Porphyrin Immobilized Nanographene Oxide for ... - ACS Publications

58 downloads 57 Views 7MB Size Report
Jul 8, 2016 - enhanced and targeted photothermal therapy for brain cancer. ... The photothermal therapy effects of PNG-RGD completely eliminated the ...
Article pubs.acs.org/journal/abseba

Porphyrin Immobilized Nanographene Oxide for Enhanced and Targeted Photothermal Therapy of Brain Cancer Siheng Su,† Jilong Wang,† Evan Vargas,† Junhua Wei,† Raul Martínez-Zaguilán,‡ Souad R. Sennoune,‡ Michelle L. Pantoya,† Shiren Wang,§ Jharna Chaudhuri,† and Jingjing Qiu*,† †

Department of Mechanical Engineering, Texas Tech University, 2500 Broadway, Lubbock, Texas 79409, United States Department of Cell Physiology and Molecular Biophysics, Texas Tech University Health Sciences Center, 3601 Fourth Street, Lubbock, Texas 79430, United States § Department of Industrial & Systems Engineering, Texas A&M University, 400 Bizzell Street, College Station, Texas 77843, United States ‡

S Supporting Information *

ABSTRACT: Brain cancer is a fatal disease that is difficult to treat because of poor targeting and low permeability of chemotherapeutic drugs through the blood brain barrier. In a comparison to current treatments, such as surgery followed by chemotherapy and/or radiotherapy, photothermal therapy is a remarkable noninvasive therapy developed in recent years. In this work, porphyrin immobilized nanographene oxide (PNG) was synthesized and bioconjugated with a peptide to achieve enhanced and targeted photothermal therapy for brain cancer. PNG was dispersed into the agar based artificial tissue model and demonstrated a photo-to-thermal conversion efficiency of 19.93% at a PNG concentration of only 0.5 wt %, with a heating rate of 0.6 °C/s at the beginning of irradiation. In comparison, 0.5 wt % graphene oxide (GO) indicated a photo-to-thermal conversion efficiency of 12.20% and a heating rate of 0.3 °C/s. To actively target brain tumor cells without harming healthy cells and tissues surrounding the laser path, a tripeptide L-arginyl-glycylL-aspartic (RGD) was further grafted to PNG. The photothermal therapy effects of PNG-RGD completely eliminated the tumor in vivo, indicating its excellent therapeutic effect for the treatment of brain cancer. KEYWORDS: graphene oxide, porphyrin, photothermal, targeting

1. INTRODUCTION According to the American Cancer Society, in 2015, 22 850 malignant brain tumors were diagnosed.1 Current therapy methods for brain cancers are surgery followed by chemotherapy and/or radiation therapy. However, with these treatments, the 2-year survival rate of brain cancer is still very low (i.e., 30% for the malignant brain cancer glioblastoma) due to the complex structure of the brain, poor targeting, and low permeability of drugs through the blood brain barrier.1 In a comparison to current therapy methods with disadvantages and severe side effects, photothermal therapy (PTT) is a promising strategy in brain cancer therapy.2,3 PTT utilizing photoabsorbing agents to convert light energy to thermal energy has been explored as a noninvasive treatment for fatal diseases, especially cancer.4,5 Nanomaterials, such as silver nanoparticles, gold nanorods, gold nanomatryoshkas, and carbon nanotubes, are considered to have a large surface to volume ratio and high absorbance in the near-infrared (NIR) window. Therefore, nanomaterials attract intensive research interests in the biomedical field.1,6−8 However, metal nanomaterials, especially gold nanomaterials, undergo instability or melting of nanostructures under NIR irradiation, even though © 2016 American Chemical Society

they have strong surface-plasmonic absorption in the NIR region.9−12 In contrast, it was reported that graphene or graphene oxide shows a negligible change of optical absorbance even with continuous NIR laser irradiation for hours, exhibiting superior stability for PTT.13 Moreover, in contrast to sodium dodecylbenzenesulfonate solubilized single-wall carbon nanotubes (SDS-SWNT), graphene sheets can generate more heat under the same power of NIR radiation and perform better PTT effects in in vitro studies.14 Graphene has been demonstrated as a potential PTT agent in previous studies.14−18 Ultrasmall reduced graphene oxide (nRGO) was first demonstrated as a highly efficient PTT agent for in vitro cancer cells.19 Reduced graphene oxide nanomesh with RGD conjugation also showed promising tumor targeted PTT effects.20 Although great efforts have been made in recent years, the applications of graphene based nanosheets as photothermal agents are still limited by some technical challenges. Tumors in mice were found to regrow and achieve Received: May 27, 2016 Accepted: July 8, 2016 Published: July 8, 2016 1357

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Scheme 1. Schematic Illustration of (a) the PNG-RGD Synthesis and Its PTT Process in Vivo and (b) the Enhanced Photothermal Effect of PNG, Including the Light Absorption of Porphyrin, Photo-Induced Energy Transfer from Porphyrin to GO, and Heat Dissipation of GO

2. EXPERIMENTAL SECTION

a 5.4-fold increment in size after 22 days postinjection of GOPEG, indicating that it did not completely eliminate tumors.21 Nanographene oxide was shown to decrease fertility in Balb/C mice at a high concentration.22 Thus, it is crucial to improve the PTT efficiency of graphene based nanosheets at a low concentration and avoid the possible long-term toxicity of graphene. In order to improve the properties of graphene for enhanced PTT efficiency, some recent efforts have been focused on magneto-PTT and photosensitizer enhanced PTT.23,24 Photosensitizers, such as chlorin e6 (Ce6) and cyanine, have been grafted onto graphene sheets for PTT therapy to achieve more efficient photo-to-thermal conversion. Photosensitizers are used often as photodynamic therapy (PDT) agents for cancers that require oxygen during treatment. Therefore, PDT is not suitable for hypoxia tumor therapy, while PTT can overcome this limitation.24−26 Recently a pH-responsive cyanine-grafted graphene oxide was developed on the basis of a fluorescence resonance energy transfer (FRET) mechanism to enhance its PTT effect in lysosomes.27 However, the results showed there was more than 50% reduction of absorbance around 780 nm of cyanine-grafted graphene oxide after 10 min irradiation of 785 nm laser, which hinders its future application. Because of the limitations of the current treatments for brain cancer, we developed porphyrin immobilized nanographene oxide (PNG) conjugated with the tripeptide L-arginyl-glycyl-Laspartic (PNG-RGD) as a PTT agent with high photostability and active targeting. GO is an excellent energy acceptor of organic dyes, without depending on the emission spectra of the donor.28 In PNG, the photon energy absorbed by porphyrin effectively transfers to GO to enhance the photothermal effects. Moreover, PNG-RGD recognizes specifically the overexpressed αvβ3 integrin on brain cancer cells without harming the healthy cells during therapy. In this paper, NIR light-absorbing porphyrin enhanced the photothermal effects of GO through photoinduced energy transfer mechanism, resulting in a phototo-thermal conversion efficiency of 19.93% of agar artificial tissue with 0.5 wt % of PNG, which is about 1.6 times that with 0.5 wt % of GO. Efficient in vitro PTT effects of PNG-RGD were also demonstrated on a glioblastoma cell line U87-MG. For in vivo study, U87-MG cells were injected into the mice to grow tumors. Subsequently, the mice were injected with PNGRGD intratumorally and irradiated by an 808 nm laser (Scheme 1). This treatment completely eliminated tumors, indicating PNG-RGD is an excellent therapeutic agent for brain cancer.

2.1. Materials. Graphite was purchased from Asbury Carbons, and meso-tetra(4-methylphenyl) porphine (porphyrin) was obtained from Frontier Scientific. Nitric acid (HNO3), sodium chlorate (NaClO3), Nmethyl-2-pyrrolidone (NMP), tetrabutylammonium hydroxide (TBA), N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), Arg-Gly-Asp (RGD), 4′,6diamidino-2-phenylindole (DAPI), and dimethyl sulfoxide (DMSO) were from Sigma-Aldrich. Cyanine 7 NHS ester was purchased from Lumiprobe. 2.2. Cells. U87-MG cells were purchased from ATCC (cat. HTB14), and the human brain microvascular endothelial cells (HBMEC, ACBRI 376) are from Cell systems. U87-MG cells were cultured in Eagle’s minimum essential medium (EMEM, cat. 30-2003) from ATCC supplemented with 10% FBS and 100 IU mL−1 penicillin and streptomycin. HBMEC were cultured in complete serum-free medium (Cell Systems, cat. SF-4Z0-500) supplemented with culture boost (Cell Systems, 4Z0-210). All cells were cultured at 37 °C in 5% CO2. 2.3. Synthesis of GO, PNG, and PNG-RGD. GO was prepared by the modified Brodie’s method.29 Specifically, 500 mg of graphite was oxidized by a mixture of 80 mL of HNO3 and 4.25 g of NaClO3. The mixture was stirred at room temperature for 24 h. The sample was diluted with water and then neutralized with sodium hydroxide. After that, the sample was washed and collected by centrifugation to obtain graphite oxide. Graphite oxide was mixed with ammonia hydroxide, and was exfoliated into GO by ultrasonication for 2 h. The exfoliated GO was collected by centrifugation at 5000 rpm for 30 min to remove the large sediment. PNG was prepared as described in our previous work.30 Specifically, 10 mL of NMP was mixed with 2 mL of TBA, and then 1.8 mg of GO and 1.8 mg of porphyrin were added into the above solvent. The mixture was stirred at 200 rpm at room temperature for 7 days. The excess porphyrin was removed by dialysis (MW: 14 000 Da) in water for 1 week. Finally, PNG was obtained by centrifugation at 5000 rpm for 30 min to remove the large flakes of graphite oxide. PNG-RGD was prepared by a carbodiimide cross-linker reaction described as follows: 2 mg of PNG was dissolved in 20 mL of water by tip sonication for 2 h, and then 2 mg of EDC was added into the solution, followed by 3 mg of NHS. The mixture was then stirred for 30 min at 200 rpm to activate the carboxyl group. Then, 0.3 mg of RGD was added into the above solution and reacted with PNG by stirring overnight. The obtained PNG-RGD was further purified by dialysis (MW: 14 000 Da) in water for 1 week to remove the residues. For in vitro cell imaging, PNG-RGD was conjugated with Rhodamine B (RhB) through a carbodiimide cross-linker reaction. Specifically, 0.5 mg of PNG was dissolved into 10 mL of water by tip sonication for 2 h, and then 2 mg of EDC was added into the solution, followed by 3 mg of NHS. The mixture was then stirred for 30 min at 200 rpm to activate the carboxyl group. Then, 0.5 mg of Rhodamine B was added into the above solution and reacted with PNG by stirring overnight. The obtained PNG-RGD-RhB was further purified by dialysis (MW: 14 000 Da) in water for 1 week to remove the residues. 1358

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Figure 1. TEM images of (a) GO and (b) PNG. (c) UV−vis spectra, (d) Raman spectra, and (e) TGA curves of GO, porphyrin, and PNG. PNG-RGD conjugated with cyanine 7 (PNG-RGD-cy7) was used for in vivo fluorescence imaging. The as-fabricated PNG-RGD (1 mg) was dispersed in 0.1 M sodium bicarbonate aqueous solution (10 mg) by tip-sonication for 1 h. Then cyanine 7 NHS ester (0.1 mg) in DMSO (1 mL) was added into the solution, and the mixture was stirred overnight. PNG-RGD-cy7 was purified by dialysis (MW: 14 000 Da) in water for 1 week to remove the residues. 2.4. Materials Characterizations. Fourier transform infrared (FTIR) spectra were achieved with a Nicolet IS10 FT-IR spectrometer (Thermo Scientific) under transmission mode by a KBr pellet method. The morphology of GO and PNG was examined by transmission electron microscopy (TEM, HITACHI 8100, operated at 75 kV), while UV−vis absorption spectra and the absorbance of crystal violet at 590 nm were measured by an Infinite M1000 Pro plate reader (Tecan). Fluorescence emission spectra of samples were measured on a FluoroMax-3. Thermogravimetric analysis (TGA) was performed using a TGA Q50 instrument in a nitrogen inert atmosphere. Raman scattering measurements were performed at room temperature using a SENTERRA microscope from Bruker, with an excitation at 514 nm. The composition of GO was analyzed by PHI5000 Versa Probe X-ray photoelectron spectroscopy (XPS). The beam size was 100 μm with a 45° takeoff angle. The X-ray source provided the monochromatic Al Kα radiation of 1486.7 eV. 2.5. Laser Irradiation and Temperature Measurement. Agar hydrogel was used to simulate tissue to measure the photo-to-thermal conversion phenomenon of PNG. Specifically, PNG was dispersed in deionized (DI) water with concentration of 100, 50, and 20 μg/mL; subsequently, 20 mg/mL agar was added in each sample. The mixture was heated to 95 °C in an oil bath with continuous stirring to dissolve the agar. After that, the mixture was injected to a mold with a dimension of 15 mm × 15 mm × 5 mm or 50 mm × 15 mm × 5 mm and cooled down at 4 °C for 2 h. Then the gel was released from the mold to perform the measurement. As control samples, we used 100 μg/mL of GO embedded agar gel, 100 μg/mL of porphyrin embedded agar gel, and pure agar gel. PNG100/agar, PNG50/agar, PNG20/agar, GO100/agar, and porphyrin100/agar represent the 100 μg/mL of PNG, 50 μg/mL of PNG, 20 μg/mL of PNG, 100 μg/mL of GO, and 100 μg/mL of porphyrin embedded agar gel, respectively.

The temperature variation of those gels under continuous 808 nm laser irradiation (2.5W/cm2) was monitored by a FLIR SC8303 highspeed infrared camera. The camera was placed directly in front of the agar gels, recording two-dimensional transient temperature distribution images (Figure S1). The viewing area of the samples and holder was set at a 640 × 360 pixel window frame size, with an acquisition rate of 2 frames per second. The heating rate of the gel was calculated as the temperature change versus time, while the photo-to-thermal conversion efficiency was calculated as follows14,31 η=

Q cmΔT = W p×t

(1)

where η represents the thermal efficacy, Q is the heat released by the PNG/GO embedded agar gel, W is the work induced by the input power, c is the heat capacity of the gel, m is the mass of the gel, ΔT is the average temperature change of the gel, p is the input power of the laser to the gel, and t represents the irradiation time. Heat capacity (c) of those different tissue models was calculated according to the following equation,32 where cagar, magar, cPNG, and mPNG represent the heat capacity of agar gel, weight of agar gel, heat capacity of PNG, and weight of PNG, respectively:

c=

magar × cagar + mPNG × c PNG magar + mPNG

(2)

Since the weight of PNG is only 100 μg in each milliliter of the gel, far less than the weight of agar (20 mg), mPNG × cPNG is negligible. Also, eq 2 can be modified into

c=

magar × cagar magar + mPNG

(3) 33

where cagar = 3.9 J/(g K). 2.6. Cytotoxicity Evaluation of PNG-RGD. Cytotoxicity of PNG-RGD was evaluated by a crystal violet staining method.34 HBMEC cells were seeded at 2 × 104 cells/well onto an attachmentfactor-coated 24-well plate. Cells were cultured in 1 mL of CSF medium for 24 h. The medium was replaced by 1 mL of sterilized fresh 1359

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Figure 2. (a) Infrared images of temperature distribution in PNG100/agar, GO100/agar, porphyrin100/agar, and agar under NIR laser irradiation (808 nm, 2.5 W/cm2) for 300 s. (b) Infrared images of temperature distribution in PNG100/agar PNG50/agar, and PNG20/agar under NIR laser irradiation (808 nm, 2.5 W/cm2) for 300 s. medium with 0, 10, 20, 50, 100, 150, or 200 μg/mL of PNG-RGD, respectively. Cells were cultured with or without PNG-RGD for another 24 h and then fixed with 1 mL of 2% glutaraldehyde in triplicate for 10 min. After that, the medium was aspirated, and the cells were washed with PBS twice. Then, 1 mL of 0.1% crystal violet was added into the plate, and the cells were stained for 40 min. The crystal violet was then washed extensively using water. The only crystal violet left was inside the stained cells. The plate was further drained inversely overnight. The cytotoxicity was evaluated by reading the absorbance (590 nm) of crystal violet. 2.7. PTT Effects for U87-MG Brain Tumor Cells in Vitro. U87MG brain tumor cells were seeded onto a 96-well plate at 5 × 104 cells/well in the cell culture medium and cultured for 24 h. Then the medium was replaced by 100 μL of sterilized fresh medium with PNGRGD at different concentration. After 2 h of culturing in a CO2 incubator at 37 °C, the tumor cells were washed by fresh warm medium and irradiated using an 808 nm NIR laser with a power density of 2.5 W/cm2 for 0, 2, 4, 6, 8, and 10 min, respectively. Cells in the fresh medium without irradiation were used as a control. Cell proliferation of the cells was evaluated by the crystal violet staining method as described before. 2.8. In Vivo Fluorescence Imaging and in Vivo Photothermal Therapy. Female athymic nude mice purchased from Charles River Laboratories (Strain Code 490) were injected with U87-MG cells. The glioblastoma tumor models were generated by subcutaneous injection of 2 × 106 U87-MG cells suspended in 100 μL of PBS into the right

hind flank of the nude mice. The mice were used for in vivo fluorescence imaging and photothermal therapy when the diameter of the tumors reached ∼6 mm. For in vivo fluorescence imaging, the mice were anesthetized with 3% isoflurane/air and injected with PBS or PNG-RGD-cy7 (100 μL) intratumorally. The mice were monitored by an IVIS spectrumCT in vivo imaging system at 5 min, 24 h, and 48 h postinjection at the excitation and emission wavelengths of 745 and 800 nm with the following parameters: exposure time, 1 s; binning, medium; F/stop, 2. For in vivo photothermal therapy, the mice bearing glioblastoma tumors were intratumorally injected with PNG-RGD with a dosage of 10 mg/kg. Tumor-bearing mice with or without injection of nanoparticles were anaesthetized with 3% isoflurane/air and irradiated by 808 nm laser (2.5 W/cm2) for 5 min (Figure S2). The tumor size was measured by a caliper every 2 days and calculated as the volume = (tumor length) × (tumor width)2/2. The relative tumor volumes were calculated as V/V0 (V0 is the tumor volume when the treatment is initiated). All the animal protocols used in this study were approved by the Animal Care and Use Committee at Texas Tech University.

3. RESULTS AND DISCUSSION PNG was prepared by π−π conjugation between GO and porphyrin, and the morphologies of GO and PNG are shown in Figure 1. The TEM images (Figure 1a,b) indicate that the synthesized PNG consists of planar nanosheets with porphyrin 1360

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Figure 3. (a) Maximum temperature change and (b) heating rate of PNG100/agar, GO100/agar, porphyrin100/agar, and agar.

Table 1. Thermal Efficiency of Test Samplesa factors

a

sample

c (J/g K)

magar (g)

mparticles (g)

ΔT (K)

P (W)

t (s)

η (%)

PNG100/agar PNG50/agar PNG20/agar GO100/agar prophyrin100/agar agar

3.9 3.9 3.9 3.9 3.9 3.9

1.06946 1.19068 1.07175 1.05094 1.01451 0.99553

10−4 5 × 10−5 2 × 10−5 10−4 10−4 0

21.5 9.70 8.38 13.40 6.42 1.10

1.5 1.5 1.5 1.5 1.5 1.5

300 300 300 300 300 300

19.92 10.01 7.78 12.21 5.65 0.95

ΔT represents the mean temperature change of the entire gel.

porphyrin were embedded into agar gel, and the temperature variations of the gels under laser irradiation were recorded by an IR camera. Figure 2 shows the infrared images of temperature distribution in PNG100/agar, PNG50/agar, PNG20/agar, porphyrin100/agar, GO100/agar, and agar under 2.5W/cm 2 NIR laser irradiation. In particular, PNG100/agar showed the highest temperature increment than the other materials, and the heating rate of PNG is concentration dependent. Figure 3 shows the statistical analysis of heat generation effects in all the agar gels. Figure 3a shows that the maximum temperature increment of PNG100/agar reaches 43 °C after an irradiation of 60 s, at which point tumor cells are effectively killed. Furthermore, maximum temperature increments (Δt) in PNG100/agar, PNG50/agar, PNG20/agar, agar, GO100/agar, and porphyrin100/agar are 40.25, 18.42, 14.42, 1.76, 25.8, and 14.40 °C, respectively. In Figure 3b, the temperature reaches a steady state (heating rate = 0) in all kinds of gels after the irradiation for about 180 s. The trend of heating rate is similar to the temperature increment, while PNG100/agar shows an increment of about 100% in the heating rate at the beginning compared to GO100/agar. The photo-to-thermal conversion efficiency of various gels is calculated by eq 1, while the heat capacity of the gels is obtained by eq 3. The input power of the laser is 1.5 W with the spot size of around 10 mm × 6 mm. As shown in Table 1, PNG100/agar has the highest photo-to-thermal conversion efficiency, 19.93%. It indicates that PNG100/agar harvests phonon energy better than GO100/agar, which is because porphyrin in PNG can effectively absorb NIR light. For free porphyrin molecules, energy normally is released via fluorescence and singlet-oxygen generation.40 While these processes are greatly suppressed in PNG, the excited electrons triggered by NIR light in porphyrin as a donor instead transfer to GO as an acceptor. In GO, electrons transit from the shortlived excited states to the ground states through nonradiative

particle aggregations, unlike the clean planar GO nanosheets. Additionally, broad absorbance in the NIR region of PNG can be observed in the UV−vis spectra (Figure 1c), suggesting that porphyrin conjugation improves the ability of GO absorbing NIR light, where porphyrin aggregates on graphene nanosheets are used as antennas to harvest energy from the photons. The fluorescence spectra (Figure S4) verified the energy transfer between porphyrin and GO, which show that the fluorescence of porphyrin was quenched by GO in porphyrin/GO mixture due to the direct contact between porphyrin and GO sheets. The Raman spectra (Figure 1d) of GO show a D band (defects/disorder-induced mode) at 1339 cm−1 and a G band (in-plane stretching tangential mode) at 1584 cm−1 in GO. In comparison, the G band of PNG red shifts to 1592 cm−1, suggesting the conjugation between GO and porphyrin.35 Moreover, ID/IG increases from 1.14 in GO to 1.37 in PNG, indicating an increase in the number of sp3 carbons formed on the graphene nanosheet during the functionalization.36 Moreover, 2D/G ratios suggest the number of layers of stacked graphene sheets. I2D/G values of single-, double-, triple-, and multilayered (>4) samples are demonstrated to be >1.6, ∼0.8, ∼0.3, and ∼0.07, respectively.37−39 The Raman spectra show I2D/IG of both GO and PNG are about 0.09, indicating the presence of multilayer sheets in GO and PNG. TGA curves (Figure 1e) of GO, porphyrin, and PNG further verified the immobilization of porphyrin on GO. Below 100 °C, the absorbed water was removed, and the oxygen-containing functional groups and carbon skeleton on GO and PNG started pyrolysis. The weight loss of porphyrin was observed around 300 °C. At 800 °C, porphyrin, GO, and PNG retained 41.9%, 56.4%, and 47.8% of original mass, respectively, suggesting that the percentage of porphyrin in PNG is about 59.4%. In order to study the temperature variation and distribution inside the tissue under laser irradiation, PNG, GO, and 1361

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Figure 4. (a) One-dimensional temperature distribution in GO (above) and PNG (bottom) at the laser irradiation direction and (b) temperature distribution of PNG100/agar (insert: photograph of PNG100/agar gel with size 50 mm × 15 mm).

stretching mode of OH, respectively. The peak at 1070 cm−1 is attributed to the stretching mode of CO, while CO presents at 1626 cm−1. All of these oxidation groups are present in GO, PNG, and PNG-RGD. Moreover, the presence of amide II (∼1580 cm−1) and amide III (1287 cm−1) in PNG-RGD verifies a successful covalent bonding between PNG and RGD.44,45 Cytotoxicity of PNG-RGD was evaluated before in vitro PTT effects studies. Figure 6a shows high proliferation of the HBMEC cells cultured with PNG-RGD, indicating a low cytotoxicity of PNG-RGD even at high concentration (200 μg/ mL). In in vitro PTT studies, the number of cells incubated with 100 μg/mL of PNG-RGD decreased to 46.13% ± 0.19% after 4 min irradiation with 2.5 W/cm2 of 808 nm laser (Figure 6b). After a 10 min irradiation, 78.07% ± 0.1% of U87-MG cells were killed, indicating the effective PTT effects of PNGRGD. Moreover, the concentration dependent PTT effect is demonstrated in Figure 6c. Without PNG-RGD or with 10 μg/ mL of PNG-RGD, cells were mostly alive. The number of living cells decreased to 61.48% ± 2.71% and 47.74% ± 3.87% after culturing cells with 20 or 50 μg/mL PNG-RGD under NIR irradiation for 10 min. In a comparison to PNG-RGD, PNG was washed away by PBS after 2 h culturing with cells since there was no ligand to recognize U87-MG cells (Figure 6d). This further demonstrated the successful conjugation of RGD onto PNG. Thus, the number of U87-MG cells did not decrease after culturing cells with PNG even with same irradiation time (10 min). These results suggest that the addition of PNG-RGD with laser irradiation can effectively kill brain cancer cells. The observation that PNG-RGD was effective in killing cancer is surprising since RGD is a general cell attachment sequence. RGD was originally identified as part of the sequence motif in fibronectin involved in cell attachment.46 Integrins are cell adhesion molecules and also recognize the RGD motif to allow binding of cells to the extracellular matrix and for cell− cell interactions. Integrins are an important constituent of focal adhesions. Further, tumor cells have a greater number of focal adhesions and focal adhesion kinases than normal cells. Focal adhesion kinase (FAK) has been implicated in cell adhesion, growth, cell differentiation, and cell migration. Not surprisingly, in mouse models of cancer, FAK have been found to be involved in tumor formation and progression.47 In human tumors, FAK expression was significantly elevated in 17 out of 17 invasive and metastatic colonic lesions and in 22 out of 25 invasive and metastatic breast cancer when compared to normal tissue from the same patients.48 FAKs are also overexpressed in

decay, converting light energy to heat through the photoinduced energy transfer mechanism.41 The NIR absorption significantly increased in PNG mainly due to the π−π interaction between porphyrin and GO, which enhances the conversion efficiency from NIR light to heat. The penetration depth was also studied in both PNG and GO integrated in vitro tissue models. A 100 μg/mL of PNG embedded agar gel (50 mm × 15 mm) was used to study the penetration depth of an 808 nm laser. Figure 4 shows the onedimensional temperature distribution of PNG100/agar gel on the laser irradiation direction where the maximum temperature of the gel occurs. In the region 1.70−12.83 mm from the left edge of the PNG100/agar tissue model, temperature was above 43 °C after 300 s of irradiation, indicating that tumor cells can be killed in this region. So, the penetration depth of the 808 nm laser can be up to about 13 mm in PNG integrated tissue models, 1.5 mm deeper than the one in GO integrated tissue models, which is sufficient for a brain tumor since the vertex of an adult is about 6 mm.42 Moreover, the highest temperature of PNG100/agar is remarkably higher than GO100/agar when the depth is less than 18 mm. In order to enable the PTT agent with selective targeting properties, RGD was covalently conjugated to PNG nanosheets. The FTIR (Figure 5) spectra demonstrated the covalent

Figure 5. FTIR spectra of GO, PNG, and PNG-RGD.

bonding between RGD and PNG. In Figure 5, a peak at 1633 cm−1 is observed for GO, PNG, and PNG-RGD, indicating the CO stretching vibration of COOH groups.43 This carboxyl group is beneficial for further covalent functionalization with biomarkers such as antibodies and ligands through amide reaction for active targeting. Furthermore, the peaks at 1380 and 3390 cm−1 are associated with the deformed mode and 1362

DOI: 10.1021/acsbiomaterials.6b00290 ACS Biomater. Sci. Eng. 2016, 2, 1357−1366

Article

ACS Biomaterials Science & Engineering

Figure 6. (a) Cell viability of HBMEC cultured with PNG-RGD at different concentrations (P > 0.05). (b) Cell viability of U87-MG cells cultured with 100 μg/mL of PNG-RGD under irradiation for various times. (c) Cell viability of U87-MG cultured with PNG-RGD at different concentration under irradiation for 10 min. (d) Cell viability of U87-MG cultured with 100 μg/mL of PNG or PNG-RGD under irradiation for 10 min (*, P < 0.05; **, P < 0.01).

Figure 7. Confocal microscopy images of U87-MG after coculturing with PNG-RGD-RhB. DAPI (blue) was used to stain the nucleus of the cells.

laser irradiation (