Two Related Strains of Feline Infectious Peritonitis Virus Isolated from Immunocompromised Cats Infected with a Feline Enteric Coronavirus

JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 1996, p. 3180 3184 Vol. 34, No. 12 0095-1137/96/$04.00 0 Copyright 1996, American Society for Microbiology Two Related Strains of Feline Infectious Peritonitis Virus Isolated from Immunocompromised Cats Infected with a Feline Enteric Coronavirus AMY M. POLAND, 1 HARRY VENNEMA, 1 JANET E. FOLEY, 1 AND NIELS C. PEDERSEN 1,2 * Center for Companion Animal Health 1 and Department of Medicine and Epidemiology, 2 School of Veterinary Medicine, University of California, Davis, California 95616 Received 18 March 1996/Returned for modification 10 July 1996/Accepted 30 July 1996 Two groups of cats were experimentally infected orally with the cat-passaged RM strain of feline enteric coronavirus (FECV-RM). One group of cats (n 19) had been chronically infected with feline immunodeficiency virus (FIV) for over 6 years, while a second control group (n 20) consisted of FIV-naive siblings. Fecal virus shedding of FECV occurred in both groups starting on day 3 postinfection, nearly ceased by 4 weeks in FIV-uninfected cats, but remained at high levels in FIV-infected animals. FIV-infected cats shed virus for a longer period of time and at levels 10 to 100 times greater than those for FIV-uninfected cats. The coronavirus antibody response of the FIV-infected cats was delayed and of reduced titer compared with that of the FIV-uninfected animals. Cats in both groups remained asymptomatic for the first two months following FECV-RM infection; however, 8 to 10 weeks postinfection two cats in the FIV-infected group developed feline infectious peritonitis (FIP). The FIP viruses (designated FIPV-UCD9 and -UCD10) isolated from these two cats had almost complete genetic homology to each other and to the infecting FECV-RM. However, unlike FECV- RM, they readily induced FIP when inoculated intraperitoneally into specific-pathogen-free cats. This study confirms that FIPVs are frequently and rapidly arising mutants of FECV. Immunosuppression caused by chronic FIV infection may have enhanced the creation and selection of FIPV mutants by increasing the rate of FECV replication in the bowel and inhibiting the host s ability to combat the mutant viruses once they occurred. Pedersen and Floyd (17) first postulated that feline infectious peritonitis virus (FIPV) was a simple mutation of feline enteric coronavirus (FECV) that occurs commonly during the course of primary FECV replication. Evidence supporting this hypothesis has been compelling but largely circumstantial (10). FIPVs and FECVs always coexist in an environment and are antigenically, serologically, and morphologically identical (13, 14). Individual FIPV strains also show much more genetic homology to FECVs obtained from the same environment than to FIPV or FECV strains isolated from geographically distant catteries (23). The most compelling evidence for the mutational origin of FIPVs from FECVs comes from an accidental outbreak of FECV infection in a large barrier-maintained specific-pathogen-free breeding colony, in which almost 1,000 cats seroconverted without signs of disease. Over the next two years, six young cats raised in this colony developed FIP (8). The FECV isolated from this cattery, FECV-RM, produced asymptomatic seroconversion after oral or intraperitoneal inoculation into specific-pathogen-free cats, while the FIPV isolated from this cattery, designated FIPV-UCD8, faithfully reproduced FIP upon animal inoculation. FECV-RM and FIPV-UCD8 were greater than 97% homologous by genetic sequencing of the spike gene and were genetically distinct from FIPV and FECV strains isolated from cats * Corresponding author. Mailing address: School of Veterinary Medicine, Center for Companion Animal Health, University of California, Davis, CA 95616. Phone: (916) 752-7295. Fax: (916) 752-7701. Electronic mail address: ncpedersen@ucdavis.edu. Present address: Virology Division, Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, 3508 TD Utrecht, The Netherlands. in other environments (23). Unfortunately, FIPV-UCD8 was isolated almost 2 years after the primary FECV epidemic and it was therefore impossible to make a strong temporal linkage between the FECV-RM outbreak and the appearance of FIPV-UCD8. The present study was designed to document that FIPVs arise as mutations within cats undergoing FECV infection. In order to maximize the opportunity for such a mutation to occur, chronically immunocompromised cats were infected orally with FECV-RM. It was hypothesized that the immunocompromised cats would replicate FECV at much higher levels than nonimmunocompromised cats, thereby increasing the likelihood that FIP mutants would be generated. The diminished immunity would also compromise host resistance to the mutant coronavirus and thereby increase the chances of developing FIP. The appearance of FIP would be the most tangible evidence for the generation of an FIPV. Our group of immunocompromised cats were in the mid to late stages of feline immunodeficiency virus (FIV) infection. These cats had been FIV infected for almost 6 years and had significant reductions in CD4 T-cell numbers and certain T-cell-dependent immune responses but were asymptomatic (19). Although outwardly healthy, they responded to feline calicivirus and herpesvirus infections with more severe disease, higher levels of virus replication, and diminished specific primary antibody responses (20, 21). Thus, a similar response to an FECV challenge was anticipated. MATERIALS AND METHODS Experimental animals. Thirty-nine cats, 7 years of age, were used in the study; 19 of these cats were FIV infected, and 20 were FIV naive. The conditions of FIV infection and monitoring have been previously reported (19). Both groups of cats had been sequentially exposed to Haemobartonella felis, feline calicivirus, Chla- 3180

VOL. 34, 1996 TWO RELATED STRAINS OF FIPV FROM CATS WITH FECV 3181 FIG. 1. Feline coronavirus antibody responses of FIV-negative and -positive cats infected with FECV-RM. mydia psittaci, feline herpesvirus type I, Toxoplasma gondii, and Bartonella henselae over a 5-year period. All 39 cats were negative for feline coronavirus antibodies at the start of the study. Experimental cats were housed with three to four animals per cage, with three cages in one room, and with FIV-negative or -positive animals in separate rooms. Animal holding facilities were barrier contained, and caretakers changed disposable outer garments, shoe covers, and gloves before entering a new room. Animal care was under the supervision of the Animal Resources Services, University of California, Davis. Virus inocula. Cats were orally infected with a single dose (1 ml) of fecal extract containing FECV-RM. The origin of FECV-RM and its characterization and preparation of infectious fecal extracts have been previously reported (8). This virus has never caused FIP directly, either after oral or intraperitoneal inoculation. Cats recovering from FECV-RM infection developed antibodyenhanced FIP when they were challenge exposed later with related strains of FIPV, such as FIPV-UCD8. The inoculum for FIPV-UCD9 consisted of 1 ml of whole pleural fluid collected from a cat with the effusive form of FIP. The fluid was aliquoted in 2-ml vials and stored at 70 C. The inoculum for FIPV-UCD10 was made by grinding affected tissues from a cat with noneffusive FIP in a mortar and pestle; the mortar contained equal volumes of sterile silica sand, tissue, and Hank s buffered saline solution (Gibco BRL, Gaithersburg, Md.). After the tissue was thoroughly ground, the mixture was centrifuged at 1,000 g at room temperature for 20 min to remove all sand and gross tissue debris. The resultant supernatant was centrifuged at 10,000 g at 4 C for 20 min to pellet subcellular particulate material but not virus. The final supernatant was aliquoted in 2-ml vials and frozen at 70 C. The 50% tissue culture infective doses of both inocula were not determined, because none of these virus strains would grow in tissue culture. However, the infectious doses used of FIPV-UCD9 and FIPV-UCD10 were near 1, because only one of two cats in each inoculation seroconverted and became ill. In the FECV-RM inoculations, all cats became infected, indicating a 50% tissue culture infective dose of 1. Preparation of feces for PCR assay. Feces were collected from a single litter box placed in each animal cage, pooled, and frozen at 20 C. Each cage contained from three to four cats from the same experimental group; therefore, fecal virus levels and duration of shedding were representative of the cats in a cage and not individual animals. Feces were collected daily for the first two weeks, and every other day from week 2 to 4. Frozen feces were thawed and mixed with equal parts (wt/vol) of phosphatebuffered saline (PBS; ph 7.4), shaken vigorously, and then centrifuged at 3,000 g for 20 min. RNA was extracted from the supernatant by the method of Boom et al. (1), with modifications (3), by using 20 l of silica mixed with 100 l of fecal extract and 900 l of L6, a guanidine-based lysis buffer. The RNA was eluted in 25 l of diethyl pyrocarbonate-treated 1 mm EDTA in sterile distilled water. RT-PCR for FECV-RM. PCR primers were chosen from a relatively conserved region of the coronavirus 7b gene by amplifying a 355-bp fragment. The primers used were 177 (reverse primer; 5 CAC TTA CAA TAT AGA AAT TAT CTA C3 ) and C202 (forward primer; 5 GGG TTT TCC TGC TAT ACA TTG 3 ). Reverse transcription was performed by standard techniques with MMLV reverse transcriptase (RT; Promega Corp., Madison, Wis.). RNA (5 l) was mixed with 5 pmol of primer, denatured at 95 C for 5 min, and then quick-cooled on wet ice. Four microliters of the RT mixture was added to the RNA for a final concentration of 50 mm Tris-HCl, 75 mm KCl, 3 mm MgCl 2, 10 mm dithiothreitol, 1 mm each deoxynucleoside triphosphate (dntp), 10 U of RNasin, and 50 U of MMLV RT. This mixture was incubated for 1hat37 C. At the end of the hour, double-stranded DNA contaminants were cleaved by incubating cdna and the PCR mixture for 1 h with 2.5 U of Sau3A (5) and PCR was performed with Taq DNA polymerase. A 90- l PCR mixture containing a final concentration of 50 mm KCl, 10 mm Tris-HCl, 0.1% Triton X-100, 2 mm MgCl 2, 0.2 mm each dntp, 20 pmol of each forward and reverse primer, and 2UofTaq DNA polymerase was added to the cdna. PCR conditions were as follows: 95 C for 2 min; 35 cycles of 95 C for 30 s, 55 C for 1 min, and 72 C for 2 min; 7 min at 72 C; and then cooling to 4 C. Positive samples were visualized on a 1% agarose gel stained with ethidium bromide. Quantitative and competitive (QC) RT-PCR. A naturally occurring deletion mutation of 85 bp in the same region of the 7b gene as the diagnostic PCR was identified in a cloned plasmid DNA; except for this deletion, the nucleotide sequence was identical to the wild-type sequence (Fig. 1). The deletion mutant clone was digested with EcoRI and PstI, recloned into the pgem-3z vector (Promega), linearized with HindIII, and purified. The DNA was transcribed to RNA by Promega s standard T7 RNA polymerase protocol. This RNA was extracted with acid phenol to remove DNA, precipitated, and resuspended in diethyl pyrocarbonate-treated 1 mm EDTA in water. The quantity of total RNA in the preparation was determined with a Beckman DU-70 spectrophotometer reading at 260 nm. For the QC-RT-PCR, serial 10-fold dilutions of competitor were made and 5 l of each dilution was added to 100 l of raw fecal extract at the RNA isolation stage to control for variability from the point of RNA isolation and afterward. Effective competition between the wild type and mutant competitor occurred with the 10 3,10 4, and 10 5 dilutions. Reverse transcription and PCR were carried out in the presence of the competitor in the same way as described above for the regular RT-PCR. The amount of native viral RNA in the sample corresponded to the amount of competitor which yielded a PCR band with an intensity equal to or slightly greater than that of the band generated by the competitor. Cloning and sequencing. PCR fragments were cloned with a TA cloning kit (Invitrogen Corp., San Diego, Calif., or Novagen, Inc., Madison, Wis.) according to the manufacturer s instructions. The resulting plasmid DNA was sequenced with U.S. Biochemical s Sequenase kit, [ 32 P]dATP (Amersham Corp., Arlington Heights, Ill.), and standard or reverse M13 sequence primers or PCR primers. Sequence data were analyzed with the University of Wisconsin s Genetics Computer Group sequence analysis software package (4, 6) on a VAX computer. Coronavirus serology. Feline coronavirus serology was performed by an indirect immunofluorescence assay (11). The substrate was acetone-fixed FIPV- UCD1-infected Felis catus whole fetus 4 cells grown on eight-well Teflon-templated microscope slides. Serum was serially diluted in PBS at 1:25, 1:100, 1:400, 1:1,600, and 1:3,200. Twenty-five microliters of sample was placed in each well and incubated at 37 C for 45 min in a moist incubator. Slides were washed three times in PBS and blotted dry, and 30 l of pretitrated rabbit anti-cat fluoroisothiocyanate immunoglobulin G conjugate was added to each well. The slides were then incubated for an additional 45 min at 37 C. The slides were once again washed in PBS, counterstained with Evans blue, and visualized by indirect immunofluorescence microscopy. The absolute titer was the highest serum dilution that still produced detectable levels of fluorescence in foci of virus-infected cells. The titers were then translated into titer class for statistical analysis, with a negative at 1:25 designated as 0 and with positives at 1:25, 1:100, 1:400, 1:1,600, and 1:3,200 designated as 1, 2, 3, 4, and 5, respectively. Serology was performed on postinfection days 0, 7, 9, 14, 21, and 28. RESULTS Clinical signs. No signs of illness were noted in any of the 39 FECV-RM-infected cats during the initial 8-week postinoculation period. Stools remained normal, appetite was unaffected, complete blood count remained normal, and no fever was detected. Cats 121 and 123 from the FIV-infected group became ill with signs of FIP at 8 and 10 weeks following FECV-RM infection, respectively, while all other cats remained healthy. Clinical signs of FIP may include fever, malaise, dyspnea, anorexia, abdominal or pleural effusion, neutrophilia and lymphopenia, neurologic abnormalities, especially anterior uveitis, and/or palpable abdominal lesions. The FIP diagnosis was confirmed by pathology and PCR of infected tissues. Serology. FECV antibodies appeared later and were at a lower titer in FIV-infected cats than in FIV-negative cats (Fig. 1). In the FIV-negative cats, 8 of 19 cats seroconverted by 10

3182 POLAND ET AL. J. CLIN. MICROBIOL. FIG. 2. Amounts of FECV-RM RNA shed in the feces of FIV-negative and -positive cats as determined by QC-RT-PCR. days postinfection and all 19 had seroconverted by day 14. In contrast, only 1 of 18 FIV-positive cats seroconverted by day 10; only 9 cats had seroconverted by day 14. Uniform seroconversion among FIV-infected cats was not observed until the fourth week postinfection. The mean titer levels were lower in the FIV-positive cats than in the FIV-negative cats at all time points, and the peak coronavirus antibody titers were more variable, with the standard deviation sometimes exceeding the mean titer. Fecal virus shedding. Both FIV-infected and -uninfected cats became PCR positive for FECV-RM in their feces by day 3 postinoculation and remained PCR positive for the next 3 to 4 weeks. Virus shedding either ceased or reached low levels by weeks 3 to 4 in the FIV-naive group; the cats in one cage were negative at the 3- and 4-week time points, and the cats in four cages were only weakly positive by the fourth week. In contrast, FIV-infected cats in all six cages were still strongly fecal virus positive at week 4. Although the timing of virus shedding was similar in FIV-infected and -noninfected cats, fecal virus shedding was 10 to 100 times greater in the FIV-positive cats than in the FIV-negative cats (Fig. 2). By QC-RT-PCR with competitor dilutions of 10 3,10 4, and 10 5, the levels of fecal virus shedding were approximately the same in both groups of cats at day 3 (less than 2.3 pg of competitor RNA per 100 l of fecal extract). However, by days 5 to 7, the levels of fecal virus shedding were at 230 pg/100 l of fecal extract in the FIVinfected cats and approximately 2.3 pg/100 l of fecal extract in non-fiv-infected animals. Fecal virus shedding in the FIVpositive cats had decreased to approximately 23 pg/100 l of fecal extract by day 9, at which level it remained for the rest of the study. The level of fecal virus shedding in the FIV-naive cats was always about 2.3 pg/100 l of fecal extract or less. Plasma viremia. Plasma was tested for coronavirus RNA on days 0, 2, 5, 7, 9, and 12 on five cats from the FIV-infected group, including the two animals who subsequently succumbed to FIP, and three cats from the FIV-uninfected group. Neither of the two FIP-affected cats had detectable levels of virus in their plasma at days 0 to 12 and 28 postinfection (no measurements were made at the time of death, 4 to 6 weeks later). Four of six remaining cats, two from each of the FIV-infected and noninfected groups, were viremic at one of the time points that were tested. Three of the four cats tested positive on day 5, while the fourth cat was positive on day 7 postinoculation. This period of viremia corresponded with peak viral shedding in the feces. QT-RT-PCR on the positive serum samples indicated effective competition at a dilution of 10 5 in one cat and less than 10 6 for the other three cats. Therefore, the level of virus in plasma was much lower than that in the feces. The occurrence of FIP in FECV-RM-infected cats. Two cats infected with FECV-RM, both from the FIV-infected group, developed signs compatible with FIP at about 8 and 10 weeks postinfection. Cat 121 was febrile, depressed, anorexic, and had a characteristic yellow-tinged, high-level-protein thoracic effusion. Cat 123 had similar signs but was without fluid effusions. The first cat was confirmed to have the thoracic form of effusive FIP on necropsy and histopathologic examination. Cat 123 had noneffusive FIP, with granulomatous lesions in the omentum, spleen, and mesenteric lymph nodes. Characterization of coronavirus isolates from cats with FIP. Pleural exudate from cat 121 or a homogenized cell-free suspension (1-g equivalent of tissue per ml) of omentum from cat 123 was each inoculated intraperitoneally into two healthy specific-pathogen-free cats. One of two cats given each inoculum developed effusive FIP within 10 to 14 days; the second cat in each group remained seronegative and healthy. The cat given infectious pleural fluid became febrile on day 7 and was euthanized on day 14, while the cat given homogenized omentum developed a fever on day 16 and was euthanized on day 20. The two FIPV isolates were designated FIPV-UCD9 (cat 121 with effusive FIP) and FIPV-UCD10 (cat 123 with noneffusive FIP). The 7b genes (previously 6b genes) of FIPV-UCD9 and -UCD10 were sequenced, and the sequences were compared with those of the 7b genes of FECV-RM and a group of geographically and temporally unrelated feline and canine coronaviruses (Table 1; Fig. 3) (7). FIPV-UCD9 and -UCD10 were closely related to each other and to FECV-RM. FIPV- UCD8, which had arisen from the same cattery several years

VOL. 34, 1996 TWO RELATED STRAINS OF FIPV FROM CATS WITH FECV 3183 Strain TABLE 1. Percent homologies of several representative FECV and FIPV strains % Homology with: FECV-79-1683 FECV-UCD FIPV-UCD3 FIPV-UCD10 FIPV-UCD9 FIPV-UCD8 FECV-RM FECV-RM 87.21 89.86 89.32 99.39 99.52 99.03 100.00 FIPV-UCD8 86.16 88.89 88.35 99.03 99.19 100.00 FIPV-UCD9 87.21 89.32 88.84 99.51 100.00 FIPV-UCD10 87.21 89.32 88.84 100.00 FIPV-UCD3 86.68 99.35 100.00 FECV-UCD 86.89 100.00 FECV-79-1683 100.00 earlier, was also closely related to FECV-RM but to a lesser degree than it was related to FIPV-UCD9 and -UCD10. Although these three isolates were closely related to each other (all four strains were greater than 99% homologous), they formed a genetically distinct clade from previously described FECV and FIPV isolates (Fig. 3). FIG. 3. Phylogenetic tree of several feline and canine coronavirus (CCV) strains based on the nucleotide sequences of their 7b genes. Uncorrected genetic differences were calculated with the program DISTANCES in the Genetics Computer Group package, and the output was used to draw a phylogenetic tree with GROWTREE. DISCUSSION The benign acute clinical course of FECV infection in FIVnaive cats in this study closely resembled infection with FECV- 79-1683 (16) and FECV-UCD (15). In the earlier study (16), FECV-79-1683 was also detected in feces in increasing amounts between days 1 and 7 postinfection, noticeably decreased by day 10, and was largely absent by day 17. However, fecal virus shedding in cats given FECV-RM was still seen at 28 days. The longer shedding period observed for FECV-RM may have reflected its greater virulence; FECV-RM is maintained by cat-to-cat fecal-oral passage, while FECV-79-1683 was tissue culture adapted and passaged. FECV-RM was detected in the blood from days 5 to 7 following oral inoculation. Systemic spread was also observed in the earlier FECV-79-1683 study; the highest levels of virus were found in the small intestine and mesenteric lymph nodes, and low levels of virus were also isolated from tonsils, thymus, lung, spleen, liver, and kidney. FIV-infected cats showed a significant impairment in their ability to clear FECV infection compared with their non-fivinfected siblings, as evidenced by the 100-fold higher levels of fecal virus and the increased duration of fecal shedding. Despite this greatly enhanced level and duration of intestinal virus replication, FIV-infected cats were also outwardly normal following FECV-RM infection. The enhanced level of virus shedding in FIV-infected cats was best explained by a delay in the formation of and a decrease in the titers of coronavirus antibodies in sera. Decreased viral antibody titers, with a concomitant enhancement of virus shedding, have also been described for chronically FIV-infected cats that were challenge-exposed to feline calicivirus (21) or feline herpesvirus (20). FIV, as well as other immunodeficiency lentiviruses, impairs the ability of the host to mount a primary antibody response to T-cell-dependent antigens and impedes immunoglobulin M to immunoglobulin G class switching (22). Two FECV-RM-infected cats from this study, both from the FIV-infected group, later developed FIP. The viruses isolated from lesions of these cats were determined to be FIPVs on the basis of their biologic behavior in animal inoculation studies. The temporal relationship between FECV-RM infection and the appearance of these two FIPV strains, coupled with their close genetic relationship, provides strong direct evidence that FIPVs are mutations of FECVs. If FIPVs arise as mutations during bouts of primary or secondary FECV replication, it stands to reason that certain events might favor both the mutation and subsequent disease. Virus mutations are much more likely to occur if the rate of virus replication is high, such as occurred in the FIV-infected cats in this study. Coronaviruses have a high degree of inherent mutability, and a number of coronavirus mutations have led to pronounced changes in disease potential (9). The detection of these two FIP-inducing mutants of FECV-RM was probably enhanced by the immunocompromised status of the FIV-infected cats that allowed the disease to appear (17). Normal cats may immunologically contain many of these mutant viruses and curtail disease (12). Therefore, two circumstances in this study, increased virus replication and immunoincompetence, may have maximized the chances for both generation and clinical expression of FECV-RM mutants that possessed the FIP phenotype. The nature of the genetic mutation(s) that leads to the FIP phenotype has yet to be determined. This determination will involve extensive sequence comparisons of all important re-

3184 POLAND ET AL. J. CLIN. MICROBIOL. gions of the parent FECV-RM and mutant FIPV-UCD9 and FIPV-UCD10 genomes; such studies are under way (22a). Mutations leading to phenotypically different coronavirus strains in other animal models have involved point mutations, deletions, and recombinations. The porcine respiratory coronavirus apparently arose from transmissible gastroenteritis virus as a result of one major and two minor deletions in the spike gene of RNA 2 and a nucleotide substitution rendering RNA 3 nontranscribable (18, 24). Tremendous strain variation has occurred over time and between distant regions in avian infectious bronchitis virus by base substitutions and genome insertions and/or deletions (2). Variation in avian infectious bronchitis virus isolates can also involve recombination between field and vaccine strains (10). Whatever the FIP-inducing mutations in FECV might be, they appear to be frequent in occurrence. This indicates that the relevant mutations are relatively simple and probably occur in a distinct and hypervariable region of the genome. ACKNOWLEDGMENTS We are grateful to Solvay Animal Health, Inc., Mendota Heights, Minn., for their financial support of the Center for Companion Animal Health, School of Veterinary Medicine, University of California, Davis, Calif., and also to numerous individual donors. We also acknowledge Jeff Carlson, Steve Ramirez, Tami Citron, and Kim Floyd-Hawkins for their excellent technical support and Richard Kienle for computer assistance. REFERENCES 1. Boom, R., C. J. A. Sol, M. M. M. Salimans, C. L. Jansen, P. M. E. Wertheimvan Dillen, and J. van der Noordaa. 1990. Rapid and simple method for purification of nucleic acids. J. Clin. Microbiol. 29:1804 1811. 2. Cavanagh, D., and P. J. Davis. 1993. Sequence analysis of strains of avian infectious bronchitis virus isolated during the 1960s in the U.K. Arch. Virol. 130:471 476. 3. Cheung, R. C., S. M. Matsui, and H. B. Greenberg. 1994. Rapid and sensitive method for detection of hepatitis C virus RNA by using silica particles. J. Clin. Microbiol. 32:2593 2597. 4. Devereux, J., P. Haeberli, and O. Smithies. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387 395. 5. Dougherty, R. M., P. E. Phillips, S. Gibson, and L. Young. 1993. Restriction endonuclease digestion eliminates product contamination in reverse transcribed polymerase chain reaction. J. Virol. Methods 41:235 238. 6. Genetics Computer Group. 1994. Program manual for the Wisconsin package, version 8. Genetics Computer Group, Madison, Wis. 7. Herrewegh, A. A. P. M., H. Vennema, M. C. Horzinek, P. J. M. Rottier, and R. J. de Groot. 1995. The molecular genetics of feline coronaviruses: comparative sequence analysis of the ORF7a/7b transcription unit of different biotypes. Virology 212:622 631. 8. Hickman, M. A., J. G. Morris, Q. R. Rogers, and N. C. Pedersen. 1995. Elimination of feline coronavirus infection from a large experimental specific pathogen-free cat breeding colony by serologic testing and isolation. Feline Pract. 23:96 102. 9. Horzinek, M. C., A. Herrewegh, and R. J. de Groot. 1995. Perspectives on feline coronavirus evolution. Feline Pract. 23:34 39. 10. Kusters, J. G., E. J. Jager, et al. 1990. Sequence evidence for RNA recombination in field isolates of avian coronavirus infectious bronchitis virus. Vaccine 8:605 608. 11. Pedersen, N. C. 1976. Serologic studies of naturally occurring feline infectious peritonitis. Am. J. Vet. Res. 37:1449 1453. 12. Pedersen, N. C. 1995. An overview of feline enteric coronavirus and infectious peritonitis virus infections. Feline Pract. 23:7 20. 13. Pedersen, N. C. 1995. The history and interpretation of feline coronavirus serology. Feline Pract. 23:46 51. 14. Pedersen, N. C., and J. F. Boyle. 1980. Immunologic phenomenon in the effusive form of feline infectious peritonitis. Am. J. Vet. Res. 41:868 876. 15. Pedersen, N. C., J. F. Boyle, et al. 1981. An enteric coronavirus infection of cats and its relationship to feline infectious peritonitis. Am. J. Vet. Res. 42:368 377. 16. Pedersen, N. C., J. F. Evermann, A. J. McKeirnan, and R. L. Ott. 1984. Pathogenicity studies of feline coronavirus isolates 79-1146 and 79-1683. Am. J. Vet. Res. 45:2580 2585. 17. Pedersen, N. C., and K. Floyd. 1985. Experimental studies with three new strains of feline infectious peritonitis virus: FIPV-UCD2, FIPV-UCD3, and FIPV-UCD4. Comp. Cont. Edu. 7:1001 1011. 18. Rasschaert, D., M. Daurte, and H. Laude. 1990. Porcine respiratory coronavirus differs from transmissible gastroenteritis virus by a few genomic deletions. J. Gen. Virol. 71:2599 2607. 19. Reubel, G. H., G. A. Dean, J. W. George, J. E. Barlough, and N. C. Pedersen. 1994. Effects of incidental infections and immune activation on disease progression in experimentally feline immunodeficiency virus infected cats. J. Acquired Immune Defic. Syndr. 7:1003 1015. 20. Reubel, G. H., J. W. George, J. E. Barlough, J. Higgins, N. C. Pedersen, and C. K. Grant. 1992. Interaction of acute feline herpesvirus-1 and chronic feline immunodeficiency virus infections in experimentally infected specific pathogen free cats. Vet. Immunol. Immunopathol. 35:95 120. 21. Reubel, G. H., J. W. George, J. Higgins, and N. C. Pedersen. 1994. Effect of chronic feline immunodeficiency virus infection on experimental feline caliciviral-induced disease. Vet. Microbiol. 39:335 351. 22. Torten, M., M. Franchini, J. E. Barlough, J. W. George, E. Mozes, H. Lutz, and N. C. Pedersen. 1990. Progressive immune dysfunction in cats experimentally infected with feline immunodeficiency virus. J. Virol. 65:2225 2230. 22a.Vennema, H., and N. C. Pedersen. Unpublished observations. 23. Vennema, H., A. Poland, K. Floyd-Hawkins, and N. C. Pedersen. 1995. A comparison of the genomes of FeCVs and FIPVs and what they tell us about the relationships between feline coronaviruses and their evolution. Feline Pract. 23:40 44. 24. Wesley, R. D., R. D. Woods, and A. K. Cheung. 1991. Genetic analysis of porcine respiratory coronavirus, an attenuated variant of transmissible gastroenteritis virus. J. Virol. 65:3369 3373.