12 Multiple Systems
The agents considered in this group appear to affect multiple systems. However, this may simply reflect the present lack of understanding of these host-parasite relationships. Included in the group are three parvoviruses (Kilham rat virus, H-1 virus, and minute virus of mice), polyoma virus, and the hantaviruses. The hantaviruses are a group of closely related viruses that occur primarily in wild rodents but that also have been found in laboratory rats, mainly in Japan and Europe. Some of the hantaviruses cause serious zoonotic infections in people.
Kilham Rat Virus
There are few reports of natural disease or interference with research due to this virus, but its high prevalence in rats and propensity for damaging populations of replicating cells in vivo and in vitro make it a significant pathogen.
Kilham rat virus (KRV)* was discovered by Kilham and Olivier (1959) during attempts to isolate a suspected oncogenic virus from rats with ex-
perimental tumors (a Cysticercus fasciolaris-induced sarcoma in F344 rats and a transplantable leukemia in OM rats). In subsequent years research with this agent emphasized molecular biology of the virus (Tattersall and Cotmore, 1986) and experimental animal models for producing lesions such as cerebellar hypoplasia (Kilham and Margolis, 1975; Margolis and Kilham, 1975) while relatively little attention was given to understanding the natural history and pathogenesis of KRV in rat colonies (Jacoby et al., 1987).
KRV is a single-stranded DNA virus, family Parvoviridae, genus Parvovirus. KRV is the type species of the genus (Siegl et al., 1985). It is synonymous with rat virus (RV) and parvovirus r-1. Approximately a dozen strains of the virus have been isolated (Tattersall and Cotmore, 1986). The virion is non-enveloped, and measures 18-30 nm in diameter. The genus Parvovirus presently contains 13 distinct serotypes, three of which occur in rodents: KRV, H-1 virus, and Minute Virus of Mice. Serotyping is based on the neutralization (NT), complement fixation (CF), and hemagglutination inhibition (HAI) tests (Jacoby et al., 1979; Tattersall and Cotmore, 1986). KRV can be propagated in primary rat embryo, 324K, and BHK-21 cells (Jacoby et al., 1979, 1987).
Parvoviruses are remarkably resistant to environmental conditions. Infectivity is retained after heating at 80°C for 2 hours or 40°C for up to 60 days. They also are resistant to dessication, pH 2 to 11, chloroform, ether, and alcohol (Toolan, 1968; Jacoby et al., 1979; Tattersall and Cotmore, 1986).
Laboratory and wild rats (Rattus norvegicus) are the natural hosts. Syrian hamsters and several other species have been infected experimentally (Kilham, 1966; Jacoby et al., 1979; Tattersall and Cotmore, 1986).
KRV is a common natural infection of wild and laboratory rats. Kilham (1966) found antibodies to KRV in 40% to 62% of wild rats at four different locations near Hanover, New Hampshire, and in 89% of a population of laboratory rats. More recent serological surveys of laboratory rats reported prevalences of 44% (Lindsey et al., 1986a) and 71% (Parker, 1980) in the United States, 74% in Canada (Lussier and Descoteaux, 1986), 60% in the United Kingdom (Gannon and Carthew, 1980), and 32% in West Germany (Kraft and Meyer, 1986).
Transmission of KRV is primarily by the horizontal route, either through
direct contact or fomites. Contaminated fomites are probably important because parvoviruses are extremely stable to dessication. The efficiency of horizontal transmission may depend on strain of virus and age at which rats are infected. Virus has been reported to be shed in urine, feces, milk, and nasal secretions (Lipton et al., 1973; Kilham and Margolis, 1974a,b; Robinson et al., 1974; Tattersall and Cotmore, 1986; Jacoby et al., 1987, 1988). Rats infected by the oronasal route at two days of age transmitted the virus to cagemates for 10 weeks, whereas rats infected at 4 weeks of age were able to infect their cagemates for only 3 weeks post infection (Jacoby et al., 1988). Transplacental infection has been suggested (Kilham and Margolis, 1966, 1969; Robey et al., 1968; Novotny and Hetrick, 1970), but its occurrence has been questioned on the basis of more recent experimental data (Tattersall and Cotmore, 1986; Jacoby et al., 1988). Persistent infection has been shown to occur for up to 14 weeks in rats infected as infants (Paturzo et al., 1987).
KRV has been shown to persistently infect a number of cultured cell lines (Kilham et al., 1968; Wozniak and Hetrick, 1969; Lum, 1970; Bass and Hetrick, 1978) and to be a frequent contaminant of transplantable tumors (Kilham and Olivier, 1959; Lum and Schreiner, 1963; Kilham and Moloney, 1964; Tattersall and Cotmore, 1986). Thus contaminated cells and tumors are important sources of KRV infection.
Natural infections nearly always are subclinical. Clinical disease is a rare event, and few examples have been reported (Kilham and Margolis, 1966; Coleman et al., 1983).
Natural disease due to KRV was observed in 13-day pregnant rats by Kilham and Margolis (1966). They reported increased numbers of uterine resorption sites in the dams and runting, ataxia, cerebellar hypoplasia, and jaundice in pups from several litters. KRV was isolated from the pups, as well as from tissues, feces, and milk of the dams.
Spontaneous deaths, scrotal cyanosis, abdominal swelling, dehydration, and other signs of severe illness occurred in juvenile and 7-week-old male rats within 2 weeks after these animals were added to a colony of adult rats that were serologically positive for KRV (Coleman et al., 1983)
KRV infects actively replicating (S-phase) cells and thus, has a predilection for proliferating and growing tissues. Also, productive infection is always lytic and thus, infection usually leads to cell destruction (Margolis
et al., 1971; Tattersall and Cotmore, 1986). For these reasons, the major pathologic effects of KRV infection would be expected to occur during fetal development and early life. Although the number of reports is small, this has been borne out in natural infections. Although transplacental infection is probably not important as a mode of transmission (Jacoby et al., 1988), fetal deaths and teratologic effects have been attributed to KRV in a few instances (Kilham and Margolis, 1966; Jacoby et al., 1979). Natural infections of KRV also have been the cause of severe disease in rats from birth to 9 weeks of age. The findings in this group have included stunting, cerebellar hypoplasia, jaundice, and hemorrhagic infarction with thrombosis in multiple organs (including brain, spinal cord, testes, and epididymis). Amphophilic intranuclear inclusions occurred in the endothelium and other cells of affected organs (Coleman et al., 1983). Focal necrosis, hypertrophy and vacuolar degeneration of hepatocytes, cholangitis, and biliary hyperplasia can occur in the liver (Kilham and Margolis, 1966; Jacoby et al., 1979; Coleman et al., 1983). Hemorrhagic encephalopathy was reported to occur in one group of naturally infected adult LEW rats following the administration of 100-150 mg/kg of cyclophosphamide (El Dadah et al., 1967).
Experimental infections of KRV, in which large doses of virus were inoculated by unnatural routes, have been used in developing a large array of experimental animal models (reviewed by Kilham and Margolis, 1975; Margolis and Kilham, 1975). These have included models of teratologic effects (Kilham and Ferm, 1964), cerebellar hypoplasia (Margolis and Kilham, 1968; Margolis et al., 1971), hepatitis (Bergs and Scotti, 1967; Margolis et al., 1968), and hemorrhagic encephalopathy (Cole et al., 1970; Margolis and Kilham, 1970; Nathanson et al., 1970; Baringer and Nathanson, 1972) in rats, and of cerebellar hypoplasia, "mongolism," and periodontal disease in Syrian hamsters (Kilham, 1961a,b; Baer and Kilham, 1962a,b, 1964; Toolan, 1968; Lipton and Johnson, 1972).
Serologic methods are used for routine health monitoring. The enzyme-linked immunosorbent assay (Singh and Lang, 1984) and the indirect fluorescent antibody test are considered most sensitive, but do not discriminate between infections due to different serotypes of parvoviruses in rats (i.e., KRV or H-1 virus). The HAI, CF, and NT tests are used for serotype discrimination (Cross and Parker, 1972; Tattersall and Cotmore, 1986).
For the diagnosis of natural disease outbreaks due to KRV, virus serology, demonstration of typical pathologic lesions, and virus isolation should be carried out (Coleman et al., 1983). The virus can be propagated in primary rat embryo, 324K, and BHK-21 cells (Jacoby et al., 1979, 1987).
The most practical approach to control of KRV infection is to obtain only rats demonstrated to be free of the virus by regular serologic monitoring. Also, transplantable tumors, cell lines, and virus stocks should be tested for KRV infection by the mouse antibody production (MAP) test before being brought into an animal facility (Rowe et al., 1959a).
For those breeding colonies that have the infection, the most practical solution may be to eliminate the colony and repopulate with KRV-free stock. To eliminate the infection from valuable breeding stocks of rats, cesarean derivation should be successful because transplacental transmission is not considered important (Jacoby et al., 1987, 1988; Paturzo et al., 1987). Alternatively, it might be possible to isolate individual breeding pairs of rats from KRV-infected populations in separate containment devices such as filter-top cage systems, with subsequent selection of seronegative progeny as breeders (Lipman et al., 1987; Weir et al., 1987).
Interference with Research
Although transplacental transmittal is generally considered unimportant, early work did suggest such transmission (Kilham and Margolis, 1966). If transplacental transmittal does occur, KRV infection might interfere with studies of fetal development and might have teratogenic effects.
KRV can contaminate transplantable tumors (Kilham and Olivier, 1959; Lum and Schreiner, 1963; Kilham and Moloney, 1964; Tattersall and Cotmore, 1986) and rat cell tissue cultures (Dawe et al., 1961; Wozniak and Hetrick, 1969; Lum, 1970; Bass and Hetrick, 1978; Jacoby et al., 1979).
KRV has been reported to suppress the development of leukemia due to Moloney virus (Bergs, 1969).
KRV infection has been reported to alter in vitro lymphocyte responses (Campbell et al., 1977a,b) and cytotoxic lymphocyte activity (Darrigrand et al., 1984).
Inapparent KRV infection can be exacerbated to cause clinical disease by immunosuppression (El Dadah et al., 1967).
KRV infection has been reported to induce interferon production in vivo (Kilham et al., 1968).
Although quite prevalent in contemporary rat stocks, this virus is not known to cause natural disease and has been shown to interfere with research
in very few instances. It has been found useful for producing a few experimental infection models in animals.
MVM was isolated from a human tumor cell line (HEp-1) after it had been passaged in rats (Toolan et al., 1960; Toolan, 1961), and therefore, the virus was probably of rat origin (Tattersall and Cotmore, 1986). Like Kilham rat virus, H-1 virus has been used to produce experimental malformations of the central nervous system, skeleton and teeth, particularly in rats and hamsters (reviewed by Toolan, 1968; Kilham and Margolis, 1975; Margolis and Kilham, 1975). The natural history of H-1 virus in rat populations has been studied very little.
H-1 virus is a single-stranded DNA virus, family Parvoviridae, genus Parvovirus. The virion is non-enveloped and measures 18-30 nm in diameter. Approximately five strains of H-1 virus have been characterized (Jacoby et al., 1979; Tattersall and Cotmore, 1986). The genus Parvovirus presently contains 13 distinct serotypes, three of which occur in rodents: H-1 virus, Kilham rat virus and Minute Virus of Mice. Serotyping is based on the neutralization (NT), complement fixation (CF) and hemagglutination inhibition (HAI) tests (Tattersall and Cotmore, 1986). Parvoviruses can be propagated in primary rat embryo, 324K, and BHK-21 cells (Jacoby et al., 1979, 1987). They are remarkably resistant to environmental conditions. Infectivity is retained after heating at 80°C for 2 hours or 40°C for up to 60 days. They also are resistant to dessication, pH 2 to 11, chloroform, ether, and alcohol (Toolan, 1968; Jacoby et al., 1979; Tattersall and Cotmore, 1986).
Laboratory and wild rats (Rattus norvegicus) are the natural hosts. Syrian hamsters and several other species have been infected experimentally (Kilham, 1966; Kilham and Margolis, 1975; Margolis and Kilham, 1975; Tattersall and Cotmore, 1986).
H-1 virus is a common natural infection of wild and laboratory rats. Kilham (1966) found antibodies to KRV in none to 100% of wild rats at four different locations near Hanover, New Hampshire, and in 80% of a population of laboratory rats. More recent serological surveys of laboratory rats have given prevalences of 11% (Lindsey et al., 1986a) and 52% (Parker, 1980) in the United States, 71% in Canada (Lussier and Descoteaux, 1986), and 27% in West Germany (Kraft and Meyer, 1986).
The epizootiological characteristics of H-1 virus infection in rat colonies are generally assumed to be similar to those of Kilham rat virus infection which has been investigated much more extensively (Robey et al., 1968; Lipton et al., 1973; Kilham and Margolis, 1974a,b; Robinson et al., 1974; Tattersall and Cotmore, 1986; Jacoby et al., 1987, 1988). Transmission is primarily horizontal, and virus is shed in urine, feces, nasal secretions, and milk. Transplacental infection in rats has been demonstrated only after parenteral inoculation of pregnant dams (Kilham and Margolis, 1969). Viral persistence and latency have not been adequately studied (Tattersall and Cotmore, 1986).
Clinical signs have not been reported to occur in natural infections of H-1 virus.
Natural infections of H-1 virus are not known to cause disease.
As with other rodent parvoviruses, H-1 virus infection requires replicating (S-phase) cells and productive infection is always lytic (Margolis et al., 1971; Margolis and Kilham, 1975; Tattersall and Cotmore, 1986). Because of these basic virus-host relationships, experimental infections of fetal and infant hamsters and rats with H-1 virus have provided models of disease including cerebellar hypoplasia, runting, "mongolism," bone and tooth abnormalities, and multiple organ malformations (Toolan, 1960, 1968; Ferm and Kilham, 1964, 1965; Margolis and Kilham, 1968, 1975; Kilham and Margolis, 1975).
The diagnosis of H-1 infection is usually made during health surveillance testing as natural infections are inapparent. The enzyme-linked immunosorbent assay and indirect fluorescent antibody test are usually used
for initial screening, followed by either the HAI, CF, or NT tests for discriminating between H-1 virus and Kilham rat virus infections (Cross and Parker, 1972; Tattersall and Cotmore, 1986). For virus isolation primary rat embryo, 324K, and BHK-21 cells can be used (Jacoby et al., 1979, 1987).
The same measures as recommended for Kilham rat virus should be effective in the control of H-1 virus.
Interference with Research
It may be possible for H-1 virus to alter the results of studies concerned with fetal development or teratogenesis (Kilham and Margolis, 1969); but this has not been reported to occur as a result of natural infection.
H-1 virus has been reported to cause hepatocellular necrosis in infected rats when they are subjected to liver injury by hepatotoxic chemicals, parasitism, or partial hepatectomy (Ruffolo et al., 1966; Kilham et al., 1970).
H-1 virus infection has been found to inhibit experimental tumor induction by adenovirus 12 (Toolan and Ledinko, 1968) and dimethylbenzanthracene (Toolan et al., 1982) in hamsters.
Minute Virus of Mice
The significance of minute virus of mice (MVM) is uncertain. It has little significance for most studies, but may be highly significant for studies involving mouse transplantable tumors, leukemias, and in vitro immunoassays.
MVM was first recognized as a contaminant in a stock of mouse adenovirus by Hartley and Rowe (1960). It was later described and named by Crawford (1966), and further characterized by Crawford et al. (1969). Subsequently, it was identified as a prevalent indigenous infection in mouse colonies (Parker et al., 1970b), and as one of the most common contaminants of mouse leukemia virus stocks and transplantable tumors (Collins and Parker, 1972). Two strains of the virus MVM(p) and MVM(i) have been of particular interest to molecular biologists (Ward and Tattersall, 1982; Tattersall and Cotmore, 1986).
MVM is a single-stranded DNA virus, family Parvoviridae, genus Parvovirus. The genus Parvovirus presently contains 13 distinct serotypes, three of which occur in rodents: MVM, H-1 virus, and Kilham rat virus. Serotyping is based on the neutralization (NT), complement fixation (CF), and hemagglutination inhibition (HAI) tests (Siegl et al., 1985; Mengeling et al., 1986; Tattersall and Cotmore, 1986).
The virion of MVM measures 26 nm in diameter, has an icosahedral capsid with 32 capsomeres, and is non-enveloped. The complete nucleotide sequence of the genome is 5081 nucleotides long. The genome encodes for three proteins that make up the viral capsid, VP-1, VP-2, and VP-3, plus two nonstructural proteins, NS-1 and NS-2 for which the function is uncertain (Ward and Tattersall, 1982; Astell et al., 1983; Cotmore et al., 1983; Cotmore and Tattersall, 1986).
Recognized strains of MVM include MVM(CR), the original strain of Crawford (1966); MVM(p), the prototype strain derived by plaque purification from MVM(CR); MVM(i), the immunosuppressive variant discovered by Bonnard et al. (1976); MVM-890; and CZ-7 (Tattersall and Cotmore, 1986). Two strains have been studied extensively, MVM(i) and MVM(p). MVM(i) grows lytically in mouse T-lymphocytes (thus, sometimes referred to as the lymphotropic strain), inhibits various functions mediated by T-lymphocytes in vitro, and is restricted in fibroblasts. MVM(p) grows in mouse fibroblasts (the fibrotropic strain), is restricted in T-lymphocytes, and does not inhibit T-lymphocyte functions (McMaster et al., 1981; Spalholz and Tattersall, 1983; Tattersall and Bratton, 1983). The nucleotide sequences of the genomes of both MVM(p) (Astell et al., 1983) and MVM(i) (Sahli et al., 1985) have been determined. The genomic organization of the two strains is identical but there are 29 amino acid differences in their putative viral gene products (Sahli et al., 1985). Both the lymphotropic and fibrotropic determinants have been found to map to the same 237-nucleotide sequence within the coding region of the virus structural gene (Gardiner and Tattersall, 1988b).
MVM can be propagated in A9 fibroblast, 324K (SV40 transformed human newborn kidney), and EL4 mouse T-cell lymphoma cells, and rat and mouse embryo tissue cultures (Parker et al., 1970a; Gardiner and Tattersall, 1988a).
Parvoviruses are remarkably resistant to environmental conditions. Infectivity is retained after heating at 80°C for 2 hours or 40°C for up to 60 days. They also are resistant to dessication, pH 2 to 11, chloroform, ether, and alcohol (Toolan, 1968; Tattersall and Cotmore, 1986).
Laboratory and wild mice (Mus musculus) are the natural hosts (Tattersall and Cotmore, 1986).
MVM is a common infection of wild and laboratory mice. Parker et al. (1970b) found 76 (20%) of 390 sera from wild Mus musculus trapped in Florida, Georgia, Maryland, and Indiana to be positive for MVM by the HAI test; positives were found in mice from all four states. In laboratory mice Parker et al. (1970b) found serological evidence of the infection in 38 (86%) of 44 conventional colonies, three (38%) of eight "specific pathogen-free" colonies, and none of five germfree mouse colonies. More recent surveys of laboratory mice have given prevalences for colonies infected with MVM of 73% (Lindsey et al., 1986) and 81% (Parker, 1980) in the United States, 50% in Canada (Lussier and Descoteaux, 1986), 33% in West Germany (Kraft and Meyer, 1986), and 3% in France (Van Der Logt, 1986).
MVM is also a common contaminant of mouse leukemia virus stocks and transplantable tumors. In one study, Parker et al. (1970b) found 79 (40%) of 195 mouse leukemias and transplantable tumors to be contaminated. Later, Collins and Parker (1972) demonstrated MVM in 151 (32%) of 465 mouse leukemia virus stocks and transplantable tumors. In the latter study, MVM ranked second to lactic dehydrogenase-elevating virus in occurrence as a viral contaminant.
MVM is highly contagious. Virus is shed in the urine and feces. Fomites such as contaminated food and bedding are particularly important because the virus is very resistant to environmental conditions. Nasal-oral contact is also effective in transmission but aerosols are probably not important because uninfected mice in cages eight inches away from infected mice do not become infected. Maternal antibody is protective until 6 to 8 weeks of age but most mice in infected colonies become infected by two or three months of age. Transplacental infection is not considered important (Parker et al., 1970b).
Natural infections are inapparent (Parker et al., 1970b).
Natural infections of MVM are not known to produce disease (Parker et al., 1970a,b; Ward and Tattersall, 1982; Tattersall and Cotmore, 1986).
Experimental infections of MVM(CR) in fetal and neonatal mice, rats, and hamsters produce lesions such as runting, cerebellar hypoplasia, and periodontal disease (Kilham and Margolis, 1970, 1971; Baer and Kilham, 1974). However, the lesions due to MVM(CR) are relatively mild by comparision to the lesions produced experimentally by Kilham rat virus or H-1 virus (Kilham and Margolis, 1975; Margolis and Kilham, 1975).
Serologic methods are used for routine health surveillance. The enzymelinked immunosorbent assay and the immunofluorescent antibody test are considered most sensitive (Tattersall and Cotmore, 1986). Mice of the MRL/MpJ strain are reported to give frequent false positive HAI test results for MVM (Jonas, 1984). Transplantable tumors and other biologic materials from mice can be screened by the mouse antibody production (MAP) test (Rowe et al., 1959) and/or virus isolation (deSouza and Smith, 1989). The HAI, CF, and NT tests can be used in situations where discrimination between MVM, Kilham rat virus, and H-1 virus infection is desired (Cross and Parker, 1972; Tattersall and Cotmore, 1986).
MVM can be eliminated from stocks of mice by cesarean derivation, but elimination of infected mice followed by replacement with MVM-free mice is often more practical. Maintenance of MVM-free mice can be difficult because the virus is highly contagious. Adherence to strict barrier procedures is required. Careful attention to the exclusion of wild mice and prompt elimination of MVM-infected stocks of laboratory mice in other parts of the facility are essential. Regular serologic monitoring of all mouse subpopulations must be practiced for early detection (Parker et al., 1970b).
Infection can become established in a colony through introduction of contaminated transplantable tumors, virus stocks, and other biologic materials that are passaged in mice. All such materials should be monitored for MVM and other infectious agents before admission to a facility (Rowe et al., 1959; Parker et al., 1970b; de Souza and Smith, 1989).
Interference with Research
Direct evidence that wild type MVM in contemporary mouse stocks interferes with research is lacking. However, it is clear that MVM is a common contaminant of mouse leukemia virus stocks, transplantable tumors, hybridomas, and other cell lines (Parker et al., 1970a; Collins and Parker, 1972;
de Souza and Smith, 1989). Thus, it seems highly probable that studies involving these materials are at times compromised because of inadvertent MVM contamination.
Considerable evidence that MVM can interfere with research has come from studies of MVM(i), a single variant of the virus that may or may not occur as a natural infection in contemporary mice. MVM(i) grows lytically in cytotoxic T-lymphocyte clones, abrogates cytotoxic T-lymphocyte responses, suppresses T-lymphocyte mitogenic responses and suppresses T helper-dependent B-lymphocyte responses, in vitro (Bonnard et al., 1976; Herbermann et al., 1977; Engers et al., 1981; McMaster et al., 1981; Tattersall and Cotmore, 1986).
The intramuscular inoculation of MVM(p) into mice suppresses the growth of Ehrlich ascites tumor cells given intraperitoneally (Guetta et al., 1986).
A parvovirus serologically related to MVM has been reported as a contaminant of calf serum (Nettleton and Rweyemamu, 1980).
Polyoma virus has limited significance as a naturally occurring infection of mice likely to interfere with research. Its major importance has been for use in experimental model systems, including in vivo viral carcinogenesis (Eddy, 1982), in vitro transformation of cells (Topp et al., 1980), and certain aspects of human polyoma virus infection (Shah and Christian, 1986).
1953: Gross (1953a,b) and Stewart (1953) independently observed sarcomas in mice that had been inoculated with extracts of murine leukemia virus.
1957: Stewart et al. (1957) demonstrated that polyoma virus (which had caused the sarcomas in the 1953 study described above) replicates in tissue culture.
1958: Eddy et al. (1958) discovered that inoculation of the virus into newborn hamsters led to tumor induction with a shorter latent period than in mice.
1961: Wild mice were shown to have a high prevalence of polyoma virus infection (Rowe, 1961; Rowe et al., 1961).
1972: The virus was found to be a common contaminant of transplantable murine tumors and stocks of murine leukemia viruses (Collins and Parker, 1972).
Polyoma virus is a DNA virus, family Papovaviridae, genus Polyoma virus. Mouse polyoma virus is the type species of the genus. It is related to K virus, simian virus 40, rabbit kidney vacuolating virus, and others. The virion is spherical, measures 44 nm in diameter, and icosahedral. It hemagglutinates guinea pig, hamster, sheep, and human O erythrocytes at 4°C. It can be propagated in primary cultures of mouse embryo or mouse kidney, and in 3T3 and 3T6 mouse fibroblasts. It is highly resistant to environmental conditions and to physical and chemical agents (Eddy, 1982; Shah and Christian, 1986).
Wild and laboratory mice (Mus musculus) are the natural hosts. Wild mice serve as reservoir hosts (Rowe, 1961).
The prevalence of polyoma virus infection is poorly understood. The virus has been demonstrated to occur focally in populations of wild mice (Rowe et al., 1959b, 1961) and to occur in colonies of laboratory mice in the United States, Europe, and Japan (Rowe et al., 1959b; Parker et al., 1966), primarily in colonies housed in close proximity to experimentally infected mice (Parker et al., 1966). It is not likely to be seen in mice from commercial barrier breeding facilities in the United States.
The virus is highly contagious. It is shed in urine, saliva, and feces, but excreted in higher titer and for a longer periods of time in urine. In persistently infected dams, the titer of virus in the kidney increases during late pregnancy (McCance and Mims, 1979). The virus has a propensity for airborne dissemination and intranasal infection. Transmission occurs readily within and between cages in the same room. Contaminated feed and bedding can be important sources of infection. Transplacental transmission can occur but only if mice are infected during gestation (Rowe et al., 1958; Rowe, 1961; McGarrity et al., 1976; McCance and Mims, 1977; McGarrity and Dion, 1978).
The respiratory tract is generally accepted as the natural route of infection of mice with polyoma virus as smaller amounts of virus are required to establish the infection by this route than by the alimentary route (Rowe et al., 1959a, 1960). Dubensky et al. (1984) used an in situ hybridization technique to follow the pattern of infection after intranasal infection of neonatal mice. Primary replication of the virus occurred in the nasal mucosa, submaxillary salivary glands, and lungs. This was followed by a
systemic phase in which viral replication predominated in lung and liver, and also occurred in kidney and colon around six days post infection. The lungs, kidney, and colon appeared to be the main portals for exit of the virus.
Natural infections are usually inapparent (polyoma virus-induced tumors have not been reported in naturally infected immunocompetent mice). Tumors caused by the virus occur more readily in immunocompromised hosts (Eddy, 1982; Rubino and Walker, 1988).
Athymic (nu/nu) mice given heterotransplants of human tumors contaminated with the virus have been reported to develop a syndrome characterized by wasting and paralysis of the rear legs and tail (Sebesteny et al., 1980; McCance et al., 1983).
Morphologic lesions, including polyoma virus-induced tumors, usually are not found in naturally infected mice (Eddy, 1982).
In experimental infection, factors that favor tumor formation include infection at a very young age, inoculation of very large doses of the virus, inoculation of the virus directly into tissues (i.e., use of either the subcutaneous, intravenous, or intracerebral routes), and immunosuppression of the host (Shah and Christian, 1986). Experimental inoculation of the virus into adult mice does not result in tumor induction because an efficient immune response prevents virus-transformed cells from growing to form tumors. A polyoma tumor-specific transplantation antigen (TSTA) is induced in the early phase of polyoma virus infection prior to DNA synthesis, and protection from tumor development is T-cell-dependent and directed toward this antigen (Allison, 1980; Ito, 1980; Eddy, 1982; Ramquist et al., 1988).
In neonatal mice inoculated with high doses of virus, transformed cells multiply to numbers that cannot be controlled effectively by the immature immune system. Tumors appear around 2-3 months and most develop by 4-7 months postinoculation. Infection of neonates can be prevented by maternal antibody in infected colonies (Ito, 1980; Eddy, 1982; Shah and Christian, 1986).
Pleomorphic salivary gland tumors are the most common lesions occurring in experimentally infected neonatal mice. Sarcomas and carcinomas of the kidney, subcutis, mammary gland, adrenal gland, bone, cartilage, blood vessels, and thyroid also occur. Many are transplantable to syngeneic hosts. Strains of mice differ in their susceptibilities, with those of C57BL and
C57BR/cd lineage being among the most resistant and athymic (nu/nu) mice being the most susceptible (Eddy, 1982).
Athymic (nu/nu) mice heterotransplanted with a human tumor contaminated with polyoma virus have been reported to show progressive weight loss and flaccid paralysis of the rear legs and tail. The brain and spinal cord had multifocal demyelination associated with virus particles in the cytoplasm and nuclei of oligodendroglia. The disease was attributed to primary destruction of oligodendroglia by the virus. Intracytoplasmic and intranuclear inclusions also were present in epithelium of the bronchi, ureters, and renal pelvis (Sebesteny et al., 1980; McCance et al., 1983). Subsequent investigations of this syndrome by Harper et al. (1983) gave a different explanation. Most of their nu/nu mice inoculated with polyoma virus developed vertebral tumors resulting in compression of the spinal cord, and they concluded that the lesions in the nervous system occurred secondarily.
Malignant tumors also occur following experimental inoculation of polyoma virus into neonatal hamsters, rats, guinea pigs, mastomys, and ferrets. Newborn rabbits develop benign tumors that subsequently regress. Infections in these species and in mice are extremely important animal models of virus-induced tumors (Barthold and Olson, 1974; Eddy, 1982; Shah and Christian, 1986).
The enzyme-linked immunosorbent assay and the hemagglutination inhibition test are commonly used in routine health monitoring. The mouse antibody production (MAP) test can be used for screening tumor lines and other biologic materials (Rowe et al., 1959a,b, 1961). Tissue sections can be examined for presence of the virus using an immunoperoxidase test for the major capsid polypeptide (Gerber et al., 1980) or in situ DNA hybridization (Dubensky et al., 1984). Virus isolation using primary mouse embryo or mouse kidney cultures, or 3T3 or 3T6 mouse fibroblasts can be useful in selected situations (Shah and Christian, 1986).
Control of polyoma virus is not likely to be a problem except in those animal facilities where mice are given experimental infections of the virus, transplantable tumors or other biological materials contaminated with the virus are introduced into the facility, or wild mice gain access to the facility and transmit the infection to laboratory mice. Once established in a mouse facility, the infection can spread rapidly to adjacent rooms. If experimental infections of polyoma virus are to be carried out in a facility, rigorous containment measures such as the use of plastic film isolators may be
necessary to prevent spread of the infection (McGarrity et al., 1976; Eddy, 1982). Transplantable tumors and other biologic specimens should be screened for the virus before entry into an animal facility.
Cesarean derivation and barrier maintenance are considered very effective in eliminating the virus because transplacental transmission probably is of little or no importance in natural infections (McCance and Mims, 1977). However, more practical methods such as isolation and quarantine of individual breeding pairs with subsequent selection of seronegative progeny or removal of neonates and fostering them on polyoma virus-free dams should be effective (Lipman et al., 1987). Strict isolation of virus-free mice from polyoma virus-infected stocks and exclusion of wild mice are essential for preventing reinfection. Laminar flow units and filter-top cages are helpful in reducing the spread of infection in laboratories in which the agent is used experimentally (McGarrity et al., 1976).
Interference with Research
Polyoma virus can complicate research by contaminating tumor lines, stocks of other viruses, and other biologic materials that are passaged in mice (Collins and Parker, 1972).
Polyoma virus-contaminated tumors can cause paralysis and wasting in athymic (nu/nu) mice (Sebesteny et al., 1980; McCance et al., 1983) due to vertebral tumors and compression of the spinal cord (Harper et al., 1983).
Neonatal mice from naturally infected colonies can be refractory to experimental infection with polyoma virus because of maternal antibody (Eddy, 1982).
Mice subjected to neonatal thymectomy later developed parotid gland tumors because the colony had inadvertently become infected with polyoma virus (Law, 1965).
The genus Hantavirus includes several recently recognized viruses from different regions of the world (LeDuc, 1987). Some of these agents have caused persistent, subclinical infections in laboratory rats (and other rodents introduced into laboratories), resulting in serious illness in research personnel (Kawamata et al., 1980, 1987; Tsai, 1987a). The most important human pathogen in the group, Hantaan virus, has been called "the most significant zoonotic pathogen of laboratory rodents since the discovery of lymphocytic choriomeningitis virus" (Johnson, 1986).
In the period from the early 1940s to the early 1970s there were recognized across Eurasia a large assortment of clinical syndromes in humans, most commonly referred to as hemorrhagic fevers (H. W. Lee et al., 1979). These included hemorrhagic nephrosonephritis in Russia, recognized around 1944 (Smorodintsev, 1959), epidemic hemorrhagic fever (EHF) in China, recognized about 1942-1944 (H. W. Lee et al., 1979), nephropathia epidemica (NE) in Scandinavia (Myhrman, 1951; Lahdevirta, 1971), Korean hemorrhagic fever (KHF) in Korea (Smadel, 1953; Earle, 1954), EHF in Eastern Europe (Gajdusek, 1962), and EHF in Japan (Tamura, 1964).
During the Korean War thousands of cases of KHF, a syndrome often characterized by acute high fever, shock, hemorrhage, and renal failure, occurred in United Nations' forces and attracted worldwide attention (Smadel, 1953; Earle, 1954; Tsai, 1987b). H. W. Lee et al. (1978) isolated the causative agent of KHF from the field mouse, Apodemus agrarius, and named it Hantaan virus (H. W. Lee et al., 1981). Following the discovery of successful cell cultures for Hantaan virus (French et al., 1981; Kitamura et al., 1983), a large number of Hantaan-related viruses were isolated and assigned to the genus, Hantavirus (McCormick et al., 1982; White et al., 1982; Schmaljohn and Dalrymple, 1983).
Naturally infected laboratory rats have been the source of Hantavirus infections in research personnel in Japan (Umenai et al., 1979; Kawamata et al., 1980, 1987), Belgium (Van Ypersele de Strihou, 1979; Desmyter et al., 1983), the United Kingdom (Lloyd et al., 1984), and France (Dournon et al., 1984). In Japan alone, at least 126 human cases and one death have occurred (Kawamata et al., 1987). As of 1987, hantaviruses still had not been eradicated from all animal facilities in Japan (Kawamata et al., 1987). Wild rodents brought into the laboratory have caused human infections of hantaviruses in Korea (H. W. Lee and Johnson, 1982) and Russia (Tsai, 1987a).
The agents are single-stranded RNA viruses, family Bunyaviridae, genus Hantavirus. This is a new genus based on morphological, genetic, antigenic, and physiochemical similarities (McCormick et al., 1982; White et al., 1982; Hung et al., 1983; Schmaljohn and Dalrymple, 1983; Schmaljohn et al., 1983, 1985). Collectively, they have been grouped together by the World Health Organization as the viruses of hemorrhagic fever with renal syndrome (HFRS) in humans (W.H.O., 1983). Hantaan virus, the cause of KHF, is the prototype member of the genus. Thus far the genus has not been separated into species but several strains of hantaviruses have been isolated: Hupei-I
from a patient with EHF in China; SR-11 from a laboratory rat associated with an EHF outbreak in Japan; Tchoupitoulas from a wild Norway rat in New Orleans, La.; Hallnas-I and Puumala from Clethrionomys glareolus trapped in areas where NE occurs in Sweden and Finland, respectively; and Prospect Hill-I from Microtus pennsylvanicus in Frederick, Md.; Girard Point from a Norway rat trapped near Philadelphia, Pa.; and others (Schmaljohn and Dalrymple, 1983; Schmaljohn et al., 1983, 1985; LeDuc et al., 1984; Childs et al., 1985; Goldgaber et al., 1985; LeDuc et al., 1985a).
In addition to Hantavirus, genera within the Bunyaviridae are Bunyavirus (prototype, Bunyamwera virus), Nairovirus (prototype, Crimean/Congo hemorrhagic fever virus), Phlebovirus (prototype, Sandfly fever virus), and Uukuvirus (prototype, Ukuniemi virus). The five genera are serologically, morphologically, and biochemically distinct. The Bunyaviridae are generally spherical, measure 80-120 nm in diameter, have surface projections 510 nm in length that are anchored in a lipid bilayered envelope, and have a genome of single-stranded RNA consisting of three segments (Martin et al., 1985).
Hantaan virus virions include round particles that measure 98 nm in diameter and oval particles that measure 110 x 173 nm in length. Negatively stained virions have an enveloped surface with a square grid-like pattern. The virus is stable at pH 7.0-9.0 but is inactivated at pH 5.0. It is relatively stable at 4-20°C but is inactivated rapidly at 37°C. It can be stored at -60°C (H. W. Lee, 1982; Martin et al., 1985). It can be grown in A-549 and Vero E6 cells (H. W. Lee, 1982; Schmaljohn et al., 1983).
The natural hosts of all hantaviruses appear to be small mammals, primarily rodents. Multiple species may serve as hosts in a given geographical area, and the strain of virus and likelihood of causing disease in man vary from region to region. Thus far, there are about five dominant virus-rodent-human disease (if known) associations: (i) Hantaan virus—the field mouse Apodemus agrarius—KHF in Korea (H. W. Lee, 1982; H. W. Lee et al., 1978) and the severe form of EHF in China (Song et al., 1983); (ii) Puumala virus—the bank vole Clethrionomys glareolus—NE in Eastern Europe and Scandinavia (Lahdevirta, 1971; Chumakov et al., 1981; Lahdevirta et al., 1984; Yanagihara et al., 1985a); (iii) Urban and laboratory rat viruses—Rattus norvegicus—moderate disease in people, mostly in Asia but occasionally in Europe (Umenai et al., 1979; Van Ypersele de Strihou, 1979; H. W. Lee and Johnson, 1982; H. W. Lee et al., 1982; Desmyter et al., 1983; Kitamura et al., 1983; Dournan et al., 1984; Lloyd et al., 1984; Kawamata et al. 1987); (iv) Girard Point and other viruses from North and South America—Rattus norvegicus—no disease recognized in humans although serological evidence of infection has been found (LeDuc et al.,
1984, 1985a; Childs et al., 1985, 1988; Tsai, 1985; Yanagihara et al., 1985b); and (v) Prospect Hill virus—the meadow vole Microtus pennsylvanicus—no disease recognized in people (LeDuc et al., 1984).
Hantaviruses are transmitted to humans from persistently infected rodents and other small mammals. In laboratory settings, this has usually been from laboratory rats or their tumors to people. The major mode of transmission is respiratory infection by aerosols of urine, feces, and saliva containing infectious virus. Animal contact is not necessary as many visitors to infected animal facilities have contracted the infection. Animal bites can transmit the infection but appear to be of relatively minor importance (LeDuc, 1987; Tsai, 1987a).
Reservoir hosts of Hantaan virus show no clinical signs, but the virus appears in their lungs about ten days after infection and subsequently appears in the urine and saliva. Peak virus shedding occurs about three weeks after infection, but virus can be detected in the lungs for six months and occasionally for up to two years. Aerosols are the main method of transmission (H. W. Lee, 1982; W.H.O., 1983; Morita et al., 1985). The epizootiology of other Hantavirus infections is poorly understood.
Although hantaviruses appear to be worldwide in distribution, it should be emphasized that human disease due to these agents has been reported only in Europe and Asia. Serologic surveys of rats and other small mammals have given evidence of Hantavirus infections in many areas of the world where disease due to hantaviruses is not known to occur, including North and South America (P. W. Lee et al., 1981b; LeDuc et al., 1984, 1985a; Childs et al., 1985, 1987a,b; Tsai, 1985).
Reservoir hosts remain asymptomatic during infection (H. W. Lee, 1982; W.H.O., 1983).
The clinical severity of Hantavirus infections in people varies according to geographic distribution of virus strains and, possibly, to other factors. KHF caused by Hantaan virus ranges from severe to mild. The incubation period is usually 2-3 weeks. Signs include fever, headache, muscular pains, hemorrhages (cutaneous petechiae or ecchymoses, hemoptysis, hematuria, hematemesis, melena), and proteinuria (Earle, 1954; Giles et al., 1954; Sheedy et al., 1954). About 20% of patients develop shock, severe hemorrhages, and renal failure (the "hemorrhagic fever with renal syndrome"). Mortality can be 5-10% (H. W. Lee, 1982; W.H.O., 1983).
Cases of NE in human patients in Scandinavia tend to be relatively mild,
have few hemorrhagic features, and show a mortality of less than 0.5%. There is acute onset of fever, headache, nausea, and vomiting followed in 3-6 days by proteinuria, oliguria, hematuria, and thrombocytopenia. Oliguria persists only a few days and is followed by polyuria. Most patients recover within three weeks (W.H.O., 1983; Lahdevirta et al., 1984).
Lesions have not been observed in reservoir hosts infected with hantaviruses. Lungs and other tissues may contain large amounts of virus without morphologic lesions (H. W. Lee et al., 1981; P. W. Lee et al., 1981a; H. W. Lee, 1982; Kurata et al., 1983; LeDuc et al., 1984).
Characteristic lesions in human patients that die of KHF are retroperitoneal edema; diffuse myocardial hemorrhage in the right atrium of the heart; severe congestion, hemorrhage, and necrosis in the renal medulla; and hemorrhage and necrosis in the anterior lobe of the pituitary gland (Lukes, 1954).
Mice and rats have been used for studies of experimental Hantaan virus infections. This virus is fatal for newborn mice (P. W. Lee et al., 1981a; Kurata et al., 1983; Yamanouchi et al., 1984; Nakamura et al., 1985).
Laboratory rats and biological materials such as transplantable tumors from laboratory rats constitute the greatest concern, particularly those from Eurasian sources. Since Hantavirus infections in rodents are inapparent, the diagnosis is most likely to be made through health surveillance. The recommended serological tests for this purpose are indirect fluorescent antibody test, enzyme-linked immunosorbent assay, and hemagglutination inhibition. Non-infectious antigen should be used and care should be taken to work with animals and blood products in a biological safety cabinet. P3 conditions are needed for working with unconcentrated virus in small amounts (Johnson, 1988).
For testing transplantable tumors and other biologic materials for Hantavirus contamination, the rat antibody production (RAP) test is recommended. Use 6-week-old pathogen-free rats; they should be prebled (to obtain control serum) and blood for serological testing should be collected 6 weeks after inoculation. Again, it is necessary that P3 containment conditions be maintained because the virus is excreted in urine and feces (Johnson, 1988).
Rats of the LOU strain and the transplantable plasmacytomas from that strain, which originated from a colony in Belgium, have been found to be infected with a member of the genus Hantavirus, and a number of personnel
contracted this infection in Belgium and the United Kingdom (van Ypersele de Strihou, 1979; Desmyter et al., 1983; Lloyd et al., 1984). Therefore, rats and tumors from these lines should be screened before use. A large number of rat cell lines maintained by the American Type Culture Collection in Rockville, Md., have been tested and found to be free of hantaviruses (LeDuc et al., 1985b). Also, LOU rats obtained by the National Institutes of Health from the University of Louvain in Belguim have been tested and found free of these viruses (Johnson, 1986).
The hantaviruses can be isolated and propagated in A-549 and Vero E6 cells (French et al., 1981; H. W. Lee, 1982; Kitamura et al., 1983; Schmaljohn et al., 1983; Yamanishi et al., 1983). Again, P3 conditions should be maintained.
The best approach is to prevent Hantavirus infections by obtaining only animals, transplantable tumors, and other biologic materials that have been tested and found to be free of the infection. Also, contamination of rodent stocks by wild rodents must be prevented.
Laboratory rodent stocks found to be infected with a Hantavirus should be destroyed and replaced with pathogen-free stock (W.H.O., 1983; HFRS, 1988). Although not proven to be effective, cesarean derivation has been recommended for eliminating the infection in valuable genetic stocks (Johnson, 1986; HFRS, 1988).
Interference with Research
The hantaviruses are important zoonoses (Johnson, 1986; Tsai, 1987b).