Search

Noroviruses are major pathogens associated with acute gastroenteritis. They are diverse viruses, with at least six genogroups (GI-GVI) and multiple genotypes defined by differences in the major capsid protein, VP1. This diversity has challenged the development of broadly cross-reactive vaccines as well as efficient detection methods. Here, we report the characterization of a broadly cross-reactive monoclonal antibody (MAb) raised against the capsid protein of a GII.3 norovirus strain. The MAb reacted with VLPs and denatured VP1 protein from GI, GII, GIV and GV noroviruses, and mapped to a linear epitope located in the inner shell domain. An alignment of all available VP1 sequences showed that the putative epitope (residues 52-56) is highly conserved across the genus Norovirus. This broadly cross-reactive MAb thus constitutes a valuable reagent for the diagnosis and study of these diverse viruses.

The application of molecular technologies, such as the expression of viral proteins in baculovirus, has provided a powerful approach to the diagnosis of human calicivirus (HuCV) infections. The baculovirus-expressed HuCV capsid protein self-assembles into virus-like particles, providing excellent reagents for immunologic assays, such as enzyme immunoassays (EIAs). Following the expression of the capsid protein of Norwalk virus, the capsid proteins of 8 other HuCV strains have been expressed in baculovirus. The unlimited supply of baculovirus-produced reagents for HuCVs allows these EIAs to be applied in large-scale clinical and epidemiological studies. Both the antigen and antibody-detection EIAs are highly sensitive. The antigen-detection EIAs are highly specific, but the antibody-detection EIAs are more broadly reactive. This article reviews baculovirus expression techniques used to produce HuCV capsid antigens, development of EIAs using these antigens, and application of these EIAs in studies of HuCV infection and illness.

Feline calicivirus (FCV) strains can show significant antigenic variation when tested for cross-reactivity with antisera produced against other FCV strains. Previous work has demonstrated the presence of hypervariable amino acid sequences in the capsid protein of FCV (designated regions C and E) that were postulated to constitute the major antigenic determinants of the virus. To examine the involvement of hypervariable sequences in determining the antigenic phenotype, the nucleotide sequences encoding the E regions from three antigenically distinct parental FCV strains (CFI, KCD, and NADC) were exchanged for the equivalent sequences in an FCV Urbana strain infectious cDNA clone. Two of the three constructs were recovered as viable, chimeric viruses. In six additional constructs, of which three were recovered as viable virus, the E region from the parental viruses was divided into left (N-terminal) and right (C-terminal) halves and engineered into the infectious clone. A final viable construct contained the C, D, and E regions of the NADC parental strain. Recovered chimeric viruses showed considerable antigenic variation from the parental viruses when tested against parental hyperimmune serum. No domain exchange was able to confer complete recognition by parental antiserum with the exception of the KCD E region exchange, which was neutralized at a near-homologous titer with KCD antiserum. These data demonstrate that it is possible to recover engineered chimeric FCV strains that possess altered antigenic characteristics. Furthermore, the E hypervariable region of the capsid protein appears to play a major role in the formation of the antigenic structure of the virion where conformational epitopes may be more important than linear in viral neutralization.

The capsid protein of the Hawaii strain of human calicivirus was expressed in the transient MVA/bacteriophage T7 polymerase hybrid expression system in order to examine its processing in mammalian cells. Selected amino acid modifications (an insertion, deletion, and substitution) at the predicted amino terminus of the capsid protein as well as the presence or absence of the ORF3 gene were examined for their effect on capsid expression. The protein was expressed efficiently in cell lines derived from three different species, with most of the expressed protein remaining localized within the cells. There was no evidence for N-linked glycosylation or myristylation of the 57 kDa capsid protein. Hawaii virus-like particles (HV VLPs), efficiently produced in the baculovirus expression system, were not observed in this expression system under the conditions in this study.

Feline calicivirus (FCV), a member of the Caliciviridae, produces its major structural protein as a precursor polyprotein from a subgenomic-sized mRNA. In this study, we show that the proteinase responsible for processing this precursor into the mature capsid protein is encoded by the viral genome at the 3'-terminal portion of open reading frame 1 (ORF1). Protein expression studies of either the entire or partial ORF1 indicate that the proteinase is active when expressed either in in vitro translation or in bacterial cells. Site-directed mutagenesis was used to characterize the proteinase Glu-Ala cleavage site in the capsid precursor, utilizing an in vitro cleavage assay in which mutant precursor proteins translated from cDNA clones were used as substrates for trans cleavage by the proteinase. In general, amino acid substitutions in the P1 position (Glu) of the cleavage site were less well tolerated by the proteinase than those in the P1' position (Ala). The precursor cleavage site mutations were introduced into an infectious cDNA clone of the FCV genome, and transfection of RNA derived from these clones into feline kidney cells showed that efficient cleavage of the capsid precursor by the virus-encoded proteinase is a critical determinant in the growth of the virus.

The Norwalk and Hawaii viruses are antigenically distinct members of the family Caliciviridae and are considered to be important etiologic agents of epidemic gastroenteritis, with most studies focusing on the role of Norwalk virus. To further investigate the importance of Hawaii virus, Hawaii virus-like particles (VLPs) were produced by expression of its capsid protein in the baculovirus system and these VLPs were used as the antigen in an enzyme-linked immunosorbent assay that was efficient in the detection of a serologic response to Hawaii virus. The ready availability of Hawaii VLPs should enable larger-scale epidemiological studies to further elucidate the importance of this agent.

Snow Mountain virus (SMV) is the reference strain for serotype 3 as determined by immune electron microscopy of the human caliciviruses that are associated with epidemic gastroenteritis. In order to establish the genetic relationship of its capsid protein with those from other human caliciviriuses, the sequence of the open reading frame 2 (ORF2) encoding the SMV capsid protein was determined. The SMV ORF2 sequence was 1626 nucleotides in length and the deduced protein of 542 amino acids had a calculated molecular weight of 59.2 kD. The SMV capsid sequence showed approximately 48 and 77% amino acid sequence identity with the capsid proteins of the Norwalk (serotype 1) and Hawaii (serotype 2) human calicivirus reference strains, respectively, a finding consistent with its serotypic distinctiveness. Furthermore, the predicted amino acid sequence of the SMV capsid was found to share highest sequence identity (98%) with the Melksham human calicivirus in database searches.

Toronto virus (TV), previously called "minireovirus", a human calicivirus classified as genogroup 2 and phylogenetic type P2-A, was originally described in association with diarrhea in children. The second open reading frame, encoding the capsid protein of TV24, was expressed in a baculovirus recombinant. The recombinant baculovirus produced a protein (rTV) with an apparent molecular mass of 58 kDa that self-assembled into virus-like particles approximately 30 nm in diameter with a density of 1.29 g/ml. Antigenic and immunogenic characteristics of these particles were determined by protein immunoblot, immunoprecipitation, and enzyme immunoassay. Seroconversion to the rTV protein was detected in 6 of 8 (75%) patients from a recent outbreak of gastroenteritis associated with a virus of similar phylogenetic type. These results confirm and extend the previous reports of the expression of the Norwalk and Mexico virus capsid proteins.

Norwalk virus (NV) infection was recently found to be associated with gastroenteritis in U.S. military troops stationed in Saudi Arabia during the 1990 Desert Shield Operation. We identified a Norwalk-like virus in the stools of two military personnel with gastroenteritis by ELISA and IEM. By RT-PCR and sequence analysis, the nucleotide sequence of part of the polymerase region of each of these two "Desert Shield" strains (DSV275 and DSV395) was found to be 73% identical to the corresponding region of NV. In addition, one of the strains (DSV395), which underwent sequence analysis of approximately 2900 consecutive bases, had a genomic organization characteristic of the Caliciviridae. Comparison of the DSV395 amino acid sequence of the capsid region with that of three other viruses in the Norwalk group (Norwalk, Southampton, and Toronto viruses) showed amino acid identity of 47-68%. Consensus sequence analysis of these capsid proteins identified two regions of conserved amino acids that flanked an area of variable amino acids. In addition, the proteins corresponding to the capsid regions of DSV395 and NV were expressed in an in vitro translation system. Immunoprecipitation studies using the expressed capsid proteins and paired DSV395 or NV infection sera indicated the presence of shared antigenic sites between the capsid proteins of DSV395 and NV. However, hyperimmune sera specific for the self-assembled recombinant NV capsid protein did not react with DSV stool antigen in an ELISA, suggesting that there may also be unique antigenic sites not shared between DSV395 and NV.

Since the discovery of the Norwalk virus (NV) by immune electron microscopy (IEM) in 1972, serologic studies with this virus have relied on particle-positive fecal material from infected volunteers as the source of antigen because it has not been possible to propagate this virus in cell culture. However, the recent cloning of the NV (strain 8FIIa) genome and expression of the capsid protein in a baculovirus system to form "virus-like particles" has provided a consistent source of antigen (designated rNV). The purpose of the present study was to compare the antigenicities of these rNV particles with those of native NV antigen derived from human fecal material by using well-characterized sera obtained from earlier studies. In IEM studies, the rNV antigen reacted with NV-specific antibodies in a manner similar to that observed previously when particle-positive fecal material was used as antigen. In addition, a direct enzyme-linked immunosorbent assay, in which the rNV antigen was used as antigen, proved efficient and specific for the detection of serologic responses to NV compared with the previously established techniques of IEM and blocking antibody immunoassays in which particle-positive fecal material was used as the antigen. The availability of an unlimited source of antigen will enable serologic studies that will greatly increase our understanding of the epidemiology of NV and its role in human enteric illness.

To further study the molecular characteristics of group (gp) C rotaviruses, we produced, cloned, and sequenced cDNA to gene 5 of the Shintoku strain of bovine gp C rotavirus. The resulting clone was specific for gene 5 and was genetically related to the human and porcine gp C rotaviruses, as demonstrated by Northern blot hybridization analysis. The Shintoku gene 5 is 1352 nucleotides in length and has one open reading frame encoding a polypeptide of 395 amino acids with a predicted molecular mass of 44.5 kDa. Comparative sequence analysis indicated that: (i) the Shintoku gene 5 protein shared 88.4 to 90.6% homology with the VP6 of the human (Bristol and 88-220) and porcine (Cowden) strains of gp C rotaviruses, but only low homology with the VP6 of bovine gp A (RF) and human gp B (ADRV) rotaviruses (41.3 and 16.3%, respectively); (ii) the predicted secondary structure was highly conserved among the gene 5 proteins of the bovine, porcine, and human gp C rotaviruses; and (iii) seven highly conserved regions were identified for the first time in the deduced primary amino acid sequences of gene 5 of gp C and gene 6 of gp A rotaviruses. However, only three of these highly conserved areas were present in the regions of VP6, where the secondary structure was predicted to be similar for the rotavirus strains examined. These three regions may contribute to common epitopes between the two groups of rotaviruses. Our results, in comparison with data for other rotaviruses, indicate that gene 5 of the bovine gp C rotavirus codes for the major inner capsid protein (VP6).

Sera from 17 of 18 adult volunteers challenged with a virulent serotype 1 rotavirus strain (D) were examined for prechallenge antibody levels against several well-defined rotavirus VP7 and VP4 neutralization epitopes by a competitive epitope-blocking immunoassay (EBA) in order to determine whether correlates of resistance to diarrheal illness could be identified. The presence of prechallenge serum antibody at a titer of greater than or equal to 1:20 that blocked the binding of a serotype 1 VP7-specific monoclonal antibody (designated 2C9) that maps to amino acid residue 94 in antigenic site A on the serotype 1 VP7 was significantly associated with resistance to illness or shedding (P less than 0.001) or illness and shedding (P less than 0.01) following challenge with the serotype 1 virus. In addition, an EBA antibody titer of greater than or equal to 1:20 in prechallenge serum against a serotype 3 VP7-specific epitope (defined by monoclonal antibody 954/159) that maps to amino acid 94 on the serotype 3 VP7 was also significantly associated with resistance to illness or shedding (P = 0.02), with a trend for protection against illness and shedding. A trend was also noted between the presence of EBA antibody against a cross-reactive VP4 epitope common to many human rotavirus strains, including the challenge virus, or a rhesus monkey rotavirus strain-specific VP4 antigenic site, and resistance to illness or shedding. These data confirm that the presence of serum antibody correlates with resistance to rotavirus illness or shedding but, in addition, demonstrate the association of antibody to a specific epitope with resistance to illness or shedding. These data also suggest that antigenic site A on the rotavirus VP7, composed of amino acids 87 to 96, may be involved in the formation of a major protective epitope. Further study of the role of this epitope in the development of homotypic and heterotypic immunity to rotaviruses following natural or vaccine-induced infection may be important in the development of strategies for control of rotavirus diarrheal disease.

Two partial cDNA clones of genes 5 (encoding the major inner capsid protein VP 6) and 6 (encoding a nonstructural protein) of the porcine group (Gp) C rotavirus (Cowden strain) were radiolabeled with 32P and used individually as probes in Northern and dot blot hybridization assays. The specificity of each probe was tested against genomic dsRNA from: (1) porcine Gp A, B, and C rotaviruses; (2) Gp C rotaviruses from different species; and (3) porcine Gp C rotavirus field strains with varying electropherotype patterns. Neither probe hybridized with ds RNA from the porcine Gp A and B strains under the stringency conditions employed in the study. However, the gene 5 probe hybridized with the corresponding gene from the homologous porcine and the heterologous human and bovine Gp C rotaviruses tested. The gene 6 probe hybridized with the corresponding gene from the homologous Cowden strain, but hybridized weakly with gene 6 from the human and bovine Gp C rotaviruses. Both probes recognized all six different porcine Gp C field strains, although with varying intensities. Our results demonstrate that the gene 5 and 6 probes used in this study are specific for Gp C rotaviruses. However, evidence for greater genetic variation in the gene 6 among porcine, bovine and human Gp C strains suggested that the gene 5 probe may prove more broadly reactive among Gp C strains from different species. cDNA probes used in our study should prove useful for the detection of Gp C rotaviruses in feces and facilitate epidemiologic studies.

The complete nucleotide sequence of the gene 6 from the porcine group (Gp) C rotavirus strain Cowden was determined from gene 6-specific clones selected from a cDNA library and from viral transcript RNA. The gene is 1348 nucleotides in length with a potential initiation codon beginning at nucleotide 25 and a stop codon at nucleotide 1231. The deduced protein contains 402 amino acids. Comparison of the gene 6 from this Gp C strain with sequences in the GenBank data base indicated that this gene shared homology with gene 7 of Gp A rotavirus strain SA-11 (22.9%) and gene 9 of Gp A rotavirus strain UK (22.6%), both of which encode the NS34 protein. In vitro translation products produced from transcripts generated from a gene 6 clone containing the entire open reading frame were not immunoprecipitated with either hyperimmune serum specific for the Gp C Cowden strain or a monoclonal antibody directed against the group antigen (VP6) of the Cowden strain. However, products generated from a full-length gene 5 clone of the Cowden strain were immunoprecipitated by each of these antibodies. These data suggest that in contrast to the Gp A viruses in which the gene 6 encodes the major inner capsid protein VP6, the gene 6 of the Cowden Gp C strain encodes a nonstructural protein corresponding to the NS34 of Gp A rotaviruses.

The genetic diversity of gene 8 (encoding the outer capsid glycoprotein VP7) among group (Gp) C rotaviruses was examined by Northern and dot blot hybridization. A cDNA clone of the porcine Gp C Cowden strain gene 8 was labeled with 32P by nick translation and used as a probe. The gene 8 probe hybridized with the corresponding gene of one human (88-196) and four porcine (Cowden, NB, WH, and Wi) strains of Gp C rotaviruses under both moderate (50% formamide, 5X SSC, and 42 degrees) and high (50% formamide, 5X SSC, and 52 degrees) stringency conditions. However, under high stringency conditions little or no hybridization was detected with the corresponding gene of one bovine (Shintoku) and three other porcine (Ah, HF, and KH) strains of Gp C rotaviruses. In control experiments, the Cowden gene 8 probe did not hybridize with Gp A (Gottfried strain) or Gp B (Ohio strain) rotaviruses. These data demonstrate that the Cowden gene 8 probe is Gp C rotavirus-specific and that genetic diversity exists among Gp C rotaviruses in the gene encoding the outer capsid glycoprotein VP7. Our gene 8 probe may be useful in hybridization assays for serotyping Gp C rotaviruses, analogous to the use of VP7 probes for serotyping Gp A rotaviruses. However, final confirmation of our genetic approach to serotype Gp C rotaviruses awaits the serologic analysis of these viruses.

The group A rotavirus strain 69M has recently been classified as a new rotavirus serotype, G8, based on the antigenic specificity of its VP7 outer capsid glycoprotein. The fourth gene of 69M, which encodes the other outer capsid protein VP4, was sequenced. The gene is 2362 nucleotides in length and contains one long open reading frame beginning at the 10th nucleotide from the 5' end and terminating with a stop codon 22 bases from the 3' end. The encoded VP4 contains 776 amino acids. Comparative analysis of the deduced amino acid sequence with other rotavirus strains demonstrated that the VP4 of strain 69M shares 68.9 to 74.5% amino acid identity with the VP4 from strains known to represent four human rotavirus VP4 genetic alleles and 74.9 to 86.2% identity with VP4 from various animal rotaviruses. Northern blot hybridization with either radiolabeled ssRNA transcripts from 69M or cloned 69M gene 4 cDNA as probes indicated that the fourth gene of strain 69M is genetically distinct from any group A rotavirus gene 4 described thus far. These data suggest that at least one additional antigenic specificity of VP4 exists in circulating human group A rotavirus strains.

Sequence analysis of the gene encoding the major neutralization glycoprotein (VP7) was performed on 27 human and animal rotavirus strains of serotype 3 in order to examine genetic variation within strains of identical serotype. Comparisons of the deduced amino acid sequences of the VP7s showed overall sequence identities of 85% or higher. A higher degree of overall VP7 sequence similarity was observed among strains from the same animal species when compared to strains from different animal species, suggesting that there are species-specific sequences in the VP7 protein. Alignment of the amino acid sequences demonstrated that amino acid sequence divergence among serotype 3 strains from different species was located primarily in previously established VP7 serotype-specific regions where genetic variation was identified among strains of different serotype. These regions were highly conserved among serotype 3 strains derived from the same species. The varying reactivities of three anti-VP7 monoclonal antibodies with the 27 strains was consistent with the occurrence of antigenic variation among serotype 3 strains. Moreover the reactivity of monoclonal antibodies correlated with the amino acid sequence found in two serotype-specific regions (VR5 and VR8). A computer-derived predicted phylogenetic tree suggests that rotavirus strains from different animal species belonging to serotype 3 are more closely related to each other than to rotavirus strains of different serotypes.

Hybridomas producing monoclonal antibodies to VP7, a major neutralizing protein of serotype 9 rotavirus (strain W161), were prepared. One monoclonal antibody, W161-6A1, was shown to neutralize only serotype 9 rotavirus strains and reacted specifically with serotype 9 rotaviruses in an enzyme-linked immunosorbent assay. The development of an immunoassay for detection of serotype 9 rotaviruses should facilitate epidemiologic studies.

The group A rotaviruses are composed of at least seven serotypes. Serotype specificity is defined mainly by an outer capsid protein, VP7. In contrast, the other surface protein, VP3 (775 amino acids), appears to be associated with both serotype-specific and heterotypic immunity. To identify the cross-reactive and serotype-specific neutralization epitopes on VP3 of human rotavirus, we sequenced the VP3 gene of antigenic mutants resistant to each of seven anti-VP3 neutralizing monoclonal antibodies (N-MAbs) which exhibited heterotypic or serotype 2-specific reactivity, and we defined three distinct neutralization epitopes on VP3. The mutants sustained single amino acid substitutions at position 305, 392, 433, or 439. Amino acid position 305 was critical to epitope I, whereas amino acid position 433 was critical to epitope III. In contrast, epitope II appeared to be more dependent upon conformation and protein folding because both amino acid positions 392 and 439 appeared to be critical. These four positions clustered in a relatively limited area of VP5, the larger of the two cleavage products of VP3. At the positions where amino acid substitutions occurred, there was a correlation between amino acid sequence homology among different serotypes and the reactivity patterns of various viruses with the N-MAbs used for selection of mutants. A synthetic peptide (amino acids 296 to 313) which included the sequence of epitope I reacted with its corresponding N-MAb, suggesting that the region contains a sequential antigenic determinant. These data may prove useful in current efforts to develop vaccines against human rotavirus infection.

The sequence of the VP7 gene from 19 rotavirus strains recovered from asymptomatically infected newborn infants was determined by direct analysis of transcript RNAs synthesized from virus present in the stool. For five viruses the entire VP7 gene was sequenced, whereas in the remaining instances only a portion of the gene could be sequenced. In 19 specimens collected over a 4-year period, only five nucleotide substitutions were detected. None of them resulted in an amino acid substitution. Examination of a 306-nucleotide segment of gene 4 in 11 specimens yielded similar results. These results suggest that the mutation rate of rotaviruses in nature is lower than that of single-stranded RNA viruses such as poliovirus and influenza virus.

Human rotavirus strains 69M and WI61 are distinct from human rotavirus serotypes 1, 2, 3, and 4 and from each other by plaque reduction neutralization and have been proposed as new human rotavirus serotypes (serotype 8 and serotype 9, respectively). The nucleotide sequence of the gene encoding the serotype-specific capsid glycoprotein, VP7, of strains 69M and WI61 was determined. In addition, the sequence of the VP7 gene of strain F45 (serotypically indistinguishable from WI61) was determined. Comparative analyses of the nucleotide and deduced amino acid sequences with those of reference strains from serotypes 1,2,3,4,5, and 6 demonstrated that WI61 and F45 share a high degree of sequence similarity with each other and that strains 69M, WI61, and F45 are distinct from established serotypes 1,2,3,4,5, and 6 in nine defined regions of the VP7 which are variable across rotavirus serotypes.