Canine parvovirus type 2 (CPV-2) is a highly contagious infectious agent that causes gastroenteritis in dogs. CPV-2 emerged in 1978 as a new dog disease agent and rapidly spread around the world in domestic and wild dogs causing high morbidity (100%) and frequent mortality (up to 10%) . In the 1980s, two antigenic variants of CPV-2, distinguishable using monoclonal antibodies (MAbs), emerged almost simultaneously and were termed as CPV-2a and CPV-2b respectively . Currently, CPV-2a is the prevalent field strain in Italy and Germany, while CPV-2b is common in USA, Taiwan and Japan [2, 14]. CPV-2c having a change of D426E occurring in a strategic residue responsible for the antigenicity of CPV-2b has been detected in Vietnam, Italy, Spain, Germany, United Kingdom and South America [7, 17]. CPV-2 was isolated for the first time in India by Ramadass and Khadher in 1982  and after that several incidences of the disease have been reported from different parts of the country. The prevalence of CPV-2a has been documented in 2001 in India by Narayanan et al . It was also reported that CPV-2b is more common when compared to CPV-2a in Northern India . The prevalence of CPV-2c has also been reported in India by Nandi et al .
The laboratory diagnosis of CPV-2 infection mostly relies on detection of virus in faecal samples or antibody in serum samples of suspected animals. The antibody in the sera can be detected preferably by competition ELISA (c-ELISA) and sometimes by agarose gel immunodiffusion (AGID) but unable to differentiate vaccinated from infected animals. Considering, the extreme severity of the disease, the prompt, sensitive and accurate diagnosis of the canine parvoviral disease on the basis of antigen detection is of paramount importance. It can be done by demon-stration of viral antigens in the faecal samples by AGID, virus isolation in cell culture, CIE (counter immunoelectrophoresis) test , virus neutralization test, haemagglutination (HA) test, immunoelectron microcopy, immunofluorescent test, antigen-capture (sandwich) ELISA or genomic DNA detection by polymerase chain reaction (PCR) [15, 30]. Recently, the PCR technique has been increasingly used as a tool for the diagnosis of several viral infections. Further, real-time PCR , nucleic acid hybridization or dot blot, in situ hybridization , and loop-mediated isothermal amplification (LAMP)  etc with varying degree of sensitivity and specificity have been employed for diagnosis of CPV-2.
In India, the disease is widely prevalent and there exist about 25 million dogs with a high proportion of stray dogs capable of harbouring the CPV-2 without showing any symptoms. Since copious amounts of CPV-2 are shed in the faeces of infected dogs and the infection is easily spread among susceptible animals through faecal-oral route, rapid and early diagnosis of CPV-2 lead to proper treatment with high rate of recovery. The AC-ELISA for detection of viral antigens is an ideal choice as it is less time-consuming, easy to perform, having good sensitivity and specificity and well suited for screening of large number of samples. There is no indigenous ELISA kit available in India for CPV diagnosis and expensive ELISA kits are available only in few countries. So, the present study was undertaken to develop a polyclonal antibody (PAb) based AC-ELISA for detection of CPV antigens from faecal samples of infected dog and infected cell culture supernatant [4, 8, 12, 16, 18, 23].
Three faecal samples found positive in PCR were subjected to blind passages in MDCK cells. In the first 3 passages, no cytopathic effects (CPE) were seen. However, from 4th passage onwards, MDCK cells exhibited CPE characterized by rounding of cells, granulation and aggregation of cells after 72 hpi which increased subsequently and widely distributed in whole monolayer (Fig. 1). The results are in accordance with the Joshi et al . Then the virus was harvested and freeze-thawed thrice and stored at -20℃ until further use. The virus titre was determined by Reed and Muench method . The titre of virus was calculated from three different readings and the mean was found to be 104.8 TCID50/mL. Using this as inoculums, 2 000 mL of CPV-2 infected cell culture fluid was produced for purification of virus.
In the sucrose density gradient ultracentrifugation, the pellet obtained was resuspended in 2 mL of PBS. The protein content of the purified virus was estimated by U.V. method in nanodrops spectrophotometer (Thermo-scientific, USA) and found to be 350 μg/mL. After test bleeding, the presence of antibody was evaluated by AGPT and a distinct band was visualized between the raised hyperimmune serum in both species and purified virus (Fig. 2). Final bleeding was done after 10 days of the final immunization, serum collected, inactivated and kept at -20℃ for further use.
The optimum dilutions of capture and tracing antibodies were selected by check-board titration against fixed dilution/concentration of CPV antigens with different dilution of rabbit and guinea pig hyperimmune serum and most satisfactory result was obtained at a 1:1 600 dilution for the capture antibody and a 1:400 dilution for the tracing antibody with 1:2 dilution of the positive antigen and the negative antigen. At this dilution the total protein content of the positive antigen control was about 1.0 μg per well and the P/N ratio was > 2.5. These dilutions of reagents were followed throughout the study and subsequently for testing different samples. No excessive variations ( < 10%) between plates were detected and found statistically significant and found to be corroborated with Maree and Paweska . The AC-ELISA was able to detect different variants of CPV-2 such as CPV-2a, CPV-2b and CPV-2c. It detected various CPV-2 variants in the infected cell culture supernatant as positive and showed no reactivity with healthy MDCK cell culture derived negative control. Analytical specificity was confirmed for other canine enteric viruses such as canine coronavirus, canine adenovirus etc. and found to be nonreactive. The CPV-2 variants used in the study were typed previously on the basis of PCR and sequence analysis and maintained in the Virus Laboratory, CADRAD. The analytical sensitivity of the AC-ELISA was determined using serial 10-fold dilutions of CPV-2 infected cell culture supernatant of known titer (104.8 TCID50/mL) and the detection limit of the AC-ELISA assay was up to 10-2dilution, equivalent to 102.8 TCID50/mL. The analytical sensitivity of the AC-ELISA was compared with the pCPV-2ab primer set based PCR  and detection limit of the PCR was up to 10-4dilution, equivalent to 100.8 TCID50/mL (Fig. 3). So, PCR was found to be about 100 times more sensitive than the AC-ELISA for detection of CPV in cell culture system (Table 1).
Figure 3. Agarose gel showing amplicon of 681 bp of CPV-2 using pCPV-2ab (F & R) primer. M, 100 bp DNA marker; Lane 1 to 5, 681 bp PCR products of samples 1 to 5; 6, 681 bp PCR product of positive control; 7, Faecal sample of healthy dog as negative control.
Table 1. Relative sensitivity of AC-ELISA compared to PCR for detection of CPV-2.
At the end of the electrophoresis, the gel was visualized under the UV transilluminator and an amplicon of 681 bp of VP2 genes was seen from positive faecal samples and positive control, but not in fecal sample from healthy dog (Fig. 4).
In s-ELISA, 69 out of 152 samples were found positive, while in PCR 78 samples were found positive. Nine samples negative in AC-ELISA were found positive in PCR, 69 samples were positive in both AC-ELISA and PCR and 74 samples were found negative in both tests (Table 2). The relative sensitivity, specificity and accuracy of the AC-ELISA were found to be 88.4%, 100% and 91.4% respectively (Table 2).
Table 2. Relative performance of PCR and AC-ELISA for detection of CPV-2 in samples.