Design, Synthesis and Antitumor Activity of Novel Selenium-Containing Tepotinib Derivatives as Dual Inhibitors of c-Met and TrxR

Cellular mesenchymal–epithelial transition factor (c-Met), an oncogenic transmembrane receptor tyrosine kinase (RTK), plays an essential role in cell proliferation during embryo development and liver regeneration. Thioredoxin reductase (TrxR) is overexpressed and constitutively active in most tumors closely related to cancer recurrence. Multi-target-directed ligands (MTDLs) strategy provides a logical approach to drug combinations and would adequately address the pathological complexity of cancer. In this work, we designed and synthesized a series of selenium-containing tepotinib derivatives by means of selenium-based bioisosteric modifications and evaluated their antiproliferative activity. Most of these selenium-containing hybrids exhibited potent dual inhibitory activity toward c-Met and TrxR. Among them, compound 8b was the most active, with an IC50 value of 10 nM against MHCC97H cells. Studies on the mechanism of action revealed that compound 8b triggered cell cycle arrest at the G1 phase and caused ROS accumulations by targeting TrxR, and these effects eventually led to cell apoptosis. These findings strongly suggest that compound 8b serves as a dual inhibitor of c-Met and TrxR, warranting further exploitation for cancer therapy.


Introduction
Cellular mesenchymal-epithelial transition factor (c-Met), an oncogenic transmembrane receptor tyrosine kinase (RTK), plays an essential function in invasive growth during embryo development and liver regeneration [1,2]. c-Met signaling pathway has been implicated as a critical regulator for maintaining intracellular redox homeostasis and oxidative stress [3]. c-Met promotes the onset, proliferation, invasion and metastasis of hepatocellular carcinoma (HCC) [4]. It has been reported that tepotinib, a potent c-Met inhibitor, exhibits promising activity in advanced HCC with c-Met overexpression in clinical studies, indicating that c-Met may serve as a therapeutic target for HCC [5]. Recent studies have demonstrated that activation of c-Met can modulate the redox protective nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase-1 (HO-1) and downregulate reactive oxygen species (ROS), eventually inhibiting the death of cancer cells [6]. Treatment combined with c-Met and HO-1 inhibitors can promote ROS-induced oxidative stress and markedly reduce tumor growth [7]. Hence, promoting intracellular oxidative stress might serve as a strategy for improving anticancer efficacy and overcoming the resistance to anticancer drugs.
Thioredoxin reductase (TrxR), a selenium (Se)-dependent enzyme, is one of the most important antioxidant systems in cellular physiology [8]. TrxR is markedly upregulated

Chemistry
The synthetic route of compounds 8a-h is summarized in Scheme 1. Specifically, Suzuki-Miyaura cross-coupling of 2-chloro-5-fluoropyrimidine (10) with (3-(hydroxymethyl)phenyl)boronic acid (11) yielded compound 12. Treatment of compound 12 with SOCl2 (to give chloride 13) and subsequent substitution with 3-(6-oxo-1,6dihydropyridazin-3-yl)benzonitrile (14) gave compound 15. Finally, the reaction of selenourea with the corresponding chloroalkanes afforded the Se-containing intermediates, which were coupled with compound 15 in situ under basic conditions to afford compounds 8a-h. Inspired by these findings, as well as by the potential advantages of MTDLs, we utilized a Se-based bioisostere strategy on tepotinib 7 and consequently designed a series of Se-containing compounds 8a-h and 9a-c ( Figure 1B), with the aim of developing dual inhibitors of c-Met and TrxR. Such dual inhibitors may improve therapeutic effect while reducing toxicity to normal cells. Herein, we describe the synthesis of compounds 8a-h and 9a-c and systematic investigation into their antiproliferative activity, c-Met and TrxR inhibition, induction of ROS and cell cycle arrest/apoptosis-triggering ability.

Scheme 1. Synthesis of compounds 8a-h.
The synthetic route of compounds 9a-c is summarized in Scheme 2. Specifically, Suzuki-Miyaura cross-coupling of 5-bromo-2-iodopyrimidine (16) with compound 11 gave compound 17. Treatment of compound 17 with SOCl 2 and subsequent substitution with compound 14 gave compound 18. The Miyaura borylation of compound 18 with bis(pinacolato)diboron yielded compound 19, which was subject to sodium perborate oxidation to give compound 20. The reaction of compound 20 with corresponding dibromoalkanes afforded compounds 21a-c. Finally, the reaction of compounds 21a-c with potassium selenocyanate provided compounds 9a-c. Suzuki-Miyaura cross-coupling of 5-bromo-2-iodopyrimidine (16) with compound 11 gave compound 17. Treatment of compound 17 with SOCl2 and subsequent substitution with compound 14 gave compound 18. The Miyaura borylation of compound 18 with bis(pinacolato)diboron yielded compound 19, which was subject to sodium perborate oxidation to give compound 20. The reaction of compound 20 with corresponding dibromoalkanes afforded compounds 21a-c. Finally, the reaction of compounds 21a-c with potassium selenocyanate provided compounds 9a-c. Scheme 2. Synthesis of compounds 21a-c. Scheme 2. Synthesis of compounds 21a-c.

In Vitro Biological Evaluations
To assess the structure-activity relationship (SAR), compounds 8a-h and 9a-c were screened for their antiproliferative activity against human hepatocarcinoma cells (MHCC97H) using an MTT assay. Tepotinib 7 was used as a positive control. The antiproliferative activity of each compound is expressed as an IC 50 value (Table 1). It can be seen that Se-bearing compounds generally exhibited potent antiproliferative activity, with the IC 50 values ranging from 0.010 to 8.421 µM. Among them, 2-(dimethylamino)ethyl)selanylsubstituted compound 8b showed the most potent antiproliferative activity. A series of hydrophilic side chains were introduced to the R position. As a consequence, the compounds having a side chain of two atoms, that is, compounds 8b, 8d and 8f, had high activity. Compared with compound 8b, replacement of the terminal dimethylamino substituent with N-heterocycle groups, including piperidine (8c,d), morpholine (8e,f), pyrrolidine (8g) and azepane (8h), led to a slight decrease in the antiproliferative activity. In addition, compounds 9a-c bearing a selenocyanato linked to the aromatic skeleton through an alkoxy chain exhibited modest activity, and the activity was found to be dependent on the length of the alkoxy linkers, as supported by the successive decrease in the activity ongoing from 9a to 9b to 9c.

c-Met and TrxR Inhibitory Effects in MHCC97H Cells
Based on the cytotoxicity results, we chose compounds 8a, 8b, 8d, 8f and 8g to assess their inhibitory activity toward c-Met and TrxR by means of an ADP-Glo kinase assay kit and colorimetric assay (Table 1). Notably, compound 8b displayed potent inhibition on both c-Met and TrxR, with the IC 50 values being 0.010 and 0.099 µM, respectively. Compared with compound 8b, replacement of the R substituent with piperidine (8d), morpholine (8f) and pyrrolidine (8g), respectively, led to a slight decrease in the inhibitory activity toward c-Met and TrxR. These results are in accordance with those obtained from the cytotoxicity assay.

Selectivity for Cancer Cells over Noncancer Cells
To evaluate the cytotoxic selectivity of compound 8b for cancer cells over noncancer cells, we determined the cytotoxicity of compound 8b against hepatocellular carcinoma cell line (HCCLM3) and human normal liver cells (LO2) using a standard MTT assay. The cytotoxicity and selectivity index (SI) of compound 8b are listed in Table 2. The 375-fold higher SI for LO2 than for cancer cells strongly suggests that compound 8b is more cytotoxic to hepatic cancer cells than to normal cells.  (7) 0.016 ± 0.001 0.020 ± 0.003 1.786 ± 0.208 99 a IC 50 values are indicated as the mean ± SD of three independent experiments. b SI is defined as a ratio of IC 50 (noncancer)/average IC 50 (hepatic cancer cell).

Effect of Compound 8b on ROS Generation
It has been reported that TrxR inhibition is closely related to a significant increase in ROS production [10,28]. Thus, we detected the peroxidation of cellular lipid by using a fluorescent C11-BODIPY probe. Pronounced green fluorescence was observed in compound-8b-treated MHCC97H cells, indicating that this compound effectively causes the accumulation of cellular lipid peroxides ( Figure 2A). The intracellular lipid ROS generation was further evaluated with flow cytometry. The significant lipid ROS formation appeared in the compound-8b-treated group in a dose-dependent manner compared with that of the tepotinib-7-treated group, which is attributed to the TrxR inhibitory activity of compound 8b ( Figure 2B). Moreover, the levels of intracellular ROS were measured using a 2,7-dichlorofluorescein diacetate (DCF-DA) probe. As a result, compound 8b markedly induced the production of ROS, and the production of intracellular ROS was associated with TrxR inhibition ( Figure 2C). To investigate the time-dependent effect of compound 8b on the formation of ROS, we detected the levels of intracellular ROS by exposing compound 8b to MHCC97H cells for 12, 24 and 36 h and analyzed the cells by using a DCF-DA probe. As shown in Figure 2D, the formation of ROS increased with time and reached its peak at 24 h. To assess whether the production of ROS plays a crucial role in the death of cancer cells induced by compound 8b, we treated the MHCC97H cells in the presence or absence of a free radical scavenger (N-acetyl cysteine, NAC). As shown in Figure 2E, the cell death induced by compound 8b in MHCC97H could be rescued by NAC. These results indicate that compound 8b induced oxidative stress in MHCC97H cells by augmenting the levels of ROS and caused the death of cancer cells ultimately.

Cellular Apoptosis Analysis and Cell Cycle Study
It is reported that increasing the levels of intracellular ROS is highly related to the induction of cellular apoptosis [28]. As compound 8b increased the production of ROS, we then investigated its effect on apoptosis in MHCC97H cells. Thus, MHCC97H cells were treated with this compound at varying concentrations for 24 h and 48 h, respectively, and analyzed by means of an Annexin V-FITC/PI assay. As shown in Figure 3, compared with the vehicle group, incubation with compound 8b led to a significant increase in the proportion of apoptotic cells. Specifically, upon the treatment with compound 8b for 24 h, the proportions of apoptotic cells were 12.5% at 10 nM, 18.4% at 20 nM and 35.5% at 40 nM, respectively. After 48 h treatment with compound 8b, the corresponding proportions of apoptotic cells reached 26.5%, 33.4% and 61.4%, respectively. These results illustrated that compound 8b effectively induced the apoptosis of MHCC97H cells in both dose-and time-dependent manners. Moreover, immunoblotting analysis demonstrated that compound 8b markedly upregulated the expression of cleaved caspase 3, suggesting that

Cellular Apoptosis Analysis and Cell Cycle Study
It is reported that increasing the levels of intracellular ROS is highly related to the induction of cellular apoptosis [28]. As compound 8b increased the production of ROS, we then investigated its effect on apoptosis in MHCC97H cells. Thus, MHCC97H cells were treated with this compound at varying concentrations for 24 h and 48 h, respectively, and analyzed by means of an Annexin V-FITC/PI assay. As shown in Figure 3, compared with the vehicle group, incubation with compound 8b led to a significant increase in the proportion of apoptotic cells. Specifically, upon the treatment with compound 8b for 24 h, the proportions of apoptotic cells were 12.5% at 10 nM, 18.4% at 20 nM and 35.5% at 40 nM, respectively. After 48 h treatment with compound 8b, the corresponding proportions of apoptotic cells reached 26.5%, 33.4% and 61.4%, respectively. These results illustrated that compound 8b effectively induced the apoptosis of MHCC97H cells in both dose-and timedependent manners. Moreover, immunoblotting analysis demonstrated that compound 8b markedly upregulated the expression of cleaved caspase 3, suggesting that ROS induced by compound 8b caused an increase in caspase-mediated apoptosis (Figure 4). ROS induced by compound 8b caused an increase in caspase-mediated apoptosis ( Figure  4).   It is known that inhibitors that block c-Met activation are able to arrest the cell cycle at the G1 phase [29]. Therefore, we performed a flow cytometric analysis to assess the cellcycle distribution induced by compound 8b, using propidium iodide (PI) staining in MHCC97H cells. As shown in Figure 5, the proportions of cells at the G1 phase in the presence of compound 8b were 79.4% at 10 nM, 83.9% at 20 nM and 86.2% at 40 nM, respectively. This gradual accumulation of MHCC97H cells at the G1 phase with the increase in the concentration of compound 8b strongly suggests that this compound induced cell cycle arrest at the G1 phase in MHCC97H cells. In total, the above-mentioned findings clearly demonstrate several advantages of Secontaining compound 8b for cancer treatment. Firstly, compound 8b exhibited potent inhibitory activity toward TrxR, without any effect on its inhibitory activity toward c-Met. It is known that inhibitors that block c-Met activation are able to arrest the cell cycle at the G 1 phase [29]. Therefore, we performed a flow cytometric analysis to assess the cell-cycle distribution induced by compound 8b, using propidium iodide (PI) staining in MHCC97H cells. As shown in Figure 5, the proportions of cells at the G 1 phase in the presence of compound 8b were 79.4% at 10 nM, 83.9% at 20 nM and 86.2% at 40 nM, respectively. This gradual accumulation of MHCC97H cells at the G 1 phase with the increase in the concentration of compound 8b strongly suggests that this compound induced cell cycle arrest at the G 1 phase in MHCC97H cells.  It is known that inhibitors that block c-Met activation are able to arrest the cell cycle at the G1 phase [29]. Therefore, we performed a flow cytometric analysis to assess the cellcycle distribution induced by compound 8b, using propidium iodide (PI) staining in MHCC97H cells. As shown in Figure 5, the proportions of cells at the G1 phase in the presence of compound 8b were 79.4% at 10 nM, 83.9% at 20 nM and 86.2% at 40 nM, respectively. This gradual accumulation of MHCC97H cells at the G1 phase with the increase in the concentration of compound 8b strongly suggests that this compound induced cell cycle arrest at the G1 phase in MHCC97H cells. In total, the above-mentioned findings clearly demonstrate several advantages of Secontaining compound 8b for cancer treatment. Firstly, compound 8b exhibited potent inhibitory activity toward TrxR, without any effect on its inhibitory activity toward c-Met. In total, the above-mentioned findings clearly demonstrate several advantages of Se-containing compound 8b for cancer treatment. Firstly, compound 8b exhibited potent inhibitory activity toward TrxR, without any effect on its inhibitory activity toward c-Met. Secondly, compound 8b markedly induced the formation of ROS and the production of intracellular ROS was associated with TrxR inhibition. In contrast, no obvious ROS formation was detected in the tepotinib-treated group. This result strongly suggests that the Se bioisosteric modification of tepotinib 7 was the cause of ROS generation induced by compound 8b. Thirdly, the synergistic effect of TrxR and c-Met inhibition by compound 8b might accelerate the redox imbalance in cancer cells. In combination with the dual inhibition of c-Met and TrxR, compound 8b exhibited more potent cell apoptosis than tepotinib. Fourthly, compound 8b is more cytotoxic to hepatic cancer cells than to normal cells. This selectivity is considered to be due to the Se replacement of tepotinib 7, which is in agreement with previous reports that Se-containing compounds show high selectivity for cancer cells [30,31]. Hence, these advantages of compound 8b ensure that Se replacement might be a promising strategy for the rational design of novel drugs in cancer treatment.

General Methods (Chemistry)
General methods are described in the Supplementary Material.

General Procedures for the Preparation of Compounds 15 and 18
A solution of compound 12 (or 17, 5 mmol) in SOCl 2 (20 mL) was stirred at room temperature. After 2 h, the mixture was evaporated and the resulting residue was dissolved in anhydrous toluene. The solution was filtered and the filtrate was concentrated to afford chlorides, which were used for the next step without any further purification. A mixture of compound 14 (5.5 mmol, 1.08 g), K 2 CO 3 (7.5 mmol, 1.04 g) and the above-mentioned chlorides in N,N-dimethylformamide (DMF, 10 mL) was stirred at 70 • C. After 18 h, the reaction mixture was cooled to room temperature and H 2 O was added. The mixture was extracted with ethyl acetate. The combined organic layers were washed with brine, dried over anhydrous Na 2 SO 4 and evaporated. The residues were purified by silica gel chromatography using CH 2 Cl 2 /CH 3 OH (100/1, v/v) as eluents to afford compound 15 (or 18).

General Procedures for the Preparation of Compounds 8a-h
A solution of selenourea (0.5 mmol, 61 mg) and corresponding chloroalkylamines (0.5 mmol) in anhydrous EtOH (2 mL) was stirred at 80 • C under nitrogen. After 7 h, the reaction mixture was cooled to room temperature and concentrated in vacuo. To a solution of the resulting residue in a mixed solvent of DMF and H 2 O (2/1, v/v, 3 mL) compound 15 (0.3 mmol, 115 mg) and NaOH (2.5 mmol, 111 mg) were added. The reaction mixture was stirred at 60 • C under nitrogen. After 3 h, the reaction mixture was extracted with ethyl acetate. The combined organic layers were washed with brine, dried over anhydrous Na 2 SO 4 and evaporated. The residues were purified by silica gel chromatography using CH 2 Cl 2 /CH 3 OH (20/1, v/v) as eluents to afford compounds 8a-h.

General Procedures for the Preparation of Compounds 21a-c
A mixture of compound 20 (1 mmol, 381 mg), corresponding dibromoalkane (1.5 mmol) and K 2 CO 3 (1.5 mmol, 207 mg) in CH 3 CN (10 mL) was stirred at 85°C. After 10 h, the reaction mixture was cooled to room temperature and concentrated in vacuo. The resulting residue was partitioned between ethyl acetate and water. The organic layer was separated, washed with brine, dried over anhydrous Na 2 SO 4 and concentrated. The residues were purified by silica gel chromatography using ethyl acetate/petroleum (4/1, v/v) as eluents to afford compounds 21a-c.

General Procedures for the Preparation of Compounds 9a-c
A solution of 21a-c (0.5 mmol) and KSeCN (0.75 mmol, 108 mg) in CH 3 CN (3 mL) was stirred at 85 • C. After 10 h, the reaction mixture was cooled to room temperature and concentrated in vacuo. The resulting residue was partitioned between ethyl acetate and water. The organic layer was separated, washed with brine, dried over anhydrous Na 2 SO 4 and concentrated. The residues were purified by silica gel chromatography using ethyl acetate/petroleum (2/1, v/v) as eluents to afford compounds 9a-c.

Conclusions
In this study, a series of novel selenium-containing tepotinib derivatives were designed and synthesized as dual inhibitors of c-Met and TrxR. Among these compounds, compound 8b exhibits potent antiproliferative activity against MHCC97H cells, with an IC 50 value of 0.010 µM. In addition, compound 8b induces the accumulation of intracellular ROS via inhibiting TrxR. Studies on the mechanism of action reveal that compound 8b arrests the cell cycle at the G 1 phase and induces cellular apoptosis. The present findings strongly suggest that compound 8b may serve as a potent inhibitor of c-Met and TrxR and deserves further study.