The synthesis of an amine-bearing polymerizable phospholipid

Article abstract of DOI:10.1016/j.tetlet.2011.06.031

Exchange of the natural glycerol phospholipid backbone with an artificial 1,3-diaminopropanol backbone led to the first synthetic 1,3-diaminophospholipid. The amines in the polar head group region were reacted to acrylamides giving a UV polymerizable phospholipid. Preliminary experiments demonstrate that formulated vesicles can be polymerized into large aggregates of vesicles.

Full text of DOI:10.1016/j.tetlet.2011.06.031

Tetrahedron Letters 52 (2011) 4215–4217
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
The synthesis of an amine-bearing polymerizable phospholipid
⇑
Pierre-Léonard Zaffalon, Etienne Stalder, Illya A. Fedotenko, France Favarger, Andreas Zumbuehl
Department of Organic Chemistry, University of Geneva, Quai Ernest-Ansermet 30, 1211 Geneva, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Exchange of the natural glycerol phospholipid backbone with an artificial 1,3-diaminopropanol backbone
led to the first synthetic 1,3-diaminophospholipid. The amines in the polar head group region were
reacted to acrylamides giving a UV polymerizable phospholipid. Preliminary experiments demonstrate
that formulated vesicles can be polymerized into large aggregates of vesicles.
Received 3 May 2011
Revised 31 May 2011
Accepted 7 June 2011
Available online 14 June 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Artificial phospholipids
Phospholipid synthesis
Amine-bearing phospholipids
Polymerizable phospholipids
Polymeric vesicles
Nature provides a wealth of thousands of chemically distinct li-
pid structures.1 However, a closer look at glycerophospholipids and
sphingolipids reveals that this variety is mainly based on combina-
tions of different specific head groups and tails.2,3 The backbone
linking the hydrophilic and hydrophobic parts is conserved and
consists either of glycerol or 2-amino-1,3-propanediol. Expanding
the set of available backbone structures allows chemistries that
are not possible with natural phospholipids.
Here, we introduce a 1,3-diaminophospholipid with acrylamide
functionalities at the backbone. We show that these phospholipids
form vesicles and undergo rapid UV polymerization to form giant
vesosome-like structures (vesicles inside vesicles).4 This approach
overcomes a limitation of diacetylene lipids which only polymerize
in a solid ordered state5 and expands the reactivity of dienoyl
phospholipids that polymerize into individual vesicles.6
Polymerizable phospholipids were intensely studied in the
1980s. One of the main targets was to increase the stability and
to control the properties of liposomes in order to use them as ther-
apeutic tools and membrane models. The concept was introduced
in 1980 in a series of papers by Chapman, Regen, and Rings-
dorf.7–9 Whereas Regen’s initial solution was a polymerizable
long-chain ammonium salt, Ringsdorf and Chapman presented
the synthesis of two very similar polymerizable diacetylene lipids
which greatly resemble natural phosphatidylcholines. The diacety-
lenic moieties were less reactive compared to free diacetylenic
fatty acids but nonetheless significant polymerization could be in-
duced at 254 nm as long as temperatures were below the Tm. In the
following decades many more polymerizable phospholipids have
been synthesized and the work has been reviewed in several excel-
lent articles.10–14
Besides using diacetylene-modified acyl chains, polymerization
has been achieved with reactive groups placed at the end of the
acyl chains, near the backbone, and at the head group of phospho-
lipids. With this the chapter of polymerizable phospholipids has
been closed under the notion that both head group and tail poly-
merizations significantly influence membrane biophysics. How-
ever, there have been no reports on polymerizations from the
backbone of a phospholipid. We reasoned that only placing the
polymerizable groups at the backbone between the hydrophobic
and hydrophilic parts of a phospholipid would yield polymeric
membranes that retain much of the original lipid’s biophysical
properties. Furthermore, by replacing the diacetylene moiety with
the more tolerant acrylamide group would allow UV polymeriza-
tion at room temperature. Also, by switching from a 1,2 to a 1,3
pattern, the polymerizable groups should be further placed in an
optimal alignment without changing the physical properties of
the phospholipids.15 Overall, we believe that the chapter of poly-
merizable phospholipids has been closed prematurely before all
possibilities of the technology have been properly assessed (this
notion is corroborated by a recent review on phosphate-containing
polymers).16 This Letter summarizes what we think is a promising
new direction in the field: polymerizations based on the artificial
phospholipid PanAc-PC-PanAc (6 in Scheme 1).
We have recently started the synthesis of different artificial
phospholipids and found it necessary to amend the current
nomenclature for phospholipids.17 Along with others we have sug-
gested keeping the standard naming scheme but introducing a
two-letter code designating the nature of the chemical bond
through which the fatty acid is bound to the glycerol-type back-
⇑
Corresponding author.
E-mail address: andreas.zumbuehl@unige.ch (A. Zumbuehl).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.06.031

4216
P.-L. Zaffalon et al. / Tetrahedron Letters 52 (2011) 4215–4217
Scheme 1. Synthesis of PanAc-PC-PanAc (6). (a) Benzoyl chloride, Et3N, CH2Cl2, 12 h, 93%; (b) LiAlH4, THF, 14 h, 94%; (c) epichlorohydrin, EtOH, 1 h, 150 °C, microwave, 97%;
(d) (1) Benzyloxydichlorophosphine (BODP), Et3N, THF, 20 h; (2) Bromoethanol, 1 h; (3) m-CPBA, 1 h; 2–10% over three steps; (e) CF3COOH, CH2Cl2, 2 h; (f) NMe3, CHCl3,
iPrOH, CH3CN, 24 h, 56% over two steps; (g) H2, Pd/C, MeOH, 24 h, 88%; (h) acryloyl chloride, Et3N, THF, 14 h, 43%.
bone.18 The molecule reported in this communication is PanAC-PC-
PanAc (6), a derivative of the 1,3-substituted C16 (P) diamino (an)
Pan-PC-Pan (5).
times through a 100 nm filter at 60 °C.22 The formation of vesicles
(that can be nontrivial for artificial amphiphiles23) was supported
by transmission electron microscopy experiments and dynamic
light-scattering measurements (see Supplementary data). Upon
standing for several hours, a precipitate formed and vesicle aggre-
gates with a mean diameter of 600 nm were visible with dynamic
light scattering. Here, we, therefore, worked with freshly extruded
populations of vesicles. We deposited an aliquot of 18 lL of the
vesicle suspension onto a scanning electron microscopy (SEM) alu-
minum stub fitted with double-sided carbon tape. To this was
added 2 lL of Irgacure 2959 and the aluminum stub was placed
The synthesis started with the assembly of the phospholipid tail
region: benzyl protected hexadecylamine was prepared using a lit-
erature protocol19 and reacted with rac-epichlorohydrin under
microwave conditions to give the hydrophobic part (2) of the tar-
get molecule (see Scheme 1). The unprotected Pan-OH-Pan is insol-
uble. Several methods for the installation of phosphate head group
were investigated. Typical conditions like POCl3 or 2-bromoethyl
phosphorodichloridate did not yield any product.20 Similarly,
PCl3, 2-chloro-1,3,2-dioxaphospholane, and phosphoramidite re-
agents did not lead to success. Finally, benzyloxydichlorophos-
phine (BODP) was found to be both reactive and surprisingly
stable, requiring only moderately dry working conditions on the
vacuum-line. For ether and ester phospholipids we have obtained
good yields for the phosphorylation of secondary alcohols.21 Using
the same conditions on amino-lipids we found much lower yields,
due to the steric hindrance from the benzyl protecting groups, N-
oxitative cleavage of the benzyl protecting group, and also due to
repetitive purifications we performed in order to obtain clean ana-
lytical spectra. In the same reaction vessel, BODP was reacted first
with alcohol 2 followed by bromoethanol. After benzyl deprotec-
tion of the phosphate 3, the bromoethanol was substituted by tri-
methylamine using an established protocol.19 Both amines of the
intermediate 4 were then deprotected and because of the low yield
of the reaction, the free Pan-PC-Pan (5) was directly reacted with
acryloyl chloride to give the final compound PanAc-PC-PanAc (6).
Of course, Pan-PC-Pan (5) will be an interesting phospholipid on
its own, and its biophysics will be addressed in a future report.
In order to assess the potential of PanAc-PC-PanAc (6) for mak-
ing polymerized vesicles, preliminary experiments were per-
formed: vesicles were prepared using standard conditions: a thin
film was formed when the PanAc-PC-PanAc (6) was taken up in
chloroform and the solvent was removed under reduced pressure.
The organic solvent was removed under high-vacuum overnight.
Then the thin film was hydrated with buffer at a concentration of
under a 100 W UV light (365 nm) for 15 min. The polymerized
Figure 1. (A) Schematic of the polymerization of single vesicles into vesosome-like
structures. The gray area represents polymerized material from destroyed vesicles.
The vesicles in the aggregate on the right are not drawn to scale. (B and C)
Polymerized PanAc-PC-PanAc (6) giant vesicle aggregates.
10 mg lipid per 500
lL buffer (107 mM NaCl, 10 mM HEPES,
pH = 7.4), submitted to five freeze-thaw cycles and extruded 11

P.-L. Zaffalon et al. / Tetrahedron Letters 52 (2011) 4215–4217
4217
In conclusion we have presented the synthesis of the first arti-
ficial 1,3-diaminophospholipid and have shown that it readily
undergoes UV polymerization into polymerized aggregates of ves-
icles. Polymerizable and nonpolymerizable phospholipids can be
formulated into vesicles to give an interesting, tunable
nanomaterial.
Acknowledgments
The authors thank the University of Geneva and the Swiss Na-
tional Science Foundation (200021-121718) for the financial sup-
port and acknowledge the contributions of the Bioimaging, Mass
Spectrometry, and NMR platforms at the Faculty of Sciences, Uni-
versity of Geneva.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2011.06.031.
Figure 2. (A and B) Inner geometry of the polymer revealed after ripping away
parts of the surface with scotch tape. Individual vesicular structures are visible. (C
and D) Vesicles in the process of being polymerized into a superstructure.
References and notes
1. Wenk, M. R. Cell 2010, 143, 888–895.
2. Zumbuehl, A. Chimia 2009, 63, 63–65.
3. Fahy, E.; Subramaniam, S.; Brown, H. A.; Glass, C. K.; Merill, A. H., Jr.; Murphy, R.
C.; Raetz, C. R. H.; Russell, D. W.; Seyama, Y.; Shaw, W.; Shimizu, T.; Spener, F.;
van Meer, G.; van Nieuwenhze, M. S.; White, S. H.; Witztum, J. L.; Dennis, E. A. J.
Lipid Res. 2005, 46, 839–862.
material was then washed with pure water and left to dry before it
was gold-coated and studied under SEM conditions (see Fig. 1).
The polymerization did not lead to individual 100 nm surface
polymerized vesicles but to large aggregates (ca. 150
eter), many times the size of individual vesicles, reminding us of
vesosomes (vesicles inside vesicles) or aggregates of vesicles.4
l
m in diam-
4. Walker, S. A.; Kennedy, M. T.; Zasadzinski, J. A. Nature 1997, 387, 61–64.
5. Morigaki, K.; Baumgart, T.; Offenhausser, A.; Knoll, W. Angew. Chem., Int. Ed.
Engl. 2001, 40, 172–174.
6. Tsuchida, E.; Hatashita, M.; Makino, C.; Hasegawa, E.; Kimura, N.
Macromolecules 1992, 25, 207–212.
7. Johnston, D. S.; Sanghera, S.; Ponsand, M.; Chapman, D. Biochim. Biophys. Acta
1980, 602, 57–69.
8. Regen, S. L.; Czechand, B.; Singh, A. J. Am. Chem. Soc. 1980, 102, 6638–6640.
9. Hub, H.-H.; Hupfer, B.; Kochand, H.; Ringsdorf, H. Angew. Chem., Int. Ed. Engl.
1980, 19, 938–940.
A
simple test using scotch tape that was ripped off a polymerized
SEM sample revealed that indeed we had a polymeric material
formed from hundreds of cross-linked vesicles (see Fig. 2A and
B). Nonpolymerized vesicles would not be visible in the dry,
high-vacuum conditions used. Vesosomes consist of discrete vesi-
cles encapsulated in a larger vesicle. Here, we do not find discrete
vesicles anymore but parts of each vesicle’s outer membrane layer
seem to have reacted with other vesicles. The gap between two
vesicles is filled with a polymer that must come from destroyed
vesicles. Also, the size of the holes is larger than the initial vesicle
size hinting at a dynamic process between the vesicles during
polymerization.
10. Ringsdorf, H.; Schlarb, B.; Venzmer, J. Angew. Chem., Int. Ed. Engl. 1988, 27, 113–
158.
11. Armitage, B.; Bennett, D.; Lamparski, H.; O’Brien, D. In Advances in polymer
science; Springer: Berlin/Heidelberg, 1996; vol. 126, 53-84.
12. O’Brien, D. F.; Armitage, B.; Benedicto, A.; Bennett, D. E.; Lamparski, H. G.; Lee,
Y.; Srisri, W.; Sisson, T. M. Acc. Chem. Res. 1998, 31, 861–868.
13. Nakaya, T.; Li, Y. J. Prog. Polym. Sci. 1999, 24, 143–181.
14. Cashionand, M. P.; Long, T. E. Acc. Chem. Res. 2009, 42, 1016–1025.
15. Seelig, J.; Dijkman, R.; Haas, G. H. D. Biochemistry 1980, 19, 2215–2219.
16. Monge, S.; Canniccioni, B.; Graillot, A.; Robin, J.-J. Biomacromolecules 2011, 12,
1973–1982.
A series of polymerizations was carried out at higher dilution
(0.1 mg lipid per 500
l
L of buffer). At this concentration we could
17. Fedotenko, I. A.; Zaffalon, P. L.; Favarger, F.; Zumbuehl, A. Tetrahedron Lett.
2010, 51, 5382–5384.
18. Huang, Z.; Szoka, F. C. J. Am. Chem. Soc. 2008, 130, 15702–15712.
19. Parks, B. W.; Gilbertson, R. D.; Domaille, D. W.; Hutchison, J. E. J. Org. Chem.
2006, 71, 9622–9627.
20. Hirt, R.; Berchtold, R. Pharm. Acta Helv. 1958, 33, 349–356.
21. Zaffalon, P. L.; Zumbuehl, A. Synthesis 2011, 5, 778–782.
22. Olson, F.; Hunt, C. A.; Szoka, F. C.; Vail, W. J.; Papahadjopoulos, D. Biochim.
Biophys. Acta 1979, 557, 9–23.
not observe the formation of vesosome-like structures. Instead we
found structures that possibly capture the very initial step where
individual vesicles are polymerized into a larger aggregate (see
Fig. 2C and D). Together with the above findings, this could also
show that the polymers in Figure 1 contain vesicles. Currently, a
detailed survey of the actual geometry of the polymeric material
is ongoing.
23. Zepik, H. H.; Walde, P.; Ishikawa, T. Angew. Chem., Int. Ed. 2008, 47, 1323–1325.