The interaction of the respiratory syncytial virus (RSV) Matrix (M) protein with the plasma membrane was investigated using polyclonal and monoclonal antisera raised against recombinant M expressed in bacteria. M bound mainly to the plasma membrane, although a significant proportion bound to internal membranes. However, no localisation of M with the Golgi was observed, suggesting that transport of M to the plasma membrane was independent of the transport mechanism for the viral glycoproteins. Expression from a recombinant baculovirus demonstrated the ability of M to bind membranes in the absence of viral glycoprotein expression. When cell-surface expression of the viral glycoproteins was prevented using Brefeldin A, M was still found in association with the plasma membrane, but the characteristics of M's membrane-binding ability were different to that found in untreated infected cells. In the presence of normal glycoprotein expression, M was sorted into lipid rafts and, in addition, formed structures that could only be disrupted by treatment with high salt buffers, a feature suggesting an interaction with the cytoskeleton or the formation of strong intramolecular associations. Brefeldin A prevented M from being sorted into lipid rafts or from forming strong intramolecular associations. Brefeldin A also affected the stability of M bound to the plasma membrane, as M was more readily dissociated in the presence of the inhibitor. Coexpression of M and F resulted in the incorporation of M into lipid rafts but did not cause the formation of the strong intramolecular bonds, suggesting that additional factors are required for this phenomena.

Bacterially expressed nucleocapsid (N) protein, from respiratory syncytial virus (RSV), was used to investigate RNA binding in a modified North-Western blotting protocol. The recombinant protein demonstrated no sequence specificity in binding RNA representing either the antigenomic leader sequence or the nonspecific sequence derived from a plasmid vector. When recombinant N was purified on CsCl gradients, two types of structure, both with densities indicating that they contained RNA, could be visualised by negative-stain electron microscopy. Structures similar to nucleocapsids (NC) from RSV-infected cells were observed, as were ring structures. A small fragment of the N (amino acids 1-92) was all that was required for the production of NC-like structures. Another mutant with an internal deletion could form rings but not NC-like structures. This suggests that this domain (amino acids 121-160) may be important for maintaining helical stability. Further analysis has also identified a potential site in the amino-terminus that may be involved in an interaction with the phosphoprotein. A domain model of the RSV N protein is presented which, similar to that of other paramyxoviruses, supports the idea that the amino-terminus is important for NC assembly.

Iodination of intact Pasteurella haemolytica serotype A2 cells labelled a sub-set of total cellular proteins. Comparison of the autoradiographic patterns obtained from iodinated cells grown on complete medium and on iron-depleted medium showed that expression of three proteins, of 100,70 and 35 kDa, respectively, was increased by growth under iron-depleted conditions. Of these proteins, that of 35 kDa had not been reported previously. Like the 100 and 70 kDa proteins, the 35 kDa protein was expressed in natural infections, since it was recognized by antiserum from sheep that had recovered from an experimental infection with P. haemolytica A2. The 35 kDa protein was partially purified by reverse-phase HPLC and was found to be antigenic in both sheep and mice. A monoclonal antibody that was specific for the 35 kDa protein was used to identify the cellular location of the protein by immunoblotting of cell fractions enriched for particular cellular components. This demonstrated that the 35 kDa protein was located mainly in the periplasm.

To investigate structure and biological properties of the nucleocapsid (N) protein of respiratory syncytial virus (RSV), we have generated a panel of 16 monoclonal antibodies, raised against recombinant N protein, and epitope mapped seven of these to three antigenic sites (Site I aa 16-30; Site II aa 341-350; Site III aa 351-365). Characterization by immunofluorescence and by immunoprecipitation assay demonstrated that a monoclonal antibody to antigenic site I can detect N protein complexed with phospho (P) protein. Antibodies to antigenic sites II and III, which are adjacent to each other near the carboxyl terminus of the N protein, have distinct properties. A site III monoclonal antibody detected N protein in cytoplasmic inclusion bodies and in the cytosol, but not when N was complexed to P protein, while the site II antibody reacted with N protein in the nucleocapsid fraction but did not detect cytosolic N protein. Further investigation into the reactivities of the antibodies after binding of P to N in vitro demonstrated that antigenic sites II and III were blocked by the interaction, indicating an involvement for the carboxy domain of N in the N-P interaction. This was confirmed by the ability of peptides from the carboxy terminus of N to inhibit the N-P interaction in vitro.

No abstract