OEI800 polyconjugates linked with ketalized glycolic acid for use as gene vectors

Article abstract of DOI:10.1039/c1jm12849c

Endosomal escape of DNA polyplexes is one prominent bottleneck involved in the transfection process. Purposely against the low pH level in the endosome compartment, a series of acid-cleavable gene vectors constructed from oligoethyleneimine OEI800 polycon

Full text of DOI:10.1039/c1jm12849c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2011, 21, 15305
www.rsc.org/materials
PAPER
OEI800 polyconjugates linked with ketalized glycolic acid for use as gene
vectors†
*
Xiao-hua Luo, Chen-Wei Liu, Ze-yong Li, Si-yong Qin, Jun Feng, Xian-zheng Zhang and Ren-xi Zhuo
Received 20th June 2011, Accepted 19th July 2011
DOI: 10.1039/c1jm12849c
Endosomal escape of DNA polyplexes is one prominent bottleneck involved in the transfection process.
Purposely against the low pH level in the endosome compartment, a series of acid-cleavable gene
vectors constructed from oligoethyleneimine OEI800 polyconjugates linked with ketalized glycolic acid
were designed herein and termed OEI-GKs. Their potential as gene vectors was comparatively
evaluated by investigating the properties including DNA binding ability, polyplexes zeta potential,
particle size, acid-triggered degradation, buffer capability before and after degradation, cytotoxicity,
and transfection efficiency. The resultant data indicate that the transfection efficiency and cell-
biocompatibility are dependent on the polymer architecture and molecular weights, which can be
tailored by adjusting the charge ratio of OEI800 versus the linking agent. OEI-GK(1 : 1) can be
potentially developed as efficient vectors for the gene delivery in terms of their transfection activity even
higher than PEI25k as well as the negligible cytotoxicity. Those improved properties are believed to
have association with ketal-associated degradation of OEI-GK under acid conditions in the endosome,
which lead to not only easy unpacking of DNA from the hydrolyzed polyplexes but also, interestingly,
substantially enhanced buffer capability.
ester or disulfide cross-linked OEI polyconjugates could mediate
modest transfection whilst holding improved cytotoxicity.8–10
Introduction
As one important kind of promising gene vector, polycations
Furthermore, it is expected that those polyconjugates would
have attracted great attention in recent years due to many
ultimately break down into OEIs and be readily excreted from
advantages including facile preparation, large payload, tolerance
the body, leading to a long-term safety profile.
towards gene sizes and low immunogenic risk as compared to
The gene transfection process consists of several biological
viral vectors.1,2 Although steady developments have been ach-
steps including extracellular and intracellular transfer, vector
ieved in this field, polycation vectors continue to be optimized to
unpacking as well as the final gene expression in the nucleus.
acquire promoted transfection activity and reduced cytotoxicity.
After endocytosis, endosomal escape and following dissociation
Up to date, the commercially available branched poly-
of nucleic acids from vectors have been recently disclosed to be
ethyleneimine (PEI) with a molar mass of 25 kDa (PEI25k) is still
the crucial efficiency-determining steps.11,12 As is well-estab-
regarded as one of the most effective non-viral vectors, and often
lished, polycations could provide tight package and efficient
acts as a ‘‘golden standard’’ to evaluate other newly designed
protection of DNA from undesired interaction, and mediate its
polymeric vectors.3,4 Unfortunately, its rather high cytotoxicity
uptake into the cells via electrostatic interaction.13 Nevertheless,
hampered the application in human gene therapy trials. In
such a strong interaction may be unfavorably a limiting factor
contrast to it, low molecular weight oligoethyleneimines (OEI)
for the intracellular transfer process. It would lead to difficult
such as OEI800 (Mw ¼ 800 Da) only display poor transfection
DNA detachment from vectors, inhibitory access by the tran-
activity despite their negligible cytotoxicity.5–7 Against this catch-
scription machinery and consequently low transfection effi-
22 problem, one reasonable resolution has been proposed by
ciency.14 At this point, a satisfactory gene delivery system may
coupling short PEI chains into one entity using biodegradable
preferably be constructed via the realization of multifunction
linkers. This strategy has been partly validated by studies where
that maintains extracellular stability of polymer/DNA poly-
plexes whilst allowing the endosomal escape and intracellular
disassembly of polyplexes with relatively little difficulty, in order
for the DNA to be released into the cytoplasmic and nuclear sites
and become biologically active.4
Key Laboratory of Biomedical Polymers (The Ministry of Education),
Department of Chemistry, Wuhan University, Wuhan, 430072, China.
E-mail: fengjun@whu.edu.cn; Fax: +86 27 6875 4509; Tel: +86 27 6875
4509
A future aim for gene delivery will be the development of
‘‘smart’’ polyplexes termed artificial viruses.15 Those polyplexes
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c1jm12849c
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15305

View Article Online
could sense the biological environments and subsequently
undergo programmed changes, which would make them more
compatible at each step during the transfection.16 In compar-
ison to the extracellular milieus, the endosomal environment is
quite different, where the pH is from about 7.4 in the early
period to 5.0 in the late stage.11 According to such specific pH
gradients, many efforts have been devoted to the design of
pH-dependent gene vectors, which can generally undergo pH-
responsive changes regarding the charge, phase and confor-
mation to provide a membrane-disruptive driving force, for
instance the ‘‘proton sponge’’ effect.17–20 The proton sponge
nature of PEI is thought to cause osmotic swelling and physical
disruption of the endosome, thus resulting in the escape of the
vector from the degradable lysosomal trafficking pathway.20
Also, such an effect can possibly destroy the pH gradient that
in turn disturbs endosome maturation and function. We
hypothesize that if OEI polyconjugates decompose into OEI
fragments triggered by the local acid environment in the
endosome, the membrane-disruptive capability might be
enhanced due to the enhanced water-solubility after degrada-
tion as well as the colligative feature of the osmotic pressure,
possibly leading to a strengthened ‘‘proton sponge’’ effect.
Furthermore, the decomposition of those polyconjugates would
somewhat lessen the interaction between the vector and DNA,
which would help free the condensed DNA for the following
transfer process. Such a strategy may also permit the gene
delivery system with high gene expression in solid tumor tissues
since the local pH levels are lower than those in blood or other
normal tissues.21
Scheme 1 Illustration of the syntheses of OEI-GKs.
Experimental section
Materials
Branched PEI25k and PEI800 were purchased from Sigma-
Aldrich. Glycolic acid, benzyl bromide, 2-methoxypropene, N,N-
dimethylformamide (DMF) and petroleum ether were obtained
from Shanghai Chemical Reagent Co. (China). DMF was dried
under CaH2 before use. QIAfilterꢀ plasmid purification Giga Kit
(5) was purchased from Qiagen (Hilden, Germany) and GelRedꢀ
was purchased from Biotium (CA, USA). Fetal bovine serum
(FBS), Dulbecco’s Modified Eagle’s Medium (DMEM), Hoechst
33258 and Dulbecco’s phosphate buffered saline (PBS), penicillin-
streptomycin, trypsin, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphe-
nyltetrazolium bromide (MTT), were purchased from Invitrogen
Corp. A Micro BCA protein assay kit was obtained from Pierce.
Other reagents were analytical grade and used as received.
Given those considerations, the design of acid-cleavable OEI
polyconjugates might be attractive for the purpose of improved
transfection performance. Nevertheless, the relevant researches
are rarely reported.22–24 Wagner’s group has recently done pio-
neering attempts by introducing acetal or ketal groups into the
linking reagents.22 Those acid-sensitively degradable poly-
conjugates showed enhanced transfection efficacy and biocom-
patibility. Note that in addition to the acid-cleavable groups,
those designed linkers contained ester or maleimido groups
which are subject to serious aminolysis reaction with PEI.3,25,26
Hence, low preparation (below 25 ꢀC) and storage temperatures
(ꢁ80 ꢀC) together with a short reaction time (45 min) were
required in order to abate the side reactions to a certain extent.
Even so, the side reaction eventually led to a large molecular
weight distribution of the obtained polyconjugates (Mw/Mn > 4),
incomplete polyconjugate degradation, as well as difficulties in
yielding compositionally and architecturally consistent products.
It may also raise confusion about the function of endosome-
specific degradation because the improved properties appeared
to have strong association with the significant amide linkage
originating from the side reaction.
Synthesis of 2-hydroxylbenzyl acetate
2-Hydroxylbenzyl acetate was prepared according to the refer-
ence.27 10 g of Cs2CO3 dispersed in 40 mL H2O was dropped into
a mixed solvent of MeOH and H2O (20 : 3, v/v, 115 mL) con-
taining 5 g of glycolic acid. The solution was then neutralized.
After stirring for 10 min at room temperature, the solvent was
thoroughly removed under reduced pressure and the residual was
dissolved in 150 mL DMF. Benzyl bromide (8 mL, 67.5 mmol)
was added into that solution slowly. After stirring at room
temperature for 24 h, the reaction was quenched by saturated
NaCl solution. The solution was extracted with EtOAc (3 ꢂ
40 mL). After drying over MgSO4 followed by concentration, the
product was purified by column chromatography (EtOAc :
petroleum ether ¼ 1 : 4).
In the current study, a novel polycation vector composed of
OEI800 units linked via an acid-cleavable ketal linkage was
fabricated but excluding any amine-labile groups. By design, the
linker reagent is derived from glycolic acid (Scheme 1), which can
be metabolized in vivo and has been widely utilized to prepare
biomedical materials such as poly(glycolic acid) (PGA) and poly
(lactic-co-glycolic acid) (PLGA). The prepared polyconjugates
termed OEI-GKs are expected to ultimately degrade into
OEI800 and GA in vivo, resulting in high biocompatibility.
Synthesis of benzyl-2,20-(propane-2,2-diylbis(oxy)) diacetate
(GK)
According an amended method,28 using an ice bath, 2-
methoxypropene (1.015 mL, 20 mmol) was slowly dropped to
15306 | J. Mater. Chem., 2011, 21, 15305–15315
This journal is ª The Royal Society of Chemistry 2011

View Article Online
2-hydroxylbenzyl acetate (3.5192 g, 20 mmol), which was firstly
suspended in 80 mL of dry petroleum ether containing p-tol-
uenesulfonic acid (38 mg, 0.4 mmol). After stirring for 2 h,
another 20 mmol of 2-methoxypropene was added. The reac-
tion process was monitored by thin layer chromatography
(TLC). Upon completion of the reaction, 6 mL of triethylamine
was added to neutralize the p-toluenesulfonic acid and the
product was purified by column chromatography (EtOAc :
petroleum ether ¼ 1 : 6).
(w/v) agarose gel which was put in Tris-acetate (TAE) running
buffer at 80 V for 1 h. Finally the complexes were observed by
a UV lamp with a Vilber Lourmat imaging system (France).
Amplification and purification of plasmid DNA
pGL-3 was used in the experiment. It was transformed as the
luciferase reporter genes in Escherichia coli JM109 and amplified
ꢀ
in the terrific broth media by acid–base titration at 37 C over-
night. The amplified plasmid was first purified by an EndoFree
QiAfilterꢀ Plasmid Giga Kit (5) then dissolved in TE buffer
Syntheses of OEI-GKs
ꢀ
solution and stored at ꢁ20 C. Finally the purity and concen-
An appropriate amount of benzyl-2,20-(propane-2,2-diylbis
(oxy)) diacetate was dissolved in DMF at a concentration of
0.1 g mLꢁ1. One molar equivalent of OEI800 (1.6 g mLꢁ1) and
two equiv. of 2-hydroxylpyridine were dissolved in DMF. Under
stirring at 50 ꢀC, the former solution was slowly dropped into the
OEI800 solution. After standing for 48 h, the final product OEI-
GK(1 : 1) was collected as yellow oil via the precipitation method
using petroleum ether as the poor solvent. Likewise, OEI-GK
(1 : 2) and OEI-GK(2 : 1) were obtained at different charge
ratios in feed. The overall yield was around 80% for all the
reactions.
tration of plasmids was determined by ultraviolet (UV) absor-
bance at 260–280 nm.
In vitro transfection of luciferase assay
Transfection of pGL-3 plasmids mediated by OEI-GKs in
COS7 and HeLa cells was studied and PEI25k was used as the
positive control. Cells were seeded at the density of 6 ꢂ 104
cells/well in the 24-well plate and incubated at 37 ꢀC for 24 h
respectively. The complexes at various w/w ratios ranging from
5 to 30 were prepared by adding 1 mg of pGL-3 DNA to an
appropriate volume of OEI-GK solution (in 150 mM NaCl
solution). Then the polyplexes solution were vortexed for 30 s
Characterization of OEI-GKs
ꢀ
before being incubated at 37 C for 0.5 h and then diluted to 1
Fourier transformed infrared (FTIR) spectra were recorded on
a Perkin Elmer-2 spectrophotometer. Prior to the measurement,
the sample was pressed into potassium bromide (KBr) pellets. 1H
nuclear magnetic resonance (1H NMR) spectra were recorded on
a Varian Mercury-VX 300 spectrometer using D2O as solvent.
Size exclusion chromatography and multi-angle laser light scat-
tering (SEC-MALLS) analysis was utilized to determine the
molecular weights of the polymers. A dual detector system,
consisting of a MALLS device (DAWN EOS, Wyatt Tech-
nology) and an interferometric refractometer (a differential
refractive index detector) (Optilab DSP, Wyatt Technology) was
used. The polymer concentration was 10 mg mLꢁ1. DMF was
used as the eluent for OEI-GK and ultrapure water was used for
OEI-GKs at a flow rate of 0.3 mL minꢁ1. The column tempera-
ture was fixed at 25 ꢀC and the MALLS detector was operated at
a laser wavelength of 690 nm.
mL by serum-free DMEM before transfection. After being
incubated for 4 h at 37 ꢀC, the serum-free DMEM was replaced
by ꢀfresh DMEM containing 10% FBS and further incubated at
37 C for 44 h. The luciferase assay was followed according to
manufacture’s protocols. Relative light units (RLU) were
measured by chemiluminometer (Lumat LB9507, EG&G
Berthold, Germany). The total protein was determined
according to the BCA protein assay kit (Pierce). Luciferase
activity was expressed as RLU/mg Protein. Data were shown as
mean ꢃ standard deviation (SD) based on three independent
measurements.
Confocal laser scanning microscope
To study cellular uptake of the complexes, fluorescent confocal
microscopic images of complexes in HeLa cells were further
obtained. HeLa cells were seeded at the density of 1 ꢂ 105 cells/
well in the 24-well plate with 1 mL of DMEM containing 10%
Cell culture
ꢀ
FBS and incubated at 37 C for 24 h. 1 mg of pGL-3 was fluo-
African Green Monkey SV40-transfected kidney fibroblast cell
line (COS7) and human cervix carcinoma (HeLa) cells were
incubated with DMEM containing 1% antibiotics (penicillin-
streptomycin, 10 000 U mLꢁ1) and 10% FBS at 37 ꢀC in
a humidified atmosphere containing 5% CO2.
rescently labeled with 5 mL (10 ꢂ 10ꢁ6 M) YOYO-1 iodide at
ꢀ
37 C for 15 min before the complexes were added and further
incubated for 15 min. Then 900 mL of DMEM containing 10%
FBS was added and incubated at 37 ꢀC for 4 h. After that, cells in
each well were washed five times with PBS and 40 mL Hoechst
(0.2 mM in DMEM containing 10% FBS) was then added and
incubated at 37 ꢀC for 0.5 h. Then the cells were photographed by
a confocal laser scanning microscope (Nikon C1-si TE2000,
Japan) and finally recorded by EZ-C1 software.
Agarose gel retardation assay
OEI-GKs/pDNA complexes at different w/w ratios were
prepared by adding an appropriate volume of OEI-GK solution
(in 150 mM NaCl solution) to 0.1 mg of pGL-3. The obtained
complex solution was diluted by 150 mM NaCl solution to the
total volume of 8 mL and vortexed for 30 s before being incubated
at 37 ꢀC for 0.5 h. Then 1 mL GelRedꢀ was added to the
complexes respectively prior to be electrophoresed on the 0.7%
Cytotoxicity assay
The cell toxicity of OEI-GKs was examined by MTT assay. The
COS7 cells and HeLa cells were seeded in the 96-well plate with
a density of 6000 cells/well and cultured 24 h in 100 mL DMEM
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15307

View Article Online
containing 10% FBS respectively. After incubating the polymers
for 48 h, 20 mL of MTT (5 mg mLꢁ1) solution was added in and
further incubated at 37 ꢀC for 4 h. Then the medium was
removed and 150 mL DMSO was added in. A microplate reader
(BIO-RAD 550) was used to measure the absorbance of the
medium at 570 nm. The relative cell viability was calculated as
follows:
conditions in the endosome. The linking reagent benzyl-2,20-
(propane-2,2-diylbis(oxy)) diacetate (GK) was specifically
designed (Scheme 1) starting from glycolic acid, providing long-
term safety in vivo. GK is constructed around a centrally
located ketal bond, and can react with primary amines to form
covalently stable linkages. Those molecule designs avoid the
possible side-reactions and, thus, a useful model can be estab-
lished to exclusively investigate the relationship between
endosome-associated polymer degradation and gene trans-
fection efficacy.
À
Á
OD570;sample ꢁ OD570;background
À
Á
Relative cell viability ð%Þ ¼
OD570;control ꢁ OD570;background
ꢂ 100%
By adjusting the charge ratio of OEI800 versus GK, three OEI-
GKs were prepared in DMF (Table 1). As reported, the linking
reaction of OEIs would sometimes lead to the formation of
insoluble gels.29 In the present preparation, by slowly introducing
the linking reagent into the OEI800 solution, the reaction
mixture remained transparent throughout the whole process,
even when [GK] : [OEI800] approached 2 : 1. All the products
were soluble in water.
where OD570,sample was obtained in the presence of polymers,
OD570,control was obtained in the absence of polymers and
OD570,background was obtained in the absence of polymers and
cells. Data were shown as mean ꢃ standard deviation (SD) based
on triplicate independent experiments.
Degradation test
The chemical structures of OEI-GKs were identified by IR†
and 1H NMR analyses (Fig. 1). All the characteristic resonances
from the GK linker and OEI800 moieties are clearly detectable.
In the IR spectra, a strong absorbance of carbonyl groups (nC]
O) appears at 1662 cmꢁ1 and the peak at 1386 cmꢁ1 is attributed
to the stretching frequency of amide C–N (nC–N) (Fig. S1†). As
For the degradation study, polymers were dissolved in solution
with different pH values at 5.5 and 7.2 for a period of time. The
biodegradation was evaluated by monitoring the DNA binding
ability of the degraded polymers using an agarose gel retardation
assay. In the buffer capability test, the samples were incubated
for 8h before acid–base titration.
1
shown in the H NMR spectrum of OEI-GK(1 : 1) (Fig. 1B),
besides the characteristic signals in the region d 2.49–2.61 ppm
(signal e) belonging to OEI800, two separate single peaks at
d 3.90 ppm (signal b) and at d1.30 ppm (signal a) were clearly
detectable, which should be attributed to the –COCH2– and
–CH3– in GK, respectively. Moreover, it is observed that the
signals from the benzyl group of GK were almost lost after the
linking reaction, suggesting that the reaction was near comple-
tion under the given conditions. By comparing the area integrals
between signal b and a, it could be concluded that nearly no or
insignificant hydrolysis of the ketal groups took place
throughout the preparation process. Taken together, the spec-
troscopic analyses strongly demonstrate that by our method,
OEI-GKs have been successfully prepared.
Acid–base titration
The buffer capability of samples was evaluated by acid–base
titration assay over the pH range from 10.0 to 2.50. Briefly,
0.2 mg mLꢁ1 of each sample solution was prepared in 30 mL of
150 mM NaCl solution before titration. Firstly the sample
solution was titrated by 0.1 M NaOH to the pH of 10.0. Then
0.1 M HCl was gradually dropped into the solution to obtain
mixtures with different pH values which were measured by
a microprocessor pH meter simultaneously until the pH value
decreased to 2.50. Data were shown as mean ꢃ standard devia-
tion (SD) based on three independent measurements.
The molecular weights of OEI-GK products purified by the
precipitation method were determined by SEC-MALLS and the
resultant data were collected in Table 1. For the poly-
condensation between OEI and GK, the molar feed ratio would
affect not only the molecular weights but also the polymer
architectures, which are well known to play an important role in
the transfection process.1,2 Provided that OEI800 can be taken
for a special repeat unit, according to the feed ratio, OEI-GK
(1 : 2) could be viewed as a polyconjugate adopting the relatively
more branched structure due to the presence of excessive NH2
groups in PEI.30 By contrast, we can approximately look on OEI-
GK(1 : 1) and OEI-GK(2 : 1) as linear alternative copolymers
composed of OEI800 and GK units with different chain lengths.
Particle size and zeta potential measurement
Particle size and zeta potential were evaluated by dynamic light
scattering (DLS) at 25 ꢀC by Nano-ZS ZEN3600 (Malvern
Instruments). The complexes at various w/w ratios ranging from
5 to 30 were prepared by adding 1 mg of pGL-3 DNA to an
appropriate volume of OEI-GK solution (in 150 mM NaCl
solution). Then the complexe solutions were vortexed for 30 s
ꢀ
before being incubated at 37 C for 0.5 h and then diluted to 1
mL by 150 mM NaCl solution prior to being measured. Data
were shown as mean ꢃ standard deviation (SD) based on three
independent measurements.
Results and discussion
In vitro cytotoxicity
Syntheses and characterization of OEI-GKs
The biocompatibility is one critical prerequisite of gene vectors
for in vivo application. Although high molecular weight PEI such
as PEI25k displays high transfection efficiency towards most cell
lines, its clinical application appears to hardly be realized unless
the toxicity issues can be addressed.
In the present study, ketalized OEI polyconjugate vectors
termed OEI-GKs were prepared (Scheme 1), aiming at efficient
cytoplasmic localization and relatively easy disassociation of
the polyplexes through the hydrolysis of ketal bonds under acid
15308 | J. Mater. Chem., 2011, 21, 15305–15315
This journal is ª The Royal Society of Chemistry 2011

View Article Online
Table 1 Reaction condition and results of OEI-GK preparations
Feed molar
ratio (OEI/GK)
b
Sample
Composition (OEI/GK)a
Mw/g molꢁ1
PDI (Mw/Mn)b
OEI-GK(1 : 1)
OEI-GK(1 : 2)
OEI-GK(2 : 1)
1 : 1
1 : 2
2 : 1
1 : 1
1 : 2.1
1.3 : 1
21 300
30 500
9400
1.10
1.02
1.47
a
Data were determined by 1H NMR analyses. b Data were obtained from SEC-MALLS analyses.
those with PEI25k. Specifically, no IC50 can be detected
throughout the OEI-GK(1 : 2) working concentrations up to
800 mg mLꢁ1. In sharp contrast to this, 80% cells have died at
a concentration as small as 50 mg mLꢁ1 in the case of PEI25k.
Such a marked deviation in cell-biocompatibility between OEI-
GK(1 : 2) and PEI25k, despite their similar structures and the
higher Mw of OEI-GK(1 : 2), may be attributed to the lower
amine content of the former as well as its biodegradation into
smaller fragments.7 The above results clearly indicate that OEI-
GKs designed in the present study are much safer than PEI25k
for in vivo application, and can be used in a broad concentra-
tion range.
In addition to the acute toxicity aspect, the long-term fate of
the polymeric carrier has to be considered in an organism.
Therefore, implanted biomaterials, if they can be degradable
in vivo, would be more advantageous. OEI-GK vectors are
expected to ultimately degrade in vivo into low toxicity compo-
nents, providing a reasonably long-term safety profile.
Fig. 1 1H NMR spectra of GK linker and OEI-GK(1 : 1).
Many studies have revealed that the free polycations rather
than the bonded polycations are indeed the major cause of the
acute toxicity of the polycation/DNA gene transfer system.31
In order to preliminarily evaluate the safety profiles of OEI-
GKs, in vitro cytotoxicity tests were conducted in COS7 and
HeLa cells by MTT assay using PEI25k and OEI800 as the
control. Fig. 2 shows that OEI-GKs exhibited a remarkably
lower cytotoxicity than PEI25k in two different cell lines. For
one specified cationic polymer, the cytotoxicity is suggested to
be usually proportional to the molecular weights.6,7 Thus the
highest cell-biocompatibility of OEI-GK(2 : 1) can be under-
stood as a consequence of its lowest molecular weight. Both
OEI-GK(2 : 1) and OEI-GK(1 : 1) displayed better or compa-
rable cell-biocompatibility compared to OEI800. Relatively,
OEI-GK(1 : 2) presented slightly higher cytotoxicity. Due to
the relatively more branched and compact architecture, OEI-
GK(1 : 2) may possess a larger volume charge density. In
comparison, linear OEI-GK(1 : 1) adopts a relatively flexible
conformation and the positive charges are more dispersed.
Possibly owing to such structural difference, OEI-GK(1 : 2)
displayed a higher cytotoxicity despite the relatively lower
amine contents. Even so, the viabilities of both cells after co-
incubation with OEI-GK(1 : 2) were dramatically higher than
Fig. 2 Cytotoxicity of OEI-GKs in COS7 cells (top) and HeLa cells
(bottom) in comparison to that of PEI25k and OEI800. Data were shown
as mean ꢃ SD (n ¼ 3).
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15309

View Article Online
DNA binding assay
different w/w ratios and the pGL-3 gene was used as the
reporter gene.
Binding with DNA is one of the most important properties for
cationic vectors to provide an effective payload and protection of
DNA from DNase digestion. An agarose gel retardation assay
was performed to investigate whether the synthesized OEI-GKs
were able to condense pGL-3 plasmids. As shown in Fig. 3, the
mobility of the pGL-3 bands was gradually retarded as the w/w
ratio of OEI-GKs/DNA increased. The lowest retarding ratio,
where the mobility of pGL-3 could be entirely retarded, was used
as the indicator to evaluate the binding ability of the investigated
polymers. It is found that the binding ability is dependent on the
composition of the OEI-GKs. Amongst the three vectors, OEI-
GK(1 : 1) displayed the significantly strongest binding strength
as reflected by the rather low retarding ratio at 0.8. OEI-GK
(2 : 1) in the case of the ratio at 30 could barely inhibit pGL-3
migration in agarose gel electrophoresis. The capability of OEI-
GK(1 : 2) to integrate with plasmid DNA fell between those two.
The higher amine content of OEI-GK(1 : 1) may be a main
factor to account for its higher DNA binding ability as
compared to the case of OEI-GK(1 : 2). As for OEI-GK(2 : 1),
the poor binding ability should be attributed to its lower
molecular weight. In addition, the architecture of the obtained
polyconjugates may somewhat affect the binding between DNA
and polymers. Relatively, linear OEI-GK(1 : 1) is more flexible
and thus more OEI moieties per OEI-GK molecule could
readily access the DNA as compared to the hyperbranched
OEI-GK(1 : 2), which allows compact condensation at a lower
retarding ratio.
As seen from Fig. 4, the particle sizes of OEI-GK(1 : 1)
polyplexes seemed unchanged irrespective of the measured w/w
ratios ranging from 5 to 50. In comparison, after the ratio
reached about 20 for OEI-GK(1 : 2) and 30 for OEI-GK(2 : 1),
the size of the polyplexes tends to level off. Generally, those
results correlate well with the condensation efficiency of OEI-
GKs revealed by the DNA binding assay. When the negatively
charged DNAs were almost completely complexed and
condensed by the positively charged OEI-GKs, further increase
in the w/w ratio would exert an insignificant effect on the particle
size. Relatively, OEI-GK(1 : 1)/DNA polyplexes presented
a smaller particle size than the other two polyplexes. This can be
explained as the consequence of its stronger DNA binding
ability. The polyplexes formed from OEI-GK(2 : 1) may be of
a loose structure due to the weak vector/DNA interaction,
leading to the largest particle size among three investigated
polyplexes. As for branched OEI-GK(1 : 2) polyplexes, in addi-
tion to the effect of binding ability, the strong steric hindrance
might also confer the relatively larger particle size.
All of the polyplexes measured had particle sizes above
350 nm. Although nano-sized polyplexes at around 100ꢄ150 nm
are assumed to be advantageous for cellular uptake through non-
receptor-mediated endocytosis, this opinion is still debatable.32
The opposing results have shown that some larger particles could
mediate a higher transfection efficiency than smaller particles.33
From another viewpoint, the large particle size of the OEI-GK
polyplexes, to some extent, implies the relatively loose structure
within the polyplexes. Properly loosening the polyplex structure
may be beneficial for the improved gene transfection since it
would facilitate the easy approach of transcription-factors to the
swollen polyplexes. Recent studies have also revealed that in
both polyplexes and liposomes, the plasmid DNA needs to be
freely accessible for cell membrane interaction.7,14 Those repor-
ted findings suggest that the particle size may not be the primary
factor determining the transfection efficacy.
Particle size of OEI-GK polyplexes
To mediate endocytosis through a cell membrane, cationic
polymers need to condense DNA into compact particles. The
particle size of polycation/DNA polyplexes is assumed to be an
important factor for the achievement of efficient transfection.32
In the current study, the dynamic light scattering technique
(DLS) was utilized to measure the size of the polyplexes at
Zeta potential of OEI-GK/DNA polyplexes
Effective gene transfection mediated by polycation vectors
depends upon its ability to condense DNA into polyplexes with
positive charges, so as to easily access the negatively charged cell
Fig. 4 Particle size of OEI-GK/DNA polyplexes at different w/w ratios.
Data were shown as mean ꢃ SD (n ¼ 3).
Fig. 3 Agarose gel electrophoresis retardation assay of OEI-GKs.
15310 | J. Mater. Chem., 2011, 21, 15305–15315
This journal is ª The Royal Society of Chemistry 2011

View Article Online
membrane and promote the endocytosis by increasing either the
uptake rate or the endocytosis amount.34,35 Nevertheless, an
excessively high surface charge would lead to a detrimental effect
on the cells, which results in a compromised transfection efficacy.
Similar to the analyses of polyplex particle sizes, the surface
charges of the OEI-GK polyplexes also largely rely on the
polymer composition and architecture (Fig. 5). OEI-GK(1 : 2)
showed an apparently higher zeta potential than the other two
polyplexes, presumably due to more bound polymers per pDNA
and/or high volume charge density associated with the highly
branched structure. Note that although the binding assay indi-
cates that the DNA migration could be completely retarded at
the w/w ratio above 30 for OEI-GK(2 : 1), the polyplexes in the
case of relatively lower w/w ratios possessed positive charges. It
suggests that the formed polyplexes of OEI-GK(2 : 1) are not
stable, which somewhat agrees with their larger particle size as
measured by DLS.
At pH ¼ 7.2, OEI-GK(1 : 1)/DNA polyplexes are disclosed to
possess reasonably good stability as can be evidenced by slight
increment of the retarding ratio even after 48 h of incubation
(Fig. 6). Such structural stability of polyplexes is advantageous
for the protection of DNA from DNase degradation before
entering into cells. In sharp contrast to it, upon exposure to PBS
solution at pH ¼ 5.5, a marked decline of the binding strength
was distinctly observed. And the decline of DNA binding ability
was presented more profoundly as prolonging the incubation
time. It is noted that at pH ¼ 5.5, the protonated OEI-GK should
indeed have a stronger binding ability than its unprotonated
counterpart due to the higher charge density, provided that no
degradation occurs. Evidently, such an apparently reduced
affinity for DNA should be attributed to the acid-triggered
cleavage of ketal bonds, resulting in the formation of low
molecular weight OEI-GK fragments.
Those results suggest that the condensed DNA may be rela-
tively easily released from OEI-GK polyplexes after the acid-
triggered degradation in endosome.37 Lauffenburger and co-
workers established a computational model for intracellular gene
delivery, indicating that rather low dissociation rates of poly-
plexes would lead to low levels of transfer expression.38 Although
the appropriate interaction strength between DNA and vector
for the intracellular transfection process is not yet clear, it is
assumed that the partial decomposition of OEI-GK(1 : 1) and
the consequently reduced DNA–vector interaction occurring at
the endosome-escape step, may help free the DNA from the
hydrolyzed polyplexes and benefit the following transfection
procedures.
Binding ability assay after acid-triggered degradation
Viral vectors typically represent environment-responsive meta-
stable structures that simultaneously possess the stability to
withstand environmental damages as well as the instability to
respond to environmental cues, thus releasing the nucleic acid at
the appropriate sites.15,36 For bio-simulation purposes, poly-
plexes had better respond in an ‘‘intelligently dynamic’’ manner
to their microenvironment and in turn trigger correspondingly
adaptable functions at transfer steps. Responding to the unique
pH gradients in an endosome,11 the cleavage of ketal bonds in
OEI-GKs can be locally acid-triggered, leading to the partial
decomposition of OEI-GK polyconjugates into smaller frag-
ments. In addition to the expected long-term safety, such
biodegradation may also affect the binding ability and
membrane-disruptive capability as discussed below.
Comparison on the buffer capability before and after degradation
Endosomal escape is considered as one critical barrier with
regard to the efficient delivery of nucleic acids to cytoplasmic and
nuclear sites. Unlike viral vectors having fusion peptides at their
surface and being capable of escape from the endosome by
membrane fusion, the buffer capability of PEI gives rise to the so-
called ‘‘proton sponge‘‘ effect, which is assumed to be triggered
by an osmotic imbalance of the endosome upon PEI proton-
ation.20 The effect is believed to contribute much to the desta-
bilization of the endosome and subsequent cytoplasmic release of
internalized polyplexes. Therefore, the buffer capability can act
OEI-GK(1 : 1) was chosen as the model to identify the pH-
responsive biodegradation and the resulting decline of DNA
binding ability. The hydrolyses of OEI-GK(1 : 1) in endosomal
and physiological environments were simulated by incubating
the polymer in PBS with different pH values of 7.2 and 5.5,
respectively. After the treatment, the vectors were subjected to
agarose gel electrophoresis assay to investigate the DNA binding
ability (Fig. 6).
Fig. 5 Zeta potential of OEI-GK/pDNA polyplexes at different w/w
ratios. Data were shown as mean ꢃ SD (n ¼ 3).
Fig. 6 Agarose gel electrophoresis assay of OEI-GK(1 : 1) at different
pH values.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15311

View Article Online
as a useful indicator to roughly reflect the strength of the ‘‘proton
sponge’’ effect.
proposed conception termed ‘‘masked endosomolysis’’, where
the pH-labile bonds were designed to mask endosomolytic agents
but would undergo chemical bond cleavage to unmask a com-
pound’s endosomolytic activity.40–42 Such an ‘intelligent’ mask
would permit efficient in vivo delivery without sacrificing endo-
somolytic function for release into the cytoplasm. In addition,
the strong buffer capability would prevents the acidification of
endosomes and their fusion with lysosomes, by which the
possible enzymatic degradation of DNA is retarded.
In the present work, acid–base titration was used to evaluate
the buffer capability. The concentration of OEI800 moieties was
identical for all the investigated polyconjugate solutions before
titration. As shown in Fig. 7A, the buffer capability of the OEI-
GKs dropped substantially as compared to OEI800. Such
a decline of the buffer capability is usually reported for the
modified PEI derivatives and is attributed to the loss of free
amino groups after modification. Nevertheless, as seen in
Fig. 7B, upon acid-triggered degradation, OEI-GK(1 : 1) and
OEI-GK(2 : 1) exhibited insignificant difference in the titration
curve compared with OEI800. The interesting restoration of
buffer capability can be ascribed to the degradation of OEI-GKs
into small fragments with much better water-solubility. We thus
propose that the decline of buffer capability upon GK linking
should be more possibly associated with the decreased water-
solubility after polyconjugation. This explanation can be
demonstrated by the fact that alkylated PEI derivatives modified
with longer alkyl chains showed a relatively weaker buffer
capability despite the same substitution degrees.39 For this
reason, OEI-GK(2 : 1) with lowest molecular weight but better
water-solubility displayed much better buffer capability whereas
the other two were almost identical.
Another important finding is noted that although the buffer
capability of OEI-GK(1 : 2) had an apparent promotion after
degradation relative to the raw polymer prior to degradation, it
was still lower than that of OEI800. This implies that the linear
OEI-GK(1 : 1) may be more sensitive to acid-triggered degra-
dation than OEI-GK(1 : 2) with a three-dimensional hyper-
branched architecture.43 Possibly, the ketal groups in linear
OEI-GKs with a relatively flexible conformation in solution are
easily exposed to the exterior acid environment, leading to
a faster degradation rate.
In vitro transfection
The transfection mediated by OEI-GKs was investigated in
COS7 and HeLa cells, respectively. In the luciferase assay,
plasmid pGL-3 served as the reporter gene, and PEI25k, well-
established as the golden standard, was used as the control for
comparison purpose at its optimal w/w ratio of 1.3.3
The results suggest that the membrane-disruptive capability
would be greatly enhanced after OEI-GK degradation. This
degradation-associated change agrees somewhat with the
As clearly observed from Fig. 8, OEI-GK(1 : 1)/DNA poly-
plexes displayed the highest transfection efficiency amongst the
Fig.
8 Transfection efficiency of OEI-GKs/pDNA polyplexes at
Fig. 7 Acid–base titration profiles of OEI800 and OEI-GKs before (A)
and after degradation (B) in 0.15 M NaCl solution. Data were shown as
mean ꢃ SD (n ¼ 3).
different w/w ratios in COS7 cells (bottom) and HeLa cells (top)
compared with PEI25k control at w/w 1.3. Data were shown as mean ꢃ
SD (n ¼ 3).
15312 | J. Mater. Chem., 2011, 21, 15305–15315
This journal is ª The Royal Society of Chemistry 2011

View Article Online
Fig. 9 Fluorescent confocal microscopic images of OEI-GK/pGL-3 distribution in HeLa cells.
three OEI-GK polyplexes in both cell lines. At the w/w ratio of
about 10ꢄ15, the maximal luciferase expression was observed in
two cell lines transfected by OEI-GK(1 : 1)/DNA polyplexes.
OEI-GK(1 : 1) can mediate the transfection at the level even
higher than that of PEI25k by a factor of up to 10. Condensation
of DNA by a gene vector plays a key role in determining the
extracellular stability, endocytosis of polyplexes, and final gene
expression.44 OEI-GK(2 : 1)/DNA polyplexes were the least
effective in gene transfection despite having the highest cell-
biocompatibility, possibly due to the rather low DNA binding
ability. Hypothetically, there is a possibility that OEI-GK(2 : 1)/
DNA may potentially act as the vectors for the siRNA delivery in
terms of the much smaller size of siRNA and the need for only
cytoplasmic release as compared to pDNA delivery.45
a similar structure to PEI25k did not display the similarly high
transfection activity as anticipated in spite of its higher molecular
weight. Possibly, in addition to the buffer capability for endosomal
escape, the need for proper DNA binding ability before and after
degradation (responding to extracellular and intracellular trans-
fer) should be considered to account for this finding.
Overall, the luciferase assay suggests that the rational struc-
tural control on OEI-based polyconjugate vectors is of great
importance for a desirable transfection performance. It appears
that the architecture of OEI-GK(1 : 1) combining the branched
structure of OEI and the linear structure adopted by OEI-GK
(1 : 1) is superior to the thoroughly hyperbranched structure of
OEI-GK(1 : 2).
Provided that the higher transfection activity of OEI-GK(1 : 1)
over PEI25k is attributed to the lower cytotoxicity and the
degradation-induced structural destabilization of the polyplexes,46
it would be difficult to understand why OEI-GK(1 : 2) with
Intracellular fate of DNA/OEI-GK polyplexes
To visualize the intracellular localization of DNA after cellular
uptake, cancerous HeLa cells were incubated with pGL-3
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15313

View Article Online
ꢀ
polyplexes at 37 C for 4 h and then observed by confocal laser
cytoplasm and depart from the polyplexes to accomplish bio-
logical functions. In comparison, such a degradation-induced
improvement is not effective for OEI-GK(1 : 2). Thus the
endosomal escape and cytoplasmic release is somewhat
hampered, resulting a relatively lower transfection efficiency.
scanning microscopy. The pGL-3 plasmids in polyplexes were
labeled with YOYO-1 iodide, presenting green fluorescence. To
obtain further information concerning the cellular uptake, the
nuclei were stained by Hoechst 33258 and presented blue to
distinguish with the green fluorescence.
Tracking the DNA/OEI-GKs provides images similar to those
observed for PEI25k/DNA polyplexes (Fig. 9). Indeed, green
pGL-3 plasmids appear as discrete spots located in the cytoplasm
and blue-labeled nuclear areas. In comparison with other
vectors, OEI-GK(1 : 1) was evidently more effective in trans-
porting pGL-3 plasmids into the nuclei in terms of the stronger
florescence intensity. In contrast, very small amounts of green
spots could be detectable in the cells transfected with OEI-GK
(2 : 1) polyplexes. This indicates that the OEI-GK(2 : 1) poly-
plexes are difficult to enter into cells due to the instability of the
polyplexes leading to premature release and degradation of DNA
outside the cells. All the observations were in quite good
consistency with the quantitative analyses by luciferase assay.
Conclusions
Current developments of polycationic vectors involve two major
aims: to generate vectors that mediate a higher transfection
efficiency, and to make vectors less toxic, more biocompatible,
and biodegradable. The purpose of our study was to develop
a desirable gene vector specifically designed according to the
unique pH decline in the endosome compartment. OEI800 with
negligible toxicity but also poor transfection activity was linked
with ketalized glycolic acid to accomplish the enhanced DNA
delivery whilst retaining the high cell-biocompatibility. Three
kinds of acid-cleavable cationic polymers, termed OEI-GKs,
were prepared at different charge ratios, resulting in different
architectures and molecular weights of the obtained polymers.
By such a design, the multifunction of OEI-GKs was realized in
a controllable manner with improved acute cytotoxicity and
expected long-term biocompatibility, acid-responsive degrada-
tion as well as enhanced buffer capability and relatively easy
disassembly of polyplexes after endocytosis.
Hypothesis regarding OEI-GKs mediated transfection
Regarding the OEI-GKs mediated transfection, it is believed that
several main factors should be comprehensively taken into
account including the degradation rate of OEI-GKs, the DNA
binding ability and the buffer capability before and after
degradation. Based on all the results obtained in this study, we
may explain the difference in the transfection efficiency among
OEI-GKs in such a way as illustrated in Scheme 2. Due to the low
binding ability, the premature extracellular release of DNA is the
main cause for the low transfection efficiency of OEI-GK(2 : 1).
As for OEI-GK(1 : 1), the strong DNA binding ability confers
extracellular protection of DNA and effective endocytosis into
the cells. At the same time, acid-triggered degradation in the
endosome after endocytosis could lead to substantially enhanced
buffer capability and a somewhat reduced interaction between
DNA and vectors. Consequently, the internalized DNA is easier
to escape from the membrane-disrupted endosome into the
The OEI-GK(1 : 1) combining the branched structure of
OEI800 and the linear structure adopted by OEI-GK was found
to potentially act as an efficient gene vector in terms of the
transfection activity being even higher than PEI25k as well as
negligible cytotoxicity. Although great challenges still remain to
probe how acid-triggered degradation in the endosome affects
the gene transfection inside the cell by using existing techniques,
those encouraging results are believed to be associated with the
enhanced buffer capability and decreased DNA binding strength
of OEI-GKs upon acid-triggered decomposition in the
endosome.
The design of OEI-GKs provides us with a new way to deliver
DNA effectively and to remarkably reduce the cytotoxicity
usually associated with PEIs. Furthermore, the relatively stable
linkage amongst OEI-GK structures besides ketal groups, which
is different from the reported acid-cleavable PEI analogues, can
help us to develop a useful model for the further in-depth
investigation on the relationship between endosome-associated
polymer degradation and gene transfection efficacy, thus
acquiring more information in future about the transfer process
in the endosome.
Acknowledgements
This work was financially supported by the National Key Basic
Research
Program
of
China
(2011CB606202
and
2009CB930301) and National Natural Science Foundation of
China (Grant No. 20874075).
References
1 S. Han, R. I. Mahato, Y. K. Sung and S. W. Kim, Mol. Ther., 2000, 2,
302.
2 D. J. Glover, H. J. Lipps and D. A. Jans, Nat. Rev. Genet., 2005, 6,
299.
Scheme 2 Schematic illustration of the transfection process of OEI-
GKs.
15314 | J. Mater. Chem., 2011, 21, 15305–15315
This journal is ª The Royal Society of Chemistry 2011

View Article Online
3 F. W. Huang, H. Y. Wang, C. Li, H. F. Wang, Y. X. Sun, J. Feng,
X. Z. Zhang and R. X. Zhuo, Acta Biomater., 2010, 6, 4285.
4 H. Petersen, P. M. Fechner, A. L. Martin, K. Kunath, S. Stolnik,
C. J. Roberts, D. Fischer, M. C. Davies and T. Kissel, Bioconjugate
Chem., 2002, 13, 845.
5 P. J. Tarcha, J. Pelisek, T. Merdan, J. Waters, K. Cheung, K. Von
Gersdorff, C. Culmsee and E. Wagner, Biomaterials, 2007, 28, 3731.
6 D. Fischer, T. Bieber, Y. X. Li, H. P. Elsasser and T. Kissel, Pharm.
Res., 1999, 16, 1273.
7 W. T. Godbey, K. K. Wu and A. G. Mikos, J. Biomed. Mater. Res.,
1999, 45, 268.
8 M. L. Forrest, J. T. Koerber and D. W. Pack, Bioconjugate Chem.,
2003, 14, 934.
9 Y. T. Wen, S. R. Pan, X. Luo, X. Zhang, W. Zhang and M. Feng,
Bioconjugate Chem., 2009, 20, 322.
10 C. H. Ahn, S. Y. Chae, Y. H. Bae and S. W. Kim, J. Controlled
Release, 2002, 80, 273.
11 L. A. Bareford andP. W. Swaan, Adv. Drug DeliveryRev., 2007, 59, 748.
12 N. D. Sonawane, F. C. Szoka and A. S. Verkman, J. Biol. Chem.,
2003, 278, 44826.
25 F. W. Huang, J. Feng, J. Nie, S. X. Cheng, X. Z. Zhang and
R. X. Zhuo, Macromol. Biosci., 2009, 9, 1176.
26 C. F. Wang, Y. X. Lin, T. Jiang, F. He and R. X. Zhuo, Biomaterials,
2009, 30, 4824.
27 M. Van Dijk, M. L. Nollet, P. Weijers, A. C. Dechesne, C. F. Van
Nostrum, W. E. Hennink, D. T. S. Rijkers and R. M. J. Liskamp,
Biomacromolecules, 2008, 9, 2834.
28 S. E. Paramonov, E. M. Bachelder, T. T. Beaudette, S. M. Standley,
C. C. Lee, J. Dashe and J. M. J. Frechet, Bioconjugate Chem., 2008,
19, 911.
29 M. George and T. E. Abraham, J. Controlled Release, 2006, 114, 1.
30 W. J. Feast and N. M. Stainton, J. Mater. Chem., 1995, 5, 405.
31 W. C. Tseng and C. M. Jong, Biomacromolecules, 2003, 4, 1277.
32 S. Son and W. J. Kim, Biomaterials, 2010, 31, 133.
33 R. S. Burke and S. H. Pun, Bioconjugate Chem., 2008, 19, 693.
34 M. R. Park, K. O. Han, I. K. Han, M. H. Cho, J. W. Nah, Y. J. Choi
and C. S. Cho, J. Controlled Release, 2005, 105, 367.
35 H. L. Jiang, Y. K. Kim, R. Arote, J. W. Nah, M. H. Cho, Y. J. Choi,
T. Akaike and C. S. Cho, J. Controlled Release, 2007, 117,
273.
13 N. A. Jones, I. R. C. Hill, S. Stolnik, F. Bignotti, S. S. Davis and
M. C. Garnett, Biochim. Biophys. Acta, 2000, 1517, 1.
14 D. J. Chen, B. S. Majors, A. Zelikin and D. Putnam, J. Controlled
Release, 2005, 103, 273.
36 M. Uchida, M. T. Klem, M. Allen, P. Suci, M. Flenniken, E. Gillitzer,
Z. Varpness, L. O. Liepold, M. Young and T. Douglas, Adv. Mater.,
2007, 19, 1025.
37 D. V. Schaffer, N. A. Fidelman, N. Dan and D. A. Lauffenburger,
Biotechnol. Bioeng., 2000, 67, 598.
15 E. Wagner, Pharm. Res., 2004, 21, 8.
16 E. Mastrobattista, M. A. E. M. Van der Aa, W. E. Hennink and
D. J. A. Crommelin, Nat. Rev. Drug Discovery, 2006, 5, 115.
17 M. Lucas, U. Karrer, A. Lucas and P. Klenerman, Int. J. Exp.
Pathol., 2001, 82, 269.
38 C. M. Varga, K. Hong and D. A. Lauffenburger, Mol. Ther., 2001, 4,
438.
39 L. Aravindan, K. A. Bicknell, G. Brooks, V. V. Khutoryanskiy and
A. C. Williams, Int. J. Pharm., 2009, 378, 201.
18 J. O. You and D. T. Auguste, Nano Lett., 2009, 9, 4467.
19 M. S. Shim and Y. J. Kwon, J. Controlled Release, 2009, 133, 206.
20 A. Akinc, M. Thomas, A. M. Klibanov and R. Langer, J. Gene Med.,
2005, 7, 657.
40 D. B. Rozema, K. Ekena, D. L. Lewis, A. G. Loomis and J. A. Wolff,
Bioconjugate Chem., 2003, 14, 51.
41 W. A. Blattler, B. S. Kuenzi, J. M. Lambert and P. D. Senter,
Biochemistry, 1985, 24, 1517.
21 J. O. You and D. T. Auguste, Biomaterials, 2010, 31, 6859.
22 V. Knorr, V. Russ, L. Allmendinger, M. Ogris and E. Wagner,
Bioconjugate Chem., 2008, 19, 1625.
23 Y. L. Lin, G. H. Jiang, L. K. Birrell and M. E. H. El-Sayed,
Biomaterials, 2010, 31, 7150.
42 J. A. Wolff and D. B. Rozema, Mol. Ther., 2008, 16, 8.
43 M. S. Shim and Y. J. Kwon, Bioconjugate Chem., 2009, 20, 488.
44 D. D. Dunlap, A. Maggi, M. R. Soria and L. Monaco, Nucleic Acids
Res., 1997, 25, 3095.
45 Y. Y. Won, D. J. Gary and N. Puri, J. Controlled Release, 2007, 121,
64.
46 M. S. Shim and Y. J. Kwon, Biomaterials, 2010, 31, 3404.
24 Y. H. Kim, J. H. Park, M. Lee, Y. H. Kim, T. G. Park and S. W. Kim,
J. Controlled Release, 2005, 103, 209.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15305–15315 | 15315