Journal of Medical Virology 38292-297 (1992)

Detection and Serotyping of Rotaviruses in Stool Specimens by Using Reverse Transcription and Polymerase Chain Reaction Amplification Hiroshi Ushijima, Hinako Koike, Atsushi Mukoyama, Ayako Hasegawa, Shuichi Nishimura, and Jon Gentsch National Institute of Health, Tokyo, Japan (H.U., H.K., A.M., A.H., S.N.); Centers for Disease Control, Atlanta, Georgia (J.G.) Direct rotavirus serotyping (VP7, G type) in stool specimens was carried out by reverse transcription and polymerase chain reaction amplification (RT-PCR) and compared t o serotyping b y enzyme immunoassay with monoclonal antibodies (EIA-MAb). The methods used for doublestranded (ds) RNA extraction, RT-PCR amplification, and the primers used were modified from previous reports [Gouvea et al.: Journal of Clinical Microbiology 29:519-523, 1990; Gentsch et al.: Journal of Clinical Microbiology, 19921. For samples that were positive by both methods, the serotypes obtained were identical, however RTPCR typing was found t o be considerably more sensitive (70.4% samples serotyped) than EIAMAb (35.6% of samples serotyped). The overall sensitivities for detection of rotavirus in stool samples b y latex agglutination, enzyme immunoassay, electron microscopy, polyacrylamide gel electrophoresis, and RT-PCR were essentially the same. The results confirm that RT-PCR typing (genotyping) is extremely valuable for G typing of samples which cannot be typed by EIA-MAb. We also developed a PCR confirmation technique for serotypes 1, 2, and 4. (c) 1992 Wiley-Liss, Inc.

KEY WORDS: rotavirus, PCR, serotyping

INTRODUCTION Rotaviruses are important in the etiology of acute gastroenteritis in children, with group A rotaviruses causing the largest number of cases. The outer capsid proteins of rotavirus, VP4 and VP7, induce neutralizing antibodies that are important for defense against disease and are the basis for the proposed dual serotyping system that accounts for antigenic specificities [Kapikian et al., 1988; Gorziglia et al., 19901. These serotypes have been tentatively designated as P types and G types, respectively [Estes and Cohen, 19891. Fourteen G serotypes of group A rotaviruses have been 0 1992 WILEY-LISS, INC.

described on the basis of antigenic differences in the VP7 polypeptides [Browning et al., 19911. Serotypes 1 4 are the predominant strains, while serotypes 6, 8, 9, and 12 are minor causes of gastroenteritis in humans. Serotypes 5, 7, 10, 13, and 14 have not been reported in humans [Estes and Cohen, 1989; Browning e t al., 1991; Gerna et al., 19921. Although neutralization tests (plaque reduction, tube neutralization, or fluorescent focus reduction) have been used for serotyping, these methods are cumbersome because viruses must first be cultured from stool samples. An enzyme immunoassay with monoclonal antibodies (EIA-MAb) has been developed to determine serotypes directly from stool samples [Taniguchi et al., 19871. However, EIA-MAb assays cannot always identify rotavirus serotypes. In about one-third to one-fourth of rotavirus samples, serotypes cannot be determined because of the lack of double-shelled particles [Heath et al., 1986; Nakagomi et al., 19881. In addition, strains of serotypes 1 and 4 have been identified which do not bind to antibodies commonly used in EIA-MAb serotyping [Gerna et al., 1988; Bishop et al., 1991; Ward et al., 19911, and monoclonal antibodies specific for serotypes 8, 9, and 12 are not routinely available. Because of these limitations i t is clear that other methods which could broaden the range of direct serotyping of rotaviruses in stool samples would be useful. Recently, reverse transcription-polymerase chain reaction (RT-PCR)amplification and hybridization methods [Gouvea et al., 1990; Flores et al., 1990; Sethabutr e t al., 19901for determining of VP7 serotypes (G types) a t the genetic level have been described. We examined rotavirus G types from stool specimens in Japanese children by RT-PCR typing and demonstrated the importance of this method for stool collections yielding a lower percentage of serotypeable samples by EIA-MAb.

Accepted for publication June 30, 1992. Address reprint requests to Hiroshi Ushijima MD, National Institute of Health, Gakuen 4-7-1, Musashimurayama-shi, Tokyo 208, Japan.

Rotavirus and PCR

MATERIALS AND METHODS Viruses Strains of Wa, S2, YO, Hochi, 69M, and F45 were used as representatives of serotypes 1, 2, 3, 4, 8, and 9, respectively. These viruses were grown in MA104 cells and harvested from culture supernatants and cell fractions. For positive control double standard (ds) RNAs, 200 pl of each supernatant was mixed with one-tenth volume of a rotavirus-negative stool specimen and extracted as described below. One hundred twenty-nine stool samples were collected from children with acute gastroenteritis in Tokyo, Japan from December 1990 to March 1991. These samples were examined previously by latex agglutination test (LA), enzyme immunoassay (EIA) or electron microscopy (EM) [Shinozaki e t al., 1986; Ushijima et al., 19891.Furthermore, in this report, all samples were examined by polyacrylamide gel electrophoresis (PAGE) of ds-RNA [Ushijima e t al., 19891, enzyme immunoassay (EIA) using type specific monoclonal antibodies [Akatani and Ikegami, 1987; Nakagomi et al., 19881and reverse transcription-polymerase chain reaction (RT-PCR) amplification.

293 polymerase, PCR was conducted for 30 cycles (94°C 1 min denature, 47°C 2 min annealing, 72°C 3 min extension, and 72°C 7 min final extension).

Agarose Gel Analysis PCR products (10 pl) were mixed with 3 p1 of dye mixture [30% glycerol, 0.25% bromophenol blue, 0.25% xylene cyanol FF] and loaded on a 1.5% submarine agarose gel in Tris-borate buffer (0.089 M Tris-0.089 M boric acid-0.002 M EDTA, [pH 8.01) containing 0.5 pg of ethidium bromide per ml. After electrophoresis at 100 V for 1 hr, the gels were photographed under UV light with Polaroid type 667 film.

Primers The oligonucleotide primers which we used, Beg9, End9, aAT8, aBT1, aCT2, aDT4, aET3, and aFT9 were described previously [Gouvea et al., 19901. RV1248 was selected for confirming serotypes 1, 2, 4, and 8 in a mixture with the plus sense primers aAT8, aBT1, aCT2, and aDT4. The oligonucleotide sequence of RV1248 (made to the strain Wa sequence) is TGTGTATTTAGTGGACATACTTT ( a primer complementary to a sequence conserved among different rotavirus G serotypes [nucleotides 650-6281). VP7-1' was made Viral RNA a s a universal primer for detecting group A rotaviruses Viral RNA was extracted from stool specimens or in combination with Beg9. The sequence of VP7-1' is positive controls by a modification of a published proce- ACTGATCCTGTTGGCCATCCTTT (made against the dure [Yamada e t al., 1990; Gentsch et al., 19921. sequence of a bovine rotavirus, complementary to G Briefly, 200 pl of 6 M guanidine thiocyanate was added nucleotides 395-373) [Glass et al., 19851. The mixture to the same volume of a 20% stool suspension in a mi- consisting of aAT8, aBT1, aCT2, aDT4, aET3, aFT9 and crocentrifugation tube. After mixing, 10 pl of RNaid End 9 (or RVP1248) was used for detecting or confirm(BIO 101 Inc, La Jolla, CAI was added. The sample was ing serotype specificity. mixed and incubated for 10-15 min on a rocker device. Enzyme Immunoassay With Monoclonal After pelleting 1 min a t low speed (about 500-700g), Antibody the supernatant was aspirated and washed three times with the RNaid kit washer buffer. The sample was cenSerotype 1-,2-, 3-, and 4-specific MAbs (BH49, BW36, trifuged a t 7,OOOg after the last wash. Next, the rotavi- BC5, and BE18, respectively) and serotype 1-4common rus dsRNA bound to RNaid, was dried briefly in a vac- MAb (BB4) were biotinylated before use. Unbiotinyuum centrifuge to remove all the residual alcohol, lated serotype common MAb (AH6) was used as solid resuspended in 17.5 pl sterile, deionized water and vor- phase (capture) antibody. EIA serotyping was carried texed until the RNaid was completely in suspension. out as reported previously [Nakagomi e t al., 19881. The RNaid was incubated for 10 min at 65°C and pel- Briefly, the wells of microtiter plates were coated with leted at 7,OOOg. The supernatant was kept a t -30°C 50 pl of a 1 : l O O dilution serotype common MAb, or until used for the RT-PCR experiments. subgroup specific MAb in carbonate-bicarbonate buffer (pH 9.6). After overnight incubation a t 4"C, the plates RT-PCR From dsRNA were washed three times with PBST (phosphate-buffThe following is our first RT-PCR method: 0.5 pl of ered saline [PBS] containing 0.05% Tween 20) and then dimethyl sulfoxide was added to 9.5 pl of the extracted 50 p1 of SMP-PBST (PBST containing 2.5% skim milk RNA in a 600 PI microcentrifuge tube and heated at powder) was added, followed by 20 pl of stool suspen97°C for 5 min. After immediate cooling on ice, the sions (six wells for each sample). After 90 min incubareaction buffer [3 p1 of 1 M Tris-HC1 (pH 8.3), 5.5 pl of tion a t room temperature or overnight incubation a t 1M KC1,0.35 p1of 1 M MgCl,, 5 pl of 1%gelatin, 0.5 pl 4"C, the plates were washed three times with PBS and of 100 mM DTT and 63.65 p1of H20], 3 pl of the desired 50 pl of 1:100 diluted biotinylated serotype-specific primer mixture (33 pM each), 2 pl of deoxynucleoside MAbs was added. After 90 rnin incubation at room temtriphosphate mixture containing 10 mM each dATP, perature, the plates were washed again, and 50 pl of dCTP, dTTP, and dGTP, and 1.5 pl ( 3 8 . 6 ~ of ) avian 1:150 diluted peroxidase conjugated streptavidin was myeloblastosis virus reverse transcriptase XL (Life Sci- added. After 2 min, the plates were washed again, 50 pl ence Inc., St. Petersburg, FL) were added. The sample of the substrate o-phenylenediamine was added and was incubated a t 37°C for 1 hr. After adding 1U Taq color was allowed to develop for 30 min at room temper-

294

Ushijima et al. Figure 1B shows the electrophoretic patterns of RTPCR products from the same strains as in Figure 1A using a one-amplification procedure with the primer mixture containing aAT8, aBT1, aCT2, aDT4, and RV1248. The specific products were 337 base pairs (bp), 240 bp, 171 bp, and 473 bp with (GI serotype 1 , 2 , 4 ,and 8 prototype strains, respectively (Fig. l B , lanes 1, 2 , 4 , and 8). In addition, specific products of 650 bp were recognized by using the primers Beg 9 and RV1248 for RT-PCR with serotypes 1, 2, 4, and 8 (Fig. IB, lane T). As expected from its location on gene 9 [Gouvea et al., 19901, RV1248 did not work for serotype 3 and 9 (Fig. l B , lanes 3 and 9).

RT-PCR Typing of Rotavirus From Stool Specimens To type rotavirus present in stool specimens, each extracted RNA was subjected to RT-PCR using the same primer mixtures a s used for prototype strains. Minor amounts of cross-reactive bands were observed but the serotype specific bands were always predominant. To validate the PCR typing method for stools the results of EIA-MAb serotyping were compared with the molecular sizes of the serotype-specific products. Of the samples shown below seventeen serotype 1, thirteen serotype 2, nine serotype 3, and seven serotype 4 samples, including standard viruses had been identified. The left side of the uppermost panel shows eight examFig. 1. Electrophoretic patterns of RT-PCR products from cultured rotaviruses. A: From the left side, HaeII1-digested QX 174 DNA size ples of stools containing serotype 1, the left side of the middle upper panel shows eight examples of stool specmarkers (M); RT-PCR product of full length rotavirus gene 9 RNA from strain Wa (primersBeg 9 and End 9) (T);and RT-PCR products of imens containing serotype 2, the left side of the middle strains Wa ((211,S2 ((321, YO (G3),Hochi iG4), 69M ((38)and F45 iG9) lower panel shows seven examples of stool specimens (lanes 1-4, 8, 9) using primer mixture aBT1, aCT2, aET3, aDT4, aAT8, aFT9, and End 9. The molecular weights of the samples in 1, 2, containing serotype 3 and the right side of middle lower 3, 4, 8, and 9 were 749, 652, 374, 583, 885, and 306, respectively. B: panel shows six examples of stool specimens containing From the left side, HaeIII-digested @X 174 DNA size markers (M); serotype 4. RT-PCR product of strain Wa (primer Beg 9 and RV1248) (T);RT-PCR products from standard samples using primer mixture aBT1, aCT2, To confirm these serotype assignments (serotypes 1, aDT4, aAT8, and RV1248 (lanes 1-4,8,9). The molecular weights of 2, and 4) the same samples were subjected to RT-PCR the samples in 1,2,4, and 8 were 337, 240, 171, and 473, respectively. with the mixture in which RV 1248 was substituted for End 9. The right side of the uppermost panel shows ature. The reaction was then stabilized by the addition eight examples of stool specimens containing serotype of 50 pl of 4 M sulfuric acid. The optical density was 1;the right side of the middle upper panel shows eight examples of stool specimens containing serotype 2; the measured a t 492 nm with a micro-EIA reader. left side of the lowermost panel shows six examples of RESULTS stool specimens containing serotype 4. This method was RT-PCR Typing of Cultured Rotavirus especially important in confirming of several serotype 2 Prototype Strains and 3 samples t h a t give very faint products after RTFigure 1A shows the electrophoretic patterns of RT- PCR with the End 9-containing primer mixture (Fig. 2). PCR products from standard serotype 1 , 2 , 3 , 4 , 8 , and 9 As expected, stool samples containing serotype 3 did strains (Wa, S2, YO, Hochi, 69M and F45, respectively) not yield a n RT-PCR product with this primer mixture using a one-amplification procedure with the primer (data not shown). mixture containing aAT8, aBT1, aCT2, aDT4, aET3, The right half of the lowermost panel shows the RTaFT9, and End 9. The product molecular weights and PCR products of RNA from seven examples of rotavirus serotype specificities were identical to those described stool samples containing serotype 1 using universal previously IGouvea e t al., 19901, thus allowing us to primers VP7-1' and Beg 9. Moreover, not only stool serotype rotaviruses by PCR typing (Fig. l A , lanes 1 , 2 , samples containing serotypes 2 , 3 , and 4, but also strain 3, 4, 8, and 9). In addition, all six strains produced 69M (serotype 8 ) and F45 (serotype 9) showed RT-PCR full-length gene 9 DNA after RT-PCR with Beg 9 plus products (data not shown). This primer pair has been End 9 (Fig. l A , lane T) [Gouvea et al., 19901. used to obtain the data in Table 11.

295

Rotavirus and PCR

TABLE I. Comparison of Methods for Detection of Rotavirus (Total 135 Samples)" EIAiLAiEM PAGE PCRb Samples

+ + + -

+ + + +-

+ + + -

2 6

-

-

9

+

~

+ -

88 11 14 1 4

-

-

"Six cultured rotaviruses, Wa (Gl),S2(GZ),YO (G3), Hochi (G4),69M (G8) and F45 ((29) were included in the total number of samples analyzed, The other 129 were stool specimens. bUniversal primer pair (Beg 9 and VP7-1').

TABLE 11. Determination of Serotypes by EIA-MAb and RT-PCR RT-PCR typing

EIA MAb tvoing

G1

G1 G2 G3 G4 G8

17"

G2

G3

G4

G8

UntvDeable

1 0

42

13" 9 7 1

G9

UntvDeable

G9

40b

0

5

2b

0

"A sample containing both serotypes 1and 2. hA sample containing both serotypes 1 and 4. Standard cultivated rotaviruses, Wa (Gl),S2 (G2),YO (G3),Hochi (G4),69M (G8)and F45 (G9)were included in this analysis (total number of samples was 135).

methods, while 38 samples were positive by two or one methods. Nine samples were negative by all three methods. RT-PCR amplification was not more sensitive or more specific in comparison with the other two methods.

Fig. 2. Electrophoretic patterns of rotavirus RT-PCR products from stool specimens using serotypic primer mixtures. Left side of the uppermost panel: Products of amplification of RNA from eight serotype G1 containing stool specimens using primer mixture aBT1, aCT2, aET3, aDT4, aAT8, aFT9, and End9. The expected size of the G1 specific product (amplified by aBTl and End91 is 749 bp; RT-PCR products of RNA from same G1 containing stool specimens with primer mixture aBT1, aCT2, aET3, aDT4, aAT8, aFT9, and RV1248 are shown in the right uppermost panel. The expected size of the G1-specific amplification product (amplified by aBTl and RV1248) is 337 bp. The expected sizes of G2 and G4 specific products after amplification with these primer mixtures are shown in the bottom three panels. The right half of the bottom panel shows the RT-PCR products of RV from stool samples containing serotype 1 using primers Beg 9 and VP7-1'. HaeIII-digested @X 174 DNA size markers were shown in the center or at the corner of each panel.

Incidence of Rotavirus b y EM/EIA/LA, RNA-PAGE, and RT-PCR Table I shows the incidence of rotavirus by EMiEIAi LA, RNA-PAGE, and RT-PCR (universal primers). Eighty-eight samples of 135 were positive by all three

Serotyping b y EIA-MAb and RT-PCR Table I1 shows the results of serotyping by the EIAMAb and RT-PCR methods. The serotypes obtained by EIA-MAb were identical to those obtained by RT-PCR for 47 samples. One sample contained serotypes 1and 2 by both methods (footnote a in Table 11). Among 88 samples that were nontypeable by EIA-MAb method, 46 could be serotyped by RT-PCR. Forty samples were serotype 1 and five samples were serotype 3. One sample contained serotypes 1and 4 (footnote b in Table 11). Forty-two samples could not be serotyped by either method. DISCUSSION The results of a n RT-PCR method for determining rotavirus VP7 serotypes (G types) are described. The sensitivity for detecting rotavirus ds RNA has been reported to be very high in some systems after removal of inhibitory substances from fecal specimens using chromatographic cellulose fiber powder (CF11) [Wilde e t al., 19901, hydroxyapatite extraction [Gouvea et al., 19911, Isogene kit [Xu et al., 19901, or RNaid [Gentsch

296 et al., 19921. However, the sensitivity we observed for detecting the rotavirus genome was similar to that of PAGE of the viral dsRNA. This result is in agreement with a previous study for PCR typing of rotavirus [Gouvea et al., 1990).The use of a two amplification system, “nested PCR” has been reported to improve the sensitivity in detection of RV [Gouvea et al., 1990; Gentsch e t al., 19921. We attempted first “nested PCR’, but the sensitivity was similar compared to a one-amplification RT-PCR. The reason we did not obtain a n increase in sensitivity is not clear. In our experiments, we used the primers previously reported [Gouvea et al., 19901 and also designed two new primers to detect all group A rotaviruses that we have tested (VP7-1’ in combination with Beg 9) and to confirm typing reactions for serotypes 1 , 2 , 4 , and 8 (consensus primer RV1248 in combination with aBT1, aCT2, aDT4, and aAT8). Using the latter primer mixture we were able to confirm all typing reactions, some of which were very weak (i.e., serotype 2, the right side of middle upper panel in Fig. 2). In another recent report utilizing PCR for G typing, it was noted that G2 and G4 were detected less efficiently by RT-PCR compared with EIA-MAb [Nakagomi et al., 1991 I. The use of our new primer (RV1248)as a replacement for End 9 in the typing mixture increased the sensitivity for detection and confirmation of serotype 2. MAb-EIA is a convenient method for differentiating serotypes; however, in about 30% of samples, serotypes cannot be determined. Some samples do not react while others react for several VP7 serotype-specific MAbs. The results showed that 46 of 88 samples which could not be determined by EIA-MAb, were serotypeable by PCR. In another 47 samples which could be serotyped by EIA-MAb, the correlation demonstrated between the RT-PCR typing and the former method supported a previous report that PCR genotyping (G Typing) [Gouvea et al., 19901 can be a reliable means of identifying rotavirus serotypes, and extremely valuable for sample sets in which a low percentage of specimens are serotypeable by EIA. The untypeables determined by PCR were almost all serotype 1. Coulson 119871 reported different monotypes of serotype 1 which reacted with some serotype 1-VP7-specific MAbs but not others. It will be interesting to determine if any of the 46 PCRtypeable strains react with other type 1-VP7-specific MAbs that recognize other monotypes. In this regard preliminary results in our laboratory with another stool collection using two series of MAbs [Akatani and Ikegami, 1987; Taniguchi et al., 19871 demonstrated that 46 of 360 stool samples were serotype 1 by both MAbs, another 24 of 360 were serotype 1by a MAb from Akatani and coworkers and the other 21 of 360 were serotype 1 by a MAb from Taniguchi and coworkers [Hasegawa et al., in preparation]. In some samples from the current, although rotavirus was found by RNAPAGE they could neither be classified by EIA-MAb nor by serotype specific PCR. These samples will be tested further by neutralization tests and sequence analysis of gene 9 after culture adaptation.

Ushijima et al.

ACKNOWLEDGMENTS We thank Drs. K. Akatani and N. Ikegami for kindly providing anti-rotavirus mouse monoclonal antibodies, and Drs. S. Metazoan and R. Glass for suggestions on how to prepare this manuscript, and John O’Connor for editorial help. This study was partially supported by a grant from the Expanding Program of Immunization, World Health Organization, and by a grant from the Japan Health Science Foundation. REFERENCES Akatani K , Ikegami N (1987): Typing of fecal rotavirus specimens by enzyme-linked immunosorbent assay using monoclonal antibodies. Clinical Virology 156:61-68 (in Japanese). Bishop R, Unicomb L, Barnes G (1991): Epidemiology of rotavirus serotypes in Melbourne, Australia, from 1973-1989. Journal of Clinical Microbiology 29:862-868. Browning GF, Fitzgerald TA, Chalmers RM, Snodgrass DR (1991):A novel group A rotavirus G serotype: Serological and genomic characterization of equine isolate F 123. Journal of Clinical Microbiology 29:2043-2046. Coulson BS (1987): Variation in neutralization epitopes of human rotaviruses in relation to genomic RNA polymorphism. Virology 159:209-216. Estes MK, Cohen J (1989): Rotavirus gene structure and function. Microbiological Review 53:410-449. Flores J, Sears J, Schael IP, Lanata C, Kapikian AZ (1990):Identification of human rotavirus serotype by hybridization to polymerase chain reaction-generated probes derived from a hyperdivergent region of the gene encoding outer capsid protein VP7. Journal of Virology 64:4021 4 0 2 4 . Gentsch J , Glass RI, Woods P, Gouvea V, Gorziglia M, Flores J , Das BK, Bhan MK (1992):Identification of group A rotavirus gene 4 types by polymerase chain reaction. Journal of Clinical Microbiology 30:1365-1373. Gerna G, Sarasini A, di Matteo A, Parea M, Orsolini P, Battaglia M (1988).Identification of two subtypes of serotype 4 human rotavirus by using VP7-specific neutralizing monoclonal antibodies. Journal of Clinical Microbiology 28:1388-1392. Gerna G, Sarasini A, Parea M, Arista S, Miranda P, Brussow H, Hoshino Y, Flores J (1992): Isolation and characterization of two distinct human rotavirus strains with G6 specificity. Journal of Clinical Microbiology 30:9-16. Glass RI, Keith J, Nakagomi 0, Nakagomi T, Askaa J, Kapikian AZ, Chanock RM, Flores J (1985): Nucleotide sequence of the structural glycoprotein VP7 gene of Nebraska calf diarrhea virus rotavirus: Comparison with homologous genes from four strains of human and animal rotaviruses. Virology 141:292-298. Gorziglia M, Larralde G, Kapikian AZ, Chanock RM (1990):Antigenic relationship among human rotaviruses as determined by outer capsid protein VP4. Proceedings of the National Academy of Sciences USA 87:7155-7159. Gouvea V, Glass RI, Woods P, Taniguchi K, Clark HF, Forrester B, Fang ZY (1990):Polymerase chain reaction amplification and typing of rotavirus nucleic acid from stool specimens. Journal of Clinical Microbiology 28:276-282. Gouvea V, Allen JR, Glass RI, Fang ZY, Bremont M, Cohen J, McCare MA, Saif U,Sinarachatant P, Caul EO (19911: Detection of group Band C rotaviruses by polymerase chain reaction. Journal ofClinical Microbiology 29519-523. Heath R, Birch C, Gust I (1986):Antigenic analysis of rotavirus isolates using monoclonal antibodies specific for human serotypes 1, 2 , 3 , and 4, and SA 11.Journal of General Virology 67:2455-2466. Kapikian AZ, Flores J, Midthun K, Hoshino Y, Green KY, Gorziglia M, Taniguchi K, Nishikawa K, Chanock RM, Potash L, PerezSchael I, Dolin R, Christy C, Santosham M, Halsey NA, Clements ML, Levine MM, Losonsky GA, Rennels MB, Gothefors L, Wadell G, Glass RI, Vesikari T, Anderson EL, Belshe RB, Wright PF, Usasawa S (1988):“Development of a Rotavirus Vaccine by a ‘Jennerian’ Approach. Cold Spring Harbor Symposium on New Vac-

Rotavirus and PCR cines. Cold Spring Harbor, NY: Cold Spring Harbor Laboratories, pp 151-159. Nakagomi T, Akatani K, Ikegami N, Katsushima N, Nakagomi 0 (1988): Occurrence of changes in human rotavirus serotypes with concurrent changes in genomic RNA electropherotypes. Journal of Clinical Microbiology 26:25862592. Nakagomi 0, Oyamada H, Nakagomi T (1991):Experience with serotyping rotavirus strains by reverse transcription and two-step polymerase chain reaction with generic and type-specific primers. Molecular and Cellular Probes 5:285-289. Sethabutr 0, Unicomb LE, Holmes IH, Taylor DN, Bishop RF, Echeverria P (1990):Serotyping of human group A rotavirus with oligonucleotide probes. Journal of Infectious Diseases 162:368-372. Shinozaki T, Araki K, Ushijima H, Kim B, Tajima T, Fujii R (1986): Comparison of five methods for detecting human rotavirus in stool specimens. European Journal of Pediatrics 144:513. Taniguchi K, Urasawa T, Morita Y, Greenberg HB, Urasawa S (1987): Direct serotyping of human rotavirus in stools by an enzymelinked immunosorbent assay using serotype 1-,2-, 3-, and 4-specific monoclonal antibodies to VP7. Journal of Infectious Disease 155:1159-1166.

297 Ushijima H, Honma H, Mukoyama A, Shinozaki T, Fujita Y, Kobayashi M, Ohseto M, Morikawa S, Kitamura T (1989):Detection of group C rotaviruses in Tokyo. Journal of Medical Virology 27:299303. Ward RL, McNeal MM, Clemens J D , Sack DA, Rao M, Huda N, Green KY, Kapikian AZ, Coulson BS, Bishop RF, Greenberg HB, Gerna G, Schiff GM (1991): Reactivities of serotyping monoclonal antibodies with culture-adapted human rotaviruses. Journal of Clinical Microbiology 29:449-456. Wilde J, Eiden J , Yolken R (1990): Removal of inhibitory substances from human fecal specimens for detection of group A rotaviruses by reverse transcriptase and polymerase chain reactions. Journal of Clinical Microbiology 28:1300-1307. Xu L, Harbour D, McCrae MA (1990): The application of polymerase chain reaction to the detection of rotaviruses in faeces. Journal of Virological Methods 27:29-38. Yamada 0, Matsumoto T, Nakashima M, Hagari S, Kamahora T, Ueyama H, Kishi Y, Uemura H, Kurimura T (1990): A new method for extracting DNA or RNA for polymerase chain reaction. Journal of Virological Methods 27:203-209.