Poly(choloyl)-based amphiphiles as pore-forming agents: Transport-active monomers by design

Article abstract of DOI:10.1021/ja053527q

This paper describes the design and synthesis of a family of pore-forming amphiphiles. Two of these amphiphiles, which are derived from cholic acid, lysine, and p-phenylenediamine, can produce pores in lipid bilayers as individual molecules. In sharp cont

Full text of DOI:10.1021/ja053527q

Published on Web 08/17/2005
Poly(choloyl)-Based Amphiphiles as Pore-Forming Agents:
Transport-Active Monomers by Design
Wen-Hua Chen, Xue-Bin Shao, and Steven L. Regen*
Contribution from the Department of Chemistry, Lehigh UniVersity,
Bethlehem, PennsylVania 18015
Received May 30, 2005; E-mail: slr0@lehigh.edu
Abstract: This paper describes the design and synthesis of a family of pore-forming amphiphiles. Two of
these amphiphiles, which are derived from cholic acid, lysine, and p-phenylenediamine, can produce pores
in lipid bilayers as individual molecules. In sharp contrast, analogous amphiphiles that do not contain a
rigid 1,4-phenylenediamide moiety favor the formation of dimer-based pores. Kinetic evidence in support
of monomer- and dimer-based pores has been obtained from Na+ transport measurements across bilayers
made from 1-palmitoyl-2-oleoyl-2-sn-glycero-3-phosphocholine (POPC). Structure-activity studies that have
been carried out with pore-forming, dimer-based amphiphiles have also revealed a significant activity
dependence on their overall compactness. The practical potential of pore-forming amphiphiles with
controllable supramolecular properties is briefly discussed.
Introduction
In recent studies we have shown that certain conjugates
derived from cholic acid, putrescine, spermidine, spermine, and
The ability to fine tune the pore-forming properties of
synthetic amphiphiles offers the promise of a multitude of
applications. Pore-forming amphiphiles that can selectively insert
into the plasma membrane of bacterial and fungal cells relative
to mammalian cells, for example, may lead to new classes of
antibiotics having minimal susceptibility toward drug resis-
tance.1,2 The inclusion of pore-forming amphiphiles in liposomes
loaded with enzymes or nonbiogenic catalysts has the potential
for use as nanoreactors and chemical sensors.3 The recent
demonstration that a pore-forming protein can recognize a
specific base along an individual strand of DNA suggests that
large families of pores of different sizes and shapes could result
in devices for DNA sequencing.4,5 Pore-forming amphiphiles
that can enhance the release of agents from thermally sensitive
liposomes (i.e., thermally gated liposomes) could also be
considered for applications ranging from the targeted delivery
of drugs to the controlled release of flavors in foods.6 Despite
such potential, the ability to create pore-forming amphiphiles
with perfect control over the size, shape, and supramolecular
structure of the resulting pores remains as a major synthetic
challenge.7
lysine form dimer-based pores in lipid bilayers having Na+
transport activities that increase, exponentially, with increasing
numbers of choloyl units up to 12 sterols per conjugate (e.g.,
1, Chart 1).6,8 Because of this unusually wide range in activity,
we have become keenly interested in developing this class of
compounds in the broadest possible sense. In the present
investigation our primary goal was to design and synthesize
pore-forming poly(choloyl)-based amphiphiles which do not rely
on supramolecular behavior (i.e., aggregation) for pore forma-
tion; that is, they can form pores as indiVidual molecules. A
secondary objective of this work was to measure the sensitivity
of ion transport activity of poly(choloyl)-based amphiphiles with
respect to their overall compactness.
Since a poly(choloyl)-based amphiphile that can form a pore
as an individual molecule does not involve monomer-aggregate
equilibria, its activity is expected to be more controllable. For
an analogous amphiphile that requires aggregation for pore
formation, however, the number, size, and composition of the
resulting pores is expected to be a function of the microenvi-
ronment that exists within the membrane. Thus, depending upon
the composition of a given target or host membrane, aggregation
and pore formation may or may not be favored. In certain
instances, such a dependency may be desirable. For example,
an amphiphile that favors pores in bacterial membranes that
are devoid of sterols, relative to mammalian membranes, which
are rich in cholesterol, could lead to useful antibiotics. For other
applications, where control over pore size and the number of
pores is important (e.g., nanoreactors, chemical sensors, and
DNA sequencing), unimolecular pore-forming amphiphiles that
do not depend on aggregation would clearly be preferred.
(1) Stadler, E.; Dedek, P.; Yamashita, K.; Regen, S. L. J. Am. Chem. Soc.
1994, 116, 6677-6682.
(2) Yamashita, Y.; Janout, V.; Bernard, E. M.; Armstrong, D.; Regen, S. L. J.
Am. Chem. Soc. 1995, 117, 6249-6253.
(3) Treyer, M.; Walde, P.; Oberholzer, T. Langmuir 2002, 18, 1043-1050.
(4) Ashkenasy, N.; Sanchez-Quesada, J.; Bayley, H.; Ghadiri, M. R. Angew.
Chem., Int. Ed. 2005, 44, 1401-1404.
(5) Walt, D. R. Science 2005, 308, 217-219.
(6) Chen, W.-H.; Regen, S. L. J. Am. Chem. Soc. 2005, 127, 6238-6239.
(7) For reviews of synthetic ionophores, see: (a) Matile, S. Tetrahedron 2004,
60, 6405-6435. (b) Gokel, G. W.; Mukhopadhyay, A. Chem. Soc., ReV.
2001, 30, 274-286. (c) Kobuke, Y. AdVances in Supramolecular Chemistry;
Gokel, G. W., Ed.; JAI Press: Greenwich, CT, 1997; Vol 4, pp 163-210.
(d) Fyles, T. M.; van Straaten-Nijenhuis, W. F. ComprehensiVe Supra-
molecular Chemistry; Reinhoudt, D. N., Ed.; Elsevier Science: Amsterdam,
New York, 1996; Vol. 10, pp 53-77.
(8) Zhang, J.; Jing, B.; Regen, S. L. J. Am. Chem. Soc. 2003, 125, 13984-
13987.
9
10.1021/ja053527q CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 12727-12735
12727

A R T I C L E S
Chart 1
Chen et al.
layer. After washing with water, the chloroform layer was dried over
anhydrous sodium sulfate. Removal of solvent under reduced pressure
and purification by column chromatography on a silica gel column,
In addition to controlling the supramolecular properties of a
poly(choloyl)-based amphiphile, we have also become interested
in the possibility of modulating ion transport activity by
adjusting the overall compactness of such molecules. To our
knowledge, such a parameter has not previously been examined
for any synthetic ionophore reported to date.7
1
eluting with CHCl3/MeOH (40/1, v/v), afforded 7 (369 mg, 79%). H
NMR (500 MHz, CD3OD, 298 K) δ: 4.01 (br, 1 H), 3.93 (br, 1 H),
3.20 (m, 2H), 3.02 (t, 3J ) 6.78 and 6.91 Hz, 2 H), 1.79-1.30 (m, 39
H).
Compound 8. To a solution of 7 (364 mg, 0.642 mmol) and
3-hydroxy-1,2,3-benzotriazine-4(3H)-one (DHBT) (115 mg, 0.705
mmol) in anhydrous THF (10 mL) was added DCC (165 mg, 0.80
mmol). After the mixture was stirred for 5 h, a solution of p-
phenylenediamine (31 mg, 0.287 mmol) and triethylamine (0.5 mL) in
anhydrous THF (1.0 mL) was added. After stirring at room temperature
for 12 h, another portion of DCC (165 mg) was added and the stirring
continued at room temperature for 12 h. The urea that formed was
removed by filtration. The filtrate was then concentrated and purified
by chromatography on a silica gel column, eluting with CHCl3/MeOH
Experimental Section
1
General Methods. H NMR spectra were recorded in either CD3-
OD or CDCl3 using a Bruker 360 or 500 spectrometer, using the solvent
as a reference. Silica gel 60 (Fluka, particle size 0.035-0.070 mm,
220-440 mesh) was used for all column chromatographic purifications.
Thin-layer chromatography (TLC) was performed on a glass plate
precoated with silica gel and a fluorescence indicator (Whatman).
Detection on TLC was made by use of sulfuric acid 10% in water,
iodine, and UV (254 or 365 nm). All reagents and chemicals were
obtained from commercial sources and used as received unless
otherwise stated. Conjugates 1 and 3 were prepared according to the
procedures recently described by us.6,8
1
(40/1, v/v), to afford 8 (334 mg, 95%). H NMR (500 MHz, CD3OD,
318 K) δ: 7.50 (s, 4 H), 4.11 (br, 2 H), 3.94 (br, 2 H), 3.22-3.18 (m,
3
4 H), 3.01 (t, J ) 6.74 and 6.85 Hz, 4 H), 1.69-1.33 (m, 78 H).
L-Lysine-Dicholamide. The procedure used for the synthesis of
L-lysine-dicholamide was similar to that previously described.6,8,9 Thus,
to a solution of cholic acid (6.08 g, 14.9 mmol) and N-hydroxy-
succinimide (1.90 g, 16.5 mmol) in anhydrous THF (50 mL) was added
DCC (3.60 g, 17.4 mmol). The resulting mixture was stirred at room
temperature. After 4 h the product mixture was filtered to remove the
insoluble urea. The filtrate was then added, dropwise, to a solution of
L-lysine (950 mg, 6.50 mmol) and triethylamine (6.2 mL) in H2O (20
mL) and stirred overnight. The solution was concentrated under reduced
pressure and poured into 1 M hydrochloride acid (200 mL). The
resulting solid was collected by filtration. The residue was then purified
by chromatography on a silica gel column, eluting with CHCl3/MeOH/
H2O (40/10/1, v/v/v), to afford L-lysine-dicholamide (4.51 g, 75%)
MALDI-TOF MS m/z 1244 ([M + Na]+).
Conjugate 2. A solution of 8 (60 mg, 0.049 mmol) in trifluoroacetic
acid (3 mL) was stirred at room temperature for 3 h and then
concentrated under reduced pressure. The residue was dried under
reduced pressure for 4 h and used directly in the following reaction
without further purification. The complete removal of Boc, yielding 9,
was confirmed by 1H NMR (500 MHz, CD3OD, 318 K) δ: 7.57 (s, 4
3
3
H), 4.02 (t, 2 H, J ) 6.68 and 6.44 Hz), 3.83 (t, 2 H, J ) 7.14 and
3
6.58 Hz), 3.27-3.20 (m, 2H), 2.92 (t, 4 H, J ) 7.63 and 7.77 Hz),
2.04-1.82 (m, 8 H), 1.70-1.45 (m, 16 H).
To a solution of L-lysine-dicholamide (340 mg, 0.367 mmol) and
N-hydroxysuccinimide (60 mg, 0.522 mmol) in anhydrous DMF (2 mL)
was added DCC (95 mg, 0.46 mmol). After the mixture was stirred
for 3 h, a solution of 9, which was prepared from 8 (60 mg, 0.049
mmol), and 4-(dimethylamino)pyridine (180 mg, 1.473 mmol) in
anhydrous DMF (2.0 mL) was added. The reaction mixture was stirred
at room temperature for 24 h, concentrated under reduced pressure,
and poured into 1 M hydrochloride acid (200 mL). The resulting
precipitate was collected by filtration and purified twice by chroma-
tography on a silica gel column, eluting with CHCl3/CH3OH/H2O (30/
10/1, v/v), to give 2 (69 mg, 23%). 1H NMR (500 MHz, CD3OD, 328
K) δ: 7.66 and 7.55 (s, 4 H), 4.54-4.25 (m, 10 H), 3.93 (s, 12 H),
3.78 (s, 12 H), 3.37-3.34 (m, 12 H), 3.15 (m, 20 H), 2.31-0.91 (m,
1
having the expected H NMR spectrum.6,8,9
Compound 7. A solution of Boc-L-Lys-OH (200 mg, 0.812 mmol)
and NRNꢀ-di-Boc-L-lysine hydroxylsuccinimide ester (Boc-L-Lys(Boc)-
OSu) (317 mg, 0.834 mmol) in THF-H2O (16.5 mL, 10/1, v/v) was
stirred at room temperature for 2 h, and the product mixture was then
concentrated under reduced pressure. The resulting residue was
partitioned between water and chloroform, and the organic layer was
separated and combined with chloroform extracts (3×) of the aqueous
(9) Bandyopadhyay, P.; Bandyopadhyay, P.; Regen, S. L. Bioconjugate Chem.
2002, 13, 1314-1318.
9
12728 J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005

Poly(choloyl)-Based Amphiphiles
A R T I C L E S
396 H), 0.69 (s, 36 H). MALDI-TOF MS m/z 6099 ([M + Na]+). The
sample was judged to be homogeneous based on the appearance of a
single TLC spot (developed by two or more eluents) and a comparison
of the integrated NMR signals for the lysine protons (4.54-4.25 ppm)
relative to the 3-position protons of choloyl units (3.93 ppm).
After removal of solvent under reduced pressure, the deprotected
product (appearing as an oil) was used directly in the next step. To a
solution of cholic acid (410 mg, 1 mmol) in anhydrous DMF (5 mL)
was added DHBT (200 mg, 1.2 mmol) and DCC (240 mg, 1.2 mmol).
After stirring the mixture for 3 h at room temperature, the deprotected
form of 13 and triethylamine (0.1 mL) were added. The reaction mixture
was stirred overnight at room temperature and then poured into diluted
aqueous HCl (50 mL). The resulting precipitate was collected by
filtration and purified by chromatography on a silica gel column, eluting
with CHCl3/CH3OH/H2O (65/25/4, v/v/v), to give 5 (112 mg, 36%).
1H NMR (CD3OD, 500 MHz) δ: 4.26 (br, 4 H), 3.95 (br, 8 H), 3.81
(s, 8 H), 3.35-3.30 (br, 16 H), 3.18 (br, 12 H), 2.30-1.40 (m, 242 H),
1.05-0.92 (m, 64 H), 0.72 (s, 24 H). MALDI-TOF MS m/z 3862 ([M
+ Na]+). The sample was judged to be homogeneous based on the
appearance of a single TLC spot (developed by two or more eluents)
and a comparison of the integrated NMR signal for the lysine protons
(4.26 ppm) relative to the 3-position protons of choloyl units (3.95
ppm).
Compound 10. To a solution of 1,4-diaminobutane (57 mg, 0.65
mmol) and triethylamine (0.5 mL) in anhydrous DMF-THF (4 mL,
1/1, v/v) was added Boc-L-Lys(Boc)-OSu (620 mg, 1.40 mmol). The
resulting suspension was stirred at room temperature for 24 h. Removal
of solvent under reduced pressure and purification by chromatography
on a silica gel column, eluting with CHCl3/MeOH (10/1, v/v), gave 10
1
(432 mg, 90%). H NMR (500 MHz, CDCl3, 298K) δ: 6.89 (br, 2H,
CONH), 5.29 (br, 2 H, CONH), 4.67 (br, 2 H, CONH), 4.06
(unresolved, 2 H), 3.36 (unresolved, 2 H), 3.08-3.07 (unresolved, 6
H), 1.75-1.28 (m, 52 H). MALDI-TOF MS m/z 767 ([M + Na]+).
Conjugate 4. A solution of 10 (50 mg, 0.067 mmol) and trifluoro-
acetic acid (1 mL) was stirred at room temperature for 1 h. The reaction
mixture was concentrated under reduced pressure, and the residue (11)
was dried under vacuum for 4 h and used directly in the following
reaction without further purification. The complete removal of Boc
Compound 14. To a solution of 1,4-phenylenediamine (20 mg, 0.19
mmol) and triethylamine (0.30 mL) in anhydrous THF (2 mL) was
added Boc-L-Lys(Boc)-OSu (205 mg, 0.46 mmol). After the reaction
mixture was stirred at room temperature for 24 h, another portion of
Boc-L-Lys(Boc)-OSu (100 mg, 0.23 mmol) was added. The reaction
was stirred for an additional 48 h. The reaction mixture was concen-
trated under reduced pressure, and the residue was purified by
chromatography on a silica gel column, eluting with CHCl3/MeOH
1
groups, giving 11, was confirmed by H NMR (500 MHz, CD3OD,
3
298 K) δ: 3.84 (t, 2 H, J ) 6.63 and 6.63 Hz), 3.26-3.20 (m, 4 H),
2.93 (t, 4 H, 3J ) 7.66 and 7.79 Hz), 1.93-1.80 (m, 4 H), 1.73-1.68
(m, 4 H), 1.59-1.53 (m, 4 H), 1.49-1.43 (m, 4 H).
To a solution of L-lysine-dicholamide (311 mg, 0.335 mmol) and
DHBT (68 mg, 0.417 mmol) in anhydrous DMF (1 mL) was added
DCC (108 mg, 0.523 mmol). After stirring for 3 h, a solution of 11,
which was prepared from 10 (50 mg, 0.067 mmol), and triethylamine
(0.5 mL) in anhydrous DMF (1 mL) was added. The mixture was stirred
for 24 h and then poured into 1 M hydrochloride acid (200 mL). The
resulting precipitate was collected by filtration and purified twice by
chromatography on a silica gel column, eluting with CHCl3/CH3OH/
1
(40/1, v/v), to give 14 (125 mg, 88%). H NMR (500 MHz, CD3OD,
323 K) δ: 7.49 (s, 4 H), 4.12 (br, 2 H), 3.04 (t, 3J ) 6.66 and 6.76 Hz,
4 H), 1.79-1.37 (m, 48 H). MALDI-TOF MS m/z 787 ([M + Na]+).
Conjugate 6. A solution of 14 (50 mg, 0.065 mmol) in trifluoroacetic
acid (2 mL) was stirred at room temperature for 2 h and then
concentrated under reduced pressure. Removal of solvent under reduced
pressure and drying under vacuum for 4 h gave 15, which was used
directly without further purification. The complete removal of the Boc
groups was confirmed by 1H NMR (500 MHz, CD3OD, 298 K) δ: 7.59
1
H2O (40/10/1, v/v), to give 4 (120 mg, 45%). H NMR (500 MHz,
CD3OD, 323 K) δ: 4.27 (m, 6 H), 3.94 (s, 8 H), 3.79 (s, 8 H), 3.36-
3.39 (m, 8 H), 3.20-3.15 (m, 16 H), 2.29-0.91 (m, 280 H), 0.70 (s,
24 H). MALDI-TOF MS m/z 4004 ([M + Na]+). The sample was
judged to be homogeneous based on the appearance of a single TLC
spot (developed by two or more eluents) and a comparison of the
integrated NMR signals for the lysine protons (4.27 ppm) relative to
the 3-position protons of choloyl units (3.94 ppm).
3
3
(s, 4H), 4.03 (t, 2 H, J ) 6.60 and 6.50 Hz), 2.93 (t, 4 H, J ) 7.55
and 7.85 Hz), 2.01-1.95 (m, 4 H), 1.74-1.69 (m, 4 H), 1.561.51 (m,
4 H).
To a solution of L-lysine-dicholamide (290 mg, 0.31 mmol) and
DHBT (64 mg, 0.39 mmol) in anhydrous DMF (2 mL) was added DCC
(100 mg, 0.51 mmol). After the mixture was stirred for 3 h, a solution
of 15, which was prepared from 14 (50 mg, 0.065 mmol), and
triethylamine (0.5 mL) in anhydrous DMF (0.5 mL) was added. The
reaction mixture was stirred at room temperature for 24 h and then
poured into 1 M hydrochloride acid (150 mL). The resulting precipitate
was collected by filtration and purified twice by chromatography on a
silica gel column, eluting with CHCl3/CH3OH/H2O (40/10/1, v/v/v),
Compound 12. To a solution of L-lysine-dicholamide (3.7 g, 4
mmol) in anhydrous DMF (10 mL) was added DHBT (1 g, 6.4 mmol)
and DCC (825 mg, 4 mmol). After stirring for 3 h at room temperature,
Boc-L-Lys-OH (790 mg, 3.2 mmol) and triethylamine (0.2 mL) were
added. The reaction mixture was stirred for 12 h at room temperature
and then poured into diluted aqueous HCl (50 mL). The resulting
precipitate was collected by filtration and purified by chromatography
on a silica gel column, eluting with CHCl3/CH3OH/H2O (65/25/4, v/v/
1
to give 6 (147 mg, 56%). H NMR (500 MHz, CD3OD, 328 K) δ:
1
v), to give 12 (2.51 g, 68%). H NMR (CD3OD, 500 MHz) δ: 4.25
7.69 (unsolved, 2H), 7.56 (unsolved, 2H), 4.44 (unresolved, 2 H), 4.31
and 4.23 (unresolved, 4 H), 3.94 (s, 8 H), 3.79 (s, 8 H), 3.36-3.34 (m,
8 H), 3.15 (m, 12 H), 2.29-0.91 (m, 276 H), 0.70 (s, 24 H). MALDI-
TOF MS m/z 4024 ([M + Na]+). The sample was judged to be
homogeneous based on the appearance of a single TLC spot (developed
by two different eluents) and a comparison of the integrated NMR
signals for the lysine protons (4.44-4.23 ppm) relative to the 3-position
protons of choloyl units (3.94 ppm).
Vesicle Formation and Na+/K+ Transport Measurements. Typi-
cally, 2.5 mL of a 20 mg/mL solution of 1-palmitoyl-2-oleoyl-2-sn-
glycero-3-phosphocholine (POPC) in chloroform was transferred to a
Pyrex test tube. The desired amount of pore-forming conjugate was
then added from a stock solution in methanol. While rotating the tube,
the organic solvents were removed under a stream of nitrogen, resulting
in a thin lipid film. The last traces of solvent were then removed under
reduced pressure (25 °C, 12 h, <0.2 Torr). To the dried film was added
1.0 mL of a 150 mM KCl solution that was 10% D2O and 90% H2O,
and the mixture was vortexed for 1 min. The dispersion was then
(m, 1 H), 4.00 (br, 1 H), 3.94 (s, 2 H), 3.78 (s, 2 H), 3.31 (m, 2 H),
3.16-3.14 (m, 4 H), 2.29-1.38 (m, 75 H), 1.02-0.91 (m, 16 H), 0.70
(s, 6 H). MALDI-TOF MS m/z 1178 ([M + Na]+).
Compound 13. To a solution of 12 (302 mg, 0.26 mmol) in
anhydrous DMF (5 mL) was added DHBT (100 mg, 0.64 mmol) and
DCC (100 mg, 0.5 mmol). After stirring for 3 h at room temperature,
spermine (26 mg, 0.13 mmol) and triethylamine (0.1 mL) were added.
The mixture was stirred overnight at room temperature and then poured
into diluted aqueous HCl (50 mL). The resulting precipitate was
collected by filtration and purified by chromatography on a silica gel
column, eluting with CHCl3/CH3OH/H2O (65/25/4, v/v/v), to give 13
(212 mg, 66%). 1H NMR (CD3OD, 500 MHz) δ: 4.23 (br, 2 H), 3.95
(br, 6 H), 3.80 (s, 4 H), 3.35-3.17 (m, 4 H), 3.16 (br, 8 H), 2.60 (br,
8 H), 1.97-1.41 (m, 166 H), 1.10-0.92 (m, 32 H), 0.71 (s, 12 H).
MALDI-TOF MS m/z 2478 ([M + H]+).
Conjugate 5. A solution of 13 (200 mg, 0.08 mmol) in TFA and
chloroform (5 mL, 1:1, v/v) was stirred for 4 h at room temperature.
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12729

A R T I C L E S
Chen et al.
Chart 2
incubated for 5 min, followed by another 1 min of vortexing and 20
min of incubation at ambient temperature. The sample was subjected
to five freeze/thaw cycles (77/325 K), followed by sequential extrusion
through a 400 nm Nuclepore membrane (10 times) and a 200 nm
membrane (10 times). After extrusion the dispersion was incubated at
room temperature for 1.25 h. In a quartz NMR tube 1.5 mL of a 150
mM NaCl solution in 10% D2O plus 90% H2O was mixed with 0.3
mL of a shift reagent solution (10 mM DyCl3; 30 mM Na5P3O10 in
10% D2O plus 90% H2O). To this solution was added 0.75 mL of the
vesicle dispersion, and the resulting mixture was vortexed for 30 s.
23Na NMR spectra were recorded continuously at 35 °C overnight on
a Bruker AMX 360 MHz NMR instrument. Pseudo-first-order rate
constants were calculated from the change in the percentage of
encapsulated Na+ as a function of time using a curve-fitting procedure.
Results and Discussion
Design Principles. The design principle that we have
previously used in devising 1 and related homologues is shown
in Chart 2.6,8,9 In essence, a collection of facial amphiphiles is
covalently attached to a flexible backbone. When inserted into
a lipid bilayer, such a conjugate is expected to favor conforma-
tion B over A. In the case of B, the hydrophobic faces (darkened
rectangles) lie in contact with the acyl chains of neighboring
phospholipids (not shown) and the hydrophilic faces (lightly
shaded rectangles) point toward each other. Subsequent dimer-
ization across the bilayer is then expected to produce a “barrel
stave” structure (C). Kinetic studies that have been carried out
for the transport of Na+ across liposomal membranes made from
1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) for
conjugates of this type have provided support for such a
model.6,8
Chart 3
An alternate design principle that we have recently become
attracted toward is illustrated in Chart 3. Here, a small, rigid,
and planar linker is used to connect two “half barrels” to yield
a “whole barrel”. It should be noted that this design principle
bears a strong resemblance to that previously used in creating
an artificial ionophore via the attachment of two hydroxylated
sterols to a terephthaloyl diester spacer.10 In that study it was
postulated that such a molecule would favor a membrane-
spanning orientation in the bilayer. On the basis of a kinetic
analysis for Na+ transport, approximately three such amphiphiles
appear to combine to form a transport-active species.10 A key
difference with the approach reported herein is that our
conjugates have the potential for acting as indiVidual molecules;
that is, aggregation is not required. We also note that the use
of small, rigid, and planar moieties has been used in the design
of certain other artificial ionophores, where the transport-active
species appears to exist in an aggregated form.11-15
that is, 4 and 5 (Chart 5). Thus, whereas 3 consists of four
lysine-dicholamide moieties that are covalently attached to a
spermine backbone, conjugate 4 has the same four lysine-
dicholamides attached to a putrescine backbone, which has been
extended by two lysine groups. In the case of conjugate 5, the
eight choloyl units are now spaced further apart by use of a
spermine core that has been extended on both ends by two lysine
units. Finally, as a further test of our new design principle,
conjugate 6 was chosen as a synthetic target since it can be
directly compared with 4 (Chart 6). In this case, the flexible
putrescine core of 4 has been replaced by a 1,4-phenylenedia-
mide moiety.
Conjugate Synthesis. The synthetic approach that was used
for the preparation of 2 is outlined in Scheme 1. Thus, acylation
of Boc-L-Lys-OH with NRNꢀ-di-Boc-L-lysine hydroxylsuccin-
imide ester (Boc-L-Lys(Boc)-OSu) afforded 7. Subsequent
activation of the carboxylic group and condensation with 1,4-
phenylenediamine produced 8, which was then deprotected and
acylated with L-lysine-dicholamide to give 2. The synthesis of
4 was carried out by a similar sequence of reactions. Specifi-
cally, acylation of putrescine with Boc-L-Lys(Boc)-OSu, fol-
lowed by deprotection, and acylation with L-lysine-dicholamide
produced the desired conjugate (Scheme 2). Conjugate 5 was
synthesized by acylating L-lysine-dicholamide with Boc-L-Lys-
OH, followed by activation, condensation with spermine,
deprotection, and, finally, condensation with an activated form
of cholic acid (Scheme 3).
Synthetic Targets. To test our design principle we first chose
conjugate 2 as a synthetic target (Chart 4). Here, a 1,4-
phenylenediamide moiety was selected as a small, rigid, and
planar linker. To explore the sensitivity of ion transport activity
toward the compactness of a poly(choloyl)-based amphiphile,
we compared the activity of 3 with two less compact analogues;
(10) De Riccardis, F.; Di Filippo, M.; Garrisi, D.; Izzo, I.; Mancin, F.; Pasquato,
L.; Scrimin, P.; Tecilla, P. Chem. Commun. 2002, 3065-3067.
(11) Yoshii, M.; Yamamura; M.; Satake, A.; Kobuke, Y. Org. Biomol. Chem.
2004, 2, 2619-2623.
(12) Goto, C.; Yamamura, M.; Satake, A.; Kobuke, Y. J. Am. Chem. Soc. 2001,
123, 12152-12159.
(13) Eggers, P. K.; Fyles, T. M.; Mitchell, K. D. D.; Sutherland, T. J. Org.
Chem. 2003, 68, 1050-1058.
(14) Kobuke, Y.; Nagatani, T. J. Org. Chem. 2001, 66, 5094-5101.
(15) Cameron, L. M.; Fyles, T. M.; Hu, C.-W. J. Org. Chem. 2002, 67, 1548-
1553.
9
12730 J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005

Poly(choloyl)-Based Amphiphiles
A R T I C L E S
Chart 4
Chart 5
Finally, the synthesis of 6 was accomplished by acylating
1,4-phenylenediamine with Boc-L-Lys(Boc)-OSu, followed by
deprotection and acylation with L-lysine-dicholamide (Scheme
4).
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12731

A R T I C L E S
Chart 6
Chen et al.
are thermodynamically favored and their concentration can be
approximated by the total concentration of 1 that is present,
and (iii) dimers are responsible for ion transport.16 Thus, for
the general case of transport-active aggregates, the rate of ion
flow is expected to obey eq 1, where kobsd is a pseudo-first-
order rate constant that is the product of the aggregate
concentration and a rate constant, k2 (eq 2). If the aggregate
concentration is expressed in terms of the monomer concentra-
tion and the dissociation constant (K) that defines the aggregate-
monomer equilibrium (eq 3), then kobsd will vary with the
monomer concentration to the nth power (eq 4). Thus, the
second-order dependence of kobsd on the mol % of 1 in bilayers
of POPC supports the existence of transport-active dimers.
rate ) kobsd[Na+]
(1)
(2)
kobsd ) k2[aggregate]
K) [monomer]n/[aggregate]
kobsd ) k2[monomer]n/K
(3)
(4)
Kinetics of Na+ Transport. The pore-forming properties of
each of these conjugates were examined by measuring their
ability to promote Na+ transport across POPC-based liposomes
(ca. 200 nm diameter, unilamellar), as determined by 23Na+
NMR spectroscopy, using methods similar to those previously
described.8 In all cases, the rate of entry of Na+ into the aqueous
compartment of the liposomes obeyed pseudo-first-order kinet-
In contrast to 1, kobsd for 2 was found to be directly
proportional to the concentration of the conjugate (Figure 1B).
This finding supports the existence of a monomer-active species
and is fully consistent with our new design principle (Chart 3).
Similar plots that were made for 3-5 showed the expected
second-order dependence on the mol % of each conjugate, which
also supports the existence of dimer-active species and our
earlier design principle (Figure 1C-E). Finally, similar to 2, a
first-order dependence of kobsd on the mol % of the conjugate
was found for 6 (Figure 1F). Thus, of the six conjugates that
were investigated, those four that had a flexible core (1, 3, 4,
ics. For 1 the observed pseudo-first-order rate constant (kobsd
)
was found to have a strong dependence on its concentration in
the membrane (Figure 1A). Specifically, kobsd showed a second-
order dependence on the mol % of 1. As discussed elsewhere,
such a dependence lends strong support for a model in which
(i) monomers of 1 are in equilibrium with dimers, (ii) monomers
Scheme 1
9
12732 J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005

Poly(choloyl)-Based Amphiphiles
A R T I C L E S
Scheme 2
Scheme 3
Scheme 4
and 5) showed kinetic features indicative of dimer-active species;
the two that possessed a rigid, planar core (2 and 6) exhibited
kinetic features indicative of a monomer-active species.
On the basis of a nonlinear least-squares fit of the data in
Figure 1, values of k2/K for 3, 4, and 5 were found to be 260,
21, and 85 (min-1mol %-2), respectively. Although we are
presently unable to separate these kinetic and thermodynamic
terms, we believe that the much higher value of k2/K for 3
relative to 4 is a likely consequence of a pore structure that is
less deformable due to a higher degree of compactness. In other
words, we believe that this difference is largely a reflection of
the difference in the intrinsic activities of the pores (i.e., k2 is
dominant). The reason k2/K for 5 is greater than that of 4 is not
presently clear. One possibility is that the greater conformational
(16) Otto, S.; Osifchin, M.; Regen, S. L. J. Am. Chem. Soc. 1999, 121, 7276-
7277.
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12733

A R T I C L E S
Chen et al.
Figure 1. Plots of kobsd versus (mol % conjugate) for (A) 1, (B) 2, (C) 3, (D) 4, (E) 5, and (F) 6 in POPC liposomes at 35 °C. The solid lines are nonlinear
least-squares fit of the data according to kobsd ) k0 + k2(mol % conjugate)n/K, where n ) 2.0, 1.1, 2.1, 2.2, 2.0, and 0.9, respectively; insets represent similar
plots where n has been fixed at 2.0, 1.0, 2.0, 2.0, 2.0, and 1.0 for 1, 2, 3, 4, 5, and 6, respectively.
flexibility of 5 allows for more effective matching of the two
halves of the barrel stave; that is, K may be more important
than k2. Despite these uncertainties, the fact that the overall
activity of a poly(choloyl)-based amphiphile can be modified
to a significant extent by adjusting its compactness adds a new
dimension to the design of pore-forming agents in a general
sense. It is also noteworthy that since each molecule of 2 (a
monomer-active amphiphile) contributes, in effect, six choloyl
units per monolayer, one may compare its activity with a dimer-
active amphiphile (16) containing six choloyl groups linked to
a spermidine backbone that we have reported recently.8 Thus,
whereas 0.11 mol % of 16 results in a half-life for the Na+
transport equaling 70 min, a similar level of activity requires
only 0.0004 mol % of 2. From an operational standpoint,
therefore, 2 is ca. 300 times more active than 16.8 The greater
the activity per mole of 1 (a dimer-active amphiphile) is not
very different from that of 2 (a monomer-active amphiphile)
transport activity found with 2 (k1 ) 31.0 min-1mol %-1
)
relative to 6 (k1 ) 1.7 min-1mol %-1), where kobsd ) k1[2 or
6], is a likely consequence of a larger pore size resulting from
the larger number of sterols per conjugate. Finally, the fact that
9
12734 J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005

Poly(choloyl)-Based Amphiphiles
A R T I C L E S
implies that the unimolecular pore is less active for cation
translocation. The lower activity of the latter is a likely result
of a smaller pore size, due to the presence of fewer sterols in
each half of the bilayer.
Thermally Gated Release of Carboxyfluorescein. Although
a precise determination of the pore sizes that are created by
these conjugates lies beyond the scope of the present study,
preliminary experiments that we have carried out with the most
compact member of this series (i.e., 3) indicate an effective
diameter that is greater than that of carboxyfluorescein (CF),
estimated to be ca. 0.96 nm.17 Thus, using experimental
procedures similar to those previously described, CF was
encapsulated in liposomes made from 1,2-dipalmitoyl-sn-
glycero-3-phosphocholine (DPPC) containing 3.6 By raising the
temperature of the dispersion from 37 to 43 °C, a rapid efflux
of CF was observed (Figure 2).18
Figure 2. Plot of percentage of CF release from liposomes made from
DPPC containing 0.01 mol % of 3 as a function of time at 37 and 43 °C.
Complete release of CF was measured after destroying the liposomes with
Triton X-100.
Conclusions
A series of poly(choloyl)-based pore-forming amphiphiles
have been synthesized in which core flexibility and overall
compactness have been modified by synthetic design. Kinetic
results obtained from Na+ transport measurements across POPC
bilayers provide strong support for a design principle in which
a poly(choloyl)-based conjugate bearing a rigid core can act as
a transporter as an individual molecule. Related studies with
analogous amphiphiles having a flexible core indicate that
overall compactness is an additional variable that can be used
to adjust transport activity. Efforts aimed at gaining further
control over pore formation by poly(choloyl)-based conjugates
are currently in progress with a view toward the design of new
antibiotics.
Acknowledgment. We are grateful to the National Science
Foundation (CHE-0345248) for support of this research.
Supporting Information Available: Typical 23Na+ NMR
spectrum used for quantifying internal and external (shifted)
Na+ (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
(17) Rudnick, D. E.; Noonan, J. S.; Geroski, D. H.; Prausnitz, M. R.; Edelbauser,
H. F. InVest. Ophth. Vis. Sci. 1999, 40, 3054-3058.
(18) Dynamic light scanning measurements indicated the DPPC liposomes
containing 0.01 mol % of 3 had, essentially, the same size and size
distribution before and after CF release at 43 °C. The average diameters
for the dispersions were ca. 172 and 179 nm, respectively.
JA053527Q
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12735