Constitutive phosphorylation of the vesicular stomatitis virus P protein modulates polymerase complex formation but is not essential for transcription or replication.

As a subunit of both the P-L polymerase complex and the P-N assembly complex, the vesicular stomatitis virus (VSV) P protein plays a pivotal role in transcription and replication of the viral genome. Constitutive phosphorylation of this protein is currently thought to be essential for formation of the P-L complex. We recently identified the three relevant phosphate acceptor sites in the VSV Indiana serotype P protein (R. L. Jackson, D. Spadafora, and J. Perrault, Virology 214:189-197, 1995). We now report the effects of substituting Ala at these acceptor sites on transcription reconstitution in vitro and replication of defective interfering virus (DI) templates in vivo. The singly substituted S60A, T62A, and S64A mutants and the doubly substituted S60A/T62A and T62A/S64A mutants, all of which retain some constitutive phosphorylation, were nearly as active as the wild type in both assays. Surprisingly, the nonphosphorylated S60A/S64A protein was also active in transcription (> or = 28%)) and replication (> or = 50%) under optimal conditions. However, this mutant was much less active in in vitro transcription (< or = 5% of wild type) at low P concentrations (<27 nM). In addition, S60A/S64A required higher concentrations of L protein than did the wild type for optimal DI replication in vivo. DI replication efficiency and intracellular accumulation of L, P, and N proteins in the transfected system were very similar to those in VSV-infected cells. We conclude that P protein constitutive phosphorylation is not essential for VSV RNA synthesis per se but likely plays an important role in vivo in facilitating P multimerization and possibly P-L complex formation.

Stabilization of vesicular stomatitis virus L polymerase protein by P protein binding: a small deletion in the C-terminal domain of L abrogates binding.

We showed previously that cells expressing the vesicular stomatitis virus (VSV) L polymerase gene via the vaccinia-T7 RNA polymerase system accumulated 2- to 5-fold more L protein when the P protein was coexpressed (Canter et al., 1993, Virology 194, 518-529). The results presented here provide an explanation for this phenomenon. Pulse-chase analysis revealed that L was unstable with a half-life of 3 to 6 hr if expressed in the absence of P protein, but was stable for at least 16 hr when coexpressed with a 10- to 15-fold molar excess of P. The P protein, in contrast, was stable under both conditions. Stabilization correlated with formation of a P:L polymerase complex evidenced both by coimmunoprecipitation and by glycerol gradient sedimentation analyses. A mutant L protein, lacking amino acids 1638 to 1673, was not stabilized by coexpression and showed no binding to P protein. Its anomalous sedimentation, however, suggested misfolding and/or aggregation as the cause for the failure to bind P. Transcription reconstitution in vitro, using extracts from cells expressing excess of P over L protein, strongly depended on coexpression of the proteins for optimal activity. We propose that the coexpression dependence for polymerase reconstitution documented here for VSV, as well as that reported previously for the Sendai paramyxovirus, reflects the protective effect of P protein on L protein stability.

Faithful and efficient in vitro reconstitution of vesicular stomatitis virus transcription using plasmid-encoded L and P proteins.

We demonstrate here that plasmid-expressed polymerase proteins of a negative-strand RNA virus can faithfully reconstitute all aspects of the transcription process carried out by virion cores in vitro. The assay is based on adding purified nucleocapsid templates of vesicular stomatitis virus to extracts of cells expressing L and P viral polymerase proteins via the vaccinia-T7 RNA polymerase recombinant virus. No significant differences were seen between the native virion core reaction and the optimally reconstituted system including ratio of transcripts produced, polyadenylation, net synthesis per template, amounts of polymerase proteins and template, and competence to initiate infection in vivo. Reconstitution was not dependent on cotranslation of P and L proteins in the same cell since nearly as much activity was obtained by mixing extracts expressing each protein individually. Cotransfection with P plasmid, however, stimulated L protein accumulation two- to fivefold relative to transfection with L alone. Surprisingly, deleting a small region in the C-terminal half of the L polymerase protein (amino acids 1638 to 1673) abolished transcription as well as stimulation by P coexpression. Since the polymerase domain of L presumably lies in the N-terminal half of the protein, these results suggest that the putative nucleotide-binding motif in the deleted segment may be involved in an accessory function essential for the transcription process.