Rigorous control of vesicle-forming lipid pKa by fluorine-conjugated bioisosteres for gene-silencing with siRNA

Article abstract of DOI:10.1016/j.jconrel.2018.12.044

While the influence of pK_a provided by amine-containing materials in siRNA delivery vectors for use in gene-silencing has been widely studied, there are little reports in which amine pK_a is controlled rigorously by using bioisosteres and its effect on gene-silencing. Here, we report that amine pK_a could be rigorously controlled by replacement of hydrogen atom(s) with fluorine atom(s). A series of mono- and di-amine lipids with a different number of fluorine atoms were synthesized. The pK_a of the polyamine lipids was shifted to a lower value with an increase in the number of fluorine atoms. The optimal pK_a for high gene-silencing efficiency varied according to the number of amine residues in the polyamine lipid. Whereas the endosomal escape ability of mono-amine lipid-containing lipid vesicles (LVs) depended on the pK_a, that of all tested di-amine lipid-containing LVs showed equal membrane-destabilizing activity. LVs showing moderately weak interactions with siRNA facilitated cytoplasmic release of siRNA, resulting in strong gene-silencing. These findings indicate that appropriate amine pK_a engineering depending on the number of amines is important for the induction of effective RNA interference.

Full text of DOI:10.1016/j.jconrel.2018.12.044

Accepted Manuscript
Rigorous control of vesicle-forming lipid pKa by fluorine-
conjugated bioisosteres for gene-silencing with siRNA
Ayaka Okamoto, Hiroyuki Koide, Naoki Morita, Yusuke Hirai,
Yuji Kawato, Hiromichi Egami, Yoshitaka Hamashima, Tomohiro
Asai, Takehisa Dewa, Naoto Oku
PII:
S0168-3659(18)30749-1
DOI:
https://doi.org/10.1016/j.jconrel.2018.12.044
COREL 9594
Reference:
To appear in:
Journal of Controlled Release
Received date:
Revised date:
Accepted date:
12 October 2018
8 December 2018
25 December 2018
Please cite this article as: Ayaka Okamoto, Hiroyuki Koide, Naoki Morita, Yusuke Hirai,
Yuji Kawato, Hiromichi Egami, Yoshitaka Hamashima, Tomohiro Asai, Takehisa Dewa,
Naoto Oku , Rigorous control of vesicle-forming lipid pKa by fluorine-conjugated
bioisosteres for gene-silencing with siRNA. Corel (2018), https://doi.org/10.1016/
j.jconrel.2018.12.044
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
Rigorous control of vesicle-forming lipid pK by
a
fluorine-conjugated bioisosteres for
gene-silencing with siRNA
a
a
b
a
b
Ayaka Okamoto , Hiroyuki Koide , Naoki Morita , Yusuke Hirai , Yuji Kawato ,
b
b
a
c
Hiromichi Egami , Yoshitaka Hamashima , Tomohiro Asai , Takehisa Dewa , and
a,
Naoto Oku *
a
b
Department of Medical Biochemistry and Synthetic Organic Chemistry, School of
Pharmaceutical Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka
4
22-8526, Japan
c
Department of Life Science and Applied Chemistry, Nagoya Institute of Technology,
Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan
*Corresponding author. Tel: +81 54 264 5701, Fax: +81 54 264 5705.
E-mail address: oku@u-shizuoka-ken.ac.jp. (N. Oku)
1

ACCEPTED MANUSCRIPT
Graphical Abstract
2

ACCEPTED MANUSCRIPT
Abstract
While the influence of pK provided by amine-containing materials in siRNA
a
delivery vectors for use in gene-silencing has been widely studied, there are little
reports in which amine pK is controlled rigorously by using bioisosteres and its effect
a
on gene-silencing. Here, we report that amine pK could be rigorously controlled by
a
replacement of hydrogen atom(s) with fluorine atom(s). A series of mono- and di-amine
lipids with a different number of fluorine atoms were synthesized. The pK of the
a
polyamine lipids was shifted to a lower value with an increase in the number of fluorine
atoms. The optimal pK for high gene-silencing efficiency varied according to the
a
number of amine residues in the polyamine lipid. Whereas the endosomal escape ability
of mono-amine lipid-containing lipid vesicles (LVs) depended on the pK , that of all
a
tested di-amine lipid-containing LVs showed equal membrane-destabilizing activity.
LVs showing moderately weak interactions with siRNA facilitated cytoplasmic release
of siRNA, resulting in strong gene-silencing. These findings indicate that appropriate
amine pK engineering depending on the number of amines is important for the
a
induction of effective RNA interference.
KEYWORDS:lipid nanoparticle, siRNA, pKa, fluorine
3

ACCEPTED MANUSCRIPT
1
. Introduction
Pharmaceuticals such as small-molecule and biomedical drugs have made
great contribution to longevity. However, there are many diseases untreatable by these
pharmaceuticals [1,2]. Oligonucleotides such as antisense, siRNA and miRNA have
received much interest as novel therapeutic agents capable of treating such diseases,
because they specifically lead to cleavage or functional inhibition of their target
mRNA[3] and can be designed and synthesized according to the sequence on demand.
As certain types of oligonucleotides act in the cytoplasm, they face several hurdles to
get there: 1) being taken up into cell, 2) escaping from the endosomes, and 3) being
released into the cytoplasm [4]. Unfortunately, the physicochemical properties of
oligonucleotides such as their high molecular weight, hydrophilicity, and high
negative-charge density make it difficult for them to break these barriers by themselves.
In addition, oligonucleotides are decomposed by nucleases and eliminated rapidly after
entrance into the bloodstream [5]. Therefore, various materials such as liposomes [6],
lipid nanoparticles [7], polymer nanoparticles [8], dendrimers [9], and nanogels [10]
have been studied widely as their carriers.
Polyethylenimine (PEI) has been widely used as a polycationic material for
the holding of oligonucleotides such as siRNAs in the carriers. Although transfection
with siRNA in PEI-containing nanocarriers results in a significant gene-silencing effect
by increasing siRNA uptake into the cell and providing endosomal escape ability [11],
its strong cytotoxicity has been a critical problem [12]. Therefore, a number of studies
have been carried out to reduce the toxicity and enhance the knockdown effect by
improving polycation structure. For example, Kataoka’s group prepared a series of the
N-substituted polyaspartamides possessing repeating aminoethylene units [13]; Hope’s
group synthesized 56 kinds of amino lipids that include primary, secondly or tertiary
amine residues [14]; Anderson’s group prepared a large library of lipidoids containing a
secondary or a tertiary amine residue [15]; Harashima’s group synthesized
dimethylamino lipids containing different numbers of methylene units [16]; and
Siegwart’s group synthesized 139 lipocationic polyesters containing dialkyl amine or
heterocyclic amine polymer [17]. These reports indicate that amine pK (pK = 6.2–6.5)
a
a
is important for efficient gene-silencing and reduced cytotoxicity. However, precise
4

ACCEPTED MANUSCRIPT
regulation of the amine pK without a significant change in the chemical structure is
a
considered to be extremely difficult, since adjustment of the amine pK usually requires
a
conjugation with bulky groups such as alkyl chain, benzene, cycloalkane or
bicyclo-compound to the amine group [14-17]. These modification changes the steric
bulk of the molecules along with the pK . It is known that steric bulk has a significant
a
effect on gene-silencing [15]. Therefore, rigorous control of amine pK without drastic
a
structural change, and investigation of the amine pK influence on gene-silencing is still
a
a great challenge and important for the designing of siRNA carriers.
In the present study, we focused on the use of fluorine atoms for rigorous
control of amine pK . As the fluorine atom is a strong electron-withdrawing substituent
a
and has an atomic size similar to that of hydrogen [18], the pK s of conjugate acids
a
derived from neighboring nitrogen atoms decrease in accordance with the number of
fluorine atoms without causing a significant change in chemical structure or steric bulk
[
19]. We used ethylenediamine (EtDA) as a model amine group. In addition,
diethylenetriamine (DiETA) was used to examine the effect of the number of amines on
the gene-silencing.
5

ACCEPTED MANUSCRIPT
2
. Material and methods
2
.1 Titrationof polyamine lipid
Preparation procedures for DCP-polyamine-CH R (Et-CH , Et-CH F,
2
3
2
Et-CHF , Et-CF , Di-CH , Di-CH F, Di-CHF and Di-CF ) were described in
2
3
3
2
2
3
Supporting Information. DCP-polyamine-CH R and DSPE-PEG2000 (1/0.1 as a molar
2
ratio) were dissolved in t-butanol and lyophilized. Liposomes were produced by
hydration of the lipid mixture with RNase-free water. The liposomes were diluted (final
lipid conc.: 2 mM) with 5 mL of ultra pure water containing 150 mM NaCl, and basified
with 0.1 M NaOH. Five µL of 0.1 M HCl was added to the vial by use of micropipette
while stering the solution. pH change was monitored by use of pH meter HM-31P
(
DKK-TOA, Tokyo, Japan). After stabilized the pH, additional HCl was dropped.
2
.2 Preparationof lipid vesicles
DOPE, Cholesterol, DPPC and polyamine lipid (1/1/0.5/0.5 as a molar ratio)
were dissolved in t-butanol and lyophilized. Lipid vesicles (LVs) were produced by
hydrating the lipid mixture with RNase-free water. LVs were freeze-thawed 3-times to
prepare unilamellar vesicles.
2
.3 TNS assay
Ten µL of 1 mM LVs were diluted in 480 µL of assay buffer containing 20
mM sodium phosphate, 25 mM citrate, 20 mM ammonium acetate, 150 mM NaCl (pH
.0-12.0), and incubated for 20 min at room temperature. Then, 10 µL of 0.3 mM TNS
2
in was mixed with the LVs solutions. The fluorescence intensity of TNS was determined
with a Tecan Infinite M200 microplate reader (Salzburg, Austria) operated according to
the manufacturer's instructions (ex. 322 nm, em. 431 nm).
2
.4 Preparationof siRNA-encapsulated LVs
6

ACCEPTED MANUSCRIPT
siRNA-encapsulated lipid vesicles (sLVs) was prepared by freeze-thawing of
siRNA and LVs complex as described previously [21]. LVs and siRNA were mixed
(
protonable amine groups in polyamine lipid / phosphorus atoms in siRNA = 20 as a
molar ratio) and incubated for 20 min at room temperature in 1 mM citric
acid/RNase-free water (pH 3.0) to form siRNA/LVs complexes. Then, the complex was
frozen in liquid nitrogen and thawed in a water bath at 45°C with vortexing to prepare
sLVs. Cell-penetrating peptides (CPP)-DOPE conjugate (6 mol% to total lipids) was
modified to sLVs by incubating them at 50°C for 30 min (sLVs-CPP). The amino acid
sequence of CPP is RRRRRRGGRRRRG.
2
.5 Cell culture
HT1080 human fibroblast cells and HT1080 expressing enhanced green
fluorescent protein (HT1080-EGFP) cells were cultured with D-MEM/Ham’s F-12
containing 10% fetal bovine serum, 100-units/mL penicillin G and 100-μg/mL
streptomycin in a CO incubator.
2
2
.6 siRNA transfection
Cells were seeded onto a culture plate and pre-cultured overnight. The
medium was changed to a fresh one supplemented with FBS and free of antibiotics
adjusted pH 7.4) before transfection. Et-sLVs-CPP or Di-sLVs-CPP was added to the
(
culture medium, and the cells were then incubated for 24 h at 37°C in a 5% CO2
incubator. After a medium change, the cells were incubated for the desired time as
described for each experimental procedure.
2
.7 siRNA uptake test
HT1080 cells were seeded onto 24-well plates (BD Bioscience, San Jose, CA)
7

ACCEPTED MANUSCRIPT
4
at the density of 1.5×10 cells/well. FITC-labeled siRNA formulated in LVs-CPP were
added to the cells (5 pmol/500 µL; 10 nM as siRNA) and incubated for 24 h. After
having been washed with PBS containing 30 units/mL heparin, the cells were lysed with
1
w/v% n-octyl-β-D-glucoside containing the following protease inhibitors: 1 mM
phenylmethylsulfonyl fluoride, 2 μg/mL leupeptin, 2 μg/mL aprotinin, and 2 μg/mL
pepstatin A. The fluorescence intensity of FITC was determined with a Tecan Infinite
M200 microplate reader (ex. 495 nm, em. 535 nm). Total protein contents were
measured by bicinchoninic acid assay using a Pierce™ BCA Protein Assay Reagent Kit
according to the manufacturer's instructions.
2
.8 Gene-silencing effect
4
HT1080-EGFP cells (1.5×10 cells/well) were seeded onto 24-well plates (BD
Bioscience) and transfected with siRNA against GFP (siGFP) formulated in each
LVs-CPP for 24 h at a final siGFP concentration of 10 nM (5 pmol/500 µL). After the
medium had been changed to a fresh one, the cells were cultured for an additional 48 h.
Cell viability was determined by WST-8 assay; the media was changed to WST-8 assay
reagent (Cell Counting Kit-8 : medium = 7.5 µL : 292.5 µL), and then the cells were
incubated for 1 h at 37°C. To determine cell viability, absorbance of supernatant at 450
nm was measured using Tecan Infinite M200 microplate reader. Then, the cells were
washed with PBS and lysed with 1 w/v% n-octyl-β-D-glucoside containing the protease
inhibitors. Gene silencing against EGFP was determined as a reduction rate of the
fluorescence intensity measured with a Tecan Infinite M200 microplate reader (ex. 485
nm, em. 535 nm). Total protein content measured with a Pierce™ BCA Protein Assay
Reagent Kit. RNA interference efficiency was determined as follows: Knockdown (%) /
siRNA uptake (pmol/well).
2
.9 Hemolysis assay
For preparation of erythrocyte, 500 µL of caw blood was washed by gently
vortexing with 1 mL of PBS, and was centrifuged at 10,000 xg for 10 min at 4°C. After
8

ACCEPTED MANUSCRIPT
repeating the wash with PBS five times, the pellet was suspended with 0.3 M sucrose in
ultra-pure water. sLVs-CPP diluted with 10 mM phosphate buffer (pH 7.4 or 5.5)
containing 0.3 M sucrose were mixed with 10 µL of the erythrocyte (100 µM as amine
moiety), and incubated at 37°C for 1 h in a shaking container. After centrifugation
(
10,000 xg for 10 min at room temperature), the liberated hemoglobin was determined
by colorimetric analysis of the supernatant at 405 nm with a Tecan Infinite M200
microplate reader. The value for 100% hemolysis was set from the erythrocytes treated
with 0.1% Triton X-100. The ζ-potential of the sLVs-CPP diluted with 10 mM
phosphate buffer (pH = 5.5 or 7.4) was measured by use of Zetasizer Nano ZS
(
Malvern).
2
.10 Isothermal titrationcalorimetric assay
Isothermal titration calorimetry was performed with MicroCal PEAQ-ITC
(
Malvern) at 25°C. Injection cell was filled with 300 nM siRNA solution in 1 mM
phosphate buffer (pH 7.4 or 5.5). Et-LVs or Di-LVs was titrated into the cell according
to the manufacturer's program. Data was analyzed by use of the MicroCal PEAQ-ITC
Analysis (Malvern).
2
.11 Confocal laser-scanning microscopy
HT1080 cells were seeded onto glass bottom 24-well plate (AGC Techno
4
Glass Co., Ltd., Shizuoka, Japan) at a density of 1.5×10 cells/well. FITC-labeled
siRNA-encapsulating Et-LVs-CPP or Di-LVs-CPP was added to the cells (siRNA
concentration was 10 nM; 5 pmol/500 µL/well). Distribution of FITC-labeled siRNA in
+
the HT1080 cells was observed under an A1R confocal laser-scanning microscope
(
Nikon, Tokyo, Japan). Individual cells within a single field of view were imaged every
min for 24 h. siRNA diffusion time after transfection of LVs-CPP and the proportion
5
of siRNA-diffusing cell at selected time points were determined. The proportion of
siRNA-diffusing cells was calculated as follows: the number of FITC-distributed cells /
the number of all the cells.
9

ACCEPTED MANUSCRIPT
2
.12 Statistical analysis
Differences within a group were evaluated by analysis of variance (ANOVA)
with the Tukey post-hoc test.
1
0

ACCEPTED MANUSCRIPT
3
. Results and discussion
3
.1 Rigorous control of amine pKa by replacement of hydrogen atom(s) with fluorine
atom(s).
Procedures for the synthesis of dicetylphosphate (DCP)-EtDA and -DiETA are
shown in Figure 1. As for (DCP)-EtDA-CH R (R indicates CH ([Et-CH ], CH F
2
3
3
2
(
[Et-CH F]), CHF2 ([Et-CHF ], CF3 [Et-CF ]), reductive amination of the
2 2 3
corresponding fluorine-containing primary amines with N-Boc-2-aminoacetaldehyde (a)
was first conducted. Removal of the Boc protection with 4 M HCl proceeded smoothly
to furnish the diamine backbone. Fortunately, in line with our previous reports [6], the
amidation with dicetyl chlorophosphate (b) occurred chemoselectively at the primary
amine under typical conditions. Though a tedious reaction sequence was required, a
similar method was applicable for the synthesis of (DCP)-DiETA-CH R (Di-CH ,
2
3
Di-CH F, Di-CHF , Di-CF , respectively). Initially, with the same amines as a starting
2
2
3
material, N-Boc-N-fluoroalkyl-2-aminoacetal-dehydes were prepared through a reaction
set consisting of reductive amination with (tert-butyldimethylsiloxy)a-cetaldehyde (c),
Boc protection, and desilylation followed by oxidation of the primary alcohol. Requisite
elongation to
the
triamine backbone was readily performed with
N-Boc-ethylenediamine (d), and the resulting primary amine was conjugated to
chlorophosphate.
To determine the pK of the lipids, acid-base titration was performed with 0.1
a
M HCl. While the titration curves of EtDA-lipids showed a single equivalence point and
DiETA-lipids showed a two-phase equivalence curve (Figure S1), the pK of
a
EtDA-lipids was gently shifted to lower value by increasing the number of fluorine
atoms (Table 1). In addition, both pK and pK of DiETA-lipids were also shifted to
a1
a2
lower values in response to the substitution with fluorine. The pK of Et-CF and pK
a2
a
3
of Di-CF could not be determined due to loss of buffering capacity over the range of
3
pH 2–11. From these results, we succeeded in rigorous control of amine pK by fluorine
a
substitution without causing any significant change in chemical structure or steric bulk.
11

ACCEPTED MANUSCRIPT
Figure 1. Synthesis of fluorine-conjugated lipids.
A) Dicetylphosphate-ethylenediamine-CH R
(
(Et-R)
and
(B)
2
Dicetylphosphate-diethylenetriamine-CH R (Di-R) conjugates were synthesized by
2
reductive amination.
Table 1. pK values of polyamine lipids
a
pKa1
pKa2
pKa1
pKa2
Et-CH3
Et-CH2F
Et-CHF2
Et-CF3
8.2
5.3
-
-
-
-
Di-CH3
Di-CH2F
Di-CHF2
Di-CF3
9.2
8.0
7.5
7.1
5.4
4.9
4.4
3.9
< 3.3
< 3.0
1
2

ACCEPTED MANUSCRIPT
3
.2 Effect of the number of amines and fluorine atoms on gene-silencing
We
next
prepared
lipid
vesicles
(DOPE),
(LVs)
composed
of
dioleoylphosphatidylethanolamine
cholesterol,
dipalmitoylphosphatidylcholine (DPPC), and EtDA-lipid or DiETA-lipid (Et-LVs,
Di-LVs, 1/1/0.5/0.5 as a molar ratio). Charged percentage of the LVs within the pH
2
–12 range was measured by using 6-(p-toluidino)-2-naphthalenesulfonic acid [20].
Although the curve of Et-LVs shifted to low pH side by increasing the number of
fluorine atom, that of Di-LVs did not show any considerable change (Figure 2A). These
results suggest that the surface charge of mono-amine-containing nanoparticles
depended on the pK of the amino group whereas that of di- or more amine-containing
a
nanoparticles did not depend on the amine pKa
To evaluate the gene-silencing activity of siRNA delivered with these LVs, we
prepared siRNA-encapsulated LVs (sLVs) by freeze-thawing siRNA and LV complexes
3
times in 1 mM citric acid (pH 3.0, Figure S2) [21]. In this study, FITC-conjugated
siRNA was used for the uptake test, and siRNA against green fluorescent protein (GFP)
was used for the gene-silencing assay. Importantly, un-encapsulated siRNA was not
detected by performing an electrophoretic assay after the freeze-thawing (Figure S3).
Modification of functional peptides has often been used to improve cellular uptake of
nanocarriers [22]. In the present study, we modified the sLV surface with
cell-penetrating peptide (CPP) [23] (sLV-CPP, Figure S3). Then, the cellular uptake
ratio and gene-silencing effect of each sLV-CPP were examined (Figure 2B, Figure
S4-S6). No cytotoxicity was observed with any sLV-CPP (Figure S6). Surprisingly, the
optimal amine pK for a high knockdown effect depended on the number of amines.
a
Thus, sLV-CPP containing Et-CH (pK = 8.2) showed a high knockdown effect in the
3
a
EtDA (mono-amine) group; and sLV-CPP containing Di-CF , (pK = 7.1, pK < 3.0) in
3
a1
a2
the DiETA (di-amine) group. These results indicate that the balance between the number
of amines and fluorine atoms was crucial to achieve the high knockdown effect.
1
3

ACCEPTED MANUSCRIPT
Figure 2. Effect of the number of amines and fluorine atoms on gene-silencing.
(
A) TNS was mixed with lipid vesicles in a series of buffers with pH ranging between 2
and 12. Fluorescence intensity was normalized by the TNS fluorescence value at pH 2.0.
B) HT1080-EGFP cells were transfected with siGFP-encapsulating LV-CPP. RNA
interference efficiency was determined as follows: Knockdown (%) / siRNA uptake
pmol/well). Significant differences :**P<0.01, ***P<0.001
(
(
1
4

ACCEPTED MANUSCRIPT
3
.3 Endosomal escape ability of sLV-CPP
It is known that the ability to escape endosomes is important for
gene-silencing induction [24]. Hemolysis assay is known as a surrogate assay to test for
endosomal escape ability because of similarities in their lipid bilayer and glycocalyx
compositions [25]. As endosomal pH is known to decrease with endosome maturation
[
26], caw erythrocytes were incubated with each sLV-CPP in 10 mM phosphate buffer
adjusted pH at 5.5 or at 7.4 for 1 h (Figure 3A). The pH-responsive hemolytic activity
was observed with both Et- and Di-sLV-CPP. The hemolytic activity decreased as the
pK in the Et-sLV-CPP decreased. It is known that positively charged nanoparticles tend
a
to induce strong hemolytic activity [27]. Since the surface charge of Et-sLV-CPP
decreased with increasing fluorine number (Figure 3B), the hemolytic activity should
also be decreased accordingly. These results indicate that Et-CHF - and Et-CF -sLVs
2
3
could not break the endosomal membrane even if the endosome was acidified. Then, the
siRNA should remain in the endosome and be enzymatically degraded [28]. Importantly,
pH-dependent change of ζ-potentials of sLVs-CPP was similar to sLVs (without CPP
modification). In contrast, there was no correlation between the number of fluorine
atoms and the hemolytic activity of Di-sLV-CPP. This data is following the result that
Di-sLVs-CPPs have similarly charged percentage and the surface charges. Surface
charges of Di-sLVs-CPP were not significantly changed even if increasing of fluorine
atom number because pKa2 is still above 5.5 (Figure 3B). Therefore, hemolytic
activities of Di-sLVs-CPP were not changed with the fluorine atom number unlike
Et-sLVs. The data suggest that the impact of pK on membrane-destabilizing activity
a
was small in the di- or more-amine-containing LVs.
1
5

ACCEPTED MANUSCRIPT
Figure 3. Membrane-destabilizing activity of sLVs-CPP.
(
A) sLVs-CPP were incubated with RBCs at 37°C for 1 h in 10 mM phosphate buffer
containing 0.3 M sucrose at pH 7.4 or 5.5. After centrifugation, absorbance of
leaked-hemoglobin was determined at 405 nm. (B) The ζ-potential of each sLVs-CPP in
1
0 mM phosphate buffer was measured. Bars with oblique line indicate the ζ-potentials
of sLVs-CPP at pH 7.4. Filled bars indicate those at pH 5.5.
1
6

ACCEPTED MANUSCRIPT
3
.4 Effect of binding affinityof Di-LVs for siRNA on gene-silencing
We next measured the binding affinity of siRNA for LVs. siRNA needs to
form RISC in the cytoplasm after the endosomal escape for gene-silencing. Therefore,
effective hand-off of siRNA from LVs to the RISC should be important for a strong
gene-silencing effect. We hypothesized that binding affinity of siRNA for Di-CF -LVs,
3
which showed a strong gene-silencing effect, would be weak in the cytoplasm compared
with that of other Di-LVs. Apparent dissociation equilibrium constants (K ) of siRNA
d
for LVs were measured by isothermal titration calorimetry (Figure S7, Table 2). The Kd
of Et-CH - or Et-CH F-LVs at pH 7.4 was 1.27 μM or 2.43 μM, respectivery. The K
d
3
2
of Et-CHF - and Et-CF -LVs were unable to determine. siRNA showed stronger
2
3
binding affinity for all Di-LVs at pH 5.5 than at pH 7.4, due to their positive charge at
pH 5.5 (Figure 2A, 3B). The K values of Di-CH -, Di-CH F-, or Di-CHF -LVs were
d
3
2
2
in the range of 100–500 nM at pH 7.4. In contrast, the K of Di-CF -LVs was 3–10
d
3
times larger (1.57 μM) than that of other Di-LVs. Since the pK of Di-CF is lower than
a
3
that of other Di-lipids, the amine of the Di-CF -LVs should be less protonated than that
3
of the other Di-LVs at pH 7.4. As described above, each Di-LVs had equal hemolytic
activity (Figure 3A). Therefore, the weak interaction of siRNA with LVs in the
cytoplasm should be an important for the release of siRNA from LVs, resulting in strong
gene-silencing. In fact, the K of Et-CH - and Et-CH F-LVs, which showed strong
d
3
2
gene-silencing effect, was similar to that of Di-CF -LV at pH 7.4.
3
Table 2. K values of polyamine liposomes
d
pH 7.4
pH 5.5
pH 7.4
pH 5.5
Et-CH3
Et-CH2F
Et-CHF2
Et-CF3
1.27 µM
2.43 µM
N.D.
1.71 µM
273 nM
1.58 µM
N.D.
Di-CH3
Di-CH2F
Di-CHF2
Di-CF3
327 nM
581 nM
111 nM
1.57 µM
24.3 nM
75.0 nM
41.6 nM
320 nM
N.D.
1
7

ACCEPTED MANUSCRIPT
3
.5 Effect of siRNA release time from LVs in the cytoplasm on gene-silencing
To demonstrate the effect of siRNA release time from LVs into the cytoplasm
on gene-silencing, we measured siRNA release time into the cytoplasm from LVs by
time-lapse imaging using a confocal laser-scanning microscope (Figure 4, Figure S8,
Movie S1-S8.). FITC-conjugated siRNA was used for the experiment, and each
sLV-CPP was used to transfect HT1080 cells. Then, siRNA-releasing time into the
cytoplasm (siRNA-spreading throughout the cytoplasm) was recorded for the first 10
cells up to 24 h after the transfection (Figure 4A). Et-CH -LVs-CPP showed the fastest
3
siRNA release (2.94 h) among the Et-LVs. Although Di-LVs showed relatively fast
siRNA release compared with the Et-LVs, Di-CF -LVs-CPP had the fastest (2.05 h)
3
among the Di-LVs. These results indicate that fast siRNA release from the nanocarriers
is important for the effective gene-silencing. To elucidate the relationship between
proportion of siRNA-diffusing cells and gene-silencing effect, the proportion of cells
with siRNA that had diffused throughout the cytoplasm at a selected time point was
measured (Figure 4B). The diffusion of siRNA was observed in about 70% of cells 24 h
after the transfection with Et-CH - or Di-CF -LV-CPPs. However, proportion of
3
3
siRNA-diffused cells was less than 10% for the other LV-CPPs. siRNA delivered by
LVs containing Et-CH F, Et-CHF or Et-CF did not spread out in the cytoplasm, but
2
2
3
remained in a compact form and appeared to have difficulty escaping from the
endosomes because of low pK . Therefore, the diffusion of siRNA might not be
a
observed. Since the interaction of siRNA to LVs containing Et-CH -LVs was weak,
3
siRNA might be immediately released from the LVs after the endosomal escape and
spread into the cytoplasm. In terms of the DiETA-series, it seems that every
Di-sLV-CPP could escape from the endosome because hemolytic activities of the
Di-sLV-CPP were same. However, affinity of siRNA for LVs containing Di-CF was
3
relatively weak at pH 7.4 compared with that for other Di-LVs (Table 2). Therefore, the
Di-CF -LVs could release siRNA all over the cytoplasm effectively after the endosomal
3
escape compared with other Di-sLV-CPPs. These results indicate that endosomal escape
ability and binding affinity of siRNA for the nanocarrier are important for the siRNA
spreading in the cytoplasm. High proportion of siRNA-diffusing cells should show high
gene-silencing effect.
1
8

ACCEPTED MANUSCRIPT
Figure 4. Intracellular behavior of siRNA delivered with LVs-CPP. HT1080 cells
were transfected with FITC-siRNA-encapsulating LVs-CPP and observed by confocal
laser-scanning microscopy. (A) siRNA diffusion time after transfection with LV-CPPs.
Significant differences: *P<0.05, *** P<0.001. (B) Proportion of siRNA-diffusing cells
at selected time points. The ratio was calculated as follows: the number of
FITC-distributed cells / the number of all the cells.
1
9

ACCEPTED MANUSCRIPT
4
. Conclusion
We succeeded in rigorous control of amine pK by fluorine substitution
a
without causing a significant change in chemical structure or steric bulk. The pK of
a
EtDA (mono-amine)-lipids and both pK and pK of DiETA (di-amine)-lipids were
a1
a2
gently shifted to lower value by increasing the number of fluorine atoms. The optimal
amine pK for the efficient gene-silencing depended on the number of amines. sLV-CPP
a
containing Et-CH (pK = 8.2) showed a high knockdown effect in the EtDA group; and
3
a
sLV-CPP containing Di-CF , (pK = 7.1, pKa2 < 3.0) in the DiETA group. The
3
a1
hemolytic activity decreased as the pK in the Et-sLV-CPP decreased. However, the
a
hemolytic activities of Di-sLVs-CPP were not changed with the fluorine atom number.
The binding affinity of siRNA for the LVs changed even if the LVs showed similar
surface charge and hemolytic activity. LVs having a high RNAi efficiency might be
superior in the cytoplasmic release of siRNA. Moderate interaction between siRNA and
nanocarrier was important for effective gene-silencing. Because sLV-CPP containing
Et-CH or Di-CF showed fast siRNA release time from LVs in the cytoplasm and high
3
3
proportion of siRNA-diffusing cells compared with other sLVs. Since cellular uptake
amount and intracellular behavior of liposome and siRNA may vary depending on the
cancer cell lines, knockdown effect may also vary according to cell lines. Therefore,
adequate pK for strong gene-silencing may differ depending on the cell lines. Although
a
further examination will be needed, these results indicate that it is important to control
the pK of the carrier accurately to achieve the ideal multifunctionalitiy of the vector.
a
We believe that these data provide general information for the design of oligonucleotide
delivery vectors and will prompt the development of novel delivery vector for practical
application of oligonucleotides.
2
0

ACCEPTED MANUSCRIPT
Supporting Information. Information on pK of amine lipids, cellular uptake,
a
knockdown effect, cell viability, interaction of siRNA for LVs, intracellular behavior of
siRNA and NMR data is available in the Supporting Information
Conflict of Interest: The authors declare no competing financial interest
Acknowledgements
This research was supported by the program Grants-in-Aid for Scientific
Research from the Japan Society for the Promotion of Science. The authors thank
Nippon Fine Chemical Co. for the gifts of DPPC, cholesterol, DOPE, and DSPE-PEG.
REFERENCES
[
1] M. Toyoshima, H.L. Howie, M. Imakura, R.M. Walsh, J.E. Annis, A.N. Chang, J.
Frazier, B.N. Chau, A. Loboda, P.S. Linsley, M.A. Cleary, J.R. Park, C. Grandori,
Functional genomics identifies therapeutic targets for MYC-driven cancer, Proc
Natl Acad Sci U S A, 109 (2012) 9545-9550.
[
2] C.V. Pecot, S.Y. Wu, S. Bellister, J. Filant, R. Rupaimoole, T. Hisamatsu, R.
Bhattacharya, A. Maharaj, S. Azam, C. Rodriguez-Aguayo, A.S. Nagaraja, M.P.
Morelli, K.M. Gharpure, T.A. Waugh, V. Gonzalez-Villasana, B. Zand, H.J. Dalton,
S. Kopetz, G. Lopez-Berestein, L.M. Ellis, A.K. Sood, Therapeutic silencing of
KRAS using systemically delivered siRNAs, Mol Cancer Ther, 13 (2014)
2
876-2885.
[
3] E. Bernstein, A.A. Caudy, S.M. Hammond, G.J. Hannon, Role for a bidentate
ribonuclease in the initiation step of RNA interference, Nature, 409 (2001)
3
63-366.
[
[
[
4] S.F. Dowdy, Overcoming cellular barriers for RNA therapeutics, Nature
biotechnology, 35 (2017) 222-229.
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] a) T. Dewa, T. Asai, Y. Tsunoda, K. Kato, D. Baba, M. Uchida, A. Sumino, K.
Niwata, T. Umemoto, K. Iida, N. Oku, M. Nango, Liposomal polyamine-dialkyl
phosphate conjugates as effective gene carriers: chemical structure, morphology,
and gene transfer activity, Bioconjug Chem, 21 (2010) 844-852. b) T. Asai, S.
2
1

ACCEPTED MANUSCRIPT
Matsushita, E. Kenjo, T. Tsuzuku, N. Yonenaga, H. Koide, K. Hatanaka, T. Dewa,
M.
Nango,
N.
Maeda,
H.
Kikuchi,
N.
Oku,
Dicetyl
phosphate-tetraethylenepentamine-based liposomes for systemic siRNA delivery,
Bioconjug Chem, 22 (2011) 429-435.
[
7] S.C. Semple, S.K. Klimuk, T.O. Harasym, N. Dos Santos, S.M. Ansell, K.F. Wong,
N. Maurer, H. Stark, P.R. Cullis, M.J. Hope, P. Scherrer, Efficient encapsulation of
antisense oligonucleotides in lipid vesicles using ionizable aminolipids: formation
of novel small multilamellar vesicle structures, Biochim Biophys Acta, 1510
(
2001) 152-166.
[
[
[
[
8] D.W. Bartlett, M.E. Davis, Physicochemical and biological characterization of
targeted, nucleic acid-containing nanoparticles, Bioconjug Chem, 18 (2007)
4
56-468.
9] B. Ghosn, S.P. Kasturi, K. Roy, Enhancing polysaccharide-mediated delivery of
nucleic acids through functionalization with secondary and tertiary amines, Curr
Top Med Chem, 8 (2008) 331-340.
10] S. Toita, N. Morimoto, K. Akiyoshi, Functional cycloamylose as a
polysaccharide-based biomaterial: application in a gene delivery system,
Biomacromolecules, 11 (2010) 397-401.
11] O. Boussif, F. Lezoualc'h, M.A. Zanta, M.D. Mergny, D. Scherman, B. Demeneix,
J.P. Behr, A versatile vector for gene and oligonucleotide transfer into cells in
culture and in vivo: polyethylenimine, Proc Natl Acad Sci U S A, 92 (1995)
7
297-7301.
[
[
12] S.M. Moghimi, P. Symonds, J.C. Murray, A.C. Hunter, G. Debska, A. Szewczyk, A
two-stage poly(ethylenimine)-mediated cytotoxicity: implications for gene
transfer/therapy, Mol Ther, 11 (2005) 990-995.
13] H. Uchida, K. Miyata, M. Oba, T. Ishii, T. Suma, K. Itaka, N. Nishiyama, K.
Kataoka, Odd-even effect of repeating aminoethylene units in the side chain of
N-substituted polyaspartamides on gene transfection profiles, J Am Chem Soc, 133
(
2011) 15524-15532.
[
14] M. Jayaraman, S.M. Ansell, B.L. Mui, Y.K. Tam, J. Chen, X. Du, D. Butler, L.
Eltepu, S. Matsuda, J.K. Narayanannair, K.G. Rajeev, I.M. Hafez, A. Akinc, M.A.
Maier, M.A. Tracy, P.R. Cullis, T.D. Madden, M. Manoharan, M.J. Hope,
Maximizing the potency of siRNA lipid nanoparticles for hepatic gene silencing in
2
2

ACCEPTED MANUSCRIPT
vivo, Angew Chem Int Ed Engl, 51 (2012) 8529-8533.
[
[
[
15] C.A. Alabi, K.T. Love, G. Sahay, H. Yin, K.M. Luly, R. Langer, D.G. Anderson,
Multiparametric approach for the evaluation of lipid nanoparticles for siRNA
delivery, Proc Natl Acad Sci U S A, 110 (2013) 12881-12886.
16] Y. Sato, H. Hatakeyama, M. Hyodo, H. Harashima, Relationship Between the
Physicochemical Properties of Lipid Nanoparticles and the Quality of siRNA
Delivery to Liver Cells, Mol Ther, 24 (2016) 788-795.
17] J. Hao, P. Kos, K. Zhou, J.B. Miller, L. Xue, Y. Yan, H. Xiong, S. Elkassih, D.J.
Siegwart, Rapid Synthesis of a Lipocationic Polyester Library via Ring-Opening
Polymerization of Functional Valerolactones for Efficacious siRNA Delivery, J Am
Chem Soc, 137 (2015) 9206-9209.
[
[
18] J.D. Dunitz, A. Gavezzotti, Attractions and repulsions in molecular crystals: What
can be learned from the crystal structures of condensed ring aromatic
hydrocarbons?, Accounts Chem Res, 32 (1999) 677-684.
19] J. Olsen, P. Seiler, B. Wagner, H. Fischer, T. Tschopp, U. Obst-Sander, D.W.
Banner, M. Kansy, K. Muller, F. Diederich, A fluorine scan of the
phenylamidinium needle of tricyclic thrombin inhibitors: effects of fluorine
substitution on pKa and binding affinity and evidence for intermolecular C-F...CN
interactions, Org Biomol Chem, 2 (2004) 1339-1352.
[
[
20] A.L. Bailey, P.R. Cullis, Modulation of membrane fusion by asymmetric
transbilayer distributions of amino lipids, Biochemistry, 33 (1994) 12573-12580.
21] H. Koide, A. Okamoto, H. Tsuchida, H. Ando, S. Ariizumi, C. Kiyokawa, M.
Hashimoto, T. Asai, T. Dewa, N. Oku, One-step encapsulation of siRNA between
lipid-layers of multi-layer polycation liposomes by lipoplex freeze-thawing,
Journal of controlled release : official journal of the Controlled Release Society,
2
28 (2016) 1-8.
22] W. Tai, X. Gao, Functional peptides for siRNA delivery, Adv Drug Deliv Rev,
10-111 (2017) 157-168.
[
[
1
23] T. Asai, T. Tsuzuku, S. Takahashi, A. Okamoto, T. Dewa, M. Nango, K. Hyodo, H.
Ishihara, H. Kikuchi, N. Oku, Cell-penetrating peptide-conjugated lipid
nanoparticles for siRNA delivery, Biochemical and biophysical research
communications, 444 (2014) 599-604.
[
24] G. Sahay, W. Querbes, C. Alabi, A. Eltoukhy, S. Sarkar, C. Zurenko, E. Karagiannis,
2
3

ACCEPTED MANUSCRIPT
K. Love, D. Chen, R. Zoncu, Y. Buganim, A. Schroeder, R. Langer, D.G. Anderson,
Efficiency of siRNA delivery by lipid nanoparticles is limited by endocytic
recycling, Nature biotechnology, 31 (2013) 653-658.
[
25] W.H. Evans, W.G. Hardison, Phospholipid, cholesterol, polypeptide and
glycoprotein composition of hepatic endosome subfractions, Biochem J, 232
(
1985) 33-36.
[
[
26] J. Huotari, A. Helenius, Endosome maturation, EMBO J, 30 (2011) 3481-3500.
27] W.G. Light, E.T. Wei, Surface charge and hemolytic activity of asbestos, Environ
Res, 13 (1977) 135-145.
[
28] L. Stalder, W. Heusermann, L. Sokol, D. Trojer, J. Wirz, J. Hean, A. Fritzsche, F.
Aeschimann, V. Pfanzagl, P. Basselet, J. Weiler, M. Hintersteiner, D.V. Morrissey,
N.C. Meisner-Kober, The rough endoplasmatic reticulum is a central nucleation
site of siRNA-mediated RNA silencing, EMBO J, 32 (2013) 1115-1127.
2
4

ACCEPTED MANUSCRIPT
Figure legends
Figure 1. Synthesis of fluorine-conjugated lipids.
A) Dicetylphosphate-ethylenediamine-CH R
(
(Et-R)
and
(B)
2
Dicetylphosphate-diethylenetriamine-CH R (Di-R) conjugates were synthesized by
2
reductive amination.
Figure 2. Effect of the number of amines and fluorine atoms on gene-silencing.
(
A) TNS was mixed with lipid vesicles in a series of buffers with pH ranging between 2
and 12. Fluorescence intensity was normalized by the TNS fluorescence value at pH 2.0.
B) HT1080-EGFP cells were transfected with siGFP-encapsulating LV-CPP. RNA
interference efficiency was determined as follows: Knockdown (%) / siRNA uptake
pmol/well). Significant differences :**P<0.01, ***P<0.001
(
(
Figure 3. Membrane-destabilizing activity of sLVs-CPP.
A) sLVs-CPP were incubated with RBCs at 37°C for 1 h in 10 mM phosphate buffer
(
containing 0.3 M sucrose at pH 7.4 or 5.5. After centrifugation, absorbance of
leaked-hemoglobin was determined at 405 nm. (B) The ζ-potential of each sLVs-CPP in
1
0 mM phosphate buffer was measured. Bars with oblique line indicate the ζ-potentials
of sLVs-CPP at pH 7.4. Filled bars indicate those at pH 5.5.
Figure 4. Intracellular behavior of siRNA delivered with LVs-CPP.
HT1080 cells were transfected with FITC-siRNA-encapsulating LVs-CPP and observed
by confocal laser-scanning microscopy. (A) siRNA diffusion time after transfection with
LV-CPPs. Significant differences: *P<0.05, *** P<0.001. (B) Proportion of
siRNA-diffusing cells at selected time points. The ratio was calculated as follows: the
number of FITC-distributed cells / the number of all the cells.
2
5

Figure 1

Figure 2

Figure 3

Figure 4