Synthesis, Structural, and Cytotoxic Properties of New Water-Soluble Copper(II) Complexes Based on 2,9-Dimethyl-1,10-Phenanthroline and Their One Derivative Containing 1,3,5-Triaza-7-Phosphaadamantane-7-Oxide

A series of water-soluble copper(II) complexes based on 2,9-dimethyl-1,10-phenanthroline (dmphen) and mixed-ligands, containing PTA=O (1,3,5-triaza-7-phosphaadamantane-7-oxide) have been synthesized and fully characterized. Two types of complexes have been obtained, monocationic [Cu(NO3)(O-PTA=O)(dmphen)][PF6] (1), [Cu(Cl)(dmphen)2][PF6] (2), and neutral [Cu(NO3)2(dmphen)] (3). The solid-state structures of all complexes have been determined by single-crystal X-ray diffraction. Magnetic studies for the complex 1–3 indicated a very weak antiferromagnetic interaction between copper(II) ions in crystal lattice. Complexes were successfully evaluated for their cytotoxic activities on the normal human dermal fibroblast (NHDF) cell line and the antitumor activity using the human lung carcinoma (A549), epithelioid cervix carcinoma (HeLa), colon (LoVo), and breast adenocarcinoma (MCF-7) cell lines. Complexes 1 and 3 revealed lower toxicity to NHDF than A549 and HeLa cells, meanwhile compound 2 appeared to be more toxic to NHDF cell line in comparison to all cancer lines. Additionally, interactions between the complexes and human apo-transferrin (apo-Tf) using fluorescence and circular dichroism (CD) spectroscopy were also investigated. All compounds interacted with apo-transferrin, causing same changes of the protein conformation. Electrostatic interactions dominate in the 1/2 – apo- Tf systems and hydrophobic and ionic interactions in the case of 3.


Introduction
Coordination chemistry of copper compounds has been an important subject of intensive investigations by researchers for many years [1][2][3]. Coordination polymers or discrete complexes constructed of copper metal ions and organic ligands/linkers have been investigated for the design of a new generation for luminescent [4][5][6][7], magnetic [8], catalytic [9][10][11][12], host-guests [13], and biological [14][15][16][17][18][19] properties. Thus, recently, there is great attention on the coordination chemistry of copper(I/II) complexes, due to their structural and physicochemical properties, in particular biological. On the one hand, an important role of copper in various biological processes, such as photosynthesis and dinitrogen metabolism as well as oxidative stress protection, etc., was reported by many scientists [14,15]. On the other hand, recent biological studies demonstrated that copper(I/II) coordination compounds represent various biomedical applications [16] that contain antifungal [17] and antibacterial [18] activities. Moreover, some of these complexes show also a relatively high activity as antiviral, anticancer, and antiproliferative agents and lower toxicity than cisplatin [15,[17][18][19].

Synthesis and Characterization
The reaction of Cu(NO 3 ) 2 with a stoichiometric amount of PTA=O, in EtOH solution under reflux conditions, followed by the addition of a stoichiometric amount of dmphen (Cu:PTA=O: dmphen molar ratios of 1:1:1), led to [Cu(NO 3 )(PTA=O)(dmphen)][PF 6 ] (1) discrete coordination compound (Scheme 2). Treatment of CuCl 2 with dmphen in the presence of KPF 6 in a Cu:dmphen: KPF 6 molar ratio (1:1:1) under the same conditions afforded [Cu(Cl)(dmphen) 2 ][PF 6 ] (2) complex (Scheme 2), whereas the use of Cu(NO 3 ) 2 salt with a stoichiometric amount of dmphen in MeCN under reflux conditions and molar ratio (1:1) gave rise to the formation of [Cu(NO 3 ) 2 (dmphen)] ·MeCN (3). Compounds 1-3 were isolated as air stable, green microcrystalline solids in ca. 36-85% yields based on appropriate copper salt, and characterized by IR spectroscopy, elemental analyses, and single-crystal X-ray diffraction. Although during the preparation of this manuscript the X-ray structure of 2 was published but described by us as one-pot reaction that was faster and yielded a cleaner product than the previously published method [49].

Scheme 2.
Schematic representation of the synthesis and structural formulae for 1-3.

Magnetic Properties
The magnetic properties of 1-3 were investigated over the temperature range of 1.8-300 K. Plots of magnetic susceptibility χ m T product vs. T (χ m is the molar magnetic susceptibility per one Cu(II) ion) are given in Figure 4. For 1-3, χ m T was essentially constant (~0.4 cm 3 mol −1 K) in the whole temperature range. It was consistent with one unpaired electron in magnetically diluted copper(II) complexes [63]. Only a slight decrease of the value χ m T in the low-temperature range (below 10 K) was caused by occurrence of weak antiferromagnetic interactions in the crystal lattice.  (1) and (2). The inset shows field dependences of the magnetization (M per one Cu II atom) at 2 K for 1-3. The solid line is the Brillouin function curve for one uncoupled spin with S = 1 / 2 and g = 2.0.
The variation of the magnetization (M) with respect to the field (H), at 2 K, also confirmed the nature of the ground state in 1-3 ( Figure 4, see inset). As the magnetic field increased, the M vs. H curves are linear in the whole field range and indicate values of magnetization close to 1 µ B at 5 T. The magnetization curves for 1-3 were reproduced by the equation M = gβSNB s (x) (S = 1·S Cu ), where B s (x) is the Brillouin function and x = gβH/kT [63]. The experimental values closely followed the Brillouin function for one uncoupled spin with S = 1 / 2 and confirmed our previous assumption.
From the magnetic point of view, 1, 2, and 3 were considered as mononuclear compounds. In such situations the magnetic data were fitted using the susceptibility equation for S = 1 / 2 (Equation (1)). To elucidate the significance of exchange between copper(II) ions in the crystal lattice, a molecular field correction term was also included (Equations (1) and (2)) [63,64].
where zJ ' is the intermolecular exchange parameter, z is the number of the nearest neighbors, and the others have their usual meaning. A least-squares fitting of the experimental data led to the following values: zJ' = − 0.51 (1) Figure 4. The calculated curves reproduced the magnetic data very well in the whole temperature range ( Figure 4). The criterion used in determination of the best fit was based on minimization of the sum of squares of the deviation: R = Σ(χ exp Tχ calc T) 2 /Σ(χ exp T) 2 . Small zJ' exchange parameters (below 1 cm −1 ) were consistent with the crystal structures of 1-3. Since the intermolecular Cu···Cu distances in all complexes were all rather long, it was to be expected that the coupling between electrons of the copper(II) ions in the system were weak. This fact was expected in mononuclear magnetically diluted Cu(II) compounds [65][66][67][68].
The EPR spectra of solid samples 1, 2, and 3 recorded in the X-band at room temperature and 77 K were essentially similar and additionally confirmed the properties detected by the direct magnetic measurements. The EPR spectra of 1-3 resembled monomeric copper(II) species with poorly resolved hyperfine features and g || > g ⊥ > g e . The spectral features were characteristic of a distorted square-planar geometry and a d x 2_ y 2 ground state for the copper(II) center [69][70][71][72][73]. It was in fairly good agreement with the copper(II) geometries obtained from the crystal structures.

Cytotoxic Assays
On the one hand, toxicity levels of 1-3 on NHDF cells were investigated, which reflected normal cells of a body, and on the other hand, the ability of these compounds to kill the human lung (A549), breast (MCF7), colon (LoVo), and cervical (HeLa) cancer cells. These results of in vitro cytotoxicity tests are demonstrated in Table 2.  It may be interesting to compare the activity of compounds 1-3 with the copper(I)-dmphenphosphine complexes described in literature, especially as the influence of copper(I) compounds on cancer cells is well known. They usually show strong or moderate anticancer effect. However, they are not always safe for normal cells. According to Komarnicka et al. reported data, IC 50 value for [Cu(iodide)(dmphen)(P(p-OCH 3 -Ph) 2 CH 2 OH)] showed 62.01 µM for A549 cell line, while, at the same time, already 32-43 µM was toxic for normal cell lines like MRC5, HEK293T, or HaCat [38]. In our case, not only that complexes 1 and 3 were less toxic for NHDF (normal cells) than A549, but also smaller than in the case of [Cu(iodide)(dmphen)(P(p-OCH 3 -Ph) 2 CH 2 OH)] doses (0.29 and 0.43 µM, respectively) were cytotoxic for cancer cells. A similar situation was observed in regard to [Cu(iodide)(dmphen)(P(p-OCH 3 -Ph) 2 CH 2 -SarGly)], where IC 50 value for A549 was >100 µM. On the other hand, in the case of MCF7 cell line, both mentioned-above copper-iodide complexes showed better anticancer properties than 1 and 3, which confirmed that the same compounds may be toxic for one type of neoplasia and nontoxic for others [38].
The bioactivities of the compounds is often compared with values of their logarithm of 1-octanol/water partition coefficient (log(P), see Experimental).
This method describes hydrophobic/hydrophilic properties of the compounds. For the accurate biological activity and bioavailability of potential drugs, a balanced solubility in both water and nonpolar compounds such as lipids is essential [79]. Indeed, the different activities of the copper complexes 1-3 could be related to their positive, low log(P) values (0.90-1.75), in contrast to cisplatin with strongly negative log(P) factor (−2.21) [80]. Figure 5 shows the fluorescence emission spectra of apo-Tf -copper complexes systems registered under physiological conditions and the excitation wavelength at 280 nm. It was evident that apo-Tf showed a strong fluorescence band at 325.5 nm and titration of the protein with small amounts of the complexes causing distinct decrease in fluorescence intensity of the protein. Moreover, the maximum band position was red shifted to ca. 330.5 nm. All the observations indicate that the apo-Tf conformation was changed and the protein's chromophores were moved to a more polar environment [81]. There are two mechanisms of fluorescence quenching, dynamic and static, that can be distinguished based on the temperature dependency. Dynamic quenching is due to collisions between quencher and fluorophore and static quenching results from the formation of the ground state complex between the molecules [82]. A modified Stern-Volmer equation (Equation (3)) was used to calculate Stern-Volmer quenching constants (K SV ) and quenching rate constants (Kq) of the interactions between apo-transferrin and copper complexes and to determine the mechanism of fluorescence quenching.

Apo-Transferrin Interactions
In the equation, Fo and F are the fluorescence intensities of protein in the absence and presence of the quencher in the concentration (Q), respectively; K SV is the Stern-Volmer quenching constant; and fa is the fraction of the fluorophore accessible to the quencher.
Kq, the quenching rate constant of the biomolecules, is expressed as: where τo is the fluorescence lifetime of protein without quencher. For apo-transferrin, the τo value is 2.5 ns [83]. The modified Stern-Volmer plots are shown in Figure 6A, and Table 3 lists quenching parameters. As shown, the Ksv values increased with the temperature increases in the case of apo-Tf -complex 1 and complex 2 systems (2.50-5.69·10 4 M −1 ). Simultaneously, the proper Kq values (10 13 − 10 14 (M −1 s −1 )) were higher than 2 × 10 10 M −1 s −1 , the maximum Kq value expected for dynamic processes of the biopolymers [84,85]. Therefore, we concluded that some specific interactions were involved in the reaction between apo-Tf-copper complexes 1/2 that mad Kq greater, and the fluorescence quenching was initiated by both static and dynamic processes. In the case of apo-Tf -complex 3 system, the Ksv values (5.37 × 10 4 and 3.51 × 10 4 in 300 K and 310 K, respectively) were reversely correlated with the temperature, indicating static fluorescence quenching mechanism. Association constants (Ka) and number of binding sites (n) were determined based on the Equation (5): where Fo and F are the fluorescence intensities in the absence and presence of the quencher in the concentration (Q), respectively, Ka is the binding constant, and n is the number of binding sites. All binding data are collected in Table 4 and appropriate plots are shown in Figure 6B. It is clear that under tested conditions only one binding site (n) in protein for all copper complexes existed. The association constants (Ka) decreased with the temperature increases for 1 and 2, suggesting forming of the unstable complexes. The binding constant of apo-Tf -3 system increased with temperature increase, indicating the formation of the stable adduct and endothermic process.
The interactions of the tested compounds with apo-Tf were definitely stronger than that of cisplatin (10 6 -10 7 M −1 vs. 0.20-0.35 M −1 , respectively) [86,87]. Moreover, obtained association constants were higher than association constant of chromium(III)-phen -apo-Tf system (1.5 × 10 5 M −1 ) [88] or those obtained for NAMI-A (-Imidazolium-trans-tetrachloro(dimethylsulfoxide)imidazoleruthenium(III)name explained in [89]) and its reduced form (1.28 × 10 4 M −1 and 1.36 × 10 4 M −1 , respectively) [89].  It is known that interactions between proteins and small molecules include hydrogen bonds, van der Waals, and electrostatic forces as well hydrophobic interactions [90]. Based on thermodynamic parameters, we determined the type of interaction of copper complexes with apo-transferrin. Table 5 shows all ∆H 0 , ∆S 0 , and ∆G 0 values obtained from van't Hoff plots and Equations (6) and (7): where Ka is the bimolecular binding constant at the corresponding temperature (T); R is the gas constant; and ∆H 0 , ∆S 0 , and ∆G 0 are enthalpy, entropy, and free energy change, respectively. The positive ∆S 0 and negative ∆H 0 for complexes 1 and 2 suggest that electrostatic interactions were involved in the reaction between protein and complexes. Both positive ∆H 0 and ∆S 0 for 3 indicate hydrophobic and ionic interactions. Negative values for all systems point to spontaneous processes.
Circular dichroism (CD) measurements in the far UV region were performed to determine the effect of copper complexes on the apo-transferrin secondary structure. A characteristic for α-helical structure of protein negative bands was visible in the CD spectrum at wavelengths of 210 and 220 nm ( Figure 7) and were assigned n-π* transitions peptide bonds. All the tested complexes did not generate CD signal in the measured range. The α-helix content the free protein was equal to 19.80%.
The interaction of 1 and 2 with apo-Tf had an insignificant effect on its secondary structure and, upon binding, the complexes α-helix content decreased to 19.07% and 18.00%, respectively, when the molar ratio apo-Tf:1/2 was 1:20. In contrast to 1 and 2, binding 3 complex to apo-transferrin caused extensive changes in conformation of the protein reducing α-helix content to 17.05% and 8.16% at the molar ratios apo-Tf:3 equal 1:10 and 1:20, respectively.

Cytotoxic Properties on Normal and Cancer Cell Lines -Quantitative Suspension Test According to EN 14476
The 100 µL of suspension of NHDF, A549, HeLa, LoVo, and MCF7 cells at a density of 4 × 10 4 cells mL −1 were incubated in a 96-well polystyrene plate (NUNC, Denmark) for 24 h [93]. Product test solutions were prepared in DMEM, DMEM-F12, or EMEM supplemented with additional 2% FBS and 2 mM l-glutamine. Solutions of the reagents at concentrations from 460 (for 1), 454 (2), 690 (3), 1242 (Cu(NO 3 ) 2 ), 1732 (PTA=O), and 1440 µM (dmphen) to 4.60 × 10 −6 , 4.54 × 10 −6 , 4.69 × 10 −6 , 1.24 × 10 −5 , 1.73 × 10 −5 , and 1.44 × 10 −5 µM, respectively, were prepared and transferred (100 µL) into cell culture units (wells of microtiter plates) containing monolayer of cells. Eight units were inoculated with each dilution. Plates were incubated in 37 • C, 5% CO 2 and observed daily for 4 days for the development of cytotoxic effect, using an inverted microscope (Olympus Corp., Hamburg, Germany, Axio Observer, Carl Zeiss MicroImaging GmbH). Then the wells were washed in phosphate buffered saline (PBS) and stained using DAPI (4 ,6-diamidino-2-phenylindole) and propidium iodide (Merck, Darmstadt, Germany). The study used the fact that DAPI is able to be combined with cellular DNA, permeate through the membrane of cell, rapidly enter the nucleus of living cells and bind with DNA to form a DAPI-DNA complex. Propidium iodide does not pass through the membranes in healthy cells, but as a result of its damage it can penetrate into the cell. In cell culture these are cells that are apoptotic or damaged by any physical or chemical agent. Therefore, if these two fluorochrome are given at the same time, DAPI will get to healthy cells faster, while propidium iodide will stain dead or dying cells. The calculation of the ratio of cells counted on a specific area, allows to determine the potency of a substance in inhibiting a specific biological function (IC 50 ).

Fluorescence Spectroscopy
Emission fluorescence spectra were recorded on Jasco 8200 spectrofluorimeter (Tokyo, Japan) in the range of 300-500 nm using 1.0 cm quartz cells. The λex was set to 280 nm and the excitation and emission slit widths were set to 5 nm. All fluorescence intensities were corrected (according to Equation (8)) for the inner filter and dilution effects and the corrected values were used to determine the quenching mechanism and binding data. Moreover, the copper complexes showed a fluorescence signal in the measured range. Therefore, all spectra are shown as different spectra of (apo-Tf-copper complex)−(copper complex). F corr = exp(A 280 + A em /2) × F exp (8)

CD Spectroscopy
Circular dichroism measurements were carried out on a Jasco J-715 spectropolarimeter (Tokyo, Japan) in the range of 190-250 nm using 0.1-cm cuvettes. The α-helical content of the protein was calculated from the Equations (9) and (10): where MRE is the mean residue ellipticity, C p is the molar concentration of the protein, n the number of amino acid residues (679), and l is the path length (0.1 cm).

Synthesis and Analytical Data
The stirring for 1 h, the color of the solution changed from blue to green. Solid dmphen (208.25 mg, 1 mmol) with KPF 6 (184.1 mg, 1 mmol) were added and the obtained mixture was refluxed for 2 h and then filtered off. The filtrate was left to slowly evaporate at~4 • C for several days, producing green X-ray quality single crystals of 1 and red byproduct, determined as (Cu(dmphen) 2 ) (PF 6

Stability/Solubility Tests and Octanol-Water Partition Coefficient Determination
The copper complexes 1-3 were air stable in the solid state and at least for several days in H 2 O solutions. In a general procedure, the complex was dissolved in H 2 O in air atmosphere. IR spectra of the solid samples after evaporation of the solvent under vacuum showed that no evident changes were produced in several days at room temperature. Also UV-Vis spectra of water solutions of 1-3 confirm their stability in several days. In a stepwise procedure of the solubility determination, increasing volumes of water were added at 25 • C to approximately 5 mg of the compound in a 5 mL glass tube. After each addition of an amount of water (100 µL), the mixture was shaken for several minutes and then visually checked for any undissolved parts of the sample.
The log P values consistent to the octanol-water partition coefficient were adjusted to the solubility properties of the compounds [94]. Complexes 1-3 were dissolved in water, previously saturated with octanol with concentrations 10 −5 M. Into a 25-mL flask at room temperature with a magnetic stir bar was introduced initially 5 mL of octanol previously saturated with water and then 5 mL of the complex solutions in water. A two-phase mixture was stirred vigorously for 15 min and samples, which were measured by UV-Vis spectroscopy, were taken from the separated phases. Values of log P have been found as 0.90, 1.75, and 1.11, for 1, 2, and 3, respectively.

X-ray Crystallography
Single crystal data collection was performed on KUMA diffractometer with Sapphire CCD detector, equipped with an Oxford Cryosystems open-flow nitrogen cryostat, using ω-scan and a graphite-monochromated Mo Kα (λ = 0.71073 Å) radiation. Cell refinement, data reduction, analysis, and absorption correction were carried out with CRYSALISPro (Rigaku Oxford Diffraction, Wrocław, Poland) software. The structures were solved by direct methods with SHELXS [95], and refined with full-matrix least-squares techniques on F 2 with SHELXL [96]. The C-bonded hydrogen atoms were calculated in idealized geometry riding on their parent atoms. In the case of 1, fluoride atoms were disordered over two equally occupied positions, while for 3, acetonitrile molecule was disordered over two positions with occupation factor 0.5. As the crystals structure of compound 2 was already published (FEYTUB, CCDC 1819088) only preliminary data collection was performed to confirm identity [49].
The molecular structure plots were prepared using Diamond (

Magnetic Measurement
The magnetization of powdered samples 1, 2, and 3 was measured over the temperature range 1.8-300 K using a Quantum Design SQUID-based MPMS-XL-5-type magnetometer (San Diego, USA). The superconducting magnet was generally operated at a field strength ranging from 0 to 5 T. Measurements were made at magnetic field 0.5 T. The SQUID magnetometer was calibrated with the palladium rod sample. Corrections were based on subtracting the sample-holder signal and the contribution of χ D was estimated from the Pascal's constants [97].

EPR Spectra
Electron Paramagnetic Resonance (EPR) spectra of powdered samples 1, 2, and 3 were recorded at room temperature and 77 K on a Bruker ELEXSYS E 500 CW-EPR (continuous-wave EPR) spectrometer (Billerica, MA, USA) operating at X-band frequency and equipped with an ER 036TM NMR Teslameter and E41 FC frequency counter.

Conclusions
A series of water-soluble and air-stable copper(II) discrete complexes based on 2,9-dimethyl-1,10-phenanthroline (dmphen) and mixed-ligands, containing PTA=O (1,3,5triaza-7-phosphaadamantane-7-oxide), The solid-state structures of all complexes were determined by single-crystal X-ray diffraction. Compound 1 extended a still poor family of Cu(II)-based metal-organic architectures assembled from a versatile, water-soluble, and cagelike aminophosphine oxide (PTA=O). The magnetic susceptibility measurements, as well as a relationship between the magnetization and magnetic field strength in 1-3, revealed very weak antiferromagnetic interactions between magnetic centers of copper(II) ions in crystal lattice at low temperature.
Complexes were successfully evaluated for their cytotoxic activities on the normal human dermal fibroblast (NHDF) cell line and the antitumor activity using the human lung carcinoma (A549), epithelioid cervix carcinoma (HeLa), colon (LoVo), and breast adenocarcinoma (MCF-7) cell lines.