Synthesis, Characterization and Biological Evaluation of Metal Adamantyl 2-Pyridylhydrazone Complexes

Four new complexes derived from adamantly containing hydrazone (APH) ligand with Cu(II) (1), Co(II) (2), Ni(II) (3) and Zn(II) (4), have been synthesized and characterized using different physicochemical methods. The structure of the ligand APH and its copper complex 1 have been established by single-crystal X-ray diffraction direct methods, which reveal that complex 1 has distorted square-pyramidal geometry. Complexes 1–4 are screened against seven human cancer cell lines namely, breast cancer cell lines (MCF7, T47D, MDA-MB-231), prostate cancer cell lines (PC3, DU145) and the colorectal cancer cell line Coco-2, for their antiproliferative activities. Complex 1 has shown a promising anticancer activity compared to the other ones. The structural and spectroscopic analysis of APH and its complexes are confirmed by DFT calculations.


Introduction
The potential of having complexes as anti-cancer agents stemmed from the accidental discovery of the cis Platine, cis-[Pt(II)(NH 3 ) 2 Cl 2 ], in the 1960s [1]. However, due to its toxicity and narrow selectivity, the search for more selective and potent complexes has exponentially increased since then, and a new era of complex research has evolved [2,3]. In the last decades, the search was directed towards finding platinum-based complexes, which are less toxic, with a better and wider range of selectivity. As a result of such hunt, a wide range of platinum complexes were introduced, like Pt(II) cisplatin, carboplatin, oxaliplatin and Pt(IV) satraplatine and diazides [4,5]. In addition to treatment, complexes hold great applications in the therapeutic and diagnostic of cancer. Other complexes containing zinc, gold, copper, nickel, titanium, rhodium, vanadium, cobalt, ruthenium and other metals are being given great attention as potential anticancer agents, due to their unique physiochemical properties [4,6,7].
The ability of complexes to adopt numerous three-dimensional structures allows them to have a diverse formation of functionalized targets of unique molecular structures, geometries and kinetic properties [4], which makes them promising candidates as anticancer agents. The geometry, oxidation and electronic state of the metal, due to partially filled d-orbitals, are the main factors in the activities and selectivities of the designed complexes. Another factor that also effects both selectivity and the activities of complexes is the existence of biologically relevant moieties as a counter ligand to the metal of interest.

Synthesis and Characterization
The ligand Adamantane-1-carboxylic acid pyridin-2-ylmethylene hydrazide (APH) was obtained via condensation of 2-pyridinecarbaldehyde with adamantane-1-carboxylic acid hydrazide according to published procedures [33,34] as described in Scheme 1. All complexes obtained in good yield from the reaction of the metal chloride salts with ligand APH in refluxing ethanol for appropriate time, as shown in Scheme 2 All complexes were characterized using different spectroscopic techniques. High resolution mass spectra were obtained for all synthesized compounds, and the results confirmed the exact molecular formula and purity of the desired compounds. This mass spectrometric data supported the other data for the full assignments of the synthesized compounds.

FT-IR Spectra
IR spectra of complexes 1, 2, 3 and 4 analyzed in comparison to that of APH, recorded from 4000 to 400 cm −1 . Table 1 shows the vibrations for selected groups within the ligand, as well as the complexes. Compared with the ligand spectra APH, all complexes exhibit a shifted band from 1651 to 1588-1644 cm −1 assigned to ν(C=O). The calculated ν(C=O) is 1764 cm −1 where the hydrophobic interactions between adamantyl moieties are not included, as all calculations are performed on one APH unit. The ν(C=O) in the staggered conformer is lower in energy than the eclipsed conformation, due to it having less steric hindrance, as expected. The shift of the band to a lower wavenumber in the spectra of the complexes indicated the coordination of oxygen with metal ion, and the formation of M-O bond. The ν(C=N) band of the azomethine in the ligand at 1546 cm −1 is shifted to a lower wave number in the spectra of the complexes, 1523-1529 cm −1 , suggesting its coordination with metal ion, and was also attributed to delocalization of the lone pair, and the formation of the Cu-N bond [40]. Low-energy in-plane and out-of-plane pyridine ring vibrations were shifted to higher frequencies in complexes 1, 2 and 3 (from 670-684 cm −1 ), compared to that of APH (624-407 cm −1 ), which may be attributable to the rigidity of the system due to complexation [41][42][43][44]. All such observations were confirmed by the DFT calculations.
The absence of the carbonyl band in the spectra of complex 2 indicated that the hydrazone ligand had been coordinated to metal center through carbonyl oxygen after deprotonation. This evidence was strengthened via the presence of new stretching vibrations in the range 1090-1150 cm −1 , attributed to the C-O in 2. The ν(C-N) band at 640 cm −1 in 4 of the pyridine nitrogen in the complex is slightly shifted from that of the ν(C-N)py band (624 cm −1 ), indicating the absence of coordination of nitrogen with metal ion. The presence of a new band at about 3153 cm −1 was attributed to the ν(H-N), which indicated the formation of a hydrogen bond between the hydrogen atom of the NH and the nitrogen in the pyridine ring.
The difference in the calculated corrected Gibbs free energy for complex 1 (eclipsed) for complex 1 (staggered) is −166 kcal/mole ( Figure 1). Such lower energy of the eclipsed complex is due to the greater separation between the carbonyl oxygen from the chlorides, hence minimizing electron-electron repulsion. The HOMO-LUMO prsentation ( Figure 2) shows relatively similar charge distributions, where the molecular orbital from the adamantyl group makes no contribution to similar molecular orbitals.

Magnetic Measurements and Molar Conductivities
The magnetic moment obtained at room temperature for complex 1 shows an µ eff value of 1.94 B.M., corresponding to one unpaired electron and is indicative of a d 9 system, which reveals the paramagnetic nature of the complex, indicating that copper in this complex is in the +2 oxidation state. The measured molar electrical conductivity of compound 1 is 277 Ω −1 ·cm 2 ·mol −1 in water, indicating that the complex behaves as a 2:1 electrolyte. These results confirmed the lability of Cl − anion in aqueous solution [45]. In the solid state, the Cl − anion is bonded to the Cu(II) ion as confirmed by X-ray analysis described below.
Magnetic moments of the complex 2 shows µ eff value of 4.64 B.M., which is indicative of a high spin d 6 system, and reveals the paramagnetic nature of the complex; hence, cobalt, in this complex, is in the +3 oxidation state. The measured molar electrical conductivity of compound 2 in water is 121 Ω −1 ·cm 2 ·mol −1 , indicating that the complex behaves as a 1:1 electrolyte in aqueous solution.
The magnetic moment of complex 3 shows µ eff value of 2.21 B.M., corresponding to two unpaired electrons, which is an indicative of d 8 system, which reveals the paramagnetic nature of the complex; accordingly, nickel in this complex is in the +2 oxidation state. The measured molar electrical conductivity of compound 3 is 217 Ω −1 ·cm 2 ·mol −1 in water, indicating that the complex behaves as a 2:1 electrolyte. These results attested the lability of Cl − anion in aqueous solution. In the solid-state, a chloride anion is coordinated to the Ni(II) ion.
The magnetic moment of the complex 4 is zero and the behavior is diamagnetic, as expected for d 10 configuration. The measured molar electrical conductivity of compound 4 is 78 Ω −1 ·cm 2 ·mol −1 in methanol, indicating that the complex behaves as a 1:1 electrolyte in methanolic solution: this result confirmed the lability of Cl − anion in methanolic solution.  A molecular structure view of APH with labeling and numbering of atoms, which is shown in Figure 3, along with selected crystalline structure parameters, is given in Table 2. The adamantyl groups constituting APH are disordered by 30 • rotation (Figure 3), giving rise to two structures. The crystallized ligand with a molecule of water is essential for the H-bonding formation that is clearly connecting every other two ligands, as is shown in Figure 4. Figure 3 shows that the APH, in the free state, exists in the keto tautomeric form, and the structural conformation of the molecule is E, with respect to the imine double bond. The unit cell of the ligand ( Figure 5) shows the head-to-head arrangements of the adamantyl moieties within the backed crystals, which can be attributed to the Van der Wall forces between the hydrophobic parts of the ligand units.
. Molecular plot of APH shows the different intermolecular interactions between APH and water moieties. Only one asymmetric unit interaction is shown, as the other unit is symmetrically equivalent.

Crystal Structure of 1
A crystal structure of 1 with an atom numbering scheme is shown in Figure 6. The bonding parameters are presented in Table 2. The crystal structure reveals that complex 1 crystallizes in a triclinic system, with a space group Pī. The unit cell of the crystal contains two molecules of (C 17 H 18 Cl 2 CuN 3 O). Complex 1 adopts a distorted square pyramidal structure. The Coordination sphere around copper(II) ion is satisfied with the five coordination environment, which is formed from the tridentate ligand; a pyridyl nitrogen atom (N 1 ), an imino nitrogen (N 2 ) and a keto oxygen atom (O 1 ), along with two chloride ions. In the complex, N 1 , N 2 , O 1 and Cl 1 occupy the basal plane of a distorted square pyramid, where the Cl 2 is located in the apical position. The bond distances and bond angles of compound 1 are listed in Table 3. In the ligand APH, the C=O bond length is 1.229(6) Å, while in complex 1, this bond is 1.234 (7)   Applying continuous shape measures (CShM) on Cu(II) ion by comparing with all reference standard 5-coordinate polyhedrons were calculated using SHAPE 2.0 program [46]. This five-coordination sphere and geometry can be best described as a square pyramid (C 4v ) with CShM = 1.545, in comparison with the closest value for a vacant octahedron (a regular polyhedron with one or two vertices removed) with CShM = 3.186.
Six different cancer cell lines (MCF7, T47D, MDA-MB-23, PC3, DU145 and Caco-2) are used to assess the effects of those compounds on cell proliferation. Two different concentrations of the test compounds are used: 12.5 µM and 50 µM. The % of cell growth inhibition is assessed by an MTT assay. The obtained results are compared with the ligand APH. In addition, doxorubicin (a well-known antitumor drug that is used in the treatment of several types of cancers by activating apoptotic pathways, leading to the death of these cells [60]), is used as a positive control.
As shown in Table 4, all of the tested compounds have shown an antiproliferative activity against the used cell lines when they are compared with the parent ligand APH at both treatment doses of 12.5 and 50 µM. Compounds 1 and 2 are the most cytotoxic of them all. Compound 1 has shown activities of 93.30 ± 1.20%, 88.60 ± 1.17%, 90.26 ± 0.55%, 89.88 ± 0.39%, 89.85 ± 0.73%, 87.06 ± 1.57% of cell viability inhibition in MCF7, T47D, MDA-MB-23, PC3, DU145 and Caco-2 cell lines, respectively. Compound 2, with a concentration of 50 µM (Table 4) has shown cytotoxic activities of 89.35 ± 1.22%, 86.03 ± 0.99%, 84.52 ± 2.54%, 88.60 ± 0.44%, 87.78 ± 1.32%, 85.22 ± 0.65% in MCF7, T47D, MDA-MB-23, PC3, DU145 and Caco-2 cell lines, respectively. Compounds 3 and 4 are also found to be toxic to cells, but with lower impacts on cell viability inhibition than compounds 1 and 2, with a % of growth inhibition of 37.29 ± 2.71 for MCF7, 44.78 ± 2.10 for MDA-MB-231, 62.47 ± 0.14 for PC3, and 64.28 ± 2.78 for CaCo2 for compound 3. Compound 4 of 50 µM concentration has cytotoxic activities of 64.83 ± 1.18 for MCF7, 32.31 ± 1.95 for MDA-MB-231, 84.93 ± 0.53 for PC3, 71.62 ± 1.46 for DU145 and 60.32 ± 1.43 for the Caco-2 cell line. Other studies have reported that copper (II) complexes with 1-adamantoyl hydrazone bearing pyridine rings are found to induce apoptosis and inhibit agiogenesis in some tumor cell lines [61].  There was no significant difference in the % of cell viability inhibition values between the different cell lines used in this study, following the treatment with compounds 1 and 2. On the other hand, compounds 3 showed a stronger antiproliferative effect against the two prostate cancer cell lines and Caco-2 (the colon cancer cell line), compared to their effect on the three breast cancer cell lines. Compound 4 caused more cell viability inhibition in the two prostate cancer cell lines than the other ones.
Doxorubicin, an anthracycline antibiotic and one of the most effective anticancer drugs against breast cancer since 1970 [62], showed an inhibition of cell viability ranging from 35.41% to 65.68%, and from 41.06% to 72.23% at 12.5 µM and 50 µM treatment doses, respectively. At a concentration of 12 µM, doxorubicin was showing stronger cytotoxic effects on cells, compared with our four test compounds, except for the CaCo-2 cell line, where compound 1 was found to be more cytotoxic to these cells with a 2-fold increase in the % of viability inhibition than doxorubicin (Table 4). At 50 µM treatment dose, compounds 1 and 2 showed a stronger antiproliferative activity on the three breast cancer cell lines used in this study than doxorubicin; compounds 1, 2, and 3 were more cytotoxic to PC3 cells compared to doxorubicin; and all four tested compounds were more cytotoxic to DU145 and CaCo-2 cells when compared with doxorubicin (Table 4).
It is concluded that the cytotoxicity effect of APH against different tumor cell lines has been improved through its complexation with different metals, hence, making these complexes possible candidates as anticancer drugs. Furthermore, future studies shall be conducted on these newly synthesized bioactive compounds, in order to understand the mechanisms of the cytotoxic and mutagenic changes they may cause cancer cells.

General Methods
Starting materials and solvents used were analytical grades and provided from Scharlau, Fluka and Sigma Aldrich (St. Louis, MO, USA). All reactions were monitored by thin layer chromatography (TLC) (Merck, Darmstadt, Germany) using Merck aluminum plates, which are pre-coated with silica gel PF254. The 1 H− and 13 C-NMR spectra were recorded on 300 MHz Bruker AVANCE (Bruker BioSpin, Ettlingen, Germany). Chemical shifts were reported as δ values in ppm. Spectra were acquired in either DMSO-d 6 or CDCl 3 (Scharlau, Fluka and Sigma Aldrich (St. Louis, MO, USA). High-resolution mass spectra (HRMS) were measured (in positive/or negative ion mode) using the electrospray ion trap (ESI) technique by collision-induced dissociation on a Bruker APEX-IV (7 Tesla), (Bruker Daltonics, Bremen, Germany) instrument. Melting points (mp) were determined on an Electrothermal Melting point Apparatus, and were reported as acquired without corrections. The infrared spectra were recorded from 4000 to 400 cm −1 region with a Nicolet Impact-400 FT-IR spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) using KBr pellets.

Synthesis of the Hydrazone Ligand (APH)
The hydrazone ligand was prepared by adding 2-pyridinecarbaldehyde (1.1 mmol) to a solution of 1 mmol of adamantane-1-carboxylic acid hydrazide in absolute ethanol containing six drops of glacial acetic acid [33,34]. The mixture was refluxed for 4 h, and the reaction progress was monitored by TLC. The mixture was then poured into cold water and the formed precipitate was filtered, washed with petroleum ether, and then recrystallized from ethanol to produce white crystals (90% Yield, mp 136-137 • C).

Computational Methods
Structures of all complexes were optimized in silico gas phase using DFT methods with the Pople valence triple zeta basis set with added diffuse and polarization functions 6-311 + G (2d, p). No negative or imaginary frequencies were obtained for any structures. All calculations were performed on G16 running HPC cluster. Calculations were on conducted on distorted square pyramidal on complex 1, seesaw geometry on complex 2, distorted square pyramidal on complex 3 and distorted tetrahedral geometry on complex 4. The optimized structure of complex 1 is in complete agreement with the obtained single crystal structure one. In order to validate the calculations on molecules using the previously mentioned methods, another DFT calculation was conducted on complex 1 with Ahlrich valence triple zeta basis set with added polarization functions DefTZVP. Results from both basis sets eluded the same frequencies as assigned in Table 3. The discrepancies between the obtained experimental frequencies with the calculated ones can be attributed to the solvent molecules that are trapped in the crystals.

Viability Assay
Cells from each cell line were seeded at a density of 10 × 10 3 cells/well in 96-well plates and allowed to attach overnight. Cells were treated with two different concentrations (12.5 µM and 50 µM) of the tested compounds 1, 2, 3, 4 and doxorubicin, which was used as a positive control. The experiment was done in 4 replicas. After 72 h treatment, MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) assay was performed according to the cell proliferation assay kit (Promega, Madison, WI, USA). Absorbance (OD) was measured at 570 nm, with background subtraction at 630 nm. DMSO was used as a negative control, and the percentage of viability inhibition was measured as an indication of the cytotoxicity of the tested compounds.

Conclusions
In the present work, we have reported the synthesis, characterization and anticancer activities of four new adamantylpyridylhydrazone complexes. All prepared complexes contain the new tridentate N,N,O-donor Schiff base, which is derived from theadamantylpyridylhydrazone nucleus. One crystalline complex (1) was analyzed using its X-ray single structure analysis, and it was found that it exists in a square pyramidal geometry. Complex 2 exists in the form of Co(III) ions. Further investigations of the structure activity-relationship studies of our new complexes are in progress in our laboratory, and the results will be reported in due course.