Design, synthesis and biological evaluation of novel diosgenin–benzoic acid mustard hybrids with potential anti-proliferative activities in human hepatoma HepG2 cells

Abstract To discover new lead compounds with anti-tumour activities, in the present study, natural diosgenin was hybridised with the reported benzoic acid mustard pharmacophore. The in vitro cytotoxicity of the resulting newly synthesised hybrids (8–10, 14a–14f, and 15a–15f) was then evaluated in three tumour cells (HepG2, MCF-7, and HeLa) as well as normal GES-1 cells. Among them, 14f possessed the most potential anti-proliferative activity against HepG2 cells, with an IC50 value of 2.26 µM, which was 14.4-fold higher than that of diosgenin (IC50 = 32.63 µM). Furthermore, it showed weak cytotoxicity against GES-1 cells (IC50 > 100 µM), thus exhibiting good antiproliferative selectivity between normal and tumour cells. Moreover, 14f could induce G0/G1 arrest and apoptosis of HepG2 cells. From a mechanistic perspective, 14f regulated cell cycle-related proteins (CDK2, CDK4, CDK6, cyclin D1 and cyclin E1) as well mitochondrial apoptosis pathway-related proteins (Bax, Bcl-2, caspase 9, and caspase 3). These findings suggested that hybrid 14f serves as a promising anti-hepatoma lead compound that deserves further research.


Introduction
As of September 2019, approximately 84.3% (156 out of 185) of anticancer small molecules approved by the FDA are natural products or their derivatives 1 , which affirms the position of natural products as a focal point for finding structural and medicinal inspiration for drug discovery. Natural steroids and their synthetic derivatives are attracting increasing research interest for their promising anti-tumour activities and potential utilisation in the discovery and design of new anti-tumour agents [2][3][4] .
In-depth studies on the anti-tumour effects and mechanistic patterns of DSG have revealed that it could regulate multiple genes and several signalling pathways in various types of human cancers. For instance, DSG could arrest the cell cycle at the G2/M phase by regulating the Cdc25C-Cdc2-cyclin B pathway in human breast cancer cells 7 , exert tumour-suppressive function by inhibiting Cdc20 in osteosarcoma cells 8 , induce mitochondria-mediated apoptosis in human cholangiocarcinoma cells and apoptosis via suppression of Skp2 in human breast cancer cells 5,9 , inhibit the activation of cAMP/PKA/CREB pathway in colorectal cancer cells 6 , and so on. Although DSG possess extensive anti-cancer activity, the application of DSG for cancer therapy was limited by its moderate potency. Therefore, it is important to optimise the scaffold of DSG to obtain promising anti-tumour compounds with improved inhibitory effect.
Nitrogen mustards, which are a type of DNA bifunctional alkylating agents, are developed as clinically useful anti-cancer agents, and include compounds such as chlorambucil, mechlorethamine, melphalan, cyclophosphamide, and estramustine. They exert cytotoxicity by binding to DNA, cross-linking the two chains, and preventing cell replication 10,11 . Molecular hybridisation is a widely used strategy in drug discovery, which forms new molecular entities by incorporating two or more bioactive substructures through suitable linkages [12][13][14][15][16] . These hybridised molecules possess improved or new biological properties relative to their individual components 17 . Recently, it has been reported that the hybridisation of natural products with nitrogen mustards provides new strategies for discovering anticancer molecules with improved anti-cancer effect, selectivity, and reduced toxicity. For example (Figure 2), a series of b-carboline derivatives with nitrogen mustard moieties synthesised by Sun et al. showed potent inhibitory activities in human breast carcinoma cells (MCF-7 and MDA-MB-231); especially, compound A containing benzoic acid mustard possessed significant anti-proliferative activity against MCF-7 cells 18 . Compound B, which was obtained by introducing a benzoic acid mustard fragment displayed the highest anti-proliferative properties against cervical cancer HeLa cells 19 . In addition, Han et al. synthesised a series of novel conjugates of brefeldin A and nitrogen mustards and found that compound C was the most active derivative against Bel-7402 cells 20 . Compound D, which was synthesised by conjugating the D-ring-derived androstene oxime with benzoic acid mustard, exhibited the most outstanding effect on inhibiting the growth of ovarian cancer IGROV1 cells 21 . These findings, coupled with the anti-cancer profiles of DSG and benzoic acid mustard, promoted us to further explore the anti-tumour potential of DSG-benzoic acid mustard hybrids.
To the best of our knowledge, there are few reports on the synthesis and biological activity of DSG-benzoic acid mustard derivatives with amide-amide linkages. Therefore, in this study, with the aim of finding new kind of DSG derivatives with improved anti-tumour activity, selectivity, and reduced toxicity, and further explore their structure-activity relationships, fifteen novel DSG-benzoic acid mustard hybrids bearing diversified linkers were designed and synthesised using the molecular hybridisation strategy. The cytotoxic activities of hybrids were evaluated against three cancer cell lines (HepG-2, MCF-7, and HeLa) and normal GES-1 cells. Furthermore, in-depth anti-proliferative mechanisms of the most potent compound, 14f, including cell cycle progression, expression of cell cycle-related genes and proteins, induction of apoptosis, changes in mitochondrial membrane potential (MMP) and expression of apoptosis-related genes and proteins were explored as well. Taken together, these new results  might give researchers some inspiration for structural modification of other natural products, to facilitate discovery of novel antitumour agents.

Materials and methods
All reagents were obtained from chemical and biological companies. Anhydrous solvents were dried through routine protocols. DSG derivatives 5, 6, and 11 were prepared using our published procedures 22 . The purity of the compounds was measured by HPLC by using a Waters Symmetry C18 (4.6 Â 250 mm, 5 mm) column and its peak UV detection was 254 nm. HPLC conditions: methanol/water with 90:10; flow rate 1.0 ml/min. The purity of target compounds was >95% by the analysis of HPLC. Nuclear magnetic resonance (NMR) spectra were recorded using a AVANCE NEO 600 ( 1 H, 600 MHz; 13 13

General procedure for synthesising 15a-15f
A solution of compounds 14a-14f in aqueous NaOH (4 N) in methanol:tetrahydrofuran (1:1.5, v/v) was stirred for 6 h at room temperature. After concentration in vacuo, 30 ml of water was added to it and it was extracted with CH 2 Cl 2 (6 ml Â 3). The CH 2 Cl 2 extract was washed with aqueous saturated NaCl (25 ml) and dried over anhydrous Na 2 SO 4 , following which the solids were removed by means of filtration. The reaction mixture was evaporated under reduced pressure and purified using preparative TLC with a CH 2 Cl 2 /MeOH (10:1, v/v) system, to obtain the compounds 15a-15f.  13

Cells and cell culture conditions
Human hepatoma carcinoma HepG2, breast carcinoma MCF-7, cervical cancer HeLa, and gastric epithelial GES-1 cells were acquired from America Type Culture Collection (USA). HepG2 and GES-1 cells were cultured in Dulbecco's Modified Eagle's Medium supplemented with 10% foetal bovine serum (FBS). MCF-7 cells were incubated in Minimum Essential Medium supplemented with 10% FBS. HeLa cells were cultured in Roswell Park Memorial Institute medium supplemented with 10% FBS. All cells were maintained in a humidified incubator with 5% CO 2 , at 37 C.

MTT assay
The inhibitory activities of the target hybrids (8-10, 14a-14f, and 15a-15f) were tested using the MTT method, as described previously 23 . HepG2, MCF-7, HeLa, and GES-1 cells were seeded into 96-well plates, at a density of 5 Â 10 3 /well. After 24 h, the cells were treated with serially diluted concentrations of the test compounds. After treatment for 48 h, 10 mL of MTT (5 mg/mL) was added into each well and incubation was continued for another 4 h. The medium was then removed and 100 mL of dimethyl sulfoxide (DMSO) was added into the wells, to dissolve the formazan crystals. Finally, the absorbance was read on a microplate reader (iMark TM , Bio-Rad, USA), at the wavelength of 492 nm.

Flow cytometric analysis of cell cycle distribution
HepG2 cells were seeded into 6-well plates and allowed to grow for a period of 12 h. After incubation with different concentrations of 14f (0, 5, 10, and 20 mM) for 24 h, the cells were trypsinized, washed with phosphate-buffered saline (PBS), and fixed in 1.5 ml of 75% ethanol for the night, at 4 C, following which RNase and propidium iodide (PI) (Multi Sciences Biotech Co. Ltd.) were added to the cells. The cell cycle distribution and data processing were analysed using a flow cytometer (FACSCalibur TM , BD Biosciences, USA).

Assessment of changes in cell morphology
First, HepG2 cells were plated into 6-well plates, at a density of 1 Â 10 6 cells/well. Thereafter, 14f (0, 5, 10, and 20 mM) was added to the cells and they were incubated with it for 48 h; 0.1% DMSO was used as a vehicle control. The cellular morphology was then observed and photographed using a light microscope (Olympus, 1 Â 51, Japan). Next, HepG2 cells were incubated with 14f (0, 5, 10, and 20 mM) for 48 h, after which the cells were collected and stained with a mixture of Hoechst 33342 and PI in buffer solution, for 20 min at room temperature. Finally, the cell samples were studied and photographed under a fluorescence microscope (Obeserve.A1, Zeiss, Germany).

Assessment of apoptosis using flow cytometry
Briefly, HepG2 cells were incubated in 6-well plates for 24 h, and then treated with 14f (0, 5, 10, and 20 mM) for 48 h. Following that, the cells were washed once with pre-cold PBS, re-suspended in 300 mL of buffer solution, and stained with 10 mL of Annexin V-FITC and 5 mL of PI staining solution, under the conditions of 37 C and absence of any light source, for 10 min. Finally, the samples were immediately measured using a flow cytometer. The percentages of apoptotic cells were determined on the CellQuest TM software (BD Biosciences).

Evaluation of MMP (dwm)
HepG2 cells were plated into 6-well plates for 12 h. Following treatment with varying concentrations of 14f for 48 h, the cells were incubated with JC-1 dyeing working solution for 0.5 h at room temperature, then washed with staining buffer 3 times and re-suspended in 300 mL of staining buffer. The cells were finally assessed and analysed using a flow cytometer (FACSCalibur TM ).

Quantitative real-time polymerase chain reaction (qRT-PCR) assay
Total RNA was extracted from HepG2 cells using a Trizol TM reagent (Invitrogen, USA) and qRT-PCR was conducted on it, as documented before 24 . The purity and concentration of total RNA were checked using a spectrophotometer, in terms of A 260 nm/ A 280 nm. A PrimeScript TM RT reagent Kit (Takara Bio, Japan) was applied to synthesise cDNA, according to the manufacturer's instructions. A Detection System (Q5, ABI, USA) was used to carry out the qRT-PCR, using a TB Green TM Kit (Takara Bio) and the cycling conditions recommended for all genes in the kit instructions. Assessment of gene expression was carried out using the 2 ÀDDCt method. The primer sequences used are presented in Table 1.

Determination of caspase activities
The activities of caspase 9 and caspase 3 in the cell lysates were evaluated as described previously 25 . The harvested HepG2 cells treated with varying concentrations of 14f for 48 h were lysed using chilled 100 mL lysis buffer, to extract the total protein. The protein concentration was assessed using a Bradford Kit. The protein samples were incubated with assay buffer and colorimetric substrate (Ac-DEVD-pNA or Ac-LEHD-pNA) at room temperature for 2 h, according to the manufacturer's protocol. The released pNAs were measured at the wavelength of 405 nm using a Multiskan V R Spectrum system. The experiments were performed in triplicate.

Western blot
To analyse the effect of compound 14f on protein levels, western blot assay was employed, as elaborated previously 26 . The proteins were acquired by collecting HepG2 cells treated with varying concentrations of 14f for 48 h and adding them to lysis buffer, in an ice bath for 30 min. After quantitation of protein concentration using BCA assay, equal amounts of total protein were separated using 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and then transferred onto polyvinylidene fluoride membranes. The membranes were incubated with 5% bovine serum albumin solution for 1.5 h at 37 C, followed by incubation with primary antibody overnight, and finally secondary antibody for 2 h at 37 C. The blots were detected using enhanced chemiluminescence reagent with a fully automatic gel imaging system (ChemiDoc TM MP, Bio-Rad). The relative expression was calculated by normalising the expression of the control group to 1 for comparison.

Molecular docking study
Molecular docking studies were performed using GLIDE (2016, Schr€ odinger Suite) 27 . The crystal structures of Bcl-2 (PDB: 2O2F) 28 were retrieved from the RCSB Protein Data Bank, and further prepared using the Protein Preparation Wizard tool implemented in the Schr€ odinger Suite, by adding all hydrogen atoms as well as missing side chains of residues and deleting all bound water. The ligands were built within Maestro BUILD (2016, Schr€ odinger Suite) and prepared using the LIGPREP module (2016, Schr€ odinger Suite) 27 . The Glide Grid was built using an inner box of dimensions 15 Â 15 Â 15 Å 3 around the centroid of the ligand, assuming that the ligands to be docked were of a size similar to that of the co-crystallized ligand. This docking methodology has been validated by extracting the crystallographic bound ligand and redocking it with the Glide module using extra precision. Different docking poses of ligands were generated and analysed for interpretation of the final results.

Chemistry
The structures and reaction conditions for the preparation of the new hybrids of DSG-benzoic acid mustard are depicted in Schemes 1 and 2. The benzoic acid mustard 4 was prepared according to a previously described method 20 . DSG derivative 6, which was obtained using a procedure reported previously by our research group 22 , was saponified to obtain intermediate 7.
Intermediate 6 or 7 was coupled with benzoic acid mustard in CH 2 Cl 2 using EDCI and DMAP as catalysts, to obtain the target DSG-benzoic acid mustard hybrids 8, 9, and 10 (Scheme 1).  Next, the C-26 hydroxyl of intermediate 6 was oxidised using Jones reagent, to obtain intermediate 11. Further, the C-26 carboxyl of intermediate 11 was reacted with various N-Boc-protected amines using TBTU and DIPEA as coupling catalysts, to provide conjugates 12a-12f (Scheme 2), which were converted to 13a-13f by using CF 3 COOH for deprotection. The DSG-benzoic acid mustard hybrids 14a-14f were prepared using the same methods as those for 10. Finally, saponification of hybrids 14a-14f resulted in the corresponding target DSG-benzoic acid mustard hybrids 15a-15f. The structures of all new hybrids were fully corroborated using various spectroscopic methods, including HRMS and NMR spectroscopy ( 1 H NMR and 13 C NMR).

In vitro anti-proliferative activity
The anti-proliferative activities of the synthesised target hybrids towards three human cancer cells HepG2 (hepatoma), MCF-7 (breast cancer), HeLa (cervical cancer) and normal GES-1 (stomach) cells were evaluated using MTT assay. Mitomycin C was selected as the positive control drug, and the results of this experiment are listed in Table 2. Benzoic acid mustard (4) was firstly introduced on the scaffold of DSG derivatives 6 and 7 via an ester bond. Unfortunately, the resulting hybrids 8-10 showed decreased inhibitory activities against the three cancer cells.
Compared to DSG and 4, all the target hybrids 14a-14f and 15a-15f (except for 15d) displayed moderate to potent inhibitory activities in HepG2 cells, with IC 50 values in the range of 2.26-25.29 mM. The results showed that introducing amide-amide bonds into the two pharmacophores of DSG and 4, to afford hybrids 14a-14f and 15a-15f, could improve their anti-proliferative potency. However, changing the linkers does not seem to regularly affect activities. Among them, compound 14f (IC 50 ¼ 2.26 mM), which comprised of DSG derivative 11 substituted by the moiety of 4 through a homopiperazinyl linker, showed obvious selective anti-proliferative activity against HepG2 cells. Its  potency was approximately 15.0-fold higher than that of DSG (IC 50 ¼ 33.87 mM), and 14.4-fold higher than that of Mitomycin C (IC 50 ¼ 32.63 mM). The cytotoxic activity of 14f was greatly improved in comparison with those of the counterparts of DSG-amino acidbenzoic acid mustard trihybrids 12a-12g (IC 50 > 10.43 mM) 25 .
In MCF-7 cell line, hybrids 14a-14f (IC 50 > 11.07 mM) generally exhibited higher inhibitory activity than hybrids 15a-15f (IC 50 > 37.20 mM). These data showed that the group of acetyl substituted at the C-3 OH was beneficial for compounds with inhibitory potency against MCF-7 cell line. Among them, only hybrid 14f possessing a homopiperazinyl linker exhibited moderate antiproliferative activity, with an IC 50 value of 11.07 mM, and was about 1.7-fold more active than DSG (IC 50 ¼ 23.91 mM). The results also confirmed that the incorporation of homopiperazinyl between DSG derivative 11 and 4 results in favourable enhancement of anti-proliferative activity against MCF-7 cells, as compared to those of nitrogen-containing groups.
The selectivity index (SI), one of the important pharmaceutical parameters, was calculated to determine the toxicities of these hybrids against human normal gastric epithelial GES-1 cell line, as compared to those against hepatoma HepG2 cell line ( Table 2). Among them, the most potent anti-proliferative hybrid 14f showed considerable safety (SI > 44. 25).
Taken together, among these DSG-benzoic acid mustard hybrids, hybrid 14f, with the linkage of homopiperazinyl, showed higher anti-proliferative activity against HepG2 and MCF-7 cancer cells, which suggested that the introduction of homopiperazinyl between DSG derivative 11 and benzoic acid mustard is beneficial for anti-proliferative activity. Our findings are in good agreement with a report by Wolfram et al., which also showed that the linker of homopiperazinyl is essential for mitocanic triterpenoidic rhodamine B adducts that show high cytotoxicity against a panel of human tumour cell lines 29 . Hybrid 14f showed the lowest IC 50 value of 2.26 mM and low anti-proliferative activity against normal GES-1 cells, with an IC 50 value > 100 mM. Hence, 14f was selected for further investigation of the possible cellular mechanisms in HepG2 cell line.

Effect of 14f on the cell cycle
Uninterrupted cell cycle progression is vital to the biochemical processes surrounding cell division and replication. Thus, blockade of the cell cycle is considered as an effective strategy in cancer therapy 30 . Previous studies have demonstrated that DSG arrests cell cycle and induces apoptosis in different cancer cells 5,7,31,32 . To determine whether the anti-proliferative effects of hybrid 14f are caused by cell cycle arrest at a certain phase, the effects of different concentrations of 14f on cell cycle progression were examined in HepG2 cells. Upon treatment with different concentrations (0, 5, 10, and 20 mM) of 14f, 32.38, 33.76, 40.38, and 46.40% of the cells, respectively, were found to be in the G0/G1 phase (Figure 3). The results demonstrated that 14f arrested the cell cycle at the G0/G1 phase in HepG2 cells.
Subsequently, the specific mechanism by which 14f regulates the G0/G1 phase was further investigated at the transcription and translation levels. The effect of 14f on the expression levels of several key cell cycle-related genes in HepG2 cells was determined using qRT-PCR. It was found that 14f decreased the genes levels of CDK2, CDK4, CDK6, cyclin D1 and cyclin E1 (Figure 4(A)), all of which appear in the G0/G1 phase. In addition, upon performing western blot to detect the G0/G1 phase-related proteins, it was found that 14f could concentration-dependently downregulate the expression levels of CDK2, CDK4, CDK6, cyclin D1, and cyclin E1 (Figure 4(B)). These findings suggested that the inhibition of HepG2 cells by 14f was related to the induction of G0/G1 phase arrest.

14f promoted apoptosis in HepG2 cells
Inducing apoptosis is an important strategy for cancer treatment. To assess whether the 14f could promote the apoptosis of HepG2 cells, we detected the apoptosis of HepG2 cells upon treatment with various concentrations of 14f. First, we observed the morphological changes in HepG2 cells under an inversion microscope. As seen in Figure S1, microscopic observation showed that HepG2 cells presented a significant round shape with shrinkage of the cell membrane upon treatment with 14f, accompanied by a prominent decrease in the number of cells. These phenomena were further examined using the Hoechst 33342/PI staining assay, the results for which showed that upon treatment with 14f, there was an increase in the apoptosis of HepG2 cells, indicated by deep blue and red fluorescence, accompanied by the observation of apoptosis bodies ( Figure 5). Collectively, the above observation demonstrated that 14f treatment induced apoptosis in HepG2 cells.
To further obtain an accurate evaluation of the 14f-mediated apoptosis of HepG2 cells, cells treated with different concentrations of 14f were analysed using flow cytometry, post Annexin V-FITC/PI double staining. As seen in Figure 6, with increasing concentrations of 14f, the number of apoptotic cells increased from 3.89% to 9.80%, 14.38%, and 28.13%, while the number of   surviving cells decreased correspondingly, in a concentrationdependent manner. These data demonstrated that 14f promoted apoptosis in HepG2 cells as well.

14f decreased the MMP (dwm) of HepG2 cells
Mitochondrion plays an important role in regulating cellular functions. Loss of MMP is a sign of the apoptotic process in cells. MMP collapse represents the initiation and activation of apoptosis by the intrinsic pathway 33 . In the present study, JC-1 staining was conducted to investigate whether 14f-induced HepG2 cell apoptosis was associated with the loss of MMP. As seen in Figure 7, there was a 14f dose-dependent increase in the percentage of mitochondrial depolarisation (7.90% at 5 mM, 15.55% at 10 mM, and 40.46% at 20 mM). The result indicated that 14f induced apoptosis in HepG2 cells by destroying the mitochondrial integrity and decreasing MMP.

14f induced apoptosis in HepG2 cells through the mitochondrial pathway
Based on the results, we hypothesised that 14f might be a MMP disruptor, and further examined the effect of 14f on mitochondrial pathway-related factors (such as anti-apoptotic protein Bcl-2 and pro-apoptotic protein Bax) and apoptotic effectors (such as caspase 9 and caspase 3), which play important roles in cell apoptosis 22 . Firstly, the expression of mitochondrial pathway-related genes was evaluated using qRT-PCR. As seen in Figure 8(A), 14f treatment led to high gene expression of Bax, caspase 9, and caspase 3, while the gene expression of Bcl-2 was suppressed. Next, based on the qRT-PCR results, ELISA was further used to confirm whether the apoptosis-related enzymes, caspase 9 and caspase 3, are involved in the regulation of apoptosis. The results indicated that 14f upregulated the enzyme activity of caspase 9 and caspase 3 (Figure 8(B)). Finally, western blot was conducted to investigate the protein expression levels of Bax and Bcl-2. As shown in      (Figure 8(C)). Moreover, the quantitative analysis results were consistent with the trend of mRNA level expression (Figure 8(D)). These results suggested that 14f triggered the mitochondrial pathway of apoptosis in HepG2 cells.

Molecular docking studies
Based on the downregulation of Bcl-2 by hybrid 14f, we next explored the potential binding mode of 14f with Bcl-2. High-affinity binding of pro-apoptotic proteins to Bcl-2 is largely mediated by protein-protein interactions in the P2 and P4 hydrophobic pockets. Thus, the Bcl-2-inhibiting inhibitory potency of small molecules is determined by the interactions of these inhibitors with P2 and P4 pockets 34,35 . Venetoclax, also known as ABT-199, was the first Bcl-2-selective inhibitor 35 . Our docking research showed that the tetrahydro-1, 1 0 -biphenyl fragment and 1H-pyrrolo[ [2,3-b]] pyridine fragment of ABT-199 could occupy the P2 and P4 pockets, respectively ( Figure 9). Coincidentally, similar behaviours could be observed in the bonding mode of 14f to Bcl-2 protein. 14f could lay on the deep hydrophobic binding pocket in the Bcl-2 groove, with the benzoic acid mustard moiety accommodating the P2 pocket and the acetyl moiety entering the P4 pocket. The experimental results showed that 14f might have the ability to interact with Bcl-2, thereby inducing HepG2 cells apoptosis.
Altogether, the above results indicated that 14f arrested the cell cycle at the G0/G1 phase and induced apoptosis in HepG2 cells through the mitochondrial apoptosis pathway. To comprehend the exact regulatory mechanism of 14f against hepatoma, there is a need to carry out well-designed in vivo studies in the future.

Conclusion
In summary, fifteen novel DSG-benzoic acid mustard hybrids (8-10, 14a-14f, and 15a-15f) were designed and synthesised in the present study. This was followed by testing the cytotoxicity of these target hybrids against human cancer HepG2, MCF-7, and HeLa cell lines and human normal GES-1 cells. Most of them did not display significant cytotoxic activity. Several of these hybrids exhibited moderate to potent inhibitory activities against HepG2 cells. Among them, hybrid 14f (IC 50 ¼ 2.26 mM) exhibited the most potent anti-proliferative activity in HepG2 cells, and displayed an efficacy that was 14.4-fold higher than that of DSG (IC 50 ¼ 32.63 mM). Moreover, 14f exhibited good anti-proliferative selectivity between normal and tumour cells. SAR studies showed that the introduction of homopiperazinyl between DSG derivative 11 and benzoic acid mustard was beneficial for enhancing the anti-proliferative activity. Further investigations on the anti-tumour mechanism in HepG2 cells indicated that 14f arrested the cell cycle at the G0/G1 phase, by regulating the cell cycle-related proteins (CDK4, CDK6, cyclin D1, CDK2, and cyclin E1), and induced apoptosis by mediating the apoptosis-related proteins (Bax, Bcl-2, caspase 9, and caspase 3). Furthermore, molecular docking revealed that hybrid 14f could bind at the deep hydrophobic binding pocket in the Bcl-2 groove, indicating that the potential target of 14f might be Bcl-2. Therefore, conjugating DSG with benzoic acid mustard might be an efficient strategy for the modification of this class of natural products.

Disclosure statement
No potential conflict of interest was reported by the author(s).