Comparison of NSP4 protein between group A and B human rotaviruses: detection of novel diarrhea-causing sequences in group B NSP4

Summary. The human group B rotavirus is a causative agent of severe adult diarrhea. In this study, we analyzed the NSP4 structure of a group B rotavirus strain, CAL-1, and determined whether enterotoxin activity was present in CAL-1 NSP4. CAL-1 NSP4 was comprised of 219 amino acids which was longer than group A and C rotavirus NSP4, and the primary structures of their sequences differed considerably. However, CAL-1 NSP4 had an enterotoxin-like sequence (residues 106–127) that was only 27% identical to the enterotoxin region of NSP4 of KUN (a group A rotavirus strain) at residues 114–135. Interestingly, both of the synthetic peptides, one (residues 99–128) containing the enterotoxin-like sequence and the other (residues 191–219) containing 29 C-terminal amino acids of CAL-1 NSP4, induced diarrhea in 5.5-day-old mice, but not in 17.5-day-old mice, when administered parenterally. Thus, rotavirus “enterotoxin” sequences could be considerably divergent.     

Molecular characterization of a human group C rotavirus detected first in Turkey

The present study was done to find out the prevalence of group B and C rotavirus infections in children with diarrhea presented at two major hospitals in Ankara, Turkey. Group B rotavirus was not found in any samples. One of 122 samples was positive for group C rotavirus. Phylogenetic analysis of genes for nonstructural protein NSP4, and structural proteins VP4, VP6, and VP7 confirmed the human origin of this strain. Similar to other human group C rotaviruses, one N-glycosylation site was predicted at amino acid residue 67 on the VP7 of strain GUP188. The genes of strain GUP188 were closely related to those of human group C rotavirus strain from the UK (Bristol) for NSP4, China (208 and Wu82) for VP4 and VP6, and from Colombia (Javeriana) for VP7, indicating that the Turkish group C rotavirus was unique and can serve as an additional reference strain for the molecular epidemiology of group C rotaviruses.     

A Human Vaccine Strain of Lamb Rotavirus (Chinese) NSP4 Gene: Complete Nucleotide Sequence and Phylogenetic Analyses

A lamb strain of rotavirus has recently been licensed for use in China as a live vaccine to prevent rotavirus diarrhea in children. As rotavirus NSP4, especially the cytotoxic domain alone is considered to be associated with diarrhea, we sequenced gene segment 10, which encodes NSP4, of lamb rotavirus. Comparative analyses was performed to identify differences from human rotavirus strains, that might be associated with attenuation, and to ascertain whether the lamb rotavirus gene fits among the NSP4 of other sequenced rotavirus strains. Our comparative nucleotide sequence analysis suggests its close identity (91.17% homology) with that of group-A equine rotavirus (strain HI23). Multiple alignment of the deduced amino acid sequence of lamb NSP4 with that of other group A rotaviruses demonstrated homology ranging from 63.42% with that of porcine YM strain to 93.71% with equine HI23 strain of rotavirus. A group A-specific NSP4 monoclonal antibody recognized the glycosylated and unglycosylated forms of the protein from virus-infected lysates, suggesting a well-conserved group-specificity of the lamb NSP4. Phylogenetic analysis of the lamb rotavirus gene, with 60 other NSP4 gene sequences of human and animal rotavirus strains, demonstrated that the lamb rotavirus strain belongs to genotype A. Comparative analysis also revealed that although it is a vaccine strain, the NSP4 cytotoxic domain of lamb strain demonstrated an overall amino acid conservation similar to that of other strains, whose NSP4 alone causes diarrhea in animal models. These results taken together with our previous observations clearly reaffirm the idea that the attenuation phenotype of rotaviruses does not involve NSP4 cytotoxic domain, perhaps due to the suppression of NSP4 cytotoxic activity by other rotaviral proteins.     

Group C rotavirus NSP4 induces diarrhea in neonatal mice

Summary.  Nonstructural glycoprotein NSP4 of group A rotavirus induces diarrhea in neonatal mice by functioning as an enterotoxin. Previously, our laboratory reported that the structural features of group A and group C rotavirus NSP4 proteins are well conserved despite a lack of sequence homology between group A and group C rotavirus NSP4 proteins [Horie Y, et al., Arch Virol (1997) 142: 1865–1872]. To test whether group C rotavirus NSP4 has an enterotoxigenic activity, we expressed in Escherichia coli the carboxy two-thirds (corresponding to amino acid residues 55–150) of the NSP4 protein derived from group C rotavirus strain Ehime 9301. This truncated NSP4 protein was able to induce diarrhea in 5-day-old CD-1 mice when administered intraperitoneally. Thus, group C rotavirus NSP4 acts as an enterotoxin like group A rotavirus NSP4. Received September 22, 2000 Accepted November 18, 2000     

Whole genomic characterization of a human rotavirus strain B219 belonging to a novel group of the genus Rotavirus

Novel rotavirus strains B219 and ADRV-N derived from adult diarrheal cases in Bangladesh and China, respectively, are considered to belong to a novel rotavirus group (species) distinct from groups A, B, and C, by genetic analysis of five viral genes encoding VP6, VP7, NSP1, NSP2, and NSP3. In this study, the nucleotide sequences of the remaining six B219 gene segments encoding VP1, VP2, VP3, VP4, NSP4, and NSP5 were determined. The nucleotide sequences of the group B human rotavirus VP1 and VP3 genes were also determined in order to compare the whole genome of B219 with those of group A, B, and C rotavirus genomes. The nucleotide and deduced amino acid sequences of all B219 gene segments showed considerable identity to the ADRV-N (strain J19) sequences (87.7-94.3% and 88.7-98.7%, respectively). In contrast, sequence identity to groups A-C rotavirus genes was less than 61%. However, functionally important domains and structural characteristics in VP1-VP4, NSP4, and NSP5, which are conserved in group A, B, or C rotaviruses, were also found in the deduced amino acid sequences of the B219 proteins. Hence, the basic structures of all B219 viral proteins are considered to be similar to those of the known rotavirus groups.     

Full genomic analysis of human rotavirus strain TB-Chen isolated in China

A G2P[4]/NSP4[A] rotavirus strain TB-Chen was isolated from a 2-year-old patient hospitalized with acute gastroenteritis in Kunming, China. The strain TB-Chen was demonstrated having group A-specific antigenicity, a “short” (subgroup II) electropherotype. To investigate its overall genomic relatedness and to determine which group it belonged, the complete genome of strain TB-Chen was determined. Genomic comparison based on amino acid sequence identity and phylogenetic analysis revealed that all 11 gene segments of strain TB-Chen were highly identical (> 91.80%) with the representative G2P[4]/NSP4[A] human strains DS-1, S2, NR1 and IS2, suggesting that this rotavirus strain was derived from human host. Besides, almost all the available representative rotavirus gene segments among group A were analyzed and identified within 15 G-types, 28 P-types, and 6 NSP4 genotypes. This is the first report of group A rotavirus genomic analyses in China and the findings have important implications for rotavirus vaccine development.     

Molecular characterization of the 11th RNA segment from human group C rotavirus

The complete nucleotide sequence of genome segment 11 from the noncultivatable, human group C rotavirus (Bristol strain) was determined. Comparison of the nucleotide sequence of the segment termini with the consensus 5 and 3 terminal noncoding sequences of the human group C rotavirus genome revealed characteristic 5 and 3 sequences. Human group C rotavirus genome segment 11 is 613 bp long and encodes a single open reading frame of 450 nucleotides (150 amino acids) starting at nucleotide 39 and terminating at nucleotide 489, leaving a long 3 untranslated region of 124 nucleotides. The predicted translation product has a calculated molecular weight of 17.7 kD and contains four potential N-linked glycosylation sites. No significant homologies to other viral proteins were found in database searches. Hydropathy analysis predicted the human group C rotavirus genome segment 11 translation product has a hydrophilic carboxy terminus (amino acids 54–150) and a hydrophobic amino terminus (amino acids 1–53) that can be further subdivided into three short hydrophobic sequences—H1, H2, and H3. These features are analogous to the integral membrane glycoprotein NSP4 encoded by group A rotavirus gene 10.     

Genomic characterization of a novel group A lamb rotavirus isolated in Zaragoza, Spain

An ovine rotavirus (OVR) strain, 762, was isolated from a 30-day-old lamb affected with severe gastroenteritis, in Zaragoza, Spain, and the VP4, VP7, VP6, NSP4, and NSP5/NSP6 genes were subsequently characterized molecularly. Strain OVR762 was classified as a P[14] rotavirus, as the VP4 and VP8* trypsin-cleavage product of the VP4 protein revealed the highest amino acid (aa) identity (94% and 97%, respectively) with that of the P11[14] human rotavirus (HRV) strain PA169, isolated in Italy. Analysis of the VP7 gene product revealed that OVR762 possessed G8 serotype specificity, a type common in ruminants, with the highest degree of aa identity (95-98%) shared with serotype G8 HRV, bovine rotavirus, and guanaco (Lama guanicoe) rotavirus strains. Moreover, strain OVR762 displayed a bovine-like NSP4 (genotype E2) and NSP5/NSP6 (genotype H3), and a VP6 genotype I2, as well as a long electropherotype pattern. This is the first report of a lamb rotavirus with P[14] and G8 specificities, providing additional evidence for the wide genetic and antigenic diversity of group A rotaviruses.     

Characterization of the Structural-Protein-Coding Region of SAT 2 Type Foot-and-Mouth Disease Virus

The NSP4 protein of rotavirus is a nonstructural glycoprotein and has a crucial function in virus morphogenesis during infection of host cells. It was recently reported that NSP4 may also function as a viral enterotoxin in the induction of rotavirus diarrhea by causing Ca⁺⁺ influx in the cytoplasm of the infected cells. We sequenced and analyzed two (Wa and M strains) pairs of NSP4 genes of virulent (v) and attenuated (a) (after 30 to 40 passages in cell culture) human group A rotaviruses and a pair of NSP4 genes of virulent and attenuated porcine group C rotavirus (Cowden strain). These strains were previously identified as virulent (induce diarrhea) or attenuated (no diarrhea) in a gnotobiotic pig model of rotavirus infection [Bohl et al. (4), Saif et al. (13), Ward et al. (17)]. The NSP4 genes of the Wa, M and Cowden strains were amplified with RT-PCR using a proof reading polymerase (Tli) and the RT-PCR product was sequenced directly. Analysis of the NSP4 deduced amino acid sequences showed that only 3 (Wa) and 2 (M and Cowden) amino acids differed between the virulent and attenuated strains. For the Wa strain, the changes from the virulent to attenuated strain were in amino acids 13 (V to A), 16 (L to S) and 34 (P to L); in the M strain, the difference was in amino acids 53 (T to I) and 104 (K to E), and in the Cowden strains, amino acids 50 (L to F) and 97 (D to N) differed between virulent and attenuated strains. To our knowledge, this is the first sequence comparison between NSP4 of a virulent and attenuated pair of group C rotaviruses. The potential impact of these few amino acid changes on the pathogenesis of the NSP4 protein for piglets is unclear, relative to previous findings in mice (1), but requires further study using purified recombinant NSP4 proteins or peptides.     

Complete nucleotide sequence of a group A avian rotavirus genome and a comparison with its counterparts of mammalian rotaviruses

The nucleotide sequences encoding four structural proteins (VP1-4) and six nonstructural proteins (NSP1-6) of avian rotavirus PO-13 were determined. Based on the results of earlier sequencing studies [Ito et al., 1995, Sequence analysis of cDNA for the VP6 protein of group A avian rota viruses. Arch. Vriol. 140, 605-612; Rohwedder et al., 1997, Chicken rotavirus Ch-1 shows a second type of avian VP6 gene, Virus Genes 15, 65-71; Rohwedder et al., 1997, Bovine rotavirus 993/83 shows a third subtype of avian VP7 protein, Virus Genes 14, 147-151], determination of PO-13 genome sequence has been completed. The PO-13 genome is 18845 nucleotides in length. It is 290 nucleotides longer than the genome of SA11. The amino acid sequence homology between PO-13 and mammalian rotaviruses ranged from 76-77% (VP1) to 16-18% (NSP1). The features of gene and amino acid sequence were compared with those of the corresponding protein of mammalian rotaviruses. Based on results of the phylogenetic analyses of NSP1, we speculate that an ancestral rotavirus could have separated into groups A, B and C rotaviruses at an early evolutionary stage and that group A rotavirus separated into mammalian and avian rotaviruses with host evolution.     

Comparative sequence analysis identified mutations outside the NSP4 cytotoxic domain of tissue culture-adapted ATCC-Wa strain of human rotavirus and a novel inter-species variable domain in its C-terminus

Summary.  Complete nucleotide sequence of the tissue culture-adapted ATCC*-Wa strain of human rotavirus NSP4 was determined. Sequence analysis detected two alternate forms of the gene with a nucleotide difference at position 331 (A or G) in the coding sequence (NSP4-A or NSP4-G) leading to a change from neutral glutamine⁹⁷ in NSP4-A to a positively charged arginine⁹⁷ in NSP4-G originating from the same ATCC-Wa preparation. In addition to this, both forms of ATCC-Wa NSP4 revealed three mutations at nucleotide positions 88 (T to C), 142 (C to T) and 572 (G to A), when compared to the previously reported NSP4 sequence from virulent Wa strain. The former two mutations were in the coding sequence and resulted in a leucine¹⁶ to serine¹⁶ and a proline³⁴ to leucine³⁴ change, while the third mutation was in the 3′ non-coding region of the gene. The two amino acid changes may contribute to the ‘tissue culture-adaptation’ of ATCC-Wa strain. The ATCC-Wa NSP4 sequence was found to differ from the previously reported NSP4 sequence of attenuated Wa strain by lacking a mutation at 133 (T to C), though the mutations at 88 and 142 were present in both strains. Furthermore, comparison of deduced amino acid sequence of NSP4 from human, bovine, porcine and simian rotavirus strains identified a seven-residue (positions 135–141) inter-species variable domain in its C-terminal region. Received November 24, 1999/Accepted April 10, 2000     

Molecular characterization of VP4, VP6, VP7, NSP4, and NSP5/6 genes identifies an unusual G3P[10] human rotavirus strain

An unusual strain of human rotavirus G3P[10] (CMH079/05) was detected in a stool sample of a 2‐year‐old child admitted to the hospital with severe diarrhea in Chiang Mai, Thailand. Analysis of the VP7 gene sequence revealed highest identities with unusual human rotavirus G3 strain CMH222 at 98.7% on the nucleotide and 99.6% on the amino acid levels. Phylogenetic analysis of the VP7 sequence confirmed that the CMH079/05 strain formed a cluster with G3 rotavirus reference strains and showed the closest lineage with the CMH222 strain. Analysis of partial VP4 gene of CMH079/05 revealed highest degree of sequence identities with P[10] rotavirus prototype strain 69M at nucleotide and amino acid levels of 92.9% and 94.6%, respectively. Phylogenetic analysis of the VP4 sequence revealed that CMH079/05 and 69M clustered closely together in a monophyletic branch separated from other rotavirus genotypes. To our knowledge, this is a novel G–P combination of G3 and P[10] genotypes. In addition, analyses of VP6, NSP4, and NSP5/6 genes revealed these uncommon genetic characteristics: (i) the VP6 gene differed from the four other known subgroups; (ii) the NSP4 gene was identified as NSP4 genetic group C, an uncommon group in humans; and (iii) the NSP5/6 gene was most closely related with T152, a G12P[9] rotavirus previously isolated in Thailand. The finding of uncommon G3P[10] rotavirus in this pediatric patient provided additional evidence of the genetic diversity of human group A rotaviruses in Chiang Mai, Thailand. J. Med. Virol. 81:176–182, 2009. © 2008 Wiley‐Liss, Inc.     

Sequential analysis of nonstructural protein NSP4s derived from Group A avian rotaviruses

We determined the NSP4 sequences of turkey rotavirus strains Ty-1 and Ty-3 and a chicken rotavirus, strain Ch-1, and compared these sequences with those of a pigeon rotavirus, strain PO-13, and mammalian rotaviruses. The turkey strains and PO-13 were found to be closely related (90-97% homologies). Ch-1 NSP4 was distinctly different from other avian rotavirus NSP4s, with 78-79% homologies. The NSP4 sequences of avian rotaviruses were found to be 6-7 amino acids shorter than those of all mammalian strains and to have considerably low identities (31-37%) with them. Therefore, it seems highly likely that the NSP4 genes of avian rotaviruses are classified into two NSP4 genotypes distinct from those of mammalian rotaviruses. The enterotoxin domain in NSP4 is conserved in terms of its sequential and structural properties despite extremely low homologies in the full lengths of NSP4s in avian and mammalian rotaviruses.     

Characterization of human rotavirus serotype G9 isolated in Japan and Thailand from 1995 to 1997.

Serotyping of human rotavirus was conducted in 396 Japanese and 100 Thai rotavirus-positive fecal specimens collected from 1995 to 1997. Serotype G9 was found to be the third most common serotype with frequency of 16.2% in Thailand from 1996 to 1997. It was also detected in Japan with a low frequency (0.7%) in this year. The genetic analyses of VP4 and NSP4 genes of these G9 strains showed that 1 strain from Japan possessed P[8] genotype and NSP4 Wa-group with long electropherotype (e-type). In contrast, 5 strains from Thailand belonged to P[6] and 1 strain belonged to P[4]. All of the Thai strains were in the NSP4 KUN-group with a short e-type. Sequence analysis of their VP7 gene revealed that there was the highest homology among fecal G9 strains (> 96.3%, amino acid identity) and a relatively high degree of homology to standard viruses, F45 from Japan (95.4-96.3%, amino acid identity) and 116E from India (92-92.3%, amino acid identity). However, immunological analysis using G9 specific monoclonal antibodies (Mabs) against VP7 protein showed that the G9 strains isolated from the two countries had different antigenic specificity. It was confirmed further by intraserotypical phylogenetic analysis of VP7 amino acid. These results indicated that the prevalence of G9 rotavirus in 1996-1997 in Thailand was relative to the continuing recent emergence of it on a worldwide basis, while the Japanese G9 strain isolated in this survey was identified to have progenitors common to the F45 strain that was prevalent in 1985 in Japan. Phylogenetic analysis of VP7 amino acid of G1-14 prototype rotavirus showed that the G9 strains were most closely related to the equine G14 rotavirus FI23 strain but G3 strains, interserotypically. These findings suggest that G9 rotaviruses might be divided into two or more subtypes.     

河北省卢龙地区猪与人G9型A组轮状病毒进化关系研究

目的 了解河北省卢龙地区猪与人G9型A组轮状病毒（GARV）主要基因的分子特征及进化关系.方法 选取1株2008年河北卢龙地区G9型GARV阳性的腹泻仔猪粪便标本LLP48及4株2009年至2011年卢龙地区G9型GARV阳性的5岁以下住院腹泻患儿粪便标本进行主要基因片段的扩增,测序后对获得的基因序列通过MEGA、DNAStar等生物软件进行序列比对、同源性分析和系统进化分析.结果 卢龙猪GARV LLP48与4株卢龙人GARV编码的VP7、VP6、NSP2和NSP4基因具有高度同源性,核苷酸和氨基酸同源性分别是89.4％～94％和94.8％ ～98.2％;VP4片段的核苷酸（氨基酸）同源性较低,为71.4％～71.6％（68.2％～69.0％）.系统进化分析表明,LLP48编码的VP7、VP6、NSP2和NSP4基因与人来源的毒株关系密切,而VP4基因与猪来源的毒株关系密切.结论 卢龙猪GARV LLP48可能是猪的VP4与人的VP7、VP6、NSP2和NSP4自然重组.     

Sequence analysis of genes encoding structural and nonstructural proteins of a human group B rotavirus detected in Calcutta, India

Nucleotide sequences of RNA segments encoding structural proteins(VP4, VP6, and VP7) and nonstructural proteins(NSP1 and NSP3) of a human group B rotavirus CAL-1, which was detected in Calcutta, India, were determined and their relatedness with cognate genes of other group B rotaviruses was analyzed. The CAL-1 genes showed generally high sequence identities (more than 90%) to those of human group B rotavirus, adult diarrheal rotavirus (ADRV) in China, while identities with bovine, murine, and ovine viruses were considerably lower (58-73%). Among RNA segments analyzed, sequence identity of the VP6 gene was relatively high compared with other gene segments. In the CAL-1 VP7 sequence, many characteristics were shared by ADRV, but not by other animal group B rotaviruses. In contrast, VP4 and NSP3 of CAL-1 were single amino acid and 23 amino acids longer than those of ADRV strain, respectively, due to differences of a few nucleotides. These findings suggested that human group B rotaviruses CAL-1 and ADRV might have originated from a common ancestral virus distinct from animal group B rotaviruses reported so far, while some notable sequence differences indicated the distinct nature of these viruses.     

Molecular characterization of a rare G3P[3] human rotavirus reassortant strain reveals evidence for multiple human-animal interspecies transmissions

An unusual strain of human rotavirus G3P[3] (CMH222), bearing simian-like VP7 and caprine-like VP4 genes, was isolated from a 2-year-old child patient during the epidemiological survey of rotavirus in Chiang Mai, Thailand in 2000-2001. The rotavirus strain was characterized by molecular analysis of its VP4, VP6, VP7, and NSP4 gene segments. The VP4 sequence of CMH222 shared the greatest homology with those of caprine P[3] (GRV strain) at 90.6% nucleotide and 96.4% amino acid sequence identities. Interestingly, the VP7 sequence revealed highest identity with those of simian G3 rotavirus (RRV strain) at 88% nucleotide and 98.1% amino acid sequence identities. In contrast, percent sequence identities of both the VP4 and VP7 genes were lower when compared with those of human rotavirus G3P[3] reference strains (Ro1845 and HCR3). Analyses of VP6 and NSP4 sequences showed a close relationship with simian VP6 SG I and caprine NSP4 genotype C, respectively. Phylogenetic analysis of VP4, VP6, VP7, and NSP4 genes of CMH222 revealed a common evolutionary lineage with simian and caprine rotavirus strains. These findings strongly suggest multiple interspecies transmission events of rotavirus strains among caprine, simian, and human in nature and provide convincing evidence that evolution of human rotaviruses is tightly intermingled with the evolution of animal rotaviruses.     

Diarrhea induction by rotavirus NSP4 in the homologous mouse model system.

Comparison of the NSP4 amino acid sequences from 31 strains of mammalian rotaviruses revealed the presence of four distinct NSP4 alleles; i.e., the Wa, KUN, AU-1, and EW alleles. The EW allele consists only of NSP4s from murine rotavirus strains and is divergent from other NSP4 alleles from the evolutionary perspective. There have been conflicting reports regarding the enterotoxigenic activity of NSP4 in the mouse model system; heterologous simian and porcine rotavirus NSP4s function as an enterotoxin in mice, while a homologous EC NSP4 does not play a dominant role as an enterotoxin in the cystic fibrosis conductance regulator knockout mice. To further examine the enterotoxigenic activity of NSP4, we expressed in Escherichia coli a recombinant protein consisting of glutathione S-transferase and amino acid residues 86-175 of the EW NSP4. We found that this fusion protein caused diarrhea in the majority (8/14) of 5- to 6-day-old CD1 mice. This study confirmed and extended that group A rotavirus NSP4s were able to induce diarrhea in neonatal mice and had an enterotoxigenic activity.     

Species-specific but not genotype-specific primary and secondary isotype-specific NSP4 antibody responses in gnotobiotic calves and piglets infected with homologous host bovine (NSP4[A]) or porcine (NSP4[B]) rotavirus

Using recombinant baculoviruses expressing rotavirus NSP4 [A], [B], [C], and [D] genotypes of bovine, porcine, human, simian, or murine origin, we analyzed serum antibody responses to NSP4s in gnotobiotic calves and piglets infected by the oral/alimentary or intraamniotic route with bovine (NSP4[A]) (Wyatt, R.G., Mebus, C.A., Yolken, R.H., Kalica, A.R., James, H.D., Jr., Kapikian, A.Z., Chanock, R.M., 1979. Rotaviral immunity in gnotobiotic calves: heterologous resistance to human virus induced by bovine virus. Science 203(4380), 548-550) or porcine (NSP4[B]) (Hoshino, Y., Saif, L.J., Sereno, M.M., Chanock, R.M., Kapikian, A.Z., 1988. Infection immunity of piglets to either VP3 or VP7 outer capsid protein confers resistance to challenge with a virulent rotavirus bearing the corresponding antigen. J. Virol. 62(3), 744-748) rotaviruses. Following primary infection and challenge with virulent rotaviruses, the animals developed higher or significantly higher antibody titers to homologous host homotypic NSP4s than to heterologous host homotypic or heterologous host heterotypic NSP4s, indicating that antibody responses were species specific rather than genotype specific. Antibody responses to NSP4s corresponded closely with the phylogenetic relationships of NSP4s within a species-specific region of amino acids (aa) 131-141. In contrast, NSP4 genotypes determined by amino acid full-length sequence identity predicted poorly their "serotypes". In piglets, antibodies to NSP4 induced by previous oral infection failed to confer protection against challenge from a porcine rotavirus bearing serotypically different VP4 and VP7 but essentially identical NSP4 to the porcine rotavirus in primary infection. Thus, in an approach to immunization with a live oral rotavirus vaccine, the NSP4 protein does not appear to play an important role in protection against rotavirus disease and infection.