Fig 1.
Identification of M2-N31 targeted compounds from prototypic viroporin inhibitor classes.
A. Sequences of Eng195 M2 peptides (amino acids 18–60, corresponding to the “conductance domain” (CD)) used in study, comprising wild type (top) and N31S mutant (bottom). B. M2-N31/S31 containing Eng195 peptides were assessed for channel forming activity in DOPA/DOPC(DOPE) liposomes containing self-quenching concentrations of carboxyfluorescein. Activity relative to liposome-only, solvent and untreated M2 peptide controls (100%). At least two biological repeats were carried out, each with duplicate/triplicate technical repeats of each condition (***** p≤0.00001, paired student t-test). C. Cell culture antiviral effect of rimantadine and adamantane compound D upon Eng195 in a plaque reduction assay. Results are representative of three biological repeats, each containing two technical repeats (* p≤0.05, paired student t-test).
Fig 2.
Eng195 homology models based upon PDB: 2RLF conductance domain structure.
A. Overlapping structures of original 2RLF structure (green) and Eng195 (E195) homology model (light/dark grey corresponding to TM and C-terminal extension of CD, respectively). B. Docking of compound D into lumenal and peripheral binding site within Eng195 model. C. Alteration of Eng195 structure caused by occupation of position 31 by either serine (gold) or asparagine (pale blue). Left: top-down view showing lumenal orientation of Ser31 versus inter-helical Asn31. Middle and zoom: side view of helical displacement and location of Ser31 versus Asn31 sidechains. D. Docking of Rimantadine and compound D into lumenal (top) or peripheral (bottom) binding sites of Eng195 homology model with either Ser31 (gold) or Asn31 (pale blue).
Fig 3.
Analysis of adamantane M2WJ332 biological activity versus M2-N31.
A. Structure, sequence composition and binding site grids within M2-N31 TM (light grey) and CD (C-terminal extension in dark grey) peptides. B. PDB 2LY0 adapted from Wang et al., PNAS 2009 showing M2WJ332-bound M2 TM domains from an S31N-mutated H3N2 M2 channel complex (Udorn). C. Activity of M2WJ332 in liposome dye release assay versus M2-N31 TM and CD peptides. Data representative of multiple biological repeats, each containing two technical repeats (* p≤0.05, paired student t-test). D. Predicted docking poses for M2WJ332 at both lumenal and peripheral binding sites upon the Eng195 2RLF-based channel complex homology model, generated using e-Hits (see also S4A Fig for predicted interactions). E. Inhibition of M2-N31 CD peptide channels using M2WJ332 within interface lipid bilayers. Results from three independent experiments. Error bars represent standard error of the mean with p values determined using the student T-test (* p≤0.05, ** p≤0.01). F. As for E using M2-N31 TM peptides.
Fig 4.
Selection and characterisation of lumenally or peripherally targeted M2-N31 ligands.
A. Schematic of inhibitor assay pipeline: left—in silico compound selection via vHTS showing respective target grids for lumenal and peripheral binding sites (top-down view of M2-N31 tetramer), middle—validation of binding versus TM and CD Eng195 peptides, right–inhibition assays in Eng195 MDCK culture (Imagery designed using “Biorender” https://app.biorender.com). B. Titration of class exemplar compounds versus M2-N31 TM/CD peptides in liposome dye release assays: Lumenally targeted L1.1 showing equivalent activity versus both TM/CD; peripherally targeted DP9 showing preferential inhibition of CD peptides; DL7 showing preference for CD peptides but with activity versus TM at higher concentrations; P6.4 showing preference for TM peptides (* p≤0.05, ** p≤0.01, *** p≤0.001, paired student t-test).
Fig 5.
Activity of compounds showing specific M2-N31 activity versus Eng195 in vitro in MDCK plaque reduction assays.
A. Compounds present at 80 μM both during infection (MOI of 0.01 pfu/cell) and through a 24 hr incubation in producer cells prior to titration of secreted infectivity. Rimantadine (80 μM) negative and Zanamivir (20 μM) positive controls were included. Data are representative of at least three biological repeats containing duplicate technical repeats (* p≤0.05, paired student t-test). B. 8-point IC50 calculations across log2 concentration steps (triplicates for each point) for hit compounds L1.1, DL7 and DP9, alongside M2WJ332 in MDCK plaque reduction assays, shown with corresponding molecular structures.
Fig 6.
Evolution of Eng195 in culture with targeted compounds.
A. Schematic of serial supernatant passage of Eng195 with increasing inhibitor concentrations. Underlined Sample IDs (serial passage, “SP”1, 2, 3 etc) were stored at 4 ºC for at least 24 hr prior use. Double underlined were clarified, snap frozen and then thawed prior to use. Green text indicates that the M2 protein was sequenced (bulk direct sequencing). B. Direct sequencing of secreted vRNA throughout serial passages of M2WJ332-treated virus showing passage 5 and 14 bulk populations, as well as an example of plaque purified virus from day 5. Progressive enrichment of V27A is apparent and becomes absolute in the plaque-purified samples. C. Monitoring of viral suppression during the course of selection. Quantification of SP7 output titres after infection by equivalent doses of SP6 innoculae. D. SP14 titres following imposition of bottleneck at SP8 (* p≤0.05, ***** p≤0.005. ******* p≤0.000005, paired student t-test). E. Longitudinal quantitative whole genome deep sequencing analysis for virus treated with M2WJ332, L1.1 or DL7. Percentage prevalence of changes relative to controls are shown for SP5 and SP14. M2WJ332 selects both V27A and other minor variants within M2, yet no such changes occur in L1.1 treated cultures. Other minor variants are selected over the course of the experiment in HA and other ORFs for both compounds.
Table 1.
Synonymous (syn) and non-synonymous (non-syn/NS) changes in E195 whole genome sequencing over the course of fourteen passages of drug selection.
Virus RNA derived from supernatant passages five and fourteen was purified and sequenced, prior to the evolutionary bottleneck and the final passage, as well as a plaque-purified sample virus from passage five. Data reflects the number of discrete changes present within the sample rather than their prevalence within the population. Complete list of changes is provided in S3 Table.
Fig 7.
Antiviral effects of M2-N31 inhibitor combinations versus Eng195 MDCK culture.
Class exemplar compounds were used in combination regimens across matrices of increasing concentrations. Resultant antiviral effects were assessed, and MacSynergy used to determine potential synergistic/additive/antagonistic effects. A. Top-down views of 3D-relief plots are shown with coloured peaks/troughs representative of synergistic (i.e. >0) or B. antagonistic (i.e. <0) effects. C. Synergistic MacSynergy plot of M2-N31 inhibitor L1.1 used in combination with Zanamivir (NAi).