
==== Front
eBioMedicine
EBioMedicine
eBioMedicine
2352-3964
Elsevier

S2352-3964(24)00313-X
10.1016/j.ebiom.2024.105277
105277
Articles
Tanomastat exerts multi-targeted inhibitory effects on viral capsid dissociation and RNA replication in human enteroviruses
Lim Therese Yien May ai
Jaladanki Chaitanya K. bi
Wong Yi Hao c
Yogarajah Thinesshwary micthiy@nus.edu.sg
ad∗
Fan Hao fanh@bii.a-star.edu.sg
befg∗∗
Chu Justin Jang Hann miccjh@nus.edu.sg
acdh∗∗∗
a Laboratory of Molecular RNA Virology and Antiviral Strategies, Department of Microbiology and Immunology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
b Bioinformatics Institute, Agency for Science, Technology and Research (A∗STAR), 30 Biopolis Street, Matrix #07-01, 138671, Singapore
c NUSMed Biosafety Level 3 Core Facility, Yong Loo Lin School of Medicine, National University of Singapore, 14 Medical Drive, 117599, Singapore
d Infectious Disease Translational Research Programme, Yong Loo Lin School of Medicine, National University of Singapore, 117597, Singapore
e Synthetic Biology Translational Research Program, Yong Loo Lin School of Medicine, National University of Singapore, 10 Medical Drive, 117597, Singapore
f Department of Biochemistry, Yong Loo Lin School of Medicine, National University of Singapore, 8 Medical Drive, 117596, Singapore
g Duke-NUS Medical School, 8 College Rd, 169857, Singapore
h Institute of Molecular and Cell Biology, Agency for Science, Technology and Research (A∗STAR), 61 Biopolis Drive, Proteos #06-05, 138673, Singapore
∗ Corresponding author. Laboratory of Molecular RNA Virology and Antiviral Strategies, Department of Microbiology and Immunology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore. micthiy@nus.edu.sg
∗∗ Corresponding author. Bioinformatics Institute, Agency for Science, Technology and Research (A∗STAR), 30 Biopolis Street, Matrix #07-01, 138671, Singapore. fanh@bii.a-star.edu.sg
∗∗∗ Corresponding author. Laboratory of Molecular RNA Virology and Antiviral Strategies, Department of Microbiology and Immunology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore. miccjh@nus.edu.sg
i Both Authors contributed equally to the work.

02 9 2024
9 2024
02 9 2024
107 10527717 3 2024
23 7 2024
28 7 2024
© 2024 The Author(s)
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Summary

Background

Global cyclical outbreaks of human enterovirus infections has positioned human enterovirus A71 (EV-A71) as a neurotropic virus of clinical importance. However, there remains a scarcity of internationally approved antivirals and vaccines.

Methods

In pursuit of repurposing drugs for combating human enteroviruses, we employed a comprehensive pharmacophore- and molecular docking-based virtual screen targeting EV-A71 capsid protein VP1-4, 3C protease, and 3D polymerase proteins. Among 15 shortlisted ligand candidates, we dissected the inhibitory mechanism of Tanomastat in cell-based studies and evaluated its in vivo efficacy in an EV-A71-infected murine model.

Findings

We demonstrated that Tanomastat exerts dose-dependent inhibition on EV-A71 replication, with comparable efficacy profiles in enterovirus species A, B, C, and D in vitro. Time-course studies suggested that Tanomastat predominantly disrupts early process(es) of the EV-A71 replication cycle. Mechanistically, live virus particle tracking and docking predictions revealed that Tanomastat specifically impedes viral capsid dissociation, potentially via VP1 hydrophobic pocket binding. Bypassing its inhibition on entry stages, we utilized EV-A71 replication-competent, 3Dpol replication-defective, and bicistronic IRES reporter replicons to show that Tanomastat also inhibits viral RNA replication, but not viral IRES translation. We further showed that orally administered Tanomastat achieved 85% protective therapeutic effect and alleviated clinical symptoms in EV-A71-infected neonatal mice.

Interpretation

Our study establishes Tanomastat as a broad-spectrum anti-enterovirus candidate with promising pre-clinical efficacy, warranting further testing for potential therapeutic application.

Funding

10.13039/501100001459 MOE Tier 2 grants (MOE-T2EP30221-0005 , R571-000-068-592 , R571-000-076-515 , R571-000-074-733 ) and A∗STAR10.13039/501100012415 Biomedical Research Council (BMRC).

Keywords

Drug repurposing
Human enteroviruses
Pharmacophore
Tanomastat
VP1 hydrophobic pocket
==== Body
pmc Research in context

Evidence before this study

Human enterovirus infections contribute significantly to the global burden of childhood illnesses due to their high transmissibility within the immunocompromised paediatric population, especially in childcare and school settings. Human enterovirus A71 (EV-A71) has emerged as a neurotropic virus of clinical importance. Global outbreaks result in high annual hospitalization rates and can occasionally cause severe neuropathology and cardiopulmonary complications, with a portion of patients experiencing long-term neurological sequelae and developmental retardation. Current mitigation measures are confined to sentinel surveillance of enterovirus circulation patterns and promoting good hygiene practices. These are limited in their capacity to contain the spread of enterovirus infections. Due to the rapid shifts in seroprevalence of circulating enteroviruses, there remains an unmet need for the development of broad-spectrum antivirals against these clinically significant enteroviruses. Drug repurposing presents an ideal approach to identifying novel antivirals. Leveraging on their established safety profiles, this strategy addresses the prevalent safety concerns encountered by many antiviral candidates and accelerates the pre-clinical process. Here, we employed a pharmacophore- and molecular docking-based virtual screen targeting key EV-A71 viral proteins: capsid protein VP1-4, 3C protease, and 3D polymerase. Tanomastat was selected as the hit of interest for deeper exploration of its antiviral properties and intrigue mechanism. Tanomastat is an orally available matrix metalloproteinase inhibitor which has undergone human clinical trials for cancer therapy, with well-documented safety and pharmacokinetic profiles. Yet, there is no or limited knowledge on its antiviral properties.

Added value of this study

This study establishes Tanomastat as a potent broad-spectrum inhibitor of clinically significant human enteroviruses which belong to enterovirus species A, B, C, and D in vitro. Tanomastat was revealed to inhibit viral capsid dissociation via VP1 hydrophobic pocket binding and suppress viral RNA replication in a dual-staged manner during enterovirus replication. In neonatal mice challenged with a lethal dose of EV-A71, Tanomastat conferred protection by achieving prolonged survival and reduced clinical scores, with no significant side effects. Significant reductions in viral load and virus-induced pathology in muscle tissues were also observed.

Implications of all the available evidence

Tanomastat is a potential broad-spectrum anti-enterovirus therapeutic which has shown in vivo competency and possesses multi-targeting properties that effectively raises the barrier to antiviral resistance.

Introduction

Cyclical epidemics of hand, foot, and mouth disease (HFMD) caused by human enteroviruses has emerged as a global public health menace. In particular, human enterovirus A71 (EV-A71) is recognised as the most clinically significant causative agent. EV-A71 is also frequently associated with severe neurological and cardiopulmonary complications such as brainstem encephalitis and pulmonary edema, as well as fatalities in young children.1 Following its initial isolation in 1969 from the excrement of an encephalitis-stricken patient in California, United States, EV-A71 outbreaks were reported in several other countries within the next 20 years and throughout the 1900s.2,3 Today, EV-A71 has established its presence as an endemic virus in the Asia–Pacific region.3 Recurring outbreaks of HFMD in the area imposes huge healthcare and socio-economic burden on affected countries, including China, Japan, Malaysia, Singapore, and Taiwan.4, 5, 6, 7, 8, 9, 10

Enteroviruses are non-enveloped viruses which comprise of a single positive-sense RNA of approximately 7500 nucleotides in length encapsulated within a symmetrical icosahedral capsid that spans between 20 and 30 nm.11 The enterovirus replication cycle is initiated upon attachment of capsid proteins to host cell surface receptors such as human P-selectin glycoprotein ligand 1 (PSGL-1) and human scavenger receptor class B member 2 (SCARB2), followed by entry via receptor-mediated endocytosis.12,13 Virus uncoating, which is the release of viral genomic RNA from the capsid into the cytosol via a pore in the endosomal membrane, can be induced by receptor binding and/or pH alterations.14 The viral genomic RNA serves as a template for viral genome replication and viral polyprotein translation. A single open reading frame (ORF), flanked by the 5’ and 3’ untranslated regions (UTRs), encodes a single large precursor polyprotein of approximately 2100 amino acids which can be processed into four structural (VP1-VP4) and seven non-structural (2A-2C; 3A-3D) proteins by viral proteases, namely 2A proteinase (2Apro) and 3C protease (3Cpro).15,16 At the present, no antiviral therapy or internationally approved vaccine that confers protection against human enteroviruses is available in the market. While the search for potential antivirals is still ongoing, many of these drugs fail to exhibit effectiveness in vivo and in clinical trials. There is a pressing need for the discovery of novel broad-spectrum antivirals with high safety and efficacy profile to address the scarcity of available options against human enterovirus infections.17

Drug repurposing has been gaining traction in the search scape to expedite the discovery and development of therapeutics by utilizing existing drugs for new therapeutic indications.18, 19, 20 Our study aimed to repurpose existing drugs for the treatment of human enterovirus infections by targeting three key EV-A71 viral proteins: capsid protein VP1-4 (VP1-4), 3Cpro, and 3D polymerase (also known as RNA dependent RNA polymerase; 3Dpol). We employed pharmacophore- and molecular docking-based virtual screening approaches to identify potential candidates from the Drugbank database. Subsequently, we evaluated and characterized for the first time the broad-spectrum antiviral activity of Tanomastat against human enteroviruses in vitro. In the context of EV-A71 infection, we further elucidated its possible antiviral mechanism of action and demonstrated its inhibitory effects in murine model, which may serve as a valuable tool to support the advancement of antivirals.

Methods

Ethics

Animal care and housing were provided in accordance with the National Advisory Committee for Laboratory Animal Research (NACLAR) Guidelines (Guidelines on the Care and Use of Animals for Scientific Purposes). Mice experiments were designed and granted approval under protocol R19-01035 and R23-0865 by National University of Singapore Institutional Animal Care & Use Committee (IACUC) based on NACLAR Guidelines. This work was reported according to ARRIVE guidelines.

Protein modelling

The crystal structures of human EV-A71 capsid protein VP1-4 (PDB ID: 4CEY), 3Cpro (PDB ID: 3QZR) and 3Dpol (PDB ID: 7W9S) were obtained from the Protein Data Bank (PDB).21 These protein structures were prepared using Protein PrepWizard.22 This involved adding missing hydrogen atoms, fine-tuning bond orders and formal charges, and removing all the crystal water molecules. Furthermore, we optimized the hydrogen bond network by modifying side chains of specific amino acids including Cys, Ser, Asn, Glu, and His. To mimic physiological conditions, we employed the PROPKA23 program to generate ionization states of polar amino acids at neutral pH.

Ligand docking database generation

We collected 9712 compounds from the Drugbank database,24 including FDA-approved drugs, investigational compounds, and experimental drugs. To ensure data integrity, we systematically eliminated duplicate entries from our collection. Subsequently, the retained ligands were geometrically optimized with OPLS3e force field25 using Ligprep module.26 Chiral centres within the ligands were retained, yielding 32 low-energy conformers per compound. Employing Epik, we computed ionic states (phosphates) and tautomers of the ligands at pH 7·0 ± 2·0.

e-pharmacophore hypothesis generation and screening

e-Pharmacophore27 hypotheses were constructed using Glide28,29 and Phase30 modules, for the three crystal protein-ligand complex structures (PDB ID: 4CEY, 3QZR, 7W9S). First, the crystal ligands were redocked to their respective protein structures and scored “in place” using the Glide XP program, with the geometric centre of the ligand serving as the grid centroid. Second, ligands were further refined in torsional space within the binding pocket using the OPLS3e force field. Glide XP binding energy terms were then associated with each pharmacophoric feature, and the features with the best (lowest) binding energies were selected. Finally, receptor-based excluded volumes were incorporated into each e-Pharmacophore as steric constraints. The radius of each exclusion sphere was determined relative to the van der Waals radius of receptor atoms within a 5 Å distance range from the bound ligand. The 9712 Drugbank compounds were screened using Phase. This screening was performed with the following criterion: at least 3 out of 4 features of a given e-Pharmacophore should be matched. A distance matching tolerance of 2·0 Å was chosen to strike a balance between stringent and flexible alignment matching. The compounds selected from this screening process were ranked according to their fitness scores.

Molecular docking-based virtual screening

The molecular docking-based virtual screening was performed by the Glide module of Schrodinger suite. The Receptor Grid Generation Panel within Glide suite was used to set up receptor grid. A cubic grid box of side 15 Å and another cubic box of side 20 Å were used as the inner and outer box, at the centroid of ligand, respectively. The OPLS3e force field was employed to identify and rank the poses. For VP1-4 protein, we have observed a conserved water molecule interacting with F131, P193, and Y201, so we also employed docking with that conserved water molecule.

Cell lines, viruses, and media

The following cell lines were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Sigma–Aldrich), supplemented with 10% heat-inactivated fetal calf serum (HI-FCS; Capricorn Scientific) and 0.2% sodium hydrogen carbonate (NaHCO3), at 37 °C with 5% CO2: human rhabdomyosarcoma (RD) cells (CCL-136™, ATCC) and Vero cells (ATCC Cat# CCL-81, RRID:CVCL_0059). Mycoplasma testing has been performed on all cell lines used. The following viruses were propagated and titrated (Vero cells for CV-B5) in RD cells cultured in DMEM, supplemented with 2% HI-FCS and 0.2% NaHCO3, at 37 °C (33 °C for EV-D68) with 5% CO2: clinical isolate EV-A71 strain 5865/sin/000009 (EV-A71; GenBank accession number: AF316321.2), EV-A71 strain H (VR-1432™, ATCC; GenBank accession number: AY053402.1), EV-A71 genotype B5 (GenBank accession number: FJ461781.1), EV-A71 genotype C4 (GenBank accession number: JQ965759.1), Coxsackievirus A6 (CV-A6; GenBank accession number not available), Coxsackievirus A16 (CV-A16; GenBank accession number: U05876.1), Coxsackievirus B5 (CV–B5; GenBank accession number: JX843811.1), Echovirus 7 (Echo-7; GenBank accession number: AF465516.1), Coxsackievirus A24 (CV-A24; GenBank accession number: KF725085.1), and Enterovirus D68 (EV-D68; GenBank accession number KM851231).

Drug preparation

Tanomastat (Cayman Chemical) and Pleconaril (TargetMol) were reconstituted in 100% dimethyl sulfoxide (DMSO; Sigma Aldrich) to an initial 20 mM stock concentration and stored at −20 °C for in vitro experiments. Prior to experimental use, the compounds were further diluted in maintenance media to achieve the relevant working concentrations. 0.1% DMSO was used in conjunction with any treatment to ensure that any discernible antiviral effects were solely attributed to the compound(s), and not confounded by the solvent used.

Tanomastat (Cayman Chemical) was reconstituted in 100% DMSO to an initial 25 mg/mL stock concentration and stored at −20 °C for in vivo experiments in mice. Prior to administration via oral gavage, Tanomastat was further diluted in sterile saline to achieve 10 mg/kg or 30 mg/kg dosages. DMSO was used in conjunction with any treatment to ensure that any discernible antiviral effects were solely attributed to the compound, and not confounded by the solvent used.

Viral plaque assay

In 24-well plates, seeded cells were infected with 10-fold serially diluted virus suspension for 1 h at 37 °C with 5% CO2. After washing with phosphate-buffered saline (PBS; pH 7.4), enterovirus-infected cells were overlaid with DMEM, supplemented with 2% FCS and 0.5% agarose (Vivantis) and further incubated at 37 °C. The cells were fixed and stained with 4% paraformaldehyde (PFA) and 1% crystal violet solution (Sigma–Aldrich). Viral plaque formation was counted manually and used to derive the total infectious viral titre in plaque forming units (PFU) per mL.

Cell viability assay

In 96-well plates, seeded cells were treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 10 μM to 200 μM) for 12 h at 37 °C with 5% CO2. After washing with PBS, alamarBlue™ Cell Viability Reagent (Thermo Fisher Scientific) diluted 1:10 in maintenance media was added. The cells were further incubated for up to 4 h. The fluorescence intensity in each well was measured at an excitation and emission wavelength of 570 nm and 600 nm, respectively, using the Infinite™ 200 series microplate reader (Tecan). With the use of GraphPad Prism 9.0, the relative cell viability values at each drug concentration were plotted in a non-linear regression curve. The 50% cytotoxic concentration (CC50) was determined by interpolation.

Post-infection treatment assay

In 96- or 24-well plates, seeded cells were infected with virus at multiplicity of infection (M.O.I) of 1 for 1 h at 37 °C with 5% CO2. After washing with PBS, the infected cells were treated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (>80% cell viability; from 1 μM up to 50 μM) for 12 h. For the first set, the plate was frozen and thawed (−80 °C; 37 °C) and the total infectious viral titres were determined by viral plaque assay. For the second set, cell lysate was collected for verification of protein expression by Western blot. With the use of GraphPad Prism 9.0, the infectious viral titre values at each drug concentration were plotted in a non-linear regression curve. The 50% inhibitory concentration (IC50) was determined by interpolation.

Time-of-addition and time-of-removal assay

In 96-well plates, seeded RD cells were infected with EV-A71 at M.O.I of 1 for 1 h at 37 °C with 5% CO2. For time-of-addition assay, the infected cells were washed with PBS and incubated with maintenance medium. At 0, 2, 4, 6, 8, and 10 h.p.i, the cells were treated with 0.1% DMSO vehicle control or 25 μM Tanomastat. For time-of-removal assay, the infected cells were washed with PBS and treated with 0.1% DMSO vehicle control or 25 μM Tanomastat. At 0, 2, 4, 6, 8, and 10 h.p.i, the drugs were replaced with maintenance medium. The plates were frozen and thawed (−80 °C; 37 °C) at 12 h.p.i and the total infectious viral titres were determined by viral plaque assay.

Pre-infection treatment assay

In 96-well plates, seeded RD cells were treated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (from 0.1 μM to 50 μM) for 2 h at 37 °C with 5% CO2. After washing with PBS, the treated cells were infected with EV-A71 at M.O.I of 1 for another 1 h. The infected cells were washed with PBS and further incubated with maintenance medium for 12 h. The plate was frozen and thawed (−80 °C; 37 °C) at 12 h.p.i and the total infectious viral titres were determined by viral plaque assay.

Co-infection treatment assay

EV-A71 virus particles were incubated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (from 0.1 μM to 50 μM) for 1 h at 37 °C with 5% CO2. After centrifugation using a 100,000 molecular weight centrifugal filter (Sartorius) in accordance with the manufacturer’s instructions to remove any unbound drug molecules, the treated virus particles were washed with PBS and resuspended in maintenance medium. In 96-well plates, seeded RD cells were infected with treated virus particles at M.O.I of 1 at 37 °C with 5% CO2. The infected cells were washed with PBS and further incubated with maintenance medium for 12 h. The plate was frozen and thawed (−80 °C; 37 °C) at 12 h.p.i and the total infectious viral titres were determined by viral plaque assay.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Samples were assayed in a 25 μL reaction mixture, each consisting of the following: 250 ng sample RNA, 50 μM Oligo(dt)20 (Invitrogen), 10 mM dNTP mix (Invitrogen), M-MLV 5X reaction buffer (Promega), and 200U M-MLV reverse transcriptase (Promega) using MiniAmp™ Plus (Applied Biosystems) to generate complementary DNA (cDNA). Quantitative polymerase chain reaction (qPCR) was conducted using a SYBR green-based RT-PCR kit (Maxima) in accordance with the manufacturer’s instructions. Briefly, the samples were assayed in a 25 μL reaction mixture, each consisting of the following: Maxima SYBR Green/ROX qPCR Master Mix (2X), 0.3 μM each of forward and reverse primers (for negative and positive sense EV-A71 RNA, respectively), and sample cDNA. The forward and reverse primer sequences for detecting EV-A71 VP1 gene are 5’- ATTGTCACCATAAGCAGCCA-3’ and 5’-CCTCCGGCCCCTGAATGCGGCTAAT-3’, respectively. The forward and reverse primer sequences for detecting β-Actin gene are 5’-AGAGCTACGAGCTGCCTGAC-3’ and 5’-AGCACTGTGTTGGCGTACAG-3’, respectively. The reaction mixtures were run using the StepOnePlus™ Real-Time PCR System (Applied Biosystems) at 95 °C for 1 min, followed by 40 cycles of denaturation at 95 °C for 15 s, and annealing and extension at 60 °C for 1 min. After normalization of sample cycle threshold (Ct) values against those of the β-Actin gene, which were concurrently detected as the endogenous control, the double-delta Ct method was employed to derive the relative fold change in viral gene expression by comparing values of Tanomastat-against that of 0.1% DMSO vehicle control-treated samples.

Virus attachment assay

EV-A71 virus particles were incubated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (5 μM and 50 μM) for 1 h at 37 °C with 5% CO2. After centrifugation using a 100,000 molecular weight centrifugal filter in accordance with the manufacturer’s instructions to remove any unbound drug molecules, the treated virus particles were washed with PBS and resuspended in maintenance medium. The subsequent steps were performed on iced water and ice-cold reagents were used. In 24-well plates, seeded RD cells were pre-incubated on iced water for 30 min prior to infection with treated virus particles at M.O.I of 1 for 1 h to facilitate virus pre-adsorption only. The cells were washed with PBS. For the first set, total cellular RNA was isolated using the RNeasy Mini Kit (Qiagen) in accordance with the manufacturer’s instructions. The relative abundance of viral gene was detected by RT-qPCR. For the second set, the plate was frozen and thawed (−80 °C; 37 °C) and the total infectious viral titres were determined by viral plaque assay.

Virus internalization assay

The subsequent steps were performed on iced water and ice-cold reagents were use, unless specified otherwise. In 24-well plates, seeded RD cells were pre-incubated on iced water for 30 min prior to infection with EV-A71 at M.O.I of 1 for 1 h to facilitate virus pre-adsorption only. After washing with PBS, the infected cells were treated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (5 μM and 50 μM) for 1 h at 37 °C with 5% CO2 to facilitate virus entry. The cells were washed with PBS to remove any traces of FCS. Then, the cells were incubated with 0.5 mg/mL Pronase diluted in PBS to remove any non-internalized virus particles. After 2.5 min, Pronase activity was immediately neutralised by DMEM, supplemented with 10% HI-FCS. The cells were washed with PBS. For the first set, total cellular RNA was isolated using the RNeasy Mini Kit (Qiagen) in accordance with the manufacturer’s instructions. The relative abundance of viral gene was detected by RT-qPCR. For the second set, the plate was frozen and thawed (−80 °C; 37 °C) and the total infectious viral titres were determined by viral plaque assay.

Preparation of labelled EV-A71 virus particles

RD cells were infected with EV-A71 at M.O.I of 1 for 1 h at 37 °C with 5% CO2 and incubated with 5 μM SYTO 82 (Molecular Probes) in phenol red-free DMEM (Thermo Fisher Scientific). The infected flasks were frozen and thawed (−80 °C; 37 °C) at 12 h.p.i. After harvesting the virus supernatant by centrifugation at 2500 rpm, 4 °C for 10 min, the virus was purified with a 35–40% sucrose gradient by ultracentrifugation at 37,000 rpm, 4 °C for 16 h. Afterwards, purified SYTO 82 labelled EV-A71 was incubated with 1 mg/mL Cy5 (GE Healthcare) in 100 mM NaHCO3 buffer for 1 h at room temperature, with gentle rocking. Unbound dye was removed by buffer exchange into 50 mM Hepes buffer (pH 7·4, 145 mM NaCl) using a gel filtration column (GE Healthcare). The total infectious viral titres were determined by viral plaque assay.

Live-cell fluorescence microscopy

In 18-well chamber slides (Ibidi), seeded RD cells were pre-incubated on iced water for 30 min prior to infection with purified, labelled EV-A71 at M.O.I of 20 to facilitate virus pre-adsorption only. After washing with Hank’s Balanced Salt Solution (HBSS), the infected cells subjected to 50 μM Tanomastat treatment. 0.1% DMSO and pleconaril were used as vehicle and capsid inhibitor31 controls, respectively. Cell nuclei were stained with Hoechst 33,342 solution (Thermo Fisher Scientific). A series of images were recorded under a temperature-controlled inverted IX83 laser scanning confocal microscope (Olympus FLUOVIEW FV3000) at 37 °C with 5% CO2 for 60 min. All images were processed using Imaris V10.1 software based on Pearson’s and Maender’s coefficient calculation for colocalization of particles based on object and distance. Distance of less than or equal to 1 μm was set as dissociation between capsid and viral RNA.

Post-infection treatment assay (at 2 h timepoint after pre-adsorption)

The subsequent steps were performed on iced water and ice-cold reagents were use, unless specified otherwise. In 96-well plates, seeded RD cells were pre-incubated on iced water for 30 min prior to infection with treated virus particles at M.O.I of 1 for 1 h to facilitate virus pre-adsorption only. The infected cells were washed with PBS and incubated with maintenance medium at 37 °C with 5% CO2 to facilitate virus entry in the absence of the drug. At 2 h.p.i, the cells were treated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (from 1 μM to 50 μM). The plate was frozen and thawed (−80 °C; 37 °C) at 12 h.p.i and the total infectious viral titres were determined by viral plaque assay.

Entry bypass assay

EV-A71 RNA was purified using QIAamp Viral RNA Mini Kit (Qiagen) in accordance with the manufacturer’s instructions. In 96-well plates, seeded RD cells were transfected with 50 ng EV-A71 RNA complexed with DharmaFECT 1 transfection reagent (Thermo Fisher Scientific) for 4 h at 37 °C with 5% CO2. After washing with PBS, the transfected cells were treated with 0.1% DMSO vehicle control or non-cytotoxic Tanomastat concentrations (from 10 μM to 50 μM). The plate was frozen and thawed (−80 °C; 37 °C) at 12 h.p.i and the total infectious viral titres were determined by viral plaque assay.

SDS-PAGE and Western blot

Boiled cell lysate samples were separated on 10% acrylamide sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel at 100 V using the Mini-PROTEAN® Tetra cell electrophoresis system (Bio-Rad). Protein bands were transferred onto a methanol-activated polyvinylidene difluoride (PVDF) membrane using the Trans-Blot Turbo system (Bio-Rad). After blocking with 2% bovine serum albumin (BSA) (Sigma–Aldrich) dissolved in Tris-buffered saline-Tween 20, the membrane was incubated in the following primary antibodies: mouse anti-EV-A71 VP2 (Merck Millipore Cat#MAB979, RRID:AB_95300), rabbit anti-EV-A71 VP1 (Genetex Cat#GTX132339, RRID:AB_2886617), mouse anti-MMP-9 (Thermo Fisher Scientific Cat#MA5-15886, RRID:AB_11157246), and mouse anti-β-Actin (Sigma–Aldrich Cat#A2228, RRID:AB_476697). The membrane was then incubated in the corresponding secondary antibodies: goat anti-mouse IgG (H + L), HRP (Thermo Fisher Scientific Cat#31430, RRID:AB_228307) and goat anti-rabbit IgG (H + L), HRP (Thermo Fisher Scientific Cat#31460, RRID:AB_228341). The membrane was immersed in Immobilon Western Chemiluminescent HRP substrate (Merck Millipore), followed by chemiluminescent detection of protein bands using the C-DiGit Chemiluminescence Western Blot Scanner (LI-COR). Membrane-bound antibodies were dislodged with Restore PLUS stripping buffer (Thermo Fisher Scientific) where necessary. The protein band intensities were normalized against those of β-Actin, which was probed for as the endogenous control.

Nanoluciferase reporter assay

Purified EV-A71 replication-competent or -defective RNA replicons, which had been previously established, were utilized in this experiment. In 96-well white plates (Corning), seeded RD cells were transfected with 200 ng EV-A71 RNA replicon complexed with DharmaFECT-1 transfection reagent for 4 h at 37 °C with 5% CO2. After washing with PBS, the transfected cells were treated with non-cytotoxic Tanomastat concentrations (from 1 μM to 10 μM). 0.1% DMSO, cycloheximide (CHX) and guanidine hydrochloride (GuHCI) were used as vehicle, general translation inhibitor,32 and RNA replication-specific inhibitor33 controls, respectively. After 12 h, the Nano-Glo Kit (Promega) was used to detect luciferase activity in accordance with the manufacturer’s instructions using the GloMax 20/20 plate reader (Promega).

Bicistronic luciferase reporter assay

Purified bicistronic reporter construct, which had been previously established, were utilized in this experiment. In 96-well white plates (Corning), seeded RD cells were transfected with transfection mixture containing 200 ng bicistronic construct and jetPRIME transfection reagent (Polyplus) for 4 h at 37 °C with 5% CO2. After washing with PBS, the transfected cells were treated with non-cytotoxic Tanomastat concentrations (from 10 μM to 50 μM). 0.1% DMSO and apigenin were used as vehicle and EV-A71 IRES translation inhibitor34 controls, respectively. After 12 h, the Dual-Glo Luciferase Assay System (Promega) was used to detect luciferase activity in accordance with the manufacturer’s instructions using the GloMax 20/20 plate reader.

Small interfering RNA (siRNA) knockdown assay

Predesigned siGENOME SMARTpool MMP-9 siRNA and non-targeting siRNA pool (Dharmacon) were commercially purchased. The relevant siRNA concentrations (from 5 nM to 75 nM) and DharmaFECT 1 transfection reagent were separately diluted in minimum essential medium reduced serum (MEM-RS; Cytiva), before combining to form transfection complexes. In 24-well plates, seeded RD cells were transfected with siRNA-DharmaFECT 1 transfection complexes for 72 h at 37 °C with 5% CO2. Silenced cells were infected with EV-A71 at M.O.I of 1 for 1 h at 37 °C with 5% CO2. After washing with PBS, the infected cells were treated with non-cytotoxic Tanomastat concentrations (from 25 μM to 50 μM). For the first set, the plate was frozen and thawed (−80 °C; 37 °C) and the total infectious viral titres were determined by viral plaque assay. For the second set, cell lysate was collected for verification of protein expression by Western blot.

Toxicity and efficacy evaluation in vivo

The primary outcome measure was defined to be survival. A power analysis was conducted to determine the appropriate sample size required at 5% of significance level and 80% power. Given an effect size of −0.75, the sample size was calculated to be n ≥ 5 per group using the sample size calculation described by Charan et al. (2013).35 Secondary outcome measures assessed were body weight, clinical scoring, and infective viral load in brain and hind limb tissues.

For toxicity evaluation, DMSO, 10 mg/kg or 30 mg/kg Tanomastat prepared in saline was orally administered to 5-day-old suckling BALB/c mice daily for six days. For efficacy evaluation, DMSO, 10 mg/kg or 30 mg/kg Tanomastat prepared in saline was orally administered to 5-day-old suckling BALB/c mice 2 h pre- or post-infection with 2 × 107 PFU of EV-A71 by intraperitoneal (i.p.) injection. A second dose was administered at 24 h.p.i and further administered daily thereafter for 4 days. The survival, body weight, and clinical scoring of drug-treated mice were monitored daily for 15 days, with a scoring system in 5 categories. Activity: 0 normal, 1 lethargy/abnormal posture, 2 huddled/inactive, 3 unresponsive to stimuli/severe tremor/inability to right itself. Breathing: 0 normal, 1 rapid/shallow, 2 not applicable, 3 laboured, blue. Movement: 0 normal, 1 weakness, incoordination, 2 single limb dragging/paralysis, 3 multiple limbs dragging/paralysis. Body weight: 0 normal, 1 loss of 5% over 24h, 2 loss of more than 15% or up to 10% over 24h, 3 loss of more than 10% over 24h or 20% in total or body condition 2 or below. Dehydration: 0 normal skin tent, 1 skin tent present on dorsum, 2 moderate skin tent, 3 severe skin tent. A total of 6 or more points accumulated across all categories or any criteria in score 3 will be determined as a humane endpoint.

Quantification of viral load and histology of mouse tissues

EV-A71-infected mice were sacrificed on 5 days post-infection (d.p.i). For viral load titration studies, brain and hind limb tissues were harvested into CK14 homogenizing tubes (Bertin Corp). After homogenizing, the tissue samples were centrifugated at 10,000×g, 4 °C for 10 min to pellet tissue debris. The virus supernatant was collected, and the total infectious viral titres were determined by viral plaque assay. For histopathology studies, spinal cord and hind limb muscle tissues were fixed in 4% paraformaldehyde, decalcified by Decalcifier II solution (Leica), and embedded. Tissue pathology was evaluated by haematoxylin and eosin (H&E) staining and EV-A71 antigen was detected by immunohistochemistry staining (IHC) using mouse anti-EV-A71 VP2 antibody. Images were captured using the Leica Bond-Max system.

Statistics

GraphPad Prism 9.0 was utilized to conduct any statistical analyses. The significance of results obtained from in vitro drug treatment and mice survival experiments were derived using one-way analysis of variance (ANOVA) (or two-way ANOVA in the time-course live cell assay) with Dunnett’s post-test by comparing data of drug-treated (P < 0.05) against that of 0.1% DMSO vehicle control-treated samples, and Cox regression model, respectively. Assumptions underlying one-way or two-way ANOVA were assessed. Normality assumption was assessed using QQ plot, D'Agostino-Pearson omnibus test, Anderson-Darling test, Shapiro–Wilk test, and Kolmogorov–Smirnov test. Homogeneity of variance assumption was assessed using Brown–Forsythe test and Bartlett’s test. Proportional hazards assumption underlying Cox regression was assessed using the log minus log plot. All observations are independent of one another. The origin, start, and end times for survival analysis are 0 d.p.i., 0 d.p.i., and 14 d.p.i., respectively. The significance of results obtained from in vivo drug treatment experiments in mice were derived using Mantel–Cox test or Kruskal–Wallis test by comparing data of drug-treated (P < 0.05) against that of DMSO vehicle control-treated mice. Statistical significance is denoted by ∗ P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, along with the P-values (mean difference, 95% confidence interval (CI)) for in vitro drug treatment experiments or P-values (hazard ratio (HR), 95% CI) for mice survival experiments.

Role of funders

The funders played no part in the design, data collection, data analysis, interpretation, writing or decision to publish the results.

Results

Pharmacophore mapping

9712 compounds retrieved from Drugbank were screened with e-Pharmacophores generated from protein-ligand complex structures of EV-A71 VP1-4, 3Cpro, and 3Dpol proteins. This pharmacophore screening process selected 355, 254, and 305 compounds for VP1-4, 3Cpro, and 3Dpol, respectively. Molecular docking-based virtual screening assessed the binding affinities and interaction profiles of these compounds with the target proteins. We shortlisted five ligand candidates for each of the three target proteins based on docking scores and human inspection (Table 1).Table 1 The compounds were shortlisted after pharmacophore-based (using fitness scores) and molecular docking-based virtual screening (using docking scores).

	

All identified candidate compounds were validated at 10 μM concentration in post-infection treatment assays. The VP1-4 inhibitors showed reduction in infectious viral titres, with Tanomastat showing the highest inhibition of 1.3 log PFU/mL reduction (Fig. 1a). All 3Cpro inhibitors showed reduction in infectious viral titres, with Sesamolin showing the highest inhibition of 1.2 log PFU/mL reduction (Fig. 1b). All 3Dpol inhibitors showed reduction in infectious viral titres, with Domperidone showing the highest inhibition of 1.1 log PFU/mL reduction (Fig. 1c). Since Tanomastat led to the most prominent reduction in infectious viral titres, further studies focused on unravelling the inhibitory mechanistic action of Tanomastat in enterovirus infection.Fig. 1 The shortlisted compounds achieved significant inhibition against EV-A71 at 10 μM. RD cells were infected with EV-A71 at M.O.I of 1 and post-treated with 0.1% DMSO vehicle control or candidate compounds targeting (a) EV-A71 capsid VP1-4, (b) EV-A71 3C protease, and (c) EV-A71 3D polymerase proteins at 10 μM. The total infectious virus titres were determined by viral plaque assay. Each data point denotes the mean of triplicates, and the error bar denotes the standard deviation. One-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against 0.1% DMSO vehicle control. P-values, mean difference, and 95% CI are reported in Supplementary Table S1. Bar graphs represent mean ± standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

Tanomastat displays potent dose-dependent inhibition against Enteroviruses in vitro

The chemical structure of Tanomastat is illustrated in Fig. 2a. The cytotoxicity profile of Tanomastat was evaluated in RD cells, which are high permissibility to all enterovirus serotypes tested. The relative cell viability remained above the 80% cell viability threshold at up to 60 μM (Fig. 2b). Subsequently, dose-dependent inhibition studies were conducted using the concentrations that were non-cytotoxic (from 1 μM to 50 μM). These concentrations were carefully defined to ensure that any decrease in infectious viral titres observed was solely attributed to the antiviral effects of Tanomastat and not influenced by any detrimental effects on cell viability. Tanomastat treatment of EV-A71-infected RD cells induced significant dose-dependent inhibition of up to 3.7 log PFU/mL reduction in infectious viral titres compared to the 0.1% DMSO vehicle control (Fig. 2c). 50% cytotoxic concentration (CC50) and 50% inhibitory concentration (IC50) were calculated based on data obtained from cell viability and dose-dependent inhibition studies to be 81.39 μM and 18.12 μM, respectively, giving a selectivity index (SI) of 4.49 (Table 2). Besides, Tanomastat treatment of RD cells inoculated with other EV-A71 strains also led to significant dose-dependent reductions in infectious viral titres, with a maximum of 5.74 log PFU/mL inhibition of strain H (Fig. 2d), 2.44 log PFU/mL inhibition of genotype B5 (Fig. 2e), and 3.03 log PFU/mL inhibition of genotype C4 (Fig. 2f). The corresponding IC50 and SI were determined to be 23.78 μM and 3.42 for strain H, 11.44 μM and 7.11 for genotype B5, and 1.945 μM and 41.85 for genotype C4 (Table 2).Fig. 2 Tanomastat displays dose-dependent inhibition against a broad range of enteroviruses in vitro. (a) Chemical structure of Tanomastat. (b) RD cells were treated with Tanomastat at the relevant concentrations (from 10 μM to 200 μM) to determine the cytotoxicity profile. 0.1% DMSO was used as vehicle control. Cell viability is expressed as a percentage relative to 0.1% DMSO vehicle control. The dashed line represents the 80% relative cell viability threshold which is indicative of non-cytotoxicity. In post-infection treatment assays, RD cells were infected with (c) EV-A71 (d) EV-A71 (Strain H), (e) EV-A71 (Genotype B5), (f) EV-A71 (Genotype C4), (g) CV-A6, (h) CV-A16, (i) CV-B5, (j) ECHO-7, (k) CV-A24, and (l) EV-D68 at M.O.I of 1 and post-treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 1 μM to 50 μM). The total infectious virus titres were determined by viral plaque assay. Virus titres below the detectable limit are denoted as not detectable (n.d.). Each data point denotes the mean of triplicates, and the error bar denotes the standard deviation. One-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against 0.1% DMSO vehicle control. P-values, mean difference, and 95% CI are reported in Supplementary Table S2. Bar graphs represent mean ± standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

Table 2 CC50 (50% cytotoxic concentration), IC50 (50% inhibitory concentration) and SI (selectivity index).

Enterovirus	Species	Cell line	CC50 (μM)	IC50 (μM)	SI	
EV-A71 Strain 5865/sin/000009	Enterovirus A	RD	81.39	18.12	4.49	
EV-A71 Strain H	Enterovirus A	RD	81.39	23.78	3.42	
EV-A71 Genotype B5	Enterovirus A	RD	81.39	11.44	7.11	
EV-A71 Genotype C4	Enterovirus A	RD	81.39	1.945	41.85	
CV-A6	Enterovirus A	RD	81.39	14.58	5.58	
CV-A16	Enterovirus A	RD	81.39	4.285	18.99	
CV-B5	Enterovirus B	RD	81.39	9.270	8.78	
ECHO-7	Enterovirus B	RD	81.39	1.888	43.1	
CV-A24	Enterovirus C	RD	81.39	4.400	18.50	
EV-D68	Enterovirus D	RD	81.39	0.3843	211.79	

The broad-spectrum antiviral effects of Tanomastat against other related enteroviruses were also explored. At tested concentrations, infectious viral titres were consistently reduced in a dose-dependent manner, with a 3.15 log PFU/mL reduction of CV-A6 (Fig. 2g), 2.95 log PFU/mL reduction of CV-A16 (Fig. 2h), 3.25 log PFU/mL reduction of CV-B5 (Fig. 2i), 4.60 log PFU/mL reduction of ECHO-7 (Fig. 2j), 2.81 log PFU/mL reduction of CV-A24 (Fig. 2k), and 3.54 log PFU/mL reduction of EV-D68 (Fig. 2l) at the maximum Tanomastat concentration tested. The corresponding IC50 and SI were determined to be 14.58 μM and 5.58 for CV-A6, 4.285 μM and 18.99 for CV-A16, 9.270 μM and 8.78 for CV-B5, 1.888 μM and 43.1 for ECHO-7, 4.4 μM and 18.5 for CV-A24, and 0.3843 μM and 211.79 for EV-D68 (Table 2). Taken together, the results demonstrate the broad-spectrum antiviral activity of Tanomastat against different Enteroviruses.

Tanomastat inhibits an early stage(s) of the EV-A71 replication cycle

To assess the antiviral efficacy of Tanomastat in inhibiting enterovirus replication, EV-A71 was selected as the model for subsequent experiments. A series of time-course studies were employed in vitro. Time-of-addition and time-of-removal studies were carried out to identify the critical time window in the EV-A71 replication cycle at which Tanomastat exerts its antiviral effects. In the time-of-addition experiment, RD cells were treated at selected time points relative to EV-A71 inoculation. The time-of-removal experiment involved treatment at 0 h.p.i, followed by its removal at the same time points. Taken together, the intersection between both curves show that Tanomastat is most efficacious when administered within the initial 2 h of infection (Fig. 3a and b). This suggests that Tanomastat targets early EV-A71 replication processes occurring between 0 and 2 h.p.i.Fig. 3 Tanomastat targets viral internalization, an early stage during EV-A71 replication. In time-of-addition and removal studies, RD cells were infected with EV-A71 at M.O.I of 1 and treated with (a) 0.1% DMSO vehicle control or (b) 25 μM Tanomastat at 2-h intervals from 0 to 10 h time points. (c) In pre-infection treatment assay, RD cells were pre-treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 0.1 μM to 50 μM) and infected with EV-A71 at M.O.I of 1. (d) In co-infection treatment assay, EV-A71 virus particles were pre-incubated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 0.1 μM up to 50 μM), and then used to infect RD cells at M.O.I of 1. In a–d, the infectious virus titres were determined by viral plaque assay. (e and f) In virus attachment assay, EV-A71 virus particles were pre-incubated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (5 μM and 50 μM), and then used to infect RD cells at M.O.I of 1 at 4 °C. (g and h) In virus internalization assay, RD cells were infected with EV-A71 at M.O.I of 1 at 4 °C and treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (5 μM and 50 μM) at 37 °C. In e–h, the relative fold change in VP1 gene expression and infectious virus titres were determined by RT-qPCR and viral plaque assay, respectively. The relative fold change in VP1 gene expression was normalized against β-Actin. Each data point denotes the mean of triplicates, and the error bar denotes the standard deviation. One-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against 0.1% DMSO vehicle control. P-values, mean difference, and 95% CI are reported in Supplementary Table S3. Line and bar graphs represent mean ± standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

Pre-infection treatment assay was performed to investigate if Tanomastat binds to host cell surface receptors involved in virus engagement to inhibit viral entry. There were no notable differences in infectious viral titres compared to the 0.1% DMSO vehicle control at all concentrations tested (Fig. 3c). Co-treatment assay was performed to investigate if Tanomastat binds to viral surface proteins to inhibit downstream process(es) during EV-A71 replication. Tanomastat treatment resulted in significant reduction in infectious viral titres from 5 μM onwards compared to the 0.1% DMSO vehicle control (Fig. 3d). These suggest that Tanomastat may bind directly to the EV-A71 virus particles to affect entry processes.

To verify these findings, viral attachment assay was performed to determine whether Tanomastat interferes with viral binding to the associated host cell surface receptors. EV-A71 virus particles were incubated with Tanomastat for 1 h. After purification to remove any unbound drugs, RD cells were infected with Tanomastat-treated EV-A71 virus particles at 4 °C for 1 h to facilitate virus pre-adsorption, but not entry into host cells. Tanomastat treatment failed to reduce infectious viral titres (Fig. 3e) or produce significant fold change in relative VP1 gene expression after normalization against β-Actin (Fig. 3f) compared to the 0.1% DMSO vehicle control. This suggests that Tanomastat is unlikely to affect EV-A71 attachment to host cell surface receptors prior to viral entry. Next, virus internalization assay was performed to determine whether Tanomastat affects virus internalization into host cells. RD cells were infected with EV-A71 at 4 °C for 1 h to allow only virus pre-adsorption. Then, the infected cells were treated at 37 °C for 1 h to facilitate virus internalization. As shown, a significant dose-dependent reduction in infectious viral titres (Fig. 3g) and fold change in relative VP1 gene expression after normalization against β-Actin (Fig. 3h) were observed upon Tanomastat treatment compared to the 0.1% DMSO vehicle control. This suggests that Tanomastat possibly targets EV-A71 internalization into host cells.

Tanomastat acts by targeting viral capsid dissociation upon EV-A71 entry into host cells

To explore the mechanism of action of Tanomastat during EV-A71 internalization into host cells, a method previously published for imaging of poliovirus entry into live cells was adopted to allow monitoring of viral RNA release, with slight modifications.36 Dual-labelled EV-A71 virus particles were generated by a two-step process: (1) introduction of viral RNA label (SYTO 82) into virus culture, followed by (2) incorporation of viral capsid label (Cy5) into purified single-labelled virus particles. Live images of infected RD cells were captured over a time course of 60 min post-infection using confocal microscopy. Initiation of capsid-viral RNA dissociation was observed at 20 min post-infection, as indicated by the overlap between capsid (green) and viral RNA (red), forming a yellow particle (Fig. 4a). Complete dissociation and release of viral RNA was observed at 60 min post-infection. To investigate Tanomastat’s involvement in capsid-viral RNA dissociation, infected RD cells were treated and imaged over a similar time course of 60 min post-infection. Pleconaril, a picornavirus capsid protein inhibitor,31 was included as an experimental positive control. Fewer viral RNA particles were observed with Pleconaril and Tanomastat treatment at 20 min, 30-, and 60-min post-infection compared to the 0.1% DMSO vehicle control (Fig. 4b). Capsid-viral RNA dissociation is indicated by the close alignment of capsid (green) and viral RNA (red) (Fig. 4c). Pleconaril and Tanomastat treatment had significantly lower (p < 0·0001, Dunnet’s multiple comparison test) viral RNA mean counts compared to the 0.1% DMSO vehicle control (Fig. 4d). However, there were no significant (Pleconaril: P = 0.77 and Tanomastat: P = 0.53, Dunnet’s multiple comparison test) differences in viral capsid mean counts for all treatments (Fig. 4d). Following the viral RNA mean counts, significantly lower (p < 0.0001, Dunnet’s multiple comparison test) viral particle dissociation was observed upon Pleconaril and Tanomastat treatment compared to the 0.1% DMSO vehicle control (Fig. 4d). In another competitive analysis, 0.1% DMSO vehicle control treatment showed significantly higher (p < 0.0001, Dunnet’s multiple comparison test) viral RNA mean counts compared to viral capsid mean counts, whereas Pleconaril and Tanomastat treatment showed significantly lower (p < 0.0001, Dunnet’s multiple comparison test) viral RNA mean counts compared to viral capsid mean counts (Fig. 4e). The consistent viral capsid mean counts across all treatments suggests that there was no effect on virus entry into cells. Instead, significantly lower viral RNA mean counts compared to viral capsid mean counts upon Tanomastat treatment suggests that it limits viral capsid dissociation, leading to an inhibition on viral genome release and hence, viral replication.Fig. 4 Tanomastat limits capsid-viral RNA dissociation upon internalization of EV-A71 into RD cells. (a) Time-course live cell assay of capsid-viral RNA dissociation was observed for a prolonged period of 60 min post-infection at M.O.I of 20. Tracking of single viral capsid showed dissociation of the capsid (green) and release of viral RNA (red). Overlapping viral RNA and capsid (yellow) indicates the start of dissociation. The purple/white arrows over the time course tracks a single virus particle up to capsid-viral RNA dissociation. (b) Snapshot of time-course live cell assay of capsid and viral RNA intracellular at 20-, 30-, and 60-min post-infection at M.O.I of 20.10 μM Pleconaril served as capsid-viral RNA dissociation inhibitor positive control. Images were processed using Imaris 10·1 software indicated by capsid (green), viral RNA (red), and nucleus (blue). (c) Capsid-viral RNA dissociation is indicated by colocalization of capsid to viral RNA with a distance of less than or equal to 1 μm (white arrow). (d) Mean of viral RNA, capsid and viral particle dissociation counts for 0·1% DMSO, positive control, pleconaril, and 50 μM Tanomastat treatment upon infection at M.O.I of 20. (e) 0·1% DMSO, positive control, pleconaril, and 50 μM Tanomastat treatment upon infection at M.O.I of 20. Green arrow indicates lower or higher mean viral RNA to mean viral capsid counts. Two-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against 0.1% DMSO vehicle control. P-values, mean difference, and 95% CI are reported in Supplementary Table S4. Line graphs represent mean ± standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

Tanomastat was docked to EV-A71 capsid protein VP1-4 (PDB ID: 4CEY). The binding pocket of VP1 can be delineated into two distinct hydrophobic regions: the entrance region containing I113, Y201, and W203; and the deep inside region comprised of I24, F135, F137, F155, V179, V190, and V192. The Tanomastat molecule contains three aromatic rings (R1, R2, and R3) and an acidic functional group between R1 and R2 (Supplementary Fig. S1). The R1 aromatic ring of Tanomastat formed π-π stacking interactions with W203 and hydrophobic interactions with I113. The biaryl group (R2 and R3) occupied the deep inside hydrophobic region (Supplementary Fig. S1a). The aromatic ring R2 was intercalated between F135 and F155, forming edge-to-face π-π stacking interactions. The third aromatic ring R3 formed parallel-displaced π-π stacking interactions with F137 and F233. Interestingly, the carboxylic acid moiety of Tanomastat, although not directly engaged with any protein residues, is bound in the proximity of a conserved water molecule located within the binding site. This conserved water molecule was found to interact with Y201 sidechain, F131 and P193 backbones across various VP1-4 ligand-bound crystal structures. When docked to this binding site with that conserved water, the carboxylic acid group of Tanomastat formed hydrogen bond interactions both with the conserved water molecule and Y201 (Supplementary Fig. S1b), yielding a slightly better docking score of −11.56 kcal/mol than that from docking without this water molecule (Table 1). This observation suggests that the binding of the central hydrophilic region of Tanomastat in the hydrophobic pocket of VP1 could be reinforced through interactions with this conserved water molecule. Hence, it can be deduced that Tanomastat binds to VP1 leading to inhibition of viral capsid dissociation in vitro.

Tanomastat exhibits some inhibitory effects on EV-A71 RNA replication

To determine the ability of Tanomastat to retain its antiviral effects beyond its time period of maximal efficacy described in earlier observations (Fig. 3a and b), RD cells were infected with EV-A71 for 1 h at 4 °C. From 0 to 2 h.p.i, the infected cells were incubated with maintenance media at 37 °C, followed by Tanomastat treatment for another 10 h. A maximum of 1.7 log PFU/mL reduction in infectious viral titres were observed upon Tanomastat treatment compared to the 0.1% DMSO vehicle control (Fig. 5a), which is a 2 log PFU/mL difference from that observed in dose-dependent inhibition studies (Fig. 2c). The limited inhibition observed when Tanomastat was introduced after the 2 h time window is in agreement with the time-of-addition and time-of-removal studies which strongly suggested that the time period of maximal efficacy is between 0 and 2 h.p.i of the EV-A71 replication cycle. However, Tanomastat’s antiviral activities were not completely abolished even at lower drug concentrations. To verify the possibility that Tanomastat also inhibits late stage(s) of the EV-A71 replication cycle (viral RNA replication, protein translation, virion packaging and release) which occur after viral genome release into host cytosol, entry bypass assay was performed. The introduction of EV-A71 RNA directly into RD cells bypasses early stages (viral entry, internalization and uncoating). Subsequent Tanomastat treatment led to a dose-dependent reduction in infectious viral titres (Fig. 5b). These observations imply that Tanomastat’s inhibitory effects were maintained after viral uncoating, which allow it to also acts on later process(es) of EV-A71 replication.Fig. 5 Tanomastat targets EV-A71 RNA replication. (a) RD cells were infected with EV-A71 at M.O.I of 1 at 4 °C. After pre-adsorption, the infected cells were incubated for 2 h at 37 °C and post-treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 1 μM to 50 μM). (b) In entry-bypass assay, RD cells were transfected with EV-A71 RNA and treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 10 μM to 50 μM). In a and b, the infectious virus titres were determined by viral plaque assay. (c–e) RD cells were infected with EV-A71 at M.O.I of 1 and post-treated with 0.1% DMSO vehicle control or Tanomastat at the relevant concentrations (from 1 μM to 40 μM). Protein bands were separated by SDS-PAGE, followed by Western blot analysis using anti-EV-A71 VP2 monoclonal antibody, anti-EV-A71 VP1 polyclonal antibody, and anti-β-Actin monoclonal antibody. Band intensities below the detectable limit are denoted as not detectable (n.d.). The relative VP2 and VP1 band intensities were normalized against β-Actin. Results are representative of two independent experiments. (f and g) RD cells were transfected with EV-A71 3D polymerase replication competent or defective RNA replicons and treated with Tanomastat at the relevant concentrations (from 1 μM to 10 μM). 0.1% DMSO, CHX and GuHCl served as vehicle, general translation inhibitor and RNA replication-specific inhibitor controls, respectively. (h) RD cells were transfected with the EV-A71 bicistronic luciferase construct and treated with Tanomastat at the relevant concentrations (from 10 μM to 50 μM). 0.1% DMSO and apigenin served as vehicle and EV-A71 IRES translation inhibitor controls, respectively, Luminescence readings were used to derive the normalized F Luc/R Luc ratio, which is reflective of IRES activity. Each data point denotes the mean of triplicates, and the error bar denotes the standard deviation. One-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against 0.1% DMSO vehicle control. P-values, mean difference, and 95% CI are reported in Supplementary Table S5. Bar graphs represent mean ± standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

To evaluate the inhibitory effects of Tanomastat on the initial round of viral protein translation and/or successive exponential cycles of viral RNA replication and protein translation, genomic and proteomic assays were conducted. Upon Tanomastat treatment of EV-A71-infected RD cells, a dose-dependent reduction of EV-A71 viral protein expression was observed (Fig. 5c). Normalization against β-Actin revealed VP0, VP1, and VP2 protein expression to be reduced to under detection limit at 40 μM Tanomastat (Fig. 5d and e). Then, we exploited EV-A71 replication-competent or -defective constructs, which had been previously established,37 to examine Tanomastat’s effects on viral RNA replication and translation processes. In both replicons, structural protein genes are replaced with gene encoding for Nanoluciferase (NanoLuc). The replication-competent construct has an intact 3D region that encodes for 3Dpol, which is responsible for RNA replication. As such, it retains the ability to self-replicate and transfected cells will exhibit luciferase expression. Conversely, the replication-defective construct contains a 53-amino acid deletion from the C-terminus of the 3Dpol gene, which produces non-functional 3Dpol protein and effectively averts viral RNA replication. Transfected cells will express luciferase after IRES translation of original replicons only. Two experimental positive controls were included: cycloheximide (CHX), a general translation inhibitor32 and guanidine hydrochloride (GuHCl), an RNA replication-specific inhibitor.33 In cells transfected with the replication-competent replicon, a significant decrease in luminescence readings was observed upon Tanomastat, CHX and GuHCl treatment compared to the 0.1% DMSO vehicle control (Fig. 5f). In cells transfected with the replication-defective replicon, a significant decrease in luminescence readings was observed only upon CHX treatment compared to the 0.1% DMSO vehicle control (Fig. 5g). This suggests that Tanomastat is capable of inhibiting viral RNA replication but not translation.

To ascertain this finding, RD cells were transfected with a previously established bicistronic reporter construct.38 The construct consists of a human cytomegalovirus (CMV) promoter positioned upstream of a Renilla luciferase (R Luc) gene, followed by an EV-A71 internal ribosomal entry site (IRES) positioned upstream of a firefly luciferase (F Luc) gene. This configuration allows for cap-dependent translation of R Luc by the CMV promoter, while cap-independent translation of F Luc is facilitated by the EV-A71 IRES. The IRES activity can be accurately determined by interpretation of the ratio of F Luc to R Luc luminescence readings (F Luc/R Luc). Apigenin, an EV-A71 IRES inhibitor,34 was included as an experimental positive control. Tanomastat treatment did not result in any notable decrease in normalized F Luc/R Luc ratio, suggesting that Tanomastat has no effect on EV-A71 IRES translation (Fig. 5h).

After establishing the viral targets, we were keen to determine whether wild-type EV-A71 will acquire mutations that confer drug resistance. Repeated passaging of wild-type EV-A71 in the presence of increasing Tanomastat concentrations did not give rise to drug-resistant mutants (data not shown). In view of the characteristic matrix metalloproteinase (MMP) inhibitory properties of Tanomastat,39 we opted to investigate the effect of diminished MMP-9 expression on the drug’s antiviral effects. In fact, elevated MMP-9 expression or activity has been reported in mouse brain and patient cerebrospinal fluid (CSF) samples upon EV-A71 infection. In these cases, treatment with an MMP-9 inhibitor mitigated the virus-induced brain edema in mice.40 MMP-9 gene knockdown revealed no significant differences in EV-A71 viral protein expression and infectious viral titres compared to the non-targeting control (NTC)-silenced cells (Supplementary Fig. S2). The P-values, mean difference, and 95% CI are reported in Supplementary Table S6. This observation suggests that MMP-9 is unlikely to be involved in Tanomastat’s antiviral activities.

Tanomastat rescues mortality in EV-A71-infected suckling mice

We demonstrated the in vivo safety and efficacy of Tanomastat using our established 5-day-old BALB/c murine model of EV-A71 infection.41 To evaluate the therapeutic toxicity, 10 mg/kg or 30 mg/kg Tanomastat treatment was orally administered to neonatal mice daily for six days (Fig. 6a). Tanomastat achieved an excellent safety profile of 100% survival (Fig. 6b). The neonates were clinically scored based on physical symptoms of activity, breathing, movement, body weight, and dehydration. There were no significant clinical scores observed at both dosages tested (Fig. 6c and d), apart from inactivity, rapid breathing and skin tent observed in a subset of neonates in the 30 mg/kg Tanomastat treatment group (Supplementary Fig. S3). Prophylactic efficacy was evaluated by infecting 5-day-old BALB/c mice with 2 × 107 PFU of EV-A71 via intraperitoneal (i.p.) injection. In our treatment regime, 10 mg/kg or 30 mg/kg Tanomastat treatment was commenced 2 h prior to EV-A71 injection, followed by further administration daily for five days (Fig. 7a). Infected neonates treated with 10 mg/kg and 30 mg/kg Tanomastat achieved improved survival rates of 71% (n = 7, P = 0.013, Mantel–Cox test) and 85% (n = 7, P = 0.0043, Mantel–Cox test), respectively as compared to 14% in those treated with DMSO vehicle control (Fig. 7b). The hazard ratio (HR) measures the relative risk of survival in infected neonates treated with 10 mg/kg and 30 mg/kg Tanomastat as compared to those treated with DMSO vehicle control over the study period. The reportable HR may guide assessment of the survival benefit associated with Tanomastat treatment in our EV-A71-infected murine model. Infected neonates in the 10 mg/kg Tanomastat treatment group had a significantly higher risk of having shorter survival time than those in the DMSO vehicle control group (HR = 5.34, 95% CI 1.27 to 22.86). Infected neonates in the 30 mg/kg Tanomastat treatment group had a significantly higher risk of having shorter survival time than those in the DMSO vehicle control group (HR = 10.47, 95% CI 2.25 to 48.82) (Supplementary Table S7). Similarly, infected neonates were clinically scored based on physical symptoms of activity, breathing, movement, body weight, and dehydration. The clinical scores of infected neonates treated with 10 mg/kg and 30 mg/kg Tanomastat were substantially lower as compared to those treated with DMSO vehicle control (Fig. 7a and b), most of which developed observable clinical symptoms of inactivity, rapid breathing, hind limb paralysis, and loss of body weight (Supplementary Fig. S4). To examine the therapeutic impact on viral infection and pathology, 5 d.p.i brain and hind limb muscle tissues were harvested (Fig. 7a). A maximum of 1.3 log PFU/mL (n = 7, P = 0.0017, Kruskal–Wallis test) and 2.8 log PFU/mL (n = 7, P = 0.0012, Kruskal–Wallis test) reduction in EV-A71 viral titres was obtained in brain and hind limb muscle tissues, respectively compared to DMSO vehicle control (Fig. 7e and f). H&E staining revealed mild necrosis in Tanomastat-treated mice compared to severe necrosis observed in DMSO vehicle control (Fig. 7g). IHC staining muscle tissue showed low viral presence in Tanomastat-treated mice compared to extensive antigen positive in DMSO vehicle control (Fig. 7h).Fig. 6 Tanomastat does not display in vivo toxicity at tested drug doses. 5-day-old BALB/c neonatal mice were challenged with 10 mg/kg and 30 mg/kg Tanomastat via oral gavage daily for six days. DMSO was used as treatment vehicle control. The (a) percent survival, (b) body weight, and (c) clinical scores were recorded daily for up to 15 days. One-way ANOVA followed by Dunnett’s test was used to determine the statistical significance of the treatments when compared against DMSO vehicle control, with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.

Fig. 7 Tanomastat displays in vivo prophylactic efficacy and viral load inhibition at tested drug doses in sucking BALB/c mice challenged with lethal dose of EV-A71. (a) 5-day-old BALB/c neonatal mice were infected with EV-A71 at a dose of 2 × 107 per mice via i.p. At 0 d.p.i, a single dose of 10 mg/kg or 30 mg/kg Tanomastat was administered via oral gavage 2 h pre-infection. A second dose was administered 24 h post-infection, and subsequent doses were administered daily up to 120 h.p.i. DMSO was used as treatment vehicle control. The (b) percent survival, (c) body weight, and (d) clinical scorings were recorded for up to 14 d.p.i. EV-A71-infected mice were sacrificed on 5 d.p.i for quantification of viral titres in (e) brain and (f) hind limb muscle tissues, as well as for histopathology evaluation in spinal cord and hind limb muscle tissues. H&E staining showed severe necrosis in (g) DMSO control-treated mice, whereas mild necrosis was observed in 10 mg/kg and 30 mg/kg Tanomastat-treated mice. IHC staining muscle tissue using anti-EV-A71 VP2 monoclonal antibody showed extensive antigen positive in (h) DMSO control-treated mice, whilst similarly low antigen distribution was present in 10 mg/kg and 30 mg/kg Tanomastat-treated mice. Mantel–Cox test or Kruskal–Wallis test was used to determine the statistical significance of the treatments when compared against DMSO vehicle control. P-values, HR, and 95% CI are reported in Supplementary Table S7. Horizontal lines in scatter dot plots represent geometric mean ± geometric standard deviation with ∗ P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001. Data displayed are representative images of each group (n ≥ 7). Magnification for H&E and IHC staining are conducted at 10X.

To further assess the therapeutic efficacy, a second treatment regime involving the commencement of 30 mg/kg Tanomastat treatment 2 h after EV-A71 injection, followed by further administration daily for five days, was conducted (Supplementary Fig. S5a). Infected neonates treated with 30 mg/kg Tanomastat achieved improved survival rate of 80% (n = 5, P = 0.049, Mantel–Cox test) as compared to 16% in those treated with DMSO vehicle control (Supplementary Fig. S5b). Infected neonates in the 30 mg/kg Tanomastat treatment group had a significantly higher risk of having shorter survival time than those in the DMSO vehicle control group (HR = 4.80, 95% CI 0.97 to 23.80) (Supplementary Table S8). Notably lower clinical scores were observed in infected neonates treated with 30 mg/kg Tanomastat as compared to those treated with DMSO vehicle control (Supplementary Fig. S5c–i). EV-A71 viral titres were also reduced in brain and hind limb muscle tissues (Supplementary Fig. S5j and k). Importantly, 30 mg/kg Tanomastat treatment managed to suppress the EV-A71 viral titre in brain tissues by at least 2.4 log PFU/mL in 40% of infected neonates as compared to the mean EV-A71 viral titre in brain tissues of DMSO vehicle control (Supplementary Fig. S5j). These results demonstrate the in vivo efficacy of Tanomastat against EV-A71 infection with minimal adverse effects in our murine model.

Discussion

Despite the shift in seroprevalence from EV-A71 infections to CV-A6 and CV-A16 infections in recent years, EV-A71 remains a prevalent public health threat worldwide. The immunocompromised nature of the target paediatric population, combined with the potentially associated neurological complications, highlights an urgent need to establish safe and potent antiviral therapeutics against enterovirus infections. The successful repurposing of drugs for antiviral applications underscores the utility of computational approaches in streamlining the process of drug discovery. In the present study, integration of pharmacophore- and molecular docking-based virtual screening enabled the efficient identification of compounds with favourable binding characteristics to target EV-A71 proteins. The experimental validation of shortlisted compounds, particularly the notable activity of Tanomastat, confirms the effectiveness of this approach (Fig. 1).

A notable dose-dependent decrease in the EV-A71 infectious viral titres was observed in EV-A71-infected RD cells treated with non-cytotoxic concentrations of Tanomastat (Fig. 2). Moreover, comparable antiviral efficacies were observed in cells infected with other EV-A71 strains and related enteroviruses. This is evident of its safety and potency which makes it a suitable candidate for further drug development in vitro.42 Time-course studies clarified that the drug exerts its antiviral effects during the 0–2 h.p.i. time window of the EV-A71 replication cycle, which according to EV-A71 replication kinetics, coincides with viral attachment, internalization and uncoating stages (Fig. 3).43 Indeed, virus internalization assay revealed a reduction in EV-A71 relative viral gene expression and infectious viral titres in a dose-dependent manner. The intricate mechanistic involvement of Tanomastat during virus internalization into host cells was explored by live tracking of labelled EV-A71 virus particles during early infection, whereby Tanomastat was shown to impede viral capsid dissociation (Fig. 4). Molecular docking analysis offered insights into the binding mechanistic of Tanomastat to the VP1-4 protein of EV-A71. Tanomastat's distinct molecular features facilitate its interaction with different regions of the VP1 binding pocket, reminiscent to the known inhibitor NLD (PDB ID: 4CEY).44 In this docking pose, Tanomastat's aromatic groups (R1 and R2/3) bind similarly as NLD's pyridine ring and phenoxy ring in the entrance and deep inside regions, respectively. Notably, Tanomastat’s central hydrophilic carboxylic acid group can interact with a conserved water molecule within the hydrophobic pocket of VP1-4, facilitating its binding within the hydrophobic pocket (Supplementary Fig. S1). Typically, a “pocket factor” binds to this pocket on the surface of the “canyon” to confer stability to the virion. During viral entry, VP1 binding to specific host cell receptors (including PSGL-1 and SCARB2) promotes the displacement of the “pocket factor” located at the base of the “canyon”, causing irreversible structural changes to the viral capsid.45 In turn, the destabilized virion undergoes a two-step uncoating process to facilitate genome release into the cytosol.46 Small molecule inhibitors generally confer capsid stability by outcompeting the “pocket factor” to bind more strongly to the hydrophobic pocket, or by modifying the receptor-binding interface to reduce the receptor’s affinity for the virus such that the “pocket factor” cannot be dislodged.47 Hence, it would be worthwhile to investigate the specific interaction sites of VP1 to affirm Tanomastat’s mechanism of action against picornaviruses.

Phylogenetically, enterovirus serotypes are defined by genetic sequence divergence in the VP1 protein coding region.48,49 Members within the same serotype possess at least 75% nucleotide and 88% amino acid sequence homology, respectively. Conversely, those with below 70% nucleotide and 85% amino acid sequence homology are classified into separate serotypes.49,50 However, a recent study which analysed the protein structure and sequence data across 500 enteroviral sequences shed light on the presence of 28 highly conserved amino acid residues in the VP1 hydrophobic pocket structure, most of which are found within its interior. These residues are critical in maintaining virion structure stability and potentially interact with other capsid proteins.51 Previous studies showed that ICA135 displays broad-spectrum antiviral activities against human enteroviruses by targeting the interior of the VP1 hydrophobic pocket based on in silico docking models.52 However, the researchers did not explore the underlying mechanisms by which the interaction between ICA135 and the VP1 hydrophobic pocket may contribute to the inhibitory effects of the drug. In contrary, our study showed that Tanomastat impedes viral capsid dissociation upon entry into host cells. The structural and sequence similarities in the VP1 hydrophobic pocket coding region across various enteroviruses strengthens the notion that the observed broad-spectrum activity is likely a result of Tanomastat targeting a conserved region within the VP1 hydrophobic pocket, hence offering versatility against picornaviruses.

Interestingly, subsequent experiments found Tanomastat to retain some of its inhibitory effects after the 2 h time window. In particular, luciferase assays involving replication-competent and defective replicons with fused Nanoluciferase reporter revealed that Tanomastat inhibits viral RNA replication (Fig. 5). Together, these findings suggest that Tanomastat may target both early and late stages of the EV-A71 replication cycle. The multi-targeting activity allows it to provide a more comprehensive suppression of viral replication while significantly reducing the chances of the virus acquiring mutations that confers resistance to the drug’s mechanism of action.53 The development of resistant mutants is a common obstacle that has complicated the administration of antivirals in clinical settings.54 However, wild-type EV-A71 remained sensitive to Tanomastat after multiple passages and did not acquire genetic mutations that enabled it to evade the drug’s effects or reduce its susceptibility to the drug, which may be attributed to the multi-targeting capacity that Tanomastat possesses. One limitation of this study is that in vitro experiments were conducted in a cancer cell line, which may not be entirely representative of the drug’s therapeutic capacity. The implication of Tanomastat in inhibiting extracellular matrix degradation creates the possibility of off-target effects. Nonetheless, our findings provide valuable insights into potential of Tanomastat as a potent multi-targeting agent against clinically significant human enteroviruses in vitro.

The oversimplification of in vivo conditions in cell culture poses a challenge in accurately translating the findings to clinical relevance. While cell culture experiments provide valuable data for establishing basic efficacy of drugs in specific cell types, their limited scope may lead to over-reliance on statistical significance rather than on broader physiological implications. To bridge this gap, the therapeutic efficacy of Tanomastat was evaluated in our EV-A71-infected murine model, which offers a more complex and physiologically relevant environment. This will allow us to gauge the potential clinical importance and make more informed decisions regarding further development. Several enterovirus capsid binders have exhibited promising therapeutic efficacy in murine models55, 56, 57, 58 and some have advanced to clinical trials.47,59,60 However, most of these drugs failed to demonstrate sufficient efficacy and caused unwanted side effects when evaluated in Phase I and II clinical trials. Remarkably, Tanomastat treatment offered protection against the lethal EV-A71 challenge in neonatal mice, as evident from the improvement of survival rates and alleviation of symptoms (Fig. 7 and Supplementary Fig. S5). Despite Tanomastat treatment groups conferring improved survival rates in both treatment regimes, HR > 1 was derived when comparing 10 mg/kg and 30 mg/kg Tanomastat treatment groups to DMSO vehicle control treatment group. This could be attributed to the small sample size, which may lack the statistical power to detect small to moderate differences in HR (Supplementary Tables S7 and S8). Furthermore, differential censoring due to the death event not occurring within the study period in Tanomastat treatment groups can distort the HR, given that HR relies on the assumption that censored observations have the same probability of experiencing the event as those that remain in the study.61 While these findings provide insights into the treatment’s effects on primary and secondary outcome measures, the HR should be interpreted with caution. It should serve as a guide for sample size estimation and study design in future research to explore the drug’s therapeutic efficacy. In agreement with its in vitro properties, the drug induced a significant reduction in EV-A71 viral load and tissue damage at sites where the virus is known to cause severe manifestations. In particular, fatal EV-A71 encephalomyelitis often exhibits a stereotypical pattern of inflammation, most strikingly seen in the spinal cord, brainstem, hypothalamus, and cerebellar dentate nucleus.62 However, such appearances were substantially reduced upon our treatment regime. On top of its pharmacokinetic profile that has been established in phase I to III trials either as a monotherapy or in conjunction with other agents in cancer research, the favourable in vivo efficacy against EV-A71 infection and convenient oral route of administration makes it a promising candidate for subsequent development.39,63, 64, 65, 66 Nevertheless, interindividual variability in immune competence and body weight at experimental onset may affect disease progression in infected neonates. Accurate and cautious preclinical profiling must be done to define the feasibility of Tanomastat for clinical development against EV-A71 infection.

In conclusion, our work demonstrates that Tanomastat is a potent multi-targeting antiviral against enterovirus infection in vitro and confers protection in EV-A71-infected neonatal mice. We suggest that Tanomastat’s inhibitory effects on EV-A71 replication can be derived by its ability to hinder viral capsid dissociation via binding to the VP1 hydrophobic pocket and limit viral RNA replication. Considering its antiviral activities exhibited against a broad range of clinically relevant enteroviruses, Tanomastat demonstrates a remarkable aptitude for its development as a broad-spectrum anti-enterovirus agent for the treatment of HFMD. Overall, enteroviruses remain a growing public health threat and Tanomastat may serve as a valuable tool to support the advancement of broad-spectrum antiviral therapeutics.

Contributors

TY, HF, and JJHC contributed to the conception of Tanomastat antiviral mechanism against enteroviruses. TYML, TY, and JJHC contributed to the design of all in vitro and in vivo experiments. TYML, and TY performed all in vitro and in vivo experiments. CKJ and HF contributed to the design of all computational experiments. CKJ performed all the computational experiments. TYML, YHW, and TY contribute to the harvesting and processing of mice tissues for immunohistochemistry. TYML, CKJ, TY, HF, and JJHC contributed to the writing of the manuscript. TY, HF and JJHC have verified the underlying data. All the authors read and approved the final manuscript.

Data sharing statement

The data supporting the conclusions of this study are available within the article. Raw data files will be made available by inquiries to the corresponding authors T.Y. (micthiy@nus.edu.sg), H.F. (fanh@bii.a-star.edu.sg), and J.J.H.C. (miccjh@nus.edu.sg).

Declaration of interests

HF received grants from Synthetic Biology Translational Research Program, School of Medicine, National University of Singapore.

Appendix ASupplementary data

Supplementary material

Supplemental western blots_Figure 5c

Supplemental western blots_Figure S2d

Supplemental western blots_Figure S2a

Acknowledgements

This study is supported by the following 10.13039/501100001459 MOE grants: R571-000-068-592 , R571-000-076-515 , R571-000-074-733 . HF and JCK were supported by funding from the 10.13039/501100012415 Biomedical Research Council of A∗STAR in Singapore.

Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.ebiom.2024.105277.
==== Refs
References

1 Solomon T. Lewthwaite P. Perera D. Cardosa M.J. McMinn P. Ooi M.H. Virology, epidemiology, pathogenesis, and control of enterovirus 71 Lancet Infect Dis 10 2010 778 790 10.1016/S1473-3099(10)70194-8 20961813
2 Schmidt N.J. Lennette E.H. Ho H.H. An apparently new enterovirus isolated from patients with disease of the central nervous system J Infect Dis 129 1974 304 309 10.1093/infdis/129.3.304 4361245
3 Puenpa J. Wanlapakorn N. Vongpunsawad S. Poovorawan Y. The history of enterovirus A71 outbreaks and molecular epidemiology in the asia-Pacific region J Biomed Sci 26 2019 75 10.1186/s12929-019-0573-2 31627753
4 Yang B. Liu F. Liao Q. Epidemiology of hand, foot and mouth disease in China, 2008 to 2015 prior to the introduction of EV-A71 vaccine Euro Surveill 22 2017 16-00824 10.2807/1560-7917.ES.2017.22.50.16-00824
5 Hagiwara A. Tagaya I. Yoneyama T. Epidemic of hand, foot and mouth disease associated with enterovirus 71 infection Intervirology 9 1978 60 63 10.1159/000148922 202573
6 Chan L.G. Parashar U.D. Lye M.S. Deaths of children during an outbreak of hand, foot, and mouth disease in sarawak, Malaysia: clinical and pathological characteristics of the disease. For the Outbreak Study Group Clin Infect Dis 31 2000 678 683 10.1086/314032 11017815
7 McMinn P.C. An overview of the evolution of enterovirus 71 and its clinical and public health significance FEMS Microbiol Rev 26 2002 91 107 10.1111/j.1574-6976.2002.tb00601.x 12007645
8 Chan K.P. Goh K.T. Chong C.Y. Teo E.S. Lau G. Ling A.E. Epidemic hand, foot and mouth disease caused by human enterovirus 71, Singapore Emerg Infect Dis 9 2003 78 85 10.3201/eid0901.020112 12533285
9 Wu Y. Yeo A. Phoon M.C. The largest outbreak of hand; foot and mouth disease in Singapore in 2008: the role of enterovirus 71 and coxsackievirus A strains Int J Infect Dis 14 2010 e1076 e1081 10.1016/j.ijid.2010.07.006 20952237
10 Ho M. Chen E.R. Hsu K.H. An epidemic of enterovirus 71 infection in Taiwan. Taiwan enterovirus epidemic working group N Engl J Med 341 1999 929 935 10.1056/NEJM199909233411301 10498487
11 Yuan J. Shen L. Wu J. Enterovirus A71 proteins: structure and function Front Microbiol 9 2018 286 10.3389/fmicb.2018.00286 29515559
12 Nishimura Y. Shimojima M. Tano Y. Miyamura T. Wakita T. Shimizu H. Human P-selectin glycoprotein ligand-1 is a functional receptor for enterovirus 71 Nat Med 15 2009 794 797 10.1038/nm.1961 19543284
13 Yamayoshi S. Yamashita Y. Li J. Scavenger receptor B2 is a cellular receptor for enterovirus 71 Nat Med 15 2009 798 801 10.1038/nm.1992 19543282
14 Baggen J. Thibaut H.J. Strating J. van Kuppeveld F.J.M. The life cycle of non-polio enteroviruses and how to target it Nat Rev Microbiol 16 2018 368 381 10.1038/s41579-018-0005-4 29626210
15 Yang C.H. Li H.C. Jiang J.G. Enterovirus type 71 2A protease functions as a transcriptional activator in yeast J Biomed Sci 17 2010 65 10.1186/1423-0127-17-65 20682079
16 Shih S.R. Chiang C. Chen T.C. Mutations at KFRDI and VGK domains of enterovirus 71 3C protease affect its RNA binding and proteolytic activities J Biomed Sci 11 2004 239 248 10.1007/BF02256567 14966374
17 Wang J. Hu Y. Zheng M. Enterovirus A71 antivirals: past, present, and future Acta Pharm Sin B 12 2022 1542 1566 10.1016/j.apsb.2021.08.017 35847514
18 Mercorelli B. Palu G. Loregian A. Drug repurposing for viral infectious diseases: how far are we? Trends Microbiol 26 2018 865 876 10.1016/j.tim.2018.04.004 29759926
19 Pushpakom S. Iorio F. Eyers P.A. Drug repurposing: progress, challenges and recommendations Nat Rev Drug Discov 18 2019 41 58 30310233
20 Parvathaneni V. Kulkarni N.S. Muth A. Gupta V. Drug repurposing: a promising tool to accelerate the drug discovery process Drug Discov Today 24 2019 2076 2085 31238113
21 RCSB protein Data Bank Available from: https://www.rcsb.org
22 Schrödinger release 2020-3: protein preparation wizard 2020 Epik, Schrödinger, LLC New York, NY Impact, Schrödinger, LLC, New York, NY; Prime, Schrödinger, LLC, New York, NY, 2020. Available from: https://www.schrodinger.com/life-science/learn/white-papers/protein-preparation-wizard/
23 Bas D.C. Rogers D.M. Jensen J.H. Very fast prediction and rationalization of pKa values for protein-ligand complexes Proteins: Struct, Funct, Bioinf 73 2008 765 783 10.1002/prot.22102
24 DrugBank Available from: https://go.drugbank.com
25 Roos K. Wu C. Damm W. OPLS3e: Extending force field coverage for drug-like small molecules J Chem Theory Comput 15 2019 1863 1874 30768902
26 Schrödinger release 2020-3 2020 LigPrep, Schrödinger, LLC New York, NY Available from: https://www.schrodinger.com/platform/products/ligprep/
27 Salam N.K. Nuti R. Sherman W. Novel method for generating structure-based pharmacophores using energetic analysis J Chem Inf Model 49 2009 2356 2368 10.1021/ci900212v 19761201
28 Schrödinger release 2020-3: Glide 2020 Schrödinger, LLC New York, NY Available from: https://www.schrodinger.com/platform/products/glide/
29 Friesner R.A. Banks J.L. Murphy R.B. Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy J Med Chem 47 2004 1739 1749 15027865
30 Schrödinger release 2020-3: phase 2020 Schrödinger, LLC New York, NY Available from: https://www.schrodinger.com/platform/products/phase/
31 Pevear D.C. Tull T.M. Seipel M.E. Groarke J.M. Activity of pleconaril against enteroviruses Antimicrob Agents Chemother 43 1999 2109 2115 10.1128/AAC.43.9.2109 10471549
32 Schneider-Poetsch T. Ju J. Eyler D.E. Inhibition of eukaryotic translation elongation by cycloheximide and lactimidomycin Nat Chem Biol 6 2010 209 217 10.1038/nchembio.304 20118940
33 Murray K.E. Nibert M.L. Guanidine hydrochloride inhibits mammalian orthoreovirus growth by reversibly blocking the synthesis of double-stranded RNA J Virol 81 2007 4572 4584 10.1128/JVI.02106-06 17301147
34 Lv X. Qiu M. Chen D. Zheng N. Jin Y. Wu Z. Apigenin inhibits enterovirus 71 replication through suppressing viral IRES activity and modulating cellular JNK pathway Antiviral Res 109 2014 30 41 10.1016/j.antiviral.2014.06.004 24971492
35 Charan J. Kantharia N.D. How to calculate sample size in animal studies? J Pharmacol Pharmacother 4 2013 303 306 10.4103/0976-500X.119726 24250214
36 Brandenburg B. Lee L.Y. Lakadamyali M. Rust M.J. Zhuang X. Hogle J.M. Imaging poliovirus entry in live cells PLoS Biol 5 2007 e183 10.1371/journal.pbio.0050183
37 Wu K.X. Chu J.J. Antiviral screen identifies EV71 inhibitors and reveals camptothecin-target, DNA topoisomerase 1 as a novel EV71 host factor Antiviral Res 143 2017 122 133 10.1016/j.antiviral.2017.04.008 28427827
38 Gunaseelan S. Wong K.Z. Min N. Prunin suppresses viral IRES activity and is a potential candidate for treating enterovirus A71 infection Sci Transl Med 11 2019 10.1126/scitranslmed.aar5759
39 Rowinsky E.K. Humphrey R. Hammond L.A. Phase I and pharmacologic study of the specific matrix metalloproteinase inhibitor BAY 12-9566 on a protracted oral daily dosing schedule in patients with solid malignancies J Clin Oncol 18 2000 178 186 10.1200/JCO.2000.18.1.178 10623708
40 Wang H. Yuan M. Yang E. Chen D. Su A. Wu Z. Enterovirus 71 infection induced Aquaporin-4 depolarization by increasing matrix metalloproteinase-9 activity Neurosci Lett 759 2021 136049 10.1016/j.neulet.2021.136049
41 Tan Y.W. Yam W.K. Sun J. Chu J.J.H. An evaluation of chloroquine as a broad-acting antiviral against hand, foot and mouth disease Antiviral Res 149 2018 143 149 10.1016/j.antiviral.2017.11.017 29175128
42 Cena-Diez R. Singh K. Spetz A.L. Sonnerborg A. Novel naturally occurring dipeptides and single-stranded oligonucleotide act as entry inhibitors and exhibit a strong synergistic anti-HIV-1 profile Infect Dis Ther 11 2022 1103 1116 10.1007/s40121-022-00626-8 35391633
43 Lu J. He Y.Q. Yi L.N. Zan H. Kung H.F. He M.L. Viral kinetics of enterovirus 71 in human abdomyosarcoma cells World J Gastroenterol 17 2011 4135 4142 10.3748/wjg.v17.i36.4135 22039330
44 De Colibus L. Wang X. Spyrou J.A.B. More-powerful virus inhibitors from structure-based analysis of HEV71 capsid-binding molecules Nat Struct Mol Biol 21 2014 282 288 10.1038/nsmb.2769 24509833
45 Rossmann M.G. Viral cell recognition and entry Protein Sci 3 1994 1712 1725 10.1002/pro.5560031010 7849588
46 Shingler K.L. Yoder J.L. Carnegie M.S. The enterovirus 71 A-particle forms a gateway to allow genome release: a cryoEM study of picornavirus uncoating PLoS Pathog 9 2013 e1003240 10.1371/journal.ppat.1003240
47 Anasir M.I. Zarif F. Poh C.L. Antivirals blocking entry of enteroviruses and therapeutic potential J Biomed Sci 28 2021 10 10.1186/s12929-021-00708-8 33451326
48 Oberste M.S. Maher K. Flemister M.R. Marchetti G. Kilpatrick D.R. Pallansch M.A. Comparison of classic and molecular approaches for the identification of untypeable enteroviruses J Clin Microbiol 38 2000 1170 1174 10.1128/JCM.38.3.1170-1174.2000 10699015
49 Oberste M.S. Maher K. Kilpatrick D.R. Pallansch M.A. Molecular evolution of the human enteroviruses: correlation of serotype with VP1 sequence and application to picornavirus classification J Virol 73 1999 1941 1948 10.1128/JVI.73.3.1941-1948.1999 9971773
50 Brown B.A. Maher K. Flemister M.R. Resolving ambiguities in genetic typing of human enterovirus species C clinical isolates and identification of enterovirus 96, 99 and 102 J Gen Virol 90 2009 1713 1723 10.1099/vir.0.008540-0 19264596
51 Huang Y.L. Lin T.M. Wang S.Y. Wang J.R. The role of conserved arginine and proline residues in enterovirus VP1 protein J Microbiol Immunol Infect 55 2022 590 597 10.1016/j.jmii.2022.01.004 35232679
52 Chen J. Ye X. Zhang X.Y. Coxsackievirus A10 atomic structure facilitating the discovery of a broad-spectrum inhibitor against human enteroviruses Cell Discov 5 2019 4 10.1038/s41421-018-0073-7 30652025
53 Debing Y. Neyts J. Delang L. The future of antivirals: broad-spectrum inhibitors Curr Opin Infect Dis 28 2015 596 602 10.1097/QCO.0000000000000212 26524332
54 Pevear D.C. Hayden F.G. Demenczuk T.M. Barone L.R. McKinlay M.A. Collett M.S. Relationship of pleconaril susceptibility and clinical outcomes in treatment of common colds caused by rhinoviruses Antimicrob Agents Chemother 49 2005 4492 4499 10.1128/AAC.49.11.4492-4499.2005 16251287
55 Ho J.Y. Chern J.H. Hsieh C.F. In vitro and in vivo studies of a potent capsid-binding inhibitor of enterovirus 71 J Antimicrob Chemother 71 2016 1922 1932 10.1093/jac/dkw101 27098012
56 Cao Y. Lei E. Li L. Antiviral activity of Mulberroside C against enterovirus A71 in vitro and in vivo Eur J Pharmacol 906 2021 174204 10.1016/j.ejphar.2021.174204
57 Lin W.Y. Yu Y.J. Jinn T.R. Evaluation of the virucidal effects of rosmarinic acid against enterovirus 71 infection via in vitro and in vivo study Virol J 16 2019 94 10.1186/s12985-019-1203-z 31366366
58 Zhang G. Zhou F. Gu B. In vitro and in vivo evaluation of ribavirin and pleconaril antiviral activity against enterovirus 71 infection Arch Virol 157 2012 669 679 10.1007/s00705-011-1222-6 22245989
59 Rotbart H.A. Antiviral therapy for enteroviruses and rhinoviruses Antivir Chem Chemother 11 2000 261 271 10.1177/095632020001100402 10950388
60 Hayden F.G. Herrington D.T. Coats T.L. Efficacy and safety of oral pleconaril for treatment of colds due to picornaviruses in adults: results of 2 double-blind, randomized, placebo-controlled trials Clin Infect Dis 36 2003 1523 1532 10.1086/375069 12802751
61 Clark T.G. Bradburn M.J. Love S.B. Altman D.G. Survival analysis part I: basic concepts and first analyses Br J Cancer 89 2003 232 238 10.1038/sj.bjc.6601118 12865907
62 Ong K.C. Wong K.T. Understanding enterovirus 71 neuropathogenesis and its impact on other neurotropic enteroviruses Brain Pathol 25 2015 614 624 10.1111/bpa.12279 26276025
63 Hirte H. Goel R. Major P. A phase I dose escalation study of the matrix metalloproteinase inhibitor BAY 12-9566 administered orally in patients with advanced solid tumours Ann Oncol 11 2000 1579 1584 10.1023/a:1008347630465 11205466
64 Hirte H. Vergote I.B. Jeffrey J.R. A phase III randomized trial of BAY 12-9566 (tanomastat) as maintenance therapy in patients with advanced ovarian cancer responsive to primary surgery and paclitaxel/platinum containing chemotherapy: a National Cancer Institute of Canada Clinical Trials Group Study Gynecol Oncol 102 2006 300 308 10.1016/j.ygyno.2005.12.020 16442153
65 Molina J.R. Reid J.M. Erlichman C. A phase I and pharmacokinetic study of the selective, non-peptidic inhibitor of matrix metalloproteinase BAY 12-9566 in combination with etoposide and carboplatin Anti Cancer Drugs 16 2005 997 1002 10.1097/01.cad.0000176504.86551.5c 16162976
66 Erlichman C. Adjei A.A. Alberts S.R. Phase I study of the matrix metalloproteinase inhibitor, BAY 12-9566 Ann Oncol 12 2001 389 395 10.1023/a:1011183905848 11332153
