Bioreducible poly (β-amino esters)/shRNA complex nanoparticles for efficient RNA delivery

Article abstract of DOI:10.1016/j.jconrel.2010.12.014

RNA interference (RNAi) mediating gene silencing is a promising approach in the area of gene therapy, but it still is a major challenge to find new non-viral vectors with high transfection efficiency and low toxicity until today. In this work, three novel bioreducible poly (β-amine esters) (PAEs) with different amino monomers in the main chain were designed and synthesized by Michael addition polymerization. All PAEs could condense shRNA into complex nanoparticles with particle size (60-200 nm) and positive surface charges (>+10 mV). The PAEs/shRNA complex nanoparticles (PAENs) were stable under the extracellular physiological condition, while it would degrade in the reductive environment due to the cleavage of the disulfide bonds in the PAEs main chain. PAENs could achieve efficient cellular uptake and EGFP silencing in HEK-293 cells and U-87 MG cells with low cytotoxicity. The high accumulation of PAENs in tumor and high silencing efficiency of intra-tumor EGFP expression occurred when PAENs were intravenously injected into BALB/c mice bearing U-87 MG-GFP tumor. The relationship between the polymer structure and RNAi efficiency and cytotoxicity showed that the density of nitrogen atoms in PAEs backbone and the existence of disulfide bonds demonstrated the remarkable influence on in vitro and in vivo gene silencing efficiency and cytotoxicity. These experimental results suggested that the PAENs could be a promising non-viral vector for efficient RNA delivery.

Full text of DOI:10.1016/j.jconrel.2010.12.014

Journal of Controlled Release 151 (2011) 35–44
Contents lists available at ScienceDirect
Journal of Controlled Release
journal homepage: www.elsevier.com/locate/jconrel
Bioreducible poly (β-amino esters)/shRNA complex nanoparticles for efficient
RNA delivery
⁎
Qi Yin, Yu Gao, Zhiwen Zhang, Pengcheng Zhang, Yaping Li
Center of Pharmaceutics, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China
a r t i c l e i n f o
a b s t r a c t
Article history:
RNA interference (RNAi) mediating gene silencing is a promising approach in the area of gene therapy, but it
still is a major challenge to find new non-viral vectors with high transfection efficiency and low toxicity until
today. In this work, three novel bioreducible poly (β-amine esters) (PAEs) with different amino monomers in
the main chain were designed and synthesized by Michael addition polymerization. All PAEs could condense
shRNA into complex nanoparticles with particle size (60–200 nm) and positive surface charges (N+10 mV).
The PAEs/shRNA complex nanoparticles (PAENs) were stable under the extracellular physiological condition,
while it would degrade in the reductive environment due to the cleavage of the disulfide bonds in the PAEs
main chain. PAENs could achieve efficient cellular uptake and EGFP silencing in HEK-293 cells and U-87 MG
cells with low cytotoxicity. The high accumulation of PAENs in tumor and high silencing efficiency of intra-
tumor EGFP expression occurred when PAENs were intravenously injected into BALB/c mice bearing U-87
MG-GFP tumor. The relationship between the polymer structure and RNAi efficiency and cytotoxicity showed
that the density of nitrogen atoms in PAEs backbone and the existence of disulfide bonds demonstrated the
remarkable influence on in vitro and in vivo gene silencing efficiency and cytotoxicity. These experimental
results suggested that the PAENs could be a promising non-viral vector for efficient RNA delivery.
© 2011 Elsevier B.V. All rights reserved.
Received 20 August 2010
Accepted 24 December 2010
Available online 16 January 2011
Keywords:
Poly (β-amine esters)
Bioreducible
Non-viral vector
RNA delivery
Nanoparticles
1. Introduction
lular environment and fast degradation inside the cell during the
design of polymers.
RNA interference (RNAi) mediated gene silencing is a promising
approach in the area of gene therapy for the treatment of
tumorigenesis and cancer therapy [1,2]. However, many natural
barriers have to be overcome for the safe and efficient delivery
of therapeutic siRNA or short hairpin RNA (shRNA) to the target site
[3–6]. Although non-viral vectors show the potential safety with easy
manufacturing and economic benefits, and attract significant efforts
to investigate and develop this carrier, their low transfection
efficiency represents a major challenge for the clinical applications
[7,8].
Cationic polymers, such as polyethyleneimine (PEI), poly-L-lysine
(PLL), poly (amide amine) (PAMAM) et al., which can spontaneously
bind and condense RNA, have received intense attention due to their
versatile properties and in vitro success in gene expression [9,10].
Unfortunately, most of these cationic polymers were not readily
biodegradable, which induced high toxicity and necrosis [11–13].
Concerning about the safety of the patients, none of above cationic
polymers was used for clinical application up to today. So it is
necessary to give consideration to hydrolytic stability in the extracel-
Poly (β-amino esters) (PAEs) is a kind of polymer synthesized via
Michael addition reactions between compounds containing primary
amine or secondary amine and diacrylates [14]. PAEs, which carry the
positive charge at physiological pH and can condense nucleic acid,
contain biodegradable ester bonds and can degrade into non-toxic
small molecules under basic or acidic conditions. It was reported that
PAEs-modified gold nanoparticles could be used for siRNA delivery
[5], but the degradation of PAEs in the weak acidic environment in
endosome and lysosome is not rapid enough to release RNA into
cytoplasm timely so that the RNAi efficiency was not as high as
expected.
Recently, it has been well-established that the disulfide bonds
tended to be cleaved under the high concentration of reductive
glutathione and thioredoxin reductase in the intracelluar circum-
stance [15–17], which can trigger the degradation of disulfide linkage-
containing polymers inside the cells. Based on this concept, some
bioreducible polymers had been synthesized as RNA carriers, but
these polymers did not show very high in vitro RNAi efficiencies,
which were lower than 80% [18–20], and the in vivo RNAi effects by
systemic administration with other bioreducible polymers were not
reported.
In this work, a series of PAEs with new structure and disulfide
linkages was designed and synthesized. It was the first attempt to
combine poly (β-amino esters) and the bioreducible characteristic of
⁎
Corresponding author. 501 Haike Road, Shanghai 201203, China. Tel./fax: +86 21
2023 1979.
E-mail address: ypli@mail.shcnc.ac.cn (Y. Li).
0168-3659/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2010.12.014

36
Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
disulfide bond for RNA delivery, which could be favorable due to the
rapid degradation of PAEs with prompt RNA release, and was
expected to have positive influence on RNAi efficacy. In addition,
the relationship between the polymer structures and their in vitro and
in vivo RNAi efficiency and cytotoxicity also were investigated.
stirred at 25 °C for 20 h, triethylamine hydrochloride was removed by
filtration, and the solvent was removed by rotary evaporation. The
product was dissolved in dichloromethane (DCM) and washed three
times with an aqueous solution of 0.1 M Na2CO3, then washed twice
with distilled water and once with brine, and dried using anhydrous
Na2SO4 for 1 h. The final DSA product was obtained as a slightly yellow
liquid (2.25 g, 53%) by removing the solvent under reduced pressure.
The synthesis of hexylene glycol diacrylate (HGA) was analogous as
DSA from 1.9 g hexylene glycol (16.2 mmol) and 8.8 g acryloyl
chloride (97.2 mmol). The yield of HGA was 62%.
2. Materials and methods
2.1. Materials
Bis(2-hydroxylethyl) disulfide and acryloyl chloride were purchased
from Alfa Aesar (Tianjin, China). 2-dimethylaminoethylamine
(DMAEA), diethylenetriamine (DETA), and triethylenetetramine
(TETA), hexyleneglycol were obtained from Sinopharm GroupChemical
Reagent Co., Ltd. (China). Trypsin–EDTA and agarose were obtained
from Gibco-BRL (Burlington, ON, Canada). The Dulbecco's modified
Eagle medium (DMEM), YOYO-1, TOTO-3, sodium azide and fetal
bovine serum (FBS) were purchased from Invitrogen GmbH (Karlsruhe,
Germany). Ethidium bromide (EB), 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT), sodium ascorbate, trypan blue,
1,4-dithiothreitol (DTT) and branched polyethyleneimine (PEI, 25 kDa)
were purchased from Sigma (St. Louis, MO). Other chemicals if not
mentioned were of analytical grade.
The pGPU6/Neo-GFP22-shRNA-expressing pDNA that targets the
sequence GCAGCACGACTTCTTCAAG was obtained from GenePharm
Co. Ltd. (Shanghai, China) and purified with the Plasmid Maxi Kit
(Qiagen GmbH, Hilden, Germany) in accordance with the manufac-
turer's instructions. The purity was confirmed by UV spectropho-
tometry (A260/A280), and its concentration was determined from
the absorbance at 260 nm.
Synthesis of poly[bis(2-hydroxylethyl)-disulfide–diacrylate-β-(2-
dimethylaminoethylamine)] (PAA) was carried out by Michael
addition reaction between DSA and DMAEA. Briefly, 0.5 g (1.9 mmol)
of DSA dissolved in 2 mL of anhydrous DCM was added into a solution
of 0.251 g (2.85 mmol) of DMAEA in 2 mL of anhydrous DCM. The
reaction was performed at 45 °C in the dark under nitrogen for 5 days.
The resulted solution was precipitated into 50 mL of hexane under
vigorous stirring. The polymer was dried under vacuum for 1 day at
room temperature and dialyzed against deionized water for 72 h.
Water was removed by rotary evaporation and the polymer was dried
under vacuum over night. The yield of the polymer was 25%.
The synthesis of poly[bis(2-hydroxylethyl)-disulfide–diacrylate-β-
diethylenetriamine] (PAD) and poly[bis(2-hydroxylethyl)-disulfide–
diacrylate-β-triethylenetetramine] (PAT) were analogous as PAA from
0.5 g DSA (1.9 mmol) and 0.294 g DETA (2.85 mmol) or 0.417 g TETA
(2.85 mmol), respectively. The yield of PAD and PAT was 33% and 36%,
respectively.
Poly(hexylene glycol diacrylate-β-triethylenetetramine) (PHT)
was synthesized as control PAEs without disulfide bonds in backbone.
The synthesis of PHT was analogy as PAT from 0.224 g PHT (1.9 mmol)
and 0.417 g TETA (2.85 mmol). The yield of PHT was 29%.
2.2. Cell culture
2.5. Characteristics of PAEs
The cell lines HEK-293 (Human embryonic kidney cell) and U-87
MG (human glioblastoma cell) cells were obtained from the American
Type Culture Collection (ATCC, Manassas, VA). The HEK-293 and U-87
MG cells expressing stable enhanced green fluorescence protein
(HEK-293-GFP and U-87 MG-GFP cells) were kindly provided by
National Laboratory of Oncogenes and Related Genes, Shanghai
Cancer Institute. EGFP-negative cells were cultured in DMEM contain-
ing 10% fetal bovine serum (FBS), 100 Units/mL penicillin G sodium
and 100 μg/mL streptomycin sulfate (complete DMEM medium).
EGFP-positive cells were grown in complete DMEM medium with
50 μg/mL G418. Cells were maintained at 37 °C in a humidified and 5%
CO2 incubator.
The structure of the polymer was confirmed by 1H NMR spectra
recorded on Varian Mercury Plus-400 NMR spectrometer (Varian,
USA) operated at 400 MHz. The molecular weight (Mw) and
polydispersity (Mw/Mn) of PAEs were measured by a gel permeation
chromatography (GPC) system relative to dextran standards
(Mp=4400–401,000 Da, American Polymer Standards Corporation,
USA). GPC was performed on a Waters 2695 controller equipped with
Ultrahydrogel columns (250 PKGD and 1000 PKGD, 30 °C) and a
refractive index detector (model 2414). 0.05% NaN3 aqueous solution
was used as mobile phase with a flow rate of 0.5 mL/min.
The buffer capacity of PAEs was determined by acid–base titration
[22]. An amount equaling to 0.25 mmol of protonable amine groups of
PAEs was dissolved in 5 mL of 150 mM NaCl solution. The pH of the
polymer solution was adjusted with 0.1 M NaOH to 10.0. Then, the
solution was titrated with 0.1 M HCl, and the pH value of the solution
was measured using an acidimeter (Sartorius, Germany). As control,
PEI was titrated using the same method. The buffering capacity was
defined as the percentage of the protonated amine groups from pH 7.4
to 5.0 and calculated according to following equation:
2.3. Animals
Male BALB/c mice aged 6 weeks (20–25 g) were provided by Sino–
British Sippr/BK Lab Animal Ltd. (Shanghai, China) and kept under a
12 h light/dark cycle at the Animal Care Facility. The animals were
given daily fresh diet with free access to water and acclimatized for at
least 5 days prior to the experiments. The in vivo experiments were
carried out under the guideline approved by the Institutional Animal
Care and Use Committee (IACUC) of the Shanghai Institute of Materia
Medica (SIMM), Chinese Academy of Sciences (CAS).
ΔVHCl × 0:1M
Buffer capacityð%Þ =
× 100%
N mol
2.4. Synthesis of poly(β-amino ester)s (PAEs)
wherein ΔVHCl is the volume of 0.1 M HCl titrated to lower the pH
value of the polymer solution from 7.4 to 5.0, N mol is the total moles
of protonable amine groups in the polymers.
To know the hydrolytic degradation of PAEs, 10 mg of PAE was
dissolved in 2 mL PBS (pH 7.4) and incubated at 37 °C. 150 μL of the
solution were taken every 24 h and analyzed by GPC. To estimate the
reductive degradation, 0.35 mg DTT was dissolved in 1 mL PBS and
added to 1 mL PBS containing 10 mg PAE. The resulting solution was
Synthesis of a new series of poly(β-amino ester)s was showed in
Fig. 1A. The disfulfide-based diacrylate (DSA) was synthesized using
Hong's method [21]. Briefly, TEA (5.7 mL, 81 mmol) and bis (2-
hydroxyethyl) disulfide (2.5 g, 16.2 mmol) were added to a 100 mL
flask containing 30 mL of anhydrous THF. Then the flask was
immersed in an ice bath, and acryloyl chloride (8.8 g, 97.2 mmol)
was added dropwise. After the resulting heterogeneous mixture was

Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
37
Fig. 1. (A) Synthesis of PAEs: (a) PAA, (b) PAD, (c) PAT and (d) PHT. (B)1H NMR spectra (400 MHz) of (a) PAA (D2O), (b) PAD (DMSO-d6), (c) PAT (D2O). (C) Titration curves of PAEs
and PEI in 150 mM NaCl aqueous solution with 0.1 M HCl.
homogenized and incubated at 37 °C. 150 μL of the solution were
2.7. In vitro dissociation of PAENs
taken every 5 min and analyzed by GPC.
To investigate the hydrolytic dissociation, 100 μL phosphate buffer
solution (PBS, pH 7.4) with shRNA (20 μg) was added into 400 μL PBS
with different concentrations of bioreducible PAEs, respectively,
followed by vortexing for 30 s at room temperature to prepare
PAENs with N/P ratio of 5, 10 and 20. The solution was stirred at 37 °C
for 24 h, applied onto 1.0% agarose gel and subjected to electropho-
resis for 45 min at 80 V.
To investigate the reductive dissociation, PAENs were prepared as
described above. After 2.5 μmol DTT was added, the solution was
stirred at 37 °C for 1 h and performed for agarose gel electrophoresis.
PHT/RNA complex nanoparticles (PHTNs) with N/P ratio of 20 and PEI
25 kDa/RNA complex nanoparticles (PEINs) with N/P ratio of 5 were
used in the dissociation experiment as control.
2.6. Preparation and characterization of PAE/RNA complex nanoparticles
(PAENs)
PAENs were prepared by following procedure, briefly, 120 μL
distilled water containing different concentrations of PAEs was added
to 30 μL distilled water with GFP-shRNA (6 μg) to obtain the desired
polymer nitrogen to shRNA phosphate (N/P) ratio, followed by
vortexing for 30 s and incubated at room temperature for 30 min.
The formation of PAENs was confirmed by agarose gel retardation.
PAENs containing 0.4 μg shRNA at various N/P ratios were loaded on
1.0% agarose gel in Tris-Acetate-EDTA (TAE) buffer containing 0.5 μg/
mL EB with loading buffer and subjected to electrophoresis for 45 min
at 80 V. The resulting shRNA migration pattern was visualized with a
UV illuminator.
2.8. Celluar uptake experiment
The particle size and zeta-potential of PAENs were determined by
the laser light scattering measurement using a Nicomp 380/ZLS zeta
potential analyzer (Particle Sizing System, USA).
HEK-293 cells and U-87 MG cells were seeded in 24-well plates at
a density of 1×105 cells per well in 500 μL of complete DMEM
medium and incubated for 24 h. shRNA was fluorescently labeled

38
Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
with a nucleic acid stain YOYO-1 by mixing shRNA and YOYO-1 with a
ratio of 1 nM dye molecule/μg shRNA and incubating for 1 h at room
temperature in the dark. The polymers/YOYO-1-labeled shRNA
complex nanoparticles were prepared as described for PAENs
above at N/P ratio 10 for PAEs and N/P ratio 5 for PEI 25 kDa, and
added onto the cells at a shRNA concentration of 2.5 μg/well for
different incubation time with 40 min, 100 min and 220 min at 37 °C
followed by adding Hoechst 33342 (6 μg/mL, Invitrogen, USA) and
Lyso Tracker Red DND-99 (10 nmol/mL, Invitrogen, USA) to the cells,
and incubation at 37 °C for additional 20 min. The extracellular
fluorescence was quenched with 0.4% trypan blue for 2 min [23].
After the medium was removed, cells were washed three times with
PBS and visualized under a fluorescence microscope (Olympus,
Japan) to observe the internalization of the nanoparticles. Cells were
trypsinized and resuspended in the medium. Subsequently, the cells
were washed three times with PBS, and the intracellular fluorescence
was determined using a FACSCalibur system (Becton Dickinson,
USA).
complex nanoparticles as described in Section 2.8 at N/P ratio 20 for
PAENs and N/P ratio 5 for PEINs.
BALB/c mice bearing U-87 MG-GFP tumors were intravenously
injected with TOTO-3 labeled PAENs or PEINs at a dose of 12.5 μg
shRNA per mouse (in 200 μL PBS) via the tail vein. At 1 h after
injection, the mice were sacrificed. The tissue samples including heart,
liver, spleen, lung, kidney and tumor were taken out, washed with
cold saline and observed using a Maestro™ 2 in vivo fluorescence
imaging system (CRi, USA).
2.10.2. RNAi experiment in mice
PAENs at N/P ratio of 20 or PEINs at N/P ratio of 5 were
intravenously injected at a dose of 12.5 μg shRNA per mouse (in
200 μL PBS), and 200 μL saline was injected as control. After 48 h, the
mice were sacrificed and the tumors were taken out, washed with
cold saline and imbedded in OCT. The frozen sections of the tumors
were prepared and visualized under the fluorescence microscope. For
quantitative determination of the gene silencing efficiency, the
tumors were homogenized with PBS (pH 7.4). The homogenates
were filtered through nylon membranes, centrifuged at 2000×g for
5 min, washed twice with PBS and resuspended in 0.5 mL PBS. The
fluorescent intensity of the cells was measured by FACS.
2.9. In vitro RNAi and cell viability assays
RNAi experiment was performed with HEK-293-GFP cells and U-
87 MG-GFP cells by using the GFP shRNA as reporter gene. For
silencing experiment, the cells were seeded in 24-well plates at a
density of 1×105 cells per well in 500 μL of DMEM medium and
allowed to attach for 24 h prior to transfection. Then, the media was
replaced with fresh medium containing PAENs with 2.5 μg/well of
shRNA at various N/P ratios, and the cells were incubated for an
additional 48 h. The efficiency of silencing was visualized under a
fluorescence microscope and measured by FACS. As positive control,
PEI 25 kDa/shRNA complexes nanoparticles (PEINs) with N/P ratio of
5 were used and examined as described above. All RNAi experiments
were performed in triplicate.
2.11. Statistical analysis
Student's t-test (two-tailed) was applied to test the significance of
the difference between two groups. Differences were considered
statistically significant when pb0.05.
3. Results and discussion
3.1. Synthesis and characterization of PAEs
Cells were seeded in 96-well plates at a density of 1×104 cells per
well in 200 μL of DMEM medium and incubated overnight. After the
media was replaced with fresh medium, PAENs with 1 μg/well of
shRNA at various N/P ratios was added onto the cells except for those
used as control and incubation was performed for 24 h. 20 μL/well of
MTT solution (5 mg/mL in PBS) was added and followed by additional
4 h incubation. Then, the medium was replaced by 200 μL of DMSO,
and the formazan was dissolved sufficiently by mildly shaking the
plates for 10 min. The absorbance was measured at 570 nm on
microplate reader (Bio-RAD, model 550). The MTT value for untreated
cells was taken as 100% cell viability. Experiments were performed in
sextuple. In the study on RNAi and cell viability, PHTNs with N/P ratio
of 20 and PEINs with N/P ratio of 5 were used and examined as
described above as non-bioreducible control and positive control,
respectively.
To evaluate the cytotoxicity of PAENs at higher concentrations,
cells were seeded in 96-well plates at a density of 1×104 cells per well
in 200 μL of DMEM medium and incubated overnight. After the media
was replaced with fresh medium, PAENs with shRNA at various final
concentrations at N/P ratio of 20 was added onto the cells except for
those used as control, followed by incubation for 24 h. MTT assays
were performed similarly to the experiments mentioned above.
PHTNs with N/P ratio of 20 and PEINs with N/P ratio of 5 were also
used as control.
Three different bioreducible poly(β-amino ester)s with various
numbers of nitrogen atoms in the repeating units were synthesized
via Michael addition reaction between DSA and DMAEA, DETA or TETA
(Fig. 1A). The Michael-type polymerization can directly accomplish by
a single synthetic step with simple purification because of no
byproducts during polymerization. The polymerization reactions
were performed for 5 days by DSA and 1.5-fold molar amine
monomers. The diversity of density of nitrogen atoms in PAEs
backbone was obtained through the variation of amine monomers.
The excess of amine monomers was necessary to ensure that the PAEs
were amino-ended so as to possess good water-solubility and low
toxicity. The 1H NMR spectra of the polymers in D2O for PAA, PAT and
in DMSO-d6 for PAD were in accordance with the expected structures
(Fig. 1). The methylene protons close to disulfide were present at
δ=2.65–3.13 ppm. The presence of secondary amines and tertiary
amines was confirmed by the signal at δ=1.59–1.80 ppm for amine
protons and δ=2.40–2.92 ppm for methylene protons, respectively.
The signals at δ=4.43–4.65 ppm and 2.20–3.00 ppm corresponded to
the methylene protons close to the oxygen atoms and carbonyl of the
ester bonds, respectively. For all polymers, no signals appeared
between δ=5 and δ=7 ppm, assigned to the acryl group, which
indicated that these copolymers were end-capped with amino groups.
The molecular weight and polydispersity of PAEs measured by GPC
were 10.5 kDa with Mw/Mn 1.87 for PAA, 9.8 kDa with Mw/Mn 1.45
for PAD, 9.9 kDa with Mw/Mn 1.59 for PAT, respectively.
2.10. In vivo experiment
The buffer capacity of cationic polymeric gene vectors has been
implicated as a factor influencing the escape of nucleic acid from
endosomes following endocytosis [24]. The buffer capacity of PAEs,
defined as the percentage of amine groups being protonated on
decrease of pH, was determined from the acid–base titration curves
ranging from pH 7.4 to pH 5.0. All PAEs showed buffer capacity
ranging from 11.6% to 13.6%, slightly lower than that of PEI (14.4%)
(Fig. 1C), which demonstrated that the buffer capacity was improved
2.10.1. Biodistribution in BALB/c mice
Male BALB/c mice aged 6 weeks were inoculated subcutaneously
in the right flank with 2×106 U-87 MG-GFP cells. All animal
experiments were performed when the tumor reached 150–
200 mm3. The polymer/TOTO-3-labeled shRNA complex nanoparti-
cles were prepared similarly to the polymer/YOYO-1-labeled shRNA

Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
39
with the densities of nitrogen atoms in the polymers backbone
increasing.
overall efficiency of RNAi [25]. To assess the release of RNA from
complexes, the PAENs were incubated in PBS (pH7.4) in the presence
or absence of DTT, which simulated intracellular and extracellular
redox environment. The reductive potential-triggered release of RNA
from disulfide-containing PAENs (SS-PAENs including PAANs, PADNs
and PATNs) was visualized by gel electrophoresis (Fig. 2C, a–c). For
PAENs incubated without DTT, the fluorescence of the bands for
PADNs and PATNs was barely detected, which could result from PAEs'
tightly condensation of RNA so that the binding between EB and RNA
was weakened. For PAANs, the RNA migration was also retarded,
which suggested that PAENs were stable in the extracellular
physiological condition. The migration of the bands for SS-PAENs
incubated with DTT indicated that RNA was released due to the
breaking of disulfide bonds under the reductive conditions. As a result,
it could be concluded that SS-PAENs would degrade after being
internalized in the reductive intracellular environment and release
their cargo efficiently. In contrast, bands for PHTNs and PEINs did not
show migration when incubated with or without DTT (Fig. 2C, d),
which revealed that they should be stable under extracellular
physiological and intracellular reductive condition.
To determine the profiles of the hydrolytic and reductive
degradation, PAEs were incubated in PBS (pH 7.4) with or without
the reduction agent DTT at 37 °C, which simulated intracellular or
physiological conditions, respectively. The result showed that the Mw
of PAEs decreased extremely slowly with only 3.8% after 4 days and
10% after 20 days, respectively, when PAEs were incubated in
physiological environment. On the contrary, when the PAEs were
incubated with DTT, the Mw decreased very rapidly so that PAEs could
not be detected after 5 min, which revealed that PAEs was very
sensitive under the reductive circumstance, but relatively stable in
physiological condition.
3.2. Physicochemical characteristics of PAENs
RNA must be compacted into stable nanosized complexes for
efficient delivery. Gel retardation assay was performed to analyze
binding behavior between PAEs and shRNA at different N/P ratios
from 5 to 50. As shown in Fig. 2A, PAA demonstrated a relatively weak
RNA condensing capability, and complete retardation was observed at
N/P ratio of 20 (Fig. 2A, a). Total retardation of complexes was
detected for both PAD/shRNA complexes (PADNs) and PAT/shRNA
complexes (PATNs) at N/P ratios of 5 (Fig. 2A, b–c), which indicated
that PAES with higher densities of nitrogen atoms exhibited enhanced
capability to condense shRNA.
Dynamic light scattering and zeta potential measurements showed
that PAEs could condense shRNA into polyplexes with small particle
size (60–200 nm) with positive surface charges (N+10 mV) except
for PAA/shRNA complexes (PAANs) at N/P ratio of 5 (Fig. 2B). The
surface charge increased with the increase in density of nitrogen
atoms of the polymer, while the particle size decreased, which
presented that the condensation ability of PAEs showed the following
tendency: PATNsNPADNsNPAANs.
3.3. Celluar uptake of PAENs
The cellular uptake of PAENs was performed using shRNA labeled
with YOYO-1 on HEK-293 cells and U-87 MG cells. YOYO-1 has been
used in cellular uptake as an intercalating dye with high affinity to
DNA [20], and Hoechst 33342 and Lyso Tracker Red DND-99 were
used for staining the nucleus and lysosomes, respectively. HEK293
cells incubated with PAENs containing YOYO-1-labeled shRNA for 2 h
demonstrated gradual increase of green fluorescence in cytoplasm
with the density of nitrogen atoms in PAEs backbone increasing
(Fig. 3A, a–c). There was little fluorescence detected in PAANs treated
cells, while fluorescence was distributed in the cytoplasmic com-
partment in PADNs and PATNs treated cells. This distributive
difference was also observed in U-87 MG cells (Fig. 3B, e–g). PHTNs
and PEINs with YOYO-1-labeled shRNA were used as control, and the
RNA dissociation was one of the rate-limiting steps during RNA
delivery, and the efficiency of dissociation was closely related to the
Fig. 2. (A) Gel retardation electrophoresis of PAEs/shRNA complex nanoparticles. (a) PAA, (b) PAD, (c) PAT. Lane 1: naked shRNA, Lanes 2–7, polyplexes prepared at N/P ratio of 5, 10,
20, 30, 40, 50, respectively. (B) Mean particle size (a) and zeta potential (b) of PAENs. (C) Agarose gel electrophoresis of PAANs (a), PADNs (b) and PATNs (c) after incubated without
or with DTT. Lane 1: naked shRNA; Lanes 2–4, PAENs prepared at N/P ratios of 5, 10 and 20, respectively, incubated in PBS with pH7.4 at 37 °C for 24 h without DTT. Lanes 5–7, PAENs
prepared at N/P ratios of 5, 10 and 20, respectively, incubated in PBS with pH7.4 at 37 °C for 1 h with DTT. PHTNs and PEINs were used as control (d), Lane 1: naked shRNA; PHTNs
(Lane 2) prepared at N/P ratios of 20 and PEINs (Lane 4) prepared at N/P ratio of 5, respectively, incubated in PBS with pH7.4 at 37 °C for 24 h without DTT. PHTNs (Lane 3) prepared
at N/P ratios of 20 and PEINs (Lane 5) prepared at N/P ratio of 5, respectively, incubated in PBS with pH7.4 at 37 °C for 1 h with DTT.

40
Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
Fig. 3. Cellular uptake of PAENs. Fluorescent images of cellular uptake of PAANs (a, f), PADNs (b, g), PATNs (c, h), PHTNs (d, i), and PEINs (e, j) in HEK-293 cells (A) and U-87 MG cells
(B). Higher magnification images at different time points for cellular uptake of PAENs (C).(Green: PAENs with YOYO-1 labeled shRNA; red: Lyso Tracker Red stained lysosomes; blue:
Hoechst 33342 stained nucleus.) Quantitative analysis of cellular uptake of PAENs and PEINs for HEK-293 cells (D) and U-87 MG cells (E) using FACSCalibur. The FCM pictures of
control cells (a, black), cells treated with PAANs (b, green), PADNs (c, red), PATNs (d, blue), PEINs (e, purple) and PHTNs (f, yellow) of collected HEK-293 cells (F) and U-87 MG cells
(G) were shown.
intracellular fluorescence of the cells treated with PHTNs and PEINs
was comparable to that of PATNs counterpart (Fig. 3A d, e, B i, j). In
40-fold magnification images (Fig. 3C), it could be found that the
uptake of all polymers/YOYO-1-labeled shRNA complex nanoparti-
cles were hard to be seen at 1 h. After incubation for 2 h, more
nanoparticles could be detected, but most of them adhered on the cell
membrane. At 4 h, large numbers of nanoparticles could be seen
inside the cells. Though some had co-localization with the lysosomes,
most nanoparticles either dispersed in the cytoplasm or co-localized
with the nucleus, which suggested a rapid escape from endosomes.
Quantitative cellular uptake was estimated using a FACSCalibur,
which further confirmed the result observed by fluorescent micro-
scope (Fig. 3D–G). The experimental results revealed that the
electrostatic interaction between positive charged polymers and
negative charged cell membrane played a prevailing role in the
internalization of PAENs, which was inferred from the phenomenon
that the cellular uptake of PAENs was enhanced while their surface
positive charge was stronger, and the amounts of RNA delivered and
internalized into cells by PAENs were less than that by PEINs, which
possessed the higher density of positive charge.

Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
41
3.4. Silencing of EGFP expression in vitro and cell viability assays
low toxicity was contributed to the cleavage of the disulfide bonds
so that the PAEs could be rapidly degraded in the intercellular
circumstance.
The in vitro RNAi efficiency was detected by model system of
shRNA-based reporter EGFP knock-down on HEK-293 cells and U-87
MG cells, and the cell viability was assessed by MTT assays.
Transfection efficiency was evaluated at N/P ratios of 5, 10 and 20,
respectively, and PEINs at N/P ratio of 5 were used as comparison.
Fig. 4A and B illustrated that the stable expressed EGFP in both cells
was scarcely decreased after treated with PAANs, but obviously
decreased when treated with PADNs or PATNs. HEK-293 cells
transfected with PEINs showed comparable EGFP intensity with that
of PADNs, while in U-87 MG cells, EGFP was barely expressed.
The silencing efficiency measured by FACSCalibur increased with
N/P ratios increasing, which showed a similar result as above (Fig. 4C
and D). For HEK-293 cells, silencing achieved with PAANs was only
10% at N/P ratios of 5 and 10 due to poor RNA condensation, but 72%
silencing efficiency occurred at N/P ratio of 20 because RNA could be
completely condensed at this ratio and showed a higher cellular
uptake. Although the cellular uptake of PAENs was lower than that of
PEINs, a relative high silencing efficiency of 88%–97% resulted from
PADNs and PATNs, which were higher than that from PEINs (87%).
These results could attribute to the reductive triggered degradation of
disulfide-containing PAEs (SS-PAEs including PAA, PAD and PAT) and
subsequent dissociation of SS-PAENs in the cellular interior, which
helped RNA to be released in time. For U-87 MG cells, silencing
efficiency ranged from 60% to 96%, which also appeared that PAEs
with more positive surface charges brought about higher silencing
efficiency, which was consistent with the results of cellular uptake. In
both cell lines, the interference efficiency of PHTNs were lower than
those of PADNs and PATNs, which suggested that the reductive
cleavage of disulfide bonds indeed had a positive influence on gene
delivery process.
The concentration of shRNA used in in vitro RNAi experiments was
5×10−3 mg/mL. Correspondingly, the concentration of PAEs was
rather low even at N/P ratio of 20. To obtain a more convincing data
about the cytotoxicity of PAENs, cell viability was determined after the
cells were treated with higher concentrations of PAENs. In both kinds
of cells, the cell viability was more than 70% with PAENs at all tested
concentrations. Cells treated with non-bioreducible PHTNs and non-
degradable PEINs at the lowest concentration showed the cell viability
of 70% and 50–60%, respectively, and the data decreased to 40%–55%
and 27%–31% at the highest concentration. These results further
confirmed that SS-PAENs had the advantage with much lower toxicity
than PHTNs and PEINs.
3.5. Biodistribution in BALB/c mice
The tissue distribution of PAENs and PEINs with TOTO-3 labeled
shRNA in BALB/c mice bearing U-87 MG-GFP tumors at 1 h after
intravenous injection was presented in Fig. 6. All of the three kinds of
PAENs and PEINs localized principally in the liver. The accumulation of
the polymer/shRNA polyplexes, which was hardly detected in the
spleen and heart, was mainly distributed to the liver and kidney. The
accumulation of PADNs and PATNs in the tumor was comparable to
and slightly less than that in lung and kidney. The concentration of
PAANs in tumor was much lower than that of PADNs and PATNs.
The enhanced permeability and retention (EPR) effect combined
with the 60–200 nm nanoparticle sizes that exceed the 5 nm cutoff for
clearance by the kidney could potentially explain the preferential
accumulation of the PADNs in the tumor with lower concentration in
the kidney as observed in systematic administration [26]. The amount
of PATNs accumulated in the tumor was slightly less than that in the
kidney. PEINs showed a little lower accumulation in the tumor than
PATNs, which could be on account of the excessively high positive
charge of PEINs so that led to their interaction with plasma proteins
and cleared by mononuclear phagocyte system.
The cytotoxicity profiles of the polymers and the polyplexes were
shown in Fig. 5. In MTT assays, only 70% and 50% of cells survived
when treated with PHTNs and PEINs, respectively, while SS-PAENs
showed essentially low cytotoxicity with cell viabilities of approxi-
mately 80%–98% at all N/P ratios, and SS-PAEs showed higher cell
viabilities than PHT and PEI25K, which could be speculated that the
Fig. 4. In vitro RNAi experiment of PAENs. Fluorescent pictures of RNAi of PAANs (a, f), PADNs (b, g), PATNs (c, h), PHTNs (d, i) at N/P ratio of 20 and PEINs (e, j) in HEK-293-GFP cells
(A) and U-87 MG-GFP cells (B). EGFP silencing efficiency of PAENs at different N/P ratios in HEK-293-GFP cells (C) and U-87 MG-GFP cells (D). PHTNs at N/P ratio of 20 and PEINs at
N/P ratio of 5 were used for comparison. ( pb0.05 vs PAANs group, pb0.05 vs PADNs group, #pb0.05 vs PHTNs.).
⁎
⁎⁎

42
Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
Fig. 5. Cell viability determined by MTT assay for PAANs, PADNs, PATNs, bioreducible PAEs (SS-PAEs including PAA, PAD and PAT), PHTNs at N/P ratio of 20, PHT, PEINs at N/P ratio of
⁎
⁎⁎
5 and PEI25K in HEK-293-GFP cells (A) and U-87 MG-GFP cells (B). ( pb0.05 vs PATNs group, pb0.05 vs PHTNs group) and cell viability determined by MTT assay for PAANs,
PADNs, PATNs, PHTNs and PEINs at higher concentrations in HEK-293-GFP cells (C) and U-87 MG-GFP cells (D). The N/P ratio were 20 for ss-PAENs and PHTNs and 5 for PEINs.
3.6. In vivo RNA interference
The pictures for the frozen sections of tumors in Fig. 7A qualitatively
showed the in vivo RNAi effect after BALB/c mice bearing U-87 MG-GFP
tumors were injected with PAENs and PEINs. Although PAANs showed
somewhat weak interference efficiency, PADNs and PATNs brought about
marked decrease of EGFP expression in tumors. PADNs demonstrated
comparable effect to PEINs on EGFP silencing, PATNs revealed even higher
interference efficiency, and both displayed higher interference efficiency
than PHTNs. This result could come from the higher accumulation of
PAENs in tumor and the sufficient release of RNA into cytoplasm due to the
bioreducible disulfide bonds. The quantitative result of RNAi in vivo
corresponded with the one in vitro (Fig. 7B), which implied that the
density of nitrogen atoms in PAEs backbone was also important for the
gene silencing efficiency in vivo. Although very little amount of shRNA
was recognized in tumor (Fig. 6), EGFP expression in tumor was
significantly inhibited by intravenous administration of the PAEs
complexes, especially, PATNs completely silenced EGFP expression in
tumor (Fig. 7B). The discrepancy between biodistribution and silencing
effect of PAEs complexes could result from that the total amount of
injected shRNA was superfluous, and the amount accumulated in the
tumor was enough to obtain high RNAi efficiencies in the tumor.
4. Conclusions
Three kinds of bioreducible disulfide linkages-containing PAEs
with different amino monomers as polymeric nonviral RNA vector
were designed and synthesized in this work. PAEs could tightly
condense shRNA into complex nanoparticles with positive surface
charges. The cellular uptake and RNAi efficiency of PAENs were
remarkably affected by the density of nitrogen atoms in PAEs
backbone. The relationship between the polymer structure and
Fig. 6. Biodistribution of TOTO-3 labeled PAANs (b), PADNs (c), PATNs (d) and PEINs (e)
in BALB/c mice bearing U-87 MG-GFP tumors. Lanes 1–6: tumor, lung, spleen, kidney,
heart, and liver, respectively. Saline was applied for control (a).

Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
43
Fig. 7. In vivo RNAi experiment of PAENs. (A) Frozen sections of tumors taken out from BALB/C mice treated with saline (a: under fluorescence and b: under visible light), PAANs (c),
PADNs (d), PATNs (e), PHTNs (f) and PEINs (g). (B) Quantitative analysis of intra-tumor GFP silencing efficiency in BALB/c mice bearing U-87 MG-GFP tumor of PAENs by FACS. The
FCM pictures of cells collected from tumors in mice treated with saline (a, black), PAANs (b, green), PADNs (c, red), PATNs (d, blue), PEINs (e, purple) and PHTNs (f, yellow) were
⁎
⁎⁎
shown (C). ( pb0.05 vs PHTNs groups, pb0.05 vs PEINs group).
RNAi efficiency and cytotoxicity showed that PAENs could efficiently
and safely achieve EGFP silencing in HEK-293 cells and U-87 MG cells,
and obtain higher level of intra-tumor gene silencing efficiency in
BALB/c mice bearing U-87 MG-GFP tumors than PEINs. The cleavage
of disulfide bonds in reductive environment initiated the degradation
of PAEs and rapid release of RNA, which contributed to the high RNAi
efficiency and low cytotoxicity of PAENs. The density of nitrogen
atoms in PAEs backbone and the existence of disulfide bonds in
polymers showed the remarkable influence on the gene silencing
efficiency in vitro and in vivo. It was proved that the combination of
poly (β-amino esters) and the bioreducible property of disulfide
bond could bring the advantages of both into play and achieve very
high in vitro and in vivo RNAi efficiency (more than 90%), which were
more superior than the present other bioreducible polymers/RNA
complexes. These experimental results suggested that PAENs could
be a promising non-viral vector for efficient RNA delivery.
“Key New Drug Creation and Manufacturing Program” (No.
2009ZX09501-024 and 2009ZX09301-001), and Shanghai Nanomedicine
Program (1052nm06300) are gratefully acknowledged for financial
support.
References
[1] W.M. Merritt, M. Bar-Eli, A.K. Sood, The dicey role of dicer: implications for RNAi
therapy, Cancer Res. 70 (2010) 2571–2574.
[2] Y. Arthanari, A. Pluen, R. Rajendran, H. Aojula, C. Demonacos, Delivery of
therapeutic shRNA and siRNA by Tat fusion peptide targeting bcr–abl fusion gene
in Chronic Myeloid Leukemia cells, J. Control. Release 145 (2010) 272–280.
[3] H.Q. Cao, X. Jiang, C. Chai, S.Y. Chew, RNA interference by nanofiber-based siRNA
delivery system, J. Control. Release 144 (2010) 203–212.
[4] M. Andersen, A. Lichawska, A. Arpanaei, S.M.R. Jensen, H. Kaur, D. Oupicky, F.
Besenbacher, P. Kingshott, J. Kjems, K.A. Howard, Surface functionalisation of
PLGA nanoparticles for gene silencing, Biomaterials 31 (2010) 5671–5677.
[5] J.S. Lee, J.J. Green, K.T. Love, J. Sunshine, R. Langer, D.G. Anderson, Gold, poly(beta-
amino ester) nanoparticles for small interfering RNA delivery, Nano Lett. 9 (2009)
2402–2406.
[6] M.S. Huh, S.Y. Lee, S.J. Park, S.K. Lee, H.J. Chung, S.J. Lee, Y.S. Choi, Y.K. Oh, J.H. Park,
S.Y. Jeong, K.W. Choi, K.M. Kim, I.C. Kwon, Tumor-homing glycol chitosan/
polyethylenimine nanoparticles for the systemic deliverynext term of siRNA in
tumor-bearing mice, J. Control. Release 144 (2010) 134–143.
[7] X. Zhang, A. Kovtun, C.M. Palomares, M. Oulad-Abdelghani, F. Fioretti, S.
Rinckenbach, D. Mainard, M. Epple, N. Benkirane-Jessel, SiRNA-loaded multi-
Acknowledgement
The National Basic Research Program of China (2010CB934000 and
2007CB935800), the National Natural Science Foundation of China
(30925041, 30873169), National Science & Technology Major Project

44
Q. Yin et al. / Journal of Controlled Release 151 (2011) 35–44
shell nanoparticles incorporated into a multilayered film as a reservoir for gene
silencing, 10.1016/j.biomaterials.2010.04.024.
previous term gene delivery next term system for tumor targeted transfection,
Biomaterials 31 (2010) 4771–4780.
[8] I.D. Kim, C.M. Lim, J.B. Kim, H.Y. Nam, K. Nam, S.W. Kim, J.S. Park, J.K. Lee,
Neuroprotection by biodegradable PAMAM ester (e-PAM-R)-mediated HMGB1
siRNA delivery in primary cortical cultures and in the postischemic brain, J.
Control. Release 142 (2010) 422–430.
[9] T. Mori, A. Ishikawa, Y. Nemoto, N. Kambe, M. Sakamoto, Y. Nakayama,
Development of a novel nonviral gene silencing system that is effective both in
vitro and in vivo by using a star-shaped block copolymer (star vector), Bioconjug.
Chem. 20 (2009) 1262–1269.
[10] H.J. Kim, A. Ishii, K. Miyata, Y. Lee, S.R. Wu, M. Oba, N. Nishiyama, K. Kataoka,
Introduction of stearoyl moieties into a biocompatible cationic polyaspartamide
derivative, PAsp(DET), with endosomal escaping function for enhanced siRNA-
mediated gene knockdown, J. Control. Release 145 (2010) 141–148.
[11] K. Itaka, T. Ishii, Y. Hasegawa, K. Kataoka, Biodegradable polyamino acid-based
polycations as safe and effective gene carrier minimizing cumulative toxicity,
Biomaterials 31 (2010) 3707–3714.
[18] H.Y. Nam, A. McGinn, P.H. Kim, S.W. Kim, D.A. Bull, Primary cardiomyocyte-
targeted bioreducible polymer for efficient gene delivery to the myocardium,
Biomaterials 31 (2010) 8081–8087.
[19] S.H. Kim, M. Ou, D.A. Bul, S.W. Kim, Reductive degradation behavior of
bioreducible poly(disulfide amine) for enhancing SiRNA efficiency, Macromol.
Biosci. 10 (2010) 898–905.
[20] M.I. Nounou, K. Emmanouil, S. Chung, T. Pham, Z.Y. Lu, M. Bikram, Novel reducible
linear L-lysine-modified copolymers as efficient nonviral vectors, J. Control.
Release 143 (2010) 326–334.
[21] C.Y. Hong, Y.Z. You, D.C. Wu, Y. Liu, C.Y. Pan, Thermal control over the topology of
cleavable polymers: from linear to hyperbranched structures, J. Am. Chem. Soc.
129 (2007) 5354–5355.
[22] C. Lin, Z.Y. Zhong, M.C. Lok, X.L. Jiang, W.E. Hennink, J. Feijen, J.F.J. Engbersen,
Novel bioreducible poly(amino amine)s for highly efficient gene delivery,
Bioconjug. Chem. 18 (2007) 138–145.
[12] D. Jere, H.L. Jiang, R. Arote, Y.K. Kim, Y.J. Choi, M.H. Cho, J.S. Park, J.K. Lee,
Degradable polyethylenimines as DNA and small interfering RNA carriers, Expert
Opin. Drug Deliv. 6 (2009) 827–834.
[13] M.I. Nounou, K. Emmanouil, S. Chung, T. Pham, Z.Y. Lu, M. Bikram, Novel reducible
linear L-lysine-modified copolymers as efficient nonviral vectors, J. Control.
Release 143 (2010) 326–334.
[14] D.M. Lynn, D.G. Anderson, D. Putnam, R. Langer, Accelerated discovery of
synthetic transfection vectors: parallel synthesis and screening of a degradable
polymer library, J. Am. Chem. Soc. 123 (2001) 8155–8156.
[15] S. Son, K. Singha, W.J. Kim, Bioreducible BPEI-SS-PEG-cNGR polymer as a tumor
targeted nonviral genenext term carrier, 1016/j.biomaterials.2010.04.047.
[16] Y.W. Won, H.A. Kim, M. Lee, Y.H. Kim, Reducible poly(oligo-d-arginine) for
enhanced gene expression in mouse lung by intratracheal injection, Mol. Ther. 18
(2010) 734–742.
[23] M. Hashimoto, M. Morimoto, H. Saimoto, Y. Shigemasa, T. Sato, Lactosylated
chitosan for DNA delivery into hepatocytes: the effect of lactosylation on the
physicochemical properties and intracellular trafficking of pDNA/chitosan
complexes, Bioconjug. Chem. 17 (2006) 309–316.
[24] O. Boussif, F. Lezoualc'h, M.A. Zanta, M.D. Mergny, D. Scherman, B. Demeneix, J.P.
Behr, A versatile vector for gene and oligonucleotide transfer into cells in culture
and in vivo: polyethylenimine, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 7297–7301.
[25] M. Köping-Höggård, K.M. Vårum, M. Issa, S. Danielsen, B.E. Christensen, B.T. Stokk,
P. Artursson, Improved chitosan-mediated gene delivery based on easily
dissociated chitosan polyplexes of highly defined chitosan oligomers, Gene
Ther. 11 (2004) 1441–1452.
[26] A.S. Paraskar, S. Soni, K.T. Chin, P. Chaudhuri, K.W. Muto, J. Berkowitz, M.W.
Handlogten, N.J. Alves, B. Bilgicer, D.M. Dinulescu, R.A. Mashelkare, S. Sengupta,
Harnessing structure–activity relationship to engineer a cisplatin nanoparticle for
enhanced antitumor efficacy, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 12435–12440.
[17] X. Zeng, Y.X. Sun, W. Qu, X.Z. Zhang, R.X. Zhuo, Biotinylated transferrin/avidin/
biotinylated disulfide next term containing PEI bioconjugates mediated p53