Synthesis and evaluation of cationic lipids bearing cholesteryl groups for gene delivery in vitro

Article abstract of DOI:10.1246/bcsj.75.2207

A series of cationic lipids were designed and synthesized as vectors for gene delivery. The lipids contain a ketal moiety as a linker and a cholesteryl group as a hydrophobic tail. The framework of cholesteryl derivatives could increase the stability of liposomes by stabilizing the bilayers and their complexes with DNA to improve the transfection efficiency. The ketal bonds in lipids should easily degrade in an acidic environment in a cell (pH = 2-5) after transfection, resulting in little toxicity; in the neutral environment outside of cells (pH = 7), they should be stable as gene carriers. Gene transfer experiments in vitro with BL-6, 3LL, 293, 3T3, and Hela cells were performed. The results show that the gene transfection activity of three lipids is quite high, with least toxicity to cells under the experimental conditions.

Full text of DOI:10.1246/bcsj.75.2207

c. Jpn.
e Chemical © 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 2207–2213 (2002) 2207
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Synthesis and Evaluation of Cationic Lipids Bearing Cholesteryl Groups
for Gene Delivery In Vitro
Cationic Lipids for Gene Delivery
M. Z. Zhu et al.
†
†
,†
Man-Zhou Zhu, Qi-Hua Wu, Guisheng Zhang, Tan Ren, Dexi Liu,* and Qing-Xiang Guo*
0
2
0
1
7
3
Department of Chemistry, University of Science and Technology of China, Hefei 230026, China
School of Pharmacy, University of Pittsburgh, Pittsburgh, Pennsylvania 15261, USA
†
SJA8
0
9-2673
44
(
Received December 17, 2001)
4
0
0
A series of cationic lipids were designed and synthesized as vectors for gene delivery. The lipids contain a ketal
1
2
moiety as a linker and a cholesteryl group as a hydrophobic tail. The framework of cholesteryl derivatives could increase
the stability of liposomes by stabilizing the bilayers and their complexes with DNA to improve the transfection efficien-
cy. The ketal bonds in lipids should easily degrade in an acidic environment in a cell (pH = 2–5) after transfection, re-
sulting in little toxicity; in the neutral environment outside of cells (pH = 7), they should be stable as gene carriers.
Gene transfer experiments in vitro with BL-6, 3LL, 293, 3T3, and Hela cells were performed. The results show that the
gene transfection activity of three lipids is quite high, with least toxicity to cells under the experimental conditions.
4
With the rapid developments in the human genome sequenc-
ing project and the discovery of disease genes, gene therapy
will be one of the major methods to cure diseases in the future.
However, a great deal of progress still needs to be achieved for
this to be possible. One of the key points is to prepare appro-
the acidic environment in a cell (pH = 2–5), the ketals should
be easy to break up, resulting in little toxicity, while in the neu-
tral environment outside of cells (pH = 7), they should be sta-
ble as gene carriers. We have designed and synthesized a se-
ries of novel cationic lipids 1a–h with cholesteryl groups as a
hydrophobic tail and ketal as a linker bond. The preparations
of 1a–h are outlined as follows (Scheme 1).
Ketalization of 3-chloro-1,2-propanediol with cholest-4-en-
3-one 2 gave cholest-5-en-3-one 3- chloropropylene acetal 3.
The reaction of 3 with piperidine, pyrrolidine, dimethylamine,
(asym)N,N-dimethylethylenediamine, ethylenediamine, re-
spectively, gave 1a–e. The reaction of 3 with trimethylamine
at 120 °C for 24 h in an autoclave, or 80 °C for 7 d gave 1c.
1,2
priate vectors for gene delivery.
Ideally, a gene vector
should have the following properties: 1) it should protect the
gene drug from inactivation by body fluids during travel from
the site of administration to the site of action; 2) it should de-
liver the gene drug into the target cells with high efficiency; 3)
it should be safe and easy to produce on a large scale; and 4) it
3
should deliver the gene to the site where treatment is needed.
One of the vector systems developed during the past few
years having great potential to satisfy these requirements is the
3
The reaction of 1a–c with iodomethane in CHCl /DMSO, re-
4,5
cationic lipid-based delivery system. Since the initial use of
spectively, gave 1f–h.
6
a such vector system in 1987 by Felgner and his colleagues,
Results
many new cationic lipids have been synthesized, such as
DOTMA[2,3-bis(oleoyloxy)propyl]trimethylammonium chlo-
ride analogues, complex alkylamine/alkanecarboxamides, cho-
Transfection Activity. Four of the synthesized com-
pounds (1a, 1b, 1c, 1f,) were used to evaluate the activity in
gene delivery. Standard transfection procedures were followed
with BL-6 cells (murine melanoma). The data in Fig. 1A show
that the transfection activity of these new cationic lipids is de-
pendent on the cationic lipid to the DNA ratio. Using 1 µg of
plasmid DNA per well, the transfection activities of 1b, 1c,
2,7
lesterol derivatives, and aromatic ring-based derivatives.
Several groups have been working on the establishment of a
structure-function relationship for a cationic-based gene deliv-
8
–15
ery system.
Some cationic lipids have been used in clinical
16–19
20,21
trials for treating cancer
and cystic fibrosis.
2
5
Cholesterol derivatives can increase the stability of lipo-
somes by stabilizing the bilayers and their complexes with
and 1f are comparable to that of DC-Chol.
The optimal
transfection activity of these lipids follows the trend tertiary
amine 1c > quaternary amine 1f. Meanwhile, tertiary amines,
1b and 1c, showed a constant activity in a broader range of cat-
ionic lipid to the DNA ratio, while DC-Chol, as well as our
cationic lipid 1f exhibited roughly bell-shaped lipid dose-re-
sponse curves. The peak activity for 1b and 1c was seen at a
ratio of 5:1 (nmol:µg), compared to that of 1f, which exhibit-
ed the peak level at a ratio of 7.5:1 (nmol:µg). Compound 1a
did not exhibit any transfection activity under the experimental
conditions. The data in Fig. 1B show the total amount of pro-
2
2
DNA to improve the transfection efficiency. To date, two of
the most efficacious cytofectins that have been reported are
9
23
cholesterol derivatives: polyamine lipid 67 and CTAP. From
the standpoint of gene delivery, a good vector should be stable
outside of cells and degradable after entering them. Most of
the previously reported lipids use an ether or ester bond as a
linker bond. However, the ether bond as a linkage is too stable
in vivo, and may cause cytotoxicity. On the other hand, the es-
2
4
ter bond is unstable under biological conditions. Considering

2
208 Bull. Chem. Soc. Jpn., 75, No. 10 (2002)
Cationic Lipids for Gene Delivery
Scheme 1.
tein recovered from the cells. Protein recovery is commonly
(Fig. 4A). The total proteins recovered from each well showed
that these lipids appear to be the least toxic to cells; they did
not decrease along with an increase in the amount of cationic
lipids (Fig. 4B). In Hela cells, the optimal transfection activity
of these lipids showed 1c > 1f > 1b. The peak activity for 1c
and 1b was seen at a ratio of 5:1 (nmol:µg), while 1f at 7.5:1
(nmol:µg). All three lipids showed a constant activity over a
broader range for the cationic lipid to DNA ratio (Fig. 5A).
The total protein recovery from each well decreased along
with an increase in the amount of cationic lipids. 1c appears to
be the most toxic to cells (Fig. 5B).
2
6
used as an indicator for toxicity. It is evidenced from Fig. 1B
that cationic lipids 1b, 1c, and 1f appear to be the least toxic to
cells under the experimental conditions.
Transfection Activity in Different Cell Lines. To deter-
mine whether these new cationic lipids are capable of trans-
fecting other types of cells, four additional cell lines were se-
lected to test the transfection activity of the cationic lipids. Be-
cause 1a did not exhibit any transfection activity, we excluded
its test in further experimentation. 3LL (lewis lung carcino-
ma), NIH3T3 (murine embryonic fibroblast), Hela (human cer-
vical adenocarcinoma), and 293 (human emvryonic kidney)
cells were transfected using the same lipid formations as those
used in the experiments summarized in Fig. 1. In 3LL cells,
the optimal transfection activity of these lipids showed 1c >
Discussion
In the synthesis of the cationic lipids, demethylation is
worth mentioning. The demethylation of N-methyl quaternary
−
−
1
b > 1f. The peak activity for 1b, 1c, and 1f was seen at a ra-
ammonium compounds via S
an alcohol solvent has been reported previously.
N
2 displacement by Br or I in
2
7–30
tio of 5:1 (nmol:µg). Like in BL-6 cells, 1c and 1b showed a
constant activity over a broader range for the cationic lipid to
DNA ratio, while 1f exhibited a roughly bell-shaped lipid dose
response curve (Fig. 2A). The total proteins recovered from
each well decreased with an increase in the amount of cationic
lipid. Lipid 1f appears to be the most toxic to cells, followed
by that 1c and 1b appear to be the least toxic to cells (Fig. 2B).
In 293 cells, the optimal transfection activity of these lipids
showed 1c > 1f > 1b. The peak activity for 1c was seen at a
ratio of 2.5:1 (nmol:µg), while 1b and 1f at 5:1 (nmol:µg).
All three lipids showed a constant activity over a broader range
for the cationic lipid to DNA ratio (Fig. 3A). The total proteins
recovered from each well were very low (Fig. 3B). In 3T3
cells, the optimal transfecton activity of these lipids showed 1c
In this
work, compound 3 and trimethylamine in methanol were heat-
ed at 120 °C for 24 h; alternatively, at 80 °C for 7 d, product 1c
was obtained. It was rationalized that the quaternary N,N,N-
trimethyl ammonium salt was formed first. The ammonium
salt is a typical surfactant which aggregates to form a micelle
in methanol. One of the methyls in the ammonium salt was at-
tacked by halide ions in a methanol solution, and the amine
N
moiety was eliminated as a leaving group by a S 2 substitution
mechanism.
Since the first cationic lipid DOTMA was employed as a
gene vector in 1987, many cationic lipids have been synthe-
sized and shown to be active in vitro and some in vivo. Most
of these reported lipids had an ether or ester bond as a linker
bond. Compared to the structure of these preexisting cationic
lipids, cationic lipids designed and synthesized in present work
have a unique structure of ketal as a linker bond, which is bio-
>
1f > 1b. The peak activity for 1c and 1f was seen at a ratio
of 7.5:1 (nmol:µg). 1c and 1b showed a constant activity with
an increase in the amount of cationic lipid up to 7.5 nmol

M. Z. Zhu et al.
Bull. Chem. Soc. Jpn., 75, No. 10 (2002) 2209
Fig. 2. Effect of cationic Lipid to DNA ratio on the transfec-
tion activity. Cells (3LL) were transfected with 1 µg of
PCMV Luc plamid DNA various amounts of cationic lip-
ids. The level of luciferase gene expression (A) and total
protein recovery (B) in cells transfected with cationic lip-
ids 1b, 1c, and 1f at 2.5, 5.0, 7.5, and 10.0 nmol/well, re-
spectively.
Fig. 1. Effect of cationic Lipid to DNA ratio on the transfec-
tion activity. Cells (BL-6) were transfected with 1 µg of
PCMV Luc plamid DNA various amounts of cationic lip-
ids. The level of luciferase gene expression (A) and total
protein recovery (B) in cells transfected with cationic lip-
ids 1b, 1c, and 1f at 2.5, 5.0, 7.5, and 10.0 nmol/well, re-
spectively. Data represent mean ± SD (n = 3).
activity over a broader range of the cationic lipid to DNA ratio.
In conclusion, a series of novel cationic lipids bearing a ket-
al as a linker bond non-viral vector were designed and synthe-
sized for gene delivery. Transfection results suggest that they
are useful transfection reagents for in vitro gene transfer.
3
1
degradable. As a gene vector they are stable outside of cells
and are degradable after entering cells. An in vitro experiment
verified that they had good transfection activity and low toxici-
ty. The head group of a cationic lipid is also important for a
gene vector. In vitro transfection showed that three (1b, 1c, 1f)
of the tested four lipids bearing a cholesteryl group as a hydro-
phobic tail and a ketal as a linker bond exhibit good transfec-
tion activity, while 1a does not exhibit any transfection activi-
ty. This may be due to a steric effect. Piperidine does not ef-
fectively bind with DNA, but pyrrolidine does well. Tertiary
amine 1c gave a higher transfection activity than that of quater-
nary amines 1f. This observation is consistent with results ob-
Experimental
Cholesterol was purified by recrystallization in binary solvents
3
of CHCl /MeOH. All other reagents were purified by distillation.
3
-Chloro-1,2-propanediol was used a non-chiral reagent. Each of
these compounds, 1a–h and 3, is a mixture of diastereoisomers.
The melting points were determined on a Yanaco melting appara-
1
tus and not corrected. H NMR was recorded with a Bruker
3
DMX-500 spectrometer using CDCl as a solvent; the chemical
3
2
served by other groups. Although the luciferase activity ob-
tained in five cells tested implies that different cell lines have
different sensitivity in responding to a cationic lipid-based
transfection reagent, our cationic lipids, especially 1c, have
good transfection activity in any cell, and showed a constant
shifts were given in ppm downfield from TMS. IR spectra were
recorded on a Bruker Vector 220 Infrared Spectrometer using a
KBr pellet. An elemental analysis was performed on a Perkin-
Elmer 240C analytical instrument. MS were recorded with a
Hewlett Packard HP 5988A instrument (70 eV).

2
210 Bull. Chem. Soc. Jpn., 75, No. 10 (2002)
Cationic Lipids for Gene Delivery
Fig. 3. Effect of cationic Lipid to DNA ratio on the transfec-
tion activity. Cells (293) were transfected with 1 µg of
PCMV Luc plamid DNA various amounts of cationic lip-
ids. The level of luciferase gene expression (A) and total
protein recovery (B) in cells transfected with cationic lip-
ids 1b, 1c and 1f, at 2.5, 5.0, 7.5, and 10.0 nmol/well, re-
spectively.
Fig. 4. Effect of cationic Lipid to DNA ratio on the transfec-
tion activity. Cells (3T3) were transfected with 1 µg of
PCMV Luc plamid DNA various amounts of cationic lip-
ids. The level of luciferase gene expression (A) and total
protein recovery (B) in cells transfected with cationic lip-
ids 1b, 1c, and 1f at 2.5, 5.0, 7.5, and 10.0 nmol/well, re-
spectively.
Cholest-5-en-3-one 3-Chloropropylene Acetal (3).
A so-
lution containing cholest-4-en-3-one 2 (10 g, 26 mmol), was pre-
pared with high purity (>99%) via the oxidation of cholesterol by
cyclohexone and recrystallization from methanol; 3-chloro-1,2-
propanediol (4 mL, 47 mmol), benzene (100 mL, solvent), and p-
toluenesulfonic acid (catalyst) were refluxed with 2 for 7 h using a
water separator to remove the produced water. The reaction was
monitored by TLC (GF254 silica gel plates), and neutralized by a
476 (0.19), 384 (0.08), 269 (0.13), 229 (0.25), 197 (0.21), 147
(100), 119 (2.26), 95 (3.89), 69 (3.86), 55 (14.57), and 43 (8.08);
Anal. Calcd for C30
H 10.26%.
2
H49ClO : C 75.50, H 10.37%; found C 75.67,
Cholest-5-en-3-one 3-Piperidinopropylene Acetal (1a).
Compund 3 (3 g, 6.3 mmol) dissolved in piperidine (10 mL) was
refluxed for 8 h. After filtration and removal of excess piperidine,
the residue was extracted with water and chloroform. The organic
layer was separated by a silica-gel column chromatography, eluted
2 3
K CO solution (5%). The resulting solution was washed twice
with water and the organic layer was separated. After removal of
the solvent, the residue was isolated by a silica-gel column chro-
matograph using mixed solvents of petroleum ether and ethyl ace-
3
with chloroform, followed by CHCl /MeOH (20:1 v/v). 1a was
separated and recrystallized in absolute ethanol (65% yield). Mp
1
tate (50:1 v/v) as an eluent; 3 was obtained (20% yield). Mp 98–
3
148–150 °C; H NMR (500 MHz, CDCl ) δ 0.68 (s, 3H, H-19),
1
1
00 °C; H NMR (500 MHz, CDCl
3
) δ 0.69 (s, 3H, H-19), 0.86 (d,
0.86 (d, J = 2.2 Hz, 3H, H-26), 0.87 (d, J = 2.2 Hz, 3H, H-27),
0.92 (d, J = 6.5 Hz, 3H, H-21), 1.02 (s, 3H, H-18), 1–2 (strong
J = 2.2 Hz, 3H, H-26), 0.88 (d, J = 2.2 Hz, 3H, H-27), 0.92 (d, J
6.5 Hz, 3H, H-21), 1.05 (s, 3H, H-18), 1–2 (strong coupling,
8H, CH , CH), 3.46 (m, 1H, H-3ꢀ), 3.59 (m, 1H, H-1ꢀ), 3.91, 3.93
m, 1H, H-3ꢀ integral ratio 25:75), 4.11, 4.13 (m, 1H, H-1ꢀ integral
ratio 25:75), 4.33 (m, 1H, H-2ꢀ), and 5.34 (t, J = 2.5 Hz, 1H, H-
=
2
coupling, 34H, CH , CH), 2.46 (br.s, 4H, H-4ꢀ, H-8ꢀ), 2.55 (m, 1H,
2
2
H-3ꢀ), 2.65 (m, 1H, H-3ꢀ), 3.59, 3.61 (m, 1H, H-1ꢀ integral ratio
12:88), 4.05, 4.10 (m, 1H, H-1ꢀ integral ratio 88:12), 4.33 (m, 1H,
H-2ꢀ), and 5.34 (t, J = 2.5 Hz, 1H, H-6); IR (KBr) 2936, 1364,
(
); IR (KBr) 2947, 1097, and 1116 cm− ; EIMS m/z 478 (0.07),
1
and 1091 cm ; EIMS m/z 525 (0.08), 384 (0.07), 269 (0.07), 229
−1
6

M. Z. Zhu et al.
Bull. Chem. Soc. Jpn., 75, No. 10 (2002) 2211
found C 79.51, H 11.07, N 2.56%.
Cholest-5-en-3-one 3-(Dimethylamino)propylene Acetal
(
1c). An ethanol solution of compound 3 (3 g, 6.3 mmol) and
dimethylamine (4 g, 89 mmol) in autoclave was heated at 120 °C
for 10 h. After ethanol was removed, the residue was extracted
with water and CHCl
over a silica-gel column with CHCl
(
(
3
. The organic layer was chromatographed
/MeOH (15:1), giving 1c
60% yield). Mp 140–141 °C; H NMR (500 MHz, CDCl ) δ 0.68
s, 3H, H-19), 0.86 (d, J = 2.2 Hz, 3H, H-26), 0.88 (d, J = 2.2 Hz,
3
1
3
3
2
2
H, H-27), 0.92 (d, J = 6.5 Hz, 3H, H-21), 1.04 (s, 3H, H-18), 1–
(strong coupling, 28H, CH , CH), 2.27 (s, 6H, H-4ꢀ and H-5ꢀ),
.35 (m, 1H, H-3ꢀ), 2.48 (m, 1H, H-3ꢀ), 3.57, 3.62 (m, 1H, H-1ꢀ in-
2
tegral ratio 59:41), 4.04, 4.11 (m, 1H, H-1ꢀ integral ratio 41:59),
4
2
4
.24 (m, 1H, H-2ꢀ), and 5.34 (t, J = 2.5 Hz, 1H, H-6); IR (KBr)
936, 2884, 2867, 2766, 1380, 1365, and 1099 cm− ; EIMS m/z
85 (0.10), 382 (0.13), 269 (0.15), 229 (0.37), 156 (2.93), 102
1
(
11.10), 84 (7.79), 58 (100), and 43 (5.39); Anal. Calcd for
: C 79.10, H 11.43, N 2.88%; found C 78.87, H 11.19,
32 2
C H55NO
N 3.05%.
A methanol solution of compound 3 (2 g, 4.2 mmol) and trime-
thylamine (5.9 g, 0.1 mol) in an autoclave was heated at 120 °C
for 24 h. Upon removing methanol, the residue was separated
with a silica-gel column; interestingly, a demethylation product
1
c, rather than 1i, was obtained as the product. Alternatively,
when the reaction was carried out at 80 °C for 7 d, the same results
were obtained.
Cholest-5-en-3-one 3-[2-(Dimethylaminoethylamino)]pro-
pylene Acetal (1d). Compound 3 (3 g, 6.3 mmol) dissolved in
(
asym) N,N-dimethylethylenediamine (10 mL) was refluxed for 24
h. After removing excess (asym) N,N-dimethylethylenediamine,
the residue was separated by a silica-gel column chromatography
3
using CHCl /MeOH (10:1 v/v) as an eluent, and recrystallized in
petroleum ether; 1d was obtained (50% yield). Mp 119–120 °C;
1
H NMR (500 MHz, CDCl
.2 Hz, 3H, H-26), 0.88 (d, J = 2.2 Hz, 3H, H-27), 0.92 (d, J =
6.5 Hz, 3H, H-21), 1.02 (s, 3H, H-18), 1–2 (strong coupling, 28H,
CH , CH), 2.22 (s, 6H, H-6ꢀ, and H-7ꢀ), 2.41 (t, J = 6.2 Hz, 2H, H-
3
): δ 0.68 (s, 3H, H-19), 0.86 (d, J =
2
Fig. 5. Effect of cationic Lipid to DNA ratio on the transfec-
tion activity. Cells (Hela) were transfected with 1 µg of
PCMV Luc plamid DNA various amounts of cationic lip-
ids. The level of luciferase gene expression (A) and total
protein recovery (B) in cells transfected with cationic lip-
ids 1b, 1c, and 1f at 2.5, 5.0, 7.5, and 10.0 nmol/well, re-
spectively.
2
5ꢀ), 2.72 (m, 3H, H-3ꢀ, and H-4ꢀ), 2.76 (m, 1H, H-3ꢀ), 3.63, 3.67
(m, 1H, H-1ꢀ integral ratio 47:53), 4.02, 4.05 (m, 1H, H-1ꢀ integral
ratio 53:47), 4.24 (m, 1H, H-2ꢀ), and 5.34 (t, J = 2.5 Hz, 1H, H-
6); IR (KBr) 2946, 2819, 2769, 1332, 1137, and 1104 cm−
1
;
EIMS m/z 485 (0.16), 384 (0.06), 269 (0.08), 229 (0.32), 156
7.85), 102 (32.35), 84 (20.27), 58 (100), and 43 (8.48); Anal. Cal-
cd for C34 : C 77.20, H 11.46, N 5.30%; found C 76.96, H
11.16, N 5.55%.
(
(
0.28), 196 (0.81), 142 (3.17), 98 (100), 69 (3.27), and 55 (5.56);
Anal. Calcd for C35 : C 79.92, H 11.33, N 2.66%; found C
9.64, H 11.19, N 2.83%.
Cholest-5-en-3-one 3-(1-Pyrrolidinyl)propylene Acetal (1b).
60 2 2
H N O
H59NO
2
7
Cholest-5-en-3-one 3-(2-Aminoethylamino)propylene Ace-
tal (1e). A mixture of compound 3 (2 g, 4.2 mmol) and ethyl-
enediamine (10 mL) was refluxed for 10 h under nitrogen. After
removing excess ethylenediamine, the residue was separated by a
Compound 3 (3 g, 6.3 mmol) dissolved in pyrrolidine (10 mL)
was refluxed for 6 h. After removing excess pyrrolidine, the resi-
due was extracted with water and CHCl
separated by a silica-gel column with CHCl
3
. The organic layer was
/MeOH (20:1 v/v) to
silica-gel column chromatography using CHCl
an eluent to give 1e (56% yield). Mp 98–101 °C; H NMR (500
MHz, CDCl ): δ 0.67 (s, 3H, H-19), 0.86 (d, J = 2.2 Hz, 3H, H-
26), 0.87 (d, J = 2.2 Hz, 3H, H-27), 0.91 (d, J = 6.5 Hz, 3H, H-
21), 1.02 (s, 3H, H-18), 1–2 (strong coupling 28H, CH , CH), 2.42
3
/MeOH (8:1) as
1
3
1
give 1b (76% yield). Mp 104–106 °C; H MNR (500 MHz,
CDCl
3
3
) δ 0.68 (s, 3H, H-19), 0.86 (d, J = 2.2 Hz, 3H, H-26), 0.87
d, J = 2.2 Hz, 3H, H-27), 0.92 (d, J = 6.5 Hz, 3H, H-21), 1.01 (s,
(
2
3
H, H-18), 1–2 (strong coupling 32H, CH
2
, CH), 2.40 (br.s, 4H,
(t, J = 6.3 Hz, 2H, H-5ꢀ), 2.70 (m, 3H, H-3ꢀ, and H-4ꢀ), 2.72 (m,
1H, H-3ꢀ), 3.65, 3.68 (m, 1H, H-1ꢀ integral ratio 52:48), 4.07, 4.11
(m, 1H, H-1ꢀ integral ratio 48:52), 4.25 (m, 1H, H-2ꢀ), and 5.33 (t,
H-4ꢀ, H-7ꢀ), 2.43 (m, 1H, H-3ꢀ), 2.51 (m, 1H, H-3ꢀ), 3.55, 3.64 (m,
H, H-1ꢀ integral ratio 41:59), 4.08, 4.10 (m, 1H, H-1ꢀ integral ra-
1
−
1
tio 59:41), 4.28 (m, 1H, H-2ꢀ), and 5.33 (t, J = 2.5 Hz, 1H, H-6);
J = 2.5 Hz, 1H, H-6); IR (KBr) 3441, 2938, 2884, and 1096 cm
EIMS m/z 470 (4.92), 384 (0.69), 229 (0.89), 211 (30.45), 156
(30.96), 113 (100), and 55 (17.6); Anal. Calcd for C32 : C
76.75, H 11.27, N 5.59%; found C 76.48, H 11.15, N 5.72%.
;
IR (KBr) 3440, 2939, 1374, and 1095 cm− ; EIMS m/z 511 (0.08),
1
3
84 (0.08), 229 (0.21), 178 (2.95), 128 (4.59), 84 (100), and 55
56 2 2
H N O
(
6.54); Anal. Calcd for C34 : C 79.78, H 11.22, N 2.74%;
H57NO
2

2
212 Bull. Chem. Soc. Jpn., 75, No. 10 (2002)
Cationic Lipids for Gene Delivery
Cholest-5-en-3-one 3-(Trimethylammonio)propylene Ace-
tal Iodide (1f). A CHCl /MeOH (1:1, v/v) solution of com-
3
pound 1c (1 g, 2.1 mmol) and iodomethane (1 mL, 16 mmol) was
stirred in darkness for 24 h. After removing the solvent, 1f could
µL. DNA/lipid complexes in appropriate ratios were prepared by
mixing equal volumes of a diluted DNA solution and lipid suspen-
sion. The mixture (250 µL) was incubated for 5–10 min at room
temperature before being added to cells.
be directly crystallized in a mixed solution of CHCl
3
and absolute
ethanol (75% yield). Mp >300 °C; H NMR (500 MHz, CDCl ):
δ 0.68 (s, 3H, H-19), 0.86 (d, J = 2.2 Hz, 3H, H-26), 0.88 (d, J =
Cell Culture. Five cell lines derived from different origins
were used for this study. Murine melanoma BL-6 cells were cul-
tured in a RPMI Medium supplemented with 10% fetal bovine
sencm (FBS). Human emlryonic kidney (293 cells), Hela cells,
NIH3T3, and 3LL were cultured in a DMEM medium with 10%
FBS.
1
3
2
1
3
1
.2 Hz, 3H, H-27), 0.92 (d, J = 6.5 Hz, 3H, H-21), 1.04 (s, 3H, H-
8), 1–2 (strong coupling, 28H, CH , CH ), 3.44 (m, 1H, H-3ꢀ),
2
.54 (s, 9H, H-5ꢀ, H-6ꢀ, H-7ꢀ), 3.72 (m, 1H, H-3ꢀ), 4.31 (m, 1H, H-
ꢀ), 4.43 (m, 1H, H-1ꢀ), 4.72 (m, 1H, H-2ꢀ), and 5.33 (t, J = 2.5
Transfection In Vitro. For a standard transfection, cells were
−
1
4
Hz, 1H, H-6); IR (KBr) 2939, 2884, and 1113 cm ; Anal. Calcd
: C 63.13, H 9.33, N 2.23%; found C 63.39, H
seeded at a density of 5 × 10 cells per well in 48-well plates 24 h
for C33
H58INO
2
before the addition of DNA/lipid complexes. After removing the
medium, the cells were incubated with a DNA/lipid mixture
(250 µL/well) for 5 h, followed by the addition of 27.5 µL of fetal
bovine serum (FBS) to each well. The transfection solution was
replaced with a fresh medium containing 10% FBS 24 h post ex-
posure to the DNA/lipid mixture. Cells were collected after addi-
tional incubation for 24 h, and a cell lysate was prepared by PBS
washing (3×) and the addition of 100 µL of lysis a buffer (0.1 M
tris-HCl 0.1 TritonX-100, 2 mM EDTA, pH = 7.80) per well. The
cell lysates were collected and centrifuged in a microcentrifuger
(12,000 rpm, 5 min, 4 °C), and the supernatant was used for mea-
suring the luciferase activity. A luciferase assay kit purchased
from promega (Madison, WI) was used in a Luminometerc Autol-
9
.17, N 2.15%.
Cholest-5-en-3-one 3-(N-Methylpiperidimio)propylene Ace-
tal Iodide (1g). A CHCl /MeOH (1:1, v/v) solution of com-
3
pound 1a (1.9 mmol) and iodomethane (1 mL, 16 mmol) was
stirred in darkness for 24 h at room temperature. After removing
the solvent, 1g was directly crystallized in a mixed solution of
CHCl
3
and absolute ethanol (70% yield). Mp 280 °C (decomp);
1
H NMR (500 MHz, CDCl
Hz, 3H, H-26), 0.87 (d, J = 2.2 Hz, 3H, H-27), 0.92 (d, J = 6.5
3
) δ 0.68 (s, 3H, H-19), 0.86 (d, J = 2.2
Hz, 3H, H-21), 1.02 (s, 3H, H-18), 1–2 (strong coupling, 30H,
CH
br.s, 6H, H-4ꢀ, H-8ꢀ, and H-3ꢀ), 3.74 (m, 1H, H-1ꢀ), 4.06 (m, 1H,
H-1ꢀ), 4.18 (br.s, 1H, H-2ꢀ), and 5.73 (br.s, 1H, H-6); IR (KBr)
2
, CH), 2.37 (m, 4H, H-5ꢀ, and H-7ꢀ), 3.00 (br.s, 3H, H-9ꢀ), 3.30
(
33
umat according to the established protocol. Tenal of cell lysate
from transfected cells was used in each assay, and the lumines-
cence was measured for 10 s. The protein concentration of the su-
pernatant was determined by a standard protein assay using a pro-
tein assay reagent purchased from Biorad (Hercules, CA)
−
1
2
936, 2868, 1366, 1331, 1112, and 1089 cm ; Anal. Calcd for
: C 64.74, H 9.38, N 2.10%; found C 64.50, H 9.12, N
36 2
C H62INO
2
.26%.
Cholest-5-en-3-one 3-(N-Methylpyrrolidimio)propylene Ac-
etal Iodide (1h). A mixture of compound 1b (1.95 mmol) and
iodomethane (1 mL) in CHCl /DMSO (1:1, v/v) binary solvents
was stirred for 24 h in the dark. After removing the solvents, 1h
was crystallized from CHCl /MeOH solvents (63% yield). Mp
64 °C (decomp); H NMR (500 MHz, CDCl ): δ 0.68 (s, 3H, H-
3
This work was supported by the Ministry of Science and
Technology and National Natural Science Foundation of China
3
(
No. 29772031).
1
2
1
2
3
9), 0.86 (d, J = 2.2 Hz, 3H, H-26), 0.87 (d, J = 2.2 Hz, 3H, H-
7), 0.92 (d, J = 6.5 Hz, 3H, H-21), 1.02 (s, 3H, H-18), 1–2
References
(
strong coupling, 28H, CH
2
, CH), 2.39 (m, 4H, H-5ꢀ, and H-6ꢀ),
1
a) R. G. Crystal, Science 270(20), 404 (1995). b) M. D.
3
.02 (br.s, 3H, H-8ꢀ), 3.35 (br.s, 6H, H-4ꢀ, H-7ꢀ, and H-3ꢀ), 3.76
Brown, A. G. Schätzlein, and I. F. Uchegbu, Int. J. Pharm., 229, 1
(
(
m, 1H, H-1ꢀ), 4.10 (m, 1H, H-1ꢀ), 4.18 (br.s, 1H, H-2ꢀ), and 5.62
(2001).
br.s, 1H, H-6); IR (KBr) 3439, 2938, 1376, and 1112 cm− ; Anal.
1
2
A. D. Miller, Angew. Chem. Int. Ed. Engl., 37, 1768
Calcd for C35
H60INO
2
: C 64.30, H 9.25, N 2.14%; found C 64.57,
(1998).
H 9.43, N 2.11%.
Preparation of Lipid Formulation and DNA/Lipid Com-
3
4
5
6
D. Liu, J. Liposome Res., 7(2&3), 187-205 (1997).
X. Gao and L. Huang, Gene Ther., 2, 710 (1996).
P. L. Felgner and G. M. Ringold, Nature, 337, 387 (1989).
P. L. Felgner, T. R. Gadek, M. Holm, R. Roman, H. W.
plexes. Lipid formulations were prepared as previously de-
3
3
scribed. In brief, a mixture of appropriate amounts of cationic
lipid and DOPE (molar ratio 1:1) was dissolved in chloroform.
After removing chloroform under a stream of nitrogen gas, the re-
sulting lipid films were dried under a vacuum for 2 h to remove
trace amounts of the organic solvent, and then hydrated in phos-
phate buffered saline (PBS pH = 7.4) at a cationic lipid concentra-
tion of 5 mM. Hydration proceeded for 2 h at room temperature
and the lipid suspension was then sonicated for 5 min using a bath
sonicator.
Chan, M. Wenz, J. P. Northrop, G. N. Ringold, and M. Denielsen,
Proc. Natl. Acad. Sci. U.S.A., 84, 7413 (1987).
7
a) T. Ren, G. Zhang, Y. K. Song, and D. Liu, J. Drug
Targeting, 7(4), 285 (1999). b) N. Mamoru and N. Ari, Adv. Drug
Delivery Rev., 52, 197 (2001). c) K. G. Yamuna, V. Sandhya, and
S. B. Santanu, FEBS Lett., 473, 341 (2000).
8
J. H. Felgner, R. Kumar, C. N. Sridhar, C. J. Wheller, Y. J.
Tsai, R. Border, R. Ramsey, P. Martin, and P. L. Felgner, J. Biol.
Chem., 269, 2550 (1994).
pCMV-Luc plasmid DNA containing the firefly luciferase gene
driven by the cytomegalovirus immediate early promoter (CMR)
was diluted in a serum-free CHO-S-SFM medium (life Technolo-
gy inc.) to give a DNA concentration at 1 µg/125 µL. The cationic
lipid suspension (5 mM) was diluted with Hank’s balanced salt so-
lution (HBSS) to generate different cationic lipid concentrations:
9
E. R. Lee, J. Marshall, C. S. Siegel, C. Jiang, C. S. Yew, M.
R. Nichols, J. B. Nietupski, R. J. Ziegler, M. Lane, K. X. Wang, N.
C. Wan, R. K. Scheule, D. J. Harris, A. E. Smith, and S. H. Cheng,
Hum. Gene Ther., 7, 1701 (1996).
10 R. P. Balasubramaniam, M. J. Bennett, A. M. Aberle, I. G.
Malone, M. H. Nantz, and R. W. Malone, Gene Ther., 3, 163
2
.5 nmol/125 µL; 5 nmol/125 µL; 7.5 nmol/125 µL; 10 nmol/125

M. Z. Zhu et al.
Bull. Chem. Soc. Jpn., 75, No. 10 (2002) 2213
(
1996).
M. J. Bennett, A. M. Aberle, R. P. Balasubramaniam, J. G.
Malone, R. W. Malne, and M. H. Nantz, J. Med. Chem., 40, 4069
1997).
G. Byk, C. Dubertret, V. Escriou, M. Frederic, G. Jaslin, R.
(1997). b) N. Mamoru and N. Ari, Adv. Drug Delivery Rev., 52,
197 (2001).
23 R. G. Cooper, G. H. Etheridger, L. Steuart, J. Maarshall, S.
Rudginsk, S. H. Cheng, and A. D. Miller, Chem. Eur. J., 4, 137
(1998).
1
1
(
12
Rangara, B. Pitard, J. Crouzet, P. Wils, B. Schwarz, and D.
Scherman, J. Med. Chem., 41, 224 (1998).
24 X. Gao and L. Huang, Biochem. Biophys. Res. Commun.,
179, 280 (1991).
1
3
J. Wang, X. Guo, Y. Xu, L. Barron, and F. C. Szooka, J.
25 T. Ren, G. S. Zhang, F. Liu, and D .Liu, Bioorg. Med.
Chem. Lett., 2000, 1.
Med. Chem., 41, 224 (1998).
1
1
1
4
5
6
T. Ren and D. Liu, Bioorg. Med. Chem. Lett., 1999, 1247.
T. Ren and D. Liu, Tetrahedron Lett., 40, 209 (1999).
G. J. Nabel, A. E. Chang, E. G. Nabel, G. E. Plautz, W.
26 F. Liu, J. Yang, L. Huang, and D. Liu, Pharm. Res., 13,
1642 (1996).
27 H. O. House, H. C. Muller, C. G. Pitt, and P. P. Wickham,
J. Org. Chem., 28, 2407 (1963).
Engminger, B. A. Fox, P. Felgner, S. Shu, and K. Cho, Hum. Gene
Ther., 5, 57 (1994).
28 S. Saito and E. L. May, J. Org. Chem., 26, 4536 (1962).
29 D. A. Jaeger, J. Jamrozi, T. G. Golich, M. W. Clernan, and
J. Mohebalian, J. Am. Chem. Soc., 111, 300 (1989).
30 A. K. Chattopadhyay, M. Drifford, and C. Treiner, J. Phy.
Chem., 89, 1537 (1985).
31 As a chemical evidence for the degradation properties of
the lipids synthesized in this work, hydrolysis experiments were
carried out. The results showed that the lipids containing ketal as
linker used in this research were completely degrade in 0.01 mol/
L HCl ethanol solution at room temperature for 3 days. However,
the corresponding ether remained no change under conditions. Al-
though there is difference in nature between the chemical system
and biological system, it demonstrated that the ketal compounds
are degradable in acidic medium.
1
7
G. J. Nabel, E. G. Nabel, Z. Y. Yang, B. A. Fox, G. E.
Plautz, X. Gao, L. Huang, S. Shu, D. Gordon, and A. E. Chang,
Proc. Natl. Acad. Sci. U.S.A., 90, 11307 (1993).
1
8
A. Sobrero, D. Kerr, B. Glimelius, E. Van Cutsem, G.
Milano, D. M. Pritchard, P. Rougier, and M. Aapro, Eur. J. Can.,
6, 559 (2000).
M. Abner, C. Sunil, R. Jack, and A. R. Rajagopal, Biotech.
Adv., 19, 279 (2001).
E. J. Sorscher, J. J. Logan, R. A. Frizzel, R. K. Lyrene, Z.
3
1
9
2
0
Belok, J. Y. Dong, M. D. Durall, P. L. Felgner, S. Matalon, L.
Walker, and B. J. Wiatrak, Hum. Gene Ther., l5, 1259 (1994).
2
1
N. J. Caplen, E. W. F. W. Alton, P. G. Middleton, J. R.
Doring, B. J. Stevenson, X. Gao, Sr. Durham, K. Jeffery, M. E.
Hodson, C. Coutelle, L. Huang, D. J. Proteous, R. Williamson,
and D. M. Geddes, Nature Med., 1, 39 (1995).
32 H. Farhood, R. Bottega, R. M. Epand, and L. Huang, Bio-
chim. Biophys. Acta., 1111, 239 (1992).
2
2
a) N. S. Templeton, D. D. Lasic, P. M. Frederik, H. H.
33 F. Liu, H. Qi, L. Huang, and D. Liu, Gene Ther., 4, 517
(1997).
Strey, D. D. Roberts, and G. N. Pavlakis, Nat. Biotechnol., 15, 651