Co-Circulation of the Rare CPV-2c with Unique Gln370Arg Substitution, New CPV-2b with Unique Thr440Ala Substitution, and New CPV-2a with High Prevalence and Variation in Heilongjiang Province, Northeast China

To trace evolution of canine parvovirus-2 (CPV-2), a total of 201 stool samples were collected from dogs with diarrhea in Heilongjiang province of northeast China from May 2014 to April 2015. The presence of CPV-2 in the samples was determined by PCR amplification of the VP2 gene (568 bp) of CPV-2. The results revealed that 95 samples (47.26%) were positive for CPV-2, and they showed 98.8%–100% nucleotide identity and 97.6%–100% amino acid identity. Of 95 CPV-2-positive samples, types new2a (Ser297Ala), new2b (Ser297Ala), and 2c accounted for 64.21%, 21.05%, and 14.74%, respectively. The positive rate of CPV-2 and the distribution of the new2a, new2b and 2c types exhibited differences among regions, seasons, and ages. Immunized dogs accounted for 48.42% of 95 CPV-2-positive samples. Coinfections with canine coronavirus, canine kobuvirus, and canine bocavirus were identified. Phylogenetic analysis revealed that the identified new2a, new2b, and CPV-2c strains in our study exhibited a close relationship with most of the CPV-2 strains from China; type new2a strains exhibited high variability, forming three subgroups; type new2b and CPV-2c strains formed one group with reference strains from China. Of 95 CPV-2 strains, Tyr324Ile and Thr440Ala substitutions accounted for 100% and 64.21%, respectively; all type new2b strains exhibited the Thr440Ala substitution, while the unique Gln370Arg substitution was found in all type 2c strains. Recombination analysis using entire VP2 gene indicated possible recombination events between the identified CPV-2 strains and reference strains from China. Our data revealed the co-circulation of new CPV-2a, new CPV-2b, and rare CPV-2c, as well as potential recombination events among Chinese CPV-2 strains.


Introduction
Canine parvovirus-2 (CPV-2) causes a highly contagious and often fatal disease characterized by vomiting and hemorrhagic gastroenteritis in dogs of all ages, and by myocarditis in pups of less than three months of age [1]. The vaccine could provide protective immunity against the classic CPV-2 infection in canine populations, reducing the mortality of animals and spreading of the virus. However, the long-term interaction of the virus and its host may have led to the emergence of CPV-2 antigenic variants. During the past decades, new antigenic types of the CPV-2a, CPV-2b, CPV-2c, new CPV-2a, and new CPV-2b were frequently reported in the dog population, and they replaced the original CPV-2, which has caused widespread concern [2][3][4][5][6][7][8][9]. Therefore, molecular epidemiological investigation of CPV-2 is an undergoing hotspot issue worldwide.
The VP2 protein of CPV-2 is the major capsid protein, and it is highly antigenic, playing important roles in determining viral host ranges and tissue tropisms [10,11]. At present, VP2 is widely used for the identification and monitoring of the types of CPV-2 circulating in the canine population based on the amino acid residues of the VP2 protein at positions 426 Asp/ Glu/Asn and 297 Ser/Ala [10,[12][13][14][15]. In the current study, the molecular epidemiology of the CPV-2 strains circulating in northeast China was investigated by polymerase chain reaction (PCR) targeting the partial VP2 gene. Moreover, potential recombination events of the identified CPV-2 strains were analyzed using the selected entire VP2 gene. Our aim was to provide insights into the epidemiology and genetic diversity of the CPV-2 strains circulating in northeast China, which can be useful for preventing and controlling CPV-2 infections.

Ethics statement
All animal experiments were conducted according to the Guide for the Care and Use of Laboratory Animals of Harbin Veterinary Research Institute (HVRI), Chinese Academy of Agricultural Sciences, China. Sampling and data publication also were approved by animal's owners. The field study did not involve endangered or protected species. No specific permissions were required for locations of samples because the samples were collected from public areas or nonprotection areas.

Sampling
Commercial virus sampling tubes (YOCON Biological Technology Co. Ltd., Beijing, China) containing 3.5 mL of Hank's balanced salt solution, streptomycin, and penicillin, were used to collect fecal samples. A total of 201 fecal samples were collected in the form of rectal swabs from animal hospitals of the Harbin, Daqing, and Mudanjiang districts of Heilongjiang province in northeast China from May 2014 to April 2015. These animals, which were aged between 1 month and 1 year, included animals with symptoms of different degrees of diarrhea, animals with or without a vaccination history, and animals from different breeds and both genders. Of the 201 samples, 141 were collected in Harbin, 20 were collected in Daqing, and 40 were collected in Mudanjiang. All rectal swab samples were stored at −80°C, and they were also used for etiological investigations in our other studies.
China) according to the manufacturer's instructions. The extracted genomic DNAs were stored at −40°C. A commercial vaccine was used as a positive control, and distilled water was used as a negative control.

Phylogenetic analysis
For the phylogenetic analysis, the VP2 gene sequences of CPV-2, CPV-2a, CPV-2b, CPV-2c, new CPV-2a, and new CPV-2b strains from different geographical locations within China and the rest of the world were retrieved from the NCBI nucleotide database. A nucleotide phylogenetic tree of the VP2 fragments (507 bp) of CPV-2 was generated from the ClustalX-generated alignments by MEGA6.06 software using the neighbor-joining method [16]. Neighbor-joining phylogenetic tree was built with the p-distance model and 1000 replicates of interior-branch test; otherwise, the default parameters in MEGA 6 were used. The support of reliability (above 50%) was shown in the obtained phylogenetic tree.
Sequencing of the entire selected VP2 genes A total of 13 entire VP2 genes identified in our study, representing different genetic diversity in the phylogenetic tree of the 507 bp VP2 fragments of CPV-2, were chosen for recombination analysis. The entire VP2 gene of CPV-2 was amplified using the primers CPVvp2-F (5'-AAC TAAAAGAAGTAAACCAC-3') and CPVvp2-R (5'-TGTAATAAACATAAAAACAT-3'), which resulted in a 1882 bp fragment covering the open reading frame of the VP2 gene. PCR amplifications were conducted in a 50 μL reaction volume containing 0.1 μM of forward primer, 0.1 μM of reverse primer, 4 μL of genomic DNA, 5 μL of 10 × buffer, 0.5 U of ExTaq DNA polymerase (TaKaRa), and an appropriate volume of ddH 2 O. The mixtures were amplified by 35 cycles of 94°C for 1 min, 50°C for 1 min, and 72°C for 2 min, with a final extension at 72°C for 10 min using an automated Applied Biosystems GeneAmp PCR System 9700 thermal cycler. The VP2 PCR products were purified using the AxyPrep DNA Gel Extraction kit (Corning), and then purified products were directly subjected to Sanger sequencing. Each sample was sequenced three times. All sequences were submitted to the NCBI Genbank (http://www.ncbi.nlm.nih.gov) under the accession numbers KT156825-KT156837.

Recombination analysis
Thirteen entire VP2 sequences in this study and 25 VP2 gene reference sequences from different areas in China were used for the recombination analysis ( Table 1). The 38 VP2 genes were aligned in the ClustalX (1.83) program, and then were screened for possible recombination by using the Recombination Detection Program (RDP) [17], GENECONV [18], BOOTSCAN [19], MaxChi [20], CHIMAERA [21], and SISCAN [22] methods embedded in RDP4. Only potential recombination events detected by two or more of the programs, coupled with phylogenetic evidence of recombination, were considered significant using the highest acceptable P-value cutoff of 0.05.

Investigation and genotyping of CPV-2
A total of 201 samples from the Harbin, Daqing, and Mudanjiang districts of Heilongjiang province in northeast China were detected via PCR amplification of 568 bp of the VP2 gene of CPV-2. Nucleotide sequences of the partial VP2 of CPV-2 strains identified in our study were shown in Supporting Information (S1 Fig). The characteristics of the CPV-2-positive dogs, the genotypes of CPV-2 strains, and the amino acid substitutions of the VP2 protein are shown in Table 2, and a further analysis of the CPV-2-positive samples is shown in Table 3. Of the 201 samples, 47.26% (95/201) were positive for CPV-2, and three types, new2a (Ser297Ala), new2b (Ser297Ala), and 2c, were found, accounting for 64.21%, 21.05%, and 14.74%, respectively. The new2a type was identified in Harbin, Daqing, and Mudanjiang; the new2b type was identified in Harbin and Daqing; and the 2c type was identified in Harbin and Mudanjiang. Sequence comparisons revealed 99.0%-100%, 99.8%-100%, and 100% nucleotide identities within the new2a, new2b, and 2c strains, respectively, and 98.8%-100% nucleotide identities between the new2a, new2b, and 2c types, respectively. At the amino acid level, 98.8%-100%, 100%, 100%, 97.6%-100% identities were revealed within the new2a, new2b, and 2c strains, and between the new2a, new2b, and 2c types, respectively. A total of five substitutions of amino acid residues occurred in all 95 CPV-2-positive samples based on an analysis of partial VP2 sequences. Residue 324 Ile was present in all 95 CPV-2-positive samples when compared with the reference strains, and residue 370 Arg was specific for type CPV-2c when compared with CPV-2c reference strains. Only three new CPV-2a strains, HRB-K3, HRB-G0, and HRB-F8, showed amino acids substitutions (Asn!Ser) at position 419. At position 440, all new2b strains and most of the new2a strains identified in our study showed amino acids substitutions (Thr!Ala) when compared with the reference strains.

Phylogenetic analysis
To analyze the genetic diversity of the new2a, new2b, and 2c strains identified in northeast China, partial VP2 gene sequences of CPV-2, CPV-2a, CPV-2b, CPV-2c, new CPV-2a, and      strains from northeast China were divided into two different groups: GI and GII (Fig 1). All 14 CPV-2c strains and 20 new CPV-2b strains identified in our study were divided into the GI group. Meanwhile, the GI group included the new2a, new2b and 2c reference strains from China. The GII group, consisting of four subgroups, GIIa, GIIb, GIIc, and GIId, was specific for the new CPV-2a strains, except for one new CPV-2b strain (HRB-J8). The new CPV-2a strains identified in northeast China were divided into the three subgroups GIIa, GIIb, and GIIc, and the new CPV-2a reference strains from other countries were divided into the subgroup GIId. The GIV group consisted of different genotypes of CPV-2 reference strains from China and other countries.

Recombination analysis
The recombination analysis based on the VP2 gene revealed that two potential recombination events occurred between the selected 13 strains in this study and the 25 reference strains from China, but no potential recombination events were found within the selected 13 strains in this study. Nucleotide sequences of the entire VP2 of CPV-2 strains identified in our study were shown in Supporting Fig 2A, which used HRB-F8 and G1 as minor parent strain, and major parent strain, respectively. The recombination event was confirmed by fast neighbor-joining trees that were constructed using the regions derived from minor parent strain (1-133 and 558-1755) (Fig 2B), the recombinant region (134-557) ( Fig  2C), and the non-recombinant region (Fig 2D). Fig 3 shows the recombination event that occurred between MDJ-20 strain (KT156829) and a G1 strain (KF482468), which led to the recombinant 08/09 strain (GU380305). The bootscan plot of this recombination event is shown in Fig 3A, which used MDJ-20 and G1 as major parent strain, and minor parent strain, respectively. The recombination event was confirmed by fast neighbor-joining trees that were constructed using the regions derived from minor parent strain (1-809 and 1708-1755) ( Fig  3B), the recombinant region (810-1707) (Fig 3C), and the non-recombinant region (Fig 3D). The two recombinant strains MDJ-20 and 80/09 were divided into different clusters in the recombination region-based phylogenetic trees when compared with the non-recombination region-based phylogenetic trees.

Discussion
It has been reported that a high proportion (29.2%-84%) of fecal samples from diarrheaaffected dogs tested positive for CPV-2 [11,29]. In our study, 201 samples were collected from diarrhea-affected dogs in three districts in Heilongjiang province in northeast China, and 95 CPV-2-positive samples were identified, accounting for 47.26% of all samples. The results were in line with recent reports in China [15,30]. The CPV-2-positive rate differed among the three districts (Harbin, 48.94%; Daqing, 30%; Mudanjiang, 50%). In our study, mixed infections of CCoV (22.11%), CaKV (14.74%), and CBoV (7.37%) were found in the CPV-2-positive samples. A coinfection of CCoV, CBoV, and CaKV with CPV-2 was also found in one sample. Cavalli et al. (2014) reported that mixed infections of CPV-2 and CCoV reached 49.12% in diarrheic dogs in Albania [31]. CaKV and CBoV had been reported to be associated with severe enteritis in a litter of puppies [28,32]. The high prevalence of co-infections of CPV-2 with three enteric viruses may be associated with viral diarrhea in dogs in northeast China.  -KT074269 and KT074272 from Mudanjiang; KT074273 and KT074274 from Daqing; KT074278, KT074279, KT074280-KT074287, KT074289-KT074291, KT074294-KT074298, KT074300-KT074302, KT074305, KT074307-KT074310, KT074312, KT074313, KT074316, and KT074318 from Harbin. doi:10.1371/journal.pone.0137288.g001 Co-Circulation of CPV-2c, New CPV-2b and New CPV-2a in Northeast China months, which is similar to the report described by Cavalli et al. (2014) [31]. At present, most dogs in China are vaccinated for CPV-2. However, in the CPV-2-positive samples identified in this study, 48.42% came from dogs with a vaccination history. This data demonstrated that the CPV-2 vaccines used in northeast China may fail to generate protective antibody titers against heterogeneous CPV antigenic types.
The Ser297Ala mutation may have had a remarkable influence on the process of host adaptation [33]. This mutation was used as a marker of the new CPV-2a/2b variant. This study indicated that the new CPV-2a (Ser297Ala) variant, accounting for 64.21% of CPV-2 infections, was the predominating strain circulating in the three districts of Heilongjiang province in northeast China. The new CPV-2b (Ser297Ala) variant, accounting for 21.05% of CPV-2 infections, was the second most predominant strain circulating in the Harbin and Daqing districts of Heilongjiang province. In our study, the prevalence of the new CPV-2a is in agreement with most of the reports that were recently conducted in China [8,15,30,34,35], African countries [36], other Asian countries, including Korea, Thailand, Taiwan, and Turkey, and Australia [6,[37][38][39]. However, the prevalence of the new CPV-2b variant reported here in CPV-2-positive samples was higher than those in other reports in China. Interestingly, the presence of the rare CPV-2c, accounting for 14.74% of the CPV-2-positive samples, was confirmed in the Harbin and Mudanjiang of Heilongjiang province. The co-circulation of the CPV-2c, new CPV-2a, and new CPV-2b has previously not been reported in China. It has been reported that the CPV-2c variant is widely distributed and co-exists with other CPV types in Europe and North and South America [40][41][42][43][44][45]. Compared with an earlier report that detected the CPV-2, new CPV-2a, and CPV-2b variants in China [46], the emergence of the CPV-2c in northeast China suggests that CPV-2 strains circulating in China exhibit genetic variations. Additionally, our results are similar to those reported for the genotypes of CPV-2 strains circulating in India, where the new CPV-2a and new CPV-2b were the predominant strains [14], and the co-circulation of the CPV-2c was also found [47].
In the VP2 protein of CPV-2, the Tyr324Ile mutation is adjacent to residue 323, which affects binding to canine transferrin receptors, resulting in changes in the host range of canine parvovirus [13]. Presumably, this mutation may result in stronger receptor binding. In our study, all 95 identified CPV-2 strains, including the new2a, new2b, and 2c types, contained the unique Tyr324Ile mutation. The Tyr324Ile mutation in CPV-2a strains has been reported in China [8,15,30,46], South Korea [38], Thailand [48], Uruguay [49], Japan [50], Taiwan [6], and India [51]. These data support the view that the Try324Ile substitution in CPV-2 strains may be a common amino acid alteration in Asia countries, especially in China. This conclusion needs to be validated by an extensive epidemiological survey in a future study. In addition, further studies should be conducted to understand the relationship between the Try324Ile substitution and the severity of clinical symptoms.
In the VP2 protein of CPV-2, the amino acid residue at position 440 is located in the GH loop. It has been reported that the high levels of the Thr440Ala substitution in VP2 are associated with the evolution of antigenic variants in circulating parvovirus types [52,53]. In our study, the Thr440Ala substitution was found in 64.21% of the 95 CPV-2-positive samples, of which 100% of the new2b strains had the Thr440Ala substitution, 67.22% of the new2a strains exhibited the Thr440Ala substitution, while the type 2c strains did not have the Thr440Ala substitution. At present, the Thr440Ala substitution has been frequently reported in China and other countries [8,14,15,36]. Therefore, we suggest further studies to better understand the relationship between this mutation and the severity of clinical symptoms.
In this study, both new CPV-2b strains and CPV-2c strains exhibited 100% amino acid identity on the basis of the deduced sequence of the partial VP2 protein among different districts. These data demonstrated that type new2b and 2c strains of CPV-2 circulating in northeast China exhibited genetic stability. However, compared with CPV-2 reference strains, the type new2b strains showed a unique Thr440Ala substitution (100%), and the type 2c strains exhibited a unique Gln370Arg substitution (100%). The amino acid substitutions resulted in phylogenetic changes in the new2b and 2c strains identified in our study. The new2b and 2c strains (except one strain) differed genetically from the reference strains of other countries, forming one group (GI) with Chinese reference strains of the new2b and 2c types. In our study, the close relationship between the new 2b and 2c types suggests that these two types share a similar evolutionary history in China. For the identified type 2c strains, the unique Gln370Arg substitution, which was previously not described, was highlighted as evidence of a potential CPV-2c variant or new CPV-2c. Further studies regarding potential variant CPV-2c strains should be conducted to understand the relationship between the Gln370Arg substitution and viral pathogenicity.
The phylogenetic analysis revealed that the type new2a strains identified in our study exhibited high variability, forming three subgroups. Of the 61 new2a strains, a total of two amino acid substitution sites were revealed, Asn419Ser and Thr440Ala. The two amino acid substitutions, as well as some non-synonymous substitutions, accounted for the genetic diversity of the new2a strains in the phylogenetic tree. In our study, the Asn419Ser substitution was detected in only three of the 61 new2a strains. We suggest that the Asn419Ser substitution, first described here, should be the focus of further epidemiological investigations of CPV-2. Accumulating reports have indicated that the new CPV-2a strains, accounting for high positive rates among CPV-2 types, are the predominating strains circulating in China. The high variability of the new CPV-2a strains in our study may be attributed to long-term interactions between the new CPV-2a strains and their hosts. It is suggested that vaccinations against CPV-2 should be updated on the basis of these new CPV-2a strains.
Although point mutations are considered the main mechanism for generating genomic diversity in CPV, co-infection and recombination have also been explored as variability-inducing mechanisms [54]. Pérez et al. (2014) revealed that a recombinant strain arose from inter-genotypic recombination between CPV-2c and CPV-2a strains within the VP1/VP2 gene boundary [55]. In our study, two inter-genotypic recombination events between the 13 CPV-2 strains in this study and the Chinese reference strains were identified in the VP2 gene. The two recombination events occurred between type new2a and 2c strains, resulting in two potential recombinant type new2a strains. Our finding is in line with the occurrence of a possible recombination event in the CPV-2 strain 364-rec, as described by Pérez et al. (2014) [55]. Although the evidence of naturally occurring recombination events was not obtained in our study, possible recombination events among Chinese CPV-2 strains suggest that co-infections with CPV-2 strains with different genotypes, especially types 2a and 2c, as well as potential recombinant strains of CPV-2, should be monitored in China by extensive molecular epidemiology investigations in the future.
Since 1983, the epidemiological information of CPV-2 strains circulating in China has been frequently reported [8,15,30,34,35,46]. Accumulating reports indicated that circulation of CPV-2, CPV-2a, CPV-2b, new CPV-2a, and new CPV-2b strains was identified in different districts of China, of which new CPV-2a was predominant strains in China, followed by new CPV-2b. Compared with the previous epidemiological investigation of CPV-2 in China, the detailed information of samples used in our study, including collection date, location, animal breed, animal gender, animal age, animal immunization, and coinfection with other enteric viruses, resulted in the generation of more precise and valuable epidemiological data which not only enriches basic epidemiological data of Chinese CPV-2 strains, but also increases the understanding of epidemic characteristics of CPV-2 strains circulating in northeast China. In addition, our findings revealed novel epidemiological information of CPV-2 in China, involving in the co-circulation of new CPV-2a with high variation, new CPV-2b, and rare CPV-2c in Heilongjiang province of northeast China, the unique amino acids substitutions including Gln370Arg substitution in identified CPV-2c strains, and Thr440Ala substitution in identified new CPV-2b strains, and the potential recombination events which could happen among type new2a and 2c CPV-2 field strains circulating in China.

Conclusions
Taken together, in the current study, the resulting data increase our understanding of the genetic evolution of CVP-2 strains in China, and they also provide valuable information for further studies of CPV-2 in other countries. However, some speculations resulting from our data still need to be validated by further and extensive epidemiological investigation in the future.