The Veterinary Record, August 4, 2007 Aetiological role of viruses in puppies with diarrhoea K. Ye¸silbag ˘ Z. Yilmaz, A. Özkul, A. Pratelli DIARRHOEA is an important problem in dogs during the first four months of life. Electron microscopy of faeces has enabled researchers to investigate the viral agents responsi- ble for diarrhoea in dogs, and up to eight viruses have been detected by this method (Finlaison 1995). Among them, particular importance has been attributed to canine parvo- virus type 2 (CPV-2) and canine distemper virus (CDV). CPV-2 first emerged in the late 1970s and rapidly spread around the world. The original virus was soon replaced by antigenic mutants CPV-2a and CPV-2b in 1978 and 1984, respectively. Infection with CDV results in multisystemic disease that affects the respiratory, gastrointestinal and central nerv- ous systems. It occurs worldwide in both unvaccinated and vaccinated populations (Ek-kommonen and others 1997), although widespread vaccination programmes have dra- matically reduced its incidence. Canine coronavirus (CCoV) has had an increasing role in cases of diarrhoea in pup- pies (Mochizuki and others 2001), causing a gastroenteritis similar to that caused by parvovirus. There are two distinct genotypes, of which CCoV type 1 is genetically more closely related to feline coronavirus type 1 than to the classical strains of CCoV (Pratelli and others 2003). Other viruses that may cause canine enteritis include rota-, calici-, astro- and adenoviruses. This study investigated the role of CPV-2, CDV, CCoV and rotaviruses in puppies with diarrhoea. In addition, Turkish CPV-2 isolates were characterised for their antigenic types. A total of 104 dogs from different breeds were examined, selected from clinical cases submitted to the Small Animal Clinics of the Veterinary Faculty of Uludag ˘ University, Bursa, Turkey, between November 2002 and October 2003. During this period there were 691 dogs presented to the clinics with different clinical disorders. Stool samples or rectal swabs were collected from 90 puppies younger than four months of age, which had either diarrhoea alone or diarrhoea plus vomiting and fever. All the puppies were the offspring of domestic dogs. The 90 samples were tested for both CPV-2 and CCoV; 25 of them were also screened for CDV and rota- virus. Stool samples from an additional 14 dogs with gas- trointestinal and respiratory signs were also collected and tested for CDV. Some of the animals had been vaccinated, starting at eight weeks of age, but none had been vaccinated against CCoV. Swab samples were transferred to the laboratory in buff- ered saline, and stored at –30°C until testing. Detection of CPV-2 in stool samples was performed by haemagglutination (HA) reaction with 1 per cent porcine red blood cells at +4°C. Samples were also inoculated on to freshly trypsinised A72 cells, and virus growth was confirmed by HA. All the isolates that gave a HA titre greater than 1:8 were subjected to mono- clonal antibody (mAb) typing. A panel of four mAbs to CPV-2, (A4E3, B4A2, C1D1 and B4E1) was used for antigenic typing in a haemagglutination inhibition (HI) test using six HA units of the virus. This analysis was performed on both faecal sus- pensions and on viruses after three passages in cell cultures. Differential mAb reactivity allowed the characterisation of CPV isolates as CPV types 2, 2a and 2b. A total of 39 samples, 25 from puppies with diarrhoea and 14 from animals with diarrhoea and respiratory signs, were subjected to reverse transcriptase-PCR (RT-PCR) to detect a CDV-specific RNA sequence, 372 base pairs (bp) in length, located in the F gene of the viral genome. RT-PCR was car- ried out as described by Gassen and others (2000), using the primer set CDF1 (5'-AATTTAGTCCTCGAACCAATCAACCC-3', posi- tions 662 to 686) and CDF-2 (5'-TAACCCTAATCTCTGCCCAACTAA T-3', positions 1033 to 1009). CCoV was detected by RT-PCR as described by Yes ¸ilbag˘ and others (2004), using the primer set CCoV1 to CCoV2 to amplify a 409 bp fragment of the M gene. To detect atypical strains of CCoV (CCoV type 1), samples positive for CCoV RNA were sub- jected to a second round of RT-PCR using the primer set CCoV1a to CCoV2: CCoV1 (5'-TCCAGATATGTAATGTTCGG-3', positions 6729 to 6748), CCoV2 (5'-TCTGTTGAGTAATCACCAGCT-3', positions 7138 to 7118) and CCoV1a (5'-GTGCTTCCTCTTGAAGGTACA-3', positions 6900 to 6920). Polyacrylamide gel electrophore- sis (PAGE) was used to detect rotavirus genomic segments (Herring and others 1982). Student’s t test was used for sta- tistical analysis. In total, 32 of the 90 stool samples (35·5 per cent) were positive for CPV-2 using the HA assay. Of the 27 samples sub- jected to mAb analyses, 17 (62·9 per cent) were positive for CPV-2a, while 10 (37·1 per cent) were positive for CPV-2b (Table 1). Total positivity for CDV was 38·4 per cent (15 of 39 sam- ples) by RT-PCR. Of the 25 samples taken from puppies with gastrointestinal signs six (24 per cent) were positive for CDV, and nine (64·2 per cent) were positive in dogs with both res- piratory and gastrointestinal signs. Three stool samples from puppies that had previously been vaccinated with a combined vaccine (DHPPi) were also positive for CDV. CCoV findings were as previously reported (Yes ¸ilbag˘ and others 2004). CCoV-spe- cific genome sequences were detected in 14 (15·5 per cent) of the 90 samples by the first RT-PCR. Five of these 14 positive samples were also positive by the second RT-PCR. None of the samples was positive in rotavirus-specific migration patterns by PAGE, and mixed infection with the agents investigated was not detected. The positivity rates for both CPV-2 and CDV were signifi- cantly higher than for CCoV (P<0·01), but there was no sig- nificant difference between the positivity rate for CDV and CPV-2 (P>0·05). The prevalence of CPV-2a was significantly higher than that of CPV-2b (P<0·001). There was no signifi- cant difference between the prevalence of CCoV types 1 and 2 (P>0·05). Most previous studies investigating the viral aetiol- ogy of canine diarrhoea have used electron microscopy and have suggested that CPV is the major agent responsible (Hammond and Timoney 1983, Rimmelzwaan and others 1991). The findings of the present study confirm this to be the case: the positivity rate of CPV-2 (35·5 per cent) was fol- lowed by that for CDV (24 per cent) and CCoV (15·5 per cent) in acute cases of diarrhoea in puppies (P<0·01). However, recent data have suggested that HA is relatively insensitive compared with PCR for the detection of CPV-2 (Desario and others 2005), and an even higher rate of positivity for CPV-2 may have been detected if PCR had been used to detect the virus in the present study. Veterinary Record (2007) 161, 169-170 K. Ye¸ silba˘ g, DVM, PhD, Department of Microbiology, Virology, Z. Yılmaz, DVM, PhD, Department of Internal Medicine, Faculty of Veterinary Medicine, Uludag ˘ University, 16059, Bursa, Turkey A. Özkul, DVM, PhD, Department of Virology, Faculty of Veterinary Medicine, Ankara University, Ankara, Turkey A. Pratelli, DVM, PhD, Department of Animal Health and Welfare, Faculty of Veterinary Medicine, University of Bari, 70010 Valenzano, Bari, Italy Viral Number of Number (%) of agent Samples tested Positives (%) CPV-2 antigenic types* CCoV genomic types CPV-2 90 32 (35·5) 2a: 17 (62·9) Type 1: 5 (35·7) CCoV † 90 14 (15·5) 2b: 10 (37·1) Type 2: 9 (64·3) CDV 39 15 (38·4) Rotavirus 25 0 * From positive isolates that gave a titre >1:8 by haemagglutination assay † Yes¸ilbag ˘ and others (2004) CPV-2 Canine parvovirus type 2, CCoV Canine coronavirus, CDV Canine distemper virus TABLE 1: Positivity rates detected in faecal samples tested and typing results Short Communications group.bmj.com on March 5, 2013 - Published by veterinaryrecord.bmj.com Downloaded from