Design and Synthesis of Acridine-Triazole and Acridine-Thiadiazole Derivatives and Their Inhibitory Effect against Cancer Cells

We report herein the design and synthesis of a series of novel acridine-triazole and acridine-thiadiazole derivatives. The newly synthesized compounds and the key intermediates were all evaluated for their antitumor activities against human foreskin fibroblasts (HFF), human gastric cancer cells-803 (MGC-803), hepatocellular carcinoma bel-7404 (BEL-7404), large cell lung cancer cells (NCI-H460), and bladder cancer cells (T24). Most of the compounds exhibited high levels of antitumor activity against MGC-803 and T24 but low toxicity against human normal liver cells (LO2), and their effect was even better than the commercial anticancer drugs, 5-fluorouracil (5-FU) and cis-platinum. Further, pharmacological mechanisms such as topo I, cell cycle, cell apoptosis, and neovascularization were all evaluated. Only a few compounds exhibited potent topo I inhibitory activity at 100 μM. In addition, the most active compounds with an IC50 value of 5.52–8.93 μM were chosen, and they could induce cell apoptosis in the G2 stage of MGC-803 or mainly arrest T24 cells in the S stage. To our delight, most of the compounds exhibited lower zebrafish cytotoxicity but could strongly inhibit the formation of zebrafish sub-intestinal veins, indicating a potential for clinical application.


Introduction
Today, cancer is one of the major health problems in the world. With the development of molecular biology and molecular pharmacology, the pathogenesis of cancer is being explored at the gene level. Pharmacological mechanisms such as signal transduction, neovascularization, telomerase, topoisomerase, cell cycle and cell apoptosis have major impacts on cancerous cells and can be used as targets in cancer therapy [1].
Acridines are an important classe of nitrogen-containing heterocyclic compounds. Due to their structural characteristics as planar tricyclic aromatic molecules, acridines intercalate tightly but reversibly to the DNA helix [2,3]. These compounds reveal a wide variety of biological activities, including anticancer [4], antimicrobial [5,6], anti-acetylcholinesterase [7], etc. A number of acridine derivatives serve as chemotherapeutic agents, especially in the field of antitumor DNA-binding agents [8]. An example of one such compound is 9-amsacrine, which has been clinically used for the treatment of leukemia [9].
Due to their beneficial characteristics, triazole and thiadiazole derivatives can serve as potential antitumor agents and thus are of pharmaceutical interest. In drug development, the triazole ring is often used to replace the amino group to reduce the resistance of some anticancer drugs and enhance their anticancer activity [10]. Thiadiazole groups are commonly introduced in the design of anticancer drugs because of their high anticancer activity. Kumar et al. recently reported the synthesis and anticancer activity of a series of benzpyrole-thiadiazole derivatives and revealed the important role of the thiadiazole ring in cytotoxicity [11].
Designing hybrid drugs with multiple effects is a common strategy in the recent search for new anticancer drugs [12]. In recent years, many structurally diverse hybrid molecules at the 9-position of the acridine skeleton have been reported for the enhancement of anticancer activity. Examples of such compounds include acridine-mycophenolic acid hybrid (a) [13], acridine-thiazolidinedione hybrid [14] (b), and acridine-chlormethine hybrid (c) [15] ( Figure 1). such compound is 9-amsacrine, which has been clinically used for the treatment of leukemia [9].
Due to their beneficial characteristics, triazole and thiadiazole derivatives can serve as potential antitumor agents and thus are of pharmaceutical interest. In drug development, the triazole ring is often used to replace the amino group to reduce the resistance of some anticancer drugs and enhance their anticancer activity [10]. Thiadiazole groups are commonly introduced in the design of anticancer drugs because of their high anticancer activity. Kumar et al. recently reported the synthesis and anticancer activity of a series of benzpyrole-thiadiazole derivatives and revealed the important role of the thiadiazole ring in cytotoxicity [11].
Designing hybrid drugs with multiple effects is a common strategy in the recent search for new anticancer drugs [12]. In recent years, many structurally diverse hybrid molecules at the 9-position of the acridine skeleton have been reported for the enhancement of anti-cancer activity. Examples of such compounds include acridinemycophenolic acid hybrid (a) [13], acridine-thiazolidinedione hybrid [14] (b), and acridine-chlormethine hybrid (c) [15] (Figure 1). Considering these facts, our strategy was to couple an acridine and a triazole or thiadiazole nucleus to obtain a new class of compounds such as the acridine-triazole hybrid or acridine-thiadiazole hybrid ( Figure 2). The anticancer activities of the synthesized compounds were assessed based on various mechanisms of action and molecular docking.

Chemistry
The general synthetic approach for aroyl thiourea derivatives (4), acridinyl 1,2,4triazole derivatives (5) and acridinyl 1,2,4-thiadiazole derivatives (6) is illustrated in Scheme 1. Considering these facts, our strategy was to couple an acridine and a triazole or thiadiazole nucleus to obtain a new class of compounds such as the acridine-triazole hybrid or acridine-thiadiazole hybrid ( Figure 2). The anticancer activities of the synthesized compounds were assessed based on various mechanisms of action and molecular docking.
Designing hybrid drugs with multiple effects is a common stra search for new anticancer drugs [12]. In recent years, many structura molecules at the 9-position of the acridine skeleton have been enhancement of anti-cancer activity. Examples of such compounds mycophenolic acid hybrid (a) [13], acridine-thiazolidinedione hyb acridine-chlormethine hybrid (c) [15] (Figure 1).
The target compounds of 1,2,4-triazolethiones (5) and 1,2,4-thiadiazoles (6) were synthesized by means of a ring closure reaction using aroyl thiourea derivatives (4) in sodium carbonate or concentrated sulfuric acid conditions, respectively. The synthesis of aroyl thiourea derivatives (4) was carried out according to the known procedure of the addition of substituted hydrazides to acridin-9-yl isothiocyanate (3). It is important to note that the precipitate 3a is formed at room temperature, while 3b needs to be cooled in an ice bath. The key intermediates (4) were obtained in 95% EtOH without purification with a yield of 73-92% w /w.
As expected, auto-condensation cyclization proceeded effectively in the refluxing condition of 5% Na2CO3 or 98% concentrated sulfuric acid in an ice bath. It is reported that acridinyl 1,2,4-triazole derivatives (5) possibly exist in one of two tautomeric forms ( Figure 3), thione (a) or thiol (b) [16]. And the thione form (a) was established by comparison of the HSQC and HMBC spectra and DFT calculations. To further confirm the structure of our synthesized products, a single crystal of compound 5b was cultivated in absolute ethyl alcohol, and the molecular structure was confirmed as indicated in Figure  3c. The corresponding single crystal structural data for compound 5b is provided in the supporting information (CCDC 2214949).  The target compounds of 1,2,4-triazolethiones (5) and 1,2,4-thiadiazoles (6) were synthesized by means of a ring closure reaction using aroyl thiourea derivatives (4) in sodium carbonate or concentrated sulfuric acid conditions, respectively. The synthesis of aroyl thiourea derivatives (4) was carried out according to the known procedure of the addition of substituted hydrazides to acridin-9-yl isothiocyanate (3). It is important to note that the precipitate 3a is formed at room temperature, while 3b needs to be cooled in an ice bath. The key intermediates (4) were obtained in 95% EtOH without purification with a yield of 73-92% w /w.
As expected, auto-condensation cyclization proceeded effectively in the refluxing condition of 5% Na2CO3 or 98% concentrated sulfuric acid in an ice bath. It is reported that acridinyl 1,2,4-triazole derivatives (5) possibly exist in one of two tautomeric forms ( Figure 3), thione (a) or thiol (b) [16]. And the thione form (a) was established by comparison of the HSQC and HMBC spectra and DFT calculations. To further confirm the structure of our synthesized products, a single crystal of compound 5b was cultivated in absolute ethyl alcohol, and the molecular structure was confirmed as indicated in Figure  3c. The corresponding single crystal structural data for compound 5b is provided in the supporting information (CCDC 2214949). The target compounds of 1,2,4-triazolethiones (5) and 1,2,4-thiadiazoles (6) were synthesized by means of a ring closure reaction using aroyl thiourea derivatives (4) in sodium carbonate or concentrated sulfuric acid conditions, respectively. The synthesis of aroyl thiourea derivatives (4) was carried out according to the known procedure of the addition of substituted hydrazides to acridin-9-yl isothiocyanate (3). It is important to note that the precipitate 3a is formed at room temperature, while 3b needs to be cooled in an ice bath. The key intermediates (4) were obtained in 95% EtOH without purification with a yield of 73-92% w / w .
As expected, auto-condensation cyclization proceeded effectively in the refluxing condition of 5% Na 2 CO 3 or 98% concentrated sulfuric acid in an ice bath. It is reported that acridinyl 1,2,4-triazole derivatives (5) possibly exist in one of two tautomeric forms ( Figure 3), thione (a) or thiol (b) [16]. And the thione form (a) was established by comparison of the HSQC and HMBC spectra and DFT calculations. To further confirm the structure of our synthesized products, a single crystal of compound 5b was cultivated in absolute ethyl alcohol, and the molecular structure was confirmed as indicated in Figure 3c. The corresponding single crystal structural data for compound 5b is provided in the supporting information (CCDC 2214949).  Figure 3. The molecular structure of compound 5b.
The success of the cyclization of compound 6 mainly depended on reaction temperature and reaction time. The reaction temperature had to be maintained below 0 °C. When R2 was an electron-withdrawing group such as pyridyl and nitrophenyl, the reaction time had to be extended almost to 48 h. Interestingly, the final structure of compound 6 was not the desired acridine skeleton (a, Figure 4) for the compound. The N-10 atom of the acridinyl moiety captured a proton and thus resulted in the formation of a 9′,10′-dihydroacridine structure (b, Figure 4), which was verified through X-ray crystallographic analysis (c). The corresponding single crystal structural data of compound 6d is provided in the supporting information (CCDC 2214923). The exchangeable NH protons of acridine thiosemicarbazides are reported in the literature ( Figure 5) [16].

In-Vitro Anticancer Activity Assay and Structure-Activity Analysis
All newly synthesized acridinyl derivatives (4-6) were screened for their anticancer activities in comparison to the reference compounds, 5-FU and cis-platinum. Compounds The success of the cyclization of compound 6 mainly depended on reaction temperature and reaction time. The reaction temperature had to be maintained below 0 • C. When R 2 was an electron-withdrawing group such as pyridyl and nitrophenyl, the reaction time had to be extended almost to 48 h. Interestingly, the final structure of compound 6 was not the desired acridine skeleton (a, Figure 4) for the compound. The N-10 atom of the acridinyl moiety captured a proton and thus resulted in the formation of a 9 ,10 -dihydroacridine structure (b, Figure 4), which was verified through X-ray crystallographic analysis (c). The corresponding single crystal structural data of compound 6d is provided in the supporting information (CCDC 2214923). The exchangeable NH protons of acridine thiosemicarbazides are reported in the literature ( Figure 5) [16]. The success of the cyclization of compound 6 mainly depended on reaction temperature and reaction time. The reaction temperature had to be maintained below 0 °C. When R2 was an electron-withdrawing group such as pyridyl and nitrophenyl, the reaction time had to be extended almost to 48 h. Interestingly, the final structure of compound 6 was not the desired acridine skeleton (a, Figure 4) for the compound. The N-10 atom of the acridinyl moiety captured a proton and thus resulted in the formation of a 9′,10′-dihydroacridine structure (b, Figure 4), which was verified through X-ray crystallographic analysis (c). The corresponding single crystal structural data of compound 6d is provided in the supporting information (CCDC 2214923). The exchangeable NH protons of acridine thiosemicarbazides are reported in the literature ( Figure 5) [16].

In-Vitro Anticancer Activity Assay and Structure-Activity Analysis
All newly synthesized acridinyl derivatives (4-6) were screened for their anticancer activities in comparison to the reference compounds, 5-FU and cis-platinum. Compounds The success of the cyclization of compound 6 mainly depended temperature and reaction time. The reaction temperature had to be mainta °C. When R2 was an electron-withdrawing group such as pyridyl and nit reaction time had to be extended almost to 48 h. Interestingly, the fina compound 6 was not the desired acridine skeleton (a, Figure 4) for the comp 10 atom of the acridinyl moiety captured a proton and thus resulted in the f 9′,10′-dihydroacridine structure (b, Figure 4), which was verified th crystallographic analysis (c). The corresponding single crystal struct compound 6d is provided in the supporting information (CCDC 2 exchangeable NH protons of acridine thiosemicarbazides are reported in ( Figure 5) [16].

In-Vitro Anticancer Activity Assay and Structure-Activity Analysis
All newly synthesized acridinyl derivatives (4-6) were screened for th activities in comparison to the reference compounds, 5-FU and cis-platinum

In-Vitro Anticancer Activity Assay and Structure-Activity Analysis
All newly synthesized acridinyl derivatives (4-6) were screened for their anticancer activities in comparison to the reference compounds, 5-FU and cis-platinum. Compounds 4-6 were tested for their in vitro antitumor activities against HFF, MGC-803, BEL-7404, NCI-H460, and T24 tumor cell lines, and human normal liver cells (LO2), and the results are shown in Table 1. Most of the compounds had strong selective potency against MGC-803 and T24 cancer cells. In the MGC-803 cell line assay, almost all of the compounds displayed better cytotoxicity than the positive control 5-FU (IC 50 = 30.45 ± 2.87 µM), with an IC 50 of 5.52-34.99 µM. This indicates that the introduction of the triazole and thiadiazole groups on the acridine skeleton could improve the antitumor activity against MGC-803. In addition, except for compounds 4c, 4d, 4e, 4i, 4j, 5a, 5g, 6b, 6d, and 6i, almost all of the compounds demonstrated better cytotoxicity inhibition than cis-platinum (IC 50 = 15.97 ± 1.53 µM). Particularly, the IC 50 values of compounds 5d, 5g, 5i, 6g, 6e, and 6h were all below 10 µM, and the IC 50 of them were 5.52 ± 1.04 µM, 8.5 ± 1.85 µM, 8.92 ± 0.99 µM, 9.01 ± 1.32 µM, 9.95 ± 1.03 µM, and 6.85 ± 0.84 µM, respectively. In the T24 cell line assay, many compounds, especially the series of compound 4, had significant activity against T24. This implies that there is a significant increase in potency after the introduction of the aroyl thiourea group. Among these compounds, R 1 = -CH 3 and R 2 = -OCH 3 might help to improve the antitumor activity of acridine nuclear, such as compounds 4h, 5h and 6h, all of which exhibited the best inhibition compared with other analogues, with IC 50 values of 8.05 ± 1.06, 11.25 ± 1.16, and 8.93 ± 1.25 µM, respectively. In particular, compounds 4h and 6h had better antitumor activities than the two commercial anticancer drugs 5-FU (IC 50 = 32.04 ± 1.23) and cis-platinum (IC 50 = 9.13 ± 1.54 µM). To our delight, most 1,2,4-triazolethiones (5) and 1,2,4-thiadiazoles (6) have low toxicity to LO2 compared with the positive control. Compounds 5d and 6h were the most active but had lower toxicities than 5-FU and cis-platinum. Therefore, compounds 5d and 6h or 4h and 6h exhibited good cytotoxicity inhibition against MGC-803 or T24 cancer cells and were selected for further exploration to identify their mechanisms of cancer cell growth inhibition.

Apoptosis and Cell-Cycle Analysis
Apoptosis and the cell-cycle play a central role in cancer, since their induction in cancer cells is critical to a successful therapy [17,18]. Therefore, the most active compounds, including 5d and 6h or 4h and 6h were selected to study their effect on apoptosis and cell cycle profiles in the MGC80-3 or T24 cell lines, respectively.
The apoptosis ratios of MGC80-3 or T24 cell lines induced by the selected compounds at the concentration of IC 50 and 0.5 IC 50 were quantitatively determined by flow cytometry. Four quadrant images (Q1, Q2, Q3 and Q4) were observed by flow cytometric analysis. The results of apoptosis ratios (including the early and late apoptosis ratios) after 12 h are presented in Figure 6 (MGC80-3) and Figure 7 (T24). Figure 6 revealed that compounds 5d and 6h could induce apoptosis in MGC80-3 cells in a concentration dependent manner. The apoptosis percentage of compound 5d measured at different concentrations were found to be 6.616% (2.76 µM) and 17.51% (5.52 µM), while the value for control was 0.586%. Treatment was also accompanied by a decrease in the percentage of live cells, with values of 93.0% in control and 81.2% in treated cells. After treatment with compound 6h, 5.62% (3.43 µM) and 14.25% (6.85 µM) of the cells were apoptotic. These were higher percentages than the one observed in the control (0.586%). These results further demonstrate that apoptosis was induced by compounds 5d and 6h in addition to cell proliferation inhibition. From the results of Figure 7, compounds 4h and 6h led to an increase in the number of apoptotic cells in T24 with the increase of the concentration (from 0.5 IC 50 to IC 50 ), and their apoptosis ratios at their IC 50 concentrations were increased to 12.377% and 10.749%, respectively, when compared with the control (1.18%). All compounds had little effect on late apoptosis of MGC80-3 or T24, and some normal cells were found to be necrotic in Q1 region. The results evidently illustrate that representative compounds 5d and 6h or 4h and 6h could suppress cell proliferation by inducing apoptosis in the early apoptotic period.    The cell cycle distributions of T24 and MGC80-3cells after 48 h of treatment with the most active compounds, 4h and 5d, at their IC50 concentrations are shown in Figure 8. Compared to control, both compounds 4h and 5d interfered with the cell cycles of T24 and MGC80-3 cells, respectively. As shown in Figure 8a,b, the S-phase population of T24 cells increased by 30.04% compared to the control cells (22.89%), indicating that compound 4h might inhibit the growth of tumor cells by arresting the cells in S phase during the DNA synthesis period. However, compound 5d could induce a significant cell cycle arrest in the G2 phase, resulting in a concomitant population increase (13.32%) compared with the control cells (8.91%) at a concentration of 5.52 μM (Figure 8c,d). These results suggest that compound 5d may inhibit the growth of tumor cells by arresting cells in the G2 phase in the late stage of DNA synthesis. The cell cycle distributions of T24 and MGC80-3cells after 48 h of treatment with the most active compounds, 4h and 5d, at their IC 50 concentrations are shown in Figure 8. Compared to control, both compounds 4h and 5d interfered with the cell cycles of T24 and MGC80-3 cells, respectively. As shown in Figure 8a,b, the S-phase population of T24 cells increased by 30.04% compared to the control cells (22.89%), indicating that compound 4h might inhibit the growth of tumor cells by arresting the cells in S phase during the DNA synthesis period. However, compound 5d could induce a significant cell cycle arrest in the G2 phase, resulting in a concomitant population increase (13.32%) compared with the control cells (8.91%) at a concentration of 5.52 µM (Figure 8c,d). These results suggest that compound 5d may inhibit the growth of tumor cells by arresting cells in the G2 phase in the late stage of DNA synthesis.

Evaluation of Topo I Inhibitory Activity
DNA topoisomerase I (topo I) has become the main molecular target in anticancer drugs on account of its significance in all living organisms, participating in replication, transcription, recombination, and repair in many cellular metabolic processes. The topo I inhibitory activity of the compounds with the known topo I inhibitor camptothecin (CPT) is depicted in Figure 9. Only compounds 4e, 5c, and 6h exhibit potent topo I inhibitory activity at 100 μM. Compounds that have little to no inhibitory activity may have other mechanisms for their anticancer effects. Molecular docking studies of the selected compounds were carried out by the Surflex-Dock algorithm of Sybyl-X 2.0 (Tripos Inc., St. Louis, MI, USA). The molecular docking approach was verified by our previously published methods (RMSD (root-mean-square deviation) value was 0.4438 Å) [19]. The binding affinities of protein-ligand complexes were expressed as a total score and shown in Figure 10. compounds 4e, 5c and 6h exhibited good binding affinities, with total scores of 9.79, 7.81 and 9.66, respectively. Potent Topo I inhibitory activity of these compounds

Evaluation of Topo I Inhibitory Activity
DNA topoisomerase I (topo I) has become the main molecular target in anticancer drugs on account of its significance in all living organisms, participating in replication, transcription, recombination, and repair in many cellular metabolic processes. The topo I inhibitory activity of the compounds with the known topo I inhibitor camptothecin (CPT) is depicted in Figure 9. Only compounds 4e, 5c, and 6h exhibit potent topo I inhibitory activity at 100 µM. Compounds that have little to no inhibitory activity may have other mechanisms for their anticancer effects. Molecular docking studies of the selected compounds were carried out by the Surflex-Dock algorithm of Sybyl-X 2.0 (Tripos Inc., St. Louis, MI, USA). The molecular docking approach was verified by our previously published methods (RMSD (root-mean-square deviation) value was 0.4438 Å) [19]. The binding affinities of proteinligand complexes were expressed as a total score and shown in Figure 10. compounds 4e, 5c and 6h exhibited good binding affinities, with total scores of 9.79, 7.81 and 9.66, respectively. Potent Topo I inhibitory activity of these compounds may be attributed to the formation of hydrophobic residue, hydrogen bond, and π-π stacking with the same amino acid residue DA113, DC112, TGP11 as CPT. may be attributed to the formation of hydrophobic residue, hydrogen bond, and π-π stacking with the same amino acid residue DA113, DC112, TGP11 as CPT.

Toxicity and Anti-angiogenesis in the Zebrafish Model
Many antitumor drugs inevitably have side effects on normal cells, such as bone marrow suppression, liver and kidney injury, and abnormal blood cells. Therefore, in order to improve the possibility of clinical application of acridine-heterocyclic derivatives, the effective and low-toxicity antitumor drugs were screened using a zebrafish model. In this experiment, 2% DMSO was used to dissolve the target products, and the abnormal rate (MAR) and mortality rate (MOR) of zebrafish embryos (72 hpf) were used for statistics. At different concentrations (1-2 mg/mL) of the selected compounds, various deformities were observed, such as failure to hatch, embryo necrosis, severe angulation of the spine and severe pericardial edema ( Figure 11). The mortality and malformation rates of embryos increased with a dose-effect relationship ( Table 2). Almost all compounds in the compound 4 series were toxic. Particularly, compounds 4a, 4b, 4f and 4h had a total mortality and malformation rate of 100% at 2.0 mg.L −1 , exhibiting the strongest embryonic toxicity. It is worth mentioning that compounds 5d and 5h displayed high levels of antitumor activities but were less toxic to zebrafish embryos. At the highest concentration of 2.0 mg.L −1 , the mortality rate of zebrafish embryos was close to 0% and the malformation rate was less than 15%. Moreover, there was hardly any toxicity observed in compound 6 at lower concentrations (1.0 mg.L −1 ). Compound 6i exhibited very low toxicity at a high concentration of 2.0 mg.L −1 with 0% mortality rate and 25% malformation rate.  Figure 10. The best pose of the binding mode of compounds (4e, 5c, 6h and CPT) with DNA Topo I complex (PDB:1T8I).

Toxicity and Anti-angiogenesis in the Zebrafish Model
Many antitumor drugs inevitably have side effects on normal cells, such as bone marrow suppression, liver and kidney injury, and abnormal blood cells. Therefore, in order to improve the possibility of clinical application of acridine-heterocyclic derivatives, the effective and low-toxicity antitumor drugs were screened using a zebrafish model. In this experiment, 2% DMSO was used to dissolve the target products, and the abnormal rate (MAR) and mortality rate (MOR) of zebrafish embryos (72 hpf) were used for statistics. At different concentrations (1-2 mg/mL) of the selected compounds, various deformities were observed, such as failure to hatch, embryo necrosis, severe angulation of the spine and severe pericardial edema (Figure 11). The mortality and malformation rates of embryos increased with a dose-effect relationship ( Table 2). Almost all compounds in the compound 4 series were toxic. Particularly, compounds 4a, 4b, 4f and 4h had a total mortality and malformation rate of 100% at 2.0 mg.L −1 , exhibiting the strongest embryonic toxicity. It is worth mentioning that compounds 5d and 5h displayed high levels of antitumor activities but were less toxic to zebrafish embryos. At the highest concentration of 2.0 mg.L −1 , the mortality rate of zebrafish embryos was close to 0% and the malformation rate was less than 15%. Moreover, there was hardly any toxicity observed in compound 6 at lower concentrations (1.0 mg.L −1 ). Compound 6i exhibited very low toxicity at a high concentration of 2.0 mg.L −1 with 0% mortality rate and 25% malformation rate.  Currently, the zebrafish has emerged as a valuable model organism to substitute traditional models for studying angiogenesis inhibitors [20]. The genes of zebrafish show 70-80% similarity to humans, and the vascular structure of zebrafish has high similarity to that of other vertebrates [21][22][23]. Therefore, the subintestinal veins (SIVs) in the zebrafish embryos are used as evaluation indicators for anti-angiogenesis inhibitors. In this study, NBT/BCIP vascular staining was used to observe the angiogenesis effect of representative drugs (4h, 4f, 5d, 5h, 6g and 6h) in a zebrafish model. As shown in Figure 12, SIVs grew well in the blank group, naturally extending into a network in the abdomen with many branches. The length of SIVs of zebrafish was measured by Image J software and is shown in Figure 10b. Compared with the blank control group, the area of the meshed pattern vessel and the number of vascular branches in the network decreased after the administration of compounds 4h, 5d, 5h and 6h. Among these compounds, compounds 5d and 5h exhibited the strongest antiangiogenic effects that led to a nearly 50% reduction in the vessel length compared to the mean vessel length for the controls. Compounds 4a and 6g could reduce the area of blood vessels, but at the same time, additional blood vessels were formed on the blood vessel edge.
Currently, the zebrafish has emerged as a valuable model organism to substitute traditional models for studying angiogenesis inhibitors [20]. The genes of zebrafish show 70-80% similarity to humans, and the vascular structure of zebrafish has high similarity to that of other vertebrates [21][22][23]. Therefore, the subintestinal veins (SIVs) in the zebrafish embryos are used as evaluation indicators for anti-angiogenesis inhibitors. In this study, NBT/BCIP vascular staining was used to observe the angiogenesis effect of representative drugs (4h, 4f, 5d, 5h, 6g and 6h) in a zebrafish model. As shown in Figure  12, SIVs grew well in the blank group, naturally extending into a network in the abdomen with many branches. The length of SIVs of zebrafish was measured by Image J software and is shown in Figure 10b. Compared with the blank control group, the area of the meshed pattern vessel and the number of vascular branches in the network decreased after the administration of compounds 4h, 5d, 5h and 6h. Among these compounds, compounds 5d and 5h exhibited the strongest antiangiogenic effects that led to a nearly 50% reduction in the vessel length compared to the mean vessel length for the controls. Compounds 4a and 6g could reduce the area of blood vessels, but at the same time, additional blood vessels were formed on the blood vessel edge.   4f, 4h, 5d, 5h, 6g and 6h) and control on the subintestinal veins (SIVs) length of 72 hpf zebrafish embryos (x ± s, n = 12), p < 0.05.

Materials and Methods
All commercially available chemicals were reagent grade and bought from Aladdin Reagent Co., LTD (Shanghai, China); NBT/BCIP kit was bought from Tiangen Biochemical Technology Co., LTD (Shanghai, China); AnnexinV-FITC apoptosis detection kit was bought from Nanjing KGI Biotechnology Development Co., Ltd. (Nanjing, China); The spectra such as NMR, MS, and IR were all evaluated and recorded on a Bruker DRX-400 ( 1 H: 400 MHz, 13  The synthesis of N-phenyl-o-aminobenzoic acid (1) and 9-chlorine acridine (2) was carried out according to our previously published procedure, with slight modifications [24]. Compound 1 could proceed to the next step without further purification.
2-methoxyl-9-chlorine acridine (  (3) To a solution of chlorine acridine 2 (5 mmol) in acetone (50 mL), NaSCN (0.81 g, 10 mmol) and tetrabutylammonium bromide (0.32 g, 1 mmol) were added, and the mixture was then refluxed at 60 • C for 1 h. After cooling to room temperature, crystals of 3a were immediately precipitated in the reaction mixture, and crystals of 3b were precipitated in an ice bath. At the end of the procedure, the crystals were filtered, washed with water, and dried under vacuum, and no further purification was carried out. To a solution of 9-isothiocyanatoacridine 3 (2mmol) in absolute ethyl alcohol (60 mL), the appropriate substituted hydrazides (2 mmol) were added, and the reaction mixture was refluxed until the reactants had been consumed (monitored by TLC). The precipitate of 4a-4f was prepared, filtered off, washed with 95% ethyl alcohol, and dried at room temperature. Test samples, including compounds 4-6 and commercial classical anticancer drugs (5-FU and cis-platinum), were screened for their anti-cancer activity against HFF human foreskin fibroblast cells, MGC-803 human gastric cancer cells, BEL-7404 human hepatocellular carcinoma cells, NCI-H460 human large cell carcinoma cells, and T24 human bladder carcinoma cells using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assay method cited in the literature [27]. The initial concentration of all the test samples was 100 µg/mL, which was serially diluted in complete medium with ten-fold dilutions to give six concentrations per compound. Their cytotoxicity was determined in 96-well flat bottomed microtiter plates. All the test samples were tested in triplicate. The results were expressed as the drug concentration that inhibited cell growth by 50% as compared to the controls (IC 50 ). The IC 50 values were calculated from regression lines obtained from the percent cell growth inhibition plotted as a function of the logarithm of the dose.

Apoptosis and Cell Cycle Analysis
The apoptosis assay and the cell cycle analysis were carried out by cytometry (FACS-Verse, BD, Piscataway, NJ, USA) at an excitation wavelength of 488 nm according to the method described in the literature with slight modifications [28]. The cells were seeded at 2 × 10 6 /well and washed by cold PBS. The buffer solutions were prepared using 0.1 M pH 7.4 Hepes/NaOH, 1.4 M NaCl, and 25 mM CaCl 2 .

Topo I Inhibitory Activity
Topo I and pBR322 were obtained commercially from Takara Bio Inc. (Shiga, Japan). And the enzyme inhibitory activity was determined by our previous methods [19].
3.2.4. Anti-Angiogenic Effect Using the Zebrafish Model (a) Zebrafish toxicity assay The zebrafish embryos were collected at 6 hpf and randomly divided into naive control (embryos maintained in distilled water), vehicle control (embryos treated with 2% DMSO), and drug groups. Stock solutions of all drugs were prepared in 2% DMSO as a solubilizing agent and diluted to three concentrations (0.5 mg/mL, 1 mg/mL, and 2 mg/mL). Each group had 20 embryos per test concentration. The zebrafish embryos were maintained in an incubator at 28 • C and read at 72 hpf for their mortality and teratogenicity (including non-hatching, egg condensation, spinal curvature, pericardial enlargement, etc.). Each compound was evaluated in three independent biological experiments.

(b) Angiogenesis assay
The 24 hpf zebrafish embryos were dechorionated with a 1 gL −1 pronase treatment and maintained in distilled water in 12-well cell culture plates (each well contained 20 embryos). A negative control group containing zebrafish embryos in distilled water and a vehicle treatment group that was treated with 2% dimethyl sulfoxide (DMSO) were prepared. The anti-angiogenic compounds were diluted to 1 mg/mL. After incubating at 28 • C for 72 h, the embryos were immersed in 4% paraformaldehyde and dehydrated by gradient ethanol. Subsequently, the embryos were balanced in NTMT buffer (5 M NaCl + 1 M Tris [pH 9.0-9.5] + 1 M MgCl 2 + 10% Tween), and nitrotetrazolium blue chloride (NBT) and p-toluidine salt (BCIP) staining were performed. To evaluate the effect of compounds on the angiogenesis of zebrafish embryos, the growth of embryonic sub-intestinal veins (SIVs) at 72 hpf was observed using an IX71 Olympus microscope (Hamburger, Germany). The length of the SIVs was calculated using the image J 1.8.0 software (Bethesda, MD, USA).

Conclusions
A new series of acridine-triazole and acridine-thiadiazole derivatives were synthesized and characterized by spectral studies. All the synthesized compounds were evaluated for their in vitro cytotoxic activities against HFF, MGC-803, BEL-7404, NCI-H460, and T24 by the MTT assay method. Most of the compounds were sensitive to MGC-803 and T24 cell lines. Compared to all the prepared compounds, 4a, 5d and 6h exhibited the best anticancer activity against MGC-803 cell lines, and compounds 4h, 5h and 6h showed the most excellent antitumor activity against T24. Preliminary studies of antitumor mechanisms revealed that the representative compounds (5d and 6h or 4h and 6h) could suppress cell proliferation by inducing apoptosis in the Q3 period of MGC-803 or T24 cell lines. Compound 5d might inhibit the growth of tumor cells by arresting cells in the G2 phase, while compound 4h had a great effect on the S phase. In the zebrafish experiment, compound 5d displayed a superior antiangiogenic effect and lower toxicity than other compounds. Therefore, compound 5d has the potential to be an antitumor drug with high efficiency and low toxicity.