Design, synthesis, and biological evaluation of chalcones for anticancer properties targeting glycogen synthase kinase 3 beta

Chalcones compounds have been investigated to exhibit anticancer activity through various physiological modes of action. In order to develop chalcone compounds with novel anticancer-related modes of action, diverse chalcone compounds were designed and synthesized. Variously substituted poly-methoxy chalcone compounds 1–17 were prepared, and their structures were identified using high-resolution mass spectrometry (HR/MS) and nuclear magnetic resonance (NMR) experiments. Long-term survival clonogenic assay was applied to evaluate their anti-cancer abilities and revealed that their GI50 values ranged between 1.33 and 172.20 μM. When MCF-7SC cells were treated with various concentrations of compound 14, reduced cell viability and induced apoptosis in MCF-7SC cells were observed in a dose-dependent manner. Wound healing assay demonstrated that compound 14 prevented the MCF7-SC migrated cells at non-lethal concentrations after 12 and 24 h of exposure. The efficiency of compound 14 on the levels of Epithelial-mesenchymal transition (EMT) markers was accessed by the western blot analysis. For the concrete understanding of anticancer properties at the molecular level, in vitro kinase assays on 12 cancer related proteins were carried out. Glycogen synthase kinase 3 beta (GSK3β) was most effectively inhibited by compound 14 with 89% inhibitory activity at 10 µM against GSK3β. The binding mode of compound 14 with GSK3β was reinforced through in silico experiments, which demonstrated compound 14 binds with GSK3β at binding affinity ranged between − 7.5 kcal/mol and − 6.8 kcal/mol. SwissADME analysis provided the druggability and leadlikeness of compound 14, which unveiled drug development possibilities of chalcone compound 14.

Breast cancer is recognized as the most common type of cancer for women worldwide, and it is estimated that by 2040, breast cancer cases will increase by more than 46% [11]. Although many therapeutic strategies have been established for the treatment of breast cancer, including surgery, chemotherapy, radiation therapy, and hormone therapy, however, these existing therapies have some limitations [12]. The main causes of breast cancer treatment failure, such as resistance to chemotherapy and radiotherapy, recurrence, metastasis, and weak immune monitoring, can be explained by the characteristics of cancer stem cells (CSC) [13,14]. Therefore, CSC has become a promising therapeutic target in the treatment of breast cancer [15,16]. There are several reports to identify CSC signatures, such as CD34 + /CD38 − , CD44 − / CD24 + population, the enrichment of abnormal signaling like Wnt/β-catenin, NF-ĸB, or Notch pathways, promoting metastasis and tumor progression [17]. Furthermore, members of the ATP binding cassette (ABC) transporter subfamily, such as multidrug resistance protein 1 (ABCB1/MDR1) or multidrug resistance-associated protein 1 (ABCC1/MRP1) are overexpressed in different cancers, driving chemotherapy resistance and CSC biological function [18]. The stem-like MCF-7/SC cells, which is isolated from human breast carcinoma MCF-7 cells, possessed an up-regulated CD44 + /CD24 population, overexpressed drug efflux protein MRP1, and possessed features such as increased migration capacity [19]. The anticancer activity of the compound was investigated by measuring the effects on cell proliferation, apoptosis induction, epithelial-mesenchymal transition, and stem cell potential of MCF-7SC cells, which have higher stem cell characteristics compared to the parental MCF-7 cells.
To understand the results obtained from the biological experiments mentioned above at the molecular level, in vitro kinase assays were carried out. Of several kinases, glycogen synthase kinase 3 beta (GSK3β) was most effectively inhibited by compound 14. To elucidate whether compound 14 binds to GSK3β, in silico docking experiments were performed. The results obtained in this study demonstrated the compound 14 can be developed as a chemotherapeutic agent.

Synthesis
Variously substituted poly-methoxy acetophenones (I) and substituted benzaldehydes (II) were used to obtain poly-methoxy substituted chalcone derivatives 1-17 as shown in Scheme 1. Typical procedure for preparation of compound 9 was described as an example.

Spectroscopic methods
The structures of the chalcone compounds synthesized here were determined using NMR spectroscopy and HR/ MS spectrometry. The concentrations of the compounds dissolved in deuterated dimethyl sulfoxide were adjusted to approximately 50 mM. They were transferred into 2.5 mm NMR tubes to generate good spectra using short acquisition times. All NMR experiments were carried out on a Bruker AVANCE 400 spectrometer (9.4 Tesla) at room temperature. The detailed NMR experimental method followed the previous report [20]. HR/MS data were collected according to a previously published method in [M + H] + mode using Waters ACQUITY UPLC (Waters, Milford, MA, USA) [21].

Biological experiments Clonogenic assay
The clonogenic assay was used to access the long-term survival of cells as previously described [22]. Briefly, cells were seeded on 24-well plates and treated with increasing concentrations of different compounds. After seven days of incubation, colonies were fixed with 6% glutaraldehyde, followed by staining with 0.1% crystal violet. Densitometry (MultiGuage, Fujifilm, Japan) was used to determine the cell growth inhibitory dosage values, and the SigmaPlot tool was used to calculate GI 50 values.

MTT assay
The MTT assay was used to access the cell viability. Briefly, cells (2 × 104 cells/mL) were seeded into 96-well plates. After that, cells were treated with increasing concentration of compound 14 following 24 or 48 h of incubation. MTT reagent (1 mg/mL) was added to each well (ratio 1:4 with media) and incubated for 2-3 h at 37 °C. Then, 100 µL of Dimethyl sulfoxide (DMSO) was added to solubilize formazan crystals. The absorbance was recorded at 570 nm by a micro-plate reader (Tecan Group, Ltd., Salzburg, Austria).

Flow cytometric assay for Annexin V/PI staining
Cells (1 × 10 5 cells/mL) were seeded onto 60 mm plates and then treated with compound 14 at different doses. After 24 h of incubation, cells were washed with PBS and stained with an Annexin V-FITC Apoptosis Detection Kit (BD Pharmingen, San Diego, CA, USA) for 15 min at 37 °C to detect apoptotic population. FACSCalibur flow cytometer was used to access apoptotic cells.

Wound healing assay
Cells were cultured in 6-well cell culture plates and allowed to grow until cell density reached 95% confluence. A uniform scratch was made in each well by a 10 μL sterile pipette tip and then, floating cells were washed with PBS. After that, cells were exposed to compound 14 at non-lethal dosage for 24 h. The widths of scratches were quantified under a phase contrast microscope (magnification at 100x).

Western blotting analysis
Cells were harvested by the radioimmunoprecipitation assay buffer (RIPA) buffer comprising protease inhibitor cocktail (PIC) and Phenylmethylsulfonyl fluoride-PMSF (Sigma-Aldrich, Missouri, USA). Protein concentration was quantified by the bicinchoninic acid assay (BCA) kit (Thermo Fisher Scientific, Waltham, MA, USA). After quantification, lysate was separated using sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). Samples were then electrophoretic transferred to PVDF membranes, followed by blocking with 5% non-fat milk overnight. Primary antibodies were purchased from Cell Signaling Technology, Beverly, MA, USA, except the mouse anti-E-cadherin and β-catenin antibody (BD Transduction Laboratories, San Jose, CA, USA). The secondary antibodies (Vector Laboratories, Burlingame, CA, USA) were diluted at 1:5000 following the manufacturer's recommendations.

In vitro kinase assay
Kinase assays were performed according to a previously published method [22].

In silico docking
In silico docking experiments were carried out using AutoDock Vina (The Scripps Research Institute, La Jolla, San Diego, USA) [23]. Dozens of structures of GSK3β were deposited in the protein data bank. Of them, 5f95. pdb was selected for in silico docking experiments because it contains 2-[(cyclopropylcarbonyl)amino]-N-(4-phenylpyridin-3-yl)pyridine-4-carboxamide including biphenyl group as a ligand [24]. The organism of 5f95. pdb was Homo sapiens. The protein was overexpressed on baculovirus system. Its 3D structure was determined by X-ray crystallography. The apo-protein was prepared using the Chimera program [25]. The binding site was determined using the grid box provided by AutoDock Tools. The 3D structures of chalcone compounds were determined based on the X-ray crystallographic structure obtained from the previous work [26].

Statistical analysis
GraphPad Prism 8.0 software was used for statistical analysis in this present study. Data are presented as the mean ± standard deviation (SD) of three independent experiments. One way analysis of variance (Oneway ANOVA) with Dunnett's post hoc test was used for accessing the differences between groups. In all analyses, * p < 0.05 was considered to indicate a statistically significant.

Results and discussion
1 H NMR, 13 C NMR and 2D NMR were used to identify the chemical structure of the synthesized chalcone compounds. The complete assignments of chemical shift values, multiplicity, and coupling constants for each hydrogen present in each compound are summarized in Table 1. All synthesized chalcone compounds contain a conjugated enone group at the center of the molecule. In a C=C double bond, the carbon near the carbonyl C=O is assigned an α position and the carbon farther away is called the β position. For all compounds, the values of coupling constants between α hydrogen and β hydrogen (J Hα-Hβ ) range from 15.3 to 16 ppm, indicating that the carbon-carbon double bond is in the trans configuration. To measure anticancer effects of chalcone compounds, cytotoxicity assays in cancer cell lines have been used. Here, a long-term survival clonogenic assay was used because it can distinguish activities caused by small structural differences. In order to obtain the half-maximal cell growth inhibitory concentration (GI50) values, different concentrations of compounds (0, 10, 20, and 40 uM) were treated in cancer cells. In the case of compounds 13 and 14, which have good cell growth inhibitory effects, efficacy was confirmed at a low concentration of 10 uM or less. The GI 50 data are shown in Fig. 2.         Table 3. The GI 50 values with error bars were plotted against chalcone compounds (Additional file 1: Fig. S1).
To characterize why the compounds exhibit cell growth inhibitory effects against cancer cells, further biological experiments were performed. Because compound 14 showed the best GI 50 value, it was selected for further biological experiments. Compound 14 showed a remarkable capacity on the cell viability inhibition of MCF-7SC after 24 and 48 h of treatment (Fig. 3A). Therefore, compound 14 was chosen to investigate the anti-cancer effects on stem like MCF-7SC cells in further experiments. Notably, Fig. 3B demonstrated that compound 14 could not cause cytotoxicity on fibroblast cells.
In addition, we checked whether compound 14 could induce apoptosis in MCF-7SC. Notably, compound 14 enhanced the proportion of the apoptotic population in MCF-7SC by 6.47% ± 1.21% at a dose of 10 μM, as shown by annexin V/propidium iodide staining (Fig. 4A,  B). Furthermore, compound 14-treated MCF-7SC cells displayed a decrease in the levels of caspase 7 and Bcl-2 while showing an increase of c-PARP, c-caspase 7, and Bax (Fig. 4C, D). Taking together, compound 14 exhibited the inhibitory potential on proliferation and apoptosis promotion of stem-like cell MCF-7SC.
MCF-7SC cells were previously reported to display greater stem cell characteristics and migratory potential compared to parental MCF-7 cells [19]. We next tested the effect of compound 14 on the migration of MCF-7SC cells. Figure 5A, B showed that compound 14 could prevent the MCF-7SC migrated cells at non-lethal concentrations after 12 and 24 h of exposure, as represented by wound healing assay. Epithelial-mesenchymal transition (EMT) has been involved in the stemness characteristic of CSCs. By losing adherent junctions and breaking down cell-cell interaction, EMT supported the metastatic initiating potential to further site in tumor cells [27]. Therefore, the efficiency of compound 14 on the levels of EMT markers was accessed by the western blot experiment. Markedly, compound 14 suppressed the levels of wellknown mesenchymal markers such as Snail and Vimentin while increasing the level of an epithelial marker, E-cadherin dose-dependently, as shown in Fig. 5C, D.
CD44 + /CD24-population is considered as an important signature for breast CSC identification [28]. Interestingly, a dramatic decrease of this population was observed in MCF-7SC at dose 10 µM by performing FACS analysis (77.94% ± 4.51%) (Fig. 6A, B). Moreover, the levels of common stemness markers such as CD44, MRP1, and β-catenin were reduced after compound 14 treatment for 24 h, by 0.36 ± 0.16, 0.72 ± 0.11, and 0.52 ± 0.07-fold, respectively (Fig. 6C, D). Abnormal activation of the transcriptional factor β-catenin, which is a  Table 3 The half-maximal cell growth inhibitory concentrations (GI 50 ) and mass spectra data of compounds 1-17. All HR/MS data were collected in [M + H] + mode crucial member of the Wnt signaling pathway, drives the pluripotency, self-renewal, and differentiation ability in CSCs, indicating the key pathway inhibitor as an attractive therapeutic target in cancer treatment [29]. Collectively, these results indicate that compound 14 could attenuate the stemness features of MCF-7SC cells.
To understand the results mentioned above at the molecular level, in vitro kinase assays were carried out using compound 14. Several kinases known as targets for chemotherapy were selected: AMP-activated protein kinase, aurora kinase A, aurora kinase B, serine/ threonine-protein kinase B-Raf, epidermal growth factor receptor, glycogen synthase kinase 3 beta, kinase insert domain receptor, mitogen-activated protein kinase, mammalian target of rapamycin, protein kinase A, and phosphoinositide 3-kinase. Compound 14 demonstrated 89% inhibitory activity at 10 µM against glycogen synthase kinase 3 beta (GSK3β).
To elucidate whether compound 14 binds to GSK3β, in silico docking experiments were performed. As a 3D structure of GSK3β, the X-ray crystallographic structure deposited in the protein data bank as 5f95.pdb was used. It consists of two polypeptide chains, A and B. GSK3β is composed of 420 amino acids. Since chain A includes residues 36-383, and chain B does residues 36-385, chain B was selected for in silico docking. To prepare apoprotein a ligand, 2-[(cyclopropylcarbonyl)amino]-N-(4phenylpyridin-3-yl)pyridine-4-carboxamide (named as 3up) contained in 5f95.pdb was deleted using the Chimera program. The binding pocket was determined using the grid box provided by AutoDock Tools as follows: the centers of x, y, and z were 55.361, 17.528, and 167.194, respectively, and the sizes of x, y, and z were 18, 20, and 20, respectively. To confirm whether this binding pocket is correct, the ligand, 3up contained in 5f95. pdb was docked into apo-protein using the AutoDock vina program. Because the AutoDock program provides nine iteration docking process, nine complexes of GSK3β and 3up were generated. Their binding affinity ranged between -8.5 kcal/mol and − 7.6 kcal/mol. Their docking poses were superposed with the X-ray crystallographic structure well. Next, compound 14 was tried to be docked into GSK3β apo-protein. Like the original ligand, 3up, nine complexes between GSK3β and compound 14 were generated. Their binding affinity ranged from − 7.5 to − 6.8 kcal/mol. Even their values were worse than those of the original ligand, as shown in Fig. 7A, B, the binding pose of compound 14 was superposed in a similar way.
The interactions between the residues of GSK3β and 3up (Fig. 8A) were compared with those between the residues of GSK3β and compound 14 (Fig. 8B) obtained using the LigPlot program. Eight residues were observed in both 3up and compound 14: Ala83, Lys85, Leu132, Asp133, Tyr134, Val135, Leu188, and Asp200 where Asp200 shows a hydrogen bond with 3up but it does a hydrophobic interaction with compound 14. While three residues including Ile62, Phe67, Val70 were observed in the GSK3β-compound 14 complex only, five residues including Pro136, Thr138, Arg141, Gln185, and Cys199 were observed in the GSK3β-3up complex. As a result, 13 residues participated in the interaction between GSK3β and 3up, but 11 residues did in compound 14.
Different residues in the binding pockets of 3up and compound 14 may cause different binding poses as shown in Additional file 1: Fig. S2A, B. In addition, the original ligand, 3up made a hydrogen bond with the protein. These results demonstrate that the binding affinity of the original ligand should be lower than that of compound 14.
The druggability of compound 14 was evaluated using the SwissADME web tool provided by the Swiss Institute The anticancer-related mode of action was investigated through in vitro kinase assays, and it was confirmed that compound 14 inhibited the GSK3β enzyme. The binding mode of compound 14 to GSK3β was verified through in silico experiments, demonstrating that compound 14 binds to GSK3β with a binding affinity ranging from -7.5 kcal/mol to -6.8 kcal/mol.