Journal of General Virology (2014), 95, 1469–1478
DOI 10.1099/vir.0.064675-0
Relationship between genotypes and serotypes of genogroup 1 recoviruses: a model for human norovirus antigenic diversity Tibor Farkas,1,2 Cindy Wong Ping Lun13 and Brittney Fey14 Correspondence Tibor Farkas [email protected]
Received 13 February 2014 Accepted 31 March 2014
1
Cincinnati Children’s Hospital Medical Center, OH, USA
2
University of Cincinnati College of Medicine, Cincinnati, OH, USA
Human norovirus (NoV) research greatly relies on cell culture-propagable surrogate caliciviruses, including murine NoVs and the prototype ‘recovirus’ (ReCV), Tulane virus. However, the extreme biological diversity of human NoVs cannot be modelled by a uniform group of viruses or single isolate. Based on a diverse group of recently described ReCVs, a more advanced model reflecting human NoV biological diversity is currently under development. Here, we have reported the genotypic and serotypic relationships among 10 G1 ReCV isolates, including Tulane virus and nine other recent cell culture-adapted strains. Based on the amino acid sequences of virus capsid protein, VP1, and classification constraints established for NoVs, G1 ReCVs were separated into three genotypes, with variable organization of the three open reading frames. Interestingly, crossneutralization plaque assays revealed the existence of four distinct serotypes, two of which were detected among the G1.2 strains. The amino acid (aa) difference between the two G1.2 ReCV serotypes (12%) was less than the minimum 13 % difference established between NoV genotypes. Interestingly, one of the G1.3 ReCVs was equally neutralized by antisera raised against the G1.3 (6 % aa difference) and G1.1 (25 % aa difference) representative strains. These results imply the existence of a large number of human NoV serotypes, but also shared crossneutralization epitopes between some strains of different genotypes. In conclusion, the newly developed ReCV surrogate model can be applied to address biologically relevant questions pertaining to enteric CV diversity.
INTRODUCTION Caliciviruses (CV) are non-enveloped, positive-sense, single-stranded RNA viruses with polyadenylated genomes of ~6.5–8.5 kb. The Caliciviridae family consists of five established genera: Norovirus, Sapovirus, Lagovirus, Vesivirus and Nebovirus. Three recently proposed CV genera, ‘Recovirus’, ‘Valovirus’ and ‘chicken calicivirus’, are awaiting ICTV approval (Green, 2013). Depending on the genus, CV genomes are organized into two or more ORFs. Noroviruses (NoVs) and vesiviruses encode nonstructural (NS) and capsid (VP1) proteins in separate ORFs (ORF1 and ORF2, respectively), while sapoviruses and lagoviruses encode both NS and capsid proteins in a continuous ORF1. A separate ORF encoding a minor structural protein (VP2), designated ORF3 in the noro- and vesivirus genomes and ORF2 in the sapo- and 3Present address: University of Cincinnati, Cincinnati, OH, USA. 4Present address: Northern Kentucky University, Highland Heights, KY, USA. The GenBank/EMBL/DDBJ accession number for the sequences determined in this study are KC662363–KC662370 and KF431831.
064675 G 2014 The Authors
Printed in Great Britain
lagovirus genomes, is present in all CVs (Green, 2013). In murine NoV genomes, an additional ORF (ORF4) overlapping ORF2 encodes a protein (VF1) that may play a role in virulence (McFadden et al., 2011). CVs cause a variety of diseases in animals. In humans, NoVs and sapoviruses, commonly referred to as ‘human CVs’, are two of the leading causes of acute gastroenteritis. In the United States alone, NoVs cause an estimated 23 million cases of acute gastroenteritis, leading to .70 000 hospitalizations and 800 deaths each year (Lopman et al., 2011; Scallan et al., 2011; Hall et al., 2012). The prototype NoV (Norwalk virus) was discovered more than 40 years ago (Kapikian et al., 1972). However, due to the lack of an efficient cell culture system and robust animal model, human NoV research still relies on surrogate CVs and/or hosts (e.g. murine NoV or gnotobiotic pig/calf) that do not necessarily reflect the essential biological features and diversity of human NoVs and their hosts (Table 1). The chimpanzee model is a promising alternative for the preclinical evaluation of NoV vaccines and antivirals (Bok et al., 2011). However, the use of chimpanzees for biomedical research is restricted in most countries worldwide, including 1469
T. Farkas, C. Wong Ping Lun and B. Fey
Table 1. Comparison of human NoVs and surrogate caliciviruses, including ReCVs HuNoV Genus Genome organization Diversity HBGA binding Efficient cell culture Efficient reverse genetics Transmission Diarrhoea in immunocompetent host Shedding in stool Zoonotic transmission Availability from ATCC
Norovirus 3 ORFs High Yes No No Faecal–Oral Yes Yes Suggested No
MuNoV
PoNoV
BoNoV
PECV
Norovirus Norovirus Norovirus Sapovirus 4 ORFs 3 ORFs 3 ORFs 2 ORFs Low Low Low Low No No No No Yes No No Yes Yes No No Yes Faecal–Oral Faecal–Oral Faecal–Oral Faecal–Oral No No Yes Yes Yes No Yes
Yes Suggested No
Yes Suggested No
Yes No No
FCV Vesivirus 3 ORFs Low No Yes Yes Respiratory No No* No Yes
RHDV
ReCV
Lagovirus Recovirus 2 ORFs 3 ORFs Low High Yes Yes No Yes Yes Yes Faecal–Oral Faecal–Oral No Yes Yes No No
Yes Suggested No
HuNoV: human norovirus; MuNoV: Murine norovirus; PoNoV: porcine norovirus; BoNoV: bovine norovirus; PECV: porcine enteric calicivirus; FCV: feline calicivirus; RHDV: rabbit haemorrhagic disease virus; ReCV: rhesus enteric calicivirus. *FCV primarily sheds in oral and oculonasal discharge, but can be found in all body secretions, including faeces, during acute disease.
the USA. At present, licensed vaccines or antiviral strategies to prevent NoV disease are not available. One of the major challenges in NoV vaccine design is the tremendous diversity of human NoVs (Green, 2013). According to phylogenetic differences of the capsid protein, VP1, NoVs are classified into five genogroups (G1–G5), with several genetic types in each genogroup (e.g. G1.1–G1.8). Based on analysis of 164 human and animal NoV VP1 proteins, amino acid (aa) sequence differences established between strains (0–14.07 %) within a genetic type, between genotypes (14.26–43.78 %) and between genogroups (44.91–64.41 %) are used for the genetic classification of new isolates (Zheng et al., 2006; Kroneman et al., 2013). Genogroups G1, 2 and 4 mainly contain human NoVs, while G3 includes bovine and G5 murine NoVs. Over 30 human NoV genetic types have been described to date (Green, 2013). Antigenic characterization of human NoVs mainly relies on virus-like particles (VLP) and antibody binding assays (ELISA). Several studies have demonstrated the existence of cross-reactive conserved epitopes among the different NoV genotypes and even genogroups, predominantly located in the S-domain or Cterminal region of the P1 domain (Parker et al., 2005; Oliver et al., 2006; Parra et al., 2013). The role of histo-blood group antigens (HBGA) in human NoV binding/attachment and susceptibility to infection, and diversity of the HBGA binding properties of different human NoV strains are well established (Lindesmith et al., 2003; Huang et al., 2005; Frenck et al., 2012). HBGA binding sites have been mapped to the most variable P2 subdomain of the NoV capsid (Choi et al., 2008; Shanker et al., 2011). The surface-exposed P2 subdomain probably also contains the epitopes responsible for major antigenic differences (serotypes). Antibodies that block adherence of human NoV VLPs to HBGAs in binding assays (ELISA) have been proposed as virus-neutralizing (VN) antibodies. Recent results based on blocking assays performed mainly 1470
with G2.4 human NoVs suggested that antigenic changes facilitating escape from herd immunity may also drive changes in HBGA binding affinities and lead to altered population susceptibility (Debbink et al., 2012; Lindesmith et al., 2012). However, due to the lack of a human NoV cell culture system, evaluating the roles of HBGAs as attachment ligands or receptors and direct assessment of human NoV antibody neutralization, including the identification of neutralizing epitopes and relationship between genotypes and serotypes, remains a significant challenge. Murine NoVs are closely related to their human counterparts, replicate well in cell culture and represent a costeffective small animal model (Wobus et al., 2006). However, the murine NoV model also has limitations highlighted by significant differences compared with human NoVs. Murine NoVs exhibit limited genetic and antigenic diversity, have apparent immune cell tropism and cause persistent asymptomatic infection in normal mice. The prototype ‘recovirus’ (ReCV), Tulane virus, was discovered in stool samples of juvenile rhesus macaques (Farkas et al., 2008). This initial discovery was followed by the identification of several other ReCVs, including one strain in human stool samples (Farkas et al., 2010a; Handley et al., 2012; Smits et al., 2012). Since then, several additional ReCV strains described in our previous reports (Farkas et al., 2010a) have been cell culture-adapted. Studies on Tulane virus and other isolates established that ReCVs are evolutionarily and biologically closely related to human NoVs, including genetic, antigenic and HBGA binding diversities and clinical symptoms of gastroenteritis observed in experimentally infected macaques (Farkas et al., 2008, 2010 a, b; Sestak et al., 2012). Based on the biologically diverse ReCV isolates, we are currently developing a cell culture and animal model that can effectively replicate the biological features of human NoVs. Here, we examined the relationship between Journal of General Virology 95
Relationship between ReCV geno- and serotypes
genotypes and serotypes within G1 ReCVs as part of this ongoing project.
GTCATAGACATG), 3 nt in the G1.2 ReCV genomes (TGAGCTATG) and 2 nt in the G1.3 ReCVs (TGATCATG). ORF2 start and ORF3 stop codons of all NoVs and ReCVs overlapped by 1 nt (TAATG).
RESULTS G1 ReCV serotypes
Phylogeny and genome organization Phylogenetic analysis of complete ORF2 (VP1) sequences of the three ReCVs available in public databases (Tulane virus, WUHARV and Bangladesh/289/2007) and nine tissue culture-adapted ReCVs obtained in this study confirmed our earlier classification based on short RdRp sequences (Farkas et al., 2010a). Genogroup 1 ReCVs were clearly segregated into three genotypes with evolutionary distances comparable to those between the human NoV genotypes. The amino acid differences between strains of the three genotypes ranged from 24 to 30 % (Table 2). In addition to G2 rhesus isolates, the human ReCV isolate, Bangladesh/ 289, represents a third genogroup (G3) (Fig. 1). Analysis of the ReCV genomes revealed three ORFs with significant variations in arrangement. When ORF1 was adjusted as frame 1, all G1.1 and G1.2 ReCVs exhibited an order of frames 1, 1 and 3, while the order for G1.3 ReCVs was frames 1, 3 and 2 for ORF1, ORF2 and ORF3, respectively. Human NoVs, murine NoVs and ReCV isolate Bangladesh/289 exhibited an order of frames 1, 2 and 1, with an additional ORF4 in-frame 3 in murine NoV genomes (Fig. 2). The Norwalk virus ORF1 stop and ORF 2 start codons overlapped by 11 nt (ATGATGATGGCGTCTAA), the murine NoV by 8 nt (ATGAGGATGAGTGA) and ReCV Bangladesh/289 by 146 or 113 nt, depending on which of the two possible start codons was utilized (ATGAGCTCCCAAAAGGAAAATCTGTCAAACAAAATGGAGGAAGTGGTGCAACCAGCAACTGGAGGGGTTGGCGAAGTCATAACCGGAGCAGCCGCCCCACTACCAGAGAATCCAAGAGGACTGGAATTGGCCCCAACAATCAATCCCATTGA). The ORF1 stop and ORF 2 start codons were separated by 21 nt in the G1.1 ReCV genomes (TGATTGATCACAATT-
Hyperimmune sera against the Tulane virus (G1.1), FT285 (G1.2) and FT7 (G1.3) ReCVs exhibited low-level crossreactivity. The difference in titres between the homologous and heterologous antisera ranged between 102- and 2560fold, and was greater for all strains than the 20-fold cutoff used to define serotypes (Table 3). Thus, the three G1 strains used for generating hyperimmune sera represent not only three genotypes but also three serotypes. Serotypic evaluation of the 10 G1 ReCV isolates was performed in a standardized cross-neutralization plaque assay with 20 virus neutralization units (VNU) of serotypespecific hyperimmune sera. Anti-Tulane virus hyperimmune sera neutralized all G1.1 isolates and FT65, a G1.3 strain, but not the other two G1.3 or the G1.2 strains. Surprisingly, anti-FT285 (G1.2) hyperimmune sera only neutralized FT285, but not the two other G1.2 isolates (FT157 and FT499). Anti-FT7 (G1.3) hyperimmune sera neutralized only the G1.3 strains (Fig. 3). Based on these results, the 10 G1 ReCVs were grouped into four serotypes, specifically, serotype 1 (G1.1), serotype 2 (FT285, G1.2), serotype 3 (FT7 and FT400, G1.3) and serotype 4 (FT157 and FT499, G1.2). The exact serotype of FT65 (G1.3) ReCV could not be determined with the antibodies used in this study. Virus neutralization epitope prediction Several discontinuous B cell epitopes were predicted in almost all of the 10 ReCV strains in this study. These included PPT in the hinge region (aa positions 206–208 in the FT285 VP1 sequence), AITDKP-RR in loop A’-B’ of the P2 domain (aa positions 283–288) and M-VSG (aa
Table 2. VP1 pairwise amino acid difference and homology values (%) of ReCVs TV
FT205
FT218
FT494
FT285
FT499
WUHARV
FT7
FT65
FT400
Bangladesh
– 9 9 9 26 26 25 25 25 25 62
91 – 3 3 25 25 25 26 26 25 62
91 97 – 2 25 24 24 24 25 24 61
91 97 98 – 26 25 25 25 26 25 61
74 75 75 74 – 12 10 29 30 29 60
74 75 76 75 88 – 8 29 29 29 60
75 75 76 75 90 92 – 29 29 29 60
75 74 76 75 71 71 71 – 6 3 60
75 74 75 74 70 71 71 94 – 5 59
75 75 76 75 71 71 71 97 95 – 61
38 38 39 39 40 40 40 40 41 39 –
G1.1-TV G1.1-FT205 G1.1-FT218 G1.1-FT494 G1.2-FT285 G1.2-FT499 G1.2-WUHARV G1.3-FT7 G1.3-FT65 G1.3-FT400 G3.1-Bangladesh
Pairwise amino acid difference and homology scores are shown in the lower and upper diagonal half of the table, respectively. http://vir.sgmjournals.org
1471
T. Farkas, C. Wong Ping Lun and B. Fey
(a)
87 99 99 98 99 96 99 99
99 93
‘Recovirus’ 99
47 99
Norovirus
99
85
99
60 61
0.5
0.4
0.3
98
99 83 0.2
0.1
FT221 FT412 FT218* FT216 FT363 FT226 FT236 FT205* G1.1 FT457 FT494* FT472 FT422 FT473 FT234 FT187 Tulane virus* WUHARV FT285* FT293 FT499* FT213 FT176 G1.2 FT15 FT157* FT150 FT138 FT135 FT143 FT172 FT413 FT411 FT80 FT65* FT191 FT405 FT269 FT219 FT294 FT305 FT299 FT350 G1.3 FT7* FT242 FT30 FT301 FT394 FT390 FT416 FT454 FT400* FT177 FT36 FT41 ReCVBangladesh G3.1 FT286 FT310 FT443 G2.1 FT349 FT399 FT437 G5 MNV (AY228235) G2.1 Hawaii (U07611) G2.2 SMV (AY134748) G2.3 SU201 (AB039782) G4.1 (JQ613567) G3.1 BoNoVJena (AJ011099) G3.2 BoNoVNewbury (AF097917) G1.3 DSV (U04469) G1.1 Norwalk (M87661) G1.2 Southampton (L07418)
(b)
70 100 100 52 100 82
100
‘Recovirus’
100 79 100
100
100 98
Norovirus
100 100
100
99 99
100 89 0.6
0.5
0.4
0.3
0.2
0.1
FT218 FT494 FT205 Tulane virus (EU391643) FT65* FT400 FT7 FT285 WUHARV (JX627575) FT157 FT499 ReCV Bangladesh (JQ745645) G5.1 MNV (AY228235) G5.2 TF5WM (JQ408737) G2.1 Hawaii (U07611) G2.3 SMV (AY134748) G2.3 SU201 (AB039782) G4.1 (JQ613567) G3.1 Jena (AJ011099) G3.2 Newbury (AF097917) G1.3 DSV (U04469) G1.1 Norwalk (M87661) G1.2 Southampton (L07418)
G1.1
G1.3
G1.2 G3.1
0.0
0.0
Fig. 1. (a) Genetic classification of ReCVs based on partial (268 nt) RdRp sequences (Farkas et al., 2010a). ReCV strains used in Fig. 1(b) are in bold type. ReCVs used in tissue culture-based assays are marked with an asterisk. (b) Genetic classification of ReCVs based on complete VP1 amino acid sequences. Cell culture-adapted ReCV isolates described in this study are in bold type. GenBank accession numbers are KC662363–KC662370 and KF431831. The scale bars indicate the evolutionary distance represented by nucleotide or amino acid substitutions per site, respectively. The numbers at the nodes indicate the percentage of times that a node was supported during 1000 bootstrap analyses.
positions 385–389). The hinge epitope was predicted in all strains except TV, loop A’-B’ epitope in all strains except FT205 and FT494, and M-VSG epitope in all strains except TV, FT7 and FT400. Analysis of multiple sequence alignments highlighted the loop A’-B’ epitope (Fig. 4) as the most likely candidate responsible for the VN differences among the strains (Fig. 3).
DISCUSSION Since the discovery of the prototype Tulane virus (Farkas et al., 2008), a number of similarities have been identified between ReCVs and human NoVs, including close evolutionary relatedness, genetic diversity, HBGA binding properties, epidemiology and disease spectrum (Farkas 1472
et al., 2010a, b, 2012; Sestak et al., 2012). Human NoVs cannot be grown in cell culture. For this reason, other CVs are used as surrogates for human NoV research (Vashist et al., 2009; Richards, 2012). Cell culture-propagable murine NoVs have emerged as the most commonly utilized human NoV surrogates (Wobus et al., 2006). However, there remains a need for more robust surrogates, as evident from the ongoing NoroCore projects that form part of the USDANIFA Food Virology Initiative (http://norocore.ncsu.edu). These efforts have yielded several recently published studies with the prototype Tulane virus (DiCaprio et al., 2012; Hirneisen & Kniel 2013a, b; Li et al., 2013). One of the major challenges in human NoV research, particularly for vaccine development, is the enormous diversity (genetic, antigenic, HBGA binding) of human NoVs. These features cannot be Journal of General Virology 95
Relationship between ReCV geno- and serotypes
5364
NORWALK VIRUS (All human noroviruses) p48
NTPase
6947 VP1
p22 VPg
Prot
Pol
69 nt (A)n
VP2 5371
5
6950
7585
MNV p38
NTPase
p12 VPg
Prot
Pol 5066
6
76 nt (A)n
VP2
5069 5707
7304
6681
VF VP1
5062
6647 TULANE VIRUS (G1.1 and G1.2 Recoviruses: FT205, FT218, FT494. FT157, FT285, FT499 and WUHARV) p26
NTPase
p16 VPg
Prot
Pol
VP1 4355 4380
15
5981 VP2 5984
FT7 (G1.3 Recoviruses: FT65, FT400)
1
2343 730
Prot
2340
2735
4384
2969
VP1 VP2
Pol 2883
1
73 nt (A)n
VP1
736 RECOVIRUS/BANGLADESH/289 p16 VPg
6637
VP2 Pol
74 nt (A)n
4387
73 nt (A)n 4992
Fig. 2. Genome organization of ReCVs and Noroviruses. ORFs are shown as boxes. ORF1 encodes non-structural proteins, ORF2 the capsid protein (VP1), and ORF3 the minor structural protein (VP2). In the murine NoV genome, a fourth ORF that overlaps with ORF2 has been identified that encodes virulence factor (VF1) protein (McFadden et al., 2011).
duplicated by a uniform group of viruses or single isolate. Based on the availability of a large collection of diverse ReCVs in our laboratory, we recently initiated the development of a cell culture and non-human primate model able
to duplicate the diverse biological features of human NoVs and their hosts. In the current study, we evaluated the relationship between the genotypes and serotypes of an enteric CV for the first time using this unique model.
Table 3. Cross-reactivity of anti-ReCV hyperimmune mouse sera
Partial RdRp, whole capsid protein (VP1) and whole minor structural protein (VP2)-encoding sequences were obtained for nine cell culture-adapted G1 ReCVs. Phylogenetic analysis of ORF2 nt and VP1 aa sequences confirmed our previous classification based on short (268 nt) RdRp sequences (Farkas et al., 2010a) and the existence of three genotypes among G1 ReCVs. Amino acid differences among the ReCV genotypes (24–30 %) were within the range (14.26–43.78 %) established for human NoV genotypes (Zheng et al., 2006). ReCV WUHARV was grouped with the G1.3 strains, while the human ReCV isolate, Bangladesh/289, was clearly separated from rhesus isolates with 59–62 % aa differences, equivalent to those reported between the human NoV genogroups (44.91–64.41 %) (Table 2) (Zheng et al., 2006). Genogroup 2 ReCV VP1 sequences have not been published yet. However, based on analyses of published VP1 and/or RdRp sequences, ReCVs
Reciprocal of VN titres
TV FT285 FT7 Homologous/ heterologous*
Anti-TV
Anti-FT285
Anti-FT7
5120 10 50 102-fold
10 25 600 ,10 2560-fold
40 40 12 800 320-fold
VN end titres were calculated based on repeated experiments with twofold dilutions, starting from 1 : 10 and 1 : 50. *The least difference of titres of an antiserum against homologous and heterologous strains. Twenty-fold or greater difference in both directions was used to distinguish serotypes. http://vir.sgmjournals.org
1473
T. Farkas, C. Wong Ping Lun and B. Fey
S2
(a) G1/1/S 1 FT205
S3
S 1+3
G1/2
FT218
FT494
FT285
S3
G1/3
FT157
FT499
FT7
FT65
FT400
a-FT7
a-FT285
a-Tulane
Mouse sera (20 VNU)
Control
Tulane
S4
(b) 100
100
80
80
80
60
60
60
40
40
40
* *
*
*
20 *
*
0 T FT V 20 FT 5 2 FT 18 4 FT 94 28 FT 5 15 FT 7 49 9 FT 7 FT 40 FT 0 65
0
Anti-TV serum
*
*
*
*
*
*
20 *
*
* 0
Anti-FT285 serum
* *
*
*
*
*
T FT V 20 FT 5 2 FT 18 4 FT 94 28 FT 5 49 FT 9 15 7 FT FT 7 40 FT 0 65
20
T FT V 20 FT 5 2 FT 18 49 FT 4 28 FT 5 15 FT 7 49 9 FT FT 7 40 FT 0 65
Plaque reduction (%)
100
Anti-FT7 serum
Fig. 3. Serotyping of G1 ReCVs. (a) Serotyping plaque assay. Twenty virus neutralization units (VNU) of each hyperimmune mouse serum were used to cross-neutralize 50–100 p.f.u. of ReCV. (b) Mean plaque reduction (%) of repeat experiments. Error bars represent standard deviation. (*) Shows statistically significant difference (P,0.05), compared with neutralization by homologous virus/serum. ($) Indicates neutralization by genotype-matched sera. (#) Indicates lack of neutralization by genotype-matched sera. (e) Indicates neutralization by sera raised against a heterologous genotype. Serotypes were established based on ¢80 % plaque reduction.
were classified into three genogroups (G1–G3), with three genotypes within G1 (G1.1–G1.3) (Fig. 1). Considering that most known ReCV sequences originated from samples collected from the same colony of captive macaques between April and July of 2008 (Farkas et al., 2010a), the observed genetic diversity indicates highly diverse, human NoV-like evolution of ReCVs. Analysis of the ReCV genomes revealed variable arrangement of the three ORFs among the different ReCV genotypes that was not observed among NoVs (Fig. 2). Further studies using different geographical locations and primate species, including humans, are necessary to determine the full extent of ReCV diversity. To standardize for cell culture-based serotyping assays, only ReCV strains isolated in the LLC-MK2 cell line were 1474
involved in this study. Mouse hyperimmune sera generated against one representative strain of each of the three G1 genotypes were used for serotyping. End titration with the three representative strains used for immunization based on requirements established for picornavirus and rotavirus serotyping (Committee on Enteroviruses, 1962; Kapikian et al., 1967; Wyatt et al., 1982) revealed that these strains represent three different serotypes. The cross-reactivity between Tulane virus (serotype 1) and FT7 (serotype 3) was stronger than between FT285 (serotype 2) or other serotypes. However, the difference in titres in both directions was still at least fourfold higher between all strains than the 20-fold minimum cut-off value established for separating serotypes (Table 3). Journal of General Virology 95
Relationship between ReCV geno- and serotypes
(a)
Tulane FT205 FT218 FT494 FT285 FT157 FT499 FT7 FT65 FT400
(b) C’-D’ Norwalk
A’-B’ E’-F’
Tulane
FT285
FT7
Hinge
(c)
MD145
V-SHD
FT285
AITDKP-R
Due to the limited volume of mouse hyperimmune sera, antigenic relationships between the 10 ReCV isolates were evaluated in a standardized cross-neutralization plaque assay. To adjust for differences in VN antibody levels in individual hyperimmune serum samples, the test was performed with a single dilution of each serum (representing 20 VNUs) against the homologous strain (Cowen & Hitchner, 1975; Smith et al., 1998). The cross-neutralization assay highlighted statistically significant differences among the ReCV isolates. All the G1.1 isolates belonged to serotype 1. The anti-FT285 (serotype 2) serum was not able to neutralize the other two G1.2 strains (FT157 and FT499), leading to separation of the G1.2 isolates into two serotypes (serotypes 2 and 4). All G1.3 strains were neutralized by anti-FT7 (serotype 3) serum, and consequently grouped as serotype 3. One of the G1.3 isolates (FT65) was also neutralized by serotype 1-specific serum (Fig. 3). This finding is of significant interest, since FT65 and FT7 differ only by 6 %, while FT65 and Tulane virus differ by 25 % in terms of VP1 aa positions (Table 2). Heterologous sera exhibited significantly lower neutralization for all strains, except FT65, (P,0.05). To our knowledge, this is the first study to correlate genetic and serotypic differences for an enteric CV and link the genetic classification scheme to a biologically relevant http://vir.sgmjournals.org
Fig. 4. Predicted ReCV neutralization B cell epitope. (a) Multiple sequence alignment. Amino acid positions according to Tulane virus VP1 from 275 to 293 aa. Linear B-cell epitope predicted by BepiPred 1.0 in this region is underlined. Discontinuous B-cell epitope predicted by the DiscoTope 2.0 server in this region is presented in bold type. (b) Structural comparison of Norwalk virus and ReCVs. Three-dimensional structures of capsid proteins were generated using the Phyre 2 server. The S-domain is in gold, P1 domain in green and P2 domain in red. Surface-exposed loops in the P2 domain are labelled as A’-B’, C’-D’ and E’-F’. The predicted neutralization epitope located in loop A’-B’ of the P2 domain is in yellow. Amino acids in the Norwalk virus structure interacting with H-type pentasaccharide (D327, H329, Q342, D344, W375, S377 and P378) are in blue (Choi et al., 2008). (c) Structural overlap of the predicted ReCV neutralization epitope and highly variable key blockade epitope (epitope A; V-SHD) of G2.4 human NoVs (Lindesmith et al., 2013). Three-dimensional structures of MD145 (G2.4 human NoV) and FT285 were generated using the Phyre 2 server. Amino acids comprising the epitopes are depicted in yellow.
measure. In general, a strong correlation between ReCV genotypes and serotypes was observed. In an earlier study by Zheng et al. (2006), the minimum aa difference detected between NoV genotypes was 14.26 %, while in a more recent study including newly described NoVs, a difference of 13 % was reported (Kroneman et al., 2013). The smallest aa difference observed in this study between ReCV serotypes (FT285 and the other two G1.2 strains) was 12 %. On the other hand, we identified cross-neutralization between strains with a 25 % aa difference. FT157 or FT499 (serotype 4) and FT65-specific hyperimmune sera were not generated in this study. Follow-up studies that include mAbs and more ReCV isolates and sera are critical. While we were unable to evaluate whether protection against infection is serotype-specific in the rhesus macaque model, the implication of our findings for human NoV research, especially vaccine development, is significant. Further clarification of the mechanism of ReCV antibody neutralization is important for human NoV research. Meanwhile, since the VP1 sequences and cross-VN results disclose valuable information, we performed computational B cell epitope predictions on modelled 3D structures of ReCVs. A candidate epitope on loop A’-B’ of the P2 domain was identified (Fig. 4) that was also predicted at a similar location on several human NoV structures (data not shown). A literature search 1475
T. Farkas, C. Wong Ping Lun and B. Fey
revealed that an equivalent region on loop A’-B’ of the murine NoV P2 domain has been identified as one of the contact sites for a neutralizing mAb (A6.2) (Katpally et al., 2008). This epitope structurally overlaps with the highly variable key blockade epitope A that is reported to change with new G2.4 human NoV emergence (Lindesmith et al., 2013) (Fig. 4c). In summary, our findings clearly demonstrate the value of the ReCV model as a surrogate for analysing the diverse biological features of human noroviruses, and pave the way for detailed evaluation of enteric CV antibody neutralization and escape as well as the relationship with HBGA binding through cell culture-based, structural and in vivo studies.
were submitted to the GenBank database under accession numbers KC662363–KC662370 and KF431831. BLAST analyses were run against NCBI databases. Multiple sequence alignments of nt and aa sequences were created using Omiga v2.0 software (Oxford Molecular). Phylogenetic trees were constructed using the unweighted pair group method with arithmetic mean (UPGMA) and the neighbour-joining clustering methods of the Molecular Evolutionary Genetics Analysis (MEGA version 5.1) software with Jukes-Cantor and Poisson correction distance calculations for nt and aa sequence alignments, respectively (Tamura et al., 2011). The confidence values of the internal nodes were obtained by performing 1000 bootstrap analyses. Accession numbers of the TV, WUHARV, Recovirus Bangladesh/289 and human NoVs included in the alignments are listed in Fig. 2a.
Sequence and phylogenetic analyses.
Cross-titration of mouse hyperimmune sera. A cytopathic effect
METHODS Viruses and cell lines. Genogroup I ReCVs were isolated from rhesus stool samples collected during our previous studies (Farkas et al., 2008, 2010a). All strains were isolated in LLC-MK2 cell lines (ATCC CCL-7). Virus stocks were prepared from viruses plaquepurified three times and stored at 280 uC in aliquots. Anti-ReCV mouse hyperimmune sera. Mice were maintained at
the animal care facility of Cincinnati Children’s Hospital Medical Center (CCHMC), and all animal procedures complied with National Institutes of Health guidelines. Our animal protocol was approved by the Institutional Animal Care and Use Committee of CCHMC (Animal Welfare Assurance Number A3108-01). Hyperimmune sera were generated against one representative strain of each genotype of G1 ReCVs, including the prototype TV (G1.1), FT285 (G1.2) and FT7 (G1.3). Three female 6- to 8-week-old BALB/c mice were injected intraperitoneally with 106–108 TCID50 of CsCl density gradientpurified ReCVs mixed with an equal volume of Sigma Adjuvant (Sigma Aldrich). Animals were boosted twice at 3-week intervals and anaesthetized/sacrificed via carbon dioxide (CO2) inhalation on day 14 after the last immunization. Serum was collected and stored in aliquots at 280 uC. RNA extraction and genome amplification. Viral RNA was
extracted with TRIzol LS (Life Technologies) according to the manufacturer’s instructions, resolved in 20 ml RNase- and DNase-free water and stored at 280 uC. cDNA was synthesized using the SuperScript III First-Strand Synthesis System (Life Technologies) according to the manufacturer’s protocol, with a lock-docking oligodT primer containing a 59 overhang [GGCCACGCGTCGACTAGTAC(T)12VN]. The 39 portions of ReCV genomes (~3 kb) were amplified with gene-specific primers designed based on partial RNA-dependent RNA polymerase (RdRp) sequences described in our previous study (Farkas et al., 2010a) and the 59 overhang primer, GGCCACGCGTCGACTAGTAC, using the GoTaq Green Master Mix (Promega). DNA sequencing. PCR products were excised from 1.5 % agarose
gels, recovered using the Wizard SV Gel and PCR-clean up system, and cloned into pGEM-T vector according to the manufacturer’s protocols (Promega). Positive clones were identified using PCR. Plasmid DNA was extracted from small-scale bacterial cultures using the Wizard Plus SV Miniprep DNA-purification system (Promega), according to the manufacturer’s instructions, and sequenced using M13 forward and reverse and gene-specific primers (primer walking) with the chain-termination method on an ABI PRISM 3730 DNA Analyser (Applied Biosystems). Each sample was sequenced in both directions from at least two independent clones. Consensus sequences 1476
(CPE)-based virus neutralization assay was performed with heatinactivated (56 uC, 30 min) serum samples, as described previously (Farkas et al., 2010a, b, 2012). Briefly, twofold serial dilutions of hyperimmune antisera were titrated against 100 TCID50 of TV, FT285 and FT7 ReCVs. The virus/serum mix was incubated for 1 h at 37 uC and transferred to duplicate wells of 96-well tissue culture plates seeded with 104 LLC-MK2 cells per well. Virus and mock-inoculated control wells were included. Plates were stained with crystal violet at 72 h post-inoculation. By this time, all cells in the virus control wells were rounded and detached. End titrations were repeated in several experiments. End titres were calculated as mean values based on the highest dilution at which the cell monolayer was .50 % intact in both wells. This dilution against the homologous ReCV strain was taken as one VNU. Similar to entero-, rhino- and rotavirus serotyping (Committee on Enteroviruses, 1962; Kapikian et al., 1967; Wyatt et al., 1982), a 20-fold or greater difference in VN titres was required in both directions for establishing a ReCV serotype. Serotyping plaque assay. Due to the limited volume of the mouse hyperimmune sera (~0.5 ml/animal), serotyping of the 10 G1 ReCV isolates was performed in a cross-neutralization plaque assay with ~50 and ~100 p.f.u. of virus and 20 VNUs of each hyperimmune serum. A similar method was previously used for the serotyping of several viral agents, including avian infectious bronchitis viruses (Cowen & Hitchner 1975) and vesiviruses (Smith et al., 1998). Briefly, serum/ virus mix and controls were incubated for 1 h at 37 uC and adsorbed to LLC-MK2 monolayers in six-well tissue culture plates (Corning Life Sciences) for 1 h at 37 uC. Plates were washed twice with PBS and overlaid with culture medium containing 1.2 % Avicel (FMC BioPolymer). After 72 h of incubation, plates were stained with crystal violet and plaques counted. Each assay was repeated at least twice. ReCV strains with ¢80 % of plaque reduction were considered to be neutralized. Strains neutralized by a single specific antiserum were grouped as a serotype. Virus neutralization epitope prediction. Computational predic-
tion of linear B cell epitopes was performed with BepiPred 1.0 (Larsen et al., 2006). Three-dimensional structures of ReCV capsid proteins were generated using the Phyre 2 server (Kelley & Sternberg, 2009). Discontinuous B-cell epitopes were predicted with the DiscoTope 2.0 server (Kringelum et al., 2012). The predicted B-cell epitopes were evaluated based on analyses of multiple sequence alignments and cross-neutralization results (regions with serotype-specific differences) and the predicted three-dimensional structures (internal or surface-exposed regions) of ReCVs. Statistical analysis Statistical significance of cross-VN titres among
the ReCV isolates was evaluated with Fisher’s exact test. Journal of General Virology 95
Relationship between ReCV geno- and serotypes
ACKNOWLEDGEMENTS
Huang, P. W., Farkas, T., Zhong, W., Tan, M., Thornton, S., Morrow, A. L. & Jiang, X. (2005). Norovirus and histo-blood group antigens:
We thank Dr Peter Dickie (CCHMC) for critical reading of the manuscript. This study was supported by the Infectious Diseases Scholar Fund of CCHMC. A percentage of salary support (10%) was provided by NIH grant P01HD13021 to T. F.
demonstration of a wide spectrum of strain specificities and classification of two major binding groups among multiple binding patterns. J Virol 79, 6714–6722.
REFERENCES Bok, K., Parra, G. I., Mitra, T., Abente, E., Shaver, C. K., Boon, D., Engle, R., Yu, C., Kapikian, A. Z. & other authors (2011). Chimpanzees as an
animal model for human norovirus infection and vaccine development. Proc Natl Acad Sci U S A 108, 325–330. Choi, J. M., Hutson, A. M., Estes, M. K. & Prasad, B. V. (2008). Atomic
resolution structural characterization of recognition of histo-blood group antigens by Norwalk virus. Proc Natl Acad Sci U S A 105, 9175– 9180. Committee on Enteroviruses (1962). Classification of human
enteroviruses. Virology 16, 501–504. Cowen, B. S. & Hitchner, S. B. (1975). Serotyping of avian infectious
bronchitis viruses by the virus-neutralization test. Avian Dis 19, 583– 595. Debbink, K., Lindesmith, L. C., Donaldson, E. F. & Baric, R. S. (2012).
Norovirus immunity and the great escape. PLoS Pathog 8, e1002921. DiCaprio, E., Ma, Y., Purgianto, A., Hughes, J. & Li, J. (2012).
Internalization and dissemination of human norovirus and animal caliciviruses in hydroponically grown romaine lettuce. Appl Environ Microbiol 78, 6143–6152. Farkas, T., Sestak, K., Wei, C. & Jiang, X. (2008). Characterization of
a rhesus monkey calicivirus representing a new genus of Caliciviridae. J Virol 82, 5408–5416. Farkas, T., Cross, R. W., Hargitt, E., III, Lerche, N. W., Morrow, A. L. & Sestak, K. (2010a). Genetic diversity and histo-blood group antigen
interactions of rhesus enteric caliciviruses. J Virol 84, 8617–8625. Farkas, T., Dufour, J., Jiang, X. & Sestak, K. (2010b). Detection of
norovirus-, sapovirus- and rhesus enteric calicivirus-specific antibodies in captive juvenile macaques. J Gen Virol 91, 734–738. Farkas, T., Falkenstein, K. P., Bohm, R. P., Pecotte, J. & Sestak, K. (2012). High incidence of rhesus enteric calicivirus infections and
diarrhea in captive juvenile macaques: a likely association. J Med Primatol 41, 325–328. Frenck, R., Bernstein, D. I., Xia, M., Huang, P., Zhong, W., Parker, S., Dickey, M., McNeal, M. & Jiang, X. (2012). Predicting susceptibility to
norovirus GII.4 by use of a challenge model involving humans. J Infect Dis 206, 1386–1393. Green, K. Y. (2013). Caliciviridae: the noroviruses. In Fields Virology,
Kapikian, A. Z., Conan, R. M., Hamparian, V. V., Chanock, R. M., Chapple, P. J., Dick, E. C., Fenters, J. D., Gwaltney, J. M. & Hamre, D. (1967). Rhinoviruses: a numbering system. Nature 213, 761–763. Kapikian, A. Z., Wyatt, R. G., Dolin, R., Thornhill, T. S., Kalica, A. R. & Chanock, R. M. (1972). Visualization by immune electron microscopy
of a 27-nm particle associated with acute infectious nonbacterial gastroenteritis. J Virol 10, 1075–1081. Katpally, U., Wobus, C. E., Dryden, K., Virgin, H. W., IV & Smith, T. J. (2008). Structure of antibody-neutralized murine norovirus and
unexpected differences from viruslike particles. J Virol 82, 2079–2088. Kelley, L. A. & Sternberg, M. J. E. (2009). Protein structure prediction
on the Web: a case study using the Phyre server. Nat Protoc 4, 363– 371. Kringelum, J. V., Lundegaard, C., Lund, O. & Nielsen, M. (2012).
Reliable B cell epitope predictions: impacts of method development and improved benchmarking. PLOS Comput Biol 8, e1002829. Kroneman, A., Vega, E., Vennema, H., Vinje´, J., White, P. A., Hansman, G., Green, K., Martella, V., Katayama, K. & Koopmans, M. (2013).
Proposal for a unified norovirus nomenclature and genotyping. Arch Virol 158, 2059–2068. Larsen, J. E., Lund, O. & Nielsen, M. (2006). Improved method for predicting linear B-cell epitopes. Immunome Res 2, 2. Li, X., Ye, M., Neetoo, H., Golovan, S. & Chen, H. (2013). Pressure
inactivation of Tulane virus, a candidate surrogate for human norovirus and its potential application in food industry. Int J Food Microbiol 162, 37–42. Lindesmith, L., Moe, C., Marionneau, S., Ruvoen, N., Jiang, X., Lindblad, L., Stewart, P., LePendu, J. & Baric, R. S. (2003). Human
susceptibility and resistance to Norwalk virus infection. Nat Med 9, 548–553. Lindesmith, L. C., Beltramello, M., Donaldson, E. F., Corti, D., Swanstrom, J., Debbink, K., Lanzavecchia, A. & Baric, R. S. (2012).
Immunogenetic mechanisms driving norovirus GII.4 antigenic variation. PLoS Pathog 8, e1002705. Lindesmith, L. C., Costantini, V., Swanstrom, J., Debbink, K., Donaldson, E. F., Vinje´, J. & Baric, R. S. (2013). Emergence of a
norovirus GII.4 strain correlates with changes in evolving blockade epitopes. J Virol 87, 2803–2813. Lopman, B. A., Hall, A. J., Curns, A. T. & Parashar, U. D. (2011).
Increasing rates of gastroenteritis hospital discharges in US adults and the contribution of norovirus, 1996–2007. Clin Infect Dis 52, 466–474.
6th edn, vol. 1, pp 582–608. Edited by D. M. Knipe & P. M. Howley. Philadelphia: Lippincott Williams & Wilkins.
McFadden, N., Bailey, D., Carrara, G., Benson, A., Chaudhry, Y., Shortland, A., Heeney, J., Yarovinsky, F., Simmonds, P. & other authors (2011). Norovirus regulation of the innate immune response
Hall, A. J., Curns, A. T., McDonald, L. C., Parashar, U. D. & Lopman, B. A. (2012). The roles of Clostridium difficile and norovirus among
and apoptosis occurs via the product of the alternative open reading frame 4. PLoS Pathog 7, e1002413.
gastroenteritis-associated deaths in the United States, 1999–2007. Clin Infect Dis 55, 216–223.
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
Handley, S. A., Thackray, L. B., Zhao, G., Presti, R., Miller, A. D., Droit, L., Abbink, P., Maxfield, L. F., Kambal, A. & other authors (2012).
genotype 2 bovine noroviruses are antigenically distinct but share a crossreactive epitope with human noroviruses. J Clin Microbiol 44, 992–998.
Pathogenic simian immunodeficiency virus infection is associated with expansion of the enteric virome. Cell 151, 253–266.
Parker, T. D., Kitamoto, N., Tanaka, T., Hutson, A. M. & Estes, M. K. (2005). Identification of Genogroup I and Genogroup II broadly
Hirneisen, K. A. & Kniel, K. E. (2013a). Norovirus surrogate survival
reactive epitopes on the norovirus capsid. J Virol 79, 7402–7409.
on spinach during preharvest growth. Phytopathology 103, 389–394. Hirneisen, K. A. & Kniel, K. E. (2013b). Comparing human norovirus
Parra, G. I., Azure, J., Fischer, R., Bok, K., Sandoval-Jaime, C., Sosnovtsev, S. V., Sander, P. & Green, K. Y. (2013). Identification of
surrogates: murine norovirus and Tulane virus. J Food Prot 76, 139– 143.
a broadly cross-reactive epitope in the inner shell of the norovirus capsid. PLoS ONE 8, e67592.
http://vir.sgmjournals.org
1477
T. Farkas, C. Wong Ping Lun and B. Fey Richards, G. P. (2012). Critical review of norovirus surrogates in food safety research: rationale for considering volunteer studies. Food Environ Virol 4, 6–13.
human 1195.
Scallan, E., Hoekstra, R. M., Angulo, F. J., Tauxe, R. V., Widdowson, M. A., Roy, S. L., Jones, J. L. & Griffin, P. M. (2011). Foodborne illness acquired
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Kumar, S. (2011). MEGA5: molecular evolutionary genetics analysis using
in the United States–major pathogens. Emerg Infect Dis 17, 7–15.
maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 2731–2739.
Sestak, K., Feely, S., Fey, B., Dufour, J., Hargitt, E., Alvarez, X., Pahar, B., Gregoricus, N., Vinje´, J. & Farkas, T. (2012). Experimental ino-
A. D. (2012). Calicivirus from novel Recovirus genogroup in
diarrhea,
Bangladesh.
Emerg
Infect
Dis
18,
1192–
Vashist, S., Bailey, D., Putics, A. & Goodfellow, I. (2009). Model
culation of juvenile rhesus macaques with primate enteric caliciviruses. PLoS ONE 7, e37973.
systems for the study of human norovirus Biology. Future Virol 4, 353–367.
Shanker, S., Choi, J. M., Sankaran, B., Atmar, R. L., Estes, M. K. & Prasad, B. V. (2011). Structural analysis of histo-blood group antigen
Wobus, C. E., Thackray, L. B. & Virgin, H. W., IV (2006). Murine
binding specificity in a norovirus GII.4 epidemic variant: implications for epochal evolution. J Virol 85, 8635–8645. Smith, A. W., Berry, E. S., Skilling, D. E., Barlough, J. E., Poet, S. E., Berke, T., Mead, J. & Matson, D. O. (1998). In vitro isolation and
characterization of a calicivirus causing a vesicular disease of the hands and feet. Clin Infect Dis 26, 434–439. Smits, S. L., Rahman, M., Schapendonk, C. M., van Leeuwen, M., Faruque, A. S., Haagmans, B. L., Endtz, H. P. & Osterhaus,
1478
norovirus: a model system to study norovirus biology and pathogenesis. J Virol 80, 5104–5112. Wyatt, R. G., Greenberg, H. B., James, W. D., Pittman, A. L., Kalica, A. R., Flores, J., Chanock, R. M. & Kapikian, A. Z. (1982). Definition of
human rotavirus serotypes by plaque reduction assay. Infect Immun 37, 110–115. 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.
Journal of General Virology 95
DOI 10.1099/vir.0.064675-0
Relationship between genotypes and serotypes of genogroup 1 recoviruses: a model for human norovirus antigenic diversity Tibor Farkas,1,2 Cindy Wong Ping Lun13 and Brittney Fey14 Correspondence Tibor Farkas [email protected]
Received 13 February 2014 Accepted 31 March 2014
1
Cincinnati Children’s Hospital Medical Center, OH, USA
2
University of Cincinnati College of Medicine, Cincinnati, OH, USA
Human norovirus (NoV) research greatly relies on cell culture-propagable surrogate caliciviruses, including murine NoVs and the prototype ‘recovirus’ (ReCV), Tulane virus. However, the extreme biological diversity of human NoVs cannot be modelled by a uniform group of viruses or single isolate. Based on a diverse group of recently described ReCVs, a more advanced model reflecting human NoV biological diversity is currently under development. Here, we have reported the genotypic and serotypic relationships among 10 G1 ReCV isolates, including Tulane virus and nine other recent cell culture-adapted strains. Based on the amino acid sequences of virus capsid protein, VP1, and classification constraints established for NoVs, G1 ReCVs were separated into three genotypes, with variable organization of the three open reading frames. Interestingly, crossneutralization plaque assays revealed the existence of four distinct serotypes, two of which were detected among the G1.2 strains. The amino acid (aa) difference between the two G1.2 ReCV serotypes (12%) was less than the minimum 13 % difference established between NoV genotypes. Interestingly, one of the G1.3 ReCVs was equally neutralized by antisera raised against the G1.3 (6 % aa difference) and G1.1 (25 % aa difference) representative strains. These results imply the existence of a large number of human NoV serotypes, but also shared crossneutralization epitopes between some strains of different genotypes. In conclusion, the newly developed ReCV surrogate model can be applied to address biologically relevant questions pertaining to enteric CV diversity.
INTRODUCTION Caliciviruses (CV) are non-enveloped, positive-sense, single-stranded RNA viruses with polyadenylated genomes of ~6.5–8.5 kb. The Caliciviridae family consists of five established genera: Norovirus, Sapovirus, Lagovirus, Vesivirus and Nebovirus. Three recently proposed CV genera, ‘Recovirus’, ‘Valovirus’ and ‘chicken calicivirus’, are awaiting ICTV approval (Green, 2013). Depending on the genus, CV genomes are organized into two or more ORFs. Noroviruses (NoVs) and vesiviruses encode nonstructural (NS) and capsid (VP1) proteins in separate ORFs (ORF1 and ORF2, respectively), while sapoviruses and lagoviruses encode both NS and capsid proteins in a continuous ORF1. A separate ORF encoding a minor structural protein (VP2), designated ORF3 in the noro- and vesivirus genomes and ORF2 in the sapo- and 3Present address: University of Cincinnati, Cincinnati, OH, USA. 4Present address: Northern Kentucky University, Highland Heights, KY, USA. The GenBank/EMBL/DDBJ accession number for the sequences determined in this study are KC662363–KC662370 and KF431831.
064675 G 2014 The Authors
Printed in Great Britain
lagovirus genomes, is present in all CVs (Green, 2013). In murine NoV genomes, an additional ORF (ORF4) overlapping ORF2 encodes a protein (VF1) that may play a role in virulence (McFadden et al., 2011). CVs cause a variety of diseases in animals. In humans, NoVs and sapoviruses, commonly referred to as ‘human CVs’, are two of the leading causes of acute gastroenteritis. In the United States alone, NoVs cause an estimated 23 million cases of acute gastroenteritis, leading to .70 000 hospitalizations and 800 deaths each year (Lopman et al., 2011; Scallan et al., 2011; Hall et al., 2012). The prototype NoV (Norwalk virus) was discovered more than 40 years ago (Kapikian et al., 1972). However, due to the lack of an efficient cell culture system and robust animal model, human NoV research still relies on surrogate CVs and/or hosts (e.g. murine NoV or gnotobiotic pig/calf) that do not necessarily reflect the essential biological features and diversity of human NoVs and their hosts (Table 1). The chimpanzee model is a promising alternative for the preclinical evaluation of NoV vaccines and antivirals (Bok et al., 2011). However, the use of chimpanzees for biomedical research is restricted in most countries worldwide, including 1469
T. Farkas, C. Wong Ping Lun and B. Fey
Table 1. Comparison of human NoVs and surrogate caliciviruses, including ReCVs HuNoV Genus Genome organization Diversity HBGA binding Efficient cell culture Efficient reverse genetics Transmission Diarrhoea in immunocompetent host Shedding in stool Zoonotic transmission Availability from ATCC
Norovirus 3 ORFs High Yes No No Faecal–Oral Yes Yes Suggested No
MuNoV
PoNoV
BoNoV
PECV
Norovirus Norovirus Norovirus Sapovirus 4 ORFs 3 ORFs 3 ORFs 2 ORFs Low Low Low Low No No No No Yes No No Yes Yes No No Yes Faecal–Oral Faecal–Oral Faecal–Oral Faecal–Oral No No Yes Yes Yes No Yes
Yes Suggested No
Yes Suggested No
Yes No No
FCV Vesivirus 3 ORFs Low No Yes Yes Respiratory No No* No Yes
RHDV
ReCV
Lagovirus Recovirus 2 ORFs 3 ORFs Low High Yes Yes No Yes Yes Yes Faecal–Oral Faecal–Oral No Yes Yes No No
Yes Suggested No
HuNoV: human norovirus; MuNoV: Murine norovirus; PoNoV: porcine norovirus; BoNoV: bovine norovirus; PECV: porcine enteric calicivirus; FCV: feline calicivirus; RHDV: rabbit haemorrhagic disease virus; ReCV: rhesus enteric calicivirus. *FCV primarily sheds in oral and oculonasal discharge, but can be found in all body secretions, including faeces, during acute disease.
the USA. At present, licensed vaccines or antiviral strategies to prevent NoV disease are not available. One of the major challenges in NoV vaccine design is the tremendous diversity of human NoVs (Green, 2013). According to phylogenetic differences of the capsid protein, VP1, NoVs are classified into five genogroups (G1–G5), with several genetic types in each genogroup (e.g. G1.1–G1.8). Based on analysis of 164 human and animal NoV VP1 proteins, amino acid (aa) sequence differences established between strains (0–14.07 %) within a genetic type, between genotypes (14.26–43.78 %) and between genogroups (44.91–64.41 %) are used for the genetic classification of new isolates (Zheng et al., 2006; Kroneman et al., 2013). Genogroups G1, 2 and 4 mainly contain human NoVs, while G3 includes bovine and G5 murine NoVs. Over 30 human NoV genetic types have been described to date (Green, 2013). Antigenic characterization of human NoVs mainly relies on virus-like particles (VLP) and antibody binding assays (ELISA). Several studies have demonstrated the existence of cross-reactive conserved epitopes among the different NoV genotypes and even genogroups, predominantly located in the S-domain or Cterminal region of the P1 domain (Parker et al., 2005; Oliver et al., 2006; Parra et al., 2013). The role of histo-blood group antigens (HBGA) in human NoV binding/attachment and susceptibility to infection, and diversity of the HBGA binding properties of different human NoV strains are well established (Lindesmith et al., 2003; Huang et al., 2005; Frenck et al., 2012). HBGA binding sites have been mapped to the most variable P2 subdomain of the NoV capsid (Choi et al., 2008; Shanker et al., 2011). The surface-exposed P2 subdomain probably also contains the epitopes responsible for major antigenic differences (serotypes). Antibodies that block adherence of human NoV VLPs to HBGAs in binding assays (ELISA) have been proposed as virus-neutralizing (VN) antibodies. Recent results based on blocking assays performed mainly 1470
with G2.4 human NoVs suggested that antigenic changes facilitating escape from herd immunity may also drive changes in HBGA binding affinities and lead to altered population susceptibility (Debbink et al., 2012; Lindesmith et al., 2012). However, due to the lack of a human NoV cell culture system, evaluating the roles of HBGAs as attachment ligands or receptors and direct assessment of human NoV antibody neutralization, including the identification of neutralizing epitopes and relationship between genotypes and serotypes, remains a significant challenge. Murine NoVs are closely related to their human counterparts, replicate well in cell culture and represent a costeffective small animal model (Wobus et al., 2006). However, the murine NoV model also has limitations highlighted by significant differences compared with human NoVs. Murine NoVs exhibit limited genetic and antigenic diversity, have apparent immune cell tropism and cause persistent asymptomatic infection in normal mice. The prototype ‘recovirus’ (ReCV), Tulane virus, was discovered in stool samples of juvenile rhesus macaques (Farkas et al., 2008). This initial discovery was followed by the identification of several other ReCVs, including one strain in human stool samples (Farkas et al., 2010a; Handley et al., 2012; Smits et al., 2012). Since then, several additional ReCV strains described in our previous reports (Farkas et al., 2010a) have been cell culture-adapted. Studies on Tulane virus and other isolates established that ReCVs are evolutionarily and biologically closely related to human NoVs, including genetic, antigenic and HBGA binding diversities and clinical symptoms of gastroenteritis observed in experimentally infected macaques (Farkas et al., 2008, 2010 a, b; Sestak et al., 2012). Based on the biologically diverse ReCV isolates, we are currently developing a cell culture and animal model that can effectively replicate the biological features of human NoVs. Here, we examined the relationship between Journal of General Virology 95
Relationship between ReCV geno- and serotypes
genotypes and serotypes within G1 ReCVs as part of this ongoing project.
GTCATAGACATG), 3 nt in the G1.2 ReCV genomes (TGAGCTATG) and 2 nt in the G1.3 ReCVs (TGATCATG). ORF2 start and ORF3 stop codons of all NoVs and ReCVs overlapped by 1 nt (TAATG).
RESULTS G1 ReCV serotypes
Phylogeny and genome organization Phylogenetic analysis of complete ORF2 (VP1) sequences of the three ReCVs available in public databases (Tulane virus, WUHARV and Bangladesh/289/2007) and nine tissue culture-adapted ReCVs obtained in this study confirmed our earlier classification based on short RdRp sequences (Farkas et al., 2010a). Genogroup 1 ReCVs were clearly segregated into three genotypes with evolutionary distances comparable to those between the human NoV genotypes. The amino acid differences between strains of the three genotypes ranged from 24 to 30 % (Table 2). In addition to G2 rhesus isolates, the human ReCV isolate, Bangladesh/ 289, represents a third genogroup (G3) (Fig. 1). Analysis of the ReCV genomes revealed three ORFs with significant variations in arrangement. When ORF1 was adjusted as frame 1, all G1.1 and G1.2 ReCVs exhibited an order of frames 1, 1 and 3, while the order for G1.3 ReCVs was frames 1, 3 and 2 for ORF1, ORF2 and ORF3, respectively. Human NoVs, murine NoVs and ReCV isolate Bangladesh/289 exhibited an order of frames 1, 2 and 1, with an additional ORF4 in-frame 3 in murine NoV genomes (Fig. 2). The Norwalk virus ORF1 stop and ORF 2 start codons overlapped by 11 nt (ATGATGATGGCGTCTAA), the murine NoV by 8 nt (ATGAGGATGAGTGA) and ReCV Bangladesh/289 by 146 or 113 nt, depending on which of the two possible start codons was utilized (ATGAGCTCCCAAAAGGAAAATCTGTCAAACAAAATGGAGGAAGTGGTGCAACCAGCAACTGGAGGGGTTGGCGAAGTCATAACCGGAGCAGCCGCCCCACTACCAGAGAATCCAAGAGGACTGGAATTGGCCCCAACAATCAATCCCATTGA). The ORF1 stop and ORF 2 start codons were separated by 21 nt in the G1.1 ReCV genomes (TGATTGATCACAATT-
Hyperimmune sera against the Tulane virus (G1.1), FT285 (G1.2) and FT7 (G1.3) ReCVs exhibited low-level crossreactivity. The difference in titres between the homologous and heterologous antisera ranged between 102- and 2560fold, and was greater for all strains than the 20-fold cutoff used to define serotypes (Table 3). Thus, the three G1 strains used for generating hyperimmune sera represent not only three genotypes but also three serotypes. Serotypic evaluation of the 10 G1 ReCV isolates was performed in a standardized cross-neutralization plaque assay with 20 virus neutralization units (VNU) of serotypespecific hyperimmune sera. Anti-Tulane virus hyperimmune sera neutralized all G1.1 isolates and FT65, a G1.3 strain, but not the other two G1.3 or the G1.2 strains. Surprisingly, anti-FT285 (G1.2) hyperimmune sera only neutralized FT285, but not the two other G1.2 isolates (FT157 and FT499). Anti-FT7 (G1.3) hyperimmune sera neutralized only the G1.3 strains (Fig. 3). Based on these results, the 10 G1 ReCVs were grouped into four serotypes, specifically, serotype 1 (G1.1), serotype 2 (FT285, G1.2), serotype 3 (FT7 and FT400, G1.3) and serotype 4 (FT157 and FT499, G1.2). The exact serotype of FT65 (G1.3) ReCV could not be determined with the antibodies used in this study. Virus neutralization epitope prediction Several discontinuous B cell epitopes were predicted in almost all of the 10 ReCV strains in this study. These included PPT in the hinge region (aa positions 206–208 in the FT285 VP1 sequence), AITDKP-RR in loop A’-B’ of the P2 domain (aa positions 283–288) and M-VSG (aa
Table 2. VP1 pairwise amino acid difference and homology values (%) of ReCVs TV
FT205
FT218
FT494
FT285
FT499
WUHARV
FT7
FT65
FT400
Bangladesh
– 9 9 9 26 26 25 25 25 25 62
91 – 3 3 25 25 25 26 26 25 62
91 97 – 2 25 24 24 24 25 24 61
91 97 98 – 26 25 25 25 26 25 61
74 75 75 74 – 12 10 29 30 29 60
74 75 76 75 88 – 8 29 29 29 60
75 75 76 75 90 92 – 29 29 29 60
75 74 76 75 71 71 71 – 6 3 60
75 74 75 74 70 71 71 94 – 5 59
75 75 76 75 71 71 71 97 95 – 61
38 38 39 39 40 40 40 40 41 39 –
G1.1-TV G1.1-FT205 G1.1-FT218 G1.1-FT494 G1.2-FT285 G1.2-FT499 G1.2-WUHARV G1.3-FT7 G1.3-FT65 G1.3-FT400 G3.1-Bangladesh
Pairwise amino acid difference and homology scores are shown in the lower and upper diagonal half of the table, respectively. http://vir.sgmjournals.org
1471
T. Farkas, C. Wong Ping Lun and B. Fey
(a)
87 99 99 98 99 96 99 99
99 93
‘Recovirus’ 99
47 99
Norovirus
99
85
99
60 61
0.5
0.4
0.3
98
99 83 0.2
0.1
FT221 FT412 FT218* FT216 FT363 FT226 FT236 FT205* G1.1 FT457 FT494* FT472 FT422 FT473 FT234 FT187 Tulane virus* WUHARV FT285* FT293 FT499* FT213 FT176 G1.2 FT15 FT157* FT150 FT138 FT135 FT143 FT172 FT413 FT411 FT80 FT65* FT191 FT405 FT269 FT219 FT294 FT305 FT299 FT350 G1.3 FT7* FT242 FT30 FT301 FT394 FT390 FT416 FT454 FT400* FT177 FT36 FT41 ReCVBangladesh G3.1 FT286 FT310 FT443 G2.1 FT349 FT399 FT437 G5 MNV (AY228235) G2.1 Hawaii (U07611) G2.2 SMV (AY134748) G2.3 SU201 (AB039782) G4.1 (JQ613567) G3.1 BoNoVJena (AJ011099) G3.2 BoNoVNewbury (AF097917) G1.3 DSV (U04469) G1.1 Norwalk (M87661) G1.2 Southampton (L07418)
(b)
70 100 100 52 100 82
100
‘Recovirus’
100 79 100
100
100 98
Norovirus
100 100
100
99 99
100 89 0.6
0.5
0.4
0.3
0.2
0.1
FT218 FT494 FT205 Tulane virus (EU391643) FT65* FT400 FT7 FT285 WUHARV (JX627575) FT157 FT499 ReCV Bangladesh (JQ745645) G5.1 MNV (AY228235) G5.2 TF5WM (JQ408737) G2.1 Hawaii (U07611) G2.3 SMV (AY134748) G2.3 SU201 (AB039782) G4.1 (JQ613567) G3.1 Jena (AJ011099) G3.2 Newbury (AF097917) G1.3 DSV (U04469) G1.1 Norwalk (M87661) G1.2 Southampton (L07418)
G1.1
G1.3
G1.2 G3.1
0.0
0.0
Fig. 1. (a) Genetic classification of ReCVs based on partial (268 nt) RdRp sequences (Farkas et al., 2010a). ReCV strains used in Fig. 1(b) are in bold type. ReCVs used in tissue culture-based assays are marked with an asterisk. (b) Genetic classification of ReCVs based on complete VP1 amino acid sequences. Cell culture-adapted ReCV isolates described in this study are in bold type. GenBank accession numbers are KC662363–KC662370 and KF431831. The scale bars indicate the evolutionary distance represented by nucleotide or amino acid substitutions per site, respectively. The numbers at the nodes indicate the percentage of times that a node was supported during 1000 bootstrap analyses.
positions 385–389). The hinge epitope was predicted in all strains except TV, loop A’-B’ epitope in all strains except FT205 and FT494, and M-VSG epitope in all strains except TV, FT7 and FT400. Analysis of multiple sequence alignments highlighted the loop A’-B’ epitope (Fig. 4) as the most likely candidate responsible for the VN differences among the strains (Fig. 3).
DISCUSSION Since the discovery of the prototype Tulane virus (Farkas et al., 2008), a number of similarities have been identified between ReCVs and human NoVs, including close evolutionary relatedness, genetic diversity, HBGA binding properties, epidemiology and disease spectrum (Farkas 1472
et al., 2010a, b, 2012; Sestak et al., 2012). Human NoVs cannot be grown in cell culture. For this reason, other CVs are used as surrogates for human NoV research (Vashist et al., 2009; Richards, 2012). Cell culture-propagable murine NoVs have emerged as the most commonly utilized human NoV surrogates (Wobus et al., 2006). However, there remains a need for more robust surrogates, as evident from the ongoing NoroCore projects that form part of the USDANIFA Food Virology Initiative (http://norocore.ncsu.edu). These efforts have yielded several recently published studies with the prototype Tulane virus (DiCaprio et al., 2012; Hirneisen & Kniel 2013a, b; Li et al., 2013). One of the major challenges in human NoV research, particularly for vaccine development, is the enormous diversity (genetic, antigenic, HBGA binding) of human NoVs. These features cannot be Journal of General Virology 95
Relationship between ReCV geno- and serotypes
5364
NORWALK VIRUS (All human noroviruses) p48
NTPase
6947 VP1
p22 VPg
Prot
Pol
69 nt (A)n
VP2 5371
5
6950
7585
MNV p38
NTPase
p12 VPg
Prot
Pol 5066
6
76 nt (A)n
VP2
5069 5707
7304
6681
VF VP1
5062
6647 TULANE VIRUS (G1.1 and G1.2 Recoviruses: FT205, FT218, FT494. FT157, FT285, FT499 and WUHARV) p26
NTPase
p16 VPg
Prot
Pol
VP1 4355 4380
15
5981 VP2 5984
FT7 (G1.3 Recoviruses: FT65, FT400)
1
2343 730
Prot
2340
2735
4384
2969
VP1 VP2
Pol 2883
1
73 nt (A)n
VP1
736 RECOVIRUS/BANGLADESH/289 p16 VPg
6637
VP2 Pol
74 nt (A)n
4387
73 nt (A)n 4992
Fig. 2. Genome organization of ReCVs and Noroviruses. ORFs are shown as boxes. ORF1 encodes non-structural proteins, ORF2 the capsid protein (VP1), and ORF3 the minor structural protein (VP2). In the murine NoV genome, a fourth ORF that overlaps with ORF2 has been identified that encodes virulence factor (VF1) protein (McFadden et al., 2011).
duplicated by a uniform group of viruses or single isolate. Based on the availability of a large collection of diverse ReCVs in our laboratory, we recently initiated the development of a cell culture and non-human primate model able
to duplicate the diverse biological features of human NoVs and their hosts. In the current study, we evaluated the relationship between the genotypes and serotypes of an enteric CV for the first time using this unique model.
Table 3. Cross-reactivity of anti-ReCV hyperimmune mouse sera
Partial RdRp, whole capsid protein (VP1) and whole minor structural protein (VP2)-encoding sequences were obtained for nine cell culture-adapted G1 ReCVs. Phylogenetic analysis of ORF2 nt and VP1 aa sequences confirmed our previous classification based on short (268 nt) RdRp sequences (Farkas et al., 2010a) and the existence of three genotypes among G1 ReCVs. Amino acid differences among the ReCV genotypes (24–30 %) were within the range (14.26–43.78 %) established for human NoV genotypes (Zheng et al., 2006). ReCV WUHARV was grouped with the G1.3 strains, while the human ReCV isolate, Bangladesh/289, was clearly separated from rhesus isolates with 59–62 % aa differences, equivalent to those reported between the human NoV genogroups (44.91–64.41 %) (Table 2) (Zheng et al., 2006). Genogroup 2 ReCV VP1 sequences have not been published yet. However, based on analyses of published VP1 and/or RdRp sequences, ReCVs
Reciprocal of VN titres
TV FT285 FT7 Homologous/ heterologous*
Anti-TV
Anti-FT285
Anti-FT7
5120 10 50 102-fold
10 25 600 ,10 2560-fold
40 40 12 800 320-fold
VN end titres were calculated based on repeated experiments with twofold dilutions, starting from 1 : 10 and 1 : 50. *The least difference of titres of an antiserum against homologous and heterologous strains. Twenty-fold or greater difference in both directions was used to distinguish serotypes. http://vir.sgmjournals.org
1473
T. Farkas, C. Wong Ping Lun and B. Fey
S2
(a) G1/1/S 1 FT205
S3
S 1+3
G1/2
FT218
FT494
FT285
S3
G1/3
FT157
FT499
FT7
FT65
FT400
a-FT7
a-FT285
a-Tulane
Mouse sera (20 VNU)
Control
Tulane
S4
(b) 100
100
80
80
80
60
60
60
40
40
40
* *
*
*
20 *
*
0 T FT V 20 FT 5 2 FT 18 4 FT 94 28 FT 5 15 FT 7 49 9 FT 7 FT 40 FT 0 65
0
Anti-TV serum
*
*
*
*
*
*
20 *
*
* 0
Anti-FT285 serum
* *
*
*
*
*
T FT V 20 FT 5 2 FT 18 4 FT 94 28 FT 5 49 FT 9 15 7 FT FT 7 40 FT 0 65
20
T FT V 20 FT 5 2 FT 18 49 FT 4 28 FT 5 15 FT 7 49 9 FT FT 7 40 FT 0 65
Plaque reduction (%)
100
Anti-FT7 serum
Fig. 3. Serotyping of G1 ReCVs. (a) Serotyping plaque assay. Twenty virus neutralization units (VNU) of each hyperimmune mouse serum were used to cross-neutralize 50–100 p.f.u. of ReCV. (b) Mean plaque reduction (%) of repeat experiments. Error bars represent standard deviation. (*) Shows statistically significant difference (P,0.05), compared with neutralization by homologous virus/serum. ($) Indicates neutralization by genotype-matched sera. (#) Indicates lack of neutralization by genotype-matched sera. (e) Indicates neutralization by sera raised against a heterologous genotype. Serotypes were established based on ¢80 % plaque reduction.
were classified into three genogroups (G1–G3), with three genotypes within G1 (G1.1–G1.3) (Fig. 1). Considering that most known ReCV sequences originated from samples collected from the same colony of captive macaques between April and July of 2008 (Farkas et al., 2010a), the observed genetic diversity indicates highly diverse, human NoV-like evolution of ReCVs. Analysis of the ReCV genomes revealed variable arrangement of the three ORFs among the different ReCV genotypes that was not observed among NoVs (Fig. 2). Further studies using different geographical locations and primate species, including humans, are necessary to determine the full extent of ReCV diversity. To standardize for cell culture-based serotyping assays, only ReCV strains isolated in the LLC-MK2 cell line were 1474
involved in this study. Mouse hyperimmune sera generated against one representative strain of each of the three G1 genotypes were used for serotyping. End titration with the three representative strains used for immunization based on requirements established for picornavirus and rotavirus serotyping (Committee on Enteroviruses, 1962; Kapikian et al., 1967; Wyatt et al., 1982) revealed that these strains represent three different serotypes. The cross-reactivity between Tulane virus (serotype 1) and FT7 (serotype 3) was stronger than between FT285 (serotype 2) or other serotypes. However, the difference in titres in both directions was still at least fourfold higher between all strains than the 20-fold minimum cut-off value established for separating serotypes (Table 3). Journal of General Virology 95
Relationship between ReCV geno- and serotypes
(a)
Tulane FT205 FT218 FT494 FT285 FT157 FT499 FT7 FT65 FT400
(b) C’-D’ Norwalk
A’-B’ E’-F’
Tulane
FT285
FT7
Hinge
(c)
MD145
V-SHD
FT285
AITDKP-R
Due to the limited volume of mouse hyperimmune sera, antigenic relationships between the 10 ReCV isolates were evaluated in a standardized cross-neutralization plaque assay. To adjust for differences in VN antibody levels in individual hyperimmune serum samples, the test was performed with a single dilution of each serum (representing 20 VNUs) against the homologous strain (Cowen & Hitchner, 1975; Smith et al., 1998). The cross-neutralization assay highlighted statistically significant differences among the ReCV isolates. All the G1.1 isolates belonged to serotype 1. The anti-FT285 (serotype 2) serum was not able to neutralize the other two G1.2 strains (FT157 and FT499), leading to separation of the G1.2 isolates into two serotypes (serotypes 2 and 4). All G1.3 strains were neutralized by anti-FT7 (serotype 3) serum, and consequently grouped as serotype 3. One of the G1.3 isolates (FT65) was also neutralized by serotype 1-specific serum (Fig. 3). This finding is of significant interest, since FT65 and FT7 differ only by 6 %, while FT65 and Tulane virus differ by 25 % in terms of VP1 aa positions (Table 2). Heterologous sera exhibited significantly lower neutralization for all strains, except FT65, (P,0.05). To our knowledge, this is the first study to correlate genetic and serotypic differences for an enteric CV and link the genetic classification scheme to a biologically relevant http://vir.sgmjournals.org
Fig. 4. Predicted ReCV neutralization B cell epitope. (a) Multiple sequence alignment. Amino acid positions according to Tulane virus VP1 from 275 to 293 aa. Linear B-cell epitope predicted by BepiPred 1.0 in this region is underlined. Discontinuous B-cell epitope predicted by the DiscoTope 2.0 server in this region is presented in bold type. (b) Structural comparison of Norwalk virus and ReCVs. Three-dimensional structures of capsid proteins were generated using the Phyre 2 server. The S-domain is in gold, P1 domain in green and P2 domain in red. Surface-exposed loops in the P2 domain are labelled as A’-B’, C’-D’ and E’-F’. The predicted neutralization epitope located in loop A’-B’ of the P2 domain is in yellow. Amino acids in the Norwalk virus structure interacting with H-type pentasaccharide (D327, H329, Q342, D344, W375, S377 and P378) are in blue (Choi et al., 2008). (c) Structural overlap of the predicted ReCV neutralization epitope and highly variable key blockade epitope (epitope A; V-SHD) of G2.4 human NoVs (Lindesmith et al., 2013). Three-dimensional structures of MD145 (G2.4 human NoV) and FT285 were generated using the Phyre 2 server. Amino acids comprising the epitopes are depicted in yellow.
measure. In general, a strong correlation between ReCV genotypes and serotypes was observed. In an earlier study by Zheng et al. (2006), the minimum aa difference detected between NoV genotypes was 14.26 %, while in a more recent study including newly described NoVs, a difference of 13 % was reported (Kroneman et al., 2013). The smallest aa difference observed in this study between ReCV serotypes (FT285 and the other two G1.2 strains) was 12 %. On the other hand, we identified cross-neutralization between strains with a 25 % aa difference. FT157 or FT499 (serotype 4) and FT65-specific hyperimmune sera were not generated in this study. Follow-up studies that include mAbs and more ReCV isolates and sera are critical. While we were unable to evaluate whether protection against infection is serotype-specific in the rhesus macaque model, the implication of our findings for human NoV research, especially vaccine development, is significant. Further clarification of the mechanism of ReCV antibody neutralization is important for human NoV research. Meanwhile, since the VP1 sequences and cross-VN results disclose valuable information, we performed computational B cell epitope predictions on modelled 3D structures of ReCVs. A candidate epitope on loop A’-B’ of the P2 domain was identified (Fig. 4) that was also predicted at a similar location on several human NoV structures (data not shown). A literature search 1475
T. Farkas, C. Wong Ping Lun and B. Fey
revealed that an equivalent region on loop A’-B’ of the murine NoV P2 domain has been identified as one of the contact sites for a neutralizing mAb (A6.2) (Katpally et al., 2008). This epitope structurally overlaps with the highly variable key blockade epitope A that is reported to change with new G2.4 human NoV emergence (Lindesmith et al., 2013) (Fig. 4c). In summary, our findings clearly demonstrate the value of the ReCV model as a surrogate for analysing the diverse biological features of human noroviruses, and pave the way for detailed evaluation of enteric CV antibody neutralization and escape as well as the relationship with HBGA binding through cell culture-based, structural and in vivo studies.
were submitted to the GenBank database under accession numbers KC662363–KC662370 and KF431831. BLAST analyses were run against NCBI databases. Multiple sequence alignments of nt and aa sequences were created using Omiga v2.0 software (Oxford Molecular). Phylogenetic trees were constructed using the unweighted pair group method with arithmetic mean (UPGMA) and the neighbour-joining clustering methods of the Molecular Evolutionary Genetics Analysis (MEGA version 5.1) software with Jukes-Cantor and Poisson correction distance calculations for nt and aa sequence alignments, respectively (Tamura et al., 2011). The confidence values of the internal nodes were obtained by performing 1000 bootstrap analyses. Accession numbers of the TV, WUHARV, Recovirus Bangladesh/289 and human NoVs included in the alignments are listed in Fig. 2a.
Sequence and phylogenetic analyses.
Cross-titration of mouse hyperimmune sera. A cytopathic effect
METHODS Viruses and cell lines. Genogroup I ReCVs were isolated from rhesus stool samples collected during our previous studies (Farkas et al., 2008, 2010a). All strains were isolated in LLC-MK2 cell lines (ATCC CCL-7). Virus stocks were prepared from viruses plaquepurified three times and stored at 280 uC in aliquots. Anti-ReCV mouse hyperimmune sera. Mice were maintained at
the animal care facility of Cincinnati Children’s Hospital Medical Center (CCHMC), and all animal procedures complied with National Institutes of Health guidelines. Our animal protocol was approved by the Institutional Animal Care and Use Committee of CCHMC (Animal Welfare Assurance Number A3108-01). Hyperimmune sera were generated against one representative strain of each genotype of G1 ReCVs, including the prototype TV (G1.1), FT285 (G1.2) and FT7 (G1.3). Three female 6- to 8-week-old BALB/c mice were injected intraperitoneally with 106–108 TCID50 of CsCl density gradientpurified ReCVs mixed with an equal volume of Sigma Adjuvant (Sigma Aldrich). Animals were boosted twice at 3-week intervals and anaesthetized/sacrificed via carbon dioxide (CO2) inhalation on day 14 after the last immunization. Serum was collected and stored in aliquots at 280 uC. RNA extraction and genome amplification. Viral RNA was
extracted with TRIzol LS (Life Technologies) according to the manufacturer’s instructions, resolved in 20 ml RNase- and DNase-free water and stored at 280 uC. cDNA was synthesized using the SuperScript III First-Strand Synthesis System (Life Technologies) according to the manufacturer’s protocol, with a lock-docking oligodT primer containing a 59 overhang [GGCCACGCGTCGACTAGTAC(T)12VN]. The 39 portions of ReCV genomes (~3 kb) were amplified with gene-specific primers designed based on partial RNA-dependent RNA polymerase (RdRp) sequences described in our previous study (Farkas et al., 2010a) and the 59 overhang primer, GGCCACGCGTCGACTAGTAC, using the GoTaq Green Master Mix (Promega). DNA sequencing. PCR products were excised from 1.5 % agarose
gels, recovered using the Wizard SV Gel and PCR-clean up system, and cloned into pGEM-T vector according to the manufacturer’s protocols (Promega). Positive clones were identified using PCR. Plasmid DNA was extracted from small-scale bacterial cultures using the Wizard Plus SV Miniprep DNA-purification system (Promega), according to the manufacturer’s instructions, and sequenced using M13 forward and reverse and gene-specific primers (primer walking) with the chain-termination method on an ABI PRISM 3730 DNA Analyser (Applied Biosystems). Each sample was sequenced in both directions from at least two independent clones. Consensus sequences 1476
(CPE)-based virus neutralization assay was performed with heatinactivated (56 uC, 30 min) serum samples, as described previously (Farkas et al., 2010a, b, 2012). Briefly, twofold serial dilutions of hyperimmune antisera were titrated against 100 TCID50 of TV, FT285 and FT7 ReCVs. The virus/serum mix was incubated for 1 h at 37 uC and transferred to duplicate wells of 96-well tissue culture plates seeded with 104 LLC-MK2 cells per well. Virus and mock-inoculated control wells were included. Plates were stained with crystal violet at 72 h post-inoculation. By this time, all cells in the virus control wells were rounded and detached. End titrations were repeated in several experiments. End titres were calculated as mean values based on the highest dilution at which the cell monolayer was .50 % intact in both wells. This dilution against the homologous ReCV strain was taken as one VNU. Similar to entero-, rhino- and rotavirus serotyping (Committee on Enteroviruses, 1962; Kapikian et al., 1967; Wyatt et al., 1982), a 20-fold or greater difference in VN titres was required in both directions for establishing a ReCV serotype. Serotyping plaque assay. Due to the limited volume of the mouse hyperimmune sera (~0.5 ml/animal), serotyping of the 10 G1 ReCV isolates was performed in a cross-neutralization plaque assay with ~50 and ~100 p.f.u. of virus and 20 VNUs of each hyperimmune serum. A similar method was previously used for the serotyping of several viral agents, including avian infectious bronchitis viruses (Cowen & Hitchner 1975) and vesiviruses (Smith et al., 1998). Briefly, serum/ virus mix and controls were incubated for 1 h at 37 uC and adsorbed to LLC-MK2 monolayers in six-well tissue culture plates (Corning Life Sciences) for 1 h at 37 uC. Plates were washed twice with PBS and overlaid with culture medium containing 1.2 % Avicel (FMC BioPolymer). After 72 h of incubation, plates were stained with crystal violet and plaques counted. Each assay was repeated at least twice. ReCV strains with ¢80 % of plaque reduction were considered to be neutralized. Strains neutralized by a single specific antiserum were grouped as a serotype. Virus neutralization epitope prediction. Computational predic-
tion of linear B cell epitopes was performed with BepiPred 1.0 (Larsen et al., 2006). Three-dimensional structures of ReCV capsid proteins were generated using the Phyre 2 server (Kelley & Sternberg, 2009). Discontinuous B-cell epitopes were predicted with the DiscoTope 2.0 server (Kringelum et al., 2012). The predicted B-cell epitopes were evaluated based on analyses of multiple sequence alignments and cross-neutralization results (regions with serotype-specific differences) and the predicted three-dimensional structures (internal or surface-exposed regions) of ReCVs. Statistical analysis Statistical significance of cross-VN titres among
the ReCV isolates was evaluated with Fisher’s exact test. Journal of General Virology 95
Relationship between ReCV geno- and serotypes
ACKNOWLEDGEMENTS
Huang, P. W., Farkas, T., Zhong, W., Tan, M., Thornton, S., Morrow, A. L. & Jiang, X. (2005). Norovirus and histo-blood group antigens:
We thank Dr Peter Dickie (CCHMC) for critical reading of the manuscript. This study was supported by the Infectious Diseases Scholar Fund of CCHMC. A percentage of salary support (10%) was provided by NIH grant P01HD13021 to T. F.
demonstration of a wide spectrum of strain specificities and classification of two major binding groups among multiple binding patterns. J Virol 79, 6714–6722.
REFERENCES Bok, K., Parra, G. I., Mitra, T., Abente, E., Shaver, C. K., Boon, D., Engle, R., Yu, C., Kapikian, A. Z. & other authors (2011). Chimpanzees as an
animal model for human norovirus infection and vaccine development. Proc Natl Acad Sci U S A 108, 325–330. Choi, J. M., Hutson, A. M., Estes, M. K. & Prasad, B. V. (2008). Atomic
resolution structural characterization of recognition of histo-blood group antigens by Norwalk virus. Proc Natl Acad Sci U S A 105, 9175– 9180. Committee on Enteroviruses (1962). Classification of human
enteroviruses. Virology 16, 501–504. Cowen, B. S. & Hitchner, S. B. (1975). Serotyping of avian infectious
bronchitis viruses by the virus-neutralization test. Avian Dis 19, 583– 595. Debbink, K., Lindesmith, L. C., Donaldson, E. F. & Baric, R. S. (2012).
Norovirus immunity and the great escape. PLoS Pathog 8, e1002921. DiCaprio, E., Ma, Y., Purgianto, A., Hughes, J. & Li, J. (2012).
Internalization and dissemination of human norovirus and animal caliciviruses in hydroponically grown romaine lettuce. Appl Environ Microbiol 78, 6143–6152. Farkas, T., Sestak, K., Wei, C. & Jiang, X. (2008). Characterization of
a rhesus monkey calicivirus representing a new genus of Caliciviridae. J Virol 82, 5408–5416. Farkas, T., Cross, R. W., Hargitt, E., III, Lerche, N. W., Morrow, A. L. & Sestak, K. (2010a). Genetic diversity and histo-blood group antigen
interactions of rhesus enteric caliciviruses. J Virol 84, 8617–8625. Farkas, T., Dufour, J., Jiang, X. & Sestak, K. (2010b). Detection of
norovirus-, sapovirus- and rhesus enteric calicivirus-specific antibodies in captive juvenile macaques. J Gen Virol 91, 734–738. Farkas, T., Falkenstein, K. P., Bohm, R. P., Pecotte, J. & Sestak, K. (2012). High incidence of rhesus enteric calicivirus infections and
diarrhea in captive juvenile macaques: a likely association. J Med Primatol 41, 325–328. Frenck, R., Bernstein, D. I., Xia, M., Huang, P., Zhong, W., Parker, S., Dickey, M., McNeal, M. & Jiang, X. (2012). Predicting susceptibility to
norovirus GII.4 by use of a challenge model involving humans. J Infect Dis 206, 1386–1393. Green, K. Y. (2013). Caliciviridae: the noroviruses. In Fields Virology,
Kapikian, A. Z., Conan, R. M., Hamparian, V. V., Chanock, R. M., Chapple, P. J., Dick, E. C., Fenters, J. D., Gwaltney, J. M. & Hamre, D. (1967). Rhinoviruses: a numbering system. Nature 213, 761–763. Kapikian, A. Z., Wyatt, R. G., Dolin, R., Thornhill, T. S., Kalica, A. R. & Chanock, R. M. (1972). Visualization by immune electron microscopy
of a 27-nm particle associated with acute infectious nonbacterial gastroenteritis. J Virol 10, 1075–1081. Katpally, U., Wobus, C. E., Dryden, K., Virgin, H. W., IV & Smith, T. J. (2008). Structure of antibody-neutralized murine norovirus and
unexpected differences from viruslike particles. J Virol 82, 2079–2088. Kelley, L. A. & Sternberg, M. J. E. (2009). Protein structure prediction
on the Web: a case study using the Phyre server. Nat Protoc 4, 363– 371. Kringelum, J. V., Lundegaard, C., Lund, O. & Nielsen, M. (2012).
Reliable B cell epitope predictions: impacts of method development and improved benchmarking. PLOS Comput Biol 8, e1002829. Kroneman, A., Vega, E., Vennema, H., Vinje´, J., White, P. A., Hansman, G., Green, K., Martella, V., Katayama, K. & Koopmans, M. (2013).
Proposal for a unified norovirus nomenclature and genotyping. Arch Virol 158, 2059–2068. Larsen, J. E., Lund, O. & Nielsen, M. (2006). Improved method for predicting linear B-cell epitopes. Immunome Res 2, 2. Li, X., Ye, M., Neetoo, H., Golovan, S. & Chen, H. (2013). Pressure
inactivation of Tulane virus, a candidate surrogate for human norovirus and its potential application in food industry. Int J Food Microbiol 162, 37–42. Lindesmith, L., Moe, C., Marionneau, S., Ruvoen, N., Jiang, X., Lindblad, L., Stewart, P., LePendu, J. & Baric, R. S. (2003). Human
susceptibility and resistance to Norwalk virus infection. Nat Med 9, 548–553. Lindesmith, L. C., Beltramello, M., Donaldson, E. F., Corti, D., Swanstrom, J., Debbink, K., Lanzavecchia, A. & Baric, R. S. (2012).
Immunogenetic mechanisms driving norovirus GII.4 antigenic variation. PLoS Pathog 8, e1002705. Lindesmith, L. C., Costantini, V., Swanstrom, J., Debbink, K., Donaldson, E. F., Vinje´, J. & Baric, R. S. (2013). Emergence of a
norovirus GII.4 strain correlates with changes in evolving blockade epitopes. J Virol 87, 2803–2813. Lopman, B. A., Hall, A. J., Curns, A. T. & Parashar, U. D. (2011).
Increasing rates of gastroenteritis hospital discharges in US adults and the contribution of norovirus, 1996–2007. Clin Infect Dis 52, 466–474.
6th edn, vol. 1, pp 582–608. Edited by D. M. Knipe & P. M. Howley. Philadelphia: Lippincott Williams & Wilkins.
McFadden, N., Bailey, D., Carrara, G., Benson, A., Chaudhry, Y., Shortland, A., Heeney, J., Yarovinsky, F., Simmonds, P. & other authors (2011). Norovirus regulation of the innate immune response
Hall, A. J., Curns, A. T., McDonald, L. C., Parashar, U. D. & Lopman, B. A. (2012). The roles of Clostridium difficile and norovirus among
and apoptosis occurs via the product of the alternative open reading frame 4. PLoS Pathog 7, e1002413.
gastroenteritis-associated deaths in the United States, 1999–2007. Clin Infect Dis 55, 216–223.
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
Handley, S. A., Thackray, L. B., Zhao, G., Presti, R., Miller, A. D., Droit, L., Abbink, P., Maxfield, L. F., Kambal, A. & other authors (2012).
genotype 2 bovine noroviruses are antigenically distinct but share a crossreactive epitope with human noroviruses. J Clin Microbiol 44, 992–998.
Pathogenic simian immunodeficiency virus infection is associated with expansion of the enteric virome. Cell 151, 253–266.
Parker, T. D., Kitamoto, N., Tanaka, T., Hutson, A. M. & Estes, M. K. (2005). Identification of Genogroup I and Genogroup II broadly
Hirneisen, K. A. & Kniel, K. E. (2013a). Norovirus surrogate survival
reactive epitopes on the norovirus capsid. J Virol 79, 7402–7409.
on spinach during preharvest growth. Phytopathology 103, 389–394. Hirneisen, K. A. & Kniel, K. E. (2013b). Comparing human norovirus
Parra, G. I., Azure, J., Fischer, R., Bok, K., Sandoval-Jaime, C., Sosnovtsev, S. V., Sander, P. & Green, K. Y. (2013). Identification of
surrogates: murine norovirus and Tulane virus. J Food Prot 76, 139– 143.
a broadly cross-reactive epitope in the inner shell of the norovirus capsid. PLoS ONE 8, e67592.
http://vir.sgmjournals.org
1477
T. Farkas, C. Wong Ping Lun and B. Fey Richards, G. P. (2012). Critical review of norovirus surrogates in food safety research: rationale for considering volunteer studies. Food Environ Virol 4, 6–13.
human 1195.
Scallan, E., Hoekstra, R. M., Angulo, F. J., Tauxe, R. V., Widdowson, M. A., Roy, S. L., Jones, J. L. & Griffin, P. M. (2011). Foodborne illness acquired
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M. & Kumar, S. (2011). MEGA5: molecular evolutionary genetics analysis using
in the United States–major pathogens. Emerg Infect Dis 17, 7–15.
maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 2731–2739.
Sestak, K., Feely, S., Fey, B., Dufour, J., Hargitt, E., Alvarez, X., Pahar, B., Gregoricus, N., Vinje´, J. & Farkas, T. (2012). Experimental ino-
A. D. (2012). Calicivirus from novel Recovirus genogroup in
diarrhea,
Bangladesh.
Emerg
Infect
Dis
18,
1192–
Vashist, S., Bailey, D., Putics, A. & Goodfellow, I. (2009). Model
culation of juvenile rhesus macaques with primate enteric caliciviruses. PLoS ONE 7, e37973.
systems for the study of human norovirus Biology. Future Virol 4, 353–367.
Shanker, S., Choi, J. M., Sankaran, B., Atmar, R. L., Estes, M. K. & Prasad, B. V. (2011). Structural analysis of histo-blood group antigen
Wobus, C. E., Thackray, L. B. & Virgin, H. W., IV (2006). Murine
binding specificity in a norovirus GII.4 epidemic variant: implications for epochal evolution. J Virol 85, 8635–8645. Smith, A. W., Berry, E. S., Skilling, D. E., Barlough, J. E., Poet, S. E., Berke, T., Mead, J. & Matson, D. O. (1998). In vitro isolation and
characterization of a calicivirus causing a vesicular disease of the hands and feet. Clin Infect Dis 26, 434–439. Smits, S. L., Rahman, M., Schapendonk, C. M., van Leeuwen, M., Faruque, A. S., Haagmans, B. L., Endtz, H. P. & Osterhaus,
1478
norovirus: a model system to study norovirus biology and pathogenesis. J Virol 80, 5104–5112. Wyatt, R. G., Greenberg, H. B., James, W. D., Pittman, A. L., Kalica, A. R., Flores, J., Chanock, R. M. & Kapikian, A. Z. (1982). Definition of
human rotavirus serotypes by plaque reduction assay. Infect Immun 37, 110–115. 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.
Journal of General Virology 95
