Norovirus recombination - CiteSeerX

Journal of General Virology (2007), 88, 3347–3359

DOI 10.1099/vir.0.83321-0

Norovirus recombination Rowena A. Bull, Mark M. Tanaka and Peter A. White Correspondence Peter A. White

School of Biotechnology and Biomolecular Sciences, Faculty of Science, University of New South Wales, Sydney 2052, Australia

[email protected]

Received 16 July 2007 Accepted 23 August 2007

RNA recombination is a significant driving force in viral evolution. Increased awareness of recombination within the genus Norovirus of the family Calicivirus has led to a rise in the identification of norovirus (NoV) recombinants and they are now reported at high frequency. Currently, there is no classification system for recombinant NoVs and a widely accepted recombinant genotyping system is still needed. Consequently, there is duplication in reporting of novel recombinants. This has led to difficulties in defining the number and types of recombinants in circulation. In this study, 120 NoV nucleotide sequences were compiled from the current GenBank database and published literature. NoV recombinants and their recombination breakpoints were identified using three methods: phylogenetic analysis, SimPlot analysis and the maximum x2 method. A total of 20 NoV recombinant types were identified in circulation worldwide. The recombination point is the ORF1/2 overlap in all isolates except one, which demonstrated a double recombination event within the polymerase region.

INTRODUCTION The family Caliciviridae contains four genera: Lagovirus, Vesivirus, Norovirus and Sapovirus. The latter two genera are known to cause gastroenteritis in humans. Over the last few years, norovirus (NoV)-associated gastroenteritis has reached unprecedented levels (Bull et al., 2006; Tu et al., 2007; Widdowson et al., 2004) and NoV infection is now regarded as the most common cause of acute gastroenteritis in people of all ages (Atmar & Estes, 2006). In addition, the incidence of NoV infection is predicted to rise further (Widdowson et al., 2005a). NoV is a small, round virion of 27–35 nm in diameter and possesses a single-stranded, positive-sense, polyadenylated RNA genome of 7400–7700 nt. The NoV genome consists of three open reading frames (ORFs). ORF1 encodes the non-structural proteins, including an NTPase, protease and RNA-dependent RNA polymerase (RdRp). ORF2 encodes the major capsid protein of 58–60 kDa (Hardy, 2005) and overlaps the 39 end of ORF1 by 17–20 bp, depending on the genogroup. ORF3 encodes a minor capsid protein of 22–29 kDa (Glass et al., 2000; Pletneva et al., 2001; Seah et al., 1999). NoVs are divided into five genogroups (GI–GV), but only GI, GII and GIV are known to infect humans (Vinje et al., 2004), with GII the most prevalent in acute cases of adult gastroenteritis (Atmar & Estes, 2006). GII contains porcine as well as human strains, GIII contains bovine strains and GV contains murine strains (Zheng et al., 2006). GI is further subdivided into at least eight genotypes, GII into 17 genotypes and GIII into two genotypes (Zheng et al., 2006). 0008-3321 G 2007 SGM

Printed in Great Britain

RNA recombination is one of the major driving forces of virus evolution (reviewed by Lai, 1992; Worobey & Holmes, 1999). Virus recombination can affect phylogenetic groupings, increase the virulence of the virus, confuse molecular epidemiological studies and have major implications in vaccine design. A recombinant NoV can be defined as one that clusters with two distinct groups of NoV strains when two different regions (normally the capsid and polymerase) of the genome are subjected to phylogenetic analysis (Bull et al., 2005). Naturally occurring NoV recombinants have been reported within three of the five genogroups, GI, GII and GIII (Han et al., 2004; Hardy et al., 1997; Katayama et al., 2002a). Recently, we reported a number of unique recombinant NoV GII types, where a NoV recombinant type defines a unique polymerase/capsid combination, e.g. GII.4/GII.3 (Bull et al., 2005). Until recently, NoV genotyping was performed solely on the polymerase region of ORF1 (Katayama et al., 2002b). However, later studies showed better segregation of the different strains into their respective genotypes by phylogenetic analysis of nucleotide sequences within the capsid region (Vinje et al., 2004). However, genotyping based solely on the capsid sequence may not be sufficient with the growing identification of naturally occurring recombinant NoVs, particularly as the recombination breakpoint is close to or within the ORF1/2 overlap (Bull et al., 2005). Thus, recombinants would be missed by sequencing only the capsid region. Moreover, genotyping of recombinants has been further complicated as recombinants may cluster with a capsid genotype but the polymerase sequence does not cluster with polymerases from characterized strains. 3347

R. A. Bull, M. M. Tanaka and P. A. White

These unclassified polymerase clusters therefore need a classification system that separates them from wild-type polymerases that have the same genotype as the corresponding capsid genotype. One such polymerase system has been proposed and used for the novel Hilversumlike polymerase cluster (Gallimore et al., 2004a), which was termed GII.b (Buesa et al., 2002). For consistency, we have adhered to this system with the novel polymerase regions categorized in this study, GII.a, GII.c and GII.d. Increased awareness and identification of naturally occurring recombinants worldwide has led to a rise in their report (Ambert-Balay et al., 2005; Bon et al., 2005; Bull et al., 2005; Etherington et al., 2006; Gallimore et al., 2004a, 2004b, 2005; Han et al., 2004; Hansman et al., 2004a, 2004b; Iritani et al., 2003; Katayama et al., 2002b; Martinez et al., 2002; Oliver et al., 2006; Phan et al., 2006a, 2006c; Reuter et al., 2005, 2006; Sasaki et al., 2006; Tsugawa et al., 2006; van den Berg et al., 2005; Wang et al., 2005). However, difficulties in genotyping and classifying the polymerase region of NoV GII has led to confusion, with many groups claiming a recombinant is novel when it has already been published as a recombinant. For example, we first reported the recombinant Hu/NoV/Picton/03/AU as novel (Bull et al., 2005). This recombinant NoV has the novel GII.b polymerase and a GII.1 capsid. Subsequently, almost identical strains with the same polymerase and capsid genotypes were reported as novel recombinants by three other groups (Ambert-Balay et al., 2005; Reuter et al., 2006; Vidal et al., 2006). In addition, another recombinant type, also with the novel GII.b polymerase but with a GII.3 capsid, was first reported in 2001 (Buesa et al., 2002), but because Buesa et al. (2002) did not release the capsid sequence on GenBank, the same recombinant type was subsequently reported as novel in four separate publications (Ambert-Balay et al., 2005; Bull et al., 2005; Gallimore et al., 2004a; Reuter et al., 2006). The GII.b/GII.3 NoV recombinant caused hundreds of outbreaks of gastroenteritis across Europe in 2000 and 2001 (Ambert-Balay et al., 2005; Bon et al., 2005; Gallimore et al., 2004a, 2005; Reuter et al., 2006) and subsequently multiple outbreaks across Australasia and Asia (Bull et al., 2005; Phan et al., 2006a, 2006c). The first full-length GII.b polymerase sequence reported in the database was NoV/SydneyC14/ 02/AU (GenBank accession no. AY845056) (Bull et al., 2005). The GII.b polymerase has now been associated with four different capsid genotypes, GII.1, GII.2, GII.3 and GII.4 (Reuter et al., 2006), none of which are speculated to be the GII.b polymerases matching native capsid sequence (Reuter et al., 2006). The aim of this study was to clarify some of the confusion in NoV recombination identification by characterizing the number of known NoV recombinants from all five genogroups circulating worldwide. In addition, we have promoted a system for tentatively classifying the polymerase clusters that cannot be linked to a capsid genotype. The genetic relationship among all identified recombinant types 3348

was explored and the recombination breakpoints accurately determined using three methods.

METHODS Literature and database searches were performed to identify NoV recombinant strains. One hundred and twenty single sequences, comprising 30, 60, 19, four and seven sequences for NoV GI, GII, GIII, GIV and GV, respectively, were downloaded into MEGA 3.1 (Kumar et al., 2004) from the GenBank/EMBL sequence databases. The breakpoint for NoV recombinants has been identified previously within the ORF1 and ORF2 overlap (Bull et al., 2005). Therefore, strains were defined as recombinant if their polymerase and capsid regions segregated into different genotypic clusters in the two separate trees: one of the 39 end of ORF1 and the other of the 59 end of ORF2. ORF1 and ORF2 sequences were only included if they had been obtained from a single amplicon. The size of the ORF1 and ORF2 regions analysed was dependent on the length of the sequences available. The ORF1 tree comprised 277, 420, 210, 819 and 1003 bp of the 39 end of ORF1 for NoV GI, GII, GIII, GIV and GV, respectively. The ORF2 tree comprised 295, 550, 898, 1648 and 1000 bp of the 59 end of ORF2 for NoV GI, GII, GIII, GIV and GV, respectively. Nucleotide sequence alignments were performed using CLUSTAL W (Thompson et al., 1994) in MEGA 3.1 (Kumar et al., 2004) and were visualized in GeneDoc (Nicholas et al., 1997). Phylogenetic trees were constructed using the neighbour-joining method and were based on the Tajima– Nei distance algorithm (Tajima & Nei, 1984). Bootstrap values were determined from 100 bootstrap resamplings of the original data. Two further methods were used to confirm the NoV recombinants identified by phylogenetic analysis and included the maximum x2 method (Maynard Smith, 1992) and SimPlot (version 2.5) (Lole et al., 1999). The maximum x2 method and SimPlot are also able to identify putative recombination breakpoints. The maximum x2 method uses the distribution of polymorphic sites between a probable recombinant and its putative ‘parents’ to estimate a recombination junction. The maximum x2 method of Maynard Smith has been recognized as one of the most accurate when compared independently with 13 other methods (Posada, 2002). SimPlot is a diversity plot and was the simplest of the methods used. SimPlot uses a graphical window to display the genetic distance comparisons between a chosen sequence and comparison sequences.

RESULTS NoV strains were only defined as recombinants if they met the criteria, described below, for all three methods used in this study. For phylogenetic analysis, strains were defined as recombinants if different regions of their sequence segregated into separate phylogenetic groups, with bootstrap values of the branches for each cluster of .60 %. SimPlot analysis was the second method utilized and strains were defined as recombinants if a crossover event occurred between the two putative parental strains. Parental strains were defined as non-recombinant strains with closest related sequence using BLAST searches of the two separate polymerase and capsid regions. The maximum x2 method was the third analysis performed on the suspected recombinants and strains were defined as recombinants if the crossover event was found to be significant (P,0.01). Analysis of 120 NoV strains with these three methods identified 29 recombinants that could be grouped into 20 unique, Journal of General Virology 88

Norovirus recombination

Table 1. Representatives of the 20 NoV recombinant types NoV Recombinant Prototype strain Polymerase Capsid Breakpoint: Estimated First published genogroup group genotype genotype SimPlot/x2 P value I

1

WUGI/01/JP

GI.2

GI.6

211/+1

II

1

Picton/03/AU

GII.b

GII.1

26/246

2

GII.c

GII.2

26/2104

GII.4 GII.b

GII.2 GII.2

232/217 235/27

5

Snow Mountain2/ 76/US E3/97/Crete Pont de Roide 673/ 04/Fr SydneyC14/02/AU

GII.b

GII.3

+22/+23

6

Sydney2212/98/AU

GII.a

GII.3

7

SaitamaT66e/02/JP

GII.d

GII.3

8

Chiba1/04/JP

GII.4

GII.3

9

771/05/IRL

GII.4/GII.d

GII.4

3 4

10

GII.b

GII.4

GII.2 GII.d

GII.5 GII.5

13

Nyiregyhaza/1057/ 02/HUN* S63/99/Fr Hokkaido133/03/ JP Kunming/04/Ch

GII.6

GII.7

14

Mc37/01/Th

GII.4

GII.10

15

SaitamaU1/02/JP

GII.4

GII.12

16

Minato14/99/JP

GII.6

GII.15

17 1 2

VannesL23/99/US B-1SVD/03/US CV521-OH/02/US

GII.5 GIII.2 GIII.1

GII.15 GIII.1 GIII.2

11 12

III

Related strains reported in the literature

,0.0001 Katayama et al. (2002b) ,0.0001 Bull et al. (2005) Ambert-Balay et al. (2005); Bon et al. (2005); Phan et al. (2006b); Reuter et al. (2006); Vidal et al. (2006) ,0.0001 Hardy et al. (1997) ,0.0001 Bull et al. (2005) Ambert-Balay et al. (2005) ,0.0001 Bon et al. (2005) Reuter et al. (2006)

,0.0001 Buesa et al. Ambert-Balay et al. (2005); (2002); Lole et al. Bon et al. (2005); Bull et al. (1999) (2005); Gallimore et al. (2004a, 2005); Phan et al. (2006a, 2006c); Reuter et al. (2006) Bull et al. (2005); Gallimore 214/2104 0.0001 Jiang et al. et al. (2004b); Iritani et al. (1999) (2003); Martinez et al. (2002); Tsugawa et al. (2006) 222/26 ,0.0001 Phan et al. (2007) 211/+24 ,0.0001 Vidal et al. Gallimore et al. (2004a); (2006) Phan et al. (2006c) 2594/2596 ,0.0001 Waters et al. 2203/2221 (2007) Gallimore et al. Bon et al. (2005); Reuter et (2004a) al. (2006) +38/+32 ,0.0001 Bull et al. (2005) Ambert-Balay et al. (2005) 229/26 ,0.0001 Phan et al. (2007) 29/298 0.0013 Phan et al. (2006d) 216/+23 ,0.0001 Hansman et al. (2004b) 230/247 ,0.0001 Katayama et al. Bull et al. (2005); (2002b) Etherington et al. (2006) 260/246 ,0.0001 Sasaki et al. (2006) 215/246 ,0.0001 Bull et al. (2005) Ambert-Balay et al. (2005) 227/213 ,0.0001 This study 27/214 ,0.0001 Han et al. (2004); Oliver et al. (2004)

*Sequence data for the polymerase and capsid was not available in the database for this strain and it was therefore not confirmed as a recombinant in this study.

naturally occurring NoV recombinant types: one NoV GI, 17 NoV GII and two NoV GIII (Table 1). All GIV and GV strains available in GenBank, four and seven strains, respectively, were analysed for recombination but none was identified as recombinant (data not shown). http://vir.sgmjournals.org

NoV GI recombinants Thirty NoV GI strains representing all eight genotypes were analysed for recombination. Two of these strains did not have available ORF1 sequence but were still included in the 3349


capsid sequence tree as representatives of GI.7 and GI.8 (Fig. 1). Using phylogenetic analysis, seven NoV GI recombinants were identified, including the previously reported GI recombinant NLV/WUGI/01/JP (Katayama et al., 2002b) (Fig. 1). All seven recombinants were related (93–98 % nucleotide identity), with a GI.2 polymerase region and a GI.6 capsid, and therefore all belonged to the same recombinant type, NoV GI.2/GI.6 (Fig. 1) (Table 1). NoV GII recombinants Seventeen genetically diverse NoV GII recombinant types were identified by the three methods: phylogenetic analysis, SimPlot analysis and the maximum x2 method. Due to the large number of NoV GII sequences in the GenBank database, only recombinant prototypes (Table 1) were chosen for analysis, and duplicates of these recombinant types were not included in the analysis of the GII recombinants. All 17 prototype GII recombinants were

a combination of one of eight different polymerase genotypes and one of nine different capsid genotypes (Fig. 2, Table 1). Genotyping of the polymerase region from the 17 recombinants revealed that five belonged to GII.4 and four belonged to the novel GII.b polymerase cluster (Buesa et al., 2002). It should be noted that due to a double recombination event, the sequence from one of the GII.4 polymerases (NoV/771/05/IRL) was a combination of two polymerase genotypes, GII.4 and GII.d; however, as the majority of the sequence was GII.4, it was defined as a GII.4 polymerase. The remaining eight polymerases belonged to GII.2, GII.5, GII.6 and three novel genotypes, termed in this study GII.a, GII.c and GII.d (Table 1). The novel polymerases had greater than 15 % nucleotide difference compared with other classified polymerase genotypes. The capsid genotypes associated with recombinant NoV were more diverse with one GII.1, three GII.2, four GII.3, two GII.4, two GII.5, one GII.7, one GII.10, one GII.12 and two GII.15 (Fig. 2) (Table 1).

Fig. 1. Phylogenetic analysis of NoV GI recombinants. Phylogenetic analysis of the nucleotide sequences of polymerase and capsid regions of one NoV GI recombinant type in relation to 30 known strains and prototype strains. The left tree is obtained from a 277 bp region of the 39 end of the polymerase region. The right tree corresponds to 295 bp of the 59 end of the capsid sequence. Suspected recombinants described in this study are represented in bold. The percentage bootstrap values in which the major groupings were observed among 100 replicates are indicated. The branch lengths are proportional to the evolutionary distance between sequences and the distance scale, in nucleotide substitutions per position, is shown. The polymerase and capsid clustering is shown in bold and is based on the classification of Zheng et al. (2006). 3350

Journal of General Virology 88


Fig. 2. Phylogenetic analysis of NoV GII recombinants. Phylogenetic analysis of the nucleotide sequences of 19 identified recombinant NoV GII strains in relation to 38 polymerase sequences and 37 capsid sequences from known strains and prototype strains. The 19 NoV GII recombinants group into 16 unique recombinant types; the 17th NoV GII recombinant type, NoV/Nyiregyhaza/1057/00/HUN, is not included in the figure. The left tree is based on a 420 bp region from the 39 end of the polymerase region. The right tree is based on 550 bp from the 59 end of the capsid sequence. Suspected recombinants described in this study are represented in bold. The percentage bootstrap values in which the major groupings were observed among 100 replicates are indicated. The branch lengths are proportional to the evolutionary distance between sequences and the distance scale, in nucleotide substitutions per position, is shown. The polymerase and capsid clustering is shown in bold and is based on the classification of Zheng et al. (2006).

Due to the large genomic diversity within the NoV GII strains, some recombinant strains were difficult to define. For example, it was difficult to determine phylogenetically whether NoV/Kunming/04/Ch was a recombinant or a divergent NoV GII strain. NoV/Kunming/04/Ch has a GII.6 polymerase and a GII.7 capsid, but due to close clustering of these two genotypes in the polymerase and capsid genes, it is difficult to distinguish this as a recombinant from the two phylogenetic trees (Fig. 2). In the present study, NoV/Kunming/04/Ch was defined as a recombinant as it met the criteria for all three methods; although, the P value for its recombinant breakpoint was higher than the P values calculated http://vir.sgmjournals.org

for the other recombinants, it was still less than 0.01 (Table 1). NoV/Nyiregyhaza/1057/02/HUN (Reuter et al., 2006) is also a suspected recombinant, with a GII.b polymerase and GII.4 capsid, but the polymerase and capsid sequence for this strain were not available for analysis. However, it was defined as a recombinant prototype in this study because, in contrast to Hu/Noguchi-8/00/GH (described below), the GII.b/GII.4 recombinant type has been reported elsewhere (Bon et al., 2005). Additionally, its polymerase and capsid regions grouped into two phylogenetically distinct and well-characterized clusters: GII.b and GII.4 (Fig. 2). 3351


Unconfirmed NoV GII recombinants In this study, there were four suspected recombinants that failed to meet the recombination analysis criteria for the three methods utilized. Hu/Noguchi-8/00/GH (GenBank accession no. DQ013132) is a previously reported recombinant (Armah et al., 2006). However, Hu/Noguchi-8/00/ GH could not be confirmed as a recombinant in this study as its polymerase sequence has not been released on GenBank/EMBL and therefore we were unable to perform our own analysis. Armah et al. (2006) reported that Hu/ Noguchi-8/00/GH clusters with GII.8 in the polymerase region and GII.14 in the capsid region. Our analysis of the capsid region classified this capsid sequence as GII.13 (according to Zheng et al., 2006) (data not shown). Additionally, in the polymerase region, GII.8 clusters closely together with GII.6, GII.7, GII.8, GII.9, GII.13 and GII.14 (Fankhauser et al., 2002; Fig. 2). Consequently, Hu/ Noguchi-8/00/GH may not be a true recombinant, but could be a divergent GII.13 wild type. Phylogenetic and SimPlot analysis of NoV/Yuri/02/JP and NoV/OsakaNI/04/JP indicated that they may also be recombinants, with a putative crossover event around the

ORF1/2 overlap (Fig. 3a, b). Both Yuri/02/JP and OsakaNI/ 04/JP had a GII.d polymerase (Fig. 2). However, genotyping of their capsids revealed that OsakaNI/04/JP clustered with GII.2 strains, whilst Yuri/02/JP did not group with any of the 17 genotypes published for NoV GII (Zheng et al., 2006) and differed by 28 % over full-length ORF2 nucleotide sequences compared with any other sequence in GenBank. Therefore, in this study, it was tentatively termed GII.18 for its capsid genotype. However, the maximum x2 analysis failed to find a significant breakpoint (P.0.05) and therefore they were not classified as recombinants in this study. The lack of statistical significance is probably due to the inability to find a suitable wild-type parental strain related to the recombinant in the polymerase region. However, the divided clustering of their capsids indicates that at least one, if not both, of these strains is a recombinant. NoV strain NoV/Pont de Roide 671/04/Fr was similar to Yuri/02/JP and OsakaNI/04/JP, with phylogenetic and SimPlot analysis indicating that its capsid and polymerase regions clustered separately (Fig. 3c). Its polymerase clustered with GII.4 and its capsid clustered with GII.14 (Fig. 2). However, the maximum x2 method was unable to

Fig. 3. Similarity plot of NoV GII recombinants: (a) Yuri/02/JP, (b) OsakaNI/04/JP, (c) Pont de Roide 671/04/Fr, (d) 771/05/ IRL. The graph represents as a percentage the identity of the two putative parental strains (black and grey lines) with their respective recombinant. The genomic position is numbered so that the first nucleotide of ORF2 is equal to 1, with numbering increasing towards the 39 end and decreasing towards the 59 end of the genome. The window size was 100 bp with a step size of 10 bp. The site where the two parental strains have equal identity to the recombinant (i.e. where the lines cross) is the predicted site of recombination. 3352



find a statistically significant breakpoint, probably also due to the unavailability of a non-recombinant parental strain with a related polymerase gene. Therefore, it was not included as a recombinant type in this study. Intragenotype recombination in NoV GII All 20 recombinants defined in this study were intergenotypic recombinants, i.e. recombinants formed from two different genotypes. Whilst intragenotypic recombination, i.e. recombination between strains from the same genotype, is possible, it is difficult to identify and detection would require highly sensitive methods. Nevertheless, nine intragenotype NoV GII recombinants have been published. Intragenotype recombination has been reported for Hu/ NLV/Saitama U3/02/JP (GenBank accession no. AB039776) (Etherington et al., 2006) and Hu/NLV/GII/ MD145-12/87/US (GenBank accession no. AY032605) (Etherington et al., 2006). In the present study, Saitama U3/02/JP clustered with GII.6 strains in both the polymerase and capsid genes and MD145-12/87/US clustered with GII.4 strains in both the polymerase and capsid genes (Fig. 2). SimPlot analysis and the maximum

x2 method were not able to detect a recombination

breakpoint (P.0.05) within Saitama U3/02/JP and MD145-12/87/US. Therefore, we cannot confirm either strain as a recombinant NoV. Recently, seven other GII NoVs were reported as recombinants with crossover points distributed across the capsid gene (Phan et al., 2006a; Rohayem et al., 2005). They were all defined as intragenotypic recombinants with less than 5 % variation between the two parental strains and the putative recombinant. Analysis of these seven strains failed to identify them as recombinants by either SimPlot analysis or the maximum x2 method (P.0.05). Therefore, as our analysis was not able to detect any significant recombination events, these seven strains were not defined as recombinants. NoV GIII recombinants Nineteen NoV GIII strains representing two genotypes were compiled from GenBank and were analysed for recombination (Fig. 4). Three recombinants were identified: Bo/NoV/B-1SVD/03/US, Bo/NoV/CV521-OH/02/ US and Bo/NoV/Thirsk10/00/UK (Fig. 5) (Table 1). CV521-OH/02/US and Thirsk10/00/UK both had a GIII.1

Fig. 4. Phylogenetic analysis of NoV GIII recombinants. Phylogenetic analysis of the nucleotide sequences of polymerase and capsid regions identified two NoV GIII recombinant types in relation to 19 known strains and prototype strains. The left tree is based on a 210 bp region from the 39 end of the polymerase region. The right tree is based on 898 bp from the 59 end of the capsid sequence. Suspected recombinants described in this study are represented in bold. The percentage bootstrap values in which the major groupings were observed among 100 replicates are indicated. The branch lengths are proportional to the evolutionary distance between sequences and the distance scale, in nucleotide substitutions per position, is shown. The polymerase and capsid clustering is shown in bold and is based on the classification of Zheng et al. (2006). http://vir.sgmjournals.org

3353


Recombination breakpoint The site of recombination was measured from the start codon of the capsid for all three genogroups, with the first nucleotide of ORF2 equal to 1, with numbering increasing towards the 39 end and decreasing towards the 59 end of the genome. The crossover points for all 20 recombinant types, as determined by SimPlot and the maximum x2 method, were plotted in relation to the NoV GII prototype strain, Lordsdale (GenBank accession no. X86557) (Fig. 6). All crossover sites, with the exception of the crossover sites for recombinant 771/05/IRL, calculated by SimPlot analysis and the maximum x2 method were between –104 and +32, and had a mean position of 216 and 219, respectively. Therefore, the mean crossover point lies between 16 and 19 nt upstream of the start of the ORF1/2 overlap.

Fig. 5. Similarity plots of NoV GIII recombinants CV521-OH/02/ US, B-1SVD/03/US and Thirsk10/00/UK. The graphs represent as a percentage the identity of the two putative parental strains, Newbury 2 and Jena, with the two recombinants B-1SVD/03/US (a) and CV521-OH/02/US (b). The percentage identity of the three recombinants B-1SVD/03/US, CV521-OH/02/US and Thirsk10/00/UK were also included in the plot. The genomic position is numbered so that the first nucleotide of ORF2 is equal to 1, with numbering increasing towards the 39 end and decreasing towards the 59 end of the genome. The window size was 100 bp with a step size of 10 bp. The site where the two parental strains have equal identity to the recombinant (i.e. where the lines cross) is the predicted site of recombination.

polymerase and a GIII.2 capsid. B-1SVD/03/US had a GIII.2 polymerase and a GIII.1 capsid. Therefore, the three recombinants were categorized into two recombinant types, NoV GIII.1/GIII.2 and NoV GIII.2/GIII.1.

Recombinant 771/05/IRL was not included with the breakpoint analysis of the other recombinants because two recombination sites were identified for 771/05/IRL by SimPlot and maximum x2 analysis, 2203/2594 and 2221/ 2596, respectively (Fig. 3d). Both of these recombination points in 771/05/IRL occurred well upstream of the ORF1/2 overlap. Recombinant 771/05/IRL shares over 90 % sequence identity over 1063 bp with GII.4 strain NoV/Sakai/04-179/ 05/JP except for the region between approximately –595 and –212 where identity drops to approximately 60 % and identity to another GII.d strain, OsakaNI/04/JP, is over 90 %. Except for this 383 bp region, sequence identity to OsakaNI/ 04/JP is less than 84 % (Fig. 3d). Due to the low sequence identity between the two putative parental strains, OsakaNI/ 04/JP and NoV/Sakai/04-179/05/JP, but the high sequence identity of the recombinant to its two putative parental strains, the two breakpoints, 2590 and 2216, were able to be determined visually from the sequence alignment (Fig. 7). The breakpoints corresponded to the breakpoints determined by SimPlot analysis and the maximum x2 method. Circulation of related polymerases with genetically diverse capsid types Analysis of the 17 confirmed GII recombinants and three of the suspected recombinants, Pont de Roide 671/04/Fr, Yuri/02/JP and OsakaNI/04/JP, revealed that 12 (60 %) of

Fig. 6. Comparison of the breakpoints for all 20 NoV recombinants as determined by SimPlot and maximum x2 analysis. The breakpoint for each of the 20 recombinant types identified was plotted in relation to the NoV GII prototype strain, Lordsdale (GenBank accession no. X86557). The mean breakpoint site was determined for both methods and is represented by a vertical line. Recombinant 771/05/IRL was not included in determining the mean and appears as two outliers upstream of the ORF1/2 overlap. 3354



Fig. 7. Nucleotide sequence alignment of NoV recombinant 771/05/IRL with its two putative parental strains, Sakai/04-179/ 05/JP (Sakai/04) and OsakaNI/04/JP (OsakaNI/04). The two recombination sites, approximately at nt 4545 and 4920, are marked with boxes and the numbering system is based on the full-length genome sequence of OsakaNI/04/JP (GenBank accession no. DQ366347). The alignment illustrates that up to nt 4545 and after nt 4920, the recombinant 771/05/IRL and the parental strain Sakai/04-179/05/JP have a higher nucleotide identity than OsakaNI/04/JP. However, between nt 4545 and 4920, OsakaNI/04/JP and recombinant 771/05/IRL have a higher nucleotide identity than Sakai/04-179/05/JP. It should be noted that, despite putative parental strain OsakaNI/04/JP being a suspected recombinant, it was used as a parental strain primarily because of the lack of a suitable non-recombinant strain and also because the recombination crossover point in OsakaNI/04/JP is in a different genomic location to the recombination crossover points in 771/05/IRL and therefore should not influence the analysis of recombinant 771/05/IRL.

these 20 strains belonged to three (38 %; GII.b, GII.4 and GII.d) of the eight polymerase clusters associated with recombinant strains. Therefore, it is possible that some polymerases are more prone to recombine than other polymerases. The GII.b polymerase was associated with four genotypically different capsids: GII.1, GII.3, GII.4 and GII.14, indicating four separate recombination events for viruses with a GII.b polymerase. The three GII.b recombinant-type polymerases associated with GII.1, GII.3 or GII.14 capsid shared greater than 95 % identity across 420 nt of the 39 end of the polymerase region. http://vir.sgmjournals.org

The GII.4 polymerase was associated with five genotypically different capsids: GII.2, GII.3, GII.10, GII.12 and GII.14 (Table 1, Fig. 2). Four of the five recombinant GII.4 polymerases clustered together (98 % nucleotide identity) and independently of the GII.4 polymerases that have been associated with four pandemics in the last decade, Farmington Hills, US95/96 (representative strains of US95/96 in Fig. 2 include Burwash Landing and Miami Beach 326), Hunter and 2006a (Fig. 2). The GII.d polymerase was associated with four genotypically different capsids: GII.2, GII.3, GII.5 and GII.18 (Fig. 2). Furthermore, this study indicated that all of the 3355


GII.d polymerases in the database are either confirmed or suspected recombinants.

DISCUSSION Recombination is a driving force of viral evolution and has been reported for many single-stranded RNA viruses, including the emerging RNA virus, NoV. The large genomic diversity, the lack of a prototype for each recombinant type and the use of different methods to define recombinants has confused their classification and resulted in duplication of the reporting of novel recombinant types. In order to address this confusion, we have nominated a prototype strain for each recombinant type (Table 1) and proposed a genotyping system for novel polymerase clusters. NoV strains were only defined as recombinants in this study if they were identified as recombinant NoV using all three methods: phylogenetic, SimPlot and maximum x2 analysis. Phylogenetic analysis is useful when a parental strain cannot be identified for one of the regions being examined. SimPlot analysis and the maximum x2 method are useful for identifying the breakpoint, but require parental strains. For NoV, this is a problem due to the large genetic diversity of circulating strains. However, phylogenetic analysis cannot be solely relied on to identify recombinants, as the nonantigenic region (ORF1) does not cluster as definitively as the antigenic (ORF2) region. Thus, low bootstrap support for some polymerase genotypes may result in artefactual identification of some recombinants (Vinje et al., 2004). Therefore, it is important for recombinant analysis also to use other methods, such as SimPlot and maximum x2, that are independent of phylogeny. In this study, we analysed NoV strains from all five NoV genogroups for recombination and confirmed 20 unique naturally occurring NoV recombinant types: one NoV GI, 17 NoV GII and two NoV GIII (Table 1). The GI recombinant, WUGI/01/JP (GenBank accession no. AB081723), has been identified previously (Katayama et al., 2002b) and is the prototype for this recombinant type. There were six other WUGI/01/JP-like strains identified from GenBank and they were isolated in Japan and the USA between 2000 and 2004, suggesting that this recombinant type virus is prevalent. Two GIII recombinant types were identified. The GIII.2/ GIII.1 recombinant type was identified for the first time in this study. The second recombinant type, GIII.1/GIII.2, includes the previously published strains Thirsk10/00/UK and CV521-OH/02/US (Han et al., 2004; Oliver et al., 2004). Interestingly, Thirsk10/00/UK and CV521-OH/02/ US had not been recognized previously as being the same recombinant type, probably because the two strains have low sequence similarity (84 % nucleotide identity over ORF1 and ORF2), thereby suggesting that recombination may have occurred some time ago and that both isolates have since diverged. 3356

Of the 17 NoV GII recombinant types identified, GII.4 and GII.b were the two most common polymerases regardless of their corresponding capsid genotypes. The term GII.b was first published by Buesa et al. (2002) and arose as its polymerase forms a novel cluster with no known matching capsid sequence. The GII.b polymerase has been associated with outbreaks across the world (Bon et al., 2005; Buesa et al., 2002; Bull et al., 2006; Lindell et al., 2005; Maunula & Von Bonsdorff, 2005; Nygard et al., 2003; Reuter et al., 2006). Furthermore, since its first isolation in 2000 with a GII.3 capsid, it has been associated with three further capsid types (Bon et al., 2005; Reuter et al., 2006). Whether this is a result of recent recombination events or more thorough epidemiological investigations is unknown. The GII.4 and GII.b polymerases are the two most prevalent polymerases circulating around the world (Bull et al., 2006; Gallimore et al., 2004a; Phan et al., 2006b; Reuter et al., 2005). Therefore, the probability of finding a GII.4 or GII.b recombinant is higher as recombination requires co-infection with two NoV strains. However, this is also true of the capsid and, despite GII.4 being the most prevalent capsid type worldwide (Bull et al., 2006), there were only two recombinant types with a GII.4 capsid (not including recombinant 771/05/IRL). The lack of multiple polymerase genotypes associated with GII.4 capsids and the fact that the GII.b and GII.4 polymerases are each associated with four and five different capsid types, respectively, suggests that the polymerase may be a driving factor in recombination. All of the NoV recombinants from GI, GII and GIII, with the exception of one NoV GII recombinant, had a crossover point either within or close to the ORF1/2 overlap. Recombination between animal RNA viruses has been shown to require sequence homology, poor processivity of the RdRp, or complex secondary structures, such as stem–loop structures, that will lead to interaction of the parental strains (Kim & Kao, 2001). The ORF1/2 overlap includes the subgenomic promoter (Asanaka et al., 2005) and consequently has a stem–loop structure that is 100 % conserved within each genogroup (Bull et al., 2005). This is consistent with the proposed model (Bull et al., 2005), which suggests that recombination occurs when the polymerase switches templates mid-transcription due to complex secondary structure at the start of ORF2. Consequently, polymerases with poor processivity would switch templates at a higher frequency than other RdRps. The ability of polymerases to switch templates at the start of ORF2 is advantageous because it can help viruses escape evolutionary bottlenecks (Coyne et al., 2006; Muller, 1964). This is because ORF2 encodes the capsid protein, VP1, which contains the antigenic regions. Therefore, viruses that are able to swap their capsid coat are able to escape immune responses and possible viral extinction. Unlike the GII.b and GII.4 recombinants, the other recombinants do not seem to have diffused as widely throughout the community. The reason why some strains Journal of General Virology 88


dominate over others is unknown, but the fact that the GII.b polymerase is associated with four different capsid types suggests that it may use recombination as a way of escaping population bottlenecks in the host population, as evidenced by the retention of polymerase genes but the loss of their equivalent capsid types. Whilst there are clearly recombination hotspots within the picornaviruses and coronaviruses, recombination has been seen to occur in vitro throughout the entire genome of both (reviewed by Lai, 1992). However, recombination has not been detected within the capsid regions encoding VP1 and VP3 in picornaviruses (reviewed by Lai, 1992). Whilst recombination within the capsid gene has been suggested for NoV GII (Etherington et al., 2006; Phan et al., 2006a; Rohayem et al., 2005), analysis of those strains in this study did not yield any statistically significant breakpoints. For picornaviruses, it has been speculated that recombination in the capsid gene could lead to non-functional or unstable products and is consequently selected against (reviewed by Lai, 1992). Recently, the first NoV recombinant with a breakpoint outside the ORF1/2 overlap was detected (Waters et al., 2007). Further analysis of this strain, 771/05/IRL, in this study revealed that it had a double breakpoint, which was not reported by Waters et al. (2007). Our study is the first report of a double recombination event in NoV. The region where the genetic exchange occurred could clearly be seen in the nucleotide alignments and matched the crossover sites identified by SimPlot and the maximum x2 method. The second putative crossover site occurred in a region with high sequence similarity between the two parental strains; the putative crossover point at the 59 end of the genome, however, shared very little sequence similarity between the putative parental strains. The lack of a visible RNA promoter or secondary structure in this region suggests that this recombination event may have arisen by other mechanisms to those that induce a breakpoint in or around the ORF1/2 overlap. However, the lack of insertions or deletions at the putative crossover sites suggests that it is a result of homologous recombination and lends support to the template switching model proposed for NoV (Bull et al., 2005). NoV has been detected in a wide range of mammals, including humans, mice, cows and pigs (Wang et al., 2005), with strong evidence of zoonotic transmission (Widdowson et al., 2005b). Interspecies exchange has been reported for many RNA viruses, the most notable being reassortment (equivalent to recombination in segmented genomes) between the human and avian influenza strains to produce a highly virulent virus (de Jong et al., 1997). In NoV, intergenogroup recombination could result in recombination between different mammalian NoVs. Indeed, intergenogroup recombination in a closely related calicivirus, sapovirus, has already been reported (Hansman et al., 2005). Therefore, understanding recombination in NoV is important, as recombination between the different http://vir.sgmjournals.org

mammalian NoVs may result in the emergence of new NoV variants, with potentially different pathogenesis and virulence.

REFERENCES Ambert-Balay, K., Bon, F., Le Guyader, F., Pothier, P. & Kohli, E. (2005). Characterization of new recombinant noroviruses. J Clin

Microbiol 43, 5179–5186. Armah, G. E., Gallimore, C. I., Binka, F. N., Asmah, R. H., Green, J., Ugoji, U., Anto, F., Brown, D. W. & Gray, J. J. (2006). Characterisation

of norovirus strains in rural Ghanaian children with acute diarrhoea. J Med Virol 78, 1480–1485. Asanaka, M., Atmar, R. L., Ruvolo, V., Crawford, S. E., Neill, F. H. & Estes, M. K. (2005). Replication and packaging of Norwalk virus

RNA in cultured mammalian cells. Proc Natl Acad Sci U S A 102, 10327–10332. Atmar, R. L. & Estes, M. K. (2006). The epidemiologic and clinical

importance of norovirus infection. Gastroenterol Clin North Am 35, 275–290. Bon, F., Ambert-Balay, K., Giraudon, H., Kaplon, J., Le Guyader, S., Pommepuy, M., Gallay, A., Vaillant, V., de Valk, H. & other authors (2005). Molecular epidemiology of caliciviruses detected in sporadic

and outbreak cases of gastroenteritis in France from December 1998 to February 2004. J Clin Microbiol 43, 4659–4664. Buesa, J., Collado, B., Lopez-Andujar, P., Abu-Mallouh, R., Rodriguez Diaz, J., Garcia Diaz, A., Prat, J., Guix, S., Llovet, T. & other authors (2002). Molecular epidemiology of caliciviruses

causing outbreaks and sporadic cases of acute gastroenteritis in Spain. J Clin Microbiol 40, 2854–2859. Bull, R. A., Hansman, G. S., Clancy, L. E., Tanaka, M. M., Rawlinson, W. D. & White, P. A. (2005). Norovirus recombination in ORF1/ORF2

overlap. Emerg Infect Dis 11, 1079–1085. Bull, R. A., Tu, E. T., McIver, C. J., Rawlinson, W. D. & White, P. A. (2006). Emergence of a new norovirus genotype II.4 variant

associated with global outbreaks of gastroenteritis. J Clin Microbiol 44, 327–333. Coyne, K. P., Reed, F. C., Porter, C. J., Dawson, S., Gaskell, R. M. & Radford, A. D. (2006). Recombination of Feline calicivirus within an

endemically infected cat colony. J Gen Virol 87, 921–926. de Jong, J. C., Claas, E. C., Osterhaus, A. D., Webster, R. G. & Lim, W. L. (1997). A pandemic warning? Nature 389, 554. Etherington, G. J., Dicks, J. & Roberts, I. N. (2006). High throughput

sequence analysis reveals hitherto unreported recombination in the genus Norovirus. Virology 345, 88–95. Fankhauser, R. L., Monroe, S. S., Noel, J. S., Humphrey, C. D., Bresee, J. S., Parashar, U. D., Ando, T. & Glass, R. I. (2002).

Epidemiologic and molecular trends of ‘‘Norwalk-like viruses’’ associated with outbreaks of gastroenteritis in the United States. J Infect Dis 186, 1–7. Gallimore, C. I., Cubitt, D., du Plessis, N. & Gray, J. J. (2004a).

Asymptomatic and symptomatic excretion of noroviruses during a hospital outbreak of gastroenteritis. J Clin Microbiol 42, 2271–2274. Gallimore, C. I., Lewis, D., Taylor, C., Cant, A., Gennery, A. & Gray, J. J. (2004b). Chronic excretion of a norovirus in a child with cartilage

hair hypoplasia (CHH). J Clin Virol 30, 196–204. Gallimore, C. I., Cheesbrough, J. S., Lamden, K., Bingham, C. & Gray, J. J. (2005). Multiple norovirus genotypes characterised from an

oyster-associated outbreak of gastroenteritis. Int J Food Microbiol 103, 323–330. 3357


Glass, P. J., White, L. J., Ball, J. M., Leparc-Goffart, I., Hardy, M. E. & Estes, M. K. (2000). Norwalk virus open reading frame 3 encodes a

Maunula, L. & Von Bonsdorff, C. H. (2005). Norovirus genotypes

minor structural protein. J Virol 74, 6581–6591.

causing gastroenteritis outbreaks in Finland 1998–2002. J Clin Virol 34, 186–194.

Han, M. G., Smiley, J. R., Thomas, C. & Saif, L. J. (2004). Genetic

Maynard Smith, J. (1992). Analyzing the mosaic structure of genes.

recombination between two genotypes of genogroup III bovine noroviruses (BoNVs) and capsid sequence diversity among BoNVs and Nebraska-like bovine enteric caliciviruses. J Clin Microbiol 42, 5214–5224.

Muller, H. J. (1964). The relation of recombination to mutational

Hansman, G. S., Doan, L. T., Kguyen, T. A., Okitsu, S., Katayama, K., Ogawa, S., Natori, K., Takeda, N., Kato, Y. & other authors (2004a).

Detection of norovirus and sapovirus infection among children with gastroenteritis in Ho Chi Minh City, Vietnam. Arch Virol 149, 1673–1688. Hansman, G. S., Katayama, K., Maneekarn, N., Peerakome, S., Khamrin, P., Tonusin, S., Okitsu, S., Nishio, O., Takeda, N. & Ushijima, H. (2004b). Genetic diversity of norovirus and sapovirus in

hospitalized infants with sporadic cases of acute gastroenteritis in Chiang Mai, Thailand. J Clin Microbiol 42, 1305–1307. Hansman, G. S., Takeda, N., Oka, T., Oseto, M., Hedlund, K. O. & Katayama, K. (2005). Intergenogroup recombination in sapoviruses.

Emerg Infect Dis 11, 1916–1920. Hardy, M. E. (2005). Norovirus protein structure and function. FEMS

Microbiol Lett 253, 1–8. Hardy, M. E., Kramer, S. F., Treanor, J. J. & Estes, M. K. (1997). Human

calicivirus genogroup II capsid sequence diversity revealed by analyses of the prototype Snow Mountain agent. Arch Virol 142, 1469–1479.

J Mol Evol 34, 126–129. advance. Mutat Res 106, 2–9. Nicholas, K. B., Nicholas, H. B., Jr & Deerfield, D. W., II (1997).

GeneDoc: analysis and visualization of genetic variation. EMBnet News 4, 1–4. http://www.embnet.org/files/shared/EMBnetNews/ embnet_news_4_2.pdf Nygard, K., Torven, M., Ancker, C., Knauth, S. B., Hedlund, K. O., Giesecke, J., Andersson, Y. & Svensson, L. (2003). Emerging

genotype (GGIIb) of norovirus in drinking water, Sweden. Emerg Infect Dis 9, 1548–1552. Oliver, S. L., Brown, D. W., Green, J. & Bridger, J. C. (2004). A

chimeric bovine enteric calicivirus: evidence for genomic recombination in genogroup III of the Norovirus genus of the Caliciviridae. Virology 326, 231–239. Oliver, S. L., Batten, C. A., Deng, Y., Elschner, M., Otto, P., Charpilienne, A., Clarke, I. N., Bridger, J. C. & Lambden, P. R. (2006). Genotype 1 and genotype 2 bovine noroviruses are

antigenically distinct but share a cross-reactive epitope with human noroviruses. J Clin Microbiol 44, 992–998.

Iritani, N., Seto, Y., Kubo, H., Murakami, T., Haruki, K., Ayata, M. & Ogura, H. (2003). Prevalence of Norwalk-like virus infections in cases

Phan, T. G., Kuroiwa, T., Kaneshi, K., Ueda, Y., Nakaya, S., Nishimura, S., Yamamoto, A., Sugita, K., Nishimura, T. & other authors (2006a). Changing distribution of norovirus genotypes and

of viral gastroenteritis among children in Osaka City, Japan. J Clin Microbiol 41, 1756–1759.

genetic analysis of recombinant GIIb among infants and children with diarrhea in Japan. J Med Virol 78, 971–978.

Jiang, X., Espul, C., Zhong, W. M., Cuello, H. & Matson, D. O. (1999).

Phan, T. G., Takanashi, S., Kaneshi, K., Ueda, Y., Nakaya, S., Nishimura, S., Sugita, K., Nishimura, T., Yamamoto, A. & other authors (2006b). Detection and genetic characterization of norovirus

Characterization of a novel human calicivirus that may be a naturally occurring recombinant. Arch Virol 144, 2377–2387. Katayama, K., Horikoshi-Shirato, H. & Takeda, N. (2002a).

Phylogenetic analysis of 14 strains of Norwalk-like viruses: identification of the region in the genome for genotyping. Nippon Rinsho 60, 1165–1174 (in Japanese). Katayama, K., Shirato-Horikoshi, H., Kojima, S., Kageyama, T., Oka, T., Hoshino, F., Fukushi, S., Shinohara, M., Uchida, K. & other authors (2002b). Phylogenetic analysis of the complete genome of 18 Norwalk-

like viruses. Virology 299, 225–239. Kim, M. J. & Kao, C. (2001). Factors regulating template switch in vitro

by viral RNA-dependent RNA polymerases: implications for RNARNA recombination. Proc Natl Acad Sci U S A 98, 4972–4977. MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinform 5, 150–163.

Kumar, S., Tamura, K. & Nei, M. (2004).

Lai, M. M. (1992). RNA recombination in animal and plant viruses.

Microbiol Rev 56, 61–79. Lindell, A. T., Grillner, L., Svensson, L. & Wirgart, B. Z. (2005).

Molecular epidemiology of norovirus infections in Stockholm, Sweden, during the years 2000 to 2003: association of the GGIIb genetic cluster with infection in children. J Clin Microbiol 43, 1086–1092. Lole, K. S., Bollinger, R. C., Paranjape, R. S., Gadkari, D., Kulkarni, S. S., Novak, N. G., Ingersoll, R., Sheppard, H. W. & Ray, S. C. (1999).

strains circulating among infants and children with acute gastroenteritis in Japan during 2004–2005. Clin Lab 52, 519–525. Phan, T. G., Yagyu, F., Kozlov, V., Kozlov, A., Okitsu, S., Muller, W. E. & Ushijima, H. (2006c). Viral gastroenteritis and genetic character-

ization of recombinant norovirus circulating in Eastern Russia. Clin Lab 52, 247–253. Phan, T. G., Yan, H., Li, Y., Okitsu, S., Muller, W. E. & Ushijima, H. (2006d). Novel recombinant norovirus in China. Emerg Infect Dis 12,

857–858. Phan, T. G., Kaneshi, K., Ueda, Y., Nakaya, S., Nishimura, S., Yamamoto, A., Sugita, K., Takanashi, S., Okitsu, S. & Ushijima, H. (2007). Genetic heterogeneity, evolution and recombination in

noroviruses. J Med Virol 79, 1388–1400. Pletneva, M. A., Sosnovtsev, S. V. & Green, K. Y. (2001). The genome

of Hawaii virus and its relationship with other members of the Caliciviridae. Virus Genes 23, 5–16. Posada, D. (2002). Evaluation of methods for detecting recombina-

tion from DNA sequences: empirical data. Mol Biol Evol 19, 708–717. Reuter, G., Krisztalovics, K., Vennema, H., Koopmans, M. & Szucs, G. (2005). Evidence of the etiological predominance of norovirus in

gastroenteritis outbreaks – emerging new-variant and recombinant noroviruses in Hungary. J Med Virol 76, 598–607.

Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence of intersubtype recombination. J Virol 73, 152–160.

Reuter, G., Vennema, H., Koopmans, M. & Szucs, G. (2006).

Martinez, N., Espul, C., Cuello, H., Zhong, W., Jiang, X., Matson, D. O. & Berke, T. (2002). Sequence diversity of human caliciviruses

Rohayem, J., Munch, J. & Rethwilm, A. (2005). Evidence of

recovered from children with diarrhea in Mendoza, Argentina, 1995–1998. J Med Virol 67, 289–298.

Sasaki, Y., Kai, A., Hayashi, Y., Shinkai, T., Noguchi, Y., Hasegawa, M., Sadamasu, K., Mori, K., Tabei, Y. & other authors (2006). Multiple

3358

Epidemic spread of recombinant noroviruses with four capsid types in Hungary. J Clin Virol 35, 84–88. recombination in the norovirus capsid gene. J Virol 79, 4977–4990.


Norovirus recombination viral infections and genomic divergence among noroviruses during an outbreak of acute gastroenteritis. J Clin Microbiol 44, 790–797. Seah, E. L., Gunesekere, I. C., Marshall, J. A. & Wright, P. J. (1999).

Variation in ORF3 of genogroup 2 Norwalk-like viruses. Arch Virol 144, 1007–1014. Tajima, F. & Nei, M. (1984). Estimation of evolutionary distance

Vinje, J., Hamidjaja, R. A. & Sobsey, M. D. (2004). Development and

application of a capsid VP1 (region D) based reverse transcription PCR assay for genotyping of genogroup I and II noroviruses. J Virol Methods 116, 109–117. Wang, Q. H., Han, M. G., Cheetham, S., Souza, M., Funk, J. A. & Saif, L. J. (2005). Porcine noroviruses related to human noroviruses. Emerg

between nucleotide sequences. Mol Biol Evol 1, 269–285.

Infect Dis 11, 1874–1881.

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994).

CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.

Waters, A., Coughlan, S. & Hall, W. W. (2007). Characterisation of a

Tsugawa, T., Numata-Kinoshita, K., Honma, S., Nakata, S., Tatsumi, M., Sakai, Y., Natori, K., Takeda, N., Kobayashi, S. & Tsutsumi, H. (2006).

Widdowson, M. A., Cramer, E. H., Hadley, L., Bresee, J. S., Beard, R. S., Bulens, S. N., Charles, M., Chege, W., Isakbaeva, E. & other authors (2004). Outbreaks of acute gastroenteritis on cruise ships and

Virological, serological and clinical features of an outbreak of acute gastroenteritis due to recombinant genogroup II norovirus in an infant home. J Clin Microbiol 44, 177–182.

Widdowson, M. A., Monroe, S. S. & Glass, R. I. (2005a). Are

Tu, E. T.-V., Nguyen, T., Lee, P., Bull, R. A., Musto, J., Hansman, G., White, P. A., Rawlinson, W. D. & McIver, C. J. (2007). Norovirus

novel recombination event in the norovirus polymerase gene. Virology 363, 11–14.

on land: identification of a predominant circulating strain of norovirus – United States, 2002. J Infect Dis 190, 27–36. noroviruses emerging? Emerg Infect Dis 11, 735–737.

GII.4 strains and outbreaks, Australia. Emerg Infect Dis 13, 1128–1129.

Widdowson, M. A., Rockx, B., Schepp, R., van der Poel, W. H., Vinje, J., van Duynhoven, Y. T. & Koopmans, M. P. (2005b). Detection of serum

van den Berg, H., Lodder, W., van der Poel, W., Vennema, H. & de Roda Husman, A. M. (2005). Genetic diversity of noroviruses in raw

antibodies to bovine norovirus in veterinarians and the general population in the Netherlands. J Med Virol 76, 119–128.

and treated sewage water. Res Microbiol 156, 532–540.

Worobey, M. & Holmes, E. C. (1999). Evolutionary aspects of

Vidal, R., Roessler, P., Solari, V., Vollaire, J., Jiang, X., Matson, D. O., Mamani, N., Prado, V. & O’Ryan, M. L. (2006). Novel recombinant

norovirus causing outbreaks of gastroenteritis in Santiago, Chile. J Clin Microbiol 44, 2271–2275.

http://vir.sgmjournals.org

recombination in RNA viruses. J Gen Virol 80, 2535–2543. Zheng, D. P., Ando, T., Fankhauser, R. L., Beard, R. S., Glass, R. I. & Monroe, S. S. (2006). Norovirus classification and proposed strain

nomenclature. Virology 346, 312–323.

3359