Introduction

Porcine circovirus type 2 (PCV2) is a highly infectious pathogen that causes immune suppression in pigs. It has resulted in tremendous economic losses in the swine industry. Taxonomically, Porcine circovirus 2 is a member of the genus Circovirus, within the family Circoviridae. PCV2 is one of the smallest DNA viruses infecting mammals, possessing a non-enveloped virion particle of 12–23 nm in diameter [34]. PCV2 is the primary pathogen underlying several syndromes collectively known as porcine circovirus-associated disease (PCVAD), which include postweaning multisystemic wasting syndrome (PMWS) [1, 7] and porcine dermatitis and nephropathy syndrome (PDNS) [2, 35]. The clinical signs of PCV2-infected pigs include weight loss, proliferative and necrotizing pneumonia, enteritis, reproductive disorders in sows, fetal myocarditis, and respiratory disease in weaned and fattening pigs [36].

PCV2 has four major open reading frames (ORFs). These are ORF1–ORF4, which encode the replicase, capsid protein (Cap), viral pathogenesis-associated protein, and apoptosis-suppressing protein, respectively [8, 9, 36]. The Cap protein is the major immunogenic molecule, and virus-like particles of the Cap protein provide effective protection [9]. Due to the lack of a viral envelope, Cap is exposed on the surface of the virion. This has led to remarkable genetic diversity of Cap proteins among viruses. As a result, the Cap protein has proved very useful for epidemiological and phylogenetic studies of PCV2 [5, 25]. Phylogenetic analyses indicated that PCV2 strains could be divided into five genotypes (PCV2a–2e) based on pairwise sequence comparisons of PCV2 isolates [6]. PCV2a could be further subdivided into five clusters (2A–2E), while PCV2b could be subdivided into three clusters (1A–1C) [28]. Previous studies revealed that the majority of Chinese PCV2 strains are of genotype 2a or 2b. A few strains could be classified into genotype 2d. However, no strains of genotype 2c could be isolated [13]. Based on these observations, it has been suggested that the PCV2b genotype has become the dominant viral strain in China in recent years [15, 21, 39].

Although the majority of pigs are vaccinated against PCV2 using killed virus, the incidence of clinical disease in China is still on the rise [37]. In this study, we obtained clinical samples from dead pigs from 16 different provinces. Using sequencing technologies, we analyzed the genetic diversity of these PCV2 strains and identified possible recombination events. Cumulatively, this work helps elucidate important aspects of the molecular genetic evolution of this virus, which is a prerequisite for the future development of effective disease control and prevention strategies.

Materials and methods

Clinical samples

Two hundred and thirty-five tissue samples (lymph nodes) were collected from 235 dead pigs on different farms in 16 provinces in China. Pigs were 4–10 weeks of age, and were suspected to be infected by PMWS and/or PDNS based upon evidence of growth retardation, dyspnea, paleness of the skin, and enlarged lymph nodes upon necropsy (Table 1).

Table 1 List of PCV2 samples isolated in this study
Full size table

PCR amplification and DNA sequencing of ORF2

Tissues were analysed using published PCR methods to identify samples positive for PCV2 [32], porcine reproductive respiratory syndrome virus (PRRSV) [43], porcine parvovirus [26], classical swine fever virus (CSFV) [18], and Mycoplasma hyopneumoniae [16]. From the PCV2-positive samples, ORF2 was amplified using primers: 5′-TGAGTCTTTTTTATCACTTCGT-3′ (position 993–1014 bp) and 5′-CTTACAGCGCACTTCTTTCGT-3′ (position 1743–1763 bp). Thermal cycling conditions were: 95 °C for 5 min; followed by 35 cycles of 94 °C for 1 min, 52 °C for 1 min and 72 °C for 1 min; and a final elongation step at 72 °C for 10 min. PCR products were run on a 1% agarose gel and imaged under ultraviolet light. The positive PCR products (777 bp) were purified with an E.Z.N.A.™ Gel Extraction Kit (OMEGA, Georgia, USA), and cloned using the pMD18-T Vector System (Takara, Dalian, China). Four individual clones for each insert were sequenced by Sanger sequencing (Life Technologies, Shanghai, China), and the consensus sequence was obtained using Vector NTI Suite 9 (InforMax Inc., Maryland, USA).

Phylogenetic analysis

In addition to the sequences obtained from dead animals, we downloaded 496 ORF2 sequences from GenBank, which were identified in 28 provinces in the principal pig farming areas of China. This gave us a total of 648 Chinese PCV2 strains to examine. Representative sequences for PCV2 and PCV1 (GenBank accession no. FJ475129) were used as references. Sequence alignment was carried out using MEGA software (v. 6.0) (Pennsylvania, USA) and the ClustalW algorithm; the identity among sequences, at the nucleotide or amino acid level, was determined using BioEdit (v7.0.5) (California, USA). A phylogenetic tree was constructed using MEGA software based on the cap nucleotide sequence, using the neighbor-joining (NJ) method with the Kimura two parameter model for nucleotide substitution.

Recombination analysis

To detect putative recombination breakpoints in the PCV2 ORF2 gene of the complete (cumulative) dataset, and to identify sequences that might have originated from a recombination event, six methods (RDP, GeneConv, BootScan, MaxChi, Chimaera, and SiScan) were implemented using the RDP program (v. 4.46) (Cape Town, South Africa) [24]. We employed the following settings in these analyses: window size = 20, highest multiple comparison-corrected P value = 0.01, Bonferroni correction, finding consensus daughter sequences, and polishing breakpoints. Only putative recombination events detected by more than one method were considered. Base to base analysis was used to confirm the recombination events detected by the SimPlot program (v. 3.5.1) (Maryland, USA) when compared to the parental strain sequences, as described previously [22].

Results

Sample screening and identification

We collected 235 tissue samples from dead pigs that were suspected to have suffered from PMWS, and we found that 152 were positive for PCV2 infection. Of these, 25 samples were positive for both PCV2 and PRRSV infections, 20 were positive for both PCV2 and CSFV infections, 17 were positive for both PCV2 and pseudorabies virus (PRV) infections, and five were positive for triple infection with PCV2, PRRSV, and CSFV.

Identification of PCV2 genotypes

The NJ phylogenetic tree based on ORF2 sequences from the 648 PCV2 strains showed that all of the Chinese PCV2 strains belonged to four genotypes: PCV2a, PCV2b, PCV2c, and PCV2d (Fig. 1). Among the Chinese PCV2 strains obtained from GenBank, 75 strains (11.6%) belonged to the PCV2a genotype, 246 strains (38.0%) belonged to the PCV2b genotype, and 327 strains (50.4%) to the PCV2d genotype. Only one strain, collected in 2010, was genotyped as PCV2c (GenBank accession no. KC823058). The phylogenetic distances between the genotypes ranged from 0.057 (genotype 2b vs. genotype 2d) to 0.098 (genotype 2d vs. genotype 2a). Within each genotype, the average distances ranged from 0.004 (PCV-2a) to 0.021 (PCV-2d).

Fig. 1
figure 1

Phylogenetic tree of Chinese PCV2 isolates. PCV2 strains (n = 648) were used to construct trees based on the ORF2 gene with MEGA software. The neighbor-joining method was used as statistical method, with the Tamura-Nei model and 1000 bootstrap replications, to assess the reliability of the tree. Strains detected in our lab are indicated by triangles. Genotypes and clusters are indicated by square brackets

Full size image

Mutational analysis of the ORF2-encoded Cap protein

Sequence analysis of the ORF2 gene in the 648 PCV2 strains revealed that nucleotide variation ranged from 89.6–100% and predicted amino acid identity ranged from 70.8–100%. Furthermore, amino acid alignments of the Cap protein encoded by ORF2 indicated that there are five major regions of variation among the PCV2 strains (Fig. 2). These include residues 57–91, 121–151, 181–191, 206–215, and 230–233. The ORF2 amino acid variations at two positions (residue 53: F to I; and residue 68: A to N) were unique to genotype PCV2d. In addition, the amino acid variations at four positions (residue 47: T/A/G to S; residue 130: V/L to F; residue 133: A/V to S; and residue 191: G/A/R/E to K) were unique to genotype PCV2a-2F in the PCV2a cluster.

Fig. 2
figure 2figure 2

Alignment of the capsid protein of PCV2 strains. The ORF2 amino acid variations at positions 53 (F to I) and 68 (A to N) were unique to genotype PCV2d. In addition, the amino acid variations at positions 47 (T/A/G to S), 130 (V/L to F), 133 (A/V to S), and 191 (G/A/R/E to K) were unique to genotypes PCV2a–2F in the PCV2a cluster. Mutation of an amino acid is indicated by a block

Full size image

Detection and analysis of recombinants

Within the 648 PCV2 ORF2 sequences, six were identified as potential recombinant strains (Table 2). The possible breakpoints for recombination were determined (Fig. 3). The nucleotide sequences before the putative breakpoints were very similar to the sequences of the minor parent, while the regions after the breakpoint were most similar to the major parent sequences. Amino acid sequence similarities between recombinants and their parents ranged from 96.4–99.8%.

Table 2 Six recombination events identified among 649 PCV2 ORF2 genes
Full size table
Fig. 3
figure 3

Recombination analysis of the PCV2 ORF2 gene. Recombination events were analyzed by Simplot software (v. 3.5.1). The similarity among 648 ORF2 genes was calculated using the Kimura-2-parameter method with a transition-transversion ratio of 2. The y-axis indicates the percentage identity within a sliding window of 200 bp centered on the position plotted, with a step size between plots of 20 bp. The red vertical line shows the potential breakpoint

Full size image

Discussion

In China, PCV2 infection is very common, and PMWS has been a major problem for the swine industry since a 2002 outbreak caused substantial economic losses to farmers [42]. Since then, PMWS outbreaks have occurred frequently, along with an increase in the incidence of other swine diseases. For example, previous studies reported higher rates of PCV2 comorbidity with porcine parvoviruses [30, 37]. Furthermore, numerous studies have identified an increase in cases of coinfection with viruses including PRRSV, porcine epidemic diarrhea virus (PEDV), PRV, and with Mycoplasma hyopneumoniae [29]. Consistent with this, our results show that a variety of pathogens enhanced PCV2 lesions and disease. It is possible that the comorbidity of combinations of PCVADs varies from region to region, but similar mechanisms might underlie the observed enhancement of the disease phenotype.

Based on phylogenetic studies, a classification scheme for PCV2 was proposed, which divides the viral strains into several major groups based on genotypes identified in different countries [41]. In Malaysia, amino acid sequence analysis of the PCV2 capsid protein (ORF2) revealed that the PCV2b genotype constituted a major subgroup of viral strains [19]. Phylogenetic analysis and comparison with reference sequences demonstrated that PCV2b was most prevalent between 2007 and 2014 in northern Italy [11]. In Taiwan, PCV2a was the most common strain isolated in 2001, but, since 2003, PCV2b has become the predominant subgroup found on pig farms [38]. In Brazil, the results revealed remarkable genetic diversity: all four genotypes currently recognized were detected, including PCV2a, PCV2b, PCV2c and PCV2d [10]. In India, the molecular characterization of PCV2 revealed that individual pigs could harbor multiple genotypes simultaneously, including combinations of PCV2a-2D and PCV2d [3]. Between 2004 and 2008, phylogenetic analyses indicated that PCV2 strains isolated in China could be divided into four genotypes (PCV2a, PCV2b, PCV2d, and PCV2e), and PCV2b was the most common [41]. Between 2009 and 2010, PCV2b became the predominant genotype in mainland China. Base-by-base comparisons of the ORF2 gene sequences indicated that PCV2 evolution traced from PCV2a to PCV2b to PCV2d.

PCV2d was initially identified in 1999 in samples collected in Switzerland. This strain now appears to be widespread in China and has been present in North America since 2010 [42]. From 2012 to 2013, 37% of all PCV2 sequences isolated from US pigs were classified as PCV2d. The study of Mu et al. showed that the approximate percentages of genotypes PCV2a, PCV2b and PCV2c in Henan Province, between 2005 and 2011, were 6.5% (2/31), 93.5% (29/31) and 0%, respectively [27]. Five and 61 of the 66 PCV2 strains belonged to genotypes PCV2a and PCV2b, respectively, indicating that PCV2b was the predominant genotype circulating in southern China from 2011 to 2012 [40]. Our results in this study revealed that the PCV2d genotype constituted 50.4% of all collected samples. Cumulatively, the data suggest an ongoing genotype shift from PCV2b to PCV2d is occurring in pig populations in China. There are four major regions (57–91, 121–151, 181–191, and 230–233) of amino acid variation among the PCV2 strains in our study, which were identified as dominant immunoreactive areas [20]. One study reported that a PCV2 vaccine based on genotype PCV2b was more effective in protecting pigs against the effects of PCV2b infection than those based on genotype PCV2a [31]. Commercial PCV2 vaccines in China are mainly based on the PCV2a and PCV2b genotypes. Vaccines may have become less effective in recent years because of the antigenic variability of PCV2.

The genotype PCV2a may be divided into different clusters. Phylogenetic analysis revealed that new PCV2a isolates were not included in clusters 2A-2E. The new cluster was called PCV2a 2F, or, more simply, PCV2e. PCV2e was first reported in 2009 in China, and constituted approximately 50.0% of all isolates in the PCV2a cluster [39]. However, in our study, strains obtained were predominately of the PCV2e genotype (75.6%) within the PCV2a cluster. The amino acid substitutions specific to the genotype PCV2e were mainly localized to positions 47 (T/A/G to S), 130 (V/L to F), 133 (A/V to S), and 191 (G/A/R/E to K) in the PCV2 capsid protein. These amino acid substitutions are more diverse than those observed in 2009.

We demonstrated that recombination in the ORF2 gene occurred between the PCV2a/2b or PCV2a/2d strains, yielding different recombinants. Some examples of recombination identified by Huang et al. (2013) were not found again here, because the sequence analyses were different (full-length genome in Huang et al. and only the ORF2 gene in our study [17]). The diversity of PCV2 is closely related to these virus recombinants. Intergenotype recombination was found in several countries and regions, including South America, China and India [3, 17, 33]. It also occurred between different genotypes. For example, natural recombination was observed among different lineages of PCV2 strains from Hong Kong [23] and between PCV2a and PCV2b [4]. Gagnon et al. found a type 1 and type 2 PCV recombinant genome that contained the ORF1 of PCV1 and the ORF2 of PCV2a [12]. Recombinant mutants may enhance viral replication and alter antigenicity in vitro [14]. Nevertheless, the recombination between PCV2 strains likely contributes to the genetic variation and diversity that we observed for this virus in the field.

Conclusion

We conducted a comprehensive molecular analysis of Chinese PCV2 strains based on the ORF2 gene. The predominant genotype has shifted from PCV2b to PCV2d. This is of great concern, because current vaccines target only PCV2a and PCV2b strains. Furthermore, there have been shifts within the PCV2a subgroup itself, since we found that PCV2a-2F is now the main genotype. These findings highlight the importance of understanding the composition and dynamics of genetic diversity within the Chinese PCV2 strains because this information is critical for developing effective vaccines and control strategies.