ARTERIVIRUSES ( ARTERIVIRIDAE)

Abstract

History Lactate dehydrogenase-elevating virus (LDV) was discovered by accident in 1960 during a study on the development of methods for early detection of tumors in mice. Inoculation of mice with Ehrlich carcinoma cells resulted in a 5 to 10-fold increase in lactate dehydrogenase (LDH) levels in the serum in 4 days. Because this early rise in LDH levels was also observed in mice inoculated with cell-free extracts or sera from tumor-bearing mice, the presence of an infectious agent was suspected. LDV is now known to produce an asymptomatic, persistent infection in mice characterized by a lifelong viremia and elevation of the levels of at least seven serum enzymes in addition to LDH. Equine arteritis virus (EAV) was first isolated in 1957 in Bucyrus, Ohio from the lung tissues of an aborted fetus during an epidemic of abortions and arteritis in pregnant mares. However, equine disease consistent with that caused by EAV, was first observed in the late 1800s. At the time of its discovery EAV was distinguished from equine (abortion) influenza virus. Serological evidence has indicated that, although EAV is widespread in the horse population, it rarely causes clinical disease. EAV can produce persistent infections in seropositive horses. Simian hemorrhagic fever virus (SHFV) was first isolated in 1964 during outbreaks of a fatal hemorrhagic fever in macaque colonies in the United States, Russia and Europe. A number of additional SHFV outbreaks in macaque colonies have occurred since the 1960s. The most ‘famous’ of these was the one in Reston, Virginia, which occurred in conjunction with a Reston Ebola virus outbreak. SHFV causes asymptomatic acute or persistent infections in several genera of African monkeys that are thought to be the natural hosts for this virus. Porcine respiratory and reproductive syndrome was first detected in North America in 1987 and in Europe in 1990. This disease has also been referred to as porcine epidemic abortion and respiratory syndrome (PEARS), swine infertility and respiratory syndrome (SIRS) and mystery swine disease (MSD). The causative agent of this disease is now referred to as porcine respiratory and reproductive syndrome virus (PRRSV). The extent of the sequence divergence observed between American and European strains of PRRSV suggests that the ability to cause porcine disease emerged independently in these two virus populations. Taxonomy and Classification On the basis of the size and morphology of their virions as well as the infectious nature of their RNA genomes, LDV and EAV were originally classified within the Togaviridae family. After it was reported that EAV transcribed multiple subgenomic mRNAs, EAV was reclassified as the sole member of the Togaviridae genus Arterivirus. LDV was subsequently designated as a member of this genus. The genus Arterivirus was then designated as a ‘floating genus’. SHFV was first classified as a togavirus and then as a flavivirus. In 1996, EAV, LDV, SHFV and PRRSV were designated as members of a new family, the Arteriviridae. At the same time, the family Arteriviridae was classified with the family Coronaviridae in the new Order Nidovirales. The genomes of EAV, LDV, PRRSV and SHFV have been sequenced and their genome organization is similar to that of the coronaviruses (Fig. 1 ). Although the genomes of the four known arteriviruses contain the general features and conserved motifs characteristic of coronavirus genomes, they are only about half as long as coronavirus genomes. In addition, arterivirus particles are about half the size of coronavirus particles and differ from them morphologically; the surface of arterivirus envelopes is fairly smooth, and arteriviruses have icosahedral nucleocapsids. Figure 1 Organization of the genomes of LDV, EAV, PRRSV and SHFV. The nonstructural polyproteins are encoded by open reading frame (ORF) 1a/1b, which is located in the 5′ portion of the genome. ORF 1b is translated only when a −1 frameshift occurs. The conserved motifs in ORF 1a/1b are indicated by black boxes. From left to right these are: one to three papain-like cysteine proteases (stippled boxes), a cysteine protease, a serine protease, an RNA polymerase, a zinc finger, a helicase and a nidovirus-specific motif. The 3′ portion of the genome encodes the structural proteins. Adjacent 3′ ORFs are in different frames. (The genome sequences utilized are from: EAV, den Boon et al (1991) J. Virol. 65: 2910; PRRSV (Lelystad), Meulenberg et al (1993) Virology 192: 62; LDV, Godeny et al (1993) Virology 194: 585, Smith et al (1997) Gene 19: 205 and Brinton, unpublished data.) Properties of the Virion Virions are spherical, enveloped and 50–60 nm in diameter (Fig. 2 ). Unfixed virions undergo distortion and disintegration during standard negative staining procedures. The virion surface appears smooth and is thought to contain short, if any, spike-like projections. The virion capsid is icosahedral and about 30–35 nm in diameter. The general appearance of these viruses is similar to that of the alpha togaviruses and the flaviviruses and this morphological similarity resulted in some of the arteriviruses initially being classified as members of these virus families. Buoyant densities of 1.13–1.17 g cm−3 have been reported for arteriviruses in sucrose. The spherical capsid is composed of the nucleocapsid (N) protein. Virion envelopes contain the M protein, one major glycoprotein, and one to three minor glycoproteins. The major glycoprotein and the M protein are present as heterodimers in virions. Figure 2 An electron micrograph of a positively stained thin section of gradient-purified LDV. Properties of the Genome The genomes of the arteriviruses are single-stranded RNAs of positive polarity. They contain a 3′ poly (A) tract of approximately 50 nucleotides in length and a 5′ type I cap. The length of the genome of EAV is 12.7 kb, that of LDV is 14.1 kb, that of PRRSV is 15.1 kb and that of SHFV is 15.7 kb (Fig. 1). The nonstructural polyproteins are encoded by ORFs 1a and 1b. The ORF 1b proteins are expressed only when a −1 frameshift occurs. The frameshift region is characterized by a slippery sequence upstream of a pseudoknot. Three different types of virion proteases are encoded in ORF 1a and these post-translationally cleave the ORF 1a and ORF 1a/1b polyproteins into 12 or more mature nonstructural proteins. There are six 3′ ORFs in the genomes of EAV, PRRSV and LDV, whereas SHFV encodes nine 3′ ORFs. Limited sequence homology suggests that the SHFV ORFs 2a, 2b and 3 may be duplications of the SHFV ORFs 4, 5, and 6, respectively (Fig. 1). Adjacent arterivirus 3′ ORFs are in different reading frames and in most cases overlap. Only the 5′ terminal ORF of most of the subgenomic mRNAs is translated. However, mRNA 2 of EAV, LDV, and PRRSV and mRNAs 2 and 4 of SHFV are bicistronic. Properties of the Proteins Two polyproteins are translated from the ORF 1a/1b region of the arterivirus genome. The ORF 1a polyprotein contains one or more papain-like cysteine proteases, a cysteine protease, and a serine protease. The serine protease cleaves at five sites within the ORF 1a polyprotein and eight sites within the ORF 1a/1b polyprotein. Three of the ORF 1a encoded proteins have hydrophobic domains which may be important for membrane association of the viral replication–transcription complexes. The ORF 1a/1b polyprotein is translated only after a −1 frameshift occurs. Motifs characteristic of a viral RNA polymerase, a zinc-finger, a helicase, and a conserved nidovirus-specific region are found in ORF 1b (Fig. 1). The capsid protein is encoded by ORF 7 (ORF 9 in SHFV), the nonglycosylated, triple-membrane spanning envelope M protein is encoded by ORF 6 (ORF 8 in SHFV), and the major virion glycoprotein is encoded by ORF 5 (ORF 7 in SHFV). The remaining 3′ ORFs each encode proteins of about 20 kDa in size that contain a number of putative N-linked glycosylation sites and may function as minor virion envelope glycoproteins. The ORF 2 (ORF 4 in SHFV) protein is a class I integral membrane protein. A soluble, non-virion-associated form of the ORF 3 protein is released from infected cells. Physical Properties Virions can be stored indefinitely at −70°C. At −20°C, samples of LDV lost half of their infectivity by 4 weeks. At 4°C, LDV in mouse plasma decreased in titer by about 3.5 logs in 32 days; at room temperature, plasma virus was stable for 24 h, whereas infectivity was completely inactivated by heating at 58°C for 1 h. LDV suspended in media supplemented with 10% serum was more heat labile than virus in plasma. Virions are fairly stable between pH 6 and pH 7.5, but are rapidly inactivated at high or low pH. Virus is efficiently inactivated by lipid solvents and is very sensitive to nonionic detergent treatment. A brief incubation with low concentrations (0.01%) of nonionic detergents, such as NP40 and Triton X-100, efficiently disrupts the virion envelope. Replication Viral replication occurs entirely in the cytoplasm. Infected cells contain full-length genome RNA, subgenomic viral mRNAs and full-length and subgenomic viral minus-strand RNAs. The subgenomic mRNAs form a 3′ coterminal nested set (Fig. 3 ) and are thought to be transcribed by an uncharacterized discontinuous mechanism which utilizes the conserved junction sequences that precede each 3′ ORF. Each subgenomic mRNA contains a common 5′ leader sequence which is identical to the 5′ terminal sequence of the genome RNA. Conserved RNA structures and cis-acting signals for plus-strand RNA replication have been identified in the 3′ noncoding region of the complementary full-length minus-strand RNA. The 3′ (−)non-coding region has also been shown to bind specifically to four cellular proteins thought to be involved in viral RNA replication. Figure 3 EAV proteins encoded at the 5′ end of the genome are translated from the genome RNA whereas proteins encoded at the 3′ end are translated from subgenomic mRNAs. The EAV genomic and subgenomic mRNAs are designated by numbers 1–6. The autoradiographs show Northern-blot hybridizations obtained with mock-infected (M) and EAV-infected (V) cell extracts reacted with either (A) a 3′ probe or (B) a 5′ probe. The open boxes represent ORFs, and the closed boxes represent the 5′ leader sequence. The bicistronic nature of mRNA2 is not indicated. (Reproduced with permission from Spaan et al (1989) In: Brinton MA and Heinz FX (eds) New Aspects of Positive Strand Viruses, p. 12. Washington: ASM Press.) Assembly, Uptake and Cytopathology Evidence for a specific saturable, but as yet unidentified receptor for LDV on a subpopulation of murine macrophages has been reported. LDV-immune complexes are infectious and apparently infect target cells via Fc receptors. Ventral motor neurons in susceptible mouse strains can be infected by ‘free’ LDV-C, but not by viral immune complexes. A heparin-like molecule on the surface of MARC-145 cells may serve as a receptor for PRRSV. PRRSV has been reported to enter cells via a low pH-dependent, endocytic pathway. Soon after infection, virus particles have been observed in small vesicles that appear to be clathrin-coated. Virus replication occurs in the perinuclear area of infected cells. Virions are observed to bud through endoplasmic reticular membranes into intracellular vesicles. Virions presumably move to the exterior of the cell via these vesicles. Assembled nucleocapsids have not been observed in the cytoplasm, except in association with budding virions. Infection of primary macrophages with EAV, SHFV, PRRSV and, probably, also LDV is cytocidal. The formation of double-membrane cytoplasmic vesicles is a characteristic of virus-infected cells. Laboratory strains of EAV, SHFV, and PRRSV cause obvious cytopathology in the continuous cell cultures which they infect. Cells become rounded and are released from the tissue culture flask. Apoptosis has been observed in PRRSV-infected porcine alveolar macrophages, MA-104 cells and testicular germ cells. Geographic and Seasonal Distribution Viruses with biological properties identical to those of LDV have been isolated from small groups of wild mice (Mus musculus) in Australia, Germany, the US and UK. EAV-infections and EAV-induced disease have been well documented in North America and Europe and antibodies to EAV have been detected in horse sera from Africa and South America indicating that EAV infection is geographically widespread. Natural PRRSV infections have been reported in North America and Europe. SHFV infection in captive patas monkeys has been documented and SHFV has been detected in the blood of wild patas monkeys, African green monkeys, and baboons suggesting that these African primates are the natural hosts for SHFV. Host Range and Virus Propagation LDV replicates efficiently in all strains of laboratory mice and wild Mus musculus and somewhat less efficiently in the Asian mouse Mus caroli. Numerous attempts to infect other rodents such as rats, hamsters, guinea pigs, rabbits, deer mice (Peromyscus maniculatus) and dwarf hamsters (Phodopus sungorus) with LDV have not been successful. LDV grows very efficiently in mice reaching titers of 1011 ID50 ml−1 of plasma within the first day of infection. By a week after infection, viral titers have declined to a steady-state level of about 106 ID50 ml−1. In vitro, LDV replicates only in an as yet uncharacterized subpopulation of primary murine macrophages. In peritoneal macrophage cultures obtained from 1–2-week-old mice, up to 70% of the cells are infectible. However, only 40% or less of the macrophages are virus susceptible when the cells are obtained from older mice. The highest number of infected cells was observed when peritoneal cell cultures were infected during the first 24 h plating. Thereafter, the number of infectible cells progressively declined, until by 7 days, the cultures were no longer infectible. However, incubation of fresh peritoneal macrophage cultures with colony-stimulating factor resulted in the continued differentiation of virus-susceptible cells from stem cells. Attempts to infect numerous continuous cell lines with LDV, including all attempts to infect murine macrophage cell lines, have been unsuccessful with one exception. A line of mouse macrophage–human hybrid cells was reported to be transiently permissive for LDV replication. EAV infects horses and donkeys. Field isolates are rarely obtained and are difficult to propagate in cell culture. However, laboratory strains of EAV have been successfully grown in primary cultures of horse macrophages and kidney cells, rabbit kidney cells and hamster kidney cells and also in cell lines, such as BHK-21, RK-13, MA-104 and Vero. Natural infections with PRRSV were thought to be restricted to pigs. However, a recent report suggested that chickens and mallard ducks may also be susceptible to infection with PRRSV. PRRSV replicates in primary cultures of porcine alveolar macrophages. Some, but not all, isolates of PRRSV can be adapted to replicate in vitro in MA-104 cells. SHFV naturally infects several species of African primates, namely Erythrocebus patas, Cercopithecus aethiops, Papio anuibus and Papio cynocephalus. SHFV infection of members of the genus Macaca has occurred in a number of primate facilities. Some isolates of SHFV can replicate in primary cultures of rhesus aveolar lung macrophages and MA-104 cells. Genetics Evidence for the existence of virulence mutants of LDV, SHFV and EAV has been obtained. Virulent and avirulent mutants of EAV have been identified on the basis of the severity of the diseases they cause. A number of temperature-sensitive mutants of EAV have also been selected. One strain of LDV, designated LDV-C, was shown to induce neurologic disease in a few susceptible inbred mouse strains. SHFV isolates that produce acute infections and ones that cause persistent infections in patas monkeys have been reported. EAV and PRRSV infectious clones have recently been constructed. Evolution Comparison of the genomes of EAV, LDV, PRRSV and SHFV (Fig. 1) suggest that these genomes have evolved via both point mutations and copy choice RNA recombination mechanisms. Sequence comparisons of various field isolates of either PRRSV or EAV indicated that the sequences of the M and N proteins are more conserved than those of the glycoproteins. The extent of the divergence between the sequences of European and North American PRRSV isolates indicate that these two virus populations represent subspecies. Phylogenic analyses have shown that PRRSV is most closely related to LDV and that SHFV is more closely related to these two viruses than to EAV. The organization of the genomes of LDV, EAV, SHFV and PRRSV is very similar to that of the coronaviruses (Fig. 1). Both types of viral genomes have the same set of conserved nonstructural protein motifs arranged in the same order within ORF 1a/b. Although there are significant differences in the structural protein regions of the arterivirus and the coronavirus genomes, both families of viruses encode their capsid protein as the terminal 3′ gene and a nonglycosylated envelope M protein as the penultimate 3′ gene. Coronaviruses have helical nucleocapsids, whereas LDV, SHFV, PRRSV and EAV have icosahedral nucleocapsids. The capsid protein of the coronaviruses is about twice as large as the one encoded by the arteriviruses. Coronavirus particles have large peplomer spikes on their surface, whereas LDV, SHFV, PRRSV and EAV particles have ‘smooth’ surfaces. The coronaviruses and the arteriviruses most likely evolved as two distinct lineages from a common progenitor. The prototype virus may originally have had an iscosahedral nucleocapsid, and then via a copy choice recombinational event acquired a helical capsid gene from another virus. This change in the type of capsid utilized would have removed the packaging restrictions on the genome size. The additional genetic information found in the current coronavirus genomes could then have been gained during further evolution via recombination. Serologic Relationships and Variability All attempts to demonstrate antigenic crossreactivity between EAV, LDV, PRRSV and SHFV have been unsuccessful. Even though antiviral antibodies are detected in infected animals by a week after infection, all four viruses can cause persistent infections. The mechanism by which these viruses evade clearance by the immune system is not known. Neutralization escape virus variants have been reported to arise during persistent LDV infections. Antiserum obtained from animals chronically infected with LDV and PRRSV contains infectious viral immune complexes. Plasma from LDV-infected mice has a higher nonspecific binding activity than plasma from uninfected mice; virus-specific binding measured by an ELISA usually cannot be detected until the plasma has been diluted at least 1:400. Transmission No insect vector has been associated with LDV or SHFV transmission in wild animal populations. In the laboratory, LDV infection is rarely transmitted between mice housed in the same cage, even though infected mice excrete virus in their feces, urine, milk and saliva, unless the cage mates are fighting males. Since LDV in mice and SHFV in patas monkeys is produced throughout the lifetime of an infected animal, the transfer of fluids or tissues from an infected animal to an uninfected one results in the inadvertent transfer of infection. Historically, the most frequent mode of transmission of LDV among laboratory mice and of SHFV from patas monkeys to macaques has been through experimental procedures such as the use of the same needle for sequential inoculation of several animals. Currently, the most frequent sources of LDV contamination are pools of other infectious agents or tumor cell lines that have been repeatedly passaged in mice, especially those first isolated in the 1950s. Such materials should be checked for the presence of LDV and, if found to be contaminated, can be ‘cured’ of LDV by repeated passage in culture. It has been suggested, but not proven, that SHFV can be transmitted between macaques via the respiratory route. No evidence for transmission of EAV and PRRSV by insect vectors in domestic animal populations has been reported. Horizontal transmission of EAV and PRRSV can occur via the respiratory route as well as via the sexual route in semen. Vertical transmission of PRRSV in utero has been reported. Tissue Tropism The primary target cells of all four arteriviruses are macrophages. LDV replicates in an uncharacterized subpopulation of murine macrophages. Virus target cells are located in tissues as well as in the blood. Cells containing LDV-specific antigen have been identified in sections of liver and spleen by indirect immunofluorescence. In spleen, the virus-infected cells were nucleated and located in the red pulp. In liver, only Kupffer cells contained LDV-specific antigen. In C58 and AKR mice infected with a neurotropic strain of LDV, designated LDV-C, virus replication was demonstrated in ventral motor neurons. Measurement of the amount of virus in various tissues during natural EAV infections indicated that lung macrophages were the first host cells to be infected. Bronchial lymph nodes subsequently became infected and then the virus spread throughout the body via the circulatory system. In fatally infected horses, lesions are found in subcutaneous tissues, lymph nodes and viscera. PRRSV has been reported to replicate in testicular germ cells which could result in excretion of virus into the semen. Pathogenicity and Clinical Features of Infection Avirulent and virulent strains of EAV and PRRSV have been isolated. Attenuated vaccine strains of EAV and PRRSV have been selected and used as vaccines. Although these vaccines induce immunity against disease, immunized animals are not protected from reinfection. Horses infected with the virulent Bucyrus strain of EAV develop a high fever, lymphopenia and severe disease symptoms. LDV-C differs from other LDV isolates in its ability to infect ventral motor neurons in immunosuppressed C58 and AKR mice and induce poliomyelitis. Isolates of SHFV that induce persistent, asymptomatic infections and ones that cause acute, asymptomatic infections in patas monkeys have been reported. All SHFV isolates cause fatal hemorrhagic fever in macaque monkeys. Mice infected with LDV usually display no overt symptoms of disease. A distinguishing feature of LDV infections is the chronically elevated levels of seven serum enzymes, LDH (8–10-fold), isocitrate dehydrogenase (5–8-fold), malate dehydrogenase (2–3-fold), phosphoglucose isomerase (2–3-fold), glutathione reductase (2–3-fold), aspartate transaminase (2–3-fold) and glutamate–oxaloacetate transaminase (2–3-fold). The elevated LDH levels result primarily from a decreased rate of clearance of normal turnover enzyme from the blood. A permanent reduction in the population of Kupffer cells involved in receptor-mediated endocytosis of LDH has been shown to occur soon after infection of mice with LDV. A decrease in the humoral and cellular immune response to non-LDV antigens is observed during the first 2 weeks after LDV infection. Thereafter, the immune response to other antigens is normal. In immunosuppressed C58 and AKR mice, one isolate of LDV, LDV-C, can induce a sometimes fatal poliomyelitis. Immunosuppression is required to delay antibody production so that virus can reach the central neurons system (CNS) and infect the susceptible ventral motor neurons. LDV-infected neurons become the targets of an inflammatory response. In mice 6 months of age or older, paralysis of one or both hind limbs and sometimes a fore limb is observed. In younger C58 mice, poliomyelitis is usually subclinical. Both EAV and PRRSV can cause either asymptomatic infections or induce various disease symptoms such as respiratory disease, fever, necrosis of small muscular arteries and abortion. The severity of disease caused by EAV and PRRSV depends on the strain of virus as well as the condition and age of the host animal. The most common symptoms of natural EAV infections in horses are anorexia, depression, fever, conjunctivitis, edema of the limbs and genitals, rhinitis, enteritis, colitis and necrosis of small arteries. If clinical symptoms occur, they are most severe in young animals and pregnant mares. Infections in pregnant mares are often inapparent, but can result in a high percentage (50%) of abortions. Young animals sometimes develop a fatal bronchopneumonia, but natural infections are usually not life-threatening. In contrast, about 40% of pregnant mares and foals experimentally inoculated with EAV die as a result of the infection. SHFV causes asymptomatic acute or persistent infections in patas monkeys, but a fatal hemorrhagic fever in macaques. Infected macaques develop fever and mild edema followed by anorexia, dehydration, adipsia, proteinuria, cynosis, skin petechia, bloody diarrhea, nose bleeds and occasional hemorrhages in the skin. The pathological lesions consist of capillary-venous hemorrhages in the intestine, lung, nasal mucosa, dermis, spleen, perirenal and lumbar subperitoneum, adrenal glands, liver and periocular connective tissues. These signs and symptoms are not unique to SHFV-infected animals, since they are also observed after infection with other types of viruses that cause hemorrhagic fevers. Although the SHFV-induced lesions are widespread, the level of tissue damage is not severe. Even so, the mortality of SHFV infections in macaques approaches 100%. Pathology and Histopathology Although most LDV infections are inapparent in mice, some histopathogenic changes are observed in infected animals. As described above, the serum levels of seven enzymes are chronically elevated in LDV-infected mice. Normally, an increase in the serum levels of tissue enzymes is the result of tissue damage, but in LDV-infected animals little tissue damage is observed. Although there are five naturally occurring LDH isozymes in mouse plasma, only the level of isozyme LDH 5 is elevated. Studies have indicated that the increase in enzyme levels is primarily the result of a decreased rate of enzyme clearance. A recent report showed that a subpopulation of Kupffer cells involved in receptor-mediated endocytosis of LDH is severely diminished in mice by 24 h after LDV infection. It has been postulated that LDV replication in these cells causes their death and the depletion of these cells results in the slower turnover of LDH. Splenomegaly, characterized by a greater than 30% increase in spleen weight, occurs in about 40% of the mice infected with LDV. The increase in spleen weight is observed by 24 h after infection and persists for up to a month. A marked necrosis of lymphocytes in thymic-dependent areas occurs during the first 4 days after LDV infection together with a transient decrease in the number of circulating T lymphocytes between 24 and 72 h after infection. A transient decrease in peritoneal macrophages is also observed between the first and 10th day of infection. Despite the lifelong presence of circulating viral immune complexes and the demonstration of LDV antibody deposits in the kidneys of LDV-infected mice as early as 7 days after infection, infected animals do not develop kidney disease. It has been suggested that nephritis does not develop in chronically infected mice because of the inability of the majority of the LDV-antibody complexes to bind C1q. Low levels of C1q-binding activity can only be detected between days 10 and 18 after LDV infection. The central nervous system lesions in LDV-C-infected C58 and AKR mice are located in the gray matter of the spinal cord and, occasionally, in the brainstem and consist of focal areas of inflammatory mononuclear cell infiltrate in the ventral horn. Virus-specific protein and nucleic acid have been detected and maturing virions have been observed in ventral motor neurons. In horses experimentally or fatally infected with EAV, the most common gross lesions are edema, congestion and hemorrhage of subcutaneous tissues, lymph nodes and viscera. Microscopic investigation of tissues from chronically infected horses, which have mildly swollen lymph nodes and slightly increased volumes of pleural and peritoneal fluids, revealed extensive lesions consisting of generalized endothelial damage to blood vessels of all sizes and severe glomerulonephritis. Both types of lesions are thought to be caused by the deposition of viral immune complexes. Extensive capillary necrosis leads to a progressive increase in vascular permeability and volume, hemoconcentration and hypotension. During the terminal stages of the disease, lesions are also found in the adrenal cortex, and degenerative changes are observed in the bone marrow and liver. Virus infection causes focal myometritis which is thought to be the cause of deficiencies in the fetal and placental blood supply. The resulting anoxia is probably the cause of abortion. Immune Response The persistence of LDV infectivity in the plasma of infected mice and the failure of initial attempts to demonstrate the presence of neutralizing antibodies led investigators to postulate that LDV-specific antiviral antibodies were not produced. It was subsequently demonstrated that antiviral antibodies were efficiently produced in LDV-infected animals, but were complexed with virions. Whereas 99% of the infectivity in sera collected 24 h after infection could be neutralized with ether-extracted murine anti-LDV immunoglobulin, no neutralization of the LDV infectivity in sera from chronically infected mice could be demonstrated unless antibody to mouse immunoglobulin was also used. Anti-LDV antibody not complexed to virus can be detected by 15 days after infection, indicating that antibody is present in excess of virus in chronically infected mice. Studies with nude mice suggest that LDV is a T-independent antigen. LDV-infected mice display a polyclonal humoral response, but anti-LDV antibody apparently accounts for only a small portion of this polyclonal response. The mechanism by which LDV infection activates B cells polyclonally is currently not known, but mice immunized with inactivated virus do not develop a polyclonal response. Autoantibodies to a variety of cellular components (autoimmune antibodies) have been detected in mice chronically infected with LDV. LDV-infected animals develop cytotoxic T cells that can specifically lyse virus-infected macrophages after infection. However, the cytotoxic response does not clear the infection. A T cell response has also been detected in PRRSV-infected pigs. Anti-EAV antibodies can be detected in horses 1 to 2 weeks after infection with virulent or avirulent strains of the virus by either serum-neutralization or complement-fixation assays. Virus neutralization is enhanced by the presence of fresh complement. Newborn foals with mothers that are immune receive protective antibodies in the colostrum. Complement-fixing, antiviral antibodies peak at 2–3 weeks after the initiation of infection and then decline. Neutralizing antibody levels peak between 2 and 4 months after infection. Often after 8 months, anti-EAV antibody can no longer be detected by complement-fixation or neutralization assays, but in some animals the virus persists and viral immune complexes may continue to circulate. Neutralizing antibodies in sera obtained from EAV- and LDV-infected animals are ORF 5 protein-specific and the neutralizing epitope has been mapped to the ectodomain of this protein. The number of gylcosylation sites in the ectodomain of the ORF 5 protein varys in different LDV strains and it has been postulated that antibodies bind less efficiently to virions with extensive glycosylation in this region. The induction of neutralizing antibodies by the ORF 4 protein of PRRSV has also been reported. SHFV isolates that induce acute infections in patas monkeys induce high levels of neutralizing antibody, whereas SHFV isolates that induce persistent infections induce low titers of non-neutralizing antibody. Antibodies to virus that causes acute infection do not cross-neutralize virus that causes persistent infection. Prevention and Control The most effective means of prevention of LDV, SHFV, PRRSV and EAV infection is interruption of animal to animal transmission. Infected animals should be destroyed or isolated. However, the current lack of rapid diagnostic assays for the detection of LDV and SHFV in persistently infected animals means that it is still a time-consuming task to identify animals with inapparent infections. Materials obtained from animals that might be infected with an arterivirus should be checked for viral contamination before they are injected into a susceptible animal. Multiple animals should not be injected using the same needle. Killed and attenuated live vaccines for EAV and PRRSV are commercially available. The live vaccines are more effective and induce a longer-lasting immunity than the killed vaccines. Although animals immunized with the live vaccines are protected from disease, they are not protected from reinfection and can spread virus. Current serological assays can not distinguish field strains from vaccine strains. Future Perspectives Arteriviruses have so far been isolated from mice (LDV), horses (EAV), pigs (PRRSV) and monkeys (SHFV). It seems likely that other host species, including humans, may harbor additional members of this virus family. Such viruses will be especially difficult to find in natural hosts that develop asymptomatic infections. Because of the inapparent and persistent nature of infections caused by LDV, PRRSV, SHFV and EAV it is important to develop rapid and reliable diagnostic tests for these viruses to easily identify infected animals. The availability of complete genomic sequences for EAV, PRRSV, SHFV and LDV should facilitate the development of new diagnostic assays and vaccines and may also provide the means for detecting additional arteriviruses. See also: IMMUNE RESPONSE | Cell Mediated Immune Response; IMMUNE RESPONSE | General Features; VACCINES AND IMMUNE RESPONSE.