Xubo Zhao and
Peng Liu*
State Key Laboratory of Applied Organic Chemistry and Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000, China. E-mail: pliu@lzu.edu.cn; Fax: +86 0931 8912582; Tel: +86 0931 8912582
First published on 16th April 2014
A novel graphene oxide (GO)-based nanocarrier has been designed for the targeting and pH-responsive controlled release of anti-cancer drugs via the classic amidation of the carboxyl groups of carboxylated graphene oxide (CG) with the amine end-groups of functional poly(ethylene glycol) (PEG) terminated with an amino group and a folic acid group (FA–PEG–NH2). The carboxylated graphene oxide conjugated folate-terminated poly(ethylene glycol) (CG–PEG–FA) nanocarrier containing 44.4 wt% of functional PEG brushes exhibits stable dispersibility in PBS media, outstanding cytocompatibility, high drug-loading capacity (0.3993 mg mg−1 for DOX) via π–π stacking interactions, perfect folate receptor-targeting and pH-activated controlled release properties, demonstrating that the nanocarrier could be a promising drug delivery system (DDS) for cancer therapy.
In recent years, many experiments have been done on optimizing the properties of GO to obtain new derivatives for biomedical applications.6 Among them, grafting polymer moieties seems to be an efficient approach.7 Dai et al. reported the first pioneering paper on six-armed PEG-amine star polymer grafted GO as a novel drug nanocarrier for water-insoluble anti-cancer drugs via hydrophobic interactions and π–π stacking interactions in 2008.8 Meanwhile, the in vivo behaviors of PEGylated GO derivatives after intravenous injection or inhalation, and the surface coating & size dependent bio-distribution and toxicology profiles were investigated. This important fundamental study has offered a deeper understanding of the in vivo behavior and toxicology of the functionalized graphene nanomaterials in animals, depending on their different administration routes.9 The most intriguing properties of these GO derivatives are their remarkable solubility and stability in physiological media and their biocompatibility, making them promising substrate biomaterials for controlled drug delivery.
The targeting function of drug delivery systems (DDSs) has recently attracted considerable attention because most of the commonly used anti-cancer drugs have serious side effects due to their non-specific activity on healthy cells.10 By utilizing antibodies or specific ligands, DDSs along with the therapeutic agent can selectively bind to targeted cells, and then be delivered to the interior of a given type of cells via receptor mediated endocytosis.11 Folic acid (FA), which has a very high affinity for folate receptors (FRs), is a promising candidate for cancer-cell targeting towards several human cancer cells which over-express FRs, such as breast, ovary, lung, kidney, and endometrium cancers.12 Moreover, owing to its high stability, low cost and ability to conjugate with large amounts of molecules, FA has received considerable attention as a targeting agent for the imaging and therapy of cancer.13 It was also reported that the multivalent targeting effect could dramatically enhance the biological targeting ability.14
In the present work, a carboxylated graphene oxide conjugated folate-terminated poly(ethylene glycol) (CG–PEG–FA) nanocarrier was designed for the targeted controlled release of anti-cancer drugs by grafting folate-terminated poly(ethylene glycol) (FA–PEG–NH2) onto the carboxylated graphene oxide nanosheets (CG) via classic amidation (Scheme 1). Herein, PEG brushes were introduced to the carboxylated graphene oxide nanosheets (CG), which rendered them stable under phosphate buffer saline (PBS), and folic acid (FA) moieties were attached to the CG for targeting specific cells with folate receptors. Through the transformation of the hydroxyl groups of the GO nanosheets into carboxylic acid groups, more PEG brushes as well as FA targeting moieties could be immobilized onto the functional graphene derivative. Thus, excellent dispersibility, cytocompatibility, and targeting specificity are expected for the nanocarrier for the specifically targeted controlled release of anti-cancer drugs.
![]() | ||
Scheme 1 Schematic illustration of the preparation of the biocompatible graphene oxide nanocarrier (CG–PEG–FA). |
Bifunctional PEG (NH2–PEG–NH2) and monofunctional PEG (PEG–NH2) of 2000 Daltons molecular mass were provided by Beijing Kaizheng Biological Engineering Development Co., Ltd. (Beijing, China).
1-Ethyl-3-(3-dimethyl aminopropyl) carbodiimide hydrochloride (EDC·HCl) was purchased from Fluorochem. N-Hydroxylsuccinimide (NHS) was purchased from Aladdin Chemistry Co. Ltd. Doxorubicin hydrochloride (DOX·HCl) was purchased from Beijing Huafeng United Technology Co. Ltd. Dimethyl sulfoxide (DMSO), folic acid (FA), potassium permanganate (KMnO4), sodium nitrate (NaNO3), phosphorus pentoxide (P2O5), potassium dichromate (K2Cr2O7), and sulfuric acid (H2SO4, 98%) were analytical reagent grade and were purchased from Tianjin Chemical Company, China. Double distilled water was used throughout.
This pretreated graphite was then subjected to oxidation by the Hummers method. The pretreated graphite powder (1.01 g) was placed in two round bottom flasks with NaNO3 (1.00 g) and concentrated H2SO4 (50.1 mL) at 0 °C. KMnO4 (4.01 g) was added gradually with stirring over about 1 h while keeping the temperature of the mixture around 0 °C in an ice-water bath. After the mixture was stirred vigorously for 2 days at room temperature, an aqueous H2SO4 solution (5 wt%, 100 mL) was added over about 1 h with stirring, and the temperature was kept at 98 °C. The resultant mixture was further stirred for 2 h at 98 °C. The temperature was then reduced to 60 °C and 30% H2O2 (3.2 mL) was added, the color of the mixture changed to bright yellow, as reported by Kovtyukhova et al.16 and the mixture was stirred for 2 h at room temperature. For purification, the mixture was centrifuged and washed with 10% HCl solution and then with deionized water to remove residual metal ions until the solution became pH neutral, after which individual GO nanosheets were stably dispersed in deionized water. The resulting GO nanosheets were dried at 65 °C in vacuum.
The aqueous suspension of GO (100 mL, 2 mg mL−1) was ultrasonicated for 1 h to obtain a clear suspension. NaOH (10.00 g) and chloroacetic acid (ClCH2COOH) (10.00 g) were then added to the GO suspension and ultrasonicated for another 3 h to convert the hydroxyl and epoxy groups on the GO nanosheets into carboxyl groups. The resulting carboxylated graphene oxide (CG) suspension was neutralized, and purified by repeating the cycle of rinsing and filtration. Then, the CG suspension was dialyzed against distilled water for over 48 h to remove any ions, followed by drying at 65 °C in vacuum.17
For comparison, monofunctional PEG (PEG–NH2) was also conjugated onto the CG nanosheets with the same procedure. The product (CG–PEG) was used to examine the folate receptor-targeting function of the CG–PEG–FA nanocarrier.
Cumulative release (%) = cumulative amount released/total mass loaded × 100% |
For drug-loading, solutions of doxorubicin (DOX) (1.0 mg mL−1) were prepared in deionized water. The CG–PEG–FA nanocarrier (∼10.0 mg) was added into 5.0 mL doxorubicin solutions (pH 5.0, 6.5, 7.4 or 8.5) for drug loading. After being swung on a table concentrator for 24 h, the different doxorubicin-loaded CG–PEG–FA nanocarriers (CG–PEG–FA/DOX) were centrifuged to remove the excess DOX. Compared with the CG–PEG–FA nanocarrier, the doxorubicin (DOX) loaded CG nanocarrier was studied in a DOX solution at pH 7.4. The drug concentration in the supernatant solution was monitored using an ultraviolet (UV) spectrophotometer at 233 nm to assess the drug-loading capacities. The drug-loading capacities of the GO and the CG–PEG–FA nanocarriers were calculated from the drug concentrations before and after loading.
As for their controlled release performance, the CG–PEG–FA/DOX dispersions (10 mL) in phosphate-buffered saline (PBS at pH 7.4, 6.5 or 5.0) were transferred to dialysis tubes (molecular weight cutoff of 10000), and immersed in 120 mL PBS at 37 °C at pH 7.4, 6.5 or 5.0, respectively. Aliquots (5.0 mL) of solution were removed at certain intervals, and the drug concentrations in the dialysates were analyzed using a UV spectrophotometer to assess the cumulative release of the drug-loaded CG–PEG–FA nanocarrier. 5.0 mL fresh PBS was added after each sampling to keep the total volume of the solution constant. The cumulative release was expressed as the total percentage of drug that was released from the drug-loaded CG–PEG–FA nanocarrier and was transported through the dialysis membrane over time.
The drug release data obtained from the in vitro release study was analyzed for the rate of release, using the Higuchi drug release equation given below:
Mt = k·t1/2 |
When a plot of cumulative drug release of t1/2 yields a straight line with a slope which possesses a value ≥1, a system is considered to follow Higuchi kinetics of drug release.19
Mt/M∞ = k·tn |
For the MTT assay, the cells were seeded into 96-well plates at densities of 1 × 105 cells per well for 24 h. Then, the CG–PEG–FA nanocarrier with different concentrations, drug-loaded CG–PEG–FA and CG–PEG nanocarriers, and DOX were added to the cells and incubated for 48 h. Thereafter, the cells were washed three times with phosphate buffered saline (PBS) and processed for the MTT assay to determine the cell viability. 100 μL of pH 7.4 PBS solution containing 20 μL of 5 mg mL−1 MTT was added to each well, and incubated for an additional 4 h. Cell bound dye was dissolved with 100 μL DMSO in each well cell culture plate and these were swung on a table concentrator for 20 min, The absorbance of each well was read on a microplate reader using a test wavelength of 490 nm.
Raman spectra were obtained with a Horiba Jobin-Yvon LabRAM HR 800 UV apparatus using an excitation laser with a wavelength of 532 nm.
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 800 °C under a nitrogen atmosphere.
The morphologies of the biocompatible CG–PEG–FA nanocarriers were characterized with a JEM-1200 EX/S transmission electron microscope and an SPA-300HV atomic force microscope. The samples were dispersed in water and then deposited on a copper grid covered with a perforated carbon film and deposited silicon wafer, followed by drying at 45 °C in vacuum.
The GO was synthesized using a modified Hummers method from natural graphite powder15 and this was proved by the appearance of the absorbance peak at 1706 cm−1 of the CO stretch band of the carboxyl groups (Fig. 1). The absorbance peaks at 1041, 1391, 1623 and 1706 can be attributed to C–O stretching (epoxy or alkoxy), O–H stretching (carboxyl), C
C skeletal vibrations of unoxidized graphite domains, and C
O in carboxylic acid and carbonyl moieties, respectively.
There are plentiful hydroxyl and epoxy groups in the GO nanosheets. It is necessary to convert these groups into COOH groups to improve the aqueous solubility and reaction sites of the graphene derivatives and to facilitate chemical binding of the functional PEG to the CG via EDC chemistry. In the present study, the hydroxyl and epoxy groups of the GO were converted into COOH groups by mixing the GO with ClCH2COOH under strongly alkaline conditions according to the literature.17 It is interesting that the color of the GO suspension changed from dark-brown to black during the conversion process, this may be due to the partial reduction of the GO under strongly alkaline conditions.22 The presence of –CH2COOH groups in the carboxylated graphene oxide (CG) was confirmed by the FT-IR spectra (Fig. 1). A new peak is found in the IR spectrum of the CG at 1088 cm−1 for the stretching vibration of the C–O–C groups, compared with the graphene oxide (GO). This indicated that the hydroxyl and epoxy groups of the GO were successfully converted into –COOH groups. The peak at 1706 cm−1 corresponding to CO of –COOH on the GO shifted to 1700 cm−1 in the CG because the proportion of –CH2COOH grew in the CG.
Next, a hydrophilic biocompatible polymer, PEG conjugated with a targeting ligand (folic acid (FA)), was synthesized for the surface modification of the CG nanosheets to improve their cytocompatibility and targeting specificity. The presence of FA moieties in the FA–PEG–NH2 was confirmed by FT-IR measurements. The characteristic peak of FA at 1585 cm−1 can be clearly observed in Fig. S1, ESI†, and is slightly shifted in the FT-IR spectrum of FA–PEG–NH2. Also, an obvious peak at 1645 cm−1 corresponding to the characteristic peak of N–O in the FT-IR spectrum of the folic acid active ester was observed.23 This finding suggested that FA was successfully conjugated with the NH2–PEG–NH2.
After the CG was conjugated with FA–PEG–NH2, the peak at 1700 cm−1 shifted to 1696 cm−1 in the CG–PEG–FA spectrum due to the hydrogen bonds formed between the carboxyl groups of folic acid moieties and carboxyl groups on the GC. At the same time, the peak at 1088 cm−1 corresponding to the C–O–C characteristic absorbance might emerge in the FT-IR spectra of the biocompatible and specific targeting CG–PEG–FA nanocarriers. Meanwhile, the conjugation of FA–PEG–NH2 onto the CG through the formation of an amide bond was confirmed by the strong NH–CO stretching vibration (1634 cm−1). These results suggested that a biocompatible and specifically targeted CG–PEG–FA nanocarrier was successful designed via the covalent attachment of functional polymer brushes.
Two new absorbance bands at 283 and 380 nm, which can't be seen in the UV-vis absorption spectrum of the CO nanosheets, appeared in that of the CG–PEG–FA (Fig. 2), attributed to the characteristic absorbance of the pterin ring in FA.24 However, the GO only displayed a small absorbance peak at about 233 nm due to the π–π* of CC.25 This also verified the successful grafting of the folate-terminated poly(ethylene glycol) (FA–PEG–NH2) onto the CG nanosheets via facile amidation (Scheme 1).
The TGA curves of the functional graphene nanosheets are shown in Fig. 3. The GO showed a large weight loss (more than 30%) at temperatures lower than 200 °C, attributed to the removal of the water that is held in the material, and the functional groups (–OH and –COOH), from the GO.26 As for the CG nanosheets, the weight loss in this temperature range decreased to about 15%, and a sharp weight loss near 20% occurred around 200 °C, resulting from the decomposition of the organic groups (hydroxyl, epoxy and carboxyl groups) of the CG nanosheets.27
There was a 60 wt% weight loss at 450 °C for the CG–PEG–FA, whereas the GO and CG nanosheets exhibited weight losses of 37 wt% and 46 wt%, respectively. Moreover, a weight loss of about 40 wt% occurred at 220–450 °C in the curve of the CG–PEG–FA, which may be due to the decomposition of the FA–PEG–NH2 polymer brushes on the CG–PEG–FA nanocarrier. So we can calculate that the CG–PEG–FA nanocarrier contained about 44.4 wt% FA–PEG–NH2 polymer moieties from the TGA results (Fig. 3). This value is twice that of the result from grafting PEG400 onto the GO directly,28 although PEG2000 was used in the present work, for which the bigger steric hindrance resulting from the higher molecular weight hinders the grafting. This revealed that carboxylation of the GO is an efficient method to increase the degree of functionalization of the GO nanosheets.
Transmission electron microscopy (TEM) and atomic force microscopy (AFM) images (Fig. 4) provide morphological information on the GO and the CG–PEG–FA nanocarrier. A typical TEM image of the GO showed the see-through flaky material with a size of less than 2 μm in lateral width. This is characteristic of GO as a single or two layered sheets.29
A large number of new groove-like and ridge-like morphologies were found in the AFM image of the CG–PEG–FA compared with the GO. The graphene oxide (GO) sheets had very sharp edges and flat surfaces. In contrast, the edges of the CG–PEG–FA nanocarrier appeared relatively groove-like and some ridge-like structures were observed on the surfaces, which were formed by the polymer wrapping and folding on the surfaces.6 This also showed that functional PEG brushes had been successfully grafted onto the surface of the GO nanosheets.
Owing to the incorporation of FA–PEG–NH2 polymer brushes, the CG–PEG–FA can be readily dissolved in phosphate buffer saline (PBS, pH = 7.4) with the aid of ultrasonication. Fig. 5(A) and (B) show the dispersion states of the GO and the CG–PEG–FA in phosphate buffer saline (PBS) at the same concentration (0.4 mg mL−1) after 0.5 and 48 h at 37 °C, respectively. It was found that GO could not be well dispersed in PBS (pH 7.4) and precipitated 0.5 h after stopping ultrasonication. Interestingly, the solubility of the CG–PEG–FA nanocarrier in PBS (pH 7.4) was good due to the good solubility of the –PEG–FA polymer brushes. The excellent stability of the dispersion in PBS (pH 7.4) at 37 °C encouraged us to explore the applications of CG–PEG–FA nanocarriers in controlled loading and drug delivery.
The highest loading capacity of the CG–PEG–FA was lower than that of the GO, which was 0.4546 ± 0.0287 mg mg−1 under the same conditions (Fig. S2B, ESI†). It may be because the larger functional PEG brushes on the CG reduced the interaction between the drug molecules and the nanocarrier. In the Raman analysis (Fig. 6), the G band (1581 cm−1) corresponding to the sp2 hybridized carbon shifted to 1606 cm−1 after the drug-loading, indicating that the drug-loading might occur via π–π stacking interactions between DOX and the aromatic structure on the CG–PEG–FA.31 It could also explain the drug-loading difference between the GO and the CG–PEG–FA.
![]() | ||
Fig. 7 Cumulative release of DOX from the DOX-loaded CG–PEG–FA at phosphate buffer (pH 5.0) (a), (pH 6.5) (b) and (pH 7.4) (c) at 37 °C. |
The DOX release data was analyzed using the Higuchi and Korsmeyer–Peppas equations. The release rates k and n of each model were calculated by linear regression analysis. Coefficients of correlation (R2) were used to evaluate the accuracy of the fitting. The plots for the Higuchi equation of the CG–PEG–FA/DOX nanocarrier (pH 7.4, 6.5 or 5.0 at 37 °C) resulted in linearity with an R2 value of 0.9421, 0.9762 and 0.9524, and a k value of 0.1542, 0.2004 and 0.5592 (Fig. S3 ESI†), respectively. However, the k values of the Higuchi equation were less than 1, so Fickian diffusion could not be used to describe the drug release of the CG–PEG–FA/DOX nanocarrier at pH 7.4. But even more crucially, the plots of the Korsmeyer–Peppas equation were used to described DOX release mechanism of the CG–PEG–FA/DOX nanocarrier in PBS (pH 7.4, 6.5, or 5.0 at 37 °C) (Fig. S3 ESI†). The plots for the Korsmeyer–Peppas equation of the CG–PEG–FA/DOX resulted in linearity with R2 values of 0.9705, 0.9828 and 0.9840, and n values of 0.2644, 0.1605 and 0.1674, respectively. The Korsmeyer–Peppas equation yielded comparatively good linearity (R2 = 0.9705, 0.9828 and 0.9840) and perfect release exponent (n = 0.2644, 0.1605 and 0.1674). The results revealed that the release mechanism was diffusion-controlled drug release.33
To evaluate the folic acid (FA) group-mediated targeting function of the CG–PEG–FA, CG–PEG–FA/DOX and CG–PEG/DOX (with a DOX-loading capacity of 0.4015 ± 0.0114 mg mg−1 at pH 7.4) were used for in vitro study with HepG2 cells and LSECs. The two graphene-based nanocarriers had a similar structure except that the PEG brushes on the CG–PEG–FA were terminated with FA moieties. The DOX-loading capacity of the CG–PEG–FA was 99.5% of the CG–PEG under the same drug-loading conditions, and the marginal difference between the two nanocarriers was caused by the FA moieties in the CG–PEG–FA. So it could be concluded that the two nanocarriers must have the same drug-loading mechanism, as well as the same releasing mechanism.
As shown in Fig. 8 (B), the anticancer activity tests showed that the free DOX and the CG–PEG–FA/DOX displayed similar anticancer activity toward HepG2 cell lines. In the absence of folic acid, the anticancer activity of the CG–PEG/DOX was dramatically decreased. The order of efficacy as a killing agent is the free DOX, then the target-specific CG–PEG–FA/DOX, and finally the non-target-specific CG–PEG/DOX. This revealed that the target-specific CG–PEG–FA/DOX showed more obvious cell inhibition than the non-target-specific CG–PEG/DOX, which is mainly attributed to the plentiful FA receptors (FRs) on the surface of the tumor cells. Therefore, we can deduce that the DOX-loaded CG–PEG–FA biocompatible and target-specific nanocarriers show no obvious difference in therapeutic effects against cancer cells as compared with free DOX, but show much lower cytotoxicity.
As for the LSEC lines, a type of human liver sinusoidal endothelial cells, the target-specific CG–PEG–FA/DOX and the non-target-specific CG–PEG/DOX displayed similar anticancer activity, much lower than that of the free DOX with increasing DOX dosage (Fig. 8(C)). With the same DOX dosage of 12 μg mL−1, the relative cellular viability in the CG–PEG–FA/DOX and CG–PEG/DOX experiments reached about 74% within 48 h, which was much higher than that of the free DOX (37.65%). This meant that the cell toxicity of DOX was decreased significantly by the two drug carriers. Furthermore, the similar anti-tumor efficacy of the two nanocarriers indicated that the folic acid groups had no target-specificity to the normal cells. These results demonstrated that folic acid segments played an important role in generating receptor-mediated specificity for the selective killing of cancer cells.35
Footnote |
† Electronic supplementary information (ESI) available: FTIR spectra of the NH2–PEG–NH2 and NH2–PEG–FA, the DOX-loading capacity of the CG–PEG–FA nanocarrier at different pH values, and kinetics models of the drug release performance. See DOI: 10.1039/c4ra02466d |
This journal is © The Royal Society of Chemistry 2014 |