Synthesis and Biological Evaluation of 2-Methyl-4,5-Disubstituted Oxazoles as a Novel Class of Highly Potent Antitubulin Agents

Antimitotic agents that interfere with microtubule formation are one of the major classes of cytotoxic drugs for cancer treatment. Multiple 2-methyl-4-(3′,4′,5′-trimethoxyphenyl)-5-substituted oxazoles and their related 4-substituted-5-(3′,4′,5′-trimethoxyphenyl) regioisomeric derivatives designed as cis-constrained combretastatin A-4 (CA-4) analogues were synthesized and evaluated for their antiproliferative activity in vitro against a panel of cancer cell lines and, for selected highly active compounds, interaction with tubulin, cell cycle effects and in vivo potency. Both these series of compounds were characterized by the presence of a common 3′,4′,5′-trimethoxyphenyl ring at either the C-4 or C-5 position of the 2-methyloxazole ring. Compounds 4g and 4i, bearing a m-fluoro-p-methoxyphenyl or p-ethoxyphenyl moiety at the 5-position of 2-methyloxazole nucleus, respectively, exhibited the greatest antiproliferative activity, with IC50 values of 0.35-4.6 nM (4g) and 0.5–20.2 nM (4i), which are similar to those obtained with CA-4. These compounds bound to the colchicine site of tubulin and inhibited tubulin polymerization at submicromolar concentrations. Furthermore, 4i strongly induced apoptosis that follows the mitochondrial pathway. In vivo, 4i in a mouse syngeneic model demonstrated high antitumor activity which significantly reduced the tumor mass at doses ten times lower than that required for CA-4P, suggesting that 4i warrants further evaluation as a potential anticancer drug.

Scientific RepoRts | 7:46356 | DOI: 10.1038/srep46356 with IC 50 values ranging from 33 to 702 nM in the five cell lines examined but was comparable to CA-4 as an inhibitor of tubulin polymerization.
In our ongoing effort to discover novel and potent antimicrotubule agents, these results led us to start a pharmacophore exploration and optimization effort around the 2-methylthiazole derivatives with general formula 3. Here we describe replacing the thiazole nucleus with the less aromatic and basic bioisosteric equivalent oxazole ring 22 , by the preparation of two different regioisomeric series of 2-methyl-4,5-disubstituted oxazole derivatives with general structures 4 and 5. In these two series of designed analogues, obtained by interchanging the substitution pattern of ring A and B, we fixed one of the aryl groups as the 3′ ,4′ ,5′ -trimethoxyphenyl moiety, identical to the A-ring of CA-4, and examined several substitutions with electron-withdrawing (F and Cl) or electron-releasing (Me, OMe, and OEt) groups (EWG or ERG, respectively) on the other aryl moiety, corresponding to the B-ring of CA-4. In addition, for compounds 4a and 5a, the B-ring of CA-4 was replaced with the bulky and lipophilic naphth-2-yl moiety.
It has been previously reported that the replacement of the meta-hydroxy group of ring B of CA-4 with halogens such as fluorine or chlorine increased tubulin affinity as well as antiproliferative potency 23 . Since the methoxy and ethoxy groups proved to be favorable for bioactivity, we maintained one of these substituents at the para-position and introduced an additional substituent (F and Cl) at the meta-position of the phenyl ring.
Relative to the activity of 4e, the insertion of an additional EWG on the meta-position of the p-methoxyphenyl ring had varying effects on antiproliferative activity. The marked influence of an additional fluorine at the meta-position of 4e, to furnish the m-F-p-OMe derivative 4g, led to a 4-276-fold increase in antiproliferative activity, which was most pronounced in the A549 cells. An opposite effect occurred with replacement of m-fluorine with m-chlorine, to furnish derivative 4h, which led to a 60-213-fold reduction of activity relative to 4g. Compound 4h was also 5-60-fold less active than 4e against six of the seven cancer cell lines, the exception being the A549 cells.
The p-ethoxyphenyl homologue 4i was 2-to 358-fold more potent than its methoxy counterpart 4e. The greatest difference in activity was 358-fold against the A549 cells. Since the p-ethoxy group of 4i was favorable for potency, the introduction of an additional electron-withdrawing chlorine group at the meta-position of the p-ethoxyphenyl ring, resulting in compound 4j, had variable effects, producing a 5-40-fold reduction in antiproliferative activity against five of the cell lines and increased activity against MCF-7 and HT-29 cells. Compound 4i was from 2-to 12-fold more potent than the isomeric derivative 5f in five of the cell lines, the exceptions being the HT-29 and MCF-7 cells, in which 5f was 11-and 3-fold more active than 4i, respectively.
Effects of test compounds 4a, 4i and 5f in non tumoral cells. To obtain a preliminary indication of the cytotoxic potential of these derivatives in normal human cells, some of the most active compounds (4a, 4i and 5f) were evaluated in vitro against peripheral blood lymphocytes (PBL) from healthy donors. All compounds showed an IC 50 greater than 10 μ M both in quiescent lymphocytes and in lymphocytes in an active phase of proliferation induced by phytohematoagglutinin (PHA) a mitogenic stimulus (Table 2). Moreover, we also evaluated the effects of these compounds on primary cultures of human umbilical endothelial cells (HUVECs), and we found that the three compounds were practically inactive, having IC 50 values > 100 μ M. These results indicate that these compounds have very low toxicity in normal cells in comparison to tumor cells, suggesting potential for an excellent therapeutic index.
Inhibition of tubulin polymerization and colchicine binding. A subset of compounds (4a, 4d,e, 4g, 4i,j, 5a, 5e,f) were evaluated for their in vitro inhibition of tubulin polymerization and for inhibitory effects on the binding of [ 3 H]colchicine to tubulin (Table 3). CA-4 was also examined in contemporaneous experiments. In the assembly assay, with 10 μ M tubulin, one of the most active antiproliferative agents (4i), along with compound 4d, were the best inhibitors of tubulin polymerization, with IC 50 values of 0.56 and 0.66 μ M, respectively, having twice the potency of CA-4 (IC 50 :1.3 μ M). Derivatives 4a, 4g and 5f showed comparable antitubulin activity to that   of CA-4, while compound 5e was about half as potent as CA-4. For these new compounds and CA-4, the order of inhibitory effects on tubulin assembly was 4i > 4d > 4e > 4g = 4a = 5f = CA-4 > 5a > 4j > 5e.
In the colchicine binding studies, the same compounds potently inhibited the binding of [ 3 H]colchicine to tubulin, since 54-86% inhibition occurred with these agents and colchicine both at 5 μ M. Specifically, derivatives 4d and 4i were slightly less active than CA-4 (86 and 82% inhibition, respectively), which in this experiment inhibited colchicine binding by 99%. Inhibition of colchicine binding by compounds 4j, 5a and 5e fell into the 53-64% range.
While this group of compounds were all highly potent in the biological assays (inhibition of cell growth, tubulin assembly and colchicine binding), correlations between these assay types were imperfect. Thus, while compound 4g was half as active as an assembly inhibitor as 4d, these two compounds were equipotent in the colchicine binding assay. Moreover, compounds 4d and 4i were nearly equipotent as inhibitors of tubulin assembly, while 4i was 2-185-fold more active than 4d in its effects on cell growth.
Nevertheless, these studies identified tubulin as the molecular target of these compounds, since those with the greatest inhibitory effects on cell growth strongly inhibited tubulin assembly and the binding of colchicine to tubulin. Molecular modelling. The binding mode in the colchicine site of tubulin of the newly prepared 2-methyloxazole derivatives was elucidated performing a series of molecular docking simulations, following a previous reported procedure 21 . The binding observed for all the derivatives is closely related to the one found for the co-crystallized DAMA-colchicine, and it is consistent with those previously reported for different tubulin polymerization inhibitors 21,23 . The trimethoxyphenyl ring, in both the 4-(para-ethoxyphenyl) and isomeric 5-(para-ethoxyphenyl)-2-methyloxazole derivatives 4i and 5f, respectively, is in close contact with Cys241, while the second substituted phenyl ring occupies a small hydrophobic subpocket (Fig. 3, Panels A and B), with the ethoxy substituents lying deep in this pocket. Hydrophobic interactions with the surrounding amino acids (e.g., β Met259, β Thr314, β Val181, etc.) of the subpocket stabilize the binding of the molecules in the colchicine site.
Analysis of cell cycle effects. The effects of a 24 h treatment with different concentrations of 4a, 4i and 5f on cell cycle progression in Jurkat, and HeLa cells were determined by flow cytometry (Fig. 4, Panels A-F). All three compounds caused a significant G2/M arrest in a concentration-dependent manner in the two cell lines examined, with a rise in G2/M cells occurring at a concentration as low as 50 nM, especially with compound 4i, while at higher concentrations more than 60% of the cells were arrested in G2/M. The cell cycle arrest in G2/M phase was accompanied by a corresponding reduction in cells in the other phases (G1 and S) of the cell cycle. This ability of 4i to induce G2/M arrest correlates directly with its strong inhibition of tubulin polymerization. With the purpose of evaluating whether 4i arrested cells in mitosis, Hela cells were stained with an immunofluorescent antibody to p-histone H3, a well known mitotic marker 24 , as well as with propidium iodide (PI), and analyzed by flow cytometry. As shown in Fig. 5 (Panel A), in which representative histograms are presented, cells arrested in M phase by treatment with 4i are readily distinguished from G2 cells by the higher level of p-histone H3. Compound 4i induced, after a 24 h incubation, a dose-dependent increase in the percentage of mitotic cells, from 1.3% observed in the untreated cells to about 32% and 46% at 50 and 100 nM 4i, respectively.
Compound 4i induced alteration of cell cycle checkpoint proteins and induced DNA damage. We investigated the effects of 4i on the expression of proteins involved in regulation of the cell cycle and in spindle assembly. Cyclin B1 is involved in the G2 to M transition as a complex with cdc2, and the activation of the cdc2/cyclin B1 complex through cdc25c-dependent dephosphorylation of phospho-cdc2 and phosphorylation of cyclin B1 triggers cells to enter mitosis 25,26 . As shown in Fig. 5 (Panel B) a marked increase of cyclin B1 occurred in a concentration dependent manner following 24 and 48 h treatments with 4i. On the other hand, total cdc25c expression was reduced both at 24 and 48 h after treatment with 100 nM 4i, and we observed the simultaneous appearance of a slowly migrating form of cdc25c, indicating changes in its phosphorylation state. Furthermore, in good agreement, the expression of phosphorylated cdc2 decreased, most noticeable after the 24 h treatment with 100 nM 4i. These findings are in good agreement with previous results 21, 23 obtained with other antimitotic derivatives and indicate that cdc2/cyclin B1 complexes failed to be activated, preventing cells from exiting mitosis, which would eventually lead to apoptotic cell death.
Moreover, since it is well known that prolonged mitotic arrest induces DNA damage 27,28 we also examined the expression of phosphorylated histone H2A.X at Ser139 (γ H2A.X), a marker of DNA damage 29 . We observed (Fig. 5, Panel B) a great increase of the phosphorylation of γ H2A.X, after a 48 h treatment, suggesting that DNA damage occurred following treatment with 4i.
Compound 4i induced apoptosis. To evaluate the mode of cell death induced by 4i, we used an annexin-V/PI assay. We treated two cell lines (HeLa and Jurkat) with the test compound with concentrations ranging from 15 to 125 nM for 24 or 48 h. As shown in Fig. 6, both Jurkat (Panels A, B) and HeLa (Panels C, D) cells treated with 4i showed a significant accumulation of annexin-V positive cells in a concentration dependent manner after a 24 h treatment, and the proportions of apoptotic cells further increased at 48 h. Note that 4i caused the appearance of 70% apoptotic cells at 60 nM in Hela cells, while in the leukemic cell line we observed a lower value (40%) after a 24 h treatment, in good agreement with its cytotoxic activity.
Compound 4i induced apoptosis through the mitochondrial pathway. Since many combretastatin analogues cause apoptosis following the mitochondrial pathway [19c, refs 20 and 23] we investigated if 4i also induced mitochondrial depolarization. Mitochondrial potential was monitored by flow cytometry using the fluorescent dye JC-1. We treated both Hela and Jurkat cells with 4i at 50 or 100 nM for 24 or 48 h. As shown in Fig. 7 (Panels A and B), both Jurkat and HeLa cells treated with 4i exhibited a marked increase in the percentage of cells with low Δ ψ mt in a time dependent manner, paralleling the results obtained with the annexin-V apoptotic assay. Since it is well known that dissipation of mitochondrial potential is associated with mitochondrial production of reactive oxygen species (ROS) 30,31 , we also evaluated whether ROS production increased after treatment with 4i. We utilized the dye 2,7-dichlorodihydrofluorescein diacetate (H 2 -DCFDA), which is oxidized to the fluorescent compound dichlorofluorescein (DCF) upon ROS induction.
The results shown in Fig. 7 (Panels C and D) indicate that 4i induced ROS production in comparison with the amounts observed in control cells, in both Jurkat and HeLa cells, moving from about 4% of DCF positive cells in untreated samples to about 25-30% in treated cells at 100 nM, the highest concentration examined. These results are in excellent agreement with the dissipation of Δ ψ mt described above. We also investigated the expression of two anti-apoptotic proteins, Mcl-1 and XIAP. Mcl-1 is a member of the Bcl-2 family of anti-apoptotic proteins. Mcl-1 is overexpressed in many cancers, and it has been reported that sensitivity to antimitotic drugs is regulated by Mcl-1 levels 33 . As shown in Fig. 8, the expression of Mcl-1 was only slightly decreased at the highest 4i concentration used (100 nM).
On the other hand, XIAP, also a member of the IAP family (inhibitors of apoptosis protein) 34 , was significantly reduced after both the 24 and 48 h treatments, suggesting that 4i treatment induced downregulation of these proteins to disable their anti-apoptotic function.
Evaluation of antivascular activity of 4i and 5f. Since many tubulin binding agents, including CA-4, are endowed with vascular disrupting activity [9c], we investigated in vitro the potential anti-vascular activity of compounds 4i and 5f. We tested the anti-vascular effects in HUVECs, evaluating the ability of the compounds to i) interfere with angiogenesis by inhibiting endothelial cell migration and ii) to interfere with the process of capillary-like tube formation. Confluent HUVEC monolayers were scraped with a pipette tip and cellular migration induced to repair the wound was followed by optical microscopy, and the percentage of reduction of wound healing was calculated at different times 35 . As shown in Fig. 9 (Panels A and B), after only a 6 h treatment, cell migration was significantly reduced at 100 nM compound 4i or 5f. This significant reduction was maintained for a 24 h treatment as well. In contrast, CA-4 was significantly active even at 10 nM, after both the 6 and 24 h treatments.
Endothelial cells seeded on Matrigel are able to form a capillary network miming the first angiogenesis steps 36 . This network is a useful experimental model to assess the action of molecules on vascular morphogenesis.
The antivascular effect induced by the tested compounds is shown in Fig. 9 (Panels C and D). In this case, we evaluated the pictures taken after a 24 h incubation by optical microscopy (Fig. 9, Panel C), and a quantitative analysis was carried out studying two dimensional (percent area covered by HUVECs and total length of HUVECs network per field) and two topological parameters (number of meshes and branching points per field).
Compounds 4i and 5f did not show a significant antivascular effect on Matrigel preformed tubular structures (Fig. 9, Panel D), even at the highest concentration used (100 nM). In this assay, CA-4 showed, as expected, a significant antivascular activity even at 10 nM. Our findings suggest that these compounds do not possess a good antivascular profile in comparison to the well known activity of CA-4, since some effects, in particular cellular migration, occurred at a concentration higher than that required for CA-4.  37 . This model consists of the use of B16 murine melanoma cells that are injected in the flank of mice. Consequently, in preliminary experiments we wanted to verify the effectiveness of the compound 4i on this murine tumor line. The compound had an IC 50 of 15.6 ± 1.9 nM measured by the MTT assay, indicating that its cytotoxic potency was similar to that found in human tumor cell lines (see Table 1). In addition, we also evaluated, if in this cell line, the compound was able to arrest cells in the G2/M phase of the cell cycle and to induce apoptosis. As shown in Fig. 10 (Panels A and B), 4i, as was observed in the human tumor cell lines, induced a G2/M arrest and a strong apoptotic response at low concentrations (50-100 nM). These results indicated that the allogeneic mouse model  used was fully relevant to the evaluation of 4i in vivo. Thus, 4i was administered by the intraperitoneal route every other day, at two different doses (3.0 and 7.5 mg/kg). As reference compound, CA-4P (1b) was used at 30 mg/kg.
As shown in Fig. 10 (panel C), after a six day treatment (doses administered on days 9, 11 and 14), 4i was able to significantly reduce tumor burden in a dose-dependent manner, even at the lowest dose tested (3.0 mg/kg). We observed reduction in tumor mass of 34.9, and 52.5% at the doses of 3.0 and 7.5 mg/kg, respectively. The reference compound CA-4P at 30 mg/kg induced only a 28.0% reduction in tumor mass. Notably, the in vivo efficacy clearly indicate an increased antitumor efficacy of 4i as compared with CA-4P (in both weight and molar terms). Even at the highest dose, 4i did not show any sign of toxicity and did not cause a decrease in animal body weight (data not shown).

Conclusions
The isomerization of the cis-double bond of CA-4 to its trans-form in solution is one of the major disadvantages of this molecule. The instability of the Z-double bond of CA-4 has been resolved by incorporating the stilbene double bond into the structure of five-member heterocyclic rings. The bioisosteric equivalence between oxazole and thiazole prompted us to synthesize by a three-step procedure two novel series of 2-methyl-4,5-disubstituted oxazole derivatives with general formulas 4 and 5, in which the oxazole ring replaced the thiazole system of previously published analogues with general structure 3 and could serve as a suitable mimic to retain the bioactive configuration afforded by the cis-double bond present in CA-4. For both these series of compounds, the 3′ ,4′ ,5′ -trimethoxyphenyl and 2-methyloxazole rings mimic the ring A and cis-double bond of CA-4, respectively, while a naphth-2-yl or phenyl ring substituted with electron-releasing or electron-withdrawing groups was utilized as a B-ring surrogate to mimic the 3′ -hydroxy-4′ -methoxyphenyl group in CA-4. Comparing compounds with the same aryl substitution, the 2-methyl-4-(3′ ,4′ ,5′ -trimethoxyphenyl)oxazole derivatives were more active than their isomeric 2-methyl-5-(3′ ,4′ ,5′ -trimethoxyphenyl)oxazole counterparts. The results indicated that 2-methyloxazole derivatives 4a, 4d-e, 4i and 5e exhibited more highly potent antiproliferative activity than the corresponding 2-methylthiazole analogues 3a-e previously described. These marked differences were maintained in all the biological evaluations performed. In particular, it is important to underline that 4i has very low toxicity in non tumoral cell lines such as PBLs and HUVECs. Although preliminary investigations regarding the potential antivascular activity of these new oxazoles derivatives indicated that they are not potent vascular disrupting agents, in vivo experiments demonstrated that 4i had excellent antitumor activity that was evident at lower doses than CA-4P and in the absence of obvious toxicity. In summary, the biological characterization of compound 4i provides compelling evidence to support its further development as an anticancer drug.

Experimental Section
Chemistry. Materials and Methods. 1 H and 13 C NMR data were obtained with a Varian VXR 200 spectrometer and a Varian Mercury Plus 400 spectrometer, respectively. Peak positions are given in parts per million (δ) downfield, and J values are given in hertz. Positive-ion electrospray ionization (ESI) mass spectra were recorded on a double-focusing Finnigan MAT 95 instrument with BE geometry. Melting points (mp) were determined on a Buchi-Tottoli apparatus and are uncorrected. The purity of tested compounds was determined by combustion elemental analyses conducted by the Microanalytical Laboratory of the Chemistry Department of the University of Ferrara with a Yanagimoto MT-5 CHN recorder elemental analyzer. All tested compounds yielded data consistent with a purity of at least 95% as compared with the theoretical values. TLC was carried out using glass plates coated with silica gel 60 F 254 by Merck, and compounds were visualized by UV detection or with aqueous KMnO 4 . Flash column chromatography was performed using 230-400 mesh silica gel and the indicated solvent system. Organic solutions were dried over anhydrous Na 2 SO 4 . Solvents and reagents that are commercially available were purchased from Aldrich (Sigma-Aldrich) or Alfa Aesar (Johnson Matthey Company) and were used without further purification unless otherwise noted.
General procedure A for the preparation of compounds 8 and 9a-f. A mixture of the appropriate 2-bromoacetophenone 6 or 7a-f (4 mmol) and acetamide (708 mg, 12 mmol) was heated to 150 °C for 2 h. After this time, the mixture was cooled to room temperature, treated with a 2 M aqueous solution of Na 2 CO 3 (10 mL), and the suspension was carefully adjusted to pH 12 with Na 2 CO 3 . The mixture was extracted with EtOAc (2 × 20 mL), the combined organic phase was washed with water (10 mL) and brine (10 mL), dried (Na 2 SO 4 ) and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel.      General procedure B for the preparation of compounds 10 and 11a-f. A solution of the appropriate 2-methyl-4-aryloxazole 8 or 9a-f (4 mmol) in anhydrous CHCl 3 (20 mL) was cooled to 0 °C, then N-bromosuccinimide (783 mg, 4.4 mmol) was added in small portions. The reaction mixture was allowed to warm slowly to room temperature. After 2 h, the resulting mixture was diluted with CH 2 Cl 2 (20 mL), washed with a saturated solution of NaHCO 3 (10 mL), brine (10 mL), dried (MgSO 4 ) and evaporated. The residue was purified by column chromatography on silica gel.

Materials and Methods. Cell growth conditions and antiproliferative assay.
Human T-cell leukemia (Jurkat) and human B-cell leukemia (SEM and RS4;11) cells were grown in RPMI-1640 medium (Gibco, Milano, Italy). Breast adenocarcinoma (MCF-7), human cervix carcinoma (HeLa), human lung adenocarcinoma (A549) and human colon adenocarcinoma (HT-29) cells were grown in DMEM medium (Gibco, Milano, Italy), all supplemented with 115 units/mL penicillin G (Gibco, Milano, Italy), 115 μg/mL streptomycin (Invitrogen, Milano, Italy), and 10% fetal bovine serum (FBS; Invitrogen, Milano, Italy). Stock solutions (10 mM) of the different compounds were obtained by dissolving them in DMSO. Individual wells of a 96-well tissue culture microtiter plate were inoculated with 100 μ L of complete medium containing 8 × 10 3 cells. The plates were incubated at 37 °C in a humidified 5% CO 2 incubator for 18 h prior to the experiments. After medium removal, 100 μ L of fresh medium containing the test compound at different concentrations was added to each well in triplicate and incubated at 37 °C for 72 h. The percentage of DMSO in the medium never exceeded 0.25%. This was also the maximum DMSO concentration in all cell-based assays described below. Cell viability was assayed by the MTT test as previously described. 21 The IC 50 was defined as the compound concentration required to inhibit cell proliferation by 50%, in comparison with cells treated with the maximum amount of DMSO (0.25%) and considered as 100% viability.
Peripheral blood lymphocytes (PBL) from healthy donors were obtained from human peripheral blood (leucocyte rich plasma-buffy coats) from healthy volunteers using the Lymphoprep (Fresenius KABI Norge AS) gradient density centrifugation.
Buffy coats were obtained from the Blood Transfusion Service, Azienda Ospedaliera of Padova and provided at this institution for research purposes. Therefore, no informed consent was further needed. In addition, buffy coats were provided without identifiers. The experimental procedures were carried out in strict accordance with approved guidelines.
After extensive washing, cells were resuspended (1.0 × 10 6 cells/mL) in RPMI-1640 with 10% FBS and incubated overnight. For cytotoxicity evaluations in proliferating PBL cultures, non-adherent cells were resuspended at 5 × 10 5 cells/mL in growth medium, containing 2.5 μg/mL PHA (Irvine Scientific). Different concentrations of the test compounds were added, and viability was determined 72 h later by the MTT test. For cytotoxicity evaluations in resting PBL cultures, non-adherent cells were resuspended (5 × 10 5 cells/mL) and treated for 72 h with the test compounds, as described above.
Effects on tubulin polymerization and on colchicine binding to tubulin. To evaluate the effect of the compounds on tubulin assembly in vitro 41 , varying concentrations of compounds were preincubated with 10 μ M bovine brain tubulin in 0.8 M monosodium glutamate (pH adjusted to 6.6 with HCl in a 2.0 M stock solution) at 30 ˚C and then cooled to 0 ˚C. After addition of 0.4 mM GTP, the mixtures were transferred to 0 °C cuvettes in a recording spectrophotometer and warmed to 30 °C. Tubulin assembly was followed turbidimetrically at 350 nm. The IC 50 was defined as the compound concentration that inhibited the extent of assembly by 50% after a 20 min incubation. The capacity of the test compounds to inhibit colchicine binding to tubulin was measured as described 42 , except that the reaction mixtures contained 1 μ M tubulin, 5 μ M [ 3 H]colchicine and 1 or 5 μ M test compound.
Flow cytometric analysis of cell cycle distribution. 5 × 10 5 HeLa or Jurkat cells were treated with different concentrations of the test compounds for 24 h. After the incubation period, the cells were collected, centrifuged, and fixed with ice-cold ethanol (70%). The cells were then treated with lysis buffer containing RNase A and 0.1% Scientific RepoRts | 7:46356 | DOI: 10.1038/srep46356 Triton X-100 and then stained with PI. Samples were analyzed on a Cytomic FC500 flow cytometer (Beckman Coulter). DNA histograms were analyzed using MultiCycle for Windows (Phoenix Flow Systems).
Apoptosis assay. Cell death was determined by flow cytometry of cells double stained with annexin V/FITC and PI. The Coulter Cytomics FC500 (Beckman Coulter) was used to measure the surface exposure of PS on apoptotic cells according to the manufacturer's instructions (Annexin-V Fluos, Roche Diagnostics).
Western blot analysis. HeLa cells were incubated in the presence of 4i and, after different times, were collected, centrifuged, and washed two times with ice cold phosphate buffered saline (PBS). The pellet was then resuspended in lysis buffer. After the cells were lysed on ice for 30 min, lysates were centrifuged at 15000× g at 4 °C for 10 min. The protein concentration in the supernatant was determined using the BCA Protein Assay (Pierce, Italy). Equal amounts of protein (10 μ g) were resolved using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Criterion Precast, BioRad, Italy) and transferred to a PVDF Hybond-P membrane (GE Healthcare). Membranes were blocked with a bovine serum albumin solution (5% in Tween PBS 1X), the membranes being gently rotated overnight at 4 °C. Membranes were then incubated with primary antibodies against caspase-9, PARP, cdc25c, p-H2AX Ser139 , cyclin B, p-cdc2 Tyr15 (all from Cell Signaling) or β -actin (Sigma-Aldrich) for 2 h at room temperature. Membranes were next incubated with peroxidase labeled secondary antibodies for 60 min. All membranes were visualized using ECL Select (GE Healthcare), and images were acquired using an Uvitec-Alliance imaging system (Uvitec, Cambridge, UK). To ensure equal protein loading, each membrane was stripped and reprobed with anti-β -actin antibody.
Antivascular activity. HUVECs were prepared from human umbilical cord veins, as previously described 43 . The adherent cells were maintained in M200 medium supplemented with LSGS (Low Serum Growth Supplement), containing FBS, hydrocortisone, hEGF, bFGF, heparin, gentamycin/amphotericin (Life Technologies, Monza, Italy). Once confluent, the cells were detached by a trypsin-EDTA solution and used in experiments from the first to sixth passages.
The motility assay for HUVECs was based on "scratch" wounding of a confluent monolayer 35 . Briefly, HUVECs (1x10 5 ) were seeded onto 6-well plates coated with 0.1% collagen type I (BD Biosciences, Italy) in complete medium until a confluent monolayer was formed. The cells were wounded using a pipette tip, and wells were washed with PBS to remove undetached cells. Then, the cells were treated with the test compounds, and, at different times from the scratch, the cells were photographed under a light microscope. At all indicated time points, the wound width was measured in four areas and compared with the initial width.
Matrigel matrix (Basement Membrane Matrix, BD Biosciences, Italy) was kept at 4 °C for 3 h, when 230 μ L of Matrigel solution was added to each well of a 24-well plate. After gelling at 37 °C for 30 min, gels were overlaid with 500 μ L of medium containing 6 × 10 4 HUVECs. The cells were incubated over Matrigel for 6 h to allow capillary tubes to form. Different concentrations of test compound were added in the cultures and incubated for different times, and the disappearance of existing vasculature was monitored and photographed (five fields for each well: the four quadrants and the center) at a 10x magnification. Phase contrast images were recorded using a digital camera and saved as TIFF files. Image analysis was carried out using ImageJ image analysis software, and dimensional (percent area covered by HUVECs and total length of HUVECs network per field) and topological parameters (number of meshes and branching points per field) were determined 43 . Values were expressed as percent change from control cultures grown with complete medium.
In vivo animal studies. Animal experiments were approved by our local animal ethics committee (OPBA, Organismo Preposto al Benessere degli Animali, Università degli Studi di Brescia, Italy) and were executed in accordance with national guidelines and regulations. Procedures involving animals and their care conformed with institutional guidelines that comply with national and international laws and policies (EEC Council Directive 86/609, OJ L 358, 12 December 1987) and with "ARRIVE" guidelines (Animals in Research Reporting In Vivo Experiments). Six week old C57BL/6 mice (Charles River, Calco) were injected subcutaneously into the dorsolateral flank with 2.5x10 5 BL6-B16 murine melanoma cells in 200 μ L of PBS. When tumors were palpable, animals were treated intraperitoneally every other day with different doses of test compounds dissolved in 50 μ L of DMSO. Tumors were measured in two dimensions, and tumor volume was calculated according to the formula V = (D × d 2 )/2, where D and d are the major and minor perpendicular tumor diameters, respectively. Statistical analysis. Unless indicated differently, the results are presented as mean ± SEM. The differences between different treatments were analyzed, using the two-sided Student's t test. P values lower than 0.05 were considered significant.