Solution equilibrium, structural and cytotoxicity studies on Ru(η6-p-cymene) and copper complexes of pyrazolyl thiosemicarbazones

Solution chemical properties of two bidentate pyrazolyl thiosemicarbazones 2-((3-methyl-1-phenyl-1H-pyrazol-4-yl)methylene)hydrazinecarbothioamide (Me-pyrTSC), 2-((1,3-diphenyl-1H-pyrazol-4-yl)methylene)hydrazinecarbothioamide (Ph-pyrTSC), stability of their Cu(II) and Ru(  6 - p -cymene) complexes were characterized in aqueous solution (with 30% DMSO) by the combined use of UV-visible spectrophotometry, 1 H NMR spectroscopy and electrospray neutral pH. Me-pyrTSC and Ph-pyrTSC exhibited moderate cytotoxicity against human colonic adenocarcinoma cell lines (IC 50 = 33-76  M), while their complexation with Ru(  6 - p cymene) (IC 50 = 11-24  M) and especially Cu(II) (IC 50 = 3-6  M) resulted in higher cytotoxicity. Cu(II) complexes of the tested thiosemicarbazones were also cytotoxic in three breast cancer and in a hepatocellular carcinoma cell line. No reactive oxygen species production was detected and the relatively high catalase activity of SUM159 breast cancer cells was decreased upon addition of the ligands and the complexes. In the latter cell line the tested compounds interfered with the glutathione synthesis as they decreased the concentration of this cellular reductant. medium (DMEM) for HepG2 and breast cancer cell lines supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L -glutamine, 1 mM sodium pyruvate and 100 mM HEPES. The cells were incubated at 37°C, in a 5% CO 2 , 95% air atmosphere. All cell lines were detached with Trypsin-Versene (EDTA) solution for 5 min at 37 °C. MTT and EZ4U assays: Me-pyrTSC, Ph-pyrTSC, Ph-pyrSC and their copper(II) and Ru(  6 - p -cymene) complexes were dissolved in a 90% (v/v) DMSO/H 2 O mixture first using 10 mM ligand and 0, 5 or 10 mM metal ion concentrations. The metal salts without ligands were also tested. Cisplatin (Teva) was used as a positive control. Then stock solutions were diluted in complete culture medium, and two-fold serial dilutions of compounds were prepared in 100  L of RPMI 1640, horizontally. The semi-adherent colonic adenocarcinoma cells were treated with Trypsin-Versene (EDTA) solution. They were adjusted to a density of 1×10 4 cells in 100  L of RPMI 1640 medium, and were added to each well, with the exception of the medium control wells. The final volume of the wells containing compounds and cells was 200  L. The culture plates (Colo205, Colo320, MRC-5) were incubated at 37°C for 72 h; at the end of the incubation period, 20  L of MTT solution (from a stock solution of 5 mg/mL) were added to each well. After incubation at 37°C for 4 h, 100  L of SDS solution (10% in 0.01 M HCI) were added to each well and the plates were further incubated at 37˚C Solution stability of anticancer Ru(II)(  6 - p -cymene) and Cu(II) complexes of two bidentate pyrazolyl thiosemicarbazones (TSC) was determined, and three structures were resolved by X-ray diffraction. [(Ru(  6 - p -cymene)) 2 (TSC) 2 ] 2+ and [Cu(TSC) 2 ] complexes are the dominant species at neutral pH. The metal complexes are more cytotoxic than the ligands. Graphical abstract


Introduction
Thiosemicarbazones (TSCs) are versatile compounds and known for their wide pharmacological activity including anticancer properties [1,2]. Triapine (3-aminopyridine-2carboxaldehyde thiosemicarbazone) is the best-known representative of this family and was extensively investigated in numerous clinical phase I and II trials in mono or combination therapies [3,4]. Two new promising TSCs, namely N'- (6,7-

dihydroquinolin-8(5H)-ylidene)-4-
(pyridin-2-yl)piperazine-1-carbothiohydrazide (COTI-2, an orally available third generation TSC) and di-2-pyridylketone-4-cyclohexyl-4-methyl-3-thiosemicarbazone (DpC) also entered human clinical studies in the last years [5,6]. These compounds belong to the family of -Npyridyl TSCs, therefore are considered as tridentate ligands with a typical (N pyridyl ,N,S) coordination mode with strong affinity towards various metal ions including Fe(II/III) and Cu(II) [7]. The anticancer properties of Triapine are connected to the inhibition of the ironcontaining ribonucleotide reductase enzyme and the formation of redox-active iron complexes in the cells is suggested to be a crucial step of the mechanism of action [8]. On the other hand, other essential metal ions (especially copper) besides iron have been associated with the mechanism of action of TSCs [9][10][11]. -N-pyridyland salicylaldehyde-type TSCs are able to Although solution speciation of numerous Cu(II) and Fe(II)/III) complexes of -N-pyridyland salicylaldehyde-type TSCs was determined in our previous works [12][13][14]24], no available stability data are published for the complexes of bidentate TSCs so far. In this work our aim was to reveal the differences of the complex formation with Cu(II) ions between two bidentate pyrazolyl TSCs Me-pyrTSC, Ph-pyrTSC (Chart 1) and tridentate TSCs, and to characterize the complexation of these ligands with the half-sandwich organometallic [Ru( 6 -p-cymene)(H 2 O) 3 ] 2+ . The bis-ligand Pd(II) complexes of them were reported to be rather cytotoxic against MCF7 breast cancer cell lines (IC 50 = 0.57-1.24 M) [25], but no cytotoxicity data were published for their Cu(II) and Ru( 6 -p-cymene) complexes. Therefore, solution speciation, solid phase structures of Cu(II) and Ru( 6 -p-cymene) complexes of the selected bidentate pyrazolyl TSCs were investigated by different methods such as UV-visible (UV-vis) spectrophotometry, 1 H , 13 C NMR and electron paramagnetic resonance (EPR) spectroscopy, electrospray ionization mass spectrometry (ESI-MS) and Xray crystallography. Additionally, anticancer activity via cytotoxicity assays, ROS production, catalase activity and L-glutathione (GSH) levels were also monitored. for Me-pyrTSC and 23% for Ph-pyrTSC complexes, respectively). The isolated complexes were characterized by ESI-MS, 1 H and 13 C NMR spectroscopy and UV-vis spectrophotometry; see experimental data and spectra (Figures S1-S3) in SI.
[Cu(Me-pyrTSCH -1 ) 2   Numerical (for crystal I) or empirical (for crystal II and III) absorption correction [29] were carried out using the program CrystalClear [30]. Sir2014 [31] and SHELXL [32] under WinGX software [33] were used for structure solution and refinement, respectively. The structures were solved by direct methods. The models were refined by full-matrix least squares on F 2 .
Refinement of non-hydrogen atoms was carried out with anisotropic temperature factors.
Hydrogen atoms were placed into geometric positions (except for water hydrogens). They were included in structure factor calculations but they were not refined. The isotropic displacement J o u r n a l P r e -p r o o f 7 the platelet crystal shape prevented the collection of high quality data. The summary of data collection and refinement parameters are collected in Table S2. Selected bond lengths and angles of compounds were calculated by PLATON software [34]. The graphical representation and the edition of CIF files were done by Mercury [35] and PublCif [36]  Samples and cells were prepared in the similar way as in case of the traditional MTT assay.
After the 24 h incubation the compounds were diluted in a volume of 100 µL medium, then the culture plates were incubated at 37 C for 24 h; at the end of the incubation period, 20 µL of MTT solution was added to each well and incubated for 2 h incubation. The cell growth was determined using both methods by measuring the optical density (OD) at 450 nm (ref.
620 nm) with a Multiscan EX ELISA reader. Inhibition of the cell growth (expressed as IC 50 : inhibitory concentration that reduces by 50% the growth of the cells exposed to the tested compounds) was determined from the sigmoid curve where 100 -((OD sample -OD medium control ) / (OD cell control -OD medium control ))×100 values were plotted against the logarithm of compound concentrations. Curves were fitted by GraphPad Prism software [42] using the sigmoidal dose-response model (comparing variable and fixed slopes).
Reactive oxygen species production assay: The ROS measurement was performed in Catalase activity and GSH level assays: For both assays cells were prepared in a same manner. The SUM159 and MCF-7 were seeded in 6-well plates at density of 5×10 5 cell/well, and were allowed 24 h to attach to the well. Then, cells were treated with 1 µM compound and left for 24 h after which they were harvested, and the dry pellet was stored at -80 C until analysis. For analyses, cells were lysed in phosphate buffered saline (PBS) by 4 freeze/thaw cycles, and total protein content was measured by Bradford method [43]. Total GSH was measured spectrophotometrically at 450 nm (ref. 620 nm) by modified Tietze method based on reduction of DTNB (Ellman's reagent) to 2-nitro-5-thiobenzoate (TNB anion) by GSH [44]. Catalase activity was assayed by measuring H 2 O 2 decomposition by catalase in the whole cell lysate by Góth method [45].

Synthesis, solid and solution phase characterization of Me-pyrTSC, Ph-pyrTSC and Ph-pyrSC
Compounds Me-pyrTSC, Ph-pyrTSC and Ph-pyrSC (Chart 1) were prepared by the condensation reaction of the corresponding 4-formyl pyrazoles (pyrazole-4-carbaldehyde) and thiosemicarbazide or semicarbazide hydrochloride as described in references [27,46], although some modifications were applied, namely microwave-assisted reactions were performed (Scheme S1). Acid-catalyzed condensation reaction of acetone (1a) or acetophenone (1b) with phenylhydrazine in refluxing ethanol led to the corresponding hydrazones (2a and 2b) in excellent yields. Subsequent cyclization and simultaneous formylation with the Vilsmeier-Haack reagent (POCl 3 /DMF) afforded 4-formyl-pyrazoles (3a and 3b) [46], which were then converted to thiosemicarbazones (Me-pyrTSC and Ph-pyrTSC) J o u r n a l P r e -p r o o f 11 and a semicarbazone (Ph-pyrSC), respectively, in the presence of acetic acid (Me-pyrTSC and Ph-pyrTSC) or sodium acetate (Ph-pyrSC) [27] in ethanol under MW irradiation. 1  molecules. The acidic hydrogen, H2N4 is coordinating to a water oxygen, and water hydrogens are connected to N2 and S1 atoms. The packing arrangement is shown in Figure S7. Selected H-bond distances and angles are collected in Table S3. Apart from hydrogen bonds a … (off-centered parallel stacking) interaction could be observed between the two five membered rings for which the distance of 3.8687(14) Å can be measured. defined previously [52,53]. CEHHEE also crystallized with one water molecule, while PIKRUX and XEBCIS crystallized without solvent inclusion in their crystal lattices. When the pyrazole rings are overplayed high flexibility of the hydrazine-carbothioamide side chain can be seen for the different crystals ( Figure S8).
Since the TSCs and their metal complexes showed much more significant cytotoxicity than the semicarbazone Ph-pyrSC (vide infra), our solution studies were focused on the behaviour of Me-pyrTSC and Ph-pyrTSC. The purity of Me-pyrTSC and Ph-pyrTSC was also checked via 1 H NMR spectroscopic measurements using maltol as internal standard. The solid ligand was dissolved in DMSO-d 6 and different amounts of maltol were added ( Figure S9), then the peak integrals of the ligand were compared to those of maltol of known concentrations. The compounds were found to be adequately pure (98.5-100%) and the exact concentrations of their stock solutions prepared on a weight-in-volume basis were calculated taking notice of these values.
The studied compounds have limited water solubility that hindered the use of pHpotentiometric titrations, therefore the proton dissociation processes of Me-pyrTSC and Ph-pyrTSC were studied by UV-vis spectrophotometry at low ligand concentration in 30% (v/v) DMSO/water solvent mixture, similarly to our former works [12][13][14]24] to obtain comparable data. In addition Bz-TSC was involved as a simple bidentate TSC model for comparison. No spectral changes were observed up to pH 10 ( Figure 2), and the appearance of a well-defined isobestic point at higher pH values demonstrates that two species (HL, L -) are involved in the equilibrium. pK a values and the spectra of the individual ligand species were calculated on the basis of deconvolution of recorded UV-vis spectra ( Table 1). The deprotonation process can be attributed to the hydrazinic-NH moiety. These TSCs have high and similar pK a values, and we can conclude that they are present in their HL form at neutral pH. These pK a values are somewhat higher than those of the -N-pyridyl TSCs such as Triapine or formaldehyde TSC [24]. 1 H NMR spectra were recorded at various pH values (see the representative example for Bz-TSC in Figure S10) and the pH-dependence of the chemical shifts confirmed the high pK a values of the ligands.  Table 1). The logD 7.4 values represent the strong lipophilic feature of the compounds. Table 1 pK a values determined by UV-vis titrations,  max , molar absorbance () and logD 7 [26,54]. In addition the hydrolysis constants were also determined in 20% (v/v) DMSO/water solvent mixture by UV-vis spectrophotometric titrations in our former work using 0.2 M KCl ionic strength [55]. Based on the reported constants (  Figure 3, which refer to the presence of only one type of hydrolysis product appearing at pH > 5. This species was identified as the -hydroxido bridged dinuclear [(Ru( 6 -p-cymene)) 2 (-OH) 3 ] + as it was suggested in the previous works [26,54].     Table 2). The complex characterized by the constant chemical shift is assumed to be a dinuclear species [(Ru( 6 -p-cymene)) 2 (L) 2 ] 2+ in which both ligands coordinate via an (N,S -) donor set while the sulphur atoms act as -bridging ligands.
The suggested structures of the various complexes are presented in Chart 2. X-ray structures of dimeric -thiolato bridged Ru( 6 -p-cymene)-TSC complexes in solid phase have been already reported in the literature [56]. Based on the 1 H NMR spectral data stepwise stability constants for the formation of the dimeric complexes from the mononuclear species could be computed (Table 2). Notably, the ratio of the peak integrals of the mononuclear and dinuclear species was not changed by varying the metal-to-ligand ratios at pH 2.7 and no other species appeared (besides the unbound organometallic cation or the free ligand at metal or ligand excess, respectively).
However, new peaks were observed at pH > 10.6 and precipitate was also formed most probably due to the formation of a neutral mixed hydroxido species [Ru( 6 -pcymene)(L)(OH)] as it was observed in case of the complexes of other bidentate ligand [57]. A low intensity doublet peak appeared at 7.2 ppm that is typical to the free pcymene representing the release of the arene ligand.
In order to further corroborate the co-existence of the mononuclear and dinuclear species, a dilution series was prepared in the concentration range 11.3-450 M at pH 7.4 and UV-vis spectra were recorded from which molar absorbance spectra were calculated ( Figure S13). As level of dissociation increases with dilution, formation of higher fraction of mononuclear species is favoured at lower concentration compared to the dimer. It was found that the spectral characteristics altered significantly upon dilution suggesting the changed ratio of the mononuclear and dimeric complexes. In addition ESI-MS spectra were recorded for the [Ru( 6 -p-cymene)(H 2 O) 3 ] 2+ -Ph-pyrTSC system at 4 equivalents of ligand excess ( Figure 5). The major species is the mononuclear complex, and some dinuclear and bis-ligand complexes were also found, although formation of the latter species could not be proved by the 1 Figure S14). The predominant species is [(Ru( 6 -p-cymene)) 2 (L) 2 ] 2+ at neutral pH. No significant differences in the solution speciation were found between the two studied TSC ligands. been determined by single crystal X-ray diffraction. Notably, the isolated complexes contain the neutral ligands coordinated. II crystallized with one chloride counter ion in monoclinic P 2 1 /a space group and its ORTEP representation is depicted in Figure 6 and the unit cell is shown in Figure S15. Disordered structures have been found with disordered positions of the coordinated chlorido ligand as well as the isopropyl group and the phenyl-pyrazol moiety of the ligand (Figure 6, S16-S18).   Table S4. Notably single crystals of [Ru( 6 -p-cymene)(Ph-pyrTSC)Cl]Cl were also obtained, however the level of disordered structures was even higher (not shown).

Solution equilibrium and structural studies of copper(II) complexes
Solution stability of Cu(II) complexes of numerous TSCs has been already characterized in our previous works [12][13][14]24]; however these compounds were tridentate ligands, namely -  Figure 7 (and in Figure S19 for Bz-TSC complexes).

N-pyridyl
Although fairly low ligand concentrations were used (10-44 M), precipitation was observed at pH > ~6 using both 1:1 and 1:2 metal-to-ligand ratios. Notably in the Cu(II) -Bz-TSC (1:2) system the precipitate occurred at somewhat higher pH (~9). Spectra were also recorded at constant pH (4.6) and constant ligand (10 M) or metal ion (10 M) concentrations varying the metal-to-ligand ratios (1:5-5:1).  Figure 7). These data reveal the considerably higher stability of the pyrazolyl complexes compared to that of Bz-TSC species. On the other hand all these bidentate TSCs preferably form bis-ligand complexes, which are neutral species with rather low water solubility.
Significant fraction of bis complexes are present in the solution even at 1:1 ratio (Figure 8.a).
On the contrary the salicylaldehyde TSCs form exclusively mono complexes [14]. However, bis and dinuclear complexes are also formed with -N-pyridyl TSCs besides the predominating mono complexes, but only at ligand excess, and at 1:1 ratio they do not appear  Table 3 Overall stability constants (log) of the Cu(II) complexes formed with the studied ligands for

In vitro cytotoxicity and antioxidant properties
In   Since the studied TSCs were highly synergistic with CuCl 2 , and the higher cytotoxic activity of certain TSCs is associated with induction of reactive oxygen species [17,18] Table 5. Similarly to the Colo205 and Colo320 cell lines, the ligands exhibited undoubtedly synergistic effect with the Cu(II) ions. The Cu(II) salt and the ligands were less toxic against these cell lines compared to Colo205/320. The complexation resulted in higher cytotoxicity in the SkBr3 and SUM159 cells.   Table S7).   Figure   10.a. Different effect was observed for the two kinds of cell lines, namely the ligands and the Cu(II) complexes revealed similar and low catalase activity in MCF7 cells, while increased catalase activity was seen in SUM159 cells. However, comparing the behaviour of the compounds to solvent control, a decreased activity was detected in the latter cells. Notably, the compounds were more cytotoxic in SUM159 than in MCF7 cells (see IC 50 values in Table 5), thus the increased catalase activity did not protect the SUM159 cells against the toxicity of the tested compounds. Disturbances in GSH homeostasis are often involved in progression of cancer [58]. The decrease in the GSH level can lead to an increased susceptibility to oxidative stress, while the elevated GSH concentration increases the antioxidant capacity and can consequently increase the resistance to oxidative stress. Based on the data obtained for the GSH levels ( Figure 10.b) the monitored two untreated cell lines exhibited similar values.
Addition of both the solvent DMSO mixture and CuCl 2 resulted in increased GSH content in SUM159 cells, and they did not affect the GSH level in MCF7. The tested compounds (Me-pyrTSC, Ph-pyrTSC and their Cu(II) complexes) had different effects, since the GSH level was decreased in SUM159 cells, while increased GSH concentration was measured in the MCF7 cells compared to the controls. No significant changes were seen upon addition of Cu(II) to the ligands.
These results suggest that the tested compounds may cause changes in antioxidant transcription factor Nrf2, which is responsible for transcription of enzymes needed for GSH synthesis [59] independently of ROS production. Certainly, as these two cell lines represent breast cancer of different malignancy with different metabolic capacities, it would be interesting to investigate in the near future how these differences occur in the light of therapy resistance.

Conclusions
J o u r n a l P r e -p r o o f 28 NH moiety takes place only in the basic pH range (pK a : 11.53 (Me-pyrTSC), 11.78 (Ph-pyrTSC)). This feature also contributes to the strongly lipophilic character of the ligands. The stoichiometry and stability of the Cu(II) and Ru( 6 -p-cymene) complexes were studied in 30% (v/v) DMSO/H 2 O solvent mixture with a focus on the most plausible species that emerged at physiological pH. For the solution speciation studies mainly UV-vis spectrophotometry was used and 1 H NMR spectroscopy and ESI-MS were also applied for the Ru( 6 -p-cymene) complexes. The complex formation with Cu(II) was found to be fast, while diminished upon addition of the tested compounds. The compounds resulted in a decreased GSH level in SUM159 cells, while higher GSH concentration was seen in the MCF7 cells comparing to the controls. It suggests that the studied compounds interfere with the GSH synthesis without ROS production. However, the Cu(II) complexation did not bring differences in the GSH levels.