Discovery of a Novel Non-Nucleoside Inhibitor of RNA-Dependent RNA Polymerase Against Dengue Virus

Abstract

Dengue virus (DENV) is an acute infectious pathogen worldwide, for which no effective therapeutics are available. The RNA-dependent RNA polymerase (RdRp) displays an important role during DENV replication and is therefore a promising target in the development of antiviral drugs. However, there are still no clinically approved RdRp inhibitors available. In this study, we identified a natural small molecule, 12β-hydroxydammar-3-one-20(S)-O-β-d-glucopyranoside (PN-1), using a surface plasmon resonance-based screening assay. Biochemical and structural analyses revealed that PN-1 selectively targets the RdRp of DENV NS5 protein by covalently modifying residues Glu255, Met387, Glu479, and Ala507. Mechanistic studies involving tryptophan scanning and hydrogen-deuterium exchange mass spectrometry revealed that PN-1 binding regulates RdRp conformational transitions. This allosteric mechanism leads to suppression of enzymatic activity and inhibition of DENV replication. Consequently, PN-1 exhibited potent antiviral activity across various cell lines and conferred significant protection in both ICR suckling and AG129 mouse models. Taken together, our results show that PN-1 functions as a novel non-nucleoside inhibitor to suppress DENV replication by targeting RdRp. These findings highlight PN-1 as a promising anti-DENV lead compound, while revealing conserved RdRp residues as actionable targets for rational drug design.

Keywords: AG129 mice; ICR suckling; RNA‐dependent RNA polymerase; dengue virus; target.

FIGURE 1

PN‐1 directly targets NS5 RdRp. (A) Schematic representation of SPR screening for compounds interaction with RdRp. (B) K _D values for the binding of fourteen compounds to RdRp. The black dashed line represents the K _D below 10⁻⁶. Red points represent compounds that may interact with RdRp. (C) Chemical structure of PN‐1. (D) The binding affinity of PN‐1 with RdRp determined by SPR assay. (E) The enzymatic activity of RdRp after treatment with PN‐1. (F) The binding kinetics of PN‐1 with RdRp were calculated by ITC analysis. (G–I) Thermal stability and proteolytic susceptibility of NS5 in the presence or absence of PN‐1 determined by CETSA and DARTS assays. (J) Chemical structure of Bio‐PN‐1. (K) Viral RNA copies of E and NS1 with or without Bio‐PN‐1 treatment by qRT‐PCR assay. (L) Co‐localization of Bio‐PN‐1 (Biotin, green) and NS5 (red) detected by immunofluorescence analysis (scale bars = 50 µm). BHK‐21 cells were infected with DENV‐2 for 45 h, then 10 µM Bio‐PN‐1 was added and the culture was continued for additional 3 h (48 h in total), followed by fixation for immunofluorescence experiments. (M, N) Pull down assay was performed using DENV‐2 infected cell lysates. ^*** p < 0.001 versus DENV‐2 model.

FIGURE 2

PN‐1 selectively targets RdRp. (A) Schematic of different functional domains of NS5 and subdomains of RdRp. (B) PN‐1 selectively interacted with RdRp domain. (C) Thermal stability of RdRp with or without PN‐1 treatment assessed by CETSA assay. (D, E) Resistance of RdRp to protease by DARTS assay at different concentrations of pronase or PN‐1. (F, G) The binding mode of PN‐1 with DENV‐2 RdRp using molecular docking. (H–K) Pull down assay was performed in RdRp transfected cell lysates or recombinant RdRp with or without PN‐1. (L) Co‐localization of PN‐1 (red) and RdRp (green) detected by immunofluorescence analysis (scale bars = 50 µm). (M) PN‐1 mainly interacted with palm and thumb domains of RdRp.

FIGURE 3

PN‐1 allosterically regulates conformation of RdRp and inhibits its enzyme activity. (A, B) LC‐MS/MS identified PN‐1 covalent binding to site E255 on the peptide EGGAMYADDTAGWDTR. (C, D) LC‐MS/MS revealed the binding site of PN‐1 at the residue A507 on peptide LAANAICSAVPSHWVPTSR. A and C: DMSO control; B and D: PN‐1 treated. Recombinant RdRp protein was incubated with DMSO (left panels) or PN‐1 (right panels) at 4°C for 24 h. Green fragment ion peaks correspond to b‐ions; orange fragment ion peaks represent y‐ions. (E) Pull down analysis was used to detect that E255, M387, E479, and A507 mutation attenuated the interaction between PN‐1 and RdRp. (F) The tryptophan fluorescence intensity of RdRp upon PN‐1 treatment. (G) PN‐1 induced the conformational changes of RdRp determined by HDX‐MS analysis. The peptides with lower levels of hydrogen–deuterium exchange after PN‐1 treatment have been marked in red.

FIGURE 4

PN‐1 inhibits DENV replication. (A) Inhibition rate of PN‐1 on BHK‐21 cells against DENV‐2 intracellular replication. After DENV‐2 infection and PN‐1 treatment, BHK‐21 cell viability was quantified using the CCK‐8 assay, and viral inhibition rate was calculated. (B) PN‐1 improved the CPE of BHK‐21 cells induced by DENV‐2. The cells were photographed under an IX 53 light microscope. (C, D) Viral progeny synthesis in DENV‐2‐infected BHK‐21 cells with or without PN‐1. (E‐G) Quantification of viral E and NS1 RNA copies and protein levels in DENV‐2‐infected BHK‐21 cells by qRT‐PCR and western blot analysis. (H, I) Quantification of viral E and NS1 RNA copies when PN‐1 was added at different points. PN‐1 was administered at specific timepoints post‐infection (1, 5, 9, and 13 hpi), and cellular RNA was harvested at 48 hpi for qRT‐PCR analysis. (J–L) The expression of E (green), NS1 (red), and dsRNA (orange) detected by immunofluorescence assay (scale bars = 50 µm). ^** p < 0.01, ^*** p < 0.001 versus DENV‐2 model.

FIGURE 5

PN‐1 exerts anti‐DENV activities in various DENV serotypes and cell lines. (A, B) Quantification of DENV‐1 NS1 and DENV‐3 NS5 RNA copies after PN‐1 treatment in BHK‐21 cells. (C–F) Effects of PN‐1 on DENV‐2 E RNA copies in Huh7, HepG2, 293T, and Vero cells. ^** p < 0.01, ^*** p < 0.001 versus DENV model.

FIGURE 6

Protective effects of PN‐1 against DENV‐2 infection in ICR suckling mice. (A) The scheme of DENV‐infected ICR suckling mice experiment: mice were intracranially and intraperitoneally inoculated with DENV‐2 and further intraperitoneally administered with PN‐1 for 6 consecutive days. (B) Daily body weight of infected or mock‐infected mice. (C) Clinical scores of infected or mock‐infected mice. (D) Survival rates after infection with DENV‐2 and treatment with PN‐1. (E) Representative images of H&E staining in brain cortex and hippocampus (scale bars = 400 µm). The black arrow represents the perivascular cell infiltration. (F, G) Quantification of viral E and NS1 RNA copies in DENV‐2 infected ICR mice brain. (H) Viral protein levels of E and NS1 in ICR mice brain determined by western blot analysis. (I) Representative images of IHC staining for E in brain cortex and hippocampus (scale bars = 400 µm). ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001 versus DENV model.

FIGURE 7

Protective effects of PN‐1 against DENV‐2 infection in AG129 mice. (A) The scheme of DENV‐infected AG129 mice experiment. (B) Daily body weight of infected or mock‐infected AG129 mice. (C) Survival rates of AG129 mice after infecting with DENV‐2 and treating with PN‐1. (D–F) Viral RNA copies detected by qRT‐PCR assay on days 3, 5, and 7 post‐inoculation. (G–I) Representative H&E staining images of liver, small intestine, and colon tissues (scale bars = 400/200 µm). ^### p < 0.001 versus control group, ^* p < 0.05, ^*** p < 0.001 versus DENV model.