Discussion
In recent studies, Rotavirus A G18P[17] had been confirmed as a primary cause of YPDS-like diseases in domestic pigeons (Rubbenstroth et al., 2020). Furthermore, PiCV is also supposed to be an etiological agent of YPDS mostly affecting young pigeons worldwide (Raue et al., 2005). PiCV has been reported to be prevalent in at least fourteen countries. In China, PiCV was first detected in meat pigeons in 2009 (Yu et al., 2009). In recent years, the epidemiological survey showed that the positive rates of PiCV infection in Chinese meat pigeons were 19.67% and 75.3% in the poultry farms of eastern China in 2009 and 2015 (Wang et al., 2017; Zhang et al., 2015), implying that PiCV is widely distributed among meat pigeon populations in eastern China. However, the epidemiology and distribution of PiCV in the racing pigeons is unknown. The main objective of this work was to evaluate the genetic diversity and epidemiology of PiCV strains circulated in the racing pigeons of China. Positive samples were detected from seven provinces and the prevalence rates among these provinces were variant. Overall, our data implied that PiCV was also widely distributed in diseased racing pigeons and healthy racing pigeons in China for the first time.
Previous studies showed that the length of PiCV genome was 2031-2043 nt (Loiko et al., 2018; Sarker et al., 2019; Wang et al., 2017). However, our findings first showed that the 2030 nt (n=1), 2044 nt (n=2), and 2045 nt (n=1) complete genome were present in the PiCV positive samples. Despite many genetic diversities were found in the genome of PiCV strains, no mutation was found in the conserved nonanucleotide motif (TAGTATTAC) (Mankertz et al., 2000; Todd et al., 2001) located at the apex of a potential stem-loop which was putatively associated with the initiation of rolling circle replication (RCR) (Mankertz et al., 2000; Todd et al., 2001). Previous studies have shown many genetic diversities in the cap gene (Cságola et al., 2012; Stenzel et al., 2014a; Wang et al., 2017). Our findings showed that the 90 identifiedcap genes exhibited a higher diversity as compared with the reference strains. Some unique amino acid substitutions at 28 different positions were observed among the Cap proteins. In addition, thecap nucleotide sequence with 828 nt in length encoding a novel Cap protein of 275 amino acids was first identified in two Chinese PiCV strains, TY3/SN/2016/MW181931 (clade E) and WL4/SN/2018/MW181959 (clade E). Higher diversity of Cap protein versus the Rep protein is due to the fact that the Cap protein, as the protein shell of the virus, is exposed to the host’s immune cells, resulting in a stimulation of a cascade of immune responses. An interaction with the immune system may result in adaptive mutations in the Cap protein. Some mutations can change the structure of Cap protein. This may lead to a better binding of virus with the receptors of target cells, which increases the infection ability of virus in cells. Moreover, these mutations may also protect virus from being neutralized by antibodies (Bassami et al., 1998; Bennett et al., 2006; Stenzel et al., 2017). In this study, we identified many mutations in both N- and C-terminus of the Cap protein. According to the investigations on Cap protein of BFDV, the arginine-rich N-terminus has two functions (nuclear localization and nucleic acid binding) that enable as penetration of the viral genome into the host cell nucleus through nuclear pore complex (Heath et al., 2006). Like BFDV, the N-terminus of PiCV Cap protein is also rich in arginine and has been predicted to be a nuclear localization signal and a nucleic acid binding domain by software, which means that this region is highly possible to be functionally similar to the corresponding region of the Cap protein of BFDV (Lai et al., 2014). Moreover, the results of this study showed that most of the amino acid deletion sites in Cap protein were concentrated in the N-terminus and many novel mutations also have been identified in this region. However, due to the lack of relevant experimental evidence, we cannot infer whether these mutations and deletions could affect the function of Cap protein. The Cap protein has been used as a coating antigen in ELISA due to its antigenic activity and can be recognized by PiCV specific antibodies (Daum et al., 2009; Stenzel et al., 2016). If the N-terminal region of the PiCV Cap protein, which is located within the capsid, resembles the corresponding region in BFDV, the N-terminal amino acid mutation and deletion may not affect the ability to bind to the antibody as a coating antigen (Daum et al., 2009). In addition, there are also many mutations in the C-terminus where most sites are composed of hydrophilic amino acids. It means that this region may locate on the outside of the protein and may have ability for binding with antibody. Therefore, mutations of this region may lead to changes in antigenicity. Cap protein, which contains neutralizing antibody epitopes, can be considered as a potential antigen candidate in sub-unit vaccine development (Gai et al., 2020; Stenzel et al., 2018). Moreover, the sub-unit vaccines based on PCV2 recombinant capsid proteins are successfully used in the prevention of PCV2-SD (Blanchard et al., 2003; Fort et al., 2009; Li et al., 2015; Zhu et al., 2016). It means that the development of a sub-unit vaccine may protect pigeons from infection with PiCV (Stenzel et al., 2018). Due to a high genetic diversity of Cap proteins may lead to differences in antigenicity between different strains, sub-unit vaccines may also be diversified in future. To sum up, since there is little research on the function of PiCV Cap protein, the specific meaning of these mutations cannot be accurately explained. Therefore, further experiments are needed to explain if these mutations and deletions affect the function of Cap protein. According to the obtained nucleotide sequences of cap gene, we found that ATT and GTG existed in the position of the start codon site through sequence alignment. In addition, several ATG was found inside cap genes. Since no available cell lines have been found to culture PiCV in vitro (Huang et al., 2021; Santos et al., 2020; Todd, 2004) and lack of commercial antibodies against PiCV Cap protein, it is hard to verify that indeed a full-length natural Cap protein is translated from the described alternative start codons, rather than a truncated version from a downstream in frame of ATG by experiments now. Therefore, the ATT and GTG were highly suspected to be the start codons of the Cap protein through sequence analysis in this study. A similar approach has been employed for identifying the alternative start codon of Cap protein in other circoviruses, such as GuCV (Todd et al., 2007) and BFDV (Bassami et al., 2001). Moreover, previous studies had confirmed that codons differing from ATG in a single position can support translation initiation in eukaryotes (Peabody, 1989; Wei, J. et al., 2013). Thus, it still needs more experiments to identify if the two non-ATG start codons can be actually used for translation initiation by Cap protein of PiCV. In all 68 identified Rep proteins, three amino acid motifs named FTLNNP (position 41-46), HLQGF (position 78-82), and YCSK (position 116-119) that putatively associated with RCR (Mankertz et al., 2000; Mankertz et al., 1998) were completely conserved. In addition, a fourth motif, which was putatively associated with dNTPase activity, namely GKS (position 197-199) (Mankertz et al., 2000; Mankertz et al., 1998), was also completely conserved. These sequence characteristics help understand the genetic diversity of PiCV. However, the pathogenicity of different PiCV strains still need to be validated in future.
In the present study, the 90 PiCV strains were divided into seven clades (A, B, C, E, G, H, and I) based on a phylogenetic analysis of capgene. Additionally, we found that PiCV strains isolated from the same club belonged to different clades and shared a low sequence identity. These data suggested that the infection and evolution of PiCV in Chinese racing pigeons might have different evolutionary origins. This may result from the fact that the import of racing pigeons from all around the world to China is very significant and a lack of oversight in the international pigeon trade (racing pigeons mainly) (Ashton, 1984; Stenzel et al., 2012). Interestingly, two isolates,QYQX1/HE/2018/MW181909 (clade A) and QYQX2/HE/2018/MW181910 (clade C), with lower identity (75.5% nucleotide identity and 76.8% amino acid identity) in the cap gene were detected from the same pigeon, suggesting a horizontal transmission occurred among the racing pigeons in the same racing clubs (Gerdes, 1993; Woods et al., 1993). The complexity of epidemic PiCV strains in China may cause difficulties to protect pigeon from infection by vaccines in future.
Viral recombination had been proved to play a significant role in the evolution of many ssDNA viruses (Lefeuvre et al., 2009). The extensive recombination events had been reported in PiCV genomes and other circoviruses, such as PCV (Kleymann et al., 2020; Wei, C. et al., 2019) and BFDV (Julian et al., 2013; Varsani et al., 2011). Similar to the previous reports, 31 recombination events were detected in the 67 identified PiCV strains in this study. Thus, the recombination seemed to be a key mechanism for PiCV evolution (Cságola et al., 2012; Loiko et al., 2018; Sarker et al., 2019; Stenzel et al., 2014c).
In conclusion, our study demonstrated that PiCV infection in racing pigeons was widespread in northern China and revealed the characteristics of the PiCV genome. Furthermore, the identified PiCV strains displayed a high genetic diversity. Our data also demonstrated that PiCV in Chinese racing pigeons had an extensive recombination for the first time. Thus, these data could increase our understanding of the epidemiology and genetic variation of PiCV circulated in northern China and evolutionary relationships among different strains.