Introduction

Abstract Introduction Materials and Methods Results Discussion References

Surfactant protein A (SP-A) is a major constituent of pulmonary surfactant, a mixture of phospholipids and proteins that reduces surface tension within the alveoli of the lung. SP-A has multiple activities in systems in vitro, including enhancement of phosphatidylcholine (PC) adsorption to an air-water interface (1, 2); extracellular formation of tubular myelin (3, 4); inhibition of surfactant secretion (5, 6); enhancement of lipid uptake into alveolar type II cells (7); and enhancement of phagocytosis (8, 9), bacteria killing (10), and chemotaxis of alveolar macrophages (11). Transgenic mice harboring null alleles for SP-A provide evidence that the gene for SP-A is not essential; mice that fail to express SP-A exhibit almost no abnormality in surfactant secretion or function and appear to have adequate host defense under routine laboratory conditions (12). It is not yet clear whether there are redundant or compensatory systems that execute SP-A function in the absence of this protein.

The gene for SP-A is expressed primarily in alveolar type II cells and bronchiolar Clara cells (15). SP-A is a secreted protein, and the primary translation product contains a signal peptide that directs the protein to the lumen of the endoplasmic reticulum (ER). Subsequently, the protein is transported to the Golgi apparatus, where it undergoes oligosaccharide maturation and perhaps other biochemical modifications (18). It has been generally considered that SP-A is transported to lamellar bodies and stored until appropriate secretory stimuli induce exocytosis. This concept is derived from several observations. There is a high ratio of SP-A/protein in lamellar bodies (LBs) compared with that of the total cell (21). Some immunocytochemical studies also identify SP-A as a major component of LBs (16, 22). SP-A and disaturated PC showed similar in vivo radiolabeling kinetics in type II cells and alveolar lavage fluid (23).

However, there are conflicting reports about the relationship between SP-A trafficking and its accumulation in LBs. Some reports support the idea that the major portion of SP-A is transported to LBs and stored (24). In contrast are reports that SP-A is constitutively secreted via pathways not involving LBs (27). The reason for this discrepancy is not exactly known. We have addressed this issue by separating newly synthesized secreted SP-A, cellular SP-A, and SP-A localized in LBs. In addition, we have used brefeldin A (BFA) to block intracellular protein transport at the level between the ER and the Golgi apparatus (30, 31). Results from these experiments indicate that newly synthesized SP-A initially bypasses the LB compartment but is eventually taken up from the extracellular space and assembled into the secretory organelle.

	Materials and Methods

Abstract Introduction Materials and Methods Results Discussion References

Chemicals

[³⁵S]methionine and [³⁵S]cysteine (Tran³⁵S-label) were purchased from ICN Biochemicals (Irvine, CA). [³H]Methylcholine chloride was purchased from DuPont NEN (Boston, MA). Aprotinin, leupeptin, phenylmethylsulfonyl fluoride (PMSF), and protein A-sepharose CL-4B were from Sigma (St. Louis, MO). BFA was from Epicentre Technologies (Madison, WI). Goat antirabbit immunoglobulin G conjugated with horseradish peroxidase was from Bio-Rad (Hercules, CA). Other common chemicals and equipment for cell culture, sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE), immunoblotting, autoradiofluorography, and other procedures were purchased from Sigma, GIBCO BRL (Grand Island, NY), Novex (San Diego, CA), and Kodak (Rochester, NY).

Cells

Alveolar type II cells were isolated from male Sprague- Dawley rats weighing 180 to 250 g by elastase digestion and metrizamide density-gradient separation as previously described (30). The cells were plated in plastic culture dishes at 3 × 10⁶ cells/ml in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS), or in DMEM lacking methionine for radiolabeling of SP-A. In radiolabeling experiments, the cells were supplemented with 250 µCi of Tran³⁵S-label TM (as [³⁵S]methionine) and incubated in a 10% CO₂ incubator at 37° C for 6 h. For the 24-h incubation, the methionine-cysteine-free DMEM was supplemented with 10% complete DMEM and 10% FBS. The FBS was previously dialyzed against phosphate-buffered saline (PBS). When indicated, BFA in methanol was added to a final concentration of 2 µg/ ml at 15 min before adding Tran³⁵S-label TM. At this concentration of BFA there was no difference in cell viability between treated and untreated cells after 24 h incubation, on the basis of a trypan blue dye exclusion test. For labeling with [³H]choline, the cells were suspended at 3 × 10⁶ cells/ml of DMEM with 10% FBS and 5 µCi/ml of [³H]choline for 6-h labeling, or with 1 µCi/ml of [³H]choline for 24-h labeling.

Isolation of LBs

After the indicated culture time, cells were kept at 4°C. The cells were washed twice with PBS, resuspended in 10 mM N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid) (Hepes) (pH 7.4), and homogenized with a Dounce glass homogenizer (10 strokes with pestle A, 100 strokes with pestle B). When the cells were cultured for 24 h prior to harvesting, they typically adhered to the dish. The adherent cells were washed with PBS and harvested in 10 mM Hepes, using a cell scraper. The previously adherent cells were then homogenized as described previously. The homogenates were immediately adjusted to contain 2 mM MgCl₂ and 0.9 M sucrose, and were loaded on a discontinuous sucrose density gradient (from bottom to top: 1.5 M, 0.9 M, 0.8 M, 0.7 M, 0.6 M, 0.5 M, 0.4 M, 0.3 M, 0.2 M sucrose in 10 mM Hepes/2 mM MgCl₂, pH 7.4). The sucrose gradients were centrifuged at 100,000 × g for 3 h in an SW28 rotor (Beckman) (32, 33). LBs were isolated from the 0.4 to 0.6 M sucrose fractions. After pooling, the LB fraction was diluted in 0.25 M sucrose with 10 mM Hepes/ 2 mM MgCl₂/2 mM CaCl₂, and centrifuged at 100,000 × g for 1 to 2 h. The pellets were resuspended in a small volume of 10 mM Hepes, pH 7.4, and used for further processing. Marker-enzyme analysis of the LB fraction indicated less than 1% contamination by cytochrome c oxidase or nicotinamide adenine dinucleotide phosphate cytochrome c reductase, and less than 5% contamination by alkaline phosphatase. Analysis of proteins from these fractions was performed as described subsequently. Lipids were extracted according to the method of Bligh and Dyer (34).

Immunopurification

Individual 35-mm plastic dishes were used for ³⁵S radiolabeling of 3 × 10⁶ cells in 1 ml of medium. After the indicated labeling time, the cells and medium were separated by low-speed centrifugation, and the cells were washed twice with PBS. Where indicated, an aliquot of the cell suspension was lysed with 0.5 ml of lysis buffer (50 mM Hepes, 150 mM NaCl, 10% glycerol, 3% Triton X-100, 1.5 mM MgCl₂, 1 mM ethyleneglycol-bis-(-aminoethyl ether)- N,N'-tetraacatic acid, 10 µg/ml aprotinin, 1 mM PMSF, and 10 µg/ml leupeptin, pH 7.4), and centrifuged at 100,000 × g for 30 min. The resulting supernatant was referred to as the cell lysate. For analysis of media, PMSF, leupeptin, aprotinin, and Triton X-100 were added to the same concentration as in the lysis buffer. For analysis of LBs, the organelle pellet was lysed with 0.5 ml of the lysis buffer. Small aliquots of cell lysates were used for 10% trichloroacetic acid (TCA) precipitation. The total radioactivity present in TCA-precipitated ³⁵S-labeled proteins was not different for BFA-treated and untreated cells after 6 h incubation, but after 24 h incubation the BFA-treated cells incorporated into proteins 60 to 70% of the ³⁵S incorporated by the untreated cells. The amount of cell lysate used for immunoprecipitation was adjusted to be equal among samples on the basis of TCA-precipitable ³⁵S-radiolabeled proteins, and similar methods were used to adjust sample amounts of medium and LBs. In some experiments, the amount of the LB fraction loaded on electrophoretic gels was greater than the amount of LBs present in a given volume of homogenate loaded on the same gel, and this difference was expressed as a sample-volume factor. For example, a sample-volume factor of 10 for LBs meant that the LB lane of a gel contained 10 times the corresponding amount of LBs present in the homogenate lane of the same gel. For immunoprecipitations, 3 µg of mouse antirat SP-A monoclonal antibody (1D6) (6) was added to the samples and incubated for 2 h at 4°C. Next, 25 µl of protein A-sepharose CL-4B beads were added and incubated for another 2 h, using a rotary shaker. Subsequently, the beads were centrifuged and washed five times with the lysis buffer. Each wash set included a 15-min incubation on a rotary shaker. The antigen-antibody complex on the beads was eluted twice with 20 µl of 2 × SDS-PAGE sample buffer containing 0.4 M dithiothreitol. The samples were electrophoresed on an 8 to 16% Tris-glycine gel. The gels were fluor-impregnated, dried, and used to expose X-ray film. The films were read and quantified with a densitometer (Shimadzu CS900, Kyoto, Japan).

Enzyme-Linked Immunosorbent Assay

SP-A was quantified with a sandwich-type enzyme-linked immunosorbent assay (ELISA) method (6). First, 0.5 µg of rabbit antirat SP-A polyclonal antibody in 100 µl of 0.1 M NaHCO₃ was adsorbed overnight onto wells of Microtiter immunoassay plates (Dynatech, Chantilly, Virginia). The wells were washed and then blocked with 5% skim milk/ 1% Triton X-100/PBS. Several dilutions of samples from sucrose-gradient fractions, medium and cell lysates, and SP-A standards (0 to 50 ng), were added and incubated for 1 h at 37°C. After washing the wells, 2 µg of horseradish peroxidase-conjugated rabbit antirat SP-A polyclonal antibody was added and further incubated for 90 min at 37°C. The wells were next washed with 1% Triton X-100/ PBS. Substrate and buffer (10 mg o-phenylenediamine dihydrochloride, 10 µl of 30% H₂O₂ in 10 ml of 0.1 M citrate buffer, pH 4.6) were added, and color development was allowed to proceed for 5 to 10 min in the dark. The reaction was stopped by adding 2 N H₂SO₄. Product formation was quantified by measuring A490 with an Automated Microplate Reader (Bio-tek, Winooski, VT).

	Results

Abstract Introduction Materials and Methods Results Discussion References

Most Newly Synthesized PC and ELISA-Detectable SP-A Was Colocalized to the LB Fraction of Freshly Isolated Type II Cells

Freshly isolated alveolar type II cells synthesize, transport, and store large amounts of [³H]PC. Figure 1 shows the accumulation of [³H]PC in the LB fraction isolated from cultured type II cells incubated with a [³H]choline precursor for either 6 h or 24 h. LBs initially present in the homogenate at 0.9 M sucrose migrated to the 0.4 to 0.6 M sucrose zone in high-gravitational force fields, as a consequence of their high phospholipid/protein ratio and very low density. The migration position of LBs was essentially identical for both the 6 h and 24 h cultures. The position of the main peak of phospholipid found in these gradients, which contain 2 mM Mg²⁺, was the same as previously reported for gradients that do not contain Mg²⁺ (32).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Accumulation of [3H]PC in the LB fraction. Freshly isolated type II cells were labeled with 5 µCi/ml or 1 µCi/ml of [3H]choline for 6 h (triangles) and 24 h (squares), respectively. Cells were homogenized and separated by discontinuous sucrose density-gradient centrifugation. The radioactivity of [3H]PC in lipid extracts from each gradient fraction was quantified by liquid scintillation counting. A representative one of three experiments is shown.

Total immunoreactive SP-A, including both newly synthesized and stored SP-A, in type II cells was quantified through ELISA, using a rabbit antirat SP-A antibody. Figure 2 shows the distribution of the immunoreactive SP-A in the sucrose-gradient fractions. The LB fraction (0.4 to 0.6 M sucrose gradients) contained 38.5 ± 4.3% (mean ± SE, n = 4) of total cellular SP-A at the 6-h incubation time and 52.9 ± 6.7% (n = 4) of total SP-A at the 24-h incubation time. The difference in the SP-A content in the 6-h and 24-h fractions was primarily due to a reduced SP-A content in 0.9 to 1.5 M sucrose-gradient fractions, which consisted of dense membrane compartments that included the endoplasmic reticulum and Golgi apparatus (32, 33, 35). The lower SP-A content in the high-density fractions at 24 h was probably due to the termination of SP-A synthesis shortly after isolation (36), and transport of SP-A out of these compartments by 24 h. The absolute amount of SP-A in the LB fractions was almost identical in the 6 h and 24 h preparations.

View larger version (15K): [in this window] [in a new window]   Figure 2.   Distribution of SP-A in sucrose gradients measured with ELISA. Freshly isolated type II cells (12 × 106 cells) were cultured in plastic dishes for 6 h (triangles) and 24 h (squares). The cells were washed, homogenized, and separated on sucrose density gradients as in Figure 1. The SP-A concentration in each gradient fraction was quantified with ELISA. A representative one of three experiments is shown.

Initially, we could not detect significant amounts of SP-A in the LB fraction, but detected most of the SP-A in 0.9 to 1.5 M fractions of the sucrose gradients when we did not include divalent cations, according to the original method (32) (data not shown). In preliminary experiments, freshly isolated type II cells were homogenized in 10 mM Hepes, pH 7.4. The sample was adjusted to 0.25 M sucrose and divided into two samples either lacking divalent cations or containing 2 mM MgCl₂, which were centrifuged at 100,000 × g for 1 h. In the absence of divalent cations, 30 to 40% of SP-A was present in the supernatant. Because intracellular SP-A should be compartmentalized within cellular organelles in either the secretory or the endocytic pathways, the amount of the protein found in nonsedimentable form in the absence of divalent cations appears to have been the result of leakage from organelles. Inclusion of 2 mM MgCl₂ immediately after homogenization reduced the percentage of nonsedimentable SP-A to less than 5% of the total, and greatly improved that recovered in the LB fraction. Since Mg²⁺, unlike Ca²⁺, does not promote binding of SP-A to phospholipid, we interpret this result to mean that Mg²⁺ preserves organelle integrity and compartmentalization of SP-A.

Newly Synthesized [³⁵S]SP-A Was Efficiently Secreted and Does Not Localize to the LB

When type II cells were labeled with [³⁵S]methionine and [³⁵S]cysteine for 6 h, only trace amounts of [³⁵S]SP-A were present in the LB fraction, and 17.1 ± 5.1% (mean ± SE, n = 3) was secreted (Figure 3). The secreted [³⁵S]SP-A had a dominant molecular mass of the highly glycosylated form (36 to 38 kD), whereas intracellular [³⁵S]SP-A had a lower molecular mass (26 to 36 kD) consistent with forms containing immature oligosaccharide. The nonglycosylated form of SP-A migrated with an apparent mass of approximately 26 kD, and was detected as a minor component in both media and cells. When cells were incubated with 2 µg/ml of BFA, no [³⁵S]SP-A was recovered from the media. BFA-treated cells also failed to make species of SP-A containing mature oligosaccharides, possibly because of the inability of the protein to reach the trans-Golgi compartment. Thus, the amount of [³⁵S]SP-A in the LB was significantly less than that secreted by the cell. Although there was only a trace amount of newly synthesized [³⁵S]SP-A in the LB fraction after 6 h incubation, as much as 38.5 ± 4.3% (n = 4) of the cellular pool of SP-A was detected in LBs when measured with ELISA (Figure 2).

View larger version (70K): [in this window] [in a new window]   Figure 3.   Trafficking of newly synthesized [35S]SP-A in type II cells radiolabeled for 6 h. Freshly isolated type II cells were labeled with [35S]methionine and [35S]cysteine for 6 h with or without 2 µg/ml of BFA in methionine-deficient DMEM. Media and cells were separated, and the LB fraction was purified by sucrose density-gradient centrifugation as shown in Figures 1 and 2. SP-A was immunoprecipitated from these three samples (media, cells, and LBs), separated on SDS-PAGE, and visualized by autoradiography. Lane 1: medium, BFA; lane 2: medium, BFA+; lane 3: cell lysate, BFA; lane 4: cell lysate, BFA+; lane 5: LB, BFA; lane 6: LB, BFA+. Sample-volume factors were equal between medium samples and between cell samples (lanes 1 to 4), and for LBs (lanes 5 and 6) were 1.1 times that of lanes 3 and 4. A representative one of three experiments is shown.

When freshly isolated type II cells were incubated and labeled with [³⁵S]methionine and cysteine for 24 h, most of the labeled [³⁵S]SP-A (70 ± 10% of the total, in medium plus cells [mean ± SE, n = 4]) was secreted into the medium, and only a minor amount remained in cells (Figure 4). If cells were incubated with 2 µg/ml of BFA, newly synthesized [³⁵S]SP-A was not secreted and remained in the cells. The [³⁵S]SP-A recovered from BFA-treated cells did not show the highly glycosylated form of the protein, a finding consistent with disassembly of the Golgi apparatus. The total amount of [³⁵S]SP-A found in the media plus cells was not different for BFA-treated and untreated cells. The sample volumes loaded on the SDS-PAGE were adjusted on the basis of the amounts of total TCA-precipitable ³⁵S-labeled proteins. Only very low amounts, 6.8 ± 2.9% (mean ± SE, n = 4) and 4.2 ± 2.8% (n = 3), of [³⁵S]SP-A were found in LBs in BFA-treated and untreated cells, respectively.

View larger version (65K): [in this window] [in a new window]   Figure 4.   Trafficking of newly synthesized [35S]SP-A in type II cells radiolabeled for 24 h. Freshly isolated type II cells were labeled with [35S]methionine and [35S]cysteine for 24 h with or without 2 µg/ml of BFA in methionine-deficient DMEM supplemented with 10% complete DMEM and 10% dialyzed FBS. Samples were treated as in Figure 3. Lane 1: medium, BFA; lane 2: medium, BFA+; lane 3: cell lysate, BFA; lane 4: cell lysate, BFA+; lane 5: LB, BFA; lane 6: LB, BFA+. Sample-volume factors were equal between media samples and between cell samples (lanes 1 to 4), and for LBs (lanes 5 and 6) were 1.7 times that of lanes 3 and 4. A representative one of four experiments is shown.

Newly Secreted SP-A Was Internalized by Type II Cells and Selectively Localized to the LB Compartment

Newly synthesized [³⁵S]SP-A, along with other secreted proteins, was collected from the culture media of type II cells incubated with [³⁵S]methionine and cysteine for 24 h. The media, containing [³⁵S]SP-A and other proteins, were centrifuged to remove cells, and were dialyzed against DMEM to remove ³⁵S-labeled amino acid precursors. These [³⁵S]SP-A-containing medium preparations were next incubated for 6 h with primary cultures of nonradiolabeled type II cells that had been maintained in culture for 24 h. In some of the cultures, the medium containing [³⁵S]SP-A was supplemented with 5 µg/ml of purified (unlabeled) rat SP-A. The results of these experiments provided evidence that SP-A secreted from cultured type II cells could be taken up and incorporated into LBs (Figure 5). At the end of a 6-h incubation, the cultured type II cells internalized a trace amount of [³⁵S]SP-A when it was added alone. The uptake of [³⁵S]SP-A could be increased to 4.7 + 1.1% (mean ± SE, n = 3) by the inclusion of 5 µg/ml of purified rat SP-A in the culture medium. Lanes 5 to 7 in Figure 5 contain 10 times the amount of LBs found in Lanes 2 to 4, to enhance visualization. BFA did not inhibit this incorporation of [³⁵S]SP-A into the cells. If the secreted [³⁵S]SP-A and unlabeled SP-A were equally incorporated into the type II cells, it is estimated that approximately 44 ng of SP-A was taken up per 1 × 10⁶ cells during the 6-h period. Analysis of the subcellular distribution of exogenous SP-A on sucrose-density gradients revealed that 84 ± 20% (mean ± SE, n = 3) of the cell-internalized [³⁵S]SP-A was found in the LB fraction. When the [³⁵S]SP-A preparations containing 5 µg/ml of purified SP-A were incubated with type II cells at 4°C, no [³⁵S]SP-A was incorporated into the cells (data not shown). These results indicated that significant amounts of SP-A could be incorporated into LBs from the extracellular compartment.

View larger version (51K): [in this window] [in a new window]   Figure 5.   Uptake of newly synthesized and secreted [35S]SP-A by type II cells. Freshly isolated type II cells were labeled with [35S]methionine and [35S]cysteine and allowed to secrete [35S]SP-A for 24 h. The radiolabeled media were collected, dialyzed, and added to dishes of nonradiolabeled type II cells that had been cultured for 24 h. After addition of the radiolabeled media, cells were incubated for 6 h either with or without 5 µg/ml of exogenous SP-A and 2 µg/ml of BFA. The cells were washed and homogenized, and the LB fraction was purified as described in Figure 1. SP-A was immunoprecipitated from cell homogenates and LB preparations as in Figure 3. Lane 1: radiolabeled media; lanes 2 to 4: cell homogenates; lanes 5 to 7: LB. In lanes 2 and 5, the cultures received no additions to media; in lanes 3 and 6, the cultures received 5 µg/ml of exogenous SP-A; in lanes 4 and 7, the cultures received 5 µg/ml of exogenous SP-A and 2 µg/ml of BFA. Sample-volume factors were equal for lanes 1 to 4, but for lanes 5 to 7 were 10 times that of lanes 1 to 4. A representative one of three experiments is shown.

	Discussion

Abstract Introduction Materials and Methods Results Discussion References

This study shows that SP-A is present in significant quantities in the LB fraction isolated from freshly isolated and 24-h-cultured type II cells. In contrast, newly synthesized SP-A is secreted into the culture medium, with little evidence for its accumulation in LBs. The secreted form of SP-A is primarily the high-molecular-weight form containing mature oligosaccharides, whereas the intracellular forms of the protein contain both immature and mature oligosaccharides. These observations suggest that newly synthesized SP-A is not the immediate source of the protein found in the LB fraction. An alternative source of SP-A in LBs may be the extracellular pool, and the results presented in this and other reports (27, 37, 38) provide evidence that SP-A can be taken up by type II cells and incorporated into LBs. The surprising finding in the present investigation was that the extracellular route may be the predominant pathway for assembly of SP-A into LBs. Although such pathways have been described for lysosomal enzymes, and LBs are modified lysosomes, the apparent complete bypass of the LB is unexpected.

Conflicting data exist in the literature for several of the findings made in this study. Some investigators have found that SP-A is a minor component of LB proteins, and that addition of SP-A to LBs facilitates their rapid transformation to tubular myelin (35, 39). An immunohistochemical study found that LBs contained relatively little SP-A (40). Other investigators, however, using similar immunohistochemical techniques, found significant levels of SP-A in LBs (16, 22, 24). Examination of type II cells with confocal immunofluorescence microscopy showed that SP-A was localized primarily but not exclusively in LBs (41). In preliminary experiments in the present study, we observed that homogenization conditions could have a profound effect on SP-A compartmentalization. In the absence of divalent cations, significant SP-A (as much as 40%) was recovered in high-speed centrifugation (100,000 × g for 1 h) supernatants. LBs contain a high concentration of Ca²⁺ that may contribute to the retention of SP-A (42). The amount of nonsedimentable SP-A may increase with the time of LB preparation (i.e., during homogenization, loading of sucrose gradients, and centrifugation) in the absence of divalent cations, especially if the integrity of the LB is compromised. This is a critical observation, since SP-A is both a secretory and endocytic protein and is always present in membrane compartments that sequester it from the cytosol. Thus, the appearance of soluble SP-A is a likely indicator of loss of compartmental integrity. The simple inclusion of Mg²⁺ immediately after homogenization yields approximately 95% sedimentable SP-A in cell homogenates. Since SP-A binding to lipid is Ca²⁺ (but not Mg²⁺) dependent (43), we interpret this result to mean that the Mg²⁺ probably stabilizes membrane structure and prevents SP-A leakage into the soluble compartment. Our findings suggest that a number of previous observations should be reevaluated with regard to the ionic conditions used for cell and tissue preparations.

The rapid secretion of radiolabeled SP-A from type II cells, and the extremely poor labeling of SP-A in the LB fraction, suggest that a unique process is occurring in these cells. The secretion of newly synthesized SP-A in our experiments appears to be legitimate and not a consequence of artifacts such as cell lysis, because the process was inhibited by BFA. BFA, a fungal metabolite from Eupenicillium brefeldinum, blocks the movement of newly synthesized membrane and secretory proteins from the ER to the Golgi apparatus, through redistribution of cis/medial Golgi components to the ER (30). BFA prevents the assembly of a non-clathrin coat (COPI) and adenosine diphosphate ribosylation factor (ARF)-1 on buds formed from Golgi cisternae membrane (44, 45). Recent studies have clarified that BFA inhibits guanine nucleotide exchange of ARF-1 and ARF-3 by affecting Golgi-membrane-associated guanine nucleotide exchange protein (31, 46, 47). In addition, the secreted SP-A is predominantly the high-molecular-weight form that contains mature oligosaccharides. It is also noteworthy that when cells were treated with BFA for 6 h in our study, the form of newly synthesized [³⁵S]SP-A found in the LB fraction reflected that found in the whole cell (i.e., the lower-molecular-weight proteins that primarily contain immature oligosaccharides). This latter observation suggests that the trace amounts of [³⁵S]SP-A found in LBs are due to low levels of contamination by other organelles. Our findings that SP-A secretion initially bypasses the LB is consistent with other reports that SP-A secretion appeared to be a constitutive process that occurred independently of lipid secretion (27- 29), although these studies did not separate newly synthesized SP-A from other SP-A.

Although our results indicate that nascent SP-A bypasses the LB, they also show that significant amounts of the protein are associated with this organelle compartment. Our results provide clear evidence that secreted [³⁵S]SP-A can be taken up from the medium and incorporated into LBs. This uptake process was significantly augmented by the addition of exogenous SP-A. This latter observation may be related to the positive cooperativity found for SP-A binding to its receptor on type II cells (48). Type II cells, like many primary cells, undergo numerous phenotypic changes upon adaptation to culture (49, 50). Thus, some caution is always warranted in interpreting the behavior of type II cells in vitro. In vivo labeling of preterm lamb lungs with intratracheally instilled [³⁵S]methionine indicated a substantial lag in the appearance of [³⁵S]SP-A in the LB fraction relative to its early appearance in the extracellular compartment (27). Extension of the work in preterm lambs to either newborn or adult rabbits also suggests that newly synthesized SP-A enters the extracellular compartment prior to entering the LB (51). These previous findings make it unlikely that the processes described in this report are merely a consequence of in vitro manipulations of type II cells. Thus, findings in both in vivo studies and in our present in vitro study are consistent with an extracellular routing of SP-A prior to its accumulation in LBs.

The function of the two different pools of SP-A (i.e., newly synthesized and LB-associated) is at present not clear. It is likely that the SP-A associated with LBs remains with the lipid components during exocytosis. In contrast, the newly synthesized and secreted SP-A pool may be relatively depleted of lipid. The "lipid-poor" SP-A may be required for effective tubular myelin formation after interaction with secreted LBs. Alternatively, lipid-poor SP-A may be necessary for efficient uptake of phospholipid for recycling. Although our results do not yet indicate a clear function for the differential compartmentalization of SP-A, they reveal that type II cells utilize a novel strategy for assembling SP-A into LBs.