Archaeosomes based on synthetic tetraether-like lipids as novel versatile gene delivery systems

Article abstract of DOI:10.1039/b618568a

Novel cationic liposomes, termed "archaeosomes", based on mixtures of neutral/cationic bilayer-forming lipids and archaeobacterial synthetic tetraether-type bipolar lipids show efficient in vitro gene transfection properties and represent a new approach f

Full text of DOI:10.1039/b618568a

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Archaeosomes based on synthetic tetraether-like lipids as novel versatile
gene delivery systems{
G. Re´thore´,a T. Montier,b T. Le Gall,b P. Dele´pine,b S. Cammas-Marion,a L. Lemie`gre,a P. Lehnb and
T. Benvegnu*a
Received (in Cambridge, UK) 20th December 2006, Accepted 27th March 2007
First published as an Advance Article on the web 11th April 2007
DOI: 10.1039/b618568a
‘‘archaeosomes’’, characterized by membranes which don’t exhibit
a bilayer structure as in conventional liposomes but a monolayer
(or a combination of monolayer and bilayer) arrangement (Fig. 1).6
The monolayer membrane organization results here from the
presence of archaeobacterial-like bipolar lipids that span the
membrane from polar headgroups on one side to headgroups on
the other side. In contrast to the double-layer cell membranes of
eukaryotes, the single-layer membranes formed by these unusual
lipids have a high degree of physical rigidity and chemical/
enzymatic stability.7 Fine-tuning of the membrane rigidity via
modulation of its lipid composition (bilayer/monolayer-forming
lipids) as in natural archaebacteria7 represents a new approach in
nonviral gene delivery.
Novel cationic liposomes, termed ‘‘archaeosomes’’, based on
mixtures of neutral/cationic bilayer-forming lipids and archae-
obacterial synthetic tetraether-type bipolar lipids show efficient
in vitro gene transfection properties and represent a new
approach for modulating the lipidic membrane fluidity of the
complexes they form with DNA.
Since the discovery of lipofection,1 cationic lipid–DNA complexes
(lipoplexes) have been widely used for gene transfection in vitro
and in vivo.2 A typical cationic lipid (CL) is basically composed of
a cationic headgroup and a lipophilic moiety connected via a
linker. Current cationic lipids include double-chained amphiphiles
and cholesterol derivatives which are generally formulated as
The archaeosomes studied in this work are based on a set of
synthetic lipids that include the cationic tetraether lipid GRcat and
the neutral tetraether GR (Fig. 2) whose structures are typical of
natural archaeobacterial membrane lipids. These artificial bipolar
lipids are actually characterized by 1) a hemicyclic lipid core
composed of two phytanyl chains and a 31 atom-long bridging
chain containing a cyclopentane ring (supposed to increase the
lipid dispersion into water)8 linked together by 2 glycerol moieties
to which they are attached via ether bonds and 2) two neutral
hydroxyl groups (GR) or two positive quaternary ammonium
groups provided by glycine beta¨ıne residues linked, via an amide
bond, to each end of the backbone (GRcat). The presence of
oxygen atoms within the spanning chain of lipids GRcat and GR
results from our aim to simplify the synthetic pathway initially
cationic liposomes characterized by
a bilayered membrane
(Fig. 1a).3 Numerous combinations of DNA with suspensions of
vesicles composed of cationic lipids and neutral ‘‘helper’’ lipids
such as DOPE (dioleoylphosphatidylethanolamine) or cholesterol,
were developed for transfection3 but their use in clinical gene
therapy trials was relatively unsatisfactory.4 This was in particular
due to the inability of conventional bilayered lipidic systems to
meet two conflicting requirements: (1) ensure a sufficient rigidity of
the lipoplexes in vivo and (2) maintain the membrane fluidity
required at various cellular stages of the transfection process.5
Taking into account the importance of membrane stability,
we have developed in the present work liposomes, termed
Fig. 1 Membrane models of conventional bilayered liposomes (a) and
monolayered (b) or mixed bilayered–monolayered (c) archaeosomes.
aENSCR, UMR CNRS 6226 Sciences Chimiques de Rennes, ‘‘Equipe
Synthe`se Organique et Syste`mes Organise´s’’, Avenue du Ge´ne´ral
Leclerc, 35700 Rennes, France.
E-mail: thierry.benvegnu@ensc-rennes.fr
bUnite´ INSERM 613, Institut de Synergie des Sciences et de la Sante´,
Faculte´ de Me´decine et des Sciences de la Sante´, Universite´ de Bretagne
Occidentale, Avenue Foch, 29220 Brest, France
{ Electronic supplementary information (ESI) available: Experimental
details. See DOI: 10.1039/b618568a
Fig. 2 Structures of the synthetic dicationic tetraether lipid GRcat,
neutral tetraether lipid GR, monocationic double-chained lipid MM1810
and standard helper lipid DOPE.
2054 | Chem. Commun., 2007, 2054–2056
This journal is ß The Royal Society of Chemistry 2007

View Online
formulations containing exclusively a cationic lipid (GRcat or
MM18) or a combination of a co-lipid (DOPE or GR) and an
aforementioned cationic lipid were prepared, the quantity of co-
lipid ranging from 0% to 15% (w/w). Low percentages of co-lipids
were used preferentially in the liposomal formulations since
addition of large amounts of a co-lipid like DOPE may result in
some cytotoxicity.10b Liposome and archaeosome sizes were
determined as mean diameter using a 3000 Zetasizer Malvern
Instrument. Clearly, MM18 liposomes were smaller (80 nm) than
the GRcat archaeosomes (240 nm), whereas the sizes of co-lipid
containing vesicles ranged from 180 nm (MM18/DOPE 95/5) to
470 nm (GRcat/DOPE 95/5) with a polydispersity index between
0.5 and 0.9 (Table 1). Thus, archaeolipids led generally to larger
vesicles than conventional lipids. It is noteworthy that the
incorporation of co-lipids also resulted in a significant size
increase. For both MM18-based and GRcat-based vesicles,
addition of a co-lipid resulted in a significant increase in their
surface charge, the zeta potential of GRcat vesicles being however
generally more positive than that of MM18 liposomes (Table 1).
The precise composition of the formulations was assessed by high
performance thin layer chromatography (HPTLC) on Silica gel 60
(Merck) using the Camag ATS4 for automatic sample application.
After plate development in an adapted mobile phase, lipids were
quantified by fluorescence at 366 nm (primuline derivatization of
plates) using the Camag scanner 3. The results showed that the
observed archaeosome lipid composition was only slightly different
from its expected theoretical composition (see ESI).
developed for tetraethers including exclusively methylene groups
within the bridging chain.9
Thus, we herein describe the synthesis of dicationic tetraether
GRcat and neutral tetraether GR and report the transfection
activity of the complexes (termed ‘‘archaeoplexes’’) they form with
plasmid DNA when used either alone or in combination with 1)
the neutral colipid DOPE in the case of the cationic lipid GRcat or
2) the typical bilayer-forming monocationic aliphatic lipid
MM1810 (Fig. 2) in the case of the neutral tetraether GR. The
aim of our approach was therefore to get a first insight into the
transfection properties of these highly innovative formulations and
thereby ascertain the potential of such monolayer-forming reagents
as cationic lipid GRcat or co-lipid GR.
First, synthesis of bipolar cationic tetraether GRcat involved
alkylation of the (S)-glycerol derivative 49 containing a phytanyl
chain with ditriflate 5 and subsequent hydrogenolysis of the
benzyloxy groups to provide the corresponding neutral tetraether
GR (Scheme 1). A double O-alkylation of 12-O-benzyl-dodecane-
1-triflate 1 with cis-1,3-bis(hydroxymethyl)cyclopentane 2 using
Proton-Sponge1 as the base allowed an efficient and rapid
generation of the corresponding bridging chain 3 (75% yield).
Finally, introduction of the cationic glycine beta¨ıne head groups
was performed by simultaneous N-acylation of diamine 6 resulting
from azidation of diol GR and subsequent hydrogenolysis, with
the N-acyl thiazolidine-2-thione derivative 710 of glycine beta¨ıne
(for detailed spectral data, see ESI).
Liposomes and archaeosomes were prepared by hydrating the
lipids with sterilized water during 12 h at 4 uC followed by
vortexing for 2 min and sonication for 10 min. Thus, various
Cationic lipid–DNA complexes were prepared by mixing the
appropriate amount of aqueous liposome or archaeosome
suspensions with plasmid DNA (pTG11033, 9.7 kb) expressing
the luciferase reporter gene and analyzed after 30 min of
incubation at room temperature. Practically, to a fixed amount
of DNA (4 mg), we added increasing amounts of cationic MM18,
MM18/GR, MM18/DOPE, GRCat/DOPE or GRCat formula-
tions to obtain lipoplexes with increasing (¡) mean theoretical
charge ratios R. We next measured the size and the zeta potential
of the resulting complexes in the culture medium (OptiMem).
Schematically, the size of the lipoplexes was not dramatically
modified in comparison with the corresponding unreacted
liposomes and, as expected, the zeta potential values varied in
agreement with the charge ratios of the complex (data not shown).
In addition to these physicochemical studies, we also performed
in vitro transfection experiments to get some insight into the
potential of archaeosome-containing formulations for gene
transfection. Practically, transfection experiments were carried
out with the human alveolar epithelial cell line A549, a model
Table 1 Mean diameters and zeta potentials of liposomes and
archaeosomes measured by using a 3000 Zetasizer Malvern
Instrument at 20 uC after an appropriate dilution of the formulations
Average Polydispersity Zeta potential
(mV)
diameter (nm) index
MM18
MM18/DOPE 5%
MM18/DOPE 15% 300
MM18/GR 5%
MM18/GR 15%
GRcat
80
180
0.5
0.6
0.9
0.7
0.6
0.4
0.7
0.5
16
45
45
43
58
42
61
60
Scheme 1 Reagents and conditions: (i) Proton Sponge, CH2Cl2, reflux,
75%; (ii) H2, Pd/C, cyclohexane, 90%; (iii) KH, THF, 0 uC, 50–55%; (iv)
H2, Pd/C, AcOEt, quantitative; (v) MsCl, NEt3, CH2Cl2, 90–95%; (vi)
NaN3, TBABr, DMF, reflux, 85–90%; (vii) Pd/C, H2, THF–EtOH (1/1 v/
v), 90–95%; (viii) 7, NEt3, DMF, 75–80%. TBABr: tetrabutylammonium
bromide; MsCl: methanesulfonyl chloride.
250
190
240
470
GRcat/DOPE 5%
GRcat/DOPE 15% 310
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 2054–2056 | 2055

View Online
(such as membrane fluidity) required for efficient gene transfection.
Taken altogether, these results strongly suggest that an optimal
lipofection activity may require the reaching of a compromise
between fluidity and stability of the lipoplex membranes and that
archaeobacterial-like lipids may constitute interesting tools for
reaching such a compromise. Accordingly, we are at present
studying GRcat formulations containing a variety of archaeobac-
terial-like neutral co-lipids in order to assess more precisely the
transfection activity of archaeosomes characterized by highly rigid
membranes. More generally, a better understanding of the
mechanisms underlying the transfection activity of such systems
is clearly necessary and will probably require sophisticated
physicochemical experiments as well as additional biological
investigations, such as studies on the interaction between
archaeoplexes and the cellular machinery.
Fig. 3 Transfection efficiency for formulations containing 0, 5 or 15 wt%
of neutral co-lipid on A549 cells for 4 mg DNA delivered (charge ratio
R = 4 to 8).
system for lung gene therapy.11 The commercially available
cationic lipid Lipofectamine was used as positive control whereas
unreacted (‘‘naked’’) plasmid DNA and untreated cells were used
as negative controls. Expression of the transfected luciferase
reporter gene was quantified by luminometry (MLX Dynex), the
results being expressed as Relative Light Units (RLU) per mg of
total protein (Fig. 3).
In summary, we have developed archaeobacterial-like lipids
which can be used as cationic lipids or co-lipids for in vitro gene
transfection. Our work demonstrates the potential of combining
conventional bilayer-forming lipids with monolayer-forming
lipids as a new strategy to modulate the membrane properties of
CL–DNA complexes. In ongoing studies, we are at present
evaluating the potential of our novel archaeoplexes for in vivo gene
transfection into the airway epithelium by nasal instillation or
aerosolization, with a view to lung-directed gene therapy for cystic
fibrosis.
The results showed that archaeobacterial-like lipids can be used
for lipofection either as co-lipids or cationic lipids, the transfection
activity being dependent on the nature of the lipids and
archaeolipids present in the formulation, the cationic lipid/neutral
co-lipid (w/w) ratio, and the (¡) charge ratio, clearly positive
charge ratios being assumed to allow efficient interaction of the
complexes with negative cell surface residues (Fig. 3). First,
liposomal formulations of MM18 with small amounts of the
neutral tetraether GR were more efficient than pure MM18
liposomes and MM18 formulations containing the classical co-
lipid DOPE. The MM18/GR 95/5 (w/w) formulation was actually
as efficient as the widely used reagent Lipofectamine (at a +8
charge ratio i.e. at a high amount of cationic lipids). These data
demonstrate that the neutral lipid GR can be used as a co-lipid
and it represents the most favourable helper lipid in these in vitro
transfection experiments. Here, although the addition of DOPE or
GR to MM18 led to similar increases in zeta potential values and
consequently possibly of transfection activity, the fact that the
highest transfection activity was observed with the GR-containing
formulations suggests that addition of GR may provide MM18-
based complexes with additional properties beneficial for transfec-
tion. These results were quite unexpected since this bipolar lipid
should have a lower tendency to form an inverted hexagonal phase
(HII) at acidic pH, which is usually indicative of an efficient escape
from lysosomal degradation and cytoplasmatic release of DNA.12
Next, as also shown in Fig. 3, the cationic tetraether GRcat
could mediate significant gene transfection when used as a cationic
liposome formulation with DOPE. GRcat was, however, ineffec-
tive alone, although it could efficiently bind plasmid DNA as
shown by gel retardation (see ESI) and although GRcat
archaeoplexes exhibited a clearly positive zeta potential (Table 1).
Interestingly, these data suggest that the addition of DOPE
provides the GRcat archaeosomes with physicochemical properties
We are grateful to the Re´gion Bretagne for grants to GR and
TLG. This work was also supported by ‘‘Vaincre La
Mucoviscidose’’ (Paris, France).
Notes and references
1 P. L. Felgner, T. R. Gadek, M. Holm, R. Roman, H. W. Chan,
M. Wenz, J. P. Northrop, G. M. Ringold and M. Danielsen, Proc. Natl.
Acad. Sci. U. S. A., 1987, 84, 7413.
2 (a) B. Martin, M. Sainlos, A. Aissaoui, N. Oudrhiri, M. Hauchecorne,
J. P. Vigneron, J. M. Lehn and P. Lehn, Curr. Pharm. Des., 2005, 11,
375; (b) A. D. Miller, Angew. Chem., Int. Ed., 1998, 37, 1769; (c) X. Gao
and L. Huang, Gene Ther., 1995, 2, 710.
3 S. Zhang, Y. Xu, B. Wang, W. Qiao, D. Liu and Z. Li, J. Controlled
Release, 2004, 100, 165.
4 Y. Barenholz, Curr. Opin. Colloid Interface Sci., 2001, 6, 66.
5 K. Matsui, S. Sando, T. Sera, Y. Aoyama, Y. Sasaki, T. Komatsu,
T. Terashima and J. I. Kikuchi, J. Am. Chem. Soc., 2996, 128, 3114.
6 (a) M. de Rosa, A. Gambarcorta and A. Gliozzi, Microbiol. Rev., 1986,
50, 78; (b) T. Benvegnu, M. Brard and D. Plusquellec, Curr. Opin.
Colloid Interface Sci., 2004, 8, 469.
7 G. B. Patel and G. D. Sprott, Crit. Rev. Biotechnol., 1999, 19, 317.
8 M. Brard, W. Richter, T. Benvegnu and D. Plusquellec, J. Am. Chem.
Soc., 2004, 126, 10003.
9 T. Benvegnu, G. Rethore, M. Brard, W. Richter and D. Plusquellec,
Chem. Commun., 2005, 5536.
10 (a) V. Floch, N. Legros, S. Loisel, C. Guillaume, J. Guilbot,
T. Benvegnu, V. Ferrieres, D. Plusquellec and C. Ferec, Biochem.
Biophys. Res. Commun., 1998, 251, 360; (b) D. Gilot, M.-L. Miramon,
T. Benvegnu, V. Ferrieres, O. Loreal, C. Guguen-Guillouzo,
D. Plusquellec and P. Loyer, J. Gene Med., 2002, 4, 415.
11 M. T. Cushion, J. J. Ruffolo and P. D. Walzer, Lab. Invest., 1988, 58,
324.
12 (a) C. R. Safinya, Curr. Opin. Struct. Biol., 2001, 11, 440; (b) D. Hirsch-
Lerner, M. Zhang, H. Eliyahu, M. E. Ferrari, C. J. Wheeler and
Y. Barenholz, Biochim. Biophys. Acta, 2005, 1714, 71.
2056 | Chem. Commun., 2007, 2054–2056
This journal is ß The Royal Society of Chemistry 2007