1. Introduction
Influenza is an infectious disease caused by the influenza virus which is a RNA virus of the family Orthomyxoviridae. Influenza spreads around the world in seasonal epidemics, with an estimated three to five million cases of severe illness and 250,000 to 500,000 deaths per annum [
1]. Four major influenza pandemics occurred in the 20th century that caused more than 20–50 million deaths, and influenza virus infection remains one of the leading causes of mortality [
2,
3]. A new H1N1 influenza A virus, also called the 2009 H1N1 pandemic influenza virus (2009 H1N1 virus), had spread throughout the world and caused a serious influenza pandemic in 2009 [
4,
5]. Over 17,000 reported deaths have been caused by 2009 H1N1 virus infection since its identification in Mexico in April 2009, so drugs and vaccines against 2009 H1N1 virus infection are urgently needed [
6]. However, 2009 H1N1 virus, like many other influenza virus stains, has developed resistance to commercially available anti-influenza drugs. Currently the neuraminidase (NA) inhibitor oseltamivir, which can interfere with the enzymatic activity of the NA of the influenza virus, is mainly used for the treatment of influenza patients, but the 2009 H1N1 virus has been reported to be resistant to it [
7,
8]. It has been recently reported that over 160 sporadic viral isolates of 2009 H1N1 virus show resistance to oseltamivir due to the NA H275Y genotype mutation [
8,
9]. On the other hand, though the vaccines against 2009 H1N1 virus infection have been developed and used in clinical practice, the safety of theses vaccines remains one of the major public concerns in most of countries [
10,
11,
12,
13], as deaths and serious side effects of vaccines against 2009 H1N1 virus have been reported [
14].
The haemagglutinin (HA) on the surface of influenza virus particles is a major viral membrane glycoprotein molecule, which is synthesized in the infected cell as a single polypeptide chain precursor (HA0) with a length of approximately 560 amino acid residues and subsequently cleaved by an endoprotease into two subunits called HA1 and HA2 and then be covalently attached by the disulfide bond [
15,
16]. The crystallographic structure of the HA shows a long tightly intertwined fibrous stem domain at its membrane-proximal base, a globular head which contains the sialic acid receptor binding site (RBS) and five antigenic sites surrounding the RBS [
17]. The mature HA on the viral surface is a trimeric rod-shaped molecule with the carboxy terminus inserted into the viral membrane and the hydrophilic end forming the spike of the viral surface [
18,
19,
20]. Although the amino acid sequence identity of different virus strains can be less than 50%, the structure and functions of these HAs are highly conserved [
16]. The major functions of the HA are as the receptor-binding ligand, leading to endocytosis of the virus into the host cell and subsequent membrane-fusion events in the infected cells [
16,
21]. Influenza virus initiates infection by binding to sialic acids on the surface of target cells. After endocytosis, the endosome acquires a lower pH value, mainly because of the activity of the Vacuolar-type H+-ATPase (V-ATPase) [
22]. In the acid environment of the endosome, the HA molecule is cleaved into HA1 and HA2 subunits and then undergoes a conformational change which resulting in the exposure of the fusion peptide at the
N-terminus of the HA2 subunit [
23,
24]. The fusion peptides insert into the endosomal membrane, while the transmembrane domains remain anchored in the viral membrane. Finally, the fusion peptide brings the endosomal membrane and the viral membrane into juxtaposition, leading to fusion. Subsequently, a pore is opened up by this structural change of more than one haemagglutinin molecule and then the contents of the virion are released into the cytoplasm of the cell. This completes the uncoating process [
25]. Because of the conformational change of viral HA protein is indispensable for the membrane fusion process between influenza virus and the endosome of the host cell, this makes it a new target for anti-influenza virus drug development. Recently, some small compounds acting as HA conformational change inhibitors have been reported [
26,
27].
Herbal extracts have been reported to have an important role in controlling virus infections by serving as immuno-modulators during influenza virus infection [
28] or blocking the interaction of virus with target cells or having virucidal activity through direct interaction with the virus [
29,
30]. Most importantly, accumulating evidence has suggested that treatment of herbal extracts might be able to reduce the risk of drug-resistant virus emergence [
31].
Melaleuca alternifolia Concentrate (MAC), which is an essential oil derived from the leaves or terminal branches of the native Australian tea tree,
Melaleuca alternifolia, is a heterogeneous mixture of approximately 100 chemically defined components that mainly contains terpinen-4-ol (56%–58%), γ-terpinene (20.65%), and α-terpinene (9.8%) [
32]. The ability of MAC to induce anti-inflammatory effect [
33,
34] and inhibit infection of various microbial species, such as bacteria [
35,
36], viruses [
37,
38,
39] and fungi [
40,
41] makes it a promising candidate for development of therapeutics against 2009 H1N1 virus infection.
The purpose of this study was to determine the antiviral effect against 2009 H1N1 virus using an in vitro test of cytopathic effect (CPE) inhibition of MAC. As previously described, terpinen-4-ol was the main component of MAC, so here we also assessed the feasibility and sensitivity of interaction of terpinen-4-ol with the viral haemagglutinin protein through in silico prediction to confirm the drug target and the characterization of the protein changes after treatment with MAC.
3. Experimental
3.1. Bio-Safety
All experiments involving pathogenic influenza A viruses were performed in a bio-safety level 2 (BSL2) laboratory of Zhongshan School of Medicine of Sun Yat-sen University, Guangzhou, China.
3.2. Cells and Virus
Madin-Darby Canine Kidney (MDCK) cells maintained by our laboratory were grown in Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen Corporation, New York, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Thermo Scientific HyClone product line, Logan, UT, USA) at 37 °C, 5% CO2 (Heracell 150i, Thermo Scientific, Langenselbold, Germany). No antibiotics or anti-mycotic agents were used in cell or virus culture. 2009 H1N1 pandemic influenza virus strain, A/GuangzhouSB/01/2009(H1N1) (GZ01/09 for short) was a gift from the Guangdong Centers for Disease Control and Prevention that was propagated from clinical isolates and maintained in our laboratory. The virus strain was propagated in MDCK cells that were cultured in 0.02% TPCK- trypsin (Amresco Inc., Solon, OH, USA) at 37 °C, 5% CO2. The supernatant containing virus particles in MDCK cell culture was collected when 75%–100% CPE was observed. The virus was stored at −80 °C in aliquots until use.
3.3. Melaleuca Alternifolia Concentrate (MAC)
Hundred percent MAC (batch 270409) was provided by NeuMedix Biotechnology Pty Ltd, North Sydney, Australia. Preliminary experiments established the optimal solubility into dimethyl sulfoxide (DMSO) (Beijing Dingguo Changsheng Biotechnology Co. Ltd., Beijing, China) and the concentration of stock solution was 10% (v/v). For testing, the MAC stock solution was diluted by serum free DMEM media for working solutions with various concentrations.
3.4. Virus Titrations
The virus strain was titrated by standard Tissue Culture Infectious Dose
50 (TCID
50) Assay in MDCK cells. Briefly, MDCK cells were seeded in 96-well culture plates (about 5 × 10
4 cell/well) in DMEM with 10% fetal bovine serum (FBS) for 12–24 h at 37 °C with 5% CO
2. After cell propagation, growth medium was removed and 10 fold serial dilutions of the GZ01/09 virus suspension in DMEM media with 1 μg/mL TPCK-trypsin were added to the wells. The plate was incubated at 37 °C with 5% CO
2, and morphological changes on the MDCK cells were observed microscopically every 12 h. The final CPE was recorded after 72 h. TCID50 was calculated by counting all the wells with 1–4 CPE as being positive. TCID50 was calculated by the Reed-Muench method [
50].
3.5. MTT Assay to Determine the Cellular Viability of MDCK Cells
The cellular viability of MDCK cells was measured quantitatively by the reduction of formazan dye using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Beijing Dingguo Changsheng Biotechnology Co. Ltd., Beijing, China) assay. Briefly, confluent MDCK cell monolayer in 96-well culture plates was washed with sterile PBS and incubated for 3 h at 37 °C after 40 μL/well of MTT solution (5 mg/mL) was added into each well. When a purple precipitate was clearly visible, the liquid was carefully withdrawn without touching the sediment or the cells. DMSO at 100 μL/well was added to dissolve the purple formazan, and the absorbance at A490 nm was read with an Absorbance Microplate Reader (Gene Co. Ltd., Hong Kong, China).
3.6. Bioimaging in 96 Well Plates
The influence to entering host cell of the influenza virus by MAC treatment was determined by an immunofluorescence assay on MDCK cells in a 96 well plate. Briefly, MDCK cells were plated in a sterile 96-well plate about 10,000 cells/well. The influenza virus suspension treated with MAC of final concentration of 0.010% for 0.5 and 1 h at room temperature, virus suspension and maintain media for cell control were inoculated to the cell monolayer respectively, for 5 h in order for sufficient viral protein synthesis in the host cell. The cells were incubated at room temperature in the 3.7% formaldehyde 10 min for fixation; 0.1% Triton X-100 5 min for permeabilization and 3% fetal bovine serum 30 min for blocking. The influenza virus was stained with influenza A m1 (matrix protein 1) antibody (Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA) followed by Alexa Fluor® 488 Goat Anti-Mouse IgG (H + L) (Molecular Probes, Invitrogen, Carlsbad, CA, USA). Finally, 50 µL per well of Fluoroshield™ with 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich, Inc., St. Louis, MO, USA) was added and analyzed using an imaging instrument (Leica DMI4000B, Meyer Instruments, Inc., Houston, TX, USA).
3.7. Electron Microscopy Observation of the Influenza Virus Morphology
MDCK cells with or without treatment with MAC were observed under an electron microscope. The concentration and the treatment time of MAC were indicated in the figure legends. Each 10 μL of MAC-treated and untreated virus suspension was placed on a clean slide. Two copper grids were applied to float on the drops of virus suspensions using fine, clean forceps for 2 min. The bulk of the fluid was removed with the edge of the copper grid vertically on a strip of filter paper. Air dried the copper grid for 1 min. The copper grids were applied to float on a drop of 2% potassium phosphotungstate, using fine clean forceps, for 1 min. The bulk of the fluid was removed with the edge of the copper grid vertically on a strip of filter paper. Air dried the grid and examined in the electron microscope.
3.8. Statistical Analysis
The cell survival result in each group was expressed as the mean ± S.D. and the data was statistically compared with the relative control group using one-way analysis of variance (ANOVA), SPSS 17.0 for Windows software. p < 0.05 was considered to be statistically significant.
3.9. Molecular Docking
The structure of HA (PDB: 3AL4) [
51] was used in the docking calculations. The program Autodock 4.0 [
52] with a Lamarckian genetic algorithm is used to carry out the molecular docking. To evaluate the binding energies between the ligand and receptor, the AutoGrid program was used to generate the grid map with 80 × 80 × 80 points spaced equally at 0.375 Å is using. The number of GA runs is 200 and the energy evaluation is 25,000,000, other docking parameters were set to default values. At the end of the run, all docked conformations were clustered using a tolerance of 2 Å for root mean square deviations (RMSDs) and ranked based on docking energies.
3.10. Molecular Dynamics Simulations
The Amber 11.0 simulation suite [
53] was used in molecular dynamics (MD) simulations and data analysis. An all-atom model of HA was generated using the tleap module on the basis of the initial model. To release conflicting contacts among residues, energy minimization was performed with steepest descent method for 500 steps, followed by conjugated gradient method for 500 steps. The protein was then solvated with water in a truncated tetrahedral periodic box (76.096 × 76.096 × 76.096 nm). The TIP3P [
54] water model was used, and five Na
+ counterions were added to neutralize the system. Prior to the production phase, the following equilibration protocol was applied. First, the solvent was relaxed by energy minimization while restraining the protein atomic positions with a harmonic potential. The system was then energy-minimized without restraints for 2,500 steps using a combination of steepest descent and conjugated gradient methods. The system was gradually heated from 0 to 300 K over 20 ps using the NVT enemble. Finally, 20-ns MD simulation was conducted at 1 atmosphere and 300 K with the NPT ensemble. During the simulation, the SHAKE [
55] algorithm was applied to constrain the covalent bonds to hydrogen atoms. A time step of 2 fs and a non-bond interaction cutoff radius of 12.0 Å were used. Coordinates were saved every 1 ps during the entire process. The ff03 all atom Amber force field (AMBER ff03) developed by Duan
et al. [
56], which shows a good balance in the balance between helix and sheet, was used for the protein and the AMBER GAFF [
57] was used for the ligand. The parameters for terpinen-4-ol were developed as follows: the electrostatic potential of terpinen-4-ol was obtained at the HF/6-31G basis set from GAUSSIAN 2003 [
58] after a geometry optimization at the same level. The partial charges were derived by fitting the gas-phase electrostatic potential using the restrained electrostatic potential (RESP) method [
59]. The missing interaction parameters in the ligand were generated using antechamber tools in Amber. The long-range electrostatic were calculated by the particle-mesh ewald (PME) method [
60]. Then we used molecular mechanics generalized Born surface area (MM-GBSA) to estimated the binding energies at 192 AMD Opteron (tm) Processor CPUs (2.0 GHz) were used in the simulation process.
3.11. Binding Free Energy Calculation
The binding free energies (ΔG
bind) were calculated using the MM-GBSA approach [
61] inside the AMBER program. The first step of MM-GBSA method was the generation of multiple snapshots from an MD trajectory of the protein-ligand complex and a total 0f 50 snapshots were taken from the last 5 ns trajectory with an interval of 100 ps. For each snapshot, the free energy was calculated for each molecular species (complex, receptor, and ligand) using the following equation [
62]:
where G
com, G
rec, and G
lig were the free energies for the complex, receptor, and ligand, respectively. ΔE
mm was the molecular mechanics energy of the molecule expressed as the sum of the internal energy of the molecule plus the electrostatics and van der Waals interactions; ΔG
solv was the solvation free energy of the molecule; T was the absolute temperature; and ΔS is entropy of the molecule. ΔE
elec was the Coulomb interaction, ΔE
vdw was the van der Waals interaction, and ΔE
ini was the sum of the bond, angle, and dihedral energies; in this case, ΔE
ini = 0. ΔG
GB is polar solvation contribution calculated by solving the GB equation [
63] for MM_GBSA method. ΔG
np was the nonpolar solvation term; γ was the surface tension that was set to 0.0072 kal/(mol Å
2);. ΔSASA is the solvent accessible surface area (Å
2) that was estimated using the MOLSURF algorithm and β was a constant that was set to 0. The solvent probe radius was set to 1.4 Å to define the dielectric boundary around the molecular surface. The vibrational entropy contributions were estimated by NMODE analysis [
64] and 50 snapshots were used in the NMODE analysis. To obtain the contribution of each the binding energy, MM_GBSA was used to decompose the interaction energies to each residue involved in the interaction but only considering molecular mechanics and solvation energies without the contribution of entropies.