New Parvovirus in Child with Unexplained Diarrhea, Tunisia

A divergent parvovirus genome was the only eukaryotic viral sequence detected in feces of a Tunisian child with unexplained diarrhea. Tusavirus 1 shared 44% and 39% identity with the nonstructural protein 1 and viral protein 1, respectively, of the closest genome, Kilham rat parvovirus, indicating presence of a new human viral species in the Protoparvovirus genus.


The Study
Using metagenomic deep sequencing, we analyzed fecal samples from 180 infants and children ages 7 days-96 months (mean 18.7 months) in Tunisia who had unexplained diarrhea that tested negative for rotavirus, norovirus, astrovirus, sapovirus, adenovirus types 40 and 41, and Aichi virus by reverse transcription PCR (7). The University of California at San Francisco committee on human research approved the study.
The fecal supernatants were filtered through a 0.45μm filter (Millipore, Darmstadt, Germany) to remove bacterium-sized particles, and the filtrates were digested with a mixture of DNases (Turbo DNase from Ambion, Carlsbad, CA, USA; Baseline-ZERO from Epicenter, Madison, WI, USA; and Benzonase from Novagen, San Diego, CA, USA) and RNase (Fermentas, Pittsburgh, PA, USA) to digest unprotected nucleic acids. Enriched viral nucleic acids (RNA and DNA) were then extracted and amplified by using ScriptSeq v2 RNA-Seq Library Preparation Kit (Epicenter) and analyzed in pools of 10 specimens in 2 Illumina MiSeq run of 250-bp end reads, yielding 20,693,619 unique sequences. We compared the Illumina sequences with the GenBank nonredundant protein databases using BLASTx (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
The individual fecal sample within the pool containing this sequence was then identified by using PCR and underwent further deep sequencing as above, generating 11 more parvovirus sequences. No other eukaryotic viral sequences were identified from 260,000 unique sequence reads from this patient. The near complete parvovirus genome was then acquired by filling genome gaps by PCR and amplifying 5′ and 3′ extremities using RACE (Rapid Amplification of cDNA Ends, Life Technologies). Amplicons were directly sequenced by primer walking. We named this virus Tusavirus 1 for Tunisian stool-associated parvovirus.

DISPATCHES
New Parvovirus, Tunisia panel A). The phospholipase A 2 (PLA2) motif was identified in VP1 N-termini with expected calcium-binding site and catalytic residues. The methionine codon of VP2 was located upstream of glycine-rich sequence (GGGARAG-GVG). An unusual serine-rich sequence (SSSDSGPSSS) was also seen near VP1 N-termini.
Protein sequence alignments were made by using ClustalX version 2.0.3 (http://www.clustal.org) with the default settings; a phylogenetic tree with 100 bootstrap resamples of the alignment datasets was generated by using the neighbor-joining method based on the Jones-Taylor-Thornton matrix-based model in MEGA5 (http:// www.megasoftware.net). Bootstrap values for each node are given if >70%. Resulting trees were examined for consistency with published phylogenetic trees.
Phylogenetic analysis showed that Tusavirus 1 was distinct from known members of the Protoparvovirus genus ( Figure, panel B). NS1 shared the highest identity to the Kilham rat parvovirus (44%). VP1 and VP2 shared identities of 39% and 37%, respectively, to those of Kilham rat parvovirus. According to the International Committee on Taxonomy of Viruses, the members of the same parvovirus genus should share >30% and members of the same species >85% aa identity in NS1. The Protoparvovirus genus currently comprises species infecting carnivores, rodents, pigs, and humans (1). Tusavirus 1 is proposed as prototype for primate protoparvovirus 2 species that would join bufaviruses as human viruses in this genus.

Conclusions
We detected fecal shedding of a previously uncharacterized parvovirus in a child with unexplained diarrhea. The 18-month-old girl showed twice daily liquid and greenish feces over 3 days but no fever (37.4°C), vomiting, or dehydration. To identify other viral infections in this patient, the Tusavirus-positive sample was individually analyzed by the same metagenomics method. No other mammalian virus was detected, suggesting a possible role for Tusavirus 1 in this patient's gastrointestinal illness, although the lack of testing for pathogenic bacteria and parasites does not enable us to exclude these alternative explanations. Wider geographic sampling of human samples, including casecontrol PCR studies of unexplained diarrhea and serologic tests, are needed to define the prevalence and disease association of this new parvovirus species in different age groups and populations.