Introduction

Pigeon circovirus (PiCV), also known as columbid circovirus (CoCV), is a small, circular, non-enveloped, single-stranded DNA virus that belongs to the genus Circovirus of the family Circoviridae [1]. PiCV was first recognized in the USA in 1993 [2], and it was later reported in Australia [3] and several European countries, including Northern Ireland [4], England [5], Germany [6], France [7], Italy [8] and Belgium [9]. PiCV infection has been identified in other parts of the world in recent years [10], but little is known about the pathogenesis of the virus. It was reported recently that PiCV is responsible for a complex and variable immune suppression syndrome that mostly affects young pigeons, called “young pigeon disease syndrome” (YPDS) [11]. PiCV infection is generally characterized by a wide range of clinical symptoms, including lethargy, anorexia, weight loss, respiratory diseases, diarrhoea, and other non-characteristic behaviours [12, 13].

The complete genome of PiCV was first cloned from pigeon tissues in 2000 [14]. The average PiCV genome size is 2.0 kb, with minor length variations among different isolates [15]. Viral genomes are characterized by two major open reading frames (ORF), namely ORF V1, composed of 318 amino acids (nt 41–994), encoding the replication-associated protein (Rep), and ORF C1 in the opposite orientation, composed of 274 amino acids (nt 1165–1988) [16], encoding the only structural protein of the virion—capsid protein (Cap) [17]. An additional ORF, ORF C2, is composed of 250 amino acids (nt 411–791), encoding a hypothetical protein [17, 18]. The functions of the other ORFs are unknown.

The family Circoviridae includes two genera: Gyrovirus and Circovirus [19]. Presently, the chicken anemia virus (CAV) is the only member of the genus Gyrovirus, while the genus Circovirus includes several members, including pigeon circovirus (PiCV), porcine circovirus (PCV), psittacine beak and feather disease virus (BFDV) and canary circovirus (CaCV) [20, 21]. Thus far, there are no available culture methods for PiCV, but a PCR assay based on specific PiCV DNA sequences can be used to detect the presence of the virus [22]. In this study, we detected PiCV in Chinese pigeons by PCR and amplified complete PiCV genomes by inverse primer PCR (IP-PCR). We used bioinformatics software to align and analyse the complete PiCV genome sequences from the samples collected in this study as well as those of nine other members of the family Circoviridae available in the GenBank database. We also estimated a phylogeny of PiCV based on analysis of Rep and Cap genes.

Materials and methods

Sample sources

A total of 109 infected pigeons displaying clinical symptoms (such as respiratory signs, weight loss, and diarrhea) and 49 healthy pigeons were collected from poultry farms in the six provinces (Shandong, Anhui, Zhejiang, Shanghai, Hebei and Jiangsu) with the highest pigeon population densities in eastern China (Table 1). Serum or tissue samples were collected from pigeons and immediately frozen at −80 °C.

Table 1 Pigeon serum and tissue samples used for PiCV detection
Full size table

Reagents

DNA and gel extraction kits were purchased from Geneaid (USA). DL2000 DNA Marker, Pfu DNA polymerase, Taq DNA polymerase, 10× buffer, dNTP mixture and a pMD18-T vector kit were purchased from TaKaRa (Dalian, China). DH5α competent cells were purchased from Tiangen (Beijing, China). All reagents were high-purity and intended for analytical use.

Primer design and synthesis

Two pairs of primers, one for PiCV detection (PiCV-1/2) and the other for amplification of complete genomes (PiCV-JST1/2), were designed according to the published genome sequence of PiCV available in the GenBank database (Table 2). Primers were synthesized by Invitrogen (Shanghai, China).

Table 2 Primer sequences used in this study
Full size table

PCR detection of PiCV

Genomic DNA was extracted from 200 μl of serum (10 %, w/v) or from approximately 25 mg of tissue samples using a commercial DNA extraction kit according to the manufacturer’s instructions. The extracted DNA was eluted twice with 25 μl of sterile distilled water to give a final extract volume of 50 μl.

The genomic DNA was used as template to perform PCR for the detection of PiCV. PCR reaction mixtures (25 μL) contained 18 μL of ddH2O, 2.5 μL of 10× buffer, 1 μL each of primer pairs PiCV-1 and PiCV-2 (10 μM each), 2 μL of DNA template, and 0.5 μL of Taq DNA polymerase (5 U/μL). PCR was initiated by heating for 5 min at 94 °C, followed by 40 cycles of 30 s at 94 °C, 30 s at 60 °C, 40 s at 72 °C and a final extension for 10 min at 72 °C. The amplification products (2 μL) were analyzed by electrophoresis in a 2  % TAE agarose gel and visualized under ultraviolet light.

Cloning and sequencing of the PiCV genome

Complete PiCV genomes were amplified by IP-PCR using the inverse primers PiCV-JST1 and PiCV-JST2. Templates were randomly chosen PiCV-positive DNA samples from six poultry farms (one from each geographic region). IP-PCR reaction mixtures (50 μL) contained 21 μL of Pfu DNA polymerase, 4 μL of DNA template, 2 μL each of the primers, and 21 μL of ddH2O. The PCR protocol was the same as described above. PCR amplification products with the expected sizes were purified using a commercial gel extraction kit according to the manufacturer’s instructions.

The recovered DNA was cloned into the pMD18-T vector and used to transform DH5α competent cells using the heat shock method. The positive clones were confirmed by PCR and Sanger sequencing.

Sequence alignment and analysis

Sequences were first identified by BLAST search against the NCBI database. DNASTAR 7.1 and ClustalX 1.8 software were used to perform alignments and sequence comparisons [23]. Phylogenies were inferred from distance matrices using the neighbor-joining method with a bootstrap test of 1000 iterations in MEGA5 [24].

Results

PCR detection of PiCV infection

An approximately 338-bp specific fragment was amplified. The specific PCR products were identified as PiCV by sequencing and BLAST analysis. Of the 158 samples, 119 were PiCV positive (positive rate, 75.3 %). Among the samples, 88 samples from the 109 unhealthy pigeons (80.7 %) and 31 samples from the 49 healthy pigeons (63.3 %) were PiCV positive (Table 3). PiCV was detected in samples from both diseased and healthy pigeons.

Table 3 PiCV infection results
Full size table

Amplification of PiCV genomes

An approximately 2040-bp complete genome fragment was obtained from PiCV-positive samples from six poultry farms. BLAST alignment results confirmed that six of the recovered genomes were PiCV, and they were designated as the AHBZ, HBLF-E2, JSJN, NJPK-21, SDDZ and SHWH-AB4 strains. The GenBank IDs of these six PiCV complete genomes were KJ704801, KJ704802, KJ704803, KJ704804, KJ704805, and KJ704806, respectively.

Sequence analysis of PiCV genomes

PiCV is a member of the genus Circovirus and the family Circovirdae, so we included nine publicly available genomes to represent the diversity of the family Circovirdae. The genomes were downloaded from GenBank and included chicken anemia virus (CAV), porcine circovirus (PCV), duck circovirus (DuCV), goose circovirus (GoCV), beak and feather disease virus (BFDV), canary circovirus (CaCV), gull circovirus (GuCV), starling circovirus (StCV), raven circovirus (RaCV), and six PiCV strians (PiCV-zj1, NI7050, Ger1, Bel936, FraA4042, US93A) from China, the USA, and four European countries (Table 4).

Table 4 GenBank sequences used in this study
Full size table

Sequence analysis demonstrated that the PiCV genome contained two main ORFs, ORF V1 (ORF 1) and ORF C1 (ORF2). As in other members of the genus Circovirus, these genes encode the viral replication-associated protein (Rep) and the viral capsid protein (Cap), respectively. ORF V1 corresponds to nt 41–994 in the nucleotide sequence and encodes 318 amino acids. ORF C1 corresponds to nt 1165–1988 in the nucleotide sequence and encodes 274 amino acids. Notably, the PiCV coding sequence region of ORF V1 and ORF C1 had a conserved nine-nucleotide sequence (TAGTATTAC; Table 5).

Table 5 Characteristics of ORFs from different PiCV strains
Full size table

Sequence similarity of PiCV genomes

The 12 complete PiCV genome sequences were analysed to assess sequence similarity. The percent sequence similarity among the six complete PiCV genomes in this study ranged from 88.0 % to 100 %, and from 86.9 % to 100 % in a comparison with six other PiCV genomes (Fig. 1), suggesting that PiCV strains maintain an evolutionary signal despite their distribution, and there were no obvious geographical differences in these strains.

Fig. 1
figure 1

Percent identity of 12 PiCV genome sequences based on amino acids. Samples 1–6 are PiCV sequences obtained in this study. Samples 7–12 are PiCV sequences from GenBank

Full size image

We also assessed the sequence similarity of Rep and Cap genes of the 12 complete PiCV genomes. The sequence similarity of the six Rep and Cap datasets generated in this study ranged from 93.9 % to 100 % and 83.1 % to 98.5 %, respectively. The sequence similarity ranged from 93.8 % to 100 % and 79.1 % to 100 % in a comparison with the six other Rep and Cap sequences available from GenBank (Fig. 2 and Fig. 3). The PiCV Cap sequences were more variable than the relatively conserved Rep sequences.

Fig. 2
figure 2

Percent identity of 12 PiCV strains based on Rep protein amino acid sequences. Samples 1–6 are PiCV sequences obtained in this study. Samples 7–12 are PiCV sequences from GenBank

Full size image
Fig. 3
figure 3

Percent identity of 12 PiCV strains based on Cap protein amino acid sequences. Samples 1–6 are PiCV sequences obtained in this study. Samples 7–12 are PiCV sequences from GenBank

Full size image

Sequence analysis of circoviruses

The sequence similarity rate of the 11 circovirus genome sequences from GenBank (Table 4) ranged from 30.2 % to 82.7 %, and the sequence similarity between the 11 circoviruses and the 12 PiCV genomes ranged from 31.4 % to 65.0 % (Fig. 4). It was notable that the sequence similarity between the CAV genome and 22 other circovirus genomes ranged from 30.2 % to 35.7 %, indicating that the only member of the genus Gyrovirus, CAV, is distantly related to the other circoviruses.

Fig. 4
figure 4

Percent identity of 23 circovirus genomes. Samples 1–6 are PiCV strains obtained in this study. Samples 7–12 are PiCV strains and samples 13–23 are circovirus sequences from GenBank

Full size image

In addition, we constructed a phylogenetic tree based on the 23 complete genomic sequences of the family Circoviridae listed in Table 4 (Fig. 5). In the phylogenetic tree, the 23 circovirus genomes formed two distinct groups. The first branch included the 22 genomes of circoviruses from different hosts, suggesting that they had a close genetic relationship. As the only member of the genus Gyrovirus, one CAV strain was grouped into the second branch alone.

Fig. 5
figure 5

Neighbor-joining tree of 23 circovirus genomes. “●”indicates PiCV genomes obtained in this study, “○” indicates PiCV sequences from GenBank. The 12 PiCV strains are within the shaded box

Full size image

Discussion

As with other avian circoviruses, PiCV was detected in a variety of tissues and is therefore likely to be present in many secretions [25]. However, there is little information on the prevalence of PiCV infection [26]. We found that PiCV is widely distributed in eastern Chinese pigeons, and there are no obvious differences in the geographical distribution of this virus. The fact that PiCV was detected in both healthy and diseased pigeons implies a potential threat to the pigeon-breeding industry. Histopathological examinations have revealed lymphocyte depletion in the primary and secondary lymphoid tissues in PiCV-infected pigeons, and PiCV infection is thought to cause immune suppression, like other circovirus infections [27]. When the immune system is compromised, secondary infections determine the clinical signs of illness [28, 29], and the initial PiCV infection may be overlooked.

There are currently no serological tests to diagnose PiCV, and the virus has not yet been grown in vitro [30]. Original diagnoses of PiCV infection were dependent on histopathological examination and electron microscopy, but recent diagnostic approaches have focused on detecting the specific PiCV DNA, and a number of PCR methods have been developed and applied [31–33]. The pairwise comparison of 12 complete PiCV genomes as well as Cap and Rep genes showed that regardless of where the strains were collected, the level of nt sequence identity was very similar, and differences in geographic origin play only a minor role in their genetic relationships. It is notable that the sequence similarity of Rep genes (93.8 %–100 %) was much higher than that of Cap genes (79.1 %–100 %) among the PiCV genomes. It is well known that ORF V1 is located on the sense strand, while ORF C1 is located on the complementary strand [34]. The 5′ intergenic region is located between the start sites of the two ORFs and contains a highly conserved nucleotide sequence (TAGTATTAC), the initiation site of rolling-circle replication [35]. Thus, conserved nucleotide sequences and 5′ intergenic regions within the PiCV Rep genes would be suitable for inclusion in diagnostic PCR primers, while variable sequences within the Cap genes should be considered for phylogenetic studies and for selecting virus isolates for vaccine development [36].

Our study confirms that PiCV should be included as a member of the family Circoviridae based on the close relationship of PiCV strains to other circoviruses, the size of the virus, and associated pathologic features [37]. Further research on the family Circoviridae will likely identify more members of this interesting viral lineage. In this study, we included 10 members of the family Circoviridae whose genome sequences were available in GenBank, including nine circoviruses and one gyrovirus [38, 39]. Members of the genus Circovirus formed a group in the phylogenetic tree, and there were no discernible evolutionary patterns related to host differences or geographical distribution (Fig. 5). The two genera of the Circoviridae family, Circovirus and Gyrovirus, are reciprocally monophyletic according to these preliminary data (Fig. 4) and have a different evolution process. The 12 PiCV samples formed a clade and had a close relationship to each other regardless where they were collected.

PiCV infection affects healthy and diseased young pigeons and is thought to be responsible for immune suppression, resulting in complex and variable syndromes. This virus affects pigeons across the world, and outbreaks may have a devastating impact on poultry farming if the virus is left unchecked. Our study provides information about the prevalence of PiCV in eastern China, reveals characteristics of PiCV genomes, and lays the foundation for further research about genetic variation and molecular evolutionary relationships in the family Circoviridae. Future research should aim to include more samples of PiCV from across its distribution area and should explore the pathogenic mechanism of PiCV and methods for culturing this virus.