From 30 available paired serum samples 17 (56.7%) were obtained from privately‐owned dogs (group A). Five of these 17 samples (29.4%) showed a significant increase in anti‐CRCoV antibody titres. The antibody titres of two dogs increased more than 128‐fold. CRCoV‐specific nucleic acid from nasal or tonsillar swabs was detected in these two dogs.

Another 13 paired serum samples (43.3%) were collected from a population of 15 kennelled working dogs with an acute episode of CIRD (group B). As two of the 15 dogs were non‐compliant with blood sampling, in these cases no serum samples were obtained. In 10 of the 13 paired serum samples (76.9%), a significant increase in anti‐CRCoV antibodies was found. Six dogs revealed a 16‐ to 128‐fold antibody titre increase and concurrent evidence of CRCoV RNA in sample material from the nose or tonsils. No further causative viral agent was detected in these cases (Table 4).

Focussing on upper respiratory tract samples (nasal and tonsillar swabs), viral nucleic acids were detected in 31 of 214 diseased dogs (14.5%). Sixteen dogs tested positive for CRCoV (7.5%), 14 dogs for CPiV (6.5%) and one of these dogs additionally for CAV‐2‐specific nucleic acid (0.5%). One single dog tested positive for CDV‐specific nucleic acid (0.5%). In none of the obtained samples from the upper respiratory tract was CIV‐specific nucleic acid detected. Of those 31 positive dogs, 21 were privately owned (group A), and 10 kept in shelters (group B). They consisted of five puppies, 12 adolescent dogs and 14 adult dogs. Twenty‐seven of the 31 positive dogs (87.1%) showed acute onset of signs, three suffered from chronic disease (9.7%) and for one diseased dog this information was not available (Table 2).

Furthermore, upper respiratory tract samples from two dogs (4.0%) of the clinically healthy control group C tested positive for CRCoV‐specific nucleic acid (Table 2).

Nine dogs from group A (5.2%) and seven dogs out of group B (17.0%) tested positive for CRCoV in either nasal, tonsillar or both samples at one time. One of these dogs belonged to the subgroup of puppies; nine dogs were from the adolescent subgroup and six animals from the subgroup of adult dogs. With one exception, all these animals showed acute onset of CIRD (93.7%).

Fourteen diseased dogs (6.5%) tested positive for CPiV. From those, 11 belonged to group A and three to group B. They all harboured CPiV‐specific nucleic acid in sample material from the nose and one dog concurrently from the tonsils. Four of these dogs were classified as puppies; three dogs were from the adolescent subgroup and seven dogs were adults. Twelve out of these 14 animals showed acute onset of clinical signs (85.7%), one dog was chronically ill, and for another dog this information was not available. Seven dogs (50.0%) were regularly vaccinated‐including against CPiV.

In one of these 14 CPiV‐positive dogs, CAV‐specific nucleic acid was detected concurrently. This dog was privately owned (group A) and tested positive for CAV in both nasal and tonsillar swabs and CAV‐2 strain (Toronto) was confirmed by DNA sequencing. Belonging to the subgroup of adults this dog had been irregularly vaccinated and received its latest vaccine 45 days before sample collection. It presented with a several week history of clinical signs including severe coughing, nasal and ocular discharge, dyspnoea and fever. Apart from that case, in no other dog was viral nucleic acid of two or more different viruses detected. In addition, no other proband of the study tested positive for CAV.

One dog from group A tested positive for CDV‐specific nucleic acid in a sample retrieved from the tonsils. RNA sequencing enabled the identification of a CDV vaccine strain (Onderstepoort). The dog was an adult and presented with chronic respiratory disease but no other signs consistent with CDV infection. The vaccination status of this dog was unknown.

Additional information regarding all PCR‐positive dogs is summarised in Table 3.

All BALF samples collected from 31 chronically ill dogs revealed negative PCR results.

Seropositivity ratios were evaluated by χ

2 test (Minitab 12.0). P < 0.05 value was taken to indicate statistical significance.

All twelve dogs that were tested for CPV at the initial presentation were PCR-negative (Table 5). Vector-borne infections were detected in 4 dogs (31 %, Table 5): infection with Babesia spp. was detected in Dogs 3 and 8; infection with L. infantum was diagnosed in Dog 4 and infection with Dirofilaria immitis was found in Dog 13. Dog 13, which tested positive in D. immitis antigen and Knott tests, had received a certificate from a laboratory in Budapest, Hungary, that stated a negative result in the Knott test in August 2013. None of the rescue dogs tested positive for Ehrlichia canis (Table 5).

Canine adenovirus indirect ELISA Kit (EVL/European Veterinary Laboratory-Netherlands, catalogue no. D1003-AB01) was used for detecting CAV antibodies in dogs. The test was performed as per the manufacturer's instructions. The plates were then read on an automatic plate reader at 450 nm.

The radiographic examinations of the thorax revealed moderate-to-severe interstitial lung changes with variable bronchial thickening in four dogs (Dogs 1 - 4). The changes were generalised and most pronounced in the dorsal (Dog 2), perihilar (Dogs 2 and 4) and caudal lung areas (Dog 3, Fig. 2). The radiographic findings were compatible with bronchopneumonia in all four dogs, and all of the dogs tested CDV PCR-positive. The thoracic radiographs of Dog 13 revealed mild right-sided cardiomegaly and mild generalised bronchointerstitial lung changes; echocardiography showed mild tricuspid and aortic regurgitation but no signs of pulmonary hypertension or right ventricular pressure overload. Dog 13 tested D. immitis-positive but was negative for CDV.

Because of the clinical signs, the tiger was initially diagnosed with rhinotracheitis, and only trachea samples were obtained for laboratory examination after it died. The viral pathogen of felidae include canine/feline distemper virus (CDV/FeDV), feline calicivirus (FCV) and FHV-1, we first considered these three agents. Polymerase chain reaction (PCR) and reverse transcription polymerase chain reaction (RT-PCR) were used to detect these viruses. The PCR/RT-PCR results indicated the presence of FHV-1.

The target fragments observed by agarose gel electrophoresis (AGE) were extracted by a DNA Gel Extraction Kit (Axygen, Hangzhou, China), and were cloned into a pGEM-T Easy vector with routine methods. The sequencing of the cloned plasmid was performed by BGI Sequencing, and the results were submitted to GenBank for alignment analysis. The phylogenetic tree was constructed using MEGA-5 software.

Nasal swab, urine, fecal and blood samples from each affected panda were collected at the time of clinical disease onset. Viral DNA and RNA were isolated from samples using the AxyPrep Multisource Genomic DNA Miniprep kit (AXYGEN, Union City, USA) and RNeasy Mini kit (QIAGEN, Germantown, MD) according to manufacturer’s protocols. Extracted nucleic acids were tested by RT-PCR for CDV using primers specific for CDV H gene (P1:5′-CGAGTCTTTGAGATAGGGTT-3′ and P2: 5′-CCTCCAAAGGGTTCCCATGA-3′). RT-PCR and PCR testing for other viruses threatening giant pandas (canine adenovirus, canine herpesvirus, canine coronavirus, and canine parainfluenza virus) were performed using previously reported methods1730. RT-PCR testing for CDV was also performed on samples collected from the heart, liver, spleen, lungs, kidneys, intestines, and brain of each deceased giant panda, with the exception of Chengcheng for whom tissue samples were not available. Serum samples were collected from the giant pandas during the outbreak to measure SN antibody titers against CDV.

Fluid replacement, systemic antibiotic administration, antinausea medicines, antidiarrhea medicines, and a rigorous diet combined with monoclonal antibodies are the main treatment methods for CPV-2-infected dogs; however, the recovery rates vary from 27.8 to 93.5% (39, 53, 55, 56, 64, 65). Both disease and pathology of the infected animal differ depending on the age. CPV-2 infection in adult dogs results in temporary panleukopenia or lymphopenia; CPV-2 infection in neonatal animals causes myocarditis (12, 66). CPV-2 monoclonal antibodies are highly therapeutic in a short period of time and have treatment well effect (53). The cure rate for a CPV-2 single infection is higher than that of a coinfection with other viruses (41). With improvements in medical treatments, CPV-2 cure rates have been improved using specific drugs or other treatments.

At present, the emergence of new pathogens and the continuous circulation of common etiological agents in dogs have made canine diseases more complex and difficult to diagnose. Dog infectious diseases mainly include respiratory and intestinal viral diseases, including CRV (CAV-2, CDV, CIV and CPIV) and CEV (CAV-2, CanineCV, CCoV and CPV). However, the traditional methods of virus identification and isolation are time consuming, causing delays in treatment initiation. A few methods for detecting virus-induced respiratory or enteric disease have been developed [4, 27, 28, 34], but no previous study had developed a systematic way to detect both CRV and CEV in dogs. Here, we developed two mPCR methods for detection of the most frequently coinfected viruses; these methods could be performed to diagnose dogs according to their clinical symptoms.

Primer design is the first and most important step in the process of establishing a detection method, and the following conditions must be satisfied: primers were designed to bind to conserved sequence regions, to have similar annealing temperatures, and to lack dimers or hairpin structures. In these novel mPCR methods, the primer combination produced amplicons that were easy to distinguish from each other, the primer annealing temperatures were similar, and degenerate bases were required only infrequently. The specificity, sensitivity and reproducibility tests all showed good results.

The mPCR methods were tested on 20 NS and 20 AS samples collected from dogs with symptoms of respiratory disease or enteric disease. The ratio of positive samples to total samples was 80% (16/20) for CRV detection and 85% (17/20) for CEV detection. Because the sample number was insufficient, these results were not statistically significant. However, CPV and CDV clearly remain two of the more serious and epidemic diseases in dogs in worldwide at present [35–38]. Epidemiological monitoring of CPV is particularly important because CPV evolves at a rapid rate, similar to that of Porcine Circovirus 3 [39, 40]. Because a small number of dogs were negative for the viruses tested by the CRV or CEV detection assays, although they suffered respiratory illness or intestinal problems, we suggest that some viruses with low prevalence and pathogenic bacteria may also cause disease in dogs [2, 41]. A variety of pathogenic bacteria are often present along with viruses in canine infections [42, 43], and thus, it is essential to expand the coverage of mPCR detection in the future. For example, CIRD also include CHV-1, canine reovirus, and Bordetella bronchiseptica and so on. At the same time, other pathogens causing serious zoonotic diseases, such as pseudorabies virus, should also be monitored in future [44, 45].

In this study, the detection of CanineCV was added to an mPCR method for the first time, because coinfection of this pathogen with other pathogens is common. Though the pathogenic mechanism of CanineCV is unclear, epidemiological testing is important for future research. CanineCV was not detected from the AS clinical samples; perhaps the limited source of these clinical samples was responsible for this result. We didn’t get a lot of clinical samples because it was not easy to get disease samples. CAV-2 mostly replicates in the lower respiratory tract and was detected in the NS samples; however, the CAV-2 primer pair used in this study was probably able to amplify the CAV-1 DNA virus despite the optimization performed. Notably, the live vaccine strains used may have an unavoidable impact on disease detection using the methods developed in this study. Additionally, discriminating between wild-type infections and vaccines is important, and therefore, a trend exists toward later development of broad-spectrum and accurate mPCR detection methods. Sometimes, cross contamination may lead to experimental failure. It is worth noting that PCR pretreatment and post-treatmen performed in different isolation zones can effectively avoid pollution. Besides, regular air spray cleaning will also play a role.

In conclusion, these newly established mPCR methods provide an efficient, sensitive, specific and low-cost testing tool for the detection of CRV (CAV-2, CDV, CIV and CPIV) and CEV (CAV-2, CanineCV, CCoV and CPV). The use of Taq Master Mix makes the detection process more convenient and reduces the chance of contamination during the process of sample addition; PCRs can be initiated by simply adding enzyme, ddH2O, premixed primers, and template, and thus, this method is superior to other mPCR detection methods. Here, detection of CanineCV was added to mPCR for the first time, making this method suitable for the further study of coinfection by CanineCV and other pathogens. This study provides a novel tool for systematic clinical diagnosis and laboratory epidemiological surveillance of CRV and CEV among dogs.

Necropsies were performed on all deceased giant pandas at which time tissue samples were collected for histologic examination. Lung samples from the giant panda named Fengfeng, who experienced a long illness duration, were selected for histologic review using routine methods. Tissue samples were fixed in 10% phosphate-buffered formalin, embedded in paraffin wax, sectioned, and stained with hematoxylin and eosin prior to analysis.

Antibody (Ab) detection was only performed for 2011 serum samples for CPV and CDV using an indirect ELISA kit (Ingezim Canine Parvo 15.CPV.K1® and Ingezim Moquillo 1.5.CDG.K.1®-Ingenasa), for specific IgG detection, according to the manufactures instructions.

This method allowed the quantification of the Ab titer for CPV, using a formula provided by the kit, although it did not specify the correspondence with the hemagglutination inhibition assay for quantification of anti-CPV antibodies. The values were organized in three classes: 1) 100–1000 ELISA Units (EU); 2) 1000–10000 EU and 3) > 10000 EU.

For CDV the OD values had correspondence with the indirect immunofluorescence (IIF) method, and the values were divided into three categories: 1) low titer (IIF values: 1/20-1/40), 2) medium titer (IIF values: 1/80-1/160) and 3) high titer (IIF values: ≥ 1/320).

Infectious causes of respiratory disease are common in dogs; canine distemper virus, adenovirus 2, parainfluenza, influenza, herpesvirus, pneumovirus, respiratory coronavirus, Bordetella bronchiseptica, various Mycoplasma spp., and Streptococcus equi var. zooepidemicus are documented causes.1 Molecular diagnostic assays to detect viral and bacterial pathogens are available for these agents. In the United States, modified live vaccines (MLVx) for intranasal (IN) administration are currently available for adenovirus 2, B. bronchiseptica, and parainfluenza. These vaccines do not induce sterilizing immunity, and vaccinated dogs can still develop clinical signs of disease if exposed to virulent strains of the organisms.2 It is currently unknown if IN administration of MLVx against these agents results in positive molecular diagnostic assay results in dogs without previous vaccination. If transient positive molecular diagnostic assay results are common after vaccination, the positive predictive value of the diagnostic assays to predict disease caused by these agents in dogs would be decreased.

The purpose of this study was to determine the impact of administration of a single IN dose of a commercially available MLVx adenovirus 2, B. bronchiseptica, and parainfluenza containing vaccine,1 included as part of a facility standard initial vaccination series with a parenteral administration of MLVx containing adenovirus 2, canine distemper virus, and parvovirus, on the results of a commercially available polymerase chain reaction (PCR) panel that amplifies the RNA or DNA of the agents.2

The study was completed with Institutional Animal Care and Use approval. Beagle puppies housed at a commercial breeding facility were used.3 The puppies were housed in a closed facility without contact with other dogs and staff members followed facility barrier precautions over the course of the study. A sterile cotton swab was gently rubbed at the entrance to the external nares, and a second swab gently rubbed against the mucosa of the oropharynx in nonsedated puppies. The swabs were stored separately at 4°C in sterile plastic tubes and stored until shipped by overnight express on cold packs for performance of the molecular assays.2

A total of 12 puppies were screened twice as described, 1 week apart, and all were negative for nucleic acids of the target organisms. Eight puppies were randomly selected for the study and housed in a separate room at the breeding facility for the duration of the study. The puppies were approximately 9 weeks of age when samples were collected on Day 0 before the SQ administration of a MLVx containing adenovirus 2, canine distemper virus, and parvovirus4 and the IN administration of a MLVx1 containing adenovirus 2, B. bronchiseptica, and parainfluenza following manufacturer's instructions (approximately ½ mL per nares). Nasal and pharyngeal swabs were then collected on days 1, 2, 3, 4, 5, 6, 7, 10, 14, 17, 21, 24, and 28 for molecular analysis.2

Sneezing or coughing which have been associated with IN MLVx administration was not noted over the course of the study. Adverse effects associated with the collection of the nasal and oropharyngeal swabs were not noted. At the time the study was performed, the PCR panel utilized also included primers for canine distemper virus RNA; and none of the samples collected over the course of the study were positive. In contrast, nucleic acids of adenovirus 2, B. bronchiseptica, and parainfluenza were amplified from both sampling sites, from all 8 puppies, on multiple days after vaccine administration (Table 1). Because adenovirus 2 was administered in both vaccine types, source of that virus cannot be determined. Increasing numbers of positive samples after vaccination suggest local replication of the vaccinal strains. Decreasing numbers of positive samples over time suggest immune responses inhibiting organism replication. However, quantitative PCR assays normalized to total DNA/RNA on the swab would be needed to confirm or deny these hypotheses. The PCR laboratory adheres to standard operating procedures including use of positive and negative controls thus erroneous results are unlikely.

Agents considered most common for kennel cough syndrome include canine distemper virus, adenovirus 2, parainfluenza, and B. bronchiseptica. However, emerging pathogens include influenza, herpesvirus, respiratory coronavirus, pantropic coronavirus, pneumovirus, and others.1 All of these agents, as well as S. equi var. zooepidemicus and Mycoplasma spp., have been identified as causes of canine infectious respiratory disease. Determination of the agent is important for targeting treatment, particularly for dogs who fail to respond to standard treatment recommendations.2 In animal shelter environments, agent identification is critical for outbreak control and individual case management.3

Bacterial and viral shedding postvaccine administration complicates diagnostic testing and treatment. This is especially problematic in shelter environments as dogs are routinely vaccinated on intake. Viral shedding after vaccination has been detected in cats,4 people,5 cattle,6 pigs,7 and dogs.8 A vaccine strain of B. bronchiseptica was detected via nasal culture up to 4 weeks after IN vaccination of 2–week‐old puppies.9

Commercially available respiratory PCR panels are a relatively cost and time effective diagnostic method for identifying multiple respiratory pathogens. However, amplification of nucleic acids may inherently lead to inaccurate clinical diagnosis because small amounts can be amplified from some animals even though the agent may not be present in sufficient quantity to cause disease. In this study, nucleic acids of all 3 organisms contained in the IN vaccine were amplified from both sites on multiple days via PCR, although no clinical signs of respiratory disease were observed. Thus, interpretation of PCR panel results for diagnoses should include consideration of recent vaccination status and clinical signs of disease. Use of quantitative PCR and wild‐type sequence differences may be able to differentiate between vaccine and pathogenic agent shedding and may be used diagnostically in the future.

Real‐time reverse transcriptase PCR has been used to amplify canine distemper virus RNA in blood, urine, and conjunctival swabs after administration of SQ MLVx.10 In this study, the PCR panel did not amplify distemper virus RNA from nasal or pharyngeal swabs. Further studies are needed to determine whether the negative result is because this strain of vaccine virus does not reach the nasal or pharyngeal tissues or was present at levels below the detectable limit of the assay used.

All RNA extracts were subjected to a previously-established RT-PCR assay for detection of CnPnV RNA, with minor modifications. Briefly, a one-step method was adopted using SuperScript™ One-Step RT-PCR for Long Templates (Invitrogen srl, Milan, Italy), according to the manufacturer’s instructions, and primers SH1F/SH187R that amplify a 208-bp of the small hydrophobic (SH) protein gene (Table 1). The following thermal protocol was used: reverse transcription at 50°C for 30 min, inactivation of Superscript II RT at 94°C for 2 min, 40 cycles of 94°C for 30 s, 54°C for 30 s, 68°C for 60 s, with a final extension at 68°C for 10 min. The PCR products were detected by electrophoresis through a 1.5% agarose gel and visualisation under UV light after ethidium bromide staining.

In addition to the gel-based RT-PCR, a real-time RT-PCR assay based on the TaqMan technology was developed for the rapid detection and quantification of the CnPnV RNA in all clinical samples. Reactions were carried out using Platinum® Quantitative PCR SuperMix-UDG (Invitrogen srl) in a 50-µl mixture containing 25 µl of master mix, 300 nM of primers CnPnV-For and CnPnV-Rev, 200 nM of probe CnPnV-Pb (Table 1) and 10 µl of template RNA. Duplicates of log10 dilutions of standard RNA were analyzed simultaneously in order to obtain a standard curve for absolute quantification. The thermal profile consisted of incubation with UDG at 50°C for 2 min and activation of Platinum Taq DNA polymerase at 95°C for 2 min, followed by 45 cycles of denaturation at 95°C for 15 s, annealing at 48°C for 30 s and extension at 60°C for 30 s.

CnPnV positive samples were inoculated into semiconfluent canine fibroma (A-72) cells, as previously described. Inoculated cells were maintained in D-MEM supplemented with 5% FCS and monitored daily for the occurrence of cytopathic effect (CPE). After 6 days of incubation, the monolayers were tested for CnPnV antigen by an immunofluorescence (IF) assay using a monoclonal antibody targeting HRSV (Monosan®, Sanbio BV, Uden, The Netherlands). The cells were sub-cultured every 6–8 days for 5 consecutive passages.

To evaluate the reproducibility of the assay, the detection mPCRs for both CRV and CEV were performed as three independent mPCR assays by using three different PCR instruments at different times. Three premixed plasmids for CRV (Fig 5A) and CEV (Fig 5B), with different dilutions, could be amplified under different conditions and showed similar results among the assays.

There are currently five other known canine parvovirus species belonging to two genera of the Parvoviridae family. Canine parvovirus 2 (CPV2) in the Carnivore protoparvovirus 1 species is a highly pathogenic virus that is closely related to feline parvovirus (FPV), the cause of feline panleukopenia, and can infect other carnivores such as coyotes, wolfs, raccoons and pumas. Canine bufavirus, a second protoparvovirus (in the species Carnivore protoparvovirus 2) was reported in 2018 in fecal and respiratory samples from both healthy and dogs with signs of respiratory illness. That same protoparvovirus was recently reported as a frequent component of juvenile cats fecal and respiratory samples. The canine minute virus (CnMV) in the Carnivore bocaparvovirus 1 species is less pathogenic than CPV2 but can cause diarrhea in young pups and is frequently found in the context of co-infections. Distantly related to CnMV, a second canine bocavirus in the Carnivore bocaparvovirus 2 species was sequenced in dogs with respiratory diseases. A third bocavirus was then characterized from the liver of a dog with severe hemorrhagic gastroenteritis.

Here, we describe the near complete genomes of two closely related cachaviruses, members of a new tentative species (Carnivore chapparvovirus 1) in a proposed genus Chapparvovirus, the third genera of viruses from the Parvoviridae family now reported in canine samples. The chapparvovirus was found in only two animals of the initial nine sampled. Many of the dogs in the outbreak analyzed were sampled more than 10 days after onset of clinical signs, increasing the possibility that they were no longer shedding viruses. Additionally, diarrhea is one of the top reasons for veterinary visits and some patients may have coincidentally presented with diarrhea from some other cause.

The two samples positive for CachaV-1 presented in the same week and were in the group of patients with the most severe clinical signs, requiring plasma transfusion and more aggressive supportive care. One of the two dogs, sampled at nine days after onset, died two days later. Because of the variable and often delayed feces sampling, it was therefore not possible to determine a clear disease association in this small group of diarrheic dogs (i.e., not all affected animals were shedding cachavirus).

A possible role for the cachavirus infection in canine diarrhea was further tested by comparing cachavirus DNA PCR detection in larger groups of healthy and diarrheic animals including a group of animals with bloody diarrhea. A statistically significant difference (p = 0.037) was seen when diarrhea samples from 2018 were compared to the feces from healthy animals collected the same year. When 2017 diarrheic samples were compared to e 2018 healthy samples, the p-value was 0.08. When 2017 and 2018 diarrhea samples were combined and compared to the healthy samples, the p-value was 0.05. The association of cachavirus with diarrhea is therefore borderline and the detection of viral DNA remains limited to ~4% of cases of diarrhea. The limited number of healthy samples available for PCR limited the statistical power of this analysis and a larger sample size will be required for further testing of disease association. The absence of detectable cachavirus DNA in 83 other cases of bloody diarrhea was unexpected given the similar signs that developed in the initial outbreak. Detection of viral DNA in feces may be related to timing of sample collection as shedding of the intestinal lining during hemorrhagic diarrhea may preclude viral replication and fecal shedding.

The detection of this virus in multiple fecal samples, the absence of prior cachavirus reports from tissues or fecal samples from other animals, and the confirmed vertebrate (murine) tropism of another chapparvovirus (mouse kidney parvovirus), support the tentative conclusion that cachavirus infects dogs. Given its relatively low viral load and only borderline association with diarrhea, this virus’ possible role in canine diarrhea or other diseases will require further epidemiological studies. Because viral nucleic acids in fecal samples may also originate from ingestion of contaminated food (rather than replication in gut tissues), the tropism of cachavirus for dogs will require further confirmation such as specific antibody detection, viral culture in canine cells, and/or evidence of replication in vivo such as RNA expression in enteric tissues of dogs shedding cachavirus DNA.

An RT-RPA assay with high analytical sensitivity and specificity was successfully developed for the rapid detection of CDV, which could be completed within 20 min. More importantly, the portable feature of the RT-RPA assay makes it applicable at quarantine stations, ports or the site of outbreak. The rapid, sensitive and feasible RT-RPA assay would be a useful tool in CDV control, especially in the resource-limited settings.

The percentage of positive results from in vitro transcribed RNA dilutions were 100% (21/21), 100.0% (21/21), 95% (20/21), 25.0% (4/16), and 0% (0/16) for 103, 102, 101, 100, and 10-1 copies of the RNA standard, respectively. The ≥95% detection rate and 95% confidence limits, calculated by probit regression analysis, were 10.1 and 7.6–15.6 copies per reaction, respectively, indicating that the accuracy of the established reaction was comparable to that of the real-time RT-PCR for the detection of CDV.

Dilutions of the Onderstepoort reference strain, when tested with both real-time RT-PCR and RT-iiPCR, generated positive results with all 10-2 dilution samples. At the 10-3 dilution, 2/3 and 1/2 samples showed positive signals in RT-iiPCR and real-time RT-PCR, respectively (Table 1), indicating that the detection limit of RT-iiPCR was comparable to that of real-time RT-PCR.

During specificity testing, positive signals were generated only from CDV (Snyder Hill strain) but not from the other canine pathogens tested (Table 2), indicating that the established reaction could detect CDV specifically.

In this study, we developed a RT-RPA method based on exo probe for the rapid and sensitive detection of CDV. Specificity analysis revealed that the RT-RPA assay could only detect the CDV, but not other viruses (Fig. 1). Other CDV genotypes were not included in the assay except for the genotype America-2, which is deficiency of the study. The RPA is tolerant to 5–9 mismatches in primer and probe showing no influence on the performance of the assay [12–14], and there were only 2–4 mismatches in the primers and probe in this study with other CDV genotypes. It is assumed the assay would detect all genotypes of CDV, based on targeting a conserved region, but this was not confirmed by testing validated genotypes. The detection limit of the RT-RPA was 9.4 copies, which was 10 times lower than the real-time RT-PCR described previously. We further evaluated this method using clinical samples, and the diagnosis agreement of the RT-RPA and real-time RT-PCR was 100%. Interpretations were limited by the small sample-size, but the results suggested that the developed RT-RPA performed well in CDV detection. RT-RPA assay should be further tested to more CDV strain RNA extracts or clinical samples from various regions worldwide to evaluate sensitivity of the assay for detecting various strains of the virus circulating.

In recent years, a number of isothermal DNA amplification methods have been developed as a simple, rapid alternative to PCR-based amplification. In the RT-LAMP assay for CDV, four primers were needed and the optimal reaction condition was 60 min at 65 °C. The developed iiPCR could detect as low as 7.6 copies of CDV RNA in approximately 1 h. For the RT-RPA assay developed in this paper, it could detect 9.4 copies of CDV RNA in 12 min, which was more rapid than the above assays. Compared to other isothermal amplification techniques, RPA requires no initial heating for DNA denaturation, and the results could be obtained in less than 20 min; RPA demonstrates a certain tolerance to common PCR inhibitors, and could tolerate a wide range of biological samples; RPA reagents in the lyophilized pellet form could be delivered and stored without cold chain, which could perform satisfactory at 25 °C for up to 12 weeks and at 45 °C for up to 3 weeks. Thus, RPA may be the most applicable approach for the field and point-of-care diagnosis of infectious diseases. A noteworthy feature of the developed RT-RPA assay, i.e., the use of the tube scanner Genie III makes on-site CDV detection feasible, which is especially important for CDV detection and epidemiological surveillance in the field.

Nine canine diarrheal samples from an unexplained outbreak of diarrhea were analyzed by viral metagenomics using three pools of three diarrhea samples each. Based on the BLASTx results, one of the three pools showed the presence of viral sequences most closely related to different chapparvoviruses reported from different vertebrates (0.05% of all reads). Other eukaryotic viral sequences observed were from Gyrovirus 4 (0.0003% of all reads), which has been reported in both chicken meat and human stool, indicating that it likely represents a dietary contaminant, and Torque teno canis virus (0.002% of reads), a common commensal canine blood virus.

Using de novo assembly and PCR paired with Sanger sequencing, a near complete genome of 4,123 bases containing the two main open reading frames of chapparvoviruses was generated (Figure 1, panel A). The available genome consisted of a 516 bases partial 5’UTR followed by an ORF encoding a 663 aa non-structural protein (NS) possessing the ATP binding Walker loop motif GPSNTGKS followed by a second ORF encoding a 505 aa viral capsid (VP) finishing with a 108 bases partial 3’UTR (Figure 1, panel A). When NS1 and VP1 proteins were compared to all available parvovirus sequences, the closest relative was from a Cameroonian fruit bat chapparvovirus (MG693107.1) with an amino acid identity of 61 and 63% respectively (Table S1). A 210 amino acid ORF that is missing a start codon and is overlapping the NS1 ORF was also detected showing 57% identity to its homologue protein in mouse kidney parvovirus (AXX39021) (Figure 1, panel A). This NP ORF is widely conserved among chapparvoviruses. The 5’ UTR DNA sequence was 68% identical to that of the bat parvovirus sequence (MG693107.1)). The virus was named cachavirus (canine chapparvovirus) strain 1A (CachaV-1A).

Distance matrices of the NS1 showed that the cachavirus is sufficiently divergent based on ICTV criteria (members of same species showing >85% NS1 identity) to qualify as a member of a tentative new species Carnivore chapparvovirus species 1 in the proposed Chapparvovirus genus (Table S1). A phylogenetic analysis of the NS1 ORF confirms its closest currently known relative is from a Cameroonian fruit bat (Figure 1, panel B).

Using a nested PCR, the other 8 samples were tested for the presence of this virus which was detected in a second diarrheic sample from that outbreak.

A larger set of canine fecal samples were then tested using a real-time PCR assay. Of 2,053 fecal samples tested, a total of 80 were positive (Table 1). Fecal sample submissions from the same time frame as the outbreak (Sept-Oct 2017) were tested in order to determine the prevalence of CachaV-1 during that time. Healthy samples from fecal flotation samples submitted in 2018 for preventive care screening were available. A second set of diarrhea samples that were collected during the same time frame as the healthy samples was also analyzed to check for differences in prevalence across time, as was a set of 83 bloody diarrhea samples.

Three stool samples out of 203 healthy animals tested positive, 32 were positive out of 802 diarrhea submissions from September to October of 2017, and 45 were positive out of 965 diarrhea submissions from September to October of 2018. None of the 83 bloody diarrhea samples tested were positive (Table 1). When the fraction of PCR positive fecal samples was compared between the healthy animals (1.47% positive) and those with diarrhea, a statistically significant difference (p < 0.05) could be detected with the 965 diarrhea cases from 2018 (4.66% positive; p = 0.037), but not with the 803 diarrhea cases collected in 2017 (4.0% positive; p = 0.08). When 2017 and 2018 diarrhea samples were combined (4.35% positive) and compared to the healthy group (1.47% positive), we measured a p-value of 0.05.

Cachavirus viral load as reflected by the Ct value of the real-time PCR were low across all four cohorts, with Ct values ranging from 29 to 39 with an average value of 36 for all positive groups. The five dog samples with the lowest Ct values (highest viral load) were then analyzed by viral metagenomics. All five samples yielded cachavirus reads, but one yielded a near complete genome (cachavirus [1B]). This sample also yielded 0.001% reads that were related to anelloviruses. None of the other four animals showed the presence of other known mammalian viruses. The cachavirus-1B genome showed 98% overall nucleotide identity with the index IDEXX-1A strain. The NS1 and VP encoded protein showed 99 % identity.

A clinical evaluation was further performed to determine whether the CDV RT-iiPCR was suitable for detection of CDV viruses from clinical samples. For this purpose, 110 clinical samples submitted to the University of Tennessee, College of Veterinary Medicine (UTCVM), Clinical Virology Lab from 2010 to 2014 were tested by RT-iiPCR and real-time RT-PCR. These samples included strains from 4 different clades of the phylogenetic trees derived from sequences of the H gene and M-F intergenic region. 100% agreement was found between the two reactions (56 positive and 54 negative) (Table 3). No discrepancy was found in samples (13/56) containing low levels of CDV (Ct > 35, real-time RT-PCR, indicating that the established method reliably detects low amounts of CDV in clinical samples. Positive samples had real-time RT-PCR Ct values ranging from 14.31 to 39.41, indicating that the RT-iiPCR was able to detect viral RNA across the entire range of the assay.

The proportion of CPV-2-positive unvaccinated dogs varies between 32.63 and 84.98%, whereas CPV-2 positive vaccinated dogs (vaccinated at least once) varies between 15.02 and 48.42% (Figure 7). Both vaccinated and unvaccinated dogs can be affected by CPV-2 (60). The positive rates of CPV in unvaccinated dogs were significantly higher than those in vaccinated animals (41–43, 47, 50, 55, 56, 58, 60). The above studies show that vaccination is vitally important for the prevention and control of CPV-2. In China, insufficient attention is often paid to dogs, and therefore, individuals have not yet realized the importance of immunizing their pets. This in turn leads to a large proportion of unimmunized sick dogs. In conclusion, the key to preventing canine parvovirus is to establish a timely and functional immune response.

Between June 2012 and June 2014, a total of 207 specimens were obtained from 207 enrolled dogs. Of these 207 cases, 74 (35.7 %) were positive for at least one gastrointestinal pathogen, as determined by a commercial kit and real-time PCR. Among those dogs with positive results, most were positive for DogCV (58/74, 78.4 %), followed by CPV-2 (22/74, 29.7 %), giardia (8/74, 10.8 %), CCV (4/74, 5.4 %) and CDV (3/74, 4.1 %). The correlations between the severity of diarrhea and positive tests for the different canine gastrointestinal pathogens are summarized in Table 2.