Synthesis and Anticancer Activity of Mitotic-Specific 3,4-Dihydropyridine-2(1H)-thiones

Most anticancer drugs target mitosis as the most crucial and fragile period of rapidly dividing cancer cells. However the limitations of classical chemotherapeutics drive the search for new more effective and selective compounds. For this purpose structural modifications of the previously characterized pyridine analogue (S1) were incorporated aiming to obtain an antimitotic inhibitor of satisfactory and specific anticancer activity. Structure-activity relationship analysis of the compounds against a panel of cancer cell lines allowed to select a compound with a thiophene ring at C5 of a 3,4-dihydropyridine-2(1H)-thione (S22) with promising antiproliferative activity (IC50 equal 1.71 ± 0.58 µM) and selectivity (SI = 21.09) against melanoma A375 cells. Moreover, all three of the most active compounds from the antiproliferative study, namely S1, S19 and S22 showed better selectivity against A375 cells than reference drug, suggesting their possible lower toxicity and wider therapeutic index. As further study revealed, selected compounds inhibited tubulin polymerization via colchicine binding site in dose dependent manner, leading to aberrant mitotic spindle formation, cell cycle arrest and apoptosis. Summarizing, the current study showed that among obtained mitotic-specific inhibitors analogue with thiophene ring showed the highest antiproliferative activity and selectivity against cancer cells.


Introduction
The key function of microtubules and mitotic spindles in cell division make them an attractive target in cancer therapy. Therefore, microtubule-targeting agents (MTAs), also named "spindle poisons" remain the most important of classical chemotherapeutics [1]. These drugs are divided into two main groups: microtubule stabilizing agents (including taxanes, epothilones, laulimalide) and microtubule destabilizing agents (colchicine, vinca alkaloids, combretastatins) with five different tubulin binding sites [2]. Microtubule polymerization and likewise depolymerization are necessary for correct cell division. Therefore, despite different actions, both types of MTAs disrupt microtubule dynamics, leading to achieved by thionation of oxygen analogue O7, obtained by addition of benzylmagnesiate to N-Me 5-phenylpyridin-2-one. Product S6 was gained by thionation of oxygen analogue O6, obtained as by-product during benzylation conducted with bigger amount of MeLi in comparison to a standard procedure. The structures of all compounds were elucidated with the aid of 1D NMR ( 1 H, 13 C, 13 C-DEPT-135) and basic 2D NMR spectroscopy, low (GC-MS) and high (HRMS) resolution mass spectrometry and elemental analyses. The dihydropyridinones (O-serie) and dihydropyridinethiones (S-serie) were obtained in the form of racemates. Synthesis and spectroscopic data of S1-S4 and S6-S22 is described in the Materials and methods (4.1. Experimental part), while synthesis and spectroscopic data of their precursors is presented in the Appendix A (Synthesis of precursors of compounds S1-S22). Figure 1. Compounds S1-S22 obtained and tested within a three-step optimisation process. Figure 1. Compounds S1-S22 obtained and tested within a three-step optimisation process. Scheme 1. Sequence of the synthesis of targeted compounds S1-S4 and S8-S22.

The Antiproliferative Activity
The compound S1 with previously described promising anticancer activity [9] has been chosen as a starting point for the current study. It should be emphasized that biological activity of S1 was evaluated at single high concentration (100 µ M) in the previous study. The Cell Proliferation Reagent (WST-1 assay) was used to evaluate the antiproliferative activity of S1 and its analogues against selected cancer cell lines including breast (MCF7), melanoma (A375), colon (HT-29), ovarian (SK-OV-3) and prostate (PC-3). A combretastatin (CA-4) one of the most extensively researched CBSIs [16], was used as reference control as it was predicted that S1 may bind to colchicine-binding side (which is presented and discussed below). The results are summarized as IC50 values in Table 1. The data revealed that from the following N-H and N-substituted derivatives compound S1 with N-H moiety and S2 with N-Methyl group were the most promising (IC50 values equal 4.33 ± 1.00 µ M and 12.55 ± 2.16 µ M against A375, respectively). Moreover, structure-activity relationship study of the first step showed that benzene ring is essential for the cytostatic activity of the studied structure since the analogue without 5-phenyl group (S5) lost anticancer activity (IC50 > 100 µ M). The importance of that group was previously confirmed in case of dihydropyrimidinethiones derivatives [17]. In spite of increase lipophilic properties, compounds S6 and S7 (isomers of S1 and S2, respectively) were markedly less active, probably as a result of different position of nitrogen atom in relation to phenyl group.
During the second task of the study compounds S1 and S2 were further derivatized by adding appropriate groups to A-ring. Selected groups were introduced on the basis of their π parameter, determining influence of different substituents on Log p value: two derivatives (S8/S13 and S9/S14) containing respectively one or two fluorine atoms (each of F atoms increases Log P by 0.13), one derivative (S10/S15) containing chlorine, increasing Log p value by 0.51, two derivatives (S11/S16 and S12/S17) containing one or two methoxy groups, respectively (each of the groups decreases Log p value by 0.26 compared to the original compound). In the present study we did not observe direct correlation between Log p values and antiproliferative activity, as it was noted in our previously published data, where the most lipophilic pyrimidine analogues were also the most potent growth inhibitors [17]. The introduction of substituents at the para-position of the phenyl ring (compounds S8-S17) resulted in a reduction of antiproliferative activity against all selected cancer cell lines, as compared with the activity of the unsubstituted A-ring analogue (S1 and S2). Scheme 1. Sequence of the synthesis of targeted compounds S1-S4 and S8-S22.

The Antiproliferative Activity
The compound S1 with previously described promising anticancer activity [9] has been chosen as a starting point for the current study. It should be emphasized that biological activity of S1 was evaluated at single high concentration (100 µM) in the previous study. The Cell Proliferation Reagent (WST-1 assay) was used to evaluate the antiproliferative activity of S1 and its analogues against selected cancer cell lines including breast (MCF7), melanoma (A375), colon (HT-29), ovarian (SK-OV-3) and prostate (PC-3). A combretastatin (CA-4) one of the most extensively researched CBSIs [16], was used as reference control as it was predicted that S1 may bind to colchicine-binding side (which is presented and discussed below). The results are summarized as IC 50 values in Table 1. The data revealed that from the following N-H and N-substituted derivatives compound S1 with N-H moiety and S2 with N-Methyl group were the most promising (IC 50 values equal 4.33 ± 1.00 µM and 12.55 ± 2.16 µM against A375, respectively). Moreover, structure-activity relationship study of the first step showed that benzene ring is essential for the cytostatic activity of the studied structure since the analogue without 5-phenyl group (S5) lost anticancer activity (IC 50 > 100 µM). The importance of that group was previously confirmed in case of dihydropyrimidinethiones derivatives [17]. In spite of increase lipophilic properties, compounds S6 and S7 (isomers of S1 and S2, respectively) were markedly less active, probably as a result of different position of nitrogen atom in relation to phenyl group.
During the second task of the study compounds S1 and S2 were further derivatized by adding appropriate groups to A-ring. Selected groups were introduced on the basis of their π parameter, determining influence of different substituents on Log p value: two derivatives (S8/S13 and S9/S14) containing respectively one or two fluorine atoms (each of F atoms increases Log P by 0.13), one derivative (S10/S15) containing chlorine, increasing Log p value by 0.51, two derivatives (S11/S16 and S12/S17) containing one or two methoxy groups, respectively (each of the groups decreases Log p value by 0.26 compared to the original compound). In the present study we did not observe direct correlation between Log p values and antiproliferative activity, as it was noted in our previously published data, where the most lipophilic pyrimidine analogues were also the most potent growth inhibitors [17]. The introduction of substituents at the para-position of the phenyl ring (compounds S8-S17) resulted in a reduction of antiproliferative activity against all selected cancer cell lines, as compared with the activity of the unsubstituted A-ring analogue (S1 and S2).  1 Caclulated using the Ghose-Crippen approach [18] When the effect of electron-releasing (ERG's) or electron-withdrawing groups (EWG's) at the phenyl ring was compared, compounds S11 and S16 with electron-donating methoxy groups were generally the most potent among -N-H and -N-Me analogues, respectively. Our data are compliant with Romagnoli et al. study, where the presence of electrondonating group (like methyl, methoxy) rather than electron-withdrawing groups on the phenyl moiety at the position of the 2-alkoxycarbonyl-3-(3 ,4 ,5 -trimethoxyanilino)thiophene system contributed to increase biological activity [19]. Surprisingly, in the current study two methoxy group (S12 and S17) contributed to decrease anticancer potential. In general, during the second stage of the study, -N-H analogues (S8-S12) were more potent than -N-Me counterparts (S13-S17). Therefore, the antiproliferative activity of -N-H derivatives with aromatic rings at C4/C5 position were evaluated at the third stage. As shown in Table  1, additional modifications of aryl rings at C4 and C5 contributed to increased anticancer activity (IC 50 values of compounds S18-S22 ≤ 10 µM), however, an analogue with thiophene ring at C5 pyridine-2(1H)thione (S22) was the most effective one, with IC 50 value of 1.71 ± 0.58 µM and 1.67 ± 1.47 µM against A375 and SK-OV-3 cell line, respectively. These findings correspond with several previous studies where thiophene, benzothiophene and fused thienopyridine or thienopyrimidine rings are presented in a variety of anticancer compounds [19][20][21], including CBSIs [22][23][24].

Selectivity-Index
The starting compound (S1) and its two derivatives defined as the most active compounds in the antiproliferative study (S19 and S22) were selected for further biological study, using A375 cells, that presented the highest sensitivity to the tested compounds. At first, the antiproliferative assay (WST-1) was used to estimate the selectivity-index (SI) defined as ratio of IC 50 for normal cells to IC 50 for cancer cells. Cancer cell line A375 and a non-tumor human melanocyte cell line (HEM) were used for that purpose. As shown in Table 2, all the newly evaluated compounds showed better selectivity against cancer cells when compared with reference: CA-4. Similar to our study,  as non-selective cytotoxic agent. Namely, the authors revealed that CA-4 showed higher antiproliferative activity against human lymphocytes than tumor cell lines, except for HL-60 and OVCAR-8 [25]. Whereas, Blay et al. reported that in spite of the combination of ombrabulin (CA-4 analogue) and cisplatin significantly improved progression-free survival in patients with advanced soft-tissue sarcomas, serious adverse effect excluded this strategy from anticancer therapy [26]. According to the "selectivity criteria" (SI > 10 [27]) only the compound S22 could be considered as selective. Similarly, Hura et al. revealed that among twenty-one tested CA-4 derivatives, C-13 analogue with thiophene as ring B was the most potent and selective against cancer cells [24].

Apoptosis Detection
Since apoptosis is identified as one of the most common mechanism of drug induced cell death, an apoptotic marker was evaluated using flow cytometry in the next step of the study. The assay used was based on the detection of phosphatidylserine (PS) on the surface of apoptotic cells. A dead cell marker (7-AAD) is also used as an indicator of the cell membrane structural integrity. Simultaneous staining with two dyes allows to distinguish (after debris excluding) viable (Annexin V-PE− and 7-AAD−), early apoptotic (Annexin V-PE+ and 7-AAD−) and late apoptotic/already dead cells (Annexin V-PE+ and 7-AAD+). The A375 and HEM cells were treated with the compounds: S1, S19 and S22 at 10 µM for 48 h. CA-4 was used as reference. As shown in Figure 2A, the population of living (non-apoptotic) cancer cells significantly decrease after treatment. Moreover, compound S22 showed the most comparable pattern to CA-4 with the percentage of early and late apoptotic cells equal to 27.42 ± 3.20% and 43.58 ± 3.30% for S22 and 31.57 ± 3.16% and 43.42 ± 3.43% for CA-4, in A375 and HEM cells, respectively. As for S1, S19 and S22, more apoptotic cells could have been expected after treatment with 10 µM concentration (based on IC 50 values from WST-1 assay), obtained data from apoptosis test suggest that those compounds possess stronger antiproliferative than direct cytotoxic potential. Simultaneously, analysis was performed in non-cancerous cells what allowed to confirm selectivity observed during antiproliferative study ( Figure 2B). Namely, only CA-4 and S1 significantly increased the percentage of early apoptotic cells to 19.74 ± 2.19% and 16.43 ± 4.88%, respectively when compared to untreated control (8.8 ± 1.1%). Surprisingly, during apoptosis detection in both cell lines, the differences between reference CA-4 and our compounds seemed less distinct than during IC 50 evaluation. Representative individual experiments (scatter diagrams with cell populations divided into quadrant dot plots specific for Annexin V/7-AAD staining) are shown in Figures S1 and S2. Additionally, optical microscopy imaging of A375 and HEM cells was performed just before apoptosis analysis of harvested cells. As shown in Figure S3, A375 treated cells had lower density in comparison with non-treated cells, while increased the number of dead cells could be observed only in CA-4 and S22 exposed cells, which was consistent with flow cytometry results. In the HEM cells, the most striking difference between control and treated cells was in cell morphology, especially in case of CA-4 and S22 exposed cells ( Figure S4).

Cell Cycle Analysis
Since growth inhibition may be a result of arrest cell cycle at a particular point, specific antiproliferative activity of the selected compounds was further evaluated by flow cytometry, for the impact on the cell cycle progression in A375 cells. The assay utilizes PIbased staining of DNA content to discriminate and measure the percentage of cells in each cell cycle phase (i.e., G0/G1, S, and G2/M) [28]. In present study, the cells were treated with tested compounds at 10 µ M concentration for 6 and 24 h, aimed to detect cell cycle arrest leading to apoptosis at 48 h of treatment. As expected, S1, S19 and S22 compounds caused accumulation of A375 cells in a tetraploid (4N) state, simultaneously decreasing the percentage of cells in G0/G1 phase after 6 h of treatment ( Figure 3A) and Figure S5 A-E. The percentage of A375 cells in G2/M phase was 54.46 ± 2.94%, 51.43 ± 4.77% and 56.95 ± 3.09% for S1, S19 and S22, respectively, compared to 33.64 ± 1.97% in control. These data were comparable with reference CA-4, where the percentage of arrested cells was 60.00 ±

Cell Cycle Analysis
Since growth inhibition may be a result of arrest cell cycle at a particular point, specific antiproliferative activity of the selected compounds was further evaluated by flow cytometry, for the impact on the cell cycle progression in A375 cells. The assay utilizes PI-based staining of DNA content to discriminate and measure the percentage of cells in each cell cycle phase (i.e., G0/G1, S, and G2/M) [28]. In present study, the cells were treated with tested compounds at 10 µM concentration for 6 and 24 h, aimed to detect cell cycle arrest leading to apoptosis at 48 h of treatment. As expected, S1, S19 and S22 compounds caused accumulation of A375 cells in a tetraploid (4N) state, simultaneously decreasing the percentage of cells in G0/G1 phase after 6 h of treatment ( Figure 3A) and Figure  S5A-E. The percentage of A375 cells in G2/M phase was 54.46 ± 2.94%, 51.43 ± 4.77% and 56.95 ± 3.09% for S1, S19 and S22, respectively, compared to 33.64 ± 1.97% in control. These data were comparable with reference CA-4, where the percentage of arrested cells was 60.00 ± 2.54%. Surprisingly, after 24 h, when CA-4 induced mitotic block increased to 72.36 ± 1.10%, the percentage of G2/M phase cells treated with tested compounds was much lower and not significantly different from control. However, it should be pointed that simultaneously the number of hypodiploid DNA (sub-G0/G1) and/or polyploid DNA (population after G2/M phase) markedly increased ( Figure 3B). As shown in Figure S5F-J, the sub-G0/G1 population increased significantly in S1, S19 and S22 treated cells which is considered as characteristic feature of cells undergoing apoptosis [29], whereas in CA-4 treated cells, a DNA index dramatically increased as a result of DNA polyploidization. Polyploidy is also typical for cells blocked in mitosis by colchicine (described as "c-mitosis" or "colchicine-mitosis"), as a results of partial or complete absence of spindle apparatus following the breakdown of nuclear envelope, condensed chromosomes, and undivided centromeres [30]. Our data are consistent with Gilson et al. study, where after initial accumulation of H358 and HeLa cells in mitosis as a result of pyrrolopyrimidine (PP-13) treatment, the percentage of G2/M arrested cells was gradually decreasing, concomitantly with the accumulation in the sub-G1 and polyploid fraction [31]. As mentioned above, inhibition of tubulin polymerization and mitotic spindle formation may be one of the reasons for polyploid nuclei formation. Therefore, in the further study, the impact of S1, S19 and S22 compounds on mitotic spindle formation and tubulin polymerization was evaluated. 2.54%. Surprisingly, after 24 h, when CA-4 induced mitotic block increased to 72.36 ± 1.10%, the percentage of G2/M phase cells treated with tested compounds was much lower and not significantly different from control. However, it should be pointed that simultaneously the number of hypodiploid DNA (sub-G0/G1) and/or polyploid DNA (population after G2/M phase) markedly increased ( Figure 3B). As shown in Figure S5 F-J, the sub-G0/G1 population increased significantly in S1, S19 and S22 treated cells which is considered as characteristic feature of cells undergoing apoptosis [29], whereas in CA-4 treated cells, a DNA index dramatically increased as a result of DNA polyploidization. Polyploidy is also typical for cells blocked in mitosis by colchicine (described as ''c-mitosis'' or ''colchicine-mitosis''), as a results of partial or complete absence of spindle apparatus following the breakdown of nuclear envelope, condensed chromosomes, and undivided centromeres [30]. Our data are consistent with Gilson et al. study, where after initial accumulation of H358 and HeLa cells in mitosis as a result of pyrrolopyrimidine (PP-13) treatment, the percentage of G2/M arrested cells was gradually decreasing, concomitantly with the accumulation in the sub-G1 and polyploid fraction [31]. As mentioned above, inhibition of tubulin polymerization and mitotic spindle formation may be one of the reasons for polyploid nuclei formation. Therefore, in the further study, the impact of S1, S19 and S22 compounds on mitotic spindle formation and tubulin polymerization was evaluated.

Impact on Mitotic Spindle Formation
As G2/M cell-cycle arrest is strongly correlated with microtubule disruption [32], in the next step the effect of tested compounds on mitotic spindle formation was evaluated using confocal microscopy imaging. After 6 h-incubation of A375 cells with S1, S19 and S22, the cells were fixed and α-tubulin and chromosomes were stained. At first it should be emphasized that the morphological differences between treated and untreated cells were noticeable only during mitosis, confirming mitotic-specific activity of tested compounds. Similarly, spindle microtubules are more sensitive to colchicine than the interphase microtubules [33], whereas taxanes and vinca-alkaloids also affect interphase cells [34]. Cell-permeable small molecules that perturb mitosis without effects on microtubule/cytoskeleton in interphase may be useful in anticancer therapy and contribute to less pronounced adverse effects. For example, as differentiated neurons are in a postmitotic state and are not be able to re-enter the cell cycle [35], this phase-specific activity may contribute to reduced neurotoxicity, which is one of the drawbacks of classical MTAs.
As depicted in Figure 4A, the control cells in mitosis showed bipolar spindle formation with chromosomes arrangement in the central metaphase plate. In contrast, compound S1, S19 and S22 caused improper changes in mitotic spindle organization ( Figure 4B-D). We could describe the three most common abnormal phenotypes: (1) bipolar spindles with chromosomes arrangement in the central metaphase and additional "cloud" of chromosomes around of centrioles (misalignment chromosomes), (2) monoastral spindles, (3) multipolar asymmetric asters ( Figure 5). Observed aberrations were previously associated with strong defects in chromosome congression and alignment, and may be connected with prometaphase blockade [31]. Multipolar and "bipolar misaligned" spindles were described by Jaunky et al. in cells treated with colchicine and C75 (new CBSIs), depending on concentration and (in the case of C75) cell line used. The authors supposed that the discrepancy in mitotic phenotypes caused by C75 between different cell lines might be connected with genetically-dependent sensitivity to spindle disruption [36]. However, observed aberrations are not only specific for tubulin inhibitors but also other compounds targeting mitotic spindle. Namely, bipolar spindles with chromosomes aligned at metaphase and misaligned chromosomes located close to the spindle poles are also characteristic for inhibitors of kinesin-like protein (KIFC1) and centromere-associated protein E (CENP-E), kinesins essential for bipolar spindle formation and correct chromosome alignment, respectively [37,38]. Therefore, it is not possible to precisely define the molecular target of currently tested compounds on the base of observed phenotypes.
In agreement with the flow cytometry analysis, the number of huge polynucleated cells and apoptotic bodies increased after 24 h of treatment ( Figure 6). Observed changes are characteristic for mitotic cell death (MCD) often called a mitotic catastrophe. Typical features of this process include block in mitosis, mitotic spindle disorganization, failed chromosome segregation, and formation of multinucleated cells [39,40]. Mitotic catastrophe has been described as predominant form of cell death mediated by CA-4 and CA-4 analogue [41,42].

Tubulin Polymerization in Cells and Cell-Free Conditions
The observed aberrations in mitotic spindle morphology encouraged us to verify whether this effect was a result of interaction with microtubule dynamics, which is essential for a proper mitotic process [43]. A direct effect of S1, S19 and S22 derivatives on tubulin behavior in cell free conditions was evaluated using a fluorescence-based tubulin polymerization assay. The activity of the tested compounds was compared with paclitaxel (PTX), CA-4 and 0.2% DMSO as references and control, respectively. As shown in Figure 7, DMSO in the control sample had no direct effect on spontaneous tubulin self-assemble, where three typical phases of microtubule formation (namely: nucleation, growth and steady state inhibition) could be observed. The reference PTX (10 µM) eliminated the nucleation phase and enhanced the V max (maximal slope values for the growth phase), whereas CA-4 (10 µM) inhibited tubulin polymerization, resulting in a decrease in V max and reduction in final polymer mass protein. The overlapping curves of CA-4 (10 µM) and tested compounds at 50 µM and 100 µM indicated that S1, S19 and S22 act as tubulin polymerization inhibitors. However, inhibition of tubulin assembly by the tested compounds observed only at higher concentrations points that they are less active than reference drug in this regard. At 10 µM (green curves in Figure 7) S1 was the most effective among the tested compounds, as delayed start of tubulin assembly, reduced microtubule polymerization speed and total polymer mass could still be clearly seen at that concentration. At 1 µM none of the investigated derivatives had any significant effect on tubulin polymerization (all purple curves finally overlapped with control in Figure 7). These observations were confirmed by IC 50 evaluation presented in Table 3.   Representative confocal microscopy images of A375 cells after 6 h incubation with medium and 0.2% DMSO (A), compound S1 (B), S19 (C), S22 (D) at concentration of 10 µ M, with stained α-tubulin (green) and chromosomes (blue); magnification ×20 plus zoom ×3 during acquisition.  . Representative confocal microscopy images of A375 cells after 24 h incubation with compound S1 (A), S19 (B), S22 (C) at concentration of 10 μM, with stained α-tubulin (green) and chromosomes (blue); magnification ×40 plus zoom ×2 during acquisition.

Tubulin Polymerization in Cells and Cell-Free Conditions
The observed aberrations in mitotic spindle morphology encouraged us to verify whether this effect was a result of interaction with microtubule dynamics, which is essential for a proper mitotic process [43]. A direct effect of S1, S19 and S22 derivatives on tubulin behavior in cell free conditions was evaluated using a fluorescence-based tubulin polymerization assay. The activity of the tested compounds was compared with paclitaxel (PTX), CA-4 and 0.2% DMSO as references and control, respectively. As shown in Figure  7, DMSO in the control sample had no direct effect on spontaneous tubulin self-assemble, where three typical phases of microtubule formation (namely: nucleation, growth and steady state inhibition) could be observed. The reference PTX (10 µ M) eliminated the nucleation phase and enhanced the V max (maximal slope values for the growth phase), whereas CA-4 (10 µ M) inhibited tubulin polymerization, resulting in a decrease in V max and reduction in final polymer mass protein. The overlapping curves of CA-4 (10 µ M) and tested compounds at 50 µ M and 100 µ M indicated that S1, S19 and S22 act as tubulin polymerization inhibitors. However, inhibition of tubulin assembly by the tested compounds observed only at higher concentrations points that they are less active than reference drug in this regard. At 10 µ M (green curves in Figure 7) S1 was the most effective among the tested compounds, as delayed start of tubulin assembly, reduced microtubule polymerization speed and total polymer mass could still be clearly seen at that concentra- Figure 6. Representative confocal microscopy images of A375 cells after 24 h incubation with compound S1 (A), S19 (B), S22 (C) at concentration of 10 µM, with stained α-tubulin (green) and chromosomes (blue); magnification ×40 plus zoom ×2 during acquisition. 1.08 ± 0.33 Figure 7. Effects of S1 (A), S19 (B) and S22 (C) derivatives on tubulin polymerization in cell free conditions. Increasing relative fluorescence units is indicative of tubulin polymerization. Paclitaxel (PTX) and combretastatin (CA-4) were used as references and DMSO (0.2%) as a vehicle control. Curves of tested compounds represent the mean from at least three independent experiments. Error bars are not shown for the sake of clarity.
The obtained data are in line with the results from another tubulin polymerization assay used (based on western blot method). Analysis was limited to the single concentration of the tested compounds (10 µ M). As expected, CA-4 and PTX (used as references), increased the amount of tubulin in the soluble or polymeric fraction, respectively (Figure 8 A). As shown in Figure 8 B, α-tubulin polymer fraction was slightly but not significantly reduced in S1-treated A375 cells when compared with the control (24.84% vs. 33.36%). One could expect that the observed effect would be more distinct at higher concentrations, Figure 7. Effects of S1 (A), S19 (B) and S22 (C) derivatives on tubulin polymerization in cell free conditions. Increasing relative fluorescence units is indicative of tubulin polymerization. Paclitaxel (PTX) and combretastatin (CA-4) were used as references and DMSO (0.2%) as a vehicle control. Curves of tested compounds represent the mean from at least three independent experiments. Error bars are not shown for the sake of clarity.
The obtained data are in line with the results from another tubulin polymerization assay used (based on western blot method). Analysis was limited to the single concentration of the tested compounds (10 µM). As expected, CA-4 and PTX (used as references), increased the amount of tubulin in the soluble or polymeric fraction, respectively ( Figure 8A). As shown in Figure 8B, α-tubulin polymer fraction was slightly but not significantly reduced in S1-treated A375 cells when compared with the control (24.84% vs. 33.36%). One could expect that the observed effect would be more distinct at higher concentrations, as our previously published data showed a reduction of polymerized tubulin fraction to 0.58% in S1 exposed cells at 100 µM [9]. Compounds S19 and S22 did not influence tubulin polymerization as evaluated by a western blot method. as our previously published data showed a reduction of polymerized tubulin fraction to 0.58% in S1 exposed cells at 100 µ M [9]. Compounds S19 and S22 did not influence tubulin polymerization as evaluated by a western blot method.

Colchicine-Binding Site Study
The aforementioned common behavior of our compounds and colchicine encouraged us to verify the possible binding mode. For this purpose a molecular docking study was performed. Colchicine binds to tubulin at the interface between the α and β subunits. The majority of the compound is buried in the β subunit. As revealed by X-ray analysis of the tubulin-DAMA/colchicine complex structure, the A and C colchicine rings interact with the β-subunit of tubulin, while the B ring of the side chain interacts with the α-subunit. One of its methoxy oxygens of trimethoxyphenyl (TMP) moiety (A ring) is involved in hydrogen bonding with the thiol of Cys241β [2,44]. The Cys241β has been previously indicated as an important target for colchicine site agents [45]. According to Nguyen et al., all of the 15 structurally diverse CBSIs formed hydrogen bond with Cys241β (Cys239β in that study) [46]. Therefore, we focused on this crucial residue during molecular docking study. As shown in Figures S7 and S9 the B-ring of S1 and S22 (first possible mode of action) or A-ring (thiophene) of S22 ( Figure S10, second possible docking model) were directed toward Cys239β, but did not form hydrogen bond with sulphur atom of this amino acids. Bhattacharyya et al. indicated that the pi electron clouds of the thiophene and the phenyl rings of nocodazole (well known CBSIs) can act as hydrogen bond acceptors by forming a weak hydrogen bond coupled with hydrogen bond donor such as the thiol group of Cys239β [33]. However, recently data revealed that nocodazole did not interact with the α-tubulin and was situated deeper in β-tubulin binding pocket overlapping very little with colchicine in the regard of binding site [47]. Our findings confirm a crucial

Colchicine-Binding Site Study
The aforementioned common behavior of our compounds and colchicine encouraged us to verify the possible binding mode. For this purpose a molecular docking study was performed. Colchicine binds to tubulin at the interface between the α and β subunits. The majority of the compound is buried in the β subunit. As revealed by X-ray analysis of the tubulin-DAMA/colchicine complex structure, the A and C colchicine rings interact with the β-subunit of tubulin, while the B ring of the side chain interacts with the α-subunit. One of its methoxy oxygens of trimethoxyphenyl (TMP) moiety (A ring) is involved in hydrogen bonding with the thiol of Cys241β [2,45]. The Cys241β has been previously indicated as an important target for colchicine site agents [46]. According to Nguyen et al., all of the 15 structurally diverse CBSIs formed hydrogen bond with Cys241β (Cys239β in that study) [47]. Therefore, we focused on this crucial residue during molecular docking study. As shown in Figures S7 and S9 the B-ring of S1 and S22 (first possible mode of action) or A-ring (thiophene) of S22 ( Figure S10, second possible docking model) were directed toward Cys239β, but did not form hydrogen bond with sulphur atom of this amino acids. Bhattacharyya et al. indicated that the pi electron clouds of the thiophene and the phenyl rings of nocodazole (well known CBSIs) can act as hydrogen bond acceptors by forming a weak hydrogen bond coupled with hydrogen bond donor such as the thiol group of Cys239β [33]. However, recently data revealed that nocodazole did not interact with the α-tubulin and was situated deeper in β-tubulin binding pocket overlapping very little with colchicine in the regard of binding site [48]. Our findings confirm a crucial role of sulphur atom and -NH group of the pyridine-2(1H)thione (C-ring) moiety in hydrogen bond formation. Namely, the sulphur atom of S1 showed hydrogen bonding interaction with amino acids of tubulin (Asn101α) ( Figure S7). The S19 formed two potential hydrogen bonds between sulphur atom and -NH group of compound and Gln247α and Ala354β, respectively ( Figure S8). Whereas S22 could form only one hydrogen bond between -S and Gln247β ( Figure S9) or -NH and Val238β ( Figure S10). Our data are in line with previous studies that pointed to the sulphur atom as an essential factor for anticancer activity of monastrol and other thio-derivaties (3,4-dihydropyrimidin-2(1H)-thiones). The substitution of sulphur by an oxygen led to the loss of the cytotoxic activity, probably due to the "soft nature" of the sulphur atom which makes compounds more nucleophilic [49].
These molecular docking results were further supported by a fluorescence-based competitive binding assay. Namely, we evaluated the ability of compounds S1, S19 and S22 to compete with colchicine which intrinsic fluorescence increases upon binding to tubulin. We chose two the most critical concentration of tested compounds, selected during the cell free tubulin polymerization study (10 µM and 50 µM). As shown in Figure 9 the fluorescence of colchicine-tubulin complex was reduced in the presence of S1, S19 and S22 in dose-dependent manner suggesting that these compounds bind at the colchicine site and may belong to CBSIs. Vinblastine (VBL, used as negative control) did not compete with colchicine because it inhibits tubulin polymerization by binding at different site, located on β-tubulin chain. However, the investigated S1, S19, and S22 compounds had lower affinity to colchicine-binding site than reference CA-4, and those findings were also in line with the results of the tubulin polymerization study described above. role of sulphur atom and -NH group of the pyridine-2(1H)thione (C-ring) moiety in hydrogen bond formation. Namely, the sulphur atom of S1 showed hydrogen bonding interaction with amino acids of tubulin (Asn101α) ( Figure S7). The S19 formed two potential hydrogen bonds between sulphur atom and -NH group of compound and Gln247α and Ala354β, respectively ( Figure S8). Whereas S22 could form only one hydrogen bond between -S and Gln247β ( Figure S9) or -NH and Val238β ( Figure S10). Our data are in line with previous studies that pointed to the sulphur atom as an essential factor for anticancer activity of monastrol and other thio-derivaties (3,4-dihydropyrimidin-2(1H)-thiones). The substitution of sulphur by an oxygen led to the loss of the cytotoxic activity, probably due to the "soft nature" of the sulphur atom which makes compounds more nucleophilic [48]. These molecular docking results were further supported by a fluorescence-based competitive binding assay. Namely, we evaluated the ability of compounds S1, S19 and S22 to compete with colchicine which intrinsic fluorescence increases upon binding to tubulin. We chose two the most critical concentration of tested compounds, selected during the cell free tubulin polymerization study (10 µ M and 50 µ M). As shown in Figure 9 the fluorescence of colchicine-tubulin complex was reduced in the presence of S1, S19 and S22 in dose-dependent manner suggesting that these compounds bind at the colchicine site and may belong to CBSIs. Vinblastine (VBL, used as negative control) did not compete with colchicine because it inhibits tubulin polymerization by binding at different site, located on β-tubulin chain. However, the investigated S1, S19, and S22 compounds had lower affinity to colchicine-binding site than reference CA-4, and those findings were also in line with the results of the tubulin polymerization study described above. Figure 9. Colchicine competitive binding assay. Results have been shown as intrinsic fluorescence of colchicine-tubulin complex normalized to control (0.2% DMSO). CA-4 was used as a positive control whereas vinblastine (VBL), which binds at vinblastine site, was used as negative control. Data are expressed as the mean and SD from three independent experiments for S1, S19, S22, CA-4 and two independent experiments for VBL; * p < 0.05 vs. ctrl; # p < 0.05 vs. CA-4. Figure 9. Colchicine competitive binding assay. Results have been shown as intrinsic fluorescence of colchicine-tubulin complex normalized to control (0.2% DMSO). CA-4 was used as a positive control whereas vinblastine (VBL), which binds at vinblastine site, was used as negative control. Data are expressed as the mean and SD from three independent experiments for S1, S19, S22, CA-4 and two independent experiments for VBL; * p < 0.05 vs. ctrl; # p < 0.05 vs. CA-4.

Conclusions
The current study is a part of wider research related to an increasing interest in nitrogen-and sulphur-based heterocycles as promising anticancer agents [10]. A structureactivity relationship study resulted in identification and detailed characterization of 3,4dihydropyridine-2(1H)-thione derivative with thiophene ring at C5 pyridine-2(1H)-thione (S22) with higher antiproliferative activity (IC 50 equal 1.71 ± 0.58 µM) and better selectivity (SI = 21.09) against A375 cells than the starting compound S1 (IC 50 = 4.33 ± 1.00 µM, SI = 6.65). Moreover, all three the most active compounds from the antiproliferative study, namely S1, S19 and S22 showed better selectivity against A375 cancer cells than reference CA-4, suggesting their possible lower toxicity and wider therapeutic index. Additionally, mitotic-specific activity of the tested compounds (observed during cell cycle analysis and confocal microscopy imaging), might potentially contribute to decreased side effects compared with classical microtubule targeted agents (MTAs). Further study revealed that all the three tested compounds transiently arrested cells in mitosis, leading to apoptotic cell death. Moreover, we proved that compounds S1, S19 and S22 disrupted mitotic spindle organization and inhibited tubulin polymerization via colchicine binding site in dose-dependent manner. However, direct activity against tubulin was not as high as could be expected from results of cell-based study. Hence, obtained findings may suggest more complexed mechanism of action. The higher anticancer selectivity of S22 compared with the reference microtubule inhibitor (CA-4) may be also connected with multitarget mode of action, including not only microtubules but also kinesins and kinases involved in mitosis and overexpressed in cancer cells [50,51]. The cooperation of kinesins and kinases in microtubule assembly is well documented and pointed as a new opportunity to design potential multitarget antimitotic inhibitors [52]. For example nocodazol, a colchicine binding site agent was found to simultaneously inhibit various cancer-related kinases, including ABL1, c-KIT, BRAF, MEK1 (MAP2K1), MEK2 (MAP2K2), and MET [53]. Frezzato et al., demonstrated that a remarkable ability of nocodazole to discriminate between leukaemic and normal cells was connected with inhibition of Src-kinase Lyn, overexpressed and crucially involved in the neoplastic B-cell survival [54]. We supposed that anticancer selectivity of compound S22 may be also connected with the fact that distribution of βtubulin isotypes varies among different tumor and normal cells, which modulates their sensitivity to chemotherapeutic drugs [55,56]. However, both hypothesis should be verified in a further study.

Experimental Part
Melting points were determined on a Boetius hot stage apparatus. 1 H and 13 C NMR spectroscopic measurements were performed on a Bruker DPX 400 Avance III HD spectrometer (Billerica, MA, USA, operating at 400.2 and 100.6 MHz, respectively. TMS was used internal standard, δ H,C = 0 ppm). For NMR analyses MestReNova (version: 12.0.4 Santiago de Compostela, Spain) program was used. For detailed peak assignments, 2D spectra were acquired using Bruker software ( 1 H, 1 H DFQCOSY; 13

Synthesis of Compounds S1-S22
General procedure for the synthesis of 5-aryl-4-benzylpyridine-2(1H)-thiones S1-S22 by the oxygen-sulphur exchange: To a solution of aryl-and benzyl-substituted pyridin-2(1H)-one (0.308 g, 1.2 mmol) in anhydrous toluene (40 mL) Lawesson reagent (Fluorchem, Derbyshire, United Kingdom; 0.27 g, 0.66 mmol) was added and the mixture was stirred for 4 h at 80 • C (oil bath temperature). After this time the mixture was cooled to room temperature and most of the solvent (2/3 volume) was distilled off in vacuo and the remaining mixture was twice chromatographed on silica gel. The obtained solids were crystallized from appropriate solvents.

Antiproliferative Activity
The antiproliferative activity of S1 and S1 analogues (S2-S22), with combretastatin (CA-4, Sigma-Aldrich Merck Group, St. Louis, MO, USA) as reference, were evaluated using the Cell Proliferation Reagent WST-1 assay (Sigma-Aldrich Merck Group, St. Louis, MO, USA). It should be noted that this kit is commonly used for proliferation measurement, but it does not directly measure DNA replication, and may be influenced by cell metabolic activity. WST-1 assay is based on the reduction of the tetrazolium salt to formazan which amount is directly correlated to the number of metabolically active cells. In the present study, MCF7, A375, HT-29, SK-OV-3 and PC-3 cells were seeded in 96-well plate (at plating densities depending on the doubling time of individual cell lines) and then cultured in appropriate medium in standard conditions. After 24 h the culture medium was removed and replaced with fresh medium containing compounds at final concentrations equal: 1, 10, 50, 100 µM for 48 h. All the tested compounds were dissolved in DMSO purchased from Sigma-Aldrich Merck Group, St. Louis, MO, USA (final concentrations of DMSO did not exceed 0.2%; initially impact of 0.2% DMSO on cell viability was negatively verified). The cells without tested compounds (but in medium with DMSO) were used as control and tested compounds in medium without cells as blank. After 48 h, WST-1 reagent was added, incubated with the cells for 30 min and absorbance was measured at 450 nm (with 620 nm background correction), using a spectrophotometric microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland). The cell viability was calculated using the following formula: [(Atest − Ablank)/(Acontrol − Ablank)] × 100%. The readings were acquired from at least three independent experiments (each conducted in triplicate). The IC 50 values (the inhibitory concentration causing 50% growth inhibition) were evaluated using an online calculator (AAT Bioquest, Inc., Quest Graph™ IC50 Calculator (v.1). Retrieved from: https://www.aatbio.com/tools/ic50-calculator-v1 (accessed on: 10 September 2020).

Selectivity-Index
The malignant melanoma (A375) and non-tumor melanocytes (HEM) cells were used to estimate the anticancer selectivity of the three the most active compounds from the earlier study (S1, S19 and S22; CA-4 was used as reference). The WST-assay and IC 50 calculation were performed, as described above. Then the selectivity index (SI) was calculated according to the following formula: SI = IC 50 normal cells/IC 50 cancer cells.

Apoptosis Detection
A375 and HEM cells were seeded in 24-well plates (1 × 10 4 and 2 × 10 4 cells/well, respectively). After 24 h, the culture medium was removed and the cells were treated with S1, S19, S22 and CA-4 (as reference) at final concentrations of 10 µM and incubated for another 48 h. Untreated cells growing in medium with DMSO (0.2%) was used as control. After treatment the cells were washed with PBS (both the cell medium and PBS were collected) and harvested by trypsinization. After centrifugation, the supernatant was discarded and cells pellet was resuspended in appropriate medium. Next, equal volume of the Muse Annexin V & Dead Cell Reagent (Luminex Corporation, Austin, TX, USA) was added, mixed thoroughly by pipetting up and down and stained for 20 min (at RT in the dark). Samples were analyzed by Muse Cell Analyzer (Luminex Corporation, Austin, TX, USA). The assay is based on the detection of phosphatidylserine (PS) on the surface of apoptotic cells by binding with Annexin V-PE. A dead cell marker (7-AAD) is also used as an indicator of cell membrane structural integrity. Simultaneous staining with these two dyes allow to distinguish four populations of cells: viable (Annexin V-PE− and 7-AAD−), early apoptotic (Annexin V-PE+ and 7-AAD−), late apoptotic/already dead cells (Annexin V-PE+ and 7-AAD+), mostly nuclear debris (Annexin V-PE− and 7-AAD+). According to the manufacturer's protocol (on the base of cell size index) debris were excluded from further analysis. The readings were acquired from at least three independent experiments. Additionally, optical microscopy imaging of A375 and HEM cells was performed just before apoptosis analysis of harvested cells using Smart Fluorescent Cell Analyzer Microscope JuLi (Seoul, Korea).

Cell Cycle Analysis
For cell cycle analysis, A375 cells were seeded in 6-well plates (1 × 10 5 cells/well) and cultured for 24 h. Then the medium was removed and replaced with fresh medium containing S1, S19, S22 and CA-4 at final concentrations of 10 µM and incubated for additional 6 and 24 h. Afterwards, cells were washed with PBS and harvested by trypsinization. Then, cell pellet was washed with PBS and fixed with 70% cold ethanol at −20 • C for at least 3 h. Before the analysis ethanol-fixed cells were centrifuged, washed with PBS and stained with Muse Cell Cycle Reagent (Luminex Corporation, Austin, TX, USA), containing propidium ioide (PI) and RNAse, for 30 min at RT in the dark. The assay utilizes PI-based staining of DNA content to discriminate and measure the percentage of cells in each cell cycle phase (G0/G1, S, and G2/M) using a Muse Cell Analyzer (Luminex Corporation, Austin, TX, USA). The readings were acquired from three independent experiments.

Confocal Microscopy Imaging
To visualize the microtubule and DNA, A375 cells were seeded on cover slips at density of 1 × 10 5 cells/well and cultured as above for 24 h. Afterwards, 10 µM of compounds: S1, S2 and S22 were added, and incubated for additional 6 h. Next, the cells were fixed in 4% buffered formalin for 5 min at 37 • C, washed twice with PBS and permeabilized with 0.01% Triton X-100 (Sigma-Aldrich Merck Group, St. Louis, MO, USA) in PBS for 10 min at RT. Nonspecific antibody binding was blocked (0.01% Triton X-100, 2% BSA, 1.5% FBS in PBS) for 10 min (RT) and subsequently incubated with mouse anti-α-tubulin monoclonal antibody (Sigma-Aldrich Merck Group, St. Louis, MO, USA), diluted 1:2000 for overnight (4 • C). The cover slips were then washed tree times with 0.05% Triton X-100 in PBS and the cells were incubated with secondary antibodies anti-mouse IgG-FITC (Sigma-Aldrich Merck Group, St. Louis, MO, USA) diluted 1:64 for 1 h at RT, washed and counterstained with DAPI (Sigma-Aldrich Merck Group, St. Louis, MO, USA). Finally, cells were washed with PBS and mounted with mounting medium (Dako-Agilent, CA, USA) on glass slides. The slides were examined with a FV1000 confocal microscope (Olympus, Hamburg, Germany) in two separated channels: for DAPI (405 nm laser), FITC (488 nm laser). To calculate the proportion of different mitotic spindle phenotypes cells in mitosis from ten random fields (×20 objective magnification) of the slide were counted (from two independent experiments).

Cell Free Tubulin Polymerization Assay
Tubulin polymerization was monitored using a >99% pure tubulin and fluorescencebased tubulin polymerization kit (Cytoskeleton, Denver, CO, USA), where tubulin polymerization is followed by fluorescence enhancement due to the incorporation of a fluorescence reporter into microtubules as polymerization proceeds. Compounds S1, S19 and S22 (at final concentrations: 100, 50, 10 and 1 µM) were evaluated for their effect on tubulin polymerization (in cell free conditions). Paclitaxel (Cytoskeleton, Denver, CO, USA) and CA-4 (10 µM) were used as references and DMSO (0.2%) as vehicle control. According to the manufacturer's protocol after incubation of tested compounds at 37 • C for 1 min, the icy tubulin reaction mixture (2 mg/mL tubulin, 1.0 mM GTP, 15% glycerol in buffer containing 80 mM PIPES, 2.0 mM MgCl2, 0.5 mM EGTA, pH 6.9 and 10 µM fluorescent reporter) was added. The samples were mixed and tubulin assembly was monitored (excitation: 360 nm, emission: 450 nm) at 1 min intervals for 1 h at 37 • C using a spectrophotometric microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland). The inhibition of tubulin polymerization was expressed as IC 50 values and quantified according to Cytoskeleton, Inc. instruction. Firstly, maximal slope (V max) of the growth phase expressed as RFU/min was calculated using a software installed on the aforementioned plate reader. Next, V max of each polymerization curve was normalized to V max of appropriate control's curve and IC 50 values were evaluated using an online calculator (AAT Bioquest, Inc. Quest Graph™ IC50 Calculator. Retrieved from https://www.aatbio.com/tools/ic50-calculator (accessed on: 16 February 2021).

Cell-Based Tubulin Polymerization Assay
Western blot based tubulin polymerization assay was modified by previously described method [57,58]. A375 cells were seeded in 6-well plates (5 × 10 4 cells/well), grown to subconfluency and next additional 6 h with 10 µM of compounds S1, S19, S22, CA-4 and PTX (as references), and DMSO (0.2%) as control. After treatment, cells were washed twice with warm PBS, lysed shaking for 5 min with hypotonic buffer (20 mM Tris-HCl pH 6.8, 1 mM MgCl 2 , 2 mM, EGTA, 0.5% NP-40; Sigma-Aldrich Merck Group, St. Louis, MO, USA) containing 1X protease inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA, USA), then scraped and transferred to tubes. The samples were centrifuged at 13,200 rpm for 10 min at RT, and supernatants containing soluble (cytosolic) tubulin were separated from the pellets containing polymerized (cytoskeletal) tubulin. The pellets were resuspended in equal volumes of hypotonic buffer. Total protein concentration was then determined in each fraction by BCA method (Thermo Fisher Scientific, Waltham, MA, USA). The equal amount of protein (6 µg per well) from supernatant and pellet fractions were resuspended in 2× Laemmli buffer (Bio-Rad, Hercules, CA, USA), heated for 5 min at 95 • C and analyzed by SDS-PAGE using 4-20% mini-PROTEAN electrophoresis system (Bio-Rad, Hercules, CA, USA), then transferred to PVD membrane (Bio-Rad, Hercules, CA, USA). Immunoblotting was performed using a primary mouse anti-α-tubulin monoclonal antibody (Sigma-Aldrich Merck Group, USA) diluted 1:1000 for overnight (4 • C) and secondary mouse IgGκ BP conjugated to horseradish peroxidase (HRP) (Santa Cruz Biotechnology, Dallas, TX, USA) diluted 1:2000. The obtained products were detected using the Western Bright Sirius Chemiluminescent Detection Kit (Advansta, San Jose, CA, USA) and bands were subsequently visualized in UVP camera (UVP BioImaging system EpiChemi 3, Houston TX, USA). Equal loading in the lanes was evaluated by stripping the blots according to manufacturer's protocol (Restore Western Blot Stripping Buffer, Thermo Fisher Scientific, Waltham, MA, USA) and reprobing them with mouse β-actin monoclonal antibody 1:1000 (Santa Cruz Biotechnology, Dallas, TX, USA) and then secondary HRP-conjugated antibody (as above). Percentage of α-tubulin in the pellet fractions was calculated as a densitometric value of the pellet band divided by the total densitometric value of the pellet and supernatant bands using ImageJ Software 1.53c (Maryland, VA, USA). Each densitometric value of tubulin was before normalized to the densitometric value of β-actin The readings were acquired from four independent experiments.

Colchicine Binding-Site Assay
Colchicine binding-site assay was performed as previously described [44] by using >99% pure tubulin protein (Cytoskeleton, Denver, CO, USA) reconstituted in buffer containing 80 mM PIPES pH 6.9, 2 mM MgCl 2 , 0.5 mM EGTA supplemented with 1 mM GTP. The compounds S1, S19 and S22 at 50 and 10 µM were incubated with 3 µM tubulin in the presence and absence of 3 µM colchicine (Sigma-Aldrich Merck Group, St. Louis, MO, USA) in 30 mM Tris buffer for 60 min at 37 • C. CA-4 (10 µM) was used as a positive control whereas vinblastine (VBL, Sigma-Aldrich Merck Group, St. Louis, MO, USA) binding at distinct site on tubulin (10 µM) was used as negative control. This method is based on the fact that the intrinsic fluorescence of colchicine increases upon binding to tubulin. Hence, tubulin binding site inhibitors that compete with colchicine may decrease fluorescence of colchicine-tubulin complex. After incubation fluorescence was determined (excitation 350 nm, emission 435 nm) by using a spectrophotometric microplate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland). The raw fluorescence values were normalized first by subtracting the fluorescence of the blank (tubulin with tested compounds without colchicine) and setting the fluorescence of 3 µM tubulin with 3 µM colchicine to 100% (control with 0.2% DMSO).

Molecular Docking
Computational docking was carried out using the genetic algorithm-based ligand docking program Chimera and AutoDock Vina version 1.1.2 (San Diego, CA, USA). The X-ray crystal structure of αβ-tubulin complexed colchicine (PDB code: 4o2b) was used in this study. The stathmin-like domain and the C and D subunits were removed from the model. Binding modes for each colchicine analogue were investigated to understand the steric, electrostatic, and hydropathic features of the colchicine binding site. The top 10 conformations with different scores (between −5 to −9) were acquired, and the best ranking pose was visualized.

Statistical Analysis
Results are presented as mean ± standard deviation (SD). Statistical analysis was carried out using Statistica 13.3 (StatSoft Inc., Tulsa, OK, USA) and ANOVA test followed by a post hoc Tukey's multiple comparison. A p-value level of < 0.05 was considered statistically significant.
Supplementary Materials: The following are available online at https://www.mdpi.com/1422-0 067/22/5/2462/s1, Figure S1. Representative scatter diagrams of A375 cells exposed to medium with 0.2% DMSO (A), CA-4 (B), S1 (C), S19 (D), S22 (E) at 10 µM for 48 h. Figure S2. Representative scatter diagrams of HEM cells exposed to medium with 0.2% DMSO (A), CA-4 (B), S1 (C), S19 (D), S22 (E) at 10 µM for 48 h; Figure S3: Optical microscopy images of A375 cells, controls (A) and after 48 h incubation with 10 µM of compounds: CA-4 (B), S1(C), S19 (D) and S22 (E); Figure S4: Optical microscopy images of HEM cells, controls (A) and after 48 h incubation with 10 µM of compounds: CA-4 (B), S1(C), S19 (D) and S22 (E); Figure S5. Representative individual experiment of cell cycle phase distribution of A375 cells exposed to medium with 0.2% DMSO (A), CA-4 (B), S1 (C), S19 (D), S22 (E) for 6 h and DMSO (F), CA-4 (G), S1 (H), S19 (I), S22 (J) for 24 h. Figure   2-Methoxy-5-phenylpyridine [59] (1) was prepared on multigram scale as follows: To the mixture of DMF (300 mL) and water (300 mL) in a 1 l flask 5-bromo-2methoxypyridine (20 g, 0.106 mol), arylboronic acid (14.3 g, 0.117 mol), (K 3 PO 4 . 5H 2 O (22 g; 0.0827 mol) and PdCl 2 (0.374 g, 2.1 mmol). The resulting solution was stirred for 22 h at room temperature in open flask. After this time brine (300 mL) was added and the aqueous layer was extracted with diethyl ether (4 × 300 mL) and the organic layer were washed with brine (100 mL) and dried with MgSO 4 . The mixture was filtered, and the solvents were evaporated under reduced pressure. The crude product was purified by vacuum distillation (bp 68-70 • C, 0.01 mbar) afforded 1 in 95% yield (18.57 g) as colorless oil. 1  To a 100-mL Schlenk flask, equipped with a magnetic stir bar and a crowned with argon balloon, charged with 40 mL of degassed DMF, 2-iodo (Apollo Scientific, 1 g, 4.76 mmol), 2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxa yl)pyridin [67] (1.679 g, 7.14 mmol, 1.5 eqiv.) and 0.275 g tetrakis(triphenylphos ladium(0) (Sigma-Aldrich Merck Group, St. Louis, MO, USA) was added, and t was stirred at rt for 60 min. Next 21.5 mL of degassed 1.0 M aqueous Na2CO3 so added, and the reaction mixture was heated under argon at 80 °C for 3 h. Afte the reaction mixture was cooled to rt. and solvent was removed under reduced The residue was extracted by ethyl acetate, and the organic layer was washed and dried over MgSO4. The crude product was purified by column chromato silica gel using a mixture n-hexane and ethyl acetate (9:1 v/v) as the eluent, yi yellow oil 0.827g (90%). 1  To a 100-mL Schlenk flask, equipped with a magnetic stir bar and a condenser crowned with argon balloon, charged with 40 mL of degassed DMF, 2-iodothiophene (Apollo Scientific, 1 g, 4.76 mmol), 2-methoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)pyridin [67] (1.679 g, 7.14 mmol, 1.5 eqiv.) and 0.275 g tetrakis(triphenylphosphino)palladium(0) (Sigma-Aldrich Merck Group, St. Louis, MO, USA) was added, and the mixture was stirred at rt for 60 min. Next 21.5 mL of degassed 1.0 M aqueous Na 2 CO 3 solution was added, and the reaction mixture was heated under argon at 80 • C for 3 h. After this time the reaction mixture was cooled to rt. and solvent was removed under reduced pressure. The residue was extracted by ethyl acetate, and the organic layer was washed with brine, and dried over MgSO 4. The crude product was purified by column chromatography on silica gel using a mixture n-hexane and ethyl acetate (9:1 v/v) as the eluent, yielding 8 as yellow oil 0.827g (90%). 1  The residue was extracted by ethyl acetate, and the organic layer wa and dried over MgSO4. The crude product was purified by column silica gel using a mixture n-hexane and ethyl acetate (9:1 v/v) as the yellow oil 0.827g (90%  13  Method B: A solution of 2-methoxy-5-arylpyridine (0.0106 mol), LiCl (2.24 g, 0.053 mol) and PTSA (9.1 g, 0.053 mol) in DMF (35 mL) was stirred at 120 • C for 3 h. After this time, next portion of LiCl (2.97 g, 32.5 mmol) and PTSA (9.1 g, 0.053 mol) was added and the mixture was heated next 3 h at 120 • C. After cooling the reaction mixture was quenched by the addition of water (15 mL) and extracted with EtOAc (100 mL × 3) and washed with brine (15 mL × 3). The combined organic phase was dried over Na 2 SO 4 , filtered and concentrated. The residue was purified by column chromatography (SiO 2 , ethyl acetate). 5-(3-Methoxyphenyl)pyridin-2(1H)-one [69] [5(NH)]. Yield 80%. 1 H and 13 C NMR spectra matched those previously reported [69]. Method B: A solution of 2-methoxy-5-arylpyridine (0.0106 mol), LiCl (2.24 g, 0.053 mol) and PTSA (9.1 g, 0.053 mol) in DMF (35 mL) was stirred at 120° C for 3 h. After this time, next portion of LiCl (2.97 g, 32.5 mmol) and PTSA (9.1 g, 0.053 mol) was added and the mixture was heated next 3 h at 120 °C. After cooling the reaction mixture was quenched by the addition of water (15 mL) and extracted with EtOAc (100 mL × 3) and washed with brine (15 mL × 3). The combined organic phase was dried over Na2SO4, filtered and concentrated. The residue was purified by column chromatography (SiO2, ethyl acetate). 5-(3-Methoxyphenyl)pyridin-2(1H)-one [69] [5(NH)]. Yield 80%. 1 H and 13 C NMR spectra matched those previously reported [69]. The solution prepared in the first Schlenk flask was then transferred to a cooled (0 • C) solution prepared in the second Schlenk flask. The resulting solution was stirred for 15 min at 0 • C and 120-180 min at rt (TLC or GC-MS control). After this time, the mixture was carefully quenched with saturated aqueous NH 4 Cl (30 mL). The aqueous layer was extracted with ethyl acetate (3 × 90 mL) and the combined organic layer were dried with MgSO 4 . The mixture was filtered, and the solvents were evaporated under reduced pressure. The crude product was purified by column chromatography on silica gel using appropriate solvents. NOTE: BnMgCl (2.0 M in THF) was commercially available, while α-NaphCH 2 MgCl (0.05 M in THF) was prepared and titrated according to the procedure described earlier. [70] p-MeOC 6 H 4 CH 2 MgCl in THF (A) and p-PhC 6 H 4 CH 2 MgCl in THF (B) were prepared as follows: A Schlenk flask (100 mL) equipped with a 3 cm long stir bar was heated twice with a heat gun under argon. Then magnesium turnings (Sigma-Aldrich Merck Group, St. Louis, MO, USA) (A: 1.19 g, 0.0489 mol, B: 1.52 g, 0.0624 mol), porcelain finely broken (about half the volume of magnesium used) and iodine (catalytic amount) was added and heated again three short times, next vigorous stirring was started to grind the magnesium turnings (1 h at the speed of 600 rpm). Next dry THF (A: 42 mL B: 52 mL) was added (rt). In a second Schlenk flask the solution of [A: p-MeOC 6 H 4 CH 2 Cl (1.31 g, 8.87 mmol) in THF (5 mL), B: p-PhC 6 H 4 CH 2 Cl (2.1 g, 0.01 mol) in THF (7 mL)] was prepared, and the content of the second flask was added dropwise with a syringe to a suspension of Mg in THF over 120 min stirred at the speed of 500 rpm. The content of the flask was then allowed to stand for 30 min, and after this time a clear Grignard solution was transferred into separate