First detection and genetic characterization of canine bufavirus in domestic dogs, Thailand

Canine bufavirus (CBuV) was reported in domestic dogs worldwide. We conducted a survey of canine bufavirus in domestic dogs in Thailand from September 2016 to October 2022. Rectal swab samples (n = 531) were collected from asymptomatic dogs and dogs with gastroenteritis signs. The samples were tested for CBuV using PCR with specific primers to the VP1/VP2 gene, and 9.42% (50/531) was CBuV positive. Our findings showed that CBuVs could be detected in both symptomatic and healthy dogs. The Thai CBuVs were found in dogs from different age groups, with a significant presence in those under 1 year (12.60%) and dogs aged 1–5 years (7.34%) (p < 0.05), suggesting a high prevalence of Thai CBuVs in dogs under 5 years of age. We performed complete genome sequencing (n = 15) and partial VP1/VP2 sequencing (n = 5) of Thai CBuVs. Genetic and phylogenetic analyses showed that whole genomes of Thai CBuVs were closely related to Chinese and Italian CBuVs, suggesting the possible origin of Thai CBuVs. The analysis of VP1 and VP2 genes in Thai CBuVs showed that 18 of them were placed in subgroup A, while only 2 belonged to subgroup B. This study is the first to report the detection and genetic characterization of CBuVs in domestic dogs in Thailand. Additionally, surveillance and genetic characterization of CBuVs in domestic animals should be further investigated on a larger scale to elucidate the dynamic, evolution, and distribution of CBuVs.


Phylogenetic and genetic analysis of CBuVs
In this study, we successfully sequenced the CBuVs, and the genome size of Thai CBuVs was 4214 bp.The genome structure of the virus contains non-structural protein (NS1) and viral capsid proteins (VP1 and VP2).Comparing the genome structure to other parvoviruses (canine parvovirus type-2 (CPV-2), canine bocavirus type 1 and 2(CBoV type 1 and 2), the genome structure of Thai CBuVs were similar to reference CBuV but diverse from CPV-2 and CBoV (Supplement Fig. 2).
For NS1 gene, phylogenetic analysis of the NS1 gene showed that Thai CBuVs (n = 15) were grouped into BuVs of the Protoparvovirus genus with CBuVs from Canada, China, Hungary, and Italy but separated from other canine enteric parvoviruses of Protoparvovirus genus (CPV-2) and Bocaparovirus genus (CBoV) (Fig. 2).It is noted that based on NS1 phylogenetic analysis, Parvovirinae contains 10 genera including Amdoparvovirus, Aveoparvovirus, Artiparvovirus, Bocaparvovirus, Copiparvovirus, Dependoparvovirus, Erythroparvovirus, Loliparvovirus, Protoparvovirus and Tetraparvovirus.For NS1 nucleotide comparison, Thai CBuVs were highly  3).For VP1 and VP2 genes, based on phylogenetic analysis, BuVs can be divided into subgroups based on host species, e.g., human, canine, swine, and bat subgroups.The CBuVs can be further divided into 2 subgroups: A and B. The Thai CBuVs (n = 18) were grouped into subgroup A, which were closely related to CBuVs from China and Italy.While 2 Thai CBuVs (CU_FS 28678 and CU_FS 28683) were grouped into subgroup B, which were like Italy CBuVs (35/ITA and 9AS/ITA) (Supplement Fig. 3 and 4).VP1 and VP2 genes of Thai CBuVs possessed the highest nucleotide identities to those of reference CBuVs (98.70-99.80%).Thai CBuVs (subgroup A) possessed the highest nucleotide identities to China CBuVs (Henan38; 99.80%) but low percentages of nucleotide identities to CPV-2 (51.40 Genetic analysis of the NS1 gene showed that the NS1 gene of CBuVs contains 1917 nucleotides (639 amino acids).The conserved replication initiator motifs (GLHFHVLLQ and IVRYFLTKQP) were observed to be identical in all reference CBuVs but were not present in human bufavirus (GLHIHVLVC and IANYFLIKKP).The conserved amino acids at the walker loop ATP or GTP binding motifs of Thai CBuVs (GPASTGKS) were observed in both CBuVs and other bufaviruses (Table 2).
VP1 gene of CBuVs contains 2130 nucleotides (710 amino acids).Thai CBuVs contained three potential sites for PLA2 activity, which is required for viral entry.At one site, the calcium binding loop, Thai CBuVs contained amino acid residues (YLGPG) similar to other bufaviruses from dogs, bats, and pigs but were not present in human bufaviruses (YLGPF).The other, two catalytic sites contain amino acid residues (HDLEY and D) similar to all reference BuVs (Table 2).Interestingly, amino acid residues of VP1 related to host preference (human specific) were observed in Thai CBuVs, including PTNRP3-6AIRKA, G22F, T24Q, N71D, K86R, and K89R.This observation could suggest preference characteristics of BuVs to human hosts (Table 3).
VP2 gene of CBuVs contains 1704 nucleotides (568 amino acids).A glycine-rich motif (G-rich) was observed at the N-terminus of VP1, which was similar to other parvoviruses (Table 2).This motif is speculated to be associated with the cellular entry of the virus.Thai CBuVs subgroup B (CU_FS28678 and CU_FS28683) contained 16 unique amino acid residues, which were similar to CBuVs strain 9AS and 35 from Italy, suggesting unique

Glycine-rich sequence GGG GGG GSGVG
This study  4).Moreover, amino acid residues related to host preference were observed.Amino acid insertion at 12-13 and 370 were observed in human BuVs and bat BuVs (Table 3).
The selective pressure of CBuVs was analyzed by the alignment of all genes of CBuVs using the statistical parameters (FEL, FUBAR, and MEME).The potential positive selection sites of NS1 (n = 9), VP1 (n = 8), and VP2 (n = 7) were detected in this study (Supplement Table 6).The overall mean difference of dN/dS was 0.306 for the NS1 gene, 0.125 for the VP1 gene, and 0.118 for the VP2 gene, suggesting all gene of CBuVs was under negative selection.For the NS1 gene, only one amino acid at position 512 of NS1 was confirmed to be a positive selection (p < 0.1 by MEME and posterior probability of 0.9 by FUBAR).For the VP1 gene, the amino acid position at 256 was confirmed to be a positive selection site (p < 0.1 by MEME and FEL, posterior probability of 0.9 by FUBAR).For the VP2 gene, amino acid positions at 22 and 113 were found as positive selection (p < 0.1 by FEL and posterior probability of 0.9 by FUBAR).A potential positive selection at 113 of the VP2 gene located at the predicted B-cell epitopes of Thai CBuV (strain CU_FS20141) at the location 99-118.

Discussion
Bufavirus (BuV) is a novel member of the family Parvoviridae.It was first reported in humans with gastroenteric symptoms in 2012 8 .Bufavirus infection in dogs was first described in Italy in 2016 and subsequently reported in several countries 20,23,24,27 .However, CBuVs have not been reported in Thailand.This study is the first to detect CBuVs in domestic dogs in Thailand.Our survey showed CBuV positivity at 9.42% (50/531), which was comparable to other previous studies (2.5-8.8%positivity) 20,[22][23][24] .However, there was no significant difference in CBuV positivity by season.Our result showed that CBuVs was highly detected in gastroenteritis dogs, although there was no statistical significance between symptomatic and asymptomatic dogs.The previous studies showed that the CBuVs have been detected in both healthy dogs and dogs with respiratory and gastroenteric symptoms 20,24,27 .Some studies supported that BuVs may associated with gastroenteritis and cause systemic infection in humans and dogs 10,11,23,24 .However, the pathogenesis of CBuV infection is still not clear.Thai CBuVs showed a significant presence in both younger than 1 year (< 1 year) and dogs aged 1-5 years with statistical significance (p < 0.05) (Supplement Table 1).This observation agreed with the previous study, which suggested that CBuV tends to be more prevalent for dogs under 5 years of age than the older age group 20 .In this study, we observed co-infection of CBuVs with other enteric viruses.For example, CPV-2 and CECoV were co-infection with CBuV, which was similar to other studies 23,27 .However, the severity level of clinical signs of co-infection between CBuVs and other pathogens should be further investigated.
Currently, only 17 nearly complete genome sequences of CBuVs are available in the GenBank Database.Our study provided an additional 15 complete genomes of CBuVs in the database.In this study, the complete genome size of CBuVs encoding 3 ORFs, including NS1, VP1, and VP2 (Supplement Fig. 2).Based on the phylogenetic analysis of the complete genome, Thai CBuVs belong to the bufavirus of the Protoparvovirus genus (canine group) and were closely related to CBuVs from China (Henan38), Hungary (HUN/2012/126) and Italy This study Human and non-human primate (ITA/2015/297).It is noted that the classification of parvovirus genus was classified by NS1 gene sequencing 28 .
The phylogenetic analysis of the NS1 gene showed that Thai CBuVs were grouped into the Protoparvovirus genus (CBuV group) but were separated from canine parvovirus type -2 (CPV-2).Thai CBuVs have high nucleotide identities to reference CBuVs at 98.70-99.90%but were distinct from CPV-2 viruses at only 57.10% nucleotide identities.The phylogenetic analysis of VP1 and VP2 genes showed that CBuVs clustered into 2 major subgroups, subgroups A and B, which agreed with the previous study 23 .Thai CBuVs (n = 18) were grouped into subgroup A with CBuVs from Italy and China.The other Thai CBuVs (strain CU_FS 28678 and CU_ FS 28683; n = 2) were grouped into subgroup B with CBuV from Italy.It is noted that Thai CBuVs (subgroups A and B) might share the same common ancestor with CBuVs from China and Italy.Thai CBuVs contained conserved regions of the parvovirus-conserved replication initiator motifs.For NS1, Thai CBuVs contained a helicase motif walker (GPASTGKS), which was also observed in all reference bufaviruses 8,20 .For the VP1 gene, Thai CBuVs posed a unique calcium-binding loop site (YLGPG) and two catalytic sites (HDLEY and D), which were also observed in most reference bufaviruses, suggesting unique characteristics 11,18 .Previous studies indicated that the calcium-binding loop site and phospholipase A2 (PLA2) region may be associated with viral entry to the host cell of the viruses 29 .The N-terminus of the VP1/VP2 had a glycine-rich motif, which may be associated with the cellular entry of parvovirus 30,31 .Moreover, human-specific amino acid residues of VP1 were observed at 3-6AIRKA, 22F, 24Q, 71D, 86R, and 89R.Unique amino acid insertion at positions 12-13 and 370 of VP2 were observed in human BuVs and Bat BuVs, suggesting potential antigenic markers for human and bat BuVs.Moreover, in the comparison of the genetic analysis between CBuVs subgroup A and B, there were 16 unique amino acids of CBuVs subgroup B, suggesting the genetic characteristics of the CBuV subgroup.
Our study showed that the VP1 and VP2 genes of Thai CBuVs have higher variations than the NS1 gene, which were similar to the previous study 32 .The VP2 gene of parvoviruses may have functions relating to viral entering, receptor binding, and immunogenicity and contains many major epitopes [32][33][34] .Moreover, B-cell epitopes are recognized as an immunogenicity classification.Thus, the mutations of this epitope may affect immunogenicity or generate a novel serotype of the virus 35 .The top list of antigenic epitope predictions on the VP2 gene of Thai CBuVs (CU_FS 20141 and CU_FS 28678) were observed in this study at 239-258 and 90-109, respectively (Supplement Table 5).In the selective pressure analysis based on the VP2 gene, two positive selections (22,  113) were observed in this study.A residue at 113 of VP2 was located at predicted B-cell epitopes of Thai CBuV (CU_FS20141) (location 99-118).Notably, the possibility of an antigenic shift to escape host immune responses of Thai CBuVs should be considered.However, due to the limited number of isolates in this study, further investigation on an extensive sample scale is necessary.
The positive selection pressure and genetic recombination are the factors that affect to high evolution rate of parvovirus evolution 36,37 .Previous study showed that natural recombination events of protoparvovirus between bufaviruses strains WUHARV and MgBuV1 can occur, suggesting cross-species transmission or sharing a common ancestor between bat, swine, and non-human primate bufavirus 15 .In this study, the recombinant Thai CBuV (CU_FS28678) was observed between Thai CBuV (CU_FS28696) and Italy CBuVs (9AS), suggesting a possible common ancestor of Thai and Italy CBuVs and inducing genetic diversity of CBuVs.
In conclusion, this is the first to report and genetically characterize the complete genome of CBuVs in domestic dogs in Thailand.Our result showed that Thai CBuVs were detected in both healthy and dogs with gastroenteric signs.The phylogenetic analysis showed that Thai CBuVs might share a common ancestor with CBuVs from Italy and China.However, the genetic database of CBuVs is still limited.Thus, surveillance and genetic characterization of CBuVs in domestic animals should be further investigated on a larger scale to elucidate the dynamic, evolution, and distribution of CBuVs.

Canine samples
In this study, we conducted a cross-sectional sample collection from Chulalongkorn University's Veterinary Teaching Hospital and private small animal hospitals in Thailand from September 2016 to October 2022.A total of 531 rectal swab samples were collected from dogs with asymptomatic (n = 216) and gastroenteritis (n = 315) symptoms, including vomiting, diarrhea, and dehydration.The samples were collected from dogs of young age (< 1 year) (n = 254), adults (1-5 years) (n = 177), and seniors (> 5 years) (n = 100).The animal demographic data, including age, sex, breed, and vaccination history, were recorded.This study was performed in accordance with the Chulalongkorn University, Animal Care and Uses Protocol (CU-VET IACUC#2031050, CU-VET IACUC#2031035) guidelines and regulations.

Canine bufavirus identification
All rectal swab samples were subjected to DNA extraction using DNA/RNA GENTi Automated Nucleic Acid Extraction System (GeneAll® Seoul, Korea) following the manufacturer's recommendations.For canine bufavirus identification, DNA samples were screened using PCR with specific primers for the VP1/VP2 gene.The primers used in this study were previously described, including CPPV 165F (5′-CTG GTT TAA TCC AGC AGA CT3′), CPPV 371R (5′-TGA AGA CCA AGG TAG TAG GT3′) corresponded to the position 2872-2891 and 3060-3079 of CBuV, respectively 22 .The PCR was performed in a final volume of 50 μl comprising 4 μl of template DNA, 25 μl of 2 × reaction buffer of the HotStarTaq® Master Mix (Qiagen, Hilden, Germany), 0.2 μM of each forward and reverse primers and distilled water to a final volume of 50 μl.The PCR condition included an initial denaturation step at 95 °C for 15 min, following 40 cycles of denaturation at 94 °C for 30 s, annealing at 52 °C for 30 s, and extension at 72 °C for 1 min, as well as a final extension step at 72 °C for 10 min.PCR products were run on a 1.5% agarose gel, which was mixed with RedSafe™ (iNtRON Biotechnology, Inc., Korea) at 100 V for 45 min.The expected size of the CBuV-positive sample was 208 bp.In this study, other canine viral enteric pathogens, including Canine Parvovirus, Canine Coronavirus, and Canine Rotavirus, were also tested in all samples [38][39][40] .

Canine bufavirus characterization
Representatives of positive CBuV were selected for whole genome sequencing (n = 15) and VP1/VP2 sequencing (n = 5).The CBuVs were selected based on epidemiological and demographic data such as age, collection date, breed, and vaccination history.Whole genome sequencing was conducted by amplification of each gene of the viruses by using modified oligonucleotide primer sets as previously described and new primer sets designed by using the Primer 3 plus program (https:// www.bioin forma tics.nl/ cgi-bin/ prime r3plus/ prime r3plus.cgi) (Supplement Table 8) 22,23,25,41 .Nucleotide amplification was conducted in 50 μl PCR reaction comprising 4 μl of template DNA, 25 μl of 2 × reaction buffer of the HotStarTaq® Master Mix (Qiagen, Hilden, Germany), 0.2 μM of each forward and reverse primer and distilled water to a final volume of 50 μl.The PCR condition was set as initial denaturation step at 95 °C for 15 min; 40 cycles of denaturation at 94 °C for 30 s, annealing at 50-55 °C for 30 s and extension at 72 °C for 1-1.30min, and final extension step at 72 °C for 10 min.PCR products were then purified and sequenced at 1st Base Laboratories Sdn Bhd, Malaysia.The nucleotide sequences were edited, validated, and assembled using SeqMan software v.5.03 (DNASTAR Inc.; Wisconsin, USA).In this study, whole genome and VP2 sequences of Thai CBuVs were deposited into the GenBank database under the accession numbers OQ730240-OQ730259 (Table 1).

Phylogenetic and genetic analyses of canine bufavirus
Phylogenetic and genetic analyses of CBuV were conducted by comparing nucleotide sequences of Thai CBuVs with those of reference viruses of the Parvoviridae family available from the GenBank database, including canine parvovirus, canine bocavirus, human bufavirus, bat bufavirus, rat bufavirus, and swine bufavirus.The reference nucleotide sequences of CBuV were also included.It should be noted that only 17 complete genome sequences of domestic dog and canid wildlife BuVs have been reported and were available from the GenBank database.Phylogenetic trees of WGS and VP1 gene of CBuVs were constructed using MEGA v.7.0 (Tempe, AZ, USA) with the neighbor-joining method applied with the Kimura 2-parameter and 1000 bootstrap replicates.For NS1 and VP2, phylogenetic tree and best models were generated by using IQ-TREE version 2.1.3(http:// www.iqtree.org/) with the TVMe + IG4 model of nucleotide substitution, default heuristic search options, and ultrafast

Figure 1 .
Figure 1.Phylogenetic analysis of the whole genome sequences of Thai CBuV.The phylogenetic tree was constructed using MEGA v.7.0 with a neighbor-joining algorithm with a Kimura-2 parameter model with 1000 replications of bootstrap analysis.The pink circle represents Thai CBuVs.

Figure 3 .
Figure 3. Phylogenetic analysis and recombination analysis of Thai CBuV (CU_FS 28678).(a) Phylogenetic analysis of CBuVs was constructed using MEGA v.7.0 with a neighbor-joining algorithm with a Kimura-2 parameter model with 1000 replications of bootstrap analysis.The pink circle represents Thai CBuV strain CU_FS 28678.The blue and green triangles showed putative major and minor parents.(b) Similarity and Bootscan analysis of Thai CBuV showed the recombinant CU_FS 28678.(c) Genome organization of potential recombinant of Thai CBuV strain CU_ FS 28678.

Table 1 .
Detailed description of Thai CBuVs characterized in this study.

Table 2 .
Genetic analysis of Thai CBuV and reference viruses.

Table 3 .
Genetic analysis of the primate and human-specific residues of Thai CBuV and reference BuVs.