Novel Chemical Ligands to Ebola Virus and Marburg Virus Nucleoproteins Identified by Combining Affinity Mass Spectrometry and Metabolomics Approaches

Abstract

The nucleoprotein (NP) of Ebola virus (EBOV) and Marburg virus (MARV) is an essential component of the viral ribonucleoprotein complex and significantly impacts replication and transcription of the viral RNA genome. Although NP is regarded as a promising antiviral druggable target, no chemical ligands have been reported to interact with EBOV NP or MARV NP. We identified two compounds from a traditional Chinese medicine Gancao (licorice root) that can bind both NPs by combining affinity mass spectrometry and metabolomics approaches. These two ligands, 18β-glycyrrhetinic acid and licochalcone A, were verified by defined compound mixture screens and further characterized with individual ligand binding assays. Accompanying biophysical analyses demonstrate that binding of 18β-glycyrrhetinic acid to EBOV NP significantly reduces protein thermal stability, induces formation of large NP oligomers, and disrupts the critical association of viral ssRNA with NP complexes whereas the compound showed no such activity on MARV NP. Our study has revealed the substantial potential of new analytical techniques in ligand discovery from natural herb resources. In addition, identification of a chemical ligand that influences the oligomeric state and RNA-binding function of EBOV NP sheds new light on antiviral drug development.

Figure 1. Identification of chemical ligands bound to Ebola virus and Marburg virus NPs from Chinese licorice.

(A) The workflow of combining affinity MS and metabolomics approaches for ligand discovery towards NPs. (B) OPLS-DA score plots of the crude extracts, the NP incubation samples and the control samples show clear separation of three groups in the data sets of EBOV NP (left) and MARV NP (right). (C) VIP and S/N plots of features detected in the protein incubation sample and control from the EBOV NP experiment (left) and MARV NP experiment (right). Each grey symbol represents a feature detected in four independent replicates (RSD of S/N ratios across replicates < 30%). Each feature also matches its accurate mass with the Chinese licorice compound database. Red symbols annotate two putative ligands bound to NPs (GC7 and GC13). Blue symbols annotate the rest of 11 constituents in Chinese licorice that do not bind NPs. All color-coded compounds are identified by HRMS and MSMS analysis according to reference standards.

Figure 2. Structural elucidation of GC7 (A) and GC13 (B) by MS and MSMS analysis.

The proposed fragmentation pathway is shown below mass spectra.

Figure 3. Validation of NP ligands in compound mixture screens.

(A) LC-HRMS chromatograms of the protein incubation sample and the control from the EBOV NP experiment (upper) and MARV NP experiment (lower). The NP protein was incubated with a mixture of 14 pure compounds known to be licorice constituents and then ligands were identified by affinity MS analysis. (B) S/N ratios of specific compounds from the mixture screen experiment. Note that GC2/GC12 and GC7/GC10 show identical S/N ratios as they are isomeric pairs that cannot be distinguished under the LC condition. GC7/GC10 and GC13 with S/N > 2 are considered ligands to both NPs.

Figure 4. Potential interaction of new ligands with the RNA-binding groove of Ebola virus NP.

(A) The crystal structure of EBOV NP_core containing the N-lobe (in blue) and C-lobe (in brown). The RNA-binding pocket surrounded by several residues is marked by a circle and enlarged in the side stereoview. (B) Primary sequence alignment of members of the filoviridae family. Residues in the RNA-binding groove are indicated by yellow arrows. Docking model of GC7 (C) and GC13 (D) interacting with the RNA-binding groove of EBOV NP.

Figure 5. K160 is engaged in ligand binding and shift of NP oligomeric state.

(A) Binding efficiency of GC7 or GC13 to a specific EBOV NP mutant normalized to ligand binding to wild-type NP (defined as 100%) measured by affinity MS. Error bars represent SD from independent experiments in triplicate. (B) SEC chromatograms of wild-type EBOV NP alone or incubated with a specific ligand. (C) SEC chromatograms of EBOV NP K160A mutant alone or incubated with a specific ligand.

Figure 6. GC7 disrupts ssRNA association with oligomeric EBOV NP complexes.

(A) Measurement of ssRNA binding to oligomeric EBOV NP 36–450 at increasing concentrations by fluorescence anisotropy. (B) Competitive binding between GC7 and fluorophore-labeled ssRNA to oligomeric EBOV NP 36–450. IC₅₀ is derived at the concentration of GC7 that inhibits ssRNA binding to NP oligomers by 50%. Error bars represent SD from independent experiments in triplicate.