Synthesis, Characterization And Antitumor Activity Of Copper(II) Complexes, [CuL2] [HL1-3=N,N-Diethyl-N'-(R-Benzoyl)Thiourea (R=H, o-Cl and p-NO2)]

The copper (II) complexes (CuL2) were prepared by reaction of Cu(CH3COO)2 with the corresponding derivatives of acylthioureas in a Cu:HL molar ratio of 1:2. Acylthiourea ligands, N,N-diethyl-N'-(R-benzoyl) thiourea (HL1-3) [R=H, o-Cl and p-NO2] were synthesized in high yield (78-83%) and characterized by elemental analysis, infrared spectroscopy, 1H and 13C NMR spectroscopy. The complexes CuL2 were characterized by elemental analysis, IR, FAB(+)-MS, magnetic susceptibility measurements, EPR and cyclic voltammetry. The crystal structure of the complex Cu(L2)2 shows a nearly square-planar geometry with two deprotonated ligands (L) coordinated to CuII through the oxygen and sulfur atoms in a cis arrangement. The antitumor activity of the copper(II) complexes with acylthiourea ligands was evaluated in vitro against the mouse mammary adenocarcinoma TA3 cell line. These complexes exhibited much higher cytotoxic activity (IC50 values in the range of 3.9-6.9 μM) than their corresponding ligands (40-240 μM), which indicates that the coordination of the chelate ligands around the CuII enhances the antitumor activity and, furthermore, this result confirmed that the participation of the nitro and chloro substituent groups in the complex activities is slightly relevant. The high accumulation of the complexes Cu(L2)2 and Cu(L3)2 in TA3 tumor cells and the much faster binding to cellular DNA than Cu(L1)2 are consistent with the in vitro cytotoxic activities found for these copper complexes.


INTRODUCTION
Whilst most of the investigations in the treatment of cancer diseases are oriented to synthesize new metal complexes analogous to cis-diamminedichloroplatinum(II) (cisplatin) (antineoplastic agent of clinical use)/1, 2/, there is a growing number of non-platinum metal complexes which also exhibit remarkable anticancer activities/3-5/. Thus, the bis(acetate) bis(imidazole)copper(II) complex has shown a high cytotoxic activity against to the mouse B16 melanoma cancer cell line, and determinations realized with d X174 RF DNA indicate that the target of this complex may be the guanine residues of the DNA helix/6/. In addition, the copper (II) complexes with thiosemicarbazone derivatives (O,N,S-tridentate chelate ligands) are well-known by their biological applications as antiviral/7/, antimicrobial/8/, antitubercular and antitumor agents/9-11/.
In this sense, Antholine et  In our previous work, we reported the in vitro antitumor activity of the platinum (II) complexes with analogousligands against mouse mammary adenocarcinoma TA3 cells, where these complexes showed to be more cytotoxic at low micromolar concentrations than their corresponding ligands/23/. Since our earlier work had revealed that platinum complexation with acylthiourea ligands enhances the antitumor activity, we were motivated to use copper as different central atom in order to obtain a cytotoxic behaviour similar to the platinum complexes mentioned before. In the present work, we report about synthesis and characterization of copper(lI) complexes with ligands N,N-diethyl-N'-R-benzoylthiourea [R=H, o-C1 and p-NO2] and their in vitro cytotoxic effect, copper accumulation and binding to cellular DNA on mouse TA3 mammary adenocarcinoma.

Materials
Copper(II)acetate monohydrate, sodium sulphate, sodium bicarbonate, benzoylchloride, ochlorobenzoylchloride, p-nitrobenzoylchloride and diethylamine from Aldrich were used as received. All solvents were reagent grade and were purified by standard procedures before use.
Eagle's Minimum Essential Medium [MEM(E)] was supplied by Sigma (USA) and the fetal bovine serum [FBS] was provided by Difco (Detroit, MI). The penicillin, streptomycin and sodium chloride 0.9% were obtained from Sanderson's laboratory (Chile).

Measurements
Elemental analyses were carried out on a Fisons-Carlo Erba 1108 elemental microanalyser. Melting points were determined on a Boetius melting-point apparatus. Magnetic susceptibilities at room temperature (296 K) were measured using the Gouy method on a Johnson-Mathey MSB-MKI balance with HgCo(NCS)4 as standard and diamagnetic corrections were made using Pascal's constants/24/. The infrared (IR) spectra were recorded in solid state (KBr pellets) on a Bruker FT-IR IFS 55 Equinox spectrophotometer in the range 4000 400 cm . FAB(+) mass spectra were obtained on a ZAB-HSQ (V.G. Analytical Ltd.) spectrometer.
Determinations of copper concentration in biological samples were performed on a Perkin Elmer 3110 atomic absorption spectrometer. NMR (H and 3C) spectra of the ligands were recorded on a Bruker Advance DRX-300, using CDCI3 as solvent and TMS as internal standard. Cyclic voltammetric measurements were made in DMSO solvent on a Bioanalytical System BAS CV-50 W with an X-Y recorder using a carbon disk (6 mm diameter) as working electrode vs. SCE as reference and a Pt filament as auxiliary electrode. The solutions of the complexes (1 x 10 .3 M ) in DMSO with tetrabutylammonium perchlorate (TBAP) (0.1 M) as supporting electrolyte were deoxygenated by a stream of dry nitrogen for at least 10 min. EPR spectra of the copper complexes solutions in CH2CI at room temperature were measured using a Brukcr ESP 300 E spectrometer operating at X-band t'requency.

Synthesis of the copper(ll) complexes
The complexes were prepared according to the following general procedure.
To a stirred solution of N,N-diethyl-N'-(R-benzoyl)thiourea (1 mmol) in methanol (30 mL) was added dropwise a solution of Cu(CH3COO)2 HO (0.5 mmol) in methanol (30 mL). The reaction mixture was refluxed for 1 h and then stirred at room temperature for 24 h. The green solution was filtrated and the filtrate was evaporated to dryness under reduced pressure to yield a dark green solid. The solid was washed several times with small portions of cold ethanol and dried in vacuo.
Synthesis, Characterization And Antitumor Activity of Copper(lI) Complexes

Crystal structure determination
The crystallographic data of Cu(L2)2 were obtained on a Siemens CCD Smart Diffractometer (Mo Kc radiation, =0.71073 A, graphite monochromator, T 218(2) K). The intensities were corrected for Lorentz and polarization effects and for absorption using SADABS. The structure was solved by direct methods, which revealed the positions of all non-hydrogen atoms and refined on F z by a full-matrix least-squares procedure using anisotropic displacement parameters. The hydrogen atoms were located from difference Fourier syntheses and refined isotropically. All calculations were carried out using the SHELXS-97 and SHELXL-97 programs/27-28/. Crystal data collection and refinement details for the complex Cu(L2)2 are summarized in Table 1. Table 1 Crystal data and refinement summary for the copper(II) complex Cu(L2)z. The mouse mammary adenocarcinoma TA3 cell line was obtained from ascites fluid of young adult male CAF Jax mice. The cells were cultured at 37 C in a growth medium consisting of Eagle's Minimum Essential Medium [MEM(E)] with Earle's Salts, L-glutamine and 25 mM HEPES, supplemented with 10% fetal bovine serum [FBS], 44 mM sodium bicarbonate, 100 U mL "1 penicillin and 100 lag mL "1 streptomycin.
For these experiments, cells at a concentration of 2 x 105 mL 1 were seeded by 96 h in 20 mL of culture medium/29/. After 24 h of tumor cell growth, the compounds were dissolved in DMSO just before the experiments and a calculated amount of drug solution was added to each growth medium to obtain the desired concentrations of either the ligands ( 20, 40 and 60 laM)or their copper complexes (1, 5 y 10 laM).
The final DMSO concentration in the culture medium, 0.5%, did not show an appreciable effect on cell growth for these assays. Parallel cultures were used as control. Viable cells were determined using a Neubauer counting chamber every 24 h. The concentrations of every compound were plotted with respect to the percentage of cell survival. The IC.s0 values were obtained by graphical interpolation at 48 h of exposure on each one of the compounds. These values represent the drug concentration (laM) required to inhibit cell growth by 50%. All assays were performed in triplicate cultures and repeated three times in independent pattern.
2.6.2. Cellular copper accumulation Cells (5 x 106 mL) were incubated at 37 C with each one of the copper complexes (125 laM). After various time intervals, an aliquot of the mixture was removed and assayed according to the reported method /30/. Briefly, the cells were pelleted by centrifugation and washed three times with ice-cold PBS (0.1 M phosphate buffer in 0.15 M NaC1, pH 7.4). Pellets were dried and digested overnight in concentrated nitric acid, then diluted with distilled water to give a final HNO3 concentration of 20%. Cellular copper concentrations were measured by atomic absorption spectroscopy and the amount of accumulate0 Cu was expressed as nmol Cu/10 6 cells.

Binding to cellular DNA in TA3
Each one of the copper complexes (20-100 tM) was incubated with the TA3 tumor cells (5 x 106/mL ) at 37 C for 2 h of exposure. Then, cells were isolated by centrifugation, suspended in mL of PBS and centrifuged again. This washing step was repeated twice. After centrifugation cells were taken up in 500 tL of TEN buffer, pH 8 (10 mM Tris, 10 mM EDTA and 150 mM NaCI), and subsequently 5 laL of proteinase K and 50 laL of 10% SDS were added. Lysed cells were kept at 55 C for h. Proteins were then removed by chloroform/phenol (1:1) extraction followed by an extraction with chloroform alone. DNA was precipitated from the aqueous layer by adding an equal volume of isopropyl alcohol. Finally, precipitated DNA was removed from the solution, washed with 70% ethanol and dissolved in 1 mL of water/31/.
The DNA concentration was determined by measuring the UV absorption at 260 nm and concentration of base pairs was calculated using mean molar extinction coefficient per base pair 260 16800 M 1 cm1. instead of acetonitrile and the intermediary product, isothiocyanate, was not removed from the reaction mixture. With these changes it was possible to obtain a high yield, which was compared with those obtained for related acylthioureas described in the literature/32, 33/. The ligands I-IL and HL were recrystallized from mixed solvents of CH2CI2-CH3OH (2:1, v/v) and CHC13-EtOH (2:1, v/v), respectively, whereas HL 3 was recrystallized from CH2C12. The ligands were obtained in satisfactory yield (78-83 %) and characterized by elemental analysis, infrared spectroscopy and 1H and 13C NMR spectroscopy.
The copper (II) complexes Cu(L)2, Cu(L2)2 and Cu(La)2 were prepared by refluxing the methanolic mixture of the ligand (HL) and Cu(CH3COO)2-H20 in the molar ratio of 2:1. With this employed technique, the yields of copper(II) complexes were improved as related to other methods that carry out the synthesis during several days at room temperature/34/. Recrystallization of the complex Cu(L)2 from hot ethanol yielded crystals suitable for structural determination by X-ray diffraction. All three copper (II) complexes were characterized by elemental analysis and IR, FAB(+)-mass, magnetic susceptibility, EPR and cyclic voltammetry. The elemental and spectroscopic analysis of the ligands and their copper (II) complexes are consistent with the proposed structures given in Scheme 1. In general, the methods used for the preparation of these compounds were adopted because they are fast, easy and give higher yields than other methods, which contain some modifications in the synthetic route, such as, temperature, reaction time, solvents and separation procedures of the product.

IR and H-NMR spectra
In the IR spectra, al'l acylthiourea ligands exhibited NH stretching bands in the range of 3200-3284 cm1, which disappeared after coordination. This indicates the loss of the proton originally bonded to nitrogen atom of the (NH-CO) amide group. The vibrational frequencies due to the carbonyl (1648-1690 cm) and thiocarbonyl (1230-1233 cm 1 ) groups in the free ligands are shifted (65-68 and 21-25 cm1, respectively) towards lower frequencies upon complexation, confirming that the deprotonated ligands are coordinated to Cu u ion through the oxygen and sulfur donor atoms/35-37/.
In the H-NMR spectra for the ligands, the presence of the chloro-and nitrosubstituent groups on the benzoyl moiety (HL z and HL3), causes no significant changes in the chemical shifts of the N-H group (8.32 and 8.41 ppm, respectively) relative to the ligand I-IL (8.35 ppm) with unsubstituted benzoyl moiety. The aromatic protons signals (ortho and meta ) are shifted upfield (0.49 and 0.24 ppm, respectively) for I-IL 2 and downfield (0.17 and 0.86 ppm, respectively) for HL 3 with respect to HL . These results are consistent with the mesomeric/inductive effects expected for these substituents. On the other hand, the resonances of the methylene protons bonded to the nitrogen of the respective thioamide group appeared as two separated signals at 3.61-3.69 and 3.97-4.04 ppm, respectively, showing in each signal a well-resolved quartet. The magnetic inequivalence of the CH2 protons can be attributed to the restricted rotation around the C-N bond between the thiocarbonyl group and the amine nitrogen due to the partial double character of this bond /38, 39/.

Magnetic data and EPR spectra
The room temperature magnetic moments for the copper (II) complexes Cu(L)2, Cu(L2)z and Cu(L3)2 resulted to be 2.0, 1.93 and 1.82 BM, respectively. These values are consistent with the presence of one unpaired electron in mononuclear d 9 copper (II) complexes. Furthermore, these magnetic data .are ,in agreement with those of the cis-bis(acylthioureato) copper (II) complexes (tci-f= 1.8-2.00 BM) reported in the literature/32, 40/.

Electrochemical studies
The redox processes of the copper (II) complexes were electrochemically confirmed by cyclic voltammetry. The voltammograms of all free ligands do not present any oxidation or reduction peak in the potential range studied, except for the ligand HL3, which show a reduction peak at-1.05 V, which may correspond to the reduction of the p-nitrophenyl group. Cyclic voltammograms of the copper (II) complexes in DMSO show a quasi-reversible voltammetric response for the Cu(II)/Cu (1) (Figure 2a), respectively. The ratio of anodic (ipa) and cathodic (/pc) peak currents is in the range of 0.83-0.91 at 100 mVs scan rate. The AEp values for these complexes were lower than those of related copper complexes (AEp >200 mV in DMF) considered in the literature as irreversible processes /34/. Furthermore, as we can observe in Figure 2b, the cyclic voltammogram for the complex Cu(L3)2 shows that the metal-based potential peaks change with the scan rate (50-500 mV s) indicating that the redox process becomes more irreversible as reported for thc related copper complexes too/40/.
The presence of the nitro and chloro substituent groups on the bcnzoyl moiety has a weak influence on the electrochemical properties of these copper (II) complexes /32/. Only small differences found in the Complexes cathodic potentials indicate that Cu(L)2 must be more difficult to reduce than Cu(L2) and Cu(L3)z with the chloro and nitro substituent groups on the benzoyl moiety.

Structural data
The molecular structure of the complex Cu(L), together with the atom numbering scheme adopted, is shown in Fig. 1. Selected bond distances and bond angles are listed in Table 2. As can be seen from Fig. 1, the structure shows that the copper(II) ion has a slightly distorted squareplanar geometry indicated by the angle between the planes CulS101 and CulS202 of 12.1 (1) .T he structural determination of this complex confirms that two deprotonated ligands are coordinated bidentately to the copper(II) ion through the sulphur and oxygen donor atoms in a cis arrangement /44,45/. The coordination around the Cu ion generates two six-membered chelate rings, which form angles with the adjacent phenyl groups of 48.2(1) and 43.0(1) .T his result probably indicates that the presence of the chloro substituent on the benzoyl moiety leads to a steric repulsion between chelate rings and phenyl rings. A) are shorter compared to I-IL (N1-C1 1.428(2) and N1-C2 1.360(2) A). This indicates that the decrease of the bond orders of the thiocarbonyl and carbonyl groups together with the changes in the N-C bond lengths on coordination, originates from a n-electron delocalization over the chelate ring of the complex Cu(L)z /33,46/. Table 2 Selected bond lengths (A) and angles (o) for Cu(L2)e Bond Lengths Cul-S1 Cul-O1 S1-C1 Cul-S1-C1

129.7(2)
Cul-S2 Cul-O2  (2) 92.99 (4) 105.38 (6) 124.3(2) 126.5 (1) 130.1 (2) 3.6. Antitumor evaluation The acylthiourea ligands and the complexes Cu(L)2, Cu(L2)z and Cu(L3h were tested in vitro for their cytotoxic activity against mouse mammary adenocarcinoma TA3 cell line. Figures 3 and 4 show the concentration-dependent inhibitory effect of ligands and their respective copper (II) complexes on the percentage of cell survival at 48 h of exposure on culture medium. In Figure 3, we can observe that the ligand HL a showed a major cytotoxic activity (---40 tM) than ligands HL and HL (> 160 laM) in the survival range of 50-60 %. This high cytotoxicity that HL 3 presents could be due to the biotransformation of the nitrophenyl group to its nitrophenyl radical anion (PhNOi') catalyzed by the enzyme NADPHcytochrome P450 reductase, together with the formation of the superoxide radical anion (O2")/47-49/. As we can see in Figure  4, the copper (II) complexes turned out to be highly more cytotoxic than their ligands, at concentration range IC50 values for the tested compounds are given in tM) evaluated in the same tumor cell line. These results demonstrate that the bis-chelate copper complexes, with a slightly distorted square-planar geometry play an important role in the inhibition of TA3 tumor cell growth.  3.7. Accumulation of copper in TA3 cells and binding to cellular DNA As we can see from Figure 5, the complexes Cu(L2) and Cu(L3)2 enter into the cells more rapidly than Cu(LI)z. After 2.5 h of exposure, copper accumulated in TA3 cells from complexes Cu(LI), Cu(L) and Cu(L3)z were 7.12, 12.1 and 14.5 nmol Cu/106 cells, respectively. Significantly, 2.3 times higher Cu(L3) than Cu(L) was accumulated in TA3 cells, whereas Cu(L3)z cellular accumulation was only slightly higher than Cu(LZ)z. Complexes Cu(L)2 and Cu(L3) are incorporated in TA3 cells at similar concentrations than platinum(ll) complex, PtNH3Cb(L) [L=2-Phenyipyridine] (100 tM, 3 h of exposure) in mouse sarcoma 180 cells (13.5 nmol Pt/106 cells)/52/. These results indicate that the hydrophobicity of these copper complexes together with the presence of the nitro and chloro substituent groups in the benzoyl moiety seems to expedite the transport of these complexes through the cellular membrane/52, 53/. Figure 6 shows the number of drug molecules bound per base pair (rb) based on different concentrations of the copper complexes for 2 h of exposure. The complexes Cu(LZ)z and Cu(L3) displayed a higher DNAbinding activity than Cu(LI). At the concentration range of 20-100 laM, the complex Cu(L3): binds 2 times more quickly to DNA than Cu(LI)z. These results are in accordance with the cytotoxic activities in vitro found for these complexes and the differences of binding to DNA that are presented by these copper(II) complexes may be caused by their lipophylic properties, incorporation rate and their square-planar geometry

CONCLUSIONS
In summary, we have prepared the bis-chelate copper (II) complexes from acylthiourea derivatives. The crystal structure of Cu(LZ)2 shows that the copper atom presents a nearly square-planar geometry with two bidentate ligands having sulfur and oxygen as donor atoms in cis positions.
In vitro cytotoxic activity tests against the mouse mammary adenocarcinoma TA3 cell line showed that the copper (II) complexes were more cytotoxic at low micromolar concentrations with respect to the free ligands. In particular the complex Cu(L3h was able to induce a notable decrease of cell survival with a IC.s(i value very similar to those of the related platinum (II) complexes tested under the same experimental conditions. In addition, the complexes Cu(L2)2 and Cu(L3)2 have shown remarkable copper accumulation in TA3 cells and higher binding to cellular DNA than Cu(L1)2. We hope the present results may be complemented with other additional biological tests involving the assessment of their cytotoxic effects on different tumor cell lines, their toxicity in vivo and the biodistribution of these compounds in diverse organs with the purpose to lead to the development of a new class of antitumor agents.