First-in-Class Colchicine-Based Visible Light Photoswitchable Microtubule Dynamics Disrupting Agent

Compounds that disrupt microtubule dynamics, such as colchicine, paclitaxel, or Vinca alkaloids, have been broadly used in biological studies and have found application in clinical anticancer medications. However, their main disadvantage is the lack of specificity towards cancerous cells, leading to severe side effects. In this paper, we report the first synthesis of 12 new visible light photoswitchable colchicine-based microtubule inhibitors AzoCols. Among the obtained compounds, two photoswitches showed light-dependent cytotoxicity in cancerous cell lines (HCT116 and MCF-7). The most promising compound displayed a nearly twofold increase in potency. Moreover, dissimilar inhibition of purified tubulin polymerisation in cell-free assay and light-dependent disruption of microtubule organisation visualised by immunofluorescence imaging sheds light on the mechanism of action as microtubule photoswitchable destabilisers. The presented results provide a foundation towards the synthesis and development of a novel class of photoswitchable colchicine-based microtubule polymerisation inhibitors.


Introduction
Photopharmacology is an emerging method based on incorporation of photoswitchable component-molecular switches into the skeleton of a parent compound with expected biological activity. The goal of photopharmacology is to reduce the effects of drug substances apart from the cellular target and severe systemic/environmental side effects by establishing an external and selective means of controlling the activity of these compounds with time and spatial precision. It involves the design, synthesis, research, and application of drugs in the form of photochromic molecular switches which can be regulated by light [1][2][3][4]. Although photopharmacology is a relatively new technique that has not yet found clinical application, recent years have abounded with outstanding research towards the development of novel photoresponsive bioactive compounds including G proteincoupled receptors (GPCRs) agonists [5][6][7], ion channels activity modulators [8][9][10], and enzyme inhibitors [11]. Azobenzenes are photoswitches that can have many applications in photopharmacology. They can be switched between the (E) and (Z) configurations by light [12,13]. Due to their small size, high quantum efficiency, high extinction coefficients, low photobleaching factor, and easy synthesis, they are perfect structural elements for creating complex optical tools-they require low-intensity light and, because of their stability, they can be switched many times in a large number of cycles [14][15][16]. For example, the action of photopharmaceuticals containing skeletal fragments of azobenzene as a functionalizing unit allows for control of biological functions with precision in space and time [17][18][19][20]. Classical azobenzenes also have disadvantages that limit their practical use in biological sciences. The first is the necessity to use destructive and harmful-to-cells UV In this study, our goal was to develop novel visible light photoswitchable colchicinebased microtubule disrupting agents and assess their antiproliferative activity against selected tumorous cell lines.

Synthesis
Compound characterisation and copies of NMR spectra are provided in the Supporting Information.
General procedure A: the aqueous solution (35 mL) of oxone ® (3.5 mmol) was added, dropwise, to the solution of aniline derivative 1a-d (1 mmol) in dichloromethane (20 mL). The reaction mixture was vigorous stirred at room temperature. After disappearance of the starting material (analysed by TLC), the reaction was quenched by addition of Na-HCO3. After separation, the aqueous phase was extracted twice with DCM. The combined organic layers were dried over MgSO4 and concentrated in a vacuum. The residue containing nitrosoarene 2a-d was dissolved in acetic acid (50 mL) and appropriate isomer of aminobenzoic acid (1 mmol) was added. The reaction was stirred at room temperature for 24 h and then poured into water. The crude product was collected by filtration and recrystallized from ethyl acetate to afford the analytically pure product. For soluble products, the solvent was evaporated, and the residue was subjected to column chromatography with 1% of acetic acid in DCM used as eluent.
General procedure B: a solution of appropriate azobenzene m-, p-3a-d obtained from procedure A (0.3 mmol) in DMF (5 mL) was added to (R)-N-deacetyl colchicine (0.15 mmol), HATU (0.15 mmol), and DIPEA (0.9 mmol) under argon atmosphere. The mixture was stirred at room temperature for 4 h and then diluted with ice-cooled water and extracted with ethyl acetate (2 × 40 mL). Combined organic layers were washed with brine

Synthesis
Compound characterisation and copies of NMR spectra are provided in the Supporting Information.
General procedure A: the aqueous solution (35 mL) of oxone ® (3.5 mmol) was added, dropwise, to the solution of aniline derivative 1a-d (1 mmol) in dichloromethane (20 mL). The reaction mixture was vigorous stirred at room temperature. After disappearance of the starting material (analysed by TLC), the reaction was quenched by addition of NaHCO 3 . After separation, the aqueous phase was extracted twice with DCM. The combined organic layers were dried over MgSO 4 and concentrated in a vacuum. The residue containing nitrosoarene 2a-d was dissolved in acetic acid (50 mL) and appropriate isomer of aminobenzoic acid (1 mmol) was added. The reaction was stirred at room temperature for 24 h and then poured into water. The crude product was collected by filtration and recrystallized from ethyl acetate to afford the analytically pure product. For soluble products, the solvent was evaporated, and the residue was subjected to column chromatography with 1% of acetic acid in DCM used as eluent.
General procedure B: a solution of appropriate azobenzene m-, p-3a-d obtained from procedure A (0.3 mmol) in DMF (5 mL) was added to (R)-N-deacetyl colchicine (0.15 mmol), HATU (0.15 mmol), and DIPEA (0.9 mmol) under argon atmosphere. The mixture was stirred at room temperature for 4 h and then diluted with ice-cooled water and extracted with ethyl acetate (2 × 40 mL). Combined organic layers were washed with brine dried over MgSO 4 and evaporated under reduced pressure. The residue was purified by silica gel column chromatography (DCM/MeOH 9:1).
General procedure C: thionyl chloride (5.0 mmol) was added to a solution of 2aminobenzoic acid (1.05 mmol) in toluene (5 mL), and the mixture was refluxed for 4 h. Next, the solvent was evaporated under reduced pressure to obtain the crude acid chloride as yellow oil, which was used immediately in the next step without any purification. Et 3 N (1.05 mmol) was added to a solution of (R)-N-deacetyl colchicine (1.0 mmol) in DCM (10 mL) at 0 • C and stirred for 15 min. The solution of acid chloride in methylene chloride (5 mL) was added dropwise to the latter mixture at 0 • C and stirred overnight. Thereafter, the solvent was removed under reduced pressure and the residue was purified by column chromatography (ethyl acetate/acetone 4:1) to afford intermediate S1. In subsequent reactions, nitrosobenzenes 2a-d (obtained as in general procedure A, 0.2 mmol) was dissolved in acetic acid (5 mL) and intermediate S1 (0.15 mmol) was added in DCM (5 mL). The reaction was stirred at room temperature for 24 h and then the solvent was evaporated. The residue was dissolved in ethyl acetate (10 mL), washed with NaHCO 3 (2 × 2 mL) water (2 mL), dried over MgSO 4 , and concentrated in a vacuum. The product was purified by silica gel column chromatography (DCM/MeOH 95:5).

Tubulin Polymerisation Assay
Tubulin from porcine brain was purified according to a protocol published previously [54]. The tubulin polymerisation reaction was conducted at 3.5 mg/mL tubulin, in a tubulin polymerisation buffer (80 mM piperazine-N,N -bis(2-ethanesulfonic acid) (PIPES) pH = 6.9; 0.5 mM EGTA; 2 mM MgCl 2 ), in a 96-well plate (100 µL), in a EnSpire ® multimode plate reader (PerkinElmer, Turku, Finland) with temperature maintained at 37 • C. Tubulin was initially preincubated for 30 min at room temperature with (Z) enriched isomer (green light pre-illuminated) or with all (E) isomer (thermally adapted in dark) of o-AzoCol26DF (10 µM) in buffer with 1% DMSO, without GTP. A sample with 1% DMSO alone was used as a control. GTP was added to the concentration 1 mM, and the change in absorbance at 340 nm was monitored at 15 s intervals for 20 min. A solution of colchicine (5 µM) or cosolvent (DMSO) was used as a control.

Cell Culturing
Cells were cultured in a humidified incubator at 37 • C under 5% CO 2 . Human breast adenocarcinoma (MCF-7) cells were maintained in phenol red-free Dulbecco's Modified Eagle's Medium (Thermo Fisher Scientific, Waltham, MA, USA, 11054020) supplemented with 2 mM L-glutamine, and human colorectal carcinoma (HCT116) cells were maintained in the same medium but supplemented with 4.5 g/L glucose and 4 mM L-glutamine. All culture media were supplemented with 1% Penicillin-Streptomycin (Sigma-Aldrich, Burlington, MA, USA, P4333) and 10% fetal bovine serum (Gibco, Billings, MT, USA, 10270-106). Cells were sub-cultured at approximately 70-90% confluency to maintain the culture in the logarithmic growth phase.

MTT Cytotoxicity Assay with Green and Blue Light Irradiation
Cells were seeded in a 96-well plate at the density of 7 × 10 3 cells per well and allowed to grow for 24 h. Afterwards, the medium was aspirated, and a fresh medium was added (200 µL) with serial dilutions of tested compounds or DMSO at corresponding concentrations as a control. Following 48 h of incubation under 500 ms pulsed green or blue light irradiation every 15 s, the medium was replaced with a medium (100 µL) containing MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide, 0.5 mg/mL) and incubated at 37 • C for 4 h. The formed formazan was dissolved in DMSO (100 µL) and incubated at 37 • C for 10 min. The absorbance was measured at 540 nm. After blank subtraction, the half maximal effective concentration (IC 50 ) was calculated by GraphPad Prism software version 7 (GraphPad Software Inc., San Diego, CA, USA). Each independent experiment was performed in triplicate.

Immunofluorescence
Cells were seeded on coverslips on a 24-well plate at a density of 5 × 10 4 cells per well and allowed to grow overnight. Next, the medium was replaced with a medium containing the o-AzoCol26DF or the colchicine (without illumination) or DMSO as control and incubated for 24 h under 500 ms pulsed green or blue light irradiation every 15 s. Then, cells were fixed and permeabilised with 100% methanol at −20 • C for 15 min and subsequently washed three times with PBS at room temperature. After 1 h blocking with 3% BSA/PBS at 4 • C, slides were incubated with anti-α tubulin 12G10 antibody (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, USA) (diluted 1:300 in 3% BSA/PBS) and anti-acetylated αtubulin antibody (Cell Signalling, Danvers, MA, USA) (diluted 1:1000 in 3% BSA/PBS) overnight at 4 • C. After 3 × 10 min washing with PBS, slides were incubated with AlexaFluor555-conjugated anti-rabbit and AlexaFluor488conjugated anti-mouse secondary antibodies (diluted 1:400 in 3% BSA/PBS) (Thermo Fisher Scientific, Waltham, MA, USA, A31570) and with DAPI (50 ng/mL) for 1 h at room temperature. After washing (3 × 10 min with PBS), slides were mounted in Fluoromount-G (Southern Biotech., Birmingham, AL, USA). Images were recorded using Leica TCS SP8 (Leica Microsystems, Wetzlar, Germany) confocal microscope and analysed using ImageJ 1.53t software.

Cell Cycle Stages Analysis
Approximately 10 6 cells were plated and treated with DMSO, colchicine, or o-AzoCol26DF. Next, the medium was replaced with a medium containing solution of the o-AzoCol26DF or solution of the colchicine (without illumination) or DMSO as control and incubated for 24 h under 500 ms pulsed green or blue light irradiation every 15 s, as described above. After 24 h, cells were harvested, fixed with 70% ethanol, and stained with 50 µg/mL propidium iodide with presence of 50 µg/mL RNAse A. Stained cells were immediately analysed in Cytometer BD FACSCalibur (BD, Franklin Lakes, NJ, USA).

The Computational Study
As mentioned in the introduction, the implementation of fluorine atoms makes it possible to separate the n → π* absorption bands in the UV-VIS spectrum and to separate them from the π → π* band. This is why we first computed and analysed the geometry proposed by using compounds, o-, m-, p-3a-d for (E) and (Z) isomers, using the density functional theory (DFT). In the calculations, the B3LYP functional, 6-31G*, and basis set was employed and the continuum model (PCM; Gaussian 03W, see Supporting Information) was used to simulate the effects of the solvent, DMSO [55,56]. This method successfully reproduces the relative energies of the isomers of many azobenzene derivatives, including bridged azobenzenes [57,58]. The SCF energy for the (E) and (Z) isomers of o-, m-, p-3a-d is presented in Table S1 in the Supporting Information. For all compounds, the (E) isomer has a lower energy (mostly 55.8-59.8 kJ/mol). The smallest differences were observed for (E)-p-3d/(Z)-p-3d and (E)-m-3d/(Z)-m-3d at 43.38 kJ/mol and 45.98 kJ/mol, respectively. Moreover, for compounds (E)-o-3b/(Z)-o-3b and (E)-o-3a/(Z)-o-3a, due to the large spherical hindrance, differences are the largest at 81.08 kJ/mol and 80.61 kJ/mol, respectively. In the case of (E)-o-3d/(Z)-o-3d and (E)-o-3c/(Z)-o-3c, the hydrogen bonds between fluor of first ring and hydrogen of second ring can be observed, such bonds reducing the energy difference between the (E) and (Z) isomers (56.13 kJ/mol and 56.78 kJ/mol, respectively, see Supporting Information). The geometry of the respective photoswitch has a strong influence on the n → π* excitation energies. To determine whether the conformational changes of the photoswitch structures provide shifts in the excitation energies, we calculated the energy of (E) and (Z) orbitals HOMO and LUMO for switches o-, m-, p-3a-d (Table S2). It is known from the framework of MO theory that the lowest excited states of azobenzenes can be quite well described using singly excited n → π* and π → π* configurations [59]. The separation of n and π orbitals by symmetry is easy for the planar (E) isomers, and the relevant orbitals, i.e., π, n, and π*, are readily recognised in MO calculations also for the differently shaped (Z) isomer. Analysing the obtained data, we can observe that, for all compounds, the π* orbital level is much higher in the (Z) isomer relative to the (E) isomer. This relates to the fact that in the (Z) isomers the π-electron delocalisation is reduced due to the large dihedral angles about the N−C single bonds. The n-orbital energy level in the (Z) isomer is also much higher than in the E-isomer. This effect is connected with the linearity of the (E) isomer and the interaction of the lone pair orbitals on the two neighboring N atoms through bonds. In the case of the nonlinear (Z) isomer, the lone pair orbitals interact much more strongly through space. As described by Hecht [24,26], the repulsive interaction of the nitrogen lone pairs increases the n-level in azobenzenes and the introduction of a fluorine atom (σ-electron withdrawing groups) to aromatic ring; especially in ortho position, it should lower the n-orbital energy. It is worth noting that the n → π* excitation energies are very similar for both (E) and (Z) isomers. This conclusion is consistent with the results of Ali et al. and Hecht et al. obtained for other fluorinated compounds [24,59]. The energy differences of the (E) and (Z) orbitals HOMO and LUMO are small, ranging from 3.597 eV to 3.935 eV. The information obtained theoretically was verified by synthesis of o-, m-, p-3a-d compounds ( Figure 2) and by measuring their UV-Vis spectra (see Supporting Information).
(E) isomers, and the relevant orbitals, i.e., π, n, and π*, are readily recognised in MO c culations also for the differently shaped (Z) isomer. Analysing the obtained data, we c observe that, for all compounds, the π* orbital level is much higher in the (Z) isomer re tive to the (E) isomer. This relates to the fact that in the (Z) isomers the π-electron deloc isation is reduced due to the large dihedral angles about the N−C single bonds. The orbital energy level in the (Z) isomer is also much higher than in the E-isomer. This eff is connected with the linearity of the (E) isomer and the interaction of the lone pair orbit on the two neighboring N atoms through bonds. In the case of the nonlinear (Z) isom the lone pair orbitals interact much more strongly through space. As described by He [24,26], the repulsive interaction of the nitrogen lone pairs increases the n-level in azob zenes and the introduction of a fluorine atom (σ-electron withdrawing groups) to a matic ring; especially in ortho position, it should lower the n-orbital energy. It is wo noting that the n → π* excitation energies are very similar for both (E) and (Z) isome This conclusion is consistent with the results of Ali et al. and Hecht et al. obtained for ot fluorinated compounds [24,59]. The energy differences of the (E) and (Z) orbitals HOM and LUMO are small, ranging from 3.597 eV to 3.935 eV. The information obtained th retically was verified by synthesis of o-, m-, p-3a-d compounds ( Figure 2) and by meas ing their UV-Vis spectra (see Supporting Information).

Design and Synthesis of Azocolchicines
Our synthetic strategy towards photoswitchable azobenzamides-colchicnes AzoC is based on the replacement of the acetoamide group of a well-known and potent mic tubule disrupting agent, colchicine, with an azobenzene unit ( Figure 2). Previous d suggest that substitutions at this position are well tolerated and do not lead to loss of b activity [60,61]. Hence, we anticipated that this approach would maintain the antipro erative activity and simultaneously allow for precise spatiotemporal control of its activ with light irradiation. Recently, we explored the various synthetic methods to obtain a compounds. Utilising the optimised conditions, we focused on a one-step method w oxone synthesis [62]. We started from aniline derivatives substituted with a fluorine at at various positions 1a-d and oxidised them to corresponding nitroso compounds 2a by reaction with potassium peroxymonosulfate in biphasic dichloromethane/water so tion. Obtained nitroso derivatives were used in Baeyer−Mills reactions with ortho-, me and para-aminobenzoic acid affording azobenzenes o-, m-, p-3a-d (Scheme 1) [14]. A o pot condensation reaction between N-deacetycolchicine 4 and meta-or para-3a-d isom

Design and Synthesis of Azocolchicines
Our synthetic strategy towards photoswitchable azobenzamides-colchicnes AzoCols is based on the replacement of the acetoamide group of a well-known and potent microtubule disrupting agent, colchicine, with an azobenzene unit ( Figure 2). Previous data suggest that substitutions at this position are well tolerated and do not lead to loss of bioactivity [60,61]. Hence, we anticipated that this approach would maintain the antiproliferative activity and simultaneously allow for precise spatiotemporal control of its activity with light irradiation. Recently, we explored the various synthetic methods to obtain azo compounds. Utilising the optimised conditions, we focused on a one-step method with oxone synthesis [62]. We started from aniline derivatives substituted with a fluorine atom at various positions 1a-d and oxidised them to corresponding nitroso compounds 2a-d by reaction with potassium peroxymonosulfate in biphasic dichloromethane/water solution. Obtained nitroso derivatives were used in Baeyer−Mills reactions with ortho-, meta-, and paraaminobenzoic acid affording azobenzenes o-, m-, p-3a-d (Scheme 1) [14]. A one-pot condensation reaction between N-deacetycolchicine 4 and metaor para-3a-d isomers allowed for a straightforward synthesis of photoswitchable azobenzamides-colchicines m-, p-AzoCols. Unfortunately, reactions with ortho-3a-d resulted in a complicated, inseparable products mixture. Thus, we decided to react 4 with 2-aminobenzoyl chloride in the presence of triethylamine and, in the following step, with nitrosobenzene 2a-d. This synthetic route resulted in the desired target compounds ortho-AzoCols (Scheme 2). allowed for a straightforward synthesis of photoswitchable azobenzamides-colchicines m p-AzoCols. Unfortunately, reactions with ortho-3a-d resulted in a complicated, insepar ble products mixture. Thus, we decided to react 4 with 2-aminobenzoyl chloride in th presence of triethylamine and, in the following step, with nitrosobenzene 2a-d. This sy thetic route resulted in the desired target compounds ortho-AzoCols (Scheme 2).

Photochemical Characterisation
Photochemical properties of AzoCols photoswitches are attributed to azobenzen 3a-d moiety incorporated into the parent pharmacophore structure; therefore, we eval ated the photochemical properties of obtained azobenzenes by NMR spectroscopy an UV-Vis spectrophotometry. As a solvent, we chose DMSO due to its ability to dissolv polar and nonpolar molecules, which is crucial for the analytical methods used in th study. Additionally, its intermediate polarity allows for good approximation of organ and aqueous solvents. Furthermore, in the context of photopharmacology, DMSO is use for stock solution preparation which is illuminated and then diluted into aqueous system for biological activity assessment [63][64][65]. The photoisomerisation of azobenzenes 3a-d not altered upon condensation with colchicine (see Supporting Information Figures  and S3). We assumed that ultraviolet light would cause photocatalyzed degradation azobenzamides-colchicines AzoCols. Indeed, irradiation of p-AzoCol4F with UV lig (365 nm) lead to the formation of a complex mixture of products (see Supporting Info mation Figure S4). This result confirms that ultraviolet light is incompatible with ph toswitchable ligands based on the colchicine structure. Therefore, we determined the di tribution of (E) and (Z) isomers for azobenzenes o-, m-, p-3a-d at the photostationary sta (PSS) under constants illumination with selected wavelengths of the visible spectru (390-610 nm) by 1 H or 19 F NMR analysis (see Supporting Information Figures S6-S17). Th obtained results are summarised in Figure 3a allowed for a straightforward synthesis of photoswitchable azobenzamides-colchicines m-, p-AzoCols. Unfortunately, reactions with ortho-3a-d resulted in a complicated, inseparable products mixture. Thus, we decided to react 4 with 2-aminobenzoyl chloride in the presence of triethylamine and, in the following step, with nitrosobenzene 2a-d. This synthetic route resulted in the desired target compounds ortho-AzoCols (Scheme 2).

Photochemical Characterisation
Photochemical properties of AzoCols photoswitches are attributed to azobenzene 3a-d moiety incorporated into the parent pharmacophore structure; therefore, we evaluated the photochemical properties of obtained azobenzenes by NMR spectroscopy and UV-Vis spectrophotometry. As a solvent, we chose DMSO due to its ability to dissolve polar and nonpolar molecules, which is crucial for the analytical methods used in this study. Additionally, its intermediate polarity allows for good approximation of organic and aqueous solvents. Furthermore, in the context of photopharmacology, DMSO is used for stock solution preparation which is illuminated and then diluted into aqueous systems for biological activity assessment [63][64][65]. The photoisomerisation of azobenzenes 3a-d is not altered upon condensation with colchicine (see Supporting Information Figures S2  and S3). We assumed that ultraviolet light would cause photocatalyzed degradation of azobenzamides-colchicines AzoCols. Indeed, irradiation of p-AzoCol4F with UV light (365 nm) lead to the formation of a complex mixture of products (see Supporting Information Figure S4). This result confirms that ultraviolet light is incompatible with photoswitchable ligands based on the colchicine structure. Therefore, we determined the distribution of (E) and (Z) isomers for azobenzenes o-, m-, p-3a-d at the photostationary state (PSS) under constants illumination with selected wavelengths of the visible spectrum (390-610 nm) by 1 Figures S6-S17). The obtained results are summarised in Figure 3a

Photochemical Characterisation
Photochemical properties of AzoCols photoswitches are attributed to azobenzene 3a-d moiety incorporated into the parent pharmacophore structure; therefore, we evaluated the photochemical properties of obtained azobenzenes by NMR spectroscopy and UV-Vis spectrophotometry. As a solvent, we chose DMSO due to its ability to dissolve polar and nonpolar molecules, which is crucial for the analytical methods used in this study. Additionally, its intermediate polarity allows for good approximation of organic and aqueous solvents. Furthermore, in the context of photopharmacology, DMSO is used for stock solution preparation which is illuminated and then diluted into aqueous systems for biological activity assessment [63][64][65]. The photoisomerisation of azobenzenes 3a-d is not altered upon condensation with colchicine (see Supporting Information Figures S2 and S3). We assumed that ultraviolet light would cause photocatalyzed degradation of azobenzamidescolchicines AzoCols. Indeed, irradiation of p-AzoCol4F with UV light (365 nm) lead to the formation of a complex mixture of products (see Supporting Information Figure S4). This result confirms that ultraviolet light is incompatible with photoswitchable ligands based on the colchicine structure. Therefore, we determined the distribution of (E) and (Z) isomers for azobenzenes o-, m-, p-3a-d at the photostationary state (PSS) under constants illumination with selected wavelengths of the visible spectrum (390-610 nm) by 1 H or 19 F NMR analysis (see Supporting Information Figures S6-S17). The obtained results are summarised in Figure 3a After selection of optimal wavelengths, we acquired UV-Vis absorption spectra for azobenzenes o-, m-, p-3a-d in dimethyl sulfoxide at 500 µM concentrations for visualisation and analysis of weak n-π* bands. Spectroscopic data are summarised in Table 1. In general, the strong π-π* transition band was observed at 315-330 nm and the weaker n-π* band at approximately 445 nm for (E) isomers. After illumination with green light (causing photoconversion to the (Z) isomer), a decrease in π-π* band intensity and an increase in n-π* bands were observed. Most importantly, for compounds o-, m-, p-3d, due to the introduction of an ortho-fluoro substituent, causing stabilisation of nonbonding electron pairs of the azo-bond, significant separation of the n → π* transition band of the (E) and (Z) isomers (around 30 nm) was observed (Figure 3b), allowing for selective addressing of both isomers with visible light. Multiple cycles of photoreversible switching under alternating green and blue light irradiation without noticeable photobleaching or degradation confirmed repeatable and robust photochromic conversion of the obtained azobenzenes (Figure 3c). Moreover, we checked photochemical stability of o-AzoCol26DF and corresponding azobenzene o-3d in cell growth media at conditions similar to photopharmacological assays (high glucose medium supplemented with 4 mM L-glutamine, 500 ms pulses of green or blue light every 15 s for 48 h, 10% of DMSO). No changes in spectrum indicating degradation were observed. The only change in spectrum was attributed to (E) → (Z) photoconversions (see Supporting Information Figure S5). All photoswitches displayed substantially slower spontaneous (Z) → (E) relaxation (the half-life at 37 • C varied from 5 h to >48 h) than the biological assays timescale. Stability in organic and aqueous media, near-ideal photochemical characteristics, and bidirectional photoswitching showed that the obtained molecular switches can be used in photopharmacological assays or in vivo. → (Z) photoconversions, affording the highest PSS ratios. On the other hand, blue light (390-430 nm) induced reverse (Z) → (E) photoconversion, affording low PSS compositions. The lowest PSS percentages were obtained for compounds without a fluorine substituent, i. e., o-, m-, p-3a (46, 35, and 35%, respectively), while introducing a fluorine atom at 4-and 2,4-positions in compounds o-, m-, p-3b-c caused only a slight increase in PSS ratios. The azobenzenes bearing an ortho-fluorine substituent, i.e., o-, m-, p-3d, displayed the highest PSS compositions (88, 77, and 71%, respectively). After selection of optimal wavelengths, we acquired UV-Vis absorption spectra for azobenzenes o-, m-, p-3a-d in dimethyl sulfoxide at 500 μM concentrations for visualisation and analysis of weak n-π* bands. Spectroscopic data are summarised in Table 1. In general, the strong π-π* transition band was observed at 315-330 nm and the weaker n-π* band at approximately 445 nm for (E) isomers. After illumination with green light (causing photoconversion to the (Z) isomer), a decrease in π-π* band intensity and an increase in n-π* bands were observed. Most importantly, for compounds o-, m-, p-3d, due to the introduction of an ortho-fluoro substituent, causing stabilisation of nonbonding electron pairs of the azo-bond, significant separation of the n → π* transition band of the (E) and (Z) isomers (around 30 nm) was observed (Figure 3b), allowing for selective addressing of both isomers with visible light. Multiple cycles of photoreversible switching under alternating green and blue light irradiation without noticeable photobleaching or degradation confirmed repeatable and robust photochromic conversion of the obtained azobenzenes (Figure 3c). Moreover, we checked photochemical stability of o-AzoCol26DF and corresponding azobenzene o-3d in cell growth media at conditions similar to photopharmacological assays (high glucose medium supplemented with 4 mM L-glutamine, 500 ms pulses of green or blue light every 15 s for 48 h, 10% of DMSO). No changes in spectrum indicating degradation were observed. The only change in spectrum was attributed to (E) → (Z) photoconversions (see Supporting Information Figure S5). All photoswitches displayed substantially slower spontaneous (Z) → (E) relaxation (the half-life at 37 °C varied from 5 h to >48 h) than the biological assays timescale. Stability in organic and aqueous media, near-ideal photochemical characteristics, and bidirectional photoswitching showed that the obtained molecular switches can be used in photopharmacological assays or in vivo.

Photocontrollable in Cellulo Studies
Colchicine is a potent microtubule polymerisation inhibitor leading to mitotic arrests and, as a consequence, cell death [66]. We expected photoswitchable azobenzamidescolchicnes AzoCols to show similar antiproliferative activity. Therefore, we decided to evaluate in cellulo cytotoxicity for twelve obtained colchicine analogues at the most favourable illuminating conditions (430 nm for predominantly (E) and 535 nm for predominantly (Z) isomer, respectively). For irradiation of cell cultures, we used self-built arrays of 24 low-power light-emitting diodes (LED) controlled by the Adurino board (see Supporting Information). Such an automated system allowed for precise pulsed illumination (500 ms pulses of light every 15 s in our experiments) during long-term assays. Most importantly, it has been proven to be compatible with cell culturing conditions [37,40,42,45,46]. We first screened for bioactivity on an MCF-7 cell line as a model for human breast adenocarcinoma (N = 1). For lead compound exhibiting light-dependent cytotoxicity, we further expanded our research to include a HCT116 cell line (human colorectal carcinoma) and a HKE293 cell line (human embryonic kidney). The obtained results are summarised in Table 2. Notably, all AzoCols showed potent antiproliferative activity (IC 50 ranging from 28 to 187 nM), clearly demonstrating that incorporation of azobenzene moiety neither suppress binding to tubulin nor disrupt permeation through the cytoplasmic membrane. The meta-AzoCols set of compounds displayed equipotent bioactivity under illumination with green or blue light. In contrast, throughout paraand orthoisomers, p-AzoCol24DF and o-AzoCol26DF showed dissimilar cytotoxicity, dependent on the irradiation conditions. However, for o-AzoCol and o-AzoCol4F, we obtained somewhat inconsistent results. In the predominantly (Z) isomer state (green light illumination), p-AzoCol24DF and o-AzoCol26DF showed higher potency than in the predominantly (E) isomer state (blue light illumination). The most promising compound, o-AzoCol26DF, displayed c.a. a double potency shift on HCT116 cells upon illumination with green light.   -a without illumination ref. [67].; b without illumination ref. [68].; nc-not calculated.

o-AzoCol26DF Disrupt Tubulin Polymerisation and Cellular Microtubule Organisation in Light-Dependent Manner
To further explore the molecular mechanism of o-AzoCols light-dependent cellular activity, we examined the influence of o-AzoCol26DF on tubulin polymerisation in cell-free assays with purified tubulin. In the mainly (Z) isomer state (green light), 10 µM solution of o-AzoCol26DF resulted in ca. 40% inhibition of polymerisation over control (referenced to DMSO as 0%), while the thermally adapted state yielded only a 30% polymerisation inhibition. In comparison, 5 µM solution of colchicine caused ca. 70% of tubulin polymerisation inhibition (Figure 4b). It is worth noticing that this is a highly nonlinear assay in an environment far from cellular conditions. Thus, these results cannot be used for evaluating potencies, but rather to shed light on the mechanism of action as microtubule destabilisers, as the parent colchicine is. To verify the effect of o-AzoCol26DF on the microtubular cytoskeleton in vivo, we incubated HTC116 cells for 24 h with either DMSO (control) or o-AzoCol26DF. During the incubation, all cell samples were illuminated with green or blue light as described in the Material and Methods section. As an additional positive control, we used cells treated with 20 nm colchicine without light illumination. In both control samples, non-dividing cells had a dense, well-expanded network of interphase microtubules, while dividing cells formed a bipolar mitotic spindle (Figure 5a,b). A similar cell phenotype was observed in o-AzoCol26DF-treated, blue-light-illuminated cells (Figure 5c). In cells treated with 100 nM o-AzoCol26DF and activated by the green light, the interphase microtubular cytoskeleton was only slightly less prominent with respect to the control. However, we observed more dividing cells, and the mitotic spindle structure was frequently abnormal, with more than two spindle poles or misaligned microtubules (Figure 5d). This phenotype was even more pronounced in 120 nM o-AzoCol26DF-treated and green-light-illuminated cells (Figure 5f). Importantly, similar changes were observed in cells treated with 20 nM colchicine (Figure 5g).  To verify these observations, we determined the distribution of cell cycle stages using FACS analysis ( Figure 6). Under the control treatment, within the HTC116 samples, ~31-34% cells were in G1 phase, 22-26% in S phase, and 25-26% in G2/M phases. We also observed some polyploid and apoptotic cells ( Figure 6). In 100 nM o-AzoCol26DF-treated and blue-light-illuminated cells, the distribution of cell cycle stages was similar to that in control cells; however, we observed a higher number of apoptotic cells. In contrast, in green-light-illuminated 100 nM o-AzoCol26DF-treated cells, the number of G2/M cells increased to ~40%, while the number of apoptotic cells was the same as that in blue-lightilluminated cells. Treatment of cells with 20 nM colchicine raised the number of G2/M cells to ~45%, indicating that, likely, 100 nM o-AzoCol26DF has similar activity to 20 nM colchicine. Importantly, the increase in the concentration of o-AzoCol26DF to 120 nM elevated the number of G2/M cells to ~60% in green-light-illuminated cells, while in the corresponding blue-light-treated cells G2/M cell numbers remained at 29%.
To summarise, under used conditions, we were able to control the activity of o-Azo-Col26DF and, consequently, disrupt microtubular cytoskeleton leading to mitotic arrest  To verify these observations, we determined the distribution of cell cycle stages using FACS analysis ( Figure 6). Under the control treatment, within the HTC116 samples, 31-34% cells were in G1 phase, 22-26% in S phase, and 25-26% in G2/M phases. We also observed some polyploid and apoptotic cells ( Figure 6). In 100 nM o-AzoCol26DF-treated and blue-light-illuminated cells, the distribution of cell cycle stages was similar to that in control cells; however, we observed a higher number of apoptotic cells. In contrast, in green-light-illuminated 100 nM o-AzoCol26DF-treated cells, the number of G2/M cells increased to~40%, while the number of apoptotic cells was the same as that in blue-lightilluminated cells. Treatment of cells with 20 nM colchicine raised the number of G2/M cells to~45%, indicating that, likely, 100 nM o-AzoCol26DF has similar activity to 20 nM colchicine. Importantly, the increase in the concentration of o-AzoCol26DF to 120 nM elevated the number of G2/M cells to~60% in green-light-illuminated cells, while in the corresponding blue-light-treated cells G2/M cell numbers remained at 29%.

Conclusions
In this study, we successfully designed and synthesised a set of novel photoswitchable colchicine-based microtubule dynamics disrupting agents. The developed photoswitches can be photoisomerised by visible light instead of UV light, which is used in classical photopharmaceutical reagents. This is crucial due to colchicine instability under UV light irradiation. For the lead compound o-AzoCol26DF, we demonstrated light- Figure 6. Analysis of cell cycle stages of HTC116 cells upon treatment with DMSO, o-AzoCol26DF, or colchicine and blue or green light exposure. Graphs show representative flow cytometry analysis for each sample. In the table, the mean of at least two experiments for each sample are included. The increase with respect to the control is marked by the red color and upwards arrow, while the decrease is marked by the blue color and downwards arrow.
To summarise, under used conditions, we were able to control the activity of o-AzoCol26DF and, consequently, disrupt microtubular cytoskeleton leading to mitotic arrest and, eventually, cell death.

Conclusions
In this study, we successfully designed and synthesised a set of novel photoswitchable colchicine-based microtubule dynamics disrupting agents. The developed photoswitches can be photoisomerised by visible light instead of UV light, which is used in classical photopharmaceutical reagents. This is crucial due to colchicine instability under UV light irradiation. For the lead compound o-AzoCol26DF, we demonstrated light-dependent cytotoxicity in both the HCT116 and the MCF-7 cancerous cell lines. Inhibition of purified tubulin polymerization, as well as disruption of microtubule organization, support proofof-concept of using AzoCols as photoswitchable microtubule dynamics disrupting agents. In summary, we have proven that AzoCols provide the basis for further improvement and development of the novel class of photoswitchable colchicine-based microtubule polymerisation inhibitors that could be used in future studies and applications.