Dokument

Summary:
Viruses of the genus Ebola virus cause severe fevers with unusually high case fatality rates, and as of today, no licensed antiviral drugs are available. The Ebola virus matrix protein VP40 plays a major role during budding of new viral particles and is also involved in the regulation of viral genome replication and transcription. VP40 exists in three different homo-oligomeric forms, namely dimers, octamers and polymeric filaments. Dimers are transported towards the plasma membrane where the interaction with lipids triggers the formation of VP40-filaments. Filaments represent a flexible row of dimers which enable budding of viral particles and line the inner layer of the new virions. Upon binding of cellular RNA, VP40 dimers turn into ring-shaped octamers that exert an inhibitory effect on viral RNA synthesis. For the present work, the resolution of the VP40 crystal structure was improved and a structure-guided drug design was employed with the aim to impair VP40’s essential homo-oligomerization. The study investigated VP40 of two Ebola virus strains Zaire (zVP40) and Sudan (sVP40). Residues L117 and W95 are so-called hot spot amino acids of the dimeric and octameric interface of sVP40 that were characterized by mutational analyses. As expected, both sVP40 wildtype (WT) and W95A formed dimers nearly exclusively whereas sVP40 L117A formed mainly monomers when expressed in E. coli. Surprisingly though, the structure of both sVP40∆43 mutant oligomers was similar to sVP40 WT. As a dimeric crystal packing was surprising for monomeric sVP40 L117A, crystallographic artefacts were considered which prompted the structural analysis of the mutants in solution. Using hydrogen-deuterium exchange mass spectrometry and thermal shift assays, it could be shown that both mutants exhibited increased fluidity and decreased stability in solution and it could be confirmed that sVP40 L117A was indeed a monomer. In cellulo, the ability of sVP40 L117A to form virus-like particles (VLPs) and inhibit viral genome replication and transcription was completely abolished, whereas sVP40 W95A exhibited a gain of function as this protein released more VP40-VLPs into the cellular supernatant and also showed a stronger inhibition of viral RNA synthesis. This data suggests that targeting homo-oligomerization is a promising strategy to impair VP40 functionality, but demands interdisciplinary methods, especially regarding structure determination. Based on a high-resolution crystal structure of sVP40∆43 WT, the structure of the C-terminus of sVP40 could be analyzed. The C-terminus contains the only two cysteines of the molecule which were oxidized and formed a disulfide bridge in the crystal. When VP40 was expressed in mammalian cells and released into the supernatant, the cysteines were also oxidized by post-translational modifications such as glutathionylation and nitrosylation. In vitro, VP40 could be reduced again by human thioredoxin. While the overall structure and oligomeric state of sVP40 was preserved, mutation of the cysteins resulted in altered phenotypes with regard to VP40’s ability to regulate viral RNA synthesis and to induce budding and particle formation. In an attempt to disrupt homo-oligomerization directly, interface-mimicking peptides were designed and tested in both functional cell culture and biochemical assays. While their binding to VP40 could be demonstrated, the peptides were unable to influence VP40’s functions or self-assembly. Further, the dimeric interface of VP40 should be targeted with small molecules. To this end, disulfide tethering was established as an alternative approach using a sVP40 variant with a cysteine residue near the dimeric interface (N67C). This method exploits the formation of a covalent disulfide bridge between the introduced cysteine at position 67 and thiol-containing fragments. Upon incubation under reducing conditions, only fragments with additional interactions to VP40 were bound favourably and could be detected via intact protein mass spectrometry, yielding several fragment hits. While no structural information of one of the sVP40 N67C-fragment complexes could be determined to assess binding mode and location, this strategy proved to be highly successful in identifying promising lead-like molecules. Using a library of 96 preselected fragments and crystal soaking, salicylic acid (SA) was identified as a crystallographic binder of VP40. The binding to VP40 could be confirmed in solution. As expected, the weak binding resulted in only minor effects on VP40’s function in RNA synthesis and budding. The characterization of residues of VP40 involved in the interaction with SA (L158 and R214) suggested that the binding pocket between the N- and the C-terminal domain is a highly vulnerable target site as mutation of these residues resulted in a decreased ability to regulate viral genome replication and transcription for sVP40 R214A as well as decreased budding for both sVP40 L158A and R214A. This prompted the testing of SA derivatives and the identification of four further crystallographic binders (3-amino-SA; 4-fluoro-SA, 4-fluoro-2-hydroxybenzamide and 5-amino-SA). Further structure-guided drug design led to the design, synthesis and testing of LL060, a compound that was also able to impair the formation of VP40-VLPs. In summary, a drug design process from scratch to target the function of Ebola virus VP40 is described. Here, we characterized and validated VP40 homo-oligomerization as a target and identified several lead-like molecules originating from a site-directed ligand discovery screening. The highly promising lead compound LL060, which was identified via crystal soaking represents the starting point for the development of a potent Ebola virus inhibitor.

Das Dokument ist im Internet frei zugänglich - Hinweise zu den Nutzungsrechten