Use of Viral Entry Assays and Molecular Docking Analysis for the Identification of Antiviral Candidates against Coxsackievirus A16

Summary

The goal of the protocol is to illustrate the different assays relating to viral entry that can be used to identify candidate viral entry inhibitors.

Abstract

Antiviral assays that mechanistically examine viral entry are pertinent to discern at which step the evaluated agents are most effective, and allow for the identification of candidate viral entry inhibitors. Here, we present the experimental approaches for the identification of small molecules capable of blocking infection by the non-enveloped coxsackievirus A16 (CVA16) through targeting the virus particles or specific steps in early viral entry. Assays include the time-of-drug-addition analysis, flow cytometry-based viral binding assay, and viral inactivation assay. We also present a molecular docking protocol utilizing virus capsid proteins to predict potential residues targeted by the antiviral compounds. These assays should help in the identification of candidate antiviral agents that act on viral entry. Future directions can explore these possible inhibitors for further drug development.

Introduction

Hand, foot, and mouth disease (HFMD) is a disease most commonly caused by coxsackievirus A16 (CVA16) and enterovirus 71 (EV71) in young children. Recently across the Asia-Pacific region, there has been a significant uptick in CVA16-induced HFMD. While symptoms can be mild, severe complications can occur that affect the brain and the heart, with potential fatality^1,2. At present, there are no licensed antiviral therapies or vaccinations available for CVA16, and thus there is a pressing need to develop antiviral strategies to curb future outbreaks and the associated complications.

CVA16 is a non-enveloped virus which has an icosahedral capsid assembled from pentamers that each contain 4 structural proteins namely VP1, VP2, VP3, and VP4. Encircling each five-fold axis in the pentamer is a 'canyon' region that shows as a depression and is noted for its role in receptor binding³. At the bottom of this canyon lies a hydrophobic pocket in the VP1 region that contains a natural fatty ligand named sphingosine (SPH). Cellular receptors, such as human P selectin glycoprotein ligand 1 (PSGL-1) and scavenger receptor class B member 2 (SCARB2), have been suggested to play a role in viral binding by displacing this ligand which results in conformational changes to the capsid and the subsequent ejection of viral genome into the host cell^4,5,6. Identifying possible inhibitors that block the successive events in the viral entry process could provide potential therapeutic strategies against CVA16 infection.

The steps in the virus life cycle can be dissected through experimental approaches as targets to help identify mode-specific antiviral agents. A time-of-drug-addition analysis examines the drug treatment effect at different times during the viral infection, including pre-entry (added prior to the virus infection), entry (added concurrent to the virus infection), and post-entry (added following the virus infection)⁷. The impact can be assessed using a standard plaque assay by quantitating the number of viral plaques formed in each of the treatment conditions. The flow cytometry-based viral binding assay determines if the drug prevents viral attachment to host cells. This is achieved by shifting the temperature from 37 °C, at which the majority of human virus infections occur, to 4 °C, where the virions are able to bind to the host cell surface but are unable to enter the cells⁷. The cell membrane-bound virus particles are then quantified through immunostaining against viral antigens and assessed by flow cytometry. The viral inactivation assay on the other hand helps to assess potential physical interactions of the drug with free virus particles, either shielding or neutralizing the virions, or causing aggregations or conformational changes that render them inactive for subsequent interactions with the host cell surface during the infection^8,9. In this experiment, the viral inoculum is allowed to first incubate with the drug before being diluted to titrate out the drug prior to infecting the host cell monolayer and performing a standard plaque assay⁸. Finally, molecular docking is a powerful tool to predict potential drug interaction sites on the virion surface, including the viral glycoproteins from enveloped viruses and the viral capsid proteins from non-enveloped viruses, by using computational algorithms. This helps to mechanistically pinpoint targets of the drug's mode of action and provide useful information that can be further validated by downstream assays.

We recently employed the above described methods to identify antiviral compounds that efficiently blocked infection by the non-enveloped CVA16⁹. Herein, the detailed protocols that were used are described and discussed.

Protocol

NOTE: All cell culture and virus infections must be conducted in certified biosafety hoods that are appropriate for the biosafety level of the samples being handled. The two tannin-class of small molecules chebulagic acid (CHLA) and punicalagin (PUG), that were observed to efficiently block CVA16 infection⁹, are used as examples of candidate inhibitory agents. For basic principles in virology techniques, virus propagation, determination of virus titer, and concepts of plaque forming units (PFU) or multiplicity of infection (MOI), the reader is referred to reference¹⁰.

1. Cell Culture, Virus Preparation, Compound Preparation, and Compound Cytotoxicity

1. Human rhabdomyosarcoma (RD) cells are host cells permissive to CVA16 infection¹¹. Grow the RD cells in 10 mL of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 200 U/mL penicillin G, 200 µg/mL streptomycin, and 0.5 µg/mL amphotericin B in T-75 flasks at 37 °C in a 5% CO₂ incubator.
2. Prepare CVA16 by propagating the virus in RD cells and determine viral titer in PFU/mL. For optimized protocol, please refer to reference¹¹.
3. Prepare the test compounds and controls using their respective solvents: for example, dissolve CHLA and PUG in dimethyl sulfoxide (DMSO). For all infection steps, the basal medium consisted of DMEM plus 2% FBS and antibiotics.
   NOTE: The final concentration of DMSO in the test compound treatments is equal to or below 0.25% in the experiments; 0.25% DMSO is included as a negative control treatment in the assays for comparison.
4. Perform cytotoxicity assay of the test compounds on the RD cells using a cell viability determining reagent such as XTT ((2,3-bis[2-methoxy-4-nitro-5-sulfophenyl]-5-phenylamino)-carbonyl]-2H-tetrazolium hydroxide). For detailed protocol, please refer to reference¹². Determine the cytotoxicity concentrations of the test compounds using an analytical software such as GraphPad Prism according to manufacturer’s protocol. Drug concentrations that do not significantly influence the cell viability ( ≥95% viable cells) are used for the remainder of the study.

2. Time-of-drug-addition Assay

1. To evaluate the influence of drugs on host cells prior to viral infection (pretreatment)
   1. Seed RD cells in 12-well plates at a seeding density of 2 x 10⁵ cells/well. Incubate overnight at 37 °C in a 5% CO₂ incubator to obtain a monolayer.
   2. Treat RD cells with test compounds at non-cytotoxic concentrations (determined from step 1.4) in 1 mL of basal media volume for 1 h or 4 h.
   3. Wash cells with 1-2 mL of PBS before adding 50 PFU/well of virus in basal medium (final volume of the inoculum is 300 μL) for 1 h. Rock the plate every 15 min.
   4. Following the infection, wash the monolayers again using PBS, then overlay with 1 mL of basal media containing 0.8% methylcellulose for further incubation at 37 °C in a 5% CO₂ incubator.
   5. After 72 h of incubation, remove the overlay media and wash the wells using 2 mL of PBS.
   6. Fix the wells using 0.5 mL of 37% formaldehyde for 15 min.
   7. Remove the supernatant and wash again using PBS.
   8. Stain the wells using 0.5 mL of 0.5% crystal violet solution. Then remove the stain solution within 2 min and wash the wells with a gentle stream of water before air drying.
   9. Count the viral plaques by placing the plate on a white-light box. Calculate the percent (%) CVA16 infection as follows: (Mean # of plaque _virus+drug / Mean # of plaque _{virus+DMSO control}) × 100%.
2. To evaluate the effect of adding the drugs and the virus concurrently (co-addition)
   1. Seed RD cells in 12-well plates at a seeding density of 2 x 10⁵ cells/well. Incubate overnight at 37 °C in a 5% CO₂ incubator to obtain a monolayer.
   2. Treat RD cells with the test compounds at the appropriate concentrations and 50 PFU/well of CVA16 (final volume of the inoculum is 300 μL) simultaneously for 1 h. Rock the plate every 15 min.
   3. Wash cells with 1 mL of PBS and then overlay with 1-2 mL of basal media containing 0.8% methylcellulose for further incubation at 37 °C in a 5% CO₂ incubator.
   4. Stain viral plaques with crystal violet after 72 h post infection and determine the% CVA16 infection as described above.
3. To evaluate drug treatment effect after viral entry (post-infection)
   1. Seed RD cells in 12-well plates at a seeding density of 2 x 10⁵ cells/well. Incubate overnight at 37 °C in a 5% CO₂ incubator to obtain a monolayer.
   2. Inoculate the RD cells with 50 PFU/well of CVA16 (final volume of the inoculum is 300 μL) for 1 h. Rock the plate every 15 min.
   3. Wash the wells with 1-2 mL of PBS and overlay the cells with basal media containing 0.8% methylcellulose and the appropriate concentrations of test compounds.
   4. Stain viral plaques with crystal violet and count after 72 h post infection and determine the% CVA16 infection incubations described above.
      NOTE: Perform all PBS washes gently to avoid lifting the cells.

3. Flow Cytometry-based Binding Assay

1. Seed RD cells in 12-well plates at a seeding density of 2 x 10⁵ cells/well. Incubate overnight at 37 °C in a 5% CO₂ incubator to achieve a monolayer.
2. Pre-chill the cell monolayer at 4 °C for 1 h.
3. Infect RD cells with CVA16 (MOI = 100) in the presence and absence of the test compounds for 3 h at 4 °C.
   NOTE: Perform the viral inoculation on ice and the ensuing incubation in a 4 °C refrigerator to maintain the temperature at 4 °C, which permits viral binding but not entry.
4. Remove virus inoculum and wash once with 1-2 mL of ice-cold PBS.
5. Lift cells by adding 1 mL of ice-cold dissociation buffer to the wells on ice for 3 min, before collecting the cells and resuspending them in ice-cold flow cytometry buffer (1x PBS plus 2% FBS).
6. Wash cells twice using the ice-cold flow cytometry buffer and fix the cells with 0.5 mL of 4% paraformaldehyde for 20 min on ice.
7. Wash the cells using PBS to remove any unbound or weakly bound viruses, and then stain the cells with 1 mL of anti-VP1 antibody (1:2,000; diluted in PBS containing 3% BSA) on ice for 1 h, followed by incubation with a secondary Alexa 488-conjugated anti-mouse IgG (1:250; diluted in PBS containing 3% BSA) on ice for 1 h. Perform PBS washes (3 times) following each antibody treatment.
8. Resuspend the cells in 0.5 mL of the ice-cold flow cytometry buffer and perform flow cytometry analysis on a flow cytometer using standard procedures. Present data in histograms using the associated software and quantitate for bar graph representation.

4. Viral Inactivation Assay

1. Perform the viral inactivation assay as previously described¹² using the following conditions:
   - A starting concentration of 10⁶ PFU/mL of CVA16.
   - RD cell monolayers in 12-well plates from a seeding density of 2 x 10⁵ cells/well.
   - 50-fold dilution for titrating out the drug compounds resulting in a final virus concentration of 50 PFU/well.
   - Wash steps using 1-2 mL of PBS.
   - Overlay media containing 0.8% methylcellulose.
2. Perform final readout of viral infection using the crystal violet staining of viral plaques procedure as detailed above.

5. Molecular Docking Analysis

1. Download 3D molecules of test compounds from PubChem (https://pubchem.ncbi.nlm.nih.gov/). If molecules do not have a 3D structure uploaded, download the 2D structures or use the SMILES string sequence and transform into 3D molecules via a molecular program (e.g. CORINA).
2. Download viral biological assembly unit from RCSB Protein Data Bank (https://www.rcsb.org/) and prepare the viral structure model using a biocomputing program (e.g. UCSF Chimera). For example, in the case of CVA16 mature virion crystal structure (PDB: 5C4W)³, delete solvents from the PDB file, replace incomplete side chains using data from the Dunbrach 2010 Rotamer Library, and add hydrogens and charges to the structure as previously reported¹³. Docking targets can be any relevant viral proteins for the intended analysis with a biological assembly information (Protein Data Bank).
3. Dock test compounds onto the prepared virus unit using for example UCSF Chimera, and analyze the output files with a visualization software (e.g., Autodock Vina, PyMol):
   1. Upload the test compound file into UCSF Chimera as the ‘ligand’ and perform blind docking by selecting the whole prepared viral protein as the ‘receptor’. Use the computer mouse or trackpad to resize the search volume to the entire ‘receptor’. In ‘Advanced options’, allow the number of binding modes to be at maximum. Docking frames will be automatically ranked from highest to lowest binding energy.
   2. (Optional) Further to blind docking, confine the docking site onto the viral protein in regions of interest derived from the blind docking results using the mouse or trackpad again to reduce the search volume (e.g., 100 Å x 100 Å x 100 Å). This step helps confirm the blind docking results and increases specificity.
   3. Use a molecular graphics system (e.g., PyMol) to analyze the binding modes’ positions by uploading the docking file. Find polar contacts from the compound to the viral protein by selecting the ‘ligand’ and identifying polar contacts with the option ‘to any atoms’; examine the results.

Representative Results

The time-of-drug-addition assay is indicated in Figure 1 and shows the influence from treatment using the small molecules CHLA and PUG on CVA16 infection either pre-viral entry (pretreatment), during viral entry (co-addition), or post-viral entry (post-infection). Both small molecules only produced marginal impact against CVA16 infectivity whether in the pretreatment of the host cells prior to viral infection (Figure 1A) or in the post-infection treatment (Figure 1C). In contrast, CHLA and PUG efficiently abrogated the CVA16 infection by >80% in the co-addition treatment (Figure 1B). These observations therefore suggest that the two compounds are most effective when they are concurrently present with the virus particles on the host cell surface during the infection.

In Figure 2, the flow cytometry-based binding analysis (schematically illustrated in Figure 2A) confirms that the two tannins prevent CVA16 entry by preventing the viral particle binding to the host cells. The quantification data in Figure 2B shows that the amount of virus detected on the RD cell surface in the presence of the two drugs, is less than 10%, similar to the heparin positive control which is known to prevent CVA16 attachment¹⁴. Figure 2C, 2D, and 2E depict the associated flow cytometry histograms where the band shift due to detection of CVA16 on the RD cell surface is significantly reduced when CHLA and PUG are present.

Figure 3A depict how the viral inactivation experiment was performed. The drug compound was either mixed with the CVA16 virus particles and incubated for 1 h (long-term) prior to the dilution step, or mixed and immediately diluted (short-term) prior to the infection. As shown in Figure 3B, a pre-incubation of the CV16 particles with the test agents for 1 h led to a near complete protection of the RD cells against the viral infection compared to short-term incubation and the DMSO control. The results therefore suggest that both CHLA and PUG interact with the CVA16 particles and are able to render them inactive in the subsequent infection.

Since our data indicate that the drug compounds can directly inactivate CVA16 particles, and hence identifying the virion itself as a plausible target of their antiviral activity, we used molecular docking to predict the potential interaction(s) between these agents and the CVA16 capsid pentamer. Figure 4A shows a surface projection of the CVA16 pentamer which makes up the icosahedral capsid of the CVA16 virion. Molecular docking of the tannins CHLA (Figure 4B; green) and PUG (Figure 4C; blue) indicate that they both are predicted to bind in the canyon region of the CVA16 pentamer. Specifically, both small molecules bound just above the pocket entrance (Figure 4B and 4C, zoomed panels), which holds the pocket factor and plays an important role for mediating CVA16 binding and entry into the host cell. Both CHLA and PUG therefore appear to mask the pocket entrance region, which theoretically would obstruct interactions between the virus particles and the host cell receptors. Figure 4D and 4E indicate the unique residues predicted from the polar contacts of CHLA and PUG, respectively, around the pocket entrance, with most of these interactions occurring with VP1 for both compounds and the 3 amino acids Asn⁸⁵, Lys²⁵⁷, and Asn⁴¹⁷ being in common between the two tannins.

Figure 1: Time-of-drug-addition effect of CHLA and PUG against CVA16 infectivity. RD cells were treated with CHLA (20 µM) or PUG (25 µM) at different times of CVA16 inoculation (50 PFU/well). DMSO (0.25%) treatment was included as negative control and all assays were analyzed by plaque assay using crystal violet staining 72 h after incubation. (A) For pretreatment, cells were incubated with the test compounds for 1 h or 4 h and then were washed before CVA16 infection. (B) For co-addition assays, cells were administered with drugs and virus simultaneously for 1 h and then washed. (C) In post-infection, cells were infected with CVA16 for 1 h, washed, and then treated with test compounds. Data shown are the means ± standard deviation (SD) from three independent experiments. *p < 0.05 compared to the respective ‘virus only’ group. Statistical analysis was performed using one-way analysis of variance. This figure has been adapted from reference⁹. Please click here to view a larger version of this figure.

Figure 2: CHLA and PUG abolish CVA16 binding to the host cell. (A) Schematic of the flow cytometry- based binding assay. (B) RD cells (2 x 10⁵ cells/well) were infected with CVA16 (MOI = 100) in the presence or absence of CHLA (20 µM), PUG (25 µM), soluble heparin (500 µg/mL, positive control), or DMSO (0.25%, negative control) for 3 h at 4 °C. Inocula from wells were collected into tubes, washed with PBS twice, fixed, and stained with anti-VP1 antibody followed by Alexa 488-conjugated secondary antibody for flow cytometry detection of surface-bound viruses. Quantified data from the detected fluorescence signals were plotted as the means ± SD from three independent experiments in bar graph as ‘Virus binding (%)’. *p < 0.05 compared to the ‘DMSO’ control treatment. Statistical analysis was performed using one-way analysis of variance. The representative flow cytometry histograms of CHLA (C), PUG (D), and heparin (E) treatments are shown. This figure has been adapted from reference⁹. Please click here to view a larger version of this figure.

Figure 3: CHLA and PUG inactivate cell-free CVA16 virus particles. (A) Schematic of the viral inactivation assay. (B) CVA16 (10⁶ PFU/well) was treated with CHLA (20 µM) or PUG (25 µM) and mixed immediately for short-term inactivation or incubated for 1 h at 37 °C for long-term inactivation before being diluted 50-fold to a non-effective concentration of test compounds before inoculating on RD cells (final virus concentration = 50 PFU/well). DMSO (0.25%) was used as a negative control. Experiments were analyzed by plaque assay using crystal violet staining 72 h post-infection. Data shown are the means ± SD from three independent experiments. *p < 0.05 compared to the respective ‘virus only’ group. Statistical analysis was performed using one-way analysis of variance. This figure has been adapted from reference⁹. Please click here to view a larger version of this figure.

Figure 4: CHLA and PUG target the CVA16 capsid near the pocket entrance. Surface projection of the CVA16 virion particle with the monomeric structural pentamer delineated by red lines (A). Additional pentamers on the virion are shown in cyan, magenta, indigo, bronze, and green. Molecular docking analysis of CHLA (B, green) and PUG (C, blue) on the CVA16 pentamer (PDB: 5C4W); zoomed-in panels are demarcated in yellow. VP1 = orange, VP2 = gray, VP3 = white; polar contacts are shown as black dashes. Residues that make-up pocket entrance are colored red (Ile⁹⁴, Asp⁹⁵, Gln²⁰⁷, Met²¹², Met²¹³, Lys²⁵⁷, Thr²⁵⁸). D, E. Close-up side view into the canyon where the pocket entrance is located and where CHLA (D) and PUG (E) bind to. Unique residues that are polar contacts from the compounds’ polar contacts on the pentamer are labeled in yellow (VP1), white (VP2), and in black (VP3) fonts. The white dashed line indicates the pocket entrance region. This figure has been adapted from reference⁹. Please click here to view a larger version of this figure.

Discussion

In this report, we described the protocols that are useful for the identification of antiviral candidates that target viral entry, in particular against the non-enveloped CVA16. The assays are designed in ways to dissect the early events during viral entry, which is helpful to clarify the mechanism(s) of action and potential target(s) of the test agents' antiviral activity. The 'time-of-drug-addition assay' permits to broadly determine the potential target of the test compounds, for instance the uninfected host cells (pretreatment analysis), the virus particles or its interactions with the host cell surface (co-addition analysis), or the virus-infected host cell during the viral replicative phase (post-infection analysis). This assay alone can determine the method of interaction from the compounds (e.g., co-addition) that leads to the subsequent assays described in this protocol (e.g., viral inactivation assay and binding analysis). Wash steps are critical to ensure that the treatment method examined is specific to the one analyzed. The use of the 'flow cytometry-based binding assay' helps to assess the influence of the compounds specifically on virus binding to the host cell. Maintaining the temperature of the experiment at 4 °C is important to the final detection of the virions on the cell surface, as this temperature permits viral binding but not entry. The 'viral inactivation assay' can aid to determine potential physical interaction of the drug compounds with the cell-free virus particles. The critical step is the dilution for titrating out the drug compounds following incubation with the viral inoculum, as this is necessary to prevent any meaningful interaction of the drugs with the host cell surface in the subsequent infection step¹².

Since viral entry is a multi-step event, a viral entry inhibitor class of antiviral agents could possibly exert several types of mechanisms, including: (1) modulating cell surface entry factors/receptors or its associated signaling pathways; (2) affecting cell membrane fluidity or integrity; (3) targeting electrostatic or van der Waals interactions between the virus particles and the host cell surface; (4) inducing physical changes to the virions such as particle breakage or aggregation; (5) binding to viral glycoproteins or capsid proteins and preventing their functions or conformational changes; (6) blocking fusion associated mechanisms; and (7) prevent release of viral genome inside the host cell. The analyses described in this report can therefore help point to the above-listed potential modes of action that can be further validated by additional experiments. Lastly, the 'molecular docking analysis' described here is instrumental to predict potential interaction regions between the drug compounds and the virus particles, and as such can help identify candidate viral capsid or glycoprotein binders and the targeted residues on the virus particles. However, these predictions are dependent on the docking software, and the resolution and accuracy of the viral protein crystal structures. It is important to note that the optional confined docking method in step 5.3.2 was added because oftentimes when using viral structural proteins as the 'receptor' molecule, the ligand can possibly bind to regions normally not accessible or exposed on the surface (e.g., under surface of the virion capsid facing the inside of the virion, transmembrane regions of envelope glycoproteins, etc.). Confining the search box allows only accessible regions of the viral protein to be targeted and rules out any unrealistic interactions. Molecular docking is dependent on crystallized structures, but recent advances in homology modeling have enabled analysis of non-crystallized structures by fitting its amino acid sequence onto a closely related crystallized structure¹⁵. This has allowed more structures to be analyzed and the information acquired can be useful for further studies including mutational analyses that can help validate the predicted interactions.

In conclusion, the assays and protocols described in this report are specifically catered to identify candidate antiviral agents that target viral entry, and provide information on which step of the viral entry process the test agent targets to, whether they interact with free virus particles, and predicting possible drug interaction sites on the virions. These types of assays can be repeated on other non-enveloped viruses or adapted to enveloped viruses as a method of screening antiviral drug compounds for possible inhibitors of viral entry. Using such mechanism-driven approach to identify antiviral candidates could help expedite the drug development process and expand the scope of antiviral therapeutics.

Materials

Name	Company	Catalog Number	Comments
4% Paraformaldehyde	Sigma	AL-158127-500G
Alexa 488-conjugated anti-mouse IgG	Invitrogen	A11029
Amphotericin B	GIBCO	15290-018
Anti-VP1 antibody	Merck-Millipore	MAB979	Anti-Enterovirus 71 Antibody, cross-reacts with Coxsackie A16, clone 422-8D-4C-4D
Beckman Coulter Cytometer	Beckman Coulter	FC500
Corina	Molecular Networks GmbH
Crystal violet	Sigma	C3886-100G
DMEM	GIBCO	11995-040
DMSO	Sigma	D5879
FBS	GIBCO	26140-079
Formaldehyde	Sigma	F8775
Graphpad Prism	GraphPad
Heparin sodium salt	Sigma	H3393
In vitro toxicology assay kit, XTT-based	Sigma	TOX2
Methylcellulose	Sigma	M0512-100G
PBS pH 7.4	GIBCO	10010023
Penicillin-Streptomycin	GIBCO	15070-063
PyMol	Schrödinger
UCSF Chimera	University of California, San Francisco