The role of disulfide-bridge on the activities of H-shape gemini-like cationic lipid based siRNA delivery

Article abstract of DOI:10.1016/j.jconrel.2016.05.051

In our previous study, a H-shape gemini-like cationic lipid (ssGLCL, formerly named as CLD), composed of two hydrophilic lysine heads and two hydrophobic oleyl alcohol tails with a bridge of the redox-active disulfide-bond, had been synthesized and used as a nanocarrier for delivering small interfering RNAs (siRNAs) into cells. In order to further elucidate the role of disulfide ( - S - S - ) bridge on the activity of ssGLCL based siRNA delivery, a comparable ccGLCL bridged with a non-reducible carbon-carbon bond was synthesized and used as control in this study. Both two H-shape GLCL molecules could individually self-assemble into cationic nanoparticles in water phase and complex with negatively-charged siRNA into nanoplexes with particle size of ~ 200 nm and zeta potential of ~ + 30 mV, and exhibit effective siRNA delivery both in vitro and in vivo. Investigation of internalization pathway displayed that both ssGLCL/siRNA and ccGLCL/siRNA nanoplexes were predominantly internalized into MCF-7 cells by the clathrin-mediated endocytosis pattern. Although a lower cellular uptake of siRNA was found in the human breast cancer MCF-7 cells, the ssGLCL/siRNA nanoplexes could exhibit similar or even stronger down-regulation effects on the targeted EGFR mRNA and protein in MCF-7 cells when compared to the ccGLCL/siRNA nanoplexes. Furthermore, mechanistic study showed that the enhanced down-regulation effects of ssGLCL/siRNA nanoplexes on targeted mRNA and protein were probably attributed to the increased release of siRNA from lysosomes to cytoplasm following the cleavage of redox-active disulfide-bridge in ssGLCL. Therefore, we believed that the redox-active H-shape ssGLCL could be a potential nanocarrier towards improving siRNA delivery.

Full text of DOI:10.1016/j.jconrel.2016.05.051

The role of disulfide-bridge on the activities of H-shape gemini-like cationic
lipid based siRNA delivery
Xiao-Fei Ma, Jing Sun, Chong Qiu, Yi-Fan Wu, Yi Zheng, Min-Zhi Yu,
Xi-Wei Pei, Lin Wei, Yu-Jie Niu, Wen-Hao Pang, Zhen-Jun Yang, Jian-Cheng
Wang, Qiang Zhang
PII:
DOI:
Reference:
S0168-3659(16)30330-3
doi: 10.1016/j.jconrel.2016.05.051
COREL 8294
To appear in:
Journal of Controlled Release
Received date:
Revised date:
Accepted date:
8 September 2015
17 May 2016
23 May 2016
Please cite this article as: Xiao-Fei Ma, Jing Sun, Chong Qiu, Yi-Fan Wu, Yi Zheng,
Min-Zhi Yu, Xi-Wei Pei, Lin Wei, Yu-Jie Niu, Wen-Hao Pang, Zhen-Jun Yang, Jian-
Cheng Wang, Qiang Zhang, The role of disulfide-bridge on the activities of H-shape
gemini-like cationic lipid based siRNA delivery, Journal of Controlled Release (2016), doi:
10.1016/j.jconrel.2016.05.051
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

ACCEPTED MANUSCRIPT
The role of disulfide-bridge on the activities of H-shape gemini-like cationic
lipid based siRNA delivery
Xiao-Fei Ma1, Jing Sun1, Chong Qiu1,Yi-Fan Wu, Yi Zheng, Min-Zhi Yu, Xi-Wei Pei, Lin Wei,
Yu-Jie Niu, Wen-Hao Pang, Zhen-Jun Yang*, Jian-Cheng Wang**, Qiang Zhang
State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences,
Peking University, Beijing, 38 xueyuan Road, 100191, China.
*Corresponding Author:
Jian-Cheng Wang, Ph.D, Tel/Fax: 86-10-82805932, Email: wang-jc@bjmu.edu.cn; &
Zhen-Jun Yang, Ph.D, Tel/Fax: 86-10-82802503, Email: yangzj@bjmu.edu.cn;
1These authors equally contributed to this work.
Abstract
In our previous study, a H-shape gemini-like cationic lipid (ssGLCL, formerly named as
CLD), composed of two hydrophilic lysine heads and two hydrophobic oleyl alcohol tails
with a bridge of the redox-active disulfide-bond, had been synthesized and used as a
nanocarrier for delivering small interfering RNAs (siRNAs) into cells. In order to further
elucidate the role of disulfide (-S-S-) bridge on the activity of ssGLCL based siRNA delivery,
a comparable ccGLCL bridged with a non-reducible carbon-carbon bond was synthesized and
used as control in this study. Both two H-shape GLCL molecules could individually
self-assemble into cationic nanoparticles in water phase and complex with negatively-charged
siRNA into nanoplexes with particle size of ~ 200 nm and zeta potential of ~ +30 mV, and
exhibit effective siRNA delivery both in vitro and in vivo. Investigation of internalization
pathway displayed that both ssGLCL/siRNA and ccGLCL/siRNA nanoplexes were
predominantly internalized into MCF-7 cells by the clathrin-mediated endocytosis pattern.
Although a lower cellular uptake of siRNA was found in the human breast cancer MCF-7
cells, the ssGLCL/siRNA nanoplexes could exhibit similar or even stronger down-regulation
effects on the targeted EGFR mRNA and protein in MCF-7 cells when compared to the
ccGLCL/siRNA nanoplexes. Furthermore, mechanistic study showed that the enhanced
down-regulation effects of ssGLCL/siRNA nanoplexes on targeted mRNA and protein were
probably attributed to the increased release of siRNA from lysosomes to cytoplasm following
the cleavage of redox-active disulfide-bridge in ssGLCL. Therefore, we believed that the
redox-active H-shape ssGLCL could be a potential nanocarrier towards improving siRNA
delivery.
Keywords: siRNA delivery, gemini-like lipid, disulfide bond, biodegradation, nanocarrier.
1. Introduction.
Nowadays small interference RNA (siRNA) is regarded as a significant potential
therapeutic agent for diverse diseases, especially tumors [1-4]. However, its inherent defects
1

ACCEPTED MANUSCRIPT
impede the further clinic progress [5-7]. Hence, a lot of nanocarrier-based delivery systems
have emerged immensely for relieving this dilemma [8-10]. Therein, cationic liposomes have
been deemed as the most promising gene vectors for their efficient transfection and facile
formulation [11, 12].
As one distinct representative, gemini-like cationic lipids (GLCLs) have particular
superiority and evoke interest in biological and biomedical applications [13, 14]. These
structures of H-shape or X-shape GLCLs are generally composed of three basic ingredients:
two cationic heads, two aliphatic tail chains and a rigid or flexible linker [15, 16]. These
distinctive properties endowed GLCLs not only the higher encapsulation efficiency but also
the lower cytotoxicity than those of corresponding monomeric lipids, which displayed
advantages for delivery DNA and antisense oligonucleotide (AS-ODN) drugs [17-19]. In our
previous studies [20], a novel H-shape gemini-like cationic lipid (ssGLCL, formerly named
as CLD), composed of two hydrophilic lysine heads and two hydrophobic oleyl alcohol tails
linked with disulfide (-S-S-) bridge, had been synthesized and used as a nanocarrier for
siRNA delivery.
In recent years, the concept of “environment-response” was fused in the design of siRNA
delivery systems [21, 22]. These stimuli-responsive features (such as pH [23], enzyme [24],
redox [25], and hypoxia [26] etc.) promote the disassembly of nanoplexes and increase the
release of siRNA from endosomes or lysosomes, consequently improving gene silencing
effects on targeted mRNA. During the past years, disulfide-bond structures were embedded
into multiple delivery systems based on the intracellular redox environment [27-32]. The
redox imbalance between extra- and intracellular environments in tumor tissues endowed it
biodegradable characteristics in the present of excess reductive substances in cytoplasm, thus
mediating redox-degradation assisted release of gene drugs and improving transfection
efficacy.
Our previous study have demonstrated that the siRNA targeting MEK1 mRNA formulated
with ssGLCLs into lipoplexes could exhibit an effective gene silencing of MEK1 expression
[20]. However, there were still some important mechanistic issues need be further investigated:
i) the cationic nanoparticles self-assembled with amphiphilic ssGLCL molecules would
possess a higher capacity for siRNA package and protection; ii) the prepared ssGLCL/siRNA
nanoplexes would exhibit a lower cytotoxicity; iii) enhanced silencing on targeted mRNA
would be obtained by a redox-assisted siRNA releasing from lysosomes to cytoplasm
attributed to the cleavage of disulfide-bridge in H-shape ssGLCL.
To elucidate the mechanisms of the role of disulfide-bridge on the activities of H-shape
ssGLCL based siRNA delivery, the analogous of H-shape gemini-like cationic lipid
(ccGLCL) bridged with an non-reducible carbon-carbon bond, instead of disulfide-bridge,
were also synthesized as an important reference. In addition, monomeric lipid units (named as
sGLCL) were also synthesized as the mimic residue after cleavage of H-shape ssGLCL.
2. Materials and methods.
2.1 Materials and Cell Lines
1-Ethyl-3-(3-dimethyllaminopropyl)-carbodiimide hydrochloride (EDCI), 4-dimethyl
aminopyridine (DMAP), N-Hydroxybenzotrizole (HOBt) and L-2, 6-diaminopimelic acid
2

ACCEPTED MANUSCRIPT
were obtained from J&K Scientific Ltd. (Beijing, China). Boc-Lys(Boc)-OH, alanine,
cysteine, oleyl alcohol, and Boc2O were purchased from GL Biochem Co., Ltd. (Shanghai,
China). Lipofectamine2000, OPTI-MEM, and LysoTracker Red DND-99 were purchased
from Invitrogen (NY, USA); Agarose was obtained from GENE COMPANY (Hong Kong,
China); Hoechst 33258 was purchased from Molecular Probes Inc. (Oregon, USA);
sulforhodamine B (SRB) was purchased from Sigma Aldrich. Co. (St. Louis, MO).
Anti-EGFR siRNA (sense strand: 5’-AGGAAUUAAGAGAAGCAACAUdTdT-3’; antisense
strand: 5’-AUGUUGCUUCUCUUAAUUCCUdTdT-3’, named as siEGFR), Anti-VEGF
siRNA (sense
strand:
GGAGUACCCUGAUGAGAUCdTdT,
antisense
strand:
5’-GAUCUCAUCAGGGUACUCCdTdT-3’, named as siVEGF), Anti-GFP siRNA (sense
strand: 5’-GGCUACGUCCAGGAGCGCACC-3’; antisense strand: 5’-UGCGCUCC
UGGACGUAGCCUU, named as siGFP), negative control siRNA (sense strand: 5’-UUC
UCC GAA CGU GUC ACG UTT-3’; antisense strand: 5’-ACG UGA CAC GUU CGG AGA
ATT-3’. named as siNC) and fluorescein-labeled siRNA (5’end of the sense strand,
FAM-siRNA) were synthesized and purified with HPLC by GenePharma Co. Ltd (Shanghai,
China). Human breast cancer MCF-7 cells, human cervical cancer HeLa cells and those who
can express GFP (Green Fluorescence Protein) stably (GFP-HeLa cell line) were obtained
from the Institute of Basic Medical Science, Chinese Academy of Medical Sciences (Beijing,
China). The cells were cultured in RPMI-1640 medium (Macgene, Beijing, China)
supplemented with 10% fetal bovine serum (FBS), 100 units/mL penicillin and 100 mg/mL
streptomycin at 37℃ in a humidified atmosphere containing 5% CO2. The cells for all
experiments were in the logarithmic phase of growth.
2.2 Synthesis of GLCLs Materials.
Three novel GLCLs materials were synthesized in a similar scheme. The synthetic route
of ssGLCL material was described here as the representative procedure. In brief, the
Boc-protected cystine 2 was firstly prepared and then esterificated with oleyl alcohol under
the coupling reagents EDC and DMAP to obtain intermediate 3 as colorless gel. Compound 3
was deprotected by TFA to produce compound 4 as yellow grease. Intermediate 5 was formed
as colorless wax in amidation between compound 4 and Boc-Lys (Boc)-OH under activator
EDCI and HOBt. Then compound 5 was deprotected again and salified by HCl gas to obtain
1
the ssGLCL (compound 6) as a white solid. All products were confirmed by H nuclear
magnetic resonance (AVANCE III, 400 MHz, Bruker, Billerica, MA) and ESI-Ms
spectrometry (VG Scienta, England). The total synthetic yield was 67.6% and the detailed
processing was seen in supplementary information.
2.3 Preparation and Characterization of GLCL/siRNA Nanoplexes.
GLCL nanoparticles were prepared using thin film hydration method. Briefly, GLCL
materials were dissolved into methanol in a flask with appropriate concentration, and then
organic solvent was removed by rotary evaporation method at 37℃ for 15 min. The formed
film was hydrated by 5% glucose solution pretreated with diethy pyrocarbonate (DEPC) and
sterilization at 60℃, subsequently sonicated at 70℃ for 40 min. As a result, GLCL
nanoparticles were obtained. To form the GLCL/siRNA nanoplexes, the GLCL nanoparticles
and siRNA were mixed together with N/P = 10/1 or other ratios, and the nanoplexes were
3

ACCEPTED MANUSCRIPT
incubated at 37℃ for 20 min, and then diluted to certain concentration for use. The particle
sizes and zeta potentials of nanoplexes were measured at 25℃ using dynamic light scattering
(Malvern Zetasizer Nano ZS, Malvern, UK). The morphology of nanoplexes was observed by
transmission electron microscope (TEM, JEOL, Japan) and atomic force microscope (AFM,
Veeco Metrology, USA).
2.4 Gel Retardation Assay.
Various GLCL/siRNA nanoplexes were prepared at different N/P ratios, the final siRNA
concentration was fixed at 1 μM. Then these prepared nanoplexes were tackled with loading
buffer and electrophoresed with 1％ agarose gel containing 0.5 mg/mL Exred (a special
luminant dye for siRNA staining). Electrophoresis was performed at 80 mV for 3 min,
subsequently 100 mV for 10 min, and these resulting gels were photographed under
UV-illumination. Free siRNA was used as the control.
2.5 Cytotoxicity Assay.
In vitro cytotoxicity of GLCL nanoparticles on MCF-7 cells was performed by
sulforhodamine B (SRB) assay. Briefly, MCF-7 cells were seeded in 96-well plates at a
density of 5000 cells per well. After 24 h proliferation, cells were treated with 200 μL
OPTI-MEM containing serial concentrations of each nanoparticle. The medium was replaced
by RPMI-1640 medium containing 10％ FBS after 4h incubation, and then followed by
another 48 h incubation. Removed the culture medium carefully, fixed cellular protein by the
addition of 100 μL of 10％ TCA (trichloroacetic acid) at 4℃ for 1 h, the 96-well plates were
then washed with deionized water for five cycles and air dried. The SRB solution (200 μL)
was added into each well and allowed staining for a 30 min in dark condition. Then, the SRB
solution was removed, and plates were washed with 1％ acetic acid for five times. After
air-dried, 100 μL of 10 mM Tris base solution was added into each well to solubilize the
protein-bound dye on a gyratory shaker for 30 min. Absorbance values were read on a
microplate reader at the wavelength of 540 nm. The cell viability (％) was calculated
according to the following formula: Cell viability (％) = [OD540 (sample)/OD540 (control)] ×
100％, where OD540 (sample) is the absorbance from the cells treated with various
nanoparticles and OD540 (control) is that from the cells treated with 5％ glucose.
2.6 Hemolysis Assay.
Various GLCL nanoparticles were separately diluted to a final concentration of 0.96
mg/mL with 5％ glucose. Erythrocytes (RBC) were isolated from rat blood by centrifugation
at 1500 g for 10 min at 4℃ and were washed with physiological saline solution three times at
the same condition as above. The cell pellet was resuspended into a 2％ (v/v) erythrocyte
suspension with pre-chilled PBS (0.1 M, pH = 7.4), and then the erythrocyte suspensions (180
μL) were loaded in each well of 96-well plates. Aliquot 20 μL of nanoparticle suspensions
was added into each well of the plates. Each sample was separated in two groups of
erythrocyte suspension without or with DTT reagent (final concentration was set as 10 mM).
The 5％ glucose solution was used as a negative control, and Triton X-100 (1％, v/v) was
used as a positive control. At predetermined time points, each sample was centrifuged and the
supernatant was transferred to a new plate, and the absorbance was measured at 540 nm using
4

ACCEPTED MANUSCRIPT
a microplate reader.
2.7 Cellular Uptake of GLCL/siRNA Nanoplexes.
MCF-7 cells and HeLa cells (3 × 105 per well) were seeded in six-well plates. After 24 h
proliferation, various formulations containing FAM-siRNA with different N/P ratios at the
fixed final concentration of 100 nM were exposed to cells and incubated for an additional 4 h
at 37 °C. After incubation, the cells were harvested and washed three times with pre-cooled
PBS solution and the uptake of FAM-siRNA was detected by a FACS Calibur flow cytometer
(Becton Dickinson, San Jose, CA, USA) immediately. Confocal microscopy was also used to
observe the internalization efficacy of GLCL/siRNA nanoplexes. After transfected for 4 h, the
cells were washed three times with PBS followed by fixation with 4％ paraformaldehyde for
15 min at room temperature. Then, F-actin and nucleus were individually stained by
rhodamine labeled phallacidin and Hoechst 33258 as the manufacturer’s instructions. The
cells were visualized under a Leica TCS SP5 confocal fluorescence microscope (Leica
Microsystems, Heidelberg, Germany).
2.8 Endocytic Mechanism of GLCL/siRNA Nanoplexes
HeLa or MCF-7 Cells were seeded in 12-well plates at 2×105 cells/well for 24 hours
before transfection. Transfections were performed with the presence of inhibitors in cell
culture medium (50 μg/mL genistein, 10 g/mL chlorpromazine, 0.25 mM amiloride). Firstly,
prior to transfection, cell culture medium was replaced with serum-free medium containing
the desired inhibitors for 30 minutes. Then, various FAM-siRNA loaded formulations with
N/P = 10/1 at the final siRNA concentration of 100 nM were exposed to cells and incubated
for an additional 4 hours at 37℃. In the process of incubation, the concentration of inhibitors
was consistent with the pre-treatment. After incubation, the cells were harvested and washed
with pre-cooled PBS solution three times and the intracellular fluorescence intensities were
detected by a FACS Calibur flow cytometer (Becton Dickinson, San Jose, CA, USA)
immediately.
Confocal microscopy was also used to observe the co-localization of channel markers
with fluorescence-labeled GLCL/siRNA nanoplexes. MCF-7 Cells were seeded in confocal
dishes (2×105 cells/dish) for 24 hours prior to transfection. Then, various formulations
containing Cy5-siRNA with N/P = 10/1 at the final concentration of 200 nM were exposed to
cells and incubated for an additional 2 hours at 37℃. Before the end of incubation, transferrin
labled with Alexa Fluor-488 (10 μg/mL, 30 min), Cholera Toxin Subunit B labled with Alexa
Fluor-488 (2 μg/mL, 30 min) and Dextran labled with FITC (1 mg/mL,1 hour) were dropped
into the culture medium respectively. After treated, the cells were washed with PBS three
times followed by fixation with 4％ paraformaldehyde for 15 min and cell punching with 0.1％
Tition-PBS for 1 min at room temperature. Then, F-actin and nucleus were individually
stained by rhodamine labeled phallacidin (1 U/mL, 10 min) and Hoechst 33258 (5 μg/mL,10
min) according to the manufacturer’s instructions. Cells were visualized by a Leica TCS SP8
confocal fluorescence microscope (Leica Microsystems, Heidelberg, Germany).
2.9 Intracellular Distribution of GLCL/siRNA Nanoplexes.
MCF-7 cells (2×105 cells per well) were seeded into confocal dishes. After 24 hours
5

ACCEPTED MANUSCRIPT
incubation, the GLCL/FAM-siRNA nanoplexes suspension (the final siRNA concentration
equal to 100 nM, dispersed in OPTI-MEM medium) was added into each dish for different
time periods. At the end of transfection, LysoTracker Red (Invitrogen, Carlsbad, USA) (50
nM) was added in cell culture medium and incubated for 30 min at 37℃ for staining
endosomes/lysosomes. Then the cells were washed with PBS for three times, and fixed in 4％
paraformaldehyde. Intracellular fluorescence distribution of FAM-siRNA was observed with a
Leica SP 5 confocal microscope (Leica Microsystems, Heidelberg, Germany).
2.10 In vitro Gene Silencing Effects.
The levels of mRNA and protein in the cells treated with different nanoparticles (N/P =
10/1, final siRNA concentration were 100 nM) were analyzed by RT-PCR and western blot
technologies. To detect relative EGFR mRNA expression, MCF-7 cells were treated similarly
as the process in Cellular Uptake Section, except for the usage of siEGFR. Then cells were
incubated by another 24 h of incubation. Total RNA was extracted using TRIOL® reagent
method and reverse transcribed with ReverTra Ace qPCR RT Kit (TOYOBO, Japan). The
resulting cDNAs were used for PCR using THUNDERBIRD qPCR Mix (TOYOBO, Japan)
in triplicates. PCR and data collection were performed on Real-Time PCR amplifier
(MX3005P, Stratagene, USA). All quantitation were normalized to an endogenous control
GAPDH. The relative quantitation value for each target gene compared to the calibrator for
that target is expressed as 2-(Ct-Cc) (Ct and Cc are the mean threshold cycle differences after
normalizing to GAPDH). For western blot analysis of EGFR protein, similar cell culture and
treatments were performed as above described. After transfection for 48 h, the cells were
washed with pre-cooled PBS and then resuspended in 100 mL RIPA lysis buffer supplemented
with 1％ proteinase inhibitor cocktail. The cell lysates were incubated on ice for 30 min and
vortexed every 10 min. The total protein was gathered by centrifugation at 12,000 g for 5 min
and their concentration was determined with BCA protein assay kit. Total protein (30 mg) was
loaded on 10% sodium dodecyl sulfate polyacrylamide gel and electrophoresized at 80 mV
for 30 min and 120 mV for 2 h. Then the proteins were transferred to Polyvinylidene Fluoride
(PVDF) membranes at 200 mA for 90 min then blocked with 5％ skimmed milk on a
horizontal shaker for 1h. The membranes were incubated with the 1:500 EGFR monoclonal
antibody (Santa Cruz Biotechnology) overnight at 4 ℃, followed by incubation with
HRP-conjugated goat anti-rabbit antibodies (1:15,000; Zhongshan Golden-Bridge Co. Ltd.,
Beijing, China) at room temperature for 2 h. Finally, the membranes were exposed using a
Bio-rad ChemiDoc XRS System. β-Actin was used as endogenous control. In addition, the
fluorescence intensities of GFP in HeLa cells (GFP-HeLa cell line) treated with different
nanoplexes were detected. For GFP detection, similar cell culture and treatments were
performed as above described except for the usage of siGFP. After 48 h incubation, the cells
were harvested and washed three times with pre-cooled PBS solution and the fluorescence
intensity of GFP was detected by a FACS Calibur flow cytometer (Becton Dickinson, San
Jose, CA, USA) immediately. Another important VEGF protein in MCF-7 cell culture
medium was also analyzed using a Human VEGF immunoassay kit (Raybiotech, USA).
Similar cell culture and treatments were performed as above described except for the usage of
siVEGF. The final concentrations of siVEGF were 30, 50, 100 nM. After 4 h incubation, the
OPTI-MEM medium was replaced with fresh complete medium, followed by 8 h incubation.
6

ACCEPTED MANUSCRIPT
After that, the medium was changed to fresh complete medium and incubated for additional
16 h. Finally, the medium was collected and analyzed using a human VEGF immunoassay kit
(RayBiotech, USA) according to the manufacturer’s instructions. The untreated cells and the
cells treated with Lipofectamine2000 were used as negative and positive controls,
respectively.
2.11 In Vivo Anti-Tumor Study.
Six-week-old female Balb/c nude mice were housed in an air-conditioned room. All care
and handling of animals was performed with the approval of Institutional Authority for
Laboratory Animal Care of Peking University. MCF-7 cells (2×106 cells of each injection)
were inoculated subcutaneously in the right flank of each Balb/c nude mouse. Tumor growth
was monitored by measuring the tumors perpendicular diameter using a caliper. Estimated
tumor volume was calculated using the formula: volume (mm3) = (length×width2)/2. The
treatment was started at 7th day post-implantation when the tumor volume reached
approximately 80 mm3. The female balb/c nude mice xenografted MCF-7 tumors were
randomly divided into five groups and injected intratumorally with different siRNA
formulations at a dose of 1 mg/kg every other day for four times. 5％ glucose was used as
control. Body weight and tumor size of each mouse were also measured every day. Two days
after the last injection, the tumor tissues were excised for further investigation.
2.12 Immunofluorescence Detection of EGFR Protein in Tumor Tissue.
After sacrifice, tumor tissues collected from the nude mice were immediately embedded
in Tissue-Tek OCT (Sakura Finetek, Tokyo, Japan), frozen rapidly and subsequently cut into
5-mm cryo-sections. Sections of tumor tissues were fixed in pre-cooled acetone for 10 min at
4℃, followed by incubated with 1％ BSA for 30 min and subsequently with 1:50 anti-EGFR
rabbit monoclonal antibody (Abcam, Cambridge, United Kingdom) overnight at 4℃. After
being washed with PBS, the sections were treated with a 1:100 dilution of FITC-conjugated
goat anti-rabbit IgG (Zhongshan Golden-Bridge Co. Ltd., Beijing, China) at 37℃ for 1 hour.
After staining of the nucleus with Hoechst 33258 (2 mg/mL) for 25 min, the cryosections
were observed by Leica SP8 confocal microscopy.
2.13 Statistical Analysis
For statistical analysis between two groups, Student’s t-test for independent means was
applied. The differences between any two groups out of several groups were analyzed by
one-way analysis of variance (ANOVA) followed by LSD multiple comparisons. Statistical
analysis was performed with the SPSS software (Version 16.0, SPSS Inc, Chicago).
3. Results
3.1 Synthesis and Structural Confirmation of GLCLs.
The representative synthesis procedure of GLCLs materials was shown in Scheme 1.
7

ACCEPTED MANUSCRIPT
Scheme 1. Synthetic procedures of three GLCL materials. Reagents and Conditions: i. Boc2O/
dioxane, KHCO3 aq. r.t.; ii. EDC, DMAP/DCM, r.t.; iii. TFA/CH2Cl2, r.t.; iv. EDCI, HOBt,
DIEA/ CH2Cl2, r.t.; v. HCl(gas)/EtOAc, r.t.
The structures of three GLCLs materials (Figure 1) and all intermediates (see in
1
Supporting Information) were confirmed by HNMR (400 MHz, DMSO-d6) and ESI-Ms
spectrometry and the data were listed as below: ssGLCL: 1H NMR δ 9.33(d, 2H, -NH-C=O),
+
+
8.40(s, 6H, N*H3 -CH2-), 8.08(s, 6H, N*H3 -CH-), 5.32(m, 4H, -CH=CH-), 4.62(m, 2H,
+
O=C-C*H- NH3 ), 4.08(m, 4H, -O-CH2-), 3.88(m, 2H, -NH-C*H-C=O), 3.09(m, 4H,
NH3 -C*H2-), 2.75(s, 4H, -S-CH2-), 1.99(m, 8H, -C*H2-CH=CH-), 1.81(m, 4H,
+
+
NH3 -CH-C*H2-), 1.60-1.24(m, 56H, CH2 on long chains), 0.85 (t, 6H, -CH3). ESI-MS: m/z:
[M+H]+ calcd C54H104N6O6S2, 998.57; found 998.24. ccGLCL: H NMR δ 9.15 (d, 2H,
1
-NH-); 8.42 (m, 6H, -N*H3); 8.11 (m, 6H, -N*H3); 5.36-5.29 (m, 4H, -CH=CH-); 4.24-4.20 (m,
2H, -C*H-NH3); 4.05 (m, 4H, -O-CH2-); 3.87 (m, 2H, -NH2-C*H-C=O); 2.76 (s, 4H,
NH3-C*H2-); 1.99 (m, 6H, -NH-CH-C*H2- and -NH-CH-CH2-C*H2-); 1.59 (m, 8H,
-C*H2-CH=CH-C*H2-); 1.24 (m, 58H, CH2 on long chain); 0.86 (t, 6H, -CH3). ESI-MS: m/z:
[M+H]+ calcd C55H106N6O6, 948.47; found 948.25; 949.23. sGLCL: H NMR δ 9.06 (d, 1H,
1
-N*H-C=O); 8.34 (s, 3H, -N*H3); 8.07 (s, 3H, -N*H3); 5.37-5.29 (m, 2H, -CH=CH-);
4.33-4.30 (m, 1H, NH3-C*H-); 4.09-3.82 (m, 2H, -O-CH2-); 3.81 (s, 1H, -NH-C*H-CH3); 2.71
(s, 2H, NH3-C*H2-); 1.99-1.98 (m, 4H, -C*H2-CH=CH-C*H2-); 1.58 (m, 2H, NH3-CH-C*H2-);
1.56 (m, 4H, NH3-CH2-C*H2- and -O-CH2-C*H2-); 1.34 (m, 2H, -O-CH2-CH2-C*H2-);
1.33-1.26 (d, 3H, -CH-C*H3); 1.24 (m, 22H, CH2 on long chain); 0.86 (t, 3H, -CH2-C*H3).
ESI-MS: m/z: [M+H]+ calcd for C27H55N3O3, 468.73; found 468.73; 469.75.
8

ACCEPTED MANUSCRIPT
Figure 1. Structures of ssGLCL/ccGLCL/sGLCL materials.
3.2 Characterization of GLCL/siRNA Nanoplexes.
As seen from Table 1, all GLCL nanoparticles could complex with siRNA effectively and
form the uniform nanoplexes at N/P = 10/1 with a mean size around 200 nm. The
ssGLCL/siRNA nanoplexes and the ccGLCL/siRNA nanoplexes shared similar zeta
potentials of 30.5±1.8 mV and 33.3±1.5 mV, respectively. However, zeta potential of
sGLCL/siRNA nanoplexes was only 13.9 ± 1.5 mV, which was significantly lower than those
of the other nanoplexes.
Table 1. Characteristics of various nanoplexes
Nanoplexes
Particle size
(d, nm)
Polydispersity Zeta potential (mV)
index(PDI)
ssGLCL/siRNA
ccGLCL/siRNA
sGLCL/siRNA
232.1±4.2
231.7±30.1
199.0±4.7
0.14±0.067
0.27±0.102
0.31±0.034
30.5±1.8
33.3±1.5
13.9±1.5
The morphology was observed by TEM and AFM. It showed that GLCLs nanoparticles
and corresponding siRNA nanoplexes were uniform spherical vesicles without obvious
aggregation (Figure 2A and B). Compared to GLCL nanoparticles, these GLCL/siRNA
nanoplexes were observed as darker blots in TEM images and smaller ones in AFM images,
which indicated that GLCLs nanoparticles and siRNAs were compacted together more tightly
due to the electrostatic interaction.
9

ACCEPTED MANUSCRIPT
Figure 2. Characterization of the prepared nanoparticles. TEM (A) and AFM (B) images of
GLCL nanoparticles and their siRNA complexes. Size and zeta potential of nanoplexes
detected by DLS (C).
The results of gel retardation assay further confirmed that siRNA could be completely
compacted by all cationic GLCL nanoparticles at an N/P ratio as low as 3/1, with no
difference among these nanoplexes (Figure 3A). Cytotoxicity assay (Figure 3B) showed that
there were no significant difference between ssGLCL and ccGLCL nanoparticles, however,
the nanoparticles composed of the monomeric lipid sGLCL presented markedly cytotoxicity
to MCF-7 cells than other two H-shape GLCLs nanoparticles.
10

ACCEPTED MANUSCRIPT
Figure 3. Gel retardation assay for various nanoplexes at different N/P ratios (A).
Cytotoxicity of various GLCL nanoparticles in MCF-7 cells (B).
3.3 Cellular Uptake of GLCL/siRNA Nanoplexes.
With cellular uptake of siRNA, higher fluorescence intensities were found in cells after 4
h incubation with various GLCL/FAM-siRNA nanoplexes with different N/P ratios (Figure
4A). In MCF-7 cells, ccGLCL/FAM-siRNA group showed much higher fluorescence
intensity than ssGLCL/FAM-siRNA at almost all N/P ratios. When the ratio of N/P up to 10/1
or more, a significantly higher cellular uptake of siRNA was found in the
ccGLCL/FAM-siRNA nanoplexes than that of lipofectamine2000. In CLSM images (Figure
4B), ccGLCL group displayed a higher green fluorescence intensity than ssGLCL group.
During observation by CLSM, the rhodamine-labeled phallacidin (red fluorescence) was used
to display the cytoskeleton. Surprisingly, much more incomplete cell membrane was observed
in MCF-7 cells after 4 h incubation with ccGLCL/FAM-siRNA nanoplexes, while the smooth
and intact cell membrane was still observed in the ssGLCL/FAM-siRNA group. Interestingly,
this phenomenon was hardly found in HeLa cells when these two GLCL/siRNA nanoplexes
were incubated. Therefore, the N/P ratio of 10/1 was selected as an optimal ratio for both of
MCF-7 and HeLa cells in the following experiments.
11

ACCEPTED MANUSCRIPT
Figure 4. Cellular uptake of various nanoplexes in MCF-7 cells (A) and HeLa cells (B). The
intracellular fluorescence intensities were detected by flow cytometric after 4h incubation
with different nanoplexes with different N/P ratios. The data was shown as mean ± SD (n = 3).
(C) Confocal laser scanning microscope images of MCF-7 after 4h incubation with
ssGLCL/siRNA nanoplexes and ccGLCL/siRNA nanoplexes with a final siRNA
concentration of 100 nM at N/P = 10/1. The Rhodamine-labeled phallacidin (red) was used to
show the cytoskeleton and Hoechst 33258 (blue) for nucleus.
3.4 Endocytic Mechanism of GLCL/siRNA Nanoplexes
Endocytosis mechanisms were generally divided into four classes: clathrin-mediated
endocytosis (CME), caveolae-mediated endocytosis (CvME), macropinocytosis and
CME/CvME-independent endocytosis. In our study, two approaches were performed to
elaborate the endocytosis pathways of the GLCL/siRNA nanoplexes: i) Specific blocking
agents were used to block internalization channels. Amiloride, chlorpromazine (CPZ),
genistein and NaN3 are known as the inhibitors of macropinocytosis, CME-mediated
endocytosis, CvME-mediated endocytosis and ATP-dependent endocytosis respectively. The
changes of intracellular fluorescence intensities were detected by flow cytometry; ii) Specific
markers were used to indicate the co-localization of endocytosis pathway of the
fluorescence-labeled nanoplexes and the fluorescence distribution was detected by CLSM
12

ACCEPTED MANUSCRIPT
image.
As shown in Figure 5A1, the intracellular fluorescence intensities of ssGLCL/siRNA
nanoplexes were about 25% decrease induced by amiloride, CPZ and NaN3 respectively; For
ccGLCL/siRNA nanoplexes, the fluorescence intensity decreased by CPZ was about 50%. In
Figure 5B1, a significant reduction of fluorescence intensity (~ 50%) was observed in both
ssGLCL/siRNA and ccGLCL/siRNA nanoplexes when they were blocked with genistein and
NaN3. Seen from the results of various channel rates shown in Figure 5A2 and 5B2, both
ssGLCL/siRNA and ccGLCL/siRNA nanoplexes were internalized into MCF-7 cells or
HeLa cells predominantly through the CME-mediated endocytosis pathway (50% ~ 70%).
In CLSM images, the fluorescence-labeled transferrin, cholera toxin B and dextran (green
color) were used as the channel markers to trace intracellular trafficking routes of
GLCL/siRNA nanoplexes (Cy5-labeled siRNA, red) by co-localization observation. As shown
in Figure 5C, both ssGLCL/siRNA nanoplexes and ccGLCL/siRNA nanoplexes were
mainly co-localized with transferrin (yellow color); while no obvious co-localization were
found between ssGLCL/siRNA nanoplexes and CT-B, as well as seen from the
co-localization between ccGLCL/siRNA nanoplexes and dextran. These results were
consistence with those results detected by FCM (Figure 5A).
Figure 5. Internalization of various nanoplexes after treated with different inhibitors in
MCF-7 cells (A) and HeLa cells (B). The intracellular fluorescence intensities were detected
13

ACCEPTED MANUSCRIPT
by flow cytometry after 4 h incubation. The mock (100%) was the fluorescence intensity of
various nanoplexes without inhibitors. The data was shown as mean ± SD (n = 3). (C)
Internalization pathways of various nanoplexes in MCF-7 cells by confocal microscopy.
Hoechst 33258 (Blue) and rhodamine-labeled phallacidin (Purple) were used to show the
nucleus and membrane respectively. The final concentration of Cy5-labeled siRNA (Red) was
100 nM.
3.5 Evaluation of Disulfide-Bond Function.
Firstly, the cleavage of disulfide-bond in ssGLCL was investigated by comparing the
different hemolysis effects of GLCLs on red cells with or without DTT addition. As seen from
Figure 6A, before the addition of DTT, only sGLCL nanoparticles showed significant
hemolysis effects (about 65%), while there were no difference on the other groups compared
with 5% glucose, which was used as a negative control. However, a significant hemolysis
effect was found in ssGLCL nanoparticles (up to almost 40%) after the addition of DTT
reagent, while ccGLCL still did not show any hemolysis effect. This difference between
ssGLCL and ccGLCL nanoparticles may indicate the special property of disulfide-bond
under the reductive condition.
Figure 6. A. Relative hemolysis effects of different nanoplexes in the absence/presence of
DTT with different incubation time at 37℃. Data was shown as mean ± SD (n=3). ** p < 0.01.
B. CLSM images of MCF-7 cells after incubated with various GLCL/siRNA nanoplexes at
N/P = 10/1 for different time periods. The LysolTracer Red DND-99 (red) was used to show
the endosomes/lysosomes. The final FAM-siRNA concentration is 100 nM.
Secondly, the endosome/lysosome escape of ssGLCL and ccGLCL in MCF-7 cells were
evaluated by CLSM imaging to estimate that whether the cleavage of disulfide-bond in
14

ACCEPTED MANUSCRIPT
lysosomes would promote the release of siRNA from ssGLCL nanoplexes into cytoplasm,
which resulted in higher gene silencing activities. As seen from Figure 6B, at 4 h time point,
ssGLCL/FAM-siRNA and ccGLCL/FAM-siRNA nanoplexes were all internalized into
cytoplasm as green discrete dots. As time went by, the co-localization of nanoplexes and
endosomes/lysosomes was gradually presented as yellow dots. At 8 h point (Figure 7A),
either in ssGLCL group or in ccGLCL group, more and more yellow dots appeared in
cytoplasm, that meant the similar endosomal/lysosomal internalization pathway was happened
for these two nanoplexes, More interesting, in ssGLCL/FAM-siRNA group, a lot of green
dots appeared again in cytoplasm at 12 h point, that meant large percentage of
green-fluorescence labeled siRNA was released from lysosomes into cytoplasm. In contrast,
there were many yellow dots still displayed in cytoplasm in ccGLCL/FAM-siRNA group.
This change was further confirmed by the quantitative analysis of Pearson’s correlation
coefficient. As seen in Figure 7B, it was shown that the Pearson’s correlation coefficient
decreased from 0.7007 (at 8 h point) to 0.4643 (at 12 h point) in ssGLCL group, while the
ccGLCL group changed from 0.7067 (at 8 h point) to 0.7981(at 12 h point). These results
suggested that the disulfide cleavage of ssGLCL under the reductive condition played an
important role for the release of siRNA from lysosome escape, which was consistent with the
results of hemolysis assay in Figure 6A.
Figure 7. (A) CLSM images of MCF-7 cells after incubation with ssGLCL/siRNA
nanoplexes and ccGLCL/siRNA nanoplexes at N/P = 10/1 for 8 h and 12 h periods. The
LysolTracker Red DND-99 (red) was used to stain endosomes/lysosomes. The final
FAM-siRNA concentration is 100 nM. (B) The quantitative analysis of Pearson’s correlation
15

ACCEPTED MANUSCRIPT
coefficient was performed by the co-localization computation module of Leica SP8 confocal
microscopy. Data was given as mean ± SD (n = 3).
3.6 In Vitro Gene Silencing Effects.
The levels of EGFR mRNA in MCF-7 cells were detected by RT-PCR analysis after
transfected with GLCL/siEGFR nanoplexes (Figure 8A). It was shown that both
ccGLCL/siEGFR and ssGLCL/siEGFR nanoplexes could significantly knock down EGFR
mRNA effectively as lipofectamine 2000, even though ssGLCL/siRNA showed a much lower
cellular uptake than ccGLCL/siRNA in MCF-7 cells. Meanwhile, the control GLCL/siNC
nanoplexes barely had any efficacy. These effects on mRNA level were perfect accordance
with the results of EGFR protein expression in western-blot assay (Figure 8B). Nevertheless,
a significant higher GFP silence effect was observed in ssGLCL/siGFP nanoplexes when
compared to that of ccGLCL/siGFP nanoplexes (Figure 8C). It was worthy to mention that
the same results of cellular uptake were found in HeLa cells after treated with these two
GLCL/siRNA nanoplexes.
Figure 8. Relative levels of EGFR mRNA in MCF-7 detected by RT-PCR analysis (A), EGFR
protein expression detected by Western-blot assay (B), relative fluorescence intensity (RFI) in
GFP-HeLa cells detected by flow cytometry (C), after transfection of different nanoplexes
16

ACCEPTED MANUSCRIPT
with a final siRNA concentration of 100 nM at the N/P ratio of 10/1. Data was shown as mean
± SD (n=3). ** p < 0.01, * p < 0.05.
To further confirm the concentration-dependent gene silencing efficiency, the levels of
VEGF protein expression in cell culture medium were detected by using a Human VEGF
immunoassay kit after the cells transfected with various nanoplexes (N/P ratio was 10/1)
containing different concentrations of siRNA (30 nM, 50 nM, 100 nM), respectively. From the
results in Figure 9, it was shown that VEGF protein could be significantly knocked down by
siVEGF at three concentrations in both MCF-7 and HeLa cells. When the final siVEGF
concentration was 100 nM (Figure 9C) or 50 nM (Figure 9B), a similar silencing activity on
VEGF was observed between ssGLCL/siRNA nanoplexes and ccGLCL/siRNA nanoplexes.
While at the siVEGF concentration of 30 nM (Figure 9A), ssGLCL/siRNA nanoplexes
displayed a significantly higher inhibition effects on targeted VEGF protein in both HeLa
cells and MCF-7 cells than ccGLCL/siRNA nanoplexes (P < 0.01).
Figure 9. Concentration-dependent gene silencing effects on VEGF expression. The levels of
VEGF expression in culture medium were detected by Elisa VEGF kit after tranfected with
different nanoparticles in MCF-7 cells (1) and HeLa cells (2). The final siRNA concentration
17

ACCEPTED MANUSCRIPT
of siRNA was 30 nM (A), 50 nM (B), 100 nM (C), respectively. N/P ratio was 10/1. Data was
shown as mean ± SD (n = 3). **p < 0.01.
3.7 In vivo Anti-Tumor Study.
Nude mice xenografted with human breast MCF-7 tumor were injected intratumorally for
a total of four treatments at the days indicated as arrows in Figure 10A. After inoculation,
tumor growth inhibition efficacy was calculated from the changes of tumor size after
intratumorally administration of different siRNA formulations, and 5% glucose was used as
control. Both ccGLCL/siEGFR nanoplexes and ssGLCL/siEGFR nanoplexes exhibited
significantly higher tumor growth inhibition effects than GLCL/siNC nanoplexes, through
intratumoral injection after injection for 3 doses (P < 0.05). However, there was no significant
difference of tumor growth inhibition effects between ccGLCL/siEGFR nanoplexes and
ssGLCL/siEGFR nanoplexes. The in vivo antitumor efficacies of these two formulations were
consistent with the in vitro gene silencing effects very well. The body weight was measured
every day for safety evaluation. As shown in Figure 10B, there was no significant change of
body weight in all of mice during the whole experimental period, as similar as that of 5%
glucose group.
Figure 10. Anti-tumor effects of various nanoplexes. Female Balb/c nude mice xenografted
MCF-7 tumors were injected intratumorally with different siRNA complexes (each dose of 1
mg siRNA/kg) for a total of 4 doses at the day indicated as arrows. A. The changes of tumor
18

ACCEPTED MANUSCRIPT
size after treated with GLCL/siRNA nanoplexes. B. The changes of body weight. (** p <
0.01)
3.8 Immunofluorescence Detection of EGFR Protein in Tumor Tissues.
As seen from Figure 11, both ssGLCL/siEGFR and ccGLCL/EGFR group showed
thorough down-regulation of EGFR protein which stained as green by FITC-labeled EGFR
antibody, while siNC groups had no such effect. These results indicated that both
ssGLCL/siEGFR nanoplexes and ccGLCL ones exhibited the efficient antitumor activity
with a gene sequence-specific pattern.
Figure 11. Immunofluorescence detection of EGFR expression in tumor tissue cryo-sections
by confocal microscopy. Green color represents for EGFR binding with FITC-labeled
antibody, and blue color represents for nuclei stained with Hoechst 33258.
4. Discussions.
With greater understanding of the difference between normal and pathological tissues and
cells, temperature, pH, enzyme and hypoxia are examples of "triggers" at the diseased site that
could be exploited with stimuli-responsive nanocarriers. A nanocarrier system incorporated
with stimuli-responsive property (e.g. pH or redox potential) would be used to overcome
some of the systemic and intracellular delivery barriers [23-27].
In the current study, we incorporated disulfide (-S-S-) linkage into the H-shape structure
of gemini-like cationic lipid to make the gemini lipid biodegradable as well. Here we
presented for the first time the transfection properties of this comparable non-reducible
ccGLCLs and elucidated the role of disulfide-bridge on the activity of redox-response
ssGLCL/siRNA nanoplexes in different cell lines (cervical cancer HeLa cell line and breast
cancer MCF-7 cell line) (Scheme 2). Transfection activities have been performed with three
kinds of siRNA sequences targeted EGFR, VEGF and GFP, respectively.
19

ACCEPTED MANUSCRIPT
Scheme 2 Schematic illustration of the redox-assisted release of siRNA from ssGLCL
nanoplexes compared to that of ccGLCL nanoplexes. The GLCL/siRNA nanoplexes were
prepared with electrostatic interaction. After cell internalization, the ssGLCLs were split into
two single chain lipid units with the cleavage of disulfide-bridge by abundant reductive
glutathione (GSH) in cells, then the resulted monomeric lipid unit exhibited higher membrane
fusion and facilitated release of siRNA from lysosomes into cytoplasm, thus an enhanced
gene silencing effects on targeted mRNA was obtained.
Biosafety and gene silencing activity are the two essential requirements of nanocarriers
applied for gene delivery [33-35]. As reported in previous studies, gemini-like cationic lipid,
comprising a dimeric form of two symmetrical monomeric lipid units via a suitable linker,
exhibited not only higher encapsulation efficiency but also lower cytotoxicity than those
corresponding monomeric lipid units, which displayed advantages for delivery DNA and
antisense oligonucleotide (AS-ODN) drugs [17-19]. In our study, it was demonstrated that
both two H-shape gemini cationic lipids (ssGLCLs and ccGLCLs) could individually
self-assemble into cationic nanoparticles in water phase and compact negatively-charged
siRNA into nanoplexes with particle size of ~ 200 nm and zeta potential of ~ +30 mV (Table
1 and Figure 2). Gel retardation assay also displayed both GLCL nanoparticles exhibited high
encapsulation efficiency of siRNA even at the N/P ratio of 3/1 (Figure 3A). Compared to
monomeric lipid unit sGLCLs, both ssGLCLs and ccGLCLs showed significantly lower
cytotoxicity in MCF-7 cells (Figure 3B). Furthermore, in vitro and in vivo gene silencing
experiments confirmed both ssGLCL/siRNA nanoplexes and ccGLCL/siRNA nanoplexes
could exhibit a high cellular uptake and the sequence-specific down-regulation effects on
targeted mRNA and protein expression, verified by RT-PCR analysis, western-blot assay and
20

ACCEPTED MANUSCRIPT
in vivo tumor growth inhibition (Figure 8-11). All above results demonstrated that both two
H-shape GLCLs-based systems would be potential carriers towards improving siRNA
delivery. Furthermore, it was revealed that the internalization pathway of ssGLCLs
nanoplexes and ccGLCLs nanoplexes were predominantly dependent on the CME-mediated
endocytosis (Figure 5), and there was no significant difference between ssGLCLs and
ccGLCLs found in both MCF-7 cells and HeLa cells.
Surprisingly, an unexpected outcome of gene silencing efficacy was observed in
ssGLCLs, in comparison with ccGLCLs. During in vitro gene silencing experiments, it was
displayed that the ssGLCL/siEGFR nanoplexes exhibited similar down-regulation effects on
both levels of EGFR mRNA and proteins expression when compared to the ccGLCL/siRNA
nanoplexes (Figure 8A and 8B), although a significantly lower cellular uptake of siRNA was
found in the human breast cancer MCF-7 cells treated with ssGLCLs (Figure 4A and 4C).
Interestingly, although a similar cellular uptake was found in GFP-HeLa cells (Figure 4B),
the ssGLCL/siGFP nanoplexes showed significantly higher down-regulation effects on GFP
expression compared to that of ccGLCL/siRNA nanoplexes (Figure 8C). Moreover, the
ssGLCL/siRNA nanoplexes displayed a significantly higher inhibition effects on targeted
VEGF protein in both HeLa cells and MCF-7 cells than ccGLCL/siRNA nanoplexes (P <
0.01) at the siVEGF concentration of 30 nM (Figure 9A). These evidences demonstrated that
the ssGLCL/siRNA nanoplexes exhibited stronger down-regulation effects on the levels of
targeted mRNA and/or proteins expression in cells, although similar or even lower cellular
uptake of siRNA was found in ssGLCLs when compared to those of ccGLCL/siRNA
nanoplexes. Why this phenomenon happened? What happened during transfection of
ssGLCL/siRNA nanoplexes?
Obviously, this interesting story most likely occurred in the intracellular trafficking
process. The role of reducible disulfide-bridge in ssGLCLs might be a key point for its higher
gene silencing efficacy in comparison with ccGLCLs. The mechanism was hypothesized as
follow: after cellular uptake of nanoplexes, the ssGLCL was broken into monomeric lipid
units (sGLCL) with the cleavage of disulfide-bridge by reductive substances glutathione
(GSH) [36, 37], which provided a reductive intracellular environment [38,39]; then the single
chain sGLCL exhibited a high membrane fusion and induced further disassembly of
nanoplexes. With a promoted siRNA escaping from lysosomes into cytosol, the enhanced
gene silencing effects on target mRNA occurred consequently. This hypothesis was confirmed
in our experiments. As seen in CLSM images (Figure 6B and Figure 7), at 4 h time point,
ssGLCL/FAM-siRNA nanoplexes and ccGLCL/FAM-siRNA nanoplexes were all
internalized into cytoplasm as green discrete dots after 4h incubation in cells. With the
prolonged incubation time, the co-localization of nanoplexes and endosome/lysosomes was
gradually presented as yellow dots. At 8 h or 12 h time point, much more green dots appeared
again in ssGLCL/FAM-siRNA group, that meant large percentage of green-fluorescence
labeled siRNA were released from lysosomes into cytoplasm. While the
ccGLCL/FAM-siRNA group still displayed yellow dots mostly. It was suggested that the
cleavage of disulfide-bond in lysosomes could promote the release of siRNA from ssGLCL
nanoplexes into cytoplasm and initiation gene silencing activities.
To further elucidate the mechanisms of enhanced release of siRNA induced by
biodegradation of ssGLCLs, we performed a hemolysis assay to mimic the enhanced
21

ACCEPTED MANUSCRIPT
membrane fusion based on the cleavage of disulfide-bond (Figure 6A).
1,4-dithio-D,L-threitol (DTT) was selected as a reductive reagent in in vitro environment [40].
It was observed that a significant hemolysis effect (up to almost 40％) was found in ssGLCL
nanoparticles after addition of DTT reagent, while ccGLCL still did not show any hemolysis
effect. This significant enhancement of hemolysis effect in ssGLCL was probably resulted
from disruption of cell membrane by the surfactant-based membrane fusion, like as the
hemolysis effects (about 65％) of sGLCL. The rationale behind this approach was to ensure
disassembling of the ssGLCL/siRNA complexes inside the cytoplasm after reduction of
disulfide-bond by the intracellular glutathione environment. Glutathione (GSH) is an
important molecule involved in the function of disulfide-bond in cells. It has been well known
that a significantly high GSH concentration exists inside cells (~ 2 mM) as compared to
outside cells (2 μM) [37]. It was suggested that the vast different of GSH concentration
between inside and outside could provide a potent mechanism for release of siRNA from
disulfide-bridge based ssGLCL/siRNA nanoplexes. Therefore, we believed the difference of
gene silencing efficiency between ssGLCL and ccGLCL nanoplexes was likely resulted from
the special property of disulfide-bond within the reductive condition.
By insight into the mechanism of the role of disulfide-bridge in gemini like lipids,
ssGLCLs indeed displayed its predominance of siRNA delivery in vitro, compared to
ccGLCLs. Nevertheless, there was a discrepancy between in vitro gene silencing effects and
in vivo anti-tumor efficiency. Although higher in vitro gene silencing effect and in vivo tumor
growth inhibition were happened in ssGLCL/siRNA nanoplexes, no significant difference
between ssGLCLs and ccGLCLs was observed in in vivo tumor growth inhibition (Figure
10A) and down-regulation of EGFR expression level (Figure 11). These conflicting results
might be influenced from the over-dose of siRNA and the unsatisfactory administration
frequency. We are progressing to meliorate the inadequacies mentioned above and will report
the detail results of in vivo tumor therapeutic efficacy in the near future.
5. Conclusion
Based on the above mentioned mechanistic study, it was suggested that the enhanced
down-regulation effects of ssGLCL/siRNA nanoplexes on targeted mRNA and protein
expression was probably resulted from the increased release of siRNA from
endosomes/lysosomes to cytoplasm, which was promoted by the enhanced lysosomal escape
capacity of monomeric lipid units (sGLCL) after the disulfide-bridge cleavage of ssGLCLs
in lysosomes induced by intracellular reductive substances glutathione (GSH). Therefore, we
believe the redox-active disulfide bridge plays a key role on the activity of H-shape
gemini-like lipid based siRNA delivery.
Acknowledgments
The authors acknowledge the financial support from the Ministry of Science and
Technology of China (Grant No. 2013CB932501, 2012CB720604, 2012AA022501), the
National Natural Science Foundation of China (Grant No. 81273455, 81473158 and
20932001), and Programs from Ministry of Education of China (NCET-11-0014 and
BMU20110263).
22

ACCEPTED MANUSCRIPT
References
[1] A. Fire, S.Q. Xu, M.K. Montgomery, S.A. Kostas, S.E. Driver, C.C. Mello. Potent and
specific genetic interference by double-stranded RNA in Caenorhabditis Elegans. Nature,
391(1998) 806-810.
[2] M. Ghildiyal, P.D. Zamore. Small silencing RNAs: an expanding universe. Nat. Genet.,
10(2009), 94-108.
[3] S.M. Hammond, E. Bernstein, D. Beach, G.J. Hannon. An RNA-directed nuclease
mediates post-transcriptional gene silencing in Drosophila cells. Nature, 404(2000), 293-295.
[4] Y. Dorsett, T. Tuschl. siRNA: applications in functional genomics and potential as
therapeutics. Nat. Rev. Drug Discov., 3(2004), 318-329.
[5] K.A. Whitehead, R. Langer, D.G. Anderson. Knocking down barriers: advances in siRNA
delivery. Nat. Rev. Drug Discov., 8(2009), 129-138.
[6] J. Wang, Z. Lu, M.G. Wientjes, J.L. Au. Delivery of siRNA therapeutics: barriers and
carriers. Aaps J., 12(2000), 492-503.
[7] B.L. Davidson, P.B. Jr-McCray. Current prospects for RNA interference-based therapies.
Nat. Genet., 12(2011), 329-340.
[8] H.M. Ali, G. Urbinati, M. Raouane, L. Massaad-Massade. Significance and applications of
nanoparticles in siRNA delivery for cancer therapy. Expert Rev. Clin. Pharmacol., 5(2012),
403-412.
[9] P. Kesharwani, V. Gajbhiye, N.K. Jain. A review of nanocarriers for the delivery of small
interfering RNA. Biomaterials, 33(2012), 7138-7150.
[10] X.Z. Yang, S. Dou, Y.C. Wang, H.Y. Long, M.H. Xiong, C.Q. Mao, Y.D. Yao, J. Wang.
Single-step assembly of cationic lipid polymer hybrid nanoparticles for systemic delivery of
siRNA. ACS Nano, 6(2012), 4955-4965.
[11] Y. Wang, L. Huang. A window onto siRNA delivery. Nat. Biotechnol., 31(2013),
611-612.
[12] B. Ozpolat, A.K. Sood, G. Lopez-Berestein. Liposomal siRNA nanocarriers for cancer
therapy. Adv. Drug Deliver. Rev., 66(2014), 110-116.
[13] M. Damen, J. Aarbiou, S.F. van Dongen, R.M. Buijs-Offerman, P.P. Spijkers, M. van den
Heuvel, K. Kvashnina, R.J. Nolte, B.J. Scholte, M.C. Feiters. Delivery of DNA and siRNA by
novel gemini-like amphiphilic peptides. J. Control. Release, 145(2010), 33-39.
[14] A.J. Kirby, P. Camilleri, J.B. Engberts, M.C. Feiters, R.J. Nolte, O. Soderman, M.
Bergsma, P.C. Bell, M.L. Fielden, C.L. Garcia Rodriguez, P. Guedat, A. Kremer, C. McGregor,
C. Perrin, G. Ronsin, M.C. van Eijk. Gemini surfactants: new synthetic vectors for gene
transfection. Angew. Chem. Int. Ed., 42(2003), 1448-1457.
[15] S.K. Misra, M. Munoz-Ubeda, S. Datta, A.L. Barran-Berdon, C. Aicart-Ramos, P.
Castro-Hartmann, P. Kondaiah, E. Junquera, S. Bhattacharya, E. Aicart. Effects of a
delocalizable cation on the headgroup of gemini lipids on the lipoplex-type nanoaggregates
directly formed from plasmid DNA. Biomacromolecules, 14(2013), 3951-3963.
[16] J. Biswas, A. Bajaj, S. Bhattacharya. J. Phys. Chem. B, 115(2011), 478-486.
[17] S. Perrone, M. Usai, P. Lazzari, S.J. Tucker, H.M. Wallace, M. Zanda. Efficient cell
transfection with melamine-based gemini surfactants. Bioconjugate Chem. 24(2013),
176-187.
[18] C. Bombelli, F. Faggioli, P. Luciani, G. Mancini, M.G. Sacco. Efficient transfection of
23

ACCEPTED MANUSCRIPT
DNA by liposomes formulated with cationic gemini amphiphiles. J. Med. Chem., 48(2005),
5378-5382.
[19] M. Munoz-Ubeda, S.K. Misra, A.L. Barran-Berdon, S. Datta, C. Aicart-Ramos, P.
Castro-Hartmann, P. Kondaiah, E. Junquera, S. Bhattacharya, E. Aicart. How does the spacer
length of cationic gemini lipids influence the lipoplex formation with plasmid DNA?
Physicochemical and biochemical characterizations and their relevance in gene therapy.
Biomacromolecules, 13(2012), 3926-3937.
[20] Y. Zheng, Y.J. Guo, Y.T. Li, Y. Wu, L.H. Zhang, Z.J. Yang. A novel gemini-like cationic
lipid for the efficient delivery of siRNA. New J. Chem., 38(2014), 4952-4962.
[21] D. Ma. Enhancing endosomal escape for nanoparticle mediated siRNA delivery.
Nanoscale, 6(2014), 6415-6425.
[22] T. Sun, Y.S. Zhang, B. Pang, D.C. Hyun, M. Yang, Y.N. Xia. Engineered nanoparticles
for drug delivery in cancer therapy. Angew. Chem. Int. Ed., 53(2014), 12320-12364.
[23] Y. Sato, H. Hatakeyama, Y. Sakurai, M. Hyodo, H. Akita, H. Harashima. A pH-sensitive
cationic lipid facilitates the delivery of liposomal siRNA and gene silencing activity in vitro
and in vivo. J. Control. Release, 163(2012), 267-276.
[24] L. Zhu, F. Perche, T. Wang, V.P. Torchilin. Matrix metalloproteinase 2-sensitive
multifunctional polymeric micelles for tumor-specific co-delivery of siRNA and hydrophobic
drugs. Biomaterials, 35(2014), 4213-4222.
[25] J. Beloor, C.S. Choi, H.Y. Nam, M. Park, S.H. Kim, A. Jackson, K.Y. Lee, S.W. Kim, P.
Kumar, S.K. Lee. Arginine-engrafted biodegradable polymer for the systemic delivery of
therapeutic siRNA. Biomaterials, 33(2012), 1640-1650.
[26] F. Perche, S. Biswas, T. Wang, L. Zhu, V.P. Torchilin. Hypoxia-targeted siRNA delivery.
Angew. Chem. Int. Ed., 53(2014), 3362-3366.
[27] G. Saito, J.A. Swanson, K.D. Lee. Drug delivery strategy utilizing conjugation via
reversible disulfide linkages: role and site of cellular reducing activities. Adv. Drug Deliver.
Rev., 55(2003), 119-125.
[28] C.J. Chen, J.C. Wang, E.Y. Zhao, L.Y. Gao, Q. Feng, X.Y. Liu, Z.X. Zhao, X.F. Ma, W.J.
Hou, L.R. Zhang, W.L. Lu, Q. Zhang. Self-assembly cationic nanoparticles based on
cholesterol-grafted bioreducible poly(amidoamine) for siRNA delivery. Biomaterials,
34(2013), 5303-5316.
[29] A. Fraix, T. Le Gall, M. Berchel, C. Denis, P. Lehn, T. Montier, P.A. Jaffres. Cationic
lipophosphoramidates with two disulfide motifs: synthesis, behaviour in reductive media and
gene transfection activity. Org. Biomol. Chem., 11(2013), 1650-1658.
[30] L. Brulisauer, N. Kathriner, M. Prenrecaj, M.A. Gauthier, J.C. Leroux. Tracking the
bioreduction of disulfide containing cationic dendrimers. Angew. Chem. Int. Ed., 51(2012),
12454-12458.
[31] M. Oua, R.Z. Xu, S.H. Kim, D.A. Bull, S.W. Kim. A family of bioreducible
poly(disulfide amine)s for gene delivery. Biomaterials, 30(2009), 5804-5814.
[32] V.A. Sethuraman, M.C. Lee, Y. H. Bae. A biodegradable pH-sensitive micelle system for
targeting acidic solid tumors. Pharm. Res., 25(2008), 657-666.
[33] Y. Kodama, S. Harauchi, S. Kawanabe, N. Ichikawa, H. Nakagawa, T. Muro, N. Higuchi,
T. Nakamura, T. Kitahara, H. Sasaki. Safe and effective delivery of small interfering RNA
with polymer- and liposomes-based complexes. Biol. Pharm. Bull., 36(2013), 995-1001.
24

ACCEPTED MANUSCRIPT
[34] S.H. Lee, B.H. Chung, T.G. Park, Y.S. Nam, H. Mok. Small-interfering RNA
(siRNA)-based functional micro- and nanostructures for efficient and selective gene silencing.
Accounts Chem. Res., 45(2012), 1014-1025.
[35] P. Kesharwani, V. Gajbhiye, N.K. Jain. A review of nanocarriers for the delivery of small
interfering RNA. Biomaterials, 33(2012), 7138-7150.
[36] I. Rebrin, S. Zicker, K.J. Wedekind, I. Paetau-Robinson, L. Packera, R.S. Sohal. Effect of
antioxidant-enriched diets on glutathione redox status in tissue homogenates and
mitochondria of the senescence-accelerated mouse. Free Radic. Biol. Med., 39(2005),
549-557.
[37] B. Khorsand, G. Lapointe, C. Brett, J.K. Oh. Intracellular drug delivery nanocarriers of
glutathione-responsive degradable block copolymers having pendant disulfide linkages.
Biomacromolecules, 14(2013), 2103-2111.
[38] T. Wang, D.Y. Ng, Y. Wu, J. Thomas, T. Tam-Tran, T. Weil. Bis-sulfide bioconjugates for
glutathione triggered tumor responsive drug release. Chem. Commun., 50(2014), 1116-1118.
[39] M.H. Lee, Z. Yang, C.W. Lim, Y.H. Lee, S. Dongbang, C. Kang, J.S. Kim. Chem. Rev.,
113(2013), 5071-5109.
[40] S. Li, Q. He, T. Chen, W. Wu, K. Lang, Z.M. Li, J.S. Li. Controlled co-delivery
nanocarriers based on mixed micelles formed from cyclodextrin-conjugated and cross-linked
copolymers. Colloid. Surface B, 123(2014), 486-492.
25

ACCEPTED MANUSCRIPT
Table 1 Characteristics of various nanoplexes (n=3)
Nanoplexes
Particle size
(d, nm)
Polydispersity index
(PDI)
Zeta potential
(mV)
ssGLCL/siRNA
ccGLCL/siRNA
sGLCL/siRNA
232.1±4.2
231.7±30.1
199.0±4.7
0.14±0.07
0.27±0.10
0.31±0.03
30.5±1.8
33.3±1.5
13.9±1.5
26

ACCEPTED MANUSCRIPT
Figure Captions
Scheme 1. Synthetic procedures of three GLCL materials. Reagents and Conditions: i. Boc2O/
dioxane, KHCO3 aq. r.t.; ii. EDC, DMAP/DCM, r.t.; iii. TFA/CH2Cl2, r.t.; iv. EDCI, HOBt,
DIEA/ CH2Cl2, r.t.; v. HCl(gas)/EtOAc, r.t.
Scheme 2 Schematic illustration of the redox-assisted release of siRNA from ssGLCL
nanoplexes compared to that of ccGLCL nanoplexes. The GLCL/siRNA nanoplexes were
prepared with electrostatic interaction. After cell internalization, the ssGLCLs were split into
two single-tailed sGLCL with the cleavage of disulfide-bridge by abundant reductive
glutathione (GSH) in cells, then the resulted monomeric lipid unit exhibited higher membrane
fusion and facilitated release of siRNA from lysosomes into cytoplasm, thus an enhanced
gene silencing effects on targeted mRNA was obtained.
Figure 1. Structures of ssGLCL/ccGLCL/sGLCL materials.
Figure 2. Characterization of the prepared nanoparticles. TEM (A) and AFM (B) images of
GLCL nanoparticles and their siRNA complexes. Size and zeta potential of nanoplexes
detected by DLS (C).
Figure 3. Gel retardation assay for various nanoplexes at different N/P ratios (A).
Cytotoxicity of various GLCL nanoparticles in MCF-7 cells (B).
Figure 4. Cellular uptake of various nanoplexes in MCF-7 cells (A) and HeLa cells (B). The
intracellular fluorescence intensities were detected by flow cytometric after 4h incubation
with different nanoplexes with different N/P ratios. The data was shown as mean ± SD (n = 3).
(C) Confocal laser scanning microscope images of MCF-7 after 4h incubation with
ssGLCL/siRNA nanoplexes and ccGLCL/siRNA nanoplexes with a final siRNA
concentration of 100 nM at N/P = 10/1. The Rhodamine-labeled phallacidin (red) was used to
show the cytoskeleton and Hoechst 33258 (blue) for nucleus.
Figure 5. Internalization of various nanoplexes after treated with different inhibitors in
MCF-7 cells (A) and HeLa cells (B). The intracellular fluorescence intensities were detected
by flow cytometry after 4 h incubation. The mock (100%) was the fluorescence intensity of
various nanoplexes without inhibitors. The data was shown as mean ± SD (n = 3). (C)
Internalization pathways of various nanoplexes in MCF-7 cells by confocal microscopy.
Hoechst 33258 (Blue) and rhodamine-labeled phallacidin (Purple) were used to show the
nucleus and membrane respectively. The final concentration of Cy5-labeled siRNA (Red) was
100 nM.
Figure 6. A. Relative hemolysis effects of different nanoplexes in the absence/presence of
DTT with different incubation time at 37℃. Data was shown as mean ± SD (n=3). ** p < 0.01.
B. CLSM images of MCF-7 cells after incubated with various GLCL/siRNA nanoplexes at
27

ACCEPTED MANUSCRIPT
N/P = 10/1 for different time periods. The LysolTracer Red DND-99 (red) was used to show
the endosomes/lysosomes. The final FAM-siRNA concentration is 100 nM.
Figure 7. (A) CLSM images of MCF-7 cells after incubation with ssGLCL/siRNA
nanoplexes and ccGLCL/siRNA nanoplexes at N/P = 10/1 for 8 h and 12 h periods. The
LysolTracker Red DND-99 (red) was used to stain endosomes/lysosomes. The final
FAM-siRNA concentration is 100 nM. (B) The quantitative analysis of Pearson’s correlation
coefficient was performed by the co-localization computation module of Leica SP8 confocal
microscopy. Data was given as mean ± SD (n = 3).
Figure 8. Relative levels of EGFR mRNA in MCF-7 detected by RT-PCR analysis (A), EGFR
protein expression detected by Western-blot assay (B), relative fluorescence intensity (RFI) in
GFP-HeLa cells detected by flow cytometry (C), after transfection of different nanoplexes
with a final siRNA concentration of 100 nM at the N/P ratio of 10/1. Data was shown as mean
± SD (n=3). ** p < 0.01, * p < 0.05.
Figure 9. Concentration-dependent gene silencing effects on VEGF expression. The levels of
VEGF expression in culture medium were detected by Elisa VEGF kit after tranfected with
different nanoparticles in MCF-7 cells (1) and HeLa cells (2). The final siRNA concentration
of siRNA was 30 nM (A), 50 nM (B), 100 nM (C), respectively. N/P ratio was 10/1. Data was
shown as mean ± SD (n = 3). **p < 0.01.
Figure 10. Anti-tumor effects of various nanoplexes. Female Balb/c nude mice xenografted
MCF-7 tumors were injected intratumorally with different siRNA complexes (each dose of 1
mg siRNA/kg) for a total of 4 doses at the day indicated as arrows. A. The changes of tumor
size after treated with GLCL/siRNA nanoplexes. B. The changes of body weight. (** p <
0.01)
Figure 11. Immunofluorescence detection of EGFR expression in tumor tissue cryo-sections
by confocal microscopy. Green color represents for EGFR binding with FITC-labeled
antibody, and blue color represents for nuclei stained with Hoechst 33258.
28

ACCEPTED MANUSCRIPT
Scheme 1
29