After massive parallel sequencing of the nucleic acids obtained from fur seal scats, a wide variety of invertebrate and vertebrate viral hosts assignations with low nucleotidic and amino-acidic identities were obtained, most of them corresponding to animal species not described before in Antarctica. These results make us reconsider the use of closed RefSeq databases for viral discovery, especially because the studied area was a remote geographical area where a high number of new viral species is expected to occur22.
After repeating the analysis of the contigs obtained using BLASTn, a high number of miss-assignments was observed, corresponding almost entirely to contigs newly assigned as unclassified Eukaryotic Circular Rep-Encoding Single-Stranded DNA (CRESS-DNA) viral sequences. CRESS viruses have been detected ubiquitously in many different animals without any recognised role in the development of any disease23,24,25,26.
These results are in accordance with the recent reporting of CRESS sequences also being ubiquitous in a wide variety of environments and at high proportions, including Antarctica, where they have been described to represent more than 50% of sequences obtained from glacier waters27.
Viral-host distribution
Virome studies in other Arctocephalus species from subantarctic and South American regions revealed a 5% of viral sequences with predominance of bacteriophages followed by viruses from the Parvoviridae family28. The methodology here applied provided an increase of 12–25% viral reads when probe-based Target Enrichment Sequencing (TES) was applied, that in comparison with Untargeted Viral Metagenomics (UVM) approaches conducted in these type of samples28 could be considered an optimal result.
Most of the viral species detected in feces corresponded to unknown viruses, 83.59% from the total of sequences, followed by viruses that infect invertebrates, 8.75%, bacteriophages, 4.46%, and vertebrate viruses, 3.11% (Fig. 1).
Host distribution of viral assignations sequenced from fecal (A) and serum (B) samples collected from male A. gazella.
As expected, when applying both targeted and untargeted sequencing methodologies, TES approach resulted in a recovery of many vertebrate viral assignations (Table 1) whereas untargeted sequencing enabled a better detection of viruses known to infect invertebrates (Table 2). To describe the complete A. gazella fecal virome, sequences obtained by both sequencing methodologies were considered all together, representing a total of 2.62 million reads.
A. gazella viruses
Fur seal picorna-like virus
Fur seal picorna-like virus was firstly described in a fecal sample obtained from A. gazella in King George Island in the South Shetland Islands, Antarctica by Krumbholz and co-workers16.
In this study, we report a total of 19 contigs resulting after assembling 2671 reads obtained from 4/4 fecal pools analysed being the most prevalent virus described in this study. One of the contigs covered 96.91% of the fur seal picorna-like virus genome and presented a nucleotide homology of 99.38% with the reference strain described in 2017. The other contigs coverage ranged from 19.75 to 21.22% with a 45.92 to 90.5% nucleotide identity with reference strain NC_035110. Four contigs matching the ORF2 polyprotein are represented in Fig. 2 where differences among them and with the reference strain are showed.
Nucleotide alignment of ORF2 sequences from the A. gazella picorna-like contigs compared to the ORF2 from RefSeq NC_0351110. In consensus strain, position 1 represents position 6523 from RefSeqs genome.
Picornaviruses are known to cause a wide variety of diseases in vertebrate hosts, especially mammals29, but the role of Fur seal picorna-like virus in pathogenesis development is still unknown30. Many picornaviruses are transmitted horizontally via fecal–oral or airborne routes29. The fact that these sequences were detected in all the fecal pools obtained from animals with no evidence of disease may that suggest the virus may have a stable endemic relationship within that seal population.
Torque teno pinniped virus
Lambdatorquevirus is a genus within the Anelloviridae family. The genus comprises 8 species named Torque teno pinniped virus 2 to 9 isolated from different pinniped species: A. gazella (Torque teno pinniped virus 6 and 7)17, Phoca vitulina (Torque teno pinniped virus 2, 3, 4)31, Zalophus californianus (Torque teno pinniped virus 5)32 and Leptonychotes weddellii (Torque teno pinniped 8 and 9)33.
One contig with a nuleotide similarity of 95.12% against Torque teno pinniped virus 7 was obtained from one of the fecal pools. This virus had been described in these animals inhabiting Livingston Island in 2016, using rolling circle amplification and subsequent Sanger sequencing from buccal swabs17. However, sequences obtained in this study belong to partial ORF2 which is not the optimal genome region for typing purposes or phylogenetic analysis.
These members of the Anelloviridae represent the more abundant viruses found in human, animals and environmental samples although their etiological role in any disease has not been clearly identified being considered a persistent virus ubiquitous to several different tissues34,35
No Torque teno virus sequences were detected in serum samples which agree with what was observed for Zalophus californianus anellovirus prevalently detected in different tissues, like lung and liver, but not in blood samples. Interestingly, other known anelloviruses are typically found in blood or plasma samples32.
Mamastrovirus
Two of the fecal pools analyzed presented Mamastrovirus sequences. The presence of these viruses in humans and other mammals is widely known, as well as their involvement in gastroenteritis development36. The four contigs obtained (comprising 1008 sequences) showed homologies against reference genomes, ranging from 45.70% to 59.37% when compared at nucleotide level and 36.69% to 46.69% when compared at aminoacidic level. Phylogenetic analysis of partial OFR2 regions of these contigs indicate its closer similarity with sequences from California Sea Lion astroviruses, a virus that was determined as to be the most prevalent in fecal samples from these animals (Z. californianus)37. This finding suggests that these sequences may belong to a yet unknown virus like Z. californianus astrovirus and may indicate that such virus is prevalent in the sampled area (detected in 2/4 fecal pools studied) and the second more abundant virus (1008 reads) in the studied fecal samples (Fig. 3).
Phylogenetic consensus tree based on partial ORF2 sequences from the Mamastrovirus contigs sequenced from A. gazella scats (in bold). Bootstrap resampling with 1000 replicates.
Adeno associated virus 2
Two of the studied fecal pools presented 138 sequences, forming 3 contigs with nucleotide identities ranging from 46.91 to 48.04% (Table 1), that matched adeno associated viruses previously described in Z. californianus, humans and other mammals with and unknow etiologic role (Fig. 4). The detected sequences probably correspond to fur seal adeno associated viruses never described before. The detection of these viruses is quite common in other mammals suggesting they could cause persistent infections in their hosts, but no etiological role has been attributed to them38.
Phylogenetic consensus tree of the Adeno-associated virus contigs sequenced from A. gazella scats (in bold). Bootstrap resampling with 1000 replicates.
Norovirus
A norovirus contig was obtained in one of the four pools analyzed. Noroviruses are the most relevant non-bacterial gastroenteritis etiological agents in humans39, with its presence widely described in other mammals40. The contig detected in the fecal samples, represented the 4.43% of the viral genome, was in the VP1 region and comprised 56 reads with an identity > 99% to California sea lion norovirus described by Teng and collaborators in 201841 (Fig. 5). Results obtained suggest these sequences belong to a putative new norovirus specie.
Phylogenetic consensus tree of the Norovirus contig sequenced from A. gazella scats (in bold). Bootstrap resampling with 1000 replicates.
Viruses in serum samples
All the viral sequences obtained from serum samples (970 reads) matched to CRESS-DNA viral sequences from unknown hosts.
The fact that no other viruses were identified in serum samples suggests the animals tested were not under active viremia at the time of sample collection or it was not detectable by the applied methodology.
Diet related viruses
Several virus sequences similar to viruses known to have invertebrate animals as hosts were detected in fecal pools, mainly by UVM although some also by TES. These viruses are probably present in fur seal feces because of dietary habits although, since scats were collected from the ground nearby the animals, environmental cross-contamination should not be ruled out.
Sequences with high coverage or similarities to any described virus are showed in Table 2.
The high prevalence of virus sequences from crustaceans in the feces analyzed is hardly surprising because A. gazella inhabiting the Antarctic peninsula and the Atlantic sector of the Southern Ocean feed mostly on Antarctic krill Euphasia superba during the summer months42,43,44,45,46,47,48. Sequences from cephalopod viruses were also detected, although were much scarcer than those from crustaceans. This also agrees with current knowledge about the diet of A. gazella in the Atlantic sector of the Southern Ocean, where octopuses and squids are regularly consumed, although in low numbers44,45,46. It is worth noting than not cephalopod beak was recovered from the scats analyzed here48. Among all invertebrate viruses identified, some sequences present low identities with genomes from available databases, probably because Antarctica wildlife has been scarcely explored, forcing bioinformatic analysis to match them with the most similar viruses from these databases.
No fish viruses were found in this study. Hard skeletal remains of fishes are often recovered from the scats of A. gazella from the Atlantic sector of the Southern Ocean42,43,44,45,46,47 and occurred indeed in the samples analysed here48, but stable isotope analysis of blood and whiskers revealed a negligible contribution of fish to the assimilate diet of juvenile and subadult male A. gazella49, which likely explain the absence of fish viruses in the samples analized here. Additionaly, no data on the virome present in the fish species regularly consumed by A. gazella has been published to our knowledge, with information limited to the bacteriome32, so even in case fish viruses were sequenced, it might not be correctly assigned to a fish host. Nevertheless, the methodology applied in this study had been successfully applied to the identification of the virome of Atlantic fishes50. Furthermore, Li and coworkers.37 and Wille and coworkers.22 also observed viral sequences probably corresponding to fish when analyzing the fecal virome of the California sea lions and Antarctic penguins.
On the other hand, sequences highly similar to Coelho and Khabarov viral polymerases (greater than 98% of aminoacid identity), previously described in chinstrap penguins (Pygoscelis antarcticus) by Wille and coworkers22, were found in this study. The consumption of penguins by A. gazella during the summer months has been reported widely51,52,53,54,55, penguins feathers were reported from the scats analyzed in this study48 and stable isotope analysis of blood and whiskers revealed penguins as the second most relevant prey from juvenile and subadult male A. gazella in the population studied here49. This evidence is consistent with the presence of virus from chinstrap penguins in the samples analysed here. All in all, the study of fecal virome constitutes a very promising tool to explore the consumers’ diet.
Source: Ecology - nature.com
