SiRNA delivery into tumor cells by cationic cholesterol derivative-based nanoparticles and liposomes

Article abstract of DOI:10.1248/bpb.b14-00526

Previously, we reported that cationic nanoparticles (NP) composed of diamine-type cholesteryl-3-carboxamide (OH-Chol, N-(2-(2-hydroxyethylamino)ethyl)cholesteryl-3-carboxamide) and Tween 80 could deliver small interfering RNA (siRNA) with high transfection efficiency into tumor cells. In this study, we synthesized new diamine-type cationic cholesteryl carbamate (OH-C-Chol, cholesteryl (2-((2-hydroxyethyl)-amino)ethyl)carbamate) and triamine-type carbamate (OH-NC-Chol, cholesteryl (2-((2-((2-hydroxyethyl)-amino)ethyl)amino)ethyl)carbamate), and prepared cationic nanoparticles composed of OH-C-Chol or OHNC-Chol with Tween 80 (NP-C and NP-NC, respectively), as well as cationic liposomes composed of OH-CChol or OH-NC-Chol with 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) (LP-C and LP-NC, respectively) for evaluation of their possible use as siRNA delivery vectors. LP-C and LP-NC/siRNA complexes (lipoplexes) exhibited larger gene silencing effects than NP-C and NP-NC/siRNA complexes (nanoplexes), respectively, in human breast tumor MCF-7 cells, although the NP-C nanoplex showed high association with the cells. In particular, LP-NC lipoplex could induce strong gene suppression, even at a concentration of 5 nM siRNA. From these results, cationic liposomes composed of OH-NC-Chol and DOPE may have potential as gene vectors for siRNA transfection to tumor cells.

Full text of DOI:10.1248/bpb.b14-00526

30
Biol. Pharm. Bull. 38, 30–38 (2015)
Vol. 38, No. 1
Regular Article
siRNA Delivery into Tumor Cells by Cationic Cholesterol Derivative-
Based Nanoparticles and Liposomes
,a
Yoshiyuki Hattori,* Eriko Hara,a Yumiko Shingu,a Daiki Minamiguchi,b Ayako Nakamura,a
Shohei Arai,a Hiroaki Ohno,b Kumi Kawano,a Nobutaka Fujii,b and Etsuo Yonemochia
a Institute of Medicinal Chemistry, Hoshi University; 2–4–41 Ebara, Shinagawa-ku, Tokyo 142–8501, Japan: and
b Graduate School of Pharmaceutical Sciences, Kyoto University; Sakyo-ku, Kyoto 606–8501, Japan.
Received July 21, 2014; accepted November 3, 2014
Previously, we reported that cationic nanoparticles (NP) composed of diamine-type cholesteryl-3-car-
boxamide (OH-Chol, N-(2-(2-hydroxyethylamino)ethyl)cholesteryl-3-carboxamide) and Tween 80 could
deliver small interfering RNA (siRNA) with high transfection efficiency into tumor cells. In this study, we
synthesized new diamine-type cationic cholesteryl carbamate (OH-C-Chol, cholesteryl (2-((2-hydroxyethyl)-
amino)ethyl)carbamate) and triamine-type carbamate (OH-NC-Chol, cholesteryl (2-((2-((2-hydroxyethyl)-
amino)ethyl)amino)ethyl)carbamate), and prepared cationic nanoparticles composed of OH-C-Chol or OH-
NC-Chol with Tween 80 (NP-C and NP-NC, respectively), as well as cationic liposomes composed of OH-C-
Chol or OH-NC-Chol with 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) (LP-C and LP-NC, re-
spectively) for evaluation of their possible use as siRNA delivery vectors. LP-C and LP-NC/siRNA complexes
(lipoplexes) exhibited larger gene silencing effects than NP-C and NP-NC/siRNA complexes (nanoplexes),
respectively, in human breast tumor MCF-7 cells, although the NP-C nanoplex showed high association with
the cells. In particular, LP-NC lipoplex could induce strong gene suppression, even at a concentration of 5nM
siRNA. From these results, cationic liposomes composed of OH-NC-Chol and DOPE may have potential as
gene vectors for siRNA transfection to tumor cells.
Key words cationic nanoparticle; cationic liposome; small interfering RNA (siRNA) delivery; transfection;
cholesteryl carbamate
RNA interference (RNAi) is a powerful gene-silencing pro- effect must be prepared.
cess that holds great promise in the field of cancer therapy.1)
For efficient gene silencing in tumor cells, the intracellular
Synthetic small interfering RNAs (siRNAs), which are small trafficking of siRNA after endocytosis must be controlled.
double-stranded RNAs, are substrates for the RNA-induced si- Polyethylenimines (PEIs) are also used as siRNA vectors and
lencing complex. siRNA suppresses the expression of a target have been shown to have the ability to escape the endosome
gene by triggering specific degradation of the complementary by buffering capacity.10) Therefore, to facilitate the endosomal
mRNA sequence.2) In siRNA delivery, non-viral vectors such escape of siRNA transfected by cationic nanoparticles or li-
as cationic liposomes or nanoparticles have been more com- posomes, we designed new cationic cholesterol derivative, tri-
monly used than viral vectors.3,4) For efficient siRNA delivery amine-type cholesteryl carbamate (OH-NC-Chol, cholesteryl
into cells by cationic liposome, many different cationic lipids (2-((2-((2-hydroxyethyl)amino)ethyl)amino)ethyl)carbamate),
have been synthesized. In particular, cholesterol- and glycerol- which has 2 units of aminoethylene between the amino head
based cationic lipids have been extensively used for cationic group and the cholesterol skeleton. In this study, we synthe-
liposome-mediated gene delivery.
sized cationic triamine-type cholesteryl carbamate (OH-NC-
For gene delivery with cationic liposomes, cationic cho- Chol) and diamine-type carbamate (OH-C-Chol, cholesteryl
lesterol derivatives are useful due to their high transfection (2-((2-hydroxyethyl)amino)ethyl)carbamate), and examined
efficiency and low toxicity.5–7) Cationic cholesterol derivatives the formulation of cationic liposomes and nanoparticles for
are composed of three distinct parts: a cholesteryl skeleton, siRNA delivery. We prepared cationic nanoparticles composed
a cationic amino group and a linker arm between the cho- of OH-C-Chol or OH-NC-Chol with Tween 80 (NP-C and
lesteryl skeleton and the cationic amino group. Previously, NP-NC, respectively) and cationic liposomes composed of
we reported that cationic nanoparticles (NP) composed of OH-C-Chol or OH-NC-Chol with DOPE (LP-C and LP-NC,
N-(2-(2-hydroxyethylamino)ethyl)cholesteryl-3-carboxamide respectively), and evaluated their possible use as siRNA deliv-
(OH-Chol) and Tween 80 could efficiently deliver siRNA into ery vectors.
tumor cells without any helper lipids such as 1,2-dioleoyl-sn-
glycero-3-phosphoethanolamine (DOPE) and cholesterol.8,9)
MATERIALS AND METHODS
OH-Chol is a cationic cholesterol derivative with a hydroxy-
ethyl group at the amino terminal and a carboxamide-type
Synthesis of Cationic Cholesterol Derivatives N-(2-
linker. However, to obtain a large gene silencing effect by (2-Hydroxyethylamino)ethyl)cholesteryl-3-carboxamide (OH-
NP, the presence of sodium chloride in the formulation of NP/ Chol) (Fig. 1) was synthesized as previously described.11)
siRNA complex (nanoplex) was needed, but it often induced
Diamine-type cholesteryl carbamate 7, cholesteryl (2-((2-
the formation of a large complex. For in vivo application, hydroxyethyl)amino)ethyl)carbamate (OH-C-Chol), was syn-
injectable-sized nanoplexes showing a large gene silencing thesized as shown in Fig. 2. To a mixture of commercially
*To whom correspondence should be addressed. e-mail: yhattori@hoshi.ac.jp
© 2015 The Pharmaceutical Society of Japan

Vol. 38, No. 1 (2015)
Biol. Pharm. Bull.
31
Fig. 1. Structure of Cationic Cholesterol Derivatives: (3S)-N-(2-(2-Hydroxyethylamino)ethyl)cholesteryl-3-carboxamide (OH-Chol); Cholesteryl
(2-((2-Hydroxyethyl)amino)ethyl)carbamate (OH-C-Chol); Cholesteryl (2-((2-((2-Hydroxyethyl)amino)ethyl)amino)ethyl)carbamate (OH-NC-Chol)
Fig. 2. Synthetic Scheme for Diamine-Type Cationic Cholesteryl Carbamate: Cholesteryl (2-((2-Hydroxyethyl)amino)ethyl)carbamate (OH-C-Chol)
Ns, 2-Nitrobenzenesulfonyl; THP, Tetrahydropyranyl; PMBA, p-Mercaptobenzoic acid.
available cholesteryl chloroformate (1) (2.56g, 5.7mmol) and 70% yield).
mono-nosylated ethylenediamine (2)12) (1.4g, 5.7mmol) in
A mixture of 3 (2.64g, 4.02mmol), commercially available
CH2Cl2 (28.5mL) was added Et3N (1.03mL, 7.41mmol) at 2-(2-chloroethoxy)tetrahydro-2H-pyran (4) (0.79g, 4.82mmol),
room temperature. The mixture was stirred overnight at this K2CO3 (1.11g, 8.04mmol) and KI (0.13g, 0.80mmol) in N,N-
temperature. The mixture was diluted with CHCl3 (50mL) and dimethylformamide (DMF) (40mL) under argon was stirred
H2O (50mL). The aqueous phase was extracted with CHCl3 overnight at 90°C. After cooling to room temperature, the
(50mL, twice), and the combined organic phases were washed mixture was diluted with CHCl3 (200mL) and H2O (200mL).
with brine (50mL), dried over MgSO4 and evaporated to give The aqueous phase was extracted with CHCl3 (100mL, twice),
crude 3. Purification of the crude sample by column chroma- and the combined organic phases were washed with brine
tography over silica gel (hexane/EtOAc) gave pure 3 (2.64g, (50mL), dried over MgSO4 and evaporated to give crude 5.

32
Biol. Pharm. Bull.
Vol. 38, No. 1 (2015)
Fig. 3. Synthetic Scheme for Triamine-Type Cationic Cholesteryl Carbamate: Cholesteryl (2-((2-((2-Hydroxyethyl)amino)ethyl)amino)ethyl)carbamate
(OH-NC-Chol)
Purification of the crude sample by column chromatography ture was diluted with CHCl3 (20mL) and H2O (20mL). The
over silica gel (hexane/EtOAc) gave pure 5 (1.22g, 39% yield). aqueous phase was extracted with CHCl3 (40mL, twice), and
To a mixture of 5 (1.22g, 1.55mmol) in MeOH (5.85mL) the combined organic phases were washed with brine (20mL),
and CHCl3 (1.9mL) was added p-TsOH·H2O (3.0mg, dried over MgSO4 and evaporated to give crude 9. Purification
0.016mmol) at room temperature under stirring. After of the crude sample by column chromatography over silica gel
completion of the reaction monitored by TLC (SiO2, hexane/ (hexane/EtOAc) gave pure 9 (3.02g, 87% yield).
EtOAc=1/1), the reaction was quenched by the addition of
By using a procedure identical to that described for the
saturated NaHCO3 (5mL). Concentration of the mixture gave preparation of 5, 9 (3.02g, 3.4mmol) was converted to 10
residual oil, which was diluted with CHCl3 (70mL) and satu- (1.12g, 32% yield) by the reaction with 4 (0.67g, 4.08mmol),
rated NaHCO3 (60mL). The aqueous phase was extracted with K2CO3 (1.08g, 7.8mmol) and KI (0.13g, 0.78mmol) in DMF
CHCl3 (60mL, twice), and the combined organic phases were (34 mL).
dried over MgSO4 and evaporated to give crude 6. Purification
of the crude sample by column chromatography over silica gel preparation of 6, 10 (1.12g, 1.1mmol) was converted to 11
(hexane/EtOAc) gave pure 6 (964mg, 89% yield). (777mg, 76% yield) by the reaction with p-TsOH·H2O (2.1mg,
A mixture of 6 (964mg, 1.375mmol), p-mercaptobenzoic 0.011mmol) in MeOH (4.1mL) and CHCl3 (1.38 mL).
acid (PMBA) (233mg, 1.51mmol) and K2CO3 (418mg, By using a procedure identical to that described for the
By using a procedure identical to that described for the
3.03mmol) in DMF (13.8mL) under argon was stirred over- preparation of 7, 11 (777mg, 0.835mmol) was converted to
night at 60°C. Further PMBA (233mg, 1.51mmol), K2CO3 12 (134mg, 29% yield) by the reaction with PMBA (566mg,
(418mg, 3.03mmol) and DMF (6.9mL) were added to the 3.68mmol) and K2CO3 (1.01g, 7.36mmol) in DMF (12.5mL).
mixture, and the mixture was stirred for one day. After cool-
siRNA siRNAs targeting nucleotides of firefly pGL3
ing, the mixture was diluted with CHCl3 (50mL), 2N NaOH luciferase (Luc siRNA), Cy5.5-labeled Luc siRNA (Cy5.5-
(10mL) and H2O (10mL). The aqueous phase was extracted siRNA) and nonsilencing siRNA (Cont siRNA) as a negative
with CHCl3 (50mL, twice), and the combined organic phases control were synthesized by Sigma Genosys (Tokyo, Japan).
were dried over MgSO4 and evaporated to give crude 7. Puri- The siRNA sequences of the Luc siRNA were as follows:
fication of the crude sample by column chromatography over sense strand: 5′-GUGGAUUUCGAGUCGUCUUAA-3′, and
NH-silica gel (CHCl3/MeOH, twice) gave pure 7 (294mg, 41% antisense strand: 5′-AAGACGACUCGAAAUCCACAU-3′. In
yield).
Cy5.5-siRNA, Cy5.5 dye was conjugated at the 5′-end of the
Triamine-type cholesteryl carbamate 12, cholesteryl sense strand. The siRNA sequences of the Cont siRNA were as
(2-((2-((2-hydroxyethyl)amino)ethyl)amino)ethyl)carbamate follows: sense strand: 5′-GUACC GCA CGUCAUUCGUAUC-3′,
(OH-NC-Chol), was synthesized as shown in Fig. 3. To a and antisense strand: 5′-UACG AAUGAC GUGCG GUACGU-3′.
mixture of cholesterol chloroformate (1) (1.66g, 3.71mmol) Alexa Fluor®488-labeled AllStars Negative Control siRNA
and bis-nosylated diethylenetriamine (8) (1.76g, 3.71mmol) in (AF-siRNA) was obtained from Qiagen (Valencia, CA, U.S.A.).
CH2Cl2 (18.6mL) was added Et3N (616µL, 4.45mmol) at room
Preparation of Nanoparticles and Nanoplexes The cat-
temperature. The protected diethylenetriamine derivative 8 ionic cholesterol derivative-based nanoparticles, NP, NP-C
was prepared according to a reported procedure with a slight and NP-NC, were prepared from OH-Chol/Tween 80, OH-
modification (2eq of Et3N were used for bis-nosylation).13) The C-Chol/Tween 80 and OH-NC-Chol/Tween 80 at a molar
mixture was stirred overnight at this temperature. The mix- ratio of 95/5 by a thin-film hydration method, respectively.13)

Vol. 38, No. 1 (2015)
Biol. Pharm. Bull.
33
Cationic liposomes, LP, LP-C and LP-NC, were prepared sacrificed and the tissues were frozen on dry ice and sliced at
from OH-Chol/DOPE, OH-C-Chol/DOPE and OH-NC-Chol/ 16 µm. The localization of Cy5.5-siRNA was examined using
DOPE at a molar ratio of 3/2 by a thin-film hydration method, an Eclipse TS100-F microscope (Nikon, Tokyo, Japan).
respectively. The nanoparticle/siRNA complex (nanoplex) or
Statistical Analysis Data were compared using analysis
the liposome/siRNA complex (lipoplex) at charge ratios (+/−) of variance and evaluated by Student’s t-test. A p value of
of 1/1, 3/1, 4/1, 7/1 and 10/1 of cationic lipid to siRNA was 0.05 or less was considered significant.
formed by the addition of each nanoparticle or liposome to
siRNA with gentle shaking and left at room temperature for
15min. The charge ratio (+/−) of nanoparticles/siRNA or lipo-
somes/siRNA is expressed as the molar ratio of cationic lipid
to siRNA phosphate.
RESULTS AND DISCUSSION
Synthesis of Cationic Cholesterol Derivatives F o r e f fi -
cient gene silencing in tumor cells, the intracellular trafficking
Size and ζ-Potential of Nanoparticles and Nanoplexes of siRNA after endocytosis must be controlled. Previously,
The particle size distributions of nanoparticles, nanoplexes, we reported that cationic nanoparticles (NP) composed of
liposomes and lipoplexes were measured by the cumulant diamine-type cholesteryl-3-carboxamide, OH-Chol (Fig. 1)
method using a light-scattering photometer (ELS-Z2, Otsuka and Tween 80 could deliver siRNA with high transfection
Electronics Co., Ltd., Osaka, Japan), at 25°C after diluting efficiency into tumor cells.8,9) Furthermore, to facilitate the
the dispersion to an appropriate volume with water. The endosomal escape of siRNA transfected by cationic nanopar-
ζ-potentials were measured using ELS-Z2 at 25°C after dilut- ticles, we synthesized another cationic cholesterol deriva-
ing the dispersion to an appropriate volume with water.
tive, triamine-type cholesteryl-3-carboxamide (OH-N-Chol,
Cell Culture Human breast cancer MCF-7-Luc (TamR- N-(2-(2-(2-hydroxyethylamino)ethylamino)ethyl)cholesteryl-3-
Luc#1) cells stably expressing firefly luciferase (pGL3) were carboxamide) (Supplemental Fig. S1), which has 2 units of
donated by Dr. Kazuhiro Ikeda (Division of Gene Regula- aminoethylene between the amino head group and the choles-
tion and Signal Transduction, Research Center for Genomic terol skeleton, and then prepared NP-N composed of OH-N-
Medicine, Saitama Medical University, Saitama, Japan). The Chol and Tween 80. In the transfection for tumor cells, NP-N
cells were cultured in Dulbecco’s modified Eagle’s medium nanoplex with a small size exhibited high gene knockdown by
(DMEM), supplemented with 10% heat-inactivated fetal bo- siRNA.13) In NP-N, we expected that OH-N-Chol may show
vine serum (FBS), 100µg/mL kanamycin and 0.5mg/mL G418 single protonation at neutral pH due to the strong electrostatic
at 37°C in a 5% CO2 humidified atmosphere.
repulsion between two protonated amines in OH-N-Chol,
Luciferase Activity MCF-7-Luc cells were plated into while exhibiting double protonations at an endosomal pH of
6-well culture dishes at a density of 3×105 cells per well. 5.5 (proton sponge effect). However, the carboxamide-type
For transfection, each lipoplex and nanoplex of 100pmol Luc cholesterol derivatives such as OH-Chol and OH-N-Chol had
siRNA or Cont siRNA formed at various charge ratios (+/−) a main drawback in their synthesis, especially for large-scale
was diluted in 2mL of medium supplemented with 10% FBS synthesis. Thus, OH-Chol and OH-N-Chol were prepared from
and then the mixture was added to the cells. Forty-eight hours cholesteryl carboxylic acid as the starting material, which was
after the transfection, luciferase activity was measured as obtained by CO2 treatment of Grignard reagent (cholesteryl
counts per second (cps)/µg protein using the luciferase assay magnesium bromide) derived from cholesteryl bromide.13) This
system (Pica Gene, Toyo Ink Mfg. Co., Ltd., Tokyo, Japan) step is very sensitive to moisture and air, which decreases
and BCA reagent (Pierce, Rockford, IL, U.S.A.), as previously the reproducibility of the reaction. Formation of the minor
reported.14) Luciferase activity (%) was calculated as relative isomer in this step is also problematic. We expected that, if
to the luciferase activity (cps/µg protein) of untransfected the cholesteryl carboxamide moiety (cholesteryl-CONHR) can
cells.
be replaced by a cholesterol-type structure (cholesteryl-OR),
Flow Cytometric Analysis MCF-7-Luc cells were pre- the synthesis costs would be considerably reduced. Thus, we
pared by plating on 6-well culture dishes at a density of 3×105 designed cholesteryl carbamates, (2-((2-hydroxyethyl)amino)-
cells per well 24h prior to each experiment. Each nanoparticle ethyl)carbamate (OH-C-Chol) and (2-((2-((2-hydroxyethyl)-
and liposome was mixed with 50pmol AF-siRNA at a charge amino)ethyl)amino)ethyl)carbamate (OH-NC-Chol) (Fig. 1).
ratio (+/−) of 7/1. The nanoplexes and lipoplexes were then These carbamates of diamine and triamine can be easily
diluted in 1mL of medium containing 10% FBS and the mix- prepared from cholesteryl chloroformate (Figs. 2, 3), which
ture was added to the cell. After 3-h incubation, the dish was is easily available on the order of one hundred grams from
washed 2 times with 1mL of PBS to remove any unbound several companies.
nanoplex. The amount of AF-siRNA in the cells was deter-
Preparation of Nanoplexes and Lipoplexes First, we
mined by examining fluorescence intensity on a FACSCalibur examined formulation of cationic cholesterol-based liposomes
flow cytometer (Becton Dickinson, San Jose, CA, U.S.A.), as and nanoparticles for siRNA delivery. Here, we used OH-
previously described.13)
Chol, OH-C-Chol and OH-NC-Chol as cationic cholesterol
Biodistribution of siRNA after Intravenous Injection derivatives for the preparation of cationic liposomes and
of Cationic Nanoplex and Lipoplex to Mice All animal nanoparticles. Cationic cholesterol derivatives can self-assem-
experiments were performed with approval from the Institu- ble into cationic nanoparticles without any helper lipids. Pre-
tional Animal Care and Use Committee of Hoshi University. viously, we reported that cationic nanoparticles composed of
Cationic nanoplexes or lipoplexes of 50µg of Cy5.5-siRNA cationic cholesterol derivative and Tween 80 could efficiently
were intravenously administered via lateral tail veins into deliver siRNA into tumor cells.9,13) Therefore, as cationic
female BALB/c mice (8 weeks of age; Sankyo Lab Service cholesterol derivative-based nanoparticles, we prepared NP,
Corp., Tokyo, Japan). One hour after injection, the mice were NP-C and NP-NC composed of OH-Chol/Tween 80, OH-C-

34
Biol. Pharm. Bull.
Vol. 38, No. 1 (2015)
Table 1. Particle Size and ζ-Potential of Cationic Cholesterol-Based Nanoplexes and Lipoplexes at Various Charge Ratios (+/−)
Formulation Charge ratio
ζ-Potentiala)
Formulation Charge ratio
ζ-Potentiala)
Nanoparticle
Sizea) (nm)
111.7 1.6
Liposome
Sizea) (nm)
(molar ratio)
(+/−)
(mV)
(molar ratio)
(+/−)
(mV)
Without
siRNA
Without
siRNA
45.7 5.7
98.8 4.3
49.6 6.4
1
3
Aggregation
237.5 20.9
182.5 11.4
165.5 3.1
190.5 9.9
N.D.
1
3
173.9 11.0 −37.4 4.4
OH-Chol/
Tween 80
=95/5
OH-Chol/
DOPE
=60/40
40.0 3.8
47.5 1.7
46.9 3.2
47.9 1.4
237.5 21.0
283.2 32.2
237.6 9.0
221.5 12.0
43.5 1.2
40.5 2.1
43.3 1.8
44.7 1.5
NP
LP
4
4
7
7
10
10
Without
siRNA
Without
siRNA
115.5 1.1
51.0 1.7
70.6 0.1
53.0 0.8
1
3
184.3 8.8
226.7 2.1
208.3 14.1
186.4 8.0
174.4 3.8
−30.2 0.6
47.1 3.9
48.7 1.0
47.4 1.1
48.8 1.3
1
3
168.9 0.6
−47.5 0.6
N.D.
OH-C-Chol/
Tween 80
=95/5
OH-C-Chol/
DOPE
=60/40
Aggregation
Aggregation
174.0 14.0
162.3 11.0
NP-C
LP-C
4
4
N.D.
7
7
43.0 0.2
46.1 1.1
10
10
Without
siRNA
Without
siRNA
116.3 2.7
49.3 6.3
109.0 1.6
38.7 5.1
1
3
271.6 7.8
211.0 6.4
185.2 3.1
162.2 5.6
172.4 15.5
4.6 0.3
35.3 0.4
37.2 2.9
42.5 1.0
44.9 6.4
1
3
181.2 18.3 −18.6 9.7
OH-NC-Chol/
DOPE
OH-NC-Chol/
Tween 80
=95/5
Aggregation
N.D.
NP-NC
LP-NC
=60/40
4
4
207.2 6.7
33.4 0.7
38.6 1.1
40.0 2.0
7
7
247.0 24.3
217.4 10.5
10
10
a) In water. N.D.: not determined. Values represent mean S.D. (n=3).
Fig. 4. Effect of Charge Ratio (+/−) in the Formation of NP Nanoplex and LP Lipoplex of siRNA on Gene Silencing Activity in MCF-7-Luc Cells
NP nanoplexes and LP lipoplexes were formed at charge ratios (+/−) of 1/1, 3/1, 4/1, 7/1 and 10/1. They were added to MCF-7-Luc cells at 50nM siRNA, and the lucif-
erase assay was carried out 48h after incubation of the nanoplexes and lipoplexes. Each column represents the mean S.D. (n=3). **p<0.01, compared with Cont siRNA.
Chol/Tween 80 and OH-NC-Chol/Tween 80 at a molar ratio +50, 53 and 39mV, respectively (Table 1). When the nano-
of 95/5, respectively. In liposomal formulation, cationic lipo- plexes and lipoplexes were prepared with siRNA at various
somes composed OH-Chol and DOPE were previously used charge ratios (+/−), large-sized aggregates were found at
for gene delivery.15,16) Therefore, as cationic cholesterol deriv- around neutral ζ-potential in NP, LP-C and LP-NC. However,
ative-based liposomes, LP, LP-C and LP-NC were prepared beyond a charge ratio (+/−) of 7/1, their sizes were about
from OH-Chol/DOPE, OH-C-Chol/DOPE and OH-NC-Chol/ 200nm and their ζ-potentials were about 39−47 mV.
DOPE at a molar ratio of 3/2, respectively.
Effect of Charge Ratio (+/−) for Formation of Nanoplex
The sizes of NP, NP-C and NP-NC were about 110–120nm, and Lipoplex on Gene Knockdown Efficacy We examined
and their ζ-potentials were 46, 51 and 49mV, respectively the effect of charge ratio (+/−) on the gene knockdown effect
(Table 1). In contrast, the sizes of LP, LP-C and LP-NC were by the nanoplex or the lipoplex at a concentration of 50nM
about 100, 70 and 110nm, and their ζ-potentials were about siRNA using a luciferase assay system with MCF-7-Luc cells.

Vol. 38, No. 1 (2015)
Biol. Pharm. Bull.
35
Fig. 5. Effect of Charge Ratio (+/−) in the Formation of NP-C Nanoplex and LP-C Lipoplex of siRNA on Gene Silencing Activity in MCF-7-Luc
Cells
NP-C nanoplexes and LP-C lipoplexes were formed at charge ratios (+/−) of 1/1, 3/1, 4/1, 7/1 and 10/1. They were added to MCF-7-Luc cells at 50nM siRNA, and the
luciferase assay was carried out 48h after incubation of the nanoplexes and lipoplexes. Each column represents the mean S.D. (n=3). *p<0.05 and **p<0.01, compared
with Cont siRNA.
Fig. 6. Effect of Charge Ratio (+/−) in the Formation of NP-NC Nanoplex and LP-NC Lipoplex of siRNA on Gene Silencing Activity in MCF-7-Luc
Cells
NP-NC nanoplexes and LP-NC lipoplexes were formed at charge ratios (+/−) of 1/1, 3/1, 4/1, 7/1 and 10/1, respectively. They were added to MCF-7-Luc cells at 50nM
siRNA, and the luciferase assay was carried out 48h after incubation of the nanoplexes and lipoplexes. Each column represents the mean S.D. (n=3). *p<0.05 and
**p<0.01, compared with Cont siRNA.
For OH-Chol-based formulae, NP nanoplex of Luc siRNA did erase activity were observed in transfection of Cont siRNA by
not exhibit suppression of luciferase activity at any charge LP, NP-NC or LP-C compared with untreatment, indicating
ratio (+/−) (Fig. 4A). However, for LP lipoplex, moderate sup- that their lipoplexes or nanoplexes might induce changes in
pression of luciferase activity was observed beyond a charge expression of luciferase mRNA by a target-independent fash-
ratio (+/−) of 3/1 (Fig. 4B). For OH-C-Chol-based formulae, ion (off-target effect). The liposomes or nanoparticles without
NP-C nanoplex of Luc siRNA moderately suppressed lucifer- siRNA did not significantly affect the luciferase activities
ase activity (Fig. 5A), and LP-C lipoplex could strongly sup- with the exception of the same amount of NP and LP with
press luciferase activity at charge ratios (+/−) of 1/1, 3/1, 4/1 NP nanoplex and LP lipoplex at charge ratios (+/−) of 10/1
and 7/1, compared with those of Cont siRNA (Fig. 5B). For and 1/1, respectively (Supplemental Fig. S2). From the results
OH-NC-Chol-based formulae, NP-NC nanoplex of Luc siRNA of lipoplex size and gene silencing effects, LP-C and LP-NC
exhibited slight suppression of luciferase activities at charge lipoplex formed at a charge ratio (+/−) of 7/1 exhibited high
ratios (+/−) of 3/1 and 4/1 (Fig. 6A). In contrast, LP-NC lipo- gene knockdown by siRNA without aggregation. Therefore,
plex could strongly suppress luciferase activity at all charge in subsequent experiments, we used 7/1 as an optimal charge
ratios (+/−) (Fig. 6B). However, the slight reductions of lucif- ratio (+/−) for siRNA transfection.

36
Biol. Pharm. Bull.
Vol. 38, No. 1 (2015)
Fig. 7. Dose Dependence of siRNA-Mediated Inhibition of Luciferase Activity 48h after Incubation with LP, LP-C and LP-NC Lipoplexes Prepared
at a Charge Ratio (+/−) of 7/1
MCF-7-Luc cells were treated with each lipoplex at a final concentration of 20, 10 and 5nM siRNA. Each value represents the mean S.D. (n=3). *p<0.05 and
**p<0.01, compared with Cont siRNA.
Fig. 8. Cellular Association at 3h after Transfection of Nanoplex and Lipoplex
NP nanoplex and LP lipoplex (A), NP-C nanoplex and LP-C lipoplex (B), NP-NC nanoplex and LP-NC lipoplex (C), were formed by mixing with AF-siRNA at a charge
ratio (+/−) of 7/1. The association of nanoplex or lipoplex with MCF-7 cells was determined on the basis of Alexa Fluor®488-fluorescence by flow cytometry.
The persistence of gene silencing is a key factor when of siRNA, LP-NC lipoplex could significantly suppress lucif-
considering the therapeutic uses of siRNA. We thus studied erase activity (39% of untreated cells) even at 5nM siRNA,
the dose dependence of inhibition produced by LP, LP-C and compared with that of Cont siRNA (Fig. 7C). Furthermore,
LP-NC lipoplexes (Figs. 7A–C). In terms of the concentration Cont siRNA transfected by LP, LP-C or LP-NC at low concen-

Vol. 38, No. 1 (2015)
Biol. Pharm. Bull.
37
trations did not induce changes in luciferase activity.
indicating that a carbamate-type linker might affect proton-
Association of Nanoplexes with the Cells To clarify the ation of aminoethylene units of OH-NC-Chol in the endosome.
relationship between gene silencing effect and cellular uptake However, for liposomal formulae, LP-NC lipoplex could in-
in the transfection of nanoplexes and lipoplexes, we examined duce a larger gene silencing effect than LP-C lipoplex at low
the cellular association by flow cytometric analysis. Cellular concentrations of siRNA (Figs. 7B, C). These findings sug-
associations with LP and LP-NC lipoplexes were higher than gested that liposomal formulation might improve the double
with NP and NP-NC nanoplexes, respectively (Figs. 8A, C). protonation of the aminoethylene units of OH-NC-Chol and
This result corresponds to those of gene knockdown effica- increase gene silencing effect in the cells by LP-NC lipoplex.
cies (Figs. 4, 6). In contrast, NP-C nanoplex exhibited higher However, in this study, we have no evidence that LP-NC
cellular association than LP-C lipoplex (Fig. 8B), although lipoplex could facilitate endosomal escape after internaliza-
the gene silencing effect by NP-C nanoplex was lower than tion into the cells; therefore, further study must be performed
that by LP-C lipoplex (Figs. 5A, B). This might indicate that to investigate the mechanism to improve the gene silencing
the cellular association could be increased by the inclusion effect.
of OH-C-Chol into the nanoparticles, but siRNA transduced
In a preliminary study, we intravenously injected the
by NP-C could not be efficiently released from the nanoplex. lipoplexes or nanoplexes of Cy5.5-siRNA into mice and ob-
From the results, the LP-NC composed of OH-NC-Chol and served the biodistribution of siRNA 1h after injection by
DOPE could deliver siRNA via efficient cellular association, fluorescent microscopy (Supplemental Fig. S3). Injections of
and induce a large gene silencing effect by siRNA.
NP and NP-C nanoplex and LP-C lipoplex produced the ac-
Previously, we reported that NP nanoplex formed in NaCl cumulation of siRNA both in lung and in liver, and those of
solution exhibited marked RNAi activity in human prostate LP and LP-NC lipoplexes, and NP-NC nanoplex in the liver.
tumor PC-3 cells.8,9) In vitro transfection activity increased These findings indicate that their nanoparticles and liposomes
in parallel with ζ-potential and the size of the lipoplex.17,18) In might be useful for siRNA delivery to lung or liver tumor
plasmid DNA transfection, large lipoplexes over 700nm in metastasis. Further in vivo application should be performed to
mean diameter induced efficient transfection; in contrast, lipo- examine the gene knockdown effect in tumors by intravenous
plexes of less than 250nm were inefficient.19) The presence of injection of the nanoplex or lipoplex.
NaCl during the formation of NP nanoplex increased the size
In this study, we synthesized new cholesteryl derivatives
of the nanoplex (ca. µm in size) and affected the gene silenc- with a carbamate-linked spacer as cationic lipids for lipid-
ing effect by siRNA.8,9) However, in this study, the NP nano- based siRNA vectors. LP-NC composed of OH-NC-Chol and
plex was formed in water to prepare small nanoplexes (Table DOPE could form siRNA lipoplexes of 200nm in size, and
1); therefore, the NP nanoplex might not be able to induce could induce a high level of gene knockdown in the cells.
a gene knockdown effect (Fig. 4A). In contrast, LP lipoplex These findings suggest that OH-NC-Chol may have potential
could induce the gene silencing effect in the cells (Fig. 4B). as a cationic lipid to transfect siRNA efficiently into tumor
To overcome poor endosomal escape and subsequent degrada- cells.
tion of siRNA in the late lysosomes, pH-sensitive lipids such
as DOPE have been used as a component of liposomal siRNA
Acknowledgments This project was supported in part
vectors.20,21) The role of DOPE is not fully understood, but it by a Grant-in-Aid for Scientific Research (C), Japan Soci-
might affect the structural transition of cationic nanoparticles ety for the Promotion of Science (KAKENHI Grant Num-
at the acidic pH of late endosomes in the cells.22) Although the ber 26460046), and the Science Research Promotion Fund
sizes and ζ-potentials of LP lipoplexes were similar with those
of NP nanoplex (Table 1), LP lipoplex exhibited gene silenc-
ing effect in the cells (Fig. 4). These findings indicate that
LP lipoplex might increase the gene silencing effect by the
improvement of intracellular trafficking.
from the Promotion and Mutual Aid Corporation for Private
Schools of Japan.
Conflict of Interest The authors declare no conflict of
interest.
Upon comparing the transfection efficiency between OH-
Chol- and OH-C-Chol-based liposomes, LP-C lipoplex exhib-
Supplementary Materials The online version of this ar-
ited a larger gene silencing effect than LP lipoplex. In plasmid ticle contains supplementary materials.
DNA transfection, it has been reported that the linker of cat-
ionic cholesterol derivatives affected transfection efficiency by
cationic lipoplex.9,23) However, it was not clear why the OH-C-
Chol with a carbamate-type linker could induce higher gene
suppression by siRNA than OH-Chol with a carboxamide-type
linker.
To facilitate the endosomal escape of siRNA transfected by
cationic nanoparticles, we prepared NP-NC that had amino-
ethylene units on the surface of the nanoparticles; however,
NP-NC nanoplex could not induce a silencing effect in the
cells (Fig. 6A). Previously, we reported that NP-N composed
of OH-N-Chol and Tween 80 could efficiently deliver siRNA
into cytoplasm via efficient endosomal escape and induce a
large gene silencing effect.13) The only difference between
OH-N-Chol and OH-NC-Chol is in the structure of the linker,
REFERENCES
1) Pai SI, Lin YY, Macaes B, Meneshian A, Hung CF, Wu TC. Pros-
pects of RNA interference therapy for cancer. Gene Ther., 13,
464–477 (2006).
2) Kim DH, Behlke MA, Rose SD, Chang MS, Choi S, Rossi JJ. Syn-
thetic dsRNA Dicer substrates enhance RNAi potency and efficacy.
Nat. Biotechnol., 23, 222–226 (2005).
3) Ozpolat B, Sood AK, Lopez-Berestein G. Liposomal siRNA nano-
carriers for cancer therapy. Adv. Drug Deliv. Rev., 66, 110–116
(2014).
4) Gao Y, Liu XL, Li XR. Research progress on siRNA delivery with
nonviral carriers. Int. J. Nanomedicine, 6, 1017–1025 (2011).
5) Inoh Y, Furuno T, Hirashima N, Kitamoto D, Nakanishi M. Rapid
delivery of small interfering RNA by biosurfactant MEL-A-con-

38
Biol. Pharm. Bull.
Vol. 38, No. 1 (2015)
taining liposomes. Biochem. Biophys. Res. Commun., 414, 635–640
(2011).
with herpes simplex virus thymidine kinase. Cancer Gene Ther.,
12, 796–809 (2005).
6) Ivanova EA, Maslov MA, Kabilova TO, Puchkov PA, Alekseeva
AS, Boldyrev IA, Vlassov VV, Serebrennikova GA, Morozova NG,
Zenkova MA. Structure–transfection activity relationships in a
series of novel cationic lipids with heterocyclic head-groups. Org.
Biomol. Chem., 11, 7164–7178 (2013).
7) Maslov MA, Morozova NG, Chizhik EI, Rapoport DA, Ryabchiko-
va EI, Zenkova MA, Serebrennikova GA. Synthesis and delivery
activity of new cationic cholesteryl glucosides. Carbohydr. Res.,
345, 2438–2449 (2010).
15) Hasegawa S, Hirashima N, Nakanishi M. Comparative study
of transfection efficiency of cationic cholesterols mediated by
liposomes-based gene delivery. Bioorg. Med. Chem. Lett., 12,
1299–1302 (2002).
16) Igarashi S, Hattori Y, Maitani Y. Biosurfactant MEL-A enhances
cellular association and gene transfection by cationic liposome. J.
Control. Release, 112, 362–368 (2006).
17) Ross PC, Hui SW. Lipoplex size is a major determinant of in vitro
lipofection efficiency. Gene Ther., 6, 651–659 (1999).
8) Hattori Y, Yoshizawa T, Koga K, Maitani Y. NaCl induced high
cationic hydroxyethylated cholesterol-based nanoparticle-mediated
synthetic small interfering RNA transfer into prostate carcinoma
PC-3 cells. Biol. Pharm. Bull., 31, 2294–2301 (2008).
18) Takeuchi K, Ishihara M, Kawaura C, Noji M, Furuno T, Nakanishi
M. Effect of zeta potential of cationic liposomes containing cat-
ionic cholesterol derivatives on gene transfection. FEBS Lett., 397,
207–209 (1996).
9) Hattori Y, Hagiwara A, Ding W, Maitani Y. NaCl improves siRNA
delivery mediated by nanoparticles of hydroxyethylated cationic
cholesterol with amido-linker. Bioorg. Med. Chem. Lett., 18, 5228–
5232 (2008).
10) Zintchenko A, Philipp A, Dehshahri A, Wagner E. Simple modifica-
tions of branched PEI lead to highly efficient siRNA carriers with
low toxicity. Bioconjug. Chem., 19, 1448–1455 (2008).
11) Hattori Y, Kubo H, Higashiyama K, Maitani Y. Folate-linked
nanoparticles formed with DNA complexes in sodium chloride solu-
tion enhance transfection efficiency. Journal of Biomedical Nano-
technology, 1, 176–184 (2005).
19) Turek J, Dubertret C, Jaslin G, Antonakis K, Scherman D, Pitard B.
Formulations which increase the size of lipoplexes prevent serum-
associated inhibition of transfection. J. Gene Med., 2, 32–40 (2000).
20) Spagnou S, Miller AD, Keller M. Lipidic carriers of siRNA: differ-
ences in the formulation, cellular uptake, and delivery with plasmid
DNA. Biochemistry, 43, 13348–13356 (2004).
21) Mével M, Kamaly N, Carmona S, Oliver MH, Jorgensen MR,
Crowther C, Salazar FH, Marion PL, Fujino M, Natori Y, Thanou
M, Arbuthnot P, Yaouanc JJ, Jaffres PA, Miller AD. DODAG; a
versatile new cationic lipid that mediates efficient delivery of pDNA
and siRNA. J. Control. Release, 143, 222–232 (2010).
12) Dioury F, Sambou S, Guéné E, Sabatou M, Ferroud C, Guy
A, Port M. Synthesis of a tricyclic tetraazatriacetic ligand for
gadolinium(III) as potential contrast agent for MRI. Tetrahedron,
63, 204–214 (2007).
13) Hattori Y, Nakamura T, Ohno H, Fujii N, Maitani Y. siRNA de-
livery into tumor cells by lipid-based nanoparticles composed of
hydroxyethylated cholesteryl triamine. Int. J. Pharm., 443, 221–229
(2013).
22) Mochizuki S, Kanegae N, Nishina K, Kamikawa Y, Koiwai K, Ma-
sunaga H, Sakurai K. The role of the helper lipid dioleoylphospha-
tidylethanolamine (DOPE) for DNA transfection cooperating with
a cationic lipid bearing ethylenediamine. Biochim. Biophys. Acta,
1828, 412–418 (2013).
23) Kawaura C, Noguchi A, Furuno T, Nakanishi M. Atomic force mi-
croscopy for studying gene transfection mediated by cationic lipo-
somes with a cationic cholesterol derivative. FEBS Lett., 421, 69–72
(1998).
14) Hattori Y, Maitani Y. Folate-linked nanoparticle-mediated suicide
gene therapy in human prostate cancer and nasopharyngeal cancer