2014-Hyaluronic - Acid - Conjugated - Graphene - Oxide - For Targeted - Drug - Delivery

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

Available at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/carbon

Hyaluronic acid conjugated graphene oxide


for targeted drug delivery

Huixia Wu a,*, Haili Shi a, Yapei Wang a, Xiaoqing Jia a, Caizhi Tang a, Jiamin Zhang b,
Shiping Yang a,*
a
The Key Laboratory of Resource Chemistry of Ministry of Education and the Shanghai Key Laboratory of the Rare Earth Functional Materials,
Department of Chemistry, College of Life and Environmental Science, Shanghai Normal University, Shanghai 200234, People’s Republic of
China
b
Shanghai (Red Cross) Blood Center, Shanghai Institute of Blood Transfusion, Shanghai 200051, People’s Republic of China

A R T I C L E I N F O A B S T R A C T

Article history: Nano-sized graphene oxide (GO) is functionalized with adipic acid dihydrazide to introduce
Received 3 July 2013 amine groups, and then hyaluronic acid (HA) is covalently conjugated to GO by the forma-
Accepted 11 December 2013 tion of amide bonds. The resulting HA-grafted GO (GO–HA) has negligible hemolytic activity
Available online 18 December 2013 and very low cytotoxicity towards HeLa and L929 cells, and it can be effectively taken up by
cancer cells through receptor-mediated endocytosis. The histological, hematological and
biochemical analysis results suggest no perceptible toxicity of GO–HA in mice at a high
exposure level of 10 mg kg1 and at an exposure time of up to 10 days. Doxorubicin
(DOX) can be efficiently loaded on the GO–HA, and the resulting GO–HA/DOX exhibits nota-
ble cytotoxicity to HeLa cells. The in vivo drug delivery capability of GO–HA is demonstrated
by following the tumor growth in mice after intravenous administration of GO–HA/DOX.
The results indicate that GO–HA can efficaciously deliver DOX to the tumors and suppress
tumor growth.
 2013 Elsevier Ltd. All rights reserved.

1. Introduction to biological applications [16]. Importantly, both sides of a GO


sheet can be available for drug loading, which contributes to
As a two-dimensional sp2 carbon networking material iso- the high drug-loading amount. Moreover, the dynamic bond-
lated for the first time in 2004, graphene has attracted great ing interactions (for example, p–p stacking, hydrophobic,
attention because of its extraordinary chemical, optical, elec- hydrogen bonding and electrostatic interactions) between
trical and mechanical properties and potential applications GO and the drugs show controlled response to external stim-
for nanoelectronic devices, transparent conductors and com- uli (such as pH, temperature, chemical substances and elec-
posite materials [1–3]. Very recently, graphene has also been tric fields) [17,18]. Therefore, the controlled drug release
used in biomedical fields, such as drug/gene delivery, biolog- from GO can be achieved by various routes. All these positive
ical sensing, molecular imaging and cancer therapy [4–11]. attributes make GO much more efficacious than other carbo-
In particular, functionalized graphene oxide (GO) and its naceous nanomaterials as drug carriers for in vitro and in vivo
derivatives have emerged as promising materials for drug biological applications.
delivery [4,6,10–15]. GO possesses unique features, such as As a naturally occurring polysaccharide composed of alter-
easy synthesis, high dispersibility in water as well as in phys- nating units of N-acetyl-D-glucosamine and D-glucuronic acid,
iological environments, excellent biocompatibility and easily hyaluronic acid (HA) has unique and excellent physicochem-
tunable surface functionalization, which are highly propitious ical properties, such as biodegradability, biocompatibility and

* Corresponding authors.
E-mail addresses: wuhuixia@shnu.edu.cn (H. Wu), shipingy@shnu.edu.cn (S. Yang).
0008-6223/$ - see front matter  2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carbon.2013.12.039
380 CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

O H
NH2 N
N NH2
H O
OH OH
O O

HOOC (a) O
NH
HN
H
N
O NH2
O

O OH
OH
OH
O
O OH O
O O O
NH HO
NH n
OH
NH O
CH3
O
O
NH (b)
OH OH NH OH
O O
O O OH O O OH O
[ HO O O O ]n
NH HO NH
OH OH
O O
CH3 CH3

O OH O
OH

(c) OH

O O OH O OH
NH2
O

O
O OH
OH
O OH
OH

O HO O
O OH

O NH2
O
NH

NH
O
O (d)
NH
OH OH NH OH
O O
O O OH O O OH O
[ HO O O O ]n
NH HO NH
OH OH
O O
CH3 CH3

Fig. 1 – Schematic diagram of the synthesis of GO–HA and succedent DOX loading: (a) Functionalizing GO with ADH to
produce GO–ADH; (b) conjugation of HA with GO–ADH by the formation of amide bonds to produce GO–HA; (c) loading DOX
onto GO–HA and (d) intravenous administration of GO–HA/DOX by tail vein. (A colour version of this figure can be viewed
online.)

nonimmunogenicity [19,20]. HA is widely present in the extra- expressing cancer cells has shown great application potential
cellular matrix, connective tissues and bodily fluids, and it in biomedical fields such as cancer diagnosis and therapy [28–
plays important roles in many pathophysiological processes 30]. In view of the above-mentioned targeting characteristics
[21,22]. It is also known that HA has a strong affinity with of HA, the conjugation of HA to GO is a promising method
cell-specific surface markers such as cluster determinant 44 for targeting tumor cells.
(CD44) and receptor for hyaluronate-mediated motility In this contribution, to integrate the advantages of GO and
[23–25]. These HA receptors, especially the CD44, show HA, the nano-sized GO was functionalized with adipic acid
dramatically higher levels on the surface of various tumor dihydrazide (ADH), and then the introduced amino groups
cells [25]. CD44 overexpression is also associated with cancer- were used to conjugate HA. The in vitro and in vivo toxicity
ous angiogenesis and other types of tumor progression studies of the resulting HA-grafted GO (GO–HA) were carried
[26,27]. Such an interesting selectivity of HA to CD44 over- out by conducting 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-
CARBON 6 9 (2 0 1 4) 3 7 9–38 9 381

tetrazolium bromide (MTT) assays, hematological and serum cultured in regular growth medium consisting of Roswell Park
biochemistry measurements, and histological assessments. Memorial Institute medium (RPMI) 1640 supplemented with
Doxorubicin hydrochloride (DOXÆHCl), a hydrophilic drug, 10% fetal bovine serum at 37 C under a 5% CO2 atmosphere.
can be loaded onto the functionalized GO in a controlled The cells were routinely harvested by treating with a trypsin–
way (Fig. 1). The in vitro selective targeting and cytotoxic effect ethylene diamine tetraacetic acid (EDTA) solution (0.25%).
of the DOX-loaded GO–HA (GO–HA/DOX) to CD44 over-
expressing HeLa cells were examined. Moreover, the chemo- 2.5. In vitro cytotoxicity and hemolysis assay
therapy effects of HeLa tumor-bearing mice intravenously in-
jected with GO–HA/DOX were studied. Cytotoxicity: HeLa and L929 cell lines were used to evaluate the
in vitro cytotoxicity of GO and GO–HA. The cells, which were
2. Experimental seeded in a 96-well cell culture plate at a density of 1 · 104
cells per well, were cultured under 5% CO2 at 37 C for 24 h.
2.1. Chemicals and materials Then the cells were exposed to serial concentrations of the
materials, and further incubated for 12 or 24 h. Finally, the cell
ADH (P98%) was provided by TCI. Sodium hyaluronate (ex- viability was assessed using the MTT assays. The cytotoxicity
tracted from cockscomb) was obtained from Shanghai Haoran was expressed as the percentage of cell viability compared to
Biological Technology Co., Ltd. DOXÆHCl (>99%) was acquired that of untreated control cells.
from Beijing Hua Feng United Technology Co., Ltd. Other chem- Hemolysis assay: Human red blood cells (HRBCs) stabilized
icals and reagents were purchased from Sinopharm Chemical with EDTA were kindly provided by Shanghai Blood Center.
Reagent Co., Ltd. Deionized water with resistivity higher than The in vitro hemolysis assays were carried out following the
18 MX cm was used in all experiments. GO was prepared by a literature method [32]. The concentration of the test material
modified Hummers’ method [31] using natural graphite pow- is 50, 100, 200, 300 and 400 lg mL1.
der as a starting material (Supporting information).
2.6. In vivo toxicity studies
2.2. Synthesis of GO–HA
The healthy male Kunming (KM) mice (body mass 20 g) were
GO (50 mg) was dispersed in distilled water (50 mL), and then obtained from Shanghai Laboratory Animal Center, Chinese
N-(3-dimethylaminopropyl)-N 0 -ethylcarbodiimide hydrochlo- Academy of Science, Shanghai, China (Warrant No. SCXK
ride (EDC, 250 mg) and N-hydroxysuccinimide (NHS, 750 mg) [Shanghai] 2012-0002). All animal operations were conducted
were introduced to activate the carboxylic acid groups of in accordance with the guidelines of the Institutional Animal
GO. ADH (520 mg) was added to the activated GO solution, fol- Care and Use Committee. GO–HA dispersed in physiological
lowed by another 24 h of stirring. The produced GO–ADH was saline (0.2 mL) at a total dose of 10 mg kg1 was injected into
separated by centrifugation and rinsed repeatedly with water. KM mice (n = 3 per group) by the tail vein. KM mice (n = 3) with
Sodium hyaluronate (15 mg), EDC (38 mg) and NHS (113 mg) injection of only physiological saline were chosen as the con-
were dissolved in phosphate buffered saline (PBS, pH 5.6, trol group.
15 mL), then the above-mentioned GO–ADH was introduced, Histology and hematology studies: Blood samples and tissues
and the mixture was stirred at room temperature for 24 h. were harvested from mice 4 h and 10 days post-injection.
The resulting GO–HA was collected by centrifugation, washed Blood was collected from the orbital sinus by quickly remov-
repeatedly with water, and then dispersed in PBS (pH 7.4). ing the eyeball from the socket with a pair of tissue forceps.
Two indicators for kidney functions (creatinine and blood
2.3. Characterization urea nitrogen (BUN)) and three important hepatic indicators
(alanine aminotransferase (ALT), aspartate aminotransferase
Transmission electron microscopy (TEM) images were ob- (AST) and total bilirubin (TBIL)) were determined. For the
tained on a JEOL JEM-2100 high-resolution transmission elec- preparation of blood smear, a drop of blood was placed on
tron microscope. Morphological analyses were performed one end of a slide and dispersed along the length of the slide
with a Veeco Multimode IIIa atomic force microscope (AFM). by another slide. The slide was left to air dry, after which the
Fourier transform infrared (FT-IR) spectra were recorded on blood was stained with hematoxylin–eosin (HE) to distinguish
a Nicolet Avatar 370 FT-IR spectrophotometer using KBr pel- the cells from each other. Upon completion of the blood col-
lets. X-ray diffraction (XRD) patterns were obtained on a Rig- lection, the mice were sacrificed. The heart, liver, spleen, lung
aku DMAX 2000 diffractometer using Cu-Ka radiation and kidneys were removed and fixed in 4% formaldehyde,
(k = 0.15405 nm) (40 kV, 40 mA). Raman spectra were collected then dehydrated and embedded in paraffin. 5 mm transverse
with a super LabRam II confocal microscopic Raman spec- sections were cut (Leica RM2235 microtome) and stained with
trometer instrument (Dilor, France). Ultraviolet–visible HE for histological analysis.
(UV–vis) absorption spectra were obtained with a BeckMan
coulter DU 730 spectrophotometer. 2.7. Luminescence imaging

2.4. Cell culture Confocal luminescence imaging was performed with a Leica
TCS SP5 II laser scanning confocal microscope and a 63· oil-
HeLa and L929 cell lines, which were provided by the Institute immersion objective lens. The cells were seeded on coverslips
of Biochemistry and Cell Biology, SIBS, CAS (China), were and incubated with 30 lg mL1 of GO–HA/DOX and GO/DOX
382 CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

for 3 h. The cells were washed 3 times with PBS and then nanosheets have a thickness of 0.8–1.3 nm (Fig. 2c), suggest-
stained with 4,6-diamino-2-phenyl indole (DAPI) solution be- ing the complete exfoliation of GO sheets down to individual
fore measurements. ones. The determined thickness of GO is larger than the the-
oretical value of a single layer graphene because of the pres-
2.8. Drug loading, release and cytotoxicity assay ence of covalently bound oxo-groups [33]. The GO sheets
show good separation, and their lateral width changes from
Loading DOX onto GO–HA was performed by mixing DOXÆHCl 60 to 500 nm and more than 80% of the GO sheets are smaller
with the material (5 mg) in PBS (pH = 7.4, 10 mL) for 20 h un- than 300 nm (Fig. 2d).
der dark conditions at a certain drug concentration. The The FTIR spectrum (Fig. 2e) of the GO reveals the presence
resulting GO–HA/DOX was collected by centrifugation, and of –OH (3380 cm1), C@O (1725 cm1), C@C (1622 cm1) and C–
washed with PBS repeatedly to remove the excess drug. The O (1060 cm1) [14,34]. The as-synthesized GO was further
loading ratios of DOX were estimated by UV–vis spectroscopy investigated by XRD analysis. As shown in Fig. 2f, a sharp
at 490 nm. peak is observed at 11.3 (d = 0.78 nm), which usually appears
To study the release behavior of DOX from GO–HA, GO–HA/ in the XRD patterns of GO samples [14,35]. This peak is due to
DOX (5 mg) was suspended in a PBS buffer (2.0 mL, pH 7.4, 6.3 the peak position of stacked GO analogs formed during the
and 5.2) and sealed in dialysis membranes (Mw cutoff = 3500). process of preparation of XRD samples. The characteristic dif-
The dialysis bags were incubated in PBS buffer (8.0 mL) of the fraction peak of graphite at 26.5 disappears due to the com-
same pH value at 37 C under shaking (200 rpm min1). The plete oxidation of graphite [14]. The Raman spectrum of GO
amount of the released DOX was measured at different time (Fig. 2g) shows two strong peaks at 1333 and 1592 cm1 that
points by a UV–vis spectrophotometer (490 nm). Each experi- are assigned to the D and G bands, respectively. The D/G ratio
ment was repeated three times. of GO is about 1.2, which is higher than that of as-prepared
The cytotoxicity of free DOX and GO–HA/DOX against HeLa graphite oxide (0.79) (Fig. S1). The significant increase of the
cells was determined using MTT assays. The determination D/G ratio for GO compared to that of graphite and graphite
procedure was similar to the cytotoxicity assay of GO–HA. oxide suggests the introduction of a higher level of disorder
onto the graphene layer and the decrease in the size of the
2.9. In vivo tumor therapy in-plane sp2 domains after oxidation and succedent exfolia-
tion [14,35]. Compared with that of graphite oxide, the 2D
The HeLa tumor models were generated by subcutaneous band of GO is shifted from around 2700 to 2650 cm1. The
injection of 5 · 106 cells in RPMI 1640 (0.2 mL) into the left lat- 2D band is the second order band of zone-boundary phonons.
eral abdominal wall of the nude mice. Treatment was initi- The shift of the 2D peak depends on the number of graphene
ated when the tumors reached an approximate size of layers. Single-layer graphene shows a single sharp 2D peak
100 mm3. The mice were divided into three groups, and each below 2700 cm1, while bilayer or more sheets have a broader
experimental group included 3 mice. The mice were adminis- and upshifted 2D peak located at about or more than
tered with (a) physiological saline (200 lL), (b) GO–HA (200 lL, 2700 cm1. The Raman spectrum of GO exhibits a 2D peak
10 mg kg1), (c) GO–HA/DOX (200 lL, 6 mg kg1 in terms of which is lower than 2700 cm1, suggesting the single layer
DOX, with the same amount of GO–HA as that of GO–HA crystal structure of the GO sheets [14].
group) intravenously through tail veins every three days. In order to introduce amine moieties for further graft of
The tumor volume and body weight were monitored every HA, the as-prepared GO was first conjugated with ADH
other day. The tumor volume was calculated using the follow- through carbodiimide coupling. Fig. 3A(a) shows the FT-IR
ing formula: tumor volume = (tumor length) · (tumor width)2/ spectrum of the obtained GO–ADH, which reveals the disap-
2. Relative tumor volumes were calculated as V/V0 (V0 is the pearance of the C@O stretching of –COOH (1725 cm1) and
tumor volume when the treatment was initiated). the appearance of a strong band at 1630 cm1 associated with
the amide bonds [36]. Meanwhile, the amine content present
2.10. Statistical analysis in GO–ADH was determined to be about 5.9 · 105 mol g1
using the standard FMOC (9-fluorenylmethyl chloroformate)
Statistical analysis was performed using the unpaired two- quantification protocol [37], so the weight percent of conju-
tail Student’s t-test. The data are presented as means ± stan- gated ADH is about 1.0%. Afterwards, the primary amine
dard deviation. A P value of less than 0.05 was considered sta- groups in GO–ADH were allowed to react with HA through
tistically significant. an amide linkage. The FT-IR spectrum of the resulting GO–
HA is present in Fig. 3A(b). Several new absorbance character-
3. Results and discussion istics of HA appear at 1627 cm1 (amide I band), 1405 cm1
(carboxylate symmetric stretching), 1155 cm1 (skeletal acetal
3.1. Synthesis and characterization of GO–HA valence band) and 1045 cm1 (C–O stretching), confirming the
successful conjugation of HA on GO [38]. The weight content
The water-soluble GO was prepared by the oxidation of graph- of HA in GO–HA has been estimated by TGA method (Fig. S2).
ite according to a modified Hummers’ method [31] and the The thermogram of pristine GO shows a little mass loss
following sonication treatment. The obtained GO was stable (10%) below 120 C, which is associated with the adsorbed/
in pure water for over a month and no agglomeration trapped water molecules. A significant mass loss (31%) in
occurred. TEM (Fig. 2a) and AFM (Fig. 2b) images provide the the range of 120–260 C is attributed to the decomposition of
morphological information of the as-prepared GO. The oxygen-containing groups [39]. From the TGA curve of GO–
CARBON 6 9 (2 0 1 4) 3 7 9–38 9 383

(a) (b) 1.077 nm (c)

40

30
(d)

Percentage %
20

10

0
0 100 200 300 400 500
Size (nm)

(e) (f) 26.5


o (g)
D

G
Intensity (a.u.)
Transmittance

Intensity (a.u.)
1725 Graphite powder
1060
1622
o
11.3

2D
3380 GO

4000 3500 3000 2500 2000 1500 1000 10 20 30 40 50 60 70 80 1000 1500 2000 2500 3000
-1
Wavenumber (cm )
-1 2θ (degree) Raman shift (cm )

Fig. 2 – TEM image (a), AFM image (b) and height histogram (c), size distribution obtained from AFM images (d), FT-IR
spectrum (e), XRD pattern (f), and Raman spectrum (g) of GO. The XRD pattern of original graphite powder was also shown in
(f) for comparison. (A colour version of this figure can be viewed online.)

(A) a
(B)

1630
b
1155
Transmittance

c 1405
1627 1045

4.925 nm

4000 3500 3000 2500 2000 1500 1000 500


-1
Wavenumber (cm )

Fig. 3 – (A) FT-IR spectra of GO–ADH (a), GO–HA (b) and HA (c); (B) AFM image and height histogram of GO–HA. (A colour version
of this figure can be viewed online.)

HA, it is observed that about 14% weight loss occurs between GO due to the sonication treatment during the synthesis.
120 and 220 C, which is caused by the removal of remaining Meanwhile, the measured thickness of GO-HA increases to
oxygen-containing groups on GO. The notable weight loss of 2.0–5.0 nm, mainly due to the attachment of HA to the GO
49% in the range of 230–680 C mainly corresponds to the sheet [40]. The initial GO has a zeta potential of 30.0 mV,
complete degradation of grafted HA segments [30,39]. which changes to 11.4 mV when modified with ADH. After
According to the TGA data of GO and GO–HA, the weight HA grafting, the zeta potential of the resulting GO–HA is
percentage of grafted HA chains is 49% by ignoring the significantly reduced to 34.2 mV (Fig. S3). The zeta-poten-
conjugated amount of ADH. Fig. 3B shows the AFM image tial measurement results further indicate the successful
of the resulting GO–HA. The GO-HA remains well dispersed, grafting of HA on GO. The as-prepared GO–HA can be
which is critical for biological applications. The GO-HA colloidally stable in water and physiological saline at room
sheets have sizes (40–350 nm) lower than that of the original temperature (Fig. S4).
384 CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

3.2. In vitro cytotoxicity of GO–HA 2.0 2.0

1.5 GO
Water
The cytotoxicity effects of GO-HA against the HeLa and L929 GO-HA

Hemolysis %
1.0
cell lines were investigated by conducting MTT assays. As 1.5
0.5
shown in Fig. 4a and b, these two cell lines give similar results

Absorbance
for GO-HA, and the difference in cell viabilities after 12 and 0.0
1.0 0 100 200 300 400
24 h of incubation is negligibly small. No obvious cytotoxicity -1
Concentration (μg mL )
against HeLa and L929 cells was found after 24 h of treatment
at a concentration as high as 200 lg mL1. These results indi-
0.5
cate that GO–HA has low cytotoxicity, which is an important
PBS, GO-HA
character for drug carriers. GO-HA
H2O 400 300 200 100 PBS
0.0
3.3. In vitro blood compatibility 500 550 600 650 700
Wavelength (nm)
For successful intravenous administration and transport of
anticancer drug, the carriers must have excellent blood com- Fig. 5 – UV–vis absorption spectra to detect the presence of
patibility, such as a very low hemolytic effect. Very recently, hemoglobin in the supernatant of GO–HA suspensions in
several groups have researched the hemolytic effect of GO, PBS at concentrations of 50, 100, 200, 300 and 400 lg mL1
but they get the opposite results [41–44]. Sasidharan et al. using water and PBS as the positive and negative controls,
found that HRBCs treated with carboxyl-functionalized respectively. The insets: photographic image for direct
graphene did not show any hemolysis up to 75 lg mL1 [41], observation of hemolysis by GO–HA (lower right) and
while Singh et al. reported that GO exhibited significant hemolysis percentages of HRBCs by GO and GO–HA at
hemolytic activity [42]. Herein, hemolysis assays were utilized different concentrations (upper right). (A colour version of
to evaluate the blood compatibility of GO and GO–HA at differ- this figure can be viewed online.)
ent concentrations, as presented in Fig. 5. No hemolytic effect
of GO and GO–HA can be observed visually in the experiment
over a broad concentration range of 0–400 lg mL1. The GO–HA. At the time point of 4 h and 10 days after injection,
hemolytic effect of the samples was further quantitatively blood was drawn for blood smear analysis and biochemistry
determined by measuring the absorbance of the supernatant assays. Upon completion of blood collection, the mice were
at 541 nm (hemoglobin) using UV–vis spectroscopy. Even at a sacrificed, and the main organs were harvested for histologi-
high concentration of 400 lg mL1, the mean hemolytic cal study.
activity of GO and GO–HA has been detected to be as low as Nanoscale materials, whose sizes are within the size range
1.5% and 0.7%, respectively. Consequently, both GO and GO– of viruses and large proteins, may induce inflammatory re-
HA have negligible hemolytic activity. sponses and alter the activity of the immune system and
the hematological factors [45,46]. Therefore, hematological
3.4. In vivo toxicity studies and biochemical analyses were carried out to quantify the po-
tential toxicity of GO–HA after administration. Fig. 6a and b
The GO–HA was intravenously injected into KM mice through show the photomicrographs of the blood smears from blood
tail vein with a dose of 10 mg kg1. Over the entire study per- in vivo treated with physiological saline (control) and with
iod, the observed behaviors of all the treated mice were nor- GO–HA, respectively. The blood smears reveal that the GO–
mal, and no infection, impaired mobility, or reduced food HA treatment does not alter the number and shape of red
taking was observed, indicating a negligible acute toxicity of blood cells, platelet and white blood cells in the test mice as

(a) 120
(b) 120
12 h 24 h 12 h 24 h
100 100
Cell viability %

80 80
Cell viability %

60 60

40 40

20 20

0 0
0 5 25 50 100 200 0 5 25 50 100 200
-1
-1
Concentration (μg mL ) Concentration (μg mL )

Fig. 4 – Viability of HeLa (a) and L929 (b) cells incubated with different concentrations of GO–HA for 12 and 24 h. (A colour
version of this figure can be viewed online.)
CARBON 6 9 (2 0 1 4) 3 7 9–38 9 385

compared to the control mice. Meanwhile, more quantitative widely-used chemotherapeutic anticancer drugs, but its
evaluation about the influence of GO–HA on the exposed mice administered dosage is strongly limited by the severe side ef-
was studied by using the established serum biochemistry as- fects. In view of the unique structure and properties of GO,
says. Three important hepatic indicators, viz. ALT, AST and the disadvantages of DOX can be efficaciously resolved by
TBIL, are at similar levels for the mice exposed to GO–HA using functionalized GO as drug carriers. Therefore, herein
and for the control mice (Fig. 6c). The test groups (4 h and DOX was used as a model drug to examine the drug loading
10 days post-injection) also show no significant difference in and release properties of GO–HA.
the two indicators for kidney functions (creatinine and BUN) The drug loading of GO–HA was started by mixing GO–HA
as compared with the control group (Fig. 6c, P > 0.05). These with a PBS solution of DOXÆHCl. The UV–vis absorbance at
results suggest no toxicity of GO–HA in mice at a high expo- 490 nm (characteristic of DOX) was measured to estimate
sure level of 10 mg kg1 and at an exposure time of up to the amount of the loaded drug. As indicated in Fig. 7a, the
10 days. drug loading ratio (the weight ratio of the loaded drug to
To determine whether or not the GO–HA cause tissue dam- GO–HA) for DOX increases with the increase in initial drug
age, inflammation or lesions, the mouse organs including concentration and finally reaches its saturation value when
heart, lung, liver, spleen and kidneys were harvested, sec- the drug concentration is higher than 1000 lg mL1, suggest-
tioned into thin slices, and H&E stained for histological ing that the loading of DOX onto GO–HA can be well con-
assessment. In comparison with an untreated control group, trolled. The maximum loading ratio of DOX is up to 81.5%
no obvious abnormality can be found in various organs from (815 mg DOX per gram of GO–HA).
the GO–HA-exposed mice (Fig. S5). Although further long- The release profile of DOX from the GO–HA/DOX (60.0%
term studies are needed to fully illustrate the toxicology pro- loading ratio of DOX) in PBS buffer (pH 7.4, 6.3 and 5.2) at
files of GO-HA, our preliminary results show that GO-HA has 37 C is presented in Fig. 7b. Only a little of loaded drug
good biocompatibility. (6.8% for pH 7.4 and 10.9% for pH 6.3) has been released after
65 h of immersion in PBS at pH 7.4 and 6.3. However, when
3.5. In vitro loading and release of DOX the pH of the solution is decreased to 5.2, much more DOX
is released, and the amount of released DOX within 65 h
Cancer chemotherapy agents with extended p-structures lar- markedly increases to 26%. This is attributed to the increased
ger than one aromatic ring can be efficiently loaded on GO by hydrophilicity and solubility of DOX at this pH [6]. The pH-
strong p–p stacking and hydrophobic interaction [4,6]. Com- dependent drug release from the GO–HA is important in prac-
pared with their CNT analogs, graphene derivatives are more tical bio-applications, since the microenvironments in can-
attractive as drug carriers because both sides of a single sheet cerous tissue and intracellular lysosomes and endosomes
can be accessible for drug binding. DOX is one of the most are acidic, which potentially facilitates active drug release
from the carriers [47,48]. In addition, it should be noted that
the minimal drug leakage from the GO–HA under normal
physiological conditions will be of great significance because
(a) (b) of the minimized side-effects to normal organs.

3.6. Cellular uptake and cytotoxicity of drug-loaded GO–


HA

HeLa cell line shows very high levels of CD44 protein [30],
while L929 cells are low-expressing for CD44, as indicated in
Fig. S6. Moreover, CD44 on normal cells is usually in a quies-
(c) Control
150 cent state and does not have the binding activity to HA, while
GO-HA 4 h
GO-HA 10 days the tumor cells overexpressing CD44 are in a highly activated
120
state capable of binding HA [49]. Herein, these two cell lines
Concentration

90 were selected to evaluate the targeting effect of GO–HA to tu-


mor cells. The GO–HA/DOX (60.0% loading ratio of DOX) was
60 incubated with HeLa and L929 cells at 37 C for 3 h, and the
30
cell uptake efficiency was analyzed by confocal fluorescence
microscopy. As shown in Fig. 8a, strong red fluorescence can
0.3
be seen in HeLa cells after incubation with GO–HA/DOX. In
0.0 contrast to HeLa cells, very weak fluorescence signal is de-
Creatinine BUN ALT AST TBIL
(μM) (mM) -1
(IU L )
-1
(IU L ) (μM) tected in L929 cells (Fig. 8b), indicating that GO–HA/DOX is
up-taken by HeLa cells more efficiently than by L929 cells.
Fig. 6 – Blood smear images of the mice at 10 days post- To further confirm the dominant receptor-mediated endocy-
injection with physiological saline (a) and with GO–HA (b); (c) tosis mechanism of the GO–HA/DOX, HeLa cells were pre-
serum biochemistry results obtained from mice injected treated with excess free HA and then incubated with
with physiological saline (10 days post-injection) and with GO–HA/DOX under the same conditions. As a result, HA-
GO–HA (4 h and 10 days post-injection). (A colour version of pre-incubated HeLa cells no longer show appreciable fluores-
this figure can be viewed online.) cence (Fig. 8c), demonstrating the specific uptake of GO–HA/
386 CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

(a) 100 (b) 40


5.2
80 6.3
30
7.4

DOX loading ratio (%)

Drug-released (%)
2.0

60
1.5
20

Absorbance
Incubated
Original solution with GO-HA
40 1.0

0.5 10
20
0.0
300 400 500 600
Wavelength (nm) 0
0
0 500 1000 1500 2000 0 10 20 30 40 50 60 70
-1 Time (h)
Concentration of DOX (μg mL )

Fig. 7 – (a) Plot of loading ratio of DOX to GO–HA versus the drug concentration; (b) release profiles of DOX from GO–HA at pH
7.4, 6.3 and 5.2. The inset in (a): Typical UV–vis spectra of DOX solutions before and after interaction with GO–HA, the
solutions were diluted with PBS prior to the measurements. (A colour version of this figure can be viewed online.)

(a) (b) (c) (d) (e)


a1 b1 c1 d1 e1

a2 b2 c2 d2 e2

a3 b3 c3 d3 e3

Fig. 8 – Confocal fluorescence images of GO–HA/DOX (30 lg mL1) in HeLa (a), L929 (b), and free HA pre-treated HeLa (c) cells
incubated at 37 C for 3 h; (d) confocal fluorescence images of GO–HA/DOX in HeLa cells incubated at 4 C for 3 h and (e)
confocal fluorescence images of GO/DOX (30 lg mL1, with the same concentration of DOX) in HeLa cells incubated at 37 C for
3 h. Scale bar (in c3) corresponds to 20 lm for all images. Red fluorescence (a1–e1): DOX molecules; blue fluorescence (a2–e2):
nuclei stained with DAPI; (a3–e3): merged images of blue and red fluorescence with the bright-field image. (A colour version
of this figure can be viewed online.)

DOX could be blocked by the excess free HA. The above con- GO/DOX show a very weak fluorescence. Therefore, the up-
focal fluorescence imaging results are consistent with the take of GO–HA/DOX to the cancer cells overexpressing HA
data obtained by flow cytometric determination (Fig. S7). Fur- receptors is associated with the HA-receptor-mediated endo-
thermore, when HeLa cells were treated with GO–HA/DOX at cytosis [50]. The grafted HA segments on GO sheets controls
4 C to block the endocytosis process, negligible red the targeted drug delivery to HeLa cells by selective binding
fluorescence can be observed (Fig. 8d). Confocal fluorescence to the HA receptors. The significant selectivity of drug-loaded
images of nontargeted GO/DOX in HeLa cells incubated at GO–HA towards cancer cells with high-expressing HA recep-
37 C for 3 h are also shown in Fig. 8 for comparison tors is vital for chemotherapeutic drug delivery. The contact
(Fig. 8e). Compared to GO–HA/DOX, HeLa cells treated with of CD44 receptor with HA is dominated by hydrogen bonds
CARBON 6 9 (2 0 1 4) 3 7 9–38 9 387

DOX HA receptor-mediated endocytosis has been blocked by free


120
GO-HA/DOX HA. In our case, a similar cytotoxicity to HeLa cells is observed
GO-HA/DOX + free HA
100 for GO–HA/DOX and free DOX in the concentration range of
* * * 0.1–10.0 lg mL1. Several papers have also reported that the
Cell viability %

80 *
DOX-loaded nanocarriers with targeting effect show similar
60 or lower cytotoxicity to the cancer cells in comparison with
the same incubation concentration of free DOX [6,52,53].
40 The similarities in cytotoxicity might be attributed to the dif-
20
ferent uptake mechanisms between free DOX and GO–HA/
DOX. Free DOX molecules directly enter the HeLa cells
0 through diffusion [54], whereas the entry route into the cells
0 0.1 1.0 2.0 5.0 10.0
-1
for GO–HA/DOX is associated with HA-receptor-mediated
Concentration (μg mL )
endocytosis. Although the conjugation of HA facilitates the
Fig. 9 – In vitro cytotoxicity of free DOX and GO–HA/DOX uptake of GO–HA/DOX by HeLa cells, it would take additional
against HeLa cells after 12 h of incubation, together with time for DOX to be released from the surface of GO–HA com-
viabilities of HA-pre-treated HeLa cells incubated with GO– pared to free DOX [52,53]. To support this, the time-dependent
HA/DOX (marked as GO–HA/DOX + free HA) for 12 h. *P < 0.01 intracellular uptake and distribution of DOX for GO–HA/DOX
(GO–HA/DOX group versus GO–HA/DOX + free HA group). (A and free DOX were observed by confocal fluorescence analy-
colour version of this figure can be viewed online.) sis (Figs. S8 and S9). At an incubation time of 0.5 h, HeLa cells
treated with free DOX show red fluorescence, while no red
fluorescence signal is detected for the cells treated with GO–
and van der Waals forces rather than electrostatic or aromatic HA/DOX. When the incubation time is prolonged to 1.5 h,
stacking interactions [51]. The carboxyl groups (hexasaccha- strong red fluorescence due to the released DOX from GO–
ride) and the N-acetyl groups in the HA molecule are related HA is observed mainly in the cytoplasm. The released DOX
with its binding to HA receptors [51]. The confocal microscopy from GO–HA diffuses through the nuclear membrane and
observations also suggest that the binding ability of HA with accumulates in the nuclear, as indicated by the intense red
the receptors on cancer cells has not been inhibited after its fluorescence located in the nucleus after 3 h of incubation.
grafting to GO and the subsequent drug loading. Although the toxicity of GO–HA/DOX is comparable to that
The target-specific delivery of GO–HA/DOX to the cells of free DOX, the good selectivity to the cancer cells high-
with high levels of HA receptors encouraged us to evaluate expressing HA-receptors and minimal drug leakage under
the anticancer efficacy of GO–HA/DOX by determining its normal physiological conditions will be of great significance
in vitro cytotoxicity against HeLa cell line using MTT assays. for practical bio-applications. Targeted drug delivery is be-
Fig. 9 presents the viabilities of HeLa cells against GO–HA/ lieved to improve the therapeutic efficacy while reducing
DOX and against free DOX after incubation for 12 h. The the toxic side effects of the drugs [55].
GO–HA itself has negligible cytotoxic effects on the cancer
cells (Fig. 4), while notable cytotoxicity to the HeLa cells is 3.7. In vivo cancer therapy
observed for GO–HA/DOX. The cytotoxic effects of GO–HA/
DOX increase with the increase of drug concentration. How- To validate the practical chemotherapy application of drug-
ever, when the GO–HA/DOX was incubated with HA-pre-trea- loaded GO–HA in cancer treatments, in vivo anticancer exper-
ted HeLa cells, the cytotoxicity becomes significantly lower at iments were carried out using the HeLa tumor model on nude
DOX concentrations of 1.0–10.0 lg mL1 (P < 0.01), because the mice. The HeLa tumor-bearing mice were divided into 3

(a) 8 (b) 110

7
105
Body weight change (%)
Relative tumor volume (V/V0)

6
100
5

4 95

3
** 90 Control
2 * ** ** GO-HA
* Control 85
GO-HA/DOX
1 GO-HA
GO-HA/DOX
0 80
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
Days Days

Fig. 10 – In vivo antitumor effects of GO–HA/DOX in tumor-bearing nude mice: (a) Tumor growth curves for mice treated with
GO–HA/DOX and GO–HA, in comparison with the blank control. The tumor volumes were normalized to their initial sizes.
*P < 0.05, **P < 0.01 (GO–HA/DOX group versus saline injection group). (b) Body weight (normalized to day 0) curves of different

groups during the experiments. The insets in (a): Representative photos of mice from control group (left) and GO–HA/DOX
group (right) taken at the end of treatment. (A colour version of this figure can be viewed online.)
388 CARBON 6 9 ( 2 0 1 4 ) 3 7 9 –3 8 9

groups. The first group which was given intravenous injection R E F E R E N C E S


of saline was used as the control. The second and third groups
were intravenously injected with GO–HA and GO–HA/DOX,
respectively. The mice were monitored for 16 days for their [1] Geim AK, Novoselov KS. The rise of graphene. Nat Mater
tumor growth pattern. The change of relative tumor volume 2007;6:183–91.
as a function of time is plotted in Fig. 10a, and the typical pho- [2] Geim AK. Graphene: status and prospects. Science
2009;324:1530–4.
tographs of mice from GO–HA/DOX and control groups are
[3] Allen MJ, Tung VC, Kaner RB. Honeycomb carbon: a review of
shown in the insets of Fig. 10a. Mice treated only with the sal- graphene. Chem Rev 2010;110:132–45.
ine experience a rapid growth of tumor volume over the en- [4] Liu Z, Robinson JT, Sun XM, Dai HJ. PEGylated nanographene
tire experiment period. The mice treated with GO–HA oxide for delivery of water-insoluble cancer drugs. J Am
(material control) do not show tumor regression either, and Chem Soc 2008;130:10876–7.
the tumor kept on growing. However, the tumor growth of [5] Lu CH, Yang HH, Zhu CL, Chen X, Chen GN. A graphene
platform for sensing biomolecules. Angew Chem Int Ed
the mice treated with GO–HA/DOX is suppressed considerably
2009;48:4785–7.
with an inhibition of 40% by day 16. Because of the enhanced
[6] Zhang LM, Xia JG, Zhao QH, Liu LW, Zhang ZJ. Functional
permeability and retention effect and the specific binding be- graphene oxide as a nanocarrier for controlled loading and
tween the HA and the CD44 receptor expressed on the surface targeted delivery of mixed anticancer drugs. Small
of tumor cells, the GO–HA/DOX can selectively accumulate in 2010;6:537–44.
the malignant tumor issues and inhibit the tumor growth [7] Peng C, Hu WB, Zhou YT, Fan CH, Huang Q. Intracellular
[56,57]. We also measured the body weight of the mice for imaging with a graphene-based fluorescent probe. Small
2010;6:1686–92.
all groups during the treatments (Fig. 10b). For GO–HA and
[8] Chen BA, Liu M, Zhang LM, Huang J, Yao JL, Zhang ZJ.
GO–HA/DOX treated groups, no noticeable weight loss is ob-
Polyethylenimine-functionalized graphene oxide as an
served, indicating that the toxicity of treatments is not severe. efficient gene delivery vector. J Mater Chem
The results of the above cancer treatments reveal that anti- 2011;21:7736–41.
cancer drug loaded GO–HA has great application potential [9] Yang K, Zhang SA, Zhang GX, Sun XM, Lee ST, Liu Z.
for in vivo cancer chemotherapy. Graphene in mice: ultrahigh in vivo tumor uptake and
efficient photothermal therapy. Nano Lett 2010;10:3318–23.
[10] Yang K, Feng LZ, Shi XZ, Liu Z. Nano-graphene in
4. Conclusions biomedicine: theranostic applications. Chem Soc Rev
2013;42:530–47.
HA has been covalently conjugated to GO using ADH as a linker [11] Feng LY, Wu L, Qu XG. New horizons for diagnostics and
to produce GO–HA, and the application of GO–HA in cancer therapeutic applications of graphene and graphene oxide.
treatment has been studied. The in vitro and in vivo toxicity Adv Mater 2013;25:168–86.
studies show that the resulting GO–HA exhibits very low cyto- [12] Zhang LM, Lu ZX, Zhao QH, Huang J, Shen H, Zhang ZJ.
Enhanced chemotherapy efficacy by sequential delivery of
toxicity, good blood compatibility and no evident toxic effects
siRNA and anticancer drugs using PEI-grafted graphene
in mice. Cellular uptake experiments demonstrate that the oxide. Small 2011;7:460–4.
GO–HA can targetedly deliver the anticancer drugs into the [13] Wang CS, Li JY, Amatore C, Chen Y, Jiang H, Wang XM. Gold
cells by receptor-mediated endocytosis. The GO–HA shows a nanoclusters and graphene nanocomposites for drug
high loading capacity for the anticancer drug DOX, and the delivery and imaging of cancer cells. Angew Chem Int Ed
resulting GO–HA/DOX exhibits high cytotoxicity to HeLa cells. 2011;50:11644–8.
[14] Liu KP, Zhang JJ, Cheng FF, Zheng TT, Wang CM, Zhu JJ. Green
Moreover, the chemotherapeutic effects of tumor-bearing mice
and facile synthesis of highly biocompatible graphene
intravenously injected with GO–HA/DOX have been studied.
nanosheets and its application for cellular imaging and drug
GO–HA/DOX can selectively accumulate in the malignant delivery. J Mater Chem 2011;21:12034–40.
tumor issues and inhibit the tumor growth. All these positive [15] Ma D, Lin JT, Chen YY, Xue W, Zhang LM. In situ gelation and
results suggest that the GO–HA is a promising candidate as sustained release of an antitumor drug by graphene oxide
anticancer drug carriers for cancer treatment. nanosheets. Carbon 2012;50:3001–7.
[16] Li D, Mueller MB, Gilje S, Kaner RB, Wallace GG. Processable
aqueous dispersions of graphene nanosheet. Nat
Acknowledgements Nanotechnol 2008;3:101–5.
[17] Pandey H, Parashar V, Parashar R, Prakash R, Ramteke
This work is supported by the National Natural Science Foun- PW, Pandey AC. Controlled drug release characteristics
dation of China (Grant no. 21271130), the Shanghai Natural and enhanced antibacterial effect of graphene
Science Foundation (13ZR1430300), the Shanghai Municipal nanosheets containing gentamicin sulfate. Nanoscale
2011;3:4104–8.
Education Commission (14ZZ128), and the Opening Project
[18] Zhu YW, Stoller MD, Cai WW, Velamakanni A, Piner RD, Chen
of State Key Laboratory of High Performance Ceramics and
D, et al. Exfoliation of graphite oxide in propylene carbonate
Superfine Microstructure (SKL201212SIC). and thermal reduction of the resulting graphene oxide
platelets. ACS Nano 2010;4:1227–33.
Appendix A. Supplementary data [19] Lapcik Jr L, Lapcik L, De Smedt S, Demeester J, Chabrecek P.
Hyaluronan: preparation, structure, properties, and
applications. Chem Rev 1998;98:2663–84.
Supplementary data associated with this article can be found,
[20] Necas J, Bartosikova L, Brauner P, Kolar J. Hyaluronic acid
in the online version, at http://dx.doi.org/10.1016/j.carbon.201
(hyaluronan): a review. Vet Med-Czech 2008;53:397–411.
3.12.039.
CARBON 6 9 (2 0 1 4) 3 7 9–38 9 389

[21] Knudson CB, Knudson W. Hyaluronan-binding proteins in [41] Sasidharan A, Panchakarla LS, Sadanandan AR, Ashokan A,
development, tissue homeostasis, and disease. FASEB J Chandran P, Girish CM, et al. Hemocompatibility and
1993;7:1233–41. macrophage response of pristine and functionalized
[22] Entwistle J, Hall CL, Turley EA. HA receptors: regulators of graphene. Small 2012;8:1251–63.
signaling to the cytoskeleton. J Cell Biochem 1996;61:569–77. [42] Singh SK, Singh MK, Kulkarni PP, Sonkar VK, Gracio JJA, Dash
[23] Aruffo A, Stamenkovic I, Melnick M, Underhill CB, Seed B. D. Amine-modified graphene: thrombo-protective safer
CD44 is the principal cell surface receptor for hyaluronate. alternative to graphene oxide for biomedical applications.
Cell 1990;61:1303–13. ACS Nano 2012;6:2731–40.
[24] Laurent TC. Biochemistry of hyaluronan. Acta Otolaryngol [43] Yang K, Wan JM, Zhang SA, Zhang YJ, Lee ST, Liu Z. In
Suppl 1987;442:7–24. vivo pharmacokinetics, long-term biodistribution, and
[25] Li F, Bae BC, Na K. Acetylated hyaluronic acid/photosensitizer toxicology of PEGylated graphene in mice. ACS Nano
conjugate for the preparation of nanogels with controllable 2011;5:516–22.
phototoxicity: synthesis, characterization, [44] Liao KH, Lin YS, Macosko CW, Haynes CL. Cytotoxicity of
autophotoquenching properties, and in vitro phototoxicity graphene oxide and graphene in human erythrocytes and
against HeLa cells. Bioconjug Chem 2010;21:1312–20. skin fibroblasts. ACS Appl Mater Interfaces
[26] Mok H, Park JW, Park TG. Antisense oligodeoxynucleotide- 2011;3:2607–15.
conjugated hyaluronic acid/protamine nanocomplexes for [45] Singh P, Prasuhn D, Yeh RM, Destito G, Rae CS, Osborn K,
intracellular gene inhibition. Bioconjug Chem 2007;18:1483–9. et al. Bio-distribution, toxicity and pathology of cowpea
[27] Lee H, Mok H, Lee S, Oh YK, Park TG. Target-specific mosaic virus nanoparticles in vivo. J Control Release
intracellular delivery of siRNA using degradable hyaluronic 2007;120:41–50.
acid nanogels. J Control Release 2007;119:245–52. [46] Hauck TS, Anderson RE, Fischer HC, Newbigging S, Chan
[28] Lee YH, Lee H, Kim YB, Kim JY, Hyeon T, Park H, et al. WCW. In vivo quantum-dot toxicity assessment. Small
Bioinspired surface immobilization of hyaluronic acid on 2010;6:138–44.
monodisperse magnetite nanocrystals for targeted cancer [47] Liu Z, Sun XM, Nakayama-Ratchford N, Dai HJ.
imaging. Adv Mater 2008;20:4154–7. Supramolecular chemistry on water-soluble carbon
[29] Bhang SH, Won N, Lee TJ, Jin H, Nam J, Park J, et al. nanotubes for drug loading and delivery. ACS Nano
Hyaluronic acid quantum dot conjugates for in vivo 2007;1:50–6.
lymphatic vessel imaging. ACS Nano 2009;3:1389–98. [48] Sun XM, Liu Z, Welsher K, Robinson JT, Goodwin A, Zaric S,
[30] Ma M, Chen HR, Chen Y, Zhang K, Wang X, Cui XZ, et al. et al. Nano-graphene oxide for cellular imaging and drug
Hyaluronic acid-conjugated mesoporous silica nanoparticles: delivery. Nano Res 2008;1:203–12.
excellent colloidal dispersity in physiological fluids and [49] Bachar G, Cohen K, Hod R, Feinmesser R, Mizrachi A, Shpitzer
targeting efficacy. J Mater Chem 2012;22:5615–21. T, et al. Hyaluronan-grafted particle clusters loaded with
[31] Hummers WS, Offeman RE. Preparation of graphitic oxide. J Mitomycin C as selective nanovectors for primary head and
Am Chem Soc 1958;80. 1339–1339. neck cancers. Biomaterials 2011;32:4840–8.
[32] Wu HX, Liu G, Zhuang YM, Wu DM, Zhang HQ, Yang H, et al. [50] Jiang G, Park K, Kim J, Kim KS, Hahn SK. Target specific
The behavior after intravenous injection in mice of intracellular delivery of siRNA/PEI-HA complex by receptor
multiwalled carbon nanotube/Fe3O4 hybrid MRI contrast mediated endocytosis. Mol Pharm 2009;6:727–37.
agents. Biomaterials 2011;32:4867–76. [51] Banerji S, Wright AJ, Noble M, Mahoney DJ, Campbell ID, Day
[33] Stankovich S, Dikin DA, Piner RD, Kohlhaas KA, AJ, et al. Structures of the CD44-hyaluronan complex provide
Kleinhammes A, Jia Y, et al. Synthesis of graphene-based insight into a fundamental carbohydrate–protein interaction.
nanosheets via chemical reduction of exfoliated graphite Nat Struct Mol Biol 2007;14:234–9.
oxide. Carbon 2007;45:1558–65. [52] Yang XQ, Grailer JJ, Rowland IJ, Javadi A, Hurley SA, Steeber
[34] Shan CS, Yang HF, Han DX, Zhang QX, Ivaska A, Niu L. Water- DA, et al. Multifunctional SPIO/DOX-loaded wormlike
soluble graphene covalently functionalized by biocompatible polymer vesicles for cancer therapy and MR imaging.
poly-L-lysine. Langmuir 2009;25:12030–3. Biomaterials 2010;31:9065–73.
[35] Krishnamoorthy K, Veerapandian M, Yun K, Kim SJ. The [53] Abdullah-Al-Nahain, Lee JE, In I, Lee H, Lee KD, Jeong JH,
chemical and structural analysis of graphene oxide with et al. Target delivery and cell imaging using hyaluronic acid-
different degrees of oxidation. Carbon 2013;53:38–49. functionalized graphene quantum dots. Mol Pharm
[36] Liu L, Liu DR, Wang M, Du GC, Chen J. Preparation and 2013;10:3736–44.
characterization of sponge-like composites by cross-linking [54] Tang Y, Lei TJ, Manchanda R, Nagesetti A, Fernandez-
hyaluronic acid and carboxymethylcellulose sodium with Fernandez A, Srinivasan S, et al. Simultaneous delivery of
adipic dihydrazide. Eur Polym J 2007;43:2672–81. chemotherapeutic and thermal-optical agents to cancer cells
[37] Yoon TJ, Yu KN, Kim E, Kim JS, Kim BG, Yun SH, et al. Specific by a polymeric (PLGA) nanocarrier: an in vitro study. Pharm
targeting, cell sorting, and bioimaging with smart magnetic Res 2010;27:2242–53.
silica core–shell nanomateriats. Small 2006;2:209–15. [55] Wang ZH, Yu Y, Dai WB, Lu JK, Cui JR, Wu HN, et al. The use
[38] Khachatryan G, Khachatryan K, Stobinski L, Tomasik P, of a tumor metastasis targeting peptide to deliver
Fiedorowicz M, Lin HM. CdS and ZnS quantum dots doxorubicin-containing liposomes to highly metastatic
embedded in hyaluronic acid films. J Alloys Compd cancer. Biomaterials 2012;33:8451–60.
2009;481:402–6. [56] Zhang W, Guo ZY, Huang DQ, Liu ZM, Guo X, Zhong
[39] Li F, Park S, Ling D, Park W, Han JY, Na K, et al. Hyaluronic HQ. Synergistic effect of chemo-photothermal therapy
acid-conjugated graphene oxide/photosensitizer using PEGylated graphene oxide. Biomaterials
nanohybrids for cancer targeted photodynamic therapy. J 2011;32:8555–61.
Mater Chem B 2013;1:1678–86. [57] Yang Y, Zhang YM, Chen Y, Zhao D, Chen JT, Liu Y.
[40] Bao HQ, Pan YZ, Ping Y, Sahoo NG, Wu TF, Li L, et al. Construction of a graphene oxide based noncovalent
Chitosan-functionalized graphene oxide as a nanocarrier for multiple nanosupramolecular assembly as a scaffold for drug
drug and gene delivery. Small 2011;7:1569–78. delivery. Chem Eur J 2012;18:4208–15.

You might also like