Novel 5-Substituted 2-(Aylmethylthio)-4-chloro-N-(5-aryl-1,2,4-triazin-3-yl)benzenesulfonamides: Synthesis, Molecular Structure, Anticancer Activity, Apoptosis-Inducing Activity and Metabolic Stability

A series of novel 5-substituted 2-(arylmethylthio)-4-chloro-N-(5-aryl-1,2,4-triazin-3-yl) benzenesulfonamide derivatives 27–60 have been synthesized by the reaction of aminoguanidines with an appropriate phenylglyoxal hydrate in glacial acetic acid. A majority of the compounds showed cytotoxic activity toward the human cancer cell lines HCT-116, HeLa and MCF-7, with IC50 values below 100 μM. It was found that for the analogues 36–38 the naphthyl moiety contributed significantly to the anticancer activity. Cytometric analysis of translocation of phosphatidylserine as well as mitochondrial membrane potential and cell cycle revealed that the most active compounds 37 (HCT-116 and HeLa) and 46 (MCF-7) inhibited the proliferation of cells by increasing the number of apoptotic cells. Apoptotic-like, dose dependent changes in morphology of cell lines were also noticed after treatment with 37 and 46. Moreover, triazines 37 and 46 induced caspase activity in the HCT-116, HeLa and MCF-7 cell lines. Selected compounds were tested for metabolic stability in the presence of pooled human liver microsomes and NADPH, both R2 and Ar = 4-CF3-C6H4 moiety in 2-(R2-methylthio)-N-(5-aryl-1,2,4-triazin-3-yl)benzenesulfonamides simultaneously increased metabolic stability. The results pointed to 37 as a hit compound with a good cytotoxicity against HCT-116 (IC50 = 36 μM), HeLa (IC50 = 34 μM) cell lines, apoptosis-inducing activity and moderate metabolic stability.


Introduction
For decades chemotherapeutics have played the most important role in the fight against cancer. Unfortunately, the serious toxicities of conventional cytotoxic medicines have impelled and are still impelling researchers to focus on the development of new potent and selective anticancer drugs.

Introduction
For decades chemotherapeutics have played the most important role in the fight against cancer. Unfortunately, the serious toxicities of conventional cytotoxic medicines have impelled and are still impelling researchers to focus on the development of new potent and selective anticancer drugs.
In vitro tests for metabolic stability, which is one of the important parameters characterizing pharmacokinetic properties of a molecule, were performed on selected compounds. The experiments were run in the presence of human liver microsomes and NADPH.
For the most active compounds apoptosis-inducing activities on HCT-116, HeLa and MCF-7 cell lines were further investigated.
In vitro tests for metabolic stability, which is one of the important parameters characterizing pharmacokinetic properties of a molecule, were performed on selected compounds. The experiments were run in the presence of human liver microsomes and NADPH.
Compound 59 crystallizes in the triclinic system in the space group P1 . The asymmetric unit contains two sulfonamide molecules and four solvating dimethylformamide (DMF) molecules. The unit cell of the crystal contains two asymmetric parts (Z = 2). The two independent sulfonamide molecules have the same topological atom connectivity (constitution) but cannot be superimposed on each another (they are almost related by a mirror plane passing through SO2 group and bisecting the S-N-C and S-C-C planes). The solvent DMF molecules are not only filling the voids in the structure, but also participate in hydrogen bonding-two of carbonyl groups are acceptors of hydrogen from N-H of 1,2,4-triazine residue (O14-N4, O16-N9, see Table 1). The other two DMF molecules use their CO groups as acceptors for amidic NH groups (O13-N6, O15-N1, Table 1). Other intermolecular interactions responsible for crystal packing are π-π stacking operating between R1 (N3, C22, N4, N5, C23, C24) and R5 (C25-C30) with the distance between ring geometry centers (centroids) of 3.8252(3)Å and R6 (N8, C54, N9, N10, C55, C56) and R10 (C57-C52) 3.8952(3)Å. All other rings are separated by more than 4.4Å and interactions among them were neglected. The structures of the final compounds 27-60 were confirmed by IR, 1 H-NMR and 13 C-NMR spectroscopy. 1 H-NMR spectra showed the absence of characteristic of aminoguanidine NH signals at 4 and 6 ppm, confirming the success of the condensation reactions between the amine and carbonyl groups of the reagents leading to the heterocyclic ring formation. In addition, the presence in the 1 H-NMR spectra of a singlet at about 9 ppm corresponding to the H-6 proton in the 1,2,4-triazine fragment was observed. Moreover, X-ray analysis was done to confirm the proposed structures using the representative compound 59.
Compound 59 crystallizes in the triclinic system in the space group P1. The asymmetric unit contains two sulfonamide molecules and four solvating dimethylformamide (DMF) molecules. The unit cell of the crystal contains two asymmetric parts (Z = 2). The two independent sulfonamide molecules have the same topological atom connectivity (constitution) but cannot be superimposed on each another (they are almost related by a mirror plane passing through SO 2 group and bisecting the S-N-C and S-C-C planes). The solvent DMF molecules are not only filling the voids in the structure, but also participate in hydrogen bonding-two of carbonyl groups are acceptors of hydrogen from N-H of 1,2,4-triazine residue (O14-N4, O16-N9, see Table 1). The other two DMF molecules use their CO groups as acceptors for amidic NH groups (O13-N6, O15-N1, Table 1). Other intermolecular interactions responsible for crystal packing are π-π stacking operating between R1 (N3, C22, N4, N5, C23, C24) and R5 (C25-C30) with the distance between ring geometry centers (centroids) of 3.8252(3)Å and R6 (N8, C54, N9, N10, C55, C56) and R10 (C57-C52) 3.8952(3)Å. All other rings are separated by more than 4.4Å and interactions among them were neglected. The sulfonamide group seems to be deprotonated in the solid state, as there is a C-H bond from a DMF molecule directed to the sulfonamidic N ( Figure 2) which would not be beneficial without assumed ionization of the -SO 2 NH-fragment. Additionally, protonation of the nitrogen atom in position 2 of the 1,2,4-triazine is required by electrical neutrality and by formation of hydrogen bond to the carbonyl group from the neighbor DMF molecule. Thus, two charge-assisted hydrogen bonds are created: C-H¨¨¨N(´)-S and (+)N-H¨¨¨O between the sulfonamide and the solvating DMF.  The sulfonamide group seems to be deprotonated in the solid state, as there is a C-H bond from a DMF molecule directed to the sulfonamidic N ( Figure 2) which would not be beneficial without assumed ionization of the -SO2NH-fragment. Additionally, protonation of the nitrogen atom in position 2 of the 1,2,4-triazine is required by electrical neutrality and by formation of hydrogen bond to the carbonyl group from the neighbor DMF molecule. Thus, two charge-assisted hydrogen bonds are created: C-H···N(−)-S and (+)N-H···O between the sulfonamide and the solvating DMF. Bond lengths within the triazine ring are not helpful to attribute double bonds in the substructure. The 1,2,4-triazine residue is flat and almost coplanar with 3,4-dimethoxyphenyl which may be a consequence of beneficial π-π interactions with the neighbor molecule in crystal. The second solvent Bond lengths within the triazine ring are not helpful to attribute double bonds in the substructure. The 1,2,4-triazine residue is flat and almost coplanar with 3,4-dimethoxyphenyl which may be a consequence of beneficial π-π interactions with the neighbor molecule in crystal. The second solvent DMF molecule forms hydrogen bond with the amide N-H proton donor in the other branch of the main molecule. The other functional groups are not unusual and their geometry will therefore not be further analyzed.
a Analysis was performed using the MTT assay after 72 h of incubation. Values are expressed as the mean˘SD of at least three independent experiments; * Viability of cell lines at 100 µM of tested compounds were approximately 100%.

Investigation of Apoptotic Activity
The ability to induce apoptosis in cancer cells is a desired feature of a potential chemotherapeutic agent. Thus, the apoptosis-inducing activity was studied through biochemical markers, such as: DNA fragmentation, loss of mitochondrial membrane potential (∆ψ m ), phosphatidylserine translocation and caspase activation. Apoptotic-like changes in morphology of tested cell lines were also evaluated. The experiments were performed with the most active compounds (37 and 46) exhibited the highest activity in MTT tests.

Cell Morphology
To evaluate the changes in morphology of the treated cells, they were incubated with increasing concentrations of 37 (HCT-116, HeLa) or 46 (MCF-7) for 24 h and observed using light microscope. The most characteristic morphological changes (shrinkage of the cells, detachment from the surface) occurred in the HCT-116 and HeLa cells treated with 37 ( Figure 3).

Cell Cycle Analysis
One of the most common mechanisms in anticancer drug treatment is changes in cell cycle, which can be measured by DNA content [37]. Cell cycle distribution in HCT-116, HeLa and MCF-7 cells was examined to determine cycle arrest and/or presence of sub-G1 after treatment with compounds 37 and 46. Cells were coincubated with trazines and a reference drug (cisplatin) for 24 h, respectively. Cell populations data are expressed as the mean˘SD of at least three independent experiments.

Cell Morphology
To evaluate the changes in morphology of the treated cells, they were incubated with increasing concentrations of 37 (HCT-116, HeLa) or 46 (MCF-7) for 24 h and observed using light microscope. The most characteristic morphological changes (shrinkage of the cells, detachment from the surface) occurred in the HCT-116 and HeLa cells treated with 37 ( Figure 3).

Cell Cycle Analysis
One of the most common mechanisms in anticancer drug treatment is changes in cell cycle, which can be measured by DNA content [37]. Cell cycle distribution in HCT-116, HeLa and MCF-7 cells was examined to determine cycle arrest and/or presence of sub-G1 after treatment with compounds 37 and 46. Cells were coincubated with trazines and a reference drug (cisplatin) for 24 h, respectively. Cell populations data are expressed as the mean ± SD of at least three independent experiments.
Coincubation of HCT-116, HeLa and MCF-7 with appropriate triazines (37 and 46) caused appearance of sub-G1 stage and high concentrations led to fragmentation of DNA, resulting in a sub-G0/G1 peak of cell cycle. A significant accumulation of the HCT-116, HeLa and MCF-7 population in the sub-G1 phase indicated that the cells had undergone programmed cell death (apoptosis) for all of the analyzed compounds. Triazine coincubation did not cause significant differences in the cell cycle phases of the analyzed cell lines. Sub-G1 cell population increased in dose dependent manner (HCT-116 from 3.6%˘1.66% (control) to 48.6%˘13.5%; HeLa from 5.7%˘3.4% (control) to 21%˘3.4%) after treatment with compound 37 for 24 h. Figure 4 shows one representative experiment. Similar results were obtained for MCF-7 cells coincubated with compound 46 for 24 h (from 5.0%˘3.8% (control) to 20%˘5.6%) (Figure 4).
Coincubation of HCT-116, HeLa and MCF-7 with appropriate triazines (37 and 46) caused appearance of sub-G1 stage and high concentrations led to fragmentation of DNA, resulting in a sub-G0/G1 peak of cell cycle. A significant accumulation of the HCT-116, HeLa and MCF-7 population in the sub-G1 phase indicated that the cells had undergone programmed cell death (apoptosis) for all of the analyzed compounds. Triazine coincubation did not cause significant differences in the cell cycle phases of the analyzed cell lines.
As shown in Figure 5 HCT-116, HeLa and MCF-7 cells in the control group, high red fluorescence (J aggregates) were observed. However, exposure of cells even to low and moderate concentrations of compounds 37 and 46 remarkably decreased ∆ψ m which is the earliest indicators of an apoptotic cell death.
Loss of the mitochondrial membrane potential (Δψm) is one of the earliest indicators of an apoptosis [38]. It can be detected using fluorescent dye JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′tetraethylbenzimidazolylcarbocyanine iodide) exited with blue laser. JC-1 spontaneously forms dimer complexes (J aggregates) with red fluorescent or remain as a monomers with green fluorescence in cells with high and low Δψm, respectively. Appearance of increased population of PI-low/FITC-high (early apoptotic cells), PI-high/FITC-high (late apoptotic cells) gates was noticed for HCT-116 and HeLa cells coincubated with compound 37 (Figure 6). The increased levels of apoptotic cells were concentration dependent. For MCF-7 cells treated with compound 46 the results were not conclusive, as the differences between treated cells and control cells were statistically significant only for low 25 μM of 46 ( Figure 6). Appearance of increased population of PI-low/FITC-high (early apoptotic cells), PI-high/FITC-high (late apoptotic cells) gates was noticed for HCT-116 and HeLa cells coincubated with compound 37 ( Figure 6). The increased levels of apoptotic cells were concentration dependent. For MCF-7 cells treated with compound 46 the results were not conclusive, as the differences between treated cells and control cells were statistically significant only for low 25 µM of 46 ( Figure 6).

Caspase Activation
Caspase activation plays a central role in the execution of apoptosis and is required for the occurrence of its biochemical and morphological hallmarks, such as DNA fragmentation, formation of apoptotic bodies and chromatin condensation. The ability of the examined compounds to induce caspase activity was determined with the use of a fluorescent labeled caspase inhibitor, a carboxyfluorescein (FAM) derivative of valylalanylaspartic acid (VAD) fluoromethyl ketone (FMK).
FAM-VAD-FMK contains a target sequence recognized by active caspases (caspases 1 through 9). Binding to this sequence inhibits the enzymatic activity of caspases and allows for the determination of their activity through the direct measurement of fluorescent intensity of the bound inhibitor. The results of the research showed that the tested compounds 37 and 46 induced caspase activity in the examined cells lines, as shown by an increase in FAM-VAD-FMK fluorescence in the cell population with activated caspases (Figure 7). The strongest influence on caspase activation was noticed for the HeLa cell line, where that cell population increased by 31% in the presence of 37 at a concentration of 100 µM.  their activity through the direct measurement of fluorescent intensity of the bound inhibitor. The results of the research showed that the tested compounds 37 and 46 induced caspase activity in the examined cells lines, as shown by an increase in FAM-VAD-FMK fluorescence in the cell population with activated caspases (Figure 7). The strongest influence on caspase activation was noticed for the HeLa cell line, where that cell population increased by 31% in the presence of 37 at a concentration of 100 μM.

Metabolic Stability
Metabolism as one of the drugs' pharmacokinetic characteristics can be evaluated during the early preclinical stage to assessment of degree of drug candidate conversion into a set of metabolites. Metabolic stability can be assessed by incubation of a potent drug in a presence of liver microsomes and NADPH to give an insight to metabolic properties [39][40][41].
For further evaluation of metabolic stability the compounds with outstanding activities and varied structural features including R 1 , R 2 and Ar substituents (30, 31, 34-38, 46, 47, 52) were selected. Human liver microsomes were chosen as a model enzymatic system and in vitro metabolic half-life was assessed (Table 3).
Based on the enzymatic test results some structure-metabolic stability relationships can be noted. The most stable compound 34 bears a 4-CF 3 -C 6 H 4 moiety in both the R 1 and Ar positions. Replacement of Ar = 4-CF 3 -C 6 H 4 (34) by 4-MeO-C 6 H 4 (compound 35) resulted in decrease of the t 1/2 value from >60 to 42.4 min, while an analogous change between 46 and 47 caused a significant decrease of t 1/2 from >60 to 17.5 min. On the other hand, taking into account the results for 30 (Ar = 3-MeO-C 6 H 4 ) and 31 (Ar = 3,4-diMeO-C 6 H 3 ), it was found that an additional methoxy (MeO) group in the Ar substituent slightly decreased the metabolic stability. These facts suggest that the methoxy substituent is a major soft spot in the described series of compounds and moreover, its undesirable influence can be diminished by incorporation of CF 3 instead of a MeO substituent. Comparison of t 1/2 for 34 (R 1 = 4-CF 3 -C 6 H 4 ) and 37 (R 1 = 1-naphthyl) indicated that 1-naphthyl (R 1 ) also seems to decrease metabolic stability. We assumed that the undesirable influence of the 1-naphthyl (R 1 ) and 4-MeO-C 6 H 4 (Ar) substituents accelerated in the case of compound 38, which was characterized as the least stable derivative. Further synthesis in this group of compounds should avoid incorporation of the two abovementioned substituents. Although a 4-CF 3 -C 6 H 4 (R 1 ) substituent is preferable for metabolic stability in comparison with a 1-naphthyl moiety, the high biological response observed for 37 predisposes it to be a lead compound, as it shows a good balance between stability and cytotoxic properties.
In order to explain the differences between the metabolic stability of distinctive compounds (34-35, 37-38) we applied a tool accessible on-line for accurate prediction of xenobiotic metabolism sites, called XenoSite Cytochrome P450 Prediction Models [42]. Among the available models, one is able to predict which atoms on a molecule are likely to be oxidized by human liver microsomes.
In silico results showed that the least stable derivative 38 has three more sites vulnerable for metabolic biotransformation than 34 (see Figure 8). Moreover, a slight decrease of metabolic stability of 35 and 37 compared to 34 may resulted from presence of additional sites of oxidation on methoxy group and methylene linker (35, Figure 8) as well as on methylene linker and naphthalene ring (37, Figure 8).  Comparison of t1/2 for 34 (R 1 = 4-CF3-C6H4) and 37 (R 1 = 1-naphthyl) indicated that 1-naphthyl (R 1 ) also seems to decrease metabolic stability. We assumed that the undesirable influence of the 1-naphthyl (R 1 ) and 4-MeO-C6H4 (Ar) substituents accelerated in the case of compound 38, which was characterized as the least stable derivative. Further synthesis in this group of compounds should avoid incorporation of the two abovementioned substituents. Although a 4-CF3-C6H4 (R 1 ) substituent is preferable for metabolic stability in comparison with a 1-naphthyl moiety, the high biological response observed for 37 predisposes it to be a lead compound, as it shows a good balance between stability and cytotoxic properties.
In order to explain the differences between the metabolic stability of distinctive compounds (34-35, 37-38) we applied a tool accessible on-line for accurate prediction of xenobiotic metabolism sites, called XenoSite Cytochrome P450 Prediction Models [42]. Among the available models, one is able to predict which atoms on a molecule are likely to be oxidized by human liver microsomes. In silico results showed that the least stable derivative 38 has three more sites vulnerable for metabolic biotransformation than 34 (see Figure 8). Moreover, a slight decrease of metabolic stability of 35 and 37 compared to 34 may resulted from presence of additional sites of oxidation on methoxy group and methylene linker (35, Figure 8) as well as on methylene linker and naphthalene ring (37, Figure 8).  [42]. Yellow color indicates more vulnerability to biotransformation than blue. Some significant differences are additionally pointed out by red arrows.

X-ray Structure Determination
Crystals of 59 were grown from DMF solution. All specimens obtained were in form of very thin needles, having poor diffraction power. The copper radiation (almost no diffraction was observed with the Mo lamp) and long exposure times (4 min) were applied. The obtained signal to noise ratio was still smaller than usually, so final R int and R 1 indices are above the regular standards for small molecules. Nevertheless, structure solution is with no doubts generally correct. Diffraction intensity data were collected on an IPDS 2T dual-beam diffractometer (STOE & Cie GmbH, Darmstadt, Germany) at 120.(2) K with Cu-Kα radiation of a microfocus X-ray source (50 kV, 0.6 mA, λ = 154.186 pm, GeniX 3D Cu High Flux, Xenocs, Sassenage, France), The crystal was thermostated in nitrogen stream at 120 K using CryoStream-800 device (Oxford CryoSystem, Oxford, UK) during the entire experiment. Data collection and data reduction were controlled by X-Area 1.75 program [43]. An absorption correction was performed on the integrated reflections by a combination of frame scaling, reflection scaling and a spherical absorption correction. Outliers have been rejected according to Blessing's method [44].
The structure was solved using direct methods with SHELXS-13 program and refined by SHELXL-2013 [45] program run under control of WinGx [46]. All C-H type hydrogen atoms were attached at their geometrically expected positions and refined as riding on heavier atoms with the usual constraints. The N-H hydrogen atoms were found in the differential Fourier electron density map and were refined without constraints.
Crystallographic data for the analysis have been deposited with the Cambridge Crystallographic data Centre, CCDC reference number is 1471508. Copies of this information may be obtained free of charge from CCDC, 12 Union Road, Cambridge, CB21EZ, UK (Fax: +44-1223-336033; E-mail: deposit@ccdc.cam.ac.uk or www: http://www.ccdc.cam.ac.uk).

Cell Culture and Cell Viability Assay
All chemicals, if not stated otherwise, were obtained from Sigma-Aldrich (St. Louis, MO, USA). The MCF-7 cell line was purchased from Cell Lines Services (Eppelheim, Germany), the HeLa and HCT-116 cell lines were obtained from the Department of Microbiology, Tumor and Cell Biology, Karolinska Institute (Stockholm, Sweden). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum, 2 mM glutamine, 100 units/mL penicillin, and 100 µg/mL streptomycin. Cultures were maintained in a humidified atmosphere with 5% CO 2 at 37˝C in an incubator (HeraCell, Heraeus, Langenselbold, Germany).
Cell viability was determined using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl -tetrazoliumbromide) assay. Stock solutions of the studied compounds were prepared in 100% DMSO. Working solutions were prepared by diluting the stock solutions with DMEM medium, the final concentration of DMSO did not exceed 0.5% in the treated samples. Cells were seeded in 96-well plates at a density of 5ˆ10 3 cells/well and treated for 72 h with the examined compounds in the concentration range 1-100 µM (1, 10, 25, 50 and 100 µM). Following treatment, MTT (0.5 mg/mL) was added to the medium and cells were further incubated for 2 h at 37˝C. Cells were lysed with DMSO and the absorbance of the formazan solution was measured at 550 nm with a plate reader (1420 multilabel counter, Victor, Jügesheim, Germany). The optical density of the formazan solution was measured at 550 nm with a plate reader (Victor 1420 multilabel counter). The experiment was performed in triplicate. Values are expressed as the mean˘SD of at least three independent experiments. 4-chloro-2-(R 2 -methylthio)-5-R 1 -N-(5-aryl-1,2,4-triazin-3-yl)benzenesulfonamides. The results pointed to 37 as a hit compound with a good biological response against HCT-116 (IC 50 = 36 µM) and HeLa (IC 50 = 34 µM) cell lines with apoptosis-inducing activity, and satisfactory metabolic stability.