Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Pulmonary surfactant, a mixture of phospholipids (PL) and proteins lining the alveolar epithelium, is essential for normal alveolar function (1). Phosphatidylcholine (PC) of whole lung is highly saturated and largely reflects the composition of alveolar surfactant and of type II alveolar epithelial cells that synthesize surfactant (1). The presence of PL and protein in bronchial secretions has been well documented, and one of their major postulated functions in such secretions is to inhibit adhesion of the ciliae to the mucous gel and, by accelerating ciliary beat frequency, to make the gel phase glide on top of the sol phase toward the pharynx (2). However, the source of this material is not known. Although it is clear that no cell type of the conductive airway epithelium contains lamellar inclusion bodies characteristic of surfactant stored in type II alveolar epithelial cells, Clara cells synthesize surfactant proteins A, B, and D (SP-A, SP-B, and SP-D) (7). No previous analysis has characterized conductive airway surfactant comprehensively in terms of PL classes, PC molecular species, surfactant proteins, and surface-tension function. The first aim of this study was therefore to compare in detail the PL and PC molecular species composition of tracheal aspirate samples with that of bronchoalveolar lavage (BAL), as well as that of surfactant purified from tracheal aspirates (Surf_TrachAsp) with those of two fully active surfactant preparations from BAL (Surf_BAL) and the 27,000 × g pellet of BAL (Surf_P27000). The reasoning for this was that if these compositions differed significantly, then they might have different cellular sources. Conversely, if Surf_TrachAsp and surfactant preparations from BAL were similar, this result would be consistent with the concept that airway surfactant is derived from overflow of alveolar material. To assess the activity of conductive airway surfactant as a potentially physiologically important component of pulmonary surfactant, we included functional analysis of Surf_TrachAsp in comparison with Surf_BAL and Surf_P27000 in this study. SP-A, SP-B, and SP-C concentrations were measured in samples to provide evidence for mechanisms of any differences in surface-tension function between Surf_TrachAsp and alveolar surfactant.

The second aim was to investigate the relationship between the PL compositions of the conductive airway epithelium and those of both tracheal aspirates and Surf_TrachAsp. The cell type-specific composition of membrane PL is determined by a combination of phenotypic expression and cellular lipid nutrition. Although eicosanoids were described in airways as inflammatory mediators (10, 11), the molecular compositions of major PL components in airway epithelia as precursors for eicosanoids have never been determined. Presumably, bronchial and alveolar epithelial cells are exposed to comparable parameters of cellular lipid nutrition derived from the circulation, and any differences in PL composition will be due to cell-specific aspects of PL synthesis and turnover. We therefore investigated the biochemical composition of PL classes and PC molecular species in tracheal aspirates and Surf_TrachAsp of adult pigs in comparison with the underlying epithelium, BAL, surfactant preparations from BAL (Surf_BAL, Surf_P27000), and lung parenchyma.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

Macroscopically healthy pig lungs were obtained from local slaughter facilities. High-performance liquid chromatography (HPLC)-grade solvents were supplied by Baker (Deventer, the Netherlands). All other solvents and chemicals were of analytical grade and from various commercial sources.

Harvesting of Raw Tracheal Aspirates, Bronchoalveolar Lavage Fluids, and Tissue Specimens

For harvesting tracheal aspirates, the esophagus and heart were removed from the lungs, and the trachea was cut open along the membranaceous part. Tracheal mucus was aspirated by means of a disposable syringe, the trachea rinsed three times with 2 ml of 154 mmol/liter NaCl, and the aspirate then added to the original aspirate sample (total volume: 6 to 8 ml). A 6.5-mm blockable catheter was then placed in the periphery of the left lung and the lower lobe was lavaged three times with a total volume of 500 ml of 154 mmol/liter NaCl (recovery: 350 ml). For the removal of cells, BAL fluid was centrifuged for 15 min at 270 × g and 4°C, and the pellet was discarded.

Cell-free BAL fluid was then centrifuged for 3 h at 27,000 × g and 4°C in a Dupont Sorvall Rc-5B centrifuge (Du Pont De Nemours, Bad Homburg, Germany) to generate a raw surfactant pellet (P₂₇₀₀₀; 15 to 25 µmol PL/ml). The resulting 27,000 × g supernatant was then recentrifuged at 60,000 × g for 1 h. The 60,000 × g supernatant (S₆₀₀₀₀) was collected and the 27,000 × g and 60,000 × g pellets (P₂₇₀₀₀, P₆₀₀₀₀) were resuspended in 154 mmol/liter NaCl/1.5 mmol/ liter CaCl₂ for biochemical and functional analysis (see the subsequent Discussion). P₂₇₀₀₀, P₆₀₀₀₀, and S₆₀₀₀₀ contained 74.1 ± 3.7 mol%, 3.8 ± 1.2 mol%, and 22.1 ± 2.9 mol% of the total BAL phospholipids, respectively. Functional analysis revealed that the active surfactant was harvested from BAL as P₂₇₀₀₀, since it reached surface tensions below 5 mN/m in a pulsating bubble surfactometer (Electronetics Co., Amherst, NY) whereas P₆₀₀₀₀ did not.

For analysis of lung parenchyma, 2 to 3 g of lavaged and nonlavaged lung parenchyma was cut from the left and right lower lobe with a pair of scissors. Airway mucosa was prepared from the underlying tissue at the level of the trachea, carina, and peripheral bronchi, carefully avoiding any contamination of smaller conductive airway specimens by lung parenchyma (Figure 1). All samples were stored at 20°C until further analysis. Additional samples for histology were obtained from the trachea, carina, and bronchi, using the same procedure. The samples were fixed in 4% formalin, dehydrated, and embedded in paraffin. Sections were prepared and stained with hematoxylin/ eosin (H&E) for light microscopy.

View larger version (70K): [in this window] [in a new window]   Figure 1.   Conductive airway mucosa for PL analysis. Microphotographs of H&E-stained sections of the mucosa from: (a) trachea, (b) carina, and (c) bronchus of pig respiratory tract. All specimens include the epithelium (E) and lamina propria (LP). (a and b) Broad abundance of glandular tissue can be seen. (c) Smooth-muscle tissue is included. Bar represents 200 µm.

Surfactant Preparation

Surfactant was prepared from tracheal aspirates, original BAL, and P₂₇₀₀₀ according to the method of Shelley (12), and was designated Surf_TrachAsp, Surf_BAL, and Surf_P27000, respectively. Briefly, a lower phase of 16% NaBr was generated by mixing a 3.75-ml sample with 1.25-ml of 64% NaBr in 154 mmol/liter NaCl. This solution was overlayered sequentially with 5 ml 13% NaBr (wt/vol) in 154 mmol/liter NaCl, and then with 2.5 ml of 154 mmol/liter NaCl. After ultracentrifugation for 1.4 h at 114,000 × g and 4°C, the surfactant in the samples accumulated as a visible white band between the 13% NaBr and the 154 mmol/liter NaCl layers (density = 1.085 g/ml) (12). The band was aspirated by means of a disposable syringe, resuspended in double-distilled water to a total volume of 25 ml, and centrifuged for 1 h at 27,000 × g and 4°C in a Beckman L8-70 ultracentrifuge using a 70Ti rotor (Beckman Instruments, Mountain View, CA). The pellet was resuspended in 154 mmol/ liter NaCl supplemented with 1.5 mmol/liter CaCl₂ and stored at 80°C until further analysis. Further centrifugation of the resulting supernatant at 27,000 × g for 1 h did not precipitate additional material. Density-gradient residues and supernatants were stored for PL measurements.

Phospholipid Analysis

For PL analysis, lung parenchyma and airway epithelium were extracted with chloroform/methanol according to the method of Folch and colleagues (13), and other samples were extracted according to Bligh and Dyer (14). Lipid extracts were dried under nitrogen and dissolved in chloroform/methanol (10:1, vol/vol). Quantitation of PL phosphorus was done according to Bartlett (15) after digestion of the organic compounds at 190°C for 35 min in the presence of 500 µl 70% perchloric acid (wt/vol) and 200 µl 30% hydrogen peroxide (wt/vol). Distribution of total PL classes was determined by normal-phase HPLC as described previously, using ultraviolet absorption at 205 nm and subsequent fluorescence emission (excitation wavelength = 340 nm, emission wavelength = 460 nm) after postelution formation of mixed micelles in the presence of 1,6-diphenyl-1,3,5-hexatriene (DPH) (16). The fluorescence response was quantitative for the amount of PL, whereas the UV response was proportional to the degree of fatty acyl unsaturation. For the analysis of individual molecular PC species, a PC fraction was isolated from the total lipid extract on a 100-mg Varian Bondelut^® NH₂ disposable cartridge (Varian, Hamburg, Germany). PC molecular species were then resolved by reverse-phase HPLC and eluted PC species were quantified by postelution fluorescent-derivative formation as outlined earlier (17).

Measurement of Surface Tension

For surface-tension measurements, the PL concentration of Surf_TrachAsp, Surf_BAL, or Surf_P27000 was adjusted to 4.00, 1.33, and 0.33 µmol/ml, respectively, with 154 mmol/liter NaCl/1.5 mmol/liter CaCl₂. Adsorption, minimal surface tension (_min), and maximal surface tension (_max) were then determined with the pulsating bubble surfactometer (18). Briefly, a bubble was created in a surfactant suspension at 37°C and the static adsorption determined as the surface tension 10 s after formation of the bubble. The bubble was then pulsated for 5 min at a frequency of 20 oscillations per min between a minimal bubble radius of 0.4 mm and a maximal radius of 0.55 mm. Adsorption was determined after 10 s, whereas _min and _max were calculated, using the LaPlace equation, after 1, 3, 5, 7, 9, 21, 50, and 100 pulsations.

Protein Analysis

Proteins were quantified according to the method of Lowry and colleagues (19) in 96-well microtiter plates at 630 nm, using bovine serum albumin (BSA) as a standard. Sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was done according to Laemmli (20) on a 12.5% polyacrylamide gel with a commercial low-molecular-weight calibration kit (alpha-lactalbumin, 14.4 kD; trypsin inhibitor, 20.1 kD; carbonic anhydrase, 30 kD; ovalbumin, 43kD; albumin, 67 kD; phosphorylase b, 94 kD) (Pharmacia-Biotech, Uppsala, Sweden). After overnight fixation in 30% methanol/10% glacial acetic acid/ 0.8% formaldehyde, the gels were silver-stained according to Bloom (21) and stored in 30% methanol. Immunoblot analysis of SP-A in surfactant samples was done on a 12 % polyacrylamide gel using a 1:1,000 dilution of a polyclonal rabbit anti-rat-SP-A antibody and goat anti-rabbit IgG, conjugated with horseradish peroxidase, as the second antibody. SP-B and SP-C were analyzed from butanol extracts of the surfactant samples. SP-B and SP-C were separated from 200 to 400 nmol PL using a Sephadex LH-60 column (Pharmacia-Biotech), with UV detection of the proteins at 228 nm, essentially as described previously (22).

Statistics

Statistical analysis was done by one-way analysis of variance (ANOVA) and the two-tailed Student's t test, using commercial software (GraphPad InStat Version 1.1, San Diego, CA). Statistical values were corrected for multigroup comparisons (ANOVA) using the method of Bonferroni.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Amount of Phospholipids

The PL concentration was substantially lower in airway mucosa (8.7 ± 1.7, 11.0 ± 1.4, and 11.2 ± 0.4 µmol/g in trachea, carina, and bronchi, respectively) than in nonlavaged lung parenchyma (27.10 ± 1.36 µmol/g, P < 0.001) (Table 1). Representative sections of the airway mucosa samples are shown in Figure 1. Tracheal aspirates and BAL samples contained 1.10 ± 0.17 and 0.23 ± 0.03 µmol PL/ml, respectively (Table 1). Recovery of Surf_TrachAsp by density-gradient centrifugation was 11.6 ± 4.8 mol% of the total PL in tracheal aspirates. By comparison, recoveries of Surf_BAL from BAL and of Surf_P27000 from the highly concentrated P₂₇₀₀₀ were 36.4 ± 5.9 mol% and 92.2 ± 1.4 mol%, respectively. Surf_P27000 contained 68.3 ± 5.2% of total BAL PL.

Composition of Phospholipid/Classes

The PL composition of tracheal aspirate samples was similar to that of BAL and lung parenchyma samples, but not to that of the underlying epithelium (Table 1). Airway epithelium contained less PC and, with the exception of tracheal epithelium, apparently no phosphatidylglycerol (PG). In contrast, the phosphatidylethanolamine (PE) concentration was increased in airway epithelium as compared with both lavaged and nonlavaged lung parenchyma (Table 1). Moreover, the high UV-to-fluorescence ratio (16) indicated an increased concentration of unsaturated fatty acids in PC from airway epithelium, as compared with the highly saturated PC composition of tracheal aspirates, BAL, and lung parenchyma (Table 1).

The concentrations of sphingomyelin (SPH), PE, and phosphatidylinositol (PI) were higher in tracheal aspirates than in BAL fluid (Table 1). In contrast, Surf_TrachAsp had the same composition of PL classes as the alveolar surfactant preparations Surf_BAL and Surf_P27000 (Table 2). Although the UV-to-fluorescence ratio of PC was higher in Surf_TrachAsp than in Surf_BAL and Surf_P27000 (Table 2; P < 0.05), the PC composition of Surf_TrachAsp was nevertheless more similar to that of Surf_BAL and Surf_P27000 than to that of the underlying airway epithelium. Therefore, we subsequently reinvestigated our samples for PC molecular species.

Composition of Phosphatidylcholine Molecular Species

The major PC species present in tracheal aspirates were dipalmitoyl-PC (PC16:0/16:0), palmitoyloleoyl-PC (PC16: 0/18:1), palmitoylmyristoyl-PC (PC16:0/14:0), palmitoyl- palmitoleoyl-PC (PC16:0/16:1), and palmitoyllinoleoyl-PC (PC16:0/18:2) (Table 3). This composition was similar but not identical to the PC molecular species of BAL fluid (Table 3). PC16:0/16:0 was the predominant PC molecular species in tracheal aspirates (49.6 ± 0.8%), BAL (56.2 ± 1.2%), and lavaged lung parenchyma samples (32.7 ± 1.2%). This value was significantly lower in tracheal aspirates than in BAL samples (P < 0.001). In contrast, the PC16:0/16:0 content of airway mucosa was considerably lower. Mucosal preparations from trachea, carina, and bronchi, containing both surface epithelium and glandular tissue with minor contaminations from connective tissue (Figure 1), had essentially identical PC compositions, with PC16:0/ 16:0 contributing consistently less than 10% (Table 3). The major species present in airway mucosa were PC16:0/18:1, PC16:0/18:2, and stearoyllinoleoyl PC (PC18:0/18:2). In addition, airway mucosa contained more palmitoylarachidonoyl PC (PC16:0/20:4), stearoylarachidonoyl PC (PC18: 0/20:4), and palmitoyldocosahexaenoyl PC (PC16:0/22:6) than did tracheal aspirates, BAL, and lung parenchyma.

Surfactant isolated from tracheal aspirates and BAL (Surf_TrachAsp, Surf_BAL, and Surf_P27000) contained, respectively, 59.7 ± 1.1, 62.9 ± 1.0, and 63.2 ± 1.8 mol% PC16:0/ 16:0, and there were only minor differences in the concentrations of other PC species (Table 4).

Surface Activity of Surf_TrachAsp, Surf_BAL, and Surf_P27000

To correlate the lipid biochemical findings with functional data, we measured static adsorption, _min, and _max of Surf_TrachAsp, Surf_BAL, and Surf_P27000. Static adsorption of Surf_TrachAsp was much slower than that of Surf_BAL and Surf_P27000 at all concentrations measured (0.33, 1.33, and 4.00 µmol PL/ml, respectively), resulting in higher surface-tension values after 10 s (Table 5). At these PL concentrations, _min after 5 min of cycling was 24.6 ± 2.0 mN/m, 21.8 ± 0.5 mN/m, and 18.2 ± 1.1 mN/m, respectively (Figure 2a). In no analysis of Surf_TrachAsp did _min fall to a value below 5 mN/m. At the PL concentrations evaluated, _max initially ranged from 56.6 ± 1.3 to 69.4 ± 1.6 mN/m (Figure 2b). Values of 35 to 40 mN/m, which are characteristic of alveolar surfactant preparations, were not achieved even at high PL concentrations (4 µmol/ml). Nevertheless, surface tension after static adsorption for 10 s, _min during the first bubble pulsation, and the number of pulsations necessary for achieving stable _min values all remained concentration-dependent with Surf_TrachAsp (Table 5, Figure 2a).

View larger version (44K): [in this window] [in a new window]   Figure 2.   Surface activity of SurfTrachAsp in comparison with alveolar surfactant preparations. (a, c, e) Minimal (closed symbols) and (b, d, f) maximal (open symbols) surface tensions of SurfTrachAsp (a, b), SurfBAL (c, d) and SurfP27000 (e, f) determined at 0.33 (squares), 1.33 (triangles), and 4.00 (circles) µmol PL/ml as described in MATERIALS AND METHODS. Data are means ± SE of four to six experiments. If no error bar is shown, SE was smaller than the symbol.

These characteristics of conductive airway surfactant function were compared with those of alveolar surfactant preparations of the same lungs. In contrast to Surf_TrachAsp, Surf_BAL and Surf_P27000 showed rapid adsorption (Table 5), _min values below 5 mN/m ( Figures 2c and 2e), and _max values around 30 to 40 mN/m (Figures 2d and 2f). All of these parameters were concentration-dependent, and were identical for both alveolar surfactant preparations (Figures 2c through 2f). Even at a concentration of 0.33 µmol PL/ml, minimal surface tension was 15.4 ± 5.3 mN/m and 15.9 ± 2.8 mN/m for Surf_BAL and Surf_P27000, respectively.

Surfactant Protein Analysis

Since the composition of PL classes or PC molecular species did not offer any explanation for the functional differences between Surf_TrachAsp and either Surf_BAL or Surf_P27000, we analyzed the various surfactant preparations for protein concentration and composition. The protein concentration of Surf_TrachAsp (0.28 ± 0.05 mg protein/µmol PL, n = 13) was significantly higher (P < 0.001) than that of either Surf_BAL (0.11 ± 0.01 mg protein/µmol PL, n = 4) or Surf_P27000 (0.10 ± 0.01 mg protein/µmol PL, n = 11). SDS-PAGE analysis (Figure 3a) confirmed that Surf_BAL, Surf_P27000, and P₂₇₀₀₀ all contained two protein bands at 32 to 36 kD, which is the molecular weight range of glycosylated surfactant apoprotein A monomers (1), whereas P₆₀₀₀₀ and S₆₀₀₀₀ (see Materials and Methods) were not. Although these protein bands of 32 to 36 kD were also present in Surf_TrachAsp, other protein bands were more prominent in these preparations than in Surf_BAL and Surf_P27000. We therefore determined SP-A semiquantitatively in Surf_TrachAsp as compared with Surf_P27000 by immunoblot analysis, applying identical amounts of PL (35 nmol) per lane to a 12% polyacrylamide gel. This result (Figure 3b) demonstrated clearly that SP-A is considerably decreased in Surf_TrachAsp as compared with alveolar surfactant (Surf_P27000). Moreover, since surface adsorption strongly depends on the presence of SP-B and SP-C, we analyzed these proteins in Surf_P27000 and Surf_TrachAsp using Sephadex LH 60 gel filtration (Figure 4). Surf_P27000 (n = 3) contained 10.8 ± 1.0 µg SP-B and 16.2 ± 3.6 µg SP-C/ µmol PL. These values are comparable with the respective concentrations in surfactant from BAL of pig, dog, and sheep (data not shown). By contrast, Surf_TrachAsp (n = 3 from two to five pooled samples each) did not contain measureable amounts of SP-B and SP-C (Figure 4).

View larger version (93K): [in this window] [in a new window]   Figure 3.   SDS-PAGE and immunoblot analysis of SP-A in SurfTrachAsp and SurfP27000 preparations. SDS-PAGE was done with (a, b) unstained and (b) prestained molecular weight markers. (a) Separation of the indicated surfactant preparations was done on a 12% polyacrylamide gel and the proteins (5 µg/lane) were silver-stained, as indicated in MATERIALS AND METHODS. The arrow indicates the range of 32 to 36 kD, representing glycosylated surfactant apoprotein A monomer. For abbreviations see text. (b) Blotting was done with a polyclonal rabbit-anti-rat-SP-A antibody and subseqent incubation with horseradish peroxidase-conjugated goat antirabbit antibody. Lane 1: prestained molecular weight standards; Lane 2: 1 µg porcine SP-A; Lanes 3, 5, 7: SurfTrachAsp samples (35 nmol PL applied to gel); Lanes 4, 6, 8: SurfP27000 samples (35 nmol PL applied to gel); Lane 9: 0.1 µg porcine SP-A; Lane 10: unstained molecular weight standards.

View larger version (18K): [in this window] [in a new window]   Figure 4.   Representative HPLC analyses of SP-B and SP-C in SurfTrachAsp compared with SurfP27000. Surfactant was prepared from tracheal aspirates (SurfTrachAsp) and the 27,000 × g pellet of BAL fluid (SurfP27000). The samples were extracted with n-butanol and analyzed for SP-B and SP-C as described in MATERIALS AND METHODS. (a) SurfP27000. (b) SurfTrachAsp. PL = phospholipid.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Our analyses of material from adult porcine lungs showed that the composition of both PL classes and, particularly, PC molecular species in tracheal aspirate samples, were similar to those in BAL, and were markedly different from the PL and PC composition of airway epithelium. The PC of airway mucosa was characteristic of the molecular composition of mucosa from other tissues (23), whereas the tracheal aspirate PC composition was characteristic of lung surfactant (1). The higher content of SPH, PE, and PI in initial tracheal aspirate samples as compared with BAL suggests either a significant contribution of cell membrane material or secretion of material enriched in these phospholipids from the airway epithelium. It is not possible to distinguish between these putative processes, as little evidence has been presented to support active PL secretion in the airways (24). Although the trachea may also contribute some PG to the airway mucus (27), smaller airway epithelia do not contain PG. Since airway mucus and its constituents are obviously transported from more distal parts of the airways by mucociliary activity of the epithelium, PG secretion by the tracheal mucosa is unlikely to be a major contributor to conductive airway surfactant. Moreover, conductive airway surfactant isolated from tracheal aspirate samples by sodium bromide density-gradient centrifugation (Surf_TrachAsp) (12) exhibited a PL and PC molecular species composition essentially identical to those of BAL and of surfactant prepared from BAL (Surf_BAL and Surf_P27000). This result strongly suggests that, whatever their cellular source, the increased contents of SPH, PE, and PI in initial tracheal aspirate samples were not integral components of conductive airway surfactant (Surf_TrachAsp).

The simplest explanation for our results is that Surf_TrachAsp is derived from alveolar material rather than being secreted by airway epithelial or glandular tissue. Although it is evidently not possible to provide conclusive proof of this hypothesis solely by compositional analysis, it is supported by considerable circumstantial evidence. Airway mucosa samples, consisting of surface epithelium together with glandular tissue and minor contaminations by connective tissue, contained only half the concentration of total PL, only minor amounts of PC16:0/16:0 and, with the exception of tracheal epithelium, no PG as compared with lung parenchyma. Instead of the high concentration of PC16:0/ 16:0 measured in all the various surfactant fractions including Surf_TrachAsp, airway mucosal PC was enriched in the mono- and diunsaturated species PC16:0/18:1, PC16:0/ 18:2, and PC18:0/18:2, together with the highly unsaturated species PC16:0/20:4, PC18:0/20:4, and PC16:0/22:6. Consequently, the overall PL content of airway mucosa is unlikely to have been the source of surfactant enriched in PC16:0/16:0, although it remains a potential substrate for release of the eicosanoid precursor arachidonic acid (10, 28, 29). It is, however, possible that a specialized organelle or cell type within the airways may be involved in secretion of such surfactant material. Although small amounts of PL material may be present in the granules and secretions of glandular epithelial cells, specific secretion of PC16:0/16:0 and PG from such cells has never been demonstrated (26, 30). However, since no detailed analyses have been made of the PC molecular compositions of either individual cell types or isolated cellular subfractions, neither possibility can be assessed directly.

Comparison with other cell types that actively secrete PL, such as hepatocytes, type II alveolar cells, and gastric mucosal cells may prove useful in this regard. PL is secreted by hepatocytes both in bile and in lipoprotein complexes (31, 32), and by the gastric mucosa to form a protective hydrophobic barrier (23, 33). In all cases, although a degree of molecular species selectivity has been demonstrated in these regulated secretion processes, the PC species secreted were always also major components of the total cellular PL extract. Airway mucosal PL classes and PC molecular species compositions were, however, directly comparable with those in our previous report of the composition of the gastric mucosa (23), which secretes principally PC16:0/18:1 and PC16:0/18:2 rather than PC16: 0/16:0. In contrast, although secreted alveolar surfactant is enriched in PC16:0/16:0 as compared with lung tissue PC, PC16:0/16:0 remains a major component of PC in this tissue. Because PC16:0/16:0 contributed only 9% to airway mucosal PC, rather than the more than 30% in lavaged lung tissue, this indirect evidence suggests that secretion of PL enriched in PC16:0/16:0 from a putative specialized organelle is unlikely. In support of this view is the lack of lamellar bodies characteristic of surfacant secretion in the airway epithelium and glandular tissue (26, 34). Moreover, significant contributions of a specialized airway cell type to the PC16:0/16:0 content of conductive airway surfactant would imply increased synthesis and storage of PC16:0/ 16:0, which should be detectable by labeling in vivo with radioactive precursors, followed by autoradiography. Although such a specific study has not been performed, the early experiments of Chevalier and Collet (35) that confirmed the type II cell as the cellular source of alveolar surfactant did not report any such cell in conductive airways.

The similarity of both PL classes and PC molecular species compositions of airway mucosa to those of gastric mucosa rather than lung parenchyma suggests that aspects of the metabolism of lipid mediators may be similar in conductive airways and the gastric mucosa. PC from both tissues contains significant amounts of highly unsaturated fatty acids. The present study has determined, for the first time in terms of individual PC molecular species, the PC composition of conductive airway mucosa, and has demonstrated concentrations of PC16:0/20:4 and PC18:0/20:4 identical to gastric mucosal PC (23). Docosahexaenoic acid was additionally present as PC16:0/22:6. These data support the concept that airway epithelia are potential sources for release of the eicosanoid precursor arachidonic acid from PC16:0/20:4 and PC18:0/20:4 (10, 28, 29). Moreover, in addition to cellular PC16:0/20:4 and PC18:0/20:4, the minor amounts of these molecules present in airway mucus may serve as a source for eicosanoid or other mediator synthesis by resident or invading inflammatory cells.

Conductive airway surfactant may be important for mucociliary clearance (2, 4, 6, 36), and an impaired surface-tension function of this material has been associated with pathologic mucociliary clearance in chronic suppurative airway inflammation (41, 42). However, although the PL and PC compositions of conductive airway surfactant (Surf_TrachAsp) were identical to those of original BAL, P₂₇₀₀₀, Surf_BAL, and Surf_P27000, the adsorption velocity and minimal surface tension (_min) of Surf_TrachAsp from the healthy porcine lungs in the present study were impaired as compared with those of Surf_BAL and Surf_P27000. This is in contrast to the situation in full-term newborn infants, in which highly active surfactant can be isolated from pharyngeal aspirates through the same technique as used in our study (43, 44). However, the perinatal period is a different physiologic setting, since at birth considerable amounts of surfactant are squeezed out from the lung periphery rather than transported by mucociliary clearance.

Possible explanations for the differences in surface-tension function in Surf_TrachAsp from adult lungs as compared with alveolar surfactant are that the surface pressure of the surfactant layer at the alveolar air-liquid interface leads to a preferential squeezing out of individual surfactant components from that interface, and that fully active surfactant does not enter the airways. Moreover, surfactant proteins may be retained or degraded in the periphery of the lungs. Electron microscopic studies of adult lungs have shown that the surfactant structures present in the sol phase of the airway mucus are composed primarily of small unilamellar rather than large multilamellar vesicles (2). Within the alveolus, such a distribution would imply a preponderance of surface-inactive surfactant destined for recycling by the type II alveolar epithelial cell (1, 45, 46). Assuming a comparable structure-function relationship in the airways, this could suggest that adult conductive airway surfactant is derived from surface-inactive alveolar surfactant. Our data support this concept, since Surf_TrachAsp shows a different protein composition in terms of decreased SP-A levels and the absence of measurable amounts of SP-B and SP-C. Additionally, the lower surface activity and unilamellar structure of conductive airway surfactant may be due to the presence of inhibitory proteins derived from airway secretions, since the total protein concentration was higher in Surf_TrachAsp than in Surf_BAL and Surf_P27000. This is in accord with earlier reports on the tight association of surfactant to mucus components (40, 47).

In summary, the biochemical results of our study suggest that the mucosal cells of larger conductive airways of adults are probably not major sources of newly secreted intact surfactant particles, although they probably contribute individual PL components to conductive airway secretions. The substantial amounts of PL and PC molecular species recovered from surfactant preparations of airway secretions may have been derived from alveolar surfactant, as suggested by their identical compositions and by the completely different PL and PC composition of the underlying airway mucosa. Nevertheless, final proof of this hypothesis can only be achieved with additional studies, such as with isolated tracheal preparations. The surface-tension function of the conductive airway surfactant was substantially impaired in comparison with alveolar surfactant, even after purification by density-gradient centrifugation, and the concentrations of surfactant proteins SP-A, SP-B, and SP-C were decreased. Tracheal aspirate samples from adult subjects may be an accessible source of conductive airway surfactant (Surf_TrachAsp) with a PL composition comparable to that of alveolar surfactant (Surf_BAL, Surf_P27000). However, although such samples may prove useful for assessing the functional adequacy of conductive airway surfactant, they are unlikely to provide results representative of alveolar surfactant function and protein composition.