At present, small non-enveloped DNA circoviruses of concern to animal and human health are classified in two genera (Circovirus and Cyclovirus) in the Circoviridae family. The genus Circovirus consists of 27 species, while the genus Cyclovirus has 43 species [1]. Porcine circoviruses are members of the genus Circovirus with two recognized species, porcine circovirus type 1 (PCV1) and porcine circovirus type 2 (PCV2). PCV1 can infect and replicate in swine, but is not known to cause any obvious disease [2]. While PCV2 infection can result in significant production and economic losses to the worldwide swine industry [3, 4]. Clinically, PCV2 can cause porcine circovirus disease (PCVD) or porcine circovirus-associated disease (PCVAD), which is characterized by postweaning multisystemic wasting syndrome (PMWS), porcine dermatitis and nephrotic syndrome (PDNS), interstitial pneumonia, and reproductive disorders [5, 6].

Recently, a novel circovirus significantly divergent from PCV1 and PCV2, provisionally designated porcine circovirus type 3 (PCV3), was identified in U.S. swine herds [7, 8]. The full-length genome sequence of newly found PCV3 was 2000 nucleotides, which was considerably larger than those of PCV1 and PCV2 (~1760 and 1780 nucleotides, respectively) [9]. Actually, prior to identification of PCV3, PCV3-like sequences were also reported in different farmed animals several years ago, which revealed possible early distribution of PCV3 in animal species [10, 11]. It is interesting that PCV3, similar to PCV2, has a potential association with multi-systemic disease, PDNS and reproductive failure in pigs [7, 8]. At present, PCV3 also has been reported in China, South Korea and Poland, accompanying with widely geographical distribution [12–17]. Despite good progress in PCV3 epidemiology, the circulating status of PCV3 in pigs with different clinical presentations (especilly asymptomatic or diarrhea) was not well established. Therefore, the present study aims to perform comparative epidemiology of PCV3 in pigs with different clinical presentations.

In 2016, two small-scale fattening pig farms (about 200 or 300 American Landrace pigs in total), locating in Guangdong province (southern China) and Jilin province (northern China) separated by more than 3, 000 km in distance (Fig. 1), suffered from severe respiratory disease (SRD). Pigs (2-3 months old) on both farms have received immunizations against PCV2, pseudorabies virus (PRV), classical swine fever virus (CSFV) and type 2 porcine reproductive and respiratory syndrome virus (PRRSV). However, the morbidity rate of pigs on the two farms was approximately 60%, the mortality in affected pigs was about 80%. From autopsy results of 18 individual dead pigs (n = 8 for farm A, n = 10 for farm B) (Additional file 1: Table S1), the similar macroscopic lesions were concentrated in the lungs and lymph nodes, which was characterized by PRRSV- or PCV2-like lesions including pulmonary interstitial widening, congestion and lymphadenopathy bleeding. These 18 pig lung samples were defined as the samples with SRD. To compare the detection results of PCV3 based on the same organ and same age of pigs, additional 80 swine (with SRD, 5-8 weeks old) serum samples, 216 asymptomatic weaned swine (5-8 weeks old) serum samples and 175 diseased weaned swine (5-8 weeks old) (with mild respiratory disease [MRD] including cough, softly panting, abdominal breathing) serum samples (Additional file 1: Table S1) were collected from Shandong province (n = 26), Gansu province (n = 120), Guangxi zhuang autonomous region (n = 34), Guangdong province (n = 130), Sichuan province (n = 65) and Yunnan province (n = 96) (Fig. 1). Moreover, 35 diarrheal weaned pig (6-7 weeks old) samples and 35 non-diarrheal weaned pig (6-7 weeks old) samples (Additional file 1: Table S1) were collected from Gansu province (n = 40) and Neimenggu autonomous region (n = 30) (Fig. 1). The above-mentioned samples were stored −80 °C until further use.

Prior to quantitative PCR (qPCR), 200 μL tissue supernatants or serum samples were used to perform viral DNA extraction according to the manufacturer’s instructions (TIANGEN Biotech Co., Ltd. Beijing, China). Viral DNA samples were stored at −80 °C until use. A previously established qPCR protocol was used, consisting of two specific primers (Forward, 5′–AGTGCTCCCCATTGAACG–3′; Reverse, 5′–ACACAGCCGTTACTTCAC–3′) and one modified probe (5′–FAM–ACCCCATGGCTCAACACATATGACC–BHQ1–3′) [8]. Moreover, the detection of PRRSV was performed according to a previous method [18]. The full-length genome sequence of PCV3 was amplified using the two pairs of overlapping PCR primers according to a previous description [13]. These two PCR fragments were cloned and sequenced. The corresponding sequencing results were spliced using SeqMan program (DNAStar software version 7). Sequence alignment and phylogenetic analysis based on three species of PCV sequences were performed using Clustal W program implemented in DNAStar software and MEGA 5.1 software, respectively. In addtion, qPCR detection results in pigs with different clinical presentations were analyzed by chisquare (χ²) test (Ziyue software), the value of P < 0.05 and P < 0.01 was considered significant and very significant, respectively.

Previous studies suggested that PCV3 was common in PDNS cases and that it might play a role in reproductive failure and multi-systemic inflammation in U.S swine herds [7, 8]. Although there were other swine pathogens (including torque teno sus virus 1, TTSuV1; porcine hemagglutinating encephalomyelitis virus, PHEV; porcine astrovirus 4, PAstV4) present in PCV3-infected tissues, high numbers of PCV3 sequence reads by metagenomic sequencing suggested that PCV3 might play a potentially pathogenic role in those clinical cases [7, 8]. Similarly, immunohistochemistry localized PCV3 in typical PDNS lesions in the absence of PCV2 [8]. However, in four latest studies from China, single PCV3 infection and/or co-infection with PCV2 were reported in clinical cases [12, 13, 15, 16]. Surprisingly, during one of four studies, PCV3 had an abnormally high positive rate (85.7%, 12/14) in pig samples with productive failure, but PCV2, PRRSV and PRV were negative in these pigs [15]. In this study, we identified co-infection of PCV3 with PRRSV in SRD pigs, not in other pigs. Co-infection of pigs with PCV2 and PRRSV is known to exacerbate disease severity [19, 20]. A possible synergy between PCV3 and PRRSV or PCV2 co-infection with disease severity warrants further investigation. Additionally, low Ct values of PCV3 were frequently found in diseased pigs, not in asymptomatic pigs (Fig. 2b), which revealed that the high titers of PCV3, similar to PCV2, could contribute to the occurrence of clinical diseases in pigs [21]. Another interesting finding was that PCV3 was detected in fecal samples, which may suggest for a fecal-oral transmission by PCV3.

To conclude, the present study further extends evidence that newly described PCV3 widely circulates in six additional provinces of Southern and Northern China and has high similarity to previously reported isolates. As an emerging virus of swine, although the present case-control study reveals that PCV3 has a potential association with swine respiratory disease and diarrhea, further investigations into the pathogenesis are needed to ascertain the role of PCV3 in swine health. Similar to PCV2 [22], the usage of viral infectious clone accompanying by clinical and histological methods could resolve this scientific issue.