Journal of Medical Virology 67:282–288 (2002)

Epidemiology of Human Sapporo-like Calicivirusesin the South West of England: MolecularCharacterisation of a Genetically Distinct Isolate

Samantha Robinson,1 Ian N. Clarke,1 I. Barry Vipond,2 E. Owen Caul,2 and Paul R. Lambden1*1Division of Molecular Microbiology and Infection, University Medical School, Southampton General Hospital,Southampton, United Kingdom2Virus Laboratory, Bristol PHLS, Bristol, United Kingdom

Human enteric caliciviruses have been assignedto two distinct genera: the Norwalk-like viruses(NLVs) and the Sapporo-like viruses (SLVs).During a 3-year surveillance of gastroenteritis inthe South West of England during November1997–2000, a total of 27 clinical samples contain-ingSLVswere collected. PCRamplicons coveringa region of the RNA polymerase gene were ob-tained from 18 of the SLV samples. Sequenceanalysis of the PCR products indicated that theSLV isolates could be assigned to one of the twomajor genetic groups represented by Sapporoand London/92 caliciviruses. One of these iso-lates belonging to the London/92 group (Bristol/98) was subjected to a complete genome se-quence analysis. The full genomic sequence ofthe Bristol/98 isolate was determined from RNAextracted froma single stool sample and consistsof 7490 nucleotides, excluding the poly(A) tail.The genome is organised into two open readingframes (ORFs), similar to that of Manchester SLValthough the small ORF overlapping the regionencoding the capsid protein observed in Man-chester SLV is absent in Bristol/98 SLV. Thepolyprotein (ORF1) of Bristol/98 SLV consists of2,280 amino acids and, as observed in all SLVs,the structural protein is encoded in frame andcontiguous with the 30 terminus of the ORF1.Phylogenetic studies based on complete capsidsequences and genome arrangements within theSLVs indicate that the human enteric viruseswithin the ‘‘Sapporo-like’’ virus clade should bedivided into two distinct genetic groups analo-gous to the assignment of the Norwalk-likeviruses. J. Med. Virol. 67:282–288, 2002.� 2002 Wiley-Liss, Inc.

KEY WORDS: gastroenteritis; calicivirus;sequence; phylogeny; epide-miology

INTRODUCTION

Members of the Caliciviridae family have beenidentified as the causative agents of a wide range ofdiseases, both human and veterinary. The Caliciviridaefamily has recently been classified into four genera[Green et al., 2000]: Vesivirus, Lagovirus, ‘‘Sapporo-likeviruses’’ (SLVs), and ‘‘Norwalk-like viruses’’ (NLVs).

Human enteric caliciviruses are included in the‘‘Sapporo-like’’ and ‘‘Norwalk-like’’ genera. The ‘‘Nor-walk-like viruses,’’ previously termed small roundstructured viruses (SRSVs) [Caul and Appleton, 1982],have an amorphous structure and a ragged edge, asviewed by negative-stain electron microscopy. In con-trast, SLVs display the characteristic ‘‘Star of David’’morphology typical of the Lagoviruses and Vesiviruses.In addition to the differing virion morphologies, thetwo human calicivirus genera differ epidemiologically;NLVs are associated with sporadic outbreaks of gas-troenteritis in all ages, with a distinct seasonalitylinked to the winter months [Zahorsky, 1929; Wrightet al., 1998; Pang et al., 1999], whereas the SLVs occurless frequently, are associated predominantly withpaediatric infection, and are endemic with no season-ality reported. A recent study of Mexican childrenshowed that NLVs and SLVs co-circulated in thecommunity and that SLVs were present in 40% ofdiarrhoeic and nondiarrheic stools, suggesting thatthere is a high rate of asymptomatic infection inMexican children [Farkas et al., 2000].

The EMBL accession number of the sequence reported in thisarticle is AJ249939.

*Correspondence to: Paul R. Lambden, Mail Point 814, Divisionof Molecular Microbiology and Infection, University MedicalSchool, Southampton General Hospital, Southampton SO166YD, UK. E-mail: prl1@soton.ac.uk

Accepted 21 November 2001

DOI 10.1002/jmv.2219

Published online in Wiley InterScience(www.interscience.wiley.com)

� 2002 WILEY-LISS, INC.

At present, the human enteric caliciviruses cannot becultured in vitro. Molecular studies have thereforerelied on the availability of stool samples collectedduring outbreaks, or from volunteers. Currently, com-plete nucleotide sequences are available for only threeNLVs; Southampton NLV [Lambden et al., 1993, 1995],Lordsdale NLV [Dingle et al., 1995], and the prototypeNorwalk virus [Jiang et al., 1993] and one SLV,Manchester [Liu et al., 1995, 1997]. Complete nucleo-tide sequences are also available for two animal entericcaliciviruses: Jena virus [Liu et al., 1999a], which isphylogenetically related to the NLVs, and a porcineenteric calicivirus (PEC), which is related to humanSLVs [Guo et al., 1999].

The human enteric caliciviruses have single-stran-ded (ss) positive sense RNA genomes of 7,400–7,700nucleotides. The nonstructural proteins of calicivirusesare encoded in a large polyprotein, which undergoesproteolytic cleavage by a 3C-like protease to producethe functional proteins. A small basic protein is encodedat the 30 end of all calicivirus genomes. Calicivirusesencode a single major structural protein of 58–62 kDa.In the NLVs, this protein is encoded in a separate openreading frame (ORF), ORF2, to the nonstructuralproteins (ORF1), whereas in the SLVs it is encoded aspart of a large polyprotein (ORF1) contiguous with thenonstructural proteins. A further small ORF over-lapping the 50 terminus of the capsid protein is predictedin some SLVs, but it is uncertain whether this ORF isfunctional in vivo [Liu et al., 1995; Noel et al., 1997].

Recent studies of SLVs have resulted in the identi-fication of Parkville virus [Noel et al., 1997] andHouston 90 virus [Berke et al., 1997; Jiang et al.,1997]. Parkville SLV showed 80% amino acid identityto the Manchester SLV capsid protein and also contain-ed the predicted overlapping ORF. A distinct ‘‘Sapporo-like’’ calicivirus, London/92 SLV [Berke et al., 1997;Jiang et al., 1997], shared only 40% amino acid identitywith the capsid of Manchester SLV and this isolate didnot encode the predicted overlapping ORF. A compar-ison study of a region of the 3D RNA polymerase hastentatively divided the SLVs into three genetic groups[Jiang et al., 1997; Farkas et al., 2000]. Recently, astudy of SLV isolates from the United Kingdom,Sweden, and the Netherlands suggested that theremay be further genetic diversity among SLVs [Vinjeet al., 2000].

This report describes the incidence and molecularepidemiology of SLV infection in the South West ofEngland and suggests that human viruses in the SLVgenus can be divided into two major genogroupsanalogous to the current classification of the Norwalk-like viruses.

MATERIALS AND METHODS

Sapporo-like Viruses

Approximately 20,000 stool samples were screenedduring a study of gastroenteritis outbreaks in the SouthWest of England between November 1997 and Novem-

ber 2000: the presence of SLVs was detected by electronmicroscopy. Twenty-seven samples from this studycontained calicivirus particles with the typical ‘‘Starof David’’ morphology, and all but one were from chil-dren <3 years of age. Polymerase chain reaction (PCR)products were obtained from 18 of the 27 samples.

PCR and Sequencing

Oligonucleotide primers used in this study are listedin Table I. Stool samples containing candidate SLVswere used for synthesis of cDNA, as described by Liuet al. [1999a]. Viral RNA was extracted directly fromstool samples using a Trizol (Gibco-BRL)/chloroformprocedure [Chomczynski and Sacchi, 1987] and theRNA concentrated by adsorption onto a silica matrix(RNaid, Bio 101). cDNA was synthesised from thepurified RNA in a 50-ml reaction by priming with a 30

degenerate primer T25VN and using reverse transcrip-tase (Superscript, Gibco-BRL), according to manufac-turer’s instructions. Single-stranded cDNA (sscDNA)was produced by removal of the RNA strand by alkalinehydrolysis. In a 50-ml reaction, 1 M NaOH (5 ml) wasadded directly to the newly synthesised cDNA andincubated at 658C for 30 min to destroy the RNA; thesscDNAwas then neutralised by the addition of 5 ml 1MHCl and 8 ml of 1 M Tris-HCl pH 7.5. This cDNA wasused directly as a template to amplify a 216-bp fragmentof the RNA-dependent RNA polymerase gene, using thedegenerate inosine-containing primer pair GL1 andPK1. These primers were based, respectively, on theconserved GLPSGM and PTAADK motifs obtained byaligning the RNA polymerase genes from all publishedSLVs. Amplification was performed using a Perkin-Elmer-Cetus 9600 thermal cycler and consisted of 35cycles of 948C for 20 sec, 468C for 20 sec, and 728C for20 sec, using Bio-x-Act polymerase (Bioline, London,UK) in a 50-ml volume, containing 1� Optibuffer(Bioline) 2 mM MgCl2 and 200 mM each dATP, dGTP,dTTP, and dCTP. Amplicons (216 bp) were obtainedfrom 18 of the 27 samples. The amplicons were purifiedfrom the reaction mix using a PCR preps column(Qiagen) and the nucleotide sequence determined

TABLE I. Oligonucleotide Primers Used for PCRAmplification and Determination of the Genomic 50 Terminus

Sequence

Primer Sequence 50!30

T25VN T25-A/G/C A/G/C/TLinker N7 TAGTACATAGTGGATCCAGCT(N)7GL1 GGICTICCITCIGGIATGPK1 YTTRTCIGCIGCIGTIGGBV3 6651TGGCDGTBTTCAATGTKGAAAC6672

BV4 6672GGTTTCMACATTGAAVACHGCCAT6649

BV5 4553GTTGTCAACTCTCTTAAC4570

BV49 727GTTAAATTGTGTGCGCAT710

BV53 1633CACACCAATCATTGTCTC1616

BV64 536CTAACGGCAGTTCCTTAAA518

BV65 642GATGTACACTTTACCAGG625

BV67 269CAGGGTTAATGTCCACTGG251

BV68 319TGGGAATTGTTCATGTGG302

BV93 142CATGGGCGTACATGTCCTG125

Sapporo-like Virus 283

directly with primers GL1 and PK1 using a model ABI373 or ABI 377 automated DNA Sequencer (AppliedBiosystems). One sample (Bristol/98 sample 80238)showed 94% sequence identity to London/92 SLV andsufficient sample was available for a complete genomeanalysis and permit full-length comparison of theLondon/92-like group of SLVs with Manchester SLVand porcine enteric calicivirus.

The amino acid sequences of the 30 regions of Man-chester SLV and London/92 were aligned and degen-erate primers BV3 and BV4 were built as forward andreverse primers based on the conserved MAVFNETmotif contained within the capsid region. Primer pairBV5 and BV4 were used to produce a 2,116-bp ampliconand BV3 with T25VN produced an 800-bp amplicon. Theamplicons were purified from the reaction mix usingQiaquick PCR purification columns and sequenceddirectly using the PCR primers in each case. Thisprimary sequence was used to build additional specificprimers that were used to sequence directly the initialPCR product. Comparisons with London/92 SLV sug-gested that Bristol/98 SLVwas very similar and furtherprimers were designed, based on the published London/92 sequence (accession no. U67858). This allowed therapid production of a consensus sequence for this 2,963-bp region of Bristol/98 SLV.

Sequencing of the 50 Regionof the Bristol/98 SLV Genome

Oligonucleotide primers are listed in Table I.Sequence upstream of the GL1, PK1 amplified regionwas obtained by the random PCRmethod [Dingle et al.,1995] adapted from the procedures of Grothues et al.[1993] and Froussard [1992]. Single-stranded cDNAwas generated using specific primers and then con-verted into double-stranded (ds) cDNA using a randomprimer Linker N7 and Klenow polymerase. The ran-domly primed viral dscDNA was amplified in twosuccessive nested rounds of PCR using specific primersand a primer based on the linker component of LinkerN7. Bio-x-Act was used in all rounds of PCR for 35cycles of 948C for 20 sec, 508C for 20 sec, and 728C for1 min. Reaction products were purified and sequenceddirectly using further specific nested primers. Thesequence obtained was used to design new oligonucleo-tides for further rounds of random primer extensiontowards the 50 terminus.Using this procedure, sequencewas obtained to within 500 nucleotides of the authenticgenomic 50 terminus.

The remaining 500 bp of Bristol/98 SLV, includingthe genomic 50 terminus, was defined by homopolymertailing and PCR, using a commercial kit (50 RACE;Gibco-BRL) and a number of defined custom-made oli-gonucleotide primers. cDNA was synthesised from thepurified viral RNA and a single primer BV53. ThecDNA was tailed with A residues at the 30 terminususing terminal deoxynucleotide transferase (TdT) andamplified by two successive rounds of PCR using Bio-x-Act polymerase with primer T25VN and the specificprimers for nested PCR BV49 and BV65: a further

nested primer BV64 was used to sequence the amp-licon. This initial 50 RACE gave sequence to within200 bp of the authentic 50 terminus when compared toManchester SLV. A further round of 50 RACE wasperformed using primer BV65 for producing the cDNA.The cDNA was tailed with either C or A residues at the30 terminus using TdT and again amplified by twosuccessive rounds of PCR using Bio-x-Act polymerasewith either T25VN or the 50 RACE-abridged anchorprimer/abridged universal primer (Gibco-BRL) toge-ther with the specific primers BV64 and BV68 fornested PCR. Primer BV67 was used to sequence theamplicon and primer BV93 was synthesised, based onthe sequence obtained from BV67. Sequencing of theamplicons obtained using nested primers BV64/68against T25VN and the abridged anchor primer, withBV93, enabled the authentic 50 genomic terminus ofBristol/98 SLV to be defined.

The sequence upstream of GL1 was confirmed bysynthesising a number of specific primers that wereused to produce a series of overlapping amplicons enab-ling direct sequencing of both strands of amplifiedcDNA and a consensus sequence to be obtained.

Sequencing was carried out with Applied Biosystemsmodels 373A and 377 automated sequencers using Taqcycle dideoxy terminator chemistry. Computer ana-lyses of the sequence data were carried out usingLasergene software (DNASTAR, Madison, WI). Oligo-nucleotides were obtained from Cruachem.

Expression of RecombinantCapsid in Insect Cells

A 2.4-kb PCR amplicon, representing the 30 ter-minus of Bristol/98 SLV (nt 5090–7515), including thepolyadenylate tail, was amplified directly from Bristol/98 SLV cDNA using specific primers BV18 (5090CACC-TACGAAGCGTGGTT5110) and BvcapA (AAGC[T]26C-CAGAGGGTCATGGC7476) and cloned into pFastbac(Gibco-BRL), producing BV98cap, this constructincluded 83 nucleotides of viral sequence in front ofthe initiator codon of the capsid coding region.Transfection of the bacmid DNA into Sf9 cells wascarried out using a liposome-mediated transfectionsystem and 3-ml volumes of Bacmid DNA. The cellswere incubated at 288C for 72 hr and the cells andsupernatants harvested and analysed by sodiumdodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE).

Phylogenetic Analysis

Analyses were performed as previously described(Liu et al., 1999a), using the PHYLIP software[Felsenstein, 1993].

RESULTS AND DISCUSSION

Molecular Epidemiology of SLV Infection

During a 3-year study of gastroenteritis in the SouthWest of England 27 stool samples were identified

284 Robinson et al.

containing caliciviruses showing the typical morphol-ogy. This detection rate is similar to that reported in aseparate study of SLV occurrence in Sweden, TheNetherlands, and the United Kingdom [Vinje et al.,2000]. A similar survey in Melbourne Australia ofchildren admitted to hospital suffering from acutegastroenteritis showed that the incidence of SLV infec-tion was significantly less than that due to NLVs[Kirkwood and Bishop, 2001]. Our survey showed thatbased on sequence analysis of a 216-bp fragment of theRNA polymerase region of the SLV genome, SLVisolates could be divided into three groups. Ninesamples could be ascribed to the Sapporo/Manchestergroup; seven samples were assigned to the London/92group; two samples were identical to the Parkville SLV.One of the Parkville-like samples was detected in thestool of a 40-year-old patient, whereas all other sampleswere obtained from infants of < 3 years of age.Interestingly, the original Parkville isolate was asso-ciated with an outbreak of gastroenteritis affectingadults [Noel et al., 1997], suggesting that a subset ofSLVs may be associated with disease in all age groups.

Sequence Analyses of the Bristol/98 SLV Genome

Bristol/98 SLV has a genome of 7490nt, excluding thepolyadenylate tail (Fig. 1A). The first translationinitiation codon for a large uninterrupted ORF islocated at nucleotide 14 and occurs within a favourablecontext (GGTATGGCT) for translation initiation by theribosome scanning model [Kozak, 1991]. At the 30 end ofthe genome, there are 140 untranslated nucleotidespreceding the polyadenylate tail. Sequence analysis ofBristol/98 SLV indicated two potential ORFs with thecapsid in frame and contiguous with the nonstructuralproteins as observed for Manchester SLV [Liu et al.,1995]. The additional small ORF seen in ManchesterSLV overlapping the 50 terminal region of the capsidsequence was not present in the Bristol strain, as notedin London/92 SLV [Berke et al., 1997; Jiang et al.,

1997]. The large ORF1 polyprotein of Bristol/98 SLVhas a coding potential of 2280 amino acids and includesthe conserved motifs typical of the 2C NTPase (GPP-GIGKT), 3C protease (GDCG) and 3D RNA dependentRNA polymerase (YGDD and GLPSG).

A second ORF of 495 nucleotides coding for a smallprotein of 164 amino acids is predicted in frame �1 andis situated at the 30 end of the genome. This small ORFhas a counterpart in all caliciviruses and the predictedprotein for Bristol/98 SLV is basic and hydrophilic and,as is the case with Manchester SLV, does not containcysteine residues.

The Bristol/98 SLV 50 terminus has 13 nucleotidesbefore the first translation initiation codon, which isone more than the 12 reported for Manchester SLV [Liuet al., 1997] and the 30 non-coding region of Bristol/98SLV is 61 nt longer than Manchester SLV. Both Man-chester and Bristol/98 SLVs have a�1 frameshift at theORF1/ORF2 junction; however, Manchester SLV ORF2overlaps ORF1 by 4 nt, whereas in Bristol/98 SLV theoverlap is 1 nt.

Manchester [Liu et al., 1997] and Bristol/98 SLV(Fig. 1B) contain a deletion in the area surroundingthe putative start of the subgenomic message of 3 and4 nucleotides, respectively, as compared with the 50

Fig. 1. Diagrammatic representation of reading frame usage in Bristol/98 SLV. A: The nucleotide co-ordinates of the translation products are numbered on the open boxes. B: Alignment of the repeat motifsat the 50 genomic terminus and the predicted 50 terminus of the putative subgenomic RNA.

Fig. 2. Sequence comparisons between Bristol/98 SLV, ManchesterSLV and PEC of the 50 termini of the (A) genomic and (B) putativesubgenomic RNAs. Asterisks indicate identical nucleotides andhyphens are used to adjust alignment spacing.

Sapporo-like Virus 285

terminus of the genomic RNA. This is not the case,however, in PEC, where the start of the putative sub-genomic message shares 11/12 nucleotides (92%) withthe genomic 50 terminus. The genomic 50 termini ofManchester and Bristol/98 SLV share 87.5% sequenceidentity in the first 15–16 nucleotides, whereas PECshares only 8/15 and 8/16 nucleotides with ManchesterSLV and Bristol/98 SLV, respectively (Fig. 2A). Theputative start of the subgenomic RNA of Manchesterand Bristol/98 SLV share 91.6% sequence identity (11/12 nucleotides) over the 12 bases surrounding the ATGcodon, only 9/12 nucleotides are identical betweenBristol/98 SLV and PEC (Fig. 2B).

Alignment of Manchester SLV and Bristol/98 SLVsequences revealed only 61.3% sequence identity acrossthe entire genome and 62% amino acid identity acrossthe large polyprotein. Alignment of Bristol/98 SLV and

PEC showed only 42.2% amino acid identity across thepolyprotein. Comparisons of the amino acid sequencesof the large ORF1 polyproteins of Bristol/98 SLV andManchester SLV (X86560) revealed many conservedregions, including the known conserved motifs of theNTPase, protease and RNA dependent RNA polymer-ase, with highest amino acid identities observed in theNTPase (71.5%) and RNA polymerase (72%) regions(Table II). Amino acid sequence comparisons (Table II)of Bristol/98 SLV with Southampton NLV (L07418),Lordsdale NLV (X86557), RHDV (M67473), FCV F9(M86379) and PEC (AF182760) showed around 25–55%identity in the RNA polymerase region and around 25to 46% identity in the NTPase region. The putativecleavage positions of the Bristol/98 non-structuralproteins were determined by alignment of ManchesterSLV and Bristol/98 SLV sequences with Southampton

TABLE II. Amino Acid Sequence Identity Between Bristol/98 SLV Nonstructural Proteinsand Those of Other Caliciviruses*

RegionAminoacids

ManchesterSLV (%) PEC (%)

SouthamptonNLV (%)

LordsdaleNLV (%)

FCV(%)

RHDV(%)

N-terminus 1–325 57 27 NS NS 22 NSNTPase 326–641 71.5 46 28 31 40 443A 642–940 58 27 NS NS 23 NSVPg 941–1055 72 57 37.5 35 40 38Protease 1056–1246 65 48 NS NS 24 24Polymerase 1247–1722 72 55.5 27 28 38 32

NS, no significant identity at < 20%.*The functional regions of the Bristol/98 SLV polyprotein were arbitrarily defined by alignments with thepolyproteins of caliciviruses for which cleavage sites have been experimentally determined.

Fig. 3. Unrooted phylogenetic tree constructed from amino acidsequences of complete calicivirus structural capsid proteins. The treeshows the relationship of Bristol/98 SLV to the other ‘‘Sapporo-like’’caliciviruses and to other members of the Caliciviridae. The twogenogroups of Sapporo-like viruses and Norwalk-like viruses arehighlighted with shaded ellipses. Accession numbers (in parentheses)for caliciviruses are as follows: Manchester SLV (X86560), DCC(Houston DCC) SLV (U95643), Parkville SLV (U73124), Houston/90

SLV (U95644), London/92 SLV (U95645), Stockholm/97 (AF 194182),porcine enteric calicivirus (AF182760), Southampton NLV (L07418),Norwalk virus (M87661), Desert Shield NLV (U04469), bovine Jenavirus (AJ011099), Lordsdale NLV (X86557), Hawaii NLV (U07611),EBHSV (Z69620), RHDV (M67473), Pan-1 (AF091736), SMSV4(P36285), SMSV1 (P36284), and FCVF9 (M86379). Phylogeneticanalyses were performed with PHYLIP software [Felsenstein, 1993].

286 Robinson et al.

NLV and RHDV, where cleavage positions and sequen-ces have been determined [Wirblich et al., 1995, 1996;Liu et al., 1996, 1999b; Konig et al., 1998].

Expression of the Bristol/98 SLV StructuralProtein in Insect Cells

Expression of the capsid genes of Sapporo virus[Numata et al., 1997] and more recently Houston 90SLV [Jiang et al., 1999] and PEC [Guo et al., 2001] ininsect cells using recombinant baculovirus vectorsresulted in the production of virus-like particles (VLPs)that could be detected in the cell culture supernatant.Expression of recombinant Bristol/98 ORF2 led to theproduction of a 60-kDa protein corresponding to thepredicted size of the capsid (data not shown). However,VLPs were not observed in the supernatants of Sf9 cellsor High Five (Trichoplusia ni) cells infected at anmoi¼10 (data not shown). The reason for the lack ofspontaneous assembly into VLPs is not clear but it hasbeen suggested that the sequence upstream of theinitiator codon may influence expression and formationof VLPs [Jiang et al., 1999].

Phylogenetic Analysis of Bristol/98 SLV

The region of sequence encoding the entire structuralprotein of Bristol/98 SLV was compared, as an aminoacid sequence, to the published amino acid sequences ofLondon/92 SLV [Jiang et al., 1997] and ManchesterSLV [Liu et al., 1995, 1997]. The entire capsid of Bristol/98 SLV shared 45% amino acid identity with that ofManchester SLV and 94.3% identity to London/92 SLVover the 560aa region. Three regions of extensivedifference were shown in comparing the capsid ofBristol/98 SLV with London/92. These differences couldbe attributable to frameshift sequencing errors inLondon/92 and suggest that these two viruses wouldshare an even higher identity across the capsid. The560-amino acid sequence of Bristol/98 SLV was furthercompared with the published capsid amino acidsequences of other ‘‘Sapporo-like viruses,’’ the porcineenteric calicivirus, ‘‘Norwalk-like viruses,’’ includingJena virus, Lagoviruses, and Vesiviruses. An unrootedphylogenetic tree (Fig. 3) was constructed as describedpreviously [Liu et al., 1999a].

Based on complete capsid sequences, both the porcineand bovine enteric caliciviruses possibly form distinctgenogroups within the SLV and NLV genera, respec-tively. The results suggest the presence of two distinctgenogroups of human ‘‘Sapporo-like’’ viruses withParkville SLV and Houston/90 SLV clustered with thegroup I Sapporo/Manchester SLV genogroup, andLondon/92 SLV and Bristol/98 SLV forming a distinctgenogroup (group II) analogous to the situation withenteric caliciviruses assigned to the NLV genus.

ACKNOWLEDGMENT

S.R. thanks the Medical Research Council for aresearch studentship.

REFERENCES

Berke T, Golding B, Jiang X, Cubitt DW, Wolfaardt M, Smith AW,Matson DO. 1997. Phylogenetic analysis of the caliciviruses. J MedVirol 52:419–424.

Caul EO, Appleton H. 1982. The electron microscopical and physicalcharacteristics of small round human fecal viruses: an interimscheme for classification. J Med Virol 9:257–265.

Chomczynski P, Sacchi N. 1987. Single-step method of RNA isolationby acid guanidinium thiocyanate-phenol-chloroform extraction.Analytical Biochemistry 162:156–159.

Dingle KE, Lambden PR, Caul EO, Clarke IN. 1995. Human entericCaliciviridae: the complete genome sequence and expression ofvirus-like particles from a genetic group II small round structuredvirus. J Gen Virol 76:2349–2355.

Farkas T, Jiang X, Guerrero ML, Zhong WM, Wilton N, Berke T,Matson DO, Pickering LK, Ruiz-Palacios G. 2000. Prevalence andgenetic diversity of human caliciviruses (HuCVs) in Mexicanchildren. J Med Virol 62:217–223.

Felsenstein J. 1993. PHYLIP (Phylogeny Inference Package) version3.5c. Distributed by author.

Froussard P. 1992. A random-PCR method (rPCR) to construct wholecDNA library from low amounts of RNA. Nucleic Acids Res 20: 2900.

Green KY, Ando T, Balayan MS, Berke T, Clarke IN, Estes MK,Matson DO, Nakata S, Niell JD, Studdert MJ, Thiel H-J. 2000.Taxonomy of the caliciviruses. J Infect Dis 181:S322–S330.

Grothues D, Cantor CR, Smith CL. 1993. PCR amplification ofmegabase DNA with tagged random primers (T-PCR). NucleicAcids Res 21:1321–1322.

Guo M, Chang KO, Hardy ME, Zhang O, Parwani AV, Saif LJ. 1999.Molecular characterisation of a porcine enteric calicivirus geneti-cally related to Sapporo-like human caliciviruses. J Virol 73:9625–9631.

Guo M, Qian Y, Chang KO, Saif LJ. 2001. Expression and self-assembly in baculovirus of porcine enteric calicivirus capsids intovirus-like particles and their use in an enzyme-linked immuno-sorbent assay for antibody detection in swine. J Clin Microbiol39:1487–1493.

Jiang X, Wang M, Wang K, Estes MK. 1993. Sequence and genomicorganization of Norwalk virus. Virology 195:51–61.

Jiang X, Cubitt WD, Berke T, Zhong W, Dai XM, Nakata S, PickeringLK, Matson DO. 1997. Sapporo-like human caliciviruses aregenetically and antigenically diverse. Arch Virol 142:1813–1827.

Jiang X, Zhong WM, Kaplan M, Pickering LK, Matson DO. 1999.Expression and characterization of Sapporo-like human caliciviruscapsid proteins in baculovirus. J Virol Methods 78:81–91.

Kirkwood CD, Bishop RF. 2001. Molecular detection of humancalicivirus in young children hospitalized with acute gastroenter-itis in Melbourne, Australia, during 1999. J Clin Microbiol 39:2722–2724.

Konig M, Thiel H-J, Meyers G. 1998. Detection of viral proteins afterinfection of cultured hepatocytes with rabbit hemorrhagic diseasevirus. J Virol 72:4492–4497.

Kozak M. 1991. Structural features in eukaryotic mRNAs thatmodulate the initiation of translation. J Biol Chem 266:19867–19870.

Lambden PR, Caul EO, Ashley CR, Clarke IN. 1993. Sequence andgenome organization of a human small round-structured (Nor-walk-like) virus. Science 259:516–519.

Lambden PR, Liu BL, Clarke IN. 1995. A conserved sequence motif atthe 50 terminus of the Southampton virus genome is characteristicof the Caliciviridae. Virus Genes 10:149–152.

Liu BL, Clarke IN, Caul EO, Lambden PR. 1995. Humanenteric caliciviruses have a unique genome structure and aredistinct from the Norwalk-like viruses. Arch Virol 140:1345–1356.

Liu BL, Clarke IN, Lambden PR. 1996. Polyprotein processing inSouthampton virus: identification of 3C-like protease cleavagesites by in vitro mutagenesis. J Virol 70:2605–2610.

Liu BL, Clarke IN, Caul EO, Lambden PR. 1997. The genomic 50

terminus of Manchester Calicivirus. Virus Genes 15:25–28.

Liu BL, Lambden PR, Gunther H, Otto P, Elschner M, ClarkeIN. 1999a. Molecular characterization of a bovine enteric calici-virus: relationship to the Norwalk-like viruses. J Virol 73:819–825.

Sapporo-like Virus 287

Liu BL, Viljoen GJ, Clarke IN, Lambden PR. 1999b. Identification offurther proteolytic cleavage sites in the Southampton caliciviruspolyprotein by expression of the viral protease in E. coli. J GenVirol 80:291–296.

Noel JS, Liu BL, Humphrey CD, Rodriguez EM, Lambden PR, ClarkeIN, Dwyer DM, Ando T, Glass RIM, Monroe SS. 1997. Parkvillevirus: a novel genetic variant of human Calicivirus in the Sapporovirus clade, associated with an outbreak of gastroenteritis inadults. J Med Virol 52:173–178.

Numata K, Hardy ME, Nakata S, Chiba S, Estes MK. 1997. Molecularcharacterization of morphologically typical human calicivirusSapporo. Arch Virol 142:1537–1552.

Pang XL, Joensuu J, Vesikari T, 1999. Human calicivirus-associatedsporadic gastroenteritis in Finnish children less than two years of

age followed prospectively during a rotavirus vaccine trial. PediatrInfect Dis J 18:420–426.

Vinje J, Deijl H, van der Heide R, Lewis D, Hedlund K-O. Svensson L.Koopmans MPG. 2000. Molecular detection and epidemiology ofSapporo-like viruses. J Clin Microbiol 38:530–536.

Wirblich C, Sibilia M, Boniotti MB, Rossi C, Thiel H-J, Meyers G.1995. 3C-like protease of rabbit hemorrhagic disease virus:identification of cleavage sites in the ORF1 polyprotein andanalysis of cleavage specificity. J Virol 69:7159–7168.

Wirblich C, Thiel H-J, Meyers G. 1996. Genetic map of the calicivirusrabbit hemorrhagic disease virus as deduced from in vitrotranslation studies. J Virol 70:7974–7983.

Wright PJ, Gunesekere IC, Doultree JC, Marshall JA. 1998. Smallround-structured (Norwalk-Like) viruses and classical humancaliciviruses in southeastern Australia, 1980–1996. J Med Virol55:312–320.

Zahorsky J. 1929. Hyperemesis hiemis or the winter vomiting disease.Arch Pediatr 46:391–395.

288 Robinson et al.