In‐Cell Activation of Organo‐Osmium(II) Anticancer Complexes

Abstract The family of iodido OsII arene phenylazopyridine complexes [Os(η6‐p‐cym)(5‐R1‐pyridylazo‐4‐R2‐phenyl))I]+ (where p‐cym=para‐cymene) exhibit potent sub‐micromolar antiproliferative activity towards human cancer cells and are active in vivo. Their chemical behavior is distinct from that of cisplatin: they do not readily hydrolyze, nor bind to DNA bases. We report here a mechanism by which they are activated in cancer cells, involving release of the I− ligand in the presence of glutathione (GSH). The X‐ray crystal structures of two active complexes are reported, 1‐I (R1=OEt, R2=H) and 2‐I (R1=H, R2=NMe2). They were labelled with the radionuclide 131I (β−/γ emitter, t1/2 8.02 d), and their activity in MCF‐7 human breast cancer cells was studied. 1‐[131I] and 2‐[131I] exhibit good stability in both phosphate‐buffered saline and blood serum. In contrast, once taken up by MCF‐7 cells, the iodide ligand is rapidly pumped out. Intriguingly, GSH catalyzes their hydrolysis. The resulting hydroxido complexes can form thiolato and sulfenato adducts with GSH, and react with H2O2 generating hydroxyl radicals. These findings shed new light on the mechanism of action of these organo‐osmium complexes.


Synthesis of [Os(η 6 -p-cym)(5-EtO-AZPY)OH]PF6 (1-OH·PF6).
To a stirred solution of [Os(η 6 -p-cym)Cl2]2 (100 mg, 126.5 µmol) in methanol (3 mL), silver nitrate (58.87 g/L, 1459 µL) in water was added. The mixture turned yellow and a white precipitate formed, which was removed via filtration. A solution of 5-EtO-AZPY (60.4 mg, 265.6 µmol) in methanol (5 mL) was added drop-wise to the yellow solution and it turned brown. The mixture was stirred for 18 h at ambient temperature, then ammonium hexafluorophosphate (206.2 mg, 1.26 mmol) was added. The product was extracted with DCM (10 mL) and washed with water (2 x 10 mL). DCM was removed under reduced pressure and the product was re-dissolved in a minimum amount of methanol (~2 mL), and placed in a freezer overnight. A brown precipitate formed, which was collected via vacuum filtration then washed with ice-cold ethanol (2 x 1 mL), diethyl ether (2 x 5 mL), and dried overnight in a vacuum desiccator. Yield: 160.6 mg (89%).

NMe2)Cl]PF6 (2-Cl·PF6).
These were synthesized and characterised as described previously. [2] Crystal growth. Crystals of 1-I (as 1-I·PF6·0.5EtOH) suitable for x-ray crystallography were grown by dissolving 1-3 mg of solids in ethanol (1 mL) and cooling solutions in a freezer at ca. 253 K. Crystals of 2-I (as 2-I·PF6) were grown in a similar manner in methanol.
X-ray crystallography. Diffraction data were collected on an Oxford Diffraction Gemini four-circle system with a Ruby CCD area detector. All structures were refined by full-matrix least squares against F 2 using SHELXL 97 and were solved by direct methods using SHELXS(TREF) with additional light atoms found by Fourier methods.
Hydrogen atoms were added at calculated positions and refined using a riding model. EdenTerm V1.21 software. The cellular protein content was determined for digested S8 cells using a calibration curve (Fig. S6). The data were processed using MS excel and standard deviations were calculated.
In vitro growth inhibition assay. Approximately 5000 A2780 or MCF-7 cells were seeded per well in 96-well plates. The cells were pre-incubated in drug-free media at 310 K for 48 h before adding various concentrations of the osmium complex. Cells were exposed to complexes for 24 h at 310 K. The supernatants were removed by suction and each well was washed with PBS. The cells were allowed to recover for 72 h in a drug-free medium at 310 K. The SRB assay was used to determine cell viability. was used to solubilise the complexes to 1 mM. The EPR spectra were recorded immediately after sample preparation. EPR spectra were analysed and simulated using the EASYSPIN software. [3] pKa determination. shifts of the peaks of complex 1-OH were followed by 1 H NMR. The pH* values were corrected using the equation; pH = 0.929pH* + 0.41, [4] and the data were fitted to the Henderson-Hasselbalch equation using Origin 8.5. [a] The retention times listed here were determined on a different HPLC column from that used in the main script.