Gold nanoclusters as a GSH activated mitochondrial targeting photosensitizer for efficient treatment of malignant tumors

Abstract

Gold nanoclusters (Au NCs), which have the characteristics of small size, near infrared (NIR) absorption and long triplet excited lifetime, have been used as a new type of photosensitizer for deep tissue photodynamic therapy (PDT). However, the therapeutic efficiency of the nano-system based on Au NCs still needs to be improved. Herein, we proposed a strategy using Mito-Au₂₅@MnO₂ nanocomposites to achieve enhanced PDT. Au₂₅(Capt)₁₈⁻ nanoclusters were applied as photosensitizers and further modified with peptides to target mitochondrial and MnO₂ nanosheets to consume glutathione (GSH). In the presence of GSH, Mito-Au₂₅@MnO₂ dis-integrated and Mito-Au₂₅ nanoparticles realized accurate mitochondrial targeting. Under the irradiation of 808 nm light, the nanocomposite ensured highly efficient PDT both in vitro and in vivo via oxidation pressure elevation and mitochondrial targeting in cancer cells.

This is the first example of mitochondrial targeting Au NCs capable of improving the efficiency of photodynamic therapy. Mito-Au₂₅@MnO₂ can be activated by consuming GSH and elevating oxidation pressure in cancer cells.

Introduction

Gold nanoclusters (Au NCs), consisting of a few to hundreds of gold atoms, exhibit discrete energy levels as their size (∼2 nm) approaches the Fermi-wavelength of electrons.^1,2 Due to the unique molecule-like properties such as optical absorption and photoluminescence, Au NCs were widely used in biosensing,³ imaging^4,5 and detection.⁶ Furthermore, the photophysical property of Au NCs could be regulated by ligands to achieve strong near infrared (over 800 nm) absorption⁷ and long lifetime of excited triplet state.⁸ As we know, near infrared (NIR) excitation could reduce light scattering and absorption of endogenous molecules when interact with biological tissues, thus it's promising in tissue penetration and biological safety.⁹ Consequently, Au NCs have been used as a new type of photosensitizers (PSs) for photodynamic therapy (PDT) which impart cytotoxicity via exciting photosensitizers in appropriate wavelengths of light.¹⁰ More importantly, compared with PSs loaded on upconversion nanoparticles,¹¹ Au NCs improve the utilization efficiency of NIR excitation directly, without considering loading capacity and energy transfer efficiency.

In the meanwhile, the recent research suggests that MnO₂ nanosheets cooperating with PSs can elevate cellular uptake of PSs and upgrade oxidation pressure of cancer cells.¹² On the one hand, high concentration of glutathione (2–10 mM) alleviates the PDT effect via consuming reactive oxygen species (ROS) generated by PSs.^13,14 MnO₂ nanosheets can be reduced to Mn²⁺ by GSH, which will deplete GSH and strengthen oxidation pressure in cancer cells.¹² On the other hand, Au NCs with small size are prone to accumulate at the liver and spleen instead of tumor tissue.¹⁵ MnO₂ nanosheets are suitable for enlarging the size of nanomaterials (up to 100–200 nm), which will improve the passive targeting ability of nanomaterials by enhanced permeability and retention effect (EPR). Unfortunately, there are still two main disadvantages of the assembled Au NCs and MnO₂ nanosheets. Firstly, complex charge conversion operation is needed before the negatively charged MnO₂ nanosheets are assembled with Au NCs. Secondly, this kind of nanocomposites still lack of mitochondrial targeting ability which is of significant importance to elevate the PDT efficiency.¹⁶ Due to the short lifetime (<200 ns) and limited movement distance (∼20 nm), ROS are difficult to reach the core area of cancer cells and can't provide efficient tumor treatment capabilities.¹⁷ Mitochondria play an important role in providing energy through oxidation of carbohydrates, fat and amino acids.¹⁸ Besides, they are extremely fragile organelles in eukaryocyte cells because single-stranded mito-DNA is easier to be destroyed and thus accelerating the necrosis and apoptosis of cancer cells.¹⁹ Therefore, nanocomposites which can target mitochondria will make full use of ROS and enhance PDT effect. Hence, constructing a new nano-system which can not only realize mitochondrial targeting but also facilitate direct electrostatic interaction with MnO₂ nanosheets is conducive to elevate PDT efficiency.

Herein, we utilized Au₂₅(Capt)₁₈⁻ nanoclusters, which have good chemical stability, anti-photobleaching ability and long excitation wavelength at NIR I region as PSs. Subsequently, mitochondrial targeting peptides (Fr-(Cha)-K-NH₂) were connected onto the surface of Au₂₅(Capt)₁₈⁻ nanoclusters to obtain Mito-Au₂₅ nanoparticles. Finally, the Mito-Au₂₅@MnO₂ nanocomposites were obtained through electrostatic interaction between Mito-Au₂₅ nanoparticles and MnO₂ nanosheets. As illustrated in Fig. 1, the Mito-Au₂₅@MnO₂ nanocomposite was an inactive photosensitizer before MnO₂ nanosheets were consumed by GSH. After cellular uptake, MnO₂ depleted GSH and the oxidation pressure increased in cancer cells. In the meanwhile, the nanocomposites disintegrated and Mito-Au₂₅ was released to achieve specific targeting of mitochondrial, thus elevating the utilization efficiency of ROS. Under the irradiation of 808 nm laser, we successfully realized high-efficient, deep seated and reliable PDT on HepG2 cells and 4T1 tumor bearing mice. Compared with traditional methods,^20,21 our study highlights the significance of subcellular targeting ability and GSH consumption efficiency.

Fig. 1. Schematic diagram of assembling process and the mechanism of Mito-Au₂₅@MnO₂ for in vivo PDT.

Results and discussion

Characterization and photophysical properties of Au₂₅(Capt)₁₈⁻ nanoclusters

The Au₂₅(Capt)₁₈⁻ nanoclusters (abbreviated as Au₂₅NCs) were prepared using a NaBH₄ reduction method at room temperature.²² As shown in Fig. 2a, the obtained Au₂₅NCs had good dispersity and uniformed size. The average hydrodynamic diameter of Au₂₅NCs was 2.515 ± 0.546 nm (Fig. 2a). Characteristic UV-Vis absorption spectra displayed at 670 nm, 450 nm and 400 nm (Fig. 2b) were in good accord with former research.²³ With the excitation at 561 nm, the maximum fluorescence emission wavelength was at 740 nm (Fig. S1†). X-ray photoelectron spectroscopy (XPS) was employed to demonstrate the elemental composition and valence state of Au₂₅NCs. Binding energy of Au 4f_5/2 and Au 4f_7/2 were 87.6 eV and 84.1 eV, respectively (Fig. S2a†). The orbital binding energy of Au 4f_7/2 (84.1 eV) was between the binding energy of Au(0) (83.8 eV) and Au(i) (86.0 eV), which confirmed that Au(0) and Au(i) both existed in Au₂₅NCs.²⁴ These results were consistent with former core–shell structure theory.²⁵ In addition, the binding energy of S 2p was 163.4 eV different from 168 eV (SO₃⁻), which confirmed the formation of Au–S bound (Fig. S2b†).^26,27 Therefore, we successfully synthesized Au₂₅NCs.

Fig. 2. (a) TEM image and size distribution of Au₂₅NCs. (b) UV-Vis absorption spectra of Au₂₅NCs. (c) Singlet oxygen generation of Au₂₅NCs. Time dependent absorption changes at 412 nm of DMF solutions of DPBF (200 μM) and Au₂₅NCs (0 or 200 μg mL⁻¹) mixture upon irradiation at 808 nm light. (d) Singlet oxygen generation efficiency liner fitting of Au₂₅NCs and indocyanine green (ICG).

Singlet oxygen (¹O₂) generation ability of Au₂₅NCs was monitored using singlet oxygen probe DPBF.²⁸ The UV-Vis spectrum of DPBF and Au₂₅NCs mixture solution was recorded after each 30 s of irradiation with 808 nm light. As shown in Fig. 2c, with increasing irradiation time, the absorption value at 412 nm decreased gradually, indicating the formation of ¹O₂. To our delight, the ¹O₂ quantum yield of Au₂₅NCs was 9.12 times as much as indocyanine green (ICG), suggesting that Au₂₅NCs could be used as excellent PSs to conduct PDT (Fig. 2d).

Characterization of MnO₂ nanosheets

Two dimensional MnO₂ nanosheets were prepared by ultrasonicating of bulk MnO₂.²⁹ The ultrathin MnO₂ nanosheets had a broad optical absorption spectrum (250–700 nm) and its maximum absorption wavelength was 371 nm (Fig. S3†). Transmission electron microscopy (TEM) and field emission scanning electron microscope (FESEM) images showed that MnO₂ nanosheets dispersed uniformly with sheet morphology (Fig. S4a, c and d†). The average hydrodynamic diameter of MnO₂ nanosheets was 187.5 ± 71.36 nm (Fig. S4b†), and Fig. 3a indicated that these nanosheets were 1.2–1.6 nm in height, which were nearly two layers of MnO₂ nanosheet (the theoretical thickness of the single-layered MnO₂ nanosheet is 0.69–0.72 nm (ref. 30)). XPS graph showed that there were Mn 2p and O 1s in the spectrum (Fig. S5a†). It could be found that the binding energy of Mn 2p_3/2 and Mn 2p_1/2 were 641.3 eV and 653.2 eV, respectively (Fig. S5b†), with a spin-energy separation of 11.9 eV, which confirmed the formation of Mn(iv)O₂ nanosheets.³¹ The reaction between MnO₂ and GSH was tested. With the increasing concentration of GSH, the absorption of MnO₂ at 371 nm decreased gradually (Fig. S6†), which indicated the reaction of MnO₂ and GSH was concentration dependent.

Fig. 3. (a) Atomic force microscopy (AFM) image and corresponding height profile of MnO₂ nanosheets. Scale bar: 250 nm. (b) UV-Vis absorption spectra of mitochondrial targeting peptide, Au₂₅NCs and Mito-Au₂₅. (c) Zeta potential of MnO₂, Au₂₅NCs, Mito-Au₂₅ and Mito-Au₂₅@MnO₂. (d) GSH (0 or 200 μM) dominated activable ¹O₂ generation of Mito-Au₂₅@MnO₂. Time dependent absorption changes at 412 nm of DPBF (200 μM) and Mito-Au₂₅@MnO₂ (200 μg mL⁻¹) mixture solution upon irradiation at 808 nm.

Characterization and photophysical properties of Mito-Au₂₅ nanoparticles and Mito-Au₂₅@MnO₂ nanocomposites

The Mito-Au₂₅ nanoparticles were synthesized through biological coupling reaction between mitochondrial targeting peptide and MnO₂ nanosheets. Compared with Au₂₅NCs, the size of these nanoparticles increased obviously (Fig. S7†), which was probably due to charge inversion and electrostatic interaction between Mito-Au₂₅ and Au₂₅NCs. Furthermore, a new UV-Vis absorption peak emerged at 276 nm after mitochondrial targeting peptide was modified onto the surface of Au₂₅NCs, which was a little red shift referring to the characteristic absorption at 260 nm of mitochondrial targeting peptide (Fig. 3b). Fourier-transform infrared spectroscopy (FTIR) further proved the change of functional group at 3424 cm⁻¹ (Fig. S8†). We monitored the ¹O₂ generation ability of Mito-Au₂₅ to test whether the modification of peptide would influence the photophysical property of Au₂₅NCs. As shown in Fig. S9a,† the absorption at 412 nm of DPBF and Mito-Au₂₅ mixture solution decreased via a time dependent mode indicating the formation of ¹O₂. Three parallel experiments were conducted and confirmed the efficient ¹O₂ generation of Mito-Au₂₅ nanoparticles (Fig. S9b†).

Zeta potential of Au₂₅NCs, Mito-Au₂₅ and MnO₂ nanosheets were measured before assembling (Fig. 3c). Au₂₅NCs were negatively charged (−41.9 mV). After modification of mitochondrial targeting peptide, Mito-Au₂₅ showed the average zeta potential of 17.2 mV. The average zeta potential of MnO₂ nanosheets was −30.4 mV, which presented an ideal charge state for electrostatic assembling. Finally, Mito-Au₂₅@MnO₂ nanocomposites were obtained through electrostatic interaction between positively charged Mito-Au₂₅ nanoparticles and negatively charged MnO₂ nanosheets. The average zeta potential of nanocomposites was 14.1 mV. UV-Vis absorption spectrum in Fig. S10† showed no characteristic absorption peak of Au₂₅ NCs in supernatant after centrifugation procedure, which meant that 2.5 mg Mito-Au₂₅ was completely assembled with 0.08 mg MnO₂ nanosheets.

TEM was utilized to observe the morphology of Mito-Au₂₅@MnO₂ nanocomposites. As shown in Fig. S11a,† the nanocomposites presented good dispersity and no evident aggregation occurred. But the Mito-Au₂₅ nanoparticles on the surface of MnO₂ nanosheets were not clear. From the high-magnification image (Fig. S11b†), the Mito-Au₂₅ nanoparticles were distinctively observed on the surface of nanocomposites and the nanomaterials presented clear sheet morphology. Furthermore, freely dispersed Mito-Au₂₅ nanoparticles didn't exist. These results further confirmed the formation of Mito-Au₂₅@MnO₂. The ultimately obtained Mito-Au₂₅@MnO₂ displayed sheets morphology. Compared with mesoporous³² and nanoparticle³³ structures, ultrathin nanosheets have more sensitive responsibility to tumor microenvironment (TME),³⁴ which can not only lead to fast GSH consumption and oxidation pressure elevation in cancer cells, but also efficiently release photosensitizer to generate ¹O₂. Therefore, nanosheets morphology of Mito-Au₂₅@MnO₂ nanocomposites is benefit to the improvement of PDT efficiency. In the meanwhile, Mito-Au₂₅@MnO₂ nanocomposites maintained good thermal stability (Fig. S12†), which was also conducive to better PDT efficiency.

As expected, Mito-Au₂₅@MnO₂ nanocomposites could hardly generate ¹O₂ under 808 nm light irradiation (Fig. 3d). However, after GSH consuming part of MnO₂ nanosheets (Fig. S13†), the Mito-Au₂₅ nanoparticles released from nanocomposites still could generate ¹O₂ (Fig. 3d). The results verified the GSH activated PDT property of Mito-Au₂₅@MnO₂ nanocomposites evidently.

In virto mitochondrial targeting ability and PDT efficiency

In order to get ready for subsequent cellular experiments, we first tested the cellular uptake efficiency. To HepG2 cells, the optimum cellular uptake time of Au₂₅NCs and Mito-Au₂₅@MnO₂ were 2 h and 4 h, respectively (Fig. S14–S16†). The difference of cellular uptake time was probably related to the charge state, particle size and morphology of nanomaterials.^35–37 Targeting ability of Mito-Au₂₅ might also prolong cell retention time of nanocomposites. Mito Tracker Green (MTG) was used to compare mitochondrial colocalization efficiency of Au₂₅NCs and Mito-Au₂₅. A small part overlap of fluorescence signal of Au₂₅NCs and MTG was observed (Fig. S17†), indicating that Au₂₅NCs didn't have mitochondria specificity. On the country, thanks to the mitochondrial targeting peptide, the red fluorescence from Mito-Au₂₅ (Fig. 4a) and the green fluorescence from MTG (Fig. 4b) showed significant overlap (Fig. 4c and e). The colocalization degree also intuitively showed in Fig. 4d with a Pearson's correlation factor of 0.965. All these statistics proved that Mito-Au₂₅ could efficiently localized in mitochondria after cellular uptake.

Fig. 4. Fluorescence signal of (a) Mito-Au₂₅ (160 μg mL⁻¹, λ_ex = 488 nm, λ_em = 650–760 nm) and (b) Mito Tacker Green (0.2 μM, λ_ex = 488 nm, λ_em = 500–550 nm) in HepG2 cells. (c) Overlay of (a) and (b) in bright field. Scale bar: 10 μm. Pearson's colocalization coefficient (d) and profile gray value (e) of Mito-Au₂₅ and Mito Tracker Green. (f) Cell viability of HepG2 cells treated with Au₂₅NCs or Mito-Au₂₅@MnO₂ and irradiated at 808 nm (0 or 2.4 W cm⁻²) through CCK-8 assay. (g) Confocal images of Calcein-AM and PI stained HepG2 cells with different treatments (PBS, 808 nm light, Mito-Au₂₅@MnO₂, Au₂₅NCs + 808 nm light and Mito-Au₂₅@MnO₂ + 808 nm light). Scale bar: 50 μm.

DCFH-DA as a kind of ROS probe was used to detect ¹O₂ generated in HepG2 cells.³⁸ After oxidization by ¹O₂ to DCF, it showed bright green fluorescence under 488 nm excitation. As shown in Fig. S18,† HepG2 cells treated with Mito-Au₂₅@MnO₂ or 808 nm light nearly didn't exhibit green fluorescence signal. This phenomenon was similar with control group. While the cells treated with Mito-Au₂₅@MnO₂ followed by 808 nm light irradiation showed significant green fluorescence emission, indicating the generation of ¹O₂ in HepG2 cells.

To estimate the PDT efficiency of the prepared nanocomposites, four groups of HepG2 cells were treated under different conditions for 24 h and the cell viability were measured by CCK-8 assay. As shown in Fig. 4f, Au₂₅NCs, Mito-Au₂₅@MnO₂ and 808 nm light had favourable biocompatibility because cells treated in these groups remained a high viability over 80%. However, when first incubated with Au₂₅NCs or Mito-Au₂₅@MnO₂ and then accepted 808 nm light treatment, the cell viability sharply reduced to 40% and 4%, respectively (Fig. 4f). The result demonstrated superior cytotoxicity of Mito-Au₂₅@MnO₂ under 808 nm light irradiation. Furthermore, Calcein-AM and PI staining assay intuitively revealed the killing effect of different treatments (Fig. 4g). Taken together, the above results had confirmed that Mito-Au₂₅@MnO₂ had higher PDT efficiency than Au₂₅NCs in vitro.

In vivo PDT efficiency

Before investigating the in vivo PDT effect, the biocompatibility of Mito-Au₂₅@MnO₂ (0, 5, and 10 mg kg⁻¹) was evaluated. The results illustrated that the increasing trend of body weight from different group of mice had no difference during 14 days after injection (Fig. S19†), and the biochemical indicators of three groups of mice were all in normal range (Fig. S20†). H&E staining figures indicated few obvious injuries in major organs (Fig. S21†). These results suggested that Mito-Au₂₅@MnO₂ had favourable biocompatibility.

The 4T1 tumor bearing mice model was constructed to compare the therapeutic efficiency of Au₂₅NCs and Mito-Au₂₅@MnO₂. When the tumor volume reached 100–150 mm³, the mice were divided into five groups randomly and were given different treatments: (1) PBS; (2) 808 nm light irradiation (1.46 W cm⁻²); (3) Mito-Au₂₅@MnO₂; (4) Au₂₅NCs + 808 nm light irradiation and (5) Mito-Au₂₅@MnO₂ + 808 nm light irradiation. All these materials (PBS, Au₂₅NCs and Mito-Au₂₅@MnO₂) were directly injected into the tumor (Fig. 5a). The body weight of mice in five groups reflected the same increasing trend (Fig. S22†), and the H&E staining results indicated few damages in major organs (Fig. S23†), verifying that all the experimental treatments had hardly adverse effect to the mice. As shown in Fig. 5b and c, 808 nm light or Mito-Au₂₅@MnO₂ didn't affect the growth rate of tumor tissues and these two groups showed the same growing trend with blank control. On the contrary, the tumor growth rate of mice treated with nanomaterials together with 808 nm laser slowed down, and Mito-Au₂₅@MnO₂ was more effective than Au₂₅NCs in inhibiting tumor growth. Besides, the tumor tissues in different group were excised for H&E and TUNEL staining. As expected, the group treated with Mito-Au₂₅@MnO₂ and 808 nm light showed the highest level of tumor tissue damage (Fig. 5d), indicating that GSH consumption ability and mitochondrial targeting ability of nanocomposites could remarkably enhance the therapeutic efficiency.

Fig. 5. (a) Schematic illustration of in vivo antitumor study. (b) Relative tumor volume of mice with various treatments. (c) Representative photographs of the corresponding excised tumor (from left to right were given PBS, 808 nm light, Mito-Au₂₅@MnO₂, Au₂₅NCs + 808 nm light and Mito-Au₂₅@MnO₂ + 808 nm light, respectively). (d) Images of H&E and TUNEL stained tumor tissues. Scale bar: 100 μm.

Conclusions

In summary, we constructed an efficient nanocomposite system named Mito-Au₂₅@MnO₂ and fulfilled enhanced PDT. The nano-system consisting of Au₂₅(Capt)₁₈⁻, mitochondrial targeting peptide and GSH consumable MnO₂ nanosheets was structured through chemical modification and electrostatic interaction. Mito-Au₂₅@MnO₂ nanocomposite was an inactive photosensitizer before GSH decomposed MnO₂ nanosheets. Once MnO₂ nanosheets were depleted, the nanocomposite turned into Mito-Au₂₅ nanoparticles. With the consumption of GSH, oxidation pressure elevated and Mito-Au₂₅ anchored mitochondria in cancer cells. Cell viability decreased remarkably to 4% under 808 nm light irradiation by fully utilization of ROS. Eventually, we realized efficient therapy of 4T1 tumor bearing mice. These results confirmed that GSH consumption and mitochondrial targeting were promising strategies for PDT. We conceive that this study will encourage further exploitation on PDT based on Au NCs.

Ethical statement

All animal studies were performed according to the Guidelines for the Care and Use of Laboratory Animals of the Chinese Animal Welfare Committee and approved by the Institutional Animal Care and Use Committee, Wuhan University Center for Animal Experiment, Wuhan, China.

Conflicts of interest

The authors declare no competing financial interest.