A tumor-targeting drug delivery system based on cyclic NGR-modified, combretastatin A4-loaded, functionalized graphene oxide nanosheets

Abstract

Graphene oxide has shown great potential in drug delivery. In this study, we developed a novel tumor-targeting drug carrier based on aminopeptidase N (APN or CD13)-targeting peptide (NGR) functionalized graphene oxide (GP–cNGR/PVP). For the preparation, graphene oxide (GO) was functionalized using polyethylenimine (PEI) covalently linked with cyclic NGR (cNGR), and then the nanosheets were conjugated with polyvinylpyrrolidone (PVP) via non-covalent interactions. The results showed that an efficient loading of combretastatin A4 (CA4) on GP–cNGR/PVP (0.5630 ± 0.0132 mg mg⁻¹) was obtained. In vitro cytotoxicity and cellular uptake studies using two tumor cells (HT-1080, MCF-7) indicated that the GP–cNGR/PVP nanosheets could recognize certain tumor cells and enhance the uptake by cells, especially for cells overexpressing CD13 receptors. These results demonstrate that GP–cNGR/PVP could be a potential vehicle to delivery anticancer agents for specific cancer therapy.

---

1. Introduction

The efficacy of many hydrophobic antitumor agents is often hindered by their poor solubility, severe toxic side effects, drug resistance, and nonspecific action. Various drug delivery systems have been developed to overcome these limitations, among which graphene oxide (GO), a one-atom-thick two-dimensional carbon material, has attracted considerable attention. GO has shown exciting efficacy due to its large theoretical specific surface area, well-defined surface properties, low systemic toxicity and loading capacity of aromatic molecules via π–π stacking and hydrogen-bonding interactions.^1–3 Antitumor compounds, such as 5-fluorouracil, methotrexate, camptothecin, and doxorubicin, can be highly loaded onto GO nanosheets via π–π stacking and van der Waals interaction.^4–7 However, GO can easily produce aggregation in solution under physiological conditions, which makes further processing difficult.⁸ To improve stability and biocompatibility in aqueous phase, hydrophilic groups, such as PEI, polyethylene glycol have been grafted onto GO.^9,10

To increase the tumor-specific drug accumulation, great attention has been given to the unique molecular markers that specially overexpress in cancerous tissues.¹¹ The aminopeptidase N (APN or CD13), a transmembrane zinc-dependent metalloprotease, has contributed in tumor invasion and angiogenesis.¹² Besides that, CD13 is also an important biomarker, which is highly overexpressed at endothelial cell surface in the neo-angiogenic blood vessels, and especially in tumor cells.¹³ CD13 is selectively recognized by peptides containing the Asn–Gly–Arg (NGR) sequence, a tumor-homing motif discovered by phage display technologies.¹⁴ According to the report, NGR as a potent tumor-targeting ligand, has been certified useful for delivering chemotherapeutic drugs, apoptotic peptides and liposomes to tumor vessels.^15–18 Apart from that, it is found that cyclic NGR motif with disulfide bridge constraint is not only critical for stabilizing the bent conformation but also can increase the tumor-targeting efficiency.¹⁹ Thence, a novel tumor-targeting drug delivery system was constructed by modifying GO nanosheets with the cyclic NGR.

Combretastatin A4 (CA4), isolated from the South African tree Combretum caffrum, is a small molecule natural anticancer drug. It is an antimitotic agent that strongly inhibits the polymerization of tubulin by attaching to the colchicine-binding site of the β-tubulin subunit.²⁰ CA4 can act as a vascular disrupting agent to elicit irreversible vascular shutdown with in solid tumors and leave the normal vasculature intact.²¹ In addition, CA4 also exhibits strong cytotoxicity against a broad spectrum of cancer cell lines, such as murine melanoma, human ovarian and colon cancer cells, even those with multidrug resistance.^22–24 However, the poor solubility and the limited bioavailability of CA4 significantly hinder the clinical application. The phosphate CA4P pro-drug has been developed to increase the water solubility, which is not the optimum approach because of the short half-life and undesirable side effects in normal tissues, such as cardiotoxicity and ataxia.²⁵

In this work, we developed a GP–cNGR/PVP/CA4 nanosystem to induce selective cytotoxicity in CD13-overexpressing cancer cells. As schematically depicted in Fig. 1, the cyclic NGR peptide and functionalized GO were conjugated by a classic amidation reaction to improve the specificity of cell targeting by nanocarriers. GP–cNGR was decorated with polyvinylpyrrolidone (PVP) to improve physiological stability,²⁶ and then CA4 was loaded onto GP–cNGR/PVP via simple π–π stacking interaction. The resultant GP–cNGR/PVP nanosystem showed high stability, dispersibility, high drug loading efficiency and biocompatibility, and it was capable of sustained drug release. The in vitro selective cytotoxicity of the GP–cNGR/PVP/CA4 nanosystem induced in targeted tumor cells compared to non-targeted cells was investigated using cell viability analysis. HT-1080 cells (human fibrosarcoma cell line, high level of expression of CD13) and MCF-7 cells (human breast adenocarcinoma cell line, low level of expression of CD13) were chosen as the model for tumor cells.^15,27,28 Furthermore, the cellular uptake of the GP–cNGR/PVP nanosystem was revealed using fluorescent imaging. It could be expected that the cNGR-modified nanosystem may have great potential clinical applications for cancer therapy.

Fig. 1 Schematic representation of the synthesis of cNGR-modified functionalized GO nanosheets and preparation of drug delivery system GP–cNGR/PVP/CA4. Top: scheme (a); bottom: scheme (b).

2. Experimental

2.1. Chemicals and materials

Graphite (purity > 95%), N-(3-dimethylaminopropyl-N′-ethylcar-bodiimide) hydrochloride (EDC), 2-(4-morpholino) ethane sulfonic acid (MES), branched polyethylenimine (PEI, M_w = 1800 Da) and N-hydroxysuccinimide (NHS) were purchased from Adamas Reagent Co., Ltd. (Shanghai, China). Sulfuric acid (H₂SO₄, 98%), phosphoric acid (H₃PO₄, 85%), potassium permanganate (KMnO₄), hydrochloric acid (HCl), dihydrogen peroxide (H₂O₂, 30%), 3,3-dithiodipropionic acid (DTPA), dimethyl sulfoxide (DMSO) and polyvinylpyrrolidone (PVP, M_w = 30 000 Da) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Cyclic NGR peptides (c[CNGRCK]) were synthesized by Shanghai GL peptide Ltd. (Shanghai, China). Combretastatin A4 (CA4) was provided by Shanghai Ecust Biomedicine Co., Ltd. (Shanghai, China). The dialysis bags (MWCO: 8000 Da) were purchased from Spectrum Laboratories Inc. Fluorescein isothiocyanate (FITC), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and Hoechst 33242 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Fetal bovine of GIBCO Invitrogen Corp. (Carlsbad serum (FBS) and cell culture medium were the products, California, USA).

2.2. Preparation of GO

Graphene oxide (GO) was synthesized from natural graphite using a reported procedure.²⁹ Graphite powder (0.5 g) was added to a mixture of H₂SO₄ (18 mL) and H₃PO₄ (2 mL) with constant stirring. KMnO₄ (3.0 g) was added gradually to the above solution while keeping the temperature less than 20 °C. The mixture was stirred at 35 °C for 12 h, and deionized water (80 mL) was added to the resulting solution with vigorous stirring. The reactant became bright yellow by addition of 30% H₂O₂ (2 mL). For purification, the resultant mixture was washed with 5% HCl solution and deionized water several times until the solution became neutral. GO was obtained after centrifugation and drying.

2.3. Preparation of PEI functionalized GO (GP)

Branched PEI was covalently conjugated to the carboxylic acid group of GO using EDC chemistry.³⁰ EDC (50 mg) and NHS (75 mg) were added to GO solution (20 mL, 1 mg mL⁻¹) in a vial and then stirred gently for 3 h at room temperature. The PEI (100 mg) dissolved in MES buffer solution (2 mL, pH 6.0) was added to the above GO solution. The mixture was stirred at 25 °C for 48 h, followed by sonication during 24 h. The resulting solution was dialyzed against deionized water for 48 h to remove the excess reagents. GP was obtained after high-speed centrifugation.

2.4. Preparation of cNGR-modified GP (GP–cNGR)

GP–cNGR was prepared by an amidation reaction using a cross-linking reagent, i.e. 3,3-dithiodipropionic acid (DTPA).^19,31 DTPA (50 mg) was dissolved in DMSO (10 mL), EDC and NHS were added with stirring for 3 h. After the prescribed time, the above mixture solution was added to GP aqueous solution (10 mL, 1 mg mL⁻¹) and stirred for 24 h. The excess reagents were removed through centrifugation. The precipitate (GP–NHS) was dispersed in PBS solution and the pH was adjusted to 8.0 by an addition of N-methyl morpholine. cNGR (5 mg) was added and stirred for 24 h at room temperature. The unreacted peptide was removed by dialyzing against deionized water for over 48 h. The conjugation efficiency was quantified by a gradient reverse-phase high-performance liquid chromatography (RP-HPLC) method with a flow rate of 1.0 mL min⁻¹. The mobile phase consisted of H₂O-0.1% trifluoroacetic acid (eluent A) and acetonitrile-0.1% trifluoroacetic acid (eluent B). The eluted gradient was set from 2% eluent B/98% eluent A to 27% eluent B/73% eluent A over 25 min. The detected wavelength was 220 nm and the column temperature was maintained at 30 °C.

2.5. Loading of PVP onto the GP–cNGR (GP–cNGR/PVP)

PVP (40 mg) was added to the GP–cNGR solution (20 mL, 0.5 mg mL⁻¹), and the mixture was kept stirring vigorously at room temperature for 48 h. The resultant GP–cNGR/PVP was well dispersed in PBS solution.²⁶ GP/PVP was prepared in a similar manner.

2.6. Drug loading and release of GP–cNGR/PVP

CA4 (5 mg) was dispersed in 3 mL of methanol and stirred at room temperature for 30 min. Then GP–cNGR/PVP (20 mL) was dropwise added into the CA4 solution. The temperature was rapidly raised from room temperature to 60 °C, the mixture was continuously stirred for 3 h. After stirring overnight in the dark, the suspension was dialyzed against 300 mL PBS solution (pH 7.4) for 24 h to remove the methanol, and the residual CA4. The concentration of CA4 in the solution was estimated by RP-HPLC analysis, using methanol and water (68 : 32, v/v) as the mobile phase, and the detection wavelength was 295 nm with the flow rate at 30 °C.³² The CA4 loading efficiency of GP–cNGR/PVP was calculated according to the following equation:

Loading efficiency (f) = (W₀ − W_s)/W₀

where W₀ and W_s represent the initial CA4 mass and the CA4 mass in the PBS solution, respectively.

The in vitro release behaviour was evaluated by the dialysis method. 20 mL of GP/PVP/CA4, GP–cNGR/PVP/CA4 and free CA4 (0.45 mM, pH 7.4) were placed inside three dialysis bags, which were immersed in 200 mL PBS solution (pH 7.4) containing 10% fetal bovine serum (FBS), respectively. The temperature of the solution was maintained at 37 °C. The released CA4 outside of the dialysis bag was sampled at defined time period and assayed by a validated RP-HPLC method as described above. Cumulative release was expressed as the total percentage of drug released through the dialysis membrane over time.

2.7. Cytotoxicity in vitro

The biocompatibility of GP–cNGR/PVP and the selectively cytotoxicity of GP–cNGR/PVP/CA4 were evaluated by a standard MTT assay. 1 × 10⁴ cells per well of HT-1080 cells and MCF-7 cells were seeded onto a 96-well plate and then incubated for 24 h.

For toxicity studies, cells were incubated with GP–cNGR/PVP, GP/PVP/CA4, GP–cNGR/PVP/CA4 and free CA4 for 48 h at 37 °C. Then MTT was added and the plates were incubated for 4 h in the dark. 100 μL of DMSO was added into each well to dissolve the insoluble formazan crystals. The absorbance was measured using a microplate render (Biorad, 680, America) at 540 nm and the values were compared with respect to control cells.

2.8. Cellular uptake

For cell uptake examination, fluorescence probe FITC with structure of aromatic nucleus was loaded onto GP–cNGR/PVP through strong π–π stacking, by mixing FITC solution with GP–cNGR/PVP for 12 h in the dark. HT-1080 cells and MCF-7 cells were seeded in 6-well plates (2 × 10⁴ cells per well) and incubated for 24 h. GP–cNGR/PVP/FITC (5 μg mL⁻¹) was added to each well and incubated for another 3 h at 37 °C. The medium was then removed and the cells were washed with cold PBS (0.1 M, pH 7.4) three times, followed by fixing with 4% paraformaldehyde for 15 min. Nuclear staining was performed by Hoechst 33242 for 10 min at room temperature. The fluorescent images of cells were analyzed using a laser scanning confocal microscope (Leica, TCS SP2, Germany).

2.9. Characterizations

Fourier-transform infrared spectra were recorded using a Nicolet IS10 FT-IR Spectrometer (Thermo Scientific) under the transmission mode by a KBr pellet method. Transmission electron microscopy (TEM) images were acquired by a JEM-2000 EX (JEOL) electron microscope operating at 200 kV. Thermogravimetric analysis (TGA) results were obtained with a TA Instrument 2050 thermogravimetric analyzer at a heating rate of 10 °C min⁻¹ from 25 to 600 °C under a nitrogen atmosphere. The Raman signal was determined by a Lab RAM Aramis Raman Spectrometer with a DPSS laser at an excitation wavelength of 514 nm. The zeta potential and hydrodynamic size of nanosheets were made on a Malven Zetasizer Nano ZS instrument. Powder X-ray diffraction was carried out on an X-ray diffractometer (XRD, X' Pert PW3040/60) by Ni-filtered CuKα radiation with 4 degree min⁻¹ scanning rate. The contents of CA4 and cNGR were determined by a high performance liquid chromatograph equipped with an LC-20AT pump, SIL-20A autosampler, SPD-20AT PDA detector, and CTO-20A column oven (Shimadzu Corporation, Kyoto, Japan).

3. Results and discussion

3.1. Preparation and characterization of GO and GP

The water-soluble GO nanosheets were successfully synthesized according to a previously reported method,²⁶ which was confirmed by the presence of the oxygen-containing functional groups. Fig. 2c shows the FT-IR spectra of the GO nanosheets in the range from 4000 cm⁻¹ to 500 cm⁻¹. The absorbance peaks at 3424 cm⁻¹ and 1732 cm⁻¹ corresponding to O–H and C=O stretching vibration of the formation of carboxylic structures were observed in the FT-IR spectrum. The peaks at 1092 cm⁻¹, 1260 cm⁻¹, 1383 cm⁻¹ and 1625 cm⁻¹ could be attributed to the C–O–C stretching vibrations, C–O stretching, C–OH stretching and C–C stretching mode of the sp² carbon, respectively.^10,33

Fig. 2 (a) TEM image of GO. (b) TEM image of GP. (c) FT-IR spectra of GO and GP. (d) TGA curves of GO and GP. Digital photographs of GO and GP dispersed in PBS (pH 7.4).

There are a few carboxylic groups in the GO nanosheets. Further modified GO with –NH₂ is necessary to improve the water solubility and reaction sites of the graphene derivatives and to facilitate chemical binding of the targeting ligand to GP via EDC chemistry. The characteristic amide-carbonyl for the stretching vibration of C=O vibration for the NH–CO group at 1636 cm⁻¹ was consistent with the PEI molecules on GO nanosheets (Fig. 2c).³⁴ The new peaks at 2958 cm⁻¹ and 1563 cm⁻¹ could be assigned to the typical stretching vibration of C–H and the well-defined characteristic absorbance band of primary amine, respectively.

The morphological information on the obtained GO and GP was investigated by TEM observation. Fig. 2a shows the pristine GO precursor is a planar backbone conformation and transparency but with a small amount of wrinkles. TEM image indicated the quality of our GO nanosheets was very good.³⁵ Functionalization of GO truncated it from several hundreds of nanometers to less than 200 nm in lateral dimensions (Fig. 2b), which could be due to the size reduction by the repeated sonication and the folding and re-forming of the GO nanosheets during the EDC/NHS-mediated reaction process.^36,37

The relative amount of PEI grafted onto the surface of GO was tested by TGA (Fig. 2d). At 400 °C, GO and GP showed about 34.8% and 59.3% weight losses, respectively, thus the relative amount of PEI grafted onto GO was calculated to be 24.5%.³⁸ Most importantly, owing to the incorporation of PEI moieties, GP could be readily dispersed in PBS (pH 7.4) with the aid of slight ultrasound. Compared with GP, it was interestingly found that GO nanosheets could not be well dispersed and aggregated at once. These findings implied that the obtained GP had excellent water solubility.

3.2. Preparation and characterization of GP–cNGR

To modify GP with cNGR peptide, DTPA was attached to the amino groups on GP surface (GP–NHS), and cNGR was conjugated to the attached DTPA through the amide bonds between carboxyl of DTPA and amino of cNGR. Fig. 1a shows the preparation of GP–cNGR. The successful conjugation of cNGR to GP was confirmed by the RP-HPLC analysis (Fig. 3a). For the reaction between peptide and nanosheets at a weight ratio of 1 : 2, the peak area of cNGR significantly decreased after 24 h, indicating that a considerable amount of free peptide was conjugated. The conjugation efficiency was calculated to be about 80.4%.

Fig. 3 (a) HPLC analysis of unconjugated peptide: (1) standard of cNGR in mobile phase, (2) the reaction mixture. (b) FT-IR and (c) Raman spectra of GO, GP–NHS and GP–cNGR. (d) Zeta potentials of the different nanosystem. Error bars were based on triplet samples.

The attachment of DTPA was confirmed by the FT-IR spectra (Fig. 3b). A new peak was found in the IR spectrum of the GP–NHS at 1249 cm⁻¹ for the stretching vibration of the C–S and C–O groups of activated DTPA, indicating that DTPA was successfully grafted onto GP. After GP–NHS was conjugated with cNGR, the peak at 1566 cm⁻¹ corresponding to N–H of CO–NH shifted to 1559 cm⁻¹ in the GP–cNGR spectrum because the proportion of CO–NH grew in the GP–NHS. Meanwhile, the peak at 1401 cm⁻¹ (C–N, stretching vibration) might emerge in the FT-IR spectra of GP–cNGR. These results suggested the successful cNGR conjugation on the surface of the GP nanosheets.

Two distinct peaks were observed for each sample at around 1589 cm⁻¹ and around 1345 cm⁻¹ corresponding to the G band and D band (Fig. 3c), respectively.³⁹ In the Raman spectra, the existence of the G-band suggests the well-defined graphitic domains, and the I_G/I_D ratio is nearly proportional to graphitization degree.⁴⁰ As observed, the I_G/I_D ratio changed slightly after cNGR conjugated to the functionalized GO, suggesting that further minor modification in GP–NHS moiety would take place of chemical interaction with the peptide. The I_G/I_D ratio for GP–cNGR suggested the graphitic structure was well preserved, and thus GP–cNGR also could load drugs via π–π stacking and hydrogen-bonding interactions as similar as GO.

3.3. Characterization of GP–cNGR/PVP

To make GP–cNGR more stable under physiological conditions, PVP was deposited onto the GP–cNGR nanosheets via non-covalent interaction. The surface modification on GP–cNGR could be reflected by the change of the zeta potential. The zeta potential is the mutual repulsion between particles or measuring the intensity of attraction. As shown in Fig. 3d, the zeta potential decreased from −25.7 ± 1.5 mV (GP–cNGR) to −31.6 ± 1.8 mV after PVP loaded onto the GP–cNGR, mainly due to the PVP could serve as a biocompatible stabilizer of GP–cNGR in PBS solution.

3.4. Drug loading and release of GP–cNGR/PVP

The loading and release profile of GP–cNGR/PVP is the most important performance indictor of the drug delivery materials. The loading efficiency of CA4 on GP–cNGR/PVP was found to be 0.5630 ± 0.0132 mg mg⁻¹. This could be mainly ascribed to π–π stacking interaction between GO and aromatic rings drug molecules. Meanwhile, there may be hydrogen-bond or hydrophobic interaction between CA4 and PVP. The result of XRD analysis is shown in Fig. 4c. It was clear that the physical mixture of CA4 and GP–cNGR/PVP blank delivery system showed significant drug crystal peak as compared to that of CA4 powder. However, the XRD characteristic peak between GP–cNGR/PVP and GP–cNGR/PVP/CA4 was identical.⁴¹ These findings indicated that CA4 could exist as the amorphous or molecule state in the GP–cNGR/PVP nanosheets.

Fig. 4 (a) TEM image of GP–cNGR/PVP/CA4. (b) Size distributions of GP–cNGR/PVP/CA4 in PBS of pH 7.4 in time measured by DLS. (c) XRD patterns. (1) CA4 powder. (2) GP–cNGR/PVP. (3) Physical mixture of CA4 and GP–cNGR/PVP. (4) GP–cNGR/PVP/CA4 nanosystem. (d) In vitro release in PBS (pH 7.4) containing 10% FBS. Error bars were based on triplet samples.

The small size is very important to ensure materials as drug delivery system for cancer therapy. Visualized in TEM (Fig. 4a), the GP–cNGR/PVP/CA4 nanosheets were seen to be spherical nanoparticles, which could be due to the hydrophilic chain surrounding the nanosheets. The average particle size of the nanosystem was about 204.6 ± 2.7 nm and there was no distinct change in the size of the GP–cNGR/PVP/CA4 in PBS solution of pH 7.4 in 72 hours (Fig. 4b), suggesting its excellent dispersion stability under physiological condition.

In vitro release experiments were carried out at phosphate buffered saline (PBS, pH 7.4) solution with 10% fetal bovine serum (FBS) solution at 37 °C. As shown in Fig. 4d, compared to the quick release of CA4 nearly 93% within 12 h, the cumulative releases of CA4 molecules from GP–cNGR/PVP and GP/PVP were calculated to be about 53.7% and 51.9% within 24 h respectively. These were ascribed to PEI and PVP could offer a diffusion barrier to retard CA4 release, as well as the strong π–π stacking interaction between CA4 and GO. These results indicated that GP–cNGR/PVP could effectively load and release CA4 with a sustained-release manner.

3.5. Cytotoxicity in vitro

The potential nonspecific toxicity of GP–cNGR/PVP is always a great concern for drug delivery materials used for in vitro and in vivo application. It was found that the cell viability of GP–cNGR/PVP was not less than 85% after 48 h of incubation, even when the concentration was increased up to 40 μg mL⁻¹ (Fig. 5a). These results suggested that GP–cNGR/PVP did not have any obvious cytotoxicity on either HT-1080 or MCF-7 cells, indicating favorable cytocompatibility, which was due to the excellent biocompatibility of PVP and GO.

Fig. 5 The cell viability of HT-1080 cells (target cells) and MCF-7 cells (control cells) after treatment with GP–cNGR/PVP, GP/PVP/CA4, GP–cNGR/PVP/CA4 and free CA4 at different concentrations for 48 h. Error bars were based on triplet samples.

The cytotoxicity of free CA4, GP–cNGR/PVP/CA4 and GP/PVP/CA4 to HT-1080 cells and MCF-7 cells was next determined in vitro. All nanosystem showed a dose-dependent cytotoxicity against respective cells with concentration from 0 to 10 ng mL⁻¹ of CA4. As shown in Fig. 5b, the cell viability tests showed that GP–cNGR/PVP/CA4 displayed obvious difference between HT-1080 cells and MCF-7 cells. At a CA4 concentration of 10 ng mL⁻¹, the HT-1080 cells viability of GP–cNGR/PVP/CA4 was 37.4%, while the MCF-7 cells viability was 55.0%. This revealed that GP–cNGR/PVP/CA4 was much more sensitive to HT-1080 cells as compare to MCF-7 cells.

Due to cNGR has small molecular weight, affording the formed nanoparticle with approximately similar conformation or size to the non-modified ones. Therefore, GP/PVP/CA4 and GP–cNGR/PVP/CA4 had the same drug-loading and release mechanism according to the release experiments. As for MCF-7 cells (Fig. 5c), GP–cNGR/PVP/CA4 and free CA4 displayed similar anticancer activity, slightly lower than GP/PVP/CA4 with increasing CA4 concentration. This meant that the cNGR peptide had no obvious target-specific to the CD13 low-expressed tumor cells.

As shown in Fig. 5d, with the same concentration of CA4 administration (2.5 ng mL⁻¹), free CA4 and GP/PVP/CA4 had similar cytotoxicity effect on HT-1080 cells and the respective cell viabilities were 80.1% and 86.0%, while in the presence of cNGR peptide, the cell viability of GP–cNGR/PVP/CA4 was dramatically decreased to 63.0%. These results suggested that the targeted GP–cNGR/PVP/CA4 showed a remarkable higher cytotoxicity than free CA4 and the non-targeted GP/PVP/CA4, even though the release amount of CA4 from GP–cNGR/PVP/CA4 was much lower than that of free CA4 at 48 h, which might be mainly ascribed to the plentiful of CD13 receptors on the surface of HT-1080 cells. Therefore, we concluded that the CA4-loaded GP–cNGR/PVP nanosystem could significantly enhance the specific antitumor activity at the same CA4 concentration. Meanwhile, cNGR peptide played an important role in generating receptor-mediated specificity for the selective cytotoxicity of HT-1080 cells.

3.6. Cellular uptake assay

To evaluate the specific affinity for HT-1080 cells, the fluorescent GP–cNGR/PVP/FITC was used to measure cellular uptake in MCF-7 cells and HT-1080 cells by a laser scanning confocal microscope (Fig. 6). Here, the blue fluorescence of nuclei was employed to localize the position of nanosystem (green fluorescence) and representing that the nanosystem accumulated in membrane or cytoplasm or nuclei. The results showed that, in HT-1080 cells, GP–cNGR/PVP/FITC exhibited higher fluorescence intensity, suggested that a great amount of nanosheets had entered tumor cells. However, the fluorescence intensity in MCF-7 cells was much weaker as compared to HT-1080 cells at the same conditions. In brief, the effective cellular uptake and internalization suggested the potential of GP–cNGR/PVP as a transmembrane drug delivery carrier to promote the cellular specific uptake, decrease the drug side-effect, increase the drug intracellular accumulation and enhance local killing effect. Along with the cytotoxicity assay as shown in Fig. 5, we could demonstrate specific uptake of the GP–cNGR/PVP nanosystem by HT-1080 cells via receptor-mediated endocytosis.

Fig. 6 Fluorescence images of HT-1080 cells and MCF-7 cells incubated with 5 μg mL⁻¹ of GP–cNGR/PVP/FITC for 3 h.

4. Conclusions

In summary, we have successfully developed the targeted drug delivery system functionalized with cNGR peptide as the ligand for CD13 receptor overexpressed in tumor cells. The GP–cNGR/PVP nanosystem possessed excellent physiological stability, high drug loading efficiency and sustained release of anticancer drugs. In vitro cytotoxicity study showed that the cell viability of GP–cNGR/PVP/CA4-treated HT-1080 cells was lower about 20% compared with MCF-7 cells. On the other hand, the HT-1080 cell viability of GP–cNGR/PVP/CA4 was much lower than free CA4 and GP/PVP/CA4 at the same CA4 concentration. In vitro cellular uptake result showed that GP–cNGR/PVP could specifically uptake by HT-1080 cells, with a much weaker fluorescence intensity to MCF-7 cells, confirming their ability of targeting delivery. The overall investigations demonstrate that there is a great potential of GP–cNGR/PVP as a drug delivery system for targeting the CD13-overexpressed specific tumor cells.

Acknowledgements

The authors acknowledge Shanghai Excellent Young Program (no. 4521ZK11YQ02) and the National Science Foundation of China (NSFC) (no. 21172148, 21472126) funding support.

Notes and references

1. H. Pandey, V. Parashar, R. Parashar, R. Prakash, P. W. Ramteke and A. C. Pandey, Nanoscale, 2011, 3, 4104.
2. W. Miao, G. Shim, S. Lee, Y. S. Choe and Y. K. Oh, Biomaterials, 2013, 34, 3402.
3. H. X. Wu, H. L. Shi, Y. P. Wang, X. Q. Jia, C. Z. Tang, J. M. Zhang and S. P. Yang, Carbon, 2014, 69, 379.
4. X. Fan, G. Jiao, L. Gao, P. Jin and X. Li, J. Mater. Chem. B, 2013, 1, 2658.
5. J. An, Y. Gou, C. Yang, F. Hu and C. Wang, Mater. Sci. Eng. C, 2013, 33, 2827.
6. Y. Pan, H. Bao, N. G. Sahoo, T. Wu and L. Li, Adv. Funct. Mater., 2011, 21, 2754.
7. X. Y. Yang, X. Y. Zhang, Z. F. Liu, Y. F. Ma, Y. Huang and Y. Chen, J. Phys. Chem. C, 2008, 112, 17554.
8. Z. Liu, J. T. Robinson, X. M. Sun and H. J. Dai, J. Am. Chem. Soc., 2008, 130, 10876.
9. G. C. Wei, R. H. Dong, D. Wang, L. Feng, S. Dong, A. Song and J. C. Hao, New J. Chem., 2014, 38, 140.
10. X. B. Zhao and P. Liu, RSC Adv., 2014, 4, 24232.
11. X. Jiang, X. Sha, H. Xin, L. Chen, X. Gao, X. Wang, K. Law, J. Gu, Y. Chen, Y. Jiang, X. Ren, Q. Ren and X. Fang, Biomaterials, 2011, 32, 9457.
12. P. Mina-Osorio, Trends Mol. Med., 2008, 14, 361.
13. L. Guzman-Rojas, R. Rangel, A. Salameh, J. K. Edwards, E. Dondossola, Y.-G. Kim, A. Saghatelian, R. J. Giordano, M. G. Kolonin and F. I. Staquicini, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 1637.
14. W. Arap, M. G. Kolonin, M. Trepel, J. Lahdenranta, M. Cardó-Vila and R. J. Giordano, Nat. Med., 2012, 8, 121.
15. Y. F. Yang, Y. Yang, X. Y. Xie, X. S. Cai and H. Zhang, Biomaterials, 2014, 35, 4368.
16. L. M. Luo, Y. Huang, B. X. Zhao, X. Zhao and Y. Duan, Biomaterials, 2013, 34, 1102.
17. X. Wang, Y. Wang, X. Chen, J. C. Wang, X. Zhang and Q. Zhang, J. Controlled Release, 2009, 139, 56.
18. L. Wang, J. J. Shi, X. Jia, R. Y. Liu, H. H. Wang, Z. Z. Wang, L. L. Li, J. Zhang, C. F. Zhang and Z. Z. Zhang, Pharm. Res., 2013, 30, 2757.
19. M. Gupta, G. Chashoo, P. R. Sharma and A. K. Saxena, Mol. Pharm., 2014, 11, 697.
20. G. M. Tozer, C. Kanthou, C. S. Parkins and S. A. Hill, Int. J. Exp. Pathol., 2002, 83, 21.
21. J. Griggs, R. Hesketh and G. A. Smith, Br. J. Cancer, 2001, 84, 832.
22. S. M. Nabha, N. R. Wall, R. M. Mohammad, G. R. Pettit and A. M. Al-Katib, Anticancer Drugs., 2000, 11, 385.
23. A. El-Zayat, D. Degen, S. Drabek, G. M. Clark, G. R. Pettit and D. D. Von Hoff, Anticancer Drugs, 1993, 4, 19.
24. N. J. Lawrence, D. Rennison, M. Woo, A. T. Mcgown and J. A. Hadfield, Bioorg. Med. Chem. Lett., 2001, 11, 51.
25. S. L. Young and D. J. Chaplin, Expert Opin. Invest. Drugs, 2004, 13, 1171.
26. X. C. Qin, Z. Y. Guo, W. Liu, W. Zhang, M. M. Wan and B. W. Yang, J. Photochem. Photobiol., B, 2013, 120, 156.
27. Y. Van Hensbergen, H. J. Broxterman, Y. W. Elderkamp, J. Lankelma, J. C. Beers and M. Heijn, Biochem. Pharmacol., 2002, 63, 897.
28. A. Von Wallbrunn, J. Waldeck, C. Höltke, M. Zühlsdorf, R. Mesters and W. Heindel, J. Biomed. Opt., 2008, 13, 011007.
29. D. C. Marcano, D. V. Kosynkin, J. M. Berlin, A. Sinitskii, Z. Sun and A. Slesarev, ACS Nano, 2010, 4, 4806.
30. G. J. Xu, S. J. Liu, H. Niu, W. P. Lv and R. A. Wu, RSC Adv., 2014, 4, 33986.
31. H. Chen, Z. Y. Wang, S. F. Zong, L. Wu, P. Chen, D. Zhu, C. L. Wang, S. H. Xu and Y. P. Cui, ACS Appl. Mater. Interfaces, 2014, 6, 17526.
32. J. Wang, Q. Q. Tian, F. Ding, Y. Y. Yu and F. H. Wu, RSC Adv., 2015, 5, 40258.
33. Y. F. Jia and K. M. Thomas, Langmuir, 2000, 16, 1114.
34. G. C. Wei, M. M. Yan, R. H. Dong, D. Wang, X. Z. Zhou, J. F. Chen and J. C. Hao, Chem.–Eur. J., 2012, 18, 14708.
35. K. Hyunwoo, L. Duhwan, K. Jinhwan, K. Tae-il and K. J. Won, ACS Nano, 2013, 7, 6735.
36. D. Yin, Y. Li, H. Lin, B. F. Guo, Y. W. Du, X. Li, H. J. Jia, X. J. Zhao, J. Tang and L. Zhang, Nanotechnology, 2013, 24, 105102.
37. J. T. Robinson, S. M. Tabakman, Y. Y. Liang, H. L. Wang, H. S. Casalongue, D. Vinh and H. J. Dai, J. Am. Chem. Soc., 2011, 133, 6825.
38. J. Shi, H. Zhang, L. Wang, L. Li, H. Wang and Z. Wang, Biomaterials, 2013, 34, 251.
39. K. N. Kudin, B. Ozbas, H. C. Schniepp, R. K. Prud'homme, I. A. Aksay and R. Car, Nano Lett., 2008, 8, 36.
40. I. K. Moon, J. Lee, R. S. Ruoff and H. Lee, Nat. Commun., 2010, 1, 73.
41. T. Y. Yang, Y. G. Wang, Z. Q. Li, W. B. Dai, J. Yin, L. Liang, X. Ying, S. F. Zhou, J. C. Wang, X. Zhang and Q. Zhang, Nanomedicine, 2012, 8, 81.

---