Antigenic Diversity of Human Sapoviruses

Correspondence between virus antigenicity and capsid (VP1) genogrouping and genotyping is likely.

The SaV genomes are predicted to contain either 2 or 3 main open reading frames (ORF1-3). SaV ORF1 encodes for nonstructural proteins and the major capsid protein, and ORF2 (VP2) and ORF3 encode proteins of yet unknown functions. On the basis of VP1 nucleotide sequences, SaVs have been divided into 5 genogroups (GI-GV), of which GI, GII, GIV, and GV strains infect humans and GIII strains infect porcine species (8). SaV genogroups can be further subdivided into genotypes. Recently, we identifi ed several recombinant SaV strains (8,9).
Human SaV and NoV strains are uncultivable, but expression of a recombinant subgenomic-like construct (i.e., VP1 to the end of the genome) or VP1 alone in insect or mammalian cells results in the formation of viruslike particles (VLPs) that are morphologically similar to native SaV (10)(11)(12)(13)(14)(15)(16). However, production of VLPs of SaV remains diffi cult, usually only resulting in low yields of VLPs compared with norovirus (10,12,16,17). Cryoelectron microscopy and x-ray crystallography analyses of NoV VLPs identifi ed the shell (S) and protruding domains (subdomains P1-1, P1-2, and P2) (18). Also, Chen et al. described strictly and moderately conserved amino acid residues in the capsid protein among the 4 genera in the family Caliciviridae (13).
Previously, we reported that SaV GI/1 (strain Mc114) and GV/1 (strain NK24) were antigenically distinct (5,10). More recently, we discovered that SaV GI/5 (strain Yo-kote1) VLPs were antigenically distinct from SaV GI/1 Mc114 and GV/1 NK24 VLPs (19). Other than these few studies, little is known about the genetic and antigenic relationships among the 4 human SaV genogroups. For classifi cation of NoV, distinct genotypes have been defi ned as having bootstrap values >950 (VP1 sequences); at least 14 GI and 17 GII genotypes have been identifi ed (20). For SaV, genogroups have only been vaguely defi ned, mostly because 2 of them (GIV/1 and GV/1) were only recently identifi ed, few sequences exist in the database, and antigenic relationships among all genogroups are unknown. In addition, genetic recombination was only recently discovered and appears to be common within the genus Sapovirus.
The purpose of this study was to examine cross-reactivities among the 4 human SaV genogroups and compare results with those of genetic analysis. For this purpose, 2 other SaV strains, GII/3 Syd53 and GIV/1 Syd3, were expressed and antisera were produced against their purifi ed VLPs. A total of 5 SaV strains (GI/1 Mc114, GI/5 Yokote1, GII/3 Syd53, GIV/1 Syd3, and GV/1 NK24) that include all 4 human genogroups and 2 GI genotypes were compared. Our results show that SaV genogroups were antigenically distinct and corresponded with results of genetic classifi cation on the basis of full-length VP1 nucleotide sequences. Proper genetic classifi cation of SaV strains is required, and a consensus of genogroups and genotypes that represent genetically and antigenically diverse strains, which include recombinant SaV strains, should be established to avoid confl icting grouping.

Specimens
Virus-positive stool specimens were collected from several sources. SaV strain Mc114 (GenBank accession no. AY237422) was isolated from an infant hospitalized with acute gastroenteritis in Chiang Mai, Thailand, in 2001 (21). SaV strain NK24 (AY646856) was isolated from an infant with gastroenteritis in Nong Khai, Thailand, in 2003 (22). SaV strain Yokote1 was isolated from an outbreak of gastroenteritis at a kindergarten in Yokote City, Japan, in 2006 (19). SaV strains Syd53 and Syd3 were isolated from infants hospitalized with acute gastroenteritis in Sydney, New South Wales, Australia, in 2001 (23). NoV strain Osaka659 was isolated from an outbreak of gastroenteritis in Japan, in 2006 (unpub. data). RNA extraction and RT-PCR were performed as described (24).

Sequence Analysis
Nucleotide sequences were determined by using the Terminator Cycle Sequence Kit version 3.1 and an ABI 3130 sequencer (both from Applied Biosytems, Boston, MA, USA). Nucleotide sequences were aligned with Clust-alX (www.embl.de/∼chenna/clustal/darwin) and the distances were calculated by the Kimura 2-parameter method (24). Phylogenetic trees with bootstrap analysis from 1,000 replicas were generated by the neighbor-joining method as described (20). Amino acid VP1 secondary structure predictions were made by using PSIPRED secondary structural prediction software (25).

Expression of Viruslike Particles
For the expression of VP1 in insect cells, all SaV constructs were designed to begin from the predicted VP1 start AUG codon and included the ORF2 and poly(A) sequences. SaV strains Syd53 and Syd3 were cloned as described (10) for strains SaV Mc114, NK24, and Yokote1 according to the protocol of the Baculovirus Expression System using Gateway Technology (Invitrogen, Carlsbad, CA, USA). Briefl y, strains Syd53 and Syd3 were amplifi ed with specific sense primers Syd53attb1 (5′-GGGGACAAGTTTGTA CAAAAAAGCAGGCTTCGAAGGAGATAGAACCAT GGAGGGTGTGTCCCACCCAGA-3′) and Syd3attb1 (5′-G G G G A C A A G T T T G T A C A A A A A A G C A G G CTTCGAAGGAGATAGAACCATGGAGGGCAATGG CCTACCCCAGGCTG-3′) and antisense primer TX30SXN (5′-GACTAGTTCTAGATCGCGAGCGGCCGCCCT TTTTTTTTTTTTTTTTTTTTTTTTTTTTT-3′). PCR fragments were purifi ed after electrophoresis on a 1.0% agarose gel. Fragments were cloned into donor vector pDONR201 (Invitrogen) and transferred into a baculovirus transfer vector pDEST8 (Invitrogen).
The recombinant pDEST8 was purifi ed and used to transform Escherichia coli DH10Bac-competent cells (Invitrogen), which produced recombinant bacmids (baculovirus shuttle vectors) containing the VP1 gene. Recombinant bacmids were then transfected into Sf9 cells (Riken Cell Bank, Ibaraki, Japan), and recombinant baculoviruses were isolated. Recombinant baculoviruses were used to infect ≈3 × 10 6 confl uent Tn5 cells (Invitrogen) at a multiplicity of infection of 5-10 in 1.5 mL of Ex-Cell 405 medium (JRH Biosciences, Lenexa, KS, USA), and the infected cells were incubated at 26°C. The culture medium was harvested 5-6 d postinfection, centrifuged for 10 min at 3,000× g, and further centrifuged for 30 min at 10,000× g. VLPs were concentrated by ultracentrifugation for 2 h at 45,000 rpm at 4°C (Beckman TLA-55 rotor; Beckman Coulter, Fullerton, CA, USA), and resuspended in 30 μL of Grace's medium. Samples were examined for VLP formation by electron microscopy as described (10), and large-scale production of VLPs was performed as described (24).

Antibody Production
Hyperimmune sera to newly expressed VLPs of SaV (Syd53 and Syd3) were prepared in rabbits and guinea pigs. The fi rst subcutaneous injection was performed with purifi ed VLPs (≈10 μg) in Freund complete adjuvant. After 3 weeks, the animals received 1 booster injection (intravenously in rabbits and subcutaneously in guinea pigs) of 10 μg of VLPs without adjuvant. Blood was collected from the animals 1 week after their last booster injection.

Antibody ELISA
Cross-reactivities among antiserum samples against SaV were examined by using an antibody ELISA with hyperimmune rabbit antibodies against VLPs. Briefl y, wells of 96-well microtiter plates (Maxisorp; Nunc, Roskilde, Denmark) were each coated with 100 μL of purifi ed VLPs (≈1.0 μg/ mL in carbonate-bicarbonate buffer, pH 9.6) (Sigma, St. Louis, MO, USA) and incubated overnight at 4°C. Wells were washed twice with phosphate-buffered saline (PBS) containing 0.1% Tween 20 (PBS-T) and blocked with PBS containing 5% skim milk (PBS-SM) for 1 h at room temperature. Wells were then washed 4 times with PBS-T, 100 μL of 2-fold-diluted hyperimmune rabbit antibodies from an initial concentration of 1:500 in PBS-T-SM was added to each well, and the plates were incubated for 1 h at 37°C. Wells were then washed 4 times with PBS-T, and 100 μL of a 1:1,000 dilution of horseradish peroxidase-conjugated goat antirabbit immunoglobulin G diluted in PBS-T-SM was added to each well. Plates were incubated for 1 h at 37°C. Wells were then washed 4 times with PBS-T, and 100 μL of substrate (o-phenylenediamine) and H 2 O 2 were added to each well, and the plates were left in the dark for 30 min at room temperature. The reaction was stopped by the addition of 50 μL of 2N H 2 SO 4 to each well, and the absorbance was measured at 492 nm (A 492 ). The optical density (OD) cutoff point was determined to be 0.15, which was equal to 3 times the mean OD of preimmune serum (5).

Antigen ELISA
Cross-reactivities among VLPs were also examined by using an antigen ELISA. Briefl y, wells were coated with 100 μL of a 1:8,000 dilution of hyperimmune rabbit antiserum diluted in PBS (except for Syd3, for which a 1:3,000 dilution was used), and the plates were incubated overnight at 4°C. Wells were washed 4 times with PBS-T and blocked with PBS-SM for 1 h at room temperature. Wells were then washed 4 times with PBS-T, 100 μL of VLPs (≈1.0 μg/mL in carbonate-bicarbonate buffer, pH 9.6) (Sigma) was added to duplicate hyperimmune rabbit wells, and the plates were incubated for 1 h at 37°C. Wells were then washed 4 times with PBS-T, 100 μL of a 1:8,000 dilution of hyperimmune guinea pig antibody diluted in PBS-T-SM was added to each well (except for Syd3, which used a 1:3,000 dilution), and the plates were incubated for 1 h at 37°C. Wells were washed 4 times with PBS-T, and 100 μL of a 1:1,000 dilution of horseradish peroxidase-conjugated rabbit antiguinea pig immunoglobulin G diluted in PBS-T-SM was added to each well. The plates were then processed as described above. On the basis of our previous study, a specimen with an A 492 (P -N) >0.1 and a P/N ratio >1.34 (where P is hyperimmune antiserum and N is preimmune antiserum) was considered signifi cantly positive (4).

Sequence Analysis
The sequence of the 3′ end of the genome (≈2,600 nt), i.e., VP1 to poly(A), for the newly expressed SaV strains (Syd53 and Syd3) was determined. Genetic analysis was performed with only complete VP1 sequences, which included sequences from our epidemiologic studies and other sequences available on the database (Figure 1). Five SaV GI and 6 GII genotypes were observed, but only 1 genotype for SaV GIV and 1 for GV was found. This result suggests that SaV GI and GII strains were more genetically diverse, prevalent, or more virulent than SaV GIV and GV strains. However, because the SaV GIV and GV strains were only recently detected (26,27), this result may refl ect only the specifi city and sensitivity of the detection methods used.

Expression of VP1
We previously expressed SaV GI/1 Mc114, GI/5 Yo-kote1, and GV/1 NK24 in insect cells, which resulted in the formation of VLPs morphologically similar to native SaV (5,10). In this study, we newly expressed SaV GII/3 Syd53 and GIV/1 Syd3 in insect cells to analyze the cross-reactivity among all human SaV genogroups. SaV GII/3 Syd53 and GIV/1 Syd3 successfully formed VLPs with a diameter of 41 to 46 nm and were morphologically similar to native SaV ( Figure 2). Hyperimmune sera against these purifi ed VLPs were prepared in rabbits and guinea pigs.

Amino Acid Alignment and Secondary Structure Prediction
An amino acid alignment of the 5 SaV VP1 sequences (GI/1 Mc114, GI/5 Yokote1, GII/3 Syd53, GIV Syd3, and GV NK24) showed that the shell domain contained more conserved residues than the predicted P domains ( Figure  4). However, SaV GI/1 Mc114 and GI/5 Yokote1 shared more conserved continuous residues in the predicted P2 subdomain than other genogroups. The NoV P2 subdomain is thought to contain the determinants of strain specifi city, cell binding, and antigenicity. For example, monoclonal antibodies that recognize regions in the P2 subdomain inhibit binding of NoV VLPs to cells (28,29). In a recent study, we analyzed cross-reactivities among 26 different NoV VLPs (6 GI and 12 GII genotypes) (30)    broad-range cross-reactivities for several NoV antisera.
Our results suggested that these cross-reactivities were due to conserved amino acid residues located outside the P2 domain. Conversely, secondary structure predictions made by using PSIPRED secondary structural prediction software showed that helix structures could also infl uence the crossreactivity among the NoV VLPs. In the current study, we determined the secondary structure of the 5 SaV VP1 amino acid sequences. Overall, SaV VP1 structures appear to be similar (online Appendix Figure, available from www. cdc.gov/EID/content/13/10/1519-appG.htm). The location of 3 helix structures in the shell domain and 1 helix structure in the C-terminal region were nearly identical for the 5 SaV VP1 sequences. Only SaV GV/1 NK24 was predicted to have a single helix structure in the P2 subdomain. These results suggested that the amino acid sequence, particularly the P2 subdomain, plays a major role in determining crossreactivity among SaV strains. However, additional studies, including high-resolution VLP structural analysis, are needed.

Discussion
In this study, we analyzed genetic and antigenic relationships for 4 human SaV genogroups (GI, GII, GIV, and GV). We observed weak 2-way cross-reactivity with SaV GI/1 Mc114 and GI/5 Yokote1 antisera against the heterologous GI/5 Yokote1 and GI/1 Mc114 VLPs, respectively, by using an antibody ELISA. However, when we used an antigen ELISA, we found that GI/1 Mc114 and GI/5 Yo-kote1 VLPs were antigenically distinct. These weak crossreactivities identifi ed by using the antibody ELISA may have been infl uenced by several factors, including unfolded VLPs on the microtiter plates at the high pH (carbonatebicarbonate buffer, pH 9.6) (31) or conserved continuous residues outside the predicted P2 domain. Therefore, these 2 SaV genotypes (GI/1 and GI/5) are for the most part antigenically distinct. Likewise, we found that the 4 human SaV genogroups were antigenically distinct in the antigen ELISA. To our knowledge, these new fi ndings provide the fi rst evidence that SaV antigenicity corresponded well with VP1 genogrouping and genotyping.   (13). Asterisks indicate conserved amino acids among these 3 VP1 sequences.