2,2′‐Bipyridine‐Modified Tamoxifen: A Versatile Vector for Molybdacarboranes

Abstract Investigations on the antitumor activity of metallacarboranes are sparse in the literature and limited to a handful of ruthena‐ and molybdacarboranes. In this study, the molybdacarborane fragment [3‐(CO)2‐closo‐3,1,2‐MoC2B9H11] was combined with a vector molecule, inspired by the well‐known drug tamoxifen or 4,4′‐dihydroxytamoxifen (TAM‐diOH). The molybdacarborane derivative [3,3‐{4‐[1,1‐bis(4‐hydroxyphenyl)but‐1‐en‐2‐yl]‐2,2′‐bipyridine‐κ2 N,N′}‐3‐(CO)2‐closo‐3,1,2‐MoC2B9H11] (10), as well as the ligand itself 4‐[1,1‐bis(4‐hydroxyphenyl)but‐1‐en‐2‐yl]‐2,2′‐bipyridine (6) showed cytotoxic activities in the low micromolar range against breast adenocarcinoma (MDA‐MB‐231, MDA‐MB‐361 and MCF‐7), human glioblastoma (LN‐229) and human glioma (U‐251) cell lines. In addition, compounds 6 and 10 were found to induce senescence and cytodestructive autophagy, lower ROS/RNS levels, but only the molybdacarborane 10 induced a strong increase of nitric oxide (NO) concentration in the MCF‐7 cells.


Instrumentation
NMR spectra were acquired at room temperature with a Bruker AVANCE III HD 400 spectrometer. 1 H (400.13 MHz) and 13 C (100. 16 MHz) NMR spectra were referenced to tetramethylsilane (TMS) as internal standard. 11 B (128.38 MHz) NMR spectra were referenced to the unified Ξ scale. [6] Mass spectrometry measurements were carried out in the ESI-MS mode using a Bruker ESQUIRE 3000 (Benchtop LC Iontrap) spectrometer. IR spectra were obtained with a PerkinElmer system 2000 FTIR spectrometer, scanning between 400 and 4000 cm −1 using KBr pellets, which were prepared in a glovebox under nitrogen atmosphere. Elemental analyses were performed with a Hereaus VARIO EL oven. The single crystal X-Ray data were collected on a Gemini-CCD diffractometer (RIGACU INC.) using Mo-Kα radiation (λ = 0.71073 Å), ω-scan rotation. Data reduction was performed with CrysAlis Pro [7] including the program SCALE3 ABSPACK [8] for empirical absorption correction. The structure solution for 6 and 9b was performed with SHELXT (dual-space method). [9] The graphical user interface ShelXle [10] was used for SHELXL. [11] The anisotropic full-matrix least-squares refinement on F 2 of all non-hydrogen atoms was performed with SHELXL-97. Except for disordered solvent molecules, all non-hydrogen atoms were refined with anisotropic thermal parameters and the HFIX command was used to locate all hydrogen atoms for non-disordered regions of the structure. The C2 unit within the carborane cluster was located with bond lengths analysis. Structure figures were generated with Mercury (version 4.0.0) [12] or UCSF Chimera (version 1.14). [13] CCDC-1944134 (6) and CCDC-1944251 (9b) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via https://summary.ccdc.cam.ac.uk/structure-summary-form (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44)1223-336-033; or deposit@ccdc.cam.uk). UV-Vis absorption spectra were measured with a PerkinElmer UV/VIS/NIR Lambda 900 spectrometer, equipped with tungsten-halogen and deuterium lamps, using quartz cuvettes (V = 3 cm 3 , l = 10 mm). Spectra were recorded in the range 240-800 nm, at 1.0 nm resolution. Steady-state fluorescence emission and Rayleigh Light Scattering (RLS) spectra were measured with a PerkinElmer LS-50b spectrometer, equipped with a xenon-arc lamp, using quartz cuvettes (V = 1 cm 3 , l = 5 mm). Nanoparticle Tracking Analysis data were recorded using a NanoSight LM10 (Malvern Instruments Ltd, Worcestershire, UK), containing a sample chamber of about 0.25 mL, and equipped with a 532 nm-laser, a microscope LM14B and a camera sCMOS. The NTA 3.0 analytical software (NanoSight Ltd) was used for both capture and processing. Acquisition and processing parameters were optimized for each sample and the respective blank.

Rct.
Catalyst   (6) In a Schlenk flask 5 (0.180 g, 0.456 mmol, 1.0 equiv.) was dissolved in CH2Cl2 (10 mL) and cooled to -65 °C. BBr3 (0.26 mL, 2.74 mmol, 6.0 equiv.) was added slowly to the mixture via syringe and the mixture was slowly warmed up to ambient temperature and then stirred for 12 h (the color of the reaction mixture changed from orange/red to light orange and the formation of a precipitate was observed). The reaction was quenched with H2O (20 mL), the two phases were separated, and the aqueous phase was extracted with EtOAc (3 x 20 mL). All combined organic phases were dried over MgSO4, filtered and the solvent was removed under reduced pressure. Flash column chromatography on silica gel (22 cm x 3 cm) using a CHCl3/MeOH/NH3(aq.) 8:1:0.1 (v/v) eluent system yielded a paleyellow sticky solid, which was recrystallized from hot CHCl3 with a few drops of MeOH to give pale yellow crystals of 6 (0.143 g, 80%).

[3,3-{4-[1,1-Bis(4-hydroxyphenyl)but-1-en-2-yl]-2,2'-bipyridine-κ 2 N,N'}-3-(CO)2-closo-3,1,2-MoC2B9H11] (10)
The same procedure was applied as for 8, but using  CCDC-1944134 (6) and CCDC-1944251 (9b) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk. Figure S1. Packing of 6 in the solid state stabilized by intermolecular hydrogen bonds. The thermal ellipsoids are presented at 30% probability level and the hydrogen atoms not involved in hydrogen bonds are omitted for clarity. One molecule of 6 is highlighted. Figure S2. Molecular structures of the two isomers of 9b in the solid state (not the packing in the crystal is shown here). Thermal ellipsoids are presented at 50% probability level and hydrogen atoms are omitted for clarity. CCDC number 1944251.

NMR Spectroscopy
In the 1 H NMR spectra of the ligands alone (3, 5 and 6) compared to the chemical shifts in complexes 8, 9 and 10, there is a clear coordination shift detectable (Figures S3−S5). The CHcluster signals split into two broad singlets due to the non-symmetric N,N-chelating ligand. Importantly, the coordination of 3, 5 and 6 at molybdenum(II) generates always two isomers, which cannot be distinguished via NMR spectroscopy. The carbonyl groups give two slightly different signals at 255.3 and 254.5 ppm (8, CD2Cl2), 256.3 and 255.5 ppm (9, CDCl3) or 260.3 and 260.0 ppm (10, CD3CN) in the 13 C{ 1 H} NMR spectra of the respective complexes according to the C1 symmetry of the whole molecule.
For in vitro cell culture tests, the stock solutions of sparingly water-soluble compounds are typically prepared in DMSO (ethanol or methanol are good alternatives) and stored below 4 °C. For that purpose, the chemical stability of 3, 5, 6 and 8−10 was tested in a solution of water-containing DMSO-d6 in air for at least 36 days (3, 5, 6, 8, 9) and for 14 days for complex 10. In all cases, 1 H and 11 B{ 1 H} NMR spectra, where appropriate, revealed that the ligands 3, 5 and 6 can be stored in a DMSO stock solution for at least a month without decomposition, and the molybdacarboranes for 14 days up to a month with minor decomposition (where the decomposition products are the free ligands 3, 5 or 6, the nido-ortho-carborane ([nido-C2B9H12] − ) and most likely a molybdenum species in higher oxidation state) (see Figures S6−S14). Figure S3. 1 H NMR spectra (DMSO-d6) of free ligand (3, top) and the respective molybdacarborane complex (8, bottom). Clear coordination shift of the aromatic protons of the 2,2'-bipyridine unit (aromatic region) and a small shift of the ethyl group (aliphatic region). Additionally, for 8 the BH signals (0.66-3.71 ppm) and the CH protons of the carborane cluster split into two broad singlets (2.87 or 2.94 ppm) due to the non-symmetric ligand 3. * Indicates the water residual peaks, # indicates the DMSO residual peak. Figure S5. 1 H NMR spectra (DMSO-d6) of free ligand (6, top) and the respective molybdacarborane complex (10, bottom). Clear coordination shift of the aromatic protons of the 2,2'-bipyridine unit (aromatic region) and a small shift of the 1,4substituted phenyl rings and small shift of the ethyl group (aliphatic region). Also, in DMSO-d6 the OH groups are detectable and shift upon coordination. Additionally, for 10 the BH signals (0.66-3.61 ppm) and the CH protons of the carborane cluster split into two broad singlets (2.76 or 2.91 ppm) due to the non-symmetric ligand 6. * Indicates the water residual peaks, # indicates the DMSO residual peak.        S12. Stability test for 9 in DMSO-d6 stock solution via 11 B{ 1 H} NMR spectroscopy, over 37 days. Minor changes could be detected. The slowly growing broad peak at ca. 20 ppm is due to agglomeration and the changes observed are not related to decomposition, but rather to reorganization of the molecules in solution.  S14. Stability test for 10 in DMSO-d6 stock solution via 11 B{ 1 H} NMR spectroscopy, over 14 days. Minor changes could be detected. The slowly growing broad peak at ca. 20 ppm is due to agglomeration and the changes observed are not related to decomposition, but rather to reorganization of the molecules in solution.

UV-Vis, Fluorescence and Resonance Light Scattering (RLS) Spectroscopy
All working solutions were kept at a controlled temperature of 25 °C for the whole duration of the experiments. UV-Vis spectra were recorded at 25 °C, in the range of 190-800 nm, at 1.0 nm resolution. All measurements were corrected by subtracting the respective blank: PBS + 2.0 vol% DMSO for measurements in buffer. Fluorescence emission spectra were acquired in the range 290-550 nm, with cut-off (emission) filter at 290 nm. Three excitation wavelengths (λexc) were used, λ280 (Trp and Tyr), λ295 (Trp) and λ320 (reference λexc for [BSAnoMg-warfarin] complex), with excitation and emission slits of 5.0 and 2.5 (λ280), 2.5 and 2.5 (λ295), 5.0 and 2.5 (λ320) nm, respectively. Each measurement was corrected subtracting the blank (PBS + 2 vol% DMSO) at the respective λexc. Raw data (number of scans = 3) were smoothed with 10 points moving average function, integrated in the software FLWINLab. The two-and three-components systems, i.e. [BSAnoMg-6/10], [BSAnoMgsite marker], [BSAnoMg-site marker-6/10] and [BSAnoMg-6/10-site marker], as well as the respective reference solutions were measured over 24 h with UV-Vis and fluorescence spectroscopy (spectra see Figure S15−S16). RLS spectra were measured after 1 h from sample preparation and were acquired with Δλ = λem -λexc = 0 nm, as reported, [17] closed slits, and 1% transmittance attenuator filter. Each measurement was corrected subtracting the blank (PBS + 2 vol% DMSO). Raw data (ns = 3) were smoothed with 10 points moving average function, integrated in the software FLWINLab.    1). In contrast, for 10, no improvement of the solubility could be detected as formulated with BSAnoMg (1:1), the RLS intensity even increases. The blanks consist of PBS + 2% DMSO.

Nanoparticle Tracking Analysis (NTA)
The possible self-assembling behavior of 6 and 10 in PBS/DMSO mixture (pH 7.4) was analyzed by Nanoparticle Tracking Analysis (NTA), with and without BSAnoMg, applying an analogous procedure as reported previously. [18] Samples of 6 and 10 in PBS/DMSO were prepared as described for UV-Vis, fluorescence and RLS measurements and measured 0.  Antiestrogen receptor alpha antibody (ab3575), antiestrogen receptor beta antibody (ab3576) and pre-analyzed with CyFlow® Space Partec using the PartecFloMax® software. For the measurement of the intracellular nitric oxide (NO) quantity, cells were treated with an IC50 dose of the experimental compounds for 72 h, washed, trypsinized and stained with 5 μM 4-amino-5-methylamino-2',7'difluorofluorescein diacetate (DAF-FM diacetate) for 1 h at 37 °C in phenol red-free RPMI 1640. Thereafter, cells were washed and additionally incubated for 15 min in fresh RPMI 1640 without phenol red and serum, to finish the reaction of de-esterification. Analysis was done as indicated for DHR 123. For the determination of the β-galactosidase activity, cells were treated with an IC50 dose of 6, 10 or Fc-diOH for 72 h, and then stained with β-galactosidase substrate FDG (fluorescein-di--Dgalactopyranoside) to a final concentration of 1 mM. After 1 min incubation at 37 °C, cells were analyzed by flow cytometry as described. Channels FL1 (green emission), FL2 (orange emission) and/or FL3 (dark red emission) were used for fluorescence detection, according to the specific staining agent. Experiments were run in three independent replicates.

Biological Studies
Fluorescence microscopy. For DAPI staining, cells were cultivated on chamber slides overnight, then treated with 6, 10 or Fc-diOH. Cells were then fixed with 4% (w/w) paraformaldehyde (15 min, room temperature), and the chamber slides were covered with DAPI Fluoromount-G (Southern Biotech, Birmingham, AL, USA) before analysis. The slides were analyzed with a Zeiss AxioObserver Z1 inverted fluorescence microscope (Carl Zeiss AG, Oberkochen, Germany) at 200× magnification. Morphological signs of apoptosis (irregular nuclei shape, condensed chromatin, apoptotic bodies) were examined in three independent experiments.
Statistical analysis. Analysis of variance (ANOVA) followed with a Students−Newman−Keuls test was used for significance of the differences between treatments, and a p value less than 0.05 was taken as statistically significant.  In order to draw interferences between the activity of the novel molecules, which are inspired by tamoxifen, a drug for hormone-dependent breast cancer treatment, and the presence of the ER receptors, the receptor status for ER-α and ER-β was tested. This is reasonable, due to the fact that the two transcriptions factors (ER-α and ER-β) have a very important role in health and disease, and the real ER status of the tested cell lines should always be doublechecked to draw a connection to the tested compounds. [19] Therefore, bridging the gap between the tested cell lines and the estrogen receptors alpha (ER-α) and beta (ER-β), western blot analysis was performed ( Figure S21). It revealed that the tested cell lines express the ER-α -MDA-MB-361, MCF-7 and LN-229 in higher amounts, MDA-MB-231 in medium amounts and U-251 to a lesser extent. Concerning the expression of the ER-β, also all cell lines express this receptor, but to varying extent. MDA-MB-231 and U-251 show low expression of ERβ, whereas MDA-MB-361, MCF-7 and LN-229 express higher amounts of ER-β. Vivid discussion is found in the literature about the validity of the commercially available ER-α/ER-β antibodies: their selective binding is doubted, and in turn, the estrogen receptor status of the cells is found to be contradicting throughout the literature. [20] However, it seems that the anticancer potential of our newly designed compounds is not in correlation with ER receptor expression, under the applied concentrations, indicating the existence of other targets for these molecules.