Synthesis and Evaluation of Anticancer Activity of New 4-Acyloxy Derivatives of Robustic Acid

In the present study, a series of 4-acyloxy robustic acid derivatives were synthesized and characterized for evaluation of their anti-cancer activity. The structures of these derivatives were elucidated by mass spectra (MS) nuclear magnetic resonance spectra (NMR). The single-crystal X-ray diffraction structure of one of these compounds was obtained, for further validation of the target compound structures. The anticancer activities of the target products were evaluated against human leukemic cells HL-60, human non-small cell lung carcinoma cells A-549, human hepatic carcinoma cells SMMC-7721, human hepatocellular carcinoma cells HepG2, and human cervical carcinoma cells Hela. Three compounds among them exhibited potent in-vitro cytotoxicity and excellent DNA topoisomerase I inhibitory activity, even at 0.1 mM concentrations. The most noteworthy observation was the minor toxicity of two of these compounds to normal cells, with an activity similar to the positive control in cancerous cells. A Surflex-Dock docking study was performed to investigate the topoisomerase I activity of all compounds. Of all the other compounds, the most sensitive compound was selected for further investigation of its effect on apoptosis induction and cell cycle regulation in HL-60 cells. Our results suggest that the anticancer effects of these compounds can be attributed to their pharmacological effects on topoisomerase I, cell apoptosis, and cell cycle. These findings suggest that robustic acid derivatives could be used as potential antitumor drugs.


Introduction
Coumarin compounds are widely distributed in nature, especially in higher plants [1]. They are proven to possess a wide range of biological activities, with minimal side effects [2]. In particular, pyranocoumarins (including linear type and angular type) are an interesting class of fused coumarins. They are important as photoactive drugs with important pharmacological effects that include anti-inflammatory [3], anti-HIV-1 [4], anti-hepatitis B virus [5], anti-cancer [6], and antinociceptive properties [7]. As the major constituents from Clausena or Angelica genus plants [8][9][10][11], some linear-type are reported to exhibit potent cytotoxicity against human non-small cell lung carcinoma cells (A549), human breast adeno carcinoma cells (MCF-7), human nasopharyngeal carcinoma cells (KB), and multidrug-resistant nasopharyngeal carcinoma cells (KB-VIN) [5]. Compound 4 had significant antitumor activity on MCF-7, human hepatocellular carcinoma cells (HepG2), and human prostate carcinoma cells (PC-3 and LNCaP) through causing the induction of apoptosis [12][13][14]. Compounds 5 and 6 are more cytotoxic to leukemia cancer cells [15,16]. However, detailed information on the cytotoxic mechanism and structure-activity relationships of these linear-type pyranocoumarins is still limited in the literature. Considering the wide spectrum of biological activities of these compounds, the synthesis of new linear pyranocoumarins has drawn significant attention in recent years. Recently, our group isolated and identified a new linear pyranocoumarin was from Dalbergia benthamii Prain: robustic acid (RA; Figure 2). RA constitutes to about 3.5% w/w of its dry matter and was considered to be one of the major bioactive phytoconstituents of D. benthamii [17]. D. benthamii is used commonly in traditional medicine for the treatment of traumatic injury, as a pain reliever and anti-inflammatory agent, and for regulating menstrual cycle [18]. To further explore and continue our earlier investigation on linear pyranocoumarins, the synthesis of a new series of 4-acyloxy compounds of robustic acid derivatives was carried out. Preliminary cytotoxic activities and related mechanisms such as DNA topoisomerase I (Topo I), cell cycle, and apoptosis were also investigated in this study.  Recently, our group isolated and identified a new linear pyranocoumarin was from Dalbergia benthamii Prain: robustic acid (RA; Figure 2). RA constitutes to about 3.5% w/w of its dry matter and was considered to be one of the major bioactive phytoconstituents of D. benthamii [17]. D. benthamii is used commonly in traditional medicine for the treatment of traumatic injury, as a pain reliever and anti-inflammatory agent, and for regulating menstrual cycle [18]. To further explore and continue our earlier investigation on linear pyranocoumarins, the synthesis of a new series of 4-acyloxy compounds of robustic acid derivatives was carried out. Preliminary cytotoxic activities and related mechanisms such as DNA topoisomerase I (Topo I), cell cycle, and apoptosis were also investigated in this study.
benthamii Prain: robustic acid (RA; Figure 2). RA constitutes to about 3.5% w/w of its dry matter and was considered to be one of the major bioactive phytoconstituents of D. benthamii [17]. D. benthamii is used commonly in traditional medicine for the treatment of traumatic injury, as a pain reliever and anti-inflammatory agent, and for regulating menstrual cycle [18]. To further explore and continue our earlier investigation on linear pyranocoumarins, the synthesis of a new series of 4-acyloxy compounds of robustic acid derivatives was carried out. Preliminary cytotoxic activities and related mechanisms such as DNA topoisomerase I (Topo I), cell cycle, and apoptosis were also investigated in this study.

Chemistry
The scheme used for the synthesis of 4-acyloxy robustic acid derivatives (2a-2j) is illustrated in Scheme 1. The starting material RA was isolated from D. benthamii in our previous study [17]. The target compounds 2a-2j were synthesized by nucleophilic substitution from RA and derivatives of acyl chloride. We found that the yields of aromatic acyl chloride compounds were higher than the fatty acyl chloride compounds, which ranged from 41-85%. With the attachment of the withdrawing group on aromatic ring, the reaction time was shortened to less than 1 h and the yield was greatly improved. The structures of compounds were elucidated by, MS, 1 H NMR, and 13 C NMR. All spectral data were in accordance with the assumed structures, and are listed in the procedure. The ESI-MS indicated that the molecular weights were in accordance with the calculated values. In 1 H NMR spectra, the single peaks of two methoxyl groups of the original nucleus appeared at δ 3.79 and δ 3.56, and six hydrogen atom signals of two methyl groups were observed at δ 1.47-1.59 ppm. Two pyran ring hydrogen signals were in the ranges δ 6.02-6.58 and δ 5.50-8.74, respectively. Further verification of the structures of the target products was carried out, and a single crystal for 2f(CCDC 1944893, these data can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html or the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.) was obtained in ethanol solution. The molecular structure of 2f is shown in Figure 3, and the packing diagram of 2f is depicted in Figure 4. The formation of a twin crystal was observed in the crystal structure of 2f. Most bond lengths in the system fell in the range of single and double bonds and the C-C bond distances in all rings ranged between 1.316(5) and 1.498(6) Å. These values are almost equal to the values of typical bonds of aromatic structure. The crystal packing of the compound is stabilized by π-π stacking. The scheme used for the synthesis of 4-acyloxy robustic acid derivatives (2a-2j) is illustrated in Scheme 1. The starting material RA was isolated from D. benthamii in our previous study [17]. The target compounds 2a-2j were synthesized by nucleophilic substitution from RA and derivatives of acyl chloride. We found that the yields of aromatic acyl chloride compounds were higher than the fatty acyl chloride compounds, which ranged from 41-85%. With the attachment of the withdrawing group on aromatic ring, the reaction time was shortened to less than 1 h and the yield was greatly

Anti-Proliferative Activity Screening
All the synthesized compounds were evaluated for their cytotoxicity effects in five human cancer cell lines and two normal cells. The cell lines included: leukemia cells HL-60, human non-small cell lung carcinoma cells A549, human hepatic carcinoma cells SMMC-7721, hepatocellular carcinoma cells HepG2, cervical carcinoma cells Hela, human normal liver cells LO2, and human normal lung epithelial cells BEAS-2B. The obtained IC50 values of the synthesized compounds, compared to RA and the reference drug (cisplatin and camptothecin (CPT)), are shown in Table 1. Interestingly, the activities of most of the target products were superior to the parent nucleus. Compounds 2d, 2g, and 2i exhibited promising biological activity in almost all cell lines. Of all the cell lines selected, they were more sensitive to HL-60, with IC50 values of 21.04 ± 0.43, 16.63 ± 0.12, and 16.38 ± 0.27 μM, respectively. The obtained IC50 values were close to the positive control cisplatin (14.23 ± 0.56) and CPT (12.25 ± 1.06). In addition, 2d and 2g exhibited moderate activity in Hela cells, with IC50 values of 37.03 ± 0.97 and 27.45 ± 0.38 μM, respectively. Their potential antitumor activity might be related to the attracting group on benzene ring of 4-aroyl. It should be noted that most target compounds

Anti-Proliferative Activity Screening
All the synthesized compounds were evaluated for their cytotoxicity effects in five human cancer cell lines and two normal cells. The cell lines included: leukemia cells HL-60, human non-small cell lung carcinoma cells A549, human hepatic carcinoma cells SMMC-7721, hepatocellular carcinoma cells HepG2, cervical carcinoma cells Hela, human normal liver cells LO2, and human normal lung epithelial cells BEAS-2B. The obtained IC50 values of the synthesized compounds, compared to RA and the reference drug (cisplatin and camptothecin (CPT)), are shown in Table 1. Interestingly, the activities of most of the target products were superior to the parent nucleus. Compounds 2d, 2g, and 2i exhibited promising biological activity in almost all cell lines. Of all the cell lines selected, they were more sensitive to HL-60, with IC50 values of 21.04 ± 0.43, 16.63 ± 0.12, and 16.38 ± 0.27 μM, respectively. The obtained IC50 values were close to the positive control cisplatin (14.23 ± 0.56) and CPT (12.25 ± 1.06). In addition, 2d and 2g exhibited moderate activity in Hela cells, with IC50 values of 37.03 ± 0.97 and 27.45 ± 0.38 μM, respectively. Their potential antitumor activity might be related to the attracting group on benzene ring of 4-aroyl. It should be noted that most target compounds

Anti-Proliferative Activity Screening
All the synthesized compounds were evaluated for their cytotoxicity effects in five human cancer cell lines and two normal cells. The cell lines included: leukemia cells HL-60, human non-small cell lung carcinoma cells A549, human hepatic carcinoma cells SMMC-7721, hepatocellular carcinoma cells HepG2, cervical carcinoma cells Hela, human normal liver cells LO2, and human normal lung epithelial cells BEAS-2B. The obtained IC 50 values of the synthesized compounds, compared to RA and the reference drug (cisplatin and camptothecin (CPT)), are shown in Table 1. Interestingly, the activities of most of the target products were superior to the parent nucleus. Compounds 2d, 2g, and 2i exhibited promising biological activity in almost all cell lines. Of all the cell lines selected, they were more sensitive to HL-60, with IC 50 values of 21.04 ± 0.43, 16.63 ± 0.12, and 16.38 ± 0.27 µM, respectively. The obtained IC 50 values were close to the positive control cisplatin (14.23 ± 0.56) and CPT (12.25 ± 1.06). In addition, 2d and 2g exhibited moderate activity in Hela cells, with IC 50 values of 37.03 ± 0.97 and 27.45 ± 0.38 µM, respectively. Their potential antitumor activity might be related to the attracting group on benzene ring of 4-aroyl. It should be noted that most target compounds could significantly inhibit the proliferation of selected cancerous cells with no damage to normal cells (LO2 and BEAS-2B (IC 50 > 100 µM)). However, compound 2g has detrimental effects on LO2 cells, as it exhibited an IC 50 value of 20.79 ± 0.97 µM. All compounds were less toxic on non-cancerous cells than positive control. This indicates that these compounds (especially 2d and 2i) have highly selective activity against cancerous cells and are non-toxic to normal human cell lines. They can thus be suggested as potential candidates for development into anti-tumor drugs with high efficiency and low toxicity.

Evaluation of DNA Topoisomerase I (Topo I) Inhibitory Activity and Molecular Docking Study
DNA Topo I is an important enzyme in all living organisms, and it participates in many cellular metabolic processes, such as replication, transcription, recombination, and repair [19]. DNA Topo I has been established as a molecular target of anticancer drugs. A combined study of synthesis and docking for the anti-cancer activity of novel coumarins concluded that coumarin derivatives are potent cytotoxic agents. Docking studies further suggest that all coumarins possess higher binding affinities with DNA Topo I [20]. Currently, discovering novel non-camptothecin inhibitors for DNA Topo I has become an important area of research. Thus, a DNA Topo I inhibitory activity assay was also carried out using the supercoiled DNA unwinding method in this study [21]. Figure 5A shows the conversion of supercoiled plasmid DNA Topo I in the presence of compounds RA and 2a to 2j with camptothecin (CPT) as a positive control. Compounds 2d, 2g, and 2i showed significant inhibitory activities at 10 mM concentration. Compounds 2b, 2c, 2f, and 2h showed moderate inhibitory activities with obscure strips in electrophoresis. Compounds 2a, 2e, and 2j did not show any considerable Topo I inhibitory activity. Therefore, compounds 2d, 2g, and 2i were selected for further quantitative investigation (depicted in Figure 5B). It was found that the inhibition weakened gradually as the concentration of RA reduced. At critical concentrations lower than 10 mM, no inhibitory activity was found. In agarose gel electrophoresis, DNA fragmentation was evident with treatment of 2d, 2g, and 2i, even at low concentrations of 1 mM or 0.1 mM. Note that 2g and 2i exhibited Topo I inhibitory activities higher than that of compound 2d. This correlates well with the cytotoxicity to HL-60, A-549, and SMMC-7721 cells. Thus, we infer that the anti-proliferative activity of 4-acyloxy derivatives of RA to these three cells may be related to Topo I.
To understand the mechanism of Topo I inhibitory activity of the compounds synthesized, molecular docking studies were carried using the Surflex-Dock algorithm of Sybyl-X 2.0 (Tripos Inc., St. Louis, MO, USA). The crystal structure of DNA Topo I in complex with CTP was downloaded from the Protein Data Bank (PDB code: 1T8I). To validate the molecular docking approach, the crystallographic pose of CTP (derived from complex) and the top-ranked docking pose obtained in this study were compared. The two binding poses of CTP had almost perfect superposition, with an RMSD (root-mean-square deviation) value of 0.4438 Å ( Figure 6). The obtained RMSD value is far below the well-established tolerance level of 2.0 Å, thus validating the adopted docking methodology [22,23]. To understand the mechanism of Topo I inhibitory activity of the compounds synthesized, molecular docking studies were carried using the Surflex-Dock algorithm of Sybyl-X 2.0 (Tripos Inc., St. Louis, MO, USA). The crystal structure of DNA Topo I in complex with CTP was downloaded from the Protein Data Bank (PDB code: 1T8I). To validate the molecular docking approach, the crystallographic pose of CTP (derived from complex) and the top-ranked docking pose obtained in this study were compared. The two binding poses of CTP had almost perfect superposition, with an RMSD (root-mean-square deviation) value of 0.4438 Å ( Figure 6). The obtained RMSD value is far below the well-established tolerance level of 2.0 Å, thus validating the adopted docking methodology [22,23]. After running Surflex-Dock, the binding affinities of protein-ligand complexes were obtained and expressed as total score. All 10 inhibitors were docked into the active site of Topo I, as shown in Figure 7. The predicted binding affinities (total score) of all the compounds containing hydrophobic residue, hydrogen bond, and π-π stacking are listed in Table 2. The results indicate that all synthetic compounds except 2b had better combination with Topo I than the RA mother nucleus. Compound 2g exhibited the highest total score value of 10.02, which was close to the positive control (CPT). The most important residues on the 2g-Topo I complex ( Figure 4h) were ILE427, TGP11, DT10, DA113, ARG364, DC112, PRO357, GLU356, LYS374, GLU418, and TYR426. As depicted, the oxygen atom of acyl group of compound 2g made a hydrogen bonding interaction with the hydrogen of the -NH2  To understand the mechanism of Topo I inhibitory activity of the compounds synthesized, molecular docking studies were carried using the Surflex-Dock algorithm of Sybyl-X 2.0 (Tripos Inc., St. Louis, MO, USA). The crystal structure of DNA Topo I in complex with CTP was downloaded from the Protein Data Bank (PDB code: 1T8I). To validate the molecular docking approach, the crystallographic pose of CTP (derived from complex) and the top-ranked docking pose obtained in this study were compared. The two binding poses of CTP had almost perfect superposition, with an RMSD (root-mean-square deviation) value of 0.4438 Å ( Figure 6). The obtained RMSD value is far below the well-established tolerance level of 2.0 Å, thus validating the adopted docking methodology [22,23]. After running Surflex-Dock, the binding affinities of protein-ligand complexes were obtained and expressed as total score. All 10 inhibitors were docked into the active site of Topo I, as shown in Figure 7. The predicted binding affinities (total score) of all the compounds containing hydrophobic residue, hydrogen bond, and π-π stacking are listed in Table 2. The results indicate that all synthetic compounds except 2b had better combination with Topo I than the RA mother nucleus. Compound 2g exhibited the highest total score value of 10.02, which was close to the positive control (CPT). The most important residues on the 2g-Topo I complex (Figure 4h) were ILE427, TGP11, DT10, DA113, ARG364, DC112, PRO357, GLU356, LYS374, GLU418, and TYR426. As depicted, the oxygen atom of acyl group of compound 2g made a hydrogen bonding interaction with the hydrogen of the -NH2 After running Surflex-Dock, the binding affinities of protein-ligand complexes were obtained and expressed as total score. All 10 inhibitors were docked into the active site of Topo I, as shown in Figure 7. The predicted binding affinities (total score) of all the compounds containing hydrophobic residue, hydrogen bond, and π-π stacking are listed in Table 2. The results indicate that all synthetic compounds except 2b had better combination with Topo I than the RA mother nucleus. Compound 2g exhibited the highest total score value of 10.02, which was close to the positive control (CPT). The most important residues on the 2g-Topo I complex (Figure 4h) were ILE427, TGP11, DT10, DA113, ARG364, DC112, PRO357, GLU356, LYS374, GLU418, and TYR426. As depicted, the oxygen atom of acyl group of compound 2g made a hydrogen bonding interaction with the hydrogen of the -NH 2 group of LYS425 (O-H-LYS425, 2.76 Å). The oxygen atoms in the -NO 2 group made three binding interactions with hydrogens of the -NH 2 group of MET428 (O-H-MET428, 2.72, 3.06, and 3.33Å). Compounds 2d and 2i also exhibited good binding affinities, with total scores of 8.76 and 8.89. All these compounds exhibited remarkable Topo I inhibitory activity even at low concentration, which is nearly in agreement with our observations of in vitro studies. However, the results of the docking assay did not exactly correlate with the Topo I inhibitory activity. The binding affinity of all compounds were decreasing in sequence of 2g, 2h, 2i, 2c, 2d, 2e, 2f, 2a, 2j, 2b. However, the inhibitory activities of 2h and 2c were not as good as the prediction data. This may be related to other combination modes or other pharmacological mechanisms. π-π ASP533 Hydrogen Bond TGP11 π-π a Total score represents the binding affinity in units of -logK d .

Apoptosis and Cell-Cycle Analysis
Cell-cycle arrest and apoptosis are widely accepted as the main mechanisms of inhibition in tumor formation [19,20]. As HL-60 cells were the most sensitive to all tested compounds, they were selected for testing the most active compound 2i (at 6.25-, 12.5-, and 25-µM doses), for their effects on apoptosis and cell cycle profile after 48 h. The induction of apoptosis by 2i was evident from the nuclear morphologic alterations. These were observed by confocal microscopy after labeling the cells with acridine orange (AO)/ethidium bromide (EB). Representative images from four independent experiments are shown in Figure 8. Compared to the control group, chromatin condensation, nuclear fragmentation, and apoptotic bodies were observed in treated cells at concentrations of 6.25, 12.5, and 25 µM. At the highest concentration of 25 µM, the necrosis cell volume was increased, showing uneven orange-red fluorescence and an unapparent outline.  The extent of apoptosis by flow cytometric analysis of the HL-60 cells labeled with Annexin V-APC/7-AAD was further quantified. For this study, treatment with 2i at doses of 6.25, 12.5, and 25 μM for 48 h was carried out. Such treatments resulted in 34.61%, 52.52%, and 66.17% apoptotic cells, respectively ( Figure 9). The apoptosis ratio for control was found to be 1.46%. The obtained results thus indicate that compound 2i suppressed cell proliferation by inducing apoptosis, and a considerable dose dependence was observed at concentrations of 6.25, 12.5, and 25 μM.
Since the induction of apoptosis may be mediated through the regulation of cell cycle, we also examined the effect of 2i on cell cycle perturbations labeled with propidium iodide (PI). Compared to control, 2i treatment resulted in an appreciable arrest of HL-60 cells in the G1 phase of the cell cycle, after 48 h of the treatment (Figure 10). The treatment caused an arrest of 56.48% cells in G1 phase of the cell cycle at a 6.25 μM dose, which further increased to 63.40%and 68.29% at higher doses of 12.5 and 25 μM. However, this increase in G1 cell population was accompanied by a decrease of cell number in S and G2 phases. These results indicated that compound 2i induced a significant cell cycle arrest in G1 phase in a concentration-dependent manner. The extent of apoptosis by flow cytometric analysis of the HL-60 cells labeled with Annexin V-APC/7-AAD was further quantified. For this study, treatment with 2i at doses of 6.25, 12.5, and 25 µM for 48 h was carried out. Such treatments resulted in 34.61%, 52.52%, and 66.17% apoptotic cells, respectively ( Figure 9). The apoptosis ratio for control was found to be 1.46%. The obtained results thus indicate that compound 2i suppressed cell proliferation by inducing apoptosis, and a considerable dose dependence was observed at concentrations of 6.25, 12.5, and 25 µM.
Since the induction of apoptosis may be mediated through the regulation of cell cycle, we also examined the effect of 2i on cell cycle perturbations labeled with propidium iodide (PI). Compared to control, 2i treatment resulted in an appreciable arrest of HL-60 cells in the G1 phase of the cell cycle, after 48 h of the treatment (Figure 10). The treatment caused an arrest of 56.48% cells in G1 phase of the cell cycle at a 6.25 µM dose, which further increased to 63.40%and 68.29% at higher doses of 12.5 and 25 µM. However, this increase in G1 cell population was accompanied by a decrease of cell number in S and G2 phases. These results indicated that compound 2i induced a significant cell cycle arrest in G1 phase in a concentration-dependent manner. Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 11 of 18

Materials and Methods
All chemicals used in the study were reagent grade and are commercially available. All yields refer to the yields of isolated products after purification. NMR spectra were measured on a Bruker (E)

Materials and Methods
All chemicals used in the study were reagent grade and are commercially available. All yields refer to the yields of isolated products after purification. NMR spectra were measured on a Bruker DRX-500 ( 1 H: 500 MHz, 13 C: 126 MHz) using CDCl 3 as solvent. Chemical shifts (d) are expressed in parts per million (ppm), and J values are given in hertz (Hz). The mass spectra (MS) were obtained on an Agilent 6210 ESI/TOF MS instrument. Melting points were determined using an X-4 apparatus, and were uncorrected. Samples of compounds 2a-2i are available from the authors.

General Procedure for the Synthesis of 4-Acyloxy Derivatives of Robustic Acid 2a-2j
Robustic acid (0.19 g, 0.5 mmol) was dissolved in CH 3 CN (10 mL), followed by the addition of acyl chloride derivative (1 mmol) and pyridine (0.161 µL). After 15-210 min refluxing with stirring (monitored by thin-layer chromatography), the reaction mixture was cooled to room temperature. The solvent was removed and recrystallization with ethanol was carried out. The formed crystals (2a-2j) were filtered and dried under vacuum.

Antiproliferative Activity
The MTS (3-(4,5-dimethylthiazole-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium) (Promega, Madison, WI, USA) assay was used to determine the IC 50 value for all compounds, according to [26]. We used a panel of five human carcinoma cell lines and two normal cells including HL-60, A549, SMMC-7721, HepG2, Hela, LO2 and BEAS-2B. Cells were seeded in 96-well plates at a density of 3 × 10 3 -1.5 × 10 4 cells/well in 100 µL medium (DMEM or RMPI1640) and incubated for 12-24 h at 37 • C. Subsequently, cells were exposed to drugs in an additional 100 µL medium and were further incubated at 37 • C for 48 h. In brief, 100 µL MTS reagent supplement with 100 µL medium was added directly to the wells and cells were incubated at 37 • C for a minimum of 2 h. Colorimetric measurements were recorded at 492 nm. Background absorbance was corrected using triplicate sets of wells, and cells were incubated at 37 • C for a minimum of 2 h. Colorimetric measurements were recorded at 492 nm. Background absorbance was corrected using triplicate sets of wells containing medium only (no cells), and MTS reagent was added per experimental well. Each experiment was performed at least three times, and values in results are represented as mean ± SD.

Topo I Inhibitory Activity
Supercoiled pBR32 DNA was used as a substrate to determine the Topo I catalytic activity. Commercial samples of Topo I and pBR322 were obtained from Takara Bio Inc. The buffer solutions were prepared using 500 mM KAc, 200 mM Tris-Ac, 100 mM Mg(Ac) 2 , and 1 mg/mL BSA. The procedure used for determination of enzyme inhibitory activity was similar to one reported by Xu et al. [27].

General Procedure for AO/EB Staining
Apoptosis was determined by nuclear morphology. Cells were fixed and stained with AO/EB according to the manufacturer's instruction (KGA501, KeyGEN BioTECH, Nanjing, China). HL-60 cells were seeded in six-well plates with sterile cover slips. The concentration of cells used per well was 5 × 10 5 -6 × 10 6 cells/2 mL. To facilitate growth, the medium was replaced with fresh medium (RPMI1640) plus 10% fetal bovine serum. The fresh medium was supplemented with compound 2i (6.25, 12.50, and 25 µM). Post treatment, 25 µL of cell suspension was collected on a glass slide by inverting the cover slip and stained with 10 µL of AO/EB stain (100 mg/mL). Finally, stained nuclei were observed immediately under a fluorescence microscope (BX41, Nikon, Tokyo, Japan).

General Procedure for Apoptosis Ratio Determination
For determination of apoptosis ratios by application of flow cytometry analysis, the manufacturer's protocol (KGA1024, KeyGEN BioTECH, Nanjing, China) for Annexin-V APC/7-AAD double-stain assay was used. The HL-60 cells were grown at a density of 5 × 10 5 cells/mL in six-well plates (Costar), and were treated with 2i (6.25, 12.50, and 25 µM) for 72 h. The cells were trypsinized and washed with PBS. Subsequently, the prepared HL-60 cells (5 × 10 5 cells/mL) were collected and resuspended in 500 µL of binding buffer containing 5 µL of Annexin V-APC. The suspension was shaken well, and 5 µL of 7-AAD was added to it. The suspension was then incubated for 5-15 min in the dark at room temperature. The suspensions were then analyzed by flow cytometry (FACS Cali-bur, Becton Dickinson, San Jose, CA, USA).

General Procedure for Cell Cycle Analysis
The intercalating blue-excited dye propidium iodide (PI) was used in this study. HL-60 cell cultures were seeded within six-well plates (Costar). Then, the cells were treated with 2i (6.25, 12.50, and 25 µM) in DMEM complete medium for 72 h. After the incubation time elapsed, the harvest procedure was performed as follows: the supernatant medium was transferred to centrifuge tubes. To the medium, 0.25% trypsin (Sigma) was added, until all cells were detached. The cell suspension was then washed twice with PBS and resuspended. About 1 mL of 70% ethanol was quickly added to the cell suspension (5 × 10 5 cells/mL), which was then incubated at 4 • C overnight. The samples were treated with 100 µL RNase A at 37 • C for 30 min and then stained with 400 µL PI solution. After another 30 min of incubation at 4 • C, the resulting nuclei suspension was analyzed with a flow cytometer (FACS Verse, Becton Dickinson, San Jose, CA, USA) at 488 nm excitation for red fluorescence of PI.

Molecular Docking
Surflex-Dock in Sybyl 2.0 was used for molecular docking. This system performs molecular docking functions aided by the generation of an idealized active site (Protocol), consisting of dummy atoms that guide the docking process. The crystal structures of DNA-Topo I complex was downloaded from the RCSB website (https://www.rcsb.org/structure/1T8I) (PDB ID: 1T8I). The proteins were then imported into the Surflex-Dock and prepared for docking using the biopolymer preparation tool, according to the following criteria: H-Addition, H-Bond, removal of water molecules, termini treatment, charge, protonation.
The structures of the compounds were sketched using ChemBioOffice version 14.0 by PerkinElmer (Waltham, MA, USA). The alignments of the training set molecules were derived using FlexS in SYBYL. All values were filled with valence, and Gasteiger-Marsili charges were calculated for each compound. Ultimately, ligand docking under the Surflex-Dock GeomX precision was performed to generate a grid of proteins and ligands. The docking results were then imported into LigPlot + 2.1, and the combinations between compounds and proteins were discussed in terms of H-Bond, conjugate action, and hydrophilic or hydrophobic action.

Conclusions
A new series of linear pyranocoumarin derivatives were synthesized, and their anti-cancer activities were evaluated. The biological screening results indicated that most of the synthesized compounds exhibited potent in vitro anti-cancer activity, which were better than original nucleus robustic acid. Among them, compounds 2d, 2g, and 2i were showed excellent anti-cancer activity with respective IC 50 values of 21.04 ± 0.43, 16.63 ± 0.12, and 16.38 ± 0.27 µM against HL-60 cells. These values were estimated in comparison to the positive control cisplatin (14.23 ± 0.56 µM) and CPT (12.25 ± 1.06 µM). And these compounds were much less toxic to non-cancerous cells when comparing with the two positive controls. In addition to this, docking studies of 2d, 2g, and 2i were performed, and they all exhibited good binding affinities with relatively high total scores of 8.76, 10.02, and 8.89. The inhibitory activities of the compounds against Topo I indicated that 2g and 2i were showed significant inhibitory activity, even at the low concentration of 0.1 mM. Compound 2d aslo exhibited moderate inhibitory activity. These results indicate that the anti-tumor activity of these three compounds may be related to Topo I inhibitory activity. However, the results of the docking assay of other compounds do not exactly correlate with the Topo I inhibitory and anti-tumor activity. This may be the result of other pharmacological mechanisms. Later, compound 2i was selected as the representative compound for further investigation, due to its properties of suppressing cell proliferation by inducing apoptosis and arresting HL-60 cells in the G1 stage. These findings may contribute to the identification of novel linear pyranocoumarins that can serve as the new lead compounds for anti-tumor drugs.
Author Contributions: L.H. and J.W. contributed to the separation and synthesis of compounds. R.C., J.Z., X.X., and Y.L. performed pharmacological activity testing and docking study of the target products. L.H. and R.C. wrote the manuscript. D.L. contributed to the project proposal and executing the experiment. Y.J. and L.W. contributed to improving the manuscript, providing intellectual input, and editing the language of the manuscript.
Funding: This research received no external funding.