tqL JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1992, p. 781-785 Vol. 30, No. 4 0095-1137/92/040781-05$02.00/0 Copyright © 1992, American Society for Microbiology Distribution of Serotypes of Human Rot/virus INISA in Different Populations Nob PATRICIA A. WOODS,' JON GENTSCH,' VERA GOUVEA,' LEONARDO MATA4,2?"19V194141. 1111 ALBERTO SIMHON,2 MATHURAM SANTOSHAM,3 ZHI-SHENG BAI,4 SHOZO URASAWA,5 AND ROGER 1. GLASS' Viral Gastroenteritis Unit, Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control, Atlanta, Georgia 30333'; Instituto de Investigations en Salud, Universidad de Costa Rica, Ciudad Universitaria Rodrigo Facio, San Jose, Costa Rica2; Department of International Health, Johns Hopkins School of Hygiene and Health, Baltimore, Maryland 212053; Lanzhou Institute of Biomedical Products, Lanzhou, China"; and Department of Hygiene and Epidemiology, Sapporo Medical College, Sapporo 060, Japans Received 29 July 1991/Accepted 30 December 1991 Serotyping is a useful tool to study the cpidemiologic characteristics of rotaviruses in large populations and to assess the need for a vaccine to protect against all strains. By using an enzyme immunoassay with serotype-specific monoclonal antibodies to the four most common rotavirus serotypes, we analyzed 1,183 rotavirus-positive specimens from 16 stool collections in eight countries on four continents that were obtained from 1978 to 1989. Of the 926 strains (78%) that could be serotyped, 48% were serotype 1, 8% were serotype 2, 15% were serotype 3, and 7% were serotype 4. Twenty-two percent had insufficient numbers of double- shelled virus particles to react with the monoclonal antibody of the VP4 rotavirus protein and therefore could not be serotyped. Our results indicate that vaccines being developed must provide the greatest coverage against serotype 1 and that the serotype distribution cannot be predicted currently by the geographic area or prevalence in the preceding year. dookyibAii logo Group A rotaviruses are the most important cause of hospitalizations and deaths due to diarrhea in children under 2 years of age in both developed and developing countries (7, 9, 21, 22). Virtually all children are infected by 5 years of age. While the detection of rotaviruses by enzyme immu- noassays (EIAs) has become a simple, routine procedure, the serotyping of individual strains can provide additional insights into the epidemiologic features of rotavirus diarrhea. From recent studies of rotavirus surveillance, geographic and temporal patterns of disease in the United States and throughout the world have been identified; these might be clarified if more detailed knowledge of the distribution of the strains were available (26). Similarly, while early vaccine trials with the bovine rotavirus vaccine (RIT 4237, serotype 6) demonstrated good heterotypic protection against rotavi- ruses of serotypes 1 and 2 (41), subsequent studies with the monovalent rhesus rotavirus vaccine (serotype 3) suggested that homotypic protection was probably important as well (10). Consequently, the distribution of serotypes in a popu- lation could alter the effectiveness of vaccines being tested. Four major and two minor serotypes of human rotavirus have been identified on the basis of the major neutralization glycoprotein, VP7, which is present in the outer capsid of the virus. In the past, the serotyping of rotavirus has been dependent on the ability to grow different viral strains in cell culture and neutralize them with type specific antiserum (43). The recent development of monoclonal antibodies (MAbs) that specifically and selectively bind to the VP7 glycoprotein of each of the four most-common rotavirus serotypes (serotypes 1 to 4) has made it possible to serotype large collections of specimens by EIA (35). To examine the global distribution of rotavirus serotypes * Corresponding author. in different geographic locations and at different times, we have used an EIA with serotype-specific MAbs reactive to rotavirus specimens from 16 separate studies conducted in eight different countries on four continents. These results are compared with those reported in 28 other studies from 23 different countries reported in the literature. Together, these findings suggest that vaccines being developed must provide particularly good coverage for serotype 1, the most common serotype, and that year-to-year variations in the distribution of serotypes may provide clues to the reservoirs and trans- mission of rotaviruses. MATERIALS AND METHODS Specimens. Stool specimens for the serotyping of rotavi- ruses were received frozen and were stored at -70°C until used. An approximately 20% stool suspension was made in Tris-buffered saline, pH 7.2, (Vogts TBS) and clarified by low-speed centrifugation. Samples from seven collections were obtained from children hospitalized with diarrhea in the following locations: Beer-Sheva, Israel (1986-1987) (8); Seoul, Korea (1987-1988) (23); Puriscal, Costa Rica (1983- 1987) (34a); Mexico City, Mexico (1986-1987); Lanzhou, China (1982-1986); Seattle, Wash. (1986-1987); and the White River Indian Reservation, Ariz. (1981-1985). Other collections came from children with diarrhea in day-care. centers in Phoenix, Ariz. (1981-1983) (2), from specimens submitted to a diagnostic laboratory that serves the south- eastern United States (1987-1988), and from newborns with both symptomatic and asymptomatic infections in South Africa (1978) (29). The five remaining collections from the United States were serotyped as part of an investigation to examine the distribution of rotaviruses in the United States for the period from 1987 to 1989 (19). For 3 of the 16 collections (Costa Rica, China, and White River, Ariz.), 781 782 WOODS ET AL. 1. CLIN. MICROBIOL. TABLE 1. Analysis of rotavirus serotypes from different geographic locations and different years Place Yr No. of samples tested % (n) belonging to serotype: 2 3 4 NS° United States Denver, Colo. 1987-1989 60 75 (45) 2 (1) 23 (14) 0 (0) 0 (0) Seattle, Wash. 1986-1989 40 93 (37) 0 (0) 0 (0) 0 (0) 7 (3) Phoenix, Ariz. 1981-1983 29 55 (16) 7 (2) 21 (6) 0 (0) 17 (5) San Diego, Calif. 1988-1989 35 63 (22) 3 (1) 34 (12) 0 (0) 0 (0) Houston, Tex. 1987-1988 23 26 (6) 0 (0) 74 (17) 0 (0) 0 (0) White River, Ariz. 1981-1985 39 36 (14) 15 (6) 28 (11) 8 (3) 13 (5) Philadelphia, Pa. 1987-1989 30 77 (23) 3 (1) 17 (5) 0 (0) 3 (1) Buffalo, N.Y. 1987-1989 26 84 (22) 4 (1) 8 (2) 0 (0) 4 (1) Southeast region 1987-1988 196 83 (163) 1 (1) 8 (16) (0) 8 (16) Mexico 1986-1989 166 24 (40) 16 (26) 31 (51) 6 (11) 23 (38) Costa Rica 1983-1987 184 8 (15) 7 (12) 6 (11) 28 (52) 51 (94) Peru 1986 10 50 (5) 20 (2) 0 (0) 30 (3) 0 (0) South Africa 1978 31 0 (0) 16 (5) 52 (16) 0 (0) 32 (10) China 1982-1986 80 47 (38) 29 (23) 14 (11) 0 (0) 10 (8) Korea 1988 138 52 (71) 4 (6) 1 (1) 0 (0) 43 (60) Israel 1987 97 55 (52) 10 (10) 0 (0) 19 (19) 16 (16) Total 1,183 48 (569) 8 (97) 15 (173) 7 (87) 22 (257) Positive for VP4 but not for serotypes 1 to 4. specimens were available for a 5-year period and were analyzed both together and for individual years to evaluate changes in the distribution of serotypes in the same area over time. Several points should be noted about these collections: (i) the specimens were collected from several different groups of children with diarrhea including newborns, children in day-care centers, and ambulatory and hospitalized patients; (ii) some collections span a limited time period; (iii) some collections are small, such as those from South Africa and Peru; and (iv) some of the collections which span several years have limited numbers of samples. They are not meant to be representative of patterns in a specific place or time but to demonstrate the diversity of strains seen in a large reference laboratory. EIA for serotyping. Stool specimens were serotyped with an EIA by using MAbs KU-4, S2-2010, YO-1E2, and ST-2G7, which recognize specific serotype 1, 2, 3, and 4 neutralization epitopes, respectively, in the outer capsid protein VP7, and YO-2C2, a MAb which recognizes the VP4 outer capsid protein of all four serotypes. The EIA was performed as described previously with some modifications (35). Microtiter plates were coated with a 1:10,000 dilution of ascitic fluid in phosphate-buffered saline (PBS), pH 7.2. After overnight incubation, the plates were washed three times with PBS-T (PBS containing 0.05% Tween) and incu- bated with 1% BSA-PBS (2 h at 37°C). The 1% BSA-PBS was removed, and 50 1.1.I of a 20% stool suspension diluted 1:3 in 1% BSA-PBS was added in duplicate to the wells; the plates were then incubated overnight at 4°C. After the plates were washed three times with PBS-T, pooled rabbit serum made against the four rotavirus serotypes was added at a 1:50,000 dilution in 1% BSA-PBS. After three additional washes in PBS-T, 50 pi of horseradish peroxidase-conju- gated goat anti-rabbit immunoglobulin G (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, Md.) was added at a dilution of 1:12,000 in 1% BSA-PBS. The plates were incubated for 1 h at 37°C and washed three times with PBS-T, and 100 IA of the chromogen tetramethylbenzidine was added. After a 10-min incubation at room temperature, the reaction was stopped with 25 µI of 2 M H2SO4, and the optical density at 450 nm (013450) was determined with a microtiter plate reader (Dynatech MR600). A serotype was assigned when the OD value for the reaction with the MAb corresponding to that serotype was higher than 0.2 and the OD value for the reaction corresponding to that serotype was at least twice the mean OD of the other three MAbs reactive to the same sample_ Comparison with published results. To examine global trends in the distribution of rotavirus serotypes, we identi- fied studies of rotavirus diarrhea from a Medline search of papers published in English from 1983 to 1991, by using the keywords "rotavirus" and "serotypes." Studies that had used any serotyping method and had reported serotyping results for more than 25 strains were reviewed. Vaccine studies were not included because of a possible serotype- specific protection induced by the vaccine_ RESULTS A total of 1,183 rotavirus-positive specimens from the 16 collections were assayed for serotype. Of these, 926 (78%) had sufficient double-shelled virus particles to react with the VP4 MAb so that a serotype could be assigned (Table 1). No mixed-serotype samples were observed in this study. Over- all, serotype 1 was the most prevalent serotype (48%) and VOL. 30, 1992 DISTRIBUTION OF HRV SEROTYPES 783 TABLE 2. Changes in rotavirus serotypes during 5 years in three countries Place and yr % (n) belonging to serotype: No. of samples tested 1 2 3 4 NS' Costa Rica 1983 0 (0) 0 (0) 0 (0) 77 (31) 22 (9) 40 1984 0 (0) 0 (0) 0 (0) 62 (18) 38 (11) 29 1985 0 (0) 14 (5) 14 (5) 3 (1) 68 (24) 35 1986 0 (0) 19 (7) 17 (6) 5 (2) 58 (21) 36 1987 34 (15) 0 (0) 0 (0) 0 (0) 66 (29) 44 Total 8 (15) 7 (12) 6 (11) 28 (52) 51 (94) 184 Lanzhou, China 1982 76 (16) 19 (4) 0 (0) 0 (0) 5 (1) 21 1983 4 (1) 69 (16) 4 (1) 0 (0) 22 (5) 23 1984 62 (10) 0 (0) 31 (5) 0 (0) 6 (1) 16 1985 56 (9) 12 (2) 31 (5) 0 (0) 0 (0) 16 1986 50 (2) 25 (1) 0 (0) 0 (0) 25 (1) 4 Total 47 (38) 29 (23) 14 (11) 0 (0) 10 (8) 80 Ariz. 1981 0 (0) 0 (0) 89 (8) 0 (0) 11 (1) 9 1982 33 (2) 0 (0) 50 (3) 0 (0) 17 (1) 6 1983 80 (8) 0 (0) 0 (0) 0 (0) 20 (2) 10 1984 9 (1) 54 (6) 0 (0) 27 (3) 9 (1) 11 1985 100 (3) 0 (0) 0 (0) 0 (0) 0 (0) 3 Total 36 (14) 15 (6) 28 (11) 8 (3) 13 (5) 39 ° Positive for VP4 but not for serotypes 1 to 4. the predominant serotype in 12 of the 16 collections. The South African collection, which came from an outbreak of rotavirus in a newborn unit, contained samples with both serotype 2 and 3 specificities and no strains of serotype 1. Serotype 4 was the least common serotype (7%), present in specimens from children in Latin America, Israel, and the White River Indian Reservation, Ariz., but not those from Asia, Africa, and the rest of the United States. Three of the collections represent samples obtained every year for 5 years and provide some insight into the variability of rotavirus strains in a single location over time (Table 2). In Costa Rica, serotype 4 was the only strain identified for 2 years, after which it was displaced during the next 2 years by serotypes 2 and 3, which gave way the last year to serotype 1. A large number of these strains (51%) could not be serotyped, and this inability to serotype was not related to the age of the specimens in this collection. In Lanzhou, China, serotype 1 was the predominant serotype for 4 of the 5 years, serotype 2 was predominant for a single year, serotype 3 strains were preyalent in a third of cases for two consecutive years, and serotype 4 was absent. In Arizona, with the smallest number of strains, serotypes 1, 2, and 3 shared predominance from year to year. Data from 28 studies of rotavirus serotypes published previously were identified in the Medline search and re- viewed (Table 3). A total of 5,419 specimens were tested, and 74% were serotyped. Serotypes were determined in 26 studies by EIAs with MAbs, in 1 study (Italy) by solid-phase immunoelectron microscopy, and in 1 study (Kenya) by neutralization. Overall, serotype 1 was the most prevalent serotype (61%), and it was the predominant serotype in TABLE 3. Global review of serotyping studies Place Yr Refer- ence No. of samples % Belonging to serotype 1 2 3 4 NS' N. America Canada 1984 3 41 61 10 29 0 7 Mexico 1984-1987 31 132 43 17 15 25 33 United States 1979-1989 28, 30 1,676 59 3 31 7 26 S. America Argentina 1983-1986 16 123 41 19 14 26 44 Brazil 1984-1985 27 57 29 21 10 40 9 Peru 1983 3 54 20 62 6 12 7 Venezuela 1981-1983 12 134 49 16 20 15 24 Africa Africa 1983-1985 13 152 72 15 13 0 6 Gambia 1982-1984 33 24 41 36 23 0 8 Kenya 1982-1983 40 16 59 33 8 0 25 Asia Bangladesh 1988 1 123 42 30 3 25 71 Bangladesh 1985-1986 42 143 31 16 15 38 0 Burma 1986 3 17 25 19 37 19 6 India 1983-1985 6 117 48 9 6 37 73 Indonesia 1978-1979 5 111 2 9 53 36 48 Japan 1984 38 16 63 12 25 0 33 Japan 1986-1988 39 562 51 31 11 7 25 Pakistan 1985 3 12 0 0 0 100 0 Sri Lanka 1987 3 5 0 0 0 100 0 Thailand 1983-1984 32 88 31 29 0 40 34 1987-1988 Australia 1986-1987 37 344 95 4 .5 .5 38 1973-1986 36 647 67 9 9 15 37 1973-1989 4 943 75 7 3 15 29 Europe England 1983-1988 3 353 55 20 16 9 2 Finland 1984 3 135 87 5 3 5 4 1986-1987 Italy 1981-1985 14 129 72 14 0 14 3 Sweden 1983 3 32 50 34 16 0 0 Total 5,419 61 10 16 13 26 Percentage of HRV-positive samples whose serotypes could not be determined. North America, Africa, Australia, and Europe in 19 of the 28 studies analyzed. Serotypes 2, 3, and 4 were evenly distrib- uted, but serotype 4 was absent in studies from Africa. DISCUSSION Serotyping is a useful tool to study the epidemiologic features of rotaviruses in large populations and to assess the need for a vaccine to protect against all strains. This study indicates that serotype 1 is the most common global serotype in circulation and any vaccine must provide good protection against it. At the same time, from three collections spanning 5 years and from many collections with 1 year of data alone, great variations in the distribution of serotypes occur over time and from place to place. These indicate that, to be highly effective, a vaccine must protect against all of these serotypes at the same time. Furthermore, collections having 784 WOODS ET AL. a high proportion of nonserotypeable strains (i.e., samples that were VP4-positive by EIA but could not be serotyped) must be examined further for other less-common serotypes (S8, S9, and S12) or new serotypes yet to be described. While 78% of all strains could be serotyped using the four MAbs, the rates of nontypeable isolates varied greatly, from 50% in Costa Rica to none in Peru. The serotyping technique described herein is easily repro- ducible and suitable to study large collections of epidemio- logic interest. At the same time, it is based on MAbs binding to single VP7 epitopes. Recent studies have demonstrated that subtypes of rotavirus serotypes 1 (4) and 4 (15) exist, on the basis of the binding of their VP7 polypeptides to some neutralizing, serotype-specific MAbs but not others. These subtypes have been designated monotypes by some workers (4). Although insufficient intact virions may account for most nontypeable strains (about 30% in most studies), data that some isolates of rotaviruses do not bind to all MAbs com- monly used in serotyping have been presented (42). Due to these epitope variations of isolates, it has been suggested that the range of this method could be enhanced by the use of MAbs specific for more than one VP7 epitope (42). The range of the method may also be enhanced by the addition of MAbs for serotypes 8, 9, and 12, which appear to be uncommon except in certain geographic regions. Other techniques for serotyping nontypeable strains include the polymerase chain reaction (18), nucleotide sequencing (20), hybridization methods (11, 34), and solid-phase immunoelec- tron microscopy (15). The polymerase chain reaction is also amenable to the identification of new serotypes through the addition of new primers as sequence information becomes available. Recent studies have delineated the human rotaviruses into three VP4 serotypes and one subtype (17). These results suggest that a more-complete characterization of strains in the future will depend upon the neutralization associated with VP4 as well as VP7 polypeptides. Techniques already developed for VP7 serotyping may also be applicable for VP4. Neutralizing MAbs that can be used to separate human rotaviruses into four overlapping VP4-antigenic groups have already been reported (24). Although these antibodies are not specific enough to distinguish all VP4-antigenic groups in a solid-phase assay, it is plausible that these and other antibodies may be used to develop an EIA-based serotyping method for VP4. Similarly, a hybridization-based method for the discrimination of gene 4 types has been reported (25). When reduced to practice, these methods will allow a more complete understanding of the prevalence of rotavirus genes important for neutralization and help in the design of future vaccine candidates. ACKNOWLEDGMENTS We thank Kyung-Hee Kim, Guillermo Ruiz-Palactos, Ron Dagan, Bob Gilman, and Martin Smith for submitting collections; James Allen, Judy Lew, and Koki Taniguchi for helpful discussions and critical reading of the manuscript; and Diane Mott and Cesar Carcamo for valuable assistance in preparing the manuscript. REFERENCES 1. Ahmed, M. U., K. Taniguchi, N. Kobayashi, T. Urasawa, F. Wakasugi, M. Islam, H. Shaikh, and S. Urasawa. 1989. Charac- terization by enzyme-linked immunosorbent assay using sub- group- and serotype-specific monoclonal antibodies of human rotavirus obtained from diarrheic patients in Bangladesh. J. Clin. Microbial. 27:1678-1681. 2. Bartlett, A. V_, M. Moore, G. W. Gary, K. M. Stark°, J. J. Erben, and B. A. Meredith. 1985. Diarrhea] illness among J. CLEN. MICROBIOL. infants and toddlers in day care centers. 1. Epidemiology and pathogens. J. Pediatr. 107:495-502. 3. Beards, G., U. Desselberger, and T. Flewett. 1989. Temporal and geographical distributions of human rotavirus serotypes, 1983 to 1988. J. Clin. Microbiol. 27:2827-2833. 4. Bishop, R. F., L. E. Unicomb, and G. L. Barnes. 199L Epidemi- ology of rotavirus serotypes in Melbourne, Australia, from 1973 to 1989. J. Clin. Microbiol. 29:862-868. 5. Bishop, R. F., L. E. Unicomb, Y. Soenarto, H. Suwardii, •. •. Risanto, and G. L. Barnes. 1989. Rotavirus serotypes causing acute diarrhoea in hospitalized children in Yogyakarta, Indone- sia during 1978-1979. Arch. Virol. 107:207-213. 6. Brown, D. W. G., M. M. Mathan, M. Mathew, R. Martin, G. M. Beards, and V. I. Mathan. 1988- Rotavirus epidemiology in Vellore, South India: group, subgroup, serotype, and electro- phoretype. J. Clin. Microbial. 26:2410-2414. 7. Cook, S. M. R. I. Glass, C. W. LeBaron, and M.-S. Ho. 1990. Global seasonality of rotavirus infections. Bull. W.H.O. 68:171- 177. 8. Dagan, R., and R. I. Glass. 1989. Epidemiology and clinical aspects of rotavirus diarrhea in Bedouin and Jewish infants and young children in the Negev region of Israel. Pediatr. Infect. Dis. J. 9:314-321. 9. De Zoysa, I., and R. G. Feachem. 1985. Interventions for the control of diarrhoea) diseases among young children: rotavirus and cholera immunization. Bull. W.H.O. 63:569-583. 10. Flores, J., I. Perez-Schael, M. Gonzalez, D. Garth, M. Perez, N. Mond, W. Cunto, R. M. Chanock, and A. Z. Kapikian. 1987. Protection against severe rotavirus diarrhoea by rhesus rotavi- rus vaccine in Venezuelan infants. Lancet i.:1882-1884. 11. Flores, J., J. Sears, I. P. Schael, L. White, D. Garcia, C. Lomita, and A. Z. Kapildan. 1990. Identification of human rotavirus serotype by hybridization to polymerase chain reaction-gener- ated probes derived from a hyperdivefgent region of the gene encoding outer capsid protein VP7. J. Virol. 64:4021-4024. 12. Flores, J., K. Taniguchi, K. Green, I. Perez-Schael, D. Garcia, J. Sears, S. Urasawa, and A. Z. Kapildan. 1988. Relative frequen- cies of rotavirus serotypes 1, 2, 3, and 4 in Venezuelan infants with gastroenteritis. J. Clin. Microbiol. 26:2092-2095. 13. Georges-Courbot, M. C., A. M. Berand, G. M. Beards, A. D. Campbell, J. P. Gonzalez, A. J. Georges, and T. H. Hewett. 1988. Subgroups, serotypes, and electrophoretypes of rotavirus isolated from children in Bangui, Central African Republic. J. Clin. Microbial. 26:668-671. 14. Germ, G., S. Arista, N. Passarani, A. Sarasini, and M. Battaglia. 1987. Electropherotype heterogeneity within serotypes of hu- man rotavirus strains circulating ie Italy. Arch. Vim!. 95:129- 135. 15. Gerna, G., A. Sarasini, A. di Matte°, M. Pares, P. Orsolini, and M. Battaglia. 1988. Identification of two subtypes of serotype 4 human rotavirus by using VP7-specific neutralizing monoclonal antibodies. J. Clin. Microbiol. 28:1388-1392. 16. Gomez, J., M. K. Estes, D. 0. Matson, R. Bellinzoni, A. Alvarez, and S. Grinstein. 1990. Serotyping of human rotaviruses in Argentina by ELISA with monoclonal antibodies. Arch_ Virol. 112:249-259. 17. Gorziglia, M., G. Larralde, A. Z. Kapikian, and R. M. Chanock. 1990. Antigenic relationships among human rotaviruses as de- termined by outer capsid protein VP4. Proc. Natl. Acad. Sci. USA 87:7155-7159. 18. Gouvea, V., R. I. Glass, P. Woods, K. Tanigulchi, H. F. Clark, B. Forrester, and Z. Y. Fang. 1990. Polymerase chain reaction amplification and typing of` rotavirus nucleic acielt from stool specimens. J. Clin. Microbiol. 28:276-282. 19. Gouvea, V., M.-S. Ho, R. I. Glass, P. Woods, B. Forrester, C. Robinson, R. Ashley, M. Riepenhoff-Talty, H. F. Clark, K. Taniguchi, E. Meddix, B. McKellar, and L. Pickering. 1990. Serotypes and electrophoretypes of human rotavirus in the USA: 1987-1989. J. Infect. Dis. 162:362-367. 20. Green, K. Y., J. F. Sears, K. Taniguchi, K. Midthun, Y. Hoshirm, M. Gorziglia, K. Nishikawa, S. Urasawa, A. Z. Kapikian, R. M. Chanock, and J. Flores. 1988. Prediction of human rotavirus serotype by nucleotide sequence analysis of VoL. 30, 1992 DISTRIBUTION OF HRV SEROTYPES 785 the VP7 protein gene. J. Virol. 62:1819-1823. 21. Ho, M. S., R. 1. Glass, P. F. Pinsky, and L. J. Anderson. 1988. Rotavirus as a cause of diarrhea' morbidity and mortality in the United States. J. Infect. Dis. 158:1112-1116. 22. Kapikian, A. Z., and R. M. Chanock. 1985. Rotaviruses, p. 863-906. In B. N. Fields (ed.), Virology, Raven Press, New York. 23. Kim, K.-H., I.-S. Suh, J. M. Kim, C. W. Kim, and Y.J. Cho. 1989. Etiology of childhood diarrhea in Korea. J. Clin. Micro- biol. 27:1192-1196. 24. Kobayashi, N., K. Taniguchi, and S. Urasawa. 1990. Identifica- tion of operationally overlapping and independent cross-reac- tive neutralization regions on human rotavirus VP4. J. Gen. Virol. 71:2615-2623. 25. Larralde, G., and J. Flores. 1990. Identification of gene 4 alleles among human rotaviruses by potymerase chain reaction-derived probes. Virology 179:469-473. 26. LeBaron, C. W., J. Lew, R. I. Glass, J. M. Weber, G. M. Ruiz-Palacios, and The Rotavirus Study Group. 1990. Annual rotavirus epidemic patterns in North America: results of a five-year retrospective survey of 88 centers in Can: da, Mexico, and the United States. JAMA 264:983-988. 27. Linhares, A. C., Y. B. Gabbay, J. D. P. Mascarenhas, R. B. Freitag, T. H. Flevvett, and G. M. Beards. 1988. Epidemiology of rotavirus subgroups and serotypes in Belem, Brazil: a three- year study. Ann. Inst. Pasteur Viral. 139:89-99. 28. Matson, D. 0., M. K. Estes, J. W. Burns, H. B. Greenberg, K. Taniguchi, and S. Urasawa. 1990. Serotype variation of human group A rotaviruses in two regions of the USA. J. infect. Dis. 162:605-614. 29. Oelofsen, M. J., E. H. Venter, and M. S. Smith. 1985. Noso- comial rotavirus infections in a South African hospital nursery for white infants. S. Air. Med. J. 68:394-396. 30. O'Ryan, M., D. Matson, M. Estes, A. Bartlett, and L. Pickering. 1990. Molecular epidemiology of rotavirus in children attending day care centers in Houston. J. Infect. Dis. 162:810-816. 31. Padilla-Noriega, L., C. Arias, S. Lopez, F. Puerto, D. Snodgrass, K. Taniguchi, and H. Greenberg. 1990. Diversity of rotavirus serotypes in Mexican infants with gastroenteritis. J. Clin. Mi- crobiol. 28:1114-1119. 32. Pongsuwanne, Y., K. Taniguchi, M. Choonthanom, M. Chi- wakul, T. Susansuok, S. Saguanwongse, C. Jayavasu, and S. Urasawa. 1989. Subgroup and serotype distributions of human, bovine, and porcine rotavirus in Thailand. J. Clin. Microbiol. 27:1956-1960. • 33. Rowland, M., S. Gob, K. Williams, A. Campbell, G. Beards, R. Sanders, and T. Flewett. 1985. Epidemiological aspects of rota- virus infection in young Gambian children. Ann. Trop. Paediatr. 5:23-28. 34. Sethabutr, 0., L. E. Unicomb, I. H. Holmes, D. N. Taylor, R. F. Bishop, and P. Echeverria. 1990. Serotyping of human group A rotavirus with oligonucleotide probes. J. Infect. Dis. 162:368- 372. 34a.Simhon, A., L. Mats, M. Vives, L. Rivera, S. Vargas, G. Ramirez, L. Lizano, G. Catarinella, and J. Azofeifa. 1985. Low endemicity and low pathogenicity of rotavirus among rural children in Costa Rica. J. Infect. Dis. 152:1134-1142. 35. Taniguchi, K., T. Urasawa, Y. Morita, H. B. Greenberg, and S. Urasawa. 1987. Direct serotyping of human rotavirus in stools using serotype 1-, 2-, 3-, and 4-specific monoclonal antibodies to VP7. J. Infect. Dis. 155:1159-1166. 36. Unicomb, L., B. Coulson, and R. Bishop. 1989. Experience with an enzyme immunoassay for serotyping human group a rotavi- ruses. J. Clin. Microbiol. 27:586-588. 37. Unicomb, L. E., and R. F. Bishop. 1989. Epidemiology of rotavirus strains infecting children throughout Australia during 1986-1987: a study of serotype and RNA electropherotype. Arch. Viral. 106:23-34. 38. Urasawa, S., T. Urasawa, K. Taniguchi, and S. Chiba. 1984. Serotype determination of human rotavirus isolates and anti- body prevalence in pediatric population in Hokkaido, Japan. Arch. Viral. 81:1-12. 39. Urasawa, S., T. Urasawa, K. Taniguchi, F. Wakasugi, N. Koba- yashi, S. Chiba, N. Sakurada, M. Morita, 0. Morita, M. Tokieda, H. Kawamoto, Y. Minekawa, and M. Obesto. 1989. Survey of human rotavirus serotypes in different locales in Japan by enzyme-linked immunosorbent assay with monoclonal antibodies. J. Infect. Dis. 160:44-51. 40. Urasawa, T., S. Urasawa, Y. Chiba, K. Taniguchi, N. Kobayashi, L. N. Mutanda, and P. M. Tukei. 1987. Antigenic characteriza- tion of rotaviruses isolated in Kenya from 1982 to 1983. J. Clin. Microbial. 25:1891-1896. 41. Vesikari, T., E. Isolauri, and E. D'Hondt. 1984. Protection of infants against rotavirus diarrhea by RIT 4237 attenuated bovine rotavirus strain vaccine. Lancet 1:977-981. 42. Ward, R., M. McNeal, J. Clemens, D. Sack, M. Rao, N. Huda, K. Green, A. Kapikian, B. Coulson, R. Bishop, H. Greenberg, G. Gerna, and G. Schiff. 1991. Reactivities of serotyping monoclo- nal antibodies with culture-adapted human rotaviruses. J. Clin. Microbiol. 29:449-456. 43. Wyatt, R. G., H. B. Greenberg, and W. D. James. 1984. Definition of human rotavirus serotypes by plaque reduction assay. Infect. Immun. 37:110-115.