Expression of interferon gamma by recombinant rabies virus strongly attenuates virus pathogenicity and enhances safety and efficacy of post-exposure treatment vectors

Animal model experiments have shown interferon gamma (IFNgamma) expression correlates with survival from infection with attenuated rabies virus and reduction of neurological sequelae. Therefore, we hypothesized that rapid production of IFNgamma by the rabies virus itself would induce a more robust antiviral response than would occur naturally, thereby attenuating the virus and enhancing its safety and efficacy as pre- and post-exposure treatment for rabies. We reverse-engineered IFNgamma-expressing recombinant RABV: SPBNgamma, GASgamma and GASgammaGAS. Morbidity and mortality were monitored in mice infected intranasally with SPBNgamma or control virus to determine the degree of attenuation attributable to viral IFNgamma. Incorporation of IFNgamma into the rabies genome highly attenuated the virus. SPBNgamma has an LD50 more than 100 fold greater than SPBN(-). Subsequent experiments were designed to assess how viral IFNgamma expression could attenuate SPBNgamma, and determine the ability of virus-expressed IFNgamma to improve the safety and immunogenicity of existing vaccine vectors for use as pre- and post-exposure treatment (PET). In vitro and in vivo mouse experiments show that SPBNgamma infection enhances type I IFN expression, in the absence of increased viral replication. Furthermore, knockout mice lacking the type I interferon receptor (IFNAR-/-) rapidly die from SPBNgamma infection. Since SPBNgamma retains residual pathogenicity in adult mice, we used GASgamma and GASgammaGAS, constructed on attenuated backbones, for subsequent studies. Mortality and morbidity were monitored in suckling mice and models of pre- and post-exposure treatment. Behavioral testing was also used to measure neurological deficits in mice surviving PET. We demonstrate that GASgamma and GASgammaGAS are significantly attenuated in suckling mice compared to the highly attenuated GASGAS vaccine. GASgamma better protects mice from lethal DRV4 RABV infection in both pre- and post-exposure models compared to GASGAS. Finally, GASgammaGAS significantly reduces post-infection neurological sequelae, compared to its control vector, during PET of DRV4 infection. We conclude that expression of IFNgamma by a vaccine vector can further enhance its safety profile while increasing its efficacy as a traditional vaccine and PET. Furthermore, the mechanism of attenuation is due, at least in part, to increased induction of type I IFN.