Vol. 31, No. 8JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1993, p. 2010-20150095-1137/93/082010-06$02.00/0Copyright © 1993, American Society for Microbiology

Characterization of Field Strains of Group A BovineRotaviruses by Using Polymerase Chain Reaction-Generated

G and P Type-Specific cDNA ProbestANIL V. PARWANI, HUSSEIN A. HUSSEIN, BLAIR I. ROSEN, ALEJANDRO LUCCHELLI,

LORENZO NAVARRO, AND LINDA J. SAIF*Food Animal Health Research Program, Ohio Agricultural Research and Development Center,

The Ohio State University, Wooster, Ohio 44691

Received 1 March 1993/Accepted 5 May 1993

Dot and Northern blot hybridization assays were used to analyze field strains of group A bovine rotaviruses(BRVs) by using nucleic acid probes representing P and G type specificities. The probes were prepared bypolymerase chain reaction amplification of hyperdivergent regions of the cloned VP4 (nucleotides 211 to 686)and VP7 (nucleotides 51 to 392) genes from four serotypically distinct (in P or G types) strains of rotaviruses:NCDV (G6, P1), IND (G6, P5), 69M (G8, P10), and Cr (G10, P11). The P and G type cDNA probes wereradiolabeled with [32P]dCTP and hybridized with RNA extracted from reference cell culture-passagedrotavirus strains or the field samples. The field samples were obtained from young diarrheic calves from Ohio,Nebraska, Washington State, and Canada. The cDNA probes were specific for their respective G or P types onthe basis of analysis of known P and G type reference strains. The G typing analysis of 102 field samplesrevealed that 36.3% (37 of 102) were G6, 2.9% (3 of 102) were G8, 12.7% (13 of 102) were G10, and 23.5%(24 of 102) were untypeable. The P typing results for 93 samples indicated that 2.2% (2 of 93) were P1(NCDV-like), 20.4% (19 of 93) were P5 (UK-like), 9.3% (10 of 93) were P11 (B223-like), and 40.8% (38 of 93)were untypeable. This is the first report of the identification among BRV strains in North America of a G typeother than G6 or G10. Our report further confirms that G6, P5 rotaviruses are predominant among the BRVfield strains that we examined, and the P ypes of these strains differ from that of the BRV vaccine strain usedin the United States (G6, P1). The large number of untypeable G (23.5%) and P (40.8%) types suggests thatother or new P and G types exist among BRV field strains.

Group A bovine rotaviruses (BRVs) are a major cause ofenteric disease in young calves (1 to 3 weeks of age) (13, 20,27). Affected calves may die as a result of severe dehydra-tion or secondary bacterial infections (27). No highly effec-tive vaccines for the prevention or control ofBRV infectionsare available in the United States (20, 21). The serotypicclassification of rotaviruses is based on the reactivities of theouter capsid proteins, VP4 and VP7, which determine the P(for protease-susceptible) and G (for glycoprotein) sero-types, respectively (4, 10). Group A rotaviruses are classi-fied into at least 14 G serotypes and 12 P serotypes (2-4, 10,25). The information on the diversity of BRV P and G typesin the field is important for the future development andmonitoring of effective vaccines, because rotaviruses whichpossess distinct G and P types may not induce heterotypiccross-protection (9, 20, 27, 29).

Isolates of at least four G serotypes (Gl, G6, G8, and G10)infect cattle (15, 24), of which serotypes G6 and G10(represented by prototype strains NCDV and B223, respec-tively) have been identified in the United States (15, 29).Recently, a Gl BRV isolate, T449, was identified in the fecesof a calf with diarrhea in Argentina (1). Others have reportedG8 BRV isolates in the United Kingdom (24). At least four Ptypes (P1 [NCDV], P5 [UK], P11 [B223], and P12 [678]) havebeen reported among BRV strains (16, 25). It has beensuggested that the difference between P types may havecontributed to the lack of homotypic (G type-specific) pro-

* Corresponding author.t Journal article 36-93 of The Ohio Agricultural Research and

Development Center.

tection observed between the G6 strains NCDV (P1) andB641 (P5) (9, 29). To better understand the epidemiology ofBRV infections and to elucidate the antigenic and molecularrelationships that exist among BRV strains, sensitive andreliable methods for the characterization of P and G types ofBRV are very important.

Nucleic acid probes and the polymerase chain reaction(PCR) have been useful for the detection and serotypicclassification of human and animal rotaviruses in clinicalspecimens (5, 7, 12, 15-18, 20). Full-length cDNA probesprepared from rotavirus gene segments coding for VP7 orpartial-length PCR probes amplified from hyperdivergentregions of the VP7 gene have been used for the G typing ofhuman (7), bovine (15), and porcine (18) rotaviruses. Similarmethods have been reported for the P typing of human (12),bovine (16), and porcine (17) group A rotaviruses.The objective of the present study was to clone the

full-length VP4 and VP7 genes of the BRV strains NCDV,IND (UK-like), and Cr (B223-like) and to generate partial-length PCR probes for the P and G type characterization of93 and 102 BRV field strains, respectively, from Ohio,Nebraska, Washington State, and Ontario, Canada.

MATERIALS AND METHODS

Viruses and cells. The reference viruses were grown inrhesus monkey kidney (MA-104) cells in roller bottles andtitrated by a plaque assay as described previously (22). Thestrains and serotypes (P and G) of the reference human andanimal rotaviruses used in the present study are listed inTable 1, and their sources have been described previously

2010

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

G AND P TYPING OF BOVINE ROTAVIRUSES 2011

TABLE 1. Cell culture-adapted and gnotobiotic calf-passagedrotaviruses used for nucleic acid hybridization with BRV

G and P type PCR-derived cDNA probes

Rotavirus . G Pstrain Group Speces typea typeb

NonbovineWa A Human 1 8DS-1 A Human 2 4ST3 A Human 4 6VA70 A Human 4 869M A Human 8 10SA1l A Simian 3 2Gottfried A Porcine 4 6OSU A Porcine 5 7

BovineK124c A Bovine 6d leT1309f A Bovine 6d leC486 A Bovine 6d leIND A Bovine 6d SeOK A Bovine 6d SeID A Bovine 6d SeATI1 A Bovine 6d SeRV584 A Bovine 6d SeB223 A Bovine lod lleCr A Bovine lod lleATIB B Bovine ND) NDShintoku C Bovine ND NDa According to the unified serotyping scheme on the basis of virus neutral-

ization assays originally proposed by Hoshino et al. (10).b According to the tentative P serotyping scheme originally proposed by

Estes and Cohen (4) and extended by Snodgrass et al. (25).c Nebraska calf diarrhea virus (NCDV), Cody strain.d G serotype as determined in the present study on the basis of hybridiza-

tion analysis with G-type (G6, IND; G10, Cr) cDNA probes and numberedaccording to the uniform serotyping scheme (10). The G type cDNA probesdid not cross-react with heterologous G types.

Pp serotype as determined in the present study on the basis of hybridizationdata obtained with P-type (P1, NCDV; P5, IND; P11, Cr) cDNA probes andnumbered according to the P typing system proposed by Estes and Cohen (4)and extended by Snodgrass et al. (25) for group A bovine rotaviruses. The Ptype cDNA probes did not cross-react with heterologous P types.J Nebraska calf diarrhea virus, Lincoln strain.g ND, not determined.

(15-18, 20). Ten reference group A BRV laboratory strainswere also used as controls and are listed in Table 1. Inaddition to the BRV strains described previously (15, 16),strain C486 was obtained from D. Benfield, South DakotaState University, strain ATIl was obtained from the Agri-cultural Technical Institute (ATI) herd in Wooster, Ohio

(20a), and the cell culture-adapted strain RV584 was ob-tained from Salah Hammami, University of California atDavis. The G serotypes of these BRV strains were previ-ously characterized by using G6- and G10-neutralizing sero-

type-specific monoclonal antibodies (MAbs) or by two-waycross-neutralization tests (20a). A group C BRV, Shintokustrain, previously adapted to growth in cell culture was usedas a negative control (28). The group A BRV strains NCDV(Cody), Cr, OK, ID, B223, and IND (15) and a bovine groupB rotavirus (ATI B strain) (19) were also passaged ingnotobiotic or colostrum-deprived calves as described pre-

viously (21) (Table 1).Field strains of BRV. Two rotavirus-negative and 102

BRV-positive samples for the hybridization assays were

obtained from Ohio (n = 57), Nebraska (n = 27), Washington(n = 3), and Ontario, Canada (n = 15). The majority of thesamples were provided as calf diarrhea survey samples(Ohio) (13) or as diagnostic samples from calves with diar-rhea (1 to 30 days of age) and were collected between 1988and 1992. Before testing with nucleic acid probes, all thesamples were tested for group A rotaviruses by polyacryl-amide gel electrophoresis (PAGE) (18) and at least one ormore of the following assays: enzyme-linked immunosorbentassay (ELISA) (13) and immune electron microscopy (13).

Extraction and electrophoresis of RNA. Rotavirus double-stranded RNA (dsRNA) was extracted from cell culture-propagated viruses or fecal samples from the field by previ-ously described procedures (18). The dsRNA in extractedsamples was analyzed by PAGE to confirm the presence ofdsRNA and to examine the genomic electropherotypes asdescribed previously (15). Electrophoresis was conducted at12 mA for 14 to 16 h. The dsRNA bands were visualized bysilver staining or staining with ethidium bromide (0.5 ,ug/ml;Sigma Chemical Co., St. Louis, Mo.).

Cloning of BRV VP4 and VP7 genes. The oligonucleotideprimers used in the present study are listed in Table 2. Thesingle-stranded RNA (ssRNA) was produced by in vitrotranscription of purified viral cores by previously describedprocedures (6). Amplification of ssRNA templates requiredan initial reverse transcription step. First-strand cDNA was

synthesized by incubating heat-denatured ssRNA (0.2 to 1,ug) for 30 min to 1 h at 42°C in a reaction mixture containing25 mM Tris (pH 8.3), 25 mM KCl, 5 mM MgCl2, 1 mMdithiothreitol, 200 ,uM (each) dATP, dCTP, dTIP, anddGTP, 200 ng of primers 7B (VP7 gene) or 4B (VP4 gene),and 10 U of avian myeloblastosis virus reverse transcriptase(Boehringer Mannheim Biochemicals, Indianapolis, Ind.)(Fig. 1).

TABLE 2. Sequences of oligonucleotides used for PCR amplification of full- or partial-length VP4 and VP7 genes

Primer Gene Location Strand' Sequence

4A VP4b 5' end Sense 5'-CCC GGG ATC CGA ATT CGG CTA TAA AAT GGC TTG GCT-3'4B VP4 3' end Antisense 5'-TCOG CGA ATT CTG CAG GTC ACA TCC TCC AGA AGC-3'4C VP4 Nucleotides 211 to 230 Sense 5'-CCG TAT CAG CCOG GOCG CCG TT-3'4D VP4 Nucleotides 677 to 686 Antisense 5'-GGC GGC AGC CCOG TTG TTT AT-3'7A VP7c 5' end Sense 5'-CCC GGG ATC CAT GGC CGG CTT TAA AAG CGA GAA TTT-3'7B VP7 3' end Antisense 5'-CGA TCG CGA ATT CTG CGG CAG GTC ACA TCA TAC AAC TCT A-3'7C VP7 Nucleotides 51 to 71 Sense 5'-GTA TGG TAT TGA ATA TAC CAC-3'7D VP7 Nucleotides 376 to 392 Antisense 5'-GAT CCT GTT GGC CAT CC-3'

a Strand denotes the sense or antisense of the primer, which indicates that the sequence of the primer is either identical or complementary to the bovinerotavirus mRNA (VP4 and/or VP7 genes), respectively.

b The sequences of the oligonucleotides used for partial- and full-length amplification of the VP4 genes were based on the published sequence for the VP4 geneof the UK strain of group A bovine rotaviruses (11).

c The sequences of the oligonucleotides used for partial- and full-length amplification of the VP7 genes were based on the published sequence for the VP7 geneof the NCDV strain of group A bovine rotaviruses (8).

VOL. 31, 1993

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

2012 PARWANI ET AL.

l

a 9fi mQ

APrimer 4A Primer 4D

211 686 .,*I z

serotyping"Primer4C region

BPrimer 7A

Primer 4B

Primer 7D51 392 _

"serotyping" PPrimer 7C region Primer 7B

FIG. 1. Schematic representation of primers VP4 (A) and VP7(B) used for PCR amplification of full- or partial-length cDNAfragments in the hybridization studies. For amplification of VP7partial-length cDNA segments, primers 7C and 7D were used. Foramplification of VP7 full-length cDNA segments, primers 7A and 7Bwere used. For amplification of VP4 partial-length cDNA segments,primers 4C and 4D were used, and for amplification of VP4full-length cDNA segments, primers 4A and 4B were used.

The PCR mixture contained 200 ,uM (each) dATP, dCTP,dGTP, and dTTP, 20 mM Tris-HCl (pH 8.3), 2.5 mM MgCl2,0.05% gelatin, 200 ng of each primer (VP4 gene, primers 4Aand 4B; VP7 gene, primers 7A and 7B), 1 to 10 ng of DNAtemplate, and 2.5 U of Taq polymerase (Boehringer Mann-heim Biochemicals). Thirty amplification cycles were used,with each cycle consisting of 94°C for 1 min (denaturation),42°C for 1.5 min (annealing), and 72°C for 3.5 min (exten-sion). The full-length PCR products were purified by elec-troelution and analyzed on 1% agarose gels by standardprocedures (23).The PCR-amplified full-length VP4 and VP7 genes of BRV

strains IND, NCDV, and Cr were cloned into the plasmidpGEM 3zf (Promega, Madison, Wis.) or pBS (Stratagene, LaJolla, Calif.) by standard recombinant DNA procedures (23).Recombinant plasmids were identified by colony blot hybrid-ization with 3 P-labeled PCR probes. The specificities of theclones were confirmed by Northern blot hybridization bypreviously described procedures (15).The full-length VP7 gene of human rotavirus strain 69M

(G8) cloned into the plasmid pUC 13 was obtained fromJorge Flores, National Institutes of Health. The recombinantplasmids were used as templates for the production by PCRof partial-length cDNA segments for use as probes in sero-typing assays.PCR amplification and labeling. Partial-length VP7 cDNA

(nucleotides 51 to 392) and VP4 cDNA (nucleotides 211 to686), encompassing areas of major sequence diversity, wereprepared by PCR amplification with specific primers (Fig. 1).For amplification of VP7 cDNA segments (341 bp) fromrotavirus strains IND (G6), 69M (G8), and Cr (G10), primer7C (homologous to nucleotides 51 to 71 of the NCDV VP7gene) and primer 7D (complementary to nucleotides 376 to392 of the NCDV VP7 gene) were used (Table 2). For PCRamplification of VP4 partial-length cDNA segments (475 bp)from strains NCDV (P1), IND (P5), and Cr (P11), primer 4C(sequence homologous to nucleotides 211 to 630 of the UKVP4 gene) and primer 4D (sequence complementary tonucleotides 677 to 686 of the UK VP4 gene) were used (Table

(A) (B) (C)

FIG. 2. (A) Agarose gel electrophoresis of the amplificationproduct (PCR) of the full-length VP4 gene of the NCDV strain ofbovine rotavirus. Lane MW contains molecular weight markers.The arrow indicates the PCR product representing the full-lengthgene 4 of the NCDV BRV strain. (B and C) The specificity of thecloned PCR product was confirmed by Northern blot hybridizationassay against RNA from BRV strains NCDV, B223, and ID. (B)Ethidium bromide-stained dsRNA electropherotypes ofBRV strainsrepresenting P types 1 (NCDV), 5 (ID), and 11 (B223); (C) autora-diograph corresponding to the gel in panel B obtained with the32P-labeled NCDV gene 4 clone. Arrowheads, gene segment 4.

2). These primers corresponded to highly conserved regionsof the VP4 and VP7 genes and flanked the target "sequence-divergent" regions (Fig. 1).The purified PCR products were radiolabeled by using a

nick-translation kit (Bethesda Research Laboratories, Gaith-ersburg, Md.) and [32P]dCTP (specific activity, 650 Ci/mmol;ICN Biomedicals Inc., Irvine, Calif.) as described previ-ously (16).

Northern and dot blot hybridization assays. The Northernand dot blot hybridization assays were performed by usingmodifications of previously described procedures (16). Hy-bridization was performed under high-stringency conditions(520C, 5x SSC [lx SSC is 0.15 M NaCl plus 0.015 M sodiumcitrate] 50% formamide) in 3 to 5 ml of hybridization buffercontaining 4.5% dextran sulfate and 3 x 106 to 5 x 106 cpmof heat-denatured PCR probe (-1.2 x 107 cpm/pg of DNAtemplate). The hybridizations were performed for 16 to 24 hat 52°C. The membranes were washed four times at roomtemperature in 2x SSC-0.1% sodium dodecyl sulfate (SDS)and two times at the hybridization temperature in 0.4xSSC-0.1% SDS. The washed membranes were rinsed oncewith water, blotted, and exposed to Kodak XAR films withintensifying screens at -700C for 1 to 7 days.

RESULTS

Cloning of BRV VP4 and VP7 genes. Full-length BRV VP4genes were amplified from the BRV strains NCDV (P1), IND(PS), and Cr (P11). Figure 2A shows the full-length PCRproduct (2,362 bp) amplified from ssRNA obtained from the

I I

I I

J. CLIN. MICROBIOL.

9 <0la 03

10 9> goiz M lau C4

M IZz 0-0

w IMPAW

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

G AND P TYPING OF BOVINE ROTAVIRUSES 2013

TABLE 3. Identification of bovine rotavirus G (G6, G8, and G10) types from reference and clinical specimens by PAGE andnucleic acid hybridization assays

No. (%) of strains

Sample origin Total no. PAGE Probe type Total Total

tested positive IND VP7 69M VP7 Cr VP7 positivea untypeableb(G6) (G8)C (G10)

Reference strainsd 10 10 8 (80) 0 (0) 2 (20) 10 (100) 0 (0)Field strainseOhio 57 46 15 (26) 2 (3.5) 8 (14) 25 (54) 21 (46)Nebraska 27 19 12 (44) 1 (3.7) 3 (11) 16 (84) 3 (16)Washington 3 2 2 (67) 0 (0) 0 (0) 2 (67) 0 (0)Canada 15 10 8 (53) 0 (0) 2 (13) 10 (100) 0 (0)

Total 102 77 (75.5) 37 (36.3) 3 (2.9) 13 (12.8) 53 (51.9) 24 (23.5)

a These samples tested positive by PAGE and with cDNA probes.b These samples tested positive by PAGE, but were negative with G6, G8, and G10 cDNA probes.This represents a G8 probe prepared from the cloned full-length VP7 gene of the human rotavirus strain 69M (14).

d Ten laboratory strains of bovine rotavirus propagated in MA104 cells (Table 1).I Fecal samples from calves with diarrhea diagnosed positive for group A BRV.

BRV strain NCDV. The PCR products were purified andcloned into recombinant plasmids for use as templates insubsequent experiments and for further studies. The speci-ficities of the clones were confirmed by dot and Northernblot hybridization assays. Figure 2B shows the ethidiumbromide-stained electropherotypes of BRV strains NCDV,B223, and ID (short genome pattern). The dsRNA waselectrotransferred onto nylon membranes which were hy-bridized with the radiolabeled NCDV VP4 clone. The clonewas specific and hybridized only with the VP4 gene from theBRV strain NCDV and not with the heterologous P types(B223 [P11] and ID [P5) (Fig. 2C). Similarly, full-lengthBRV VP7 genes from strains IND (G6), NCDV (G6), and Cr(G10) were generated by PCR and cloned into the recombi-nant plasmid pGEM. Under high-stringency conditions, theG6 clones hybridized only with the dsRNA samples obtainedfrom G6 rotaviruses (IND and NCDV) (data not shown).Similarly, the G10 clone (Cr strain) hybridized only with theRNA extracted from G10 samples (B223 and Cr) (data notshown).

Specificities of the partial-length G and P type PCR probes.Dot and Northern blot hybridization assays with the cellculture-passaged rotavirus samples and the BRV VP4 andVP7 cDNA probes showed that the three VP4 (P1, P5, andP11) and the three VP7 (G6, G8, and G10) cDNA probes didnot hybridize with other group A rotaviruses, including thehuman, porcine, and simian strains, with different G and Pspecificities. Furthermore, none of the probes showed anyreaction with other rotavirus serogroups (groups B and C)(data not shown).

Characterization of the G and P types of the BRV fieldstrains. Rotavirus dsRNA was extracted from the samplesobtained from the field. The dsRNA was heat denatured andtested by dot blot hybridization assays. All of the 102rotavirus-positive and 2 rotavirus-negative samples weretested by dot blot hybridization by using the three G type-specific partial-length cDNA probes. For P typing, 93 rota-virus-positive and 2 rotavirus-negative samples were tested.The results of the G typing of the 102 BRV field samples

and the 10 reference BRV strains by using the partial-lengthPCR-generated cDNA probes are presented in Table 3. The24 PAGE-positive field samples were untypeable becausethey failed to hybridize with any of the three G type-specificprobes. For P typing, a total of 93 BRV field samples and the

10 reference BRV strains were examined, and the results ofthe P typing assays are presented in Table 4. The 38PAGE-positive field samples were considered untypeablebecause they tested negative with all three P type probes.The G and P typing results were compared to investigate

the various combinations of G and P types that occurred inthe field specimens tested. The most frequently detectedcombination of G and P types in the samples that weexamined was G6, P5 (n = 14). Other P type combinationswith G6 were G6, P1 (n = 2) and G6, P11 (n = 4). There wereno P1, P5, or P11 types detected among the three type G8field samples. For the G10 strains, the combinations wereG10, P5 (n = 2) and G10, P11 (n = 4), but there were no G10,P1 strains.

DISCUSSION

In the present study, we cloned full-length VP4 and VP7genes from three serotypically distinct (in P or G types)strains of rotaviruses: NCDV (G6, P1), IND (G6, P5), and Cr(G10, P11). The cloned BRV VP4 and VP7 genes were usedas templates in PCR to generate partial-length cDNA seg-ments representing serotype-specific regions of the genes.Similarly, a human rotavirus (69M) G8 cDNA partial-lengthprobe was also prepared from the recombinant plasmid madeavailable for the present study. These partial-length seg-ments were radiolabeled and used to characterize fieldstrains of group A bovine rotaviruses.The serotypic classification of BRV has also been defined

on the basis of VP7 specificities by ELISA with VP7-specificMAbs (20, 24). The neutralization specificities of rotavirus,however, are also dependent on VP4, and the VP4 genesegregates independently from the VP7 gene (4). It hastherefore been proposed that serotypic classification ofrotaviruses should account for both VP4 (P) and VP7 (G)specificities (4, 10, 25). Previous investigators have reportedthat it is difficult to prepare VP4-specific MAbs (26). There-fore, at least for P typing, cDNA probes offer great promise.In addition, the cDNA probes have the potential to detectgenomic variants or monotypes not detectable by MAbs (5).We demonstrated in the present study that cDNA probes canbe used as a reliable means of characterizing the G and Ptype specificities of BRV field specimens.Two previous studies have investigated the frequencies of

VOL. 31, 1993

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

2014 PARWANI ET AL.

TABLE 4. Identification of bovine rotavirus P (P1, P5, and P11) types from clinical specimens by PAGE and nucleic acidhybridization assays

No. (%) of samples

Sampleorigin ~~~~~~~~~~~~~ProbetypeSample origin Total no. PAGE Total Totaltested positive NCDV VP4 IND VP4 Cr VP4 positivea untypeableb

(P1) (P5) (P11)Reference 10 10 3 (30) 5 (50) 2 (20) 10 (100) 0 (0)

strainscField strainsdOhio 48 38 1 (2) 8 (17) 3 (6) 12 (32) 26 (68)Nebraska 27 19 1 (4) 7 (26) 4 (15) 12 (63) 7 (37)Washington 3 2 0 (0) 0 (0) 0 (0) 0 (0) 2 (67)Canada 15 10 0 (0) 4 (27) 3 (20) 7 (70) 3 (30)

Total 93 69 (74.2) 2 (2.2) 19 (20.4) 10 (9.3) 31 (33.3) 38 (40.8)a These samples tested positive by PAGE and with cDNA probes.b These samples tested positive by PAGE, but were negative with the P1, P5, and P11 cDNA probes.c Ten laboratory strains of bovine rotavirus propagated in MA-104 cells (Table 1).d Fecal samples from calves with diarrhea diagnosed positive for group A BRV.

infection of calves with serotypes G6 and G10. Woode et al.(29) used two-way neutralization assays and demonstratedthat 89% of the isolates were G6 but that only 7.4% wereG10. Snodgrass et al. (24) used a MAb-based ELISA for Gtyping and reported that 66% of the isolates that they testedwere G6 but that only 7.4% were typed as G10. Those resultsare in accordance with the results of our study (although ouroverall percentage was lower), in which the predominant Gtype was G6 (37 of 102 samples tested; 36.3%), whereas 13 of102 (12.7%) were G10. The reason for these discrepanciesmay reflect the more diverse (in geographic location) fieldsamples used in our study. Furthermore, we have estab-lished in our laboratory that this discrepancy is not due todifferences in antigenic versus genetic typing with MAbsversus cDNA probes, respectively, because the percentobserved agreement between MAbs and probes for thesamples described in the present study for G6 and G10typing was 90.6% (12a).There is very limited information on the frequency and

distribution of BRV P types. Such information is very usefulfrom an epidemiological standpoint. Furthermore, there hasbeen no systematic study of the P types of BRV field cases.In our studies, the predominant P type was P5 (19 of 93samples tested; 20.4%). The P type found in the vaccinestrain (NCDV, P1) was detected in only 2.2% of the samples.Comparison of the G and P typing results further confirmsthat the field strains with the G and P combination G6, P5were detected most frequently (n = 14) among the BRV fieldstrains examined. Therefore, a BRV strain with the G6, P5combination may be a better vaccine candidate strain thanthe BRV vaccine strain used in the United States (G6, P1).Investigators have observed a lack of cross-protection be-tween the two BRV G6 strains B641 (G6, P5) and NCDV(G6, P1) (29). Because the P types of these strains aredistinct, it has been suggested that differences in P types maybe a contributing factor to this lack of cross-protection (9).

In the present study, we identified at least three isolatesthat tested positive with the G8 cDNA probe. Snodgrass etal. (24) demonstrated that the two bovine strains J2538 and678 from the United Kingdom belong to G8. Interestingly,Ohshima et al. (14) reported a moderate level of homologybetween the human G8 strain 69M and some bovine rotavi-rus strains. To our knowledge, ours is the first report of G8rotaviruses identified among cattle in North America. There

are at least three possibilities that may explain the occur-rence of human G types (Gl and G8) (1, 24) in the bovinepopulation. First, they may be the result of direct transmis-sion between animals and humans. Second, the humanstrains may have evolved ancestrally from animal strains,and lastly, there may be natural reassortants between humanand animal strains.The large number of untypeable BRV field samples (G

untypeables, 23.5%; P untypeables, 40.8%) remains a majorarea of concern (Tables 3 and 4). These samples previouslytested positive by ELISA and PAGE. Since it is possible thatany combination of P and G types may occur in nature, theuntypeable samples may represent other existing or new Gor P types. These G or P types may escape classification ifthere are no suitable diagnostic reagents available. Thesesamples may also be classified as untypeable because theywere tested only with cDNA probes representing the Gand/or P types traditionally associated with the bovinepopulation. In future studies, the samples that tested nega-tive with the BRV G type (G6, G8, and G10) and P type (P1,P5, and P11) cDNA probes will be retested with human andother animal G and P type cDNA probes that are notcommonly associated with the bovine population.

ACKNOWLEDGMENTS

The synthetic primers used in the present study were kindlyprovided by Mario Gorziglia, and the human rotavirus strain 69MVP7 clone was provided by J. Flores, both of the National Institutesof Health. Various rotavirus strains were kindly provided by SusanLance and Paul Bartlett (13), The Ohio State University; RamMohan, Ohio Department of Agriculture; Fernando Osorio, Univer-sity of Nebraska; Thomas Besser, Washington State University; andSusan Carman, Ontario Ministry of Agriculture and Food, Ottawa,Ontario, Canada.

Salaries and research support were provided by state and federalfunds appropriated to the Ohio Agricultural Research and Develop-ment Center, The Ohio State University. This study was supportedin part by Animal Health Competitive Research grant 86-CSRS-2-2910 from the Science and Education Administration, CSRS, U.S.Department of Agriculture.

REFERENCES1. Blackhall, J., R. Bellinzoni, N. Mattion, M. K. Estes, J. L.

LaTorre, and G. Magnusson. 1992. A bovine rotavirus serotype1: serologic characterization of the virus and nucleotide se-

J. CLIN. MICROBIOL.

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from

G AND P TYPING OF BOVINE ROTAVIRUSES 2015

quence determination of the structural glycoprotein VP7 gene.Virology 189:833-837.

2. Browning, G. F., R. M. Chalmers, T. A. Fitzgerald, and D. R.Snodgrass. 1991. Serological and genomic characterization ofL338, a novel equine group A rotavirus G serotype. J. Gen.Virol. 72:1059-1064.

3. Browning, G. F., T. A. Fitzgerald, R. M. Chalmers, and D. R.Snodgrass. 1991. A novel group A rotavirus G serotype: sero-logical and genomic characterization of equine isolate F123. J.Clin. Microbiol. 29:2043-2046.

4. Estes, M. K., and J. Cohen. 1989. Rotavirus gene structure andfunction. Microbiol. Rev. 53:410-449.

5. Estes, M. K., and T. Tanaka. 1989. Nucleic acid probes forrotavirus detection and characterization, p. 79-100. In F. Ten-over (ed.), DNA probes for infectious diseases. CRC Press,Inc., Boca Raton, Fla.

6. Flores, J., H. B. Greenberg, J. Myslinski, A. R. Kalicia, R. G.Wyatt, A. Z. Kapikian, and R. M. Chanock. 1982. Use oftranscription probes for genotyping rotavirus reassortants. Vi-rology 121:288-295.

7. Flores, J., J. Sears, I. Perez-Schael, L. White, D. Garcia, C.Lanata, and A. Z. Kapikian. 1990. Identification of humanrotavirus serotypes by hybridization to polymerase chain reac-tion-generated probes derived from hyperdivergent region of thegene encoding outer capsid protein VP7. J. Virol. 64:4021-4024.

8. Glass, R. L, J. Keith, 0. Nakagomi, T. Nakagomi, J. Askaa,A. Z. Kapikian, R. M. Chanock, and J. Flores. 1985. Nucleotidesequence of the structural glycoprotein VP7 gene of Nebraskacalf diarrhea virus rotavirus: comparison with homologousgenes from four strains of human and animal rotaviruses.Virology 141:292-298.

9. Hardy, M. E., G. N. Woode, Z. Xu, and M. Gorziglia. 1991.Comparative amino acid sequence analysis of VP4 for VP7serotype 6 bovine rotavirus strains NCDV, B641, and UK. J.Virol. 65:5535-5538.

10. Hoshino, Y., R. G. Wyatt, H. B. Greenberg, J. Flores, and A. Z.Kapikian. 1984. Serotypic similarity and diversity of rotavirusesof mammalian and avian origin as studied by plaque reductionneutralization. J. Infect. Dis. 149:694-702.

11. Kantharidis, P., M. Dyall-Smith, G. W. Tregear, and I. H.Holmes. 1988. Nucleotide sequence of the UK bovine rotavirussegment 4: possible host restriction of VP3 genes. Virology166:308-315.

12. Larralde, G., and J. Flores. 1990. Identification of gene 4 allelesamong human rotaviruses by polymerase chain reaction-derivedprobes. Virology 179:469-473.

12a.Lucchelli, A., et al. Unpublished data.13. Lucchelli, A., S. E. Lance, P. B. Bartlett, G. Y. Miller, and L. J.

Saif. 1992. Prevalence of bovine group A rotavirus sheddingamong dairy calves in Ohio. Am. J. Vet. Res. 53:169-174.

14. Ohshima, A., T. Takagi, T. Nakagomi, S. Matsuno, and 0.Nakagomi. 1990. Molecular characterization by RNA-RNA hy-bridization of a serotype 8 human rotavirus with "super-short"RNA electropherotype. J. Med. Virol. 30:107-112.

15. Parwani, A. V., B. I. Rosen, J. Flores, M. A. McCrae, M.Gorziglia, and L. J. Saif. 1992. Detection and differentiation of

bovine group A rotavirus serotypes using polymerase chainreaction-generated probes to the VP7 gene. J. Vet. DiagnosticInvest. 4:148-158.

16. Parwani, A. V., B. I. Rosen, M. A. McCrae, and L. J. Saif. 1992.Development of cDNA probes for typing group A bovinerotaviruses on the basis of VP4 specificity. J. Clin. Microbiol.30:2717-2721.

17. Rosen, B. I., A. V. Parwani, M. Gorziglia, G. Larralde, and L. J.Saif. 1992. Characterization of full-length and polymerase chainreaction-derived partial length Gottfried and OSU gene 4 probesfor serotypic differentiation of porcine rotaviruses. J. Clin.Microbiol. 30:2644-2652.

18. Rosen, B. I., L. J. Saif, D. J. Jackwood, and M. Gorziglia. 1990.Hybridization probes for the detection and differentiation of twoserotypes of porcine rotavirus. Vet. Microbiol. 24:327-339.

19. Saif, L. J. 1990. Non group A rotaviruses, p. 73-95. In L. J. Saifand K. W. Theil (ed.), Viral diarrheas of man and animals. CRCPress, Inc., Boca Raton, Fla.

20. Saif, L. J. 1991. Bovine rotavirus, p. 126-130. In A. E. Castroand W. P. Heuschele (ed.), Diagnostic veterinary virology: apractitioner's guide. The Williams & Wilkins Co., Baltimore.

20a.Saif, L. J., et al. Unpublished data.21. Saif, L. J., D. R. Redman, K. L. Smith, and K. W. Theil. 1983.

Passive immunity to bovine rotavirus in newborn calves fedcolostrum supplements from immunized or nonimmunizedcows. Infect. Immun. 41:1118-1131.

22. Saif, L. J., B. I. Rosen, S. Y. Kang, and K. L. Miller. 1988. Cellculture propagation of rotaviruses. J. Tissue Culture Methods11:147-156.

23. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, 2nd ed. Cold Spring HarborLaboratory Press, Cold Spring Harbor, N.Y.

24. Snodgrass, D. R., T. Fitzgerald, I. Campbell, F. M. Scott, G. F.Browning, D. L. Miller, A. J. Herring, and H. B. Greenberg.1990. Rotavirus serotypes 6 and 10 predominate in cattle. J.Clin. Microbiol. 28:504-507.

25. Snodgrass, D. R., Y. Hoshino, T. Fitzgerald, M. Smith, G. F.Browning, and M. Gorziglia. 1992. Identification of four VP4serological types (P serotypes) of bovine rotavirus using viralreassortants. J. Gen. Virol. 73:2319-2325.

26. Taniguchi, K., Y. Morita, T. Urasawa, and S. Urasawa. 1987.Cross-reactive neutralization epitopes on VP3 of human rotavi-ruses: analysis with monoclonal antibodies and antigenic vari-ants. J. Virol. 61:1726-1730.

27. Theil, K. W. 1990. Group A rotaviruses, p. 35-72. In L. J. Saifand K. W. Theil (ed.) Viral diarrheas of man and animals. CRCPress, Inc., Boca Raton, Fla.

28. Tsunemitsu, H., L. J. Saif, B. Jiang, M. Shimizu, M. Hiro, H.Yamaguchi, T. Ishiyama, and T. Hirai. 1991. Isolation, charac-terization, and serial propagation of a bovine group C rotavirusin a monkey kidney cell line (MA104). J. Clin. Microbiol.29:2609-2613.

29. Woode, G. N., N. E. Kelso, T. F. Simpson, S. K. Gaul, L. E.Evans, and L. BabiuL 1983. Antigenic relationships amongsome rotavirus: serum neutralization and cross-protection ingnotobiotic calves. J. Clin. Microbiol. 18:358-364.

VOL. 31, 1993

on April 13, 2018 by guest

http://jcm.asm

.org/D

ownloaded from