Membrane Transport, Structure, Function,

and Biogenesis:
Cholesterol Rules: DIRECT
OBSERVATION OF THE
COEXISTENCE OF TWO FLUID
PHASES IN NATIVE PULMONARY
SURFACTANT MEMBRANES AT
PHYSIOLOGICAL TEMPERATURES

Jorge Bernardino de la Serna, Jesus Perez-Gil,

Adam C. Simonsen and Luis A. Bagatolli
J. Biol. Chem. 2004, 279:40715-40722.

Downloaded from http://www.jbc.org/ by guest on May 18, 2014

doi: 10.1074/jbc.M404648200 originally published online July 1, 2004

Access the most updated version of this article at doi: 10.1074/jbc.M404648200

Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites.

Alerts:
• When this article is cited
• When a correction for this article is posted

Click here to choose from all of JBC's e-mail alerts

This article cites 44 references, 6 of which can be accessed free at

http://www.jbc.org/content/279/39/40715.full.html#ref-list-1
THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2004 by The American Society for Biochemistry and Molecular Biology, Inc.
Cholesterol Rules
DIRECT OBSERVATION OF THE COEXISTENCE OF TWO FLUID PHASES IN NATIVE PULMONARY
SURFACTANT MEMBRANES AT PHYSIOLOGICAL TEMPERATURES*
Received for publication, April 27, 2004, and in revised form, June 23, 2004
Published, JBC Papers in Press, July 1, 2004, DOI 10.1074/jbc.M404648200
Jorge Bernardino de la Serna‡, Jesus Perez-Gil‡, Adam C. Simonsen§, and Luis A. Bagatolli¶储
From the ‡Departmento de Bioquı́mica, Facultad de Biologı́a, Universidad Complutense, 28040 Madrid, Spain
and the MEMPHYS-Center for Biomembrane Physics, the §Department of Physics and the ¶Department of Biochemistry
and Molecular Biology, University of Southern Denmark, 5230 Odense, Denmark
Pulmonary surfactant, the lipid-protein material that
stabilizes the respiratory surface of the lungs, contains
approximately equimolar amounts of saturated and un-
saturated phospholipid species and significant propor-
tions of cholesterol. Such lipid composition suggests that
the membranes taking part in the surfactant structures
could be organized heterogeneously in the form of in-
plane domains, originating from particular distributions
of specific proteins and lipids. Here we report novel re-
sults concerning the lateral organization of bilayer mem-
branes made of native pulmonary surfactant where the
coexistence of two distinct micrometer sized fluid phases
(fluid ordered and fluid disordered-like phases) is ob-
served at physiological temperatures by using fluores-
cence microscopy and atomic force microscopy. Addi-
tional experiments using fluorescent-labeled proteins
SP-B and SP-C show that at physiological temperatures
these hydrophobic proteins are located exclusively in the
fluid disordered-like phase. Most interestingly, the micro-
scopic coexistence of fluid phases is maintained up to
37.5 °C, where most fluid ordered phases melt. This obser-
vation suggests that the particular composition of this
material is naturally designed to be at the “edge” of a
lateral structure transition under physiological condi-
tions, likely providing particular structural and dynamic
properties for its mechanical function. The observed lat-
eral structure in native pulmonary surfactant membranes
is dramatically affected by the extraction of cholesterol,
an effect not observed upon extraction of the surfactant
proteins. Furthermore, the spreading properties of the
native surfactant material at the air-liquid interface were
also greatly affected by cholesterol extraction, suggesting
a connection between the observed lateral structure and
a physiologically relevant function of the material. We
suggest that the particular lipid composition of surfac-
tant could be finely tuned to provide, under physiological
conditions, a structural scaffold for surfactant proteins to
act at appropriate local densities and lipid composition.
Pulmonary surfactant is the main secretory product of the
alveolar type II pneumocytes and is required to stabilize the
* This work was supported by Dirección General de Educación Supe-
rior e Investigaciones Cientı́ficas Grant BIO2003-09056 (to J. B. S. and
J. P.-G.), Comunidad Autónoma de Madrid Grant 08.2/0054, and by
Danish Natural Science Research Council Grant 21-03-0569 (to
L. A. B.). The costs of publication of this article were defrayed in part by
the payment of page charges. This article must therefore be hereby
marked “advertisement” in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
储 To whom correspondence should be addressed: MEMPHYS-Center for
Biomembrane Physics, Dept. of Biochemistry and Molecular Biology, Uni-
versity of Southern Denmark, 5230 Odense M, Denmark. Tel.: 45-65-50-
34-76; Fax: 45-66-15-87-60; E-mail: bagatolli@memphys.sdu.dk.
This paper is available on line at http://www.jbc.org
Vol. 279, No. 39, Issue of September 24, pp. 40715–40722, 2004
Printed in U.S.A.
lungs in air-breathing animals (1, 2). This surfactant material
is mainly composed of lipids, mainly phospholipids, and small
amounts of specifically associated proteins. Among the phos-
pholipids significant amounts of dipalmitoylphosphatidylcho-
line (DPPC)1 and phosphatidylglycerol are present, both of
which are unusual species in most animal membranes. Mono-
unsaturated phosphatidylcholines (PC), phosphatidylinositol,
and neutral lipids including cholesterol are also present in
pulmonary surfactant in varying proportions. The surfactant
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
proteins (SP) SP-A, SP-B, SP-C, and SP-D constitute about 8%
of lung surfactant by weight. SP-A and SP-D are large glyco-
proteins (⬎500 kDa) belonging to the superfamily of collectins,
and they serve important functions in innate defense mecha-
nisms of the lung (3). The hydrophobic proteins SP-B and SP-C
modulate the surface-active properties of surfactant lipids and
are strictly required to establish an operational respiratory
surface (4). Pulmonary surfactant lipid-protein complex is se-
creted into the thin aqueous alveolar lining of the lung as
multilamellar assemblies, which spontaneously transform into
nanotubular membrane-based structures called tubular mye-
lin. These structures are considered reservoirs of highly sur-
face-active components in the pathway that forms surface-
active films at the lung air-water interface. Such films can
reduce the surface tension to nearly 0 mN/m and in doing so
prevent alveolar collapse at low lung volumes (1, 5). Dysfunc-
tions of the surfactant system are relevant in diseases includ-
ing neonatal and acute respiratory distress syndrome, cystic
fibrosis, and pneumonia (6). Several studies (7, 8) have shown
that interfacial films made with the hydrophobic fraction (lip-
ids ⫹ SP-B ⫹ SP-C) of pulmonary surfactant undergo phase
separation under lateral compression. In addition, the partic-
ular phenomenon of gel/fluid phase separation induced by tem-
perature was also reported in bilayers composed of part of the
hydrophobic fraction of pulmonary surfactant at physiological
temperatures (9). Even though the particular composition of
the lung surfactant suggests that native surfactant membrane-
based structures could exhibit lateral segregation phenomena
at physiological temperatures, this aspect has been merely
speculative up to now. For example, the presence of high
amounts of DPPC (⬃40% weight of the total material) with a
melting phase transition above mammalian physiological tem-
peratures (41.5 °C) indicates the possibility of the coexistence
of a solid/fluid-like phase (10). On the other hand, the presence
1
The abbreviations used are: DPPC, 1,2-dipalmitoyl-sn-glycero-3-
phosphocholine; AFM, atomic force microscopy; Bodipy-PC, 2-(4,4-
difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-pentanoyl)-1-
hexadecanoyl-sn-glycero-3-phosphocholine; DiIC18, 1,1⬘-dioctadecyl-
3,3,3⬘,3⬘-tetramethylindocarbocyanine perchlorate; Me2SO, dimethyl
sulfoxide; DOPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine; GUV, giant
unilamellar vesicle; NPSM, native pulmonary surfactant material; SP,
surfactant proteins; PC, phosphatidylcholine.
40715
40716 Phase Separation in Native Surfactant Membranes
of cholesterol along with other unsaturated phospholipid spe-
cies in the native material suggests the possibility of a fluid
ordered/fluid disordered-like phase coexistence (distinguished
in part by their cholesterol content) under physiologically rel-
evant conditions (10). Even though lung surfactant has choles-
terol concentrations of up to 20 –22 mol % (11, 12), there is no
clear understanding of how this molecule impacts the lateral
structure of the native material. The presence of cholesterol
could thus have important consequences for the lateral organi-
zation of the native pulmonary surfactant material, illustrat-
ing another example of the fundamental role of cholesterol in
modulating the lateral structure of membranes, as the raft
hypothesis proposes (13, 14). Furthermore, there is presently
no clear link between membrane lateral heterogeneity and the
physiological function of native pulmonary surfactant material
(10).
In recent years, a new experimental strategy, based on the
direct visualization of free standing lipid bilayers using giant
unilamellar vesicle technology in conjunction with confocal and
two photon excitation fluorescence microscopy techniques, has
opened the possibility of correlating morphological and dynam-
ical information on lipid membranes at molecular and supra-
molecular levels. Important novel information such as the mor-
phology of different coexisting lipid phases (such as gel/fluid
and fluid ordered/fluid disordered), mechanical properties of
membranes displaying phase coexistence, local hydration, and
molecular diffusion in lipid domains can be extracted directly
from the fluorescence images (9, 15–23). The present study
uses the GUV technology in addition to other experimental
techniques to provide evidence that a particular lateral struc-
ture occurs in native membranes of pulmonary surfactant at
physiological temperatures. Additionally we demonstrate that
cholesterol plays a critical role in promoting such membrane
organization. On the other hand, it is important to note that
preparation of GUVs was previously limited to membranes
composed of artificial and natural lipid extracts (from lipid
solutions in organic solvents). In this study we also present a
method to prepare GUVs composed of native membranes with-
out using organic solvents or detergents. To our knowledge,
GUV experiments performed with native membrane prepara-
tions to further evaluate the concurrent effect of lipids and
proteins in lateral segregation phenomena have not yet been
reported.
EXPERIMENTAL PROCEDURES
Materials—1,1⬘-Dioctadecyl-3,3,3⬘,3⬘-tetramethylindocarbocyanine
perchlorate (DiIC18), 2-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-
indacene-3-pentanoyl)-1-hexadecanoyl-sn-glycero-3-phosphocholine
(Bodipy-PC), and isothiocyanate derivatives of Alexa Fluor® 488 and
Texas Red® were from Molecular Probes Inc. (Eugene, OR). Methyl-␤-
cyclodextrin was from Aldrich. DPPC, 1,2-dioleoyl-sn-glycero-3-
phosphocholine (DOPC), and cholesterol were purchased from Avanti
Polar Lipids (Alabaster, AL) and were used without further purifica-
tion. Native pulmonary surfactant material was obtained from pig
lungs as described previously (24). The cholesterol quantitation kits
INFINITYTM and IVD were purchased from Sigma and Spinreact
(Girona, Spain), respectively.
Preparation of Surfactant and Proteins—Stock suspensions of native
pulmonary surfactant material (NPSM) were in 5 mM Tris buffer, pH 7,
containing 150 mM NaCl. The total concentration of phospholipid was
0.5 mg/ml, estimated by phosphorus quantitation upon phospholipid
mineralization (25). For fluorescence experiments the fluorescent
probes were incorporated into NPSM stock suspensions before prepar-
ing the giant vesicles. In the case of the lipophilic probes (DiIC18 and
Bodipy-PC) a small aliquot (0.5 ␮l) of an Me2SO solution containing the
fluorescent probes (0.125 mM) was added to the NPSM stock suspension
(final volume 500 ␮l) and incubated for 20 min at room temperature. In
all cases the percentage of fluorescent probes in the sample was less
than 0.1 mol %.
Fluorescently labeled surfactant proteins were prepared by labeling
purified SP-B and SP-C solutions in organic solvent, as described pre-
viously (26). This procedure allows attachment of a single fluorophore to
the N-terminal amine group of the protein. In our studies Alexa 488 and
Texas Red were used for SP-C and SP-B, respectively. Labeled proteins
were incorporated into NPSM by addition of small aliquots of fluores-
cent SP-B and SP-C (0.5% protein to lipid by weight) in Me2SO.
Giant Vesicle Preparation—GUVs composed of NPSM organic ex-
tract, its lipid fraction, or the ternary mixture DOPC/DPPC/cholesterol
were prepared as described previously (16, 27) by using the electrofor-
mation method originally developed by Angelova and Dimitrov (28) and
Angelova et al. (29). A previously described (16, 27) special tempera-
ture-controlled chamber was used for this purpose. Briefly the process
can be described in the following three steps. 1) ⬃3 ␮l of the stock
solution of the lipid or lipid/protein organic solution were spread on the
surface of each platinum wire. The chamber was located under a stream
of N2 during this procedure and then placed under vacuum overnight to
remove the organic solvent. 2) Aqueous solution was added to the
chamber (200 mosM sucrose solution prepared with Millipore-filtered
water, 17.5 megohms/cm). The sucrose solution was previously heated
to the desired temperature (above the lipid mixture phase transition,
we used 45 °C), and then sufficient volume was added to cover the
platinum wires (⬃300 ␮l). 3) The platinum wires were connected im-
mediately to a function generator (Digimess® FG 100, Germany), and a
low frequency AC field (sinusoidal wave function with a frequency of 10
Hz and an amplitude of 3 V) was applied for 120 min. After vesicle
formation, the AC field was turned off, and the vesicles were collected
with a pipette and transferred to a plastic tube.
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
NPSM GUVs were also prepared using the electroformation method
but with some modifications in step 1 of the protocol described above. In
this case NPSM GUVs were formed from NPSM suspended in buffer
solution (no organic solvents in this case). Briefly, ⬃3 ␮l of the stock
suspension in buffer was spread on the surface of each platinum wire as
small drops. The chamber was then placed under a stream of N2 and
subsequently under low vacuum for 30 min to allow the native material
to adsorb onto the platinum wire. An important point in this step is to
avoid dehydration of the sample to maintain the integrity of the native
membranes. Once the material was adsorbed to the Pt wire, we proceed
with steps 2 and 3 as described above. As a control experiment, GUVs
composed of DOPC/DPPC/cholesterol were also prepared from multila-
mellar suspensions in aqueous solution as described above for NPSM
and then compared with those obtained from the lipid organic extract.
No differences in the morphology of the final resulting GUVs were
observed between the two described protocols.
Observation of Giant Vesicles—Aliquots of giant vesicles suspended
in sucrose were added to an equi-osmolar concentration of glucose
solution. Because of the density difference between the two solutions,
the vesicles precipitate at the bottom of the chamber which facilitates
observation of the GUVs in the inverted confocal microscope. GUV
preparations were observed in 8-well plastic chambers (Lab-Tek Brand
Products, Naperville IL). The chamber was located in an inverted
confocal microscope (Zeiss LSM 510 META) for observation. The exci-
tation wavelengths were 488 (for Alexa 488 and Bodipy-PC) and 543 nm
(for DiIC18 and Texas Red). The temperature was controlled from a
water bath connected to a homemade device into which the 8-well
plastic chamber was inserted. The temperature was measured inside
the sample chamber using a digital thermocouple (model 400B, Omega
Inc., Stamford, CT) with a precision of 0.1 °C.
Differential Scanning Calorimetry—Experiments were performed in
a Science Corp. N-DSCII calorimeter using 15 mM total lipid concen-
tration of native pulmonary surfactant suspension. The scan rate was
0.5 K/min versus buffer. Five temperature scans were collected from
each sample between 25 to 70 °C. Neither change in the total concen-
tration nor in the scan rate gave significant changes in the obtained
results.
Cholesterol Extraction Experiments—Methyl-␤-cyclodextrin/glucose
solutions were prepared in Millipore water 17.5 megohms/cm. The
osmolarity of these solutions was measured in an osmometer (The
Advance Osmometer, model 3D3, Advance Instruments Inc., Norwood,
MA) and adjusted to match the osmolarity of the vesicles containing
sucrose solution. Vesicles were diluted in the methyl-␤-cyclodextrin/
glucose-containing solution and observed afterward. Alternatively, ali-
quots of methyl-␤-cyclodextrin solutions were added to the vesicles
already dispersed in glucose solution. No differences in the effect of
methyl-␤-cyclodextrin on the vesicles were observed between the two
procedures. Cholesterol was also removed from the lipid fraction of
NPSM by hydrophobic chromatography in Sephadex LH-20 (30).
Phospholipid and Cholesterol Determination—Total phospholipid
concentration in the different samples was determined by phosphorus
analysis as we described above (25). The amount of cholesterol in NPSM
 Phase Separation in Native Surfactant Membranes
FIG. 1. Giant vesicles from native pulmonary surfactant mem-
branes. Giant unilamellar (A–C) and multilamellar (D) vesicles formed
from native pulmonary surfactant material. A, transmission image;
B–D, fluorescent images. The confocal planes in the fluorescent images
were obtained at the polar (B) and center (C and D) regions of the giant
vesicles. The temperature was 37 °C. The giant vesicles are labeled with
the lipophilic fluorescence probes DiIC18 (red) and Bodipy-PC (green). In
the NPSM the rounded green areas in the images correspond to fluid
disordered regions, and the red background corresponds to fluid ordered
phase. The scale bars correspond to 10 ␮m. Note that the different
domains span the lipid bilayer (C and D).
and in their different fractions was quantitatively estimated using the
following: (i) a colorimetric assay based on the activity of cholesterol
oxidase (IVD kit from Spinreact, Girona), and (ii) the determination of
peroxide products resulting from cholesterol oxidation (INFINITYTM
kit from Sigma). Cholesterol content in the samples is given as choles-
terol-to-phospholipid molar ratio. Four different samples were used for
total cholesterol quantitation from each of at least three different
batches of each of the fractions studied.
Protein Extraction from NPSM—The hydrophobic fraction of NPSM,
containing all the lipid species plus the hydrophobic proteins SP-B and
SP-C but lacking hydrophilic proteins SP-A and SP-D, was obtained by
chloroform/methanol extraction. Separation of lipid from protein com-
ponents in the organic mixture was achieved by chromatography of the
surfactant organic extract through a Sephadex LH-20 column, as de-
scribed previously (31).
Atomic Force Microscopy of NPSM Bilayers—NPSM aqueous suspen-
sions were spread on top of an ultrapure water subphase in a Langmuir-
Blodgett trough (Kibron, ␮-trough) until a surface pressure of 1 mN/m
was obtained. The film was allowed to equilibrate for 10 min and was
subsequently compressed to 42 mN/m. While maintaining this pressure,
a film was transferred to a freshly cleaved mica substrate previously
immersed into the subphase by lifting the support at a constant speed
of 1.5 mm/min. Following the initial deposition, a second layer was
transferred onto the first by re-immersion of the coated mica into the
subphase at 1.5 mm/s (also at a constant surface pressure of 42 mN/m).
Consequently, the supported membranes used in the atomic force mi-
croscopy (AFM) experiments are bilayer structures formed by double
transfer of the surface film (including any additional associated struc-
ture to the monolayer), which is obtained by spreading aqueous sus-
pensions of native surfactant at the air-water interface. Topographical
AFM images were obtained under aqueous conditions in a PicoSPM
(Molecular Imaging) microscope operated in magnetically activated tap-
ping mode using MAClevers.
Spreading Kinetics—Surface activity of surfactant preparations was
assayed by running ␲-t interfacial adsorption isotherms. In a typical
experiment, 3 ␮l of surfactant suspension (10 mg/ml phospholipid con-
centration) were spread by direct deposition on top of a subphase 5 mM
Tris, pH 7, containing 150 mM NaCl in the Teflon trough (20 cm2) of a
thermostated surface balance built by Nima (Coventry, UK), and at 5
cm of a filter paper Wilhelmy plate, before monitoring surface pressure
against time. Spreading experiments were done either at 25 or 37 °C.
Spreading experiments in the presence of methyl-␤-cyclodextrin were
carried out by spreading equivalent amounts of native surfactant on top
of 5 mM Tris, pH 7, subphases containing 150 mM NaCl, and different
concentrations of methyl-␤-cyclodextrin. Data shown are representative
of three different experiments for two different batches of each assayed
surfactant preparation.
RESULTS
Fig. 1 shows images of single GUVs composed of NPSM
doped with the fluorescent probes Bodipy-PC and DiIC18.
Based on the particular round shape of the domains (18, 20, 21)
and the partition properties of the different fluorescent probes
40717
(15, 19), we conclude that the lateral organization of the NPSM
vesicles corresponds to a fluid ordered/fluid disordered-like
phase coexistence. The coexistence of two fluid phases persists
from 22 to 37.5 °C. In general, when fluid domains are embed-
ded in a fluid environment, circular domains are formed be-
cause both phases are isotropic, and the line energy (tension),
which is associated with the rim of two demixing phases, is
minimized by optimizing the area-to-perimeter ratio. As ob-
served in Fig. 1C, the domains span the bilayer in agreement
with previous observations done in artificial and natural lipid
mixtures (15–18). NPSM frequently gave rise to multilamellar
giant vesicles (see Fig. 1D), whose morphology resembles the
structure of lamellar bodies that are the multilamellar assem-
blies storing and secreting pulmonary surfactant in the type II
pneumocytes of the lung. Coexistence of two fluid phases was
observed in every bilayer of these giant multilamellar assem-
blies. Fig. 2 shows an AFM image of an NPSM-supported
membrane, without exogenous fluorescent probes. Such mem-
branes are part of the surface-active film of pulmonary surfac-
tant that spontaneously forms upon adsorption at the interface
(32). It is important to emphasize that the samples visualized
by AFM are not simple monolayers, as those studied previously
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
in the literature, but bilayers as we perform a double deposition
of surfactant films onto mica (see “Experimental Procedures”).
AFM scanning of supported membranes demonstrates the
presence of circular-shaped domains with sizes comparable
with those seen in GUVs, confirming that two fluid phases
coexist in the native material. This observation represents an
important control experiment because it indicates that the
membrane lateral structure in the GUVs is really present in
the original material and is not because of the incorporation of
fluorescent probes or the way the samples are prepared. AFM
also reveals the coexistence of nano-scale structures inside the
circular-shaped domains. Topological analysis shows that the
distribution of heights observed in the round domains (i.e.
higher versus lower heights) is more heterogeneous than the
height distribution observed in the more homogeneous inter-
vening background (Fig. 2). In particular, the round domains
contain regions of lower heights compared to the height ob-
served in the continuous region. This last finding suggests that
1) the round domains may correspond to a liquid disordered-
like phase state and that 2) the region surrounding the round
domains may display a liquid ordered-like phase state. The
higher heights in the round domains may be due to the pres-
ence of surfactant proteins. The complex topography of the
liquid-disordered domains is likely connected with particular
organization of the many different molecular species (lipids
and proteins) in the plane of the membrane. Given the compo-
sitional complexity of the material, further experiments to clar-
ify this issue are warranted but are outside of the scope of the
present study.
The particular domain pattern observed on the GUV surface
is likely to be relevant for modulating the lateral distribution of
the surfactant hydrophobic proteins SP-B and SP-C in both the
native membranes and those formed from surfactant organic
extract, as shown in Fig. 3. When fluorescently labeled SP-B
and SP-C were incorporated into NPSM GUVs or into GUVs
formed from the hydrophobic fraction of NPSM, the fluores-
cence of these proteins colocalized with the Bodipy-PC-enriched
areas (from 22 to 37.5 °C). This last fact indicates that the
hydrophobic proteins segregate into fluid disordered regions, in
agreement with the observations obtained in the AFM experi-
ments (Fig. 2). The partitioning of SP-B and SP-C in NPSM and
in bilayers made of surfactant organic extract resemble the
distribution of the proteins observed in monolayer experi-
ments, where both SP-B and SP-C preferentially situate in
40718 Phase Separation in Native Surfactant Membranes
FIG. 2. AFM scanning image from a
pulmonary surfactant membrane. A,
AFM scanning image of membranes
formed by spreading NPSM on top of an
air-liquid interface, at a lateral pressure
of 42 mN/m. Film transfer was performed
using the Langmuir-Blodgett method
onto a mica support. The supported mem-
brane was scanned in a liquid cell using
tapping mode. B, magnification of the re-
gion marked in A. C, topographic profile
along the white line included in B. The
temperature was 25 °C.
FIG. 3. Lateral distribution of surfactant proteins SP-B and
SP-C in pulmonary surfactant membranes. Fluorescence images of
GUVs composed of NPSM (A) or NPSM organic extract (B) upon incor-
poration of fluorescently labeled proteins SP-B and SP-C (0.5% protein/
lipid by weight). The two color images (bottom images) are split into
those corresponding to Alexa 488-labeled SP-C (green) and Texas Red-
labeled SP-B (red). Scale bars are 10 ␮m. Images were taken at 37 °C.
liquid-expanded areas (33). This distribution is particularly
remarkable in the case of SP-C, because this protein should, in
principle, have a different local environment in bilayers with
respect to monolayers simply because of the different mem-
brane thickness (SP-C in bilayers accommodate in a transmem-
brane orientation (1)). Our results seem also to discard the
potential association of SP-C with cholesterol-enriched liquid
ordered regions of the membranes, in contrast with that de-
scribed for other palmitoylated transmembrane proteins (34).
Additionally, we observed that removal of SP-A from NPSM
does not affect the partition properties of either SP-B or SP-C
proteins (Fig. 3B).
Fig. 4 illustrates the effect of temperature on the lateral
structure of NPSM. The liquid disordered domains showed
dynamically fluctuating borderlines when the temperature was
raised close to 37.5 °C followed by a dramatic lateral structure
change above 37.5 °C. Fig. 4A shows that at 38 °C, most of the
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
liquid ordered phase has melted, yielding less than 10% of the
surface covered by a network of filamentous shapes, probably
ordered-like phase. Fig. 4B shows a representative differential
scanning calorimetry experiment obtained from NPSM solu-
tions. The thermogram is characterized by a broad thermal
transition showing the end of the melting process above
37.5 °C, in agreement with the data obtained in GUVs and that
previously reported for alveolar surfactant (crude lung wash)
dispersion from mongrel dogs (35). Additionally, a very similar
effect is observed by using fluorescent-labeled proteins SP-B
and SP-C instead of the lipophilic fluorescent probes. From our
experimental data we can observe that the proteins remain in
the fluid disordered areas at temperatures above the lateral
structure change as observed at 37.5 °C (Fig. 4C).
The similarities between the fluorescent images obtained in
the NPSM and those reported previously (18, 20 –22) in choles-
terol-containing ternary lipid mixtures are remarkable and
suggest a potentially important role for cholesterol in the lat-
eral structural organization of NPSM. However, because of the
compositional complexity of NPSM, the potential influence of
other components, in particular surfactant proteins, on the
lateral arrangement of this membranous system had to be
evaluated. To address this point we used different experimen-
tal strategies involving the extraction of particular components
of NPSM. Cholesterol extraction with methyl-␤-cyclodextrin
and extraction of surfactant proteins (SP-A, SP-B and SP-C)
were performed on the NPSM, and the results are summarized
in Fig. 5A. Neither the absence of the water-soluble fraction
(including the collectin SP-A) nor the extraction of the hydro-
phobic proteins SP-B and SP-C from the NPSM changed the
circular shape of the lipid domains between 22 and 37.5 °C.
This result indicates that the presence of the surfactant pro-
teins is not critical to maintain the phase coexistence pattern
observed in NPSM. The cholesterol-to-phospholipid molar ratio
in all these materials was similar (between 18 and 21 mol %),
as summarized in Table I. Fig. 5A also shows that after iso-
thermal cholesterol extraction with methyl-␤-cyclodextrin, the
phase coexistence pattern characterized by round domains in
NPSM changes to elongated irregularly shaped domains, sim-
 Phase Separation in Native Surfactant Membranes
FIG. 4. Effect of temperature on the
structure of native pulmonary sur-
factant membranes. Representative
confocal fluorescence images of giant
unilamellar vesicles composed of native
pulmonary surfactant material at the in-
dicated temperatures obtained with the
lipophilic fluorescent probes DiIC18 and
Bodipy-PC (A) and fluorescently labeled
SP-B and SP-C (C). The fluorescent image
in C is split into two colors corresponding
to Alexa 488-labeled SP-C (green) and
Texas Red-labeled SP-B (red). Scale bars
are 20 ␮m. Differential scanning calorim-
etry thermogram of NPSM (B).
ilar to those reported previously in artificial ternary mixtures
at very low cholesterol levels (20 –22) and in DPPC-containing
phospholipid binary mixtures (15, 16). The lateral structure
observed in these systems corresponds to the coexistence of gel
and fluid phases. To explain further the observed cholesterol-
dependent phase changes in NSPM at constant temperature,
cholesterol-extraction experiments using methyl-␤-cyclodex-
trin were performed in the well characterized DOPC/DPPC/
cholesterol mixture (20 –22). This particular lipid mixture was
chosen as a model system because the presence of DPPC and
cholesterol is similar to that of NPSM. Fig. 5B shows two
different phase transitions as a result of increasing methyl-␤-
cyclodextrin concentration at constant temperature, i.e. fluid
ordered 3 fluid ordered/fluid disordered and fluid ordered/fluid
disordered 3 gel/fluid. The last phase transition occurs in a
remarkably similar fashion to that observed in NPSM (com-
pare Fig. 5, A and B). The domain compositional information
already reported for DOPC/DPPC/cholesterol in the fluid or-
dered/fluid disordered phase coexistence regime (20 –22) sug-
gests that the DPPC-containing fluid ordered phase is enriched
in cholesterol compared with the DOPC-enriched fluid disor-
dered phase where the cholesterol content is low. Based on
these observations we propose that the fluid ordered phase in
NPSM may be enriched in DPPC and cholesterol compared
with the fluid disordered phase. Additionally, from our exper-
iments we notice that although the shape of the lipid domains
is quite similar for NPSM and the cholesterol-containing ter-
nary mixtures in the same phase coexistence regime (compare
Fig. 5A, top left image, with B, center image, and Fig. 5A,
bottom left image, with B, bottom image), there are some dif-
ferences in the partition properties of the fluorescent probes.
This observation suggests that the molecular composition of
these membrane domains (i.e. fluid ordered, gel, fluid disor-
dered) could play an important role in modulating specific
molecular interactions in the plane of the membrane in agree-
ment with previous results observed in lipid mixtures (18).
We found that the changes observed as the cholesterol con-
centration is decreased (Fig. 5A) can be reversed by re-intro-
ducing cholesterol as shown in Fig. 5C. GUVs formed from
surfactant lipid extract partially depleted of cholesterol (cho-
lesterol to phospholipid ratio 10 mol %, instead of 20 mol % as
in NPSM) show a similar pattern of gel/fluid phase coexistence
as that observed in NPSM GUVs treated with a high concen-
40719
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
tration of methyl-␤-cyclodextrin (compare Fig. 5, A and C).
Increasing the cholesterol concentration to 15 mol % induces
the appearance of numerous circular-shaped domains. Further
increase in cholesterol (up to 30 mol % relative to phospholip-
ids) yields an increase in the area occupied by the fluid-ordered
phase (Fig. 5C, red area in the GUVs containing 15–30 mol %
cholesterol). Note that size and morphology of the fluid disor-
dered domains become comparable with those observed in
NPSM (Fig. 5A) upon addition of 8 –10 mol % of cholesterol,
which raises the cholesterol content to values in the same
range as those in native pulmonary surfactant. From this ex-
periment, we estimate that a cholesterol-to-phospholipid ratio
of about 13–15 mol % triggers a phase change from gel/fluid to
fluid ordered/fluid disordered phase coexistence. These exper-
iments clearly demonstrate that the cholesterol concentration
is a critical parameter in modulating the lateral structure of
pulmonary surfactant membranes. Moreover, the cholesterol-
dependent phase coexistence pattern reported here may ex-
plain why the phase coexistence pattern reported previously in
GUVs composed of the bovine lipid extract surfactant BLES®,
a clinical preparation, corresponds to gel/fluid phase coexist-
ence (9). BLES® is almost devoid of cholesterol, and in this
respect it is similar to NPSM at relatively high methyl-␤-
cyclodextrin concentration.
One key question is whether the lateral structure of the
surfactant material may affect its functional properties. Rapid
spreading and adsorption of lung surfactant material to form
surface-active films at the air-water interface is an important
property that can be used to evaluate the proper function of a
given surfactant. Fig. 6 presents spreading-at-interface ␲-t iso-
therms of different surfactant suspensions directly deposited
on top of open buffered subphases. Both native surfactant as
purified from porcine lungs and its reconstituted organic frac-
tion (lipids ⫹ SP-B ⫹ SP-C) have optimal adsorption activities,
very rapidly producing equilibrium surface pressures around
45 mN/m (Fig. 6, A and B). Suspensions lacking proteins show
much lower adsorption rates and do not reach surface pres-
sures higher than 30 mN/m after 50 min, indicating the essen-
tial role of hydrophobic proteins to catalyze phospholipid inter-
facial transference. Most interestingly, spreading isotherms of
suspensions prepared from the surfactant lipid fraction par-
tially depleted of cholesterol are further impaired, suggesting
that cholesterol-containing surfactant structures have some-
40720 Phase Separation in Native Surfactant Membranes
FIG. 5. Fluorescent images of GUVs made from different sur-
factant fractions and lipid mixtures. A, GUV composed of NPSM
(top left) and GUV composed of the organic extract of NPSM that
contains all the hydrophobic components but no hydrophilic proteins
(top right); GUV composed of the lipid fraction of the NPSM organic
extract, from which hydrophobic proteins SP-B and SP-C have also been
removed (bottom right), and GUV of NPSM treated with 40 mg/ml
methyl-␤-cyclodextrin (bottom left). Images were taken at 25 °C. The
cholesterol-to-phospholipid molar ratio for the samples that display
coexistence of two fluid phases (round-shaped domains) are 21% (top left
image), 19% (top right image), and 18% (bottom right). B, fluorescence
images of GUVs composed of DOPC/DPPC/cholesterol (1:1:40 mol %)
before (top) and after exposure to 10 (center) and 40 mg/ml (bottom) of
methyl-␤-cyclodextrin. Images were taken at 25 °C. All the GUVs in the
figure have been labeled with the lipophilic fluorescence probes DiIC18
(red) and Bodipy-PC (green). C, fluorescence images of GUVs composed
of a lipid fraction of NPSM that contains no surfactant proteins at
different cholesterol molar ratio (with respect to phospholipids). Images
from left to right are from GUVs containing 10, 15, 18, 21, and 30 mol %
cholesterol. Images were taken at 25 °C. The scale bars correspond to 10
␮m.
TABLE I
Cholesterol content in NPSM and its fractions
% cholesterol/phospholipid
molar ratioa
NPSM 21.3 ⫾ 2
NPSM organic extract 19.2 ⫾ 3
Protein-free NPSM organic extract 18.5 ⫾ 3
Cholesterol-depleted, protein-free NPSM 9.4 ⫾ 2
organic extract
a
Values are means ⫾ S.D. after averaging four different samples of
each material.
how better interfacial dynamics. Spreading kinetics were al-
ways faster at 37 than at 25 °C, but the difference in the ability
of the different samples to reach the highest surface pressures
persisted, suggesting that such effect is because of intrinsic
structural differences among the samples. Fig. 6C shows
spreading kinetics of NPSM suspensions spread on top of sub-
phases containing different concentrations of methyl-␤-cyclo-
dextrin. Native surfactant spread rapidly at the first instance,
independent of the amount of methyl-␤-cyclodextrin present in
the subphase. However, the higher the amount of methyl-␤-
cyclodextrin in the subphase (i.e. the higher amount of choles-
terol extracted from the spreading surfactant), the lower the
maximum pressure reached at the late spreading stages. Op-
timal insertion of phospholipids against high pressures seems
therefore to require the action of surfactant-specialized struc-
tures such as the one described here when both cholesterol and
hydrophobic proteins are present.
DISCUSSION
The present study offers a novel view on the importance of
cholesterol for the relationships in the structure-function of
pulmonary surfactant, and suggests a rationale for the pres-
ence of a significant amount of this lipid species in the compo-
sition of all the known surfactants. In this paper we demon-
strate that the presence of cholesterol promotes a defined
lateral structure in surfactant bilayers at physiological temper-
atures. The particular lateral organization of lipids and pro-
teins in surfactant membranes would be essential to support
not only a rapid interfacial adsorption to equilibrium surface
pressures (as shown in the Fig. 6), but for the formation of well
defined surface-associated structures (such as that shown in
the Fig. 2). This last scenario would be competent to support
proper dynamic surface behavior when the surfactant material
Downloaded from http://www.jbc.org/ by guest on May 18, 2014
is subjected to rapid dynamic cycling.
Several studies have analyzed previously the effects of cho-
lesterol on the surface activity of surfactant, mostly using
model lipid mixtures. However, the conclusions about a poten-
tial role of cholesterol in surfactant function have been contra-
dictory. Some studies have concluded that cholesterol has a
general beneficial effect on surfactant interfacial adsorption,
mostly due to a “fluidizing” effect on DPPC-enriched surfactant
lipid systems (36). On the other hand, the surface behavior
studies of cholesterol-containing monolayers using model lipid
mixtures has come to the conclusion that cholesterol impairs
the ability of any lipid mixture to reach and sustain very high
surface pressures during compression (37–39). As a conse-
quence of these studies, many of the clinical surfactants used
today are depleted of cholesterol, which is considered as a sort
of “contaminant” for a good surface activity. Our results sug-
gest that last idea should be re-examined by using native
material as model system.
The dependence of the lateral structure of surfactant mem-
branes on cholesterol concentration provides a framework for the
physiological meaning of rapid changes in the cholesterol-to-
phospholipid ratio reported previously (11) in pulmonary surfac-
tant in response to various physiological stimuli. For example, in
homeothermic mammals, the cholesterol content changes with
exercise, suggesting that changes in the mechanical properties of
surfactant membranes are required to accommodate the changes
in ventilation. The cholesterol content in surfactant also varies
with the temperature of the environment in both heterothermic
species (40) and in animals that enter torpor (41). The proportion
of cholesterol correlates with differences in lung structure, both
from a phylogenetic and an ontogenetic point of view (42). The
amount of cholesterol could be finely tuned in the context of the
presence, in given amounts, of other lipid species, such as satu-
rated and unsaturated phospholipids, to provide a structural
“scaffold” for surfactant membranes at precise conditions of tem-
perature and breathing dynamics. Potential physiological mech-
anisms providing tight regulatory control of surfactant composi-
tion, especially regarding cholesterol/phospholipid ratio, should
be investigated as they could be evolutionarily conserved to fit
surfactant material properties to required performance. For ex-
ample, compositional data reported for surfactant in the litera-
ture include remarkable differences in terms of cholesterol-to
phospholipid ratio, both among species and among individuals.
The possibility that such differences could originate, at least
 Phase Separation in Native Surfactant Membranes 40721

FIG. 6. Interfacial spreading ␲-t isotherms of NPSM and its fractions. ␲-t isotherms at 25 (A) and 37 °C (B) of NPSM (closed squares), its
reconstituted organic extract containing all the lipids plus the hydrophobic proteins (open squares), its reconstituted lipid fraction without proteins
(closed circles), and its reconstituted lipid fraction without proteins and partly depleted of cholesterol (10 mol % instead of 20%) (open circles). C,
␲-t isotherms of NPSM spread at 25 °C on top of subphases containing different concentrations of methyl-␤-cyclodextrin (indicated in the figure,
in mg/ml). All the experiments were done by spreading 3 ␮l of an aqueous suspension of the different materials (phospholipid concentration 10
mg/ml) on top of a 20-cm2 thermostated trough containing buffer (5 mM Tris, 150 mM NaCl, pH 7), with or without methyl-␤-cyclodextrin, as a
subphase. The subphase was equilibrated for 4 min at the desired temperature before initiating sample spreading.

partly, from environmental effects should be explored. These and hydrophobic proteins are involved in initiating monolayer-

Downloaded from http://www.jbc.org/ by guest on May 18, 2014

ideas need still to be fully established in future experiments, for
example by comparing different surfactant preparations ex-
tracted from particular animal species at different environmen-
tal conditions.
The thermotropic behavior of NPSM (Fig. 4) suggests that
the composition of this material needs to be optimized to be at
the “edge” of the liquid ordered/liquid disordered phase coex-
istence regime at physiological temperatures, which is likely to
provide particular structural and dynamical properties for its
mechanical function under particular environmental condi-
tions. It is interesting to note that the phase transition event
observed for NPSM is different from that observed for lipid
mixtures displaying coexistence of two fluid phases (18, 21). In
the artificial system the fluid ordered/fluid disordered phase
pattern (that is characterized for the presence of circular do-
mains) changes to a single homogeneous phase (the fluid dis-
ordered phase), an effect that is not observed in NPSM. The
possible maintenance of a minor proportion of ordered-like
phase at temperatures a few degrees higher than 38 °C, such as
those reached in feverish states, as well as the correlation
between the structural features and the functional properties
of surfactant material remain to be established. For instance,
Inoue et al. (43) showed that lung compliance was consistent
with surfactant activity remaining intact above 42 °C. One
could speculate that possible regulatory mechanisms might
exist to compensate for environmental factors through modifi-
cation of the relative proportion in surfactant of certain key
species, with cholesterol a probable candidate.
Cholesterol has been reported to induce a compression-
driven remixing behavior of coexisting phases in monolayers
composed of pulmonary surfactant organic extracts (8, 44).
Such behavior is probably related to the existence of a cholesterol-
dependent critical miscibility pressure in the two-dimensional
phase diagrams of simple model systems (45). A possible cor-
relation between the compression-driven remixing and the
temperature-induced melting of liquid ordered regions in sur-
factant bilayers observed in our experiments is difficult to
attain because there is not a clear link between lateral pressure
conditions in both systems (monolayers and bilayers). How-
ever, both phenomena are indicative of surfactant composition
being, perhaps, evolutionarily tuned to support a highly dy-
namic behavior of the surfactant operative structures.
It has been proposed that the regions in pulmonary surfac-
tant interfacial films accumulating unsaturated lipid species
to-bilayer transitions upon compression (during lung deflation)
and in forming structures that re-spread efficiently upon ex-
pansion (during inspiration) (1, 46). Similarly, we speculate
that phase separation observed in surfactant membranes may
be relevant to provide particular regions (with particular com-
position and lateral arrangement) where accumulated SP-B
and SP-C may initiate bilayer-to-monolayer molecular trans-
fer, a process that is essential in establishing the surface-active
surfactant film in the lung. This last assumption is supported
by the connection between the results of the spreading exper-
iments (Fig. 6) and the particular lateral structure of NPSM at
different cholesterol-to-phospholipid molar ratios. We propose
that in order to show optimal spreading behavior, the surfac-
tant must contain hydrophobic surfactant proteins while si-
multaneously having a lipid composition permitting the coex-
istence of fluid phases under physiological conditions. The role
of the phase coexistence-containing structures in other surface-
active or mechanical properties of surfactant still has to be
explored in detail, but the present data indicate that composi-
tion of clinical surfactants designed to treat respiratory pathol-
ogies such as acute respiratory distress syndrome could still be
largely improved.
Finally, our results also provide a clear example of the notion
that specialized regions with particular lateral packing prop-
erties coexist in natural membranes and control specific molec-
ular interactions in agreement with the raft hypothesis. The
raft hypothesis has been widely spread in recent years (13, 14)
and suggests that factors modulating lipid segregation and
domain dynamics may indirectly regulate specific membrane-
associated functions. There are remarkable aspects in this
mechanism as we learned from our experiments with pulmo-
nary surfactant membranes. First, the coexistence of defined
membrane regions with particular lateral arrangements is
triggered by the presence of a few key lipid species (cholesterol,
DPPC instead of sphingomyelin, as in the raft hypothesis, and
unsaturated phospholipids), independent of the compositional
complexity of the material. This last aspect could possibly be
connected with events at supramolecular levels such as the
mechanical properties of the membrane. Second, the composi-
tion of a particular membrane region displaying a particular
lateral structure (for example the liquid ordered or liquid dis-
ordered regions) would modulate specific molecular interac-
tions in the plane of the membrane (as seen in Fig. 5 with the
different distribution of the lipophilic probes between artificial
40722 Phase Separation in Native Surfactant Membranes
and native membranes displaying the coexistence of two fluid
phases), possibly being connected with events occurring at the
molecular level. Most interestingly, our results show that the
surfactant proteins are not participating substantially in trig-
gering the particular lateral structure observed in this partic-
ular membrane. Even though this is a new observation in the
framework of the raft hypothesis, additional experiments in
other membrane systems are required to generalize this sug-
gestion further.
CONCLUSIONS
We have obtained unambiguous evidence that at physiolog-
ical temperatures the lateral phase separation observed in
native pulmonary surfactant membranes is entirely dependent
on the concentration of a few key lipid species, with a partic-
ular involvement of cholesterol. Remarkably, this phenomenon
is independent of the presence of surfactant proteins. Choles-
terol extraction affects the spreading properties of the native
material suggesting an interesting correlation between the lat-
eral structure and one of the relevant physiological functions of
surfactant. Additionally, the dramatic lateral structure transi-
tion observed at 37.5 °C underlines the critical properties as-
sociated with surfactant full composition under physiological
3. Wright, J. R. (2003) J. Clin. Investig. 111, 1453–1455
4. Whitsett, J. A. & Weaver, T. E. (2002) N. Engl. J. Med. 347, 2141–2148
5. Perez-Gil, J. & Keough, K. M. (1998) Biochim. Biophys. Acta 1408, 203–217
6. Lewis, J. F. & Veldhuizen, R. (2003) Annu. Rev. Physiol. 65, 613– 642
7. Discher, B. M., Maloney, K. M., Schief, W. R., Jr., Grainger, D. W., Vogel, V. &
Hall, S. B. (1996) Biophys. J. 71, 2583–2590
8. Nag, K., Perez-Gil, J., Ruano, M. L., Worthman, L. A., Stewart, J., Casals, C.
& Keough, K. M. (1998) Biophys. J. 74, 2983–2995
9. Nag, K., Pao, J. S., Harbottle, R. R., Possmayer, F., Petersen, N. O. & Baga-
tolli, L. A. (2002) Biophys. J. 82, 2041–2051
10. Piknova, B., Schram, V. & Hall, S. B. (2002) Curr. Opin. Struct. Biol. 12,
487– 494
11. Orgeig, S. & Daniels, C. B. (2001) Comp. Biochem. Physiol. A Mol. Integr.
Physiol. 129, 75– 89
12. Casals, C., Arias-Diaz, J., Valino, F., Saenz, A., Garcia, C., Balibrea, J. L. &
Vara, E. (2003) Am. J. Physiol. 284, L466 –L472
13. Edidin, M. (2003) Annu. Rev. Biophys. Biomol. Struct. 32, 257–283
14. Simons, K. & Ikonen, E. (1997) Nature 387, 569 –572
15. Korlach, J., Schwille, P., Webb, W. W. & Feigenson, G. W. (1999) Proc. Natl.
Acad. Sci. U. S. A. 96, 8461– 8466
16. Bagatolli, L. A. & Gratton, E. (2000) Biophys. J. 78, 290 –305
17. Bagatolli, L. A. & Gratton, E. (2000) Biophys. J. 79, 434 – 447
18. Dietrich, C., Bagatolli, L. A., Volovyk, Z. N., Thompson, N. L., Levi, M.,
Jacobson, K. & Gratton, E. (2001) Biophys. J. 80, 1417–1428
19. Feigenson, G. W. & Buboltz, J. T. (2001) Biophys. J. 80, 2775–2788
20. Veatch, S. L. & Keller, S. L. (2002) Phys. Rev. Lett. 89, 268101(-1)–268101(-4)
21. Veatch, S. L. & Keller, S. L. (2003) Biophys. J. 85, 3074 –3083
22. Scherfeld, D., Kahya, N. & Schwille, P. (2003) Biophys. J. 85, 3758 –3768
23. Baumgart, T., Hess, S. T. & Webb, W. W. (2003) Nature 425, 821– 824
24. Casals, C., Herrera, L., Miguel, E., Garcia-Barreno, P. & Municio, A. M. (1989)
Biochim. Biophys. Acta 1003, 201–203
25. Rouser, G., Fkeischer, S. & Yamamoto, A. (1970) Lipids 5, 494 – 496

Downloaded from http://www.jbc.org/ by guest on May 18, 2014

conditions. Apparently, this material has been evolutionarily 26.
optimized to sustain a structure that is at the borderline of a
27.
lateral structure transition, likely contributing to essential lev-
els of mechanical plasticity. We believe that our results point to
28.
an entirely new perspective on the establishment of structure-
function relationships in pulmonary surfactant, which may 29.
significantly facilitate the search for new pulmonary surfactant
30.
materials effective in the treatment of pathologies such as
acute respiratory distress syndrome. In addition, the relevance 31.
of our results goes beyond the pulmonary surfactant field, into 32.
the general field of biomembranes. According to our results,
pulmonary surfactant could be one of the first membranous 33.
systems reported where the coexistence of specialized mem- 34.
brane domains exists as a structural basis for its function.
35.
Acknowledgments—We thank Dr. I. Plasencia for technical assist- 36.
ance with protein and lipid analysis of surfactant fractions and Dr. Lars 37.
Duelund for the assistance with differential scanning calorimetry ex- 38.
periments. We also thank Dr. David Jameson (University of Hawaii, 39.
Manoa), Dr. Kevin M. W. Keough (Memorial University of Newfound- 40.
land), Dr. Ole Mouritsen (University of Southern Denmark), Dr. Felix
Goñi (University of Basque Country), Dr. Enrico Gratton (University of 41.
Illinois, Urbana-Champaign), and Drs. Sandra Orgeig and Chris
Daniels (University of Adelaide) for their useful comments and sugges- 42.
tions on a critical reading of the manuscript. MEMPHYS-Center for
43.
Biomembrane Physics is supported by The Danish National Research 44.
Foundation.
45.
REFERENCES
1. Perez-Gil, J. (2002) Biol. Neonate 81, Suppl. 1, 6 –15 46.
2. Daniels, C. B. & Orgeig, S. (2003) News Physiol. Sci. 18, 151–157
Plasencia, I., Cruz, A., Lopez-Lacomba, J. L., Casals, C. & Perez-Gil, J. (2001)
Anal. Biochem. 296, 49 –56
Düzgünes, N., Bagatolli, L. A., Meers, P., Oh, Y. & Straubinger, R. M. (2003)
in Liposomes (Torchiling, V. P. & Weissig, V., eds) Vol. 2, pp. 105–147,
Oxford University Press, London
Angelova, M. I. & Dimitrov, D. S. (1986) Faraday Discuss. Chem. Soc. 81,
303–311
Angelova, M. I., Soléau, S., Meléard, P. H., Faucon, J. F. & Bothorel, P. (1992)
Colloid Polym. Sci. 89, 127–131
Discher, B. M., Maloney, K. M., Grainger, D. W., Sousa, C. A. & Hall, S. B.
(1999) Biochemistry 38, 374 –383
Perez-Gil, J., Cruz, A. & Casals, C. (1993) Biochim. Biophys. Acta 1168,
261–270
Schurch, S., Green, F. H. & Bachofen, H. (1998) Biochim. Biophys. Acta 1408,
180 –202
Nag, K., Taneva, S. G., Perez-Gil, J., Cruz, A. & Keough, K. M. (1997) Biophys.
J. 72, 2638 –2650
Morrow, I. C., Rea, S., Martin, S., Prior, I. A., Prohaska, R., Hancock, J. F.,
James, D. E. & Parton, R. G. (2002) J. Biol. Chem. 277, 48834 – 48841
Trauble, H., Eibl, H. & Sawada H. (1974) Naturwissenschaften 61, 344 –354
Yu, S. H. & Possmayer, F. (2003) J. Lipid Res. 44, 621– 629
Diemel, R. V., Snel, M. M., Van Golde, L. M., Putz, G., Haagsman, H. P. &
Batenburg, J. J. (2002) Biochemistry 41, 15007–15016
Taneva, S. & Keough, K. M. (1997) Biochemistry 36, 912–922
Yu, S. H. & Possmayer, F. (2001) J. Lipid Res. 42, 1421–1429
Daniels, C. B., Barr, H. A., Power, J. H. & Nicholas, T. E. (1990) Exp. Lung Res.
16, 435– 449
Langman, C., Orgeig, S. & Daniels, C. B. (1996) Am. J. Physiol. 271,
R437–R445
Orgeig, S., Daniels, C. B., Johnston, S. D. & Sullivan, L. C. (2003) Reprod.
Fertil. Dev. 15, 55–73
Inoue, H., Inoue, C. & Hildebrandt, J. (1982) J. Appl. Physiol. 53, 567–575
Discher, B. M., Schief, W. R., Vogel, V. & Hall, S. B. (1999) Biophys. J. 77,
2051–2061
McConnell, H. M. & Vrljic, M. (2003) Annu. Rev. Biophys. Biomol. Struct. 32,
469 – 492
Schief, W. R., Antia, M., Discher, B. M., Hall, S. B. & Vogel, V. (2003) Biophys.
J. 84, 3792–3806