Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Alveolar type II cells synthesize and secrete pulmonary surface-active material that is required to reduce surface tension at the air-liquid (A/L) interface of the lung, stabilize alveolar units, and diminish the work of breathing. Studies of surfactant secretion in vivo are extremely difficult to perform because of the need for lavage to recover secreted surfactant, and because of the large individual variations in amount of surfactant recovered by lavage in normal animals (1). A great deal of information is available on the regulation of phospholipid secretion because type II cells in primary culture synthesize and secrete phosphatidylcholine (PC), the major lipid component of surface-active material (2, 3). However, very little is known about the regulation of surfactant-protein secretion, since the ability of type II cells to synthesize the proteins of surfactant decreases rapidly when the cells are cultured on tissue-culture plastic containers.

Maintenance of surfactant-protein gene expression in type II cells in vitro has been difficult to achieve. Type II cells plated on plastic tissue culture containers spread rapidly and lose both their cuboidal morphology and the expression of the genes for the surfactant proteins (4, 5). The spreading can be reduced by culturing the cells in the absence of serum, but the ultimate transdifferentiation into flat type I-like cells is not reversed. As the type II cells spread, they acquire some of the morphologic and biochemical features of type I cells (6). However, this process appears reversible because flattened cells can recover surfactant-protein gene expression if they are transferred back to gels made of Engelbreth-Holm-Swarm (EHS) tumor matrix and cultured at an A/L interface (5, 9). Type II cells cultured directly on collagen gels also flatten and lose many of their original phenotypic features (5). In contrast, if collagen gels are detached from the edge of the culture vessel, the type II cells assume a more cuboidal morphology and surfactant-protein gene expression is enhanced. In some culture systems, the collagen gel contains irradiated fibroblasts, and these fibroblasts may promote contraction of the gels and differentiation of the type II cells (4). Another variation of substratum is to culture the cells directly on EHS tumor basement-membrane gels. Although EHS promotes differentiation and maintains surfactant-protein gene expression (10), the cells form closed aggregates in which their apical surface is oriented inward, precluding direct studies of surfactant secretion. Recently, EHS added directly to culture medium was reported to increase the differentiation of hepatocytes (11, 12). Corticosteroids and keratinocyte growth factor (KGF) also promote differentiation of type II cells. Hydrocortisone increases the synthesis of the phospholipid components of surfactant by type II cells cultured on EHS gels in defined, serum-free medium (13, 14), and dexamethasone increases the expression of surfactant protein-A (SP-A) in rat lung in vivo (15). KGF is an epithelial-cell-specific growth factor that is a potent stimulator of type II cell proliferation in vivo (16) and in vitro (17). Recently, KGF has been shown to increase SP-A and SP-B expression in type II cells cultured on EHS gels (18). Additionally, culture of epithelial cells from large airways at an A/L interface promotes differentiation (19, 20), and this type of culture system has been used to express differentiated markers in type II cells in vitro (9).

A major objective of the present study was to develop a long-term culture system for type II cells that would maintain epithelial polarity and provide apical access for secreted proteins. We chose to culture type II cells on a collagen substratum; to form a compact monolayer by increasing the number of type II cells with KGF; to promote differentiation by including EHS, KGF, and hydrocortisone in the media; and to maintain the cells with a limited amount of apical media and to rock the cultures to ensure an A/L interface during culture. This paper describes this culture system and the initial studies of SP-A and SP-D secretion. An abstract of this work has been published (21).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Isolation and Culture of Alveolar Type II Cells

Alveolar type II cells were isolated from specific pathogen-free adult male Sprague-Dawley rats by elastase dissociation and purification on a discontinuous metrizamide gradient as described elsewhere (22). Cells were seeded on a filter insert (Millicell-CM insert, 30-mm diameter; Millipore Corp., Bedford, MA) that had been coated with rat-tail collagen (4). This unit is referred to in this report as a transwell filter. The type II cells were plated at a density of 3.5 × 10⁶ cells/well in 2 ml of Dulbecco's modified Eagle's medium (DMEM)/F12 medium (GIBCO BRL, Gaithersburg, MD) containing 10% fetal bovine serum (FBS) (Irvine Scientific, Santa Ana, CA), 2 mM glutamine, 2.5 µg/ml amphotericin B, 100 µg/ml streptomycin, and 100 U/ml penicillin G (Sigma Chemical Co., St. Louis, MO), and 2.5 ml of medium was added to each well outside the insert. The cells were cultured at 37°C in humidified 5% CO₂ in air. After attachment for 20 to 24 h, the cells were washed and cultured in the same medium, but containing 2% FBS and 5 ng/ml KGF (Promega, Madison, WI). The medium was changed daily until a monolayer of cells was achieved (usually 3 d). For the next 2 d, cells were cultured in medium without serum and KGF, but with the addition of 5% EHS (vol/vol) (~ 650 µg protein/ml), 1 mg/ml bovine serum albumin (BSA), 100 nM hydrocortisone (Sigma Chemical Co.), 5 ng/ml insulin, 5 ng/ml transferrin, and 5 pg/ml selenous acid (Collaborative Biomedical Products, Bedford, MA). We refer to this medium as DF12 plus EHS. On Day 6 of culture, the cells were rinsed with DF12 basal medium and then cultured in 0.4 ml of different media at the A/L interface on a rocking plate. Media were changed every 24 h. The media are designated DF12 (serum-free medium alone), DF12/EHS (DF12 plus 5% EHS), DF12/ KGF (DF12 plus 5 ng/ml KGF), and DF12/EHS/KGF (DF12 plus both EHS and KGF). The EHS component was added only to the 0.4 ml apical medium, and not to the basolateral medium.

RNA Analysis

Total cellular RNA was extracted from cultured cells using 4 M guanidine isothiocyanate, 0.5% sarkosyl, and 0.1 M -mercaptoethanol in 25 mM sodium citrate buffer, pH 7.0 (GITC buffer) (23). Cells from two or three 30-mm millicells were pooled and brought up to a volume of 3 ml with GITC buffer, after which they were vortexed, sheared by three passages through a 26-gauge needle, and stored at 70°C. Each 3-ml RNA sample in GITC was purified by centrifugation through a 2-ml cushion of 5.7 M cesium chloride in 25 mM citrate buffer (pH 7.0) with 0.1 M ethylene diamine tetraacetic acid (EDTA) (pH 8.0). Samples were centrifuged at 150,000 × g for 18 h at 20°C. The resulting RNA pellet was dissolved in ribonuclease (RNase)- free H₂O with 400 U/ml RNasin (Promega, Madison, WI), and was quantitated. Equal loads of RNA were size-fractionated by electrophoresis through a 1% agarose gel under denaturing conditions, transferred to a Nytran membrane (Schleicher and Schuell, Keene, NH) by capillary action overnight, and then dried in a heated vacuum oven. Each Northern blot was prehybridized for 2 h at 42°C in a mixture of 15 ml deionized formamide (IBI, New Haven, CT), 7.5 ml 20× sodium chloride-sodium phosphate-EDTA (SSPE), 3 ml 50× Denhardt's solution, 0.3 ml of 10 mg/ml salmon-sperm DNA (5Prime-3Prime, Boulder, CO), 0.3 ml 10% sodium dodecyl sulfate (SDS), and 3.9 ml RNase-free H₂O. Blots were hybridized for 18 h at 42°C in 15 ml prehybridization buffer with the addition of 10% dextran sulfate (Pharmacia, Uppsala, Sweden) and 5 × 10⁵ cpm/ml of ³²P-labeled probe for SP-A, SP-B, SP-C, SP-D, glyceraldehyde-6-phosphate dehydrogenase (GAPDH), or 28S oligonucleotide (5, 9, 24). All ³²P-labeled surfactant and GAPDH probes were made from inserts (10) by random-primed second-strand synthesis using a commercially available kit (GIBCO BRL). ³²P-labeled 28S oligonucleotide (5, 9, 24) was made with a 5' end-labeling kit (GIBCO BRL). After hybridization, blots were washed four times for 30 min each in 500 ml 2× standard saline citrate (SSC) containing 0.1% SDS at 42°C, once in 500 ml 0.2× SSC with SDS at 55°C, and once in 0.2× SSC without SDS at 55°C for 2 min. For reprobing, blots were stripped for 2 h at 65°C in a mixture of 50% formamide, 30% 20× SSC, and 20% H₂O, and then rinsed in 2× SSC. Autoradiography was done at 70°C, using Kodak XAR film (Kodak, Rochester, NY) in a cassette equipped with intensifying screens. Quantitation was done on a phosphor screen using ImageQuant software (Molecular Dynamics, Sunnyvale, CA).

Sample Collection for Protein Secretion

Media in the inserts were collected every 24 h and centrifuged for 10 min at 800 × g to remove cells. Supernatant was transferred to a new tube and stored at 20°C until assayed.

Antibodies and ELISA for SP-A and SP-D

SP-A and SP-D were quantitated with a standard enzyme-linked immunosorbent assay (ELISA) with rabbit antirat SP-A and antirecombinant SP-D polyclonal antibodies as primary antibodies, as reported elsewhere (25). Native rat SP-A and recombinant rat SP-D expressed in Chinese hamster ovary (CHO) cells were employed as standards.

Lysozyme Assay

Regulated Phosphatidylcholine and Protein Secretion

Cells were labeled with [³H]choline overnight on Day 9 of culture, which was after 3 d at the A/L interface. Secretion experiments were performed by incubating the cells with secretagogues for 3 h on Day 10 of culture, which was the fourth day under A/L conditions. Secretion was calculated as a percent of the total PC recovered in the cells plus medium, as described previously (28). Protein secretion was also measured after a 3-h incubation conducted on Day 10 of culture. Cytotoxicity was evaluated in selected experiments by measurement of lactate dehydrogenase (LDH) in the medium and the cells after 3-h incubations for secretion of SP-A and SP-D. In these experiments, LDH was measured spectrophotometrically (Sigma Chemical Co.).

Phase Microscopy, Histochemistry, and Electron Microscopy

Phase microscopy was performed with an inverted-phase microscope using standard techniques. Alkaline phosphatase was detected in paraformaldehyde-fixed cells embedded in glycol methacrylate as described previously (29, 30). Electron microscopy was performed with standard techniques (31).

Immunocytochemistry

Different fixations and immunocytochemical methods were used to maximize detection of specific antigens, as reported previously, with a standard indirect immunoperoxidase method (9, 32). SP-A and SP-D were detected in cells fixed in acid alcohol (96% ethanol, 1% acetic acid, and 3% H₂O) overnight, transferred to 70% ethanol, and then embedded in paraffin. Sections were cut, dried, extracted with ethanol to remove paraffin, washed, treated with hydrogen peroxide in phosphate-buffered saline (PBS) to inhibit endogenous peroxidase, incubated with the primary rabbit IgG antibody to either rat SP-A or rat SP-D, and then detected with a biotinylated goat antirabbit IgG, streptavidin-conjugated horseradish peroxidase (HRP), and diaminobenzidine (DAB). For detection of pro-SP-C, cells were fixed in 4% paraformaldehyde overnight at 4°C. Detection was enhanced by treating sections with 6 M guanidine hydrochloride for 30 min at room temperature (33), and then with 0.1 mg/ml trypsin for 30 min at 37°C. After endogenous peroxide was inhibited with H₂O₂ in PBS, the primary antibody was applied and detected with the streptavidin-HRP system (32). For detection of SP-B, we used two polyclonal rabbit antibodies (No. 4633 [full length recombinant pre-pro-human SP-B] and No. R28031 [mature bovine SP-B]) kindly provided by Dr. Timothy Weaver, Dr. Jeffery Whitsett, and Dr. Susan Wert of the University of Cincinnati. The samples were fixed in 10% formalin, and the only modification of the standard protocol was to include 0.02% saponin in all of the buffers. For the 3F9 mouse monoclonal antibody, which detects an apical surface antigen on type II cells and Clara cells (34), the cultured cells were fixed in 4% paraformaldehyde. However, the step to block endogenous peroxidase with hydrogen peroxide in PBS was eliminated because this treatment greatly diminishes antigen detection with the 3F9 monoclonal antibody, and the antigen was detected by cobalt/nickel enhancement of the HRP-DAB reaction product as described by Wert and coworkers (35). We used the V9 antibody (Boehringer Mannheim, Indianapolis, IN) to detect the intermediate filament vimentin, and the monoclonal antibody MAB 1435 (clone ED-1) (Chemicon, Temecula, CA) (36) to detect macrophages. The cells were fixed with acid alcohol and processed as described earlier for SP-A and SP-D, except that the secondary antibody was a biotinylated horse antimouse IgG.

Statistics

Values are from more than three independent experiments. Mean phospholipid secretion levels were compared across our four different media and three different treatments, using repeated-measures analysis of variance (ANOVA), which was followed by pairwise contrasts of means within media or within treatments. Repeated measures ANOVA was also used to compare mean mRNA levels across four different media, followed by pairwise contrasts of means. Paired t tests were used to compare levels of protein secretion under basal and 12-O-tetradecanoyl-phorbol-11-acetate (TPA) plus terbutaline treatments. A value of P < 0.05 was considered statistically significant. Means and SEs are reported.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Morphology

Our goal was to have a monolayer of differentiated cuboidal type II cells with their apical surface facing the medium so that we could study secretion. The system that we developed is complex, and the timing of the addition of both KGF and EHS is important. If EHS is added before or simultaneously with KGF, the cells form aggregates in which the apical side faces the lumen of the aggregate, and the cultures appear similar to those of cells cultured directly on EHS (5). In such a system, the apical surface is pointing inward, toward the center of these spherules, precluding ready access for the study of secretion. In the culture system that we designed, the apical surface faces the medium and the morphologic appearance is shown in Figures 1-4. A monolayer was formed after overnight attachment in medium containing 10% FBS and subsequent culture for an additional 3 d in 2% FBS and 5 ng/ml KGF (Figure 1a). KGF stimulated cell proliferation and produced an apparently more tightly packed monolayer on Days 3 and 4 than that of cells without KGF (data not shown). After the addition of EHS, the type II cells appeared to flatten slightly, but after washing the cells showed little change (Figure 1b). Under A/L conditions (rocking with 0.4 ml of apical fluid), the cells cultured in DF12 or DF12/EHS remained flattened (Figures 1c and 1d), whereas those cultured with KGF (DF12/KGF or DF12/EHS/KGF) developed a variable number of clumps of tightly packed cuboidal cells whose cellular outlines appeared refractile (Figures 1e and 1f). Most of the subsequent studies were done in the DF12/EHS/KGF medium.

View larger version (216K): [in this window] [in a new window]   Figure 1.   Phase micrographs of cultures at different times and under different conditions. Type II cells were allowed to adhere to the collagen-coated filters for 1 d, and were then cultured for 3 d with KGF (a) and an additional 2 d with EHS (b). The cells were then cultured with a reduced apical fluid (400 µl) on a rocking platform with four different apical media. The media were DF12 alone (c), DF12/EHS (d), DF12/KGF (e) and DF12/EHS/ KGF (  f  ). Only in the presence of KGF did clumps or islands of refractile epithelial cells appear after 3 d under air-liquid conditions. Phase microscopy suggested little difference between DF12/KGF and DF12/EHS/KGF conditions.

View larger version (188K): [in this window] [in a new window]   Figure 2.   Electron micrographs of cells cultured for 3 or 9 d under air-liquid conditions. (a) Type II cells cultured for 3 d with DF12/EHS/KGF medium. These cells contain lamellar inclusions and have apical microvilli and an array of intracellular organelles. Cells cultured for 3 d under air-liquid conditions with DF12/KGF medium show in addition electron-lucent areas that are areas of glycogen (G) deposition (b). After 9 d under air-liquid conditions there is also G present in cells grown with DF12/EHS/KGF medium (c). In addition, at Day 9 there is tubular myelin found in the apical fluid adjacent to the type II cells. The bar on the micrographs indicates 1 µm.

View larger version (82K): [in this window] [in a new window]   Figure 3.   Immunocytochemistry and alkaline phosphatase histochemistry. Type II cells were cultured under air-liquid (A/L) conditions in DF12/EHS/KGF medium and fixed as stated in MATERIALS AND METHODS. The cells stained for SP-A (a), SP-D (b), SP-C (c), vimentin (d), reactivity to the antibody 3F9 (e), and alkaline phosphatase (  f   ). The methods for immunostaining and enzyme histochemistry for alkaline phosphatase are stated in the text. These cells were cultured under A/L conditions for 3 d for detection of SP-A (a), SP-D (b), reactivity to 3F9 antibody (e), and alkaline phosphatase (  f  ). However, the cells were cultured for 6 d under A/L conditions for detection of SP-C (c) and 9 d for detection of vimentin (  f   ). Longer times were required to obtain staining for SP-C and to obtain more fibroblasts for vimentin staining.

View larger version (45K): [in this window] [in a new window]   View larger version (42K): [in this window] [in a new window]   Figure 4.   Detection of SP-B by immunocytochemistry. Type II cells were cultured for 3 d (a) and 9 d (b) under air-liquid (A/L) conditions in DF12/EHS/KGF medium, as stated in MATERIALS AND METHODS. The filter inserts were washed and fixed overnight in 10% buffered formalin. The filters were then processed, and sections were stained immunocytochemically for SP-B with an antibody to recombinant human pre-pro-SP-B (antibody No. 4633) as described. After 3 d in A/L culture, SP-B was detected within the cuboidal cells (a). After 9 d under A/L conditions, there were more cuboidal cells and the cells remained positive for SP-B. The clear cytoplasm in the apical portion of the cells is an area which, as shown in other studies, contains glycogen (b).

Cells cultured in DF12/EHS/KGF medium displayed ultrastructural evidence of differentiated type II cell function (Figures 2 and 3). By electron microscopy, the cuboidal cells were polarized, as documented by apical microvilli, and contained lamellar inclusions (Figure 2). In addition, after 9 d under A/L conditions (total time in culture: 15 d), tubular myelin could be visualized along the apical surface. Electron microscopy also demonstrated that there were some fibroblasts beneath the monolayer. These cells were identified further by staining for vimentin, an intermediate filament (Figure 3d). Macrophages can also be a contaminant cell type in preparations of type II cells. However, staining with the monoclonal antibody, MAB 1435, which readily identifies macrophages in rat lung, showed macrophages to be very rare in these cultures (data not shown).

Differentiation was also assessed with immunocytochemistry. Only cells cultured in DF12/KGF or DF12/EHS/KGF medium stained positively for SP-A, pro-SP-C, or SP-D. Some of the cuboidal cells stained intensely (Figure 3). However, there were also some cells in which these antigens were not detected. The cuboidal cells tended to stain positively, and the more flattened cells appeared negative. SP-A was detected in many cells at 3 d under A/L culture conditions (Figure 3a), appeared to increase at Days 6 to 9, and then appeared to decrease on Days 12 and 15 (data not shown). There was some immunostaining of SP-A under the monolayer, especially on Days 6 through 12 under A/L conditions. We have previously observed this pattern of immunostaining for SP-A in type II cells cultured on contracted collagen gels. Staining for SP-D (Figure 3b) gave results similar to that for SP-A, and was prominent as early as 3 d under A/L culture conditions. However, the immunostaining was sustained for 15 d under A/L conditions, and was not detected under the monolayer. The staining for pro-SP-C was more variable, and it was unusual to find staining in these cultures as intense as in lung tissue. We seldom detected any immunostaining for SP-C on Day 3, and only modest staining for pro-SP-C on Day 6 under A/L conditions (Figure 3c). Alkaline phosphatase and the antigen recognized by the 3F9 antibody are additional differentiation markers of type II cells and were used to assess cellular polarity (29, 30, 34). Type II cells, cultured for 3 d under A/L conditions with EHS and KGF showed apical staining for alkaline phosphatase (Figure 3f). In these cultures, the staining was not absolutely restricted to the apical surface, and also occurred along the basolateral side. Although not all cells stained with the 3F9 antibody, the cells that stained positively had immunostaining only along the apical surface (34). Additional studies were performed for detection of SP-B because the mRNA of this particular surfactant protein was well maintained in these cultures. The immunostaining for SP-B after Days 3 and 9 under A/L conditions is shown in Figure 4. We chose Day 3 because at that time there is secretion of SP-A and SP-D, and Day 9 because there are more cuboidal cells but also more glycogen within the epithelial cells. As can be seen in Figure 4, the staining appears to be associated with granules within these cells, and was more localized to the basal portion of the cell than were the immunostains for SP-A and SP-D. Importantly, the SP-B protein can be detected on both Days 3 and 9 of A/L culture.

The appearance of glycogen within type II cells in the electron micrographs was unexpected (Figures 2b and 2c). Both periodic acid-Schiff (PAS) staining and removal of the positive stain with diastase (salivary amylase) confirmed the presence of glycogen. Electron microscopy and toluidine-blue-stained sections showed areas of apical cytoplasm devoid of organelles, which had an appearance similar to that of Day 21 fetal rat type II cells (37). The amount of glycogen present, as identified by PAS/diastase treatment, was diminished but not eliminated by culturing the cells at lower glucose concentrations or without insulin (data not shown). There was more glycogen in cells cultured with DF12/KGF or DF12/EHS/KGF medium than in those cultured with DF12 or DF12/EHS. Only with the DF12/KGF medium did we consistently observe glycogen with PAS staining after 3 d under A/L conditions (Day 9 of culture), although it was present in cells cultured with DF12/EHS/KGF medium at later times.

Differentiation was also assessed by measurement of mRNA levels for the surfactant proteins. The mRNA levels for the surfactant proteins were normalized to the level of 28S ribosomal RNA, and the results for different culture conditions are shown in Figure 5. Cells cultured in DF12 or DF12/EHS medium for 3 d under A/L conditions showed a low amount of mRNA for all four surfactant proteins. In contrast, cells cultured in DF12/KGF and in DF12/EHS/KGF medium under A/L conditions had more mRNA for the surfactant proteins on Day 3. The mRNA levels for SP-B and SP-D in DF12/KGF- and DF12/EHS/ KGF-cultured cells were 60 to 70% of those in freshly isolated type II cells, whereas the level of SP-A mRNA was about 30% and that of SP-C about 10% of the respective mRNA levels found in freshly isolated type II cells. The results for the different conditions were also normalized to the mRNA for GAPDH, with similar results (data not shown). However, GAPDH expression increases in primary culture, and could not be used to compare freshly isolated with cultured type II cells.

View larger version (42K): [in this window] [in a new window]   Figure 5.   Expression levels of mRNA for SP-A, SP-B, SP-C, and SP-D under different conditions. The cells were cultured for 3 d with an air-liquid interface in four different media: DF12, DF12/ EHS, DF12/KGF, and DF12/EHS/KGF. Day 0 refers to cells on the day of isolation. Total RNA was extracted in guanidinium isothiocyanate and isolated by centrifugation through a CsCl cushion. RNA was size-fractionated by electrophoresis, transferred to a Nytran membrane, and probed with radiolabeled complementary DNAs (cDNAs) for SP-A, SP-B, SP-C, SP-D, and an oligonucleotide complementary to 28S ribosomal RNA. Levels of mRNA were quantified with Molecular Dynamics ImageQuant software. All data were normalized to the level of 28S RNA, expressed as arbitrary units, and normalized to 100% for values found in freshly isolated type II cells. The mRNAs for the surfactant proteins increased in culture with DF12/KGF and DF12/ EHS/KGF, but were not statistically different from each other with these two media. The culture media and respective individual values (mean ± SE) for SP-C mRNA levels as a percentage of freshly isolated cells were: DF12 (3.4 ± 0.8%), DF12/EHS (3.5 ± 1.4%), DF12/KGF (7.4 ± 2.1%), and DF12/EHS/KGF (10.4 ± 2.0%) in three independent experiments. The SP-C mRNA level for the DF12/EHS/KGF medium was significantly different from that for the DF12 or DF12/EHS medium (P < 0.02).

Basal Secretion of SP-A, SP-D, and Lysozyme

This culture system allows us to study surfactant-protein secretion into the apical medium in long-term primary cultures of type II cells. We chose to study SP-A, which is thought to be secreted both by exocytosis, as a constituent of lamellar bodies, and by some undefined constitutive pathway (38, 39); SP-D, which is not found in lamellar bodies (40) and is presumed to be secreted independently of surfactant phospholipids; and lysozyme, which is found in lamellar bodies (41) and is thought to be secreted as a constituent of lamellar bodies. The initial study was conducted to determine whether these proteins were secreted into the media under the four different culture conditions. The results are shown in Figure 6. The concentration of these proteins in the apical fluid remained at low levels from Day 1 to Day 3 for cells cultured in DF12 or DF12/ EHS. In contrast, the concentration of SP-A, SP-D, and lysozyme in the apical fluid of cells cultured in DF12/KGF and DF12/EHS/KGF increased markedly for 3 d under A/ L conditions. This observation indicates that KGF increases the expression and secretion of SP-A, SP-D, and lysozyme in type II cells in vitro. KGF has previously been reported to increase SP-A and SP-B mRNA levels in type II cells cultured on EHS gels (18).

View larger version (16K): [in this window] [in a new window]   Figure 6.   Basal secretion of SP-A, SP-D, and lysozyme by cells cultured under different conditions. After overnight attachment, proliferation for 3 d in media containing 5 ng/ml KGF, and maintenance for 2 d, in medium containing 5% EHS, the cells were cultured in four different media at air-liquid (A/L) interfaces: DF12, DF12/EHS, DF12/KGF and DF12/EHS/KGF. Apical media in the inserts were collected on Days 1, 2, and 3 of A/L conditions and centrifuged at 800 × g for 10 min to remove cells and debris. The resultant supernatants were subjected to ELISA for SP-A and SP-D, and spectrophotometric assay for lysozyme, as described in MATERIALS AND METHODS. There was an increase in SP-A, SP-D, and lysozyme in the media under DF12/KGF and DF12/EHS/KGF conditions, but the increases under these two conditions did not differ from each other. Values represent means ± SE of four independent experiments.

Unstimulated secretion of SP-A and SP-D was measured for an additional 12 d with rocking, and the reduced apical fluid was cultured in DF12/EHS/KGF medium. The results are shown in Figure 7. The amount of SP-A and SP-D continued to increase, and then decreased slightly after 15 d at the A/L interface (Day 21 of culture). The maximal levels occurred at Days 6 through 9 under A/L conditions. For these experiments, the medium was changed daily. On the day of analysis, the amount of protein in the medium reflected the balance of that which was secreted over the prior 24 h plus any in the residual fluid that remained when the medium was changed, and any uptake that occurred.

View larger version (15K): [in this window] [in a new window]   Figure 7.   Secretion of SP-A and SP-D with time in culture. Media were collected daily and analyzed on Days 3, 6, 9, 12, and 15 under air-liquid conditions. The samples were processed as described in MATERIALS AND METHODS. There was sustained constitutive secretion of SP-A, SP-D, and lysozyme. Values represent means ± SE of seven independent experiments.

Stimulated Phospholipid and Protein Secretion

In addition to the prolonged basal secretion of SP-A and SP-D over 15 d, we also evaluated secretion in response to secretagogues over a shorter period. Type II cells in primary culture secrete phospholipid in response to secretagogues such as TPA and terbutaline (28). Since the secretory response to different classes of agonists is additive, we chose to evaluate the combination of SP-A and terbutaline in order to maximize the opportunity of observing a response in these long-term primary cultures. As shown in Table 1, type II cells cultured in DF12, DF12/EHS, or DF12/KGF showed only a 2-fold stimulation of phospholipid secretion in response to TPA and terbutaline. However, secretion by cells cultured in DF12/EHS/KGF was stimulated nearly 5-fold by the addition of TPA and terbutaline. Additionally, for type II cells cultured in DF12/ EHS/KGF medium, we found that secretion of SP-A and lysozyme, but not of SP-D, could be stimulated by the combination of TPA and terbutaline (Table 2). This is consistent with the concept that TPA and terbutaline stimulate the secretion of lamellar inclusion bodies, which contain phospholipid, SP-A, and lysozyme, but not SP-D (40). In additional experiments, we evaluated the effect of TPA and terbutaline individually on SP-A and SP-D secretion under A/L conditions on Day 4. TPA at 10 nM and 100 nM increased secretion of SP-A by 178 ± 27% and 190 ± 19%, respectively, of the control values, whereas terbutaline alone at 100 µM did not stimulate SP-A secretion (96 ± 10% of the control value) (n = 3). None of these additions increased the secretion of SP-D. In these experiments, less than 2% of total cellular LDH was released into the medium.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Type II cells cultured with KGF and EHS on rat-tail-collagen gels in membrane inserts continued to demonstrate their differentiated phenotype, as shown by the presence of apical microvilli, lamellar inclusions, staining for alkaline phosphatase, and secretion of SP-A, SP-D, and lysozyme. In addition, on Day 10 of culture (after 3 d with the reduced apical fluid and under A/L conditions), type II cells secreted SP-A, lysozyme, and phospholipid, but not SP-D, in response to the combination of TPA and terbutaline. The presence of tubular myelin in the apical fluid, seen in electron micrographs of cells cultured for 9 d under A/L conditions, suggests secretion of SP-B in addition to that of SP-A and phospholipid. Cells cultured for 3 and 9 d under A/L conditions contained SP-B as determined immunocytochemically. Hence, a long-term primary culture system has been established for studying the secretion of individual components of pulmonary surfactant. In addition, this is the first report of quantified SP-D secretion by type II cells in vitro, and demonstrates the importance of KGF as a factor that increases mRNA levels for SP-D in vitro. This system should be useful for investigating type II cell functions that require a more sustained differentiated expression and direct apical access. Although type II cells can also retain differentiated function, when they are cultured on EHS, their apical surface is at the small lumen of the spherule, and the secretory product is inaccessible without disrupting the junctional complexes (10). The apical access of the current system should allow not only study of secretion and processing of the surfactant proteins, but also of the uptake of labeled proteins from the medium for studies of endocytosis.

TPA and terbutaline are well known for their ability to stimulate secretion of the surfactant phospholipids in lamellar bodies (2, 28, 40, 41), but it was not known whether they would stimulate secretion of SP-A and SP-D, under these conditions. SP-A, but not SP-D, is found in lamellar bodies (41). We observed SP-A secretion in response to TPA alone or the combination of TPA and terbutaline. Froh and colleagues reported that in cultured fetal human type II cells, secretion of SP-A was not stimulated by TPA, by the calcium ionophore A23187, or by terbutaline, and they concluded that secretion of SP-A was independent of lamellar-body secretion. The interpretation of their observation is difficult, because when type II cells are placed on tissue-culture plastic, they rapidly lose their ability to synthesize SP-A, whereas they can continue to transport newly labeled PC to their lamellar bodies (42). Hence, the lamellar bodies formed in vitro may not contain SP-A. Rooney and coworkers were unable to demonstrate stimulated secretion of SP-A by freshly isolated type II cell or cells cultured in plastic for 18 to 20 h (45). Their baseline secretion was high (20% in 3 h), and they concluded that SP-A and lipid secretion are independently regulated. Difficulties with these conclusions are that freshly isolated cells may be damaged to some degree and may lack polarity, which may be important for secretion. Cells cultured on plastic do not maintain their mRNA levels for the surfactant proteins. However, other investigators have reported stimulation of secretion of SP-A by TPA in primary cultures of adult rat type II cells (3). The in vivo metabolic labeling studies of Ikegami and associates demonstrated that newly synthesized SP-A appears first in lung lavage fluid and later in the lamellar-body fraction. These data strongly support the concept of secretion of SP-A by a pathway distinct from that of exocytosis of lamellar bodies (46). Altogether, the data indicate that SP-A is secreted both by exocytosis of lamellar bodies and by some other, more direct route. If TPA stimulates lamellar-body secretion, it should also stimulate some SP-A and lysozyme secretion. Because of the similar secretion of phospholipid and lysozyme, the simplest explanation is that the SP-A measured in the medium during the 3 h incubation comes in part from lamellar bodies. However, our studies do not address the issue of whether all SP-A is transported intracellularly directly to lamellar bodies or travels there only by endocytosis of SP-A from the medium. The levels of SP-A and SP-D in the medium in our study reached 6 µg/ml and 0.8 µg/ml, respectively, which may seem high for cultured cells but is likely to be less than is present in alveolar fluid in vivo. In normal rats, one can recover about 50 µg of SP-A and 5 to 10 µg of SP-D per rat by lavage, which, if one assumes an alveolar-fluid volume of 100 µl (47), calculates to an alveolar-fluid concentration of 500 µg/ml for SP-A and 50 to 100 µg/ml for SP-D.

Although the culture system used in our study is an improvement over previous systems for studying secretion of the surfactant proteins, it has some limitations and produces some differences from type II cells in vivo. The mRNA level for the surfactant proteins normalized to 28S ribosomal RNA was lower than that found in freshly isolated cells. This was especially evident for the mRNA for SP-C, which fell progressively in long-term cultures. In our culture system, as well as those used in other in vitro studies (5, 10, 14, 18), there appears to be independent regulation of the surfactant proteins. The expression of both SP-B and SP-D is preserved at 60% of the value for freshly isolated type II cells as measured by mRNA levels normalized to 28S ribosomal RNA, whereas the levels of SP-A are 35% and those of SP-C, only 10% of those for freshly isolated type II cells. The differences among surfactant-protein mRNA levels may in part be a consequence of the pharmacologic effect of KGF. When adult rat type II cells are cultured on EHS, KGF increases the expression of SP-A and SP-B, but not that of SP-C (18). In transgenic mice expressing human KGF driven by the SP-C promoter, there is also less mRNA and immunostaining for SP-C than in littermate controls during development (48). Preservation of the expression of SP-C mRNA at the levels found in freshly isolated type II cells remains a challenge. In the current studies, we did not vary the concentration of corticosteroid in the media, nor did we evaluate other supplements that might enhance the expression of individual surfactant-protein mRNA levels. The expression of individual surfactant-protein mRNA levels might be increased by increasing the concentration of corticosteroid in the medium or adding other supplements such as cyclic adenosine monophosphate (cAMP). We chose to develop a system to study secretion of SP-A and SP-D, and refining the culture conditions for maximal expression of each of the surfactant proteins was beyond the scope of this study. In addition, we measured only the relative abundance of mRNAs and did not evaluate whether the changes were due to transcription or message stability. Under our culture conditions, there may not be complete segregation of plasma-membrane proteins to apical and basolateral domains. Alkaline phosphatase staining was not restricted to the apical surface, but extended to the basolateral surface. We have seen similar changes in type II cells proliferating in vivo, as identified by double staining for alkaline phosphatase with histochemistry and for bromodeoxyuridine with immunocytochemistry. There is also some heterogeneity of type II cells in these cultures, such that some are cuboidal and express markers of their differentiated state and some are flattened and no longer express these markers. Moreover, the cultures are also not pure epithelial-cell cultures; there are fibroblasts under the monolayers, as seen in the electron microscope and identified by immunostaining for vimentin. The presence of fibroblasts may contribute to maintenance of the differentiation of type II cells in these cultures. For example, fibroblasts produce KGF as well as other growth factors and cytokines (17). Type II cells cultured on top of collagen gels containing irradiated fibroblasts display improved differentiated function (5). There are many reports of the positive influence of mesenchymal cells on epithelial differentiation in the fetal lung. However, there were only very rare macrophages in our cultures, as contrasted with the large number of macrophages that appear when type II cells are cultured on tissue-culture plastic for an extended period (49). Hence, our system is a significant improvement for studying secretion of SP-A and SP-D, although other refinements will be necessary to study other functions of type II cells.

The presence of glycogen in type II cells cultured from adult animals was an unexpected finding. Glycogen was present only in long-term cultures that included KGF in the medium. Although the amount of glycogen appeared to be slightly less in cultures with low glucose levels and without insulin, it was still present. In transgenic mice that overexpress KGF under the control of the SP-C promoter, fetal epithelial cells also appear to contain large amounts of glycogen (48). We have not yet investigated the enzymatic basis for the glycogen accumulation nor measured the amount of glycogen directly. However, the PAS-positive material in our preparations was removed by treatment with salivary amylase. Additional studies will be required to determine whether KGF alone is responsible for the glycogen deposition, and the biochemical basis for its accumulation. Our routine culture system has high levels of glucose, insulin, and dexamethasone, all of which can influence carbohydrate metabolism.

Our conditions were slightly different from the A/L interface conditions reported for tracheal and bronchial epithelial cells. Airway epithelial cells are cultured routinely with no added apical fluid (19, 20), whereas we chose to add 0.4 ml of apical fluid and to incubate the cultures on a rocking platform. We added apical fluid so that we could monitor secretion on a daily basis in culture. We performed a few experiments with cells at the A/L interface in the absence of added apical medium. In these experiments we could find very little difference in the apical immunostaining for SP-A, SP-C, or SP-D or mRNA levels, compared with our usual rocked cultures with 400 µl of apical media. Specifically, SP-C mRNA levels were not significantly increased by culturing the cells without apical fluid.

In summary, we have developed a culture system for investigating the secretion of surfactant proteins that provides apical access for investigating the flux of materials into and out of type II cells. The critical component for our cultures appears to be added KGF. This should complement studies of type II cells on EHS gels, which provide ready access to the basolateral side.