Mycobacteria, particularly tuberculosis-causing Mycobacterium tuberculosis (M. tb), pose a serious threat to global health [1–4]. The current first-line treatment heavily relies on antibiotics, thereby resulting in the emergence of multidrug-resistant M. tb and high mortality in humans [5–8]. Therefore, an urgent need exists for the rational development of antimicrobial agents with novel mechanisms of action. In the past decade, antibacterial nanomaterials have become contenders as antibacterial treatments for use when traditional antibiotics fail. This emerging area of antibiotics includes many nanostructured materials, including nanocarriers loaded with antibiotics , noble metals or metal oxide nanoparticles [9–12], graphene or graphene oxide [13–15], organic–inorganic nanocomposites [16, 17], carbon nanotubes/dendrimers , self-assembled or unimolecular micelles [19, 20], and supramolecular nanoparticles [21, 22]. Although different types of antibacterial nanomaterials show substantial differences in their antimicrobial behavior, they share antimicrobial advantages such as minimal invasiveness, broad-spectrum antibacterial properties, and low drug resistance.
Specifically, drug-free nanosystems based on photodynamic therapy (PDT) are considered an emerging noninvasive modality for antibacterial treatment [23–27]. PDT uses photosensitizers (PSs) that absorb light and transfer tissue oxygen to reactive oxygen species (ROS) to kill bacteria. However, the short lifetime (<0.04 μs) and small active regions (<0.02 μm) of ROS are formidable challenges in PDT. In addition, the “always-active” model and the low selectivity of the currently used PSs can induce unexpected phototoxicity in both tumor and normal tissues, thus potentially damaging normal tissues and causing long-lasting cutaneous photosensitivity [28, 29]. To overcome these limitations, “smart” PSs have been developed to provide an additional level of control (beyond light and PS presence) of ROS generation. These smart PSs are designed to be quenched (“OFF’’ state) before administration. After accumulating in target tumor tissues, PSs are turned on (dequenched) via predetermined stimuli, such as protein, pH, DNA, or thiol responses [29, 30–34]. Among these stimuli, near-infrared (NIR) light lasers have attracted great interest because of their noninvasiveness and deep tissue penetration. NIR-sensitive nanocarriers, such as Au nanorods and graphene oxide, have been developed to turn off PSs [35, 36]. After NIR laser irradiation, the PS is released from NIR-sensitive nanocarriers and becomes highly phototoxic. Despite research progress, the strategy often has drawbacks of complicated molecular design, time-consuming synthesis, and relatively low antibacterial efficacy. The development of a simple approach for highly efficient activatable PDT is urgently needed.
Here, we present the design of a smart PS for photothermal-activatable PDT. The platform was constructed by functionalization of a peony-shaped FeSe2 with typical zinc (II) phthalocyanine (ZnPc) PSs ( Scheme 1a ). Effective energy transfer occurs between ZnPc PSs and FeSe2 because of their proximity, thus resulting in a nonphototoxic state of PSs (OFF state). Under NIR irradiation, the photothermal effect of FeSe2 promotes the release of PSs from the FeSe2 surface, thus turning on the photodynamic effect of PSs (ON state). Notably, FeSe2 not only acts as a smart carrier of PSs but also provides a highly efficient photothermal therapy (PTT) effect for enhanced antibacterial therapy. This activatable platform has robust photodynamic/photothermal antibacterial capability and therefore has great potential for broad-spectrum eradication of mycobacteria ( Scheme 1b ).
2. MATERIALS AND METHODS
FeCl3·6H2O was purchased from Aldrich. Se powder was purchased from Alfa Aesar. Oleylamine was purchased from Acros Organics Chemicals. Zinc phthalocyanine (ZnPc) and Tween 80 were purchased from Sigma-Aldrich. Tetrahydrofuran (THF) was obtained from J&K Scientific Ltd (San Jose, CA, USA). Singlet oxygen sensor green (SOSG) was obtained from Meilunbio. Difco™ Middlebrook 7H9 medium, 7H10 medium, and oleic acid–albumin–dextrose–catalase (OADC) were purchased from BD Biosciences (Rutherford, NJ, US). Glycerol was obtained from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). Deionized water (Millipore Milli-Q grade) with a resistivity of 18.2 MU was used in the experiments. All chemicals were used as received without further purification.
2.2 Synthesis of FeSe2-ZnPc
FeCl3·6H2O (1.0 mmol) and Se powder (2.0 mmol) were mixed with 10 mL of oleylamine in a 100 mL three-neck flask. Under mild stirring, the reaction mixture was heated to 120°C for 1 h under a nitrogen atmosphere. Subsequently, the temperature was increased to 180°C for another 1 h. The flask was then allowed to cool to room temperature. The FeSe2 product (black precipitate) was harvested via centrifugation and then washed three times with THF. For generation of FeSe2-ZnPc, 1 mL of ZnPc was added to 1 mL of FeSe2 solution. The mixture was stirred at room temperature for 24 h. The resulting FeSe2-ZnPc products were collected by centrifugation, washed with THF and ethanol several times, and dispersed in deionized water at 4°C for further use.
The size and morphology of the products were recorded with scanning electron microscopy (SEM) (Hitachi SU8220) and high revolution transmission electron microscopy (TEM) (JEM-1200EX). The ζ-potentials were determined by electrophoretic mobility measurement (Zetasizer Nano ZS). The absorption spectra were measured with a UV–vis spectrometer (V-750, JASCO). The fluorescence emission spectra were measured with a F-4500 fluorescence spectrometer (F-4500, Hitachi). X-ray diffraction (XRD) was performed on a D/MAX-TTRIII (CBO) series XRD instrument at 40 kV and 200 mA. Infrared-radiation thermal images were monitored through an infrared-radiation thermal camera (TiS65, Fluke).
2.4 NIR-laser-induced photothermal conversion
The temperature curves of FeSe2-ZnPc solutions were recorded under 808 nm NIR laser irradiation. Briefly, different concentrations of FeSe2-ZnPc solutions (0.25, 0.5, and 1.0 mg/mL) were prepared and irradiated by an 808 nm NIR laser (BWT Beijing Ltd, Beijing, China) with a power density of 1.0 W/cm2. The temperatures of the FeSe2-ZnPc solutions were monitored every 30 s with a thermal imaging camera. When the solution reached a steady-state temperature (6 min), the laser was shut off. The temperature of distilled water was used as a control.
The photothermal conversion efficiency (η) of FeSe2-ZnPc was calculated as follows: FeSe2-ZnPc solution was irradiated by an 808 nm NIR laser (1.0 W/cm2) for 6 min. Subsequently, the laser was shut off. The temperatures during the heating and cooling process were recorded. The time constant was determined by application of linear time data versus ln θ from the cooling stage of the FeSe2-ZnPc solution.
2.5 ZnPc release from FeSe2-ZnPc
NIR light-triggered release of ZnPc was investigated. The FeSe2-ZnPc solution was kept in the dark or irradiated by an 808 nm NIR laser (1.0 W/cm2). Subsequently, the solutions were submitted to centrifugation at several time points (0, 1, 2, 3, 4, or 5 min). The supernatant containing free ZnPc was collected for quantitative analysis with a UV–vis spectrophotometer.
2.6 NIR-activatable ROS generation of FeSe2-ZnPc
The ROS generation of FeSe2-ZnPc was detected with SOSG as an indicator in an aqueous environment. Typically, 1 mL of FeSe2-ZnPc solution was incubated with 5 μL SOSG stock solution (5 mM). The mixture was irradiated by an 808 nm light laser (1.0 W/cm2) and/or 640 nm light laser (0.1 W/cm2). The fluorescence emission of SOSG was recorded by a fluorescence spectrophotometer with an excitation wavelength of 488 nm.
2.7 Antibacterial activity testing
The M. smegmatis mc2–155 strain (M. smeg) was from our laboratory. The strains were cultured in 7H9 medium supplemented with 10% OADC, 0.05% (v/v) Tween 80 detergent for protein solubilization, and 0.2% (v/v) glycerol at 37°C. M. smeg was incubated in 7H9 medium with PBS (1:100, V/V) as a negative control, or FeSe2-ZnPc was added at a final concentration 25 μg/ml for 6 h, and photoirradiation was performed with an 808 nm laser (1.0 W/cm2) and/or 640 nm laser (0.1 W/cm2) (with a 10 min break after 2 min irradiation). For the FeSe2-ZnPc-808–640 group, the strain suspension was subjected to 808 nm laser irradiation for 20 min followed by 640 nm light for 20 min. After another 24 h of incubation, the strains were serially diluted 10-fold. Finally, the selected dilutions were plated on 7H10 medium supplemented with 10% OADC and 0.2% (v/v) glycerol at 37°C for 3 days. The colony forming units (CFU) were subsequently enumerated. Escherichia coli (E. coli) DH5α was purchased from TransGen Biotech. Ltd. Luria Bertani (LB) medium with or without bacterial agar was used for the liquid culture steps and CFU counts. The bacteria were cultured in LB broth and incubated overnight in a shaker at 220 rpm/min at 37°C. On the second day, the bacteria were inoculated in fresh LB broth at a 1:50 (V/V) dilution. Subsequently, FeSe2-ZnPc (10 μg/mL), or an equal volume of LB as a control, was added. Cultures were incubated for 2 h in a shaker 220 rpm/min at 37°C. After exposure to light and another 4 h of incubation, the bacteria were serially diluted 10-fold and plated on LB solid medium. The subsequent treatment and irradiation steps were as described for M. smeg.
3. RESULTS AND DISCUSSION
3.1 Successful synthesis and characterization of FeSe2-ZnPc
FeSe2-ZnPc was synthesized according to the process depicted in Scheme 1 . First, FeSe2 was prepared through a simple solution-phase method. SEM and TEM images indicated that the as-prepared FeSe2 had a uniform size (1.48 μm) and peony-shaped morphology ( Figure 1a-c , Figure S1 ). The XRD pattern showed sharp peaks corresponding to the orthorhombic structure of FeSe2, thereby confirming the high crystallinity of the as-synthesized FeSe2 ( Figure 1d ). The UV–vis–NIR absorption spectrum of FeSe2 showed broad absorption from the UV to the NIR-II range ( Figure 1e ), in agreement with the black color of the FeSe2 solution. Subsequently, the ZnPc molecules were loaded on the FeSe2 surface. The loading content of ZnPc was determined to be 3.2%. The absorption spectrum of FeSe2-ZnPc exhibited characteristic peaks of ZnPc ( Figure 1e ). In addition, zeta potential measurement indicated that FeSe2 was positively charged (7.5 mV) and became negatively charged (−12.5 mV) after surface modification with ZnPc ( Figure 1f ). In addition, the FTIR spectra of FeSe2 and FeSe2-ZnPc confirmed the successful attachment of ZnPc on FeSe2 ( Figure S2 ).
3.2 Photothermal performance of FeSe2-ZnPc
The strong NIR absorption of FeSe2 motivated us to study its NIR-triggered photothermal properties. NIR light (808 nm light) was chosen for excitation, because of its minimal heating effects and high tissue penetrability [37, 38]. As shown in Figure 2a , FeSe2-ZnPc had good photothermal properties under 808 nm NIR light irradiation. The temperature of the FeSe2-ZnPc dispersions increased to 69.5°C in 6 min, whereas the temperature of pure water increased to only 27.2°C. In addition, the temperature of FeSe2-ZnPc increased in a manner dependent on concentration and irradiation duration ( Figure 2b ), thus demonstrating the highly efficient phototothermal conversion ability of FeSe2-ZnPc. The photothermal conversion efficiency (η) of FeSe2-ZnPc was calculated to be 26.4% ( Figure 2c ). Furthermore, the photothermal effects were sustained during four ON/OFF cycles of 808 nm NIR light irradiation (an 808 nm light laser at 1.0 W/cm2, with 6 min laser irradiation for each cycle) without apparent changes in the peak temperature and the time to reach a plateau ( Figure 2d ). The effective photothermal performance and excellent photostability of FeSe2-ZnPc make it a favorable agent for PTT.
3.3 NIR-activatable PDT
To test the 808 nm NIR light-triggered release of ZnPc from FeSe2-ZnPc, we first prepared a standard curve, given the linear relationship between ZnPc concentration and absorption intensity ( Figure 3a ). Subsequently, the release profiles of ZnPc were monitored. Briefly, the FeSe2-ZnPc solution was kept in the dark or irradiated by an 808 nm NIR laser (1.0 W/cm2) for 1, 2, 3, 4, or 5 min. The supernatants were collected from the suspension by centrifugation, and the amount of ZnPc released was quantified with UV–vis spectroscopy. After 808 nm NIR laser irradiation, a rapid burst release (60%) of the loaded ZnPc was observed in the first 1 min, and 75.3% of ZnPc was released from FeSe2-ZnPc within 5 min ( Figure 3b ). In contrast, less than 18% of ZnPc was released without NIR irradiation. We speculated that the rapid heat generation of FeSe2 weakened the interaction between ZnPc and FeSe2, thus leading to higher release efficiency of ZnPc. Furthermore, the pH release profiles of ZnPc from FeSe2-ZnPc did not differ between pH 5.5 and 7.4 ( Figure S3 ), thus confirming that pH had no effect on ZnPc release.
This high release efficiency of ZnPc under 808 nm NIR irradiation encouraged us to test the controlled ROS generation of FeSe2-ZnPc. SOSG, whose fluorescence increases after oxidation by ROS, was chosen as an ROS indicator to assess the ROS generation ability of FeSe2-ZnPc. The fluorescence intensity of SOSG in the FeSe2-ZnPc solution increased under 640 nm light irradiation ( Figure 3c, d ). In contrast, the FeSe2-ZnPc solution without light irradiation did not generate ROS, thus indicating the photodynamic effect of FeSe2-ZnPc. Notably, the increase in the SOSG fluorescence when FeSe2-ZnPc was subjected to 808 nm light irradiation followed by 640 nm light irradiation (808 nm (+), 640 nm (+)) was much higher than that under the same 640 nm light irradiation conditions without 808 nm light irradiation (808 nm (−) and 640 nm (+)). These results indicated that the release of ZnPc from FeSe2, triggered by 808 nm light, effectively enhanced the photodynamic effect of PSs, thereby increasing the ROS generation capability of FeSe2-ZnPc.
3.4 Strong antimicrobial effect of FeSe2-ZnPc
Inspired by the robust photodynamic/photothermal capability of FeSe2-ZnPc, we evaluated the antimicrobial characteristics of FeSe2-ZnPc. M. smeg was chosen as the model bacterium for the antibacterial testing. Briefly, M. smeg was incubated with PBS or FeSe2-ZnPc for 6 h, and 808 nm and/or 640 nm laser photoirradiation was subsequently performed. After another 24 h of incubation, the strains were serially diluted 10-fold. The antibacterial results of different groups were determined through a typical colony counting strategy. After incubation with FeSe2-ZnPc for 6 h, the number of CFUs was similar to that in the control group ( Figure 4a ), thereby indicating a negligible effect of FeSe2-ZnPc on M. smeg. Subsequently, 808 nm and/or 640 nm photoirradiation was applied. Treatment with PBS followed by light irradiation did not significantly decrease the viability of M. smeg ( Figure 4b, c ), thereby allowing us to exclude the effect of light irradiation on the CFU of M. smeg. The CFU decreased to a greater extent after 24 h treatment with FeSe2-ZnPc than the negative control. Importantly, after irradiation with only an 808 nm or 640 nm laser, moderate inhibition of viability was observed in M. smeg treated with FeSe2-ZnPc, thus confirming the potential of FeSe2-ZnPc as a PTT or PDT agent. Notably, the highest cytotoxicity toward mycobacteria was achieved with treatment with FeSe2-ZnPc plus 808 nm and 640 nm light irradiation, thereby confirming that the combination of PTT and PDT increased antimycobacterial potency. Furthermore, no significant differences in the growth of M. smeg were observed when equal proportions of PBS and ddH2O were added to the 7H9 medium ( Figure S4 ). The antimicrobial effect of FeSe2-ZnPc was confirmed in E. coli, which showed the same results as M. smeg ( Figure S5 ). The above findings also indicated that the developed FeSe2-ZnP material has a universal bactericidal effect.
PDT is a promising strategy for the treatment of pathogenic microorganisms. At a specific light irradiation wavelength, PSs transfer photon energy to the surrounding molecular oxygen and consequently generate cytotoxic ROS. Compared with conventional antibiotic therapy, PDT has markedly greater selectivity and fewer adverse effects. ZnPc, a widely used PS, is promising for PDT because of the high quantum yield of ROS and low toxicity. However, the ZnPc inevitably is distributed to normal tissues, particularly the skin, as observed in clinical practice. Under sunlight, ZnPc generates ROS and can subsequently induce skin photosensitivity, thus resulting in unavoidable damage and long-lasting cutaneous photosensitivity. One possible reason for the undesirable phototoxicity is that ZnPc is always in the ON (or active) state, even in normal tissues.
To solve this problem, we report the design and synthesis of a smart PDT system based on a combination of peony-shaped FeSe2 particles and ZnPc PSs. ZnPc PSs are designed to be quenched (OFF state) before administration, because of the proximity between ZnPc and FeSe2. After accumulating in target tissues, ZnPc PSs can be turned on (dequenched) by 808 nm NIR light irradiation, thereby avoiding the potential adverse effects of active PSs. In addition, FeSe2 produces heat under 808 nm NIR irradiation and consequently could be used as a PTT agent. Therefore, our designed liposomes containing ZnPc and FeSe2 may be used as an activatable PDT combined with PTT for the eradication of pathogenic bacteria.