1-Hydroxyanthraquinones Containing Aryl Substituents as Potent and Selective Anticancer Agents

A series of 1,2-, 1,4-disubstituted or 1,2,4-trisubstituted anthraquinone-based compounds was designed, synthesized, characterized and biologically evaluated for anticancer efficacy. 2- or 4-arylated 1-hydroxy-9,10-antraquinones (anthracene-9,10-diones) were prepared by Suzuki–Miyaura cross-coupling reaction of 1-hydroxy-2-bromoanthraquinone, 1-hydroxy-4-iodoanthraquinone or 1-hydroxy-2,4-dibromoanthraquinone with arylboronic acids. The cross-coupling reaction of 2,4-dibromo-9,10-anthraquinone with arylboronic acids provide a convenient approach to 2,4-bis arylated 1-hydroxyanthraquinones with a variety of aryl substituent in the 2 and 4 position. The cytotoxicity of new anthraquinone derivatives was evaluated using the conventional MTT assays. The data revealed that six of the aryl substituted compounds among the entire series 3, 15, 16, 25, 27, 28 were comparable potent with the commercially available reference drug doxorubicin on the human glioblastoma cells SNB-19, prostate cancer DU-145 or breast cancer cells MDA-MB-231 and were relatively safe towards human telomerase (h-TERT)immortalized lung fibroblasts cells. The results suggested that the in vitro antitumor activity of synthesized 2-aryl, 4-aryl- and 2,4-diaryl substituted 1-hydroxyanthraquinones depends on the nature of the substituent within the cyclic backbone. Docking interaction of 2-, 4-substituted and 2,4-disubstituted 1-hydroxyanthraquinones indicates intercalative mode of binding of compounds with DNA topoisomerase. The interaction with the DNA of 4-aryl-13, 15, 16 and 4-(furan-3-yl)-23 1-hydroxyanthraquinones was experimentally confirmed through a change in electroforetic mobility. Further experiments with 1-hydroxy-4-phenyl-anthraquinone 13 demonstrated that the compound induced cell cycle arrest at sub-G1 phase in DU-145 cells in the concentration 1.1 μM, which is probably achieved by inducing apoptosis. 4-Arylsubstituted 1-hydroxyanthraquinones 13 and 16 induced the enhancement of DNA synthesis on SNB19 cell lines.


Introduction
The anthraquinone (anthracene-9,10-dione), a polycyclic aromatic core, is an important structural motif in a large number of organic molecules, is prevalent in nature. Several functionalized anthraquinones are well known for their diverse and profound biological activities. Since the discovery of the synthetic anthraquinone antitumor drug mitoxantrone, that is clinically used for the treatment of a variety of cancers [1], various 9,10-anthraquinone derivatives have been investigated and used as antiviral [2], antibacterial [3], anticancer [4] and anti-inflammatory agents [5]. 2-Aryl and 4-aryl substituted anthraquinones were isolated from the natural sources [6]. From these compounds the natural 4-arylanthraquinones-knipholones -are of interest. Knipholone and isoknipholone have recently been reported to exhibit good antitumoral activities against several cancer cells, some of them comparable to that of etoposide [4,7]. Knipholone and its related derivatives have been reported to exhibit significant activities against the malaria parasite, Plasmodium falciparum [6]. Based on these grounds, the search for efficient and versatile synthetic methodologies leading to variously substituted 9,10-anthraquinones deserves great attention.
In the framework of our studies dealing with the development of convenient routes to functionalization of some plant metabolites or their derivatives [41][42][43][44], we report herein the synthesis of a range of 1-hydroxy substituted anthraquinones containing an aryl substituent in the 2 or 4 (or 2 and 4 simultaneously) position of the anthraquinone core. As a starting compound, we used the 1-hydroxy-4-iodoanthraquinone (1), 1-hydroxy-2-bromoanthraquinone (2) or 1-hydroxy-2,4-dibromoanthraquinones (3) which were obtained from 1-hydroxy-9,10-anthraquinone or 4-amino-1-hydroxy-9,10-anthraquinone by the known procedures [45,46]. The Pd-catalyzed Suzuki-Miyaura cross-coupling reaction of the mentioned compounds with aryl boronic acids was the main route of synthesis. Taking into account the interest to substituted 4-arylanthraquinones as anticancer agents [4,6,7], we evaluated the cytotoxicity of the synthesized compounds toward a panel of cancer cell lines in vitro and also obtained some data about the potential mechanism of action of the new compounds.
Antracenedione drugs are known to exert their cytotoxic effects through interaction with DNA resulting in modification of its structure hence inhibition of its replication.
Anthraquinone mitoxantrone is a potent synthetic anticancer drug which blocks DNA synthesis by inhibiting the function of DNA topoisomerase II. This compound inhibits the activities of both enzyme isoforms: topoisomerase IIα [47,48] and topoisomerase IIβ [1,49]. Several anthraquinone pharmacophores can realize their anticancer activity by affecting other molecular targets, such as proteins.
Purpurin (1,2,4-trihydroxy-9,10-anthraquinone) is a non-competitive inhibitor of adipocyte-derived leucine aminopeptidase (A-LAP) which play a crucial role in angiogenesis [50]. Emodin (1,3, was characterized as a significant inhibitor of cell proliferation, presumably via down regulation of excision repair cross-complementary 1 (ERCC1) and DNA recombinase protein Rad51 [51], but its 2,4-dibromo derivatives exert their anti-proliferative activity at least in part, by inhibition of ATP citrate lyase (ACL), plays a critical role in generating cytosolic acetyl CoA [34]. Emodin and 2-chloroemodin were also considered as potential targets of dioxygenases (ALKBH 2, 3 proteins and FTO) overexpression blockers [52]. There was therefore value in a targeted preparation and investigation of novel hydroxyl-aryl substituted anthraquinones. The hydroxy substituent in the anthraquinones will be necessary for further improve the low druggability of the anthraquinone core.

Cytotoxicity Studies
One of the necessary steps in the study of the biological activity of potential pharmacological agents synthesized as oncolytics is the study of their cytotoxic profile in tumor cell cultures. This allows to evaluate the feasibility of their further research at the next stages of screening.
The cytotoxicity of the synthesized series of 4-substituted 5, 13-17, 22, 23, 2-substituted 25-30 and 2,4-disubstituted 1-hydroxy-9,10-anthraquinones 33-38, 40-45 was evaluated against a panel of seven different human cancer cell lines (glioblastoma cancer cells, human prostate cancer cells, T-cellular human leucosis, breast cancer cells) and also a normal cell line of hTERT-immortalized lung fibroblasts, using conventional MTT assay [55]. Doxorubicin (DOX) is clinically used to treat cancer as drug in world and have a very wide antitumor spectrum. That we use them as positive control compounds. The cytotoxicity was determined by measuring the concentration inhibiting human tumor cell viability by 50% (GI 50 ). The results are presented in Table  1,4-Diaryl substituted 1-hydroxyanthraquinones 37 and 45 shown selective cytotoxicity towards prostate cancer cells LNCAP (GI 50 6.2-7.2 µM). The OMe-and CF 3 -groups in the 4-aryl substituent provided the selectivity for prostate cancer activity.
A remarkable increase in activity and selectivity towards prostate cancer cell line DU-145 was observed for 4-aryl substituted compounds 13, 15, 16, 17 and 2,4-diaryl substituted derivatives 35, 37, 38, 40; all these compounds demonstrated inhibition against this prostate cancer cells in the micromolar concentration which is comparable or higher than that of the drug Doxorubicin. Characteristically, that the 4-phenylsubstituted anthraquinone 13 possess the best activity in DU-145 cell lines (GI 50 1.1 µM).
Studying the comparable effect of the compounds on viability of human cancer lines revealed that compounds with an aryl substituent in the 4 position demonstrate the increase of potency compared with compounds having a furyl substituent in this position (22,23). Additionally, both C-2 aryl and C-4 aryl series were less cytotoxic towards the normal cell line than the 2,4-diarylated 1-hydroxyanthraquinone derivatives 34-38, 45 and also the bromo-aryl substituted compounds 40-42. Characteristically, the methoxy substitution is more favorable than the fluoro-or CF 3 -substitution in the aromatic rings in the C-4 and C-2 arylsubstituted series; these compound we less toxic towards the normal cell line model.

Electrophoretic Mobility
The interaction of compounds 13, 15, 16, 23, 25 27, 28, 35, 40 and 44 with DNA was experimentally confirmed by study of the electrophoretic mobility. For this the gel retardation assay was performed. Doxorubicin in different concentration was used as a positive control. In order to avoid destruction of the possible complex of DNA with compounds, the electrophoresis was performed at low voltage.
The results are presented at Figure 3. It was shown that 4-aryl and 4-(furan-3yl) substituted compounds 13, 15, 16, 23, 2-(4-methoxyphenyl)-27 and also 2,4-(4-dimethoxyphenyl)-35 substituted anthraquinones cause a retardation of plasmid DNA which could indicate the formation of complexes. The results taken into suggest that DNA interaction is a necessary component for mediating aryl substituted 1-hydroxy-9,10-anthraquinone-induced cell death but can not account for the differences in their cytotoxic potential entirely.

Cell Cycle and DNA Synthesis Analysis
Due to the fact that the implementation of the antitumor potential of the currently existing antitumor agents is carried out by acting on various biological targets, at the next stage we performed the cell cycle analysis and DNA synthesis. Cell cycle analysis of 1-hydroxy-4-phenyl-9,10-anthraquinone 13, 1-hydroxy-4-(2,3-dimethoxyphenyl)-9,10anthraquinone 16, 1-hydroxy-4-(furan-3-yl)-9,10-anthraquinone 23 and 1-hydroxy-2-(2,3-dimethoxyphenyl)-9,10-anthraquinone 28 against SNB-19 cells for 24 h produced interesting results. After examining the data for compounds 13 and 16 (Table 3, Figure 4C,D), good number of cells are distributed in S phase, i.e., initiated DNA replication mechanism and also in G0/G1 phase, i.e., initial phase of cell cycle. G1 and G2 are the growth phases in cell cycle analysis, whereas S phase is a synthetic phase wherein DNA replication and DNA synthesis take place. The 2-aryl substituted compound 28 initiated the G0/G1 phase in analogy with doxorubicin and mitomycin. Direct attack on cell regulatory protein is suggested. Further, compounds play a vital role in controlling the regulation of Sub-G1 and G2/M phases. Some difference in the mechanism of operation of 4-aryl substituted 13,16 and 2-aryl substituted 9,10-anthraquinones 28 has been observed during the progression of cell cycle like: S phase was highly disturbed by 13; G0/G1 phase was highly disturbed by 28. Interestingly, 4-arylsubstituted 1-hydroxyanthraquinones 13,16 enhancement of DNA synthesis on SNB-19 cells (Table 3). This observation can be explained by the influence on the cell repair systems that respond to the intercalation of compounds in DNA. It is well established that the anthraquinone drug mitoxantrone arrests G1 and G2 phases at cell cycle progression and ultimately inhibits the cell growth [57]. Further, mitoxantrone promotes the arrest of S phase of cell cycle. Hence, cell goes apoptosis upon treatment with mitoxantrone due to inhibition at cell growth phases G1 and G2 along with inhibiting the DNA replication/duplication process (S phase).
Characteristically, the induction of apoptosis by the cell cycle arrest at G1 phase in MDA-MB-231 breast cancer cells was also established for the 1-arylanthraquinone containing fraction of the medicinal plant Bulbine frutescens [58]. Compound 13 in accordance to cytotoxic studies on DU-145 cell lines shows better effect over Doxorubicin, i.e., 1.1 and 2.0 µM, respectively (Table 1). We have carried out flow cytometric study of compounds 13, 16 and the isomeric compound 25 (Table 4, Figure 4G,H). Concentrations of 16 and 25 were taken as 5.4 and 14.5 µM, respectively, as per cytotoxic studies (Table 1). After examining the data for 13 ( Figure 4G, Table 4), a good number of cells are arrested in Sub-G1, G0/G1 and S phase. Biosynthetic activity is very high during Sub-G1 phase; this is probably achieved by enhancing of apoptosis. The isomeric molecule 25 arrest G0/G1 and S phases, i.e., affecting the DNA synthesis/replication mechanism. Similarly, for 25 a different pattern is observed, i.e., variations at S and G2/M phases were observed, in comparison to the control. Hence, both the isomeric molecules 13 and 25 deregulate the cell cycle which is the primary condition for any drug candidate to be cytotoxic. Some difference in the mechanism of operation of 13 and 25 has been observed during the progression of cell cycle like: Sub-G1 phase was highly disturbed by 13 and S and G0/G1 phases were highly disturbed by 25.

Conclusions
A straightforward methodology has been developed for the introduction of an aryl substituent at C-2, C-4 or C-2,4 positions in the anthraquinone core in a two-step procedure starting from 1-hydroxyanthraquinone by Suzuki-Miyaura cross coupling reaction of the subsequent halogen substituted anthraquinones with aryl (hetaryl) boronic acids. The cytotoxicity of twenty-six novel compounds was tested against a panel of seven human tumor cell lines and also towards hTERT-immortalized lung fibroblast cells in the MTT assay. Cytotoxicity studies revealed that six of the aryl substituted compounds among the entire series 3, 15, 16, 25, 27, 28 are more potent than the commercially available reference drug doxorubicin against human glioblastoma SNB-19, prostate cancer DU-145 or breast cancer MDA-MB-231 cells and relatively safe towards hTERT-immortalized lung fibroblasts cells. The structure-cytotoxicity investigations implied that the phenyl, 4-methoxyphenyl, 2,3-dimethoxyphenyl or 3,5-difluorophenyl substituted 1-hydroxyanthraquinones exhibited the higher cytotoxicity in glioblastoma cancer cell lines. Another observed effect is the enhancement of DNA synthesis in SNB-19 cells for 4-aryl 1-hydroxyanthraquinones 13, 16 compared with doxorubicin and especially mitomycin C can be explained by the strengthening of cell repair systems that respond to the intercalation of compounds in DNA. and Avance 600 ( 1 H: 600.30 MHz, 13 C: 150.95 MHz) spectrometers (Bruker BioSpin GmbH, Rheinstetten, Germany), using tetramethylsilane (TMS) as an internal standard. NMR signal assignments were carried out with the aid of a combination of 1D and 2D NMR techniques that included Heteronuclear Single Quantum Correlation (HSQC) and Heteronuclear Multiple Bond Correlation (HMBC). Chemical shifts are reported in parts per million (ppm) and coupling constants are expressed in Hz. HRMS spectra were recorded on a Thermo Scientific DFS mass spectrometer (evaporator temperature 200-230 • C, EI ionization at 70 eV). Melting points were determined using termosystem Mettler Toledo FP900 (USA). The analytical and spectroscopic investigations were carried out at the Collective Use Center for Chemical Services of the Siberian Branch of the Russian Academy of Sciences.

General Information
The reaction progress was monitored by TLC on Silufol UV-254 plates (Kavalier, Czech Republic), CHCl 3 -EtOH, 100:1; detection under UV light. Column chromatography was performed by using silica gel (0.070-0.230 mm, Acros-Organics). Purity of all compounds was checked by TLC.

Molecular Docking Study
Molecular modeling was carried out in the Schrodinger Maestro visualization environment using applications from the Schrödinger Small Molecule Drug Discovery Suite 2016-1 package [59]. Three-dimensional structures of the derivatives were obtained empirically in the LigPrep application using the OPLS3 force field [60]. For the calculations, the XRD model of topoisomerase IIβ-DNA complex inhibited by mitoxantrone from Protein Data Bank was chosen (PDB ID 4G0V) [61]. To model a possible mechanism of inhibition of selected target, molecular docking of new compounds was performed at the binding site of topoisomerase IIβ-DNA complex using Glide [62]. The search area for docking was selected according to the size of inhibitor. Docking was performed in comparison with mitoxantrone and doxorubicin. The three-dimensional structures of inhibitors were obtained in the PubChem database and prepared in the LigPrep application. Non-covalent interactions of molecules in the binding site were visualized using Biovia Discovery Studio Vizualizer.

Cell Culture and Determination of Cytotoxicity
The human cancer cells of the glioblastoma (U-87MG, SNB-19, T98G), prostate cancer cell line (LNCAP,, the cells of T-cellular human leucosis (MT-4) and human breast cancer cells (MDA-MB-231) were used in this study. The cells were cultured in the RPMI-1640 medium that contained 10% embryonic calf serum, L-glutamine (2 mmol/L), gentamicin (80 mg/mL) and lincomycin (30 mkg/mL) in a CO 2 incubator at 37 • C. The tested compounds were dissolved in DMSO and added to the cellular culture at the required concentrations. Three wells were used for each concentration. The cells which were incubated without the compounds were used as a control. Cells were placed on 96-well microliter plates and cultivated at 37 • C in 5% CO 2 /95% air for 72 h. The cell viability was assessed through an MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-phenyl-2H-tetrazolium bromide] conversion assay [55]. A total of 1% MTT was added to each well. Four hours later, DMSO was added and mixed for 15 min. Optical density (D) of the samples was measured on a BioRad 680 multi-well spectrophotometer (USA) at the wavelength of 450 nm. The 50% cytotoxic dose GI 50 ) of each compound (i.e., the compound concentration that causes the death of 50% of cells in a culture, or decreases the optical density twice as compared to the control wells) was calculated from the data obtained. Statistical processing of the results was performed using the Microsoft Excel-2007, STATISTICA 6.0 and GraphPad Prism 5.0 programs. The results are given as an average value ± a deviation from the average (mean ± standard error of the mean (SEM)). Reliability of differences (p) was estimated using the Student t test. The differences with p < 0.05 were considered as reliable. The experimental results are given as the data average values obtained from three independently conducted experiments.