Enhanced anticancer potency with reduced nephrotoxicity of newly synthesized platin-based complexes compared with cisplatin

As a platinum-containing anticancer drug, cisplatin is the keystone for treating many malignancies. Nephrotoxicity is the main dose-limiting toxicity, and several hydration therapies and supplementary strategies are utilized to reduce cisplatin-induced kidney damage, so the discovery and development of effective and safe antitumor drugs are still on the path of human health. Herein, a new four-coordinated Pt complex [Pt(TSC)Cl] using N(4)-phenyl-2-formylpyridine thiosemicarbazone (HTSC) was synthesized and characterized by single-crystal X-ray diffraction, 1HNMR, FT-IR, LC/MS and CHN elemental analysis. The Pt(TSC)Cl complex revealed antiproliferative activity against A549, MCF-7 and Caco-2 cell lines with a low micromolar IC50 (200–1.75 µM). Specifically, the Pt(TSC)Cl complex displayed more selectivity in Caco-2 cells (IC50 = 2.3 µM) than cisplatin (IC50 = 107 µM) after 48 h of treatment. Moreover, compared with cisplatin, a known nephrotoxic drug, the Pt(TSC)Cl complex exhibited lower nephrotoxicity against Hek293 normal cells. We also found that the Pt(TSC)Cl complex can effectively prevent cancer cell propagation in sub-G1 and S phases and induce apoptosis (more than 90%). Real time PCR and western analysis demonstrated that the expression pattern of apoptotic genes and proteins is according to the intrinsic apoptosis pathway through the Bax/Bcl-2-Casp9-Casp3/Casp7 axis. Collectively, our findings indicated that the Pt(TSC)Cl complex triggers apoptosis in Caco-2 cell lines, while low nephrotoxicity was shown and may be considered a useful anticancer drug candidate for colorectal cancers for further optimization and growth.


Materials and methods
All chemicals and reagents were obtained from Sigma Aldrich and used without purification.
FT-IR spectra were recorded on a Bruker Tensor 27 FT-IR spectrophotometer with KBr disks in the range 4000-400 cm −1 (Columbia, MD, USA). A QTOF LC/MS spectrometer was used for mass spectrometry (Billerica, MA, USA). 1HNMR spectroscopy was performed at room temperature on an NMR Bruker Advance 400 MHz magnet (Agilent, USA) using DMSO-d6. Elemental analysis, C H N, was performed by an elemental analyzer Vario EL III (Sydney, Australia). X-ray crystallography was performed on an Xtal AB Synergy Dualflex Hypix diffractometer using CuKα radiation, λ = 1.54184 Å at 100 K. The structure was solved by direct methods and different Fourier syntheses 24 . A thermal ellipsoid plot was generated using the Mercury program 25 integrated within the WINGX 26 suite of programs.
A Facs Canto II (Bio-Rad) flow cytometry system (Mexico, USA) with the Flow LogicTM program was used for cell cycle Annexin V and PI evaluation. Real time PCR test was performed by Mic real time PCR, Bio Molecular system (Santa Clara, USA).
Cell viability assay. All  Investigation of cell cycle progression. Evaluation of cell cycle progression was performed by flow cytometry of propidium iodide and can identify the fraction of time spent by cells in various stages of the cell cycle. First, A549, MCF-7, and Caco-2 cells were seeded in 6-well plates at 25*10 4 cells per well and exposed to a concentration at the IC 50 of Pt(TSC)Cl complex and cisplatin for 48 h. Cell cycle distribution was determined by flow cytometry by the protocol described previously 28 . The control condition consisted of cells cultured in the absence of treatment but in the presence of PI. After harvesting the cells with trypsin, they were rinsed and fixed with 70% ethanol and kept for 3 days at 4 °C. Then, the samples were centrifuged, and after removal of supernatants after removing supernatants, the cells were dispersed in 300 µL of PBS. Next, after adding 10 µL of ribonuclease-A (10 mg/mL) and 45 min of incubation, 10 µL propidium iodide (1 mg/mL) was added to each of the samples and vortexed. After 10 min of incubation at room temperature and in the dark, the cells were examined by a FACSCalibur flow cytometer to estimate the cell cycle phases. FlowJo 7.6 software was used to analyze the cell cycle.
RNA isolation and cDNA synthesis. Three cancer cell lines were seeded in 6-well plates at 10 6 cells per well and incubated with the IC 50 values of the Pt(TSC)Cl complex and cisplatin at 48 h. Then, the cells were rinsed with cold PBS and harvested. Total RNA was isolated from MCF-Z, A549, and Caco-2 cell lines by the TRIzol method (TRIzol™ Reagent (Invitrogen, Catalog number: 15596026)). Briefly, the cells were centrifuged at 500 g at 4ºC, the plates were processed for cell lysate by 750 μl of RiboEx; then 200 μl chloroform was subjoined to the cell lysate and after incubating for 2 min at room temperature, they were centrifuged at 12,000 g for 20 min (4 °C). Following collection of aqueous phases, one volume of isopropyl alcohol was dispensed for precipitation of total RNA at 12,000 g for 20 min (4 °C Assessment of apoptotic protein level alterations by western blot analysis. Caco-2 cancer cells, as the most sensitive cells to Pt(TSC)Cl, were seeded in 6-well plates at a population of 10 6 cells per well and treated with cisplatin and Pt(TSC)Cl for 48 h. Assessment of apoptotic protein level alterations by western blot analysis by the protocol described previously 28 . The control included cells without treatment. To provide the cell lysate, cells were collected and treated with the following materials: RIPA buffer [500 µL of Tris-HCl (pH = 8), one tablet of protease inhibitor cocktail, 0.003 g EDTA, 0.08 g NaCl, 0.025 g sodium deoxycholate, 0.01 g SDS and 10 µl of Triton NP40 (1%)] used at 4 °C. After that, the cells were centrifuged at 12,000 rpm for 10 min at 4 °C. To measure total protein, the supernatant of cells was determined by the Bradford assay (Bio-Rad pro- www.nature.com/scientificreports/ tein assay, Bio-Rad Laboratories, USA) using a spectrophotometer (Bibby Scientific Ltd, Beacon Rd, UK). The portions of target protein received from the sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were carried over the polyvinylidene difluoride (PVDF) membrane and were blocked with TBST buffer containing 5% (w/v) skim milk (0.1% v/v Tween®20-tris buffered saline: TBST). The blocked PVDF membranes that contained the target proteins were incubated with specific primary antibodies: Bax (B-9) mouse monoclonal antibody (Santa Cruz) (1:1000), Bcl-2 sc-492 (Santa Cruz) rabbit monoclonal antibody (1:1000), caspase-7 (C7) (Cell Signaling) rabbit polyclonal antibody (1:1000), Caspase-3 Cell Signaling rabbit monoclonal antibody (1:1000), caspase-9 (Cell Signaling) rabbit polyclonal antibody (1:1000), GAPDH (Santa Cruz) mouse monoclonal antibody (1:1000) and P27 (Elabscience) rabbit polyclonal antibody (1:400) that were diluted with blocking buffer) overnight at 4 °C. The membrane was washed with TBST and incubated with secondary antibodies (m-IgGκ BP-HRP (Santa Cruz) for Bax (B-9), Bcl-2 and GAPDH, and mouse anti-rabbit IG-HRP (Santa Cruz) was used for Caspase-3, 7, 9 and P27, which were diluted using blocking buffer (1:1000).) for 1 h at room temperature. The target protein bands were visualized by an enhanced chemiluminescence detection kit (Thermo Fisher Scientific, Breda, the Netherlands) and Amersham® Imager 600 system (GE Healthcare Life Sciences, Eindhoven, Netherlands). Finally, the western blotting results were normalized to the expression of GAPDH as the loading control. All the protein bands were quantified by ImageJ software (v. 1.52n).

Results and discussion
The reaction of platinum (II) with ligand (HTSC) yielded the corresponding complexes. On the basis of elemental analysis and X-ray structure determination, the complex was formulated as Pt (TSC)Cl. In the complex, the ligand acts as monoanionic through deprotonation of the hydrazonic hydrogen atom (Scheme 2).

Structural characterization of the Pt(TSC)Cl complex with 1 HNMR and FTIR.
In the 1 HNMR spectrum of the complex (Fig. S1), the NH-N proton signal at 12.00 ppm in the free ligand was omitted upon coordination with platinum. In addition, the CH=N proton shifted from 8.20 in the free ligand to the downfield region, 8.65 ppm, as a strong singlet in the complexation. In comparison with the IR spectra of the free ligand (HTSC) and Pt(TSC)Cl complex (Fig. S2), the displacement of the bands assigned to ν C-H and ν N-H in the www.nature.com/scientificreports/ 2927-3440 cm -1 region and the bands assigned to ν C=N and ν C=C in the 1440-1700 cm -1 region could be indicative of coordination of the ligand through its nitrogen atoms, as the crystal structure shows.
X-ray crystal structure of Pt(TSC)Cl complex. Relevant data for the collection and structure solutions are listed in Table 2, and Fig. 1 shows the thermal ellipsoid plot of the complex molecules with an atom numbering scheme. The Pt(TSC)Cl complex crystallizes in the orthorhombic space group Pna2 1 with two independent molecules in the asymmetric unit, which form a π-stacked dimer where the distance between each of the planes is defined by atoms C1, C7, N1, N3, and Pt1, and the Pt atom of the second molecule is 3.521(4) Å. Selected bond distances and bond angles are summarized in Tables 3 and 4, respectively. Small differences in the geometrical parameters for the two molecules presumably arise due to differences in their packing environments in the crystal. These two molecules are essentially identical, and herein after, the values for molecule 1 and molecule 2 are presented in brackets. The most significant difference is in the torsional angle C7-N4-C8-C9 (7(3) 30 and is indicative of substantial single-bond characters.
The C7-N3 bond lengths, (1.28(3) [1.33 (2)]Å, are decreased compared to the corresponding distance in the free ligand (1.3685(2) Å 3 ). The shortening of the mentioned C-N bonds and lengthening of the C-S bonds in the complex molecules compared to the corresponding bonds in the free ligand revealed that complexation occurs by losing the hydrogen atom attached to N3 and N3A, as shown in Fig. 1. The shortening of C7-N3 is more than www.nature.com/scientificreports/   www.nature.com/scientificreports/ the shortening of C7A-N3A, and vice versa, the length of S1-C7 is more than the lengths of S1A-C7A. Therefore, the double and single characters of C7-C3 and S1-C7 bonds, respectively, are more than those of C7A-C3A and S1A-C7A bonds, so the N3 atom has a more negative charge than the N3A atom. There is no interaction between the two molecules in one asymmetric unit cell, but each molecule is in insignificant short contact with adjacent molecules in the neighboring asymmetric unit cells. The Cl1 atom in the molecule is linked with N4A of an adjacent molecule in the neighboring asymmetric unit cell by hydrogen bonding (Fig. 2

Cell viability assay in vitro.
To investigate the potency of the Pt(TSC)Cl complex as an anticancer drug, three different cancer cell lines, MFC-7, A549 and Caco-2 (10,000 cells per well), were exposed to increasing concentrations of Pt(TSC)Cl and cisplatin (up to 200 µM) to determine their IC 50 values/µM ( Fig. S3 and Table 5). All of cell lines were treated for three different duration times (24, 48 and 72 h). The Caco-2 cells were very sensitive to Pt(TSC)Cl in comparison with the two others in all three different durations of treatment. We also used cisplatin, a commercially available platinum-based compound, as a reference. In the Caco-2 cell line, the Pt(TSC)Cl complex was highly toxic, with an IC 50 < 10 µM, and even the toxicity was higher than that of cisplatin in all durations of the cell viability assay (****, p < 0.00001). After two days of treatment for 48 h, the  www.nature.com/scientificreports/ Pt(TSC)Cl complex showed 46-fold more toxicity than cisplatin in the Caco-2 cell line. In the MCF-7 and A549 cell lines, cisplatin showed more toxicity than the Pt(TSC)Cl complex (***, P ≤ 0.001). Cisplatin's lower toxicity to Caco-2 cell lines compared to MCF7 and A549 cell lines was confirmed in our previous work 31 .
Since renal toxicity is a major problem with platinum-based drugs, the toxic effect of the Pt(TSC)Cl complex was evaluated on Hek293 kidney cells in comparison to cisplatin. The Pt(TSC)Cl complex, with the higher cytotoxic potency for the Caco-2 cell line, revealed lower nephrotoxicity in Hek293 cells (***p < 0.0001) (Fig. S3) than cisplatin (cisplatin was 15-fold more toxic to Hek293 cells than the Pt(TSC)Cl complex). A summary of all cytotoxicity data is listed in Table 5.
Three different cancer cell lines, MCF-7, A549, Caco-2, and normal Hek293 cells, were treated with different concentrations of cisplatin and Pt(TSC)Cl complex for 24, 48, 72 h. Cell viability was assessed with MTT and is presented relative to untreated cells (set to 100%). Fig. 3 and Table 6, untreated A549, MCF-7 and Caco-2 cells in the G0/G1, S or G2/M phases showed normal growth conditions. In the MCF-7 cells treated with cisplatin (S: 37.2%) and Pt(TSC)Cl complex (S: 28%), the amount of S phase significantly increased compared with control cells (p < 0.0001). Therefore, S phase arrest was dominant in MCF-7 cells treated with both cisplatin and the Pt(TSC) Cl complex. Emerging sub G1 phase as a marker for apoptotic cells occurred in the Caco-2 cell lines treated with cisplatin and Pt(TSC)Cl complex. Compared to the nontreated Caco-2 cell lines (control group), in groups that were treated with cisplatin and Pt(TSC)Cl, approximately 10 and 51% of the cells were observed in the subG1 phase, respectively. The difference between subG1 arrest in the presence of the Pt(TSC)Cl complex and cisplatin  www.nature.com/scientificreports/ and between the Pt(TSC)Cl complex and control cells was significant (p < 0.0001). These data are also in line with the results of Annexin V and P analysis. However, in A549 cells, just in the presence of Pt(TSC)Cl complex was the sub G1 phase observed (subG1:22%). The increase in cell number at the subG1 phase can be associated with apoptotic cell death, leading to reduced cell entry to the G2/M and S phases 32 . Observing approximately 50% cell arrest of Caco-2 cells also proved that in the presence of the Pt(TSC)Cl complex, apoptosis was a dominant pathway of cell death ( Fig. 3 and Table 6). In support of this claim, we noted that the percentage of proliferating cells at the G2/M phase was diminished in groups that were treated with Pt(TSC)Cl complex.

Apoptosis induction assessed by flow cytometry of Annexin V, PI. An Annexin-V/PI assay was
performed via flow cytometry to measure the percentage of cells with early and late apoptotic changes. The percentages of apoptotic and necrotic cells for three cancer cell lines in the presence of both Pt(TSC)Cl complex and cisplatin were compared with those for vehicle control (stained viable cells) and are reported in Fig. 4 and Table 7. To gain insight into the chosen pathway in the presence of Pt(TSC)Cl complex and cisplatin, the percentage of necrotic cells was compared to the percentage of apoptotic cells. The results showed that Caco-2 cells treated with Pt(TSC)Cl complex were mostly killed by the apoptosis pathway provided by 89.7% apoptotic vs 4.79% necrotic cells after 48 h. It seems that apoptosis is the hallmark of Caco-2 cell lines death, in which approximately 37 and 52% of cells showed early and late apoptosis, respectively. Consistent with our data, late www.nature.com/scientificreports/ apoptotic changes were more evident cytotoxic effects in cells that were treated with Pt(TSC)Cl complex. In the presence of cisplatin, 21.17 and 2.91% of cells were killed by apoptosis and necrosis, respectively. In the A459 cells exposed to cisplatin and Pt(TSC)Cl complex, the percentages of apoptotic and necrotic cells were 15.33 and 1.73 and 10.86 and 3.90, respectively, which confirmed that they did not show significant toxicity to A549 cell lines. In MCF-7 cell lines treated with Pt(TSC)Cl complex, 31.28 and 2.37% of cells were killed by apoptosis and necrosis, respectively, and this percentage decreased to 18.88 and 3.24 when MCF-7 cells were treated with cisplatin. The results showed that in Caco-2 cell lines, the most common pathway in the presence of the Pt(TSC) Cl complex is apoptosis, so it can be shown that the Pt(TSC)Cl complex has one or more targets in Caco-2 cancer cells, and by targeting them, programmable cell death and apoptosis can be modulated.

Apoptosis induction assessed by real time PCR.
With the aim of precise investigation of the apoptosis pathway, real time PCR was performed. Bax, Bcl2, and different caspase subsets participate in regulating the apoptotic response. To this end, we measured the gene expression of different factors using real time PCR analysis. The results of the real-time PCR test are presented as a heatmap of the change in gene expression related to the control group (Fig. 5). Upregulation of Caspase 9 is related to mitochondrial injury and can trigger an internal apoptosis signaling pathway. In the next step, Caspase 9 activates the common apoptosis pathway consisting of Caspases 3 and 7. The data showed that the expression of apoptotic pathway genes, such as Bax and Caspases 9, 3, and 7, was upregulated in the cell treatment group that received Pt(TSC)Cl complex and cisplatin compared to the control. Low level of Bcl-2 mRNA is in agreement with apoptosis. Additionally, apoptotic gene expression was more dominant in all three cell lines treated with the Pt(TSC)Cl complex compared to cisplatin, which is in agreement with the cell cycle and Annexin-V analysis. The real time PCR results for apoptosis pathway genes showed that the expression pattern benefits from the intrinsic apoptosis pathway through the Bax/Bcl-2-Casp9-Casp3/Casp7 axis.
Western blotting to evaluate the mitochondrial mediator apoptotic pathway. Some studies have shown that mitochondrial DNA damage is evident in cisplatin-treated cells 31 . To check the activation of the mitochondria-dependent apoptosis pathway, the expression of BAX, a proapoptotic protein, and BcL-2, an antiapoptotic protein, was evaluated by western blotting. Overexpression of the BAX gene at the protein level in the presence of cisplatin and the Pt(TSC)Cl complex was observed. This protein can oligomerize and permeabilize the mitochondrial outer membrane to some apoptogenic factors. However, antiapoptotic proteins such as Bcl-2 can prevent the release of mitochondrial apoptogenic factors. The higher amount of Bcl-2 protein in the presence of the Pt(TSC)Cl complex shows that this compound, like cisplatin, can stimulate apoptosis by regulating the antiapoptotic and proapoptotic Bcl-2 protein families. Western blot analysis was performed to assess the executive apoptotic protein levels. Western blot analysis proved that the apoptotic process occurred via cascade activation of Bax, caspase 9, caspase 3 and caspase 7 (Figs. 6 and S4). The protein levels of Bax, caspase 9, caspase 3, and caspase 7 were upregulated and enlarged 1.9-, 3.2-, 1.97-and 1.92-fold, respectively, in Caco-2 cells treated with Pt(TSC)Cl complex compared to the control group. Furthermore, the abovementioned procedure was proven by the downregulation of Bcl2 protein levels. This result was in agreement with real time PCR outcomes. As reported previously, p27 protein has two duties in the determination of cell futures: pro-apoptotic function and cell cycle arrest 33 . According to our results, a 2.14-fold increase in p27 protein in Caco-2 cells treated with Pt(TSC)Cl complex compared to the control group showed the occurrence of apoptosis in Caco-2 cell lines treated with Pt(TSC)Cl complex. The fold changes of the mentioned proteins in Caco-2 cells treated with cisplatin were not increased significantly compared to untreated Caco-2 cells. The western blot results were in accordance with the cell cycle, Annexin-V and real time PCR results, which showed the highest apoptotic performance of CaCo-2 cells treated with Pt(TSC)Cl complex compared to cisplatin. Despite treating CRC with different kinds of therapies, it remains one of the most aggressive cancers, with a poor prognosis and high mortality. Platinum drugs, mostly cisplatin, have been conventionally used in chemotherapy against CRC. To fight against cancer, attempts have to be focused on the design and fabrication of very effective antitumor compounds with less toxicity to normal cells and tissues 10,34 . However, because of the www.nature.com/scientificreports/ substantial adverse effects of platinum-based drugs, as well as tumor resistance to treatment, new thiosemicarbazone derivatives in complex with platin (Pt(TSC)Cl complex) have been designed, and the data gathered and displayed in this work indicate that the Pt(TSC)Cl complex could be a promising substitute in the future. The Pt(TSC)Cl complex displayed potent antiproliferative activity against MCF-7, A549 and Caco-2 cell lines. Previous studies showed that platinum metal complexes of 2-acetylpyridine thiosemicarbazones showed slight antitumor potency against Ehrlich ascites tumor cells in vitro (because of the low solubility of the complex) but potent cytotoxic activity in vivo 35 . Thiosemicarbazone derivative (2-acetylpyridine thiosemicarbazones and heterocyclic thiosemicarbazones)-based complexes with transition metal (palladium (II), platinum (II), ruthenium (III), rhodium (III) and iridium (III) copper (II)) appeared to be more toxic in vivo 35 . This had also been observed earlier for complexes with similar ligands 36 . The representative Pt(TSC)Cl complex exhibited highly potent antiproliferative activity toward the Caco-2 cancer cell lines and lower nephrotoxicity against normal kidney cells, so a higher selectivity for a tumor mass than noncancerous cells was displayed. In agreement with our results, pyridine-2-carbaldehyde thiosemicarbazone and pyridine-2-carbaldehyde 4 N-methylthiosemicarbazone and their complexes with different transition metal ions have antitumor activity in colon cancer cells (HT-29 and SW-480) 37 . Therefore, these kinds of complexes could be replaced with cisplatin since nephrotoxicity is a serious and dose-limiting toxicity of cisplatin 5 . Our results confirmed that the Pt(TSC)Cl complex induced apoptosis by sub-G1 arrest, leading to blockade of cell cycle progression. Chemotherapeutic agents are mostly used to settle cellular hemostasis by damaging nuclear DNA and inducing cell death by apoptosis. A number of studies checked the DNA interactions of platinum-based compounds, but few of them checked the effect of these compounds on the activation of mitochondrial mediator apoptotic pathways 31 . Apoptotic gene expression was more dominant in all three cell lines treated with the Pt(TSC)Cl complex compared to cisplatin, which is in agreement with the cell cycle and Annexin-V analysis. The real time PCR results for apoptosis pathway genes revealed that the expression pattern benefits from the intrinsic apoptosis pathway via the Bax/Bcl-2-Casp9-Casp3/Casp7 axis. Moreover, more Bcl-2 protein in the presence of the Pt(TSC) Cl complex showed that this compound, like cisplatin, can stimulate apoptosis by regulating the antiapoptotic and proapoptotic Bcl-2 protein families. According to our results, a 2.14-fold increase in p27 protein in Caco-2 cells treated with Pt(TSC)Cl complex compared to the control group showed the occurrence of apoptosis in Caco-2 cell lines treated with Pt(TSC)Cl complex. The western blot results were in accordance with the cell cycle, Annexin-V and real time PCR results, which showed the higher apoptotic performance of Caco-2 cells treated with Pt(TSC)Cl complex compared to cisplatin. Collectively, the Pt(TSC)Cl complex showed selective toxicity against Caco-2 cancer cells, blocking cell cycle progression and inducing apoptosis via the intrinsic pathway. www.nature.com/scientificreports/ Our findings suggested that the Pt(TSC)Cl complex with the structure of thiosemicarbazone in complex with platinum may be a suitable anticancer agent for colon cancers for further optimization and development. Figure 6. The cropped blots/gels of proteins expression obtained from western blotting of the Caco-2 cells treated with IC 50 value of Cisplatin and Pt(TSC)Cl complex. Un-treated cells were considered as the control group. Proteins: Bcl-2, Bax, p27, pro-Caspase-3, Cleaved-Caspase-3, pro-Caspase-7, Cleaved-Caspase-7, pro-Caspase-9, Cleaved-Caspase-9, and GAPDH as internal control. Full-length blots/gels are presented in Supplementary Fig. S4

Conclusion
In the present study, a new Pt(II) metal complex [Pt(TSC)Cl] of N(4)-phenyl-2-formylpyridine thiosemicarbazone (HTSC) was synthesized and characterized. The X-ray crystal structure shows deprotonation of the ligand upon complexation. The coordination geometry around the four-coordinated Pt atoms in the complex molecule with one Cl atom, two N atoms and one S atom is a slightly distorted square-planar geometry. We found that this complex exerts increased cytotoxicity in Caco-2 cell lines and has reduced nephrotoxic potential compared to cisplatin. The new Pt(TSC)Cl complex was toxic to cancer cell lines in a cell cycle-dependent manner and induced cell death via caspase-mediated apoptosis, as proven at the gene and protein levels. Last, the Pt complex designed in this study may be motivating nominees for the improvement of a new category of anticancer drugs, especially for colorectal cancers. Forthcoming studies should focus on investigating their effectiveness on cisplatin-resistant tumors and their specificity, safety and pharmacokinetic behaviors in vivo.