Glutathione modified low molecular weight PEI for highly improved gene transfection ability and biocompatibility

Article abstract of DOI:10.1039/c9nj02396h

The efficient delivery of therapeutic genes remains a major challenge in realizing a feasible gene-based treatment. Herein, a versatile oligopeptide, glutathione, was introduced to construct novel non-viral cationic gene vectors. Reduced/oxidized forms of

Full text of DOI:10.1039/c9nj02396h

NJC
View Article Online
View Journal
PAPER
Glutathione modified low molecular weight PEI
for highly improved gene transfection ability
and biocompatibility†
Cite this:DOI: 10.1039/c9nj02396h
Yu-Rong Zhan, Qing-Ying Yu, Ji Zhang, * Yan-Hong Liu, Ya-Ping Xiao,
Ju-Hui Zhang, Xi He and Xiao-Qi Yu
*
The efficient delivery of therapeutic genes remains a major challenge in realizing a feasible gene-based
treatment. Herein, a versatile oligopeptide, glutathione, was introduced to construct novel non-viral
cationic gene vectors. Reduced/oxidized forms of glutathione (GSH/GSSG) and relevant amino acids
(Glu, Cys, and Gly) were used to modify low molecular weight PEI through surface modification or
crosslinking. These polymers could bind well and condense DNA into spherical nanoparticles, which
were stable in the presence of serum. The disulfide bonds within the crosslinked polymer GSSG-PEI may
facilitate polymer degradation and DNA release under a reductive environment. In vitro transfection
experiments reveal that the modification could largely improve the gene transfection efficiency of low
molecular weight PEI, especially in the presence of serum. In HeLa cells, GSSG-PEI could even give up
to 150 times higher efficiency than PEI 25 kDa. TEM and serum concentration effect assay also
demonstrate the good serum tolerance of the polymers. Flow cytometry results show that GSSG-PEI
might induce cellular uptake with higher efficiency than PEI 25 kDa, especially in the presence of serum.
Results reveal that GSSG is a good candidate for the crosslinking of small cationic molecules to form
polymeric gene vectors with improved transfection efficiency and biocompatibility.
Received 9th May 2019,
Accepted 4th July 2019
DOI: 10.1039/c9nj02396h
rsc.li/njc
transfection efficiency resulting from its effective endosomal
escape through the ‘‘proton-sponge’’ mechanism.11 However,
1. Introduction
With the development of molecular biology and gene transfer high molecular weight PEI shows not only high transfection
techniques, therapeutic interventions at the genetic level have efficiency but also high cytotoxicity, while low molecular weight
gained more attention and achievement in recent decades, and PEI shows low cytotoxicity but poor transfection efficiency.12,13
gene therapy is regarded as a potential clinical treatment.1–5 Therefore, development of PEI derivatives with high efficiency
To achieve efficient gene delivery, it is necessary to overcome together with low toxicity via appropriate modification is of
several extracellular and intracellular barriers including gene special significance. Some strategies have been applied for the
loading, systemic long circulation, cellular uptake, endosomal modification of PEI: (1) crosslinking low molecular weight PEI
escape and cargo release. Thus, a safe and efficient vector is with biodegradable bonds, such as ester,14,15 disulfide,16,17
essential in gene therapy.6 Although some viral vectors exhibited ketal,18 imine19 and hydrazine,20 which are cleavable under an
high efficiency, it is difficult to control their immunogenicity and intracellular acid and reductive environment, facilitating the
tropism.7,8 In contrast, non-viral vectors are more advantageous in release of gene cargo and the decrease of cytotoxicity; (2) attaching
biosafety and preparation, making them the research focus of gene biocompatible components, including amino acids,21–23 PEG,24,25
vectors.9,10
and polysaccharides,26,27 to decrease the cytotoxicity of PEI and
Among the numerous non-viral vectors, cationic polymer is extend systemic circulation; (3) grafting targeting ligands, such as
the most studied type for its stability, high nucleic acid loading, RGD peptides, folic acid, and hyaluronic acid, to promote the
and easy modification. Polyethylenimine (PEI) is one of the specificity of cellular uptake;13,18,28–31 (4) introducing stimuli-
promising non-viral cationic polymeric vectors due to its high responsive motifs to achieve effective controlled release;28–32
(5) adding short hydrophobic substituents to endow the vector
with amphiphilicity, which could promote interaction with the cell
Key Laboratory of Green Chemistry and Technology (Ministry of Education),
membrane and polyplex stability.33
College of Chemistry, Sichuan University, Chengdu 610064, P. R. China.
Glutathione (GSH), L-g-glutamyl-L-cysteinyl-glycine, as
a
E-mail: jzhang@scu.edu.cn, xqyu@scu.edu.cn
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9nj02396h versatile oligopeptide widely existing in cells, plays significant
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
roles in free radical scavenging, detoxification, liver protection could be obtained. For comparison, PEI modified with relevant
and anti-cancer.34,35 In cells, glutathione is simultaneously amino acids including glutamic acid, cysteine, and glycine
present in another form, glutathione disulfide (GSSG), which was also prepared (named as Glu-PEI, Cys-PEI, and Gly-PEI
is formed upon oxidation and is also referred to as ‘‘oxidized respectively). Different feed ratios (mole ratio of amino acid or
glutathione’’.36 According to previous reports, GSH has been GSH to amino groups in PEI 1800 Da) were used to obtain a
applied as a capping agent to impart biocompatibility and series of modified PEI derivatives with different substitution
monodispersity to cell-imaging quantum dots and bimodal degrees (SDs), which could be calculated according to the ratio
imaging probes.37,38 Besides, the GSSG-modified nanoparticle of characteristic peak integrals in the 1H NMR spectra. The data
(GSSG@Fe3O4) was found to show high biocompatibility and shown in Table S1 (ESI†) indicate that the actual SDs were lower
effective cellular uptake.39 It was also reported that the addition of than the feed ratio because of the steric hindrance and different
GSH could reduce the cytotoxicity of PEI-coated nanoparticles.40 reactivity of the various amino groups. According to previous
At present, GSH is mainly used as a biocompatible shield on the results,22 the modified PEI with a SD of B10% was chosen
surface of nanoparticles. Therefore, we herein introduced GSH/ for subsequent study. Besides, the oxidized form of GSH (GSSG)
GSSG moieties into low molecular weight PEI though covalent was also used to connect PEI to form crosslinked polymers
surface modification (GSH) or crosslinking (GSSG) to create novel (GSSG-PEI). Since several PEI units may exist in one polymer
gene vectors. The unique nature of GSH was expected to endow molecule, a lower molecular weight (600 Da) was used for the
vectors with excellent biocompatibility and transfection efficiency. crosslinking. Similarly, the actual mole ratio of PEI 600 Da to
Furthermore, the disulfide bond in GSSG is a classical reduction- GSSG in the polymer was estimated to be B3 according to the
1
responsive motif whose cleavage under an intracellular reducing characteristic peak integrals in the H NMR spectrum, slightly
environment contributes to cargo release during gene delivery.
lower than the feed mole ratio of 4. GPC was used to measure
the molecular weight (Mw) and polydispersity indexes (PDIs) of
polymers, which are shown in Table S2 (ESI†). The Mws
of single amino acid modified PEI were found to be around
5 kDa, while GSH-PEI and GSSG-PEI exhibited a higher Mw
(7–7.5 kDa). The molecular weight of all polymers was distinctly
lower than that of PEI 25 kDa, which would be beneficial for the
biocompatibility.
2. Results and discussion
2.1. Preparation of target polymers
The target polymers were prepared according to the route
shown in Scheme 1 and Scheme S1 (ESI†). PEI with a molecular
weight of 1800 Da was chosen for the modification. GSH was
grafted on PEI through several steps starting from dithiodipyr-
idine, which first reacted with mercaptopropionic acid to give
compound 1. The carboxyl of 1 was subsequently condensed
2.2. Formation and characterization of polymer/DNA
complexes
with the amino group of PEI to yield material 2. After the It is essential for gene vectors to have the ability to bind and
mercapto-exchange reaction with GSH, target polymer GSH-PEI condense DNA into nanoparticles with suitable size and surface
Scheme 1 Synthetic route of target polymers.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
potential. Agarose gel retardation assay was carried out to
assess the capacity of polymers to bind DNA. As shown in
Fig. 1A, all the polymers could hinder the migration of DNA
entirely at polymer/DNA weight ratios (w/w) of 0.8, which was
just slightly higher than PEI 25 kDa (0.4), suggesting their good
cationic capabilities to bind and condense DNA.
The stability of polymer/DNA complexes (polyplexes) was
subsequently investigated, and heparin was used to bind the
cationic polymers and induce the DNA release. The results
shown in Fig. 1B indicate that DNA release could be observed
at a relatively higher heparin/DNA weight ratio of 8–16, reflecting
the good stability of the polyplexes, which is important for DNA
protection. It could be found that PEI 25 kDa bound to DNA so
tightly that heparin could not release DNA from its polyplex even
at high dosage. Such tight binding may hinder the DNA release in
the transfection process, leading to a negative effect on transfec-
tion efficiency.
Fig. 2 DTT assisted DNA release from the polyplexes at various weight
ratios (DTT: 20 mM, heparin: 0.4 mg mLꢀ1).
The disulfide bond in GSSG-PEI was expected to be cleaved
under the reductive environment in cells to facilitate the release
of DNA. Thus, DTT was used to verify the reduction respon-
siveness of GSSG-PEI. As shown in Fig. 2, upon incubation with
DTT, DNA could be released by heparin more easily. In other
words, DTT could promote the release of DNA. However, since
PEI 600 Da could also interact and bind with DNA, total release
could not be achieved.
The physical properties of the polyplexes, such as particle
size and zeta potential, were evaluated by using the dynamic
light scattering (DLS) method. As shown in Fig. 3A, with the
increase of w/w, the particle size of polyplexes gradually
decreased and tended to be stable at 200 to 400 nm, which is
a size suitable for cellular uptake.41 Meanwhile, the zeta
potential of the polyplexes underwent charge reversal from
negative to positive with a w/w increase and ultimately stabi-
lized at around +30 mV (Fig. 3B), which is indispensable for
polyplexes to facilitate their interaction with the cell membrane
and subsequent cellular uptake. In addition, the morphology of
the polyplex was observed directly via TEM. Taking GSSG-PEI as
an example, the polyplex particles under serum-free or serum-
containing conditions are shown in Fig. 3C. The polyplex
existed in the form of approximately spherical particles that
were seldom affected by serum, indicating their good stability
and serum tolerance.
2.3. In vitro gene transfection
To evaluate the gene transfection efficiency of these polymers,
experiments on luciferase reporter gene pGL-3 plasmid expres-
sion were carried out in two cell lines, and the results are shown
in Fig. 4. In the absence of serum in HeLa cells (Fig. 4A), all the
polymers performed better than PEI 1800 Da, demonstrating
the advantage of modification. Nevertheless, compared to PEI
25 kDa, only GSSG-PEI showed higher efficiency, suggesting
that crosslinking might be a better strategy than surface
modification for low molecular weight PEI. In 7702 cells, all
the polymers could give higher transfection efficiency than PEI
25 kDa (Fig. 4C). Unfortunately, although the crosslinked
polymer GSSG-PEI showed good results, the GSH modified
material GSH-PEI did not exhibit superior performance compared
to the single amino acid decorated PEIs. On the other hand, in the
presence of serum, all the studied polymers exhibited higher
transfection efficiency relative to PEI 25 kDa. For GSSG-PEI in
HeLa cells, up to 150 times higher transfection efficiency than PEI
25 kDa could be achieved (Fig. 4B). Without the presence of
serum, only 2.7 times higher efficiency was obtained. Other amino
acid/GSH modified PEIs also showed such a trend. Besides, the
enhanced green fluorescent protein expression experiment results
(Fig. S1, ESI†) also visually reveal that all the modified polymers
Fig. 1 (A) DNA retardation by polymers at different weight ratios. (B) DNA release assay by the addition of heparin.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Fig. 3 Particle size (A) and zeta potential (B) of polyplexes at various weight ratios. Data represent mean ꢁ SD (n = 3). (C) TEM images of GSSG-PEI/DNA
complex (w/w = 16) without serum (left) or with serum (right), scale bar: 50 nm.
could induce more green fluorescence than PEI 1800 Da both in be found that serum has a less negative effect on the modified
tumor and normal cells, and GSSG-PEI showed better transfection PEIs than that on PEI 25 kDa, especially for the MFI. The dramatic
behavior than PEI 25 kDa, especially in the presence of serum. decrease of transfection efficiency of PEI 25 kDa could be attrib-
These results suggest that the modification PEI could not only uted to the largely reduced amount of internalized DNA. There-
enhance the transfection efficiency of the cationic materials, but fore, protecting DNA from extracellular degradation by nuclease
also improve the serum tolerance of the vectors. The higher or early release by serum proteins and other components is
transfection efficiency of GSSG-PEI might also come from its significant for the design of PEI derived vectors. Besides, confocal
reduction responsiveness, which may facilitate DNA release in cells. laser scanning microscopy (CLSM) was also used to visualize the
Furthermore, the effect of serum concentration on the transfection effective internalization of Cy5-labeled DNA delivered by the
efficiency of GSSG-PEI was also investigated. The results in Fig. 5 polymer vectors in the presence of serum (Fig. S2, ESI†).
show that the increase of serum concentration only caused a slight
Cytotoxicity is also an important factor to evaluate non-viral
decrease of efficiency. On the contrary, the PEI 25 kDa-mediated gene vectors. The cell viability of the polyplexes prepared at various
transfection efficiency gave rise to a dramatic inhibition, further weight ratios was measured by MTS assay in HeLa cells, and the
verifying the excellent serum tolerance of GSSG-PEI.
results are shown in Fig. 7. The modification led to no obvious
Flow cytometry was performed to measure the cell interna- change in the cell viability, except for the polymers with higher
lization of Cy5-labeled DNA carried by the polymers using the molecular weight at high weight ratios. All of the polymers exhi-
fluorescence activated cell sorting (FACS) technique. After 4 h bited higher cell viability than PEI 25 kDa, indicating that mole-
of incubation with the polyplexes in HeLa cells, the percentage cular weight is still the main factor in the cytotoxicity of cationic
of positive cells for Cy5-labeled DNA and mean fluorescence polymers. Combined with the transfection results, it is proved that
intensity (MFI) could be calculated. Among the two quantities, the rational modification of low molecular weight PEI with amino
the MFI may reflect the amount of internalized Cy5-labeled acids or small peptides is a promising strategy to enhance the gene
DNA more accurately. As shown in Fig. 6A, GSSG-PEI exhibited transfection efficiency while reducing the toxicity of the vectors.
slightly lower cellular uptake percentages than PEI 25 kDa both
in the absence and presence of serum. However, it could induce
higher fluorescence intensity (Fig. 6B), suggesting that GSSG-PEI
may induce cellular uptake with higher efficiency. This might be
due to the enhanced biocompatibility via the GSSG modification,
3. Experimental section
3.1. Materials and methods
which made the polyplexes enter cells more easily, especially in All reagents and chemicals were purchased from commercial
the presence of serum. Similar to the transfection results, it could providers and used without further purification unless specially
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
Fig. 4 Luciferase gene expression mediated by the polyplexes at different weight ratios in HeLa (A: without serum; B: with 10% serum) and 7702 cells
(C: without serum; D: with 10% serum). PEI 1.8 and 25 kDa as control.
Sigma-Aldrich (St. Louis, MO, USA) and low molecular weight PEI
(PEI 1800 Da, Mw = 1800; PEI 600 Da, Mw = 600) was purchased
from Aladdin (Shanghai, China). N-Boc-glycine, cysteine, g-benzyl-
glutamate, reduced glutathione and 2,20-dithiodipyridine were
purchased from Energy Chemistry (Shanghai, China). Oxidized
glutathione was purchased from J&K Scientific Ltd. The plasmids
used in this work were pEGFP-N1 (Clontech, Palo Alto, CA, USA)
coding for Enhanced Green Fluorescent Protein (EGFP) DNA and
pGL-3 (Promega, Madison, WI, USA) coding for luciferase DNA.
HeLa and 7702 cell lines were purchased from the Shanghai
Institute of Biochemistry and Cell Biology, Chinese Academy
of Sciences. Cy5TM was supplied by Molecular Probe (Mirus,
Madison, WI, USA). DMEM, 1640 medium and fetal bovine
serum (FBS) were supplied by Invitrogen Corp (Carlsbad, CA,
Fig. 5 Effect of serum concentration on the transfection efficiency of
GSSG-PEI (w/w = 16) and PEI 25 kDa (w/w = 1.4) in HeLa cells.
USA). The MicroBCA protein assay kit was obtained from Pierce
(Rockford, IL, USA). The luciferase assay kit and MTS (3-(4,5-
dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfo-phenyl)-
2H tetrazolium, inner salt) were obtained from Promega (Madison,
mentioned. Anhydrous dichloromethane (DCM) was distilled WI, USA).
after being dried with calcium hydride (CaH2). All aqueous
The 1H NMR and 13C NMR spectra were measured on a
solutions were prepared using deionized water. Column chroma- Bruker AM400 NMR spectrometer (Bruker Corporation, Billerica,
tography was performed using 200–300 mesh silica gel. Branched MA, USA). HRMS spectral data were recorded on a Bruker
polyethylenimine (PEI 25 kDa, Mw = 2.5 ꢂ 104) was supplied by Daltonics Bio TOF mass spectrometer. The molecular weights
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
Fig. 6 Cellular uptake percentage (A) and mean fluorescence intensity (B) of Cy5-labeled DNA induced by the polymers without (ꢀ) or with serum (+) in
HeLa cells. w/w (polymer/DNA) = 1.4 for PEI 25 kDa, and 16 for other materials.
and 30%, respectively) dissolved in anhydrous DCM (20 mL)
was added to the reaction system dropwise. Here, different feed
ratios were chosen to obtain a series of polymers with different
substitution degrees. After stirring for 24 h at room temperature,
the excess solvent was removed under reduced pressure and
the residue was dialyzed (MWCO 1000 Da) against DMF and
deionized water for 3 days. Compound 2 was then obtained as a
pale yellow solid after lyophilization.
Compound 2 was dissolved in 20 mL of solvent (methanol/
H2O = 1/1) with acetic acid (100 mL), and GSH was added to
the reaction mixture. After stirring overnight, methanol was
removed under reduced pressure and the residue was dialyzed
(MWCO 1000 Da) against deionized water for 3 days. The
product GSH-PEI was obtained as a white solid after lyophilization.
Total yield of the two steps: 23% (calculated by comparing the
Fig. 7 Cytotoxicity of complexes at various weight ratios in HeLa cells.
mass of the product to starting PEI, and the same below).
Compound 2: 1H NMR (D2O, 400 MHz) d 8.34, 7.84–7.78,
of the polymers were determined via gel permeation chromato-
graphy (GPC), which consisted of a Waters 515 isocratic HPLC
pump, a Linear 7.8 ꢂ 300 mm column (Waters Corp, Milford, MA,
USA), an 18-angle laser scattering instrument (Wyatt Technology
Corporation, USA), and an OPTILAB DSP interferometric refracto-
meter (Wyatt Technology Corporation, USA). 0.5 M HAc/NaAc
buffer solution was used as the mobile phase at a flow rate
of 1 mL minꢀ1. Molecular weights were calculated against
poly(ethylene glycol) standards, with average molecular weights
ranging from 601 to 106 000.
7.27–7.22 (d, J = 4 Hz, m, m, Ar-H), 3.03–2.63 (m, PEI-H,
-S-CH2CH2-COOH).
1
GSH-PEI: H NMR (D2O, 400 MHz) d 3.71–3.68 (t, J = 12 Hz,
COOH-CH2-NHCO-,
-CH2CH2-CH(NH2)-COOH),
3.44–2.45
(m, PEI-H, -CH2-SH, -NHCO-CH2CH2-CH(NH2)-COOH), 2.12–2.06
(m, -NHCO-CH2CH2-CH(NH2)-COOH).
Preparation of Glu/Cys/Gly-PEI. g-Benzyl-glutamate (16 mmol)
and TEA (32 mmol) were dissolved in 100 mL of solvent (THF/
H2O = 2/1), di-tert-butyl dicarbonate ((Boc)2O, 19.2 mmol) in
50 mL of THF was added to the mixture dropwise in an ice bath
and the system was stirred overnight at room temperature. THF
was removed under reduced pressure and the aqueous solution of
3.2. Preparation of target polymers
Preparation of GSH-PEI. Compound 1 was prepared according the crude product was adjusted to pH 2 with HCl hydrous solution
to a previous report.42 Compound 1 (2 mmol), N-hydroxy- (2N). The resulting solution was extracted with EA three times and
succinimide (NHS, 2.4 mmol), N-(3-dimethylaminopropyl)-N0- the combined extracts were washed with brine. After drying over
ethylcarbodiimide hydrochloride (EDCI, 2.4 mmol), and anhydrous Na2SO4, the organic phase was concentrated and
triethylamine (TEA, 2.4 mmol) were dissolved in anhydrous purified via column chromatography (silica gel, PE/EA = 1/1).
DCM (30 mL) and stirred for 3 h at 0 1C. Then, the desired Compound 3 was obtained as a pale yellow viscous liquid.
amount of PEI 1800 Da (0.86, 0.43 and 0.28 g, mole ratios of In the same way, compound 4 could be obtained as a light red
compound 1 to amino groups in PEI 1800 Da were 10%, 20% viscous liquid from cysteine.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
Compound 3 (Scheme S1, ESI†): 1H NMR (CDCl3, 400 MHz) 1% (w/v) agarose gel containing GelRed and Tris-acetate (TAE)
d 7.35 (s, 5H, Ph-H), 5.12 (s, 2H, Ph-CH2-), 4.37–4.25 (m, 1H, running buffer at 130 V for 30 min. DNA was visualized under
Boc-NH-CH-COOH), 2.57–2.44 (m, 2H, -CH-CH2-CH2-COOPh), an ultraviolet lamp using a Bio-Rad Universal Hood II (Berkeley,
2.25–2.05 (m, 2H, -CH-CH2-CH2-COOPh), 1.43 (s, 9H, Boc-H). CA, USA).
13C NMR (CDCl3, 100 MHz) d 176.09, 172.72, 155.68, 135.85,
The ability of polymers to bind DNA could be studied by the
128.67, 128.30, 128.25, 80.27, 66.62, 52.76, 30.53, 28.41, 27.38. heparin sodium replacement assay. At first, the desired amount
HRMS (DCM): m/z: calcd: 360.1423, found: 360.1421 [M + Na]+. of polymers and 0.125 mg of pUC-19 were incubated to obtain
Compound 4 (Scheme S1, ESI†): 1H NMR (CDCl3, 400 MHz) the complexes (w/w = 2). Then, a series of heparin sodium
d 4.64 (s, 1H, Boc-NH-CH-COOH-), 3.07–2.96 (m, 2H, -CH2-SH), solutions with different concentrations were added to the
1.46 (s, 9H, Boc-H). 13C NMR (CDCl3, 100 MHz) d 174.69, complexes and the mixture was incubated at 37 1C for another
155.48, 80.80, 54.40, 28.27, 27.04. HRMS (DCM): m/z: calcd: 30 min. Similarly, gel retardation assay was carried out as above.
244.0619, found: 244.0614 [M + Na]+.
Besides, DTT was used to investigate the reduction respon-
N-Boc-amino acid (2 mmol), EDCI (2.4 mmol), 1-hydroxy- siveness of GSSG-PEI. Two groups of experiments were under-
benzotriazole (HOBT, 2.4 mmol) and ethyldiisopropylamine taken; in one group, DTT solution (20 nM) was added to the
(DIEA, 2.4 mmol) were dissolved in anhydrous DCM (30 mL) polymer/DNA complexes, while in the other group, water
and stirred for 3 h at 0 1C. Then, the desired amount of PEI was added as a control. Then, the mixtures were incubated
1800 Da (0.86, 0.43, 0.28 and 0.17 g, mole ratios of N-Boc-amino for 3 h at 37 1C. Heparin sodium (0.4 mg mLꢀ1) was added to
acid to amino groups in PEI 1800 Da were 10%, 20%, 30% and replace DNA.
50%, respectively) dissolved in anhydrous DCM (20 mL) was
added to the reaction system dropwise. After stirring for 48 h
3.4. Morphology study of polymer/DNA complexes
at room temperature, HCl hydrous solution (1N) was added The size and zeta potential of complexes at various weight
to remove the protecting group. The excess solvent was ratios were evaluated by dynamic light scattering (DLS) at 25 1C
removed under reduced pressure and the residue was dialyzed by using a Nano-ZS ZEN 3600 (Malvern Instruments, Malvern,
(MWCO 1000 Da) against deionized water for 3 days. The Worcestershire, UK). The complexes were prepared by adding
product amino acid–PEI was obtained as a pale yellow solid 1 mg of pUC-19 to the appropriate volume of polymer solution.
after lyophilization.
After incubating for 30 min at 37 1C, the solution was diluted to
Glu-PEI: yield 26%. 1H NMR (D2O, 400 MHz) d 3.91–3.85 1 mL with deionized water before measurement. All data were
(m, -NH-CH-COOH), 3.15–2.74 (m, PEI-H).
averaged and a variance was obtained after 3 parallel experiments.
The morphology of the complexes was also observed via
TEM (Tecnai G2 F20 S-TWIN, Hillsboro, OR, USA) with an
Cys-PEI: yield 9%. 1H NMR (D2O, 400 MHz) d 3.58 (s, NH2-CH-
CONH-), 3.23–2.76 (m, PEI-H, -CH2-SH).
1
Gly-PEI: yield 19%. H NMR (D2O, 400 MHz) d 3.77 (d, J = acceleration voltage of 100 kV. 2 mg of pUC-19 was added to
8 Hz, NH2-CH2-CONH-), 3.20–2.73 (m, PEI-H).
the appropriate volume of polymer (weight ratio of GSSG-PEI
Preparation of the crosslinked polymer GSSG-PEI. Compound relative to DNA, w/w = 16), and the mixture was incubated at
5 could be obtained according to a previous report.43 Compound 5 37 1C for 30 min. 15 min before measurements, the complex
(0.5 mmol), DIEA (2.4 mmol), HOBT (2.4 mmol), and EDCI solution was diluted with deionized water (or PBS containing
(2.4 mmol) were dissolved in anhydrous DCM (20 mL) and stirred 10% FBS) to 200 mL. A drop of complex suspension was placed
for 3 h at 0 1C. Then, PEI 600 Da (1.2 g, mole ratio of PEI 600 Da to onto the copper grid. After 3 min, the excess solution was
compound 5 was 4) dissolved in anhydrous DCM (30 mL) was removed with filter paper. Then, a drop of 2% (w/v) phospho-
added to the reaction system dropwise. After stirring for 48 h tungstic acid was placed on the above grid. The grid was dried
at room temperature, HCl hydrous solution (1N) was added to at room temperature at atmospheric pressure for several minutes
remove the protecting Boc group. The excess solvent was removed before observation.
under reduced pressure and the residue was dialyzed (MWCO
1000 Da) against deionized water for 3 days. The product
3.5. Cell culture
GSSG-PEI was obtained as a pale yellow solid after lyophilization HeLa cells were cultured in DMEM medium containing 10%
with a yield of 34%.
GSSG-PEI: 1H NMR (D2O, 400 MHz) d 4.57–4.53 (m, -CH-CH2-S-), while 7702 cells were incubated in 1640 medium containing
3.92 (s, -CH2CH2-CH(NH2)-CONH-), 3.74–3.73 (m, -NHCO-CH2- 10% FBS and 1% antibiotics (penicillin–streptomycin, 10 kU mLꢀ1
FBS and 1% antibiotics (penicillin–streptomycin, 10 kU mLꢀ1),
)
NHCO-), 3.39–2.73 (m, PEI-H, -CH-CH2-S-), 2.50 (s, -CH2CH2-C at 37 1C in a humidified atmosphere containing 5% CO2.
H(NH2)-CONH-), 2.14–2.13 (m, -CH2CH2-CH(NH2)-CONH-).
3.6. Gene transfection assay in vitro
3.3. Gel retardation assay
Gene transfection efficiency of complexes was investigated in
Polymer/DNA complexes at different w/w ratios ranging from HeLa and 7702 cells. For the expression of the EGFP gene, HeLa
0.25 to 16 were prepared by adding an appropriate volume and 7702 cells were seeded in 48-well plates (5 ꢂ 104 cells per well)
of polymer solution to 5 mL of pUC-19 (0.025 mg mLꢀ1). The and incubated for 24 h at 37 1C in 5% CO2. Before transfection,
resulting complex solutions were diluted to 10 mL and incubated at the medium in every well was exchanged for fresh serum-free or
37 1C for 30 min. After that, the complexes were electrophoresed on 10% serum-containing medium containing polymer/DNA (0.4 mg)
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.

View Article Online
Paper
NJC
complexes at various weight ratios. After incubation for 4 h, the 4. Conclusion
medium was replaced with 10% serum-containing medium. After
another 20 h incubation, cells were observed using an inverted
fluorescence microscope (Nikon TS100, Tokyo, Japan).
In summary, low molecular weight PEI was modified by reduced/
oxidized glutathione and three relevant amino acids to obtain a
series of target polymers. Gel retardation and heparin replace-
ment assays proved that all of these polymers could bind and
condense DNA into stable polyplexes. Besides, the crosslinked
polymer GSSG-PEI showed a reduction responsive property, which
was beneficial to the DNA release in an intracellular reductive
environment. The modification could largely improve the gene
transfection efficiency of low molecular weight PEI, and their
transfection efficiencies were even higher than PEI 25 kDa.
In particular, in the presence of serum, GSSG-PEI had 150 times
higher efficiency than PEI 25 kDa. TEM, serum concentration
effect assay and flow cytometry assay also demonstrated the good
serum tolerance of the target polymer. GSSG-PEI might induce
cellular uptake with higher efficiency than PEI 25 kDa, especially
in the presence of serum. These results revealed that GSSG is a
For luciferase expression, the cells were transfected with
complexes containing pGl-3 plasmids. After a similar transfec-
tion process as described above, cells were washed with PBS
and lysed with 60 mL 1ꢂ Lysis reporter buffer (Promega,
Madison, WI, USA). The luciferase activity was measured by
using a microplate reader (Imark, Bio-Rad, Berkeley, CA, USA).
The protein content of the lysed cell was determined by BCA
protein assay (Pierce, Rockford, IL, USA). Gene transfection
efficiency was shown as the relative fluorescence intensity per
mg protein (RLU per mg protein). All the experiments were
carried out in triplicate.
3.7. Cellular uptake of plasmid DNA (flow cytometry)
The cellular uptake efficiency of polymer/fluorescein labeled good candidate for the crosslinking of small cationic molecules to
DNA complexes was evaluated via flow cytometry (Accuri C6, form polymeric gene vectors with improved transfection efficiency
BD, Franklin Lakes, NJ, USA). Here, polymers and DNA labeled and biocompatibility.
using a Label IT Cy5 Labeling Kit were used to form complexes
at a weight ratio of 16. After 4 h transfection, the cells were
Conflicts of interest
washed with 1ꢂ PBS and digested with trypsin to obtain a
dispersive cell suspension. Cy5-labeled plasmid DNA uptake
was measured in the FL4 channel using a red diode laser
(633 nm). The mean fluorescence intensity was analyzed using
a flow cytometer (Accuri C6, BD, USA).
There are no conflicts to declare.
Acknowledgements
This work was financially supported by the National Natural
Science Foundation of China (No: 21672155 and 21877082).
We also acknowledge the Comprehensive Training Platform
of Specialized Laboratory, College of Chemistry, Sichuan
University for sample analysis.
3.8. Confocal laser scanning microscopy (CLSM) analysis
HeLa cells were seeded at a density of 2.5 ꢂ 105 cells per well in
a 35 mm confocal dish, 24 h prior to transfection. Then, the
medium was exchanged with fresh serum-containing medium.
Complexes of polymers and Cy5-labeled pGL-3 at a weight ratio
of 24 were added to each well. After incubation at 37 1C for 4 h,
the medium was removed and cells were rinsed three times Notes and references
with PBS (pH = 7.4). Then, cells were fixed with 4% para-
1 U. Lachelt and E. Wagner, Chem. Rev., 2015, 115,
11043–11078.
formaldehyde (dissolved with PBS solution) for 15 min and
nuclear staining was done using Hoechst 33342. The CLSM
observation was performed via confocal laser scanning microscopy
(LSM 780, Zeiss, Jena, Germany) at excitation wavelengths of
361 nm for Hoechst 33342 (blue) and 633 nm for Cy5 (red),
respectively.
2 R. Voelker, JAMA, J. Am. Med. Assoc., 2018, 319, 434.
3 C. H. Evans, S. C. Ghivizzani and P. D. Robbins, Nat. Rev.
Rheumatol., 2011, 7, 244–249.
4 J. C. T. van Deutekom and G. J. B. van Ommen, Nat. Rev.
Genet., 2003, 4, 774–783.
5 A. D. Miller, Nature, 1992, 357, 455–460.
6 S. M. Dizaj, S. Jafari and A. Y. Khosroushahi, Nanoscale Res.
Lett., 2014, 9, 252–260.
7 R. Waehler, S. J. Russell and D. T. Curiel, Nat. Rev. Genet.,
2007, 8, 573–587.
8 M. A. Kay, Nat. Rev. Genet., 2011, 12, 316–328.
9 M. A. Mintzer and E. E. Simanek, Chem. Rev., 2009, 109,
259–302.
3.9. Cell viability assay
The cytotoxicity of the complexes towards HeLa and 7702 cells
was studied by using the MTS Kit. After 24 h incubation, the
medium containing complexes was removed and MTS in PBS
was added for another 1 h incubation. Then, the absorbance of
each sample was measured using an enzyme-linked immuno-
sorbent assay (ELISA) plate reader (Model 680, Bio-Rad, Berkeley,
CA, USA) at a wavelength of 490 nm. The cell viability (%) was 10 M. S. Al-Dosari and X. Gao, AAPS J., 2009, 11, 671–681.
obtained according to the manufacturer’s instructions. The 11 R. V. Benjaminsen, M. A. Mattebjerg, J. R. Henriksen,
untreated cell controls were taken as 100% cell viability.
All the experiments were carried out in triplicate.
S. M. Moghimi and T. L. Andresen, Mol. Ther., 2013, 21,
149–157.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019

View Article Online
NJC
Paper
12 Y. H. Wang, M. Zheng, F. H. Meng, J. Zhang, R. Peng and 28 J. H. Wang, H. He, X. Xu, X. Wang, Y. B. Chen and L. C. Yin,
Z. Y. Zhong, Biomacromolecules, 2011, 12, 1032–1040. Biomaterials, 2018, 171, 72–82.
13 H. Y. Tian, L. Lin, J. Chen, X. S. Chen, T. G. Park and 29 H. Kim, J. Kim, M. Lee, H. C. Choi and W. J. Kim, Adv.
A. Maruyama, J. Controlled Release, 2011, 155, 47–53. Healthcare Mater., 2016, 5, 1918–1930.
14 Q. Liu, R. C. Su, W. J. Yi and Z. G. Zhao, Molecules, 2017, 22, 30 G. Q. Lin, W. J. Yi, Q. Liu, X. J. Yang and Z. G. Zhao,
566–577. Molecules, 2018, 23, 2061–2072.
15 X. F. Hao, Q. Li, H. N. Wang, K. Muhammad, J. T. Guo, 31 W. X. Song, D. A. Gregory, H. Al-janabi, M. Muthana,
X. K. Ren, C. C. Shi, S. H. Xia, W. C. Zhang and Y. K. Feng,
Z. Q. Cai and X. B. Zhao, Int. J. Pharmaceut., 2019, 555,
ACS Biomater. Sci. Eng., 2018, 4, 2870–2878.
322–336.
16 G. Y. Zhang, X. L. Jiang, J. Liu, Q. Z. Yang and R. X. Zhuo, 32 M. Bez, T. J. Kremen, W. Tawackoli, P. Avalos, D. Sheyn,
J. Controlled Release, 2013, 172, E118–E119.
17 L. X. Xie, X. Ding, R. Budry and G. Z. Mao, Int. J. Nanomed.,
2018, 13, 4943–4960.
G. Shapiro, J. C. Giaconi, S. Ben David, J. G. Snedeker,
Z. Gazit, K. W. Ferrara, D. Gazit and G. Pelled, Mol. Ther.,
2018, 26, 1746–1755.
18 H. He, Y. G. Bai, J. H. Wang, Q. R. Deng, L. P. Zhu, 33 C. C. Shi, F. L. Yao, J. W. Huang, G. L. Han, Q. Li, M. Khan,
F. H. Meng, Z. Y. Zhong and L. C. Yin, Biomacromolecules,
2015, 16, 1390–1400.
Y. K. Feng and W. C. Zhang, J. Mater. Chem. B, 2014, 2,
1825–1837.
19 X. M. Li, X. S. Guo, Y. Cheng, X. T. Zhao, Z. W. Fang, 34 A. Meister and M. E. Anderson, Annu. Rev. Biochem., 1983,
Y. L. Luo, S. J. Xia, Y. Feng, J. J. Chen and W. E. Yuan, Front.
Oncol., 2018, 8, 354–365.
52, 711–760.
35 B. G. Hill and A. Bhatnagar, IUBMB Life, 2007, 59, 21–26.
20 G. Fang, F. Zeng, C. Yu and S. Wu, Colloids Surf., B, 2014, 36 H. Sies, Free Radical Biol. Med., 1999, 27, 916–921.
122, 472–481.
37 Y. Zheng, S. Gao and J. Y. Ying, Adv. Mater., 2007, 19,
376–380.
38 D. K. Jha, K. Saikia, S. Chakrabarti, K. Bhattacharya,
K. S. Varadarajan, A. B. Patel, D. Goyary, P. Chattopadhyay
and P. Deb, Mater. Sci. Eng., C, 2017, 72, 415–424.
21 Q. Y. Yu, Y. R. Zhan, J. Zhang, C. R. Luan, B. Wang and
X. Q. Yu, Polymers, 2017, 9, 362–373.
22 Q.-F. Zhang, C.-R. Luan, D.-X. Yin, J. Zhang, Y.-H. Liu,
Q. Peng, Y. Xu and X.-Q. Yu, Polymers, 2015, 7, 2316–2331.
23 G. R. Dhanya, D. S. Caroline, M. R. Rekha and 39 Y. J. Ma, Z. B. Yan, H. Xu and H. C. Gu, Chin. Sci. Bull., 2012,
K. Sreenivasan, Int. J. Biol. Macromol., 2018, 120, 999–1008. 57, 3525–3531.
24 J. Z. Deng, N. N. Gao, Y. A. Wang, H. Q. Yi, S. T. Fang, 40 K. Strojan, J. Lojk, V. B. Bregar, P. Veranic and M. Pavlin,
Y. F. Ma and L. T. Cai, Biomacromolecules, 2012, 13,
3795–3804.
25 R. Devulapally, T. Lee, A. Barghava-Shah, T. V. Sekar, K. Foygel,
Toxicol. In Vitro, 2017, 41, 12–20.
41 H. Aldawsari, R. Edrada-Ebel, D. R. Blatchford, R. J. Tate,
L. Tetley and C. Dufes, Biomaterials, 2011, 32, 5889–5899.
S. V. Bachawal, J. K. Willmann and R. Paulmurugan, Nano- 42 C. Y. Ang, S. Y. Tan, S. Wu, Q. Qu, M. F. E. Wong, Z. Luo,
medicine, 2018, 13, 1051–1066.
P.-Z. Li, S. Tamil Selvan and Y. Zhao, J. Mater. Chem. C, 2016,
4, 2761–2774.
43 K. Mitamura, M. Sogabe, H. Sakanashi, S. Watanabe, T. Sakai,
Y. Yamaguchi, T. Wakamiya and S. Ikegawa, J. Chromatogr. B:
Anal. Technol. Biomed. Life Sci., 2007, 855, 88–97.
26 Y. Zhang, L. Lin, L. Liu, F. Liu, A. Maruyama, H. Tian and
X. Chen, Carbohydr. Polym., 2018, 201, 246–256.
27 X. L. Zhao, Z. Y. Li, H. B. Pan, W. G. Liu, M. M. Lv, F. Leung
and W. W. Lu, Acta Biomater., 2013, 9, 6694–6703.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2019
New J. Chem.