Langmuir−Blodgett Films Formed by Continuously ... - ACS Publications

4950

Langmuir 2007, 23, 4950-4958

Langmuir-Blodgett Films Formed by Continuously Varying Surface Pressure. Characterization by IR Spectroscopy and Epifluorescence Microscopy Lin Wang,† Antonio Cruz,‡ Carol R. Flach,† Jesu´s Pe´rez-Gil,*,‡ and Richard Mendelsohn*,† Department of Chemistry, Olson Hall, Newark College, Rutgers UniVersity, 73 Warren Street, Newark, New Jersey 07102, and Dept. Bioquı´mica, Fac. Biologı´a, UniVersidad Complutense de Madrid, 28040 Madrid, Spain ReceiVed October 26, 2006. In Final Form: February 6, 2007 Monolayer films of phospholipids at the air-water interface have been transferred to solid substrates under conditions of continuously varying surface pressure, an approach termed COVASP. The molecular and supramolecular properties of the film constituents have been characterized with two complementary techniques. IR spectroscopy was used to monitor chain conformation as a function of transfer surface pressure. Results were compared to those from Langmuir films determined directly at the A/W interface by IR reflection-absorption spectroscopy (IRRAS). The methylene stretching frequencies for both proteated and acyl chain perdeuterated 1,2-dipalmitoylphosphatidylcholine (DPPC and DPPC-d62) in the transferred molecules indicate that the phospholipids retain at least, in part, their surface pressuredependent chain-conformational order characteristics. The line widths of these modes are somewhat reduced, suggestive of slower rates of reorientational motion in the Langmuir-Blodgett (LB) films. Epifluorescence microscopy reveals a progressive condensation gradient, including nucleation and growth of probe-excluding condensed domains along the transfer line. DPPC condensation, observed along a single LB film, was qualitatively comparable to compressiondriven condensation as observed in situ or in conventional LB films transferred at constant pressures. However, condensation along the compression isotherm in COVASP-LB films was reduced by 15-20% as compared to films equilibrated at different constant pressures, probably the result of kinetic differences in equilibration processes. As a preliminary demonstration of the utility of this new approach, the monolayer f multilayer transition known to occur (Eur. Biophys. J. 2005, 34, 243) in a four-component model for pulmonary surfactant has been examined. IR parameters from both the lipid and the protein constituents of the film all indicate that the transition persists during the transfer process. This new approach for the study of transferred films will permit the efficient characterization of lipid-protein interactions and structural transitions occurring in pulmonary surfactant films subjected to dynamic compression.

Introduction Langmuir-Blodgett (LB) and Langmuir films of phospholipids are widely used as models for biological interfaces. For biophysical applications, these molecules exhibit interesting surface pressure-dependent polymorphism and domain formation at the air/water interface. For biological applications, phospholipid films are useful mimics of cell membranes and of the films known to form at the air/alveolar lining layer in the mammalian lung.1 Practical applications for lipid films span a variety of areas including molecular electronics and sensors.2 In addition, a variety of methods have been reported in which phospholipid molecules are grafted onto solid substrates to increase surface biocompatibility.3,4 Several physical techniques have been used to characterize the structure and homogeneity of both LB and Langmuir films at a variety of distance scales. For LB films, Brewster angle microscopy and fluorescence microscopy are useful for visualizing * To whom correspondence should be addressed. Telephone: (973) 3535613 (R.M.); (34)-91-3944994 (J.P.-G.). Fax: (973) 353-1264 (R.M.); (34)91-3944672 (J.P.-G.). E-mail: [email protected] (R.M.); [email protected] (J.P.-G.). † Rutgers University. ‡ Universidad Complutense de Madrid. (1) Wu¨stneck, R.; Perez-Gil, J.; Cruz, A.; Fainerman, V. B.; Pison, U. AdV. Colloid Interface Sci. 2005, 117, 33. (2) Ulman, A. An Introduction to Ultrathin Organic Films: From LangmuirBlodgett to Self Assembly; Academic Press: Boston, 1991. (3) Kim, K.; Kim, C.; Byun, Y. Biomaterials 2004, 25, 33. (4) Heiden, A. P.; Goebbels, D.; Pijpers, A. P.; Koole, L. H. J. Biomed. Mater. Res. 1997, 37, 282.

domain microstructure,5 while atomic force microscopy is useful for monitoring nanoscale properties,6 and time-of-flight secondary ion mass spectrometry provides7,8 a powerful, relatively recent approach to monitor domain composition in LB films. For Langmuir films, π-A isotherms are routinely used to monitor surface thermodynamics, while further characterizations may be accomplished with X-ray reflection to track lipid supramolecular organization. Because LB films are generally formed by transfer of Langmuir films onto a solid substrate, one issue of interest is the fidelity of molecular structure and phase state in LB films as compared to Langmuir films of the same material. A priori, although it seems improbable that the molecular properties of Langmuir films could be retained upon transfer to a solid substrate, limited evidence exists6-9 indicating similar domain formation in both instances. However, detailed comparisons of LB with Langmuir films over the whole range of surface pressures are lacking. The problem is rendered awkward by the fact that structural measurements are generally performed in a “discontinuous” mode; that is, one experiment (with a single film) is performed at each desired surface pressure. Comparison of the properties of LB film constituents with those of Langmuir films requires techniques that directly monitor (5) Cruz, A.; Worthman, L.-A.; Serrano, A. G.; Casals, C.; Keough, K. M. W.; Pe´rez-Gil, J. Eur. Biophys. J. 2000, 29, 204. (6) Cruz, A.; Va´zquez, L.; Ve´lez, M.; Pe´rez-Gil, J. Biophys. J. 2004, 86, 308. (7) Harbottle, R. R.; Nag, K.; McIntyre, N. S.; Possmayer, F.; Petersen, N. O. Langmuir 2003, 19, 3698. (8) Bourdos, N.; Kollmer, F.; Benninghoven, A.; Ross, M.; Sieber, M.; Galla, H.-J. Biophys. J. 2000, 79, 357. (9) Hui, S. W.; Yu, H. Biophys. J. 1993, 64, 150.

10.1021/la063139h CCC: $37.00 © 2007 American Chemical Society Published on Web 03/28/2007

LB Films Formed by Varying Surface Pressure

film molecular structure and supramolecular organization. IR spectroscopy and epifluorescence microscopy, taken together, are well-suited for this purpose. The advantage of the IR measurement is that direct molecular structure information (e.g., chain conformational order, protein secondary structure) can be acquired. The technique of IRRAS was developed10-12 specifically to monitor lipid/protein interaction in models for pulmonary surfactant in Langmuir films.13,14 The LB experiments are tedious, because, as noted above, they require films to be transferred at a whole series of discrete surface pressures. The COVASP-LB approach reported here facilitates the experiment. Epifluorescence microscopy provides an important parallel measurement to IR through its ability to monitor domain structure at the sub-micrometer level. Thus, microscopic observation of lipid and lipid/protein interfacial monolayers doped with traces of fluorescently labelled molecules has provided new understanding of the parameters governing lateral structure in biomembrane systems.15 Epifluorescence microscopy applied to Langmuir films has revealed segregation of ordered and disordered regions in both simple lipid systems15,16 and complex mixtures mimicking real membrane layers.17,18 Of particular importance is the ability of the technique to characterize domain formation and reorganizations associated with the physiological function of the pulmonary surfactant system when subjected to compression-expansion dynamics. In addition, the observation of immobilized LB films under a fluorescence microscope provides advantages related to the lateral resolution achievable and improved sensitivity at low fluorescence levels. The current study has two goals. First, we demonstrate the feasibility of transferring lipid or lipid/protein films from the air/water interface to solid substrates while continuously varying the surface pressure in a controlled fashion. The molecular behavior of film constituents over the entire range of surface pressures is examined in a single transferred film with transmission IR spectroscopy, while domain formation within the films is evaluated by epifluorescence microscopy. The second goal of the study demonstrates the utility of the COVASP-LB approach for the study of lung surfactant films. Preliminary IR data are reported for a four-component film of 1,2-dipalmitoylphosphatidylcholine(DPPC)/1,2-dipalmitoylphosphatidylglycerol (DPPG)/cholesterol/SP-C (a pulmonary surfactant-specific protein). This film undergoes a reversible and probably physiologically relevant monolayer-to-multilayer transition at high surface pressures.19 Experimental Section Materials. 1,2-Dipalmitoylphosphatidylcholine (DPPC), acylchain perdeuterated DPPC (DPPC-d62), 1,2-dipalmitoylphosphatidylglycerol (DPPG), 1-palmitoyl-2-{12-[(7-nitro-2,1,3-benzoxadiazol4-yl)amino]dodecanoyl}phosphatidylcholine (NBD-PC), and cholester(10) Mitchell, M. L.; Dluhy, R. A. J. Am. Chem. Soc. 1988, 110, 712. (11) Flach, C. R.; Brauner, J. W.; Taylor, J. W.; Baldwin, R. C.; Mendelsohn, R. Biophys. J. 1994, 67, 402. (12) Mendelsohn, R.; Brauner, J. W.; Gericke, A. Annu. ReV. Phys. Chem. 1995, 46, 305. (13) Meister, A.; Nicolini, C.; Waldmann, H.; Kuhlmann, J.; Kerth, A.; Winter, R.; Blume, A. Biophys. J. 2006, 91, 1388. (14) Cai, P.; Flach, C. R.; Mendelsohn, R. Biochemistry 2003, 42, 9446. (15) Mohwald, H.; Dietrich, A.; Bohm, C.; Brezesinski, G.; Thoma, M. Mol. Membr. Biol. 1995, 12, 29. (16) Nag, K.; Boland, C.; Rich, N.; Keough, K. M. W. Biochim. Biophys. Acta 1991, 1068, 157. (17) Discher, B. M.; Maloney, K. M.; Schief, W. R., Jr.; Grainger, D. W.; Vogel, V.; Hall, S. B. Biophys. J. 1996, 71, 2583. (18) Nag, K.; Perez-Gil, J.; Ruano, M. L.; Worthman, L. A.; Stewart, J.; Casals, C.; Keough, K. M. W. Biophys. J. 1998, 74, 2983. (19) Wang, L.; Cai, P.; Galla, H.-J.; He, H.; Flach, C. R.; Mendelsohn, R. Eur. Biophys. J. 2005, 34, 243.

Langmuir, Vol. 23, No. 9, 2007 4951

Figure 1. Schematic of the COVASP-LB film transfer. A CaF2 crystal is partially immersed through the air/water interface prior to spreading a Langmuir film. The initial position of the crystal at the A/W interface is marked on the crystal, which is then raised through the interface at a known (constant) rate. The barrier is moved so that the area is changed at a constant rate. From the isotherm, the pressure at each area (thus, at each time point) is known. We therefore know the pressure at each position on the crystal (since the area at each position is known). The IR beam (apertured to 150 µm in the transfer direction) is ∼1/60th the length of the 10 mm film. We estimate the uncertainty in terms of pressure at ∼1 mN/m, which, on average, corresponds to a length of ∼160 µm on the crystal. The residual clean crystal surface serves as a reference channel. The IR aperture is 150 × 200 µm, the former in the direction of transfer. ol were purchased from Avanti Polar Lipids (Alabaster, AL) and were used without further purification. Chloroform, methanol, EDTA, and HPLC-grade water were obtained from Fisher Scientific (Pittsburgh, PA). Trizma [tris(hydroxymethyl)aminomethane] hydrochloride, tris base (Trizma), and sodium chloride were purchased from Sigma (St. Louis, MO). Highly purified porcine pulmonary surfactant SP-C was the generous gift of Prof. Kevin Keough and Ms. June Stewart (Memorial University of Newfoundland). Monolayer Preparation for IR Measurements. DPPC-d62, DPPC, and cholesterol solutions were prepared in chloroform, while DPPG and SP-C solutions were prepared in chloroform/methanol (10:1 and 2:1 v/v, respectively) at ∼1 mg/mL concentrations. Lipid/ cholesterol/peptide solutions for studies of the monolayer to multilayer transition in a four-component model system consisted of DPPC/ DPPG at a 4/1 molar ratio with 7 mol % cholesterol and 4 mol % SPC. Isotherm Acquisition and Langmuir-Blodgett Film Formation for IR. A model 611 LB trough (Nima Technology, Inc., Coventry, England) equipped with a PS4 surface pressure sensor and a D1L-75 linear dipper was used for the π-A isotherm acquisition and monolayer transfer experiments. A 25 mm diameter circular CaF2 IR window was clamped onto the dipper and was partially immersed in the subphase prior to spreading the film. The subphase consisted of 150 mM NaCl, 5 mM Tris, and 0.1 mM EDTA at pH ∼7.0 in HPLC water. The temperature of the subphase was held constant at ∼20 °C. Typically for monolayer formation, ∼35 µL of a 1 mg/mL solution was spread dropwise onto the clean surface, and 30 min was allowed for solvent evaporation and film relaxation/ equilibration. Films were transferred onto the CaF2 substrate at a rate of substrate movement of 1 mm/min while the film was compressed at a rate of area change of 15 cm2/min. The pressure varied continuously, but not linearly, during the transfer. The entire transferred film typically occupied a length of ∼10 mm on the substrate. π-A isotherms were simultaneously recorded during the transfer. Film positions on the substrate were physically marked at the beginning and end of transfer, because there was insufficient contrast between clean substrate and monolayer-covered substrate in the visible image acquired from the IR instrument to give useful positioning information. A schematic representation of the experiment is given in Figure 1.

4952 Langmuir, Vol. 23, No. 9, 2007 IR Spectroscopy. IR spectra from particular film locations along a line on the substrate were collected with the single point detector of the Perkin-Elmer “Spotlight” system. Samples were positioned with a high-precision computer-controlled motorized XY sample stage. Film lengths were typically 10 mm, and spectra were acquired with an aperture size of 200 × 150 um (the latter was in the transfer direction). 128 scans were co-added for each spectrum using 4 cm-1 spectral resolution with one level of zero filling for presentation. Typically, 30-60 spectra were acquired from along a line in each film. Total collection time was ∼1-3 h; separate backgrounds were taken every 1-2 data points. It is emphasized that to achieve the excellent spectral quality reported below requires careful control of the relative humidity. Relative humidity levels of 10% or lower in the sample chamber yielded the best spectra. The IR stage was purged in a home-built chamber. A reviewer has raised the issue of possible air oxidation of cholesterol in studies of the monolayer-to-multilayer transition. Although we routinely use a dry air purge, we have performed some experiments in a dry N2 environment and observed no differences when compared to the dry air purge. IRRAS. Our IRRAS instrumentation for examination of Langmuir films has been described in detail elsewhere.20 Briefly, IRRAS spectra are acquired with a Bruker Instruments Equinox 55 spectrometer equipped with an external variable angle reflectance accessory, the XA511. The accessory is coupled to a custom-designed Langmuir trough constructed by Nima Technology Ltd. (Coventry, England). The IR beam is directed through the external port in the spectrometer, reflected by three mirrors in a rigid mount, and focused on the water surface. The reflected light is directed onto a mercury cadmium telluride detector. The entire experimental setup is enclosed and purged to keep the relative humidity levels both low and as constant as feasible. Epifluorescence Microscopy. To permit observation of the films by epifluorescence microscopy, 1 mol % of the fluorescent lipid NBD-PC was added to the lipid in organic solvent, and the films were transferred onto glass slides while continuously monitoring surface pressure and dipper position. Epifluorescence microscopy of the supported LB films was then performed as explained elsewhere21 using a Zeiss Axioplan II fluorescence microscope (Carl Zeiss, Jena, Germany) equipped with the appropriate fluorescence filters to allow for the observation of NBD-PC fluorescence (maximum fluorescence emission at 520 nm). Images at different pressures were obtained at different positions from the COVASPLB films and subsequently assigned to the surface pressures obtained during film transfer from the Langmuir trough. Data Analysis. (i) IRRAS. Spectra were initially examined with Grams/32 (Galactic Industries Corp., Salem, NH). IR frequencies were calculated with a center-of-gravity algorithm written by D. Moffatt and provided by the National Research Council of Canada. Band areas, peak heights, and half-widths (FWHH) were calculated with the same program or with ISys (Malvern Instruments, Olney, MD). (ii) Epifluorescence. Digitally recorded images from epifluorescence microscopy were quantitatively analyzed using the program Scion Image (Scion Corporation, MD). Data shown are averaged values with standard deviations obtained after analysis of at least five different frames.

Results (A) IR of COVASP-LB Langmuir Films of DPPC and DPPC-d62. A typical isotherm generated for DPPC monolayers during the film transfer process to the IR substrate (CaF2) is shown in the inset to Figure 2. Although material is differentially transferred as surface pressure is varied, the horizontal axis is still reporting an area/molecule because the actual amount of (20) Flach, C. R.; Bi, X.; Xu, Z.; Brauner, J. W.; Mendelsohn, R. Appl. Spectrosc. 2001, 55, 1060. (21) Cruz, A.; Worthman, L. A.; Serrano, A. G.; Casals, C.; Keough, K. M. W.; Perez-Gil, J. Eur. Biophys. J. 2000, 29, 204.

Wang et al.

Figure 2. Typical COVASP-IR spectra of DPPC and acyl chainperdeuterated DPPC. The negative-going peaks marked with asterisks (*) arise from uncompensated vibrations originated from within the spectrometer. Some specific vibrational assignments are marked on the figure (ν ) stretching, δ ) bending, sym ) symmetric, asym ) asymmetric). Residual atmospheric CO2 absorbs near 2350 cm-1. Inset: Surface pressure-area isotherm acquired for DPPC during the transfer process. Various states of the monolayer are indicated.

film transferred onto a crystal area of ∼2.5 cm2 is less than 2% of the utilized area of the trough (150 cm2). Thus, the isotherm should be very similar to that acquired in the absence of material transfer, as it indeed is. The isotherm itself has been widely studied,22 and the various phases noted are labeled in the figure. Of primary current importance is the liquid expanded-liquid condensed (LE-LC) transition detected as a near horizontal line in the isotherm at 7-9 mN/m. Transfer of DPPC was generally terminated near the collapse point at ∼59 mN/m. Typical IR spectra for both DPPC and acyl-chain perdeuterated DPPC (DPPC-d62) from COVASP-LB films deposited on a CaF2 substrate are shown in Figure 2. The spectral features have been discussed many times previously (e.g., ref 23). Some vibrational modes of interest are labeled on the figure. Typical RMS noise values of ∼6 × 10-6 units were observed. The asterisks near 2929, 2912, 2858, 1550, and 1645 cm-1 in Figure 2A indicate the position of weak spectral features generated from components within the spectrometer. The origin of these bands is uncertain, although under normal instrument usage (i.e., absorbances . 1 mA) they are adequately compensated. We suggest that these very weak residual features originate from the parylene coating of the KBr port windows. The imperfect compensation presumably arises from the different optical paths taken by the sample beam (through the film-covered crystal) and the reference beam (through a different clean spot on the crystal). The effect of these residual bands is mainly to distort the C-H stretching region at peak intensities of less than ∼0.3 milliabsorbance units. In addition, a generally small variable level of residual CO2 is noted near 2350 cm-1. (B) Molecular Characterization of COVASP-LB Films by IR. To demonstrate that the transferred films exhibit continuously varying molecular properties, the surface pressure-dependence of the DPPC symmetric CH2 stretching intensity for IRRAS measurements from Langmuir films on water and for COVASPLB films on CaF2 is plotted in Figure 3A. Although the intensity scales for IRRAS (a reflection experiment) and COVASP-LB (a transmission experiment) cannot be directly compared, the surface pressure-induced intensity variations in the two are remarkably parallel. Areas measured in the COVASP-LB films (22) Pallas, N. R.; Pethica, B. A. Langmuir 1985, 1, 509.

LB Films Formed by Varying Surface Pressure

Langmuir, Vol. 23, No. 9, 2007 4953

Figure 4. CH2 stretching region 2800-3000 cm-1, overlaid at various indicated surface pressures, of COVASP-LB films of DPPC on a CaF2 substrate.

Figure 3. (A) Surface pressure-dependence of the integrated area of the symmetric CH2 stretching vibration of DPPC in the COVASPLB experiment (left-hand scale, b) and in the IRRAS experiment (0, right-hand scale). (B) Surface pressure-dependence of the integrated area of the CdO stretching vibration of DPPC in the COVASP experiment (b).

at the very lowest surface pressures may be lower than the true value because of the underlying uncompensated negative spectral features discussed above. The measured intensity variation with surface pressure is a manifestation of the density, orientation, and conformation of the lipid chains in the film. The intensity variation of the DPPC CdO stretching vibration with surface pressure as plotted in Figure 3B parallels that for the CH2 stretching mode. The similar variation with surface pressure in the two cases suggests that the molecular density rather than changes in molecular conformation or orientation controls the overall intensity, because it is unlikely that the two DPPC CdO groups would have surface pressure-induced orientation/conformation changes that exactly parallel those of the methylene chains. The lipid CdO peaks are not affected by residual spectral features arising from components within the spectrometer. The COVASP-LB IR measurements provide a direct means for comparison of LB with Langmuir film structure. To achieve this goal, it is necessary to compare the surface pressure-induced changes in conformation-sensitive regions of the COVASP-LB IR spectrum with IRRAS data. The CH2 stretching regions for DPPC COVASP-LB films at a series of surface pressures (corresponding to ∼25% of the collected spectra in each instance) are overlaid (and offset) in Figure 4. The most intense bands in each spectrum of DPPC near 2850 and 2920 cm-1 correspond to the CH2 symmetric and asymmetric stretching modes, respectively. Vertical lines as a guide to the eye are drawn at 2850 and 2918.2 cm-1. Changes in peak positions and halfwidths as the surface pressure increases are evident. The frequencies of the CH2 asymmetric and symmetric stretching modes are plotted in Figure 5A and B, respectively, and are

compared to the same parameters obtained from IRRAS in each instance. Equivalent data were acquired for DPPC-d62 (not shown). Most interesting in Figure 5 is the observation that the surface pressure-induced conformational order change in the lipid acyl chains at surface pressures corresponding to the LE-LC transition is mostly preserved upon film transfer. As is well documented,23 the methylene stretching frequencies decrease upon introduction of conformational order into the chains. Frequency shifts in the symmetric CH2 stretching mode from ∼2855.5 to ∼2851.5 cm-1 are noted (Figure 5B) between surface pressures of ∼4 and 9 mN/m in the IRRAS data. The changes persist in the COVASPLB IR data in which the frequency shifts from ∼2852.7 to ∼2850.3 cm-1 occur between surface pressures of ∼5 and 8 mN/m, close to, but somewhat broadened from, the surface pressure of the LE-LC transformation in the inset to Figure 2. Thus, the conformational order change in the acyl chains during the transition is substantially preserved on the CaF2 substrate, although the actual values of the measured methylene stretching frequencies in both the initial and the final states are lower in the COVASP-LB IR data, suggesting more conformational order at all pressures in the transferred films. Similar patterns were observed (Figure 5A) for the asymmetric CH2 stretching modes and for the CD2 stretching vibrations in DPPC-d62 (data not shown). To further characterize the lipid physical state in the transferred films, the half-widths (FWHH) of the symmetric methylene stretching mode for COVASP-LB IR were compared to the IRRAS data. Substantial differences are noted between the two sets of measurements (Figure 5C). The COVASP-LB line widths are significantly narrower than the IRRAS values at corresponding pressures, indicating substantially decreased rates of reorientational motion in the transferred films. This result is consistent with immobilization of the transferred films. However, surface pressure-induced line-width changes characteristic of the LELC transition are still evident in the COVASP-LB data. Thus, the COVASP-LB line widths are decreased from ∼13 to 10 cm-1 over the range ∼6-9 mN/m, while the IRRAS line widths (23) Mendelsohn, R.; Mantsch, H. H. In Progress in Protein-Lipid Interactions 2; Watts, A., de Pont, J. J. H. H. M., Eds.; Elsevier: Amsterdam, 1986; pp 103-146.

4954 Langmuir, Vol. 23, No. 9, 2007

Wang et al.

Figure 6. (A) Compression π-A isotherms of a DPPC monolayer, either undisturbed (black line) or obtained during simultaneous transfer onto a glass support (dashed line). (B) Epifluorescence microscopy images of DPPC films containing 1% (molar) of the fluorescent probe NBD-PC transferred onto glass supports either at constant pressure (left-hand frames) or at constantly varying surface pressure (right-hand image). Each of the left-hand frames corresponds to an independent DPPC film, compressed to the target pressure indicated by the central scale, and transferred at the same constant pressure onto a glass support. The right-hand image has been composed by overlapping different frames acquired from the same film, prepared by transfer of a single DPPC/NBD-PC monolayer while simultaneously compressed. All of the experiments were carried out in a subphase composed of Tris, 5 mM, pH 7, and NaCl, 150 mM.

Figure 5. (A) Surface pressure-dependence of the asymmetric CH2 stretching frequency of DPPC in the COVASP-LB experiment (9) on a CaF2 substrate and in the IRRAS experiment (0). (B) Surface pressure-dependence of the symmetric CH2 stretching frequency of DPPC in the COVASP-LB experiment (9) on a CaF2 substrate and in the IRRAS experiment (0). (C) Surface pressure-dependence of the symmetric CH2 stretching half-width (full width at half-height) for DPPC in the COVASP-LB experiment (9) on a CaF2 substrate and in the IRRAS experiment (0) from an aqueous surface.

decrease from ∼18 to 15.5 cm-1 over the surface pressure range 5-9 mN/m. The line-width variation in the transition region is less cooperative in the Langmuir film than in the COVASP-LB film. Similar results were obtained for the CH2 asymmetric stretching vibration as well as for the CD2 stretching modes of DPPC-d62 (data not shown). (C) Domain Structure of DPPC COVASP Films by Epifluorescence Microscopy. Studies of domain formation by epifluorescence complement the IR measurements described above. Inclusion of a trace of fluorescently labelled lipid allows evaluation of the effect of compression on the lateral structure of the films, through direct observation of the differential distribution of the probe between condensed and expanded areas of the monolayer. We have previously shown that transference onto either glass or mica supports does not alter the microstructure of interfacial DPPC films as studied by epifluorescence or atomic force microscopy (AFM).24 (24) Cruz, A.; Vazquez, L.; Velez, M.; Perez-Gil, J. Biophys. J. 2004, 86, 308.

Figure 6a compares π-A compression isotherms at 24 °C of DPPC films containing 1% of the fluorescent lipid probe NBDPC compressed at a relatively rapid speed, obtained either from a non-disturbed interface or from an interface subjected to simultaneous extraction of the interfacial film onto an LB glass support. The two isotherms in Figure 6a are very similar, including lift-off pressures at ∼75-78 Å2/molecule and a conspicuous liquid-expanded to liquid-condensed transition plateau at 1012 mN/m. As with the COVASP films transferred to CaF2 for IR measurements, it is safe to conclude that extraction of the film onto glass supports during compression does not significantly alter the main features of the isotherm, consistent with continuous compression-driven reorganization of the lipid molecules at the interface. Figure 6b presents fluorescence microscopy images of DPPC films that contain 1% NBD-PC compressed to different pressures and transferred onto glass supports. The figure compares the microstructure of two types of supported DPPC films. The frames on the left side correspond to microscopic images taken from films prepared by transfer of films compressed to and maintained during transfer at constant defined surface pressure. Each of these pictures, therefore, was taken from a different DPPC film, compressed to the indicated pressure, then transferred onto a glass support, forming a LB film with very homogeneous structure as observed under the microscope. This is the traditional way used in the literature25 to prepare supported lipid films. On the other hand, the image shown on the right side of Figure 6b has been composed by overlapping pictures taken from a single (25) Ariga, K.; Nakanishi, T.; Michinobu, T. J. Nanosci. Nanotechnol. 2006, 6, 2278.

LB Films Formed by Varying Surface Pressure

Figure 7. Condensation curve of DPPC films calculated from LB films transferred either at constant pressure (3) or at constantly varying surface pressure (O), or from images recorded in situ from films observed at the air-water interface (b). Percent average of dark, condensed area with respect to the total surface is plotted versus the surface pressure. Films contained 1% (molar) NBD-PC and were always formed, compressed, and transferred at 25 °C. Data points represent mean and standard deviations after averaging five frames obtained at each pressure.

supported film prepared upon transference that was simultaneous with the acquisition of the compression isotherm. The surface pressure at which each segment of the film was transferred to the support was therefore continuously varying along the line of transference. Observation of this LB film under the microscope reveals a different and progressively changing morphology along the transference axis. The surface pressure scale in Figure 6B indicates the corresponding surface pressures sustained by the different films (frames on right) or segments of the film (image on left) at the time of their extraction. The film extracted during compression could therefore be calibrated from the data on dipper position and surface pressure acquired by the computer during the transfer, to identify the surface pressure at which each segment of the film was transferred. The structure of the film transferred during compression, sampled along the transference axis, illustrates in a single film the nucleation and growth of lipidcondensed domains that occurs in DPPC monolayers along the compression isotherm. In contrast, sampling of images in a direction perpendicular to the transfer line produces frames of the film at identical pressure, which can therefore be averaged or compared to analyze the structure of the film at any given surface pressure in qualitative or quantitative terms. While the conventional LB technique requires the preparation and transfer of many different films to analyze the structure of a given interfacial system at a few defined pressures, the observation of this (COVASP)-LB film allows a full structural characterization of the isotherm in a single experiment, with the possibility of sampling any given surface pressure along the desired segment of the isotherm. Figure 7 compares the full compression-driven condensation curves calculated either from constant pressure LB or from COVASP-LB DPPC/NBD-PC films. Notice that the condensation curve from the COVASP-LB film can be easily prepared taking as many points as desired with just sampling the same film at many different positions along the transfer line. As was qualitatively deduced from the images, the film is 10-20% less condensed when it is simultaneously transferred than when analyzed at different independent constant pressures. In addition, the higher density of points in the COVASP-LB films facilitates the observation of the LC-LE transition near 10 mN/m. Data points on condensation obtained from the COVASP-LB film include also horizontal error bars, which indicate the error in

Langmuir, Vol. 23, No. 9, 2007 4955

surface pressure assignment to any given condensation frame. As pressure is continuously varying along the transfer line, each frame in the quantitative analysis includes condensation information over a range of pressures, which is equally wide along the isotherm. Interestingly, the condensation curve obtained from COVASP-LB films is reasonably comparable to the one calculated from images recorded directly from a monolayer compressed at the air-liquid interface (see black points in Figure 7). These results confirm that the extent of condensation is more sensitive to the existence of periods of equilibration before transfer to the support than to the transfer process itself and that the COVASPLB procedure efficiently captures qualitative and quantitative pressure-dependent features occurring in interfacial films. (D) Initial Biological Application of COVASP-LB IR: Monolayer-to-Multilayer Transition in a Four-Component Lung Surfactant Model. Pulmonary surfactant, a lipid-protein complex located at the air/alveolar lining of the mammalian lung, functions in vivo to prevent alveolar collapse during exhalation by lowering surface tension to near-zero values. In previous studies,5,6,8,19,26,27 supramolecular structure changes upon compression were observed in films that mimic the alveolar lining layer. A four-component model surfactant film of DPPC/ DPPG, cholesterol, and the pulmonary surfactant specific protein SP-C (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %) displayed a biologically relevant reversible transition from monolayers to surface-associated multilayers at high surface pressures upon compression. In our previous study,19 AFM measurements on LB films verified the formation of multilayers, while IRRAS was used to evaluate the structure and orientation of SP-C. IRRAS measurements of the monolayer-to-multilayer transition19 were undertaken at surface pressures >50 mN/m. Under these conditions, it was difficult to quantitatively analyze the chemical composition or structure changes in the film as the pressure was continuously changing due to relaxation events. It was anticipated that the monolayer-to-multilayer transition would be detectable in the COVASP-LB films, thereby rendering this process amenable to a variety of useful experiments. The feasibility of the approach is shown here. An isotherm for the system during a COVASP-LB experiment is shown in the inset to Figure 8 and reveals a generally expanded film at low pressures. A near horizontal region commences at ∼54 mN/m, indicative of the onset of multilayer formation. Prior studies of the reversibility of the process19 showed that the excluded material in the film respreads into the monolayer upon expansion of the film. COVASP-LB IR spectra at several surface pressures over the range 23-56 mN/m are shown in Figure 8. The excellent spectral quality permits detection of the amide I mode (labeled in Figure 8) at ∼1655 cm-1 in addition to the phospholipid bands alluded to previously. Several spectroscopic parameters relevant for investigating the formation of multilayer films are plotted in Figures 9 and 10. In Figure 9A, the pressure-induced variation in the methylene symmetric stretching intensity shows a major increase with a sharply defined onset pressure of ∼54 mN/m, in excellent agreement with the appearance of the near horizontal line in the isotherm (Figure 8, inset). Similar behavior is exhibited by the asymmetric methylene stretching mode and the lipid CdO stretch (data not shown). In addition, the symmetric CH2 stretching frequency exhibits (Figure 9B) a small (0.1 cm-1) but significant (26) Diemel, R. V.; Snel, M. M. E.; Waring, A. J.; Walther, F. J.; van Golde, L. M. G.; Putz, G. T.; Haagsman, H. P.; Batenburg, J. J. J. Biol. Chem. 2000, 277, 21179. (27) Breitenstein, D.; Batenburg, J. J.; Hagenhoff, B.; Galla, H.-J. Biophys. J. 2006, 91, 1347.

4956 Langmuir, Vol. 23, No. 9, 2007

Figure 8. COVASP-LB IR spectra from the four-component film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %) at various surface pressures, as indicated on the right-hand side of the figure. Of particular interest is the amide I mode near 1656 cm-1. (Inset) Surface pressure-area isotherm acquired for a four-component film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %) during transfer to a CaF2 substrate under conditions of continuously varying surface pressure. The near horizontal region between ∼54 and 56.7 mN/m reflects multilayer formation.

decrease between ∼54 and 56 mN/m. The band intensity increase (Figure 9A) is consistent with multilayer formation as the number of lipid molecules interrogated by the IR beam is increased in the presence of multilayer. The decreased frequency in the highpressure region (Figure 9B) is consistent with slightly increased conformational ordering in the multilayer film. A central theme in the organization of these multilayers is the role of protein in their formation and stabilization. Figure 10A depicts increased intensity of the amide I mode (peptide bond CdO stretch) commencing at the onset of multilayer formation. Significant frequency shifts are also noted in the amide I region upon multilayer formation. As seen in Figure 10B, an increase of ∼0.7 cm-1 is observed between 54 and 56 mN/m.

Discussion LB films are routinely formed at constant surface pressures and are subsequently structurally characterized and/or utilized for particular applications. The examination of films as a function of surface pressure is time-consuming, as it requires the preparation of individual films (with variable initial properties) at each pressure. The current study explores the feasibility of transferring films under controlled conditions of continuously varying surface pressure. In searching the literature, we have seen no reports attempting the characterization (or in fact the preparation) of COVASP-LB films. As demonstrated in this study, IR spectroscopy and epifluorescence microscopy provide complementary and effective characterizations of these films. The IR experiments are fairly straightforward with modern instrumentation. The RMS noise in the current spectra is ∼6 microabsorbance units. For typical spectral features (∼1 milliabsorbance unit), the S/N level of ∼160 thus achieved permits the precise measurement of frequency

Wang et al.

Figure 9. (A) Surface pressure-dependence of the integrated area of the symmetric CH2 stretching vibration in the COVASP-LB film spectra from the four-component film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %). (B) Surface pressuredependence of the frequency of the symmetric CH2 stretching vibration in the COVASP-LB film spectra from the four-component film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %).

shifts as small as ∼0.1 cm-1. The relationship between S/N and the precision of frequency measurements in IR was discussed 30 years ago in a useful but rarely cited study by Cameron et al.28 Such precision is needed to track small frequency shifts in either the amide I (Figure 10B) or the CH2 stretching regions (Figure 9B) upon multilayer formation. An important characteristic of COVASP films is to determine the extent to which the transferred films reflect the physical properties and phase behavior of the Langmuir films. Transfer ratios, commonly used to assess the efficiency of film transfer, cannot be conveniently measured when film pressure and area characteristics are both continually changing during the transfer. However, the results reported above for both techniques suggest that the transferred film properties continuously vary with pressure and reflect, to a significant extent, the properties of the original Langmuir films. With regard to the transfer process itself, the isotherms obtained for DPPC during the transfer processes (Figures 2 and 6 for the IR and epifluorescence, respectively) are in excellent agreement with several in the literature,22 indicating that monolayer thermodynamics are not substantially altered during transfer. The molecular structure information available from the IR also speaks to the fidelity of transfer. The surface pressureinduced variations in the CH2 stretching frequencies in COVASP(28) Cameron, D. G.; Kauppinen, J. K.; Moffatt, D. J.; Mantsch, H. H. M. Appl. Spectrosc. 1982, 36, 245.

LB Films Formed by Varying Surface Pressure

Figure 10. (A) Surface pressure-dependence of the integrated area of the amide I vibration in the COVASP-LB film spectra from the four-component film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %). (B) Surface pressure-dependence of the amide I frequency in the COVASP-LB film spectra from the fourcomponent film (DPPC/DPPG 4:1 mol ratio; cholesterol, 7 mol %, SP-C, 4 mol %).

LB (Figure 5A,B) films mimic those observed for Langmuir films by IRRAS,29 indicating that the lipid conformational order change during the LE-LC transition persists to a substantial extent in the transferred films. As a complement to the molecular structure information from IR spectral parameters, the epifluorescence data (Figures 6 and 7) directly monitor film domain structure. The transference of a lipid interfacial film onto a solid support during continuous compression evidently preserves most of the structural features associated with progressive condensation of the films in situ, including nucleation and growth of condensed lipid domains excluding the fluorescent probe. Thus, we have demonstrated (Figures 6 and 7) that qualitative and quantitative analysis of a single COVASP-LB film allows a complete characterization of the structural transformations occurring in the film along the entire isotherm. From a strictly quantitative point of view, packing of liquidcondensed areas seems to progress to a lesser extent in COVASPLB films than in conventional LB films transferred at defined constant pressures. A possible reason is that simultaneous transference during compression may introduce a local expanding component, which affects the whole isotherm. It is also probable that the difference between the two types of transferred films is related to differences in time-dependent processes associated with equilibrated and non-equilibrated systems. Films transferred at constant pressure are equilibrated for some time at the desired (29) Flach, C. R.; Cai, P.; Mendelsohn, R. In Emerging Techniques in Biophysics; Arrondo, J. L., Alonso, A., Eds.; Springer-Verlag: Heidelberg, Germany, 2006; pp 49-71.

Langmuir, Vol. 23, No. 9, 2007 4957

pressure before transference is initiated. Such equilibration time could help to complete structural reorganizations that might occur at rates comparable to the speed of compression. The condensation curves obtained from both COVASP films and those obtained from images taken in situ from the air-liquid interface are qualitatively similar, confirming that transfer under COVASP conditions preserves pressure-dependent structural features existing at the interface. Transference of films at variable pressure could likely trap transient structures still far from equilibrium. The observation of transient structures may better reflect processes occurring during dynamic transformations of the films in situ than films transferred with relatively long equilibration times. Analysis of COVASP-LB films therefore provides the possibility of analyzing not only structure but kinetics of lateral reorganizations. Proper selection of the range of pressures to be scanned and the speed of compression/transference would permit access to detailed analysis of any desired segment of the isotherm under potentially different kinetic regimes. The fact that lipid layers immobilized in solid supports may permit detection and analysis of transient structures occurring with variable lifetimes in freestanding membranes has been previously proposed.30 The excellent agreement between the onset pressure of the monolayer-to-multilayer transition seen in the isotherm (Figure 8) of the four-component system and in various COVASP-LB parameters (Figures 9 and 10A) suggests that Langmuir film properties are retained in the transferred film at high surface pressures. This result is also consistent with our previous studies,19 in which the AFM measurements of a film transferred to mica at various surface pressures revealed extensive multilayer formation only in the high-pressure regime in the presence of protein. The measured conformational properties of SP-C in the four-component system as measured with COVASP-LB IR parallel those measured in our prior IRRAS studies.19 An amide I frequency of ∼1650 cm-1 for the SP-C was previously reported in the IRRAS measurements. This frequency (from a sample in a D2O environment) indicated that the predominant conformation of the protein in these films is helical, consistent with solution NMR measurements.31 In addition, a small increase in the frequency (∼2-3 cm-1) of the amide I band was observed19 when IRRAS spectra were acquired at pressures where multilayers formed. Exactly parallel behavior is observed in the current COVASP-LB films. Thus, the amide I frequency (in a film transferred from H2O) increases from ∼1656.5 to 1657.4 cm -1 during multilayer formation (Figure 10B). The origin of the frequency increase in the IRRAS experiment was attributed to either a slight increase in the R-helical content in a bilayer versus monolayer environment or an increase in the hydrophobicity of the environment as compared to the monolayer phase. Both factors (increased helicity and hydrophobicity) are reasonable. Because COVASP-LB films have not previously been characterized (to our knowledge), direct comparison with prior studies in the literature is not possible. However, indications do exist that suggest that some properties of phosphatidylcholine phases are preserved in supported systems. Naumann et al.32 studied single bilayers formed on solid silicon beads of 320 nm radius. The gel-to-liquid crystal phase transition of DPPC was preserved but occurred with a transition temperature 2 °C lower than that found for multilamellar vesicles. The ripple (Pβ′) phase was inhibited in single bilayers on a solid support. The experiment (30) Nielsen, L. K.; Bjornholm, T.; Mouritsen, O. G. Nature 2000, 404, 352. (31) Johansson, J.; Szyperski, T.; Curstedt, T.; Wuthrich, K. Biochemistry 1994, 33, 6015. (32) Naumann, C. D.; Brumm, T.; Bayeral, T. M. Biophys. J. 1992, 63, 1314. (33) Rana, F. R.; Widayati, S.; Gregory, B. W.; Dluhy, R. A. Appl. Spectrosc. 1994, 48, 1196.

4958 Langmuir, Vol. 23, No. 9, 2007

showed that the supramolecular properties of lipid assemblies were partially preserved on this solid support. The current results complement and extend the study of Naumann et al.32 in that we now observe that the domain and molecular structure characteristics of monolayers are substantially preserved upon transfer to a CaF2 surface. The current approach offers the advantage of convenient examination of physical properties over a wide variety of transferred surface pressures, an option not possible in bilayer studies. In other relevant experiments, Langmuir films of simple lipids6 or bovine lipid surfactant extract32 have been investigated using fluorescence microscopy, and domain sizes and shapes have been compared to those observed with AFM or TOF-SIMS in films deposited on mica. The domain sizes and shapes compared well between methods, suggesting that films retained the same morphology as the original Langmuir films at the air-liquid interface. In a similar investigation of molecular ordering and orientation in monolayer domains, Hui and Yu9 used electron diffraction to study monolayers of L-R-dipalmitoylmonomethylphosphatidylethanolamine and L-R-dipalmitoyldimethyl-phosphatidylethanolamine transferred to EM grids. Examination of diffraction patterns with a 10 µm sampling area enabled the characterization of individual domains (previously seen in fluorescence as dark fields) as solids with the presence of pseudo long-range order. It is interesting to consider the potential range of applications of COVASP-LB IR. Evidently, the method will complement structurally oriented studies of LB films, because the molecular conformation information is unique to IR. As an extension of this, in experiments3,4 where films are modified on a substrate, either by polymerization or by other chemical methods, COVASP

Wang et al.

will easily be able to track surface pressure (i.e., molecular density) dependence of the extent of modification. There are limitations to COVASP, of course. As noted above, the half-widths of the methylene stretching modes measured in COVASP and IRRAS differ and reflect some suppression of motion in the chains on the substrate, although remnants of the LE-LC transition persist in this parameter. It might be also argued that transfer under simultaneous compression makes the study of defined features in equilibrated films difficult. However, analysis of COVASP-LB films provides a detailed molecular picture of the structure of the films caught while undergoing dynamical processes of clear physiological relevance such as compression-expansion cycling. In more fundamental applications, studies of the transfer process itself may be better understood with COVASP-LB. For example, the rate of deposition was shown by Rana et al.33 to influence chain conformation in the final film. The COVASP process will facilitate the examination of the surface pressure-dependence of this process. In addition, permanence of the film on the substrate (films are stable for at least 3 days under our conditions) will permit the detection of slow kinetic processes, which may alter the molecular structure of the film components. Acknowledgment. This work was supported by Grant GM29864-25 from the U.S. Public Health Service to R.M. Some of the funds for the IR microscope were provided by Rutgers University. J.P.-G. is supported by grants from the Spanish Ministry of Science (BIO2006-03130), Community of Madrid (S0505/MAT/0283), and the Marie Curie Programme (CT04007931 and CT04-512229). LA063139H