The isolation and characterization of a Norwalk virus specific cDNA

The isolation and characterization of a Norwalk

virus-specific cDNA.

S M Matsui, … , L S Oshiro, G R Reyes

J Clin Invest.

1991;

87(4)

:1456-1461.

https://doi.org/10.1172/JCI115152

.

Norwalk virus, an important cause of epidemic, acute, nonbacterial gastroenteritis in adults

and children, has eluded adaptation to tissue culture, the development of an animal model,

and molecular cloning. In this study, a portion of the Norwalk viral genome encoding an

immunoreactive region was cloned from very small quantities of infected stool using

sequence-independent single primer amplification. Six overlapping complementary DNA

(cDNA) clones were isolated by immunologic screening. The expressed recombinant

protein from a representative clone reacted with six of seven high titer. Norwalk-specific,

postinfection sera but not with corresponding preinfection sera. Nucleic acid sequence for

all clones defined a single open reading frame contiguous with the lambda gt11-expressed

beta-galactosidase protein. Only oligonucleotide probes specific for the positive strand

(defined by the open reading frame) hybridized to an RNaseA-sensitive, DNaseI-resistant

nucleic acid sequence extracted from Norwalk-infected stool. Furthermore, RNA extracted

from serial postinfection, but not preinfection, stools from three of five volunteers hybridized

to a Norwalk virus cDNA probe. Clone-specific oligonucleotide probes hybridized with

cesium chloride gradient fractions containing purified Norwalk virion. In conclusion, an

antigenic, protein-coding region of the Norwalk virus genome has been identified. This

epitope has potential utility in future sero- and molecular epidemiologic studies of Norwalk

viral gastroenteritis.

Research Article

Rapid Publication

The Isolation and Characterization

of

a

Norwalk Virus-specific

cDNA

Suzanne M. Matsui,* Jungsuh P.Kim,*HarryB.Greenberg,* WanchuangSu,*Qiming Sun,t

Philip C.Johnson,' Herbert L.DuPont,$Lyndon S. Oshiro,11and_GregoryR. Reyest

*DepartmentsofMedicine(Gastroenterology) and MicrobiologyandImmunology,StanfordUniversity,Stanford, California94305; and

VeteransAdministrationMedicalCenter,Palo Alto, California 94304;

WMolecular

VirologyDepartment, Genelabs, Incorporated,

RedwoodCity,California94063;ODepartmentofMedicine, TheUniversityofTexasHealthScience CenteratHouston, Houston, Texas 77030;andI"CaliforniaDepartmentofHealthServices, Viral and RickettsialDisease Laboratory, Berkeley, California94704

Abstract

Norwalk virus, an important causeofepidemic,acute,

nonbac-terial

gastroenteritis

in adults andchildren,haseluded

adapta-tion totissue culture,thedevelopmentof an animalmodel, and molecularcloning.Inthis study,aportion ofthe Norwalkviral

genome encoding animmunoreactive region was cloned from

very smallquantities of infectedstoolusing

sequence-indepen-dentsingle primer

amplification.

Sixoverlapping

complemen-tary DNA (cDNA) clones were isolated by

immunologic

screening.Theexpressedrecombinantproteinfrom a

represen-tativeclonereactedwithsixofsevenhigh

titer,

Norwalk-speci-fic,

postinfection

sera but not withcorresponding

preinfection

sera.Nucleic acidsequenceforallclones definedasingleopen

readingframecontiguous withtheXgtll-expressed

ft-galacto-sidaseprotein. Only oligonucleotide probes specificfor the

posi-tivestrand

(defined

bythe open

reading frame) hybridized

to an

RNaseA-sensitive, DNaseI-resistantnucleic acid sequence

ex-tracted from Norwalk-infected stool. Furthermore, RNA

ex-tracted from serial

postinfection,

but not

preinfection,

stools

from three of five volunteers hybridized to a Norwalk virus cDNAprobe.

Clone-specific oligonucleotide probes hybridized

with cesium chloride

gradient

fractions

containing

purified

Norwalkvirion. In

conclusion,

an

antigenic, protein-coding

re-gion ofthe Norwalk virus genome has been identified. This

epitopehas

potential utility

infuturesero-and molecular

epide-miologic studies of Norwalk viral

gastroenteritis. (J.

Clin. In-vest. 1991.

87:1456-1461.)

Keywords:

sequence-independent

single. primer amplification

*

gastroenteritis

virus * viral

diar-rhea*cloning

Introduction

NorwalkandNorwalk-likeviruses

comprise

agroup of

serologi-cally

diverse,

small round-structuredviruses

implicated

in common sourceoutbreaks of

gastroenteritis (1, 2).

It is

esti-Addressreprint requeststoDr.SuzanneM.Matsui, Division

ofGastro-enterology, S069,StanfordUniversitySchoolofMedicine, Stanford,

CA 94305.

Receivedfor publication 25September1990and in revisedform

27December1990.

matedthatup to 65%ofacute,nonbacterial gastroenteritisin

the United States is attributabletothese agents(3). Norwalk

virus, namedaftera 1968 outbreakofgastroenteritisat aschool

in Norwalk, Ohio,(4) was thefirst of this heterogeneousgroup

of viruses to beidentified by immune electron microscopy

(IEM)1

(5). Italone appears to cause at least one-third ofall casesof

gastroenteritis

inU.S.

epidemics

(6). While this illness

typically

lastsonly1-2days, its

impact

ondayslostfrom work

orschoolmay be

substantial

(7).

Thestudyoftheseviruseshasbeenrestrictedby the inabil-ity to cultivate these agents in vitro and lack ofan animal

model.Muchofwhatisknownabouttheimmuneresponse to

this virus has comefromhuman volunteer studies (8-10)or

epidemiologic

surveys (6). Positive identification ofthese

vi-rusesrelieson IEM, atechniquenotwell-suitedtolarge-scale screening. Seroconversion and detection of viral antigens in fecal

specimens

are also usedto establish the association of these viruses with

episodes

ofacute

gastroenteritis (11,

12). Unequivocal

diagnosis,

however, requires theuseofpaired hu-man sera(pre-and

postinfection).

The molecular characteriza-tion of thisagent haslikewise beendifficult,becausethevirus is shed inverysmallamounts instool.

Wereport thecloning ofanimmunoreactive regionof the Norwalkvirusfrom infected fecal material. Sequence-indepen-dent

single

primer amplification (SISPA)

(13;

Reyes, G. R., and J. P. Kim, manuscript submitted for publication), a

re-cently described and

generally applicable technique

of

nonse-lective cDNA amplification,wasused toovercomethe restric-tionsimposed by theverylimitedamountofvolunteerclinical material containing viral particles andthelack of existing

nu-cleotidesequenceinformation. Thespecificity of these clones for Norwalk viruswasshown by(a) the immunoreactivity of theexpressed

peptide

for only

postinfection

serafrom several volunteers; (b)thehybridization ofaclone-specificprobe

exclu-sivelyto RNAextractedfrom Norwalk

antigen-positive

volun-teerstools;and(c)theconvergenceofthe

clone-specific

probe

hybridization

peak andthe Norwalkviral antigen peakin a

cesium chloride

gradient

fraction ofbuoyant

density

1.37

g/ml.

Methods

Virusandantisera. LTI-8, theeighth diarrheal stoolspecimen col-lected from volunteerLT(14)

after

oral inoculation with infectious

1.Abbreviations used in thispaper:IEM, immuneelectron micros-copy;PCR,polymerasechainreaction;SISPA,sequence-independent single primer amplification.

1456 Matsui, Kim, Greenberg, Su, Sun, Johnson, DuPont, Oshiro,andReyes J.Clin.Invest.

©The AmericanSocietyforClinicalInvestigation,Inc.

0021-9738/91/04/1456/06 $2.00

fecalfiltrate 8FIIa(8), wasusedto construct acDNAlibrary (see be-low). Occasional Norwalk virusparticlesweredemonstratedbyIEMin this specimen and in LT1-3,thethirdpostinfectionstoolfrom thesame volunteer. Particle countswereestimated between 10'and 106 viral particles per g of feces.LTO,thepreinfectionstoolfrom thisvolunteer,

was IEMnegative.

Serial stool samples from other volunteers(I-V)inoculated with 8FIlawere usedfor dot blothybridization(seebelow).Ineach case, viral antigen was not detected byRIA(I1)and/orELISA(12) in prein-fection specimens, butwaspresent inone or morepostinfection

speci-mensfromsymptomatic volunteers. Viralparticlesweredetectedby

IEMonly inpooledspecimensfrom volunteer IV.

Sevenpairsof volunteerserademonstratedasignificantrise in titer between the pre- and thepostinfection specimens byELISA.Apairof serathat showednochangeintiter(1:100)between pre- and

postinfec-tion sampleswasusedasthenegativecontrol.

Viruspurification andnucleicacid extraction. 7.5 g ofspecimen

LTl-8waspartiallypurifiedusingfluorocarbon extraction and

pellet-ing througha sucrosesuspensionontoacesium chloridecushion,as

describedpreviously (15).Viralantigenwasmonitoredateach stepby

ELISA.LTOcontrol stool andLTI-3wereprocessed in tandem with LTi-8.

Serial stool samples from volunteers I-V,includinguninfected

con-trols,werepartiallypurified byamodifiedprocedure.500 Mlofa10% fecalsuspension(in PBS)wascombined with 125

1l

of 40%

polyethyl-eneglycol(PEG,molwt8,000)andprecipitatedovernightat40C.The

suspensionwaspelletedat10,000g for 20 min.

Nucleicacid was extracted from half of thepartially purifiedstool byaone-stepguanidinium/phenolextractionprocedure(16). Aliquots

of nucleic acid were tested forsensitivityto treatmentwith DNase-free, RNaseAorRNase-free, DNaseI.Theremainder of the partially puri-fied stoolwaslayeredonto a 1.2-1.6 g/mldiscontinuouscesium chlo-ride gradient(15).The density and Norwalk antigenic activity (by ELISA; 12)of each gradient fraction was determined.

cDNAcloningandanalysis. -I0

,g

ofnucleic acid obtained from 1.5 g oforiginalfecal material was reverse transcribed into cDNA ( 17), usinganoligo(dT) primer. Double-stranded

linker-primeroligonucleo-tidesweredirectionally ligatedontothe cDNAamplified by SISPA(13;

Reyes,G. R., and J. P. Kim, submitted for publication) and then di-gested with EcoRI. Excess linkerswereremovedby passagethrough Sephacryl 300, beforeinsertingtheamplified cDNA into bacteriophage

Xgtl1.

Libraries were screenedimmunologically(18) with postinfection serum from volunteerDD(NorwalkELISAtiter>1:25,600). The serum waspreadsorbed with Xgtl1 and used at a 1:200 dilution for

immunoscreening. After three immunoscreenings, a purified plaque for each clonewasobtained and the specificity assessed usingDD preinfection sera (ELISA titer= _1:100).

cDNAfrom clones of interest was tested by Southern blot(19)or dotblothybridizationforsimilaritytohumanlymphocyteand

Esche-richiacoli strain 1088 DNA and to other enteric viruses investigated in ourlaboratories (NCDV rotavirus and astrovirus type 1). Direct se-quencecomparisons with the hepatitisEgenome (Tam, A. W., et al.,

manuscriptsubmittedfor publication) were also made.

Subcloningandsequencing. cDNA wassubclonedintoBluescript

SK+ (Stratagene Inc., La Jolla, CA) for sequencing (20).7-deaza-dGTP

was used in confirmatory sequencing reactions. Nucleic and amino acid sequences were compared with the

Genbank1>

database.

Dotblothybridization. Nucleic acidwas preparedfordot blotas recommended by the manufacturer of the apparatus (Schleicher and

Schuell,Keene, NH). Hybridization was carried out in 50% formamide and lx hybridization buffer (5X Denhardt's reagent, 5x SSC, 50 mM NaH2PO4, 1 mMsodiumpyrophosphate/Na2HPO4, 100

_jg/ml

sal-monspermDNA, 100

Mg/ml

ATP).Filters were subsequently washed in2xSSCatroom temperature, then 0.lxSSCand0.1%SDS at 65°C forIh.

Radiolabeledprobes. Probes were made using direct polymerase

chain reaction

_{(PCR) (21)}

and

primers

of known

_Xgtl

1 sequence

(5'

GGCAGACATGGCCTGCCCGG3' and 5' TCGACGGTTTCCA-TATGGOG 33.ThetypicalPCR cycle(denaturation, 940C, 30 s;

an-nealing, 500C, I min;extension, 720C, 2 min)wasrepeatedfor 30

cycles. TheNorwalk-specific amplified fragmentwasseparated from

Xgtl1 sequences by EcoRI digestion, purified by preparative 1.5% aga-rosegel electrophoresis, and radiolabeled by randompriming.

AnN49-specificfragmentwasamplified by PCR, using

oligonucleo-tideprimersshowninFig. 1 Aand modified PCRconditions(each

segment 30s). Radiolabeled, random-primed probesweresynthesized. Single-strandedoligonucleotide probesweremadeby 3'-end-label-ing N49-specific, synthetic, oligonucleotide primers with'y-[32P]ATP.

Hybridization was carried out in 30% formamide and I xhybridization

bufferat420C.Filterswerewashedin2xSSC at room temperature for 30 min.

Results

Our

cloning

strategywasdictated

by

the smallamountof

Nor-walk-infected,IEM-positive,fecalspecimen available.The

orig-inal stool contained I05-IO6 virion

particles

per gasestimated

by

IEMand 1.5gof

original

fecalmaterialwasused in cDNA

synthesis.

cDNAwassynthesizedfrompartially purifiedvirion andtheheterogeneous population nonselectively amplified by SISPA

(13;

Reyes, G.R.,and J. P.

Kim,

manuscriptsubmitted

for

publication).

The cDNA

library

wasconstructedin bacterio-phage Xgtl 1 forcloneidentification.

The first three rounds ofimmunoscreening were accom-plished usingDD

postinfection

serum.Inthefirst screening,41

reactive plaqueswereidentifiedfromatotal of 1.2 x 105 recom-binantphage.Aftertwo morerounds ofimmunoscreeningand

plaquepurification, afinal screening yieldedsix clones(N28,

N35,N40, N48, N49,and N51)thatwerechosenbyvirtue of

their

specific reactivity

with DD postinfection, but not DD

preinfection serum.

The initial characterization of these clones demonstrated

the absence ofhybridization to Southern blots (19) ofeither humanorE.coliDNA(datanotshown).Inaddition,the

repre-sentative N49 probe (see below) did not hybridize toNCDV rotavirus RNA orastrovirus type 1 cDNA(data notshown).

N49nucleotide and peptidesequences(seebelow)weredistinct

fromthefull length hepatitisEvirussequences(Tam,A.W.,et al., manuscript submitted for publication).

Two approaches establishedthese immunoreactive clones

as derived from only Norwalk-infected but notpreinfection fecal specimens. First, we demonstrated that probes from all cloneshybridized onlytotheSISPA-expandedcDNA popula-tion from LT1-8 andnot to asimilarlyprepared LTO preinfec-tion cDNA specimen (data not shown). In a second test, a

[32P]-labeled random-primed probe derived from clone N49 hybridized only to RNA extracted from two Norwalk

IEM-positive postinfection specimens, LT1-8 and LT1-3. This probedidnothybridize with nucleic acid extracted from

nega-tivecontrols, LTOorNCDVrotavirus (datanotshown).

The relationship of the six immunoreactive clones toone

another was nextexamined byhybridization and sequence

analysis. The clones ranged in size from 172 bp (N40) to 380 bp (N35)(Fig. 1 A) and all six cross-hybridized (data not shown). Eachofthe cloneshadasingleopenreading frame,contiguous

with the

_{fl-galactosidase}

reading frame in Xgtl 1. Regions of overlap were verified by sequencing and are shown in Fig. 1. Five different start sites were represented among six isolated clones. Two regions ofdisparity between the N35 reference sequenceandthe other clone sequences were found. The

C CAA GGG GGC TTT GAT AAC CAA GGGAAT ACC CCG TTT GGT AAG GGT GTG ATG AAG 55

1

-C ACC ACC ATA TTA ATC CAG GCT GTA GCC TTG ACG ATG GAG AGA 109

___ -- - 55

1 .52

CAG GAT GAG TTC CAA CTC CAGGGG CCT ACG TAT GAC TTT GAT ACT GAC AGA GTA 163

_ -_--- 10 9

.-_---106

1 .46

1 40

GCT GCG TTC ACG AGG ATG GCC CGA GCC AAC GGG TTG GGT CTC ATA TCC ATG GCC 217

.-- __- -163

.-- _ -- --- --- 160

_ --- ₁₀₀

________________ _ _ ~~~~~~~~~~94

TCC CTA GGC AAA AAG CTA CGC AGT GTC ACC ACT ATT GAA GGA TTA&AAG A-ATGT 271

.- 217

.- 214

.___--- - 154

.-.148___ ___ ___- ___-

-CTA TCA GGC TAT AAA ATA TCA AAA TGC AGT ATA CAA TGG CAG TCA AGG GTG TAC 325

.- 271

.-_--_-268

.-_.208___- ___ ___- ___-

--- --- --- --A- AA- --- 172

ATT ATA GAA TCA GAT GGT GCC AGT GTA CAA ATC AAA GAA GAC AAG CAA GCT TTG 379

--- --- --- --- --- --- --- --- --- --- --- --- A-- A-A - 314

--- --- --- --- --- --- --- --- --- --- --- --- A-- A-A - 311

--- --- --- --- --- --- --- --- --- --- --- --- A-- A-A - 251

A 380

Q G G F D N Q G N T P F G K GV M K P T T I N R L L I Q A V A L T M E R Q 37

1 19

1 18

D E F Q L QG P T Y D F D T D R V A A F T R M A R A N G LG L I S M A S L

G K K L R S V T T I E G L K N A L S G Y K I S K C S I Q W Q S R V Y I I E

- - - K K - 57

S D G A S V Q I K E D K Q A L 126

- - - K K 104

- - - K K 103

- - - K K 83

poly (A)tailwaspresent in these clones derived fromtheoligo

(dT)-primedlibrary. Thesesequenceswereuniqueatboth the nucleic and amino acid levels whencomparedtothe Gen-bank( database. Two of these clones,N49andN51, had

iden-74 Figure1.Thenucleotide (A) and

56 predictedamino acid(B)sequences

55 ofantibody-selectedNorwalk cDNA

35

33 Clones. The sequence of thelargest clone, N35, is presentedasthe

refer-encesequence.

Regions

ofoverlap 93 and identityareshownby dashed

92 _lines(-

_-).

_Nonidentical

nucleo-tides andamino acidsare specifi-callyindicated.N49-specificprimers areunderlined inA.These sequence dataareavailable from

EMBL/Gen-Bank/DDBJ under accession num-berM62825.

tical nucleotidesequences and N49was selectedfor closer

study.

Immunological screeningof N49 withanexpanded panel

ofseven pairedNorwalk volunteerseraprovidedfurther

sup-1458 Matsui, Kim, Greenberg, Su, Sun, Johnson, DuPont, Oshiro, andReyes

A

N35 1

N49/N51 N48 N28 N40

N35

N49/N51 N48

N28 N40

N35

N49/N51 N48

N28 N40

N35

N49/N51

N48 N28 N40

N35

N49/N51 N48

N28

N40 N35 N49/N51 N48 N28 N40 N35 N49/N51 N48 N28 N40

N35

N49/N51 N48 N28 N40

B N35 1

N49/N51 N48 N28 N40

N35

N49/N51 N48 N28

N40

N35 N49/N51 N48 N28

N40

N35 N49/N51 N48 N28

port that these clones expressedan immunogenic epitope of Norwalk virus. As shown in Table I, theexpressed product fromclone N49 reacted with six of the sevenhightiter, postin-fection sera. Only preinfection serum AT, which hadhigh preexisting levels of reactivityto Norwalkantigen by ELISA, alsoreacted with the N49 cloneproduct.No~other preinfection

sera or thenegativecontrolpairedsera(EM)reacted withthe N49 clone product.

Asasecond proofthat clone N49representedaportion of theNorwalk virusgenome,RNA was extracted from preinfec-tionandserialpostinfectionstools from five additional volun-teers who became ill afterreceiving 8FIIa.As shown inFig.2, the N49probehybridizedtonucleic acid fromtestspecimens 1-4and -5, III-4, -5, -6, and -7, and IV-8 and -9. These repre-sented the third through eighth stools collected from three

symptomaticvolunteers. In each case, N49hybridization

iden-tifiedapeak of viral sheddinginthese serialpostinfectionstools thatcontainedviralantigen.Hybridizationof thisprobe could notbedemonstratedinthespecimensof volunteers II andV

who had only one antigen-positive stool specimen each. No

hybridizationcould be demonstrated withnucleicacid isolated

frompreinfectionstools.

Furthermore, the clone N49-specific probe hybridized most prominently to sample LT1-8 cesium chloride density gradient fractions 7 through 10 (Fig. 3). Peak hybridization

signaland peakNorwalk antigen signal byELISAwereboth found in fraction 8. Buoyantdensityof this fractionwas 1.37 g/ml.Simultaneouslyprepared LTOcesium chloridegradient fractions showed neither hybridization with the N49-specific probe norNorwalkantigenic activity.

Preinfection

atool Postinfection.Antigen-Positive. Fecal Specimens

3 4 5..

1 3 4 5

0

0

O

*

0

1 4 5 6 7

III

V.

O

0

'

O

1 4 5 6 7 8

0

0

0

0-5'

0

9 10 12 14 16 17

V

0

0

1 5

Figure2. Hybridization of an N49-specific, radiolabeled probe to RNAfromserial, Norwalk volunteer, fecal specimens. Specimensare numberedin order of collection, with specimen Irepresenting the antigen-negative, preinfection stool for each. All postinfection fecal samples testedwereantigen positive. Viral particlesweredetectedby IEMonly in pooled, postinfection stools from volunteer IV (*). N49-specific hybridizationwasfound in nucleic acidextractsof stools from volunteers I (specimens 4 and 5), III (specimens 4, 5,6, and 7), and IV(specimens 8 and 9).

TableLSummaryofImmunologicReactivityofNorwalk Clone

N49with PairedSerafromInfected Volunteers

Reactivityofsera to Immunoreactivity of Serum Norwalk byELISA* cloneN49*

DDpreinfection 1:100

DDpostinfection 1:25,600 + ATpreinfection 1:1,600 + ATpostinfection 1:102,400 + SLpreinfection <1:100

SLpostinfection 1:6,400 + DFpreinfection 1:400

DFpostinfection 1:6,400 + JYpreinfection <1:100

JYpostinfection 1:25,600 + CSpreinfection <1:100

CSpostinfection 1:6,400 519preinfection 1:100

519postinfection 1:25,600 + EMpreinfection <1:100

EMpostinfection <1:100

* The Norwalk ELISA was performed as described previously

(12).

tAssayperformed by mixing plaque purified N49 with nonrecom-binant lambda gtII(internalnegative control) and plating at a 1: 1 ratio. Reactive plaques were detected after first antibody incubation ("serum")usinganalkaline phosphatase-conjugated anti-IgG second antibody. The absence (-) or presence (+) of immunoreactivity was determined by two independent readers. Selected reactive plaques wereconfirmed to contain the N49-specific insert by PCR (21).

1.42 E 1.40--

1.38-z

nU 1.36

-,l 1.34

-M 1.32

-1

HYBRIDIZATION WITH

CLONE N49 PROBE:

14-mr 12

P:W 10 ui>

LU Z

S2ij

zp 6

X O

c 4-zu

2-5 6 7 8 9 10 11 12

5 6 7 8 9 10 11 FRACTION NUMBER

1 2

Figure 3.Correlation of Norwalk antigenic activity with N49-specific, radiolabeled probe hybridization in cesium chloride gradient fractions of buoyantdensity1.31-1.40 g/ml. Peak antigenic activity and peak

hybridizationsignal were detected in fraction 8, buoyant density 1.37 g/ml.

.- AM&-

-0.40

0

Finally, we examined the nature of the extracted nucleic

acid from whichthe cDNA was synthesized. Asshown in Fig. 4 A, RNaseA treatment ofLT1-8andLT1-3nucleic acid extracts

eliminated hybridization with the N49 probe, while DNaseI treatmenthad no effect.Only single-strandedprobes of nega-tivepolarity (as defined by open reading frame analysis)

hybrid-izedtoLT1-8 nucleicacid (Fig. 4 B).

Discussion

We conclude that we have cloned animmunoreactive portion of the Norwalkvirusgenome. The isolatedclones were derived from nucleic acid extracted from partially purified fecal

mate-rialthat contained a very small number ofintact viral particles. TheSISPA technique(13;Reyes, G. R., and J.P. Kim,

manu-script submitted for publication)was used tosuccessfully

am-plify the LT1-8 cDNA, before the identification of a set of

overlapping immunoreactive clones specific to postinfection sera specimens. SISPA coupled toconventional cloning and

immunoscreening techniquesappears to be agenerally applica-ble strategy for identifying cDNA clones offastidious

in-fectious pathogens derived from limited amounts of starting

material.

The expressedrecombinantprotein of representative clone

N49wasrecognized specificallyby a panel ofhigh titer

postin-fectionserafromseveral volunteers, but not by

antibody-nega-A

2 3

LT-0

LT1 -3

LTt-8

0

0

0

0

0

0

0

000

_0

B

1 2

Figure4. The Norwalk genome iscomposedofsingle-strandedRNA ofpositive polarity.(A) TheN49-specific,radiolabeled probe failed tohybridizewith RNaseA-treated nucleic acid (column 1), but hy-bridized with bothDNaseI-treated(column 2) and untreated (column 3) nucleic acid from Norwalk-infectedstools(LTI-3 andLTI-8).(B) Anegativesense, N49 oligonucleotide probe (N49 residues 208to 227,Fig.1 A) hybridizedto RNAextractedfrom infected stool LT1-8,

indicatingthat the RNAis of positive polarity (spot 1).Apositive senseprobe (N49 residues71 to91,Fig. IA) didnothybridizewith

LT1-8RNA(spot 2).

tive preinfection sera or by paired sera without antibody to Norwalk virus (Table I). The lack ofreactivity ofthe N49 prod-uct with CS postinfection serum (ELISA titer = 1:6,400) is possibly due to antibodies in the CS specimen being directed to other Norwalk virusepitope(s).

The N49-specific cDNA probe specifically hybridized to nucleic acid from only the postinfection stools of three differ-ent volunteers (Fig. 2). The positive hybridization signals

correlated wellwiththeperiod ofsymptomatic illnessand anti-gen detection inthe stool by RIA (11) and/or ELISA (12). This

cDNA-derived probewas moresensitive than IEMfor

detect-ing smallquantities of virus in stool.Generally, the limit of detectionbyIEMinfeces isin the range of

105-106

particles per ml(5). RIA and/or ELISA, however, maybe able to detect viral antigens in the stool before and after the period of shedding of

intact viral particles.

Analysis ofLT1-8 cesium chloridedensitygradient frac-tions by ELISA and dot blot hybridization provided further support that theisolatedclones representportions ofthe Nor-walk viral genome (Fig. 3). Peak Norwalk antigenic activity

and peak clone N49-specific hybridization signalwere both observed infraction8. Thebuoyantdensity of fraction 8, 1.37

g/ml,is inthe rangeofthatwhichhasbeenreportedfor Nor-walk(15)andNorwalk-like viruses (2).

Theviralproteinstructureof Norwalk virus (15) and

Nor-walk-like viruses, such as Snow Mountain agent (22), most

closely resembles calicivirusesthathavea

single

structural

pro-tein ofmolecular mass62 kD (23).

Furthermore,

a one-way

serologic relationshipbetween caliciviruses and Norwalk virus has beendemonstrated byRIA(24).Theseobservationshave led to thehypothesisthat Norwalkvirusmay bea

plus-sense,

single-stranded

RNAviruswitha3'-end

poly (A) tail,

like

cali-civiruses. We demonstratehere that the Norwalk virus hasa

single-stranded

RNA genome of

positive

polarity (Fig. 3).

These findings lend supporttothe

speculation

that Norwalk

virusmay be related tocalciviruses (23),but further

character-izationwillbe

required

toconfirm this classification. An

au-thentic poly

(A)

tail was not identified in these clones. The most likely

explanation

for the 11 to

12-bp

stretches of

(A)

residuesatthe 3'end of clones

N49/N5

1,

N48, N28,

and

N40,

isthat the

(A)-rich regions

ofthe sequence

(Fig.

1

A)

servedas

the initiation sites ofcDNA

synthesis by oligo (dT).

Theidentified immunoreactiveclonescan nowbeusedto

obtain a complete setof

overlapping genomic

cDNA clones and nucleotide sequence of the Norwalk virus. Both will be

helpfulin

studying

Norwalkvirus and

establishing

its related-ness to otherviruses. Withthe PCR and sequence information now

available,

the

development

ofusefuland

highly

sensitive

diagnostic

reagents should be

relatively straightforward.

TheNorwalksequenceidentified inthesestudiesappearsto encodea common immunoreactive viral

epitope.

Itseems likelythatthisclonecanbe

employed

to

produce

a recombi-nant

antigen

forusein

large

scale,

antibody prevalence,

sero-epidemiologic

studies. Such

analyses

should

help

further clar-ify thedistributionandimportance of Norwalk infection.

Acknowledgments

We would liketothank Drs. K.Fryand E. Mackow for

helpful

discus-sions.We alsogratefully acknowledgetheassistance ofR.Lo,K.Yun,

J.Fernandez, and the Genelabs VisualArtsDepartment.

This work wassupported inpartby Public Health Servicegrants DK-01811(to S.Matsui), DK-38707 (to the Stanford Digestive Disease Center), and RR-02558 (to the University of Texas Clinical Research Center), anda VeteransAdministration Merit Review grant(to H. Greenberg).

Noteaddedinproof:During the revision of thismanuscript, apaper describing thecloning ofaportion of the Norwalk viruswaspublished elsewhere(Jiang, X.,D. Y.Graham,K.Wang, andM. K.Estes. 1990. Science(Wash. DC). 250:1580-1583).Wehaveidentifiedan immuno-reactive regionof the Norwalk virus that is different from the polymer-aseregion reported by Jiangetal.

References

1.Dolin,R., J. J. Treanor, and H. P. Madore. 1987. Novel agents of viral

enteritis in humans.J.Infect.Dis. 155:365-373.

2. Greenberg,H. B., M.Skaar,and S. S. Monroe. 1989. The 22 to 30 nm

gastroenteritisagentsofman. InViral Diarrheas ofMan and Animals. L. J. Saif

andK. W.Thiel, editors. CRC Press,Inc. Boca Raton, FL. 137-159.

3. Blacklow,N. R.,andG. Cukor. 1982. Norwalkvirus:a major cause of

epidemic gastroenteritis.Am.J.Public Health. 72:1321-1323.

4.Adler,J. L., and R.Zickl. 1963.Winter vomitingdisease.J. Infect. Dis.

119:668-673.

5.Kapikian,A.Z.,R.G. Wyatt, R.Dolin,D.S.Thornhill,A.R.Kalica,and R. M.Chanock. 1972. Visualizationby immune electronmicroscopyofa27 nm

particle associated withacute infectious non-bacterialgastroenteritis.J. Virol. 10:1075-1081.

6. Kaplan,J.E., G.W.Gary,R.C. Baron,N.Singh,L. B. Schonberger, R.

Feldman, andH. B.Greenberg. 1982. Epidemiology ofNorwalk gastroenteritis

and theroleofNorwalkvirus in outbreaks ofacutenonbacterialgastroenteritis. Ann.Intern.Med. 96:756-761.

7.National Centerfor Health Statistics,U. S. Department of Health,

Educa-tion, and Welfare. 1973.Current estimates.From The Health Interview Survey.

Series 10,no.85, United States- 1972. Government Printing Office,Washington,

DC.

8.Dolin, R.,N. R.Blacklow, H.DuPont,S.B. Formal, R. J. Buscho, R. G. Wyatt, J. A. Kasal, R. P. Chames, R. Hornick, and R. M. Chanock. 1971.

Trans-mission ofacutenonbacterial gastroenteritisto volunteers by oral administration

ofstoolfiltrates. J.Infect.Dis. 123:307-312.

9.Levy, A.G.,L.Widerlite,C. J. Schwartz, R. Dolin, N. R. Blacklow, J. D.

Gardner,D. V.Kimberg,and J.S.Trier.1976. Jejunal adenylate cyclase activity

in human subjects during viral gastroenteritis. Gastroenterology. 70:321-325.

10.Parrino,T.A.,D.S.Schreiber, J. S. Trier,A. Z.Kapikian,and N. R.

Blacklow. 1977. Clinicalimmunityin acutegastroenteritis caused bythe Nor-walk agent. N.Engl.J. Med. 297:86-89.

11.Greenberg, H. B.,R.G. Wyatt,J.Valdesuso,A.R.Kalica,W. T.London,

R. M.Chanock,and A. Z.Kapikian. 1978. Solid-phase microtiter

radioimmuno-assayfor detectionofthe Norwalkstrainofacutenonbacterialepidemic gastroen-teritisand its antibodies.J.Med. Virol. 2:97-108.

12. Gary, G. W., Jr.,J. E.Kaplan, S.E.Stine,and L. J. Anderson. 1985.

Detection of Norwalk virus antibodiesandantigenwith a biotin-avidin immuno-assay.J. Clin.Microbiol.22:274-278.

13.Reyes,G. R.,M. A.Purdy, J.P.Kim, K.-C. Luk,L. M.Young,K. E.Fry, andD. W.Bradley. 1990.Isolation of a cDNA from the virusresponsible for

enterically transmitted non-A, non-B hepatitis.Science(Wash. DC).

247:1335-1339.

14.Johnson,P.C.,J. J.Mathewson, H.L.DuPont,and H.B. Greenberg. 1990.Multiple-challenge study ofhost susceptibilitytoNorwalkgastroenteritisin U.S. adults.J.Infect.Dis. 161:18-21.

15.Greenberg,H.B.,J.Valdesuso,A. R.Kalica,R. G.Wyatt,V. J.

McAu-liffe,A. Z.Kapikian, andR.M.Chanock.1981. Proteins of Norwalk virus.J. Virol. 37:994-999.

16. Chomczynski,P.,andN.Sacchi. 1987. Single-stepmethodofRNA isola-tion by acidguanidinium thiocyanate-phenol-chloroform extraction. Anal.

Bio-chem. 162:156-159.

17.Gubler, U., andB.J.Hoffmann. 1983.Asimpleand veryefficient method forgeneratingcDNAlibraries. Gene (Amst.). 25:263-269.

18.Young,R. A., and R. W.Davis. 1983. Efficient isolationofgenesusing antibody probes.Proc.Nail.Acad. Sci. USA. 80:1194-1196.

19.Southern,E. M.1975.Detection of specificsequences among DNA

frag-mentsseparated by gelelectrophoresis. J. Mol. Biol. 98:503-517.

20.Sanger, F., S.Nicklen, andA.R.Coulson. 1977. DNAsequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA. 74:5463-5467.

21.Saiki,R.K., S.Scharf, F. Faloona,K. B.Mullis, G.T.Horn,H. A.Erlich, andN.Arnheim. 1985.Enzymatic amplification offglobingenomicsequences

andrestriction siteanalysis for diagnosis of sickle cell anemia. Science (Wash. DC). 230:1350-1354.

22. Madore,H.P.,J. J.Treanor,and R.Dolin. 1986. Characterization ofthe

SnowMountainagentof viral gastroenteritis. J. Virol. 58:487-492.

23.Terashima, H., S. Chiba,Y.Sakuma,R.Kogasaka,S.Nakata,R.Minami,

K.Horino,andT.Nakao. 1983. Thepolypeptidesofahuman calicivirus. Arch.

Virol. 78:1-7.