Synthesis and Anticancer Properties of New 3-Methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones

Quinolinones have been known for a long time as broad-spectrum synthetic antibiotics. More recently, the anticancer potential of this group of compounds has been investigated. Following this direction, we obtained a small library of 3-methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones with various substituents at positions 1, 2, 6, and 7 of the quinolinone ring system. The cytotoxic activity of the synthesized analogs was tested in the MTT assay on two cancer cell lines in order to determine the structure–activity relationship. All compounds produced high cytotoxic effects in MCF-7, and even higher in HL-60 cells. 2-Ethyl-3-methylidene-1-phenylsulfonyl-2,3-dihydroquinolin-4(1H)-one, which was over 5-fold more cytotoxic for HL-60 than for normal HUVEC cells, was selected for further tests. This analog was shown to inhibit proliferation and induce DNA damage and apoptosis in HL-60 cells.


Introduction
Natural products, defined as chemical substances produced by living organisms, were generated for various purposes, the main of which was to develop and maintain different forms of life. According to the published data, 73% of currently available medications are either natural compounds isolated from plants, bacteria, or marine invertebrates or their chemically modified analogs [1]. The remaining 27% are totally synthetic drugs found by random screening.
It is well known that the presence, position, and character of substituents play an important role in defining the biological activity of quinolinone molecules [11]. Substitution on nitrogen atom at position 1 is essential for overall potency, as well as the presence of a carbonyl group at position 4. The nature of substituents at positions 5-8 can affect the configuration of quinolinone molecules and influence anticancer activity. The substituent at position 3 should be co-planar with quinolinone moiety and the position 2 substituent should not disturb this co-planarity. One group of quinolinones, in which the condition of co-planarity is evidently fulfilled, are 3-alkylidenequinolin-4-ones 3. Furthermore, an It is well known that the presence, position, and character of substituents play important role in defining the biological activity of quinolinone molecules [11]. Subst tion on nitrogen atom at position 1 is essential for overall potency, as well as the prese of a carbonyl group at position 4. The nature of substituents at positions 5-8 can affect configuration of quinolinone molecules and influence anticancer activity. The substitu at position 3 should be co-planar with quinolinone moiety and the position 2 substitu should not disturb this co-planarity. One group of quinolinones, in which the condit of co-planarity is evidently fulfilled, are 3-alkylidenequinolin-4-ones 3. Furthermore exo-cyclic alkylidene moiety conjugated with a carbonyl group present in analogs 3 pharmacophoric unit known to be responsible for the cytotoxic properties of many natu products, including sesquiterpene lactones [12][13][14]. The cytotoxicity of compounds w an α,β-unsaturated carbonyl group is the result of their ability to alkylate cellular th in enzymes, other functional proteins, and in free intracellular glutathione, which disru some major processes in the cell, leading to inhibition of proliferation and the induct of apoptosis [15].
Encouraged by these results, herein, we describe the synthesis and cytotoxic activ of a series of 3-methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones 5 ( Figure 1) wh were obtained in order to establish how various substituents in positions 1, 2, 6, and the quinolinone molecule can influence the in vitro cytotoxicity against promyelocytic l kemia HL-60 and breast cancer adenocarcinoma MCF-7. Normal human umbilical v endothelial cells HUVEC were used for comparison. The most promising analog in ter of cytotoxicity and selectivity, 2-ethyl-3-methylidene-1-(phenylsulfonyl)-2,3-dihyd quinolin-4(1H)-one (5a), was evaluated for its ability to inhibit cell proliferation, ind DNA damage and apoptosis, and influence the mRNA level of an ABCB1 transporter t may be responsible for multidrug resistance in cancer cells.
Encouraged by these results, herein, we describe the synthesis and cytotoxic activity of a series of 3-methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones 5 ( Figure 1) which were obtained in order to establish how various substituents in positions 1, 2, 6, and 7 of the quinolinone molecule can influence the in vitro cytotoxicity against promyelocytic leukemia HL-60 and breast cancer adenocarcinoma MCF-7. Normal human umbilical vein endothelial cells HUVEC were used for comparison. The most promising analog in terms of cytotoxicity and selectivity, 2-ethyl-3-methylidene-1-(phenylsulfonyl)-2,3-dihydroquinolin-4(1H)-one (5a), was evaluated for its ability to inhibit cell proliferation, induce DNA damage and apoptosis, and influence the mRNA level of an ABCB1 transporter that may be responsible for multidrug resistance in cancer cells.

Chemistry
The target compounds were synthesized by applying a modified methodology originally reported for the synthesis of 3-methylidene-1-tosyl-2,3-dihydroquinolin-4(1H)-ones [16]. Starting materials, methyl 2-(arylsulfonylamino)benzoates 7a-e, were synthesized by Nsulfonylation of the corresponding methyl 2-aminobenzoates 6 with arylsulfonyl chlorides. After crystallization with methanol, we obtained pure products 7a-e in good-to-moderate yields (58-85%) (Scheme 1). Next, we performed acylation of diethyl methylphosphonate (8) with obtained benzoates 7a-e in the presence of three equivalents of LDA and expected diethyl 2-oxo-2-[(2-arylsulfonylamino)phenyl]ethylphosphonate 9a-e were formed in good yields (62-87%). Condensation of 9a-e with selected alkyl and aryl aldehydes followed by the spontaneous intramolecular aza-Michael addition delivered 2-substituted 3-diethoxyphosphoryl-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones 10a-t. The piperidine acetate was used as a catalyst and the reaction mixtures were stirred for 24-48 h at room temperature. Progress of the reaction was monitored by 31 P-NMR. The crude products were purified by column chromatography. 1 H, 13 C, and 31 P NMR spectra showed that obtained compounds were formed as mixtures of transand cis-2,3-dihydroquinolin-4(1H)-ones transand cis-10a-t and 3-diethoxyphosphoryl-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ols 10a-t, enol-10a-t, with the enol form strongly predominating (70-99%). Ketones 10 were formed as single trans-isomers or as mixtures of cisand trans-isomers. Careful analysis of the 1 H NMR spectra of the obtained compounds revealed characteristic singlets or doublets with chemical shifts in the range of 10.57-10.90 ppm and 4 J H-p = 0.9-1.1 Hz, which were attributed to the protons of the hydroxyl group of the enol form. In case of ketones, configurational assignments were made on the bases of characteristic 3 J H2-H3 coupling constants which were in the range of 1.1-1.5 Hz for trans isomers and 4.5-5.8 Hz for cis isomers. Thorough NMR analysis of 2-substituted 3-diethoxyphosphoryl-2,3-dihydroquinolinones was presented in our previous paper [16], and the results obtained in this work are with full agreement with that analysis. Table 1 shows transand cis-10a-t and enol-10a-t ratios, as well as yields of the synthesized products.
form. In case of ketones, configurational assignments were made on the bases of characteristic 3 JH2-H3 coupling constants which were in the range of 1.1-1.5 Hz for trans isomers and 4.5-5.8 Hz for cis isomers. Thorough NMR analysis of 2-substituted 3-diethoxyphosphoryl-2,3-dihydroquinolinones was presented in our previous paper [16], and the results obtained in this work are with full agreement with that analysis. Table 1 shows trans-and cis-10a-t and enol-10a-t ratios, as well as yields of the synthesized products.

In Vitro Cytotoxic Activity
The cytotoxicity of a series of 3-methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)ones 5a-t was tested against HL-60 and MCF-7 cell lines using the MTT assay. Carboplatin was used as a reference compound. The obtained results are summarized in Table 2. After 48 h incubation, all new analogs showed cytotoxic activity in low µM range against both cancer cell lines. The tested compounds were more cytotoxic for HL-60 than for MCF-7 cells. Analysis of the structure activity relationship revealed that quinolin-4(1H)-ones with alkyl substituent in position 2 were more potent than those bearing an aryl substituent in that position. In general, with one exception (5i), the highest activity was observed for analogs containing i-propyl substituent in position 2. In HL-60 cells, the most active analogs 5b, 5f, and 5r had IC 50 values below 0.3 µM. Compounds 5m-p, with chlorine at position 6, were less but still very cytotoxic. The substituent on the nitrogen atom did not significantly affect the activity of the tested compounds. Table 2. In vitro cytotoxic activity of 3-methylidene-1-sulfonyl-2,3-dihydroquinolin-4(1H)-ones 5a-t tested on two cancer (MCF-7, HL-60) and one normal (HUVEC) cell lines.  1 Compound concentration required to inhibit metabolic activity by 50%. Values are expressed as mean ± SEM from concentration-response curves of at least three experiments using a nonlinear estimation (quasi-Newton algorithm) method.
Compound 5a was chosen for a more detailed evaluation of its anticancer potential in the HL-60 cell line. The MTT assay was performed for 5a after 24 h incubation and the IC50 value was 0.91 ± 0.03 µ M (Figure 2c). Based on this data, three concentrations (0.9, 1.35 and 1.8 µ M) were chosen for further experiments. In order to evaluate the influence of the new analogs on normal, non-cancerous cells, the selected compounds were tested on the HUVEC cell line. In most cases, the differences in IC 50 values between normal and cancer cells were not very significant. Analog 5a (Figure 2a) showed the highest selectivity and was over 5-fold more cytotoxic for HL-60 than for HUVEC cells (Figure 2b). For comparison, the HUVEC/HL-60 IC 50 ratio for carboplatin was 1.41. Compound 5a was chosen for a more detailed evaluation of its anticancer pot in the HL-60 cell line. The MTT assay was performed for 5a after 24 h incubation an IC50 value was 0.91 ± 0.03 µ M (Figure 2c). Based on this data, three concentration 1.35 and 1.8 µ M) were chosen for further experiments.  Compound 5a was chosen for a more detailed evaluation of its anticancer potential in the HL-60 cell line. The MTT assay was performed for 5a after 24 h incubation and the IC 50 value was 0.91 ± 0.03 µM (Figure 2c). Based on this data, three concentrations (0.9, 1.35 and 1.8 µM) were chosen for further experiments.

Inhibition of Cell Proliferation, Generation of DNA Damage, and Induction of Apoptosis by 5a
To evaluate the anticancer potential of 5a, the ability of this compound to inhibit cell proliferation, induce apoptotic cell death, and generate DNA damage was examined by flow cytometry using the 'Apoptosis, DNA Damage, and Cell Proliferation Kit' (BD Bioscience). After 24 h treatment with 5a in increasing concentrations, HL-60 cells were exposed to bromodeoxyuridine (BrdU) for additional 8 h. Then, the cells were simultaneously stained with fluorochrome-labeled anti-BrdU, anti-cleaved PARP (Asp214) and anti-H2AX (pS139). The representative results of multiparameter flow cytometry analysis are shown in Figure 3. To evaluate the anticancer potential of 5a, the ability of this compound to inhibit proliferation, induce apoptotic cell death, and generate DNA damage was examined flow cytometry using the 'Apoptosis, DNA Damage, and Cell Proliferation Kit' (BD B science). After 24 h treatment with 5a in increasing concentrations, HL-60 cells were posed to bromodeoxyuridine (BrdU) for additional 8 h. Then, the cells were simulta ously stained with fluorochrome-labeled anti-BrdU, anti-cleaved PARP (Asp214) a anti-H2AX (pS139). The representative results of multiparameter flow cytometry analy are shown in Figure 3.  BrdU is an analog of thymidine, and its incorporation into DNA was used as an in of cell proliferation. 5a dose-dependently inhibited cell proliferation. At 1.8 µM con tration, this analog caused a 2.2-fold decrease in BrdU incorporation (Figure 4a).
To check whether the cytotoxic effect of 5a was associated with the induction of ap tosis, the pro-apoptotic activity of this compound was investigated. Apoptosis regula is triggered by the activation of caspases [17]. Caspase-3, an effector caspase, cleav number of vital proteins, leading to cell death. The activity of caspase-3 was assessed flow cytometry using fluorochrome-labeled antibodies recognizing 89 kDa-cleaved PA fragments released from PARP by caspase-3 in the executive stage of apoptosis. Incu tion of HL-60 cells with 10a (at 1.35 and 1.8 µ M) for 24 h led to a significant increase in BrdU is an analog of thymidine, and its incorporation into DNA was used as an index of cell proliferation. 5a dose-dependently inhibited cell proliferation. At 1.8 µM concentration, this analog caused a 2.2-fold decrease in BrdU incorporation (Figure 4a). number of apoptotic cells, whose population raised from 1.4% for control to 20.6% and 26.4%, respectively (Figure 4b).
It is well documented that apoptosis induced by many anticancer drugs may be a consequence of DNA damage [18]. Chemical genotoxins that target DNA can inhibit DNA replication, which leads to the collapse of replication forks and the formation of DNA double strand breaks (DSBs). The generation of DNA damage was evaluated using anti-H2AX (pS139) antibodies directed against the phosphorylated form of human H2AX protein at the pS139 residue that is considered to be a biomarker for DNA DSBs. The treatment of HL-60 cells with 5a increased the levels of phosphorylated H2AX by 4.7-and 7.8fold (to 1.35 µ M and 1.9 µ M, respectively), indicating the dose-dependent genotoxic effect of the tested compound (Figure 4c). Pro-apoptotic activity of 5a was also confirmed using FITC-Annexin V and PI double-staining, based on the loss of plasma membrane asymmetry (phosphatidylserine externalization), which is a marker of the earlier stages of apoptosis. Flow cytometry analysis showed that 5a (1.8 µM) increased early and late apoptotic cell fractions by up to 13.9% and 13.7% of the cell population, respectively ( Figure 5). To check whether the cytotoxic effect of 5a was associated with the induction of apoptosis, the pro-apoptotic activity of this compound was investigated. Apoptosis regulation is triggered by the activation of caspases [17]. Caspase-3, an effector caspase, cleaves a number of vital proteins, leading to cell death. The activity of caspase-3 was assessed by flow cytometry using fluorochrome-labeled antibodies recognizing 89 kDa-cleaved PARP fragments released from PARP by caspase-3 in the executive stage of apoptosis. Incubation of HL-60 cells with 10a (at 1.35 and 1.8 µM) for 24 h led to a significant increase in the number of apoptotic cells, whose population raised from 1.4% for control to 20.6% and 26.4%, respectively (Figure 4b).
It is well documented that apoptosis induced by many anticancer drugs may be a consequence of DNA damage [18]. Chemical genotoxins that target DNA can inhibit DNA replication, which leads to the collapse of replication forks and the formation of DNA double strand breaks (DSBs). The generation of DNA damage was evaluated using anti-H2AX (pS139) antibodies directed against the phosphorylated form of human H2AX protein at the pS139 residue that is considered to be a biomarker for DNA DSBs. The treatment of HL-60 cells with 5a increased the levels of phosphorylated H2AX by 4.7-and 7.8-fold (to 1.35 µM and 1.9 µM, respectively), indicating the dose-dependent genotoxic effect of the tested compound (Figure 4c).
Pro-apoptotic activity of 5a was also confirmed using FITC-Annexin V and PI doublestaining, based on the loss of plasma membrane asymmetry (phosphatidylserine externalization), which is a marker of the earlier stages of apoptosis. Flow cytometry analysis showed that 5a (1.8 µM) increased early and late apoptotic cell fractions by up to 13.9% and 13.7% of the cell population, respectively ( Figure 5).

Influence of 5a on ABCB1 Gene Expression
Overexpression of ATP-binding cassette (ABC) transporters causing multidrug resistance (MDR) in cancer cells is one of the major problems associated with poor therapeutic outcome of anticancer drugs [19]. ABC transporters reduce the cellular uptake of drugs into cancer cells, defending them from medical interventions. In leukemia cells ABCB1 (P-glycoprotein), overexpression has been observed [20,21]. In our previous study [16], we demonstrated that some selected 3-methylidenequinolin-4-ones significantly de-

Influence of 5a on ABCB1 Gene Expression
Overexpression of ATP-binding cassette (ABC) transporters causing multidrug resistance (MDR) in cancer cells is one of the major problems associated with poor therapeutic outcome of anticancer drugs [19]. ABC transporters reduce the cellular uptake of drugs into cancer cells, defending them from medical interventions. In leukemia cells ABCB1 (P-glycoprotein), overexpression has been observed [20,21]. In our previous study [16], we demonstrated that some selected 3-methylidenequinolin-4-ones significantly decreased the expression of ABCB1 gene in MCF-7 cells. We observed a 2-fold down regulation of this gene as compared with control, indicating that compounds with such a skeleton might have potential as ABCB1 transporter inhibitors, able to prevent MDR in MCF-7 cells. It seemed interesting to determine if 3-methylidenequinolin-4-one 5a would also down-regulate ABCB1 mRNA levels in leukemia HL-60 cells. Here, we tested the ABCB1 gene expression of 5a at two concentrations, 0.9 and 1.8 µM. Compound 5a decreased the ABCB1 mRNA level at a higher concentration ( Figure 6).

General
Reagents and starting materials were purchased from commercial vendors and used without further purification. All organic solvents were dried over appropriate drying agents and distilled prior to use. Standard syringe techniques were used for transferring dry solvents. NMR spectra were recorded on a Bruker UltraShield 700 instrument, running at 700 MHz for 1 H, 176 MHz for 13 C, and 283 MHz for 31 P. Chemical shifts (δ) are reported in ppm relative to residual solvent signals (CDCl3: 7.26 ppm for 1 H, 77.16 ppm for 13 C NMR). 31 P NMR spectra were recorded using broadband proton decoupling. Melting points were determined in open capillaries and are uncorrected. Column chromatography was performed on Aldrich ® silica gel 60 (230-400 mesh). Thin-layer chromatography was performed with precoated TLC sheets of silica gel 60 F254 (Aldrich ® ) and visualized by ultraviolet irradiation.
General procedures and characterization data for compounds 9a-e and 10a-t as well as 1 H-and 13 C NMR spectra of compounds 5a-t are given in Supplementary Materials.

General
Reagents and starting materials were purchased from commercial vendors and used without further purification. All organic solvents were dried over appropriate drying agents and distilled prior to use. Standard syringe techniques were used for transferring dry solvents. NMR spectra were recorded on a Bruker UltraShield 700 instrument, running at 700 MHz for 1 H, 176 MHz for 13 C, and 283 MHz for 31 P. Chemical shifts (δ) are reported in ppm relative to residual solvent signals (CDCl 3 : 7.26 ppm for 1 H, 77.16 ppm for 13 C NMR). 31 P NMR spectra were recorded using broadband proton decoupling. Melting points were determined in open capillaries and are uncorrected. Column chromatography was performed on Aldrich ® silica gel 60 (230-400 mesh). Thin-layer chromatography was performed with precoated TLC sheets of silica gel 60 F254 (Aldrich ® ) and visualized by ultraviolet irradiation.
General procedures and characterization data for compounds 9a-e and 10a-t as well as 1 H-and 13 C NMR spectra of compounds 5a-t are given in Supplementary Materials.
The tested compounds were dissolved in sterile dimethyl sulfoxide (DMSO) and further diluted with culture medium. The final concentration of DMSO in cell cultures was less than 0.1% v/v. Controls without and with 0.1% DMSO were performed in each experiment. At the used concentration, DMSO had no effect on the observed parameters.

In Vitro Cytotoxicity Assay (MTT)
The MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, which measures the activity of cellular dehydrogenases, was performed according to the Mosmann method [22]. Exponentially growing cells were seeded into 24-well plates at a density of 8 × 10 4 cells/mL and left to grow for 24 h. After 24 h or 48 h of incubation with various concentrations of the tested compounds (dissolved in DMSO and diluted with complete culture medium), cells were incubated with 0.1 mL of MTT solution (5 mg/mL in phosphate buffered saline) for 1.5 h. The metabolically active cells reduced MTT to blue formazan crystals. The plates were centrifuged, and the supernatant was discarded. DMSO (1 mL) was added to each well to dissolve the crystals, whose absorbance was measured at 560 nm using FlexStation 3 Multi-Mode Microplate Reader (Molecular Devices, LLC, San Jose, CA, USA). The untreated cells were used as control. The IC 50 values were calculated from the concentration-response curves. The data were expressed as mean ± SEM of three independent experiments.

Statistical Analysis
All statistical calculations were performed using Prism 5.0 software (GraphPad Software Inc., San Diego, CA, USA). Results from three independent experiments performed in triplicate were expressed as mean ± SEM. Statistical significance was assessed using Student's t-test (for comparisons of two treatment groups) or one-way ANOVA followed by a post-hoc multiple comparison Student-Newman-Keuls test (for comparisons of three or more groups). p-values < 0.05 were considered statistically significant.

Conclusions
Quinolinones have been known for years as efficient, broad spectrum antibiotics. In the last decade, yet another important activity of these compounds, their ability to inhibit cancer cell proliferation, has been investigated. In order to obtain a wide spectrum of diversely substituted quinolinone derivatives, the development of synthetic methods leading to such compounds has been pursued. Here, we used the well-recognized Horner-Wadsworth-Emmons (HWE) olefination in order to introduce the exo-methylidene group onto the heterocyclic ring of dihydroquinolin-4(1H)-ones. The α,β-unsaturated carbonyl function is a structural element responsible for interaction with various cellular proteins, which disturbs many cellular processes.
The aim of this study was to establish the structure-activity relationship of the new series of dihydroquinolin-4(1H)-ones. Using the routine MTT assay, the cytotoxicity of the analogs was assessed on two cancer cell types: an adherent breast cancer MCF-7 and a non-adherent leukemia HL-60. All new analogs were more cytotoxic for HL-60 than for MCF-7 cells. Analysis of the structure-activity relationship revealed that introducing any groups into the aromatic ring (R 1 and R 2 ) was not advantageous for activity. The substituent on the nitrogen atom (R 3 ) did not significantly affect the activity of the tested compounds. The most important turned out to be position 2 (R 4 ). The quinolin-4(1H)-ones with an alkyl substituent in that position were more potent than those bearing an aryl group. In general, iso-propyl as R 4 generated more cytotoxic analogs than ethyl.
The ideal anticancer drug candidates should be highly cytotoxic for cancer cells but much less for normal, healthy cells. The selectivity of the new analogs, expressed as the IC 50 ratio of HUVEC/HL60 cells, was calculated for the selected analogs, and values between 1 and 5.5 were obtained. These values showed that very high cytotoxicity against cancer cells did not go hand-in-hand with high selectivity. However, even approved drugs used for cancer treatment are in most cases strongly cytotoxic for normal cells. The difference between healthy and cancer cells is that the latter proliferate much faster and therefore are more susceptible to the action of drugs. The analog 5a, 2-ethyl-3-methylidene-1-phenylsulfonyl-2,3-dihydroquinolin-4(1H)-one, which was 5-fold more cytotoxic for HL-60 than for HUVEC cells, was further evaluated in terms of its mode of action. This compound inhibited proliferation and induced apoptosis in HL-60 cells, which could be attributed to its ability to induce DNA damage. Additionally, the 5a down-regulated ABCB1 mRNA level decreased the risk of MDR development. Based on these findings, the described structurally diversified dihydroquinolin-4(1H)-ones can serve as lead structures and can be further optimized with the purpose of finding new possibilities for cancer treatment.