Discovery of novel conjugates of quinoline and thiazolidinone urea as potential anti-colorectal cancer agent

Abstract Based on the obtained SARs, further structural optimisation of compound BC2021-104511-15i was conducted in this investigation, and totally ten novel quinoline derivates were designed, synthesised and optimised for biological activity. Among them, compound 10a displayed significant in vitro anticancer activity against COLO 205 cells with an IC50 value of 0.11 μM which was over 90-fold more potent than that of Regorafenib (IC50>10.0 μM) and Fruquintinib (IC50>10.0 μM). Furthermore, compound 10a exhibited over 90-fold selectivity towards COLO 205 relative to human normal colorectal mucosa epithelial cell FHC cells. Flow cytometry study demonstrated that compound 10a could induce apoptosis in COLO 205 cells, however, it could not induce cell cycle arrest in COLO 205 cells. The results of preliminary kinase profile study showed that compound 10a was a potential HGFR and MST1R dual inhibitor, with IC50 values of 0.11 μM and 0.045 μM, respectively.


Introduction
Colorectal cancer (CRC) is one of the most predominant malignancies with a high mortality rate globally 1 . It is estimated that the number of CRC patients will reach 2.5 million in 2035 2 . Approximately 25% of CRC patients presented metastatic disease at diagnosis, while almost 50% of them will develop metastases. According to the statistics, the 5-year survival rate ranged from 90% to 14% if CRC is diagnosed at a localised or metastatic stage 3 .
Nowadays, chemotherapy is the most extensively applied approach for the treatment of primary CRC and/or metastatic CRC (mCRC) 4 . The drugs used in chemotherapy were divided into cytotoxic drugs, tyrosine kinase inhibitors (TKIs), monoclonal antibodies, and programmed cell death protein-1/programmed cell death 1 ligand 1 (PD1/PD-L1) inhibitors, etc 5,6 . Among them, TKIs could significantly improve major efficacy parameters which included response rate (RR), progression free survival (PFS) and overall survival (OS). Unfortunately, only Regorafenib and Fruquintinib are successfully utilised in clinic as TKIs for the treatment of patients with mCRC ( Figure 1). Regorafenib is an orally bioavailable multitarget TKI which mainly inhibits the activity of vascular endothelial growth factor receptor1-3 (VEGFR1-3), tunica intima endothelial kinase 2 (TIE2), rearranged during transfection (RET), mast/stem cell growth factor receptor (Kit), and platelet-derived growth factor receptor (PDGFR), etc 7 . It has been approved by US Food and Drug Administration (US FDA) and European Medicines Agency (EMA) for the treatment of mCRC patients who had already been treated with fluoropyrimidine, oxaliplatin, anti-VEGF therapy and/or irinotecan-based chemotherapy 8 . Fruquintinib also is a small molecule multitarget TKI with high affinity for VEGFR1-3 9 . In 2018, it was approved by National Medical Products Administration (NMPA) for the treatment of mCRC patients who had suffered at least two unsuccessful standard therapies 10 . Despite the recent advances in the chemotherapy of primary CRC and mCRC, the survival benefit is still limited due to the high heterogeneity, resistance and severe side effects. Accordingly, there is still urgent need to develop alternative and potential therapeutic strategies with high efficacy and acceptable side effects for the treatment of CRC.
Based on the above survey, a study on developing novel HGFR/MST1R dual inhibitors as anti-CRC agents was carried out by our group. As shown in Figure 1, HGFR/MST1R dual inhibitor BC2021-104511-15i was discovered by our group 26,27 . It exhibited potential in vitro anticancer activity against several cancer cell lines, especially human colorectal carcinoma cell line HT-29 cells. In order to obtain a more potent HGFR/MST1R dual inhibitor as an agent for the treatment of CRC, further modification on the fragments I, II, and III of BC2021-104511-15i was performed ( Figure 2). The details of design, synthesis, biological evaluation, docking study and anticancer mechanism were all discussed in the following sections.

The modification of lead compound
As shown in Figure 2, preliminary SARs were summarised based on the biological evaluation in our previous research. The SARs indicated that the groups attached to the piperidine ring (I) could significantly influence the HGFR and MST1R kinases inhibitory activity, in vitro anticancer activity and water solubility. Thus, compounds 10a-h bearing variant substituents on piperidine rings were designed and synthesised in the beginning of this work. Docking study showed that two significant H-bonds were formed by quinoline ring and urea moiety with Met1160 and Lys1110, respectively ( Figure 2). The terminal difluoro-substituted phenyl ring reached into a hydrophobic pocket, and H-p interaction was formed. We assumed that the spatial position change of the H-bond donors and acceptors in the moiety of thiazolidine-4-one urea might strengthen the H-bond and/or lead to additional H-bonds. Thus, the oxygen atom linked quinoline ring and 2-fluorophenyl ring (III) was replaced to deflect the dihedral angel formed by the aromatic rings, and compound 10i was designed and synthesised. Additionally, a quinazoline derivate 19 was designed and prepared to investigate the influence on the kinase inhibitory activity when the electron density distribution changed in the quinoline ring.
Based on the above assumption, totally ten novel compounds were designed, synthesised and evaluated for their biological activity in the present work. Moreover, the anticancer mechanism was also investigated preliminarily.

Chemistry
Target compounds 10a-i and 19 were successfully prepared by the synthetic routes outlined in Scheme 1 and Scheme 2 26 . Commercially available 7-(benzyloxy)-4-chloro-6-methoxyquinoline was reacted with 2-fluoro-4-nitrophenol in refluxing chlorobenzene to afford 4-aryloxyquinoline 1a. Subsequently, intermediate 1 was cleanly debenzylated by 33% HBr in acetic acid to obtain phenol 2a, which was alkylated with 1-Boc-4-methanesulfonyloxypiperidine in the presence of Cs 2 CO 3 to provide 3a. The amine 4a was achieved by reduction of nitro group in 3a. Then, the semicarbazide 6a was acquired using a two-step procedure involving acylation reaction with phenyl chloroformate in the presence of pyridine and subsequent hydrazinolysis reaction with 50%   hydrazine hydrate in xylene with vigorous agitation. Condensation of 6a with 2,6-difluorobenzaldehyde in favour of catalytic HOAc was carried out to provide 7a. The N-Boc group in the semicarbazone 7a was deprotected by CF 3 COOH to afford piperidine derivate 8a, which was then acylated or alkylated with corresponding acyl chlorides or chlorides to give intermediates 9a-h. Finally, the target compounds 10a-g were prepared by cyclisation reaction with mercaptoacetic acid in the presence of SiCl 4 . Target compound 10h was obtained by the hydrolysis of compound 10g under basic condition in MeOH.

Structure-Activity relationship
Lead compound BC2021-104511-15i, Fruquintinib, Regorafenib, Cabozantinb and Foretinib were chosen as positive controls in the study of biological evaluation. In our previous research, it was revealed that the substituent on the piperidine ring could significantly influence the inhibitory activity against both kinases and cancer cells. Thus, eight novel compounds bearing diversified R groups (10a-h) were designed and synthesised. Biological activity study indicated that no obvious differences on kinase inhibitory activity could be found between the heterocyclic groups (10e-f) and catenoid groups (10a-d and 10g-h). As a general trend, the introduction of N-CH 2 -CO fragment was beneficial for the HGFR and MST1R kinase inhibitory activity, such as compounds 10a, 10c and 10f. Docking study showed that additional H-bond formed by the carbonyl group and the residue His1094 might lead to the increased activity ( Figure 3). Among the eight compounds, compound 10a was identified as the most potent HGFR and MST1R inhibitor, with IC 50 values of 0.11 lM and 0.045 lM, respectively (Table 1).
Compared with compound 10a, replacement of the oxygen atom linked the quinoline ring and 2-fluorophenyl ring by NH (10i) led to a significant decrease in anticancer activity (HT-29 IC 50 >10.0 lM and HCT-116 IC 50 >10.0 lM). Additionally, the quinazoline derivate 19 also showed weaker anticancer activity (HT-29 IC 50 ¼8.6 lM and COLO 205 IC 50 ¼5.3 lM, Table 1). The decrease of the biological activity might result from the conversion of electron density distribution in the quinoline ring which might weaken the H-bond between the nitrogen atom in quinoline and the residue Met1160.

The cytotoxicity against FHC cells
In order to investigate the cell selectivity index, the cytotoxicity of potent compounds against human normal colorectal mucosa epithelial cell FHC cells was determined. As could be seen in Table 2, all the anticancer agents 10a-b, 10d, 10f and 10h displayed no obvious cytotoxicity against FHC cells (IC 50 >10.0 lM). Notably, the

Molecular docking study
Docking of the most potent compound 10a into HGFR was performed by Molecular Operating Environment (MOE). As shown in Figure 3, compound 10a adopted an extended conformation as type II kinase inhibitor exemplified by Cabozantinib and Foretinib. Totally three key hydrogen bonds were formed: nitrogen atom in quinoline with residue Met1160, oxygen atom in urea moiety with residue Lys1110, and oxygen atom in the terminal amido group with residue His1094. The hydrophobic pocket was occupied by the terminal 2,6-difluorophenyl ring, and weak H-p interaction was formed between 2,6-difluorophenyl fragment with residue Phe1134. Additionally, weak Harene interaction was also formed between the quinoline ring and residue Val1092.

In vitro kinase profile
To further investigate the kinase selectivity of novel target compounds, the inhibitory activity of compounds 10a and 10b against another seven kinases was evaluated, including ABL, PDGFRb, AXL, FLT3, RET, c-Src and VEGFR-2. As indicated in Table 3, compounds 10a and 10b showed much weaker inhibitory activity against the above kinases. The above results suggested that compounds 10a and 10b were potential HGFR and MST1R dual inhibitors. Certainly, only preliminary kinase profile was studied in this work, and further study will be conducted in the following structural modification.

Cell apoptosis assay by flow cytometry
Cell apoptosis assay was conducted to investigate whether the cytotoxic activity of the most potent compound 10a was caused by the activation of cellular apoptosis in COLO 205 cells. Quantitative analysis of early-apoptotic cells, advanced-apoptotic cells and necrotic cells was determined. COLO 205 cells were stimulated with different concentrations of compound 10a for 72 h. As depicted in Figure 4, compound 10a could effectively induce COLO 205 cells apoptosis in a dose-dependent manner. The total apoptosis including early apoptosis and advanced apoptosis accounted for 58.8%, 29.3% and 21.0% (the mean value of three independent determinations) when COLO 205 cells were treated with compound 10a at the concentration of 1.0 lM, 0.33 lM and 0.11 lM, respectively.

Cell cycle arrest assay by flow cytometry
The antiproliferative activity of compound 10a was evaluated by flowcytometry analysis of COLO 205 cells. COLO 205 cells were treated with different concentrations (1.0 lM, 0.33 lM and 0.11 lM) of compound 10a for 24 h. As is shown in Figure 5, compound 10a could not arrest the cell-cycle progression at the concentration of 0.33 lM and 0.11 lM. At the concentration of 1.0 lM, cell cycle arrest could not be determined due to the massive dead cells (data not shown). The above results indicated that the anticancer mechanism of compound 10a might be cytotoxicity rather than antiproliferation.

Chemistry
Unless otherwise noted, all chemicals were obtained from commercial vendors and used directly without further purification. Analytical reagent (AR) grade solvents were used for all reactions. Reaction progress was monitored by TLC on pre-coated silica plates (Huanghai HSGF254, 0.20 mm, pH 6.2-6.8) and spots were visualised by UV (254 nm). Flash column chromatography was done using silica gel (Qingdao Ocean Chemical Company, 200-300 mesh). 1 H NMR and 13 C NMR spectra were recorded on a Bruker AVANCE neo 600. High resolution ESI-MS were recorded on Orbitrap Exploris 240 (Thermo Fisher Science, MA, USA).

General procedure for the synthesis of intermediates (4a-b)
The suspension of intermediates 3a-b (0.02 mol), powered iron (0.06 mol) and concentrated HCl (2 drops) in 90% EtOH (100 ml) was refluxed for 4-6 h with vigorously stirred. After the reaction was completed, the hot mixture was filtered through celites, and the filtrate was evaporated under reduced pressure to afford the title intermediates.

General procedure for the synthesis of intermediates (6a-b)
Phenyl chloroformate (20.0 mmol) was added to a solution of amines 4a-b (10.0 mmol) and dry pyridine (30.0 mmol) in dry CH 2 Cl 2 (40 ml) at 0 C. After the addition was completed, the mixture was allowed to warm to room temperature for another 2 h, and then saturated NaHCO 3 aqueous solution was added to the solution. The CH 2 Cl 2 phase was separated, washed with water, dried over anhydrous MgSO 4 , and concentrated under reduced pressure to afford intermediates 5a-b, which were immediately used in the following step without further purification.
To a mixture of esters 5a-b in 20 ml xylene was added hydrazine monohydrate (50%, 20 ml). The reaction mixture was stirred vigorously at 70 C for 2 h. The solvent and excessive hydrazine monohydrate were evaporated under reduced pressure, and the residue was purified by flash chromatography (eluent with 1-10% MeOH in DCM, 1% Et 3 N) to afford semicarbazides 6a-b.

General procedure for the synthesis of intermediates (8a-b)
The solution of 7a-b (0.015 mol) and CF 3 COOH (0.15 mol) in CH 2 Cl 2 (50 ml) was stirred for 2 h at room temperature Evaporation of the solvent and excessive CF 3 COOH provided yellow oil. The residue was diluted with 100 ml CH 2 Cl 2 , and 20% NaOH aqueous solution was added until the pH reached to 9. The organic phase was washed with water, dried over anhydrous MgSO 4 , concentrated in vacuo, and the residue was used for the next step without further purification. 4.1.9. General procedure for the synthesis of intermediates (9a-h) Method A for the preparation of 9b, 9d, 9e and 9g. To a cold solution of 8a (0.01 mol) and Et 3 N (0.015 mmol) in dry 30 ml CH 2 Cl 2 was added acyl chloride (0.013 mol) dropwise. Then, the reaction mixture was stirred at room temperature for 4-5 h, saturated NaHCO 3 aqueous solution was added, and the CH 2 Cl 2 phase was separated, dried over anhydrous MgSO 4 , and concentrated in vacuum to afford the title intermediates.
Method B for the preparation of 9a, 9c, 9f and 9h. To a suspension of 8a-b (0.01 mol) and Cs 2 CO 3 (0.015 mol) in 30 ml DMF was added 2-chloroamides (0.015 mol). The resulting mixture was stirred for 6-8 h at 90 C, and then poured into cold water. The precipitate was filtered off, washed with water, and dried to afford the title intermediate. To a suspension of intermediates 9a-h (0.3 mmol) in dry CH 2 Cl 2 (5 ml), mercaptoacetic acid (0.3 ml) and SiCl 4 (15 drops) were added subsequently at 0 C. The reaction mixture was allowed to cooled to room temperature, and refluxed for 6-8 h. The mixture was cooled to room temperature, and quenched by 2 ml cold water. After stirred for 5 min, 10% NaOH aqueous solution was added until pH reached to 10. The CH 2 Cl 2 phase was separated and washed with water (2 Â 5 ml), concentrated under reduced pressure to yield crude products which were purified by flash chromatography (eluent with 5-10% MeOH in DCM, 1% Et 3 N) to give target compounds. 4.1.12. 7-(Benzyloxy)-4-(2-fluoro-4-nitrophenoxy)-6-methoxyquinazoline (11) The mixture of 7-(benzyloxy)-4-chloro-6-methoxyquinazoline (9.0 g, 0.03 mol) and 2-fluoro-4-nitrophenol (6.3 g, 0.04 mol) in 60 ml chlorobenzene was refluxed for 13 h. The reaction mixture was cooled to room temperature, and the solvent was evaporated under reduced pressure. The residue was dissolved in 300 ml CH 2 Cl 2 , washed with 10% NaOH aqueous solution (3 Â 30 ml) and 50 ml water. The CH 2 Cl 2 phase was dried over anhydrous MgSO 4 and concentrated under reduced pressure to give 8. Phenyl chloroformate (28.2 mmol) was added to a solution of amine 14 (6.9 g, 14.2 mmol) and dry pyridine (42.6 mmol) in dry CH 2 Cl 2 (350 ml) at 0 C. After the addition was completed, the mixture was allowed to warm to room temperature for another 2 h, and then saturated NaHCO 3 aqueous solution was added to the solution. The CH 2 Cl 2 phase was separated, washed with water, dried over anhydrous MgSO 4 , and concentrated under reduced pressure to afford brown oil, which were immediately used in the following step without further purification.

MTT assay
Taking lead compound BC2021-104511-15i, Fruquintinib and Regorafenib as positive controls, the cytotoxic activity against HT-29, HCT-116, COLO-205 and FHC cell lines by MTT assay. Detailed operation could be found in our previous study 29 .

Mobility shift assay
Kinase inhibitory activity against HGFR, MST1R, ABL, PDGFRb, AXL, FLT3, RET, c-Src, and VEGFR-2 was evaluated by the mobility shift assay. Detailed operation could be found in our previous research 29 .

Molecular docking study
Docking study were conducted by Molecular Operating Environment 2018.01 (MOE, Chemical Computing Group ULC, Montreal, QC, Canada) using default settings. The structure of HGFR kinase was prepared (protonation, modelling of missing elements) from the original PDB files using Quickprepare. The binding site was defined within 5.0 Å of the cocrystallized ligands coordinates. The docking forcefield was Amber10: EHT. Ligand conformations were placed in the site with the Triangle Matcher method and ranked using the London dG scoring function.

Cell-Cycle analysis
COLO 205 cells (2.5 Â 10 5 ) were seeded in two 12-well plates and treated with DMSO and compound 10a (0.11 lM, 0.33 lM and 1.0 lM) for 24 h (37 C, 5% CO 2 ). Cells were collected, centrifuged at 1000 rpm for 5 min, washed with cold PBS for twice, and then fixed with 500 lL 75% cold ethanol at 4 C. The cells were washed with cold PBS and stained with propidium iodide for 30 min in the dark. Cell-cycle analyses were conducted with Cytoflex S (Beckman Coulter).

Annexin V-FITC/PI apoptosis assay
COLO 205 cells (2.5 Â 10 5 ) were seeded in two 12-well plates and treated with DMSO and compound 10a (0.11 lM, 0.33 lM and 1.0 lM) for 72 h (37 C, 5% CO 2 ). The cells were collected, centrifuged at 1000 rpm for 5 min, and then washed with cold PBS for twice. Apoptosis assays were conducted with CytoFLEX S (Beckman Coulter).

Conclusions
Starting from the obtained HGFR and MST1R dual inhibitor BC2021-104511-15i, ten novel quinoline derivates were designed, synthesised and evaluated for their biological activity. More detailed SARs were summarised based on the kinase inhibitory activity and in vitro anticancer activity. Among these compounds, 10a was identified as the most potent HGFR/MST1R dual inhibitor (HGFR IC 50 ¼0.11 lM and MST1R IC 50 ¼0.045 lM) with excellent anti-colorectal cancer activity (COLO 205 IC 50 ¼0.11 lM). Furthermore, it exhibited over 90-fold selectivity towards COLO 205 cells relative to human normal colorectal mucosa epithelial cell FHC cells. Docking study indicated that compound 10a adopted an extended conformation as type II kinase inhibitor. Hbond, hydrophobic interaction and H-p interaction were the key contributors led to the strong binding affinity to kinase. Flow cytometry study demonstrated that compound 10a could induce apoptosis in COLO 205 cells; however, it could not induce cell cycle arrest in COLO 205 cells. The results indicated that the anticolorectal cancer activity against COLO 205 cells mainly depended on its cytotoxicity rather than antiproliferation. Preliminary kinase profile study showed that compound 10a was a potential HGFR and MST1R dual inhibitor, its inhibitory activity against HGFR and MST1R was more potent than that of ABL, PDGFRb, AXL, FLT3, RET, c-Src, and VEGFR-2 kinases.

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