Structural information on gigantol (CAS:67884-30-4) was obtained from NCBI PubChem (https://pubchem.ncbi.nlm.nih.gov/) [25]. Swiss absorption, distribution, metabolism and excretion (SwissADME), was used to identify the bioavailability and metabolism of gigantol [26]. Additionally, the protein targets of this compound were retrieved from SwissTargetPrediction with a probability value $\ge$0.5. Data on hepatocellular carcinoma-related targets were integrated from five databases using R Software (Lucent Technologies, Murray Hill, NJ, USA): GeneCards (https://www.genecards.org) [27], OMIM (http://omim.org) [28], PharmGKB (https://www.pharmgkb.org) [29], TTD (http://db.idrblab.net/ttd) [30], and DrugBank (https://go.drugbank.com) [31]. Common targets for gigantol and HCC were obtained using R packages (“Venn”) (https://cran.r-project.org/web/packages/VennDiagram/index.html).

The common targets were uploaded to the STRING database (https://string-db.org/), and the protein interaction network was constructed with the following conditions: the species was “Homo sapiens”, the minimum interaction score (medium confidence) was 0.4, and the free protein sites were removed [32]. The obtained protein interaction networks were imported into Cytoscape 3.8.2 software (Institute of Systems Biology, Boston, MA, USA) for visualization and processing analysis using the Network Analyzer module [33]. Finally, the core genes were screened by the cytoNCA plug-in. The cytoNCA plug-in performs topology analysis with a 2-fold average degree value as a filtering condition.

GO and KEGG pathway enrichment analysis were performed using the R package (“clusterProfiler” “org.Hs.eg.db” “enrichplot” “ggplot2” “pathview”). The significance values for GO terms and KEGG pathways were set at FDR 0.05 and p 0.05, respectively.

The two-dimensional structure of gigantol were available from the PubChem website in structured data format (SDF) format [25]. The 3D structure of the core target protein was downloaded from the PDB database (https://www1.rcsb.org/) and saved in PDB format [34]. Water molecules and small molecule ligands were removed using PyMOL software (Schrödinger, New York, USA) and hydrogenated using AutoDockTools software (Scripps Research, La Jolla, CA, USA) (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC) [35]. Molecular docking of the receptor and ligand was performed using Vina 1.1.2 (Scripps Research, La Jolla, CA, USA) [36]. The binding activity was evaluated by docking score values.

The human hepatocellular carcinoma cell lines Hep3B, SMMC-7721, and HCC-LM3 were obtained from the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% FBS (Gibco, Grand Island, NY, USA) and 100 U/mL Penicillin/Streptomycin (Gibco, Grand Island, NY, USA) in a 5% CO${}_2$ incubator. Cells undergoing a logarithmic growth period were used for follow-up experiments.

A CCK-8 assay kit was used to evaluate cell viability following the manufacturer’s instructions (HY-K0301, MedChemExpress, Monmouth Junction, NJ, USA). Cells in the logarithmic growth phase were inoculated into 96-well plates (7 $\times$ 10${}^3$ cells/well). After 12 h, different concentrations of gigantol (HY-N2523, MedChemExpress, Monmouth Junction, NJ, USA) were added, and 6 replicate wells were added for each group. Cells were incubated at 37 ${}^{\circ }$C for 24, 48 and 72 h. The supernatant was then discarded, and the cells were washed twice with PBS. A solution of 10% CCK-8 was then added and incubated at 37 ${}^{\circ }$C for 2 h. The absorption was measured at 450 nm by an microplate reader (Multiskan™ FC Microplate Photometer, ThermoFisher, Waltham, MA, USA).

A EdU detection kit (Us Everbright, Suzhou, Jiangsu, China) was used, as stated by the manufacturer’s instructions, to identify the proliferating cells. SMMC-7721 and Hep3B were intervened with different concentrations of gigantol for 24 h after apposition. Appropriate amounts of cells were inoculated in 96-well plates and cultured in the logarithmic growth phase. EdU solution was diluted with complete medium at a 1000:1 ratio, 100 $\mu$L EdU medium per well, and incubated in an incubator for 2 h. After washing twice with PBS, 50 $\mu$L of 4% paraformaldehyde was added to each well and incubated at room temperature for 30 min. After washing with PBS for 5 min, 100 $\mu$L of 0.5% Triton X-100 permeate was added to each well. Next, 100 $\mu$L of Apollo staining reaction solution was added to each well. The staining solution was then incubated for 30 min at room temperature and discarded. Finally, Hoechst 33324 staining solution was added to each well and incubated for 15 min, followed by 3 washes with PBS for 5 min each. Fluorescence microscopy was performed to observe the EdU-positive cells, which were counted and statistically analyzed.

Hep3B and SMMC-7721 cells were seeded onto cell inserts (Culture-Insert 4 well, ibidi, Gräfelfing, Germany) in a 12-well plate at a density of 5 $\times$10${}^5$ cells/mL. The insert was removed after 12 h. Subsequently, the cells were imaged under a microscope in which the 0 h results were used as control conditions. Next, the cells were processed with different concentrations of gigantol (0, 25, 50, 100 $\mu$M) for 24 h. The area ratios were imaged by microscopy after 24 hand calculated using Image J software (National Institutes of Health, Bethesda, MD, USA).

For the transwell invasion assay, matrigel chambers (BD Biosciences, San Jose, CA, USA) were carried out according to manufacturer’s instructions. Matrigel was melted at 4 ${}^{\circ }$C followed by dilution with DMEM exclusive of FBS at a concentration of 1 mg/mL. Then, 100 $\mu$L of Matrigel (1 mg/mL) was placed into the upper chamber, where it was incubated for 4 h at 37 ${}^{\circ }$C until it aggregated. Hep3B and SMMC-7721 cells were cultured in serum free medium for 12 h. Cells in 200 $\mu$L of serum free medium that contained different concentrations of gigantol (0, 25, 50, 100 $\mu$M) were placed in the upper chamber. The lower chamber was filled with 600 $\mu$L of medium containing 15% FBS. Following 24 h of incubation, cells were fixed with 4% paraformaldehyde and subsequently stained with 0.1% crystal violet. Furthermore, the upper chamber was carefully cleaned with a cotton swab. The number of cells migrating to the lower surface was counted in five randomly selected fields of view ($\times$200). The results were quantified using Image J software (National Institutes of Health, Bethesda, MD, USA).

Cells in logarithmic growth phase were seeded in 6-well plates at No. 1 $\times$ 10${}^6$ cells/cm${}^2$and then treated with gigantol (25, 50, 100 $\mu$M) and blank control (DMSO) for 48h at 37 ${}^{\circ }$C with 5% CO${}_2$ in an incubator. Subsequently, the cells were collected and lysed using radio immunoprecipitation assay (RIPA) on ice for 30 min and centrifuged at 4 ${}^{\circ }$C for 20 min at 12,000 $\times$ g. The supernatant was transferred to an EP tube. The protein concentration was measured with a BCA protein concentration assay kit (KGP902, KeyGEN BioTECH, Nanjing, Jiangsu, China). Cell lysates were then loaded into 10% SDS-PAGE gels. Proteins were transferred to polyvinylidene difluoride (PVDF) (IPVH00010, Millipore, Burlington, MA, USA) membranes and blocked with 5% skim milk in Tris-buffered saline containing 1% Tween 20 for 1 h. The PVDF membranes were then incubated with primary antibody at 4 ${}^{\circ }$C overnight. N-cadherin primary antibody (1:1000, #13116, rabbit), XIAP primary antibody (1:1000, #2045, rabbit), HSP90 primary antibody (1:1000, #4877, rabbit), ESR1 primary antibody (1:1000, #8644, rabbit), PCNA primary antibody (1:1000, #13110, rabbit), phospho-Akt primary antibody (1:2000, #4060, rabbit), Akt primary antibody (1:1000, #4685, rabbit), $\beta$-actin primary antibody (1:1000, #4970, rabbit), E-cadherin primary antibody (1:1000, #3195, rabbit) and Vimentin primary antibody (1:1000, #5741, rabbit) were purchased from Cell Signaling technology, Inc., Danvers, MA, USA. CDK1 primary antibody (1:10000, ab133327, rabbit) and MCM2 primary antibody (1:1000, ab108935, rabbit) was purchased from Abcam (San Francisco, CA, USA). The membranes were thoroughly washed and incubated with horseradish peroxidase-conjugated secondary antibodies (1:2000, KGAA35, goat anti-rabbit IgG, KeyGEN BioTECH, Nanjing, China). Once washed, the protein bands were visualized by enhanced chemiluminescence (ECL), and the gray values of the proteins were analyzed using Image J software (National Institutes of Health, Bethesda, MD, USA).

GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA, USA) for Windows was used for statistical analysis. Statistically significant differences were calculated using ANOVA. A p-value 0.05 was accepted as statistically significant.

The SwissADME results indicated that gigantol not only undergoes passive gastrointestinal absorption but also crosses the BBB (Table 1) (http://www.swissadme.ch/index.php). A total of 102 gigantol-associated targets were retrieved (Supplementary Table 1). Subsequently, a total of 1184 hepatocellular carcinoma related target genes were compiled from the OMIM, PharmGKB, GeneCards, TTD and DrugBank databases (Fig. 1A). Thus, we integrated the drug targets and disease targets using the R package “Venn”. After eliminating duplicates, we identified 32 common targets, which are shown in Fig. 1B.

Fig. 1.

Analysis of the predicted genes. (A) Intersection of HCC target genes predicted by five types of databases. (B) Venn diagram of HCC-related genes and gigantol target genes. Using R packages (“Venn”).

The 32 intersecting targets were imported into the STRING database to obtain the protein interaction network and the corresponding network (Fig. 2A–B). The data were loaded into Cytoscape for visual topological analysis, and 32 nodes with 148 edges and an average degree value of 9 were obtained, in which ESR1, ERBB2, HSP90AA1, MAPK8, CDK4, XIAP, RPS6KB1, PTGS2, CDK1, IGF1R, ABL1, PARP1, and CHEK1 were selected (Fig. 2C) (Table 2). Then, we further analyzed the high score proteins using the cytoNCA plug-in and obtained 13 nodes with 58 edges and an average degree value of 8.9. Three targets, HSP90AA1, XIAP, and ESR1, were identified as occupying a central position in the PPI network and were expected to serve as key core targets (Fig. 2D).

Fig. 2.

Network pharmacology of gigantoland HCC. (A) The PPI network of common genes. (B) The “HCC-targets-gigantol” network. (C) Visualization of the first 13 core genes was obtained using R software. (D) The top 3 potential effective targets in core genes. Obtained by Cytoscape and cytoNCA plug-in analysis for mapping.

GO enrichment analysis was performed on 32 intersecting targets using the R language package. A total of 707 biological process (BP) entries were obtained, involving peptidyl-serine modification, regulation of cell cycle G2/M phase transition, peptidyl-serine phosphorylation, regulation of protein kinase B signaling, signal transduction in response to DNA damage and other processes. Thirty seven cellular component (CC) entries were obtained, involving the pigment granule membrane, nuclear chromosome, telomeric region, cyclin-dependent protein kinase holoenzyme complex, and other components. Forty two molecular function (MF) entries involved nuclear receptor binding, transmembrane receptor protein tyrosine kinase activity, and cyclin-dependent protein kinase activity. The top 10 entries of the 3 cluster analyses were plotted according to the number of genes sorted (see Fig. 3A). KEGG enrichment analysis was performed on 32 intersecting targets with the R package. A total of 106 entries were obtained, and the results suggested that genes were predominantly enriched in the PI3K-Akt signaling pathway, endocrine resistance, ErbB signaling pathway, IL-17 signaling pathway, HIF-1 signaling pathway, and hepatocellular carcinoma, as shown in Fig. 3B (Supplementary Table 2). The results clearly revealed that the PI3K-Akt signaling pathway was more abundant among HCC-related pathways and had a critical role in the treatment of HCC (Fig. 4).

Fig. 3.

Enrichment analysis of the 32 genes by GO and KEGG. (A) GO functional annotation of 32 genes of gigantol. BP, CC and MF were analyzed using R software. (B) KEGG pathway analysis of 32 genes and column plot for top 30 pathways. GO and KEGG analysis were performed using the R package (GO, KEGG: “clusterProfiler”, “org.Hs.eg.db”, “enrichplot”, “ggplot2”).

Fig. 4.

The PI3K-Akt signaling pathway. The nodes marked in red are target genes. The red dashed box highlights our predicted key drug targets in the PI3K-Akt pathway. HSP90 may be one of the targets for the action of gigantol. HSP90/Akt/CDK1 is the likely downstream pathway of interest. Using R packages (“pathview”).

Whether a small molecule can bind to a large protein is mainly evaluated by the binding energy, which if it is less than 0, the ligand and the receptor can bind spontaneously. Smaller values indicate stronger binding capacity where the active ingredient binds to the receptor more easily. In this experiment, the average docking score of gigantol was used as the threshold value. The binding energy of the target protein to small molecules with a threshold value less than –5 kcal/mol was calculated. Three proteins were successfully docked with gigantol. The molecular docking results revealed that the binding energies of gigantol to the proteins ESR1, HSP90AA1, and XIAP were –6.5, –6.3, and –6.2 kcal/mol, respectively, indicating that the three molecules were highly active (Table 3). Affinity values –7 suggest a relatively high binding activity [37]. Gigantol binds to ESR1 protein residues GLY521, HIS524, and LEU346 (Fig. 5A–B); to HSP90AA1 protein residues PHE138 and ASP93 by hydrogen bonding and van der Waals forces (Fig. 5C–D); and to XIAP protein residues GLN1074, TYR75, GLN3074, and ARG1072 (Fig. 5E–F). From the valence bonding point of view, these three proteins bind to gigantol via multiple valence bonds, indicating a high binding stability.

Fig. 5.

Molecular docking. (A) 3D structural diagrams of the interactions between ESR1, HSP90AA1 (C), XIAP (E) with gigantol, respectively. (B) Plot of the distribution of the interaction forces between ESR1, HSP90AA1 (D), XIAP (F) and gigantol, individually.

The inhibition rates of gigantol on Hep3B, SMMC-7721, and HCC-LM3 cells are shown in Fig. 6A. For SMMC-7721 cells, the IC${}_{50}$ was 84.18 $\mu$M at 24 h, 66.90 $\mu$M at 48 h, and 57.95 $\mu$M at 72 h. For HCC-LM3 cells, the IC${}_{50}$ was 70.2 $\mu$M at 24 h, 46 $\mu$M at 48 h, and 60.21 $\mu$M at 72 h. For Hep3B cells, the IC${}_{50}$ was 70.2 $\mu$M at 24 h. Surprisingly, the IC${}_{50}$ decreased to 20.57 $\mu$M at 48 h and to 16.01 $\mu$M at 72 h. Thus, we selected Hep3B and SMMC-7721 cells for the next assays. EdU can be incorporated into newly synthesized DNA as a substitute for thymine during DNA synthesis. The newly synthesized DNA was labeled with the corresponding fluorescent probe, and proliferating cells were detected using fluorescent detection equipment. EdU staining assays further examined the proliferation inhibition ability of gigantol on hepatocellular carcinoma cells and demonstrated that EdU-positive cells were significantly reduced after treatment with different concentrations of gigantol compared to the control group (Fig. 6B–C).

Fig. 6.

Gigantol inhibits the proliferation of HCC cells. (A) The viability of SMMC-7721, Hep3B and HCC-LM3 cells treated with different concentrations of gigantol for 24 h, 48 h and 72 h was examined by CCK-8 assay. (B) The multiplicative capacity of SMMC-7721 and Hep 3B cells was decreased after different concentrations of gigantol intervention for 24 h. (C) Bar graphs show the percentage of EdU-positive cells in SMMC-7721 (p 0.001) and Hep3B cells (p 0.01).

As shown in Fig. 7A–B, wound healing assays demonstrated that the scratches of gigantol treated Hep3B and SMMC-7721 cells were significantly larger than those of the control cells after 24 h. Additionally, the Matrigel invasion assay demonstrated that gigantol inhibited the invasion rate of Hep3B and SMMC-7721 cells compared to the control group (Fig. 7C–D). The aforementioned results demonstrated that gigantol could effectively inhibit the migration and invasion of HCC cells.

Fig. 7.

Gigantol suppressed migration and invasion of cells. (A) The representative micrographs of “wound healing assay” of SMMC-7721 cells and Hep3B cells after 24 h are shown. (B) Bar graphs show the percentage of the area for the scratched region in SMMC-7721 (p 0.001) and Hep3B cells (p 0.01). (C,D) After intervention with gigantol, the invasive ability and number of invaded SMMC-7721 and Hep3B cells were evaluated by Transwell invasion assay.

To further determine the efficacy of gigantol, we assayed proliferation, migration and invasion related biomarkers. Following gigantol intervention, PCNA and MCM2 secretion was significantly diminished in the experimental group compared to the control group. Stimulation with different concentrations of gigantol (0, 25, 50, 100 $\mu$M) showed a pronounced decrease in a dose-dependent manner. For migration and invasion related proteins, the experimental group significantly inhibited N-cadherin and vimentin protein expression after treatment with different doses of gigantol, while the level of E-cadherin was significantly increased (Fig. 8).

Fig. 8.

Effects of gigantol on biomarkers of Phenotypes. (A,B,C) (D,E,F) Detection of proliferation and EMT biomarkers using western blot analysis. Data of quantifiable western blot assays.

To verify the regulatory effect of gigantol on target proteins, we examined the changes of protein levels after gigantol intervention. As can be seen from Fig. 9, the expression levels of ESR1, HSP90AA1, and XIAP were gradually depressed with the increase of drug concentration. Combined with the results obtained in Fig. 4, we found that the drug target HSP90 protein is the upstream regulator of Akt. Therefore, we tentatively examined the changes of downstream p-Akt, Akt and CDK1 protein levels. As expected, we found that the levels of p-Akt and CDK1 were subsequently altered after drug intervention. This is consistent with the altered levels of the 3 targets.

Fig. 9.

Effects of gigantol on drug targets and HSP90/Akt/CDK1 signaling pathway. (A,B,C,D,E,F) Detection of ESR1, HSP90AA1, XIAP, p-Akt, Akt and CDK1 using western blot analysis. Data of quantifiable western blot assays.