Introduction

Canine parvovirus (CPV) is a small, non-enveloped virus, containing a single-stranded DNA (~5 kb). The genome contains two major ORFs, one encoding two non-structural proteins NS1 and NS2, the other, two structural proteins (VP1 and VP2). VP1 and VP2 are translated by alternative splicing from the same primary transcript [1]. Canine parvovirus type 2 was first identified in 1978, and was soon recognized as a major killer of puppies, causing severe haemorrhagic diarrhoea, vomiting, and dehydration. CPV-2 has been shown to be closely related to feline panleukopenia virus (FPLV) and FPLV-like parvoviruses from wild carnivores, such as mink, fox, and red panda.

A variant virus, CPV-2a, with five amino acid (aa) substitutions, emerged in 1979 and replaced CPV-2 worldwide in approximately 1 year [2, 3]. Another antigenic variant, CPV-2b, with a single additional substitution (Asn426Asp) in the VP2 protein, appeared in 1984 and spread globally. Since then, gene sequencing and genetic analyses have been applied frequently and many new CPV variants throughout the world have been detected, such as the “new CPV-2a” and “new CPV-2b” with an additional 297-Ala mutation [4], CPV-2c(a) and CPV-2c(b) with an additional 300-Asp mutation [5], and CPV-2c with an additional 426-Glu mutation [6].

In China, the genetic relationships among parvoviral isolates have not been reported. Here, the VP2 genes of 22 strains of canine parvovirus, isolated from domestic dogs in China between 1983 and 2008, have been sequenced and analyzed in comparison with the sequences of 30 other Chinese strains of CPV (from GenBank) and eight reference CPV isolates.

Materials and methods

Virus isolation

Twenty-two PCR-positive fecal samples from clinical cases, collected since the early 1980s in various areas of China, had been stored at −80°C. For virus isolation, the samples were filtered through a 0.22-μm Millipore filter. Cat kidney F81 cells (obtained from the China Institute of Veterinary Drugs Control) were inoculated with the filtrates and cultured in Eagle’s Minimal Essential Medium (MEM) containing 10% fetal bovine serum and antibiotics (100 U/ml penicillin G and 100 μg/ml streptomycin sulfate). The cells were incubated at 37°C and observed daily for 4 to 5 days for cytopathic effects (CPE). Five blind passages were carried out for each sample.

VP2 gene amplification and sequencing

Viral DNA was extracted using the AxyPrepTM viral DNA miniprep kit (Axygen Scientific, USA) from viral isolates, according to the manufacturer’s instructions. PCR amplification was performed using ExTaq polymerase (TaKaRa Biotechnology Co., Dalian, China) and primers P-forward: 5′-GGATCCCCAATGAGTGATGGAGCAGTTCAACCAGAC-3′ (located on the CPV-d VP1 gene sequence at position 499 to 528), and P-reverse: 5′-CTCGAGTTAATATAATTTTCTAGGTGCTAGTTGA-3′ (located at position 2,229 to 2,257), designed according to the DNA sequence of CPV-d in GenBank (M23255).

Following purification, the DNA amplicons were cloned into the pMD18-T vector (TaKaRa Biotechnology Co.) and their sequences determined by Sangon Biological Engineering Technology & Service Co., Shanghai, China.

Phylogenetic analysis

Fifty-two VP2 sequences [the 22 sequences determined here and the 30 from GenBank (Table 1)] were compared with the VP2 sequences of 8 reference strains of CPV (CPV-b, CPV-15, CPV-V120, CPV-39, CPV-Taichung, LCPV-V139, LCPV-V203, and CPV-G7/97) retrieved from the GenBank database. Nucleotide sequences and deduced aa sequences were aligned with the MegAlign program of DNASTAR (DNASTAR, Madison, WI, USA). Phylogenetic analysis was performed using the MEGA program, version 4.0 (http://www.megasoftware.net/). The neighbor-joining method was chosen to draw a nucleotide phylogenetic tree. The reliability of the phylogenetic tree obtained for the VP2 region was evaluated by running 1,000 replicates in the bootstrap test.

Table 1 China-isolated CPV sequences retrieved from GenBank and analyzed in this study
Full size table

Results

Sequence comparisons showed nucleotide identities of 98.3–100% among the CPVs isolated in China since the early 1980s. Nucleotide sequences were translated into aa sequences to identify the isolates as CPV-2, CPV-2a, or other types. Critical positions of the CPV VP2 gene products of isolates sequenced in this study are summarized in Table 2.

Table 2 Amino acid substitutions in the VP2 gene of CPV isolates analyzed in this study
Full size table

As shown, all CPV-2a/2b strains, except one original CPV-2a (without the 297-Ala mutation) collected in 1986, were identified as “new CPV-2a” or “new CPV-2b”, all retaining the aa changes of the variants from the original CPV-2 (Met87Leu, Ile101Thr, Ala301Gly, Asp305Tyr) and displaying the mutation Ser → Ala at site 297 of the VP2 protein. CPV-2 type viruses were identified in one isolate of 1983 and three collected in 2008. There were 12 isolates of “new CPV-2a”, indicating that this has been the predominant CPV in China since the late 1990s, while “new CPV-2b” was isolated with a lower frequency (five isolates). Neither CPV-2c(a), CPV-2c(b) (with 300Asp mutation) nor CPV-2c (with 426Glu mutation) was found.

A mutation Phe → Tyr was encountered at position 267 in CPV-GZ0201, CPV-YN0201, CPV-GZ0202, CPV/WH02/06, CPV-JB1, CPV-JB3 and CPV-08-05-WH. There was a unique mutation within the VP2 of Chinese and Korean strains of “New CPV-2a” at Tyr324Ile. Interestingly, only a single isolate of this variant (CPV-B-2004; Table 1) was detected among all viruses collected before 2005, while 27 of the 33 CPVs collected since 2006 have the mutation Tyr324Ile, caused by mutation TAT → ATT at nt 970–972 of the VP2 gene.

To examine the phylogenetic relationships of the Chinese isolates with representative CPVs, a tree based on the nucleotide sequence from 1 to 1,750 of the VP2 gene was constructed. As shown in Fig. 1, most of the CPVs isolated in China formed a major monophyletic cluster, with one Korean isolate also clustering in this group. A group of four Korean isolates clustered close to the Chinese isolates, indicating a close relationship between the isolates from China and Korea. Eleven of the 13 CPVs collected before 2002 in China fell into a cluster consisting of CPVs isolated from Italy, Vietnam, Thailand, and USA.

Fig. 1
figure 1

Phylogenetic tree based on the VP2 gene sequence (1,750 nucleotides) of the parvoviruses. The last five nucleotides of the VP2 sequences were deleted from the VP2 gene alignment, as some of the sequences were incomplete, resulting in an alignment of 1,750 nt. The sequences from regions outside China used to construct phylogenetic trees were retrieved from GenBank. CPV type 2 isolates: inverted triangle; CPV types 2a and 2b isolates (with 297Ala mutation): filled circle and filled square, respectively; CPV type 2c isolates (with 426Glu mutation): filled triangle; CPV-2c(a) with 300Ala mutation: filled rhombus; the original CPV-2a type (without 297Ala mutation): open circle. Statistical support was provided by bootstrapping over 1,000 replicates and bootstrap values >75% are indicated at the correspondent nodes. Horizontal branch lengths are drawn to scale (nucleotide substitutions per site). Following the names of sequences, in parentheses, are year of isolation, GenBank accession number and the locations. NA not available

Full size image

Discussion

The analysis has indicated that “new CPV-2a” and “new CPV-2b” are co-circulating in China, with the former being isolated with a much higher frequency. “New CPV-2a/2b” appear to have replaced the prototype CPV-2a/2b strains and become the predominant types in many countries [4, 7–10]. It has been reported that mutation Ala297 does not change the viral antigenic type [11], and so the Ala297 variant cannot be distinguished serologically. However, the emergence and spread of this variant indicates that the Ala297 mutation potentially has had a marked influence on the process of continuing host adaptation, and previous research has shown that site 297 is under strong positive selection [12].

This study detected one site mutation among the CPV-2a/2b isolates in China: Tyr → Ile at position 324. This mutation was first detected in China in 2004 and also reported in three isolates from Korea in the same year [13]. Twenty-seven of the 30 CPV-2a/2b isolates collected after 2006 contained this mutation, which was not found in the 17 isolates collected before 2002. Since our study consisted of a limited number of isolates, however, it is not possible to state with certainty that this mutation did not exist in CPVs in China before 2000. The substitution at residue 324 has not been previously identified. Previous studies have shown that residue 324 is subject to strong positive selection in all parvoviruses of carnivores [14]. Residue 324 is adjacent to residue 323, which affects canine transferrin receptor (TfR) binding and, together with residue 93 [15–19], the canine host range. The 324 mutation is likely to have had an effect on the parvovirus host range.

In our study, three isolates collected in 2008 were found to be CPV-2 type, which have a close relationship with three most commonly used vaccine strains indicated by the phylogenetic tree. The vaccination histories of the three puppy hosts are unavailable. Since CPV-2 viruses are no longer considered to circulate in dog populations worldwide, it is likely that the virus detected in the puppies was CPV-2 vaccine virus [20]. Reversion of the CPV vaccine strains to virulence is unlikely, as the attenuation of virulence has proved to be highly stable [21, 22]. When a vaccine strain was detected in a diarrhoeic fecal sample, it was generally present together with CPV field strain that infects the dog shortly before or after the vaccine administration, or with other pathogens commonly associated with enteritis in dogs [23].

As the phylogenetic tree shows, most of the viruses isolated in China formed a large cluster, while some strains clustered together with viruses from regions outside China. Most of the CPVs isolated in China formed a specific cluster and certain mutations detected in Chinese CPVs probably arose during the process of local adaptation, as indicated by previous surveys [12, 24].

There are evidences to suggest that complete immunity may not be provided to pups if CPV2 vaccines are used, considering that the CPV2 vaccine appears to provide a comparatively lower and shorter immunity against heterologous CPVs [4, 25, 26]. In many countries in Europe, such as United Kingdom, Germany, and Italy, CPV-2a has been overtaken by CPV-2b or, recently appeared, CPV-2c, many researches have been conducted to evaluate the immunity conferred by CPV-2 vaccine against the CPV-2b type [26–28]. In Europe, CPV2b-based vaccines have been developed and are available now. Nevertheless, our study showed that CPV-2a is most predominant in China, and the vaccines available in China are CPV-2 type vaccine. The effectiveness of CPV-2 vaccine against CPV-2a type has not been evaluated in China.

In summary, “new CPV-2a” is the prominent type of CPV in China. Most Chinese isolates contained the mutation 324Ile, which probably arose around 2004 when this variant was first detected in China and Korea [13]. CPV-2c, which has displayed an exceptional ability to spread rapidly through the canine population in Italy [8, 29] and other European countries [30], Asia [10] and America [31–33], has not been detected in China. We conclude that due to the continuing evolution of CPV, monitoring of field CPV isolates, and detection of genetic mutation and antigenic changes, will be necessary to control CPV infection in China.