Am J Physiol Lung Cell Mol Physiol
280: L938–L946, 2001.
Deformation-induced lipid trafficking
in alveolar epithelial cells
NICHOLAS E. VLAHAKIS,1 MARK A. SCHROEDER,1
RICHARD E. PAGANO,1,2 AND ROLF D. HUBMAYR1,3
1
Thoracic Diseases Research Unit, Division of Pulmonary and Critical Care Medicine, Department
of Internal Medicine, and Departments of 2Biochemistry and Molecular Biology and 3Physiology
and Biophysics, Mayo Clinic and Foundation, Rochester, Minnesota 55905
Received 27 July 2000; accepted in final form 30 November 2000
mechanical ventilation; BODIPY lipids; plasma membrane;
lung injury; deforming stress
is the only proven treatment for
patients with acute respiratory distress syndrome (28).
In these patients, dependent alveolar lung units are
collapsed or edema filled, whereas nondependent units
are air filled (13). High tidal volume ventilation has
been used as a means to recruit dependent alveolar
units and thereby improve patient oxygenation. However, in aerated lung units, this leads to alveolar overdistension and, in turn, high lung tissue strains. In
animal models, it is readily apparent that high tidal
volume mechanical ventilation directly results in lung
injury, edema, and inflammation, referred to as ventilator-induced lung injury (VILI) (10, 17, 18). Morpho-
MECHANICAL VENTILATION
Address for reprint requests and other correspondence: N. E.
Vlahakis, Rm. 4-411 Alfred Bldg., Mayo Clinic, 200 First St. SW,
Rochester, MN 55905 (E-mail: [email protected]).
L938
logical analysis of the cells that compose the alveolus
reveals significant epithelial and endothelial cell injury
characterized by blebbing, intercellular and intracellular gap formation, and the resultant denudation of
the alveolar basement membrane (11, 12). The aim of
the present study was to investigate the effect of
basement membrane deformation on alveolar epithelial cell injury. More specifically, we sought to investigate the biological mechanisms used by the alveolar
epithelial cell to ensure the integrity of its plasma
membrane (PM).
The response of cells to mechanical forces can be
viewed very simply by considering the cell as a structural network (cytoskeleton) enveloped by a lipid bilayer (PM). It is the cytoskeleton that determines cell
shape and modulates its structure to coordinate shape
changes, whereas the PM responds to these cytoskeletal changes to maintain an intact and functional cellular barrier. The cytoskeleton is integrally associated
with the extracellular matrix or basement membrane
via focal adhesion complexes such that deformations of
the basement membrane are transmitted to the overlying cell through these connections. As lung volume
increases from functional residual capacity, the alveolar septum unfolds, and at volumes closer to total lung
capacity, the septum is placed under stress (3, 14).
Thus, with alveolar overdistension, there is a resultant
deformation of the basement membrane and a concurrent strain applied to the adherent epithelial cell. Once
deformed, the epithelial cell must increase its surface
area. This increase in surface area leads to PM stress
and, inevitably, stress failure unless the PM is able to
“grow.” As a means to adapt to this stress, the cell
might “unfold” the ruffles of its PM or the lipid bilayer
itself might stretch laterally. Although the PM possesses a high shear and bending capacity, it can accommodate lateral stretch of only 3–4%, and thus the
protective contribution of the elasticity of the membrane is limited (24). We hypothesized that in response
to basement membrane deformation, alveolar epithelial cells translocate lipid molecules to the PM as a
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
1040-0605/01 $5.00 Copyright © 2001 the American Physiological Society
http://www.ajplung.org
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
Vlahakis, Nicholas E., Mark A. Schroeder, Richard
E. Pagano, and Rolf D. Hubmayr. Deformation-induced
lipid trafficking in alveolar epithelial cells. Am J Physiol
Lung Cell Mol Physiol 280: L938–L946, 2001.—Mechanical
ventilation with a high tidal volume results in lung injury
that is characterized by blebbing and breaks both between
and through alveolar epithelial cells. We developed an in
vitro model to simulate ventilator-induced deformation of the
alveolar basement membrane and to investigate, in a direct
manner, epithelial cell responses to deforming forces. Taking
advantage of the novel fluorescent properties of BODIPY
lipids and the fluorescent dye FM1-43, we have shown that
mechanical deformation of alveolar epithelial cells results in
lipid transport to the plasma membrane. Deformation-induced lipid trafficking (DILT) was a vesicular process, rapid
in onset, and was associated with a large increase in cell
surface area. DILT could be demonstrated in all cells; however, only a small percentage of cells developed plasma membrane breaks that were reversible and nonlethal. Therefore,
DILT was not only involved in site-directed wound repair but
might also have served as a cytoprotective mechanism
against plasma membrane stress failure. This study suggests
that DILT is a regulatory mechanism for membrane trafficking in alveolar epithelia and provides a novel biological
framework within which to consider alveolar deformation
injury and repair.
DEFORMATION-INDUCED LIPID TRAFFICKING
METHODS
Cell culture. Human A549 cells (American Type Culture
Collection, Manassas, VA), passages 78–88, were grown on
six-well culture plates (BioFlex, Flexcell International, McKeesport, PA). The base of each culture plate consisted of a
flexible collagen I-impregnated silicoelastic membrane with a
surface area of 9.6 cm2. Ham’s F-12K containing L-glutamine
(2 mM), FCS (10%), and penicillin-streptomycin-amphotericin B (100 U/ml, 100 ␮g/ml, and 25 ␮g/ml, respectively;
Sigma, St. Louis, MO) was used as a growth medium. A549
cells were passaged and seeded at a density of 20,800 cells/
cm2 (200,000 cells/well) 48 h before each experiment, resulting in an 85–90% confluent monolayer of cells.
Cell strain device. A portable vacuum stretch device was
developed for cell stretching and imaging by confocal microscopy. The BioFlex wells were placed in a Teflon-like (Delrin)
rubber-sealed mold with six 1.5-cm-diameter loading posts
and then fixed to the confocal microscope movable stage
mount for easy imaging maneuverability (Fig. 1). The loading
posts were lubricated with Braycote grease 804 (Castrol,
Irvine, CA) to minimize friction between the membrane and
loading post. The membrane and, in turn, the attached cells
were then deformed from below by a portable vacuum device
connected to the Delrin mold. Cells could be visualized both
Fig. 1. Cell strain device for laser confocal microscopy (LCFM) imaging of alveolar epithelial cells. Diagram shows relationship to the
confocal X,Y stage and water-immersion lens. Photo at top, representative single, 1-␮m confocal images of cells demonstrating typically observed changes in cell cross-sectional area with deformation.
in their undeformed and deformed states with LCFM. A
single stretch of 7 or 25% was applied to the cells and held on
average for 1–2 min to allow for focusing and image acquisition.
Correlation of vacuum pressure to membrane strain amplitude. To determine the strain amplitude applied to the
BioFlex membranes by the portable vacuum device, five ink
marks in a cross configuration were imprinted on the membrane overlying the loading post. The membrane was imaged
with an analog video camera (charge-coupled device TRV82,
Sony, New York, NY) in both the undeformed and strained
states at vacuum pressures varying from 0 to 70 kPa. With a
PC-based digital frame grabber (Imaging Technology,
Woburn, MA), specific video frames were digitized. Individual marker positions were identified and saved with a user
interactive PC-based software package (Thoracic Diseases
Research Unit, Mayo Clinic, Rochester, MN). Percent membrane strain was calculated by tracking the changes in distance between all dot pair combinations and then averaging
these results. Thus stretch amplitude refers to the percent
change in the length of any line element in the membrane of
the well.
Cell imaging. Cells were imaged with an argon ion laser
scanning confocal microscope (Bio-Rad MRC 600, Hercules,
CA) mounted on an upright microscope (BH2, Olympus,
Melville, NY) immediately after membrane lipid labeling.
Optical sections (1 ␮m) were obtained with a ⫻40 waterimmersion objective lens (numerical aperture ⫽ 0.75) (Carl
Zeiss, Thornwood, NY) and by moving the stage in one
direction alone to eliminate backlash in the stepper motor.
Fluorescent labels were excited with blue [wavelength (␭) ⫽
488 nm] laser light, and emission wavelengths were collected
simultaneously by green (␭ ⫽ 520–560 nm) and red (␭ ⬎ 590
nm) filters (Omega Optical, Brattleboro, VT) for BODIPY
and by a single long band-pass filter block (␭ ⬎ 515 nm)
for 1,1-dioctadecyl-3,3,3⬘3⬘-tetramethylindocarbocyanine perchlorate (DiI), FM1-43, and FITC. At each optical section
plane, images were digitized at eight-bit resolution, averaged
over three scans with Kalman filtering to reduce background
noise, and stored in arrays of 768 ⫻ 512 pixels.
Cell surface area and fluorescence quantitation. Cell surface area, membrane fluorescence characteristics, and computer-generated three-dimensional images of cells were calculated with an image display and manipulation package
(ANALYZE, Mayo Foundation, Rochester, MN) (16). For surface area analysis, the entire cell was scanned; for fluorescence intensity measurements, a single image slice through
the middle of the cell was acquired. In both cases, a single cell
in both the undeformed and deformed states was scanned
and analyzed. For surface area, the scan images were reconstructed into three-dimensional images and median filtered
(5 ⫻ 5 ⫻ 5) to remove broad backgrounds and maintain small
cell features (7). Surface area measurements were verified
with established computer-based stereological techniques
(16). Seed points for regions of interest were utilized and then
fit to the cell by hand-drawn pixel intensity limitations.
Labeling of cell membranes with lipid probes. Three different fluorescent lipid labels were used for membrane labeling:
DiI, BODIPY-FL C5-lactosylceramide (LacCer), and FM1-43
(Molecular Probes, Eugene, OR). For surface area calculations, cells were washed with 1⫻ PBS and then incubated
with 0.1% DiI (in ethanol) at 37°C for 10 min to label the PM.
For lipid trafficking experiments, cell PMs were labeled with
BODIPY-LacCer by washing with ice-cold 1⫻ HEPES-buffered MEM (HMEM; without glucose) and incubation with
3 nM BODIPY-LacCer-defatted BSA complex (6) for 30 min
at 4°C. Cells were then washed again with ice-cold HMEM
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
further adaptive mechanism to facilitate membrane
growth and ultimately prevent membrane rupture.
To study this hypothesis, we used an in vitro cell
deformation system to reduce the complex in vivo situation and at the same time simulate alveolar basement membrane strain. With the use of laser confocal
microscopy (LCFM), the alveolar epithelial cell surface
area was found to increase as a result of substratum
deformation. In association with this morphological
change, we report that alveolar epithelial cell deformation induced lipid trafficking to the PM in a stretch
amplitude-dependent fashion. Surprisingly, cell deformation did not result in a large number of PM breaks,
suggesting that deformation-induced lipid trafficking
(DILT) might be an important mechanism not only for
accommodation of increases in cell surface area but
also in the maintenance of cell integrity.
L939
L940
DEFORMATION-INDUCED LIPID TRAFFICKING
RESULTS
Validation of membrane strain amplitude and uniformity. The degree of membrane strain produced by
the vacuum pump at increasing pressures and its reproducibility and uniformity were analyzed. As demonstrated in Fig. 2, this strain-pressure relationship
was nonlinear and varied on average by 2–3% at any
given pressure. Cell deformation experiments were
performed at 20 and 40 kPa (150 and 300 mmHg)
pressure, resulting in membrane strains of 7 ⫾ 2.3 and
25 ⫾ 2.8%, respectively. The values for percent membrane strain represent the mean results of strain values from each of 6 ink-marked pairs in each experiment (12 in total). Using multiple regression analysis
of variance, we determined that membrane strain was
uniform when analyzed across all 12 ink-marked pairs
(P ⫽ 0.7).
Cell surface area increases as a result of membrane
strain. To determine whether cell surface area increased after membrane deformation, the PMs of cells
were labeled with DiI and imaged by LCFM in the
undeformed and deformed states. Five separate cells
Fig. 2. Relationship between applied transmembrane pressure and
resultant membrane strain in the cell strain device. Membrane
strain was calculated from changes in distance between pairs of ink
marks printed on the deformable membrane. Each data point represents mean ⫾ SD strain value of 12 ink-marked pairs (6 separate
pairs from each experiment). Applied pressures of 20 kPa (150
mmHg; arrow at left) and 40 kPa (300 mmHg; arrow at right) were
used in all subsequent experiments.
were then analyzed after either 7 or 25% strain. As a
result of membrane deformation, cells were found to
flatten and elongate, with the greatest change in cell
cross-sectional area occurring in the basalmost portions of the cells (Fig. 3A). Figure 1 (top) shows a single
LCFM image of cells labeled with BODIPY and qualitatively demonstrates typical observed changes in
basal cell cross-sectional area after deformation. The
application of 7 or 25% membrane strain resulted in an
average increase (P ⬍ 0.01) in cell surface area of 18 ⫾
6 or 35 ⫾ 9%, respectively (Fig. 3B). Thus changes in
cell surface area were strain amplitude dependent.
Cell strain reduces PM lipid label concentration. In
these experiments, BODIPY-LacCer was used to label
the PM. The concentration of BODIPY label within the
PM can be quantitated by the red-to-green fluorescence
emission ratio (R/G). With decreasing molar concentrations in lipid membranes, BODIPY lipids exhibit a
concentration-dependent spectral shift in fluorescent
wavelength emission (from red to green) (21). The R/G
of each pixel within the PM and an average value of the
entire PM were measured for cells in single-image
slices in the undeformed and deformed states. The
effect of 25% stretch on the R/G of the PM is illustrated
in the representative frequency distribution of all PM
pixels in a single cross-sectional image of the cell (Fig.
4A). A leftward shift in the R/G distribution was apparent after 25% strain, representing a decrease in the
molar density of labeled lipid in the PM. The photo at
the top of Fig. 1 demonstrates this qualitatively. The
average decrease in the R/G of the PM of cells
stretched by 25% compared with their prestretched
state in five separate experiments (15 individual cells)
was 59%. Compared with the unstretched time-control
cell population, stretch resulted in a 3.2 times greater
decrease (P ⬍ 0.002) in the average R/G of the PM (Fig.
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
before LCFM imaging. Finally, the intracellular membranes
were labeled with 2 ␮M FM1-43 for 4 h at 37°C and then
washed with 1⫻ PBS before being imaged. In separate experiments, cells were ATP depleted by incubating them at
37°C for 30 min in the presence of 5 mM sodium azide
(Sigma) and 50 mM 2-deoxy-D-glucose (Sigma).
Fluorescence spectrophotometry. The FM1-43 fluorescence
(FFM1-43) in the supernatant from culture wells of A549 cells
that were undeformed or strained by 25% was measured in
an SLM 8000C spectrofluorometer (Urbana, IL). Cells were
loaded with FM1-43 as described in Labeling of cell membranes with lipid probes and washed three times with PBS.
Cells were exposed to no strain or to 25% strain, and the
supernatant from the culture wells was collected. Twenty
milliliters of supernatant were combined with 180 ml of 1%
SDS (Sigma) in a cuvette, and the fluorescence was measured
(␭ex ⫽ 488 nm; ␭em ⫽ 565 nm).
Determination of PM breaks. To determine whether cellular deformation was associated with breaks in the PM, cells
were incubated with FITC-dextran (Dx) (molecular mass ⫽ 4,
70, or 150 kDa; Fluka, Milwaukee, WI) for 10 min at 37°C
and then stretched using the same parameters as described
in Cell strain device. Cells with homogeneous intracellular
distribution of FITC-Dx were counted in five low-power (⫻40)
microscope fields in each culture well for each experiment
and tabulated as a percent of the total cells counted (between
200 and 300 cells in total). Six separate experiments were
performed for each Dx size.
Statistical analysis. All measurements are presented as
means ⫾ SD. Statistical comparisons between experimental
conditions were made with Student’s t-test for paired observations. Significance was assumed at P ⬍ 0.05 with respect to
a two-tailed probability distribution. For stretch analyses,
each cell served as its own static control. Quantification of
lipid trafficking was also compared against a time-control cell
population that was exposed to the same environmental
conditions including laser light exposure time. Analysis of
variance was used to determine 1) membrane strain uniformity resulting from applied transmembrane pressures and 2)
strain effect on percent injured cells across different molecular mass FITC-Dx molecules.
DEFORMATION-INDUCED LIPID TRAFFICKING
L941
Fig. 3. Deformation-induced changes in cell surface area. A: representative cross-sectional area-height graph of a single cell. Abscissa,
1-␮m LCFM image slices taken from the base to the apex of the cell.
After deformation, cells flattened and elongated, with the most
pronounced increase in cross-sectional area occurring in the basalmost portions of the cell. B: each data point is the change in cell
surface area of a single cell from the undeformed to the deformed
state. Strain increased cell surface area in an amplitude-dependent
fashion with cells exposed to a 7 or 25% strain, increasing their
surface area by 18 ⫾ 6 and 35 ⫾ 9%, respectively. P ⬍ 0.01.
4B); this decrease suggested trafficking of unlabeled
lipid to the PM.
Substratum deformation induces lipid trafficking to
the plasma membrane in a strain amplitude- and energy-dependent fashion. To directly measure whether
deformation resulted in trafficking of lipid molecules,
we fluorescently labeled intracellular lipid membranes
with the styryl lipid label FM1-43 (Fig. 5A). FM1-43 is
an amphiphilic molecule that is characterized by an
⬃50-fold loss in fluorescence when transferred from
the membrane bilayer to the aqueous phase (4). Thus
the change in whole cell fluorescence intensity after
stretch served as a measure of lipid trafficking to the
PM. The leftward shift in the frequency distribution of
pixel intensity for a representative single-image slice of
a cell labeled with FM1-43 after 25% strain is shown in
Fig. 5B. In three separate experiments, cells labeled
with FM1-43 were deformed by either 7 or 25% compared with unstretched time-control cells (15 separate
cells in each group). Deformation resulted in lipid
transport to the PM, and a DILT response was determined in which 25% strain resulted in a twofold
Fig. 4. Deformation-induced changes in plasma membrane (PM)
BODIPY concentration. A: representative frequency distribution
graph of BODIPY-lactosylceramide (LacCer) concentration [red-togreen ratio (R/G)] in the PM of a single cell before and after 25%
strain. Each data point is the percentage of PM pixels with a given
fluorescent R/G. There was a shift to the left in the R/G of the cell PM
after deformation, representing a decrease in the molar density of
lipid label within the PM. B: each data point is the mean percentage
decrease in PM BODIPY concentration of a single cell compared with
itself over time alone (time-control cells) or after 25% strain. *Cells
strained by 25% demonstrated a 3.2 times greater decrease in PM
label concentration compared with time-control cells (n ⫽ 5), P ⬍
0.002.
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
greater (P ⬍ 0.001) trafficking response than that in
cells with 7% strain (33 ⫾ 12.3 vs. 16 ⫾ 7.7%) and a
sevenfold greater (P ⬍ 0.001) trafficking response than
that in the time-control cells (33 ⫾ 12.3 vs. 4.9 ⫾ 3.5%;
Fig. 5C). After deformation, cells were found to exclude
trypan blue and could be recultured with normal morphology and growth characteristics over 48–120 h
(data not shown). In two further experiments, cells
were ATP depleted by incubation with sodium azide
and 2-deoxy-D-glucose. The DILT response was not
apparent in these ATP-depleted cells compared with
that in the time-control cell population (Fig. 5D), suggesting that DILT is an energy-dependent process.
To provide further evidence that FM1-43 was indeed
exocytosed as a result of substratum deformation, samples of supernatant from undeformed and 25% strained
culture wells were analyzed by fluorescence spectrofluorometry (two separate experiments; six wells each
from undeformed and 25% strained cells). The supernatant from deformed culture wells contained on average 13% more (P ⫽ 0.007) detectable FFM1-43 compared
with supernatant from undeformed culture wells.
Cell stretch induces a small number of PM breaks.
After a 10-min incubation at 37°C with FITC-Dx, the
distribution of intracellular fluorescence was found to
L942
DEFORMATION-INDUCED LIPID TRAFFICKING
be diffuse and homogeneous (representative of PM
breaks) or granular and heterogeneous (representative
of noninjurious endocytosis) (19). Cells were stretched
by the same parameters as in the previous experiments, and FITC-Dx was found to be homogeneously
distributed throughout the cell (Fig. 6A) in only a small
percentage (0.6–2.7%) of the cells overlying the middle
of the loading post (Fig. 6B). In five separate experiments at 25% strain, 2.8% of cells demonstrated PM
injury in the presence of 70-kDa FITC-Dx. This was 4.9
times greater (P ⬍ 0.02) than that in cells exposed to a
7% strain (Fig. 6B). This strain amplitude dependence
of PM injury was apparent after 25% strain for all Dx
sizes (P ⬍ 0.003). Seven percent strain did not result in
a significantly greater cell injury (P ⬎ 0.05) when
compared with that in unstrained cells. The degree of
PM injury was not dependent (P ⫽ 0.2) on the size of
Dx studied.
We also wished to determine whether the straininduced breaks were lethal to the cells and whether
they were reversible in nature. First, cells were recultured after deformation experiments, and those that
had taken up FITC-Dx were found to be adherent to
the culture plate and to have normal morphology 36–
72 h after reculturing. This suggests that the PM
injury was nonlethal. Second, to determine whether
the breaks were reversible, cells were incubated with
FITC-Dx immediately after the stretch was released
(n ⫽ 4 cells). The number of cells taking up FITC-Dx in
a homogeneous fashion was no different (P ⬎ 0.05)
when compared with uptake in unstrained cells, suggesting that the induced PM breaks were reversible.
DISCUSSION
We report a heretofore undescribed response of alveolar epithelial cells to deformation, namely, increased
lipid transport to the PM. This DILT response was
rapid in onset (observed within 60–90 s of substratum
strain) and vesicular in nature and may represent an
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
Fig. 5. Deformation-induced lipid trafficking (DILT) to the PM. A: representative confocal image of A549 cells
labeled with FM1-43 over 4 h at 37°C showing numerous labeled intracellular vesicles in all cells. B: representative
frequency distribution graph of the change in FM1-43 fluorescence (FFM1-43) of a single confocal image from a single
cell before deformation and during 25% strain. There was a shift to the left in the graph after deformation,
representing a decrease in total fluorescent intensity within the cell and thus exocytosis of FM1-43. C: mean
percent decrease in FFM1-43 intensity compared with itself over time alone (time control) or after 7 or 25% strain.
Bars, means ⫾ SD of all the cells within each group; n ⫽ 3 separate experiments. *P ⬍ 0.001 vs. time-control cell
population. **P ⬍ 0.001 vs. 7% substratum strain. D: DILT is an energy-dependent process that is abrogated in the
presence of ATP-depleting medium. Each data point represents the mean percent decrease in FFM1-43 intensity in
the presence of azide and 2-deoxy-D-glucose compared with itself over time alone (time control) or after 25% strain.
There was no significant difference between time-control and strained cells, P ⫽ 0.5.
DEFORMATION-INDUCED LIPID TRAFFICKING
important regulatory mechanism for membrane transport in alveolar epithelial cells. Cell deformation induced by substratum strain resulted in significant increases in overall cell surface area. However, in spite of
the significant changes in cell shape, the proportion of
cells sustaining PM breaks was small. These breaks
were nonlethal and reversible in nature, suggesting
that at least one important role of DILT is reparative,
site-directed lipid transport. However, DILT was observed in all cells, not just those sustaining PM breaks,
suggesting that DILT might also represent a cytoprotective response against PM stress failure.
The present study was performed with an in vitro
alveolar epithelial cell system and was established to
achieve two goals: 1) to simulate in vivo basement
membrane strain that occurs during mechanical ventilation-induced alveolar distension and 2) to allow
LCFM imaging of cells in the undeformed and deformed states. In achieving these goals, we were able to
reduce complex in vivo conditions to a meaningful
experimental setup in which direct biological consequences of definable mechanical strains could be accurately measured. We chose to study the effects of a
maximal substratum strain value of 25%. The rele-
vance of this strain amplitude in the setting of VILI
has been addressed previously (33). Briefly, despite
numerous studies (3, 20, 32) using different measurement techniques, in situ strain of alveolar epithelial
cells has never been precisely quantified. Most evidence suggests that the alveolus deforms little during
quiet breathing. However, there is reason to think that
injurious strains are imposed on the lung parenchyma
during mechanical ventilation with high tidal volumes
(10). Using the parenchymal marker technique, Rodarte and colleagues (25) measured the strains of ⬇1cm3 lung regions in intact dogs and reported values as
high as 40% during inspiratory capacity maneuvers
(25). Although these estimates cannot be directly extrapolated to basement membrane or alveolar wall
strain, they nevertheless place the 25% substratum
strain used in our present study in a biologically relevant context.
Type I epithelial cells cover ⬎90% of the alveolar
surface and are believed to be exposed to a greater
strain than type II cells, which are cuboidal in shape,
reside in alveolar corners, and might be sheltered from
deforming stresses (32). Type I cells might therefore be
a more relevant substrate for deformation-induced
lung injury research. Unfortunately, primary type I
cell culture systems have been incompletely characterized to date and are not readily available (22). Thus
A549 cells were used as a representative alveolar epithelial cell model. This epithelial cell line exhibits
intercellular adhesion molecule-1 surface receptor expression (a characteristic of the type I phenotype) (22,
36) and possesses functioning proinflammatory signaling pathways (33). Because A549 cells do not form
polarized monolayers and might not communicate with
each other through gap junctions, the biological relevance of our findings might be questioned. However,
the DILT responses of primary rat type II alveolar
epithelial cells grown for 5 days in culture were virtually identical to those of the A549 cells described in Fig.
5 (Vlahakis and Hubmayr, unpublished observations).
This finding supports the notion that A549 cells serve
as a representative model system for our experimental
questions.
By taking advantage of the unique fluorescent characteristics of BODIPY-labeled lipid analogs and of the
lipophilic dye FM1-43, we were able to demonstrate
that deformation induces lipid transport to the PM. To
control for the effects of temperature, time, and light
exposure on the membrane lipids of the cell, we compared the strained cell population with cell populations
treated in an identical fashion except for deformation
(time-control cells). Using BODIPY labeling of the PM,
we showed that deformation induced a decrease in the
molar concentration of lipid label within the PM. This
finding suggested that unlabeled endogenous lipids
were transported to the PM, thereby reducing the
molar density of the fluorescent lipid in the membrane.
Strain-induced unfolding of cell surface ruffles would
not have affected the BODIPY fluorescence spectrum
because unfolding does not alter the intermolecular
distance of BODIPY label within the plane of the lipid
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
Fig. 6. Deformation-induced PM breaks. A: representative LCFM
image of an epithelial cell strained by 25% demonstrating homogeneous uptake of 70-kDa FITC-dextran (Dx) in a single cell, thus
identifying breaks in its PM. Surrounding cells showed only their
normal autofluorescence pattern, suggesting no sustained PM
breaks. B: percentage of cells with PM breaks with the use of
different molecular weight FITC-Dx after no strain, 7% strain, or
25% strain. Only 25% strain (*) resulted in a significantly greater
number of cells with PM breaks when compared with no strain (F ⫽
21.19; *P ⬍ 0.001) and also resulted in more breaks when compared
with 7% strain (**P ⬍ 0.05). These results were not dependent on the
size of FITC-Dx that was used (F ⫽ 1.74; P ⫽ 0.18).
L943
L944
DEFORMATION-INDUCED LIPID TRAFFICKING
cells have been physically stressed, these molecules
might indeed be markers of PM disruptions rather
than cell death.
Although the percentage of cells with breaks was
found to be small, the potential biological relevance for
the lung might still be great. Even injury of a single cell
can be a sufficient trigger for the initiation of tissue
inflammation. Gap junction and local paracrine communication between epithelial cells would potentially
allow PM injury of a single cell to be magnified and
lung inflammation to be potentiated (27). Both alveolar
overdistension in vivo (30) and alveolar epithelial cell
deformation in vitro result in the production of numerous proteins including interleukin-8, one of the key
inflammatory molecules associated with VILI (23, 33).
As these protein molecules are “chaperoned” by associated lipid vesicles to the cell surface, PM surface area
might increase as a result. Thus DILT might simply be
a secondary phenomenon of protein expression or secretion by the cell. This point is underscored when
specifically considering the alveolar type II cell in
which surfactant exocytotic events occur regularly. Indeed, studies (2, 35) have shown that cell deformation
stimulates type II rat alveolar epithelial cells to release
surfactant. Therefore, the findings in our system that
recycling vesicles are transported to the PM as a result
of deformation speak to a more unique underlying
relevance for DILT in the alveolar epithelium other
than that of surfactant secretion alone in type II cells.
Our data using FFM1-43 support the notion that lipid
transport to the PM results from vesicular transport.
The rapidity of this response suggests that the cell
might either store vesicles in the immediate sub-PM
region or that vesicles early in other trafficking pathways might preferentially be targeted to the PM. Studies in which cells were deformed in the presence of
ATP-depleting media demonstrated that lipid trafficking was abrogated, suggesting that DILT is an energydependent process. The upstream transducer(s) of
DILT remains to be defined, in particular, the role of
the interaction of extracellular matrix (alveolar basement membrane) and integrins (of the alveolar epithelial cell). It is likely that the cytoskeleton (5, 34) and
focal adhesion complexes (8) are essential for transduction of these mechanical forces. Sheetz and Dai (26)
also found, in fibroblasts, that PM tension might play
an important role in transducing cell morphology, motility, and membrane trafficking within the cell. In
addition, they elegantly showed that cells maintain a
PM reservoir that serves to buffer mechanical stressinduced changes in PM tension. This reservoir increases with cytoskeletal disruption or increases in PM
tension (24). Although we did not measure PM tension,
our results do support the concept of a PM reservoir in
alveolar epithelial cells and provide direct evidence
that, in part, intracellular vesicles make up this reservoir and might be utilized in the setting of mechanical
deformation.
In conclusion, this study provides novel in vitro evidence that basement membrane deformation induces a
rapid, strain amplitude-dependent transport of lipids
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
bilayer (9). However, the BODIPY data could also be
interpreted as a selective loss of fluorescent label from
a physically strained PM. Therefore, to specifically
demonstrate lipid transport to the PM in association
with cell deformation, we used the lipophilic dye FM143. FM1-43 is known to label intracellular vesicles,
which “recycle” to the PM, and has been used extensively as a means to study exocytosis (4). Epithelial cell
deformation resulted in exocytosis of these labeled vesicles, and this process occurred in a strain amplitudedependent fashion. Thus these results provided direct
confirmation that cell deformation induces lipid transport to the PM.
Substratum deformation was also found to increase
epithelial cell surface area. This increase in surface
area provides further evidence for DILT and suggests
that the cell might grow PM to accommodate increases
in surface area. Based on lipid tether force-displacement data, it has been argued that fibroblasts and
neuronal cells maintain a PM “reservoir” consisting of
membrane ruffles that unfold to accommodate increases in cell surface area (24). Unfortunately, the
confocal microscope is unable to resolve subtle PM
invaginations and ruffles that are readily apparent on
electron micrographs of A549 cells (images not shown).
As a result, our reported surface area values are an
“apparent” measure of the true cell surface areas, and
their changes with deformation do not reflect the relative contributions of PM unfolding and lipid transport
to the PM. Nonetheless, these findings do support the
concept of a membrane reservoir and suggest that
alveolar epithelial cells use intracellular lipid vesicles
as a result of basement membrane deformation.
In contrast to DILT, which occurred in all deformed
cells, we found that substratum deformation resulted
in only a small percentage of cells with PM breaks,
even at the greatest strain (25%). This was surprising
in light of the large increases in cell surface area that
resulted from this deformation. We believe that this
finding lends further support to the adaptive role of
DILT in the accommodation of cell surface area. In
addition, the reversible and nonlethal nature of these
breaks further speaks to the cytoprotective role of
DILT and provides evidence for the first time that
rapid transport of vesicles to the site of PM injury is a
deformation-induced repair process in alveolar epithelial cells. The varying molecular size of FITC-Dx and
its homogeneous pattern of cellular uptake after cell
stress has been used by investigators as a marker of
PM breaks or “porosity” (19). It is important to note,
however, that a homogeneous pattern of FITC-Dx uptake depends on the nature of the induced PM breaks.
If the breaks are small in dimension, large in number,
and open for a short period of time, the cell might still
take up low molecular mass molecules, although not in
a homogeneous fashion. This raises an important issue
with regard to the validity of techniques that use the
cellular permeability of fluorescent and low molecular
mass (⬍1 kDa) molecules (such as ethidium homodimer and propidium iodide) to determine cell death
(31). Our results suggest that in experiments where
DEFORMATION-INDUCED LIPID TRAFFICKING
We thank Dr. G. C. Sieck and Dr. Y. S. Prakash for help with the
confocal microscopy, Jon J. Camp for assistance with image analysis,
and L. L. Oeltjenbruns for preparing the manuscript.
This work was supported by National Institutes of Health Grants
R01-HL-63178 and GM-22942, a Glaxo Wellcome Pulmonary Fellowship Research grant, and institutional support from the Clinician
Investigator Program and the Thoracic Diseases Research Unit
(Mayo Clinic).
REFERENCES
1. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes compared with traditional tidal
volumes for acute lung injury and the acute respiratory distress
syndrome. N Engl J Med 342: 1301–1308, 2000.
2. Ashino Y, Ying X, Dobbs LG, and Bhattacharya J. [Ca2⫹]i
oscillations regulate type II cell exocytosis in the pulmonary
alveolus. Am J Physiol Lung Cell Mol Physiol 279: L5–L13,
2000.
3. Bachofen H, Schurch S, Urbinelli M, and Weibel ER. Relations among alveolar surface tension, surface area, volume, and
recoil pressure. J Appl Physiol 62: 1878–1887, 1987.
4. Betz WJ, Mao F, and Bewick GS. Activity-dependent fluorescent staining and destaining of living verterbrate motor nerve
terminals. J Neurosci 12: 363–375, 1992.
5. Bi G-Q, Morris RL, Liao G, Alderton JM, Scholey JM, and
Steinhardt RA. Kinesin- and myosin-driven steps of vesicle
recruitment for Ca2⫹-regulated exocytosis. J Cell Biol 138: 999–
1008, 1997.
6. Chen C-S, Martin O, and Pagano RE. Changes in the spectral
properties of a plasma membrane lipid analog during the first
seconds of endocytosis in living cells. Biophys J 72: 37–50, 1997.
7. Chen H, Swedlow JR, Grote M, Sedat JW, and Agard DA.
The collection, processing, and display of digital three-dimensional images of biological specimens. In: Handbook of Biological
Confocal Microscopy (2nd ed.), edited by Pawley JB. New York:
Plenum, 1995, p. 201–203.
8. Chicurel ME, Singer RH, Meyer CJ, and Ingber DE. Integrin binding and mechanical tension induce movement of
mRNA and ribosomes to focal adhesions. Nature 392: 730–733,
1998.
9. Dahim M, Mizuno NK, and Brockman HL. A free volume
model for BODIPY fluorescence in monolayers (Abstract). Biophys J 78: 180A, 2000.
10. Dreyfuss D and Saumon G. Ventilator-induced lung injury:
lessons from experimental studies. Am J Respir Crit Care Med
157: 294–323, 1998.
11. Dreyfuss D, Soler P, Basset G, and Saumon G. High inflation pressure pulmonary edema: respective effects of high airway pressure, high tidal volume and positive end-expiratory
pressure. Am Rev Respir Dis 137: 1159–1164, 1988.
12. Fu Z, Costello ML, Tsukimoto K, Prediletto R, Elliott AR,
Mathieu-Costello O, and West JB. High lung volume increases stress failure in pulmonary capillaries. J Appl Physiol
73: 123–133, 1992.
13. Gattinoni L, D’Andrea L, Pelosi P, Vitale G, Pesenti A, and
Fumagalli R. Regional effects and mechanism of positive endexpiratory pressure in early adult respiratory distress syndrome. JAMA 269: 2122–2127, 1993.
14. Gil J, Bachofen H, Gehr P, and Weibel ER. Alveolar volumesurface area relation in air- and saline-filled lungs fixed by
vascular perfusion. J Appl Physiol 47: 990–1001, 1979.
15. Grembowicz KP, Sprague D, and McNeil PL. Temporary
disruption of the plasma membrane is required for c-fos expression in response to mechanical stress. Mol Biol Cell 10: 1247–
1257, 1999.
16. Hanson DP, Robb RA, Aharon S, Augustine KE, Cameron
BM, Camp JJ, Karwoski RA, Larson AG, Stacy MC, and
Workman EL. New software toolkits for comprehensive visualization and analysis of three-dimensional multimodal biomedical
images. JAMA 10: 229–230, 1997.
17. International Consensus Conference in Intensive Care
Medicine. Ventilator-associated lung injury in ARDS. Am J
Respir Crit Care Med 160: 2118–2124, 1999.
18. Lecuona E, Saldı́as F, Comellas A, Ridge K, Guerrero C,
and Sznajder JI. Ventilator-associated lung injury decreases
lung ability to clear edema in rats. Am J Respir Crit Care Med
159: 603–609, 1999.
19. McNeil PL and Steinhardt RA. Loss, restoration, and maintenance of plasma membrane integrity. J Cell Biol 137: 1–4,
1997.
20. Mercer RR, Laco JM, and Crapo JD. Three-dimensional
reconstruction of alveoli in the rat lung for pressure-volume
relationships. J Appl Physiol 62: 1480–1487, 1987.
21. Pagano RE, Martin OC, Kang HC, and Haugland RP. A
novel fluorescent ceramide analog for studying membrane traffic
in animal cells: accumulation at the Golgi apparatus results in
altered spectral properties of the sphingolipid precursor. J Cell
Biol 113: 1267–1279, 1991.
22. Paine R III and Simon RH. Expanding the frontiers of lung
biology through the creative use of alveolar epithelial cells in
culture. Am J Physiol Lung Cell Mol Physiol 270: L484–L486,
1996.
23. Pugin J, Dunn I, Jolliet P, Tassaux D, Magnenat J-L,
Nicod LP, and Chevrolet J-C. Activation of human macrophages by mechanical ventilation in vitro. Am J Physiol Lung
Cell Mol Physiol 275: L1040–L1050, 1998.
24. Raucher D and Sheetz MP. Characteristics of a membrane
reservoir buffering membrane tension. Biophys J 77: 1992–2002,
1999.
25. Rodarte JF, Hubmayr RD, Stamenovic D, and Walters BJ.
Regional lung strain in dogs during deflation from total lung
capacity. J Appl Physiol 58: 164–172, 1985.
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017
to the PM of alveolar epithelial cells. This DILT response serves, in part, to repair PM breaks in order to
maintain PM integrity and cell viability. At the same
time, DILT can be viewed as a cytoprotective mechanism against PM stress failure and represents an attractive target for pharmacotherapy of VILI. The recently concluded Acute Respiratory Distress Syndrome
Network Trial (1) conclusively showed that smaller
ventilator-delivered tidal volumes improve patient survival. In a number of animal models, mechanical ventilation with high tidal volumes has been shown to
raise vascular permeability and promote lung inflammation. PM stress failure of alveolar lining cells is a
plausible mechanotransduction event underlying these
cardinal manifestations of VILI. The inability of epithelial and endothelial cells to accommodate large
basement membrane strains can lead to cell injury that
manifests as gap formation within and across the alveolar-capillary barrier. The consequent leakage of
serum into the alveolar space impairs surfactant function and alters alveolar micromechanics. The accumulation of serum proteins can trigger phagocytosis and
proinflammatory signaling. Another potential inflammatory trigger is the calcium influx through transient
endothelial and epithelial PM wounds (29). Woundinginduced cell activation has been extensively studied in
other systems. It is one of many possible mechanisms
through which cell deformation can initiate nuclear
factor-␬B activation, proinflammatory gene transcription, and cytokine release (15, 19). Thus DILT represents a newly described regulatory mechanism for
membrane traffic in alveolar epithelium and provides a
novel biological framework within which to consider
alveolar deformation injury and repair.
L945
L946
DEFORMATION-INDUCED LIPID TRAFFICKING
26. Sheetz MP and Dai J. Modulation of membrane dynamics and
cell motility by membrane tension. Trends Cell Biol 6: 85–89,
1996.
27. Simon RH and Paine R III. Participation of alveolar epithelial
cells in lung inflammation. J Lab Clin Med 126: 108–118, 1995.
28. Slutsky AS. Mechanical ventilation. American College of Chest
Physicians’ Consensus Conference. Chest 104: 1833–1859, 1993.
29. Terasaki M, Miyake K, and McNeil PL. Large plasma membrane disruptions are rapidly resealed by Ca2⫹-dependent vesicle-vesicle fusion events. J Cell Biol 139: 63–74, 1997.
30. Tremblay L, Valenza F, Ribeiro SP, Li J, and Slutsky AS.
Injurious ventilatory strategies increase cytokines and c-fos
mRNA expression in an isolated rat lung model. J Clin Invest 99:
944–952, 1997.
31. Tschumperlin DJ and Margulies SS. Equibiaxial deformation-induced injury of alveolar epithelial cells in vitro. Am J
Physiol Lung Cell Mol Physiol 275: L1173–L1183, 1998.
32. Tschumperlin DJ and Margulies SS. Alveolar epithelial surface area-volume relationship in isolated rat lungs. J Appl
Physiol 86: 2026–2033, 1999.
33. Vlahakis NE, Schroeder MA, Limper AH, and Hubmayr
RD. Stretch induces cytokine release by alveolar epithelial cells
in vitro. Am J Physiol Lung Cell Mol Physiol 277: L167–L173,
1999.
34. Wang N, Butler JP, and Ingber DE. Mechanotransduction
across the cell surface and through the cytoskeleton. Science 260:
1124–1127, 1993.
35. Wirtz HR and Dobbs LG. Calcium mobilization and exocytosis
after one mechanical stretch of lung epithelial cells. Science 250:
1266–1269, 1990.
36. Yu ML and Limper AH. Pneumocystis carinii induces ICAM-1
expression in lung epithelial cells through a TNF-␣-mediated
mechanism. Am J Physiol Lung Cell Mol Physiol 273: L1103–
L1111, 1997.
Downloaded from http://ajplung.physiology.org/ by 10.220.33.3 on June 18, 2017