Discovery of novel quinoline-based analogues of combretastatin A-4 as tubulin polymerisation inhibitors with apoptosis inducing activity and potent anticancer effect

Abstract A new series of quinoline derivatives of combretastatin A-4 have been designed, synthesised and demonstrated as tubulin polymerisation inhibitors. These novel compounds showed significant antiproliferative activities, among them, 12c exhibited the most potent inhibitory activity against different cancer cell lines (MCF-7, HL-60, HCT-116 and HeLa) with IC50 ranging from 0.010 to 0.042 µM, and with selectivity profile against MCF-10A non-cancer cells. Further mechanistic studies suggest that 12c can inhibit tubulin polymerisation and cell migration, leading to G2/M phase arrest. Besides, 12c induces apoptosis via a mitochondrial-dependant apoptosis pathway and caused reactive oxygen stress generation in MCF-7 cells. These results provide guidance for further rational development of potent tubulin polymerisation inhibitors for the treatment of cancer. Highlights A novel series of quinoline derivatives of combretastatin A-4 have been designed and synthesised. Compound 12c showed significant antiproliferative activities against different cancer cell lines. Compound 12c effectively inhibited tubulin polymerisation and competed with [3H] colchicine in binding to tubulin. Compound 12c arrested the cell cycle at G2/M phase, effectively inducing apoptosis and inhibition of cell migration.


Introduction
Cancer is a disease of an uncontrolled growth and abnormal division of cells, which leads to death. In recent years, targeted antineoplastic agents have become an effective treatment choice for cancer 1 , with pharmaceutical companies focussing on targeted therapies against different and special cancer types 2 . Tubulin polymerisation inhibitors represent one of the most well-known and potential examples of such targeted cancer therapies 3 . Tubulin is a globular protein that performs a substantial function in cell mitosis. Microtubules (MTs), which represent the basic constituents of eukaryotic cell, are cytoskeletons constructed by the association of aand b-tubulin heterodimers with a head and tail pattern to form hollow cylindrical tubes (nearly 25 nm in diameter) [4][5][6][7] . MTs play a crucial role in many fundamental cellular activities, such as motility, cell formation, cell secretion, signalling, maintenance of cell shape, regulation of intracellular transport and cell division [8][9][10] . Due to these multiple functions, microtubule system has become an attractive target for cancer chemotherapy 11,12 . Disruption of MTs or tubulin dynamics exposes the cell to mitotic arrest of the cell cycle at G 2 /M phase, and consequently induction of cellular apoptosis 13,14 . Several microtubule-interfering agents (MIAs) have been identified, e.g. paclitaxel, vincristine, and colchicine that are obtained from the natural products, taxol, vinca and colchicine, respectively. MIAs are known to bind to tubulin at specific binding sites that are classified as taxol, vinca and colchicine sites to either enhance or inhibit tubulin polymerisation. 15 For example, microtubule stabilisers, e.g. paclitaxel stimulate microtubule polymerisation 16 , while microtubule destabilizers, e.g. colchicine, and the vinca alkaloids vinblastine and vincristine inhibit polymerisation of microtubules 17 . Significant attention is now focussed on colchicine binding site inhibitors due to their positive impact on ABC-transporter-mediated drug resistance 18,19 . Combretastatin A-4 (CA-4), (1, Figure 1) has been reported as the most potent antimitotic agent of this family against several tumour cells 20 . CA-4 was first isolated from the bark of the willow tree Combretum caffrum from South Africa in 1989 21 . CA-4 has a vascular disrupting activity against tumour cell vasculature by preventing blood supply to solid tumour, resulting in apoptosis [22][23][24] . Given its structural simplicity, CA-4 has been studied as a lead pharmacophore for deciphering tubulin functions and properties 25 . Phases II and III clinical studies are currently ongoing with tubulin-targeted drugs 24,26 . Structure activity relationships (SAR) studies with CA-4 have revealed three important structural features (Figure 1). These include: (i) a 3,4,5-trimethoxy moiety on ring A that is essential for activity; (ii) a cisconfiguration of both aromatic rings that is essential for activity (trans-orientation is inactive); (iii) the presence of small substituent on ring B, e.g. methoxy group that is important for activity. The cis-alkene configuration in CA-4 allows the aromatic rings to assume optimal binding orientation for interactions with the colchicine binding site. Unfortunately, the cis configuration of CA-4 has a propensity for undergoing transformation to the inactive trans configuration upon storage and during in vivo metabolism. To overcome this, many structural modifications of CA-4 have been undertaken where the cis double bond is replaced with heterocycles, either monocyclic, such as oxadiazole, isoxazole and imidazole, resulting in compounds, such as 1, 2 and 3 respectively (Figure 1) [27][28][29][30][31][32][33] or fused heterocyclic, such as pyrazolopyridines 34 , triazolopyridines 35 and triazolothiadiazine derivatives 36 . These compounds, like CA-4 showed pronounced activity against a panel of cancer cell lines.
Quinoline derivatives are popular for the treatment of malaria 37,38 . Moreover, quinoline heterocyclic containing compounds demonstrate potent anticancer activities with different modes of actions, including inhibition of proteasome, tyrosine kinases, and tubulin polymerisation [39][40][41] . Previous studies have reported the antiproliferative activity of CA-4, isoCA-4 or chalcone compounds containing quinoline scaffold, either as ring A bioisoster, e.g. 4a 23 and 4b 42 or ring B bioisoster, e.g. 4c 43-45 and 4d 46 . These compounds demonstrate the potential of the quinoline ring as a template for developing more promising tubulin polymerisation inhibitors and antiproliferative agents.
In this work, we optimised CA-4 into a series of novel hybrid quinoline derivatives as potent tubulin inhibitor, which involves introducing a rigid oxazolone or imidazolone between rings A and B to maintain the cis configuration, as well as targeting the quinolyl moiety (ring B), by varying the electronic substituents effect while maintaining the 3,4,5-trimethoxyphenyl moiety as present in ring A of CA-4 ( Figure 2). Following, we synthesised several analogues that constitute two classes of compounds: the oxazolones (Compounds 12a-h) and the imidazolones (Compounds 13a-h). The compounds have been screened for their antiproliferative activities against a variety of cancer cell lines, as well as studied for their mechanism of action. We expect the results to lead to better understanding of the mechanistic mode of the compounds' activity against tubulin and provide guidance for further development of potent anticancer drugs.

Chemistry
Melting points were determined with a Gallenkamp (London, UK) melting point apparatus and are uncorrected. IR spectra (KBr, cm À1 ) were recorded on Bruker Vector, 22FT-IR [Fourier Transform Infra-red (FTIR), Germany] spectrometer. Unless otherwise specified, proton ( 1 H) and carbon ( 13 C) NMR spectra were recorded at room temperature in base filtered CDCl 3 on a spectrometer operating at 400 & 300 MHz for proton and 100 & 75 MHz for carbon nuclei. The signal due to residual CHCl 3 appearing at d H 7.26 and (CH 3 ) 2 SO appearing at d H 2.5 and the central resonance of the CDCl 3 "triplet" appearing at d C 77.0 and for (CD 3 ) 2 SO "multiplet" appearing at d C 39.0 were used to reference 1H and 13 C NMR spectra, respectively. 1 H NMR data are recorded as follows: chemical shift (d) [multiplicity, coupling constant(s) J (Hz), relative integral] where multiplicity is defined as s ¼ singlet; d ¼ doublet; t ¼ triplet; q ¼ quartette; and m ¼ multiplet or combinations of the above. Elemental analyses were determined using Manual Elemental Analyser Heraeus (Germany) and Automatic Elemental Analyser CHN Model 2400 Perkin Elmer (Waltham, MA, USA) at Microanalytical Centre, Faculty of Science, Cairo University, Egypt. All the elemental analyses results corresponded to calculated values within experimental error. Progress of reactions was monitored by thin-layer chromatography (TLC) using precoated TLC sheets with Ultraviolet (UV) fluorescent silica gel (Merck 60F254), and spots were visualised by iodine vapours or irradiation with UV light (254 nm). All chemicals were purchased from Sigma-Aldrich or Lancaster Synthesis Corporation (UK). Intermediates 6-8a-i were prepared according to reported procedure 47,48 .

2.1.2.
General procedure for preparation of (13a-h) The appropriate oxazolones 12a-h (1 mmol) was stirred and heated under reflux in ethanol (10 ml) containing ammonium hydroxide (10 ml), and the reaction monitored by TLC. After completion of the reaction in 24 h, the solvent was concentrated and cooled, and the precipitate was filtered off and crystallised from ethanol.  13

Biochemical evaluation of activity
All biochemical assays were performed in triplicate on at least three independent occasions for the determination of mean values.

Cell culture
The four human tumour cell lines MCF-7, HCT-116, HL-60 and HeLa used in this study were obtained from the VACSERA (Giza, Egypt) cell culture unit that were originally acquired from ATCC (Manassas, VA, USA). All the human tumour cell lines were cultured in Dulbecco's Modified Eagle's Medium (DMEM) with 10% foetal bovine serum, 2 mM L-glutamine and 100 mg/mL penicillin/ streptomycin. Cells were maintained at 37˚C in 5% CO 2 in a humidified incubator. All cells were sub-cultured 3 times/week by trypsinisation using TrypLE Express (1X).

Cell viability assay
The quinoline compounds were evaluated for antiproliferative effect using the MTT viability assay of four cancer cell lines (MCF-7, HCT-116, HL-60 and HeLa) and normal breast cells MCF-10A to calculate the relative IC 50 values for each compound. Cells were seeded in triplicate in 96-well plates at a density of 10 Â 10 3 cells/ml in a total volume of 200 ml per well. 0.1% of DMSO was used as a vehicle control. Following, the cells were treated with 2 ml test compound (from stock solutions in ethanol) to furnish the concentration range of study, 1 nM to 50 mM, and re-incubated for a further 72 h. The culture medium was then removed, and the cells washed with 100 mL phosphate buffered saline (PBS) and 50 mL MTT added, to reach a final concentration of 1 mg/mL. Cells were incubated for 2 h in darkness at 37 C. Solubilisation was begun through the addition of 200 ml DMSO, and the cells maintained at room temperature in darkness for 20 min to ensure thorough colour diffusion before reading the absorbance. Plates were incubated for 72 h at 37 C þ 5% CO 2 . The MTT (5 mg/mL in PBS) was added and incubated for another 4 h, and the optical density was detected with a microplate reader at 570 nm. Results were expressed as percentage viability relative to vehicle control (100%). Dose response curves were plotted and IC 50 values (concentration of drug resulting in 50% reduction in cell survival) were obtained using the commercial software package Prism (GraphPad Software, Inc., La Jolla, CA, USA). All the experiments were repeated in at least three independent experiments.

Tubulin polymerisation assay
The assembly of purified bovine tubulin was monitored using a kit, BK006, purchased from Cytoskeleton Inc., (Denver, CO, USA). The assay was carried out in accordance with the manufacturer's instructions using the standard assay conditions 49 . Briefly, purified (>99%) bovine brain tubulin (3 mg/mL) in a buffer consisting of 80 mM PIPES (pH 6.9), 0.5 mM EGTA, 2 mM MgCl 2 , 1 mM GTP and 10% glycerol was incubated at 37 C in the presence of either vehicle (2% (v/v) ddH 2 O), CA-4, or the quinoline compounds. Light is scattered proportionally to the concentration of polymerised microtubules in the assay. Therefore, tubulin assembly was monitored turbidimetrically at 340 nm in a Spectramax 340 PC spectrophotometer (Molecular Devices, Sunnyvale, CA, USA). The concentration that inhibits tubulin polymerisation by 50% (IC 50 ) was determined using area under the curve (AUC). The AUC of the untreated controls were considered as 100% polymerisation. The IC 50 value for each compound was computed using GraphPad Prism Software.

Colchicine site competitive binding assay
The affinity of compounds 12c to colchicine binding site was determined using Colchicine Site Competitive Assay kit CytoDYNAMIX Screen15 (Cytoskeleton, Inc., Denver, CO, USA) using the standard protocol of the manufacturer to determine Ki. Biotin-labelled tubulin (0.5 mg) in 10 mL of reaction buffer was mixed with [3H]colchicine (0.08 mM, PerkinElmer, Waltham, MA) and the test compounds (positive control colchicine, negative control vinblastine, G-1, fluorescent G-1, or 2-ME) in a 96-well plate (final volume: 100 mL). After incubating for 2 h at 37 C with gentle shaking, streptavidin-labelled yttrium SPA beads (80 mg in 20 mL reaction buffer, PerkinElmer, Waltham, MA) were added to each well and incubated for 30 min at 4 C. The plates were then read on a scintillation counter (Packard Instrument, Topcount Microplate Reader) and the percentage of inhibition was calculated 50,51 .

Cell cycle analysis
MCF-7 cells were seeded at a density of 1 Â 10 5 cells/well in 6well plates and treated with CA-4 (50 nM) and compound 12c (50 and 250 nM) for 24, 48 and 72 h. The cells were collected by trypsinisation and centrifuged at 800Âg for 15 min. Cells were washed twice with ice-cold PBS and fixed in ice-cold 70% ethanol overnight at À20 C. Fixed cells were centrifuged at 800Âg for 15 min and stained with 50 mg/mL of PI, containing 50 mg/mL of DNasefree RNase A, at 37 C for 30 min. The DNA content of cells (10,000 cells/experimental group) was analysed by flow cytometer at 488 nm using a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA) and all data were recorded and analysed using the CellQuest Software (Becton-Dickinson).

Annexin V/PI apoptotic assay
Apoptotic cell death was detected by flow cytometry using Annexin V and propidium iodide (PI). MCF-7 Cells were seeded in 6 well plated at density of 1 Â . Samples were analysed immediately using the BD accuri flow cytometer and prism software for analysis the data. Four populations are produced during the assay Annexin V and PI negative (Q4, healthy cells), Annexin V positive and PI negative (Q3, early apoptosis), Annexin V and PI positive (Q2, late apoptosis) and Annexin V negative and PI positive (Q1, necrosis).

Evaluation of expression levels of anti-apoptotic proteins bcl-2, pro-apoptotic proteins bax and caspase 9
The level of the anti-apoptotic marker and apoptotic marker BAX were assessed using Bcl-2 Elisa kit and human Bax ELISA Kit purchased from Zymed laboratories, invitrogen and Cloud-Clone Crop. (Katy, TX, USA), respectively, following the manufacturer's instructions. Briefly, Treated MCF-7 cell lysate with 250 nM of compound 12c were prepared, and equal amount of cell lysates were loaded and propped with specific antibodies. The samples were measured and analysed at 450 nm in ROBONEK P2000 ELISA reader 52 . In Vitro Caspase-9 Activation Assay was performed using human active caspase-9 Invitrogen EIA kit according to the manufacturer's instructions. Compound 12c at concentrations of 50 and 250 nM and CA-4 (50 nM) were prepared in dH 2 O up to a final volume of 50 mL/well followed by addition of 5 mL of active caspase-9. Following, the cells were mixed and 50 mL of the Master Mix was added to each well and allowed to react at 37 C for 1 h. The fluorescence intensity of the test samples was recorded and analysed in a fluorescence plate reader at 400 nm excitation and 505 nm emissions. All experiments were conducted in triplicates.

Colony formation assay
MCF-7 cells (600 cells per well) were seeded in 6-well plates and incubated for 24 h before being then treated with different doses of the compound 12c (50 and 250 nM) for 14 days. Following, the cells were washed with PBS twice and subsequently fixed with 4% paraformaldehyde and stained with 0.05% crystal violet for 30 min. Finally, cells were visualised using an inverted microscope.
2.2.9. Wound healing assay MCF-7 were grown in 6-well plates for 24 h, and scratches were made using pipette tip and washed with PBS to remove non-adherent cell debris. Subsequently, the cells were treated with different concentrations of 12c for 24 h. The migrations across the wound area were photographed under a phase contrast microscopy. Fluorescence spectra (510 À 600 nm) were monitored using an excitation wavelength of 488 nm 53,54 .

Design and chemistry
The natural product, Combretastatin A-4 (CA-4; Figure 1) exhibits significant antiproliferative activities against several tumour cells by binding to the colchicine site of tubulin to inhibit the protein polymerisation 20 . However, the cis double bond of CA-4 has a propensity to isomerise into the inactive trans configuration, leading to reduction in the molecule's pharmacologic activity. Several structural modifications of the CA-4 pharmacophore have subsequently been undertaken to overcome this disadvantage, e.g. replacing the cis double bond with a heterocycle, oxadiazole, isoxazole and imidazole, resulting in compounds, such as 1, 2 and 3 respectively (Figure 1) [27][28][29][30][31][32] . In this work, we undertook a rational design approach of introducing chalcone system (ring C) in the form of either oxazolone or imidazolone between the two rings A and B, as well as isosterically replace ring B with quinolone. Specifically, the rigidity of the molecules was increased by introducing 1,3-oxazol-5-ones and 1,3-imidazol-4-ones to the cis-olefinic bond of CA-4, which we anticipate would create a desirable conformational and configurational restriction to prevent isomerisation of CA-4 into the inactive trans-isomer, as well as improve on the anticancer activities of these compounds since chalcones are well known for their anticancer properties 55,56 . The second design step involves varying the electronic substituents effect on the quinolyl moiety (ring B), while maintaining the natural active compound 3,4,5-trimethoxyphenyl moiety, which we anticipate will increase the potency of these compounds.
The syntheses of the proposed quinoline compounds 12a-h (oxazolones) and 13a-h (imidazolones) ( Table 1) are shown in Schemes 1 and 2 and involve two core structural components: (i) 2-methoxyquinolyl-3-carbaldehyde nucleus 8a-h, and (ii) 3,4,5-trimethoxyphenyl moiety 11. A concise (three-step) synthesis was used for the synthesis of the first core structure, 2-  methoxyquinoline-3-carbaldehyde derivatives 8a-h, as shown in Scheme 1. The synthesis was initiated with acetylation of the starting aniline derivatives 5a-h using acetic anhydride and glacial acetic acid at 0 C. The produced amides 6a-h were subjected to Vilsmeier-Haack reaction to give the corresponding quinoline-3aldehyde derivatives 7a-h. Addition of methoxy substituent to 7a-h to give 8a-h was achieved through the use of sodium methoxide at 40 C in methanol 47,48 . The synthesis of the second core 3,4,5-trimethoxyphenyl moiety 11 started with acylation of the acid 9 under highly acidic condition using SOCl 2 to give acyl benzotriazole 10 (Scheme 2) 57,58 . Following, treatment of the acyl benzotriazole 10 with glycine in aqueous acetonitrile gave the acyl glycine 11. Condensation of 11 with the appropriate quinoline aldehydes 8a-h in the presence of acetic anhydride and catalytic amount of sodium acetate resulted in the formation of the oxazolones 12a-h. Aminolysis of 12a-h via condensation reaction with ammonia led to the formation of the imidazolones 13a-h. It seems the nucleophilic ammonia attacks the carbonyl group of the oxazolone ring, followed by immediate intramolecular condensation and cyclisation to give the imidazolones 13a-h. In summary, two classes of compounds, 12a-h (oxazolones) and 13a-h (imidazolones) were synthesised and used for further functional and biological studies.

In vitro antiproliferative activities
All the synthesised compounds (with CA-4 as a positive reference) were evaluated for their antiproliferative activities using MTT assay with four different cancer cell lines -MCF-7 breast adenocarcinoma, HL-60 leukaemia, HCT-116 colorectal carcinoma, and HeLa cervical adenocarcinoma. As shown in Table 1, most of the compounds demonstrated moderate to highly potent antiproliferative activities. In the oxazolone analogues (12a-h), compound 12a without any substituent on the quinoline ring was the least active when compared with quinoline ring substituted compounds. The relative position of the substituent on the quinoline ring also seemed to be critical for antiproliferative activity. Compound 12c with methyl group at the 7-position ring displayed impressive non-selective potency in nanomolar range against HL-60, MCF-7, HCT-116 and HeLa cell lines with IC 50 of 0.019, 0.010, 0.022 and 0.042 mM, which compared to 0.076, 0.019. 0.026 and 0.064 mM for CA-4, respectively. In contrast, both the 6-CH 3 analog (12b) and 8-CH 3 analog (12d) were 3-to 15-fold less active than 12c. The nature of the substituents on the quinoline ring of the oxazolone compounds was also found to significantly influence the biological activity. For example, replacement of the methyl group in 12b and 12c with a stronger electron-releasing methoxy group yielded compounds 12e and 12f, respectively, which resulted in better antiproliferative activities. The methoxy-containing compound 12e was 2.7-to 13-fold more active than the methyl-containing compound 12b with the four cancer cell lines (IC 50 of 0.068, 0.056, 0.031 and 0.010 mM in HL-60, MCF-7, HCT-116 and HeLa cancer cell lines, respectively). Compound 12f had a similar effect as 12c against MCF-7 and HCT-116 cells (0.052 and 0.066 mM, respectively), but with reduction in activity against the other two cell lines, HL-60 and HeLa (0.352 and 0.138 mM, respectively). Introducing larger substituents at the quinoline ring as in 12g (7tert-butyl) and 12h (7-benzyloxy) led to a dramatic decrease in activity compared to their corresponding analog 12f (7-methoxy). Summarily, adding smaller and/or polar groups to the quinoline ring of the oxazolone resulted in significant improvement in the antiproliferative activity. The imidazolones (Compound 13a-h) also resulted in impressive antiproliferative activity with IC 50 values ranging from 0.04-8.21 mM in all four cell lines. In general, the imidazolones showed similar antiproliferative activities as the oxazolones (Table  1), which could be due to similar electronic effects of the oxazolone and imidazolone rings. Like the oxazolone, lack of substituent on the quinolone ring as in compound 13a led to reduction in activity, with IC 50 values of more than 1 mM in all four cell lines, similar to the results obtained with the oxazolone derivative 12a. Methyl substitution on the quinoline ring, e.g. 6-CH 3 13b, 7-CH 3 13c and 8-CH 3 13d led to potent activity in submicromolar range in all four cancer cell lines.
The position of the methoxy substituent on the quinoline heterocycle also influenced the antiproliferative activity of the  compounds against the cancer cell lines. For example, the antiproliferative activity of 6-methoxy-substituted 13e was better than its analog 7-methoxy-substituted 13f against MCF-7, HCT-116 and HeLa cells with IC 50 values of 0.042, 0.085 and 0.062 mM, which compare to 0.092, 0.187 and 0.101 mM for 13f, respectively. However, in HL-60, 13e exhibited less antiproliferative activity with IC 50 value of 0.272 mM. In a similar trend as the oxazolone derivatives, bulky substituents on the quinoline ring 13g (7-tert-butyl) and 13h (7-benzyloxy) resulted in drastic decrease in activity in all four cancer cell lines with 14-to 125-fold loss in potency compared to their corresponding 13f (7-methoxy containing) compound. In summary, both oxazolone and imidazolone compounds displayed potent antiproliferative effects, strengthening our hypothesis that nitrogen-containing heterocycles, such as quinoline, are beneficial surrogates for the ring B of CA-4. The different biological activities of the compounds are likely the result of differences in their mode of interaction with the colchicine binding site. Due to its excellent antiproliferative activity, compound 12c was studied in more details as described below.

3.2.2.
In vitro inhibition of tubulin polymerisation and colchicine binding Trimethoxyphenyl (TMP) containing stilbenoid derived compounds, such as colchicine, resveratrol and CA-4 bind to tubulin at the colchicine binding site, resulting in inhibition of microtubule polymerisation 59,60 . To confirm whether the quinoline compounds similarly target the tubulin-microtubule system, representative quinoline compounds, including four oxazolone analogues (12a, 12c, 12e and 12g) and two imidazolone analogues (13c and 13e), as well as the reference compound CA-4, were evaluated for their antitubulin polymerisation activities and the results presented in Table 2. The methyl and methoxy substituted oxazolone compounds 12c and 12e, respectively strongly inhibited tubulin assembly with IC 50 of 1.21 and 2.26 mM, respectively compared to that of CA-4 (IC 50 of 2.17 mM), while the  The methyl analogue 13c was inactive in the tubulin polymerisation assay (IC 50 of 20.29 mM), and is 16-fold less active compared to its corresponding oxazolone derivative 12c, which is in agreements with the poor cell growth inhibitory activity of 13c compared to 12c. Compound 12c was also examined at two different concentrations (1 and 5 mM) for its ability to compete with colchicine for binding to tubulin using a [ 3 H] colchicine binding assay. Compound 12c strongly inhibited colchicine binding to tubulin by 79% and 87% at 1 and 5 mM respectively, which compares with 86% and 97% inhibition by CA-4, respectively. These results suggest that compound 12c is involved in tubulin polymerisation inhibition through the colchicine-binding site.

Cell cycle analysis
Induction of cell cycle arrest at G 2 /M phase is strongly accompanied with tubulin polymerisation inhibition. It is well established that CA-4 arrests cell cycle at G 2 M phase [61][62][63] . To further gain insight into compound 12c potent antiproliferative activity, cell cycle analysis of MCF-7 cells was performed at two concentrations of 50 nM and 250 nM and at different time points of 0, 24, 48 and 72 h. Figure 3(A) clearly demonstrates that 12c caused a significant arrest in G 2 /M phase and apoptosis in a dose-and time-dependent manner. After 48 h, the percentage of G 2 /M phase arrested cells were 28.4% and 38.3% at 50 nM and 250 nM, respectively compared to 9.2% of untreated cells (Figure 3(B)). Moreover, there was an increase in the number of cells in G 2 /M phase after 72 h (33.0% and 40.8% at 50 nM and 250 nM, respectively) with a concomitant decrease of cells in G 0 /G 1 phase (40.3% and 29.8% at 50 nM and 250 nM, respectively) compared to the control (57.3%). In a comparable finding, CA-4 (50 nM) also significantly arrested G 2 /M phase at 24, 48 and 72 h (40.3%, 43.8% and 47.7%, respectively). Accordingly, a concomitant decrease of MCF-7 cells was detected in G0 phase (Figure 3(C)). Furthermore, compound 12c induced a gradual increase in apoptosis (16.2%, 23.4% and 32.7%) at 250 nM as the proportion of cells in the sub-G1 phase increased at 24, 48 and 72 h, respectively compared to untreated cells (1.5%) (Figure 3(D)). Similarly, 23.5%, 31.1% and 37.4% increase in apoptosis was observed for CA-4 at 24, 48 and 72 h, respectively. These findings are in agreement with previously reported for a series of related quinoline analogues, which significantly induced apoptosis and G 2 /M cycle arrest in MCF-7 cells 23,42,46,64 .

Cell apoptosis
We investigated whether cell death induced by compound 12c treatment was related to apoptosis using the Annexin-V/PI double staining flow cytometric assay (Figure 4(A,B)). MCF-7 cells were treated with three different concentrations (0, 50 and 250 nM) of compound 12c at different time points (24, 48 and 72 h). Compound 12c caused a significant accumulation of annexine-V positive cells and induced both early and late apoptosis in a doseand time-dependent manner compared to the untreated cells. As shown in Figure 4(B), when the cells were treated with 12c (0 and 250 nM) or CA-4 (50 nM) for 48 h, the percentage of Annexin Vstaining positive cells significantly increased from 1% in untreated cells to 15%, 21% and 29% respectively. The percentage of early and late apoptotic cells in the presence of 12c increased after 72 h to 17.6% and 29.3% at 1 and 5 mM respectively when compared to the untreated cells (2%). Based on the cell cycle arrest and apoptosis findings (Figure 3(B-D)), it appears that compound 12c could efficiently induce apoptosis cell death in MCF-7 cells in a dose-and time-dependent manner.

Assessment of toxicity to non-tumorigenic human cells
To assess the cytotoxicity and selectivity of 12c towards cancer cells, normal epithelial breast MCF-10A cell viability study was carried out. As shown in Figure 5(A), the IC 50 value of 12c was more than 50 mM in MCF-10A cells, which was significantly higher than the IC 50 values of 19, 10, 22 and 42 nM in MCF-7, HL-60, HCT-116 and HeLa cancer cell lines, respectively. Remarkably, 12c was found to be less toxic in normal MCF-10A (IC 50 >50 mM) when compared to CA-4 (IC 50 ¼ 6.1 mM) ( Figure 5(B)), suggesting 12c to have better selective toxicity against cancer cells.

Expression of the apoptotic proteins in MCF-7 cell lines
The previous data clearly demonstrate that 12c is an effective anti-mitotic quinoline compound in MCF-7 cell lines. Herein, the effect of 12c on the expression of apoptosis pathway markers, Bcl-2 anti-apoptotic protein and Bax pro-apoptotic protein was investigated. MCF-7 cells treated with 12c at 250 nM for 48 h decreased the expression level of the anti-apoptotic protein Bcl-2, and correspondingly up-regulated the expression of the pro-apoptotic protein Bax ( Figure 6(A,B)). Activation of caspases initiates apoptosis, and in particular caspase-9 is considered an important effector caspase responsible for This finding confirms that compound 12c like CA-4 enhanced the rate of apoptosis in MCF-7 cell through caspase-9 activation.

Inhibition of colony formation
Colony formation assay is one of the effective techniques for the determination of long-term cell proliferation upon anticancer drug exposure. The inhibitory potential of 12c on MCF-7 cells colony formation is displayed in Figure 7. Compound 12c suppressed the clonogenic formation potential of MCF-7 cells in a dose dependent manner when compared to CA-4.  Figure 9. Assessment of mitochondrial membrane potential (Dw mt ) after treatment of MCF-7 cells with 12c. Cells were treated with indicated concentration of compound 12c for 6, 12 and 24 h and then stained with fluorescent DiOC2(3) for analysis of mitochondrial potential. Cells were then analysed by flow cytometry as described in the experimental section. Data are presented as mean ± SEM of three independent experiments. Statistical analysis was performed using two-way ANOVA ( ÃÃ p < 0.01; ÃÃÃ p < 0.001).

Wound healing assay
Migration and motility of cancer cells are considered as critical factors in tumour progression and metastasis 67,68 . In order to investigate the effect of compound 12c on the migration of MCF-7, wound healing assay was performed. As illustrated in Figure  8(A,B), the untreated cells migrated to the scraped area while in 12c-treated wells, cell migration was significantly inhibited in a dose-dependent manner. This significant difference in the wound area confirms that 12c suppressed MCF-7 cell migration, an important event in tumour metastasis.

Mitochondrial membrane potential
Mitochondria membrane potential plays a crucial role in the propagation of apoptosis. Specifically, loss of mitochondrial membrane potential Dw mt (MMP) is characteristic of early stage of apoptosis [69][70][71] . To confirm whether compound 12c could decrease the MMP of MCF-7 cancer cells, MMP was monitored by the fluorescence of the dye DiOC2(3). MCF-7 cells treated with 12c at 50 and 250 nM exhibited significant decrease in MMP in a dose-and time-dependant manner (Figure 9(A)). This depletion in MMP was associated with an increase of annexin-V positive early apoptotic cells. Maximum decrease in MCF-7 MMP was detected after 24 h treatment with 12c in which the percentage of apoptotic cells increased from 1.1% to 24.9% and 31.3% at 50 and 250 nM, respectively (Figure 9(B)). This indicates that compound 12c induces mitochondrial dysfunction in MCF-7, which eventually triggered apoptotic cell death. These results are in agreement with previously reported CA-4 analogues study that were shown to cause apoptosis through the mitochondrial pathway 68,72,73 .

Intracellular reactive oxygen species (ROS) production
The dissipation of mitochondrial potential is strongly associated with mitochondrial production of reactive oxygen species (ROS) 71,73 . The production of ROS after 12c treatment at 50 and 250 nM, as well as CA-4 (50 nM) with hydrogen peroxide H 2 O 2 was followed with 2,7-dichlorofluorescin diacetate (H 2 -DCFDA). As shown in Figure 10, after 24 h of 12c treatment, the levels of ROS in MCF-7 cells were 22.7 and 26.6% at 50 and 250 nM, respectively. The level in untreated MCF-7 cells was 1.0%, while it increased only to 18.3% in CA-4-treated cells. This result along with the significant loss of mitochondrial membrane potential above clearly suggests that compound 12c induced apoptosis via the mitochondrial pathway.

Conclusion
In this study, we designed, synthesised and evaluated two classes of novel quinoline compounds combretastatin A-4 derivatives as potential inhibitors of tubulin polymerisation. Several other studies have also reported derivatisation of the CA-4 pharmacophore with varying success [27][28][29][30][31][32] . Unlike the previous compounds, we for the first time introduced a chalcone system, including oxazolones and imidazolones to the cis bond of CA-4 to give more rigidity to the required active conformation. The chalcone system is well known for its anticancer activities. Our design also kept the essential natural trimethoxyphenyl pharmacophore (found in CA-4), while varying the electronic substituents effect on the quinolyl moiety (ring B) that were expected to enhance the potency of the compounds. Most of the compounds showed significant and, in some instances, comparable antiproliferative activities against different cancer cell lines as the previously studied combretastatin A-4 compound, CA-4. One of the most promising compound 12c showed potent anti-proliferative activities against HL-60, MCF-7, HCT-116 and HeLA cancer cell lines with IC 50 values of 0.019, 0.010, 0.022 and 0.042 mM, respectively, and simultaneously low cytotoxicity towards MCF-10A non-cancer cells. The microtubule polymerisation inhibitory effect of 12c was confirmed with an in vitro tubulin polymerisation and colchicine inhibition assays. Compound 12c effectively block the G 2 /M phase at the cell cycle and induce MCF-7 cell apoptosis together with significant change of Bax/Bcl expression ratio indicating involvement of mitochondrial apoptosis pathway. Further cellular mechanistic studies confirmed that 12c inhibited MCF-7 cell migration and colony formation. In conclusion, these results highlight our novel quinoline compounds and particularly 12c as promising anti-tubulin agent for the treatment of MCF-7 breast cancer cells. Moreover, the results point to a direction for rational development of potent tubulin polymerisation inhibitors for the treatment of cancer. Statistical analysis was performed using two-way ANOVA ( ÃÃÃ p < 0.001).