Search

Arteriviruses are positive-stranded RNA viruses that infect mammals. They can cause persistent or asymptomatic infections, but also acute disease associated with a respiratory syndrome, abortion or lethal haemorrhagic fever. During the past two decades, porcine reproductive and respiratory syndrome virus (PRRSV) and, to a lesser extent, equine arteritis virus (EAV) have attracted attention as veterinary pathogens with significant economic impact. Particularly noteworthy were the 'porcine high fever disease' outbreaks in South-East Asia and the emergence of new virulent PRRSV strains in the USA. Recently, the family was expanded with several previously unknown arteriviruses isolated from different African monkey species. At the molecular level, arteriviruses share an intriguing but distant evolutionary relationship with coronaviruses and other members of the order Nidovirales. Nevertheless, several of their characteristics are unique, including virion composition and structure, and the conservation of only a subset of the replicase domains encountered in nidoviruses with larger genomes. During the past 15 years, the advent of reverse genetics systems for EAV and PRRSV has changed and accelerated the structure-function analysis of arterivirus RNA and protein sequences. These systems now also facilitate studies into host immune responses and arterivirus immune evasion and pathogenesis. In this review, we have summarized recent advances in the areas of arterivirus genome expression, RNA and protein functions, virion architecture, virus-host interactions, immunity, and pathogenesis. We have also briefly reviewed the impact of these advances on disease management, the engineering of novel candidate live vaccines and the diagnosis of arterivirus infection.

Lactate dehydrogenase-elevating virus (LDV) exacerbates mouse susceptibility to endotoxin shock through enhanced tumour necrosis factor (TNF) production by macrophages exposed to lipopolysaccharide (LPS). However, the in vivo enhancement of TNF production in response to LPS induced by the virus largely exceeds that found in vitro with cells derived from infected animals. Infection was followed by a moderate increase of Toll-like receptor (TLR)-4/MD2, but not of membrane CD14 expression on peritoneal macrophages. Peritoneal macrophages from LDV-infected mice unresponsive to type I interferons (IFNs) did not show enhanced expression of TLR-4/MD2 nor of CD14, and did not produce more TNF in response to LPS than cells from infected normal counterparts, although the in vivo response of these animals to LPS was strongly enhanced. In contrast, the virus triggered a sharp increase of soluble CD14 and of LPS-binding protein serum levels in normal mice. However, production of these LPS soluble receptors was similar in LDV-infected type I IFN-receptor deficient mice and in their normal counterparts. Moreover, serum of LDV-infected mice that contained these soluble receptors had little effect if any on cell response to LPS. These results suggest that enhanced response of LDV-infected mice to LPS results mostly from mechanisms independent of LPS receptor expression.

In 2006-2007, equine viral arteritis (EVA) was confirmed for the first time in Quarter Horses in multiple states in the USA. The entire genome of an equine arteritis virus (EAV) isolate from the index premises in New Mexico was 12 731 nt in length and possessed a previously unrecorded unique 15 nt insertion in the nsp2-coding region in ORF1a and a 12 nt insertion in ORF3. Sequence analysis of additional isolates made during this disease occurrence revealed that all isolates from New Mexico, Utah, Kansas, Oklahoma and Idaho had 98.6-100.0 % (nsp2) and 97.8-100 % (ORF3) nucleotide identity and contained the unique insertions in nsp2 and ORF3, indicating that the EVA outbreaks in these states probably originated from the same strain of EAV. Sequence and phylogenetic analysis of several EAV isolates made following an EVA outbreak on another Quarter Horse farm in New Mexico in 2005 provided evidence that this outbreak may well have been the source of virus for the 2006-2007 occurrence of the disease. A virus isolate from an aborted fetus in Utah was shown to have a distinct neutralization phenotype compared with other isolates associated with the 2006-2007 EVA occurrence. Full-length genomic sequence analysis of 18 sequential isolates of EAV made from eight carrier stallions established that the virus evolved genetically during persistent infection, and the rate of genetic change varied between individual animals and the period of virus shedding.

Strains of Equine arteritis virus (EAV) differ in the severity of the disease that they induce in horses. Infectious cDNA clones are potentially useful for identification of genetic determinants of EAV virulence; to date, two clones have been derived from a cell culture-adapted variant of the original (Bucyrus) isolate of EAV, and it has previously been shown that recombinant virus derived from one of these (rEAV030) is attenuated in horses. A complete cDNA copy of the genome of the virulent Bucyrus strain of EAV has now been assembled into a plasmid vector. In contrast to rEAV030, recombinant progeny virus derived from this clone caused severe disease in horses, characterized by pyrexia, oedema, leukopenia, high-titre viraemia and substantial nasal shedding of virus. The availability of infectious cDNA clones that produce recombinant viruses of different virulence to horses will facilitate characterization of the virulence determinants of EAV through reverse genetics.

Equine arteritis virus (EAV) causes a persistent infection of the reproductive tract of carrier stallions. The authors determined the complete genome sequences of viruses (CW96 and CW01) that were present 5 years apart in the semen of a carrier stallion (CW). The CW96 and CW01 viruses respectively had only 85.6 % and 85.7 % nucleotide identity to the published sequence of EAV (EAV030). The CW96 and CW01 viruses had two 1 nt insertions and a single 1 nt deletion in the leader sequence, and a 3 nt coding insertion in ORF1a; thus their genomes included 12 708 nt as compared to the 12 704 nt in EAV030. Variation between viruses present in the semen of stallion CW and EAV030 was especially marked in the replicase gene (ORF1a and 1b), and the greatest variation occurred in the portion of ORF1a encoding the nsp2 protein. The ORFs 3 and 5, which respectively encode the GP3 and GP5 envelope proteins, showed greatest variation amongst ORFs encoding structural EAV proteins. Comparative sequence analyses of CW96 and CW01 indicated that ORFs 1a, 1b and 7 were highly conserved during persistent infection, whereas there was substantial variation in ORFs 3 and 5. Although the variation that occurs in ORF5 results in the emergence of novel phenotypic viral variants as determined by neutralization assay, all variants were neutralized by high-titre polyclonal equine antisera, suggesting that immune evasion is unlikely to be responsible for the establishment of persistent EAV infection of carrier stallions. Northern blot analyses of RNA extracted from cell culture propagated viruses isolated from 10 different persistently infected stallions failed to demonstrate any large genomic deletions, suggesting that defective interfering particles are also unlikely to be important in either the maintenance or clearance of persistent EAV infection of the reproductive tract of carrier stallions.

Equine arteritis virus (EAV) causes a systemic infection in equids with variable outcome, ranging from subclinical infections to severe disease, and also has the capacity to induce abortion in pregnant mares and persistent infections in stallions. The serum virus-neutralizing antibody response that invariably develops in the infected animal lasts for many months or years and is believed to play an important role in virus clearance. However, very little is known about cellular immunity against EAV because of a lack of methods for evaluating these immune responses. In the present study, we describe methods for detecting cytotoxic T lymphocyte (CTL) precursors in the peripheral blood of EAV-convalescent ponies using a (51)Cr release cytolysis assay. Primary equine dermal cells, used as CTL targets, were shown to express MHC I but not MHC II and to retain (51)Cr efficiently and support EAV replication. Peripheral blood mononuclear cells (PBMC) collected from EAV-convalescent ponies that had been incubated with or without live EAV were used as effectors. EAV-induced PBMC cultures showed evidence of expansion and activation of lymphoblasts, with an increase in the CD8(+)/CD4(+) ratio in comparison with mock-induced PBMC. The cytotoxicity induced by EAV-stimulated PBMC was virus specific, showed genetic restriction, was mediated by CD8(+) T lymphocytes and could be detected for periods of 4 months to more than 1 year post-infection. These findings and methods will hopefully contribute to an understanding of virus-host interactions in horses, in particular the mechanisms of virus clearance occurring during EAV infection.

Early after infection, lactate dehydrogenase-elevating virus (LDV) alters the immune system by polyclonally activating B lymphocytes, which leads to IgG2a-restricted hypergammaglobulinaemia, and by suppressing the secretion of Th2 cytokines. Considering that these alterations may involve cells of the innate immune system and cytokines such as interferon-gamma (IFN-gamma), we analysed the effect of LDV on natural killer (NK) cells. Within a few days of infection, a strong and transient NK cell activation, characterized by enhanced IFN-gamma message expression and cytolysis, was observed. LDV triggered a large increase in serum IFN-gamma levels. Because NK cells and IFN-gamma may participate in the defence against virus infection, we analysed their possible role in the control of LDV titres with a new agglutination assay. Our results indicate that neither the activation of NK cells nor the IFN-gamma secretion affect the early and rapid virus replication that follows LDV inoculation.

An imported carrier stallion (A) from Europe was implicated in causing an extensive outbreak of equine viral arteritis (EVA) on a Warmblood breeding farm in Pennsylvania, USA. Strains of equine arteritis virus (EAV) present in the semen of two carrier stallions (A and G) on the farm were compared to those in tissues of foals born during the outbreak, as well as viruses present in the semen of two other stallions that became persistently infected carriers of EAV following infection during the outbreak. The 2822 bp segment encompassing ORFs 2-7 (nt 9807-12628; which encode the G(S), GP3, GP4, G(L), M and N proteins, respectively) was directly amplified by RT-PCR from semen samples and foal tissues. Nucleotide and phylogenetic analyses confirmed that virus present in the semen of stallion A initiated the outbreak. The genomes of viruses present in most foal tissues (10/11) and serum from an acutely infected mare collected during the outbreak were identical to that of virus present in the lung of the first foal that died of EVA. Virus in the placenta of one foal differed by one nucleotide (99.9% identity) from the predominant outbreak virus. The relative genetic stability of viruses that circulated during the outbreak contrasts markedly with the heterogeneous virus populations variously present in the semen of persistently infected stallions on the farm. These findings are consistent with the hypothesis that the carrier stallion can be a source of genetic diversity of EAV, and that outbreaks of EVA can be initiated by the horizontal aerosol transmission of specific viral variants that occur in the semen of particular carrier stallions.

Equine arteritis viruses (EAV) from Europe and America were compared by phylogenetic analysis of 43 isolates obtained over four decades. An additional 22 virus sequences were retrieved from GenBank. Fragments of the glycoprotein G(L) and the replicase genes were amplified by RT-PCR, prior to sequencing and construction of phylogenetic trees. The trees revealed many distinctive lineages, consistent with prolonged diversification within geographically separated host populations. Two large groups and five subgroups were distinguished. Group I consisted mainly of viruses from North America, whilst group II consisted mainly of European isolates. In most instances, where the geographic origin of the viruses appeared to be at variance with the phylogenetically predicted relationships, the horses from which the viruses were recovered had been transported between Europe and America or vice versa. Analysis of the replicase gene revealed similar phylogenetic relationships although not all of the groups were as clearly defined. Virus strains CH1 (Switzerland, 1964) and S1 (Sweden, 1989) represented separate 'outgroups' based on analysis of both genomic regions. The results of this study confirm the value of the G(L) gene of EAV for estimating virus genetic diversity and as a useful tool for tracing routes by which EAV is spread. In addition, computer-assisted predictions of antigenic sites on the G(L) protein revealed considerable variability among the isolates, especially with respect to regions associated with neutralization domains.

The subgenomic mRNAs (sg mRNA) of porcine reproductive and respiratory syndrome virus (PRRSV) were characterized. The number of sg mRNAs, which form a 3'-coterminal nested set in PRRSV-infected cells, varied from six to seven among PRRSV isolates with differing virulence. The additional species of sg mRNA in some isolates of PRRSV was designated as sg mRNA 3-1. The leader-mRNA junction sequences of sg mRNAs 3, 3-1 and 4 were found to contain a similar six nucleotide sequence motif, U(G)UA(G/C)ACC. By comparing the 5'-terminal sequence of sg mRNA 3-1 with the genomic sequence of two isolates, ISU79 and ISU1894, it was found that a point mutation, from U in isolate ISU1894 to C in isolate ISU79, led to the acquisition of a new leader-mRNA junction sequence (UUGACC) in isolate ISU79, and therefore an additional species of sg mRNA 3-1. A small ORF (3-1) was identified at the 5' end of sg mRNA 3-1.

The nucleotide sequence of a 3266 bp region encompassing open reading frames (ORFs) 2 through 7 of the porcine reproductive and respiratory syndrome virus (PRRSV) was determined for 10 isolates recovered from the midwestern United States. Pairwise comparisons showed that genetic distances between isolates ranged from 2.5% to 7.9% (mean 5.8% +/- 0.2%) whereas the Lelystad strain from Europe was, on average, 34.8% divergent from US clones. Thus, US and European PRRSV isolates represent genetically distinct clusters of the same virus. ORF 5, which encodes the envelope glycoprotein, was the most polymorphic [total nucleotide diversity (pi) = 0.097 +/- 0.007] and ORF 6, encoding the viral M protein, was the most conserved (pi = 0.038 +/- 0.003). The substantial differences in nucleotide diversity among ORFs suggests that the virus is evolving by processes other than simple accumulation of random neutral mutations. In support of this hypothesis, statistical analyses of the nucleotide sequence provided strong evidence for intragenic recombination or gene conversion in ORFs 2, 3, 4, 5 and 7, but not in ORF 6. An excess of synonymous (silent) substitutions was observed in all six ORFs, indicating an evolutionary pressure to conserve amino acid sequences. Taken together, the data indicate that despite intragenic recombination among extant PRRSV isolates, purifying selection has acted to maintain the primary structure of individual ORFs.

This report describes the preparation of six monoclonal antibodies (MAbs) raised against a British isolate of porcine reproductive and respiratory syndrome virus (PRRSV), their characterization in terms of protein specificity and their reactivity with different PRRS viruses from Europe and the USA. Radioimmunoprecipitation and Western blotting studies of MAb reactivity with proteins from cell lysates of infected cells and purified virus revealed that four of the six MAbs (WBE1 and WBE4-6) precipitated a 15 kDa viral protein. Further studies using in vitro translated products of the Lelystad virus genome showed that this protein was the product of ORF7, the putative nucleocapsid protein. The specificity of another MAb, WBE2, was found to be for a 45 kDa protein, determined to be the product of ORF3 and demonstrated to be present in purified virion preparations. The protein specificity of the sixth MAb, WBE3 could not be determined. Thirty-three PRRSV isolates from Europe and the USA were grown in alveolar macrophages and examined by immunoperoxidase staining, using the panel of six MAbs. All European isolates were recognized by the four MAbs specific for the putative nucleocapsid, but the viruses showed different patterns of reactivity with WBE2 and WBE3. Furthermore, these two MAbs stained only a small proportion of the cells infected with certain isolates, suggesting that a single isolate may be antigenically heterogeneous. No MAbs bound to US isolates, indicating a consistent antigenic difference between the putative nucleocapsid of US and European isolates. Detergent extraction of cell lysate antigen abrogated the binding of WBE1-3, suggesting that the epitopes are conformation dependent.

The initial replication of lactate dehydrogenase-elevating virus (LDV) in mice, its invasion of the central nervous system (CNS) and infection of anterior horn neurons in C58 and AKXD-16 mice were investigated by Northern and in situ hybridization analyses. Upon intraperitoneal injection, LDV replication in cells in the peritoneum was maximal at 8 h post-infection (p.i.). Next, LDV infection was detected in bone marrow cells and then in macrophage-rich regions of all tissues investigated (12 to 24 h p.i.). By 2 to 3 days p.i., LDV RNA-containing cells had largely disappeared from all non-neuronal tissues due to the cytocidal nature of the LDV infection of macrophages. In the CNS at 24 h p.i. LDV replication was very limited and confined to cells in the leptomeninges. LDV replication in the cells of the leptomeninges should result in the release of progeny LDV into the cerebrospinal fluid and thus its dissemination throughout the CNS. However, in C58 and AKXD-16 mice, which are susceptible to paralytic LDV infection, only little LDV RNA and few LDV-infected cells were detectable in the spinal cord until at least 10 days p.i. Extensive cytocidal infection of anterior horn neurons occurred only shortly before the development of paralytic symptoms between 2 and 3 weeks p.i. The reason for the relatively long delay in LDV infection of anterior horn neurons is not known. No LDV RNA or LDV RNA-containing cells were detected in the brain, except in the leptomeninges at early times after infection.