Anticancer Ruthenium Complexes with HDAC Isoform Selectivity

The histone deacetylase (HDAC) enzymes have emerged as an important class of molecular targets in cancer therapy, with five inhibitors in clinical use. Recently, it has been shown that a lack of selectivity between the 11 Zn-dependent HDAC isoforms may lead to unwanted side-effects. In this paper, we show that piano stool Ru complexes can act as HDAC inhibitors, and variation in the capping arene leads to differences in HDAC isoform selectivity.

S2 NMR spectra ( 1 H, 13 C) were recorded on a Varian VXR-400 spectrometer ( 1 H at 399.97 Hz, 13 C at 100.57 MHz) or a Varian VNMRS-700 spectrometer ( 1 H at 699.73 MHz, 13 C at 175.95 MHz). Spectra were recorded at 295 K in commercially available deuterated solvents and referenced internally to the residual solvent resonances. The multiplicity of each signal is indicated by s (singlet); d (doublet); t (triplet); q (quartet); quin (quintet) or sept (septet). The number of protons (n) for a given resonance signal is indicated by nH. Coupling constants (J) are quoted in Hz and are recorded to the nearest 0.1 Hz. Identical proton coupling constants (J) are averaged in each spectrum and reported to the nearest 0.1 Hz. The coupling constants are determined by analysis using MestreNova software. Spectra were assigned using COSY, HSQC, HMBC and NOESY experiments as necessary.
Both electrospray and high-resolution mass spectrometry were performed on a Thermo-Finnigan LTQ FT system using methanol as the carrier solvent. m/z values are reported in Daltons with specific isotopes identified.
(4-methylcyclohexa-1,4-dien-1-yl)methanamine 5 p-Toluidine (2.00 g, 16.2 mmol) was added to liquid ammonia (15 ml), precooled to -78 °C. Sodium (1.30 g, 56.8 mmol) was added in small portions and the mixture stirred for 2 h. The reaction was quenched by the addition of methanol (7 ml) before the mixture was warmed to room temperature to remove excess ammonia. The volume of the reaction mixture was reduced to around 2 ml, before water (15 ml) was added and the mixture extracted with Et2O (3 × 15 ml). The organic layers were combined, dried over MgSO4, filtered and the solvent removed under reduced pressure to give the title compound as a crude yellow oil, which was used without further purification (1.37 g, 45%). δH (CDCl3) 5. 63 -5.59 (1H, m, H 2 ), 5.48 -5.44 (1H, m, H 5

Aqueous Stability Assay
To assess stability with respect to hydrolysis of the Ru-Cl bond, 1 H-NMR (D2O, 298 K, 400 MHz) studies were carried out, monitoring the resonances of the capping p-cymene group. To distinguish the Ru-Cl complex (1) and the Ru-OD2 complex (1a), NMR spectra were run in 100 mM NaCl solution (Fig S2 iii) and AgNO3 (Fig S2 iv), respectively. Complex 1 was dissolved in D2O and hydrolysis of the Ru-Cl bond was monitored over the course of 96 h. As shown in Fig S2, after 1 h, the complex S18 remains intact as the Ru-Cl complex. After 96 h, approximately 90% of the complex remains as the Ru-Cl complex, with 10% hydrolysis occurring.

Enzyme Inhibition Assay
HDAC enzyme inhibition assays were carried out for as stated using commercially available assay kits (BPS Bioscience) and undertaken in a 96 well plate. All measurements were carried out in triplicate and standard deviation errors are included in the plots. Human recombinant enzymes HDAC1 FLAG-tag His-tag and HDAC6 GST-tag were used. Potential inhibitors were dosed at the required concentration (either 1 µM/0.1µM for the two concentration point assays or 10 to 0.004 µM for IC50 measurements) from original DMSO stock (10 mM) and diluted using assay buffer (25 mM Tris/Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2). A master mixture was made using assay buffer, acetylated substrate specific to HDAC enzyme (200 µM) and BSA (1 mg/mL) in a ratio of buffer:substrate:BSA (6:1:1) and 40 µL added to each well plate. Buffer (10 µL) was added to all blank wells and inhibitor (5 µL) was added to all potential inhibitor wells. HDAC enzymes were diluted based on their stated activity and 5 µL was added to all wells except blanks and the plate incubated at 37 °C for 30 min. Developer was added and the mixture incubated at room temperature for a further 15 min. Fluorescence was measured using a Synergy H4 microplate reader (λex = 360 nm, λem = 460 nm). Percentage HDAC activity is determined by quantification of fluorescence relative to control, with no added inhibitor. Data from the two point assays with errors from standard deviation are shown in Table S1 (and shown graphically in

Cell Viability Assays
Cellular behaviour of complexes 1, 8a-8f, SAHA and L 1 were conducted using MCF7 human breast adenocarcinoma cells using fluorescence and laser scanning confocal microscopy. Cells were maintained in exponential growth as monolayers in F-12/DMEM (Dulbecco's Modified Eagle Medium) ºC. Cell suspensions were pelleted by centrifugation at 1000 rpm for 3 min and were re-suspended by repeated aspiration with a sterile plastic pipette. Microscopy cells were seeded in 12-well plates on 13mm 0.17mm thick standard glass cover-slips or un-treated iBibi 100 uL live cell channels and allowed to grow to 40% -60% confluence, at 37 ºC in 5% CO2. At this stage, the medium was replaced and cells were treated with compounds in varying concentrations (0.80 -100 µM dissolved in cell culture media) and co-stains as appropriate.
After 96 h incubation, cell toxicity measurements were conducted using a ChemoMetec A/S NucleoCounter3000-Flexicyte instrument with Via1-cassette cell viability cartridge. The cellular stain Acridine Orange was used for cell detection and the nucleic acid stain DAPI was used to detect nonviable cells. The experiments were carried out in triplicate. Cell viability was corrected using a proliferation factor (Equation S1) where: vib I = corrected viability; nx = number of cells after incubation; nc = number of cells on loading; vibx = measured viability.
= (S1) To investigate cell uptake mechanism, MCF7 cells were plated using the same method described in Cell Viability Assay and exposed to complexes 1 at the measured IC50 concentrations (1.53 μM). Cells were allowed to incubate at room temperature for 1 h before being cooled to 4 °C for 24 h. After this time, cell number viability was 98% compared to the initial loading, suggesting no accumulation of complexes in cells at this temperature and, therefore, an uptake mechanism driven by active transport.

Cell Uptake Assay
In cellular uptake studies, cells were seeded in 6-well plates and allowed to grow to 80% -100% confluence, at 37 ºC in 5% CO2. At this stage, the media was replaced with media containing complexes 1 and 8a-8f at their EC50 concentration and total cellular ruthenium was determined using inductively coupled plasma mass spectrometry (ICP-MS).

S25
Cells used for ICP-MS studies were prepared as follows. 1 Cells were cultured in a 6-well plate to 90% confluence. Cells were then counted (10 7 cells based on a cell volume of 4000 µm 3 ) and incubated with medium containing the complex and promoter/inhibitor for 96 h before being washed three times with phosphate-buffered saline (PBS). The cells were then treated with trypsin and harvested and diluted to 1 mL with PBS. Concentrated nitric acid (0.6 mL) was added and the samples were digested for 24 h at 37 °C. These digested samples were submitted for ICP-MS measurements. The samples were run against a series of Ru standards, and the measured concentration was back calculated to find the total Ru concentration present in the original counted cells. In Fig. S11 cell uptake is compared to the inverse of the cell viability EC50 value for all Ru complexes, showing the clear trend between the two values.
Fig. S11 Cell uptake (blue bars) and inverse of cell viability EC50 (yellow circles with black line).

Computational Docking Studies
DFT optimisation of complex 1 was carried out using B3LYP hybrid functionals. For atoms H, C, N and O, the 6-31G* basis set was used, for Ru, the SDD basis set was required. Optimization calculations were carried out in the gas phase. The optimised structure was validated to be the lowest energy structure through vibrational frequencies calculations. The NMR spectra were also calculated and compared to the experimental spectra. All calculations were carried out with Gaussian09. The remaining structures (L 1 , complexes 8c and 8f) were adapted from this optimized compound using Avogadro and bond length data. 2 The 3D model of SAHA was obtained from the ZINC compound database. 3 Genetic Optimisation of Ligand Docking 4 (GOLD) was used to perform molecular docking of SAHA, L 1 and complexes 1, 8c and 8f to the human HDAC1 and HDAC6 proteins (PDB: 4BKX, 5EDU). The Hermes Visualiser was used to prepare the ligands and proteins before docking was undertaken, the ruthenium ion was treated as a dummy atom by GOLD and all water/ligand molecules removed. The region of interest was specified as within 10 Å of the zinc ion present in the binding site. The complex was subjected to 10 genetic algorithm runs, looking for diverse solutions using the ChemScore fitness function. These were then analysed using the ChemScore components and modes of predicted binding.
The docking poses were evaluated, and figures produced, using UCSF Chimera. 5