Antiproliferative activity of platinum(II) and copper(II) complexes containing novel biquinoxaline ligands

Abstract Nowadays, cancer represents one of the major causes of death in humans worldwide, which renders the quest for new and improved antineoplastic agents to become an urgent issue in the field of biomedicine and human health. The present research focuses on the synthesis of 2,3,2ʹ,3ʹ-tetra(pyridin-2-yl)-6,6ʹ-biquinoxaline) and (2,3,2ʹ,3ʹ-tetra(thiophen-2-yl)-6,6ʹ-biquinoxaline) containing copper(II) and platinum(II) compounds as prodrug candidates. The binding interaction of these compounds with calf thymus DNA (CT-DNA) and human serum albumin were assessed with UV titration, thermal decomposition, viscometric, and fluorometric methods. The thermodynamical parameters and the temperature-dependent binding constant (Kʹb) values point out to spontaneous interactions between the complexes and CT-DNA via the van der Waals interactions and/or hydrogen bonding, except Cu(ttbq)Cl2 for which electrostatic interaction was proposed. The antitumor activity of the complexes against several human glioblastomata, lung, breast, cervix, and prostate cell lines were investigated by examining cell viability, oxidative stress, apoptosis-terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling, in vitro migration and invasion, in vitro-comet DNA damage, and plasmid DNA interaction assays. The U87 and HeLa cells were investigated as the cancer cells most sensitive to our complexes. The exerted cytotoxic effect of complexes was attributed to the formation of the reactive oxygen species in vitro. It is clearly demonstrated that Cu(ttbq)Cl2, Pt(ttbq)Cl2, and Pt(tpbq)Cl2 have the highest DNA degradation potential and anticancer effect among the tested complexes by leading apoptosis. The wound healing and invasion analysis results also supported the higher anticancer activity of these two compounds.


Introduction
Chemotherapy, one of the alternatives that exists to fight cancer, relies on the discovery of chemicals that can kill rapidly dividing cells, thereby slowing and stopping the growth and spread of cancer. 1 For decades, metal-based anticancer drugs have remained of great interest as a potential therapeutic agent against treatments 2 or to detect disease. 3It goes without saying that a high level of success in this field owes itself to the application of cisplatin in hard tumor cases, which kills harmful cells and clears "the body of oncogenes."Known as "cancer penicillin" 4 because of its high potential to treat various forms of this disease, the use of cisplatin in chemotherapy has proven effective in the treatment of testicular cancer-among others-as it can rapidly alter the prognosis and cure up to 80% of cases within the early stages. 5However, despite this success, the use of cisplatin faces serious side effects that limit the dose administered to patients.In addition, this substance does not have sufficient efficacy in other cancer types, and tumors can often develop resistance to the drug over time. 6These limitations encourage experts to keep searching for metal-based drug alternatives to fight cancer by mimicking cisplatin which are now widely used in treatment and play a major role in medical tumors. 7 -12Another strategy used to increase the efficacy of the antinoplastic agents is the use of chelating ligands.Combined chelated copper-based compounds, which are good candidates for this purpose, have been shown to possess improved antineoplastic properties compared to cisplatin when it comes to laboratory and clinical interventions. 13 , 14As a result, there are 2-and 3-Clip-Phen, which contain two phenanthroline ligands linked by a serinol bridge at the second or third position to better explain the coordination of the two phenanthroline units around the same copper ion.The oxidative nuclease processes of these compounds and molecules are far more active than phenanthroline.Clip-Phen compounds have been reported to have enhanced activity on DNA cleavage. 15n the light of the previously mentioned information, quinoxaline ligands were used in this study to obtain new metal-based anticancer compounds due to their increasing use nowadays. 16 -18his effort could lead to numerous biological and therapeutic uses for quinoxaline or benzopyrazines, including their byproducts. 19 , 20As for quinoxaline cores, they have the ability to fight cancer, thus making them key ingredients in drugs for this purpose. 21For example, a new set of 2-alkylcarbonyl and 2benzoyl-3-trifluoromethylquinoxaline-1,4-di-N-oxide derivatives was developed and evaluated for its antitumor properties against in vitro MCF7 (breast), NCIH 460 (lung), and SF-268 (CNS) models. 22ome previous research has already shed light on the effectiveness of pyridine platinum(II) compounds mimicked by cisplatin.These studies included peridyl-quinoxaline as a metal-based agent. 23Additionally, there is ample evidence to suggest that the interactions between platinum(II) compounds and quinoxaline ligands are key not only to the biological, but also the biochemical processes, thus paving the way for large-scale pharmaceutical production in recent years. 24sing the strategies summarized above, for the first time in this field as far as the literature is concerned, the present research combines two quinoxaline units (phenyl or thenyl quinoxalines) and, then, prepares their metal complexes to enhance the known anticancer effects of a single quinoxaline unit.For this purpose, the study reports on the synthesis, identification, and anticancer activity of 2,3,2 ʹ,3 ʹ-tetra(pyridin-2-yl)-6,6 ʹ-biquinoxaline (tpbq or pyridyl quinoxaline) 25 and 2,3,2 ʹ,3 ʹ-tetra(thiophen-2yl)-6,6 ʹ-biquinoxaline (ttbq or thenyl quinoxaline) ligands and their new copper(II) and platinum(II)chloride complexes with the closed formula of M(tpbq)Cl 2 and M(ttbq)Cl 2 (M: Cu(II) or Pt(II)) (see Fig. 1 ).Here, we determined the actual binding mode of the action between the compounds and DNA as well as blood transfer protein [human serum albumin (HSA)] using thermal denaturation and spectroscopic (electronic, fluorescence) and viscometric measurements.In addition, the bioefficacy of these complexes were investigated by testing the cytotoxicity, oxidative stress, apoptosis, invasion and migration inhibition capacity, and gonotoxicity potential using in vitro systems.

Platinum complexes
The synthesis of the platinum(II) complexes of tpbq and ttbq followed two steps.In the first step, 163.7 mg (0.964 mmol) AgNO 3 was directly added into a 5 ml aqueous solution of 100 mg (0.241 mmol), K 2 PtCl 4 in darkness and left overnight.In the second step, after removing the precipitated AgCl, 5 ml DMF solution of 0.136 g (0.482 mmol) (tpbq) or 0.1413 g (0.482 mmol) (ttbq) was added dropwise to the Pt-solution while stirring.Next, the solution was refluxed at 40°C for 24 h, and the precipitate of the Ptcomplexes was collected under vacuum and dried at room temperature.The percentage yields obtained for the Pt(tpbq)Cl 2 was 16.45% and 3.19% for Pt(ttbq)Cl

DFT calculations
Density functional theory calculations were performed by utilizing the hybrid functional B3LYP 26 -29 and lanl2dz basis set for the computational analysis of the metal ligand complexes as described in our previous study. 30 , 31Grimme's empirical dispersion corrections with Becke-Johnson damping were performed for all DFT calculations by using the Gaussian 09 Rev. D.01 software package. 32

Binding studies
The DNA and HSA binding studies were carried out using the methods described in our previous study. 31The details of the synthesis and the identification of the compounds can be found elsewhere. 33 -36he change in the electronic absorption spectrum of the complexes in the presence and the absence of CT-DNA was followed using an HP Agilent8453 Spectrophotometer.During the measurements, the concentration of the compounds was kept constant [Cu(tpbq)Cl 2 = 3.0 × 10 −3 M; Cu(ttbq)Cl 2 = 0.50 × 10 −3 M; Pt(tpbq)Cl 2 = 3.0 × 10 −3 M; Pt(ttbq)Cl 2 = 3.0 × 10 −3 M)], while the CT-DNA concentration was changed depending on the R (R = [DNA]/[complex]) value of 0 to 10.The stock solutions of all the complexes were prepared in DMF and, then, diluted to the desired concentrations using a 5 mM Tris HCl-50 mM NaOH buffer (1:1) at pH = 7.11 prior to the UV titrations, where the percentage ratio of DMF in the final solutions did not exceed 0.5%.The optimal incubation time was determined as 30 min for Cu(ttbq)Cl 2 , 60 min for Pt(tpbq)Cl 2 , and 45 min for Cu(tpbq)Cl 2 and Pt(ttbq)Cl 2 at 37°C spectroscopically.The intrinsic binding constant (K b ) and the thermodynamic parameters of our compounds were calculated according to the change in the intensity of the electronic absorption band observed at 375 nm for Cu(tpbq)Cl 2 , 412 nm for Cu(ttbq)Cl 2 , 376 nm for (Pt(tpbq)Cl 2 , and 421 nm for (Pt(ttbq)Cl 2 using the methods available in literature. 1 -31 , 33 , 34 Similar experiments were carried out to show the strength of binding of our complexes (K b ) to HSA and to investigate the related thermodynamic parameters in the same buffer solution as well. 31In these studies, 2.12 × 10 −5 M HSA titrated with the complexes (0 and 2.12 × 10 −6 M) and the spectral changes in the electronic absorption spectrum were monitored at 280 nm.The optimal incubation time for HSA binding titrations was determined as 30 min for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , and 60 min for Cu(tpbq) Cl 2 and Pt(tpbq)Cl 2 spectroscopically. 37he thermal decomposition of 60 μM CT-DNA in the presence of the compounds was monitored spectroscopically at 260 nm and between 30°C and 90°C with 10 to 160 μM of the complexes in a HAAKE temperature-monitored circular bath.Similarly, the change in the viscosity of 60 μM CT-DNA was studied with the help of an AND SV-10 VIBRO Viscometer by varying the concentration of the compounds between 10 and 70 μM at room temperature. 35he HSA thermal decomposition was followed at 280 nm upon increasing the temperature from 30°C to 130°C at a 1°C min −1 rate from the spectral change.In order to better examine the processes involving the interaction of metal complexes and protein structure, viscometric titration was carried out in the same manner as for the nucleic acid.
In order to obtain the linear Stern-Volmer quenching constant (K sv ), a Thermo-Scientific Lumina Fluorescence Spectrometer was used to conduct the flourometric measurements of 10 μM ethidium bromide (EtBr)-pretreated CT-DNA solutions in the presence of 0 to 160 μM compounds. 31 , 36The excitation wavelength was adjusted at 478 nm, and emissions ware observed at about 610 nm.
Tests involving fluorescence titration were conducted by maintaining the concentration of 1 μM HSA and changing the concentrations of the complexes between 0 and 50 μM in 5 mM Tris HCl buffer.The fluorescence spectral measurements were made at 200 to 500 nm after exciting the solution at 280 nm.

In vitro bioefficacy studies
All the experiments conducted in this section were performed precisely as explained in our previous study. 31The cytotoxicity, oxidative stress, apoptotic effects, and antitumor activities of the synthesized complexes were investigated by using several human glioblastomata, prostate, breast, and lung cell lines and noncancer cell lines used as control.

MTT cell viability assay
The used cell lines (A172, LN229, U87, HeLa, MDA-231, PC-3, A549, and CHO-K1) were harvested and seeded (6 × 10 3 cells/well) into the 96-well plates and incubated at 37°C and 5% CO 2 for 24 h.After that, the synthesized Cu(II) and Pt(II) complexes, at a concentration range of 6.25 to 100 μM in DMEM-F12, were added to the cell monolayers and incubated at 37°C and 5% CO 2 for 24, 48, and 72 h.The dose-response relationships of the synthesized complexes were determined by using an MTT colorimetric agent (3-[4,5dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide, #M5655, Sigma-Aldrich), which is reduced by the mitochondrial enzymes of the metabolically active live cells and converted into a purple formazan product.The MTT stock solution was later dissolved in a phosphate buffer solution (PBS, 5 mg/ml, w/v) and, then, 5 μg/ml of MTT in DMEM was added into each well and incubated at 37°C and 5% CO 2 for 4 h.DMSO (#D 8418, Sigma-Aldrich), as a dissolving solution for formazan crystals, was added into each well and placed in a shaker for 2 h.The absorbance at a test wavelength of 570 nm and a reference wavelength of 630 nm was measured on a microplate reader.

Oxidative stress testing-DCFDA assay
The reactive oxygen species (ROS) formation on the selected cell lines upon exposure to the synthesized Cu(II) and Pt(II) complexes was investigated using the oxidation-sensitive dye, 2 ʹ,7 ʹdichlorofluorescein diacetate (DCFDA, #D6883 Sigma-Aldrich).The cell lines (A172, LN229, U87, HeLa, MDA-231, PC-3, A549, and CHO-K1) were seeded (6 × 10 3 cells/well) into the 96-well plates and incubated at 37°C and 5% CO 2 for 24 h.After that, the Cu(II) and Pt(II) complexes, at a concentration range of 6.25 to 100 μM, were added to the cell monolayers.Hydrogen peroxide (H 2 O 2 ), at concentrations from 50 to 500 μM, was utilized as a positive control.The following day, 5 μM of DCFDA in PBS was added to the cell monolayers and, then, the plate was incubated for 30-45 min.Consequently, the fluorescence intensity was measured with 485 nm excitation and 535 nm emission wavelengths using a microplate reader.The fluorescence values of each treatment were normalized to the negative control, including only growth medium without any treatment.The assay was repeated in triplicate.

Tunnel assay
The sensitive U87 and HeLa cell lines, selected according to the results of the IC 50 values of the examined Cu(II) and Pt(II) complexes, were seeded at a seeding cell number of 5 × 10 4 cells on the 12 mm round cover slips placed in the 24-well plate and incubated at 37°C and 5% CO 2 for 24 h.The following day, the Cu(dpq)Cl 2 , Cu(dtq)Cl 2 , Pt(dpq)Cl 2 and Pt(dtq)Cl 2 complexes were freshly prepared and added at a range of 50 and 100 μM to the cell monolayers.After a 24 h incubation time, the cell monolayers were fixed with 4% paraformaldehyde for 15 min at room temperature and, then, the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL) method was applied to detect the DNA damage associated with apoptosis.A commercial assay kit, In Situ Cell Death Detection Kit (Roche, #11767291910), was used for this purpose in accordance to the manufacturer's instructions.The experiment was repeated twice with the tested cell lines, and images were recorded with using a Leica DMI 6000 fluorescence microscope with 40 × magnification.

Matrigel invasion analysis
The analysis was completed with U87 and HeLa cell lines to assess the metastatic potential of tumor cells using transwell inserts with an 8 μm pore size polycarbonate membrane (Corning Costar, Cambridge, MA, USA) coated with 1/5 of Matrigel (Beckton Dickinson, Bedford, MA, USA) and air-dried at room temperature.After seeding the U87 and HeLa cell lines (3 × 10 5 cells/insert), the complex concentrations were selected as 6.25 or 12.5 μM and incubated at 37°C in a humidified incubator with 5% CO 2 for 24 h.The cell monolayers in the transwell inserts were fixed with ice-cold methanol and stained with 0.05% crystal violet for the analysis.The invasion of the cells to the basolateral side of the inserts was investigated using a light microscope with 40 × magnification.

In vitro scratch wound healing
In vitro scratch wound healing experiments were carried out with HeLa cell lines (3 × 10 4 cells/insert), which were seeded into 6-well plates.When they reached 80-90% confluence, scratches were created using AutoScratch (BioTek) in a straight line.The same Cu(II) or Pt(II) complex concentrations (6.25 or 12.5 μM) were used and incubated in the presence of 5% CO 2 at 37°C in a humidified incubator for 24, 48, and 72 h.Images were obtained at the Paper | 5 beginning of the exposure (time 0), and repeated at the end of 24, 48, and 72 h incubation periods.

In vitro comet assay
Commet assay was performed with U87 and HeLa cell lines (4 × 10 4 cells/well) to detect DNA damage in the individual eukaryotic cells, where 25 or 50 μM Cu(II) or Pt(II) complexes were added to the cells and incubated for 24 h.Subsequently, the comet assay was performed as described before. 31The DNA damage was evaluated by calculating the DNA tail percentage using the following equation: DNA tail % = 100 × tailing DNA density/cell DNA density.The assay was repeated twice, and 50 individual cells were involved for each treatment.

DNA cleavage activities
The DNA helix cleavage effect of the synthesized Cu(II) or Pt(II) complexes was evaluated by determining their ability to convert the plasmid DNA in the supercoiled form (SC) to its nick circular form (NC) and linear form (LF). pBI-CMV1 plasmid (3.1 kb) was grown in Escherichia coli and, then, purified using a Machery Nagel DNA isolation kit.The Cu(II) or Pt(II) complexes, at concentrations of 100 or 200 μM in the water, were incubated with 100 ng of plasmid DNA in double-distilled water (ddH 2 O) for 16 h at ambient temperature in a total volume of 20 μl.After that, the samples under investigation were electrophoresed on 1% agarose gels by means of a tris-acetate-EDTA (TAE) buffer at 100 V for 1 h.The gel was stained using EtBr, and images of the bands were captured by a ChemiDoc imaging system (BioRad).
Statistical analyses were performed by ANOVA and Tukey's post hoc test at a P < 0.05 significance level.
All complexes and ligands are stable in DMF and buffer solutions for 24, 48, and 72 h and longer.The stability of the complexes was checked using electronic absorption spectrum before the experiments.

Synthesis and characterization
The ligands used in the present study, tpbq and ttbq, contain two quinoxaline groups attached from their 6-6 position, as depicted in Fig. 1 .The ligands and their copper(II) and platinum(II) complexes were stable at room temperature, sparingly soluble in ethanol and methanol but very soluble in DMF.The elemental analysis and mass spectrum results support the composition of the compounds.Furthermore, the change in the position of the proton signals of tpbq and ttbq in DMSO-d 6 in the 1 H-NMR spectrum of the metal complexes with metal ions was a good indication of metal ion-ligand coordination.In those spectra, the proton signals of tpbq, observed between δ: 7.29 and 8.65 ppm, deviate to δ: 7.42 and 8.34; likewise, the proton signals of ttbq at around δ: 7.24 and 8.54 ppm shift toward δ: 7.15 and 8.60 ppm upon metal coordination.
The electronic absorption spectrum of the metal complexes had two main absorption bands in DMF ( Supplementary Fig. S19).The band, which appeared at about 284-287 nm, was attributed to the π → π * charge transfer transitions. 38Furthermore, the band observed at around 373 nm for the Cu(tpbq)Cl 2 and Pt(tpbq)Cl 2 , and at 415 nm for the Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 complexes were assigned to the n→ π * transitions. 39 , 40he aromatic C-H vibration absorption peaks appeared in the FTIR spectrum of tpbq at approximately 3100, 3054, and 3003 cm −1 .Following the coordination of Cu(II) and Pt(II) to tpbq, those absorption bands shifted to 3200, 3055, and 2924 cm −1 .
The aromatic (C-H) vibrations in the FTIR spectrum of ttbq were identified as the source of the absorption peaks, which appeared at approximately 3095, 3062, and 2921 cm −1 .Here, the ν (C-24) vibrations shifted to 3053, 2924, and 2856 cm −1 due to metal-ttbq coordination.Similarly, for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , the aromatic ν (C = N) vibration frequency at 1600 cm −1 was observed at 1607 cm −1 , and the ν (ph) vibration frequency of ttbq, observed at roughly 1477 cm −1 , appeared at 1473 cm −1 .The absorption peak at around 833 cm −1 was ascribed to the ν (C-S-C ring) for ttbq and was seen almost at the same frequency (835 cm −1 ) for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , revealing no bond formation between metal and sulfur atom.
Additionally, the Raman spectrum of the complexes was taken in order to clarify the structure of our complexes by determining possible band vibrations between the metal and -N, -S or -Cl atoms.The vibrational modes, appearing at around 112 and 118 cm −1 in the Raman spectrum, were assigned to ν (Cl-Pt-Cl) stretching for all Pt complexes.The Cl-Cu-Cl absorption, on the other hand, was observed at about 113-186 cm −1 .These vibrations clearly indicated direct coordination of chlorine atom with the metal ions.Similarly, the ν (N-Cl) band vibrations were obtained at around 340-381 cm −1 for both the tpbq and ttbq complexes.The Cl-M-N vibrations were detected at around 200-220 cm −1 , while the absorption bands of the M-N vibrations for symmetric and asymmetric modes were seen at around 394 to 528 cm −1 .Likewise, the N-M-N vibrational modes were obtained at around 114 cm −1 for the Cu(II) complexes and at about 240 cm −1 for the Pt(II) complexes.The data suggested that all the ligands coordinated with the metal ions through the N-atom of quinoxaline units, and that the chlorine atoms interacted with the metal centers directly in the inner shell of the complexes.It was very interesting to observe that no absorption band appeared at around 265 cm −1 for the Cu(ttbq)Cl 2 and at about 280 cm −1 for the Pt(ttbq)Cl 2 complexes.The absence of these bands could be attributed to Cu-S and Pt-S vibrations, respectively.These results clearly showed that the thenyl-quinoxaline ligands were not coordinated with the metal ions through S-atoms in our complexes.Similar results were obtained from the FTIR measurements.
In the complexes mentioned earlier, Pt 2 + interacts with the chlorine atoms through a longer distance compared to Cu 2 + .The Pt-N distances were calculated as 2.03 and 2.04 Å, while the Cu-N distances were found to be 2.10 and 2.02 Å.The optimized geometry calculations for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , on the other hand, exhibited coordination between one of the H and N atom of the quinoxaline unit and the metal ions.The chlorine atoms filled the empty coordination sites as depicted in Fig. 2 .The interaction of the metal ions with the H-atom led the deviation from planarity, which was also proved by a high N1-Cl4-N2-Cl3 dihedral of 59.8°and 59.5°for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , respectively.For both complexes, the M 2 + -N and M 2 + -H distances were found  to be almost equal, while longer metal ion-Cl distances were observed for the platinum complex than that of the copper complex (Table 1 ).It is evident from Fig. 2 , no metal-sulfur atom coordination is seen in the optimized structure of Cu(ttbq)Cl 2 , and Pt(ttbq)Cl 2 .This was also parallel with the experimental findings.

DNA-binding studies
Electronic absorption titration was performed to examine the type of the binding interaction of the complexes to the DNAcovalent or non-covalent.For this purpose, the electronic absorption spectra of the complexes were compared in the existence and the absence of CT-DNA.Cu(II) and Pt(II) complexes exhibited n→ π * transitions 39 , 40 at around 373 and 415 nm in the UV region.Upon increasing the concentration of CT-DNA, a hyperchromic change in the band intensity was seen, as illustrated in Fig. 3 A and Supplementary Figs.S20a-S22a, while the peak position remained constant.Given that hyperchromism is often correlated with non-covalent binding, 41 such a rise in the intensity of the electronic absorption band was most probably representative of electrostatic interactions occurring between the compound cations and the phosphate groups in the CT-DNA duplex 42 after the chloride ions had been freed from the outer sphere.The intrinsic binding constants (K b ) for the present complexes were evaluated based on the spectral changes taking place in the compounds (Table 2 ; Fig. 3 A, inset).The Pt(II) and Cu(II) complexes having tpbq show three times higher affinity toward CT-DNA compared to those with ttbq.This might be attributed to the different donation ability of pyridine-and thienyl-containing quinoxaline ligands to the metal ions.It was also observed that the Cu(II) tpbq and ttbq complexes maintained stronger interactions with the CT-DNA than those of Pt(II) (Table 2 ).These K b values fall even lower compared to the value for commonly used intercalators, 43 -46 likely suggesting the presence of a rather weak non-covalent interaction between the compounds and CT-DNA.
To confirm the type of binding affinity of the complexes to CT-DNA, temperature-dependent binding constants and thermodynamic parameters were also calculated upon varying the temperature between 310 and 340 K and using spectroscopic findings (Table 2 ).Accordingly, the negative value for the standard Gibbs free energy change represents the spontaneous interaction of the complexes with the double helix of DNA.Furthermore, the negative H o and S o values of the Pt(tpbq)Cl 2 , Pt(ttbq)Cl 2 , and Cu(tpbq)Cl 2 suggest their van der Waals interaction and/or hydrogen bonding, whereas the positive S o with the negative H o value of Cu(ttbq)Cl 2 is attributed to an electrostatic interaction with CT-DNA. 47n the present study, several spectroscopic and viscosometric techniques were used to demonstrate the degree of DNA binding affinity of Cu(II) and Pt(II) biquinoxaline compounds.The intrinsic binding constant (K b ) values were calculated based on the changes in the spectrum of the compounds upon the addition of CT-DNA (Fig. 3 A, inset) and in accordance to Equation 2 presented in the experimental section.The Pt(II) and Cu(II) complexes with tpbq exhibited 3-fold greater affinity for CT-DNA than those containing ttbq.This might be attributed to the different donating ability of the pyridine-and thienyl-containing quinoxaline ligands to the metal ions.It was also observed that the Cu(II) tpbq and ttbq complexes maintain stronger interactions with the CT-DNA than those of Pt(II) (Table 1 ).These K b values are even lower compared to commonly used intercalators, 43 -46 suggesting the presence of a rather weak non-covalent interaction between the compounds and CT-DNA.
The temperature-dependent binding constants were measured spectroscopically to confirm the type of binding affinity for  47 The CT-DNA was examined for viscometric behaviors in the presence of the Cu(II) and Pt(II) compounds, and the obtained values pose obvious questions concerning the nature of these processes; for instance, a rise in the DNA viscosity due to the binders can be explained by the intercalative nature of the effect as the DNA helix begins to elongate. 48As shown in Fig. 3 B and Supplementary Figs.S20b-S22b, the viscosity and the slope of the plots related to CT-DNA slightly increase upon the action of the compounds, thus pointing to the electrostatic nature of the groove binding here.The values obtained for the relative viscosity slopes ranged from 0.0266 to 0 0.0443 but, in the case of the intercalators, this value increased up to 1. 49 , 50 In the next stage, the conformational changes in the DNA helix were investigated according to the fluctuations in the electronic absorption band of the DNA at 260 nm in the presence of the compounds between 30°C and 90°C in a Tris HCl buffer.The melting temperature, T m , of DNA was measured based on the data collected during the denaturation of the double helix structure to a single-strand form by heating.The magnitude of the T m value of DNA in the presence of any binder provides additional clues as to the nature of the binding. 51In light of these observations, an increase in T m represents either an intercalative or a phosphate binding, while a decline in the temperature represents basebinding. 52Intercalation binding reinforces double helix formation, causing T m to increase by 5°C to 8°C; conversely, non-intercalative binding does not cause a significant increase in T m . 53In the same manner, the denaturation detected on the DNA duplex triggered by the Cu(II) and Pt(II) compounds caused only minor changes in the T m of DNA (see Fig. 3 C and Supplementary Figs.S20c-S22c).The extent of change in the DNA melting temperature ( T m ) in the presence of the complexes did not exceed 1.5°C, implying electrostatic interaction with the phosphate groups within the grooves. 48 , 54 , 55o shed light on the nature of the bonding in the metalbased compounds, additional fluorometric analyses were performed.To this end, we used EtBr-pretreated CT-DNA to determine the binding affinity between the compounds and DNA, later sketching a Stern-Volmer plot according to Fig. 3   Supplementary Figs.S20d-22d.The resulting quenching constants K SV for each compound ranged from 0.0082 to 0.0143 (Table 1 ).The K SV values for CT-DNA also highlight electrostatic interaction via groove binding 56 , 34 as observed in the previously obtained spectroscopic and viscometric results.Although the Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 complexes have a stronger binding affinity compared to their tpbq counterparts, all complexes

HSA-binding studies
The binding affinity of the compounds toward a plasma transport protein HSA was examined using EAS.For this purpose, the spectral changes were followed at the representative absorption band of the protein (280 nm) during UV titration, which was performed via a fixed value HSA (2.12 × 10 −5 M) with varying concentrations of compounds (2.12 × 10 −5 to 2.12 × 10 −4 M). 58 The decline seen within the band intensity could be attributed to the formation of a surface adduct of compound and protein.Such formation can also imply high polarity and reduced hydrophobicity around the tryptophan residue, 59 resulting in bonds between the compounds and the hydrophobic domains of the protein as well as a change in the HSA configuration. 60Conversely, an increase in absorption (hyperchromic effect) may appear if the compounds are exposed to HSA, thereby causing an adduct formation on the protein through external contact and possibly through electrostatic interactions in the protein's secondary structure.Figure 4 A and Supplementary Figs.S23a-S25a demonstrate the spectral change that takes place during the titration tests.The observed hypochromic effect on the addition of compounds may be due to the hydrophobic interactions triggered by the ππ stacking relationship between the aromatic rings in the compounds and the phenyl rings at the tryptophan, tyrosine, and phenylalanine residues in the protein binding groove. 61he intrinsic binding constant, K b , is measured at about 10 4 M −1 for all HSA-compound adducts (see Table 3 ).This result revealed the elevated affinity of all of our compounds toward HSA compared to reported complexes in other studies having binding constants between 10 5 and 10 6 . 62ests of UV titration at four different temperatures ranging from 310 to 340 K indicated the spontaneous affinity of the compounds toward HSA with a negative G o value. 63As the thermodynamic studies revealed in Table 2 , the positive value of S o and H o imply the presence of entropy-driven electrostatic coupling interactions between compounds and HSA. 64iscosity tests are another useful tool for determining the mode of action between Cu(II) and Pt(II) compounds and albumin.The electrostatic interactions between these compounds and HSA are likely to increase the relative viscosity of the mixture. 65inor changes in the related figures can be observed upon the addition of compounds that form bonds in the HSA groove. 66inding with HSA can be attributed to hydrophobic processes if there is an insignificant increase in the relative specific viscosity with increasing compound concentrations. 67Figure 4 B and Supplementary Figs.S23b-S25b show changes in the relative viscosity of HSA in the presence of compounds.A small increase Fig. 5 The curves for the Cu(II) and Pt(II) complexes using cell lines from different origins (A172, LN229, U87, A549, HeLa, MDA-MB-231, PC3, CHO-K1) obtained upon exposure for 24, 48, and 72 h using MTT cell viability assay.(0.0495-0.0657) in the slope of the graph for relative viscosity is indicative of interactions in the hydrophobic domain. 68hermal denaturation of proteins is a major challenge in their separation/repositioning, biotransformation, and biosensing as well as drug production and food production. 69Generally, when a compound binds to a protein in its native state, this causes the temperature to remain constant and the corresponding index improves.However, the thermal denaturation analyses conducted here indicate that the melting point of HSA declined from 114°C to 110°C, 111°C, 109°C, and 105°C upon the addition of Cu(tpbq)Cl 2 , Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 , respectively (Fig. 4 C, Supplementary Figs.S23c-S25c).Consequently, it can be stated that once albumin binds with our complexes, it loses its stability.
Fluorescent quenching assessments for albumin are crucial as they can reveal the interactions that HSA can have with any drug. 70Quite commonly, fluorescence emerges thanks to three inherent properties: tryptophan, tyrosine, and phenylalanine residues of HSA.Under real conditions, pure fluorescence in most proteins is mainly mediated by tryptophan alone Nonetheless, given the very low quantum yields of phenylalanine, the fluorescence of a tyrosine can be completely eliminated by ionization. 71he changes in the emission band of HSA at 346 nm were also monitored thoroughly after the compounds were administered to obtain information about how the compounds affect the structural conformation of serum albumin.On this basis, it became evident that our complexes reduced the emission intensity without any significant shifts in the peak position; thereby, suggesting that they interacted with HSA by producing non-fluorescent adducts. 72In other words, the Cu(II) and Pt(II) complexes interacted with serum albumin via the hydrophobic region located inside the protein (Fig. 4 D, Supplementary Figs.S23d-S25d). 73

Cytotoxicity of Cu(II) and Pt(II) complexes
The human cancer cell lines, glioblastoma (A172, LN229, U87), cervix (HeLa), breast (MDA-231), lung (A-549), prostate (PC-3) and a non-cancer Chinese hamster ovary CHO-K1 cell line as a control were used to determine the cytotoxicity of the Cu(tpbq)Cl 2 , Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 and Pt(ttbq)Cl 2 complexes by using an MTT assay.The dose-response relationship between the examined Cu(II) and Pt(II) complexes and the used cell lines were obtained from the percentage cell viability vs. exposed concentrations (Fig. 5 ).The 50% inhibition concentration (IC 50 ) for each compound for 24, 48, and 72 h exposure period was obtained from the dose-response curve as presented in Table 4 .Overall, Cu(ttbq)Cl 2 was found to have the most anticancer potential among the examined complexes across the tested cell lines, followed by Pt(tpbq)Cl 2 > Pt(ttbq)Cl 2 > Cu(tpbq)Cl 2 in a descending order.The cytotoxic effects of the Cu(II) and Pt(II) complexes did not increase parallel to their derivatives; in another words, the ttbq derivative of Cu(II) was more cytotoxic than the tpbq derivatives.Conversely, the tpbq derivative of Pt(II) was more cytotoxic than the ttbq derivatives.The most sensitive cancer cells upon exposure to the complexes were found to be U87 and HeLa cell lines, which were set aside for use in other assays.The MDA231 cell line was also seen to be sensitive, but only against Cu(ttbq)Cl 2 and not any other complex.As expected, the non-cancer control, CHO-K1, cell line was less responsive than the others.
In another study, novel Cu(II)-2-(2 -pyridyl) quinoxaline complexes were found more cytotoxic on human breast (MCF-7) and human embryonic kidney (HEK-293) cell lines compared to the cisplatin used as a positive control. 74The Cu-complex was also evaluated by another study group against six different cancer cell lines (HL-60 cells, PC-3M-1E8 human prostate tumor cells, BGC-832 cells, MDA cells, Bel-7402 human hepatoma cells, and Hela human cervix cancer cells), showing a considerable inhibitory rate and cytotoxic specificity. 75Another study investigated Pt(II)terpyridine complexes, which were found to be strongly cytotoxic to human cancer cell lines HCT116 (colorectal), SW480 (colon), Fig. 7 The apoptotic potency of the U87 and HeLa cells treated with the Cu(II) and Pt(II) complexes at concentrations of 50 and 100 μM for 24 h evaluated by applying TUNEL assay.The Negative control was treated with only a growth medium, and the positive control was treated with Dnase.The images were taken with 40 × magnification using a fluorescence microscope.NCI-H460 (non-small cell lung), and SiHa (cervix).Their IC 50 values ranged from 0.05 to 4.4 μM. 76In a different research, several platinum-based complexes, owning various types of ligands and revealing excellent anticancer activity, showed higher cytotoxic effects against MCF-7, A549, and HCT-116 cell lines in comparison with clinically used cisplatin and oxaliplatin. 77Overall, these previous studies and our study on Cu(II) and Pt(II) complexes demonstrate high anticancer potential against cancer cell lines from a variety of origins.

ROS production-DCFDA assay
Oxidative stress in an alive system is generally measured by using the fluorescent marker dihydrodichlorofluorescein diacetate (H(2)DCF-DA), which is de-esterified and oxidized to fluorescent DCF (2 ʹ,7 ʹ-dichlorofluorescein). 78 , 79 In order to determine DCFDA activity that increases with increased ROS generation in response to oxidative stress, ROS generation is determined as a DCFDA signal for a particular condition referring to the corresponding proliferation counts.In this study, the Cu(II) and Pt(II) complexes at concentrations from 6.25 to 100 μM were administered to the used A172, LN229, U87, A549, Hela, PC-3, and MDA-231 cell monolayers for 24 h and, then, the ROS formation was measured using a DCFDA fluorescence agent with a fluorescence spectrophotometer.Figure 6 shows a comparison of the level of ROS generation in the cells against the negative control including growth medium in response to the synthesized complexes.As a result of this assay, the dithenyl derivatives of Cu(II) and Pt(II) complexes generated more ROS than the dipyridyl derivatives on the used cell lines.Overall, Cu(ttbq)Cl 2 caused the highest increase in the level of ROS formation, followed by Pt(ttbq)Cl 2 and Cu(tpbq)Cl 2 , in the used cell lines.To elaborate, Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 increased ROS production 3-to 5-fold in human glioblastoma (A172, LN229, U87) and almost 2-fold in the lung A549, cervix HeLa, and breast MD-A231 cells, while no formation in prostate PC3 cells.Furthermore, Pt(tpbq)Cl 2 did not induce any ROS formation on any of the used cell lines.Human glioblastoma cell lines produced more ROS than the other used cell lines.H 2 O 2 was tested as a positive control and caused 3-fold ROS generation at 500 μM concentration.It was reported earlier in another study that the five binuclear Pt(II) complexes enhanced oxidative stress formation in fibroblasts by means of increased ROS generation and decreased antioxidant property. 80

Apoptosis-TUNEL assay
Apoptosis is one of the well-known main cell death patterns which is a cellular suicide process triggered by specific proteins. 81he effects of the synthesized Cu(II) and Pt(II) complexes on the apoptotic potential of the U87 and HeLa cell lines were

In vitro cell invasion and migration assays
Cancer cells must migrate and invade through the extracellular matrix to enter the bloodstream, attach to a distant site, and spread out of a vessel.Cell migration studies related to metastatic progression have gained importance in cancer research since this progression is considered as the main cause of death in patients.Tumor metastasis, the migration and invasion of tumor cells, is one of the biggest obstacles in anticancer treatment. 83 , 84In vitro cell invasion assay is a widely used method to provide an assessment of cell invasive capacity.In the assay used for this study, the Corning Matrigel matrix acts as an in vitro membrane that Fig. 9 An in vitro cell migration assay using a human cervix HeLa cell line after 24 h treatment with Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 at concentrations of 6.25 and 12.5 μM.The images were acquired at 0, 24, 48, and 72 h with an inverted microscope.
prevents noninvasive cells from migrating, but invading cells (malignant and nonmalignant) enzymatically degrade the matrix and spread through the membrane pores. 85he anti-invasive capacity of the synthesized Cu(II) and Pt(II) complexes on the U87 and HeLa cells across the Matrigel matrix was observed under microscope (Fig. 8 ).
The Cu(ttbq)Cl 2 and Pt(tpbq)Cl 2 complexes at concentrations of 6.25 and 12.5 μM showed high anti-invasive effects on the HeLa cell monolayer as compared with the negative control, but not on the U87 cell monolayer.As seen in Fig. 8 , there were no differences between the anti-invasive capacity of either concentrations of the Cu(ttbq)Cl 2 and Pt(tpbq)Cl 2 complexes on the HeLa cells, except Pt(tpbq)Cl 2 which was only effective at the high concentration (12.5 μM).The Cu(tpbq)Cl 2 complex did not have any anti-invasive effect on either of the cell lines (data not presented here).Similarly, Gu et al. reported that copper compounds inhibited the invasion abilities of HeLa cells, whereas, there was no significant difference found in the U87 cell line. 86he scratch wound healing assay is widely used for screening novel anticancer candidate drugs.It may provide useful information to understand how well a particular cell type can spontaneously migrate or respond to a chemo-attractant and directionally migrate toward it.As a result of the cell invasion assay, Hela was found the most sensitive cell line against the tested compounds; hence, its selection to use for the cell migration assay.The potential inhibitory effects of Cu(II) and Pt(II) complexes on the migration of HeLa are presented in Fig. 9 .
The number of migrating HeLa cells toward the scratch zone upon treatment by Cu(ttbq)Cl 2 , Pt(ttbq)Cl 2 , and Pt(tpbq)Cl 2 at concentrations of 6.25 and 12.5 μM declined during 24, 48, and 72 h treatments.However, the inhibitory effects of the tested complexes did not improve with increasing concentrations.As a result of the cell migration assay, the examined Cu(II) and Pt(II) complexes were shown to have an anti-migration potential on the HeLa cell line, and they could serve as anti-cancer drug candidates for further investigation.

Genotoxicity In vitro comet assay
To assess DNA damage, a crucial confounding factor in defining the genotoxic potential of the synthesized complexes was investigated by means of a comet assay, which precisely detected the DNA damage and revealed the genotoxic potential. 87The effective anticancer potential of the Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 complexes was determined from previous assays.Therefore, their genotoxicity at concentrations of 25 and 50 μM was examined on the HeLa and U87 cell lines using an in vitro comet assay.The results demonstrated that Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 caused DNA damage on the U87 cells at both concentrations; whereas, Cu(ttbq)Cl 2 was only genotoxic on the Hela cells at a high concentration (50 μM), while Pt(ttbq)Cl 2 was not genotoxic on either of the used cell lines.The DNA fragmentation on the U87 cell line upon Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 treatments for 24 h increased with elevated concentrations.Briefly, Cu(ttbq)Cl 2 showed more genotoxic potential than the other synthesized complexes-a finding that is supported by the results obtained for other assays in this study (Fig. 10 ).

Plasmid DNA interaction assay
The genotoxic effects of the synthesized Cu(II) and Pt(II) complexes at concentrations of 100 and 200 μM were analysed with plasmid DNA interaction assay.The migration pattern of the treated plasmid DNA with the Cu(II) and Pt(II) complexes was evaluated by agarose gel electrophoresis.The pure plasmid DNA as a negative control in the absence of any external agents was examined and the change in migration pattern was evaluated in the gel (Fig. 11 ).The DNA cleavage activities of the complexes were evaluated by determining their ability to convert the SC plasmid DNA to open circular form and linear form.The pure plasmid DNA is mainly composed of SC and NC forms.The conversion process mainly led to a decline in the intensity of the NC bands, possibly due to the accumulation of too many nicks leading to the degradation of SC-DNA.Pt(II) treatment led to plasmid DNA degradation, which was observed in the increased intensity of NC bands  Although the synthesized Cu(II) complexes did not interact with plasmid DNA in this study, previous studies have shown that other copper complexes may act by different mechanisms, such as ROS generation, which bring about oxidative cell damage and consequently trigger cancer cell death through an apoptotic mechanism. 88Indeed, in our study, synthesized Cu(II) complexes caused an increase in the level of ROS generation.
Similar results have been previously reported.For instance, novel Cu(II)-2-(2 ʹ-pyridyl) quinoxaline complexes to cause DNA cleavage and become more cytotoxic on human breast MCF-7 and human embryonic kidney HEK-293 cell lines compared to cisplatin used as a positive control. 74Likewise, another study on platinum(II)-quinoxaline compounds revealed that they interacted with calf-thymus DNA and demonstrated high levels of cytotoxic effect on L1210 murine leukemia cell lines. 89

Conclusion
In the present study, Cu(II) and Pt(II) complexes of tpbq and ttbq were synthesized and characterized.The molecular structure of the complexes was determined theoretically.Calculations at the B3LYP/LANL2DZ level of theory showed that Cu(tpbq)Cl 2 and Pt(tpbq)Cl 2 display a deviation from the square planar structure around the metal center.This deviation was significant in Cu(tpbq)Cl 2 , while slight in Pt(tpbq)Cl 2 .The optimized geometry calculations for Cu(ttbq)Cl 2 and Pt(ttbq)Cl 2 , on the other hand, exhibited a coordination between the metal center and an H-atom on the quinoxaline, as well as a nitrogen atom coordination of the same unit.The interaction of the metal ions with the H-atom led to a deviation from planarity.
The nature of the binding of the complexes on DNA and HSA were determined as electrostatic and hydrophobic interactions, respectively.
The cytotoxicity of the complexes indicated that Cu(ttbq)Cl 2 and Pt(tpbq)Cl 2 were the most effective complexes, and the U87 and HeLa cell lines were determined to be the most effected cell lines upon exposure to the four compounds.DNA damage was generated by compounds dependent on ROS formation, and the cell death pattern was identified as apoptosis for the U87 and HeLa cell lines upon the addition of Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 .
Lastly, the Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 complexes demonstrated higher potency to prevent the invasion and migration of HeLa cells.Thus, in accordance with our findings, Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 have a higher potential to serve as anticancer drug candidates and deserve further and more comprehensive studies.

Fig. 3 (
Fig. 3 (A) Electronic absorption spectra of Cu(tpbq)Cl 2 in the absence and in the presence of increasing amount of CT-DNA (0-1.25 × 10 −3 ) (inset: the plot used for calculating K b ).(B) Thermal denaturation plots obtained for Cu(tpbq)Cl 2 and CT-DNA.(C) The changes in the relative viscosity of the CT-DNA in the presence of Cu(tpbq)Cl 2 .(D) The change in the Fluorescence spectrum of EtBr-bound CT-DNA in the presence of Cu(tpbq)Cl 2 (10-160 μM) (inset: the plot used for calculating K SV ).
D and

Fig. 4 (
Fig. 4 (A) Electronic absorption spectra of Cu(tpbq)Cl 2 in the presence of HSA (2.12 × 10 5 M) (inset: the plot used for calculating K b ).(B) Thermal denaturation plots obtained for Cu(tpbq)Cl 2 and HSA.(C) The changes in the relative viscosity of the HSA in the presence of Cu(tpbq)Cl 2 .(D) The change in the fluorescence spectrum of HSA in the presence of Cu(tpbq)Cl 2 (10-50 μM) (inset: the plot used for calculating K SV ).

Fig. 6
Fig.6 The ROS generation of the used cell lines from different origins (A172, LN229, U87 A549, Hela, PC3, and MDA231) upon exposure to the Cu(II) and Pt(II) complexes for 24 h in the DCFDA assay.H 2 O 2 was used as a positive control at the indicated doses.

Fig. 8
Fig. 8 The anti-invasive effects of Cu(ttbq)Cl 2 , Pt(tpbq)Cl 2 , and Pt(ttbq)Cl 2 at concentrations of 6.25 and 12.5 μM on U87 and HeLa cells across the Matrigel during 24 h treatment by Matrigel invasion assay as compared with the negative control.The images were taken with 40 × magnification.

Fig. 10
Fig. 10 Single-cell gel electrophoresis of U87 and HeLa cell lines following treatment with the Cu(II) and Pt(II) complexes for 24 h.Negative control: growth medium; Positive control: ethyl methane sulfonate (EMS)-treated cells.Representative microscopic images of cells were taken at 40 × magnification with fluorescence microscope.

Table 1 .
Selected geometrical parameters for the metal-ligand complexes (bond lengths are in Å, bond angles and dihedrals are given in °)

Table 1 .
Accordingly, the negative value for the standard Gibbs free energy change represented the spontaneous bias toward the double helix of DNA within compounds.Furthermore, the negative H o and S o values of the Pt(tpbq)Cl 2 , Pt(ttbq)Cl 2 , and Cu(tpbq)Cl 2 suggested van der Waals and hydrogen bond interactions, whereas the positive S o with the negative H o value of Cu(ttbq)Cl 2 was linked to electrostatic interaction with DNA.

Table 4 .
TheIC 50 values of the synthesized Cu(II) and Pt(II) complexes calculated from the dose-response curves obtained using MTT cell viability assay and cell lines from a variety of origins (A172, LN229, U87 A549, Hela, MDA231, PC3 and CHO-K1) upon exposure for 24, 48, and 72 h