Copper(I) complexes of modified nucleobases and vitamin B3 as potential chemotherapeutic agents: In vitro and in vivo studies
N J M Sanghamitraa, M K Adwankarb, A S Juvekarb, V Khurajjamb & C Wycliff a, A G Samuelsona
aDepartment of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore 560 012, India Email: ashoka@ipc.iisc.ernet.in
bDivision of Chemotherapy, Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Centre, Navi Mumbai 410 208, India
Received 27 October 2010; accepted 3 January 2011
Three new complexes of Cu(I) have been synthesized using ancillary ligands like thiopyrimidine (tp) a modified nucleobase, and nicotinamide (nic) or vitamin B3, and characterized by spectroscopy and X-ray crystallography. In vitro cytotoxicity studies of the complexes on various human cancer cell lines such as Colo295, H226, HOP62, K562, MCF7 and T24 show that [Cu(PPh3)2(tp)Cl] (1) and [Cu(PPh3)2(tp)]ClO4 (2) have in vitro cytotoxicity comparable to cisplatin.
Complex [Cu(nic)3PPh3]ClO4 (3) is non-toxic and increases the life span by about 55 % in spontaneous breast tumor model.
DNA binding and cleavage studies show that complex (3) binds to calf thymus DNA with an apparent binding constant of 5.9 × 105 M and completely cleaves super-coiled DNA at a concentration of 400 µM, whereas complexes (1) and (2) do not bind DNA and do not show any cleavage even at 1200 µM. Thus, complex (3) may exhibit cytotoxicity via DNA cleavage whereas the mechanism of cytotoxicity of (1) and (2) probably involves a different pathway.
Keywords: Bioinorganic chemistry, Metallodrugs, Copper, Antitumor activity, Thiopyrimidine, Nicotinamide, Cytotoxicity, Lipophilicity
The FDA approved gold phosphine based antiarthritic drug, auranofin, and antitumor drug, cisplatin, have created immense interest in inorganic chemists to develop new metallodrugs.1,2 Soft metal centers like Cu(I) and Au(I) are known to possess biological activity3-5 but need to be stabilized in biological media with sulfur or phosphorous containing ligands. In many instances the biological properties of the ligands are enhanced on complexation with metals, as reported for 1,2-bis(diphenylphosphinoethane) (DPPE) where the anticancer activity of DPPE increases on coordination to Cu(I) and Au(I).5 DPPE based Cu(I) complexes have shown promising antitumor activity in a range of cell lines like PA1, CHO and human ovarian carcinoma cell lines.3 The thioglucose derivative auranofin is also an effective cytotoxic agent against p388 leukemia.2 The Au(I) trialkylphosphine complexes of 2-thiopyridine and 2-thiopyrimidine are reported to have carcinogenic and antitumor properties and have been used as drugs.6-8 Nicotinamide (vitamin B3) is known for its cytostatic activity9,10 achieved by maintaining mitochondrial membrane potential, but metal complexes of nicotinamide do not appear to have been used as drugs.
In this report, we have described the synthesis of three Cu(I) complexes of PPh3 with heterocyclic thione thiopyrimidine and nicotinamide. The in vitro antitumor activity in different cell lines such as colon carcinoma (Colo205), human lung carcinoma (H226 and HOP62), human erythroid leukemia (K562), human breast carcinoma (MCF7) and human bladder carcinoma (T24) was studied by semi-automated sulforhodamine-B assay (SRB assay)11 while in vivo activity was carried out in murine tumor models, viz., spontaneous mouse-mammary tumor, L1210-murine leukemia and P388 leukemia. Previous study on the detailed mechanism of action of a Cu(I) complex of DPPE reveals that the complexes bind to calf thymus (CT) DNA and cleave DNA in vitro and in vivo.12 Hence, we have studied the DNA binding ability of these complexes with CT DNA and in vitro cleavage of supercoiled (SC) DNA. The lipophilicity of the complexes was studied by UV-visible spectroscopic technique to understand their activity profiles.
Materials and Methods
The solvents, acetonitrile, dichloromethane and petroleum ether, were dried and distilled over calcium hydride, P2O5 and sodium ketyl radical respectively.
Thiopyrimidine and calcium hydride were obtained from Aldrich, USA. PPh3, nicotinamide and acetic acid were obtained from SD Fine Chem. Ltd., India.
Cu(CH3CN)4ClO4 and CuCl were synthesized by literature methods.13,14 The calf thymus DNA, agarose (molecular biology grade), carboxymethyl cellulose (CMC) and ethidium bromide, were purchased from Sigma, USA. Supercoiled pUC19 (cesium chloride purified) DNA was from Bangalore Genei, India.
RPMI 1640 and fetal bovine serum (FBS) were procured from GIBCO-BRL, Invitrogen Life Technologies, USA. Cisplatin was purchased from SD Fine Chem. Ltd., India. Sterile plasticware were purchased from Nunc, Denmark. Tris-HCl buffer solution was prepared by using milli-Q water.
1H-NMR spectra were recorded either on a Bruker ACF 200 MHz or AMX 400 MHz spectrometer with tetramethylsilane (TMS) as the internal reference.
31P{1H} NMR spectra were recorded either on a Bruker AMX 400 MHz spectrometer operating at 162.2 MHz or a Bruker ACF 200 MHz operating at 81.1 MHz with H3PO4 (85 %) as the external reference. IR spectra were recorded in the solid state as KBr pellets on a Bruker Equinox 55 spectrometer.
UV-vis spectra were recorded with Perkin Elmer (Lambda 55) spectrometer. Emission spectra were recorded with Perkin Elmer (LS 50B) fluorescence spectrometer. ELISA plates of SRB assay were read on the Sunrise model of Tecan, Austria.
Synthesis and characterization of the complexes Cu(thiopyrimidine)(PPh3)2Cl (1)
CuCl (0.099 g, 0.001 mole) was reacted with PPh3
(0.524 g, 0.002 mole) in acetonitrile (30 mL) for 10 min before adding thiopyrimidine (0. 112 g, 0.001 mole). A bright orange yellow solution was formed, which gave a deep yellow precipitate after 1 hour. The precipitate was filtered and re-dissolved in dichloromethane and layered with petroleum ether to give orange yellow crystals suitable for single crystal XRD. The complex was obtained in nearly quantitative yield. 31P and 1H NMR was recorded in CDCl3. 6.6 (t, 1H, tp), 8.1(s, 2H, tp), 7.18 – 7.46 (m, 30H, PPh3). 31P NMR gives a single peak at –3.41ppm.
Cu(thiopyrimidine)(PPh3)2ClO4 (2)
Cu(CH3CN)4ClO4 (0.327 g, 0.001 mole) was reacted with PPh3 (0.524 g, 0.002 mole) in acetonitrile (30 mL) for 10 min. To this solution thiopyrimidine (0.112 g, 0.001 mole) was added. A bright orange
yellow solution was formed. The solvent was removed under vacuum and the volume of the solution was reduced to half and layered with diethyl ether, to give orange yellow crystals, suitable for single crystal XRD. The complex was obtained in nearly quantitative yield. 31P and 1H NMR was recorded in CDCl3. 1H NMR (CDCl3): 6.6(t, 1H, tp), 8.07(s, 2H, tp), 7.04 – 7.57(m, 30H, PPh3). 31P NMR gives a single peak at 0.9ppm. IR (KBr pellet): 1095 (ν ClO4).
Cu(nic)3PPh3ClO4 (3)
Cu(CH3CN)4ClO4 (0.327 g, 0.001 mole) was reacted with PPh3 (0.262 g, 0.001 mole) in dichloromethane (20 mL) for 1 hr, and then nicotinamide (0. 366 g, 0.003 mole) was added. After 30 min, an off-white precipitate was formed. This was stirred for another 24 h to ensure the completion of reaction, which was checked by the absence of PPh3
in the supernatant solution by TLC. The precipitate was dissolved in hot ethanol, which on cooling gave colorless crystals and a blue solution. These crystals were found to contain three molecules of nicotinamide, one PPh3 coordinated to Cu(I) when analyzed by XRD. (Yield 88 %). 31P and 1H NMR was recorded in acetone d6. 6.95(s, 3H, nic), 7.4-7.47 (m, 15 H PPh3 and 3 H nic), 7.7 (s, 3H nic) 8.4 (s, 3H nic). 31P NMR (0.04). IR (KBr pellet): 1090 (ν ClO4).
X-ray crystallography
Single crystals of complexes (1), (2) and (3) were separately glued to the tip of glass fibers along the largest dimension. Data were collected on a Bruker AXS single crystal diffractometer equipped with Smart Apex CCD detector and a sealed Mo-Kα source working at 2.2 KW and 50/35 (kv/mA). Intensity data were collected at room temperature. Crystallographic computations were performed using the WINGX (1.63.02) package.15 Data was corrected for Lorentz and polarization effects. The structures were solved by the combination of Patterson and Fourier techniques and refined by full-matrix least-squares method using the SHELX program.16 The hydrogen atoms of the complex (1) were geometrically fixed and the hydrogen atoms for complex (2) were located from the difference Fourier map and refined.
Description of biological assays SRB assay17
Human tumor cell lines were cultured in RPMI- 1640 medium supplemented with (FBS) (10 %) at
37 °C and were maintained in a CO2 incubator in an atmosphere of 5 % CO2 in tissue culture flasks. The confluent cultures (70 %) were used to determine the cytotoxic effects of the test compounds. Single cell suspension of these tumor cells was made and cell count was adjusted to 1 × 105 to 5 × 105 cells/mL. Cell number for seeding was derived from a calibration curve set up with known number of cells, for each cell line. 96-well plate was seeded with this cell suspension, each well receiving 90 µl of it. The plate was then incubated at 37 °C temperature in CO2
incubator for 24 h to ensure adequate cell-growth prior to determination of cell growth inhibition. The drugs (10 µl) were then added at appropriate concentrations, followed by further incubation for 48 h. Experiment was terminated by gently layering the cells in the wells with 30 % chilled TCA. The plates were kept in refrigerator for 1 h, following which they were washed thoroughly with tap water, dried and stained with 0.4 % SRB in 1 % acetic acid.
Excess SRB dye was removed by washing the plates, 3 to 4 times, with 1 % acetic acid. The bound SRB was eluted with Tris (10 mM, pH 10.5). Absorbance was read at 540 nm with 690 nm reference wavelength, in the ELISA-plate reader. Optical density of drug-treated cells was compared with that of control cells and growth inhibition was calculated as percent values. Each compound was tested at four different concentrations (10, 20, 40 & 80 µg/mL), in triplicate, on the human malignant cell lines.
Concentration for 50 % growth-inhibition (IC50)
≤ 10 µg/mL was considered to indicate activity. For each of the experiments, a known anticancer drug cisplatin was used as a positive control.
In vivo xenograft studies
Anticancer activities of complexes (1) and (2) were evaluated in the P388 leukemia, L1210 and Lewis lung carcinoma xenograft models. Male BDF-1 mice was used in all the carcinoma models except in P388 leukemia, where female BDF-1 mice was used. All of the studies using laboratory animals were approved by the Institutional Ethics Committee for ‘Animal Care and Use’ of the Advanced Centre for Treatment, Research and Education, Navi Mumbai, India and all the applicable institutional and governmental guidelines for the human care and use of laboratory animals were adhered to.
In vivo anticancer activity of complexes (1) and (2) was evaluated in lewis lung carcinoma, P388 murine leukemia and L1210 mouse model. In the case of
lewis lung carcinoma model, on day 0 of each experiment, the tumor was removed and minced.
Normal saline was added and 0.2 mL was administered to each animal by intramuscular injection. On day 1, all mice were randomized and then divided into three groups each group contain six mice. Mice weights were taken on day 1 and 5.
Treatment schedule for the mice groups was from day 7 to day 15 by 1 trough 9 schedule via intraperitonial (IP) injections. Drugs were prepared in 10 % DMSO (first weight of drug was taken and dissolved in 100 % DMSO and then distilled water was added to make final concentration of DMSO to 10 %). The mice received IP injections of 0.1 mL of the compounds (100 mg/kg) at morning from day 1 to day 9. On day 7, day 11, day 15 and day 19 tumor volume was measured in cc.
In the case of P388 murine leukemia model, 1×106 cells/mouse were injected via IP into BDF1 female mice on day 0. On day 1, all mice were randomised and then divided into five groups, each group containing six mice. Weights of mice were taken on day 1 and day 5. Treatment schedule for the mice groups was from day 1 through day 9. Drugs were prepared in 10 % DMSO in one experiment and CMC in a second experiment. The mice received IP injections of 0.2 ml of the compounds (100 mg/kg) at morning from day 1 to day 9. On day 7 and day 9 tumor volume was measured in cc. For L1210 model, 1 × 105 cells/mouse were injected via IP into the BDF1 male mice on day 0. On day 1, all mice were randomised and then divided into seven groups, each group containing six mice. Weights of mice were taken on 1st and 5th days. Treatment schedule for the mice groups was from day 1, day 5 and day 9. Drugs were prepared in CMC and the mice received IP injections of 0.2 mL of the compounds (75, 50 and 25 mg/kg) at morning from day 1, day 5 and day 9.
On day 1, day 5, day 7, day 11, day 15 and day 19, tumor volume was measured in cc.
Anticancer activity of complex (3) was evaluated in spontaneous breast carcinoma and L1210 mouse model. In both the models, cells were injected intraperitonially into BDF1 female mice on day 0. On day 1, all mice were randomised and then divided into five groups, each group containing six mice.
Treatment schedule for the mice groups was from day 1 through day 9. The drug was prepared in water. The mice received IP injections of 0.2 mL of the compounds (30 mg/kg) for L1210 model and
50 mg/kg for spontaneous breast carcinoma model at morning on day 1, day 5 and day 9. From day 5 through day 35, tumor volume was measured in cc.
Tumor volume (Tvol) was calculated in accordance with the equation; Tvol = L × W2 × 0.5, where L is the maximum length of the tumor and W is the minimum length. At the end of study the mice were sacrificed by cervical dislocation and the tumors were excised and fixed in 10 % phosphate buffered formalin until further evaluation. The tumor volumes are expressed as mean ± standard error for three mice in each group.
One-way analysis of variance (ANOVA) was used for multiple comparisons followed by Student’s t-test to find out difference between individual treatments.
Body weights and water consumption were measured in allthe experiments to assess toxicity.
Lipophilicity measurements
The lipophilicity of the complexes were measured by standard shake flask technique.18 A chloroform solution of each of the complexes (1 mg/mL) was separately prepared. This solution 40 µL was added to 6 mL of CHCl3 and divided into two parts. The absorbance of one 3 mL part was recorded to get A0. To another part, 3 mL of milli-Q water was added and stirred for 4 h. The CHCl3 layer was separated by separatory funnel and the solution was centrifuged at 16000 rpm for 15 min and water particles were removed. Absorbance of the CHCl3 solution was recorded to obtain A1. A0 is ACHCl3 and A0 – A1 gives AH2O. Since the complexes were not soluble in water, the partition coefficient in CHCl3-water system, i. e., log PCHCl3 was measured and then log Poct was calculated using the following regression equation18, log Poct = (1.343 + log PCHCl3)/1.126.
DNA binding and cleavage
The concentration of CT DNA was determined from the absorption intensity19, assuming an ε value of 6600 M-1 cm –1 at 260 nm. Binding of the complexes were studied by ethidium bromide (EtBr) displacement method by monitoring the fluorescence change of CT DNA-bound-EtBr in Tris-HCl (5 mM)/NaCl (5mM). Excitation wavelength was set to 510 nm and emission was monitored at 600 nm.
Binding constant of complex (3) with CT DNA was calculated using the reported procedure.20 Supercoiled pUC19 DNA (5 µL, ∼500 ng) in 50 mM Tris-HCl buffer (pH 7.2) was treated with the metal complex (5 µL of respective concentration). The total volume
of solution loaded on gel was 10 µL. After incubation for 1 h at 37 °C, the samples were added to the loading buffer containing 25 % bromophenol blue, 0.25 % xylene cyanol, and 30 % glycerol (3 µL), and the solution was finally loaded on 0.8 % agarose gel containing 1.0 µg/mL EtBr. Electrophoresis was carried out for 2 h at 100 V in TBE buffer. Bands were visualized by UV light and photographed using Kodak Gel Documentation system.
Results and Discussion
The reactions of CuCl with thiopyrimidine (tp) and PPh3 in the ratio of 1:1:2 gives complex (1) in which tp coordinates in monodentate fashion through S (Scheme 1). The complex obtained with CuCl in the presence of NaH or pyridine by deprotonating tp also resulted in the same complex. This suggests that the tp is more stable in its protonated form in the complex. The chloride ion coordinated to the copper center is involved in strong hydrogen bonding with the pyrimidine N-H. The reaction of Cu(CH3CN)4ClO4
with PPh3 and thiopyrimidine in 1:2:1 ratio in acetonitrile gives complex (2), which was found to be coordinated to tp through both N and S with a ClO4-
ion sitting outside the coordination sphere (Scheme 2).
Figure 1 shows the ORTEP view of complexes (1), (2) and (3) (50 % probability thermal ellipsoids).
Hydrogen atoms were omitted for clarity. The bond distances (angstrom) of complexes (1), (2) and (3) are given in Table 1 and the corresponding bond angles (degrees) are given in Table 2. The crystallographic data for complexes (1), (2) and (3) are given in Table 3.
Thus, copper (I) interacts with thiopyrimidine in different ways and might be responsible for bringing about anticancer activity in a unique way.
Lipophilicity is an important physicochemical parameter that contributes to the toxicity and effectiveness of a drug. The in vivo distribution of a drug involves partitioning between the extracellular aqueous medium and the cell membrane mostly made up of the lipid molecules. Hence, lipophilicity of a drug directly correlates with its affinity for the cell membrane. Since the best mimic of a cell membrane is n-octanol, according to Hansch and Fujita convention, lipophilicity of a drug is expressed as the logarithm of its octanol-water partition coefficient21,22, log Poctanol = log (Coctanol/ Cwater), where Coctanol is concentration of drug in octanol and Cwater is concentration of drug in water. Lipophilicity of these complexes was studied by shake flask technique.23 Since complexes (1) and (2) were not soluble in
Fig. 1Molecular structure of complexes (1), (2) and (3). [50 % probability thermal ellipsoid].
water, the partition coefficient was first measured in CHCl3 and log Poctanol was calculated indirectly as described in the experimental section. The concentrations
of the complexes were determined by UV-visible absorption spectroscopy by monitoring the absorbance at 260 nm. The absorption of complex (1) and (3) was monitored at 260 nm and for complex (2) at 290 nm. The lipophilicities of the complexes were found to be 3.02, 2.76 and 0.75 for the complexes (1), (2) and (3) respectively. The high toxicity shown by complexes (1) and (2) in vivo could be attributed to the high lipophilicity.
Cytotoxicity of the copper(I) complexes
Thio derivatives of DNA bases are known to induce cellular sensitization to ultraviolet A (UVA) and combination of non lethal doses of UVA and thiobases show cooperative cytotoxicity to cultured human cells.24 Thiopurines have been used as antileukemic agents.24 Niacin deficiency is also known to impair cell cycle arrest and DNA damage induced apoptosis, leading to the survival of cells with leukemogenic potential.9 Previously Ru(II) complexes of thiopurines and thiopyrimidines have shown potent activity against ovarian cancer cells.25 Hence, it was of interest to evaluate the anticancer activities of the complexes by studying the potential inhibitory effects of the complexes on the growth of cancer cells in various cancer cell lines.
All three complexes showed consistent in vitro cytotoxicity in human malignant cell lines. The complexes were found to inhibit more than 80 % growth of the cells at 20 µg /mL concentration after 48 h similar
Table 3Crystallographic data for complexes (1), (2) and (3)
(1) (2) (3)
Empirical formula C40H34ClCuN2P2S C40H34ClCuN2O4P2S C36H33ClCuN6O4P
Formula weight 735.68 799.68 791.66
Crystal system Monoclinic Monoclinic Tetragonal
Space group P21/c P21/n R -3
a (Å) 14.427(5) 10.0774(15) 15.28(4)
b (Å) 10.127(3) 15.400(2) 15.28(4)
c (Å) 24.357(8) 25.138(4) 37.78(12)
α (deg.) 90 90.000(3) 90.00
β (deg.) 94.491(5) 78.655(2) 90.00
γ (deg.) 90 90.000(3) 120.00
Volume (Å3) 3547.7(19) 3825.0(10) 7639(37)
Z 4 4 9
Density (calc.) (mg/m3) 1.377 1.389 1.549
Crystal size (mm3) 0.12 × 0.13 × 0.1 0.22 × 0.07 × 0.16 0.21 × 0.16 × 0.26
Reflections collected 27478 32759 1663
Independent reflections 7231 [R(int) = 0.0644] 9004 [R(int) = 0.0766] [R(int) = 0.1436]
(Goodness-of-fit on F2)c 1.113 1.101 1.306
Final R1a , wR2b [I > 2σ(I)] R1 = 0.0635, wR2 = 0.1245 R1 = 0.0880, wR2 = 0.1775 R1 = 0.0675, wR2 = 0.1589 Final R1a, wR2b (all data) R1 = 0.0953, wR2 = 0.1366 R1 = 0.1456, wR2 = 0.2026 R1 = 0.1026, wR2 = 0.1762
aR1 = (Σ||Fo| - |Fc||) / (Σ|Fo|); bwR2 = [Σ(w|Fo|2-|Fc|2)2 / Σw|Fo|2)2]1/2; cGOF = [w(Fo2-Fc2) 2]/(n-p)1/2. Table 1Bond distances for complexes (1), (2) and (3)
Bond Bond distance (Å)
(1) (2) (3)
Cu(1)-P(1) 2.2848(11) 2.2137(13) 2.186(3) Cu(1)-P(2) 2.2947(12) 2.2607(13)
Cu(1)-Cl(1) 2.3633(12) Cu(1)-S(1) 2.3800(13)
Cu(1)-N(1) 2.121(4) 2.105(8)
Cu(1)-S(2) 2.4913(15)
Table 2Bond angles for complexes (1), (2) and (3)
Angles Bond angle (o)
(1) (2) (3)
P(1)-Cu(1)-P(2) 122.41(4) 128.95(5) P(1)-Cu(1)-Cl(1) 111.81(4)
P(2)-Cu(1)-Cl(1) 98.65(4) P(1)-Cu(1)-S(1) 102.18(5) P(2)-Cu(1)-S(1) 112.87(4) Cl(1)-Cu(1)-S(1) 108.69(4)
N(1)-Cu(1)-P(1) 126.28(10) 118.87(5)
N(1)-Cu(1)-P(2) 99.72(10)
N(1)-Cu(1)-S(2) 68.21(11)
P(1)-Cu(1)-S(2) 112.58(5)
P(2)-Cu(1)-S(2) 103.58(5)
N(1)-Cu(1)-N(1) 98.64(7)
to cisplatin, except in K562 cell line in which these complexes show better inhibition than cisplatin. The IC50
values of the complexes are given in Table 4 which shows that complexes (1) and (2) are more potent in vitro in the cell lines Colo205, H226, HOP62 and K562 than complex (3), whereas the activity is similar for MCF7 and T24. Complexes (1) and (2) are found to be equivalent to or more active than cisplatin in all the cell lines. Previously Ru(II) complexes of thiopyrimidines and thiopurines have been shown to have an IC50 of 18 µM in ovarian cancer cell lines25, whereas in this study in a different set of cell lines we have observed IC50 values of ~ 5-7 µM for Cu(I) complexes of thiopyrimidines (1) and (2).
Since these three complexes showed consistent activity in a range of cell lines their in vivo activity was studied. Unlike the in vitro behavior, the in vivo activity of the complexes was different. The in vivo results and % ILS of the complexes are given in Table 5. These results show that complex (3) is active and non-toxic in spontaneous breast carcinoma and L1210 whereas complexes (1) and (2) are active and non-toxic in P388 leukemia. However, complexes (1) and (2) were very toxic in Lewis lung carcinoma and
P388 when the vehicle was 10 % DMSO, and so further experimentation was not carried out. On the other hand, with carboxymethyl cellulose (CMC) as a vehicle, these two complexes were active and non-toxic, although the
% ILS was only 50 % and 37.5 % for complexes (1) and (2) respectively at 75 mg/kg concentration. Unlike complexes (1) and (2), complex (3) was non-toxic in both spontaneous breast carcinoma model and murine leukemia L1210. Complex (3) did not show any activity in L1210 but showed activity at 50 mg/kg on the 35th day with 55 % ILS in spontaneous breast carcinoma upon intravenous injection. Higher toxicity of complexes (1) and (2) could be partly due to the observed higher log Poct value of more than 2.6. As we have discussed before the non-toxicity of complex (3) could be ascribed to its higher water solubility having a log Poct value of 0.75.
Metal complex-DNA interactions
Since it is reported that the key step in the mechanism of the anticancer activity of thiopurine based drugs is the incorporation of thiobases into the nucleic acids,24,26, 27 we have studied the DNA binding and DNA cleavage activity of the complexes. Intercalation of EtBr in CT DNA enhances its fluorescence intensity as the fluorescence quenching by solvent molecules is prevented. Binding of small molecules to CT DNA can be conveniently studied by monitoring the fluorescence of EtBr, since it is displaced from DNA and the fluorescence intensity is reduced. A plot of fluorescence intensity versus concentration of drug/small molecule permits evaluation of the apparent binding constant. The apparent binding constant of complex (3) with calf thymus DNA is 5.9 × 105 M. Complexes (1) and (2) also reduce the EtBr fluorescence intensity but Kapp could not be obtained as
Table 5The % ILS of the complexes in different mouse models
Cell lines Dose/route Schedule Vehicle % ILSa/ % tumor growth inhibitionb Comments
(1)a (2)a (3)b
P388 100mg/kg IP 1 – 9 10 % DMSO 10 00 NDc Inactive & toxic
L 1210 75 mg/kg IP
50 mg/kg IP 25 mg/kg IP
1, 5, 9 1, 5, 9 1, 5, 9
CMC
00 14.3
7.1
14.3 7.1 7.1
NDc
Both inactive
P388 75 mg/kg IP
25 mg/kg IP
1, 5, 9 1, 5, 9
10 % DMSO 23.5
5.8 17.6 5.8
NDc Both inactive
P388 75 mg/kg IP
50 mg/kg IP 25 mg/kg IP
1, 5, 9 1, 5, 9 1, 5, 9
CMC CMC CMC
37.5 25.0 25.0
50 37.5 25.0
NDc 1 and 2 active and non toxic at all these doses.
Spontaneous breast tumor 50 mg/kg, IV 1, 5, 9 Water - - 55.0 Active on day 35 Lewis lung carcinoma 100 mg/kg. IP
100 mg/kg IP 1 – 9
10 % DMSO Toxic Toxic NDc Expt. terminated
a % increase in lifespan (% ILS); b % tumor growth inhibition; c Not done.
Table 4In vitro IC50 values for the complexes for the different cell lines
Complex IC50 (µg/mL)a
Colo205 H226 HOP62 K562 MCF7 T24
(1) 5 5 7 10 5 5
(2) 5 5 5 10 5 5
(3) 15 16 18 18 16 8
Cisplatin 9 7 8 20 7 5
aThe IC50 values are derived from graphs plotted using the average
% inhibition values obtained from 3 different sets of experiments.
50 % reduction in fluorescence intensity of EtBr-DNA complex was not achieved. Due to the high hydrophobicity of the complexes a precipitate was formed on further addition of the complexes to the EtBr-DNA solutions. However, before the onset of precipitate formation, the EtBr fluorescence was decreased to 13.2 % and 22.5 % after initial addition of 1200 µM of complexes (1) and (2) respectively. The fluorescence quenching of the EtBr-DNA solution on addition of these complexes is shown in Fig. 2.
Interestingly, complexes (1) and (2) are weakly fluorescent when excited at 250 nm and emit at 364 nm in 5 mM tris-HCl and 5 mM NaCl. As the emission spectra are not affected by the addition of CTDNA, this emission could not be used for binding studies.
Since a previous study12 showed that the Cu(I) DPPE complexes cleave DNA in vivo, SC DNA cleavage study was carried out. Figure 3 shows the cleavage of SC DNA by the complexes (1), (2) and (3).
Complex (3) shows complete cleavage at 400 µM but in the case of complexes (1) and (2) no cleavage was observed even at 1200 µM as shown in Fig. 3.
However, these two complexes showed retardation in movement, which might be due to the reduction in the super helical density of the complex bound DNA.
Potent anticancer drugs, adriamycin and cisplatin, behave in a similar fashion and do not cleave SC DNA.28 It has also been observed that adriamycin and actinomycin D result in elongation of DNA and reduction in DNA synthesis respectively.29,30 So, in vitro DNA cleavage does not always translate to cytotoxicity. Complexes (1) and (2) show potent in vitro cytotoxicity at < 10 µg/mL concentration, whereas complex (3) shows cytotoxicity only at < 20 µg/mL dose level. These complexes probably follow a different pathway for the observed antitumor activity.
Interaction with biologically relevant molecules
While elucidation of the mechanism of action of the metallodrugs, it is also crucial to ascertain the active species in vivo, after intravenous injection of the metal complexes. Especially since copper complexes readily undergo ligand substitution and redox reactions, it is difficult to pinpoint the nature of the active species [Cu(I) or Cu(II)] with certainty. Once intravenously administered, metallodrugs can react with human serum albumin (HSA) and intracellularly abundant thiols like glutathiones and undergo redox reactions.
Hence, in order to understand the structural integrity under physiological conditions, it is important to study the binding of the metal complexes with plasma
proteins, such as human serum albumin (HSA) and intracellularly abundant thiols like glutathiones. The interaction of the most active complex (3) with bovine serum albumin (BSA), reduced glutathione (GSH), oxidized glutathione (GSSG) and methionine (meth) was studied by 31P NMR.
In the case of complex (3), there was no ligand exchange reaction even in the presence of 5 equivalents of GSH as observed from the unchanged peak position of PPh3 in the 31P NMR spectrum. As a control experiment, similar reactivity studies were done with the complex [Cu(PPh3)4]ClO4. It was observed that
31P NMR value was completely shifted to –5 ppm (free PPh3 value) from –0.7 ppm (complex), and most
Fig. 2Fluorescence quenching of EtBr-DNA solution upon addition of different concentrations of the complexes in 5 mM Tris-HCl and 5 mM NaCl. [■, (1); ●, (2); ▲, (3). λex = 510 nm; λem = 600 nm].
Fig. 3Gel electrophoresis diagrams for the cleavage of supercoiled (SC) pUC19 DNA with (A) complexes (1) and (2) and (B) complex (3). [Cleavage of supercoiled pUC19 DNA (0.5 µg) by 400 µM of complex (3) in a 5 mM tris HCl/NaCl buffer (pH 7.2) at 37 °C containing 1 % DMF. Forms I and II are SC and nicked circular (NC) DNA. 1200 µM of complexes (1) and (2) were used].
importantly, the intensity of the peak at –0.7 ppm was completely diminished while the intensity of the peak at –5 ppm (free PPh3) and +30 ppm (phosphine oxide) increased with increasing concentrations of GSSG and GSH. These results point to the fact that complex (3) retains its structural integrity in the presence of HSA under physiological conditions, and suggests that the activity of complex (3) could be due to the Cu(I) species.
Conclusions
In the present work, we have shown potent anticancer activities of thionucleobase and nicotinamide containing copper(I) phosphine complexes. By using the differences in the fluorescence intensities of bound and free EtBr as a probe, we have monitored DNA binding.
In the presence of these complexes, there was decrease in the fluorescence intensity showing release of EtBr from the DNA-EtBr complex. The % quenching of EtBr-DNA fluorescence was highest about 55 % for complex (3) whereas the quenching is only 13 % and 22 % for complex (1) and (2) respectively. So complex (3) binds to CT DNA more efficiently than complexes (1) and (2). Similarly in an in vitro DNA cleavage studies with SC DNA we have also showed that complex (3) cleaves about 90 % of SC DNA at 0.4 mM concentration whereas complex (1) and (2) showed only retardation in the mobility of bound SC-DNA. Since complexes (1) and (2) showed potent in vitro cytotoxicity but they do not exhibit DNA binding and DNA cleavage activity, the results indicate that complexes (1) and (2) exert cytotoxicity by some other mechanism which may not involve direct cleavage of DNA. However, their in vivo activity was not encouraging owing to the toxicity observed. Hence, our experimental results suggest that complex (3) holds forth promise as a drug candidate, with potent in vitro and in vivo antitumor activity, having optimum lipophilicity, leading to reduced toxicity.
Supplementary Data
The X ray crystallographic files (in CIF format) for the structure determination of the complexes have been deposited with the Cambridge Crystallographic Data Centre under CCDC 767500 (complex 1), CCDC 767499 (complex 2), CCDC 767501 (complex 3). These can be obtained free of charge from the CCDC via www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgement
Generous financial support from DBT, New Delhi, India, is gratefully acknowledged.
References
1 Marzilli L G, Ano S O, Intini F P & Natile G, J Am Chem Soc, 121 (1999) 9133.
2 Simon T M, Kunishima D H, Vilbert G J & Lorber A, Cancer Res, 41 (1981) 94.
3 Adwankar M K, Wycliff C & Samuelson A G, Indian J Exp Biol, 35 (1997) 810.
4 Berners-Price S J, Girard G R, Hill D T, Sutton B M, Jarrett P S, Faucette L F, Johnson R K, Mirabelli C K & Sadler P J, J Med Chem, 33 (1990) 1386.
5 Snyder R M, Mirabelli C K, Johnson R K, Sung C M, Faucette L F, McCabe F L, Zimmerman J P, Whitman M, Hempel J C
& Crooke S T, Cancer Res, 46 (1986) 5054.
6 Colacio E, Romerosa A, Ruiz J, Roman P, Gutierrez-Zorilla J M, Vegas A & Martinez- Ripoll M, Inorg Chem, 1991 30 3743.
7 Cookson P D & Tiekink E R T, J Chem Soc Dalton Trans, 1993 251.
8 Stocco G, Gattuso F, Isab A A & Shaw III C F, Inorg Chim Acta, 1993 209 129.
9 Kirkland J B, Mol Cancer Ther, 8 (2009) 725.
10 1Nicola G, In vivo, 17 (2003) 169.
11 Skehan P, Storeng R, Scudiero D, Monks A, McMahon J, Vistica D, Warren J T, Bokesch H, Kenney S & Boyd M R, J Natl Cancer Inst, 82 (1990) 1107.
12 Sanghamitra N J M, Phatak P, Das S, Somasundaram K &
Samuelson A G, J Med Chem, 48 (2005) 977.
13 Hathaway, J Chem Soc Dalton Trans, (1961) 3215.
14 Keller R N & Wycoff H D, Inorg Synth, 50 (1946) 1.
15 WINGX, An Integrated System of Windows Program for the Solution Refinement and Analysis of Single Crystal X-Ray Diffraction, ver. 1.63.02, (Department of Chemistry, University of Glasgow, Glasgow).
16 SHELX-97, A Program for Crystal Structure Refinement, release 97-2, (University of Goettingen, Germany) 1997.
17 Skehan P, Storeng R, Scudiero D, Monks A, McMahon J, Vistica D, Warren J T, Bokesch H, Kenney S & Boyd M R, J Natl Cancer Inst, 82 (1990) 1107.
18 Leo A, Hansch C & Elkins D, Chem Rev, 71 (1971) 525.
19 Reichmann M E, Rice S A, Thomas C A & Doty P, J Am Chem Soc, 76 (1954) 3047.
20 Lee M, Rhodes A.L, Wyatt M.D, Forrow S & Hartly J H, Biochemistry, 32 (1995) 4237.
21 Hansch C & Fujita T, J Am Chem Soc, 86 (1964) 1616.
22 Hansch C, Maloney P P & Fujita T J, Nature, 194 (1962) 178.
23 Leo A, Hansch C & Elkins D, Chem Rev,71 (1971) 525.
24 Massey A, Xu Y Z & Karan P, Curr Biol, 11 (2001) 1142.
25 Cini R, Tamasi G, Defazio S, Corsini M, Zanello P, Messori L, Marcon G, Piccioli F & Orioli P, Inorg Chem, 42 (2003) 8038.
26 Hruza G, Health News, 8 (2002) 3.
27 Shigeta S, Mori S, Watanabe F, Takahashi K, Nagata T, Koike T S N & Wakayama M, Antivir Chem Chemother, 13 (2002) 67.
28 Reichmann M E, Rice S A, Thomas C A & Doty P, J Am Chem Soc, 76 (1954) 3047.
29 Bunte T, Novak U, Friedrich R & Moelling K, Biochim Biophys Acta, 610 (1980) 241.
30 Simpkins H & Pearlman L F, Biochim Biophys Acta, 783 (1984) 293.
