Inorg. Chem. 2007, 46, 740−749
DNA Binding and Cytotoxicity of Ruthenium(II) and Rhenium(I) Complexes of 2-Amino-4-phenylamino-6-(2-pyridyl)-1,3,5-triazine Dik-Lung Ma,†,‡ Chi-Ming Che,*,† Fung-Ming Siu,† Mengsu Yang,§ and Kwok-Yin Wong‡ Department of Chemistry and Open Laboratory of Chemical Biology of the Institute of Molecular Technology for Drug DiscoVery and Synthesis, The UniVersity of Hong Kong, Pokfulam Road, Hong Kong SAR, China, Department of Biology and Chemistry, City UniVersity of Hong Kong, 83 Tat Chee AVenue, Kowloon, Hong Kong SAR, China, and Department of Applied Biology and Chemical Technology and Central Laboratory of the Institute of Molecular Technology for Drug DiscoVery and Synthesis, The Hong Kong Polytechnic UniVersity, Hunghom, Kowloon, Hong Kong, China Received August 11, 2006
[Ru(tBu2bpy)2(2-appt)](PF6)2 [1‚(PF6)2, tBu2bpy ) 4,4′-di-tert-butyl-2,2′-bipyridine, 2-appt ) 2-amino-4-phenylamino6-(2-pyridyl)-1,3,5-triazine] and [Re(CO)3(2-appt)Cl] (2) were prepared and characterized by X-ray crystal analysis. The binding of 1‚(PF6)2 and 2 to calf thymus DNA (ct DNA) led to increases in the DNA melting temperature (∆Tm ) +12 °C), modest hypochromism (29% and 5% of the absorption bands at λmax ) 450 and 376 nm, respectively), and insignificant shifts in the absorption maxima. The binding constants of 1‚(PF6)2 and 2 with ct DNA, as determined by absorption titration, are (8.9 ± 0.5) × 104 and (3.6 ± 0.1) × 104 dm3 mol-1, respectively. UV−vis absorption titration, DNA melting studies, and competition dialysis using synthetic oligonucleotides [poly(dA−dT)2 and poly(dG−dC)2] revealed that 1‚(PF6)2 and 2 exhibit a binding preference for AT sequences. A modeling study on the interaction between 1 or 2 and B-DNA revealed that the minor groove is the most favored binding site and an extensive hydrogen-bonding network is formed. As determined by MTT assays, 1‚(PF6)2 and 2 exhibited moderate cytotoxicities toward several human cancer cell lines (KB-3-1, HepG2, and HeLa), as well as a multi-drug-resistant cancer cell line (KB-V-1). According to confocal microscopic and flow cytometric studies, 1‚(PF6)2 and 2 induced apoptosis (50−60%) in cancer cells with 3σ(I) R wR GOF
1‚(PF6)2‚MeCN‚0.5Et2O
2
C54H67F12N11O0.5P2Ru 1269.20 triclinic P1h 13.356(3) 14.821(3) 17.251(4) 77.47(3) 75.69(3) 67.17(3) 3021.5(12) 2 1308 1.395 300(2) 0.7107 662 8189 6047 0.096 0.26 1.13
C17H12N6O3ClRe 569.98 monoclinic P2/c (No. 13) 13.800(2) 8.022(2) 18.044(3) 90 102.84(2) 90 1947.6(6) 4 1088 1.944 273.2 0.7107 253 3660 3170 0.047 0.061 2.33
performed on an Agilent 7500 inductively coupled plasma mass spectrometer. X-ray Structure Determination of [Ru(tBu2bpy)2(2-appt)](PF6)2 [1‚(PF6)2] and [Re(CO)3(2-appt)Cl] (2). The diffraction data were collected on a MAR diffractometer with a 300-mm image plate detector at 300 K using graphite-monochromatized Mo KR radiation (λ ) 0.7107 Å). The structures were refined by full-matrix least-squares against Fo2 using the program SHELXL-97.19 Crystal refinement data are listed in Table 1. Perspective views of 1‚(PF6)2 and 2 are presented in Figures 1 and 2, respectively. Figures S1 and S2 in the Supporting Information show crystal packing diagrams for the two complexes. Restriction Endonuclease Fragmentation Assay. Digestion of a plasmid DNA (pDR2, 10.7 kb) with a restriction enzyme, Apa I (Boehringer Mannheim), was performed by mixing pDR2 (21 nM bp-1) in 1× SuRE/Cut Buffer A with Apa I (1 unit/µL) and then incubating the mixture at 37 °C for 1 h.20 The mixture of pDR2 with ethidium bromide (4 µM), Hoechst 33342 (200 µM), cisplatin (cis-Pt(NH3)2Cl2) (200 µM), 1‚(PF6)2 or 2 (2-200 µM), and pDR2 (10.7 kbp, 21 nM bp-1) was first incubated at room temperature for 5 min, and then restriction enzyme (1 unit/µL) was added. Solutions of pDR2 both in the absence and in the presence of the Apa I restriction enzyme in digestion buffer were used as controls. All solutions were incubated at 37 °C for 1 h. After restriction enzyme digestion, the samples were analyzed by agarose gel electrophoresis. Absorption Titration. ct DNA solutions of various concentrations (0-2000 µM bp-1) were added to the metal complexes [50 µM dissolved in Tris buffer (5 mM Tris, 50 mM NaCl, pH 7.2)]. Absorption spectra were recorded after equilibration at 20.0 °C for 10 min. The intrinsic binding constant, K, was determined from a plot of D/∆ap vs D according to eq 121 D/∆ap ) D/∆ + 1/(∆ × K)
(1)
(19) Sheldrick, G. M. SHELXS-97, Program for Crystal Structure Analysis; University of Go¨ttingen: Go¨ttingen, Germany, 1997. (20) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory Press: Woodbury, NY, 1989; p 5.31. (21) Kumar, C. V.; Asuncion, E. H. J. Am. Chem. Soc. 1993, 115, 8547.
DNA Binding and Cytotoxicity of Ru(II) and Re(I) Complexes Competition Dialysis Assay.22a Dialysate solutions (6 mM Na2HPO4, 2 mM NaH2PO4, 1 mM Na2EDTA, and 41 mM NaCl, pH 7.0) containing 1 µM metal complex were used for the competition dialysis experiments. Each DNA sample (200 µL at 75 µM monomeric unit) was pipetted into an individual dialyzer unit (Pierce). Four dialysis units were placed in a beaker containing 400 mL of dialysate solution. The beaker was covered with Parafilm and wrapped in foil. The solution was equilibrated by continuous stirring at room temperature (20-22 °C) for 48 h. At the end of the equilibration period, DNA samples were transferred to microfuge tubes and treated with 1% SDS. The concentration of metal complex (Ct) within each dialysis unit and the free metal complex concentration (Cf) that remained in dialysate solution were each determined spectrophotometrically. The concentration of bound metal complex (Cb) was determined according to eq 2 Ct - Cf ) Cb
Figure 1. Perspective view of the [Ru(tBu2bpy)2(2-appt)]2+ (1) cation (50% probability ellipsoids, hydrogen atoms omitted for clarity). Selected bond lengths (Å): Ru(1)-N(1) 2.118(6), Ru(1)-N(6) 2.060(6), Ru(1)-N(8) 2.077(7), Ru(1)-N(10) 2.107(6). Selected bond angles (deg): N(6)-Ru(1)-N(8) 98.3(2), N(6)-Ru(1)-N(10) 172.6(2), N(1)-Ru(1)-N(9) 86.5(2), N(1)-Ru(1)-N(7) 97.7(2).
(2)
Viscosity Experiments. Titrations were performed by adding 1‚(PF6)2 or 2 to ct DNA in BPE buffer (6 mM Na2HPO4, 2 mM NaH2PO4, 1 mM Na2EDTA, and 41 mM NaCl, pH 7.0). The DNA concentration was approximately 1 mM (in base pairs). Mixing of the solution in the viscometer was achieved by briefly bubbling nitrogen gas through it. The relative viscosity of DNA in the presence and absence of the metal complex was calculated using eq 3 η ) (t - to)/to
(3)
where t is the flow time of the DNA-containing solution and to is the flow time of buffer alone. According to Cohen and Eisenberg,18a the relationship between the relative solution viscosity (η/ηo) and contour length (L/Lo) is given by eq 4 L/Lo ) (η/ηo)1/3
Figure 2. Perspective view of [Re(CO)3(2-appt)Cl] (2). Selected bond lengths (Å): Re(1)-N(1) 2.167(6), Re(1)-N(2) 2.200(4), Re(1)-Cl(1) 2.497(2), Re(1)-C(1) 1.887(7). Selected bond angles (deg): Cl(1)-Re(1)-N(1) 84.0(2), Cl(1)-Re(1)-C(1) 178.8(2), N(1)-Re(1)-C(1) 97.0(2), N(1)-Re(1)-N(2) 74.7(2).
where D is the concentration of DNA in base pairs; ∆ap ) |A F|; A ) Aobs/[complex]; and ∆ ) |B - F|, with B and F corresponding to the extinction coefficients of the DNA-bound complex and unbound complex, respectively. DNA Melting Study. Solutions of the 33-bp double-helical DNA molecule (50 µM bp-1) both in the absence and in the presence of the metal complexes (DNA base pair/metal complex ) 1:1) were prepared in Tris buffer. The temperature of the solution was increased at a rate of 1 °C min-1, and the absorbance at 260 nm was monitored. Tm values were determined from plots of absorbance vs temperature.
(4)
where Lo and ηo denote the apparent molecular length and solution viscosity, respectively, in the absence of the metal complex. Cytotoxicity Tests [MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5tetrazolium Bromide) Assays]. Cells (20 000 cells/well) were seeded in a 96-well flat-bottomed microplate containing 150 µL of growth medium [10% fetal calf serum (FCS, Gibco), 1% Sigma A-7292 Antibiotic and Antimycotic Solution in minimal essential medium (MEM-Eagle, Sigma)]. Complex 1‚(PF6)2, complex 2, and cisplatin (positive control) were dissolved in DMSO (dimethyl sulfoxide) and mixed with the growth medium (final percentage of DMSO e 4%). The microplate was incubated at 37 °C with 5% CO2/95% air in a humidified incubator for 48 h. After incubation, 10 µL of MTT reagent (5 mg/mL) was added to each well. The microplate was re-incubated for 4 h, with 100 µL of solubilization solution (10% SDS in 0.01 M HCl) added to each well. The microplate was then left in the incubator for 24 h, before the absorbance at 550 nm was measured using a microplate reader. The IC50 values (concentrations required to reduce the absorbance by 50% compared to the controls) for 1‚(PF6)2 and 2 were determined by dose dependence of surviving cells after exposure to the metal complexes for 48 h. Molecular Modeling. Molecular docking was performed using the ICM-Pro 3.4-8a program (Molsoft). To simplify the calculation, no force field of the metal atom was input in the calculations. In (22) (a) Sun, D.; Hurley, L. H.; von Hoff, D. D. Biotechniques 1998, 25, 1046. (b) McKeage, M. J.; Berners-Price, S. J.; Galettis, P.; Bowen, R. J.; Brouwer, W.; Ding, L.; Zhuang, L.; Baguley, B. C. Cancer Chemother. Pharmacol. 2000, 46, 343.
Inorganic Chemistry, Vol. 46, No. 3, 2007
743
Ma et al. Table 2. UV-Vis Absorption Data for 1‚(PF6)2 and 2 and Solution Emission Data for 1‚(PF6)2 in Various Solvents at 298 K 1‚(PF6)2
a
2
solvent
absorption λ/nm (max/dm3 mol-1 cm-1)
emissiona λmax/nm
MeCN MeOH EtOH CH2Cl2
215 (57000), 283 (72100), 450 (12400) 218 (66000), 283 (93000), 451 (16900) 215 (57000), 283 (72000), 451 (12900) 215 (57000), 283 (73000), 450 (12500)
646 640 635 640
φemb
absorption λ/nm (max/dm3 mol-1 cm-1)
0.001 0.0007 0.0013 0.001
266 (30000), 376 (br, 5900), 440 (sh, 1200) 266 (27800), 374 (br, 5700), 440 (sh, 1200) 268 (28900), 376 (br, 5900), 440 (sh, 1300) 269 (30200), 383 (br, 5200), 440 (sh, 1200)
Excitation at 450 nm. b Emission lifetime e 0.1 µs.
other words, a soft “static” (or “frozen”) approximation was kept for DNA and the metal atom during the calculation. According to the ICM method, the molecular system was described using internal coordinates as variables. The biased-probability Monte Carlo (BPMC) minimization procedure was used for global energy optimization. The BPMC global energy optimization method consists of the following steps: (1) a random conformation change of the free variables according to a predefined continuous probability distribution, (2) local energy minimization of analytical differentiable terms, (3) calculation of the complete energy including non-differentiable terms such us entropy and solvation energy, (4) acceptance or rejection of the total energy based on the Metropolis criterion and return to step 1. The binding between 1, 2, or 2-appt and DNA molecule was evaluated in terms of the binding energy, where the energy reflects the quality of the complex and includes grid energy, continuum electrostatic, and entropy terms. The DNA crystal structure (PDB code 423D), which contained 12 base pairs (8 G-C base pairs) and more than 800 atoms, was downloaded from Protein Data Bank. In the docking analysis, the binding site was assigned across the entire major and minor grooves of the DNA molecule. ICM docking was performed to find the most favorable orientation. The resulting 1/2-DNA complex trajectories were energy minimized, and the interaction energies were computed. Confocal Microscopy. KB-3-1 cells treated with 1‚(PF6)2 (50 µΜ) or cisplatin (22 µΜ) for 72 h were incubated in 5% CΟ2 at 37 °C. The treated and untreated (as control) cells (1 mL) were stained with an acridine orange (AO)/ethidium bromide (EB) solution (40 µL, 50 mg/mL AO, 50 mg/mL EB in phosphate buffer) and then examined by laser confocal microscopy (Zeiss Axiovert 100M). Flow Cytometric Analysis. Flow cytometry measurements were performed with an EPICS XL cytometer (Coulter Corporation, Miami, FL) equipped with an argon laser. The sheath fluid was an isotonic solution (Isoflow Coulter 8547008, Coulter Corporation). An excitation wavelength of 488 nm at 15 mW was used. About 10 000 cells were analyzed for each sample. HeLa cells were cultured to a density of ∼2 × 105 cells/mL, and then 1‚(PF6)2 (50 µM) or Staurosporin Streptomyces (as a positive control) were added. Cells were incubated in 5% CO2 at 37 °C, and were collected at 12- and 18-h intervals. The DNA was extracted according to a previously described procedure11 and evaluated using an Annexin V protocol. Cellular Uptake Studies. Cellular uptake experiments were conducted according to a literature method with some modifications.22b HeLa cells (5 × 104 cells) were seeded in 60-mm tissue culture dishes containing culture medium (2 mL/well) and incubated at 37 °C in an atmosphere of 5% CO2/95% air for 24 h. The culture medium was removed and replaced with medium containing the ruthenium or rhenium complex at a concentration of 5, 25, or 50 µM. After exposure to the ruthenium or rhenium complex for 2 h, the medium was removed, and the cell monolayer was washed four times with ice-cold PBS. Milli-Q water (500 mL) was added, and the cell monolayer was scraped off the culture dish. Samples (300
744 Inorganic Chemistry, Vol. 46, No. 3, 2007
mL) were digested in 70% HNO3 (500 mL) at 70 °C for 2 h and then diluted 1:100 in water for inductively coupled plasma mass spectrometry (ICP-MS) analysis.
Results and Discussion Mingos and co-workers previously reported the use of pyridyl-2,4-diamino-1,3,5-triazines as H-bonding scaffolds for the construction of metal complexes for crystal engineering studies.13 Following the reported procedure,13 we prepared 2-amino-4-phenylamino-6-(2-pyridyl)-1,3,5triazine (2-appt) through the condensation of 1-phenylbiguanide with 2-pyrdinecarboxamide in alkaline methanol. The reaction of 2-appt with cis-[Ru(tBu2bpy)2(acetone)2]2+ and [Re(CO)5Cl] gave 1‚(PF6)2 and 2 in yields of 70% and 84%, respectively. The structures of 1‚(PF6)2 and 2 were established by X-ray crystallography. The metal-ligand bonding parameters for both complexes are comparable to those previously reported for ruthenium(II) and rhenium(I) diimine complexes.23,24 In 2, the Re(1)-N(1) and Re(1)-N(2) distances are 2.167(6) and 2.200(4) Å, respectively (Figure 2), which are close to the related Re-N distances of 2.18(1) and 2.20(1) Å in [Re(dppz)(CO)3(py)](O3SCF3).25 As shown in the crystal structure of 2 (Figure 2), the Cl(1)-to-H(3) contact distance is 2.18 Å, suggesting the presence of an intermolecular hydrogen bond. The 1H NMR spectra of 1‚(PF6)2 and 2 contain the expected numbers of proton signals; however, extensive overlapping of the 1H signals in the aromatic region renders complete spectral assignment difficult. The absorption spectrum of 1‚(PF6)2 in acetonitrile includes a broad absorption at 450 nm that is assigned to the [dπ(Ru) f π*(diimine)] MLCT transition.26 Complex 2 exhibits two absorption bands with λmax (max/dm3 mol-1 cm-1) at 376 nm (sh, 5.9 × 103) and 266 nm (3.0 × 104) in acetonitrile. Previous studies on rhenium(I)-diimine complexes have revealed that the [ReI f π*(diimine)] transition is solventsensitive.27 However, changing the solvent from MeCN to MeOH or EtOH does not affect the absorption peak maximum of 2 at ∼376 nm (Table 2), indicating that it corresponds to an intraligand transition. Excitation of 1‚(PF6)2 (50 µM) at 450 nm in acetonitrile resulted in a weak emission with λmax at 646 nm and a (23) Yam, V. W. W.; Lo, K. K. W.; Cheung, K. K.; Kong, R. Y. C. J. Chem. Soc., Chem. Commun. 1995, 1191. (24) Farah, A. A.; Stynes, D. V.; Pietro, W. J. Inorg. Chim. Acta 2003, 343, 295. (25) Yam, V. W. W.; Lo, K. K. W.; Cheung, K. K.; Kong, R. Y. C. J. Chem. Soc., Dalton Trans. 1997, 2067. (26) Hartshorn, R. M.; Barton, J. K. J. Am. Chem. Soc. 1992, 114, 5919. (27) Wrighton, M.; Morse, D. L. J. Am. Chem. Soc. 1974, 96, 998.
DNA Binding and Cytotoxicity of Ru(II) and Re(I) Complexes
Figure 3. Inhibition of plasmid Apa I (restriction endonuclease) digestion of pDR2 (10.7 kbp, 21 nM bp-1) by various small molecules. Lane A is size marker. Lanes B and C are undigested and Apa I (1 unit/µl) digested products of pDR2, respectively. Lanes D-F are the digested products of pDR2 in the presence of ethidium bromide (4 µΜ) (lane D), Hoechst 33342 (200 mM) (lane E), and cisplatin (200 mM) (lane F). Lanes G-I are the digested products of pDR2 in the presence of 1‚(PF6)2 at 100 mM (lane G), 20 mM (lane H), and 2 mM (lane I).
quantum yield of 0.001 with a lifetime of 200 52.3 ( 3.4 43.5 ( 2.2 22.1 ( 3.6
>200 199 ( 3 195 ( 1 39.1 ( 1.7
>200 30.2 ( 6 30.9 ( 1.1 10.5 ( 0.5
>200 59.7 ( 0.5 50.3 ( 0.2 11.6 ( 0.2
greater negative electrostatic potentials than DNA stretches that are rich in GC. Molecular Modeling. The results of spectroscopic titrations and viscosity experiments suggest that 1‚(PF6)2 or 2 does not interact with DNA via intercalation. To provide further insight into the possible modes of interaction, we examined the binding of 1‚(PF6)2, 2, and 2-appt toward DNA using molecular modeling (ICM).36 Complex 1 exhibits the highest binding interaction (as revealed by the calculated binding energy of -57.84 kcal mol-1) toward B-DNA, and as shown in Figure 8a, it is located close to the AT-rich (36) Totrov, M.; Abagyan, R. Proteins, Suppl. 1997, 1, 215.
748 Inorganic Chemistry, Vol. 46, No. 3, 2007
CCD-19Lu >200 151 ( 3 112 ( 3 129 ( 1
sequence of the minor groove. This finding is consistent with the proposed specificity from the spectroscopic titrations and competitive dialysis experiments. The optimal binding conformations in the minor groove through H-bonding interactions are shown in Figure 8b. It can be seen in this figure that 1 fits the groove shape well, forming an extensive hydrogen-bonding network. Similar modes of binding were observed for the interactions of 2 and 2-appt with B-DNA, with both binding to AT-rich stretches in the minor groove and with 2 having a relatively lower calculated binding energy (-46.58 vs -56.52 kcal mol-1; Figures S5 and S6). The results indicate that the effects, due either to the
DNA Binding and Cytotoxicity of Ru(II) and Re(I) Complexes
Figure 9. Percentages of cell death for HeLa cells induced by 1‚(PF6)2 and 2 after 12 h of incubation. Staurosporin Streptomyces was used as a positive control.
ruthenium/rhenium atoms or the ancillary ligands other than 2-appt, do not significantly impede the formation of H-bonds between the complexes and DNA. Cytotoxicity Tests (MTT Assay). The cytotoxicities of 1‚(PF6)2 and 2 against four human carcinoma cell lines (KB3-1, KB-V-1, HepG2, HeLa) and a noncancerous CCD-19Lu cell line were studied, with the results listed in Table 5. Using the MTT assay, the IC50 values were calculated from the numbers of cells that survived at each particular complex concentration, after exposure for 48 h. The IC50 values (final concentration e4% DMSO) were found to be in the following order: cisplatin < 1‚(PF6)2 ≈ 2. There were no significant variations in the IC50 values for either 1‚(PF6)2 or 2 between the various cell lines. The complexes exhibited significant potency against the multi-drug-resistant KB-V-1 cell line, but were less toxic toward the noncancerous CCD-19Lu cell line. Because of the production of lactic acid, a slightly acidic milieu of pH 6.8 is frequently encountered in solid tumors, as compared to a pH of 7.2-7.4 in normal tissues. For comparative purposes, we performed cytotoxicity tests with 1‚(PF6)2 and 2 in a HeLa cell line at pH 6.8. Results indicated that there were no significant variations in the IC50 values (∼50 µM) obtained at pH 6.8, compared to those determined at pH 7.2-7.4 (∼50-60 µM). Confocal Microscopy and Flow Cytometric Analysis. The cell death induced by 1‚(PF6)2 and 2 was quantitatively examined by flow cytometry using the Annexin V protocol.11 On the basis of their overall cell morphology and cell membrane integrity, necrotic and apoptotic cells could be distinguished from one another using laser confocal microscopy.20,37 In the absence of cisplatin or 1‚(PF6)2 (Figure S7), the nuclei of living cells were stained bright green in spots. However, after treatment with either cisplatin or 1‚(PF6)2, green apoptotic cells containing apoptotic bodies, as well as red necrotic cells, were observed. As shown in Figure 9, 1‚(PF6)2 and 2 were found to induce 50-60% apoptosis, with less than 5% necrosis detected. The results show that (37) (a) Schwartzman, R. A.; Cidlowski, J. A. Endocr. ReV. 1993, 14, 133. (b) Vermes, I.; Haanen, C. AdV. Clin. Chem. 1994, 31, 177.
1‚(PF6)2, like cisplatin, can induce apoptosis in the KB-3-1 cell line. Analogous results were observed for 2. Cellular Uptake Studies. Cellular uptake studies were performed to ascertain whether the cytotoxicity results correlated with cellular levels. The accumulation of 1 and 2 in HeLa cells is shown in Figure S8. The results suggested that 1 and 2 were accumulated in HeLa cells to approximately equal extents. Both 1 and 2 could effectively enter the cancer cell, and the cellular uptake levels were dependent on the dose of each complex used. This also indicates that their cytotoxicities as determined by MTT assays were not disproportionately influenced by the complexes having different cellular uptake levels. Conclusions [Ru(tBu2bpy)2(2-appt)](PF6)2 [1‚(PF6)2, tBu2bpy ) 4,4′di-tert-butyl-2,2′-bipyridine, 2-appt ) 2-amino-4-phenylamino-6-(2-pyridyl)-1,3,5-triazine] and [Re(CO)3(2-appt)Cl] (2) bind to ct DNA with binding constants of (8.9 ( 0.5) × 104 and (3.6 ( 0.1) × 104 dm3 mol-1, respectively, at 20 °C. The data from UV-vis absorption titration, melting studies, and competition dialysis experiments all support a specific binding mode for 1‚(PF6)2 and 2 at AT-rich regions of DNA. The binding of 1‚(PF6)2 or 2 to DNA led to modest hypochromicity and insignificant shifts in the absorption maxima. Results from spectroscopic titrations and viscosity experiments were consistent with the complexes having low binding constants (∼104) and indicated that these metal complexes interact with DNA via groove binding. Modeling studies suggest that the minor groove is the favored binding site for 1 and 2. Complexes 1‚(PF6)2 and 2 exhibited moderate cytotoxicities toward several established human cancer cell lines, including KB-3-1, HepG2, and HeLa, and had significant activity against the multi-drug-resistant KB-V-1 cancer cell line. According to results from confocal microscopic and flow cytometric studies, 1‚(PF6)2 and 2 selectively induced apoptosis (50-60%) leading to cancer cell death, with necrosis levels generally below 5%. As the electronic structure and properties of metal complexes can be modulated through the metal ion and auxiliary ligands, we envision that metal complexes containing analogous H-bonding motifs have significant potential in the development of specific metal-based groove binders. Acknowledgment. This work was supported by the Areas of Excellence Scheme established under the University Grants Committee of the Hong Kong Special Administrative Region, China (AoE/P-10/01); The University of Hong Kong (University Development Fund); and the Research Grants Council of Hong Kong SAR, China (CityU 1189/01P). Supporting Information Available: Figures S1-S8 and CIF files for the crystal structures of [Ru(tBu2bpy)2(2-appt)](PF6)2 [1‚ (PF6)2] and [Re(CO)3(2-appt)Cl] (2). This material is available free of charge via the Internet at http://pubs.acs.org. IC061518S
Inorganic Chemistry, Vol. 46, No. 3, 2007
749
