Biotinylated photoactive Pt( IV ) anticancer complexes †

Novel biotinylated diazido-Pt( IV ) complexes exhibit high visible light photocytotoxicity while being stable in the dark. Photocytotoxicity and cellular accumulation of all- trans -[Pt(py) 2 (N 3 ) 2 (biotin)(OH)] (2a) were enhanced significantly when bound to avidin; irradiation induced dramatic cellular morphological changes in human ovarian cancer cells treated with 2a.

The synthetic routes for photoactive Pt(IV) complexes 2a and 2b are summarised in Scheme 1. Mono-substituted 2a was obtained by combining 1 with biotin using TBTU/DIPEA as coupling agents, while the bi-substituted 2b was synthesised by stirring 2a with DCA anhydride. Both complexes gave satisfactory elemental analyses and were also characterised by ESI-HRMS, NMR and UV-vis spectroscopy, and for purity by HPLC (Fig. S1, ESI †).  Complexes 2a and 2b exhibited excellent dark stability in RPMI-1640/DMSO (95%/5%, v/v) monitored by UV-vis spectroscopy ( Fig. 1a and Fig. S6a, ESI †). When irradiated with blue light (420 nm), the LMCT bands of both complexes at ca. 300 nm decreased in intensity, indicating release of the azide ligands ( Fig. 1b and Fig. S6b, ESI †). 27 Notably, 2b exhibited a larger decrease at the absorption maximum, probably due to release of the DCA ligand. Released azidyl radicals and singlet oxygen were trapped by DMPO and TEMP, respectively, for 2a upon irradiation (463 nm; Fig. 1c and d).
The photoreactions of 2a/2b with 5 0 -GMP were analysed by LC-MS. Pt(IV) complexes (30 mM) were incubated with 5 0 -GMP (2 mol equiv.) at 310 K in the dark for 1 h, then irradiated by blue light (420 nm) for 1 h (Fig. S7, ESI †). For both complexes, [Pt II (CH 3 CN)(py) 2 (GMP-H)] + (G1, 756.48 m/z) was detected as the major Pt-GMP product, together with two minor adducts [Pt II (HCOO)(py) 2 (GMP)] + (G2, 762.18 m/z) and [Pt II (N 3 )(py) 2 (GMP)] + (G3, 758.23 m/z). These results were similar to those obtained for 1, suggesting a negligible effect of axial substituents on the photoreaction with guanine. 38 DNA melting experiments monitored by UV-vis were carried out in phosphate buffer (1 mM, pH = 7.9) to investigate DNA binding with a drug/base pair ratio of 0.2 (Table S1, ESI †). Similar to parent complex 1, 2a and 2b interact with ct-DNA only weakly (DT m o 2 K) in the dark, while upon irradiation with blue light (420 nm) an apparent interaction was observed (DT m ca. 5 K). The increase in the DNA melting temperature suggests that 2a and 2b might form inter-strand cross-links after irradiation, similar to 1.
The binding affinity of avidin towards 2a and 2b in PBS was assessed by a displacement titration using HABA (2-(4-hydroxyphenylazo)benzoic acid) and monitored by UV-vis spectroscopy (Fig. 2). The avidin-HABA adduct displays an absorption band at ca. 496 nm, which decreases when HABA is replaced by biotin. 39 On mixing, a gradual decrease of the absorbance at 496 nm was observed upon addition of either complex to a solution containing HABA (160 mM) and avidin (8 mM), indicating their stronger binding to avidin compared with HABA (dissociation constant K d = 10 À6 M). 39 A sharp end point and same absorbance changes (DA = 1.01) with biotin were observed for 2a and 2b, indicating their similar affinity towards avidin as unmodified biotin. These results suggested the possibility of using avidin as a nanocarrier for delivery of biotinylated complexes to cancer cells.
The IC 50 values of 2a and 2b in human A2780 ovarian, A549 lung and PC3 prostate cancer cell lines were determined by the sulforhodamine B (SRB) colorimetric assay (Table 1), using parent complex 1 and CDDP (cisplatin) as references. Both complexes were relatively non-toxic towards all cancer cells and healthy MRC-5 lung fibroblasts in the dark with IC 50 values 450/100 mM, but exhibited promising photocytotoxicity with high photocytotoxicity indices (PI). In all the cancer cell lines studied, di-substituted 2b (1.3-5.9 mM) is at least twice as phototoxic as the mono-substituted 2a (11.7-21.1 mM), and 5Â more toxic than 1 (7.1-55.6 mM). Under the same conditions, cisplatin exhibited very low cytotoxicity in all cell lines (IC 50 4 100 mM) due to the short incubation time (2 h).
In A2780 ovarian cancer cells, 2a (IC 50 = 11.7 mM) was slightly less active than 1 (IC 50 = 7.1 mM), while 2b was more potent (IC 50 = 1.3 mM), potentially attributable to the synergistic anticancer activity of the released DCA ligand. In order to investigate the effect of avidin on activity, 2a was mixed with avidin (2a : avidin = 4 : 1) prior to addition to A2780 cells following the same protocol used for 2a alone. Notably, the avidin-2a complex exhibited good dark stability, being potently photocytotoxic in A2780 cells with an IC 50 value of 4.4 mM, 2.7Â lower than that of 2a, and 1.6Â more toxic than 1. In contrast, the photocytotoxicity of 2b was not enhanced in the presence of avidin.  Since biotin is expected to exhibit a preference towards cancer cells due to overexpressed receptors, 5 ICP-MS was used to quantify and compare the cellular uptake of biotinylated complexes 2a and 2b as well as the unsubstituted 1. A2780 ovarian cancer cells were treated with 10 mM prodrugs in the dark for 1 h, then the Pt content of cells was analysed by ICP-MS ( Table 2). The mono-substituted 2a exhibited lower Pt accumulation (0.4 ng per 10 6 cells) than 1 (0.64 ng per 10 6 cells) after 1 h incubation, while accumulation of di-substituted 2b (21.1 ng per 10 6 cells) was 33Â higher than that of 1 and 53Â higher than 2a. These results showed that conjugation with biotin alone did not increase the accumulation of diazido Pt(IV) complexes in A2780 cells, while the substitution of the second axial ligand with DCA resulted in significant increase in Pt cellular accumulation, perhaps due to higher lipophilicity of 2b, consistent with its longer HPLC retention time (12.3 min for 2a, 19.7 min for 2b, Fig. S1, ESI †). Notably, the amount of Pt taken up by A2780 cells incubated with 1, 2a and 2b was inversely proportional to their IC 50 values. These results suggested that cellular accumulation of Pt(IV) prodrugs plays an important role in their antiproliferative activity. Since the avidin-2a complex exhibited improved photocytotoxicity, the effect of avidin on the uptake of 2a was also investigated.
The accumulation of Pt after exposure of A2780 cells to the mixture of 10 mM 2a and avidin (4 : 1) was 10Â higher than 2a alone, which correlates with the higher photocytotoxicity of 2a-avidin complex. However, Pt accumulation for 1 and 2b was not affected by avidin, since 1 does not contain biotin and 2b is mainly taken up by passive diffusion due to its high lipophilicity.
A2780 ovarian cancer cells were treated with 2a (1 or 2Â photo IC 50 concentration) in the presence and absence of light and live-imaged using confocal microscopy and flow cytometry to investigate changes in cell morphology (Fig. 3 and Fig. S8, S10, ESI †). The cell permeant dye SYTOt 17 was used to stain the nuclei. Without irradiation, 2a exhibited low cytotoxicity (IC 50 4 100 mM) and, accordingly, no changes in cellular morphology were observed when cells were treated with 2a for 2 h in the dark. In these conditions, the cells appeared healthy with welldefined plasma membranes and intact nuclei. In contrast, A2780 cells treated with 2a exhibited dramatic morphological changes after 1 h irradiation with blue light (465 nm). The cells rounded up and the nuclei were fragmented into pieces. Damaged membranes and copious cell debris were observed when cells were treated with 2a at 2 Â IC 50 concentration ( Fig. 3 and Fig. S10, ESI †). DNA is usually regarded as the major target of platinum anticancer drugs, so the ability of 2a to fragment cell nuclei exclusively upon irradiation, indicated its potential as a photoactive prodrug with a novel mechanism of action. Complex 2a at higher concentration induced more morphological changes upon irradiation, while irradiation alone did not result in significant effects on cell morphology ( Fig. S9 and S10, ESI †).
Cellular ROS generation for 2a, 2b and 1 was monitored in A549 lung cancer cells by DCFH-DA, which exhibits switch-on fluorescence in the presence of ROS. Cells treated with complexes (10 mM) in the dark showed no apparent change, but 2b induced an increased DCF fluorescence upon irradiation, indicating its ability to generate ROS (Table S2, ESI †).

Conflicts of interest
There are no conflicts to declare.