Platinum(iv)-azido monocarboxylato complexes are photocytotoxic under irradiation with visible light†

Complexes trans,trans,trans-[Pt(N3)2(OH)(OCOR)(py)2] where py = pyridine and where OCOR = succinate (1); 4-oxo-4-propoxybutanoate (2) and N-methylisatoate (3) have been synthesized by derivation of trans,trans,trans-[Pt(OH)2(N3)2(py)2] (4) and characterised by NMR and EPR spectroscopy, ESI-MS and X-ray crystallography. Irradiation of 1–3 with green (517 nm) light initiated photoreduction to Pt(ii) and release of the axial ligands at a 3-fold faster rate than for 4. TD-DFT calculations showed dissociative transitions at longer wavelengths for 1 compared to 4. Complexes 1 and 2 showed greater photocytotoxicity than 4 when irradiated with 420 nm light (A2780 cell line IC50 values: 2.7 and 3.7 μM) and complex 2 was particularly active towards the cisplatin-resistant cell line A2780cis (IC50 3.7 μM). Unlike 4, complexes 1–3 were phototoxic under green light irradiation (517 nm), with minimal toxicity in the dark. A pKa(H2O) of 5.13 for the free carboxylate group was determined for 1, corresponding to an overall negative charge during biological experiments, which crucially, did not appear to impede cellular accumulation and photocytotoxicity.

(complexes 2 and 3) during data collection. Using Olex2, 3 the structures were solved with the XS 4 structure solution program using Direct Methods and refined with the ShelXL 5 refinement package using Least Squares minimisation. The OH and NH hydrogen atoms were located in a difference map but refined with restraints. Images representing the intermolecular forces for each complex were created using Mercury 3.3. X-ray crystallographic data for complexes 1-3 have been deposited in the Cambridge Crystallographic Data Centre under the accession numbers 1998067, 1998068 and 1998069 respectively. X-ray crystallographic data in CIF format are available from the Cambridge Crystallographic Data Centre (http://www.ccdc.cam.ac.uk/).

UV-vis absorption spectra:
were recorded on a Varian Cary 300 UV-vis spectrophotometer in 1 cm path-length quartz glass cuvettes, purchased from Starna Scientific. The spectral width was 200 -800 nm, with bandwidth 1.0 nm and a scan rate of 600 nm.min -1 .

Fluorescence measurements:
The emission spectra were recorded on a Jasco FP-6500 using the following parameters: excitation wavelength at 320 nm, excitation and emission slits at 10 nm and 5 nm respectively. The response was set to medium and the scanning rate to 200 nm. min -1 .
pKa determination of complex 1: pH values were measured in NMR tubes with a solid-state electrode pocket ISFET pH meter (Minilab Series). The meter was calibrated using standard buffer solutions (pH 7.00, pH 4.01, pH 10 Exponential rate fitting: For determining the rate of loss of LMCT bands in UV-vis spectra, the maximum for each of the graphs was normalized so that the first spectrum (t=0 h, i.e. before the irradiation) had a normalized absorbance of 100%. Then by considering only the initial degradation (up to ca. 38%) the points were fitted to an exponential curve (y=y 0 +Ae Rox ) using Origin 8.5.

Electron Paramagnetic Resonance:
The EPR spectra were recorded on a Bruker EMX (Xband) spectrometer. A cylindrical T M10 mode cavity (Bruker 4103 TM) was used and samples were placed in spectrosil quartz tubes of inner diamater 1.0 mm and outer diameter of 1.2 mm (Wilmad Labglass) sealed with J-Blue Tac. The tube was placed inside a larger quartz tube so the sample could be more accurately positioned inside the resonator. Parameters: modulation amplitude 2.0 G, microwave power 0.63 mW, 1.0 x10 4 receiver gain, sweep gain at 41.94 s and a repeated number of 10 X scans was used and resolution in Y of 24. The green LED (517 nm) was placed at a distance of 8.5 cm from the tube in the EPR cavity. Irradiations lasted for 35 min; every 7 min a spectrum was recorded. The spin adduct concentration was determined using a calibration curve (SigmaPlot) obtained from standard solutions of 4-hydroxy-2,2,6,6tetramethyl-piperidine-1-oxyl (Tempol) in the corresponding solvent and the processing was carried out using Bruker WINEPR software.
Computational details: All calculations were performed with the Gaussian 03 (G03) program 7 employing the DFT method, the PBE1PBE 8 functionals. The LanL2DZ basis set 9 and effective core potential were used for the Pt atom and the 6-31G**+ basis set 10 was used for all other atoms. Geometry optimizations of complex 1 (deprotonated) in the ground state (S 0 ) and lowest-lying triplet state (T 1 ) were performed in the gas phase and the nature of all stationary points was confirmed by normal mode analysis. For the T 1 geometry the UKS method with the unrestricted PBE1PBE functional was employed. The conductor-like polarizable continuum model method (CPCM) 11 with water as solvent was used to calculate the electronic structure and the singlet excited states of complex 1 in solution (H 2 O). Thirty-two singlet excited states with the corresponding oscillator strengths were determined with a Time-dependent Density Functional Theory (TD-DFT) 12,13 calculation.
LC-MS studies: LC-MS analysis was carried with a Dionex 3000RS UHPLC coupled with Bruker MaXis Q-TOF mass spectrometer. An Agilent Zorbax Eclipse plus column (C18, 100x2.1 mm, 1.8 µm) was used. Mobile phases were A (water with 0.1% formic acid) and B (as ACN with 0.1% formic acid). A gradient of 5% B to 100% B in 15 min was employed with flow rate at 0.2 mL.min -1 , UV was set at 240 nm. The mass spectrometer was operated in electrospray positive mode with a scan range 50-2,000 m/z. Source conditions: end plate offset at -500 V; capillary at -4500 V; nebulizer gas (N 2 ) at 1. Phototoxicity testing: Cell culture and other chemicals were obtained from Sigma Aldrich Ltd (Poole, UK). Disposable sterile cell culture plastics were obtained from Greiner Bio-One (Cambridge, UK). All procedures were carried out in a specially adapted photobiology laboratory with ambient light levels measured below 1 lux (Solatell, UK).
OE19 human oesophageal carcinoma and the paired A2780/A2780cis human ovarian carcinoma cell lines were obtained from the European Collection of Cell Cultures (Porton Down, UK) and maintained in RPMI containing 10% (v/v) foetal calf serum. Cells were free of mycoplasma and were maintained in antibiotic-free conditions in a humidified atmosphere of 5% CO 2 /95% air. For experiments, cells were seeded at a density of 6 -7 x 10 4 cells.cm -2 in 96 well plates for broadband irradiation; or 2 x 10 6 cells.mL -1 in stirred culture for monochromatic irradiation. Complexes were prepared immediately before use in Earle's Balanced Salt Solution (EBSS) and filter-sterilised, with the exception of complex 3 which was dissolved in DMSO and then diluted into EBSS. Irradiations were performed in optically clear medium and experiments were controlled for light, complex, solvent (when required), and handling. Blue light was delivered by a bank of TL03 fluorescent tubes (λ max : 420 nm) with wavelengths shorter than 400 nm blocked by filtering. Irradiance was measured with a Gigahertz Optik meter calibrated to the source using a spectroradiometer (Bentham, UK; mean irradiance 1.3 mW.cm -2 ± 0.3). Irradiances were measured through filters, and where appropriate, cell culture plate lids. Monochromatic irradiation was performed on stirred cultures using a Bentham monochromator equipped with a 450 W Xenon arc lamp and light guide. Irradiance was measured using a calibrated integrating sphere and all measurements were traceable to the National Physics Laboratory (mean irradiance at 517 ± 27 nm: 97.5 mW.cm -2 ± 3.8; at 570 ± 27 nm: 49.0 mW.cm -2 ± 1.7). Sham-irradiated cells were treated identically and in parallel with irradiated cells, except that photons were blocked.
Phototoxicity was determined by neutral red dye uptake either 24 hours after irradiation (broadband irradiation) or alternatively cells were seeded at a low density and toxicity measured 72 hours post-irradiation (monochromatic irradiation). Absorbance was read at 540 nm in a Synergy™ 2 plate reader. The concentration of complex required to inhibit dye uptake by 50% (IC 50 value) was calculated from the log-transformed cytotoxicity curves normalised to untreated cells (Graphpad Prism v.6). Goodness of fit was determined by the 95% confidence interval of the IC 50 value, and the R 2 value. The phototoxic index (PI) indicates the fold difference in cytotoxicity between cells treated with complex and either sham-irradiated or irradiated. Cells were also seeded at a low density into 35 cm dishes and stained with crystal violet (0.5%) 7-10 days post-irradiation. Experiments were performed in triplicate, and independently repeated at least once (6 observations from n=2) on cells of differing passage number. The comet assay was performed as previously described. 14 In order to reveal DNA adducts that inhibit DNA migration, the cells are briefly incubated (5 mins) with a low dose of hydrogen peroxide (25 µM) after the cells have been photoactivated and are about to be incorporated into the agarose gel and lysed in a DMSO-containing solution. To show DNA lesions that increase DNA migration (i.e. strand breaks), no peroxide is added. After electrophoresis, the nuclei were stained with ethidium bromide and scored by eye using a Nikon E600 eclipse fluorescent microscope.
Cell uptake studies: A2780 cells were seeded at a density of 5x10 5 cells.mL -1 in each well and they were allowed to adhere for 36 h. Before the complexes were added the cells were washed with PBS (1 ml). The stock concentration of the complex was confirmed via ICP-MS analysis before the complex exposure to the cells. The complexes (20 μM, 2 mL in PBS) were added and allowed to incubate for 1 h, in the dark at 37 o C. Following the uptake of the complex, cells were washed with PBS (1 mL) and a solution of Trypsin/ EDTA (2%) was added (0.5 mL) and incubated at 37 o C for 3 min, and then neutralized with cell culture media-RPMI (2 mL). Cells were transferred into falcon tubes and broken into a single cell suspension to enable counting, additional RPMI (2 mL) was added to dilute the cells. Cells were counted in duplicate and an average was taken. They were then spun down (1000 rpm, 22 C o , 5 min). The supernatant was removed and the cells were resuspended in PBS (1 mL) for a further wash. The cell pellet was kept at -20 o C until the Pt content was analysed via ICP-MS.

ICP-MS:
Digestion of the cell pellet was carried out overnight in concentrated nitric acid (73%) at 70 o C. The resulting solutions were diluted to 5% v/v HNO 3 , using doubly ionised water and then filtered via an Inorganic Membrane filter to remove any insoluble cell debris.

HPLC purity test:
The purity of the compounds was performed on Agilent 1200 system with a VWD and 100 μL injection loop. The column used was an Agilent ZORBAX Eclipse Plus C18, with dimensions of 250 x 4.6 mm and a 5 μm port size. The injected volume was 45 μL. The mobile phase used was H 2 O 0.1 % TFA/ ACN 0.1% TFA. The flow rate was set to 1mL.min -1 and the following method (Table S1) was used. Complexes 1 and 2 were dissolved in water and complex 3 in 10% MeOH at a concentration of ca 50 µM.

Synthesis and characterisation of synthetic precursors
Caution! No problems were encountered during this work, however heavy metal azides are known to be shock-sensitive detonators, therefore it is essential that platinum azides are handled with care. The Pt-diazido complexes were synthesised and handled under dim lighting conditions.

N-Methylisatoic acid (N-MIA) (5)
The open form of the N-methylisatoic anhydride was prepared by hydrolysis and decarboxylation of N-methylisatoic anhydride. 15 N-MI anhydride (1.2 g, 6.77 mmol) was dissolved in a solution of 2 M KOH (15 mL) and allowed to react for 4 h at 100 o C. The transparent solution was then allowed to cool down at room temperature, before the pH was adjusted to 6 -7 by the addition of HCl (3 M

Trans-[Pt(N 3 ) 2 (py) 2 ] (7)
This complex was prepared as previously reported. 16 Trans-[PtCl 2 (py) 2 ] (1) (3.75 g, 8.85 mmol) was suspended in water (500 mL), AgNO 3 (3.008 g, 2 mol eq) was added and the reaction was stirred at 60 o C overnight. The grey precipitate was filtered through celite over a frit to obtain a pale yellow solution which was then further passed through an IM filter. Then NaN 3 (5.75 g, 10 mol eq) was added and the reaction was stirred overnight at room temperature. The yellow precipitate was filtered and washed with ice-cold water, ethanol and diethyl ether (3.48 g, 90%). The solid was recrystallised from pre-warmed pyridine at 40 o C at a ratio of 1 g/37 mL of pyridine and following hot filtration it was allowed to crystallise at -20 o C. The yellow solid was filtered, washed and dried under vacuum (Yield = 2.44 g, 70%).

Trans, trans, trans-[Pt(N 3 ) 2 (OH) 2 (py) 2 ] (4)
The title product was obtained by oxidation of trans-[Pt(py) 2 (N 3 ) 2 ]. 16 Trans-[Pt(N 3 ) 2 (py) 2 ] (2.00 g, 4.56 mmol) was suspended in H 2 O 2 (160 mL, 30%) and allowed to react at 45 o C for 4 h, producing a yellow solution. This was filtered (IM) and transferred to a larger flask to which H 2 O (340 mL) was added. The solution was lyophilized, yielding a yellow solid which was recrystallised from 2:1 ethanol/methanol and the product precipitated by addition of diethyl ether at -20 o C. The solid was filtered and dried under vacuum (Yield = 1.6 g, 80%).   The intermolecular attractions which dictate the packing of complex 1 are shown in Figure S2 and consist of classical hydrogen bonds as well as some weak - and C-H··· interactions. Each molecule in the unit cell is able to participate in three intermolecular hydrogen bonds: two are established with an adjacent molecule through the interaction of the hydroxido ligand and the azide nitrogen atom (O2-H2···N3, O2-H2···N6 2.12 Å), while the third hydrogen bond interaction involves the OH group of the succinate ligand and the oxygen atom from the hydroxido belonging to a third molecule of complex (O-H5···O2, 1.706 Å). Based on the distance, both hydrogen bonds can be considered strong, but the interaction with the succinate is considerably stronger. Additionally, the pyridine rings of complex 1 interact via two different intermolecular forces: a C-H··· interaction between the ortho hydrogen of a pyridine ring and the -cloud of another pyridine ring from a neighbouring molecule (Ct1···H15, 3.025 Å), and also a very weak - stacking interaction between two pyridines with a centroid-centroid distance of 3.950 Å (Ct2-Ct3) ( Figure S2). In complex 2 the azido ligands are not involved in any hydrogen bonding. Each molecule establishes two intermolecular hydrogen bonds between the carbonyl group of the 4-oxo-4propoxybutanoate ligand and the hydroxido group of a nearby molecule (O2-H2···O2 1.972 Å) as shown in Figure S3. Moreover, complex 2 shows a - stacking interaction between pyridine rings (3.551 Å) in the crystal packing, which is stronger in comparison with that found for complex 1. In complex 3, each molecule in the crystallographic unit cell is hydrogen-bonded via an azido nitrogen atom (N5) and the hydroxido group (O2-H2) to another neighbouring molecule ( Figure S4). The strength of this bond can be considered as medium-to-weak as it is 2.288 Å in length. The azido ligand in complex 3 which participates in an intermolecular hydrogen bond shows a weaker bond with the Pt(IV) ion than the second azido ligand does. Complex 3 is the only complex of 1, 2 and 3 for which a strong intramolecular hydrogen bond interaction is observed. In this case the H-bond involves the carbonyl oxygen and the -NH group of the coordinated N-methylisatoate anion (N-H10···O3 1.096 Å), which also dictates the spatial orientation of the ligand.          Table S6: Mass spectra and putative assignment of the different species detected by LCMS after the irradiation of compound 1 (0.5 mM, λ irr 420 nm, 30.5 J.cm -2 ) in the presence of 5'-GMP (2 mol. eq.). The peak numbers correspond to the labelling in Figure 6.   Table S7: Mass spectra and assignment of the different species detected by LCMS after the irradiation of compound 3 (1.5 mM, λ irr 420 nm, 45 min) in the presence of 5'-GMP (2 mol. eq.). The peak numbers correspond to the labelling in Figure 7.