trans-Platinum(II) Thionate Complexes: Synthesis, Structural Characterization, and in vitro Biological Assessment as Potent Anticancer Agents

Abstract

A series of Pt(II) complexes trans-[Pt(PPh₂allyl)₂(k¹-S-SR)₂], 1, PPh₂allyl = allyldiphenylphosphine, SR = pyridine-2-thiol (Spy, 1a), 5-(trifluoromethyl)-pyridine-2-thiol (SpyCF₃-5, 1b), pyrimidine-2-thiol (SpyN, 1c), benzothiazole-2-thiol (Sbt, 1d), benzimidazole-2-thiol (Sbi, 1e), were synthesized. They were characterized by NMR, HR ESI-MS and X-ray crystallography. These complexes were treated by human cancer cell lines (A549, SKOV3, MCF-7) and shown the promising antitumor effects in comparison with cisplatin. These compounds were showed suitable selectivity between tumorigenic and non-tumorigenic (MCF-10A) cell lines. Analyses of cell cycle progression and apoptosis were conducted for 1a, the best cytotoxic compound, to screen dose/time response and to study the effects of the antiproliferative mechanism. The electrophoresis mobility shift assay was performed to assess the direct interaction of 1a with DNA and the strong genotoxic ability was indicated through comet assay method.

Graphical Abstract

To improve the therapeutic effect of platinum-based prodrugs, new trans-configured Pt(II) complexes bearing phosphane and heterocyclic thiolates are designed. They are explored as anticancer metallodrugs by screening against three human cancer cells and DNA interaction. A non-cancerous cell line is used to investigate the selectivity between cancer and the normal cell lines. Analyses of cell cycle progression and apoptosis are conducted to screen dose/time response and study about the mechanism of antiproliferative effects.

Introduction

Medicinal inorganic chemistry plays an integral role in developing metallodrugs for the treatment of cancer which is one of the leading causes of death worldwide.^[1] Platinum-based anticancer therapeutics are still among the most widely used drugs in clinical chemotherapy.^[2] Cisplatin, currently the best leading metal-based antitumor drug in the field, exhibits effective activities against a number of human malignancies, but has major limitations in its clinical applications.^[3] Therefore, continuous efforts to attain a new structure platinum-based anticancer agents with improved therapeutic index have remained intense.^[4] To design a potent metal-based therapeutic complex, it is essential to know what happens to the coordination complex under physiological conditions.^[5] DNA is considered the major pharmacological target for cisplatin and the other platinum anticancer complexes.^[6] It is well-known that the biological effectiveness of platinum complexes derives from the formation of drug-DNA cross-link adducts.^{[3c, 7]} They coordinatively interact with DNA bases via the N⁷ positions of Guanine and Adenine residues, which generate DNA lesions and subsequently cell death. The kinetic and thermodynamic interactions between platinum(II) complexes and DNA, have been widely investigated.^[8] It is generally accepted that the activity and toxicity of platinum drugs are directly attributed to the nature of the ligands around the Pt center, specifically the kinetics of exchange reactions.^[9]

Trans-configured Pt(II) complexes have been introduced as a strategy to potentially overcome the drawbacks and diminish severe side effects, drug resistance, poor selectivity and serious toxicity of cisplatin.^[10] They have displayed considerable in vitro antiproliferative effects against a wide range of cancer cells.^[11] The unique cytotoxicity profiles of trans-Pt(II) complexes with bulky planar ligands are attributed to their different structural and DNA-binding properties in comparison with cisplatin analogues.^[12] Sterically hindered Pt(II) complexes would have reduced reactivity in a substitution reaction with all potential targets, i.e., nucleophiles on DNA, proteins and small molecules. However there is some evidence that the antitumor activities of platinum(II) complexes increase in the presence of sterically hindered ligands.^[13] They can impede the substitution reactions and inhibit the undesired interactions between Pt(II) centers and cellular components before the platinum drugs bind to target DNA.^[14] Heterocyclic bulky ligands appropriately offer high steric hindrance and subsequently, thermodynamic stability and make the metal complex kinetically inert in nature.^[15] Aryl groups facilitate transportation across the cell membrane by increasing lipophilicity, thereby improving tumor uptake of drugs. Accordingly, the structure of metal-ligand complexes as well as ligand structure are thought to be significant parameters for the antitumor activity of metallodrugs.^[8]

Heterocyclic thiones or thionates are specific structural motives that combine soft and hard ends while also possessing rich coordination chemistry according to the diverse modes of action ends.^[16] Metal complexes bearing thiones or their corresponding anions have formed an expanding of research field and directing more attention to the complexes wide range of applications.^[17] Following the aim to introduce potential antitumor metal-based complexes of new structure,^{[6e, 17-18]} we have designed a class of Pt(II)-phosphane complexes containing heterocyclic thionate ligands. Further investigations have previously revealed that phosphane ligands would render the platinum complexes more stable in the physiological environment, leading to lower required drug doses and potentially reduced toxicity.^{[6e, 18c}, ^19] In addition, the presence of weakly coordinating carbon moieties in the structure of phosphane ligands can donate different steric and electronic properties to the complex, making it capable of new biological features.

In the current study, we describe the reaction of thionate ligands (pyridine-2-thiol, a; 5-(trifluoromethyl)-pyridine-2-thiol, b; pyrimidine-2-thiol, c; benzothiazole-2-thiol, d; and benzimidazole-2-thiol, e) with starting complex cis-[Pt(PPh₂allyl)₂Cl₂], A, bearing allydiphenylphosphane (PPh₂allyl) ligand. The preparation of Pt(II) thionate complexes was well confirmed by NMR spectroscopy. Structural information has been extracted by the X-ray crystallographic method. Thionates, as a source of sulfur donor ligands, coordinated to the platinum center in trans-configuration modes. To investigate the antitumor activity of the complexes, new trans-Pt(II) complexes have been evaluated against three human cancer cell lines (lung, A549, ovarian, SKOV3, and breast, MCF-7) by means of the MTT assay. Moreover, the ability of complex 1a to inhibit the proliferation of MCF-7, was assessed by measuring cell death via induction of apoptosis. Cell cycle analysis was also applied to study the mechanism of antiproliferative effects of 1a. To predict the genotoxic effect of complexes on MCF-7 cancer cells, comet assay was used. However, the interaction ability of complexes with DNA was investigated by gel mobility-shift analysis as a valuable method.

Results and Discussion

Synthesis and Characterization

Scheme 1 clearly demonstrates the general synthetic route for new complexes. The starting complex cis-[Pt(PPh₂allyl)₂Cl₂], A, was synthesized through ligand exchange, by the reaction of allydiphenylphosphane (PPh₂allyl) with known complex cis,trans-[PtCl₂(SMe₂)₂]^[20] or cis-[PtCl₂(dmso)₂]^[21]. The dimethyl sulfide (SMe₂) and dimethyl sulfoxide (dmso) displace with monodentate phosphane ligand under mild conditions. Treatment of A (in CH₂Cl₂) with ethanolic solution of potassium thionate salts (KSR) produced new platinum(II) complexes of general formula trans-[Pt(PPh₂allyl)₂(k¹-S-SR)₂], SR = deprotonated form of pyridine-2-thiol (Spy, 1a), 5-(trifluoromethyl)-pyridine-2-thiol (SpyCF₃-5, 1b), pyrimidine-2-thiol (SpyN, 1c), benzothiazole-2-thiol (Sbt, 1d) and benzimidazole-2-thiol (Sbi, 1e).

Scheme 1.

Preparation of trans-Pt(II) complexes with heterocyclic thionate ligand.

This cis-trans isomerization (from A to 1a-e) was previously detected in some Pt(II) complexes during chloride ligand replacement with bulky thiolate ligands. However, to support this observation and get more insight into this isomerization, density functional theory (DFT) was applied by the optimization of the lowest energy structures (either gas phase or CH₂Cl₂ solution, of the proposed cis and trans geometries for 1c as a case study in this calculation (Figure 1). The DFT calculations indicated that the trans isomer 1c is remarkably more stable than the suggested cis isomer 1c' in both states. In gas phase, 1c is more stable than 1c' in the amount of 23.051 kJ/mol. In CH₂Cl₂ solution, the difference between 1c and 1c' is equal to 10.787 kJ/mol (Figure 1). Additionally, the NMR spectra and X-ray crystallography (see following) were supported that the theoretical calculations correctly predicted the preferred trans configuration.

Figure 1.

The energy difference between 1c and its corresponding cis isomer 1c' in (a) gas phase and (b) CH₂Cl₂ solution.

The cis-Pt(II) phosphane precursor complex (A) together with the trans-Pt(II) complexes (1a-e) were precisely identified by analytical (elemental analysis and HR ESI-Mass) and spectroscopy methods (NMR and in the case of A and 1a-c by X-ray diffraction). The numerical data are reported in the Experimental section.

The formulae proposed for A and 1a-e were confirmed by HR ESI-Mass analysis. The ESI-Mass spectra of these complexes were recorded in the positive mode and in a dilute acetonitrile solution. The HR ESI-Mass(+) spectrum for A displayed the presence of molecular ion [M - 2CI]²⁺, whereas the HR ESI-Mass(+) spectra for 1a-e revealed the molecular ion [M + 2H]²⁺ as the most intense peak in the complexes. The NMR labeling of ligand moieties are depicted in Scheme 1, for clarifying the chemical shift assignments. The main resonances of functional groups for phosphane ligand in A are derived from ¹H NMR spectroscopy. The ³¹P{¹H} NMR spectrum of A has revealed one singlet resonance at δ = 4.6 ppm flanked by Pt sattellites (¹J_PtP= 3671 Hz) which is attributed to the PPh₂allyl moieties in cis-dipositions. This signal has been cleanly shifted to the lower fields in trans-configured Pt(II) complexes accompanied by decreasing in the coupling constant between platinum atom and phosphorus atom, for instance, in the case of complex 1c, δ = 10.8 ppm (¹J_PtP= 2766 Hz). The ¹J_PtP value is indicative of the electron-donating character of the cis- or trans-positioned ligands. The lowering of this value is good evidence to rule out the formation of the cis configuration in the thiolate complexes and is consistent with a trans feature (1a-e).^{[10d, 10e, 22]} Confirming earlier observations in the ³¹P{¹H} NMR spectra by the ¹⁹⁵Pt{¹H} NMR spectra of complexes A and 1c (selected as case study) demonstrated a triplet resonance at δ = −4420.1 ppm with ¹J_PtP value of 3679 Hz and δ = −4787.3 ppm with ¹J_PtP = 2771 Hz, respectively, corresponding to the coupling among Pt(II) center and PPh₂allyl ligand. In the ¹H NMR spectra of 1a-e, all the characteristic resonances of PPh₂allyl ligands as well as S-coordinated thionate moieties have been clearly distinguished.

The colorless (A and 1c) or yellow (1a and 1b) single crystals, suitable for X-ray diffraction, were obtained by slow diffusion of n-hexane into a CH₂Cl₂ solution of each product at room temperature. An ORTEP representation of A and 1a-c are determined in Figure 2, whereas the crystal data and structural refinement parameters are collected in Table S1. A crystallized in the monoclinic (space group P21/n), while 1a-c crystallized in the triclinic (space group P $\stackrel{‒}{1}$) crystal systems. The molecular structure of precursor A evidently shows two PPh₂allyl ligands in a cis configuration while the molecular structures of 1a-c display two PPh₂allyl ligands in a trans position. Two allydiphenylphosphanes (P1, P2) and two chloride ligands (Cl1, Cl2) in A or two PPh₂allyl (P1, P#1) and two sulfur atoms (S1, S#1) in 1a-c entirely surrounded the Pt(II) center (Figure 2). The geometry around the platinum center in A and 1a-c is considered a slightly distorted square-planar. The angle subtended by the phosphane ligands at the Pt(II) center in A (P1-R1-P2 = 99.65(8)°), deviates from the 90° indicative of a distorted planar environment. The angles of P1-Pt1-P P#1 and S1-Pt1-S#1 in 1a-c are 180.0° but other angles between thiolates and phosphane ligands are deviated from the 90°. As depicted by perspective ORTEP view of the structures (Figure 2), the allyl substituents of the PPh₂allyl phosphane ligands are obviously detected to be perpendicular to the molecule plane in each structure. The distances between Pt(II) and the PPh₂allyl moieties (Pt1-P1 = 2.241(2) Å and Pt1-P2= 2.238(3) Å) in A, (Pt1A-P1A = 2.2980(4) Å and Pt1 B-P1B 2.3129(4) Å) in 1a, or (Pt1-P1 = 2.3015(6) Å) in 1b and (Pt1-P1 = 2.3031(5) Å) in 1c, are in the same range as those found in similar Pt(II)-phosphane complexes.^[23] The distances of the Pt–S bonds are almost equal in 1a-c (~2.33 Å). It is notable that the pyridyl moiety in the k¹-S-SR ligand is approximately perpendicular to the metal plane.

Figure 2.

ORTEP plot molecular structures of A and 1a-c drawn at 50% probability level. Selected bond lengths (Å) and angles (deg) for complexes: (a) A: Pt1-Cl1 2.352(2), Ptl-Cl2 2.346(3), Pt1-P1 2.241(2), Pt1-P2 2.238(3); Cl1-Pt1-Cl2 88.50(9), P1-Pt1-P2 99.65(8), Cl1-Pt1-P1 151.57(8), Cl2-Pt1-P2 168.14(9); (b) 1a: Pt1A-P1A 2.2980(4), Pt1B-P1B 2.3129(4), Pt1A-S1A 2.3261(4), Pt1B-S1B 2.3310(4), P1A-Pt1A-P#1A 180.0, P1B-Pt1B-P#1B 180.0, S1A-Pt1A-S#1A 180.0, S1B-Pt1B-S#1B 180.0; (c) 1b: Pt1-P1 2.3015(6), Pt1-S1 2.3313(5), P1-Pt1-P#1 180.0, S1-Pt1-S#1 180.0, P1-Pt1-S1 84.840(18), P1-Pt1-S#1 95.160(18); (d) 1c: Pt1-P1 2.3031(5), Pt1-S1 2.3328(5), P1-Pt1-P#1 180.0, S1-Pt1-S#1 180.0, P1-Pt1-S1 94.710(16), P1-Pt1-S#1 85.289(16), P#1-Pt1-S1 85.291(16), P#1-Pt1-S#1 94.710(16). Hydrogen atoms and CH₂Cl₂ solvent molecules (A and 1c) were omitted for clarity.

Biological activity studies

The in vitro cytotoxic activity of A and 1a-d were evaluated on three cancer cell lines including human lung (A549), ovarian (SKOV3), and breast (MCF-7) carcinoma. As shown in Table 1, 1a, showed higher anti-proliferative activity than cisplatin on the studied cell lines. It showed a good anti-proliferative activity with IC₅₀ of 4.31, 6.23 and 4.80 μM comparing with those measured for cisplatin (9.71, 14.48 μM and 11.59 μM, against A549, SKOV3 and MCF-7 cell lines, respectively). As shown in Tables S6-S8, one-way ANOVA statistical analysis showed that the differences between 1a and cisplatin is statistically significant. 1b also displayed better in vitro cytotoxicity than cisplatin on SKOV3 cell line with IC₅₀ of 12.38 μM, however this difference is not statistically significant. 1b and cisplatin IC₅₀ on A549 cell lines (Table S6) was also not statistically significant. 1c and cisplatin IC₅₀ on SKOV3 cells was also not statistically significant (Table S7). Complexes A, 1b, 1c, and 1e also showed antitumor activitiy against MCF-7 cell line in comparison with cisplatin. It should be mentioned that, the cytotoxicity of all the ligands including Spy, Spy-5-CF₃, SpyN, Sbt, Sbi and allydiphenylphosphane (PPh₂allyl) was evaluated against A549 cell line and the IC50 of all of these ligands was more than 100 μM.

Table 1.

In vitro cytotoxicity of all the synthesized compounds against cancerous and non-cancerous cell lines.

Complex	(IC50 ± SD) μM
	A549	SKOV3	MCF-7	MCF-10A
A	22.49 ± 1.62	34.50 ± 1.53	19.01 ± 1.24	68.74 ± 1.21
1a	4.31 ±0.72	6.23 ± 0.74	4.80± 0.71	38.42 ± 2.06
1b	12.59 ± 1.09	17.49 ± 1.21	19.68± 1.59	53.26 ± 1.13
1c	16.01 ±1.12	12.38 ± 1.17	15.64 ± 1.37	44.71 ±1.42
1d	20.15 ± 1.47	22.14 ± 1.67	34.65 ± 0.83	88.42 ±1.39
1e	18.62 ± 1.55	21.73 ± 1.19	17.45 ± 1.71	47.06 ± 0.83
Cisplatin	9.71 ± 1.70	14.48 ± 1.54	11.59 ± 1.66	29.47 ± 1.03

In addition, to verify the selectivity between cancer and normal cell line, the effects of these compounds on the proliferation of non-tumoral cell line (MCF-10A; normal human epithelial breast cell line) were also determined. As shown in Table 1, all of these compounds displayed reasonable selectivity between tumorigenic and non-tumorigenic cell lines and showed less cytotoxicity than cisplatin on MCF-10A cell line. Structure-activity relationship studies revealed that, generally, the complex A which encompasses chloro group, instead of thiolated ligand, showed lower anti-proliferative activity than others. Among the thiolated ligands, 1b with a thiolated pyrimidine ring and especially, 1a with a thiolated pyridine ring showed the highest potency. In order to gain some more insight in the structure-activity relationship of the synthesized trans-Pt(II) complexes, the chemical descriptors of them such as surface area, volume, hydration energy, logP (measure of lipophilicity), refractivity (measuring the total polarizability of a mole of a compound), polarizability and mass were calculated (Table S9) using Hyperchem 8 software. The greater logP (lipophilicity) of 1a confirmed the importance of this descriptors for the cytotoxic activity of the trans-Pt(II) complexes.

Determining apoptotic effect of 1a on MCF-7 cell line

We used BioLegend's PE Annexin V Apoptosis Detection Kit with 7AAD to specifically determine the dose-dependent apoptotic effect of complex 1a on cancerous cells. To determine this, 1a with three concentrations (2.5, 5 and 10 μM) was applied onto MCF-7 cells. As illustrated in Figure 3, with the increase in the concentration of 1a from 2.5 to 10 μM, the percentage of the cells in early apoptotic phase significantly increases from 7.8% in untreated cells to 11.0%, 40.8%, and 62.9% in the treated cells. This observation indicated that compound 1a as a representative of new trans-Pt(II) series, is able to effectively induce apoptosis in cancerous cells in a dose dependent manner. Also, the observed antiproliferative/cytotoxic effect for 1a in cytotoxic assay, could be partly mediated through inducing apoptosis in cancer cells.

Figure 3.

Flow cytometric analysis of apoptotic effect of 1a. MCF-7 cells (Human breast carcinoma) were left untreated (A) or treated for 48 h with 2.5, 5 and 10 μM of 1a. Q1: necrotic cells, Q2: late apoptotic cells, Q3: early apoptotic cells, and Q4: living cells.

The potential effect of 1a on the MCF-7 cells’ cell cycle

Quantitation of DNA content using flow cytometry or cell cycle analysis, is a basic method which is commonly used to assess the mechanisms of antiproliferative effects of anticancer drugs. In this method, a fluorescent DNA binding dye (in the present study: propidium iodide, PI) is used to stain DNA and measure the amount of DNA present in the cell. The cells in the G2 phase are expected to absorb approximately twice the amount of color compared to the cells in G1 as their DNA content has been doubled during S phase. Therefore, the cells in S phase have more DNA than G1 cells but less than G2 ones. Hence, we are able to check whether our compound exerts its antitumor effects thorough modifying cell cycle or not. As could be observed in Figure 4, comparing to untreated cells, no obvious change in different cycle phases could be observed which probably shows that 1a has no clear effect on the cell cycle of cancerous cells.

Figure 4.

The effect of 1a on the cell cycle in MCF-7 cells.

Genotoxicity and DNA interaction

Here, comet assay was used as a valuable method to predict the genotoxic effect of new synthesized trans-Pt(II) complexes on cancerous cells. In this single cell microgel electrophoresis method, following the DNA damage, the migration of chromosomal DNA from the nucleus increases and resembles the shape of a tail or comet. The longer tails display the more genotoxicity, whilst untreated cells as un-fragmented cells, represent a little or no tail. In the current study, we checked the genotoxicity of 1a as the best cytotoxic compound in the 1 series through comet assay. As could be observed in Figure 5, treatment of MCF-7 cells with both low and high concentrations of 1a (10 and 50μM), results in the appearance a relatively long tail following the electrophoresed cells in the concentration of 10 μM, which shows strong genotoxic ability of 1a. In the case of a concentration of 50, as displayed in Figure 5D, no nucleus remained and only a blurry tail of degraded DNA could be seen. These observations collectively showed that 1a compound intensely targets the genome content of cancerous cells. However, in electrophoresis mobility shift assay which used to further show the direct interaction of 1a compound with DNA, a little shift was observed comparing to cisplatin, as positive control (Figure 6). It could also be seen that compared to the untreated control, 1a could make a nick in DNA and subsequently slightly shift the mobility of the plasmid in a dose dependent manner. These observations confirm the direct interaction of 1a with DNA through genotoxic effect as observed in comet assay. This clearly indicates that the effective mechanisms of these compounds have a direct interaction with DNA and probably other molecules as well. These results are collectively consistent with previous studies which have described platinum compounds as DNA-targeting metal-based anticancer agents.^{[6e, 18]}

Figure 5.

Genotoxic effect of 1a on MCF-7 cell line. The percentage of degraded DNA in the tail has remarkably increase following treatment with doxorubicin (B: Doxorubicin) as positive control and different concentrations of 1a (C and D) in comparison to untreated cells (A: negative control). For better resolution, the same pictures from CometScore software are also shown.

Figure 6.

Mobility shift assay of 1a compound. pGEM-FT plasmid in circular form was incubated with different concentrations of cisplatin (positive controls) as well as compound 1a for 24h.

In order to determine the binding mode and binding site of the trans-Pt(II) complexes in interaction with DNA, molecular docking studies was also employed. Table S10 shows the docking binding energies of the trans-Pt(II) complexes with DNA (PDB ID: 1BNA). The high negative values of the binding free energies (kcal/mol) of the trans-Pt(II) complexes suggest that they bind reasonably well to DNA. 1d shows the lowest binding energies in compared to the others. As shown in Figure S1, 1d fitted in to the minor groove of DNA and interacts through its sulfur groups via weak hydrogen bonding with G4 and C11 base pairs in minor groove of DNA.

The percentage of degraded DNA in the tail has remarkably increased following treatment with doxorubicin (Dox) as positive control. This is also effected by differences in the concentrations of 1a in comparison to untreated cells (negative control). pGEM-FT plasmid in circular form was incubated with different concentrations of cisplatin (positive controls) as well as compound 1a for 24h.

Intracellular Reactive Oxygen Species (ROS) Determination

We used a flow cytometry based method for determination of cellular reactive oxygen species (ROS) in SKOV3 cell line after treatment with 1a. In this highly sensitive method, to consider ROS formation, SKOV3 cells were treated with 1a and H₂O₂ (positive control) and stained with 2’,7’-Dichlorodihydrofluorescein diacetate (DCFH-DA), a cell-permanent non-fluorescent dye which is oxidized by cellular ROS and produce 2’,7’- dichlorofluorescein (DCF) fluorescent component. The intensity of DCF is a direct estimate of the amount of ROS within the cells. As illustrated in the Figure 7, treatment of SKOV3 cells with 1a moderately induce ROS in a relatively dose dependent manner, however the production of ROS was higher in lower concentration of 1a (10 μM) comparing to higher concentration (40 μM) (Figure 7).

Figure 7.

Generation of ROS in SKOV3 cells induced by various concentrations of 1a and H202 as positive control. Changes in ROS levels were expressed as a ratio of the mean fluorescence intensity (MFI) in each condition divided by the basal intensity of the ROS at the untreated cells (negative controls, MFI0). H₂O₂ could induce ROS production within 20 minutes (blue line). Compound 1a at all concentrations could induce ROS production in SKOV3 cell line, however in lower concentration, 10 μM (dark blue line), is more efficient than higher concentrations, 20 μM (green line) and 40 μM (orange line), in comparison to untreated cells (red filled).

Conclusions

A synthetic approach has been introduced to obtain a class of platinum complexes with general formula trans-[Pt(PPh₂allyl)₂(k¹-S-SR)₂], 1, PPh₂allyl = allyldiphenylphosphine, SR = deprotonated form of pyridine-2-thiol (Spy, 1a), 5-(trifluoromethyl)-pyridine-2-thiol (SpyCF₃-5, 1b), pyrimidine-2-thiol (SpyN, 1c), benzothiazole-2-thiol (Sbt, 1d) and benzimidazole-2-thiol (Sbi, 1e). Multinuclear ¹H, ³¹ P{¹H} and ¹⁹⁵Pt{¹H} NMR spectroscopy was applied as an important tool to accurately characterize the new platinum(II) compounds. Single crystal X-ray crystallography technique, confirmed the cis-configuration of PPh₂allyl ligands in starting complex A. The X-ray crystal structure determined, the heterocyclic thionate ligands are bound to the Pt(II) center with S-coordinating mode while trans-positioned to each other. Meanwhile, all platinum(II) complexes were tested against three human cancer cell lines including lung (A549), ovarian (SKOV3), and breast (MCF-7) which showed potent antitumor activities. In vitro studies introduced 1a as a therapeutic agent due to the inhibition growth of MCF-7 cancer cell, mediated through inducing apoptosis. Furthermore, the potential effect of 1a on the MCF-7 cell was tested by cell cycle analysis to assess the mechanisms of antiproliferative effects. To predict the genotoxic effect of 1a on cancerous cells, comet assay was used as a valuable method which showed that 1a intensely targets the genome content of MCF-7 cancerous cells and directly interact with DNA as its major target. According to the observations, 1a has the highest affinity to the DNA in vitro and considerable potential for additional development as an antitumor agent. Further studies will reflect new insights into accurately exploring the mechanism of action of platinum drugs in biological systems in order to develop new effective anticancer metallodrugs.

Experimental Section

General procedures and materials

¹H NMR (400 MHz), ¹⁹F{¹H} (376.6 MHz), ³¹P{¹H} NMR (162 MHz) and ¹⁹⁵Pt{¹H} (85.6 MHz) spectra were recorded on a Bruker Avance DPX 400 MHz instrument at room temperature. All chemical shifts (δ) are reported in ppm (part per million) relative to their corresponding external standards (SiMe₄ for ¹H, CFCl₃ for ¹⁹F{¹H}, 85% H₃PO₄ for ³¹P{¹H}, Na₂PtCl₆ for ¹⁹⁵Pt{¹H}) and the coupling constants (J) have being expressed in Hz. The microanalyses were performed using a vario EL CHNS elemental analyzer. The instrument for HR ESI-Mass measurement was a Shimadzu IT-TOF with an electrospray ionization source, which is part of the Arkansas Statewide Mass Spectrometry Facility. The allyldiphenylphosphane (PPh₂allyl), pyridine-2-thiol (HSpy), 5-(trifluoromethyl)-pyridine-2-thiol (HSpyCF₃-5), pyrimidine-2-thiol (HSpyN), benzothiazole-2-thiol (HSBt), benzimidazole-2-thiol (HSBi) and all the other chemicals were purchased from commercial resources. All the reactions were carried out under Argon atmosphere and in the common solvents and all solvents were purified and dried according to standard procedures.^[24] Precursor complexes cis,trans-[PtCl₂(SMe₂)₂]^[20] or cis-[PtCl₂(dmso)₂],^[21] were synthesized as reported in the literature.

Synthesis of the complexes

cis-[Pt(PPh₂allyl)₂CI₂], A

To a solution of cis-[PtCl₂(dmso)₂] (450 mg, 1.07 mmol) or cis,trans-[PtCl₂(SMe₂)₂] (418 mg, 1.07 mmol) in CH₂Cl₂ (20 mL), two equivalents of PPh₂allyl (461 μL, 2.14 mmol) was added. The mixture was stirred at room temperature for 3 h and then the solvent was concentrated to small volume (~ 1 mL) under vacuum, and diethyl ether (3 mL) was added to give A as a white solid, which was filtered and washed with diethyl ether (3 × 3 mL) and dried. Yield: 669 mg, 87%. Elem. Anal. Calcd. for C₃₀H₃₀Cl₂P₂Pt (718.49): C, 50.15; H, 4.21. Found: C, 50.22; H, 4.26. HR ESI-MS(+) m/z Cacld. for [M - 2Cl]²⁺ 323.5731; Found 323.5737. NMR data in CDCl₃: δ (¹H) 7.43 (dd, ³J_HH = 8.0 Hz, ³J_PH = 10.1 Hz, 8H, H°), 7.35 (t, ³J_HH = 7.6 Hz, 4H, H^p), 7.18 (t, ³J_HH = 8.0 Hz, 8H, H^m), 5.98-5.90 (m, 2H, H^b), 5.09 (d, ³J_H^c-cis_H^b = 9.7 Hz, 2H, diastereotopic H^c), 4.84 (dd, ³J_H^c-transH^b = 16.6 Hz, ⁴J_H^c-trans_H^a = 3.1 Hz, 2H, diastereotopic H^c), 3.32 (dd, ³J_HH = 7.6 Hz, ²J_PH = 12.3 Hz, ³J_PrH = 35.3 Hz, 4H, H^a); δ (³¹P{¹H}) 4.6 (s, ¹J_PtP = 3671 Hz, 2P); δ (¹⁹⁵Pt{¹H}) −4420.1 (t, ¹J_PtP = 3679 Hz, 1Pt).

trans-[Pt(PPh₂allyl)₂(k¹-S-Spy)₂], 1a

Two equimolar amount of KSpy (21 mg, 0.14 mmol) were dissolved in ethanol (15 mL) and added to a solution of A (50 mg, 0.07 mmol) in CH₂Cl₂ (1 mL). The reaction mixture was stirred at room temperature for 10 h. Then, solvent was removed under reduced pressure and the residue was extracted with CH₂Cl₂ (10 mL). The obtained green solution was filtered through celite and the filtrate was concentrated to a small volume (~ 1 mL) under vacuum, and n-hexane (5 mL) was added to give 1a as a green solid, which was filtered and washed with n-hexane (3 × 3 mL) and dried. Yield: 45 mg, 74%. Elem. Anal. Calcd. for C₄₀H₃₈N₂P₂PtS₂ (867.90): C, 55.36; H, 4.41; N, 3.23; S, 7.39. Found: C, 55.48; H, 4.38; N, 3.28; S, 7.48. HR ESI-MS(+) m/z Cacld. for [M + 2H]²⁺ 434.5874; Found 434.5865. NMR data in CDCl₃: δ (¹H) 7.93 (m, 2H, H^g), 7.62 (m, 8H, H°), 7.40-7.09 (m, 14H, H^d, H^mand H^p), 6.80 (t, ³J_HH = 7.8 HZ, 2H, H^e), 6.54 (t, ³J_HH = 7.4 Hz, 2H, H^f), 5.83-5.71 (m, 2H, H^b), 4.90 (d, ³J_H^c-cisH^b = 9.4 Hz, 2H, diastereotopic H^c), 4.71 (d, ³J_H^c-transH^b = 15.9 Hz, 2H, diastereotopic H^c), 3.38 (m, 4H, H^a); δ (³¹P{¹H}) 10.3 (s, ¹J_PtP = 2732 Hz, 2P); δ (¹⁹⁵Pt{¹H}) - 4747.0 (t, ¹J_PtP = 2737 Hz, 1Pt).

The other new complexes were made similarly using A and the appropriate potassium thionate ligands.

trans-[Pt(PPh₂allyl)₂(k¹-S-SpyCF₃-5)₂], 1b

Yield: 60 mg, 86%. Elem. Anal. Calcd. for C₄₂H₃₆F₆N₂P₂PtS₂ (1003.89): C, 50.25; H, 3.61; N, 2.79; S, 6.39. Found: C, 50.33; H, 3.67; N, 2.74; S, 6.44. HR ESI-MS(+) m/z Cacld. for [M + 2H]²⁺ 502.5747; Found 502.5741. NMR data in CDCl₃: δ (¹H) 8.07 (s, 2H, H^f), 7.58 (dd, ³J_HH = 7.6 Hz, ³J_PH = 11.1 Hz, 8H, H°), 7.29 (t, ³J_HH = 7.4 Hz, 4H, H^p), 7.19 (t, ³J_HH = 7.4 Hz, 8H, H^m), 7.08 (d, ³J_HH = 8.3 Hz, 2H, H^d), 6.93 (dd, ³J_HH = 8.3 Hz, ⁴J_HH = 2.2 Hz, 2H, H^e), 5.87-5.77 (m, 2H, H^b), 4.98 (d, ³J_H^c-cis_H^b = 10.2 Hz, 2H, diastereotopic H^c), 4.81 (d, ³J_H^c-trans_H^b = 17.0 Hz, 2H, diastereotopic H^c), 3.39 (m, 4H, H^a); δ (¹⁹F{¹H}) −61.8 (s, 6F); δ (³¹P{¹H}) 10.1 (s, ¹J_PtP = 2687 Hz, 2P); δ (¹⁹⁵Pt{¹H}) −4747.4 (t, ¹J_PtP = 2686 Hz, 1Pt).

trans-[Pt(PPh₂allyl)₂(k¹-S-SpyN)₂], 1c

Yield: 49 mg, 81%. Elem. Anal. Calcd. for C₃₈H₃₆N₄P₂PtS₂ (869.87): C, 52.47; H, 4.17; N, 6.44; S, 7.37. Found: C, 52.32; H, 4.21; N, 6.49; S, 7.41. HR ESI-MS(+) m/z Cacld. for [M + 2H]²⁺ 435.5827; Found 435.5826. NMR data in CDCl₃: δ (¹H) 7.85 (d, ³J_HH = 4.6 Hz, 4H, H^d), 7.74 (dd, ³J_HH = 7.9 Hz, ³J_PH = 10.6 Hz, 8H, H°), 7.28-7.19 (m, 12H, H^m and H^p), 6.44 (t, ³J_HH = 4.6 Hz, 2H, H^f, 6.00-5.91 (m, 2H, H^b), 5.00 (d, ³J_H^c-cis_H^b = 10.3 Hz, 2H, diastereotopic H^c), 4.88 (d, ³J_H^c-trans_H^b = 16.4 Hz, 2H, diastereotopic H^c), 3.43 (m, 4H, H^a); δ (³¹P{¹H}) 10.8 (s, ¹J_PtP = 2766 Hz, 2P); δ (¹⁹⁵Pt{¹HJ) - 4787.3 (t, ¹J_PtP = 2771 Hz, 1Pt).

trans-[Pt(PPh₂allyl)₂(k¹-S-Sbt)₂], 1d

Yield: 58 mg, 84%. Elem. Anal. Calcd. for C₄₄H₃₈N₂P₂PtS₄ (980.07): C, 53.92; H, 3.91; N, 2.86; S, 13.09. Found: C, 53.98; H, 3.87; N, 2.81; S, 13.16. HR ESI-MS(+) m/z Cacld. for [M + 2H]²⁺ 491.0594; Found 491.0598. NMR data in CDCl₃: δ (¹H) 7.68 (dd, ³J_HH = 8.1 Hz, ³J_PH = 9.8 Hz, 8H, H°), 7.55 (d, ³J_HH = 8.0 Hz, 2H, H⁹), 7.45 (d, ³J_HH = 7.9 Hz, 2H, H^d), 7.28-7.16 (m, 14H, H^f, H^m and H^p), 7.11 (t, ³J_HH = 7.9 Hz, 2H, H^e), 5.79-5.68 (m, 2H, H^b), 4.93 (d, ³J_H^c-cisH^b = 10.2 Hz, 2H, diastereotopic H^c), 4.78 (d, ³J_H^c-trans_H^b = 16.8 Hz, 2H, diastereotopic H^c), 3.47 (m, 4H, H^a); δ (³¹P{¹H}) 10.9 (s, ¹J_PtP = 2591 Hz, 2P); δ (¹⁹⁵Pt{¹H}) −4641.4 (t, ¹J_PtP = 2589 Hz, 1Pt).

trans-[Pt(PPh₂allyl)₂(k¹-S-Sbi)₂], 1e

Yield: 46 mg, 69%. Elem. Anal. Calcd. for C₄₄H₄₀N₄P₂PtS₂ (945.97): C, 55.87; H, 4.26; N, 5.92; S, 6.78. Found: C, 55.98; H, 4.31; N, 5.87; S, 6.72. HR ESI-MS(+) m/z Cacld. for [M + 2H]²⁺ 473.5983; Found 473.5966. NMR data in CDCl₃: δ (¹H) 10.32 (brs, 2H, NH), 7.45 (dd, ³J_HH = 7.7, ³J_PH = 9.6 Hz, 8H, H°), 7.24 (d, ³J_HH = 7.8 Hz, 2H, H^g), 7.21 (d, ³J_HH = 7.6 Hz, 2H, H^d), 7.18-7.04 (m, 12H, H^m and H^p), 7.01-6.87 (m, 4H, H^e and H^f), 5.74-5.63 (m, 2H, H^b), 4.97 (d, ³J_H^c-cis_H^b = 10.0 Hz, 2H, diastereotopic H^c), 4.79 (d, ³J_H^c-trans_H^b = 16.2 Hz, 2H, diastereotopic H^c), 3.48 (m, 4H, H^a); δ (³¹P{¹H}) 10.6 (s, ¹J_PtP = 2643 Hz, 2P); δ (¹⁹⁵Pt{¹H}) −4684.8 (t, ¹J_PtP = 2651 Hz, 1Pt).

Single-crystal structure determination

Intensity data for these compounds were collected using a D8 Quest k-geometry diffractometer with a Bruker Photon II ccd area detector^[25] and an Incoatec Iμs microfocus Mo Kα source (λ = 0.71073 Å). The data was corrected for absorption by the empirical method.^[26] The crystal system and the space group were determined by systematic absences and statistical tests and verified by subsequent refinement. The structure was solved by direct methods and refined by full-matrix least-squares methods on F2.^[27] The positions of hydrogens were initially determined by geometry and were refined using a riding model. Non-hydrogen atoms were refined with anisotropic displacement parameters. Hydrogen atom displacement parameters were set to 1.2 times the isotropic equivalent displacement parameters of the bonded atoms.

The molecules in the crystal structure of 1a are located on inversion centers; thus, one half of the atoms in the two molecules are unique. The selected crystal for complex 1b was slightly split (~4°), requiring that the intensity data be processed as if the sample were a twin. The molecule was found to sit on a center of symmetry; thus, only one half of the atoms were unique. Also, the metal complex 1c was located on an inversion center; thus ½ of the atoms are unique. The solvent molecule was in a general position. The CH₂Cl₂ solvent molecule was disordered. The occupancies of the CH₂Cl₂ refined to 0.73(3) and 0.27(3) for the S and T orientations, respectively. Restraints on the positional and displacement parameters of the solvent were required.

Computational Details

Density functional calculations were performed with the program suite Gaussian09^[28] using the B3LYP level of theory.^[29] The LANL2DZ basis set was chosen to describe Pt^[30] and the 6-31G(d) basis set was chosen for other atoms. The geometries of complexes were fully optimized by employing the density functional theory without imposing any symmetry constraints. In order to ensure the optimized geometries, frequency calculations were performed employing analytical second derivatives. Solvent effects have been considered by the conductor-like polarizable continuum model (CPCM).^[31]

The crystal structure of 1c was directly or indirectly employed in order to make input files for the software. The ground state (S₀) of 1c and 1c" were optimized in gas phase or CH₂Cl₂ solvent and their optimized coordinates are collected in Tables S2-S5.

Biological assay

Cell lines and cell culture

Human cancer cell lines, MCF-7 (breast cancer), SKOV3 (ovarian cancer), and A549 (non-small cell lung cancer) were purchased from National Cell Bank of Iran (NCBI, Pasteur Institute, Tehran, Iran). The cells were grown in complete culture media containing RPMI 1640 (Biosera, France), 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin-streptomycin (Biosera, France) and kept at 37 °C in a humidified CO₂ incubator. MCF10A cells (human breast epithelial cell line) were cultured in DMEM/Ham's F-12 (GIBCO-Invitrogen, Carlsbad, CA) supplemented with 100 ng/ml cholera toxin, 20 ng/ml epidermal growth factor (EGF), 0.01 mg/ml insulin, 500 ng/ml hydrocortisone, and 5% chelex-treated horse serum.

Cytotoxic activities of trans-Pt(II) compounds were investigated using a standard 3-(4,5-dimethylthiazol-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay, as previously described.^{[6e, 18a]} To do this, the cells with a density of 0.8 × 10⁴ cells per well were seeded in 96-well microplates and kept for 24h to recover. The cells were then treated with compounds series A and 1 in different concentrations ranging from 1 to 100 μM in a triplicate manner and incubated for at least 72 hours at 37 °C in humidified CO₂ incubator. Following incubation, the media was completely discarded and replaced with 150 μl of RPMI 1640 containing 0.5 mg/mL MTT solution and incubated at room temperature for 3h. To dissolve the formazan crystals, the media containing MTT was discarded again and 150 μl of DMSO was added to each well and incubated for at least 30 min at 37 °C in the dark. The absorbance of individual well was then read at 490 nm with an ELISA reader. The 50% inhibitory concentration of each compound, representing IC₅₀, was calculated with CurveExpert 1.4. Data are presented as mean ± SD.

Apoptosis assay

BioLegend's PE Annexin V Apoptosis Detection Kit with 7AAD (Biolegend, USA) was used to assess the apoptotic effect of 1a compound as previously described.^{[6e, 18a]} Briefly, 0.5 x 10⁵ cells per 1 ml of complete culture medium were seeded in a 24-well culture plate, treated with 1a compound in different concentrations (2.5, 5 and 10 μM) for 72 h. An untreated sample was also included as a negative control. Treated and untreated cells were then harvested and washed twice with cold BioLegend's Cell Staining Buffer, transferred to the polystyrene round-bottom tubes (BD Bioscience, USA) and stained with 2 μl of PE-conjugated Annexin V and 2 μl of 7-AAD solution for 15 min at room temperature in the dark. 300 μl of Binding Buffer was added to each tube and analyzed immediately by four-color FACSCalibur flow cytometer (BD Bioscience, USA) with proper setting. The data were analyzed by FlowJo software packages.

Cell cycle analysis

The MCF-7 cells in a total number of 50 × 10³ were seeded in a 24-well cell culture plate and treated with two different concentrations of 1a (1.5 and 2.5 μM). Following 72 hours’ incubation, the cells were harvested and washed in PBS 1x. The cells were then fixed in cold 70% ethanol with overnight incubation at 4°C. The fixed cells were washed two times in PBS 1x and centrifuged at 850 ×g. Afterwards, the cells were treated with 50 μl ribonuclease A (100 μg/ml) to ensure only DNA, not RNA, is stained. At the end, 200 μl Propidium Iodide (PI, 50 μg/ml) solution were added to stain DNA. The cells were finally acquired on four-color FACSCalibur flow cytometer (BD Bioscience, USA) with proper setting and analyzed by FlowJo software.

Shift mobility assay

The shift mobility assay was applied to assess the direct interaction of 1a with DNA. For this purpose, as previously described,^[18b] the same aliquots of circular form of pGEM-FT plasmid was diluted in a buffer containing Tris-HCl (pH=8.5) in the presence of different concentrations of 1a (100, 200 and 400 μm) and then incubated at 37 °C for 24h. Cisplatin in the same concentrations as well as untreated DNA were also included as positive and negative controls, respectively. Afterward, 10 μl of each sample were mixed with 5 μl KBC loading dye (Kawsar Biotec, Iran) and electrophoresed for 3 hours at 70 V in 0.5% TEA buffer in 1 % agarose gel (Invitrogen, USA) and then visualized by a UV detector.

Comet assay

We also assessed the genotoxic ability of 1a compound using comet assay. To do this, 5 × 10⁵ MCF7 cells in 2 ml complete culture medium were prepared and treated with two different concentrations of 1a (10 and 50 μM). Untreated as well as Doxorubicin treated (1 μM) cells were also included as negative and positive controls, respectively. The cells were incubated for 20 min at 37°C in a humidified incubator with 5% CO2. The cells were then participated, re-suspended in 100 μl 1× PBS, mixed with low melting point agarose (LMPA) and dropped on a slide pre-coated slide with normal melting point agarose (NMPA) layer. A coverslip was placed over the gel and set at 4°C for 15 min. The coverslip was then removed and 100 μl of LMPA was added onto the agarose gel mixture layer, covered with a new coverslip and placed at 4°C for 15 min. The coverslip was then removed and the slides were immersed into cold lysis solution and refrigerated overnight and then in fresh cold alkaline electrophoresis buffer for 40 min. The slides were electrophoresed with the adjusted voltage (24 V) and current (300 mA). Afterward, the slides were flooded with neutralizing Tris buffer (pH=7.4) and distilled water for 5 min, and then in 70%, 90% and 100% Ethanol (Merck, Germany), sequentially. The slides were lastly stained with 100 μl PI (50μg/ml) and visualized by a high resolution fluorescent microscopy (BX61, Olympus). The images were taken at 20× magnification and analyzed by the Olympus micro imaging software CellSens (Olympus, Japan).

Determination of Intracellular Reactive Oxygen Species (ROS)

Cellular Reactive Oxygen Species Detection Assay Kit (TEB PAZHOUHAN RAZI, Iran) was used to determine the oxidative stress response in SKOV3 cell lines following treatment with 1a according to manufacture instruction with a little modification. Briefly, SKOV3 cells were grown in complete culture media, harvested as a single cell suspension with a density of 300 × 10³ and treated with 10, 20 and 40 μM for 2 h at 37°C. Untreated cells were also included as negative control. After incubation, the cells were washed two times with buffer R1 and stained with R2 buffer containing 2’,7’-Dichlorodihydrofluorescein diacetate (DCFH-DA) prepared in phenol red negative RPMI culture media (Biosera, France) for 1 h at 37°C. The H₂O₂ (1000 μM) were added to the control positive tube and incubated for more 20 min. The cells were then washed with R1 buffer and subjected to flow cytometry analysis immediately. At least 15000 events were acquired on four color FACSCalibur flow cytometer (BD Biosciences, USA) and analyzed by FlowJo software v10.^[18b]

CCDC-1901055 (A), CCDC-1901054 (1a), CCDC-1901052 (1b) and CCDC-1901053 (1c) contain the supplementary crystallographic data for this paper. This data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.