Introduction

Since the successful discovery of cisplatin in the treatment of types of cancers, platinum-based anticancer drugs have attracted great attention in the chemotherapeutic studies [1, 2]. Simultaneously, however, serious side effects have been limiting the further uses of the classic platinum-based drugs. To overcome the drawbacks, various non-platinum transition metals, such as ruthenium, osmium, gold, palladium, and rhodium-based drugs have been investigated as potential anticancer therapeutics [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]. Very recently, possible biological applications of iridium complexes have received much interests [21,22,23,24,25,26,27,28,29,30], in which organoiridium anticancer agents have shown very promising antiproliferative activity toward cancer cells [31,32,33,34]. In addition, iridium complexes were found displaying novel anticancer mechanism of actions (MoAs) different from platinum drugs, and showing high antiproliferative activity against cisplatin-resistant cancer cells [35, 36], thus could be a potential candidates for treatment of platinum resistant cancers.

One of the most important reasons people select metal complexes as potential anticancer drugs is their structural diversity. A wide variety of metals and ligands with steric and electronic features can be selected to make a huge amount of combinations and incorporations. Moreover, any subtle changes in the composition or in the structure of complex may cause totally different effects on chemical and biological activities. For example, antiproliferative activity of organoiridium complexes against cancer cells can be switched on readily by replacing one nitrogen atom in 2,2′-bipyridine with one carbon atom to form ligand phenyl pyridine [37], where the new complex displayed a much stronger binding to nucleobases and higher hydrophobicity.

In the potential ligand candidates, N^N-donor chelating ligands present one of the main subjects which have been widely used in chemical and biological evaluation [38,39,40]. Iminopyridine type of ligands has been attracting much attention in catalysis [41] and may play an important role in redox procedure of transition metal complexes [42, 43]. Interestingly, anticancer agents and catalysts are sometimes inseparable [44]. Some ruthenium transfer hydrogenation catalysts with poor catalytic activity showed high anticancer activity [45]. In addition, investigation on the catalytic behavior of iridium complexes helped to understand their MoAs as anticancer agents [46]. Very recently, iminopyridine type of ligands started to be used in the design of metal-based anticancer agents. Sadler group and our group have reported a panel of osmium(II) [47] and iridium(III) [48] phenyliminopyridine complexes, respectively, which have shown very promising anticancer activities against various cancer cell lines.

On the basis of our previous work, herein we newly explored adamantyl iminopyridine and oxime pyridyl as novel N^N-chelating iminopyridine ligands, and synthesized eight iridium half-sandwich complexes containing Cp* and its phenyl substituents. Their hydrolysis chemistry, nucleobase binding, catalytic transfer hydride reactions, in vitro cancer cell toxicity, and basic mechanism of action were studied.

Experimental section

Materials

IrCl3·nH2O (99%), 1, 2, 3, 4, 5-pentamethyl-cyclopentadiene (95%), 2, 3, 4, 5-tetramethyl-2-cyclopentanone (95%), 4-bromobiphenyl (98%), butyllithium solution (1.6 M in hexane), ammonium hexafluorophosphate (NH4PF6), 9-ethylguanine, and 9-methyladenine were purchased from Sigma-Aldrich. Solvents of laboratory grade were used for syntheses without further purification. The starting Ir(III) dimers [(η5-Cp*)IrCl2]2, [(η5-Cpxph)IrCl2]2 and [(η5-Cpxbiph)IrCl2]2, and ligand L1 were prepared according to reported methods [49]. Ligand L2 (methyl 2-pyridyl ketoxime) were purchased from Alfa Aesar. For the biological experiments, phosphate-buffered saline (PBS), fetal bovine serum, DMEM medium, trypsin/EDTA, and penicillin/streptomycin mixture were purchased from Sangon Biotech. Testing compounds were dissolved in DMSO and diluted with the tissue culture medium before use. All reactions were performed under an N2 atmosphere.

Synthesis of the [(η 5-Cpx)Ir(N^N)Cl]PF6

General method

A solution of chloride bridged dimer [(η5-Cpx)IrCl2]2 (0.05 mmol) and iminopyridine ligand (0.12 mmol) in MeOH (20 ml) was heated under reflux in an N2 atmosphere for 10 h and filtered. The volume was slowly reduced to half on a rotary evaporator and NH4PF6 (0.42 mmol) was added. After standing at 298 K, a microcrystalline product formed. This was collected by filtration, washed with diethyl ether, and recrystallized from CH2Cl2/diethyl ether. The 1H NMR (500.13 MHz, CDCl3 or DMSO-d6) peak integrals of complexes 1–6 are shown in Figs. S7–S12. The high-resolution mass spectrometry data of complexes 1–6 are shown in Fig. S13.

[(η 5-Cp*)Ir(L1)Cl]PF6 (1)

Yield: 46.4 mg, 62.0%. 1H NMR (500.13 MHz, CDCl3) δ 8.98 (s, 1H), 8.72 (d, J = 5.4 Hz, 1H), 8.23 (d, J = 7.6 Hz, 1H), 8.08 (t, J = 7.2 Hz, 1H), 7.69 (m, 1H), 2.30 (s, 3H), 2.20 (d, 3H), 2.05 (d, 3H), 1.77 (s, 6H), 1.69 (s, 15H). Anal. calcd for [(η5-Cp*)Ir(L1)Cl]PF6 (748.18): C, 41.74; H, 4.72; N, 3.74. Found: C, 41.69; H, 4.76; N, 3.77. MS: m/z 603.42 [(η5-C5Me5)Ir(L1)Cl]+. Crystals suitable for X-ray diffraction were obtained by slow evaporation of a CH2Cl2/diethyl ether solution at ambient temperature.

[(η 5-Cpxph)Ir(L1)Cl]PF6 (2)

Yield: 34.5 mg, 42.6%. 1H NMR (500.13 MHz, CDCl3) δ 9.00 (s, 1H), 8.24 (dd, J = 14.5, 6.8 Hz, 2H), 8.01 (t, J = 7.7 Hz, 1H), 7.51 (m, J = 6.1 Hz, 5H), 7.36 (m, 1H), 2.26 (s, 3H), 2.20 (d, J = 11.9 Hz, 4H), 2.02 (m, 5H), 1.90 (d, J = 4.2 Hz, 6H), 1.74 (d, J = 3.0 Hz, 6H), 1.59 (s, 3H). Anal. calcd for [(η5-Cpxph)Ir(L1)Cl]PF6 (810.19): C, 45.95; H, 4.60; N, 3.46. Found: C, 45.87; H, 4.65; N, 3.51. MS: m/z 655.50 [(η5-C5Me4C6H5)Ir(L1)Cl]+.

[(η 5-Cpxbiph)Ir(L1)Cl]PF6 (3)

Yield: 62.4 mg, 70.4%. 1H NMR (500.13 MHz, CDCl3) δ 9.02 (s, 1H), 8.28 (s, 2H), 8.01 (s, 1H), 7.72 (d, 1H), 7.65 (d, 1H), 7.61 (d, 1H), 7.42–7.49 (m, 5H), 7.27 (d, 2H), 2.21 (t, 4H), 2.05 (d, 3H), 1.92 (d, J = 4.8 Hz, 3H), 1.77 (d, J = 13.8 Hz, 4H), 1.62 (s, 2H), 1.50 (s, 11H). Anal. calcd for [(η5-Cpxph)Ir(L1)Cl]PF6 (886.22): C, 50.14; H, 4.66; N, 3.16. Found: C, 50.22; H, 4.58; N, 3.24. MS: m/z 741.50 [(η5-C5Me4C6H4 C6H5)Ir(L1)Cl]+.

[(η 5-Cp*)Ir(L2)Cl]PF6 (4)

Yield: 18.2 mg, 28.3%. 1H NMR (500.13 MHz, CD3CN) δ 8.65 (d, J = 5.6 Hz, 1H), 8.02 (t, J = 7.9 Hz, 1H), 7.81 (d, J = 8.1 Hz, 1H), 7.53 (dd, J = 7.3, 5.9 Hz, 1H), 3.27 (s, 1H), 2.45 (d, J = 1.3 Hz, 3H), 1.68 (s, 15H). Anal. calcd for [(η5-Cp*)Ir(L2)Cl]PF6 (644.08): C, 31.70; H, 3.60; N, 4.35. Found: C, 31.65; H, 3.67; N, 4.42. MS: m/z 463.33 [(η5-C5Me5)Ir(L2)]2+.

[(η 5-Cpxph)Ir(L2)Cl]PF6 (5)

Yield: 19.6 mg, 27.8%. 1H NMR (500.13 MHz, CDCl3) δ 8.30 (s, 1H), 7.89 (t, 1H), 7.71 (d, 1H), 7.56 (s, 2H), 7.49 (s, 3H), 7.27 (s, 1H), 2.52 (s, 3H), 1.86 (s, 9H), 1.71 (s, 3H). Anal. calcd for [(η5-Cpxph)Ir(L2)Cl]PF6 (706.09): C, 37.42; H, 3.57; N, 3.97. Found: C, 37.45; H, 3.51; N, 3.89. MS: m/z 525.33 [(η5-C5Me4C6H5)Ir(L2)]2+.

[(η 5-Cpxbiph)Ir(L2)Cl]PF6 (6)

Yield: 20.18 mg, 25.8%. 1H NMR (500.13 MHz, CDCl3) δ 8.47 (d, J = 5.5 Hz, 1H), 8.08 (t, J = 7.9, 1.2 Hz, 1H), 7.98 (d, J = 7.6 Hz, 1H), 7.75 (d, J = 8.3 Hz, 2H), 7.70 (d, J = 8.3 Hz, 2H), 7.65 (m, 2H), 7.59 (t, 1H), 7.50 (t, J = 7.6 Hz, 2H), 7.42 (t, J = 7.4 Hz, 1H), 2.71 (s, 3H), 1.93 (s, 3H), 1.90 (s, 3H), 1.85 (s, 3H), 1.79 (s, 3H). Anal. calcd for [(η5-Cpxph)Ir(L2)Cl]PF6 (782.18): C, 43.00; H, 3.74; N, 3.58. Found: C, 43.08; H, 3.71; N, 3.62. MS: m/z 601.42 [(η5-C5Me4C6H4 C6H5)Ir(L2)]2+.

X-ray crystallography

Diffraction data of complex 1 were obtained on a Bruker Smart Apex CCD diffractometer equipped with graphite-monochromated Mo Kα radiation. SADABS program was used to correct the date for absorption effects. The crystal with the Oxford Cryosystems Cobra was kept at 100 K throughout the data collection process. The structure was solved by direct methods [SHELXS (TREF)] with additional light atoms found by Fourier methods and refined against F2 using SHELXL.

The crystal data for complex [(η5-Cp*)Ir(L1)Cl]PF6 1 are available as Fig. 1, and Tables 1 and 2, and have been deposited in the Cambridge Crystallographic Data Centre with CCDC reference number 1574955. X-ray crystallographic data in CIF format can be obtained free of charge from The Cambridge Crystallographic Data Centre via http://www.ccdc.cam.ac.uk/.

Fig. 1
figure 1

Cyclopentadienyl iridium iminopyridine complexes 1–6

Full size image
Table 1 Crystallographic data for [(η5-Cp*)Ir(L1)Cl]PF6 (1)
Full size table
Table 2 Selected bond lengths (Å) and angles (deg) for [(η5-Cp*)Ir(L1)Cl]PF6 (1)
Full size table

NMR spectroscopy

1H NMR spectroscopy was obtained in 5 mm NMR tubes at 298 K on Bruker DPX-500 spectrometers. 1H NMR chemical shifts were internally referenced to (CHD2)(CD3)SO (2.50 ppm) for DMSO-d6 and CHCl3 (7.26 ppm) for chloroform-d1 solutions. The data were processed using XWIN-NMR version 3.6 (Bruker UK Ltd).

Mass spectrometry

Electrospray ionization mass spectra (ESI–MS) were obtained on LCQ FLEET. Samples were prepared in 60% MeOH/40% H2O (v/v). The mass spectra were recorded with a scan range of m/z 200–1500 for positive ions.

UV–Vis spectroscopy

A TU-1901 UV–Vis recording spectrophotometer with 1 cm path-length quartz cuvettes (3 ml) was used. All spectra were processed with UVWinlab software. All experiments were carried out at 298 K unless otherwise stated.

Hydrolysis studies

The final concentrations of compounds 3 and 6 are 1 mM in 30% MeOD-d4/70% D2O (v/v), obtained by dissolving compounds 3 and 6 in MeOD-d4 followed by rapid dilution with D2O. All 1H NMR spectra of different time intervals were recorded at 298 K. The final concentrations of compounds 3 and 6 are 50 µM in 20% MeOH/80% H2O (v/v), obtained by dissolving compounds 3 and 6 in MeOH followed by rapid dilution with H2O. All 1H NMR spectra of different time intervals were recorded at 298 K.

Catalytic oxidation of NADH to NAD+

Complexes 3–6 were evaluated for the catalytic oxidation of NADH to NAD+ by UV–Vis spectroscopy after various time intervals in 20% MeOH/80% H2O at 298 K. The concentration of complexes that remained fixed at 1 μM and NADH concentrations is around 75 μM. The conversion of NADH to NAD+ was followed by absorption at 339 nm (εNADH = 6220 cm−1 M−1) to allow evaluation of kinetic data. As NADH is converted to NAD+ by O2 overtime, a NADH control under the same condition was used for all the UV–Vis experiments.

The reaction between complex 6 (0.25 mM) and NADH (3.5 mol equiv) in 30% MeOD/70% D2O at 298 K was monitored by 1H NMR spectroscopy.

Interaction with GSH

GSH (5 mol equiv) was added to an NMR tube containing a 2 mM solution of complex 3 and 6 in 50% MeOD-d4/50% D2O (v/v) at 298 K under N2 atmosphere. 1H NMR spectra of the resulting solution were recorded at 298 K after various time intervals.

Log P determination

Water saturated octanol (WSO) and octanol-saturated water (OSW) were prepared using 0.05 M aqueous NaCl solution (to suppress hydrolysis of the chlorido complexes) and analytical grade octanol. Mix aliquots of stock solutions of iridium complexes 3 and 6 in OSW and WSO in equal volume and shaken in an IKA Vibrax VXC basic shaker for 4 h at 500 g/min, 298 K, to allow partition at room temperature. The aqueous layer was attentively separated from the octanol layer for iridium analysis. Partition coefficients of metal complexes were calculated using the equation log P = log ([M]WSO/[M]OSW), where [M]WSO was obtained by subtraction of the metal content of the aqueous layer after partition from the metal content of the aqueous layer before partition.

Reactions with nucleobases

Interactions of complexes 3 and 6 (ca. 1 mM) with 9-ethylguanine (9-EtG) and 9-methyladenine (9-MeA) (ca. 1 mM) were studied by 1H NMR spectroscopy after various time intervals after addition of the nucleobase (s) in 30% MeOD/70% D2O (v/v) at 298 K using a Bruker DPX500 NMR spectrometer.

Cell culture

A549 human lung cancer cells were purchased from Shanghai Institute of Biochemistry and Cell Biology (SIBCB). A549 cells were cultured in Dulbecco’s modified Eagle medium (DMEM). All medium were replenished with 10% fetal bovine serum and 1% penicillin–streptomycin solution. All cells were cultured at 37 °C in a humidified incubator under a 5% CO2 atmosphere.

Cell viability assay (MTT assay)

A549 cells were seeded in 96-well plates at a density of 5 × 103 cells per well. After the cells were preincubated in drug-free medium at 37 °C for 24 h, different concentrations of the compounds tested were added. The complexes were dissolved in dimethyl sulfoxide (DMSO) to obtain the stock solution of the drug. This stock was further diluted using cell culture medium to get working concentrations. MTT solution (15 µl, 5 mg/ml) was added to form a purple formazan after 24 h of drug exposure. Then 100 μl of DMSO was added to each well to dissolve the purple formazan. Results were measured by a microplate reader (DNM-9606, Perlong Medical, Beijing, China) at an absorbance of 570 nm. All processing were performed in triplicate. IC50 values quoted are mean ± SEM.

Crystal violet assay

A549 cells were seeded in 96-well plates at a density of 5 × 103 cells per well. After the cells were preincubated in drug-free medium at 37 °C for 24 h, different concentrations of the compounds tested were added. The complexes were dissolved in DMSO to obtain the stock solution of the drug. This stock was further diluted using cell culture medium to get working concentrations. Monolayers were washed twice with PBS and fixed and stained 0.5% crystal violet for 30 min after 24 h of drug exposure. Afterward, plates were washed with hyperpure water and dried and dye was eluted with 0.1 M sodium citrate (pH 4.2) in ethanol (1:1). Results were measured by a microplate reader (DNM-9606, Perlong Medical, Beijing, China) at an absorbance of 570 nm. All processing were performed in triplicate. IC50 values quoted are mean ± SEM.

Cell cycle analysis

A549 cells were seeded in 6-well plates at a density of 1.0 × 106 cells per well. After the cells were preincubated in drug-free medium at 37 °C for 24 h, drugs were added at concentrations of 0.25 × IC50 and 0.5 × IC50. MTT solution (15 µl, 5 mg/ml) was added to form a purple formazan after 24 h of drug exposure. The supernatants were removed after 24 h of drug exposure, and the cells were washed twice with PBS. Afterward, cells were harvested using trypsin–EDTA and fixed in ice-cold 70% ethanol for 24 h. Then, cells rinsed twice with PBS and DNA were stained with propidium iodide (PI) and RNase at 37 °C for 30 min. Cell pellets were washed and resuspended in 200 µl PBS before being analyzed using a flow cytometer (ACEA NovoCyte, Hangzhou, China). Data were analyzed using NovoExpress™ software.

Induction of apoptosis

Apoptotic populations of A549 cells caused by exposure to iridium complexes were carried out using the Annexin V-FITC Apoptosis Detection Kit (Beyotime Institute of Biotechnology, China) by flow cytometry analysis. Momently, A549 cells (1.5 × 106/2 ml per well) were seeded in a 6-well plate. After the cells were preincubated in drug-free medium at 298 K for 24 h, drugs were added at concentrations of 1 × IC50, 2 × IC50, and 3 × IC50. After 24 h of complexes treatment, cells were harvested, washed twice with PBS, and resuspended in 195 μl of binding buffer which was then added to 5 μl of annexin V-FITC and 10 μl of PI, and then incubated at ambient temperature under dark conditions for 15 min. Afterward, the samples were placed in an ice bath in the dark and were analyzed by a flow cytometer (ACEA NovoCyte, Hangzhou, China).

ROS determination

ROS generation in A549 cells caused by exposure to IrIII complexes was carried out using the ROS Assay Kit (Beyotime Institute of Biotechnology, Shanghai, China) by flow cytometry analysis. Momently, 1.0 × 106 A549 cells per well were seeded in a 6-well plate. The cells were preincubated in drug-free medium at 37 °C for 24 h in a 5% CO2 humidified atmosphere, and then drugs were added at concentrations of 0.25 × IC50, 0.5 × IC50. After 24 h of drug exposure, cells were washed twice with PBS and then incubated with the DCFH-DA probe (10 µM) at 37 °C for 30 min, and then washed it three times immediately with PBS. The fluorescence intensity was analyzed by flow cytometry (ACEA NovoCyte, Hangzhou, China). Data were analyzed using NovoExpress™ software. Samples were kept in the darks all the time to avoid light-induced ROS production.

Results and discussion

Three iridium cyclopentadienyl dimers [(η5-Cp*)IrCl2]2 (dimer 1), [(η5-Cpxph)IrCl2]2 (dimer 2) and [(η5-Cpxbiph)IrCl2]2 (dimer 3), where Cp* = C5Me5, Cpxph = C5Me4C6H5, Cpxbiph = C5Me4C6H4C6H5, were synthesized according to a reported method [50]. Six IrIII half-sandwich iminopyridine complexes 1–6 of the type [(η5-Cpx)Ir(N^N)Cl]PF6 were prepared by reaction of corresponding dimers 1–3 with iminopyridine ligand L1–L2 in 1:2.4 ratio in methanol at ambient temperature followed by ion exchange from Cl− to PF6− (Fig. 2; Scheme 1). All the six complexes are newly prepared and fully characterized by mass spectroscopy, 1H NMR spectroscopy and elemental analysis.

Fig. 2
figure 2

X-ray crystal structure for [(η5-Cp*)Ir(L1)Cl]PF6 (1). Hydrogen atoms and counter ions PF6− are omitted for clarity, and thermal ellipsoids are drawn at 50% probability

Full size image
Scheme 1
scheme 1

Synthesis of IrIII complexes [(η5-Cpx)Ir(N^N)Cl]PF6

Full size image

X-ray crystal structure

Crystal of [(η5-Cp*)Ir(L1)Cl]PF6 (1) was grown from CH2Cl2/diethyl ether by vapor diffusion. The molecular structure of 1 was determined (Fig. 2). Crystallographic data and selected bond lengths and angles are listed in Tables 1 and 2. Complex 1 adopts the half-sandwich piano-stool geometry. The distance between iridium to the centroid of bound η5-cyclopentadienyl ligand is 1.8153 Å. The bond length of Ir–Cl is 2.3986(12) Å for 1, which is almost identical to the reported [38] cationic polypyridyl iridium complexes [(η5-C5Me4C6H4C6H5)Ir(bpy)Cl]PF6 (2.3840 (14) Å) and [(η5-C5Me4C6H5)Ir(bpy)Cl]PF6 (2.3859 (5) Å). The Ir–N1 (chelating ligand) bond length (2.087(4) Å) is dramatically shorter than the bond length of Ir–N2 (2.157 (4) Å). The single crystal structure of the cationic part of complex 4 has been reported [51].

Cell viability assay

The effect of cell viability of complexes 1–6 on human lung cancer A549 cells was evaluated by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, and the IC50 values (concentration at which 50% of the cell growth is inhibited) after 24 h drug exposure are listed in Table 3. The Cp* complex 1 deemed as inactive as its IC50 was > 100 µM. However, the rest five complexes were all active against A549 cell line, displaying IC50 values range from 75.9 to 2.7 μM. Basically, for those complexes containing the same chelating ligands, their cell viability decreased with the number of the phenyl ring on the Cp ring in the trend of Cpxbiph < Cpxph < Cp*, which is consistent with our previous results [50]. Introduction of one phenyl ring switched on anticancer activity from complex 1 (IC50 > 100 μM) to complex 2 (IC50 16.9 μM), and the cell viability decreased significantly from complex 2 to 3 (IC50 2.7 μM) arising from introduction of the second phenyl ring on Cp. The additional phenyl ring increases the lipophilicity of the complexes, which in turn enhances the cell membrane penetration [38]. Varying chelating ligands also caused significant differences on the cell viability. The complex 3 containing adamantyl iminopyridine ligand displayed the most potent anticancer activity with IC50 value of 2.7 μM, ca. 4 and 8 ×more active than its oxime pyridyl analog and cisplatin, respectively. Given the best anticancer activity displayed by complexes 3 and 6, the two complexes were chosen to perform more experiments to investigate their MoAs. The cell viability assay was also determined by crystal violet method (Table S1). There is a significant difference from the results of MTT assay.

Table 3 IC50 values for iridium complexes 1–6 and cisplatin against A549 cancer cells determined by MTT assay
Full size table

Hydrolysis studies

In comparison with metal chloride complexes, metal aqua complexes often display more reactive activity, and therefore the hydrolysis of the M–Cl bonds (M = metal) can represent an activation step for transition metal-based anticancer complexes, for example cisplatin [38]. The hydrolysis of complexes 3 and 6 was studied by 1H NMR spectroscopy in 30% MeOD-d4/70% D2O at 298 K, where methanol was used to avoid precipitation of the complexes in aqueous solution. However, no hydrolysis was observed by NMR for complexes 3 and 6 in 30% MeOD-d4/70% D2O at 298 K (Figure S1). The UV–Vis spectra also could not detect obvious hydrolysis for complexes 3 and 6 (50 µM) in 20% MeOH/80% H2O (v/v) (Figure S2), indicating that these complexes remained stable under the above conditions.

Interaction with nucleobases

Like cisplatin, transition metal anticancer complexes often bind to DNA and thus cause cell death. The binding of complexes 3 and 6 to nucleobase models 9-EtG (9-ethylguanine) and 9-MeA (9-methyladenine) was investigated. Nucleobases 9-EtG or 9-MeA was added to an solution containing iridium complex (1 mM) in 30% MeOD-d4/70% D2O (v/v), and the solution was recorded by 1H NMR spectra at 298 K after various time intervals. Based on 1H NMR peak integrals, the percentages of iridium nucleobase adduct formed after 24 h are shown in Table 4. Although no hydrolysis for complex [(η5-Cpxbiph)Ir(L2)Cl]PF6 (6), the addition of 9-EtG to an solution of 6 led to the appearance of one set of new peaks assignable to the 9-EtG adduct, showing that 86% of 6 reacted after 24 h (Fig. 3a). Complex 6 also showed the high affinity for 9-MeA with 40% nucleobase adducts formation after 24 h (Fig. 3b). However, no any nucleobase adduct formation was observed for complex 3 with 9-EtG and 9-MeA (Figure S3), most likely due to the steric hindrance of the adamantane. These results indicate that the chelating ligand plays a crucial role in reaction with nucleobases.

Table 4 Extent of adduct formation for complexes 3 and 6 (ca. 1 mM) with 9-EtG and 9-MeA at 298 K after 24 h
Full size table
Fig. 3
figure 3

Reactions of complex [(η5-Cpxbiph)Ir(L2)Cl]PF6 (6) (1 mM) with nucleobases a 9-EtG; b 9-MeA in 30% MeOD-d4/70% D2O (v/v) at 298 K recorded by 1H NMR spectrum at various times. Peaks labeled diamond correspond to the new formed nucleobases adduct, and peaks labeled triangle correspond to complex 6

Full size image

Catalysis of NADH oxidation

Coenzyme NADH and NAD+ is crucial in numerous biocatalyzed processes, such as electron transfer in the mitochondrial electron chain. Previously, we have shown that cyclopentadienyl IrIII aqua complexes can catalytically convert NADH to its oxidized form NAD+ by robbing hydride from NADH, and pass the hydride to oxygen to finally generate reactive oxygen species (ROS H2O2), thus provide a pathway to an oxidant MoA [46, 52]. First, we investigated the reaction between complex 6 (0.25 mM) and NADH (3.5 mol equiv) in 30% MeOD-d4/70% D2O (v/v), which was recorded by 1H NMR at 298 K. When NADH was added to a solution of complex 6, NADH was gradually converted to NAD+ and a sharp singlet correspond to Ir–H at − 11.4 ppm was detected.

To evaluate the catalytic abilities of the iridium complexes, the reaction between NADH (around 75 μM) and complexes 3–6 (fixed concentration of complex at 1.0 μM) in 20% MeOH/80% H2O (v/v) was monitored by UV–Vis spectroscopy over 8 h at 298 K (Fig. 4b; Figure S4). As NADH has an obvious absorbance at 339 nm while NAD+ does not, the decrease of absorbance intensity at 339 nm was monitored and the data were used for calculation of the turnover number (TON) for each complex (Fig. 4c). Basically, the complexes displayed good catalytic activity in conversion of NADH to NAD+, with TONs range from 6.8 of complex 5 to 19.2 of complex 6. Complexes 3 and 6 showed higher catalytic ability than did 4 and 5, which is consistent with their anticancer activity. The TONs of these complexes are similar with the half-sandwich IrIII N-heterocyclic carbene complexes [50]. The Sadler group has reported that the iridium complex could modify the NAD+/NADH ratio in A2780 cancer cells [52]. Therefore, the potent catalytic activity of these complexes in converting NADH to NAD+ may provide a basis to modulate the redox balance and generate ROS in cells [46, 53]. In a solution of 6 (0.5 mM) with NADH (3 mol equiv) in 50% MeOH/50% H2O (v/v) at 298 K, the ROS hydrogen peroxide was detected by appearing blue on hydrogen peroxide test paper (Fig. 4d). It is found that the concentration of H2O2 is about 0.18 mM, and its level may be limited by the oxygen solubility (ca. 0.23 mM at 288 K) [46].

Fig. 4
figure 4

a 1H NMR spectra recorded the reaction between complex [(η5-Cpxbiph)Ir(L2)Cl]PF6 (6) (0.25 mM) and NADH (3.5 mol equiv) in 30% MeOD/70% D2O (v/v) at 298 K. b UV–Vis spectra recorded the reaction of NADH (about 75 µM) with complex 6 (1 µM) in 20% MeOH/80% H2O (v/v) at 298 K for 8 h. c Turnover numbers of complexes 3–6. d Detection of hydrogen peroxide (H2O2) in a solution of complex 6 (0.5 mM) with NADH (3 mol equiv) in 50% MeOH/50% H2O (v/v) at 298 K. After 24 h, H2O2 (ca. 0.18 mM) was detected by peroxide-test paper

Full size image

Interaction with GSH

GSH can coordinate to the metals in anticancer drugs to form less toxic GSH conjugates, thus detoxifying anticancer drugs [46]. Therefore, the stability of Cpxbiph complexes 3 and 6 to GSH was explored. GSH (5 mol equiv) was added to an 50% MeOD-d4/50% D2O (v/v) solution containing complex 3 or 6 (2 mM) under N2 atmosphere, and the solution was recorded by 1H NMR spectra at 298 K after various time intervals. No new peaks were observed by NMR for complex 3 (Figure S5). The result shows that adamantyl iminopyridine complex 3 is stable in the solution of GSH, which may avoid inactivating before reaching its biological targets. However, 1H NMR spectra showed that 6 had reacted with GSH after 10 min (Figure S6). The less stability of complex 6 in the presence of GSH may lead to less potency than complex 3 against cancer cells.

Hydrophobicity (log P)

The octanol/water partition coefficients (log P) are usually a factor relevant for cellular uptake and anticancer activity [38]. The log P values for the complexes 3 and 6 were determined (Table 5). To make sure that the determined log P values are for the chlorido complexes, NaCl (0.05 M) was added to the solution of the compounds to inhibit hydrolysis. The log P values for complexes 3 and 6 are 1.16 and 0.70, respectively, suggesting that complex 3 possesses higher hydrophobicity than 6, and therefore contributed to the more potent anticancer activity of complex 3 than 6. This result showed that the hydrophobicity of the complexes is related to their anticancer activity.

Table 5 The log P values for complexes 3 and 6
Full size table

Apoptosis assay

Cpxbiph complexes 3 and 6 were chose to explore more details on their MoAs using flow cytometry. Apoptosis is the process where cells die in a programmed fashion under particular stimuli. A large number of publications have reported that metal-based anticancer agents could inhibit cell growth by causing apoptosis [54,55,56,57]. To investigate whether the cell viability of the complexes is a result of apoptosis, A549 cells were exposed to complexes 3 and 6 at IC50, 2 × IC50 and 3 × IC50 concentrations for 24 h, followed by annexin V/propidium iodide staining, and then determined by flow cytometry. Cell populations can be distinguished as viable (annexin V−/PI−), nonviable (annexin V−/PI+), early apoptosis (annexin V+/PI−), and late apoptosis (annexin V+/PI+). Compared to the high viability of untreated cells (93.1%), obvious induction of apoptosis was observed in those cells treated with complexes 3 and 6 (Fig. 5; Table S2). The plots showed that around 4.3%, A549 cells were in early apoptotic phase and 23.2% cells were in the late apoptotic phase after 24 h of exposure to complex 3 at a concentration of IC50. The proportion of cells in late apoptosis increased significantly with higher concentration of complex 3, where 94.3% of cells were observed in this stage upon exposure to 3 at 3 × IC50. Interestingly, complex 6 remained most cells in early apoptosis with 84.2% in early apoptosis and only 8.9% in late apoptosis at 3 × IC50.

Fig. 5
figure 5

a Apoptosis of A549 cells induced by complexes 3 and 6. Annexin V-FITC vs PI staining and determined by flow cytometry. b Bar chart showing cell populations in various phases

Full size image

Therefore, both complexes 3 and 6 dramatically induced apoptosis of A549 cells in a dose-dependent manner. However, there were big differences in their mode of action. In comparison to the complex 3 which caused most cells in late apoptosis, complex 6 was mainly concentrated in early apoptosis. The stronger ability to cause apoptosis of complex 3 is consistent with its lower the cell viability compared to complex 6.

Cell cycle arrest

Arrest of cell cycle of cancer cells has been recognized as an effective strategy in chemotherapy [57]. In this work, compounds 3 and 6 were investigated their effects on the cell cycle of A549 cells. Treating of A549 cells with compound 3 at 0.5 ×IC50 led to obvious G1 phase arrest, where the percentages of cells increased 12.6% compared to untreated cells (Fig. 6; Table S3). In addition, a sub-diploid peak corresponds to sub-G1 was detected, usually indicating cells are undergoing apoptosis. This result is consistent with the apoptosis assay. Upon exposure of the cells to complex 6 at a concentration of 0.5 × IC50, there were significant increases in the proportions of cells at S phase and G2/M phase, about 6.6% and 23.2%, respectively. Compared to the untreated control, complex 3 and 6 disturbed the cell cycle at G1 phase, and S and G2/M phase, respectively.

Fig. 6
figure 6

a Cell cycle arrest towards A549 cancer cells after exposure to complexes 3 and 6 for 24 h at 298 K, determined by flow cytometry with PI/RNase staining. Concentrations used: 0.25 and 0.5 × IC50. b Cell populations in different cell cycle phases

Full size image

ROS induction

Reactive oxygen species (ROS) play important roles as regulating cell proliferation, death and signal transduction [58, 59]. The ROS level in A549 cells induced by complexes 3 and 6 (concentrations used: 0.25 × IC50) was monitored by flow cytometry (Fig. 7; Table S4). Compared to untreated cells, a dramatic increase in the ROS levels was observed for both complexes even at low concentration (0.25 × IC50) after 24 h of drug exposure. Complex 3 caused even higher ROS levels in cells than complex 6, causing 93.4% of A549 cells in high ROS levels at concentration of 0.25 ×IC50. The higher ability to generate ROS in cells of complex 3 is correlated to its more potent anticancer activity in comparison with complex 6. The generation of ROS by the complexes may provide a basis for antiproliferative activity.

Fig. 7
figure 7

a Flow cytometry determined ROS generation in A549 cells treated by complexes 3 and 6 at concentrations of 0.25 × IC50. b Cell populations in high or low ROS levels

Full size image

Conclusion

In this work, six IrIII half-sandwich complexes [(η5-Cpx)Ir(N^N)Cl]PF6 containing two types of iminopyridine chelating ligand were designed and prepared. Four complexes exhibited superior anticancer potency toward A549 cancer cells than the clinical used cisplatin. The most potent complex 3 is 8 × more active than cisplatin against A549 cells. The anticancer activity can be tuned by varying both Cp*-based ligands and iminopyridine ligands, in the order of Cpxbiph > Cpxph > Cp*.

Basically, both complexes 3 and 6 containing different iminopyridine ligands did not show hydrolysis in aqueous condition; however, the chelating ligand played a very important role in nucleobase binding selectivity and in the reaction with glutathione. Complex 6 showed strong binding to the nucleobases, preferring 9-EtG than 9-MeA. These results suggest that DNA could be a potential target, although other targets appear to be also important because complex 3 containing adamantyl iminopyridine is a potent anticancer agent even no nucleobase binding. Complex 3 is much more stable than complex 6 in the presence of GSH, and is more hydrophobic than complex 6, which appears contribute to its higher anticancer activity. This type of complexes are effective catalysts in converting coenzyme NADH to NAD+ with formation of an Ir hydride adduct, thus possible affect the redox environment in cells. Further mechanistic studies showed that complexes 3 and 6 induced obvious apoptosis, arrested cell cycles at G1, S, and G2/M phase, respectively, and significantly increased the ROS levels in A549 cancer cells. Our study demonstrates that these iridium iminopyridine complexes could be potential anticancer agents.