INFECTION AND IMMUNITY, OCt. 1983, p. 293-3000019-9567/83/100293-08$02.00/0Copyright © 1983, American Society for Microbiology

Vol. 42, No. 1

Response of Mice to Rotaviruses of Bovine or Primate OriginAssessed by Radioimmunoassay, Radioimmunoprecipitation,

and Plaque Reduction NeutralizationPAUL A. OFFIT,* H. FRED CLARK, AND STANLEY A. PLOTKIN

The Children's Hospital of Philadelphia, Division of Infectious Diseases, University of Pennsylvania,Philadelphia, Pennsylvania 19104

Received 18 April 1983/Accepted 11 July 1983

Sera from (i) gnotobiotic BALB/c, CD-1, and CFW mice and (ii) conventionalBALB/c mice were evaluated by radioimmunoassay, radioimmunoprecipitation,and plaque reduction neutralization, using the Wa, SA-11, and WC-3 (bovine)strains of rotavirus as the detecting antigens. The gnotobiotic mice had noantirotavirus antibody detectable by radioimmunoprecipitation and no neutraliz-ing antibody at a dilution of 1:50 by plaque reduction neutralization. All sera fromthe conventional mice had rotavirus-specific antibodies detected by radio-immunoassay and by radioimmunoprecipitation at serum dilutions of 1:50 and1:10,000, respectively. The antibodies were directed against viral proteins p116,p94, p88, and p84 of all three viruses, but had no neutralizing activity againstheterologous rotaviruses at a dilution of 1:50. Conventional seropositive micewere parenterally immunized with the Wa, SA-11, or WC-3 strain of rotavirus. Anapproximate 100-fold increase in rotavirus-specific antibodies was detected byradioimmunoassay, and >20-fold selective neutralization of the immunizing strainof virus was observed. Sera from the mice immunized with Wa virus hadantibodies directed against inner and outer capsid proteins of all three rotaviruses.The mouse can be a useful model for studying the immune response to heterolo-gous rotavirus infection; preexisting antibodies presumably directed towardsmurine rotavirus do not prevent the development of a type-specific immuneresponse to a nonmurine rotavirus.

Rotaviruses have been shown to be the singlemost important group of etiological agents ofacute gastroenteritis requiring hospitalization ofinfants and young children, both in the UnitedStates and in developing countries (3, 4, 13, 29).The worldwide impact of these viruses has excit-ed interest in disease prevention via the develop-ment of a vaccine (14, 15). The development of asuccessful vaccine will be facilitated by definingthe immune response necessary for protectionagainst challenge. Some light has been shed onthe nature of the protein(s) responsible for evok-ing type-specific neutralizing antibodies by anti-sera prepared against purified viral antigens (1,16, 24). However, numerous investigators haveshown that convalescent antisera obtained fromeither animals or humans can neutralize rotavi-ruses isolated from heterologous species (5, 8,25, 31, 37, 39, 43). This fact implies that aheterotypic rotavirus may be used as a candidatevaccine strain.

In both animals and humans, the prevalenceof rotavirus-specific antibodies is extensive (10,

18, 20, 27, 30, 36, 41-43). Therefore, for thestudy of the immune response to rotaviruses, ithas been difficult to identify an experimentalanimal model without preexisting rotavirus-spe-cific antibodies. In the numerous studies of theimmunization of animals with heterologous rota-virus antigens, the effect of these preexistingantibodies on the specificity of the immuneresponse has not been elucidated. In this paper,we evaluate the polypeptide and neutralizationspecificities of sera obtained from (i) gnotobioticBALB/c, CD-1, and CFW mice; (ii) convention-al seropositive BALB/c mice obtained fromcommercial breeding laboratories; and (iii) con-ventional seropositive BALB/c mice parenteral-ly immunized with heterologous rotaviruses. Weshow that (i) structural polypeptides on both theinner and outer capsids of the Wa, SA-11, andWC-3 rotavirus strains share antigenic determi-nants and (ii) preexisting antibodies presumablydirected toward murine rotavirus do not preventthe development of a type-specific immune re-sponse to a nonmurine rotavirus.

293

294 OFFIT, CLARK, AND PLOTKIN

MATERIALS AND METHODS

Animals and production of antisera. BALB/c mice (8to 14 g) were obtained from Flow Laboratories, Inc.,(McLean, Va.), Charles River Breeding Laboratories,Inc.; (Wilmington, Pa.), and Harlan-Sprague-DawleyBreeding Laboratories (Madison, Wis.). Sera fromadult, gnotobiotic CD-1 and BALB/c mice were ob-tained from Howard Blatt (Fox Chase, Pa.), and serafrom adult, gnotobiotic CFW mice were obtained fromMorris Pollard (Notre Dame, Ind.). Blood sampleswere obtained by retroorbital capillary plexus punc-ture from all mice on the day of arrival. All animalswere housed in separate units (isolators), with eachunit containing its own air supply.

Conventional BALB/c mice were primed intraperi-toneally with 250 ,ll of a clarified virus stock (Wa [4 x106 PFU/ml], SA-11 [5.1 x 107 PFU/ml], or WC-3 [1.5X 107 PFU/ml]) and boosted 3 and 5 weeks later by theintravenous inoculation of 5 ,ug of purified virus. Serawere obtained 4 days after the second booster inocula-tion.

Cells and viruses. African green monkey kidney cells(CV-1) were grown in BHK cell medium (19), supple-mented with 10% fetal bovine serum and 25 p.g ofgentamicin per ml. Fetal rhesus monkey cells (MA-104) were grown in BHK cell medium supplementedwith 10% fetal bovine serum, 100 U of penicillin perml, and 100 ,ug of streptomycin per ml.The Wa strain of human rotavirus was obtained

from Richard Wyatt (Bethesda, Md.). A seed stock ofsimian rotavirus SA-11 was obtained from H. H.Malherbe (San Antonio, Tex.). The bovine rotavirusWC-3 was isolated from a cow in southeastern Penn-sylvania in 1981 and adapted to growth in CV-1 cells inthis laboratory.

Plaque-purified stocks of Wa, SA-11, and WC-3viruses for use in these studies were prepared in CV-1cells. Confluent monolayers of CV-1 cells grown in150-cm2 flasks were washed twice with phosphate-buffered saline (PBS) and infected at a multiplicity ofinfection of 0.1 PFU/cell. After an adsorption period of30 min at 37°C, BHK medium containing 40 ,ug oftrypsin (Flow) per ml without a serum supplement wasadded. The cells and culture fluid were harvestedwhen a cytopathic effect including -90% of the cellmonolayer was evident. The tissue culture fluids wereclarified by centrifugation for 10 min at 600 x g.

Purification of virus. Infected CV-1 cells were dis-rupted by three cycles offreezing and thawing, and theresultant suspension was clarified by centrifugation for25 min at 15,300 x g. The cell pellet was suspended inTNC buffer (0.05 M Tris-HC1 [pH 7.5], 0.15 M sodiumchloride, 0.01 M calcium chloride) and extracted twicewith 1,1,3-trifluorotrichloroethane. The aqueousphase from this extraction was combined with thesupernatant fluid from the initial clarification andcentrifuged at 5°C for 2 h at 125,000 x g. The pelletswere suspended in TNC buffer, mixed with CsCl at aninitial density of 1.36 g/cm3, and subjected to isopycniccentrifugation at 5°C for 18 h at 200,000 x g, using atype 50.1 rotor (Beckman). After centrifugation, thevisible top and bottom viral bands were collected afterbottom puncture of the tubes, suspended in TNCbuffer, and pelleted at 5°C for 2 h at 250,000 x g. Thepellets were suspended in TNC buffer, divided into 10-p.1 portions, and stored at -70°C. To estimate the

quantity of rotavirus purified, the absorbance of thefinal suspension was measured. It was assumed thatthe specific absorbance was equal to that of reovirus(5.1 absorbance U at 260 nm/mg per ml of virus) (35).

Viral infectivity plaque assay. Viral infectivity wasassayed by plaque formation in six-well plastic tissueculture plates (Flow) containing confluent monolayersof MA-104 cells. After the growth medium was re-moved, each well was washed twice with PBS. Eachwell was inoculated with 0.1 ml of a serial 10-folddilution of the virus. After an adsorption period of 30min at 37°C, 2.5 ml of overlay medium, consisting of0.5% purified agar (Agarose; Seakem) and 13 ,ug oftrypsin per ml in Eagle minimal essential medium, wasadded. The cultures were placed in a humidifiedincubator for 4 days at 37°C in 5% CO2. A secondoverlay medium, containing 0.5% purified agar and0.03% neutral red in Earle balanced salt solution, wasthen added, and the plaques were counted approxi-mately 5 h later.PRN assay. The plaque reduction neutralization

(PRN) assay was a modification of the techniquedescribed previously by Matsuno et al. (25). A virussuspension containing 500 PFU of Wa, SA-11, or WC-3 rotavirus per ml was mixed with an equal volume ofserial fivefold dilutions of serum (heat treated at 56°Cfor 30 min). The serum-virus mixture was incubated ina water bath at 37°C for 30 min. The serum-virusmixture (0.2 ml) was then inoculated onto confluentmonolayers of MA-104 cells in six-well plates andincubated for 30 min at 37°C. The plates were againwashed twice with PBS, and the addition of theoverlay medium and counting of viral plaques wereperformed as described above.RIA. A 100-ng amount of purified Wa virus in

sodium carbonate buffer (15 mM Na2CO3, 35 mMNaHCO3) was added to individual wells of round-bottomed 96-well polyvinyl plates (Costar) and kept atroom temperature overnight. Each well was thentreated with 1% bovine serum albumin in PBS. Serumsamples were diluted in 0.1% bovine serum albumin inPBS and incubated with viral immunoadsorbent for 90min at room temperature. Serum controls were per-formed by adding each serum dilution to wells withoutimmunoadsorbent. The wells were then washed withPBS and incubated for 90 min with 1251I-labeled rabbitanti-mouse F(ab')2. After a final washing procedure,the wells were separated from the plate and assayedfor radioactivity in a gamma radiation counter. Finalvalues were obtained by subtracting the serum controlvalues from the serum values obtained with immuno-adsorbent.

Preparation of lysates of [35S]methionine-labeled ro-tavirus-infected cells. Monolayers of MA-104 cells in35-mm plastic petri dishes (Flow) were washed twicewith PBS 12 h before infection, after which BHKmedium without serum was added. The cells wereeither mock infected with BHK medium or infected ata multiplicity of infection of 10 with Wa, SA-11, orWC-3 rotavirus. After a 30-min adsorption period at370C, BHK cell medium supplemented with 3 ,ug ofactinomycin D per ml and 5 ,ug of trypsin per ml wasadded. At 12 h postinfection, the cells were washedonce with PBS, and then the medium was changed tomethionine-free BHK medium supplemented with 3,ug of actinomycin D per ml and 5 ,ug of trypsin per ml.At 13 h postinfection, the cells were labeled for 20 min

INFECT. IMMUN.

MURINE IMMUNE RESPONSE TO ROTAVIRUS 295

A B C D

116 - -

8894 ;"84

60--

55-

41-_l

37 -35 _34-=29 - | W

27- ^

FIG. 1. [35S]methionine-labeled rotavirus polypep-tides synthesized in vitro were separated by electro-phoresis in l1o SDS-polyacrylamide gels, and thebands were visualized by fluorography. Lane A, WC-3virus-infected cells; lane B, SA-11 virus-infected cells,lane C, Wa virus-infected cells; and lane D, uninfectedcells.

with 500 ,uCi of [35S]methionine per ml (New EnglandNuclear Corp., Boston, Mass.). Cell lysates wereprepared by adding 200 ,ul of buffer containing 0.8 MKCI, 10 mM Tris-hydrochloride (pH 7.8), 1 mM phen-ylmethylsulfonyl fluoride and 1% Triton X-100 fol-lowed by 800 ,ul of 10 mM Tris-hydrochloride (pH 7.8),1 mM phenylmethylsulfonyl fluoride, and 1% TritonX-100. The lysates were than centrifuged for 45 min at5°C at 250,000 x g. The supernatant fluids wereremoved and used as a source of labeled rotavirusproteins for radioimmunoprecipitation (RIP) (see be-low).RIP of rotavirus proteins. Cell lysates were pread-

sorbed with staphylococcus protein A Cowan strain I(SAC I) by adding 500 1ld of a 10%o SAC I suspension to1.0 ml of lysate. After incubation for 15 min at O°C,the bacteria were removed by centrifugation for 2 minat 12,800 x g. The preadsorbed supernatants werethen divided into 5-,ul portions, to which 5 ,ul ofundiluted serum was added. After 18 h at 4°C, 75 ,u1 of10% SAC I was added to each serum-lysate mixtureand allowed to incubate at 0°C for 1 h. The bacteriawere pelleted and washed three times with PBS con-

taining 2% fetal bovine serum and 0.5% Triton X-100and three times with PBS containing 0.1% sodiumdodecyl sulfate (SDS) and 0.5% Triton X-100. Theadsorbed labeled proteins were recovered by suspend-ing the bacterial pellets in 20 R1I of sample buffercontaining 0.25 M Tris-hydrochloride (pH 6.8), 20%glycerol, 1% SDS, 2% 2-mercaptoethanol, and 0.003%phenol red and boiling the suspension for 3 min. Thebacteria were pelleted, and the supernatants wereapplied to SDS-polyacrylamide gels.

Discontinuous SDS-polyacrylamide gel electrophore-sis. Discontinuous SDS-polyacrylamide gel electro-

phoresis was performed on vertical slabs (1.5 by 100mm), using 10-mm-wide sample wells formed in thestacking gel. The stacking gel consisted of the acryl-amide monomer (5% [wt/vol]), N,N'-methylenebisacryl-amide (0.13% [wt/vol]), 0.125 M Tris base (pH 6.8)N,N,N',N'-tetramethylenediamine (0.05%) [vol/vol]),ammonium persulfate (0.04% [wt/vol]), and SDS (0.1%[wt/vol]). The resolving gel consisted of the acryl-amide monomer (10% [wt/vol]), N,N'-methylenebis-acrylamide (0.26% [wt/vol]), 0.375 M Tris base (pH8.8), tetramethylenediamine (0.05% [vol/vol]), ammo-nium persulfate (0.06% [wt/vol]), and SDS (0.1%[Wt/vol]).

Electrophoresis was performed at 30 mA per gel.Molecular weight standards prepared as describedabove (Bio-Rad Laboratories, Richmond, Calif.) weredetected by staining with silver nitrate as described byMerril et al. (28). Fluorograms were prepared asdescribed by Laskey and Mills (17).

RESULTS35S-labeling of rotavirus polypeptides for RIP.

35S-labeled Wa, SA-11, and WC-3 polypeptideswere produced in MA-104 cells. Figure 1 showsthe SDS-polyacrylamide gel electrophoresispatterns of the virus-specified proteins as deter-mined by comparison with a mock-infected con-trol lysate. To determine whether virus-speci-fied proteins were nonstructural or were locatedon the inner capsid or outer capsid of the virion,we compared our protein electropherogramswith those of previous investigators (6, 22). Themolecular weights of virus-specified proteinswere determined by comparison with proteins ofknown molecular weight (Bio-Rad) (Table 1).Slight differences in molecular weights are foundamong homologous viral proteins of rotavirusstrains.

Analysis of sera from conventional and gnotobi-otic mice. (i) RIA. Rotavirus-binding antibody

TABLE 1. Molecular weights of Wa, SA-11, andWC-3 rotavirus proteins

Mol wt (x103) of:Designationa Wa SA-11 WC-3

I 116 116 116I (p88, p84) 94 94 94O (p61, p29) 88 88 880 84 84 840 61 60 61NS 55 55 55I 41 41 42O 38 37 37NS 35 35 35NS 34 34 34O 29 29.5 30O 27 27.5 27.5

a I, Inner capsid protein; 0, outer capsid protein;NS, nonstructural protein. The proteins in parenthesesare proteolytic cleavage products (6, 22).

VOL. 42, 1983

296 OFFIT, CLARK, AND PLOTKIN

0

x

ca

CL

I-.

f-

m

0aC4

70-

60-

50-

40-

30-

20-

10

SERUM DILUTION 10-216-3 1o-4SERUM DILUTION

FIG. 2. (A) RIA of sera from 25 conventional unimmunized BALB/c (-0*) and 16 gnotobiotic BALB/c,CD-1, and CFW (0----0) mice. The titers are represented as the mean plus or minus the standard error. (B) RIAof sera from 18 conventional BALB/c mice immunized with either Wa, SA-11, or WC-3 virus as described in thetext. The titers are represented as the mean plus or minus the standard error of pre- (0----0) and post- (0-*)immunization sera.

was detected in the sera of 25 conventionalBALB/c mice at a dilution of 1:50 and wasgreater than twice the value shown in the sera of16 gnotobiotic BALB/c, CD-1, and CFW micetested at the same dilution (Fig. 2A). No signifi-cant differences were found in the levels ofapparently low nonspecific radioimmunoassay(RIA) reactivity in different strains of gnotobiot-ic mice.

(ii) RIP. All conventional BALB/c mice ob-tained from commercial laboratories were foundby RIP to have rotavirus-specific antibodiesdirected against structural proteins p94, p88,p84, and p41 at dilutions of 1:100 and 1:10,000,using 35S-labeled Wa, SA-11, or WC-3 viralpolypeptides (Fig. 3). (The bands detected at amolecular weight of approximately 150,000probably represent antibodies directed againsteither [i] cellular proteins or [ii] poorly solubi-lized viral proteins.) No rotavirus-specific anti-bodies were detected in sera from gnotobioticBALB/c, CD-1, and CFW mice by RIP.

(iii) PRN. Pre-immunization sera from nineconventional BALB/c mice tested by PRN didnot neutralize Wa, SA-11, or WC-3 virus at adilution of 1:50 despite the presence of rota-virus-specific antibodies detected by RIA andRIP. Sera from 16 gnotobiotic BALB/c, CD-1,and CFW mice tested by PRN were found tohave no neutralizing activity against Wa, SA-11,or WC-3 virus at a dilution of 1:50.

2 A B

1A B C

94

8884

__WS 441

FIG. 3. (1) RIP analysis of undiluted serum from anunimmunized conventional BALB/c mouse using 35S-labeled WC-3 (lane A), SA-11 (lane B), and Wa (laneC) viral polypeptides. (2) RIP of serum from anunimmunized conventional BALB/c mouse at dilu-tions of 1:100 (lane A) and 1:10,000 (lane B), using 35S-labeled Wa viral polypeptides.

0

2

a.

u

C-

0

0

I-

z

m

I-

In

B

IF

\U

INFECT. IMMUN.

MURINE IMMUNE RESPONSE TO ROTAVIRUS 297

TABLE 2. PRN assay of sera from conventional BALB/c mice parenterally immunized with either Wa, SA-11, or WC-3 rotavirus

Neutralization by sera(reciprocal of serum dilution)

Immunizing antigen Animal from mice immunized with:

Wa SA-11 WC-3

Wa WA R2 9,375 240 <100WA NM 3,000 170 <100WA R1 >12,500 300 100

SA-11 SA NM <100 >12,500 485SA R2 <100 11,125 100SA R1 <100 >12,500 300

WC-3 WC RL <100 300 >12,500WC RI <100 400 >12,500WC L1 <100 380 6,250

Analysis of sera from parenterally immunizedconventional mice. (i) RIA. A total of 18 BALB/cmice were immunized with either Wa, SA-11, or

WC-3 rotavirus. Pre- and post-immunizationsera were tested by RIA (Fig. 2B). Post-immuni-zation an approximately 100-fold increase inrotavirus-specific antibody titers above preim-munization levels was found. No significantdifferences in antibody titers among mice immu-nized with Wa, SA-11, or WC-3 rotavirus werefound.

(ii) RIP. Sera from mice immunized with Warotavirus were tested by RIP, using 35S-labeledWa, SA-11, or WC-3 viral polypeptides. Typicalresults are shown in Fig. 4. Sera from animalsimmunized with Wa virus had antibodies direct-ed against proteins which comprise the inner andouter capsid layers of Wa, SA-11, and WC-3viruses. Our difficulty in precipitating the poly-peptide with a molecular weight of 116,000 (I1)by RIP using sera from animals parenterallyimmunized with rotavirus has been experiencedby other investigators (6, 21).

(iii) PRN. Three sera from each group of sixmice immunized with either Wa, SA-11, or WC-3 virus were selected at random for testing byPRN (Table 2). Neutralization of the immunizingvirus at serum dilutions of 1:3,000 or greater wasshown by all of the sera tested, but neutraliza-tion of heterologous viruses was also sometimesfound at dilutions approximately 20-fold lower.

DISCUSSION

All conventional 3- to 4-week old BALB/cmice tested from three commercial laboratorieswere shown to have detectable rotavirus-specif-ic antibodies by both RIA and RIP. The RIP testwas more sensitive than the RIA, with antibod-

ies detected at a dilution of 1:10,000 by RIP ascompared with a dilution of 1:50 by RIA. Rota-virus-specific antibodies were clearly absent ingnotobiotic murine sera tested by the sameassays. The gnotobiotic BALB/c, CD-1, andCFW mice provide us with a consistently sero-negative population of adult animals, a situationapparently difficult to maintain with BALB/cmice in commercial breeding laboratories.The rotavirus-specific antibodies in the con-

ventional mice presumably represent either pre-vious exposure to murine rotavirus (epizooticdiarrhea of infant mice) or antibodies passivelyacquired from a dam infected with this virus.Sheridan et al. (34), using C57BL, CD-1, andCF-1 mice, have presented preliminary data,indicating that there is a positive correlationbetween the titers of circulating rotavirus-specif-ic immunoglobulin G (IgG) in the dam and thosein her litter as detected by enzyme-linked im-munoadsorbent assay. The IgG appeared to beof colostral milk origin as decay of antibodyoccurred after weaning. Furthermore, colostral-ly deprived animals derived by cesarean sectionfrom an IgG-seropositive dam were seronega-tive. In their study, a neonate of a dam with aserum IgG titer of approximately 1:8,000 hadtiters of 1:512 and 1:32 at 21 and 42 days of age,respectively. A titer of 1:50 as detected by RIAin our study of 3- to 4-week-old conventionalmice does not allow a distinction between thepassive or active acquisition of rotavirus-specif-ic antibodies.

Rotaviruses have been found to share one ormore antigens by various techniques, includingcomplement fixation, immunofluorescence, geldiffusion, immune electron microscopy, and en-zyme-linked immunoadsorbent assay (39, 43,45). By immune electron microscopy, heterolo-gous convalescent antiserum has been shown to

VOL. 42, 1983

298 OFFIT, CLARK, AND PLOTKIN

A B C D

9488

84

60-55-

41373534

29 -27 -

FIG. 4. RIP of serum from mouse Wa R1 pre- (laneA) and post- (lane B) immunization with Wa virus,using "S-labeled Wa (lanes A and B), SA-11 (lane C),and WC-3 (lane D) rotaviral polypeptides. The outercapsid protein (-p38) of all three rotaviruses is precip-itated by post-immunization serum as shown in lanesB, C, and D.

agglutinate rotavirus particles lacking an outercapside layer (2, 23, 40). This suggests thatrotaviruses share determinants on the inner cap-sid layer of the virus. We have shown by RIPthat conventional mice have antibodies directedagainst 35S-labeled Wa, SA-11, and WC-3 viralpolypeptides p94, p88, p84, and p41. Theseproteins are associated with viral cores purifiedfrom stocks grown in the presence of trypsin (6,22).

In conventional mice immunized parenterallywith Wa, SA-11, or WC-3 rotavirus, we wereable to induce an immune response specific forrotavirus stains as determined by the selectiveneutralization of the immunizing strain by PRN.Numerous investigators have demonstrated thisrotavirus type-specific immune response in con-ventional guinea pigs and rabbits obtained com-mercially (7-9, 11, 26, 32, 33, 39, 40, 44). Ourstudies show that the same type-specific im-mune response can be achieved in mice in thepresence of preexisting antibodies to shareddeterminants among murine and heterologousrotaviruses.

INFECT. IMMUN.

Sera from conventionally bred mice immu-nized with Wa virus had antibodies directedagainst the inner and outer capsid proteins ofWavirus, as well as against those of SA-11 and WC-3 virus by RIP. Studies of convalescent andrandom sera from humans, as well as fromvarious animal species, have consistently shownneutralizing activity against heterologous rota-viruses (5, 8, 25, 31, 37, 39, 43). However, todate, studies using polyclonal, monospecific,and monoclonal antibody preparations have notprovided a clear-cut explanation of this phenom-enon. One possible explanation for the interspe-cies neutralizing activity of convalescent anti-sera is that some degree of homology of theouter capsid protein(s) of rotaviruses responsi-ble for evoking neutralizing antibodies exists.Kalica et al. (12), in analyzing reassortants be-tween the UK bovine and human strains ofrotavirus, found that neutralizing specificity reg-ularly segregated with the ninth RNA gene seg-ment of the human rotavirus. Investigationsusing monospecific sera have shown that a hightiter of type-specific neutralizing antibodies isevoked by immunizing guinea pigs with themajor outer capsid glycoprotein (-p38) (1, 17,24). Using sera from immunized conventionalmice, we have demonstrated some degree ofhomology among the outer capsid proteins (in-cluding outer capsid protein p38) of these threerotavirus strains. The evaluation of sera fromgnotobiotic mice hyperimmunized with nonmur-ine rotavirus strains will determine whetherthese heterologous neutralizing antibodies aredue to inherent cross-reactive domains on the38-kilodalton protein among rotavirus strains orare secondary to previous exposure of the im-munized mouse to murine rotavirus. The titra-tion of the RIP activity of sera from mice hyper-immunized with different rotaviruses may clarifythe relationship between cross-reactive neutral-ization by PRN and RIP-detected shared deter-minants among homologous outer capsid pro-teins. These studies are now in progress in ourlaboratory.Using RIA, RIP and PRN, we have defined a

seropositive population of conventional mice, aswell as a seronegative population of gnotobioticmice. Our studies of parenterally immunizedconventional mice show that (i) structural poly-peptides of the inner and outer capsids of theWa, SA-11, and WC-3 rotavirus strains shareantigenic determinants; and (ii) a type-specificimmune response to a nonmurine rotavirus canbe obtained despite the presence of preexistingantibodies presumably directed against murinerotavirus. The mouse can be a useful model forevaluating the immune response to rotavirusesfrom other species.

4#401.-"O

---

-1-1

40.

:4b 40

MURINE IMMUNE RESPONSE TO ROTAVIRUS 299

ACKNOWLEDGMENTS

We thank Dianna Hannigan and Kathy Dolan for theirtechnical assistance. We also thank Walter Gerhard, WilliamStroop, Charles Hackett, Mark Frankel, and Jan Tuttlemanfor their helpful discussions and careful reading of the manu-

script.This work was supported by Public Health Service grant

F 32 AI-06733 from the National Institutes of Health to P.A.O.and, in part, by the Hassel Foundation.

LITERATURE CITED

1. Bastardo, J. W., J. L. McKimm-Breschkin, S. Sonza, L. D.Mercer, and I. H. Homes. 1981. Preparation and charac-terization of antisera to electrophoretically purified SA-11virus polypeptides. Infect. Immun. 34:641-647.

2. Bridger, J. C. 1978. Location of type-specific antigens incalf rotaviruses. J. Clin. Microbiol. 8:625-628.

3. Bryden, A. S., H. A. Davies, R. E. Hadley, and T. H.Flewett. 1975. Rotavirus enteritis in the West Midlandsduring 1974. Lancet ii:241-243.

4. Davidson, G. P., R. F. Bishop, R. R. W. Townley, I. H.Holmes, and B. J. Ruck. 1975. Importance of a new virusin acute sporadic enteritis in children. Lancet i:242-246.

5. Elias, M. M. 1977. Distribution and titres of rotavirusantibodies in different age groups. J. Hyg. 79:365-372.

6. Ericson, B. L., D. Y. Graham, B. B. Mason, and M. K.Estes. 1982. Identification, synthesis, and modification ofsimian rotavirus SA-11 polypeptides in infected cells. J.Virol. 42:825-839.

7. Estes, M. K., and D. Y. Graham. 1980. Identification ofrotaviruses of different origins by the plaque-reductiontest. Am. J. Vet. Res. 41:151-152.

8. Gaul, S. K., T. F. Simpson, G. N. Woode, and R. W.Fulton. 1982. Antigenic relationships among some animalrotaviruses: virus neutralization in vitro and cross-protec-tion in piglets. J. Clin. Microbiol. 16:495-503.

9. Greenberg, H. B., A. R. Kalica, R. G. Wyatt, and R. W.Jones. 1981. Rescue of non-cultivatable human rotavirusby gene reassortment during mixed infection with tsmutants of a cultivatable bovine rotavirus. Proc. Nati.Acad. Sci. U.S.A. 78:420-424.

10. Hall, G. A., J. C. Bridger, R. L. Chandler, and G. N.Woode. 1976. Gnotobiotic piglets experimentally infectedwith neonatal calf diarrhea reovirus-like agent (rotavirus).Vet. Pathol. 13:197-210.

11. Hoshino, Y., R. G. Wyatt, F. W. Scott, and M. J. Appel.1982. Isolation and characterization of a canine rotavirus.Arch. Virol. 72:113-125.

12. Kalica, A. R., H. B. Greenberg, R. G. Wyatt, and J.Flores. 1981. Genes of human (strain Wa) and bovine(strain UK) rotaviruses that code for neutralization andsubgroup antigens. Virology 112:385-390.

13. Kapikian, A. Z., H. W. Kim, R. G. Wyatt, and W. L.Cline. 1976. Human reovirus-like agent as the majorpathogen associated with "winter" gastroenteritis in hos-pitalized infants and young children. N. Engl. J. Med.294:965-972.

14. Kapikian, A. Z., R. G. Wyatt, H. B. Greenberg, and A. R.Kalica. 1980. Approaches to immunization of infants andyoung children against gastroenteritis due to rotaviruses.Rev. Infect. Dis. 2:459-469.

15. Kapikian, A. Z., R. G. Wyatt, M. M. Levine, and R. H.Jolken. 1983. Oral administration of human rotavirus tovolunteers: induction of illness and correlates of resist-ance. J. Infect. Dis. 147:95-106.

16. Killen, H. M., and N. J. Dimmock. 1982. Identification ofa neutralization-specific antigen of a calf rotavirus. J.Gen. Virol. 62:297-311.

17. Laskey, R. A., and A. D. Mills. 1975. Quantitative filmdetection of 3H and "4C in polyacrylamide gels by fluorog-raphy. Eur. J. Biochem. 56:335-341.

18. Lecce, J. G., M. W. King, and W. E. Dorsey. 1978.

Rearing regimen-producing piglet diarrhea (rotavirus) andits relevance to acute infantile diarrhea. Science 199:776-778.

19. MacPherson, I., and M. Stoker. 1962. Polyoma transfor-mation of hamster cell clones-an investigation of geneticfactors affecting cell competence. Virology 16:147-151.

20. McNulty, M. S., G. M. Allan, D. J. Thompson, and J. D.O'Boyle. 1978. Antibody to rotavirus in dogs and cats.Vet. Rec. 102:534-535.

21. Mason, B. B., D. Y. Graham, and M. K. Estes. 1980. Invitro transcription and translation of simian rotavirus SA-11 gene products. J. Virol. 33:1111-1121.

22. Mason, B. B., D. Y. Graham, and M. K. Estes. 1983.Biochemical mapping of the simian rotavirus SA-11genome. J. Virol. 46:413-423.

23. Mathan, M. 1977. An antigenic subunit present in rota-virus-infected faeces. J. Gen. Virol. 34:325-329.

24. Matsuno, S., and S. Inouye. 1983. Purification of outercapsid glycoprotein of neonatal calf diarrhea virus andpreparation of its antisera. Infect. Immun. 39:155-158.

25. Matsuno, S., S. Inouye, and R. Kono. 1977. Plaque assayof neonatal calf diarrhea virus and the neutralizing anti-body in human sera. J. Clin. Microbiol. 5:1-4.

26. Matsuno, S., S. Inouye, and R. Kono. 1977. Antigenicrelationship between human and bovine rotaviruses asdetermined by neutralization, immune adherence hemag-glutination and complement fixation tests. Infect. Immun.17:661-662.

27. Mebus, C. A., R. G. White, E. P. Bass, and M. J.Twiehaus. 1973. Immunity to neonatal calf diarrhea virus.J. Am. Vet. Med. 163:880-883.

28. Merril, C. R., D. Goldman, S. A. Sedman, and M. H.Ebert. 1980. Ultrasensitive stain for proteins in polyacryl-amide gels shows regional variation of cerebrospinal fluidproteins. Science 211:1437-1438.

29. Middleton, P. J., M. T. Szymanski, and M. Petric. 1977.Viruses associated with acute gastroenteritis in youngchildren. Am. J. Dis. Child. 131:733-737.

30. Petric, M., P. J. Middleton, C. Grant, J. S. Tam, andC. M. Hewitt. 1978. Lapine rotavirus: preliminary studieson epizoology and transmission. Can. J. Comp. Med.42:143-147.

31. Sato, K., Y. Inaba, T. Shinozaki, and M. Matumoto. 1981.Neutralizing antibody to bovine rotavirus in various ani-mal species. Vet. Microbiol. 6:259-261.

32. Sato, K., Y. Inaba, S. Tokuhisa, and M. Matumoto. 1982.Isolation of lapine rotavirus in cell cultures. Arch. Virol.71:267-271.

33. Sato, K., Y. Inaba, Y. Miura, S. Tokuhisa, and M.Matumoto. 1982. Antigenic relationships between rotavir-uses from different species as studied by neutralizationand immunofluorescence. Arch. Virol. 73:45-50.

34. Sheridan, J. F., R. S. Eydelloth, S. L. Vonderfecht, and L.Aurelian. 1983. Virus-specific immunity in neonatal andadult mouse rotavirus infection. Infect. Immun. 39:917-927.

35. Smith, R., H. Zweerink, and W. Joklik. 1969. Polypeptidecomponents of virus top components and cores of reovi-rus type 3. Virology 39:791-810.

36. Snodgrass, D. R., J. A. Herring, K. A. Linklater, and D. A.Dyson. 1977. A survey of rotaviruses in sheep in Scotland.Vet. Rec. 100:341.

37. Stuker, G., N. J. Schmidt, B. Forghani, and C. A. Holm-berg. 1979. Rotavirus antibody assays on monkey sera: acomparison of enzyme immunoassay with neutralizationand complement-fixation tests. Am. J. Vet. Res. 40:1620-1623.

38. Thouless, M. E., G. M. Beards, and T. H. Flewett. 1982.Serotyping and subgrouping of rotavirus strains by theELISA test. Arch. Virol. 73:219-230.

39. Thouless, M. E., A. S. Bryden, T. H. Flewett, and G. N.Woode. 1977. Serological relationships between rotavir-uses from different species as studied by complement-fixation and neutralization. Arch. Virol. 53:287-294.

VOL. 42, 1983

300 OFFIT, CLARK, AND PLOTKIN

40. Urasawa, S., T. Urasawa, and K. Taniguchi. 1982. Threehuman rotavirus serotypes demonstrated by plaque neu-tralization of isolated strains. Infect. Immun. 38:781-784.

41. Woode, G. N. 1978. Epizootology of bovine rotavirusinfection. Vet. Rec. 103:44-46.

42. Woode, G. N., and J. C. Bridger. 1975. Levels of colostralantibodies against neonatal calf diarrhea virus. J. Am.Vet. Med. 163:880-883.

43. Woode, G. N., J. C. Bridger, J. M. Jones, and T. H.Flewett. 1976. Morphological and antigenic relationships

INFECT. IMMUN.

between viruses (rotaviruses) from acute gastroenteritis ofchildren, calves, piglets, mice and foals. Infect. Immun.14:804-810.

44. Wyatt, R. G., H. B. Greenberg, W. D. James, and A. L.Pittman. 1982. Definition of human rotavirus serotypes byplaque reduction assay. Infect. Immun. 37:110-115.

45. Yolken, R. H., B. Barbour, R. G. Wyatt, and A. R. Kalica.1978. Enzyme-linked immunosorbent assay for identifica-tion of rotaviruses from different animal species. Science201:259-262.