Ebola remains a severe global public health threat and economic burden with limited therapeutic strategies. The single-stranded, negative-sense RNA virus is highly virulent, and clinical observations suggest that the high fatality rate of the virus is correlated with high viral load due to uncontrolled viral replication. Ebola viral protein 35 (VP35) contains a C-terminal interferon inhibitory domain (IID) that binds dsRNA, which forms as complementary viral RNA strands are synthesized. By binding viral dsRNA, VP35 IID suppresses RIG-I-mediated interferon production and host immune response, enabling viral replication. Promisingly, inhibition of the VP35 IID-dsRNA interaction makes Ebola avirulent. The Bowman lab previously identified a VP35 IID cryptic pocket that is allosterically coupled to its dsRNA binding and further identified mutations that control the cryptic pocket. When the cryptic pocket is open, dsRNA binding is reduced; when the cryptic pocket is closed, dsRNA binding is enhanced. Recently, we also characterized VP35 IID’s interaction with a disordered peptide derived from Ebola’s nucleoprotein, which has been shown to be critical for virion particle formation. We discovered that the cryptic pocket also controls the peptide’s binding to VP35 IID. Thus, targeting the cryptic pocket disrupts two critical, native interactions and represents a novel way to target traditionally undruggable proteins. Here, we apply the deep learning-based protein design tools RFdiffusion, ProteinMPNN, and ESMfold to design inhibitory peptides targeting the VP35 IID cryptic pocket. We subsequently use fluorescence polarization, mass spectrometry, and nuclear magnetic resonance techniques to characterize the binding affinity and site of the designed peptides, as well as their inhibitory capacity.
Bray et al. (Sun,) studied this question.