Abstract:
1. Recombinant DNA techniques have been applied to the dsRNA genome of the bovine rotavirus Nebraska Calf Scours Diarrhoea virus (NCDV). The sequence of a full - length cloned copy of genomic segment 10 of NCDV has been determined using the Sanger dideoxynucleoside sequencing technique by subcloning cDNA into M13 vectors. 2. Genomic segment 10 codes for the non-structural protein NCVP5, a protein which appears to be involved in virus maturation (Estes et al., 1983). Determination of the nucleic acid sequence of the gene has enabled the amino acid sequence o f the bovine NCVPS protein to be inferred. Comparison of the inferred amino acid sequence with homologous sequences derived from other virus strains (Both et al., 1983c; Baybutt and McCrae, 1984; Okada et al., 1984; Ward et al., 1985) has enabled conserved regions of the molecule to be identified. A small region of the NCVPS protein has been identified (residues 131-161) which exhibits considerable variability between rotavirus strains. 3. A computer-based algorithm has been utilised to predict the folding pattern of NCDV gene 10 mRNA. This reveals a 'panhandle ' structure which differs from that proposed for the related gene of strain Wa rotavirus (Okada et al., 1984) but both molecules possess a common feature in that the initiation codon falls within a potentially-stable duplex formed with a portion of the 3 untranslated region. 4. A series of four site-directed deletion mutants of the cloned gene were constructed in order to investigate the functional significance of the three N-terminal hydrophobic regions of the NCVP5 protein. Two mutants were constructed using conveniently-located HindIII and BamHI restriction enzyme sites. The other two mutants were generated using M13 vectors and synthetic oligonucleotides. These modifications yielded genes coding for proteins in which portions of the first and second hydrophobic regions had been deleted. 5. DNA corresponding to the 'wild-type' coding region was inserted into an SP6 transcription vector to enable mRNA to be produced in vitro. This mRNA, when incubated in a reticulocyte lysate, directed the synthesis of a protein of the correct size (20 K). The addition of dog pancreatic microsomes to the reaction yielded a protein product (29 K) of a size consistent with the glycosylated ('wild-type') form of the NCVP5 protein. 6. The four variant forms of the NCVP5 gene were also inserted into SP6 transcription vectors and the protein products synthesised by the resulting mRNAs studied. All four mRNAs directed synthesis of variant protein products of the anticipated size. 7. The ability of the four variant proteins to become glycosylated and to associate with membranes was investigated. The topology of the proteins in the membrane was examined by digestion with proteolytic enzymes. Variant proteins altered in the first or second hydrophobic regions retained their ability to associate with membranes, suggesting that the third hydrophobic region, which was not altered, might play a role in membrane association. 8. A model for the disposition of NCVPS in the endoplasmic reticulum is proposed in which the first hydrophobic region is located within the lumen of the endoplasmic reticulum, the second hydrophobic region spans the membrane and the third hydrophobic sequence associates independently with the membrane from the cytoplasmic side leaving the C-terminus of the molecule exposed to the cytoplasm. The model proposed accounts for the experimental observations but is in conflict with current mechanisms proposed for the insertion of proteins into membranes. (Wickner and Lodish, 1985).