Atomic Force Microscopy Imaging of Adsorbed Pulmonary Surfactant Films

Abstract

Pulmonary surfactant (PS) is a lipid-protein complex that adsorbs to the air-water surface of the lung as a thin film. Previous studies have suggested that the adsorbed PS film is composed of an interfacial monolayer, plus a functionally attached vesicular complex, called the surface-associated surfactant reservoir. However, direct visualization of the lateral structure and morphology of adsorbed PS films using atomic force microscopy (AFM) has been proven to be technically challenging. To date, all AFM studies of the PS film have relied on the model of Langmuir monolayers. Here, we showed the first, to our knowledge, AFM imaging of adsorbed PS films under physiologically relevant conditions using a novel, to our knowledge, experimental methodology called constrained drop surfactometry. In conjunction with a series of methodological innovations, including subphase replacement, in situ Langmuir-Blodgett transfer, and real-time surface tension control using closed-loop axisymmetric drop shape analysis, constrained drop surfactometry allowed the study of lateral structure and topography of animal-derived natural PS films at physiologically relevant low surface tensions. Our data suggested that a nucleation-growth model is responsible for the adsorption-induced squeeze-out of the PS film, which likely results in an interfacial monolayer enriched in dipalmitoylphosphatidylcholine with the attached multilayered surface-associated surfactant reservoir. These findings were further supported by frequency-dependent measurements of surface dilational rheology. Our study provides novel, to our knowledge, biophysical insights into the understanding of the mechanisms by which the PS film attains low surface tensions and stabilizes the alveolar surface.

Significance

A pulmonary surfactant (PS) is a lipid-protein mixture that coats the air-water surface of the lung by adsorption. Although it is generally accepted that the adsorbed PS film at the air-water surface forms multilayers, it is technically challenging to visualize the lateral structure and topography of the adsorbed PS film. We have developed a novel, to our knowledge, experimental methodology called constrained drop surfactometry. In conjunction with atomic force microscopy, we studied the lateral structure, topography, and interfacial rheology of animal-derived natural PS films at physiologically relevant low surface tensions. Our data suggested an adsorption-induced squeeze-out, which provides novel, to our knowledge, biophysical insights into understanding the mechanism by which the PS film attains low surface tensions and stabilizes the alveolar surface.

Introduction

A pulmonary surfactant (PS) is a lipid-protein complex that covers the entire air-water surface of the lung as a thin film (1,2). It consists of ∼80% phospholipids, 5–10% neutral lipids (mainly cholesterol), and 5–10% surfactant-associated proteins (SPs) by weight. The main biophysical function of the PS film is to lower the alveolar surface tension to near zero to protect alveoli against collapse, thus maintaining a large surface area of the lung for gas exchange (3,4). In addition to surface tension reduction, the PS film also acts as the first-line host defense against inhaled particles and pathogens (5,6).

The innate PS film is formed at the alveolar surface via adsorption from surfactant vesicles synthesized by the alveolar type II cells (7). Electron microscopy observations have demonstrated that the innate PS film consists of multilayers of phospholipids (7,8). Using captive bubble surfactometry, Schürch et al. have convincingly demonstrated that adsorbed PS films contain considerably more highly surface active material than can be accounted for by a single monolayer, indicating such films must adopt a multilayered conformation (9). Many subsequent investigations showed that the adsorbed PS film is composed of an interfacial monolayer at the air-water surface plus a functionally attached vesicular complex (10, 11, 12, 13), i.e., the so-called surface-associated surfactant reservoir (SASR) (9). Further research suggested that the SASR is most likely stabilized by hydrophobic surfactant proteins (SP-B and SP-C) (14,15). The combined interfacial monolayer and SASR form a dynamically stable complex that favors interexchange of surface active materials between them during the highly dynamic inhalation and exhalation cycles of normal tidal breathing (16,17). The importance of the SASR in maintaining the normal biophysical function of the PS film was demonstrated in recent studies (18, 19, 20). It was found that impaired multilayer structure or the SASR led to deteriorated in vitro biophysical properties of the PS film and induced symptoms of acute respiratory failure in animal models with nanoparticle inhalation or intratracheal instillation (18, 19, 20).

Despite its success in visualizing the conformation of the PS film, electron microscopy suffers from a few limitations, such as complications in sample preparation (e.g., requirements of fixation, staining, and vacuum) and a lack of three-dimensional topographic information, which weaken its popularity in studying monolayers and biomembranes. Atomic force microscopy (AFM), on the other hand, has been proven to be an ideal method for visualizing the lateral structure and topography of the PS film (21, 22, 23, 24, 25). AFM has helped obtain valuable biophysical knowledge about the PS film, such as phospholipid phase separation and transitions, lipid-protein interactions, nano-bio interactions, and the collapse mechanisms of the PS film (1,26).

However, to date, almost all AFM studies of the PS film rely on monolayers spread at the air-water surface with the aid of organic solvents. As illustrated in Fig. 1, such a spread film, often known as a Langmuir film, is fundamentally different from an adsorbed film, often known as a Gibbs film. This difficulty is mainly inherent to the use of the classical Langmuir trough, which has intrinsic technical limitations for studying adsorbed PS films (1,2), such as a low surface area/volume ratio that prevents rapid adsorption of the PSs, and technical difficulties of performing Langmuir-Blodgett (LB) transfer of the adsorbed PS film because LB transfer in general requires a clean liquid subphase. Otherwise, the solid substrate for LB transfer would be contaminated by surfactant vesicles in the subphase before the transfer. To the best of our knowledge, all existing AFM imaging of the PS films was performed on spread Langmuir films rather than on adsorbed PS films.

Figure 1.

Schematic of the (a) spread film and (b) adsorbed film. The spread film, often known as the Langmuir film, consists of insoluble surfactant molecules, which are usually formed by spreading, with the aid of an organic solvent, to the air-water surface via a microsyringe. After spreading, the Langmuir film usually assumes a monolayer conformation. The adsorbed film, often known as the Gibbs film, usually consists of water-soluble surfactant molecules, which are usually formed by molecular adsorption from the aqueous subphase to the air-water surface. Although largely composed of insoluble lipid molecules, the natural pulmonary surfactant (PS) film is formed by adsorption of lipid vesicles to the air-water surface. After adsorption, the PS film assumes a multilayer conformation consisting of an interfacial monolayer plus a surface-associated surfactant reservoir (SASR) that is functionally attached to the interfacial monolayer. To see this figure in color, go online.

Here, we present the first, to our knowledge, AFM imaging of adsorbed PS films under physiologically relevant conditions using a novel experimental methodology developed in our laboratory, called constrained drop surfactometry (CDS). CDS is a new generation of droplet-based surface tensiometry capable of high-fidelity biophysical simulations of the PS film under physiologically relevant conditions (18, 19, 20). We have developed a novel subphase replacement technique that allows washing and replacing the vesicular subphase of the PS suspension without disturbing the adsorbed interfacial monolayer and SASR. Together with an in situ LB transfer technique integrated with the CDS, the subphase replacement facilities AFM imaging of de novo adsorbed PS films at physiologically relevant low surface tensions, i.e., 5–20 mN/m. With these technical advances, we have studied two animal-derived clinical PSs, Infasurf and Curosurf. Differences were found for the lateral structure and topography of these two adsorbed PS films. Such differences in the film structure were further related to the interfacial rheology of these PS films, which was also determined with the CDS. Our study provides novel biophysical insights into the understanding of natural PSs and the mechanisms by which the PS films attain low surface tensions upon compression.

Materials and Methods

PSs

Infasurf was a gift from ONY Biotech (Amherst, NY). It is prepared from lung lavage of newborn calves with centrifugation and organic extraction. Infasurf contains all of the hydrophobic components of natural bovine surfactant, whereas hydrophilic surfactant proteins (SP-A and SP-D) were removed during the extraction process. Curosurf was donated by Cornerstone Therapeutics (Cary, NC). It is a modified animal PSs prepared from minced porcine lung tissue. In addition to organic extraction and centrifugation, an additional procedure for removing all neutral lipids by gel chromatography is involved in the manufacture of Curosurf. Both surfactants were stored at −20°C in sterilized vials with a total phospholipid concentration of 35 mg/mL for Infasurf and 76 mg/mL for Curosurf. They were diluted by a saline buffer (pH 7.0; 0.9% NaCl, 1.5 mM CaCl₂, and 2.5 mN HEPES) to a final concentration of 1 mg/mL on the day of the experiment.

CDS

CDS is a new, to our knowledge, generation of droplet-based tensiometry technique developed in our laboratory (18, 19, 20). As illustrated in Fig. 2, CDS uses the air-water surface of a sessile droplet (∼30 μL in volume, ∼5 mm in diameter, and ∼0.4 cm² in surface area) to accommodate the adsorbed PS film. The surfactant droplet is constrained on a carefully machined pedestal using its knife-sharp edge to prevent film leakage even at low surface tensions. The adsorbed surfactant film can be compressed and expanded periodically at physiologically relevant rates and compression ratios by controlling liquid flow out of and into the droplet using a motorized syringe. The surface tension and surface area of the PS film are determined simultaneously using newly developed closed-loop axisymmetric drop shape analysis (CL-ADSA) (27). Owing to system miniaturization, CDS enables high-fidelity simulation of physiological conditions, i.e., the core body temperature of 37°C and a relative humidity close to 100%, using an environmental control chamber.

Figure 2.

Schematic of the constrained drop surfactometry (CDS). CDS uses the air-water surface of a droplet to accommodate the adsorbed PS film. The droplet is constrained on a knife-sharp pedestal and is enclosed in an environmental control chamber that enables simulation of the physiological conditions. The surface tension and surface area of the PS film are determined simultaneously from the shape of the droplet using CL-ADSA. Subphase replacement is implemented with a coaxial pedestal connected to two motorized syringes, with one withdrawing the vesicle-containing subphase from the droplet and another one simultaneously injecting buffer into the droplet with the same volumetric rate. Consequently, the subphase replacement technique facilitates washing and removal of lipid vesicles from the droplet without disturbing the adsorbed PS film. The PS film is subsequently LB transferred from the droplet surface to a freshly peeled mica substrate for atomic force microscopy (AFM) imaging. To see this figure in color, go online.

Specifically, a 30-μL or 1 mg/mL Infasurf or Curosurf droplet was dispensed on a 5-mm CDS pedestal via a pipette. Immediately after forming the droplet, the surface tension was continuously recorded and found to be quickly (within seconds) reduced to an equilibrium value around 22–25 mN/m, indicating the rapid formation of the adsorbed PS film at the air-water surface of the droplet (1). To mimic respiration, the adsorbed PS film was compressed and expanded at 20 cycles per minute, corresponding to a compression rate of 13.3% of the area per second.

Subphase replacement and LB transfer

To transfer the adsorbed PS film for AFM imaging, as shown in Fig. 2, the subphase replacement was implemented using a coaxial CDS pedestal connected with two motorized syringes, with one withdrawing the phospholipid-vesicle-containing subphase from the droplet at a rate of 1 μL/s and the other one simultaneously injecting buffer into the droplet at the same rate. Consequently, phospholipid vesicles in the aqueous subphase (i.e., the droplet) were washed away without disturbing the adsorbed PS film at the air-water surface. After the subphase replacement, LB transfer of the adsorbed PS film was performed by first quickly inserting a freshly peeled mica sheet into the droplet followed by slowly lifting the mica across the air-water interface of the droplet at a rate of 1 mm/min. During the LB transfer process, the PS film was maintained at a constant surface pressure (±1.5 mN/m) for a prolonged period using CL-ADSA (27,28). All measurements were conducted at 37°C for at least three times. The deposition ratio of the LB transfer, defined as the ratio between the lost area of the surfactant film during the LB transfer and the total surface area of the mica sheet (29), was estimated to be 1.24 ± 0.18, which indicates a complete transfer of the surfactant film from the air-water surface to the mica surface.

AFM

The lateral structure and topography of the adsorbed PS films were imaged using an Innova AFM (Bruker, Santa Barbara, CA). Samples were scanned in the air using the tapping mode with a silicon cantilever of the spring constant 42 N/m and a resonance frequency of 300 kHz. Images were taken at multiple locations to ensure the reproducibility. Lateral structures and topography of the samples were analyzed using NanoScope Analysis (version 1.5).

Surface dilational rheology

Detailed experimental protocols for determining the surface dilational modulus $E=\left(d\gamma /dlnA\right)$ of the adsorbed PS film using CDS can be found elsewhere (30,31). Briefly, the surface area of the de novo adsorbed PS film at equilibrium surface tension (γ_e) after subphase replacement was oscillated in a sinusoidal waveform, with frequencies of 0.01, 0.1, and 1 Hz and an amplitude of 10% of the initial surface area, using CL-ADSA. The surface tension response to the surface area oscillation was recorded as the output and was compared against the surface area oscillation waveform as the input. The elastic (E_r) and viscous (E_i) components of the surface dilational modulus were determined from the phase shaft (φ) between the input and output waveforms and from the oscillation amplitudes of the surface area and the surface tension. The loss tangent angle (tanφ) was calculated as the ratio between the viscous modulus and the elastic modulus (E_i/E_r). Spread dipalmitoylphosphatidylcholine (DPPC) monolayers were also studied as a reference. All measurements were carried out at 37°C for at least three times.

Results

Subphase replacement of a dye: proof of feasibility

We first demonstrate the feasibility of the subphase replacement technique integrated into the CDS. Fig. S1 shows the washing and replacement of 1 mg/mL royal blue dye from a 45 μL droplet by simultaneously withdrawing the dye solution and replacing it with an equal amount of pure water at the rate of 1 μL/s. It can be seen that the blue dye in the droplet was washed away with pure water fourfold (4×) of its original volume (i.e., a total of 180 μL) after 180 s. During the entire subphase replacement process, the surface tension, surface area, and volume of the droplet remained constant. A video clip of the subphase replacement process can be found in the Supporting Materials and Methods.

Subphase replacement of Infasurf

We then performed the subphase replacement on a 30-μL droplet of 1 mg/mL Infasurf, using one-, two-, three-, four-, seven-, and 10-fold replacement volumes, respectively. As shown in Fig. 3, surface tension, surface area, and volume of the Infasurf droplet remained constant for all subphase replacement volumes, indicating the integrity of the interfacial surfactant monolayer at the air-water surface. However, as shown in Fig. 4, dynamic surface activity of Infasurf after 10-fold subphase replacement shows significant deterioration in comparison with the de novo adsorbed Infasurf film. The dynamic surface activity was determined at physiologically relevant conditions, i.e., 37°C, 100% relative humidity, and less than 30% compression ratio at a frequency of 20 cycles/min to mimic the exhalation and inhalation cycles during normal tidal breathing. These results, therefore, indicate that the excessive washing at 10-fold volume may impair the SASR, although not the integrity of the interfacial monolayer. At all other replacing volumes from onefold to up to sevenfold, the Infasurf film after the subphase replacement maintains a similar surface activity as that of the de novo adsorbed Infasurf film, indicating a relatively intact SASR. Reproducibility of Figs. 3 and 4 can be found in Figs. S2 and S3.

Figure 3.

Subphase replacement of a 30-μL, 1 mg/mL Infasurf droplet with various replacing volumes from onefold (1×) to 10-fold (10×) of the original droplet volume at a rate of 1 μL/s. It can be seen that during the subphase replacement process, the surface tension, surface area, and volume of the droplet remained unchanged, indicating the integrity of the interfacial surfactant monolayer at the air-water surface. To see this figure in color, go online.

Figure 4.

Comparison of the dynamic surface activity of the subphase-replaced Infasurf film with various replacing volumes from onefold (1×) to 10-fold (10×) of the original droplet volume. With onefold to up to sevenfold replacing volumes, the subphase-replaced Infasurf film maintained the surface activity of the de novo adsorbed Infasurf, indicating a relatively intact SASR. However, with a 10-fold replacement volume, the subphase-replaced Infasurf film showed impaired surface activity, indicating that the excessive washing damaged the SASR. Dynamic surface activity was determined with CDS at 20 cycles per minute, 30% compression ratio, under physiologically relevant conditions, i.e., 37°C and 100% relative humidity. To see this figure in color, go online.

Optimization of the subphase replacement technique for adsorbed Infasurf film

Fig. 5 shows the lateral structure and topography of the Infasurf film after subphase replacement with onefold to 10-fold replacing volumes. Reproducibility of these AFM images can be found in Figs. S4–S9. Fig. 5 also shows the quantified height analysis of the AFM images at various replacing volumes. It can be seen that at onefold and twofold washing, the AFM images show isolated large aggregates higher than 100 nm, which are most likely surfactant vesicular residuals adsorbed to the solid substrate before LB transfer, thus indicating inadequate washing to remove all surfactant vesicles from the subphase. At sevenfold and 10-fold washing, on the other hand, the AFM images show a rather flat, mostly monolayer structure with isolated individual protrusions of 5–14 nm in height. Especially at 10-fold washing, only few isolated protrusions can be found, indicating that the excessive washing process significantly impairs the SASR. This AFM observation of the lateral structure and topography is in line with the compromised dynamic surface activity found at 10-fold washing (Fig. 4).

Figure 5.

Lateral structure and topography of the subphase-replaced Infasurf film with various replacing volumes from onefold (1×) to 10-fold (10×) of the original droplet volume. Also shown is a quantified height analysis of the subphase-replaced Infasurf film with various replacing volumes. All AFM images have the same scanning size of 20 × 20 μm. Images in the first two columns have a z range of 100 nm, whereas the rest images have a z range of 20 nm. It is found that threefold and fourfold replacement volumes are the optimized volumes because they efficiently remove the vesicles from the droplet and, meanwhile, preserve the SASR of the adsorbed Infasurf film. To see this figure in color, go online.

At threefold and fourfold washing, the AFM images show consistent lateral structure and topography with uniformly distributed multilayer protrusions of 20–28 nm in height. Given the thickness of a fully hydrated phospholipid bilayers to be around 4 nm (32), these AFM images reveal SASR of five to seven stacked bilayers, in good agreement with previously reported electron microscopy observations (8) and AFM imaging of compressed PS monolayers to its equilibrium spreading surface pressure of around 50 mN/m (24,25), which is equivalent to the equilibrium surface tension (γ_e) of adsorbed natural PSs at around 22 mN/m. These AFM observations, together with the surface activity measurements shown in Fig. 4, therefore collectively indicate that an optimal washing procedure exists for the subphase replacement technique to balance sufficient removal of the surfactant vesicles from the subphase and efficient preservation of the functional SASR structures. It should be noted that in addition to the volume of the subphase replacement, the rate of the subphase replacement was also studied and found to have a negligible effect on the lateral structure of the adsorbed Infasurf film (Fig. S10).

Adsorbed Infasurf and Curosurf films at physiologically relevant low surface tensions

Fig. 6 shows the lateral structure and topography of adsorbed Infasurf (Fig. 6, a–d) and Curosurf (Fig. 6, e–h) films at physiologically relevant low surface tensions. Optimization of the subphase replacement technique for Curosurf can be found in Fig. S11. The adsorbed PS films at three physiologically relevant low surface tensions (i.e., 15, 10, and 5 mN/m) were prepared by compressing the subphase-replaced PS films to the target surface tensions at a physiologically relevant rate of 13.3% of the area per second, followed by LB transfer at the controlled surface tensions (±1.5 mN/m) using CL-ADSA. Fig. S12 shows a typical result of LB transfer of Curosurf films at these low surface tensions. Figs. S13 and S14 include additional AFM images that show the reproducibility of the lateral structure and topography.

Figure 6.

Lateral structure and topography of the adsorbed Infasurf (a–d) and Curosurf (e–h) films within the physiologically relevant low surface tension range, including the equilibrium surface tension (γ_e) around 22–25 mN/m and three surface tensions lower than γ_e, i.e., 15, 10, and 5 mN/m, respectively. All AFM images shown in the two middle rows have the same scanning area of 20 × 20 μm and a z range of 20 nm. Images in the top and bottom rows show the three-dimensional renderings of the zoomed-in AFM images. White arrows point to the locations where the height measurements in the AFM images were determined. To see this figure in color, go online.

As shown in Fig. 6 a, right after the de novo adsorption, i.e., at the equilibrium surface tension (γ_e) of around 22 mN/m, the Infasurf film shows uniformly distributed isolated individual multilayer protrusions as high as 28 nm, corresponding to seven stacked phospholipid bilayers. Upon further reducing the surface tension to 15 mN/m (Fig. 6 b), the multilayer protrusions are enlarged in the lateral dimension, but not significantly in height. At a surface tension of 10 mN/m (Fig. 6 c), the originally isolated individual multilayer protrusions are compacted into largely continuous multilayer structures covering a large fraction of the Infasurf film. At the surface tension 5 mN/m (Fig. 6 d), these continuous multilayer structures further overlap to form large folded structures indicated by steep increments of around 20 nm in height. Such large-scale folded structures are very similar to the multilayered SASR found with the innate PS film using electron microscopy (8).

The adsorbed Curosurf film shows a lateral structure and topography distinctly different from those of Infasurf. As shown in Fig. 6 e, compared with Infasurf, the de novo adsorbed Curosurf film shows only a few but significantly larger multilayer protrusions in the lateral dimension. When the film is compressed to a surface tension of 15 mN/m (Fig. 6 f), more multilayer protrusions appear, but their heights remain relatively unchanged. At a surface tension of 10 mN/m (Fig. 6 g), the multilayer protrusions merge into laterally large domains more than 10 μm in size. When the adsorbed Curosurf film is compressed to a very low surface tension of 5 mN/m (Fig. 6 h), large foldings of the surfactant film, as indicated by steep increments of ∼16 nm in height, are observed. To better understand the different lateral structures of adsorbed Infasurf and Curosurf films under physiologically relevant low surface tensions, next we studied the interfacial rheological properties of these two adsorbed PS films.

Surface dilational rheology of adsorbed PS films

Fig. 7 shows the elastic and storage moduli (E_r), viscous and loss moduli (E_i), and the loss tangent (tanφ = E_i/E_r) of the subphase-replaced Infasurf and Curosurf films. A DPPC monolayer at γ_e was also studied as a reference. For each film, the surface dilational rheology was studied at three frequencies of 0.01, 0.1, and 1 Hz, respectively. Detailed procedures of determining the interfacial rheological properties can be found in Figs. S15 and S16.

Figure 7.

Surface dilational rheological properties of adsorbed Infasurf and Curosurf films. These rheological properties include (a) the elastic and storage moduli (E_r), (b) the viscous and loss moduli (E_i), and (c) the loss tangent (tan φ = E_i/E_r). All rheological properties were determined at 37°C and the equilibrium surface tension (γ_e) around 22 mN/m, with three frequencies of film oscillation, i.e., 0.01, 0.1, and 1 Hz, respectively. The DPPC monolayer at γ_e was included as a reference for comparison.

It can be seen that the surface dilational rheological properties of the three films are not appreciably different from each other at the same frequency of oscillation, thus indicating that the interfacial monolayer of the adsorbed Infasurf and Curosurf may both be enriched in DPPC. Nevertheless, tanφ of adsorbed Curosurf films is significantly higher than that of Infasurf films, indicating that Curosurf forms a more viscous film than Infasurf. In addition, interfacial rheological properties of all three films demonstrate a clear frequency dependence. Although both E_r and E_i increase with the frequency of film oscillation, tanφ of these films decreases especially when the frequency is increased to the physiologically relevant value of 1 Hz.

Discussion

Because of its experimental simplicity, spread Langmuir monolayers have long been used as a standard model for studying the biophysics of PS films. Despite providing valuable biophysical knowledge of the PS film, especially on the quasiequilibrium phospholipid phase behavior and the physical chemistry of lipid-protein interactions, it is well accepted that the Langmuir monolayer model is incapable of mimicking the highly dynamic nonequilibrium biophysics of the adsorbed PS film under physiologically relevant conditions (3,8,33, 34, 35). As illustrated in Fig. 1, the spread Langmuir monolayer model differs from adsorbed PS films mainly in the lack of the multilayered SASR. Although PS films without a SASR (as in the case of spread monolayers) or with impaired SASR (as in the case of 10× replacement volume shown in Figs. 3 and 5) can still maintain low surface tensions under static or quasiequilibrium conditions, a functional SASR actively attached to the interfacial monolayer at the air-water surface appears to be necessary for maintaining the normal biophysical properties of the PS film during dynamic cycling under physiologically relevant conditions (Fig. 4).

Directly imaging an adsorbed PS film with SASR is technically challenging. Here, with a series of technological advances in conjunction with the development of CDS, including subphase replacement, in situ LB transfer from a droplet, and real-time surface tension measurements and control using CL-ADSA, we have performed the first, to our knowledge, AFM observation on adsorbed PS films. With the study of two animal-derived modified natural PSs, Infasurf and Curosurf, we have gained novel insights into the biophysical mechanisms of PS films.

Our data provide direct evidence about the multilayer structure of adsorbed PS films. After de novo adsorption, both Infasurf and Curosurf form multilayer structures that consist of an interfacial monolayer at the air-water surface plus a multilayered SASR. The specific lateral structure and topography of the SASR are different between Infasurf and Curosurf, most likely related to their different unsaturated phospholipid profiles as a result of their different animal sources. Another major compositional difference between these two surfactant preparations is cholesterol, which is known to affect the thermotropic melting profile of PS membranes (36). Although Infasurf contains 5–8 wt % cholesterol, Curosurf is essentially devoid of cholesterol (24,37). Cholesterol lowers the viscosity of phospholipid monolayers by promoting the formation of nanodomains (38), which enhance the mixing between tilted-condensed (TC) and liquid-expanded (LE) phases (22). Because it appears likely that the multilayer formation initiates at the domain boundaries (22, 23, 24, 25), this difference could explain why the SASR of Infasurf appears as uniformly distributed, isolated, laterally small protrusions, whereas the SASR of Curosurf appears as individual large protrusions (Fig. 6).

Interestingly, it is found that at physiologically relevant low surface tensions, i.e., surface tensions less than γ_e, the SASR structures are compacted, merged, and eventually folded at the very low surface tension of 5 mN/m for both surfactant preparations (Fig. 6). Although the lateral dimensions of the SASR increase with film compression, the height of the SASR does not change significantly. This finding is consistent with our previous observations of Infasurf and Curosurf using the quasistatic compressed Langmuir monolayer model (25). We found that multilayered protrusions of Infasurf and Curosurf grow in height only during the compression-driven plateau region (25). After passing the plateau, multilayers stop growing in height but only increase packing density in the lateral dimension as a consequence of surface area reduction. Such a squeeze-out behavior in a supersaturated Langmuir monolayer can be explained by a nucleation-growth model that predicts that the formation of a compression-driven multilayer from an interfacial monolayer above a critical surface pressure initiates from a single-step nucleation, followed by growth of the three-dimensional nuclei, and ends with overlapping of the growing nuclei, at which point the size of the nuclei becomes limited in all directions (39). Once the growth of the nuclei becomes restricted, the compression rate exceeds the relaxation rate of the film, thereby rendering the film a high metastability until ultimate film collapse at the collapse pressure. These data suggest that such a nucleation-growth model may also apply to adsorbed PS films, in which molecular adsorption to the air-water surface is equivalent to film compression in the Langmuir film model.

Consequently, our data suggest an adsorption-induced squeeze-out of the PS film, which results in an interfacial monolayer enriched in DPPC with the attached multilayered SASR made up of mainly fluid non-DPPC components. Although numerous biophysical studies have demonstrated the compression-induced squeeze-out (24,25), the adsorption-induced squeeze-out does not involve lateral compression of the PS film but only spontaneous adsorption of surfactant vesicles. It was found that after adsorption, an air bubble in a captive bubble surfactometer required less area reduction to reach very low surface tensions than estimated from the lipid composition of the surfactant. This was thought to be due to “selective adsorption” of DPPC at the air-liquid interface of the bubble (9,33), a process known to require surfactant proteins (40). This study suggests that DPPC is not selectively adsorbed but rather enriched by squeeze-out unsaturated phospholipids from the air-water surface during the spontaneous adsorption process. This adsorption-induced squeeze-out model is consistent with the autoradiographic studies of Yu and Possmayer (11), who found that the lipid composition of adsorbed natural surfactant films, including the interfacial monolayer plus the attached SASR, had no difference from that of the bulk surfactant in the aqueous phase.

Direct extrapolation of the in vitro AFM data to understanding lung physiology, however, should be taken with caution. The clinical surfactant preparations used here, i.e., Infasurf and Curosurf, lack SP-A, which is known to play a significant role in stabilizing the SASR in vivo (41). The adsorbed PS films studied here were formed at a relatively low surfactant concentration of 1 mg/mL, whereas the surfactant concentration in the alveolar fluid was estimated to be one or two orders of magnitude higher (1). Therefore, it is not unexpected that the SASR formed in vivo is more complex, thicker, and intrinsically more stable at high compression rates. Moreover, our AFM data do not allow us to distinguish the orientation of the SASR formation because the adsorbed PS film was LB transferred from the air-water surface onto a mica surface. It remains controversial whether the SASR orients toward the air side or the aqueous liquid side (42,43).

The adsorption-induced squeeze-out model is supported by the analysis of interfacial rheology. It was found that the rheological properties of de novo adsorbed Infasurf and Curosurf films are very close to those of the DPPC monolayer at γ_e, especially at 1 Hz (Fig. 7). Our methodological advances allow us to study the surface dilational rheological properties of de novo adsorbed PS films under physiologically relevant conditions, i.e., adsorbed films at γ_e of around 25 mN/m, body temperature of 37°C, and the frequency of oscillation at 0.1–1 Hz, covering the frequency range of normal tidal breathing. It is found that both the dilational elastic modulus (E_r) and viscous modulus (E_i) increase with increasing frequency from 0.01 to 1 Hz, whereas their ratio (tanφ = E_i/E_r) decreases with increasing frequency, thus indicating that the PS film is significantly more elastic than viscous under physiologically relevant conditions. Such a frequency dependency of the surface dilational rheology in the range between 0.1 and 1 Hz is in good agreement with what was recently reported by Bykov et al. (44) but appears to be somewhat different from that determined at a much lower frequency range. Using captive bubble surfactometry, Wüstneck et al. studied the dilational rheology of DPPC monolayers with 2 mol % SP-C at room temperature in a very low frequency range between 0.006 and 0.025 Hz (45). They found that although the elastic modulus was relatively independent of the frequency, the viscous modulus decreased significantly with increasing the frequency of oscillation. Therefore, tanφ still decreases with increasing frequency, consistent with what we found in the high frequency range (Fig. 7).

It is interesting to point out that the surface shear rheology of spread PS films showed a frequency dependency similar to that of the surface dilational rheology determined here. It was found that both the shear elastic and viscous moduli of DPPC-containing monolayers (46) and natural PS preparations (47,48) (including Infasurf and Curosurf) increase with frequency in the same frequency range as studied here. Dilational rheology and shear rheology represent different rheological properties of thin films (49). Although the dilational rheology studies changes in an area at a constant shape, the shear rheology concerns changes in shape at a constant area. In terms of PS film, the surface dilational rheology provides information on the composition of the PS film, which is important for understanding phase transitions and film collapse during normal tidal breathing. Surface shear rheology provides information on the stability of PS films, which is crucial to better understand the spreading and Marangoni flow of PS films (46, 47, 48,50). The finding that both dilational moduli and shear moduli of PS films increase with frequency may have implications in better understanding the biophysical stability of the PS films, thus contributing to the design of novel, to our knowledge, surfactant preparations for pulmonary drug delivery.

Conclusions

We have developed a novel, to our knowledge, experimental methodology called CDS. In conjunction with AFM, CDS allowed the study of lateral structure and topography of animal-derived clinical PS films at physiologically relevant low surface tensions. Our data suggested that a nucleation-growth model is responsible for adsorption-induced squeeze-out at the PS film, which likely results in an interfacial monolayer enriched in DPPC with an attached multilayered SASR. These findings were further supported by frequency-dependent measurements of surface dilational rheology. Our study has implications in understanding the mechanisms by which the PS film attains low surface tensions and stabilizes the alveolar surface.

Author Contributions

L.X. and Y.Y. carried out the experiments and data analysis. Y.Y.Z. designed the research and oversaw the experiments and analysis. Y.Y.Z. wrote the article. All authors discussed the results.

Editor: Roland Winter.

Supporting Material

Video S1. Proof of Feasibility of the Subphase Replacement Technique Using a 45-μL Droplet of Royal Blue Dye

It can be seen that the dye was washed away after fourfold (4X) subphase replacement volumes after 180s and the droplet remained unchanged during the subphase replacement process. The video is played at fivefold (5X) speed.