Amphiphilic polyether branched molecules to increase the circulation time of cationic particles

Article abstract of DOI:10.1016/j.bmc.2007.02.037

The preparation, physicochemical and biological properties of amphiphilic polyether branched molecules is described. These 'bunch shaped' molecules when inserted into cationic liposomes/DNA complexes have shown efficient surface charge shielding. As a con

Full text of DOI:10.1016/j.bmc.2007.02.037

Bioorganic & Medicinal Chemistry 15 (2007) 3176–3186
Amphiphilic polyether branched molecules to increase the
circulation time of cationic particles
Marie Garinot, Nathalie Mignet,* Celine Largeau, Johanne Seguin,
Daniel Scherman and Michel Bessodes*
Inserm, U640, Paris F-75006, France
CNRS, UMR8151, Paris F-75006, France
Universite´ Rene´ Descartes, Faculte´ de Pharmacie, Chemical and Genetic Pharmacology Laboratory, Paris F-75270, France
Ecole Nationale Supe´rieure de Chimie de Paris, Paris F-75005, France
Received 25 May 2006; revised 14 February 2007; accepted 20 February 2007
Available online 22 February 2007
Abstract—The preparation, physicochemical and biological properties of amphiphilic polyether branched molecules is described.
These ‘bunch shaped’ molecules when inserted into cationic liposomes/DNA complexes have shown efficient surface charge shield-
ing. As a consequence they efficiently inhibited the non specific interactions with blood components and significantly enhanced cir-
culation time of the particles in the blood track. Formulations containing these molecules compared positively with those containing
PEG lipids, providing a 5-fold increase in circulation time.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
inflammatory reactions.6 They are also rapidly cleared
from the blood track as they interact with cell membrane
and circulating proteins.7,8 Efforts to alleviate this prob-
lem have led to considerable amount of work. For in-
stance, coating of conventional neutral liposomes with
polyethylene glycol polymers (PEG) has been extensively
described to increase their circulation time.9–11 It is
thought that this neutral, hydrophilic polymer provides
a shielding of the particle surface that protects them from
blood proteins and macrophages. These have been
named ‘stealth liposomes’.12,13 Transposition of this
technology to the cationic complexes was very tempting
and has been proposed.14 It has been found however that
PEG coating of cationic liposomes did not completely in-
hibit the non-specific interactions mentioned above, and
that despite this treatment, these particles were rapidly
eliminated through the intervention of complement pro-
teins and macrophages.15 Furthermore, those PEG-
coated cationic lipoplexes that could escape this event,
and eventually make it through the target cell mem-
branes, could not release their therapeutic contents into
the cytoplasm.16 We have recently described the synthe-
sis and properties of new pH labile PEG lipids that are
hydrolyzed in slightly acidic environment, such as that
occurring in inflammatory or tumor tissues.17 However,
the problem of short circulation time was still not
resolved. An alternative to the use of PEG lipids in the
Administration of therapeutic DNA to correct genetic
diseases has received considerable attention these last
15 years.1–4
Cationic lipid complexes with DNA have been widely de-
scribed;5 these so-called cationic lipoplexes strongly
interact by electrostatic neutralization, thus providing
good protection and compaction of DNA. They present
however severe drawbacks to their use in human therapy:
the cationic lipids that enter into composition are toxic to
the cell and the living body, as they generate strong
Abbreviations: BOP, benzotriazolylhexamethylphosphoramide hexa-
fluorophosphate; DODA, dioctadecylamine; DOPE, dioleylphospha-
tidylethanolamine; GFP, green fluorescent protein; PEG,
polyethyleneglycol; TEA, triethylamine; TFA, trifluoroacetic acid;
THF, tetrahydrofuran; TLC, thin layer chromatography.
Keywords: Liposome/DNA complexes; Neutral amphiphilic lipid;
Surface charge shielding; Inhibition of nonspecific interactions; Blood
circulation time enhancement; Synthesis; Physical chemistry; Biolog-
ical properties.
*
´
Corresponding authors at present address: Unite de Pharmacologie
´ ´
´
Chimique et Genetique, U640/UMR8151, Faculte des Sciences
Pharmaceutiques et Biologiques, 4, avenue de l’Observatoire, 75270
Paris 06, France. Tel.: +33 01 53 73 9581/9566; e-mail addresses:
nathalie.mignet@univ-paris5.fr; michel.bessodes@univ-paris5.fr
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.02.037

M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
3177
formulation of cationic lipoplexes, in view of their
systemic administration, would then be of great interest.
character, well-defined aspect, optimized interfaces,
etc. Dendrimers consist of branched, wedge-like struc-
tures called dendrons that are attached to a multivalent
core and emerge radially toward the periphery. The syn-
thesis of such a structure requires successive reactions of
activation and coupling. For our design, we chose a tri-
valent core with one function dedicated to attach the
hydrophobic moiety, the other two allowing the growth
of the polyether.
We proposed the hypothesis that amphiphilic neutral
structures, containing ramified polyether moieties and a
lipid part, could advantageously replace the PEG. The
bunch shaped structures could provide efficient shielding
of the lipoplex surface charges, whereas their non-poly-
meric nature would not over-stabilize the complexes.
These molecules would then protect cationic lipoplexes
during blood circulation, thus increasing their plasmatic
half-life. This would provide a better chance for the par-
ticle to reach the therapeutic target, either by passive tar-
geting or through the interaction of a ligand specific of
target cells’ membrane receptors. Destabilization in the
cytosol should be facilitated by the non-polymeric struc-
ture and the lower molecular weight of these molecules as
compared to conventional PEG lipids.
Divergent18 and convergent syntheses19 have been de-
scribed for dendritic molecules: In the divergent route
the molecule grows from the center core to the outside,
the number of coupling reaction increasing with each
generation a high reactivity of the terminal groups is re-
quired. As hydroxyl groups are not highly reactive, we
´
favored the convergent route for the polyether part. Fre-
chet et al. have first described the convergent synthesis
of a new family of dendrimers with an aliphatic poly-
ether backbone.20 We have adapted this early procedure
to the synthesis of a branched polyether structure corre-
sponding to the hydrophilic part of the final molecule,
with a unique functionalized focal point (Fig. 1).
We have prepared a family of molecules containing a
bunch shaped hydrophilic structure attached to a lipidic
moiety, with different hydrophilic densities. We describe
here the synthesis, physicochemical and biological prop-
erties of these new molecules.
We took advantage of the two electrophilic sites of
methallyl dichloride reagent to grow the polyether part.
The alkylation reaction was performed according to a
procedure developed earlier by us using potassium
hydroxide and catalytic crown ether in tetrahydrofu-
ran.21 The yields varied from 74% (four benzyl terminal
groups [1]) to 66% (eight benzyl terminal groups [3]) as a
consequence of increasing sterical hindrance. Subse-
quent hydroxylation of the olefinic focal point was
achieved by regioselective hydroboration using
2. Results and discussion
2.1. Chemistry
The first step of the synthesis consisted in building the
hydrophilic branched part. In the past few decades, den-
dritic architectures have been highly investigated
because of their unique properties such as multivalent
O
O
O
O
O
O
O
O
O
O
O
O
O
O
i
ii, iii
OH
HO
1,3-dibenzyloxy
propan-2-ol
[1]
[2]
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
HO
i
ii, iii
[3]
[4]
Figure 1. Synthesis of the branched polyether by the convergent method. (i) KOH, 18-Cr-6, dichloromethallyle, THF, 20 ꢁC; (ii) BH3/THF, 0 ꢁC; (iii)
3 M NaOH, 30% H2O2, 20 ꢁC.

3178
M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
borane:tetrahydrofuran complex followed by oxidation
with hydrogen peroxide.
range (Table 1). Formulation of the particles including
the branched polyether lipids with a (gly)3 spacer (com-
pounds [12] and [14]) was relatively straight, only few
minutes of ultrasonication were necessary to obtain an
average size around 200 nm. The same size range was
obtained with the lipid tetraol [11] without drastic treat-
ment, whereas it needed 75 min of sonication to form
the particles including the DODA-octaol [13]. The diam-
eter of these particles still remained larger than those of
the other formulations, this might be due to the large
hydrophilic octaol moiety.
The hydrophobic part was composed of dioctadecyl-
amine (DODA) with or without a three-glycine spacer,
previously used in our laboratory to anchor polyethyl-
ene glycol moiety to liposomes.22 The triglycine spacer
is thought to allow a better presentation of the polyether
part on the lipoplex surface and a higher lipid insertion
in the hydrophobic lipid bilayer.
The molecules with the peptidic spacer were obtained by
coupling the Boc-protected tripeptide (Boc-NH-Gly-
Gly-Gly-COOH) with the DODA using BOP reagent
and triethylamine to give compound [5]. Subsequent
deprotection by trifluoroacetic acid (TFA) gave [6]
(Fig. 2).
When associated with DNA, at a charge ratio (N/P)
equal to 4, colloidal stabilization was achieved for all
the formulations, leading to suitable particles for sys-
temic injection.
Association of DNA to the four original formulations
was checked and compared to the RPR209120/DOPE.
Using the fluorescence of PicoGreenꢂ associated
with non complexed DNA as an index of DNA acces-
sibility (i.e. characterizing non-compacted DNA), we
observed no differences between the four formulations
containing compounds [11]–[14] when half of the fluo-
rescence was measured. This showed that the affinity
of RPR209120 to DNA was not affected by the tetra-
ol or octaol compounds (Fig. 5). A slight difference
was obtained at N/P charge ratios from 1 to 3, which
might be related to the modified accessibility of DNA
due to the large hydrophilic head of the four branched
lipids, but this might also be due to the unstability of
the colloids (aggregation due to the low lipid/DNA
ratio).
The coupling reaction between, respectively, the two
hydrophobic parts (DODA and [6]) and the two
branched molecules ([2] and [4]) was achieved by first
reacting the hydroxyl core of the dendritic molecule with
trichloromethyl chloroformate.23 The chloroformate
thus formed reacted spontaneously in the presence of
base with the amino group of the hydrophobic anchors
to give the corresponding carbamate derivatives. Prod-
ucts [7], [8], [9], and [10] were obtained in yields varying
between 35% and 70%. Deprotection of the peripheric
hydroxyls was performed by catalytic hydrogenolysis
(Pd/C) of the benzyl ethers as the last step to give [11],
[12], [13], and [14] (Fig. 3).
2.2. Physico-chemical characterization
2.2.1. DNA compaction. The newly synthesized lipids
were incorporated in a formulation as a 1:1 molar ratio
with the lipopolyamine RPR20912024 (Fig. 4) and com-
pared to RPR209120/DOPE 1:1 previously described.25
When cationic charges were used in excess (ratio P 4),
DNA was not accessible to the PicoGreenꢂ probe in
any of the formulations, thus confirming its efficient
compaction.
In the absence of DNA, the cationic particles formed by
the thin film method exhibited a diameter in the 200 nm
2.2.2. Surface charges. After showing that incorporation
of polyether branched lipids in cationic particles did not
O
O
HN
H3C CH2
H3C CH2
H3C CH2
H3C CH2
O
17
O
17
H
H
iv
N
N
O
NH
N
O
+
N
H
NH
O
O
17
17
O
[5]
HN
O
OH
H3C CH2
H3C CH2
O
O
17
H
N
NH+3
v
N
N
H
O
17
[6]
Figure 2. Synthesis of the hydrophobic part with the triglycine spacer. (iv) BOP, Et3N, CHCl3, 20 ꢁC; (v) TFA, 20 ꢁC.

M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
3179
OH
OH
[11]
O
N
O
O
O
OH
OH
OH
OH
O
O
O
O
H
H
N
N
N
N
[12]
O
H
OH
O
O
OH
OH
OH
O
OH
O
OH
O
N
[13]
O
O
O
OH
OH
O
O
OH
OH
OH
O
OH
O
OH
OH
O
O
O
H
H
N
N
[14]
N
N
O
O
H
O
O
OH
OH
O
O
OH
OH
Figure 3. Amphiphilic branched molecules synthesized and studied for their physicochemical and biological ability to stabilize DNA/cationic lipid
complexes.
NH+3
H2
N+
O
N+
H
N
4
Cl
N
H3N+
O
H
Figure 4. Cationic lipid (RPR209120) used in lipoplex preparations.
interfer with their ability to interact and associate with
DNA, we focused on their potential charge masking
effect (Table 1). First, the effect of the spacer between
the lipidic chain and the hydrophilic head was evaluated.

3180
M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
Table 1. Particle diameter (nm) and zeta potential (mV) of formulations containing the branched polyether lipids [11]–[14] and the formulation
RPR209120/DOPE, in the absence or presence of DNA (CR = charge ratio of 4) in NaCl 150 mM (mean SE of 3 determinations)
Formulation
ꢀDNA
+DNA (CR = 4)
Zeta (mV)
Size (nm)
Zeta (mV)
Size (nm)
RPR209120/DOPE
169 3.78
126 2.33
393 73.8
179 2.04
237 7.26
+74.1 11
+77.1 4.4
+48.9 8.3
+56 2.8
155 10.2
189 9.53
334 46.3
212 3.46
203 26.4
+41.6 2.5
+49 5.1
RPR209120/DODA-tetraol [11]
RPR209120/DODA-octaol [13]
RPR209120/DODA-(gly)3-tetraol [12]
RPR209120/DODA-(gly)3-octaol [14]
+30.7 12
+38.8 16.9
+23.5 4.9
+55.4 3.5
100
80
60
40
20
0
2.2.3. Particle stability in biological medium. Based on
the above results we incubated in serum containing
medium the cationic particles composed of the spacer
containing polyether branched lipids (compounds [12]
and [14]), as compared to RPR209120/DOPE
formulation.
DODA-(gly)3-tetraol
DODA-(gly)3-octaol
DODA-tetraol
DODA-octaol
As reported in Table 2, the cationic formulation taken as
a reference as well as the formulation containing the
polyether with the smallest hydrophilic head group [12]
aggregated rapidly in the presence of serum, thus sug-
gesting seric proteins induced colloidal destabilization.
On the opposite, the larger polyether branched lipid
[14] increased significantly the stability of the particles.
Colloidal destabilization started only after 2 h of incuba-
tion at 37 ꢁC, while the two other formulations already
aggregated after 15 min. This again suggests a better
masking effect of the large polyether derivative at the
particle surface, which would in turn prevent non-spe-
cific interactions with seric proteins. Moreover this result
is in correlation with the zeta potential decrease observed
for this formulation. From these data, we could expect
that formulations bearing the octaol derivative [14]
might present an increased circulation time, as compared
to RPR209120/DOPE particles.
0
2
4
6
8
ratio (+/-)
Figure 5. Fluorescence of DNA associated PicoGreen for formula-
tions including RPR209120 and the polyether branched lipids or
DOPE, and DNA at different charge ratios.
When comparing formulations containing the two lipids
bearing the smallest polyether part (DODA-tetraol [11]
and DODA-(gly)3-tetraol [13]), we observed that the
presence of the triglycine spacer in [13] led to a signifi-
cant zeta potential decrease (D = 30 mV) compared to
the cationic reference RPR209120/DOPE.
Then, the effect of the hydrophilic head of the polyether
branched lipid on the particle zeta potential was evalu-
ated ([11] vs [13] and [12] vs [14]). Comparison of the
zeta potential of the cationic formulations containing
DOPE or the two lipids without spacer indicated that
the octaol moiety was more efficient in reducing the zeta
potential than the tetraol part. This effect was equally
observed with the two lipids bearing the triglycine spacer
[12] and [14].
2.3. In vitro results
2.3.1. Transfection efficiency. In order to evaluate the
structure/activity relationship, we were interested in
the transfection efficiency of the formulations bearing
polyether branched lipids containing a triglycine spacer
and bearing a small or large head group (compounds
[12] and [14]). For this purpose, plasmid encoding the
luciferase reporter gene was associated, at different cat-
ionic lipid/DNA ratios, to the cationic formulations
containing the polyether lipids and compared to the cat-
ionic RPR209120/DOPE.
From these data, one might assume that the polyether
branched lipid bearing the larger head is the most effi-
cient to mask the cationic charges of the particles and
should be able to reduce the non-specific interactions
with circulating proteins. Also, the presence of a spacer
between the lipid and the head might help exposing the
hydrophilic part on the surface of the particles, and thus
more efficiently mask the cationic charges.
As observed in Figure 6, the three different ratios tested
showed a decrease in the transfection efficiency for the
formulations containing the polyether branched lipids
[12] or [14] as compared to the cationic control. Even
Table 2. Particle size (nm) after incubation in cell medium containing 10% fetal bovine serum
RPR209120/DOPE
RPR209120/DODA-(gly)3-tetraol
RPR209120/DODA-(gly)3-octaol
0
169
179
237
15 min
1 h
2 h
896.3
2000
2000
955.3
2000
2000
334.6
370.1
749.1

M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
3181
[14] would insure better shielding of the particle surface,
thus reducing particle to cell membrane interactions.
DOPE/209120
DODA-(gly)3-tetraol/209120
DODA-(gly)3-octaol/209120
10000000
1000000
100000
10000
1000
2.3.2. Correlation between particle internalization and
protein expression. To demonstrate that particle surface
shielding was the reason for transfection efficiency
decrease, we performed cell internalization experiments.
To this aim DOPE-Rhodamine (0.2%) was incorporated
in the different formulations as a fluorescent marker.
100
Comparison after 24 h, of internalizations of the cationic
lipid/DOPE versus cationic lipid/tetraol [12] versus cat-
ionic lipid/octaol [14], was straightforward as can be seen
in Figure 7. The particles without branched lipid (DOPE/
RPR209120) were internalized into B16 cells as large
aggregates (upper left side). A large amount of lipid
was incorporated by the cells and GFP encoding plasmid
was expressed. The use of tetraol [12] into the formula-
tion did not significantly modify the pattern of particle
internalization. Aggregates as well as punctuations were
observed inside the cells, and approximately the same
amount of cells expressed GFP (Fig. 7), which was in cor-
relation with luciferase expression quantification (Fig. 6).
10
1
2
4
8
charge ratio (+/-)
Figure 6. Transfection of B16 cells performed with the formulations
DOPE/RPR209120, DODA-(gly)3-tetraol [12]/RPR209120, and
DODA-(gly)3-octaol [14]/RPR209120 at charge ratios 2, 4, and 8.
though this effect was more pronounced with the octaol
as compared to the tetraol, these particles still induced a
good transfection level. This result was also in agree-
ment with the zeta potential and serum resistance data
as already mentioned. Indeed, increasing the hydrophilic
moiety of the lipid from tetraol [12] to octaol [14]
reduced the zeta potential of the particles and increased
their stability in serum. These data led us to conclude
that the larger bunch-shaped structure in the case of
In opposite, a significant difference in particle internali-
zation was observed when the octaol [14] was incorpo-
rated into the formulation. Even though we did not
quantify it, the amount of internalized particle was
clearly reduced by the presence of the octaol lipid [14].
Scarce isolated punctuations were observed at the cellu-
lar level, confirming the stability of this formulation in
Figure 7. Internalization experiments on B16 cells. Fluorescence microscopy was performed 24 h after the transfection. (a) DOPE/RPR209120/
DOPE-Rh, 1:1:0.2%; pCMU-GFP plasmid was mixed at a charge ratio (+/ꢀ) of 8 (b) DODA-(gly)3-tetraol [12]/RPR209120/DOPE-Rh, 1:1:0.2% ;
pCMU-GFP plasmid was mixed at a charge ratio (+/ꢀ) of 8 (c) DODA-(gly)3-octaol [14]/RPR209120/DOPE-Rh, 1:1:0.2% ; pCMU-GFP plasmid
was mixed at a charge ratio (+/ꢀ) of 8.

3182
M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
the cell medium. Moreover, GFP expression was
observed in a very small amount of cells. Hence, low
transgene expression was well correlated with poor
particle internalization, which tended to confirm our pri-
mary hypothesis of the efficient shielding provided by
the octaol lipid compared to the tetraol derivative.
tion time, PEG does not fulfill the needs. In the present
paper, we have described the synthesis and the biological
study of four different molecules, composed of a double
alkyl chain (insertion in the formulation), with or with-
out a triglycine spacer and a bunch shaped polyether
structure varying in size. We have incorporated these
molecules into cationic formulations and have obtained
stable DNA compacting complexes. We have assumed
that the presence of a spacer and a large polyether hydro-
philic part would improve the shielding of the surface
charges of the particles, thus decreasing the cationic
complexes non specific interactions. According to these
data, internalization and transfection was considerably
decreased with particles bearing the largest triglycine
derivative. The tripeptide spacer supposedly facilitated
the organization of the branched structure at the com-
plex surface. In addition, the larger the hydrophilic part
was, the more efficient was the charge shielding, leading
to a large decrease of non specific interactions. More-
over, these particles were still able to transfect as shown
with in vitro transfection experiments.
2.4. Biodistribution studies
To further evaluate the efficacy of cationic particle
shielding by the octaol lipid [14], we performed biodistri-
bution studies on mice. The lipopolyamine formulations
containing the DODA-(gly)3-tetraol and DODA-(gly)3-
octaol lipids were compared to the cationic control
RPR209120/DOPE or to the same formulation contain-
ing 5% DODA-(gly)3-PEG5000. The different cationic
particles were injected to the mice tail vein, and blood
was collected 30 min after injection. The data expressed
as the percentage of the injected dose indicated signifi-
cant differences between the formulations (Fig. 8).
Indeed, 8% and 22% were recovered when, respectively,
tetraol and octaol polyether lipids were incorporated in
the formulation, while only 2.7% of the cationic particles
were recovered in total blood. A 3-fold enhancement in
blood circulation was thus obtained with the adjunction
of the DODA-(gly)3-tetraol lipid, whereas an 8-fold fac-
tor was obtained with the DODA-(gly)3-octaol lipid. It is
noteworthy that only 10% of the injected dose was recov-
ered in the blood when DODA-(gly)3-PEG5000 was
added to the formulation, which is comparable to the
level found with the RPR209120/DODA-(gly)3-tetraol
particles, and below the 22% obtained with the octaol
lipid. It is also interesting to note that blood cells interac-
tion was significantly reduced in the case of branched lip-
ids, corroborating the shielding effect of these lipids.
However, no passive accumulation could be observed
in the 3LL tumor implanted in mice flank.
With an octaol moiety branched to a lipid moiety
through a triglycine linker, we have presently obtained
an original formulation with increased blood circula-
tion. Indeed, as compared to PEG, the insertion of the
triglycine octaol [14] in the formulation limits the cellu-
lar internalization and results in a longer blood circula-
tion time. Active targeting could then be achieved either
by coupling a ligand on the terminal hydroxyls or by
using a small amount of PEG-lipid-ligand. Work is in
progress to this aim.
3. Experimental
3.1. Materials and methods
Until now, the main strategy developed in non viral gene
delivery with lipoplexes to avoid toxicity and non specific
interactions was by coating the particles with PEG. As
we mentioned before, despite an increased blood circula-
The chemicals were obtained from Sigma–Aldrich. The
solvents were purchased from SDS and were of synthesis
grade. Analytical TLC was run on Merck silica-pre-
coated aluminum plates and silica for preparative
Associated to blood cells
30
Circulating
**
20
10
0
*
Figure 8. Distribution of the cationic particles in the blood 30 min after injection. Percentage of the injected dose was evaluated by extraction of
rhodamine-lipid in the blood after systemic injection. Statistics have been performed with Kruskal–Wallis test on two separate experiments including
three mice each. Significant differences were found between RPR209120/DOPE and RPR209120/DODA-(gly)3-tetraol, *P < 0.05, RPR209120/
DOPE and RPR209120/DODA-(gly)3-octaol, **P < 0.01.

M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
3183
chromatography was also purchased from Merck. Prod-
ucts were revealed by a 50/50 solution of 1 M iodine and
1 M H2SO4 spray. RPR209120 was synthesized as
described before.26 Dioleylphosphatidylethanolamine
(DOPE) was purchased from Avanti Polar, Picogreenꢂ
from Molecular Probes. Compounds were named using
MDL AutoNom software according to IUPAC rules,
for the sake of clarity trivial names have been defined
and appear between inverted commas. NMR spectra
were obtained from a Bruker Avance instrument.
3.1.1.3. 3-[3-(2-Benzyloxy-1-benzyloxymethyl-ethoxy)-
2-(2-benzyloxy-1-benzyloxymethyl-ethoxymethyl)-prop-
oxy]-2-[3-(2-benzyloxy-1-benzyloxymethyl-ethoxy)-2-
(2-benzyloxy-1-benzyloxymethylethoxymethyl)-prop-
oxymethyl]-propene. ‘Octabenzyl allyl’ [3]. The title
compound was prepared from compound [2] as de-
scribed for product [1] (yield 66%). 1H NMR
(400 MHz, CDCl3): d (ppm) 2.15 (m, 2H: CH); 3.3–3.6
(m, 36H: CH2AO); 3.82 (s, 4H: CH2AC@); 4.5 (s,
16H: CH2ABz); 5.05 (s, 2H: CH2@C); 7.2–7.4 (m,
40H: CH aromatic). 13C NMR (75 MHz, CDCl3): d
(ppm) 41 (CH), 68.8 (CH2AO), 70 (CH2AO), 72
(CH2AO), 73.5 (CH2ABz), 78.5 (CHAO), 106 (CH@),
128 (CH), 129 (CH), 132.5 (CH), 138 (C).
3.1.1. Chemistry
3.1.1.1. 3-(2-Benzyloxy-1-benzyloxymethyl-ethoxy)-2-
(2-benzyloxy-1-benzyloxymethyl-ethoxymethyl) propene.
‘Tetrabenzyl allyl’ [1]. To a stirred solution of 1,3-diben-
zyloxypropan-2-ol (1 g, 3.67 mmol) in tetrahydrofuran
(4 mL), fine potassium hydroxide powder (2 g,
36.7 mmol), 18-crown-6 (97 mg, 0.367 mmol), and
methallyl dichloride (193.2 lL, 1.67 mmol) were added.
The mixture was stirred at room temperature for 2 h.
A large volume of dichloromethane was added and the
mixture was washed three times with brine. The organic
phase was then dried over Na2SO4, filtered, and evapo-
rated to dryness. Chromatography of the residue on sil-
ica gel eluted with cyclohexane/ethyl acetate (7:3, v/v)
gave a viscous colorless oil (737.6 mg, 1.24 mmol, yield
3.1.1.4. 3-[3-(2-Benzyloxy-1-benzyloxymethyl-eth-
oxy)-2-(2-benzyloxy-1- benzyloxymethyl-ethoxymethyl)-
propoxy]-2-[3-(2-benzyloxy-1-benzyloxymethyl-ethoxy)-
2-(2-benzyloxy-1-benzyloxymethylethoxymethyl)-prop-
oxymethyl]-propanol. ‘Octabenzyl alcohol’ [4]. The title
compound was prepared from compound [3] as de-
scribed for product [2] (yield 50%). 1H NMR
(400 MHz, CDCl3): d (ppm) 2.05 (m, 1H: CH); 2.15
(m, 2H: CH); 3.3–3.7 (m, 38H: CHAO, CH2AO,
CH2AOH); 4.5 (s, 16H: CH2ABz); 7.2–7.4 (m, 40H:
CH aromatic). 13C NMR (75 MHz, CDCl3): d (ppm)
41 (CH), 42 (CH), 64 (CH2AOH), 68.8 (CH2AO), 70
(CH2AO), 70.3 (CH2AO), 72 (CH2AO), 73,4
(CH2ABz), 78 (CHAO), 127.6–128.4 (CH), 139 (C).
1
74%) corresponding to the expected product. H NMR
(400 MHz, CDCl3):
d (ppm) 3.52–3.62 (m, 8H:
CH2AO); 4.69–4.75 (m, 2H: CHAO); 4.15 (s, 4H:
CH2AO); 4.5 (s, 8H: CH2ABz); 5.2 (s, 2H: CH2@C);
7.2–7.4 (m, 20H: CH aromatic). 13C NMR (75 MHz,
CDCl3): d (ppm) 70.2 (CH2AO), 71 (CH2AO), 73
(CH2ABz), 77 (CHAO), 106 (CH@), 128 (CH), 129
(CH), 132.5 (CH), 138 (C).
3.1.1.5. ({[(Dioctadecylcarbamoylmethyl-carbamoyl)-
methyl]-carbamoyl}-methyl)-carbamic acid tert-butyl
ester. ‘DODA-(gly)3-Boc’ [5]. To a stirred solution of
Boc-N-gly-gly-gly-OH (1 g, 3.5 mmol) and TEA
(1.44 mL, 10.4 mmol) in chloroform (70 mL), dioctade-
cylamine (1.8 g, 3.5 mmol) was added and the mixture
was stirred until solubilization. BOP reagent (1.7 g,
3.8 mmol) was then added and the mixture was stirred
at room temperature for 4 h. It was then successively
washed with 0.5 M KHSO4 (4 · l20 mL), saturated
NaHCO3 (4 · 20 mL), and finally with brine. The organ-
ic layer was dried over MgSO4, filtered, and evaporated
to dryness. The residue was washed with acetonitrile,
centrifuged, and dried to obtain the expected product
3.1.1.2. 3-(2-Benzyloxy-1-benzyloxymethyl-ethoxy)-2-
(2-benzyloxy-1-benzyloxymethyl-ethoxymethyl) propane.
‘Tetrabenzyl alcohol’ [2]. The olefin [1] (500 mg,
0.84 mmol) was stirred in anhydrous tetrahydrofuran
in a round-bottomed flask under argon and maintained
at 0 ꢁC. A solution of boran/THF complex (1 M,
922 lL, 0.922 mmol) was then slowly added and the
mixture was stirred at 0 ꢁC for 5 h. A 3 M solution of
sodium hydroxide (1.26 mL, 3.77 mmol) and 30%
hydrogen peroxide (285 lL, 2.51 mmol) were then cau-
tiously added. The mixture was stirred at room temper-
ature for 30 min, saturated with K2CO3, and extracted
with ether. The organic layer was dried over anhydrous
Na2SO4 and the solvent was removed under vacuum.
Purification of the residue by chromatography on silica
gel with cyclohexane/ethyl acetate (7:3 v/v) gave a vis-
cous colorless oil (288 mg, 0.47 mmol, yield 66%) corre-
1
as a white powder (2.18 g, 2.75 mmol, yield 78%). H
NMR (300 MHz, CDCl3): d (ppm) 0.9 (t, J = 6.5 Hz,
6H: CH3); 1.28 (s, 64H: CH2); 1.47 (s, 9H: CH3); 3.16
(t, J = 8 Hz, 2H, CH2); 3.32 (t, J = 8 Hz, 2H: CH2);
3.89 (d, J = 7.6 Hz, 2H: CH2); 4.03 (dd, J1 = 5.2 Hz,
J2 = 8.4 Hz, 4H: CH2); 5.15 (s broad, 1H: NH); 6.76
(t, J = 5.2 Hz, 1H: NH); 7.96 (t, J = 4.8 Hz, 1H: NH).
13C NMR (75 MHz, CDCl3): d (ppm) 14 (CH3ACH2),
23 (CH2ACH3), 27 (CH2ACH2), 28.5 (CH3AC), 29.5
(CH2), 30 (CH2ACH2), 32 (CH2), 41.2 (CH2ANH), 43
(CH2ANH), 46.5 (CH2ANH), 47 (CH2AN), 167
(C@O), 168.5 (C@O), 170 (C@O).
1
sponding to the expected product. H NMR (400 MHz,
CDCl3): d (ppm) 2.11 (m, 1H: CH); 3.51 (m, 8H:
CH2AO); 3.6–3.8 (m, 8H: CHAO, CH2AO, CH2AOH);
4.52 (s, 8H: CH2ABz); 7.2–7.4 (m, 20H: CH aromatic).
13C NMR (75 MHz, CDCl3): d (ppm) 41.3 (CH), 63
(CH2AO), 69.5 (CH2AO), 70 (CH2AOH) 73 (CH2ABz),
78 (CHAO), 126.5–128.8 (CH), 137.8 (C).
3.1.1.6. 2-Amino-N-[(dioctadecylcarbamoylmethyl-
carbamoyl)-methyl]-acetamide; TFA salt. ‘DODA-(gly)3;
TFA salt’ [6]. The Boc-protected product [5] (2 g,
2.52 mmol) was deprotected at room temperature in tri-
fluoroacetic acid (10 mL) for 2 h. TFA was evaporated
under vacuum. The expected product was obtained
IR: t 1100 cmꢀ1 (CAOAC), 2858 cmꢀ1 (CH2ABz),
3460 cm (OH). MS (DCI): m/z = 632 MNH4þ, m/z =
ꢀ1
615 MH+, m/z = 290 (C17H20O3 + NH4)+.

3184
M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
(1.87 g, 2.32 mmol, yield 92%). 1H NMR (400 MHz,
CDCl3): d (ppm) 0.9 (t, J = 6.5 Hz, 6H: CH3); 1.28 (s,
64H: CH2); 3.14 (t broad, 4H, CH2); 3.5–3.6 (q,
J = 5.2 Hz, 2H: CH2); 3.8 (d, J = 4.8 Hz, 2H: CH2);
3.87 (d, J = 5.6 Hz, 2H: CH2); 7.93 (s, 3H: NH); 8.02
(t, J = 5.2 Hz, 1H: NH); 8.56 (t, J = 5.2 Hz, 1H: NH).
MS (DCI): m/z = 693 MH+.
6.77 (t, J = 1.96 Hz, 1H: NH); 6.99 (t broad, 1H: NH);
7.25 to 7.35 (m, 20H, CH). 13C NMR (75 MHz, CDCl3):
d
(ppm) 14.3 (CH3ACH2), 23 (CH2ACH3), 27
(CH2ACH2), 30 (CH2ACH2), 32 (CH2), 40.6 (CH2),
42 (CH2AN), 44.3 (CH2ANH), 46.2 (CH2ANH), 47
(CH2ANH), 62 (CH2AO), 67 (CH2AO), 69 (CH2AO),
72 (CH2ABz), 77 (CHAO), 126–128 (CH aromatic),
166 (C@O), 168 (C@O), 170 (C@O).
3.1.1.7. Dioctadecyl-carbamic acid 3-(2-benzyloxy-1-
benzyloxymethyl-ethoxy)-2-(1-benzyloxymethyl-2-ben-
zyloxy-ethoxymethyl)-propyl ester. ‘DODA-tetrabenzyl’
[7]. To a stirred solution of compound [2] (150 mg,
0.24 mmol) in toluene (15 mL), trichloromethylchloro-
formate (295 lL, 2.44 mmol) was added. The mixture
was heated to 65 ꢁC and stirred for 4 h. Successive evap-
orations under reduced pressure with toluene (3 · 5 mL)
and diethyl ether (3 · 5 mL) gave a colorless oil. The res-
idue was dissolved in THF (2 mL) and TEA (342 lL,
2.4 mmol) was added. Dioctadecylamine (116 mg,
0.22 mmol) and TEA (342 lL, 2.4 mmol) were solubi-
lized in THF (2 mL) and added to the previous mixture.
The solution was stirred at room temperature for 3 days
and solvents were evaporated under vacuum. The resi-
due was dissolved in ethyl acetate and washed with
brine. The organic layer was dried over Na2SO4, filtered,
and concentrated. Chromatography of the crude residue
on silica gel in cyclohexane/ethyl acetate (8:2, v/v) gave a
colorless oil (93 mg, 0.08 mmol, yield 37%) of the
expected product. 1H NMR (400 MHz, CDCl3): d
(ppm) 0.89 (t, J = 6.4 Hz, 6H: CH3); 1.26 (s, 64H:
CH2); 2.25 (m, 1H: CH); 3.1 (s broad, 2H: CH2); 3.2
(s broad, 2H: CH2); 3.5–3.7 (m, 14H: CH, CH2); 4.16
(d, J = 6 Hz, 2H: CH2); 4.5 (s, 8H: CH2); 7.24–7.34
(m, 20H, CH).
3.1.1.9. Dioctadecyl-carbamic acid 3-[3-(2-benzyloxy-
1-benzyloxymethyl-ethoxy)-2-(1-benzyloxymethyl-2-hydro-
xy-ethoxymethyl)-propoxy]-2-[3-(2-benzyloxy-1-benzyl-
oxymethyl-ethoxy)-2-(1-benzyloxymethyl-2-hydroxy-eth-
oxymethyl)-propoxymethyl]-propyl ester. ‘DODA-octa-
benzyl’ [9]. The title compound was prepared similarly
to product [7] starting from compound [4] (68 mg,
1
0.036 mmol, yield 74%). H NMR (400 MHz, CDCl3):
d (ppm) 0.89 (t, J = 6.4 Hz, 6H: CH3); 1.26 (s, 64H:
CH2); 2.25 (m, 3H: CH); 3.1 (s broad, 2H: CH2); 3.2
(s broad, 2H: CH2); 3.5–3.7 (m, 35H: CH, CH2); 4.16
(d, J = 6 Hz, 2H: CH2); 4.5 (s, 16H: CH2); 7.24–7.34
(m, 40H, CH).
3.1.1.10. ({[(Dioctadecylcarbamoylmethyl-carbamoyl)-
methyl]-carbamoyl}-methyl)-carbamic acid 3-[3-(2-ben-
zyloxy-1-benzyloxymethyl-ethoxy)-2-(1-benzyloxymeth-
yl-2-hydroxy-ethoxymethyl)-propoxy]-2-[3-(2-benzyloxy-
1-benzyloxymethyl-ethoxy)-2-(1-benzyloxymethyl-2-hydro-
xy-ethoxymethyl)-propoxymethyl]-propyl ester. ‘DODA-
(gly)3-octabenzyl’ [10]. The title compound was prepared
similarly to product [8] but using [4] (52 mg, 25.7 lmol,
1
yield 40%). H NMR (600 MHz, CDCl3): d (ppm) 0.89
(t, J = 6.4 Hz, 6H: CH3); 1.26 (s, 64H: CH2); 2.25 (m,
3H: CH); 3.1 (s broad, 2H: CH2); 3.2 (s broad, 2H:
CH2); 3.5–3.6 (m, 36H: CH, CH2); 3.85 (s broad, 2H,
CH2); 3.98 (s broad, 2H, CH2); 4.16 (m, 4H: CH2); 4.5
(s, 16H: CH2); 5.3 (t broad, 1H: NH); 6.77 (t,
J = 1.96 Hz, 1H: NH); 6.99 (t broad, 1H: NH); 7.24–
7.34 (m, 40H, CH).
3.1.1.8. ({[(Dioctadecylcarbamoylmethyl-carbamoyl)-
methyl]-carbamoyl}-methyl)-carbamic acid 3-(2-benzyl-
oxy-1-benzyloxymethyl-ethoxy)-2-(1-enzyloxymethyl-2-
benzyloxy-ethoxymethyl)-propylester. ‘DODA-(gly)3-tet-
rabenzyl’ [8]. To a stirred solution of product [2]
(580 mg, 0.94 mmol) in toluene (15 mL), trichlorometh-
ylchloroformate (456 lL, 3.77 mmol) was added. The
mixture was heated to 65 ꢁC and stirred for 4 h. Succes-
sive evaporation under reduced pressure with toluene
and diethyl ether gave a colorless oil. The residue was
immediately dissolved in THF (10 mL). The product
[6] (758.7, 0.94 mmol) and diisopropylamine (1.64 mg,
9.4 mmol) were dissolved in THF (10 mL) and added
to the previous mixture. The solution was stirred at
room temperature for 24 h. Solvents were evaporated
under vacuum, the residue was dissolved in ethyl acetate
and washed with brine. The organic layer was dried
over Na2SO4, filtered, and concentrated to a yellow
oil. Chromatography of the crude residue on silica gel
in cyclohexane/ethyl acetate (2:8, v/v) gave a white
gum (350 mg, 0.26 mmol, yield 28%) of the expected
product. 1H NMR (400 MHz, CDCl3): d (ppm) 0.88
(t, J = 7 Hz, 6H: CH3); 1.26 (s, 64H: CH2); 2.2 (m,
1H: CH); 3.15 (t, J = 7.6 Hz, 2H: CH2); 3.3 (t,
J = 7 Hz, 2H: CH2); 3.5–3.7 (m, 14H: CH, CH2); 3.8
(d, J = 4.6 Hz, 2H, CH2); 3.97 (d, J = 4.3 Hz, 2H,
CH2); 4 (d, J = 2.8 Hz, 2H: CH2); 4.2 (d, J = 5.6 Hz,
2H: CH2); 4.51 (s, 8H: CH2); 5.3 (t broad, 1H: NH);
3.1.1.11. Dioctadecyl-carbamic acid 3-(2-hydroxy-1-
hydroxymethyl-ethoxy)-2-(2-hydroxy-1-hydroxymethyl-
ethoxymethyl)-propyl ester. ‘DODA-tetraol’’ [11]. The
product [7] (90 mg, 77.4 lmol) was dissolved in ethyl
acetate (4 mL) and Pd on activated carbon (10%)
(20 mg, 19 lmol) was added. The mixture was placed
under H2 at atmospheric pressure and stirred at room
temperature for 1 h. The solution was then filtered and
concentrated. The expected product (50 mg, 62 lmol,
1
yield 81%) was obtained as a colorless viscous oil. H
NMR (400 MHz, CDCl3): d (ppm) 0.9 (t, J = 6.4 Hz,
6H: CH3); 1.26 (s, 64H: CH2); 2.26 (m, 1H: CH); 3.2
(s broad, 4H: CH2); 3.45–3.48 (m, 2H: CH); 3.63–3.78
(m, 12H: CH2); 4.24 (d, J = 6 Hz, 2H: CH2).
3.1.1.12. ({[(Dioctadecylcarbamoylmethyl-carbamoyl)-
methyl]-carbamoyl}-methyl)-carbamic acid 3-(2-hydroxy-
1-hydroxymethyl-ethoxy)-2-(2-hydroxy-1-hydroxymethyl-
ethoxymethyl)-propyl ester. ‘DODA-(gly)3-tetraol’ [12].
The product [8] (100 mg, 75 lmol) was dissolved in
methanol (8 mL) and Pd on activated carbon (10%)
(18 mg, 17 lmol) was added. The mixture was placed
under H2 at atmospheric pressure and stirred at room

M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
3185
temperature overnight. The solution was then filtered
and concentrated. The expected product (45 mg,
46.2 lmol, yield 62%) was obtained as a white powder.
300 mM) was added dropwise to liposomes (250 lL,
32 mM of total lipids, H2O) with constant vortexing.
The charge ratio was calculated stoichiometrically as
mole ratio of cationic lipid RPR209120 (3 positive
charges per molecule) to DNA nucleotide residue (M
330, 3 nmol of phosphate in 1 lg of DNA).
1H NMR (400 MHz, CDCl3):
d (ppm) 0.88 (t,
J = 7 Hz, 6H: CH3); 1.26 (s, 64H: CH2); 2.2 (m, 1H:
CH); 3.15 (t, J = 7.6 Hz, 2H: CH2); 3.3 (t, J = 7 Hz,
2H: CH2); 3.5–3.7 (m, 14H: CH, CH2); 3.8 (d,
J = 4.6 Hz, 2H, CH2); 3.97 (d, J = 4.3 Hz, 2H, CH2); 4
(d, J = 2.8 Hz, 2H: CH2); 4.2 (d, J = 5.6 Hz, 2H: CH2);
5.3 (t broad, 1H: NH); 6.77 (t, J = 1.96 Hz, 1H: NH);
6.99 (t broad, 1H: NH).
3.1.2.3. Light scattering experiments. Diameter and
zeta potential values of the complexes RPR209120/
DOPE or lipids [11]–[14] DNA were determined with
dynamic light scattering instruments (respectively, Mal-
vern Zetasizer Nano Series and Malvern Zetasizer
3000HSA). For size measurement, 5 lL of complexes
was diluted in 900 lL of milliQ water filtered with
0.2 lm syringe filter and for the zeta potential 200 lL
of complexes was diluted in 2 mL NaCl 20 mM.
3.1.1.13. Dioctadecyl-carbamic acid 3-[3-(2-hydroxy-
1-hydroxymethyl-ethoxy)-2-(2-hydroxy-1-hydroxymethyl-
ethoxymethyl)-propoxy]-2-[3-(2-hydroxy-1-hydroxymeth-
yl-ethoxy)-2-(2-hydroxy-1-hydroxymethyl-ethoxymeth-
yl)-propoxymethyl]-propyl ester. ‘DODA-octaol’ [13].
The title compound was prepared similarly to product
[11] but using [9] (yield 64%). 1H NMR (600 MHz,
CDCl3): d (ppm) 0.89 (t, J = 6.4 Hz, 6H: CH3); 1.26 (s,
64H: CH2); 2.15 (m, 3H: CH); 3.23 (s broad, 4H:
CH2); 3.4 to 3.7 (m, 36H: CH, CH2); 4.02 (s broad,
2H: CH2).
3.1.2.4. Evaluation of plasmid DNA compaction with
PicoGreenꢂ. Compaction of DNA in the liposomes con-
taining the polyether derivatives was verified by addition
of Picogreenꢂ (100 lL of Picogreenꢂ diluted 200-fold in
trisborate-EDTA 1 N for 40 ng of DNA) and measure-
ment of the fluorescence (kexc = 450 50 nm ; kem =
530 25 nm) which decreases when DNA is compacted.
3.1.1.14. ({[(Dioctadecylcarbamoylmethyl-carbamoyl)-
methyl]-carbamoyl}-methyl)-carbamic acid 3-[3-(2-hydro-
xy-1-hydroxymethyl-ethoxy)-2-(2-hydroxy-1-hydroxy-
methyl-ethoxymethyl)-propoxy]-2-[3-(2-hydroxy-1-hydroxy-
methyl-ethoxy)-2-(2- hydroxy-1-hydroxymethyl-etho-
xymethyl)-propoxymethyl]-propyl ester. ‘DODA-(gly)3-
octaol’ [14]. The title compound was prepared similarly
3.1.2.5. Stability in fetal bovine serum. A solution of
cationic liposomes RPR209120/DOPE or branched lip-
ids [11]–[14] (120 nmol) was incubated at 37 ꢁC in
400 lL of serum (MEM Gibco, L-glutamine 2 mM, pen-
icillin 50 U/mL, streptomycin 50 U/mL, and 10% of fetal
bovine serum). At different times (15 min, 1 h, and 2 h)
50 lL aliquots of this solution were diluted in 900 lL
of filtered water before light scattering analysis.
1
to product [12] but using [10] (yield 50%). H NMR
(300 MHz, CDCl3): d (ppm) 0.89 (t, J = 6.4 Hz, 6H:
CH3); 1.26 (s, 64H: CH2); 2.25 (m, 3H: CH); 3.1 (s
broad, 2H: CH2); 3.2 (s broad, 2H: CH2); 3.5–3.6
(m, 36H: CH, CH2); 3.8 (s broad, 2H, CH2); 3.98 (s
broad, 2H, CH2); 4.1 (m, 4H: CH2); 4.16 (s large, 2H:
CH2).
3.1.2.6. Cell culture. B16 cells were grown in Mini-
mum Essential Medium (MEM Gibco) supplemented
with 10% (v/v) fetal bovine serum (FBS), L-glutamine
2 mM, penicillin 50 U/mL, streptomycin 50 U/mL, and
a solution of non essential amino acids. Cultures were
maintained at 37 ꢁC in a 5% CO2/air incubator.
3.1.2. Physicochemistry and biology
3.1.2.1. Preparation of liposomes bearing the branched
polyether lipids or DOPE. The cationic lipid RPR209120
(2.10 mg, 2.5 lmol) and DOPE (1.86 mg, 2.5 lmol) were
dissolved with chloroform in a round-bottomed flask
(5 mL) and dried under reduced pressure until 5 mbar
to form a thin film in the flask. After 2 h at this pressure,
the film was hydrated overnight with distilled and fil-
tered (0.22 lm) water (500 lL) to obtain a final concen-
tration of 10 mmol/L. The liposomes bearing the
polyether branched lipids [11]–[14] were formulated
according to the same procedure replacing DOPE by
molecules [11]–[14]. Liposomes were then sonicated to
obtain a diameter in the 200 nm range. The size was
determined by dynamic light scattering with a Malvern
Zetasizer Nano Series.
3.1.2.7. Lipofection. The day before the experiment,
B16 cells were seeded into 24-well culture plates at a
density of 50,000 cells per well. They were incubated
at 37 ꢁC, under 5% CO2 in MEM for 24 h. Just before
transfection, cells were washed twice with fresh medium
with or without FBS. The lipoplex solutions prepared as
described above at least 30 min before lipofection were
added to the cells (0.5 lg of plasmid per well). 10% (v/
v) FBS was added to the culture wells containing ser-
um-free lipofection medium 2 h later at 37 ꢁC. The cells
were incubated for 24 h at 37 ꢁC in the presence of 5%
CO2.
3.1.2.8. Luciferase assay. Cells were washed with PBS
and lysed with 200 lL of cell culture lysis reagent (Pro-
mega). Luciferase expression was quantified on 5 lL of
centrifuged lysate supernatant using a luciferase assay
kit (Promega). Light emission was measured using a
luminometer (Multilabel counter 1420 Victor2; EG&G
Wallac), equipped with a co-injector that delivered
80 lL of luciferase substrate into 40 lL of cell extracts.
Relative light units (RLU) were calculated versus back-
3.1.2.2. Preparation of lipoplexes. Lipoplexes were
prepared in NaCl (150 mL) at different charge ratios
of cationic lipid/DNA. Plasmid DNA used contained
either the luciferase or the GFP reporter gene under
the cytomegalovirus (CMV) promoter and was prepared
as described.27 To prepare 500 lL of complexes at
charge ratio equal to 8, DNA (250 lL, 2 lg/mL, NaCl

3186
M. Garinot et al. / Bioorg. Med. Chem. 15 (2007) 3176–3186
4. Anderson, D. M.; Hall, L. L.; Ayyalapu, A. R.; Irion, V.
R.; Nantz, M. H.; Hecker, J. G. Hum. Gene Ther. 2003, 14,
191–202.
ground activity. Light emission was normalized to the
protein concentration of cell extracts, determined using
the Bradford protein assay kit (Bio-Rad).
5. Pitard, B.; Oudrhiri, N.; Vigneron, J.-P.; Hauchecorne,
M.Aguerre, O.; Toury, R.; Airiau, M.; Ramasawmy, R.;
Scherman, D.; Crouzet, J.; Lehn, J.-M.; Lehn, P. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 2621–2626.
6. Yang, J.-P.; Huang, L. Gene Ther. 1997, 4, 950–960.
7. Simberg, D.; Weisman, S.; Talmon, Y.; Faerman, A.;
Shoshani, T.; Barenholz, Y. J. Biol. Chem. 2003, 278,
39858–39865.
8. Liu, F.; Qi, H.; Huang, L.; Liu, D. Gene Ther. 1997, 4,
517–523.
9. Woodle, M. C.; Matthay, K. K.; Newman, M. S.;
Hidayat, J. E.; Collins, L. R.; Redemann, C.; Martin, F.
J.; Papahadjopoulos, D. Biochim. Biophys. Acta 1992,
1105, 193–200.
10. Awasthi, V. D.; Garcia, D.; Goins, B. A.; Phillips, W. T.
Int. J. Pharm. 2003, 253, 121–132.
11. Needham, D.; Kim, D. H. Colloids Surf., B 2000, 18, 183–
195.
12. Harasym, T. O.; Tardi, P.; Longman, S. A.; Ansell, S. M.;
Bally, M. B.; Cullis, P. R.; Choi, L. S. Bioconjugate Chem.
1995, 6, 187–194.
13. Uster, P. S.; Allen, T. M.; Daniel, B. E.; Mendez, C. J.;
Newman, M. S.; Zhu, G. Z. FEBS Lett. 1996, 386, 243–
246.
3.1.2.9. Internalization experiments. The day before
the experiment, B16 cells were seeded into 12-well cul-
ture plates containing lamella at a density of 80,000 cells
per well. They were incubated at 37 ꢁC, under 5% CO2 in
MEM supplemented with 10% (v/v) fetal bovine serum
(FBS), L-glutamine 2 mM, penicillin 50 U/mL, strepto-
mycin 50 U/mL, and a solution of non essential amino
acids, for 24 h. Wells were washed and MEM was added
(transfection with serum). The lipoplexes (0.5 lg of GFP
plasmid/well) were added to the cells. 10% (v/v) FBS was
added to the culture wells containing serum-free lipofec-
tion medium after 2 h at 37 ꢁC. The cells were incubated
in the darkness, for 24 h at 37 ꢁC in the presence of 5%
CO2. Cells were washed with 500 lL PBS and then fixed
with paraformaldehyde (500 lL 3%). After 30-min incu-
bation in the darkness, cells were washed with 500 lL
PBS and DAPI solution (500 lL 0.1 lg/mL) was added.
Cells were incubated for 30 min in the dark, then washed
with PBS, water and mounted in mowiol prior examina-
tion. Slides were then analyzed with a Zeiss Axiophot
microscope equipped with a Zeiss Neofluor 100· objec-
tive lens.
14. Harvie, P.; Wong, F. M. P.; Bally, M. B. J. Pharm. Sci.
2000, 89, 652–663.
15. Song, L. Y.; Ahkong, Q. F.; Rong, Q.; Wang, Z.; Ansell,
S.; Hope, M. J.; Mui, B. Biochim. Biophys. Acta 2002,
1558, 1–13.
16. Shi, F.; Wasungu, L.; Nomden, A.; Stuart, M. C. A.;
Polushkin, E.; Engberts, J. B. F. N.; Hoekstra, D.
Biochem. J. 2002, 366, 333–341.
17. Masson, C.; Garinot, M.; Mignet, N.; Wetzer, B.; Mailhe,
P.; Scherman, D.; Bessodes, M. J. Controlled Release
2004, 99, 423–434.
18. Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos,
G.; Martin, S.; Roeck, J.; Ryder, J.; Smith, P. Polym. J.
1985, 17, 117–132.
3.1.2.10. In vivo distribution in tumor-bearing mice.
Experiments were performed on five-week-old female
C57Bl/6 mice. Fifteen days before the experiment,
100 lL (105 cells/100 lL) of 3LL cellular solution (Lewis
lung carcinoma tumors) was implanted subcutaneously
in the right flank of the mice. The mice were anesthe-
tized by intraperitoneal injection with a mix of ketamine
(85.8 mg/kg; Centravet) and xylazine (3.1 mg/kg; Bayer)
diluted in 150 mM NaCl. A 200 lL volume of rhoda-
mine-labeled complexes (100 nmol total lipids in 5% glu-
cose) was injected into the mouse tail vein. Blood was
collected by cardiac puncture 30 min after injection.
´
19. Hawker, C. J.; Frechet, J. M. J. J. Chem. Soc. Chem.
Commun. 1990, 1010–1013; Hawker, C. J.; Frechet, J. M.
´
J. J. Am. Chem. Soc. 1990, 112, 7638–7641.
20. Jayaraman, M.; Frechet, J. M. J. A. J. Am. Chem. Soc.
1998, 12996–12997.
21. Bessodes, M.; Shamsazar, J.; Antonakis, K. Synthesis
1988, 7, 560–562.
22. Nicolazzi, C.; Mignet, N.; De la Figuera, N.; Cadet, M.;
Ibad, R. T.; Seguin, J.; Scherman, D.; Bessodes, M.
J. Controlled Release 2003, 88, 429–443.
23. Kurita, K.; Iwakura, Y. Org. Synth. 1980, 59, 195–
201.
Rhodamine-labeled lipids were extracted according to
Takeuchi et al.28 from 100 lL of blood with 3 mL chlo-
roform/methanol (1:1, v/v) by vigorous mixing during
30 min followed by centrifugation (3000 rpm, 10 min).
The fluorescence intensity was assayed on 100 lL super-
natant with a luminometer Wallac Victor2 (1420 Multi-
label Coulter, 570 nm). The amount of complexes in the
blood was evaluated with a calibration curve and ex-
pressed as the remaining percentage of injected dose.
24. Byk, G.; Dubertret, C.; Escriou, V.; Frederic, M.; Jaslin,
G.; Rangara, R.; Pitard, B.; Crouzet, J.; Wils, P.;
Schwartz, B.; Scherman, D. J. Med. Chem. 1998, 41,
229–235.
References and notes
25. Tranchant, I.; Thompson, B.; Nicolazzi, C.; Mignet, N.;
Scherman, D. J. Gene Med. 2004, S24–S35.
26. Byk, G.; Scherman, D.; Schwartz, B.; Dubertet, C. PCT
Int. Appl. WO 9718185, 1997, 81 pp.
27. Soubrier, F.; Cameron, B.; Manse, B.; Somarriba, S.;
Dubertret, C.; Jaslin, G.; Jung, G.; Caer, C. Le.; Dang, D.;
Mouvault, J. M.; Scherman, D.; Mayaux, J. F.; Crouzet,
J. Gene Ther. 1999, 6, 1482–1488.
1. Rosenberg, S. A.; Aebersold, P.; Cornetta, K.; Kasid, A.;
Morgan, R. A.; Moen, R.; Karson, E. M.; Lotze, M. T.;
Yang, J. C.; Topalian, S. L. N. Engl. J. Med. 1990, 323,
570–578.
2. Edelstein, M. L.; Abedi, M. R.; Wixon, J.; Edelstein, R.
M. J. Gene Med. 2004, 6, 597–602.
3. Goyenvalle, A.; Vulin, A.; Fougerousse, F.; Leturcq, F.;
Kaplan, J-C.; Garcia, L.; Danos, O. Science 2004, 306,
1796–1799.
28. Takeuchi, H.; Kojima, H.; Yamamoto, H.; Kawashima,
Y. J. Controlled Release 2000, 68, 195–205.