Novel coumarin-6-sulfonamides as apoptotic anti-proliferative agents: synthesis, in vitro biological evaluation, and QSAR studies

Abstract Herein, we report the synthesis of different novel sets of coumarin-6-sulfonamide derivatives bearing different functionalities (4a, b, 8a–d, 11a–d, 13a, b, and 15a–c), and in vitro evaluation of their growth inhibitory activity towards the proliferation of three cancer cell lines; HepG2 (hepatocellular carcinoma), MCF-7 (breast cancer), and Caco-2 (colon cancer). HepG2 cells were the most sensitive cells to the influence of the target coumarins. Compounds 13a and 15a emerged as the most active members against HepG2 cells (IC50 = 3.48 ± 0.28 and 5.03 ± 0.39 µM, respectively). Compounds 13a and 15a were able to induce apoptosis in HepG2 cells, as assured by the upregulation of the Bax and downregulation of the Bcl-2, besides boosting caspase-3 levels. Besides, compound 13a induced a significant increase in the percentage of cells at Pre-G1 by 6.4-folds, with concurrent significant arrest in the G2-M phase by 5.4-folds compared to control. Also, 13a displayed significant increase in the percentage of annexin V-FITC positive apoptotic cells from 1.75–13.76%. Moreover, QSAR models were established to explore the structural requirements controlling the anti-proliferative activities.


Introduction
Cancer is a very complex, widespread and lethal disease that affects approximately 14 million people every year. In 2018, the American Cancer Society predict that the new diagnosed cancer cases will reach to 1,735,350 and cancer death 595,690 cancer deaths in the United States 1 . Accordingly, the discovery of new anti-cancer agents with promising bioactivity and high therapeutic index is still an urgent need and a major challenge for researchers.
It is well established that natural and synthetic coumarin derivatives have attracted a great deal of interest due to their variety of biological and pharmacological properties, such as their antiinflammatory 2 , antibacterial 3 , antiviral 4 , and anti-cancer activities 5 . In the current medical era, much attention has been extensively paid to modify and update coumarin-based drug leads from the point of view of drug design and medicinal chemistry to fulfill more effective and safe anti-cancer agents [6][7][8] .
On the other hand, sulfonamide derivatives constitute another important class of organic compounds that displayed interesting biological activities including anti-carbonic anhydrase and anticancer activities 32,33 . It is noteworthy to mention that there is a continuing interest in the synthesis of coumarin sulfonamide derivatives as potent anti-cancer agents [34][35][36][37] . SLC-0111, Figure 1, an ureido benzenesulfonamide derivative, is currently in phase I/II clinical trials as anti-cancer drug. SLC-0111 II is characterized by its selectivity towards inhibition of the transmembrane isoforms hCA IX/XII (over the cytosolic isoforms hCA I/II). SLC-0111 was also able to block human breast cancer invasion, delay tumor growth and diminish the cancer stem cell population in vivo [38][39][40] . KCN1, Figure 1, a novel synthetic sulfonamide, is a HIF pathway inhibitor with in vitro and in vivo anti-pancreatic cancer activities and preclinical pharmacology. KCN1 specifically inhibited HIF reporter gene activity in several glioma cell lines at the nanomolar level. Also, KCN1 effectively inhibited the growth of subcutaneous malignant glioma tumor xenografts with low side effects on the host 41,42 . Moreover, Indisulam (N-(-3-chloro-7-indolyl)-1,4-benzenedisulfonamide), Figure 1, is an orally active sulfonamide anti-tumor agent that possesses anti-cancer properties through down-regulation of various cell-cycle checkpoint molecules, thereby blocking the phosphorylation of retinoblastoma protein and inducing p53 and p21. Pre-clinical and clinical studies have established synergy of Indisulam with nucleoside analogs as well as topoisomerase inhibitors [43][44][45][46][47] .
Based on the aforementioned findings, herein we report the synthesis of novel sets of coumarin sulfonamide derivatives (4a, b, 8a-d, 11a-d, 13a, b and 15a-c) and their in vitro growth inhibitory activity towards the proliferation of three cancer cell lines; HepG2 (hepatocellular carcinoma), MCF-7 (breast cancer), and Caco-2 (colon cancer). Additionally, the synthesized coumarin sulfonamides were further examined regarding their potential apoptotic induction and their effects on cell cycle progression in the Hep-G2 cancer cells to acquire a perception for the mechanism of their anti-cancer activity.

Chemistry
Melting points were determined on Electrothermal IA 9000 apparatus and were uncorrected. Elemental microanalyses were performed on Elementar, Vario EL, at the Micro-analytical Laboratory, National Research Centre, Dokki, Cairo. The 1 H NMR and 13 C NMR spectra were recorded with a BrukerAvance 400 MHz spectrometer at Turku University, Finland and JEOL ACA 500 NMR spectrometer, at the National Research Centre, Dokki, Cairo, Egypt. The mass spectra were performed on Mass Spectrometer Finnigan MAT SSQ-7000 and GCMS-QP 1000EX Shimadzu Gas Chromatography MS Spectrometer at Faculty of Science, Cairo University, Egypt. The reactions were followed by TLC (silica gel, aluminum sheets 60 F254, Merck) using chloroform/methanol (9.5:0.5 v/v) as eluent.
Light yellow crystals, mp 253-254 C, yield (34%). 1  General procedures for preparation of target compounds (11a-d and 13a, b) A mixture of thiosemicarbazone 9 (10 mmol) and the appropriate a-halocarbonyl compounds 10a-d, phenacyl bromide 12a or coumarin-3-acetylbromide 12 b (10 mmol) in dioxane (25 ml) containing catalytic amount of triethylamine was heated under reflux for 8 h and then cooled. The solution was poured onto water-ice and concentrated hydrochloric acid. The solid produced was collected by filtration and crystallized from ethanol to furnish compounds 11a-d and 13a, b, respectively.

General procedures for synthesis of thiazolidinones (15a-c)
In 50 ml round-bottom flask, thiosemicarbazone 9 (10 mmol) was dissolved in 15 ml acetic acid, followed by addition of anhydrous sodium acetate (30 mmol) under magnetic stirring and warming. After 20 min, bromoacetic acid 14a, 2-bromopropanoic acid 14b or maleic anhydride 14c (15 mmol) was added and the reaction mixture was maintained under reflux for 8 h. After cooling, the precipitate was filtered, washed with water, dried, and recrystallized from ethanol to give compounds 15a-c, respectively.

Biological evaluation
In vitro anti-proliferative activity HepG2 liver cancer, MCF-7 breast cancer and Caco-2 cancer cell lines were obtained from the National Cancer Institute (Cairo, Egypt). Caco-2 cells were grown in DMEM while HepG2 and MCF-7 were grown in RPMI-1640. Media were supplemented with 10% heat-inactivated FBS, 50 units/mL of penicillin and 50 g/mL of streptomycin and maintained at 37 C in a humidified atmosphere containing 5% CO 2 . The cells were maintained as a "monolayer culture" by serial subculturing. Cytotoxicity was determined using the sulforhodamine B (SRB) method as previously described by Skehan et al. 50 Exponentially growing cells were collected using 0.25% trypsin-EDTA and seeded in 96-well plates at 1000-2000 cells/well in supplemented DMEM medium. After 24 h, cells were incubated for 72 h with various concentrations of the tested compounds as well as doxorubicin as the reference compound. Following 72 h of treatment, the cells were fixed with 10% trichloroacetic acid for 1 h at 4 C. Wells were stained for 10 min at room temperature with 0.4% SRB dissolved in 1% acetic acid. The plates were air dried for 24 h, and the dye was solubilized with Tris-HCl for 5 min on a shaker at 1600 rpm. The optical density (OD) of each well was measured spectrophotometrically at 564 nm with an ELISA microplate reader (ChroMate-4300, FL, USA). The IC 50 values were calculated according to the equation for Boltzmann sigmoidal concentrationeresponse curve using the nonlinear regression models (GraphPad, Prism Version 5). The results reported are means of at least three separate experiments. Significant differences were analyzed by one-way ANOVA wherein the differences were considered to be significant at p < .05.

ELISA immunoassay
The levels of the apoptotic markers (Bax, caspase-3) and the antiapoptotic marker (Bcl-2) were examined using ELISA colorimetric kits per the manufacturer's protocol and referring to reported instructions 51,52 .

Cell cycle analysis
The HepG2 cells were treated with compound 13a for 24 h (at IC 50 concentration), then cells were washed twice with ice-cold phosphate-buffered saline (PBS). Subsequently, the treated cells were collected by centrifugation, fixed in ice-cold 70% (v/v) ethanol, washed with PBS, re-suspended with 0.1 mg/mL RNase, stained with 40 mg/mL PI, and analyzed by flow cytometry using FACS Calibur (Becton Dickinson, BD). The cell cycle distributions were calculated using CellQuest software (Becton Dickinson) 53 .

Annexin V-FITC apoptosis assay
Phosphatidylserine externalization was measured using Annexin V-FITC/PI apoptosis detection kit (BD Biosciences, San Jose, CA) according to the manufacturer's instructions, as reported earlier 53 . HepG2 cells were treated with 13a at defined concentrations for 24 h.

Chemistry
The synthetic routes employed to prepare the new target derivatives are depicted in Schemes 1-4. The key intermediate 2-oxo-2Hchromene-6-sulfonyl chloride 2 was obtained from a reaction of coumarin 1 with chlorosulfonic acid, which subsequently reacted with sulfanilamide 3a and sulfapyridine 3 b in refluxing ethanol to furnish the corresponding target compounds 4a, b, respectively (Scheme 1).
Refluxing of compound 6 with thiosemacarbazide in ethanol and catalytic amount of acetic acid yielded the key intermediate 9, which utilized for preparation of diverse derivatives. Treatment of intermediate 9 with the appropriate a-halocarbonyl compounds 10a-d in refluxing dioxane furnished the corresponding thiazole derivatives 11a-d. Additionally, refluxing of 9 with phenacyl bromide 12a and coumarin-3-acetylbromide 12b in dioxane afforded the corresponding thiazoles 13a, b, respectively (Scheme 3). Moreover, treatment of intermediate 9 with each of bromoacetic acid 14a, 2bromopropanoic acid 14b and maleic anhydride 14c in acetic acid in the presence of anhydrous sodium acetate afforded the target thiazolidinone derivatives 15a-c, respectively (Scheme 4).
Postulated structures of the newly synthesized coumarin sulfonamides were in full agreement with their spectral and elemental analyses data (Supplementary material).

Biological evaluation
In vitro anti-proliferative activity All the newly synthesized target coumarin sulfonamides were evaluated for their in vitro anti-proliferative activity against three human tumor cancer cell lines, HepG2 hepatocellular carcinoma, MCF-7 breast cancer, and Caco-2 colon cancer, using SRB colorimetric assay as described by Skehan et al. 50 Doxorubicin was involved in the experiments as a reference cytotoxic compound. The results were expressed as growth inhibitory concentration (IC 50 ) values which represent the compound concentrations required to produce a 50% inhibition of cell growth after 72 h of incubation compared to untreated controls. The results were summarized in Table 1.
From the displayed results, it was obvious that several of the prepared coumarin sulfonamides possess excellent to modest growth inhibitory activity towards the tested cancer cell lines. Examinations of the anti-proliferative activity towards HepG2 cells revealed that it is the most sensitive cell line to the influence of the target coumarin sulfonamide derivatives. Compounds 13a and 15a emerged as the most active coumarins against HepG2 cells through this study with IC 50 values of 3.48 ± 0.28 and 5.03 ± 0.39 mM, respectively. They displayed 1.5-and 1.1-fold increased potency than doxorubicin (IC 50 ¼ 5.43 ± 0.24 mM). Besides, compounds 4b and 15c with IC 50 values of 8.08 ± 0.51 and 7.57 ± 0.66 mM, respectively, displayed excellent activity in comparison to doxorubicin. Also, compounds 8b and 13b (IC 50 ¼ 11.84 ± 1.34 and 11.80 ± 1.16 mM, respectively) exhibited good anti-proliferative activity. Moreover, compounds 6, 8a, 8c, and 15b had moderate activity with IC 50 values ranging from 25.07 ± 2.08 to 29.80 ± 2.21 mM.
Concerning activity against MCF-7 cells, compounds 15a and 15b were the most active members that displayed potent antiproliferative activity with IC 50 values of 10.95 ± 0.96 and 10.62 ± 1.35 mM, respectively, in comparison to the standard drug doxorubicin (IC 50 ¼ 3.18 ± 0.32 mM). Moreover, compounds 8a, 11a, and 15c were moderately active against MCF-7 cells with IC 50 values: 14.30 ± 1.18, 13.86 ± 1.19 and 16.32 ± 1.48 mM, respectively. On the other hand, anti-proliferative activity evaluation in Caco-2 cells showed that compound 8a possessed the best growth inhibitory activity (IC 50 ¼ 8.53 ± 0.72), with twofold decreased activity than doxorubicin (IC 50 ¼ 4.10 ± 1.37 mM). In addition, compound 11a showed good anti-proliferative activity against Caco-2 cancer cell line (IC 50 ¼ 10.12 ± 0.90). Whilst, both compounds 8d and 11d displayed moderate activity against Caco-2 cancer cell line with IC 50 values: 16.02 ± 1.32 and 16.06 ± 1.28, respectively. These results suggest that diverse functionalities could be incorporated into the coumarin sulfonamide scaffold to obtain promising anti-proliferative activity against different types of cancer cells.

In vitro cytotoxicity towards human normal WI-38 cells
The cell growth inhibitory activity of the most potent compounds 13a and 15a was examined towards non-tumorigenic human normal lung fibroblast cell line (WI-38) to investigate the potential safety of the newly synthesized coumarin-6-sulfonamides towards the normal cells. The results were expressed as IC 50 values, and selectivity index was calculated ( Table 2). The tested coumarin-6sulfonamides (13a and 15a) showed non-significant cytotoxic action with IC 50 values of 73.20 ± 3.47 and 55.92 ± 0.39 mM, respectively, with good selectivity index range 21 and 11, thereby providing a high safety profile as anticancer agents.

Apoptosis induction in HepG2 cells
Apoptosis induction in cancer cells represents one of the most successful strategies for the development of cancer therapy [54][55][56] . As displayed above, coumarins 13a and 15a emerged as the most active ones towards HepG2 liver cancer cells. Accordingly, we investigated the ability of compounds 13a and 15a to provoke  apoptosis in HepG2 cells to define the principle mechanism for their anti-proliferative activity.

Effects on mitochondrial apoptosis pathway proteins Bcl-2 and Bax
The Bcl-2 proteins family is mainly responsible for synchronizing the mitochondrial apoptotic pathway, and classified into two major classes: anti-apoptotic proteins such as Bcl-2 protein and the counteracting pro-apoptotic proteins including Bax protein. In this study, we examined the impact of coumarins 13a and 15a on the level of the anti-apoptotic Bcl2 and the level of the proapoptotic Bax ( Table 3). As shown in Table 3, compound 13a induced the protein expression of Bax with 16.5 folds of the control while 14.3 folds were recorded for compound 15a. On the other hand, treatment of HepG2 cells with compounds 13a and 15a significantly reduced the expression levels of the anti-apoptotic protein Bcl-2 by 21.4 and 38.6%, respectively, compared to the control.
Effects on the levels of active caspase-3 (key executor of apoptosis) As a key executioner protease, caspase-3 is activated by upstream initiator caspases as caspase-9. Herein, treatment of HepG2 cells with coumarins 13a and 15a resulted in a significant elevation in the level of active caspase-3 by about 6.6 and 5.7 folds, respectively, compared to control (Table 3).

Cell-cycle analysis
The impact of compound 13a on cell cycle progression was examined in HepG2 cancer cell line after 24 h of treatment (Figure 2). This impact was illustrated by DNA flow cytometric analysis where HepG2 cells were treated with 13a at concentration equals to its IC 50 . Figure 2 showed that exposure of HepG2 cells to compound 13a induced a significant increase in the percentage of cells at Pre-G1 by 6.4-folds, with concurrent significant arrest in the G2-M phase by 5.4-folds compared to control. Alteration of the Pre-G1 phase and arrest of G2-M phase were significant remarks for compound 13a to induce apoptosis in HepG2 cells.

Annexin V-FITC apoptosis assay
The apoptotic effect of coumarin 13a was further assured by Annexin V-FITC/PI (AV/PI) dual staining assay to investigate the occurrence of phosphatidylserine externalization (Figure 3). Flow cytometric analysis revealed that HepG2 cells treated with 13a showed a significant increase in the percent of annexin V-FITC positive apoptotic cells (UR þ LR) from 1.75% to 13.76% which comprises about 7.9 folds compared to control.
2D QSAR study Development of QSAR models. With the aim to assess the structural basis for the anti-proliferative activity of the newly prepared coumarin sulfonamides (4a, b, 6, 8a-d, 9, 11a-d, 13a, b and 15a-c), 2D-QSAR analysis was carried out. This analysis was run by means of the DS 4.0 software (Discovery Studio 4.0, Accelrys, Co. Ltd) 57 . A set of the newly synthesized compounds (4a, b, 6, 8a-d, 9, 11a-d, 13a, b and 15a-c) was used as training set with their measured pIC 50 against HepG2 and Caco-2 for QSAR modeling, compounds 6, 8d, 15a and 15c were chosen as statistical outliers while building model for HepG2 whereas, compounds 6, 9, and 13a were chosen for the Caco-2 model. "Calculate Molecular Properties" module was used for calculating the molecular properties. Genetic function approximation (GFA) protocol was applied in order to choose the best descriptors that characterize the activity. Multiple linear regression (MLR) protocol was then employed to search for optimal QSAR models with the best statistical validation measures and capable of correlating bioactivity variation across the used training set collection. QSAR model was validated employing leave one-out cross-validation by setting the folds to a number much larger than the number of samples, r 2 (squared correlation coefficient value) and r 2 prediction (predictive squared correlation coefficient value), residuals between the predicted and experimental activity of the test set and training set (Table 4).
The two MLR models exhibited good correlation coefficient r 2 of 0.940 and 0.896, r 2 (adj) ¼ 0.910 and 0.865, respectively, r 2 (pred) ¼ 0.878 and 0.789, Least-Squared error ¼ 0.0168 and 0.0173 respectively, where r 2 (adj) is r 2 adjusted for the number of terms in the model; r 2 (pred) is the prediction r 2 , equivalent to q 2 from a leave-1-out cross-validation 58 .    (1) and (2) for HepG2 and Caco-2 suggested that the anti-proliferative activity of the prepared compounds was positively affected by Dipole, which is a 3D electronic descriptors describing the strength and orientation behavior of a molecule in an electrostatic field. Both the magnitude and the components (X, Y, Z) of the dipole moment are calculated (Debyes). It is estimated by utilizing partial atomic charges and atomic coordinates. Partial atomic charges are computed using Gasteiger if not present. Dipole properties have been correlated to long range ligand-receptor recognition and subsequent binding. The anti-proliferative activity towards HepG2 was also affected by the high and negative value of Shadow_YZfrac while for Caco-2 affected by Shadow_Ylength. Shadow indices are a set of geometric descriptors that characterize the shape of the molecules. They are calculated by projecting the model surface on three mutually perpendicular planes: xy, yz, and xz. These descriptors depend not only on conformation, but also on the orientation of the model. In order to calculate them, the models are first rotated to align the principal moments of inertia with the x-, y-, and z-axes. Shadow_YZ is area of the molecular shadow in the yz plane while  Shadow_Ylength is indicator for length of molecule in the y dimension 59 .
Equation (1) also showed that the anti-proliferative activity was also affected by jurs descriptors. Jurs descriptors are a group of geometric descriptors that combine both shape and electronic information which may characterize the molecules 60 . In particular, Jurs_PNSA_3, which contributes inversely to the activity, represents the atomic charge weighted negative surface area and calculated by Sum of the product of solvent-accessible surface area Â partial charge for all negatively charged atoms. This indicates increasing this value in a molecule could increase reuptake inhibition activity because it is negatively correlated with the activity. Equations (1) and (2) are also affected by topological descriptors which are a special class of descriptors that do not rely on a three-dimensional model. All calculations are derived from the two-dimensional topology of the molecule. As CHI_3_C which is connectivity indices while IAC_Mean indicate Graph-Theoretical InfoContent descriptors.

QSAR validation
The established QSAR models (1 and 2) were verified by applying; Leave-one-out (LOO) internal validation (r 2 ¼ 0.940 and 0.896, respectively). Cross-validation was also employed where q 2 , which is equivalent to r 2 (pred), was 0.878 and 0.789, respectively. In addition, validation was employed by measuring the residuals between the experimental and the predicted activities of the training set listed in Table 4. Interestingly, the predicted activities of the QSAR models were found very close to those experimentally observed, Furthermore, to evaluate the predictive ability of the developed models HepG2, compounds 6, 8d, 15a, and 15c were applied as test compounds, where they were not included in model generation, the same was also applied for the Caco-2 model using compounds 6, 9, and 13a (Table 5).

Conclusion
In our endeavor to develop potent anti-cancer agents, different sets of coumarin-6-sulfonamide derivatives (4a, b, 8a-d, 11a-d, 13a, b, and 15a-c) were synthesized. Anti-proliferative activities of the newly synthesized coumarins was invistigated in three human tumor cancer cell lines, namely, HepG2 hepatocellular carcinoma, MCF-7 breast cancer and Caco-2 colon cancer using sulforhodamine B (SRB) colorimetric assay. Compounds 13a and 15a emerged as the most active derivatives towards HepG2 cells (IC 50 ¼ 3.48 ± 0.28 and 5.03 ± 0.39 mM, respectively), with 1.5-and 1.1-fold increased activity than the reference drug, doxorubicin (IC 50 ¼ 6.9 ± 2.05 mM), respectively. Besides, compounds 15a and 15b were the most active members against MCF-7 cells with IC 50 values of 10.95 ± 0.96 and 10.62 ± 1.35 mM, respectively. Also, compound 8a possessed the best growth inhibitory activity against Caco-2 cells (IC 50 ¼ 8.53 ± 0.72), with 2-fold decreased activity than doxorubicin (IC 50 ¼ 4.10 ± 1.37 mM). Compounds 13a and 15a were able to induce apoptosis in HepG2 cells, as assured by the upregulation of the Bax and down-regulation of the Bcl-2, besides boosting caspase-3 levels. Besides, compound 13a induced a significant increase in the percentage of cells at Pre-G1 by 6.4-folds, with concurrent significant arrest in the G2-M phase by 5.4-folds compared to control. Also, 13a displayed significant increase in the percentage of annexin V-FITC positive apoptotic cells from 1.75% to 13.76%. Moreover, QSAR models were established to explore the structural requirements controlling the anti-proliferative activities.