Synthesis and transfection efficiency of cationic oligopeptide lipids: Role of linker

Article abstract of DOI:10.1021/bc2002874

In the design of new cationic lipids for gene transfection, the chemistry of linkers is widely investigated from the viewpoint of biodegradation and less from their contribution to the biophysical properties. We synthesized two dodecyl lipids with glutami

Full text of DOI:10.1021/bc2002874

Article
pubs.acs.org/bc
Synthesis and Transfection Efficiency of Cationic Oligopeptide Lipids:
Role of Linker
Vijaya Gopal,† Jennifer Xavier,† Md. Zahid Kamal,† Srinath Govindarajan,† Makoto Takafuji,‡ Shuta Soga,‡
Takayuki Ueno,‡ Hirotaka Ihara,‡ and Nalam M. Rao*,†
†Centre for Cellular and Molecular Biology (Council of Scientific and Industrial Research), Uppal Road, Hyderabad, India 500007
‡Department of Applied Chemistry and Biochemistry, Kumamoto University, Kumamoto, Japan
S
* Supporting Information
ABSTRACT: In the design of new cationic lipids for gene
transfection, the chemistry of linkers is widely investigated
from the viewpoint of biodegradation and less from their
contribution to the biophysical properties. We synthesized two
dodecyl lipids with glutamide as the backbone and two lysines
to provide the cationic headgroup. Lipid 1 differs from Lipid 2
by the presence of an amide linkage instead of an ester linkage
that characterizes Lipid 2. The transfection efficiency of
lipoplexes with cholesterol as colipid was found to be very high
with Lipid 1 on Chinese Hamster Ovary (CHO) and HepG2 cell lines, whereas Lipid 2 has shown partial transfection efficiency
on HepG2 cells. Lipid 1 was found to be stable in the presence of serum when tested in HepG2 and CHO cells albeit with lower
activity. Fluorescence-based dye-binding and agarose gel-based assays indicated that Lipid 1 binds to DNA more efficiently than
Lipid 2 at charge ratios of >1:1. The uptake of oligonucleotides with Lipid 1 was higher than Lipid 2 as revealed by confocal
microscopy. Transmission electron microscopy (TEM) images reveal distinct formation of liposomes and lipoplexes with Lipid 1
but fragmented and unordered structures with Lipid 2. Fusion of Lipids 1 and 2 with anionic vesicles, with composition similar to
plasma membrane, suggests that fusion of Lipid 2 was very rapid and unlike a fusion event, whereas the fusion kinetics of Lipid 1
vesicles was more defined. Differential scanning calorimetry (DSC) revealed a high Tm for Lipid 1
(65.4 °C) while Lipid 2 had a Tm of 23.5 °C. Surface area−pressure isotherms of Lipid 1 was less compressible compared to
Lipid 2. However, microviscosity measured using 1,6-diphenyl-1,3,5-hexatriene (DPH) revealed identical values for vesicles made
with either of the lipids. The presence of amide linker apparently resulted in stable vesicle formation, higher melting temperature,
and low compressibility, while retaining the membrane fluid properties suggesting that the intermolecular hydrogen bonds of
Lipid 1 yielded stable lipoplexes of high transfection efficiency.
Transfection efficiency by cationic lipids is strongly depend-
ent on the cell type and inclusion of colipids in formula-
tions.16−18 To obtain insight into the relationship between
transfection efficiency and the properties such as size, charge,
stability, and interaction with the biological membranes, cationic
lipids have been extensively investigated.15 These studies
indicated that compact, fluid, and positively charged lipoplexes
overcome serum destabilization, enhanced cellular associations,
and eventual escape from the endosomes by fusion.19 Amino
acids or short peptides have been used, especially as headgroup
moieties, in the design of cationic lipids.15 Cationic amino acids
lysine and arginine have been extensively used to provide the
cationic headgroup15 to facilitate complexation with the nucleic
acids. The pKa properties of amino acid histidine have been
used successfully20 to cause endosomolysis after uptake of
lipoplex. However, use of amino acids as backbone for the
cationic lipid has not been investigated. Earlier, we reported
INTRODUCTION
■
Cationic lipid-based formulations offer attractive solutions for
in vitro delivery of nucleic acids.1−4 Liposomes containing
cationic lipids electrostatically bind with the polyanionic nucleic
acids resulting in a lipoplex that protects nucleic acids from
degradation and facilitates endosomal uptake.5,6 Structurally,
cationic lipids are amphiphilic with a hydrophilic cationic
headgroup connected with a linker to hydrophobic tails, usually
alkyl chains.7−9 Extensive investigations on chemistry and trans-
fection efficiency of cationic lipids suggest that a wide variety
of molecules are suitable when appropriate amphiphilicity is
maintained.5,6 Alkyl chains
and steroid molecules11 are a
8,10
few examples that were found suitable as hydrophobic groups,
while a large variety of cationic groups such as primary or
tertiary amines together with polyamine and quarternary
ammonium,12,13 guanidino groups,14 and amino acids15 were
found to enhance transfection efficiency. The enhancement in
transfection efficiency in the presence of a neutral lipid or a
colipid such as phosphatidylethanolamine or cholesterol was
the prime motivation to successfully test a large number of
cationic lipids with colipids.
Received: June 3, 2011
Revised: September 19, 2011
Published: October 11, 2011
© 2011 American Chemical Society
2244
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
synthesis and transfection efficiencies of pyridinium-based
cationic lipids with glutamide as the backbone.21 In the design
of cationic lipids, linkers are used to connect the headgroup
or the hydrophobic portion to the backbone, for example,
glycerol.22−24 This is to provide biodegradability and thereby
safety due to decreased toxicities. Ester linkages are preferred
over ether linkers25 primarily for their biodegradability. In some
reports, the linkers are designed to enhance the release of the
nucleic acid from the lipoplex in response to a change in
conditions inside the cell. For example, low pH labile bonds
are used to dissociate the headgroup from the lipid in the
endosomal compartment and release the plasmid DNA.13
Reduction of linkers with disulfide bonds in the cytoplasmic
milieu destabilizes the complex26 thereby separating the nucleic
acid from the liposomes. Recently, in an otherwise chemically
similar lipid, the linker orientation was shown to have dramatic
effect on the transfection properties of the lipid.24 We report in
this study two lysinated-glutamide lipids which differ from each
other only in their linker chemistry, i.e., either amide or an ester
designated as Lipid 1 and Lipid 2, respectively. Lipid 1 with an
amide linker melts at a higher temperature and has higher
transfection efficiency than the low melting counterpart Lipid
2 with the ester linker. This striking structural difference
between the two lipids motivated further investigations to
address whether the linker has any role in determining the
transfection efficiency.
CH2CH2NH), 1.62 (m, 2H, CHCH2(CH2)3NH), 1.83 (q, 2H,
CH3OC(O)C*HCH2), 3.08−3.18 (m, 4H, C*H-
(CH2)3CH2NHC(O)), 3.68 (s, 3H, CH3), 4.18 (s, 1H,
CH3OC(O)C*H), 4.53 (s, 1H, NHC(O)C*HNHC(
O)), 5.07 (m, 6H, C*H2Ph), 7.35 (d, 15H, C6H5). Anal.
Found: H, 6.77%; C, 63.4%; N, 7.97%. Calcd. for C37H46N4O9:
H, 6.71%; C, 64.3%; N, 8.11%.
(B). Z-Lys(Z)-Lys(Z)-OH. Z-Lys(Z)-Lys(Z)-OMe (4.6 g, 6.60
mmol) and 1 N NaOH 19 mL (19 mmol) were dissolved in
100 mL methanol, stirred for 1 h at room temperature, and pH
was adjusted to 2 by 1 N HCl. After removing methanol from
the solution in vacuo, the residue was added to ethanol. The
obtained white precipitate was removed by filtration, and the
solution was concentrated and added to n-hexane to give a
white precipitate. The precipitate was collected by filtration and
dried in vacuo: yield 4.30 g (95% 6.35 mmol) mp 145−147 °C;
FT-IR (KBr) 3322, 2941, 2863, 1692, 1652, and 1541 cm−1; 1H
NMR (CDCl3) δ 1.35 (m, 4H, C*HCH2CH2), 1.47 (m, 4H,
CH2CH2NH), 1.63−1.86 (br, 4H, C*HCH2), 3.10 (d, 4H,
C*H(CH2)3CH2NHC(O)), 4.26 (s, 1H, HOC(O)C*H),
4.50 (s, 1H, NHC(O)C*HNHC(O)), 5.06 (m, 6H, CH2Ph),
7.29 (d, 15H, C6H5). Anal. Found: H, 6.86%; C, 63.44%; N, 8.08%.
Calcd. for C36H44N4O9: H, 6.55%; C, 63.9%; N, 8.28%.
Synthesis of Lipid 1 (2C12-Gln-Lys-Lys). (A). Synthesis of
2C12-Gln. Synthesis of 2C12-Gln was described earlier (Bio-
conjugate Chem. 2006, 17, 1530).
(B). 2C12-Gln-Lys(Z)-Lys(Z)-Z. 2C12-Gln (0.93 g, 1.93
mmol), Z-Lys(Z)-Lys(Z)-OH (1.30 g, 1.92 mmol), and
triethylamine (0.37 g, 3.84 mmol) were dissolved in dry THF
(100 mL). The solution was cooled to 0 °C, and DEPC (0.52 g,
2.88 mmol) was added to the solution and stirred for 1 day at
room temperature. After stirring, the solution was concentrated
in vacuo, and the lipid was recrystallized from ethanol, which
gave a white solid powder: yield 1.07 g (50%, 0.94 mmol); mp
184−187 °C; FT-IR (KBr) 3295, 2925, 2853, 1693, 1637, and
1543 cm−1; 1H NMR (CDCl3) δ 0.86−0.89 (t, 6H, CH3), 1.25
(m, 36H, CH3(CH2)9), 1.33 (s, 4H, CH2CH2NHC(O)),
1.48 (br, 6H, C*H(CH2)NHC(O)), 1.58−2.40 (s, 8H,
C*HCH2CH2CH2, C*HCH2CH2CH2), 2.44 (m, 2H,
C*HCH2CH2C(O)), 3.16−3.23 (br, 8H, CH2NHC(O),
CH2NHC(O)O), 4.11 (s, 1H, C*HNHC(O)), 4.35 (s,
1H, C*HNHC(O)), 4.51 (s, 1H, C*HNHC(O)), 4.92−
5.15 (br, 6H, CH2Ph) 7.28−7.34 (m, 15H, C6H5). Anal. Found:
H, 8.94%; C, 68.3%; N, 8.56%; Calcd. for C65H101N7O10: H,
8.93%; C, 68.5%; N, 8.60%.
EXPERIMENTAL PROCEDURES
■
Materials. In this work, all chemicals and solvents were
purchased and used without further purification. The chemical
structures of the lipids were identified by melting point, FTIR,
1H NMR, and elemental analysis. Melting points were deter-
mined on a micro melting point apparatus. FTIR spectra were
1
performed on a JASCO FT/IR-4000 spectrometer. H NMR
spectra were recorded by a JEOL JNM-EX400 spec-
trometer using tetramethylsilane as an internal standard.
Elemental analyses were performed with a Yanaco CHN
Corder MT-3. ESI-HRMS was performed on a QSTAR XL
hybrid MS/MS system (Applied Biosystems/MDS sciex),
equipped with an ESI source.
1,6-diphenyl-1,3,5-hexatriene, ethidium bromide, NBD-
DHPE, and Rhodamine −DHPE were purchased from Molecular
Probes (USA). 1,2-Dioleoyl-sn-glycero-3-phosphoethanolamine
(DOPE), cholesterol, L-α-phosphatidylglycerol (PG), and 1,2-
dioleoyl-sn-glycero-3-phosphocholine (DOPC) were purchased
from Avanti Polar Lipids (USA). Fluorescein (FAM) Labeling
Kits were sourced from Ambion. 3-(4,5-Dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (MTT) and other reagents
were purchased from Sigma Co.
(C). 2C12-Gln-Lys-Lys. 2C12-Gln-Lys(Z)-Lys(Z)-Z (0.93 g,
0.82 mmol) was dissolved in 100 mL ethanol with heating, and
Pd/carbon (0.5 g) was added to the solution. H2 gas was
bubbled slowly into the solution for 14 h at 70 °C. Pd/carbon
was removed by filtration, and the solution was dried in vacuo
to give a solid powder: yield 0.34 g (52%, 0.43 mmol, as
a trihydrate); mp 104−107 °C; FT-IR (KBr) 3289, 2920,
Synthesis of Z-Lys(Z)-Lys(Z)-OH. (A). Z-Lys(Z)-Lys(Z)-
OMe. Z-Lys(Z)-OH (3.42 g, 9.06 mmol), H-Lys(Z)-
OMe·HCl (3.0 g, 9.06 mmol), and triethylamine (2.76 g,
27.1 mmol) were dissolved in chloroform. The solution was
cooled to 0 °C, and DEPC (1.97 g, 1.09 mmol) was added to
the solution and stirred for 30 min at this temperature. After
stirring for 1 day at room temperature, the solution was con-
centrated in vacuo, and the residue was dissolved in chloroform.
The solution was washed with 0.2 N HCl, 5 wt % NaHCO3,
and water. The solution was dried over Na2SO4, concentrated,
and finally added to diethyl ether, which gave a white solid
powder: yield 4.56 g (73%, 6.5 mmol); mp 103−105 °C; FT-IR
1
2851, 1639, and 1557 cm−1; H NMR (CDCl3) δ 0.87−0.90
(t, 6H, CH3), 1.28 (m, 36H, CH3(CH2)9), 1.52−2.17 (br, 20H,
CH2CHNHC(O), CH2C*H, CH2CH2C*H, CH2(CH2)2C*H),
3.12−3.35 (br, 8H, CH2-NHC(O), CH2NHC(CO)O),
4.31−4.56 (br, 3H, C*HNHC(O)). Anal. Found: H, 10.9%;
C, 64.9%; N, 11.4%. Calcd. for C41H89N7O7: H, 11.3%; C,
62.2%; N, 12.4%. ESI-HRMS: m/z = +738.6569 (calcd value for
C41H84N7O4 = 738.6584).
Synthesis of Lipid 2 (2C12Glu-Lys-Lys). (A). Didodecyl-
L-Glutamate Toluene-p-Sulfonate (2C12L-Glu TosOH).
L-Glutamic acid (6.0 g, 40.0 mmol), p-toluenesulfonic acid
(KBr) 3316, 2942, 1741, 1690, 1650, and 1542 cm−1; H
NMR (CDCl3) δ 1.37 (m, 4H, C*HCH2CH2), 1.49 (m, 4H,
1
2245
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
(6.82 g, 48.0 mmol), and 1-dodecanol (17.8 g, 96.0 mmol)
were dissolved in toluene (250 mL), and refluxed for 5 h by use
of Dean−Stark apparatus. The solution was concentrated in
vacuo, and the residue was dissolved in 200 mL of diethyl ether,
and refrigerated to 4 °C for 1 day, which gave a white solid
powder: yield 13.1 g (49%), mp 65−66 °C. Anal. Found: H,
9.93; C, 65.4; N, 2.40; Calcd. for C36H65NO7S: H, 9.99; C,
65.9; N, 2.14. FT-IR (KBr) 2921 cm−1 (νC−H), 2850 cm−1
(νC−H), 1743 cm−1 (νester), 1533 cm−1 (δNH), 1182 cm−1 (νC−O).
1H NMR (CDCl3, internal reference: tetramethylsilane) δ
0.86−0.90 (m, 6H, CH3) δ 1.26 (m, 36H, CH3-(CH2)9), δ 1.55
(m, 4H, CH2CH2OC(O)), δ 2.15−2.25 (m, 2H,
C*HCH2CH2C(O)), δ 2.34 (s, 1H,CH3Ph), δ 2.43−2.58
(m, 2H, C*HCH2CH2C(O)), δ 3.95−4.11 (m, 4H,
CH2CH2OC(O)), δ 7.12−7.14 (d, 2H, C6H5), δ 7.72−
7.74 (d, 2H, C6H5).
samples were directly loaded on a 1% agarose gel with TAE
buffer and electrophoresed at 80 V for approximately 2 h. The
gel was stained with ethidium bromide (EtBr), post electro-
phoresis, and visualized under a transilluminator. DNA binding
was also carried out by EtBr fluorescence titrations as described.27
Ethidium Bromide (EtBr) Fluorescence Quenching. The
binding of DNA with Lipid 1 and 2 was studied using EtBr as
the fluorescent probe. The lipids were reconstituted with
colipids DOPE or cholesterol in deionized water using standard
protocols. Bound EtBr was expelled from the DNA−EtBr
complexes by the lipid. This displacement of EtBr is reflected
as a drop in the fluorescence signal, since unbound EtBr does
not fluoresce. All fluorescence measurements were carried
out using a Hitachi F-4500 fluorescence spectrophotometer.
The excitation wavelength, λex, was 516 nm and the emission
wavelength was kept at 598 nm (slit width 5 × 5 nm2). Briefly,
2.3 μg of pCMV-β-gal plasmid DNA was added to 500 μL of
20 mM Tris−HCl buffer (pH 7.4) in a fluorescence cuvette.
EtBr 0.23 μg, from a diluted stock solution, was added to DNA
and the baseline fluorescence was determined. The fluorescence
intensity obtained upon each addition of lipid was normalized
relative to the fluorescence signal of DNA−EtBr complex, in
the absence of the lipid, which was taken as 100%. The binding
of DNA by the lipid was recorded after each addition at time
intervals of 5 min.
Transfection. CHO, HepG2, cells were plated at a density
of 10 000 cells per well in a 96-well plate on the previous day.
Plasmid DNA pCMVβ-gal was purified by using the Qiagen Kit
(endotoxin free) using the manufacturer’s protocol and also by
CsCl−EtBr density gradient ultracentrifugation using standard
protocols. Plasmid (0.9 μg corresponding to 2.72 nmol) was
complexed with varying amounts of Lipid 1 and 2, formulated
with cholesterol at 1:1 mol ratio, to obtain charge ratios varying
from 1:1 to 9:1, DNA/Lipid, in DMEM without serum and
incubated at RT for 30 min. Complexes were added to cells
after appropriate dilution with DMEM to achieve 0.3 μg per
well, and incubated for 3 h. The medium was then replaced
with complete medium containing 10% serum.
(B). 2C12Glu-Lys(Z)-Lys(Z)-Z. 2C12Glu-Lys(Z)-Lys(Z)-Z was
synthesized by similar method with 2C12Gln-Lys(Z)-Lys(Z)-Z:
yield 1.20 g (51%); mp 130−131.5 °C; FT-IR (KBr) 3307 cm−1
2924 cm−1 (νC−H), 2853 cm−1 (νC−H), 1732 cm−1 (νCO(ester)),
1688 cm−1 (νCO(urethan)), 1640 cm−1 (νCO(amide)), 1543 cm−1
1
(δN−H), H NMR (CDCl3) δ 0.86−0.90 (m, 6H, CH3), δ 1.26
(m, 40H, CH3-(CH2)9, C*HCH2CH2CH2), δ 1.49 (m, 4H,
CH2CH2NHC(O)), δ 1.56 (m, 4H, CH2CH2OC(O)), δ
1.68−1.88 (br, 4H, CH2C*HNHC(O)), δ 1.94−2.01(m, 1H,
C*HCH2CH2C(O)), δ 2.11−2.16 (m, 1H, C*HCH2CH2C-
(O)), δ 2.36 (m, 2H, C*HCH2CH2C(O)), δ 3.16 (m, 4H,
CH2CH2NHC(O)), δ 4.03 (m, 4H, CH2CH2OC(O)), δ
4.15 (m, 1H, CH2OC(O)C*HNH C(O)), δ4.38 (q, 1H,
NHC(O)C*HNH C(O)), δ 4.53 (q, 1H, NHC(
O)C*HNH C(O)OCH2Ph), δ 5.07 (m, 6H, PhCH2OC(
O), δ 7.31 (d, 15H, C6H5). Anal. Found: H, 8.69; C, 67.8; N,
6.20; Calcd. For C65H99N5O12: H, 8.73; C, 68.3; N, 6.13.
(C). 2C12-Glu-LysLys. 2C12Glu-LysLys was synthesized by
similar method with 2C12Gln-LysLys: yield 0.40 g (49%, 0.47
mmol, as a hexahydrate); FT-IR (KBr) 3307 cm−1 and 2925
cm−1 (νC−H), 2854 cm−1 (νC−H), 1737 cm−1 (νCO(ester)), 1654
1
cm−1 (νCO(amide)), 1557 cm−1 (δN−H), H NMR (CDCl3) δ
0.86−0.90 (m, 6H, CH3), δ 1.27 (m, 36H, CH3(CH2)9), δ 1.50−
2.27 (br, 18H, CH2), δ 2.51 (m, 2H, C*HCH2CH2C(O)), δ
3.15 (m, 4H, CH2-NH2), δ 4.14−4.20 (m, 4H, CH2CH2OC(
O)), δ 4.30−4.57 (br, 3H, C*HNHC(O)). Anal. Found: H,
9.4; C, 58.1; N, 8.00. Calcd. for C41H93N5O12: H, 11.1; C, 58.1;
N, 8.26. ESI-HRMS: m/z = +740.6245 (calcd value for
C41H82N5O6 = 740.6265).
Preparation of Liposomes. Cationic amphiphiles Lipid 1,
2, and the colipid cholesterol were dissolved in chloroform
and mixed in glass vials at 1:1 mol ratio. Chloroform was then
removed with a thin flow of nitrogen and the dried lipid film
was kept under vacuum for 4−5 h. Subsequently, deionized
water was added to the dried lipid film for overnight rehydra-
tion. The vial was vortexed thoroughly at RT to produce
multilamellar vesicles (MLVs). MLVs were sonicated until
clarity to obtain small unilamellar vesicles (SUV).
DNA-Binding Assay. Agarose Gel Electrophoresis. Plas-
mid pGFPN3 was complexed with cationic lipids at charge
ratios varying from 0.3:1 to 3:1 (lipid/DNA) to measure the
DNA binding ability. In a typical binding assay, 0.6 nmol of
DNA, corresponding to 200 ng of plasmid DNA, was com-
plexed with the lipids, at indicated charge ratios, in 10 mM
phosphate buffer (pH 7.5), deionized water or 0.5× PBS in a
volume of 20 μL and incubated at room temperature for
30 min. After addition of the tracking dye bromophenol blue,
Serum stability experiments were performed as described21 in
CHO and HepG2 cells. Briefly, transient transfections, in vitro,
were done in the presence of fetal calf serum. Lipid/DNA com-
plexes were prepared in the absence of serum and incubated for
a period of 30 min at room temperature. Serum was then added
to the complex to obtain the 10% final concentration and then
added to the cells, seeded on a 96-well plate and the charge
ratios varied from 2:1 to 12:1 Lipid/DNA. Data is represented
as β-gal activity/well (20 000 cells). Reporter gene assay was
carried out 48 h post transfection and efficiency depicted as
relative β-gal activity. In these assays, cells were washed with
PBS and lysed in 50 μL of lysis buffer (0.25 M Tris.HCl, pH
8.0, and 0.5% NP40). β-Gal activity was estimated by adding
50 μL of 2× substrate (1.33 mg/mL ONPG, 0.2 M sodium
phosphate, pH 7.15, and 2 mM magnesium chloride) to an
equal volume of the lysate in a fresh 96-well plate and incubated
at 37 °C. Absorption at 405 nm was converted to β-gal units
by using a standard curve generated with pure commercial
enzyme.
Differential Scanning Calorimetry. Thermal transitions
in Lipids 1 and 2 were performed using VP-DSC MicroCal
calorimeter (VP-ITC model). The calorimetry was performed
using Lipids 1 and 2 (8.8 mg/mL) vesicles prepared in
deionized water as described above except the sonication step.
All samples (250 μL) were degassed before acquiring the scans
2246
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
which were performed between 10 °C and 95 o C at a scan rate
of 60 deg/h. Water or buffer baselines were obtained and
subtracted from lipid scans.
488−5′dUTP): RhDHPE-labeled lipid complexes were visual-
ized using the 488 nm excitation of argon laser, 405 nm
excitation of diode while Rhodamine was detected using HeNe
excitation wavelength of 543 nm. The emission collection
wavelengths were 500−540 nm for Oregon Green, 415−485
nm for Hoechst 33258, and 560−615 nm for Rhodamine. In
each panel, 5 slices of 0.5 μm were combined. Last panel
indicates overlaid images depicting nuclear staining with
Hoechst 33258.
Fluorescence Anisotropy Measurements. These meas-
urements were carried out using Fluorolog 3−22 Fluorescence
Spectrophotometer (Jobin Yvon,USA). Lipids 1 and 2 were
prepared from chloroform stocks by the appropriate addition of
the fluorescent probe DPH (1,6-diphenyl-1,3,5-hexatriene) at
300:1 mol ratio of Lipid/DPH, to obtain 2 mM stock of the
lipid. This mixture was then dried under nitrogen gas. The
residual chloroform was then removed in a vacuum for 4 h.
Lipid films were then hydrated in 1 mL of buffer (Tris. HCl,
pH 7.4, 100 mM) overnight, vortexed, and sonicated prior
to the scans between 15 and 90 °C with 2.5 °C increments.
Fluorescence anisotropy was measured by recording the DPH
fluorescence values (excitation at 354 nm and emission at
427 nm) in parallel and perpendicular polarizer positions.
Anisotropy values were calculated by instrument software.
Membrane Fusion. Membrane fusion between cationic
Toxicity Assay. Cytotoxicity of the Lipids 1 and 2 were
assessed in CHO cells using 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT). Complexes prepared
with lipid, were evaluated for their toxicity to cells by per-
forming the MTT assay as described27 in 96-well plates at
charge ratios used in the transfection experiments. Cells were
incubated with transfection complexes for 3 h at 37 °C in
serum-free DMEM. Soon after transfection, cells were washed
with PBS and replaced with 100 μL DMEM+25 μL MTT
(2 mg/mL in PBS) for 3 h. The medium was removed and
replaced with 100 μL DMSO/methanol 1:1 (v/v) and mixed to
dissolve the formazan crystals. Absorbance was then measured
at 540 nm with untreated cells serving as controls. Results were
expressed as percent viability. [A540 (treated cells) −
background/A540 (untreated cells) − background] × 100.
Small Angle X-ray Diffraction of the Lipid
Vesicles. Vesicles of Lipids 1 and 2 with cholesterol were
prepared as described above. The lipid concentration for SAXS
studies was 5 mM and vesicles were used without sonication.
Diffraction data were collected on a S3-MICROpix attached to
a Dectris 100K Platus detector (Hecus Xray Systems, Graz,
Austria) located 300 mm from the sample. Detector was
precalibrated with silver behenate. Samples were filled in 1 mm
thin-walled quartz-glass capillaries in tight thermal contact with
a programmable Peltier unit. Samples were equilibrated for 10
min at each temperature before measurement. The data were
fitted using Fit2d software.28
lipid vesicles and anionic vesicles was monitored by Forster
̈
resonance energy transfer (FRET) between NBD-PE and
Rhodamine DHPE. The liposomes containing the fluorophores
were prepared with the following composition: DOPC:DO-
PE:egg PG:NBD-PE:Rh-PE at a mole ratio of 75:20:5:0.8:0.8.
Cationic liposomes were prepared with Lipid 1 and 2 along
with cholesterol in a mole ratio of 50:50. The fusion was
initiated by mixing 50 nmol of cationic lipid vesicles with 10
nmol of fluorescent lipids in phosphate buffered saline (10 mM
sodium phosphate buffer, pH 7.4, with 150 mM of sodium
chloride) at 25 °C in stirring conditions. The increase in
fluorescence was monitored in a Hitachi F4300 spectro-
fluorimeter by exciting the sample at 460 nm and collecting the
emission at 530 nm. The total fluorescence was measured by
adding Triton X-100 at a final concentration of 0.15%. The
small but measurable quenching due to Triton X-100 was
corrected. Anionic vesicles were also prepared using phopha-
tidic acid instead of PG at the same mole ratio, and we obtained
similar results.
RESULTS
■
Surface Area−Pressure Isotherms. The measurements
were made on a Langmuir trough manufactured by Nima
(model 622D) using high-purity water as subphase. Lipids 1
and 2 were prepared as stocks in chloroform and spread at the
air−water interface in less than 5 μL. The pressure−area iso-
therms were obtained at compression rates of 50 cm2 per
minute. Since the surface area available with the trough does
not allow us to monitor the collapse point of the monolayer,
only the rate of change of the surface pressure for change in
molecular area was used in the study.
Transmission Electron Microscopy. SUVs of Lipids 1
and 2 (2 mM) were prepared as described earlier. The vesicle
solution was cast on the poly(vinyl acetate)-coated copper grid
and stained with 1.0 wt % uranyl acetate aqueous solution. The
TEM images were recorded by JEM-2000EX/FX (JEOL, Co.
Ltd.) with 80 kV accelerating voltage.
Live Uptake of DNA by Confocal Microscopy. Cham-
bered cover glass was plated with 20 000 cells (CHO) and
incubated with lipoplexes (with 6:1 lipid/DNA charge ratio)
prepared with Lipids 1 and 2 in 300 μL of DMEM containing
Oregon Green-labeled plasmid pCMVβ-gal plasmid DNA
prepared as described. After the incubation period of
3 h, DMEM was removed, and the cells were washed with
PBS twice before obtaining images (Leica TCS S52). The
cells containing labeled DNA (ChromaTide Oregon Green
Synthesis of the Lipids. To investigate the role of the
amide and ester linker in a cationic lipid on transfection
properties, 2C12Glu and 2C12-Gln were synthesized and later
coupled to Z-Lys (Z)-Lys(Z)-OH. The schemes for the
synthesis of Lipid 1 (amide lipid) and Lipid 2 (ester lipid) is
given in Scheme 1A,B. The purity was assessed by HPLC and
chemical nature of the compounds was confirmed by FTIR,
1H-NMR, and elemental analysis.
DNA Binding Properties of Lipid 1 and 2. Agarose
Gel Retardation Assay. We initially investigated the DNA-
binding ability of Lipid 1 and 2 by a gel-based assay. Both lipids
formulated with cholesterol bind to plasmid DNA, pEGFPN3
(Figure 1A). Formation of a complex and its retardation was
observed with Lipids 1 and 2 with cholesterol as the colipid at
0.3:1 charge ratio. At charge ratios greater than 2:1, the com-
plexes were either retained in the well or not amenable to EtBr
binding after electrophoresis. The binding of lipids with the
plasmid was found to be ratio-dependent. The binding of Lipid
1 was more avid compared to Lipid 2 when formulated with
DOPE (data not shown). Therefore, only formulations with
cholesterol were taken up for further biophysical investigations
and cell biological activity in vitro.
Fluorescence Quenching. Binding strengths of cationic
lipids with DNA are important to correlate the transfection
properties of lipids to its chemical structure. Exclusion of EtBr
2247
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
Scheme 1. A. Synthesis Scheme and Structure of Glutamide Lipid 1 (n = 12) with the Amide Linker. B. Synthesis Scheme and
Structure of Lipid 2 (n = 12) with Ester Linker
from DNA upon the addition of cationic lipid, and consequent
enhancement of fluorescence, provides a sensitive and
quantitative measurement for DNA binding. Figure 1B shows
the DNA binding pattern of Lipid 1 in comparison to Lipid 2,
which was similar at lower charge ratios. However, at charge
ratios above 1:1, Lipid 1 shows a slightly stronger binding with
∼25% drop in the fluorescence intensity when compared to
∼17% decrease with Lipid 2.
Uptake of Lipoplexes in CHO Cells. We examined the
cellular uptake of lipoplexes prepared with Lipid 1 and Lipid 2
formulated with cholesterol in CHO cells. An Oregon-Green
labeled plasmid probe (green) was mixed with an equal
proportion of unlabeled DNA to prepare lipoplexes labeled with
0.5 mol % of fluorescent lipid (RhDHPE), at the charge ratios
indicated and added to CHO cells followed by incubation at 37
°C for 3 h. After thorough washes with DMEM, live cells were
viewed by confocal microscopy. The upper panel, i.e., Lipid 1
(Figure 2), exhibited particulate bright red fluorescence originating
from the Rhodamine-labeled lipid that colocalized with the
labeled-plasmid DNA. Uptake of Lipid 2 was comparatively less
(see merged panel of this figure). Quantitative analysis of the
confocal image was done by measuring mean fluorescence
intensity (MFI) per cell using LAS AF software. When compared
with Lipid 2, we observed a 3-fold increase in the MFI measured
either by OG-labeled DNA or Rhodamine-labeled liposomes in
the uptake of lipoplexes prepared with Lipid 1/Chol.
Biophysical Characterization of the Lipids. Determi-
nation of the Melting Temperature of the Lipids. Hydrated
lipid samples with and without the colipid, cholesterol, were
subjected to thermal transitions in a differential scanning
calorimeter (DSC). The heat capacity changes in the lipid
phase as a function of temperature were plotted in Figure 3.
The transition midpoints were found to be 65.4 and 23.5 °C for
Lipid 1 and 2, respectively. Similar transition midpoints was
also found for these lipids in the presence of cholesterol. The
transition midpoint for lipids is known to be determined by the
chain length of the alkyl group. Since both lipids have identical
chain length (C12) and identical head groups, the higher
2248
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
Figure 3. Differential scanning calorimetric scans of lipid vesicles:
Thermograms of Lipid 1 and Lipid 2 formulated with equimolar
amounts of cholesterol. Vesicles were prepared as described in
methods.
anisotropy has been extensively used to estimate the phase
transition temperatures of lipid assemblies. Lipid 1 exhibited a
sharp phase transition around 67 °C, whereas Lipid 2 had a
shallow transition at lower temperatures (Supporting Informa-
tion Figure 1S). Thus, a higher melting temperature of Lipid 1
was confirmed by two independent methods.
Figure 1. Binding of DNA with Lipids 1 and 2 in the presence of the
colipid cholesterol: (A) 1% Agarose gel electrophoresis of plasmid
pCMVβ-gal complexed with Lipids 1 and 2 at charge ratios varying
from 0.3:1 to 3:1 Lipid/DNA, indicated above each well. Samples were
electrophoresed using TAE buffer (pH 8.0) and visualized post-
electrophoresis by staining with EtBr. Control in each panel denoted
as “C” pertains to naked plasmid DNA. (B) Fluorescence quenching
by EtBr exclusion assay of Lipids 1 and 2 formulated with colipid
cholesterol. Fluorescence titration curves depict quenching of
fluorescence due to release of EtBr from pCMVβ-gal plasmid DNA
upon lipid binding in a buffer containing 20 mM Tris·HCl (pH 7.4).
Transmission Electron Microscopy (TEM). Having deter-
mined the melting temperature of liposomes of Lipids 1 and 2,
we sought to examine the structural features of these liposomes
by transmission electron microscopy (TEM). The images
obtained with Lipid 1 vesicles prepared with cholesterol or
DOPE as the colipid at charge ratio 3:1 revealed the formation
of very well-defined spherical structures with a clear inside−
outside demarcation. The average size of the vesicles was
between 100 and 200 nm in diameter. In contrast, liposomes
prepared with Lipid 2 with either of the colipids resulted in
supramolecular assemblies that were fragmented and poorly
defined (Figure 4). The liposomal preparation with Lipid 2
resulted in structures with long continuous layers that were not
discrete. Liposomes of Lipid 1 had unaltered morphology when
the particles were prepared either in water or in the presence of
PBS (ionic strength = 0.154) (data not shown).
To confirm our observations further, we subjected both the
liposomal samples to small angle X-ray scattering (SAXS).
SAXS provides information on characteristics of long-range-
ordered structures and has been extensively used in identifying
the phase properties of the lipids in membranes. The SAXS
pattern of Lipids 1 and 2 as shown in Figure 5 indicated defined
scatter peaks for Lipid 1 with ordered assembly, and the
simulation fitted the scatter to ordered lamellar structures. The
scatter profile of liposomes prepared with Lipid 2 resulted in a
broad scatter peak indicative of an amorphous material. The
peak distances in the scattergram were used to calculate the dl
(thickness of the membrane). We obtained 3.58 nm for the
thickness of the Lipid 1. We obtained similar distance for Lipid
2 though the peak peak intensities were significantly smaller in
case of this lipid. The thickness saturated phosphatidylcholine
of 12-carbon chains gives a thickness of 3.0 nm, which is
comparable with Lipid 1. The results of SAXS confirm the
absence of organized structure in the liposomes of Lipid 2.
Compressibility of Lipid Monolayers. With a view to inves-
tigate the higher melting temperature of Lipid 1, we spread
both lipids on water subphase and monitored the changes in
surface pressure as a function of area occupied by the molecule
Figure 2. Uptake of lipoplexes and analysis by confocal microscopy:
Transfection complexes with the corresponding lipid were added to
cells in the absence of serum and incubated for 3 h. Lipid 1/Chol at
6:1 DNA/Lipid (top) and Lipid 2/Chol at 6:1 DNA/Lipid (bottom).
DNA was labeled with Oregon-Green while liposomes were labeled
with Rhodamine DHPE at 0.5 mol %. The merged panel indicate
colocalization of the fluorophores. Cell nuclei were stained with Hoechst
(blue). Panels on the extreme right depict images merged with the
transmission channel. Five optical sections in the middle region, each of
0.5 μm thickness, were combined to generate the images.
melting temperature of Lipid 1 may have been due to additional
interactions in Lipid 1. To confirm the phase transitions, we
used identical preparations of Lipid 1 and 2 with a fluorescent
probe diphenyl hexatriene (DPH) at 300:1 ratio (lipid/DPH,
mol/mol). Anisotropy of DPH is sensitive to the microviscosity
of its location. DPH resides at the center of the bilayer and its
2249
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
Figure 6. Surface pressure isotherms of Lipids 1 and 2 were generated
in the presence of high-purity water as the subphase. Isotherms are
representative of four independent measurements with very good
reproducibility between the isotherms.
cellular membrane is critical for release of DNA. We monitored
the fusion between cationic liposomes of both lipids along with
the anionic vesicles prepared with a composition similar to the
plasma membrane. The fusion kinetics was monitored using
FRET between the donor, NBD-PE, and the acceptor, Rh-PE.
Lipid 2 as seen in Figure 7 showed rapid increase in fluorescence
Figure 4. Transmission electron micrographs of Lipids 1 and 2: Lipids
formulated with DOPE or cholesterol were complexed with plasmid
pCMVβ-gal DNA in 1× PBS buffer at Lipid/DNA 6:1 charge ratio.
Lipoplexes were stained with uranyl acetate as described.
Figure 7. Membrane fusion kinetics: Membrane fusion between
anionic vesicles with Lipid 1 or Lipid 2 was monitored by fluorescence
resonance energy transfer between NBD-PE and Rhodamine PE, as
described.
Figure 5. Small-angle X-ray diffraction of lipid vesicles: Vesicles of
Lipids 1 and 2 formulated with cholesterol were prepared. The lipid
concentration was kept at 5 mM and used without sonication.
Diffraction data were collected and fitted using Fit2d software.
signal and reached a value representative of complete mixing, i.e.,
in the presence of Triton X-100, within 1 min. Many variations
employed in the experimental design such as the ratio of labeled
and unlabeled vesicle and composition of the labeled vesicles did
not change the kinetics. Since Lipid 2 has poorly defined vesicles
as seen in TEM, it is possible that these vesicles were unstable,
hence show very rapid increase in fluorescence that was not
typical of a fusion process. On the other hand, Lipid 1 shows
comparatively slower kinetics of fusion with labeled vesicles. The
fusion kinetics was similar when the labeled vesicles were prepared
with phosphatidic acid instead of phosphatidylserine (data not
reported).
(Figure 6). Surface pressure vs area isotherms of monolayers of
lipids provides information on the surface areas and
intermolecular interactions. The slow compressibility of Lipid
1 compared to that of Lipid 2 suggested that the Lipid 1
monolayer experiences additional resistance to compression
compared to Lipid 2. This resistance may possibly be attributed
to additional interactions between molecules.
Fusion Properties with Anionic Liposomes. Fusion of the
membrane components of the lipoplex membrane with the
2250
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
Figure 8. Transient transfection in vitro: (a) Transfection of CHO cells with Lipid 1 formulated with equimolar amounts of the colipid,
cholesterol, at charge ratios varying from 2:1 to 9:1 Lipid/DNA in the absence (light gray) and presence of serum (dark gray) and normalized
to milligrams protein. Graph depicts reporter gene activity at various charge ratios. Standards such as Lipofectamine (Lfamine) and DOTAP at
2:1 and 3:1 charge ratio, respectively, were included for comparison. Lipid 2 did not show any detectable reporter gene activity on CHO cells
and was not plotted. (b) Transfection in HepG2 cells with Lipid 1/Chol at the indicated charge ratios in the absence (light gray) and presence
of serum (dark gray). (c) Transfection in HepG2 cells with Lipid 2/Chol at the indicated charge ratio in the absence (light gray) and presence
of serum (dark gray).
Biological Activity and Stability in Serum. Transient
Transfection in CHO and HepG2 Cells. The lipids initially
characterized for their DNA-binding properties were then
tested for transfection efficiency in four cell lines, namely,
CHO, HepG2, MCF-7, and HeLa. Both the lipids were
formulated with DOPE or cholesterol as colipid and tested for
transfection efficiency. Formulations with DOPE as colipid
were consistently lower than formulations with cholesterol as
colipid (data not shown), when tested in CHO cells. Hence, in
all further studies only cholesterol-containing liposomes were
used. Transient transfections of the two lipids at varying charge
ratios from 2:1 to 9:1 Lipid/DNA, using pCMVβ-gal as the
reporter gene, were performed as described in CHO cells
(Figure 8a) and HepG2 cells (Figure 8b and c). When com-
pared with the cationic lipid standards, i.e., DOTAP/DOPE
and polycationic lipid, Lipofectamine, we observed slightly
better transfection than DOTAP, and the efficiency was 50% of
the transfection observed with Lipofectamine. We also tested
the 14- and 16-carbon analogues of Lipid 1, which is 12-carbon,
and found that Lipid 1 has shown higher transfection than the
14- and 16-carbon analogues, >3-fold, Supporting Information
Figure 2S. As in the case of HepG2, Lipid 1 (Figure 8b)
exhibited much higher activity than Lipid 2 (Figure 8c). Lipid 2
was inefficient in all cell lines tested. The biological activity of
Lipid 1 with the amide linkage was consistently higher than that
of Lipid 2 that is ester-based in both HeLa cells Supporting
Information Figure 3S and MCF-7 (data not reported). In all
cell lines, maximal activity was observed at 3:1 Lipid/DNA with
Lipid 1 formulated with cholesterol.
Although lipids bind DNA by charge−charge interaction,
these may not protect DNA sufficiently from degradation by
nucleases. Stability in serum is an important property when
developing protocols for in vivo gene transfer. To assess the
serum stability of Lipid 1, serum was included in the medium,
to a final concentration of 10%, at the time of incubation of
cells with the complexes. As evident from Figure 8a−c, com-
pared to the controls, i.e., in the absence of serum, Lipid 1
retained approximately 25% of its transfection efficiency in the
presence of serum at 10% final concentration in CHO and 40−
57% in the case of HepG2 cells at all charge ratios. Transfection
decreased further in the presence of serum concentrations
greater than 10%. Lipid 2 has observable but low transfection in
HepG2 cells at 3:1 charge ratio and good stability in the
presence of serum (Figure 8c). Overall, the transfection efficiency
in the presence and absence of serum has a similar bell-shaped
curve as a function of charge ratio.
Toxicity. Toxicity is an important issue to be dealt with while
generating formulations useful for gene delivery protocols. We
performed MTT-based cell viability assays at various charge
ratios, i.e., 1:1 to 9:1 in CHO cells. Percent cell viabilities upon
treatment with Lipid 1 at charge ratios 3:1 is ∼70% (Figure 9).
Figure 9. Determination of cell viability in CHO cells. Lipid 1 and 2:
DNA:Lipid complexes of Lipids 1 and 2 formulated with cholesterol
was evaluated for cytotoxicity at the indicated charge ratios varying
from 1:1 to 9:1 N/P. Following 3 h incubation, the MTT assay was
performed. Percent cell viability represents experiments performed in
triplicate as described in Materials and Methods. The absorption
values obtained using reduced formazan, in the absence of lipids, were
taken to be 100.
In contrast, Lipid 2 is inefficient at this charge ratio, although
nontoxic. The utility of Lipid 1/cholesterol offsets the observed
toxicity making it effective in vitro.
DISCUSSION
■
Two glutamic acid-based lipids that have identical head groups
and acyl chains were synthesized. These lipids differ from each
other with respect to the chemistry of the linker connecting
the acyl chains to glutamide. Lipid 1 with the amide linker
exhibited high transfection efficiency in four different cell lines,
2251
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
when compared to Lipid 2 with an ester linker. Among the
popular colipids such as DOPE and cholesterol, both lipids
exhibited higher transfection efficiency with cholesterol
compared to DOPE. The binding of lipids to DNA in EtBr
exclusion assay showed identical profiles, however, at charge
ratios >1:1, Lipid 1 shows stronger displacement of EtBr.
Similar charge ratio dependency was also observed in gel
retardation assay with these lipids. Electron microscopy images
of the lipoplexes prepared using Lipids 1 and 2 along with the
colipids cholesterol and DOPE showed clearly that Lipid 1
lipoplexes have a defined morphology, whereas Lipid 2
aggregates were irregular and amorphous. Lack of defined
liposomal structure of Lipid 2 vesicles was also confirmed by
broad scattering profile in SAXS. It is apparent that the stronger
binding and formation of defined stable complexes are
important in the interaction of Lipid 1 with cells and eventually
in their uptake. The higher uptake of the lipoplexes with Lipid
1, as seen from confocal microscopy may consequently be
responsible for observed higher transfection efficiency. Lipid 1
has also demonstrated efficient transfection in the presence of
serum when tested in both CHO and HepG2 cells. The
benefits of serum in transfection protocols have been
reported.29,30 Stability in serum suggests the potential of
these lipids for DNA delivery in vivo.
the higher melting temperature of Lipid 1 does not originate
from the hydrophobic portion of the molecule. The addition of
the slow compression rate of Lipid 1 monolayers compared to
Lipid 2 also points to the fact that Lipid 1 monolayers experience
additional intermolecular interactions. Since the head groups of
both lipids are identical, the altered isotherm may originate
intermolecular interactions. The amine linkers of Lipid 1 possibly
form a hydrogen bond with carbonyl group of the adjacent lipid
in a liposome, thus providing additional stability necessary for a
stable lipoplex (Figure 10). In our opinion, this is the first report
Biophysical Basis for Higher Transfection Efficiency
with Lipid 1. DNA binding being a prerequisite for achieving
good transfection efficiency, it must be condensed prior to cell
uptake. Cationic lipids bind electrostatically to the DNA.3,30
This process results in condensation and protection of the
DNA and leads to the formation of complexes that are
particulate in nature.3,31 DNA present in a particle is more
efficiently taken up by the cell and endocytosed than free DNA.
Characterization of lipoplexes at varying charge ratios revealed
that transfection efficiency sharply increases with DNA
condensation, change in zeta potential from negative to positive
value, and observations of ordered lipoplexes in electron
micrographs.32 Although DNA was found to be condensed
below optimal charge ratios, DNA was accessible to nucleases
leading to decrease in transfection. Stability of the lipoplex is
critical for protecting the DNA during transfection and also
enhancing the uptake of the lipoplex by the cells.32 Though
binding of Lipid 2 with the DNA was comparable to Lipid 1,
the packing of the DNA into lipoplexes may be inefficient
leading to particles that may be large and variable in size. Thus,
Lipid 2, a 12-carbon lipid, does not provide sufficient stability
for the formation of liposomes; hence, its uptake by cells and
subsequent transfection is poor. Though Lipid 1 has same chain
length and headgroup as Lipid 2, its melting temperature was
nearly 40 °C higher. Melting temperature of lipids depends
strongly on the length of the acyl chains and increases with the
increase in chain lengths. Since Lipids 1 and 2 have identical
chain lengths, the higher melting temperature may be due to
additional interactions between the molecules. The stabilizing
effect of the intermolecular hydrogen bond was observed
between the headgroups in phosphatidylethanolamines (PE),
which increases the melting temperature of PEs by nearly 20 °C
compared to phosphatidylcholines of similar alkyl chain
length.33 Such intermolecular hydrogen bonding was observed
in Gemini quaternary ammonium surfactants that contain an
amide linker but not in ester-containing surfactants.34 In
addition, the data on polarization of DPH, an indicator of
microviscosity of the membrane, suggests that both lipids possess
similar microviscosity values, corroborating the observations that
Figure 10. Intermolecular interactions of Lipid 1. Model depicting
stabilization of Lipid 1 through intermolecular interactions generated
through hydrogen bonding.
of an approach to increase the stability and transfection of
cationic liposomes by designing features favoring intermolecular
hydrogen bonding.
Unsaturated lipids, predominantly oleic acid, containing
cationic lipid formulations showed higher transfection efficiency
than formulations with saturated lipids. Investigation of
transfection efficiency with a given cationic lipid (DOTAP)
along with phosphatidylethanolamine lipids with various acyl
chains with different lengths indicated that unsaturated PEs are
essential compared to saturated PEs for efficient transfection.35
Hydrophobic moiety of cationic lipids strongly modulates their
transfection activity. A study on the chain length variants of
DOTAP suggests that DLTAP (N-[1-(2,3-dilauroyloxy)-
propyl]-N,N,N-trimethylammonium methylsulfate) was least
efficient in transfection.36 In this context, the observation of
higher transfection efficiency of Lipid 1, despite having higher
melting temperature, is interesting. It is observed that higher
melting points of long acyl chain containing lipids do not mix
with other membranes easily. Fusion between vesicles occurs
readily above the phase transition temperature of the lipids.
Below the phase transition temperature, the lipids are in
crystalline phase and have restricted fluidity, essential for
mixing of the membrane contents.37 Within a cell, transfection
processes involve membrane−membrane interactions between
lipoplex and endosomal membrane and fusion.38 Fusion is
essential for dissociation of the cationic lipids from the DNA
and release the DNA into the cytoplasm. Though Lipid 1 has
higher melting temperature, its microviscocity as measured by
DPH, is still in the range of 12-carbon lipids.
2252
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
Hence, the miscibility of Lipid 1 may not restrict fusion with
other membranes. Amide linkers have been used in the design
of cationic lipids and were found to be biodegradable and
stable.39 DOGS is a popular amide linker containing cationic
lipid, where the amide linker is between the backbone and the
headgroup.40 A series of cationic lipids with amide and ester
linkers to a pyridinium headgroup were synthesized. Amide
linker containing lipids were found to be superior in transfection
compared to the ester containing pyridinium lipids.41 However,
the amide containing lipids melt at slightly lower temperatures
than the ester lipids. In another study, a series of detergents were
synthesized with amide or ester linkers to amino acid head
groups. The amide detergents bound to the oligonucleotides
more efficiently than the ester containing detergents.42 In Lipid
1, the amide linker provides stability to the liposome and the
lipoplex by forming an intermolecular hydrogen bond which
did not seem to influence the membrane interactions crucial for
transfection. Incorporating intermolecular interactions, such as
hydrogen or ionic bonds, in cationic lipids offers yet another
strategy to design efficient cationic lipids.
REFERENCES
■
(1) Chesnoy, S., and Huang, L. (2000) Structure and function of
lipid-DNA complexes for gene delivery. Annu. Rev. Biophys. Biomol.
Struct. 29, 27−47.
(2) Li, W., and Szoka, F. C. Jr. (2007) Lipid-based nanoparticles for
nucleic acid delivery. Pharm. Res. 24, 438−49.
(3) Rao, N. M. (2010) Cationic lipid-mediated nucleic acid delivery:
beyond being cationic. Chem. Phy. Lipids 163, 245−52.
(4) Nicolazzi, C., Garinot, M., Mignet, N., Scherman, D., and
Bessodes, M. (2003) Cationic lipids for transfection. Curr. Med. Chem.
10, 1263−77.
(5) Audouy, S., and Hoekstra, D. (2001) Cationic lipid-mediated
transfection in vitro and in vivo. Mol. Membr. Biol. 18, 129−43.
(6) Koynova, R., Wang., L., Tarahovsky, Y., and MacDonald, R. C.
(2005) Lipid phase control of DNA delivery. Bioconjugate Chem. 16,
1335−9.
(7) Niculescu-Duvaz, D., Heyes, J., and Springer, C. J. (2003)
Structure-activity relationship in cationic lipid mediated gene trans-
fection. Curr. Med. Chem. 10, 1233−61.
(8) Barteau, B., Chevre, R., Letrou-Bonneval, E., Labas, R., Lambert,
̀
O., and Pitard, B. (2008) Physicochemical parameters of non-viral
vectors that govern transfection efficiency. Curr. Gene Ther. 8, 313−23.
̀
(9) Montier, T., Benvegnu, T., Jaffres, P. A., Yaouanc, J. J., and Lehn,
CONCLUSIONS
■
P. (2008) Progress in cationic lipid-mediated gene transfection: a
series of bio-inspired lipids as an example. Curr. Gene Ther. 8, 296−
312.
Stability of a lipoplex is important for achieving good trans-
fection efficiencies. This is dependent on the nature of the
cationic headgroup and DNA interactions and the hydrophobic
volume of the lipid. The overview of lipid/DNA interactions
and lipid properties of the two lipids, different in their linker
chemistry, suggests that stable formation of liposomes with
Lipid 1 was critical for superior transfection properties com-
pared to Lipid 2. Formulations made with Lipid 1 were stable
in serum, thereby implying its utility for in vivo experimenta-
tion. Since the two lipids are identical in all aspects except for
the linker, the stability of amide linker containing Lipid 1 is
due to the presence of intermolecular hydrogen bonds leading
to stable vesicles. This observation is supported by higher
transition temperatures, TEM, SAXS, and monolayer studies.
Incorporation of intermolecular interactions may be yet another
way to increase the stability of the liposomes and thereby
transfection.
(10) Tranchant, I., Thompson, B., Nicolazzi, C., Mignet, N., and
Scherman, D. (2004) Physicochemical optimization of plasmid
delivery by cationic lipids. J. Gene Med. 6 (Suppl 1), S24−35.
(11) Gao, X., and Huang, L. (1991) A novel cationic liposome
reagent for efficient transfection of mammalian cells. Biochem. Biophys.
Res. Commun. 179, 280−5.
(12) Ren, T., Zhang, G., Liu, F., and Liu., D. (2000) Synthesis and
evaluation of vitamin D-based cationic lipids for gene delivery in vitro.
Bioorg. Med. Chem. Lett. 10, 891−4.
(13) Fichert, T., Regelin, A., and Massing, U. (2000) Synthesis and
transfection properties of novel non-toxic monocationic lipids:
Variation of lipid anchor, spacer and head group structure. Bioorg.
Med. Chem. Lett. 10, 787−91.
(14) Floch, V., Loisel, S., Guenin, E., Herve,
Yaouanc, J. J., des Abbayes, H., and Ferec, C. (2000) Cation
́
A. C., Clement, J. C.,
́
substitution in cationic phosphonolipids: a new concept to improve
transfection activity and decrease cellular toxicity. J. Med. Chem. 43,
4617−28.
ASSOCIATED CONTENT
* Supporting Information
(15) Heyes, J. A., Niculescu-Duvaz, D., Cooper, R. G., and Springer,
C. J. (2002) Synthesis of novel cationic lipids: effect of structural
modification on the efficiency of gene transfer. J. Med. Chem. 45,
99−114.
■
S
[Graphs depict thermal transitions of Lipids 1 and 2 using
DPH, transfection efficiencies of Lipid analogues in comparison
to Lipid 1 in CHO cells and transfection efficiency of Lipid 1
and 2 in HeLa cells. This material is available free of charge via
the Internet at http://pubs.acs.org.
(16) Prasad, T, K., Rangaraj, N., and Rao, N, M. (2005) Quantitative
aspects of endocytic activity in lipid-mediated transfections. FEBS Lett.
579, 2635−42.
(17) Hui, S. W., Langner, M., Zhao, Y. L., Ross, P., Hurley, E., and
Chan, K. (1996) The role of helper lipids in cationic liposome-
mediated gene transfer. Biophys. J. 71, 590−9.
AUTHOR INFORMATION
Corresponding Author
*WP +91-40-27192514, Fax # +91-40-27160311, www.ccmb.
res.in, E-mail: madhu@ccmb.res.in.
■
(18) Simoes, S., Slepushkin, V., Gaspar, R., de Lima, M. C., and
̃
Duzgunes,̧ N. (1998) Gene delivery by negatively charged ternary
̈
̈
complexes of DNA, cationic liposomes and transferrin or fusigenic
peptides. Gene Ther. 5, 955−64.
(19) Zuhorn, I. S., and Hoekstra, D. (2002) On the mechanism of
cationic amphiphile-mediated transfection. To fuse or not to fuse: is
that the question? J. Membr. Biol. 189, 167−79.
ACKNOWLEDGMENTS
■
This work was supported by JSPS-DST Bilateral Joint
Project (# DST/INT/JSPS/Proj-38/2007) and a CSIR Network
project (NWP0035). Advice of Drs. R. Nagaraj and B. Raman is
gratefully acknowledged. We thank Nandini Rangaraj and Dr.
Shoeb Ahmad for help with confocal microscopy and fluorescence
anisotropy measurements, respectively. We thank Prasada Raju of
Indian Institute of Chemical Technology, Hyderabad, for HR-ESI-
MS analysis.
(20) Kumar, V. V., Pichon, C., Refregiers, M., Guerin, B., Midoux, P.,
and Chaudhuri, A. (2003) Single histidine residue in head-group
region is sufficient to impart remarkable gene transfection properties
to cationic lipids: evidence for histidine-mediated membrane fusion at
acidic pH. Gene Ther. 10, 1206−15.
(21) Gopal, V., Prasad, T. K., Rao, N. M., Takafuji, M., Rahman, M. M.,
and Ihara, H. (2006) Synthesis and in vitro evaluation of glutamide-
containing cationic lipids for gene delivery. Bioconjugate Chem. 17, 1530−6.
2253
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254

Bioconjugate Chemistry
Article
(22) Banerjee, R., Das, P. K., Srilakshmi, G. V., Chaudhuri, A., and
Rao, N. M. (1999) Novel series of non-glycerol-based cationic
transfection lipids for use in liposomal gene delivery. J. Med. Chem. 42,
4292−9.
(23) Kumar, V. V., Singh, R. S., and Chaudhuri, A. (2003) Cationic
transfection lipids in gene therapy: successes, set-backs, challenges and
promises. Curr. Med. Chem. 10, 1297−306.
(42) Santhiya, D., Dias, R. S., Shome, A., Das, P. K., Miguel, M. G.,
Lindman, B, and Maiti, S. (2009) Role of linker groups between
hydrophilic and hydrophobic moieties of cationic surfactants on
oligonucleotide-surfactant interactions. Langmuir 25, 13770−13775.
(24) Rajesh, M., Sen, J., Srujan, M., Mukherjee, K., Sreedhar, B., and
Chaudhuri, A. (2007) Dramatic influence of the orientation of linker
between hydrophilic and hydrophobic lipid moiety in liposomal gene
delivery. J. Am. Chem. Soc. 129, 11408−20.
(25) Ghosh, Y. K, Visweswariah, S. S., and Bhattacharya, S. (2002)
Advantage of the ether linkage between the positive charge and the
cholesteryl skeleton in cholesterol-based amphiphiles as vectors for
gene delivery. Bioconjugate Chem. 13, 378−84.
(26) Legendre, J. Y., Trzeciak, A., Bohrmann, B., Deuschle, U., Kitas,
E., and Supersaxo, A. (1997) Dioleoylmelittin as a novel serum-
insensitive reagent for efficient transfection of mammalian cells.
Bioconjugate Chem. 8, 57−63.
(27) Xavier, J., Singh, S, Dean, D. A., Rao, N. M., and Gopal, V.
(2009) Designed multi-domain protein as a carrier of nucleic acids
into cells. J. Controlled Release 133, 154−60.
(28) Hammersley, A. P., Svensson, S. O., Hanfland, M., Fitch, A. N.,
and Hausermann, D. (1996) Two-dimensional detector software: from
̈
real detector to idealised image or two-theta scan. High Pressure
Research 14, 235−248.
(29) Simoes, S., Slepushkin, V., Pires, P., Gaspar, R., Pedroso de
Lima, M. C., and Duzgunes, N. (2000) Human serum albumin
enhances DNA transfection by lipoplexes and confers resistance to
inhibition by serum. Biochim. Biophys. Acta 1463, 459−469.
(30) Tros de Ilarduya, C., Arangoa, M. A., and Duzgunes, N. (2003)
Transferrin-lipoplexes with protamine condensed DNA for serum-
resistant gene delivery. Methods Enzymol. 373, 342−356.
(31) Rao, N. M., and Gopal, V. (2006) Cell biological and
biophysical aspects of lipid-mediated gene delivery. Biosci. Rep. 26,
301−24.
(32) Miller, A. D. The problem with cationic liposome/micelle-based
non-viral vector systems for gene therapy. Curr. Med. Chem. 10, 1195-
1211.
(33) Blume, A. (1983) Apparent molar heat capacities of
phospholipids in aqueous dispersions: Effects of chain length and
head group structure. Biochemistry 22, 5436−5442.
́
́ ̌ ̌ ́ ́
(34) Pisarcik, M., Polakovicova, M., Pupak, M., Devinsky, F., and
Lacko, I. (2009) Biodegradable gemini surfactants. Correlation of area
per surfactant molecule with surfactant structure. J. Colloid Interface Sci.
329, 153−159.
(35) Koynova, R., Tenchov, B., Wnag, L., and MacDonald, R. C.
(2009) Hydrophobic moiety of cationic lipids strongly modulates their
transfection activity. Mol. Pharmaceutics 6, 951−958.
(36) Regelin, A. E., Fankhaenel, S., Gurtesch, L., Prinz, C., von
̈
Kiedrowski, G., and Massing, U. (2000) Biophysical and lipofection
studies of DOTAP analogs. Biochim. Biophys. Acta 1464, 151−164.
(37) Ma, B., Zhang, S., Jiang, H., Zhao, B., and Lv, H. (2007)
Lipoplex morphologies and their influences on transfection efficiency
in gene delivery. J. Controlled Release 123, 184−94.
(38) Friend, D. S., Papahadjopoulos, D., and Debs, R. J. (1996)
Endocytosis and intracellular processing accompanying transfection
mediated by cationic liposomes. Biochim. Biophys. Acta 1278, 41−50.
(39) Niculescu-Duvaz, D., Heyes, J., and Springer, C. J. (2003)
Structure-activity relationship in cationic lipid mediated gene trans-
fection. Curr. Med. Chem. 10, 1233−61.
(40) Behr, J. P., Demeneix, B., Loeffler, J. P., and Perez-Mutul, J.
(1989) Efficient gene transfer into mammalian primary endocrine cells
with lipopolyamine-coated DNA. Proc. Natl. Acad. Sci. U. S. A. 86,
6982−6.
(41) Zhu, L., Lu, Y., Miller, D. D., and Mahato, R. I. (2008)
Structural and formulation factors influencing pyridinium lipid-based
gene transfer. Bioconjugate Chem. 19, 2499−2512.
2254
dx.doi.org/10.1021/bc2002874|Bioconjugate Chem. 2011, 22, 2244−2254