Carbon-Dot-Decorated Carbon Nitride Nanoparticles for Enhanced Photodynamic Therapy against Hypoxic Tumor via Water Splitting Di-Wei Zheng,†,‡,§ Bin Li,†,§ Chu-Xin Li,† Jin-Xuan Fan,† Qi Lei,† Cao Li,‡ Zushun Xu,‡ and Xian-Zheng Zhang*,† †
Key Laboratory of Biomedical Polymers of Ministry of Education, Institute for Advanced Studies (IAS), Department of Chemistry, Wuhan University, Wuhan 430072, P.R. China ‡ Hubei Collaborative Innovation Center for Advanced Organic Chemical Materials, Key Laboratory for the Green Preparation and Application of Functional Materials of Ministry of Education, Hubei University, Wuhan, Hubei 430062, P.R. China S Supporting Information *
ABSTRACT: Hypoxia, a typical feature of solid tumors, remarkably restricts the efficiency of photodynamic therapy (PDT). Here, a carbon nitride (C3N4)-based multifunctional nanocomposite (PCCN) for light-driven water splitting was used to solve this problem. Carbon dots were first doped with C3N4 to enhance its red region absorption because red light could be used to trigger the in vivo water splitting process. Then, a polymer containing a protoporphyrin photosensitizer, a polyethylene glycol segment, and a targeting Arg-Gly-Asp motif was synthesized and introduced to carbon-dot-doped C3N4 nanoparticles. In vitro study showed that PCCN, thus obtained, could increase the intracellular O2 concentration and improve the reactive oxygen species generation in both hypoxic and normoxic environments upon light irradiation. Cell viability assay demonstrated that PCCN fully reversed the hypoxia-triggered PDT resistance, presenting a satisfactory growth inhibition of cancer cells in an O2 concentration of 1%. In vivo experiments also indicated that PCCN had superior ability to overcome tumor hypoxia. The use of water splitting materials exhibited great potential to improve the intratumoral oxygen level and ultimately reverse the hypoxia-triggered PDT resistance and tumor metastasis. KEYWORDS: water splitting, photodynamic therapy, carbon nitride nanoparticle, hypoxic tumor
P
able to enrich intratumoral O2 and induces oxygen selfenriching PDT. However, the use of perfluorohexane has limited effects in reducing intratumoral hypoxia. Catalase or MnO2 could catalyze H2O2 decomposition and produce O2, whereas the low intracellular raw material concentration (H2O2 < 50 μM) greatly limits the O2 generation amount of catalase and MnO2. In nature, plants integrate light harvesting, charge generation/separation, and catalytic reaction into chloroplast to produce O2 from abundant H2O with high efficiency. Inspired by this, scientists have used nanocomposites based on inorganic, organic, macromolecular, or hybrid materials as solar-driven water splitting materials to produce H2 and O2
hotodynamic therapy (PDT), which utilizes O2 to generate cytotoxic reactive oxygen species (ROS), is becoming a promising method in cancer treatment due to its noninvasive features, high efficiency, and ideal accuracy.1−3 However, the PDT process relies highly on the local oxygen level. Unfortunately, hypoxia is a key feature of the tumor microenvironment (pO2 ≤2.5 mmHg), and the low O2 level restricts the efficiency of PDT.4 Due to the PDT-induced oxygen consumption, worsened hypoxia would cause irreversible tumor metastasis or drug resistance.5 To solve this problem, clinical trials used hyperbaric oxygen (HBO) therapy to improve the therapeutic effect of PDT. However, serious side effects, including hyperoxic seizures and barotrauma, greatly limit the efficiency of HBO.6,7 Recent reports have indicated that O2 generation materials, such as perfluorohexane, catalase, and MnO2, could improve intratumoral O2 supplementation and enhance the efficiency of PDT, radiotherapy, and chemotherapy.8−14 Perfluorohexane is © 2016 American Chemical Society
Received: June 23, 2016 Accepted: August 17, 2016 Published: August 17, 2016 8715
DOI: 10.1021/acsnano.6b04156 ACS Nano 2016, 10, 8715−8722
Article
www.acsnano.org
Article
ACS Nano directly from water.15,16 In the past decade, water splitting materials have attracted extensive attention due to their promising applications in solving energy and environmental issues. Nevertheless, the biomedical use of water splitting materials has not been explored. Since water is also the most abundant compound in living organisms, and compared with previous O2-generating materials, using water as the source would provide unlimited row materials for in vivo O 2 generation. Among various types of water splitting materials, carbon nitride (C3N4) has attracted considerable attention for their adjustable band gap and band position. After the modification, water splitting can be driven under high penetrable red light (>600 nm), which makes C3N4 suitable for in vivo therapy.17,18 Most importantly, due to the absence of metal elements, C3N4 was considered to be a highly biocompatible material for biomedical applications. Keeping all these issues in mind, we hypothesize that enhanced PDT to fight against hypoxic tumors could be achieved by using water splitting materials. Here, metal-free C3N4 was chosen as a promising water splitting material. Since C3N4 has limited water splitting efficiency in the red light region, carbon dots were prepared and decorated into C3N4. After a ball-milling process, a carbon-dot-decorated C3N4 nanocomposite (CCN) with an enhanced red light absorption was synthesized.19 In addition, an amphipathic polymer, PpIXPEG-RGD, consisting of photosensitizer protoporphyrin IX (PpIX) and tumor-targeting sequence RGD (Arg-Gly-Asp) using the PEG (polyethylene glycol) as the linker, was assembled with CCN based on π−π stacking between PpIX and C3N4 to obtain polymer-modified, carbon-dot-doped carbon nitride nanoparticles (PCCN). When PCCN arrives and accumulates at the tumor tissue through active RGD targeting and enhanced permeability and retention (EPR) effect, a 630 nm laser was used to trigger PCCN to split water to generate O2. In addition, the photosensitizer could further transmit the energy to the produced O2 to generate cytotoxic singlet oxygen (1O2) for cancer treatment under 630 nm laser irradiation. The design of PCCN is diagramed in Figure 1, and this strategy would overcome the restriction of hypoxia in PDT.
RESULTS AND DISCUSSION Characterization of PCCN. In this study, a simple thermal decomposition method was performed to synthesize CCN derived from urea and carbon dots. Then, a ball-milling process was used to prepare CCN nanoparticles. Amphipathic polymer PpIX-PEG-RGD was synthesized, and its molecular weight was determined to be 1313 by electrospray ionization mass spectrometry. After self-assembly based on π−π stacking between PpIX and C3N4, PCCN was further prepared. The mechanism of C3N4-induced water splitting is illustrated in Figure 2a. C3N4 has a relatively small band gap (ca. 2.7 eV), and it can be activated under blue light (∼420 nm). Then, C3N4 transmits the energy to water and induces the water splitting reaction (requiring 1.23 eV theoretically).20 In this paper, carbon dots were doped to decrease the band gap of C3N4, and red light could be used to trigger the water splitting. As shown in Figure 2b, C3N4 nanoparticles had obvious absorption at around 340 nm but weak absorption at the red light region. After the doping of carbon dots, nanoparticles displayed an enhanced visible absorption over the entire wavelength range. Indicated by the dynamic light scattering analysis, PCCN showed a hydrodynamic size of 180 nm (PDI = 0.19), and transmission electron microscopy (TEM) images also suggested the formation of a nanocomposite (Figure S1). X-ray diffraction (XRD) analysis demonstrated that the C3N4 atomic structure was largely retained after various modifications (Figure S2). X-ray photoelectron spectra (XPS) were characterized with a monochromated Al Kα X-ray source. High oxygen atomic ratio and the appearance of an amide bond in PCCN indicated the successful modification of polymer (Figures 2d and S3). Thermogravimetry analysis also indicated the polymer incorporation, as shown in Figure S4. Also, the PpIX loading capacity of PCCN was calculated to be 9.6%. The interaction between PpIX-PEG-RGD and CCN was further studied, and PCCN was added into sodium dodecyl sulfate, Tween 20, urea, NaCl, and EDTA solutions. After a 2 h co-incubation, supernatants were collected and the UV−vis spectroscopy was performed. About 80% of PpIX-PEG-RGD was released after Tween 20 and sodium dodecyl sulfate treatment. This result indicated that hydrophobic competition triggered the self-assembly between PpIX-PEG-RGD and CCN. In contrast, treatments of urea (eliminating hydrogen bonds), NaCl (shielding electrostatic force), and EDTA (eliminating coordination bond) were ineffective for PCCN disassembly.21 In Vitro Study of Water Splitting and PDT Enhancement. To further prove the water splitting ability of PCCN, a dissolved oxygen meter was used to measure the O2 generation in phosphate-buffered saline (PBS) upon the irradiation of the 630 nm laser. After sodium thiosulfate treatment, O2 in PBS was consumed to obtain deoxygenated PBS. Notably, a liquid paraffin seal was also used to isolate deoxygenated PBS from air. As shown in Figure 2e, instantly after the light irradiation, increased O2 concentration could be observed in PCCN- and CCN-dispersed solutions. After 15 min light irradiation, the CCN-dispersed solution showed an oxygen concentration of 1.86 times higher than that in C3N4-dispersed solution. The enhanced water splitting effect under 630 nm laser was attributed to the carbon dot doping. PpIX would consume part of the O2 produced by CCN, thus, compared with CCN, reduced O2 generation of PCCN was observed. However, the water-splitting-triggered O2 generation of PCCN was still faster
Figure 1. Structure of PCCN and schematic diagram of 630 nm light-driven water splitting enhanced PDT; (I) carbon dot doping to prepare carbon-dot-doped C3N4; (II) PpIX-PEG-RGD selfassembly to prepare PCCN; (III) receptor-mediated endocytosis of PCCN; (IV) 630 nm light-driven water splitting to produce O2, in which O2 enhances PDT to produce ROS and kill the cancer cells. 8716
DOI: 10.1021/acsnano.6b04156 ACS Nano 2016, 10, 8715−8722
Article
ACS Nano
Figure 2. (a) Schematic illustration of the C3N4-mediated water splitting process; (b) UV−vis absorption spectra of C3N4, CCN, and PCCN; (c) hydrodynamic size of CCN and PCCN; (d) XPS spectra of CCN and PCCN; (e) O2 generation curve of C3N4, CCN, PCN, and PCCN; (f) ROS production efficiency of PCCN, PCN, CCN, and PpIX in hypoxic and normoxic environments.
than its PDT-induced O2 consumption. Polymer-modified carbon nitride (PCN) without carbon dot decoration showed no O2 generation under a 630 nm laser. Based on the experiments, a constant O2 level was found without laser irradiation. The ROS sensor, 2′,7′-dichlorofluorescein (DCFH), was used to measure the in vitro ROS generation ability of materials in both hypoxic and normoxic PBS.22 As shown in Figure 2f, in the normoxic environment, upon the 15 min laser irradiation for PpIX, PCN, and PCCN, a similar amount of ROS was produced. After 15 min of irradiation in a hypoxic environment, the ROS production of PpIX and PCN was extremely weak. Interestingly, in the hypoxic environment, remarkable fluorescence of DCFH was detected in the PCCN group, verifying the unaffected ROS generation ability under light irradiation. This result indicated that PCCN may overcome hypoxia in cancer cells and improve the PDT efficacy. After confirming light-driven water splitting and PDT enhancement of PCCN in PBS, we studied its intracellular O2 generation and ROS production in 4T1 cells. ROS-ID is a fluorogenic probe for intracellular hypoxia detection. In the normoxic environment, the ROS-ID hypoxia detection probe is a weakly fluorescent compound with a nitro group. However, in a hypoxic intracellular environment, the nitro group is reduced to hydroxylamine and amino groups, inducing the red fluorescence recovery.23 As shown in Figure 3a, in the normoxic environment, PpIX and PCN treatment displayed red fluorescence, whereas in the hypoxic environment, remarkably enhanced red fluorescence was observed in PpIX- and PCNtreated 4T1 cells after light irradiation. In contrast, weak fluorescence was observed in PCCN-treated cells after irradiation in both normoxic and hypoxic environments, indicating PCCN could overcome PDT-induced hypoxia. The overcoming of hypoxia also enhanced the ROS generation ability of PDT. Here, 2′,7′-dichlorofluorescein diacetate (DCFH-DA) was used as an intracellular ROS sensor because DCFH-DA with nonfluorescence can be rapidly hydrolyzed by esterase and oxidized with ROS to produce fluorescent 2′,7′-dichlorofluorescein (DCF). Confocal laser
Figure 3. (a) CLSM images of PDT-induced hypoxia reversion and intracellular ROS generation; (b) mean fluorescence intensity of the ROS-ID hypoxia detection probe and DCFH-DA staining 4T1 cells; (c) cell viability assay of PCN-treated 4T1 cells in hypoxic and normoxic environments; (d) cell viability assay of PpIX-treated 4T1 cells in hypoxic and normoxic environments; (e) cell viability assay of PCCN-treated 4T1 cells in hypoxic and normoxic environments.
scanning microscopy (CLSM) images confirmed that, in both hypoxic and normoxic environment, PCCN displayed a satisfactory ROS generation ability after light irradiation. Analyzed by ImageJ software, the quantitative mean fluorescence intensity study showed that the hypoxia only reduced 3% of DCF fluorescence as compared with normoxic 8717
DOI: 10.1021/acsnano.6b04156 ACS Nano 2016, 10, 8715−8722
Article
ACS Nano
Figure 4. (a) In vivo fluorescence imaging of PCCN at different time points after intravenous injection; (b) 3D reconstructed micro-CT transillumination fluorescent combination imaging of PCCN in vivo; (c) ex vivo fluorescence imaging of PCCN in the tumor, heart, liver, spleen, lung, and kidney; (d) CLSM images of the tumor, liver, spleen, and lung after PCCN treatment.
successfully accelerate ROS generation in a hypoxic environment. In Vivo Biodistribution Study of PCCN. To confirm that PCCN could achieve tumor-specific accumulation in vivo, a model of 4T1-cell-bearing mice was established. Due to the long emission wavelength of PpIX, a small animal fluorescence imaging system was used to study the in vivo distribution of PCCN. As shown in Figure 4a, before the injection of PCCN, tumor tissue displayed no fluorescence. After the injection, PpIX fluorescence started to accumulate in the tumor positon. Thirty-six hours after the injection, fluorescence intensity in the tumor tissue began to decrease. This result indicated that PCCN could accumulate in the tumor for a long time. The ex vivo fluorescence images of various organs and CLSM images of sections were also obtained at 1, 2, 4, and 8 h (Figures S9 and S10). Inevitable lung and liver accumulation was observed; however, a large amount of fluorescence still appeared in the tumor tissue, which coincided with the in vivo fluorescence imaging study. Then, transillumination (excitation light source located below the mouse) was performed to get both fluorescence intensity and depth information on the animal.30 Microcomputed tomography (micro-CT) imaging was also used in combination with fluorescent imaging based on transillumination because the combination of structural imaging (micro-CT) and functional imaging (fluorescent imaging) would further guide the therapy in vivo. As shown in Figure 4b, the 3D reconstructed micro-CT transillumination fluorescent combination image revealed that most of the fluorescence was located within the tumor.31 Ultimately, 48 h postinjection, mice were sacrificed and ex vivo fluorescence images of the various organs were performed. As presented in Figure 4c, most of the fluorescence accumulated in tumor tissue, whereas limited fluorescence was observed in the heart, liver, spleen, lung, and kidney. Furthermore, fluorescence of tumor, lung, liver, and kidney was observed in CLSM. As shown in Figure 4d, no PpIX fluorescence could be observed in metabolic organs, but strong fluorescence could be observed in the tumor region. In Vivo Overcoming Hypoxia Study of PCCN. The efficiency of PCCN to reverse tumor hypoxia was examined in
environment. The 4T1 cells treated with PpIX and PCN had negligible fluorescence in the hypoxic environment, and the quantitative study demonstrated that hypoxia decreased 52% of DCF fluorescence. Cellular Selectivity and Cytotoxicity Studies. In order to achieve a tumor-triggered targeting, the RGD sequence was introduced to PCCN. It is well-known that most cancer cells express high level of integrin αvβ3, thus PCCN could efficiently target cancer cells via the RGD sequence.24 Here, αvβ3 positive 4T1 cells and αvβ3 negative MCF-7 cells were used to study the cellular selectivity of PCCN,25−27 and the Western blot analysis of αvβ3 level in 4T1 cells and MCF-7 cells is shown in Figure S5. After 4 h co-incubation with PCCN, a larger area of fluorescence was observed in 4T1 cells (Figure S6). In contrast, red fluorescence was negligible in MCF-7 cells. Flow cytometry further indicated the addition of free RGD could inhibit the αvβ3-mediated uptake of PCCN in 4T1 cells (Figure S7). After illustrating the targeting ability of PCCN, we studied the efficiency of on-demand PDT. Since C3N4 is able to catalyze NADH regeneration, cell proliferation and cytotoxicity assays based on intracellular redox status would be affected.28 Herein, sulforhodamine B colorimetric assay based on cellular protein content was used, and the cytotoxicity of both PpIX and PCCN in hypoxic and normoxic environments was investigated.29 Without laser irradiation, no significant cytotoxicity of PCCN in 4T1 cells was observed (Figure S8). As presented in Figure 3c−e after laser irradiation, PpIX and PCN showed a significant cytotoxicity in normoxic environment (21% O2), while limited cytotoxicity in hypoxic incubator (
