Lamellar bodies form solid three-dimensional films at the respiratory air-liquid interface

Abstract

Pulmonary surfactant is essential for lung function. It is assembled, stored and secreted as particulate entities (lamellar body-like particles; LBPs). LBPs disintegrate when they contact an air-liquid interface, leading to an instantaneous spreading of material and a decline in surface tension. Here, we demonstrate that the film formed by the adsorbed material spontaneously segregate into distinct ordered and disordered lipid phase regions under unprecedented near-physiological conditions and, unlike natural surfactant purified from bronchoalveolar lavages, dynamically reorganized into highly viscous multilayer domains with complex three-dimensional topographies. Multilayer domains, in coexistence with liquid phases, showed a progressive stiffening and finally solidification, probably driven by a self-driven disassembly of LBPs from a sub-surface compartment. We conclude that surface film formation from LBPs is a highly dynamic and complex process, leading to a more elaborated scenario than that observed and predicted by models using reconstituted, lavaged, or fractionated preparations.

FIGURE 1.

Morphology of LBPs and their content of surfactant-associated proteins. A, transmission EM. LBPs appear as partially unraveled lamellated whorls, intermixed with vesicular structures and plane membrane patches. B, representative Western blots (n = 3). PS, purified surfactant; PP, purified proteins.

FIGURE 2.

Experimental and optical setups. A, top, chamber with the inverted air-liquid interface, bottom, enlarged view (dimensions not in scale). With this setup, exploration of interfacial phenomena is possible under defined (37 °C and 100% rH) conditions. B, optical setup combining synchronous dual emission (green and red) and light reflection imaging (see text; EM, emission; EX, excitation; LW, long, SW short wavelength).

FIGURE 3.

Early adsorption events. A, two different mechanisms of LBPs adsorption: Rupture (disaggregation) of LBPs when they contact an interface free of phospholipids (1st min) or at low surface coverage (2nd min), and association of LBPs with reflective surface structures at saturating conditions (after 3rd min). Simultaneous detection of Bodipy-PC and DiI fluorescence unmasks phospholipid phase organization, reflection microscopy maps interface topography, and reveals surface structures. Time of interface contact is denoted by the arrows, acquisition rate = 0.33 fps. B, time-dependent categorization of interfacial adsorption events: Disaggregating denotes LBPs with sudden loss of point-shaped fluorescence, associating LBPs those which retained a particulate appearance. C, disaggregating LBPs were further subdivided according to the velocity of disaggregation.

FIGURE 4.

Entire sequence of film formation by LBPs. A, enlarged view demonstrates optically flat (low reflectance) and highly reflective, light scattering regions. B, growth of reflective, light scattering structures with time (in % of total surface area occupancy). C, increase in overall and phase selective fluorescence. Surfactant accumulation was almost linear with a tendency to slow down during the last 10 min.

FIGURE 5.

Analysis of surface domains. A, exemplary binary image (right) derived from the original reflection image (left). Black denotes flat regions; white, light scattering regions. Binary images were used to separate dye intensities within the two domains. B, distribution/accumulation of Bodipy-PC and DiI in light scattering or flat regions. In flat regions, the two dyes almost overlap. C, fluorescence intensities (I^f_-bg) in light scattering (L.S.) regions in relation to those in flat regions demonstrate the partition of the dyes between the two domains. D, normalized intensities (see text) of DiI (nI^DiI_-bg) in flat or light scattering regions, divided by those of Bodipy-PC within the same domains. Plot of this ratio shows the varying proportion of the dyes within each region.

FIGURE 6.

Segregation of liquid ordered and expanded phases as revealed by Bodipy-PC and DiI. A, within flat regions at 15 min. B, distribution of Bodipy-PC and DiI with respect to the surface topography (reflection) at 30 min. Enlarged views are the areas within the white squares, black bar, 30 μm. C, appearance of the surface film originating from surfactant purified from lung lavages. Experimental conditions in B and C are identical.

FIGURE 7.

Line scan and defocus aberration measurements. A, Z-scans of LBPs-derived membranes stained with Bodipy-PC and DiI. Reference Z-level (0 μm) was set to the focal plane of the flat region. Thus, −2 and +2.5 μm indicate in-focus structures shifted by 2 μm toward the subphase or 2.5 μm toward the air. Lines indicate calculation of intensity profiles plotted in C and D. B, reflected light image of the same region. Focal plane was at 0 μm. C, intensity profiles at different focal planes of lines 1 and 2 in A illustrate loss of fluorescence contrast at “out-of-focus” Z-levels. The focal plane of line 1, for example, is 2.5 μm above the level of the flat interface (sharp intensity profile of gray line), whereas that of line 2 is 2 μm below (sharp intensity profile of black line). D, intensity profiles of Bodipy-PC and DiI along line 3. For the ease of comparison, the size of the images in A and B (30 × 30 μm) correspond to the black bar in the enlarged view of Fig. 6B.

FIGURE 8.

Stiffening of the surface coat. A, measurement of fluorescent bead mobility. Reflection images (top) illustrate the status of the interface at 0 and 60 min. Insets show the trajectories of surface-embedded fluorescent beads during 4 s exposure times used to analyze the covered distances. Left plot (mean ± S.D.; n = 24) reveals a slowdown of movements. Right plot, comparison at time 60 min with purified surfactant (PS) is shown in the bar chart (n = 6). B, a line crossing flat and light scattering regions along which FRAP was performed. Fluorescence recovery (%I^f) was measured in the indicated regions (n = 5) and purified surfactant (n = 6). C, microtip-induced surface mobility (see also Fig. 9). Tip movement was along dashed lines with a stop at the circles. Arrows indicate evoked movement of selected surface structures before and after tip movement.

FIGURE 9.

Estimation of local surface tension. A, puncture of the interface by a glass μtip enforces a fluid meniscus whose shape is depending on surface tension, contact angle, and gravitation (wetting is assumed to be complete). Meniscoid curvature can be analyzed by the microscope's image function upon epiillumination: Incident light from the objective is reflected to, or away from, the objective, depending on the angle of the surface. Below a critical incident angle (C.a.), light back reflection into the objective approaches zero (gray region). a and b illustrate 2 different menisci (as a result of different surface tensions), leading to the corresponding diameters a′, b′ of central dark image regions, the areas of which were used to estimate surface tension. B, line scans through the tip center (along dashed line in image) were used to define a threshold level (gray dotted line). Pixels with intensities below were used to calculate the central dark area (indicated by the bright circle) of diameter b′. According to A, a decrease in surface tension (e.g. by LBPs) leads an increase in the area below the threshold (a′). C, evaluation and results of the method, using the calculated dark areas and different surfactants, all at 37 °C and 100% rH. All surfactants yielded statistically different values (p ≤ 0.025) except DPPC compared with purified surfactant. Different concentrations within each group were not different. D, example of a measurement with LBPs. Top, only flat regions were used for C. Bottom, light scattering structures resisted penetration. They were lifted (here: 10 μm), leading to an asymmetric shading (meniscus).