Search

We have previously reported on the use of a tobacco mosaic virus (TMV) vector TMV-30B to express foreign viral antigens for use as experimental immunogens. Here we describe the development of an improved TMV-30B vector that adds a sequence of 7 histidine residues to the C-terminus of recombinant proteins expressed in the vector. We used this TMV-30B-HISc vector to express the VP8* fragment of the VP4 protein from bovine rotavirus (BRV) strain C-486 in plants. Recombinant VP8* protein was purified from N. benthamiana leaves at 7 days post-inoculation by immobilized metal affinity chromatography. The plant-produced VP8* was initially detected using anti-His tag mAb and its antigenic nature was confirmed using both monoclonal and polyclonal specific antisera directed against BRV. Adult female mice, inoculated by the intraperinoteal route with an immunogen containing 4 microg of recombinant VP8*, developed a specific and sustained response to the native VP8* from the homologous BRV. Eighty five percent of suckling mice from immunized dams that were challenged with the homologous virus at the fifth day of age were protected from virus as compared to 35% of the pups from mothers immunized with a control protein. These results demonstrate that the plant-produced VP8* was able to induce passive protection in the new born through the immunization of dams. This suggests that the technology presented here provides a simple method for using plants as an inexpensive alternative source for production of recombinant anti-rotavirus antigens.

Availability of effective oral vaccine delivery vehicles should contribute to the success of oral immunization in domestic animals. To achieve this goal, we evaluated alginate microspheres for their capacity to induce mucosal immune responses following oral and enteric immunizations. Mice were immunized with either live porcine rotavirus (PRV) or its recombinant VP6 protein, encapsulated in alginate microspheres or unencapsulated. VP6-specific IgG (but no IgA) antibodies were detected in the sera of mice after a single intraperitoneal (i.p.) immunization with either VP6 in Incomplete Freund's adjuvant (VP6-IFA), VP6 in alginate microspheres (VP6-MS) or with live PRV in incomplete Freund's adjuvant (PRV-IFA). In contrast, VP6-specific IgA (but no IgG) was detected in culture supernatants of mesenteric lymph nodes from mice immunized i.p. with either VP6-IFA or with PRV-IFA. Oral immunization with VP6-MS induced the highest level of VP6-specific fecal IgA antibody, similar to responses induced by oral immunization with live PRV. Furthermore, the VP6-specific fecal IgA could be boosted by a secondary i.p. immunization with VP6. Further experiments were performed in a sheep intestinal 'loop' model to evaluate uptake of microspheres by Peyer's patches. Microspheres containing colloidal carbon were specifically bound and transported by follicle-associated epithelium of Peyer's patches. Additionally, mucosal immune responses were detected following enteric immunization with porcine serum albumin (PSA) encapsulated in alginate microspheres. Our results confirm that alginate microspheres are an effective oral delivery vehicle for protein antigens and intestinal IgA antibody responses are induced by antigens encapsulated in alginate microspheres without any additional mucosal adjuvant. These investigations confirm that alginate microspheres have the potential as an effective delivery vehicle for oral immunization of ruminants.

The binding of the complement C3d molecule with receptors on B cells and/or follicular dendritic cells (FDCs) influences the induction of humoral immune responses. For example, C3d fused to an antigen has been shown to have a strong adjuvant effect on antibody production. We investigated the possibility that co-expression of antigen and C3d as a fusion protein could enhance antigen-specific immune responses, following plasmid immunization. One or two copies of murine C3d-cDNA, C3d or (C3d)(2), respectively, were cloned together with bovine rotavirus (BRV) VP7 or bovine herpesvirus type 1 (BHV-1) glycoprotein D (gD) genes. All constructs contained a signal peptide that resulted in the secretion of the expressed proteins. In vitro, the characterization of the chimeric proteins indicated that both VP7 and gD retained their antigenicity and the C3d remained biologically active. However, immunization with plasmids encoding VP7-C3d chimeras did not enhance rotavirus-specific antibody responses and the frequency of BRV-specific IFN-gamma secreting cells in the spleens were significantly lower in mice immunized with pVP7-(C3d)(2) when compared with mice immunized with plasmid encoding VP7. The same pattern of immune responses was observed for plasmids encoding gD-C3d. Both gD-specific antibody responses and the frequency of gD-specific IFN-gamma secreting cells were significantly lower in mice immunized with plasmid expressing gD-C3d chimeras when compared with mice immunized with plasmid encoding gD alone. These results indicate that co-expression of C3d with an antigen actually inhibit both humoral and cell-mediated antigen-specific immune responses.

One approach to understanding the physiologically relevant events during the induction of an immune response is to identify genes that are expressed when the immune system first encounters antigen. Such an investigation requires a naive but fully functional immune system, and the fetal lamb provides these conditions during the last trimester of gestation. 'Intestinal segments,' containing a jejunal Peyer's patch, were surgically prepared in fetal lambs (>120 days gestation) and individual 'intestinal segments' were injected with either culture medium or infectious bovine rotavirus. Peyer's patch tissue was collected 18 h postinfection. Histology and virus culture confirmed that bovine rotavirus had infected the mucosal epithelium. RNA was extracted from jejunal Peyer's patch tissue and mRNA differential display was used to identify genes expressed following rotavirus infection. Ten cDNAs were identified by differential display and these cDNAs were isolated, cloned, and sequenced. One of the cDNAs sequenced, displayed homology to the gene encoding the sperm surface protein Sp17. Differential expression of this gene in antigen-exposed jejunal Peyer's patches was confirmed by Northern blot and RT-PCR. The complete sequence for sheep Sp17 mRNA was obtained from a lambda cDNA library, prepared from the jejunal Peyer's patch of a young lamb. Sp17 expression was detected by RT-PCR in a variety of mucosa-associated lymphoid tissues but not in primary or other secondary lymphoid tissues. Thus, the fetal lamb model may be appropriate for identifying genes relevant to mucosal immunity.

A synthetic peptide corresponding to bovine rotavirus C486 (BRV) VP4 amino acid sequence 232-255 (VP4-peptide) was studied with the objective of defining the origin of the protective immune response reported previously by Ijaz et al. (J. Virol. 1991, 65, 3106-3113). Pretreatment of MA-104 cells with the VP4-peptide before infection with rotavirus prevented both the attachment of 35S-labelled virus and plaque formation in vitro. In vivo studies using a murine rotavirus model demonstrated that intragastric administration of VP4-peptide protected subjects from challenge with virulent rotavirus. These results clearly indicate the importance of this epitope in virus-cell interactions and their potential as a rotavirus vaccine candidate.

VP8*, the N-terminal cleavage product of rotavirus VP4, contains the virus neutralizing epitopes in the hemagglutination domain. To map the neutralizing epitope, we developed monoclonal antibodies specific for VP4 of bovine rotavirus C486 (BRV). A neutralizing escape mutant was generated by one of these monoclonal antibodies (2E8) and a point mutation (Glu-->Asp) was found at aa 116 of VP8*. To investigate the effect of this mutation on the cellular binding and hemagglutination activities, the VP8* genes of the escape mutant and wild type (WT) virus were expressed in E. coli and their functional activities were compared. Both the escape mutant and WT virus VP8* showed hemagglutination and MA-104 cell binding activities. However, hemagglutination activity of the WT virus VP8* was inhibited by 2E8, but that of the escape mutant VP8* was not. These data indicate that the neutralizing epitope is located in the HA domain but is not critical for rotavirus attachment to MA-104 cells. To understand virus neutralization, radiolabelled BRV was incubated with 2E8 and the distribution of radioactivity in a CsCI density gradient was analysed as was the morphology of the virions in peak fractions. Interaction of 2E8 with rotavirus led to virus morphological changes with a concomitant shift in buoyant density. These data suggest that aa 116 influences the binding of 2E8 which in turn may alter virus integrity.

Rotavirus VP8*, the N-terminal trypsin cleavage product of VP4, has been shown to bind to MA-104 cells and human O type erythrocytes. To examine whether bacterially expressed VP8* binds to cellular components of MA-104 cells, the VP8* (aa 1-247) was expressed in E. coli and radiolabelled with 35S-methionine. The radiolabelled rVP8* was immunoprecipitated with antiserum to bovine rotavirus C486 (BRV). The rVP8* was found to bind to MA-104 cells and its binding was competed by BRV. To study the interaction between VP8* and receptors of erythrocytes, hemagglutination (HA) and hemagglutination inhibition (HI) assays were carried out using solubilized rVP8*. rVP8* showed HA which could be inhibited by antiserum to BRV. This interaction was also inhibited by gangliosides, demonstrating a sialic acid dependent interaction. To study the contribution of the C-terminal region of VP8* to HA, a number of approaches were used. First, a peptide spanning aa 230-247 was synthesized and antisera was raised against the peptide to see whether it could inhibit HA of rVP8*. Second, a truncated form of VP8* (tVP8*: aa 1-229) was expressed to examine its hemagglutinating activity. Third, the dimerization of rVP8* and tVP8* was compared by Western-blotting following electrophoresis using native SDS-PAGE. The results indicated that antibody to aa 230-247 inhibits hemagglutination by preventing dimerization of VP8* which in turn allows the molecule to cause HA. To characterize the interaction between the HA domain and sialic acid receptors, erythrocytes were treated with sialidases of different specificities. Arthrobacter ureafaciens, Clostridium perfringens and alpha 2-8 linkage-specific neuraminidase destroyed the ability of sialic acid of erythrocytes to interact with rVP8*, indicating that bovine rotavirus C486 binding requires an alpha 2-8 linkage but acetylation of the sialic acid is not necessary.

Neonatal calf diarrhoea caused by bovine rotavirus is one of the most common diseases in cattle. VP8 portion of the rotavirus VP4 protein is known to contain neutralizing epitopes and hemagglutination activity. We expressed the VP8* portion of bovine rotavirus strain C486 (G6P1 serotype) in E. coli, and examined potential of the recombinant VP8* protein for induction of neutralizing antibody responses in host animals. One hundred twenty pregnant beef cows were selected and immunized eight weeks prior to parturition with the recombinant VP8* protein. Colostral and milk samples were collected 12 hours and 10 days post-calving, respectively, and the virus neutralizing titers elicited by the recombinant subunit antigen were determined by plaque reduction assay. Colostral antibody titres of the vaccinated group were significantly higher than those of the unvaccinated control group, and these titers were equivalent to the titers elicited by a commercial vaccine. While titers of the control group rapidly decreased to 220 after 10 days of calving, neutralizing titers in the milk of the vaccinees remained 510. Rabbit and bovine antibodies induced by the recombinant VP8* protein were also able to neutralize bovine rotavirus P5 serotype (B641) at significant level and P11 serotype (B223) moderately. Our results suggest that the E. coli-produced recombinant VP8* protein can be an useful subunit vaccine candidate to prevent rotavirus infection in new-born calves.

Immunization with naked plasmid DNA effectively induces both humoral and cell-mediated immunity to vaccine antigens and can confer protection against numerous infectious diseases. To explore the potential use of DNA immunization to induce rotavirus-specific immune responses, we used plasmid DNA encoding the VP4 gene of bovine rotavirus (BRV). Intrasmuscular injection of the plasmid encoding the VP4 gene into C57BI/6 mice induced cell-mediated immunity as measured by cytokine production. Although DNA immunization did not induce a detectable BRV-specific antibody response, DNA-immunized animals were primed for antibody production and a cellular immune response. Following viral inoculation, the immunized animals displayed an enhanced number of BRV-specific antibody-secreting cells and cytotoxic activity. The immune response induced by DNA immunization alone or followed by viral inoculation was biased toward IFN-gamma production (Th1-like). CD4+ lymphocytes were the major source of IFN-gamma production in the spleen following DNA immunization. In contrast, a balanced cytokine production was observed in the spleens of animals receiving whole virus. These experiments showed that DNA immunization with a gene encoding the VP4 protein of BRV stimulated a Th1-like immune response in mice, and this bias in the immune response persisted following exposures to whole virus.

Bovine rotavirus VP8*, N-terminal trypsin cleavage product of VP4, was produced in Escherichia coli. To examine if this antigen could induce neutralizing antibody responses, different species of animals were immunized with the recombinant VP8* protein. The VP8* antigen was found to stimulate a neutralizing immune response in rabbits. When VP8*-immunized mice were exposed to bovine rotavirus strain C486, significantly higher antibody responses were observed than if they were only exposed to C486. To stimulate a current vaccination protocol in the field with livestock, mice were exposed to live C486 virus first and then to VP8*. These mice had the elevated immune responses indicating that VP8* could boost immunity in primed mice. The immune response to VP8* was also tested in pregnant cows. The efficacy of VP8* in stimulating milk antibody was compared with a commercial inactivated vaccine. Differences in colostral antibody titres between VP8*-vaccinated and unvaccinated cows were statistically significant (P < 0.05) and equivalent to the commercial vaccine (P = 0.0569). The milk antibody titres on day 10 were comparable between VP8*- and commercial vaccine-vaccinated animals and were significantly higher (P < 0.05) than in unvaccinated controls. Furthermore, rabbit and bovine antibodies induced by VP8* were able to neutralize different P types of bovine rotaviruses to varying degrees, suggesting that serotype-specific and cross-reactive epitope(s) are present on the VP8* protein of rotavirus. Taken together, E. coli-expressed VP8* may be useful as a subunit vaccine candidate for bovine rotavirus.

A synthetic peptide corresponding to the trypsin cleavage site on the 84 k protein of bovine rotavirus was synthesized (VP4-peptide). This synthetic peptide could be cleaved by trypsin and therefore possessed the enzyme binding site present on the authentic protein. Further proof that this peptide mimicks the authentic trypsin cleavage site was the specific reaction of anti-peptide serum with the 84 k protein. The reaction of anti-peptide serum with infectious virus neutralized infectivity thereby supporting the biological importance of this site. Another interesting characteristic of this peptide was its ability to bind to the nucleocapsid protein resulting in a laddering effect on the nucleocapsid monomer (45 k), dimer (90 k) and trimer (135 k) [Gorzilia et al., J. Gen. Virol. 66, 1889-1900 (1985); Sabara et al., J. Virol. 53, 58-66 (1985); Sabara et al., J. Gen. Virol. 67, 201-212 (1986)]. Definitive proof of binding was provided by the fact that the increments in the ladder corresponded to the molecular weight of the synthetic peptide and that anti-peptide serum specifically reacted with the ladder formations. The laddering of the nucleocapsid could be eliminated by incubation with trypsin thus further supporting the formation of a synthetic peptide-nucleocapsid complex. Due to the ability of the peptide to bind to trypsin and to the nucleocapsid protein its biological activity was investigated. It appeared that increasing concentrations of the peptide reduced the rate of virus plaque formation, thereby suggesting that virus replication was inhibited. These results illustrate two features of this synthetic peptide which warrant further investigation; (1) its capacity to mimic an enzyme cleavage site and, (2) its ability to complex tightly to another protein. In protection-challenge experiments performed using a murine model, animals immunized with VP4-peptide provided protection passively, to neonates suckling on the immune dams, against a virulent rotavirus. The potential applications of this peptide in rotavirus diagnosis, therapy and synthetic peptides based vaccine is discussed.

Synthetic peptides derived from bovine rotavirus C-486 (BRV) outer capsid (VP7 and VP4) and inner capsid (VP6) proteins were tested to evaluate their ability to prime and induce an anti-rotavirus antibody response. Peptides corresponding to the amino acid residues 232-255 of VP4 (VP4-peptide), 275-295 of VP7 (VP7-peptide) and 40-60 of VP6 (VP6-peptide) of BRV were chemically synthesized. These peptides were coupled to carrier proteins (either keyhole limpet haemocyanin (KLH) or recombinant rotavirus inner capsid protein-VP6 assembled into virus-like particles (VP6-carrier) were used as carrier to link the synthetic peptides under study), and the resulting conjugates were used to immunize rotavirus seronegative mice. An enzyme-linked immunosorbent assay (ELISA) was used to determine anti-peptide and anti-rotavirus antibody titres in serum samples collected after immunization. All peptides were immunogenic in mice and induced the production of anti-peptide antibodies, but with the exception of VP6-peptide they were not able to induce anti-rotavirus antibodies as measured by ELISA. Western blot analysis indicated that antibodies against each peptide were able to react with the respective authentic viral proteins of various rotavirus serotypes. To determine if a peptide-primed animal would respond to native viral proteins, animals were subsequently injected with purified BRV. A rapid and high anti-rotavirus antibody titre, in addition to a rise in anti-peptide antibody titre, was observed in peptide-primed mice. Furthermore, the sera obtained from these mice neutralized the virus under in vitro conditions. The significance of these results in relation to a potential rotavirus synthetic peptide-based vaccine is discussed.

The molecular basis of pathogenesis in vivo for a virulent mouse rotavirus (MRV) and a less virulent bovine rotavirus (BRV) were compared under in vitro and in vivo conditions. Obvious differences in the mobility of several genomic RNA segments were observed in one-dimensional gels. Under in vitro conditions, partial proteolytic peptide mapping identified differences between the two outer capsid proteins of these virus and no difference in inner capsid protein was observed. Since it has been observed by us and others that the gene coding for VP4 protein plays a significant role in determining virulence, the variability observed in the present study between the 84 k proteins (VP4) provided a basis for further investigations in order to locate a potential virulence determinant. A comparison of the carboxypeptidase digests of the MRV- and BRV-VP4 revealed an area of variability between amino acids 307 and 407, which may represent a site of virulence determinant. Under in vivo conditions the virulence of both parenteral BRV and MRV isolates and their corresponding reassortants (with replaced gene 4) were studied in murine and bovine hosts. Like their parents, BRV and MRV isolates, reassortants obtained by replacement of gene 4 in BRV with MRV gene 4 indicated that the dose of the virus isolate used and the clinical outcome in vivo was determined by gene segment 4. The implications of these findings to elucidate the molecular basis of pathogenesis of rotaviruses are discussed.

The proteins, genomic RNA and disassembly conditions and pathogenesis in vivo for a virulent mouse rotavirus (MRV) and a less virulent bovine rotavirus (BRV) were compared. An obvious difference in the mobility of several genomic RNA segments were observed in one-dimensional gels. Reassortants obtained by replacement of gene 4 in BRV with MRV gene 4 indicated that the dose of the virus used and the clinical outcome in vivo was determined by gene segment 4. Under in vitro conditions, a comparison of the inner capsid proteins by partial proteolytic peptide mapping did not reveal any difference between corresponding proteins. However, this technique did identify differences between the two corresponding outer capsid proteins of these viruses. These differences, in turn, may account for the increased stability of MRV, as compared to BRV, when subjected to calcium-chelating and chaotropic agents and may be one of the mechanisms involved in conferring virulence on the virus. The observed variability between the 84K proteins (VP4) provided a basis for further investigations in order to locate a potential virulence determinant, since it has been observed by us and others that the gene coding for this protein plays a role in determining virulence. A comparison of the carboxypeptidase digests of the MRV and BRV VP4 revealed an area of variability between amino acids 307 and 407, which may represent the site of a virulence determinant.

Rotavirus structural proteins VP4, VP6 and VP7 from Bovine Rotavirus Strain C486 were cloned and expressed in a baculovirus expression system. Combinations of the proteins were assembled into a series of virus-like particles, and a murine model was used to determine the capacity of the recombinant proteins and particles to induce protective immunity. All of the proteins induced humoral immunity as measured by an ELISA against whole virus. However, only the antisera from animals immunized with VP4 neutralized virus and inhibited haemagglutination. Challenge of neonates born to animals immunized with VP4 protein on assembled particles or in cell lysates showed protection against challenge with both homologous (bovine C486) and heterologous (SA-11) strains of rotavirus. In contrast, the offspring of mice immunized with VP6 were only partially protected. Neonates of animals immunized with virus-like particles composed of VP7 assembled on VP6 spherical particles were protected against challenge with the homotypic virus and significantly protected from a heterotypic challenge whereas unassembled VP7 protein provided only partial protection against challenge.

A system which allows for the efficient attachment of synthetic peptides to spherical virus-like particles assembled from the VP6 rotavirus nucleocapsid protein is described. This attachment was shown to be mediated by peptide-protein interactions and did not require additional chemicals for conjugation. The resulting large macromolecular structure was highly immunogenic for both the VP6 protein and the coupled peptides. The antibody response to peptides bound to VP6 particles was of higher titre and longer duration than that induced by other carriers. In addition, the response to VP6-coupled peptides was not affected by prior exposure to rotavirus and exhibited a range of immunoglobulin subclasses in the absence of an adjuvant. These data demonstrate that assembled VP6 spherical particles are useful carriers for low doses of synthetic peptides.

Simultaneous coinfection of Spodoptera frugiperda cells with baculovirus recombinants containing the genes for VP6 and VP7 of bovine rotavirus strain C486 resulted in the production of spherical particles resembling smooth, double-shelled rotavirus. These particles were predominantly located in the cell culture supernatant instead of being cell associated. Pretreatment of infected-cell culture supernatants with CaCl2, prior to particle purification, was found to enhance the smooth appearance of the particles. The authenticity of the proteins making up the particles was demonstrated by their reactivity with antiserum specific for double-shelled C486 rotavirus.

We have evaluated the potential of two peptides derived from highly conserved regions of rotavirus outer capsid proteins (VP7 and VP4) to act as a rotavirus vaccine. The capacity of peptides coupled to rotavirus VP6 spherical particles to provide passive protection in a murine model was compared with the protection induced by peptide-keyhole limpet hemocyanin (KLH) conjugates. Female mice were immunized a total of three times before and during pregnancy. Suckling mouse pups were challenged at 7 days of age with either homologous or heterologous rotavirus serotypes. The efficacy of vaccination was determined by analyzing the clinical symptoms and measuring xylose adsorption in the intestine. In this model the VP4 peptide-VP6 conjugate provided protection equal to that obtained using bovine rotavirus (BRV) as the immunogen. The VP7 peptide-VP6 conjugate provided slightly less protection than the VP4 peptide-VP6 conjugate. A mixture of the VP4 peptide-VP6 and VP7 peptide-VP6 conjugates provided better heterologous protection than immunization with BRV. In contrast, KLH-conjugated peptides provided only partial protection. The significance of a synthetic-peptide-based rotavirus vaccine in the prevention of rotavirus infections is discussed.

A major problem in the development of useful animal subunit vaccines has been the generation of immune responses to weakly immunogenic molecules. For this purpose a new and effective delivery system has been devised. This system is based upon the inner capsid of bovine rotavirus. Under the appropriate conditions, the inner capsid protein, designated BP6, can be made to self-assemble in vitro and form spherical particles. These particles possess an inherent capacity to target to cells of the immune system. Exploitation of these properties has led to the development of technology to couple antigens to the VP6 particles such that the sphere acts as a novel immunological carrier. This is based on a "binding peptide" derived from another rotavirus peptide, VP4, as well as on more traditional techniques of chemical coupling. We have coupled peptides or proteins to this carrier via the binding peptide and have shown that every epitope tested to date gave excellent immune responses. Furthermore, using this carrier, immunity has been developed without the use of adjuvants. This has far-reaching implications for animal and human immunization.

Neuropeptides and neurohormones (neurotransmitters) have been shown to modulate immune responses in vitro and in vivo. Since reproduction and lactation are regulated by neurohormones, we investigated whether neurohormones could enhance anti-rotavirus immunity in milk. Rotavirus-free mice were immunized orally with killed bovine rotavirus (BRV) and bred 6 weeks post-immunization. Post-whelping, each group of dams (ten mice/group) was given a single injection of prolactin (PRL), estrogen, PRL and estrogen or testosterone. The effects of neuropeptides, substance P (SP) and somatostatin (SS) on serum and lactogenic anti-rotavirus humoral immune responses were also investigated. The results revealed that in the groups given PRL or estrogen, anti-rotavirus antibody titers in milk and serum were enhanced. In contrast, testosterone had a negative effect on antibody titers. The administration of neuropeptide SP resulted in some enhancement of the lactogenic anti-rotavirus antibody titer at day 9 post-whelping whereas the opposite effect was observed following administration of SS. Prolactin given at 100 micrograms/mouse, on the day after whelping, gave optimum milk and serum antibody responses. Neurotransmitters potentiated immune responses to the weaker immunogenic proteins, VP4 and VP7 as well as to the strongly immunogenic VP6. In order to verify that the enhancement of anti-rotavirus antibody production was due to PRL and not to other factor(s), bromocriptine (BCR), a selective PRL inhibitor, was used as a control. Mice given BCR exhibited a drastic reduction in anti-rotavirus antibody in serum and milk. The role of neurotransmitters in the modulation of the lactogenic immune response and its significance in protection of neonates from enteric infections is discussed.

A method was further developed to screen non-tissue-culture-adapted bovine rotaviruses for serotype, using a neutralization test with infectious fecal rotavirus. One of those rotaviruses (B223) which was not blocked by antiserum to the neonatal calf diarrhea virus (NCDV) serotype was then adapted to cell culture in the presence of the antiserum for two or more passages and hyperimmune antiserum to this isolate had a 60-fold-higher homologous neutralization titer than with the NCDV serotype rotavirus. Seventy-three isolates were serotyped and eight (11%) were not of the NCDV serotype (bovine rotavirus serotype I). Of these eight, five belonged to the new bovine rotavirus serotype II and three were not typed, indicating the existence of one or more further serotypes. Cross-protection studies in gnotobiotic calves showed that cross-protection only occurred between rotaviruses of the same serotype, and even a minor serotype difference was sufficient for the calves to show a lack of cross-protection. The serotypes (I and II and the three untyped isolates) also showed differences in the rate of migration in polyacrylamide gel electrophoresis of some of their RNA segments (no. 4, 6, 7, 8, 9, 10), indicating that they were of different electropherotypes.

Rotaviruses have been shown to be of importance as aetiologic agents of gastroenteritis in infants and in domestic animals of several species. Hemagglutinins were prepared from two Canadian isolates of bovine rotavirus and from one isolate of a simian rotavirus. A United Kingdon isolate of bovine rotavirus was shown not to possess hemagglutinating activity, indicating a strain difference between a Canadian and United Kingdom bovine rotavirus. In hemagglutination-inhibition (HAI) tests a rabbit hyperimmune (two injections) serum, prepared to one of the bovine rotaviruses, was not helpful in distinguishing the two bovine viruses because of cross-reactions between the viruses. However, it was possible to distinguish the bovine viruses from the simian virus with this serum. When guinea pig immune sera were prepared to the four rotavirus strains and tested with the three hemagglutinins in the HAI test, antigenic differences between the four strains of rotavirus were demonstrated. Hyperimmune guinea pig serum prepared to a strain of human rotavirus did not inhibit any of three hemagglutinins indicating that the human strain is different from the three rotavirus strains which gave hemagglutinins.

The genomic RNA patterns of six different bovine rotavirus isolates were analyzed on high-percentage polyacrylamide gels (12.5, 13.6, and 17.5%). In contrast to the RNA patterns exhibited by conventional gel systems, those on the high-percentage gels showed an improvement in segment resolution which consequently aided in the detection of extensive band splitting in these patterns. The ability to clone out various electrophoretically distinct virus subpopulations from each of the six isolates provided an explanation for the band splitting detected by the high-resolution gels. The significance of the coexistence of genetically distinct rotavirus populations within a single host is discussed.

The nucleocapsid protein (VP6) and outer capsid glycoprotein (VP7) of bovine rotavirus (BRV) assemble in vitro, in 0.01 M Tris-HCl, pH 8.0, 50 mM CaCl2, into smooth particles resembling double-shelled BRV. That the two proteins interact is demonstrated by the immunoprecipitation of both by antibody directed against either VP6 or VP7. The calcium-dependence, particle morphology, and immunoreactivity in ELISA suggest that VP7 is presented authentically on the outer capsid. The implications for rotavirus morphogenesis are discussed.

Bovine rotavirus grown in the presence or absence of tunicamycin was analyzed with respect to yield of infectious virus, the ratio of complete to incomplete particles, and polypeptide composition. Tunicamycin at a concentration of 1 microgram/ml reduced virus yields by 4 logs and completely prevented the incorporation of [3H]uridine into complete rotavirus particles, as determined by cesium chloride gradient analysis. Concomitant with a reduction in complete particles, three rotavirus polypeptides shifted in their relative position on polyacrylamide gels from 41,900-molecular-weight position (41.9K), 29.3K, and 16.1K to migrate at 35.5K, 22.7K, and 15.5K, respectively. Limited proteolysis indicated that the lower-molecular-weight polypeptides possessed the same constituent peptides as the larger polypeptides, suggesting that they represented the unglycosylated equivalents. These results suggest that interference with glycosylation prevents proper assembly of the outer coat proteins in bovine rotavirus.

Preliminary characterization of naturally occurring temperature-sensitive bovine rotavirus mutants, generated via gene reassortment, indicated that the product of gene segment 10 can influence virus adsorption to MA-104 cells. Cell-free translation of this gene segment in a rabbit reticulocyte system was then utilized to identify a polypeptide with a molecular weight of ca. 21,500 as the primary translation product. This identification provides a basis for further studies of rotavirus assembly and cell interactions which may, in turn, aid in elucidating the replicative strategy of the virus.

A hemagglutinin has been prepared from Nebraska calf diarrhea virus (NCDV) propagated in BS-C-1 cell line. After cesium chloride centrifugation, the hemagglutinin of the bovine rotavirus was found to be associated with intact virions (density 1.355 g/ml) but not with virions lacking an outer capsid layer (density 1.375 g/ml). In hemagglutination-inhibition (HAI) tests, the hemagglutinin reacted specifically with NCDV serum, and HAI seroconversions were detected in some sera tested. Cross-reactions were observed in complement fixation tests between the human and bovine rotaviruses but were not demonstrated by HAI, suggesting that the hemagglutinin detects a specific rather than a group antibody response.

A murine model to study enteritis induced by bovine (BRV) and murine rotavirus (MRV) has been developed. The course of infection was determined by clinical symptoms of diarrhea and virus isolation as well as histopathological, immunohistochemical, and electron microscopic methods. Both isolates were able to replicate and produce clinical symptoms in neonatal mice. Rotavirus-free neonates were orally inoculated with MRV or BRV and observed over a 192-hr postinoculation (HPI) period. Following infection with 10(4) PFU of virus, diarrhea and maximal intestinal dysfunction, as measured by xylose absorption, did not occur until beyond 20 hr postinfection even though maximal virus production occurred at 10-15 HPI. Immunohistochemically and by electron microscopy we were able to demonstrate viral antigen and virus particles in the enterocytes of villous tips at 5-8 HPI. The appearance of diarrheal symptoms was dependent on the virus dose and the type of virus isolate inoculated. The disease could be induced with doses as low as 1 x 10(2) PFU/mouse of BRV and 1 x 10(1) PFU/mouse of MRV. On the basis of these results, MRV was found to be more virulent than BRV in this model. The model should prove useful for studies designed to assess rotavirus virulence genes and for vaccine protection studies. This work emphasizes the need for early sample collection for critical evaluation of any vaccine or antiviral agent using this model.

The effect of different routes of immunization with either live or killed bovine rotavirus (BRV) on the production of lactogenic antibody response in mice was evaluated. The routes of immunization were intramuscular (IM), oral (O) or intradermal in the mammary region (IMam). Following immunization, serum antibody responses were monitored by an enzyme linked immunosorbent assay (ELISA). Following whelping, the mice were allowed to stay with their mother until sacrificed on alternate days post-parturition from day 1-11. Milk from their stomach was collected for antibody titration by ELISA and virus neutralization test. Regardless of the routes of immunization, a rapid increase in serum anti-rotavirus antibody titers was observed for the first 5 wk after immunization followed by a gradual decline. After parturition, the mean antibody titer of lacteal secretions, as determined by ELISA, increased gradually for 7 days with the greatest increase on day 9, followed by a decrease in anti-rotavirus antibody. These titers also correlated with antibody titers in milk as measured by virus neutralization test. The best lactogenic antibody response was observed when IMam X IM X 2 route of immunization was used with live BRV as the antigen. Interestingly, immunization via the oral route with killed BRV also resulted in good antibody responses. In contrast, in the group where killed BRV was used, animals receiving 3X orally had the highest antibody titer. The distribution of different antibody subtypes in milk samples revealed IgG to be the predominant antibody followed by IgM and IgA. Irrespective of the route of administration, there was an increase in IgA on day 9 as compared to day 1 in most of the groups. The significant role played by mucosal immunity in passive protection and the possible ways to modulate subtype specific lactogenic immune response are discussed.

Rotaviruses are generally difficult to isolate and culture in vitro; therefore, virus isolation has not been used as a method of diagnosing this group of agents. The present report describes a simple procedure for isolating bovine rotaviruses directly from feces after pretreatment of fecal samples with trypsin. This procedure resulted in virus isolation from five of five samples that contained virus particles, as demonstrated by electron microscopy, and four of seven samples where virus particles could not be observed but were considered positive by the presence of immunofluorescent-staining cells in feces. Virus could not be isolated from "normal" feces. If the virus was not passaged in the presence of trypsin, the infectivity was gradually lost, but infectivity could be restored again if trypsin was added, resulting in increased virus spread and concomitant increase in virus yield. The application of this technique as a diagnostic tool for bovine and other rotaviruses is briefly discussed.

Previously we reported that calcium plays an important role in the maturation of bovine rotavirus (M. S. Shahrabadi and P. W. K. Lee, 1986. Virology 152, 298-307). We now demonstrate that the formation of mature double-shelled (L) particles was strictly dependent on the concentration of calcium present in the growth medium. The formation of single-shelled (D) particles did not appear to be a calcium-mediated process. Subsequent labeling studies using 45Ca revealed that calcium was incorporated into the L particles but not the D particles. The previously noted decreased level of the outer capsid protein VP7 (42K) in calcium-deprived cultures was now found to be due to the preferential degradation, and not to the impaired synthesis, of this protein in the absence of calcium. It was further demonstrated that calcium had a stabilizing effect on VP7 and that VP7 synthesized in the presence of calcium was not degraded upon subsequent calcium deprivation. Protein degradation during calcium deprivation was apparently limited to the mature form of VP7 since the unglycosylated precursor (pVP7), formed in the presence of tunicamycin, was found to be stable under this condition. Electron microscopic examination of infected cells revealed that in the presence of calcium, virus maturation took place by the budding of viral cores through the endoplasmic reticulum (ER). No such budding was observed in calcium-deprived cells. In these cells mature virions were absent and membrane fragments could be found associated with viral cores or single-shelled particles.

Previous experiments demonstrated that an antigenic site responsible for virus neutralization and cell attachment was located on a 14,000-molecular-weight fragment of the major bovine rotavirus (BRV) glycoprotein (M. Sabara, J. E. Gilchrist, G. R. Hudson, and L. A. Babiuk, J. Virol. 53:58-66, 1985). However, it was necessary to investigate whether this fragment also had the ability to induce the production of neutralizing antibodies. Upon immunization of mice, the bovine serum albumin-conjugated 14,000-molecular-weight fragment, the unconjugated 14,000-molecular-weight fragment, and the native glycoprotein all induced a similar neutralizing antibody response, albeit to a lesser extent than did the infectious, whole virus. In addition, immuno-blot enzyme-linked immunosorbent assay analysis of the reactivity of anti-peptide serum versus anti-glycoprotein serum with the glycoprotein was very comparable. These results suggest that the 14,000-molecular-weight fragment may represent not only a biologically active region but also an immunodominant area of the glycoprotein.

The 38,200-molecular weight (unreduced)/41,900-molecular-weight (reduced) glycoprotein of bovine rotavirus, isolate C486, was identified as the major neutralizing antigen. This glycoprotein as well as the corresponding glycoprotein of another bovine rotavirus serotype also specifically attached to cell monolayers under normal conditions for virus adsorption in vitro. Further support for this glycoprotein being directly responsible for virus attachment to cells was that (i) infectious virus of both serotypes could compete with the C486 glycoprotein for cell surface receptors, and (ii) neutralizing monospecific antiserum and neutralizing monoclonal antibodies directed toward the glycoprotein could block this virus-cell interaction. Preliminary epitope mapping of the glycoprotein with monoclonal antibodies further localized the neutralization-adsorption domain to a peptide with an approximate molecular weight of 14,000. The effect of two protein modifications, glycosylation and disulfide bridging, on the reactivity of this peptide with antibodies and cell surface receptors was investigated. It was demonstrated that, whereas glycosylation did not appear to affect these reactivities, disulfide bridging seemed to be essential.

The nucleocapsid protein of bovine rotavirus was shown to exist in trimeric units in both the virus particle and in infected cells, with the subunits linked by non-covalent interactions. These trimeric units complex further by disulphide bridges into larger units which may represent the hexameric structures observed by electron microscopy. Visualization of various nucleocapsid protein complexes was also achieved on polyacrylamide gels by treating virus preparations with urea at 37 degrees C or boiling in the presence and absence of 2-mercaptoethanol. Since virus particles devoid of nucleic acid were also broken down into trimeric subunits by such treatments, assembly of virus particles appears not to require an RNA-protein interaction. Four nucleocapsid-specific monoclonal antibodies with low neutralizing ability reacted with the monomeric (45,000 mol. wt., 45K), dimeric (90K), trimeric (135K) and trimeric pair (270K) subunits, indicating that a site responsible for neutralization is probably exposed after assembly of these subunits. Analysis of radiolabelled virus revealed that a high proportion (80%) of infectious particles could be immunoprecipitated by these monoclonal antibodies, suggesting that the virus particles are either partially double-shelled or have the nucleocapsid exposed on the surface. The monoclonal antibodies also cross-reacted with the nucleocapsid proteins of simian (SA11), pig (OSU), bovine (NCDV and UK) and human (Wa and ST4) rotaviruses in an immunoblot ELISA reaction. Since these six viruses belong to two different subgroups, it is likely that the antibodies did not recognize the subgroup-specific site, but a shared exposed antigenic determinant. Due to the hexameric configuration of the nucleocapsid in virus particles the neutralizing epitope may be repeatedly presented and, therefore, may contribute to the immunogenicity of this protein.