Available online at www.sciencedirect.com
Biochimica et Biophysica Acta 1768 (2007) 2873 – 2881
www.elsevier.com/locate/bbamem
The partition of cholesterol between ordered and fluid bilayers of
phosphatidylcholine: A synchrotron X-ray diffraction study
Lin Chen a,b , Zhiwu Yu a,⁎, Peter J. Quinn b,⁎
a
Key Laboratory of Bioorganic Phosphorous Chemistry and Chemical Biology (Ministry of Education), Department of Chemistry,
Tsinghua University, Beijing 100084, P.R. China
b
Department of Biochemistry, King’s College London, 150 Stamford Street, London SE1 9NN, UK
Received 21 February 2007; received in revised form 27 July 2007; accepted 30 July 2007
Available online 16 August 2007
Abstract
The structure and composition of coexisting bilayer phases separated in binary mixtures of dipalmitoylphosphatidylcholine and cholesterol and
ternary mixtures of equimolar proportions of dipalmitoyl- and dioleoylphosphatidycholines containing different proportions of cholesterol have been
characterized by synchrotron X-ray diffraction methods. The liquid-ordered phase is distinguished from gel and fluid phases by a disordering of the
hydrocarbon chains intermediate between the two phases as judged from the wide-angle X-ray scattering profiles. Electron density distribution
calculated in coexisting bilayer phases shows that liquid-ordered phase is enriched in dipalmitoylphosphatidylcholine and cholesterol and a higher
electron density in the methylene chain region of the bilayer ascribed to the location of the sterol ring of cholesterol. The ratio of the two constituents
in the liquid-ordered phase is not constant because the stoichiometry is temperature-dependent as seen by respective changes in bilayer thickness over
the range 20° to 36 °C where coexisting phases are observed. Three coexisting phases were deconvolved in the ternary mixture at 20 °C. From an
analysis of the ternary mixtures containing mole fractions of cholesterol from 0.09 to 0.15 it was found that the liquid-crystal and gel phases each
contained about 10% of the cholesterol molecules and the liquid-ordered phase was comprised of 30% cholesterol molecules.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Liquid-ordered phase; Bilayer electron density; Phosphatidylcholine; Cholesterol partition coefficient; Membrane raft
1. Introduction
Over the past decade there has been growing interest in lipid
rafts since formulation of the raft model was proposed by
Simons and Ikonen in 1997 and possible biological functions
ascribed to these structures[1–4]. Although the detailed
structure and properties of rafts are still not completely clear,
it is generally agreed that tight acyl chain packing is a key
feature of raft lipid organization. The structure is designated as
liquid-ordered phase (Lo), which is formed by the interaction
between saturated-chain molecular species of sphingolipids and
phosphatidycholines with cholesterol [5,6]. The formation of
such liquid-ordered phases results in phase separation and
⁎ Corresponding authors. Z. Yu is to be contacted at tel.: +86 10 6279 2492;
fax: +86 10 6277 1149. P.J. Quinn, tel.: +44 20 78484408; fax: +44 20
78484500.
E-mail addresses: [email protected] (Z. Yu), [email protected]
(P.J. Quinn).
0005-2736/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.bbamem.2007.07.023
coexistence with liquid-crystal phase over a wide temperature
range [7]. Quantitative analysis has suggested that the forces that
govern the creation of liquid-ordered phase arise through the
formation of stoichiometric complexes between cholesterol and
the phospholipids [8–10]. For cholesterol-containing lipid
mixtures, knowledge of the composition of the coexisting
phases is necessary for a more detailed understanding of the
properties and functions of “membrane rafts” [11]. However,
determination of the composition of coexisting phases has
proved to be a considerable challenge [12]. Although some
studies have been undertaken using fluorescence probe
techniques and electron paramagnetic resonance spectroscopy,
a quantitative analysis of the composition of coexisting domains
has yet to be fully resolved.
To understand the physical properties of biological membranes containing cholesterol-induced liquid-ordered phases,
model systems of monolayer or giant unilamellar vesicles of
binary and ternary mixtures of phospholipids and cholesterol
have been investigated by various methods [13–15]. The liquid-
2874
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
ordered phase is conceptualized as a bilayer structure with
properties intermediate between liquid-crystal (Lα) and gel
phase (Lβ) [16]. Molecules in this phase are extended and tightly
packed, as in the gel phase, with the hydrocarbon chains being
predominantly in an all-trans conformation [17], but have a high
degree of lateral mobility within the bilayer. It was shown that
the lateral diffusion coefficient was typically 2–3 times slower in
the liquid-ordered phase than in the Lα phase but was notably
faster than in the Lβ phase [18]. This unique feature has been
explained in terms of an umbrella model [19]. There have been a
number of structural studies of model bilayer lipid systems that
are known to form Lo phase in order to define the relative
location of the component lipids and the dimensions of the
phase. Electron density calculations have been made from X-ray
scattering intensity data to show that the bilayer thickness of Lo
phase is intermediate between gel and fluid phase of the pure
phospholipid component [20]. Furthermore, the electron density
of the methylene chain region of sphingomyelin [21] and
phosphatidylcholine [22,23] bilayers is significantly greater in
the presence of cholesterol. The role of liquid-ordered phases of
phospholipids and cholesterol in cellular functions has been
reviewed by Brown and London [24].
Model membranes with three or more lipids have been used
as a more reliable model of biological membranes compared
with binary mixtures. Eukaryotic plasma membranes typically
are comprised of about 50 mol% sphingolipids and phospholipids and up to 40 mol% cholesterol. Multibilayer structures
composed of dipalmitoylphosphatidylcholine (DPPC), dioleoyl-phosphatidycholine (DOPC), and cholesterol are thought
to form micro-domains that have similar characteristics to those
of rafts isolated from cell plasma membrane [12]. In this study, a
binary system consisting of DPPC/cholesterol was used to
characterize the structure of Lo phase and the ternary system
consisting of 1:1:0.5 DPPC/DOPC/cholesterol was selected to
characterize the structural features of the liquid-ordered (Lo)
phase in fluid bilayer structures. Electron density profiles were
constructed to distinguish salient features of the bilayer such as
thickness and the dimensions of the intervening water layer. The
partition ratios of cholesterol in coexisting phases have also
been derived from the analysis of small-angle X-ray diffraction
intensity peaks. Precise values of the widths of liquid-ordered
and liquid-disordered phases and partition ratio of cholesterol
should assist in understanding mechanisms of membrane
component sorting by raft domains [25,26].
2. Materials and methods
2.1. Materials
1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), and cholesterol were purchased from Sigma
Chemicals (St. Louis, MO, USA). They were used without further purification.
The phospholipids and cholesterol dissolved in chloroform were mixed in the
desired proportions, dried under a stream of oxygen-free dry nitrogen and any
remaining traces of solvent removed by storage overnight in vacuo. The dried
lipids were fully hydrated with Tris–HCl buffer (50 mM Tris–HCl, 150 mM
NaCl, 0.1 mM CaCl2, pH = 7.2) with the ratio of lipid:buffer, 1:4, by weight. The
dispersions were vortex mixed during the freeze–thaw cycles between 20 and
70 °C of which there were more than 10 cycles to ensure homogeneous
dispersion. Mixtures of DPPC/cholesterol containing 0%, 17%, 33%, and 50%
molar percentage of sterol were examined. In addition, samples consisting of
equimolar proportions of DPPC and DOPC containing 0, 9, 10, 11, 12, 13, 14,
15 and 20 molar percentage of cholesterol were also examined. The uncertainties
of cholesterol mole fraction are 0.3%. The thermotropic phase behaviour
observed in mixtures showed characteristics consistent with full hydration e.g.
the temperature of transitions from gel to fluid phases.
2.2. X-ray diffraction
Real-time synchrotron X-ray diffraction experiments for the binary samples
of DPPC and cholesterol were performed at Station BL40B2 of Spring-8, Japan.
The small-angle X-ray scattering (SAXS)/wide-angle X-ray scattering (WAXS)
data were recorded on-line with image plates. The wavelength was 0.100 nm,
beam dimensions of 0.2 mm vertical × 0.25 mm horizontal and with a sample to
detector distance of 400 mm. A sample of silver behenate was used for spatial
calibration. A Linkam thermal stage (Linkam, UK) was used to perform
temperature scans at a rate of 1°/min. The overall time for collection of each
image including exposure, data processing and dumping was 330 s. The X-ray
diffraction data of samples containing equimolar proportions of DPPC and
DOPC were recorded on Beamline 16.1 of the Daresbury Synchrotron Radiation
Source (Warrington, UK). The lipid dispersions were sandwiched in a copper cell
of 1 mm thickness between two mica windows which was mounted on the
thermal stage. The wavelength was 0.141 nm, beam dimensions 0.2 mm
vertical × 2.5 mm horizontal and with a sample to SAXS detector distance of
1.50 m. Data analysis was performed using either the Fit2D for image plates or
the OTOKO program [27] for detection by quadrant (SAXS) and INEL (WAXS)
delay-line electronic detectors. Lorentz geometric corrections were applied to the
data collected at SPring-8 and Daresbury, this is considered in the calculation of
structure amplitude. Polarization corrections can be reasonably neglected
because the polarization factors are almost 1 in the range of small angle
diffraction region. The diffraction intensities were evaluated by the integral of
corresponding scattering intensity peaks. The analyses were performed using
PeakFit (version 4, SPSS Inc) software to fit peaks to a Gaussian + Lorentzian
distribution. The best fits were obtained using an iterative process to maximize
the R2 values and the uniqueness of the fits was checked by comparing the centres
of all the deconvolved peaks for the individual diffraction orders; in all cases the
sets of peaks conformed closely to diffraction orders of individual bilayer
structures. In multicomponent mixtures the curve fitting procedure began by
assuming a priori that the diffraction peaks consisted on a single component and
the R2 values were compared with fits of two or more components. By
performing this analysis on all four orders of the lamellar repeats it could be
demonstrated that where phase coexistence is observed the peak fit analysis was
reliable.
2.3. Electron density calculation
To obtain estimates of bilayer thickness the distance between peaks of highest
relative electron density, said to represent the phosphate residues on either side of
the phospholipid bilayer (dpp), and the thickness of the intervening water layer
(dw) between successive bilayers, the electron density profiles of the lamellar
structures were calculated. The dimensions of the water layer (dw) are obtained
by subtracting dpp from the lamellar d-spacing. It is recognized that this does not
represent the true volume of water in the unit cell because part of the lipid headgroup extends into this region [28]. Nevertheless, it is reasonable to assume that
this lipid component of the electron density profile is similar for the different lipid
mixtures so that differences in dw should reflect differences in hydration of the
bilayers.
The relative electron densities were derived from the following equation
[29,30]:
qð xÞ ¼
X
2
d gðhÞd F ðhÞ d cosð 2pxh=d Þ
d
j
j
ð1Þ
where x is the distance from the center of the bilayer, d the lamellar repeat
period, g(h) the phase for the hth order, and the summation is over diffraction
orders h, F(h) is the structure amplitude. For the unoriented X-ray diffraction pattern
the absolute value of the structure amplitude was set equal to {h2I(h)}1/2 [29].
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
Fig. 1. Static small-angle (left) and wide-angle (right) X-ray scattering intensity
patterns recorded at 30 °C from cholesterol/DPPC binary mixtures as a function
of reciprocal spacing (S). The cholesterol molar ratios are indicated in the figure.
The insert shows deconvolution of the second-order reflection from the mixture
containing 17 mol% cholesterol into two peaks. The dotted line through the
WAXS profiles indicates the change in spacing of the chain packing with
increasing proportions of cholesterol.
Four peaks of X-ray scattering intensity representing the first four-orders of a
lamellar phase of each sample were used to calculate the Fourier reconstruction of
electron intensity. The phase g(h) can only be 1 or − 1 by assuming a
centrosymmetric structure in the multibilayer liposome structure. Electron
density calculations were performed for all combinations of phase angles but the
only combination fitting a centrosymmetric diffraction profile expected of a
multibilayer structure was −, −, +, − for all the diffraction profiles examined. This
phase combination has been reported consistently for the DPPC in the gel phase.
Also, the results of the electron density profile of DPPC in gel phase (at 20 °C) are
in agreement with the published data [31]: the peak-to-peak distance between the
electron dense region on either side of the bilayer (dpp, representing bilayer
thickness) is 4.11 nm and water thickness represented by the distance between
peaks of electron density in adjacent bilayers (dw) is 2.20 nm. The definition of
dw overestimates the space exclusively occupied by water because the lipid head
groups extend 0.4–0.5 nm into this region so the true space is of the order of 0.8–
1 nm less than the calculated value.
2875
characteristic of a lamellar-gel (Lβ′) indexed by four orders of
SAXS diffraction with a d-spacing of 6.43 nm and a sharp wideangle diffraction at a spacing of 0.42 nm, in agreement with
previous results [35]. The lamellar d-spacing is increased in the
presence of 17 mol% cholesterol but decreases with higher
proportions of cholesterol up to 50 mol% the highest
concentration examined. The effect of cholesterol on the
WAXS peak is consistent with a reorientation of the
hydrocarbon chains of the Lβ′ configuration and a progressive
disordering of the chains which is proportional to the molar ratio
of cholesterol in the mixture. When the molar ratio of
cholesterol in DPPC is 17%, two lamellar phases coexist at
30 °C [34,36]. The two phases have been assigned to gel- and
liquid-ordered phases, respectively. This can be seen as an
asymmetry in the small-angle X-ray diffraction peaks and
illustrated for the second-order reflection in the insert of Fig. 1.
The X-ray scattering intensity per se does not provide
information about the size of domains within these mixed
dispersions. Where scattering peaks can be deconvolved into
different components it may be inferred that there is coexistence
of more than one structural repeat. From the higher order
reflections we can say that these repeat structures are consistent
with bilayers in the mixtures that have been examined in this
study.
The effect of cholesterol on the bilayer structure of DPPC
observed at 30 °C was examined by calculation of electron
density distributions through the unit cell and the results are
presented in Fig. 2. The peak-to-peak distance (dpp) and water
layer thickness (dw) of pure DPPC bilayers are 4.24 nm and
2.4. Differential scanning calorimetry (DSC)
Samples of lipid dispersions (20 μl) for the DSC experiments prepared in the
same way as for X-ray studies were examined in a Mettler–Toledo DSC821e
differential scanning calorimeter using a temperature scan rate of 1°/min.
Measurements were repeated at least three times to confirm reproducibility.
3. Results and discussion
3.1. Electron density profiles of DPPC/cholesterol mixtures
The effect of cholesterol on the structure and thermotropic
phase properties of hydrated DPPC bilayers has been
investigated by a number of biophysical methods [32]. Phase
diagrams of the binary lipid system have been constructed from
such studies [33,34]. In order to get the electron density profiles
of liquid-ordered phase formed in DPPC and cholesterol
mixtures, small-angle (SAXS) and wide-angle (WAXS) scattering intensity patterns were recorded by synchrotron X-ray
diffraction methods. Fig. 1 shows the representative X-ray
diffraction patterns of pure DPPC and DPPC–cholesterol
mixtures at 30 °C. The diffraction pattern of pure DPPC is
Fig. 2. Molecular models representing dimensions of bilayer thickness (dpp) and
thickness of the water layer (dw) in multilamellar dispersions of DPPC ( ) and
cholesterol ( ) derived from electron density calculations. The mol% of
cholesterol is indicated on the figure.
2876
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
It is argued that the detergent-resistant membrane fractions
represent structures in liquid-ordered phase which are enriched
in DPPC/cholesterol or SM/cholesterol [43]. It is, however,
possible that bilayer thickness may be different for DPPC/
cholesterol compared with SM/cholesterol mixtures although
both form liquid-ordered phase.
3.2. The effect of cholesterol on bilayers of DPPC:DOPC
Fig. 3. SAXS (left) and WAXS (right) data recorded from aqueous dispersions at
30 °C of 1:1, DPPC:DOPC (lower curves, 0% cholesterol) and 1:1:0.5, DPPC:
DOPC:cholesterol (upper curves, 20 mol% cholesterol). The SAXS data are
plotted as X-ray scattering intensity and, to emphasize the peaks from higherorder reflections the intensities are magnified 10 fold.
2.19 nm, respectively, and are within the range of previously
published values [37–39]. Using the deconvolved diffraction
peaks [40] of the mixture containing 17 mol% cholesterol,
electron density profiles of the two phases can be calculated
separately. The dpp and dw of the liquid-ordered phase, shown
in Fig. 2b, are 4.82 nm and 1.88 nm, respectively. The increase
in dpp compared to that of pure DPPC could result from a
decrease in the angle of tilt of the acyl chains with respect to the
bilayer normal. If the chains of DPPC in the liquid-ordered
phase are not tilted, the tilt angle of pure DPPC in the gel phase
can be estimated as (180° × {cos− 1(4.24/4.82)}/π = 28.4°). This
value is in close agreement with values of chain tilt of DPPC in
excess water at 30 °C [39]. This means that the incorporation of
cholesterol into DPPC gel phase leads to the reorientation of the
hydrocarbon chains of DPPC so they become fully extended in
a direction perpendicular to the bilayer normal in the liquidordered phase [41]. It was also noted that the bilayer thickness
of the coexisting gel phase is greater than in pure DPPC bilayers
inferring that chain tilt is reduced by the presence of domains of
liquid-ordered phase.
Increasing proportions of cholesterol in DPPC bilayers up to
25 mol% reveal the existence of only one liquid-ordered phase
[7,34]. The d-spacing of the bilayer repeat increases in a
cholesterol-dependent manner in this range of cholesterol
content. When the cholesterol molar ratio reaches to 33% and
50%, the d-spacings of the liquid-ordered phase are 6.53 nm and
6.58 nm, and the wide-angle X-ray diffraction peaks are centered
at 0.465 and 0.467 nm, respectively. Thus there is a cholesterol
dependence of both SAXS and WAXS spacings in this range of
cholesterol concentrations (see Fig. 1). The reduction in dspacing in mixtures containing 33 mol% and above is due to a
decreasing bilayer thickness rather than dw indicating that the
hydrocarbon chains of the phospholipid are not fully extended
and are more disordered, consistent with the increase in the
spacing of the WAXS peak. It is noteworthy that the peak-topeak distances of sphingomyelin (SM):cholesterol, 2:1 mol%
mixture and detergent-resistant membrane fractions insoluble in
1% or 4% Triton X-100 from such mixtures were all 4.8 nm [42].
With characterization of the structural parameters of Lo
phase formed by DPPC/cholesterol we next examined the
conditions under which Lo phase forms in the presence of a
phospholipid that forms fluid bilayers at temperatures above
0 °C. The mixture chosen for these experiments consisted of an
equimolar mixture of DPPC/DOPC. An X-ray diffraction
pattern recorded from fully hydrated dispersion of this mixture
(0% mol cholesterol) at a temperature of 30 °C is shown in
Fig. 3. It can be seen that the scattering pattern can be fitted on
the basis of at least four-orders of reflection to two lamellar
phases. The lamellar d-spacings of the phases are 5.25 and
6.11 nm, respectively. These d-spacings are significantly less
than reported for dispersions of pure DOPC ( 6.32 nm) [44] and
DPPC (6.40 nm) [39], respectively. The result suggests that the
arrangement of the phospholipids in the mixed lipid bilayer
leads to either a reduction in bilayer thickness, a decrease in
hydration of the multilamellar dispersion or a combination of
both factors. The reflection in the WAXS region is relatively
broad but with a spacing centred at 0.415 nm it may be assumed
that at least one of the two phases is a gel phase. The presence of
a broad scattering band centred at 0.46 nm expected for
disordered hydrocarbon chains could not be resolved from the
WAXS diffraction pattern. Nevertheless, the two coexisting
phases can be reasonably assigned as liquid-crystal (Lα) and gel
phase (Lβ) [45], similar to the phase separation behavior of 1:1
DOPC:SM mixtures [42].
Fig. 4. Relative electron density profiles of lipid mixtures containing 1:1, DPPC:
DOPC and 1:1:0.5, DPPC:DOPC:cholesterol calculated from the data shown in
Fig. 3. The assignments of the structure of each phase of the respective mixtures
are indicated. Arrows indicate higher electron density in the hydrophobic region
of bilayers containing cholesterol assigned as Lo phase.
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
2877
Fig. 6. Relative electron density profiles of Lα (A) and Lo phase (B) calculated
from data obtained from a ternary mixture 1:1:0.5 DPPC:DOPC:cholesterol at
20°, 28°, 36°, and 58 °C.
larger than observed in the binary phospholipid and is
consistent with a more disordered in-plane packing of the
molecules within the bilayer. On this evidence and our
understanding of the mixture in the absence of cholesterol,
the two lamellar phases can be tentatively assigned as an Lα
phase and a liquid-ordered phase Lo. Such an assignment
agrees with the results of fluorescence microscopy which has
demonstrated that in such mixtures two immiscible liquid-like
phases are formed [46–48].
To further characterise the differences between the two
coexisting phases in the two dispersions, electron density
Fig. 5. The temperature-dependence of d-spacing (A) and X-ray scattering
intensity (B) of two-coexisting phases deconvolved from the second-order
small-angle diffraction peaks. ○, Lβ phase and ●, Lα phase in 1:1 DPPC/DOPC
mixture; □, Lo phase and , Lα phase in 1:1:0.5 DPPC/DOPC/cholesterol
mixture (C) DSC heating curve of 1:1 DPPC/DOPC mixture showing
relationship to the X-ray scattering intensity of phases assigned Lα and Lβ.
▪
An X-ray scattering intensity pattern recorded from a
ternary mixture consisting of 1:1:0.5 mole ratio DPPC:DOPC:
cholesterol (20 mol% cholesterol) shows a similar set of
reflections as observed in the binary phospholipid mixture in
the SAXS region and the structure can be assigned to two
lamellar phases with d-spacings of 5.45 and 6.10 nm,
respectively. The scattering pattern in the WAXS region,
however, is different and consists of a broad scattering band
centered at a spacing of 0.428 nm. This value is slightly
Table 1
The d-spacings and peak to peak distances (dPP) in 1:1:0.5 DPPC:DOPC:
cholesterol and 1:1 DPPC:DOPC at different temperatures as shown in Figs. 4,
6A and B
1:1:0.5 DPPC:DOPC:
cholesterol
d-spacing/
nm
dpp/nm
T/°C
Lo
20
28
30
36
58
6.21 ± 0.02
6.14 ± 0.02
6.10 ± 0.02
5.90 ± 0.02
1:1 DPPC:DOPC
Lβ
30
6.11 ± 0.03 4.52 ± 0.01 5.25 ± 0.03 4.04 ± 0.01
d-spacing/ dpp/nm
nm
Lα
4.60 ± 0.01
4.54 ± 0.01
4.52 ± 0.01
4.48 ± 0.01
5.44 ± 0.02
5.44 ± 0.02
5.45 ± 0.02
5.52 ± 0.02
5.31 ± 0.02
4.04 ± 0.01
4.04 ± 0.01
4.04 ± 0.01
4.04 ± 0.01
4.04 ± 0.01
Lα
2878
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
3.3. The temperature dependence of Lo electron density profiles
Fig. 7. Deconvolved small-angle X-ray diffraction intensity peak obtained from
an aqueous dispersion of an equimolar mixture of DPPC/DOPC containing
12 mol% cholesterol recorded at 20 °C. Three peaks (B, C and A) can be
deconvoluted centred at 7.1 nm, 6.9 nm and 6.5 nm, respectively.
profiles were constructed and the results are shown in Fig. 4.
For each profile the center of the bilayer is located at the
origin, the low electron density trough in the center of the
profile corresponds to the terminal methyl groups at the ends
of the hydrocarbon chains, the medium density regions on
either side of this trough correspond to the methylene chain
regions of the bilayer, and the peaks of relatively high electron
density near the edge of the profile correspond to the lipid
head groups. This shows that the two Lα phases assigned in
each of the mixtures have almost identical electron density
profiles with peak-to-peak distances of 4.04 nm but a slightly
thicker water layer when cholesterol is present. It has been
reported that this Lα phase is enriched in the unsaturated
phospholipid and the molecules have the same diffusion
coefficient as those in pure DOPC bilayers [47]. The electron
density profiles of the other two phases are also similar. Both
phases have the same headgroup peak-to-peak distance of
4.52 nm and almost the same thickness of water layer of about
1.59 nm. The major difference between these two profiles is in
the methylene chain region of the bilayer where there is
relatively greater density in Lo phase than in the corresponding
region of Lβ, indicated by arrows in Fig. 4. This characteristic
feature can be explained by the presence of cholesterol in Lo
phase because the electron density of the sterol rings of
cholesterol is greater than that of the phospholipid methylene
chains [41]. However, in all other respects the electron density
profile of Lo is similar to that of Lβ.
The phase separation behavior of 1:1:0.5 mole ratio mixtures
of DPPC:DOPC:cholesterol has also been visualized by
confocal fluorescence microscopy [47]. One domain, assigned
as liquid-disordered phase (Lα), was typified by a relatively high
diffusion coefficient and enriched in DOPC. The other, from
which the probe DiI-C18 was excluded, had a lower diffusion
coefficient and was said to be enriched in cholesterol and
saturated DPPC. This resembles the properties of liquid-ordered
phase [46,48] and detergent-resistant fractions prepared from
biological membranes [24,49] and is thus assigned as a liquid
ordered (Lo) phase.
The temperature dependence of phase separation behaviour
observed in the binary phospholipid mixture and the effect of
cholesterol on these phase properties was examined over the
temperature range 20°–60 °C. Lamellar repeat spacings and
intensities, represented as peak areas, of second-order X-ray
scattering peaks of the 1:1 mole ratio mixture of DPPC:
DOPC in the absence and presence of cholesterol determined
during a heating scan from 20 °C to 60 °C are shown in Fig.
5A and B, respectively. It can be seen that the d-spacing of
the gel phase in the DPPC:DOPC mixture remains almost
constant with a value of 6.10 nm (open circle in Fig. 5A)
until it disappears at about 41 °C. This temperature coincides
with the gel to fluid phase transition temperature of pure
DPPC, suggesting the Lβ phase consists predominantly of
DPPC.
As noted above the X-ray scattering intensity data is not
informative about the disposition of components of coexisting
phases but the thermotropic behaviour of the mixtures can
provide an indication of the extent of phase separation. The loss
of the lamellar repeat assigned to DPPC at the phase transition
temperature of the pure phospholipid indicates that the saturated
phospholipid is largely phase separated from the liquid phase.
Furthermore, it is also clear that the dispersion is fully hydrated.
In contrast to the peak assigned to Lβ phase, the d-spacing of
Lα phase decreases at a temperature greater than about 45 °C
from 5.27 nm to 5.10 nm at 60 °C (solid circle in Fig. 5A).
Electron density profiles calculated at temperature intervals
during the heating scan (data not shown) confirmed that the
bilayer thicknesses were constant at about 4.52 nm for the Lβ
phase and 4.04 nm for the Lα phase throughout the temperature
scan.
The thermotropic behaviour of the Lβ phase contrasts with
that of Lo phase formed when cholesterol is present in the
phospholipid mixture. In Lo phase the d-spacing decreases
upon heating from 6.21 nm at 20 °C to 5.90 nm at 38 °C
whereupon the phase can no longer be detected (open square in
Table 2
Deconvolution results of SAXS peaks and the solutions obtained by the Excel
Solver
x
SB
SC
SA
m = SB / SA
n = SC / SA
y
0.09
0.10
0.11
0.12
0.13
0.14
0.15
0.2
17.7
186.0
382.0
80.9
265.2
329.8
0.4
69.4
214.6
223.8
245.9
800.6
1125.9
0.1
17.8
66.2
187.1
225.2
167.8
437.6
2.2
1.0
2.8
2.0
0.4
1.6
0.8
4.8
3.9
3.2
1.2
1.1
4.8
2.6
− 0.06
− 0.03
− 0.01
0.02
0.03
0.01
0.04
xc,β
xc,o
xc,α
p
q
∑y2
0.10
0.29
0.09
0.08
0.06
0.0083
Note. x, overall mole ratio of cholesterol in sample; SA, SB and SC are the
intergrated areas of the three phases; xc,α, xc,β, and xc,o are molar ratios of
cholesterol in liquid crystal, gel and liquid ordered phase; p, q are relative to
partition coefficients.
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
Fig. 5A). The transition temperature is in agreement with
published data for Lo phase formed by DPPC/cholesterol [48].
It is reasonable to believe that the X-ray scattering intensity,
represented as peak area, approximates to the amount of the
phase present in the mixture. Plots of the peak intensities as a
function of temperature can therefore be used to assess the
transition from one phase to another. These data are shown in
Fig. 5B. They show that with increasing temperature the
proportion of Lo and Lβ phases in the DPPC:DOPC mixtures
with and without cholesterol, respectively, decreases and there
is a corresponding increase in the proportion of Lα phase. The
apparent stability of the Lβ phase is greater than the Lo phase as
judged by the relatively sharp decrease in the proportion of Lo
phase with increasing temperature. Differential scanning
calorimetric results obtained from these dispersions show that
the phase transition enthalpy from gel to liquid-crystal phase in
1:1 DPPC:DOPC is 30 kJ per mole DPPC (Fig. 5C). The
scattering intensity of peaks assigned to Lβ and Lα phases
shown in the figure indicate that the change in specific heat
capacity coincides with the loss of Lβ phase and increase of Lα
phase. The temperature range of the transition is broader than
recorded from a dispersion of pure DPPC and the transition
enthalpy is less (36.6 kJ mol− 1 for pure DPPC). The transition
enthalpy recorded in the DOPC:DPPC mixture is about twofold greater than that observed in the liquid-ordered to liquidcrystal phase transition in 1:1:0.5 DPPC:DOPC:cholesterol
which was 16.7 kJ mol− 1.
3.4. The relative electron density profiles of Lo and Lα phase
formed in 1:1:0.5 mixture
DPPC:DOPC:cholesterol at designated temperatures show
clearly that the bilayer thickness of Lα as judged by dpp distance
is independent of temperature over the range 20° to 58 °C
(Fig. 6A). All the dpp values were constant at 4.04 nm. Only
small changes in bilayer thickness as a function of temperature
have been observed in both small-angle neutron [50] and X-ray
[44] scattering studies of hydrated phospholipids. Nevertheless,
small thickness changes would be difficult to detect in the
present study where the resolution is of the order of 1.5 nm.
Fig. 6B shows the electron density profiles of Lo phase at
different temperatures before it is converted to Lα phase. Over
the temperature range 20° to 36 °C the peak-to-peak distance
and water layer thickness both decrease with increasing
temperature, however, the water layer thickness decreases by
a greater amount than the bilayer width, 0.2 nm vs. 0.12 nm.
Table 1 presents a summary of d-spacings and peak to peak
distances (dpp) in 1:1:0.5 DPPC:DOPC:cholesterol and 1:1
DPPC:DOPC at designated temperatures as shown in Figs.
4, 6A and B. The temperature dependency of liquid-ordered
phase has been earlier reported on the basis of solid-state NMR
and X-ray diffraction studies [51]. The change in bilayer
thickness may result from a change in proportion of
phospholipid and cholesterol components constituting the Lo
phase and may reflect different partitioning preferences for
various membrane lipids and proteins into liquid-ordered
domains [52,53].
2879
3.5. The partition ratios of cholesterol in different domains
A series of ternary mixtures consisting of equimolar proportions of DPPC/DOPC containing a range of cholesterol concentrations between 9 and 15 mol% were examined to determine
the partition coefficient of cholesterol in the different phases
that form in the system. Three coexisting phases were detected
by X-ray diffraction methods in these mixtures at 20 °C. To
estimate the stoichiometry of the three phases, deconvolution of
the small-angle X-ray diffraction peaks was undertaken using
PeakFit software and fitting the peaks based on a Gaussian +
Lorenzian model. An example of the results of such a
deconvolution is presented in Fig. 7, which shows that the
second-order lamellar diffraction from a DPPC-DOPC dispersion containing 12 mol% cholesterol at 20 °C, can be deconvolved into three peaks. These peaks (peaks B, C and A)
represent three coexisting phases, which are designated gel
phase (Lβ), liquid-ordered phase (Lo) and liquid-crystal phase
(Lα), respectively.
According to the Gibbs’ phase rule, the degrees of freedom
for this system under a given pressure and temperature is 0.
This means that the stoichiometric composition of three
phases is fixed and independent of cholesterol concentration.
As the d-spacings of the diffraction peaks are similar over
the range of cholesterol concentrations examined it is reasonable to assume that the integrated areas of peaks are
proportional to the amount of respective phase with coefficients f1, f2 and f3. The areas for the deconvoluted peaks in
the mixtures examined after baseline correction are summarized in Table 2.
The total number of molecules in liquid-crystal (Lα), gel (Lβ)
and liquid-ordered phase (Lo) can be ascribed as n1 = SA⁎f1,
n2 = SB⁎f2 and n3 = SC⁎f3, respectively. Where SA, SB and SC are
the intergrated areas of the three phases, f1, f2 and f3 are the
coefficients. We can further assume that xc,α, xc,β, and xc,o are
molar ratios of cholesterol in each phase. The overall mole ratio
of cholesterol (x) in the sample can be expressed by the following equation [54]:
x¼
¼
n1 xc;a þ n2 xc;b þ n3 xc;o
n1 þ n2 þ n3
ð2Þ
SA f1 xc;a þ SB f2 xc;b þ SC f3 xc;o
SA f1 þ SB f2 þ SC f3
By further defining SB/SA = m, SC/SA = n, f2/f1 = p, f3/f1 = q
Eq. (2) can be rearranged into the following form:
x ¼ md pd ðxc;b xÞ þ nd qd ðxc;o xÞ þ xc;a
ð3Þ
p, q, xc,α, xc,β, xc,o are unknown parameters in the equation, but
they are all constants under the experimental conditions
employed.
Although the unknown parameters can be solved by five
equations with different x values, it is likely that errors
associated with sample preparation and determination of
integrated peak areas may have a significant effect on the
2880
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
solutions obtained. The SOLVER add-in non-linear program in
Microsoft Excel was used to optimize the solutions [55]. In
order to get the relevant values, we let
x md pd ðxc;b xÞ nd qd ðxc;o xÞ xc;a ¼ y
ð4Þ
The solutions should make this equation close to 0, which
means y almost equals 0. This Eq. (4), therefore, is used as the
objective function in the SOLVER program. The seven serial
values of x, m and n, shown in Table 2, are set as analysis
variables, and the unknown parameters p, q, xc,α, xc,β, xc,o are
the design variables. However, the starting values that we have
chosen in the program are based on primitive solutions obtained
from five equations. By setting the sum of the squares of each
implicit equation’s value y as 0 and no constriants, we obtain
the optimized solutions for the equations after running the
SOLVER program. The solutions are also presented in the
Table 2.
The obtained values of xc,α, xc,β, xc,o using this approach are
0.09, 0.10 and 0.29 , respectively. This suggests that the
cholesterol molecules are almost equally distributed in liquidcrystal and gel-phase at 20 °C. On the contrary, approximately
30% of the molecules in liquid-ordered phase are cholesterol.
The phase diagram of DSPC–DOPC–cholesterol shows similar
results for the partition ratio of cholesterol in liquid-crystal, gel
and liquid-ordered phases when the three phases coexist,
namely, 0.09, 0.15, and 0.25, respectively [56]. Compared to
binary DPPC–cholesterol mixtures, the ratio of cholesterol in
the coexisting gel, liquid-ordered and liquid-crystal phase is
0.07, 0.09 and 0.21, respectively, which can be obtained from
the phase diagram [34].
The liquid-ordered phase formed in mixtures of phosphatidycholines and cholesterol has been characterized by electron
density profiles in the present study. The bilayer thickness
deduced from the peak-to-peak distance in the glycerophospholipid mixture is about 4.50 nm; this is slightly smaller
than Lo phase formed in mixed aqueous dispersions of SM and
cholesterol for which values of 4.80 nm have been reported
[42]. In addition, two immiscible liquid phase separations can
be distinguished in the ternary DPPC/DOPC/cholesterol
system. One is liquid-ordered phase enriched in DPPC and
cholesterol but without a fixed stoichiometry of the components. The major difference between this liquid-ordered and
Lβ phases is the temperature-dependence of the d-spacing and
the packing of the hydrocarbon chains are in a more disordered
configuration than in Lβ phase. A higher electron density in
the methylene chain region of the bilayer is ascribed to the
location of the sterol ring system of cholesterol. The other
phase is enriched in DOPC and has an almost identical
electron density profile as that of pure DOPC. Based on the
partition ratio of cholesterol in the coexisting gel, liquid-crystal
and liquid-ordered phases calculated from the small-angle
X-ray diffraction data it can be concluded that cholesterol is
almost equally distributed in gel and liquid-crystal phases. The
cholesterol molecules represent only 10% of the total
molecules in these two phases, but they comprise 30% of
the molecules in liquid-ordered phase.
Acknowledgements
This work was supported by grants from the Natural Science
Foundation of China (NSFC: 20633080, 20505012) and a grant
from the Human Frontier Science Programme. Assistance of
Drs. Inoue and Takahashi at SPring-8 (Japan) and Dr. Anthony
Gleeson at the Daresbury Laboratory (UK) in collection of X-ray
diffraction data is gratefully acknowledged.
References
[1] K. Simons, E. Ikonen, Functional rafts in cell membranes, Nature 387 (1997)
569–572.
[2] K. Simons, D. Toomre, Lipid rafts and signal transduction, Nat. Rev., Mol.
Cell Biol. 1 (2000) 31–39.
[3] M. Edidin, The state of lipid rafts: from model membranes to cells, Annu.
Rev. Biophys. Biomol. Struct. 32 (2003) 257–283.
[4] T.P.W. McMullen, R.N.A.H. Lewis, R.N. McElhaney, Cholesterol–
phospholipid interactions, the liquid-ordered phase and lipid rafts in
model and biological membranes, Curr. Opin. Colloid Interface Sci.
8 (2004) 459–468.
[5] X. Xu, E. London, The effect of sterol structure on membrane lipid
domains reveals how cholesterol can induce lipid domain formation,
Biochemistry 39 (2000) 843–849.
[6] K. Simons, E. Ikonen, How cells handle cholesterol, Science 290 (2000)
1721–1726.
[7] J. Zhang, H. Cao, B. Jing, P.F. Almeida, S.L. Regen, Cholesterol–phospholipid
association in fluid bilayers: a thermodynamic analysis from nearest-neighbor
recognition measurements, Biophys. J. 91 (2006) 1402–1406.
[8] S.A. Pandit, D. Bostick, M.L. Berkowitz, Complexation of phosphatidylcholine lipids with cholesterol, Biophys. J. 86 (2004) 1345–1356.
[9] H.M. McConnell, A. Radhakrishnan, Condensed complexes of cholesterol
and phospholipids, Biochim. Biophys. Acta 1610 (2003) 159–173.
[10] T.G. Anderson, H.M. McConnell, Condensed complexes and the calorimetry
of cholesterol–phospholipid bilayers, Biophys. J. 81 (2001) 2774–2785.
[11] Y.W. Chiang, J. Zhao, J. Wu, Y. Shimoyama, J.H. Freed, G.W. Feigenson,
New method for determining tie-lines in coexisting membrane phases
using spin-label ESR, Biochim. Biophys. Acta 1668 (2005) 99–105.
[12] S.L. Veatch, I.V. Polozov, K. Gawrisch, S.L. Keller, Liquid domains in vesicles
investigated by NMR and fluorescence microscopy, Biophys. J. 86 (2004)
2910–2922.
[13] W.H. Pitcher III, S.L. Keller, W.H. Huestis, Interaction of nominally
soluble proteins with phospholipid monolayers at the air–water interface,
Biochim. Biophys. Acta 1564 (2002) 107–113.
[14] S.L. Veatch, S.L. Keller, Organization in lipid membranes containing
cholesterol, Phys. Rev. Lett. 89 (2002) 268101.
[15] W.H. Binder, V. Barragan, F.M. Menger, Domains and rafts in lipid
membranes, Angew. Chem., Int. Ed. Engl. 42 (2003) 5802–5827.
[16] M.D. McMullen, H. Kross, M.E. Snook, M. Cortes-Cruz, K.E. Houchins, T.A.
Musket, E.H. Coe Jr., Salmon silk genes contribute to the elucidation of the
flavone pathway in maize (Zea mays L.), J. Heredity 95 (2004) 225–233.
[17] M.B. Sankaram, T.E. Thompson, Modulation of phospholipid acyl chain
order by cholesterol. A solid-state 2H nuclear magnetic resonance study,
Biochemistry 29 (1990) 10676–10684.
[18] G. Oradd, G. Lindblom, P.W. Westerman, Lateral diffusion of cholesterol
and dimyristoylphosphatidylcholine in a lipid bilayer measured by pulsed
field gradient NMR spectroscopy, Biophys. J. 83 (2002) 2702–2704.
[19] J. Huang, G.W. Feigenson, A microscopic interaction model of maximum
solubility of cholesterol in lipid bilayers, Biophys. J. 76 (1999) 2142–2157.
[20] P.R. Maulik, G.G. Shipley, Interactions of N-stearoyl sphingomyelin with
cholesterol and dipalmitoylphos phatidylcholine in bilayer membranes,
Biophys. J. 70 (1996) 2256–2265.
[21] T.J. McIntosh, S.A. Simon, D. Needham, C.H. Huang, Structure and
cohesive properties of sphingomyelin/cholesterol bilayers, Biochemistry
31 (1992) 2012–2020.
[22] T.J. McIntosh, A.D. Magid, S.A. Simon, Cholesterol modifies the short-
L. Chen et al. / Biochimica et Biophysica Acta 1768 (2007) 2873–2881
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
range repulsive interactions between phosphatidylcholine membranes,
Biochemistry 28 (1989) 17–25.
S. Karmakar, V.A. Raghunathan, Structure of phospholipid–cholesterol
membranes: an X-ray diffraction study, Phys. Rev., E 71 (2005) 061924.
D.A. Brown, E. London, Structure and function of sphingolipid- and
cholesterol-rich membrane rafts, J. Biol. Chem. 275 (2000) 17221–17224.
T.J. McIntosh, A. Vidal, S.A. Simon, Sorting of lipids and transmembrane
peptides between detergent-soluble bilayers and detergent-resistant rafts,
Biophys. J. 85 (2003) 1656–1666.
B.Y. van Duyl, D.T. Rijkers, B. de Kruijff, J.A. Killian, Influence of
hydrophobic mismatch and palmitoylation on the association of transmembrane alpha-helical peptides with detergent-resistant membranes,
FEBS Lett. 523 (2002) 79–84.
C. Boulin, R. Kempf, M.H.J. Koch, S.M. McLaughlin, Data appraisal,
evaluation and display of synchrotron radiation experiments: hardware and
software, Nucl. Instrum. Methods Phys. Res. A 249 (1986) 399–407.
J.F. Nagle, S. Tristram-Nagle, Structure of lipid bilayers, Biochim.
Biophys. Acta 1469 (2000) 159–195.
A.E. Blaurock, C.R. Worthington, Treatment of low angle X-ray data from
planar and concentric multilayered structures, Biophys. J. 6 (1966) 305–312.
T.J. McIntosh, A.D. Magid, S.A. Simon, Range of the solvation pressure
between lipid membranes: dependence on the packing density of solvent
molecules, Biochemistry 28 (1989) 7904–7912.
P.J. Quinn, H. Takahashi, I. Hatta, Characterization of complexes formed
in fully hydrated dispersions of dipalmitoyl derivatives of phosphatidylcholine and diacylglycerol, Biophys. J. 68 (1995) 1374–1382.
A. Filippov, G. Oradd, G. Lindblom, Lipid lateral diffusion in ordered and
disordered phases in raft mixtures, Biophys. J. 86 (2004) 891–896.
I. Hatta, Temperature scanning X-ray diffraction at phase transitions of
biologically related lipid assemblies — special attention to cholesterol-rich
state, J. Therm. Anal. Calorim. 82 (2005) 189–192.
M.R. Vist, J.H. Davis, Phase equilibria of cholesterol/dipalmitoyl
phosphatidylcholine mixtures: 2H nuclear magnetic resonance and
differential scanning calorimetry, Biochemistry 29 (1990) 451–464.
Z.W. Yu, P.J. Quinn, Phase stability of phosphatidylcholines in dimethylsulfoxide solutions, Biophys. J. 69 (1995) 1456–1463.
T.P. McMullen, R.N. McElhaney, New aspects of the interaction of
cholesterol with dipalmitoylphosphatidylcholine bilayers as revealed by
high-sensitivity differential scanning calorimetry, Biochim. Biophys. Acta
1234 (1995) 90–98.
V.I. Gordeliy, M.A. Kiselev, P. Lesieur, A.V. Pole, J. Teixeira, Lipid
membrane structure and interactions in dimethyl sulfoxide/water mixtures,
Biophys. J. 75 (1998) 2343–2351.
S. Tristram-Nagle, T. Moore, H.I. Petrache, J.F. Nagle, DMSO produces a
new subgel phase in DPPC: DSC and X-ray diffraction study, Biochim.
Biophys. Acta 1369 (1998) 19–33.
W.J. Sun, S. Tristram-Nagle, R.M. Suter, J.F. Nagle, Structure of gel phase
saturated lecithin bilayers: temperature and chain length dependence,
Biophys. J. 71 (1996) 885–891.
R.G. Wu, L. Chen, Z.W. Yu, P.J. Quinn, Phase diagram of stigmasterol–
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
2881
dipalmitoylphosphatidylcholine mixtures dispersed in excess water,
Biochim. Biophys. Acta 1758 (2006) 764–771.
T.J. McIntosh, The effect of cholesterol on the structure of phosphatidylcholine bilayers, Biochim. Biophys. Acta 513 (1978) 43–58.
M. Gandhavadi, D. Allende, A. Vidal, S.A. Simon, T.J. McIntosh,
Structure, composition, and peptide binding properties of detergent
soluble bilayers and detergent resistant rafts, Biophys. J. 82 (2002)
1469–1482.
S.N. Ahmed, D.A. Brown, E. London, On the origin of sphingolipid/
cholesterol-rich detergent-insoluble cell membranes: physiological concentrations of cholesterol and sphingolipid induce formation of a detergent-insoluble,
liquid-ordered lipid phase in model membranes, Biochemistry 36 (1997)
10944–10953.
N. Kucerka, S. Tristram-Nagle, J.F. Nagle, Structure of fully hydrated fluid
phase lipid bilayers with monounsaturated chains, J. Membr. Biol. 208 (2005)
193–202.
R. Elliott, K. Katsov, M. Schick, I. Szleifer, Phase separation of saturated
and mono-unsaturated lipids as determined from a microscopic model,
J. Chem. Phys. 122 (2005) 44904.
S.L. Veatch, S.L. Keller, Miscibility phase diagrams of giant vesicles
containing sphingomyelin, Phys. Rev. Lett. 94 (2005) 148101.
D. Scherfeld, N. Kahya, P. Schwille, Lipid dynamics and domain formation
in model membranes composed of ternary mixtures of unsaturated and
saturated phosphatidylcholines and cholesterol, Biophys. J. 85 (2003)
3758–3768.
S.L. Veatch, S.L. Keller, Separation of liquid phases in giant vesicles of ternary
mixtures of phospholipids and cholesterol, Biophys. J. 85 (2003) 3074–3083.
E. London, D.A. Brown, Insolubility of lipids in triton X-100: physical
origin and relationship to sphingolipid/cholesterol membrane domains
(rafts), Biochim. Biophys. Acta 1508 (2000) 182–195.
J. Pencer, M.P. Nieh, T.A. Harroun, S. Krueger, C. Adams, J. Katsaras,
Bilayer thickness and thermal response of dimyristoylphosphatidylcholine
unilamellar vesicles containing cholesterol, ergosterol and lanosterol: a
small-angle neutron scattering study, Biochim. Biophys. Acta 1720 (2005)
84–91.
J.A. Clarke, A.J. Heron, J.M. Seddon, R.V. Law, The diversity of the liquid
ordered (Lo) phase of phosphatidylcholine/cholesterol membranes: a
variable temperature multinuclear solid-state NMR and X-ray diffraction
study, Biophys. J. 90 (2006) 2383–2393.
L.J. Pike, Lipid rafts: bringing order to chaos, J. Lipid. Res. 44 (2003) 655–667.
L.J. Pike, Lipid rafts: heterogeneity on the high seas, Biochem. J. 378 (2004)
281–292.
Y. Feng, D. Rainteau, C. Chachaty, Z.W. Yu, C. Wolf, P.J. Quinn,
Characterization of a quasicrystalline phase in codispersions of phosphatidylethanolamine and glucocerebroside, Biophys. J. 86 (2004) 2208–2217.
C. Dansirikul, M. Choi, S.B. Duffull, Estimation of pharmacokinetic
parameters from non-compartmental variables using Microsoft Excel,
Comput. Biol. Med. 35 (2005) 389–403.
G.W. Feigenson, Phase behavior of lipid mixtures, Nat. Chem. Biol. 2 (2006)
560–563.