2.1.

Chemicals and Reagents

Chemical reagents and solvents were purchased from local available sources. The calf thymus DNA and supercoiled (CsCl purified) pBR322DNA were purchased from Takara (Japan). The spin trapper N-methyl-D-glucamine dithiocarbamate (MGD) and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) were purchased from Dojindo (Japan). 2,2,6,6-Tetramethylpiperidine 1-oxyl (TEMPO) was purchased from Sigma.

Cis-$[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ and trans-$[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ complexes were synthesized following a previously described procedure.^24,25 Hydrous nitrosylruthenium trichloride (0.5 mmol) and 2-methyl-8-quinolinol (2.0 mmol) were mixed in 0.1 M acetic acid aqueous solution and the solution was refluxed for 5 h under dinitrogen in the dark. After cooling, the precipitate was collected by filtration and dried under vacuum. The crude product was chromatographically separated using a silica-gel column. The cis isomer complex was eluted from the major band with 30 vol.% $\text{ethylacetate}-{\mathrm{CH}}_2{\mathrm{Cl}}_2$. The cis complex (0.1 mmol) was dissolved in ${\mathrm{CH}}_2{\mathrm{Cl}}_2$, and the solution was irradiated with an Xe lamp through a combination of a UV cut filter and a water filter for 5 h at room temperature. After the solvent was removed, the trans isomer was chromatographically separated using $\text{ethylacetate}-{\mathrm{CH}}_2{\mathrm{Cl}}_2$ as an eluent with a yield of 20%. The cis and trans isomer complexes were confirmed through ${}^1\mathrm{H}$ NMR spectra using a Bruker 600 M spectrometer.

2.2.

Instruments and Measurements

MALDI-TOF MS measurements were performed using a BIFLEX III MALDI-TOF mass spectrometer (Bruker Daltonics) equipped with a nitrogen laser ($\lambda =337\mathrm{nm}$). The mass spectra (20 summed shots) were acquired in the reflector mode with a 19 kV accelerating voltage and a 20 kV reflector voltage. ${\mathrm{CH}}_3\mathrm{CN}$ matrix was used for the MALDI experiment. The matrix solution and analyte solution were mixed and the result was spotted on the MALDI target plate. After air-drying, the samples were analyzed.

EPR spectra were obtained at room temperature by using a Bruker ESP-300E spectrometer at 9.8 GHz, X-band, with a 100 Hz field modulation. Cis or trans isomers in DMSO at 5 mM, mixed with $2\mathrm{mMFe}{(\mathrm{MGD})}_2$, was quantitatively injected into quartz capillaries and then was illuminated in the cavity of the EPR spectrometer with an Nd:YAG laser at 532 nm (5 to 6 ns of pulse width, 10 Hz of repetition, $30\mathrm{mJ}/\mathrm{pulse}$). Nanosecond laser pulses for excitation at 355 nm were generated by an ${\mathrm{Nd}}^{3+}$: YAG laser. The excitation energy of each pulse was $\sim 3\mathrm{mJ}$. All measurements were performed at room temperature.

Real-time measurements of NO were conducted by electrochemistry on TBR4100 four-channel free radical analyzer (World Precision Instruments, USA) equipped with a wide range NO-selective electrode (WPI). The ISO-NOP NO meter (2 mm) was vertically and directly immersed inside the quartz cuvette containing $10\mu \mathrm{M}$ complex solution in 10 mM phosphate buffer at pH 7.4. The TBR4100 analyzer is interfaced with PC via Lab-Trax system. The data are acquired in real time by using the LabScribe software. To avoid any reaction before the light irradiation, the samples were protected from light by using an aluminum foil. The NO release profile was constructed by plotting the current versus time. The current of the buffer solution was set as a zero baseline to calculate the current increment. The complex solution of different isomers in the quartz cuvette was irradiated by an Xe lamp (300 W, Beijing NBeT Corp.) at the central wavelengths of 254, 420, and 550 nm with a bandwidth of 40 nm by several band-pass filters. The irradiation power measured by an optical power meter (LPE-1B, Physcience Opto-Electronics, Beijing) was kept constant at $0.2\mathrm{W}/{\mathrm{cm}}^2$ by adjusting the lamp current and the distance between the light source and sample.

2.3.

DFT Calculations

DFT calculations were performed using the Gaussian 09 program package.²⁶ The original coordinates of the cis and trans isomer atoms were obtained from the crystal structures determined via x-ray diffraction. Visualization was performed using Gaussian view 5.²⁷ All geometries were fully optimized without imposing any symmetry constraint with the Becke’s three-parameter hybrid function with the Lee-Yang-Parr correlation function.^28,29 The standard Wadt-Hay realistic effective core potentials were incorporated in the Gaussian 09 to describe the core electrons of the Ru atom. The 6-311G (d, p) was used for the ligand atoms. The frequency calculated at the DFT level was compared with the experimental frequency to check whether the stationary points from the geometry optimization were in real minima.

2.4.

Photoinduced DNA Damage

The photoinduced DNA damage by the two isomers was investigated via agarose gel electrophoresis. A supercoiled pBR322 DNA ($0.2\mu \mathrm{g}$) was mixed with a two isomer complexes solution at 50 mM, respectively, and incubated at 37°C for 30 min in the dark. The solution was then irradiated at room temperature by an Xe light source with 420-nm bandpass filters ($\sim 200\mathrm{mW}/{\mathrm{cm}}^2$) for 0.5 h. The samples were analyzed via electrophoresis for 1 h at 90 V on a 1% agarose gel in buffer (40 mM $\mathrm{Tris}─{\mathrm{CH}}_3\mathrm{COOH}$, 1 mM ethylenediaminetetraacetic acid, $\mathrm{pH}=8.0$). The gel was stained with $1{\mu \mathrm{gmL}}^{-1}$ EB and photographed under UV light. The mechanistic investigations of pBR322DNA were conducted using free radical scavengers, sodium azide (${\mathrm{NaN}}_3$, 3 mM), which were added to pBR322 DNA prior to the addition of the complex.

2.5.

Cytotoxic Activity

A HeLa tumor cell line was used in this study with standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay procedures. Cells were seeded in 96-well plates ($1\times {10}^3/\mathrm{well}$) in growth medium ($100\mu \mathrm{L}$) and incubated at 37°C under a 5% ${\mathrm{CO}}_2$ atmosphere for 24 h, and then were treated with various concentrations of complexes in a mixture of growth medium. The complexes were dissolved in DMSO and diluted with Roswell Park Memorial Institute medium 1640, and then added to the wells at a different final concentration. The concentration of DMSO was controlled $<1\%$. Control wells were prepared by the addition of a culture medium ($100\mu \mathrm{L}$) without complexes, which was also incubated at 37°C under a 5% ${\mathrm{CO}}_2$ atmosphere for 24 h. MTT ($20\mu \mathrm{l}$ of $5\mathrm{mg}/\mathrm{ml}$) was added to each well, and then the plates were further incubited for 4 h. The culture medium was finally discarded, and $150\mu \mathrm{L}$ of DMSO was added to solubilize the MTT. The solution absorbance at 490 nm was measured with a microplate-reader. The ${\mathrm{IC}}_{50}$ values of complexes were determined by plotting the viability percentage versus concentration on a logarithmic scale and reading off the concentration at which 50% of cells were viable relative to the control.

For comparing the effect of photoirradiation on the cytotoxic activity of isomer complexes, cell cultures upon photoirradiation were performed at the same conditions. After adding $20\mu \mathrm{L}$ complex isomers in a mixture of growth medium at various concentrations, the cells were incubated at 37°C under a 5% ${\mathrm{CO}}_2$ atmosphere for 4 h; then the plates were irradiated with a light source of 420 nm ($\sim 0.2\mathrm{W}/{\mathrm{cm}}^2$) for 30 min and again were incubated at 37°C in a 5% ${\mathrm{CO}}_2$ incubator for 24 h. Upon completion of the incubation, a stock MTT dye solution ($20\mu \mathrm{L}$, $5\mathrm{mg}/\mathrm{mL}$) was added to each well. After 4 h of incubation, the culture medium was finally discarded, and $150\mu \mathrm{L}$ of DMSO was added to solubilize the MTT. The optical density of each well was then measured using a microplate spectrophotometer at 490 nm. Each experiment was repeated at least three times to obtain the mean values.

3.1.

Photoinduced NO Release

3.1.2.

EPR spectroscopy

Iron(II)-N-methyl-D-glucamine dithiocarbamate [${\mathrm{Fe}(\mathrm{MGD})}_2$] is commonly used to trap NO because of its high probability of adduct formation and the high stability of its spin adduct. Spin trapping, together with EPR spectroscopy by using ${\mathrm{Fe}(\mathrm{MGD})}_2$ is considered as one of the best methods for detecting ${\mathrm{NO}}^{•}$ in real time and at its generation site.^30,31 Both $[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ isomers produced NO free radicals through photoexcitation at either 355 or 532 nm.

Figure 2 shows the characteristic triplet signal with a hyperfine splitting constant (hfsc) value of ${\mathrm{a}}_{\mathrm{N}}=12.78\mathrm{G}$ and a g-factor of $g=2.041$, which is consistent with the literature report for $\mathrm{NO}\text{-}{\mathrm{Fe}}^{2+}\text{-}\mathrm{MGD}$ adduct.^32,33 More free radical molecules were generated at the photoexcitation of 355 nm than at the photoexcitation of 532 nm, and the trans isomers produced a little more free radicals at a photoexcitation of 355 nm than the cis isomers at the same conditions.

Fig. 2

Triplet electron paramagnetic resonance signals caused from nitric oxide (NO) trapping by ${\mathrm{Fe}}^{2+}$-N-methyl-D-glucamine dithiocarbamate for the (a) cis and (b) trans isomer complexes upon irradiation with 355- (blue line) and 532-nm (green line) lasers. Dark control (black line) means no light irradiation at identical conditions.

3.1.3.

DFT calculations

DFT calculations were successfully used for investigating the reactivities of ruthenium complexes.³⁴ Figure 3 shows the energy levels and contour plots of the frontier orbitals of cis- and trans-$[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{cqn})}_2\mathrm{NO}]$ complexes. For both isomers, the highest occupied molecular orbital (HOMO) was a 2mqn ligand-based $\pi$ orbital with minimal Ru ($d$) and NO ($p$), whereas the lowest unoccupied molecular orbital (LUMO) was the antibonding overlap of Ru ($d$) and $\pi *\mathrm{NO}(p)$. The orbital analyses indicate that the direct excitation of an electron from a bonding ${\pi }_{2\mathrm{mqn}}-d\pi$ orbital to an antibonding ${\pi *}_{\mathrm{NO}}-d\pi$ orbital causes the photorelease of NO from ruthenium–nitrosyls.

Fig. 3

Contour diagrams of calculated LUMO (top) and HOMO (bottom) of (a) cis-$[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ and (b) trans-$[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$. Positive values of wave function are shown in green.

The previous studies on ruthenium polypyridyl complexes {such as [${[\mathrm{Ru}{(\mathrm{bpy})}_{3-\mathrm{n}}{(\mathrm{L})}_{\mathrm{n}}]}^{2+}$], $L=\text{diimine ligands}$} found that the HOMO was centered on the ruthenium represented by Ru ($d$) $t_{2g}$, whereas the LUMO was localized on the bpy ($\pi *$). Metal-to-ligand charge transfer transition has an important function in photochemical and photophysical processes of these ruthenium-N-donor complexes.^35,36 The different transition contributions of the HOMOs−LUMOs of Ru complexes lead to different pathways and patterns for the exited state, thereby resulting in the formation of different free radical species.

As shown in Fig. 3, the calculated energy of the HOMO–LUMO gap for the trans isomer was smaller than that of the cis isomer, which suggests that the trans isomer is more active upon photoirradiation, agreeing with the experimental observations.

3.2.

Real-Time Measurement of NO Release

Real-time NO release was measured by the electrochemistry with the free radical analyzer. The NO release increased for both $[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ isomers at the photoexcitation of 550, 420, and 254 nm, and white light. Photoirradiation wavelength positions are selected according to the measured absorption spectra: around the absorption peak in the visible region (420 nm); around the lower-absorption region (550 nm); around the absorption peak in the UV region (254 nm); and, finally, white light containing all wavelengths in the visible region. As shown in Fig. 4, the amount of released NO increased as the irradiated wavelength moving from visible region to UV region reached a maximum under white light irradiation; the amount of NO production increased as the absorbance increased in absorption spectra. Therefore, the NO release could be controlled by varying the irradiation wavelength. Moreover, the trans isomers produced slightly more NO free radicals than the cis isomers at the same conditions.

Fig. 4

Real-time measurement of NO release measured with NO-specific electrode upon photoirradiation at different wavelengths.

3.3.

Photoinduced DNA Damage

Since DNA was identified as the primary target of metal-based anticancer agents, considerable attention has been drawn into the interaction of ruthenium complexes with DNA for the potential of ruthenium complexes as chemotherapeutic agents. Further studies on the mechanism show that ruthenium complexes containing polypyridyl, imidazole, and $\beta$-carboline derivatives are antiproliferative and able to induce caspase-dependent apoptosis in cancer cells with mitochondrial dysfunction.^37–40 The photoinduced effect and damage of the $[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ isomers on plasmid DNA were monitored via agarose gel electrophoresis. When circular plasmid DNA is subjected to electrophoresis, a relatively fast migration is generally observed for the intact supercoiled Form I. While the scission occurs on one strand (nicking), the supercoil relaxes to generate a slower-moving, open-circular Form II. If DNA conformation was further changed, a broader linear Form III between Forms I and II is generated.^41–43

Figure 5 shows the photoinduced damage of pBR322 DNA by $[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ isomers. No change was observed in the control experiment (Lane 1) compared with the incubation of the plasmid DNA with the complex in the dark (Lane 2). After irradiation at 420 nm for 30 min, $50\mu \mathrm{M}$ of the complex caused an apparent conformational change of DNA because Form I disappeared and Form III (Lane 3) appeared. These results demonstrate that DNA damage occurred through the photoinduced reaction pathway. In the presence of ${\mathrm{NaN}}_3$, a well-known scavenger of singlet oxygen (${}^1{\mathrm{O}}_2$), there is no evident inhibition of DNA damage for both $[\mathrm{Ru}(\mathrm{OAc}){(2\mathrm{mqn})}_2\mathrm{NO}]$ isomers (Lane 7). By contrast, ${\mathrm{NaN}}_3$ efficiently inhibited DNA cleavage for Ru(II) polypyridyl complexes, in which ${}^1{\mathrm{O}}_2$ was detected by an EPR spin trapping technique and ${}^1{\mathrm{O}}_2$ caused DNA strand cleavage.^44,45

Fig. 5

Agarose gel electrophoresis comparison of the photocleavage of the (a) cis and (b) trans isomers on supercoiled pBR322 DNA with the addition of ${\mathrm{NaN}}_3$ upon light irradiation at 420 nm. Lane 1: DNA alone. Lane 2: $\mathrm{DNA}+\text{isomer complex}$. Lane 3: $\mathrm{DNA}+\text{isomer complex}+\mathrm{hv}$. Lane 4: $\mathrm{DNA}+{\mathrm{NaN}}_3$. Lane 5: $\mathrm{DNA}+{\mathrm{NaN}}_3+\mathrm{hv}$. Lane 6: $\mathrm{DNA}+\text{isomer complex}+{\mathrm{NaN}}_3$. Lane 7: $\mathrm{DNA}+\text{isomer complex}+{\mathrm{NaN}}_3+\mathrm{hv}$.

3.4.

Photocytotoxicity

MTT assay was conducted to evaluate the photocytotoxic potential of the isomer complexes upon photoexcitation using human cervical HeLa cancer cells as shown in Fig. 6. Upon prior incubation for 4 h in the dark and the subsequent exposure to light of 420 nm for 30 min, the trans isomer showed a moderate decrease in the cell viability with ${\mathrm{IC}}_{50}$ values of $10.0\mu \mathrm{M}$ under the light and $29.0\mu \mathrm{M}$ in the dark. The trans isomers exhibited better cytotoxicity and photocytotoxicity activity against the HeLa tumor cell line.

Fig. 6

Cell viability plots showing the photocytotoxicity of the (a) cis and (b) trans isomer complexes in HeLa cells on 4 h incubation in dark followed by exposure to light of 420 nm for 30 min, as determined from the MTT assay. The nonlinear fitted curves for dark-treated and photoexposed cells are shown by black squares and blue circles for cis isomer, and black circles and blue squares for trans isomer, respectively.

Cis-platin is known to give an ${\mathrm{IC}}_{50}$ value of $7.5\mu \mathrm{M}$ in the dark for HeLa cells after 24 h incubation. Incubation of HeLa cells with cis-platin for 4 h in the dark and subsequent exposure to UV-A light gave ${\mathrm{IC}}_{50}$ values of $68.7\mu \mathrm{M}$ under UV-A light of 365 nm, thus showing no apparent PDT effect.⁴⁶ The cis isomer gave an ${\mathrm{IC}}_{50}$ value of $>300\mu \mathrm{M}$ with the cells unexposed to light, while it gave an ${\mathrm{IC}}_{50}$ value of $100\mu \mathrm{M}$ with the cells exposed to light. Therefore, the cis isomers are less toxic and show very low cytotoxicity against the HeLa tumor cell line in dark, but its cytotoxicity activity increased to some extent upon photoirradiation.