Coronavirus and Pasteurella Infections in Bovine Shipping Fever Pneumonia and Evans' Criteria for Causation

Experimental design.

Ten of 105 and 16 of 120 cattle of two SFP epizootics occurring in 1997 and 1998 (97TXSF and 98TXSF, respectively) were subjected to prospective, experimentally designed sampling and testing at the time of assembly (day 0), after transport (day 5), and throughout the pathogenesis of fatal pneumonia. The cattle were ear-tagged and clinically examined at assembly, when initial sets of nasal swab samples for culture of infectious agents and blood samples for serum harvest were collected. Nasal swab samples were placed into transport medium for virus isolation, while dry samples were used for bacterial cultures. Nasal swab samples were placed on dry ice and subsequently transferred into −70°C freezers. All cattle were vaccinated with modified-live virus vaccines against BHV-1 and PI-3 in 1997, and only against BHV-1 in 1998 (Prevail and Reliant; Rhone Merieux, Inc., Athens, Ga.). The experimental calves also were given blackleg vaccines (Electroid 7; Mallinckrodt Vet. Inc., Mundelein, Ill.) and ivermectin (Ivomec; Merck and Co., Inc., Rahway, N.J.). They remained for 4 days at the assembly site of the order buyer and then were transported 1,932 km to the feed yard in Bushland, Tex., jointly operated by the Agricultural Research Service and the Texas Agricultural Experiment Station. They were clinically examined on the day of arrival in the feed yard (day 5) and daily thereafter. Nasal swab and blood samples also were taken at this time and subsequently on a weekly schedule. Seven and 11 cattle of the respective epizootics remained clinically normal and were also tested microbiologically and serologically as contact controls.

Necropsies, pathological, histopathological, and immunocytochemical examinations.

Necropsies were performed immediately after death, and lung samples were collected at the Texas Veterinary Diagnostic Laboratory in Amarillo, Tex. Aliquots of the lung samples were frozen at −70°C for isolation, identification, and quantification of viruses and bacteria. Pieces from each lung were used to make exfoliative cytological or cryostat preparations for immunofluorescent tests with antibodies specific for RBCV and BVDV. Lung tissue pieces were placed into neutral 10% phosphate-buffered formalin for fixation. Histological sections were prepared and stained by the hematoxylin-eosin methods.

Bacterial isolation, identification, and quantification.

Dry nasal swab samples were removed from the freezer and allowed to thaw at 25°C. Each swab was streaked over one-fourth of the surface of a plate. The lung specimens were thawed, the surfaces were seared with a hot spatula, and a 1-g sample was removed from the center of each specimen. Lung tissues were minced with scissors and homogenized in phosphate-buffered saline (PBS) at pH 7.4 by grinding with a Ten Broeck device. The resulting 10% (wt/vol) suspension was serially diluted in 10-fold steps. Aliquots of 0.1 ml of the dilutions were plated in five replicas on tryptose agar plates fortified with 5% citrated bovine blood to enumerate CFU. The culture plates were incubated at 37°C for 24 h in an atmosphere with 5% CO₂. Bacterial colonies were counted, and their numbers on each replica were averaged. Colonies of P. haemolytica and P. multocida were identified by colony morphology, Gram staining, and biochemical reactions and by use of specific antisera for serotyping P. haemolytica isolates (8, 9, 34).

Preparation of clinical samples for virus isolation.

The nasal swab samples were thawed at 25°C, and 1 ml of cold Dulbecco's modified minimum essential medium (DMEM) was added to each tube. Aliquots of these fluids were centrifuged at 2,000 × g (model TJ-6R; Beckman Instruments, Inc., Palo Alto, Calif.) for 20 min. Supernatant fluids were withdrawn and filtered through Millipore filters (Gelman Sciences, Ann Arbor, Mich.) with a pore size of 0.450 μm. The filtrates were then used to inoculate the cell cultures. Lung tissues were minced and homogenized by grinding with Ten Broeck devices in the appropriate volume of cold DMEM to make 5% (wt/vol) suspensions. The lung homogenates were centrifuged at 2,000 × g. The top half of the supernatant was withdrawn, diluted 1:1 with cold DMEM, and subjected to a 2nd cycle of centrifugation. The top half of this supernatant was withdrawn, filtered as indicated, and used for virus isolation tests or RBCV infectivity titration (28, 30).

Virus isolation methods.

Specific virus isolation tests depended on three different cell types selectively permissive for the known respiratory bovine viruses. Cell cultures included the G clone of human rectal tumor-18 (HRT-18) cells specifically permissive for RBCV (19, 28, 33), Georgia bovine kidney (GBK) cells permissive for BHV-1, PI-3, and bovine adenoviruses (BAV), as well as bovine turbinate (BT) cells highly permissive for BRSV and cytopathogenic BVDV, and, to a lesser degree, permissive for BHV-1, PI-3, and BAV (28). These cell types were propagated in 24-well cluster plates (Becton-Dickinson Labware, Franklin Lakes, N.J.) and used in virus isolation tests by inoculating four wells: two of them with 10⁻¹ and two with 10⁻² dilutions of each test sample. Cells in four wells remained as uninoculated cell controls for each test plate. The inoculated cell cultures were incubated at 37°C and examined for cytopathic changes with an inverted microscope for 2 to 5 days. The cell cultures then were frozen at −70°C and thawed at 25°C. Cultures with cytopathic changes after inoculation with specific samples were pooled, and those remaining normal were also pooled as single samples for subpassages in the respective permissive cell cultures and for detecting hemagglutinin (HA) or receptor-destroying enzyme (RDE) activities (28, 31).

Identification of virus isolates.

The virus isolates were initially differentiated by characteristics of cytopathic effects in the specifically permissive cell types. Additional virus identification tests included HA detection with rat erythrocytes (RBC) and assays for RDE functions mediated by acetylesterase (AE), a characteristic of RBCV. Tests for HA of PI-3 involved the use of bovine and chicken RBC. Virus isolates were further identified by using specific antisera in HA inhibition (HAI), infectivity neutralization tests, or negative-contrast staining and electron microscopic examinations (28).

Quantification of RBCV in the infected lung tissues.

The plaque test was used to assess the titers of PFU of RBCV in 25 available lung samples according to recently described procedures (19). The lung samples prepared for virus isolations were serially diluted in 10-fold steps, and 0.5-ml aliquots of the respective dilutions were placed on confluent G clone cell monolayers grown in six-well cluster plates. (Becton-Dickinson Labware, Franklin Lakes, N.J.). The plaque test for PI-3 infectivity involved GBK cell cultures and was otherwise identical. The plaque test for BHV-1 employed GBK cell cultures, a 2% methylcellulose overlay, and crystal violet staining.

Tests for viral HA and RDE.

Washed rat RBC at a concentration of 0.5% in PBS at pH 7.4 containing 0.05% bovine serum albumin (BSA) were used to detect HA antigens of RBCV in infected G clone cultures. The same concentrations of bovine or chicken RBC were used to detect HA of PI-3 isolates. Twofold dilutions of 50 μl of the samples were made in 96-well V-bottom microtiter plates (Becton Dickinson Labware, Franklin Lakes, N.J.), and equal volumes of RBC suspensions were added to each well (31). The plates were shaken and incubated at 6°C for 2 h, which is sufficient time for the RBC to form clear buttons in diluent control wells. The HA titers were recorded as the highest dilution with complete agglutination of RBC suspensions. The plates were then incubated at 37°C for activation of the RDE function. The RDE titer was determined by elution of virus from rat RBC in wells with previous agglutination and was recorded after 2 h as the highest dilution with settled RBC (31).

HAI tests.

Sequentially serum samples from the 26 cattle with fatal pneumonia and from 18 normal control cattle were tested for HAI antibodies against RBCV antigen. These sera were also tested for antibodies against P. haemolytica antigen in an enzyme-linked immunosorbent assay (ELISA). The serum samples were diluted 1:4 in PBS containing 0.05% BSA, heat-inactivated at 56°C for 30 min, and then diluted in twofold steps in 50-μl aliquots. An antigen extracted from RBCV-infected cell lysates was used and diluted to contain 8 to 16 U of HA and RDE. The antigen was added in 50-μl volumes to each serum dilution. The serum-antigen mixtures reacted for 30 min at 25°C, and 50 μl of rat RBC suspensions then was added. The test plates were kept at 6°C for 2 h, and the HAI titers were determined as the highest dilutions inhibiting HA of RBCV. Serum 1745 with a known HAI antibody titer and normal serum were included as positive and negative controls, respectively.

Determination of P. haemolytica antibodies.

A 32-kDa outer membrane protein of P. haemolytica was harvested from 18- to 24-h cultures in beef heart infusion broth as the supernatant after a 10,000 × g centrifugation. The antigen preparation was diluted 1:12.5, test sera were diluted 1:50, and the ELISA was performed as described previously (3).

Shedding of RBCV and Pasteurella spp. in nasal secretions before death.

Virus and bacterial isolation results from nasal swab samples of the 26 fatal cases from the two epizootics are recorded in Table 1. In the 1997 experiment, six cattle shed G clone-dependent RBCV in nasal secretions on day 0, and three additional calves contracted this infection on day 5. In the 1998 experiment, 15 of the 16 cattle shed RBCV at the beginning of the experiment, and all of them did so on day 5. Other viruses were not isolated from nasal discharges of these 26 cattle during the first 5 days of the experiments. Three cattle, surviving 14 to 36 days after onset of this epizootic, began shedding BHV-1 on day 12 and PI-3 on day 26 (data not shown). Calf 98TXSF-115 had a dual infection with RBCV and BHV-1 on day 12, and calves 98TXSF-15 and 98TXSF-100 nasally shed BHV-1 on days 12 and 19. Calf 98TXSF-15 nasally shed BHV-1 and PI-3 on day 26 (data not shown). Calf 98TXSF-100 was nasally BHV-1 isolation negative on day 26, and became PI-3 isolation positive on day 33 (data not shown). Uninoculated cell culture controls maintained normal morphological features in all of these virus isolation tests. Two of the 26 cattle shed P. haemolytica on day 0, and 17 of them carried P. haemolytica after transport on day 5. The 18 normal control cattle did not secrete RBCV in their nasal discharges during the test periods (30). None of the 18 normal control cattle had Pasteurella spp. in nasal swabs on day 0, and 3 of them nasally shed P. haemolytica on day 5 (data not shown).

Combined lung infections with RBCV and Pasteurella spp. and respective titers in the 1997 and 1998 epizootics.

Pneumonic lung samples from 9 of 10 cattle of the 1997 experiment, and from 9 of 16 cattle of the 1998 experiment were RBCV isolation positive when they died between days 5 and 7 (Table 2). The RBCV titers ranged from 1 × 10³ to 1.2 × 10⁷ PFU per g of lung tissues. BHV-1 was detected at 2 × 10³ PFU per g of lung tissue of calf 98TXSF-115, which died on day 14. Less than 10³ PFU of PI-3 per g of lung tissue were detected in calves 98TXSF15 and 98TXSF-100, which died on days 31 and 36, respectively. The last calf had the most drawn-out respiratory tract disease and was treated with antibiotics six times at different intervals. The lung of calf 98TXSF-114 was not available for virological testing. The 25 available lung tissues of 26 dead cattle tested did not yield BRSV, cytopathogenic BVDV, or BAV through inoculation of selectively permissive cell culture systems. Cells of the exfoliative or cryostat preparations from these lungs did not react with BVDV-specific fluorescent antibodies. This test was included to detect the potential presence of noncytocidal BVDV in these lungs.

P. haemolytica was cultured from the lungs of all 10 calves which died in the 1997 epizootic, as indicated in Table 2. Calf 97TXSF-15 had a pure culture of P. haemolytica A6. The lungs of the other nine calves had P. haemolytica A1 infections, and calf 97TXSF-52 also had P. multocida infections. Pasteurella spp. were isolated from 15 lungs of 16 calves which died in the 1998 epizootic (Table 2). Four calves (98TXSF-10, -71, -72, and -91) had mixed infections of P. haemolytica A1 and P. multocida. Calves 98TXSF-14, -36, -85, -100, and -110 had pure P. multocida infections of the lungs. The bacterial isolate from the lung of calf 98TXSF-102 was identified as a pure culture of P. haemolytica A2, while the bacterial isolates from the remaining five calves were identified as pure P. haemolytica A1. The lungs of calf 98TXSF-15 that died at the end of the study after repeated treatment with antibiotics were negative in bacterial cultures. The CFU of P. haemolytica varied from 4 × 10⁵ to 1.4 × 10⁹ per g of lung tissue. The CFU of P. multocida ranged from 6.0 × 10⁵ to 2.3 × 10⁹ per g of lung tissue.

Biological properties of the virus isolates.

The RBCV isolated from the pneumonic lungs were indistinguishable from the isolates detected in sequential nasal swab samples. These RBCV isolates exhibited high fusogenic functions in G clone cells in the first passage, did not replicate in GBK or BT cells, and agglutinated rat but not bovine or chicken RBC. All RBCV isolates destroyed rat RBC glycoconjugate receptors as a function of their RDE. Uniformly large RBCV plaque phenotypes with diameters of 5 to 6 mm were detected in the plaque titration of the pneumonic lungs of cattle in the 1997 and 1998 epizootics (data not shown). The viruses isolated by inoculation of GBK or BT cell cultures were identified as BHV-1 through their cytopathic changes consisting of clusters of rounded cells and neutralization by BHV-1-specific antiserum. They were not temperature sensitive like some commercial live BHV-1 vaccines because they multiplied at 36 and 39°C. The virus isolates from the lungs of calves 98TXSF-15 and 98TXSF-100 induced cell fusion in GBK or BT cells and agglutinated chicken and bovine RBC, and their HA was inhibited by PI-3-specific antiserum. These properties identified these isolates as PI-3.

Antibody responses to infections with RBCV and P. haemolytica of cattle developing fatal pneumonia and healthy resistant cattle.

The dead cattle of both epizootics had HAI antibody titers of <8 to 16 on day 0, with the exception of calf 97TXSF-63. These titers remained at these low levels during the first 5 days (Table 3). The HAI antibody titers of calf 97TXSF-63 that died on day 9 was 32 on day 0 and increased to 256 on day 5. Both nasal swab and lung samples from this calf remained negative for virus isolation. The three cattle of the 1998 experiment that died between days 27 and 36 had increasing HAI titers reaching 64. Isolation of RBCV from lung tissues of these cattle was unsuccessful. The ELISA results indicated that all except one calf were antibody negative for P. haemolytica antigen on day 0, and the antibody levels rose among 10 of 18 cattle that were tested on day 5 (Table 3). The four cattle that died between days 14 and 36 maintained such titers.

The HAI antibody titers of the 18 normal control cattle that remained clinically healthy and did not nasally shed RBCV in the 1997 and 1998 experiments ranged from 16 to 1,024 on day 0 (Table 4). Their antibody titers continued to rise and remained high on days 12 and 19. Ten of the 18 cattle were antibody negative for P. haemolytica antigen on day 0, 16 had positive titers on day 5, and all had significant titers on days 12 and 19.

Clinical and pathological findings.

Ten of 105 cattle in the 1997 experiment died on days 5 to 9, and 16 of 120 cattle in the 1998 epizootic died on days 5 to 36 (Tables 1 and 2). They all developed fever as high as 44°C and severe signs of respiratory distress. The necropsy examinations revealed that 50 to 80% of the total lung volumes had subacute exudative or subacute exudative and necrotizing lobar pneumonia primarily affecting the anteroventral areas of the lungs. The exudative component of the pneumonic process consisted of fibrin deposited to various degrees on the pleural surfaces of lungs or within distended intralobular septae and airways. Gross lesions in the gastrointestinal tracts or other organ systems of these cattle were not detected. All cattle were treated under good nutritional conditions.

Histological changes.

The microscopic changes in a majority of the pneumonic lung lobules were characterized as acute and subacute fibrinous bronchopneumonia. Typical exudative lesions associated with pneumonic pasteurellosis were not evident in isolated lung lobules. These lobules had moderate to severe bronchitis and bronchiolitis consisting of degenerating and necrotic respiratory epithelium with subepithelial and intraepithelial, nonsuppurative inflammatory cell infiltration (Fig. 1). Hyperchromatic and fused epithelial cells were present in airway epithelium adjacent to areas of necrosis, degeneration, and intraepithelial inflammation. These changes were detected at all levels of the lower respiratory tract from small bronchi to alveolar ducts. Many alveoli were no longer lined by type I pneumocytes and were partially filled with pink, protein-rich exudates and mixtures of small and large mononuclear leukocytes. Intracytoplasmic or intranuclear inclusions were not discernable. Cytoplasmic fluorescence was detected in epithelial cells of bronchi, broncheoli, and alveolar ducts when cryostat sections of the lungs of cattle that died within 7 days were tested with fluorescing antibodies against RBCV (data not shown). Cells of these lung sections did not fluoresce when treated with RBCV antibody-negative serum or with BVDV antiserum.