1. Introduction

Oxygen (O₂) within the cell plays a crucial role in regulating many physiological and pathological processes, and the abnormal changes of intracellular O₂ (icO₂) levels are usually related to diseases, such as cardiac ischemia [1,2], inflammatory diseases [3], and solid tumors [4,5]. Especially, hypoxia, which is defined as oxygen-deprived condition, is a characteristic feature of the tumor microenvironment in most solid tumors due to the limited oxygen supply from blood vessels [6]. In some hypoxic regions, the icO₂ levels can be as low as 0% [7]. Hence, it is imperative to develop an effective technology for monitoring the icO₂ levels, which can help to understand the tumor microenvironment and guide the subsequent treatments.

For many years, the detection of hypoxia has relied on some instrument techniques, including Positron Emission Tomography (PET) and functional Magnetic Resonance Imaging (MRI) [8,9]. These methods, however, have considerable limitations, such as inaccurate diagnosis and radioactive risk. In recent years, optical imaging has been demonstrated to be an effective approach for hypoxia imaging because of its significant advantages including high sensitivity and simplicity [10,11]. Among a variety of luminescent probes available for hypoxia imaging, phosphorescent transition-metal complexes (PTMCs) have unique advantages owing to their reversible phosphorescent intensity and lifetime changes sensitive to icO₂ levels, because of a diffusion-controlled collisional interaction and energy transfer between the triplet excited state of PTMCs and the triplet ground state of oxygen [10,12–20]. In addition, PTMCs exhibit some advantageous photophysical properties for biosensing and bioimaging, such as large Stokes shift, high photostability and long emission lifetime [21–24]. Especially, the long emission lifetime of PTMCs is very suitable for time-resolved luminescence imaging techniques, including photoluminescence lifetime imaging (PLIM) and time-gated luminescence imaging technology (TGLI), which can effectively eliminate the short-lived background fluorescence interference and improve the signal-to-noise ratios in the biosensing and bioimaging [24–29]. However, the examples of PTMCs for time-resolved luminescence imaging of hypoxia are very few.

In this paper, we designed and synthesized a phosphorescent dinuclear iridium(III) complex Ir1 (Fig. 1(a)) with 2-(2,4-difluorophenyl)pyridine as C^N ligand and tetrapyrido[3,2-a:2’,3′-c:3”,2”-h:2”’,3”’-j]phenazine (tpphz) as N^N ligand. Both the phosphorescent intensity and lifetime of Ir1 exhibit high sensitivity to oxygen levels. Considering its long emission lifetime, this complex has been used for time-resolved luminescence imaging of intracellular O₂ levels.

 

Fig. 1 (a) Chemical structures of complex Ir1 and Ir2. (b) UV-vis absorption and normalized phosphorescence spectra of Ir1 and Ir2 in dichloromethane at 293 K (concentration = 10 μM, excitation wavelength = 365 nm). (c and d) Phosphorescence spectra and decays of Ir1 under different O₂ concentrations.

Download Full Size | PDF

2. General experimental information

¹H NMR spectra were recorded with a Varian spectrometer at 400 MHz. Mass spectra were obtained on SHIMADZU matrix-assisted laser desorption/ionization time of flight mass spectrometer (MALDI-TOF-MASS). The UV-vis absorption spectra were recorded on a Shimadzu UV-2550 spectrometer. Photoluminescent spectra were measured on an Edinburgh LFS-920 spectrometer. Emission lifetime studies were performed on the Edinburgh LFS-920 spectrometer with a hydrogen-filled excitation source. The data were analyzed by a software package provided by Edinburgh Instruments. The absolute quantum yields of the complexes were determined through an absolute method by employing an integrating sphere.

Luminescence imaging was performed on an Olympus FV1000 laser scanning confocal microscope with a 40 × objective lens. Cells incubated with the iridium(III) complexes were excited at 405 nm with a semiconductor laser, and the emission was measured according to the spectral data. The images were accomplished using the software package provided by Olympus instruments. Photoluminescence lifetime imaging and time-gated luminescence imaging were measured on the platform afforded by Olympus FV1000 laser scanning confocal microscope and PicoQuant Company. The objective lens was 40 × and the laser repetition rate was 0.5 MHz for Ir1. The imaging time was within two minutes. The correlative calculations of the data were carried out with the software provided by PicoQuant.

Characterization data of Ir1. ¹H NMR (400 MHz, DMSO) δ 10.25 (dt, J = 8.4, 1.2 Hz, 4H), 8.51 – 8.43 (m, 4H), 8.31 (m, 8H), 7.99 (dd, J = 7.6, 1.2 Hz, 4H), 7.71 (d, J = 6.0 Hz, 4H), 7.15 – 6.96 (m, 8H), 5.74 – 5.71 (m, 4H). MALDI-TOF (M-2PF₆^-): 1674.911 (1674.465).

Characterization data of Ir2. ¹H NMR (400 MHz, DMSO) δ 9.72 (d, J = 8.0 Hz, 2H), 9.65 (d, J = 8.0 Hz, 2H), 8.69 (s, 2H), 8.46 (dd, J = 5.2, 1.2 Hz, 2H), 8.33 (d, J = 8.8 Hz, 2H), 8.19 (dd, J = 8.4, 5.2 Hz, 2H), 8.03-7.96 (m, 2H), 7.92 (d, J = 5.6 Hz, 2H), 7.87 (dd, J = 8.0, 4.0 Hz, 2H) 7.19 – 6.99 (m, 4H), 5.76 (dd, J = 8.4, 2.4 Hz, 2H). MALDI-TOF (M-PF₆^-): 956.947 (957.915).

3. Results and discussion

The N^N ligand tpphz and the chloro-bridged dinuclear cyclometalated iridium(III) precursor were synthesized according to the previous report [30]. For comparison, a mononuclear iridium(III) complex Ir2 (Fig. 1(a)) was also designed and synthesized. Both the complex Ir1 and Ir2 were synthesized according to the previous procedure [30]. The chemical structures of Ir1 and Ir2 were characterized by MALDI-TOF MS and ¹H NMR spectra.

The absorption and emission spectra of the complex Ir1 and Ir2 have been measured. Both of two complexes exhibited strong absorption below 400 nm with log[ε] > 4.0, which can be attributed to the spin-allowed singlet π-π* transition of the C^N and N^N ligands (Fig. 1(b)). Interestingly, dinuclear complex Ir1 exhibits the stronger light-absorption capability than mononuclear complex Ir2, and the molar extinction coefficient (ε) of Ir1 was about twice of that of Ir2. Moreover, intense red emission at 626 nm was observed for Ir1 upon excitation at 365 nm, and Ir2 exhibited emission maximum at 540 nm. The significant emission red-shift of Ir1 compared with Ir2 was attributed to the stabilization of the π* orbital of the N^N ligand by the two iridium centers. In addition, dinuclear complex Ir1 showed the higher quantum yields of 0.88 than mononuclear Ir1 (0.61). The emission lifetime of Ir1 (1076.8 ns) in degassed solution is also significantly longer than that of Ir2 (591.6 ns).

The long emission lifetime of Ir1 made it very sensitive to oxygen concentrations. Hence, the oxygen sensing performance of Ir1 has been investigated. As shown in Fig. 1(c), the emission intensity is very sensitive to oxygen, and it decreased significantly with the increase of oxygen concentration. Simultaneously, the emission lifetime of Ir1 also reduced from 1076.8 ns at 0% O₂ concentration to 711.6 ns at 21% O₂ concentration and 278.8 ns at 100% O₂ concentration (Fig. 1(d)). The emission lifetime of phosphorescence is much longer than that of autofluorescence (several nanoseconds). Thus, the phosphorescence signal and autofluorescence could be distinguished effectively via time-resolved luminescence technology.

For biological applications, the cytotoxicity of Ir1 was investigated. In vitro cell viability of Hela cells incubated with the complex under different concentration for 24 h and 48 h was measured by using methyl thiazolyltetrazolium (MTT) assay [31]. The results showed that Ir1 had a low toxicity even under a high concentration of 100 µM and incubation for 48 h. When incubated with the complex at the imaging concentration of 10 µM, the viability of Hela cells was higher than 90%, indicating the low cytotoxicity of Ir1. In addition, further cytotoxicity assessments (Fig. 2) by LDH release assay [32] and JC-1 mitochondrial membrane potential assay [33] also suggest that complex Ir1 shows low effect on cell growth and proliferation, implying the good biocompatibility for biological application.

 

Fig. 2 (a) Cytotoxic evaluation of complex Ir1 by LDH release. LDH release tests of Hela cells after treatment with complex Ir1 (5 μM, 10 μM, 50 μM, 100 μM, 200 μM) for 24 and 48 h. The release of LDH molecules was triggered upon disruption of the cell membrane. It was found that intracellular LDH concentrations of the cells treated with complex Ir1 showed little difference compared with the untreated control group. This indicates low adverse effects of complex Ir1 on cell membrane integrity within a time range of 48 h. (b) Cytotoxic evaluation of complex Ir1 by JC-1 staining. Hela cells were incubated with or without complex Ir1 (5, 10, 50, 100 μM) in culture dish for 48 h and then incubated with JC-1 for 30 min. For positive control experiment, Hela cells were incubated with CCCP (10 μM) for 30 min and then incubated with JC-1 for 30 min. JC-1 aggregates in healthy mitochondria and has a red fluorescence, whereas green fluorescence represents the monomeric form of JC-1, indicating dissipation of mitochondrial membrane potential (ΔΨ_m). CCCP (10 μM) treatment shows obvious fluorescence emission shift in Hela cells, while complex Ir1 (5 to 100 μM) treatment did not show obvious change.

Download Full Size | PDF

Next, the cell imaging experiments of Ir1 were carried out with confocal microscopy to investigate its cellular uptake. The excitation wavelength used was 405 nm, and the emission measured was at 580–680 nm. Figure 3 showed that Ir1 was cell permeable and mainly distributed in the cytoplasm. In order to investigate the specific staining of Ir1 to the cytoplasm, the co-staining experiment was taken with the commercial stain CMAC. Live Hela cells were incubated with the solution of Ir1 first for 1 h at 37 °C. After that, the cells were further incubated with CMAC for 10 min at room temperature. The emission of CMAC and Ir1 was collected at 420–490 nm and 580–680 nm, respectively. As shown in Figs. 3(a) and 3(e), the commercial stain CMAC distributed in the whole cells but Ir1 only stained the cytoplasm (Fig. 3(f)). The luminescence intensity profile of Ir1 (red) and CMAC (green) in Fig. 3(h) also indicated that Ir1 specifically stained the cytoplasm.

 

Fig. 3 (a) The confocal imaging of live Hela cells incubated with Ir1 for 1 h at 37 °C and then further incubated with dye CMAC for 10 min at room temperature. (b) The photoluminescence lifetime imaging of Hela cells incubated with Ir1 and CMAC. (c and d) The time-gated luminescence imaging of Hela cells incubated with Ir1 and CMAC with the delay time of 0 ns and 50 ns. (e-h) Luminescence intensity profile of Ir1 (red) and CMAC (green) across the line corresponding to cytoplasm region (1 and 3) and nuclear region (2). The concentrations of Ir1 and CMAC in incubation solution were 10 μM and 1 μM, respectively. Excitation wavelength was 405 nm.

Download Full Size | PDF

CMAC was a fluorescent organic dye with the disadvantage of photobleaching, while Ir1 was a phosphorescent dye with good photostability. Next, comparison of the photostability for Ir1 and CMAC was investiguated. Hela cells were first incubated with Ir1 and CMAC, and then the photobleaching was taken under the imaging condition. The excitation wavelength was 405 nm. The average intensity of the five circle region were recorded and analyzed after different photobleaching time. Figure 4 showed that the average intensity of CMAC in cytoplasm decreased more sharply than that of Ir1. This indicated that Ir1 exhibits higher photostability than CMAC.

 

Fig. 4 (a-c) The cell imaging of Hela cells stained with Ir1 and CMAC before photobleaching. (d-f) The cell imaging of Hela cells stained with Ir1 and CMAC after photobleaching for 220 s. (g) The average intensity of the chosen region at different photobleaching time. The concentrations of Ir1 and CMAC in incubation solution were 10 μM and 1 μM, respectively. Excitation wavelength was 405 nm. The scale bars were 20 μm.

Download Full Size | PDF

Considering the long emission lifetime of Ir1, its application in time-resolved luminescence imaging has been investigated in detail. Hela cells were incubated with Ir1 and CMAC, respectively. Herein, CMAC can act as the background interference because it exhibits short emisison lifetime (2.5 ns) close to that of autofluorescence from many biomolecules. From the confocal luminescence intensity images (Figs. 3(a) and 3(c)), we can see that it was difficult to isolate the signal of Ir1 and that of CMAC in cytoplasm. From PLIM images shown in Fig. 3(b) we can see that the long-lived phosphorescent signal from Ir1 could be distinguishable from the interference signals from CMAC because of the signifcant differences in their emission lifetimes. In addition, TGLI result (Fig. 3(d)) showed that the interference of short-lived CMAC was eliminated effectively after applying a delay time of 50 ns and the imaging with higher SNR (signal to noise ratio) was obtained. These results confirmed that with the long lifetime, Ir1 would be an excellent cytoplasm probe by using time-resolved luminescence imaging.

Considering that the phosphorescence of Ir1 was senstive to oxygen, we next carried out a series of experiments to investigate the sensitivity of Ir1 to icO₂ levels. First, Hela cells were incubated with Ir1 in two culture dishes for 30 minutes. After that, one dish was transferred into a hypoxic atmosphere of 2.5% oxygen level and then incubated for another 1 hour before luminescence cell imaging experiments. Confocal luminescence images showed that the phosphorescence of cells was enhanced evidently when the icO₂ levels reduced from 21% to 2.5% (Figs. 5(a) and 5(b)). Furthermore, the change of icO₂ levels can be monitored by PLIM. Figures 5(c), 5(d) and 5(g) showed that the emission lifetime of Ir1 in the cytoplasm was much longer under 2.5% oxygen level (771 ns) than that under 21% oxygen level (402 ns). Besides, the TGLI of Ir1 in the cytoplasm also showed that with a delay time of 450 ns, the emission signal of Ir1 in the cytoplasm was more obvious under 2.5% oxygen level than that under 21% oxygen level (Figs. 5(e) and 5(f)), achieving the monitoring of change in icO₂ levels via TGLI technique.

 

Fig. 5 Photoluminescence intensity (a and b) and lifetime (c and d) images of Ir1 in the cytoplasm under 21% and 2.5% oxygen level. Time-gated luminescence intensity images (e and f) of Ir1 in the cytoplasm under 21% and 2.5% oxygen level with 450 ns delay. (g) The representative decays corresponding to the PLIM images in cell experiments. The concentrations of Ir1 in incubation solution was 10 μM. Excitation wavelength was 405 nm.

Download Full Size | PDF

4. Conclusion

We have designed and synthesized a phosphorescent dinuclear iridium(III) complex Ir1, which exhibited excellent oxygen-sensing performance. Compared with the commercial dye CMAC, Ir1 had a better photostability, which is very beneficial for biological application. The long phosphorescent lifetime of Ir1 made it a good candidate for time-resolved luminescence imaging application. Thus, monitoring of intracellular oxygen levels has been realized successfully through photoluminescence lifetime imaging and time-gated luminescence imaging techniques, both of which can effectively eliminate the short-lived background fluorescence interference and improve the signal-to-noise ratios in bioimaging. Hence, rationally designed long-lived phosphorescent transition-metal complexes as a new generation of optical probes will be very promising for application in fields of medical optics and biotechnology.