Non-lytic extraction and characterisation of receptors for multiple strains of rotavirus.

To characterise the cellular receptors for rotavirus, we used the detergent octyl-beta-D-glucopyranoside (octyl-glucoside/OG) to extract the receptors for bovine, simian, porcine and human rotaviruses from MA104 and HT29 cells. An octyl-glucoside concentration of 0.2% dramatically reduced the susceptibility of treated cells to infection, while leaving them metabolically active, and as a result the depleted receptors were able to regenerate. Periodate treatment of the MA104 and HT29 octyl-glucoside extracts significantly decreased the ability of these extracts to neutralise rotavirus infectivity, revealing carbohydrate as component of the extracted receptors for Wa and NCDV. Treatment of MA104 cells with the metabolic inhibitors tunicamycin, deoxymannojirimycin and BenzylGalNAc suggested N-linked carbohydrate may be more important than O-linked in infection by some strains of rotavirus. Furthermore, by including cycloheximide during the regeneration of depleted receptors we found evidence that porcine rotavirus CRW8 may use a glycolipid-based receptor, while NCDV and Wa use a glycoprotein. The regenerating properties of the rotavirus receptors allowed repeated harvesting of cell surface molecules using octyl-glucoside on consecutive days, and these extracts were used to visualise virus binding in a virus overlay protein blot assay (VOPBA). Using VOPBAs, we observed both Wa and NCDV appear to recognise proteins of approximately the same molecular weight present on MA104 and HT29 cells.

Selection of rotavirus VP4 cell receptor binding domains for MA104 cells using a phage display library.

Rotavirus infection of host cells, like other viruses, is a complex process that has not been fully elucidated, and much attention has been focused on the regions of the viral attachment protein, VP4, that are involved in binding to the cellular receptor. In this study, phage display technology was employed to generate a g3p VP4 gene-targeted phage display peptide library using the porcine rotavirus strain CRW8, and a method was optimised for panning this library on adherent MA104 cells to identify receptor binding domains. Recombinant phage that displayed expressed peptides from both the rotavirus VP4 trypsin cleavage products VP8* and VP5* were selected, and while some of the phage clones contained insert sequences from regions of VP4 implicated previously in cell binding and infection, new domains were also identified. In all, four regions within VP8* and six regions of VP5* were selected by panning. To our knowledge, this paper is the first description of using a gene-targeted phage display library to identify receptor binding domains on viral proteins.

Synthesis and biological evaluation of sialylmimetics as rotavirus inhibitors.

Rotaviruses cause severe gastroenteritis in infants and are estimated to be responsible for over 600 000 deaths annually, primarily in developing countries. The development of potential inhibitors of this virus is therefore of great interest, particularly since the safety and efficacy of rotaviral vaccines has recently been questioned. This study describes the synthesis of a variety of compounds that can be considered as mimetics of N-acetylneuraminic acid thioglycosides and the subsequent in vitro biological evaluation of these sialylmimetics as inhibitors of rotaviral infection. Our results show that readily accessible carbohydrate-based compounds have the potential to act as inhibitors of rotaviral replication in vitro, presumably through inhibition of the rotaviral adhesion process.

Rotavirus infection of MA104 cells is inhibited by Ricinus lectin and separately expressed single binding domains.

Various lectins were tested for blocking rotavirus infection of MA104 cells and it was observed that galactose-specific lectins were the most inhibitory. Of these Ricinus agglutinin was able to inhibit infection (by human and animal strains) at concentrations as low as 10(-9) M. In addition, in a virus overlay protein blot assay Ricinus agglutinin competed with simian rotavirus SA11 for binding to solubilized MA104 proteins. Amino acid sequence comparisons revealed similarity between the ricin toxin B subunit (which contains two separate carbohydrate-binding motifs: single binding domains (SBD) 1 and 2) and rotavirus spike protein VP4. A filamentous phage display system was used to independently express the two binding domains and while SBD1 inhibited infection of MA104 cells by CRW8, NCDV, and to a lesser extent Wa, SBD2 blocked only CRW8 and NCDV infection. Furthermore inhibition of CRW8 infection was a direct result of phage inhibiting virus attachment to cells. When amino acid 248 within SBD2 was mutated from the ricin toxin to the Ricinus agglutinin sequence this phage clone showed reduced binding to galactose and was no longer able to inhibit virus infection. Thus, rotavirus recognizes galactose as an important component of the receptor on MA104 cells.

Haloalkane dehalogenases: structure of a Rhodococcus enzyme.

The hydrolytic haloalkane dehalogenases are promising bioremediation and biocatalytic agents. Two general classes of dehalogenases have been reported from Xanthobacter and Rhodococcus. While these enzymes share 30% amino acid sequence identity, they have significantly different substrate specificities and halide-binding properties. We report the 1.5 A resolution crystal structure of the Rhodococcus dehalogenase at pH 5.5, pH 7.0, and pH 5.5 in the presence of NaI. The Rhodococcus and Xanthobacter enzymes have significant structural homology in the alpha/beta hydrolase core, but differ considerably in the cap domain. Consistent with its broad specificity for primary, secondary, and cyclic haloalkanes, the Rhodococcus enzyme has a substantially larger active site cavity. Significantly, the Rhodococcus dehalogenase has a different catalytic triad topology than the Xanthobacter enzyme. In the Xanthobacter dehalogenase, the third carboxylate functionality in the triad is provided by D260, which is positioned on the loop between beta7 and the penultimate helix. The carboxylate functionality in the Rhodococcus catalytic triad is donated from E141. A model of the enzyme cocrystallized with sodium iodide shows two iodide binding sites; one that defines the normal substrate and product-binding site and a second within the active site region. In the substrate and product complexes, the halogen binds to the Xanthobacter enzyme via hydrogen bonds with the N(eta)H of both W125 and W175. The Rhodococcusenzyme does not have a tryptophan analogous to W175. Instead, bound halide is stabilized with hydrogen bonds to the N(eta)H of W118 and to N(delta)H of N52. It appears that when cocrystallized with NaI the Rhodococcus enzyme has a rare stable S-I covalent bond to S(gamma) of C187.

Characterisation of G serotype dependent non-antibody inhibitors of rotavirus in normal mouse serum.

Serotype specific (non-immunoglobulin) inhibitors of rotavirus have been identified in normal mouse serum obtained from BALB/c, CBA, and BL10 mice. Sialic acid was essential for the neutralising activity sera treated with the neuraminidase from Vibrio cholerae failed to neutralise rotavirus. G serotypes 4, 5, 7, 8, 9, and 10 were unaffected by the inhibitor(s) while G serotypes 1, 2, 6 and two G3 strains were neutralised to significant titres. Assessment of neutralisation of reassortants suggested that VP7 is the virus protein involved in the interaction although it remains possible that VP7 is influencing VP4 binding. Analysis of the sera by Western blot followed by virus overlay confirmed that binding is dependent on the presence of sialic acid. The human strain tested, Wa, bound to two (glyco) proteins (50 and 80 kDa) while the bovine strains tested, NCDV and UK bound to one (55 kDa) and two (36 and 55 kDa) proteins respectively. This indicates that while the bovine rotaviruses may bind to a common element, the human strain binds to clearly distinct proteins. We propose that these inhibitors interact with animal rotaviruses in a manner analogous to that by which they attach to target cells. The glycoprotein to which NCDV bound was purified and identified by N-terminal sequencing as murine alpha-1-anti-trypsin (MuAAT) and was confirmed to possess both neutralisation and anti-trypsin activity. Since MuAAT is known to possess only three N-linked glycans, identification and analysis of the actual virus-binding structure should now be possible.

Periplasmic expression of part of the major rotavirus capsid protein VP7 containing all the three antigenic regions in Escherichia coli.

Part of the porcine rotavirus outer capsid protein VP7 containing all the three antigenic regions was expressed as a chimeric protein with bacterial alkaline phosphatase (AP) in E. coli. The construct contains an ompF promoter, the DNA encoding the signal sequence and the first 12 amino acids of mature OmpF, part of vp7 and the DNA encoding mature AP. The chimeric protein is stable, retains the biological property of AP and ability to react with polyclonal antiserum against the virus, and can be exported through the bacterial inner membrane into the periplasm.

Development of rotavirus molecular epidemiology: electropherotyping.

Early in the era of rotavirology it was realized that the characteristic patterns of bands produced in polyacrylamide gels following electrophoresis of genomic dsRNA were useful for checking the identity of rotavirus isolates. However it was Romilio Espejo who first proposed the use of this technique for epidemiology, although most others did not take the suggestion seriously because the technique was then rather specialized and RNA staining methods were not very sensitive. Using samples collected by Ruth Bishop in Melbourne following the original identification of human rotaviruses, Sue Rodger recorded the "electropherotypes" of all samples available to 1979 and painstakingly compared them, side by side (since minor variations in conditions, especially temperature, alter the relative migration distances of dsRNA bands). These efforts produced the first longitudinal, extensive study of human rotavirus strain variation. Since then, technical improvements have greatly increased the sensitivity of the procedures, and electropherotyping has been recognized as a powerful and economical method for epidemiological studies of rotaviruses.

Rotavirus antigenicity is affected by the genetic context and glycosylation of VP7.

Rotavirus variants resistant to neutralization were selected using monoclonal antibodies (N-MAbs) raised to VP7 of rotavirus G types 2, 3, and 6. Their neutralization resistance patterns and deduced VP7 amino acid sequences were obtained. Variants selected by two G2-specific N-MAbs from the homologous parent virus RV-5 showed single amino acid (aa) mutations in the antigenic A region. However, variants selected from reassortant virus RV-5 x SA11 (all genes from SA11 virus except that encoding VP7, which was from RV-5 virus) fell into two neutralization resistance groups. The first group showed identical mutations to the variants selected from RV-5 virus. The second group showed antigenic C region mutations, either alone or in combination with a mutation at aa 69. Variants selected from G3 parent viruses glycosylated at position 238 had a mutation at aa 96 in the A region, otherwise a C-region mutation at 211 was selected. Mutations at amino acid positions 94 or 96 were selected by monoclonal antibodies specific for each of the three serotypes. G3-specific monoclonal antibodies also selected mutations at position 148 and the new position of 264. This latter mutation resulted in substitution of aspartic acid for glycine and was located in a highly conserved and hydrophobic region of VP7. A G2-specific N-MAb selected variants with a mutation at aa 190 producing a new, utilized glycosylation site which we propose to be in new antigenic site E. The positions of mutations in antigenic variants and their antigenicity were determined by parental background genes and VP7 glycosylation.

Avian reovirus sigma C protein contains a putative fusion sequence and induces fusion when expressed in mammalian cells.

The biological functions of the structural protein sigma C, from avian reovirus strain RAM-1, were investigated in this study. A putative fusion peptide in sigma C was recognized in the deduced amino acid sequence by homology with Pneumovirus fusion sequences, and it was thus postulated that this protein may be involved in the formation of syncytia in cells infected with RAM-1. The sigma C gene was cloned and expressed in mammalian (COS7) cells and the sigma C protein was found to induce syncytia. It was therefore concluded that this protein is indeed responsible for avian reovirus-induced cell fusion. It was also found that sigma C caused condensation of the nuclei within a syncytium, as observed in RAM-1-infected cells. On the basis that this represented condensation of the chromatin, the inhibition of cellular DNA synthesis by the virus and by the sigma C protein was measured. It was found that the virus caused a 50% reduction in cellular DNA synthesis, but the sigma C protein did not inhibit DNA synthesis. Therefore pyknosis of the nuclei and inhibition of cellular DNA synthesis by RAM-1 are likely to be separate events.

Self-injection with olive oil. A cause of lipoid pneumonia.

A 48-year-old man with unipolar depression and a psychosexual problem concerning his body image was injecting his scrotum repeatedly with olive oil to increase the size of his genitals. He developed respiratory failure following accidental intravenous injection of olive oil and was found to have lipogranulomatous lesions in the lung and the scrotum.

Comparative sequence analysis of VP4s from five Australian porcine rotaviruses: implication of an apparent new P type.

VP4s from five Australian porcine rotaviruses (CRW-8, BEN-307, BMI-1, MDR-13, and TFR-41) were sequenced and analyzed. Strains with distinct VP7s were demonstrated to have closely related VP4s similar to OSU. However, MDR-13 which manifested both G3 and G5 specificity was found to contained a unique VP4 which contained 2368 nucleotides, six bases longer than the longest VP4 ever reported.

The avian rotavirus Ty-1 Vp7 nucleotide and deduced amino acid sequences differ significantly from those of Ch-2 rotavirus.

The Vp7 gene of the avian strain Ty-1, which is classified as G7 serotype, was sequenced and the amino acid sequence deduced. The gene is 1065 nucleotides long with a long open reading frame of 987 nucleotides producing a protein 329 amino acids in length. The amino acid homology of the Ty-1 Vp7 protein to that of the avian Ch-2 Vp7 was 70%. The A, B, and C variable epitope regions of Ty-1 were unique compared to those of Ch-2 and other strains representing the 14G serotypes. The low 53% homology of the A and C regions of Ty-1 and Ch-2 would suggest that Ty-1 may be of a different serotype to the G7 reference strain Ch-2.

Isolation and characterization of a novel reassortant between avian Ty-1 and simian RRV rotaviruses.

A reassortant, TyRh, was isolated after coinfection of MA104 cells with avian Ty-1 and simian RRV rotaviruses. Hybridization and serological studies showed that the reassortant's 4th gene, which encodes Vp4, was derived from the simian RRV rotavirus parent, whereas the remaining 10 genes were derived from the avian Ty-1 rotavirus parent.

A variant serotype G3 rotavirus isolated from an unusually severe outbreak of diarrhoea in piglets.

About 80% of faecal samples from severe outbreak of porcine diarrhoea (scours) were positive for rotavirus. Rotavirus positive samples were analyzed for their antigenic properties and amino acid sequences of the glycoprotein genes. These viruses could not be assigned to any serotypes using serotyping monoclonal antibodies (MAbs) developed for porcine rotaviruses [Nagesha and Holmes: Journal of Medical Virology 35:206-211, 1991b]. When two such viruses were isolated in cell culture and analyzed by neutralization tests using hyperimmune sera they showed only one way antigenic relation with both human and porcine viruses belonging to serotype G3. In addition none of the serotyping MAbs neutralized these two virus isolates. There was no base variation between VP7 genes of faecal and cell culture isolates. Predicted amino acid sequences of the VP7 gene showed marked epitope variation from other porcine type G3 isolates with amino acid substitutions and an additional glycosylation site at residue 238. This antigenic variation seen in rotaviruses appears similar to that of influenza viruses undergoing antigenic drift.

Typing of human group A rotavirus with alkaline phosphatase-labeled oligonucleotide probes.

Rotavirus (RV) in stools of children less than 1 year of age with diarrhea in Bangkok in 1989 were serotyped by monoclonal enzyme immunoassay (MEIA). RNA extracted from these specimens was tested for hybridization with alkaline phosphatase (AP) and 32P-labeled oligonucleotides constructed from the nucleotide sequences of VP7 of human G types 1 (HuG1Ac), 2 (HuG2Ac), 3 (HuG3Ac), and 4 (HuG4Ac). Of 148 specimens that contained RV, 72% (106/148) hybridized with RV G type specific AP-labeled oligonucleotides compared to 47% (70/148) that were serotyped by MEIA (P less than 0.001). Of 68 specimens that contained only one VP7 serotype (G-type), as identified by MEIA, 94% (16/17) of G1, 90% (27/30) of G2, 57% (4/7) of G3, and 36% (5/14) of G4 RV hybridized with the AP-labeled HuG1Ac, HuG2Ac, HuG3Ac, and HuG4Ac oligonucleotides, respectively. The probes for G1, 2, 3, and 4 RV were specific for each G type. The results of hybridizing specimens with 32P- and AP-labeled oligonucleotides were similar. After transcription and amplification of cDNA of gene 9, AP-labeled RV G type specific oligonucleotides hybridized with 90% (134/148) of RV specimens. The high sensitivity of these nonimmunological techniques could be of value in identifying G types of RV during vaccine trials.

Molecular and serological analyses of two bovine rotaviruses (B-11 and B-60) causing calf scours in Australia.

Fecal specimens from 78 calves involved in outbreaks of calf diarrhea which occurred in three farms in Victoria, Australia, in 1988 were analyzed for rotaviruses. Thirty-eight samples were positive for group A virus antigen by enzyme-linked immunosorbent assay, and 20 of these contained viral double-stranded RNAs that could be detected by polyacrylamide gel electrophoresis. Two major electropherotypes could be observed, and a representative isolate of each electropherotype (isolates B-11 and B-60) was successfully adapted to grow in MA104 cells. Sequencing of the VP7 genes directly from RNA transcripts of fecal and cell culture-adapted viruses demonstrated that no base changes occurred in this gene upon adaptation to growth in MA104 cells. Sequencing also revealed that the VP7 protein of B-60 was closely related to G serotype 6 (G6) strains, whereas the B-11 sequence was significantly different from all previously published sequences except the recently reported VP7 sequences of bovine isolates 61A and B223, particularly across the antigenic regions A, B, and C. The other strains most closely related to B-11 by VP7 amino acid sequence analysis were G4 porcine strains BMI-1 and BEN-144 and G8 human strain 69M. Serotyping of B-11 and B-60 gave results that were in good agreement with the sequencing data. Hyperimmune typing sera clearly identified B-60 as a member of G6, whereas the B-11 strain reacted to moderate titers only with antisera to some G10 strains. Antiserum raised against B-11 neutralized some strains of G10 cross-reacted with porcine G4 type isolates BMI-1 and BEN-144 but not with other G4 strains or with rotaviruses of other mammalian G serotypes. Northern blot hybridization showed that B-11 was closely related to the recently reported bovine G10 strain B223, and they both possessed a similar segment 4 that was different from that of either UK bovine or NCDV rotavirus.

Direct serotyping of porcine rotaviruses using VP7-specific monoclonal antibodies by an enzyme immunoassay.

Employing a serotyping EIA test using MAbs both cell culture adapted and faecal porcine rotaviruses were classified into serotypes G3, G3/5, G4, and G5. The MAbs have confirmed and extended the serotyping results obtained using polyclonal antisera. These MAbs are therefore potential reagents for serotyping of porcine rotaviruses. Using subgroup specific MAbs serotypes G3, G3/5, and G5 were found to contain subgroup I antigens while G4 rotaviruses contained either subgroup II or subgroup I antigens.

VP4 relationships between porcine and other rotavirus serotypes.

VP4 relationship of Australian porcine rotaviruses were identified using genetic reassortants and MAbs. All porcine virus isolates except BEN-144 appeared to share VP4 antigenicity with OSU virus. VP4 and BEN-144 virus (Gottfried-like virus) showed some antigenic relationships with the human neonatal viruses ST-3 and RV-3. In addition, VP4 of porcine CRW-8 showed antigenic relationships with simian SA-11. RRV and also canine K9 viruses, while that of porcine TFR-41 showed at least one way VP4 antigenic relatedness with UK bovine rotavirus. Furthermore, BMI-1 virus which is antigenically similar to an American virus SB1-A (a naturally occurring reassortant) may have arisen similarly by gene reassortment in nature in Australia.

Expression of rotavirus VP7 antigens in fusions with bacterial proteins.

We have cloned antigenic regions of the VP7 gene from rotavirus RV-5 into the lamB gene. The insertions discussed in this paper comprise 250bp of rotavirus DNA and cover the A and B antigenic regions of the protein. The fusion proteins are expressed and are present in the outer membrane: they react with anti-rotavirus antibody. However, as yet we have not been able to demonstrate that the fusion has LamB protein functions with regard to maltodextrins or L phage, and the fusion protein appears to exist in the outer membrane as a monomer rather than as a trimer.