Targeting ROS-AMPK pathway by multiaction Platinum(IV) prodrugs containing hypolipidemic drug bezafibrate

Xin Qiao a, 1, Yu-Yang Gao a, 1, Li-Xia Zheng a, Xiao-Jing Ding a, Ling-Wen Xu a,
Juan-Juan Hu a, Wei-Zhen Gao a, b, **, Jing-Yuan Xu a, *
a Department of Chemical Biology and Tianjin Key Laboratory on Technologies Enabling Development of Clinical Therapeutics and Diagnostics, School of Pharmacy, Tianjin Medical University, Tianjin, 300070, China
b Department of Pharmacology, School of Basic Medical Sciences, Tianjin Medical University, Tianjin, 300070, China


Article history:
Received 8 February 2021 Received in revised form 13 July 2021
Accepted 27 July 2021
Available online 28 July 2021

Pt(IV) prodrugs Hypolipidemic drug Bezafibrate
ROS-AMPK pathway Oxidative stress Apoptosis


Alterations in lipid metabolism, commonly disregarded in the past, have been accepted as a hallmark for cancer. Exploring cancer therapeutics that interrupt the lipid metabolic pathways by monotherapy or combination with conventional chemotherapy or immunotherapy is of great importance. Here we modified cisplatin with an FDA-approved hypolipidemic drug, bezafibrate (BEZ), via the well-established
Pt(IV) strategy, affording two multi-functional Pt(IV) anticancer agents cis,cis,trans-[Pt(NH3)2Cl2(BE- Z)(OH)] (CB) and cis,cis,trans-[Pt(NH3)2Cl2(BEZ)2] (CP) (BEZ ¼ bezafibrate). The Pt(IV) prodrug CB exhibited an enhanced anticancer activity up to 187-fold greater than the clinical anticancer drug cisplatin. Both CB and CP had less toxicity to normal cells, showing higher efficacies and superior
therapeutic indexes than cisplatin. Mechanism studies revealed that the bezafibrate-conjugated Pt(IV) complex CB, as a representative, could massively accumulate in A549 cells and genomic DNA, induce DNA damage, elevate intracellular ROS levels, perturb mitochondrial transmembrane potentials, activate the cellular metabolic sensor AMPK, and result in profound proliferation inhibition and apoptosis. Further cellular data also provided evidence that phosphorylation of AMPK, as a metabolic sensor, could suppress the downstream HMGB1, NF-kB, and VEGFA, which may contribute to the inhibition of angiogenesis and metastasis. Our study suggests that the antitumor action of CB and CP mechanistically distinct from the conventional platinum drugs and that functionalizing platinum-based agents with lipid-modulating agents may represent a novel practical strategy for cancer treatment.

1. Introduction

Metabolic reprogramming has been regarded as one of the ten hallmarks of cancer [1]. Inactivation of tumor suppressors and activation of oncogenes can directly lead to alterations of cellular metabolism [2]. In order to support uncontrolled cell division and stress burden during malignant transformation, cancer cells develop a robust metabolic repertoire that allows the indispensable resilience to survive, proliferate, and escape drug toxicity [3].

Deregulation of lipid metabolism, commonly disregarded in the past, has been accepted as a significant aspect in cancer, which has been demonstrated to be involved in carcinogenesis, cancer metastasis, and aggressiveness in various cancers [4,5]. Moreover, alterations in lipid metabolism have tremendous therapeutic im- plications as lipids influence cell energy availability, membrane fluidity and dynamic, signal transductions, therapy response, and drug resistance [3,6].

Given the importance of lipid metabolism to cancer progression and therapy response, the development of cancer therapeutics that interrupt the lipid metabolic pathways by monotherapy or com- bination with conventional chemotherapy or immune therapies is of great importance. Indeed, the lipid-lowering drugs, statins, also known as HMG-CoA inhibitors, demonstrated antitumor effects by inhibiting inflammation, angiogenesis, and proliferation and inducing apoptosis [7,8]. The drug combinations of statins and cy- totoxins (such as cisplatin [9], doxorubicin [10], etc.) are endowed with antiproliferative and proapoptotic effects and impairing migration and invasion in cancer cells. Another group of hypolipi- demic drugs, fibrates, are well-known ligands of peroxisome proliferator-activated receptors a (PPARa), which have been proved to suppress cancer cell proliferation and migration in serval cancers [11]. In particular, lipid regulators, both statins and fibrates, were demonstrated to alleviate nephrotoxicity, ototoxicity, and vascular endothelial dysfunction induced by chemotherapeutic agents, like platinum agents [12e14].

The platinum-based anticancer drugs, represented by cisplatin, are widely employed in clinical chemotherapy, but their clinical applications are limited by systemic toxicity and drug resistance [15,16]. In exploring the next generation of platinum drugs, the Pt(IV) scaffolds have been regarded as promising alternatives to mitigate the disadvantages of conventional Pt(II) counterparts [17,18]. The kinetically inert low-spin d6 octahedral geometry in the Pt(IV) center allows for tuning the oral availability, antitumor ac- tivity, and toxicity of conventional Pt(II) drugs [16,19]. Moreover, under the intracellular reduction environment, the Pt(IV) prodrug would be activated by cellular reductants, releasing two axial li- gands and one cytotoxic Pt(II) moiety to produce DNA damage and transcription inhibition [17,18]. The adjustable axial ligands could confer desired pharmacological properties to the Pt(IV) prodrug, such as improving lipophilicity, increasing solubility, overcoming drug resistance, tumor-targeting ability, and incorporating orthogonal or complementary bioactive moieties [17,20]. In this context, the well-established Pt(IV) strategy permits to design of dual- or multi-functional candidates.

Inspired by the success of current lipid metabolism regulators in cancer therapy, we propose that interrupting the lipid metabolic pathways might be an attractive strategy to overcome the drawbacks of conventional platinum therapies. On the one hand, cancer cells are likely “addicted” to the metabolic reprogramming of lipids and, therefore, vulnerable to interruption of lipid biogenesis [4,5,21]. This addiction to lipids may provide a tumor-targeting approach for novel platinum drug development. On the other hand, the protective role of clinical lipid-lowering drugs against Pt-drug-induced nephrotoxicity, ototoxicity, and vascular endothelial damage may show significant benefits in the clinic and next-generation platinum-based drug design [12e14]. Thus, by taking advantage of the potential synergy between Pt drugs and the clinical used lipid-lowering drugs, we employed the FDA-approved hypolipidemic drug bezafibrate (BEZ) to modify platinum agents via the well-established Pt(IV) chemistry, affording two multi-functional Pt(IV) anticancer agents cis,cis,trans- [Pt(NH3)2Cl2(BEZ)(OH)] (CB) and cis,cis,trans-[Pt(NH3)2Cl2(BEZ)2] (CP) (Scheme 1) with the ability to enhance cytotoxicity in cancer cells significantly. Mechanistic studies reveal that the bezafibrate- conjugated Pt(IV) prodrugs enhance the platinum accumulation and DNA platination, induce DNA damage, increase intracellular reactive oxygen species (ROS) levels, perturb mitochondrial trans- membrane potentials, activate the cellular metabolic sensor AMPK, and inhibit cancer cell proliferation profoundly. The activation of AMPK phosphorylation might, in turn, induce suppression of ex- pressions of HMGB1, NF-kB, as well as VEGFA, which may contribute to the inhibition of angiogenesis and metastasis.

2. Results and discussion
2.1. Synthesis and characterization of bezafibrate-Pt(IV) prodrugs

Two new bezafibrate-Pt(IV) prodrugs were synthesized ac- cording to previously established protocols [22e25]. Cisplatin, as starting material, was oxidized by hydrogen peroxide to generate Pt(IV) intermediate product oxoplatin (c,c,t-[PtCl2(NH3)2(OH)2]). The condensation reaction of bezafibrate with oxoplatin was cata- lyzed by TBTU, trimethylamine, and dimethylformamide (DMF) at room temperature in DMSO to afford CB and CP, respectively (Scheme 2). The bezafibrate-conjugated Pt(IV) prodrugs were characterized by 1H and 13C NMR spectroscopy and ESI-HRMS, and the analytical purity of the compounds was verified by HPLC (>95 %) (Figs. S1eS8).

2.2. In vitro cytotoxicity evaluation of the Pt(IV) complexes

The half-maximal inhibitory concentrations, IC50 values, for 72 h treatment were determined for CB and CP by MTT assay, along with the clinically employed reference drugs cisplatin and bezafibrate for comparison purposes. Human cancer cell lines including HeLa (cervical), MCF-7 (breast), NCIeH460 (lung), A549 (lung), and normal human cell lines LO2 (liver) and HUVEC (umbilical vein endothelial) were selected to evaluate the in vitro cytotoxicity of the compounds (Table 1). As shown in Table 1, bezafibrate showed modest cytotoxicity when employed as a single agent, with IC50 values greater than 100 mM in all the cell lines tested. The other reference drug, cisplatin, displayed low micromolar toxicity, which was also very toxic to human normal liver cells LO2 and human umbilical vein endothelial cells HUVEC. We next evaluated the cytotoxicity of a mixture of cisplatin and bezafibrate at a molar ratio of 1:1 to test whether bezafibrate was able to potentiate cisplatin. In most cancer cell lines, the mixture showed increased cytotoxicities. For example, in A549 cells, cisplatin and a mixture of cisplatin and bezafibrate exhibited IC50 values of 7.49 mM and 0.83 mM, respectively. Thus, the mixture has a 9-fold increased cytotoxicity in A549 cells. In the two tested human normal human cell lines, however, the mixture’s cytotoxicities were slightly reduced compared with cisplatin. These data indicate that bezafi- brate might potentiate the anticancer activity of cisplatin in cancer cells and improve the selectivity of cisplatin, further confirming our hypothesis that incorporation of Pt(IV) prodrugs with lipids modulating agents might be an effective strategy to elevate the anticancer activity and reduce the toxicity of cisplatin.

Scheme 1. Structures of cisplatin, oxoplatin, bezafibrate, CB, and CP.

Scheme 2. Synthetic routes for CB and CP.

The cytotoxicities of CB and CP were subsequently examined. Compared with a mixture of cisplatin and bezafibrate, both CB and CP afforded significantly increased cytotoxicities in cancer cells. The dicarboxylated Pt(IV) prodrug CP was more active than cisplatin in all the tested cancer cell lines with IC50 values in the low micro- molar or submicromolar range. Notably, the monocarboxylated Pt(IV) prodrug CB was significantly more active than cisplatin or CP. The IC50 values of CB were in the nanomolar range in all the cancer cells tested. For example, the IC50 values of CP in A549 and HeLa cells were 0.15 and 0.35 mM, respectively, while those of CB were as low as 0.04 and 0.06 mM, respectively. Compared with cisplatin, CB displayed up to a 187-fold increase in cytotoxicity in the cells tested (Table 1). Unexpectedly, although CB and CP showed striking cy- totoxicities in cancer cells, their selectivity significantly increased (Table 1). The selectivity index, defined as the ratio of the IC50 value in normal liver cells LO2 to that in A549 cells, was 0.21 for cisplatin, and the values increased to 6.25 (30-fold) and 11.07 (53-fold) for CB and CP, respectively. This emphasizes the critical contribution of glucose and lipid metabolism regulator bezafibrate towards the selectivity of platinum-based agents.

2.3. Cellular uptake, DNA platination, lipophilicity, and reduction of the Pt(IV) complexes

Mechanistic investigations were performed to study the possible reasons why CB and CP have increased cytotoxicity than cisplatin. We first tested if the cytotoxic effects produced by CB, CP, and cisplatin in A549 cells correlated with the intracellular Pt content. Cellular accumulation studies were carried out to deter- mine the platinum levels in whole-cell samples. A well-defined number of A549 cells were treated with 10 mM CB, CP, cisplatin, or the combo of cisplatin and bezafibrate for 6 h, and the cell samples were lyophilized for Pt content measurement. The Pt levels in each group were determined by quadrupole inductively coupled plasma mass spectrometry (ICP-MS). As shown in Fig. 1a and Table 2, after 6 h incubation, the platinum levels in A549 cells were 24.22 and 23.89 ng Pt/106 cells for cisplatin alone and the cisplatin and bezafibrate combo group, respectively, suggesting that co- incubation with bezafibrate did not enhance nor compromise the cell uptake process of cisplatin. By contrast, CB was taken up most efficiently into A549 cells, with Pt levels reaching about 13.6-fold higher than that of cisplatin alone and combination treatment, respectively. These significantly elevated intracellular Pt accumu- lation levels are considered as one of the primary reasons for the remarkable cytotoxicity of CB. The dicarboxylated Pt(IV) prodrug CP presented less platinum accumulation in A549 cells than its mon- ocarboxylated counterpart, with cellular uptake levels about 1.3- fold higher than cisplatin suggesting that the cellular entrance of CP by passive diffusion may not be as efficient as that of CB.
Platination of genomic DNA is considered as a critical event in the cell-killing action of platinum-based anticancer agents. The elevated platinum accumulation of CB and CP in A549 cells promoted us to investigate whether the enhanced platinum uptake positively relates to their DNA platination. The Pt content in genomic DNA was measured after DNA isolation from Pt complexes treated A549 cells. As shown in Fig. 1b, A549 cells treated with 10 mM CB, CP, cisplatin, or the combo of cisplatin and bezafibrate for 6 h showed 3.7, 2.7, 2.0, and 1.9 ng Pt per mg DNA, respectively, which are consistent with the Pt content in the whole-cell samples. These data suggest that the DNA platination, as well as the cyto- toxicity, are roughly proportional to the intracellular accumulation of the Pt(IV) complexes, and the higher cytotoxicity of CB and CP could be correlated with their higher DNA platination.

Fig. 1. Intracellular (a) and genomic DNA (b) platinum content in A549 cells treated with 10 mM CB, CP, cisplatin, or the combo of cisplatin and bezafibrate for 6 h. The data are presented as the mean ± standard deviation of three independent experiments. ****P < 0.0001, or * P < 0.05 compared with the cisplatin-treated group. To further explain the difference between bezafibrate-derived Pt(IV) prodrugs and cisplatin in platinum accumulation in A549 cells, lipophilicity was considered the first parameter. It has been commonly believed that Pt(IV) complexes enter the cells mainly by passive diffusion, and lipophilicity plays a critical role in the cellular uptake of Pt(IV) compounds [26,27]. Thus, we first attempted to determine the partition coefficients (log Po/w values) by the traditional octanol-water shake-flask method to examine whether a correlation exists between lipophilicity and cellular Pt uptake. However, like other fibrate-derivative Pt(IV) complexes, the low solubility of the bezafibrate-Pt(IV) complexes led to unreliable results [20]. Therefore, we applied for the Molinspiration program ( to predict the log Po/w values of Pt (IV) complexes. As expected, the presence of lipophilic bezafibrate in Pt(IV) prodrugs dramatically increased the calculated log Po/w values (Table 2), followed the trend in lip- ophilicity as cisplatin < CB < CP < bezafibrate. This result demon- strated that bezafibrate conjugating Pt(IV) complexes could increase the lipophilicity in the Pt complexes. In particular, the improved lipophilicity in CB resulted in the profoundly enhanced cellular Pt levels and potent cytotoxic effect in A549 cells. In contrast, the dicarboxylated Pt(IV) prodrug CP displayed a higher calculated log Po/w value than CB, which might compromise the cellular uptake process and reduce cytotoxicity. The cell uptake results also indicated that the elevated cellular uptake due to increased lipophilicity might not be the only mechanism for significantly enhanced cytotoxicity for this type of Pt(IV) complexes. Under the intracellular reduction environment, the Pt(IV) pro- drug would be activated by cellular reductants, such as ascorbic acid and glutathione, which exist at a much higher intracellular concentration than in extracellular environments [28]. Reduction of the Pt(IV) complexes, in concert with the loss of the axial ligands, is generally accepted to be critical for the antitumor activity of Pt(IV) prodrugs [28,29]. To mimic the intracellular reduction environ- ment, the bezafibrate-conjugated Pt(IV) complexes were incubated with excess reductant ascorbic acid (10 eq.), and the signs of progress were monitored by HPLC (Fig. S9). From the HPLC traces at 0 h treatment, CB and CP were still kept intact in the presence of 10- fold excess ascorbic acid, with the retention time was different from that of bezafibrate (Fig. S10). This result ruled out the possibility that the bezafibrate Pt(IV) complexes might have undergone complete reduction at the earliest time point. Following the traces with chromatograms of pure bezafibrate, we found that the most potent CB could gradually release the axial ligand bezafibrate upon reduction as time progress (Fig. S9), suggesting that, under the intracellular reduction environment, bezafibrate could be easily liberated from the monocarboxylated Pt(IV) complex CB and pro- duce profound synergistic anticancer activity. In contrast, the dicarboxylated Pt(IV) prodrug CP was barely reduced by ascorbic acid after 48 h (less than 3 %), even upon treatment with 10 eq. ascorbic acid (Fig. S9). This can be expected from the reduction nature of dicarboxylato Pt(IV) center. Unlike the monocarboxylated CB, complex CP has two carboxylate bezafibrate ligands in the axial positions, which are unlikely to serve as bridging groups for elec- tron transfer [30]. Recent electrochemical and computational studies have demonstrated that the process of electron transfer and Pt ligand bond cleavage occurs in a stepwise pattern, leading to the generation of a metastable hexacoordinated Pt(III) intermediate upon addition of the first electron [31]. As a consequence of the second electron transfer, the two axial ligands are detached sequentially. Density functional theory (DFT) calculation suggests that the addition of an electron to the Pt(III) intermediate is much easier than the addition to the Pt(IV) complex [31,32]. Under the ascorbic acid reduction condition, the axial hydoxido ligand in CB might act as a bridge that interacts with the deprotonated oxygen of the ascorbic acid, and a proton shift might concurrently occur from the OH group of the ascorbate to the trans bezafibrate ligand [31,32]. As a result, the axial hydoxido ligand are released along with bezafibrate. Unlike the monocarboxylated CB, the two beza- fibrate ligands in the axial position are sterically more challenging in the case of CP. The reduction may be hindered due to the two longer bezafibrate ligands protecting the Pt(IV) center. In addition, the two bulky methyl groups adjacent to the axial ester groups of the Pt(IV) center might increase the steric hindrance of electron transfer, which may contribute to the ascorbic acid reduction resistance of CP [33,34]. This result is consistent with previous findings from other groups [35,36]. Moreover, this high resistance of CP to reduction by ascorbate was not reflected in compromised cytotoxicity (see Table 1). However, the small amount of liberated bezafibrate would likely play a powerful synergistic effect, perhaps balancing the effect of slower reduction. Unlike most of the re- ported Pt(IV) anticancer complexes, the increased resistance of CP to reduction by ascorbate acid would also be expected to lead to a great chance of bypassing many of the deactivation and reduction before arriving at the target site [37]. 2.4. DNA damage studies To further reveal whether the bezafibrate-conjugated Pt(IV) complexes CB and CP were able to cause DNA damage, comet assay and immunodetection of g-H2AX foci were performed (Fig. 2). As expected, lipid-lowing agent bezafibrate treatment (5 mM) for 24 h did not induce DNA damage (Fig. 2a), in which the nuclei were nearly round and contained like the untreated control group. However, treatment of A549 cells with 5 mM of CB, CP, cisplatin, or the combo of cisplatin and bezafibrate for 24 h resulted in typical comet tails indicating DNA double-strand breaks. The mono- carboxylated Pt(IV) complex CB induced more severe DNA frag- mentation than cisplatin, CP, and the combo group. To further quantify the extent of DNA damage, olive tail moments (OTM) of the comets were calculated by Comet Assay Software Project (CASP) image analysis program. As shown in Fig. 2b, the OTM values in CB-treated A549 cells were 4.95-fold higher than those in cisplatin-treated cells, which confirms that the monocarboxylated Pt(IV) complex CB causes more DNA damage than cisplatin at the same concentration in line with its higher DNA platination and cytotoxicity. In addition, to further evaluate the DNA damage effects of Pt complexes in A549 cells, immunofluorescence detection of g-H2AX foci was performed. Phosphorylation of histone H2AX on serine- 139 in response to DNA double-strand breaks (DSBs) can result in the g-H2AX foci, which is proportional to the amount of DSBs [38,39]. A549 cells were assessed for the presence of g-H2AX foci after 6 h of exposure to 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate, respectively. Compared with the control group, significant increases in g-H2AX fluorescence and foci formation were observed in A549 cells treated with CB, CP, cisplatin, and the combo group (Fig. 2cee), which is consistent with the comet assay data. Fig. 2. Induction of DNA damage by platinum complexes. A549 cells were treated either with PBS, or with 5 mM CB, CP, cisplatin, bezafibrate, and the combo for 24 h followed by (a) visualization of DSBs using comet assay and (b) quantification of the corresponding tail moment of the cells using comet assay software project (CASP) image-analysis program across three independent experiments. (c) Representative confocal microscopy images of A549 cells immunostained with anti-g-H2AX (green) and (d) quantification of the mean fluorescence value of g-H2AX across three independent expriments. The nuclei were stained by DAPI (blue). A549 cells were untreated or treated with 10 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 6 h. (e) Representative g-H2AX foci images. ****P < 0.0001, ***P < 0.001, or * P < 0.05 compared with the cisplatin-treated group. 2.5. Reactive oxygen species (ROS) measurement As a peroxisome proliferator-activated receptor a (PPARa) agonist, bezafibrate has been reported to induce a rapid and sus- tained generation of reactive oxygen species (ROS) with associated lipid oxidation in acute myeloid leukemia (AML) cells [40]. More- over, cisplatin has been reported to show multiple actions in cells, for example, such as elevating intracellular reactive oxygen species (ROS) levels and inducing oxidative stress, and excess accumulation of ROS has been considered as one essential mechanism for the cancer cell killing process of platinum-based agents [41]. To eluci- date the effect of the tested platinum complexes on ROS accumu- lation, we evaluated the intracellular steady-state ROS levels in platinum agents treated A549 cells using DCFH-DA assay by flow cytometry. A549 cells were incubated with control medium, cisplatin, bezafibrate, the combination of cisplatin plus bezafibrate, CB, or CP for 24 h, followed by measurement of intracellular ROS. As shown in Fig. 3a and c, incubation with 5 mM bezafibrate, cisplatin, or the mixture of cisplatin and bezafibrate slightly increased intracellular ROS in A549 cells, suggesting that co-incubation of the cells with bezafibrate produced no effect on cisplatin-induced ROS production in A549 cells. As expected, 5 mM of the mono- carboxylated Pt(IV) prodrug CB significantly elevated the intracel- lular ROS levels reaching approximately 8-fold higher than that of the cell control group. The less potent dicarboxylated Pt(IV) pro- drug CP demonstrated lower intracellular ROS accumulation than its monocarboxylated counterpart. However, the intracellular ROS levels were 1.85-fold higher than cisplatin or the combo of cisplatin and bezafibrate. The reduction of Pt(IV) compounds is per se responsible for intracellular ROS generation [42]. The conspicuous enhancement of intracellular basal ROS levels in CB or CP treated A549 cells could also be contributing to their remarkable cytotoxicity. 2.6. Mitochondrial transmembrane potential (DJm) assessment The evidence of increases in intracellular ROS accumulation in CB or CP treated A549 cells promoted us to widen our analysis to investigate the effect of Pt(IV) complexes on mitochondria depo- larization. The loss of mitochondrial transmembrane potential (DJm, MMP) is regarded as a common feature of cell death [43,44]. The fibrate lipid-lowering agents have been reported to inhibit mitochondrial respiration and impair mitochondrial function in different cancer types [45e49]. These mitochondrial actions could switch the metabolism from glycolysis to fatty acid b-oxidation and subsequently induce the AMPK activation and cytotoxicity [11,50]. Moreover, the physicochemical features of fibrates enable them to combine with some hydrophobic components of the mitochondrial electron respiratory chain and promote permeation, accumulation, and interaction with components of the inner mitochondrial membrane, which may result in oxidative metabolic stress inde- pendently with PPAR agonism [50]. Therefore, the ability of mito- chondria to maintain MMP after incubation with platinum agents was assessed in A549 cells using the fluorescent probe JC-1. In the mitochondria matrix, JC-1 aggregates as a polymer and produces red fluorescence when the MMP is high. At the same time, it exists as a monomer to give off green fluorescence when MMP is low. A549 cells were treated with 5 mM of CB, CP, cisplatin, bezafibrate, the combo of cisplatin and bezafibrate, respectively, for 24 h, stained with JC-1, and analyzed with flow cytometry to evaluate MMP. As shown in Fig. 3b and d, after 24 h incubation, A549 cells incubated with CB remarkedly decreased the DJm with an evident increase in green fluorescence (JC-1 monomer) compared with the control group, and the percentage of cells with inactivated MMP was 38.8 %. The less active CP was also capable of reducing the MMP (17.5 %), exhibiting the significant disruption of MMP in comparison with cisplatin (12.8 %), bezafibrate (5.4 %), and the combo group (14.9 %). These data suggested that CB could influence the mito- chondria function by reducing the MMP in A549 cells. Given the mitochondrial affinity features of fibrates and mitochondrial disruption effects in cancer cells [50], we speculated that the decrease of MMP induced by CB and CP might result from the mitochondrial affinity feature of bezafibrate and the elevated intracellular ROS production in A549 cells. 2.7. Inhibition of cancer cell migration and invasion by Pt(IV) complexes Since cell migration and invasion are the characteristic features of lung cancer metastasis [51], we then employed wound healing and Matrigel invasion assays to evaluate whether CB and CP affected cell migration and invasion ability. As shown in Fig. 4a and c, scraped cell images were captured at 0, 24 h, and the distance moved by wounded cell monolayer was obtained. CB significantly inhibited the migration of A549 cells with a lower healing rate (7.7 %) than those of control (42.7 %), CP (12.3 %), cisplatin (23.1 %), bezafibrate (47.2 %), and the combo of cisplatin and bezafibrate (20.4), respectively. Degradation and modification of extracellular matrix (ECM) to penetrate basement membrane are critical fea- tures for cancer invasion. The cell invasion assay showed that, after 24 h treatment with different compounds (2.5 mM), the remaining viable cells which had invaded into the Matrigel were visualized by crystal violet staining, and the percentages of invasion cells decreased dramatically (Fig. 4b and d). Compared with the control group (100 %), CB reduced the invasion rate to 28.8 %, CP to 34.4 %, cisplatin to 69.1 %, bezafibrate to 81.4 %, and the combo group to 44.3 %, respectively. These data suggested that CB most effectively suppressed A549 cells metastasis and invasion. 2.8. Influence of Pt(IV) complexes on cell cycle distribution and apoptosis Next, the impact of the compounds on cell cycle distribution was analyzed by flow cytometry. A549 cells were treated with 5 mM cisplatin, bezafibrate, the mixture of cisplatin and bezafibrate (1:1), CB or CP for 24 h, then cell cycle distribution was analyzed. Cisplatin, bezafibrate, and the mixture of cisplatin and bezafibrate arrested the cell cycle at the S phase. In the control group, 22.9 % of the cells were in the S phase, while the percentages of S phase cells increased to 29.7 % (cisplatin), 38.7 % (bezafibrate), and 30.93 % (the mixture), respectively. Cells treated with 5 mM CB or CP displayed larger S-phase populations after 24 h incubation compared with cisplatin, indicating strong S-phase arrest (Fig. 5a and c). We also carried out annexin V/propidium iodide (PI) double staining by flow cytometry to quantify apoptosis induced by the platinum- based agents in A549 cells. Results showed that CB could induce significantly more apoptotic cells (15.7 %) than cisplatin (8.7 %) (Fig. 5b and d). This result exhibited that CB effectively induced apoptosis in A549 cells. Fig. 3. (a, c) Effects of platinum compounds on intracellular ROS generation in A549 cells. A549 cells, untreated or treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h, were stained with DCFH-DA and analyzed by flow cytometry (a). Rosup (10 mg/mL) was used as a positive control. The percentage of ROS positive (red) or negative cells (black) was plotted and compared. (b, d) Effects of platinum compounds on mitochondrial membrane potential (MMP, DJm) in A549 cells. A549 cells, untreated or treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h, were stained with JC-1 and analyzed by flow cytometry (b). Carbonyl cyanide 3-chlorophenylhydrazone (CCCP, 10 mM) was used as a positive control. The percentage of cells with inactivated MMP (green) or normal MMP (red) was plotted and compared. 2.9. Pt(IV) complexes effectively regulated p-AMPK, NF-kB, HMGB1, and VEGF expression So far, the current data have proved that the bezafibrate modi- fied Pt(IV) complexes are capable of accumulating the intracellular ROS level, decreasing the MMP, inhibiting cell migration and in- vasion, arresting the cell cycle at S phase, and inducing apoptosis. These data promoted us to speculated that targeting cancer cell metabolism might play a critical role in the cell-killing process of the bezafibrate modified Pt(IV) complexes. Increasing evidence has suggested that phosphorylation of AMP-activated protein kinase (AMPK) [52] plays a critical role in the lipid biosynthesis pathway and can initiate downstream pathways to inhibit proliferation, in- vasion, and migration of human cancer cells. For example, fenofi- brate showed to inhibit the invasion and migration of oral cancer CAL 27 cells by the nuclear factor-kB (NF-kB) pathway mediated through AMPK signaling but not PPARa [53]. Thus, the expression of the cellular metabolic sensor, AMP-activated protein kinase (AMPK), was examined by Western blot. The A549 cells were treated with 5 mM Pt-drugs for 12 h, and the expression levels of AMPK and phosphorylated AMPK (p-AMPK) were assessed. As shown in Fig. 6a and c, AMPK phosphorylation was increased by CB and CP treatment, and the normalized p-AMPK levels upon Pt(IV) complex treatments were 0.99 and 0.75, respectively, while the levels were 0.47 and 0.66 for the same concentration of cisplatin and the combo group, respectively (Fig. 6a and c). This result further corroborated our hypothesis that the increased anticancer activity of CB was from AMPK activation [54,55]. Given that the elevated ROS accumulation and decreased MMP in CB treated cancer cells, it was speculated that the ROS-mediated AMPK acti- vation might contribute to the apoptosis induced by CB. Like other fibrates, the bezafibrate conjugated Pt(IV) complex CB might inherit the mitochondrial affinity features and mitochondrial disruption effects of bezafibrate, which might subsequently in- crease intracellular ROS production, activate the metabolic sensor AMPK, and lead to profound cytotoxicity in A549 cells. Moreover, inflammation plays a crucial role in cancer and metabolic diseases, such as diabetes, hyperlipidemia, and altered energy metabolism is a hallmark of cancer cells [56,57]. Given that the bezafibrate modified Pt(IV) complexes could activate the AMPK,we, therefore, examined whether AMPK activation induced by CB could regulate the downstream nuclear factor-kB (NF-kB) by Western blot and immunofluorescence evaluation. As shown in Fig. 5a and c, incubation with 5 mM CB or CP for 12 h in A549 cells demonstrated significant decreases in the expressions of NF-kB by 58.4 % (CB) and 50.7 % (CP), respectively. Immunofluorescence images also displayed that, after 6-h incubation, 5 mM Pt(IV) com- plexes treatments could induce noticeable decreases of NF-kB levels in the nucleus and cytoplasm (Fig. 6b). On the contrary, in- cubation with the same concentration of cisplatin, bezafibrate, or the combo of bezafibrate and cisplatin did not cause inhibition of NF-kB expression. Recent studies have revealed that the inflam- matory transcription factor, NF-kB, is associated with chemo- therapy response and drug resistance in lung cancer, and blockage of NF-kB is capable of impeding cancer cell survival, metastasis, and resistance to anticancer agents [56,57]. In a previous study, another fibrate lipid-lowering agent, fenofibrate, has been found to inhibit cell proliferation in A549 cells via suppression of NF-kB activity [58]. This inhibition is independent of the PPARa receptor [58]. The above results provided evidence that bezafibrate conjugated Pt(IV) may activate AMPK phosphorylation and inhibit NF-kB activity. Fig. 4. (a) Suppression of cell migration in A549 cells treated with platinum compounds. After overnight attachment, a straight scratch on the monolayer was created across the plate using a sterilized 200 mL pipette tip, and A549 cells were treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h. Representative images of closure of scratch under each treatment conditions were shown. (b) Suppression of cell invasion in A549 cells treated with platinum compounds. A549 cells were treated with 2.5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h. Representative images of crystal violet stained cells were shown. (c) Statistical analysis of wound healing rates in compound treated A549 cells. (d) Statistical analysis of cell invasion. ***P < 0.001, *P < 0.05, compared with the control group. Moreover, dysfunction of high mobility group box 1 (HMGB1) is believed to associate with the hallmarks of cancer and contribute to regulating multiple signal pathways, including proliferation, inflammation, genome stability, metastasis, metabolism, and apoptosis [59,60]. Thus, overexpression of HMGB1 may contribute to chemotherapy drug resistance and tumor growth in lung cancers [60,61]. Moreover, suppression of intracellular HMGB1 increases apoptosis and can improve the effectiveness of chemotherapy [60]. We, therefore, investigated the expressions of HMGB1 in compounds treated A549 cells. As shown in Western blot analysis, 5 mM CB incubation for 12 h profoundly suppressed HMGB1 expression in A549 cells, which led to 11.3-fold less than that of the control group (Fig. 6a and c). Fig. 5. (a) Effects of platinum compounds on the cell cycle progression in A549 cells. A549 cells, untreated or treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h, were harvested, fixed by ethanol, and stained with propidium iodide (PI). Cell cycle distribution was analyzed by flow cytometry. (b) Apoptosis induced by platinum compounds. A549 cells, untreated or treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 24 h, were then processed for Annexin V/PI double staining and analyzed by flow cytometry. (c) Comparisons of cell cycle distribution in untreated or platinum compounds treated A549 cells. (d) Comparisons of apoptosis rate in A549 cells. Annexin V-positive/PI-negative cells are in the early stages of apoptosis and double positive cells are in late apoptosis. Given that the most potent Pt(IV) compound CB could activate the ROS-AMPK pathway and induce oxidative stress, the significant suppression of HMGB1 expression could probably decrease DNA repair efficiency in response to oxidative stress, ultimately increase DNA damage. Besides, immunofluorescence observation revealed that 6 h incubation with 5 mM CB could elevate the cytoplasm HMGB1 expression, which indicated that increased endogenous ROS in CB treated A549 cells may also promote the HMGB1 trans- location and release (Fig. 6d). Once released, HMGB1 may activate the NF-kB pathway to induce proangiogenic growth factors, such as vascular endothelial growth factor (VEGF) [62]. In consistence with the decreases in HMGB1 and NF-kB expressions in CB treated A549 cells, 5 mM CB treatment in A549 cells showed a prominent down-regulation in VEGFA expression with a 13.12 % decrease compared with the control group (Fig. 6a and c). Unlike cisplatin, the less potent dicarboxylated Pt(IV) prodrug CP could also lead to significant suppression of expressions of HMGB1, NF-kB, as well as VEGFA. Using the immunofluorescence method, shorter in- cubations with 5 mM CB or CP displayed a dramatic loss of VEGFA fluorescence signals (Fig. 6b). Moreover, hypoxia and oncogene activation may trigger the overexpression of VEGF. Gabano and co-workers reported two diclofibric acid Pt(IV) complexes. The most potent complex 2 ex- hibits excellent performances under hypoxic conditions and di- minishes hypoxia-inducible factor-1a (HIF-1a) [20], a well-known regulator of angiogenesis and metastasis. These findings, as well as our results from wound healing and Matrigel invasion assays, confirmed that Pt(IV) complexes equipped with bezafibrate moiety might have great potential for inhibiting tumor angiogenesis and metastasis. Previous studies reported that fibrate hypolipidemic agents could inhibit multiple cancers by inducing apoptosis and cell cycle arrest, inhibiting tumor invasion and migration via several complex pathways, and the antiproliferative and anti-migratory effects of fibrates in different cancers were dependent or independent of PPARa [11]. Moreover, the fibrates, as PPARa agonists, could pro- mote fatty acid b-oxidation, switch energy metabolism from glucose to fatty acid utilization, and lead to activation of AMPK [11,50]. In addition, the fibrates could also induce intracellular ROS accumulation due to the elevated peroxisomal b-oxidation [50]. In this context, factionalizing Pt(IV) anticancer agents with fibrates hypolipidemic agents may result in multiple anticancer actions. Indeed, as stated above, the bezafibrate containing Pt(IV) com- plexes CB and CP had been proved to accumulate into A549 cells, lead to elevate intracellular ROS levels, perturb mitochondrial transmembrane potentials, activate the cellular metabolic sensor AMPK, and result in profound proliferation inhibition in A549 cells. The activation of AMPK phosphorylation might, in turn, induce suppression of expressions of HMGB1, NF-kB, as well as VEGFA, which may contribute to the inhibition of angiogenesis and metastasis (Fig. 7). These results confirmed that the bezafibrate modified Pt(IV) complexes could retain the multi-action properties of PPARa agonist bezafibrate, and may show synergistic anticancer effects via multiple anticancer pathways. Fig. 6. (a) Western blot analysis of AMPK, p-AMPK, NF-kB, VEGFA, and HMGB1 in A549 cells. A549 cells were untreated or treated with 5 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 12 h. Relative expression levels of p-AMPK, NF-kB, VEGFA, and HMGB1 are presented as the mean ± standard deviation of relative grayscale values from two independent experiments (c). (b) Representative confocal microscopy images of A549 cells immunostained with anti- NF-kB (green) and anti-VEGFA (red). The nuclei were stained by DAPI (blue). A549 cells were untreated or treated with 10 mM CB, CP, cisplatin, bezafibrate, or the combo of cisplatin and bezafibrate for 6 h. (d) Repre- sentative confocal microscopy images of A549 cells immunostained with anti-HMGB1 (red). Cells were co-stained with DAPI. The yellow boxes were used to compare the dif- ferences between two groups. 3. Conclusions In summary, we synthesized, characterized, and biologically evaluated two Pt(IV) prodrugs CB and CP containing lipid modu- lating agent bezafibrate. Compared with cisplatin or a mixture of cisplatin and bezafibrate, CB and CP displayed significantly increased cytotoxicity in several cancer cells, and the complexes were less toxic to normal cells, showing higher efficacies and su- perior therapeutic indexes compared with cisplatin. Mechanistic studies revealed that the bezafibrate-conjugated Pt(IV) complex CB, as a representative, could massively accumulate in A549 cells and genomic DNA, induce DNA damage, elevate intracellular ROS levels, perturb mitochondrial transmembrane potentials, activate the cellular metabolic sensor AMPK, and result in profound prolifera- tion inhibition, cell cycle arrest in S phase, and apoptosis in A549 cells. Our cellular data also provided evidence that phos- phorylation of AMPK could suppress the downstream HMGB1, NF- kB, and VEGFA, which may contribute to the inhibition of angio- genesis and metastasis. We cannot exclude the possibility that the hydrophobic feature of bezafibrate moiety in CB and CP may also contribute to the elevated cytotoxicity in cancer cells. However, the unique property of CB on ROS accumulation and MMP disruption, together with the AMPK activation and regulations of downstream effectors, clearly indicated that the hypolipidemic effect of bezafi- brate in the axial position of Pt(IV) complexes plays a vital role in the unique cytotoxicity of CB and CP, not only by facilitating the cell entrance. Nevertheless, the roles of PPARa in the antiproliferation, inhibition of tumor invasion and metastasis, and apoptosis induc- tion of bezafibrate-functionalized Pt(IV) complexes remain to be explored. The increased resistance to ascorbic acid reduction is also worthy of further investigation, which may offer great potential for targeted Pt(IV) prodrug design. Future work will also address the detailed mechanistic investigations on the mitochondrial functions of these new Pt(IV) complexes. The study has also highlighted the importance of functionalizing platinum-based agents with lipids modulating to perturb lipid metabolic pathways as a cancer- specific strategy for novel platinum anticancer candidate devel- opment. Additional in vivo evaluations of CB and CP and further mechanistic studies are currently underway. 4. Experimental section 4.1. Materials and physical measurements All of the reagents were used as received from commercial sources without further purification unless indicated otherwise. Cisplatin was purchased from Boyuan Pharmaceutical Co., Ltd (Shandong, China). Bezafibrate was obtained from Energy Chemical (Shanghai, China). All of the reactions were carried out under normal atmospheric conditions in the dark. Cell culture media (DMEM, RPMI 1640 and McCoy's 5A) and fetal bovine serum (FBS) were from Solarbio. 1H and 13C NMR spectra were recorded in CDCl3 or DMSO‑d6 on a Bruker AVANCE III 400 MHz spectrometer. Chemical shifts (d) were expressed in parts per million (ppm) relative to an internal standard (tetramethylsilane, TMS). 1H NMR data are reported in the conventional form including chemical shifts (d, ppm), multiplicity (s singlet, d doublet, t triplet, q quartet, m multiplet, br broad), coupling constants (Hz), and relative integral signal intensities. The NMR spectra were processed and analyzed using the MestReNova software package. Electrospray ionization mass spectra (ESI-MS) were performed on an Agilent 6224 ESI/TOF mass spectrometer. Analytical reversed- phase high-performance liquid chromatography (RP-HPLC) exper- iments were performed on a Shimadzu Prominence HPLC system. A Venusil XBP C18 column (5 mm, 150 Å, 250 4.60 mm) was used. Platinum contents were measured on an inductively coupled plasma mass spectrometer (ICP-MS; Thermo iCAP Q). Confocal microscopic images were captured by a laser confocal scanning microscope (Olympus FV1000). Fig. 7. Proposed mechanism of action for the bezafibrate containing Pt(IV) complexes CB and CP in A549 cells. 4.2. Synthesis and characterization All the reactions were conducted under normal atmospheric conditions and under dark conditions. c,c,t-[PtCl2(NH3)2(OH)2] was synthesized in the manner previously reported methods. Bezafibrate (110 mg, 0.30 mmol), TBTU (300 mg, 0.93 mmol), and Et3N (0.09 g, 0.93 mmol) were dissolved and stirred in ultra- dry DMSO at room temperature for 5 min. The mixture was then added by a suspension of c,c,t-[PtCl2(NH3)2(OH)2] (100 mg, 0.30 mmol) in dry DMSO (1 mL). The reaction mixture was stirred at room temperature for 48 h to get a clear solution. The final product CB was isolated by direct-phase chromatography using a solution of methanol/dichloromethane (1: 10) as eluent to get a yellow solid. Yield: 58.14 mg, 28.73 %. HPLC analytical purity: 98.38 %. 1H NMR (400 MHz, DMSO‑d6): d (ppm) 8.68 (t, J 5.5 Hz, 1H), 7.85 (d, J 8.6 Hz, 2H), 7.54 (d, J 8.5 Hz, 2H), 7.04 (d, J 8.5 Hz, 2H), 6.82 (d, J 8.5 Hz, 2H), 6.04 (br, 6H), 3.47e3.41 (m, 2H), 2.74 (t, J 7.4 Hz, 2H), 1.43 (s, 6H). 13C NMR (101 MHz, DMSO‑d6): d (ppm) 181.13, 165.03, 154.07, 135.83, 133.29, 131.36, 128.98 (d, J 10.7 Hz),128.33, 118.71, 115.07, 41.17, 34.20, 25.85. ESI-HRMS (positive-ion mode): m/z calcd for C19H27Cl3N3O5Ptþ ([M H]þ): 677.06585, found: 677.06525 (0.8 ppm). The bis-bezafibrate complex CP was obtained in the same way using a three-fold excess of activated ligand in DMSO at room temperature. The final product CP was isolated by direct-phase chromatography using a solution of methanol/dichloromethane (1: 20) as eluent to get a yellow solid. Yield: 94.16 mg, 30.78 % HPLC analytical purity: 96.94 %. 1H NMR (400 MHz, DMSO‑d6): d (ppm) 8.67 (t, J 5.5 Hz, 2H), 7.84 (d, J 8.4 Hz, 4H), 7.54 (d, J 8.4 Hz,4H), 7.06 (d, J 8.4 Hz, 4H), 6.85 (d, J 8.3 Hz, 4H), 6.58 (bs, 6H),3.43 (dd, J 13.8, 6.4 Hz, 4H), 2.76 (t, J 7.4 Hz, 4H), 1.45 (s, 12H).13C NMR (101 MHz, DMSO‑d6): d (ppm) 181.17, 165.03, 155.23,153.79, 135.83, 133.29, 131.89, 129.03, 128.33, 119.12, 79.11, 41.12,34.20, 25.87. ESI-HRMS (positive-ion mode): m/z calcd for C38H45Cl4N4O8Ptþ ([M H]þ):1022.16043, found: 1022.16028 (0.1 ppm). 4.3. Biological studies Stock solutions of bezafibrate, CP, and CP were prepared in DMF just before running the experiment, and a calculated amount of drug solution was added to the cell growth medium to a final DMF concentration of 0.5 %. Cisplatin was prepared in PBS. 4.3.1. Cell culture HeLa (human cervical carcinoma), MCF-7 (human breast carci- noma), LO2 (normal human liver), and HUVEC (human umbilical vein endothelial) cells were cultured in DMEM supported with 10 % FBS and 100 units of penicillin/streptomycin. A549 (human lung carcinoma) cells were cultured in McCoy's 5A medium with 10 % FBS and 100 units of penicillin and streptomycin. NCIeH460 (hu- man lung carcinoma) cells were maintained in RMPI 1640 medium supported with 10 % FBS. Cells were incubated in a humidified incubator at 37 ◦C with 5 % CO2 and sub-cultured every 2e3 days. 4.3.2. Cell proliferation assays The viability of different types of cells exposed to the com- pounds was evaluated by MTT assay. Briefly, cells were seeded into nonpyrogenic polystyrene 96-well plates at a density of 3000 cells/ well and allowed to attach overnight. Stock solutions were pre- pared in dry DMF and serially diluted in the culture media. Cells were treated with each compound with varying concentrations for 72 h. After continuing exposure for 72 h, the cells were incubated with MTT (10 mL, 5 mg mL—1 in PBS) for 4 h. The MTT containing culture medium was removed carefully, and DMSO (100 mL) was added to each well to dissolve the formazan. An Infinite 200 Pro NanoQuant microplate reader (Tecan, Swiss) was used to determine the absorbance at 570 nm. IC50 values were calculated from the non-linear curve fits of the dose-response curves using GraphPad Prism (La Jolla, CA) and reported as the mean ± standard deviations of at least two experiments performed in triplicate wells. 4.3.3. Determination of platinum contents in whole-cell samples and genomic DNA by inductively-coupled plasma mass spectrometry (ICP-MS) To determine the Pt content in whole-cell samples, exponentially-growing A549 cells (1 106 per well) were seeded in six-well plates with 2 mL of culture media and were allowed to attach overnight. Cells were then treated with cisplatin, CB, or CP at a concentration of 10 mM for 6 h. The cells were washed with 2 mL of ice-cold PBS three times and collected by trypsinization. Cell suspensions were centrifuged at 2000 rpm for 5 min, and the cell pellets were lyophilized using a freeze-drying system. Genomic DNA samples were incubated and collected using the same procedure, and DNA was extracted and purified using TIA- Namp Genomic DNA Kit (Tiangen Biotech, Beijing, China). The pu- rity and concentrations of the extracted genomic DNA samples were determined spectrophotometrically using Eppendorf Bio-Photometer D30 (Eppendorf, Hamburg, Germany). DNA solutions were lyophilized and stored at 80 ◦C until use. The Pt content in the lyophilized samples was detected by quadrupole inductively coupled plasma mass spectrometry (ICP- MS). Lyophilized samples were digested with 0.2 mL high-purity nitric acid for 4 h at room temperature and diluted to 2 mL with deionized water. The digested samples were centrifuged at 5000 rpm for 15 min, and an aliquot (0.2 mL) was transferred to an auto-sampler tube for intracellular platinum content determination. 4.3.4. Reduction of Pt(IV)complexes analyzed by HPLC Reduction of Pt(IV) complexes CB and CP (1 mM) was carried out using ascorbic acid (10 mM) in 0.1 M PBS/DMSO (99/1, V/V, pH 7.4), and the reduction product was monitored at 37 ◦C in the dark by RP-HPLC at different time intervals. Separation of the samples was performed on a Shimadzu Prominence HPLC system using UV de- tector at 254 nm and the following solvent system: solvent A, water/0.1 % formic acid, and solvent B, methanol/0.1 % formic acid, at a flow rate of 1.0 mL min—1 and a gradient of 95 % A—10 % A. 4.3.5. Comet assay To evaluate the extent of DNA damage induced by Pt complexes, a comet assay was performed. Briefly, A549 cells treated with 5 mM of platinum compounds or PBS for 24 h were mixed with 1 % low- melting agarose by a ratio of 1:10. The mixture was then spread on 0.8 % normal-melting agarose-coated slides. Slides were kept at 4 ◦C to solidify the low-melting agarose and lysed in ice-cold lysis buffer (0.25 M NaCl, 10 mM Na2EDTA, 1 mM Tris, 1 % Triton-X 100 and 10 % DMSO) at 4 ◦C overnight. Slides were then immersed in alkaline unwinding solution (0.1 mM EDTA, 30 mM NaOH, pH 13) at 4 ◦C for 40 min. Next, electrophoresis was performed at 300 mA, 25 V for 40 min at 4 ◦C in the same buffer, and the slides were neutralized in neutralization buffer (0.4 M Tris-HCl, pH 7.5) and dehydrated in ethanol. After stained in ethidium bromide (EB) so- lution (10 mg/mL) at room temperature, the samples were detected by a fluorescence microscope (Leica DMI3000B). The images were analyzed by the comet assay software project (CASP) image- analysis program, and the “olive tail moment” (OTM) of comets was calculated to evaluate DNA damage. 4.3.6. Detection of intracellular ROS Intracellular ROS levels were detected by using ROS fluorescence probe 20,70- dichloro-fluorescein diacetate (DCFH-DA) (Invitrogen). A549 cells were plated in six-well culture plates and allowed to attach overnight. The cells were then treated with 5 mM of platinum compounds or PBS for 24 h. The cells were harvested by trypsini- zation, washed twice with PBS, and then incubated with DCFH-DA at 37 ◦C for 20 min in the dark. The cells were analyzed immediately with BD FACS Verse flow cytometer (BD Biosciences, San Jose, Cal- ifornia, USA). The data were processed and analyzed by FlowJo 7.6 software. 4.3.7. Mitochondrial membrane potential assay The loss of mitochondrial membrane potential (MMP, DJm) was determined by flow cytometry using a JC-1 detection kit. Briefly, A549 cells were incubated with 5 mM of platinum compounds or PBS for 24 h, harvested, washed with PBS. The collected cells were then stained with JC-1 for 15 min at 37 ◦C for 20 min in the dark. Stained cells were washed, re-suspended in PBS for immediate flow cytometry analysis. The mitochondrial depolarization of the cells was evaluated by the decrease of the red/green fluorescence in- tensity ratio. 4.3.8. Wound-healing assay The cell migration ability was evaluated by wound-healing assay. Briefly, A549 cells were seeded into 12-well plates with a density of 8 105 cells/well and allowed to a monolayer. After 24 h, a straight scratch on the monolayer was created across the plate using a sterilized 200 mL pipette tip. Subsequently, the cells were washed with PBS and incubated with 5 mM of platinum compounds or PBS in fresh culture media for 24 h. The closure of the scratch was observed under a microscope (Motic AE2000), and the images were captured and processed using Mtico Images Advanced software version 3.2. Each experiment was repeated at least three times. 4.3.9. Cell invasion assay A549 cells were incubated with 2.5 mM of platinum-based compounds at 37 ◦C for 24 h and then were seeded into a Matrigel-coated 96-well plate (3500 cells/well). After 24 h of in- cubation, the cells in each well were washed twice by PBS, fixed by 4 % paraformaldehyde for 15 min, and stained with 1 % crystal violet for 15min. The images were captured on a Leica DMI3000B inverted microscope. MTT containing culture medium was added to the other plate and cultured at 37 ◦C for 4 h. The following processes were the same as the measurement of cell viability by MTT assay. The cell invasion rate was obtained using the formula: cell invasion rate ¼ (ODdrug/ODcontrol) × 100 %. 4.3.10. Cell cycle analysis A549 cells (1 × 106 cells per well) were seeded into six-well plates and incubated overnight. The cells were then incubated with 5 mM compounds. After 24 h, the cells were collected by trypsini- zation, washed with cold PBS three times, and fixed in ice-cold ethanol (70 %) overnight at 4 ◦C. Before flow cytometric detection, the cells were re-suspended and stained with propidium io- dide (PI) staining solution (50 mg/mL PI, 100 mg/mL RNase A) for 15 min at room temperature in the dark. The samples were then subjected to a BD Accuri C6 flow cytometer (BD Biosciences, San Jose, California, USA). Cell cycle distribution was analyzed by Modifit 3.1 software. 4.3.11. Annexin V-FITC/propidium iodide staining A549 cells were seeded in a six-well plate at a density of 1 × 106 cells per well. After overnight incubation, the cells were treated with 5 mM compounds and further incubated for 24 h. The cells were then collected by trypsinization (without EDTA), washed with ice-cold PBS, and stained according to Annexin V-FITC/PI double staining assay kit's instructions. The samples were analyzed by a BD FACS Verse flow cytometer (BD Biosciences, San Jose, Cal- ifornia, USA). The data were processed and analyzed by FlowJo 7.6 software. 4.3.12. Immunofluorescence (IF) Exponentially growing A549 cells (1.5 105 per well) were seeded in 15 mm glass-bottom cell culture dishes with 2 mL of cell culture media and were allowed to attach at 37 ◦C overnight. Cells were then incubated in the presence of 10 mM platinum-based compounds at 37 ◦C for 6 h and fixed with 4 % paraformaldehyde for 15 min. The cells were then permeabilized by 0.5 % Triton X-100 for 15 min and blocked with 1 % BSA solution for 1 h. Subsequently, the cells were incubated with the primary polyclonal antibody at 4 ◦C overnight. Next, the cells were washed with PBS three times and incubated with fluorophore-labeled secondary antibodies for 30 min. The samples were further washed by PBS and mounted with fluoroshield mounting buffer containing DAPI (Solarbio, China). 4.3.13. Protein extraction, quantification, SDS-PAGE, and western blotting A549 cells were plated in six-well plates at a density of 1 × 106 cells/well and incubated overnight. Then, the cells were exposed to 5 mM platinum-based compounds for 12 h at the indi- cated concentration at 37 ◦C. After the treatments, the total protein in each sample was extracted by using RIPA buffer containing 1 % PMSF and centrifuged at 12,000 rpm at 4 ◦C for 5 min. The super- natant was collected and stored at —20 ◦C. Protein concentration was determined by a BCA protein assay kit (Solarbio, China). 40 mg of total protein extracts were loaded and separated on 10 % SDS- PAGE. The gels were transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, USA) and blocked in a TBST buffer containing 5 % non-fat milk for 3 h. The membranes were then incubated with primary antibodies at 4 ◦C overnight, washed with TBST three times, and further incubated with the secondary anti- body with horseradish peroxidase for 1 h. After extensively washed, the immunoreactive bands were visualized using Pierce ECL western blotting substrate (Thermo) and recorded by a Tanon- 5200-multi imaging analysis system. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments We thank the National Natural Science Foundation of China (No.21977080) and the Tianjin Municipal Nature Science Founda- tion (Nos. 17JCZDJC33100, 17JCYBJC28500, 18JCYBJC91300) for funding support. Appendix A. Supplementary data Supplementary data to this article can be found online at References [1] P. Hainaut, A. Plymoth, Targeting the hallmarks of cancer: towards a rational approach to next-generation cancer therapy, Curr. Opin. Oncol. 25 (2013) 50e51. [2] D. Hanahan, R.A. Weinberg, Hallmarks of cancer: the next generation, Cell 144 (2011) 646e674. [3] V. Cocetta, E. Ragazzi, M. Montopoli, Links between cancer metabolism and cisplatin resistance, Int. Rev. Cell Mol. Biol. 354 (2020) 107e164. [4] S. Beloribi-Djefaflia, S. Vasseur, F. Guillaumond, Lipid metabolic reprogram- ming in cancer cells, Oncogenesis 5 (2016) e189. [5] J. Long, C.J. Zhang, N. Zhu, K. Du, Y.F. Yin, X. Tan, D.F. Liao, L. Qin, Lipid metabolism and carcinogenesis, cancer development, Am. J. Cancer Res. 8 (2018) 778e791. [6] B. Qiu, D. Ackerman, D.J. Sanchez, B. Li, J.D. Ochocki, A. Grazioli, E. Bobrovnikova-Marjon, J.A. Diehl, B. Keith, M.C. Simon, HIF2a-dependent lipid storage promotes endoplasmic reticulum homeostasis in clear-cell renal cell carcinoma, Canc. Discov. 5 (2015) 652e667. [7] E. Di Bello, C. Zwergel, A. Mai, S. Valente, The innovative potential of statins in cancer: new targets for new therapies, Front. Chem. 8 (2020) 516. [8] A. Markowska, M. Antoszczak, J. Markowska, A. Huczynski, Statins: HMG-CoA reductase inhibitors as potential anticancer agents against malignant neo- plasms in women, Pharmaceuticals 13 (2020) 422. [9] Y. Zhang, Y. Liu, J. Duan, H. Wang, Y. Zhang, K. Qiao, J. Wang, Cholesterol depletion sensitizes gallbladder cancer to cisplatin by impairing DNA damage response, Cell Cycle 18 (2019) 3337e3350. [10] A. Martirosyan, J.W. Clendening, C.A. Goard, L.Z. Penn, Lovastatin induces apoptosis of ovarian cancer cells and synergizes with doxorubicin: potential therapeutic relevance, BMC Canc. 10 (2010) 103. [11] X. Lian, G. Wang, H. Zhou, Z. Zheng, Y. Fu, L. Cai, Anticancer properties of fenofibrate: a repurposing use, J. Canc. 9 (2018) 1527e1537. [12] K.K. Nagothu, R. Bhatt, G.P. Kaushal, D. Portilla, Fibrate prevents cisplatin- induced proximal tubule cell death, Kidney Int. 68 (2005) 2680e2693. [13] S.J. Kim, C. Park, J.N. Lee, R. Park, Protective roles of fenofibrate against cisplatin-induced ototoxicity by the rescue of peroxisomal and mitochondrial dysfunction, Toxicol. Appl. Pharmacol. 353 (2018) 43e54. [14] P. Thongnuanjan, S. Soodvilai, V. Chatsudthipong, S. Soodvilai, Fenofibrate reduces cisplatin-induced apoptosis of renal proximal tubular cells via inhi- bition of JNK and p38 pathways, J. Toxicol. Sci. 41 (2016) 339e349. [15] A.J. Pickard, U. Bierbach, The cell's nucleolus: an emerging target for chemo- therapeutic intervention, ChemMedChem 8 (2013) 1441e1449. [16] T.C. Johnstone, K. Suntharalingam, S.J. Lippard, The next generation of plat- inum drugs: targeted Pt(II) agents, nanoparticle delivery, and Pt(IV) prodrugs, Chem. Rev. 116 (2016) 3436e3486. [17] D. Gibson, Multi-action Pt(IV) anticancer agents; do we understand how they work? J. Inorg. Biochem. 191 (2019) 77e84. [18] U. Basu, B. Banik, R. Wen, R.K. Pathak, S. Dhar, The Platin-X series: activation, targeting, and delivery, Dalton Trans. 45 (2016) 12992e13004. [19] M.D. Hall, H.R. Mellor, R. Callaghan, T.W. Hambley, Basis for design and development of platinum(IV) anticancer complexes, J. Med. Chem. 50 (2007) 3403e3411. [20] E. Gabano, M. Ravera, F. Trivero, S. Tinello, A. Gallina, I. Zanellato, M.B. Gariboldi, E. Monti, D. Osella, The cisplatin-based Pt(IV)-diclorofibrato multi-action anticancer prodrug exhibits excellent performances also under hypoxic conditions, Dalton Trans. 47 (2018) 8268e8282. [21] G.E. Stoykova, I.R. Schlaepfer, Lipid metabolism and endocrine resistance in prostate cancer, and new opportunities for therapy, Int. J. Mol. Sci. 20 (2019) 2626. [22] X.Q. Song, R.P. Liu, S.Q. Wang, Z. Li, Z.Y. Ma, R. Zhang, C.Z. Xie, X. Qiao, J.Y. Xu, Anticancer melatplatin prodrugs: high effect and low toxicity, MT1-ER-target and immune response in vivo, J. Med. Chem. 63 (2020) 6096e6106. [23] Z.Y. Ma, D.B. Wang, X.Q. Song, Y.G. Wu, Q. Chen, C.L. Zhao, J.Y. Li, S.H. Cheng, J.Y. Xu, Chlorambucil-conjugated platinum(IV) prodrugs to treat triple- negative breast cancer in vitro and in vivo, Eur. J. Med. Chem. 157 (2018) 1292e1299. [24] X.Q. Song, Z.Y. Ma, Y.G. Wu, M.L. Dai, D.B. Wang, J.Y. Xu, Y. Liu, New NSAID- Pt(IV) prodrugs to suppress metastasis and invasion of tumor cells and enhance anti-tumor effect in vitro and in vivo, Eur. J. Med. Chem. 167 (2019) 377e387. [25] R. Zhang, X.Q. Song, R.P. Liu, Z.Y. Ma, J.Y. Xu, Fuplatin: an efficient and low- toxic dual-prodrug, J. Med. Chem. 62 (2019) 4543e4554. [26] I.V. Tetko, H.P. Varbanov, M. Galanski, M. Talmaciu, J.A. Platts, M. Ravera, E. Gabano, Prediction of logP for Pt(II) and Pt(IV) complexes: comparison of statistical and quantum-chemistry based approaches, J. Inorg. Biochem. 156 (2016) 1e13. [27] C.A. Puckett, R.J. Ernst, J.K. Barton, Exploring the cellular accumulation of metal complexes, Dalton Trans. 39 (2010) 1159e1170. [28] D. Gibson, Platinum(IV) anticancer prodrugs - hypotheses and facts, Dalton Trans. 45 (2016) 12983e12991. [29] E. Wong, C.M. Giandomenico, Current status of platinum-based antitumor drugs, Chem. Rev. 99 (1999) 2451e2466. [30] E. Wexselblatt, D. Gibson, What do we know about the reduction of Pt(IV) pro-drugs? J. Inorg. Biochem. 117 (2012) 220e229. [31] M.C. McCormick, K. Keijzer, A. Polavarapu, F.A. Schultz, M.H. Baik, Under- standing intrinsically irreversible, non-Nernstian, two-electron redox pro- cesses: a combined experimental and computational study of the electrochemical activation of platinum(IV) antitumor prodrugs, J. Am. Chem. Soc. 136 (2014) 8992e9000. [32] E. Dabbish, F. Ponte, N. Russo, E. Sicilia, Antitumor platinium(IV) prodrugs: a systematic computational exploration of their reduction mechanism by L- ascorbic acid, Inorg. Chem. 58 (2019) 3851e3860. [33] J.D. Loya, J. Qiu, D.K. Unruh, A.F. Cozzolino, K.M. Hutchins, Co-crystallization of the anti-cholesterol drug bezafibrate: molecular recognition of a pharma- ceutical contaminant in the solid state and solution via hydrogen bonding, Cryst. Growth Des. 18 (2018) 4838e4843. [34] D. Hofer, H.P. Varbanov, M. Hejl, M.A. Jakupec, A. Roller, M. Galanski, B.K. Keppler, Impact of the equatorial coordination sphere on the rate of reduction, lipophilicity and cytotoxic activity of platinum(IV) complexes, J. Inorg. Biochem. 174 (2017) 119e129. [35] S. Jin, Y. Guo, D. Song, Z. Zhu, Z. Zhang, Y. Sun, T. Yang, Z. Guo, X. Wang, Targeting energy metabolism by a platinum(IV) prodrug as an alternative pathway for cancer suppression, Inorg. Chem. 58 (2019) 6507e6516. [36] L. Ma, N. Wang, R. Ma, C. Li, Z. Xu, M.K. Tse, G. Zhu, Monochalcoplatin: an actively transported, quickly reducible, and highly potent Pt(IV) anticancer prodrug, Angew Chem. Int. Ed. Engl. 57 (2018) 9098e9102. [37] C.K.J. Chen, P. Kappen, D. Gibson, T.W. Hambley, Trans-platinum(IV) pro-drugs that exhibit unusual resistance to reduction by endogenous reductants and blood serum but are rapidly activated inside cells: 1H NMR and XANES spectroscopy study, Dalton Trans. 49 (2020) 7722e7736. [38] E.P. Rogakou, C. Boon, C. Redon, W.M. Bonner, Megabase chromatin domains involved in DNA double-strand breaks in vivo, J. Cell Biol. 146 (1999) 905e916. [39] E.P. Rogakou, D.R. Pilch, A.H. Orr, V.S. Ivanova, W.M. Bonner, DNA double- stranded breaks induce histone H2AX phosphorylation on serine 139, J. Biol. Chem. 273 (1998) 5858e5868. [40] F.L. Khanim, R.E. Hayden, J. Birtwistle, A. Lodi, S. Tiziani, N.J. Davies, J.P. Ride, M.R. Viant, U.L. Gunther, J.C. Mountford, H. Schrewe, R.M. Green, J.A. Murray, M.T. Drayson, C.M. Bunce, Combined bezafibrate and medroxyprogesterone acetate: potential novel therapy for acute myeloid leukaemia, PloS One 4 (2009), e8147.
[41] M. Kleih, K. Bopple, M. Dong, A. Gaissler, S. Heine, M.A. Olayioye,
W.E. Aulitzky, F. Essmann, Direct impact of cisplatin on mitochondria induces ROS production that dictates cell fate of ovarian cancer cells, Cell Death Dis. 10 (2019) 851.
[42] S. Savino, V. Gandin, J.D. Hoeschele, C. Marzano, G. Natile, N. Margiotta, Dual- acting antitumor Pt(IV) prodrugs of kiteplatin with dichloroacetate axial li- gands, Dalton Trans. 47 (2018) 7144e7158.
[43] X. Xue, S. You, Q. Zhang, Y. Wu, G.Z. Zou, P.C. Wang, Y.L. Zhao, Y. Xu, L. Jia,
X. Zhang, X.J. Liang, Mitaplatin increases sensitivity of tumor cells to cisplatin

by inducing mitochondrial dysfunction, Mol. Pharm. 9 (2012) 634e644.
[44] L. Tabrizi, K. Thompson, K. Mnich, C. Chintha, A.M. Gorman, L. Morrison,
J. Luessing, N.F. Lowndes, P. Dockery, A. Samali, A. Erxleben, Novel Pt(IV) prodrugs displaying antimitochondrial effects, Mol. Pharm. 17 (2020) 3009e3023.
[45] J. Drukala, K. Urbanska, A. Wilk, M. Grabacka, E. Wybieralska, L. Del Valle,
Z. Madeja, K. Reiss, ROS accumulation and IGF-IR inhibition contribute to fenofibrate/PPARa-mediated inhibition of glioma cell motility in vitro, Mol. Canc. 9 (2010) 159.
[46] D. Han, W. Wei, X. Chen, Y. Zhang, Y. Wang, J. Zhang, X. Wang, T. Yu, Q. Hu,
N. Liu, Y. You, NF-kB/RelA-PKM2 mediates inhibition of glycolysis by fenofi- brate in glioblastoma cells, Oncotarget 6 (2015) 26119e26128.
[47] C.I. Jan, M.H. Tsai, C.F. Chiu, Y.P. Huang, C.J. Liu, N.W. Chang, Fenofibrate suppresses oral tumorigenesis via reprogramming metabolic processes: po- tential drug repurposing for oral cancer, Int. J. Biol. Sci. 12 (2016) 786e798.
[48] H.L. Jiao, B.L. Zhao, Cytotoxic effect of peroxisome proliferator fenofibrate on human HepG2 hepatoma cell line and relevant mechanisms, Toxicol. Appl. Pharmacol. 185 (2002) 172e179.
[49] Z. Zak, P. Gelebart, R. Lai, Fenofibrate induces effective apoptosis in mantle cell lymphoma by inhibiting the TNFa/NF-kB signaling axis, Leukemia 24 (2010) 1476e1486.
[50] R. Scatena, P. Bottoni, B. Giardina, Mitochondria, PPARs, and cancer: is receptor-independent action of PPAR agonists a key? PPAR Res 2008 (2008) 256251.
[51] P. Khan, A. Bhattacharya, D. Sengupta, S. Banerjee, A. Adhikary, T. Das, Aspirin enhances cisplatin sensitivity of resistant non-small cell lung carcinoma stem- like cells by targeting mTOR-Akt axis to repress migration, Sci. Rep. 9 (2019) 16913.
[52] S. Chen, X. Zhu, X. Lai, T. Xiao, A. Wen, J. Zhang, Combined cancer therapy with non-conventional drugs: all roads lead to AMPK, Mini Rev. Med. Chem. 14 (2014) 642e654.
[53] S.C. Tsai, M.H. Tsai, C.F. Chiu, C.C. Lu, S.C. Kuo, N.W. Chang, J.S. Yang, AMPK- dependent signaling modulates the suppression of invasion and migration by fenofibrate in CAL 27 oral cancer cells through NF-kB pathway, Environ. Toxicol. 31 (2016) 866e876.
[54] D. Gan, W. He, H. Yin, X. Gou, beta-elemene enhances cisplatin-induced apoptosis in bladder cancer cells through the ROS-AMPK signaling pathway, Oncol. Lett. 19 (2020) 291e300.
[55] M.M. Yung, H.Y. Ngan, D.W. Chan, Targeting AMPK signaling in combating ovarian cancers: opportunities and challenges, Acta Biochim. Biophys. Sin. 48 (2016) 301e317.
[56] Q. Li, G. Yang, M. Feng, S. Zheng, Z. Cao, J. Qiu, L. You, L. Zheng, Y. Hu, T. Zhang,
Y. Zhao, NF-kB in pancreatic cancer: its key role in chemoresistance, Canc. Lett. 421 (2018) 127e134.
[57] S.L. Ryan, S. Beard, M.P. Barr, K. Umezawa, S. Heavey, P. Godwin, S.G. Gray,
D. Cormican, S.P. Finn, K.A. Gately, A.M. Davies, E.W. Thompson, D.J. Richard,
K.J. O’Byrne, M.N. Adams, A.M. Baird, Targeting NF-kB-mediated inflammatory pathways in cisplatin-resistant NSCLC, Lung Canc. 135 (2019) 217e227.
[58] H. Liang, P. Kowalczyk, J.J. Junco, H.L. Klug-De Santiago, G. Malik, S.J. Wei,
T.J. Slaga, Differential effects on lung cancer cell proliferation by agonists of glucocorticoid and PPARa receptors, Mol. Carcinog. 53 (2014) 753e763.
[59] R. Kang, Q. Zhang, H.J. Zeh 3rd, M.T. Lotze, D. Tang, HMGB1 in cancer: good, bad, or both? Clin. Canc. Res. 19 (2013) 4046e4057.
[60] H. Zheng, J.N. Chen, X. Yu, P. Jiang, L. Yuan, H.S. Shen, L.H. Zhao, P.F. Chen,
M. Yang, HMGB1 enhances drug resistance and promotes in vivo tumor growth of lung cancer cells, DNA Cell Biol. 35 (2016) 622e627.
[61] S. Yusein-Myashkova, I. Stoykov, A. Gospodinov, I. Ugrinova, E. Pasheva, The repair capacity of lung cancer cell lines A549 and H1299 depends on HMGB1 expression level and the p53 status, J. Biochem. 160 (2016) 37e47.
[62] P. Carmeliet, VEGF as a key mediator of angiogenesis in cancer, Oncology 69 (Suppl 3) (2005) 4e10.