Human Picobirnaviruses Identified by Molecular Screening of Diarrhea Samples

The samples included in the study were diarrhea samples submitted to the Diagnostic Unit, Department of Virology, Erasmus Medical Center, Rotterdam, Netherlands, for diagnosis of gastrointestinal infections in 2007. Out of 1,025 samples screened by standard diagnostic assays for rota-, adeno-, astro-, noro-, entero-, and parechovirus, 165 stool samples were consistently found to be negative for these viruses. Of these 165 samples, 84 were available for additional studies. The percentage of patients with diarrhea of unknown etiology per month was determined, as was the percentage of patients per birth date. Data were compared using Student's t test, and differences were considered significant at a P of <0.05. In addition, 17 diarrhea samples (submitted to the Diagnostic Unit, Department of Virology, Erasmus Medical Center, Rotterdam, Netherlands, for diagnosis of gastrointestinal infections in 2007) in which a virus was previously detected were also available.

The 84 diarrhea samples were collected for the present study, made anonymous, annotated with VS numbers, and stored at −80°C until analyzed by random amplification of partially purified viral RNA, performed as described previously with modifications (1). A total of 13 diarrhea samples were suspended in phosphate-buffered saline (PBS) until their weight was 10% of the volume and centrifuged at 14,000 rpm for 5 min. Supernatants were filtered through 0.45-μm spin filters (Ultrafree-MC; Millipore), after which Omnicleave endonuclease (Epicentre Biotechnologies) and magnesium chloride (Applied Biosystems) were added to final concentrations of 2 U/μl and 5 mM, respectively, to degrade DNA and RNA. Viral nucleic acids are generally not degraded by nucleases, as they are protected by stable protein capsids and sometimes also by a lipid envelope. Samples were incubated for 1 h at 37°C. RNA was extracted from half of the samples using the Nucleospin RNA XS kit (Machery-Nagel), according to the instructions of the manufacturer. First-strand synthesis was performed by mixing RNA with 1 μM primer FR26RV-N (5′-GCCGGAGCTCTGCAGATATCNNNNNN-3′) and each deoxynucleoside triphosphate (dNTP) at 0.5 mM, which mixture was incubated at 95°C for 5 min and chilled on ice. First-strand buffer (5×; Invitrogen), dithiothreitol (DTT) (Invitrogen), RNasin (Promega), and Superscript III reverse transcriptase (Invitrogen) were added to final concentrations of 1×, 5 mM, 2 U/μl, and 10 U/μl, respectively. The reaction mixture was incubated for 5 min at 25°C and 45 min at 50°C. After a denaturation step at 94°C for 3 min and chilling on ice, 2.5 units of 3′→5′ Exo⁻ Klenow DNA polymerase (New England Biolabs) was added, and the reaction mixture was incubated at 37°C for 1 h, followed by an enzyme inactivation step at 75°C for 10 min for second-strand synthesis. The reaction mix was used as a template in a 50-μl PCR mixture containing 1× AmpliTaq Gold buffer, 2.5 mM MgCl₂, each dNTP at 0.2 mM, 0.8 μM primer FR20RV (5′-GCCGGAGCTCTGCAGATATC-3′), and 2.5 U AmpliTaq Gold (Applied Biosystems). After 10 min at 94°C, 40 cycles of amplification (94°C for 60 s, 65°C for 60 s, 72°C for 2 min) and 1 cycle of elongation (72°C for 10 min) were performed. The PCR products were separated on an agarose gel, and fragments between 200 and 400 bp, 400 and 800 bp, and 800 and 1,500 bp were excised and purified using the Invisorb Spin DNA extraction kit (Invitek GmbH) according to the manufacturer's instructions. Products were cloned into pCR4-TOPO and introduced into chemically competent Escherichia coli TOP-10 according to the manufacturer's instructions (Invitrogen).

Sequencing templates were produced directly from colonies in a 96-well format by performing a colony PCR. The 25-μl reaction mixture consisted of 1× Thermopol reaction buffer (New England Biolabs), each dNTP at 0.2 mM, 0.2 μM M13 sequencing primers, and 1 unit Taq DNA polymerase (New England Biolabs). After 3 min at 95°C, 30 cycles of amplification (95°C for 30 s, 45°C for 60 s, 72°C for 2 min 30 s) and 1 cycle of elongation (72°C for 10 min) were performed. PCR products were purified using an MSB HTS PCRapace/C kit (Invitek GmbH) according to the manufacturer's instructions. Sequencing was performed on 2-μl purified PCR products using a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) and a 3130XL genetic analyzer (Applied Biosystems).

The picobirnavirus sequence readings for segments 1 and 2 previously detected in stool filtrate VS10 and the rhinovirus sequences in sample VS25 obtained by random amplification were assembled in different contigs, and the nucleic acid between the contigs was obtained by reverse transcription (RT)-PCR (Fig. 1; Table 1). cDNA was synthesized from RNA (see above) using 0.25 μg of random primers (Promega) and each dNTP at 0.8 mM, incubated at 65°C (or 95°C for picobirnavirus) for 5 min, and chilled on ice. First-strand buffer (5×; Invitrogen), DTT (Invitrogen), RNasin (Promega), and Superscript III reverse transcriptase (Invitrogen) were added to final concentrations of 1×, 5 mM, 2 U/μl, and 10 U/μl, respectively. The reaction mixture was incubated for 5 min at 25°C, 45 min at 50°C, and 15 min at 72°C.

PCR was performed with oligonucleotides directed toward sequences determined by sequence-independent amplification from the newly identified viruses or directed toward 5′ and 3′ sequences from picobirnaviruses (Fig. 1; Table 1), using 1× AmpliTaq Gold buffer, 1.5 mM MgCl₂, each dNTP at 0.2 mM, 5 pmol primers, 2 U AmpliTaq Gold (Applied Biosystems), and 1 or 2 μl cDNA. After 5 min at 95°C, 40 cycles of amplification (95°C for 30 s, 45°C for 60 s, and 72°C for 1 min to 2 min 30 s, depending on the size of the amplicon) and 1 cycle of elongation (72°C for 10 min) were performed.

For rapid amplification of cDNA 3′ ends (RACE)-PCR of the rhinovirus genome, cDNA was synthesized using 10 pmol primer VS430 (Table 1) as described above. The first or second PCR mixtures contained 1× AmpliTaq Gold buffer, 2.5 mM MgCl₂, each dNTP at 0.2 mM, 20 pmol primers (VS459/431 or VS460/432), 2.5 U AmpliTaq Gold (Applied Biosystems), and 5 μl cDNA or first PCR products, respectively. After 2 min at 95°C, 35 cycles of amplification (95°C for 60 s, 55°C for 45 s, 72°C for 4 min) and 1 cycle of elongation (72°C for 10 min) were performed. All sequences were determined in both directions for at least two clones.

All obtained sequences were screened for homology with known sequences using the Basic Local Alignment Search Tool (BLAST), specifically BLASTN and TBLASTX (2).

All sequences were downloaded from GenBank; accession numbers are available on request. Multiple alignments were created using ClustalX (2.0.10). Sequences were dereplicated to 95% sequence identity with gaps using the online program FastGroup II (40). Phylogenetic analyses were carried out with Molecular Evolutionary Genetics Analysis (MEGA) software, version 4.1beta (33), in a manner similar to that used in previous picobirnavirus studies (7, 30, 32). The average pairwise Jukes-Cantor distance was calculated to determine the appropriateness of creating a neighbor-joining tree. Phylogenetic trees were calculated using neighbor-joining with a Jukes-Cantor model, and bootstrap analysis was performed with 1,000 replicates.

RT-PCR-grade viral RNA was extracted from aliquots of 101 samples suspended in PBS (10%, wt/vol) from patients with diarrhea, using a Magnapure LC total nucleic acid isolation kit (Roche) according to the manufacturer's instructions. cDNA synthesis was performed on 5 μl total nucleic acid using 0.25 μg of random primers (Promega) and each dNTP at 0.5 mM, which mixture was incubated at 95°C for 5 min and chilled on ice. First-strand buffer (5×; Invitrogen), DTT (Invitrogen), RNasin (Promega), and Superscript III reverse transcriptase (Invitrogen) were added to final concentrations of 1×, 5 mM, 2 U/μl, and 10 U/μl, respectively. The reaction mixture was incubated for 5 min at 25°C, 45 min at 50°C, and 15 min at 72°C. Portions of genomic segment 2 of genogroup I and genogroup II picobirnaviruses were amplified by PCR as described previously, with adaptations (3, 30). Oligonucleotides to amplify genogroup I picobirnaviruses were VS352 and VS354 and to amplify genogroup II picobirnaviruses were VS353 and VS355 (Table 1). Two microliters of cDNA was used in a PCR mixture containing 1× AmpliTaq Gold reaction buffer, 0.075% Triton X-100, 2 mM MgCl₂, 200 μM each dNTP, 2 U AmpliTaq Gold with GeneAmp (Roche), and 800 nM VS352 and VS354 or 200 nM VS353 and VS355 primers for genogroup I or II, respectively. The conditions used for PCR amplification were an initial denaturation step at 95°C for 5 min and then 40 cycles of 95°C for 30 s, 45°C for 2 min, and 72°C for 3 min, followed by a single incubation at 72°C for 10 min. All products were visualized using agarose gel electrophoresis. To verify the identities of viruses from positive assays, PCR products were cloned and sequenced (at least two clones in both directions). The percentage of patients with diarrhea that contained picobirnavirus sequences was determined per month. In addition, the birth dates of patients positive and negative for picobirnaviruses were compared using an unpaired Student t test. Differences were considered significant at a P of <0.05.

All obtained sequences were deposited in the GenBank database under accession numbers GU968923 to GU968948.

From the Diagnostic Unit, Department of Virology, Erasmus Medical Center, Rotterdam, Netherlands, we received 84 diarrhea samples of unknown etiology for further analysis of the potential presence of viral pathogens. The sample admission dates of diarrhea of unknown etiology were evenly distributed across the year (Fig. 2A). The ages of patients with diarrhea of unknown etiology were unevenly distributed, with most patients (58%) being ≤7 years of age (Fig. 2B). Initially, 13 clinical diarrhea samples were analyzed using sequence-independent amplification, with characterization of a total of 541 clones according to nucleotide and translated-nucleotide BLAST searches (2) (Fig. 3). In general, most of the sequences in each sample were unclassified or of bacterial origin. In total, ∼10% of analyzed clones showed similarity to viral sequences.

The viral sequences matched five different virus species, of which two were DNA viruses. Sequences similar to E. coli bacteriophage rv5 (family Myoviridae) were not analyzed in more detail, since it is a noneukaryotic virus. A Torque teno virus (TT-virus)-like sequence (family Circoviridae) that displayed ∼70% nucleotide identity to known TT-virus sequences was obtained. Because this group of viruses is known to ubiquitously infect humans, is genetically highly diverse, and was found in diarrhea samples previously (14, 27), the TT-virus-like sequence was not analyzed further. The Moloney leukemia virus (MoLV; family Retroviridae) sequence was nearly identical at the nucleotide level to known MoLV polymerase sequences and considered to represent a previously known virus. A potentially new human rhinovirus (HRV) genotype (family Picornaviridae) was detected, with clones displaying ∼80% nucleotide similarity to HRV14 and HRV35. The complete polyprotein-coding sequence of the rhinovirus was determined. By comparison to the recently determined polyprotein sequences of other rhinoviruses (28), we determined that our rhinovirus sequence belonged to species HRV-B and was phylogenetically most closely related to HRV72, with >95% amino acid identity (Fig. 4). In sample VS10, ∼25% of clones showed similarity to a group of unclassified, nonenveloped, small RNA viruses with genomes composed of two segments of double-stranded RNA, called picobirnaviruses. As only limited information is available on picobirnaviruses, a more in-depth study on this virus was performed.

A large part of the genome sequence of the identified picobirnavirus was determined by specific PCR targeting of the sequences detected in the sequence-independent amplification procedure. The sequences showed homology to genogroup I picobirnaviruses (3, 30) and were compared to the known complete segment 1 and 2 sequences of picobirnaviruses isolated from humans and rabbits. The sequences obtained for segment 1 of the picobirnavirus from sample VS10 showed only ∼25% amino acid identity to the segment 1 sequences of reference strain HY005102 and the rabbit picobirnavirus (see Fig. S1A in the supplemental material). Segment 2 contained one long open reading frame that showed ∼65% identity at the amino acid level to human prototype strains HY005102 and 1-CHN-97 (see Fig. S1B in the supplemental material). These data showed that the characterized Dutch picobirnavirus genome was highly divergent from the known picobirnavirus genomes.

To estimate the prevalence of picobirnaviruses, diagnostic PCR assays for genogroup I and II picobirnaviruses (3, 30) that target the polymerase gene were performed on the other 83 samples from patients with diarrhea of unknown etiology. In total, 17 samples were confirmed positive for genogroup I picobirnavirus sequences (20%). Mixed infections were detected in samples VS62 and VS142. VS62 contained two different genogroup I picobirnavirus variants. Sample VS142 contained two genogroup I variants and was also positive for the only detected genogroup II variant in our study that was highly similar (98%) to the prototype genogroup II strain 4-GA-91 (30). The medical records of the 17 patients infected with picobirnavirus were reviewed. The sample admission dates of the 17 picobirnavirus-positive samples were distributed relatively evenly over the year (Fig. 5A). Interestingly, although our sample set is skewed for young age, the average age of people infected with picobirnaviruses was significantly higher than the age of uninfected patients with diarrhea of unknown etiology (P < 0.005) (Fig. 5B). In addition, diagnostic PCR assays for genogroup I and II picobirnaviruses were performed on 17 diarrhea samples in which known pathogens were identified by standard diagnostics. Three of these samples were confirmed positive for genogroup I picobirnavirus sequences (17.5%) and also contained either adenovirus or rotavirus. Two patients were younger than 3 years of age, and one was 43 years old. The percentages of picobirnavirus-positive samples were similar between patients with diarrhea of unknown (20%) and known (17.5%) etiological agents (P > 0.8). Therefore, it remains unclear whether picobirnaviruses are the cause of diarrhea in our picobirnavirus-positive samples.

To determine the genetic relationships between the identified genogroup I picobirnaviruses in our samples and previously reported genogroup I human picobirnavirus isolates, a phylogenetic tree was constructed based on a 165-nucleotide fragment of the RNA-dependent RNA polymerase gene, as described previously (7, 30, 32). Prior to tree construction, a total of 76 groups were created from the almost 300 available picobirnavirus sequences using FastGroup II (40). The majority of groups were composed of one or two sequences; seven groups contained five or more sequences, and six groups of sequences contained more than 17 sequences. Since the average pair-wise Jukes-Cantor distance was 0.48, a neighbor-joining tree was created using the Jukes-Cantor model, with a bootstrap replication of 1,000 (Fig. 6). Twelve out of 22 of the determined genogroup I picobirnavirus sequences in this study showed <95% sequence identity to previously described picobirnavirus sequences and are shown as separate branches in the phylogenetic tree. The Dutch genogroup I picobirnavirus nucleotide sequences showed 54 to 99% similarity to each other. They belonged to different phylogenetic clades and did not group with other picobirnaviruses according to year of isolation or host species.