Photosensitizer (PS)/polyhedral oligomeric silsesquioxane (POSS)-crosslinked nanohybrids for enhanced imaging-guided photodynamic cancer therapy

Ya-Xuan Zhu a, Hao-Ran Jia a, Zhan Chen b and Fu-Gen Wu *a
aState Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing 210096, P. R. China. E-mail:
bDepartment of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109, USA

Received 31st March 2017 , Accepted 6th June 2017

First published on 8th June 2017

Photodynamic therapy (PDT) has drawn extensive attention as a promising cancer treatment modality. However, most PDT nanoagents suffer from insufficient drug loading capacity, a severe self-quenching effect, premature release of drugs and/or potential toxicity. Herein, we rationally designed an inorganic–organic nanohybrid with high drug loading capacity and superior chemical stability for enhanced PDT. Polyhedral oligomeric silsesquioxane (POSS), an amine-containing cage-shaped building block, was crosslinked with chlorin e6 (Ce6), a carboxyl-containing photosensitizer, via the amine–carboxyl reaction. Polyethylene glycol (PEG) polymers were further modified on the surface of the nanoparticle to improve the aqueous dispersibility and prolong the circulation time of the final nanoconstruct (POSS-Ce6-PEG). The as-prepared POSS-Ce6-PEG has a considerably high loading rate of Ce6 (19.8 wt%) with desirable fluorescence emission and singlet oxygen generation. Besides, in vitro experiments revealed that the nanoagent exhibited enhanced cellular uptake and a preferred intracellular accumulation within mitochondria and the endoplasmic reticulum, resulting in high anticancer efficiency under light irradiation. Furthermore, in vivo imaging-guided PDT was also successfully achieved, showing the effective tumor targeting and ablation ability of POSS-Ce6-PEG. More importantly, the nanoagent possesses negligible dark cytotoxicity and systemic side effects. Therefore, POSS-Ce6-PEG as an eligible PDT theranostic agent holds great potential in clinical applications.


Cancer, a major threat to human health, has been one of the leading causes of death in the world for decades. However, commonly used cancer treatment modalities (e.g., surgery, chemotherapy and radiotherapy) usually suffer from high invasiveness, limited therapeutic efficacy as well as severe side effects.1,2 Therefore, developing an alternative anticancer strategy is urgently needed. Photodynamic therapy (PDT) as a novel tumor-ablative treatment modality has drawn much attention in recent years because of its low systemic toxicity and high specificity.3 As a noninvasive and repeatable therapeutic method, it can also be applied in combination therapy.4,5 PDT involves a photosensitizer (PS), molecular oxygen and light irradiation to generate cytotoxic reactive oxygen species (ROS), in particular singlet oxygen (1O2), which induces cell apoptosis/necrosis and tissue destruction.6 However, since most PS molecules have poor water-solubility and are easy to aggregate in aqueous solutions due to the π–π stacking and hydrophobic interaction between the PS molecules, their ROS generation and PDT treatment outcome are not satisfactory for potential clinical applications.7 Another limitation of conventional PS molecules is their poor tumor accumulation because of their short blood circulation time after systemic administration,8 which also becomes a potential impediment to their clinical use.

To increase their water-dispersibility and tumor-targeting ability, PS molecules are usually encapsulated in nanoparticulate formulations such as liposomes,9–12 micelles,5,13,14 porphysomes,15,16 metal nanoparticles (NPs),17–20 polymeric NPs,21,22 protein-based NPs,23,24 mesoporous silica NPs,25,26 graphene oxide,27,28 upconversion NPs,29–31 and quantum dots.32,33 In general, physical encapsulation is a commonly used strategy to load PSs in different vehicles for improved therapeutic efficacy. Despite of the simplicity and generality in the fabrication of these physically formed nanocarriers, limited drug loading capacity and the uncontrollable premature release of PSs during systemic circulation still hamper their further applications.7 To address these issues, PSs are chemically conjugated with various nanocarriers to avoid their undesirable leakage and improve their in vivo performance.7 Since the loading capacity of PSs is largely dependent on the functional group density of the nanocarrier, drug carriers with multiple functional groups, such as polyethyleneimine (PEI),34 glycol chitosan21,22 and hyaluronic acid,35 have attracted extensive attention. However, because of the strong π–π interaction between PS molecules, excessive loading of PSs will possibly induce PS aggregation and the instability of nanoagents, resulting in a severe self-quenching effect and reduced PDT outcome.36,37 In addition, easy degradation and potential toxicity may also limit their further applications. It is expected that an ideal nanocarrier for PS delivery should meet the following criteria: excellent biostability, low toxicity and most importantly, high drug loading capacity without the self-quenching effect of PSs.

With the above goals in mind, herein, we selected a hybrid material, polyhedral oligomeric silsesquioxane (POSS), with unique characteristics to fabricate the PDT nanoagent. POSS is the smallest silica nanoparticle with a nanoscale cage-shaped structure. Benefiting from their tiny size and high functional group density, reactive POSS monomers have been widely used as building blocks to form hybrid inorganic–organic materials.38–49 In addition, POSS has witnessed significant advances in biomedical applications, such as bioimaging,50–54 tissue engineering,55 biodetection,41,42,53 antimicrobial treatment56–58 and drug delivery,44,45,59–62 owing to its superb chemical stability and desirable biocompatibility.63 Unfortunately, although POSS has been utilized as a nanocarrier to load PSs for PDT treatment,64 its limited loading capacity and severe quenching effect remain unsolved. In this work, we reported a POSS/chlorin e6 (Ce6)-crosslinked nanohybrid with excellent aqueous stability, low toxicity and considerably high drug loading capacity (19.8 wt%) for enhanced PDT. To be specific, the nanohybrid was first fabricated by the chemical crosslinking between POSS and Ce6 via their multiple amine groups (in POSS) and carboxyl groups (in Ce6) to form a three-dimensional network structure and then the above formed POSS-Ce6 intermediates were coated with PEG chains to construct the final nanoagent (POSS-Ce6-PEG) (Scheme 1). Different from most reported strategies for PS delivery, in our design, the Ce6 molecules serve not only as loaded drugs but also as indispensable building blocks for constructing the NPs, which ensures the considerably high drug loading capacity of the formed nanohybrid materials. Moreover, although Ce6 molecules were densely incorporated into the NPs, the fluorescence emission and 1O2 generation showed only a slight decrease. Excitingly, the POSS-Ce6-PEG NPs with enhanced cellular uptake realized remarkable cancer cell killing efficiency upon laser irradiation, while their dark cytotoxicity was negligible. In vivo imaging-guided PDT was then successfully achieved through systemic administration, demonstrating the excellent tumor accumulation, prolonged tumor retention and effective tumor ablation of the nanoagent. Besides, the carefully investigated biodistribution and systemic toxicity evaluation demonstrated that the nanoagents were safe, without noticeable side effects. Taken together, the present work reports a simple but novel strategy to fabricate POSS-based PDT nanohybrids with high drug loading capacity and good biocompatibility for enhanced anticancer therapeutic efficacy, which may promote the clinical applications of PDT.

image file: c7nr02279d-s1.tif
Scheme 1 Schematic illustration showing the synthetic method of POSS-Ce6-PEG NPs.



POSS was purchased from Hybrid Plastics (Hattiesburg, USA). Methoxyl PEG2000 succinimidyl ester (NHS-PEG2000-OMe) was bought from Nanocs, Inc. (New York, NY). Ce6 was obtained from J&K Scientific Ltd. N-Hydroxysuccinimide (NHS) and dimethyl sulfoxide (DMSO) were obtained from Aladdin Chemistry Co., Ltd. D-Serine, 3-aminopropyltriethoxysilane (APTES), ninhydrin, sodium dodecyl sulphate (SDS), sodium chloride (NaCl), urea, 1-ethyl-3-(3-(dimethylamino)propyl)carbodimide (EDC) and rhodamine 123 (Rhod-123) were purchased from Sigma-Aldrich (St Louis, MO, USA). Dialysis membranes with a molecular weight cut-off (MWCO) of 10 kDa (Spectra/Por®6 Dialysis membranes, Regenerated Cellulose) were ordered from Spectrum Labs. Deionized water (18.2 MΩ cm) was obtained from a Milli-Q system (Millipore, Billerica, MA). Cell counting kit-8 (CCK-8) was purchased from Beyotime Institute Biotechnology. Singlet oxygen sensor green (SOSG), LysoTracker Green, ER-Tracker Green and Hoechst 33342 were purchased from Invitrogen (Carlsbad, CA).

Synthesis of POSS-Ce6-PEG

To obtain POSS-Ce6-PEG, POSS was firstly conjugated with Ce6 via the EDC/NHS coupling method and then modified with PEG chains. Before synthesis, 2 mg POSS was dissolved in DMSO and mixed with 5.67 μL of triethylamine to neutralize the ammonium sites on the POSS units. Initially, 8.82 mg NHS and 7.92 mg EDC were dissolved in 1 mL of DMSO, respectively, and mixed with 3.05 mg Ce6 (5 mg mL−1 in DMSO), followed by agitation for 4 h at room temperature to activate Ce6. Then, the activated Ce6 was added to the pretreated POSS solution and the mixture was incubated for reaction under vigorous stirring overnight for the completion of the reaction to obtain POSS-Ce6. Afterward, 17.9 mg NHS-PEG2000-OMe (10 mg mL−1 in DMSO) was added to the above solution to react with the remaining amine groups of POSS moieties for 4 h. The mixture was successively dialyzed (MWCO = 10 kDa) against DMSO for 3 days and deionized water for 1 day. Finally, the POSS-Ce6-PEG dry powder was obtained after lyophilization of the above dialyzed solution.

Characterization of POSS-Ce6 and POSS-Ce6-PEG

To confirm the successful conjugation of POSS with Ce6, the infrared spectra of POSS, Ce6 and POSS-Ce6 were collected using a Thermo Scientific Nicolet iS50 Fourier transform infrared (FTIR) spectrometer. Besides, the amount of unreacted amine groups on the surface of POSS-Ce6 was quantified by a ninhydrin-based assay according to a previous report.65 To obtain the standard concentration curve of amine groups, D-serine and APTES at different concentrations were used as standard samples to conduct the ninhydrin-based assay.

Transmission electron microscopy (TEM) was used to characterize the size and morphology of the POSS-Ce6-PEG NPs. The sample was prepared by dropping POSS-Ce6-PEG NP solution (100 μg mL−1 in water) onto a glow-discharged carbon-coated grid. After being dried at room temperature, the sample was imaged by using a transmission electron microscope (JEM-2100, JEOL Ltd, Japan). The hydrodynamic diameter of POSS-Ce6-PEG NPs (200 μg mL−1) in a phosphate buffered saline (PBS) solution was characterized by dynamic light scattering (DLS) using a Zetasizer instrument (Nano ZS, Malvern Instruments, United Kingdom). The size of POSS-Ce6 (200 μg mL−1) was characterized in DMSO by DLS due to its poor water solubility. To determine the conjugation degree of Ce6 in POSS-Ce6-PEG, the UV–vis absorption spectra of Ce6 and POSS-Ce6-PEG solutions were recorded using a Shimadzu UV-2600 spectrophotometer. The concentration of Ce6 in POSS-Ce6-PEG solution was quantified by observing the absorbance at 670 nm, according to the standard concentration curve of free Ce6.

To investigate the intermolecular interactions in the POSS-Ce6-PEG NPs, 1 mg POSS-Ce6-PEG dry powder was dissolved in 1 mL of deionized water, SDS solution (20 mM), urea solution (50 mM) and NaCl solution (100 mM), respectively, and the corresponding sizes of these samples were characterized by DLS. To evaluate their long-term stability, POSS-Ce6-PEG was dispersed in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). The size of the NPs was measured by DLS every day for a week.

Fluorescence spectra of POSS-Ce6-PEG and free Ce6 were recorded using a spectrofluorophotometer (RF-5301PC, Shimadzu, Japan). SOSG was used to determine the generation of 1O2. To begin with, 1 μL of SOSG solution was added to 2 mL of free Ce6 and POSS-Ce6-PEG solutions (5 μg mL−1 Ce6 in both samples). Then, the solutions were irradiated by a laser (670 nm) at the power density of 8 mW cm−2 and the fluorescence spectra of SOSG were recorded at different time points.

Cell culture

Human cervical cancer cell line HeLa cells were cultured in DMEM supplemented with 10% FBS and 100 IU per mL penicillin–streptomycin at 37 °C under a humid atmosphere (5% CO2).

Confocal imaging

For confocal imaging, HeLa cells were seeded on 35 mm glass dishes and cultured at 37 °C for 24 h. Then, the HeLa cells were treated with free Ce6 (10 μM) or POSS-Ce6-PEG NPs (10 μM of Ce6) for 15 min, 1 h, 5 h and 12 h, respectively. Confocal images were taken by using an inverted confocal laser scanning microscope TCS SP8 (Leica, Germany) with a 63× oil immersion objective. To study the subcellular localization of POSS-Ce6-PEG NPs, the HeLa cells were treated with POSS-Ce6-PEG NPs (10 μM of Ce6) for 2 h, and then the media were carefully aspirated. Fresh media containing Rhod-123 (100 nM), LysoTracker Green (1 μM) and ER-Tracker Green (1 μM) were added to each dish, respectively. After incubation for 30 min, the culture medium in each dish was replaced by Hoechst 33342 working solution with further incubation for about 10 min. Before confocal imaging, all treated cells were washed with PBS solution for 3 times.

Flow cytometry

HeLa cells were treated with fresh culture media containing free Ce6 (10 μM) and POSS-Ce6-PEG NPs (10 μM of Ce6), respectively. After being incubated for different time periods (30 min, 1 h, 2 h, 3 h, 4 h, 5 h, 8 h, 12 h and 24 h), the cells were harvested and analyzed using a flow cytometer (NovoCyte, ACEA Bioscience, SD). The PerCP channel was selected to detect the fluorescence signal of Ce6 with excitation at 488 nm.

In vitro PDT and cytotoxicity evaluation

To study the phototoxicity of POSS-Ce6-PEG NPs, HeLa cells were seeded into 96-well plates at a density of 5 × 103 cells per well. After incubation for 24 h, the HeLa cells were treated with free Ce6 (3 μM) or POSS-Ce6-PEG (3 μM of Ce6) for 5 h. Afterward, the cells were irradiated with a laser (670 nm, 8 mW cm−2) for 1, 3, 5, 7 and 10 min. After further incubation for 12 h, the cell viabilities were determined by a CCK-8 assay: 10 μL of CCK-8 solution was added to each well and incubated with cells for 2 h at 37 °C, and the optical density (OD) at 450 nm was measured by using a microplate reader (Multiskan FC, Thermo-Scientific, USA). Cell viability was calculated as follows:
Cell viability (%) = (ODSample − ODBlank)/(ODControl − ODBlank) × 100.

To evaluate the cytotoxicity of the POSS-Ce6-PEG NPs and control samples (free Ce6 and POSS), the HeLa cells were seeded into 96-well plates at a density of 5 × 103 cells per well. After incubation for 24 h, various concentrations of free Ce6, POSS and POSS-Ce6-PEG were separately added to each well. After further incubation for 24 h, the CCK-8 assay was carried out as described above.

In vivo fluorescence imaging and PDT

For animal experiments, female BALB/c mice (15 ± 2 g) aged 4 weeks were purchased from Yangzhou University Medical Centre (Yangzhou, China) and used under protocols approved by Southeast University Laboratory Animal Center. After being acclimatized for at least 7 days, U14 tumors were inoculated by subcutaneous injection of 1 × 107 U14 cells (suspended in 100 μL of PBS) onto the dorsal side of the nude mice.

For in vivo fluorescence imaging, the U14 tumor-bearing mice were intravenously (i.v.) injected with 100 μL of PBS, free Ce6 and POSS-Ce6-PEG NPs (for free Ce6 and POSS-Ce6-PEG, the Ce6 dose was the same at 5 mg kg−1). Under deep anesthesia by continuous inhalation of a mixture of oxygen with isoflurane (5%), the treated mice were imaged by a Cri Maestroin and PerkinElmer in vivo imaging system with an excitation wavelength of 600 nm and an emission wavelength of 670 nm at different time points. To investigate the biodistribution of the POSS-Ce6-PEG NPs, the mice were sacrificed at 1 d, 3 d and 7 d postinjection. Their major organs and tumors were excised and observed under an in vivo imaging system. CRi Maestro Image software was used to quantify the fluorescence signals.

To evaluate the in vivo PDT effect of the POSS-Ce6-PEG NPs, the U14 tumor-bearing mice were prepared as described above. When the tumor volume reached approximately 50 mm3, the mice were randomly divided into 5 groups for different treatments: (1) i.v. injected with 100 μL of PBS, (2) i.v. injected with 100 μL of free Ce6 (dose: 5 mg kg−1) with laser irradiation, (3) i.v. injected with 100 μL of free Ce6 without laser irradiation, (4) i.v. injected with 100 μL of POSS-Ce6-PEG NPs (dose: Ce6 5 mg kg−1) with laser irradiation and (5) i.v. injected with 100 μL of POSS-Ce6-PEG NPs without laser irradiation. Specifically, in vivo PDT (670 nm, 30 mW cm−2) was performed at 6 h postinjection for 30 min. Tumor volumes and the weights of the mice were monitored every other day for 14 days. Tumor volumes were calculated as width2 × length/2.

Histological analysis

After in vivo PDT treatments, at day 14, the U14 tumor-bearing nude mice were sacrificed. Next, the major organs (heart, liver, spleen, lung and kidneys) and tumor tissues were collected and fixed in 4% paraformaldehyde, and paraffin sections were then made. Hematoxylin and eosin (H&E) staining of the above sections was carried out following standard protocols.

Statistical analysis

Data were expressed as the means ± standard deviations (SD). Error bars were based on the SD of at least three parallel samples. The differences among groups were determined using one-way analysis of variance (ANOVA) analysis followed by Tukey's post test (*P < 0.05, **P < 0.01, or ***P < 0.001). Image-Pro Plus (Rockville, MD, USA) was used to calculate the colocalization coefficients.66

Results and discussion

Synthesis and characterization of POSS-Ce6-PEG NPs

POSS is a three-dimensional and cage-like molecule which contains an inorganic siloxane core surrounded by aminopropyl groups at each of the eight corners. Owing to its high stability and ease of chemical functionalization, POSS serves as a promising building block to construct nanoparticles. Ce6 is a clinically used photosensitizer and has been broadly utilized to fabricate PDT nanoagents. Considering that a POSS molecule contains eight anime groups and a Ce6 molecule possesses three carboxyl groups, in this study, we innovatively constructed a photosensitive POSS-Ce6 core through the chemical crosslinking of POSS and Ce6 molecules via the amine–carboxyl reaction, benefiting from the nonspecific conjugation property of the EDC/NHS coupling method.67 Then, hydrophilic PEG polymers were covalently modified onto the surface of POSS-Ce6 conjugates to form the final POSS-Ce6-PEG nanoconstructs via the reaction between the NHS group of NHS-PEG2000-OMe and the remaining unreacted amine groups of POSS. Note that the PEG layer not only ensures the good water-solubility and aqueous stability of the formed NPs, but also endows the NPs with the “stealth” property to minimize mononuclear phagocyte system (MPS) uptake and prolong circulation time after systemic administration. Herein, POSS was selected for the following advantages: (1) the rigid cage-like structure of POSS can prevent the incorporated Ce6 molecules from self-quenching, which ensures that their fluorescence emission and 1O2 generation remain unaffected; (2) POSS can serve as building blocks to form three-dimensional network nanostructures through crosslinkage with Ce6 molecules; (3) the abundant remaining unreacted amine groups of POSS on the surface of POSS-Ce6-PEG NPs are beneficial for further reaction with NHS-PEG to construct “stealthy” nanoparticles; (4) the cage-like structure of POSS can provide a hydrophobic environment for further physical entrapment of other therapeutic agents.

We first characterized the intermediate product POSS-Ce6 before PEG modification. FTIR spectroscopy was used to confirm the successful conjugation of POSS with Ce6. As shown in Fig. S1, the strong absorption peaks at around 1110 cm−1 contributed by the asymmetric stretching vibration of Si–O–Si were observed in both POSS and POSS-Ce6, showing that the prepared POSS-Ce6 contained the Si–O–Si network. Besides, two new absorption peaks at 1644 and 1571 cm−1 (attributed to the C[double bond, length as m-dash]O stretching vibration and N–H bending vibration of the amide bond, respectively) observed in POSS-Ce6 demonstrated that the POSS was successfully crosslinked with Ce6 via the amide linkage. Furthermore, the absorption at 3410 cm−1 ascribed to the stretching vibration of –NH2 observed in both POSS and POSS-Ce6 might indicate that a part of the –NH2 groups still remained unreacted in POSS-Ce6, which ensured the subsequent PEG modification via the NH2–NHS ester reaction. To further confirm the presence of amine groups in POSS-Ce6, we carried out a ninhydrin-based assay to quantitatively measure the amount of –NH2 groups on the surface of POSS-Ce6 according to the previously reported method.65 The result showed that 1 mg mL−1 POSS-Ce6 solution contains approximately 1.2 mM –NH2 groups.

Then, the size and morphology of POSS-Ce6-PEG NPs were characterized by TEM and DLS. The TEM image confirmed the spherical shape of POSS-Ce6-PEG NPs (Fig. 1a). As shown in Fig. 1b, the average sizes of POSS-Ce6 in DMSO and POSS-Ce6-PEG in a PBS solution were 65 ± 13 nm and 70 ± 16 nm, respectively, demonstrating the successful modification of PEG polymers. It is worth noting that the polydispersity index (PDI) of POSS-Ce6-PEG NPs was below 0.1, which confirmed their excellent dispersity in aqueous solution. In addition, the time-dependent stability of the NPs in serum-containing culture medium was evaluated, which displayed negligible changes in size over a time period of 7 days (Fig. S2).

image file: c7nr02279d-f1.tif
Fig. 1 Characterization of POSS-Ce6-PEG NPs. (a) TEM image of POSS-Ce6-PEG NPs (scale bar = 100 nm). (b) Hydrodynamic size distribution of the POSS-Ce6 and POSS-Ce6-PEG NPs (measured by DLS), respectively. (c) The hydrodynamic sizes of POSS-Ce6-PEG in different solutions (water, SDS, NaCl and urea). (d) UV-vis spectra of POSS-Ce6-PEG (55 μg mL−1) and free Ce6 (10 μg mL−1). (e) Fluorescence spectra of POSS-Ce6-PEG and free Ce6 in PBS solutions at the same concentration of Ce6 (5 μg mL−1).

To further verify that POSS was covalently crosslinked rather than physically associated with Ce6, SDS (to disassociate the hydrophobic interactions in complexes), urea (to participate in the formation of competitive hydrogen bonds) and NaCl (to exert an electrostatic shielding effect) were added to the POSS-Ce6-PEG solutions, respectively. However, all these treatments were ineffective in dissociating the NPs, as evidenced by the DLS result that the hydrodynamic sizes were not affected after the addition of the above-mentioned reagents (Fig. 1c). Therefore, no physical association was involved in the POSS-Ce6-PEG nanoconstruct, which guaranteed its excellent stability under physiological conditions. Compared with free Ce6, POSS-Ce6-PEG exhibited a slightly red-shifted absorbance peak, possibly due to the changed molecular conformation of Ce6 after conjugation with POSS (Fig. 1d).

Next, we evaluated the amount of Ce6 loaded in the NPs by measuring the characteristic absorbance at 670 nm and found that POSS-Ce6-PEG achieved effective loading of Ce6 (19.8 wt%). The drug loading capacity in this system is significantly higher than that in most reported PDT strategies (generally less than 10 wt%).7 In our design, Ce6 serves as not only a photosensitizer but also an indispensable building block of the NPs, which can dramatically increase the amount of loaded Ce6. Interestingly, we noticed that the fluorescence intensity of Ce6 was not affected after being massively incorporated into the POSS-Ce6-PEG NPs (Fig. 1e). Typically, the fluorescence of porphyrins will be severely quenched when they are introduced into NPs at a high local concentration, mainly attributed to the strong π–π stacking between the neighboring porphyrin molecules. However, in view of the three-dimensional cubic structure of POSS and the rigidity of the silica cage, the crosslinked Ce6 molecules may be constrained to the corners of POSS with restricted conformational flexibility, which consequently weakens their π–π stacking. Additionally, the 1O2 generation of POSS-Ce6-PEG NPs was also evaluated by measuring the fluorescence intensity of SOSG in the course of the irradiation time (Fig. S3). Compared with free Ce6, the increase in SOSG fluorescence intensity in the POSS-Ce6-PEG group after laser irradiation was only slightly lowered.

Cellular uptake of POSS-Ce6-PEG NPs

We then investigated the cellular uptake of POSS-Ce6-PEG NPs by confocal imaging. For comparison, human cervical cancer cell line HeLa cells were incubated with POSS-Ce6-PEG NPs and free Ce6 at the same Ce6 concentration of 10 μM. As shown in Fig. 2a, the fluorescence signal of HeLa cells treated with POSS-Ce6-PEG NPs gradually increased and reached a maximum at 5 h. With respect to the free Ce6 group, the fluorescence inside the cells was very weak even after incubation for 12 h. The marked difference in the intracellular fluorescence of the two samples indicated that POSS-Ce6-PEG NPs achieved significantly enhanced cellular uptake by HeLa cells. This result was further confirmed and quantified by flow cytometry, which revealed that the amount of internalized Ce6 in the POSS-Ce6-PEG NP group was 4-fold higher than that in the free Ce6 group after incubation for 5 h (Fig. 2b).
image file: c7nr02279d-f2.tif
Fig. 2 Cellular uptake and in vitro PDT results of POSS-Ce6-PEG NPs. (a) Confocal fluorescence images of HeLa cells incubated with POSS-Ce6-PEG NPs and free Ce6 (both containing 10 μM of Ce6) at different time points, respectively. Scale bar = 10 μm. (b) Fluorescence (FL) intensities of HeLa cells treated with free Ce6 or POSS-Ce6-PEG for different time periods as measured by flow cytometry. (c) Dark cytotoxicity evaluation of HeLa cells after treatment with various concentrations of POSS, free Ce6 and POSS-Ce6-PEG NPs for 24 h. (d) Relative viabilities of HeLa cells treated with free Ce6 or POSS-Ce6-PEG NPs (3 μM of Ce6 in both samples, 5 h) after irradiation by a 670 nm laser (8 mW cm−2) for different time periods. The results were expressed as the mean ± s.d. **P < 0.01, one-way ANOVA. (e) Trypan blue staining image of HeLa cells treated with POSS-Ce6-PEG NPs after irradiation with a 670 nm laser (8 mW cm−2, 5 min). The white dotted square represents the laser irradiation area. Scale bar = 200 μm.

Dark cytotoxicity evaluation and in vitro PDT

Encouraged by the enhanced cellular uptake of the POSS-Ce6-PEG NPs, we then carried out in vitro experiments. To meet the requirements of clinical treatment, an ideal PDT agent must possess low, or even no, dark toxicity. We thus assessed the dark cytotoxicity of the POSS-Ce6-PEG NPs and the control samples (POSS and free Ce6) toward the HeLa cells (Fig. 2c). We found that the cell viability was not affected as the concentration of the NPs increased from 1 to 30 μM of Ce6. In contrast, the free Ce6 exhibited potential cytotoxicity to the HeLa cells at a relatively high concentration (30 μM), displaying only 80% cell viability. These results confirmed the low dark toxicity of the POSS-Ce6-PEG NPs, which is advantageous over the traditional formulation. To evaluate in vitro PDT efficiency, the HeLa cells were then pretreated with the POSS-Ce6-PEG NPs and the free Ce6 in culture media (both containing 3 μM of Ce6). After being incubated for 5 h, the cells were irradiated with a 670 nm laser (8 mW cm−2) for different irradiation time periods. After incubation for another 12 h, cell viabilities were measured by the CCK-8 assay. As expected, the POSS-Ce6-PEG NPs presented much higher phototoxicity as compared with free Ce6 (Fig. 2d). We found a trend of decreasing cell viability with increasing irradiation time and over 95% of cells were killed after being irradiated for 7 min. As a control, free Ce6 exhibited only moderate PDT efficacy, leaving more than 70% of viable cancer cells unaffected even after irradiation for 10 min. We believed that the enhanced cellular uptake of POSS-Ce6-PEG NPs was the key factor for the significantly improved PDT efficiency.

Trypan blue staining was also conducted to prove the cell killing effect of POSS-Ce6-PEG NPs combined with laser irradiation. The white dotted square in Fig. 2e is the laser irradiation area. A blue quadrate region matching the laser spot was observed, verifying that all the cells were effectively killed after laser irradiation. In contrast, the cells without laser irradiation remained viable. These results intuitively confirmed the potential of POSS-Ce6-PEG NPs for controlled killing of tumor cells without harming adjacent normal tissues. Taken together, POSS-Ce6-PEG NPs with enhanced in vitro PDT efficiency and decreased dark cytotoxicity were promising for in vivo PDT treatment.

Subcellular colocalization assay

To identify the subcellular distribution of POSS-Ce6-PEG NPs, the NP-treated HeLa cells were co-stained with Hoechst 33342 (to label cell nucleus), ER-Tracker Green (to label endoplasmic reticulum, ER), LysoTracker Green (to label lysosome) and Rhod-123 (to label mitochondrion). Confocal images (Fig. 3a) clearly showed that the fluorescence ascribable to POSS-Ce6-PEG NPs did not match that of Hoechst 33342 but coincided with what was observed in the cases of ER-Tracker Green, LysoTracker Green and Rhod-123. These results demonstrated that POSS-Ce6-PEG NPs were distributed throughout the entire cytoplasm after being internalized and were localized to multiple organelles including ER, lysosomes and mitochondria. To evaluate the subcellular location of POSS-Ce6-PEG more accurately, we calculated the overlap coefficient (OLC) and Pearson's correlation coefficient (PCC), and the correlation plots of POSS-Ce6-PEG with the ER-Tracker, LysoTracker and Rhod-123 were presented in Fig. 3b. The OLC and PCC values for POSS-Ce6-PEG and ER-Tracker channels were 0.91 and 0.89, respectively, which were slightly higher than those for the POSS-Ce6-PEG and Rhod-123 channels (0.86 and 0.84, respectively). In contrast, the NPs did not significantly accumulate in the lysosomes, with the OLC and PCC values lower than 0.62 and 0.67, respectively. Hence, these results clearly demonstrated that POSS-Ce6-PEG NPs were mainly distributed in the ER and mitochondria, with a partial distribution in the lysosomes. Because the migration distance of 1O2 is limited by its short lifetime,68 the PDT efficiency largely depends on the subcellular location of the PS. It has been reported that PSs localized in diverse cellular organelles may lead to different mechanisms of cell death and therapeutic effectiveness.69 For example, a PS entrapped in lysosomes will significantly lower cell killing efficiency, while mitochondria-targeted PDT is found to be more effective, and ER is one of the most vital sites for PDT as well.6,70 We believe that the multiple subcellular distribution pattern, especially the efficient targeting ability of mitochondria and ER, makes POSS-Ce6-PEG an excellent PDT agent.
image file: c7nr02279d-f3.tif
Fig. 3 Subcellular localization of POSS-Ce6-PEG NPs in HeLa cells. (a) Confocal images of POSS-Ce6-PEG-treated HeLa cells co-stained with Hoechst 33342 and one of the following organelle-specific dyes: ER-Tracker Green, LysoTracker Green and Rhod-123, respectively. Scale bar = 10 μm. (b) The intensity correlation plots of POSS-Ce6-PEG with ER-Tracker Green, LysoTracker Green and Rhod-123, respectively.

In vivo fluorescence imaging and in vivo PDT

Motivated by their remarkable in vitro PDT performance, we further evaluated the feasibility of using POSS-Ce6-PEG NPs as in vivo theranostic PDT nanoagents. Nude mice bearing subcutaneous U14 murine cervical cancer xenograft were selected as the animal model. When the tumors reached 50 mm3, the mice were intravenously (i.v.) injected with POSS-Ce6-PEG NPs and free Ce6 (dose: 5 mg kg−1) and imaged under a Cri Maestroin and PerkinElmer in vivo imaging system at different time points. As shown in Fig. 4a, the fluorescence of POSS-Ce6-PEG NPs was distributed throughout the mouse body after injection and evident tumor accumulation was observed at 2 h postinjection. Besides, the fluorescence intensity at the tumor area reached the maximum at 6 h postinjection and gradually decreased as a function of circulation time (Fig. 4b). Comparatively, no apparent fluorescence signal was detected at the tumor region in the free Ce6 group, indicating the poor tumor accumulation ability of free Ce6 during systemic circulation. The in vivo imaging results clearly proved that POSS-Ce6-PEG NPs enabled enhanced tumor-targeting ability via the enhanced permeability and retention (EPR) effect and achieved high-contrast fluorescence imaging of tumor tissues, which were very advantageous to subsequent imaging-guided PDT. To better investigate the biodistribution of POSS-Ce6-PEG NPs, nude mice were i.v. injected with the NPs and their major organs and tumor tissues were excised at 1 d, 3 d and 7 d postinjection, followed by ex vivo fluorescence imaging. As shown in Fig. 4c, POSS-Ce6-PEG NPs were mainly distributed in the tumor and liver with partial accumulation in the kidneys at 1 d postinjection. Excitingly, significant tumor accumulation of the NPs was still observed at 3 d postinjection, while the fluorescence signals in the liver and kidneys were apparently weakened. At 7 d postinjection, all major organs as well as the tumor tissue exhibited low fluorescence signals, as evidenced by the quantitative biodistribution analyses (Fig. 4d), suggesting that most POSS-Ce6-PEG NPs were cleared from the body. Based on the aforementioned results, POSS-Ce6-PEG NPs with superb fluorescence imaging properties and excellent tumor-targeting ability hold great potential for in vivo imaging-guided PDT.
image file: c7nr02279d-f4.tif
Fig. 4 In vivo and ex vivo fluorescence imaging. (a) In vivo fluorescence images of U14 tumor-bearing nude mice taken at different time points after i.v. injection of free Ce6 and POSS-Ce6-PEG NPs, respectively. The green dotted circles indicate the tumor areas. (b) Relative tumor fluorescence intensities of the free Ce6- and POSS-Ce6-PEG NP-treated mice based on the in vivo fluorescence images in (a). (c) Ex vivo fluorescence images of major organs and tumor tissue excised from the mice injected with POSS-Ce6-PEG NPs at 1 d, 3 d and 7 d postinjection. H, Li, S, Lu, K and T stand for heart, liver, spleen, lung, kidneys and tumor, respectively. (d) Quantitative biodistribution of POSS-Ce6-PEG NPs in major organs and tumor tissue at 1 d, 3 d and 7 d postinjection.

To carry out in vivo PDT, U14 tumor-bearing nude mice were i.v. injected with PBS, free Ce6 and POSS-Ce6-PEG NPs (dose: Ce6 5 mg kg−1). According to the in vivo fluorescence imaging, the injected mice were irradiated with a 670 nm laser (30 mW cm−2) for 30 min at 6 h postinjection. The mice injected with free Ce6 and POSS-Ce6-PEG NPs but without laser irradiation were set as control groups. The tumor volumes and body weights of the treated mice were monitored every other day for 14 days. The tumors on the mice in the control groups showed rapid growth. The “free Ce6 + laser” group showed only slight tumor regression, which might be attributed to its poor tumor-targeting ability. Comparatively, the tumors on the mice injected with POSS-Ce6-PEG NPs after laser irradiation treatment were effectively ablated, and no regrowth was observed within 14 days (Fig. 5a and c). These results were further supported by the H&E staining of tumor slices, where the tumor tissue of POSS-Ce6-PEG NP-treated mice was completely destructed (Fig. 5b). To evaluate the systemic toxicity of POSS-Ce6-PEG NPs, the mice were sacrificed at the 14th day and their major organs were excised for H&E staining. Compared with the control group, no pathological changes to the organs were observed (Fig. 5d). Besides, the body weights of the mice in all groups showed no significant difference (Fig. 5e). Therefore, these results clearly demonstrated that POSS-Ce6-PEG NPs achieved excellent in vivo PDT efficacy with negligible side effects.

image file: c7nr02279d-f5.tif
Fig. 5 In vivo PDT results in U14 tumor-bearing mice. (a) Representative photographs of U14 tumor-bearing mice injected with PBS, free Ce6 and POSS-Ce6-PEG NPs (dose: Ce6 5 mg kg−1) after 670 nm laser irradiation (30 mW cm−2, 30 min) at different time points, respectively. (b) H&E-stained tumor slices collected from different groups of mice at the 14th day (scale bar = 100 μm). (c) Tumor growth curves of different groups of mice after various treatments. *P < 0.05, **P < 0.01, ***P < 0.001, one-way ANOVA. (d) H&E-stained tissue slices of major organs (heart, liver, spleen, lung and kidneys) in mice injected with PBS (control) and POSS-Ce6-PEG NPs after 14 days of PDT treatment, respectively. Scale bar = 100 μm. (e) Changes of relative body weights of mice after different treatments as a function of time.

The rationally designed drug delivery system for enhanced PDT has the following advantages: (1) Ce6 serves not only as a therapeutic agent but also as a dominant constituent for constructing nanostructures, which guarantees considerably high drug loading capacity; (2) POSS-Ce6-PEG has highly stable nanostructures formed by covalent crosslinking, which can avoid unfavorable premature release of drugs during systemic circulation; (3) owing to the cubic structure of POSS, the fluorescence emission and 1O2 generation of Ce6 remain unquenched even after being densely incorporated into NPs; (4) POSS-Ce6-PEG NPs exhibit significantly enhanced cellular uptake and a preferred distribution in mitochondria and ER, ensuring their effective cell killing performance; (5) POSS-Ce6-PEG NPs can achieve imaging-guided PDT to eliminate tumors with negligible safety concerns; (6) the hydrophobic core of POSS enables further physical entrapment of drugs, making POSS-Ce6-PEG NPs an ideal platform for drug delivery.


In summary, a novel POSS-based PDT nanoagent with high PS loading capacity was fabricated for enhanced imaging-guided photodynamic cancer therapy. In our design, POSS and Ce6 were chemically crosslinked, due to their multiple amine and carboxyl groups, to form a three-dimensional network structure, which was then modified with PEG polymers on the surface. The as-synthesized POSS-Ce6-PEG NPs possessed high Ce6 (19.8 wt%) loading capacity, excellent water-solubility and long-term stability. The fluorescence emission and 1O2 generation efficiency of Ce6 in this nanostructure were scarcely affected as compared with those of free Ce6. We have successfully demonstrated that POSS-Ce6-PEG with enhanced cellular uptake and negligible dark cytotoxicity is a promising PDT nanoagent. In vivo imaging-guided PDT was further realized with a desirable therapeutic outcome, benefiting from the excellent in vivo fluorescence property, enhanced tumor-targeting ability and prolonged blood circulation of POSS-Ce6-PEG NPs. More importantly, the nanoagent exhibited low systemic toxicity without noticeable side effects. The rationally formulated PDT nanoagent may shed new light on the fabrication of functional nanomaterials and also hold great potential for combating cancers.


This work was supported by grants from the National High Technology Research & Development Program of China (2015AA020502), the National Natural Science Foundation of China (21673037), the Fundamental Research Funds for the Central Universities (2242015R30016), the Six Talents Peak Project in Jiangsu Province (2015-SWYY-003) and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry. ZC acknowledges the support from the University of Michigan.

Notes and references

  1. A. Albini, G. Pennesi, F. Donatelli, R. Cammarota, S. D. Flora and D. M. Noonan, J. Natl. Cancer Inst., 2010, 102, 14–25 CrossRef CAS PubMed .
  2. V. Pavet, M. M. Portal, J. C. Moulin, R. Herbrecht and H. Gronemeyer, Oncogene, 2011, 30, 1–20 CrossRef CAS PubMed .
  3. J. F. Lovell, T. W. B. Liu, J. Chen and G. Zheng, Chem. Rev., 2010, 110, 2839–2857 CrossRef CAS PubMed .
  4. J. Lin, S. J. Wang, P. Huang, Z. Wang, S. H. Chen, G. Niu, W. W. Li, J. He, D. X. Cui, G. M. Lu, X. Y. Chen and Z. H. Nie, ACS Nano, 2013, 7, 5320–5329 CrossRef CAS PubMed .
  5. T. T. Wang, D. G. Wang, H. J. Yu, M. W. Wang, J. P. Liu, B. Feng, F. Y. Zhou, Q. Yin, Z. W. Zhang, Y. Z. Huang and Y. P. Li, ACS Nano, 2016, 10, 3496–3508 CrossRef CAS PubMed .
  6. L. Benov, Med. Princ. Pract., 2015, 24, 14–28 CrossRef PubMed .
  7. S. S. Lucky, K. C. Soo and Y. Zhang, Chem. Rev., 2015, 115, 1990–2042 CrossRef CAS PubMed .
  8. Y. N. Konan, R. Gurny and E. Allémann, J. Photochem. Photobiol., B, 2002, 66, 89–106 CrossRef CAS .
  9. A. Yuang, X. L. Tang, X. F. Qiu, K. Jiang, J. H. Wu and Y. Q. Hu, Chem. Commun., 2015, 51, 3340–3342 RSC .
  10. C. S. Shemesh, C. W. Hardy, D. S. Yu, B. Fernandez and H. L. Zhang, Photodiagn. Photodyn. Ther., 2014, 11, 193–203 CrossRef CAS PubMed .
  11. R. Di Corato, G. Béalle, J. Kolosnjaj-Tabi, A. Espinosa, O. Clément, A. K. Silva, C. Ménager and C. Wilhelm, ACS Nano, 2015, 9, 2904–2916 CrossRef CAS PubMed .
  12. L. Z. Feng, L. Cheng, Z. L. Dong, D. L. Tao, T. E. Barnhart, W. B. Cai, M. W. Chen and Z. Liu, ACS Nano, 2017, 11, 927–937 CrossRef CAS PubMed .
  13. L. Cheng, A. Kamkaew, H. Y. Sun, D. W. Jiang, H. F. Valdovinos, H. Gong, C. G. England, S. Goel, T. E. Barnhart and W. B. Cai, ACS Nano, 2016, 10, 7721–7730 CrossRef CAS PubMed .
  14. M. Guo, H. J. Mao, Y. L. Li, A. J. Zhu, H. He, H. Yang, Y. Y. Wang, X. Tian, C. C. Ge, Q. L. Peng, X. Y. Wang, X. L. Yang, X. Y. Chen, G. Liu and H. B. Chen, Biomaterials, 2014, 35, 4656–4666 CrossRef CAS PubMed .
  15. C. S. Jin, J. F. Lovell, J. Chen and G. Zheng, ACS Nano, 2013, 7, 2541–2550 CrossRef CAS PubMed .
  16. J. F. Lovell, C. S. Jin, E. Huynh, H. L. Jin, C. H. Kim, J. L. Rubinstein, W. C. Chan, W. G. Cao, L. H. V. Wang and G. Zheng, Nat. Mater., 2011, 10, 324–332 CrossRef CAS PubMed .
  17. L. V. Nair, S. S. Nazeer, R. S. Jayasree and A. Ajayaghosh, ACS Nano, 2015, 9, 5825–5832 CrossRef CAS PubMed .
  18. B. Jang, J. Y. Park, C. H. Tung, I. H. Kim and Y. D. Choi, ACS Nano, 2011, 5, 1086–1094 CrossRef CAS PubMed .
  19. T. Simon, M. Potara, A. M. Gabudean, E. Licarete, M. Banciu and S. Astilean, ACS Appl. Mater. Interfaces, 2015, 7, 16191–16201 CAS .
  20. S. J. Wang, P. Huang, L. M. Nie, R. J. Xing, D. B. Liu, Z. Wang, J. Lin, S. H. Chen, G. Niu, G. M. Lu and X. Y. Chen, Adv. Mater., 2013, 25, 3055–3061 CrossRef CAS PubMed .
  21. S. J. Lee, H. Koo, H. Jeong, M. S. Huh, Y. Choi, S. Y. Jeong, Y. Byun, K. Choi, K. Kim and I. C. Kwon, J. Controlled Release, 2011, 152, 21–29 CrossRef CAS PubMed .
  22. S. J. Lee, H. Koo, D. E. Lee, S. Min, S. Lee, X. Y. Chen, Y. Choi, J. F. Leary, K. Park, S. Y. Jeong, I. C. Kwon, K. Kim and K. Choi, Biomaterials, 2011, 32, 4021–4029 CrossRef CAS PubMed .
  23. Q. Chen, X. Wang, C. Wang, L. Z. Feng, Y. G. Li and Z. Liu, ACS Nano, 2015, 9, 5223–5233 CrossRef CAS PubMed .
  24. X. J. Song, C. Liang, H. Gong, Q. Chen, C. Wang and Z. Liu, Small, 2015, 11, 3932–3941 CrossRef CAS PubMed .
  25. Y. Y. Li, T. Wen, R. F. Zhao, X. X. Liu, T. J. Ji, H. Wang, X. W. Shi, J. Shi, J. Y. Wei, Y. L. Zhao, X. C. Wu and G. J. Nie, ACS Nano, 2014, 8, 11529–11542 CrossRef CAS PubMed .
  26. X. Ma, Q. Y. Qu and Y. L. Zhao, ACS Appl. Mater. Interfaces, 2015, 7, 10671–10676 CAS .
  27. B. Tian, C. Wang, S. Zhang, L. Z. Feng and Z. Liu, ACS Nano, 2011, 5, 7000–7009 CrossRef CAS PubMed .
  28. S. L. Luo, Z. Y. Yang, X. Tan, Y. Wang, Y. P. Zeng, Y. Wang, C. M. Li, R. Li and C. M. Shi, ACS Appl. Mater. Interfaces, 2016, 8, 17176–17186 CAS .
  29. S. S. Lucky, N. M. Idris, Z. Q. Li, K. Huang, K. C. Soo and Y. Zhang, ACS Nano, 2015, 9, 191–205 CrossRef CAS PubMed .
  30. L. E. Liang, A. Care, R. Zhang, Y. Q. Lu, N. H. Packer, A. Sunna, Y. Qian and A. V. Zvyagin, ACS Appl. Mater. Interfaces, 2016, 8, 11945–11953 CAS .
  31. L. E. Liang, Y. Q. Lu, R. Zhang, A. Care, T. A. Ortega, S. M. Deyev, Y. Qian and A. V. Zvyagin, Acta Biomater., 2017, 51, 461–470 CrossRef CAS PubMed .
  32. S. Beack, W. H. Kong, H. S. Jung, I. H. Do, S. Han, H. Kim, K. S. Kim, S. H. Yun and S. K. Hahn, Acta Biomater., 2015, 26, 295–305 CrossRef CAS PubMed .
  33. P. Huang, J. Lin, X. S. Wang, Z. Wang, C. L. Zhang, M. He, K. Wang, F. Chen, Z. M. Li, G. X. Shen, D. X. Cui and X. Y. Chen, Adv. Mater., 2012, 24, 5104–5110 CrossRef CAS PubMed .
  34. A. S. Garcez, M. S. Ribeiro, G. P. Tegos, S. C. Núñez, A. O. Jorge and M. R. Hamblin, Lasers Surg. Med., 2007, 39, 59–66 CrossRef PubMed .
  35. H. Y. Yoon, H. Koo, K. Y. Choi, S. J. Lee, K. Kim, I. C. Kwon, J. F. Leary, K. Park, S. H. Yuk, J. H. Park and K. Choi, Biomaterials, 2012, 33, 3980–3989 CrossRef CAS PubMed .
  36. K. Liu, X. M. Liu, Q. H. Zeng, Y. L. Zhang, L. P. Tu, T. Liu, X. G. Kong, Y. H. Wang, F. Cao, S. A. G. Lambrechts, M. C. G. Aalders and H. Zhang, ACS Nano, 2012, 6, 4054–4062 CrossRef CAS PubMed .
  37. D. Zhang, M. Wu, Y. Y. Zeng, L. J. Wu, Q. T. Wang, X. Han, X. L. Liu and J. F. Liu, ACS Appl. Mater. Interfaces, 2015, 7, 8176–8187 CAS .
  38. H. Ghanbari, B. G. Cousins and A. M. Seifalian, Macromol. Rapid Commun., 2011, 32, 1032–1046 CrossRef CAS PubMed .
  39. C. Sanchez, G. J. de A. A. Soler-Illia, F. Ribot, T. Lalot, C. R. Mayer and V. Cabuil, Chem. Mater., 2001, 13, 3061–3083 CrossRef CAS .
  40. J. Han, Y. C. Zheng, S. Zheng, S. P. Li, T. N. Hu, A. J. Tang and C. Gao, Chem. Commun., 2014, 50, 8712–8714 RSC .
  41. X. Zhao, J. Du, Y. Z. Wu, H. Z. Liu and X. P. Hao, J. Mater. Chem. A, 2013, 1, 11748–11753 CAS .
  42. Y. He, H. F. Wang and X. P. Yan, Chem. – Eur. J., 2009, 15, 5436–5440 CrossRef CAS PubMed .
  43. B. L. Frankamp, N. O. Fischer, R. Hong, S. Srivastava and V. M. Rotello, Chem. Mater., 2006, 18, 956–959 CrossRef CAS .
  44. T. L. Kaneshiro and Z. R. Lu, Biomaterials, 2009, 30, 5660–5666 CrossRef CAS PubMed .
  45. Y. J. Pu, S. Chang, H. Yuan, G. Wang, B. He and Z. W. Gu, Biomaterials, 2013, 34, 3658–3666 CrossRef CAS PubMed .
  46. Y. J. Pu, L. G. Zhang, H. Zheng, B. He and Z. W. Gu, Polym. Chem., 2014, 5, 463–470 RSC .
  47. D. B. Cordes, P. D. Lickiss and F. Rataboul, Chem. Rev., 2010, 110, 2081–2173 CrossRef CAS PubMed .
  48. F. F. Du, J. Tian, H. Wang, B. Liu, B. K. Jin and R. Bai, Macromolecules, 2012, 45, 3086–3093 CrossRef CAS .
  49. K. O. Kim, B. S. Kim and I. S. Kim, J. Biomater. Nanobiotechnol., 2011, 2, 201–206 CrossRef CAS .
  50. K. Y. Pu, K. Li, X. H. Zhang and B. Liu, Adv. Mater., 2010, 22, 4186–4189 CrossRef CAS PubMed .
  51. M. Q. Tan, Z. Ye, E. K. Jeong, X. M. Wu, D. L. Parker and Z. R. Lu, Bioconjugate Chem., 2011, 22, 931–937 CrossRef CAS PubMed .
  52. C. McCusker, J. B. Carrollb and V. M. Rotello, Chem. Commun., 2005, 996–998 RSC .
  53. K. Li, Y. T. Liu, K. Y. Pu, S. S. Feng, R. Y. Zhan and B. Liu, Adv. Funct. Mater., 2011, 21, 287–294 CrossRef CAS .
  54. K. Y. Pu, K. Li and B. Liu, Adv. Mater., 2010, 22, 643–646 CrossRef CAS PubMed .
  55. M. H. Xie, J. Ge, B. Lei, Q. Zhang, X. F. Chen and P. X. Ma, Macromol. Biosci., 2015, 15, 1656–1662 CrossRef CAS PubMed .
  56. M. C. Wu, B. H. Ma, T. Z. Pan, S. S. Chen and J. Q. Sun, Adv. Funct. Mater., 2016, 26, 569–576 CrossRef CAS .
  57. M. Aflori, B. Simionescu, I. E. Bordianu, L. Sacarescu, C. D. Varganici, F. Doroftei, A. Nicolescu and M. Olaru, Mater. Sci. Eng., B, 2013, 178, 1339–1346 CrossRef CAS .
  58. I. E. Bordianu, G. David, B. Simionescu, M. Aflori, C. Ursu, A. Coroaba, G. Hitruc, C. Cotofana and M. Olaru, J. Mater. Chem. B, 2015, 3, 723–727 RSC .
  59. Q. Q. Yang, L. Li, W. Sun, Z. Zhou and Y. Huang, ACS Appl. Mater. Interfaces, 2016, 8, 13251–13261 CAS .
  60. Y. Y. Yang, X. Wang, Y. Hu, H. Hu, D. C. Wu and F. J. Xu, ACS Appl. Mater. Interfaces, 2014, 6, 1044–1052 CAS .
  61. X. J. Cai, R. R. Jin, J. L. Wang, D. Yue, Q. Jiang, Y. Wu and Z. W. Gu, ACS Appl. Mater. Interfaces, 2016, 8, 5821–5832 CAS .
  62. X. Zhang, Z. J. Zhang, X. H. Xu, Y. K. Li, Y. C. Li, Y. T. Jian and Z. W. Gu, Angew. Chem., Int. Ed., 2015, 54, 4289–4294 CrossRef CAS PubMed .
  63. A. Janaszewska, K. Gradzinska, M. Marcinkowska, B. Klajnert-Maculewicz and W. A. Stanczyk, Materials, 2015, 8, 6062–6070 CrossRef .
  64. H. I. Lee and Y. J. Kim, Colloids Surf., B, 2016, 142, 182–191 CrossRef CAS PubMed .
  65. T. Mori, T. Kubo, K. Kaya and K. Hosoya, Colloid Polym. Sci., 2009, 287, 513–523 CAS .
  66. X. Y. Zhou, Y. Luo, Y. M. Zhu, Z. H. Liu, T. A. Kent, J. G. Rong, W. Li, S. G. Qiao, M. Li, Y. Ni, K. Ishidoh and H. L. Zhang, Cell Death Dis., 2017, 8, e2618 CrossRef CAS PubMed .
  67. K. Y. Zheng, M. I. Setyawati, T. P. Lim, D. T. Leong and J. P. Xie, ACS Nano, 2016, 10, 7934–7942 CrossRef CAS PubMed .
  68. J. W. Snyder, E. Skovsen, J. D. C. Lambert, L. Poulsen and P. R. Ogilby, Phys. Chem. Chem. Phys., 2006, 8, 4280–4293 RSC .
  69. T. J. Dougherty, C. J. Gomer, B. W. Henderson, G. Jori, D. Kessel, M. Korbelik, J. Moan and Q. Peng, J. Natl. Cancer Inst., 1998, 90, 889–905 CrossRef CAS PubMed .
  70. Z. Meng, B. Yu, G. Y. Han, M. H. Liu, B. Shan, G. Q. Dong, Z. Y. Miao, N. Y. Jia, Z. Tan, B. H. Li, W. N. Zhang, H. Y. Zhu, C. Q. Sheng and J. Z. Yao, J. Med. Chem., 2016, 59, 4999–5010 CrossRef CAS PubMed .


Electronic supplementary information (ESI) available: Additional figures. See DOI: 10.1039/c7nr02279d
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2017