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

clease digestion of the PCR product for typing. Ridpath et al. (1994) utilized the specific amplification of BVDV 2 for typing, with a negative result indicating BVDV 1. Subsequently, these researchers made improvements to their PCR assays, including a positive test for BVDV 1 and specific primers for the differentiation of BVDV 1 subtypes, 1a and 1b (Ridpath and Bolin, 1998). The earlier PCR assays were followed by two multiplex PCR assays in which specific primers for BVDV 1 and BVDV 2 are used in a nested PCR format in the second round of amplification (Sullivan and Akkina, 1995; Gilbert et al., 1999). The advantage of the multiplex assay is that a specific product of differing size is produced for each genotype and differentiation occurs in a single assay. The assay of Sullivan and Akkina (1995) could type BDV as well as BVDV 1 and BVDV 2 since BDV-specific primers were included in the assay. This is seen as an advantage for typing pestiviruses from sheep, which can be infected by all three viruses, but in cattle BDV does not appear to be readily infectious. BDV has not been isolated from North American or European cattle (Paton et al., 1996; Ridpath, 1996) and only one bovine BDV isolate has ever been reported; its original isolation was thought to have been made in the 1960s (Becher et al., 1997). The PCR assay of Gilbert et al. (1999), further described in Deregt et al. (2002) could be used with conventional RNA extraction, or directly without RNA extraction, by adding the sample to the RT-PCR mixture.

Radwan et al. (1995) developed a PCR assay for testing bulk milk samples to identify dairy herds infected with BVDV. For validation of the assay, these researchers first determined the BVDV titers in milk from an experimentally (acutely) infected cow and two persistently infected cows. Virus titers of 102.5, 106.5, and 105.5 TCID50/ml were present in milk from the acutely infected cow and persistently infected cows, respectively. The virus titers in the milk of persistently infected cows were higher than those in their sera (approximately 104.5 TCID50/ml). The PCR assay was found to be about 15 times more sensitive than virus isolation in detecting BVDV in milk somatic cells. By testing bulk milk samples, BVDV was detected in 33 of 136 dairy herds by the PCR assay. In sharp contrast, virus isolation did not detect BVDV in any of the bulk milk samples. It was suggested that BVDV antibodies in the milk were responsible for the poor virus isolation results.

Drew et al. (1999) also applied PCR to bulk milk samples to identify BVDV infection among lactating cows, but did not attempt virus isolation from

these samples. Their assay was able to detect BVDV shed by 1 persistently infected cow in a herd of 162 lactating animals. Later, Renshaw et al. (2000) used a PCR assay, which employed the primers used in the assay of Radwan et al. (1995), to detect BVDV in bulk milk samples and also attempted virus isolation from these samples. They found that BVDV could be detected by both methods when milk from a single persistently infected animal was diluted 1:600 with milk from a herd of BVDV-negative animals. Of 144 bulk milk samples from 97 farms, 24 were BVDV-positive by either PCR or virus isolation: 20 were positive by PCR and 17 were positive by virus isolation. Renshaw et al. (2000) suggested that the poor virus isolation rate of Radwan et al. (1995) may have resulted from using milk somatic cells that were frozen before virus isolation was attempted, and that successful virus isolation could be obtained when freshly prepared (not frozen) milk somatic cells are utilized. Furthermore, they recommended simultaneous PCR and virus isolation testing of bulk milk samples to ensure detecting BVDV that may either not be amplified by primers currently in use or which for unknown reasons may be difficult to isolate.

McGoldrick et al. (1999) designed a real-time (Taqman) nested PCR assay for CSFV, which could be modified with different fluorescent probes to allow specific detection of BVDV 2 or BDV. Another fluorescent probe allowed detection of all pestiviruses except for some isolates of BVDV 2. The nested PCR assay was performed in a closed, single tube in which reagents for the second round of PCR were maintained in the inner lid. After reverse transcription and a primary round of PCR, the tubes were inverted to mix the second round reagents in the lid with the first round products for initiation of the second round of PCR. Later, Mahlum et al. (2002) designed a real-time (Taqman) PCR assay specifically for BVDV. The assay was found to be more sensitive than virus isolation and IPMA in detecting BVDV in sera, and more sensitive than virus isolation or immunohistochemistry in detecting BVDV in tissues.

CONTROL BY VACCINATION

Prior to the emergence of virulent BVDV 2 strains, most BVDV infections of nonbreeding animals were thought to be benign and emphasis was placed on the prevention of fetal infections. Thus, it was recommended that all breeding females be vaccinated prior to conception (Baker, 1987; Radostits and Littlejohns 1988). Other animals in the herd were

often not vaccinated. In the 1990s, with the emergence of virulent BVDV 2, the failure to vaccinate all animals in a herd proved to be disastrous, with losses of up to $40,000 to $100,000/herd when outbreaks due to these virulent viruses occurred (Carman et al., 1998). Fortunately, the vaccines at that time, which contained only BVDV 1, did provide a measure of protection against acute infections with BVDV 2 and outbreaks could be controlled (Carman et al., 1998).

Although immunization with some vaccines containing BVDV 1 alone were protective against acute BVD caused by BVDV 2 (Cortese et al., 1998, Carman et al., 1998), apparent vaccine breaks did occur (Ridpath et al., 1994) and incorporation of BVDV 2 in vaccines was considered a priority. In addition to the prevention of acute disease by BVDV 2, a continuing concern was the degree of fetal protection against BVDV provided by vaccines in general. With the emergence of BVDV 2, this concern escalated with reports that vaccines containing BVDV 1 appeared not to be protective against fetal infection with BVDV 2 (van Campen et al., 2000).

van Oirschot et al. (1999) reevaluated the results of fetal protection experiments that were conducted in the 1970s to 1990s by six different research groups. They concluded that no vaccine had shown full fetal protection and that protection varied from 33–86%. In only one study were none of the fetuses from the vaccinated cows infected after challenge; however, one fetus from six unvaccinated control cows also remained uninfected (Brownlie et al., 1995). In this type of experiment it is important that all fetuses in the control group become infected to demonstrate the ability of the challenge virus to infect the fetus under the defined experimental conditions. Thus, in the latter study the degree of protection conferred by the vaccine was calculated to be 86% because only five of six fetuses from the control group became infected rather than 100% of them.

CONTROL WITHOUT VACCINATION

In the 1990s, eradication programs for BVDV on a national level were implemented in Sweden, Norway, Finland, and Denmark (Bitsch and Ronshølt, 1995). These programs were conducted without vaccination. In Sweden, the national program began in 1993 as a voluntary program entirely financed by producers (Alenius et al., 1996). In Denmark, the BVDV eradication program was initiated by dairy farmers in 1994 followed by a government order to support the program in 1996 (Houe, 1996). Both the

Swedish and Danish programs involve classification of herds by their BVDV status, removal of persistently infected cattle from infected herds, monitoring herd status, and prevention of infection in BVDVfree herds. Since these countries do not vaccinate, antibody-positive animals act as indicators of infection. Thus, bulk milk testing for antibodies constitutes an important component of these programs and is used to classify and monitor herds as to their BVD status. In Sweden, bulk milk testing is done with an indirect ELISA and in Denmark with a blocking ELISA (Alenius et al., 1996; Houe, 1996). These programs significantly reduced the prevalence of BVDV-positive dairy herds after only a few years. For Sweden, the prevalence dropped from 51% in 1993 to 24% in 1995; for Norway, from 23% in 1993 to 14.4% in 1996; and for Denmark, from 39% in 1994 to 9% in 1999 (Alenius et al., 1997; Bitsch and Ronshølt, 1995; Bitsch et al., 2000; Waage et al., 1997). Finland, which began its program in 1994 with a very low prevalence (only 1%) in dairy herds, further reduced the prevalence to just 0.4% in 1997 (Nuotio et al., 1999).

Bitsch et al. (2000) reported that for the Danish program, legislation was needed in 1996 to ensure that no persistently infected cattle were allowed on common pastures. Movement of cattle to other herds or common pastures was allowed only after blood testing and certification that the animals were not persistently infected. At first, pregnant (non-persist- ently infected) cows were not controlled. Instead, all buyers of pregnant animals were advised to isolate and ensure a non-persistently infected status of their calves before introducing them into a new herd. Since not all farmers heeded these recommendations and herds became infected because of persistently infected calves resulting from the purchase of pregnant animals, legislation was modified to restrict the movement of all female cattle over 1 year of age, specifically that no such cattle could be moved from non-free herds to other herds or common pastures.

2000 TO THE PRESENT

RESPIRATORY DISEASE

Respiratory diseases have plagued the feedlot industry for many years and are collectively considered to be the most significant cause of mortality for feedlot calves. Several recent studies have linked BVDV with bovine respiratory disease in feedlot cattle and another study has identified BVDV strains that can cause primary respiratory disease experimentally.

Martin et al. (1999) determined the antibody titers

to BVDV, infectious bovine rhinotracheitis, parain- fluenza-3, bovine respiratory syncytial virus, and two mycoplasma strains in 32 groups of calves prior to entry, and 4–5 weeks after their arrival, at the feedlot. They then looked for an association of these infectious agents (by increases in antibody titers) with the risk of respiratory disease. Of all the agents, BVDV had the most consistent association with elevated risk of respiratory disease and lower weight gains. In another study, Haines et al. (2001) examined 49 cases of antibiotic-unresponsive, chronic disease (most often respiratory disease and/or arthritis) in Canadian feedlot cattle. By immunohistochemistry, they found that Mycoplasma bovis and BVDV were the most common pathogens persisting in the tissues of these animals. BVDV was found in lung and/or joint tissues in 20 of the 49 (41%) cases.

Shahriar et al. (2002) reported on the prevalence of pathogens in cases of chronic, antibiotic-resistant pneumonia with or without concurrent polyarthritis occurring in feedlot cattle in western Canada. They examined retrospective (1995–1998) and current cases (1999) by immunochemistry of lung and heart tissue and found that Mycoplasma bovis was present in 44 of 48 cases of the retrospective group and 15 of 16 of current cases; whereas BVDV was present in 31 of 48 retrospective and 9 of 16 current cases. From four positive virus isolations in the current group, BVDV type 1b was isolated in two cases and BVDV 2 was isolated from the other two cases. These researchers suggested that a synergism between Mycoplasma bovis and BVDV might occur in this syndrome of pneumonia with concurrent arthritis.

Later, Fulton et al. (2002) examined the prevalence of BVDV in stocker calves with acute respiratory disease and reported that BVDV type 1b was the predominant type involved. These researchers also noted that vaccines in the U.S. primarily contained BVDV type 1a, that some vaccines also included BVDV 2, but that only one vaccine contained BVDV 1b. They suggested that, for effective vaccination against BVDV type 1b, there should be demonstrated efficacy of current BVDV type 1a vaccines against type 1b. Alternatively, new components of BVDV type 1b could be included in current BVDV type 1a vaccines.

Baule et al. (1997, 2001) identified two new subtypes of BVDV 1 isolated in southern Africa and designated these as type 1c and type 1d. Type 1d viruses were found to be predominantly associated with respiratory disease. Two cytopathic viruses of this type were inoculated in calves intranasally or

intravenously. All inoculated calves developed respiratory symptoms. One of the isolates produced mainly nasal discharge and fever in calves; however, the other isolate produced a more severe disease that included ocular discharge, nasal discharge, fever, coughing, abnormal breathing, and oral erosions. Transient diarrhea was observed in only 2 of 10 calves. There was a widespread distribution of virus in various tissues and organs of infected calves, including heart muscle, skin, bone marrow, and brain. However, lesions were mainly observed in the respiratory tract (focal catarrhal bronchopneumonia and atelectasis of the lung) and lymphoid tissues. In one calf, virus was still present in tissues 31 days after infection in the absence of viremia. Furthermore, two calves were still shedding cytopathic BVDV in nasal secretions at 21 and 31 days after infection when virus was no longer detectable in blood.

MOLECULAR ACTIONS OF CYTOPATHIC AND

NONCYTOPATHIC BVDV

In recent years, the actions of cytopathic and noncytopathic BVDV on cells have been studied more intensively. For cytopathic strains of BVDV, it has been found that, like many other “lytic” viruses, they kill cells by triggering apoptosis (programmed cell death) rather than lysis and necrosis (Zhang et al., 1996; Adler et al., 1997; Hoff and Donis, 1997). Apoptosis is responsible for the elimination of cells in normal developmental processes but can be triggered by many stimuli and is characterized by condensation of chromatin, cell shrinkage, generation of apoptotic bodies, and fragmentation of chromosomal DNA with the generation of typical oligonucleosomal fragments. The mechanism(s) by which cytopathic BVDV triggers apoptosis is currently an active area of research. BVDV is an ideal model system for viral-induced apoptosis because both cytopathic and noncytopathic forms of the virus exist. In addition to the production of NS3, which is correlated with cytopathic effect, it has recently been discovered that cytopathic viruses accumulate much higher levels of viral RNA in cells than do noncytopathic viruses.

Vassilev and Donis (2000) used an infectious cDNA clone of the NADL strain to create, by deletion of the cellular insertion in the NS2 gene, an isogenic, noncytopathic virus mutant. To study viral RNA accumulation in cells, they utilized both cytopathic and noncytopathic (mutant) forms of the NADL strain and several naturally occurring cytopathic and noncytopathic viral pairs. At a multiplicity of infection of one, viral RNA accumulation was

6.5 to 23 times higher when cells were infected with the cytopathic virus than with the corresponding noncytopathic virus. Thus, in addition to generation of NS3, a second factor—increased viral RNA concentration in cells—may contribute to the ability of cytopathic BVDV to induce the death of cells.

Studies spanning several decades have shown that noncytopathic strains of BVDV do not induce type 1 (alpha/beta) interferons in cells (in vitro) and block the induction of interferons by other viruses (Diderholm and Dinter 1966; Nakamura et al., 1995; Adler et al., 1997). In contrast, cytopathic strains of BVDV do induce type 1 interferons in cells in vitro (Adler et al., 1997). Recently, Schweizer and Peterhans (2001) showed that noncytopathic BVDV inhibits the induction of apoptosis and interferons by a synthetic double-stranded RNA (poly IC). Subsequent work indicates that noncytopathic BVDV blocks an interferon regulatory factor (Baigent et al., 2002). These studies explain the basis of two somewhat obscure diagnostic tests for noncytopathic BVDV. The first, described in 1968, was the END method (enhancement/exaltation of Newcastle disease virus) (Inaba et al., 1968). Newcastle disease virus (NDV) induces, and is sensitive to, interferon. Co-infection of noncytopathic BVDV and NDV leads to an enhancement of the replication of NDV. In the END method, the presence of noncytopathic BVDV is thus shown by NDV enhancement due to the suppression of interferon by noncytopathic BVDV. The second diagnostic test was based on the suppression of an effect of poly IC (Maisonnave and Rossi, 1982). Cells were inoculated with noncytopathic BVDV, and then treated with poly IC, and subsequently inoculated with vesicular stomatis virus (VSV). Cells infected with noncytopathic BVDV were not protected against cytopathic effect by VSV, whereas cells that were not infected with noncytopathic BVDV, were protected. It is now clear that the action of noncytopathic BVDV is via inhibition of interferon induction by poly IC.

The interference with interferon production in vitro by noncytopathic BVDV has been suggested as an enabling factor in the ability of these viruses to establish persistent infections in the early fetus (Schweizer and Peterhans, 2001). To determine whether interference occurs in vivo, Charleston et al. (2001) inoculated the amniotic fluid of approximately 60-day-old fetuses with noncytopathic or cytopathic BVDV. Whereas cytopathic BVDV induced interferon production in the fetus, noncytopathic BVDV failed to do so. In contrast, the dams of the fetuses inoculated with noncytopathic BVDV did

produce interferons, as did calves in a subsequent study (Charleston et al., 2002). Thus, there appears to be a marked difference in the interferon response between the early fetus and immune-competent animals upon infection with noncytopathic BVDV.

AN OLD VIRUS, ONCE LOST, TEACHES

LESSONS

Of the old cytopathic BVDV strains—those that have been used in laboratories for many years—the NADL strain that was isolated from a case of MD in 1962 (Gutekunst, 1968) is still one of the most widely used in research and diagnostics and as a component in vaccines. As for other old cytopathic strains, the noncytopathic counterpart of the cytopathic NADL strain had been lost. However, recently, this virus has made an unexpected appearance; it was found, through in vivo studies, to be still residing in a stock of cytopathic NADL virus originally obtained from the American Type Culture Collection (Harding et al., 2002). Pregnant cows (at 90–105 days of gestation) were inoculated with virus originating from an infectious cDNA clone of NADL (i-VVNADL), or the parental NADL virus stock (termed NADL-A), and 3 or 6 weeks after inoculation the fetuses were harvested. Virus isolated from cows inoculated with iVVNADL was always cytopathic in biotype and no virus was isolated from fetuses from these cows. Surprisingly, viruses of both biotypes were initially isolated from cows inoculated with NADL-A, but after 8 days postinoculation only noncytopathic virus was recovered. Virus was isolated from 8 of 10 fetuses from dams inoculated with NADL-A. In each case, the virus was noncytopathic.

The nucleotide sequence of the NS2-3 and NS5A regions of the genome of the contaminating noncytopathic BVDV (termed NADL-1102) in the NADL-A stock was determined and the virus was found to lack the 270 base pair cellular insertion in the NS2 gene of NADL-A. Aside from the lack of the cellular insertion, which had been shown previously to be responsible for cytopathogenicity (Mendez et al., 1998), a greater than 99% homology exists between sequences of NADL-1102 and published NADL strain sequences. This level of homology indicates that NADL-1102 is the ancestral noncytopathic NADL virus from which the cytopathic NADL virus arose by genetic recombination. The authors of this report state that there had been no systematic attempts to eliminate noncytopathic BVDV from this stock by the repository personnel or the depositors (Gutekunst and Malmquist, 1963).

This study illustrates well, and confirms the findings of others, that noncytopathic BVDV, but not cytopathic BVDV, can infect the fetus and establish persistent infections (McClurkin et al., 1984; Brownlie et al., 1989). Prior to 1984, it was not appreciated that both BVDV biotypes could be isolated from cases of MD. Thus, cytopathic BVDV isolates may not have always been purified or adequately purified. For instance, plaque purification may not be adequate for purification of cytopathic BVDV without additional immunochemical or immunofluorescent staining to rule out the presence of noncytopathic BVDV. McClurkin et al. (1984), in their classic study on the production of immune-tolerant and persistently infected calves, produced persistent infections with noncytopathic BVDV but not with the cytopathic NADL strain. Apparently, their NADL stock virus was free of noncytopathic BVDV.

DIAGNOSIS

Within the past few years, several studies have examined the use of immunohistochemistry (IHC) on skin biopsies to detect BVDV infections, a technique that had been introduced earlier by Thür et al. (1996) for use on frozen specimens. The skin biopsy technique is now being used by some laboratories to screen herds for persistently infected cattle. Njaa et al. (2000) applied the technique on formalin-fixed, paraffin-embedded specimens and were able to detect positive staining in 41 of 42 skin samples from persistently infected calves. Sections of skin from all acutely infected animals were negative for staining when they were infected with 105 TCID50 of BVDV, and only 40% of acutely infected animals showed positive skin biopsies when infected with a much higher dose (108 TCID50) of BVDV. It was concluded that immunohistochemical staining of skin biopsies was an effective method for identifying persistently infected cattle.

Ridpath et al. (2002) used virus isolation as the standard to evaluate both the skin biopsy method and a serum PCR-probe test, another technique used for screening herds for persistently infected animals, for detection of acute BVDV infections. In this study, 16 animals were inoculated with 106 TCID50 of BVDV 1 or BVDV 2. Whereas virus was isolated from all animals, only three (19%) animals were positive by the PCR-probe test and none was positive by immunohistochemical (IHC) staining of skin biopsies. These researchers concluded that the skin biopsy test would usually not confuse persistently infected animals with acutely infected animals. However, both methods (IHC and PCR-probe) were

considered unreliable for detecting acute outbreaks caused by BVDV.

Recently, the use of skin biopsies as diagnostic specimens for a BVDV antigen capture ELISA was evaluated (Brodersen et al., 2002). The assay, based on the detection of the Erns protein, was reported to be equally sensitive to PCR, IHC, and virus isolation from leukocytes. The same capture ELISA had been previously validated for the detection of BVDV in serum for screening herds for persistently infected animals and for screening BVDV in sublots of fetal bovine serum for the production of the commercialized product (Plavsic and Prodafikas, 2001).

The use of pooled samples for diagnostic testing, a trend that began in the 1990s, has continued in recent years. Weinstock et al. (2001) developed a single tube PCR assay for detection of BVDV in pooled serum. The assay was compared with a microplate virus isolation method performed on individual sera. The PCR assay was sensitive enough to detect a single viremic serum sample in 100 pooled samples. Muñoz-Zanzi et al. (2000) examined leastcost strategies for using PCR/probe testing of pooled blood samples to identify persistently infected animals in herds. For a herd prevalence of 1%, the least-cost strategy for pooled sample testing was to test pools from 20 animals initially, and then pools of five blood samples for repooled testing. As herd prevalence increased beyond 3%, the competitive benefit of pooled testing diminished.

Based on earlier work of Houe (1992; 1994) and Houe et al. (1995), who reported that antibody levels can be used to predict the presence of persistently infected animals in a herd, Pillars and Grooms (2002) applied this technique to herds with unknown BVDV status. Specifically, they examined whether serological evaluation of five unvaccinated, randomly selected, 6- to 12-month-old heifers is a valid method for identifying herds that contain persistently infected cattle. A herd was classified as likely to contain persistently infected animals when at least three of the heifers had neutralizing antibody titers of 1:128 to BVDV 1 or BVDV 2. Of the 14 herds examined, 6 contained persistently infected animals. The sensitivity and specificity of the serologic method for determining which herds had persistently infected animals was 66% and 100%, respectively. In one herd with a false negative result, by chance three of the heifers examined serologically were persistently infected animals and had titers of <1:4, whereas the other two animals had titers of >1:4,096. The authors noted that this herd would have been rightly classified had virus isola-