Vol. 10, No. 2
JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1979, p. 240-244 0095-1 137/79/08-0240/05$02.00/0
Counterimmunoelectroosmophoresis for Detection of Neonatal Calf Diarrhea Coronavirus: Methodology and Comparison with Electron Microscopy SERGE DEA, RAYMOND S. ROY,* AND MICHEL E. BEGIN Departement de Pathologie et de Microbiologie, Faculte de Medecine Veterinaire, Universite de Montreal, St-Hyacinthe, Quebec, Canada J2S 7C6 Received for publication 22 May 1979
A counterimmunoelectroosmophoresis (CIE) technique is described for the detection of calf diarrhea coronavirus antigens in intestinal contents. The antibody reagent was prepared in rabbits against the Nebraska calf diarrhea coronavirus adapted to Vero cells and purified by density gradient centrifugation. The method was applied to intestinal contents of diarrheic and normal calves and compared with electron microscopy (EM). Calf coronavirus antigens were detected in intestinal contents of 44% (21/48) of the diarrheic calves and 24% (4/17) of the normal calves. Two precipitin lines could be observed in the majority of the positive samples. When compared with EM, CIE detected more positive animals. In only two cases (2/20) CIE was negative despite the visualization of coronavirus particles by EM.
Diarrhea associated with the presence of coronavirus particles in feces of neonatal calves is now frequently observed and has been reported from various parts of the world (11, 14-16, 18, 19). Based on incidence data from two studies, coronavirus infections ranked second in importance in economic losses due to infectious calf diarrhea (5). The detection of coronaviruses is presently done by electron microscopic (EM) examination of intestinal contents or feces, by in vitro cell culture, and by immunofluorescence staining of gut sections (3, 6-9, 12). Immunofluorescence staining of colon sections (9, 11, 12) is the diagnostic method of choice for most laboratories. Unfortunately, this method can be used only following the death of the calves or by sacrificing a calf for diagnostic purposes. The efficiency of negative staining for EM examination is often disputable. Some workers reported that it permitted the frequent visualization of the virus in diarrheal feces (7), whereas others stated that the examination of the specimens often revealed a very small number of particles with morphology resembling that of coronaviruses but that the structures frequently were not sufficiently well defined to establish their identity. The surface projections in particular are often badly preserved (3). However, provided the samples contain a large number of virus particles, they can be rapidly identified (12, 16). Although some bovine coronaviruses can 240
readily multiply and induce cytopathic effects in FBK (9, 15), BEK-1 (6), Vero, and MDBK cells (S. Dea et al., manuscript submitted for publication), the cultivation of the virus appears difficult, and many passages are required to obtain satisfactory yields of infectious particles. However, in most instances the presence of the virus can be demonstrated on first passage through the use of EM and immunofluorescence procedures (6, 9, 15). The purpose of the present work was to describe and compare with EM the use of counterimmunoelectroosmophoresis (CIE) for rapid detection of the coronaviruses associated with neonatal calf diarrhea. MATERIALS AND METHODS Source and preparation of intestinal samples. Jejunal and colonic contents were collected from 48 diarrheic calves within the first 5 to 6 h after the onset of diarrhea and from 17 healthy calves in 55 different herds around St. Hyacinthe, Quebec, Canada. All animals were examined clinically and autopsied; histopathological sections were made of the intestines, and their lumen contents were collected. Diarrheic animals showed intestinal lesions of bacterial or viral origin, whereas normal animals were asymptomatic and lacked intestinal lesions (M. Morin et al., Proceedings of the 2nd Symposium on Neonatal Diarrhea, Saskatoon, 1979, in press). All calves were between 1 and 21 days old when submitted for examination. The intestinal contents of each animal were combined and diluted 1:5 in 0.01 M tris(hydroxymethyl)aminomethane (Tris)-hydrochloride buffer (pH
VOL. 10, 1979
CIE FOR DETECTION OF CALF CORONAVIRUS
241
8.0), sonicated for 30 s at 20 kilocycles per s, clarified The animals were bled on day 63. Thereafter, the antiserum was inactivated at 56°C by centrifugation at 5,000 x g for 30 min, and then filtered through membrane filters (Millipore Corp.) of for 30 min and then treated to eliminate unwanted nonspecific antibodies. A volume of 9 ml of the anti200-nm pore size. Viruses used as controls. The FBK cell culture- serum was mixed with 1 ml of fetal calf serum and adapted Nebraska calf diarrhea coronavirus (NCDC) incubated at 4°C overnight. The precipitate formed (9) kindly supplied by C. A. Mebus, University of was eliminated by centrifugation at 5,000 x g for 30 Nebraska, was adapted and propagated in Vero cells. min. Then, 10 ml of the supernatant fluid was mixed The virus was used at Vero cell passage level 27. with 1 g of bovine liver powder for 1 h and then Infectious bovine rhinotracheitis (IBR), strain Colo- clarified by centrifugation at 10,000 x g for 30 min. rado, and bovine enterovirus type II (BE-II) viruses, The antiserum was subsequently washed with Vero both adapted on Vero cells, and calf rotavirus supplied cells for 30 min, recovered, tested, and then used in by C.A. Mebus and adapted on MDBK cells in our serological tests. This coronavirus antiserum was shown to be free of reactivity against calf serum and laboratory were used as controls. Virus propagation. Confluent monolayers of Vero bovine tissues by immunodiffusion and CIE. Titration of neutralizing antibodies. Twofold or MDBK cells were rinsed with Dulbecco phosphatebuffered saline (pH 7.0) and inoculated with 1 ml of dilutions of the antiserum (1:10 to 1:1,280) were made infected cell extracts. After 1 h of viral adsorption at in 0.01 M phosphate-buffered saline (pH 7.2) supple37°C, the cultures were overlaid with Eagle minimal mented with 0.1% bovine serum albumin. A 0.1-ml essential medium with Earle salt and glutamine portion of a virus dilution containing 400 TCID50 was (GIBCO) and incubated at 37°C in a 5% CO2 atmos- added to 0.1 ml of appropriate serum dilutions. The phere. When maximal cytopathic effects occurred, in- virus-serum mixtures were agitated vigorously and fected cells and cell culture fluid were frozen and allowed to remain at 37°C for 1 h. Thereafter, each thawed three times and clarified and stored at -70°C. mixture was inoculated in 0.05-ml amounts into each Viral infectivity was determined as described else- of three wells of a 96-well Microtest culture tray where (S. Dea et al., manuscript submitted for publi- (Falcon Plastics, Oxnard, Calif.) plated with Vero cells. cation), and viral titers were expressed in 50% tissue The cultures were examined for evidence of cytopathic culture infective doses (TCID5o) according to the effect after 4 to 5 days postinfection at 37°C. The titer was expressed as the reciprocal of the highest neutralmethod of Reed and Muench (13). Electron microscopy. Droplets of the clarified izing serum dilution. CIE. The CIE test was performed on Kodak projecfecal contents were placed on a 200-mesh Formvarand carbon-coated grids and allowed to remain on the tor slides (8.3 by 10.2 cm) coated with 13 ml of agarose grids for 30 s. Fluids were then blotted from the grids, in Tris-barbital buffer. After the gel had become firm and a droplet of distilled water was added for 10 s and at room temperature, two parallel rows of wells (4.0 blotted off. The specimen was stained for 10 s with 3% mm in diameter) were punched 1 cm center to center. phosphotungstate acid (pH 7.0). The excess stain was Anodal wells were filled with 15 ,ll of antiserum, and blotted off. The grids were dried and then examined cathodal wells were filled with 15 tll of antigen. The with a Philipps 300 electron microscope operating at antigens used were the clarified, infected cell culture fluids or intestinal contents obtained as described an accelerating potential of 80 kV. Preparation of coronavirus antigen. Approxi- above and diluted in Tris-barbital buffer of the same mately 300 ml of NCDC-infected Vero cell culture molarity as that used to make up the agarose. Telfa fluids were frozen and thawed three times and centri- nonadherent strips (Kendall, Chicago, Ill.) (4 by 10 fuged at 5,000 x g for 30 min at 4°C to remove coarse cm) were used as wicks. Directly after electrophoresis, debris. The clarified suspension was then concentrated the slides were washed overnight at 0.85% NaCl and to a volume of 5 ml by ultrafiltration through an XM- then immersed in 1% tannic acid. Thereafter, the 100A Diaflo membrane. The concentrate was centri- agarose slides were examined with a high-beam illufuged at 78,600 x g for 2 h at 4°C. The supernatant minator for the presence of a precipitin line between fluid was discarded, and the pellet containing the virus antigen and antiserum wells. particles was dispersed by sonication at 20 kilocycles RESULTS per s for 30 s in 0.5 ml of 0.05 M Tris-hydrochloride buffer (pH 7.4). The viral suspension was then diluted Conditions for CIE. Some parameters of the in the same buffer up to 250 ,ug of protein per ml as test were studied by using cell culture fluid CIE determined by optical density at 280 nm and used for
immunization. Preparation of antiserum. Two New Zealand
Albino rabbits (2 kg) were used for immunization. They were inoculated intramuscularly at four sites (0.5 ml per site) and in the foodpads (0.5 ml) with a mixture of equal volumes of NCDC antigen and complete Freund adjuvant. After 20 days, the rabbits received an intravenous dose of 1.0 ml of antigen along with intramuscular injections (0.50 ml per site) consisting of a mixture of 1.0 ml of antigen and 1.0 ml of incomplete Freund adjuvant. On days 35 and 49, they were reinoculated intramuscularly as described previously.
from NCDC-infected Vero cells as the source of virus. A visible double precipitin line was obtained between antibody and antigen wells when we used as antibody the anti-NCDC serum supplied by C.A. Mebus and the rabbit anti-NCDC serum produced in our laboratory (Fig. 1). Both antisera possessed a neutralizing titer of 320. No precipitin lines were observed with our antiNCDC serum against IBR, Nebraska calf rotavirus, and BE-II viruses.
Electrophoresis
was conducted at
200, 150,
242
DEA, ROY, AND BEGIN
J. CLIN. MICROBIOL.
passage 27 of NCDC virus in Vero cells against our antiserum. Precipitin lines were noted with antigen dilutions up to 1:1,000. The endpoint dilution corresponded to 10: TCIDro per ml. No precipitin lines were observed with undiluted coronavirus suspensions when antiserum was diluted more than 1:4. Comparison of EM and CIE for detecting coronavirus. Typical coronavirus particles with an average diameter of 120 nm and surrounded by bulbous or petal-shaped projections 13 to 17 nm long were visualized by EM in 16 (33%) of 48 diarrheal intestinal contents and in 4 (24%) of 17 normal intestinal contents (Fig. 2). As shown in Table 1, nine diarrheic calves were singly infected by coronavirus, whereas three others had only rotaviruses. Five calves showed mixed infection by coronavirus and rotavirus, one calf showed infection with coronavirus and reovirus, and another one showed infection with coronavirus and picornavirus. A total of 29 diarrheic calves were free of any virus by this technique. Four and two of the healthy calves were singly infected with coronavirus or with picornavirus, respectively. The other 11 calves were free of FIG. 1. (Top) CIE precipitin lines observed with any viral particles as examined by EM. As shown in Table 1, 21 (44%) of 48 clarified the cell culture fluid of passage 27 of NCDC on Vero cells and anti-NCDC (rabbit) after electrophoresis at 150 V for 90 min and staining with 1% tannic acid for 15 min. (A, B) Bovine coronavirus. (C) Nebraska calf rotavirus, passage 7 on MDBK cells. (D) BE-II virus, passage 5 on Vero cells. (E) IBR virus, passage 5 on Vero cells. (F) Nebraska coronavirus antiserum of C.A. Mebus. (G-J) Nebraska coronavirus antiserum produced in our laboratory. (Bottom) Magnification showing clearly the double-line pattern.
and 100 V for 60, 90, and 120 min at 4°C by using 0.05 M Tris-barbital buffer in the electrophoresis chamber and 0.025 M in the agarose. The best results were noted when electrophoresis was conducted at 150 V for 90 min. A buffer of pH 8.6 was shown to give better reactions than those of pH 7.0, 7.5, and 8.0. A more discernible precipitin line was obtained if antiserum and antigen wells were separated by 1 cm instead of 0.7 cm. A 1% agarose concentration gave better results than did concentrations of 0.7 and 0.5%. Very weak precipitin lines were observed if the agarose slides were not stained. A clear-cut reaction was obtained after staining the slides for 15 min in 1% tannic acid after an overnight wash in normal saline following electrophoresis. A longer exposure to tannic acid did not amplify the precipitin lines. Sensitivity of CIE. CIE was conducted with 10-fold dilutions (1:10 to 1:10,000) in 0.025 M electrophoresis buffer of the cell culture fluid of
.-j.
FIG. 2. Typical coronavirus particles observed by negative staining EM in clarified intestinal content of a diarrheic calf x288,000. Bar = 40 nm.
VOL. 10, 1979
CIE FOR DETECTION OF CALF CORONAVIRUS
243
intestinal contents of diarrheic calves and 4 TABLE 1. Comparison of EM and CIE for detection of coronavirus in intestinal contents of diarrheic (24%) of 17 intestinal contents from normal and normal calves calves contained coronavirus antigens as deNormal calves tected by CIE. No antigenic reaction occurred Diarrheic calves with the intestinal contents shown to contain CIE CIE___ Virus seen by only rotavirus or only picornavirus by EM. More EM Virus seen by EM than half of the samples that were positive for coronaviruses by this technique were positive Coronavirus (9)' 1 3 Coronavirus 0 9 only after staining with 1% tannic acid as de(4)a 2 0 scribed above. Picornavirus3 0 Rotavirus (3) like (2) As reported in Table 1, CIE was positive on 1 10 0 No virus ob5 and all intestinal contents shown to contain corona- Coronavirus served (11)h rotavirus (5) virus by EM with the exception of one diarrheic Coronavirus 1 0 and and one normal calf. Coronavirus antigens were reovirus (1) 1 0 detected by CIE in 6 of the 29 intestinal contents Coronavirus and (1) from diarrheic calves in which no viral agents Nopicornavirus 6 23 virus observed were seen by EM. From the 11 normal intestinal (29)b contents that were free of any virus particles as in parentheses indicate the number of samples determined by EM, in only one instance did we withaNumbers the indicated virus(es). observe a positive reaction by CIE. bNumber of samples in which no viruses could be observed These results indicate that CIE could detect by EM. more coronavirus-positive intestinal contents than could EM. However, a negative result obthe healthy calves were infected with coronavitained by CIE must be verified by EM. Based on the combined use of EM and CIE, rus but had not yet developed diarrhea. More46% (22/48) of the diarrheic calves and 30% (5/ over, no pathological lesions were seen in these 17) of the healthy calves studied had coronavi- normal calves after a postmortem examination (M. Morin et al., Proceedings of the 2nd Symruses in their intestinal contents. posium on Neonatal Diarrhea, Saskatoon, 1979, in press). Samples collected from other regions DISCUSSION over a long period of time are needed in Two precipitin lines were observed in the ma- to Quebec and to generalize this finding. If this confirm jority of the positive samples, indicating that, as is correct, the significance of a high percentage with other members of the coronavirus group, of asymptomatic calves containing such high one than the bovine coronavirus possesses more needs to be elucidated to of coronavirus levels precipitating antigen (1, 2, 4, 10, 17). of the correct the picture obtain As applied to intestinal contents, the reaction and the mechanism of the onset ofepidemiology the diarrhea seemed specific for coronavirus since no immu- associated with coronaviruses. It would appear noprecipitate could be detected with the con- necessary to identify the factors modifying the tents shown to contain viruses other than coroof coronavirus infections. navirus by EM. It is worth stressing the fact that clinical effects two intestinal contents were negative by CIE ACKNOWLEDGMENTS although they were positive by EM. This fact The technical assistance of Francine Breton is gratefully could be explained by a technical failure of CIE acknowledged. This study was supported by the Conseil de Recherches et or observer error by EM. However, it could also Services Agricoles du Quebec (Canada) grant no. MMV 75suggest the existence of more than one antigenic de 604. variant among bovine coronaviruses. LITERATURE CITED CIE provided a reliable, simple, and fast method for large-scale screening of intestinal 1. Bohac, J., and J. B. Derbyshire. 1975. The demonstracontents for coronavirus. However, before montion of transmissible gastroenteritis viral antigens by immunoelectrophoresis and counterimmunoelectroospecific antisera to all known bovine coronaviphoresis. Can. J. Microbiol. 21:750-753. ruses are available, it is advisable to further 2. Bohac, J., J. B. Derbyshire, and J. Thorsen. 1975. EM. examine the CIE-negative samples by The detection of transmissible gastroenteritis viral antigens by imnimunodiffusion. Can. J. Comp. Med. 39:67Unexpectedly, coronavirus particles and anti75. gens were detected by EM and/or CIE in 30% of 3. Chasey, D., and M. Lucas. 1977. A bovine coronavirus the healthy calves. To our knowledge no asympidentified by thin-section electron microscopy. Vet. Rec. tomatic coronavirus shedding has been reported 100:530-531. in this species. However, since all calves were 4. Hierholzer, J. C., E. L. Palmer, S. G. Whitfield, H. S. Kaye, and W. R. Dowdle. 1972. Protein composition subjected to necropsy, we do not know whether
244
DEA, ROY, AND BEGIN
of coronavirus OC43. Virology 48:516-527. 5. House, J. A. 1978. Economics of rotavirus and other neonatal disease agents of animals. J. Am. Vet. Med. Assoc. 173:573-576. 6. Inaba, Y., K. Sato, H. Kurogi, E. Takahashi, Y. Ito, T. Omori, Y. Goto, and M. Matumoto. 1976. Replication of bovine coronavirus in cell line BEK-1 culture. Arch. Virol. 50:339-342. 7. Marsolais, G., R. Assaf, C. Montpetit, and P. Marois. 1978. Diagnosis of viral agents associated with neonatal calf diarrhea. Can. J. Comp. Med. 42:168-171. 8. Mebus, C. A., L. E. Newman, and E. L. Stair. 1975. Scanning electron, light, and immunofluorescent microscopy of intestine of gnotobiotic calf infected with calf diarrheal coronavirus. Am. J. Vet. Res. 36:17191725. 9. Mebus, C. A., E. L. Stair, M. B. Rhodes, and M. J. Twickhaus. 1973. Neonatal calf diarrhea: propagation, attenuation and characteristics of a coronavirus-like agent. Am. J. Vet. Res. 34:145-150. 10. Mengeling, W. L. 1972. Precipitating antigens of haemagglutinating encephalitis virus demonstrated by agar gel immunodiffusion. Am. J. Vet. Res. 33:1527-1535. 11. Morin, M., P. Lamothe, A. Gagnon, and R. Malo. 1974. A case of viral neonatal calf diarrhea in a Quebec dairy herd. Can. J. Comp. Med. 38:236-243. 12. Norden Laboratories. 1973. Laboratory methods for
detecting calf diarrhea virus. Norden Laboratories, Lin-
J. CLIN. MICROBIOL. coln, Neb. 13. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty percent end points. Am. J. Hyg. 27:493497. 14. Scherrer, R., J. Cohen, R. L'Haridon, C. Feynerol, and J. C. Fayet. 1976. Reovirus-like agent (rotavirus) associated with neonatal calf gastroenteritis in France. Ann. Rech. Vet. 7:25-31. 15. Sharpee, R. L., C. A. Mebus, and E. P. Bass. 1976. Characterization of a calf diarrheal coronavirus. Am. J. Vet. Res. 37:1031-1041. 16. Stair, E. L., M. B. Rhodes, R. G. White, and C. A. Mebus. 1972. Neonatal calf diarrhea: purification and electron microscopy of a coronavirus-like agent. Am. J. Vet. Res. 33:1141-1156. 17. Tevethia, S. S., and C. H. Cunningham. 1968. Antigenic characterization of infectious bronchitis virus. J. Immunol. 100:793-798. 18. Woode, G. N., J. C. Bridger, G. Hall, and M. J. Dennis. 1974. The isolation of a reovirus-like agent associated with diarrhoea in colostrum deprived calves in Great Britain. Res. Vet. Sci. 16:102-105. 19. Zygraich, N., A. M. Georges, and E. Vascoboinic. 1975. Etiologie des diarrhees neonatales du veau. Resultats d'une enquete serologique relative aux virus reo-like et corona dans la population bovine belge. Ann. Med. Vet. 119:105-113.
JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1979, p. 240-244 0095-1 137/79/08-0240/05$02.00/0
Counterimmunoelectroosmophoresis for Detection of Neonatal Calf Diarrhea Coronavirus: Methodology and Comparison with Electron Microscopy SERGE DEA, RAYMOND S. ROY,* AND MICHEL E. BEGIN Departement de Pathologie et de Microbiologie, Faculte de Medecine Veterinaire, Universite de Montreal, St-Hyacinthe, Quebec, Canada J2S 7C6 Received for publication 22 May 1979
A counterimmunoelectroosmophoresis (CIE) technique is described for the detection of calf diarrhea coronavirus antigens in intestinal contents. The antibody reagent was prepared in rabbits against the Nebraska calf diarrhea coronavirus adapted to Vero cells and purified by density gradient centrifugation. The method was applied to intestinal contents of diarrheic and normal calves and compared with electron microscopy (EM). Calf coronavirus antigens were detected in intestinal contents of 44% (21/48) of the diarrheic calves and 24% (4/17) of the normal calves. Two precipitin lines could be observed in the majority of the positive samples. When compared with EM, CIE detected more positive animals. In only two cases (2/20) CIE was negative despite the visualization of coronavirus particles by EM.
Diarrhea associated with the presence of coronavirus particles in feces of neonatal calves is now frequently observed and has been reported from various parts of the world (11, 14-16, 18, 19). Based on incidence data from two studies, coronavirus infections ranked second in importance in economic losses due to infectious calf diarrhea (5). The detection of coronaviruses is presently done by electron microscopic (EM) examination of intestinal contents or feces, by in vitro cell culture, and by immunofluorescence staining of gut sections (3, 6-9, 12). Immunofluorescence staining of colon sections (9, 11, 12) is the diagnostic method of choice for most laboratories. Unfortunately, this method can be used only following the death of the calves or by sacrificing a calf for diagnostic purposes. The efficiency of negative staining for EM examination is often disputable. Some workers reported that it permitted the frequent visualization of the virus in diarrheal feces (7), whereas others stated that the examination of the specimens often revealed a very small number of particles with morphology resembling that of coronaviruses but that the structures frequently were not sufficiently well defined to establish their identity. The surface projections in particular are often badly preserved (3). However, provided the samples contain a large number of virus particles, they can be rapidly identified (12, 16). Although some bovine coronaviruses can 240
readily multiply and induce cytopathic effects in FBK (9, 15), BEK-1 (6), Vero, and MDBK cells (S. Dea et al., manuscript submitted for publication), the cultivation of the virus appears difficult, and many passages are required to obtain satisfactory yields of infectious particles. However, in most instances the presence of the virus can be demonstrated on first passage through the use of EM and immunofluorescence procedures (6, 9, 15). The purpose of the present work was to describe and compare with EM the use of counterimmunoelectroosmophoresis (CIE) for rapid detection of the coronaviruses associated with neonatal calf diarrhea. MATERIALS AND METHODS Source and preparation of intestinal samples. Jejunal and colonic contents were collected from 48 diarrheic calves within the first 5 to 6 h after the onset of diarrhea and from 17 healthy calves in 55 different herds around St. Hyacinthe, Quebec, Canada. All animals were examined clinically and autopsied; histopathological sections were made of the intestines, and their lumen contents were collected. Diarrheic animals showed intestinal lesions of bacterial or viral origin, whereas normal animals were asymptomatic and lacked intestinal lesions (M. Morin et al., Proceedings of the 2nd Symposium on Neonatal Diarrhea, Saskatoon, 1979, in press). All calves were between 1 and 21 days old when submitted for examination. The intestinal contents of each animal were combined and diluted 1:5 in 0.01 M tris(hydroxymethyl)aminomethane (Tris)-hydrochloride buffer (pH
VOL. 10, 1979
CIE FOR DETECTION OF CALF CORONAVIRUS
241
8.0), sonicated for 30 s at 20 kilocycles per s, clarified The animals were bled on day 63. Thereafter, the antiserum was inactivated at 56°C by centrifugation at 5,000 x g for 30 min, and then filtered through membrane filters (Millipore Corp.) of for 30 min and then treated to eliminate unwanted nonspecific antibodies. A volume of 9 ml of the anti200-nm pore size. Viruses used as controls. The FBK cell culture- serum was mixed with 1 ml of fetal calf serum and adapted Nebraska calf diarrhea coronavirus (NCDC) incubated at 4°C overnight. The precipitate formed (9) kindly supplied by C. A. Mebus, University of was eliminated by centrifugation at 5,000 x g for 30 Nebraska, was adapted and propagated in Vero cells. min. Then, 10 ml of the supernatant fluid was mixed The virus was used at Vero cell passage level 27. with 1 g of bovine liver powder for 1 h and then Infectious bovine rhinotracheitis (IBR), strain Colo- clarified by centrifugation at 10,000 x g for 30 min. rado, and bovine enterovirus type II (BE-II) viruses, The antiserum was subsequently washed with Vero both adapted on Vero cells, and calf rotavirus supplied cells for 30 min, recovered, tested, and then used in by C.A. Mebus and adapted on MDBK cells in our serological tests. This coronavirus antiserum was shown to be free of reactivity against calf serum and laboratory were used as controls. Virus propagation. Confluent monolayers of Vero bovine tissues by immunodiffusion and CIE. Titration of neutralizing antibodies. Twofold or MDBK cells were rinsed with Dulbecco phosphatebuffered saline (pH 7.0) and inoculated with 1 ml of dilutions of the antiserum (1:10 to 1:1,280) were made infected cell extracts. After 1 h of viral adsorption at in 0.01 M phosphate-buffered saline (pH 7.2) supple37°C, the cultures were overlaid with Eagle minimal mented with 0.1% bovine serum albumin. A 0.1-ml essential medium with Earle salt and glutamine portion of a virus dilution containing 400 TCID50 was (GIBCO) and incubated at 37°C in a 5% CO2 atmos- added to 0.1 ml of appropriate serum dilutions. The phere. When maximal cytopathic effects occurred, in- virus-serum mixtures were agitated vigorously and fected cells and cell culture fluid were frozen and allowed to remain at 37°C for 1 h. Thereafter, each thawed three times and clarified and stored at -70°C. mixture was inoculated in 0.05-ml amounts into each Viral infectivity was determined as described else- of three wells of a 96-well Microtest culture tray where (S. Dea et al., manuscript submitted for publi- (Falcon Plastics, Oxnard, Calif.) plated with Vero cells. cation), and viral titers were expressed in 50% tissue The cultures were examined for evidence of cytopathic culture infective doses (TCID5o) according to the effect after 4 to 5 days postinfection at 37°C. The titer was expressed as the reciprocal of the highest neutralmethod of Reed and Muench (13). Electron microscopy. Droplets of the clarified izing serum dilution. CIE. The CIE test was performed on Kodak projecfecal contents were placed on a 200-mesh Formvarand carbon-coated grids and allowed to remain on the tor slides (8.3 by 10.2 cm) coated with 13 ml of agarose grids for 30 s. Fluids were then blotted from the grids, in Tris-barbital buffer. After the gel had become firm and a droplet of distilled water was added for 10 s and at room temperature, two parallel rows of wells (4.0 blotted off. The specimen was stained for 10 s with 3% mm in diameter) were punched 1 cm center to center. phosphotungstate acid (pH 7.0). The excess stain was Anodal wells were filled with 15 ,ll of antiserum, and blotted off. The grids were dried and then examined cathodal wells were filled with 15 tll of antigen. The with a Philipps 300 electron microscope operating at antigens used were the clarified, infected cell culture fluids or intestinal contents obtained as described an accelerating potential of 80 kV. Preparation of coronavirus antigen. Approxi- above and diluted in Tris-barbital buffer of the same mately 300 ml of NCDC-infected Vero cell culture molarity as that used to make up the agarose. Telfa fluids were frozen and thawed three times and centri- nonadherent strips (Kendall, Chicago, Ill.) (4 by 10 fuged at 5,000 x g for 30 min at 4°C to remove coarse cm) were used as wicks. Directly after electrophoresis, debris. The clarified suspension was then concentrated the slides were washed overnight at 0.85% NaCl and to a volume of 5 ml by ultrafiltration through an XM- then immersed in 1% tannic acid. Thereafter, the 100A Diaflo membrane. The concentrate was centri- agarose slides were examined with a high-beam illufuged at 78,600 x g for 2 h at 4°C. The supernatant minator for the presence of a precipitin line between fluid was discarded, and the pellet containing the virus antigen and antiserum wells. particles was dispersed by sonication at 20 kilocycles RESULTS per s for 30 s in 0.5 ml of 0.05 M Tris-hydrochloride buffer (pH 7.4). The viral suspension was then diluted Conditions for CIE. Some parameters of the in the same buffer up to 250 ,ug of protein per ml as test were studied by using cell culture fluid CIE determined by optical density at 280 nm and used for
immunization. Preparation of antiserum. Two New Zealand
Albino rabbits (2 kg) were used for immunization. They were inoculated intramuscularly at four sites (0.5 ml per site) and in the foodpads (0.5 ml) with a mixture of equal volumes of NCDC antigen and complete Freund adjuvant. After 20 days, the rabbits received an intravenous dose of 1.0 ml of antigen along with intramuscular injections (0.50 ml per site) consisting of a mixture of 1.0 ml of antigen and 1.0 ml of incomplete Freund adjuvant. On days 35 and 49, they were reinoculated intramuscularly as described previously.
from NCDC-infected Vero cells as the source of virus. A visible double precipitin line was obtained between antibody and antigen wells when we used as antibody the anti-NCDC serum supplied by C.A. Mebus and the rabbit anti-NCDC serum produced in our laboratory (Fig. 1). Both antisera possessed a neutralizing titer of 320. No precipitin lines were observed with our antiNCDC serum against IBR, Nebraska calf rotavirus, and BE-II viruses.
Electrophoresis
was conducted at
200, 150,
242
DEA, ROY, AND BEGIN
J. CLIN. MICROBIOL.
passage 27 of NCDC virus in Vero cells against our antiserum. Precipitin lines were noted with antigen dilutions up to 1:1,000. The endpoint dilution corresponded to 10: TCIDro per ml. No precipitin lines were observed with undiluted coronavirus suspensions when antiserum was diluted more than 1:4. Comparison of EM and CIE for detecting coronavirus. Typical coronavirus particles with an average diameter of 120 nm and surrounded by bulbous or petal-shaped projections 13 to 17 nm long were visualized by EM in 16 (33%) of 48 diarrheal intestinal contents and in 4 (24%) of 17 normal intestinal contents (Fig. 2). As shown in Table 1, nine diarrheic calves were singly infected by coronavirus, whereas three others had only rotaviruses. Five calves showed mixed infection by coronavirus and rotavirus, one calf showed infection with coronavirus and reovirus, and another one showed infection with coronavirus and picornavirus. A total of 29 diarrheic calves were free of any virus by this technique. Four and two of the healthy calves were singly infected with coronavirus or with picornavirus, respectively. The other 11 calves were free of FIG. 1. (Top) CIE precipitin lines observed with any viral particles as examined by EM. As shown in Table 1, 21 (44%) of 48 clarified the cell culture fluid of passage 27 of NCDC on Vero cells and anti-NCDC (rabbit) after electrophoresis at 150 V for 90 min and staining with 1% tannic acid for 15 min. (A, B) Bovine coronavirus. (C) Nebraska calf rotavirus, passage 7 on MDBK cells. (D) BE-II virus, passage 5 on Vero cells. (E) IBR virus, passage 5 on Vero cells. (F) Nebraska coronavirus antiserum of C.A. Mebus. (G-J) Nebraska coronavirus antiserum produced in our laboratory. (Bottom) Magnification showing clearly the double-line pattern.
and 100 V for 60, 90, and 120 min at 4°C by using 0.05 M Tris-barbital buffer in the electrophoresis chamber and 0.025 M in the agarose. The best results were noted when electrophoresis was conducted at 150 V for 90 min. A buffer of pH 8.6 was shown to give better reactions than those of pH 7.0, 7.5, and 8.0. A more discernible precipitin line was obtained if antiserum and antigen wells were separated by 1 cm instead of 0.7 cm. A 1% agarose concentration gave better results than did concentrations of 0.7 and 0.5%. Very weak precipitin lines were observed if the agarose slides were not stained. A clear-cut reaction was obtained after staining the slides for 15 min in 1% tannic acid after an overnight wash in normal saline following electrophoresis. A longer exposure to tannic acid did not amplify the precipitin lines. Sensitivity of CIE. CIE was conducted with 10-fold dilutions (1:10 to 1:10,000) in 0.025 M electrophoresis buffer of the cell culture fluid of
.-j.
FIG. 2. Typical coronavirus particles observed by negative staining EM in clarified intestinal content of a diarrheic calf x288,000. Bar = 40 nm.
VOL. 10, 1979
CIE FOR DETECTION OF CALF CORONAVIRUS
243
intestinal contents of diarrheic calves and 4 TABLE 1. Comparison of EM and CIE for detection of coronavirus in intestinal contents of diarrheic (24%) of 17 intestinal contents from normal and normal calves calves contained coronavirus antigens as deNormal calves tected by CIE. No antigenic reaction occurred Diarrheic calves with the intestinal contents shown to contain CIE CIE___ Virus seen by only rotavirus or only picornavirus by EM. More EM Virus seen by EM than half of the samples that were positive for coronaviruses by this technique were positive Coronavirus (9)' 1 3 Coronavirus 0 9 only after staining with 1% tannic acid as de(4)a 2 0 scribed above. Picornavirus3 0 Rotavirus (3) like (2) As reported in Table 1, CIE was positive on 1 10 0 No virus ob5 and all intestinal contents shown to contain corona- Coronavirus served (11)h rotavirus (5) virus by EM with the exception of one diarrheic Coronavirus 1 0 and and one normal calf. Coronavirus antigens were reovirus (1) 1 0 detected by CIE in 6 of the 29 intestinal contents Coronavirus and (1) from diarrheic calves in which no viral agents Nopicornavirus 6 23 virus observed were seen by EM. From the 11 normal intestinal (29)b contents that were free of any virus particles as in parentheses indicate the number of samples determined by EM, in only one instance did we withaNumbers the indicated virus(es). observe a positive reaction by CIE. bNumber of samples in which no viruses could be observed These results indicate that CIE could detect by EM. more coronavirus-positive intestinal contents than could EM. However, a negative result obthe healthy calves were infected with coronavitained by CIE must be verified by EM. Based on the combined use of EM and CIE, rus but had not yet developed diarrhea. More46% (22/48) of the diarrheic calves and 30% (5/ over, no pathological lesions were seen in these 17) of the healthy calves studied had coronavi- normal calves after a postmortem examination (M. Morin et al., Proceedings of the 2nd Symruses in their intestinal contents. posium on Neonatal Diarrhea, Saskatoon, 1979, in press). Samples collected from other regions DISCUSSION over a long period of time are needed in Two precipitin lines were observed in the ma- to Quebec and to generalize this finding. If this confirm jority of the positive samples, indicating that, as is correct, the significance of a high percentage with other members of the coronavirus group, of asymptomatic calves containing such high one than the bovine coronavirus possesses more needs to be elucidated to of coronavirus levels precipitating antigen (1, 2, 4, 10, 17). of the correct the picture obtain As applied to intestinal contents, the reaction and the mechanism of the onset ofepidemiology the diarrhea seemed specific for coronavirus since no immu- associated with coronaviruses. It would appear noprecipitate could be detected with the con- necessary to identify the factors modifying the tents shown to contain viruses other than coroof coronavirus infections. navirus by EM. It is worth stressing the fact that clinical effects two intestinal contents were negative by CIE ACKNOWLEDGMENTS although they were positive by EM. This fact The technical assistance of Francine Breton is gratefully could be explained by a technical failure of CIE acknowledged. This study was supported by the Conseil de Recherches et or observer error by EM. However, it could also Services Agricoles du Quebec (Canada) grant no. MMV 75suggest the existence of more than one antigenic de 604. variant among bovine coronaviruses. LITERATURE CITED CIE provided a reliable, simple, and fast method for large-scale screening of intestinal 1. Bohac, J., and J. B. Derbyshire. 1975. The demonstracontents for coronavirus. However, before montion of transmissible gastroenteritis viral antigens by immunoelectrophoresis and counterimmunoelectroospecific antisera to all known bovine coronaviphoresis. Can. J. Microbiol. 21:750-753. ruses are available, it is advisable to further 2. Bohac, J., J. B. Derbyshire, and J. Thorsen. 1975. EM. examine the CIE-negative samples by The detection of transmissible gastroenteritis viral antigens by imnimunodiffusion. Can. J. Comp. Med. 39:67Unexpectedly, coronavirus particles and anti75. gens were detected by EM and/or CIE in 30% of 3. Chasey, D., and M. Lucas. 1977. A bovine coronavirus the healthy calves. To our knowledge no asympidentified by thin-section electron microscopy. Vet. Rec. tomatic coronavirus shedding has been reported 100:530-531. in this species. However, since all calves were 4. Hierholzer, J. C., E. L. Palmer, S. G. Whitfield, H. S. Kaye, and W. R. Dowdle. 1972. Protein composition subjected to necropsy, we do not know whether
244
DEA, ROY, AND BEGIN
of coronavirus OC43. Virology 48:516-527. 5. House, J. A. 1978. Economics of rotavirus and other neonatal disease agents of animals. J. Am. Vet. Med. Assoc. 173:573-576. 6. Inaba, Y., K. Sato, H. Kurogi, E. Takahashi, Y. Ito, T. Omori, Y. Goto, and M. Matumoto. 1976. Replication of bovine coronavirus in cell line BEK-1 culture. Arch. Virol. 50:339-342. 7. Marsolais, G., R. Assaf, C. Montpetit, and P. Marois. 1978. Diagnosis of viral agents associated with neonatal calf diarrhea. Can. J. Comp. Med. 42:168-171. 8. Mebus, C. A., L. E. Newman, and E. L. Stair. 1975. Scanning electron, light, and immunofluorescent microscopy of intestine of gnotobiotic calf infected with calf diarrheal coronavirus. Am. J. Vet. Res. 36:17191725. 9. Mebus, C. A., E. L. Stair, M. B. Rhodes, and M. J. Twickhaus. 1973. Neonatal calf diarrhea: propagation, attenuation and characteristics of a coronavirus-like agent. Am. J. Vet. Res. 34:145-150. 10. Mengeling, W. L. 1972. Precipitating antigens of haemagglutinating encephalitis virus demonstrated by agar gel immunodiffusion. Am. J. Vet. Res. 33:1527-1535. 11. Morin, M., P. Lamothe, A. Gagnon, and R. Malo. 1974. A case of viral neonatal calf diarrhea in a Quebec dairy herd. Can. J. Comp. Med. 38:236-243. 12. Norden Laboratories. 1973. Laboratory methods for
detecting calf diarrhea virus. Norden Laboratories, Lin-
J. CLIN. MICROBIOL. coln, Neb. 13. Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty percent end points. Am. J. Hyg. 27:493497. 14. Scherrer, R., J. Cohen, R. L'Haridon, C. Feynerol, and J. C. Fayet. 1976. Reovirus-like agent (rotavirus) associated with neonatal calf gastroenteritis in France. Ann. Rech. Vet. 7:25-31. 15. Sharpee, R. L., C. A. Mebus, and E. P. Bass. 1976. Characterization of a calf diarrheal coronavirus. Am. J. Vet. Res. 37:1031-1041. 16. Stair, E. L., M. B. Rhodes, R. G. White, and C. A. Mebus. 1972. Neonatal calf diarrhea: purification and electron microscopy of a coronavirus-like agent. Am. J. Vet. Res. 33:1141-1156. 17. Tevethia, S. S., and C. H. Cunningham. 1968. Antigenic characterization of infectious bronchitis virus. J. Immunol. 100:793-798. 18. Woode, G. N., J. C. Bridger, G. Hall, and M. J. Dennis. 1974. The isolation of a reovirus-like agent associated with diarrhoea in colostrum deprived calves in Great Britain. Res. Vet. Sci. 16:102-105. 19. Zygraich, N., A. M. Georges, and E. Vascoboinic. 1975. Etiologie des diarrhees neonatales du veau. Resultats d'une enquete serologique relative aux virus reo-like et corona dans la population bovine belge. Ann. Med. Vet. 119:105-113.
