A novel parvovirus (family Parvoviridae) in a freshwater fish, zander (Sander lucioperca)

In this study, a novel parvovirus (zander/M5/2015/HUN, OK236393) was detected in faecal specimens from a fish – zander or pikeperch (Sander lucioperca) – and genetically characterized using viral metagenomics and PCR methods. The NS1 and VP1 proteins of zander/M5/2015/HUN share <30% aa sequence identity, respectively, with the corresponding proteins of known members of the family Parvoviridae. Out of 62 faecal specimens collected from 13 freshwater fish species, three (4.8%) samples were positive by PCR for the novel parvovirus – all from zander. This is the second parvovirus detected in fish – after the disease-causing tilapia parvovirus of the subfamily Hamaparvovirinae – and it potentially represents a novel genus in the subfamily Parvovirinae. Supplementary Information The online version contains supplementary material available at 10.1007/s00705-022-05419-5.

Parvoviruses are capable of infecting a wide range of animals from insects to mammals. However, the first parvovirus from fish was discovered only in 2019 [6]. Tilapia parvovirus was identified in intestinal samples of tilapia (Oreochromis niloticus) and faecal samples of crocodiles fed with tilapia using next-generation sequencing (NGS) in Hainan, China [6]. Subsequently, the same parvovirus was reported in a high-mortality outbreak affecting adult farmed Nile tilapia in Hubei, China [8], and red hybrid tilapia in Thailand [14]. Tilapia parvovirus belongs to the genus Chaphamaparvovirus, subfamily Hamaparvovirinae. The only other parvovirus that has been identified in fish is syngnathid ichthamaparvovirus 1 (genus Chaphamaparvovirus, subfamily Hamaparvovirinae), which was sequenced and identified through NGS assemblies from a tissue homogenate of a gulf pipefish (Syngnathus scovelli) [9]. These pioneering studies indicate that fish could be unexplored but important hosts for parvoviruses.
In this study, we report a member of a potentially new genus of parvovirus detected in faecal samples from a freshwater fish -zander or pikeperch (Sander lucioperca) -in Hungary.
A total of 62 faecal samples were collected directly from 13 different species of freshwater fish (Table 1) living in natural and artificial open-air fishponds in the vicinity of the town of Szarvas (East Hungary) in 2015 [7]. The fish showed no clinical signs of disease during the sample collection and were released immediately after sampling.
A specimen pool containing faecal samples from three zander (Sander lucioperca) (M5, M6, and M8) were selected for viral metagenomics analysis. Briefly, 200 μl of PBSdiluted specimen was passed through a 0.45-μm sterile filter (Millipore) and centrifuged at 6,000 × g for 5 min. Then, the filtrate was treated with a mixture of DNases and RNases (Turbo DNase, Invitrogen; Baseline Zero DNase, Epicentre Biotechnologies; Benzonase Nuclease, Novagen; RNase A, Fermentas) at 37 °C for 2 hours to digest unprotected nucleic acids [11]. Virus-particle-protected nucleic acids were extracted using the QIAamp spin-column technique (QIAamp Viral RNA Mini Kit, QIAGEN) using an RNase inhibitor (RiboLock RNase Inhibitor, Fermentas) at the elution step. Sequence-independent random RT-PCR amplification [15] with 20 PCR cycles was used, and a 250-bp pairedend viral cDNA library was constructed using a Nextera XT DNA Library Preparation Kit (Illumina). The library was sequenced on a MiSeq Illumina platform according to the manufacturer's instructions. The resulting metagenomic reads were trimmed, assembled de novo [5], and analyzed using an in-house pipeline [11]. Briefly, singlets and the assembled contigs greater than 250 bp in length were compared to the GenBank [2] protein database using BLASTx (version 2.2.7) [1] using an E-value cutoff of 0.01. Candidate viral hits were then compared to a non-virus non-redundant protein database to remove false-positive viral hits. Virus-family-level categorization of all viral metagenomic sequences was based on the best BLASTx scores (E-value ≤ 10 -10 ). All of the genomic data have been deposited in NCBI BioProject PRJNA784176. Raw reads are available in the Sequence Read Archive (SRA) database under accession number SAMN23475123.
A sequence-specific screening primer pair (ZanderParvoscreen-F, GGC TAA TCA TCA AAC AGG AAA GA; Zan-derParvo-screen-R, AGC TCC CAC CAC TTA ATA TCT T) was designed to identify a 492-nucleotide-long portion of the NS1 region of the viral genome of the study strain from the specimen pool. The PCR thermocycler program consisted of 1 min at 95ºC, 40 cycles for 30 s at 95ºC, 10 s at 48ºC, and 30 s at 72ºC, and a final 10-min extension at 72ºC, using a C1000 Touch Thermal Cycler (Bio-Rad). In addition, different sets of specific primers were designed based on the sequences of the metagenomics reads/contigs and the amplified PCR products for verification of the metagenomics contig by Sanger sequencing (using a BigDye Terminator v1.1 Cycle Sequencing Kit [Thermo Fisher] on an ABI3500 Genetic Analyzer [Applied Biosystems, Hitachi, Tokyo, Japan]) and to obtain the nearly complete viral genome of the study strain. Faecal samples from each fish were tested individually by the PCR method, using the ZanderParvoscreen-F/R screening primer-pair.
ClustalX (version 2.1) and GeneDoc (version 2.7) were used to align the corresponding amino acid sequences of the helicase domain of NS1 from this study and those of representative prototype parvoviruses of the subfamily Parvovirinae [10]. For the construction of an NS1-helicase phylogenetic tree, the BEAST v. 1.10.4 software package was used with a setup similar to that described by Pénzes et al. [10]. Briefly, the substitution model LG+I+G+F with a lognormal relaxed clock and Youle process was used throughout 10 million generations.
Sequence reads/contigs corresponding to members of the family Parvoviridae were selected for further analysis. The parvovirus sequences matched best to members of an unassigned parvovirus genus (n = 14,994 reads) and nine  [17]. In addition, the NS1 protein contains two conserved replication initiator (endonuclease) motifs, xxHuHxxxx (DH 109 HMH 111 VIIP) and YxxxK ( 183 YFSKK 187 ) (conserved aa are indicated in bold letters, and "u" as a hydrophobic residue) [13, https:// talk. ictvo nline. org/ ictv-repor ts/ ictv_ online_ report/ ssdna-virus es/w/ parvo virid ae]). The NS1 protein shares 28.4% and 27.6% aa sequence identity (query coverage: 58% and 58%, respectively) with the corresponding NS1 proteins of murine adeno-associated virus 1 (NC_055485) from mice [18] and a dependovirus (MT138242) found in anal swabs of birds [Xiao et al., unpublished], both belonging to the genus Dependoparvovirus, subfamily Parvovirinae, as the closest matches by BLASTp. The VP1 was predicted to be 582 aa (1,749 nt) in length, which is similar to those of other members of the subfamily Parvovirinae (537-781 aa). The N-terminal glycine-rich region is present (34 glycine residues -28.5% -in the first 119 aa of VP1) as is the GPGN calcium-binding loop ( 35 GPGN 38 ) [16], but the phospholipase A 2 (PLA 2 ) catalytic residues (DxxAxxHDxxY + D), which are widely present in the VP1 unique part (VP1up) of many parvoviruses [16] were not identifiable in the VP1 proteins of zander/M5/2015/HUN. Interestingly, no similar VP1 protein sequence was found in GenBank using BLASTp. NS1 overlaps by 361 nt with the VP1 region, and this was confirmed by PCR and sequencing of three different strains (M3, M5, and M7). The left and right terminal sequences of the genome were partially determined. The 3' genome end is more than 475 nt long in zander/M5/2015/HUN. Similar nucleotide sequences were not found in the GenBank database. The 5' genome end could be more than 503 nt long; however, this region potentially encodes a 125-aa-long protein (Fig. 1). Similar nucleotide and protein sequences were not found in GenBank.
Comparing the 4,322-nt-long study sequence to the metagenomic reads/contigs corresponding to members of the family Parvoviridae (N = 19,998) in the specimen pool, 95.38% of these metagenomic sequences could be aligned to the nt sequence of zander/M5/2015/HUN, indicating that the vast majority of these parvovirus sequences represent one parvovirus strain.
Phylogenetic analysis based on aa sequences of the tripartite helicase domain of NS1 showed that zander/M5/2015/ HUN formed a distinct lineage -potentially representing a new genus -in the subfamily Parvovirinae (Fig. 2).
Applying the screening primer pairs, three (4.8%) of the 62 specimens (M3, M5, and M7) were PCR-positive for the study strain, all three of which were from zander (Sander lucioperca) (3 out of 7 specimens, 42.8%) ( Table 1). The nucleotide sequences of the 995-nt-long NS1/VP1 joining region were 100% and 99.6% identical (2 synonymous nt mutations) between zander/M5/2015/HUN (OK236393) and zander/M7/2015/HUN (OK236395) and between zander/ M5/2015/HUN and zander/M3/2015/HUN (OK236394), respectively. The number of known members of the family Parvoviridae is rapidly expanding [10]. At present, there are 26 parvovirus genera, and their members have been discovered in a wide range of animal host species from insects to humans. In spite of this, the first fish-origin parvovirus (a diseasecausing tilapia parvovirus from an intensively aquafarmed tilapia fish [6] for human consumption) was not reported until 2019.
This study represents the second detection and characterization of a novel parvovirus from faecal specimens of a freshwater fish, zander. According to the species demarcation criteria of the International Committee on Taxonomy of Viruses (ICTV) Parvoviridae Study Group, two parvoviruses have to share >85% aa sequence identity in the NS1 protein to belong to the same species [4]. In addition, all parvoviruses in a genus should be monophyletic and encode NS1 proteins that are >30% identical to each other at the amino acid sequence level [4]. Following these rules, since the parvovirus from zander described here has less than 30% aa identity in the NS1 protein to any presently known parvovirus NS1 proteins, it potentially represents a new genus and a new species in the family Parvoviridae. While the tilapia parvovirus is a member of the subfamily Hamaparvovirinae, the novel parvovirus from zander belongs to another parvovirus subfamily, Parvovirinae.
Although this parvovirus was identified in faecal samples from zander, the host from which this novel virus originated remains unknown. While we could detect it in more than one faecal specimen from zander, we cannot exclude the possibility of a dietary origin of the virus. The pathogenicity and impact of this novel virus on aquafarming should be investigated in further studies, especially in the light of the fact that tilapia parvovirus causes disease in fish [6,8]. It should also be noted, that one of the parvovirus-positive faecal specimen (M7) from zander also contained a potentially novel fish-origin picornavirus (family Picornaviridae) [7] as a co-infection.
Systematic investigation of samples collected from aquatic animals is necessary to explore the genetic diversity of fish-origin parvoviruses and to identify potential diseasecausing pathogens.