Design, synthesis, and transfection biology of novel cationic glycolipids for use in liposomal gene delivery

Article abstract of DOI:10.1021/jm000466s

The molecular structure of the cationic lipids used in gene transfection strongly influences their transfection efficiency. High transfection efficiencies of non-glycerol-based simple monocationic transfection lipids with hydroxyethyl headgroups recently reported by us (Banerjee et al. J. Med. Chem. 1999, 42, 4292-4299) are consistent with the earlier observations that the presence of hydroxyl functionalities in the headgroup region of a cationic lipid contributes favorably in liposomal gene delivery. Using simple sugar molecules as the source of multiple hydroxyl functionalities in the headgroup region of the transfection lipids, we have synthesized four novel simple monocationic transfection lipids, namely, 1-deoxy-1-[dihexadecyl(methyl)-ammonio]-D-xylitol (1), 1-deoxy-1-[methyl(ditetradecyl)ammonio]-D-arabinitol (2), 1-deoxy-1-[dihexadecyl(methyl)ammonio]-D-arabinitol (3) and 1-deoxy-1-[methyl(dioctadecyl)ammonio]-D-arabinitol (4), containing hydrophobic aliphatic tails and the hydrophilic arabinosyl or xylose sugar groups linked directly to the positively charged nitrogen atom. Syntheses, chemical characterizations, and the transfection biology of these novel transfection lipids 1-4 are described in this paper. Lipid 1, the xylosyl derivative, showed maximum transfection on COS-1 cells. All the lipids showed transfection with cholesterol as colipid and not with dioleoylphosphatidylethanolamine (DOPE). Radioactive quantitation of free and complexed DNA combined with ethidium bromide exclusion measurements suggest that though nearly 70% of the DNA exists as complexed DNA, the DNA may not have condensed as was observed with other cationic lipids. Presence of additional (more than two) hydroxyl functionalities in the headgroup of the cationic lipids appears to have improved the transfection efficiency and made these lipids less cytotoxic compared to two-hydroxyl derivatives.

Full text of DOI:10.1021/jm000466s

4
176
J . Med. Chem. 2001, 44, 4176-4185
Design , Syn th esis, a n d Tr a n sfection Biology of Novel Ca tion ic Glycolip id s for
Use in Lip osom a l Gen e Deliver y
†
Rajkumar Banerjee, Yenugonda Venkata Mahidhar, and Arabinda Chaudhuri*
Division of Lipid Science and Technology, Indian Institute of Chemical Technology, Hyderabad 500 007, India
Vijaya Gopal and Nalam Madhusudhana Rao*
Centre for Cellular & Molecular Biology, Hyderabad 500 007, India
Received November 1, 2000
The molecular structure of the cationic lipids used in gene transfection strongly influences
their transfection efficiency. High transfection efficiencies of non-glycerol-based simple mono-
cationic transfection lipids with hydroxyethyl headgroups recently reported by us (Banerjee et
al. J . Med. Chem. 1999, 42, 4292-4299) are consistent with the earlier observations that the
presence of hydroxyl functionalities in the headgroup region of a cationic lipid contributes
favorably in liposomal gene delivery. Using simple sugar molecules as the source of multiple
hydroxyl functionalities in the headgroup region of the transfection lipids, we have synthesized
four novel simple monocationic transfection lipids, namely, 1-deoxy-1-[dihexadecyl(methyl)-
ammonio]-D-xylitol (1), 1-deoxy-1-[methyl(ditetradecyl)ammonio]-D-arabinitol (2), 1-deoxy-1-
[dihexadecyl(methyl)ammonio]-D-arabinitol (3) and 1-deoxy-1-[methyl(dioctadecyl)ammonio]-
D-arabinitol (4), containing hydrophobic aliphatic tails and the hydrophilic arabinosyl or xylose
sugar groups linked directly to the positively charged nitrogen atom. Syntheses, chemical
characterizations, and the transfection biology of these novel transfection lipids 1-4 are
described in this paper. Lipid 1, the xylosyl derivative, showed maximum transfection on COS-1
cells. All the lipids showed transfection with cholesterol as colipid and not with dioleoylphos-
phatidylethanolamine (DOPE). Radioactive quantitation of free and complexed DNA combined
with ethidium bromide exclusion measurements suggest that though nearly 70% of the DNA
exists as complexed DNA, the DNA may not have condensed as was observed with other cationic
lipids. Presence of additional (more than two) hydroxyl functionalities in the headgroup of the
cationic lipids appears to have improved the transfection efficiency and made these lipids less
cytotoxic compared to two-hydroxyl derivatives.
In tr od u ction
with the transfection efficiency. The size, surface po-
1
2
13
14
tential, DNase sensitivity, density centrifugation,
Delivering genes into cells or tissues is an extremely
promising therapeutic tool to alleviate morbidity in
clinically challenging genetic diseases and cancers. In
the area of gene therapy, the focus of an intensely
pursued current research is centered around develop-
1
2,15
16
ethidium bromide exclusion,
circular dichroism,
1
3,14
fusogenic ability of the complexes,
electron micros-
1
7
copy, and X-ray diffraction of the lipid-DNA com-
1
8
plexes were investigated to understand the cell bio-
logical basis of transfection. Broadly, these studies made
the following observations: (a) at 1:1 charge ratio of lipid
to DNA, i.e., at charge neutrality, the ú potential of the
complex becomes positive and the particle size increases
abruptly; (b) cationic lipids condense the DNA; (c)
cationic liposomes that have fluid membranes help
transfection; (d) cone-shaped colipids facilitate endoso-
mal escape; and (e) the biology of the cell has strong
bearing on the transfection efficiency.
1
ment of efficient viral and nonviral transfection
2
-10
vectors.
mediated transfection holds considerable promise for its
a) reproducibility and simplicity in preparation, (b)
Among nonviral vectors, cationic liposome-
(
nonimmunogenic nature, and (c) efficiency in forming
stable injectable complexes even with large DNA. A
wide variety of cationic lipids have been synthesized and
9
tested for their transfection efficiencies as ingenious
formulations with success.8 Despite a considerable
Transfection efficiency strongly depends on the cell
type. The factors capable of modulating transfection
efficiencies include (a) adsorption of the lipid-DNA
knowledge base on the transfection biology of cationic
lipids, the critical factors influencing the transfection
efficiency of cationic lipids are still elusive.11
A number of physical properties of the lipid-DNA
complexes were studied in an effort to correlate them
1
1,19
complex to the cell surface;
(b) uptake of the
1
9
complex, mainly by endocytosis; (c) escape of DNA
2
0
from the endosomes; (d) the half-life of DNA in the
2
1
22
cytoplasm; and (e) entry of the DNA into the nucleus.
*
Authors to whom correspondences should be addressed: e-mail
arabinda@iict.ap.nic.in (A.C.) and madhu@ccmb.ap.nic.in (N.M.R.);
fax 91-40-7173757 (A.C.) and 91-40-7171195 (N.M.R.); telephone 91-
The molecular architecture of the cationic lipids is likely
to influence the first three of these factors. Since Felgner
4
0-7173370 (A.C.) and 91-40-7172241 (N.M.R.)
Present address: School of Pharmacy, University of Pittsburgh,
2
3
†
et al. reported their first cationic liposome-mediated
gene delivery in 1987, there has been a considerable
Pittsburgh, PA.
1
0.1021/jm000466s CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/13/2001

Cationic Glycolipids in Liposomal Gene Delivery
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4177
Ch a r t 1
are straightforward. However, given their remarkably
high transfection efficiencies, the overall yields of these
transfection lipids need to be improved in future.
Scheme 1 outlines the general synthetic strategies
adopted for preparing lipid 1. The steps involved were
(a) reacting 2,3,4,5-diisopropylidinexylose thioketal with
mercuric chloride-mercuric oxide to deprotect the
thioketal group selectively to aldehyde, converting the
resulting aldehyde to the intermediate iminium com-
pound by treating with the appropriate secondary
amine, followed by in situ reduction of the resulting
iminium compound with sodium cyanoborohydride to
produce the diisopropylidine derivative of the interme-
diate tertiary amine; (b) treating the diisopropylidine
derivative of the tertiary amine prepared in step a with
methyl iodide; and (c) sequential treatment of the
quaternized ammonium iodide intermediate obtained in
step b with trifluoroacetic acid and chloride ion ex-
changer resin, respectively.
Synthesis of lipids 2-4 (Scheme 2) essentially consists
of similar steps as outlined in Scheme 1 for the synthesis
of lipid 1 except with the use of thioketal- and diiso-
propylidine-protected arabinose as the starting material.
Highly exchangeable protons of the hydroxyl function-
alities in the sugar headgroups caused severe line
increase in design and syntheses of efficient nonviral
cationic transfection vectors.4
,7,24-44
Interestingly, many
of these reported cationic transfection lipids, such as
DOTMA, DMDHP, DMRIE, DOTAP, etc., have a com-
mon element in their molecular structure, namely, the
1
broadening of the peaks in the H NMR spectra of all
8
the final glycolipids. However, the molecular ion peaks
were observed mostly as the base peaks in LSIMS-
HRMS or FABMS as described in the Experimental
Section. Given the overall mild reaction conditions
employed in our method, stereo-integrities of the sugar
skeletons in lipids 1-4 are likely to be preserved
throughout the reaction sequences.
presence of a glycerol backbone. Among these glycerol-
based transfection lipids, the most efficient lipids, such
as DMRIE and DMDHP, contain one or two hydroxy-
ethyl groups directly attached to the positively charged
nitrogen atoms. Recently, we reported, for the first time,
a series of highly efficient non-glycerol-based simple
monocationic transfection lipids in which the most
promising lipid was demonstrated to be DHDEAB,
containing two hydroxyethyl functionalities directly
Tr a n sfection . The transfection efficiencies of cationic
lipids 1-4 were tested in COS-1 cells with pCMV-â-gal
plasmid, carrying the â-galactosidase reporter gene.
Since cationic transfection lipids work efficiently with
colipids, preferably DOPE or cholesterol, we prepared
lipid formulations containing the cationic lipids 1-4
with both of these colipids at 1:1 mole ratios. The
transfection efficiencies of the eight formulations, i.e.,
1-4 with either DOPE or cholesterol, were tested on
COS-1 cell lines along with DHDEAB (Figure 1). All the
four lipids 1-4 showed higher reporter gene activity
with cholesterol as colipid compared to DOPE. Lipids
1-4 without any colipids showed transfection efficiency
equal to or lower than transfection obtained with DOPE
as colipid (Figure 1). The charge ratios between 1 and
3 (lipid/DNA) produced maximal reporter gene activity.
Above 3:1 charge ratio and up to 9:1, the transfection
efficiency was poor as seen by the decrease in â-gal
activity (Figure 1). Lipofectin gave values lower than 5
4
1
attached to the positively charged nitrogen atom. The
promising transfection efficiency of DHDEAB motivated
us to design and synthesize non-glycerol-based cationic
amphiphiles containing more than two hydroxyl func-
tionalities in the headgroup regions and investigate
their transfection efficiencies. Herein, we report the
design, syntheses, and transfection biology of four novel
non-glycerol-based cationic glycolipids, namely, 1-deoxy-
1
1
1
1
-[dihexadecyl(methyl)ammonio]-D-xylitol (1), 1-deoxy-
-[methyl(ditetradecyl)ammonio]-D-arabinitol (2), 1-deoxy-
-[dihexadecyl(methyl)ammonio]-D-arabinitol (3), and
-deoxy-1-[methyl(dioctadecyl)ammonio]-D-arabinitol (4)
containing simple monosaccharide units as the source
of multiple hydroxyl functionalities in the headgroup
region.
Resu lts
-
4
units (10 ) at comparable charge. Lipoplexes with
charge ratios above 3 were poor in transfecting. In our
earlier studies, DHDEAB was compared with Lipofectin
and LipofectAmine and found to be equal or better than
these extensively used commercial formulations.41 In the
present study we have compared the transfection ef-
ficiencies of the four lipids 1-4 with that of DHDEAB.
Lipid 1 gave the best results with cholesterol as colipid.
Lipid 3 containing a palmityl hydrophobic chain was
better among the arabinose headgroup lipids 2-4.
Ch em istr y. The key structural elements common to
all the transfection lipids 1-4 described herein include
(a) the presence of hydrophobic group directly linked to
the positively charged nitrogen atom, (b) the presence
of either an arabinosyl or a xylosyl pentose sugar group
attached covalently to the positively charged nitrogen
atom, and (c) the absence of glycerol backbone in the
molecular architectures of the monocationic glycolipids.
The details of the synthetic procedures for all the novel
transfection lipids shown in Chart 1 are described in
the Experimental Section. As outlined in Schemes 1 and
F or m a tion of DNA-Lip id Com p lexes a n d Sen -
sitivity of th e Com p lexes to DNa se I. The binding
2
, the chemistries involved in preparing these new lipids

4
178 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Banerjee et al.
Sch em e 1. Synthesis of Lipid 1
Sch em e 2. Synthesis of Cationic Glycolipids 2-4
separate and quantitate the free DNA from bound
1
2,14,47
DNA.
Several reports used EtBr staining as a
method of viewing the free DNA, which permits quan-
1
2,17
titation of the free DNA.
To quantitate both bound
and free DNA in electrophoresis, we have used radioac-
tive plasmid DNA. This approach is simple since the
separation and quantitation of free and complexed DNA
does not involve any additional steps and is highly
reproducible. The agarose gel after electrophoresis was
exposed to a Fuji phosphorimager and the radioactive
bands were qunantitated. Nick-translated radioactive
pCMV-â-gal DNA was complexed with lipids 1-4 with
cholesterol as colipid at varying lipid-DNA charge
ratios (agarose gels containing radioactive bands of
complexed and free DNA after electrophoresis are
shown in Figure S1 of the Supporting Information). In
these studies lipid preparations with DOPE as colipid
was not included since the transfection with DOPE was
poor. Free plasmid was well resolved from the positively
charged complexes that were retarded and retained in
the well. The intensity of the band in the well was
maximum at the 3:1 lipid/DNA charge ratio, indicating
the formation of a positively charged complex that does
not migrate in an electrophoretic field. Correspondingly,
the amount of free DNA migrates at the same position
as the free (control) DNA, lane 1, albeit with decreased
band intensity. The radioactive bands in the gel were
quantitated by densitometry and the percentages of free
DNA in lipid-DNA complexes with four lipids (with
cholesterol as the colipid) are shown in Figure 2. The
amount of free DNA was maximum at low charge ratios
and lower at high charge ratios. Densitometric quanti-
tation indicates that at 3:1 charge ratio nearly 70% of
the DNA exists as lipid-DNA complex with all the four
lipids 1-4 and DOTAP, a lipid well-known for its
remarkable DNA-condensing ability (hence used as a
control in Figure 2). The rank order of lipids in terms
of binding strength was 3 ) DOTAP < 4 < 2 < 1 at
of lipids 1-4 to DNA was evaluated by quantitating
both free DNA and DNA bound in the complex. DNA-
lipid complexes were made and resolved by agarose gel
electrophoresis in order to separate the bound and free
DNA. Ethidium bromide (EtBr) staining to visualize the
DNA in a lipid-DNA complex is complicated; hence,
solvent extraction and centrifugation were used to

Cationic Glycolipids in Liposomal Gene Delivery
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4179
F igu r e 1. Transfection efficiencies of lipids 1-4 on COS-1 cells without any colipid (O), with cholesterol as colipid (9), and with
DOPE (b) as colipid. Units of â-galactosidase activity were plotted against the varying lipid to DNA (() charge ratios. The
o-nitrophenol formation (micromoles of o-nitrophenol produced per minute) was converted to units by use of a standard curve
obtained with pure (commercial) â-galactosidase. (A) Lipid 1; (B) lipid 2; (C) lipid 3; (D) lipid 4. The transfection efficiency of
DHDEAB (0) was compared with lipid 1. All the lipids were tested on the same day and the data presented are averages of three
experiments (n ) 3). DNA (0.3 µg) was complexed with lipid at various charge ratios in a volume of 50 µL for 30 min and added
to cells after the medium was diluted to 100 µL. The incubation of lipoplex with cells was allowed for 30 min before addition of
another 100 µL of DMEM medium. The reporter gene activity was assayed after 48 h by lysing the cells with NP40 (0.5%) containing
PBS.
5
0% free DNA. With lipids 2 and 4, maximum trans-
present as a complex with lipid was undigested and
remains in the well. Lipid 4 was least protective among
the four lipids to DNase I treatment. We observed some
undigested free DNA (e.g., lanes 4, 5, 8, and 12 in Figure
3). The reaction of DNase I was stopped with EDTA (30
mM) and then the agarose gel was run. EDTA, besides
inhibiting the DNase I activity, could interact with the
cationic lipids and destabilize the lipid-DNA complex.
The DNA released after interaction of EDTA with the
lipid-DNA complex could be seen as free DNA. The
ability of EDTA to destabilize the complex was con-
firmed by independently observing free DNA bands at
3:1 charge ratio with increasing amounts of added
EDTA in the absence of DNase I treatment (data not
shown).
fection was seen at a charge ratio of 1:1, where only 25%
and 5% of DNA exists as lipid-DNA complex, respec-
tively. A large amount of added DNA exists as free and
uncomplexed. These data suggest that the transfection-
competent complex formation was very much less in
percentage; when viewed in combination with uptake
studies, it appears that nearly 50% complexed DNA
enters the cell.
When the accessibility of DNA in the lipid-DNA
complex was probed with DNase I, maximum protection
was observed at charge ratios of 3:1 and some protection
was also seen at 1:1 charge ratios (Figure 3). At lower
lipid to DNA charge ratios the free DNA was highly
susceptible to degradation by DNase I and was digested
to a large extent in the time period of incubation.
Control DNA, in the absence of the lipid (Figure 3, lane
Eth id iu m Br om id e Titr a tion of th e Lip id -DNA
Com p lex. Intercalation-induced fluorescence increase
and competition with cationic lipids to bind to DNA has
2
), was also completely digested within 10 min. DNA

4
180 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Banerjee et al.
F igu r e 2. Percent of free DNA in lipid-DNA complexes
prepared with four lipids (with cholesterol as colipid) and
DOTAP. Agarose gels were scanned and quantitated by
phosphorimager. By considering the total lane intensity as
F igu r e 4. EtBr exclusion from lipid-DNA complex. Decrease
in fluorescence of EtBr was used to assess the interaction of
DNA with lipids 1 (b), 2 (O), 3 (9), 4 (0), and DOTAP (2) (n
)
2). DNA-EtBr complex was titrated with increasing amounts
1
00%, the free and complexed DNA percentages were calcu-
of lipid. The order of addition of lipid or EtBr to DNA did not
alter the observed profiles. Fluorescence in the absence of lipid
was taken to be 100%.
lated and replotted. Free DNA (percent) was plotted against
charge ratio for lipids 1 (b), 2 (O), 3 (9), 4 (0), and DOTAP
(2) (n ) 2).
F igu r e 5. Cytotoxicity (viability) of lipids 1-4 on COS-1 cells.
Toxicity of lipids 1 (b), 2 (O), 3 (9), and 4 (0) was assesssed in
COS-1 cells by the MTT assay. The absorption obtained with
reduced formazon with cells in the absence of lipids was taken
to be 100%. The data presented are averages of three inde-
pendent experiments (n ) 3).
of 3:1 was less than 20% with lipids 2-4 and about 40%
with lipid 1, whereas the presence of 1.5:1 charge ratio
DOTAP decreased 90% of the EtBr fluorescence. EtBr
data when viewed in combination with the data ob-
tained with radioactive DNA (Figure 2) indicate that,
although at 3:1 charge ratio nearly 75% of the DNA
exists as a complex and is retained in the loading well,
the DNA may not be condensed, which is essential for
squeezing out the intercalated EtBr from the DNA.
F igu r e 3. Sensitivity of lipid-DNA complexes to DNase I.
The lipid-DNA complex was treated with DNase I as indicated
in the Experimental Section and electrophoresed. The details
of treatment are indicated at the top of each lane.
Cytotoxicity. An MTT-based viability assay was
performed to assess the cytotoxicity of the four lipid
formulations at various charge ratios with COS-1 cells.
The lipid-DNA complexation and treatment were iden-
tical to transfections. Lipids 1 and 4 showed the least
cytotoxicity even at 9:1 charge ratio (Figure 5). For lipids
1-4 at 3:1 charge ratio, the toxicity varied from 15% to
25% of the treated cells. At comparable charge ratios
the two-hydroxyl derivatives showed higher cytotoxicity
made EtBr an excellent tool to study lipid-DNA inter-
actions. To investigate the interaction of lipids 1-4 with
DNA, we have also titrated the EtBr-DNA complex
with increasing amounts of cationic lipids. The data in
Figure 4 shows that lipids 1-4 interact poorly with
DNA, as seen by their relatively poor ability to exclude
ethidium bromide from DNA when compared to DOT-
AP. The decrease in EtBr fluorescence at charge ratios
4
1
than the lipids 1-4.

Cationic Glycolipids in Liposomal Gene Delivery
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4181
Discu ssion
in specific cell uptake, e.g., utilizing asialoglycoprotein
receptor for gene transfer in liver cells.
4
8
The considerable promise shown by hydroxyethyl
headgroup-containing cationic lipids reported by us
Earlier we have demonstrated that for non-glycerol-
based simple monocationic lipids, a derivative contain-
ing two hydroxyethyl headgroups was better than a
4
1
and others28,33 has prompted us to investigate the
usefulness of cationic lipids with additional headgroup
hydroxyl functionalities for improving cell transfections.
In our earlier report,41 DHDEAB [N,N-di-(n-hexadecyl)-
N,N-di(hydroxyethyl)ammonium bromide] was shown to
be the most promising among the non-glycerol-based
simple monocationic transfection lipids we synthesized
and tested. Similarly, the DORIE28 series has shown
considerable promise and was commercialized. In the
present work, toward assessing the transfection ef-
ficiencies of non-glycerol-based monocationic transfec-
tion lipids with multiple hydroxyl functionalities in the
headgroup region, we have used 5-carbon sugars, ara-
binose and xylose, as the source of four hydroxyl groups.
Lipids 2-4 were designed with arabinose as the com-
mon headgroup with varying aliphatic hydrocarbon
chain lengths (C-14, C-16, and C-18), whereas lipid 1
contains xylose as the headgroup with two C-16 satu-
rated aliphatic hydrocarbon chains. Among the four
lipids, lipid 1 showed the highest transfection efficiency.
The reporter gene activity with lipid 1 was twice that
of DHDEAB. Lipid 2 was equal to DHDEAB in trans-
fection efficiency.
4
1
single-hydroxyl derivative Hydroxyl derivatives on
different backbones have also been shown to be very
promising, viz., DMRIE/DORIE, DOTIM, etc.28 The
present study demonstrates that transfection lipids with
four hydroxyls in the headgroup area are better than
transfection lipids containing two hydroxyalkyl head-
groups on COS-1 cells. On the basis of lipid hydration
and geometric considerations, factors such as a small
polar domain or close headgroup association and un-
saturated acyl chains were considered to promote fu-
3
3
sion. Lipids used in this study, based on chemical
structure, are expected to be more cylindrical in shape
given their large polar heads and saturated chains.
Though the interaction with the DNA was shown to be
weak by EtBr exclusion, their transfection ability was
very high. No clear understanding of the dependency
of in vitro transfection efficiency on the structure of
8
cationic lipid is yet available.
To quantitate the undigested DNA in DNase I protec-
tion experiments, the lipid-DNA complex was detergent-
solubilized or extracted by organic solvents. These
methodologies require careful optimizations. Similarly,
Lipids 2-4 with arabinose as the headgroup and
containing different lengths of hydrophobic chains
demonstrated good ability to deliver genes into the cells.
A myristoyl derivative of the cationic lipid (2) showed
maximum transfection. Myristoyl derivatives of several
other classes of cationic lipids have also been reported
before. Among them, the myristoyl derivatives of im1-
to assess the extent of association of DNA with the lipid,
12,14,47
sucrose or dextran gradients were used.
In this
study, we have used radiolabled DNA, obtained by nick
translation, to visualize the free and complexed DNA,
which does not involve any extraction or fractionation
procedures. Densitometric quantitation provides a con-
venient method to evaluate the process of DNA and lipid
complexation. Lipids 3 and 1, which showed good
transfections, differ in their ability to bind DNA as
assessed by electrophoresis (Figure 2). The poor ability
to cause exclusion of EtBr from DNA by the four lipids
even at 3:1 charge ratios (Figure 4) indicates that the
lipids 1-4 do not induce the conformational changes
required for EtBr exclusion. It has been proposed that
cationic lipids must be sufficiently neutralizing the
negative charges of DNA to induce a cooperative col-
[
2-(acyloxy)ethyl]-2-alkyl(alkenyl)-3-(2-hydroxyethyl)-
imidazolinium chloride have given higher in vitro
transfection in CV-1 cells compared to oleoyl and
4
palmitoyl derivatives. Szoka and co-workers similarly
reported the highest transfection efficiences of the
3
7
myristoyl analogue of carnitines. However, no strict
dependency on chain length was observed in transfec-
3
7
tion of CV-1 cells in the study reported by Wang et al.
The 2,3-di(alkyloxy)propyl quaternary derivative of
myristoyl was better compared to oleoyl, palmitoyl, and
1
7
lapse in the DNA structure. Lipids used in this study
do not induce such cooperative collapse of DNA, even
at 3:1 charge ratio. However, the extent of complexation
as seen by electrophoresis (Figure 2) suggests nearly
60-80% of the DNA exists as a complex at 3:1 charge
ratio, which was DNase I-insensitive. These observa-
tions suggest that the DNA was ensheathed by the lipid
bilayers but that this did not cause condensation of
DNA. Similar lack of condensation was observed in
lipid-DNA interactions by binding of dye TOPRO-1, a
3
8
stearoyl derivatives in transfecting the COS-7 cells.
The transfection efficiency of the caproyl derivative of
glycine betaines was better than those of lauroyl and
3
9
myristoyl derivatives in transfecting the HT 29 cells.
High Lâ to LR transition temperatures of the lipids
would render the membrane stiff and may not allow the
fluidity required for lipid mixing events in endosomal
escape or in any other fusogenic events the complex may
be involved.39 Among the saturated acyl chains, in
general, shorter chains, less than palmitic acid, seem
to help. Mannosylated and galactosylated cholesterol
derivatives are capable of cell-specific delivery of genes
4
7
DNA intercalating dye. Transfection studies, when
compared to lipofectin or DHDEAB, indicate higher
transfection, although the binding of lipid to DNA as
assessed by EtBr exclusion was weaker than that of
lipofectin or DHDEAB. Tight lipid-DNA complexes may
not be more efficient in transfection than weaker lipid-
(into macrophages and liver cells, respectively) and the
uptake could be inhibited by either mannose or galac-
tose.4
8,49
Lipidic glycosides with aminoalkyl pendant
8
groups showed good gene delivery and low toxicity to
HeLa and 3T3 cells.50 The lipid presented in this study
were not designed with any surface receptor in consid-
eration; however, it would be interesting to assess how
far these 5-carbon lipids are similar to 6-carbon lipids
DNA complexes as suggested. An ideal lipid formula-
tion should facilitate maximal uptake of DNA by the
cell and also should dissociate from DNA inside the cell.
As a colipid DOPE was ineffective compared to
cholesterol. DOPE, a HII phase-preferring lipid at physi-

4
182 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Banerjee et al.
ological temperatures, was considered to help the fuso-
genic ability of the complex.51 The fusogenic abilities of
DNase I, calf thymus DNA, NP-40, ethidium bromide, anti-
biotics, agarose, o-nitrophenyl â-D-galactopyranoside were
4
purchased from Sigma, St. Louis, MO. T DNA polymerase and
DOPE are not always required to obtain efficient
transfection. Cholesterol enhances the transfection ef-
ficiency of all the four lipids used in this study and also
of DHDEAB, the two-hydroxyethyl-containing transfec-
tion lipid reported by us.41 Cholesterol has also been
the nick translation buffer were purchased from New England
Biolabs Ltd., U.K. Radiolabeled DNA was purified on Promega
DNA purification columns. dNTP stocks and DNA molecular
weight markers were purchased from Bangalore Genei, Ban-
32
galore, India. [R- P]dATP was purchased from Bhabha Atomic
reported as a preferred colipid in transfection formula-
Research Centre, Bombay, India. Dioleoyl-3-(trimethylammo-
nium)propane (DOTAP), cholesterol, and dioleoylphosphati-
dylethanolamine (DOPE) were purchased from Avanti Polar
Lipids, Alabaster, AL. COS-1 cell line (SV 40 transformed
African green monkey kidney, ATCC CRL 1650) was obtained
from ATCC. Diisopropylidine dithioethyl derivatives of ara-
binose and xylose were obtained as generous gifts respectively
from Dr. P. K. Sasmal and Dr. P. Chand of the Indian Institute
of Chemical Technology. Unless otherwise stated, all reagents
were purchased from local commercial suppliers and were used
without further purification.
tions in vivo.52 Biophysical characteristics of DOTAP/
DOPE and DOTAP/cholesterol show that particle di-
ameter, ú potential, and amount of adsorbed protein
were less with cholesterol as colipid.53
Con clu sion s
In summary, we have developed an efficient and novel
series of non-glycerol-based cationic glycolipids for use
in in vitro liposomal gene delivery. Similar to dihydroxyl
derivatives, the four lipids reported in this study, based
on xylose or arabinose headgroups, transfect with
cholesterol but not with DOPE as colipid. The present
synthetic protocols should, in principle, be applicable
in synthesizing various glycosylated non-glycerol-based
structural analogues of lipids 1-4 for use in targeted
gene delivery. The transfection efficiencies of these new
lipids, when used in combination with cholesterol as a
colipid, were observed to be higher than that of DH-
DEAB, which was shown by us41 to be more transfection
efficient than LipofectAmine, one of the widely used
cationic transfection lipids. Structural studies on lipid-
DNA complexes of these four cationic lipids demonstrate
that the formation of the complex was not strong and
does not condense the DNA. However, the transfection
efficiencies were higher than those lipids that form
strong complexes with the DNA.
Syn th esis of N,N-Di-(n -h exa d ecyl)a m in e. Dry dichlo-
romethane (100 mL) was cooled to 0 °C and 5.5 g of n-
hexadecylaldehyde (22.9 mmol), 5.5 g of n-hexadecylamine
(22.9 mmol), and 2.75 g of anhydrous magnesium sulfate (22.9
mmol) were added. The mixture was kept under stirring for 3
h, and the temperature of the reaction mixture was allowed
to rise gradually from 0 °C to room temperature. The magne-
sium sulfate was filtered from the reaction mixture and the
filtrate was diluted with 50 mL of methanol. The diluted
dichloromethane/methanol solution was cooled to 0 °C and 4.35
g of sodium borohydride (115 mmol) was added. The solution
was stirred for 4 h and the temperature of the reaction mixture
was gradually raised to room temperature. The reaction
mixture was then taken up in 100 mL of chloroform and
washed with water (2 × 100 mL), and the chloroform layer
was dried over anhydrous magnesium sulfate and filtered.
Chloroform was removed from the filtrate on a rotary evapora-
tor, and column chromatographic (using 60-120 mesh size
silica gel) purification of the residue with 20-50% ethyl acetate
in petroleum ether as the eluent afforded 3.30 g (30% yield)
of pure N,N-di-(n-hexadecyl)amine.
Exp er im en ta l Section
1
3
H NMR of N,N-di-(n-hexadecyl)amine (200 MHz, CDCl ):
δ (ppm) ) 0.85 [t, 6H, CH
3
(CH
14-]; 1.4-1.6 [m, 4H, N (CH CH
2
].
2 n
) -]; 1.15-1.35 [m, 56H,
Gen er a l P r oced u r es a n d Ma ter ia ls. The high-resolution
mass spectrometric (HRMS) analysis were performed on a
Micromass AUTOSPEC-M mass spectrometer (Manchester,
U.K.) with an OPUS V3.1X data system. Data were acquired
by the liquid secondary ion mass spectrometry (LSIMS)
technique with m-nitrobenzyl alcohol as the matrix. LSIMS
analysis was performed in the scan range 100-1000 amu at
+
+
-(CH
2
)
2
2 2 2
-) ]; 2.55 [t, 4H, N (CH -
CH -)
2
Syn th esis of 1-Deoxy-1-[d ih exa d ecyl(m eth yl)a m m o-
n io]-D-a r a bin itol (3): Step a . The thioketal derivative of
2,3,4,5-diisopropylidine arabinose (Scheme 2; 1 g, 2.97 mmol)
was deprotected by refluxing in 25 mL of 80/20 acetone/water
(v/v) with 2.43 g (8.9 mmol) of mercuric chloride and 2.4 g (8.9
mmol) of red mercuric oxide for strictly 1.5 h. The inorganic
materials were filtered and the residue was washed with
chloroform (2 × 20 mL). Solvents from the combined filtrate
were evaporated and the residue was taken up in chloroform
(40 mL). The chloroform extract was sequentially washed with
1 N aqueous potassium iodide solution (2 × 30 mL) and water
(2 × 30 mL). The combined aqueous solution was extracted
with chloroform (2 × 30 mL), and this chloroform extract (2 ×
30 mL) was added to the first chloroform extract (40 mL). The
combined chloroform extract was dried over anhydrous sodium
sulfate and filtered, and the chloroform from the filtrate was
evaporated in a rotary evaporator below 35 °C. The crude
aldehyde residue (636 mg, 2.97 mmol) was taken up in 10 mL
of dichloromethane, and 1.55 mg (3.2 mmol) of N,N-di-(n-
hexadecyl)amine was added to the solution. After the reaction
was stirred for 0.5 h under nitrogen in room temperature,
sodium cyanoborohydride 410 mg (2.97 mmol) was added. The
reaction mixture was stirred at room temperature for 30 h
under nitrogen. Removal of the solvent followed by silica gel
column chromatographic purification of the resulting residue
with 60-120 mesh size silica and 98/2 (v/v) hexane/acetone
as the eluent finally afforded 780 mg of the intermediate
tertiary amine (white gummy liquid, intermediate II in
1
the rate of 3 scans/s. H NMR spectra were recorded on a
Varian FT 200 MHz, Bruker 300 and 500 MHz. Unless
otherwise stated, all reagents were purchased from local
commercial suppliers and were used without further purifica-
tion. Final purification of the transfection lipids 1-4 was
carried out by semipreparative reverse-phase HPLC (on a
Whatman, Partisil 10 ODS-3 column). The LC system con-
sisted of a Shimadzu model LC10A. The progress of the
reactions was monitored by thin-layer chromatography on 0.25
mm silica gel plates. Column chromatography was performed
with silica gel (Acme Synthetic Chemicals, India, finer than
2
00 and 60-120 mesh). Elemental analyses (C, H, N) were
performed at the microanalytical laboratory of Central Salt
and Marine Chemical Research Institute, Bhavnagar, Gujarat,
India. The hydrophobic secondary amines namely, N,N-di-(n-
tetradecyl)amine, N,N-di-(n-hexadecyl)amine, and N,N-di-(n-
octadecyl)amine used in step a of Schemes 1 and 2 for the
syntheses of the lipids 1-4 were prepared by the same
synthetic procedure as described below for the synthesis of
N,N-di-(n-hexadecyl)amine. Lipids 1-4 were synthesized by
the synthetic procedure outlined in Schemes 1 and 2. As a
representative of the procedure detail, synthesis of the lipid 3
is delineated below. pCMV.SPORT-â-gal, cell culture media,
and fetal calf serum were purchased from Gibco-BRL, Rock-
ville, MD. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide (MTT), poly(ethylene glycol) 8000, Trizma base,
Scheme 2, where R) n-C16
f
H33, R 0.8 in 80/20 hexane/acetone
as the TLC developing solvent) in 40% yield.

Cationic Glycolipids in Liposomal Gene Delivery
1H NMR (200 MHz, CDCl
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4183
1H NMR (200 MHz, CDCl
) of diisopropylidine quaternary
3
3
) of intermediate tertiary amine
(
II, Scheme 2, R ) n-C16 33): δ (ppm) ) 0.9 [t, 6H, CH
H
3
(CH
2
)
)
n
-
-
ammonium iodide [intermediate III in Scheme 2 where R )
37]: δ (ppm) 0.9 [t, 6H, CH (CH -]; 1.20-1.40 [m, 60H,
15-]; 1.4 and 1.5 [s, 12H, {COC(CH OC} ]; 1.65-1.95
N ]; 3.55-3.90 [m,
]
; 1.20-1.40 [m, 52H, -(CH 13-]; 1.4-1.6 [m, 16H, COC(CH
2
)
3
2
n-C18
H
)
3
2 n
)
1
2
OC, N(CH
2
CH
2
-)
2
]; 2.3-2.6 [m, 5H, -CH H N(CH
2
CH
2
2
-)
2
]; 2.8
-(CH
2
3
)
2
+
2
1
2
+
[
dd, 1H, -CH H N(CH
2
CH
CH
CH
FABS (LSIMS) m/z 679 [M ] (calcd for C43
2
-)
-)
-)
2
]; 3.55 [t, 1H, (O)CH
]; 3.8-4.2 [m, 4H, (O)CH
].
CH(O)-
[m, 4H, N (CH
7H, (O)CH H CH(O)CH(O)CH(O)CH N (CH
4.05 [m, 1H, (O)CH H CH(O)CH(O)CH(O)CH
2
CH
2
-)
2 3
]; 3.55 [s, 3H, CH
1
2
+
CH(O)CH(O)CH
CH(O)CH(O)CH
2
N (CH
2
2
2
2
CH(O)-
2
2
CH
2
-)
2
]; 3.85-
CH -) ];
1
2
+
2
N (CH
2
2
2
2
N (CH
2
2
2
+
+
4.05-4.25 [m, 2H, (O)CH CH(O)CH(O)CH(O)CH N (CH CH -
H
85NO
4
, 38%,).
2
2
2
+
2
)
2
]; 4.35-4.50 [m, 1H, (O)CH
2
CH(O)CH(O)CH(O)CH
2
N ].
Step b. The intermediate tertiary amine (II, Scheme 2, R)
33) (210 mg, 0.3 mmol) prepared in step a was reacted
HRMS (LSIMS) m/z calcd (for C42
ion, 100%) 670.6713; found 670.6731. Anal. (C42
.5H O) C, H, N.
Syn th esis of 1-Deoxy-1-[d ih exa d ecyl(m eth yl)a m m o-
H
88NO
4
the 4° ammonium
n-C16
H
H
88ClNO ‚
4
with excess methyl iodide (7 mL) for 3 h at room temperature.
The residue after evaporation of the excess methyl iodide was
column-purified on 60-120 mesh size silica gel. The column
was eluted sequentially with 5% acetone/hexane and pure
chloroform to remove the relatively less polar impurities. The
pure quaternized ammonium iodide salt (intermediate III in
Scheme 2 where R) n-C16H33, R 0.6 in 10/90 methanol/
f
chloroform) was finally eluted with 10% methanol in chloro-
form (250 mg, 98% yield).
1
2
n io]-D-xylitol (1). As outlined in Scheme 1, the title lipid 1
(waxy semisolid) was synthesized following the same synthetic
procedure described above for preparing the lipid 3 but with
2,3,4,5-diisopropylidinexylose thioketal as the starting mate-
rial with an overall yield of 18.4%. All the isolated intermedi-
ates gave spectroscopic data in agreement with their assigned
1H NMR (200 MHz, CDCl
structures shown in Scheme 1.
3
) of the diisopropylidine quater-
1
3
H NMR (200 MHz, CDCl ) of diisopropylidine quaternary
nary ammonium iodide (intermediate III in Scheme 2 where
(CH
13-]; 1.4 and 1.5 [s, 12H, {COC(CH
ammonium iodide intermediate [Scheme 1, R ) n-C16
ppm) ) 0.9 [t, 6H, CH (CH -]; 1.20-1.40 [m, 52H, -(CH
]; 1.40-1.60 [m, 12H, {COC(CH ) OC} ]; 1.60-1.95 [m, 4H,
H33]: δ
R ) n-C16
H
33): δ (ppm) ) 0.9 [t, 6H, CH
3
2
)
n
-]; 1.20-1.40
OC} ];
N ]; 3.55-
N (CH CH -) ];
.85-4.05 [m, 1H, (O)CH H CH(O)CH(O)CH(O)CH N (CH
-) ]; 4.05-4.25 [m, 2H, (O)CH CH(O)CH(O)CH(O)CH
CH -) ]; 4.35-4.50 [m, 1H, (O)CH CH(O)CH(O)CH(O)-
(
3
2
)
n
2 13
) -
[
m, 52H, -(CH
.65-1.95 [m, 4H, N (CH
.90 [m, 7H, (O)CH H CH(O)CH(O)CH(O)CH
2
)
3
+
)
2
2
+
1
3
3
CH
(
2
CH
2
-)
2
]; 3.55 [s, 3H, CH
3
3
2
2
+
+
1
2
+
N (CH CH -) ]; 3.45 [s, 3H, -CH (CH )N (CH CH -) ] 3.50-
2
2
2
2
3
2
2
2
2
2
2
2
1 2
+
1
2
+
3.70 [m, 5H, -CH H N (CH CH -) ]; 3.90-4.00 [m, 2H, (O)-
1 2 1 2
-
2
2
2
+
2
2
+
CH H CH(O)CH(O)CH(O)CH H N (CH CH -) ]; 4.10-4.20 [m,
1 2
N -
2
2
2
2
2
2
2
1
2
H, (O)CH H CH(O)CH(O)CH(O)CH ]; 4.40-4.50 [m, 1H, (O)-
CH
2
2
2
2
+
CH CH(O)CH(O)CH(O)CH ]; 4.55-4.70 [m, 2H, (O)CH CH-
CH
2
N ].
2
2
2
+
(
2
O)CH(O)CH(O)CH N ].
HRMS (LSIMS) m/z calcd (for C44
4
H88NO the 4°-ammonium
HRMS (LSIMS) m/z calcd (for C38
H
80NO
4
the 4° ammonium
ion, 100%) 694.6713; found 694.6720.
ion, 100%) 614.6087; found 614.6104. Anal. (C38
.5H O) C, H, N.
Preparation of Plasmid DNA. pCMV-â-gal plasmid DNA
was prepared by the alkaline lysis procedure and purified by
4
H80ClNO ‚
Step c. The diisopropylidine deprotection of the diisopro-
1
2
pylidine-protected quaternized ammonium iodide salt (Scheme
2
stirring the quaternized salt (130 mg, 0.16 mmol) in 5 mL of
neat trifluoroacetic acid (TFA) at 55 °C for 48 h. Excess TFA
was evaporated and the residue was loaded over a chloride
ion-exchange resin. The title compound 3, 103 mg, waxy,
, R ) n-C16H33) obtained in step b was accomplished by
4
5
PEG-8000 precipitation according to published procedures.
The plasmid preparations showing a value of A260/A280 more
than 1.8 were used.
P r ep a r a tion of Lip osom es. Cationic amphiphiles and the
colipid (DOPE or cholesterol) in the appropriate ratios were
dissolved in chloroform in a glass vial. Chloroform was
removed with a thin flow of moisture-free nitrogen and the
dried film of lipid left in the vial was then kept under high
vacuum for 8 h. One milliliter of sterile deionized water was
added to the vacuum-dried lipid film, and the mixture was
allowed to swell for 15 h (overnight). The vial was then
vortexed for 2-3 min at room temperature and occasionally
shaken in a 45 °C water bath to produce multilamellar vesicles
f
semisolid, 100%, R 0.25 in 10/90 methanol/chloroform) was
eluted with methanol. The eluted compound was further
purified by treatment with charcoal in hot methanol. The final
HPLC purification of the title compound was carried out on a
semipreparative reverse-phase column with 95/5 methanol/
water at a flow rate of 5 mL/min (UV detector was set at 219
nm). Pure 3 was obtained in an overall yield of 39.2%.
HRMS (LSIMS) m/z calcd (for C38
ion, 80%) 614.6087; found 614.6045. Anal. (C38
.0H O) C, H, N.
Syn th esis of 1-Deoxy-1-[m eth yl(d itetr a d ecyl)a m m o-
Η80ΝÃ
4
the 4° ammonium
H
80ClNO
4
‚
1
2
(MLV). Small unilamellar vesicles (SUV) were then prepared
by sonication of the MLV placed in an ice bath for 3-4 min
until clarity was achieved, by use of a Branson 450 sonifier at
n io]-D-a r a bin itol (2). The title lipid 2 (waxy, semisolid) was
synthesized following the same procedure as described above
for synthesizing lipid 3 but with di-(n-tetradecyl)amine with
an overall yield of 40%. All the isolated intermediates gave
spectroscopic data in agreement with their assigned structures
shown in Scheme 2.
1
00% duty cycle. The formation of SUV was assured by
spinning the sonicated preparation for 10 min at 10 000 rpm
in a tabletop centrifuge. If a pellet formed, then the sample
was sonicated further until a clear solution was obtained.
Nick Tr a n sla tion of P la sm id DNA. Supercoiled plasmid
DNA, pCMV-â-gal (7.8 kb), was purified by alkaline lysis
1H NMR (200 MHz, CDCl
ammonium iodide [intermediate III in Scheme 2 where R )
-(CH
11-]; 1.4 and 1.5 [s, 12H, {COC(CH
) of diisopropylidine quaternary
3
45
followed by precipitation with poly(ethylene glycol). The DNA
n-C14
4
1
H
29]: δ (ppm) ) 0.9 [t, 6H, CH
4H, -(CH
2 2 3
.95 [m, 4H, N (CH CH -) ]; 3.55 [s, 3H, CH
3
2
)
n
-]; 1.20-1.40 [m,
OC} ]; 1.65-
N ]; 3.60-3.90
N (CH CH -) ]; 3.90-
N (CH CH -) ];
N (CH CH
obtained was resuspended in TE buffer (10 mM Tris-HCl and
2
)
3
)
2
+
2
1
mM EDTA, pH 8.0). To obtain radiolabeled DNA, pCMV-â-
+
2
gal plasmid was nick-translated with DNase I and T4 DNA
polymerase, in the presence of [R- P]dATP, dGTP, dCTP, and
1
2
+
[m, 7H, (O)CH H CH(O)CH(O)CH(O)CH
2
2
2
2
32
1
2
+
4
.05 [m, 1H, (O)CH H CH(O)CH(O)CH(O)CH
.10-4.30 [m, 2H, (O)CH CH(O)CH(O)CH(O)CH
]; 4.35-4.50 [m, 1H, (O)CH CH(O)CH(O)CH(O)CH
HRMS (LSIMS) m/z calcd (for C34 the 4° ammonium
ion, 100%) 558.5461; found 558.5448.
2
2
2
2
dTTP. The unincorporated nucleotides were removed with
Promega’s Wizard DNA clean up system. The nick-translated
DNA was finally eluted with TE buffer, analyzed by agarose
gel electrophoresis, and shown to have a specific activity of
+
4
2
2
2
+
2
-
)
2
2
2
N ].
H
72NO
4
5
2.2 × 10 cpm/µg of DNA.
Syn t h esis of 1-Deoxy-1-[m et h yl(d ioct a d ecyl)a m m o-
n io]-D-a r a bin itol (4). The title lipid 4 (waxy, semisolid) was
synthesized following the same synthetic procedure as de-
scribed above for preparing the lipid 3 but with di-(n-octade-
cyl)amine with an overall yield of 37%. All the isolated
intermediates gave spectroscopic data in agreement with their
assigned structures shown in Scheme 2.
Bin d in g of Lip id to DNA a n d Sen sitivity of Lip id -
DNA Com p lex to DNa se I. Nick-translated pCMV-â-gal
3
2
DNA, labeled with [R- P]dATP, was complexed with cationic
lipids at charge ratios varying from 0.3:1 to 3:1 (lipid/DNA) to
measure the DNA binding ability and the susceptibility of
these complexes to degradation by DNase I. Briefly, in a typical
binding assay, 0.15 nmol of DNA, corresponding to 50 ng of

4
184 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Banerjee et al.
nick-translated DNA, was complexed with the lipids, at
indicated charge ratios, in 10 mM phosphate buffer (pH 7.5),
in a volume of 15 µL and incubated at room temperature for
N-(hydroxyethyl)ammonium bromide; DOTAP, 1,2-
di(oleoyloxy)-3-(trimethylamino)propane; DHDEAB, N,N-
di-(n-hexadecyl)-N,N-di(hydroxyethyl)ammonium bro-
mide; HDEAB, N-(n-hexadecyl)-N,N-di(hydroxyethyl)-
ammonium bromide; MOOHAC, N-methyl-N-(n-octa-
decyl)-N-oleyl-N-(hydroxyethyl)ammonium chloride;
DOMHAC, N-methyl-N,N-di-(n-octadecyl)-N-(hydroxy-
ethyl)ammonium chloride; DOHEMAB, N,N-di[[(O-hexa-
decanoyl)hydroxy]ethyl]-N-(hydroxyethyl)-N-methylam-
monium bromide; DOSPA, 2,3-di(oleyloxy)-N-[2-(sper-
minecarboxamido)ethyl]-N,N-dimethyl-1-propanammin-
ium trifluoroacetate; DDAB, dioctadecyldimethylam-
monium bromide; DOPE, 1,2-dioleoylpropyl-3-phospha-
tidylethanolamine.
3
0 min on a gyratory shaker. Subsequently, the complexes
were treated with DNase I, in the presence of 5 mM MgCl
2
,
at a final concentration of 0.1 µg/0.15 nmol of DNA, and further
incubated for 10 min at room temperature. The reactions were
halted with EDTA added to a final concentration of 30 mM,
and the reaction products, after addition of the tracking dye
bromophenol blue, were directly loaded on a 1% agarose gel
with TAE buffer and electrophoresed at 100 V for approxi-
mately 2 h. The gel was subsequently dried, exposed overnight,
and scanned on a Fuji Film phosphorimager (BAS 1800), and
the bands were quantitated with Image Gauge software
version 8.01.
Exclu sion of Eth id iu m Br om id e (EtBr ) fr om DNA by
Ca tion ic Lip id s. The extent of EtBr binding to the DNA was
monitored by changes in the fluorescence. EtBr fluorescence
was monitored in a Hitachi 4500 fluorometer by setting the
excitation wavelength at 518 nm and the emission wavelength
at 585 nm. To 1 mL of TE buffer (pH 8.0) were added 0.78
nmol of DNA and 2.5 nmol of EtBr. The change in fluorescence
was monitored after addition of small volumes of lipid to the
EtBr-DNA complex. Arbitrary fluorescence values were re-
corded after sufficient time was allowed for equilibration. The
order of addition of EtBr or lipid to DNA did not alter the final
value, indicating that the equilibrium does not depend on the
order of addition and is attained in minutes. Percent fluores-
cence was calculated by considering the fluorescence value in
the absence of lipid as 100%.
Ack n ow led gm en t. Financial support for this work
from the Department of Biotechnology (to A.C.) and
from the Council of Scientific and Industrial Research
(CSIR), Government of India, New Delhi (doctoral
research fellowships to R.K.B and Y.V.M.), are gratefully
acknowledged. We sincerely acknowledge the experi-
mental assistance received from Ramadas Y. and Koush-
ik Mukherjee in preparing the revised manuscript.
Su p p or tin g In for m a tion Ava ila ble: Elemental analysis
data for lipids 1, 3, and 4; two HPLC chromatograms (with
two different mobile phases) for lipid 2; and Figure S1, showing
agarose gel electrophoresis of the susceptibility of DNA in
lipid-DNA complexes to DNases with lipids 1-4 (with cho-
lesterol as colipid) and DOTAP. This material is available free
of charge via the Internet at http://pubs.acs.org.
Tr a n sfection of Cells. Transfection of the cells was done
4
1
essentially as described earlier. Briefly, COS-1 cells were
seeded at a density of 15 000 cells/well in a 96-well plate 18 h
before the transfection. Plasmid (0.3 µg) was complexed with
varying amount of lipid (0.05-4.3 nmol) in 13 µL of plain
DMEM medium for 30 min. The charge ratios were varied from
Refer en ces
(1) Mountain, A. Gene Therapy: The First Decade. Trends Biotech-
nol. 2000, 18, 119-128 and references therein.
0
.1:1 to 9:1 (() over this range of the lipid. The complex was
(2) Felgner, P. L. Particulate systems and Polymers for in vitro and
diluted to 100 µL with plain DMEM and added to the wells.
After 3 h of incubation, 100 µL of DMEM with 10% FCS was
added to the cells. The medium was changed to 10% complete
medium after 24 h and the reporter gene activity was
estimated after 48 h. The cells were washed twice with PBS
and lysed in 50 µL of lysis buffer (0.25 M Tris-HCl, pH 8.0,
and 0.5% NP40). Care was taken to ensure complete lysis. The
â-galactosidase activity per well was estimated by adding 50
µL of 2× substrate solution (1.33 mg/mL ONPG, 0.2 M sodium
phosphate, pH 7.15, and 2 mM magnesium chloride) to the
lysate in a 96-well plate. Absorption at 405 nm was converted
to â-galactosidase units by use of a calibration curve con-
structed with pure commercial â-galactosidase enzyme. The
values of â-galactosidase units in replicate plates assayed on
the same day varied by less than 30%. The transfection
efficiency values are the average values from four replicate
transfection plates assayed on the same day. Each transfection
experiment was repeated three times on different days and
the day-to-day variation in average transfection efficiency was
found to be within 1-fold. The transfection profiles obtained
on different days were identical.
in vivo Delivery of Polynucleotides. Adv. Drug Delivery Rev.
1
990, 5, 163-187.
(3) Behr, J . P. Gene Transfer with Synthetic Cationic Am-
phiphiles: Prospects for Gene Therapy. Bioconjugate Chem.
1
994, 5, 382-389.
(
4) Solodin, I.; Brown, S. C.; Bruno, M. S.; Chow, C.-Y.; J ang, E.-
H.; Debs, R.; Heath, T. A Novel Series of Amphiphilic Imidazo-
linium Compounds for In Vitro and In Vivo Gene Delivery.
Biochemistry 1995, 34, 13537-13544.
(5) Gao, X.; Huang, L. Potential of Cationic Liposome-Mediated
Gene Delivery by Polycations. Biochemistry 1996, 35, 1027-
1
036.
(
6) Wyman, T. B.; Nicol, F.; Zelphati, O.; Scaria, P. V.; Plank, C.;
Szoka, F. C., J r. Design, Synthesis and Characterization of a
Cationic Peptide that Binds to Nucleic Acids and Permeabilizes
Bilayers. Biochemistry 1997, 36, 3008-3017.
(7) Aberle, A. M.; Tablin, F.; Zhu, J .; Walker, N. J .; Gruenert, D.
C.; Nantz, M. H. A Novel Tetraester Construct that Reduces
Cationic Lipid-Associated Cytotoxicity. Implications for the
Onset of Cytotoxicity. Biochemistry 1998, 37, 6533-6540.
8) Miller, A. D. Cationic Liposomes in Gene Therapy. Angew.
Chem., Int. Ed. Engl. 1998, 37, 1768-1785 and references
therein.
(
(
9) Li, S.; Huang, L. Nonviral Gene Therapy: Promises and Chal-
lenges. Gene Ther. 2000, 7, 31-34.
10) Shi, N.; Pardridge, W. M. Noninvasive Gene Targeting to the
Toxicity Assa y. Cytotoxicity of lipids 1-4 was assessed by
(
the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-
Brain. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 7567-7572.
4
6
mide (MTT) reduction assay as described earlier. The cyto-
toxicity assay was performed in 96-well plates by maintaining
a constant ratio of number of cells to amount of cationic lipid,
as in transfection experiments. MTT was added 3 h after
addition of the cationic lipid to the cells. Results were
expressed as percent viability ) [A540(treated cells) - back-
ground]/[A540(untreated cells) - background] × 100.
(11) Bousiff, O.; Zanta, M. A.; Behr, J .-P. Optimized Galenics Improve
in vitro Gene Transfer by Cationic Molecules up to 1000-fold.
Gene Ther. 1996, 3, 1074-1080.
(
12) Eastman, S. J .; Siegel, C.; Tousignant, J .; Smith, A. E.; Cheng,
S. H.; Scheule, R. K. Biophysical Characterization of Cationic
Lipid:DNA Complexes. Biochim. Biophys. Acta 1997, 1325, 41-
62.
(13) Harvie, P.; Wong, F. M. P.; Bally, M. B. Characterization of
Lipid-DNA Interactions: Destabilization of Bound Lipids and
DNA Dissociation. Biophys. J . 1998, 75, 1040-1051.
14) Xu, Y. H.; Hui, S.-W.; Frederik, P.; Szoka, F. C., J r. Physico-
chemical Characterization and Purification of Cationic Li-
poplexes. Biophys. J . 1999, 77, 341-353.
15) Bhattacharya, S.; Mandal, S. S. Evidence of Interlipidic Ion-
Pairing in Anion-Induced DNA Release from Cationic Am-
phiphile-DNA Complexes. Mechanistic Implications in Trans-
action. Biochemistry 1998, 37, 7764-7777.
Abbr evia tion s
(
DOTMA, 1,2-dioleyl-3-(N,N,N-trimethylamino)pro-
pane chloride; DMDHP, (()-N,N-[bis(2-hydroxyethyl)]-
N-[2,3-bis(tetradecanoyloxy)propyl]ammonium chloride;
DMRIE, N-[1,2-di(myristyloxy)prop-3-yl]-N,N-dimethyl-
(

Cationic Glycolipids in Liposomal Gene Delivery
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4185
(
(
(
16) Zuidam, N. J .; Hirsch-Lerner, D.; Margulies, S.; Barenholz, Y.
Lamellarity of Cationic Liposomes and Mode of Preparation of
Lipoplexes Affect Transaction Efficiency. Biochim. Biophys. Acta
(35) Templeton, N. S.; Lasic, D. D.; Frederik, P. M.; Strey, H. H.;
Roberts, D. D.; Pavlakis, G. N. Improved DNA:Liposome Com-
plexes for Increased Systemic Delivery and Gene Expression.
Nat. Biotechnol. 1997, 15, 647-652.
1
999, 1419, 207-220.
17) Gershan, H.; Ghirlando, R.; Guttman, S. B.; Minsky, A. Mode
of Formation and Structural Features of DNA-Cationic Lipo-
some Complexes used for Transfection. Biochemistry 1993, 32,
(
36) Hara, T.; Tan, Y.; Huang, L. In Vivo Gene Delivery to the Liver
using Reconstituted Chylomicron Remnants as a Novel Nonviral
Vector. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 14547-14552.
37) Wang, J . K.; Guo, X.; Xu, Y.; Barron, L.; Szoka, F. C., J r.
Synthesis and Characterization of Long Chain Alkyl Acyl
Carnitine Esters.Potentially Biodegradable Cationic Lipids for
use in Gene Delivery. J . Med. Chem. 1998, 41, 2207-2215.
38) Floch, V.; LeBolc’h, G. Le.; Gable-Guillaume, C.; LeBris, N.;
Yaouanc, J .-J .; Abbayes, H. Des.; Fe’rec, C.; Cle’ment, J .-C.
Phospholipids as Non-Viral Vectors for Gene Therapy. Eur. J .
Med. Chem. 1998, 33, 923-934.
7
143-7151.
(
18) Radler, J . O.; Koltover, I..; Salditt, T.; Safinya, C. R. Structure
of DNA-Cationic Liposome Complexes: DNA Intercalation in
Multilamellar Membranes in Distinct Interhelical Packing
Regimes. Science 1997, 275, 810-814.
(
(
(
(
19) Zabner, J .; Fasbender, A. J .; Moninger, T.; Poellinger, K. A.;
Welsh, M. J . Cellular and Molecular Barriers to Gene Transfer
by a Cationic Lipid. J . Biol. Chem. 1995, 270, 18997-19007.
20) Xu, Y.; Szoka, F. C., J r. Mechanism of DNA Release in Cationic
Liposome/DNA Complexes Used in Cell Transfection. Biochem-
istry 1996, 35, 5616-5623.
(39) Floch, V.; Legros, N.; Loisel, S.; Guillaume, C.; Guilbot, J .;
Benvegnu, T.; Ferrieres, V.; Plusquellec, D.; Ferec, C. New
Biocompatible Cationic Amphiphiles Derivative from Glycine
Betaine; A Novel Family of Efficient Nonviral Gene Transfer
Reagent. Biochem. Biophys. Res. Commun. 1998, 251, 360-365.
21) Lechardeur, D.; Sohn, K.-J .; Haardt, M.; J oshi, P. B.; Monck,
M.; Graham, R. W.; Beatty, B.; Squire, J .; Brodovich, H.; Lukacs,
G. L. Metabolic Instability of Plasmid DNA in the Cytosol: A
potential Barrier to Gene Transfer. Gene Ther. 1999, 6, 482-
(
40) Cameron, F. H.; Moghaddam, M. J .; Bender, V. J .; Whittaker,
R. G.; Mott, M.; Lockett, T. J . A Transfection Compound Series
based on a Versatile Tris Linkage. Biochim. Biophys. Acta 1999,
4
97.
(
22) Sebestyen, M. G.; Ludtke, J . J .; Bassik, M. C.; Zhang, G. F.;
Budker, V.; Lukhtanov, E. A.; Hagstrom, J . E.; Wolff, J . A. DNA
Vector Chemistry: The Covalent Attachment of Signal Peptides
to Plasmid DNA. Nat. Biotechnol. 1998, 16, 80-85.
1
417, 37-50.
(
41) Banerjee, R.; Das, P. K.; Srilakshmi, G. V.; Chaudhuri, A.; Rao,
N. M. Novel Series of Non-Glycerol-Based Cationic Transfection
Lipids for Use in Liposomal Gene Delivery. J . Med. Chem. 1999,
42, 4292-4299.
(
23) Felgner, P. L.; Gadek, T. R.; Holm, M.; Roman, R.; Chan, W.;
Wenz, M.; Northorp, J . P.; Ringold, G. M.; Danielsen, M.
Lipofectin: A Highly Efficient Lipid-Mediated DNA Transaction
Procedure. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 7413-7417.
24) Leventis, R.; Silvius, J . R. Interactions of Mammalian Cells with
Lipid Dispersions containing Novel Metabolizable Cationic Am-
phiphiles. Biochim. Biophys. Acta 1990, 1023, 124-132.
25) Gao, X.; Huang, L. A Novel Cationic Liposome Reagent for
Efficient Transfection of Mammalian Cells. Biochem. Biophys.
Res. Commun. 1991, 179, 280-285.
(42) Fichert, T.; Regelin, A.; Massing, U. Synthesis and Transfection
Properties of Novel Nontoxic Monocationic Lipids: Variations
of Lipid Anchor, Spacer and Headgroup Structure. Bioorg. Med.
Chem. Lett. 2000, 10, 787-791.
(
(
(
(
43) Subramanian, M.; Holopainena, J . M.; Paukkua, T.; Erikssona,
O.; Huhtaniemib, I.; Kinnunena, P. K. Characterization of Three
Novel Cationic Lipids as Liposomal Complexes with DNA.
Biochim. Biophys. Acta 2000, 1466, 289-305.
26) Stewart, M. J .; Plautz, G. E.; Del Buono, L.; Yang, Z. Y.; Xu, L.;
Gao, X.; Huang, L.; Nabel, E. G.; Nabel, G. J . Gene Transfer In
Vivo with DNA-Liposome Complexes: Safety and Acute Toxic-
ity in Mice. Hum. Gene Ther. 1992, 3, 267-275.
(
44) Ghosh, Y. K.; Visweswariah, S. S.; Bhattacharya, S. Nature of
Linkage between the Cationic Headgroup and Cholesteryl
skeleton controls gene transfection efficiency. FEBS Lett. 2000,
473, 341-344.
(
27) Alton, E. W.; Middleton, P. G.; Caplen, N. J .; Smith, S. N.; Steel,
D. M.; Munkonge, F. M.; J effery, P. K.; Stevension, B. J .;
McLachlan, G.; Dorin, J . R.; Porteous, D. J . Noninvasive
Liposome-Mediated Gene Delivery Can Correct the Ion-Trans-
port Defect in Cystic Fibrosis Mutant Mice. Nat. Genet. 1993,
(45) Sambrook, J .; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A
Laboratory Manual, 2nd ed.; Cold Spring Harbor Laboratory
Press: Cold Spring Harbor, NY, 1989.
(46) Hansen, M.-B.; Neilson, S. E.; Berg, K. Reexamination and
further Development of a Precise and Rapid Dye Method for
measuring Cell Growth/Cell Kill. J . Immunol. Methods 1989,
5
, 135-142.
(
28) Felgner, J . H.; Kumar, R.; Sridhar, C. N.; Wheeler, C. J .; Tsai,
Y.-J .; Border, R.; Ramsey, P.; Martin, M.; Felgner, P. L.
Enhanced Gene Delivery and Mechanism Studies with a Novel
Series of Cationic Lipid Formulations. J . Biol. Chem. 1994, 269,
1
19, 203-210.
(
47) Reimer, D. L.; Kong, S.; Bally, M. B. Analysis of Cationic
Liposome-Mediated Interactions of Plasmid DNA with Murine
and Human Melanoma Cells In Vitro. J . Biol. Chem. 1997, 272,
2
550-2561.
(
29) Liu, Y.; Liggitt, D.; Zhong, W.; Tu, G.; Gaensler, K.; Debs, R.
Cationic Liposome-Mediated Intravenous Gene Delivery in Mice.
J . Biol. Chem. 1995, 270, 24864-24870.
1
9480-19487.
(
48) Kawakami, S.; Sato, A.; Nishikawa, M.; Yamashita, F.; Hashida,
M. Mannose Receptor-Mediated Gene Transfer into Macro-
phages using Novel Mannosylated Cationic Liposomes. Gene
Ther. 2000, 7, 292-299.
(
30) Caplen, N. J .; Alton, E. W.; Middleton, P. G.; Dorin, J . R.;
Stevension, B. J .; Gao, X.; Durham, S. R.; J effery, P. K.; Hodson,
M. E.; Coutelle, C.; Huang, L.; Porteous, D. J .; Williamson, R.;
Geddes, D. M. Liposome-Mediated CFTR Gene Transfer to the
Nasal Epithelium of Patients with Cystic Fibrosis. Nat. Med.
(49) Kawakami, S.; Yamashita, F.; Nishikawa, M.; Takakura, Y.;
Hashida, M. Asialoglycoprotein Receptor-Mediated Gene Trans-
fer Using Novel Galactosylated Cationic Liposomes. Biochem.
Biophys. Res. Commun. 1998, 252, 78-83.
1
995, 1, 39-46.
(
31) Wheeler, C. J .; Felgner, P. L.; Tsai, Y. J .; Marshall, J .; Sukhu,
L.; Doh, G.; Hartikka, J .; Nietupski, J .; Manthorpe, M.; Nichols,
M.; Piewe, M.; Liang, X.; Norman, J .; Smith, A.; Cheng, S. H.
A. Novel Cationic Lipid Greatly Enhances Plasmid DNA Deliv-
ery and Expression in Mouse Lung. Proc. Natl. Acad. Sci. U.S.A.
(
(
(
50) Bessodes, M.; Dubertret, C.; J aslin, G.; Scherman, D. Synthesis
and Biological Properties of New Glycosidic Cationic Lipids for
DNA Delivery. Bioorg. Med. Chem. Lett. 2000, 10, 1393-1395.
51) Zelphati, O.; Szoka, F. C., J r. Mechanism of Oligonucleotide
Release from Cationic Liposomes. Proc. Natl. Acad. Sci. U.S.A.
1
996, 93, 11454-11459.
(
32) Hofland, H. E. J .; Shephard, L.; Sullivan, S. M. Formation of
Stable Cationic Lipid/DNA Complexes for Gene Transfer. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 7305-7309.
1
996, 93, 11493-11498.
52) Liu, Y.; Mounkes, L. C.; Liggitt, H. D.; Brown, C. S.; Solodin, I.;
Heath, T. D.; Debs, R. J . Factors Influencing the Efficiency of
Cationic Liposome-Mediated Intravenous Gene Delivery. Nat.
Biotechnol. 1997, 15, 167-173.
(
33) Bennett, M. J .; Aberle, A. M.; Balasubramaniam, R. P.; Malone,
J . G.; Malone, R. W.; Nantz, M. H. Cationic Lipid-Mediated Gene
Delivery to Murine Lung: Correlation of Lipid Hydration with
in Vivo Transfection Activity. J . Med. Chem. 1997, 40, 4069-
4
078.
(53) Li, S.; Rizzo, M. A.; Bhattacharya, S.; Huang, L. Characterization
of Cationic Lipid-Protamine-DNA (LPD) Complexes for Intra-
venous Gene Delivery. Gene Ther. 1998, 5, 930-937.
(
34) Liu, Y.; Mounkes, L. C.; Liggitt, H. D.; Brown, C. S.; Solodin, I.;
Heath, T. D.; Debs, R. J . Factors Influencing the Efficiency of
Cationic Liposome-Mediated Intravenous Gene Delivery. Nat.
Biotechnol. 1997, 15, 167-173.
J M000466S