MOLECULAR MECHANICS AND CALORIMETRIC STUDIES OF
PHOSPHATIDYLETHANOLS
Shusen Li, Hainan Lin, Guoquan Wang, and Ching-hsien Huang*
Department of Biochemistry and Molecular Genetics
University of Virginia School of Medicine
Charlottesville, VA 22908
Running Title: Studies on phosphatidylethanols
Subject Area: Lipids and Biophysical Chemistry
*To whom all correspondence and reprint requests should be addressed:
Ching-hsien Huang
Department of Biochemistry and Molecular Genetics
University of Virginia School of Medicine, Box 800733
Charlottesville, VA 22908
Fax: 804-924-5069, E-mail: ch9t@virginia.edu
1
ABSTRACT
Phosphatidylethanols (PEth) are negatively charged diacyl phospholipids that are
ubiquitously present in humans under the condition of alcohol intoxication. These lipids, derived
in vivo from other naturally occurring phospholipids such as phosphatidylcholines (PC) via
transphosphatidylation reaction as catalyzed by phospholipase D in the presence of ethanol, are
well known to affect many biochemical properties of the cell membranes in humans. In this
communication, we applied the combined approach of molecular mechanics (MM) simulations
and high-sensitivity differential scanning calorimetry (DSC) to investigate the structure and
phase transition behavior of PEth. We first determined the energy-minimized structures of
tetrameric C(15):C(15)PEth arranged in two types of packing motif by the MM approach. An
inwardly bent orientation of the lipid headgroup was observed; specifically, the methyl terminus
of PEth’s headgroup was juxtaposed intramolecularly to the C(2) atom of the sn-2 acyl chain.
Clearly, this headgroup conformation was rather unique among all naturally occurring
phospholipids. Subsequently, the phase transition behavior of the fully hydrated lipid bilayers
prepared individually from 11 species of saturated C(X):C(Y)PEth with the same MW was
studied by DSC, and the resulting Tm values were codified in terms of the normalized acyl chain
asymmetry (C/CL). A V-shaped Tm profile was observed in the plot of Tm versus C/CL for
each subclass of these lipids, suggesting two types of packing motif for C(X):C(Y)PEth at T <
Tm. Moreover, it was observed that within each packing motif these Tm values were, on average,
2.0 ± 0.9 ºC smaller than the Tm values of the corresponding saturated PC. However, based on
the unique headgroup conformation of PEth, we were able to predict that monounsaturated PEth
with a cis double bond near the H2O/hydrocarbon interface would exhibit a higher Tm than the
corresponding PC. Most interestingly, this prediction was indeed borne out by DSC results
obtained with C(18):C(20:15)PC and C(18):C(20:15)PEth.
Key
Words:
Differential
scanning
calorimetry;
phosphatidylcholines; phosphatidylethanols
2
molecular
mechanics
simulations;
INTRODUCTION
It is well known that when fully hydrated lipid bilayers prepared from one-component
diacyl phospholipids are subjected to heating in the differential scanning calorimeter (DSC),
multiple phase transitions are commonly observed to take place at characteristic temperatures (1).
Of these multiple transitions, the chain-melting or the gel-to-liquid crystalline phase transition
occurring at Tm is perhaps the most extensively studied one. Under the same experimental
conditions such as the ambient pressure, a fixed pH, and the same composition of the aqueous
medium, variations in Tm of the lipid bilayer prepared from different species within the same
lipid class can often be correlated to changes in the lipid structure. For a given class of
phospholipid such as phosphatidylcholine (PC) or phosphatidylethanolamine (PE), the structural
changes usually refer to variations in the acyl chain composition and the backbone linkage.
For a number of years, our laboratory has employed the high-sensitivity DSC to
investigate the effects of acyl chain composition, particularly the acyl chain asymmetry and the
chain unsaturation, on the phase transition temperature (Tm) as well as the transition enthalpy (
H) of the lipid bilayer composed of PC and PE (2-4). For instance, the Tm and H values for
alarge number of C(X):C(Y)PC (or PE) have been tabulated recently (4). Here, the abbreviations
of C(X) and C(Y) in C(X):C(Y)PC or PE designate the total numbers of carbon atoms in the sn-1
and sn-2 acyl chains, respectively, of a given saturated PC or PE molecule. In parallel, the
structures of these asymmetric phospholipids, in terms of structural parameters, in the gel-state
bilayer have also been studied extensively using the molecular mechanics (MM) approach (4-6).
For instance, the structural parameter C/CL, the normalized acyl chain asymmetry, has been
used to specify the acyl chain asymmetry in saturated phosphatidylcholines, C(X):C(Y)PC,
packed in the gel-state bilayer (2,3). Here, the numerator of the ratio C/CL is the effective chain
length difference in C-C bond length units between the two acyl chains of a C(X):C(Y)PC
molecule packed in the gel-state bilayer. The denominator of the ratio, CL, is the effective length,
also in C-C bond length units, of the longer acyl chain of the same molecule. For any
C(X):C(Y)PC, C value can be calculated from the formula: C = | X-Y+1.5|. If the sn-1 acyl
chain is longer than the sn-2 acyl chain, then CL can be calculated as CL = X – 1; if, on the other
3
hand, the sn-2 acyl chain is longer, then the following equality holds: CL = Y – 2.5. For
C(X):C(Y)PC with C/CL < 0.42, the lipids are known to packed into a partially interdigitated
motif at T < Tm. In a partially interdigitated bilayer, the sn-1 acyl chain of one C(X):C(Y)PC
molecule is juxtaposed to the sn-2 acyl chain of another lipid molecule from the opposing leaflet.
In contrast, lipids with C/CL values in the range of 0.42-0.60 are known to pack into the mixed
interdigitated motif at T < Tm. In this parking motif, the shorter acyl chain of one C(X):C(Y)PC
molecule faces the shorter chain of another C(X):C(Y)PC in the opposite leaflet, and the longer
chains of both molecules extend fully across the hydrocarbon core of the gel-state lipid bilayer.
One can thus predict the packing motif of any C(X):C(Y)PC in the gel-state bilayer simply by
knowing the lengths of the two acyl chains. Moreover, for C(X):C(Y)PC with C/CL < 0.42, the
Tm value can be related to a simple first-order function of C and N (4), where N, in C-C bond
length units, is the effective thickness of the hydrocarbon core of the gel-state bilayer units that
can be calculated readily from X and Y (4). It is thus possible to predict accurately the Tm of a
large number of saturated PC simply based on the chemical formula of the lipid species (4). The
same combined approach of DSC and MM outlined above has also been applied by our
laboratory to characterize the packing motif and to determine the Tm value of C(X):C(Y)PE (4,7).
More recently, we have extended these studies to phosphatidylglycerol (PG) (8). Now, we further
extend these studies to another class of phospholipids, phosphatidylethanol (PEth).
Phosphatidylethanols are negatively charged diacyl phospholipids. They usually are
absent in human cell membranes; however, they are ubiquitous in humans under the condition of
alcohol intoxication. Consequently, PEth is often referred to as a pathological phospholipid.
PEth has been shown by many investigators to affect significantly the biochemical properties of
many types of human cell (for an in-depth review, see ref. 9). For instance, the presence of PEth
in erythrocytes as a result of alcohol intoxication can lead to a marked decrease in intracellular
[Ca2+]. This PEth-induced effect has been attributed to an increased enzymatic activity of
membrane bound Ca2+-ATPase (10). The ability of PEth to affect cellular functions is most likely
related to its own structural and physicochemical characteristics; hence, a combined study of the
structure and the physicochemical properties of PEth is important. The combined approach of
4
MM and DSC studies to characterize the structures and phase transition behavior of a total of 13
species of saturated and monounsaturated PEth as described in this communication is interesting
in its own right; moreover, the results obtained may be of clinically relevance.
MATERIAL AND METHODS
Materials. All fatty acids and lysophosphatidylcholines used for the semi-synthesis of saturated
and monounsaturated PC were purchased from Sigma (St. Louise, MO) and Avanti Polar Lipids
(Alabaster, AL), respectively. Phospholipase D, Type I, from cabbage, which was employed as
the enzyme-catalyst for the transphosphatidylation reaction, was supplied by Sigma. Silica Gel 60
was obtained from Merck (Darmstadt, Germany). All other chemicals were of reagent grade and
organic solvents were of spectroscopic grade; they were purchased from various commercial
sources.
Lipid synthesis. In this study, 13 molecular species of PC were synthesized. Eleven of them
were saturated C(X):C(Y)PC and two were monounsaturated species, viz. C(18):C(20:15)PC
and C(18):C(22:113)PC. These lipids were semisynthesized at room temperature by acylation of
CdCl2 adducts of lysophosphatidylcholine, in dry chloroform, with fatty acid anhydride that was
prepared in situ from fatty acid and dicyclohexylcarbodiimide, in the presence of catalyst 4pyrrolidinopyridine, according to the modified procedure of Mena and Djerassi (11) as described
previously by this laboratory (12). The synthesized PC were purified by silica gel 60 (mesh size:
230-400) column chromatography (12), and were judged by TLC to be  98% pure.
The 13 molecular species of PC synthesized above were converted to the corresponding
PEth. The PC  PEth conversion was carried out in the presence of excess ethanol via the
transphosphatidylation reaction as catalyzed by phospholipase D using the detail procedure of
Omodeo-Sale et al (13). Phosphatidylethanols (50 mg) obtained from this standard procedure
were Ca2+ bound. The free and the bound Ca2+ ions were then removed by the following two
steps: the lipids were first dissolved in 30 ml of chloroform/methanol (2:1, v/v) and then washed
with 6 ml of 50 mM EDTA (pH = 7.85). The lower phase was collected and evaporated to
dryness. The residues were redissolved in 30 ml of chloroform/methanol (2:1, v/v) and washed
with 6 ml dilute HCl solution (pH = 2). The Ca2+-free lipids were then subjected to the ion-
5
exchange column chromatography using DEAE Sephadex A25. A total of 20 bed volumes of
methanol containing 10 mM sodium acetate was used as the eluent for the ion-exchange column
chromatography. The lipids in the eluate were monitored by TLC using the solvent system of
chloroform: methanol: 25% NH4OH (60:30:5, v/v/v). The lipids, after dryness, were redissolved
in 30 ml of chloroform/methanol (2:1, v/v), and the sodium acetate was removed from the
organic solvent by an extraction of a fixed volume, 6.0 ml, of H2O. At this final stage, the lipids
were in the form of sodium salts. After removal of the organic solvent, the lipids were dissolved
in a minimal volume of benzene and then lyophilized. The resulting lipid powder was kept in a
deep-freezer prior to DSC experiments.
Molecular mechanics calculations. Molecular mechanics calculations, which include the model
building of the energy-minimized structures of oligomeric assemblies of C(15):C(15)PC and
C(15):C(15)PEth arranged in different motifs and the generation of the overall steric energy
associated with the energy-minimized structure, were performed with Allinger’s MM3 force field
(version 92), implemented on an IBM RS/6000 computer workstation as described previously
(14). This MM3(92) program was provided by QCPE, Department of Chemistry, Indiana
University. The molecular graphics representations of various lipid structures arrived at by
computer-based MM calculations were displayed on a Pentium 500 platform equipped with
HyperChem 5.1 software (HyperCube, Gainesville, FL).
It should be mentioned that the MM3(92) program was originally developed for
hydrocarbons by Allinger et al. (15). Since hydrocarbon chains are by far the most predominant
moieties in membrane lipids, it is thus totally justified in employing the MM3 program to
simulate the lipid structure. However, the MM3(92) program does not contain force field
parameters for the charge group (i.e. the phosphate group) that exits in the headgroup moiety of
phospholipid. Additional force field parameters must, therefore, be added. The values of these
missing parameters, however, can be calculated as shown previously by this laboratory (16).
More recently, the optimized values of these force field parameters have been elegantly
developed for phosphosphingolipids by DuPré and Yappert (17), and these values can be
incorporated into the MM3(92) program. Consequently, the energy minimization routine can
6
now be performed conveniently using the MM3(92) program to investigate conformational
preferences of the headgroup with respect to the rest of the lipid molecule.
In modeling the various structures of PC and PEth, the monomeric C(15):C(15)PC was
used as the starting point. The atomic coordinates (or torsion angles) of C(14):C(14)PC
molecules, A and B forms, in single crystals were available (18). The torsion angles of the B
form, determined by x-ray diffraction, were used in this study as the initial set of torsion angles
for the crude structure of C(15):C(15)PC, in which a methylene unit was added, via the C-C
trans bond, to each acyl chain of C(14):C(14)PC. This crude structure was then subjected to the
energy minimization using Alliniger’s MM3(92) program. The resulting refined structure was
subsequently used as a basic structural unit to construct the energy-minimized structure of
dimeric C(15):C(15)PC with a trans-bilayer motif, which, in turn, was further expanded to the
tetrameric and octomeric assemblies with a trans-bilayer motif. The application of the MM3(92)
program in determining the structures of various oligomeric phospholipids was described
previously (5,7,10,16). In this communication, the molecular structure of tetrameric
phospholipids obtained with the MM3 energy refinement was considered, to a first
approximation, as the simple bilayer model. The energy-minimized structure of the tetramer,
however, was taken from the simulated octomer. Specifically, the two adjacent trans-bilayer
dimers located in the central section of the octomer were isolated; they were molecular
graphically illustrated to represent the simulated structure of the tetramer. The relevant torsion
angles of each lipid molecule in the tetramer obtained in this way are presented in Table I.
In addition, the crude molecular structure of monomeric C(15):C(15)PEth was initially
obtained from C(15):C(15)PC by replacing the trimethylated N(+)-group of PC with a hydrogen
atom. Furthermore, the crude torsion angles of the lipid headgroup thus obtained were adjusted
somewhat, prior to the MM3 energy refinement, to ensure that the local minimum in the potential
energy surface would not be practically reached. Hence, a considerable number of additional
rounds of energy minimization was performed prior to the final arrival of the simulated
monomeric structure of C(15):C(15)PEth. Once the MM3-minimized structure of the monomer
was determined, it was then used as a basic structural unit for the construction of the oligomeric
assembly, similar to the procedure just described for C(15):C(15)PC.
7
High-resolution differential scanning calorimetry. The phase transition behavior of the
aqueous lipid dispersion was studied using a high-resolution Microcal MC-2 differential
scanning calorimeter (Microcal, Northampton, MA). The aqueous lipid dispersion was prepared
by adding 2.0 ml of aqueous solution containing 50 mM NaCl, 5 mM phosphate buffer, pH 7.4,
and 1 mM EDTA into a test tube containing lyophilized lipid powder of ~4.0 mg. The final lipid
concentration was checked by the phosphate determination. After vortexing for several minutes,
the sample tube was sealed under N2 and stored at 4 C for a minimum of 24 h prior to the DSC
experiment. The first DSC heating scan was run over a temperature range of ~30 C from about
(Tm–15C) to about (Tm+15C). A cooling scan and then 2-3 consecutive heating scans followed
it over the same temperature range. All samples were scanned at a constant heating rate of 15 
C/h. To ensure the approximately similar thermal history for all lipid samples, only the second
heating scan is chosen to represent the DSC heating curve for each lipid sample. And this second
DSC heating curve is presented in this communication. The Tm and H values were calculated
from the second DSC heating curve using the software provided by Microcal as described
elsewhere (7,19). The Tm of each lipid sample was reproducible to within ± 0.1 ºC. The H
value, however, has a larger error (up to ± 20%), particularly for unsaturated lipids (19).
RESULTS AND DISCUSSION
The energy-minimized structures of tetrameric C(15):C(15)PEth packed in bilayer motifs.
The tetrameric C(15):C(15)PEth packed in two different motifs obtained by MM
calculations are shown in Figures 1A and B, in which two types of molecular graphics, the balland-stick and CPK space-filling models, are used to represent each assembly. In one motif, the
sn-1 and sn-2 acyl chains of one lipid molecule are juxtaposed to the sn-2 and sn-1 acyl chains,
respectively, of another lipid molecule in the opposing leaflet. In this case, the repeating
structural unit is a trans-bilayer dimer with twofold symmetry. The second motif shows,
however, that the sn-1 and sn-2 acyl chains of one lipid molecule in one leaflet are aligned endto-end with the sn-2 and sn-1 acyl chains, respectively, of two adjacent lipid molecules in the
opposing leaflet. The repeating unit is thus a trans-bilayer trimer.
8
It should be stated that the molecular graphics shown in Figure 1 are generated from the
atomic coordinates (torsion angles, bond-lengths, etc.) of tetrameric lipids confined to two
packing motifs, and these atomic coordinates are obtained from computer-based MM
calculations using Allinger’s MM3(92) program. The initial data input for carrying out the MM
calculation are the properly adjusted atomic coordinates of crystal C(14):C(14)PC, which were
determined by x-ray diffraction techniques (18). Consequently, the MM3 generated structures
shown in Figures 1A-B represent the energy-minimized structures of two modeled
C(15):C(15)PEth lipid bilayers packed in the crystalline state. These crystalline structures are
interesting in their own rights; in particular, the headgroup conformation of C(15):C(15)PEth is
rather unique and interesting in comparison with other naturally occurring phospholipids (vide
post). In addition, the structural difference between phospholipids packed in one motif and
another in the crystalline state can, to a first approximation, be extrapolated to that in the gelstate. This approximation is not unreasonable, since only restricted rotational motions with rather
small amplitudes were observed for phospholipids assembled in the gel-state bilayer (20).
However, it must be emphasized that this extrapolation is justified only if the information
obtained with MM3 calculations is used for a qualitative comparison. For instance, the steric
energies calculated for the two simulated structures shown in Figures 1A and B are virtually
identical, and we can extrapolate the results by saying that these lipid assemblies in the gel state
also are most likely to have a similar value of steric energy. However, the absolute values of
steric energy obtained with lipid assemblies in Figures 1A and B cannot be used directly for the
corresponding assemblies in the gel state. Furthermore, the MM3-minimized structures can also
be useful in the design of other experiments. In this communication, for instance, we first
compare the tetrameric structure of saturated PEth with that of tetrameric PC obtained by MM3
calculations. The structural difference thus obtained is subsequently used not only to interpret our
calorimetric data carried out with saturated PEth and PC, but also as a guide to design the
synthesis of appropriate unsaturated PEth (vide post). Now, let us first turn back to examine the
MM3-minimized structures of various lipid assemblies as illustrated graphically in Figure 1.
In the tetrameric assembly shown in Figures 1A and B, two lipid headgroups are
positioned at each of the two opposing ends of the modeled bilayer, and the hydrophobic core of
9
the lipid bilayer is sandwiched in between the two pairs of headgroup. The various torsion angles
of the headgroup sequence (i, where i = 1, 2, 3, and 4) for each C(15):C(15)PEth molecule in
the tetramer are summarized in Table 1. It is evident that the value of each i varies only slightly
when the lipid molecule changes its position or packing motif. In addition, the overall steric
energies of tetrameric C(15):C(15)PEth shown in Figures 1A and B are 169.4 and 169.1
kcal/mol, respectively, indicating the virtually identical stability. As a result, we have chosen the
trans-bilayer dimer motif depicted in Figure 1A as the representative structure model for the lipid
bilayer. For comparison, the MM3-minimized structures of tetrameric C(15):C(15)PC with two
packing motifs were also obtained. These structures are graphically illustrated in Figures 1C and
D. Again, computational results from MM calculations indicate that the steric energy of the
tetrameric C(15):C(15)PC packed according to the trans-bilayer dimer motif is virtually identical
to that characterized by the trans-bilayer trimer motif.
It should be noted that the chemical compositions of the acyl chains for all phospholipids
shown in Figures 1A-D are identical. Moreover, in each packing motif the relative orientations of
the two acyl chains are also indistinguishable between C(15):C(15)PEth and C(15):C(15)PC.
Specifically, all fully extended sn-1 acyl chains of these phospholipids in the tetramer are
observed to run in parallel to each other with their zigzag plans lying perpendicularly to the paper
plane, while the zigzag planes of all sn-2 acyl chains are oriented parallel to the paper plane. It
should be pointed out that although the sn-1 and sn-2 acyl chains in each lipid molecule have the
same total number of 15 carbons, there is an effective chain length difference between the sn-1
and sn-2 acyl chains. This difference is evident from the clear separation of the two methyl ends
along the long molecular axis in each lipid molecule. In fact, this effective chain-length
difference is one of the structural parameters that can be used to characterize the phase transition
behavior of the lipid bilayer (21). Despite the structural similarities in the hydrocarbon region,
some distinct differences between C(15):C(15)PEth and C(15):C(15)PC can be found in the polar
headgroup region.
As shown in Figure 1A, the headgroup terminus of C(15):C(15)PEth is an ethyl moiety,
designated by Eth. This Eth moiety is rather hydrophobic in nature. Interestingly, the methyl end
of Eth, labeled as M in Figure 1A, is oriented toward the hydrocarbon/H2O interface in line with
10
the long axis of the sn-2 acyl chain. The separation distance between M and the C(2)-atom of the
sn-2 acyl chain (designated as C2 in Figure 1A) is, on average, 4.3 Å. It should be mentioned that
the optimal van der Waals attractive interaction between two nonbonded (sp3)-carbons occurs at
the separation distance of 4.08 Å (6). The observed separation distance of 4.3 Å thus indicates
that there is a favorable attractive interaction between M and C2. This close and favorable van
der Waals contact distance can be most readily ascertained by looking at the CPK space-filling
model of the tetrameric assembly shown in Figure 1A. This bent down orientation of the
headgroup terminus seen in Figure 1A is a rather unique conformation for phospholipids. In
almost all other naturally occurring phospholipids, the headgroup terminus is polar and
hydrophilic. In the case of PC, for instance, the terminal end of the choline functionality is a
trimethylated N(+)-group, and this group is polar and hydrophilic due to its delocalized positive
charge. Consequently, the headgroup terminus of PC in the bilayer can be expected to point away
from the hydrocarbon/H2O interface with the charged group being in direct contact with the bulk
water. Such an orientation is indeed exhibited by lamellar C(15):C(15)PC as shown in Figures
1C and D.
The actual conformations of the headgroup termini of C(15):C(15)PEth and
C(15):C(15)PC are determined mainly by the torsion angles of 4, which specify the
energetically preferable “rotational isomeric state” of the central bond (O–CH2) in the three-bond
sequences of PO3–O–CH2–CH3 (PEth) and PO3–O–CH2–CH2 (PC). For the energy-minimized
headgroup structures of C(15):C(15)PEth and C(15):C(15)PC, shown in the expanded views
illustrated in Figures 1E & F, the averaged values of 4 are -49.3 and -162.6, respectively
(Table1). With these torsion angles in mind, one can appreciate that a counterclockwise rotation
of 113.3 about the central bond (O–CH2) can steer the headgroup terminus of C(15):C(15)PEth
from the position depicted in Figure 1E to the position corresponding to the vicinity of –
N+(CH3)3 group seen in Figure 1F. Specifically, this headgroup reorientation is accompanied by a
change of 4.3 to 6.6 Å in the separation distance between the methyl carbon of Eth and the C(2)atom of the sn-2 acyl chain. As a result, the van der Waals attractive interaction between M and
C2 is diminished almost entirely. Furthermore, this conformational change leads to a very
energetically unfavorable situation, in which the hydrophobic Eth moiety is fully exposed to the
11
aqueous medium. By contrast, the lipid in the original state has its headgroup terminus directed
towards the interface. As such, the water accessible surface area is minimal for the hydrophobic
Eth moiety (Figure 1E), and the headgroup as a whole is thus less hydrated. This conformational
state, characterized by 4 = -49.30.6°, is energetically favorable, since it is stabilized by the
increased hydrophobic effect and the attractive van der Waals contact interaction between the
methyl group of Eth and the C(2)-methylene of the sn-2 acyl chain.
In addition to the orientation, the headgroups of C(15):C(15)PEth and C(15):C(15)PC are
also quite different in other aspects. The bulky trimethylated-N(+) group in C(15):C(15)PC is
replaced by a proton in C(15):C(15)PEth. The effective cross-sectional area of the headgroup is
thus considerably smaller for C(15):C(15)PEth irrespective of the lipid state. In fact, as shown in
Figure 1A, the effective cross-sectional area parallel to the bilayer surface is virtually identical to
the overall cross-sectional area occupied by the sum of two acyl chains. In addition, the bulky
trimethylated-N(+) group in C(15):C(15)PC bears a positive charge, which is positioned
topologically in close proximity to the negatively charged phosphate group of an adjacent
C(15):C(15)PC molecule (Figure 1C). On the other hand, there is no positively charged
trimethylated-N(+) group in C(15):C(15)PEth; hence, C(15):C(15)PEth also differs from
C(15):C(15)PC in carrying a net negative charge near the neutral pH. An electrostatic repulsion
among lipids due to the negatively charged phosphate groups must, therefore, take place within
each leaflet in the C(15):C(15)PEth bilayer near the neutral pH.
The main phase transition behavior of C(15):C(15)PEth
In Figure 2, the second DSC heating curve obtained with the aqueous dispersion of
C(15):C(15)PEth in the presence of 50 mM NaCl, 5 mM phosphate buffer (pH 7.4), and 1 mM
EDTA is illustrated. A sharp, single endothermic transition with Tm = 32.0 °C and H = 7.0
kcal/mol is observed within the temperature range of 15 to 45 °C. This transition is reproducible
upon repeated reheating; hence, it is assigned as the chain-melting transition or the gel-to-liquid
crystalline phase transition. It should be mentioned that the Tm and H values obtained
calorimetrically with C(16):C(16)PEth and C(14):C(14)PEth in 50 mM NaCl were reported
earlier by Bondar and Rowe (21), and our calorimetric data obtained with C(15):C(15)PEth as
12
shown in Figure 2 lie, as expected, in between those published values. For comparison, the
second DSC heating curve for the aqueous dispersion of C(15):C(15)PC is also illustrated in
Figure 2. This DSC curve consists of two endothermic transitions: a small transition peaked at
25.3 °C (H < 1.0 kcal/mol) and a sharp, large transition peaked at 34.0 C (H = 7.1 kcal/mol).
These low- and high-temperature transitions have been previously assigned as the pretransition
and the main phase transition, respectively, for C(15):C(15)PC (22). Two characteristic features
associated with the DSC heating curves shown in Figure 2 are worth emphasizing. First, the
pretransition, which is clearly observed for C(15):C(15)PC, is not detected in the DSC heating
curve of C(15):C(15)PEth. Second, the Tm and H values associated with the chain-melting
transition of C(15):C(15)PEth are slightly smaller in magnitude than the corresponding values of
C(15):C(15)PC.
Saturated phosphatidylcholines with tilted acyl chains in the gel-state bilayer usually
exhibit calorimetrically the pretransition, and it is generally agreed that the chain tilt can be
attributed to a mismatch between the effective cross-sectional area of the polar headgroup and the
effective cross-sectional area of the sum of two acyl chains (23). The absence of the pretransition
in the C(15):C(15)PEth thermogram may, therefore, be taken as evidence suggesting that, at T <
Tm, C(15):C(15)PEth molecules are packed in the lipid bilayer without having a chain tilt. More
specifically, the absence of the pretransition implies that, in the gel-state bilayer, the headgroup
of C(15):C(15)PEth has approximately the same effective cross-sectional area as the sum of two
acyl chains. Energy-minimized structures of tetrameric C(15):C(15)PEth and C(15):C(15)PC
obtained with MM calculations indeed show that the headgroup of C(15):C(15)PEth is
discernibly smaller than that of C(15):C(15)PC (Figures 1A and C). Moreover, the effective
cross-sectional area of the headgroup in C(15):C(15)PEth is approximately identical to that of the
two acyl chains. In comparison with the (CH3)3N(+)-group containing C(15):C(15)PC, the
headgroup-headgroup steric repulsion is thus relatively insignificant in C(15):C(15)PEth. On the
other hand, C(15):C(15)PEth bears a net negative charge at the neutral pH; hence, the
electrostatic repulsion between the headgroups of C(15):C(15)PEth exists in the gel-state bilayer.
Whereas in the C(15):C(15)PC bilayer, the lipid molecules are zwitterionic without having a net
charge near the neutral pH; hence, the electrostatic repulsion among neighboring headgroups
13
does not take place. If the electrostatic repulsion present in C(15):C(15)PEth is similar in
strength to the steric repulsion in C(15):C(15)PC, then the gel-state bilayers of C(15):C(15)PEth
and C(15):C(15)PC can be expected to display a similar thermal stability. Interestingly, the
calorimetric results illustrated in Figure 2 indicate that, under the same experimental conditions,
the thermally induced phase transition behavior is comparable for bilayers prepared from these
two lipids. Nevertheless, the values of Tm and H exhibited by C(15):C(15)PEth are slightly
smaller than those displayed by C(15):C(15)PC. We suggest that the smaller values of Tm and 
H are most likely due to the possibility that the electrostatic effect in the C(15):C(15)PEth bilayer
at T < Tm is larger in magnitude than the steric effect present in the gel-state bilayer of
C(15):C(15)PC.
The phase transition behavior of a series of saturated mixed-chain phosphatidylethanols
with a common molecular weight of 648.0
In addition to C(15):C(15)PEth, a series of ten mixed-chain C(X):C(Y)PEth was semisynthesized in this laboratory. These C(X):C(Y)PEth molecules share a constant MW of 648.0.
In addition, they share a constant number of 30 as the sum of X and Y, viz. the total number of
carbon atoms in the two acyl chains of each lipid species is identical to that of C(15):C(15)PEth.
However, each lipid species in this series differs from each other in terms of C/CL, the
normalized acyl chain asymmetry, and the value of C/CL for each lipid species is listed in Table
2. High-resolution DSC experiments were performed to examine the phase transition behavior of
these lipids in the aqueous solution containing 50 mM NaCl , 5 mM phosphate buffer, pH 7.4,
and 1 mM EDTA.
In Figure 3 some representative results of the DSC experiments are shown. Specifically,
the endothermic transition profiles obtained from the second DSC heating curves for two pairs of
positional isomers, C(14):C(16)PEth (C/CL = 0.037)/C(16):C(14)PEth (C/CL = 0.233) and
C(12):C(18)PEth (C/CL = 0.290)/C(18):C(12)PEth (C/CL = 0.441), are illustrated. It is
evident that each lipid sample exhibits a single and nearly symmetrical transition. These single
transitions are assigned as the chain-melting transitions or the gel-to-liquid crystalline phase
14
transitions. The values of Tm, H, and S associated with the chain-melting transitions of all
eleven species of C(X):C(Y)PEth are summarized in Table II. Within each packing motif (vide
infra), these Tm values obtained with the series of C(X):C(Y)PEth with X+Y=30 are invariably
smaller, by 2.0 ± 0.9 C, than those observed for the corresponding C(X):C(Y)PC (22). The 11
species of C(X):C(Y)PEth with X+Y=30, shown in Table II, can be divided into two groups. In
Group I, lipids have a longer effective sn-1 acyl chain or X  Y – 1.5, and lipids in Group II are
characterized by a longer effective sn-2 acyl chain or Y – 1.5 > X. When the Tm values of lipids
within Group I and II are plotted against the normalized chain-length difference (C/CL) as
shown in Figure 4, two V-shaped Tm profiles are observed. Specifically, the Tm value is seen to
decrease almost linearly as C/CL increases initially, reaching a nadir around C/CL of 0.40.
Thereafter, the Tm value increases as the value of C/CL is increased up to about 0.60. It should
be mentioned that the biphasic Tm profiles are drawn based on similar profiles previously
observed for several series of C(X):C(Y)PC. The biphasic Tm profile for C(X):C(Y)PC has been
related to the bilayer packing motif of the acyl chains in the gel-state bilayer (3,22,24). For lipids
with C/CL < 0.42, the acyl chains of C(X):C(Y)PC are packed into the partially interdigitated
motif at T < Tm, which is similar to the one shown in Figure 1C. However, when mixed-chain
C(X):C(Y)PC are highly asymmetrical with C/CL values in the range of about 0.42-0.60, they
self-assemble at T < Tm into the mixed interdigitated motif in excess water (3,22,24). In this
mixed interdigitated packing motif, the methyl end of the short chain of one lipid molecule meets
end-to-end the methyl terminus of the short chain of another lipid molecule in the opposing
leaflet, and the long chains of both lipid molecules extend fully across the hydrocarbon core of
the gel-state bilayer. Furthermore, the headgroup of the highly asymmetric lipid in the mixed
interdigitated gel-state bilayer encompasses three all-trans acyl chains (25-28). Based on the
extensive structural and calorimetric information available for mixed-chain C(X):C(Y)PC with
mixed acyl chains (6,29) and based on the Tm profiles of C(X):C(Y)PEth observed in Figure 4,
we suggest that there are two types of packing motif at T < Tm for C(X):C(Y)PEth with X+Y=30.
Specifically, the lipids are most likely packed into the partially interdigitated motif, if C/CL <
0.40. For lipids under study with C/CL  0.40, however, the mixed interdigitated motif prevails.
15
In comparison with the published Tm values exhibited by C(X):C(Y)PC with X+Y = 30
(22), all the Tm values exhibited by PEth shown in Figure 4 are, within each packing motif, about
2.0 ± 0.9 C smaller. For lipids packed into the partial interdigitated motif, the relative Tm values
observed for C(X):C(Y)PC and C(X):C(Y)PEth at a fixed C/CL value can be interpreted as
follows: In the gel-state bilayer, the overall headgroup-headgroup charge repulsion between
phosphatidylethanols is slightly larger in magnitude than the effective headgroup-headgroup
steric repulsion between phosphatidylcholines. This larger charge repulsion gives rise to a lower
Tm (or H). The mixed interdigitated bilayer, however, is characterized by the headgroup
encompassing three acyl chains; hence, when lipid molecules are packed into the mixed
interdigitated gel-state bilayer, there is a significant increase in the lateral headgroup-headgroup
separation distance. As such, the headgroup-headgroup charge repulsion and the headgroupheadgroup steric repulsion in the bilayers of PEth and PC, respectively, must be relieved
simultaneously. The net result is that the relative magnitude of Tm remains the same as the lipid
bilayers of PEth and PC undergo the partially interdigitated  mixed interdigitated transition. It
should also be mentioned that although the headgroup-headgroup repulsion is reduced in the
mixed interdigitated bilayer, the terminal methyl group of the longer acyl chain is now partially
exposed to water in this packing mode. This means that an energetically unfavorable situation
still persists as the lipid bilayer of PEth or PC undergoes the partially interdigitated  mixed
interdigitated transition.
The phase transition behavior of monounsaturated PEth with different positions of the cis
double bond along the sn-2 acyl chain
Thus far, we have seen that the Tm value of all saturated C(X):C(Y)PEth are slightly
smaller than those of the corresponding PC obtained under the same experimental conditions;
however, the orientations and structures of these two lipids’ headgroups are, based on MM
simulations, distinctively different. An interesting question can immediately be raised as to
whether the Tm of PEth can be expected to be larger than PC for species other than the saturated
C(X):C(Y)PEth.
16
Based on our previous studies of unsaturated lipids, it is known that the phase transition
behavior of PC(or PE) with a fixed number of cis-double bonds (-bonds) is markedly affected
by the position of -bonds along the acyl chain (4). Interestingly, as the -bonds in an
unsaturated lipid are positioned very closely to the H2O/hydrocarbon interface as represented by
C(18):C(20:25,8)PE, the experimental Tm value exhibited by the lipid bilayer composed of
C(18):C(20:25,8)PE is inevitably smaller than the expected Tm value (30). This has been
interpreted as due to the penetration of water molecules into the local regions of the hydrocarbon
chains in the immediate neighborhoods of 5-bonds at T < Tm. Hence, the overall lateral chainchain van der Waals contact interaction is perturbed somewhat in the gel-state bilayer, resulting
in a smaller Tm value. If our interpretation is reasonable, such a perturbation observed for
unsaturated PE or PC should not take place in unsaturated PEth with a 5-double bond in the sn2 acyl chain. In Figure 1A, the C(2)-carbon in the sn-2 acyl chain of the monomer is shown to be
in close van der Waals contact with the methyl terminus of the headgroup within the same
monomer. This is caused by the unique bent down orientation of the Eth moiety as a result of the
averaged 4 torsion angle of -49.3 (Table I). And, importantly, the local region around the C(2)carbon is totally dehydrated. In the presence of a 5-bond in the sn-2 acyl chain, there is no
reasons to believe that the headgroup orientation of PEth will be altered. Hence, water molecules
remain excluded from the interface constituted by C(2)-carbons, and, as a consequence, the local
regions of the hydrocarbon chains surrounding the 5-bonds are also dehydrated. In the
dehydrated state, the perturbation of the overall chain-chain contact interaction by the penetrated
H2O molecules can not be assumed. We, therefore, predict that the Tm value of the lipid bilayer
composed of unsaturated PEth with a 5-bond will be larger than that of the PC counterpart.
Based on this prediction, C(18):C(20:15)PEth, C(18):C(22:113)PEth, C(18):C(20:15)PC, and
C(18):C(22:113)PC were first synthesized. Subsequently, the DSC experiments were carried out
with the aqueous dispersions prepared individually from these lipids.
Figure 5 shows the second DSC heating curves obtained from the aqueous dispersions of
C(18):C(20:15)PEth, C(18):C(22:113)PEth, C(18):C(20:15)PC, and C(18):C(22:113)PC
prepared in the presence of 50 mM NaCl, 5 mM phosphate buffer, pH 7.4, and 1 mM EDTA.
17
Clearly, the C(18):C(20:15)PEth sample shows a single endothermic transition with a Tm of
43.0 C and a H of 10.4 kcal/mol. The C(18):C(20:15)PC sample, however, exhibits a small,
broad, low-temperature pretransition as well as a large, symmetric, high-temperature transition.
Most interestingly, the values of Tm and H of the high-temperature main transition are 40.7 C
and 9.9 kcal/mol, respectively, which are smaller than the corresponding values obtained with the
C(18):C(20:15)PEth sample. To the best of our knowledge, this is the first time that the Tm (or
H) value of a PEth sample is shown to be greater than that of the PC counterpart. As discussed
already, this calorimetric result is precisely what we have predicted.
Now, let us compare the hydrocarbon region between the C(18):C(22:113)PEth bilayer
and the C(18):C(22:113)PC bilayer, at T < Tm. The chemical compositions of the two acyl
chains in these two lipids are identical; in particular, they both share a common cis double bond
at the position of C(13) in their sn-2 acyl chains. Since the 13-bond is located at the chain center
of the sn-2 acyl chain many rotational isomers with virtually identical steric energy are possible
for these monoenoic lipids in the gel-state bilayer (5). The DSC curve exhibited by the lipids are
thus broader as shown in Figure 5. Furthermore, because the 13-bond is located near the middle
of the sn-2 C22-acyl chain, water molecules are unlikely to penetrate deeply into the gel-state
bilayer to reach the 13-bond. In other words, the local hydrocarbon region in the immediate
neighborhood of the 13-bond is dehydrated in both lipid assemblies at T < Tm. The lateral chainchain contact interactions in the gel-state bilayers of C(18):C(22)PC and C(18):C(22)PEth will,
therefore, be affected equally by the introduction of a common dehydrated 13-bond into their sn2 acyl chains. As already discussed, the Tm value obtained with saturated C(X):C(Y)PEth is, on
average, 2.0  0.9 C smaller than that of the corresponding C(X):C(Y)PC, we expect that the Tm
value of the C(18):C(22:113)PEth bilayer will also be smaller than that of the C(18):C(22:1
13
)PC bilayer by a similar magnitude. In Figure 5, the second DSC heating curves exhibited by
the aqueous dispersions of C(18):C(22:113)PC and C(18):C(22:113)PEth are depicted. It is
obvious that our expectation is indeed borne out by the calorimetric results shown in Figure 5, in
which the Tm value of C(18):C(22:113)PEth is observed to be 2.6 C lower than that of
C(18):C(22:113)PC.
18
CONCLUSIONS
In summary, we have obtained the MM3-minimized structure of tetrameric
C(15):C(15)PEth packed in the bilayer motif. The computational results indicate that the lipid’s
headgroup has a rather unique conformation. Specifically, the ethyl moiety of the headgroup is
bent inwards with the methyl end being in close van der Waals contact with the initial segment of
the sn-2 acyl chain. Because of this bent orientation, the lipid bilayer composed of
phosphatidylethanols is less hydrated at the H2O/hydrocarbon interface than other naturally
occurring phospholipids such as phosphatidylcholines. In addition, our computational results
show that the effective cross-sectional area of C(15):C(15)PEth headgroup appears
approximately equal to that of the sum of two acyl chains. As a consequence, the headgroupheadgroup steric repulsion between lipids in the gel-state bilayer of C(15):C(15)PEth can be
reasonably assumed to be insignificant in comparison with that of C(15):C(15)PC. In addition to
MM calculations, we have also studied the phase transition behavior of lipid bilayers prepared
individually from 11 species of saturated X(X):C(Y)PEth by high-resolution DSC. In the plot of
Tm versus C/CL, a biphasic V-shaped Tm profile is observed, indicating two types of packing
motif of C(X):C(Y)PE at T < Tm In addition, we have performed DSC experiments on
monounsaturated PEth and their corresponding PC. These DSC experiments are designed
specifically to see whether the phase transition temperatures of monounsaturated PEth, relative to
those of the PC counterparts, can be reasonably predicted based on the unique headgroup
orientation of PEth. Most interestingly, the DSC data are indeed in complete accord with our
predicted results. We can, based on the results of the present investigation, draw the following
conclusions:
First,
the
structure
and
the
phase
transition
behavior
of
lamellar
phosphatidylethanols differ discernibly from other naturally occurring phospholipids containing
the same acyl chains. Second, the chain-melting behavior exhibited by lamellar
phosphatidylethanols can be explained reasonably well on the basis of the unique headgroup
orientation of PEth in the gel-state bilayer. Such a unique orientation has been shown by the
MM3-minimized structure of tetrameric C(15):C(15)PEth with a bilayer motif. Third, the unique
headgroup orientation of PEth in the gel-state bilayer can be attributed fundamentally to the
19
hydrophobic nature of the ethyl moiety of the lipid headgroup.
ACKNOWLEDGMENTS. This work was supported, in part, by NIH Grant GM-17452.
20
REFERENCES
1. Chapman, D. (1993) in Biomembranes: Physical Aspects (Shinitzky, M., Ed.), pp. 29-62,
Bababon Publishers, Weinheim, Germany.
2. Huang, C. , and Mason, J. T. (1986) Biochim. Biophys. Acta 864, 423-470.
3. Huang, C. (1990) Klin. Wochenschr. 68, 149-165.
4. Huang, C., and Li, S. (1999) Biochim. Biophys. Acta 1422, 273-307.
5. Li, S., Wang, H., Lin, H., and Huang, C. (1993) Biophys. J. 65, 1415-1428.
6. Huang, C., and Li, S. (1996) in Handbook of Nonmedical Applications of Liposomes, Vol. 1
( Lasic, D. D., and Barenholz, Y., Eds.), pp.173-194, CRC Press, Boca Raton, FL.
7. Wang, Q., Lin, S., Li, S., and Huang, C. (1995) J. Biol. Chem. 270, 22738-22746.
8. Durvasula, R. V., and Huang, C. (1999) Biochim. Biophys. Acta 1417, 111-121.
9. Gustavsson, L. (1995) Alcohol & Alcoholism 30, 391-406.
10. Suju, M., Davila, M., Poleo, G., Docampo, R., and Benaim, G. (1996) Biochem. J. 371, 933938.
11. Mena, P. L., and Djerassi, C. (1985) Chem. Phys. Lipids 27, 257-270.
12. Lin, H., Wang, Z., and Huang, C. (1990) Biochemistry 29, 7063-7072.
13. Omodeo-Sale’, M. F., Cestaro, B., Mascherpa, A., Monti, D., and Masserini, M. (1989)
Chem. Phys. Lipids 50, 135-142.
14. Li, S., Lin, H., Wang, G., and Huang, C. (1996) Biophys. J. 70, 2784-2794.
15. Allinger, N. L., Yuh, Y. H., and Lii, J. (1989) J. Am. Chem. Soc. 111, 8551-8582.
16. Li, S., and Huang, C. (1996) J. Comput. Chem. 17, 1013-1024.
17. DuPré, D. B., and Yappert, M. C. (1999) Theochem-J. Mol. Struct. 467, 115-133.
18. Pascher, I., Lundmark, M., Nyholm, P., and Sundell, S. (1992) Biochim. Biophys. Acta 1113,
339-373.
19. Wang, G., Li, S., Brumbaugh, E. E., and Huang, C. (1999) J. Biol. Chem. 274, 12289-12299.
21
20. Thompson, T. E., Sankaram, M. B., and Huang, C. (1997) in Handbook of Physiology: Cell
Physiology (Hoffman, J. F., and Jamieson, J. D., Eds.), pp. 23-56, Oxford University Press,
Oxford, England.
21. Mason, J. T., and Huang, C. (1981) Lipids 16, 604-608.
22. Bondar, O. P., and Rowe, E. S. (1996) Biophys. J. 71, 1440-1449.
23. McIntosh, T. J. (1980) Biophys. J. 29, 237-246.
24. Bültmann, T., Lin, H., Wang, Z., and Huang, C. (1991) Biochemistry 30, 7194-7203.
25. McIntosh, T. J., Simon, S. A., Ellinton, N. A., and Porter, N. A. (1984) Biochemistry 23,
4038-4044.
26. Hui, S. W., Mason, J. T., and Huang, C. (1984) Biochemistry 23, 5570-5577.
27. Mattai, J., Sripada, P. K., Shipley, G. G. (1987) Biochemistry 26, 3287-3297.
28. Zhu, T., and Caffrey, M. (1993) Biophys. J. 65, 939-954.
29. Mason, J. T. (1996) in Handbook of Nonmedical Applications of Liposomes, Vol. 1 ( Lasic,
D. D., and Barenholz, Y., Eds.), pp.195-218, CRC Press, Boca Raton, FL.
30. Li, S., Wang, G., Lin, H., and Huang, C. (1998) J. Biol. Chem. 273, 19009-19018.
22
Table 1. Relevant torsion angles (in degree) in the headgroup of C(14):C(14)PC2H2O,
C(15):C(15)PEth and C(15):C(15)PC.
1
2
3
4
177
-76
-47
-150
179.0
-143.1
-47.6
-165.4
C(15):C(15)PC, Mol. 2
179.6
-144.2
-45.9
-164.1
C(15):C(15)PC, Mol. 3
177.8
-107.7
-47.1
-159.3
C(15):C(15)PC, Mol. 4
176.3
-136.1
-49.4
-161.5
C(15):C(15)PEth, Mol. 1
179.0
-177.7
76.9
-48.7
C(15):C(15)PEth, Mol. 2
179.6
-179.1
79.4
-49.9
C(15):C(15)PEth, Mol. 3
177.8
179.9
78.7
-49.9
C(15):C(15)PEth, Mol. 4
176.3
-176.7
76.7
-48.6
Lipid
C(14):C(14)PC2H2O
C(15):C(15)PC, Mol. 1
Torsion angles of C(14):C(14)PC2H2O were obtained from single crystals by X-ray diffraction
(18). Torsion angles of C(15):C(15)PC and C(15):C(15)PEth were obtained from tetrameric
assemblies by MM calculations as described in the text. Mol. (1-4) are the numbers assigned to
the four lipid species in the tetrameric C(15):C(15)PEth and C(15):C(15)PC. When the tetramer
is arranged topologically as that illustrated in Figure 1, the numbers 1, 2, 3, and 4 refer to lipid
species appearing in the upper left, lower left, upper right, and lower right of the diagram,
respectively. The torsion angles of various bonds (C2-C1-O11-P-O12-C11-C12) in the lipid
headgroup refer to the energetically mostly favored rotational angle about the central bond of the
following sequences: 1 , C2 - C1 – O11 - P; 2 , C3 - O11 -P - O12; 3 , O11 - P - O12 - C11; 4 , P O12 - C11 - C12.
23
Table II: The structural parameter, the transition temperature, the transition enthalpy, and the
transition entropy for various phospholipids.
Lipid
Group*
C/CL
Tm (oC)
H(Kcal/mol)
S(cal*mol per K)
C(15):C(15)PEth
I
0.107
32.0
7.0
22.9
C(16):C(14)PEth
I
0.233
27.2
6.4
21.3
C(17):C(13)PEth
I
0.344
19.5
5.3
18.1
C(18):C(12)PEth
I
0.441
16.2
8.0
27.6
C(19):C(11)PEth
I
0.528
25.8
9.4
31.4
C(14):C(16)PEth
II
0.037
32.1
7.9
25.9
C(13):C(17)PEth
II
0.172
27.7
6.5
21.6
C(12):C(18)PEth
II
0.290
20.6
5.5
18.7
C(11):C(19)PEth
II
0.394
18.6
6.2
18.5
C(10):C(20)PEth
II
0.486
25.6
8.7
29.1
C(9):C(21)PEth
II
0.568
27.3
10.4
30.0
C(18):C(22:113)PEth
17.0
7.0
24.1
C(18):C(20:15 )PEth
43.0
10.7
33.9
*For C(X):C(Y)PEth with X  Y, these lipids belong to Group I. Lipids in Group II, however, are
characterized by X < Y.
24
FIGURE LEGENDS
Figure 1. Molecular graphics representations of the MM3-minimized C(15):C(15)PEth and
C(15):C(15)PC assemblies. (A) Tetrameric C(15):C(15)PEth with the trans-bilayer dimer motif
as represented by the ball-and-stick (top) and space-filling (bottom) models. (B) Tetrameric
C(15):C(15)PEth with the trans-bilayer trimer motif as represented by the ball-and-stick (top) and
space-filling (bottom) models. (C) Tetrameric C(15):C(15)PC with the trans-bilayer dimer motif
as represented by the ball-and-stick (top) and space-filling (bottom) models. (D) Tetrameric
C(15):C(15)PC with the trans-bilayer trimer motif as represented by the ball-and-stick (top) and
space-filling (bottom) models. Abbreviations used: C2, the C(2) atom in the sn-2 acyl chain; Eth,
the ethyl moiety of the lipid’s headgroup; M, the methyl terminal of the headgroup in PEth; P,
the phosphate moiety of the headgroup. (E) The headgroup of C(15):C(15)PEth, taken from (A).
(F) The headgroup of C(15):C(15)PC, taken from (C).
Figure 2. The second DSC heating curves obtained with the aqueous dispersions of
C(15):C(15)PEth and C(15):C(15)PC, respectively. Scan rate: 15 ºC/h.
Figure 3. The second DSC heating curves exhibited by the aqueous dispersions of
C(X):C(Y)PEth. The total number of carbons in the two acyl chains of each lipid is 30. Scan rate:
15 ºC/h.
Figure 4. Plot of Tm versus C/CL of C(X):C(Y)PEth. The total number of carbons in the two
acyl chains of each lipid is 30.
Figure 5. The relative influence of the cis carbon-carbon double bond position on the phase
transition behavior of monoenoic PEth and PC with the same acyl chain composition. Note that
the Tm value is higher for monoenoic PEth when the cis double bond lies in between C(5) and
C(6) atoms in the sn-2 C20-acyl chain. Whereas the Tm value is higher for monoenoic PC when
the cis double bond lies in between C(13) and C(14) atoms in the sn-2 C22-acyl chain
25