Diglycidyl esters cross-linked with low molecular weight polyethyleneimine for magnetofection

Article abstract of DOI:10.1071/CH14731

Magnetic polyethyleneimine (PEI) complexes have demonstrated to be simple and efficient vectors for enhancing gene transfection. However, the high cytotoxicity of PEI restricts its further application in vivo. In this study, we synthesized several low cytotoxicity biodegradable cationic polymers derived from PEI (M_w 600) linked with diglycidyl tartrate (DT-PEI) or its analogues (diglycidyl succinate (DS-PEI) and diglycidyl malate (DM-PEI); D-PEIs for all 3 polymers). Moreover, a type of biocompatible magnetic nanoparticles (MNPs) with negative charges was prepared to assemble with D-PEIs/DNA complexes via electrostatic interactions. The magnetic ternary complexes have appropriate sizes of 120-150nm and zeta potential values of ～20-25mV. The transfection ability and cell viability of D-PEIs increased as the amount of hydroxyl groups increased in the repeat unit, which indicated that increasing the hydroxyl number in the backbone of D-PEIs can enhance gene expression and decrease cytotoxicity in A549 cells. Magnetofection of DT-PEI showed similar transfection efficiency with 30min incubation; in contrast, the standard incubation time was 4h. All three magnetic complexes displayed lower cytotoxicity when compared with those of PEI complexes in COS-7 and A549. These results indicated that these series of magnetic PEI derivatives complexes could be potential nanocarriers for gene delivery.

Full text of DOI:10.1071/CH14731

CSIRO PUBLISHING
Aust. J. Chem.
Full Paper
http://dx.doi.org/10.1071/CH14731
Diglycidyl Esters Cross-Linked with Low Molecular Weight
Polyethyleneimine for Magnetofection
A C
,
B C
,
A
A
Hao Yu,
Shufeng Li,
Liandong Feng, Yucheng Liu,
A
A
A
A D
,
Xiaoliang Qi, Wei Wei, Junjian Li, and Wei Dong
A
Center for Molecular Metabolism, Nanjing University of Science and Technology,
Nanjing 210094, China.
B
Key Laboratory of Developmental Genes and Human Disease in Ministry of Education,
Department of Biochemistry and Molecular Biology, Medical School of Southeast
University, Nanjing 211189, China.
C
These authors contributed equally to this work.
D
Corresponding author. Email: weidong@njust.edu.cn
Magnetic polyethyleneimine (PEI) complexes have demonstrated to be simple and efficient vectors for enhancing gene
transfection. However, the high cytotoxicity of PEI restricts its further application in vivo. In this study, we synthesized
several low cytotoxicity biodegradable cationic polymers derived from PEI (Mw 600) linked with diglycidyl tartrate
(DT-PEI) or its analogues (diglycidyl succinate (DS-PEI) and diglycidyl malate (DM-PEI); D-PEIs for all 3 polymers).
Moreover, a type of biocompatible magnetic nanoparticles (MNPs) with negative charges was prepared to assemble with
D-PEIs/DNA complexes via electrostatic interactions. The magnetic ternary complexes have appropriate sizes of
120–150 nm and zeta potential values of ,20–25 mV. The transfection ability and cell viability of D-PEIs increased as
the amount of hydroxyl groups increased in the repeat unit, which indicated that increasing the hydroxyl number in the
backbone of D-PEIs can enhance gene expression and decrease cytotoxicity in A549 cells. Magnetofection of DT-PEI
showed similar transfection efficiency with 30 min incubation; in contrast, the standard incubation time was 4 h. All three
magnetic complexes displayed lower cytotoxicity when compared with those of PEI complexes in COS-7 and A549. These
results indicated that these series of magnetic PEI derivatives complexes could be potential nanocarriers for gene delivery.
Manuscript received: 27 December 2014.
Manuscript accepted: 29 March 2015.
Published online: 5 May 2015.
Introduction
capacity of endosomal escape. However, the high molecular
weight, charge density, and non-degradability lead to high
cytotoxicity.[16,17] By contrast, low molecular weight (LMW)
PEIs have been demonstrated to exhibit a relatively low cyto-
toxicity, associated with negligible transfection efficiency.
To improve transfection activity and reduce cytotoxicity, many
studies have been focussed on cross-linking LMW PEIs through
biodegradable disulfide or ester bonds.[18–20]
The other critical factors affecting TE is the low amount of
DNA that manages to reach the target cells. In this regard, the
development of non-viral carriers that can be guided selectively
into specific cells is very significant, and thus could improve
the chance of gene expression.[21] Magnetic nanoparticles have
been applied in targeted gene delivery using an external
magnetic field. Furthermore, magnetofection can decrease the
amount of DNA injected, shortening the time necessary to reach
the desired target cells and improving significantly gene expres-
sion efficiency.[22] Considering these advantages, magnetofec-
tion has been widely researched and has achieved tremendous
progress for gene delivery in the past few decades. However,
substantial progressand successful clinical application dependon
the efficient, specific delivery of genes without systemic toxicity.
Furthermore, gene delivered and the therapeutic efficacy can be
Gene therapy is one of the promising therapeutic strategies in
treating inherited and acquired diseases. Successful gene therapy
depends on safe and efficient gene vectors. Although displaying
high gene transfection efficiency, the application of viral
vectors is limited by severe immunogenicity and inflammation
problems.[1,2] Therefore, non-viral vectors as alternative gene
carriers have been widely investigated to avoid the aforemen-
tioned problems.[3,4] In recent years, non-viral vectors have been
chiefly divided into two types: organic component vectors and
inorganic component nanoparticles. The first type includes
polymers,[5,6] cationic lipids,[7] peptides,[8,9] and carbon nano-
tubes[10]; and the second type includes calcium phosphate
particles,[11] quantum dots,[12] gold nanoparticles,[13] and mag-
netic nanoparticles.[14]
Cationic polymers are one of the most significant non-viral
vectors for the delivery of negatively charged DNA. Among
cationic polymers, polyethyleneimine (PEI) has been widely
studied as the gold standard. However, it is difficult for PEI
to be applied in the clinical field because of its cytotoxicity,
non-degradability, and low-transfection efficiency.[15] High
molecular weight (HMW) PEIs have a relatively high transfec-
tion efficiency (TE) due to their binding affinity with DNA and
Journal compilation Ó CSIRO 2015
www.publish.csiro.au/journals/ajc

B
H. Yu et al.
accurately monitored noninvasively and spatiotemporally.[23]
Because many magnetic nanoparticles (MNPs) have been used
in clinical settings for many years, there is a high potential that
these functional targeted MNPs will be applicable in clinical gene
therapy in the future.[23] But there are also some possible
limitations of magnetic gene delivery: (1) from a medical point
of view, magnetic drug targeting makes sense mainly in localized
stages of disease and less in metastatic stages where drug action
may be required throughout the body;[24] (2) the other major
limitation of magnetofection is the residual presence of a small
amount of magnetic nanoparticles following transfection.[25] For
magnetofection, gene carriers, such as PEI, are associated with
magnetic nanoparticles. In order to absorb negatively charged
DNA on their surface, PEI or other cationic polymers have been
designed to tailor magnetic nanoparticles by physical adsorption
method[26–28] or chemical conjugation method.[29–31] However,
magnetic PEI complexes exhibit non-negligible cytotoxicity.[29]
The cytotoxicity of a gene vector is significant to the clinical
application, which powerfullydepends on the biocompatibility of
vector materials. In this paper, we synthesized biocompatible
Fe3O4 magnetic nanoparticles and a series of low cytotoxicity
cross-linked PEI 600 (D-PEIs) using diglycidyl esters with
different numbers of hydroxyl groups. The ternary complexes,
MNP@D-PEIs/DNA, were assembled by electrostatic interac-
tions. The transfection efficiency and cytotoxicity of the ternary
complexes were investigated in A549 and COS-7 cells.
when excess diesters were used.[19] The structures of the D-PEIs
were confirmed by 1H NMR and the molecular weights of the
D-PEIs were measured by gel permeation chromatography
(GPC) (Fig. S1, Table S1).
Characterization of MNPs
The structure of the MNPs was confirmed by Fourier transform
infrared (FT-IR) spectroscopy as shown in Fig. 1a. The FT-IR
spectrum of typical MNPs shows absorption bands at 1618 and
1381 cmꢀ1, which are related to the stretching vibrations of
carboxyl salt. And the peaks at 2857 and 2926 cmꢀ1 are due to
the vibrations of the symmetric and asymmetric stretching of the
(a)
1381
1047
1618
2857
2926
3312
590
4000
3000
2000
1000
0
Wavenumber [cmꢂ1
]
Results and Discussion
(b)
Synthesis and Characterization of Polymer D-PEIs
The synthesis of D-PEIs is shown in Scheme 1. In a typical
procedure, tartaric acid reacted with allyl alcohol in the presence
of sulfuric acid to obtain the diester compound 1a. Then, the
product 1a was oxidized by m-chloroperoxybenzoic acid
(m-CPBA) to give the compound 2a. The structure of 2 was
confirmed by 1H NMR (see Supplementary Material). DT-PEI
(DT ¼ diglycidyl tartrate), DM-PEI (DM ¼ diglycidyl malate),
and DS-PEI (DS ¼ diglycidyl succinate) were prepared by
cross-linking reaction between PEI 600 Da and 2a, 2b, and 2c,
respectively. In the polymerization reaction, the mole weight of
PEI 600 was slightly higher than that of the diesters, otherwise
the clear colourless reaction mixture would change to pale
yellow gel-like materials. This could be due to the uncontrol-
lable cross-linking between the diglycidyl esters and LMW PEI
100 nm
Fig. 1. (a) FT-IR spectrum of the Fe3O4 nanoparticles obtained with
Na3Cit. (b) TEM image of MNPs.
R
R
O
O
1
1
R
1
OH
O
1a, 2a: R1 ꢀ OH, R2 ꢀ OH
1b, 2b: R1 ꢀ OH, R2 ꢀ H
1c, 2c: R1 ꢀ H, R2 ꢀ H
m-CPBA
O
COOH
O
HOOC
O
O
H2SO4
O
R
O
R
2
2
R
2
O
1
2
HN
2a
DT-PEI
H
N
O
OH
R
1
H
Ethanol
Reflux
N
N
N
O
ꢁ
2b
2c
x
y
N
H
DM-PEI
DS-PEI
O
N
OH
O
R
2
NH
NH
2
Scheme 1. Preparation route of D-PEIs.

Diglycidyl Esters Cross-Linked with Polyethyleneimine for Magnetofection
C
methylene group, and the absorption band at 1047 cmꢀ1 is
probably related to the hydroxyl group from sodium citrate
(Na3Cit).[32] In addition, the peak at 590 cmꢀ1 is attributed to
(Fe–O) in the Fe3O4 nanoparticles.[33] The morphology and size
were observed by transmission electron microscopy (TEM) as
shown in Fig. 1b. The TEM image shows that the MNPs have an
average diameter of ,10 nm. Photographs of the dispersions of
the MNPs are given in Fig. 2. The MNPs were well dispersed in
water under normal conditions (Fig. 2a). If the MNP solution
is subjected to a strong magnetic field, the nanoparticles can
rapidly gather on the side of the wall within 30 s (Fig. 2b). This
result indicated that the MNPs can easily be manipulated by an
external magnetic field.
cytoplasm.[34] Therefore, ternary complexes of MNPs, D-PEIs,
and DNA were assembled here by electrostatic interactions. The
binary complexes of D-PEIs and DNA were prepared first and
then complexed with MNPs.[27,35] The size and zeta potential of
MNP@D-PEIs/DNA were determined at different weight ratios
of MNP/DNA. For the application of magnetic gene delivery,
the size of the nanoparticles prepared should be sufficiently
small (,200 nm) to improve cellular uptake and blood circu-
lation time within body.[36] As depicted in Fig. 3, the sizes of the
complexes were between 130 and 150 nm and the zeta potentials
were in the range of ,20.0–24.0 mV. The average sizes were
increased by increasing the amount of MNP, this phenomenon is
probably due to the increased aggregation by increasing the
relative concentration of MNPs. The surface charge of the
complexes were reduced by increasing the relative concentra-
tion of MNP; these results contribute to the formation of MNPs
with a negative charge surface.[37] The morphology of
MNP@DT-PEI/DNA was visualized by TEM (Fig. 3c), which
showed that the MNPs were distributed in the magnetic com-
plexes. To investigate the ability of the magnetic complexes to
bind DNA, MNP@D-PEIs/DNA complexes were assessed
using agarose gel retardation assays. As shown in Fig. 3d, the
mobility of DNA was completely reduced, which suggests that
the magnetic complexes can concentrate DNA under magne-
tofection conditions.
Assembly and Characterization of Gene Complexes
The MNPs were functionalized with positively charged PEI due
to its electrostatic interaction with negatively charged nucleic
acid and the proton sponge effect, which enables release of
MNP–PEI–nucleic acid complexes from endolysosomes into
(a)
(b)
In Vitro Transfection
Optimization of Weight Ratio of D-PEIs/DNA on
Transfection
To achieve optimized weight ratio of D-PEIs/DNA, the
transfection efficiencies of these binary complexes (without
Fig. 2. Photographs of an aqueous MNP dispersion in a vial (a) in the
absence of a magnetic fieldand(b)in thepresenceof a magnetic fieldfor30 s.
PEI/DNA
MNP@DS-PEI/DNA
MNP@DT-PEI/DNA
PEI/DNA
MNP@DS-PEI/DNA
MNP@DT-PEI/DNA
(a)
(b)
35
30
25
20
15
10
5
MNP@DM-PEI/DNA
MNP@DM-PEI/DNA
180
160
140
120
100
80
60
40
20
0
0
PEI/DNA
0.4
0.6
0.8
PEI/DNA
0.4
0.6
0.8
MNP/DNA (w/w)
MNP/DNA (w/w)
(c)
(d)
20 nm
DNA
PEI
0.8
0.6
0.4
0.8
0.6
0.4
0.8
0.6
0.4
MNP@DT-PEI
MNP@DM-PEI
MNP@DS-PEI
Fig. 3. Characterization of the magnetic complexes. (a) Size and (b) zeta potential of the ternary magneticcomplexes at varying MNP-to-DNAweight ratios
of 0.4, 0.6, and 0.8 (the weight ratio of D-PEI/DNA ¼ 4) in water (mean ꢁ s.d., n ¼ 3). (c) TEM image of the magnetic complexes. (d) Agarose gel
electrophoresis of PEI 25k and the ternary magnetic complexes at MNP-to-DNA weight ratios of 0.4, 0.6, and 0.8.

D
H. Yu et al.
60
50
40
30
20
10
0
PEI/DNA
DS-PEI/DNA
DT-PEI/DNA
Whitout magnetic
With magnetic
60
50
40
30
20
DM-PEI/DNA
10 min
30 min
1 h
4 h
10
0
Transfection time
PEI/DNA
2
3
4
5
Fig. 5. Transfection of MNP@DT-PEI/DNA (at a weight ratio of
0.6 : 4 : 1) and DT-PEI/DNA (at a weight ratio of 4 : 1) was performed under
magnetic field with different incubation times (10 min, 30 min, 1 h, 4 h) in
A549 cells (mean ꢁ s.d., n ¼ 3).
Weight ratio (w/w)
Fig. 4. Transfection of D-PEIs/DNA at weight ratios of 2, 3, 4 and 5 in
A549 cells (mean ꢁ s.d., n ¼ 3).
et al.[40] Though magnetofection exhibited nearly 38 % higher
gene expression with 30 min incubation, it showed a 3 %
increase only in gene expression using a standard incubation
time of 4 h. This result indicates that magnetofection of DT-PEI
cannot significantly enhance transfection efficiency. The reason
could be due to the fact that magnetic field itself cannot alter the
uptake mechanism. DT-PEI/DNA, but not the magnetic com-
plexes, cloud enter the nucleus and DT-PEI is crucial to
magnetofection for magnetic PEI complexes.[41]
MNP) were assessed in A549 cells. The plasmid of green
fluorescent protein (GFP) was used as a reporter gene. PEI
25 kDa was prepared at a weight ratio of 1.39 (N/P ¼ 10; N/P is
the ratio of moles of the amine groups of polymers to those of the
phosphate groups of DNA) as standard because of its high
transfection efficiency.[19] D-PEIs/DNA complexes were pre-
pared at different weight ratios including 2, 3, 4, and 5. The
results are shown in Fig. 4. All three polymers showed better
results when compared with PEI 25kDa under optimal weight
ratio condition. Using a weight ratio of 4, DT-PEI exhibited
the highest transfection efficiency, followed by DM-PEI and
DS-PEI, which is consistent with the amount of hydroxyl groups
in the polymer repeat unit. This result suggests that increasing
the hydroxyl number in the backbone of D-PEIs enhances gene
transfection. The introduction of hydroxyl groups in the back-
bone of D-PEIs may have an effect in improving the osmotic
property, and the osmotic coefficient increases with increasing
numbers of hydroxyls.[38] Furthermore, a hyperosmotic activity
was reported to significantly improve the transfection capability
by increasing the cellular uptake of the osmotically active gene
carriers.[39]
Optimization of the Amount of MNPs on Magnetofection
Sufficient amounts of MNPs are required in magnetic com-
plexes for responding to the external magnetic field. However,
an excess of MNPs would not enhance transfection.[40,43]
The amount of MNPs optimization was performed under a fixed
incubation time of 30 min with various amounts of MNPs.
As shown in Fig. 7, the transfection efficiency improved
by increasing the weight ratio (MNP/DNA) from 0.2 to
0.6 (fluorescent images of GFP expressed in A549 cells are
shown in Fig. S2 in the Supplementary Material). However,
further increases in the amount of MNPs led to a decrease in
transfection efficiency, which indicated that an MNP/DNA ratio
of 0.6 was optimal for the magnetofection of DT-PEI in A549
cells. This result indicates that low weight of MNPs is not
enough for responding to the external magnetic field, which
leads to the low efficiency. Moreover, an excess of MNPs would
increase the aggregation of complexes, resulting in the reduction
of transfection efficiency.[44]
Effect of Incubation Time on Magnetofection
The standard incubation time is approximately 4 h for PEI or
other commercial transfection regents in order to achieve high
transfection efficiency in vitro. However, a short period of gene
delivery time might be significant for applications in vivo.
Magnetofection is a way to efficiently shorten the time to
5 min to 2 h.[27,40,41] To optimize the incubation time, transfec-
tion of MNP@DT-PEI/DNA was performed under a magnetic
field with different incubation times at a fixed MNP-to-DNA
weight ratio of 0.6. The results are shown in Figs 5 and 6. The
DT-PEI/DNA complexes showed optimal incubation times over
1 h, while the MNP@DT-PEI/DNA complexes achieved maxi-
mum transfection efficiency at 30 min of incubation. The reason
could be attributed to the increasing sedimentation rates and
internalization of the magnetic complexes. However, 30 min
seems to be a little longer when compared with that displayed by
commercial reagents (i.e. 15 min).[42] The phenomenon is partly
attributed to the smaller diameters of the magnetic particles that
we prepared (,10 nm), which required a longer time to reach the
cell surface under a magnetic field. Furthermore, continuously
increasing the incubation time did not obviously promote
transfection, which is in accordance with the study by Xie
Cytotoxicity
The cytotoxicity of the MNPs, D-PEIs/DNA, and MNP@
D-PEIs/DNA were evaluated on A549 and COS-7 cells (Fig. 8).
MNPs exhibited a negligible cytotoxicity after 48 h; more spe-
cifically, over 80 % cell viability was obtained when the MNPs
concentration increased to 500 mg mLꢀ1 (Fig. 8d). Three of the
D-PEIs displayedlower cytotoxicity than 25 kDa PEI in both cell
lines (Fig. 8 a, b). Until now, many groups have indicated that the
introduction of hydroxyl groups into polycationic gene vectors
aided in the reduction of cytotoxicity.[45,46] Thus, D-PEIs
exhibited reduced cytotoxicity owing to an increased number of
hydroxyl groups, which improved the biocompatibility of the
carriers. In addition, the strong positive charge of PEI damages
cell membranes, but the positive charge of the PEI backbone in
D-PEI may be partially neutralized by electrostatically negative

Diglycidyl Esters Cross-Linked with Polyethyleneimine for Magnetofection
E
DT-PEI/DNA
MNP@DT-PEI/DNA
10 min
30 min
1 h
4 h
Fig. 6. GFP expression of MNP@DT-PEI/DNA and DT-PEI/DNA was performed under magnetic field with different incubation times (10 min,
30 min, 1h, 4h) in A549 cells (mean ꢁ s.d., n ¼ 3).
60
demonstrated low cytotoxicity under magnetofection conditions.
Theseresults indicatethat this type ofmagnetic nanoparticle may
be promising for gene delivery.
MNP@DT-PEI/DNA
50
40
30
20
10
0
Due to the fact that MNPs have been used in clinical settings
for more than a decade, we can foresee that these functional
targeted MNPs will be applicable in clinical gene therapy in the
future. To obtain an effective and low toxic transfection method
for gene delivery in vitro, future work in will mainly focus on the
development of multifunctional MNPs-vectors in vitro and in
vivo, which will make a remarkable difference in tracking and
disease diagnosis. In addition, the crude materials needed for the
synthesis ofMNPs-basedgene vectors shouldhavetheproperties
of biocompatibility, safety, and degradability in the organism.
0.2
0.4
0.6
0.8
MNP/DNA (w/w)
Fig. 7. Magnetofection efficiency of MNP@DT-PEI/DNA was performed
under varying MNP/DNA ratios of 0.2, 0.4, 0.6, and 0.8 with 30 min
incubation (DT-PEI/DNA ¼ 4) in A549 cells (mean ꢁ s.d., n ¼ 3).
Experimental
Materials
All chemicals and reagents were obtained commercially and
were used as received. Anhydrous ethanol was dried and
purified under nitrogen by using standard methods and was
distilled immediately before use. Polyethyleneimine (branched
PEI 25 kDa) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenylte-
trazolium bromide (MTT) were purchased from Sigma-Aldrich.
LMW PEI (branched, average molecular weight 600 Da, 99 %)
was purchased from Aladdin (Shanghai, China). Plasmid DNA
encoding a red-shifted variant of wild-type GFP was from
Aldevron, Inc. Dulbecco’s Modified Eagle’s Medium (DMEM),
penicillin, streptomycin, and fetal bovine serum were obtained
from GIBCO.
OH groups.[47] These results were attributed to the combination
of the degradable ester linkage and thenegativehydroxylgroups.
As shown in Fig. 8c, all three MNP@D-PEIs/DNA showed more
than 95 % cell viability in both cells under magnetofection
conditions (MNP/D-PEIs/DNA ¼ 0.6 : 4 : 1). These results
showed that these magnetic complexes did not possess obvious
toxicity because of the low cytotoxicity of MNPs and D-PEIs.
Conclusion
In this study, three polymers derived from PEI (Mw 600) linked
with diglycidyl esters were prepared to assemble with DNA,
then fabricated with biocompatible magnetic nanoparticles to
form ternary complexes by electrostatic interactions. The
magnetic ternary complexes showed good ability to condense
DNA into nanoparticles under transfection conditions. The sizes
and zeta potentials of the ternary complexes were suitable for
cell endocytosis. The results of transfection of the binary com-
plexes showed that the weight ratio of D-PEIs/DNA of 4 was the
optimized ratio. DT-PEI showed the highest transfection and
lowest cytotoxicity among the D-PEIs studied. This result
indicated that increasing the hydroxyl density in the backbone of
D-PEI can enhance gene expression and decrease cytotoxicity.
The magnetofection experiments of MNP@DT-PEI/DNA
showed that magnetofection can shorten incubation times when
compared with standard protocols and the weight ratio of MNP/
DNA of 0.6 was the optimized ratio. The MNPs exhibited
negligible cytotoxicity. The magnetic ternary complexes also
Preparation and Characterization of Diglycidyl Esters
Diglycidyl esters were prepared according to the reference.[48]
In a typical procedure, tartaric acid (0.10 mol) and allyl alcohol
(0.40 mol) were mixed and refluxed in the presence of 0.1 %
(v/v) sulfuric acid for 24 h. The reaction was neutralized with
sodium carbonate. Excess allyl alcohol was evaporated at 708C
under vacuum. The remaining solution was diluted with ethyl
acetate and dried with anhydrous Na2SO4. After filtration, the
crude product was collected by evaporation.
Diallyl tartaric acid (5 mmol) and m-chloroperoxybenzoic
acid (15 mmol, 85 % purity) were mixed in methylene chloride,
and refluxed at 558C for 8 h. After that, the by-product m-
chlorobenzoic acid was crystallized at ꢀ208C overnight and
filtered from the solution. The remaining co-products were
removed by filtration and the filtrate was poured into column

F
H. Yu et al.
Concentration [µg mLꢂ1
]
Concentration [µg mLꢂ1
]
(a)
(b)
120
100
80
60
40
20
0
120
100
80
60
40
20
0
5
10 15 25
50
5
10 15 25
50
PEI
DS-PEI DM-PEI DT-PEI
PEI
DS-PEI DM-PEI DT-PEI
A549
COS-7
100
80
60
40
20
0
(c)
(d)
104
A549
COS-7
102
100
98
96
94
92
90
88
86
84
82
10
20
50
100 200 500 1000
Concentration [µg mLꢂ1
]
Fig. 8. Cell viability: D-PEIs under various concentration on (a) A549 cells and (b) COS-7 cells; (c) magnetic complexes under
transfection conditions; and (d) MNPs with various concentrations (mean ꢁ s.d., n ¼ 3).
with ethyl acetate/petroleum ether (v/v 1 : 2) as eluent. These
compounds were identified by 1H NMR and 13C NMR.
(8000 g, 30 min) and washed with ethanol and water thrice. The
black precipitate was then dried at 608C for further use.
FT-IR spectroscopy (Nicolet iS10) was performed to analyze
the surface functionalization of the magnetic nanoparticles.
The morphology and size were characterized using TEM
(FEI TECNAI G2 20 LaB6).
Synthesis of D-PEIs
In a typical example,[19] PEI 600 (1.50 mmol), diglycidyl tar-
taric or its analogues (1.33 mmol) and anhydrous ethanol (3 mL)
were mixed in a 10-mL glass vial with a magnetic bar and sealed
with a screw cap. It was heated in an oil bath at 798C for 24 h with
stirring. The product was purified by dialysis against water for
48 h (molecular weight cut-off 3500 kDa) and isolated by
lyophilization. D-PEIs were characterized by 1H NMR and
GPC.
Assembly and Characterization of MNP@D-PEIs/DNA
DNA, D-PEIs, and MNPs were diluted in 4-(2-hydroxyethyl)-1-
piperazineethanesulfonic acid (HEPES)-buffered glucose
(HBG; 20 mM HEPES, 5 % w/v glucose, pH 7.4) for the prep-
aration of gene complexes. D-PEIs and DNA were mixed at a
w/w ratio of 4 and incubated for 15 min to form binary D-PEIs/
DNA complexes. The MNPs were added gently to the pre-
prepared complexes at different weight ratios (0.4, 0.6, or 0.8;
MNP/DNA) for 30min to form the ternary complexes of
MNP@D-PEIs/DNA. PEI 25kDa/DNA polyplexes were also
prepared at a w/w ratio of 1.39. The stability of DNA compaction
in the gene complexes was evaluated by electrophoresis on 1 %
(w/v) agarose gel in the presence of ethidium bromide (EB;
0.1 mg mLꢀ1) with a voltage of 120 V for 15min. The gel was
photographed using UV illuminator. The particle size and zeta
potential of the MNP@D-PEIs/DNA complexes were evaluated
on a Zetasizer 3000 HAS (Malvern Instrument, Inc., Worcester-
shire, UK). The size and morphology of the magnetic complexes
were observed by TEM (FEI TECNAI G2 20 LaB6, USA).
Preparation and Characterization of Magnetic
Nanoparticles
The MNPs were prepared according to the reference.[37] In brief,
anhydrous FeCl3 (2.0 mmol) was added to 2-(2-hydroxy-
ethoxy)-ethanol(DEG; 20 mL) with vigorous mechanical stirring
to form a clear solution. Then, Na3Cit (0.8 mmol) was added to
the mixture and heated at 808C with stirring to form a clear
solution. After dissolving anhydrous sodium acetate (NaOAc;
6.0 mmol) in the above solution, the mixture was transferred into
a 50-mL of Teflon-lined stainless steel autoclave and sealed in
air. After this, the autoclave was kept at 2408C for 6 h. After the
reaction, the black products were collected by centrifugation

Diglycidyl Esters Cross-Linked with Polyethyleneimine for Magnetofection
G
Cell Culture
[4] D. A. Jackson, S. Juranek, H. J. Lipps, Mol. Ther. 2006, 14, 613.
doi:10.1016/J.YMTHE.2006.03.026
COS-7 or A549 cells were incubated, respectively, in DMEM
and 1640 medium containing 10 % FBS, 100 mg mLꢀ1 strepto-
mycin, and 100 IU mLꢀ1 penicillin at 378C in an incubator with
a humid atmosphere of 5 % CO2.
[5] A. A. Eltoukhy, D. Chen, C. A. Alabi, R. Langer, D. G. Anderson, Adv.
Mater. 2013, 25, 1487. doi:10.1002/ADMA.201204346
[6] C. J. Bishop, T. M. Ketola, S. Y. Tzeng, J. C. Sunshine, A. Urtti, H.
Lemmetyinen, E. Vuorimaa-Laukkanen, M. Yliperttula, J. J. Green,
J. Am. Chem. Soc. 2013, 135, 6951. doi:10.1021/JA4002376
[7] P. J. C. Lin, Y. K. Tam, P. R. Cullis, Clin. Lipidol. 2014, 9, 317.
doi:10.2217/CLP.14.27
[8] H. O. McCarthy, J. McCaffrey, C. M. McCrudden, A. Zholobenko, A.
A. Ali, J. W. McBride, A. S. Massey, S. Pentlavalli, K.-H. Chen, G.
Cole, S. P. Loughran, N. J. Dunne, R. F. Donnelly, V. L. Kett, T.
Robson, J. Controlled Release 2014, 189, 141. doi:10.1016/J.JCON
REL.2014.06.048
[9] A. Kiselev, A. Egorova, A. Laukkanen, V. Baranov, A. Urtti, Int. J.
Pharm. 2013, 441, 736. doi:10.1016/J.IJPHARM.2012.10.020
[10] F. M. P. Tonelli, S. M. S. N. Lacerda, M. A. Silva, E. S. Avila, L. O.
Ladeira, L. R. Franca, R. R. Resende, RSC Adv. 2014, 4, 37985.
[11] K. Lee, M. H. Oh, M. S. Lee, Y. S. Nam, T. G. Park, J. H. Jeong, Int. J.
Pharm. 2013, 445, 196. doi:10.1016/J.IJPHARM.2013.01.014
[12] H. N. Yang, J. S. Park, S. Y. Jeon, W. Park, K. Na, K. H. Park,
Biomaterials 2014, 35, 8439. doi:10.1016/J.BIOMATERIALS.2014.
06.024
In Vitro Transfection
A549 cells were seeded at a density of 6 ꢂ 104 cells per well in a
24-well plate with 0.5 mL medium containing 10 % FBS and
incubated at 378C for 24 h. Then, the medium was replaced with
fresh serum-free medium, to which D-PEIs/DNA or MNP@D-
PEIs/DNA complexes were added to achieve a final DNA
concentration of 3 mg mLꢀ1. After various incubation times
(10 min, 30 min, 1 h, or 4 h), the magnetic field was moved away
and the medium was replaced by 900 mL fresh medium con-
taining serum, and the cells were further incubated for 24 h at
378C. The cells were directly observed and the microscopy
images were recorded on a fluorescence microscope (80i,
Nikon, Japan) at a magnification of 100ꢂ. The transfection
efficiency was quantified using Fluorescence Activated Cell
Shorter (FACS) (Becton Dickinson, San Jose, CA)
[13] R. K. C. Bahadur, B. Thapa, N. Bhattarai, Nanotechnol. Rev. 2014, 3,
269. doi:10.1515/NTREV-2013-0026
For magnetofection, an array of 24 neodymium-iron-boron
(Nd-Fe-B) permanent magnets (diameter ¼ 15 mm, height ¼ 3
mm; Shanghai YuHang Magnet Co., China) in the format of a
24-well plate that was placed under the cell culture plate to offer
the magnetic field.[32]
[14] H. Sun, X. Zhu, L. Zhang, X. Gu, J. Wang, J. Li, Y. Zhang, Biotechnol.
Bioprocess Eng. 2013, 18, 648. doi:10.1007/S12257-012-0720-Z
[15] S. Li, L. Huang, Gene Ther. 2000, 7, 31. doi:10.1038/SJ.GT.3301110
[16] M. Wang, J. D. Tucker, P. Lu, B. Wu, C. Cloer, Q. Lu, Bioconjugate
Chem. 2012, 23, 837. doi:10.1021/BC200674E
[17] Y. Wang, M. Zheng, F. Meng, J. Zhang, R. Peng, Z. Zhong, Bioma-
cromolecules 2011, 12, 1032. doi:10.1021/BM101364F
[18] L. Feng, A. Xie, X. Hu, Y. Liu, J. Zhang, S. Li, W. Dong, New J. Chem.
2014, 38, 5207.
[19] S. Li, Y. Wang, J. Zhang, W. H. Yang, Z. H. Dai, W. Zhu, X. Q. Yu,
Mol. BioSyst. 2011, 7, 1254. doi:10.1039/C0MB00339E
[20] H. Koo, G. W. Jin, H. Kang, Y. Lee, K. Nam, C. Z. Bai, J. S. Park,
Biomaterials 2010, 31, 988. doi:10.1016/J.BIOMATERIALS.2009.
10.004
[21] A. Ragusa, I. Garcia, S. Penades, IEEE Trans. Nanobioscience 2007, 6,
319. doi:10.1109/TNB.2007.908996
[22] A. S. Lubbe, C. Alexiou, C. Bergemann, J. Surg. Res. 2001, 95, 200.
doi:10.1006/JSRE.2000.6030
Cytotoxicity Assay
The cytotoxicity of each complex and MNPs was assessed by an
MTT assay on A549 and COS-7 cells. Cells were seeded in 96-
well plates at a density of 1 ꢂ 104 cells per well and cultured for
24 h. After that, the culture medium was replaced with 200 mL
medium containing the complexes. After 48 h of incubation, the
cells were washed with PBS once, and 50 mL of 1 ꢂ MTT buffer
was added to each well. After 4 h, the medium was removed, and
the formazan crystals were dissolved with 150 mL DMSO for
10 min on a shaker. The ultraviolet absorption of each well was
measured by a PowerWaveTM XS Microplate Reader (BioTek
Instruments, Inc.) at a wavelength of 570 nm. Cell viability ¼
(ODtreated/ODcontrol) ꢂ 100 % where OD is the optical density.
The results were expressed as mean values (ꢁ standard devia-
tion s.d.) of six repeats.
[23] R. Xing, G. Liu, J. Zhu, Y. Hou, X. Chen, Pharm. Res. 2013, 31, 1377.
[24] C. Plank, O. Zelphati, O. Mykhaylyk, Adv. Drug Delivery Rev. 2011,
63, 1300. doi:10.1016/J.ADDR.2011.08.002
[25] J. B. Sacha, D. I. Watkins, Nat. Protoc. 2010, 5, 239. doi:10.1038/
NPROT.2009.227
[26] D. Ang, C. Y. Tay, L. P. Tan, P. R. Preiser, R. V. Ramanujan, Mater.
Sci. Eng., C 2011, 31, 1445. doi:10.1016/J.MSEC.2011.05.014
[27] M. Arsianti, M. Lim, C. P. Marquis, R. Amal, Langmuir 2010, 26,
7314. doi:10.1021/LA9041919
[28] S. W. Kamau, P. O. Hassa, B. Steitz, A. Petri-Fink, H. Hofmann, M.
Hofmann-Amtenbrink, B. von Rechenberg, M. O. Hottiger, Nucleic
Acids Res. 2006, 34, e40. doi:10.1093/NAR/GKL035
Supplementary Material
1
The H NMR and 13C NMR data of diglycidyl esters, charac-
terization of D-PEIs, and GFP expression of MNP@DT-PEI/
DNA are available on the Journal’s website.
Acknowledgements
[29] Y. Zhou, Z. Tang, C. Shi, S. Shi, Z. Qian, S. Zhou, J. Mater. Sci.:
Mater. Med. 2012, 23, 2697. doi:10.1007/S10856-012-4720-5
[30] F. M. Kievit, O. Veiseh, N. Bhattarai, C. Fang, J. W. Gunn, D. Lee, R.
G. Ellenbogen, J. M. Olson, M. Zhang, Adv. Funct. Mater. 2009, 19,
2244. doi:10.1002/ADFM.200801844
[31] S. C. McBain, H. H. P. Yiu, A. El Haj, J. Dobson, J. Mater. Chem.
2007, 17, 2561. doi:10.1039/B617402G
[32] M. Lattuada, T. A. Hatton, Langmuir 2007, 23, 2158. doi:10.1021/
LA062092X
[33] C. Mi, J. Zhang, H. Gao, X. Wu, M. Wang, Y. Wu, Y. Di, Z. Xu,
C. Mao, S. Xu, Nanoscale 2010, 2, 1141. doi:10.1039/C0NR00102C
[34] S. Prijic, L. Prosen, M. Cemazar, J. Scancar, R. Romih, J. Lavrencak,
V. B. Bregar, A. Coer, M. Krzan, A. Znidarsic, G. Sersa, Biomaterials
2012, 33, 4379. doi:10.1016/J.BIOMATERIALS.2012.02.061
We are thankful for financial support from the National Natural Science
Foundation of China (Grant No. 30870625).
References
[1] M. Everts, V. Saini, J. L. Leddon, R. J. Kok, M. Stoff-Khalili, M. A.
Preuss, C. L. Millican, G. Perkins, J. M. Brown, H. Bagaria, D. E.
Nikles, D. T. Johnson, V. P. Zharov, D. T. Curiel, Nano Lett. 2006, 6,
587. doi:10.1021/NL0500555
[2] W. T. Godbey, A. G. Mikos, J. Controlled Release 2001, 72, 115.
doi:10.1016/S0168-3659(01)00267-X
[3] M. A. Mintzer, E. E. Simanek, Chem. Rev. 2009, 109, 259.
doi:10.1021/CR800409E

H
H. Yu et al.
[35] M. Arsianti, M. Lim, C. P. Marquis, R. Amal, Biomacromolecules
2010, 11, 2521. doi:10.1021/BM100748P
[42] J. Bertram, Curr. Pharm. Biotechnol. 2006, 7, 277. doi:10.2174/
138920106777950825
[36] F. N. Al-Deen, C. Selomulya, T. Williams, Colloids Surf., B 2013, 102,
492. doi:10.1016/J.COLSURFB.2012.09.026
[37] L. H. Shen, J. F. Bao, D. Wang, Y. X. Wang, Z. W. Chen, L. Ren, X.
Zhou, X. B. Ke, M. Chen, A. Q. Yang, Nanoscale 2013, 5, 2133.
doi:10.1039/C2NR33840H
[43] X. Pan, J. Guan, J. W. Yoo, A. J. Epstein, L. J. Lee, R. J. Lee, Int. J.
Pharm. 2008, 358, 263. doi:10.1016/J.IJPHARM.2008.02.020
[44] S. Ota, Y. Takahashi, A. Tomitaka, T. Yamada, D. Kami, M.
Watanabe, Y. Takemura, J. Nanopart. Res. 2013, 15, 1653.
doi:10.1007/S11051-013-1653-Y
[38] R. A. Robinson, P. K. Smith, E. R. B. Smith, Trans. Faraday Soc. 1942,
[45] F.-J. Xu, H. Li, J. Li, Z. Zhang, E.-T. Kang, K.-G. Neoh, Biomaterials
2008, 29, 3023. doi:10.1016/J.BIOMATERIALS.2008.03.041
[46] M. Ma, F. Li, Z.-F. Yuan, R.-X. Zhuo, Acta Biomater. 2010, 6, 2658.
doi:10.1016/J.ACTBIO.2010.01.024
[47] W.-S. Lee, Y.-K. Kim, Q. Zhang, T.-E. Park, S.-K. Kang, D.-W. Kim,
C.-S. Cho, Y.-J. Choi, Nanomedicine 2014, 10, 525. doi:10.1016/
J.NANO.2013.10.005
38, 63. doi:10.1039/TF9423800063
[39] M. A. Islam, C. H. Yun, Y. J. Choi, J. Y. Shin, R. Arote, H. L. Jiang, S.
K. Kang, J. W. Nah, I. K. Park, M. H. Cho, C. S. Cho, Biomaterials
2011, 32, 9908. doi:10.1016/J.BIOMATERIALS.2011.09.013
[40] L. Xie, W. Jiang, Y. Nie, Y. He, Q. Jiang, F. Lan, Y. Wu, Z. Gu, RSC
Adv. 2013, 3, 23571. doi:10.1039/C3RA43588A
[41] S. Huth, J. Lausier, S. W. Gersting, C. Rudolph, C. Plank, U. Welsch, J.
Rosenecker, J. Gene Med. 2004, 6, 923. doi:10.1002/JGM.577
[48] C. B. Gumera, Y. Wang, Adv. Mater. 2007, 19, 4404. doi:10.1002/
ADMA.200701747