Cu(II) complexes of a tridentate N,N,O-donor Schiff base of pyridoxal: Synthesis, X-ray structures, DNA-binding properties and catecholase activity

Two new Cu(II) complexes [Cu(L 1 )(N 3 )]( 1 ) and [Cu(L 1 )(NCS)] n ( 2 ), where HL 1 ((E)-4-((2(dimethylamino)ethylimino)methyl)-5-(hydroxymethyl)-2- methylpyridin-3-ol) is a N,N,O-donor Schiff base ligand, have been synthesized. These complexes were characterized on the basis of spectroscopic, electrochemical, and other physicochemical properties. X-ray crystal structure determination reveals that in complex 1 , Cu(II) is in a square planar geometry with the N,N,O-donor ligand and a terminally coordinated azide ion. In complex 2 , however, Cu(II) assumes a square pyramidal geometry, and apart from the tridentate Schiff base ligand there is a thiocyanate ion, which acts in a µ 1,3 -bridging mode, connecting adjacent Cu(II) atoms in an axial-equatorial fashion forming an one dimensional chain. Cyclic voltammetry shows that complexes undergo Cu(II)/Cu(I) reductions at −0.4 0 to − 0.55 V followed by imine reduction at around −0.7V. Cu(I)/Cu(0) reduction is observed at approximately −0.9V. Complex 1 also shows appreciable catalytic activity for the aerial oxidation of 3,5 DTBCH 2 to DTBQ (catecholase activity). Both the complexes also show strong binding affinity towards calf-thymus DNA. base ligand (HL ). Their single crystal X-ray structures have been determined. The spectroscopic and electrochemical properties of the complexes were investigated. DNA binding affinity of these complexes was also studied. 3,5-DTBC was used as a model substrate to study the catecholase activity of the complexes and only 1 was found to be catalytically active.


Introduction
Cu(II) ion plays an important role in biological systems and it is at the active site of various enzymes such as catechol oxidase, superoxide dismutase, multicopper oxidases. [1][2][3][4][5][6][7][8] Catechol oxidase is a copper-containing enzyme whose activity is similar to that of tyrosinase, a related class of copper oxidases. They are ubiquitous plant enzymes that catalyze oxidation of a broad range of ortho-diphenols to the corresponding o-quinones coupled with the reduction of oxygen to water. Krebs and co-workers reported the crystal structures of deoxy and met forms of catechol oxidase which reveal that these proteins have a dinuclear copper center. 1,9 The crystal structures of catechol oxidase have enhanced the understanding of the mechanism of catecholase activity of tyrosinase and/or catechol oxidase. Mechanisms of catechol oxidation by the neutral enzyme or synthetic model complexes have been proposed by various workers, prominent among them are the catalytic cycles proposed by Solomon , and Krebs. 1,4,10,11 Model complexes have played an important role in understanding the structural, spectroscopic and catalytic properties of the copper sites in protiens. [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26] Various workers have tried to rationalize the catalytic efficiency of the catechol oxidation reaction with variety of parameters like redox potentials, solvent effect, pH of the medium, number and nature of bridging ligands, effect of secondary coordination sphere etc., however our knowledge in this direction is still sketchy and incomplete. 12,13,[17][18][19][20][21][22][23][24][25][26] Therefore, there is need for study of new model complexes showing catecholase activity and exploration of relationship between their structural, spectroscopic and catalytic properties.
From biological point of view, study of the Schiff bases of pyridoxal and their metal complexes are attractive for multiple reasons: pyridoxal is a close analogue of pyridoxine (vitamin B6) and thus its complexes may be bio-tolerable, 27 the manganese(II) complex of dipyridoxal diphosphate (N,N′-dipyridoxylethylenediamine-N,N′-diaceticacid 5,5′-bis(phosphate)) is a contrast agent for magnetic resonance imaging of the liver, 28,29 Cu(II) complex of pyridoxal has proven to be a highly promising candidate for the treatment of diabetic complications, 30 and so on. Therefore, we though it will be meaningful to synthesize pyridoxal appended ligands and to explore catalytic and biological properties of metal complexes of such ligands. 26,31 Pseudohalide ions such as (N3 − , NCS − , NCO − , and N(NC)2 − ) are versatile ligands that can bind transition metal ions in different of ways. They can behave as monodentate or as bridging ligands leading to the formation of mononuclear and polynuclear species. Among these, the azide and thiocyanate containing metal complexes are the most investigated systems because of their diverse structures and applications in magnetic materials. [32][33][34] The azide anion can bind to the metal ion through end on (EO) bridging mode, generally leading to ferromagnetic coupling, while the end to end (EE) bridging results in antiferromagnetic coupling. Furthermore, EE and EO bridging modes may simultaneously exist in some species, leading to different topologies and materials with unprecedented, magnetic behaviour. [35][36][37] On the other hand SCN − anion is a highly versatile ambidentate ligand with two donor atoms. It can coordinate either through the N or the S atom, or both, giving rise to linkage isomers, 38,39 dimers 40 or polymers. 41 The interaction of small molecules with DNA has been an active area of research at the interface of chemistry and biology. [42][43][44][45][46][47] These small molecules are stabilized on binding to DNA through a series of weak interactions, such as the π-stacking interactions associated with intercalation of a planar aromatic group between the base pairs, hydrogenbonding and van der Waals interactions of functionalities bound along the groove of the DNA helix, 48 and the electrostatic interaction of the cation with phosphate group of DNA. 49 In recent years there is significant upsurge in the study of binding of transition metal complexes to DNA, with the aim of developing new DNA foot printing and DNA cleaving agents as well as identifying potential anticancer drugs. [50][51][52][53][54] base ligand (HL1). Their single crystal X-ray structures have been determined. The spectroscopic and electrochemical properties of the complexes were investigated. DNA binding affinity of these complexes was also studied. 3,5-DTBC was used as a model substrate to study the catecholase activity of the complexes and only 1 was found to be catalytically active.

Materials and methods
Pyridoxal hydrochloride was purchased from Aldrich and triethylenetetramine was purchased from E. Merck. All other chemicals and solvents were of reagent grade and used as such. Solvents for spectroscopic and cyclic voltammetry studies were of HPLC grade obtained from Merck or Aldrich. Elemental analyses were performed on a Perkin-Elmer 2400 C, H, N analyzer. Infrared spectra were recorded as KBr pellets on a JASCO FT-IR-460 spectrophotometer. UV-Vis spectra were recorded using a JASCO V-630 spectrophotometer. 1 H NMR spectra were recorded on a Bruker AVANCE DPX 300 MHz spectrometer using, Si(CH3)4 as internal standard. ESI-MS spectra of the samples were recorded on JEOL JMS 600 instrument. Fluorescence spectra at room temperature were recorded using PTI made QuantaMaster40 spectro-fluoremeter. EPR spectrum of complex 2 was recorded in DMF solution at 77K with a Varian E-112 EPR Spectrometer at SAIF, IIT, Bombay.

Synthesis of Ligand (HL1)
Pyridoxal hydrochloride (0.203 g, 1mmol) was dissolved in 40 mL methanol and the pH was set to 6.5-7.0 by addition of concentrated aqueous KOH solution.
On dropwise addition of methanolic N,Ndimethylethylenediamine (0.088 g, 1mmol) to the neutral pyridoxal solution, the reaction mixture turned light yellow with no formation of precipitate. The solution was refluxed for three hours and then the solvent was evaporated on a rotary evaporator, when a yellowish brown oily product was obtained. It was triturated with acetone and then recrystallized from methanol to give a yellow hygroscopic solid.

Synthesis of the complexes
A methanolic solution (5 ml) of the ligand HL1 (0.206 g, 1 mmol) was added dropwise to a solution of Cu(ClO4)2•6H2O (0.370 g, 1 mmol) in the same solvent (5 ml). After the mixture was refluxed for 1 h, an aqueous solution of NaN3 (0.130 g, 2 mmol) or KSCN (0.194 g, 2 mmol) was added, and the resulting mixture was refluxed for further 3 h. The green solution was filtered and the supernatant liquid was kept at room temperature for slow evaporation. After one day, green crystals of [Cu(L1)(N3)] (1) or [Cu(L1)(SCN)]n (2), suitable for X-ray structure determination were obtained.

X-ray crystallography
X-ray diffraction data for 1 was collected on a Agilent GV1000 Atlas diffractometer at 150(2)K using Cu-Kα radiation (λ= 1.54184 Å). Data collection and reduction were done using CrysalisPro program 55 and empirical absorption correction using spherical harmonics, were implemented in Scale3 Abspack Scaling Algorithm within CrysalisPro. Structure was solved by Direct method using SHELXS 97 and refined by fullmatrix least-squares on F 2 using SHELEXL version 2014/3. 56,57 X-ray diffraction data for 2 was collected on a Bruker SMART APEX II diffractometer at 100 (2) K, utilizing a graphite monochromated Mo-Kα radiation (λ= 0.71073 Å). Data was indexed and refined as a nonmerohedral twin with two domains using Bruker SAINT V8.18C program. 58 The ratio between the domains is 0.190 (5). Structure was solved by Direct method using SHELXS-97 and refined by full-matrix least-squares on F 2 using SHELEXL version 2017/1. 56,57 Data were corrected for absorption effects using the Bruker 'TWINABS' program. 59 The non-hydrogen atoms were refined with anisotropic displacement parameters. All hydrogen atoms were placed at calculated positions and refined as riding atoms using Please do not adjust margins Please do not adjust margins isotropic displacement parameters. A summary of the crystallographic data is presented in Table 1. Important bond distances and bond angles are collected in Table 2. Crystal data have been deposited at CCDC with deposition numbers 1817674 and 1817791 for and 1 and 2 respectively.

DNA-binding experiments 2.5.1. Absorption spectral studies
The DNA binding experiments were performed in Tris-HCl/NaCl buffer (50 mM Tris-HCl/1 mM NaCl buffer, pH 7.3) using a methanol (CH3OH) solution of complexes 1 and 2. The concentration of CT-DNA was determined from the absorption intensity at 260 nm with a ε value of 6600 M −1 cm −1 . 60 The UV absorbance ratio at 260 and 280 nm (A260/A280) of 1.85 indicated that the DNA was sufficiently free of protein contamination. 61 The DNA stock solution was stored at 4°C in the dark and used within 4 days after preparation. The concentrated stock solutions of complexes 1 and 2 were prepared by dissolving them in CH3OH and diluting suitably with Tris-HCl (pH 7.4) buffer. Complex-DNA solutions were allowed to incubate for 5 min before the absorption spectra were recorded. The intrinsic binding constant Kb was calculated from the spectroscopic titration data using the following equation: 62 where εf and εb are the extinction coefficients of the free and fully bound complex while εa is the extinction coefficient observed for the absorption band at a given DNA concentration, Kb the equilibrium binding constant. A gives Kb as the ratio of the slope to the intercept.

Fluorescence spectral studies
For fluorescence quenching experiments DNA was pretreated with ethidium bromide and kept in at 4°C for 4 hour. The complexes were then added to this mixture and their effect on the emission intensity was measured using PTI made QuantaMaster40 spectro-fluoremeter. Samples were excited at 520 nm and emission was observed between 540 to 700 nm.

Electrochemical studies
Electrochemical data of 1-2 mM DMF solutions of the complexes were collected using a CH1106A potentiostat. A three electrode configuration, consisting of a glassy carbon working electrode and a Pt-wire auxiliary electrode, TEAP as supporting electrolyte and Ag,AgCl/KCl reference electrode, with scan rate rate (20-100) Vs −1 was used. The potentials were calibrated against the ferrocene/ferrocenium couple (E0 = 0.44 V). Electrochemical titrations of copper complexes with CT-DNA were recorded by cyclic voltammetry and DPV in DMF solutions at room temperature.
The reaction was carried out spectrophotometrically by monitoring the increase in the absorbance band of 3,5-di-tert-butyl-1,2-benzoquinone (3,5-DTBQ) at 400 nm as a function of time. Absorbance vs. wavelength plots were recorded for these solutions at a regular time interval of 5 min in the range 300-500 nm. The minimum and maximum complex to substrate ratio was 1:16 and 1:500 respectively. Km, Vmax and kcat values were determined by Lineweaver-Burk plots and Michaelis-Menten equation.

Detection of Hydrogen Peroxide in the Catalytic Reaction
The presence of hydrogen peroxide in the reaction mixture was analyzed iodometrically by assaying I3 -, which has a characteristic absorption band at 353 nm (ε = 26000 M -1 cm -1 in water). The oxidation reaction of DTBCH2 by complex 1 was carried out as described in the kinetic experiment. When the formation of quinone reached a desired value at 400 nm the solution was acidified with H2SO4 to pH 2 to stop the reaction. Water (3 ml) was added and the reaction mixture was then extracted two times with CH2Cl2 to remove the DTBQ formed during the reaction. To a 2 ml aliquot of the aqueous layer 1 ml of a KI solution (0.3 M) was added with a catalytic amount of ammonium molybdate (3% solution) to accelerate the formation of I3 -. Blank experiments were performed under identical conditions in the presence of DTBQ and complex 1, but only minor formation of I3was observed.

DFT
In order to understand the electronic structure of the synthesized complex 1 and 2, we have carried out DFT calculations with the Gaussian 03 package (revision B.04). 63 Molecular orbitals were visualized using Gauss View. Single point calculations, and population analysis of the molecular orbitals were carried out using the density functional theory (DFT) with B3LYP hybrid exchange functional. 64,65 For H, C, N and O atoms 6-31G* basis set were used. 66 The LANL2DZ basis set 67 and LANL2DZ pseudopotentials of Hay and Wadt 68 were used for the Cu atom. The geometry of the complexes was taken from the crystal structure, and single point energy calculated, without further optimization, in gas phase.

Results and discussion 3.1 Synthesis of complexes
The Cu(II) complexes 1 and 2 were synthesized by refluxing a mixture of methanolic solution of Cu(ClO4)2•6H2O and aqueous solution of pseudo halides (NaN3, KSCN) with the ligand. After filtration, on slow evaporation of the filtrate at room temperature, green colored square shaped crystals of the complexes suitable for X-ray diffraction studies, appeared after 1 day. A simple scheme showing the synthesis of complex is depicted in Scheme I. The isolated complexes were characterized by elemental analysis, IR, UV-vis spectroscopy, CV, and by single-crystal X-ray crystallography.
Scheme I: Schematic representation of the synthesis of the Metal complexes.

Description of X-ray crystal structure
The ORTEP diagrams along with atom numbering scheme for complexes 1 and 2 are shown in Figures 1 and 2 respectively and the important bond lengths and angles are listed in Table 2. In the complex 1, the copper ion occupies the central position of a regular square planar arrangement. Out of its four coordination sites three are occupied by the phenolato oxygen, imine nitrogen, and the tertiary amine nitrogen of the tridentate ligand. The remaining fourth coordination site is satisfied by an azido group which is linked to Cu(II) ion through a terminal nitrogen atom. The azide group, which acts as a monodentate ligand is slightly asymmetric [N(4)-N(5)=1.184 (2)

Infrared spectra
In the complexes bands appearing at 3402 and 3228 cm −1 are assigned to the O-H stretching vibrations of coordinated hydroxymethyl group and bands at 2928 and 2920 cm −1 indicate the presence of aromatic C−H stretching vibration of pyridine moiety. The band corresponding to the C=N stretching vibration appears at 1629 cm −1 in the ligand, which shifts to higher wavelength at 1644 cm −1 in the complexes indicating imine nitrogen is coordinated to metal atom. Complex 1 shows strong sharp absorption band at 2062 cm −1 corresponds to the asymmetric stretching mode of monodentate azide ligand ( Figure S8). Free azide ion in NaN3 shows single strong absorption centred at 2133 cm −1 . 71 The shift to lower wave numbers indicates the coordination of azide ion to Cu(II) in the complex. Potassium thiocyanate shows a very strong band at 2048 cm −1 in i.r spectrum, which is caused by the CN stretching vibration. However, in complex 2 this band is shifted to higher frequency at ∼2100 cm −1 suggesting coordination of the thiocyanate group to the metal centre ( Figure S8).

Electronic spectra
The absorption spectra of the ligand (HL1), and the complexes 1 and 2, were recorded at a concentration of 10 −4 M in CH3OH, as the complexes show good solubility in CH3OH. The spectrum of the ligand exhibits two absorption bands at 287 and 333 nm attributable to the π→ π* transition associated with the azomethine chromophore and n→π* transitions involving the phenolic oxygen respectively. In the case of the complexes 1 and 2, the n→π* transition was shifted to a much greater extent, from 333 to 384 or 388 nm; this indicates the phenolic oxygen had lost a hydrogen on coordination to the metal centre. The broad transition around 615 and 634 nm for complexes 1 and 2 respectively corresponds to a d-d transition of the Cu(II) centre (Figure 3). 72 (a) (b) (c) (d) Figure 3 Electronic spectra of the complexes 1 (a and b) and 2 (c and d).

Cyclic voltammetry studies:
In cyclic voltammetry (CV) experiments, on the negative side of the reference electrode, complex 1 shows two overlapping reductive peaks at −0.66 V and −0.87 V. On scan reversal a broad

Please do not adjust margins
Please do not adjust margins anodic wave spanning −0.4 to −0.6 V was observed (Figure 4). In the differential pulse voltammetry (DPV) two well resolved peaks at −0.55 V and −0.68 V were observed. The tail of the second reductive wave extends beyond −1.0 V, but no peak could be resolved between −0.8 to −1.0 V. We believe the CV peak at −0.66 V corresponds to DPV peaks at −0.55 V and −0.68 V and they denote Cu(II)/Cu(I) and imine reductions respectively (note that β-LUMO is predominantly metal centered while α-LUMO has appreciable contribution from imine part of the ligand, vide infra). The Zn(II) complex of the ligand shows a reductive DPV peaks at −0.66 V and this supports the above assignment ( Figure S9). The CV peak at −0.87 V probably corresponds to Cu(I)/Cu(0) reduction, which is buried under the tail of the ligand reduction peak in DPV.
In   EPR spectra is an important tool to study the electronic and geometrical features of Cu(II) complexes. We have recorded the epr spectrum of complex 2, as a representative example, in DMF solution at 77K (Figure 5).

DNA binding study 3.7.1 Absorption spectroscopic studies
The  Please do not adjust margins Please do not adjust margins bonding are usually important components of this binding process, and provide stabilization. (iii) Intercalative binding / Intercalation: this association involves the insertion of planar fused aromatic ring system between the DNA base pairs, leading to significant π-electron overlap. The study of the binding capacity of the metal complexes to DNA double helix may be studied by electronic absorption spectroscopy, which follows the changes in the absorbance and shift in wavelength (Figure 6). The variation of absorption spectra with the addition of CT-DNA to the solution of the complexes was followed. It was seen that on binding of DNA by the complexes, the UV absorbance at 270 nm increases, this absorbance increase is referred to as a "hyperchromic shift" or the hyperchromic effect. The characteristic feature of hyperchromic shift is the extent of increase of the absorbance of the ligand centered band. 73 With increase of concentration of CT-DNA, both the complexes showed a hyperchromic and a slight blue-shifted LC band in the absorption spectra. The hyperchromic values observed in the presence of DNA were in the range 17.4-57.2%, and their blue shifts are by 3-4 nm ( Table 3). The intrinsic binding constant Kb calculated from the plots of [DNA]/(εa-εf) verses [DNA] can be obtained from the ratio of the slope to the intercept. Using the hyperchromism at 266 and 270 nm the intrinsic binding constant Kb for DNA binding by the complexes 1 and 2 were determined to be 5.4 × 10 8 and 1.08 × 10 5 M −1 respectively. The exceptionally high value of Kb for 1, which is relatively higher than the classical intercalator (ethidium bromide; 1.4 × 10 6 M −1 ), suggests the complex is an avid binder of DNA. Since such high binding constant is generally characteristic of DNA binding by intercalation, so we assume that the complexes 1 and 2 bind DNA predominantly by intercalation, though Hbonding, electrostatic interactions as well as interactions with the phosphate groups of DNA may also lend their contribution towards overall binding strength.

Fluorescence binding study
Ethidium bromide is a planar cationic dye which is widely used as a sensitive fluorescence probe for native DNA. Its fluorescence intensity is very weak, but it is greatly increased when EB is specifically intercalated into the base pairs of DNA. 74,75 Furthermore, this fluorescence is quenched by the addition of another molecule that displaces EB from DNA. 54 It was found that the fluorescence intensity of CT-DNA-EB decreased remarkably with the addition of the complexes 1 and 2 (Figure 7), which indicated that the complexes can bind DNA and replace EB from the CT-DNA-EB system. 76 The A kinetic treatment on the basis of the Michaelis-Menten approach was applied and the various kinetic parameters Vmax, KM and kcat were evaluated from Lineweaver-Burk doublereciprocal plots. 78 The observed rate versus [substrate] plot in methanol solution as well as Lineweaver-Burk plot are given in Figure 8. The kinetic parameters are listed in Table 4. The turnover number (kcat) at 898 h -1 is reasonably good.    A plausible mechanism for the catalysis is given in scheme II. It is suggested that DTBCH2 displaced the N3from the coordination sphere and coordinates as DTBC 2in a bidentate fashion. One electron transfer from DTBC 2converts it to a semiquinone complex having a Cu I core. The Cu I complex rapidly reacts with aerial O2 to give a Cu II -superoxo complex, containing the semiquinone ligand.
Elctron transfer from the semiquionone to superoxide results in dissociation of the quinone and conversion of superoxide to hydroperoxide. This is followed by elimination of H2O2 and the catalyst is regenerated. 21,26,79,80 The detection of H2O2 from the reaction mixture provides evidence in favour of the proposed mechanism. ESI-MS spectrum (positive ion mode) of a 1:100 mixture of 1 and 3,5-DTBCH2, recorded after 20 min of mixing of the reactants in methanol, shows that apart from the parent complex peak at 300.08, a peak with a m/z value of 543.25 which is consistent with the formula of [{Cu(L1)(3,5-DTBC 2-)}+Na] + (Figure S10 in SI), which thus adds further evidence in favour of the proposed mechanism. It may be noted that according to the mechanism proposed in scheme II, the Cu(II) complex must have a square planar geometry with a labile ligand in the square plane, for it to be catalytically active. This may be the reason why complex 2, having a chain like structure with 4+1 coordination of the Cu(II) center is inactive as a catalyst. Comparing the catalytic efficiency of

DFT calculation results
DFT calculations were performed on the complexes and the frontier orbitals of 1 are depicted in Figure 9 and while those of 2 are shown in Figure S11 in Supplementary material. For complex 1 the SOMO (α-HOMO) has only 3.17% contribution from Cu while for complex 2 (monomer) the SOMO has only 0.74% Cu contribution. The β-LUMO for complexes 1 and 2 has 48.37% and 53.38% contribution respectively from Cu. However, the unpaired spin density is predominantly centred on the Cu-atom with the contribution of the Cu atom being 48.37% and 51.51% for complexes 1 and 2 respectively (Figure 10

Conclusion.
Cu(II) complexes bearing a pyridoxal appended tridentate Schiff base ligand and azide or thiocyanate coligand are reported in this communication. It is shown that the solid state structures are different in the two complexes, with azide coordinated complex containing monomeric square planar Cu(II), whereas the thiocyanato complex has square pyramidal Cu(II) environment and it exists as a one dimensional chain in the solid state. The two compounds show different reactivity as catalyst for aerial oxidation of catechol (catecholase activity), with complex 1 showing good catalytic efficiency while complex 2 is ineffective as catalyst. The differences in the structures of these two complexes as well as the instability of the Cu(I) species in 2 compared to 1, as reflected in their electrochemical studies may be the reason for this difference in catalytic activity. Complex 1 shows much stronger DNA-binding ability than complex 2.