INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary surfactant, crucial for normal respiratory function and mechanics, is a complex mixture of lipids and three biophysically active apoproteins (1). Approximately 85 to 90% of lung surfactant by weight is made up of phospholipids, about a third of which is dipalmitoyl phosphatidylcholine (DPPC) and the remaining two-thirds are a spectrum of secondary zwitterionic and anionic phospholipids (4). The relative contributions of the multiple components of lung surfactant to the surface-active function of the overall system are not completely defined, although mechanistic understanding is substantial. DPPC is known to contribute significantly to the surface tension-lowering ability of dynamically compressed lung surfactant films. The lung surfactant proteins are understood as essential for facilitating adsorption to the air-water interface (8), and also act to improve respreading in the surface film during cycling (9, 10). The secondary phospholipids as a group are also known to improve adsorption (8, 11) and film respreading compared with DPPC (10), but the relative contributions of different molecular classes of lung surfactant phospholipids in facilitating these surface-active behaviors have not yet been determined.

Lung surfactant phospholipids can be subclassified based on whether their headgroups are zwitterionic or anionic at or near neutral pH. Zwitterionic phospholipids make up approximately 85% of total lung surfactant phospholipids (2). Phosphatidylcholines (PCs) are by far the largest group of zwitterionic phospholipids in pulmonary surfactant in all animal species studied, with small amounts of other zwitterionic phospholipids such as phosphatidylethanolamine and sphingomyelin also present. DPPC is the single most prevalent PC compound, but a host of other PCs with a broad distribution of saturated and unsaturated fatty chains are also included (6). Anionic or acidic phospholipids, particularly phosphatidylglycerol (PG) and phosphatidylinositol (PI), make up about 10 to 15% of the total phospholipids in mammalian lung surfactant (2).

Despite their significantly smaller content compared with zwitterionic phospholipids in lung surfactant, the possible functional importance of anionic phospholipids in biophysical activity has been of interest for some time (12). PG and PI are known to be biochemical markers of fetal lung maturity relevant for the neonatal respiratory distress syndrome (RDS) (12), and changes in the content of anionic phospholipids in lavage have been identified in patients with acute or adult respiratory distress syndrome (ARDS) (15, 16). However, biochemical associations with lung maturity or the development of respiratory disease do not necessarily imply a required role for anionic phospholipids in pulmonary surfactant function. In vitro biophysical studies with synthetic phospholipids have suggested that anionic phospholipids may contribute a negative charge to the surface of bilayers, facilitating molecular interactions with hydrophobic surfactant proteins (SP-B and SP-C) (17). However, studies in mixed phospholipid films have not identified specific surface-active behaviors associated uniquely with anionic as opposed to zwitterionic phospholipids (21). Liau and coworkers (25) have reported normal surface properties in PG-deficient surfactant from dogs after acute lung injury, and Beppu and coworkers (26) and Hallman and colleagues (27) showed that pulmonary mechanics were unchanged in animals fed an inositol-rich diet to deplete PG and increase PI in lung surfactant. Because the total content of anionic phospholipids was not altered in these latter studies (26, 27), they did not address the importance of acidic phospholipids as a group to surfactant activity.

To define more precisely the importance of zwitterionic compared with anionic phospholipids in adsorption and film behavior, the present study investigated chromatographically separated subfractions of the complete mix of phospholipids (PPL), and of phospholipids depleted in anionic constituents (mPPL), isolated from calf lung surfactant extract (CLSE). PPL and mPPL are studied alone and in combination with hydrophobic SP-B/C (hydrophobic SP) at the same 1.3% by weight total content found in CLSE. Experiments comparing PPL and mPPL, the latter lacking only the group of anionic phospholipids present in natural surfactant, reduce the risk of oversimplification inherent in the study of synthetic phospholipid mixtures. Relevance for lung surfactant function in vivo is also enhanced by examining a range of potentially important surface behaviors, including surface pressure-time adsorption, dynamic surface tension-lowering and respreading in cycled films on the Wilhelmy balance, and overall surface tension-lowering of dispersions on a pulsating bubble apparatus. The relative importance of zwitterionic and anionic phospholipids in the ability of phospholipid:SP mixtures to restore pressure-volume (P-V) mechanics toward normal after instillation into excised, surfactant-depleted rat lungs was also determined.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Calf Lung Surfactant Extract

CLSE, containing all of the hydrophobic constituents of endogenous surfactant, was prepared by extraction of cell-free bronchoalveolar lavage as in our prior work (11, 28, 29). Intact lungs from freshly killed calves (Conti Packing Co., Henrietta, NY) were lavaged with cold 0.15 M NaCl, and whole-lung surfactant (LS) was pelleted by centrifugation of supernatant at 12,000 × g for 30 min after low-speed centrifugation at 250 × g for 10 min to remove cells. CLSE was obtained by chloroform:methanol (C:M) extraction (30) of LS.

Purified Calf Lung Surfactant Phospholipids and Modified Calf Lung Surfactant Phospholipids

The complete mix of zwitterionic and anionic lung surfactant phospholipids (PPL) was isolated from CLSE by gel permeation column chromatography with C:M 1:1 (vol/vol) containing 5% 0.1 M HCl (31). Two passes through a Sephadex LH-20 column (Pharmacia-LKB Biotechnology, Piscataway, NJ) separated the mixed phospholipids away from SP-B and SP-C and neutral lipids. The same method, but employing nonacidic C:M (2:1, vol:vol) rather than acidified solvent was used to obtain modified phospholipids depleted in anionic components (mPPL). Acid remaining in PPL fractions was removed by C:M extraction (31). Final PPL and mPPL preparations had a protein content below the limits of detection by the assays of Lowry and colleagues (32) and Kaplan and Pederson (33). The distribution of phospholipid classes in CLSE, PPL, and mPPL measured by thin-layer chromatography with solvent system C of Touchstone and coworkers (34) and phosphate analysis (35) is given in Table 1. The phospholipid headgroup distribution in PPL and CLSE are equivalent, whereas the anionic PG and PI content in mPPL is reduced more than 30-fold.

Surfactant Proteins SP-B and SP-C

Mixed hydrophobic SP were isolated from the lipids in CLSE by two passes through an LH-20 column with nonacidified C:M (1/1) as the elution solvent (31). No phospholipid was detected in final SP isolates by phosphorus assay (35), and both SP-B and SP-C were present by ELISA, SDS-PAGE, and terminal amino acid analysis.

Reconstitution of Phospholipid-Protein Mixtures

Phospholipid-protein mixtures were formed by combining 1.3% by weight of purified hydrophobic SP with PPL, mPPL, or synthetic 1,2 dipalmitoyl-sn-3-phosphocholine (DPPC), the latter obtained from Avanti Polar Lipids (Alabaster, AL). Phospholipids in chloroform and proteins in C:M were mixed together in a tube, and solvents were evaporated under nitrogen. Dried material was either dissolved in a spreading solvent of 9:1 vol/vol hexane:ethanol (HPLC grade) for Wilhelmy balance studies or dispersed in buffered saline (10 mM HEPES, 1.5 mM CaCl₂, and 150 mM NaCl at pH 7.0) for adsorption, oscillating bubble, or excised lung experiments. Dispersion was by probe sonication (Model W-220F; Heat Systems, Plainview, NY) at a power of 40 watts, applied in three to five 15-s bursts to dispersions in an ice bath. Highly pure distilled and deionized water (Milli-Q UV Plus system; Millipore Corp., Bedford, MA) was used for all dispersions, subphases, and cleaning procedures.

Wilhelmy Surface Balance Methods

Surface pressure-area (-A) isotherms for dynamically cycled films were measured with a modified Wilhelmy balance incorporating a teflon ribbon barrier (36). Surfactants dissolved in hexane:ethanol (9:1, vol/vol) were spread dropwise from a syringe at the surface of a buffered saline subphase (see above) in the Wilhelmy balance. Dynamic cycling of the spread film was initiated after a 10-min pause to allow full evaporation of the solvent. Surface pressure (the amount by which surface tension was lowered below that of the subphase) was measured during cycling from the force on a sandblasted platinum slide dipped into the surface film enclosed by the continuous teflon ribbon. Film leakage was absent in all reported experiments as assessed by a second slide outside the barrier. -A behavior was measured for seven successive cycles of compression-expansion between maximal and minimal areas of 448 and 103 cm² (compression ratio, 4.35:1) at speeds of 1.5, 5, or 10 min per cycle. Temperature was fixed at 23 ± 1 or 37 ± 0.5° C, and humidity was maintained at the fully saturated value by open dishes of water and dampened blotting paper in the balance chamber. Respreading was determined by calculating in arbitrary but consistent scale the -A isotherm area under the second (or seventh) compression curve minus the area under the first compression curve (10), as a modification of the collapse plateau ratio criteria of Notter and coworkers (21, 37, 38). The resulting area between Compressions 2 and 1 (or 7 and 1) gave a measure of film material present on the first compression that failed to respread to participate in surface tension lowering on the second (or seventh) compression (10). An area of zero between Compressions 2 and 1 or 7 and 1 indicated complete respreading, and increased area equated to worse respreading. The poor respreading of pure DPPC films provided an effective upper limit on calculated areas. For -A isotherms where the surface pressure during Compressions 2 or 7 exceeded that found on Compression 1 at the same trough area, only the isotherm region below the first compression curve was included in respreading calculations (10).

Adsorption Methods

Adsorption was measured at 37 ± 0.3° C for surfactant dispersions in a Teflon dish containing a 70-ml buffered saline subphase (see above) stirred continuously by a Teflon-coated stirring bar and magnetic stirrer to minimize diffusion resistance (11, 29, 39). Adsorption experiments were initiated at time zero by injecting a bolus of surfactant dispersed by sonication at a concentration of 1.0 or 4.5 mg phospholipid/ 10 ml buffer into the 70-ml stirred subphase, and surface pressure was followed as a function of time from the measured force on a partially submerged, sandblasted platinum slide (11, 29, 39).

Oscillating Bubble Methods

The minimum surface tension reached by surfactant dispersions during rapid cycling at 37° C was measured with a pulsating bubble surfactometer (Electronetics Corp., Amherst, NY) based on the original design of Enhorning (40). A small air bubble, communicating with ambient air, was formed in 40 µl of a buffered dispersion of surfactant in a plastic sample chamber. The bubble was then pulsated at 20 cycles/min between maximal and minimal radii of 0.55 and 0.4 mm, respectively, and the pressure drop across the air-water interface was measured by a precision pressure transducer. Surface tension at a minimal radius was calculated as a function of time from the spherical form of the Laplace equation as recommended elsewhere (41).

Excised Lung Methods

The physiologic activity of surfactant mixtures was determined by measurements of P-V deflation mechanics in excised, lavaged rat lungs at 37 ° C (42). Lungs from adult Sprague-Dawley male rats weighing 550 to 600 g (Charles River, Wilmington, MA) were excised, degassed under vacuum, and rapidly inflated with air to 30 cm H₂O pressure. Air was intermittently delivered to maintain the lungs at this pressure during a stress relaxation period of  10 min to define TLC, and surfactant-sufficient P-V behavior was then measured during deflation from TLC at a rate of 2.47 ml/min. The lungs were then made surfactant-deficient by  15 lavages with 0.15 M NaCl, followed by degassing, reinflation and stress-relaxation at 30 cm H₂O, and measurement of a second P-V deflation curve to define surfactant-deficient mechanics. A surfactant mixture containing 30 mg phospholipid/ 2.5 ml buffered saline was then instilled into the surfactant-deficient lungs, followed by identical procedures to determine the restoration of P-V mechanics toward their original level.

Statistical Analyses

All data are in general expressed as mean ± SEM for the number of independent surface or excised lung experiments. Surface activity data at fixed times were considered significantly different if the probability of the null hypothesis was < 0.05 by Student's t test. Statistical analysis of the effects of different surfactants in excised lungs used one-way analysis of variance (ANOVA), with Scheffe's procedure applied at points of significant difference (p < 0.05).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Adsorption (-time) isotherms for dispersions of PPL and mPPL alone or combined with 1.3% hydrophobic SP are shown in Figure 1 at two different phospholipid concentrations. The lack of anionic phospholipids in dispersions containing mPPL gave rise to only small differences in adsorption relative to comparable dispersions containing PPL at low phospholipid concentration (Figure 1, top panel), and differences became even smaller when phospholipid concentration was raised (Figure 1, bottom panel). At a low concentration of 0.0125 mg phospholipid/ml subphase, PPL had a slightly higher surface pressure than did mPPL after 30 min of adsorption (15.2 ± 0.4 versus 12.8 ± 1.1 mN/m, Figure 1, top panel). Combination with hydrophobic SP resulted in a substantial increase in adsorption in PPL:SP and mPPL:SP relative to the phospholipids alone, but both SP-containing mixtures had relatively similar adsorption behavior. Adsorption was slightly more rapid for PPL:SP than for mPPL:SP over the first 20 min, but both mixtures adsorbed to the same equilibrium adsorption surface pressure of 47 mN/m after 25 min (Figure 1, top panel). At a higher phospholipid concentration of 0.0563 mg/ml, there was very little difference between the adsorption of PPL versus mPPL or of PPL:SP versus mPPL:SP (Figure 1, bottom panel).

View larger version (25K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   Figure 1.   Adsorption -time isotherms for dispersions of PPL or mPPL with and without added hydrophobic surfactant proteins. Surfactant dispersions were injected at time zero below the surface of a well-stirred buffered subphase (150 mM NaCl, 10 mM HEPES, and 1.5 mM CaCl2 at pH 7.0) at 37° C. Final surfactant phospholipid concentration was 0.0125 mg/ml (top panel ) and 0.05625 mg/ml (bottom panel ). Data are mean ± SEM for n = 4.

To complement adsorption studies on dispersions, spread films of PPL versus mPPL and of PPL:SP versus mPPL:SP were examined in Wilhelmy balance experiments at room temperature (23° C) and at body temperature (37 ° C). The -A isotherms of monolayer films of PPL and mPPL spread initially to 120 Å²/molecule were very similar at room temperature (Figure 2), with only minor differences in measured molecular areas or in collapse surface pressures (Table 2). The -A isotherms of films of PPL versus mPPL spread initially to a high surface-excess concentration of 15 Å²/molecule were also essentially equivalent (data not shown), as were the -A isotherms of surface-excess films of PPL:SP versus mPPL:SP (Figure 3). Maximum surface pressures in surface-excess films of PPL:SP were not statistically different from those found in films of mPPL:SP on the same compression over seven consecutive interfacial cycles (Figure 4), and a similar degree of equivalence in maximum pressures was present during cycling of PPL versus mPPL (data not shown).

View larger version (21K): [in this window] [in a new window]   Figure 2.   Surface pressure-area isotherms for spread monolayers of PPL and mPPL. The initial compression-expansion is shown for films spread in a Wilhelmy balance to a dilute initial area of 120 Å2/phospholipid molecule at 23° C. Rate was 10 min/complete cycle, with arrows indicating direction of area change (compression/ expansion). Subphase as in legend to Figure 1. Isotherms shown are representative, from three or more closely reproduced experiments.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Surface pressure-area isotherms of spread surface-excess films of PPL or mPPL combined with 1.3% hydrophobic SP. Shown are the first, second, and seventh compressions, and the first expansion, for films in 9:1 (vol:vol) hexane:ethanol spread in a Wilhelmy balance to a high initial concentration of 15 Å2/phospholipid molecule and cycled continuously at 1.5 min/cycle at 37° C. Subphase as in legend to Figure 1. Isotherms shown are representative, from four or more closely reproduced experiments.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Maximum surface pressures as a function of cycle number for spread films of PPL:SP and mPPL:SP. Films were spread to 15 Å2/phospholipid molecule and cycled continuously on buffer at pH 7.0 at 1.5 min/cycle at 37° C. Maximum pressures were not significantly different on any of seven consecutive compressions studied. Data are mean ± SEM for n = 4 or more.

Another parameter investigated in dynamic film studies was respreading based on -A isotherm areas between Compressions 2 and 1 and Compressions 7 and 1 at 23° C and 37 ° C (Table 3). At both temperatures, respreading in surface- excess films PPL and mPPL was significantly better than DPPC, and respreading was improved still further by the presence of hydrophobic SP in films of PPL:SP and mPPL:SP (Table 3). However, there were relatively minor differences in respreading between comparable surface-excess films containing the complete mix of zwitterionic and anionic surfactant phospholipids in PPL relative to those containing minimal anionic constituents in mPPL. At room temperature, respreading between Compressions 2 and 1 was better in films of PPL versus mPPL and PPL:SP versus mPPL:SP, but isotherm area differences became much smaller by Compression 7 (Table 3, 23° C). At body temperature, films of PPL, mPPL, PPL:SP, and mPPL:SP all showed excellent respreading, with only slight differences in calculated isotherm areas found for comparable films of PPL versus mPPL and PPL:SP versus mPPL:SP throughout seven interfacial cycles (Table 3, 37° C).

To complement adsorption and Wilhelmy balance studies, the ability of aqueous dispersions of different surfactant dispersions to lower surface tension under rapid dynamic compression was defined on a pulsating bubble surfactometer (37° C, 20 cycles/min, high humidity). Measurements of surface tension at minimal bubble radius during cycling on this instrument reflect the combined influence of adsorption and dynamic film compression at rapid rate, giving an overall assessment of surface tension-lowering ability particularly useful in correlations involving the physiologic activity of surfactants in the lungs (3, 40, 41). In bubble experiments, dispersions of PPL and mPPL had essentially equivalent surface tension-lowering behavior, reaching minimum surface tensions of ~ 20 mN/m during cycling (Figure 5). Combination of hydrophobic SP with phospholipids in PPL:SP and mPPL:SP greatly improved surface tension-lowering ability on the bubble apparatus, but both SP-containing mixtures reached the same minimum surface tension of < 1 mN/m over a very similar time course (Figure 5).

View larger version (25K): [in this window] [in a new window]   Figure 5.   Surface tension lowering under rapid dynamic compression on an oscillating bubble apparatus for PPL or mPPL alone and combined with mixed SP-B/C. Surfactant phospholipid concentration was uniform at 0.5 mM in buffer at pH 7.0. Cycling rate 20 cycles/min; temperature, 37° C. Data are mean ± SEM for n = 4 or more.

A final set of experiments utilized P-V mechanical studies in lungs excised from adult rats to provide a more direct measure of the physiologic activity of surfactant mixtures containing PPL or mPPL combined with 1.3% by weight hydrophobic SP (Figure 6). Rat lungs were characterized for quasi-static P-V deflation behavior immediately after excision to define normal mechanics, after depletion of endogenous surfactant by multiple lavage to define surfactant-deficient mechanics, and again after instillation of PPL:SP or mPPL:SP. The activity of instilled CLSE, which has been studied previously in the excised rat lung model (3, 28, 42), was also determined. The activity of all three surfactant mixtures on quasi-static deflation mechanics was found to be quite similar at the experimental dose of 30 mg phospholipid/2.5 ml saline (~ 50 to 55 mg/kg), restoring the P-V mechanics of the surfactant-deficient lungs almost to initial, prelavage levels (Figure 6). The only statistically significant difference in the effects of the three mixtures on mechanics was at zero transpulmonary pressure for mPPL:SP versus CLSE, where p  0.05. There was no statistically significant difference between the effects of PPL:SP and mPPL:SP in restoring P-V mechanics by one-way ANOVA.

View larger version (29K): [in this window] [in a new window]   Figure 6.   Effect of instilled dispersions of mPPL:1.3% SP, PPL:1.3% SP, and CLSE on pressure-volume deflation mechanics in surfactant-deficient excised rat lungs. P-V data during deflation were obtained for freshly excised lungs (surfactant-sufficient), after surfactant depletion by multiple lavage, and again after surfactant instillation (see METHODS). Instilled dose was uniform at 30 mg phospholipid/2.5 ml of buffered saline. *p < 0.05 between mPPL:1.3% SP-B/C and CLSE by one-way ANOVA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has investigated the relative importance of zwitterionic and anionic phospholipids in the surface and pulmonary mechanical activity of CLSE, which contains all the hydrophobic constituents of endogenous surfactant. Experiments measured the relative activity of the complete set of calf surfactant phospholipids in PPL, with and without added hydrophobic SP, in comparison with otherwise equivalent mixtures containing mPPL from which > 97% of the acidic phospholipids had been removed. The use of chromatographically purified PPL and mPPL subfractions allowed the relative importance of zwitterionic and anionic phospholipids in the surface-active function of CLSE to be assessed directly without recourse to simplified model mixtures. Relevance for lung surfactant activity in vivo was also enhanced by measuring a range of adsorption and dynamic film behaviors, supplemented by P-V mechanical studies in excised, surfactant-depleted rat lungs. Results were consistent in showing that zwitterionic phospholipids had a predominate influence compared with anionic phospholipids in the surface-active function of CLSE. Minimal or only minor differences were found between PPL and mPPL in adsorption behavior, in -A isotherms and film respreading on the Wilhelmy balance, and in overall dynamic surface tension-lowering ability in pulsating bubble studies at physiologic compression rates (Tables 2 and 3 and Figures 1, 2, 4, and 5). Mixtures of PPL:SP and mPPL:SP also had very similar adsorption and dynamic surface activity (Table 3 and Figures 1, 3, 4, and 5), and were essentially equivalent to each other and to CLSE in restoring P-V deflation mechanics almost to normal in lavaged excised lungs at a dose of 30 mg phospholipid (Figure 6). Although effects from the trace of anionic species left in mPPL cannot be completely ruled out, these data clearly indicate that zwitterionic phospholipids are major contributors relative to anionic phospholipids in the surface-active function of the hydrophobic constituents of calf lung surfactant.

In light of their functional importance, it is not surprising that zwitterionic phospholipids make up such a large proportion of total phospholipids in endogenous pulmonary surfactant (4). On the basis of molecular headgroup, zwitterionic PCs are by far the largest biochemical class of phospholipids in CLSE and endogenous lung surfactant as noted earlier (4), although small amounts of PE and sphingomyelin are also present (cf, Table 1). Biophysical properties as well as stoichiometry are consistent with a major contribution from DPPC and the spectrum of additional saturated and unsaturated PCs in pulmonary surfactant to the surface-active behavior of this complex material. The excellent surface tension lowering ability of DPPC, aided to some degree by other rigid saturated PCs in lung surfactant (6), is directly complemented by the ability of fluid, unsaturated PCs to facilitate adsorption (11, 39, 43) and improve film respreading (22, 38). These latter behaviors may also be contributed to by additional secondary phospholipids having non-PC headgroups (11, 44), but the secondary PCs should be particularly important both because of their greater relative concentration and their potential for specific interactions and miscibility with DPPC because of their identical headgroup.

Our data do not preclude the possibility that anionic phospholipids may have functionally important roles in the surface-activity of endogenous surfactant containing SP-A as opposed to the hydrophobic CLSE preparation studied here. Acidic surfactant phospholipids, specifically PG, have been suggested as interacting with SP-A and SP-B in the formation of tubular myelin (45), a possible precursor to the surface- active monolayer in vivo. Our findings also do not rule out additional surfactant-related roles for anionic phospholipids in vivo, including specific participation in some aspect of surfactant metabolism or its regulation. However, the utility of anionic phospholipids as biochemical markers of lung maturity (12) does not by itself imply a role in surface-active function. Also, the existence of biophysical interactions between anionic phospholipids and surfactant proteins (17) does not necessitate a functional importance unless these interactions are unique among all the hydrophobic lung surfactant components and generate a crucial surface behavior that would otherwise not be present.

Prior studies have not defined unique roles played by acidic phospholipids in the adsorption and film behavior of mixtures containing DPPC and other hydrophobic lung surfactant constituents. Unsaturated acidic phospholipids have been shown to enhance the adsorption of DPPC (46), but this is also true of zwitterionic phospholipids, including unsaturated PCs and PEs (11, 39, 43). The addition of hydrophobic SP to model mixtures containing a range of zwitterionic and anionic phospholipids is also well-known to increase adsorption (2, 3 for review), in conceptual agreement with our results for dispersions of mPPL:SP versus mPPL and of PPL:SP versus PPL (Figure 1). The almost equivalent adsorption found for mPPL:SP and PPL:SP, however, argues against a major influence on this surface property from electrostatic interactions between the anionic headgroups of acidic phospholipids and positively charged residues on the hydrophobic SP (17). Acidic phospholipids clearly have the ability to interact biophysically with all three surfactant proteins, but this is also true for zwitterionic phospholipids. Aside from headgroup-specific interactions, the hydrophobic acyl chain regions common to all phospholipids have significant interactional affinity for surfactant proteins such as SP-B and SP-C. In fact, phospholipids, including DPPC and unsaturated PCs, have extensive surface-active interactions even with nonspecific hydrophobic peptides such as homopolymers of poly-Phe and poly-Leu (47). Mixtures of DPPC with model amphipathic -helical peptides have also been shown to have substantial surface and physiologic activity in the absence of acidic phospholipid components (48, 49), and the insertion of unsaturated PC species into the interface can be facilitated by hydrophobic SP without the presence of acidic phospholipids (50).

Both zwitterionic and anionic phospholipids are also known to interact with hydrophobic SP in surface films to influence surface tension-lowering and respreading facility. In pure films, DPPC lowers surface tension to < 1 mN/m dynamic compression, but it respreads poorly on repetitive cycling at temperatures below its gel-to-liquid crystal transition of 41° C (37, 38). Respreading can be increased by forming mixed films containing DPPC plus unsaturated phospholipids with either zwitterionic or anionic headgroups (21, 38). In addition to being influenced by headgroup size and charge (44), respreading in phospholipid films is very strongly affected by fatty chain fluidity. Unsaturated fatty chains have increased fluidity and sweep out larger interactional cross sections than do saturated chains, leading to poorer packing and increased disorder within the film and in associated collapse phases above or below the interface. Interactions and molecular transport between these structures and the cycled film largely determine respreading characteristics. The majority of secondary unsaturated and saturated PCs in lung surfactant are in the fluid liquid-crystal state at body temperature (6), and these secondary PCs are much more abundant than acidic phospholipids. Although the secondary phospholipids as a group have much better film respreading than DPPC (10), the relatively similar respreading in PPL versus mPPL films found here indicates that the secondary zwitterionic phospholipids are responsible for the majority of this effect (Table 3). The similar respreading in films of PPL:SP versus mPPL:SP (Table 3), particularly at body temperature, further suggests that unique interactions between anionic phospholipids and hydrophobic SP are not required for facilitating the film respreading of the hydrophobic constituents of lung surfactant. Studies on phospholipids in the lung extracellular lining by electron paramagnetic resonance have demonstrated that unsaturated PCs as well as PGs have a pronounced effect on increasing the fluidity of DPPC (51), consistent with an action of improving respreading.

Because the ultimate importance of lung surfactant involves its effects in intact lungs, biophysical measurements of surfactant activity were complemented with physiologic studies on P-V mechanics in excised, lavaged rat lungs (Figure 6). Results in these excised lung studies correlated very well with the interfacial data. Instillation into surfactant-depleted lungs of mPPL:SP, which lacked anionic phospholipids, restored quasi-static P-V deflation mechanics almost to the same level as did mixtures of PPL:SP that contained both zwitterionic and anionic phospholipids (Figure 6). The restoration of P-V mechanics toward normal after instillation of either mPPL:SP or PPL:SP was very similar to that accompanying instillation of the parent CLSE extract, which has been shown previously to be highly active in improving P-V mechanics in this model (29, 42). Several groups have shown previously that pulmonary mechanics in animals are not substantially altered from normal by dietary manipulations to lower PG content in endogenous surfactant while raising PI content (26, 27). The beneficial mechanical results found here for mPPL:SP (Figure 6) indicate that neither PG nor PI are required in order for the hydrophobic components of lung surfactant to generate substantial physiologic activity in lungs. Although not incorporating interactions involving SP-A, hydrophobic organic solvent extracts such as CLSE are highly active in exogenous surfactant replacement therapy for RDS in premature infants (3, 28). CLSE does not form tubular myelin (52), but nevertheless it can be dispersed to display high surface and physiologic activity approaching that of endogenous surfactant (3, 28, 29).

In summary, chromatographically purified subfractions of CLSE were used to study the relative importance of zwitterionic versus anionic phospholipids in functionally relevant surface properties in vitro and on mechanical activity in excised lungs in situ. PPL, the complete mix of zwitterionic and anionic lung surfactant phospholipids, had very similar adsorption, dynamic film respreading, and surface tension-lowering when compared with mPPL, depleted more than 30-fold in anionic constituents. There were also only minor differences found between the adsorption, dynamic film properties, and surface tension-lowering ability of mixtures containing PPL versus mPPL combined with 1.3% of mixed hydrophobic SP. Dispersions of PPL:SP and mPPL:SP both adsorbed rapidly and equivalently to the air-water interface and formed films having excellent and very similar respreading and maximum surface pressure during continuous cycling at the air-water interface in a Wilhelmy balance at 37° C. Dispersions of PPL:SP and mPPL:SP also reduced surface tension to < 1 mN/m over an identical time scale during rapid cycling on a oscillating bubble (37° C, 20 cycles/min). The effects of instilled mixtures of mPPL:SP and PPL:SP on P-V mechanics in excised, lavaged rat lungs were also equivalent, with both approaching the high activity of CLSE at an instilled dose of 30 mg phospholipid. In aggregate, these data indicate that zwitterionic phospholipids are predominate over anionic phospholipids in the surface-active function of the hydrophobic components of calf lung surfactant present in CLSE.