Triphenyltin(IV) Carboxylates with Exceptionally High Cytotoxicity against Different Breast Cancer Cell Lines

Organotin(IV) carboxylates are a class of compounds explored as alternatives to platinum-containing chemotherapeutics due to propitious in vitro and in vivo results, and distinct mechanisms of action. In this study, triphenyltin(IV) derivatives of non-steroidal anti-inflammatory drugs (indomethacin (HIND) and flurbiprofen (HFBP)) are synthesized and characterized, namely [Ph3Sn(IND)] and [Ph3Sn(FBP)]. The crystal structure of [Ph3Sn(IND)] reveals penta-coordination of the central tin atom with almost perfect trigonal bipyramidal geometry with phenyl groups in the equatorial positions and two axially located oxygen atoms belonging to two distinct carboxylato (IND) ligands leading to formation of a coordination polymer with bridging carboxylato ligands. Employing MTT and CV probes, the antiproliferative effects of both organotin(IV) complexes, indomethacin, and flurbiprofen were evaluated on different breast carcinoma cells (BT-474, MDA-MB-468, MCF-7 and HCC1937). [Ph3Sn(IND)] and [Ph3Sn(FBP)], unlike the inactive ligand precursors, were found extremely active towards all examined cell lines, demonstrating IC50 concentrations in the range of 0.076–0.200 µM. Flow cytometry was employed to examine the mode of action showing that neither apoptotic nor autophagic mechanisms were triggered within the first 48 h of treatment. However, both tin(IV) complexes inhibited cell proliferation potentially related to the dramatic reduction in NO production, resulting from downregulation of nitric oxide synthase (iNOS) enzyme expression.


Introduction
Given that carcinoma is a main cause of death and a significant obstacle to raising life expectancy globally [1][2][3], it is not a surprise that the prime objective of contemporary medicinal chemistry is the creation of novel anticancer medications. After the unintentional finding of cisplatin in the 1960s [4] and the success it had treating several solid tumors types [5,6], numerous platinum-based compounds have undergone substantial research as potential chemotherapeutics [7][8][9]. However, the drawbacks related to using platinum-based therapies in clinical settings, such as systemic toxicity as well as innate and/or acquired resistance [10,11], have motivated scientists to expand the investigation also on other possible metal-based chemotherapeutic alternatives potentially exhibiting distinct modes of action. Many different metals, such as Ti, Fe, Ru, Os, Co, Rh, Ir, Pd, Cu, Au, Ga, Ge, and Sn, have spurred considerable attention in this context [12][13][14][15][16]. Among them, organotin(IV) complexes have appeared as very noteworthy non-platinum metallodrugs investigated by many research groups in the past four decades [17][18][19][20]. This is due to their reduced toxicity [17] and capacity to circumvent the medication resistance perceived for some commercially available metal-based pharmaceuticals [21].
profen (corresponding carboxylate anions), namely [Ph3Sn(IND)] and [Ph3Sn(FBP)] (Figure 1). The two compounds have been previously reported [57,58]. In this study, however, these complexes were synthesized using different precursors; furthermore, a single crystal structure analysis of [Ph3Sn(IND)] is reported and the biological activity of both triphenyltin(IV) complexes was evaluated towards four cell lines of human breast carcinoma, namely BT-474, MCF-7, MDA-MB-468, and HCC1937, using colorimetric MTT-and CVbased cell viability assays. Furthermore, flow cytometry has been employed in order to understand the mechanism of the drug-induced cytotoxicity.

Materials and Methods
Reactions were performed implementing standard Schlenk techniques to ensure nitrogen atmosphere. Chemicals (triphenyltin(IV) chloride (Sigma-Aldrich Chemie GmbH, Steinheim, Germany), indomethacin (TCI), and racemic flurbiprofen (Biozol)) were used as purchased. Triethylamine was distilled from KOH and kept over 4 Å molecular sieves previously activated at 150 °C. Toluene was obtained from an MBraun Solvent Purification System (MBraun SPS-800) and kept over 4 Å active molecular sieves. A Heraeus VARIO EL oven was used for carrying out elemental analyses. NMR spectra were recorded on a Bruker AVANCE DRX 400 spectrometer. Tetramethylsilane (TMS) was used as internal standard to reference the NMR spectra; chemical shifts are reported in parts per million; 1 H (400.13 MHz), 13 C (100.61 MHz) and 119 Sn (149.21 MHz). Mass spectra (HR-ESI-MS) were obtained with an FT-ICR-MS Bruker Daltonics ESI mass spectrometer (APEX II, 7 T). X-ray data from single crystals were collected with a Gemini-CCD diffractometer (Rigaku Oxford Diffraction, Oxford, UK).

Synthesis and Stability of Complexes
A stoichiometric amount of indomethacin or racemic flurbiprofen (0.45 g or 0.32 g, respectively), dissolved in 10 mL toluene, was added to a solution of Ph3SnCl (0.5 g, 1.3 mmol) in 10 mL toluene. After 20 min stirring, NEt3 (0.18 mL, 1.3 mmol, 1 eq.) was slowly (10 min) added to the solution and the resulting solution was stirred at room temperature overnight. The formed precipitate, (Et3NH)Cl, was filtered off. The solvent was evaporated and the solid residue was recrystallized from a 3:1 mixture of chloroform and methanol. For detailed characterization (multinuclear NMR, X-ray crystallographic data and HR-ESI-MS), see SI, Figures S1-S10. Spectroscopic data of [Ph3Sn(IND)] and [Ph3Sn(FBP)] are in agreement with those previously reported [57,58]. [

Materials and Methods
Reactions were performed implementing standard Schlenk techniques to ensure nitrogen atmosphere. Chemicals (triphenyltin(IV) chloride (Sigma-Aldrich Chemie GmbH, Steinheim, Germany), indomethacin (TCI), and racemic flurbiprofen (Biozol)) were used as purchased. Triethylamine was distilled from KOH and kept over 4 Å molecular sieves previously activated at 150 • C. Toluene was obtained from an MBraun Solvent Purification System (MBraun SPS-800) and kept over 4 Å active molecular sieves. A Heraeus VARIO EL oven was used for carrying out elemental analyses. NMR spectra were recorded on a Bruker AVANCE DRX 400 spectrometer. Tetramethylsilane (TMS) was used as internal standard to reference the NMR spectra; chemical shifts are reported in parts per million; 1 H (400.13 MHz), 13 C (100.61 MHz) and 119 Sn (149.21 MHz). Mass spectra (HR-ESI-MS) were obtained with an FT-ICR-MS Bruker Daltonics ESI mass spectrometer (APEX II, 7 T). X-ray data from single crystals were collected with a Gemini-CCD diffractometer (Rigaku Oxford Diffraction, Oxford, UK).

Synthesis and Stability of Complexes
A stoichiometric amount of indomethacin or racemic flurbiprofen (0.45 g or 0.32 g, respectively), dissolved in 10 mL toluene, was added to a solution of Ph 3 SnCl (0.5 g, 1.3 mmol) in 10 mL toluene. After 20 min stirring, NEt 3 (0.18 mL, 1.3 mmol, 1 eq.) was slowly (10 min) added to the solution and the resulting solution was stirred at room temperature overnight. The formed precipitate, (Et 3 NH)Cl, was filtered off. The solvent was evaporated and the solid residue was recrystallized from a 3:1 mixture of chloroform and methanol. For detailed characterization (multinuclear NMR, X-ray crystallographic data and HR-ESI-MS), see SI, Figures S1-S10.  and Mo-K α radiation (λ = 0.71073 Å). Data reduction and empirical absorption correction were carried out using CrysAlisPro [62] with SCALE3 ABSPACK program for the later. For solving the structure, SHELXT-2018 [63] with dual-space method were employed, while structure refinement was performed with SHELXL-2018 [64] using full-matrix leastsquares routines against F 2 . Anisotropic refinement was carried out for all non-hydrogen atoms, whereas hydrogen atoms were calculated on idealized positions. The C 19 H 15 ClNO 4 substituent is disordered on two positions with a ratio of 0.814(5):0.186 (5). All chloroform solvent molecules are disordered as well. DIAMOND-4 [65] was used to generate all structural figures. Crystallographic details are listed in Table S1 in the SI.

Cell Lines, General Conditions and IC 50 Determination
Utilizing the aforementioned cell growth medium, cells were seeded in plates with 96 wells, at a 6000/100 µL/well density. Before treatment with the investigated compounds, cells were given 24 h to adhere. Stock solutions of cisplatin, indomethacin, flurbiprofen, Three independent biological replicates and four technical replicates were conducted. After 72 h treatment, cell viability was assessed. For the MTT assay, cells were initially rinsed with PBS and then exposed for 1 h to the working solution containing 0.5 mg mL −1 MTT in culture medium. The MTT solution was then removed, the formazan formed was dissolved in DMSO and its absorbance was measured at 570 nm and 670 nm employing a multi-well plate reader SpectraMax M5 (Molecular Devices, San Jose, CA, USA). For the CV assay, cells were first washed with PBS, fixated with 4% paraformaldehyde (PFA), dried after removal of the PFA solution and only then stained for 20 min with a 10% crystal violet solution. After discarding the staining solution, stained cells were supplied with acetic acid (33% in aqua bidest.) and the absorbance was recorded by using the previously described method at the same wavelengths [67]. Cell viability is expressed as a percentage of untreated cells with the mean value being computed using a four-parametric logistic function [68]. SigmaPlot 14.0 and Microsoft Excel 2013 programs were used for data analysis and IC 50 value calculation.

Flow Cytometry
BT-474 cells were seeded in 6-well plates, with a density of 150,000 cells/well. Cells were allowed to adhere overnight, after which they were treated with IC 50 value concentrations of [Ph 3 Sn(IND)], [Ph 3 Sn(FBP)] and cisplatin and subjected to flow cytometry (BD FACSAria III) assessment. For that, several staining procedures were employed including (1) AnnV/PI to identify cells undergoing apoptosis, (2) ApoStat to detect caspase activity, (3) AO to monitor autophagy induction, (4) DHR to measure reactive oxygen species/reactive nitrogen species (ROS/RNS), (5) DAF-FM to detect intracellular NO, and (6) CFSE to track the impact on cellular proliferation.
AnnV/PI, ApoStat, AO and DAF-FM staining was carried out according to the manufacturer's recommendations after 48-h-long exposure to the examined compounds, trypsinization and washing with PBS. AnnV/PI (5% AnnV, 2% PI in PBS) [69] and AO (1 µg mL −1 PBS) staining solutions were applied for 15 min at room temperature or at 37 • C, respectively, in an environment with 5% CO 2 . For caspase activity determination, 30 min exposure to the ApoStat (1% ApoStat, 5% FCS in PBS) stain at 37 • C with 5% CO 2 was applied. For NO production analysis, cells were dyed with DAF-FM dye (5 µM DAF-FM, 10% FCS in RPMI) for 1 h at 37 • C with 5% CO 2 . To neutralize the stain, the cells were incubated in serum-free medium for 15 min. Finally, cells' detachment was performed and flow cytometry examination was conducted through measurement of fluorescein isothiocyanate (FITC) fluorescence.
For ROS/RNS production analysis and cell proliferation investigations, cells were first dyed with the appropriate stain solution before being subjected to 48 h treatment with the experimental agents and cisplatin. DHR (1 µM DHR, 0.1% FCS in PBS) and CFSE (1 µM CFSE, 0.1% FCS in PBS) [70] stains were applied overnight at 37 • C with 5% CO 2 . After dyeing and 48 h treatment, trypsin-EDTA was used to detach the cells, which were subsequently rinsed with PBS and finally analyzed by flow cytometry.
Sn1-C32 2.129(5) Å) are comparable to ones reported for analogous triorganotin(IV) carboxylate complexes [71][72][73][74][75][76]. The carboxylato ligand bridges two tin atoms which are symmetry-independent resulting in distinct Sn-O bond lengths (Table 1).     Investigations were made with two distinct cell viability assays, MTT and CV. MTT is an indirect method for cell viability determination which utilizes the capacity of living cells to catalyze tetrazolinum salt reduction in MTT to formazan [77]. This is the result of mitochondrial dehydrogenase activity taking place in the mitochondria of viable cells. Consequently, substances that alter cellular metabolism through elevation of the level of reduced nicotinamide adenine dinucleotide phosphate (NADPH) or the activity of lactate dehydrogenase (LDH) may have a considerable impact on the results of the MTT experiment [78][79][80][81]. Therefore, for higher reliabil-   Investigations were made with two distinct cell viability assays, MTT and CV. MTT is an indirect method for cell viability determination which utilizes the capacity of living cells to catalyze tetrazolinum salt reduction in MTT to formazan [77]. This is the result of mitochondrial dehydrogenase activity taking place in the mitochondria of viable cells. Consequently, substances that alter cellular metabolism through elevation of the level of reduced nicotinamide adenine dinucleotide phosphate (NADPH) or the activity of lactate dehydrogenase (LDH) may have a considerable impact on the results of the MTT experiment [78][79][80][81]. Therefore, for higher reliability of the results, also the direct, non-enzymatic CV assay was employed. The cytotoxicity of the organotin(IV) complexes was compared to that of the two NSAIDs utilized as ligands, indomethacin and flurbiprofen, as well as cisplatin which is used as standard clinical ther- (Figures S13-S16). The IC 50 values are reported in Table 2. As expected, none of the four tumor cell lines was susceptible to the COX inhibitors alone (>100 µM). The cytotoxic potential of both organotin(IV) carboxylates, on the contrary, was significantly higher compared to cisplatin's activity, exhibiting nanomolar IC 50 [57]. The investigated complexes showed much higher activity with respect to the structurally similar complex with an ibuprofenate ligand ([Ph 3 Sn(IBF)]) which has been found to be active against the Caco-2 colorectal adenocarcinoma cell line, but only remotely active or completely inactive against the DU145 (prostate carcinoma) cells and the HCT-15 (colon adenocarcinoma) cells, respectively [55]. In our previous study, flurbiprofen was used as an axial ligand of a cisplatin-based platinum(IV) conjugate and the cytotoxic activity of this cis,trans,cis-[PtCl 2 (FBP) 2 (NH 3 ) 2 ] complex was assessed against the same four breast carcinoma cell lines used in the present study [61]. The obtained results show very similar cytotoxic activity for [Ph 3 Sn(FBP)] and the cisplatin-flurbiprofenate complex suggesting that both NSAID-metal complexes with tin(IV) and platinum(IV) are highly active. Since the COX inhibitory potential of the prepared metal complexes was not assessed in these studies, the mechanism underlying the increase in cytotoxic efficacy following NSAID conjugation is unclear. The findings from the CV and MTT assays for [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] are in very good agreement with each other, while for cisplatin towards the BT-474, HCC1937 and MDA-MB-468 cells, some discrepancies were identified. This suggests that the investigated organotin(IV) carboxylates have different mechanism of action than cisplatin, not affecting the cell metabolism pathways.

Mode of Cytotoxic Activity
Flow cytometry was employed in order to comprehend the mechanism underlying the drug-induced cytotoxicity. In this method, a specific type of cell in a heterogeneous environment is recognized and physically separated using fluorescently labeled target-specific antibodies. This enables analysis of the nucleic acid material, presence of characteristic proteins, as well as phenotype-specific metabolic content and, therefore, valuable information on cell proliferation, apoptosis and autophagy, ROS/RNS, and NO production, etc., can be acquired. These data allow for conclusions regarding the drug's mechanism of action to be drawn. In the present investigation, BT-474 cells underwent treatment with IC 50 doses of [Ph 3 Sn(IND)], [Ph 3 Sn(FBP)] and cisplatin before undergoing a variety of FACS assessments. This cell line was chosen for the mechanistic studies because of the high cytotoxicity of the investigated compounds against it, which is also the greatest advancement in comparison to cisplatin.
In order to determine whether [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] induce apoptotic cell death, AnnV/PI staining was used. This method relies on the relocation of the membrane phospholipid phosphatidylserine (PS) from the internal to the external side of the membrane, causing a change of the plasma membrane asymmetry, which is a key hallmark of apoptosis. Because of its strong affinity for PS, the annexin V dye acts as a sensor for cells undergoing apoptosis including early apoptotic cells [82]. As the apoptotic process progresses, the intactness of the membrane is lost and it becomes permeable for the second dye, PI. Therefore, cells that are both AnnV-and PI-positive, are in late apoptosis or dead already. Results presented in Figure 4 show  Identification of activated caspases is another marker of apoptosis. These cytosolic cysteine proteases are implicated in the onset as well as execution of apoptosis. Therefore, detection of this enzyme's activity indicates the presence of apoptotic cells. This can be probed utilizing the ApoStat assay. Figure 5A reveals that [Ph3Sn(IND)] and [Ph3Sn(FBP)] do not activate, but rather inhibit the activity of these enzymes, whereas cisplatin has no effect on the caspase activity, although previous reports link cisplatin-induced cell death with caspase-activated apoptosis [61,[83][84][85][86][87]. These results, however, are in agreement with the ones obtained with the AnnV/PI experiments and confirm that apoptotic mechanisms in BT-474 cells could not be detected within the 48 h of treatment. Identification of activated caspases is another marker of apoptosis. These cytosolic cysteine proteases are implicated in the onset as well as execution of apoptosis. Therefore, detection of this enzyme's activity indicates the presence of apoptotic cells. This can be probed utilizing the ApoStat assay. Figure 5A reveals that [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] do not activate, but rather inhibit the activity of these enzymes, whereas cisplatin has no effect on the caspase activity, although previous reports link cisplatin-induced cell death with caspase-activated apoptosis [61,[83][84][85][86][87]. These results, however, are in agreement with the ones obtained with the AnnV/PI experiments and confirm that apoptotic mechanisms in BT-474 cells could not be detected within the 48 h of treatment.
detection of this enzyme's activity indicates the presence of apoptotic cells. This can be probed utilizing the ApoStat assay. Figure 5A reveals that [Ph3Sn(IND)] and [Ph3Sn(FBP)] do not activate, but rather inhibit the activity of these enzymes, whereas cisplatin has no effect on the caspase activity, although previous reports link cisplatin-induced cell death with caspase-activated apoptosis [61,[83][84][85][86][87]. These results, however, are in agreement with the ones obtained with the AnnV/PI experiments and confirm that apoptotic mechanisms in BT-474 cells could not be detected within the 48 h of treatment. Lack of impact was also the outcome of the DHR staining investigation conducted with the aim to assess the influence of [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] and cisplatin on the production of ROS and RNS. Although increased ROS/RNS production has been previously related to cytotoxicity brought on by cisplatin [83,88,89] as well as cisplatin-NSAID conjugates [61], in BT-474 cells treated with cisplatin or [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)], only slightly elevated ROS/RNS production could be detected ( Figure 5B). Thus, on breast BT-474 cells the tested complexes did not induce cell death by orchestrating antioxidant systems [90].
Interestingly, all three compounds have a most prominent effect on the generation of nitric oxide resulting in a significant decrease in NO. This effect, as displayed in Figure 5C, is most pronounced for the tin(IV)-based compounds with [Ph 3 Sn(FBP)] being the more potent NO suppressor, while cisplatin causes the least notable inhibition. Nitric oxide is a bioactive molecule having a profound impact on numerous physiological and pathological processes [91]. In cancer biology, the role of NO is controversial as it can exert both carcinogenic or anticancer effects, depending on the location, time, and its concentration [92][93][94][95]. NO is produced enzymatically from nitric oxide synthase (NOS). Despite the existence of three isoforms of this enzyme (neuronal, inducible, and endothelial), the inducible form (iNOS) has the most compelling association with tumor progression and metastasis [92]. Namely, elevated iNOS expression has been reported for different cancers [96][97][98][99], including breast cancer [100][101][102][103], where iNOS overexpression has been linked with tumor aggressiveness and poor prognosis for the patients. Furthermore, Chang et al. [104] have shown that treatment with iNOS inhibitors can suppress tumor cell proliferation as well as cancer stem cells' capacity for self-renewal and migration, hence lowering tumor initiation, growth, and the incidence of lung metastases from breast cancer. Considering the fact that our findings suggest that the metal complexes cause inhibition of cell proliferation (evaluated with a CFSE assay; Figure 5D) in BT 474 cells, it can be assumed that the suppression of NO production correlates to reduced cell proliferation, possibly through decreasing iNOS gene expression. A previous study reports that flurbiprofen has the ability to inhibit iNOS expression in RAW 264.7 macrophages [105]. Furthermore, indomethacin has been found to decrease iNOS gene expression and reduce tumor growth in vivo in breast tumor-bearing mice [106]. These findings regarding our ligand precursors, indomethacin and flurbiprofen, further corroborate our conclusions that the cytotoxic capacity of [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] could be due to iNOS suppression-mediated reduction in NO production.
Finally, an acridine orange (AO) assay was used to investigate if [Ph 3 Sn(IND)] and [Ph 3 Sn(FBP)] as well as cisplatin can trigger autophagy in BT-474 cells. This controlled lysosomal mechanism is engaged in the breakdown and reuse of cytoplasmic components. Amino and fatty acids, which are being created during this process, can be utilized to produce proteins or energy, which are critical for a cell's ability to survive in a starvation situation. However, autophagy can also result in cell death upon exposure to chemical therapy [107]. According to our findings ( Figure 6) sidering the fact that our findings suggest that the metal complexes cause inhibition of cell proliferation (evaluated with a CFSE assay; Figure 5D) in BT 474 cells, it can be assumed that the suppression of NO production correlates to reduced cell proliferation, possibly through decreasing iNOS gene expression. A previous study reports that flurbiprofen has the ability to inhibit iNOS expression in RAW 264.7 macrophages [105]. Furthermore, indomethacin has been found to decrease iNOS gene expression and reduce tumor growth in vivo in breast tumor-bearing mice [106]. These findings regarding our ligand precursors, indomethacin and flurbiprofen, further corroborate our conclusions that the cytotoxic capacity of [Ph3Sn(IND)] and [Ph3Sn(FBP)] could be due to iNOS suppression-mediated reduction in NO production.
Finally, an acridine orange (AO) assay was used to investigate if [Ph3Sn(IND)] and [Ph3Sn(FBP)] as well as cisplatin can trigger autophagy in BT-474 cells. This controlled lysosomal mechanism is engaged in the breakdown and reuse of cytoplasmic components. Amino and fatty acids, which are being created during this process, can be utilized to produce proteins or energy, which are critical for a cell's ability to survive in a starvation situation. However, autophagy can also result in cell death upon exposure to chemical therapy [107]. According to our findings ( Figure 6), cisplatin, [Ph3Sn(IND)] and [Ph3Sn(FBP)] did not have any significant influence on autophagy after 48 h treatment, thus neither cytoprotective nor cytotoxic effect of autophagy [108] can be expected upon treatment of B-474 cells with the investigated complexes.

Conclusions
In this study, two triphenyltin(IV) carboxylates with NSAIDs as ligands, This superior cytotoxicity, demonstrated against all cell lines involved, represents a significant enhancement in comparison to the effect of cisplatin. Various biological experiments used to investigate these compounds' mode of action suggested that neither apoptotic nor autophagy mechanisms were activated within the 48 h treatment. Ho wever, slightly elevated ROS/RNS production as well as inhibition of the cell proliferation were confirmed for these compounds; the latter might be related to the massive suppression of NO production. Gene expression experiments could be additionally performed to evaluate if the extensive NO production suppression is caused by inhibition of iNOS expression.