Aminoglycoside-derived cationic lipids for gene transfection: Synthesis of kanamycin A derivatives

Article abstract of DOI:10.1002/ejoc.200300164

Cationic lipids are currently actively investigated as an alternative approach to recombinant viruses for gene transfer studies and gene therapy applications. Basically, they rely on the formation of lipid/DNA aggregates via electrostatic interactions between their cationic headgroup and the negatively charged DNA. The development of new amphiphilic structures should allow to shed light on their still poorly understood structure/activity relationship and thereby help to design improved vectors. It appears that aminoglycosides, which are natural polyamines known to bind to nucleic acids, provide a favourable scaffold for the synthesis of a variety of cationic lipids because of their structural features and multifunctional nature. The synthesis and full characterization of a series of lipophilic derivatives of the aminoglycoside antibiotic kanamycin A, mainly kanamycin-cholesterol conjugates, are reported herein. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200300164

FULL PAPER
Aminoglycoside-Derived Cationic Lipids for Gene Transfection: Synthesis of
Kanamycin A Derivatives
Matthieu Sainlos,[a] Philippe Belmont,[a] Jean-Pierre Vigneron,*[a] Pierre Lehn,[b] and
Jean-Marie Lehn[a]
Keywords: Cationic lipids / Gene transfer / Nonviral vectors / Aminoglycosides
Cationic lipids are currently actively investigated as an al-
ternative approach to recombinant viruses for gene transfer
which are natural polyamines known to bind to nucleic acids,
provide a favourable scaffold for the synthesis of a variety of
studies and gene therapy applications. Basically, they rely on cationic lipids because of their structural features and multi-
the formation of lipid/DNA aggregates via electrostatic inter-
actions between their cationic headgroup and the negatively
charged DNA. The development of new amphiphilic struc-
functional nature. The synthesis and full characterization of
a series of lipophilic derivatives of the aminoglycoside antibi-
otic kanamycin A, mainly kanamycin−cholesterol conjugates,
tures should allow to shed light on their still poorly under- are reported herein.
stood structure/activity relationship and thereby help to de-
sign improved vectors. It appears that aminoglycosides,
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
ever, although cationic liposomes have already been used in
clinical gene therapy trials for cancer[7] and cystic fibrosis,[8]
there is a clear requirement for improved, newer generation
lipids, especially because of the relatively low gene transfer
efficiency of the current lipids when compared with viral
vectors. Thus, a major research effort is still devoted to the
discovery of new reagents and delivery systems able to carry
nucleic acids into the cell nucleus. As a part of our work
aimed at discovering such compounds,[9] we report here a
new approach based on the use of an aminoglycoside scaf-
fold.
Gene therapy consists in delivering nucleic acids into so-
matic cells of patients to treat both genetic and acquired
diseases.[1] Some positive results have already been reported,
but the clinical trials performed to date have clearly indi-
cated that a prerequisite for successful gene therapy is the
use of an efficient and safe gene delivery system.[2] Al-
though viral vectors have been shown particularly efficient
for gene transfer,[3] they suffer from a number of incon-
veniences including immunogenicity, toxicity, safety issues
and problems in industrial quality control and upscaling.[2]
To overcome these difficulties, alternative approaches have
been developed, using cationic polymers or cationic lipids.[4]
Spontaneous formation of self-assembled vector/DNA
complexes is in both cases due to electrostatic interactions
between the positively charged vector and the negatively
charged phosphate groups of DNA. Cationic lipids, which
are composed of a cationic headgroup linked via a spacer
to a hydrophobic moiety and are mostly formulated as lipo-
somes with neutral helper lipids such as dioleoylphosphati-
dylethanolamine (DOPE), are particularly attractive be-
cause it is relatively easy to synthesize a great variety of
reagents with favourable features. Thus, numerous cationic
lipids have been developed over the last decade.[5a,6] How-
Aminoglycoside antibiotics constitute a large family of
clinically important drugs used in the treatment of gram-
negative infections.[10] They effect their antibacterial activity
by interfering with ribosomal function (via binding to
rRNA), which ultimately results in the disruption of protein
biosynthesis.[11] These antibiotics have also been shown to
interact with a variety of other RNA molecules such as
group I introns,[12] hammerhead ribozymes,[13] and the
HIV-1Јs TAR[14] and RRE[15] regulatory domains. Aminog-
lycosides have in fact been shown to be rather non specific
RNA binders that recognize numerous similar three-dimen-
sional RNA structures and have therefore become an im-
portant starting point for the study of RNAϪsmall mol-
ecules recognition.[16]
Most of the naturally occurring aminoglycosides are
structurally characterized by amino sugars glycosidically
linked to an aminocyclitol which, in most cases, is 2-deoxy-
streptamine. There are actually several types of 2-deoxy-
streptamine derivatives: monosubstituted derivatives (such
as neamine), 4,5-disubstituted (neomycin type derivatives),
[a]
´
Laboratoire de Chimie des Interactions Moleculaires, CNRS
`
UPR 285, College de France,
11 Place Marcelin Berthelot, 75231 Paris cedex 05, France
Fax: (internat.) ϩ33-1-44271356
E-mail: jean-pierre.vigneron@college-de-france.fr
[b]
ˆ
´
INSERM U458, Hopital Robert Debre, AP-HP,
´
48 Bd Serurier, 75019 Paris, France
2764
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200300164
Eur. J. Org. Chem. 2003, 2764Ϫ2774

Aminoglycoside-Derived Cationic Lipids for Gene Transfection
FULL PAPER
and 4,6-disubstituted (kanamycin, tobramycin, and genta- guanidinium groups are key to natural and synthetic trans-
mycin derivatives). Aminoglycosides carry up to six amino porters; arginine oligomers and artificial peptoids have in-
groups, which are predominantly charged at physiological deed recently been shown to easily enter cells and tissues,
pH,[17] and bind with high affinity to anions and nucleic thus enhancing uptake of drugs.[24] Finally, it should be
acids via electrostatic and hydrogen bonding interactions.[18] stressed that, in KanaChol as well as in TGKC, the amino-
It has recently been shown that the 1,3-hydroxyamine mo- glycoside headgroup is linked to the cholesterol moiety via a
tifs often found in their structure interact strongly with carbamoyl bond. Indeed, carbamoyl bonds, which also link
both the phosphodiester backbone and the Hoogsteen face together the headgroup and the lipid moiety in numerous
of guanine.[19] Taken together, these features make aminog- cationic lipids including the widely used DC-Chol and our
lycosides very promising as versatile frameworks for the own BGTC,[9a] have been reported to be both chemically
synthesis of a variety of novel cationic lipids for gene trans- stable and biodegradable.[9aϪ9c,25]
fection into eukaryotic cells. Indeed, selective acylation of
Transfection studies showed that reagent KanaChol
one or several of their functional groups should provide could mediate efficient gene transfection into a variety of
aminoglycosides with the lipophilic properties required for mammalian cell lines when used either alone or as a liposo-
their use as vectors for gene transfer. In addition, the struc- mal formulation with the neutral phospholipid dioleoylpho-
tural diversity of aminoglycosides also provides a means to sphatidylethanolamine (DOPE).[26] Cationic TGKC/DOPE
investigate in detail the structure/activity relationships of liposomes were also efficient for in vitro transfection. Fi-
these new transfection reagents by allowing to evaluate the nally, kanamycin-cholesterol/DOPE lipoplexes were found
influence of specific features on transfection efficiency (nat- to be also efficient for gene transfection into the mouse air-
ure and number of the hydrophobic subunits, geometry of ways in vivo via intranasal instillation.[26] Therefore, in a
the molecules, number of positive charges, etc.).
second step of our work, in order to explore the relation-
ships between structure and transfection activity of kana-
mycin-derived lipids and so to develop more efficient vec-
tors, we prepared a series of new kanamycin-lipid conju-
gates with a spacer and/or a hydrophobic anchor different
from those in KanaChol. We report herein the synthesis
and the full characterization of these different compounds.
Along these lines, we have first prepared a cationic chol-
esterol derivative, KanaChol-6Ј, simply termed KanaChol
herein, characterized by a kanamycin A polar headgroup.
It was expected that, because of its three free amine groups,
this molecule would bind to and condense DNA at physio-
logical pH and that the cholesterol subunit would in ad-
dition facilitate the cellular uptake of the lipoplexes gener-
ated. Moreover, we have also synthesized reagent TGKC
(TriGuanidiniumϪKanamycinϪCholesterol), the fully gu-
anidinylated derivative of KanaChol.
Results and Discussion
KanaChol and TGKC
To test our approach for designing new vectors for gene
transfer, we first chose to prepare a kanamycin A derivative
because this molecule contains only one aminomethyl
group and thus a selective monoacylation of this amino
group seemed easily feasible. Indeed, within the four amine
functions present in kanamycin A, the 6Ј position is known
to be the most reactive (6Ј-NH2 Ͼ 1-NH2 Ͼ 3-NH2 Ͼ 3ЈЈ-
NH2).[27] Therefore we tried a direct reaction between kana-
mycin A and cholesteryl chloroformate that gave an insep-
arable complex mixture. We were thus led to undertake a
longer synthesis involving the selective protection of the 6Ј-
NH2 group by an efficient selective acylating agent, the pro-
tection of the amine functions in position 1, 3, and 3ЈЈ, then
the deprotection of the 6Ј-NH2 group followed by its reac-
tion with the cholesteryl chloroformate and, finally, the de-
protection of the amine functions in position 1, 3, and 3ЈЈ.
The different approaches reported for the selective pro-
tection of amino groups of aminoglycosides have been re-
Indeed, we,[9aϪ9d] as well as others,[20] have previously ex- viewed.[27] Among them, procedures based on the difference
ploited the favourable features of the guanidinium group in reactivity of the amino groups towards weak acylating
for DNA binding and recent reports have indicated that agents seem the most attractive. Particularly, the reagent,
conversion of aminoglycosides to their fully guanidinylated N-(Boc-O)-5-norbornene-endo-2,3-dicarboximide,
which
derivatives[21] or to polyarginine conjugates[22] greatly en- greatly favours reaction at unhindered amines, was claimed
hances their RNA affinity. This beneficial effect may be all ideally suited for application to aminoglycoside chemis-
the more useful considering that aminoglycosides bind try.[28] Indeed, using the commercially available Cbz deriva-
double-stranded DNA with low affinity.[18d,23] In addition, tive, instead of the Boc one, we obtained the monoprotected
Eur. J. Org. Chem. 2003, 2764Ϫ2774
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2765

M. Sainlos, P. Belmont, J.-P. Vigneron, P. Lehn, J.-M. Lehn
FULL PAPER
1
compound 2 with a 73% yield after chromatographic purifi-
cation over silica (Scheme 1). The second step of the syn-
thesis involved the full protection of the three remaining
amines.
Table 1. H Chemical shifts and peak assignments for compounds
1 and 2
[b]
H atom
2[a]
1[a]
∆
1
2a
2e
3
4
5
6
1Ј
2Ј
3Ј
4Ј
5Ј
6Јa
6Јb
1ЈЈ
2ЈЈ
3ЈЈ
4ЈЈ
5ЈЈ
6ЈЈ
2ЈЈЈ
4ЈЈЈ; 8ЈЈЈ
5ЈЈЈ; 7ЈЈЈ
6ЈЈЈ
2.86
1.10
1.83
2.65
3.18
3.56
3.23
5.13
3.55
3.64
3.25
3.75
3.59
3.26
5.01
3.52
3.03
3.35
3.90
3.76
5.09
7.43
7.43
7.43
2.83
1.16
1.90
2.83
3.25
3.60
3.19
5.27
3.52
3.64
3.24
3.71
2.93
2.71
4.98
3.44
2.94
3.26
3.85
3.71
0.03
Ϫ0.06
Ϫ0.07
Ϫ0.18
Ϫ0.07
Ϫ0.04
0.04
Ϫ0.14
0.03
0.00
0.01
0.04
0.66
0.55
0.03
0.08
0.09
0.09
0.05
0.05
[a] Chemical shift values (δ) are given in ppm. Assignments for close
lying signals are tentative and may be interchanged. [b] ∆ ϭ δ2 Ϫ δ1.
Table 2. 13C chemical shifts and peak assignments for compounds
1 and 2
[b]
C atom
2[a]
50.7
1[a]
52.9
∆
1
Ϫ2.2
Ϫ2.7
Ϫ2.0
Ϫ1.0
Ϫ2.2
Ϫ2.9
Ϫ1.1
Ϫ2.1
Ϫ2.2
Ϫ2.1
Ϫ3.3
Ϫ2.1
Ϫ2.2
Ϫ2.5
Ϫ2.1
Ϫ2.5
Ϫ2.1
Ϫ2.2
2
3
4
5
35.1
49.4
88.7
74.4
87.4
100.8
72.2
73.2
71.4
72.0
41.9
100.3
71.8
54.6
69.2
72.4
60.5
158.7
67.3
37.8
51.4
89.7
76.6
90.3
101.9
74.3
75.4
73.5
75.3
44.0
102.5
74.3
56.7
71.7
74.5
62.7
Scheme 1. Synthesis of KanaChol: (a) N-Benzyloxycarbonyloxy-5-
norbornene-2,3-dicarboximide, Et3N, DMSO/H20 (1:1), 4 days,
73%; (b) 2-(Trimethylsilyl)ethyl p-nitrophenyl carbonate, Et3N, di-
oxane/H2O (2:1), 55 °C, 48 h, 73%; (c) H2, Pd/C(10%), CH3OH,
93%; (d) Cholesteryl chloroformate, Et3N, THF/DMF (5:2), room
temp., 48 h, 77%; (e) Trifluoroacetic acid, 0 °C, 40 mn, 97%
6
1Ј
2Ј
3Ј
4Ј
5Ј
6Ј
1ЈЈ
2ЈЈ
3ЈЈ
Although the classical Boc protection was reported to
work,[29] it resulted in our hands in a complex mixture of
components bearing two to six Boc groups as shown by
MALDI-TOF mass spectroscopy experiments. On the con-
trary, the use of the protective reagent 2-(trimethylsilyl)ethyl 4ЈЈ
p-nitrophenyl carbonate, (Teoc-O-Np),[30] in slight excess at
5ЈЈ
6ЈЈ
55 °C, allowed a good retrieval of the tri-Teoc protected
1ЈЈЈ
compound 3 (73% yield). Then, the benzyloxycarbonyl
2ЈЈЈ
group of 3 was removed by a classical hydrogen reduction
on supported Pd/C (10%), giving the free aminomethyl de-
rivative 4 with yields up to 93%. The coupling of cholesteryl
chloroformate with compound 4 led to the conjugate 5
which was easily worked up (77% yield). Finally, the
3ЈЈЈ
136.8
129.2; 129.0
129.0; 128.8
128.6
4ЈЈЈ; 8ЈЈЈ
5ЈЈЈ; 7ЈЈЈ
6ЈЈЈ
[a] Chemical shift values (δ) are given in ppm. Assignments for close
kanamycinϪcholesteryl conjugate 6 (ϭ KanaChol) was ob- lying signals are tentative and may be interchanged. [b] ∆ ϭ δ2 Ϫ δ1.
2766  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2003, 2764Ϫ2774

Aminoglycoside-Derived Cationic Lipids for Gene Transfection
FULL PAPER
tained in high yield (97%) by treatment of the Teoc pro- sistent with the presence of a unique isomer. The compari-
tected compound 5 with trifluoroacetic acid. son between the chemical shifts of 2 and those of kanamy-
Although the order of reactivity of the amino groups of cin A free base 1 shows a downfield shift for the protons
kanamycin A seems well established,[27] we thought it neces- on the 6Ј position of a least 0.5 ppm which is not observed
sary to confirm the structure of 2, by spectroscopic meth- on any of the other methylene protons adjacent to an amino
ods, to make sure that the protecting group was on posi- group (3ЈЈ, 1, and 3). This difference in the chemical shifts
tion 6Ј.
signal indicates that compound 2 has only been acylated on
The complete study of compound 2 [and of kanamycin A the 6Ј position as expected.
free base (1) as a reference] through NMR spectroscopy in
To obtain the guanidinium derivative 8 of KanaChol,
D2O (1H, 13C, COSY, and HMQC on a BrukerAMX-400, named TGKC (TriGuanidiniumϪKanamycinϪCholesterol)
25 °C) enabled us to assign unambiguously almost all the by analogy with BGTC (BisGuanidiniumϪTrenϪ
signals for both protons and carbons (Tables 1 and 2, values Cholesterol)[9a]
, we used N,NЈ-bisBoc-NЈЈ-triflylguanid-
found for kanamycin A free base are consistent with those ine[33] as described by Goodman et al. and obtained the tri-
1
found in literature, see reference[31,32]). The H NMR spec- guanidino Boc-protected compound 7 as an off-white solid
trum of compound 2 revealed the presence of the character- (77%) which was subsequently treated with trifluoroacetic
istic groups of both the carbobenzyloxy moiety at chemical acid (60%yield) (Scheme 2). This last step was tricky be-
shifts (δ) of 5.09 (benzylic CH2) and 7.43 ppm (aromatic cause it is not easy to remove the six Boc groups without
protons) and the kanamycin A moiety. The latter was observing a certain amount of cleavage of the carbamate
mainly characterized by the anomeric hydrogens at chemi- function.
cal shifts of 5.01 (H-1ЈЈ) and 5.13 ppm (H-1Ј) and by the
two protons on the unsubstituted position of the deoxy-
streptamine ring (at 1.10 and 1.83 ppm). The integration
KanaCapChol, KanaLysChol, KanaLysDiChol, and
KanaSucDODA
ratio of the characteristic signals of the two moieties to-
gether with the mass spectrum results confirmed the pres- DiChol (12c) were obtained from the same protected amin-
ence of a mono-protected aminoglycoside. omethyl precursor 4, by reaction with the commercially
KanaCapChol (12a), KanaLysChol (12b), and KanaLys-
The presence of a unique signal in the carbonyl region available N-protected amino acylating agents 13aϪc fol-
(at 158.71 ppm) in the 13C NMR spectrum, as well as the lowed by straightforward deprotection and treatment with
fact that each carbon on the kanamycin A moiety has been cholesteryl chloroformate (Scheme 3).
assigned unambiguously, strongly support the hypothesis of
KanaSucDODA (12d) resulted from the reaction of 4
a unique isomer. Furthermore, the two dimensional spectra with N,N-dioctadecylsuccinamide 13d, prepared from suc-
and the coupling constants (data not shown) are fully con- cinic anhydride and dioctadecylamine[34] (Scheme 4). In all
cases, the final compound was isolated as its trifluoroacet-
ate salt by deprotection of the amine groups of the aminog-
lycoside fragment by trifluoroacetic acid.
During these synthetic studies, a secondary reaction
worth noting was observed. In some cases, mass spec-
troscopy studies of the resulting product of the catalytic
hydrogenation of N-benzyloxycarbonyl-protected amino-
glycoside revealed the presence, besides the desired com-
pound of mass M, of a derivative corresponding to M ϩ 12.
Most of the time this unexpected compound was present in
small quantities but, sometimes, it happened to be predomi-
nant. Indeed, it has previously been reported that a side-
reaction of N-alkylation of amino groups occurs during pal-
ladium-catalyzed hydrogenolytic cleavage of N-benzyloxy-
carbonyl protecting groups;[35] in the presence of palladium
catalyst and oxygen, alcohols, such as methanol, ethanol
and benzylic alcohol, are readily converted into the corre-
sponding aldehydes which combine with amino groups to
form Schiff bases; the latter giving alkylamines upon cata-
lytic hydrogenation. In our case, the formol generated by
the oxidation of methanol, necessary for the solubilization
of the aminoglycoside derivative, combines with the liber-
ated amino group to form an intermediate imine and the
favourable geometry of the aminoglycoside ring leads it
subsequently to react intramolecularly with the closest hy-
droxyl function on position 4Ј to form an oxazolidine ring.
Despite the difficulty of purifying a mixture of so similar
Scheme 2. Synthesis of TGKC: (a) N,NЈ-bis(tert-butoxycarbonyl-
NЈЈ-triflyl)guanidine, Et3N, dioxane/H2O (10:1), 4 days, 77%;
(b) Trifluoroacetic acid, 0 °C, 4 h, 60%
Eur. J. Org. Chem. 2003, 2764Ϫ2774
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2767

M. Sainlos, P. Belmont, J.-P. Vigneron, P. Lehn, J.-M. Lehn
FULL PAPER
Scheme 3. Synthesis of KanaCapChol, KanalysChol, and KanaLysDiChol. a) HOAt/EDC, 4, DMF; b) H2/Pd/C, room temp.; c) choles-
teryl chloroformate, Et3N, DMF/THF; d) CF3CO2H
Scheme 4. Synthesis of KanaSucDODA. a) HOAt/EDC, 4, DMF/CH2Cl2; b) CF3CO2H
2768
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2764Ϫ2774

Aminoglycoside-Derived Cationic Lipids for Gene Transfection
FULL PAPER
layed extraction and a reflector. It was operated at an accelerating
potential of 20 kV in reflector mode. The mass spectra shown rep-
resent an average over 256 consecutive laser shots (3 Hz repetition
rate). Peptides were used to calibrate the mass scale using the two
points calibration software 3.07.1 from PerSeptive Biosystems. The
analyte solutions were prepared in methanol or THF at a concen-
tration of 2 g·LϪ1. The matrix, 2,5-dihydroxybenzoic acid (2,5-
DHB) was from Sigma (France) and used without further purifi-
cation. It was dissolved in methanol or THF (15 g·LϪ1). Five
microliters of analyte solution were mixed with 50 µL of matrix
solution. 2 microliters of sodium iodide solution (5 g·LϪ1 in meth-
anol or THF) were added to induce cationisation in some experi-
ments. One microliter of the final solution was deposited onto the
sample stage and allowed to dry in air.
products, compound 12e has been isolated, besides 12b, in
a pure state and its structure was confirmed by NMR stud-
ies and mass results. Of note, this side reaction can be pre-
vented by adding water to methanol, removing oxygen from
the solvents and using as little as possible palladium during
the hydrogenolytic deprotection.
Conclusion
Using the natural antibiotic kanamycin A as starting ma-
terial, we have prepared and fully characterized a set of lip-
idic aminoglycoside derivatives. These various syntheses
were chiefly undertaken with a view to develop novel cat-
ionic lipids characterized by an aminoglycoside-based
headgroup for gene transfection. Indeed, it is generally
agreed that the nature of the positive headgroup plays a key
role in the transfection activity of a given cationic lipid, the
characteristics (such as the hydrophobicity/hydrophilicity
balance, length of the spacer arm, ...) of the entire lipid
molecule determining however its in toto efficiency. As indi-
cated above, transfection experiments have shown that re-
IR spectra were obtained on a Bruker Vector 22 FT-IR spec-
trometer. The microanalyses were performed at the Service De Mic-
´
roanalyse de l’Universite Pierre et Marie Curie (Paris).
6Ј-N-(Benzyloxycarbonyl)kanamycin A (2): N-Benzyloxycarbonyl-
oxy-5-norbornene-2,3-dicarboximide (Cbz-Nor, 1.29 g, 4.12 mmol,
1 equiv.) and triethylamine (575 µL, 4.12 mmol, 1 equiv.) were ad-
ded to a solution of kanamycin A free base (2 g, 4.12 mmol) in
water (20 mL) and dimethyl sulfoxide (200 mL). The mixture was
stirred at room temperature overnight, upon which it was concen-
agent KanaChol was indeed efficient for gene transfection trated in vacuo to dryness. Flash column chromatography of the
in vitro and into the mouse airways in vivo.[26] The transfec-
residue (silica gel, CH2Cl2/MeOH/NH4OH, 5:4:1) afforded the de-
sired product as a white solid (1.86 g, 72.9%). Rf ϭ 0.22 (CH2Cl2/
tion activity of the other kanamycin-derived lipids will be
reported elsewhere, the data obtained thus far suggesting
MeOH/NH4OH, 5:4:1). m.p. 179 °C dec. [α]2D5 ϭ ϩ9.73 (c ϭ 1.17,
H2O). 1H & 13C NMR: see Tables 1 and 2. IR (KBr, thin film):
that adequate modification of the spacer arm may favour-
νmax (cmϪ1) ϭ 3357 (br), 2925, 1699 (s), 1541, 1456, 1362, 1267,
ably impact on the transfection efficiency.[36]
1049 (s). MS: m/z MALDI-TOF [MHϩ] 619.3, [MNaϩ] 641.3.
FABϩ m/z [MHϩ] 619.2. C26H42N4O13·2H2O (654.6): calcd. C
47.70, H 7.08, N 8.56; found C, 47.75, H 7.09, N 8.24.
Experimental Section
1,3,3ЈЈ-Tris-N-(trimethylsilylethoxycarbonyl)-6Ј-N-(benzyloxy-
carbonyl)kanamycin A (3): To a solution of monoprotected kana-
mycin A (2) (0.373 g, 0.6 mmol) in water (5 mL) and dioxane
(12 mL) was added triethylamine (350µL, 2.5 mmol, 4 equiv.) and
2-(trimethylsilyl)ethyl p-nitrophenyl carbonate (Teoc-O-Np,
General Remarks: All commercially available chemicals were re-
agent grade and were used without further purification. Kanamy-
cin A free base was prepared from the corresponding monosulfate
salt (purchased from Aldrich Chemical Co) by use of Amberlite
0.717 g, 2.5 mmol, 4 equiv.). The mixture was stirred at 55 °C for
48 h, then the volatiles were removed in vacuo. The solid residue
was dissolved in ethyl acetate (40 mL) and washed subsequently
several times by a 1  sodium hydroxide solution (40 mL) and a
saturated sodium chloride aqueous solution. The organic layer was
concentrated to dryness and flash column chromatography of the
solid residue (CH2Cl2/MeOH, 91:9) afforded the desired product
as a white solid (0.465 g, 73.3%). Rf ϭ 0.31 (CH2Cl2/MeOH,
90:10). m.p. 222Ϫ223 °C dec. [α]2D5 ϭ ϩ5.60 (c ϭ 0.38, CH3OH).
1H NMR (300 MHz, CD3OD, 25 °C): δ ϭ 7.34 (s, 5 H, arene-H),
5.09 (m, 4 H, CH2-arene overlapping with two anomeric-H), 4.15
(m, 7 H, O-CH2-CH2 and sugar ring proton), 3.80Ϫ3.31 (over-
IRA 400 (OHϪ) strongly basic ion-exchange resin. Unless other-
wise indicated, all the reactions were performed at room tempera-
ture. Flash chromatography employed Merck silica gel [Kieselgel
60 (0.040Ϫ0.063 mm)]. Analytical TLC was performed with
0.2 mm silica-coated aluminium sheets, visualization by UV light
or by spraying either a solution of ninhydrin (0.3% in weight in n-
butanol containing 3% acetic acid in volume) or a iodine solution
(0.1  in 10% sulfuric acid aqueous solution). Melting points were
determined on an Electrothermal 9100 apparatus. 1H and 13C
NMR spectra were recorded on a Bruker Avance300 spectrometer.
MALDI/TOF stands for Matrix Assisted Laser Desorption Ioni-
sation coupled with Time Of Flight mass spectroscopy. MALDI lapping multiplets, 15 H, various ring protons), 3.19 (t app, J ϭ
mass spectra were recorded with a PerSeptive Biosystems Voyager 9.4 Hz, 1 H), 2.06 (partially resolved multiplet, J ϭ 12.6 Hz, 1 H,
Elite (Framingham MA, USA) time-of-flight mass spectrometer. 2-H equatorial), 1.53 (partially resolved multiplet, J ϭ 12.6 Hz, 1
This instrument is equipped with a nitrogen laser (337 nm), a de- H, 2-H axial), 1.00 (m, 6 H, CH2-Si), 0.06 (s, 27 H, CH3-Si) ppm.
Eur. J. Org. Chem. 2003, 2764Ϫ2774
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2769

M. Sainlos, P. Belmont, J.-P. Vigneron, P. Lehn, J.-M. Lehn
FULL PAPER
13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 158.69, 158.01, 157.53, trifluoroacetic acid (2 mL) in an ice-saline bath. After 40 min under
157.20, 136.92, 128.10, 127.62, 127.47, 101.34, 98.52, 84.34, 81.13, 0 °C, the mixture was concentrated to dryness. A foamy solid was
75.30, 73.28, 72.92, 72.65, 71.09, 70.68, 70.60, 68.21, 66.21, 62.71,
obtained after several co-evaporation with methanol. The solid
62.51, 60.94, 56.46, 50.71, 49.91, 41.17, 34.34, 29.28, 17.30, 17.27, residue was dissolved in water and subsequent lyophilyzation of the
17.16, Ϫ2.78, Ϫ2.82, Ϫ2.84 ppm. IR (KBr, thin film) νmax
(cmϪ1) ϭ 3335 (br), 2954, 1697 (vs), 1542 (s), 1457, 1419, 1251
(s), 1140, 1043 (s), 939. MS: m/z MALDI-TOF [MNaϩ] 1073.7,
[MKϩ] 1089.7.
solution afforded KanaChol as a fluffy white powder in 97% yield
(0.063 g). Rf ϭ 0.32 (CH2Cl2/MeOH/NH4OH, 5:4:1). m.p. 198 °C
dec. [α]2D5 ϭ ϩ0.68 (c ϭ 0.27, H2O). 1H NMR (300 MHz, CD3OD/
CDCl3, 25 °C): δ ϭ 5.40 (m, 1 H, 6-H from cholesteryl moiety),
5.28 (partially resolved multiplet, J ϭ 3.9 Hz, 1 H, anomeric pro-
ton), 5.11 (partially resolved multiplet, J ϭ 3.6 Hz, 1 H, anomeric
proton), 4.40 (m, 2 H), 3.93Ϫ3.29 (m, 15 H), 3.18 (t app, J ϭ
9.3 Hz, 1 H), 2.55 (partially resolved multiplet, J ϭ 12.3 Hz, 1 H),
2.34 (m, 2 H), 2.09Ϫ1.88 (m, 6 H), 1.65Ϫ0.88 (m, 33 H, various
sugar ring and cholesteryl protons), 0.73 (s, 3 H, 18-H from choles-
teryl moiety) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ
157.65, 139.86, 122.04, 100.05, 84.05, 80.75, 74.50, 73.94, 73.11,
72.04, 71.93, 71.24, 68.77, 66.17, 60.39, 56.72, 56.17, 55.52, 50.22,
49.67, 42.09, 41.35, 39.71, 39.28, 38.23, 37.98, 36.87, 36.35, 35.96,
35.70, 31.82, 27.90, 27.73, 23.89, 23.53, 21.76, 21.52, 20.75, 18.39,
17.83, 10.89 ppm. IR (KBr, thin film) νmax (cmϪ1) ϭ 3421 (br),
2951, 1679 (vs), 1530, 1437, 1383, 1206 (s), 1139 (s), 1054 (s). MS:
m/z MALDI-TOF [MHϩ] 897.9, [MNaϩ] 919.9. C52H83N4O19F9
4·5H2O (1320.3): calcd. C 47.31, H 7.02, N 4.24; found C, 47.00,
H 6.69, N 4.09.
1,3,3ЈЈ-Tris-N-(trimethylsilylethoxycarbonyl)kanamycin A (4): To a
solution of compound 3 (0.3 g, 0.29 mmol) in methanol (25 mL,
10% H2O, degazed under nitrogen) was added Pd/C (10%, 30 mg).
After 1 h stirring in hydrogen atmosphere, the suspension was fil-
tered through a pad of celite and concentrated in vacuo to dryness
to yield a white solid (0.23 g, 88%). m.p. 229Ϫ230 °C dec. [α]D25
ϭ
1
ϩ6.62 (c ϭ 0.62, CH3OH). H NMR (300 MHz, CD3OD, 25 °C):
δ ϭ 5.21 (m, 1 H, anomeric proton), 5.10 (partially resolved mul-
tiplet, J ϭ 2.7 Hz, 1 H, anomeric proton), 4.16 (m, 7 H, O-CH2-
CH2 and sugar ring proton), 3.81Ϫ3.31 (overlapping multiplets, 13
H, various ring protons), 3.19 (t app, J ϭ 9.3 Hz, 1 H), 3.01 (par-
tially resolved multiplet, J ϭ 12.3 Hz, 1 H), 2.69 (partially resolved
multiplet, J ϭ 13.3, 6.1 Hz, 1 H), 2.07 (m, 1 H, 2-H equatorial),
1.57 (partially resolved multiplet, J ϭ 12.6 Hz, 1 H, 2-H axial),
1.01 (m, 6 H, CH2-Si), 0.06 (s, 27 H, CH3-Si) ppm. 13C NMR
(75 MHz, CD3OD, 25 °C): δ ϭ 158.70, 157.56, 157.05, 100.96,
98.58, 83.91, 81.39, 75.42, 73.50, 72.96, 72.71, 72.56, 71.57, 70.59,
68.19, 62.76, 62.55, 60.96, 56.45, 50.68, 49.84, 42.34, 34.31, 17.37,
17.31, 17.20, Ϫ2.60, Ϫ2.65, Ϫ2.69 ppm. IR (KBr, thin film) νmax
(cmϪ1) ϭ 3357 (br), 2954, 1695 (vs), 1543 (s), 1419, 1251 (s), 1143,
1042 (vs), 939. MS: m/z MALDI-TOF [MHϩ] 917.7, [MNaϩ]
939.7, [MKϩ] 955.7.
Compound 7: To a solution of kanachol (6) (100 mg, 0.08 mmol) in
water (0.5 mL) was added dioxane (5 mL) and N,NЈ-bisBoc-NЈЈ-
triflylguanidine (284 mg, 0.72 mmol, 9 equiv.) in alternating por-
tions so that the solution remained clear. After 5 minutes triethyl-
amine (9 equiv., 101 µL, 0.72 mmol) was added. The solution was
stirred at room temperature for 4 days after which the volatiles
were removed in vacuo. The residue was partitioned between water
(10 mL) and chloroform (20 mL). The aqueous layer was separated
and extracted with chloroform (2 ϫ 10 mL). The combined organic
layer was washed with brine, dried with Na2SO4, and concentrated
in vacuo. Flash column chromatography (1% methanol in ethyl
acetate) afforded the desired product as a white solid (100 mg,
Compound 5: To a solution of compound 4 (1.655 g, 1.8 mmol) in
tetrahydrofuran (75 mL) and dimethylformamide (30 mL) was ad-
ded triethylamine (1.5 equiv., 377 µL, 2.7 mmol) and cholesteryl
chloroformate (1.5 equiv., 1.215 g, 2.7 mmol). The reaction mixture
became cloudy with time and was allowed to stir for 48 h. Upon
this time, evaporation under reduced pressure was performed until
only dimethylformamide remained. Water was then added to this
solution and a huge white precipitate was formed. After filtration,
flash column chromatography (silica gel, CH2Cl2/MeOH, 91:9) af-
forded the desired product as a white solid (1.857 g, 77.3%). Rf ϭ
0.39 (CH2Cl2/MeOH, 90:10). m.p. 250Ϫ251 °C. [α]2D5 ϭ ϩ4.95 (c ϭ
0.36, CHCl3). 1H NMR (300 MHz, CD3OD/CDCl3, 25 °C): δ ϭ
5.36 (m, 1 H, 6-H from cholesteryl moiety), 5.03 (m, 2 H, anomeric
protons), 4.43 (m, 1 H, 3-H from cholesteryl moiety), 4.21Ϫ4.05
(m, 7 H, CH2-O-CϭO and sugar ring proton), 3.82Ϫ3.29 (m, 15
H), 3.16 (t app, J ϭ 9.3 Hz, 1 H), 2.33 (m, 2 H), 2.17 (partially
resolved multiplet, J ϭ 11.7 Hz, 1 H), 2.04Ϫ1.86 (m, 6 H),
1.59Ϫ0.84 (m, 39 H, CH2-Si and various sugar ring and cholesteryl
protons), 0.68 (s, 3 H, 18-H from cholesteryl moiety), 0.03 (s, 27
H, CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 158.89,
157.80, 157.46, 157.04, 139.68, 122.43, 101.48, 98.69, 84.51, 81.93,
77.65, 76.08, 75.28, 74.77, 7339, 73.14, 72.62, 71.18, 70.55, 68.82,
63.05, 61.38, 56.67, 56.50, 56.11, 50.67, 50.07, 49.85, 42.21, 40.99,
39.69, 39.40, 38.44, 38.26, 36.92, 36.47, 36.07, 35.70, 31.82, 31.77,
29.47, 28.06, 27.84, 27.10, 23.66, 22.35, 22.10, 20.93, 18.97, 18.36,
17.63, 17.49, 17.44, 11.51, Ϫ1.98, Ϫ2.06 ppm. IR (KBr, thin film)
νmax (cmϪ1) ϭ 3384 (br), 2953 (s), 1697 (vs), 1542 (s), 1458, 1420,
1251 (s), 1142, 1045 (vs), 948. MS: m/z MALDI-TOF [MNaϩ]
1352.1, [MKϩ] 1368.1. C64H116N4O19Si3·2H2O (1365.9): calcd. C
56.28, H 8.86, N 4.10; found C, 56.14, H 8.62, N 3.92.
77%). Rf ϭ 0.45 (CH2Cl2/MeOH, 90:10). m.p. Ͼ 300 °C. [α]D25
ϭ
ϩ1.54 (c ϭ 0.46, CHCl3). 1H NMR (300 MHz, CD3OD, 25 °C):
δ ϭ 5.43 (m, 1 H, anomeric proton), 5.21 (m, 1 H, 6-H from choles-
teryl moiety), 5.14 (m, 1 H, anomeric proton), 4.45Ϫ4.24 (m, 3 H),
4.14Ϫ4.05 (m, 1 H), 3.86Ϫ3.28 (m, 12 H, various sugar ring and
cholesteryl protons), 3.13 (m, 1 H), 2.35Ϫ0.83 (m, 97 H, CH3 from
Boc, cholesteryl, and sugar ring protons), 0.68 (s, 3 H, 18-H from
cholesteryl moiety) ppm. IR (KBr, thin film) νmax (cmϪ1) ϭ 3327
(br), 2936, 1784, 1724 (s), 1647 (s), 1421, 1369, 1312, 1237, 1152
(vs), 1055 (s). MS: m/z MALDI-TOF [MNaϩ] 1647.2, [MNaϩ Ϫ 1
Boc] 1546.9, [MNaϩ Ϫ 2 Boc] 1447.1, [MNaϩ Ϫ 3 Boc] 1347.5,
[MNaϩ Ϫ 4 Boc] 1247.3.
Trifluoroacetate Salt 8 of the Tri-Guanidinium Kanamycin A Choles-
terol Conjugate (TGKC): Compound 7 (20 mg, 0.012 mmol) was
suspended in trifluoroacetic acid (1 mL) in an ice-saline bath. After
4 h the mixture was concentrated to dryness After several co-evap-
oration with methanol, the solid residue was dissolved in water and
subsequent lyophilyzation of the solution afforded 10.1 mg
(0.007 mmol, 60%) of TGKC. m.p. 128Ϫ129 °C dec. [α]2D5 ϭ ϩ2.17
(c ϭ 0.23, DMSO). 1H NMR (300 MHz, CD3OD, 25 °C): δ ϭ 5.39
(m, 1 H, anomeric proton), 5.26 (m, 1 H, 6-H from cholesteryl
moiety), 5.09 (m, 1 H, anomeric proton), 4.36 (m, 2 H), 4.09 (m, 1
H), 3.80Ϫ3.20 (m, 14 H, various sugar ring and cholesteryl pro-
tons), 3.28 (m, 1 H), 2.60Ϫ0.80 (m, 33 H, various sugar ring and
cholesteryl protons), 0.73 (s, 3 H, 18-H from cholesteryl moiety)
ppm. IR (KBr, thin film) νmax (cmϪ1) ϭ 3425 (br), 2936, 1677 (vs),
Trifluoroacetate Salt 6; Kanachol (KANAmycin A؊CHOLesterol
Conjugate): Compound 5 (0.102 g, 0.076 mmol) was suspended in
2770
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2764Ϫ2774

Aminoglycoside-Derived Cationic Lipids for Gene Transfection
FULL PAPER
1438, 1384, 1206 (s), 1138 (s), 1040. MS: m/z MALDI-TOF [MHϩ] 61.30, 56.68, 56.50, 56.11, 50.65, 50.10, 46.68, 42.20, 40.27, 39.70,
1023.8, [MNaϩ] 1045.9.
39.38, 38.44, 36.93, 36.44, 36.06, 35.71, 31.82, 31.75, 29.47, 29.32,
28.05, 27.82, 26.22, 25.36, 24.08, 23.65, 22.28, 22.03, 20.90, 18.88,
18.30, 17.59, 17.46, 17.41, 11.44, Ϫ2.16, Ϫ2.19 ppm. MS: m/z
MALDI-TOF: [MNaϩ] 1464.8.
Compound 9a: To a solution of amine 4 (340 mg, 0.38 mmol) in
DMF (15 mL) in an ice bath was added subsequently acid 13a
(111 mg, 0.41 mmol, 1.1 equiv.), 1-hydroxy-7-azabenzotriazole
(HOAt, 62 mg, 0.45 mmol, 1.2 equiv.), and N-(3-dimethylaminop-
Trifluoroacetate Salt 12a; KanaCapChol: Compound 11a (42 mg,
0.029 mmol) was suspended in trifluoroacetic acid (2 mL) in an ice-
saline bath. After 45 min under 0 °C, the mixture was concentrated
to dryness. After several co-evaporation with methanol, the solid
residue was dissolved in water and subsequent lyophilyzation of the
solution afforded 12a as a fluffy white powder in 92% yield (36 mg).
Rf ϭ 0.32 (CH2Cl2/MeOH/NH4OH, 5:4:1). 1H NMR (300 MHz,
CD3OD, 25 °C): δ ϭ 5.40 (m, 1 H, 6-H from cholesteryl moiety),
5.31 (partially resolved multiplet, J ϭ 3 Hz, 1 H, anomeric pro-
tons), 5.11 (partially resolved multiplet, J ϭ 2.7 Hz, 1 H, anomeric
protons), 4.39 (m, 1 H, 3-H from cholesteryl moiety), 4.10Ϫ3.32
(m, 16 H), 3.26Ϫ3.06 (m, 3 H), 2.55 (partially resolved multiplet,
J ϭ 12 Hz, 1 H), 2.34Ϫ2.22 (m, 4 H), 2.08Ϫ1.80 (m, 6 H),
1.66Ϫ0.88 (m, 38 H), 0.73 (s, 3 H, 18-H from cholesteryl moiety)
ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 175.99, 157.31,
139.87, 122.02, 100.19, 97.29, 84.02, 80.57, 74.10, 73.80, 72.97,
72.15, 71.98, 71.74, 71.19, 98.77, 66.10, 60.43, 56.73, 56.18, 55.51,
50.23, 49.76, 42.11, 40.13, 39.97, 39.73, 39.29, 38.31, 36.89, 36.34,
35.99, 35.72, 35.51, 31.82, 31.63, 29.36, 29.28, 27.89, 27.74, 26.06,
25.23, 23.91, 23.56, 21.80, 21.55, 20.76, 18.39, 17.87, 10.94 ppm.
MS: m/z MALDI-TOF [MNaϩ] 1032.7, [MHϩ] 1010.7.
ropyl)-NЈ-ethylcarbodiimide
hydrochloride (EDC,
109 mg,
0.57 mmol, 1.5 equiv.). The mixture was stirred 0 °C for 2 h, then
12 h at room temperature until the solution became incolore. The
solution was then partitioned between ethyl acetate and water and
the organic layer was washed subsequently by an aqueous solution
of hydrogen chloride (1%), a solution of sodium carbonate (0.2 ),
and a saturated sodium choride aqueous solution. The organic
layer was dried with Na2SO4, concentrated to dryness, and flash
column chromatography (7.5% MeOH in CH2Cl2) of the solid resi-
due afforded the desired product as a white powder (307 mg, 71%).
1
Rf ϭ 0.23 (CH2Cl2/MeOH, 90:10). H NMR (300 MHz, CD3OD,
25 °C): δ ϭ 7.34 (m, 5 H, arene-H), 5.15Ϫ5.06 (m, 4 H, CH2-arene
overlapping with anomeric protons), 4.25Ϫ4.10 (m, 7 H, CH2-O-
CϭO and sugar ring proton), 3.82Ϫ3.32 (m, 15 H), 3.13 (m, 3 H),
2.28 (t app, J ϭ 7.3 Hz, 2 H), 2.06 (m, 1 H), 1.68Ϫ1.30 (m, 7 H),
1.00 (m, 6 H, CH2-Si), 0.06 (m, 27 H, CH3-Si) ppm. 13C NMR
(75 MHz, CD3OD, 25 °C): δ ϭ 175.67, 158.74, 158.68, 157.52,
157.48, 157.20, 137.08, 128.11, 127.58, 127.40, 101.27, 98.57, 84.29,
81.41, 75.25, 73.10, 73.01, 72.78, 71.11, 70.88, 70.59, 68.25, 65.96,
62.74, 62.57 61.01, 56.50, 50.68, 49.90, 40.31, 39.58, 35.46, 29.24,
26.10, 25.34, 17.40, 17.31, 17.22, Ϫ2.64, Ϫ2.67 ppm. MS: m/z
MALDI-TOF [MNaϩ] 1186.5.
Compounds 9bϪ12b and 9cϪ12c were prepared according to the
procedures described for 9aϪ12a.
1,3,3ЈЈ-Tris-N-(trimethylsilylethoxycarbonyl)-6Ј-N-(6-aminohexa-
noyl)kanamycin A (10a): To a solution of 9a (118 mg, 0.10 mmol)
in methanol (10% H2O) (5 mL, degazed under nitrogen) was added
Pd/C (10%, 12 mg). The solution was degazed and after 1 h stirring
in hydrogen atmosphere, the suspension was filtered through a pad
of celite and concentrated in vacuo to dryness to yield a white solid
(85 mg, 81.4%). 1H NMR (300 MHz, CD3OD, 25 °C): δ ϭ
5.13Ϫ5.08 (m, 2 H, anomeric protons), 4.25Ϫ4.10 (m, 7 H, CH2-
O-CϭO and sugar ring proton), 3.82Ϫ3.32 (m, 15 H), 3.13 (t app,
J ϭ 9.4 Hz, 1 H), 2.72 (m, 2 H), 2.31 (t app, J ϭ 7.3 Hz, 2 H), 2.06
(m, 1 H), 1.71Ϫ1.39 (m, 7 H), 1.00 (m, 6 H, CH2-Si), 0.06 (m, 27
H, CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 174.16,
157.23, 156.06, 155.75, 99.80, 97.08, 82.83, 79.75, 73.78, 71.62,
71.48, 71.31, 69.60, 69.11, 66.71, 61.22, 61.03, 59.47, 54.95, 49.21,
48.39, 43.48, 39.15, 38.01, 29.63, 27.78, 25.15, 24.63, 24.05, 23.89,
15.87, 15.79, 15.68, Ϫ4.20, Ϫ4.25 ppm. MS: m/z MALDI-TOF
[MNaϩ] 1052.7.
1
Compound 9b: Yield 85.3%. Rf ϭ 0.28 (CH2Cl2/MeOH, 90:10). H
NMR (300 MHz, CD3OD, 25 °C): δ ϭ 7.34 (m, 5 H, arene-H),
5.17 (m, 1 H, anomeric proton), 5.08 (m, 3 H, CH2-arene overlap-
ping with anomeric proton), 4.22Ϫ4.09 (m, 8 H, CH2-O-CϭO, Hα
from linker and sugar ring proton), 3.81Ϫ3.32 (m, 15 H), 3.13 (m,
3 H), 2.07 (m, J ϭ 12.6 Hz, 1 H), 1.85Ϫ1.25 (m, 16 H), 1.01 (m, 6
H, CH2-Si), 0.06 (m, 27 H, CH3-Si) ppm. 13C NMR (75 MHz,
CD3OD, 25 °C): δ ϭ 174.49, 158.69, 157.56, 157.16 156.51, 137.04,
128.09, 127.57, 127.40, 101.02, 98.51, 83.77, 81.28, 79.29, 75.38,
73.11, 72.97, 72.71, 71.05, 70.60, 68.19, 65.68, 62.75, 62.53, 60.94,
56.45, 54.82, 50.73, 49.81, 40.06, 39.91, 34.20, 31.68, 29.17, 27.44,
22.79, 17.37, 17.27, 17.17, Ϫ2.74, Ϫ2.77 ppm. MS: m/z MALDI-
TOF [MNaϩ] 1301.6.
1
Compound 10b: Yield 89%. H NMR (300 MHz, CD3OD, 25 °C):
δ ϭ 5.17 (m, 1 H, anomeric proton), 5.09 (partially resolved mul-
tiplet, J ϭ 2.7 Hz, 1 H, anomeric proton), 4.25Ϫ4.10 (m, 8 H, CH2-
O-CϭO, Hα from linker and sugar ring proton), 3.80Ϫ3.32 (m, 15
H), 3.13 (partially resolved multiplet, J ϭ 9.4 Hz, 1 H), 2.69 (t app,
J ϭ 6.9 Hz, 2 H), 2.06 (partially resolved multiplet, J ϭ 12.3 Hz,
1 H), 1.85Ϫ1.30 (m, 16 H), 1.01 (m, 6 H, CH2-Si), 0.06 (m, 27 H,
CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 174.53,
158.72, 157.55, 157.16, 156.47, 101.05, 98.58, 83.87, 81.32, 79.26,
75.39, 73.09, 72.98, 72.72, 71.05, 70.59, 70.50, 68.17, 62.73, 62.53,
60.95, 56.39, 54.83, 50.72, 49.81, 40.64, 39.25, 34.27, 31.83, 31.35,
27.45, 22.88, 17.38, 17.29, 17.17, Ϫ2.71, Ϫ2.75 ppm. MS: m/z
MALDI-TOF [MNaϩ] 1167.8.
Compound 11a: To a solution of compound 10a (80 mg, 0.07 mmol)
in THF/DMF (9:1, 10 mL) was added triethylamine (1.5 equiv., 16
µL, 0.11 mmol) and cholesteryl chloroformate (1.5 equiv., 52 mg,
0.11 mmol). The reaction mixture was stirred for 12 h. Upon this
time, the volatiles were removed in vacuo and flash column chro-
matography of the solid residue (7.5% methanol in dichlorometh-
ane) afforded the desired product as a white solid (100 mg, 89.2%).
1
Rf ϭ 0.30 (CH2Cl2/MeOH, 90:10). H NMR (300 MHz, CD3OD/
CDCl3, 25 °C): δ ϭ 5.38 (m, 1 H, 6-H from cholesteryl moiety),
5.03 (m, 2 H, anomeric protons), 4.40 (m, 1 H, 3-H from choles-
teryl moiety), 4.21Ϫ4.06 (m, 7 H, CH2-O-CϭO and sugar ring pro- Compound 11b: Yield 86%. Rf ϭ 0.32 (CH2Cl2/MeOH, 90:10). H
1
ton), 3.82Ϫ3.32 (m, 15 H), 3.22Ϫ3.06 (m, 3 H), 2.39Ϫ2.24 (m, 4
NMR (300 MHz, CD3OD, 25 °C): δ ϭ 5.41 (m, 1 H, 6-H from
H), 2.15 (m, 1 H), 2.05Ϫ1.80 (m, 6 H), 1.66Ϫ0.80 (m, 44 H), 0.62 cholesteryl moiety), 5.17 (m, 1 H, anomeric proton), 5.09 (partially
(s, 3 H, 18-H from cholesteryl moiety), 0.04 (m, 27 H, CH3-Si)
resolved multiplet, J ϭ 3 Hz, 1 H, anomeric proton), 4.40 (m, 1 H,
ppm. 13C NMR (75 MHz, CD3OD/CDCl3, 25 °C): δ ϭ 175.73, 3-H from cholesteryl moiety), 4.22Ϫ4.06 (m, 8 H, CH2-O-CϭO,
158.84, 157.47, 157.19, 157.15, 139.79, 122.25, 101.36, 98.65, 84.41, Hα from linker and sugar ring proton), 3.80Ϫ3.32 (m, 15 H),
81.73, 77.75, 75.20, 74.25, 73.02, 72.80, 71.22, 70.57, 68.67, 62.99,
3.15Ϫ3.09 (m, 3 H), 2.34 (m, 2 H), 2.09Ϫ1.82 (m, 7 H), 1.75Ϫ0.88
Eur. J. Org. Chem. 2003, 2764Ϫ2774
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2771

M. Sainlos, P. Belmont, J.-P. Vigneron, P. Lehn, J.-M. Lehn
FULL PAPER
(m, 54 H), 0.74 (s, 3 H, 18-H from cholesteryl moiety), 0.07 (m, 27
157.10, 156.86, 139, 73, 139.56, 122.65, 122.34, 100.66, 98.64,
83.35, 82.02, 77.75, 75.49, 74.83, 74.30, 73.05, 72.68, 71.15, 70.57,
H, CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 172.99,
157.19, 156.01, 155.79, 155.64, 155.01, 138.37, 120.56, 99.49, 97.02, 68.68, 62.99, 61.31, 56.77, 56.65, 56.49, 56.26, 56.18, 55.03, 50.68,
82.27, 79.80, 77.76, 73.88, 72.56, 72.56, 71.62, 71.48, 71.22, 69.57, 50.05, 46.66, 42.23, 39.79, 39.71, 39.39, 38.46, 36.89, 36.47, 36.41,
69.10, 66.72, 61.24, 61.01, 59.46, 55.24, 54.94, 54.69, 53.34, 49.22, 36.10, 35.78, 31.91, 31.79, 29.47, 29.29, 28.09, 27.98, 27.82, 27.78,
48.72, 48.34, 40.63, 38.38, 38.25, 37.82, 36.85, 35.41, 34.86, 34.52, 24.15, 24.08, 23.74, 22.32, 22.30, 22.04, 20.94, 19.01, 18.92, 18.33,
34.25, 32.72, 30.34, 30.16, 27.75, 36.45, 26.26, 25.68, 22.45, 22.09,
17.62, 17.46, 11.56, 11.46, 8.28, Ϫ2.12, Ϫ2.19 ppm. MS: m/z
21.32, 20.36, 20.11, 19.30, 16.98, 16.43, 15.90, 15.78, 15.69, 9.51, MALDI-TOF [MNaϩ] 1892.4.
Ϫ4.21, Ϫ4.26 ppm. MS: m/z MALDI-TOF [MNaϩ] 1580.1.
Trifluoroacetate Salt 12c ; KanaLysDiChol: Yield 93.3%. Rf ϭ 0.44
(CH2Cl2/MeOH/NH4OH, 5:4:1). 1H NMR (300 MHz, CD3OD, 25
°C): δ ϭ 5.40 (m, 2 H, 6-H from cholesteryl moieties), 5.33 (m, 1
H, anomeric proton), 5.12 (m, 1 H, anomeric proton), 4.43 (m, 2
H, 3-H from cholesteryl moieties), 4.25Ϫ4.00 (m, 2 H, sugar ring
proton and Hα from linker), 3.94Ϫ3.32 (m, 15 H), 3.22Ϫ3.06 (m,
3 H), 2.54 (m, 1 H), 2.34 (m, 4 H), 2.19Ϫ1.75 (m, 12 H), 1.70Ϫ0.90
(m, 71 H), 0.75 (s, 6 H, 18-H from cholesteryl moieties) ppm. 13C
NMR (75 MHz, CD3OD, 25 °C): δ ϭ 174.64, 160.75, 157.76,
139.72, 122.28, 100.29, 98.45, 83.84, 80.92, 74.68, 74.24, 73.65,
72.75, 71.95, 70.98, 68.72, 66.10, 60.49, 60.49, 56.82, 56.67, 56.35,
56.25, 55.54, 55.17, 50.0349.86, 42.21, 39.84, 39.69, 39.36, 38.97,
38.54, 38.42, 36.95, 36.78, 36.44, 36.35, 36.09, 35.80, 31.95, 31.74,
31.53, 29.44, 28.06, 27.80, 27.71, 24.13, 24.02, 23.76, 22.15, 21.88,
20.89, 18.90, 18.75, 18.21, 11.51, 11.29, 8.01 ppm. MS: m/z
MALDI-TOF [MNaϩ] 1459.9.
Trifluoroacetate Salt 12b; KanaLysChol: Yield 96%. Rf ϭ 0.18
(CH2Cl2/MeOH/NH4OH, 5:4:1). 1H NMR (300 MHz, CD3OD, 25
°C): δ ϭ 5.40 (m, 1 H, 6-H from cholesteryl moiety), 5.35 (partially
resolved multiplet, J ϭ 3.9 Hz, 1 H, anomeric proton), 5.11 (par-
tially resolved multiplet, J ϭ 3.6 Hz, 1 H, anomeric proton), 4.38
(m, 1 H, 3-H from cholesteryl moiety), 3.99Ϫ3.32 (m, 16 H),
3.19Ϫ3.09 (m, 3 H), 2.55 (partially resolved multiplet, J ϭ 12.3 Hz,
1 H), 2.32 (m, 2 H), 2.08Ϫ1.85 (m, 8 H), 1.72Ϫ0.88 (m, 38 H), 0.74
(s, 3 H, 18-H from cholesteryl moiety) ppm. 13C NMR (75 MHz,
CD3OD, 25 °C): δ ϭ 167.94, 156.26, 138.25, 120.49, 98.74, 95.57,
84.46, 78.77, 72.66, 72.25, 71.37, 70.68, 70.34, 69.81, 69.66, 67.16,
64.53, 58.93, 55.14, 54.58, 53.89, 51.52, 48.64, 48.19, 40.51, 38.60,
38.14, 37.70, 36.71, 35.29, 34.76, 34.39, 34.13, 30.23, 30.04, 29.41,
27.64, 26.33, 26.29, 26.15, 22.32, 21.96, 20.21, 19.96, 19.17, 16.79,
16.27, 9.34 ppm. MS: m/z MALDI-TOF [MNaϩ] 1047.6.
1
Compound 9c: Yield 87.3%. Rf ϭ 0.28 (CH2Cl2/MeOH, 90:10). H
N-Succinyl(dioctadecyl)amine (13): To a solution of dioctadecylam-
ine (522 mg, 1 mmol) in pyridine (10 mL) was added succinic anhy-
dride (200 mg, 2 mmol, 2 equiv.). The solution was refluxed for
12 h. The volatiles were removed in vacuo and the solid residue
was resuspended in dichloromethane and washed with chlorhydric
acid (1 ) followed by water. Recrystallisation in acetone afforded
the desired produt as a white solid (566 mg, 90.9%). Mp found
66Ϫ68 °C, ref.[34] 56 °C. Rf ϭ 0.86 (CH2Cl2/MeOH, 90:10). 1H
NMR (200 MHz, CDCl3, 25 °C): δ ϭ 3.24 (quint., J ϭ 6.1 Hz, 4
H, CH2N), 2.65 (m, 4 H, CH2CO), 1.6Ϫ1.1 (m, 64 H), 0.88 (m, 6
H, CH3) ppm. 13C NMR (50 MHz, CDCl3, 25 °C): δ ϭ 171.47,
48.09, 46.45, 31.80, 29.56, 29.23, 28.77, 27.95, 27.58, 26.93, 26.82,
22.55, 13.92.
NMR (300 MHz, CD3OD, 25 °C): δ ϭ 7.34 (m, 10 H, arene-H),
5.20 (m, 1 H, anomeric proton), 5.10Ϫ5.06 (m, 5 H, CH2-arene
overlapping with anomeric proton), 4.19Ϫ4.10 (m, 8 H, CH2-O-
CϭO, Hα from linker and sugar ring proton), 3.80Ϫ3.32 (m, 15
H), 3.12 (m, 3 H), 2.04 (partially resolved multipet, J ϭ 12.6 Hz, 1
H), 1.80Ϫ1.30 (m, 7 H), 1.00 (m, 6 H, CH2-Si), 0.06 (m, 27 H,
CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ 172.55,
157.22, 156.09, 155.70, 135.56, 135.21, 126.70, 126.62, 126.39,
126.10, 126.02, 125.93, 92.28, 97.04, 81.97, 79.96, 73.93, 71.65,
71.55, 71.18, 69.60, 69.39, 69.13, 66.76, 64.98, 64.53, 61.29, 61.08,
59.51, 55.01, 53.75, 49.24, 48.31, 38.58, 38.08, 32.76, 30.02, 27.88,
27.66, 21.26, 15.89, 15.82, 15.73, Ϫ4.51, Ϫ4.55 ppm. MS: m/z
MALDI-TOF [MNaϩ] 1335.7.
Compound 9d: The compound was synthesized according to the
procedure used for compounds 9aϪc with slight modifications. To
a solution of amine 3 (154 mg, 0.168 mmol) in DMF/CH2Cl2 (1:1,
20 mL) in an ice bath was added subsequently the acid 13 (115 mg,
0.184 mmol, 1.1 equiv.), HOAt (27 mg, 0.201 mmol, 1.2 equiv.) and
EDC (48 mg, 0.251 mmol, 1.5 equiv.). The mixture was stirred 0
°C for 2 h, then 12 h at room temperature. The solution was washed
as described for 9a with dichloromethane instead of ethyl acetate
and flash column chromatography (CH2Cl2/MeOH, 92.5:7.5) of the
solid residue afforded the desired product as a white powder
(193 mg, 75.4%). Rf ϭ 0.37 (CH2Cl2/MeOH, 90:10). 1H NMR
(300 MHz, CD3OD/CDCl3, 25 °C): δ ϭ 5.29 (m, 1 H, anomeric
proton), 5.22 (m, 1 H, anomeric proton), 4.50Ϫ4.04 (m, 8 H, CH2-
O-CϭO and sugar ring protons), 4.00Ϫ3.38 (m, 17 H), 3.34 (t, J ϭ
1,3,3ЈЈ-Tris-N-(trimethylsilylethoxycarbonyl)-6Ј-N-(2,6-diamino-
hexanoyl)kanamycin A (10c): Yield 91%. 1H NMR (300 MHz,
CD3OD, 25 °C): δ ϭ 5.16 (m, 1 H, anomeric proton), 5.09 (m, 1
H, anomeric proton), 4.21Ϫ4.10 (m, 8 H, CH2-O-CϭO, Hα from
linker and sugar ring proton), 3.81Ϫ3.32 (m, 15 H), 3.14 (partially
resolved multiplet, J ϭ 9.4 Hz, 1 H), 2.75 (t app, J ϭ 7.0 Hz, 2 H),
2.06 (m, 1 H), 1.75Ϫ1.30 (m, 7 H), 1.02 (m, 6 H, CH2-Si), 0.06 (m,
27 H, CH3-Si) ppm. 13C NMR (75 MHz, CD3OD, 25 °C): δ ϭ
177.16, 158.73, 157.57, 157.20, 101.19, 98.62, 84.14, 81.31, 75.36,
73.09, 73.00, 72.71, 70.92, 70.61, 70.45, 68.19, 62.73, 62.55, 60.97,
56.40, 54.62, 50.71, 49.78, 40.42, 39.13, 34.74, 34.34, 30.67, 22.56,
17.38, 17.29, 17.18, Ϫ2.71, Ϫ2.74 ppm. MS: m/z MALDI-TOF
[MNaϩ] 1067.7, [MHϩ] 1045.6.
Compound 11c: Yield 60.9%. Rf ϭ 0.32 (CH2Cl2/MeOH, 90:10). 1H 9.3 Hz, 2 H), 2.89 (m, 2 H), 2.76 (m, 2 H), 2.44 (partially resolved
NMR (300 MHz, CD3OD/CDCl3, 25 °C): δ ϭ 5.37 (m, 2 H, 6-H
from cholesteryl moieties), 5.17 (m, 1 H, anomeric proton), 5.03
multiplet, J ϭ 11.1 Hz, 1 H), 1.9Ϫ1.7 (m, 5 H, -CH2- from DODA
overlapping with sugar ring proton), 1.49 (m, 60 H, -CH2- from
(partially resolved multiplet, J ϭ 3.3 Hz, 1 H, anomeric proton), DODA), 1.21 (partially resolved multiplet, J ϭ 7.9 Hz, 6 H, CH2-
4.43 (m, 2 H, 3-H from cholesteryl moieties), 4.17Ϫ4.00 (m, 8 H,
CH2-O-CϭO, Hα from linker and sugar ring proton), 3.82Ϫ3.32
Si), 1.10 (partially resolved multiplet, J ϭ 6.6 Hz, 6 H, -CH3 from
DODA), 0.26 (m, 27 H, CH3-Si) ppm. 13C NMR (75 MHz,
(m, 15 H), 3.13Ϫ3.06 (m, 3 H), 2.32 (m, 4 H), 2.13 (m, 1 H), CD3OD/CDCl3, 25 °C): δ ϭ 174.13, 171.88, 158.82, 157.46, 157.06,
2.05Ϫ1.72 (m, 12 H), 1.66Ϫ0.85 (m, 77 H), 0.70 (s, 6 H, 18-H 101.07, 98.63, 83.88, 81.82, 77.80, 75.43, 73.14, 73.04, 72.70, 71.18,
from cholesteryl moieties), 0.04 (m, 27 H, CH3-Si) ppm. 13C NMR
70.93, 70.59, 68.58, 62.94, 61.24, 56.46, 53.33, 50.75, 46.28, 39.92,
(75 MHz, CD3OD/CDCl3, 25 °C): δ ϭ 174.19, 158.85, 157.46, 31.74, 30.56, 29.68, 29.47, 29.44, 29.35, 29.22, 29.15, 28.99, 28.63,
2772
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2764Ϫ2774

Aminoglycoside-Derived Cationic Lipids for Gene Transfection
FULL PAPER
awmy, D. Scherman, J. Crouzet, J.-M. Lehn, P. Lehn, Proc.
28.18, 26.84, 26.68, 22.43, 17.59, 17.43, 17.39, 13.44, Ϫ2.26,
Ϫ2.31 ppm. MS: m/z MALDI-TOF [MNaϩ] 1543.1.
[9e]
Natl. Acad. Sci. USA 1999, 96, 2621Ϫ2626.
M. Patel, E.
Vivien, M. Hauchecorne, N. Oudrhiri, R. Ramasawmy, J.-P.
Vigneron, P. Lehn P, J.-M. Lehn, Biochem. Biophys. Res. Com-
Trifluoroacetate Salt 12d ; KanaSucDODA: Compound 11d was
deprotected according to the procedure used for 12a. Yield 95.5%.
Rf ϭ 0.39 (CH2Cl2/MeOH/NH4OH, 5:4:1). 1H NMR (300 MHz,
CD3OD/CDCl3, 25 °C): δ ϭ 5.29 (m, 1 H, anomeric proton), 5.22
(m, 1 H, anomeric proton), 4.00Ϫ3.20 (m, 19 H), 3.17 (m, 2 H),
2.68 (m, 2 H), 2.54 (m, 3 H), 1.96 (partially resolved multiplet, J ϭ
12.4 Hz, 1 H), 1.7Ϫ1.5 (m, 4 H, -CH2- from DODA), 1.30 (m, 60
H, -CH2-from DODA), 0.91 (partially resolved multiplet, J ϭ
6.6 Hz, 6 H, CH3 from DODA) ppm. 13C NMR (75 MHz,
CD3OD/CDCl3, 25 °C): δ ϭ 174.15, 172.35, 100.25, 98.03, 83.84,
80.95, 73.70, 72.96, 72.57, 71.96, 71.25, 68.76, 65.97, 60.33, 55.52,
53.71, 49.83, 48.03, 46.23, 40.20, 31.71, 30.36, 29.45, 29.42, 29.34,
29.29, 29.27, 29.13, 29.02, 28.43, 27.81, 27.36, 26.70, 26.51, 22.37,
13.11 ppm. MS: m/z MALDI-TOF [MNaϩ] 1110.8, [MHϩ] 1088.8.
[9f]
mun. 2001, 281, 536Ϫ543.
B. Pitard, N. Oudrhiri, O. Lam-
bert, E. Vivien, C. Masson, B. Wetzer, M. Hauchecorne, D.
Scherman, J.-L. Rigaud, J.-P. Vigneron, J.-M. Lehn, P. Lehn, J.
Gene Med. 2001, 3, 478Ϫ487.
[10]
H. Umezawa, I. R. Hooper (Eds.), Aminoglycoside Antibiotics.
Springer-Verlag: New York, Heidelberg, 1982.
[11] [11a]
D. Moazed, H. F. Noller, Nature 1987, 327, 389Ϫ394. [11b]
P. Purohit, S. Stern, Nature 1994, 370, 659Ϫ662.
U. Von Ahsen, H. F. Noller, Science 1993, 260, 1500Ϫ1503.
B. Clouet-d’Orval, T. K. Stage, O. C. Uhlenbeck, Biochemistry
1995, 34, 11186Ϫ11190.
[12]
[13]
[14]
H. Y. Mei, A. A. Galan, N. S. Halim, D. P. Mack, D. W. More-
land, K. B. Sanders, H. N. Truong, A. W. Czarnik, Bioorg.
Med. Chem. Lett. 1995, 5, 2755Ϫ2760.
[15] [15a] G. Werstuck, M. L. Zapp, R. M. Green, Chem. Biol. 1996,
[15b]
3, 129Ϫ137.
S. R. Kirk, N. W. Luedtke, Y. Tor, J. Am.
Chem. Soc. 2000, 122, 980Ϫ981.
[16] [16a]
F. Walter, Q. Vicens, E. Westhof, Curr. Opin. Chem. Biol.
1999, 3, 694Ϫ704. [16b] Y. To r, Angew. Chem. Int. Ed. 1999, 38,
1579Ϫ1582.
Acknowledgments
[16c]
K. Michael, H. Wang, Y. Tor, Bioorg. Med.
We would like to thank G. Blais and S. Vincent for helpful dis-
cussions. This work was supported by grants from Vaincre la Mu-
coviscidose (VLM, Paris, France), the Association Franc¸aise contre
les Myopathies (AFM, Evry, France), the Association pour la Re-
cherche contre le Cancer (ARC, Villejuif, France), and Gencell S.A.
Chem. 1999, 7, 1361Ϫ1371 and references therein.
[17] [17a]
D. E. Dorman, J. W. Paschal, K. E. Merkel, J. Am. Chem.
[17b]
Soc. 1976, 98, 6885Ϫ6888.
R. E. Botto, B. Coxon, J. Am.
[17c]
Chem. Soc. 1983, 105, 1021Ϫ1028.
L. Szilagyi, Z. S. Pusz-
tahelyi, S. Jakab, I. Kovacs, Carbohydr. Res. 1993, 247, 99Ϫ109.
[18] [18a] T. Ohyama, D. Wang, J. A. Cowan, Chem. Commun. 1998,
467Ϫ468. [18b] Q. Chen, R. H. Shafer, I. D. Kuntz, Biochemistry
[18c]
1997, 36, 11402Ϫ11407.
H. Wang, Y. Tor, J. Am. Chem.
[18d]
Soc. 1997, 119, 8734Ϫ8735.
V. Constantinou-Kokotou, G.
[1]
[1a] R. C. Mulligan, Science 1993, 260, 926Ϫ932. [1b] W. F. And-
Karikas, G. Kokotos, Bioorg. Med. Chem. Lett. 2001, 11,
1015Ϫ1018.
erson, Nature 1998, 392 (6679 Suppl), 25Ϫ30.
[2]
[19] [19a]
R. G. Crystal, Science 1995, 270, 404Ϫ410.
M. Hendrix, P. B. Alper, E. S. Priestley, C. H. Wong, An-
gew. Chem. Int. Ed. Engl. 1997, 36, 95Ϫ98. [19b] D. Fourmy, M.
I. Recht, S. C. Blanchard, J. D. Puglisi, Science 1996, 274,
1367Ϫ1371.
[3] [3a]
[3b]
D. Jolly, Cancer Gene Ther. 1994, 1, 51Ϫ64.
P. Lehn, S.
Fabrega, N. Oudrhiri, J. Navarro, Adv. Drug Deliv. Rev. 1998,
30, 5Ϫ11.
[4]
[20]
I. M. Verma, N. Somia, Nature 1997, 389, 239Ϫ242.
C. L. Densmore, T. H. Giddings, J. C. Waldrep, B. Kinsey, V.
Knight, Gene Med. 1999, 1, 251Ϫ264.
[5] [5a]
[5b]
X. Gao, L. Huang, Gene Ther. 1995, 2, 710Ϫ722.
E.
[21] [21a]
Wagner, in: Self-assembling Complexes for Gene Delivery (Eds.:
N. W. Luedtke, T. J. Baker, M. Goodman, Y. Tor, J. Am.
[21b]
A. V. Kabanov, P. L. Felgner, L. W. Seymour), John Wiley &
Chem. Soc. 2000, 122, 12035Ϫ12036.
T. J. Baker, N. W.
[5c]
Sons, Chichester, UK, 1998, pp. 309Ϫ322.
A. Kichler, J. P.
Luedtke, Y. Tor, M. Goodman, J. Org. Chem. 2000, 65,
[21c]
Behr, P. Erbacher, in: Nonviral Vectors for Gene Therapy (Eds.:
L. Huang, M. C. Huang, E. Wagner), Academic Press, San
Diego, USA, 1999, pp. 191Ϫ206.
9054Ϫ9058.
Y. Hui, R. Ptak, R. Paulman, M. Pallansch,
C.-W. T. Chang, Tetrahedron Lett. 2002, 43, 9255Ϫ9257.
[22] [22a]
A. Litovchick, A. G. Evdokimov, A. Lapidot, FEBS Let-
[6]
A. D. Miller, Angew. Chem. Int. Ed. 1998, 37, 1768Ϫ1785.
[22b]
ters 1999, 445, 73Ϫ79.
A. Litovchick, A. G. Evdokimov,
[7] [7a]
J. A. Roth, R. J. Cristiano, J. Nat. Cancer Inst. 1997, 89,
A. Lapidot, Biochemistry 2000, 39, 2838Ϫ2852.
N. W. Luedtke, Y. Tor, Angew. Chem. Int. Ed. 2000, 39,
1788Ϫ1790.
[7b]
21Ϫ39.
E. M. Hersh, A. T. Stopeck, in: Self-assembling
[23]
Complexes for Gene Delivery (Eds.: A. V. Kabanov, P. L.
Felgner, L. W. Seymour), Wiley, Chichester, UK, 1998, pp.
421Ϫ436.
[24] [24a] P. A. Wender, D. J. Mitchel, K. Pattabiraman, E. T. Pelkey,
L. Steinman, J. B. Rothbard, Proc. Natl. Acad. Sci. USA 2000,
[8] [8a]
[8b]
R. C. Boucher, J. Clin. Invest. 1999, 103, 441Ϫ445.
E.
[24b]
97, 13003Ϫ13008.
J. B. Rothbard, S. Garlington, Q. Lin,
W. F. W. Alton, M. Stern, R. Farley, A. Jaffe, S. L. Chadwick,
J. Phillips, J. Davies, S. N. Smith, J. Browning, M. G. Davies,
M. E. Hodson, S. R. Durham, D. Li, P. K. Jeffery, M. Scallan,
T. Kirschberg, E. Kreider, P. L. McGrane, P. A. Wender, P. A.
[24c]
Khavari, Nat. Med. 2000, 6, 1253Ϫ1257.
J. B. Rothbard,
E. Kreider, C. L. VanDeusen, L. Wright, B. L. Wylie, P. A.
Wender, J. Med. Chem. 2002, 45, 3612Ϫ3618.
X. Gao, L. Huang, Biochem. Biophys. Res. Commun. 1991,
179, 280Ϫ285.
P. Belmont, A. Aissaoui, M. Hauchecorne, N. Oudrhiri, L. Pe-
tit, J.-P. Vigneron, J.-M. Lehn, P. Lehn, J. Gene Med. 2002,
4, 517Ϫ526.
A. Van Schepdael, R. Busson, P. Claes, Bull. Soc. Chim. Belg.
1992, 101, 709Ϫ718.
I. Grapsas, Y. J. Cho, S. Mobashery, J. Org. Chem. 1994, 59,
1918Ϫ1922.
R. Balfour, S. J. Eastman, S. H. Cheng, A. E. Smith, D.
[8c]
Meeker, D. M. Geddes, Lancet 1999, 353, 947Ϫ954.
J. C.
[25]
[26]
Davies, D. M. Geddes, E. W. F. W. Alton, J. Gene Med. 2001,
3, 409Ϫ417.
[9] [9a]
J.-P. Vigneron, N. Oudrhiri, M. Fauquet, L. Vergely, J.-C.
Bradley, M. Basseville, P. Lehn, J.-M. Lehn, Proc. Natl. Acad.
Sci. USA 1996, 93, 9682Ϫ9686. [9b] N. Oudrhiri, J.-P. Vigneron,
M. Peuchmaur, T. Leclerc, J.-M. Lehn, P. Lehn, Proc. Natl.
N. Oudrhiri, J.-P.
Vigneron, M. Hauchecorne, R. Toury, A. I. Lemoine, M. Pe-
[27]
[28]
[29]
[9c]
Acad. Sci. USA 1997, 94, 1651Ϫ1656.
uchmaur, J. Navarro, J.-M. Lehn, P. Lehn, Biog. Amines 1998,
[9d]
14, 537Ϫ552.
B. Pitard, N. Oudrhiri, J.-P. Vigneron, M.
W. Streicher, H. Loibner, J. Hildebrandt, F. Turnowsky, Drugs
Exptl. Clin. Res. 1983, IX, 591Ϫ598.
Hauchecorne, O. Aguerre, R. Toury, M. Airiau, R. Ramas-
Eur. J. Org. Chem. 2003, 2764Ϫ2774
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2773

M. Sainlos, P. Belmont, J.-P. Vigneron, P. Lehn, J.-M. Lehn
FULL PAPER
[30]
[34]
L. A. Carpino, J.-H. Tsao, H. Ringsdorf, E. Fell, G. Hettrich,
A. Laschewsky, H. Ringsdorf, G. Schmidt, J. Schneider, J. Am.
Chem. Soc. 1987, 109, 788Ϫ796.
[35] [35a]
J. Chem. Soc., Chem. Commun. 1978, 358Ϫ359.
[31]
I. Grapsas, I. Massova, S. Mobashery, Tetrahedron 1998, 54,
F. Filira, L. Biondi, M. Gobbo, R. Rocchi, Tetrahedron
[35b]
7705Ϫ7720.
Lett. 1991, 32, 7463Ϫ7464.
J.-P. Mazaleyrat, J. Xie, M.
[32]
J. Szymoniak, B. El Mouatassim, J. Besancon, C. Mo¨ıse, P.
Wakselman, Tetrahedron Lett. 1992, 33, 4301Ϫ4302. [35c] N. L.
Benoiton, Int. J. Peptide Protein Res. 1993, 41, 611.
J.-P. Vigneron, J.-M. Lehn, P. Lehn, manuscript to be pub-
lished.
¸
Brossier, Tetrahedron 1993, 49, 3109Ϫ3116.
[33] [33a] T. J. Baker, N. W. Luedtke, Y. Tor, M. J. Goodman, J. Org.
[36]
[33b]
Chem. 2000, 65, 9054Ϫ9058.
L. Sings, M. Goodman, J. Org. Chem. 1998, 63, 3804Ϫ3805.
K. Feichtinger, C. Zapf, H.
Received March 11, 2003
2774
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2764Ϫ2774