An Approach to Transgene Expression in Liver Endothelial Cells

Using a Liposome-Based Gene Vector Coated with Hyaluronic

Acid
YUMA YAMADA,1 MASAHIRO HASHIDA,1 YASUHIRO HAYASHI,2 MAI TABATA,1 MAMORU HYODO,2 MST. NAZNIN ARA,2
NORITAKA OHGA,3 KYOKO HIDA,3 HIDEYOSHI HARASHIMA1,2
1
Laboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812,
Japan
2
Laboratory of Innovative Nanomedicine, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan
3
Division of Vascular Biology, Graduate School of Dental Medicine, Hokkaido University, Sapporo 060-0812, Japan

Received 14 December 2012; accepted 01 February 2013

Published online 7 March 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23480

ABSTRACT: Dysfunctional sinusoidal liver endothelial cells (LECs) are associated with liver
diseases, such as liver fibrosis, cirrhosis, and portal hypertension. Because of this, gene ther-
apy targeted to LECs would be a useful and productive strategy for directly treating these
diseases at the level of genes. Here, we report on the development of a transgene vector that
specifically targets LECs. The vector is a liposome-based gene vector coated with hyaluronic
acid (HA). HA is a natural ligand for LECs and confers desirable properties on particles, ren-
dering them biodegradable, biocompatible, and nonimmunogenic. In this study, we constructed
HA-modified carriers, and evaluated cellular uptake and transfection activity using cultured
LECs from KSN nude mice (KSN–LECs). Cellular uptake analyses showed that KSN–LECs
recognized the HA-modified carriers more effectively than skin endothelial cells. The transfec-
tion assay indicated that the efficient gene expression in KSN–LECs, using the HA-modified
carriers, required an adequate lipid composition and a functional device to control intracel-
lular trafficking. This finding contributes to our overall knowledge of transgene expression
targeted to LECs. © 2013 Wiley Periodicals, Inc. and the American Pharmacists Association
J Pharm Sci 102:3119–3127, 2013
Keywords: hyaluronic acid; liver endothelial cell; multifunctional envelope-type nano device;
transgene expression; non-viral gene vector; DNA delivery; liposomes; nanotechnology; plasmid
DNA; transfection

INTRODUCTION reported that the defenestration of LECs leads to the

development of hyperlipidemia because it is difficult
Dysfunction of the liver appears to be associated
for lipoproteins to reach hepatocytes, which are the
with the development of various diseases because
primary cells involved in lipoprotein uptake.2 Thus,
the organ possesses various functions including glyco-
an effective drug delivery system would be highly
gen storage, hormone production, and clearing toxic
desirable in terms of the treatment of these LECs-
molecules from the body. It has recently been demon-
related diseases.
strated that a functional defect in sinusoidal liver en-
Hyaluronic acid (HA) is a naturally occurring,
dothelial cells (LECs) is associated with liver fibrosis,
biodegradable, biocompatible, and nonimmunogenic
cirrhosis, and portal hypertension because the sinu-
anionic biopolymer and presents a wide range of an-
soidal vascular network is disrupted.1 It has also been
imal tissues from vertebrates to bacteria.3 When ex-
Additional Supporting Information may be found in the online ogenous HAs are injected intravenously, these are
version of this article. Supporting Information rapidly removed from the blood and mainly accu-
Correspondence to: Hideyoshi Harashima (Telephone: +81- mulate in LECs4,5 because these cells express HA
11-706-3919; Fax: +81-11-706-4879; E-mail: harasima@pharm
.hokudai.ac.jp)
receptors. To date, HA has been widely used as
Journal of Pharmaceutical Sciences, Vol. 102, 3119–3127 (2013)
a targeted delivery material for LECs. Polycations,
© 2013 Wiley Periodicals, Inc. and the American Pharmacists Association such as poly-L-lysine6 and polyethyleneimine,7,8 were

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013 3119

3120 YAMADA ET AL.
covalently conjugated with HA, and these materials
enabled the target-specific delivery of HA to recep-
tor positive cells in vitro7,8 and to LECs in vivo.6
We also reported that cationic liposomes modified
with HA–stearylamine effectively accumulate in the
liver when intravenously injected, whereas accumu-
lation in the lung was drastically reduced compared
with nonmodified liposomes.9 Moreover, fluorescent
microscopy observations showed that HA-modified li-
posomes preferentially accumulated in close proxim-
ity to blood vessels compared with nonmodified lipo-
somes, suggesting that the HA-modified liposomes are
ultimately distributed in LECs.
However, gene expression using an HA-coated
gene vector was not investigated, although the HA-
modified liposomes were delivered to LECs in vivo.
To achieve gene expression using a gene vector, af-
ter reaching the LECs, various phases of intracellu-
lar trafficking, including cellular uptake, endosomal
escape, and nuclear, must be regulated. We were con-
cerned that a gene vector with an anionic HA coating
might have a low transfection activity because, gen-
erally, the transfection activity of anionic carriers is
lower than that of cationic carriers, as control of the
intracellular trafficking was difficult. Thus, we con-
cluded that investigating gene expression in LECs
using the HA-coated gene vector is an important is-
sue in terms of optimizing transfection activity.
The objective of this study is to develop a liposome-
based gene vector coated with HA for the efficient
transgene expression in the LECs. The characteris-
tics of carriers used in this study are summarized
in Table 1. We first investigated the cellular uptakes
of carriers by cultured LECs (KSN–LECs) to vali-
date whether this cell line is optimal model to evalu-
ate transgene expression in the LECs via HA recep-
tor. In this experiment, we prepared HA-conjugated
polyethylene glycol liposome (HA–PEG–LP), where
PEG is present on the carrier surface as a linker. We
next attempted to package plasmid DNA (pDNA) in
HA-modified carriers by a multifunctional envelope-
type nanodevice (MEND) preparation technique. The
MEND consists of DNA particles condensed with a
polycation and a lipid envelope equipped with vari-
ous functional devices.10–14 Furthermore, we report
on an investigation of the optimal lipid composition
and functional device needed for making strong trans-
gene expression in LECs.
EXPERIMENTAL
Materials
The pDNA encoding enhanced green fluorescent
protein and luciferase protein was obtained from BD
Bioscience Clontech (Palo Alto, California). pDNA
was purified using a Qiagen EndoFree Plasmid
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
Mega Kit (Qiagen GmbH, Hilden, Germany). Egg
yolk phosphatidyl choline (EPC) was obtained from
Nippon Oil and Fats Company (Tokyo, Japan).
Cholesteryl hemisuccinate (CHEMS) and phospha-
tidic acid (PA) were purchased from Sigma (St. Louis,
Mossouri). Cholesterol (Chol), 1,2-Dioleoyl-sn-
glycero-3-phosphoethanolamine (DOPE), 7-
nitrobenz-2-oxa-1,3-diazole-labeled DOPE (NBD-
DOPE), and 1,2-distearoyl-sn-glycero-3-
phosphoethanolamine-N-[PEG(2000) maleimide]
[DSPE-PEG(2000) maleimide] were purchased from
Avanti Polar Lipids Inc. (Alabaster, Alabama).
Protamine was purchased from Calbio Chem (Darm-
stadt, Germany). Stearyl octaarginine (STR-R8)15,16
and cholesteryl GALA peptide17,18 were obtained
from Kurabo Industries Ltd. (Osaka, Japan). HA
[average molecular weight (MW) 500–700 kDa, HA
(600 kDa)] was obtained from Food Chemifa (Tokyo,
Japan). HA preparations with different molecular
weights, labeled with fluorescein [average MW
20–30 kDa, HA (25 kDa); average MW 100–300 kDa,
HA (200 kDa); average MW 600–1120 kDa, HA
(900 kDa)] were purchased from PG Research (Tokyo,
Japan). Hoechst 33342 was purchased from Dojindo
Laboratories (Kumamoto, Japan). Microvascular En-
dothelial Cell Growth Medium-2 Bullet Kit (EGM-2
MV) was purchased from Clonetics (Walkersville,
Maryland). All other chemicals were commercially
available reagent-grade products.
Synthesis of the HA Derivative
A 100-mL, one-necked, round-bottomed flask
equipped with a magnetic stirring bar was loaded
with 10 mg of HA (600 kDa) (16.6 nmol) in 25 mL of
water. 4.79 mg of 1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide (EDC) hydrochloride (25.0 :mol) and
2.88 mg of N-hydroxysuccinimide (NHS) (25.0 :mol)
was added to the reaction mixture, after which
5.63 mg of cystamine dihhydrochloride (25 :mol)
was added. The solution was then vigorously stirred
overnight at 25◦ C. The resulting solution was dia-
lyzed using a 5000 molecular weight cut-off dialytic
membrane (Spectrum Lab, California) against 1 L
of an 80:20 mixture of phosphate-buffered saline
(PBS) (−) and methanol and then three times against
water, respectively. The solvents were removed with
a centrifugal concentrator, and the white solid was
collected. The ratio for the incorporation of cystamine
was determined by 1 H NMR using D2 O as the
solvent.
Preparation of HA–PEG–LPs Labeled with NBD
Liposomes labeled with NBD were prepared by the
lipid film hydration method. Lipid films were pro-
duced on the bottom of a glass tube by the evaporation
of a chloroform solution containing 138 nmol lipids
[EPC/Chol/DSPE-PEG(2000) maleimide/NBD-DOPE
DOI 10.1002/jps
 TRANSGENE EXPRESSION IN LECS USING HA-COATED GENE VECTORS 3121

Table 1. Composition and Property of Carriers

Modification pDNA
HA R8 GALA Lipid Compositions Packaged Assay
PEG–LP PEG–LP − − − DOPE/Chol/DSPE–PEG(2000) − Cellular uptake
maleimide (7:3:1, molar ratio) (Fig. 3)
HA–PEG–LP + − −
PEG–MEND PEG–MEND − − − DOPE/CHEMS/DSPE –PEG(2000) + Transfection activity
maleimide (9:2:1, molar ratio) (Fig. S2)
HA–PEG–MEND + − −
MEND (CHEMS) MEND (CHEMS) − − − DOPE/CHEMS (9:2, molar ratio) + Transfection activity
(Fig. 4a,c)
R8–MEND (CHEMS) − + −
GALA–MEND (CHEMS) − − +
R8/GALA–MEND − + +
(CHEMS)
HA–R8–GALA–MEND + + +
(CHEMS)
MEND (PA) MEND (PA) − − − DOPE/PA (7:2, molar ratio) + Transfection activity
(Fig. 4b)
R8–MEND (PA) − + −
GALA–MEND (PA) − − +
R8–GALA–MEND (PA) − + +
HA–R8–GALA–MEND (PA) + + +

= 7:3:1:0.2 (molar ratio)]. Next, 250 :L of 10 mM
hydroxyethyl piperazineethanesulfonic acid (HEPES)
buffer (pH 7.4) was applied to the lipid film, followed
by incubation for 15 min at room temperature to hy-
drate the lipids. The lipid suspension was then soni-
cated for approximately 1 min in a bath-type sonicator
(85 W; Aiwa Company, Tokyo, Japan). For the reduc-
tion of maleimide groups on the liposome surface, a
solution of tris(2-carboxyethyl)phosphine hydrochlo-
ride (TCEP-HCl) was added to 100 :L of the lipo-
some suspension (final concentration of TCEP-HCl,
10 mM), and the suspension was incubated for 30 min
at room temperature. To attach HA to the maleimide
group of PEG on the liposome surface, a solution
of HA (600 kDa) (0.1 mol % lipids) was added, and
the resulting suspension was incubated for 30 min at
room temperature to form HA-modified PEGylated
liposomes (HA–PEG–LPs). The characteristics and
physicochemical properties of these carriers are sum-
marized in Tables 1 and S1.
Isolation of Liver Endothelial Cells and Skin Endothelial
Cells
Endothelial cells were isolated as previously
described.19–21 In brief, KSN–LECs that overexpress
HA receptors such as CD44 were isolated from the
livers of KSN nude mice. Skin endothelial cells
(KSN–SECs) were isolated from dermal tissue as
a CD44-negative cell line. ECs were isolated us-
ing a magnetic cell sorting system (Miltenyi Biotec,
Tokyo, Japan) according to the manufacturer’s in-
structions using an anti-mouse CD31 antibody (BD
Bioscience Clontech). CD31-positive cells were sorted
and plated on 1.5% gelatin-coated culture plates and
grown in EGM-2 MV and 15% FBS. KSN–LECs
and KSN–SECs were maintained in EGM-2MV me-
dia supplemented with 5% FBS, penicillin (100 units/
mL), and streptomycin (100 :g/mL). The purity of the
KSN–LECs after the isolation was confirmed by BS1-
B4 lectin (an endothelial marker) binding and the ex-
pression of endothelial cell’s marker (Fig. S1). The
results indicated that this isolated LECs were highly
pure.
Quantification of the Cellular Uptake of the HA-Coated
Liposomes by Flow Cytometry
KSN–LECs or KSN–SECs (2 × 105 cells) were seeded
on a six-well plate (Corning, New York) with EGM-2
MV containing 5% FBS under an atmosphere of 5%
CO2 /air at 37◦ C for 24 h. After washing the cells with
PBS (−), HA–PEG–LPs labeled with NBD-DOPE sus-
pended in serum-free EGM-2 MV were added to the
cells (final lipid concentration, 55 :M). The cells were
then incubated for 1 h under an atmosphere of 5%
CO2 /air at 37◦ C. After removing the medium, the cells
were washed once with ice-cold PBS (−). The cells
were then collected, suspended in EGM-2 MV with
serum isolated by centrifugation (700g, 4◦ C, 5 min)
and washed with (fluorescence activated cell sorting)
FACS buffer [0.5% bovine serum albumin and 0.1%
NaN3 in PBS (−)]. After resuspending the cells in
0.5 mL of FACS buffer, the cell suspension was filtered
through a nylon mesh to remove cell aggregates and
debris. The cells were then analyzed by flow cytometry
(FACScan; Becton Dickinson, Franklin Lakes, New
Jersey). The cells were excited with a 488- nm light
from an Ar laser for detecting NBD. The fluorescence
detection channel was set to the FL1 filter for NBD.

DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
3122 YAMADA ET AL.
Table 2. Characteristics of Various MENDs
Compositions Diameter (nm)
R8 − −
GALA − +
DOPE/CHEMS (9:2, molar ratio) 238 ± 74 183 ± 16
DOPE/PA (7:2, molar ratio) 182 ± 14 145 ± 7
Data are represented by the mean ± S.D. (n = 3–6).
The cellular uptake of the carriers was expressed as
the mean fluorescence intensity (MFI), calculated us-
ing the CellQuest software (Becton Dickinson).
Preparation of the R8–GALA–MEND and Coating the
Carriers with HA
The R8-GALA–MEND was prepared by the lipid film
hydration method, as previously reported.14 First,
condensed pDNA particles were prepared by mixing
DNA (0.1 mg/mL) and protamine in 10 mM HEPES
buffer (pH 7.4) under vortexing at a nitrogen–phos-
phate ratio of 2:2. Lipid films were produced on the
bottom of a glass tube by the evaporation of a chlo-
roform solution containing 138 nmol lipids [DOPE/
CHEMS = 9:2 (molar ratio) or DOPE/PA, = 7:2 (mo-
lar ratio)]. To equip the MEND with GALA, a so-
lution of Chol–GALA (2 mol % lipids) solution was
added to the lipid solvent. Next, 250 :L of the con-
densed pDNA particle solution was applied to the
lipid film, followed by incubation for 15 min at room
temperature to hydrate the lipids. The lipid film was
then sonicated for approximately 1 min in a bath-
type sonicator to produce the desired MEND. To at-
tach R8 to the surface of the carrier, a solution of
STR-R8 (10 mol % lipids) was incubated with the
MEND for 30 min at room temperature to produce
the R8–GALA–MEND. To coat the R8–GALA–MEND
with HA via electrostatic interaction, a solution of
HA (600 KDa) (0.1 mol % of lipids) was added to
the R8–GALA–MEND suspensions, and the result-
ing suspension was then incubated for 30 min at room
temperature (HA–R8–GALA–MEND). The character-
istics and physicochemical properties of these carriers
used in this study are summarized in Tables 1 and 2.
DNA Transfection Assay
Transfection activity was assessed by measuring
luciferase activity, as described below. KSN–LECs
(4 × 104 cells) were seeded on a 24-well plate
(Corning) with EGM-2 MV containing 5% FBS un-
der an atmosphere of 5% CO2 /air at 37◦ C for
24 h. After washing the cells, samples containing
0.4 :g of pDNA suspended in 0.25 mL of serum-
free EGM-2 MV were added to the cells. After in-
cubation under 5% CO2 at 37◦ C for 3 h, the cells
were washed with PBS (−) and then further in-
cubated for 21 h in fresh medium containing 10%
serum. The cells were then washed with PBS (−),
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
ζ Potential (mV)
+ + − − + +
− + − + − +
233 ± 64 197 ± 22 −43 ± 9 −50 ± 8 61 ± 3 53 ± 8
177 ± 10 162 ± 17 −73 ± 2 −62 ± 10 53 ± 3 42 ± 4
and luciferase activity [relative light unit (RLU)]
was measured using a Luciferase Assay System
with a Reporter Lysis Buffer kit (Promega, Madison,
Wisconsin) by a luminometer (Luminescencer-PSN;
ATTO, Tokyo, Japan). Cellular protein content was
determined using a bicinchoninic acid protein assay
kit (PIERCE, Rockford, Illinois). Transfection activity
was calculated as follows:
RESULTS AND DISCUSSION
Determination of the Cellular Uptake of HA in Cultured
Liver Endothelial Cells
We first investigated the cellular uptakes of free HA
by cultured (KSN–LECs) to validate whether this cell
line is optimal model to evaluate transgene expres-
sion in the LECs via HA receptor. The cellular up-
take of HA with high molecular weight (900 kDa) us-
ing KSN–LECs and cultured skin endothelial cells
(KSN–SECs) as negative controls was visualized us-
ing confocal laser scannig microscopy (Fig. 1). The
results confirmed that HA was internalized into the
KSN–LECs (Fig. 1a), whereas no signal was detected
in the case of KSN–SECs (Fig. 1b). The results indi-
cate that HA has the ability to recognize LECs and
be internalized. These conclusions are also supported
by previous reports that LECs are able to take up
HA both in vitro and in vivo and that the extent of
hepatic accumulation was enhanced by the HA with
higher molecular weight (Fig. S2).22,9 Therefore, we
used the HA with high molecular weight to construct
the liposomes and MEND used in this experiment.
Construction of HA–PEG–LP and the Cellular Uptake
Evaluation Between KSN–LECs and KSN–SECs
Hyaluronic acid is a polymer composed of D-
glucuronic acid and N-acetyl-D-glucosamine units
connected by $-1,3-, or $-1,4-glycosidic bonds. A selec-
tive reaction is required to form a crosslink between
HA and lipid. To accomplish this, we employed a thiol-
maleimide linkage. For the incorporation of a thiol, we
used the carboxylate moiety of D-glucuronic acid and
conjugated it with the amino group of cystamine.23,24
Cystamine can be considered to be a water-soluble
protected thiol when its disulfide bond is reduced
(Fig. 2a). We reacted HA and cystamine via an amide
linkage using EDC and NHS as coupling agents and
DOI 10.1002/jps
 TRANSGENE EXPRESSION IN LECS USING HA-COATED GENE VECTORS 3123

Figure 1. Visualization of the cellar uptake of HA in LECs. LECs (a) and skin endothelial
cells (skin ECs, b) were treated with fluorescein-labeled HA (900 kDa, green), and the cellular
uptake of HA was visualized by confocal laser scanning microscopy. Scale bars, 10 :m.

this resulted in the random incorporation of cys-
tamine into HA. After purification by dialysis, we es-
timated the incorporation ratio for cystamine by 1 H
NMR spectroscopy using the signals for the N-acetyl
group of HA and the methylene group of cystamine
(Fig. 2b). We estimated that the level of incorpora-
tion of HA in the cystamine conjugate was 10% molar
ratio. The resulting preparation was soluble in water.
To construct an HA-modified nanocarrier, we
first prepared liposomes containing DSPE-PEG(2000)

Figure 2. Characteristics of HA derivative and a scheme showing the construction of

HA–EPC–LP. (a) Preparation of HA-cystamine conjugate. (b) 1 H NMR spectrum of HA-
cystamine conjugate. (c) Scheme for the HA–EPC–LP preparation.

DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
3124 YAMADA ET AL.
Figure 3. Evaluation of the cellular uptake of
HA–PEG–LP by LECs vis-a-vis skin ECs. The graph shows
the MFI of the HA–PEG–LP labeled with NBD-DOPE taken
up by cells after transduction of the carriers. Data are rep-
resented as the mean ± SD (n = 3). ∗ Significant difference
(p < 0.01 by unpaired Student t-test).
maleimide (PEG–LP) and then conjugated them with
the HA derivative (HA–PEG–LP) as shown in Fig-
ure 2c. The diameters of the prepared liposomes
were approximately 80–120 nm, and the ξ poten-
tial of the HA–PEG–LP indicated that it had a
more negative charge than that of PEG–LP (Table 1,
Table S1), suggesting that HA was modified on
the surface of PEG–LP. We next performed a cel-
lular uptake analysis of the HA–PEG–LP using
KSN–LECs and KSN–SECs. The carriers, labeled
with NBD-DOPE lipids, were added to the cells,
and the fluorescent intensity of the HA taken up by
the cells was measured by flow cytometry (Fig. 3).
The results showed that KSN–LECs internalized the
HA–PEG–LP to a greater extent than KSN–SECs.
We also investigated the cellular uptake of HA–PEG-
modified carriers in KSN–LEC in the absence and
presence of excess HA. The results indicated that the
cellular uptake of the carriers decreased with increas-
ing concentrations of HA (Fig. S3). These results sug-
gest that KSN–LEC could be a model for evaluating
transgene expression that occurs via the HA receptor.
It was expected that the use of an HA-modified
gene vector would result in efficient transgene ex-
pression in LECs. To validate this hypothesis, we
constructed an HA-modified MEND and evaluated
its transfection activity in KSN–LECs. In this ex-
periment, we first prepared a MEND that contained
a condensed pDNA and a lipid envelope [DOPE/
CHEMS/DSPE-PEG(2000) maleimide = 9:2:1 (molar
ratio)] (PEG–MEND). PEG–MEND was then conju-
gated with HA derivative (HA–PEG–MEND). The
characteristics and physicochemical properties are
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
summarized in Table 1 and Table S2. Unexpectedly,
the transfection activity of the HA–PEG–MEND was
comparable that for naked pDNA (Fig. S4a).
Optimization for Transgene Expression in KSN–LECs
Using MEND
The transfection activity of the HA–PEG–MEND
in KSN–LECs was very low (Fig. S4a), even
though HA–PEG–LP was effectively internalized by
KSN–LECs (Fig. 3). For efficient transgene expres-
sion, it is necessary to control various aspects of in-
tracellular trafficking such as cellular uptake, endo-
somal escape, and nuclear targeting, in addition to
selective cell targeting. To identify more clearly the
cause of the low transfection activity of HA–PEG–
MEND, the cellular uptake of HA–PEG–MEND and
naked pDNA was compared in KSN–LECs (Fig. S4b).
The result showed that the cellular uptake of
naked pDNA was significantly lower than that of
the HA–PEG–MEND, although the transfection ac-
tivities of the two preparations were comparable
(Fig. S4a). These findings suggest that the low trans-
fection activity of HA–PEG–MEND in KSN–LEC can
be mainly attributed to the inefficient release of carri-
ers from endosomes. Thus, we investigated the trans-
fection activity of non-PEGylated MEND using lipid
membranes containing PA or CHEMS, which are
optimal lipid compositions for packaging pDNA in
MEND.10,14 The characteristics and physicochemical
properties of the prepared MENDs are summarized
in Tables 1 and 2.
We evaluated the transfection activity of various
types of MENDs in KSN–LECs (Fig. 4). The plain
MEND (CHEMS) showed a higher transfection activ-
ity than the PEG–MEND (less than 105 RLU/mg pro-
tein in Fig. S4a), indicating that PEGylation inhibits
the transgene expression of the MEND. Moreover,
integration of octaarginine (R8), a cellular uptake
device,11,12,16,25 and GALA, a pH-sensitive fusogenic
peptide,17,18 into the MEND (R8–GALA–MEND) dra-
matically enhanced the transfection activity of the
MEND (Fig. 4a). We also confirmed the same tendency
in the case of a MEND containing PA [MEND (PA)]
(Fig. 4b), although the value for the transfection activ-
ity of R8–GALA–MEND (PA) (1.68 × 109 RLU/mg pro-
tein) was lower than that for the R8–GALA–MEND
(CHEMS) (3.46 × 109 RLU/mg protein). A lipid com-
position [DOPE/PA, 7:2 (molar ratio)] with a high
endosome-fusogenic activity for NIH 3T3 cells and
HeLa cells10,26 might not be suitable for use in con-
junction with KSN–LECs. On the basis of the results
of these experiments, we concluded that the composi-
tion of the R8–GALA–MEND (CHEMS) was adequate
for achieving transgene expression in LECs.
We herein discuss the lipid composition needed
for transgene expression in LECs. The R8–MEND
(PA) showed a higher transfection efficiency than
DOI 10.1002/jps
 TRANSGENE EXPRESSION IN LECS USING HA-COATED GENE VECTORS 3125

Figure 4. Transfection activity of a MEND system equipped with R8 and GALA in LECs. We
prepared the MEND including the CHEMS [MEND (CHEMS)] (a) and the PA [MEND (PA)]
(b) with or without R8 and GALA, and evaluated their transfection activities by a luciferase
assay. Data are the means ± SD (n = 3–6). ∗∗ Significant differences between plane MEND and
other MEND (p < 0.01 by one-way ANOVA, followed by Bonferroni correction) was calculated. (c)
The transfection activities of R8–GALA–MEND (CHEMS), Naked pDNA, MEND (CHEMS), and
HA–R8–GALA–MEND (CHEMS) are summarized. Data are means ± SD (n = 3–6). ∗∗ Significant
differences between R8–GALA–MEND (CHEMS) and others (p < 0.01 by one-way ANOVA,
followed by Bonferroni correction) were calculated.

the GALA–MEND (PA), whereas the transfection
efficiency of the R8-MEND (CHEMS) and the GALA–
MEND (CHEMS) was similar. In previous studies,
we determined that the optimal lipid composition
for endosome-fusogenic activity was DOPE/PA/STR-
R8 (7:2:1, molar ratio) and showed that this optimal
lipid composition dramatically enhanced endosomal
escape in living cells.10,26 Thus, we considered that
the R8-MEND (PA) would have a high transfection
efficiency, even when GALA was not equipped with
the carriers.
Coating the MEND with HA Via Electrostatic
Interactions and the Transfection Activity in LECs
We next compared the transfection activity of
the R8–GALA–MEND (CHEMS) before and after

DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
3126 YAMADA ET AL.
application of the HA coating. We attempted to inte-
grate the R8–GALA–MEND (CHEMS) and HA with-
out PEGylation because PEGylation typically results
in a decrease in the transfection activity of the carrier.
We assumed that HA would coat the carrier surface by
simply mixing it with the positively charged
R8–GALA–MEND (CHEMS) via electrostatic in-
teractions between the positively charged MEND
and the negatively charged HA. The ξ potentials of
the R8–GALA–MEND (CHEMS) after mixing with
various ratios of HA are summarized in Figure S5.
The presence of HA resulted in a gradual decrease
in the carrier-surface charge, and it became negative
when the concentration of HA exceeded 0.02 mol%
suggesting that HA coated the surface of
the R8–GALA–MEND (CHEMS) and that a
HA–R8–GALA–MEND (CHEMS) had been formed.
The transfection activities for KSN–LECs using
R8–GALA–MEND (CHEMS), naked pDNA, MEND
(CHEMS), and HA–R8–GALA–MEND (CHEMS)
are summarized in Figure 4c. In this experi-
ment, HA–R8–GALA–MEND (CHEMS) containing
0.1 mol% of HA was used. The ξ potentials of carriers
are summarized in Table S3. The transfection activ-
ity of R8–GALA–MEND (CHEMS) was significantly
higher than that of naked pDNA and the MEND
(CHEMS). On the other hand, the values for the
R8–GALA–MEND (CHEMS) were not significantly
different before and after coating with HA, although
their ξ potentials were considerably different (ap-
proximately +50 mV in the case of R8–GALA–MEND
(CHEMS), approximately −30 mV in the case of
HA–R8–GALA–MEND (CHEMS). These results in-
dicate that the R8–GALA–MEND (CHEMS) had a
high transfection activity in LECs, even when HA
was coated on the carrier surface to form a negatively
charged particle.
CONCLUSIONS
Our results constitute the first report of transgene
expression in LECs using HA-coated liposome-based
gene vectors. In this study, we determined the optimal
composition of a MEND needed for efficient transgene
expression in LECs [R8–GALA–MEND (CHEMS)]
using a simple in vitro assay. We also confirmed
that HA–R8–GALA–MEND showed a high transfec-
tion activity, although, in the case of other nega-
tively charged MENDs, the transfection activity was
decreased substantially. Further studies including a
mechanism-based cellular pathway analysis of the
HA–R8–GALA–MEND and the selective transgene
expression in LECs in vivo experiments are needed.
We previously reported that an intravenously injected
MEND containing CHEMS, R8, and GALA had a high
transgene expression in the liver.27 Thus, we expected
that HA modification would be a key factor in dis-
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013
tinguishing parenchymal cell and endothelial cells in
liver transgene expression. In future studies, we plan
to compare the transfection activity of an HA-coated
MEND between parenchymal and endothelial cells
in vivo. Studies in this area are currently underway.
ACKNOWLEDGMENTS
This study was funded by Special Education and Re-
search Expenses of the Ministry of Education, Cul-
ture, Sports, Science and Technology, the Japanese
Government (MEXT), and in part by the Advanced re-
search for medical products Mining Programme of the
National Institute of Biomedical Innovation (NIBIO),
a grant-in-aid for Scientific Research (S) from MEXT.
We thank Dr. Milton Feather for his helpful advice in
writing the manuscript.
REFERENCES
1. Thabut D, Shah V. 2010. Intrahepatic angiogenesis and sinu-
soidal remodeling in chronic liver disease: New targets for the
treatment of portal hypertension? J Hepatol 53:976–980.
2. Cheluvappa R, Denning GM, Lau GW, Grimm MC, Hilmer SN,
Le Couteur DG. 2010. Pathogenesis of the hyperlipidemia of
Gram-negative bacterial sepsis may involve pathomorpholog-
ical changes in liver sinusoidal endothelial cells. Int J Infect
Dis 14:e857–e867.
3. Fraser JR, Laurent TC, Laurent UB. 1997. Hyaluronan: Its
nature, distribution, functions and turnover. J Intern Med
242:27–33.
4. Smedsrod B, Pertoft H, Eriksson S, Fraser JR, Laurent TC.
1984. Studies in vitro on the uptake and degradation of
sodium hyaluronate in rat liver endothelial cells. Biochem J
223:617–626.
5. Fraser JR, Alcorn D, Laurent TC, Robinson AD, Ryan GB.
1985. Uptake of circulating hyaluronic acid by the rat liver.
Cellular localization in situ. Cell Tissue Res 242:505–510.
6. Takei Y, Maruyama A, Ferdous A, Nishimura Y, Kawano S,
Ikejima K, Okumura S, Asayama S, Nogawa M, Hashimoto M,
Makino Y, Kinoshita M, Watanabe S, Akaike T, Lemasters JJ,
Sato N. 2004. Targeted gene delivery to sinusoidal endothe-
lial cells: DNA nanoassociate bearing hyaluronan-glycocalyx.
FASEB J 18:699–701.
7. Jiang G, Park K, Kim J, Kim KS, Hahn SK. 2009. Target
specific intracellular delivery of siRNA/PEI-HA complex by
receptor mediated endocytosis. Mol Pharm 6:727–737.
8. Kim J, Kim KS, Jiang G, Kang H, Kim S, Kim BS, Park MH,
Hahn SK. 2008. In vivo real-time bioimaging of hyaluronic acid
derivatives using quantum dots. Biopolymers 89:1144–1153.
9. Toriyabe N, Hayashi Y, Hyodo M, Harashima H. 2011. Synthe-
sis and evaluation of stearylated hyaluronic acid for the active
delivery of liposomes to liver endothelial cells. Biol Pharm Bull
34:1084–1089.
10. El-Sayed A, Khalil IA, Kogure K, Futaki S, Harashima
H. 2008. Octaarginine- and octalysine-modified nanoparti-
cles have different modes of endosomal escape. J Biol Chem
283:23450–23461.
11. Khalil IA, Kogure K, Futaki S, Hama S, Akita H, Ueno M,
Kishida H, Kudoh M, Mishina Y, Kataoka K, Yamada M,
Harashima H. 2007. Octaarginine-modified multifunctional
envelope-type nanoparticles for gene delivery. Gene Ther
14:682–689.
12. Khalil IA, Kogure K, Futaki S, Harashima H. 2006. High den-
sity of octaarginine stimulates macropinocytosis leading to
DOI 10.1002/jps
 TRANSGENE EXPRESSION IN LECS USING HA-COATED GENE VECTORS
efficient intracellular trafficking for gene expression. J Biol
Chem 281:3544–3551.
13. Kogure K, Akita H, Yamada Y, Harashima H. 2008. Multifunc-
tional envelope-type nano device (MEND) as a non-viral gene
delivery system. Adv Drug Deliv Rev 60:559–571.
14. Kogure K, Moriguchi R, Sasaki K, Ueno M, Futaki S,
Harashima H. 2004. Development of a non-viral multifunc-
tional envelope-type nano device by a novel lipid film hydra-
tion method. J Control Release 98:317–323.
15. Futaki S, Ohashi W, Suzuki T, Niwa M, Tanaka S, Ueda K,
Harashima H, Sugiura Y. 2001. Stearylated arginine-rich pep-
tides: A new class of transfection systems. Bioconjug Chem
12:1005–1011.
16. Futaki S, Suzuki T, Ohashi W, Yagami T, Tanaka S, Ueda K,
Sugiura Y. 2001. Arginine-rich peptides. An abundant source
of membrane-permeable peptides having potential as carri-
ers for intracellular protein delivery. J Biol Chem 276:5836–
5840.
17. Futaki S, Masui Y, Nakase I, Sugiura Y, Nakamura T, Kogure
K, Harashima H. 2005. Unique features of a pH-sensitive fu-
sogenic peptide that improves the transfection efficiency of
cationic liposomes. J Gene Med 7:1450–1458.
18. Kakudo T, Chaki S, Futaki S, Nakase I, Akaji K, Kawakami
T, Maruyama K, Kamiya H, Harashima H. 2004. Transferrin-
modified liposomes equipped with a pH-sensitive fusogenic
peptide: An artificial viral-like delivery system. Biochemistry
43:5618–5628.
19. Hida K, Hida Y, Amin DN, Flint AF, Panigrahy D, Morton
CC, Klagsbrun M. 2004. Tumor-associated endothelial cells
with cytogenetic abnormalities. Cancer Res 64:8249–8255.
20. Matsuda K, Ohga N, Hida Y, Muraki C, Tsuchiya K, Kurosu
T, Akino T, Shih SC, Totsuka Y, Klagsbrun M, Shindoh M,
Hida K. 2010. Isolated tumor endothelial cells maintain spe-
DOI 10.1002/jps
3127
cific character during long-term culture. Biochem Biophys Res
Commun 394:947–954.
21. Ohga N, Hida K, Hida Y, Muraki C, Tsuchiya K, Matsuda
K, Ohiro Y, Totsuka Y, Shindoh M. 2009. Inhibitory effects of
epigallocatechin-3 gallate, a polyphenol in green tea, on tumor-
associated endothelial cells and endothelial progenitor cells.
Cancer Sci 100:1963–1970.
22. Mochizuki S, Kano A, Shimada N, Maruyama A. 2009. Up-
take of enzymatically-digested hyaluronan by liver endothe-
lial cells in vivo and in vitro. J Biomater Sci Polym Ed 20:83–
97.
23. Lee H, Choi SH, Park TG. 2006. Direct visualization
of hyaluronic acid polymer chain by self-assembled one-
dimensional array of gold nanoparticles. Macromolecules
39:23–25.
24. Lee H, Mok H, Lee S, Oh YK, Park TG. 2007. Target-specific
intracellular delivery of siRNA using degradable hyaluronic
acid nanogels. J Control Release 119:245–252.
25. Nakase I, Niwa M, Takeuchi T, Sonomura K, Kawabata N,
Koike Y, Takehashi M, Tanaka S, Ueda K, Simpson JC, Jones
AT, Sugiura Y, Futaki S. 2004. Cellular uptake of arginine-rich
peptides: Roles for macropinocytosis and actin rearrangement.
Mol Ther 10:1011–1022.
26. Akita H, Kudo A, Minoura A, Yamaguti M, Khalil IA,
Moriguchi R, Masuda T, Danev R, Nagayama K, Kogure
K, Harashima H. 2009. Multi-layered nanoparticles for
penetrating the endosome and nuclear membrane via a
step-wise membrane fusion process. Biomaterials 30:2940–
2949.
27. Khalil IA, Hayashi Y, Mizuno R, Harashima H. 2011.
Octaarginine- and pH sensitive fusogenic peptide-modified
nanoparticles for liver gene delivery. J Control Release
156:374–380.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 102, NO. 9, SEPTEMBER 2013