan and with s.c.-inoculated mouse-adapted EBO-Z. Severe combined immunodeficient (SCID) mice became ill 3–4 weeks after inoculation with EBO-Z ’76, EBO Sudan or MBG Ravn, but not the other viruses. Treatment with anti-IFN-α/β antibodies markedly accelerated the course of EBO-Z ’76 infection. Antibody treatment blocked the effect of a potent antiviral drug, 3-deazaneplanocin A, indicating that successful filovirus therapy may require the active participation of the Type I IFN response. Mice lacking an IFN-α/β response resemble primates in their susceptibility to rapidly progressive, overwhelming filovirus infection. The outcome of filovirus transfer between animal species appears to be determined by interactions between the virus and the innate immune response.

The intracellular cleavage of respiratory syncytial virus (RSV) fusion (F) protein by furin was examined. In RSV-infected LoVo cells, which express an inactive form of furin, and in RSV-infected Vero cells treated with the furin inhibitor decanoyl-Arg-Val-Lys-Arg-chloromethyl ketone (dec-RVKR-cmk), the F protein was expressed as a non-cleaved 73 kDa species. In both cases the F protein was initially expressed as an endoglycosidase H (Endo H)-sensitive precursor (F0EHs) which was modified approximately 40 min post-synthesis by the addition of complex carbohydrates to produce the Endo H-resistant form (F0EHr). The size and glycosylation state of F0EHr were identical to a transient intermediate form of non-cleaved F protein which was detected in RSV-infected Vero cells in the absence of inhibitor. Cell surface biotinylation and surface immunofluorescence staining showed that F0EHr was present on the surface of RSV-infected cells. RSV filaments have been shown to be the predominant form of the budding virus that is detected during virus replication. Analysis of the RSV-infected cells using scanning electron microscopy (SEM) showed that, in the presence of dec-RVKR-cmk, virus budding was impaired, producing fewer and much smaller viral filaments than in untreated cells. A comparison of immunofluorescence and SEM data showed that F0EHr was routed to the surface of virus-infected cells but not located in these smaller structures. Our findings suggest that activation of the F protein is required for the efficient formation of RSV filaments.

H9-D3-4R2 (referred to as H9), a murine monoclonal HA1-specific IgG3, recognizes an epitope within antigenic site Cb of influenza virus A/PR/8/34 (H1N1). At 50% neutralization, inhibition of virus-mediated fusion was responsible for the majority of neutralization but, at higher antibody concentrations, the attachment of virus to target cells was also inhibited and may have contributed to neutralization. H9 FAb was also neutralizing, although the concentration needed was two orders of magnitude greater than for the IgG. Functional affinity of the IgG and affinity of the FAb were almost identical, and it is not clear why the neutralization efficiency of the FAb was so low. Unlike its IgG, H9 FAb had no detectable effect on virus attachment but inhibited virus fusion activity. It thus appears that monovalent binding by this antibody is sufficient to inhibit fusion activity and that this was directly responsible for neutralization of infectivity.

Influenza virus A/Hong Kong/1774/99, isolated from a young child with mild influenza, was shown to be similar in its antigenic and genetic characteristics to H3N2 viruses circulating in pigs in Europe during the 1990s and in particular to be closely related to viruses isolated from two children in the Netherlands in 1993. Similar viruses had previously not been identified outside Europe. Although there is little evidence as to how the child contracted the infection, it appears likely that pigs in southern China were the source of infection. Characteristics shared with the European swine viruses include resistance to the anti-influenza drugs amantadine and rimantadine. Thus not only does this incident once again highlight the potential of pigs as a source of novel human influenza viruses, but also indicates the potential for emergence of amantadine-resistant human viruses.

Herpes simplex virus type 1 (HSV-1) glycoproteins gB, gD and gHL are capable of inducing cell fusion when expressed from plasmid vectors in the absence of any other virus components. Fusion requires the expression of all four glycoproteins on the same membrane, since they are unable to cooperate in trans to induce syncytium formation. In addition, the fusion event is dependent on the expression of a gD receptor on target cell membranes and does not require the presence of cell-surface glycosaminoglycans.

Herpes simplex virus type 2 UL34 protein is expressed late in infection and is required for envelopment of nucleocapsids at the nuclear membrane and possibly at the endoplasmic reticulum (ER). It is a type II membrane protein with a C-terminal anchor that localizes mainly to the nuclear membrane in infected cells. However, in single transient expression, UL34 protein localizes predominantly to the ER. Relocation of UL34 protein from the ER to the internal nuclear membrane and the nucleus was observed in the presence of UL31 protein, a phosphoprotein known to interact physically with UL34. It is suggested here that interaction with UL31 protein is important for the nuclear targetting of UL34 protein and also that the trans-membrane region of UL34 protein is responsible for its localization at the internal nuclear membrane. The results also suggest possible sites for the interaction.