New Heterocyclic Combretastatin A-4 Analogs: Synthesis and Biological Activity of Styryl-2(3H)-benzothiazolones

Here, we describe the synthesis, characterization, and biological activities of a series of 26 new styryl-2(3H)-benzothiazolone analogs of combretastatin-A4 (CA-4). The cytotoxic activities of these compounds were tested in several cell lines (EA.hy926, A549, BEAS-2B, MDA-MB-231, HT-29, MCF-7, and MCF-10A), and the relations between structure and cytotoxicity are discussed. From the series, compound (Z)-3-methyl-6-(3,4,5-trimethoxystyryl)-2(3H)-benzothiazolone (26Z) exhibits the most potent cytotoxic activity (IC50 0.13 ± 0.01 µM) against EA.hy926 cells. 26Z not only inhibits vasculogenesis but also disrupts pre-existing vasculature. 26Z is a microtubule-modulating agent and inhibits a spectrum of angiogenic events in EA.hy926 cells by interfering with endothelial cell invasion, migration, and proliferation. 26Z also shows anti-proliferative activity in CA-4 resistant cells with the following IC50 values: HT-29 (0.008 ± 0.001 µM), MDA-MB-231 (1.35 ± 0.42 µM), and MCF-7 (2.42 ± 0.48 µM). Cell-cycle phase-specific experiments show that 26Z treatment results in G2/M arrest and mitotic spindle multipolarity, suggesting that drug-induced centrosome amplification could promote cell death. Some 26Z-treated adherent cells undergo aberrant cytokinesis, resulting in aneuploidy that perhaps contributes to drug-induced cell death. These data indicate that spindle multipolarity induction by 26Z has an exciting chemotherapeutic potential that merits further investigation.


Introduction
Microtubules are intracellular polymers comprised of α,β-tubulin heterodimers. They have essential cellular functions such as cell shape maintenance, organelle distribution, cell motility, and chromosome segregation [1]. Disruption of microtubule formation leads to G2/M phase cell cycle arrest and apoptosis [2]. Compounds that disrupt microtubule polymerization/depolymerization dynamics have been extensively studied as anticancer agents [2][3][4]. Such compounds include vinblastine, vincristine, and vinorelbine, which bind to tubulin's vinca domain, while docetaxel and paclitaxel bind to the taxane-site. [5]. Combretastatin A-4 (CA-4) binds to the colchicine binding site of tubulin. Originally isolated from the bark of the South African willow tree Combretum caffrum [6], CA-4 is one of the most potent inhibitors of tubulin polymerization, which interferes with microtubule dynamics and perturbs the mitotic cycle [7,8]. CA-4 selectively inhibits tumor neovascularization, causing rapid vascular shutdown and tumor necrosis [9,10]. When administered in vivo, CA-4 causes disruption and collapse of tumor blood vessels and subsequent necrotic cell death of tumor cells due to the lack of oxygen and nutrients [11][12][13]. The clinical usefulness of CA-4 and other vascular disrupting drugs is hampered by the formation of a viable tumor rim and by dose-limiting adverse effects, such as cardiotoxicity [14,15]. In addition, CA-4 has poor water solubility and is chemically unstable, with the active cis-isomer spontaneously converting to the thermodynamically favorable but less potent trans-isomer [16]. Numerous analogs of CA-4 have been tested to address these drawbacks [17][18][19][20]. The water-soluble phosphate prodrugs Zybrestat (2) and AVE-8082 (3) have shown promising results in pre-clinical and clinical studies, particularly in combination with other chemotherapeutic agents [21,22] against thyroid, non-small cell lung, and ovarian cancers [23][24][25]. We previously evaluated several CA-4 analogs possessing a benzoxazolone scaffold, and we identified (Z)-3-methyl-6-(3,4,5-trimethoxystyryl)-2(3H)-benzoxazolone (4, OP-107, Figure 1) as being quite potent in inducing anti-proliferative/cytotoxic and anti-angiogenic effects [19].

Single Crystal X-ray Diffraction
Single-crystal X-ray structure analysis was performed to precisely assign the E/Z conformation of the newly synthesized compounds. From the 26 new benzothiazolones, single crystals with suitable quality for data collection were grown by slow evaporation for compounds 19Z, 22Z, 22E, 23Z, 24Z, 25Z, 26E, and 27Z. ORTEP [50] views of the molecules present in the asymmetric unit (ASU) of the biologically active compounds (e.g., 25Z, 22Z, and 27Z) are shown in Figures 2 and S2.1-S2.5. The most important crystallographic and data refinement parameters from the X-ray diffraction experiment are given in Table S2.1. Compounds 25Z, 22Z, and 27Z crystallize as Z-configurational isomers in the corresponding orthorhombic Pca21, Pcab, and monoclinic C2/c space groups. Further, 27Z, 25Z, and 22Z are also positional isomers based on the styryl-benzothiazole link (C9-C5/C6 or C7), even though 25Z has only two methoxy groups, whereas 27Z and 22Z have three. The Evs. Z-configuration of stilbene derivatives is typically determined by 1 H NMR spectroscopy, based on the J-coupling constant of the olefin proton resonances. As a general rule, the Z-isomers display two doublets in the 6.33-6.75 ppm range with a J-constant around 12 Hz, while the E-isomer olefinic protons appear around 6.94-7.46 ppm as doublets, with a J-constant of 16 Hz. Uncharacteristically, the olefinic protons of the 3,5-dimethoxyphenyl (25Z) and 3,4,5-trimethoxyphenyl (26Z/E) stilbenes overlapped in CDCl 3 , resulting in singlet resonances. However, the signals were resolved in acetone-d 6 , which allowed the determination of the products' stereochemistry (J = 12.2 for the Z isomer and 16.3 Hz for the E isomer, respectively). The detailed information of NMR spectra is shown in Supplementary Materials (Figures S4.1.1-S4.14.2).

Single Crystal X-ray Diffraction
Single-crystal X-ray structure analysis was performed to precisely assign the E/Z conformation of the newly synthesized compounds. From the 26 new benzothiazolones, single crystals with suitable quality for data collection were grown by slow evaporation for compounds 19Z, 22Z, 22E, 23Z, 24Z, 25Z, 26E, and 27Z. ORTEP [50] views of the molecules present in the asymmetric unit (ASU) of the biologically active compounds (e.g., 25Z, 22Z, and 27Z) are shown in Figure 2 and Figure S2.1-S2.5. The most important crystallographic and data refinement parameters from the X-ray diffraction experiment are given in Table S2.1. Compounds 25Z, 22Z, and 27Z crystallize as Z-configurational isomers in the corresponding orthorhombic Pca2 1 , Pcab, and monoclinic C2/c space groups. Further, 27Z, 25Z, and 22Z are also positional isomers based on the styryl-benzothiazole link (C9-C5/C6 or C7), even though 25Z has only two methoxy groups, whereas 27Z and 22Z have three.  (Table S2.2). The benzothiazolone and methoxybenzene (mono-, di-, or tri-substituted) moieties are almost planar, with rmsd values < 0.03 Å. However, if we compare their relative orientation regarding the plane involving the double bond (-C-CH=CH-C-), there is a significant deviation of the angles between the mean planes of the benzothiazolone and methoxybenzene rings. The twist and fold angles between the rings and the (-C-CH=CH-C-) plane and those between the normal of the planes of the ring systems vary considerably (Table S2. 4.). Those values suggest considerable structural (conformational) flexibility ( Figure 3) of the synthesized molecules that would enhance ligand-protein interaction. Compounds 25Z, 22Z, and 27Z have non-typical donors (CH3), while the acceptors may be C-Omethoxy and/or C=O. Those molecular features result in the formation of weak intramolecular and intermolecular weak/hydrogen bonding interactions of C-Haromatic···C=O, C-Haromatic···C-Omethoxy, C-Hmethyl···C=O, and C-Hme-thyl···C-Omethoxy type (Table S2.3.), that stabilize the molecular and crystal structures. The three-dimensional packing of the molecules in the crystal structures of 25Z and 22Z shows zig-zag rearrangement along with c-and b-cell axes, respectively.  (Table S2.2). The benzothiazolone and methoxybenzene (mono-, di-, or tri-substituted) moieties are almost planar, with rmsd values < 0.03 Å. However, if we compare their relative orientation regarding the plane involving the double bond (-C-CH=CH-C-), there is a significant deviation of the angles between the mean planes of the benzothiazolone and methoxybenzene rings. The twist and fold angles between the rings and the (-C-CH=CH-C-) plane and those between the normal of the planes of the ring systems vary considerably (Table S2. 4). Those values suggest considerable structural (conformational) flexibility ( Figure 3) of the synthesized molecules that would enhance ligand-protein interaction. Compounds 25Z, 22Z, and 27Z have non-typical donors (CH 3 ), while the acceptors may be C-O methoxy and/or C=O. Those molecular features result in the formation of weak intramolecular and intermolecular weak/hydrogen bonding interactions of C-H aromatic · · · C=O, C-H aromatic · · · C-O methoxy , C-H methyl · · · C=O, and C-H methyl · · · C-O methoxy type (Table S2.3), that stabilize the molecular and crystal structures. The three-dimensional packing of the molecules in the crystal structures of 25Z and 22Z shows zig-zag rearrangement along with cand b-cell axes, respectively.  The ligand-protein interactions between the molecular models of compounds 26Z, 25Z, 22Z, and 27Z and the tubulin-colchicine:stathmin-like domain complex (PDB code: 1SA0 [54]) were investigated. The structural similarity of synthesized compounds with colchicine suggested docking in the colchicine site of the α,β-tubulin complex. The colchicine binding site lies on the border between the α and β subunits of tubulin; therefore, docking there could lead to inhibition of microtubule polymerization. Colchicine's mechanism of action is based on hydrogen bonding interactions between methoxy groups and Cys241 and Val181 amino acids [19]. However, for compounds 25Z, 22Z, and 27Z, only interactions with Cys241 are detected ( Figure 4). Crystals of compound 26Z suitable for a single crystal X-ray experiments were not obtained. The molecular model of 26Z was software generated [55] based on the crystal structure of 25Z and the benzoxazolone analog 4 (OP-107) [19], where the difference is an O atom instead of an S atom in the benzoxazolone/benzothiazole moiety. The generated molecular structure of 26Z, docked in the colchicine site, displayed interactions with The ligand-protein interactions between the molecular models of compounds 26Z, 25Z, 22Z, and 27Z and the tubulin-colchicine:stathmin-like domain complex (PDB code: 1SA0 [54]) were investigated. The structural similarity of synthesized compounds with colchicine suggested docking in the colchicine site of the α,β-tubulin complex. The colchicine binding site lies on the border between the α and β subunits of tubulin; therefore, docking there could lead to inhibition of microtubule polymerization. Colchicine's mechanism of action is based on hydrogen bonding interactions between methoxy groups and Cys241 and Val181 amino acids [19]. However, for compounds 25Z, 22Z, and 27Z, only interactions with Cys241 are detected ( Figure 4).  The ligand-protein interactions between the molecular models of compounds 26Z, 25Z, 22Z, and 27Z and the tubulin-colchicine:stathmin-like domain complex (PDB code: 1SA0 [54]) were investigated. The structural similarity of synthesized compounds with colchicine suggested docking in the colchicine site of the α,β-tubulin complex. The colchicine binding site lies on the border between the α and β subunits of tubulin; therefore, docking there could lead to inhibition of microtubule polymerization. Colchicine's mechanism of action is based on hydrogen bonding interactions between methoxy groups and Cys241 and Val181 amino acids [19]. However, for compounds 25Z, 22Z, and 27Z, only interactions with Cys241 are detected ( Figure 4). Crystals of compound 26Z suitable for a single crystal X-ray experiments were not obtained. The molecular model of 26Z was software generated [55] based on the crystal structure of 25Z and the benzoxazolone analog 4 (OP-107) [19], where the difference is an O atom instead of an S atom in the benzoxazolone/benzothiazole moiety. The generated molecular structure of 26Z, docked in the colchicine site, displayed interactions with Crystals of compound 26Z suitable for a single crystal X-ray experiments were not obtained. The molecular model of 26Z was software generated [55] based on the crystal structure of 25Z and the benzoxazolone analog 4 (OP-107) [19], where the difference is an O atom instead of an S atom in the benzoxazolone/benzothiazole moiety. The generated molecular structure of 26Z, docked in the colchicine site, displayed interactions with Cys241 The cytotoxic effect of benzothiazolone cis/trans-CA-4 analogs was evaluated using the MTT assay in the endothelial EA.hy926 cell line, since it was shown that the antitumor activity of combretastatin is mainly dependent on its anti-vascular activity [56]. After 72 h of treatment, Z-benzothiazolone analogs inhibited cell growth and showed cytotoxic activity in a concentration-dependent manner. E-analogs did not show significant activity at concentrations up to 250 µM, except for 26E. The results are shown in Table 2. All 13 cis-compounds showed growth inhibitory activity against EA.hy926 cells, with IC 50 values between 0.13 and 26 µM. 26Z exhibited the most potent cytotoxic activity with an IC 50 of 0.13 ± 0.01 µM, which was three to hundreds of times lower than the other analogs. The 26E trans-form of 26Z was also active at higher concentrations against endothelial cells with IC 50 = 38.20 ± 5.35 µM. It is well known that the affinity of E-isomer of CA-4 to α,β-tubulin is significantly less potent [57,58]. In general, cis-isomers gave much stronger antiproliferative activities over their trans-isomers, and our data confirmed it, since only the E-isomer of the most active compound (26Z) showed some activity.
The second most active compound was 27Z, with IC 50 = 0.30 ± 0.05 µM. The third most potent compound was 25Z, with an IC 50 = 1.88 ± 0.18 µM. The cytotoxic activity of the tested Z-benzothiazolones can be ranked as follows: The 26Z positional isomers 27Z, 22Z, and 18Z demonstrated loss of activity, suggesting that the position of the styryl fragment in ring B plays an essential role in the analogs' bioactivity. Amongst the most active compounds, three (22Z, 26Z, 27Z) out of four have three methoxy groups in ring A. The combination of a styryl fragment at position 6 in ring B with three OCH 3 groups in ring A showed the most prominent cytotoxic effect on EA.hy926 cells. The position of the styryl fragment in ring B relative to the compounds' activity can be ranked as 6 > 7 > 5 > 4 and is critical for the observed cytotoxic activity.
The docking analysis also confirmed the importance of methoxy groups in ring A together with the position of the styryl fragment in ring B. Compound 26Z shows a hydrogen bonding interaction with Cys241 and an additional interaction with Lys352, which could explain the greater cytotoxic effect of 26Z against EA.hy926 cells.
In A549 cancer cells, a cytotoxic effect with IC 50 = 3.01 ± 0.16 µM was observed for CA-4, while 26Z showed cytostatic activity with 45% cell viability at concentrations from 10 nM to 25 µM (Table S3.2). In control BEAS-2B cells, the cytostatic activity was around 80% (viable cells) for both CA-4 and 26Z. The greater cytotoxic activity of 26Z against A549 cells compared to CA-4 and the low toxicity in BEAS-2B control cells suggest a possible functional advantage of 26Z against lung cancer cells.
In MCF-7 breast cancer cells, we observed a cytotoxic effect with IC 50 = 2.42 ± 0.48 µM for 26Z, where CA-4 showed a cytostatic impact (40% viable cells) in the concentration range studied. Compound 26Z has IC 50  Based on the observed cytotoxic effects, we have selected the three most active compounds (26Z, 27Z, and 25Z) to conduct further studies and gain insights into the cellular and molecular mechanisms of 26Z's action on EA.hy926 endothelial cells.

25Z, 26Z, and 27Z Decrease the Clonogenic Survival of Endothelial Cells
The clonogenic assay has become the "gold standard" for assessing cellular sensitivity to cytotoxic treatments, as the ability to proliferate and form a colony is a fundamental aspect of cell survival. Cell survival assays measure the end result of treatment, which can be either cell death or survival [62]. To determine whether exposure to increasing 25Z, 26Z, and 27Z concentrations (0.01, 0.25, 2.5, 10, and 25 µM) affects the proliferation ability and colony-forming capacity of EA.hy926 cells, we evaluated the late lethality, following treatment for 72 h and 6 days of recovery.
The results indicate that 25Z, 26Z, and 27Z, as well as CA-4, significantly inhibit the proliferation and colony formation capabilities of EA.hy926 cells and show persistent effects at concentrations well below the IC 50 , as determined by MTT assay ( Figure 5). Semiquantitative analysis showed a strong reduction (about 100%) in colony formation following treatment with 2.5, 10, and 25 µM concentrations of 25Z, 26Z, and 27Z. EA.hy926 cells were more sensitive towards 26Z, 27Z, and 25Z exposure at the lowest dose of 10 nM. Moreover, a 90% decrease in colony number was observed at 250 nM 26Z, with decreases of 75% and 55% for 27Z and 25Z, respectively ( Figure 5B). The effect of CA-4 was more pronounced compared to 26Z, 27Z, and 25Z. CA-4 at 1 nM inhibited endothelial cell colony formation by 80%.
The results also showed that many cells treated with CA-4 analogs at concentrations lower than the IC 50 , as determined by the MTT assay, could not continue dividing and forming colonies when the compounds were removed and the cells were allowed to grow in a fresh medium for six days. We found that EA.hy926 cells were more sensitive to CA-4 and 26Z compared to 25Z and 27Z.
To investigate the precise mechanism responsible for the 26Z-mediated anti-proliferative effects, we examined the cell-cycle distribution profile of 26Z-treated EA.hy926 cells over time. The results also showed that many cells treated with CA-4 analogs at concentrations lower than the IC50, as determined by the MTT assay, could not continue dividing and forming colonies when the compounds were removed and the cells were allowed to grow in a fresh medium for six days. We found that EA.hy926 cells were more sensitive to CA-4 and 26Z compared to 25Z and 27Z.
To investigate the precise mechanism responsible for the 26Z-mediated anti-proliferative effects, we examined the cell-cycle distribution profile of 26Z-treated EA.hy926 cells over time.  (6) 0.01 nM of CA-4 as a positive control. Cells were allowed to form colonies in fresh medium for 6 days after treatment. (B) Number of colonies formed on day 6 after treatment normalized to control. The presented data were obtained by manual counting and expressed as mean ± SD (n = 3). * p < 0.05; ** p < 0.01; ***p < 0.001.

26Z Causes Mitotic Arrest in Treated EA.hy926 Cells
In EA.hy926 cells, 26Z treatment caused a significant inhibition of cell-cycle progression, resulting in an accumulation of treated cells in the G2/M phase compared to control cells ( Figure 6, Figure S3.2). This effect is similar to the one observed for CA-4. The G2/M population began to rise as early as 4 h post-treatment (~35%) and reached its maximum at around 24 h (~78%), and was still above 50% at 48 h and 72 h. At 48 h and 72 h, G2/M arrested cells decreased, mainly because of increased apoptotic hypodiploid DNA content peak (sub-G1) and polyploid cells. The observed cell cycle distribution shows that cell death is a direct consequence of mitotic arrest, although some cells can overpass it (polyploid > 4N cells). population began to rise as early as 4 h post-treatment (~35%) and reached its maximum at around 24 h (~78%), and was still above 50% at 48 h and 72 h. At 48 h and 72 h, G2/M arrested cells decreased, mainly because of increased apoptotic hypodiploid DNA content peak (sub-G1) and polyploid cells. The observed cell cycle distribution shows that cell death is a direct consequence of mitotic arrest, although some cells can overpass it (polyploid > 4N cells). Tubulin-binding agents cause a cell cycle blockade due to disturbances in microtubule dynamics. Previous studies on CA-4 activity have confirmed that it causes G2/M phase mitotic arrest in treated cells, as the cells are unable to complete the process of cell division [63][64][65].
Due to the structural similarity between our compounds and CA-4, which target the colchicine-binding site of tubulin, further studies were performed to investigate the specificity of 26Z interaction at the colchicine binding site of tubulin using N,N-ethylenebis(iodoacetamide) (EBI) in EA.hy926 cells. EBI forms adducts with tubulin which can be detected Tubulin-binding agents cause a cell cycle blockade due to disturbances in microtubule dynamics. Previous studies on CA-4 activity have confirmed that it causes G2/M phase mitotic arrest in treated cells, as the cells are unable to complete the process of cell division [63][64][65].
Due to the structural similarity between our compounds and CA-4, which target the colchicine-binding site of tubulin, further studies were performed to investigate the specificity of 26Z interaction at the colchicine binding site of tubulin using N,Nethylenebis(iodoacetamide) (EBI) in EA.hy926 cells. EBI forms adducts with tubulin which can be detected by Western blotting. Microtubule destabilising agents that bind at the colchicine site, such as colchicine and CA-4, prevent the formation of the β-tubulin-EBI adduct by occupying the binding site [66,67]. EA.hy926 cells were treated with vehicle control (10 µM DMSO), 10 µM colchicine, 10 µM CA-4, 25 µM vinblastine, or different concentrations of 26Z for 2 h, followed by EBI (100 µM) for an additional 2 h. Control samples show the presence of the β-tubulin-EBI adduct at a lower position, indicating that EBI has crosslinked Cys239 and Cys354 amino acids on the β-tubulin ( Figure 6B). In contrast, tubulin EBI-adduct formation was inhibited in EA.hy926 cells treated with 10 µM CA-4, 10 µM colchicine, and 25 µM 26Z. Figure 6B also shows that binding of 26Z to the tubulin is concentration dependent, since with increasing the 26Z concentration, the EBI-adduct formation was decreased. These data proved that 26Z binds specifically to the colchicine site of β-tubulin.
Cancer cell migration and motility are critical factors in tumor progression and metastasis, and the endothelial cells recruited to the tumor undergo rapid cell migration, proliferation, and differentiation. Endothelial cell migration is of crucial importance for angiogenesis. Based on the results described above, we hypothesized that EA.hy926 cells will have impaired cell motility. A wound-healing assay was performed to investigate the effect of compounds 25Z, 26Z, and 27Z on EA.hy926 cell migration.

In Vitro Cell Migration Assay
As illustrated in Figure 7, the control cells (0.01 µM DMSO) migrated to the scraped area while 25Z, 26Z, and 27Z significantly inhibited cell migration at sub-toxic doses of 50 nM, and 1 nM for CA-4.
binding site [66,67]. EA.hy926 cells were treated with vehicle control (10 µM DMSO), 10 µM colchicine, 10 µM CA-4, 25 µM vinblastine, or different concentrations of 26Z for 2 h, followed by EBI (100 µM) for an additional 2 h. Control samples show the presence of the βtubulin-EBI adduct at a lower position, indicating that EBI has crosslinked Cys239 and Cys354 amino acids on the β-tubulin ( Figure 6B). In contrast, tubulin EBI-adduct formation was inhibited in EA.hy926 cells treated with 10 µM CA-4, 10 µM colchicine, and 25 µM 26Z. Figure 6B also shows that binding of 26Z to the tubulin is concentration dependent, since with increasing the 26Z concentration, the EBI-adduct formation was decreased. These data proved that 26Z binds specifically to the colchicine site of β-tubulin.
Cancer cell migration and motility are critical factors in tumor progression and metastasis, and the endothelial cells recruited to the tumor undergo rapid cell migration, proliferation, and differentiation. Endothelial cell migration is of crucial importance for angiogenesis. Based on the results described above, we hypothesized that EA.hy926 cells will have impaired cell motility. A wound-healing assay was performed to investigate the effect of compounds 25Z, 26Z, and 27Z on EA.hy926 cell migration.

In Vitro Cell Migration Assay
As illustrated in Figure 7  These doses were chosen to avoid the effect of toxicity on cell migration. The remaining mean wound opened areas in the treated cell monolayers are as follows: CA-4 = 24.4% ± 4, 25Z = 22.4% ± 8, 26Z = 49% ± 6, and 27Z = 3% ± 1. The leading compound 26Z is the most anti-mitogenic after 24 h of treatment (Figure 7). At low concentrations, 26Z can significantly impair (p < 0.001) endothelial cell migration. This significant difference in repairing the wounded area in the cell monolayers confirms that 25Z and 26Z suppressed endothelial cell migration.
Beyond the migrating front, endothelial cells usually proliferate to generate the necessary number of cells for making new blood vessels. Combretastatins are vascular disrupting agents, and their ability to selectively disrupt already established tumor blood vessels is critical for their in vivo antitumor activity. To further investigate the effects of 25Z, 26Z, and 27Z on neovascularization, we evaluated the effect of CA-4 analogs on tubular morphogenesis. These doses were chosen to avoid the effect of toxicity on cell migration. The remaining mean wound opened areas in the treated cell monolayers are as follows: CA-4 = 24.4% ± 4, 25Z = 22.4% ± 8, 26Z = 49% ± 6, and 27Z = 3% ± 1. The leading compound 26Z is the most anti-mitogenic after 24 h of treatment (Figure 7). At low concentrations, 26Z can significantly impair (p < 0.001) endothelial cell migration. This significant difference in repairing the wounded area in the cell monolayers confirms that 25Z and 26Z suppressed endothelial cell migration.
Beyond the migrating front, endothelial cells usually proliferate to generate the necessary number of cells for making new blood vessels. Combretastatins are vascular disrupting agents, and their ability to selectively disrupt already established tumor blood vessels is critical for their in vivo antitumor activity. To further investigate the effects of 25Z, 26Z, and 27Z on neovascularization, we evaluated the effect of CA-4 analogs on tubular morphogenesis.

In Vitro Anti-Vascular Activity of Benzothiazolone CA-4 Analogs
Upon incubation for 24 h on the extracellular matrix model surface, EA.hy926 cells fuse into continuous tubules with a complete lumen to form capillary-like networks (Figure 8, time 0). After 24 h of treatment, IC 50 concentrations of 26Z and 27Z destroy formed tubes with a significant reduction of~70% in cumulative tube length, compared with length at time 0 ( Figure 8A,B). Dimensional parameters such as tube length reflect cells' ability to maintain an elongated form and are vital for in vitro vasculogenesis [68]. At the same time, topological parameters such as branch points and loop number contain important physiological information concerning the overall structural arrangement and complexity of the capillary-like tubular network [69]. Figure 8B shows the disruption of established tubes and branches, which is a good indication of a potential anti-vascular effect of the test compounds in vivo.
The results allowed us to conclude that the CA-4 analogs 25Z, 26Z, and 27Z affect endothelial capillary tubule morphogenesis in vitro. Taken together, all results shown above indicated that 26Z and 27Z directly affected EA.hy926 cells by robustly inhibiting invasion, migration, proliferation, and capillary-tube formation, which are essential attributes of potential anti-angiogenic drug candidates [70].
The next step of our study was to identify the cellular basis of the proliferation, migration, and invasion defects observed upon CA-4 analog treatment.  The results allowed us to conclude that the CA-4 analogs 25Z, 26Z, and 27Z affect endothelial capillary tubule morphogenesis in vitro. Taken together, all results shown above indicated that 26Z and 27Z directly affected EA.hy926 cells by robustly inhibiting invasion, migration, proliferation, and capillary-tube formation, which are essential attributes of potential anti-angiogenic drug candidates [70].
The next step of our study was to identify the cellular basis of the proliferation, migration, and invasion defects observed upon CA-4 analog treatment. metaphase plate (ii) followed by anaphase onset (iii), telophase, and cytokinesis are evident in Figure 9 (panel A, Control). 0.05; ** p < 0.01 *** p < 0.001 compared to 0 h. Figure 9 shows the immunofluorescence micrographs of dividing EA.hy926 cells treated with vehicle (0.01 µM DMSO, Control) and 26Z. Interphase cells (i) showed typical radial arrays of microtubules ( Figure 9A, Control, green fluorescence). The mitotic population (~25%, Figure 6) of the total number of cells in a G2/M cell-cycle phase displayed the hallmark features of a typical mitotic process. Congression of chromosomes at the metaphase plate (ii) followed by anaphase onset (iii), telophase, and cytokinesis are evident in Figure 9 (panel A, Control). Following 24 h of treatment with 26Z, large proportions of the EA.hy926 cells showed mitotic abnormalities. Typical mitotic defects included a failure of a number of chromosomes to align correctly on the metaphase plate, and the absence of two bipolar spindles with the centromeres of individual chromosomes randomly attached to either of the spindle poles (Figure 9, panel A, 26Z). Compound 26Z triggered conspicuous spindle multipolarity with a striking declustering of supernumerary centrosomes and/or apoptosis (130 nM and 300 nM). Treatment with IC50 (130 nM) 26Z for 24 h induced multipolarity, predominantly Following 24 h of treatment with 26Z, large proportions of the EA.hy926 cells showed mitotic abnormalities. Typical mitotic defects included a failure of a number of chromosomes to align correctly on the metaphase plate, and the absence of two bipolar spindles with the centromeres of individual chromosomes randomly attached to either of the spindle poles (Figure 9, panel A, 26Z). Compound 26Z triggered conspicuous spindle multipolarity with a striking declustering of supernumerary centrosomes and/or apoptosis (130 nM and 300 nM). Treatment with IC 50 (130 nM) 26Z for 24 h induced multipolarity, predominantly with three to five poles (3.9 ± 0.99), while IC 80 (300 nM) increased the number of poles, ranging between six and twelve (8.8 ± 2.3).

Aberrant Formation of Mitotic Spindles and Microtubule Network Alterations
Multinucleated interphase cells in treated cultures were also visible ( Figure 9B). Some of these cells displayed typical radial arrays of microtubules at 130 and 300 nM 26Z (Figure 9, panel  B, 26Z). Some mitotically arrested cells somehow slipped out of mitosis at 24 h and appeared as conspicuously large multinucleated cells. Most of these cells were polyploid (>4N) and had exited mitosis, probably by mitotic slippage, without proceeding through normal anaphase and cytokinesis. In adherent cells treated with 26Z at IC 50 , there were remains of the microtubule network after 24 h, but it was more disorganized, lacking the "hair-like" structural organization typical of control cells ( Figure 9B). While most control cells displayed a well-formed microtubule network, with microtubules going out from the center towards the cell's periphery, cells treated with 26Z looked more rounded. The microtubule structure was fragmented and loosely organized. There were cells with multi-lobed, segmented nuclei, and apoptotic cells with membrane blebbing and formation of membrane-bound apoptotic bodies (Figure 9, panel B, 26Z).

26Z Acts as Polymerization Inhibitor in Ex Vivo Conditions
For a better and more specific characterization of the effect of 26Z on tubulin polymerization, we performed an ex vivo polymerization experiment. Paclitaxel, CA-4, and 26Z at a concentration of 20 µM were incubated with tubulin for 45 min at 37 • C, and the obtained microtubules were sedimented and analyzed, as described in the Methods section. The results showed that 26Z acts as a tubulin polymerization inhibitor, although the effect is weaker than that of CA-4. Western blotting results showed that 26Z, similarly to CA-4, inhibits the process of tubulin polymerization, demonstrated by an increase in the soluble fraction and a decrease in the polymeric fraction of tubulin ( Figure 10A,B). On the contrary, paclitaxel, which was used as a positive control, caused a reduction in the soluble and an increase in the polymeric tubulin fraction.
Multinucleated interphase cells in treated cultures were also visible ( Figure 9B). Some of these cells displayed typical radial arrays of microtubules at 130 and 300 nM 26Z (Figure 9, panel B, 26Z). Some mitotically arrested cells somehow slipped out of mitosis at 24 h and appeared as conspicuously large multinucleated cells. Most of these cells were polyploid (>4N) and had exited mitosis, probably by mitotic slippage, without proceeding through normal anaphase and cytokinesis. In adherent cells treated with 26Z at IC50, there were remains of the microtubule network after 24 h, but it was more disorganized, lacking the "hair-like" structural organization typical of control cells ( Figure 9B). While most control cells displayed a well-formed microtubule network, with microtubules going out from the center towards the cell's periphery, cells treated with 26Z looked more rounded. The microtubule structure was fragmented and loosely organized. There were cells with multi-lobed, segmented nuclei, and apoptotic cells with membrane blebbing and formation of membrane-bound apoptotic bodies (Figure 9, panel B, 26Z).

26Z Acts as Polymerization Inhibitor in Ex Vivo Conditions
For a better and more specific characterization of the effect of 26Z on tubulin polymerization, we performed an ex vivo polymerization experiment. Paclitaxel, CA-4, and 26Z at a concentration of 20 µM were incubated with tubulin for 45 min at 37 °C, and the obtained microtubules were sedimented and analyzed, as described in the Methods section. The results showed that 26Z acts as a tubulin polymerization inhibitor, although the effect is weaker than that of CA-4. Western blotting results showed that 26Z, similarly to CA-4, inhibits the process of tubulin polymerization, demonstrated by an increase in the soluble fraction and a decrease in the polymeric fraction of tubulin ( Figure 10A,B). On the contrary, paclitaxel, which was used as a positive control, caused a reduction in the soluble and an increase in the polymeric tubulin fraction. These results confirm that 26Z acts as an inhibitor of tubulin polymerization ex vivo. This was further corroborated by fluorescence microscopy analysis of rhodamine-labeled tubulin fiber formation in the presence of the tested compounds. Fluorescently labeled porcine tubulin together with unlabeled tubulin in a 1:3 ratio formed fluorescent fibers upon polymerization. The addition of CA-4 and 26Z blocked the process of fiber formation compared to the control and paclitaxel (PTX) samples ( Figure 11). The samples treated with 26Z and CA-4 also contained visible non-polymerized tubulin aggregates. These results confirm that 26Z acts as an inhibitor of tubulin polymerization ex vivo. This was further corroborated by fluorescence microscopy analysis of rhodamine-labeled tubulin fiber formation in the presence of the tested compounds. Fluorescently labeled porcine tubulin together with unlabeled tubulin in a 1:3 ratio formed fluorescent fibers upon polymerization. The addition of CA-4 and 26Z blocked the process of fiber formation compared to the control and paclitaxel (PTX) samples ( Figure 11). The samples treated with 26Z and CA-4 also contained visible non-polymerized tubulin aggregates. Consistent with previous results, paclitaxel addition promoted the formation of microtubule fibers without visible aggregates of non-polymerized tubulin.   CA-4, and 26Z). Scale bar, 10 µm.

Benzothiazolones Depolymerizing Activity in EA.hy926 Cells
To confirm the previous observations that 26Z can inhibit tubulin polymerization ex vivo, we performed an in vitro experiment following treatment of EA.hy926 cells with 0.01 µM DMSO and 1 µM of CA-4, paclitaxel, 26Z, and 27Z for 6 h. Extraction of the soluble (S) and polymeric (P) tubulin fractions from EA.hy926 cells was performed as described in the Materials and Methods (Section 3.3.8). Following electrophoretic separation (Figure 12A), the soluble vs. polymerized tubulin ratio was determined. Figure 11. Fluorescence microscopy of ex vivo polymerized rhodamine-labeled α,β-tubulin. Representative images of formed microtubule fibers in control (DMSO 0.01 µM) and paclitaxel (PTX, 20 µM) samples, and inhibition of the process upon treatment with 20 µM CA-4 and 26Z. Aggregates of non-polymerized tubulin are also visible as points with very bright fluorescence (Control, CA-4, and 26Z). Scale bar, 10 µm.

Benzothiazolones Depolymerizing Activity in EA.hy926 Cells
To confirm the previous observations that 26Z can inhibit tubulin polymerization ex vivo, we performed an in vitro experiment following treatment of EA.hy926 cells with 0.01 µM DMSO and 1 µM of CA-4, paclitaxel, 26Z, and 27Z for 6 h. Extraction of the soluble (S) and polymeric (P) tubulin fractions from EA.hy926 cells was performed as described in the Materials and Methods (Section 3.3.8.). Following electrophoretic separation ( Figure 12A), the soluble vs. polymerized tubulin ratio was determined. The results showed ratios of 2.33 ± 0.14 in DMSO control, 0.51 ± 0.19 in paclitaxel treated cells, 3.3 ± 0.61 for CA-4, 2.74 ± 0.43 for 26Z, and 1.82 ± 0.36 for 27Z treated cells. Compared to CA-4, 26Z showed milder inhibitory activity after 6 h. The opposite effect The results showed ratios of 2.33 ± 0.14 in DMSO control, 0.51 ± 0.19 in paclitaxel treated cells, 3.3 ± 0.61 for CA-4, 2.74 ± 0.43 for 26Z, and 1.82 ± 0.36 for 27Z treated cells. Compared to CA-4, 26Z showed milder inhibitory activity after 6 h. The opposite effect was observed in cells treated with paclitaxel ( Figure 12B). The depolymerizing activity of 26Z is concentration-dependent, and the effect is clearly visible at concentrations higher than 0.05 µM ( Figure 12C). Compared to the control DMSO-treated cells, 27Z did not show a significant difference in α,β-tubulin polymerization. These results confirm the inhibitory action of 26Z on α,β-tubulin polymerization in cells.

26Z Activates Apoptotic Signaling Pathways in EA.hy926 Cells
The cell fate and pathways for initiation and execution of cell death can differ, ranging from apoptosis, necrosis, mitotic catastrophe, mitotic slippage, etc. Apoptosis, or programmed cell death, is a precisely regulated intracellular process leading to cell death without the induction of inflammation. Induction of apoptosis in proliferating endothelial cells and apoptotic changes in cell morphology are particularly important for disruption of tumor vasculature and blood flow shutdown. We observed the presence of apoptotic cells following treatment with 26Z ( Figure 6 (sub-G1) and Figure 9). To clarify the mechanism of cell death, we studied the activation of caspase signal-transduction pathways by immunoblot analysis of caspase 3, 8, and 9 protein levels. Active forms of effector caspase 3 were detected in cells treated with 26Z at IC 50 and IC 80 (0.30 ± 0.04 µM) and, to a lesser extent, in samples treated with CA-4 IC 50 and 26Z IC 20 (0.04 ± 0.01 µM, Figure 13). There was also observable proteolytic fragmentation of PARP, the downstream substrate of caspase 3, in all EA.hy926 treated cells. Our results showed that caspases 8 and 9 were also activated by 26Z treatment. The activation of initiator caspase 8 was more pronounced in cells treated with 26Z IC 80 , while caspase 9 was active in all endothelial cells treated with CA-4 and different concentrations of 26Z (Figure 13). of tumor vasculature and blood flow shutdown. We observed the presence of apoptotic cells following treatment with 26Z (Figures 6 (sub-G1) and 9). To clarify the mechanism of cell death, we studied the activation of caspase signal-transduction pathways by immunoblot analysis of caspase 3, 8, and 9 protein levels. Active forms of effector caspase 3 were detected in cells treated with 26Z at IC50 and IC80 (0.30 ± 0.04 µM) and, to a lesser extent, in samples treated with CA-4 IC50 and 26Z IC20 (0.04 ± 0.01 µM, Figure 13). There was also observable proteolytic fragmentation of PARP, the downstream substrate of caspase 3, in all EA.hy926 treated cells. Our results showed that caspases 8 and 9 were also activated by 26Z treatment. The activation of initiator caspase 8 was more pronounced in cells treated with 26Z IC80, while caspase 9 was active in all endothelial cells treated with CA-4 and different concentrations of 26Z (Figure 13).
Activating all critical components of the apoptotic cascade confirms that 26Z can induce apoptosis in EA.hy926 cells through extrinsic and intrinsic pathways.

General Information
Melting points (m.p.) were determined on a Boetius hot-stage microscope. 1 H and 13 C NMR spectra were obtained with a Bruker DRX250, Bruker DRX400, and DRX 500 spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) in CDCl3 or acetone-d6 as solvent. Chemical shifts were reported in parts per million (ppm, δ) relative to the solvent peak (7.26 ppm for 1 H; 77.16 ppm for 13 C). Coupling constants (J) were measured in Hertz Activating all critical components of the apoptotic cascade confirms that 26Z can induce apoptosis in EA.hy926 cells through extrinsic and intrinsic pathways.

General Information
Melting points (m.p.) were determined on a Boetius hot-stage microscope. 1 H and 13 C NMR spectra were obtained with a Bruker DRX250, Bruker DRX400, and DRX 500 spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) in CDCl 3 or acetone-d 6 as solvent. Chemical shifts were reported in parts per million (ppm, δ) relative to the solvent peak (7.26 ppm for 1 H; 77.16 ppm for 13 C). Coupling constants (J) were measured in Hertz (Hz). Elemental analyses (C, H, N) were carried out by a Vario III microanalyzer. The results obtained were within 0.4% of theoretical values. Thin-layer chromatography (TLC) was carried out on silica gel plates (Kieselgel 60 F254). Flash column chromatography was performed with Merck 60 silica gel (0.040-0.063 mm).

4-Bromomethyl-3-Methyl-2(3H)-Benzothiazolone (34)
The compound was prepared following the general procedure Section 3.1.3. for bromination. The product was purified by recrystallization from carbon tetrachloride. To a stirred solution of the corresponding phosphonium salt 11-14 (3 mmol) in THF/CH 2 Cl 2 (15 mL, 2:1 v/v), powdered K 2 CO 3 (10 mmol) and 18-crown-6 (0.01 g) were added, followed by the corresponding aldehyde 5-10 (3 mmol). The reaction mixture was refluxed for 4-6 h (monitored by TLC). After cooling, the inorganic salts were filtered off and washed with CH 2 Cl 2 . The filtrate was concentrated under reduced pressure and the mixture of the corresponding E/Z-stilbenes and triphenylphosphine oxide were isolated by flash column chromatography on silica gel using petroleum ether-AcOEt (9:1). The Z-stilbenes were eluted first, followed by the E-isomers. (15) Following the general procedure Section 3.1.6, diastereomers 15Z and 15E were obtained by reaction of phosphonium salt 11 and 4-methoxybenzaldehyde (5 Table S3.1. Cell cultures were maintained in an incubator at 37 • C with 95% humidity and 5% CO 2 and were routinely found free of mycoplasma.

Cellular Treatment
All tested compounds were dissolved in DMSO at 25 mM initial concentration and diluted with medium to obtain the working concentrations (0.01 nM-250 µM). Serial drug dilutions were prepared in medium immediately before each experiment. The DMSO working concentration in the medium was 0.01 µM. For the MTT assay, cells were seeded into 96-well microtitre plates at a density of 1 × 10 4 in 100 µL complete medium per well. They were further incubated for 24 h, followed by treatment with CA-4 analogs up to 72 h. Cell viability was determined using the MTT assay [78]. Ten µL of MTT solution (5 mg/mL) was added to each well and further incubated for 3 h at 37 • C. To dissolve the formazan product, 100 µL/well of 99.5% isopropanol was added. Cell viability was determined by measuring absorbance at 550/630 nm using a DTX880 plate reader (Beckman Coulter, USA). Results, expressed as concentrations that inhibit 20% (IC 20 ), 50% (IC 50 ), and 80% (IC 80 ) of cell growth, were calculated using GraphPad Prism 8.0.1.

Clonogenic Survival Assay
EA.hy926 cells were seeded at 300 cells/well on a 6-well plate and incubated for 24 h. The cells were treated with different concentrations of test compounds for 72 h. After six additional days of growth in a fresh medium to form visible colonies, cells were washed with phosphate-buffered saline (PBS, pH 7.4) and stained with crystal violet (0.5%, w/v) in 6% glutaraldehyde. After 30 min of incubation, the plates were washed with tap water, dried, photographed, and analyzed using ImageJ [79]. Raw data were normalized to the corresponding DMSO control for each independent experiment. After normalization, statistically significant differences for colony-forming efficiency (CFE) values versus 0.01 µM DMSO were calculated.

Western Blot Analysis
A total of 7 × 10 5 cells was seeded in 100-mm dishes. On the next day, cells were treated with the corresponding concentrations of CA-4 and 26Z. After 24 h, treated cells were collected and lysed with lysis buffer (50 mM HEPES, pH 7.7, 250 mM KCl, 10% glycerol, 0.1% Nonidet P-40) containing protease inhibitors (0.4 mM sodium vanadate, 0.5 mM PMSF, 1 mM DTT, 2 mg/mL leupeptin, and 2 mg/mL pepstatin) by incubation on ice for 30 min and overnight at −20 • C [80]. The concentration of obtained total cell protein lysates was determined by the method of Bradford [81]. Samples were separated on 12% polyacrylamide gels and transferred to nitrocellulose or PVDF membranes. Immunoblotting was performed by blocking in 3.5% non-fat dried milk in PBS, with subsequent incubation for 1 h with primary antibodies for caspase 3 (sc-56053, Santa Cruz Biotechnology), caspase 8 (sc-81656, Santa Cruz Biotechnology), caspase 9 (sc-8355, Santa Cruz Biotechnology), and PARP-1 (sc-8007, Santa Cruz Biotechnology), diluted in PBS containing 0.05% Tween-20. This was followed by incubation with a peroxidase-conjugated secondary antibody m-Igk BP-HRP (sc-516102, Santa Cruz Biotechnology) or a donkey anti-rabbit IgG-HRP (sc-2313, Santa Cruz Biotechnology) for 1 h, and detection of the chemiluminescent signal was accomplished with ECL Select TM Western Blotting Detection Reagent (Amersham) on X-ray film. β-actin was used as a loading control and mouse monoclonal antibody was used following the manufacturer's instruction (sc-47778, Santa Cruz Biotechnology).

In Vitro Cell Migration Assays
A total of 3 × 10 5 EA.hy926 cells were seeded in a 12-well plate and grown close to the confluence. The cell monolayer was gently scraped with a P200 pipette tip, creating a wound area around 1-mm wide. Cells were washed twice with fresh medium and treated with test compounds diluted in growth media to the corresponding sub-toxic concentrations. Images were acquired at 0 and 24 h following treatment. After incubation for 24 h, the scratch wound width was measured and recorded, and then compared with the initial scratch wound width at 0 h. Using the ImageJ image processing program (ImageJ150-WIN-Java8), the size of the denuded area in pixels was determined at each time point from the digital images.

Matrigel Tube Disruption Assay
The tube formation assay was performed in 96-well plates coated with 40 µL BD Matrigel™ (basement membrane matrix), following the manufacturer's instruction (BD Biosciences). A total of 2 × 10 4 EA.hy926 cells was re-suspended in growth medium and added to each well in a final volume of 100 µL. Tube formation was visualized directly using a light microscope. Images were captured using a phase-contrast microscope and quantitation was performed using the WimTube application of the Wimasis online image analysis platform (https://www.wimasis.com/en/products/13/WimTube; accessed on 1 June 2021).

Ex Vivo Tubulin Polymerization Assays Ex Vivo Unlabeled Tubulin Polymerization
α,β-tubulin powder (T240, Cytoskeleton) was diluted in PEM buffer (80 mM PIPES, 2 mM MgCl 2 , 0.5 mM EGTA, pH 6.9) on ice to a final concentration of 4 mg/mL. Five µL of α,β-tubulin was mixed with 0.75 µL 50% glycerol, and 0.7 µL 200 µM solution of the substances studied. Polymerization was started by adding 0.55 µL 10 mM GTP, and the samples were incubated at 35 • C for 45 min, followed by centrifugation at 13,000 × g for 10 min. The supernatant was separated and mixed with 5× sample buffer (5% SDS, 45% glycerol, 12.5% β-mercaptoethanol, and 0.05% bromophenol blue in 200 mM Tris-HCl, pH 6.8). The pellet was re-suspended in 1× sample buffer followed by separation on 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Western blot analysis was performed as described in Section 3.3.4 with the primary anti-β-tubulin antibody (sc-101527, Santa Cruz Biotechnology) diluted to 1:1000, followed by incubation with a secondary donkey anti-mouse 680 antibody (LI-COR Biosciences) and imaging with an Odyssey Imaging System (LI-COR Biosciences). The fluorescence intensity was analysed with Image Studio Lite v. 5.2.
Colchicine-Binding Site Assay N,N -ethylene-bis(iodoacetamide) (abcam, ab144980) was dissolved in DMSO to a final concentration of 50 mM. EA.hy926 cells were seeded at density of 1 × 10 5 cells/mL and incubated overnight. Cells were treated with vehicle control 10 µM DMSO, 10 µM colchicine, 10 µM CA-4, 25 µM vinblastine, or different concentrations of 26Z for 2 h, then 100 µM EBI was added and the cells were further incubated for another 2 h. Cells were lysed with Laemmli buffer and subjected to Western blot analysis for β-tubulin [66,67]. The detailed Western blot method is as shown above.
For the concentration-dependent assay, the cells were harvested by scraping and centrifuged at 1000 g for 5 min, and then lysed with the hypotonic cell lysis buffer for 10 min at room temperature [83]. Tubulin fractions were obtained by centrifugation at 12,000 × g for 10 min.
Equal volumes of the samples were subjected to 10% SDS-PAGE followed by western blotting. Detection of β-tubulin content was accomplished with an anti-β-tubulin mouse monoclonal antibody (sc-101527, Santa Cruz Biotechnology).

Cell Cycle Analysis
A total of 5 × 10 5 cells was treated with 26Z and CA-4 for time-response assays. Cells were collected by scraping, pelleted, washed with cold PBS (1×), and fixed overnight at 4 • C with 1 mL of ice-cold methanol-acetone (1:1), added dropwise to the pellet while vortexing. Pellets were resuspended and washed in 500 µL of PBS (1×), then permeabilized with a Triton X-100 (0.05-0.1%)/DNA extraction buffer (0.2 M Na 2 HPO 4 , 0.04 M citric acid, pH 7.8) in a 1:1 ratio for 15 min. After that, pelleted cells were treated with RNAse A (20 µg/mL) and labeled with 10 µg/mL of propidium iodide (Sigma-Aldrich) by incubation at 37 • C for at least 30 min, protected from light. DNA content was measured using a FACS Calibur flow cytometer (Beckman Dickenson, NJ, USA), and cell cycle analysis was performed with FlowJo software (version 8.7). At least 10,000 events were recorded for each sample measured in duplicate.

Immunocytochemistry
Indirect immunofluorescence was used to examine the effect of 26Z on α-tubulin in EA.hy 926 cells. A total of 1 × 10 4 cells grown on coverslips were washed with PBS and fixed with 3.7% buffered paraformaldehyde. Single fluorescence cell labeling was performed, as described by Apostolova et al. [84], with a primary chicken antibody against α-tubulin (ab89984, Abcam) for 2 h. Cells were washed three times with PBS for 5 min and incubated for 60 min with a goat anti-chicken secondary antibody labeled with AlexaFluor-488 (ab150173, Abcam). F-actin was detected using AlexaFluor-568 Phalloidin (Invitrogen, MA, USA). Following three washes with PBS and two in water, the slides were mounted in UltraCruz fluorescence mounting medium with DAPI (sc-24941, Santa Cruz Biotechnology, USA). Fluorescence microscopy was performed with a Carl Zeiss Axiovert 200 M Inverted Microscope and Andor Dragonfly 505 Confocal Microscope equipped with an iXon Ultra 888 EMCCD camera (Andor, Oxford Instruments, Belfast, Northern Irelands).

Statistical Analysis
The data were evaluated by analysis of variance (ANOVA) followed by Tukey's posthock test. Differences in the results at the level of p < 0.05 were considered statistically significant. The statistical analysis was carried out using the PASW 18.0 statistical software package (IBM) for Windows.

Conclusions
Here we describe the synthesis, characterization, and biological activities of a series of 26 cis-and trans-styryl-benzothiazolones. Our results showed that the number of methoxygroups in ring A and the position of the styryl fragment benzothiazolone heterocycle (4-, 5-, 6-or 7) were crucial for the biological activity of the obtained compounds 15Z-27Z. Among the reported CA-4 analogs, compound 26Z, (Z)-3-methyl-6-(3,4,5-trimethoxystyryl)-2(3H)benzothiazolone, exhibited the most potent anti-proliferative effect. We have demonstrated that its primary mechanism of action ex vivo is the inhibition of β-tubulin polymerization. In vitro activity is associated with changes in the soluble/polymerized α,β-tubulin ratio, thus favoring the soluble form present in the endothelial cells.
Additionally, the in silico studies demonstrated that these molecules could interact with tubulin in the areas of the colchicine binding site, suggesting a similar mechanism of action. The three most active compounds, 25Z, 26Z, and 27Z, induced destabilization of microtubules, cell cycle arrest, mitotic catastrophe, and cell death. We also showed that 26Z inhibits a spectrum of angiogenic events in EA.hy926 cells by interfering with endothelial cell invasion, migration, and neovascularization. Undoubtedly, future work is warranted to further explore the chemotherapeutic potential of these new CA-4 analogs.