Preparation, COX-2 Inhibition and Anticancer Activity of Sclerotiorin Derivatives

The latest research has indicated that anti-tumor agents with COX-2 inhibitory activity may benefit their anti-tumor efficiency. A series of sclerotiorin derivatives have been synthesized and screened for their cytotoxic activity against human lung cancer cells A549, breast cancer cells MDA-MB-435 using the MTT method. Among them, compounds 3, 7, 12, 13, 15, 17 showed good cytotoxic activity with IC50 values of 6.39, 9.20, 9.76, 7.75, 9.08, and 8.18 μM, respectively. In addition, all compounds were tested in vitro the COX-2 inhibitory activity. The results disclosed compounds 7, 13, 25 and sclerotiorin showed moderate to good COX-2 inhibition with the inhibitory ratios of 58.7%, 51.1%, 66.1% and 56.1%, respectively. Notably, compound 3 displayed a comparable inhibition ratio (70.6%) to the positive control indomethacin (78.9%). Furthermore, molecular docking was used to rationalize the potential of the sclerotiorin derivatives as COX2 inhibitory agents by predicting their binding energy, binding modes and optimal orientation at the active site of the COX-2. Additionally, the structure-activity relationships (SARS) have been addressed.


Introduction
Cancer has become one of the most important factors affecting human life and health in terms of incidence, mortality, and prevalence. In 2018, estimates for global statistics on cancer rates show that there were 18.1 million new cases and 9.6 million deaths. Lung cancer is the most frequent cancer and the leading cause of cancer death among males, followed by prostate and colorectal cancer. Among females, breast cancer is the most commonly diagnosed cancer and the leading cause of cancer death [1]. Cancer metastasis is the major reason of treatment failure and death [2,3]. As the research moves along, authoritative studies have shown that an inflammatory environment plays an important role in various stages of tumor development and affects the body response to chemotherapeutic agents [4,5]. Chronic inflammation is associated with tumor development, and inflammatory mediators are present in the tumor microenvironment, including cytokines, growth factors, reactive oxygen species and reactive nitrogen species [6][7][8]. These mediators also activate signaling molecules involved in inflammation and carcinogenesis, including nuclear transcription factor, inducible nitric oxide synthase and cyclooxygenase-2. All these factors together result in tumor initiation by increasing cell cycling, inhibiting tumor suppressor pathways and activating oncogenes [8][9][10].
COX-2 is a member of the cyclooxygenase family, and has long been a research focus in the treatment of inflammation. It is the key enzyme in the conversion of arachidonic Schemes 1 and 2. Twenty-four amide-derivatives (1-24) have been successfully sy sized by one-step reaction of sclerotiorin with various amines in high yields (up to The derivative 25 was obtained by hydrolysis of sclerotiorin (yields up to 85%). Follo by esterification reaction of propionic anhydride and glutaric anhydride produced 2 27 in 82% and 78% yields, respectively. The resulting extract was subjected to silic column chromatography to obtain the pure products. The structures of target compo 1-27 were confirmed by extensive spectroscopic methods including 1 H NMR, 13  Schemes 1 and 2. Twenty-four amide-derivatives (1-24) have been successfully synthesized by one-step reaction of sclerotiorin with various amines in high yields (up to 90%). The derivative 25 was obtained by hydrolysis of sclerotiorin (yields up to 85%). Followed by esterification reaction of propionic anhydride and glutaric anhydride produced 26 and 27 in 82% and 78% yields, respectively. The resulting extract was subjected to silica gel column chromatography to obtain the pure products. The structures of target compounds 1-27 were confirmed by extensive spectroscopic methods including 1 H NMR, 13

Cytotoxic Activity of Sclerotiorin Derivatives
The cytotoxic activity of all compounds was evaluated against A549 (human lung cancer) and MDA-MB-435 (breast cancer cells) by using the MTT method as described previously.
As shown in Table 1, the four yellow pigments, sclerotiorin and the other three acyl changed derivatives 25, 26, 27, present no cytotoxic activity on both cancer cell lines (IC 50 > 50 µM). Compared with sclerotiorin, most of the amine modified sclerotiorin derivatives except 1 and 2 displayed moderate to fine cytotoxic activities against the MDA-MB-435 and A549 cell lines. Particularly noteworthy, compounds 3, 7, 12, 13, 15 and 17 showed excellent cytotoxic activities with IC 50 values of 6.39, 9.20, 9.76, 7.75, 9.08 and 8.18 µM, respectively. Further SAR analysis can give the following clues: (1) a vinylogous γ-pyridone formed by the corresponding nitrogen atom substituted pyran nucleus of sclerotiorin increase the cytotoxicity; (2) a certain suitable bulky structure can obtain high cytotoxic activity, which was disclosed by comparing the IC 50 values of compounds 1, 2, 24 to those of 7, 10, 12; (3) from the observation of compounds 13, 14, 15, 16, 17, we can conclude one methylene connected aryl side chain have good effect on cytotoxic activity. In order to illustrate the underlying mechanisms for the cytotoxicity of these compounds, several rational deduces are summed up. Firstly, the amine modified sclerotiorin derivatives may be more efficiently served as a basic factor to change the acidic tumor microenvironment. In this respect, we can explain why the nitrogen substituted structures are more cytotoxic to tumor cells. Secondly, sclerotiorin derivatives possess α, β-unsaturated ketone skeleton and could be Michael receptor. Cytotoxic activity of these compounds may be attributed to the Michael addition reactions of sclerotiorin derivatives with active nucleophilic group in the biomolecules, such as amino acids, nucleic acids and other compounds in the tumor cells to irreversibly affect the functions of the biomolecules or regulate the cellular signal pathway. Thirdly, the literature addressing tumor cells always showed some high expression of cellular factors, the sclerotiorin derivatives may exhibit cytotoxicity by inhibiting some of them, such as the COX-2. Further COX-2 inhibitory activity for these compounds was conducted as following. In recent years, COX-2 inhibitors become a new target and hotspot for anticancer drug research and get a lot of attention. For further investigations the biological activity of semi-synthetic analogs of sclerotiorin, all compounds were primarily screened the COX-2 inhibitory activity at a concentration of 20 µM in vitro.
As shown in Figure 1, most of semi-synthetic derivatives and parent compound sclerotiorin displayed good inhibitory activity for COX-2. Among all the derivatives, compound 3 displayed perfect COX-2 inhibition with a ratio of 70.6%, which is comparable to the positive control indomethacin (78.9% In recent years, COX-2 inhibitors become a new target and hotspot for anticancer drug research and get a lot of attention. For further investigations the biological activity of semi-synthetic analogs of sclerotiorin, all compounds were primarily screened the COX-2 inhibitory activity at a concentration of 20 μM in vitro. As shown in Figure 1, most of semi-synthetic derivatives and parent compound sclerotiorin displayed good inhibitory activity for COX-2. Among all the derivatives, compound 3 displayed perfect COX-2 inhibition with a ratio of 70.6%, which is comparable to the positive control indomethacin (78.9%). Furthermore, compounds 7, 13, 15, 17, 25 and sclerotiorin showed moderate COX-2 inhibitory ratio of 58.7%, 51.1%, 46.7%, 47.3%, 66.1% and 56.1%, respectively. More interestingly, compared to the parent compound sclerotiorin, the inhibitory activity of the esterlysis derivative 25 increased from 56.1% to 66.1%. However, 26 and 27, the esterification product of 25, displayed low COX-2 inhibitory ratio of 42.8%, 35.1%, respectively. Based on the above results, observation was that the COX-2 inhibitory activity of sclerotiorin derivatives were associated with the cytotoxic activity. Overall, the cytotoxic activity of the compounds were positively correlated with the inhibitory activity of COX-2. Out of 28 tested compounds, 3, 7, 13, 15 and 17 showed both cytotoxic activity against two cancer cell lines and potential COX-2 inhibitory activity. Particularly, compound 3 was found to be most potent derivative with high cytotoxic activity against the breast cancer cells and superior COX-2 inhibitory activity. Otherwise, compounds 25, 26, 27 and parent compound sclerotiorin were exhibiting good COX-2 inhibitory activity, but almost have no cytotoxic activity. This find suggests sclerotiorin and its esterlysis product 25 deserve intensive investigation on anti-inflammatory activity based on COX-2 inhibition selectivity.

Molecular Docking Studies
The COX-2 inhibitory effects of compound 3 led us to perform molecular docking studies to insight understand the ligand-protein interactions in detail, and dock simulations in COX-2 (PDB ID: 5GMN) [35] were carried out in AutoDock4.2.6 [36]. Docking procedure was validated by docking of 949 (the co-crystallized ligand of COX-2 protein) in the active site of COX-2 and root-mean-square deviation (RMSD) of 0.13 Å to the X-ray structure ( Figure 2). Based on the above results, observation was that the COX-2 inhibitory activity of sclerotiorin derivatives were associated with the cytotoxic activity. Overall, the cytotoxic activity of the compounds were positively correlated with the inhibitory activity of COX-2. Out of 28 tested compounds, 3, 7, 13, 15 and 17 showed both cytotoxic activity against two cancer cell lines and potential COX-2 inhibitory activity. Particularly, compound 3 was found to be most potent derivative with high cytotoxic activity against the breast cancer cells and superior COX-2 inhibitory activity. Otherwise, compounds 25, 26, 27 and parent compound sclerotiorin were exhibiting good COX-2 inhibitory activity, but almost have no cytotoxic activity. This find suggests sclerotiorin and its esterlysis product 25 deserve intensive investigation on anti-inflammatory activity based on COX-2 inhibition selectivity.

Molecular Docking Studies
The COX-2 inhibitory effects of compound 3 led us to perform molecular docking studies to insight understand the ligand-protein interactions in detail, and dock simulations in COX-2 (PDB ID: 5GMN) [35] were carried out in AutoDock4.2.6 [36]. Docking procedure was validated by docking of 949 (the co-crystallized ligand of COX-2 protein) in the active site of COX-2 and root-mean-square deviation (RMSD) of 0.13 Å to the X-ray structure ( Figure 2). Mar. Drugs 2021, 19, x FOR PEER REVIEW 6 of 18 Compound 3 was docked into the active site of COX-2 and the interaction energy of 8.24 kcal/mol was obtained. One hydrogen bonding interaction between the ester group of compound 3 with Gln92 (2.96 Å) of COX-2 active site was observed ( Figure 3). It is clearly visible in Figure 4 that arachidonic acid and compound 3 were located deeply inside the same pocket of COX-2. Three hydrogen bonding interactions between carboxyl of arachidonic acid with Thr198 (2.92 Å, 2.95 Å) and Thr199 (2.85 Å) of COX-2 active site were observed respectively ( Figure 5). The hydrophilic groups of compound 3 including carbonyl and ester group were inclined to via a hydrogen bonding interactions with the hydrophilic amino acid of the outside of the COX-2 protein active pocket. However, alkyl chain and cyclohexene of compound 3 is easier access to the interior of hydrophobic pocket by means of hydrophobic action, which is blocking the binding site of arachidonic acid in COX-2 enzyme to some extent. This may be the reason why the compound 3 showed a perfect inhibitory activity to COX-2 cyclooxygenase ( Figure 5).  Compound 3 was docked into the active site of COX-2 and the interaction energy of 8.24 kcal/mol was obtained. One hydrogen bonding interaction between the ester group of compound 3 with Gln92 (2.96 Å) of COX-2 active site was observed ( Figure 3). It is clearly visible in Figure 4 that arachidonic acid and compound 3 were located deeply inside the same pocket of COX-2. Three hydrogen bonding interactions between carboxyl of arachidonic acid with Thr198 (2.92 Å, 2.95 Å) and Thr199 (2.85 Å) of COX-2 active site were observed respectively ( Figure 5). The hydrophilic groups of compound 3 including carbonyl and ester group were inclined to via a hydrogen bonding interactions with the hydrophilic amino acid of the outside of the COX-2 protein active pocket. However, alkyl chain and cyclohexene of compound 3 is easier access to the interior of hydrophobic pocket by means of hydrophobic action, which is blocking the binding site of arachidonic acid in COX-2 enzyme to some extent. This may be the reason why the compound 3 showed a perfect inhibitory activity to COX-2 cyclooxygenase ( Figure 5).  Compound 3 was docked into the active site of COX-2 and the interaction energy of 8.24 kcal/mol was obtained. One hydrogen bonding interaction between the ester group of compound 3 with Gln92 (2.96 Å) of COX-2 active site was observed ( Figure 3). It is clearly visible in Figure 4 that arachidonic acid and compound 3 were located deeply inside the same pocket of COX-2. Three hydrogen bonding interactions between carboxyl of arachidonic acid with Thr198 (2.92 Å, 2.95 Å) and Thr199 (2.85 Å) of COX-2 active site were observed respectively ( Figure 5). The hydrophilic groups of compound 3 including carbonyl and ester group were inclined to via a hydrogen bonding interactions with the hydrophilic amino acid of the outside of the COX-2 protein active pocket. However, alkyl chain and cyclohexene of compound 3 is easier access to the interior of hydrophobic pocket by means of hydrophobic action, which is blocking the binding site of arachidonic acid in COX-2 enzyme to some extent. This may be the reason why the compound 3 showed a perfect inhibitory activity to COX-2 cyclooxygenase ( Figure 5).  showed H-bond interactions of 3.31 Å between its oxygen atom in the ring and Glu238 residues of the COX-2 active site ( Figure 6A). However, we found that compound 26 did not located deeply inside the hydrophobic pocket of COX-2 protein as compound 25 did, Mar. Drugs 2021, 19, 12 7 of 18 but docked on the surface of the protein by hydrogen bonding ( Figure 6B). Consequently, we deduce that the presence of ester group can resulted in lower activity of compound 26.   Compound 25 was docked into the active site of COX-2 and the interaction energy of 7.66 kcal/mol was obtained. Unexpectedly, it displayed no H-bonding interaction with the amino acid residue of COX-2 protein. Compound 26 was docked into the active site of COX-2 and the interaction energy of 6.26 kcal/mol was obtained. Three hydrogen bonding interactions between ester group of 26 with His4 (3.40 Å) and Trp5 (2.53 Å, 2.74 Å) of COX-2 active site was observed. One hydrogen bonding interaction between carbonyl group of 26 with Lys169 (3.15 Å) of COX-2 active site was observed. Additionally, Compound 26 showed H-bond interactions of 3.31 Å between its oxygen atom in the ring and Glu238 residues of the COX-2 active site ( Figure 6A). However, we found that compound 26 did not located deeply inside the hydrophobic pocket of COX-2 protein as compound 25 did, but docked on the surface of the protein by hydrogen bonding ( Figure 6B). Consequently, we deduce that the presence of ester group can resulted in lower activity of compound 26.  Compound 25 was docked into the active site of COX-2 and the interaction energy of 7.66 kcal/mol was obtained. Unexpectedly, it displayed no H-bonding interaction with the amino acid residue of COX-2 protein. Compound 26 was docked into the active site of COX-2 and the interaction energy of 6.26 kcal/mol was obtained. Three hydrogen bonding interactions between ester group of 26 with His4 (3.40 Å) and Trp5 (2.53 Å, 2.74 Å) of COX-2 active site was observed. One hydrogen bonding interaction between carbonyl group of 26 with Lys169 (3.15 Å) of COX-2 active site was observed. Additionally, Compound 26 showed H-bond interactions of 3.31 Å between its oxygen atom in the ring and Glu238 residues of the COX-2 active site ( Figure 6A). However, we found that compound 26 did not located deeply inside the hydrophobic pocket of COX-2 protein as compound 25 did, but docked on the surface of the protein by hydrogen bonding ( Figure 6B). Consequently, we deduce that the presence of ester group can resulted in lower activity of compound 26.

Chemistry
3.1.1. General Information The fungus SCNU-0016 used in the study was isolated from fresh fruit of the mangrove plant Acanthus ilicifolius L, which was collected in October 2019 from Hailing island Mangrove Nature Reserve in Guangdong province, China. The fungal isolation was conducted as following. Initially, the plant fruit was washed with sterile water and surface-sterilized in a 100 mL beaker with 75% ethanol for 1 min. This was followed by dipping the sample into 5% sodium hypochlorite for 1 min, then the plant parts were rinsed with sterile water and cut into 3 mm sections and plated on potato dextrose agar (PDA, potatoes 300 mg/mL, dextrose 20 mg/mL, agar 15 mg/mL, chloramphenicol 1 mg/mL) with penicillin (100 units/mL) and streptomycin (0.8 mg/ mL). The plates were incubated at 25 ± 1 • C. The endophytic fungal strains were isolated by routine microbiological. The fungal isolates were numbered and stored at 4 • C in triplicate on PDA slants. Fungal identification was carried out using a molecular biological protocol by DNA amplication and sequencing of the ITS region [37]. The sequence data of the fungal strain have been deposited at Gen Bank with accession no. MW-309502. A BLAST search result showed that the sequence was the most similar (100%) to the sequence of P. sclerotiorum. A voucher strain was deposited in School of Chemistry, South China Normal University, Guangzhou, China, with the access code SCNU-F0016.

Fungal Culture and Sclerotiorin Extraction
The fungal strain P. sclerotiorum SCNU-0016 was cultured on autoclaved potato liquidsubstrate media (one-hundred Erlenmeyer flasks (1000 mL); each containing 400 mL of potato liquid medium, 8 g of glucose and 1.2 g artificial sea salts) at room temperature under static conditions and daylight for 28 days. Following incubation, pancake thallus grew on top of the potato liquid media. Then the air dried fungal were soaked in the solvent MeOH/CH 2 Cl 2 (v:v, 1:1). Extracts were filtered and concentrated under reduced pressure to yield 820 g of residue. The residue was fractionated by column chromatography on silica gel by eluting with a gradient of EtOAc/petroleum ether from 1:10 to 1:1 give five fractions (Fr.1-Fr.5). sclerotiorin was extracted successfully by elution with EtOAc/petroleum ether (v:v, 1:10). sclerotiorin appear good solubility in the EtOAc/petroleum ether solution. Then removal of the solvent afforded 7.82 g of yellow powder.

COX-2 Inhibitory Activity
The in vitro inhibitory of the compounds were evaluated by using COX-2 (human) inhibitor screening assay kit (Item NO:701080) supplied by Cayman chemicals USA. Indomethacin was used as a positive control and the tested compounds was dissolved in DMSO.
Each compound was tested in triplicate at 20 µM, and the percent inhibition for COX-2 was obtained for each experiment. The amount of prostaglandins produced by enzyme in the presence of the test compounds was measured and compared with the control experiments (also performed in triplicate) with enzyme inhibition = 1/concentration of prostaglandin in each enzymatic reaction (The percent inhibition of test compound was inversely proportional to the amount of prostaglandins produced by each wells). Finally, the calculations were performed as per the kit guidelines.

Molecular Docking
The X-ray crystal structure of COX-2 cyclooxygenase (PDB ID: 5GMN) enzyme was obtained from protein data bank in PDB format as initialing point. The crystal original ligand was extracted from the crystal structure prior to docking. Then, all waters were removed in the crystal structure. Hydrogens addition and Gasteiger charges were executed in turn. The protein was regarded as rigid and the conformation of the ligand was regarded as changeable. The parameter of grid box was set as 100 × 100 × 100 points and center on the protein. The Lamarckian genetic algorithm was used to dock algorithm and number of GA runs was 100. PyMOL [38] and LigPolt [39] were used to visualize and analyze results.

Conclusions
In summary, a novel series of sclerotiorin derivatives were synthesized. All compounds have been screened for their cytotoxic activity in vitro. Most of sclerotiorin derivatives showed good to great cytotoxic activity. In addition, the COX-2 inhibitory results disclosed that most of the derivatives displayed considerable inhibition of COX-2. Particularly, compound 3 displayed superior inhibitory ratio of 70.6%, a comparable inhibition ratio to positive indomethacin (78.9%) in 20 µM. Moreover, both in vitro and in silico studies showed that some of the new compounds act as promising COX-2 inhibitors. The results of this research will provide useful information for the design of novel series of anticancer agents with COX-2 inhibitory activity.  Figure S10. HRMS spectrum of compound 4; Figures S11 and S12: NMR (in CD3OD) spectra of compound 4; Figure S13: HRMS spectrum of compound 5; Figures S14 and S15: NMR (in CDCl3) spectra of compound 5; Figure S16: HRMS spectrum of compound 6; Figures S17 and S18: NMR (in CDCl3) spectra of compound 6; Figure