Introduction

Noroviruses (NoVs) belonging to Caliciviridae family are recognized as a worldwide cause of epidemic acute gastroenteritis [1, 2]. NoVs are non-enveloped viruses, 27–35 nm in diameter, with a single-stranded positive sense polyadenylated RNA genome of 7.5–7.7 kb [3]. The genome of NoVs comprises 3 ORFs. ORF1 at the 5′-end of the genome encodes for six nonstructural proteins (p48, NTPase, p22, Vpg, 3CL, and RdRp). ORF2 encodes the major structural capsid protein (VP1) and ORF3 the minor capsid protein (VP2) at the 3′-end of the genome [4, 5]. Currently, NoVs are divided into five major genogroups (GI to GV) according to the amino acid sequence diversity of the VP1 gene. GI, GII, and GIV infect humans, while GIII and GV infect bovine and murine species [6]. These five genogroups are further subdivided into at least 32 phylogenetic clusters or genotypes: Nine in GI, nineteen in GII, two in GIII, and one in GIV and GV [4].

Genotype GII-4 is the most common type causing outbreaks [7–10]. Viruses belonging to other NoV genotypes are found less consistently, causing sporadic outbreaks or temporary epidemics [11–14].

Most of the previous genetic analyses have been based on partial nucleotide sequence analyses of the genome, mainly derived from a relatively small region of either the ORF1 or ORF2 gene, differentiating a number of NoVs into their distinct lineages [15–18].

In the present study a stool specimen from a 2-year-old female child, suffering from typical clinical symptoms of viral gastroenteritis was collected in the University Hospital of Ioannina, Greece. Acute gastroenteritis was defined as consecutively episodes of diarrhea (watery or loose stools in a 24 h period), accompanied with vomiting. The detection was carried out initially by enzyme-linked immunosorbent assay (ELISA) and the subsequent detection and molecular characterization of NoV strain was achieved by reverse transcription-PCR and DNA sequencing. RT-PCR assay for GII NoVs that targets conserved regions of the genome, allowed the characterization of A6 strain at molecular level, established its genetic relationship at the sub-genogroup level and classified A6 strain at the sub-genotype level by performing phylogenetic analyses with other GII NoVs that have previously been grouped into genotypes [16]. Overall, A6 strain represents the first human NoV from Greece, whose genome has been partially (ORF1 and ORF3) and completely (ORF2) sequenced and for which its relationship with other NoVs was established.

Materials and methods

Viral RNA extraction from fecal specimen

The stool specimen was diluted into 10 ml of complete solution of phosphate-buffered saline (PBS) (pH 7.4), which was homogenized by vortex and stored at −20°C.

Viral RNA was extracted from 140 μl of the supernatant using the QIAamp® Viral RNA Mini Kit according to the manufacturer’s instructions (Qiagen, Germany).

Reverse transcription

Reverse Transcription was carried out as follows: 7 μl consisting of random primers (d N9), (Takara Biomedical group, Shiga, Japan), (50 nmol/μl), 1 μl/tube, 10 mM dNTPs, 1 μl/tube,(Ιnvitrogen, UK), ddH2O 5 μl/tube,(Sigma,USA), and 5 μl of isolated RNA were incubated at 65°C for 5 min. Reverse Transcription was carried out in a final volume of 20 μl containing the 12 μl of the previous mixture and 8 μl of a mixture containing 5× first strand buffer (4 μl/tube), 0.1 M, DTT (2 μl/tube), RNAse Out (20 U/μl, 0.5 μl/tube), ddH2O (1 μl/tube), and 0.5 μl reverse transcriptase M-MLV (200 U/μl, Ιnvitrogen, UK). The samples were sequentially incubated at 25°C for 10 min, at 37°C for 50 min and finally at 70°C for 15 min.

PCR

Three μl of the reverse transcription product was used for PCR in a total volume of 50 μl containing 10× PCR buffer, 2 mM MgCl2, 1 mM dNTPs, 0.5 μl of Paq polymerase (Stratagene), (5 U/μl) and 100 pmol of primers (Table 1). After an initial denaturation step at 95°C for 2 min, 25 amplification cycles (denaturation at 95°C for 30 s, annealing temperature according to Table 1 for 30 s and extension at 72°C for 1 min) were performed followed by a final extension step at 72°C for 5 min.

Table 1 Primers used to detect NoV by RT-PCR
Full size table

Auto Nested PCR with the same set of primers and under the same conditions for 40 cycles of amplification was used to further increase the sensitivity of the assay [19].

Primers

The primer pairs used in the present study are shown in Table 1. Initially, partial amplification of the ORF2 gene was performed using the published primer pair Mon381/Mon383 [20]. Specific primers for Norovirus were designed by Primer3 software (http://fokker.wi.mit.edu/primer3/) using published nucleotide sequences of GII viral genomes, (strains: NLV/Miami/292/1994/US, Hu/L154/2000/France, Hu/HCMC204/2006/VNM and Norwalk-like virus, GenBank accession numbers: AF414410, AY921623, EU137732 and AB078337, respectively).

Molecular cloning

After electrophoresis in 2% agarose gel in TBE buffer PCR amplicons were excised from agarose gel, purified using the QIAquick® Gel Extraction Kit (Qiagen, Germany) and cloned into the pGEM®-T Easy Vector System (Promega, USA), according to the manufacturer’s instructions.

Phylogenetic analysis

Sequence identity was achieved through BLAST (http:www.ncbi.nlm.gov/blast). The relationships among the strains were determined using the multiple alignment program CLUSTALW (http://www.ebi.ac.uk/Tools/msa/clustalw2) and the program MEGA (Molecular Evolutionary Genetic Analysis, version 4.1 Beta3).

Nucleotide sequence accession numbers

The sequences were submitted to GenBank with the accession numbers: HM172493 (A6ORF1), HM172494 (A6ORF2), and HM172495 (A6ORF3).

The GenBank accession numbers used in this study are as follows:

GI: L07418 (Southampton), AF093797 (Norwalk virus), HM87661 (CVXRNA), and AB042808 (Chiba).

GII: GU131223 (Hu/GII.9/Alingsas/p1/2009/SWE), EF190920 (Hu/NLV/DjiboutiVdG66/2003/Djibouti), AY038599 (NLV/VA97207/1997), DQ379715 (Hu/GII/GoulburnValleyG5175C/1983/AUS), AB067537 (U3GII), AB039776 (SaitamaU3), AF414407 (NLV/Florida/269/1993/US), AB067539 (U16GII), AY134748 (Snow Mountain), ×86557 (Lordsdale), and U07611 (Hawaii).

Results

Initially the primers used for the detection of A6 strain have been designed to amplify partial amplification of the ORF2 gene [21, 22]. Other regions of the genome have been also selected as targets for amplification, including the RNA-dependent RNA polymerase region and the ORF3 one.

In order to confirm the solely presence of A6 strain, nucleotide sequences were assembled in not conserved regions of the genome, using the multiple alignment program MEGA. The 3′ end of each fragment was 100% homologous with the 5′ end of the following fragment. With this approach, due to the great genetic diversity in these regions, the probability of coupling different strains that might coexist in the sample was minimized.

Four overlapping cDNAs fragments: Fragment A, from nucleotides 4529–5126; fragment Β, from nucleotides 5044–5638; fragment C, from nucleotides 5505–6501 and fragment D, from nucleotides 6466–7376, encompassing the viral RNA from the 3′ end of ORF1 to the 3′ end of ORF3, were assembled. Amplicons were cloned and three independent clones from each amplicon were sequenced to avoid any selection that might occur during cloning. The joining together of these four overlapping fragments was performed in non-conserved regions [16], using the multiple alignment program MEGA. The 3′end of each fragment was 100% homologous with the 5′ end of the following fragment.

The partial genome sequence was organized into three ORFs, with ORF2 overlapping ORF1 by 19 nucleotides and ORF3 by one nucleotide; spanning nucleotides 4527–5062, 5043–6689, and 6689–7376 for ORF1, ORF2, and ORF3, respectively.

Phylogenetic analysis

The relationship of A6 strain characterized in the present study, with other strains was assessed using sequences available from GenBank (Fig. 2). Phylogenetic analyses were conducted using the partial length of nucleotide sequences of ORF1 and the full length of nucleotide sequences of ORF2 with completely sequenced NoVs strains. The length of ORF3 was short enough to be compared without segregation. Moreover, we used partially sequenced NoVs strains that were genetically close according to BLAST clustering.

Our initial phylogenetic analysis revealed that ORF1 clustered with GII/9 strains, whereas ORF2 branched outside of the above cluster, clustering with GII/6 strains.

ORF1

The phylogenetic tree of ORF1 partial gene was constructed using the neighbor-joining method based on part of the RdRp region, corresponding to 4527–5027 nucleotides. The analysis revealed that strain A6 was genetically closed and clustered together with GII strains (Fig. 1). Specifically, the phylogenetic tree revealed two distinct phylogenetic groups with reliable bootstrap support, which matched two genogroups, GI and GII [23]. GI, contained four human NoVs from; USA (CYXRNA), UK (Southampton), Germany (Norwalk), and Japan (Chiba), whereas GII consisted of twelve strains closely related NoVs that for the purpose of discussion were divided into major subgroups: Subgroup A, contained five strains, A6ORF1, Hu/GII.9/Alingsas/p1/2009/SWE, Hu/NLV/DjiboutiVdG66/2003/Djibouti, NLV/VA97207/1997, and Hu/GII/GoulburnValleyG5175C/1983/AUS, all of which were classified to genotype GII.9. Subgroup B consisted of the remaining seven strains that were closely related to each other, but it formed a significant number of clusters and single branches. Within subgroup A strain, A6 was most closely related to NoVs strains: Hu/GII.9/Alingsas/p1/2009/SWE with nucleotide identity 94%, Hu/NLV/DjiboutiVdG66/2003/Djibouti with nucleotide identity 93%, NLV/VA97207/1997 with nucleotide identity 92%, and Hu/GII/GoulburnValleyG5175C/1983/AUS with nucleotide identity 91%.

Fig. 1
figure 1

Phylogenetic tree construct based on partial sequences of the ORF1 open reading frame of NVs. The analysis was performed using MegAlign, version 4.1 (BETA 3). The distance was calculated by the neighbor-joining method. Numbers at each branch indicate bootstrap values for the clusters supported by that branch

Full size image

ORF2

The phylogenetic analysis of complete ORF2 gene was constructed using the neighbor-joining method based upon capsid sequences available from GenBank. A total of 15 capsid sequences from NoVs genogroups GI and GII strains were used. The ORF2-based phylogenetic tree was constructed with the nucleotide sequences of four strains [34–36] that were genetically close to A6 strain ORF2 gene, plus seven other GII NoVs, in conjunction with four human GI NoVs as above for ORF1. From the phylogenetic tree presented in Fig. 2 was deduced that A6 strain belonged to the GII genogroup of NoVs and was genetically close to: NLV/Florida/269/1993/US with nucleotide identity 95%, U3GII with nucleotide identity 93%, Saitama U3 with nucleotide identity 93%, and U16GII strain with nucleotide identity 90%. It is noteworthy that the three strains (Hu/GII.9/Alingsas/p1/2009/SWE, Hu/NLV/DjiboutiVdG66/2003/Djibouti and NLV/VA97207/1997) which were most closely related to A6 strain in the ORF1-based phylogeny, belonged to genotype GII.9 regarding ORF2 gene. On the other hand, as it is pointed out in Fig. 2, four strains (NLV/Florida/269/1993/US, U3GII, SaitamaU3 and U16GII), which were most closely related to A6 strain in the phylogenies based on the ORF2 gene, belonged to genotype GII.6 regarding the ORF1 gene. These findings suggest that A6 strain is an intergenotypic recombinant within the genogroup GII between the GII.9 and GII.6 strains.

Fig. 2
figure 2

Phylogenetic tree construct based on complete sequences of the ORF2 open reading frame of NVs. The analysis was performed using MegAlign, version 4.1 (BETA 3). The distance was calculated by the neighbor-joining method. Numbers at each branch indicate bootstrap values for the clusters supported by that branch

Full size image

Recombination event

To identify the putative parent-like strains and potential recombination sites, phylogenetic profile analysis was performed using SimPlot program [24]. The plots of the nucleotide sequences are depicted in Fig. 3. When a similarity plot for the A6 strain was generated, with strains available from GenBank, a recombination breakpoint at nucleotide position 5,043 of the sequence alignment was visible between two different genotypes within the same genogroup.

Fig. 3
figure 3

SimPlot analysis for partial RdRp and capsid sequences. Vertical red line indicates the crossover point at position 5,043 in the 0RF1/ORF2 junction. The window size was 100 bp with a step size of 50 bp. The vertical axis indicates the nucleotide identities between the query sequence (A6) and the strains (listed on the window on the right of the figure), expressed as percentages. The horizontal axis indicates the nucleotide positions of the analyzed genome region

Full size image

These findings were further confirmed by boot scanning of the same genome sequences, demonstrating higher levels of phylogenetic relatedness between the A6 genome sequence and the Hu/GII.9/alingsas/p1/2009/SWE and U16GII genome sequence on the upstream and downstream side of the recombination site, respectively (Fig. 4).

Fig. 4
figure 4

Bootscan analysis of the genomic region involved in the recombination event at position 5,043 of A6 strain with eight groups of Noroviruses: Hu/GII.9/Alingsas/p1/2009/SWE, Hu/NLV/DjiboutiVdG66/2003/Djibouti, NLV/VA97207/1997, Hu/GII/GoulburnValleyG5175C/1983/AUS, U3GII, Saitama U3, NLV/Florida/269/1993/US and U16GII

Full size image

Discussion

Recombination of viral RNA is known to depend upon various immunological and intracellular constraints that may allow the emergence of viable recombinants [25]. Since its first reporting [26], many recombinant NoVs from different genotypes and genogroups have been described worldwide. Recombination intergenotype events have been reported between GII.1/GII.12, GII.b/GII.18, GII.b/GII.4, GII.b/GII.4, GII.d/GII.3, GII.3/GII.13, GII.7/GII.13, GII.b/GII.7, GII.4/GII.8, GII.5/GII.12, GI.3/GII.4, GII.1/GII.12, GII.4/GII.2, and GII4./GII.3 in RdRp and capsid gene, respectively [27–31].

Based on the sequence analysis, A6 strain was revealed belonging to the GII genogroup of NoVs. Partial ORF1 and complete ORF2 gene sequencing analysis revealed that ORF1 and ORF2 belonged to two distinct genotypes GII/9 and GII/6, respectively, and SimPlot plus boot scanning analysis confirmed that strain A6 was indeed a recombinant. The recombination site was also consistent with the fact that most of the recombinant strains have a crossover point within or around the junction of ORF1 and ORF2 [32].

To our knowledge the recombination event GII.9/GII.6 in RdRp and capsid gene, respectively, that was revealed in the present study is reported for the first time.

The circulation of a wide variety of different NoVs within a population increases the potential for mixed infections, which could result to recombination events [33]. The findings of the present report support the hypothesis that certain NoVs strains may circulate in specific areas without clinical manifestations and occasionally may change their genetic properties by recombination events.

As recombination allows the virus to increase its genetic fitness, to evolve, to spread in the population and probably escape the host immune response, our findings suggest that the huge capacity for genetic changes displayed by the NoVs will continue to generate new recombination types.