Studies of mixed liposomes with novel sorbyl functionalized head group lipids

Article abstract of DOI:10.1016/j.chemphyslip.2012.10.004

Three novel polymerizable amphiphiles with a sorbyl-substituted head group were synthesized and systematically characterized. These amphiphiles are neutral in charge. None of these molecules forms vesicles by itself, presumably due to lack of amphiphilicity and/or extensive head group interaction. Therefore, mixed vesicles were formed with other fluid lipids such as DPenPC, eggPC, or DOPC. We investigated the properties of these mixtures in both vesicles and Langmuir films. The Langmuir isotherms show formation of monolayers by all three molecules. However, the isotherms for mixed monolayers suggest that two components are largely immiscible to the mixing lipid. Under polymerization conditions, mixed vesicles of these amphiphiles form oligomers, suggesting that in spite of a larger head group, they form mesophases.

Full text of DOI:10.1016/j.chemphyslip.2012.10.004

Chemistry and Physics of Lipids 165 (2012) 809–817
Contents lists available at SciVerse ScienceDirect
Chemistry and Physics of Lipids
journal homepage: www.elsevier.com/locate/chemphyslip
Studies of mixed liposomes with novel sorbyl functionalized head group lipids
Saliya N. Ratnayaka∗, El Hadj Elandaloussi1, Bryan D. Korth, Robert B. Bates, Jeffrey Pyun,
Henry K. Hall Jr., S. Scott Saavedra
Department of Chemistry and Biochemistry, University of Arizona, Tucson, AZ 85721-0041, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 July 2012
Received in revised form
30 September 2012
Accepted 22 October 2012
Available online 2 November 2012
Three novel polymerizable amphiphiles with a sorbyl-substituted head group were synthesized and
systematically characterized. These amphiphiles are neutral in charge. None of these molecules forms
vesicles by itself, presumably due to lack of amphiphilicity and/or extensive head group interaction.
Therefore, mixed vesicles were formed with other fluid lipids such as DPenPC, eggPC, or DOPC. We inves-
tigated the properties of these mixtures in both vesicles and Langmuir films. The Langmuir isotherms
show formation of monolayers by all three molecules. However, the isotherms for mixed monolayers
suggest that two components are largely immiscible to the mixing lipid. Under polymerization condi-
tions, mixed vesicles of these amphiphiles form oligomers, suggesting that in spite of a larger head group,
they form mesophases.
Keywords:
Sorbyl groups
Polymerizable lipids
Langmuir isotherms
Mixed liposomes
© 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
are sufficiently spaced from the hydrophobic core (Ringsdorf et al.,
1988). In addition, incorporating flexible spacers between reactive
The polymerization of hydrated lipid amphiphiles is an effec-
tive way to modify the chemical and physical properties of a
supramolecular assembly. This can be achieved by introducing
reactive groups to either the tail or head group region of the
amphiphile. A diverse assortment of reactive groups, including
diacetylenyl, methacryloyl, acryloyl, dienoyl, sorbyl, and styryl
have been commonly introduced into the tail region and are well
characterized (Bader et al., 1985; Blume, 1991; Mueller and O’Brien,
2002). However, introducing reactive groups in the tail region may
change properties inherent to fluid lipid assemblies upon polymer-
ization. For example, a reactive group in the tail region is likely
to perturb the lipid packing in mesophases (Blume, 1991) and
may lead to changes in physical properties such as permeability
(Roach et al., 2004). In addition, polymerization along the tail region
has a notable effect on the lipid membrane fluidity (Bader et al.,
1985; Gros et al., 1981; Ross et al., 2003b). Hence, understand-
ing of structure–property parameters that affect lipid assembly and
properties is a critical feature necessary for the design of novel and
improved functional lipids to form micellar, or vesicular aggregates.
The incorporation of polymerizable groups in the head region
may minimally perturb the lipid packing as the reactive groups
groups and the head group region imparts conformational entropy
between the alkyl chains and the main polymer chain, which is
required to enable both efficient assembly and connectivity. These
types of reactive amphiphiles can be useful in applications that
require retention of some of the physical properties of natural fluid
lipids while enhancing the mechanical stability of the bilayer struc-
ture.
Many types of reactive groups have been incorporated into
the head group region of lipids as well, including methacryloyl,
acryloyl, styryl, nitrostyryl, vinyl groups which can be polyme-
rized via free radical processes (Bader et al., 1985; Mueller and
O’Brien, 2002; Ringsdorf et al., 1988). Most of these molecules are
cationic or anionic while few of them are neutral (phospholipids)
or uncharged. In addition to the compositional characterization of
polymerized lipid assemblies, examination of Langmuir monolayer
properties, studies of vesicle fusion mechanism, and mechanical
and structural properties of these vesicles have been reported.
These structural properties include phase behavior, electrical
resistance, packing efficiency and self-organization, permeabil-
ity behavior, and spacers to isolate molecular motions of the
hydrophobic tails (see reviews above).
The synthetic lipids used in this study (Fig. 1) differ from the
previously reported head group polymerizable lipids in that they
contain sorbyl group(s) attached to the hydrophilic head group.
Sorbyl groups inserted into the hydrophobic chains of polymeriz-
able lipids have been well characterized (see review (Mueller and
O’Brien, 2002)). The extent of the polymerization and the nature
∗
Corresponding author. Tel.: +1 520 621 9761; fax: +1 520 621 8407.
E-mail address: saliya@arizona.edu (S.N. Ratnayaka).
Current address: Université Abdelhamid Ibn Badis, Laboratoire de Valorisation
1
des Matériaux, BP 227, 27000 Mostaganem, Algerie.
0009-3084/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chemphyslip.2012.10.004

810
S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
O
O
O
O
n
O
N
H
O
O
O
O
1 n=1
2 n=3
O
HN
O
O
O
N
O
O
HN
H
O
O
O
3
Fig. 1. Structures of lipids 1–3. Lipids 1 and 2 are trans structures while 3 is a cis lipid.
of the polymers are well known. The sorbyl group can be polyme-
rized in both gel (L␤) and liquid-crystalline (L␣) phases (Lei et al.,
1999) by several different initiation methods as well as on dif-
ferent scaffolds (Conboy et al., 2003; Mueller and O’Brien, 2002;
Ratnayaka et al., 2008; Ross et al., 2003a). These reasons led us
to synthesize a series of novel lipid molecules containing sorbyl
moieties attached to the head group. These lipids are neutral in
charge. The synthesis and characterization of three of these lipids:
2-[3-(2,3-dielaidoyloxypropylcarbamoyl)propanoyloxy]ethyl 2,4-
hexadienoate (1), 2-(2-{2-[3-(2,3-dielaidoyloxypropylcarbamoyl)
propanoyloxy]ethoxy}ethoxy)ethyl 2,4-hexadienoate (2), and N-
[2,3-(dioleoyloxy)-1-propyl]succinamide 1,3-(disorboylamino)-2-
Chemistry, University of Arizona. UV–vis absorbance spectra were
recorded on a Spectral Instruments (Tucson, AZ) model 440 UV–vis
spectrometer. Size exclusion chromatography (SEC) was performed
in a THF mobile phase with a Waters 1515 isocratic pump running
three 5 ␮m PLgel columns (Polymer Labs, pore sizes 104, 103, and
2
˚
10 A) with a Waters 2414 differential refractometer and Waters
2487 dual wavelength UV–vis spectrometer. Molar masses were
calculated using Empower software (Waters) calibrating against
low polydispersity linear polystyrene standards. Dynamic light
scattering (DLS) was performed with a BI-8000-autocorrelator with
a 5 mW He–Ne polarized laser source from Brookhaven Instru-
ments Corp., and vesicle size distributions were calculated with
the accompanying software.
propyl ester (3) are reported herein (Fig. 1).
1 and 2 are
mono-substituted with sorbyl groups while 3 is bis-substituted and
has amide-linked sorbate groups.
2.3.1. Phase transition measurements
Differential scanning calorimetry (DSC) data were obtained
using a 2920 Modulated DSC (TA Instruments) running Thermal
Solutions 1.4E (TA Instruments) software. DSC measurements were
run from 10 to 80 ◦C at a ramp rate of 5 ◦C/min. Thermotropic
phase transitions were measured using a method adapted from
Torchilin and Weissig (2003). Lipids were dried from the stock ben-
zene solution into an aluminum DSC pan (10 mg). DI water (25 ␮L)
was placed into the pan dropwise using a microsyringe. The pan
was sealed hermetically with an aluminum lid. Thermal transitions
of the sample were measured against a reference of 25 ␮L water
sealed in an aluminum pan prepared similarly to the sample.
2. Experimental
2.1. Materials
l-␣-Phosphatidylcholine (eggPC), 1,2-dioleoyl-sn-glycero-3-
phosphocholine (DOPC), 1,2-dielaidoyl-sn-glycero-3-phosphocho-
line (DEPC), and 1,2-dipentadecanoyl-sn-glycero-3-phospho-
choline (DPenPC) were purchased from Avanti Polar Lipids
(www.avantilipids.com) and used without further purification. All
other chemicals and solvents were analytical grade and obtained
from Sigma–Aldrich. The water used in polymerization and making
lipid vesicles was obtained from a Barnstead Nanopure system
with a measured resistivity of 18.3 Mꢀ cm and a total organic
content specified as less than 10 ppb.
2.3.2. Langmuir isotherms
Langmuir isotherms were measured in a Nima model 611D LB
trough. The spreading solvent was either the original stock solu-
tion or, in the case of mixed monolayers, CHCl3. The subphase was
water. All measurements were carried out at 20 ◦C.
2.2. Synthesis of 1–3
Syntheses of 1–3 are shown in Schemes 1 and 2. The purity of the
lipids was evaluated by thin-layer chromatography (TLC) and visu-
alized with a UV lamp. Compounds containing UV-sensitive groups
were handled under yellow lights. Detailed synthesis procedures
and spectral information are provided in the Supplementary data.
2.4. Methods
2.4.1. Preparation of vesicles
Lipids from stock solution were dried by passing a stream of
argon over the sample followed by applying high vacuum for
ca. 4 h. The film was hydrated with deoxygenated water. Multi-
lamellar vesicles were prepared by vortexing several times and
ultra-sonication of the mixture for 1 min above the phase transi-
tion temperature (Tm) of lipids, followed by 10 freeze–thaw cycles
(−77 ◦C in dry ice/2-propanol and then warmed to 40 ◦C).
Large unilamellar vesicles of lipids (LUV) were prepared using
an extruder (Northern Lipids Inc., Vancouver, BC). The suspension
2.3. Apparatus
1H NMR spectra were measured on a Bruker spectrometer at
250 MHz in CDCl3. Mass spectroscopy data were obtained using HP
5988A GC-MS, Finnigan LCQ HPLC-MS, and JEOL HX110A FAB-MS
instruments at the Mass Spectroscopy Facility in the Department of

S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
811
O
O
O
O
a
b
O
OH
O
OH
OH
Cl
O
n
n
S1.3aO,1.3b
S1.2a,1.2b
S1.1
O
O
O
O
c
d
O
n
O
O
Cl
O
N
H
n
O
O
OH
S1.5a,1.5b
S1.4a,1.4b
O
O
O
e
O
n
O
N
H
O
O
O
O
1 n=1
2 n=3
Scheme 1. Syntheses of lipids 1 and 2. Reagents and conditions: (a) ethylene glycol or triethylene glycol, K2CO3, THF, RT, 20 h, 73% and 82%; (b) succinic anhydride, N,N-
diisopropylethylamine, THF, RT, 18 h, 88% and 93%; (c) (COCl)2, CH2Cl2, RT, 16 h, 97% and 96%; (d) 3-amino-1,2-propanediol, Et3N, THF/H2O, 13 ◦C, 20 h, 30% and 32%; (e)
elaidic acid, DCC, DMAP, CH2Cl2, RT, 20 h, 51% and 88%.
was extruded through two stacked Nuclepore polycarbonate fil-
ters (pore size 600(×3), 400(×3), and 100(×4) nm, sequentially) at
45 ◦C. The average diameter of the vesicles was determined by DLS.
spectroscopy of aliquots withdrawn periodically during the course
of the reaction and diluted with MeOH.
Redox-initiated polymerization was performed with degassed,
Ar purged solutions of 39 mM K2S2O8 and 39 mM NaHSO3. The
two solutions were mixed with the vesicle suspension to have a
final molar concentration ratio of 1:1 between the polymerizable
lipid and the oxidant/reductant (Sisson et al., 1996). Polymeriza-
tions were performed at ca. 45 ◦C under a positive argon pressure
for 20 h. All polymerizations were carried out in the dark or under
yellow light. After the polymerization, the vesicle suspension was
dialyzed (MWCO 6000–8000) to remove the salts using water as
the medium. Polymerizations were monitored by UV absorption
spectroscopy as described above.
2.4.2. Polymerization
UV polymerization was performed by exposure to light from a
high-pressure Hg/Xe arc lamp (Newport Corporation, CT) for ca. 2 h
at 45 ◦C. The lamp was equipped with a 260–400 nm FWHM band-
pass filter (U-330, Edmund Optics) to remove shorter wavelengths
of UV that was found to cause degradation of the lipid polymer
(Meier et al., 1994). In addition, the lamp was equipped with a
water filter to remove IR radiation to prevent excessive heating
of the sample. Polymerization was monitored by UV absorbance
O
HN
HN
O
O
O
O
a
b
O
OH
Cl
N
H
N
H
O
OH
O
S2.2
S2.1
S2.3
O
BocHN
O
O
d-e
c
BocHN
OH
O
OH
S2.4
O
O
HN
HN
O
O
O
N
H
O
O
O
O
3
Scheme 2. Synthesis of lipid 3. Reagents and conditions: (a) 1,3-diamino-2-propanol, CH2Cl2, RT, 91%; (b) succinic anhydride, N,N-diisopropylethylamine, THF, refluxed 5
days, 55%; (c) oleic acid, DCC, DMAP, CH2Cl2, RT, 48 h, 60%; (d) TFA, RT, 3 h; (e) S2.3, triethylamine, DEPC, CH2Cl2, RT, 7 h, 43%.

812
S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
Table 1
Table 2
Molecular formula, molecular weight, absorbance maximum (nm), and molar
Tm of the lipids measured from the DSC transitions.
absorptivity (cm−1 M−1) in methanol of 1–3.
Lipid
Tm (◦C)
Molecule
Molecular
formula
MW
Abs ꢁmax
(nm)
ε at 258 nm
DPenPC
32
11
13
7
(cm−1 M−1
)
1
2
3
1
2
3
C51H87O9N
858.26
946.34
980.41
258
258
258
23,600
23,600
50,000
C
55H95O11
N
C58H97O9N3
The Tm values of 1, 2, 3, monomers and DPenPC are reported in
Table 2.2 The measured Tm is 32 ◦C for DPenPC, which is consistent
with reported values (31–34 ◦C) (Koynova and Caffrey, 1998;
2.4.3. Analyses of lipid polymers
Following the polymerization of lipid vesicles, water was
removed by lyophilization to leave a dry, fine powder that was
resuspended in THF:MeOH:benzene (10:1:1, v/v) for analyses by
SEC. Peak integration, number- and weight-average molecular
www.avantilipids.com;
www.lipidat.chemistry.ohio-state.edu)
and shows that the method is valid despite the low water content.
Although DPenPC shows a single sharp transition, all other lipids
tested show broad transitions. This behavior is not unexpected. It
has been reported that the thermotropic transition is broadened
when there are functional groups in the tails, as well as with
bulkier head groups, by reducing the cooperative transitions
(Blume, 1991; Kusumi et al., 1983). It is thought that the tail chains
become tilted more to minimize the influence from bulky head
groups, reducing the interactions between them and producing
broader peaks.
¯
¯
¯
¯
weights (Mn and Mw), and polydispersity index (PDI = Mw/Mn)
were analyzed for each chromatogram using the supplied software.
¯
The number-average relative degree of polymerization (Xn) was
¯
calculated by dividing Mn by the molecular weight of the monomer
lipid.
3. Results and discussion
1, 2, and 3 have Tm values of 11, 13, and 7 ◦C respectively with
standard deviations of less than 1. Replicate temperature scans did
not influence the DSC curves. The Tm values of trans lipids 1 and
2 are very close to that of the corresponding PC trans lipid, DEPC
(12 ◦C) (www.avantilipids.com). They have the same tail structure.
This suggests that head groups of these polymerizable lipids exert
less influence on Tm than the trans structure of the tails. Usually
the Tm of a trans lipid is higher than that of the corresponding
cis lipid due to the larger perturbations to lipid packing from cis
double bonds than trans double bonds (Roach et al., 2004). This is
consistent with the results for the lipids measured in this study.
Increasing the length of the head group spacer in 2 vs. 1 slightly
increased the Tm. Similar results were also reported by Blume and
Walle for polymerizable lipids having methacryloyl head groups
(Blume, 1991). They found that when the methacryloyl group
is attached to the PC head group via a short spacer, the Tm is
slightly shifted to lower values relative to the unfunctionalized
A key consideration in the design of polymerizable lipids is the
choice and the location of the reactive groups. Lipids 1–3 have sor-
byl groups in the head region of the molecule. When located in the
tail region of lipids, these groups are known to form polymers by
radical- and photo-initiated polymerization (Mueller and O’Brien,
2002). 1–3 are neutral and synthetically simpler to prepare than
corresponding zwitterionic lipids. 1 has a single sorbyl group in the
head region separated by a short spacer group while 2 has a spacer
group that is three times longer that of 1 (Fig. 1). This increased
spacer length is expected to further decouple the motion of the
random polymer coil and the mesogenic tail groups of 2. In addi-
tion, it is expected to increase the hydrophilicity of the head group
regionof theneutrallipidmoleculewithaddedethyleneoxygroups,
which may increase the likelihood of bilayer formation (Evans,
1988; Evans and Ninham, 1986; Neumann and Ringsdorf, 1986).
While 1 and 2 are expected to generate linear polymers, 3 has bis-
substituted, amide-linked sorbate groups at the head region that
may generate cross-linked polymers (Fig. 1). All three molecules
contain single unsaturations in the tails. 1 and 2 are trans lipids
while 3 is a cis lipid.
lipid. Increasing the length of the spacer slightly increased the Tm
.
The stabilization of the gel phase by the increased interactions
between the chains of the bulky molecule is assumed as a result
of the increased length of the spacer.
3.3. Langmuir isotherms
3.1. Absorbance properties of the lipids
Measurements of surface pressure/area (ꢂ–A) isotherms
provide a method for studying the miscibility of two components
in mixed lipid monolayers (Bader et al., 1985) as well as studying
the influence of parameters such as head group charge and size,
chain unsaturation, and chain packing on membranes. Therefore,
ꢂ–A isotherms of 1–3, DPenPC, and mixed monolayers of 1–3 with
DPenPC were studied (Fig. 3).
The shapes of the ꢂ–A isotherms of pure 1–3 are comparable
to each other. They show a single transition point corresponding
to a collapse from the liquid-expanded (LE) state (Fig. 3a–c). The
pressure at which the collapse occurs is comparable among the
three lipids and close to 35 mN/m (Table 3). In contrast, DPenPC
The absorbance properties of each molecule are listed in Table 1.
All the lipids show an absorbance maximum at 258 nm, char-
acteristic of a sorbyl group. 1 and 2 show a molar absorptivity
(ε) of 23,600 cm−1 M−1 at 258 nm in methanol, which is consis-
tent with the values reported for mono-SorbPC lipids (Lee and
O’Brien, 1992; Tyminski et al., 1987). The bis-substituted 3 has a
ε of 50,000 cm−1 M−1 in methanol.
3.2. Thermotropic transitions by DSC
A modified procedure (Torchilin and Weissig, 2003) was used to
measure the thermotropic phase behavior, since pure 1–3 do not
form vesicles (see Section 3.4). Even though the aqueous content is
relatively small in this method, lipids with sufficiently long chains
can still form lamellar phases with a small aqueous content (Blume,
1991). The method was validated using DPenPC with known DSC
transitions. The thermotropic transitions were observed as maxima
in the excess heat capacity vs. temperature endotherms (Fig. 2).
2
A characteristic feature of lipids with polymerizable groups attached to the head
group via spacers is that the phase transition is not abolished after polymerization,
as the motions of the hydrophobic chains are decoupled from the polymer backbone
(Blume, 1991). This was not tested with lipids mentioned here due to the difficulty
of separating poly(lipids) from mixed lipid vesicles in which polymerization was
performed.

S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
813
-24.0
-15.0
-17.5
-20.0
-22.5
-25.0
b
a
-24.5
-25.0
-25.5
-26.0
25
30
35
40
0
10
20
30
Temperature (ºC)
Temperature (ºC)
0.045
0.030
0.015
0.000
0.045
0.030
0.015
0.000
c
d
0
10
20
30
40
0
5
10
15
20
Temperature (ºC)
Temperature (ºC)
Fig. 2. DSC heating thermograms: (a) DPenPC, (b) 1, (c) 2, and (d) 3.
75
60
45
30
15
0
75
(i)
(ii)
(iii)
lipid 2
lipid 2 + DPenPC mixture
(i)
(ii)
(iii)
lipid
lipid
1
a
b
1 + DPenPC mixture
60
45
30
15
0
DPenPC
DPenPC
(i)
(i)
(ii)
(ii)
(iii)
(iii)
0
50
100
150
200
0
100
200
300
400
2
2
area/molecule (Angstrom )
area/molecule (Angstrom )
75
60
45
30
15
0
(i)
(ii)
(iii)
lipid 3
c
lipid 3 + DPenPC mixture
DPenPC
(i)
(ii)
(iii)
0
75
150
225
300
2
area/molecule (Angstrom )
Fig. 3. Pressure–area isotherms for pure lipids and 50% (mol) mixtures of synthesized lipids and DPenPC at 20 ◦C: (a) 1, (b) 2, and (c) 3, with DPenPC.

814
S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
Table 3
Selected properties of pressure–area isotherms shown in Fig. 3: surface pressure at the first transition and area per molecule at the first transition for mixed monolayers,
collapse pressure, and area per molecule at the collapse pressure.
Lipid
Pressure at first
Area per molecule (Å2)
at the first transitiona
Collapse pressure
(mN/m)
Area per molecule (Å2)
at the collapse pressure
transition (mN/m)a
DPenPC
1
2
3
50% (mol) 1/DPenPC
50% (mol) 2/DPenPC
50% (mol) 3/DPenPC
17
–
–
62
–
–
55
35
34
35
68
55
55
41
82
142
130
33
31
–
–
42
34
37
66
97
93
32
a
For lipids 1–3 the first transition is the collapse pressure.
exhibits two transitions: a phase transition between LE and liquid-
condensed (LC) states at 17 mN/m and a film collapse at 55 mN/m.
As suggested by previous authors, it is possible that the lower col-
lapse pressure for 1–3, relative to DPenPC, is due to the presence
of an uncharged head group (Elbert et al., 1984) and packing per-
turbations that arise from unsaturation as well as tail disturbances
from the bulky head group (Rice et al., 1987).
In contrast to DPenPC, the 1–3 isotherms show more expanded
films at lower surface pressures. Presumably these could be due to
less hydrophilicity and bulkier groups as suggested by Cadenhead
and coworkers based on studies of esters and corresponding acidic
These can be ascribed to slight miscibility of 1 and 3 in the DPenPC
monolayers.
The acyl chain length and the Tm differences are within the sug-
gested limits for complete miscibility (Lee, 1978, 1983) between
2 and DPenPC. However, the two components are not miscible as
observed by the Langmuir isotherm as well as the absence of vesicle
formation at higher temperatures, even above Tm where both lipids
are in the L␣ phase. We speculate that the large head and spacer
groups in 2 fold back toward the tail region, creating unfavorable
interactions for mixing.
hydrocarbons (Matuo et al., 1982). The area per molecule at the
2
˚
collapse pressure for 1–3 is 82, 142, and 130 A , respectively. The
3.4. Preparation of vesicles
larger increments for 2 and 3 compared to 1 are anticipated because
these are expected to occupy a larger area and pack less efficiently
than 1. It is possible that 2, with an extended spacer occupies more
space and/or creates additional tail perturbations relative to 3.
Pressure–area isotherms of mixed monolayers prepared from
50% (mol) 1–3 and DPenPC were studied in a manner similar
to prior reports of immiscible and miscible mixed monolayers
(Bader et al., 1985; Buschl et al., 1982; Chen et al., 2000; Elbert
et al., 1984; Fukuda et al., 1979; Lee et al., 2006; Petty, 1996;
Wu and Huntsber, 1969). These mixed isotherms fall between
those of the two pure monolayers (Fig. 3a–c) and are similar in
shape. All three isotherms exhibit three distinct transition points,
corresponding to the presence of two different phases. The first
and the last transition pressures are nearly identical to the col-
lapse pressures found for pure monolayers, i.e. head group lipids
and DPenPC, respectively (Fig. 3a–c). The first transition points
for the 1/DPenPC and 3/DPenPC mixed monolayers occur slightly
above the collapse pressures of pure 1 and 3, approximately at 42
and 37 mN/m, respectively. However, the first transition point for
2/DPenPC mixed monolayer is no different than the collapse pres-
sure of pure 2, which is 34 mN/m (Table 3). According to the phase
rule, the ꢂ–A isotherm behavior for the mixed monolayers indi-
cates that the two components are immiscible with component 1
or 2 or 3 being “squeezed out” at the first transition point (Crisp,
1949).
The immiscibility observed in the mixed monolayer isotherms
is likely due to monolayer formation from different film states
(Cadenhead, 1972). At 20 ◦C, amphiphiles 1–3 are in the L␣ phase
and DPenPC is in the L␤ phase (Table 2). It has been reported that in
a two-component mixture, complete miscibility occurs if the acyl
chain lengths of the two lipids differ by less than 4 methylene units
and the difference in their Tm values is less than 20◦ (Lee, 1978,
1983). The acyl chain length difference is limited to 3 methylene
units in 1–3 and DPenPC mixtures, but not the Tm values except for
2 (Table 2). However, the first transition pressures in the 1/DPenPC
and 3/DPenPC mixed isotherms are somewhat higher than the
collapse pressures of pure 1 and 3. In addition, a slight deviation
of parallelism in the 3/DPenPC mixed isotherm after the second
transition point compared to that of pure DPenPC is observed.
In all cases, suspended particles or emulsions were visu-
ally observed in aqueous suspensions of 1–3. Ultrasonication for
extended periods above the Tm did not render clear solutions.
Extruded aqueous solutions of these lipids produced population
distributions of structures that were too small to measure reli-
ably using DLS. Similar observations were reported for polymeric
amphiphilic monomers that showed difficulty in making vesi-
cles (Johnston et al., 1983; Kunitake et al., 1981; Regen et al.,
1980; Ringsdorf et al., 1988). The reason was presumed to be
insufficient amphiphilicity (Blume, 1991; Evans and Ninham,
1986; Israelachvili, 1991; Ringsdorf et al., 1988) and/or extensive
intermolecular head group interactions that decrease hydration
(Litzinger and Huang, 1992). Attempts were then made to form
mixed vesicles using fluid lipids having tail lengths close to those
in 1–3 (Lee, 1978, 1983).
The commercially available lipids DPenPC, eggPC, or DOPC were
combined with 1–3 in various molar ratios to form vesicles.3 The
hydrodynamic mean diameter of the mixed vesicles was mea-
sured using DLS. Attempts to form mixed vesicles with 2 in any
composition were not successful and suspended particles were
visually observed. Homogeneous populations were observed by
DLS up to 50% in mole composition for both 1 and 3 with the
mixing lipid with a mean value of ca. 120 nm. Above 50% for 1
and 65% for 3, two population distributions were observed. For
example, when the composition was 70% of 1 and 30% DPenPC,
two populations with mean diameters of 50 nm and 180 nm were
observed (Fig. 4). We speculate that after a threshold concentra-
tion is reached, the vesicles rearrange to form more stable mixed
structures with appropriate hydrophilic–lipophilic balance. These
results were comparable for 1 and 3 mixed with all three fluid
lipids.
3
It is possible that the isolated double bond in either DOPC or eggPC reacts to
form oligomers during the polymerization, and this could interfere with the iden-
tification of polymer products of 1–3. Therefore, mixed vesicle studies of 1–3 with
the saturated lipid, DPenPC, were performed to eliminate this complication.

S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
815
Before UV illumination
After 30 min UV illumination
After 2.5 hours UV illumination
100
80
60
40
20
0
70% lipid
50% lipid
1
1
a
0.4
0.2
0.0
225
250
275
300
325
350
0
100
200
300
400
500
Wavelength (nm)
Diameter (nm)
Fig. 4. Representative DLS data showing diameters of the vesicles prepared from
50:50 and 70:30 (mol/mol) 1:DPenPC mixtures. For 70% lipid 1 mixture, the mean
diameter of the smaller population was 50 nm and the mean of the larger population
was ca. 180 nm. The mean diameter is ca. 120 nm for mixtures below 50% of lipid 1.
Before UV illumination
After 15 min UV illumination
After 2.5 hours UV illumination
0.4
0.2
0.0
b
3.5. Polymerization of vesicles
Mixed vesicles composed of 1 or 3 and DPenPC were subjected to
redox-initiated radical polymerization or UV photopolymerization
before measuring the molecular weight of the products. In order to
more efficiently polymerize these vesicles, the temperature of the
solution was held at 45 ◦C, above the higher of the two Tm values.
This ensured that both lipids were in the L␣ phase.
The mixed vesicles of both 1 and 3 show an absorbance maxi-
mum at 258 nm, which decreases following UV irradiation or redox
polymerization. The similarity in ꢁmax in methanol and in vesi-
cles (aqueous) indicates that the two sorbyl chromophores do
not interact or aggregate at temperatures above the Tm (Lee and
O’Brien, 1992; Tyminski et al., 1987). Fig. 5 shows a representa-
tive set of absorbance spectra obtained by sampling 1 and 3 mixed
vesicle dispersions during UV irradiation. The disappearance of
the absorbance band at 258 nm shows that molecules 1 and 3
undergo UV induced reactions and conveniently tracks the reaction
progress.
The loss of monomer absorption was also accompanied by the
progressive loss of migration of the lipid in TLC. For example, 1 and
DPenPC showed 0.88 and 0.33 Rf values, respectively, with the sol-
vent mixture 65:25:4 (v/v) CHCl3:MeOH:H2O. The UV irradiated
1/DPenPC mixture showed three Rf values: two values correspond-
ing to the pure lipids and one intermediate value (0.51). If there
is an oligomer or a polymer formed during the UV illumination,
the higher MW fragment is expected to display a slower migration
than that of the monomer. Therefore, the 0.51 Rf value is presumed
to be due to the migration of polymer. Similar observations were
reported by Lee and O’Brien for the TLC of sorbyl ether glycerophos-
phocholine lipids (Lee and O’Brien, 1992).
225
250
275
300
325
350
Wavelength (nm)
Fig. 5. Absorbance measurements of 50% (mol) mixtures of 1 and 3 in methanol
before and after UV illumination: (a) 1/DPenPC and (b) 3/DPenPC.
of the 1.85 ppm signal was clear, but the disappearance of other
proton signals was not. Mueller and O’Brien suggested that an
amide-linked sorbate monomer might follow a different polymer-
ization mechanism (Mueller and O’Brien, 2002). Further studies are
being done to identify the polymerization mechanism of 3.
3.6. DLS of polymerized mixed vesicles
The sizes of the vesicles were not significantly changed as
observed by DLS after they were subjected to polymerization condi-
tions (data not shown). The absence of larger diameter aggregates
for polymerized vesicles suggests that inter-vesicle cross-linking
polymerization or inter-vesicle aggregation was not favored under
the conditions used here.
The 1H NMR spectra of both redox and UV polymerized sam-
ples of 1 showed that all of the proton resonances due to the sorbyl
moiety disappeared. Losses of conjugated terminal methyl protons
adjacent to the sorbyl unit (1.85 ppm, doublet) and the vinyl proton
adjacent to the carbonyl (5.72–5.80 ppm, doublet) are consistent
with the formation of a polymer product via 1,4-polymerization
(Clapp et al., 1997; Lamparski et al., 1992; Lamparski and O’Brien,
1995). It has been reported that ester-linked sorbyl group poly-
merization usually follows a 1,4 mechanism (Mueller and O’Brien,
2002). However, in the 1H NMR spectra of polymerized 3, the loss
3.7. Analyses of lipid polymers by SEC
After polymerization, lyophilized mixed vesicles were partially
soluble in organic solvents. To improve the solubility and enable
SEC analysis, cosolvents (MeOH and benzene) were added to the
lyophilized samples dissolved in THF (Lamparski and O’Brien,
1995). The SEC analysis of molecular weight data for polymerized

816
S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
Table 4
Further work is required to identify the polymerization mechanism
of 3.
¯
¯
SEC data: number- and weight-average molecular weights (Mn and Mw), polydis-
persity index (PDI), and number-average relative degree of polymerization (Xn) for
¯
This report describes the first known studies of sorbyl-
substituted head group lipids. These results provide new insights
into designing novel lipid molecules for supramolecular structures.
The future possibilities include mixed lipid vesicles composed of
head group lipids that can be usefully employed to phase-separate
unreactive lipids from the newly formed polymeric domains.
the soluble fractions of polymerized mixed vesicles of DPenPC and lipids 1 and 3.
¯
¯
¯
Xn
Lipid
Polymerization
condition
Mw range
Mn range
PDI
1
1
3
3
UV
Redox
UV
1600–9700
1600–27,800
6400
1500–8300 2–10 1.1–1.2
1500–9800 2–11 1.0–2.8
4900
5200
5
5
1.3
1.4
Redox
7100
Acknowledgements
This research was supported by in its early stages by a grant from
the Department of Energy (DE-FG03-02ER15378) and in its latter
stage by the National Institute of Health under Grant no. EB007047.
The contents of this paper are the sole responsibility of the authors
and do not necessarily represent the official views of the NIH or
DOE.
mixed vesicles is summarized in Table 4. These molecular weights
are apparent MWs measured relative to PS standards. These SEC
data clearly reveal that 1 and 3 form polymers (oligomers) dur-
ing UV and radical polymerizations. However, these results reflect
¯
only the soluble fraction rather than the complete Xn range for
¯
the polymer. O’Brien and coworkers reported that the Xn is 4–51
The authors would also like to thank Dr. Gemma D’Ambruoso for
helpful discussions, Dr. Neil E. Jacobsen for NMR data processing,
Nick Laude and Dr. Eric Monroe for critical reading of the
manuscript and Prof. Vicki H. Wysocki for extending her lab facili-
ties to us.
for the transesterified samples of redox poly(mono-SorbPC) (Lei
et al., 1999) and 3–10 for UV poly(mono-SorbPC) (Lamparski and
O’Brien, 1995). The Xn value for both UV and redox polymerization
¯
of the 1 is ca. 2–10. This suggests that the sorbyl group in the head
group region of 1 behaves similarly to the sorbyl group in the tail
region of mono-SorbPC during two-dimensional polymerizations.
¯
The reported Xn for redox poly(bis-SorbPC) is 100–600 and 5–10
Appendix A. Supplementary data
for UV poly(bis-SorbPC) (Lamparski and O’Brien, 1995), compared
¯
to 5 for both redox and UV polymerized 3. The lower Xn for 3 is
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.chemphyslip.2012.10.004.
likely due to using only the soluble fraction for SEC measurements.
It has been reported that sorbyl ester groups do not readily
polymerize unless the molecules aggregate to form mesophases
(Masuda et al., 2003; Matsumoto et al., 1991; Mueller and O’Brien,
2002). The identification of oligomers from 1 and 3 under typi-
cal polymerization conditions that have been used for other sorbyl
lipids (Mueller and O’Brien, 2002; Sisson et al., 1996) indicates that,
in spite of a relatively large head group, 1 and 3 form mesophases
in mixed vesicles.
References
Bader, H., Dorn, K., Hupfer, B., Ringsdorf, H., 1985. Polymeric monolayers and lipo-
somes as models for biomembranes – how to bridge the gap between polymer
science and membrane biology. Advances in Polymer Science 64, 1–62.
Blume, A., 1991. Phase-transitions of polymerizable phospholipids. Chemistry and
Physics of Lipids 57, 253–273.
Buschl, R., Hupfer, B., Ringsdorf, H., 1982. Polyreactions in oriented systems 30.
Mixed monolayers and liposomes from natural and polymerizable lipids. Macro-
molecular Rapid Communications 3, 589–596.
Cadenhead, D.A., 1972. Film balance studies of membrane lipids and related
molecules – applying surface chemistry to molecular biology. Journal of Chem-
ical Education 49, 152–156.
Chen, K.B., Chang, C.H., Yang, Y.M., Maa, J.R., 2000. On the interaction of dipalmitoyl
phosphatidylcholine with normal long-chain alcohols in a mixed monolayer: a
thermodynamic study. Colloids and Surfaces A 170, 199–208.
Clapp, P.J., Armitage, B.A., O’Brien, D.F., 1997. Two-dimensional polymerization
of lipid bilayers: visible-light-sensitized photoinitiation. Macromolecules 30,
32–41.
Conboy, J.C., Liu, S.C., O’Brien, D.F., Saavedra, S.S., 2003. Planar supported bilayer
polymers formed from bis-diene lipids by Langmuir–Blodgett deposition and
UV irradiation. Biomacromolecules 4, 841–849.
Crisp, D.J., 1949. A two dimensional phase rule I & II. Surface Chemistry. Butterworths
Scientific Publications, London, pp. 17–35.
Elbert, R., Folda, T., Ringsdorf, H., 1984. Saturated and polymerizable amphiphiles
with fluorocarbon chains – investigation in monolayers and liposomes. Journal
of the American Chemical Society 106, 7687–7692.
Evans, D.F., 1988. Self-organization of amphiphiles. Langmuir 4, 3–12.
Evans, D.F., Ninham, B.W., 1986. Molecular forces in the self-organization of
amphiphiles. Journal of Physical Chemistry 90, 226–234.
Fukuda, K., Kato, T., Machida, S., Shimizu, Y., 1979. Binary mixed monolayers of
polyvinyl stearate and simple long-chain compounds at the air–water-interface.
Journal of Colloid and Interface Science 68, 82–95.
Gros, L., Ringsdorf, H., Schupp, H., 1981. Polymeric anti-tumor agents on a molecular
and on a cellular-level. Angewandte Chemie-International Edition in English 20,
305–325.
Israelachvili, J.N., 1991. Intermolecular and Surface Forces, 2nd ed. Academic Press,
London.
Johnston, D.S., McLean, L.R., Whittam, M.A., Clark, A.D., Chapman, D., 1983. Spectra
and physical-properties of liposomes and monolayers of polymerizable phos-
pholipids containing diacetylene groups in one or both acyl chains. Biochemistry
22, 3194–3202.
Koynova, R., Caffrey, M., 1998. Phases and phase transitions of the phosphatidyl-
cholines. Biochimica et Biophysica Acta – Biomembranes 1376, 91–145.
Kunitake, T., Nakashima, N., Takarabe, K., Nagai, M., Tsuge, A., Yanagi, H., 1981. Vesi-
cles of polymeric bilayer and monolayer membranes. Journal of the American
Chemical Society 103, 5945–5947.
4. Summary and conclusions
A series of neutral lipids with polymerizable sorbyl head groups
were synthesized. Lipids 1 and 2 are mono-substituted and 3 is bis-
substituted. There is a short spacer in 1 and longer spacer group in
2 separating the sorbyl group from the head group.
Pure solutions of all three lipids do not form stable vesicles. This
is likely to be due to a lack of amphiphilic nature of the neutral
molecules, extensive intermolecular head group interactions that
decrease hydration, and/or bulky head groups that may reduce the
cohesive tail interactions necessary for bilayer formation. However,
1 and 3 (but not 2) form mixed vesicles with commercially available
fluid lipids such as DPenPC, eggPC and DOPC.
Langmuir isotherms of mixed monolayers of 1–3 with DPenPC
suggest that they are largely immiscible with 1, 2, or 3 at 20 ◦C. How-
ever, slight deviations in the isotherms for 1 and 3 with DPenPC in
mixed monolayers are ascribed to a slight miscibility of the com-
ponents. Thus mixed vesicles can be formed with 1 and 3 but not
with 2. However, the slight miscibility of components restricts the
molar fraction of 1 or 3 in the mixed vesicles to be ≤50%, even at
temperatures above Tm
.
The disappearance of the absorbance band at 258 nm, TLC migra-
tion changes, and 1H NMR data suggest that the monomers react to
form polymer products under polymerization conditions. The SEC
studies of 1 and 3 mixed vesicles show that both form oligomers
when subjected to either UV- or redox-initiated polymerization.
These results indicate that sorbyl groups on bulky head groups
aggregate sufficiently to be polymerized. Using 1H NMR, polymer-
ization of 1 was determined to be following the 1,4 mechanism.

S.N. Ratnayaka et al. / Chemistry and Physics of Lipids 165 (2012) 809–817
817
Kusumi, A., Singh, M., Tirrell, D.A., Oehme, G., Singh, A., Samuel, N.K.P., Hyde,
J.S., Regen, S.L., 1983. Dynamic and structural-properties of polymerized phos-
phatidylcholine vesicle membranes. Journal of the American Chemical Society
105, 2975–2980.
Neumann, R., Ringsdorf, H., 1986. Peptide liposomes from amphiphilic amino-acids.
Journal of the American Chemical Society 108, 487–490.
Petty, M.C., 1996. Langmuir–Blodgett Films: An Introduction. Cambridge University
Press, Cambridge, England.
Lamparski, H., Liman, U., Barry, J.A., Frankel, D.A., Ramaswami, V., Brown, M.F.,
O’Brien, D.F., 1992. Photoinduced destabilization of liposomes. Biochemistry 31,
685–694.
Lamparski, H., O’Brien, D.F., 1995. 2-Dimensional polymerization of lipid bilayers –
degree of polymerization of sorbyl lipids. Macromolecules 28, 1786–1794.
Lee, A.G., 1978. Calculation of phase-diagrams for non-ideal mixtures of lipids, and a
possible nonrandom distribution of lipids in lipid mixtures in liquid–crystalline
phase. Biochimica et Biophysica Acta 507, 433–444.
Lee, A.G., 1983. Lipid phase transitions and mixtures. In: Aloia, R.C., Boggs, J.M. (Eds.),
Membrane Fluidity in Biology. Academic Press, New York, pp. 43–88.
Lee, Y.L., Lin, J.Y., Chang, C.H., 2006. Thermodynamic characteristics and
Langmuir–Blodgett deposition behavior of mixed DPPA/DPPC monolayers at
air/liquid interfaces. Journal of Colloid and Interface Science 296, 647–654.
Lee, Y.S., O’Brien, D.F., 1992. Chain substituted polymerizable ether lipids – synthesis
of sorbyl and diacetylenic ether glycerophosphocholine. Chemistry and Physics
of Lipids 61, 209–218.
Ratnayaka, S.N., Wysocki, R.J., Saavedra, S.S., 2008. Preparation and charac-
terization of asymmetric planar supported bilayers composed of poly(bis-
sorbylphosphatidylcholine) on n-octadecyltrichlorosilane SAMs. Journal of
Colloid and Interface Science 327, 63–74.
Regen, S.L., Czech, B., Singh, A., 1980. Polymerized vesicles. Journal of the American
Chemical Society 102, 6638–6640.
Rice, D.K., Cadenhead, D.A., Lewis, R., McElhaney, R.N., 1987.
monomolecular film study of a straight-chain phosphatidylcholine (dipalmi-
toylphosphatidylcholine) with isobranched-chain phosphatidylcholines
(diisoheptadecanoylphosphatidylcholine, diisooctadecanoylphosphatidyl-
choline, and diisoeicosanoylphosphatidylcholine). Biochemistry 26,
3205–3210.
A comparative
3
Ringsdorf, H., Schlarb, B., Venzmer, J., 1988. Molecular architecture and function of
polymeric oriented systems – models for the study of organization surface recog-
nition, and dynamics of biomembranes. Angewandte Chemie-International
Edition in English 27, 113–158.
Lei, J.T., Sisson, T.M., Lamparski, H.G., O’Brien, D.F., 1999. Two-dimensional polymer-
ization of lipid bilayers: effect of lipid lateral diffusion on the rate and degree of
polymerization. Macromolecules 32, 73–78.
Roach, C., Feller, S.E., Ward, J.A., Shaikh, S.R., Zerouga, M., Stillwell, W., 2004. Compar-
ison of Cis and Trans fatty acid containing phosphatidylcholines on membrane
properties. Biochemistry 43, 6344–6351.
Litzinger, D.C., Huang, L., 1992. Phosphatidylethanolamine liposomes – drug deliv-
ery, gene-transfer and immunodiagnostic applications. Biochimica et Biophysica
Acta 1113, 201–227.
Masuda, M., Jonkheijm, P., Sijbesma, R.P., Meijer, E.W., 2003. Photoiniti-
ated polymerization of columnar stacks of self-assembled trialkyl-1,3,5-
benzenetricarboxamide derivatives. Journal of the American Chemical Society
125, 15935–15940.
Ross, E.E., Rozanski, L.J., Spratt, T., Liu, S.C., O’Brien, D.F., Saavedra, S.S., 2003a. Planar
supported lipid bilayer polymers formed by vesicle fusion 1. Influence of diene
monomer structure and polymerization method on film properties. Langmuir
19, 1752–1765.
Ross, E.E., Spratt, T., Liu, S.C., Rozanski, L.J., O’Brien, D.F., Saavedra, S.S., 2003b. Pla-
nar supported lipid bilayer polymers formed by vesicle fusion. 2. Adsorption of
bovine serum albumin. Langmuir 19, 1766–1774.
Matsumoto, A., Horie, A., Otsu, T., 1991. Synthesis and thermal-properties
of poly(adamantyl sorbate). Macromolecular Rapid Communications 12,
681–685.
Sisson, T.M., Lamparski, H.G., Kolchens, S., Elayadi, A., O’Brien, D.F., 1996. Cross-
linking polymerizations in two-dimensional assemblies. Macromolecules 29,
8321–8329.
Matuo, H., Rice, D.K., Balthasar, D.M., Cadenhead, D.A., 1982. Mixed monolayer stud-
ies of 12-hydroxyoctadecanoic acid and its esters. Chemistry and Physics of
Lipids 30, 367–380.
Meier, H., Sprenger, I., Barmann, M., Sackmann, E., 1994. Radical polymerization of
amphiphiles in a 2-dimensional solution (mixed vesicles). Macromolecules 27,
7581–7588.
Torchilin, V., Weissig, V., 2003. Liposomes, A Practical Approach, 2nd ed. Oxford
University Press Inc., New York.
Tyminski, P.N., Ponticello, I.S., O’Brien, D.F., 1987. Polymerizable dienoyl lipids as
spectroscopic bilayer-membrane probes. Journal of the American Chemical Soci-
ety 109, 6541–6542.
Wu, S., Huntsber, J.R., 1969. Intermolecular interaction in some mixed polymer
monolayers at air water interface. Journal of Colloid and Interface Science 29,
138–147.
Mueller, A., O’Brien, D.F., 2002. Supramolecular materials via polymerization of
mesophases of hydrated amphiphiles. Chemical Reviews 102, 727–757.