You are on page 1of 11

116

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

interact with biological membranes, these carriers are often cytotoxic [29], which constitutes a limiting factor for application of
liposomes in gene delivery. Ahn et al. reported transfection up to
10% when pDNA/PEI polyplexes were delivered into rat bone marrow derived MSCs [30]. A good transfection efciency was achieved
at an N/P ratio of 16, which was comparable with that obtained with
LipofectamineTM . However, at this N/P ratio, polyplexes presented
a high cytotoxicity. In another study, human adipose tissue-derived
MSCs were employed to study the gene transfer properties of
pDNA/PEI polyplexes [31]. Several synthetic biodegradable polymers are being investigated for nucleic acid delivery, such as the
polyesters polylactic acid (PLA) and poly (lactic-co-glycolic acid)
(PLGA) [32,13,33]. Vectors of this kind have the advantage of being
eliminated after pDNA release, in the form of nontoxic degradation
products. Gwak et al. developed PLGA nanospheres as vehicles for
gene delivery to human cord blood-derived MSCs [34]. Uchimura
et al. combined colloidal gold nanoparticles to DNA/Jet-PEITM complexes in an attempt to enhance their uptake by human MSCs,
having in mind their application in cell array-based analyses and in
regenerative medicine [35]. A 2.5-fold increase in gene transfection
was obtained, as compared to the control without gold nanoparticles [36].
Even though much effort has been made in developing nonviral gene vector systems, the traditional nonviral vectors generally
cause destabilization of the cell membrane leading to a pronounced
cytotoxicity in order to achieve effective delivery of DNA. Therefore,
the lack of a widely accepted gene vector coupled with difculty in
transfection of primary cells is a signicant hurdle in the advancement of gene delivery systems. Recent ndings in nanoscience
and nanotechnology have revealed that carbon nanotubes (CNTs)
can be used as a versatile platform for a variety of biomedical
applications, including gene delivery [37]. Owing to unique chemical, physical and biological characteristics such as high surface
area, biocompatibility, low cytotoxicity, and ability to cross the
cell membrane [3840], carbon nanotubes (CNTs) have received
considerable interest in the biomedical applications such as drug
[41,42] and gene delivery [43,44], scaffolds for tissue engineering
[45], biosensing and diagnostics [46]. Many studies have reported
the intracellular transporting of biomolecules by CNTbased carrying materials [4749]. The rst work of utilization of carbon
nanotubes as a novel gene delivery vector system was reported
by Bianco et al. [50]. It is shown that the uptake of pristine CNTs
into mammalian cells is poor, which results in little transport efciency. However, the succeeding works indicated that the transport
efciency can be improved by covalent or non-covalent surface
modications of the CNTs [44,5153]. Skandani et al. [54] simulated the interaction of single-walled CNTs (SWCNTs) with a lipid
bilayer and observed that the lower chirality SWCNTs exhibit signicant adhesion with the membrane. Raffa et al. [55] reported
that the nanotube length can clearly inuence the cellular uptake
while the shorter multiwalled CNTs (MWCNTs) were easier to
be internalized than the longer ones. Besides the transportation
efciency, the cellular toxicity of the CNT-based delivery vectors
was another predominant characteristic that should be considered.
During the past decade, many works have examined in vitro toxicity of CNTs [56]. The CNTs exhibit some degree of cytotoxicity;
however, the cytotoxicity is dependent on the route of preparation
and their surface functionalization. Some studies demonstrated
that exposure of mammalian cells (human epidermal keratinocytes
[57], macrophages, human A549 lung cells lines [58], etc.) to
pristine CNTs results in oxidative stress in cells as well as induction of cellular apoptosis and necrosis which causes depletion of
total antioxidant reserves and loss of cell viability [59]. In contrast to pristine CNTs, the functionalized CNTs showed improved
biocompatibility [60]. Different types of functionalized CNTs, for
example, phenyl-SO3H-functionalized SWCNTs [61], polyethylene

glycol-modied MWCNTs [62], RNA-modied SWCNTs, proteinfunctionalized CNTs [63], etc., have been investigated in various
laboratories resulting in no signicant cell damage. On the other
hand, the cytotoxicity seems to be also dependent on the physical properties of CNTs such as size and morphology. For example,
Magrez et al. [64] and Tian et al. [65] have shown that the SWCNTs
are more toxic than MWCNTs. In general, the physical properties and surface functionalization of CNTs are the key factors that
determine the transportation efciency and cytotoxicity of CNTbased carrying materials. Furthermore, in the study of Liu et al.
multi-walled CNTs (MWCNTs) of different length were functionalized with chitosanfolic acid nanoparticles (CSFA NPs) using
ionotropic gelation process. This vector was used to transfect MCF7 and Hela cells with GFP gene. The nanotube length showed a
compromise inuence on the transfection and cytotoxicity properties of MWCNTs. In the study of Behnam et al., single-walled
carbon nanotubes (SWNTs) were functionalized by non-covalent
binding of hydrophobic moieties, which were covalently linked to
polyethyleneimines. Their results showed that PEI-functionalized
SWCNTs exhibited a good stability and dispersibility in biological
media.
One of the most relevant chemical modications to create
carboxylic acid groups on the multi-walled carbon nanotubes
(MWCNTs) is oxidation which [66,67] provides opportunity to
functionalize MWCNTs for conjugation of surfactant polymers [68].
At the same time, oxidation damages nanotubes, resulting in structural defects, shortening of tubes, accumulation of carbonaceous
impurities, and loss of small diameter nanotubes [66,67]. To oxidize nanotubes efciently and preventing signicant material loss,
oxidation conditions should be selected carefully. Nitric acid has
been the mostly utilized agent for oxidation of carbon nanotubes
[67]. It can be used solely in boiling temperature or in combination with concentrated sulfuric acid [69]. In order to maximize
the loading of nucleic acids onto the surface of carboxylated carbon nanotubes, a highly cationic-charge density polymer such as
polyethyleneimine (PEI) has been used [68,70]. Across the literature
the branched 25 kDa PEI polymer is favored as a gene transfer agent
[30]. The delivery of genes to MSCs as well as PEI-modied CNT carriers are well-documented separately, however; the application of
PEI-modied CNTs as non-viral gene delivery vectors for MSCs is
not reported. Thus, in this study, a PEI-grafted MWCNT in combination with chitosan substrate as a nanocarrier system for gene
delivery is developed and its transfection efciency in bone marrow mesenchymal stem cells (BMSCs) is evaluated using plasmid
DNA encoding enhanced green uorescent protein (pEGFP) as an
exogenous reporter gene. The PEI binding, particle size distribution,
and colloidal stability of the functionalized CNTs were analyzed by
Fourier transform infrared (FT-IR) spectra, dynamic light scattering (DLS), and zeta potential, respectively. DNA binding afnity,
cellular uptake, transfection efciency and possible cytotoxicity
were also tested by agarose gel electrophoresis, ow cytometry,
cytochemisty and MTT assay.

2. Experimental methods
2.1. Materials
PEI (bPEI-25K), Chitosan (medium molecular weight),
Hyaluronan (HA), Multi-walled carbon nanotubes, uorescein5-isothiocyanate (FITC), EDC and sulfo-NHS were obtained from
SigmaAldrich; plasmid DNA encoding enhanced green uorescent
protein (pEGFP) were purchased from Invitrogen (USA), A large
amount of these plasmids were prepared using a Qiagen Plasmid
Maxi Kit (Qiagen, Germany), Fetal bovine serum (FBS), alpha-MEM
media, opti-MEM serum-free media, PBS buffer, Trypsin/EDTA and

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

penicillin-streptomycin were purchased from GIBCO (USA); MTT


(3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide)
assay kit was obtained from Roche (Germany).
2.2. Preparation of functionalized CNTs
The MWCNTs were rstly treated with concentrated
HNO3 /H2 SO4 solution. In a typical procedure, 50 mg of MWCNTs were suspended in 40 mL of a 3:1 mixture of concentrated
H2 SO4 (98wt%)/HNO3 (16 M) and transferred into a roundbottom ask equipped with a magnetic stirring bar and a reux
condenser. The ask was immersed in an oil bath at 120 C.
The mixture was reuxed for 2 h and cooled down to room
temperature. The resultant suspension was ltered to collect
oxidized-MWCNTs (ox-MWCNTs) on a 100-nm-pore membrane
lter and washed with deionized water until the ltrate was
neutral followed by overnight drying in vacuum at 4 C. The
consequential MWCNTsCOOH were coated with PEI, following the EDC (1-ethyl-3-(3-dimethylaminopropyl) carbodiimide
hydrochloride) chemical activation of the COOH groups that yields
a PEI/MWCNT complex. The ratio of PEI to MWCNTs used for
functionalization process was from 0.1 to 0.6. Subsequently, the
complex was dispersed in 100 mM MES buffer at concentration
of 1 mg/mL using an ultrasonic bath for 1 h, at room temperature.
The suspension was then centrifuged at 22,000 g for 30 min to
remove bulky ox-MWCNTs and impurities whereas the excess
PEI was removed through ltration in Millipore Microcon 50,000
molecular weight cut-off spin lter. The obtained PEI-graftedMWCNT (PEI-g-MWCNT) solution was stored at 4 C for further
experiments. Fig. 1 represents these procedures schematically.
2.3. Characterization of PEI-g-MWCNT
2.3.1. Dispersability of PEI-g-MWCNT
To determine the best sample in terms of solubility a dispersion of functionalized nanotubes with different surfactant ratio was
diluted with deionized water in 10 mL glass vials to yield series of
concentrations ranging from 0.01 to 3.5 mg/mL. Each vial was sonicated by an ultrasonic bath for 10 min. After standing for 3 h, the
dispersions were analyzed for the presence of visible particles. The
highest concentration of functionalized nanotubes with no visible
particles is reported as the solubility value for each sample.
2.3.2. FT-IR analysis of functionalized CNTs
Fourier transform infrared (FT-IR) spectra recorded on Nicolet
400 Fourier transform infrared spectrometer (Madison, WI) for the
sample which was selected from the solubility test.
2.3.3. Dynamic light scattering and zeta potential measurements
The mean particle size of the ox-MWCNT and PEI-g-MWCNT
solutions was determined by dynamic light scattering using photon correlation spectroscopy. The measurements were performed
using a Zetasizer Nano ZS (Malvern Instruments Ltd., Malvern, UK)
equipped with a helium-neon laser at 25 C and a scattering angle
of 173 . Additionally, the zeta potential of the same solutions was
measured with the same instrument at 25 C by electrophoretic
mobility.
2.3.4. DNA binding assay
PEI-g-MWCNT/pDNA (C = 250 g/mL) complexes were prepared by mixing dilutions of PEI-g-MWCNT consist of 2, 1, 1/2,
1/4, 1/6, 1/8, 1/10, 1/20, 1/100 and 1/200 to reach different N:P
ratios (where N = number of primary amines in the PEI; P = number
of phosphate groups in the pDNA backbone). Phosphate buffered
saline (PBS) solution was used to prepare the solutions. Complex

117

solutions were gently vortexed and allowed to incubate for 30 min


at room temperature prior to experiment.
The binding of DNA to PEI-g-MWCNT was assessed by agarose
gel electrophoresis. Each PEI-g-MWCNT/pDNA dispersion with a
desired N:P ratio was mixed with 6 loading buffer (bromophenol
blue/xylene cyanol) and loaded on a 1%(w/v) agarose gel in trisacetateEDTA (TAE) buffer. The amount of pDNA loaded into each
well was 2.5 g in total volume of 10 L. The electrophoresis was
performed under 100 V for 45 min. Ethidium bromide was used for
pDNA staining and their bands were visualized under UV.

2.4. In vitro gene transfection


2.4.1. BMSCs isolation and expansion
Rat bone marrow-derived mesenchymal stem cells (BMSCs)
were isolated from long bones of 8 week-old male Wistar rats
according to a standard protocol [40]. Following euthanasia by pentobarbital 20% (v/v), the femora were aseptically excised, cleaned
of soft tissue, and washed in PBS. Subsequently, marrow was
ushed out from the tibiae and femora with 5 mL of -Minimum
Essential Medium (-MEM) using a 23-Gauge needle and syringe.
The cells were centrifuged (600 g, 5 min), suspended in fresh
medium containing 10% heat-inactivated FBS, 100 U/mL penicillin
and 100 g/mL of streptomycin and seeded in 25 cm2 asks. Flasks
were incubated at 37 C in 5% CO2 and 90% humidity. After removal
of non-adherent cells and medium exchange at day 3, colonies
became compact and cells were detached with 0.25% trypsin/EDTA
at day 7. Following cultures were passaged at 57 day intervals and
expanded to passage 5 for further experiments.

2.4.2. Preparation of chitosan-modied substrates


To improve transfection efciency, chitosan substrates were
prepared as previously described by Hsu et al. [71]. Briey, chitosan was dissolved in 1% acetic acid by gentle stirring for 12 h
at room temperature. The solution (1% chitosan) was ltered and
then coated on coverslip glass (100 mL of solution on each 15 mmdiameter glass) and air-dried for 2 days. The chitosan membranes
formed on the coverslip glass were immersed in 0.5 N NaOH solution for about 5 min, and washed extensively by distilled water. The
chitosan membranes were air-dried for further transfection experiments. These samples (1.5 cm in diameter) t the size of each well
in a 24-well plate.

2.4.3. BMSCs transfection by PEI-g-MWCNT/pDNA complexes


Plasmid DNA encoding enhanced green uorescent protein
(pEGFP) was used as an exogenous reporter gene to assess transfection efciency. Cells were cultured until 6570% conuency and
seeded at a density of 15 104 cells per well in adherent 24 well
Plates 24 h prior to transfection. Media was removed an hour before
transfection and the monolayer of cells was washed with 1 mL of
PBS per well. Subsequently, 500 l of OptiMEM serum-free media
was added to each well and incubated at 37 C for one hour. PEI-gMWCNT/pDNA complexes were produced in OptiMEM at varying
doses (1 g, 2 g, 4 g) of pDNA and incubated at room temperature for 30 min. Transfection medium (100 l) were then added
to the cells and supernatant was removed after approximately 4 h
followed by PBS wash. Complete culture media was added and the
cells were incubated at 37 C to allow expression of the transgene.
Additionally, HEK293 cells were cultured in 24-well culture plate
and transfected in the same manner as control. Non-transfected
cells and cells transfected with naked pDNA were used as negative controls. The transfection efciency of PEI-g-MWCNT/pDNA
was directly analyzed by imaging under a uorescence microscope
(Nikon, Germany).

120

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

Fig. 5. Agarose gel electrophoresis for the PEI-g-MWCNTs presenting the effective
binding of plasmid DNA with different dilution factors of 2, 1, 1/2, 1/4, 1/6, 1/8, 1/10,
1/20, 1/100 and 1/200, lane 110, respectively.

3. Results
3.1. Characterization of PEI-g-MWCNT
The solubility is dened as the maximum weight percent of
nanotubes that can be dispersed in a solvent with no particle visible
for at least 2 h. Fig. 2(A) reports the solubilities of PEI-g-MWCNTs
in water with different weight fraction of PEI to MWCNTs. The solubility of mixture enhanced rapidly at rst as the ratio of the PEI
to MWCNTs was increased. Functionalization of nanotubes with
greater amounts of PEI (>0.3 (w/w)) gave smaller increases of solubility. Fig. 2(B) shows a PEI-g-MWCNT dispersion with optimum
ratio of PEI to MWCNT (0.3 (w/w)) after 3 months.
The FT-IR was employed to characterize the formation of oxidized MWCNTs and PEI-g-MWCNTs with optimum surfactant to
nanotube ratio. As shown in Fig. 3, spectrum of the oxidized

MWCNTs (curve a) and PEI-g-MWCNTs (curve b) presented the


characteristic peaks which indicated the successful oxidation of
MWCNTs by acid treatment process and ideal assembly of the
PEI-g-MWCNTs, respectively. In curve a, spectra of oxidized MWCNTs shows four major peaks, located at 3450, 2370, 1740 and
1635 cm1 . The broad peak at 3450 cm1 refers to the OH stretch
from carboxyl groups (O COH and COH), while the peak at
2370 cm1 can be associated with the OH stretch from strongly
hydrogen-bonded COOH [72,73]. The peak at 1740 cm1 is associated with the stretch mode of carboxylic groups indicating that
carboxylic groups are formed due to the oxidation of carbon atoms
on the surfaces of the MWCNTs by nitric and sulfuric acids [74,75].
The peak at 1635 cm1 can be associated with the stretching of
the carbon nanotubes backbone [72]. In addition, peak assigned to
the CO stretching corresponds to the oxidized MWCNTs at the
1165 cm1 could be clearly observed in curve a [76]. The peaks
at 2845 and 2920 cm1 correspond to the HC stretch modes of
HC O in the carboxyl group. In curve b, the strong and broad
absorption peak located at 3450 cm1 could be attributed to the
NH bond of PEI and OH groups in PEI and MWCNTs, a NH2
bending peak at 1559 cm1 and a CN stretching vibration at
1460 cm1 were also detected [77].
Dynamic light scattering (DLS) measurement was carried out to
determine the size distribution of the formed complexes (Fig. 4(A)).
Conjugation of hydrophobic chains of PEI to MWCNTs resulted
in a signicant

120

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

Fig. 5. Agarose gel electrophoresis for the PEI-g-MWCNTs presenting the effective
binding of plasmid DNA with different dilution factors of 2, 1, 1/2, 1/4, 1/6, 1/8, 1/10,
1/20, 1/100 and 1/200, lane 110, respectively.

3. Results
3.1. Characterization of PEI-g-MWCNT
The solubility is dened as the maximum weight percent of
nanotubes that can be dispersed in a solvent with no particle visible
for at least 2 h. Fig. 2(A) reports the solubilities of PEI-g-MWCNTs
in water with different weight fraction of PEI to MWCNTs. The solubility of mixture enhanced rapidly at rst as the ratio of the PEI
to MWCNTs was increased. Functionalization of nanotubes with
greater amounts of PEI (>0.3 (w/w)) gave smaller increases of solubility. Fig. 2(B) shows a PEI-g-MWCNT dispersion with optimum
ratio of PEI to MWCNT (0.3 (w/w)) after 3 months.
The FT-IR was employed to characterize the formation of oxidized MWCNTs and PEI-g-MWCNTs with optimum surfactant to
nanotube ratio. As shown in Fig. 3, spectrum of the oxidized

MWCNTs (curve a) and PEI-g-MWCNTs (curve b) presented the


characteristic peaks which indicated the successful oxidation of
MWCNTs by acid treatment process and ideal assembly of the
PEI-g-MWCNTs, respectively. In curve a, spectra of oxidized MWCNTs shows four major peaks, located at 3450, 2370, 1740 and
1635 cm1 . The broad peak at 3450 cm1 refers to the OH stretch
from carboxyl groups (O COH and COH), while the peak at
2370 cm1 can be associated with the OH stretch from strongly
hydrogen-bonded COOH [72,73]. The peak at 1740 cm1 is associated with the stretch mode of carboxylic groups indicating that
carboxylic groups are formed due to the oxidation of carbon atoms
on the surfaces of the MWCNTs by nitric and sulfuric acids [74,75].
The peak at 1635 cm1 can be associated with the stretching of
the carbon nanotubes backbone [72]. In addition, peak assigned to
the CO stretching corresponds to the oxidized MWCNTs at the
1165 cm1 could be clearly observed in curve a [76]. The peaks
at 2845 and 2920 cm1 correspond to the HC stretch modes of
HC O in the carboxyl group. In curve b, the strong and broad
absorption peak located at 3450 cm1 could be attributed to the
NH bond of PEI and OH groups in PEI and MWCNTs, a NH2
bending peak at 1559 cm1 and a CN stretching vibration at
1460 cm1 were also detected [77].
Dynamic light scattering (DLS) measurement was carried out to
determine the size distribution of the formed complexes (Fig. 4(A)).
Conjugation of hydrophobic chains of PEI to MWCNTs resulted
in a signicant

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

121

changes reect the successful surface modication of MWCNTs


suggesting that the surface potentials of MWCNTs can be manipulated through PEI-mediated reactions.
3.2. Optimization of DNA binding
The potential of functionalized MWCNTs to neutralize, bind and
condense plasmid DNA was studied by agarose gel electrophoresis,
due to its importance for efcient gene delivery. The plasmid used
for this study was the pEGFP that encodes the enhanced green uorescent protein. Agarose gel electrophoresis assay revealed that
binding of plasmid DNA to functionalized MWCNTs inhibits ethidium bromide intercalation [78], as the pDNA is in a condensed form.
The level of binding can thereby be assessed by the measurement of
the non-bounded DNA (Fig. 5, lanes 510). The DNA binding capacity of dispersed PEI-g-MWCNTs can be estimated by reference to the
lowest concentration of nanotubes that demonstrates detectable
DNA binding (lane 5 in Fig. 5). Based on these results, a dilution
of 1/4, which can package complete pDNA was selected for further
experiments.
3.3. In vitro gene transfection
In order to determine the effectiveness of functionalized MWCNTs in transfecting mammalian cells, two different cell lines (i.e.
BMSC and HEK293) were transfected in monolayer culture with different amounts of pEGFP (2, 4 and 6 g/mL) complexed by optimum
ratio of PEI-g-MWCNTs. Subsequently, gene delivery was qualitatively studied 60 h post-transfection by morphological analysis
and expression of EGFP using optical and uorescence microscopy,
respectively (Fig. 6). The uorescent green color of the cultures is
indicative of successful transfection of PEI-g-MWCNTs/pDNA and
EGFP expression. The highest EGFP expression was observed in both
cell lines which transfected with 4 g/mL of pDNA. Furthermore,
results show that by increasing the amount of pDNA (6 g/mL),
the transfection efciency was decreased possibly because of cytotoxic effect of MWCNTs in higher doses. No uorescent green
color was observed for control groups and cells that were transfected with complexes containing 2 g of pDNA/mL. Furthermore,
the transfection efciency of cells transfected on chitosan substrate was more than cells which transfected on untreated culture
plate.
3.4. BMSCs viability assay
The MTT assay was performed to evaluate viability of BMSCs
24 and 60 h post transfection by incubation of cells with PEIg-MWCNTs and PEI-g-MWCNTs/pDNA at concentrations ranging
from 0 to 100 g/mL. As shown in Fig. 7, there was no statistically signicant difference in the cell viability of BMSCs treated
with PEI-g-MWCNTs/pDNA complexes compared to the controls at
concentrations of 037.5 g/mL (p > 0.05) after 24 h. However, PEIg-MWCNTs/pDNA start to exhibit cytotoxicity at concentrations of
50 g/mL and higher. Since free PEI-g-MWCNTs were cytotoxic at
all concentrations, it can be concluded that cytotoxicity of PEI-gMWCNT/pDNA was considerably lower than free PEI-g-MWCNT to
bone marrow mesenchymal stem cells. Furthermore, the cell viability of BMSCs treated with PEI-g-MWCNTs/pDNA complexes at
concentrations of 05 g/mL was not signicantly different compared to the controls 60 h after the transfection (p > 0.05).
3.5. Cellular uptake of PEI-g-MWCNT/pDNA
To determine the uptake of PEI-g-MWCNT/pDNA complex by
BMSCs, FITC-labeled PEI-g-MWCNTs were incubated with cells
and examined by ow cytometry. To evaluate the propensity

Fig. 7. The viability of BMSCs transfected with PEI-g-MWCNTs/pDNA. The cell viability of BMSCs is shown (A) 24 h and (B) 60 h after transfection with various
concentrations of PEI-g-MWCNTs. Each bar represents the mean standard deviation (n = 3) *p < 0.05 when compared with non-transfected cells.

of PEI-g-MWCNT/pDNA to be internalized, the intracellular uorescence values were determined by measuring non-transfected
cells uorescent (Fig. 8(A) and (C)). It can be seen that
non-transfected cells were unable to exhibit FITC-uorescent
(RN1 = 0.08%, Fig. 8(A)). Additionally, almost all cells (89.17%) displayed PEI-g-MWCNT/pDNA uptake after 4 h incubation (Fig. 8(B)).
These experiments provide quantitative data, which indicate an
effective uptake of PEI-g-MWCNTs/pDNA by the bone marrow mesenchymal stem cells.
4. Discussion
Recently, due to the limitations associated with protein delivery
[79,80], gene therapy has alternatively been applied to provide sustained protein production by transfected cells. Very little work has
been reported using functionalized-MWCNTs as the BMSCs transfection agent. In this study, we developed and evaluated the use of
PEI grafted MWCNT as a gene delivery vector for expression of a
model gene (EGFP) in primary BMSCs. Even though, MWCNTs have
been utilized in the past for gene delivery, their application has been
usually evaluated in immortal cells that are not clinically useful
[44]. On the other hand, BMSCs are multipotent cells that have the
ability to differentiate into multiple lineages and have remarkable
potential for related tissues construction and regeneration [24,25].
In view of the fact that there is no ideal vector for effective gene
transfer in BMSCs, a delivery system using PEI-g-MWCNT in corporation with chitosan substrates was successfully optimized for
BMSC transfection in monolayer cultures.

122

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125

Fig. 8. Uptake of FITC-labeled PEI-g-MWCNTs/pDNA complexes were evaluated by ow cytometry. The intracellular uorescence values were determined by measuring
non-transfected cells uorescent (A). Almost all of the cells (89.17%) displaying PEI-g-MWCNT/pDNA complex uptake after 4 h incubation (B). Population of cells was used
to study uorescent emition (C, D).

A major obstacle in stem cell transfection and gene therapy


is the lack of an efcient vector for non-viral transfection. The
unique properties of CNTs such as high surface area (theoretically
1300 m2 /g) have made this nanocarrier more attractive than any
other non-viral vectors for biomolecular delivery [81]. However,
there are still remaining problems when using carbon nanotubes
as a biological carrier. The main problem is their inherent difculty in handling as they tend to aggregate, thus their application in
aqueous media particularly in biological medium is limited [43,68].
Furthermore, the cytotoxic effect of carbon nanotubes is a major
obstacle in utilizing them for clinical applications [82]. The CNTs
have so many features which could considerably inuence their
toxicity including: surface charge, agglomeration, concentration,
and purity. Even thought, pristine CNTs are not suitable for direct
transfer of genes into biological system, it has been shown that
functionalization of CNTs with hydrophilic agents or using an optimized concentration can decrease their intrinsic toxicity and allow
their applications in biomedical eld. Therefore, development of
functionalization methods to obtain soluble carbon nanotubes with
less cytotoxic effect is primordial. Although the capability of different functionalized CNTs as non-viral vehicle has been previously
explored, the surfactant ratio and CNT doses specically for BMSCs
have not been optimized yet [68,83]. By optimizing the PEI-gMWCNT/pDNA complex specically for BMSCs this research would
nd a greater clinical relevance. Our results demonstrate that the
PEI-g-MWCNTs has the property of homogenously dispersing in
water, where it has been observed to be stable for over 3 months
at room temperature, without any solid phase aggregation. This
probably relates to the stability of the covalent bond between the
PEI polycation and the MWCNTsCOOH, which occurs through an

amidic bond. Moreover, solubility increases as the number of polar


functional groups on the nanotubes enhances. The oxidation reactions by nitric and sulfuric acids would create many carboxylic
acid functional groups on the side walls of CNTs [8486]. Although
several methods have been reported for determination of the solubility of carbon nanotubes [8789] neither the term solubility
nor a standard method for its determination has been established
yet. Most solutions of MWCNTs are in fact dispersions. We found
that the ratio of surfactant to MWCNT was also crucial in order to
optimize the dispersion [90]. As indicated in the dispersion curves
(Fig. 2(A)), MWCNTs were dispersed by conjugating with a range
of PEI concentrations. It is evident that by increasing the amount of
PEI used for grafting, the amount of dispersed MWCNTs enhances
until a maximum is reached where the optimum conditions for
dispersion are obtained. Above this optimal concentration of PEI,
the yield of dispersed nanotubes is approximately constant. This
is likely due to limited amount of MWCNTs being shared between
high concentrations of surfactants.
The length distribution of nanotubes in solution has been
measured within hydrodynamic approximations using dynamic
light scattering (DLS) [91,92]. DLS provides information for length
distribution whereas the zeta potential measurement provides
information for the degree of dispersion. The length of nanotubes
is closely related to their degree of dispersion. It is therefore necessary to combine both methods to get comprehensive information
in evaluating dispersivity and dispersion stability of nanotubes.
The FT-IR spectra and electrophoresis results showed that the
covalent binding of PEI surfactants to MWCNTs increased the efcient binding of plasmid DNA and enhanced DNA condensation. It
is evident that the covalent attachment of cationic surfactants to

H. Moradian et al. / Colloids and Surfaces B: Biointerfaces 122 (2014) 115125


[93] E.G. Tierney, et al., The development of non-viral gene-activated matrices for
bone regeneration using polyethyleneimine (PEI) and collagen-based scaffolds,
J. Control. Release 158 (2) (2012) 304311.
[94] G.S. Huang, et al., Spheroid formation of mesenchymal stem cells on chitosan
and chitosan-hyaluronan membranes, Biomaterials 32 (29) (2011) 69296945.

125

[95] B. Kang, et al., Cell response to carbon nanotubes: size-dependent intracellular


uptake mechanism and subcellular fate, Small 6 (21) (2010) 23622366.
[96] C. Bussy, et al., Intracellular fate of carbon nanotubes inside murine
macrophages: pH-dependent detachment of iron catalyst nanoparticles, Part.
Fibre Toxicol. 10 (2013) 24.

You might also like