Материал: Bovine Viral Diarrhea Virus Diagnosis, Management, and Control

virus and cytopathic virus isolated early in infection but not to the persisting noncytopathic virus nor to the cytopathic virus recovered from the carcasses at necropsy. These results again suggested that mutational changes or selection of antigenic variants in the challenge virus occurred, which resulted in a homologous cytopathic virus responsible for precipitating MD. However, subsequent studies would show that the arising homologous cytopathic virus was often the result of RNA recombination between the cytopathic challenge virus and the persisting noncytopathic virus.

ACUTE AND PERSISTENT INFECTIONS IN THE

BULL

Whitmore et al. (1978) examined BVDV shedding in the semen of nine bulls after acute infection and found detectable virus in 4 of 70 semen samples. They were also able to recover virus from the testicle of one bull. The positive samples were obtained in the first 10 days postinoculation from four of the nine bulls but the amount of virus in the semen was not quantified. Treatment of the bulls with dexamethasone at 28 or 56 days postinoculation did not result in additional virus shedding in semen. More recently, Kirkland et al. (1991) studied five bulls during an acute infection with BVDV. Semen samples were collected between 7 and 14 days after infection on four occasions from each bull. Virus was isolated from three bulls and from 9 of 12 batches of semen from these bulls. Titers of virus were low ranging from 5–75 TCID50/ml of semen. They also noted that acute infection did not appear to affect the quality of semen.

Subsequent to Coria and McClurkin’s (1978) report, discussed earlier, Barlow et al. (1986) reported on virus shedding in a persistently infected bull. Virus was shed at high titer (104.0–105.5 TCID50/0.2 ml) in the semen. This animal had a sight defect, which was discovered postmortem to be due to retinal atrophy, but otherwise the bull appeared normal. Upon examination, the semen had been of acceptable quality and the bull had successfully sired a calf. Thus, had the bull not had a vision defect, the risk it posed of transmitting BVDV might have gone undetected. Revell et al. (1988) reported on two bulls that were persistently infected. In contrast to the findings of Barlow et al. (1986), the quality of the semen of these bulls was consistently poor, as measured by density and motility. Gross abnormalities of sperm heads (“collapsed heads”) was seen in 28–45% of spermatozoa from one of the bulls (Cy 105). Paton et al. (1990) inseminated six seronegative and six

preimmunized, seropositive heifers with semen from this bull (Cy 105). Both groups had poor rates of conception but eventually all but one heifer conceived after repeated inseminations. Eleven apparently normal calves were born and none were persistently infected. Paton et al. (1989) also used virus isolated from the serum of Cy 105 to infect four Freisen bulls. Virus was isolated from the semen of one of the bulls from four collections taken 7–14 days postinoculation. The highest titer obtained was 101.4 TCID50/ml. Semen quality of the bull was found to deteriorate following infection and showed a reduction in both sperm density and motility.

Meyling and Jensen (1988) were the first to report the production of a persistently infected calf sired from the semen of a persistently infected bull. Twelve seronegative heifers were inseminated with semen from this bull containing 105–107.5 TCID50/ ml of BVDV. All 12 heifers became infected as indicated by seroconversion and all heifers gave birth to clinically normal calves. Only1 of the 12 calves was persistently infected. Later, Kirkland et al. (1994) described the results of widespread field use of semen from a persistently infected bull. Approximately 600 doses of semen had been distributed to 97 dairy farms for sire evaluation purposes and 162 cows were inseminated. The first service conception rate was only 38%. A subsequent study of 61 calves sired by the bull revealed that only two of these calves were persistently infected. From these studies, it is apparent that production of persistently infected offspring via BVDV-contaminated semen is an uncommon event.

RESPIRATORY DISEASE

During the 1980s, research into the contribution of BVDV to respiratory disease was initiated. This research followed the successful reproduction of “shipping fever” pneumonia in cattle with aerosals of infectious bovine rhinotracheitis virus and Mannheimia haemolytica in the late 1970s (Jericho and Langford, 1978). Shipping fever pneumonia, a cause of significant mortality in cattle in feedlots, was thought to be precipitated by the stress of transporting cattle from the farm to the feedlot. The pathogenesis of the disease is thought to involve dual infection of a pneumotropic virus and a colonizing bacterial species, most often M. haemolytica (Yates, 1982). As early as the first description of BVDV herd outbreaks in 1946, the virus has been implicated in causing at least mild respiratory disease with symptoms of nasal discharge and coughing. Potgieter et al. (1984, 1985) performed several

experiments examining the ability of BVDV to induce respiratory disease. BVDV infection of calves without subsequent bacterial superinfection resulted in mild respiratory tract lesions characterized by small, scattered areas of interstitial pneumonia involving 2–7% of the total lung volume. Infection of calves with M. haemolytica alone produced localized lesions involving about 15% of the lung. In contrast to the mild disease produced by these agents when given individually, inoculation of BVDV followed by M. haemolytica produced a severe fibrinopurulent bronchopneumonia and pleuritis involving 40–75% of the lung volume. Potgeiter (1985) also found that BVDV strains may differ in their pneumopathogenicity. That BVDV infection can precipitate severe respiratory disease is also supported by the observation that pneumonia, including shipping fever–like fibrinous pneumonia, was a common finding in the severe type 2 BVDV outbreaks that occurred in Ontario, Canada, in the mid1990s (Carman et al., 1998, van Dreumel, 2002).

THROMBOCYTOPENIA

Thrombocytopenia with hemorrhage associated with BVDV virus infection was first reported in 1987 within a summary of case reports for dairy herds in the northeastern United States (Perdrizet et al., 1987). In one report from 1985, 6 of 30 first-calf heifers developed a high fever, had bloody and mucoid diarrhea, and died within 2 weeks. Rebhun et al. (1989) reviewed case records of cattle admitted to the College of Veterinary Medicine at Cornell University for the years 1977–1987 and found that thrombocytopenia was reported in about 10% of clinically acute BVDV infections in adult cattle. Clinical signs included hemorrhages, red or orange bloody diarrhea, epistaxis, and abnormal bleeding from injection sites. Hemorrhages associated with BVDV infection in young veal calves were also observed with increasing frequency in the late 1980s in the northeastern United States (Corapi et al., 1990b).

Corapi et al. (1989) experimentally reproduced thrombocytopenia in young calves with a BVDV isolate (CD-87) recovered from a severe outbreak. This outbreak involved 50% of a milking herd of 100 holsteins in New York in which 20 animals died. Of the eight calves inoculated with the CD-87 isolate, three developed severe thrombocytopenia ( 5,000 platelets/µl). Two of these three calves developed hemorrhages when their platelet counts fell to 2,000/µl of blood or less for a period of 24 hours or longer. Hemorrhages were observed on the sclera of the eyes, inner surface of the eyelids, mucosal

surfaces of the cheeks, lower gingiva, tongue, and soft palate. Both calves had prolonged bleeding from venepuncture sites. Internal hemorrhages were also observed on the surfaces of various organs of one calf at necropsy. In a later report, Corapi et al. (1990b) inoculated 8 veal calves with the CD-87 isolate and 10 veal calves with CD-89, a BVDV isolate recovered from a veal calf on a farm in Pennsylvania where hemorrhages were noticed in several calves. Ten in-contact calves were included in the experiment. During the experiment, virus was isolated from all calves including in-contact calves. Severe thrombocytopenia was observed in 12 calves, and 11 of these developed hemorrhages. Calves that had a pre-exposure virus-neutralizing antibody titer of >1:32 to the Singer strain of BVDV did not develop severe thrombocytopenia. Five calves died during the course of the experiment, four of which exhibited hemorrhages, and the others recovered. Hemorrhages were observed in these experiments when platelet counts decreased below 5,000 platelets/µl. Although the majority of severely thrombocytopenic calves recovered, there was no way to determine beforehand the calf’s fate since those that died often appeared to be in relatively good physical condition just hours prior to death (Corapi et al., 1990b). A few years later, it was discovered that hemorrhagic syndrome, as the disease came to be known, was caused by a new type of BVDV, genetically distinct from the classical viruses used in BVD vaccines.

ADVANCES IN MOLECULAR BIOLOGY

In the late 1980s, significant advances were made in the molecular biology of BVDV with the first genomic sequencing of BVDV strains, the finding of a marker protein for cytopathic BVDV, and the first evidence of RNA recombination in cytopathic strains of BVDV. Monoclonal antibodies to the virus were also first produced in the late 1980s as reagents for protein studies and diagnostic test methods (see the following section, “Diagnosis”).

The first BVDV strains to be sequenced were two cytopathic strains, the North American NADL strain and the European Osloss strain (Collett et al., 1988b; Renard et al., 1987). The NADL sequence showed that the RNA genome of BVDV has one long open reading frame (ORF). Thus, proteins are produced by cotranslational and posttranslational processing of a polyprotein (Collett et al., 1988a). The original Osloss sequence was at first shown to contain two ORFs but was later corrected to consist also of a single ORF (de Moerlooze et al., 1993).

The elucidation of the genomic organization of BVDV and the pestivirus genus led to their taxonomic reclassification from the Togaviridae to the

Flaviviridae.

At about the same time, protein studies revealed that the two biotypes of the virus, cytopathic and noncytopathic, could be distinguished at a molecular level. It was found that cytopathic strains of BVDV produce in infected cells one additional nonstructural protein not observed in cells infected with noncytopathic BVDV (Donis and Dubovi, 1987; Pocock et al., 1987). The protein, which is a marker protein of cytopathic BVDV, is designated p80 or NS3. This protein is actually a smaller version of a larger nonstructural protein, p125 or NS2- 3, that is present in all BVDV-infected cells. Several mechanisms leading to expression of NS3 were later found to occur: The NS3 protein can be generated by proteolytic cleavage of NS2-3 or generated as the result of genetic duplications or deletions in the genomes of some cytopathic BVDVs (Meyers and Thiel, 1996).

Meyers et al. (1989) compared the BVDV Osloss and NADL genomic sequences with that of the pestivirus CSFV and found insertions in the NS2 gene of these cytopathic strains of BVDV. The Osloss insertion was 228 nucleotides long whereas the insertion of NADL was 270 nucleotides. The insertions were found to be different and that of NADL was unidentified. However, remarkably, the Osloss insertion was found to be derived from the cellular gene coding for ubiquitin. These findings led to a molecular model for pathogenesis of MD: In persistently infected animals, the noncytopathic virus mutates to a cytopathic virus by the incorporation of cellular sequences during a recombination event. Subsequently, it was found that genetic recombination could also occur between BVDV genomes as evidenced by insertions of viral sequences in the genomes of some cytopathic viruses.

DIAGNOSIS

Meyling (1984) described a microisolation test for detection of BVDV in serum samples that used immunoperoxidase staining of viral antigens rather than immunofluorescence. The immunoperoxidase monolayer assay (IPMA) is still commonly used today because of its ease and the ability to test many serum samples for BVDV at a time. Prior to the development of the assay for serum, BielefeldtOhmann (1983) had utilized the immunoperoxidase staining technique for the detection of BVDV in tissues of infected animals.

Panels of monoclonal antibodies were produced to both BVDV and CSFV in the late 1980s and early 1990s by several research groups. The specificity of these monoclonal antibodies was usually to the E2 (gp53) protein, which was determined to be the major neutralizing envelope protein of the virus (Donis et al., 1988), or to the nonstructural NS3 (p80) protein or less frequently, to the Erns (gp48) protein, a second envelope protein, which was discovered later to have RNase activity (Schneider et al., 1993). Edwards et al. (1988) examined a total of 38 monoclonal antibodies, of which 26 were against BVDV antigens and 12 were against CSFV antigens. The monoclonal antibodies could be divided into three panels: those that were pan-pestivirus specific; those that were CSFV-specific; and those that were selectively reactive with ruminant pestiviruses. Monoclonal antibodies to the NS3 protein tend to be cross-reactive with all pestiviruses, because the amino acid sequence of this protein is highly conserved among pestiviruses. Monoclonal antibodies to the envelope proteins E2 and Erns tend to be specific for the viral species (i.e., BVDV or CSFV) used as the immunogen in their production (Edwards et al., 1991). Corapi et al. (1990a) examined the cross-reactivity of a panel of BVDV monoclonal antibodies to 70 BVDV isolates. They found that 12 of 13 of their NS3 monoclonal antibodies reacted with 100% of BVDV isolates tested, whereas the reactivity of E2 monoclonal antibodies varied from 6–98%. Of two Erns monoclonal antibodies, one was reactive to 100% of BVDV isolates tested, whereas the other was less cross-reactive (57%). For the 70 BVDV isolates, a total of 32 distinct patterns of monoclonal antibody reactivity were observed. This demonstrated that considerable antigenic diversity exists among BVDV isolates.

CONTROL

To gain insight into the carrier state and for control strategies for BVDV, it was important to determine the prevalence of persistently infected cattle in entire populations. Meyling (1984), used the immunoperoxidase monolayer assay and found that approximately 1% of slaughter cattle in Denmark were viremic and apparently persistently infected. Subsequently, Bolin et al. (1985a) made the first attempt to determine the prevalence of persistently infected cattle in U.S. herds. The prevalence of persistent infection in a nonrandom population of 66 herds was 1.7%. Since 50% of these herds were chosen because of a history of BVDV infection, it was noted that the prevalence figure for the entire U.S. cattle

population was probably somewhat lower. However, the figure of 1.7% turned out to be similar to the prevalence of persistently infected cattle of approximately 0.5% to 2.0% obtained in later surveys conducted in different countries in the 1980s and 1990s (Houe, 1999). In their study on the prevalence of persistently infected cattle, Bolin et al. (1985a) also titrated the virus from the serum of persistently infected animals and found that these animals usually had high BVDV titers (104–105 TCID50/ml).

At about the same time as the study of McClurkin et al. (1984) on the production of immune-tolerant, persistently infected calves, Liess et al. (1984) produced congenital malformations (cerebellar hypoplasia, hydrocephalus) and persistent infections in calves by inoculation of pregnant cattle with a modified-live vaccine and demonstrated the risk of using these vaccines in breeding animals. In both studies, persistently infected calves could be produced by inoculation of pregnant cows to about 120 days of gestation, after which BVDV infection generated a fetal immune response.

By the late 1980s, modified-live vaccines, a temperature-sensitive mutant virus vaccine (Lobman et al., 1984), and killed-virus vaccines were available for use. Two elements of control were considered essential: the detection and elimination of persistently infected carriers and immunization of breeding animals before their first conception (Radostits and Littlejohns, 1988). Immunization of calves was now considered to be less important by some. Previously, before the pathogenesis of MD was understood, it was thought by some that immunization of young cattle might prevent MD from occurring because the disease usually occurs in cattle from 6–24 months of age. Radostits and Littlejohns (1988) commented that pvMD gave vaccines a poor reputation and as a result they had not been used on a regular basis. There was also a concern that modi- fied-live BVDV vaccines might cause immunosuppression and increase the risk of mortality in feedlot cattle (Martin et al., 1980; 1981).

Radostits and Littlejohns (1988) stated that there was no substantial evidence to warrant the vaccination of feedlot cattle. They further suggested that, if vaccination of the dam prior to conception is a part of the control program, vaccination of calves may be unnecessary until they approach breeding age. Thus, there was a de-emphasis of vaccination for some groups of cattle (calves and feedlot animals), at least in some circles. Baker (1987), cited studies showing an association between BVDV vaccination and increased risk of mortality in feedlot animals and a re-

port on the immunosuppressive properties of a modified-live (Singer strain) vaccine (Roth and Kaeberle, 1983) and believed that it would be advantageous to vaccinate calves in a preconditioning program before their arrival at a feedlot. If calves are vaccinated on arrival at feedlots, he believed they should be vaccinated with a killed-virus vaccine.

Bolin (1990) also stated the concerns regarding the use of modified-live vaccines: immunosuppression, the potential to produce pvMD, and the potential to adversely affect the fetus. During the late 1980s, there was also a growing awareness of the antigenic diversity among BVDV isolates. Thus, another concern for vaccines in general was their efficacy in protecting against fetal infections with antigenically variable field viruses. Bolin (1990) recommended using modified-live vaccines in large grazing herds or when handling facilities for cattle were poor, since only a single dose of vaccine is required for immunization. Modified-live vaccines were contraindicated for use in pregnant cattle and animals in contact with pregnant cattle. Instead, Bolin (1990) recommended the use of killed-virus vaccines in dairy herds where pregnant cattle are always present. Killed-virus vaccines were also recommended for bulls in semen collection centers.

Although the identification and elimination of persistently infected cattle was considered an essential element of control, it was also considered costly because of the amount of testing involved. Baker (1987) recommended testing for virus or viral antigen and determining antibody status, but stated that screening of whole herds may not be economically feasible in all situations. He suggested an alternate approach to reduce the cost of testing. Persistently infected animals are viremic, usually antibodynegative, and often exhibit a poor response to vaccination. Hence, one possibility is to vaccinate all cattle greater than 6 months of age with a killed-virus vaccine followed by a booster, and then determine their antibody titers. Cattle that remained antibodynegative or respond poorly with low levels of antibody would be suspected to be persistently infected and could be tested for virus or viral antigen. This was reminiscent of the observations of McClurkin et al. (1979) that cattle that remain seronegative while in contact with seropositive cattle are likely to be immune-tolerant and persistently infected, and these cattle had low antibody titers after vaccination. However, because the immune tolerance is specific for the persisting BVDV in these animals, screening by this method may not be highly reliable, particularly when the vaccine is antigenically very different

from the virus persisting in the herd. In the latter case, higher than expected levels of antibodies might be produced by persistently infected animals and obscure differences in antibody levels between persistently infected and normal cattle.

Baker (1987) noted several important factors that should be considered in herd screening, including that persistently infected calves may be seropositive because of colostral antibodies, and that passive immunity may interfere with virus isolation. Furthermore, any calves born in the 9 months after herd testing potentially may be persistently infected. He also noted that retesting along family lines may be worthwhile because of the possibility of persistently infected families. After a herd is free of persistently infected animals, he recommended isolating and testing all new additions to the herd and that any additions that are pregnant should also have their calves tested for virus at birth.

THE DECADE OF THE 1990s

THE EMERGENCE OF TYPE 2 BVDV

In the early and mid-1990s, in addition to outbreaks of hemorrhagic syndrome (severe thrombocytopenia with hemorrhages) (Ridpath et al., 1994), outbreaks of acute, severe BVD in which hemorrhagic syndrome was either inapparent or was not a prominent clinical feature, occurred in Canada and the United States (Pellerin et al., 1994; Carman et al., 1998; Sockett et al., 1996; Drake et al., 1996). In Canada, these BVD outbreaks were most damaging to the cattle industry in the provinces of Quebec and Ontario. In 1993, Quebec lost approximately 25% of its veal crop (overall mortality: 32,000 out of 143,000 calves) to these BVDV outbreaks (Pellerin et al., 1994). Although the Quebec outbreaks were not described in detail from a clinical standpoint, Pellerin et al. (1994) stated that herds of calves that looked healthy one day could suffer a 10% death loss the next day, and it was not a rarity to have a mortality rate of 100%. Fever, pneumonia, diarrhea, and sudden death occurred in all age groups and abortions were frequent in the Ontario outbreaks of 1993 to 1995 (Carman et al., 1998). The disease often resembled MD. However, severe acute BVD could be distinguished from MD because only noncytopathic BVDV was isolated from animals suffering severe acute BVD.

Pellerin et al. (1994) and Ridpath et al. (1994) determined that the BVDV isolates causing hemorrhagic syndrome and acute severe BVD formed a new genetic group (genotype) of BVDV distinct

from early strains such as the Oregon C24V, NADL, and Singer strains utilized in vaccines. The new group was designated type 2 (BVDV 2) and the group comprising the early strains as type 1 (BVDV 1). Pellerin et al. (1994) further subdivided BVDV 1 into two subgroups: 1a, comprising such strains as NADL, Oregon, and Singer; and 1b, which included NY-1 and Osloss strains.

In the Ontario BVDV 2 outbreaks, the initial clinical complaint was frequently of respiratory disease in calves or adults (Carman et al., 1998). Diarrhea and abortion were also listed as initial clinical signs. Postmortem lesions were generally those described for MD: gastrointestinal erosions and ulcers. Pneumonia was the most common concurrent diagnosis and was observed in all age groups.

Retrospective typing of Ontario isolates recovered from 1981–1994 proved that as early as 1981, BVDV 2 was already present in Ontario, Canada (Carman et al., 1998). Since outbreaks of severe acute BVD did not occur until much later (1993), this suggests that either earlier circulating strains of BVDV 2 were not highly virulent and that some strains had subsequently acquired virulence determinants or, alternatively, virulent strains of BVDV 2 existed early but were harbored in seropositive herds and only later caused outbreaks of severe disease when naive populations became exposed and infected.

PHYLOGENETIC STUDIES

Prior to the decade of the 1990s, BVDV, CSFV, and BDV were the three recognized pestiviruses, although it was uncertain whether BDV represented a unique viral species or whether border disease in sheep was caused by BVDV. During the 1990s, in addition to the segregation of BVDV into two genotypes, genetic characterization of isolates from sheep showed that sheep could be infected by both genotypes of BVDV and by a unique pestivirus that was referred to as the “true” border disease virus (Becher et al., 1995). The first “BDV” isolate sequenced was actually a BVDV 2 strain (Sullivan et al., 1994; Becher et al., 1995). It was also demonstrated that wildlife could be infected with pestiviruses. Genetic analysis of a pestivirus isolated from a giraffe proved this virus to be a unique genotype, whereas three deer isolates were found to belong to the BVDV 1 genotype (van Rijn et al., 1997; Becher et al., 1997).

Baule et al. (1997) reported that southern African isolates consisted of four BVDV 1 subtypes, including the subtypes 1a and 1b previously described