Fig 1.
Resistance profiling of ERDRP-0519 and GHP-88309 against MeV polymerase.
A) 2D-schematic of the MeV L protein. ERDRP-0519 (black) and GHP-88309 (green) resistance mutations are shown. Symbols denote the virus in which each resistance mutation was found (circle, MeV; triangle, CDV; square, HPIV-3; star, Sendai virus). B) 3D-homology model of the MeV L polymerase based on the structure of PIV-5 (PDBID: 6v85). The RdRP (cyan), capping (green), connector (yellow), methyltransferase (MTase; orange), and C-terminal (CTD; red) domains are shown. The GDNQ active site is highlighted by red spheres. RNA channels are shown as dotted black lines. C) Spatial organization of GHP-88309 and ERDRP-0519 resistance mutations. Shown are locations of tightly clustered GHP-88309 resistance mutations (green) and ERDRP-0519 resistance mutations (black) in the MeV L RdRP (cyan) and capping (green) domains. GDNQ is shown as dark red spheres. Predicted priming and intrusion loops [26] are shown. Proposed PRNTase and RdRP motifs are labeled with yellow squares and hexagons, respectively. The GDNQ active site is highlighted by red spheres. D-F) Characterization of resistance mutations in a cell-based MeV minigenome assay. Assessed were ERDRP-0519 induced mutations against ERDRP-0519 (D) and GHP-88309 (E), and GHP-88309 induced mutations against ERDRP-0519 (F). Results for the mutants in (F) against GHP-88309 are summarized in [36]. Symbols show sample means, curves represent 4-parameter variable slope regression models, EC50 values are presented in Table 1. The number of biological repeats (n) is specified for each construct.
Table 1.
The template PDB ID, virus, sequence identity, sequence similarity, range, coverage, GMQE score and QMEAN for the different homology models used are shown.
Table 2.
EC50 concentrations against MeV L of resistance mutations induced by viral adaptation to ERDRP-0519 and GHP-88309, respectively.
Binding kinetics (KD) between MeV L1708 or MeV L1708 harboring resistance mutations and ERDRP-0519 or GHP-88309 are shown.
Fig 2.
Target binding affinity of ERDRP-0519.
A) SDS-PAGE of purified full-length MeV L, MeV L1708 [36], and RSV L used for BLI studies. B-G) BLI of ERDRP-0519 and purified standard (WT) MeV L (B), MeV L1708 (C), RSV L (D) and MeV L1708 harboring selected ERDRP-0519 resistance mutations (E-G). Similar binding kinetics were observed for full-length MeV L and MeV L1708. KD values and goodness of fit are shown for each construct.
Fig 3.
ERDRP-0519 potently inhibits de novo RNA synthesis.
A-F) Purified recombinant WT MeV L-P complexes or complexes harboring the LT776A or LH589Y resistance mutations were incubated with either a 16-nt RNA template (A) and the represented NTPs to assess de novo RNA synthesis or a 16-nt RNA template (B), a 5’-phosphorylated 4-nt primer, and the represented NTPs to assess primer extension. Representative autoradiograms are shown for de novo initiation (C) and primer extension (D). Purified L-P complexes with an LN774A substitution in the catalytic GDNQ motif [70] served as control for specificity of the de novo initiation assay. Primer extension assays were performed in the absence of primer, template, or enzyme to control for contaminating de novo RNA synthesis driven directly by the template. Densitometry analysis was performed on elongation products of 15 to 16-nt in length (E) or 7 to 9-nt in lengths (F) and represent n = 3–4 biological repeats. EC50 values represent 4-parameter variable slope regression models, 95% confidence intervals are shown. Positions of unspecific background signals are indicated with *. G-I) Effect of ERDRP-0519 incubation on MeV primary transcription, represented in (G). Cells were infected with recMeV-Anc (MOI = 3) and incubated with vehicle (0.1% DMSO) volume control, or 0.2 μM or 0.8 μM ERDRP-0519. Infected cells were harvested 4 hours after infection. Incubation with ERDRP-0519 significantly decreased relative P-encoding mRNA amounts 4 hours post infection (H), but did not alter steepness of the MeV primary mRNA transcription gradient (I). Symbols represent individual biological repeats (n = 3), graphs show sample means. Statistical analysis through two-way ANOVA with Dunnett’s multiple comparison post-hoc test, P values are specified.
Fig 4.
Effects of ERDRP-0519 on single nucleotide addition.
A-C) Purified recombinant P-L complexes were incubated with the specified NTPs and a 25-nt RNA template driving both de novo initiation and back-priming [40] (A). A representative autoradiogram (n = 3) is shown, divided for clarity into dose-dependent inhibition of de novo RNA synthesis initiation at the promoter by ERDRP-0519 (B) and inhibition of 3’-elongation after back-priming (C). Products of less than 5-nt are not perceptible due to background from unincorporated 32P-labelled nucleotides. D) Dose-dependent inhibition of primer extension by ERDRP-0519 as in Fig 3B, but only GTP was added to visualize incorporation of the first nucleotide. EC50 values represent 4-parameter variable slope regression models, 95% confidence intervals are shown (n = 3).
Fig 5.
Photoaffinity labeling-based target mapping of ERDRP-0519.
A) Structure of photoactivatable compound ERDRP-0519az; the reactive azide moiety is highlighted (pink square). B) ERDRP-0519az is bioactive and displays no appreciable cytotoxicity. Symbols show means of three biological repeats, graphs represent 4-parameter variable slope regression models where possible. EC50 and CC50 are shown. C) 2D-schematic of the MeV L protein with locations of crosslinked peptides identified by photoaffinity labeling (top). The RdRP (cyan), capping (green), connector (yellow), MTase (orange), and C-terminal (light-red) domains, locations of known unstructured regions (LRI-LRII; LRII-LR-III), the GDNQ active site, specific amino acids motif in the capping domain, and positions of the intrusion and priming loops are shown. Sequences of peptides engaged by ERDRP-0519az (bottom). Specified are peptide location, spectrum match (PSM) score, number of ligands present, and delta mass. D) Location of peptides 1 (pink) and 2 (purple) in MeV L. E) Close-up top view of the capping and RdRP domains showing the adjacent locations of peptides 1 and 2. The priming loop (red), intrusion loop (blue), PRNTase motifs (blue spheres) and GDNQ active site (red spheres) are shown.
Fig 6.
In silico docking and ERDRP-0519 pharmacophore extraction.
A) Top view of the RdRP (cyan) and capping (green) domains of MeV L. PRNTase motifs, ERDRP-0519 resistance mutations, and ERDRP-519 are shown as blue, black, and magenta spheres, respectively. Peptide 2 (purple), the priming loop (red) and intrusion loop (blue) are labeled. The top scoring docking pose places ERDRP-0519 (orange) in close proximity to peptide 2. B) 2D-interaction projection of the top scoring ERDRP-0519 docking pose. The sulfonyl oxygen is predicted to interact with H1288 of the HR motif and the pyrazol ring hydrogen bonds with Y1155 of peptide 2. C-D) The binding pocket of ERDRP-0519 is not available in an MeV L homology model based on VSV L in initiation conformation (PDBID: 5A22) (C). Attempts to insert ERDRP-0519 result in multiple steric violations (moieties of the compound structure highlighted in red) (D). E-G) Close-up view (E), 2D-interaction map (F), and scaffold overlays (G) of the corresponding top-scoring covalent ERDRP-0519az docking pose. Positioning and main scaffold orientation resembles the pose of ERDRP-0519, although the sulfonyl group hydrogen bonds with D1378 instead of H1288. The lateral ring system containing the azide moiety of ERDRP-0519az (yellow sticks) must rotate (G) to fit.
Fig 7.
Independent ERDRP-0519 docking validation through ligand-driven 3D-QSAR modeling.
A) Development of the 3D-QSAR model. Individual data points of the training and test set, goodness of fit (R2), and slope of the best fit correlation through the origin are shown for the test set. B) Overlays of docking poses of the core ERDRP-0519 scaffold identified by the 3D-QSAR (red) and photocrosslinking-informed in silico docking (black). C) Graphical representation of the 3D-QSAR model showing space filling (grey) and pharmacophore features of the model. Proximity of features to H1288) and Y1155 is shown. D-F) Overlays of the space filling portion of the 3D-QSAR model (grey) with the L-provided binding pocket (blue). Individual views of the space-filling portion (D) and the binding pocket (E), and top and side views of the overlays (F) are shown.