Relative Frequencies of G and P Types among Rotaviruses from Indian Diarrheic Cow and Buffalo Calves

Abstract

While an increasing number of studies suggest that there is a high prevalence of rotaviruses with P8[11], a typical P type of bovine rotavirus (BRV), among human neonates or infants in India, no data are available on the distribution of G and P types of Indian BRVs. Thus, fecal specimens were collected from cow and buffalo calves under 1 month of age on organized dairy farms in India during the period between 1994 and 1997, and 36 rotavirus-positive specimens were used to determine the relative frequencies of the G and P types of Indian BRVs. As to the G type, G10 was predominant (83%), followed by G6 (6%). The majority (94%) of BRVs had P8[11], and only one isolate possessed P6[1]. The most common combination of G and P types was G10P8[11] (81%), followed by G6P6[1] (3%) and G6P8[11] (3%). The high prevalence of BRVs possessing P8[11] VP4s strongly supports the hypothesis that BRVs may cross the host species barrier and circulate among neonates in India.

Group A rotaviruses, members of the genus Rotavirus within the family Reoviridae, are the leading cause of diarrhea in cow and buffalo calves under 1 month of age on both ranches and dairy farms worldwide (22). The serotype of rotavirus was defined by the antigenicity of two independent neutralization proteins, VP7 and VP4, that constitute the outer capsid of the virion (12). The neutralization specificity carried on VP7 is termed the G serotype, and that carried on VP4 is termed the P serotype (12).

Among BRVs, there are eight G serotypes (G1, G2, G3, G6, G7, G8, G10, and G11) and four P types (P6[1], P7[5], P8[11], and P[12]) (2, 7, 13, 15, 21, 24, 25). Epidemiological studies have shown, however, that BRVs frequently detected in scouring calves carry either G6 or G10 for the G serotype and any one of P6[1], P7[5], and P8[11] for the P serotype, with BRVs that possess G6P7[5] being the most prevalent in various parts of the world (2, 7, 15, 21, 25). While the occurrence of BRV-related diarrhea in India has been documented (9, 10, 23), there are no data available on the distribution of G and P types in Indian BRVs. Such information is of clinical importance since BRV-like strains possessing serotype P8[11], but not P7[5], were reported to be prevalent in neonates and infants in India (1, 3–6, 8). This paper is the first to report the relative frequencies of G and P types among Indian BRVs.

The reference rotavirus strains used in this study were NCDV (G6P6[1]) (16), 0510 (G6P7[5]) (17), UK (G6P7[5]) (26), B223 (G10P8[11]) (27), KK-3 (G10P8[11]) (18), and KN-4 (G6P8[11]) (18). The isolation in cell culture of the BRVs, CR129 and CR156, from the same fecal collection as used in this study has been described elsewhere (11). With the same isolation procedure, two more BRVs, CR231/39 and BR65/255, from the fecal collection (described below) were also adapted to the growth in cell culture.

The cell culture-adapted Indian BRVs as well as the reference BRVs were grown in MA104 cells in the presence of 0.5 μg of trypsin (type IX; Sigma) per ml. Virus particles were purified by ultracentrifugation as described previously (20). The genomic RNAs were extracted from the purified virions with phenol-chloroform and analyzed on 10% polyacrylamide gels as described previously (19).

Fecal specimens were collected from diarrheic calves under 1 month of age during the period between 1994 and 1997 on three dairy farms in Hisar (Haryana) and on one farm each in Ambala (Haryana) and Meerut (Uttar Pradesh), India. These farms are located approximately 200 km apart. The samples were screened for the presence of rotavirus RNA by polyacrylamide gel electrophoresis followed by silver staining as previously described (10), and 36 rotavirus-positive fecal samples (31 from cow calves and 5 from buffalo calves) were identified. To prepare the RNAs for G and P genotyping, the genomic RNA extracts were further purified by using the RNaid kit (BIO 101, Inc., La Jolla, Calif.) according to the instructions of the manufacturer.

The G and P types of the 36 BRVs were determined by the reverse transcription (RT)-PCR assays described by Isegawa et al. (14). In brief, the G typing assay consisted of three steps: (i) RT of genomic RNAs with a pair of generic primers (Bov9Com5, 5′-TGTATGGTATTGAATATACCAC-3′, and Bov9Com3, 5′-TCACATCATACAACTCTAATCT-3′); (ii) the first PCR amplification, of a near-full-length VP7 gene with Bov9Com5 and Bov9Com3; and (iii) the second (nested) PCR amplification, with the 5′ generic primer (Bov9Com5) and a cocktail of typing primers specific for G6 and G10. The sequences of the G6 and G10 primers are 5′-GGTATCAGCTATTTCGTTTGAT-3′ and 5′-AACGTTCTAGTATTTGTGGTCT-3′, respectively. Since these two typing primers were selected such that each primer is located at a different distance from the 5′ end of the gene, the size of the second PCR product indicated the G type of the strain tested. The P typing assay consisted of essentially the same three steps as described above but with two generic primers (Bov4Com5, 5′-TTCATTATTGGGACGATTCACA-3′, and Bov4Com3, 5′-CAACCGCAGCTGATATATCATC-3′) and three typing primers specific for P6[1] (5′-TTAAATTCATCTCTTAGTTCTC-3′), P7[5] (5′-GGCCGCATCGGATAAAGAGTCC-3′), and P8[11] (5′-TGCCTCATAATATTGTTGGTCT-3′). In this assay, the size of the second PCR product indicated the P type of the strain tested.

Figure 1 shows the electropherotypes of the cell culture-adapted Indian BRVs and those of the reference BRVs. These electropherotypes were similar in that they all demonstrated long RNA patterns and their differences were within the range of variation of the reference BRVs. Of note is that, like NCDV, CR231/39 had a fast-moving genome segment 4, while other BRVs had slowly migrating genome segment 4s. Interestingly, as previously observed by Suzuki et al. (25), the fast-migrating genome segment 4s corresponded to P6[1] (Fig. 2). Three other strains were identified as possessing P8[11] (Fig. 2). The G types of the cell culture-adapted Indian BRVs were also unambiguously identified; CR231/39 was G6, and CR129, CR156, and BR65/255 were G10 (Fig. 2).

FIG. 1.

Electropherotypes of Indian BRV isolates and prototype BRV strains NCDV, B223, KN-4, and UK. The numbers on the left designate genome segments.

FIG. 2.

Determination of the P and G types of cell culture-adapted Indian BRV strains by PCR after RT by the method of Isegawa et al. (14). The numbers on the left are sizes (in base pairs) of the DNA fragments.

The G and P typing assays were successfully applied to the genomic RNAs extracted from the 36 fecal specimens (Table 1). Of these specimens, 31 (86%) yielded single bands upon both the G and P typing assays whereas 5 (14%) yielded doublets upon either the G or P typing assays. Those specimens that yielded doublets were interpreted as containing a mixture of BRVs carrying different serotypes. Since the RT-PCR assays used in this study detected only the coinfections with BRVs carrying different serotypes, even the mixed-infection rate of 14% may be an underestimate. In fact, Suzuki et al. (25) found that 40% of cell culture isolates from diarrheic cows in Japan contained more than one BRV.

TABLE 1.

Relative frequencies of various combinations of the G and P serotypes detected in rotavirus-positive samples from Indian diarrheic cow and buffalo calves

	No. (%) of rotaviruses identified as:
	G6	G10	G6 + G10	Total
P6[1]	1 (3)	0 (0)	0 (0)	1 (3)
P8[11]	1 (3)	29a (81)	4a (11)	34 (94)
P6[1]+P8[11]	0 (0)	1b (3)	0 (0)	1b (3)
Total	2 (6)	30 (83)	4 (11)	36 (100)

^aTwo of these rotaviruses were from buffalo calves. 

^bThis rotavirus was from a buffalo calf. 

For the relative frequencies of the G and P types among Indian BRVs, the most conspicuous findings were that over 95% of BRVs carried P8[11] VP4s and that the majority (81%) of Indian BRVs were G10P8[11] strains (Table 1). Although interpretation of our study is limited by a small sample size and sample collection in restricted regions in India, this finding is in sharp contrast to the distribution of BRV serotypes elsewhere in the world. Wherever the serotypic survey was conducted, BRVs possessing G6P7[5], represented by prototype strain UK, predominated (2, 7, 15, 21, 25). This finding is of particular interest in view of the unusual prevalence of P8[11] strains in Indian neonates and infants (1, 3–6, 8). In particular, one strain (I321) possessing this serotype combination (G10P8[11]) was isolated from a neonate in Bangalore, India, and was shown by RNA-RNA hybridization and by sequence analysis to be related to BRVs (5, 6). Moreover, the incidence of asymptomatic infection of neonates with this I321-like G10P8[11] rotavirus in Bangalore during the 7-year period from 1988 to 1994 was consistently high (about 34%) (1). Interestingly, however, serotype G10 strains were not detected in symptomatically infected children in the same area (1).

Similarly, another rotavirus strain possessing a P8[11] VP4, strain 116E, was isolated from asymptomatic neonates from Delhi, India, and characterized by RNA-RNA hybridization and by sequence analysis to be a possible reassortant between human and bovine rotaviruses (4, 8). Furthermore, the infection of human neonates with strains related to 116E was the most common (frequency, 67%) in the vicinity of the place where 116E was originally isolated (3). Given the high proportion of rotaviruses possessing P8[11] VP4s among both human rotaviruses and BRVs, it is likely that natural reassortants between human and bovine rotaviruses are emerging and circulating in neonates in India. The close interaction of the majority of the Indian population with cattle makes this possibility more plausible.