Phytochemicals Derived from Goniothalamus elegans Ast Exhibit Anticancer Activity by Inhibiting Epidermal Growth Factor Receptor

Abstract

Cancer is a major health burden and a leading cause of death worldwide, with numerous new molecules being studied and developed as therapeutic agents. In this study, the cytotoxicity of compounds derived from Goniothalamus elegans was evaluated for possible anticancer activity. It was observed that the crude methanol extract of G. elegans exerted the strongest cytotoxic activity against SW-480, AGS, and SK-LU-1 cell lines. In addition, two isolated alkaloids—namely, lysicamine and liriodenine—also showed strong inhibitory ability against similar cancer cell lines. To further investigate the compounds’ mechanism of action, a molecular docking approach was utilized to evaluate the potential of the two candidates to interact with the epidermal growth factor receptor. This assay estimated that lysicamine and liriodenine acquired protein binding affinities of −8.8 and −9.7 kcal/mol, respectively. Finally, the stabilities of the ligand–protein complexes were evaluated using molecular dynamics simulations of 100 ns each.

Introduction

Cancer is a fatal illness characterized by uncontrolled cell division and cell immortalization. Cancer cells can be aggressive and invasive, spreading to different organs. According to the American Cancer Society, approximately two million new cancer cases and 600,000 cancer-related deaths were reported in the United States in 2021, the majority of which were caused by prostate, breast, colorectal, and lung cancers. ¹ In recent decades, early cancer detection, along with the discovery and development of new technologies and drugs, has revolutionized cancer treatment. However, effective treatments for cancer are still under investigation. Carpenter and Cohen first described epidermal growth factor (EGF), EGF receptors (EGFRs), and many of their biochemical properties. ² EGFRs were the first receptors for which overexpression and mutation correlated with cancer development, with many studies demonstrating important roles for these proteins in the signal transduction and pathogenesis of various malignancies such as pancreatic, breast, stomach, and lung carcinomas.^3,4 EGFRs, also known as ErbB receptors, are members of the receptor tyrosine kinase subclass I protein family. ⁵ Given the relationship between EGFR and malignancies, the development of small-molecule tyrosine kinase inhibitors (TKIs) has revolutionized modern cancer therapy. For example, lapatinib and neratinib, which target the ErbB2 receptor, are Food and Drug Administration (FDA) approved for the treatment of breast cancer, whereas afatinib, dacomitinib, erlotinib, gefitinib, and osimertinib are EGFR/ErbB1 inhibitors used for the treatment of nonsmall cell lung cancer.^6,7 TKIs block vital pathways by targeting signaling molecules that promote cell survival. After translocation through the plasma membrane, TKIs interact with the cytoplasmic domains of tyrosine kinase receptors and inhibit their catalytic activity by interfering with adenosine triphosphate (AT)P or substrate binding. ⁸

Natural products constitute a substantial repository of potential cancer treatments.^9–11 Hence, given the effectiveness and availability of natural products, these resources are regarded as potential antineoplastic candidates. Drug development is a time-consuming and challenging process that begins with target identification and continues through progressive optimization experiments and pre-clinical studies. In addition, virtual screening analyses have facilitated the application of computational methods to pharmacological candidate discovery.^12–15 Indeed, it is well-accepted that the combination of computational and experimental approaches can improve drug development success rates. Furthermore, the development of functional foods derived from medicinal plants is a prominent area of research. Goniothalamus, a diverse genus in the Annonaceae family that contains 160 species, is partly native to tropical and subtropical regions in Southern Asia. In traditional medicine, several Goniothalamus species are used for a variety of ailments such as skin diseases, fever, typhoid, and stomachache.^16,17 Moreover, modern pharmacological approaches have demonstrated that certain extracts and compounds isolated from species of Goniothalamus possess beneficial biological properties such as antibacterial, antimalarial, antioxidant, and especially, cytotoxic activities. ¹⁸ Goniothalamus elegans Ast is primarily found in Southeast Asian countries, including Vietnam. In Vietnamese traditional medicine, G. elegans is widely used to treat heart diseases and bloody diarrhea.^16,17 Phytochemical studies of the species are limited; only 15 compounds have been isolated, most of which are styryl lactones and aristolactam alkaloids. In addition, isolated compounds from G. elegans have shown strong cytotoxicity, as well as antimalarial and antibacterial properties. ¹⁹ Specifically, our previous work reported that extracts, fractions, and isolated compounds from Vietnamese G. elegans were cytotoxic to human breast cancer cell line MCF-7, potentially due to their ability to inhibit the breast cancer genes BRCA-1 and BRCA-2, and estrogen receptor alpha. ²⁰ Goniothalamus species are often utilized in traditional medicine. However, few studies have investigated the antineoplastic properties of G. elegans-derived compounds in other cancer types. In the present study, the cytotoxic properties of crude extracts, fractions, and compounds isolated from G. elegans were tested on human colon adenocarcinoma SW-48, human gastric adenocarcinoma AGS, and human lung adenocarcinoma SK-LU-1 cell lines. Subsequently, a molecular docking method was used to investigate the potential ability of these compounds to inhibit EGFR. Finally, the stabilities of the ligand–protein complexes were evaluated with molecular dynamics simulations.

Results

Structure Determination of the Isolated Compound

The methanolic extract of the aerial parts of G. elegans was partitioned into n-hexane, dichloromethane, ethyl acetate, and aqueous fractions. Column chromatographic purification of the ethyl acetate fraction was repeated to isolate individual compounds. Finally, we performed nuclear magnetic resonance (NMR) analysis on the isolated compounds to confirm their structure, using reports in the literature as guides.

9-Deoxygoniopypyrone (1), trans-cinnamic acid (2), lysicamine ²⁰ (3), and lirodinine ²⁰ (4) were isolated and characterized from their NMR spectroscopic data, which matched the values in the literature^20–23 (Supplemental material). The structures of the compounds are shown in Figure 1.

Figure 1.

Structures of compounds isolated from Goniothalamus elegans.

In Vitro Cytotoxicity Assay

In vitro experiments were used to determine the cytotoxic effects of G. elegans-derived extracts on three cancer cell lines (Table 1). The crude methanol extract had the highest cytotoxic activity against SW-480, AGS, and SK-LU-1 cell lines with inhibitory concentration (IC₅₀) values of 14.51 ± 0.96, 19.57 ± 1.25, and 24.69 ± 2.83 μg/mL, respectively. Additionally, the ethyl acetate and dichloromethane fractions exhibited moderate inhibitory effects against the cell lines tested, with IC₅₀ values ranging from 38.03  ±  1.88 to 54.46  ±  2.51 and from 48.30 ± 1.91 to 73.42 ± 4.79 μg/mL, respectively. In contrast, the n-hexane and aqueous fractions did not exhibit cytotoxicity against any of the cancer cell lines examined (IC₅₀ values more than 100.0 μg/mL). The chemotherapeutic drug ellipticine was used as a positive control. Accordingly, ellipticine exhibited strong cytotoxic effects against all of the cancer cell lines tested, with IC₅₀ values ranging from 0.33  ±  0.04 to 0.38 ±  0.04 μg/mL. Since the ethyl acetate fraction demonstrated the strongest cytotoxicity, this fraction was subjected to column chromatography; four compounds were isolated, 9-deoxygoniopypyrone, trans-cinnamic acid, lysicamine, and liriodenine. ²⁰ Subsequently, the cytotoxic effects of these compounds were investigated (Table 1). Lysicamine and liriodenine showed strong inhibitory abilities against SW-480, AGS, and SK-LU-1 cell lines, with IC₅₀ values ranging from 13.91 to 31.72, and from 9.84 to 16.98 μg/mL, respectively.

Table 1.

Cytotoxic Activities of Crude and Fractionated Extracts of Goniothalamus Elegans Against Human Cancer Cell Lines.

Extract/compound	IC50 (μg/mL)
	SW-480	AGS	SK-LU-1
Crude methanol	14.51 ± 0.96	19.57 ± 1.25	24.69 ± 2.83
n-Hexane	>100	>100	>100
Dichloromethane	48.30 ± 1.91	55.44 ± 5.04	73.42 ± 4.79
Ethyl acetate	38.03 ± 1.88	43.80 ± 3.81	54.46 ± 2.51
Aqueous	>100	>100	>100
9-Deoxygoniopypyrone (1)	60.43 ± 4.36	60.37 ± 3.32	61.25 ± 5.96
Trans-cinnamic acid (2)	>100	>100	>100
Lysicamine (3)	23.84 ± 1.83	31.72 ± 1.61	24.83 ± 2.43
Liriodenine (4)	13.91 ± 1.18	16.98 ± 1.05	9.84 ± 0.87
Ellipticineb	0.34 ± 0.05	0.38 ± 0.03	0.38 ± 0.04

^a

IC₅₀ (concentration that inhibits 50% of cell growth).

^b

Positive control.

Molecular Docking Analysis

Our cytotoxicity experiments showed that liriodenine and lysicamine can inhibit the proliferation of different cancer cell lines. Therefore, we performed molecular mechanism studies of these compounds to further improve their efficacy. EGFR is a cell-surface receptor associated with a wide array of cancers, including breast, anal, ovarian, and lung cancers. Thus, we hypothesized that liriodenine and lysicamine inhibit cancer cells by binding to EGFR. The three-dimensional (3D) structures of these compounds were downloaded from the PubChem database and docked to the active site of EGFR. To validate the docking protocol and protein structure (PDB ID: 1XKK), we re-docked the ligand in the EGFR crystal structure and analyzed the root mean square deviation (RMSD) of the docking pose against the crystal pose. This analysis demonstrated that the docking pose recapitulated the conformation and orientation of the crystal pose, with an RMSD of 0.462 Å (Figure 2). Previous studies have considered RMSDs within 2 Å to be credible for docking experiments, ²⁴ thereby validating the docking protocol of the present study.

Figure 2.

Docking pose (yellow) and crystal pose (green) of the ligand in the crystal structures of epidermal growth factor receptor.

In addition, osimertinib is a commercial cancer drug known to inhibit EGFR. Therefore, we selected osimertinib as a reference inhibitor. The results of our docking study revealed that liriodenine and lysicamine achieved better docking scores than osimertinib (−8.0 kcal/mol) (Table 2).

Table 2.

Docking Results of Candidates Toward EGFR Protein.

Number	Compound	Docking score (kcal/mol)	Hydrogen bond	Hydrophobic interaction
1	Liriodenine	−9.7	Lys745, Cys797	Leu718, Ala743, Leu844, Val726
2	Lysicamine	−8.8		Leu718, Leu844, Ala743, Lys745, Val726
3	Osimertinib	−8.0	Gln791, Leu788, Asp855, Gly796, Leu718	Ala743, Val276, Met793, Leu792, Leu844, Arg841, Cys797, Met1002, Asp800, Leu718

According to the results of our docking analysis, liriodenine interacted with EGFR via hydrogen bonds at Lys745 and Leu792, and hydrophobic interactions at Leu718, Ala743, Leu844, and Val726 (Figure 3a). Similarly, lysicamine interacted with EGFR via hydrophobic interactions at five amino acids, including Leu718, Leu844, Ala743, Lys745, and Val726 (Figure 3b). Structurally, liriodenine and lysicamine are alkaloids that contain an oxoaporphine scaffold. The oxoaporphine alkaloid family has been used to treat a variety of diseases,^25,26 including Alzheimer’s disease and several types of cancer, and has been investigated in mouse models of depression and Leishmania infection.^27,28 We discovered that certain rings in the polyphenol structure of liriodenine and lysicamine interact with important amino acid residues of EGFR. In addition, carbonyl and hydroxyl groups interacted strongly with other amino acid residues. Interestingly, the chemical structures of both compounds contain a heterocyclic nitrogen molecule that is expected to bind strongly to amino acids in the active site of protein. Moreover, this heterocyclic nitrogen atom is reminiscent of the chemical structure of osimertinib (Figure 4). In previous studies, osimertinib strongly inhibited EGFR. ²⁹ Osimertinib binds with weaker affinity to EGFR than do liriodenine and lysicamine. However, osimertinib exhibited a stronger inhibitory ability than these two compounds in an in vitro assay. This result may indicate that the binding score is calculated based on the average number of atoms present in a molecule, and that osimertinib has a larger molecular mass and a more complex structure. In addition, the formation of hydrogen bonds between EGFR and osimertinib may explain the inhibitory activity of osimertinib. Specifically, we observed that liriodenine, lysicamine, and osimertinib interact with EGFR at similar amino acid residues. Notably, liriodenine interacted with a key residue for EGFR activity, Cys797, via a hydrogen bond.

Figure 3.

(a) Interaction between liriodenine and EGFR. (b) Interaction between lysicamine and EGFR. Abbreviation: EGFR, epidermal growth factor receptor.

Figure 4.

Interaction between osimertinib and epidermal growth factor receptor.

Molecular Dynamics Analysis

Small molecules interacting with the surfaces of proteins can elicit large changes in tertiary protein structure, which can ultimately lead to improvements in drug design. The primary advantage of molecular dynamics modeling is its ability to identify the flexibility of protein-ligand complexes. Indeed, this method has been previously used to evaluate the thermodynamics and kinetics of drug-enzyme binding. Accordingly, the 100 ns trajectories of the liriodenine and lysicamine complexes were analyzed to evaluate their stability.

In general, the RMSD of the liriodenine–EGFR complex indicated prolonged stability of both the protein and the ligand in the 100 ns period (Figure 5a). After approximately 20 ns, the system reached a stable state and maintained its pose throughout the simulation time. The overall final RMSD of 0.3 nm was within the acceptable upper threshold of 0.4 nm, confirming the validity and stability of the complex in the 100 ns simulation. The root mean square fluctuation (RMSF) is an indicator of protein residue flexibility during ligand interactions (Figure 5b). A residue with an RMSF value of <0.2 nm is generally considered as a stable amino acid. Overall, the liriodenine binding site (residues ∼700-850) remained stable in the presence of the ligand. The radius of gyration (R_g) is a measure of a protein’s volume change and compactness (Figure 5c). Moreover, the solvent-accessible surface area (SASA) helps to visualize a protein’s conformational change after binding to a ligand. Both of these values fluctuated modestly with amplitudes of 0.05 nm (for R_g) and 15 nm² (for SASA) (Figure 5d), thus confirming the stability of the protein and the complex.

Figure 5.

(a) RMSD, (b) RMSF, (c) radius of gyration, and (d) SASA of the liriodenine–EGFR complex. Abbreviations: EGFR, epidermal growth factor receptor; RMSD, root mean square deviation; RMSF, root mean square fluctuation; SASA, solvent-accessible surface area.

Overall, similar results were obtained for the lysicamine–EGFR complex. The RMSD of the lysicamine complex, approximately 0.25 nm, was lower than that of the liriodenine complex (Figure 6a). The system also acquired a stable state much more rapidly—during the first 2 ns—and remained stable for 100 ns. The RMSF for most residues fell within the accepted upper threshold of 0.2 nm, and the binding site region (residues 700-850) saw even less fluctuation than the liriodenine complex (Figure 6b). In addition, the low R_g and SASA values were similar to those of the liriodenine complex, once again proving the stability of EGFR over the course of our simulations (Figure 6c and d).

Figure 6.

(a) RMSD, (b) RMSF, (c) radius of gyration, and (d) SASA of the lysicamine–EGFR complex. Abbreviations: EGFR, epidermal growth factor receptor; RMSD, root mean square deviation; RMSF, root mean square fluctuation; SASA, solvent-accessible surface area.

Contact Frequency

The residue Cys797 plays an important role in regulating tyrosine kinase activity. ³⁰ Notably, this key EGFR residue interacted with both liriodenine and lysicamine with high-frequency contacts during the simulation period (Figure 7). In addition, residues including Leu718, Ala743, and Leu844 also contributed high-frequency contacts with both candidates during our simulations. These results suggest a role for these residues in regulating protein activity in biological pathways. However, further mutational experiments are needed to confirm these findings.

Figure 7.

(a) Contact frequency at the liriodenine docking site. (b) Contact frequency at the lysicamine docking site.

Binding Free Energy (MM/PBSA Calculations)

At the final stage of simulation, molecular mechanics Poisson–Boltzmann surface area (MM/GBSA) analysis of the complexes revealed that the binding free energies of liriodenine and lysicamine were −92.908 and −91.073 kJ/mol, respectively (Supplemental material). In particular, the binding free energy of the liriodenine complex ranged from −80 to −120 kJ/mol after the first 20 ns (the time at which the complex reached a stable state). The binding free energy of the lysicamine complex ranged from −60 to −120 kJ/mol, further explaining the potency of the G. elegans extract on cancer cell lines. This value was similar to that of the liriodenine complex; however, the liriodenine complex achieved more negative energy in the majority of the time frames. Overall, the molecular dynamics analyses of both ligands illustrate the potency of these two active compounds with respect to EGFR inhibition and explain the inhibitory activities of the G. elegans extract.

Discussion

Previous studies have shown that medicinal plants in the Goniothalamus genus possess antineoplastic activity. For example, dichloromethane extracts from the roots and stems of Goniothalamus macrophyllus were cytotoxic to cancer cells. ³¹ Sharma et al also studied the ability of fractionated extracts of Goniothalamus wynaadensis to inhibit the proliferation of human lung adenocarcinoma A-549 and human breast adenocarcinoma MDA-MB-231 cell lines. The results demonstrated that the n-hexane extract had strong efficacy on both cell lines. ³² Still other studies have shown that Goniothalamus gitingensis and Goniothalamus macrocalyx have antiproliferative activity against the lymphoblast K-562, HeLa, and human carcinoma KB cell lines. ³³ G. elegans is a plant commonly found in Vietnam and other Southeast Asian countries. However, phytochemical studies of the species are limited. Suchaichit et al ¹⁹ successfully isolated 15 compounds from G. elegans and evaluated their toxicity against the KB, MCF-7, and NCI-H187 cell lines, as well as strains of Mycobacterium tuberculosis and Plasmodium falciparum. Moreover, our previous work reported the cytotoxic activities of Vietnamese G. elegans extract, fractions, and isolated compounds against MCF-7 breast cancer cells. In this prior study, we predicted that the mechanisms of action of some compounds involved the inhibition of BRCA-1, BRCA-2, and estrogen receptor alpha. ²⁰ In the present report, the antiproliferative activity of G. elegans was further tested on SW-48, AGS, and SK-LU-1 cell lines, and evaluated for its ability to inhibit EGFR. The methanol extract showed strong cytotoxicity against SW-480 and AGS cells with IC₅₀ values < 20 μg/mL. According to the standards of the National Cancer Institute (NCI), the crude extracts and fractions from G. elegans showed remarkable inhibitory efficacy on different cancer cell lines. ³⁴

Two isolated alkaloids—namely, lysicamine and liriodenine—are oxoaporphine alkaloids in which positions 1 and 2 have been substituted for methoxy and dioxymethylene groups. ³⁵ Previous work has demonstrated that lysicamine strongly inhibits the growth of MCF-7 breast cancer cells and HCT-116 colon cancer cells. ³⁶ Unlike lysicamine, the chemical structure of liriodenine contains a methylenedioxy group. In previous studies, chemical derivatives containing methylenedioxy groups showed cytotoxicity toward human cancer cells. ³⁷ Combined with the EGFR interaction simulation results from the present study, these data show that liriodenine inhibits the growth of cancer cells more strongly than lysicamine. Specifically, liriodenine inhibited the growth of cancer cells such as human carcinoma KB, human lung adenocarcinoma A-549, human colon carcinoma HCT-8, mouse leukaemia P-388, and mouse lymphocytic leukemia L-1210 cells with IC₅₀ values of 3.6, 2.6, 2.5, 2.1, and 8.5 μM, respectively. It also inhibited the growth of MCF-7, NCI-H460, and nerve SF-268 cells with IC₅₀ values of 3.19, 2.38, and 2.19 μg/mL, respectively.^38,39 The cytotoxic activity of liriodenine can be explained by three main mechanisms. First, it inhibits the enzyme topoisomerase II, which plays an important role in unwinding DNA. This suppression leads to inhibition of DNA synthesis during the G1 phase of the cell cycle, indirectly causing cell death. ⁴⁰ In addition, liriodenine also inhibits the G2/M phase of the lung cancer cell cycle by enabling cyclin B1 accumulation and inhibiting the activity of the cyclin B1/cyclin-dependent kinase 1 complex. Furthermore, liriodenine induces apoptotic cell death. ⁴¹ Specifically, liriodenine inhibited the proliferation of CAOV-3 ovarian cancer cells after 24 h of treatment with 37.3 μM, and induced apoptosis via the mitochondrial pathway by activating caspase-3 and caspase-9. ⁴² In the present study, we demonstrated that lysicamine and liriodenine inhibit cancer cell growth by directly interacting with EGFR, an established target in cancer drug development. Our results show the potential of natural compounds to be developed into new cancer treatments, either via synthetic or semisynthetic approaches. Although further studies are needed to evaluate toxicity and molecular mechanisms, we believe that these valuable results could open up new directions for future cancer treatments.

Conclusions

Cancer is a major worldwide health burden and a leading cause of death, with numerous molecules being studied and developed as therapeutic agents. In this study, we evaluated the ability of G. elegans to inhibit cancer cell lines. According to the standards of the NCI, the crude methanol extract showed the strongest cytotoxic activity against SW-480, AGS, and SK-LU-1 cell lines, with IC₅₀ values of 14.51 ± 0.96, 19.57 ± 1.25, and 24.69 ± 2.83 μg/mL, respectively. Among the four compounds isolated from the G. elegans extract, two alkaloids—namely, lysicamine and liriodenine—showed strong inhibitory abilities against similar cancer lines. To further investigate the mechanism, a molecular docking simulation approach was used to evaluate the molecular interaction of the two candidates with EGFR. We observed that lysicamine and liriodenine have protein binding affinities of −8.8 and −9.7 kcal/mol, respectively. We also evaluated the stability of the ligand–protein complexes with molecular dynamics simulations. Our study demonstrated that G. elegans may have utility as a natural cancer treatment. Although further studies are needed to prove the efficacy and safety of these two compounds, liriodenine and lysicamine have the potential to be developed into EGFR inhibitors for cancer.

Material and Methods

Collection of Plant Material

In July 2020, the aerial parts of G. elegans were gathered from Quang Tri province, Viet Nam (N16°44′38.9″ E107°14′51.1″). The samples were verified by Quang Tri Center of Science and Technology, Vietnam Academy of Science and Technology (VAST), Vietnam. A voucher specimen (GE01) was deposited at the Faculty of Pharmacy, Hue University of Medicine and Pharmacy, Hue University, Vietnam.

General Experimental Procedures

Column chromatography was run using silica gel (60 N, spherical, neutral, 40-50 μm, Kanto Chemical Co., Inc., Tokyo, Japan), Sephadex LH-20 (Dowex® 50WX2-100, Sigma–Aldrich, USA) and YMC RP-18 (Fuji Silysia Chemical Ltd, Kasugai, Aichi, Japan). Analytical thin layer chromatography (TLC) was performed on pre-coated silica gel 60F₂₅₄ and RP-18 F_254S plates (Merck KGaA, Darmstadt, Germany). NMR spectra were recorded using a Bruker Advance 500 spectrometer (Billerica, Massachusetts) with tetramethylsilane (TMS) as the internal reference. Human colon adenocarcinoma (SW-480), gastric cancer (AGS), and lung cancer (LU-1) cell lines were used for the cytotoxic activity determinations. 96-well plates and cell culture flasks were from Corning Inc. (Corning, NY, USA). The absorbance in the cytotoxicity assay was measured using an enzyme-linked immunosorbent assay (ELISA) reader (Bio-Rad, CA, USA).

Extraction and Isolation

Dried G. elegans (4.5 kg) was extracted with MeOH at room temperature, producing 375 g of crude extract. After removing the solvent in vacuo, this extract was suspended in water and subsequently partitioned with n-hexane, CH₂Cl₂, and EtOAc (3 times). The fractions were evaporated under reduced pressure to yield fractions of n-hexane (117.6 g), CH₂Cl₂ (91.2 g), EtOAc (83.3 g), and water (74.7 g).

Based on the activity-guided fractionation method, the ethyl acetate soluble fraction (83.3 g) was applied to a silica gel column and eluted with a gradient of n-hexane-EtOAc (100:0, 95:5, 90:10,50:10, 10:10, and 0:100, v/v, 1.0 L each) to obtain six fractions (E1-E6). Fraction E5 (22 g) was separated into five subfractions (E5A-E5E) on a silica gel column eluted with n-hexane-EtOAc (5:1, v/v). Fraction E5C (2.8 g) was chromatographed on Sephadex LH-20 and eluted with MeOH to give five subfractions (E5C1–E5C5). Fraction E5C3 (450 mg) was partitioned on a YMC RP-18 column and eluted with acetone-MeOH-water (2:1:1, v/v) to obtain an additional seven fractions (E5C3A–E5C3G). Fraction E5C3D (32 mg) was partitioned on a Sephadex LH-20 column eluted with MeOH to furnish 1 (7.5 mg). Fraction E5C3B (45 mg) was chromatographed on a silica gel column using CH₂Cl₂-EtOAc (8:1, v/v) as the eluent to afford 2 (5 mg).

In Vitro Assay

The cytotoxic activity of crude extract, fractions, and compounds from the aerial parts of G. elegans was tested by an Sulforhodamine B (SRB) colorimetric assay against the growth of three human cancer cell lines, lung adenocarcinoma (SK-LU-1), gastric adenocarcinoma (AGS), and colon adenocarcinoma (SW-480). ⁴³ The above cells were cultured with 50 mL of Dulbecco‘s modified Eagle‘s medium containing 2 mM L-glutamine, 1.5 g/L sodium bicarbonate, 1.0 mM sodium pyruvate, and 10% fetal bovine serum. The media were switched every 48 h. The cells were subcultured every 3 to 5 days at a ratio of 1:3 after being dissociated with 0.05% trypsin-ethylenediaminetetraacetic acid and cultivated in a humidified atmosphere with 5% CO₂ at 37°C. Cell viability was determined using an SRB assay. Briefly, the cells at desired concentrations were added to 96-well microtiter plates and incubated for 72 h. Adherent cell cultures were fixed by adding cold 20% (w/v) trichloroacetic acid for 1 h at 4°C. SRB solution of 100 µL (0.4% wt/vol in 1% acetic acid) was added to each microtiter well, and the culture was incubated for 30 min at room temperature. Unbound SRB was removed by washing three times with 1% acetic acid, then airdrying the plates. The absorbed dye was dissolved in a Tris buffer (10 mM) and plates were gently stirred for 10 min on a mechanical stirrer. The optical density (OD) was recorded on an ELISA reader at 515 nm. Dimethyl sulfoxide (DMSO) 1% was used as the negative control, with the final concentration at 0.05%. Ellipticine was used as positive control (at concentrations of 10, 2, 0.4, and 0.08 μg/mL). The cytotoxicity was measured at doses of 100, 20, 4, and 0.8 μg/mL. The software TableCurve Version 4.0 was employed to calculate the half-maximal IC₅₀. All experiments were repeated in three time. The inhibition rates (I values) for the cell lines were calculated by the following equation I% = {100% − [(OD_t–OD₀)/(OD_c − OD₀)] × 100}, in which I is the inhibition rate of cell growth, OD_t is the average optical density at day 3, OD₀ is the average optical density at time zero, and OD_c is the average optical density of the blank DMSO control sample.

Protein and Ligand Preparation

The 3D structure of EGFR (PDB ID: 1XKK) were downloaded from Protein Data Bank. The process of removing unrelated cocrystallized ligands and water molecules was done by using PyMol software. Then, molecular hydrogens were added to the protein using the AutoDock Tool 1.5.6. Finally, the protein structure was saved into a dockable pdbqt format in preparation for molecular docking. The information on the composition of the G. elegans compounds was collected based on online databases. The 3D structures of the compounds were obtained from the PubChem library and modified or redrawn using MarvinSketch software (ChemAxon, USA). Finally, hydrogen atoms were added to all compounds and converted to the pdbqt format apted for screening using Open Babel software.

Molecular Docking

The molecular docking process of EGFR was carried out using the program AutoDock Vina 1.1.2. The grid box covering the active site of the protein was determined using the position of crystal ligand. The grid boxes of EGFR were identified with the parameters: center_x = 18.5, center_y = 35.2, center_z = 41.8, size_x = 30, size_y = 18.75, and size_z = 18.75. Docking scores were reported in kcal/mol. Results of interactions between compounds and EGFR were visualized by software Discovery Studio Visualizer 2020.

Molecular Dynamics Simulation

Molecular dynamics experiments were simulated using the GROMACS 2020.4 engine to investigate further the stability of the two ligands liriodenine and lysicamine in complex with EGFR. The complex’s topology was prepared under the CHARMM36 forcefield and solvated in a truncated octahedral box containing TIP3P water molecules. Potassium and chlorine ions were added at a concentration of 0.15M to neutralize the net charge. Energy minimization was repeated with 5000 steps using the Steepest Descent Method, with the convergence criterion being the maximum force reaching <1000 kJ/mol/nm. Afterwards, the systems were equilibrated under the isothermal-isochoric (NVT) and isothermal-isobaric (NPT) ensembles for 100 and 1000 ps, resp for each of the complex, respectively. The target value of 300 K temperature and 1 bar pressure was achieved to ensure a fully stabilized system. Finally, the production molecular dynamic simulation was run for 100 ns for each of the complexes.

Binding Free Energy (MM/PBSA Calculations)

The MM/GBSA model was utilized for calculating the relative binding free energy and implemented using MOLAICAL script. ⁴⁴ The free binding energy for a complex can be estimated as follows:

$$\mathrm{\Delta }{G}_{bind}=\mathrm{\Delta }{G}_{COMPLEX}-\mathrm{\Delta }{G}_{RECEPTOR}-\mathrm{\Delta }{G}_{LIGAND}$$

The ΔG_bind can also be represented as:

$$\mathrm{\Delta }{G}_{bind}=\mathrm{\Delta }H-T\mathrm{\Delta }S=\mathrm{\Delta }{E}_{MM}+\mathrm{\Delta }{G}_{sol}-T\mathrm{\Delta }S$$

where ΔE_MM (the gas phase molecular mechanics contribution) and ΔG_sol (the solvation Gibbs energy) are calculated as:

$$\mathrm{\Delta }{E}_{MM}=\mathrm{\Delta }{E}_{bonded}+\mathrm{\Delta }{E}_{nonbonded}=\mathrm{\Delta }{E}_{bonded}+(\mathrm{\Delta }{E}_{ele}+\mathrm{\Delta }{E}_{vdW})$$

$$\mathrm{\Delta }{G}_{sol}=\mathrm{\Delta }{G}_{pol}+\mathrm{\Delta }{G}_{no{n}_{-pol}}=\mathrm{\Delta }{G}_{GB}+\mathrm{\Delta }{G}_{\mathrm{SASA}}$$

The TΔS term refers to the conformational entropy after ligand binding, and can be calculated using the normal mode analysis method. However, this term was ignored because the enthalpy of binding is usually sufficient for comparing the relative binding energy of ligands, thus reducing the computational cost. The solvent dielectric constant of 78.5 and the surface tension constant of 0.030112 kJ/mol Å2 were used as parameters for the MM/GBSA calculations.

Supplemental Material

sj-docx-1-npx-10.1177_1934578X221138435 - Supplemental material for Phytochemicals Derived from Goniothalamus elegans Ast Exhibit Anticancer Activity by Inhibiting Epidermal Growth Factor Receptor

Supplemental material, sj-docx-1-npx-10.1177_1934578X221138435 for Phytochemicals Derived from Goniothalamus elegans Ast Exhibit Anticancer Activity by Inhibiting Epidermal Growth Factor Receptor by Linh Thuy Thi Tran, Long-Hung Dinh Pham, Nhi Yen Thi Dang, Nguyen Thao Nguyen Le, Huu Bao Nguyen and Tan Khanh Nguyen in Natural Product Communications