key: cord-291707-dzmvjh7j authors: Tupper, G. T.; Evermann, J. F.; Russell, R. G.; Thouless, M. E. title: Antigenic and biological diversity of feline coronaviruses: feline infectious peritonitis and feline enteritis virus date: 1987 journal: Arch Virol DOI: 10.1007/bf01310988 sha: doc_id: 291707 cord_uid: dzmvjh7j Antigenically related feline coronaviruses cause two distinct disease manifestations in infected cats. The diseases are feline infectious peritonitis (FIP), in which the virus is widely disseminated, and feline enteric coronavirus (FECV), a mild disease in which the virus is usually limited to the villi. These two viruses were found to differ in their growth in cell culture. FIPV grows to higher titer, forms larger plaques and switches off host cell protein synthesis more effectively than FECV. Cross neutralization studies showed antigenic differences between the strains. There also appeared to be a difference in the nucleoprotein molecular weight of the viruses causing these two different disease syndromes. Antigenically related i~line coronaviruses (FCV) cause two different disease manifestations in cats (12, 14, 15, 17) . The first is known as feline infectious peritonitis (FIP), and is characterized by peritonitis and/or pleuritis with occasional central nervous system and ocular involvement (7) . The inflammatory infiltrate consists of lymphoeytes, plasma cells, and macrophages resulting in either nonsuppurative or a granulomatous inflammarion. The second disease is caused by feline enteric eoronavirus (FECV) and is a subclinical or mild enteric infection in which, lesions are located in the upper third of the villi of the small intestine (16) . Although these viruses are associated with very different disease syndromes, evidence has suggested that the viruses are antigenicMly and biologieMly similar (12) . It was reported that both virus strains produce relatively large plaques in cell culture and grew to fairly high titers (1) . In addition, their polypeptides appeared to be the same molecular weight by immunoblotting (1) . Although the strains eausing peritonitis and enteritis in eats were antigenieally and biologicMly similar; neither protected against infection with the other (17) . Therefore, we investigated the antigenic biological and biochemical properties of these viruses in more detail. The viruses were grown in Crandell Feline Kidney (CrFK) cells. The cells were mycoplasma free by the method of Kenny (9) . Cells were grown at 37 ° C in Eagle's minimum essential medium (Auto-Pow, Flow Lab., McClean, VA) supplemented with l0 percent heat inactivated fetal calf serum, 2 mM L-Glutamine, 12 my~ sodium bicarbonate, 10 m:~ HEPES, 10 tlg/ml sm~ptomycin, and 100 Units/ml of penicillin (MEM-10). The calf serum was reduced to 1 percent for virus propagation (MEM-1). The FIP strains WSU 79-I146 and NOR 15, and the FECV strains WSU 79-1683 isolates of FCV were studied. The viruses were cloned by endpoint titration 3 times in microtiter plates. The isolation and in vivo pathogenicity of these strains has been previously reported (3, 12, 17) . Virus titers were measured by a plaque assay or a TCIDs0 endpoint. The plaque assay was done in 60 mm tissue culture petri dishes (Corning, NY). Confluent monolayers of CrFK cells were inoculated with 200 ~I of a tenfold virus dilution. After adsorption for 1 hour at 37 ° C an overlay media of 4 ml of warmed 0.7 percent, carboxyraethyl cellulose in MEM-1 was added. Monolayers were fixed with formalin and stained with 1 percent crystal violet five days post infection (PI). Plaques were counted using a light box and a plaque counting device (Seientifiea). The TCIDs0 assay tbr virus infectivity was done in 96 well microtiter plates (FALCON) by infection of 6 wells with 25 txl of a tenfold virus dilution. The monolayers were fixed and stained after 5 days. Wells were examined for cybopathie effects (CPE) and the formula of Reed and Muench was used to calculate the TCID~0 (18) . Photographs of infected monolayers in 60 mm plates containing approximately 80 distinct plaques were enlarged 11.5 times to determine the average plaque size. The diameters of the plaques were measured with a ruler. CrFK cells were grown in roller bottles at 37 ° C and infected with the feline coronavirus strains at a multiplicity of infection (MOI) of 1.0 to 0.1. The supernatant was harvested and pooled at 46 hours PI. All subsequent steps were carried out at 4 ° C similar to the procedure used by Schmidt and Kenny (22) . The suspension was clarified by centrifhgation at 4000 × g for 30 minutes. Polyethylene glycol 8000 was added to make a final concentration of 7 percent (w/v). After a 12 hour incubation the precipitate was collected by clarification at 13,200 × g for 40 minutes. The pellets were resuspended in HEPES BuS fer (20.0 rn.~( HEPES--0.15 ~ NaC1--5.0 mM EDTA-Na 2, pH 7.0) layered onto a 15 and 55 percent (w/v) sucrose step gradient, and centrifuged at 120,000 x g for 2.5 hours in a Beckman SW 28 rotor. The virus bands at the interface were collected, diluted with an equal volume of HEPES buffer and pelleted by centrifugation at 243,000 x g for 20 minutes in a Beckman SW 50.1. The virus was resuspended in the HEPES buffer for inoculation into rabbits to obtain hyperimmune sera. Further purification by isopycnic banding was undertaken to obtain purified virus for PAGE. The virus bands were concentrated on a 1 ml cushion of 1.20 gm/ml renografin (Squibb) in a SW 50.1 rotor. The concentrated virus was again diluted in the ttEPES buffer and layered onto a continuous 1.10 to 1.25 (gm/ml) renografin gradient for isopyenie banding using a Beckman type 65 rotor at 176,000 × g for 18 hours. The visible bands were collected (density 1.18 gm/ml) and then pelleted using a Beckman SW 50.1 rotor. Virus pellets were resuspended in HEPES buffer and frozen until further use. Viral infectivity was measured by TCID.50 and the protein concentrations were determined by the Lowry method (11). Rabbits were hyperimmunized with FIPV 79-1146 or FECV 79-1683 purified by sucrose rate zonal centrifugation as described above. The rabbit immunization schedule was an intramuscular injection of 1 × t08 plaque forming units (PFU)/ml in Freund's incomplete adjuvant, followed by four small intravenous boosts using doses of 0.1 to 0.4 ml of inocula containing 1 × l0 s PFU/ml (8) . The rabbits were bled l0 days after the last boost. Virus neutralization was carried out in 96 well microtiter plates (20) . Virus, (100 TCIDs0 virus in 25 ~1), was added to 26 ~1 of serial twofold dilutions of rabbit hyperimmune sera against FIPV 79-1146 or FECV 79-1683. The virus-serum mixture was incubated at room temperature for 1 hour after which 1.5 × 106 CrFK cells in 50 ~! MEM-1 was added. Plates were placed in a 2.5 percent C02 incubator at 37 ° C for 24 to 48 hours. Monolayers were fixed with formalin, stained with 1 percent crystal violet stain. PAGE was performed in 1.5 mm thick slab gels by the method of Laemmli (10) . Purified virus samples (30 l~g) in buffer (0.1 ~I Tris-HCt, pig 7.0, 2 percent mercaptoethanol, 2 percent sodium dodecyl sulfate [SDS] and l0 percent glycerol plus bromophenol blue) were boiled for 2 minutes and placed into each well. Proteins were migrated through a stacking gel containing 3.5 percent polyacrylamide and resolved using a 5 to 18 percent continuous polyacrylamide gradient gel. A modified silver stain was used to stain protein (13) . The molecular weight of virus structural proteins was determined by using molecular weight standards (Sigma, St. Louis, MO). The procedure was similar to that used to label rotavirus polypeptides (26) . Monolayers were prepared in 6 well (Costar, Cambridge, MA) tissue culture plates by adding 2 × 106 cells/well. The monolayers were washed once and infected with a MOI of 5. Virus was allowed to adsorb for 1 hour and then 5 t~g actinomycin D (Sigma) in 1 ml MEM-1 was added. The infected cells were incubated for the desired time, washed three times and i ml of 10 t~Ci/ml 35S-methionine in methionine free MEM~I was added. The plates were incubated for 1 hour at 37 ° C at which time the media was removed and the monolayers washed once. Disruption mixture (50 m~ Tris-HC1, pH 7.0, 5 percent mercaptoethanol, 2 percent SDS, and 3 percent glycerol with bromophenol blue) in a 0.5 ml volume at 80 ° C was added. Samples were treated with an ultrasonic probe and kept at -70 ° C until used. G.T. qSapper et al.: Cell controls of uninfected cells were treated in the same manner. An amount of sample containing 7000 to 10,000 counts was used per lane for PAGE. Some of the gels were stained with 0.5 percent Coomassie blue in 45 percent isopropano110 percent acetic acid and destained in 45 percent isopropanol I0 percent acetic acid. Dried gels were exposed to Kodak film for autoradiography. Virus titers were reproducible throughout the study with FIPV strains 79-1146 and NOR 15 having virus titers of -1 × l07 PFU/ml and FECV 79-1683 having a lower titer of N3 × 106 PFU/ml. The cytopathic effect produced by the three feline coronavirus isolates was characterized by syncytial formation. The plaques of the 79-1146 and NOR 15 strains measured 1.12 mm with a + 0.23 mm standard deviation while the 79-1683 strain produced smaller plaques of 0.58 mm with a _+ 0.23 mm standard deviation ( Fig. 1 ) in 5 days in CrFK. The plaque size of the FECV 79-1683 strain were significantly smaller than those of the other two FIPV strains by the T test (P = 0.01). The 79-1146 and the NOR 15 FIP isolates were distinct from the 79-1683 FECV isolate by virus neutralization using homologous and heterologous antisera prepared in rabbits against 79-1146 and 79-1683 (Table 1 ). The neutralizing titers for FIPV 79-I146 and NOR 15 were similar irrespective of the antiserum used. By comparison, the FECV 79-1683 strain had a 32 fold difference in neutralization titer using antisera against 79-1146 and a 8 fold difference when antisera against 79-1683 was used. These results suggest that 79-1146 and NOI~ 15 strains are antigenically similar by virus neutralization and are distinct from the 79-1683 strain of feline eoronavirus. Each of the three virus strains had the structural protein profile characteristic of that reported for the eoronavirus family (24) . The peplomer (P) surface protein band measured 205,000 molecular weight (MW). The mere- (Figs. 2 and 3) . This indicated the differences were consistent and were not due to maturation artitSct. There was a cell protein band just above the nucleoprotein of the FECV 79-1683 of the radiolabelled virus. This band was not present in the purified virus preparation or in radiolabelled 79-1146 where host cell synthesis was switched off. Virus proteins appeared at 6 hours PI and continued to be synthesized throughout the 14 hour study period. In the presence of actinomyein D, FIPV 79-1146 shutoff protein synthesis of CrFK cells at 8 to 10 hours post infection, NOR 15 a,t 10 to 12 hours. Celtula, r protein synthesis was reduced at 8 hours PI by FECV 79-1683 but was not completely shutoff (Fig. 4) . The feline coronavirus strains did not shutoff eellula, r protein synthesis in the absence of actinomyein D. The 79-1146, NOR 15, and 79-1683 isolates of feline coronavirus have previously been reported to be similar by their relatively high titer in cell culture, large plaque size, and by indirect immunofluorescenee with antibody-to canine coronavirus (3) . This study showed similarities between the 79-1146 and the NOR 15 isolates of FIPV and difference with the 79-1683 FECV isolate. The FIPV strains produced larger plaques in CrFK cells and half a log higher titer of virus than the FECV strain. Use of a different cell line (fcwL-4 cells) and time that the plaques were left to develop may have accounted for the fact that Boyle et al. (1) did not find a difference. However, differences in plaque size and virus titers have been reported for human, murine and porcine eoronaviruses grown in different cell lines (1, 2, 4, 5) . The FIPV strains could be distinguished from the FECV strain by cross neutralization. The low level of cross reactivity with heterologous serum compared to homologous serum suggests considerable antigenic variation between the two groups. However the sera were not very high titer despite immunization of rabbits with biologieMly cloned purified virus, perhaps due to the fragility of the peplomer protein which induces neutralizing antibody (6) . No differences were seen in the molecular weights of the peplomer but this might be undetectable in such a high molecular weight protein where the relevant epitope may be only a very small part of the entity. Previous studies with convalescent sera from naturally and experimentally infected cats did not show these differences in vitro (16, 17) but did in vivo where FIPV strains did not protect against FECV infection and prior FECV infection even seemed to enhance the pathogenicity of FIPV. The molecular weight of the nucleoproteins differed by about 1000, with the FECV strain nueleoprotein being smaller than that of the 2 FIPV strains. This was determined by co-running purified virus and pulse radiolabelled polypeptides of both strains. This is considered to be a more sensitive technique than immunoblotting used by Boyle et al. (1) , and may explain the fact that they did not observe this difference. The significance of the dit; ferenee in the molecular weight and any possible relationship to pathogenicity is unknown at the present time. However, differences in molecular weight of the nueleoproteins have been found in 5 strains of murine coronaviruses and a high degree of homology was found between them by hybridization kinetics and peptide mapping (2) . In vitro infection with the FIPV strains reduced the production of host cell proteins. This effect was enhanced by actinomycin D. The FECV strain did not switch off host cell synthesis even in the presence of aetinomyein D. Host cell protein synthesis was also shut offby infection with murine coronavirus and different strains vary in the extent to which they do it, (25) . Highly lyric strains of MCV synthesize M1 3 structurM polypeptides synchronously whereas the nucleoprotein appears earlier than the other two polypeptides in less lytic infections (19, 23) . In this study, all 3 st, ruetural polypeptides appeared synchronously in cells infected with FIPV or FECV strains. It is interesting that the more vigorously growing strains which produce more varied and severe disease (FIPV) shut off host cell synthesis more effectively than the less pathogenic FECV strain. Despite the aforementioned differences, there is currently no evidence that the in vitro differences between the virus strains studied are associated with or linked to the differences in the pathogenicity of FIPV and FECV strains in vivo. Plaque assay, poiypeptide composition, and immunochemistry of feline infectious peritonitis virus and feline enteric coronavirus isolates RNA and polypeptide homology among murine coronavirus Characterization of a feline infectious peritonitis vires isolate Replication ofmurine coronaviruses in somatic cell hybrids folTned between a mouse fibroblast cell line and either a rat schwannoma line or a rat glioina line Defective replication of porcine transmissible gastroenteritis virus in a continuous cell line Isolation of subviral components from transmissible gastroenteritis virus The virology and pathogenesis of feline infectious peritonitis hnmunogenicity of myeoplasma pneumoniae Contamination of mammalian cells in culture with myeoptasmata. In: Fogh J (ed) Contamination in tissue culture Cleavage of structural proteins during the assembly of the head of bacteriophage T 4 Protein measurement with the folin phenol reagent Isolation of feline coronavirus from two eats with diverse disease manifestations A simplified uttrasensitive silver strain for detecting proteins in potyaerylamide gels Feline infectious peritonitis and feline enteric coronavirus intbction II. Feline infectious peritonitis Pathogenic differences between various feline eoronavirus isolates An enteric eoronavirus infection and its relationship to feline infectious peritonitis Pathogenicity studies of feline eoronavirus isolates 79-1146 and 79-1683 Muench tt (I938) A simple method for estimating fifty percent endpoints Viral protein synthesis in mouse hepatitis virus strain A 59-infected cells: effect oftunieamyein Application of a microtechnique to viral serological investigation Plaque assay and improved yield of human coronaviruses in a human rhabdomyosarcoma cell line Immunogenicity and antigenicity of human coronaviruses 229 E and OC43 Coronavirus JHM: intracellular protein synthesis The structure and replication of coronaviruses The molecular biology of coronaviruses Thouless ME (1979) I~otavirus polypeptides ter Meulen V (1982) The biology and pathogenesis of coronaviruses The authors wish to expi~ss their appreciation to A. J. MeKeirnan tbr assistance with feline eoronavirus strains. t~eceived August 21, 1986