| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
HIGHLIGHTED TOPICS
Lung Growth and Repair
1Department of Experimental, Environmental Medicine and Biotechnology, and 2Department of Neuroscience and Biomedical Technologies, University of Milano-Bicocca, Monza 20052, Italy
Submitted 3 March 2004 ; accepted in final form 31 May 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
-tubulin and 
-tubulin). In endothelial cells, we observed a folding of the plasma membrane with an increase in cell surface area, a doubling of plasmalemma vesicular density, and an increase in cell volume. Minor morphological changes were observed in epithelial cells. Edema did not affect the total plasmalemma amount of PKC, -tubulin, and caveolin-1, but -tubulin and CD-31 increased. In edema, the distribution of these proteins changed between the detergent-resistant fraction of the plasma membrane (DRF, lipid microdomains) and the rest of the plasma membrane [high-density fractions (HDFs)]. PKC and tubulin isoforms shifted from the DRF to HDFs in edema, whereas caveolin-1 increased in DRF at the expense of a decrease in phosphorylated caveolin-1. The changes in cellular morphology and in plasma membrane composition suggest an early endothelial response to mechanical stimuli arising at the interstitial level subsequently to a modest (
5%) increase in extravascular water. pulmonary interstitial pressure; morphometry; air-blood barrier; mechanotransduction; plasma membrane proteins
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Pulmonary interstitial edema formation.
The right superior jugular vein was cannulated, and pulmonary interstitial edema was induced by infusing saline at a rate of 0.5 ml·kg–1·min–1 for 3 h. The experimental protocol was shown to cause a slow development of interstitial edema (12) due to a decrease in plasma colloid osmotic pressure that leads to a greater microvascular filtration rate; in addition, the increase in plasma volume (15%) also causes a greater lung perfusion and therefore a greater filtration area.
In situ lung perfusion: fixation for electron microscopy.
Three groups of animals were used for morphological and morphometric analyses: 1) animals killed immediately after anesthesia and tracheotomy (control; n = 2), 2) animals kept under anesthesia for 3 h (sham; n = 3), and 3) animals with mild interstitial edema (treated; n = 4). The chest was opened through a midsternal splitting incision to expose the pericardium; in rabbits, this allows the pleural sacs to remain intact, thus preserving the physiological lung expansion. The pericardium was opened, and the pulmonary artery was cannulated; the left atrium was sectioned to allow the drainage of blood and perfusate. Two reservoirs arranged in parallel and connected to the pulmonary artery were used for lung perfusion. They contained, respectively, saline (11.06 g NaCl/liter plus 3% dextran T-70 and 1,000 U heparin/dl, 350 mosM) and fixative (phosphate-buffered 2.5% glutaraldehyde plus 3% dextran T-70, 500 mosM, pH 7.4). The upper level of the liquids in the reservoirs was adjusted at a height of 15 cmH2O relative to left atrium level and was maintained constant during perfusion. Animals were killed by an overdose of anesthetic just before the perfusion procedure. The perfusion circuit was first primed with saline for 3 min until the outflow appeared cleared of blood cells. The circuit was then switched to fixative, which was allowed to flow for 25–30 min.
Transmission electron microscopy preparation and tissue sampling. The fixed control, sham, and treated right lungs were cut into five slices of equal thickness according to a stratified random sampling procedure (31). Five-seven small blocks were systematically obtained from each slice, immersed in 2.5% glutaraldehyde for 4 h at 4°C, and then processed for electron microscopy as previously described (4, 20). Two blocks were randomly selected from each slice and processed for morphometric analysis. A single ultrathin section, 60 nm thick, was obtained from each tissue block, mounted on uncoated 200-mesh copper grids, stained with uranyl acetate and lead citrate, and observed in a Zeiss EM900 electron microscope. Six electron micrographs were systematically obtained from each section at two sequential primary magnifications (x15,000 and x28,000, respectively) on photographic paper as positive reversals from 70-mm negative films. The highest magnification was used to better investigate the morphometry of TABB; to this aim, we examined a total of 75 fields from edematous lungs and 25 fields from control lungs, electronically acquired as positive reversals, and brought to a final magnification of x66,000 on the computer video screen.
Morphology of the TABB. TABB corresponds to regions where only a fused basement membrane separates endothelium and epithelium with no intervening cells and fibrillar matrix.
We used a multipurpose M168 grid (length of test line d = 0.174 µm). We counted the number of test points over endothelial (Pen) and epithelial cells and interstitium (Pint); point counting over a given compartment, relative to total point counting over the image, is proportional to the volume of the compartment (volume density) (31). Other techniques are available to estimate cell volume changes with good time resolution; however, these can only be used for studying cell populations or single cells (25). Using the same M168 grid, we also counted the number of intersections of test lines with luminal and interstitial surface of endothelial cells (Ien lum and Ien int, respectively) and of interstitial front of epithelial cells (Iep int). Intersection counting, for a given profile separating compartments, is proportional to the surface development of the profile (31). The mean arithmetic thickness of endothelial and interstitial compartments was obtained as 
en = (d x Pen)/[2 x (Ien lum + Ien int)] for endothelial cells and int = (d x Pint)/[2 x (Ien int + Iep int)] for the interstitial layer, respectively.
For a detailed morphometric study of the complex changes in shape of cells and their microenvironment in interstitial edema, we used a cycloidal C2 grid (8), which enables a more precise analysis.
Distribution of plasmalemmal vesicles.
Plasmalemmal vesicles (PVs) in endothelial and epithelial cells were identified by their morphology as being noncoated and 50–90 nm in diameter. We computed the numerical density of PVs on micrographs obtained at x15,000 brought to a final magnification of x36,000 in endothelial and epithelial cells [in µm–3, i.e., number of PVs per unit cell volume (Nv)], using a multipurpose test grid M168 (in this case, length of test line = 0.337 µm). Numerical density was obtained as Nv = number of PVs/unit volume x a correction factor given by (
+ T – 2h), where is the true mean diameter of the PVs [considered to average 70 nm, as commonly accepted in literature (6)], T is the thickness of the ultrathin sections (60 nm), and h is the depth by which a vesicle must penetrate the section before it is detected (5, 6, 31). We also computed the vesicular load (µm–2), i.e., number of PVs per unit cell surface profile. In this case, the micrographs were enlarged to a final magnification of x66,000, and data were acquired with the cycloidal C2 grid (8).
Chemicals.
The antibodies we used in this study were mouse monoclonal anti-caveolin-1 C2297 (1:1,000), anti-caveolin-1 phosphorylated C-91520 (1:1,000; Transduction Laboratories, Lexington, KY), mouse anti-PKC (1:200; Santa Cruz Biotechnology, Santa Cruz, CA), and mouse monoclonal anti-CD31 (1:4,000; DAKO), mouse monoclonal anti--tubulin and anti--tubulin (1:500; Sigma, St. Louis, MO).
Biochemical methods.
Samples were obtained from control animals (killed shortly after anesthesia and tracheotomy; n = 5) and from animals that received slow saline infusion through the right superior jugular vein (0.5 ml·kg–1·min–1) for 3 h to induce interstitial edema (n = 5). We then prepared the samples, as previously described: briefly, we perfused the lungs for 5 min at room temperature with mammalian Ringer solution (without calcium) containing nitroprusside (20 mg/ml). Nitroprusside is a donor of nitric oxide; however, this effect should be present in both control and treated animal samples. Thus the observed differences in membrane protein response when control animals are compared with treated animals should be due to the specific conditions caused by interstitial edema. After this, the lungs were flushed with 50 ml of solution 1 (0.25 M sucrose, 20 mM tricine, pH 7.4, and 40 µg/ml of the protease inhibitors aprotinin, chymostatin, leupeptin, and antipapain), excised from the chest, and immersed in ice-cold solution 1. The lung tissue was finely minced at 4°C and homogenated in solution 1 and then filtered sequentially through 53- and 30-µm filters. The homogenate was subjected to centrifugation (1,000 g for 10 min) at 4°C, and the supernatants were saved. The resulting pellet was resuspended in 3 ml of buffer and subjected again to centrifugation as above. The pooled supernatants were overlaid over 25 ml of 30% Percoll in buffer. After centrifugation with a SW28 rotor at 84,000 g for 45 min at 4°C, we collected a single membranous band (1 ml) readily visible at about two-thirds from the bottom of the tube. To reduce the volumes and concentrate the membranes, the bands were pelleted by first diluting the suspension threefold with PBS before centrifugation at 100,000 g for 10 min at 4°C. This membrane fraction was collected and called PMC and PME (plasma membranes control and plasma membranes edema, respectively).
Isolation of detergent-resistant fraction. The plasma membrane pellet was resuspended in 1 ml of MBS buffer (25 mM of MES buffer, pH 6.5, containing 150 mM NaCl, 1 mM PMSF, and 75 U/ml aprotinin), and we determined its protein content (BCA method). Next, we took a volume containing 4.5 mg of protein, a quantity required for each gradient procedure. To maintain a constant protein-to-detergent ratio in all experiments, we added MBS buffer containing Triton X-100 up to a volume of 2 ml to reach a final Triton X-100 concentration of 1%. This procedure was subjected to ice for 20 min to maintain the integrity of lipid rafts. Finally, the 2 ml were diluted with an equal volume of 80% (wt/vol) sucrose in MBS lacking Triton X-100 and placed at the bottom of a tube where a discontinuous sucrose concentration gradient was created (40, 30, and 5% sucrose, from bottom up) in MBS lacking Triton X-100. After centrifugation at 250,000 g for 18 h at 4°C with a TW-41 rotor (Beckman Instruments), 1-ml fractions were collected from the top of the gradient and submitted to further analyses. From now on, fraction 5 from the top will be referred as the detergent-resistant fraction (DRF), and fractions 9–12 were pooled and indicated as high-density fractions (HDFs).
Protein analysis. Aliquots of PMC and PME and all fractions collected from the gradient were submitted to TCA precipitation. The pellets, washed with acetone, were suspended in water, and protein quantity was determined by the BCA method (Sigma). Thereafter, 50 µg of PMC and PME and 10 µg of proteins collected from DRF and HDF pellets, respectively, were loaded on SDS-PAGE, 10% polyacrylamide gel, and submitted to electrophoresis. Subsequently, the proteins were transferred to membranes that were stained with Ponceau S to assess protein loading by densitometry (Bio-Rad 710 densitometry, program quantity one) (15, 24). To assess the adequacy of our loading-transfer technique, we performed densitometry of albumin loading (in the range of 5–15 µg) transferred to nitrocellulose stained with Ponceau S; the regression between densitometry and protein loading was equal to 0.5385x + 4.14 (R2 = 0.97). Furthermore, no significant difference in densitometry values was found on paired albumin loadings in the range of 5–15 µg. We compared on our samples the densitometry of the whole lane for protein loading obtained from total plasma membranes, DRF, and HDFs from control and treated animals.
Subsequently, the membranes were submitted to Western blotting. After blocking was completed, blots were incubated for 2 h with the primary antibody diluted in PBS-Tween 20-milk. For phosphorylated caveolin-1, the primary antibody was diluted in 10 mM Tris, pH 8.0, 50 mM NaCl, 0.2% Tween 20, 2% milk, and 1% BSA. Blots were then incubated for 2 h with horseradish peroxidase-conjugated anti-mouse and -goat IgG (5,000-fold to 10,000-fold diluted in PBS-Tween 20-milk). The protein samples were obtained from five control and five treated animals. Proteins were detected by enhanced chemiluminescence with the use of the SuperSignal detection kit (Pierce, Rockford, IL). We performed parallel immunoblot analyses of samples from one control and one treated animal for total plasma membrane, DRF, and HDF proteins. Immunoblot bands were analyzed by the Bio-Rad 710 densitometry.
Plasma protein concentration. In all groups of animals, blood samples were drawn to determine plasma protein concentration by optical refractometry (SPR-Atago, precision within 3%).
Statistical analysis. For morphometric analyses, primary data (point, line intersection, and vesicle counts) were summed over all the micrographs derived from each section, and the parameters were computed as the ratio of sums. The parameters were then averaged over the various section samples. Data are expressed as means ± SE. The significance of the differences among groups was determined with one-way ANOVA and t-test. To evaluate the relationships between morphometric characteristics, linear models were fitted on data in the natural or logarithmic and exponential transformation where appropriate. The linear regression parameters were estimated with the least-square method. This approach was also used for the estimate of the iso-shape curve.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Morphometry of the TABB. The low-magnification (x60) light microscopy images of Fig. 1 show the morphology of the lung in control areas and in interstitial edema. One can appreciate that in interstitial edema there is a fairly diffuse state of water imbibition spreading from the alveolar septa to the TABB.
|
|
|
|
|
|
|
|
|
|
0.
|
|
-Tubulin increased significantly in PMedema relative to PMcontrol. Furthermore, it appeared to migrate from DRF to HDF in edema as the DRF-to-HDF ratio decreased from 2.39 to 0.
-Tubulin did not change when edema plasma membrane was compared with control plasma membrane but shifted from DRF to HDF as the DRF-to-HDF ratio decreased from 0.9 to 0.18.
Caveolin-1 did not change when edema plasma membrane was compared with control plasma membrane; it was only present in DRF, and it increased significantly in this fraction in edema (Table 3). Phosphorylated caveolin-1 was not detectable in plasma membrane and was present in DRF control, and it decreased in this fraction in edema.
CD31 (also known as PECAM) increased significantly in plasma membranes in edema, reflecting its increase in the HDFs.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Recent data have clarified subcellular events in pulmonary cells in response to vascular stimuli induced by considerable elevations in capillary pressure (10, 30). In our experimental approach, the hydrostatic capillary pressure is essentially unchanged relative to normal (17), and the increase in shear stress does not exceed 30% (20). Conversely, because interstitial fluid pressure increases markedly due to matrix imbibition, from approximately –10 to 5 cmH2O (12), we may hypothesize that stimuli triggering a cellular response originate mainly on the interstitial rather than on the luminal front. This cellular response can be stimulated by either increased tissue pressure and/or the fragmentation of proteoglycans, which are known to regulate a number of dynamic cellular processes, including cell adhesion and cell-matrix interaction (18).
At large magnification, there are considerable differences in morphology of the TABB that are mostly restricted to the endothelial cells and the basement membrane (Fig. 3A). Indeed, when the relationships between the geometry of endothelial cells and their interstitial microenvironment are considered, various degrees of alterations could be described, possibly indicating a sequence of functional phases in the cellular response. Some endothelial cells appeared still similar to those under control conditions, namely flat and thin (Fig. 2B), whereas, for others, considerable changes in morphology were observed (Fig. 2, C and D), suggesting a difference in the temporal sequence for the development of cellular functional adaptations.
The increase in interstitial cell surface, due to folding of the plasma membrane, appears to be an early cellular response to developing interstitial edema, as it occurs for a negligible increase in basement membrane thickness (Fig. 4). The increase in the luminal-to-interstitial surface ratio of endothelial cells correlates to the increase in cytoplasm volume (Fig. 5) and reflects the formation of bulging processes toward the capillary lumen.
The increase in surface area in endothelial cells can only occur by creating new plasma membrane because these cells do not possess surface elements providing unfolding; this process requires lipid translocation from cytoplasm to cell surface and is fostered by increased fluidity of the plasma membrane, reflecting, in particular, the modifications of the phosphatidylcholine-to-phosphatidylethanolamine and cholesterol-to-phospholipid ratios (20).
-Tubulin and -tubulin are major subunits of microtubules whose association with plasma membrane occurs through hydrophobic interactions that are established by palmitoylation of a tubulin cystein residue (3). Palmitoylated tubulin was also found in lipid microdomains (22), and the shift of this molecule away from the DRF removes a membrane to cytoskeleton linkage, thus possibly allowing greater mobility of lipid microdomains and therefore greater deformation in these portions of plasma membrane.
CD31 is a mechanoresponsive cell-surface receptor, involved in control of microvascular permeability, and its increase was shown to occur in response to shear-induced perturbation of the endothelial plasma membrane at the luminal level (19). The present data demonstrate that an increase in CD31 may also occur in response to interstitial edema. The observed changes in PKC and in CD31 (both increase in HDFs in edema) may contribute to modulate the CD31/catenin association and therefore may be the physical link between cell surface and the nuclear envelope (19).
The increase in PVs per unit surface profile (vesicular load) was mostly observed in endothelial cells and correlates to the increase in cell surface profile (Fig. 7A). Conversely, no real correlation exists when the numerical density of PVs is plotted vs. cell volume (Fig. 7B) as it doubles for a negligible increase in volume, remaining thereafter essentially steady. It is known that caveolin-1, either in the phosphorylated or dephosphorylated form, affects the subcellular traffic of vesicles, corresponding, on morphological and biochemical grounds, to caveolae (1). Reducing caveolin-1 phosphorylation prevents the endoplasmic reticulum targeting of caveolin, favoring instead the shift of lipid microdomains to cellular surface to form vesicles (27). Therefore, our finding of an increase in the ratio of caveolin-1 to phosphorylated caveolin-1 from 2.5 to 11 in edema is in line with the increased vesicle formation occurring in the endothelial cells. PVs, besides representing a site for signaling proteins, are also involved in transcytosis. The increase in cytoplasm volume in endothelial cells could derive from the activation of plasma membrane-mechanosensitive ion channels (25) and could also correlate to the previously observed increase in AQP-1 in DRF (21). In fact, the increase in endothelial cell volume (and to a minor extent of epithelial cells) contributes to a remarkable reshuffling effect of extravascular water. We might attempt a functional interpretation of this fact considering that prevention of water accumulation in the interstitial space allows the protection of its integrity and in turn the avoidance of severe edema to develop; indeed, as long as the interstitial matrix maintains its integrity, the hydraulic interstitial pressure is positive enough to counteract microvascular filtration (14).
Mechanotransduction.
Because endothelial cells mostly respond to interstitial edema with modifications in volume and shape, a functional link between mechanical stress and protein changes is suggested. On the contrary, the morphology of the epithelial cells was less affected, possibly indicating that these cells are less sensitive to parenchymal stimuli. Endothelial cells are equipped with numerous receptors, such as CD31, that allow them to detect and to respond to mechanical forces. The cytoskeleton and other structural components have an established role in mechanotransduction, being able to transmit and modulate tension within the cell via focal adhesion sites, cellular junctions, and the extracellular matrix (7, 16). Endothelial cells display the highest surface-to-volume ratio in control conditions (Fig. 6A) because of their attachments to the extracellular matrix and to the neighboring cells and through the mechanical action of the cytoskeleton that keeps the cell in a highly deformed state. The tensional behavior of a "hard-wired" cytoskeleton (9) might put endothelial cells in a good position to respond promptly to mechanotransduction. Moreover, the rigidity of the interstitial tissue (12) adds efficiency to the cellular response because a small increase in interstitial volume results in a considerable increase in interstitial pressure. A cellular response could also depend on modifications of interstitial protein composition. Regardless of the mechanism, this paper provides evidence for a prompt endothelial cellular response to stimuli arising at interstitial level when microvascular filtration is increased. The cellular response does not imply a considerable change in total plasma membrane contents for membrane proteins that we analyzed but, instead, in a modification of their distribution between lipid microdomains and the rest of the plasma membrane. Cell swelling stimulates protein synthesis and gene expression (29); furthermore, many intracellular pathways, including the MAPK cascade, are triggered by mechanical stimuli via sequential phosphorylation, the activation of transcription factors, and gene expression.
In summary, the results of this study suggest that signs of cellular activation develop to various degrees, mostly in the endothelial cells; these cells may act as early sensors of interstitial fluid accumulation, as they respond to minor increases in extravascular water (not exceeding 5%), possibly to trigger a reparative process.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. C. Aird Phenotypic Heterogeneity of the Endothelium: I. Structure, Function, and Mechanisms Circ. Res., February 2, 2007; 100(2): 158 - 173. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. T. Yiming, K. Parthasarathi, A. C. Issekutz, and S. Bhattacharya Sequence of Endothelial Signaling during Lung Expansion Am. J. Respir. Cell Mol. Biol., December 1, 2005; 33(6): 549 - 554. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. E. Vlahakis and R. D. Hubmayr Cellular Stress Failure in Ventilator-injured Lungs Am. J. Respir. Crit. Care Med., June 15, 2005; 171(12): 1328 - 1342. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, Control data; 
, interstitial edema data. Units of volume are given by total number of points falling in each compartment.
