Trans-(−)-Kusunokinin: A Potential Anticancer Lignan Compound against HER2 in Breast Cancer Cell Lines?

Trans-(−)-kusunokinin, an anticancer compound, binds CSF1R with low affinity in breast cancer cells. Therefore, finding an additional possible target of trans-(−)-kusunokinin remains of importance for further development. Here, a computational study was completed followed by indirect proof of specific target proteins using small interfering RNA (siRNA). Ten proteins in breast cancer were selected for molecular docking and molecular dynamics simulation. A preferred active form in racemic trans-(±)-kusunokinin was trans-(−)-kusunokinin, which had stronger binding energy on HER2 trans-(+)-kusunokinin; however, it was weaker than the designed HER inhibitors (03Q and neratinib). Predictively, trans-(−)-kusunokinin bound HER2 similarly to a reversible HER2 inhibitor. We then verified the action of (±)-kusunokinin compared with neratinibon breast cancer cells (MCF-7). (±)-Kusunokinin exhibited less cytotoxicity on normal L-929 and MCF-7 than neratinib. (±)-Kusunokinin and neratinib had stronger inhibited cell proliferation than siRNA-HER2. Moreover, (±)-kusunokinin decreased Ras, ERK, CyclinB1, CyclinD and CDK1. Meanwhile, neratinib downregulated HER, MEK1, ERK, c-Myc, CyclinB1, CyclinD and CDK1. Knocking down HER2 downregulated only HER2. siRNA-HER2 combination with (±)-kusunokinin suppressed HER2, c-Myc, CyclinB1, CyclinD and CDK1. On the other hand, siRNA-HER2 combination with neratinib increased HER2, MEK1, ERK, c-Myc, CyclinB1, CyclinD and CDK1 to normal levels. We conclude that trans-(±)-kusunokinin may bind HER2 with low affinity and had a different action from neratinib.

Trans-(−)-kusunokinin, a lignan compound, isolated from Piper nigrum inhibits breast cancer cell growth, and induces cell cycle blockage, apoptosis and cell cycle arrest at the Gap 2/mitotic (G2/M) phase [18]. The synthetic compound of trans-kusunokinin consists of two forms, including trans-(−)-kusunokinin and trans-(+)-kusunokinin. This synthetic compound exhibits similar cytotoxic activity on breast and colon cancer, the same as the extract. The synthetic trans-(±)-kusunokinin is found to suppress topoisomerase II, STAT3, CyclinD1 and p21 on breast cancer and cholangiocarcinoma cells [19]. For the in vivo study, trans-(−)-kusunokinin decreases tumor growth and migration without side effects on blood parameters and the clinical chemistry of renal and liver function [20]. Interestingly, the synthetic (±)-kusunokinin inhibits the proliferation of breast cancer cells through the binding and the suppression of CSF1R, which consequently decreases AKT and the downstream proteins in cell proliferation (CyclinD1 and CDK) [21]. Using computational analysis, trans-(−)-kusunokinin binds AKR1B1, and the upstream molecules of the AKT signaling protein in migration [22]. In this recent study, we performed a computer simulation and an in vitro study in order to determine the target protein of racemic trans-kusunokinin in breast cancer.

Molecular Docking of Trans-(−)-Kusunokinin and Trans-(+)-Kusunokinin with Breast Cancer Related-Candidate Proteins
The docking scores of both trans-(−)-kusunokinin and trans-(+)-kusunokinin against 10 breast cancer associated proteins are shown in Table 1. All PDB codes of these 10 proteins and PubChem CID of all given inhibitors were summarized in the Supplementary Materials (Table S1). The docking score was given as an effective binding energy (∆Gbind) in kcal/mol although it may differ from the true binding energy. HER2 became of interest as the first prioritized target as trans-(±)-kusunokinin because HER2 was suggested as the challenging targets for breast cancer therapy [23], due to its drug resistance, cross-resistance to other HER2-targeted drugs [24,25], and drug toxicity [26,27]. These preliminary results and information encouraged further simulation study of trans-(−)-kusunokinin and trans-(+)-kusunokinin with HER2, along with the experiment of synthetic inhibitors and neratinib (HKI-272). Molecular docking could, however, only roughly guide the potential binding site of the drug towards an interested protein. MD simulation was thus undertaken in order to improve the prediction of the relative binding energy of the compounds compared to the known inhibitors, and the inspection of binding behaviors with an effect of temperature, pressure and an aqueous environment.

Atomistic Features of Trans-(−)-Kusunokinin and Trans-(+)-Kusunokinin and HER2
This study consisted of five MD simulations: four MD trajectories of ligand-protein complexes (trans-(−)-kusunokinin, trans-(+)-kusunokinin, synthetic inhibitor 03q and neratinib) and a simulation of the HER2 protein. The mentioned ligands are illustrated in Figure 1. Structural analysis was performed based on two plots: RMSD and RMSF plots ( Figure 2). The RMSD plot denoted the stability of the MD simulation after 50 ns of NPT simulation time since the RMSD values in all five simulations constantly fluctuated ( Figure 2A). Hence, taking the last 50 ns simulation was valid in terms of both structure and binding analysis. The plot of RMSD from the beginning of the simulation is shown in Supplementary Materials ( Figure S1). In addition, the RMSF pattern of the ligand-bound HER2 was illustrated together with ligand-free HER, Figure 2B; therefore, the effect of a bound ligand on flexibility could be easily detected. The difference in flexibility patterns of kusunokinin compared to ligand free HER2 was observed from the following amino acid residues: 50th to 65th, 80th to 90th, and 100th to 110th, respectively. Nevertheless, the flexibility pattern of kusunokinin only shared similarities with neratinib at some points, from the residues 50th to 65th residues, meanwhile other flexibility parts were clearly different. In Figure 3, the compound interacted towards HER2 protein via hydrophobic interaction since the cavity of the HER2 binding site was mostly composed of nonpolar amino acids, namely isoleucine, leucine, valine, methionine, alanine and phenylalanine, respectively. Trans-(−)-kusunokinin and trans-(+)-kusunokinin had bound to the same HER2 site, using hydrophobic interaction and π-π stacking due to Phenylalaine159 (Phe159). However, the hydrogen bond between oxygen atoms in the lactone ring and lysine48 (Lys48) could be a selective key to facilitate better HER2 binding of trans-(−)-kusunokinin, as shown in Figure 3A,B, respectively. Phe159 was also found to contribute interaction with the aromatic ring of all compounds, except neratinib, as shown in Figure 3A-C. The betting binding affinity (lower binding energy) from the synthetic 03q likely came from more additional hydrogen bonds, such as Lysine48 (Lys48) and arginine144 (Arg144) in the case of 03q ( Figure 3C). Furthermore, our simulation showed that the neratinib binding site consisted of cysteine100 (Cys100), methionine96 (Met96), leucine21 (Leu21) and threonine (Thr93), Figure 3D. This Cys100 could justify our binding prediction as it was a vital key residue in drug-HER2 interaction [28]. The relative binding energy was then evaluated from MM/GBSA and MM/PBSA based on the MD trajectory. To be noted, the relative binding energy was just able to predict the relative binding tendency of the ligand to the target protein, not to identify whether the compound can truly bind the studied target. Rather, in our study, we could predict the relative possibility of synthetic (±)-kusunokinin for HER2 binding with respect to the known inhibitors. The relative HER2 binding energies of trans-(−)-kusunokinin and trans-(+)-kusunokinin were then acquired, whereas neratinib and synthetic 03q were taken as the reference binding energy (Table 2). From Table 2, lower binding energies (better binding affinity) were observed from neratinib and synthetic 03q than kusunokinin. This was not surprising due to the fact that both inhibitors were selectively designed for HER2 specific binding. The calculations suggest that trans-(−)-kusunokinin to trans-(+)-kusunokinin would not compete well with neratinib and synthetic 03q in terms of HER2 binding.

Synthetic (±)-Kusunokinin Inhibited Breast Cancer Cell Proliferation through the Suppression of RAS and ERK
As suggested from the molecular simulation results, HER2 was an examined target protein in this study. HER2 and its downstream proteins (Ras, MEK1, AKT and ERK) related to cell proliferation were evaluated. The protein levels of HER2, Ras, MEK1, AKT and ERK were measured by Western blot analysis ( Figure 6A). We found that HER2 was inhibited by all treatments, except a synthetic (±)-kusunokinin. The combination of neratinib and siRNA-HER2 caused the HER2 signal to be close to the normal level of the control groups (lipofectamine and siRNA-luciferase), whereas the synthetic (±)-kusunokinin combination with siRNA-HER2 did not increase HER2 protein ( Figure 6B). Synthetic (±)-kusunokinin decreased RAS and ERK, meanwhile neratinib decreased MEK1 and ERK at 48 h after treatment. RAS was also decreased in the treatments of the synthetic (±)-kusunokinin combination with neratinib. Surprisingly, MEK1 and ERK proteins were increased, the same as in the control groups when MCF-7 cells were treated with the neratinib combination with siRNA-HER2. In addition, a combination of synthetic (±)-kusunokinin and siRNA-HER2 induced AKT and ERK levels close to the normal level ( Figure 6B-F). These results suggest that synthetic (±)-kusunokinin inhibited cell proliferation through the reduction of RAS and ERK, which is different from neratinib because it suppressed the HER2, MEK1 and ERK proteins.

Synthetic (±)-Kusunokinin Decreased CyclinB1, CyclinD1 and CDK1; Down-Stream Proteins of HER2
To determine the inhibition of breast cancer cell proliferation by synthetic (±)kusunokinin, four final proteins of the HER2 pathway were examined, including c-Myc, CyclinB1, CyclinD1 and CDK1 using Western blot analysis ( Figure 7A). We found that neratinib, the combination of neratinib and synthetic (±)-kusunokinin, and the combination of siRNA and synthetic (±)-kusunokinin decreased c-Myc, CyclinB1, CyclinD1 and CDK1. Treatment with synthetic (±)-kusunokinin alone decreased CyclinB1, CyclinD1 and CDK1. Interestingly, silencing HER2 did not cause downregulation on cell proliferation proteins. The combination of neratinib and siRNA-HER2 upregulated c-Myc, CyclinB1, CyclinD1 and CDK1 to the normal baseline, the same as in the control groups ( Figure 7B-E). Taken together, these results suggest that synthetic (±)-kusunokinin could function in the absence of HER2, which was different from neratinib, a HER2 known inhibitor.

Discussion
In our previous work, trans-(−)-kusunokokinin was a potential compound for cancer treatment. However, many points remain unclear, especially verified target proteins. To develop this compound to be a targeted drug for breast cancer, specific target identification was necessary for three reasons: (1) breast cancer comprises many receptors that relate to cancer treatment, (2) (−)-kusunokinin did not show strong binding with CSF1R and AKR1B1, and (3) synthetic (±)-kusunokinin contained two forms. Therefore, finding other target proteins from breast cancer associated proteins was the main objective of this work.
The molecular docking approach has many limitations in its representation of atomistic information for drug-protein behaviors, such as lack of temperature and dynamics behaviors, hydrogen bonding within a protein-drug complex due to the water and ionic strength of the environment [19,20,36]. MD simulation with full surroundings would provide more realistic factors, and thus the justification of binding behaviors of the compounds towards HER2 could be more consolidated. According to MD simulations, structural analysis and relative binding energy evaluation permitted the following prediction in case trans-(−)-kusunokinin and trans-(+)-kusunokinin could form direct binding with HER2: (1) HER2 preferred trans-(−)-kusunokinin to trans-(+)-kusunokinin, (2) both trans-(−)kusunokinin and trans-(+)-kusunokinin may bind HER2 differently than HER2 known inhibitors, such as neratinib or synthetic 03q, and (3) trans-(−)-kusunokinin and trans-(+)-kusunokinin exhibited significantly lower binding affinity towards HER2, considering known inhibitors such as neratinib or synthetic 03q.
The well-known HER2 inhibitor binding domain was the so-called ATP binding domain [23,28], consisting of Threonin733 to Leucine785 [37], equivalent to Threonine60 to Leucine112 in our study. Both trans-(−)-kusunokinin and trans-(+)-kusunokinin partly resided in this HER2 domain, unlike 03q and neratinib. The distinctive binding feature of synthetic HER2 inhibitors resulted from several hydrogen bonds in the domain and led to a lower binding energy than kusunokinin. Meanwhile, the trans-(−)-kusunokinin was able to apply a hydrogen bond to Lysine48 in the HER2 structure, whereas no hydrogen bond was observed from trans-(+)-kusunokinin. The reason that only the trans-(−)-form exhibited hydrogen bond formation characteristics was the different spatial orientation of both the trans-(+)-kusunokinin and trans-(−)-kusunokinin in order to fit the HER2 cavity.
The interacting amino acid residues in the ATP binding domain were hydrophobic (nonpolar) amino acids, so the hydrophobic interaction could play a role in ligand selectivity. Moreover, π-π stacking with Phe159 from the aromatic ring in kusunokinin also supported trans-(+)-kusunokinin and trans-(−)-kusunokinin bindings. The π-π stacking with Phe159 was also found in the synthetic inhibitor 03q. The synonymous binding manner of trans-(−)-kusunokinin was also previously documented in the CSF1R [19,20] and AKR1B1 [22]. In short, trans-(−)-kusunokinin and trans-(+)-kusunokinin could bind HER2 in some ATP binding regions using π-π stacking and hydrophobic interactions. Therefore, a reversible HER2 binding model for a case of trans-(−)-kusunokinin was thereby speculated as no covalent bond formation was observed. The previously reported reversible HER2 inhibitors such as lapatinib and tucatinib [38][39][40] can interact with HER2 via hydrogen bonding and π-π stacking, the same as trans-(−)-kusunokinin. Different scenarios were found in the cases of crotonamide (-C=C-C(=O)-NH-) containing inhibitors, namely neratinib and pyrotinib, which were reported for irreversible HER2 binding [23,[40][41][42][43]. The existence of the crotonamide group contributed to the irreversible HER2 binding via covalent bond formation [23,41]. The crotonamide group can react with cysteine100 (Cys100) in the ATP binding domain via Michael addition. There was a carbon-sulfur covalent bond between neratinib and Cys100 in the ATP binding domain ( Figure S2), in which the thiol group (-SH) and crotonamide acted as the Michael donor and acceptor, respectively. The mechanism of the reaction is presented in the Supplementary Materials ( Figure S2). The irreversible binding of neratinib/pyrotinib could prolong and enhance the action of HER2 inhibition, compared to reversible HER2 binding inhibitor.
Although an irreversible inhibitor like neratinib would outclass other HER2 inhibitors, since HER2 is overexpressed in 20-30% of primary tumors and associated with malignant transformation and survival rates [44,45], some problems from neratinib were noticed. Neratinib was one of the responsible factors in driving resistance to HER2-targeted therapies [24,25,46], and HER2-positive breast cancer patients usually die due to its resistance [47]. In addition, neratinib toxicity [26,27] is still a main problem. Even though the predicted efficacy of trans-(−)-kusunokinin seemed unable to compete with the irreversible inhibitor (neratinib), the drug-associated drawbacks have left room for further investigation into the anticancer effect of synthetic (±)-kusunokinin based on HER2 and HER2 related proteins in breast cancer. Thus, we decided to carry on the experiment in order to elucidate the anticancer effect of synthetic (±)-kusunokinin by targeting HER2 and its related proteins in similar pathways.
To confirm the results of our computer simulation, MCF-7 cells were treated with synthetic (±)-kusunokinin and neratinib. The achieved results of this study were consistent with the previous experiment, as described above. The synthetic (±)-kusunokinin showed less toxicity than neratinib on MCF-7 cells. In the silencing of HER2, our study confirms that neratinib strongly bound HER2 by increasing HER2 levels close to the control groups. Moreover, the combination of siRNA-HER2 and neratinib also brought the protein levels of MEK1, ERK, c-Myc, CyclinB1, CyclinD1 and CDK to normal levels. In contrast, the combination of synthetic (±)-kusunokinin and siRNA-HER2 brought only ERK levels to the normal baseline. Regarding biological assays, although trans-kusunokinin was confirmed to have cytotoxic and anti-proliferative activities, it was also shown that kusunokinin suppressed RAS and ERK but not HER2, in contrast with neratinib which is a known ligand of HER2. Therefore, kusunokinin and neratinib are likely to operate via different modes of action, suggesting that kusunokinin may affect HER2-related pathways without physically interacting with the protein, or, in the case that kusunokinin actually binds to HER2, a different binding pocket may be involved, Figure 8.

Molecular Docking
Ten proteins associated with breast cancer were chosen for investigating a possible protein target(s) of kusunokinin. The three-dimensional (3D) protein structures were retrieved from the Research Collaboratory for Structural Bioinformatics Protein Database Protein Data Bank (RCSB PDB). All co-crystallized ligands and water molecules were removed using the AutoDock Tool (ADT) version 4.1. All polar hydrogen atoms were added to mimic hydrogen bond interaction, and written into the Protein Data Bank (PDB) file format.
The 3D structures of trans-(-)-kusunokinin, trans-(+)-kusunokinin and known ligands were taken from the PubChem database. All PubChem CIDs of the ligands in this study were summarized in the Supplementary Materials, Table S1. All structure files of trans-(−)kusunokinin, trans-(+)-kusunokinin and known ligands were generated in the PDB file format using the Online SMILES Translator and Structure File Generator (https://cactus. nci.nih.gov/translate/ accessed on 21 January 2021). ADT was used to add all missing polar hydrogen atoms. Finally, both protein and ligand structures were saved in a PDB and Partial Charge (Q) and Atom Type (T), or PDBQT format file.
A molecular docking study between all compounds and the 10 selected proteins was performed using AutoDock4 version 4.2 [53]. The grid box was centered at the region covering the whole structure with the dimensions of 126 × 126 × 126 cubic angstrom (Å 3 ). Exhaustiveness was set to 50 docking runs with a population size of 200. All other parameters were set at the default value on the AutoDock4 program. The lowest binding energy (∆G bind ) was reported. The lowest energy coordinate was considered for postdocking analysis, and as the starting coordinate of molecular dynamics (MD) simulation.

Molecular Dynamics Simulation
The starting coordinate for MD simulation was adopted from the best docked conformation with the Human HER2 protein database (PDB ID 3PP0) [35] mentioned previously. First of all, the restrained electrostatic atomic partial potential (RESP) charge for all the compounds, such as trans-(+)-kusunokinin, trans-(−)-kusunokinin, synthetic 03q, and neratinib, was modeled based on the optimized geometry and electrostatic single point charges (ESP) using the B3LYP/6-31G* calculation. The B3LYP/6-31G* was implemented in the Gaussian16 package [54], kindly provided by Dr. Thanyada Rungrotemongkol (Department of Biochemistry, Faculty of Science, Chulalongkorn University, Bangkok, Thailand).
Secondly, for both ligand-free HER2 as well as a compound-HER complex structures, all polar hydrogen in the structure from the docking study was removed. All hydrogen atoms were then added again via the Leap module in the AMBER16 package [55]. All ionizable sidechains of amino acids were set as one at pH 7, namely histidine was deprotonated (HIE in AMBER name), lysine and arginine contained + 1e charge, and aspartate and glutamate contained −1e charge. The system was solvated by transferable intermolecular potential with 3 points (TIP3P) water at a distance of 14 angstrom (Å), resulting in approximately 17,400 water molecules and neutralized by either sodium (Na + ) or chloride (Cl − ) ion. The 47 NaCl pairs were applied, equivalent to the 0.15 M NaCl solution.
The MD simulation protocol was carried out similar to the previous kusunokinin study [22]. In brief, the energy minimization was completed using the Steepest Descent method for 1000 steps and the Conjugate Gradient method for 1000 steps under the periodic boundary condition. The canonical (NVT) ensemble at 310 K (37 • C), with a cut off of 16 Å, was used to handle all nonbonded/electrostatic interactions. Harmonic restraint was applied to all coordinates of the compound-protein complex with a force constant of 200, 100, 50, 25 and 10 kcal mol −1 Å −2 , respectively, with reference to the minimized structure. Each force constant lasted for 400 picoseconds (ps) with a time step of 1 femtosecond (fs), resulting in a simulation time of 2 nanoseconds (ns).
The pressure of 1.013 bar (1 atmospheric pressure) was then introduced to create the isobaric-isothermal (NPT) simulation. The temperature of 310 K and pressure were controlled using the weak-coupling algorithm [56]. All positional restraint was completely removed. The simulation was continued for 100 ns with a time step of 2 fs. The first 50-ns simulation was omitted and the 500 snapshots from the last 50-ns simulation were taken for further structural analysis and binding energy calculation.

Structural Analysis and Free Binding Energy Calculation
The structural analysis was performed via 2 parameters: the root-mean square displacement (RMSD) and the average root-mean square fluctuation (RMSF). The RMSD of the 1000 snapshots from the 100-ns-MD trajectory was illustrated to visualize the simulation stability. In addition, the RMSF of 500 snapshots from the last 50-ns MD trajectory was analyzed to investigate the sidechain flexibility in HER2 proteins, compared with ligandfree states. RMSD was obtained from the Visual Molecular Dynamics (VMD) program [57], while RMSF was calculated from the cpptraj module in the AMBER16 package.
The free binding energy of all compounds, namely trans-(+)-kusunokinin, trans-(−)-kusunokinin, synthetic 03q, and neratinib, towards HER2 was computed using the molecular mechanics/Poisson-Boltzmann surface area (MM/PBSA) approach [58]. All energetic parameters for the MM/PBSA calculation were summarized in the previous study [59]. The average binding energy (ÄG) was reported in kcal mol −1 from the last 500 MD snapshots, using the python script (MMPBSA.py) in the AMBER16 package.

Cytotoxicity Assay
Half of the maximal inhibitory concentration (IC 50 ) value was performed as previously described [60]. In brief, cells were seeded in 96 well plates and cultured for 24 h. After treatment, cells were treated at various concentrations of synthetic (±)-kusunokinin and neratinib for 72 h. Cell viability was determined using 3-(4,5-dimethyl thiazol-2 yl)-2,5diphenyltetrazolium bromide (tetrazolium salt MTT, Cat No.: M6494, Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). The absorbance was measured at 570 and 650 nM using a microplate reader (Emax ® Plus, Molecular Devices, San Jose, CA, USA). Each treatment was performed in triplicate.

In Vitro Transfection of Small Interfering RNA
HER2 gene silencing was performed using siRNA-HER2 (Invitrogen, Thermo Fisher Scientific, Cat No. AM16708A; ID 103546, Carlsbad, CA, USA). The Luciferase duplex (Photinus pyralis (American firefly) luciferase gene) (Cat No: P-002099-01-20, Dharmacon, Lafayette, CO, USA) was used as the siRNA control experiment. siRNAs were transfected using Lipofectamine ® RNAiMAX (Invitrogen, Waltham, MA, USA), according to the manufacturer's instructions, as previously described [21]. Briefly, MCF-7 cells were seeded in 24 well plates for 24 h and transfected with 100 nM siRNA-HER2 or 100 nM siRNA-Luciferase. After transfection for 6 h, the medium was removed and changed. Cells were treated with or without IC 50 concentration values of synthetic (±)-kusunokinin or neratinib for 48 h. Then, effective gene silencing was confirmed by Western blot analysis.

Dye Exclusion Assay
Neratinib (HKI-272) was obtained from Selleck Chemicals (CAS No: S2150, Houston, TX, USA). Cell viability was evaluated by a trypan blue exclusion assay, as previously described [61]. Briefly, MCF-7 cells were treated with synthetic (±)-kusunokinin, neratinib, siRNA-HER2 or a combination of compounds with siRNA-HER2. After the incubation period, floating and attached cells were harvested and stained with trypan blue dye solution at a final concentration of 0.2%. At least 150 cells were counted per treatment using phase-contrast microscopy. Assays were performed in triplicate and repeated 3 times.

Western Blot Analysis
After treatment, cells were subjected to Western blot analysis, as previously described [61]. Cells were harvested by trypsinization and cells were lysated using a RIPA buffer (Thermo Scientific, Waltham, MA, USA). According to the manufacturer's instructions, the total protein concentration was determined using the Bradford method (Bio-Rad, Hercules, CA, USA). Eighty micrograms of each protein lysate were loaded and separated on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose membrane (Millipore, Billerica, MA, USA). To prevent non-specific binding, the membranes were blocked with 5% non-fat dry milk in TBST (0.1% Tween 20, 154 mM NaCl, 48 mM Tris-base, pH 6.8). The membranes were probed with primary antibodies including anti-HER2, c-Myc, CyclinD1, Ras, AKT, MEK1 (Cell Signaling Technology, Danvers, MA, USA), CDK1, CyclinB1, ERK (Santa Cruz Biotechnology, Dallas, TX, USA) and GAPDH antibodies (Calbiochem, Darmstadt, Germany). The protein signal was visualized using the SuperSignal™ West Dura Extended Duration substrate kit (Thermo Scientific, Waltham, MA, USA), according to the protocol supplied with the kit. Subsequently, immunoblot images were quantified using the ImageJ program (NIH Image, Bethesda, MD, USA).

Statistical Analysis
Data values of 3 independent experiments are expressed as the mean ± standard deviation (SD). The student's t-test on Microsoft Excel was used for statistical analysis. A p-value of less than 0.05 was considered to indicate a statistically significant difference between groups.
Supplementary Materials: The following are available online, Table S1: PDB identification code of ten breast cancer associated protein and Pubchem CID of selected known inhibitors; Figure S1: Root mean square distance of HER2 backbone atoms. All MD simulations were carried out under canonical (NVT) ensemble in which all protein atoms of HER2 were restrained using harmonic potential. The MD simulation consisted of 200 equidistant snapshots from 1000 ps. The root mean square distance was in a unit of angstrom (Å). Figure   Sample Availability: Sample of synthetic trans-(±)-kusunokinin is available from the authors.