Introduction

Discovery of platinum (Pt)-based chemotherapy has been the front-line treatment for a variety of malignancies including testicular, lung, and ovarian cancer (Kelland 2007; Brabec et al. 2012; Schneider et al. 2010; Roos and Kaina 2013). It is reported that the activity of platinum (Pt)-based chemotherapy is related to its binding to DNA and protein. For example, the interaction of cisplatin results in the formation of covalent cross-links, and the major adduct is 1,2 intrastrand crosslink formed between the [Pt(NH3)2]2+ unit and the two adjacent guanine bases of DNA (Icsel et al. 2015). It is well known that cis-platin drugs could be utilized to remedy 70–80% of cancer patients. However, the resistance to Pt-based regimens remains a major limitation in the successful treatment for many cancers including epithelial ovarian cancer (EOC) and non-small cell lung cancer (NSCLC) (Ozols et al. 2004; Einhorn 2008). Besides, the severe toxic and side effects, such as nephrotoxicity, ototoxicity and neurotoxicity have limited the platinum (Pt)-based chemotherapy for a wide range of uses (Kelland 2007; Ta 2006; Jamieson and Lippard 1999; O’Dwyer 1999; Hanigan and Devarajan 2003). At last, the poor water-solubility is another disadvantage for the use of platinum (Pt)-based chemotherapy. Consequently, to design and synthesize new platinum (Pt)-based complexes exhibiting strong DNA and protein binding propensities are highly desirable for the development of potential chemotherapeutic agents for cancer.

DNA is one of the main targets for anticancer agents. The anticancer activity of cisplatin arises from its ability to damage DNA, with the major Pt-DNA adducts formed being intrastrand d(GpG) and d(ApG) crosslinks (Todd and Lippard 2010; Jung and Lippard 2007; Wang and Lippard 2005; Zhen et al. 1992; Lepre et al. 1990). Coordinated ligands also play an important role towards the binding of platinum complexes to DNA. Multiple modes of DNA binding can be achieved by changing different ligands. Thus, the new drug design was based on new binding modes between DNA and platinum (Pt)-based chemotherapy (Ruhayel et al. 2012; Pereira-Maia and Garnier-Suillerot 2003; Jamieson and Lippard 1999; Wheate et al. 2003; Francisco et al. 2011).

Binding to plasma proteins is regarded as a reason for the platinum (Pt)-based low general toxicity. Serum albumin is the most abundant plasma protein distributed in the blood, and more than 90% plasma protein can be combined with metal drugs in form of co-valent bonds (Huang et al. 2004; Su et al. 2008; Marzano et al. 2002). Such interactions between platinum and proteins have attracted considerable attention, consequently affecting the therapeutic dose of the drugs transferred to their target positions (Marzano et al. 2002; Matesanz et al. 2011).

So in this paper, we investigated the interaction of four novel mononuclear platinum(II) complexes with CT-DNA and bovine serum albumin (BSA) protein, supported by UV–vis, fluorescence. The complex C3 was tested for DNA nuclease activity against pBR322 plasmid DNA. The in vitro anticancer activity of the complex C3 was assessed by the MMT assay in MCF-7, HT-29 and HepG-2 cancer cells and compared to that of cisplatin.

Experimental

Materials and methods

All chemical solvents were dehydrated and distilled by standard methods before use. Bovine serum albumin, ethidium bromide (EB), pBR322 DNA and calf thymus DNA (CT-DNA) were purchased from Sigma.

Physical measurements

IR spectra were recorded on a Nicolet 380 spectrometer as KBr pellets in the range 4000–400 cm−1 with the OMNIC software. Electrospray ionization mass spectrometry (ESI-MS) was analyzed by using a Triple TOF™ 5600+ system with an ion spray source in the positive-ion mode. Electronic spectra were measured by employing a Hewlett Packard HP-8453 spectrophotometer. Fluorescence spectral data were obtained by means of a RF-5301 fluorescence spectrophotometer at the approved temperature.

DNA binding experiments

UV absorption and fluorescence spectrometry relative binding of the complex to calf thymus (CT-DNA) were studied in 5 mM Tris–HCl-50 mM NaCl buffer (pH 7.20). For CT-DNA free of protein, the ratio of UV absorbance at 260 nm to that at 280 nm should be 1.8–1.9. The CT-DNA concentration per nucleotide was determined by absorption spectrometry with a molar extinction coefficient value of 6600 M−1 cm−1 at 260 nm. With varying amounts of CT-DNA to complex, the absorption spectra of complex binding to DNA was changed. Through the McGhee–von Hippel (MvH) method (Marmur 1961), the change in the absorption intensity of the spectral bands, the intrinsic equilibrium binding constant (K b ) and the binding site size (s) of C3 to CT-DNA was obtained by regression analysis using the equation (Wolf et al. 1987):

$$\frac{{\varepsilon_{\text{a}} - \varepsilon_{f} }}{{\varepsilon_{b} - \varepsilon_{f} }} = \frac{{(b - (b^{2} - 2K_{\text{b}}^{2} C_{t} [DNA]/s){}^{1/2})}}{{2K_{b} C_{t} }}$$
(1)
$${\text{b}} = 1 + K_{\text{b}} C_{t} + K{}_{b}[DNA]/2s$$
(2)

where ε a is the absorb intensity observed at a given DNA concentration, ε f is the absorb intensity of free complex in solution, ε b is the absorb intensity of complex when fully bound to DNA (it is assumed that further addition of DNA does not change the absorb intensity), K b is the equilibrium binding constant, C t is the total DNA concentration, (DNA) is the DNA concentration, and s is the binding site size. Both K b and s are obtained from the best fit line. The non-linear least-squares analyses were conducted through the use of Origin Lab, version 7.5. By the fluorescence spectral method, the apparent DNA binding constant (K a ) values of complex were obtained by using an EB-bound CT-DNA solution in 5 mM Tris–HCl–50 mM NaCl buffer (pH 7.20) (Lakowicz et al. 1973). The sample was excited at 510 nm, and the emission range was recorded at 500–700 nm. The fluorescence intensities were recorded with increasing amounts of the added complex concentrations. Such an experiment was carried out by titrating the Pt(II) complex into EB-DNA solution at 580 nm. Due to fluorescence quenching of the free EB by the solvent molecules, EB showed no apparent emission in Tris-buffer. The non-linear least-squares analyses were also done with Origin Lab, version 7.5.

Bovine serum albumin binding

A protein binding study was performed by tryptophan fluorescence quenching experiments with BSA stock solution (BSA, 10−4 M, based on its molecular mass of 66 kDa) in 10 mM Tris–HCl buffer (pH 7.20) to keep the physiological conditions. Fluorescence emission spectra were recorded in the wavelength range 300–500 nm by an excitation wavelength of 285 nm. With the complex concentration increasing, the quenching of the emission intensity of tryptophan residues of BSA at 344 nm was monitored with the Pt(II) complex as a quencher. Fluorescence measurements were carried out by means of a RF-5301 spectrofluorimeter by keeping the concentration of BSA constant (1.0 μM). A linear fit of the data was done by using the Stern–Volmer equation: \({\text{F}}_{0} /{\text{F}} = 1 + {\text{k}}_{\text{q}} \tau_{0} \left[ {\text{Q}} \right] = 1 + {\text{K}}_{\text{sv}} \left[ {\text{Q}} \right]\) (Han et al. 2012).

MTT assay

MTT assay was used to assess the influence of the Pt(II) complexes on the growth of HepG2, MCF7 and HT29 cells. Briefly, cells (HepG2, MCF7 and HT29) at a density 4 × 104 were plated into 96-well microplates and incubated at 37 °C in an atmosphere of 5% CO2 in air overnight and three replica wells were used for controls. Those compounds with the range of concentrations (0, 12.5, 25, 50 and 100 μM) were added and incubated for 48 h. Then, 20 μL of MTT (5 mg mL−1) in phosphate buffered saline (PBS, pH 7.4) was added to each well and incubated for further 4 h. Then medium was discarded and the formazan dissolved with DMSO (100 μL per well). The absorbance was determined at 590 nm with an enzyme-linked immunosorbent assay (ELISA) reader.

DNA cleavage experiments

The DNA cleavage experiments were done by agarose gel electrophoresis (Gao et al. 2015). It was performed by incubation at 37 °C as follows: pBR322 DNA (0.1 μg μL−1) in 50 mM Tris–HCl buffer, pH 7.2, containing 20 mM NaCl treated with C3. After incubation for about 3 h, the loading buffer (0.25% bromophenol-blue, 45% glycerol and 2 mM EDTA) was added. Then the sample was electrophoresed for 2 h at 120 V on 0.9% agarose gel using Tris–boric acid–EDTA buffer and stained with 1 mg·mL−1 EB. In the end, bands were visualized by UV light and photographed.

Preparation of Pt(II) complexes

Synthesis of H2 L1 (C20H16N2O2) and PtII L1·H2O (C20H14N2O2Pt·H2O, C1 )

H2 L1 was synthesized according to the procedure described in previous literature (Ding et al. 2009). Yield: 90.6%. Electrospray ionization mass spectrometry: m/z 316.12 [H2 L1 + H]+, 317.1233. Selected IR data [KBr, max (cm−1)]: 3053 w, 2986 w, 1615 s, 1484 s, 1275 s, 1188 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C20H16N2O2: C%: 75.93; N%: 8.86; H%: 5.10, found: C%: 75.90; N%: 8.87; H%: 5.14.

One millimole of ethylenediamine solution was dropped to solution of 1 mmol H2 L1 in absolute methyl alcohol (30 mL) to make H2 L1 completely dissolved, and then 1 mmol Potassium platinochloride in 3 mL water was dropwise to the resulting solution. Leave the reaction stirring at room temperature for 6 h. The crude product of Pt(II) metal complex C1 was obtained and washed by distilled water, ethanol for three times respectively. Yield: 65.4%. Electrospray ionization mass spectrometry: m/z 509.07, [C1 + H]+, 510.0679. Selected IR data [KBr, max (cm−1)]: 3053 w, 2986 w, 1615 s, 1480 s, 1277 s, 1190 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C20H14N2O2Pt·H2O: C%: 45.54; N%: 5.31; H%: 3.06, found: C%: 45.58; N%: 5.28; H%: 3.19.

Synthesis of L2 (C22H16N2O2) and PtIIL2Cl2(C22H16Cl2N2O2Pt, C2)

L2 was prepared by following the same general procedure as described for H2 L1, except that 1,4-phthalaldehyde (10 mmol) was reacted instead of Salicylaldehyde. Yield: 39.2%. Electrospray ionization mass spectrometry: m/z 340.37 [L 2  + H]+, 341.1225. Selected IR data [KBr, max (cm−1)]: 2876 w, 1691 s, 1611 s, 1300 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C22H16N2O2: C%: 77.63; N%: 8.23; H%: 4.74, found: C%: 77.60; N%: 8.26; H%: 4.76. C2 was prepared by following the same general procedure as described for C1, except that Cis-diamminedichloro-platinum(II) (0.5 mmol) was reacted instead of Potassium platinochloride. Yield: 59.8%. Electrospray ionization mass spectrometry: m/z 352.25 [C2 + 3CH3OH + 2H]2+ 352.0230. Selected IR data [KBr, max (cm−1)]: 2876 w, 1691 s, 1611 s, 1300 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C22H16Cl2N2O2Pt: C%: 43.58; N%: 4.62; H%: 2.66, found: C%: 43.54; N%: 4.67; H%: 2.64.

Synthesis of L3 (C20H16N2) and PtIIL3Cl2·H2O (C20H16Cl2N2Pt·H2O, C3)

L3 was prepared by following the same general procedure as described for H2 L1, except that benzaldehyde was reacted instead of salicylaldehyde. Yield: 30.1%. Electrospray ionization mass spectrometry: m/z 284.35 [L3 + H]+, 285.1335. Selected IR data [KBr, max (cm−1)]: 2980 w, 1638 s, 1296 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C20H16N2: C%: 84.48; N%: 9.85; H%: 5.67, found: C%: 84.51; N%: 9.87; H%: 5.64. C3 was prepared by following the same general procedure as described for C2. Yield: 48.4%. Electrospray ionization mass spectrometry: m/z 549.03 [C3 + 2H]2+, 548.5346. Selected IR data [KBr, max (cm−1)]: 2917 s, 1631 m, 1311 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C20H16Cl2N2Pt·H2O: C%: 42.26; N%: 4.93; H%: 3.19, found: C%: 42.19; N%: 4.91; H%: 3.39.

Synthesis of L4 (C18H14N) and PtIIL4Cl2·0.4H2O (C18H14Cl2N4Pt·H2O, C4)

L4 was prepared by following the same general procedure as described for H2 L1, except that 2-pyridinecarboxaldehyde was reacted instead of Salicylaldehyde. Yield: 35.7%. Electrospray ionization mass spectrometry: m/z 286.33 [L4 + H]+, 287.1219. Selected IR data [KBr, max (cm−1)]: 1648 m, 1308 m (s, strong; m, medium; w, weak). Elemental analysis calcd for C18H14N4: C%: 75.50; N%: 19.57; H%: 4.93, found: C%: 75.49; N%: 19.56; H%: 4.95. C4 was prepared by following the same general procedure as described for C4. Yield: 25.9%. Electrospray ionization mass spectrometry: m/z 157.10 [C4 + 4H2O + 4H]4+, 157.0349. Selected IR data [KBr, mmax (cm−1)]: 1632 m, 1293 s (s, strong; m, medium; w, weak). Elemental analysis calcd for C18H14Cl2N4Pt·0.4H2O: C%: 38.64; N%: 10.01; H%: 2.67, found: C%: 38.68; N%: 9.97; H%: 2.87.

Results and discussion

Synthesis and structural characterization

As shown in Fig. 1, four Pt(II) complexes were obtained and their chemical structures were determined by elemental analyses (EAs), Electrospray ionization mass spectra (ESI–MS) (Figs. S1, S2, S3, S4, S5, S6, S7, S8) and FT-IR spectra (IR) analysis.

Fig. 1
figure 1

The structure of H2 L1–L4 and C1–C4

Full size image

DNA-binding studies

DNA binding ability is critical for DNA cleavage of effective metal based chemotherapeutic drugs (Bloodgood et al. 2013; Dhar et al. 2006; Sha et al. 2012). The potential binding ability of the complexes C1 – C4 to CT-DNA was studied by UV spectroscopy. The typical titration curve for complex C3 is shown in Fig. 2a (similar spectra of C1 and C2 and C4 are presented as Figs. S9, S10, S11), and a plot of (εa−εf)/(εb−εf) versus [DNA] for the titration of DNA to C1 – C4 is shown in Fig. 2b. The absorption peaks at 276–280 nm as these complexes are attributed to intraligand π–π* transition. With the increasing of CT-DNA, an increase in molar absorptivity (hyperchromism, 2–7 fold) for C1 – C4 was found. It is suggested that C1 – C4 binding to DNA is the intercalative mode involving a strong stacking interaction between an aromatic chromophore and the base pairs of DNA (Baldini et al. 2004; Silvestri et al. 2007; Navarro et al. 2003).

Fig. 2
figure 2

a Absorption spectra of C3 (10 μM) in the absence (dash line) and presence (solid line) of increasing amounts of CT-DNA (2.5, 5.0, 7.5, 10.0, 12.5, 15.0, 17.5, 20.0, 22.5 μM) in 5 mM Tris–HCl/50 mM NaCl buffer (pH 7.2). The arrow shows the absorbance changes on increasing DNA concentration. Inset Plot of (εa−εf)/(εb−εf) versus [DNA] for the titration of DNA to C3. b Plot of (εa−εf)/(εb−εf) versus [DNA] for the titration of DNA to C1 – C4

Full size image

In order to determine the interaction between C1 and C4 and CT-DNA, intrinsic binding constants Kb was calculated according to the Eqs. (1) and (2). As shown in Table 1, the Kb values follow the order: C3 (8.2 × 105 M−1) > C1 (6.4 × 105 M−1) > C2 (5.1 × 105 M−1) > C4 (3.6 × 105 M−1), which shows the similar binding propensity with the previously reported Pt complexes (Bielawska et al. 2010; Shao et al. 2014a, b; Ferreira et al. 2016). In short, the binding constants indicate medium binding strength of the four complexes with CT-DNA.

Table 1 Absorption spectral properties of C1 – C4 bound to CT-DNA
Full size table

Besides, to further clarify the binding of C1 – C4, fluorescence spectrum was used to analyze and verify the form of interaction between C1 and C4 and DNA. The apparent binding constant (Kapp) of the complexes was determined by fluorescence spectral technique using EB. Ethidium bromide emits intense fluorescent light in the presence of DNA due to its strong intercalation between the adjacent DNA base pairs, which could be quenched by the addition of another molecule (Gohel et al. 1999). The relative binding propensity of four complexes to CT-DNA studied with an EB bound CT-DNA solution in 5 mM Tris–HCl/50 mM NaCl buffer (pH 7.20) is shown in Fig. 3 (similar spectra of C1 and C2 and C4 are presented as Figs. S12, S13, S14, respectively). Fluorescence intensities at 580 nm (510 nm excitation) were measured at different complex concentrations. From Fig. 3, the fluorescence spectra show a significant reduction in fluorescence intensity, indicating that the EB are displaced from their DNA binding sites and are replaced by C3 under investigation (Barton et al. 1986).

Fig. 3
figure 3

Emission spectra of EB-CT-DNA in the absence (red line) and in the presence (other lines) of C3 with increasing amounts (2.5, 5.0, 7.5, 10.0, 12.5, 15, 17.5, 20.0, 22.5, 25.0, 27.5 and 30.0 μM). Inset plot of F0/F versus [C3] showing linearity. (Color figure online)

Full size image

According to the classical Stern–Volmer formula: \({\text{F}}_{0} {\text{/ F = 1 + K}}_{\text{sv}} [ {\text{Q]}}\), F0 and F of the formula are fluorescence intensity in absence and presence of DNA respectively; KSV is the Stern–Volmer binding constant; [Q] is the concentration of the complex C1 – C4. As shown in Table 1, the calculated apparent binding constant values (Kapp) follows the order: C3 (7.22 × 105 M−1) > C1 (3.52 × 105 M−1) > C2 (2.19 × 105 M−1) > C4 (1.05 × 105 M−1), which is consistent with the results obtained Kb values by UV spectroscopy. The calculated binding constant is less than 107 M−1 (Wang et al. 2006; Laitinen et al. 2006), indicating that the binding effect between C1 and C4 and DNA is medium.

BSA-binding ability

The interactions between serum albumin and metal complexes have attracted increasing research interest in recent years (Annaraj et al. 2016; Li et al. 2016; Anjomshoa et al. 2015; Lu et al. 2012; Anjomshoa et al. 2016; Mohamadi et al. 2016; Raghav et al. 2016; Easmon et al. 2001; Loganathan et al. 2014; Shao et al. 2014a, b; Galisteo-González and Molina-Bolívar 2014), since serum albumin constitutes nearly 55% of the total protein in blood plasma and it plays a pivotal role in drug transport and drug metabolism (Galisteo-González and Molina-Bolívar 2014). Bovine serum albumin is the most extensively studied serum albumin, due to its about 78% structural homology with human serum albumin (HSA) (Shi et al. 2010; Sevilla et al. 2007). As shown in Fig. 4, the strong fluorescence emission peak at 342 nm corresponds for BSA. The fluorescence intensity of BSA at 342 nm was decreased regularly with the increasing concentration of complexes C3, which confirmed that the interaction between C3 and BSA have occurred. According to the Stern–Volmer formula, \({\text{F}}_{0} /{\text{F}} = 1+ {\text{K}}_{\text{q}} \tau_{0} \left[ {\text{Q}} \right] = 1+ {\text{K}}_{\text{sv}} \left[ {\text{Q}} \right]\), F0 and F of the formula are fluorescence intensity in absence and presence of quencher respectively; Kq is rate constants; Ksv is the quenching constants; τ0 is average life expectancy without quencher (τ0 = 10−8 s) (Grzegorz 2006), [Q] is the concentration of quencher. The calculated values of KSV and kq for the interaction of the complexes with BSA are given in Table 2.

Fig. 4
figure 4

Fluorescence emission spectra of the BSA (1.0 μM) system in the absence and presence of complex C3 (0.25, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75 μM, respectively). Inset the plot of F0/F versus C3 concentration

Full size image
Table 2 The quenching constants (K SV ), binding constants (K a ), binding sites (n) of BSA-C1–C4
Full size table

As shown in Table 2, the K q of C3 was 1.61 × 1012 M−1 s−1, greater than the maximum K q numerically, showing that the quenching is not caused by dynamic quenching, but due to the static quenching of the formation of complex (Guo et al. 2009; Jiang et al. 2004). The equation, \({ \log }[({\text{F}}_{0} - {\text{F}})/{\text{F}}] = { \log }K_{a} + {\text{nlog}}\left[ {\text{Q}} \right]\), could be used to calculate the binding constants and binding sites. K a and n are the apparent binding constants and binding sites respectively (as shown in Fig. 5). The binding constants and binding sites of complexes C1–C4 and BSA could be obtained and are listed in Table 2. It is indicated that C3 exhibits higher binding constants for BSA than other three Pt(II) complexes and the K a and n values follow the order: C3 (7.08 × 105M−1, 1.21) > C1 (2.82 × 105M−1, 1.14) > C2 (0.85 × 105M−1, 0.97) > C4 (0.15 × 105M−1, 0.86). The values of n are associated with binding constants K a , which verify the conclusion that there is a direct relation between the binding constant and number of binding sites (Sathyadevi et al. 2011).

Fig. 5
figure 5

Plot of log (F0 − F)/F versus log[Q] for BSA in the presence of complex C1 – C4

Full size image

MTT assay

MTT assay is a colorimetric assay based on the conversion of the yellow tetrazolium salt to purple formazan crystals by metabolically active cells. The cytotoxic effects of C3 on the viability of several human cancer cell lines (MCF-7, HepG2 and HT29) were examined by MTT assay. A variety of tumor cells were treated with C3 and incubated for 48 h at increasing concentration, respectively. As shown in Fig. 6, C3 shows better antitumor effects on the tested cancer cells.

Fig. 6
figure 6

Cell viability of MCF-7, HepG2 and HT29 cells treated with C3 at 0, 12.5, 25, 50 and 100 μM for 48 h

Full size image

DNA cleavage studies

To explore the DNA cleavage abilities of C3, the extent of DNA cleavage was monitored by agarose gel electrophoresis. The supercoiled pBR322 plasmid DNA as a substrate was incubated with C3 in a medium of 50 mM Tris–HCl buffer containing 20 mM NaCl (pH 7.2) for 3 h. The concentration-dependent DNA cleavage activities by C3 were observed without any external agents. As shown in Fig. 7, the results carried out in the concentration range 20–80 μM for C3 and could cause DNA cleavage at 20 μM, which suggests that C3 partially degraded SC DNA into undetectable minor fragments.

Fig. 7
figure 7

The concentration dependent cleavage of pBR322 DNA (30 μM in base pair) with different concentration of C3. C, DNA; lane 1, DNA + C3 (20 μM); lane 2, DNA + C3 (40 μM); lane 3, DNA + C3 (60 μM); lane 4, DNA + C3 (80 μM)

Full size image

Conclusions

A series of mononuclear platinum(II) complexes have been synthesized and characterized. The partial intercalation between the complexes and CT-DNA has been confirmed by using absorption and emission spectral methods and medium binding strength follows the order: C3 > C1 > C2 > C4. With the single condition change such as absence of an external agent, the DNA cleavage abilities of C3 exhibit remarkable changes. Further, the ability to bind to BSA also has been explored, and the quenching mechanisms of BSA by the complexes are static procedures. The vitro cytotoxicity of C3 on cancer cell lines (MCF-7, HepG2 and HT29) has been assessed by MTT and the results indicate that C3 shows better antitumor effects on the tested cells.