Biomaterials 33 (2012) 3604e3613

Contents lists available at SciVerse ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Nano-carrier for gene delivery and bioimaging based on carbon dots with
PEI-passivation enhanced fluorescence
Changjun Liu a, b, Peng Zhang a, Xinyun Zhai a, Feng Tian b, Wenchen Li a, Jianhai Yang a, Yuan Liu a,
Hongbo Wang a, Wei Wang a, Wenguang Liu a, *
a
School of Materials Science and Engineering, Tianjin Key Laboratory of Composite and Functional Materials, Tianjin University, Tianjin 300072, PR China
b
Institute of Medical Equipment, Academy of Military Medical Sciences, Tianjin 300161, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Polyethylenimine (PEI) functionalized carbon dots (CD-PEI) were fabricated by one-step microwave
Received 16 January 2012 assisted pyrolysis of glycerol and branched PEI25k mixture where the formation of carbon nanoparticles
Accepted 29 January 2012 and the surface passivation were accomplished simultaneously. In this hybrid C-dot, PEI molecule played
Available online 16 February 2012
two key roles in the system
as a nitrogen-rich compound to passivate surface to enhance the fluo-
rescence and as a polyelectrolyte to condense DNA. This CD-PEI was shown to be water soluble and emit
Keywords:
stable bright multicolor fluorescence relying on excitation wavelength. The DNA condensation capability
Carbon dots
and cytotoxicity of CD-PEI could be regulated by pyrolysis time possibly due to the somewhat destruction
Microwave
Gene delivery
of PEI during the formation of carbon dots. CD-PEI obtained at an appropriate pyrolysis time exhibited
Photoluminescence lower toxicity, higher or comparable gene expression of plasmid DNA in COS-7 cells and HepG2 cells
Bioimaging relative to control PEI25k. Intriguingly, the CD-PEIs internalized into cells displayed tunable fluorescent
emission under varying excitation wavelength, suggesting the potential application of CD-PEI in gene
delivery and bioimaging.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction spectra, tunable emission wavelength, and excellent biocompati-

The fusion of nanotechnology and medicine has led to the
emergence of nanomedicine where a nanomaterial platform is used
for gene/drug therapy and diagnostic probe [1,2]. The central
challenge in this field is to design multifunctional nano-carriers
that combine both therapeutic and diagnostic capabilities.
Recently, many kinds of inorganic nanomaterials, such as gold
nanoparticles [3,4], iron oxide nanoparticles [5,6], silica nano-
particles [7,8], semiconductor quantum dots [9e11], carbon nano-
tubes [12,13], nanodiamonds [14] and graphene [15], have been
explored as nonviral vector for nucleic acid delivery by chemical
modifications on purpose.
Lately, photoluminescent carbon dots (C-dots), as one of new
members of carbon nanomaterials family, have drawn tremendous
attention in nanotechnology field [16]. These surface-passivated
carbon nanoparticles, with diameters lower than 10 nm, have been
utilized for the application of biolabeling owing to their remarkable
advantages in stable photoluminescence (PL), broad excitation
* Corresponding author. Tel./fax: þ86 22 27404724.
E-mail address: wgliu@tju.edu.cn (W. Liu).
bility compared with conventional organic dyes and semiconductor
quantum dots which have raised serious health and environmental
hazard concerns [17]. Routes to construct C-dots can be generally
classified into two groups: top-down and bottom-up methods. The
top-down methods, by which C-dots are generally formed through
post-treating carbon particles broken from a larger carbon structure,
consist of arc discharge [18], laser ablation [19e21], electrochemical
oxidation [22e24]. The bottom-up approaches comprise combustion
[25,26], thermal carbonization [27], acid dehydration [28] and
ultrasonic treatment [29], by which C-dots are transformed from
suitable molecular precursors. The approaches mentioned above
always involve intricate processes, expensive original materials or
great energy-consuming devices, and the yield of C-dots is very low.
The synthesized C-dots, typically, are always oxidized by nitric acid
and further surface-passivated by diamine-terminated organic
molecule to gain obvious photoluminescence properties.
More recently, strong photoluminescent C-dots have been
produced directly via microwave pyrolysis of carbohydrates solu-
tion in the presence of passivation agent, 4,7,10-trioxa-1,
13-tridecanediamine (TTDDA) using a domestic microwave oven in
our lab [30]. The strong photoluminescent performance of C-dots,
as we found, has a close connection with the introduction of N

0142-9612/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2012.01.052
 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
atoms onto the skeleton of carbon nanoparticles. Other studies also
suggested that the N-containing diamine-terminated oligomeric
poly-(ethylene glycol) (PEG1500N) played an important role in the
surface passivation of C-dots [20,28]. That was reminiscent of
polyethylenimine (PEI), a commonly used high performance
nonviral vector, consisting of much more amino groups in its
molecular structure. So we hypothesize that, the use of PEI as the
surface passivation agent will able to aid in generating the C-dots
with both enhanced fluorescent properties and gene delivery
ability. However, the traditional surface functionalization process of
nanoparticles usually involves time-consuming multiple steps.
Previous investigations have mostly focused on the synthesis
process, but the poor gene condensation capability has restricted
the use of C-dots in gene delivery itself. Herein, we report a hybrid
nano-carrier based on PEI-functionalized C-dots (CD-PEI) via one-
step microwave assisted pyrolysis of inexpensive glycerol in the
presence of branched PEI in a few minutes. The most favorable
feature of this strategy is that the formation of carbon nanoparticles
and the surface passivation with PEI can be realized simultaneously
in one pot. It is expected that the outer cationic polymer layer has
the ability to mediate plasmid DNA transfection, and the entrapped
C-dots emitting discernible fluorescence can serve as bioimaging.
2. Materials and methods
2.1. Materials and chemicals
Branched polyethylenimine (PEI) with molecular weight 25 kDa was purchased
from SigmaeAldrich (St Louis, MO, USA). Quinine sulfate (98%, suitable for fluo-
rescence) was supplied by Fluka (New York, USA). Plasmid pGL3-control with SV40
promoter and enhancer sequences encoding luciferase (5262 bp) was obtained from
Promega (Madison, WI, USA). 3-(4,5-dimethyl-2-thiazoyl)-2,5-diphenyl tetrazolium
bromide (MTT, 98%) was supplied by Alfa-Aesar (Beijing, China). The plasmids were
amplified in Escherichia coli and purified with QIAGEN Plasmid Maxi Kit (Qiagen,
Hilden, Germany). Glycerol, NaH2PO4, Na2HPO4, and H2SO4 were obtained from
Guangfu Fine Chemical (Tianjin, China). And all other chemicals were analytical
reagent grade and were used without further purification.
2.2. Synthesis of CD-PEI
The photoluminescent C-dots passivated with branched PEI was synthesized by
a facile green route of microwave assisted pyrolysis method. Briefly, 20 ml glycerol
and 6 ml 10 mM pH7.4 phosphate solution were mixed with 0.5 g PEI25k under
vigorous stirring to form a homogeneous solution in a common 100 mL beaker. Then
the beaker containing clear transparent solution was placed at the center of the
rotation plate of a domestic microwave oven (700 W) and heated for different time
intervals. When cooled down to room temperature, the color-changed solution was
then diluted and dialyzed against pure water for 4 days. Finally, a clear, light yellow-
brown aqueous solution containing CD-PEI was lyophilized to collect dry CD-PEI. In
this experiment, three CD-PEIs were designed by varying microwave irradiation time,
5 min, 10 min and 15 min, denoted as CD-PEI-A, CD-PEI-B and CD-PEI-C respectively.
2.3. Instrumentation and characterizations
Fourier Transform Infrared Spectroscopy (FT-IR) spectra were conducted on
a Nicolet 380 spectrometer (Thermo, America). The elementary compositions were
further confirmed by elemental analysis with Vanio-EL (Elementar Analysensysteme
GmbH, Germany). X-Ray diffraction (XRD) patterns were recorded on a Rigaku-D/
MAX 2500 diffractometer (Rigaku, Japan) equipped with graphite mono-
chromatized CuKa (l ¼ 0.15405 nm) radiation at a scanning speed of 4 /min in the
range from 5 to 60 . The morphologies of the CD-PEI were examined by high-
resolution transmission electron microscopy (HRTEM) on a Philips Tecnai G2F20
microscope (Philips, Netherlands) with an accelerating voltage of 200 kV. UVeVis
absorption was characterized by TU-1810 UVeVis Spectrophotometer (Pgeneral,
China). Photoluminescence (PL) emission measurements were performed using
FLS920 Fluorometer (Edinburgh Instruments, Britain). The vector/pDNA complexes
were observed under JEM-100CX II TEM (Jeol, Japan) after negatively stained with
1.5 wt% phosphotungstic acid (pH 6.7). Zeta potential of vector/pDNA complexes was
measured by using a Zetasizer Nano Z (Malvern Instruments, Malvern, UK).
2.4. Measurement of fluorescence quantum yields
The quantum yield of the C-dots was determined by a comparative method.
Quinine sulfate in 0.1 M H2SO4 (literature quantum yield: 54%) was selected as
3605
a standard sample to calculate the QYof test sample (i.e. C-dots) which was dissolved in
ultrapure water at different concentrations. All the absorbance values of the solutions
at the excitation wavelength were measured with UVeVis spectrophotometer. Pho-
toluminescence (PL) emission spectra of all the sample solutions were recorded by
FLS920 Fluorometer at an excitation wavelength of 350 nm. The integrated fluores-
cence intensity is the area under the PL curve in the wavelength range from 380 to
700 nm. Then a graph was plotted using the integrated fluorescence intensity against
the absorbance and a trend line was added for each curve with intercept at zero.
Absolute values were calculated according to the following equation:

 !
GradX h2X
FX ¼ FST (1)
GradST h2ST
Where the subscripts ST and X denote standard and test respectively, F is the
fluorescence quantum yield, Grad is the gradient from the plot of integrated fluo-
rescence intensity vs absorbance, and h is the refractive index of the solvent. In order
to minimize re-absorption effects, absorbance in the 10 mm fluorescence cuvette
should never exceed 0.1 at the excitation wavelength.
2.5. Preparation of vectors/pDNA complexes
CD-PEI and branched PEI25k were separately dissolved in ultrapure water to
form 2 mg/mL solutions which were subsequently sterilized by filtration through
a 0.22 mm filter. Vector/pDNA complexes at varied ratios were then formulated by
adding sterilized vector of prescribed concentrations to an equal volume of a defined
pDNA solution, pipetting up and down to make the mixture homogeneous thor-
oughly. Then the mixtures were incubated at room temperature for 30 min to allow
complex formation completely. In this study, the complexing ratio was expressed as
the weight ratio of vector/pDNA. Noted that the selected complexing ratios in the
following assay correspond to the maximum transfection efficiency of each vector
unless otherwise statement.
2.6. Agarose gel electrophoresis
The binding of pDNA with CD-PEI was evaluated by agarose gel electrophoresis.
The vector/pDNA complexes at different weight ratios were prepared freshly as
described above. After incubation for 30 min at room temperature, 8 mL of complex-
containing solution was mixed with 2 mL loading buffer, and loaded into a 1% agarose
gel containing ethidium bromide (0.5 mg/mL). The electrophoresis experiment was
carried out for 40 min in 1  TAE buffer at a constant voltage of 100 V. Then the
pDNA bands were visualized under a UV transilluminator at a wavelength of 365 nm.
2.7. Cell culture
COS-7 cells (African green monkey kidney cells) and HepG2 cells (human
hepatocellular liver carcinoma line) were obtained from Peking Union Medical
College (Beijing, China). COS-7 cells and HepG2 cells were separately cultured in
Dulbecco’s Modified Eagle Medium (DMEM, HyClone) with high glucose, containing
10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 mg/mL streptomycin at
37  C in 5% CO2 humidified atmosphere.
2.8. Cell viability assay
The cytotoxicity of PEI-passivated C-dots was assessed through MTT assay. COS-
7 cells and HepG2 cells were seeded in a 96-well plate, at a density of 2  104 cells/
well and incubated overnight, respectively. The following morning, the culture
medium in each well was replaced by 180 mL fresh DMEM/FBS, and 20 mL CD-PEI/
pDNA complexes solutions with various weight ratios (DNA concentration was
2 mg/mL) were then added to each well. After incubation for 24 h, the medium
containing CD-PEI/pDNA complexes was removed, and replaced with 200 mL fresh
medium containing 20 mL MTT (5 mg/ml in PBS) and incubated for another 4 h.
Finally all medium was removed and 150 mL/well DMSO was added, followed by
shaking for 15 min. The absorbance of each well was measured at 490 nm using
a Synergy HT Multi-Mode Microplate Reader (BioTek, USA) with pure DMSO as
a blank. Non-treated cell (in DMEM) was used as a control and the relative cell
viability (mean%  SD, n ¼ 3) was expressed as Abssample/Abscontrol  100%.
2.9. In vitro transfection and luciferase assay
In vitro transfection of CD-PEI-A, CD-PEI-B and CD-PEI-C were assayed in COS-7
cells and HepG2 cells, respectively. The cells were seeded at a density of 5  104 cells/
well in 24-well plates and incubated for 24 h at 37  C, in 5% CO2 humidified
atmosphere. Just prior to transfection, the medium in each well was aspirated and
replaced with 450 mL of serum-free or 10% serum-containing DMEM. CD-PEI/pDNA
complexes (50 mL, containing 1 mg DNA) at various weight ratios were then added to
each well (n ¼ 3 for each ratio). After incubating at 37  C in 5% CO2 for 3 h, the CD-
PEI/pDNA complexes that were not internalized were removed by aspirating the
medium and replacing with fresh medium. The transfected cells were re-incubated
3606 C. Liu et al. / Biomaterials 33 (2012) 3604e3613

Fig. 1. (a) CD-PEI aqueous solutions under UV light illumination (365 nm, from left to right: CD-PEI-A, CD-PEI-B and CD-PEI-C); (b) PL spectra of CD-PEI-A, CD-PEI-B and CD-PEI-C
aqueous solutions at the same concentration of 0.5 mg/ml (excitation wavelength: 350 nm, maxima emission wavelength: 470 nm); (c) HRTEM image of CD-PEI (scale bar is 50 nm;
the inset shows size distribution histogram); (d) XRD patterns for CD-PEI.

Fig. 2. FT-IR spectra of raw PEI and PEI-passivated C-dots.

 C. Liu et al. / Biomaterials 33 (2012) 3604e3613 3607

for additional 48 h. After incubation, the culture medium was removed and the cells
were washed with PBS twice. The cells in each well were treated for 15 min with
150 mL of reporter lysis buffer (RLB, Promega) followed by freezeepumpethaw
cycles to ensure complete lysis. The lysate was centrifuged for 4 min at
13,000 rpm and the supernatant was collected for luminescence measurements. The
luminescence of each sample was measured by 1420 Multilabel counter (Wallac,
USA) using Bright-GloÔ luciferase assay system (Promega, USA) according to the
manufacturer’s protocol. The results were expressed as relative light units (RLU) per
milligram of cell protein, and the protein concentration of each well was measured
by a BCA protein assay (Pierce, Rockford, IL, USA). Branched polyethylenimine
25 kDa (PEI25k) served as a positive control. PEI25k/pDNA complex with weight
ratio of 2/1 was prepared as suggested protocol.
2.10. Confocal microscopy
For confocal microscopy observation, COS-7 cells were seeded over a glass
coverslip in 6-well plate for 24 h and then transfected with sterilized CD-PEI-B/
pDNA at weight ratio of 16:1 for 3 h at 37  C. The cells were then washed with
isotonic PBS (pH 7.4) three times, and fixed with 4% paraformaldehyde solution in
PBS at 4  C overnight. The samples were observed under a Leica laser scanning
confocal microscope (Mannheim, Germany) equipped with a UV laser (351/364 nm),
an Ar laser (457/488/514 nm) and a HeNe laser (543/633 nm). Non-transfected COS-
7 cells served as a negative control.
3. Results and discussions
3.1. Characterization of CD-PEI
In this work, we presented three PEI-passivated C-dots with
different microwave irradiation time. In the following discussion,
all the samples used for characterization were prepared under the
conditions of 20 ml glycerol, 6 ml 10 mM phosphate, 0.5 g PEI25kD
and microwave treatment for 10 min (CD-PEI-B) unless specifically
mentioned. As shown in Fig. 1a, the prepared C-dots passivated
with PEI exhibited excellent water-soluble properties and emitted
blue luminescence under UV light (365 nm). The full width at a half
maximum (FWHM) at excitation of 350 nm is about 90 nm (Fig. 1b)
indicating a narrow size distribution of C-dots. Quinine sulfate
(quantum yield 54%) was selected as a standard sample to calculate
the quantum yield of C-dots and the results (Fig. S1) are CD-PEI-A
9.4%, CD-PEI-B 15.3% and CD-PEI-C 7.0%, respectively. There is an
interesting phenomenon that only a proper microwave pyrolysis
time results in high performance photoluminescence. To under-
stand this, we assume that there exist three steps in the whole
pyrolysis process, that is, (i) glycerol dehydration and nanoparticle
formation, (ii) surface passivation of nanoparticle and (iii) growth
of carbon dots. The three steps have no well-defined boundary and
might occur synchronously. Therefore, the C-dots made by 5 min
and 15 min pyrolysis showing weak photoluminescence may be
ascribed to the incomplete surface passivation before 10 min and
large particle formation after 10 min [20]. Although the origins of
photoluminescence are not yet entirely understood in C-dots, there
is mounting evidence that emission arises from the radioactive
recombination of excitations located at surface energy traps [16].
The luminescence lifetime of PEI-passivated C-dots was also
measured, and the computed results are CD-PEI-A 5.38 ns, CD-PEI-

Fig. 3. (a) Fluorescent microscopy images (all scale bars: 500 mm) of diluted aqueous solution of CD-PEI observed under different excitation wavelengths from left to right:
ultraviolet, blue and green; the pictures were taken by an Olympus BX-51 optical system microscope (Tokyo, Japan). (b) Photoluminescence emission spectra of CD-PEI
(progressively longer excitation wavelengths from 340 nm to 500 nm with a 20 nm increment) and the emission spectral intensities are normalized in the inset. (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3608 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
B 5.74 ns and CD-PEI-C 5.09 ns (Fig. S2). Additionally, the influence
of pH value of CD-PEI solution on the PL intensity was also exam-
ined (Fig. S3). The results suggest that the PL intensity of CD-PEI is
very stable in the pH range of 5e12 with a slight increase in the
acidic conditions. The high-resolution transmission electron
microscopy (HRTEM) image, as shown in Fig. 1c, indicates that the
formed carbon nanoparticles have a diameter distribution of
4e12 nm and the average value is about 7 nm. X-ray diffraction
(XRD) pattern (Fig. 1d) displays a broad diffraction peak at
2q ¼ 18.3 , suggesting an amorphous nature. The composition of C-
dots passivated with PEI was revealed by Elemental analysis and
the results are as follows: CD-PEI-A C 47.53%, N 24.71%, H 9.82% and
O (calculated) 17.94%; CD-PEI-B C 54.13%, N 22.93%, H 9.41% and O
(calculated) 13.53%; and CD-PEI-C C 57.60%, N 18.27%, H8.75% and O
(calculated) 15.38%. The result implies that more aminos in PEI
were kept with short time microwave treatment.
FT-IR spectra were recorded to identify the organic functional
groups on PEI-passivated C-dots. As shown in Fig. 2, the new peak
at 1645 cm
1, the strong peaks of 1570 cm
1, 1472 cm
1 and
1309 cm
1 are attributed to amide I C]O, NeH, CH2 and CeN
respectively. The broad bands centered at 3360 cm
1 suggests the
existence of eOH and NeH. A noticeable difference between raw
PEI and PEI-functionalized carbon dots lies in the new absorption
band located at 1645 cm
1, and the absorption of NeH located at
1570 cm
1. As the time of microwave treatment increases, the
absorption of C]O becomes stronger and stronger, while the NeH
peak shows a declining trend. This suggests more amide groups
formed through longer pyrolysis endurance time and the PEI
molecules were chemically grafted onto the surface of carbon dots
in situ. In the glycerol pyrolysis process, the carboxylic groups
converted into amide groups immediately after they have formed
on the surface of nanoparticle. The grafted PEI molecules not only
render C-dots water soluble, but also passivate the surface of CD to
enhance the photoluminescence.
Scheme 1. Depiction of the formation of CD-PEI and CD-PEI/pDNA complex.
The CD-PEI could emit multicolor luminescence under different
excitation illuminations, which was observed through fluorescent
microscopy. Typically, a tiny drop of aqueous solution containing
the CD-PEI was placed on a cover glass. Fig. 3a shows that the drop
exhibits blue, green and red luminescence under ultraviolet
(330e385 nm), blue (460e495 nm) and green (530e550 nm) light
excitation, respectively. The preeminent multicolor photo-
luminescent feature has great potential application in bioimaging,
which will be discussed in the next section. To further investigate
the wavelength-dependence feature of C-dots, the PL spectra of CD-
PEI solution was recorded by changing the excitation wavelengths.
When the excitation wavelengths changes from 340 nm to 500 nm,
the emission peak position, as shown in Fig. 3b, is red-shifted from
450 (blue) to 550 nm (green), and the PL intensity decreased
remarkably, indicating a strong dependence on the excitation
wavelengths. Although, up to date, the origins of PL in C-dots are
not very clear, we reason that this wavelength-dependent
phenomena (or multicolor PL) may come from the different
distribution of emissive energy traps on the surface of C-dots, as
suggested in the literature [13,22].
The formation of surface passivation, just as reported [20], plays
a key role in the strong fluorescence of C-dots. To reconfirm the
estimate, we also constructed a parallel control experiment. Sample
X, Y and Z were microwave-assisted pyrolysis products from
PEI25k, glycerol and the mixture of glycerol and PEI respectively. As
shown in Fig. S4, sample X has no fluorescence and sample Y shows
weak fluorescence when excited at 350 nm. The quantum yield of
sample Y is only 4.63%, the same as reported in our previous work
[30], much lower than that of sample Z. When PEI was used as
a surface passivation agent, sample Z exhibits a very strong fluo-
rescent emission owing to the successful introduction of nitrogen-
containing organic molecules onto the surface of C-dots. It is worth
mentioning that the PEI, after treated by microwave irradiation
10 min, was nonemissive in the UVeVis range. Therefore the
 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
observed bright and colorful photoluminescence should arise from
the PEI-passivated carbon nanoparticles.
The influence of pyrolysis time and PEI25k dosage on the fluo-
rescence of PEI-passivated C-dots was also investigated thoroughly
(Fig. S5 and Fig. S6). As shown in Fig. S5, the PL intensity reaches
a maximum when the microwave treating time is 10 min and then
a downward trend appears with time extension. Originally,
a certain amount of water in the raw solution was necessary to form
a transparent homogenous solution with an appropriate viscosity
for carbonization uniformity. With heating, the water was evapo-
rated before the glycerol dehydration, which was the reason why
no obvious PL was observed before 2 min. The longer the irradiation
time is, the stronger the PL spectrum emits. But when time is longer
enough to exceed 20 min, the sample shows a relative weak PL
spectrum, suggesting the change in the diameter size or surface
structure of C-dots. From Fig. S6, one can find that a proper dosage
of surface-passivation reagent is necessary for strong phtolumi-
nescent C-dots as reported in our previous work [30]. As shown in
the figure, the optimal dosage of PEI25k for the preparation of CD-
PEI with highest QY is 0.5 g.
3.2. Formation of CD-PEI/pDNA complexes
As depicted in Scheme 1, the branched PEI-functionalized C-dots
condense DNA by electrostatic interactions between positively
charged PEI on the surface of CD-PEIs and negatively charged
phosphate backbone of pDNA. The ability of PEI-passivated C-dots
to bind DNA at different complex ratios was investigated by agarose
Fig. 4. Agarose gel electrophoresis patterns of complexes: (a) PEI25k/DNA, (b) CD-PEI-
A/DNA, (c) CD-PEI-B/DNA and (d) CD-PEI-C/DNA.
3609
gel electrophoresis assay. Fig. 4 shows that the intensity of DNA
migration bands in the agarose gel decreases gradually with the
increase of weight ratio. For raw PEI25k, CD-PEI-A, CD-PEI-B and
CD-PEI-C, the electrophoretic mobility of pDNA was thoroughly
confined at or above the weight ratio of 0.2, 0.3, 0.4 and 0.5,
respectively, suggesting all the CD-PEI samples are capable of
condensing DNA at a very low weight ratio, though the condensing
ability decreases for longer pyrolysis time.
Surface charge of the complexes also reflects the ability of
cationic vectors to condense DNA. Zeta potential of various CD-PEI/
pDNA complexes was measured to determine their surface charges.
Fig. 5a shows the zeta potential data of CD-PEI-A/pDNA, CD-PEI-B/
pDNA and CD-PEI-C/pDNA complexes at different weight ratios,
respectively. The zeta potential of vector/pDNA complexes exhibits
an obvious rising trend along with an increase in weight ratio from
2 to 24. Moreover, over the selected range of complexing ratio, the
zeta potential of three complexes is in the descending order:
CD-PEI-A/DNA > CD-PEI-B/DNA > CD-PEI-C/DNA. This trend also
reflects the order of DNA condensation ability, and the result is
consistent with that obtained by electrophoresis. Microwave irra-
diation can generate a fast heating effect in the original mixture,
causing glycerol dehydration, polymerization and carbonization. In
this process, the amino groups of PEI molecules are also consumed
by taking part in the reactions; the longer irradiation time could
Fig. 5. Zeta potential data (a) and particle size (b) of vector/DNA complexes at different
weight ratio.
3610 C. Liu et al. / Biomaterials 33 (2012) 3604e3613

Fig. 6. TEM images of negatively stained vector/pDNA complexes: CD-PEI-A/pDNA at 6:1 wt/wt (a), CD-PEI-B/pDNA at 16:1 wt/wt (b) and CD-PEI-C/pDNA at 24:1 wt/wt (c). The
scale bars in the all images are 50 nm.

lead to more consumption of amino groups. Since zeta potential is The morphology and average size of CD-PEI/pDNA complexes
determined mainly by the surface charge of vector/pDNA were further inspected by both transmission electron microscope
complexes, CD-PEI-C/DNA displays the lowest potential while CD- (TEM) and dynamic light scattering (DLS). As shown in Fig. 6, at the
PEI-A/DNA shows the highest. selected vector/pDNA complex weight ratios, CD-PEI-A, CD-PEI-B

Fig. 7. Cytotoxicity testing results of vectors/pDNA complexes with a constant DNA concentration of 2 mg/mL against (a) COS-7 cells and (b) HepG2 cells from an MTT assay. The
values represent percentage cell viability (means  SD, n ¼ 3).
 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
and CD-PEI-C are all able to condense DNA into nano-sized particles
(below 50 nm), which is favorable for the efficient cellular delivery.
The corresponding hydrodynamic diameters of complexes (Fig. 5b)
are larger than TEM results due to the hydration effect in aqueous
solution. Among all three nano-carriers, CD-PEI-A forms the
smallest nanoplexes with pDNA, which is attributed to its highest
surface positive charge.
3.3. Cytotoxicity evaluation of the CD-PEI
With integration of transfection ability and effectual bioimaging
in one single C-dot, CD-PEI nanovector is required to have not only
optical merits, high transfection efficiency but also low cytotoxicity.
To evaluate the cytotoxicity of three CD-PEI vectors, the relative
viabilities of COS-7 cells and HepG2 cells exposed to CD-PEI or
complexes were measured by using the MTT assay method with
branched PEI25k as a control(Fig. 7S and Fig. 7). All the samples
exhibit a dose-dependent cytotoxicity. It is well known that the
cytotoxicity of PEI25k is attributed to its abundant positive charges
which can cause serious cell membrane damage [31]. In comparison
to the PEI25k and PEI25k/DNA controls, CD-PEI and CD-PEI/DNA
complexes exhibit remarkably less toxicity in COS-7 cells and
HepG2 cells at the same concentrations or complexing ratios,
respectively. It is probably because large amount of amino groups of
PEI molecules were destroyed in the passivation process or
embedded in the coreeshell structure of the nanoparticles. As
shown in Fig. 7a, there are more than 80% COS-7 cells viable for CD-
Fig. 8. In vitro gene transfection efficiency of the vectors (CD-PEI-A, CD-PEI-B, and CD-PEI-C)/pDNA complexes in comparison with those of PEI25k and naked DNA (ND) at various
weight ratios in COS-7 cells in the absence (a) and in the presence of serum (b); HepG2 cells in the absence (c) and in the presence of serum (d). Results are presented as the
mean  SD in triplicate.
3611
PEI-A/DNA at weight ratio 6:1, CD-PEI-B/DNA at 8:1 and CD-PEI-C/
DNA at 20:1, respectively. Similar cytotoxicity results are obtained
for HepG2 cells (Fig. 7b), and about 80% cells keep viable for CD-PEI-
A/DNA at weight ratio 4:1, CD-PEI-B/DNA at 10:1 and CD-PEI-C/DNA
at 24:1, respectively. As the ratios of vector to DNA are raised up to
18:1, the cytotoxicity of CD-PEI-A/DNA and PEI25k/DNA against both
COS-7 and HepG2 cells is enhanced remarkably, but that of CD-PEI-B
and CD-PEI-C obtained at longer microwave treatment time show
a relative mild increase at the identical complexing ratio, possibly
due to the destruction of more aminos in PEI.
3.4. In vitro transfection of CD-PEI/pDNA complexes
In order to evaluate the transfection ability of PEI-passivated C-
dots, we transfected COS-7 and HepG2 cells with CD-PEI nano-
vectors both in serum-free and serum-containing conditions, and
the results are shown in Fig. 8. When COS-7 cells were transfected
in medium without serum (Fig. 8a), the luciferase expression of
pGL-3 (RLU/mg protein) mediated by CD-PEI-A at weight ratio of
6:1 and CD-PEI-B at 16:1 are higher than or comparable to that of
PEI25k. When the complex ratio further increases, the gene
expression shows a decrease trend, which is probably due to the
increased cytotoxicity effect at high dose of CD-PEI. For the trans-
fection of COS-7 cells in serum-containing medium (Fig. 8b), CD-
PEI-A and CD-PEI-B can also achieve a gene expression level
superior to PEI25k, despite slightly lower than serum-free case. In
the presence of the negatively charged serum, the optimal complex
3612 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
Fig. 9. Laser scanning confocal microscopy images of non-transfected cells as negative controls (a) and transfected cells (b). The samples were observed under bright field and
excited at 405 nm, 488 nm and 543 nm (All scale bars: 20 mm).
ratios of CD-PEI-A and CD-PEI-B are increased to 8:1 and 18:1,
respectively. Similar results of luciferase express are obtained in
HepG2 cells (Fig. 8c,d). It is noted that CD-PEI-C remains low
transfection efficiency over the whole complexing ratio range,
though generally it exhibits an increasing trend. From the above
results, we can draw the conclusion that with longer microwave
pyrolysis time, the transfection ability of C-dots first increases and
then decreases. As discussed above, during the formation of CD-PEI,
the amino groups in PEI are involved in reaction. As for CD-PEI-A,
which was only treated under microwave for 5 min, a small
amount of the amino groups were consumed. Nonetheless, such
slight destruction of PEI molecule does not affect its DNA protection
ability and buffer effect. Conversely, a slight loss of the positive
charged molecule could improve the biocompatibility of the
cationic carrier; on the other hand, the formation of coreeshell
structured CD-PEI may be the reason for the enhanced trans-
fection level. When the microwave treatment was further increased
to 10 min, an increased amount of PEI molecules may be damaged.
Under this condition, high transfection efficiency can still be ob-
tained by simply increasing the complexing ratio. However, when
the pyrolysis time lasts more than 15 min, too much amino groups
could have been destroyed. As a result, the positive charge of the
obtained C-dots may be too weak to effectively protect the DNA
from degradation, and meanwhile the reduced buffer capability
cannot lead to efficient escape of complexes from endosome either.
3.5. Confocal microscopy
CD-PEI fabricated in this study shows multicolor fluorescence
under different laser excitation. This remarkable fluorescent feature
endows CD-PEI with an intrinsic ability to combine dual functions
of gene delivery and bioimaging. To further confirm the multi-
functions of the C-dots, COS-7 cells were selected for the trans-
fection and labeling by CD-PEI-B/pDNA complexes. Fig. 9 displays
photographs of the transfected COS-7 cells and negative controls
captured by a laser scanning confocal microscope. We can see that
the transfected COS-7 cells become quite bright owing to the strong
fluorescence emitting from CD-PEI distributed in cytosol, while the
untreated cells remain dark, suggesting large amount of nano-
particles have been internalized into the cells. After 3-h trans-
fection, most nanoparticles uptaken appear in endosomal
compartments; while some complexes have already successfully
escaped from endosome in such short time. Consistent with fluo-
rescence microscopy assay results (Fig. 3), CD-PEI shows blue, green
and red color under 405 nm, 488 nm and 543 nm laser excitation
wavelength. This multicolor emission is a great advantage of the C-
dots over other labeling agents, because it gives us much freedom
to choose the wavelength for observation. It is noteworthy that,
when the excitation wavelength was red-shifted enough, the C-
dots could even exhibit near-infrared emissions; though relatively
weak, it may also have the potential to be used in the fluorescent
tracking in vivo.
4. Conclusion
In summary, we successfully constructed a high-efficient dual
functional nano gene vector based on PEI-passivated C-dots by one-
step microwave assisted pyrolysis of glycerol in the presence of
branched PEI25k. In this hybrid nano-carrier system, PEI played
two crucial roles: surface passivation to endow the C-dots with
strong photoluminescence, DNA condensation for gene trans-
fection. The elaborately fabricated CD-PEI exhibited excellent water
solubility and bright multicolor fluorescence with great photo-
stability. CD-PEI could mediate gene transfection in COS-7 and
HepG2 cells with higher or comparable efficiency as well as lower
cytotoxicity relative to pristine PEI25k. To optimize the perfor-
mance of CD-PEI vector, proper pyrolysis time was essential to
balance between formation of C-dots and destruction of PEI. The
multicolor fluorescence of the C-dot/DNA complexes could be
clearly observed in cytosol, demonstrating the great potential of
CD-PEI in bioimaging and biolabeling.
Acknowledgement
The authors gratefully acknowledge the support for this
work from the National Natural Science Foundation of China
 C. Liu et al. / Biomaterials 33 (2012) 3604e3613
(Grant 50973082) and National Science and Technology Major
Project of China (Grant 2012ZX10004801-003-007).
Appendix. Supplementary material
Supplementary material associated with this article can be found,
in the online version, at doi:10.1016/j.biomaterials.2012.01.052.
References
[1] Roy I, Ohulchanskyy TY, Bharali DJ, Pudavar HE, Mistretta RA, Kaur N, et al.
Optical tracking of organically modified silica nanoparticles as DNA carriers:
a nonviral, nanomedicine approach for gene delivery. PNAS 2005;102(2):
279e84.
[2] Barreto JA, O’Malley W, Kubeil M, Graham B. Stephan Holger, Spiccia L.
Nanomaterials: applications in cancer imaging and therapy. Adv Mater 2011;
23:H18e40.
[3] Guo S, Huang Y, Jiang Q, Sun Y, Deng L, Liang Z, et al. Enhanced gene delivery
and siRNA silencing by gold nanoparticles coated with charge-reversal poly-
electrolyte. ACS Nano 2010;4(9):5505e11.
[4] Lee Y, Lee SH, Kim JS, Maruyama A, Chen X, Park TG. Controlled synthesis of
PEI-coated gold nanoparticles using reductive catechol chemistry for siRNA
delivery. J Control Release 2011;155(1):3e10.
[5] Veiseh O, Kievit FM, Mok H, Ayesh J, Clark C, Fang C, et al. Cell transcytosing
poly-arginine coated magnetic nanovector for safe and effective siRNA
delivery. Biomaterials 2011;32:5717e25.
[6] Namgung R, Singha K, Yu MK, Jon S, Kim YS, Ahn Y, et al. Hybrid super-
paramagnetic iron oxide nanoparticle-branched polyethylenimine magneto-
plexes for gene transfection of vascular endothelial cells. Biomaterials 2010;
31:4204e13.
[7] Radu DR, Lai CY, Jeftinija K, Rowe EW, Jeftinija S, Lin VS. A polyamidoamine
dendrimer-capped mesoporous silica nanosphere-based gene transfection
reagent. J Am Chem Soc 2004;126:13216e7.
[8] He WT, Xue YN, Peng N, Liu WM, Zhuo RX, Huang SW. One-pot preparation of
polyethylenimine-silica nanoparticles as serum-resistant gene delivery
vectors: intracellular trafficking and transfection. J Mater Chem 2011;21:
10496e503.
[9] Lee H, Kim IK, Park TG. Intracellular trafficking and unpacking of siRNA/quantum
dot-PEI complexes modified with and without cell penetrating peptide:
confocal and flow cytometric FRET analysis. Bioconjug Chem 2010;21:289e95.
[10] Zhang P, Liu W. ZnO QD@PMAA-co-PDMAEMA nonviral vector for plasmid
DNA delivery and bioimaging. Biomaterials 2010;31:3087e94.
[11] Bardi G, Malvindi MA, Gherardini L, Costa M, Pompa PP, Cingolani R, et al. The
biocompatibility of amino functionalized CdSe/ZnS quantum-dot-doped SiO2
nanoparticles with primary neural cells and their gene carrying performance.
Biomaterials 2010;31(25):6555e66.
[12] Pantarotto D, Singh R, McCarthy D, Erhardt M, Briand JP, Prato M, et al.
Functionalized carbon nanotubes for plasmid DNA gene delivery. Angew
Chem Int Ed 2004;43(39):5242e6.
3613
[13] Liu M, Chen B, Huang J, Zhang L, Zhang Z. Polyamidoamine-grafted multi-
walled carbon nanotubes for gene delivery: synthesis, transfection and
intracellular trafficking. Bioconjug Chem 2011;22:2237e43.
[14] Zhang P, Yang J, Li W, Wang W, Liu C, Griffith M, et al. Cationic polymer brush
grafted nanodiamond via atom transfer radical polymerization for enhanced
gene delivery and bioimaging. J Mater Chem 2011;21:7755e64.
[15] Kim H, Namgung R, Singha K, Oh IIK, KIM WJ. Graphene oxide-
polyethylenimine nanoconstruct as a gene delivery vector and bioimaging
tool. Bioconjug Chem; 2011. doi:10.1021/bc200397j.
[16] Baker SN, Baker GA. Luminescent carbon nanodots: emergent nanolights.
Angew Chem Int Ed 2010;49(38):6726e44.
[17] Hardman R. A toxicologic review of quantum dots: toxicity depends on
physicochemical and environmental factors. Environ Health Perspect 2006;
114(2):165e72.
[18] Xu XY, Ray R, Gu YL, Ploehn HJ, Gearheart L, Raker K, et al. Electrophoretic
analysis and purification of fluorescent single-walled carbon nanotube frag-
ments. J Am Chem Soc 2004;126(40):12736e7.
[19] Cao L, Wang X, Meziani MJ, Lu F, Wang HF, Luo PG, et al. Carbon dots for
multiphoton bioimaging. J Am Chem Soc 2007;129:11318e9.
[20] Sun YP, Zhou B, Lin Y, Wang W, Fernando KAS, Pathak P, et al. Quantum-sized
carbon dots for bright and colorful photoluminescence. J Am Chem Soc 2006;
128(24):7756e7.
[21] Wang X, Cao L, Lu FS, Meziani MJ, Li H, Qi G, et al. Photoinduced electron
transfers with carbon dots. Chem Commun 2009;25:3774e6.
[22] Zhao QL, Zhang ZL, Huang BH, Peng J, Zhang M, Pang DW. Facile preparation of
low cytotoxicity fluorescent carbon nanocrystals by electrooxidation of
graphite. Chem Commun 2008;41:5116e8.
[23] Zheng LY, Chi YW, Dong YQ, Lin JP, Wang BB. Electrochemiluminescence of
water-soluble carbon nanocrystals released electrochemically from graphite.
J Am Chem Soc 2009;131:4564e5.
[24] Zhou J, Booker C, Li R, Zhou X, Sham TK, Sun X, et al. An electrochemical
avenue to blue luminescent nanocrystals from multiwalled carbon nanotubes
(MWCNTs). J Am Chem Soc 2007;129:744e5.
[25] Liu H, Ye T, Mao C. Fluorescent carbon nanoparticles derived from candle soot.
Angew Chem Int Ed 2007;46:6473e5.
[26] Tian L, Ghosh D, Chen W, Pradhan S, Chang X, Chen S. Nanosized carbon
particles from natural gas soot. Chem Mater 2009;21:2803e9.
[27] Yang ZC, Wang M, Yong AM, Wong SY, Zhang XH, Tan H, et al. Intrinsically
fluorescent carbon dots with tunable emission derived from hydrothermal
treatment of glucose in the presence of monopotassium phosphate. Chem
Commun 2011;47:11615e7.
[28] Peng H, Travas-Sejdic J. Simple aqueous solution route to luminescent car-
bogenic dots from carbohydrates. Chem Mater 2009;21:5563e5.
[29] Li H, He X, Liu Y, Huang H, Lian S, Lee ST, et al. One-step ultrasonic synthesis of
water-soluble carbon nanoparticles with excellent photoluminescent prop-
erties. Carbon 2011;49:605e9.
[30] Liu CJ, Zhang P, Tian F, Li WC, Li F, Liu WG. One-step synthesis of surface
passivated carbon nanodots by microwave assisted pyrolysis for enhanced
multicolor photoluminescence and bioimaging. J Mater Chem 2011;21:
13163e71.
[31] Moghimi SM, Symonds P, Murray JC, Hunter AC, Debska G, Szewczyk A.
A two-stage poly(ethylenimine)-mediated cytotoxicity: implications for gene
transfer/therapy. Mol Ther 2005;11:990e5.