3,4-Dicaffeoylquinic Acid from the Medicinal Plant Ilex kaushue Disrupts the Interaction Between the Five-Fold Axis of Enterovirus A-71 and the Heparan Sulfate Receptor

Abstract

While infections by enterovirus A71 (EV-A71) are generally self-limiting, they can occasionally lead to serious neurological complications and death. No licensed therapies against EV-A71 currently exist. Using anti-virus-induced cytopathic effect assays, 3,4-dicaffeoylquinic acid (3,4-DCQA) from Ilex kaushue extracts was found to exert significant anti-EV-A71 activity, with a broad inhibitory spectrum against different EV-A71 genotypes. Time-of-drug-addition assays revealed that 3,4-DCQA affects the initial phase (entry step) of EV-A71 infection by directly targeting viral particles and disrupting viral attachment to host cells. Using resistant virus selection experiments, we found that 3,4-DCQA targets the glutamic acid residue at position 98 (E98) and the proline residue at position 246 (P246) in the 5-fold axis located within the VP1 structural protein. Recombinant viruses harboring the two mutations were resistant to 3,4-DCQA-elicited inhibition of virus attachment and penetration into human rhabdomyosarcoma (RD) cells. Finally, we showed that 3,4-DCQA specifically inhibited the attachment of EV-A71 to the host receptor heparan sulfate (HS), but not to the scavenger receptor class B member 2 (SCARB2) and P-selectin glycoprotein ligand-1 (PSGL1). Molecular docking analysis confirmed that 3,4-DCQA targets the 5-fold axis to form a stable structure with the E98 and P246 residues through noncovalent and van der Waals interactions. The targeting of E98 and P246 by 3,4-DCQA was found to be specific; accordingly, HS binding of viruses carrying the K242A or K244A mutations in the 5-fold axis was successfully inhibited by 3,4-DCQA.The clinical utility of 3,4-DCQA in the prevention or treatment of EV-A71 infections warrants further scrutiny. IMPORTANCE The canyon region and the 5-fold axis of the EV-A71 viral particle located within the VP1 protein mediate the interaction of the virus with host surface receptors. The three most extensively investigated cellular receptors for EV-A71 include SCARB2, PSGL1, and cell surface heparan sulfate. In the current study, a RD cell-based anti-cytopathic effect assay was used to investigate the potential broad spectrum inhibitory activity of 3,4-DCQA against different EV-A71 strains. Mechanistically, we demonstrate that 3,4-DCQA disrupts the interaction between the 5-fold axis of EV-A71 and its heparan sulfate receptor; however, no effect was seen on the SCARB2 or PSGL1 receptors. Taken together, our findings show that this natural product may pave the way to novel anti-EV-A71 therapeutic strategies.

Keywords: 3; 4-dicaffeoylquinic acid; 5-fold axis; enterovirus-A71; heparan sulphate.

FIG 1

3,4-DCQA suppresses EV-A71-induced cytopathic effects in RD cells. (A) Chemical structures of 3,4-DCQA and rosmarinic acid. (B) Microscopy findings revealed that 3,4-DCQA prevented EV-A71-induced cytopathic effects in RD cells. (a) Mock: mock infection with DMSO; (b) Virus: infection with EV-A71 (multiplicity of infection [MOI] = 0.1) in DMSO; (c) 3,4-DCQA: mock infection and concomitant exposure to 3,4-DCQA (50 μM); (d) infection with EV-A71 (MOI = 0.1) in DMSO and concomitant exposure to 3,4-DCQA (50 μM).

FIG 2

Anti-EV-A71 activity of 3,4-DCQA assessed by time-of-addition assays in RD cells. (A) Schematic representation of 3,4-DCQA exposure at different time points during the viral life cycle. Purple lines denote the duration of incubation with 3,4-DCQA, whereas the red double arrow indicates the duration of viral infection (between −1 and 0 hours postinfection [hpi]). (B) Bar chart showing the extent of reduction in EV-A71 titers related to different time points of 3,4-DCQA exposure. All data were normalized to control conditions (virus-only group arbitrarily set to 100%). Data are presented as means ± standard deviations of four replicates from two independent experiments; comparisons were carried out using the Student’s t test (**, P < 0.01 and ***, P < 0.001). We also showed the results in mean log viral titer (PFU/mL) to ensure a significant viral titer load reduction.

FIG 3

(A) Centrifugal filtration inactivation assay. Data are presented as means ± standard deviations from three independent experiments, and two replicates for each assay were used. Viral titers were also expressed as PFU/mL. (B) Attachment inhibition assay and (C to E) penetration inhibition assay. (B to C) RNA was extracted and analyzed with RT-qPCR to detect viral attachment and viral penetration into cells. Data are presented as means ± standard deviations from three independent experiments. (D to E) Inhibition of penetration by 3,4-DCQA by confocal immunofluorescence microscopic analysis. Bar: 10 μm. Images of endocytosed VP1 inhibited by 3,4-DCQA were analyzed with IN Cell Investigator high-content analysis software. Viral penetration into the cells was reflected by cell-associated VP1 puncta (panel D and left part of panel E). Additionally, colocalization of VP1 and EEA1 (arrows) and their relative proportions were quantified (panel D and right part of panel E). (F) Heparan sulfate pulldown assay. The wild-type 5865/sin/000009 strain as well as mutants carrying the VP1-K242A and VP1-K244A variants were pretreated with DMSO or 3,4-DCQA before binding to heparan sulfate beads. Virus attachment was assessed with RT-qPCR. RNA copy numbers for each experiment were normalized to the wild-type dimethyl sulfoxide (DMSO) control (arbitrarily set to 1). Results are from three independent experiments. *, P < 0.05, **, P < 0.01, and ***, P < 0.001. ***, P < 0.001 ns for comparisons between DMSO and 3,4-DCQA (500 μM) treatment.

FIG 4

Selection and identification of 3,4-DCQA-resistant viruses. (A) Experimental flow chart for the selection and identification of 3,4-DCQA-resistant viruses. 3,4-DCQA- or DMSO-treated viruses were plaque-purified and Sanger-sequenced (three independent clones each). (B) Growth kinetics of wild-type and mutant EV-A71 TW/2231/1998 recombinant viruses. RD cells were infected (MOI = 1) with different recombinant viruses (wild-type, E98G mutant, P246A mutant, and double-mutant) for 1 h. Samples were harvested at the reported time points and virus titers were determined using plaque assays. Data are presented as means ± standard deviations from two independent experiments; duplicates were included in each experiment.

FIG 5

Attachment, penetration, and filtration inactivation assays of wild-type and E98G/P246A mutant viruses. (A) Attachment (left graph) and penetration (right graph) assays. Subsequently, RNA was extracted and analyzed with RT-qPCR. Data are presented as means ± standard deviations from three independent experiments. **, P < 0.01 and ***, P < 0.001 compared with control (DMSO). ^##, P < 0.01 compared with the DM group. (B) Filtration inactivation assays. Data are presented as means ± standard deviations from three independent experiments. (C) 3,4-DCQA was used in combination with either PR66 (left) or 3,5-DCQA (right).

FIG 6

Receptor pulldown assays. Heparan sulfate receptor pulldown of double- or single-mutant viruses (A). 3,4-DCQA, NF449 (50 μM), or DMSO were initially mixed on ice for 1 h with either the wild-type EV-A71 recombinant virus or viruses carrying E98G and/or P246A mutations; subsequently, heparan sulfate bead pulldown was performed. 3,4-DCQA-induced inhibition was monitored with Western blotting using anti-EV71 antibodies. (B to C) PSGL-1 and SCARB2 receptor pulldown assays. Results are from three independent and reproducible experiments (control: 10% of input).

FIG 7

Structural docking of 3,4-DCQA and the EV-A71 VP1 protein. The expected binding site of 3,4-DCQA in the 5-fold axis of VP1 (PBD ID: 3ZFF) was predicted using Discovery Studio software. Amino acid residues which play crucial roles in the binding process are highlighted with different colors.