Ru(II) Complexes Bearing O, O-Chelated Ligands Induced Apoptosis in A549 Cells through the Mitochondrial Apoptotic Pathway

Two new Ru(II) complexes containing O, O-chelated ligands, Ru(dip)2(SA) (Ru-1) and Ru(dmp)2(SA) (Ru-2) (dip = 4,7-diphenyl-1,10-phenanthroline; dmp = 2,9-dimethyl-1,10-phenanthroline; SA = salicylate) were synthesized to evaluate their cytotoxicity in vitro. These complexes were found to exhibit moderate antitumor activity to different types of human cancers, including A549 (human lung carcinoma), MCF-7 (breast cancer), HeLa (human cervical cancer), and HepG2 (human hepatocellular carcinoma) cell lines, but displayed low toxicity to human normal cell lines BEAS-2B (immortalized human bronchial epithelial cells) when compared with that of cisplatin. Further studies revealed that these complexes could induce apoptosis in A549 cells, including activating caspase family proteins and poly (ADP-ribose) polymerase (PARP), reducing Bcl-2/Bax and Bcl-xl/Bad ratio, enhancing cellular reactive oxygen species (ROS) accumulation, triggering DNA damage, decreasing mitochondrial membrane potential (MMP), and leading cytochrome c release from mitochondria. Notably, complex Ru-1 showed low toxicity to developing zebrafish embryos. The obtained results suggest that these new synthetic complexes have the potential to be developed as low-toxicity agents for lung cancer treatment.


Introduction
As a new class of nonplatinum metal complexes, rutheniumbased compounds possess valuable photophysical and photochemical properties and high structural diversity, which provide more direction for designing new ruthenium anticancer drugs [1,2]. More importantly, ruthenium-based complexes often show better tumor cell selectivity than platinum metal complexes, in addition to low toxicity to normal cells and multiple anticancer mechanisms, thus making them attractive chemotherapeutic agents [3,4]. So far, there are four ruthenium complexes with different ligands, that is, [ImH][trans-RuCl 4 (DMSO)(Im)] (NAMI-A) [5,6], [IndH][trans-RuCl 4 (Ind) 2 ] (KP1019) [7], KP133910 [8], and the Ru(II)-based photosensitizer, TLD1433 [9,10], have progressed to different stages in clinical trials. In the review paper, Chen et al. pointed out that the selection of ligands plays a key role in the antitumor cell selectivity, targeting, antitumor activity, and mechanism of ruthenium compounds [3].
However, while there is a large number of studies mainly focused on Ru(II) complexes containing N, N-and C, N-chelated ligands, there is little research on the biological activities of O, O-chelating Ru(II) complexes. de Carvalho et al. found that ruthenium complex with xanthoxylin (RCX) ligand exhibited a potent cytotoxic effect in a panel of cancer cell lines in monolayer cultures and induced S-phase arrest and caused ERK1/2-mediated apoptosis in HepG2 cells through a p53-independent pathway [11]. Habtemariam et al. reported that the water-soluble Ru(II) acetylacetonate coordination compound ([Ru(acac)(pdto)]Cl, where pdto � 2,2′-[1,2-ethanediylbis-(sulfanediyl-2,1-ethanediyl)] dipyridine) showed a remarkable anti-E. histolytica activity in vitro with IC 50 values of 60 nM [12]. Toledano-Magaña et al. reported a series of neutral Ru(II) complexes containing anionic O, O-chelating ligands (acetylacetonate derivatives), and their studies suggested some of these compounds possessed reasonable activity towards A2780 cells [13]. Furthermore, Bezerra and his partners synthesized some piplartine-containing ruthenium complexes and investigated the apoptosis inducing effect in human colon carcinoma HCT116 cells and underlying mechanisms by these ruthenium complexes-induce cell death [14,15]. In our recent studies, we have synthesized three O, O-chelated ligand-bearing Ru(II) complexes, which are [Ru(bpy) 2 (SA)] (SA � salicylate), [Ru(phen) 2 (SA)] and [Ru(dmb) 2 (SA)], and found that these complexes induced apoptosis in A549 cells by targeting TrxR [16].
Notably, salicylic acid (SA) or o-hydroxybenzoic acid, a precursor of aspirin (AS), has been applied as a nonsteroidal anti-inflammatory drug (NSAID) [17]. Salicylate ligand has two hard and strong alkaline donor centers, which can promote chelation or metal bridging of medium to large cations [18]. As multifunctional ligands, SA and AS are attracting ever more attention for the treatment of cancer. For instance, the research of Wu et al. demonstrated that conjugation of SA or AS to the Ir(III) moiety could improve the cellular uptake efficacies of the Ir(III) complexes and achieve a high synergistic effect [19]. Liu and Dhar et al. reported that a novel Pt(IV) prodrug of cisplatin, asplatin, with the ligation of aspirin (c,c,t-[PtCl 2 (NH 3 ) 2 (OH) (aspirin)]), demonstrated significantly higher cytotoxicity than cisplatin towards tumor cells and nearly completely overcame the drug resistance of cisplatin-resistant cells [20][21][22]. Recently, Kowalski reported a new rhenium compound with AS as the ligand, fac-[Re(CO) 3 (phen)(aspirin)], which exhibited activity against HeLa human cancer cells, and this anticancer activity could be linked to ROS production, and cell cycle disturbance followed by triggering an apoptotic pathway of cell death [23]. Similar exciting antitumor effect has also been found in other metal-based complexes of SA. For example, silver(I) [24][25][26] and copper(II) [18,27] complexes with SA as the ligand exhibited strong antiproliferative activity against cancer cell lines.
2.6. Cellular Uptake and Nuclear Localization. A549 cells were seeded into six-well plates (5.0 × 10 5 cells per well) and grown overnight at 37°C in a 5% CO 2 incubator. e cells were incubated with the different concentrations (5, 10, 15, and 20 μM) of Ru(II) complexes for different time intervals (1, 3, and 6 h). After the incubation, the cells were harvested and washed twice with PBS. Cell Mitochondria Isolation Kit was used to extract the nuclear, mitochondrial, and cytoplasmic fractions of the A549 cells. e pellets were digested with 3 mL concentrated nitric acid and 1 mL perhydrol for 24 h and then diluted to 5 mL with ultrapure water. Finally, ICP-MS was used to determine the amount of Ru(II) complexes uptaken by A549 cells.

Lipophilicity Measurements.
Log P o/w is the partition coefficient between octanol and water. Briefly, a suitable amount of stock solution of the Ru(II) complex in aqueous NaCl was added to an equal volume of octanol, and the mixture was shaken for 48 h at 200 rpm at 25°C to allow partitioning. e aqueous layer was separated from the octanol layer after the sample was centrifuged at 3000 rpm for 10 min.
e Ru(II) content in the aqueous layer was measured by ICP-MS. Finally, Log P o/w values were calculated according to the

Apoptosis
Assay. First, A549 cells were incubated in sixwell plates for 12 h and then exposed to Ru(II) complexes for 24 h. After incubation, cell nuclei were stained with Hoechst 33342 (5 μg/mL) for 10 min, washed twice with PBS, and then photographed using an inverted fluorescence microscope. Second, after incubation with different concentrations (5, 10, and 20 μM) of Ru-1 and Ru-2 for 24 h, respectively, cells were harvested and washed twice with PBS and then resuspended in 500 μL binding buffer. e suspension was stained with 5 μL Annexin V-FITC and 10 μL PI at room temperature for 15 min and then analyzed using the flow cytometer.

MMP and ROS Determination.
e mitochondrial membrane potential (MMP) of A549 cells was analyzed by using an inverted fluorescence microscope and flow cytometry. After pretreatment with different concentrations (5, 10, and 20 μM) of Ru-1 and Ru-2 for 12 h, the A549 cells were trypsinized and washed twice with PBS. For microscope observation, the collected cells were incubated in complete medium containing JC-1 (10 μg/mL) for 30 min and washed with PBS twice and then imaged by an inverted fluorescence microscope (Nikon, Japan). For flow cytometry analysis, the cells were trypsinized and washed twice with PBS and then incubated in 500 μL PBS containing JC-1 (10 μg/mL) for 30 min at 37°C. e cells were analyzed by flow cytometer immediately. e intracellular ROS level of A549 cells was detected after DCFH-DA stain, as illustrated in our new publication [16,30].

Comet Assay and Western Blotting
Assay. Single-cell gel electrophoresis was performed to detect DNA damage, as previously described [29]. Comet assay was performed according to the manufacturer's instructions. DNA was stained with SYBR Green I (Trevigen) and imaged under an inverted fluorescence microscope. e effects of Ru-1 on the expression levels of proteins associated with caspase, Bcl-2 family proteins, and DNA damage relative proteins were examined by western blotting assay according to our previously reported method [31]. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. e protein concentrations were determined by BCA Protein Assay Kit.

Synthesis, Characterization, and Solution Stability.
Complexes Ru-1 and Ru-2 were synthesized by the reaction of cis-[Ru(N-N) 2 Cl 2 ] (N-N � dip, dmp) and salicylic acid in the presence of sodium hydroxide. e crude product was purified by chromatography over alumina. e obtained compounds were analyzed by elemental analysis, ESI-MS, 1 H NMR, FTIR, and UV-Vis absorption spectroscopy, and the results were shown in Figures S1-S7 in Supplementary Materials. e stability of Ru-1 and Ru-2 in water and CH 3 CN solution at 298 K was analyzed by UV-Vis absorption spectroscopy ( Figures S3-S5). As demonstrated in Figures S4 and S5, there is no significant change in the UV-Vis absorption spectra of Ru-1 and Ru-2 at 298 K over 24 h, indicating that Ru-1 and Ru-2 are sufficiently stable for the majority of clinical applications. In the UV-visible spectral regions, both complexes consist of two well-resolved absorption bands. In the visible region, these two complexes present a strong MLCT transition at 500-600 nm attributed to the overlap Ru(dπ)-dip or dmp (π * ) and Ru(dπ)-SA (π * ). e intense band occurring between 250 nm and 300 nm is attributed to an intra-ligand (IL) π-π * transition by comparing with the spectra of other Ru(II)-polypyridyl complexes [32][33][34].
e obtained FTIR spectra of Ru-1 and Ru-2 in KBr pellets are shown in Figures S6 and S7. In the FTIR spectrum of Ru-1 and Ru-2, the absorptions due to the stretching vibrations of the carboxylato unit have been recognized at 1597 and 1350 cm −1 [35,36]. e band at 702 cm −1 is attributed to the Ru-O stretch of Ru-1. Additionally, the appearance of the peak at 559 cm −1 corresponding to the Ru-O stretch of Ru-2 is consistent with Ru(II) coordination [37]. e bands at ca. 1462 cm −1 and 1465 cm −1 are the stretching vibrations of the C-N group of the ligand dip of Ru-1 and dmp of Ru-2, separately [38,39]. In addition, peak observed at 1365 cm −1 is attributed to the methyl of dmp of Ru-2 [38], which distinguished the IR spectrum of Ru-2 from that of Ru-1.
Investigating the hydrolysis in the presence of biorelevant molecules is very important to assess the chemical stability and biological activities of these title complexes. Hence, the stability of complexes Ru-1 and Ru-2 towards bovine serum albumin (BSA) and glutathione (GSH) was also determined. As shown in Figures S8 and S9, after 12 h of incubation with excessive GSH or BSA, respectively, there was no significant change observed in the characteristic absorption of Ru-1 and Ru-2, which suggested that Ru-1 and Ru-2 are quite stable in PBS buffer. Also, it implied that Ru-1 and Ru-2 were inactive towards GSH and BSA. is is further supported by the selectivity index (SI) assay (Table 1). Compared with Ru-2 (1.6) and cisplatin (0.8), Ru-1 (4.9) demonstrates significantly higher safety profile.

Cellular Uptake and Localization.
e lipophilicity of a drug plays a vital role in its cytotoxicity [29,[40][41][42]. According to our studies, the cellular uptake and cytotoxicity of a ruthenium complex is positively correlated with its lipophilicity [29,30]. Log P o/w values were used to estimate the lipophilicity quantitatively. As illustrated in Figure 2(a), complex Ru-1 (log P o/w � 1.4 ± 0.1) shows higher log P o/w value than that of Ru-2 (log P o/w � 0.9 ± 0.02), suggesting Ru-1 has higher lipophilicity. e above result is in line with the report that there is a positive correlation between hydrophobicity and cytotoxicity for metal-based anticancer agents [42][43][44].
It is generally acknowledged that the cellular uptake feature of transition metal-based compounds is an essential factor for their biological activities [45][46][47]. As repeatedly reported by previous ruthenium complex-related studies, the more cellular uptake results in higher cytotoxicity [29,30,48]. e cellular ruthenium concentration was determined by using ICP-MS after 1, 3, and 6 h of exposure to 5, 10, 15, and 20 μM Ru(II) complexes (Figure 2), respectively, and the results were reported as ng of ruthenium per 10 6 cell number. As expected, Ru-1 exhibits higher cellular uptake than that of Ru-2 during all the tested time points and drug concentrations (Figure 2(b) and Table S1). In a short time (1 h), the uptake levels were relatively low for Ru-1 (120.3 ± 9.5 ng) and Ru-2 (85.4 ± 9.3 ng) at low drug concentration (5 μM). With the extension of time and increase of concentration, the uptake levels for both Ru-1 and Ru-2 increased. e maximum cellular uptake was reached after 6 h of exposure to Ru-1 (432.5 ± 12.4 ng) and Ru-2 (235.2 ± 16.6 ng) at the concentration of 20 μM, and Ru-1 has a 1.8-fold better cell entry than Ru-2. e subcellular distribution of Ru-1 in A549 cells was further studied. As displayed in Figure 2(c), Ru-1 was mostly accumulated in the cell nucleus, followed by mitochondria.
ese results imply that Ru-1 may lead to mitochondria and nucleus damage and subsequent cancer cell apoptosis.

Complexes Ru-1 and Ru-2 Induced A549 Cell Apoptosis.
Apoptosis, an active form of cell death, plays a vital role in cell development and survival by clearing away damaged or unwanted cells [49]. Apoptosis is an approach by which many chemical agents exert anti-cancer effect [38,50]. Based on the apparent proliferation inhibition of Ru-1 and Ru-2 on A549 cells, their impact on apoptosis induction was further studied via Hoechst 33342 staining technique. As presented in Figure 3(a), after treatment with both Ru-1 and Ru-2, apoptotic features including chromatin condensation, nuclear fragmentation, and plasma membrane blebbing were observed. Also, at the same concentration, complex Ru-1 induced more severe apoptosis than Ru-2 in the tested cells. To further ascertain the apoptosis induced by these Ru(II) complexes, flow cytometric analysis was carried out. As shown in Figure 3 When apoptosis occurred, the expression of apoptosisrelated proteins might change. So, the effect of Ru-1 on key apoptosis protein caspase-3/8/9 and poly (ADP-ribose) polymerase (PARP) were analyzed by western blotting (Figure 3(c)). As demonstrated in Figure 3(c), after treatment with Ru-1 for 24 h, the expression of cleaved-PARP, cleaved caspase-3/8/9 increased, and the expression of total caspase-3/8 decreased in a concentration-dependent manner, which illustrated that both intrinsic and extrinsic apoptotic pathways were involved in Ru-1-induced apoptosis of A549 cells. In conclusion, complexes Ru-1 and Ru-2 induced apoptosis in A549 cells probably through caspasedependent extrinsic and intrinsic pathway.

Complexes Ru-1 and Ru-2 Induced the Decrease of MMP.
Mitochondrion, which controls the energy production in most eukaryotic cells, is one of the major cell signaling center and plays a vital role in various cellular activities [51,52]. Mitochondrial dysfunction, such as the decline of Bioinorganic Chemistry and Applications mitochondrial membrane potential (MMP), is an important indicator for apoptosis detection [53]. To test the effect of Ru-1 and Ru-2 on MMP, fluorescent dye JC-1-based MMP evaluation was employed. As Figure 4(a) illustrated, comparing with the untreated cells, after incubation with 5, 10, and 20 μM of Ru-1 or Ru-2 for 12 h, the A549 cells presented a significant red-to-green color shift in a concentrationdependent manner, suggesting the decline of MMP. Meanwhile, flow cytometry was employed to provide further information about the quantificational changes of MMP. As shown in Figure 4(b), after incubation with 5, 10, and 20 μM of Ru-1 or Ru-2, the percentage of green fluorescence underwent a dramatic increase from 5.1% to 70.5% for Ru-1 and from 5.1% to 54.0% for Ru-2, respectively. e quantification of green/red fluorescent intensity ratio by JC-1 staining after 5, 10, and 20 μM of Ru-1 and Ru-2 treatment for 12 h is shown in Figure 4(c).
To investigate whether the mitochondrial pathways involved in Ru-1-induced apoptosis, western blotting was further performed. As presented in Figure 4(d), after coincubation with Ru-1, while the proapoptotic protein Bax and Bad displayed increased expression, the expression levels of the antiapoptotic proteins Bcl-2 and Bcl-xl were both downregulated. e reduction of the ratios of Bcl-2/Bax and Bcl-xl/Bad results in the release of cytochrome c into the cytosol [54,55]. e apparent enhancement of cytochrome c expression in cytosol can be observed in Figure 4(d). e above results confirmed that Ru-1 and Ru-2 could induce the reduction of MMP, and the mitochondrial pathway was involved in Ru-1-induced apoptosis in A549 cells.

Complexes Ru-1 and Ru-2 Activated Intracellular ROS
Generation. Increasing intracellular ROS levels can trigger cell apoptosis [56]. Various Ru(II) complexes have been reported being able to enhance the intracellular ROS levels markedly [29,30]. To explore the roles of ROS in Ru(II) complex-induced cell apoptosis, we detected intracellular ROS levels by using an inverted fluorescence microscope and flow cytometry after staining A549 cells with DCFH-DA fluorescent dye. e results in Figure 5(a) displayed that, for both Ru-1 and Ru-2 groups, with the increase in the treatment concentration, a significant enhancement of DCF fluorescence signals could be observed after 12 h incubation. ese results indicated that Ru-1 and Ru-2 induced overgeneration of ROS in A549 cells in a concentration-dependent manner. e quantitative result in Figure 5(b) exhibited that, at a concentration of 20 μM, coincubation with both Ru-1 and Ru-2 for 12 h could result in significant enhancement of the mean fluorescent intensity (MFI), with about 3.0-and 1.9-fold higher than the control group for Ru-1 and Ru-2, respectively.
As reported, mitochondria are both a source and target of ROS [57]. e complexes studied in this work can induce the decline of mitochondrial membrane potential (MMP) and results in mitochondrial dysfunction, which causes the damage of respiratory chain, and generates radical and nonradical species such as superoxide anion (O 2− ) and hydrogen peroxide (H 2 O 2 ). Based on the above information, these Ru(II) complexes may stimulate the mitochondria-based ROS generation. is speculation correlates with our previous publication [31], and the Ru(II) complex-triggered ROS generation could be effectively blocked by using cyclosporine A (CsA), a confirmed mitochondrial permeability transition pore (MPTP) opening inhibitor.

Complexes Ru-1 and Ru-2 Triggered DNA Damage.
Excessive intracellular ROS can activate DNA damaged sensor proteins [58] and lead to DNA damage, an essential hallmark of cell apoptosis [59]. Indeed, the DNA damage effect has been observed in many previously developed Ru(II) complexes [29,30]. To determine whether or not these complexes can induce DNA damage, the single-cell gel electrophoresis assay (comet assay) was conducted in our study. Comet assay is proved to be a rapid, simple, convenient, and straightforward method for assessing DNA damage in single cells, and the length of the comet tail represents the level of DNA damage. Also, the comet assay (DNA fragments appearance) could serve as a mark for apoptosis assessment. As shown in Figure 6(a), in the control group, there was no comet-like cell observed. After A549 cells were incubated with Ru-1 (10 μM) and Ru-2 (10 μM) for 12 h, the cells presented well-formed comet tails, demonstrating the existence of severe DNA fragmentation. Increasing the concentration to 20 μM recorded markedly higher levels of comet tails, showing that more DNA fragmentation has occurred. For quantitative comparison,    the length of DNA tails in microscopy images was quantified by ImageJ, and the quantification of DNA tails in the comet assay was displayed in Figure 6(b). As expected, complex Ru-1 induced more DNA damage than that induced by Ru-2 based on the same external concentrations (Figure 6(b)). In addition, Ru-1-induced DNA damage was further confirmed by analyzing common DNA damage markers via western blotting assay (Figure 6(c)), as evidenced by the upregulation of the phosphorylation levels of ATM, ATR, histone, and p53 in a concentration-dependent manner. Meanwhile, Chk1 and Chk2 were activated, accompanied by the increased phosphorylation levels of Chk1 and Chk2. e induction of p53 in response to DNA damage is synergistic with ATM/ATR, which can recruit Chk1/Chk2 and subsequently activate downstream cell cycle arrest-associated effector CDC25, causing cell apoptosis [60][61][62]. erefore, these results demonstrated that Ru-1 and Ru-2 can trigger DNA fragmentation, that is, the apoptosis-mediated cell death.
e comet tail observed does not mean that the ruthenium complexes directly damage DNA. As reported, the excess intracellular ROS could attack DNA, resulting in DNA damage [58,59]. So, the observed DNA damage may result from the increased intracellular ROS levels. is has been confirmed by our previous study (the cyclometalated Ru(II) β-carboline complexes could induce DNA damage through ROS overproduction) [29]. In addition, according to our previous studies [29,63,64], we speculate that the binding affinity of the title complexes towards DNA is not very strong because they do not have a large planar aromatic ligand (intercalative), such as dppz, dpq, and pip, so the interaction between them and DNA is not supposed to be the primary reason of apoptosis. e apoptosis-mediated cell death has been commonly measured using the comet assay to detect DNA damage of cells after treatment with complexes. Comets with almost all DNA in the tail are often referred to as "hedgehog" comets and are widely assumed to represent apoptotic cells. In summary, the title complexes could induce mitochondrial dysfunction and the generation of intracellular ROS, which may indirectly lead to DNA damage.

Complex Ru-1 Showed Little Toxicity to Zebrafish
Embryos. Currently, the zebrafish model is attracting unprecedented interests in biomedical research due to its high reproductive rate, short growth period, and high homology  with human DNA [65][66][67].
us, in this work, the developing zebrafish embryos were used to assess the in vivo toxicity of Ru(II) complex. As shown in Figures 7(b) and 7(c), when treated with Ru-1 with the concentration close to IC 50 value (12.5 μM), the cumulative hatch rate and lethality rate display no difference compared with those of the control group. When the concentration of Ru-1 was lower than 50 μM, all zebrafish embryos can develop into juvenile zebrafish after 96 h. Even if the concentration of Ru-1 was up to 100 μM, the cumulative hatch rate was acceptable, and the lethality rate was lower than 35% after 96 h. Furthermore, when the concentration of Ru-1 was lower than 100 μM, no malformation was observed at 96 h (Figure 7(a)). However, when the concentration of Ru-1 was up to 200 μM, even after 72 h treatment, there were lower than 70% cumulative hatch and dramatically increased lethality rate (Figures 7(b) and 7(c)). What is more, unhealthy features such as pericardial cysts and spine curvature were notably observed at 96 h (Figure 7(a)). erefore, the in vivo toxicity of Ru-1 was concentrationdependent, and no apparent side effects were found with the treatment concentrations from 12.5 to 50 μM. e above results indicated that complex Ru-1 showed little toxicity to zebrafish embryos. Since hypotoxicity to normal cells or organs is essential for developing anticancer agents, complex Ru-1 hold great potential to be developed as a lowtoxicity agent against lung cancer cells.

Conclusions
In summary, our study extensively evaluated the anticancer effect of two novel Ru(II) complexes containing O, O-chelated ligands in A549 human lung cancer cells, which was mediated through inducing cell apoptosis. In addition, our results provided evidence that these complexes enhanced the level of intracellular ROS, induced a decrease of MMP, and indirectly led to DNA damage. Also, further studies showed that Ru-1 activated the caspase family proteins and PARP, downregulated the levels of the antiapoptotic protein Bcl-2 and Bcl-xl, upregulated the levels of the proapoptotic proteins Bax and Bad, and induced the release of cytochrome c in A549 cells. More importantly, Ru-1 exhibited low toxicity towards both normal BEAS-2B cells in vitro and zebrafish embryos in vivo. Altogether, complex Ru-1 can induce cell apoptosis via the mitochondrial pathway, which involves mitochondrial dysfunction, ROS accumulation, and caspase-related family members' activation. is work, therefore, suggested that the title Ru(II) complexes have the potential to develop into lung cancer therapeutic agents with safety profiles.

Data Availability
e data used to support the findings of this study are given in the supplementary information file.

Conflicts of Interest
e authors declare that they have no conflicts of interest.