Photothermo-chemotherapy of cancer employing drug leakage-free gold nanoshells

Biomaterials 78 (2016) 40e49

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Photothermo-chemotherapy of cancer employing drug leakage-free gold nanoshells Lu Wang a, c, 1, Yuanyuan Yuan a, 1, Shudong Lin a, 1, Jinsheng Huang a, Jian Dai a, c, Qing Jiang c, Du Cheng a, *, Xintao Shuai a, b, ** a b c

PCFM Lab of Ministry of Education, School of Chemistry and Chemical Engineering, Sun Yat-sen University, Guangzhou, 510275, China Center of Biomedical Engineering, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, 510080, China Department of Biomedical Engineering, School of Engineering, Sun Yat-sen University, Guangzhou, Guangdong, 510006, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 October 2015 Received in revised form 11 November 2015 Accepted 12 November 2015 Available online 14 November 2015

Combined photothermo-chemotherapy is a new cancer treatment modality that improves therapeutic outcome by synergistic actions of two different means. A reduction and pH dual sensitive polymeric vesicle encapsulating doxorubicin (DOX) was prepared and then decorated with a gold layer using a modified method of in situ gold seed growth. By tuning the compactness of gold layer, the gold nanoshell may possess a desirable light absorption peak for tumor photothermal therapy using near-infrared (NIR) laser irradiation, a method featuring high tissue penetrability essential for in vivo applications. The NIR light energy was converted into heat, which killed cancer cells in the vicinity and induced the rupture of nanoshell to release DOX inside tumor. Therefore, a combined photothermo-chemotherapy of tumor can be achieved precisely at tumor site. In addition, DOX released in the thermochemotherapeutic mode effectively penetrated tumor tissue, which is meaningful considering the intrinsic low tissue penetrability of nanomedicines. In nude mice bearing human Bel-7402 hepatoma, the photothermochemotherapy using DOX-loaded gold nanoshell appeared advantageous over a chemotherapy or a photothermal therapy alone. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Polymeric vesicles Gold nanoshells Near-infrared laser irradiation Triggered release Photothermo-chemotherapy

1. Introduction Chemotherapy as one of the three major approaches for clinic cancer treatment often fails due to the insufficient tumor killing and severe systemic side effects resulting from the inevitable drug resistance and nonspecific action on both the tumor and normal cells. Therefore, novel therapeutic means which may overcome the intrinsic shortages of traditional chemotherapy have drawn great attentions in recent years. For example, noninvasive photothermal therapy based on gold nanomaterials has been intensively investigated for improving cancer treatment [1,2]. The well developed blood vessels in normal human tissues can act as temperature-regulating system. The local blood vessels

* Corresponding author. ** Corresponding author. PCFM Lab of Ministry of Education, School of Chemistry and Chemical Engineering, Sun Yat-sen University, Guangzhou, 510275, China. E-mail addresses: [email protected] (D. Cheng), [email protected]. cn (X. Shuai). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.biomaterials.2015.11.024 0142-9612/© 2015 Elsevier Ltd. All rights reserved.

expand in normal tissues under hyperthermia, accelerating blood flow to take heat away and thus preventing tissue from hyperthermal damage. However, blood flow in tumor generally slows down due to vascular agglomeration, distortion and expansion. When the tumor tissue was heated, its temperature can be 5 to 10
C higher than that of the adjacent normal tissue, resulting in thermal ablation of tumor without affecting normal tissues. The retarded heat flow resulting from the abnormal vasculature provides a great opportunity for tumor thermotherapy using gold nanomaterials. Various studies have shown that the light absorption peak of conventional gold nanocrystals around 520 nm can be shifted toward the absorption region (650e900 nm) of nearinfrared (NIR) light by adjusting their morphology and structure [3,4]. Consequently, optical energy from NIR irradiation can be converted into thermal energy by the nanomaterials based on gold nanocrystals. More importantly, the NIR absorption by hemoglobin and water is fairly weak [5], making NIR irradiation highly penetrative but not harmful to human body. In addition, gold nanoparticles have very low cytotoxicity. Up to now, various gold nanomaterials including gold nanorods

L. Wang et al. / Biomaterials 78 (2016) 40e49

[6], gold nanoshells [7] and gold nanocages [8] with absorption in the NIR region have been developed to show broad application potentials in cancer treatment. In particular, gold nanoshells have strong absorption in the NIR region and good photothermal conversion efficiency [9]. They are generally fabricated using the template method of introducing preformed gold nanoparticles onto the surfaces of polymeric micelles, silicon spheres and liposomes [7,10,11]. Given that anticancer drugs were incorporated, the gold nanoshells may act as a multifunctional platform for combined photothermo-chemotherapy. However, the gold layer formed on the template surface is not microstructurally continuous and thus not drug leakage-free. Consequently, drug loss cannot be avoided when the drug-loaded gold nanoshells circulate in bloodstream before reaching the targets. In comparison, in situ growth of a continuous gold layer on the nanocarrier surface via chemical reduction of Au(Ⅲ) appears more desirable since it may yield leaktight gold nanomedicines. Unlike the former method which uses the preformed gold nanocrystals with NIR absorptions to decorate the nanomedicines, the in situ growth method requires a much more careful control of the gold shell microstructure (e.g. shell thickness, density) for NIR light absorption. On the other hand, effective photothermo-chemotherapy of cancer also requires that the drug is quickly released when the gold nanoshell is broken by NIR irradiation at the tumor site. Unfortunately, the preparation of gold nanomaterials with strong NIR absorption and meanwhile desired drug release profiles is highly challenging and rarely reported thus far. In the present study, we aimed to develop a leak-tight gold nanoshell incorporating anticancer drug for NIR light irradiationtriggered photothermo-chemotherapy of cancer. To this end, we first prepared a pH and reduction dual sensitive polymeric vesicle whose inner aqueous core was loaded with the anticancer drug doxorubicin (DOX). Then, a leak-tight gold layer with good NIR light-to-heat conversion was grown on the vesicular surface via the in situ Au(Ⅲ) reduction approach. It is expected that the DOXloaded gold nanoshell may avoid drug loss in bloodstream, and meanwhile the pH and reduction dual sensitivity may allow the vesicle to quickly release DOX by responding to the intratumoral/ intracellular microenvironments after the gold layer was broken by a NIR light irradiation. Consequently, a NIR light-triggered photothermo-chemotherapy may be achieved precisely at the tumor site to improve the therapeutic efficacy and meanwhile to lower side effects. To demonstrate this potential, the drug delivery efficiency and photothermo-chemotherapeutic effect of gold nanoshell administered through intravenous injection was evaluated in nude mice bearing human Bel-7402 hepatoma xenografts.

41

diphenyl tetrazolium bromide (MTT) was obtained from SigmaeAldrich. Annexin V-APC/7-AAD double labeling kit was obtained from BD Bioscience (5 mL per test, San Jose, CA, USA). 2.2. Preparation of DOX-loaded polymeric vesicles (DOX-loaded PNV) In brief, 10 mg of PEI-PAsp(DIP/MEA) and DOX$HCl (0.5 mg, 1.0 mg, 2.0 mg or 3.0 mg according to experimental design) were co-dissolved in 1 mL of DMSO. The solution was then emulsified by sonication in 10 mL of Phosphate Buffered Saline (PBS, pH 8 ～ 9) in an ice bath. The mixture was stirred under bubbling of an oxygen flow for 1 h to form the shell-crosslinked vesicles. Subsequently, the solution was dialyzed (MWCO: 14 kDa) against PBS (pH 7.4) for 3 days to remove free DOX. Finally, the solution was filtered through a 0.45 mm filter to remove large aggregates. 2.3. Preparation of DOX-loaded gold nanoshell (DOX-loaded GNS@PNV) As outlined in Fig. 1, the DOX-loaded gold nanoshell was prepared using the DOX-loaded PNV as template. The reactions were carried out under nitrogen and away from light [13]. First, the gold seeds were grown on the surface of the polymeric vesicles using PEI as a reducing agent. The polymeric vesicle solution (1 mg/mL, 1 mL) was diluted to 0.1 mg/mL with deionization water, followed by addition of 100 mL aqueous solution of HAuCl4 (10 mg/mL) whose pH had been adjusted to ~9 with NaOH solution. The reaction proceeded for 6 h to obtain gold seeds-decorated vesicles. Then, 10 mL of the above solution was added with 100e2500 mL of NH2OH$HCl (40 mM) and 20e500 mL of 10 mg/mL HAuCl4 solution at different Mcopolymer/MAu values of 1/0.1, 1/0.25, 1/0.5, 1/1, and 1/ 2.5 according to experimental design. The reaction was conducted for 15 min at 4
C, after which the particles were purified by centrifugation and then washed four times with pure water. PEG5kSH was added to enhance the colloidal solution stability [14]. Finally, five DOX-loaded gold nanoshells denoted as GNS@PNV-1, GNS@PNV-2, GNS@PNV-3, GNS@PNV-4 and GNS@PNV-5 respectively were obtained. DOX-free GNS@PNV was prepared by

2. Experimental section 2.1. Materials The diblock copolymer, polyethylenimine-b-poly(2diisopropylamino/2-mercaptoethylamine) ethyl aspartate denoted as PEI-PAsp(DIP/MEA), was synthesized as previously reported [12]. HAuCl4 (99.9%, SigmaeAldrich, St. Louis, MO, USA) and hydroxylamine solution (50 wt.% in H2O) (SigmaeAldrich, St. Louis, MO, USA) of analytical grade were used as received. Dimethylsulfoxide (DMSO) from SigmaeAldrich (St. Louis, MO, USA) was dried over CaH2 and distilled. Doxorubicin hydrochloride (DOX$HCl) was purchased from Zhejiang Hisun Pharmaceutical Co., Ltd., China and used as received. Human hepatoma Bel-7402 cells were purchased from the Experimental Animal Center of Sun Yat-sen University (Guangzhou, China). Cell culture media, penicillin-streptomycin, fetal bovine serum (FBS) and 0.25% trypsin were purchased from Gibco BRL (Carlsbad, CA, USA). (3-(4,5-Dimethyl-thiazol-2-yl)-2,5-

Fig. 1. The schematic diagram of the preparation of DOX-loaded polymeric vesicle (PNV) and gold nanoshell (GNS@PNV).

42

L. Wang et al. / Biomaterials 78 (2016) 40e49

similar method using the DOX-free PEI-PAsp(DIP/MEA) vesicle as template. 2.4. Characterization The transmission electron microscopy (TEM) observation of vesicles was conducted on a Philips CM120 transmission electron microscope (Philips, Eindhoven, the Netherlands) operated at an accelerating voltage of 100 kV. Samples were prepared by drying a drop (10 mL, 1 mg/mL) of the sample solution on a copper grid coated with amorphous carbon. To stain the sample, a small drop of uranyl acetate solution (2 wt% in water) was applied to the copper grid, and then blotted off with a filter paper after 1 min. The grid was finally dried overnight in a desiccator before TEM observation. The particle size was determined with dynamic light scattering (DLS). The nanoparticle solutions were filtered through a 450 nm filter. Measurements were carried out at 25
C on a 90 Plus/BI-MAS equipment (Brookhaven Instruments Corporation, Holtsville, NY, USA). Data were collected on an autocorrelator with a 90
detection angle of scattered light. The measurement result was expressed as the mean ± standard deviation (SD) of five independent measurements. To measure the fluorescence spectra (FLS), 3 mL of DOX-loaded GNS@PNV solution (DOX loading content: 2.23%) was evenly divided into three groups with solution conditions as follows: pH 7.4, pH 5.0 þ 10 mM GSH, and pH 5.0 þ 10 mM GSH þ laser. The 808 nm NIR laser was applied at a power level of 1.5 W/cm2 for 30 min. Afterwards, the solutions underwent ultrafiltration (MWCO ¼ 10 KDa) to obtain the filtrates. Each filtrate was then diluted to 5 mL with water of the same pH value for fluorescence spectrum measurement on a UV-3150 Spectrofluorophotometer (Excitation: 495 nm; emission: 510e750 nm; Slit width: 10 nm; Shimadzu, Kyoto, Japan). The UVevis absorption spectroscopy of GNS@PNV was measured on a UVevis spectrophotometer (Lambda750, PerkinElmer, Waltham, MA, USA).

antibiotics (penicillin/streptomycin) were cultured in a humidified atmosphere of 5% CO2 at 37
C. When the cells reached 80%~90% confluence, they were trypsinized using 0.25% trysin and subcultured. 2.8. Cell viability The experiments were performed in triplicates. Bel-7402 cells were seeded in 96-well plates with a density of 1  104 cells per well. After RPMI 1640 medium was added, the cells were cultured overnight in a humidified atmosphere of 5% CO2 at 37
C. The cells were then incubated in the presence of gold-free vesicles at different concentrations for 24 h. For the negative control, the cell culture medium was not added with vesicles. The medium in each well was replaced with 100 mL of fresh medium plus 10 mL of PBS containing MTT (5 mg/mL). After the cells were further incubated at 37
C for 4 h, 100 mL of DMSO was added to replace the MTTcontaining medium. After gentle agitation for 5 min, the absorbance at 570 nm for each well was recorded on an Infinite® F200 Multimode plate reader (Tecan, Crailsheim, Germany). The cell viabilities were calculated by comparison with the absorbance of negative control. The data were expressed as means ± standard deviations (SD). To evaluate the cytotoxic effect of DOX-loaded gold nanoshell under NIR irradiation, 1  104 Bel-7402 cells were seeded in 96-well plates and cultured overnight in a humidified atmosphere of 5% CO2 at 37
C. The gold nanoshell was added into the culture medium to incubate the cells for 6 h. The cell culture medium was replaced with fresh medium, and then the cells were irradiated with NIR laser at a power level of 1.5 W/cm2 for 2 min (2 min intervals, repeated by 5 cycles). The cells were further incubated for 24 h and then underwent the above mentioned MTT assay to quantify the survival rate. 2.9. Cell apoptosis

A series of 1 mL of GNS@PNV solutions with different vesicular concentrations were placed in 4 mL glass bottles, irradiated with near-infrared laser of 808 nm wavelength, photographed on an infrared thermal imaging apparatus (FLIR System SC300, Boston, MA, USA). Meanwhile, the solution temperatures were collected using thermocouples. PBS solution was used as the control. The power levels of near-infrared laser were 0, 0.7, 1.5, 3 W/cm2, respectively.

Bel-7402 cell apoptosis levels induced by different samples were evaluated by flow cytometry (FACSCalibur, BD Bioscience, San Jose, CA, USA) using Annexin V-APC/7-AAD double labeling. Bel-7402 cells were seeded in the 6-well plates (1  105 cells/well) and then cultured in RPMI 1640 medium overnight in a humidified atmosphere of 5% CO2 at 37
C. The vesicle samples were added into the culture medium to incubate the cells for 6 h. After the culture medium was replaced with fresh medium, the cells were irradiated by 808 nm laser at a power level of 1.5 W/cm2 for 2 min (2 min intervals, repeated by 5 cycles) and then cultured for another 24 h. The cells were digested with EDTA-free trypsin, washed two times with PBS, collected by centrifugation and resuspended in 500 mL binding buffer. After Annexin V-APC and 7-AAD (5 mL each) were added separately, the evenly mixed solution was reacted in dark for 15 min at room temperature and then detected by flow cytometry. The red fluorescence of Annexin V-APC was detected by FL4 channel with the excitation wavelength of 633 nm and emission wavelength of 660 nm, respectively. The red fluorescence of 7-AAD was detected by FL3 channel with the excitation wavelength of 546 nm and emission wavelength of 647 nm, respectively. The normally cultured cells were used as control group for background correction. The experiments were performed in triplicates and analyzed using Kaluza analysis software for flow cytometry.

2.7. Cell culture

2.10. Animal model

The human hepatoma Bel-7402 cells in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% double

Subcutaneous tumor model was constructed in 6e7 week old nu/nu nude mice. Briefly, the mice were anesthetized with

2.5. DOX loading content of gold nanoshells The DOX loading content was measured by detecting the UV absorbance at 480 nm on a Unico UV-2000 UVevis spectrophotometer. The DOX-loaded GNS@PNV was dispersed in PBS (pH 5.0) added with 10 mM GSH and then exposed to 808 nm laser for 30 min at a power level of 1.5 W/cm2 (BWT Beijing Ltd, Beijing, China). The solution was subjected to ultrafiltration. The absorbance of filtrate was measured at 480 nm to determine the drug concentration using a previously established calibration curve. Drug loading content in GNS@PNV was defined as the weight percentage of DOX in gold nanoshell. 2.6. Photothermal conversion of gold nanoshells

L. Wang et al. / Biomaterials 78 (2016) 40e49

ketamine (20 mg/kg) and placed in a stereotactic frame. Human Bel-7402 cells (1  106) in 200 mL of serum-free RPMI 1640 medium were subcutaneously injected in upper thighs. When tumor grew to 100 mm3, nude mice bearing the Bel-7402 xenografts were randomly divided into six groups (n ¼ 10) and treatments were started. Studies involving animals were approved by animal care and use committee of Sun Yat-Sen University.

43

power level of 1.5 W/cm2 for 2 min (2 min intervals, repeated by 5 cycles) whereas tumor on the left-side groin was un-irradiated for comparison. Then, the in vivo fluorescence imaging was recorded at different time points. The mice were sacrificed at the time point of 200 min after irradiation and the interested organs were photographed using fluorescence imaging system and infrared thermal imager, respectively. DOX fluorescence excitation: 530 nm; emission: 600 nm.

2.11. Tumor accumulation and photothermal conversion 100 mL of DOX-loaded GNS@PNV solution (in PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL) was injected into nude mice bearing tumor via the tail vein. The tumor sites were irradiated by 808 nm laser for 2 min (2 min intervals, repeated by 5 cycles) using a power level of 1.5 W/cm2 at the time points of 0, 3, 6, 9, 12, 16, and 18 h. At 1 h after irradiation, the mice under anaesthesia were imaged for DOX fluorescence (excitation: 530 nm; emission: 600 nm) on a small animal in vivo fluorescence imaging system (Carestream In-Vivo Imaging System FXPRO, Woodbridge, CT, USA). 2.12. NIR laser-triggered DOX release in tail vein injection mode Nude mice bearing two identical Bel-7402 xenografts in the groins were used in the study for easy imaging. 200 mL of serumfree suspension containing 1  106 Bel-7402 cells was subcutaneously injected into the anulus inguinalis superficialis at both sides of nude mice for tumor formation. After tumor grew to 200 mm3, the NIR laser irradiation-triggered intratumoral DOX release from GNS@PNV was investigated. In brief, 100 mL solution of DOX-loaded GNS@PNV (in PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL) was injected into mice bearing two tumors via tail vein. After 12 h, tumor on the right-side groin (leftside tumor in Fig. 6 and Fig. S9) was irradiated by 808 nm laser at a

2.13. NIR laser-triggered DOX release in intratumoral injection mode After the nude mice received intratumoral injection of DOXloaded GNS@PNV solution (100 mL, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL), the tumor site was irradiated immediately by 808 nm laser at a power level of 1.5 W/cm2. The tumor sites were examined by the in vivo fluorescence imaging system and infrared thermal imager at different time points to reveal the distribution of DOX and gold in the tumor. DOX fluorescence excitation: 530 nm; emission: 600 nm. 2.14. Anticancer effect of NIR irradiation-triggered photothermochemotherapy The nude mice bearing human Bel-7402 xenografts were randomly divided into 6 groups (n ¼ 10) to receive different treatments as follows: DOX-loaded GNS@PNV plus laser irradiation (DOX-loaded GNS@PNV þ laser), DOX-free GNS@PNV plus laser irradiation (GNS@PNV þ laser), free DOX, DOX-loaded GNS@PNV without laser irradiation (DOX-loaded GNS@PNV), PBS plus laser irradiation (PBS þ laser) and PBS only as control group. 200 mL sample solution was injected into the nude mice (dose per injection: 4.5 mg DOX per Kg body weight if required) via tail vein at the time points of 1, 3, 5 days, respectively. At 24 h after injection, the

Fig. 2. Transmission electron microscopy (TEM) images of a1) gold-free PNV at pH 7.4, a2) GS@PNV, a3) GNS@PNV-3, a4) GNS@PNV-5, a5) GNS@PNV-5 illuminated by a NIR laser (808 nm) at a power density of 1.5 W/cm2 for 10 min (GNS@PNV-3 and GNS@PNV-5 only vary in the surficial gold density, see experiment for detail), b) the fluorescence spectrum of the DOX-loaded GNS@PNV.

44

L. Wang et al. / Biomaterials 78 (2016) 40e49

Fig. 3. a) Cell viabilities at various PNV and Au concentrations. b) Effect of irradiation time of 808 nm laser on the cytotoxicity of GNS@PNV in Bel-7402 cells. c) Effect of Au concentration on the cytotoxicity of GNS@PNV in Bel-7402 cells in the presence of 808 nm laser. d) Cytotoxicity of the DOX-loaded gold nanoshell depending on both the 808 nm laser irradiation and DOX concentration. Laser irradiation: for 2 min at 1.5 W/cm2, 2 min intervals, repeated by 5 cycles. *P < 0.05 vs DOX-loaded GNS@PNV and DOX-loaded PNV, # P < 0.05 vs DOX-loaded GNS@PNV.

tumors were irradiated by 808 nm laser for 2 min (2 min intervals, repeated by 5 cycles) using a power density of 1.5 W/cm2. The treatment lasted for twenty days. The tumor size was recorded every two days. The tumor volume was measured with vernier caliper and calculated using the following equation: Volume ¼ 0.5  l  w2, where “w” and “l” are width and length of the tumor. The nude mice after treatment were killed by cervical dislocation. In situ histological and TUNEL assays of the tumor tissue sections were carried out as described in recent report [15].

Fig. 4. Detection of apoptotic Bel-7402 cells induced by DOX-loaded gold nanoshell using Annexin V and APC flow cytometry assay. The cell were treated with gold nanoshell containing 22 mM DOX and 808 nm laser irradiation. The cells were treated with various nanoparticle containing 22 mM DOX. Laser irradiation: for 2 min at 1.5 W/ cm2, 2 min intervals, repeated by 5 cycles. *P < 0.05 vs DOX-loaded GNS@PNV and DOX-loaded PNV.

3. Results and discussion 3.1. Preparation, structure and physicochemical properties of gold nanoshells The morphologies and nanostructures of the PNVs based on PEI-PAsp(DIP/MEA) were characterized by TEM and dynamic light scattering. As shown in Fig. 2-a1, the shell-crosslinked vesicles before gold decoration were spherical with clear hydrophobic membrane and uniform size distribution around 100 nm at pH 7.4. When the vesicle solution was adjusted to pH 5.0, the PAsp(DIP) segments turned to hydrophilic. Nevertheless, the interchain crosslinking via disulfide bond formation effectively prevented the vesicles from disassembly. At pH 5.0, the vesicles swelled to about 200 nm. Furthermore, typical caved-in structure resulting from the drying process of vesicle was observed (Fig. S1a). After adding 10 mM glutathione (GSH) into the solution at pH 5.0, the disulfide bond was broken, resulting in vesicle disassembly. Consequently, random polymeric aggregates were formed when drying the solution for TEM measurement (Fig. S1b). Measurement using dynamic light scattering is supportive of the TEM results. The particle sizes of the vesicle in solution were 86.6 ± 15 nm at pH 7.4, 210.5 ± 43 nm at pH 5.0, and undetectable at pH 5.0 plus GSH (10 mM) due to vesicle disassembly. Conventional preparation method of gold nanoshell by introducing preformed gold nanoparticles onto the surface of drug carrier via electrostatic interaction results in drug-leaky nanomedicines. The preparation of gold nanoshell in our study features in situ growth of gold seed on the surface of PNVs, which was

L. Wang et al. / Biomaterials 78 (2016) 40e49

45

Fig. 5. a) In vivo fluorescent imaging of nude mice after tail vein injection of DOX-loaded gold nanoshell (100 mL PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL) and tumor site irradiation with 808 nm laser at various post-injection time. Laser irradiation power level was 1.5 W/cm2, and DOX was allowed to diffuse for 1 h after laser irradiation of tumor. b) In vivo fluorescent imaging of nude mice receiving intratumor injection of DOX-loaded GNS@PNV (100 mL PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL). Right after injection, the tumor site was exposed to 808 nm laser and then the mice were imaged at various post-irradiation time points. c) In vivo infrared thermogram at different time points after intratumor injection of gold nanoshell solution (100 mL PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL) and then irradiated with 808 nm NIR laser. Laser irradiation: for 2 min at 1.5 W/cm2, 2 min intervals, repeated by 5 cycles.

expected to obtain leak-tight drug carrier. The polymeric vesicles first acted as templates for directly growing gold nanoseeds on their surface because polyethylenimine (PEI) on the surface can reduce Au(III) to form gold nanoseeds (GS@PNV) without aggregation (Fig. 2-a2) [16]. Then, further reduction of Au(III) on GS@PNV using hydroxylamine as a weak reducing agent resulted in the formation of a continuous gold nanoshell, i.e. the GNS@PNV (Fig. 2a3 and -a4). The weight percentage of gold layer is 77.68% in GNS@PNV-5. Through this method, the light absorption peak of GNS@PNV was successfully moved to the near-infrared region by varying the HAuCl4 concentrations for sample preparation. The light absorption peak of GS@PNV is around 510 nm (Fig. S2a). However, it was moved above 800 nm when the gold seeds grew to a dense layer of gold nanoshell on the surface of vesicles (Fig. S2b).

It is noteworthy that the near-infrared light has strong tissue penetrability, which is extremely favorable for the in vivo applications of a light irradiation-responsive nanomedicine [17]. The nearinfrared laser irradiation was reported to make the gold nanoshell melt and rupture [18], a property that can be utilized to enable photothermal therapy and to trigger drug release. Therefore, the gold nanoshell with near-infrared absorption peak at 808 nm (i.e. GNS@PNV-5) was used in the following experiments, and GNS@PNV refers to GNS@PNV-5 unless otherwise indicated. The rupture of GNS@PNV under the 808 nm laser radiation (1.5 W/cm2) was studied by TEM (Fig. 2-a5). GNS@PNV was coated with PEG by reacting Au with thiolated PEG to enhance its stability in aqueous solution. As shown in Fig. S3, the particle size of GNS@PNV maintained stable in PBS for at least two months. Furthermore, PEG

46

L. Wang et al. / Biomaterials 78 (2016) 40e49

Fig. 6. a) In vivo fluorescent imaging of nude mice at various post-irradiation time. Only the L tumor on the groin was irradiated with 808 nm laser 12 h after tail vein injection of DOX-loaded GNS@PNV (100 mL PBS, C[DOX] ¼ 450 mg/mL, C[GNS@PNV] ¼ 19.55 mg/mL). b) Ex vivo DOX fluorography and c) infrared thermogram showing the DOX bio-distribution in nude mice bearing Bel-7402 tumor at 200 min post-irradiation. Only the L tumor on the groin of mice was irradiated with 808 nm laser 12 h after tail vein injection of DOX-loaded GNS@PNV. Laser irradiation: for 2 min at 1.5 W/cm2, 2 min intervals, repeated by 5 cycles. L tumor means left-side tumor on the right-side groin and R tumor means right-side tumor on the left-side groin.

modification is potentially able to weaken the nonspecific adsorption of serum proteins for prolonged circulation time of GNS@PNV in bloodstream [19,20]. 3.2. DOX release from gold nanoshells Free DOX has high toxicity to heart tissue, which greatly limits its clinical applications [21]. In this study, DOX was encapsulated in the hydrophilic cavity of PEI-PAsp(DIP/MEA) vesicle decorated with a dense and continuous gold layer to prevent DOX from leaking in blood circulation. Moreover, exposure of tumor to the NIR laser was expected to break the gold shell and trigger drug release when GNS@PNV arrived at the tumor sites. The in vitro DOX release profiles of DOX-loaded GNS@PNV at various conditions were shown in Fig. 2b. The fluorescence intensities of the pH 7.4 group and pH 5 þ 10 mM GSH group were very weak, which indicates that DOX was tightly encapsulated inside the core of gold nanoshell to cause fluorescence quenching. However, when the laser radiation (808 nm, 1.5 W/cm2, 10 min) was applied at pH 5.0 þ 10 mM GSH, the DOX fluorescence intensity of the GNS@PNV solution became much stronger. In this case, laser radiation apparently led to rupture of the continuous gold layer on the GNS@PNV surface to release DOX, as demonstrated by TEM analysis (Fig. 2-a5).

increased at all the three power levels of 0.7, 1.5, 3 W/cm2. The higher the irradiation power level, the higher the solution temperature induced by the laser. In contrast, only the irradiation at the highest power level of 3 W/cm2 resulted in an obvious temperature increase for PBS containing no gold nanoshell. The moderate power level of 1.5 W/cm2 was considered proper for biological experiments since in this case NIR irradiation only caused desirable temperature increase of gold nanoshell solutions rather than PBS. Moreover, the photothermal conversion of gold nanoshell solutions showed clear dependence on the GNS@PNV concentrations. High GNS@PNV concentration was in favor of a better photothermal conversion. Consistent results were obtained using the thermal infrared imaging (Fig. S5). Finally, repeated photothermal conversion of the gold nanoshell in response to NIR illumination was achievable. As shown in Fig. S6, no significant decline of photothermal conversion was observed for the GNS@PNV solution in the 6-round NIR illumination tests using the power level of 1.5 W/cm2 and 15 min irradiation in each cycle. This result is reasonable since UVevis spectral analysis revealed no shift of the absorption peak of GNS@PNV upon NIR illumination (data not shown) even if the TEM observation detected rupture at the irradiation of 1.5 W/cm2 (Fig. 2a5). Moreover, the results obtained in phantom experiments show that the system may be still effective even at cm depths (Fig. S7).

3.4. Cytotoxicity 3.3. Photothermal conversion of gold nanoshell As shown in Figs. S4 and S5, upon illumination of NIR light at 808 nm, the temperature of gold nanoshell solutions was obviously

As shown in Fig. 3a, the PNV template and GNS@PNV without encapsulating DOX showed no obvious cytotoxicity in Bel7402 cells. For both samples, cells incubated at high nanoparticle

L. Wang et al. / Biomaterials 78 (2016) 40e49

concentration of 0.5 mg/mL still showed viabilities above 80%. The DOX-loaded PNV showed cytotoxicities depending on the DOX concentrations (Fig. 3d), with a DOX IC50 value of 17.41 mM. The DOX-loaded GNS@PNV did not show a detectable cytotoxicity when the NIR irradiation was not applied. However, in the presence of NIR laser irradiation (808 nm, 1.5 W/cm2), it showed remarkable cytotoxicity. The IC50 value of DOX reached 2.93 mM, which is even much lower than that of the DOX-loaded PNV (2.93 vs 17.41 mM). The effect of Au concentration and laser irradiation time on the cytotoxicity of DOX-free GNS@PNV was examined as well (Fig. 3b and c). Long irradiation time and high Au concentration resulted in more effective suppression of the cell growth, apparently due to higher solution temperature induced by laser irradiation in these cases (Figs. S4 and S5). Results obtained in flow cytometry are in line with that of MTT assays (Fig. 4 and S8). Cell apoptosis induced by the nonresponsive PNV depended on the concentrations of encapsulated DOX. When the Bel-7402 cells were incubated with the DOX-free PNV, the percentage of normal viable cells was as high as 93%, demonstrating again that PNV itself was roughly not cytotoxic. However, the DOX-loaded PNV induced obvious cell apoptosis and necrosis. The apoptotic plus necrotic cells accounted for 17% and 31% of the total when the DOX concentrations were 8.8 mM and 22 mM, respectively. These results demonstrated that, owing to the reduction and pH dual sensitive structure, DOX can be effectively released from the vesicular template to induce cell apoptosis inside Bel-7402 cells. The effect of gold nanoshell in inducing cell apoptosis depended on whether the NIR laser irradiation was applied. Regardless of the DOX concentrations, the GNS@PNV appeared non-cytotoxic in the absence of laser irradiation. Above 92% of cells remained viable when the laser irradiation was not applied. However, in the presence of laser irradiation, the DOXloaded GNS@PNV became highly effective in inducing cell apoptosis in a DOX concentration-dependent manner. The DOXfree GNS@PNV induced 24% cell apoptosis plus necrosis, which was solely attributed to the photothermal therapy. In comparison, the DOX-loaded GNS@PNV induced 46% and 79% cell apoptosis plus necrosis at 8.8 mM and 22 mM DOX respectively, also more effective than the DOX-loaded PNV at the same DOX concentrations (46% vs 17% at 8.8 mM; 79% vs 31% at 22 mM). The above results demonstrated that photothermal therapy and chemotherapy have synergistically acted on the cancer cells incubated with the DOXloaded GNS@PNV under intermittent NIR laser irradiation.

47

removal of released DOX from the tumor. In addition, DOX fluorescence was detected in the stomach starting from 9 h, most likely due to an infusion of free DOX since this region was not irradiated. Previous studies have also shown that free DOX administered to nude mice via IV injection underwent some intestinal metabolism [22]. Based on the above results, we chose the post-injection time point of 12 h to further investigate the intratumoral DOX release behavior of the gold nanoshell. Mice bearing two identical Bel-7402 xenografts (L tumor and R tumor in Fig. 6) on two sides of the groin were used as the tumor model. As shown in Fig. 6a and S9, the L tumor irradiated by NIR laser at 12 h after the IV injection showed gradually intensified DOX fluorescence, reaching the highest at 80 min after NIR irradiation. In contrast, R tumor without NIR irradiation did not show detectable DOX fluorescence throughout the whole process of investigation due to the aforementioned reason affecting fluorescence detection of unreleased DOX. Similar intestinal metabolism of released DOX was observed as well [22]. In the ex vivo studies (Fig. 6b and c), photothermal imaging revealed significant accumulation of gold nanoshell in liver and both tumors. However, the R tumor without laser irradiation only showed very

3.5. Tumor accumulation and NIR laser irradiation triggered DOX release of GNS@PNV in vivo To study the accumulation of DOX-loaded GNS@PNV in tumor site, the DOX-loaded GNS@PNV was injected into the nude mice via tail vein. The tumor site was irradiated by NIR laser at various postinjection time points to trigger DOX release. As the light absorption peak of gold nanoshell overlaps with the fluorescence emission peak of DOX (Fig. S2 and Fig. 2b), the light emitted by DOX can be absorbed by gold nanoshell. On the other hand, DOX fluorescence was quenched when DOX molecules were entrapped inside nanoshell. To overcome these unfavorable effects on detection, the fluorescence imaging was recorded after DOX was released and diffused far away from the gold nanoshell. That is, after the laser irradiation of tumor site, DOX was allowed to diffuse for 1 h before fluorescence imaging. As shown in Fig. 5a, tumor accumulation of the DOX-loaded GNS@PNV depended on the post-injection time. As indicated by the DOX fluorescence intensities, GNS@PNV slowly accumulated in tumor site until reaching the highest content at 12 h after intravenous (IV) injection. After that, a gradual decay of DOX fluorescence in tumor was observed, implying eventual

Fig. 7. Tumor growth curve a) and cumulative survival b) of the Bel-7402 tumorbearing nude mice injected with various samples via tail vein (dose per injection: 4.5 mg DOX per Kg body weight if required). The same amount of GNS@PNV was applied if required. Laser irradiation: for 2 min at 1.5 W/cm2, 2 min intervals, repeated by 5 cycles. Arrows indicate time points to perform injection (day 1, 3 and 5). *P < 0.05 vs DOX-loaded GNS@PNV and free DOX.

48

L. Wang et al. / Biomaterials 78 (2016) 40e49

Fig. 8. Ex vivo histological analyses of tumor sections (20 days after the first treatment). Nuclei were stained blue, and extracellular matrix and cytoplasm were stained red in H&E staining. Brown and green stains indicated apoptotic and normal cells, respectively, in TUNEL analysis. Scale bars represent 100 mm. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this article.)

weak fluorescence because DOX was not released from the gold nanoshell. More evidence indicating DOX release and diffusion triggered by laser irradiation came from the study using an intratumoral injection of DOX-loaded GNS@PNV. As shown in Fig. 5b, after the NIR laser exposure of tumor, the DOX fluorescence not only gradually intensified but also spread to vast area of tumor. Moreover, study with NIR thermal imaging showed that the gold nanoshell itself did not move much after NIR irradiation (Fig. 5c). These imaging results strongly demonstrated that DOX was released from GNS@PNV and then diffused to other intratumoral region after the NIR laser irradiation broke the gold nanoshell based on reduction and pH dual sensitive polymeric vesicle.

3.6. Synergistic photothermo-chemotherapy of cancer triggered by near-infrared light irradiation in vivo Based on the results that the DOX-loaded GNS@PNV showed synergistic photothermo-chemotherapeutic effect in Bel-7402 cells, and the gold nanoshell could effectively accumulate in tumor site after IV injection and then release DOX in response to NIR laser irradiation, we further tested whether a synergistic photothermochemotherapy could be achieved in vivo using this multifunctional nanomedicine. As shown in Fig. 7, tumors in the PBS group, PBS þ laser group and GNS@PNV group grew similarly fast and no animal survived longer than 26 days, which clearly indicated that the photothermotherapy without gold nanoshell and chemotherapy of gold nanoshell without NIR laser irradiation did not work effectively. In comparison, the GNS@PNV þ laser group displayed a clear tumor growth inhibition and prolonged animal survival time, which was even better than the free DOX group. Moreover, the DOX-loaded GNS@PNV þ laser group exhibited the best inhibition of tumor growth and the longest survival time (90% animals survived longer than 30 days). The histological studies were performed to gain further insight into the roles of gold nanoshell-mediated photothermotherapy and chemotherapy. As shown in Fig. 8, the results are highly supportive of the tumor inhibition data. The H&E staining of tumor tissue sections from the PBS group, PBS þ laser group and DOX-loaded GNS@PNV group without laser irradiation appeared highly hypercellular and showed obvious nuclear polymorphism and foci of hemorrhage. A simple NIR laser exposure had no effect on tumor growth, and DOX was not released from the gold nanoshell without laser irradiation. Among the other three therapeutic groups, tumor tissues from animals in the DOX-loaded GNS@PNV þ laser group showed the fewest tumor cells but the highest level of necrosis,

evidencing the best therapeutic effect resulting from the combined photothermo-chemotherapy. The TUNEL assay revealed the levels of cancer cell apoptosis induced by photothermal and chemotherapy, which are in line with the H&E staining assay. Similar to PBS injection, simple exposure of tumor to the NIR laser without injecting the gold nanoshell or simple injection of DOX-loaded GNS@PNV without laser exposure of tumor did not cause appreciable apoptosis of cancer cells. However, tumor tissues from the free DOX group and GNS@PNV þ laser group showed obvious cancer cell apoptosis. More importantly, the DOX-loaded GNS@PNV þ laser group showed the highest level of cancer cell apoptosis, indicating once again a synergistic photothermochemotherapy of cancer. 4. Conclusion A reduction and pH dual sensitive polymeric vesicle incorporating anticancer drug DOX was prepared and then decorated with a gold layer using a modified method of in situ gold seed growth. By adjusting the density of gold layer on the vesicle surface, the gold nanoshell showed a desirable light absorption in the NIR region and high photothermal conversion efficiency crucial for potent photothermal therapy. Owing to the compact gold layer, the DOX-loaded gold nanoshell is leak-tight to avoid drug loss in bloodstream. However, after the gold nanoshell arrived at the tumor site, a rapid DOX release inside tumor could be triggered by the tissuepenetrating NIR irradiation causing rupture of gold layer. Consequently, highly effective photothermo-chemotherapy was achieved both in vitro and in vivo using the novel multifunctional nanomedicine denoted as the DOX-loaded GNS@PNV. Acknowledgments This work was supported by National Basic Research Program of China (2015CB755500), the National Natural Science Foundation of China (51225305, U1401242, 51373203 and 21174166) and Natural Science Foundation of the Guangdong Province (2014A030312018, S2012020011070), the Guangdong Innovative and Entrepreneurial Research Team Program (2013S086). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2015.11.024.

L. Wang et al. / Biomaterials 78 (2016) 40e49

References [1] E. Boisselier, D. Astruc, Gold nanoparticles in nanomedicine: preparations, imaging, diagnostics, therapies and toxicity, Chem. Soc. Rev. 38 (2009) 1759e1782. [2] J.M. Tam, J.O. Tam, A. Murthy, D.R. Ingram, L.L. Ma, K. Travis, K.P. Johnston, K.V. Sokolov, Controlled assembly of biodegradable plasmonic nanoclusters for near-infrared imaging and therapeutic applications, ACS Nano 4 (2010) 2178e2184. [3] S.J. Oldenburg, R.D. Averitt, S.L. Westcott, N.J. Halas, Nanoengineering of optical resonances, Chem. Phys. Lett. 288 (1998) 243e247. [4] D. Suzuki, H. Kawaguchi, Gold nanoparticle localization at the core surface by using thermosensitive core-shell particles as a template, Langmuir 21 (2005) 12016e12024. [5] P. Mulvaney, Surface plasmon spectroscopy of nanosized metal particles, Langmuir 12 (1996) 788e800. [6] X. Huang, I.H. El-Sayed, W. Qian, M.A. El-Sayed, Cancer cell imaging and photothermal therapy in the near-infrared region by using gold nanorods, J. Am. Chem. Soc. 128 (2006) 2115e2120. [7] J. Kim, S. Park, J.E. Lee, S.M. Jin, J.H. Lee, I.S. Lee, I. Yang, J.S. Kim, S.K. Kim, M.H. Cho, T. Hyeon, Designed fabrication of multifunctional magnetic gold nanoshells and their application to magnetic resonance imaging and photothermal therapy, Angew. Chem. 118 (2006) 7918e7922. [8] Y.G. Sun, Y.N. Xia, Mechanistic study on the replacement reaction between silver nanostructures and chloroauric acid in aqueous medium, J. Am. Chem. Soc. 126 (2004) 3892e3901. [9] H.Y. Liu, D. Chen, L.L. Li, T.L. Liu, L.F. Tan, X.L. Wu, F.Q. Tang, Multifunctional gold nanoshells on silica nanorattles: a platform for the combination of photothermal therapy and chemotherapy with low systemic toxicity, Angew. Chem. Int. Ed. 50 (2011) 891e895. [10] M.A. Nash, J.J. Lai, A.S. Hoffman, P. Yager, P.S. Stayton, “Smart” diblock copolymers as templates for magnetic-core gold-shell nanoparticle synthesis, Nano Lett. 10 (2010) 85e91. [11] X. Chen, Y.L. An, D.Y. Zhao, Z.P. He, Y. Zhang, J. Cheng, L.Q. Shi, Core-shellcorona Au-micelle composites with a tunable smart hybrid shell, Langmuir 24

49

(2008) 8198e8204. [12] L. Wang, Y.Y. Yuan, S.D. Lin, D. Cheng, X.Y. Wang, Q. Jiang, X.T. Shuai, Codelivery of 5-fluorocytosine and cytosine deaminase into glioma cells mediated by an intracellular environment-responsive nanovesicle, Polym. Chem. 5 (2014) 4542e4552. [13] C. Graf, A. Blaaderen, Metallodielectric colloidal core-shell particles for photonic applications, Langmuir 18 (2002) 524e534. [14] Y.D. Jin, X.H. Gao, Plasmonic fluorescent quantum dots, Nat. Nanotechnol. 4 (2009) 571e576. [15] J. Dai, S.D. Lin, D. Cheng, S.Y. Zou, X.T. Shuai, Interlayer-crosslinked micelle with partially hydrated core showing reduction and pH dual sensitivity for pinpointed intracellular drug release, Angew. Chem. Int. Ed. 50 (2011) 9404e9408. [16] X. Sun, S. Dong, E. Wang, One-step synthesis and characterization of polyelectrolyte-rotected gold nanoparticles through a thermal process, Polymer 45 (2004) 2181e2184. [17] A.M. Smith, M.C. Mancini, S. Nie, Bioimaging: second window for in vivo imaging, Nat. Nanotechnol. 4 (2009) 710e711. [18] B.G. Prevo, S.A. Esakoff, A. Mikhailovsky, J.A. Zasadzinski, Scalable routes to gold nanoshells with tunable sizes and their response to near infrared pulsed laser irradiation, Small 4 (2008) 1183e1195. [19] C.L. Didychuk, P. Ephrat, M. Belton, Synthesis and in vitro cytotoxicity of mPEG-SH modified gold nanorods, Proc. SPIE 6856 (2008), 68560M1e68560M18. [20] T. Niidome, M. Yamagata, Y. Okamato, Y. Akiyama, H. Takahashi, T. Kawano, Y. Katayama, Y. Niidome, PEG-modified gold nanorods with a stealth character for in vivo applications, J. Control Release 114 (2006) 343e347. [21] A. Rahman, A. Ganjei, J.R. Neefe, Comparative immunotoxicity of free doxorubicin and doxorubicin encapsulated in cardiolipin liposomes, Cancer Chemoth. Pharm. 16 (1986) 28e34. [22] Y.M. Kang, G.H. Kim, J.I. Kim, D.Y. Kim, B.N. Lee, S.M. Yoon, J.H. Kim, M.S. Kim, In vivo efficacy of an intratumorally injected in situ-forming doxorubicin/ poly(ethylene glycol)-b-polycaprolactone diblock copolymer, Biomaterials 32 (2011) 4556e4564.