The Ebola virus glycoprotein contributes to but is not sufficient for virulence in vivo

Among the Ebola viruses most species cause severe hemorrhagic fever in humans; however, Reston ebolavirus (REBOV) has not been associated with human disease despite numerous documented infections. While the molecular basis for this difference remains unclear, in vitro evidence has suggested a role for the glycoprotein (GP) as a major filovirus pathogenicity factor, but direct evidence for such a role in the context of virus infection has been notably lacking. In order to assess the role of GP in EBOV virulence, we have developed a novel reverse genetics system for REBOV, which we report here. Together with a previously published full-length clone for Zaire ebolavirus (ZEBOV), this provides a unique possibility to directly investigate the role of an entire filovirus protein in pathogenesis. To this end we have generated recombinant ZEBOV (rZEBOV) and REBOV (rREBOV), as well as chimeric viruses in which the glycoproteins from these two virus species have been exchanged (rZEBOV-RGP and rREBOV-ZGP). All of these viruses could be rescued and the chimeras replicated with kinetics similar to their parent virus in tissue culture, indicating that the exchange of GP in these chimeric viruses is well tolerated. However, in a mouse model of infection rZEBOV-RGP demonstrated markedly decreased lethality and prolonged time to death when compared to rZEBOV, confirming that GP does indeed contribute to the full expression of virulence by ZEBOV. In contrast, rREBOV-ZGP did not show any signs of virulence, and was in fact slightly attenuated compared to rREBOV, demonstrating that GP alone is not sufficient to confer a lethal phenotype or exacerbate disease in this model. Thus, while these findings provide direct evidence that GP contributes to filovirus virulence in vivo, they also clearly indicate that other factors are needed for the acquisition of full virulence.
Author Summary Most Ebola virus species cause severe hemorrhagic fevers with high case fatality rates. However, Reston ebolavirus (REBOV) seems to be apathogenic for humans. While the reason for this is unknown, several lines of in vitro research have indicated that the viral glycoprotein (GP) may play a critical role in determining pathogenicity, although until now there was no data to support such a role in the context of an in vivo infection. In order to address this we have generated a novel reverse genetics system to facilitate rescue of REBOV entirely from cDNA, which together with a previously established full-length clone system for the highly pathogenic Zaire ebolavirus (ZEBOV) allowed us to generate chimeras in which the glycoprotein genes from these two viruses have been exchanged (rZEBOV-RGP and rREBOV-ZGP). While the exchange of the viral glycoprotein did not affect virus growth in cell culture, we could show that infection with rZEBOV-RGP resulted in decreased virulence in a mouse model of infection. Further, rREBOV-ZGP did not show any signs of virulence in this model, similar to wild-type recombinant REBOV, showing that while GP contributes significantly to filovirus virulence, it is clearly not the sole determinant of pathogenicity.