Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Pulmonary surfactant is a mixture of phospholipids and proteins that is synthesized, stored, secreted, and taken up by type II alveolar epithelium (1, 2). The most abundant surfactant protein, surfactant protein A (SP-A), is synthesized by the type II cell and the Clara cell (3, 4). The 28-kD primary translation product is post-translationally sialylated and glycosylated in the Golgi apparatus. SP-A is a member of a family of proteins known as collectins (such as mannose-binding protein), which have an N-terminal collagen-like sequence and a carboxyl-terminal globular domain with lectin-like properties (5, 6). Three monomers form an  helical structure at the N-terminal collagen-like portion of the molecule, and six of these trimers assemble into the multimeric protein of ~ 650-700 kD. The lectin-like globular domain binds to mannose as well as other sugars, and this domain also binds to type II cells in a calcium-dependent manner that does not involve the mannose-binding region (7). Binding of SP-A to type II cells is both saturable and specific, suggesting that this interaction is mediated through a receptor on the cell surface (7, 8). SP-A also specifically binds to the major phospholipid in surfactant, dipalmitoyl phosphatidylcholine (DPPC) (10). SP-A appears to be involved in the regulation of surfactant turnover (11) and pulmonary host defense (14, 15). Recent experiments using transgenic mice deficient in SP-A have shown that these animals are more susceptible to infection with group B streptococcus, and there are fewer bacteria associated with alveolar macrophages in the SP-A-deficient mice (16). Our working hypothesis is that the actions of SP-A on endocytosis of phospholipid by type II cells occur through a receptor-mediated process and that a specific receptor that regulates surfactant turnover is expressed on type II cells. The purpose of the current study was to isolate, purify, and partially characterize the receptor for SP-A expressed on type II cells in rat lung.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of Type II Cells

Type II cells were isolated from the lungs of adult Sprague-Dawley rats (Harlan, Indianapolis, IN) by the method of Dobbs and colleagues (17). Cells were cultured for 20 h before harvest for preparation of membrane fractions. To examine the binding of antiserum to the cell surface, 3 × 10⁵ cells/well were fixed in glutaraldehyde on a 96-well microtiter plate that had been in primary culture for 20 h. Phospholipid uptake assays (see below) were performed on primary cultures of cells that had been incubated for 18-20 h.

Membrane Preparations

Homogenates of whole lung or type II cells were prepared in 50 mM Tris, pH 7.4, with 5 mM ethylenediamenetetraacetic acid (EDTA) and protease inhibitors. The EDTA was used both as a protease inhibitor and to prevent calcium-dependent binding of endogenous SP-A to the membranes of lung cells. Lung cell membranes were prepared from ~ 50-75 g of whole lung obtained from adult Sprague-Dawley rats; the membranes were minced and then homogenized in 50 mM Tris (pH 7.4), 2 mM EDTA with 2.1 µM leupeptin, and 200 µM phenylmethylsulfonyl fluoride (PMSF) (wt/vol 1:10) using a Polytron homogenizer. Homogenates were filtered through two layers of cheesecloth and the filtrate was centrifuged at 1,000 × g for 10 min. The supernatant was centrifuged at 20,000 × g for 25 min and the pellets were then washed in the homogenizing buffer. Type II cell membranes were prepared from 15 × 10⁹ cells that had been in primary culture for 20 h. These were subjected to Dounce homogenization in homogenizing buffer and then centrifuged at 1,000 × g to remove unbroken cells and debris. The supernatant was then centrifuged at 29,000 × g to obtain a membrane pellet. This pellet was washed three times in homogenizing buffer.

Ligand Affinity Chromatography

Membrane preparations derived from whole rat lung and primary cultures of type II cells isolated from rat lung were used in ligand affinity chromatography to isolate calcium-dependent SP-A binding proteins. Our strategy was based on the fact that binding to the surface of the type II cell is calcium-dependent and does not involve the lectin-like mannose-binding region of SP-A (9). SP-A was purified from rat pulmonary surfactant as previously described (13), and 750 µg were conjugated to Affigel-15 beads (Bio-Rad Laboratories, Richmond, CA) according to the manufacturer's instructions. Membranes from whole lung or type II cells were solubilized in 50 mM Tris, 150 mM NaCl, 2 mM CaCl₂ (pH 7.4), and 30 mM n-octyl glucopyranoside with protease inhibitors (2.1 µM leupeptin and 200 µM PMSF) using a Polytron homogenizer. Endogenous SP-A was removed because it precipitates in saline. The suspension was stirred for 15 min at 4°C and then centrifuged at 16,000 × g for 25 min. The supernatant containing solubilized membrane proteins was collected, and the pellet was homogenized in solubilizing buffer twice more. These supernatants were pooled with the first and applied to the affinity column.

All chromatography was carried out at 4°C. The SP-A affinity column was equilibrated with 5 column volumes of column buffer (50 mM Tris, 150 mM NaCl, 2 mM CaCl₂ [pH 7.4], and 30 mM n-octyl glucopyranoside with protease inhibitors), and then the solubilized membranes were applied to the column three times. The column was washed with 3 column volumes of column buffer containing 100 mM mannose to remove mannose-containing binding proteins bound to the mannose-binding region of SP-A. Next, the column was washed with 3 column volumes of column buffer containing 2 M urea to remove weakly bound proteins. Calcium-dependent SP-A binding proteins were eluted in 3 column volumes of elution buffer (50 mM Tris, 350 mM NaCl, 5 mM EDTA [pH 7.4], and 30 mM n-octyl glucopyranoside with protease inhibitors). The eluate was monitored by ultraviolet optical density (OD) and the elution peak fractions were collected in a fraction collector. The peak fractions were pooled and dialyzed against distilled water to remove the salts and detergent. Then the fractions were dialyzed against 5 mM Tris, pH 7.4. Next, these samples were concentrated using Aquacide^® (Calbiochem, La Jolla, CA) to a volume of ~ 5 ml. Finally, these samples were further concentrated using Centricon^® (Amicon, Beverly, MA) concentrators to a volume of 100 µl. Samples were stored at 80°C until analyzed.

Protein Analysis

Purified, concentrated SP-A binding proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (18) under reducing and nonreducing conditions. To study whether any of the proteins were glycoproteins, aliquots of the purified proteins were incubated with peptide-N-glycosidase F (Oxford Glycosystems, New York, NY) according to the manufacturer's instructions, and the deglycosylated proteins were analyzed by SDS-PAGE. Proteins were visualized using a standard silver staining method.

Ligand Blots

SDS-PAGE ligand blot. Five  (nanogram amounts below the sensitivity of quantifiable protein measurement) of purified SP-A binding proteins were incubated in the absence (glycosylated) or presence of peptide-N-glycosidase F (deglycosylated) using a kit according to manufacturer's instructions. The samples were then separated by SDS-PAGE (18) under reducing conditions. The proteins were electroblotted to a nitrocellulose membrane, and the membrane was processed as described for dot blots.

Dot blot. A total of 50 µg bovine serum albumin (BSA), 50 ng SP-A, 4  of purified SP-A binding proteins, and 8  of purified SP-A binding proteins were applied to a nitrocellulose membrane. The membrane was incubated in Tris-buffered saline, pH 7.4, with 1% Tween (TBST) containing 5% nonfat milk overnight at 4°C. The membrane was then washed three times in TBST with 1 mM CaCl₂ at room temperature. Next, the blot was incubated in TBST with 1 mM CaCl₂ with 5 µg/ml of SP-A for 1 h at 37°C. After three washes with TBST with 1 mM CaCl₂, the blot was incubated with polyclonal antibody to rat SP-A in TBST with 1 mM CaCl₂ containing 1% nonfat milk for 1 h at room temperature. After three washes with TBST with 1 mM CaCl₂, the blot was exposed to goat antirabbit IgG conjugated to peroxidase (1:10,000 in TBST with 1 mM CaCl₂ plus 1% nonfat milk) for 1 h at room temperature. Bound SP-A was visualized using 3,3',5,5' tetramethylbenzidine peroxidase substrate (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD) and dextran sulfate membrane enhancer according to the manufacturer's instructions. Control experiments were performed in which CaCl₂ was omitted from all the buffers, or in which blots were not incubated with SP-A.

Antibody Generation

Purified SP-A binding proteins were used to immunize mice to generate both polyclonal and monoclonal antibodies using standard techniques (19). Briefly, primary immunizations were performed using 100 ng of antigen in complete Freund's adjuvant, and booster immunizations were given using 50 ng of antigen in incomplete Freund's adjuvant. A whole-cell enzyme-linked immunosorbent assay (ELISA) was developed using primary cultures of type II cells to screen antisera and hybridoma supernates that contained antibodies that bind to the surface of type II cells. A total of 3 × 10⁵ type II cells isolated from adult rat lung was cultured in Dulbecco's minimal essential medium in each well of a 96-well plate in 10% CO₂ at 37°C overnight. Cells were washed with phosphate-buffered saline (PBS) and fixed in 2% glutaraldehyde in cacodylate buffer. Plates were stored at 4°C until used for ELISA. Wells on the plates were incubated in PBS alone (control), preimmune serum (control) in PBS, or various dilutions of antibody in PBS (1:10,000 to 1:1) for 1 h at room temperature. After several washings, the wells were exposed to peroxidase-labeled antimouse IgG for 1 h. Color was developed and the OD were determined in an ELISA plate reader.

Competitive Binding Assay

A total of 3 × 10⁵ type II cells was plated in each well of a 96-well microtiter plate, cultured overnight, and then fixed in 0.075% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4. Wells were first incubated in TBST containing 5% nonfat milk for 1 h at room temperature. After several washes with Tris-buffered saline, cells were incubated with increasing amounts of monoclonal antibody (or preimmune serum as a control) in TBST with 2.5 mM CaCl₂. Parallel wells were first incubated with 5 µg/ml of SP-A in TBST with 2.5 mM CaCl₂ to determine whether the ligand (SP-A) could block binding of the antibody to the cells. Binding of the antibody was detected using an antimouse IgG-peroxidase conjugate followed by color development. Resulting optical densities were read on an ELISA plate reader, and data for antibody binding were analyzed after correction for results of preimmune serum. Percent inhibition of binding was calculated as: [1  (OD with SP-A/OD without SP-A)] × 100. Nonspecific binding of antibody in the presence of SP-A was subtracted from total binding of the antibody in the absence of SP-A to yield specific binding of monoclonal antibody to type II cells.

Phospholipid Uptake Assay

Primary cultures of type II cells were used to study SP-A-mediated enhancement of phospholipid uptake as previously described (13). Control cultures were incubated for 2 h in Dulbecco's minimal essential medium containing liposomes. Experimental cultures were incubated in this same media containing either SP-A alone (5 µg/ml), monoclonal antibody to the SP-A receptor alone (55 µg/ml), or monoclonal antibody (55 µg/ml) followed by addition of SP-A (5 µg/ml). Uptake was calculated as nanomoles of phosphatidylcholine per milligram of protein.

Immunoblotting/Immunofluorescence/ Immunohistochemistry

Immunoblots of proteins separated by SDS-PAGE were performed using standard methods (18). Briefly, proteins were transferred to nitrocellulose membranes, and nonspecific binding was blocked by incubation in TBST with 5% nonfat milk at 4°C overnight. Next, blots were incubated with antibody to the SP-A receptor (primary antibody) in TBST with 3% nonfat milk for 1 h at room temperature. After several washes with TBST, blots were exposed to a peroxidase conjugate of antimouse IgG (secondary antibody) for 1 h at room temperature. Controls were incubated either in nonimmune IgG instead of primary antibody or in secondary antibody alone. Bands were visualized using 3,3',5,5' tetramethylbenzidine peroxidase substrate (Kirkegaard & Perry Laboratories) and dextran sulfate membrane enhancer according to the manufacturer's instructions.

Cultures of type II cells were fixed in 1% glutaraldehyde in PBS for 45-60 min. After several washes with PBS, plates were incubated in PBS containing 10% nonfat milk for 1 h at room temperature. This was followed by incubation in PBS with 5% nonfat milk plus primary antireceptor antibody (1:500) for 1 h at room temperature. Next, the plates were incubated in PBS, 5% nonfat milk, and fluorescein-labeled antimouse IgG (1:300) for 30 min at room temperature. Controls were incubated in nonimmune serum in place of primary antibody. Cells were examined under a phase contrast-fluorescent microscope.

Whole lung was examined by immunohistochemistry after fixation in Bouin's fixative. Fixed lung blocks were embedded in paraffin and sections were mounted on slides. After washing in xylene to remove paraffin, the sections were rehydrated through graded alcohols to water, and then incubated with the HistoMark^® Black kit (Kirkegaard & Perry Laboratories) blocking solution to block endogenous peroxidase activity. Next, the slides were incubated in primary monoclonal antibody to the receptor (1:1 dilution) in PBS with 1% Tween 20 plus 5% nonfat milk for 1 h at room temperature. After washing of the slides, the samples were incubated with horseradish peroxidase-labeled antimouse IgG (1:3,000 dilution) for 1 h at room temperature. The slides were further processed using the HistoMark^® Black kit according to the manufacturer's instructions. Sections were visualized under a standard light microscope and photographs were taken.

Statistics

When appropriate, data are expressed as means ± SEM. Significant differences were analyzed by two-tailed, unpaired t test with the level of significance at P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Using ligand affinity chromatography, we isolated several calcium-dependent SP-A binding proteins from lung cell membranes. Proteins with identical mobility on SDS-PAGE were isolated both from membranes derived from type II cells and from membranes from whole lung (Figure 1a). Under nonreducing conditions there were two bands (~ 86 and > 200 kD), as depicted in Figure 1a, lanes 1 and 2. There was also a band at ~ 50-55 kD in type II cell membranes (Figure 1a, lane 2), which may represent partial reduction of the > 200-kD band. Under reducing conditions, the 86-kD band was present but the higher-molecular-weight (higher-M_r) protein was not visible; and there were four proteins purified from whole-lung cell membranes with apparent M_r of 65, 55, 50, and 40 kD (Figure 1a, lane 3). The membranes from type II cells contained 86, 65, and 50 kD proteins (Figure 1a, lane 4). Therefore, we conclude that the > 200-kD protein is formed by disulfide bonds between the lower-M_r proteins, and that there are two major calcium-dependent SP-A-binding proteins under nonreducing conditions. We isolated these SP-A-binding proteins from whole rat lung because a larger quantity of these proteins could be purified for generation of antibodies. Results of 45 separate preparations yielded identical results. We did not detect any proteins in control experiments using columns containing Affigel-15 beads that had not been linked to SP-A.

View larger version (85K): [in this window] [in a new window]   View larger version (92K): [in this window] [in a new window]   Figure 1.   Isolation of SP-A binding proteins from lung cell membranes using ligand (SP-A) affinity chromatography. (a) Calcium-dependent SP-A binding proteins isolated from whole lung are identical to those isolated from primary cultures of type II alveolar epithelial cells. Lung cell membranes prepared from whole lung or type II cells were applied to the affinity column, and calcium-dependent SP-A binding proteins were purified as described in MATERIALS AND METHODS. After separation by SDS-PAGE, proteins were visualized by silver staining. The uppermost border of the figure delineates the interface between the stacking and resolving gels. Under nonreducing conditions, whole lung (lane 1) and type II cells (lane 2) contained two major proteins with Mr > 200 and 86 kD. Under reducing conditions, the 86-kD protein was present in both whole lung (lane 3) and type II cells (lane 4). In addition, there were four proteins in whole-lung cell membranes with Mr of 65, 55, 50, and 40 kD, but the > 200-kD protein was not present. Type II cell membranes contained two proteins in addition to the 86-kD protein, with apparent Mr of 65 and 50 kD under reducing conditions (lane 4). Molecular weight standards are shown in lane 5. (b) Calcium-dependent SP-A binding proteins are different from mannose-containing SP-A-binding proteins. SDS-PAGE of SP-A-binding proteins under reducing conditions stained with silver nitrate. Proteins that bound to the mannose-binding region of SP-A were eluted using column buffer containing 100 mM mannose (lane 1). Weakly bound proteins were washed off the column using column buffer containing 2 M urea (lane 2). Specific, calcium-dependent SP-A binding proteins were then eluted in elution buffer that contained 5 mM EDTA and no calcium (lane 3). Molecular weight standards are shown in lane 4.

The carbohydrate recognition domain of SP-A binds to mannose-containing glycoproteins but is not involved in the binding to type II cells. To enhance the purification of the SP-A-binding proteins isolated from lung cell membranes, mannose-containing SP-A binding proteins were removed with column buffer containing 100 mM mannose (Figure 1b, lane 1), and the calcium-dependent SP-A-binding proteins were eluted using elution buffer without calcium containing EDTA, as shown in Figure 1b, lane 3. There were four proteins visible on SDS-PAGE gels stained with silver with apparent molecular weights of 86, 65, 50, and 31 kD. Since the 31-kD band was also present in the proteins eluted with mannose, this protein is less likely to be part of the receptor. Furthermore, this protein was not seen in gels of proteins applied to the affinity column that had not been washed with buffer containing mannose prior to elution with calcium-free buffer containing EDTA (Figure 1a). This indicates that this 31-kD band does not bind to SP-A in a calcium-dependent manner. The 86-, 65-, and 50-kD proteins are more likely candidates as subunits of the SP-A receptor.

To determine whether the purified SP-A binding proteins retained the ability to bind to SP-A, we performed ligand blots. Samples of the purified SP-A binding proteins were incubated in the presence or absence of peptide-N-glycosidase F (Oxford Glycosystems) according to manufacturer's instructions, subjected to SDS-PAGE, and transferred to a nitrocellulose membrane. This membrane was probed with SP-A in the presence of CaCl₂, as described in MATERIALS AND METHODS. Both the deglycosylated (Figure 2a, lane 1) and glycosylated (Figure 2a, lane 2) 50-kD binding proteins retain the ability to bind to SP-A, as shown in the ligand blot after gel electrophoresis (Figure 2a). The deglycosylated protein had a mobility on SDS-PAGE consistent with an apparent M_r of ~ 35 kD, demonstrating that this protein contains about 30% of its molecular mass as N-linked carbohydrate. No other proteins (i.e., the 86- and 65-kD binding proteins) were detected in ligand blots, indicating specific binding of SP-A to this protein. The specificity of binding of the purified protein to SP-A in a calcium-dependent manner was confirmed in ligand dot blots (Figure 2b, lanes 3 and 4). SP-A did not bind to BSA, which was used as a negative control (Figure 2b, lane 1). SP-A was blotted in lane 2 as a positive control. SP-A did not bind to this protein in calcium-free buffer and anti-SP-A antibody did not recognize these proteins (data not shown). Nonspecific binding to the membrane was ruled out by the lack of significant background color development (i.e., high signal-to-noise ratio). We conclude that the purification procedure does not alter the ability of this glycoprotein to bind to SP-A and that SP-A binds to the protein backbone of the 50-kD glycoprotein in a calcium-dependent manner.

View larger version (31K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   Figure 2.   Purified, calcium-dependent SP-A-binding proteins retain the ability to bind to SP-A. (a) Purified SP-A-binding proteins were incubated in the presence (lane 1) or absence (lane 2) of peptide-N-glycosidase F for 18 h. The proteins were then subjected to SDS-PAGE under reducing conditions and transferred to nitrocellulose membranes. The membranes were then incubated with 5 µg/ml SP-A in TBST containing 1 mM CaCl2, as described in MATERIALS AND METHODS. Both the 35-kD deglycosylated protein (lane 1) and 50-kD glycoprotein bound (lane 2) SP-A, indicating that the binding site for SP-A resides on the protein core of the 50-kD glycoprotein. (b) A total of 50 µg of BSA (lane 1), 20 ng SP-A (lane 2), 4  of purified binding protein (lane 3), and 8  of purified binding protein (lane 4) were dot-blotted on nitrocellulose membranes, incubated with 5 µg/ml SP-A in TBST containing 1 mM CaCl2, and then developed using anti-SP-A antibody followed by color development. SP-A did not bind to BSA (lane 1, negative control) but did bind to the purified SP-A-binding proteins (lanes 3 and 4).

Polyclonal antibodies generated in mice recognize both reduced and nonreduced proteins (Figure 3a). Under nonreducing conditions the polyclonal antibody recognizes a band at > 200 kD, a major band at 86 kD, and a band at ~ 120 kD (Figure 3a, lane 1). Under reducing conditions the antibody recognizes major bands at 86 and ~ 50 kD, and the > 200-kD protein is no longer present (Figure 3a, lane 2). This indicates that the ~ 50-kD protein assembles into multimers that are bound by disulfide bonds. The purified SP-A binding proteins were incubated with peptide- N-glycosidase F (Oxford Glycosystems) and the deglycosylated proteins were separated by SDS-PAGE under reducing conditions to determine whether the antibody was binding to either the protein or carbohydrate portions of the molecules. Immunoblots showed that the polyclonal antibody recognized the 86- and 50-kD SP-A binding proteins (Figure 3b, lane 1). This antibody also recognized a deglycosylated protein of apparent 60-kD M_r (Figure 3b, lane 2), indicating that the antibody binds to an epitope on the protein backbone of the 86-kD glycoprotein. In addition, a 35-kD band is also recognized by the antibody (Figure 3b, lane 2), which indicates that the polyclonal antibody also binds to the protein portion of the 50-kD glycoprotein in lane 1.

View larger version (77K): [in this window] [in a new window]   View larger version (56K): [in this window] [in a new window]   Figure 3.   Polyclonal antibodies to the purified SP-A-binding proteins recognize both the > 200-kD and 86-kD proteins. (a) Purified SP-A-binding proteins were separated by SDS-PAGE under nonreducing (lane 1) or reducing (lane 2) conditions. Immunoblot using polyclonal antibodies reveals that the antibody recognizes the > 200-kD protein, the 86-kD protein, and a protein with apparent Mr of ~ 120 kD. This 120-kD band was consistently present under nonreducing conditions in several experiments. Under reducing conditions, the antibody recognizes the 86- and 50-kD proteins. (b) Purified SP-A-binding proteins were incubated in the absence or presence of peptide-N-glycosidase F for 18 h. Reaction products were separated by SDS-PAGE under reducing conditions and transferred to nitrocellulose membranes. Immunoblot of glycosylated (lane 1) and deglycosylated (lane 2) SP-A-binding proteins reveals that the antibody recognizes the 60-kD protein core of the 86-kD glycoprotein. In addition, the 35-kD protein core of the 50-kD glycoprotein (lane 2) is also recognized by the antibody.

As depicted in Figure 4, lane 1, immunoblots of solubilized proteins from type II cells derived from adult rat lung detected the 86- and 50-kD proteins under reducing conditions without detection of other proteins, indicating specificity of antibody binding to these proteins. None of the proteins from either lung fibroblasts or alveolar macrophages was recognized by the antibody (Figure 4, lanes 2 and 3), indicating specific binding of the antibody to type II cell proteins.

View larger version (32K): [in this window] [in a new window]   Figure 4.   Immunoblot of solubilized proteins from type II cells derived from adult rat lung (lane 1), alveolar macrophages (lane 2), and lung fibroblasts (lane 3) using polyclonal antibody to the SP-A receptor. The antibody recognizes only the 86- and 50-kD proteins in type II cells (arrows), but it does not recognize proteins in either alveolar macrophages or lung fibroblasts.

We found saturable binding of the antibody to the surface of the type II cell (Figure 5). Incubation with secondary antibody alone (not shown) or dilutions of nonimmune serum yielded no significant binding to cells (Figure 5). Immunofluorescent staining of cultured type II cells using fluorescein-labeled antimouse IgG as the secondary antibody also showed that the primary antibody recognized the protein on the surface of the type II cell (Figure 6). Controls incubated in preimmune serum or secondary antibody alone did not stain. Primary cultures of lung fibroblasts did not stain with the antibody (data not shown). Proteins from lysates of lung fibroblasts and alveolar macrophages were not recognized by the polyclonal antibody on immunoblots. We conclude that these SP-A-binding proteins are expressed on the surface of the type II cell and not on the lung fibroblast or alveolar macrophage. Since other cells in the lung may bind SP-A, we examined whether these SP-A-binding proteins were present on other lung cells using monoclonal antibodies to these proteins.

View larger version (16K): [in this window] [in a new window]   Figure 5.   Specific binding of anti-SP-A receptor antibody to type II cells is saturable. Type II cells were cultured in 96-well dishes for 20 h, fixed to the plates, and assayed for binding of antibody to purified SP-A binding proteins as described in MATERIALS AND METHODS. Binding of polyclonal antibody (open circles) was concentration-dependent. There was no significant binding of nonimmune serum at any concentration (closed diamonds). Specific binding of the antibody to the purified SP-A-binding proteins was calculated as OD of polyclonal antibody minus OD of nonimmune serum. Results are expressed as means ± SEM of eight experiments.

View larger version (75K): [in this window] [in a new window]   Figure 6.   Anti-SP-A receptor antibody binds to the surface of cultured type II cells. Cultured type II cells were fixed and incubated with antibody to the SP-A-binding proteins. Fluorescein-tagged antimouse IgG was then applied after several washes with PBS. Top panel shows cells under phase contrast with characteristic features of type II cells. Bottom panel of same field shows immunofluorescent staining of the type II cells. Original magnification: ×400. Controls incubated with preimmune serum or only secondary antibody did not stain.

We isolated seven monoclonal antibodies from hybridomas generated using standard techniques (18) that recognized the antigen on the surface of type II cells by ELISA. These antibodies were purified over a protein G column and were IgG antibodies. Immunoblots confirmed that four of these antibodies recognized the nonreduced > 200-kD protein, one recognized the 86-kD glycoprotein, and two reacted weakly with the > 200-kD protein. Two of the antibodies that strongly reacted with the > 200-kD protein (P8E3-18 and P15G5-16) also stained only type II cells in sections of rat lung (results from P15G5-16 antibody shown in Figure 7). Controls incubated in nonimmune mouse IgG or secondary antibody alone did not stain.

View larger version (80K): [in this window] [in a new window]   View larger version (93K): [in this window] [in a new window]   Figure 7.   Anti-SP-A receptor antibody specifically binds only to type II cells. Lung sections were fixed, embedded in paraffin, and stained with monoclonal antibody to SP-A receptor. A peroxidase conjugate antimouse IgG secondary antibody was used to visualize binding of the monoclonal antibody. (a) Only type II cells in the lung sections stained with the anti-SP-A antibody. Controls, incubated in either preimmune serum, nonimmune mouse IgG, or secondary antibody only, did not stain. (b) Sections of bronchioles and capillaries did not stain with the antibody. Original magnification: ×400.

To determine whether the monoclonal antibody was binding to the receptor for SP-A, we examined the ability of the ligand (SP-A) to block binding of the antibody to type II cells. We chose this strategy because we would be able to compare the binding assay results with our data using polyclonal antibody, and because we could avoid chemical modification of SP-A (such as biotinylation or iodination), which could alter the binding characteristics of SP-A. As shown in Figure 8a, monoclonal antibody P15G5-16 significantly bound to type II cells, whereas an irrelevant mouse IgG control antibody did not. After correcting for the nonspecific control antibody binding, SP-A almost completely inhibited binding of the P15G5-16 antibody to type II cells (Figure 8b). Specific binding of P15G5-16 (binding in the absence of SP-A minus binding in the presence of SP-A) was saturable. Because there was no significant change in OD with increasing concentration of mouse IgG control antibody and no significant binding to cells, the control data for all dilutions were combined to compare controls in the absence or presence of SP-A (OD: 0.007 ± 0.004 in the absence of SP-A, 0.007 ± 0.005 in the presence of SP-A, n = 4). SP-A had no effect on inhibition of binding of the isotype control mouse IgG (Figure 8c). SP-A resulted in 93 ± 4% inhibition of binding of 17 ng of antibody; and SP-A resulted in 93 ± 7% inhibition of binding of 17 µg of antibody (Figure 8c). There was a significant decrease to 61 ± 6% inhibition of antibody binding by SP-A with 85 µg of antibody (Figure 8c), indicating that the antibody and the ligand (SP-A) were competing for the same binding site on the receptor.

View larger version (18K): [in this window] [in a new window]   Figure 8.   The ligand for the receptor (SP-A) blocks binding of monoclonal antibody to type II cells. Primary cultures of adult type II cells were fixed in glutaraldehyde and subsequently incubated in buffer with increasing amounts of P15G5-16 monoclonal antibody or irrelevant mouse IgG (as a control). Replicate wells were incubated with or without 5 µg/ml SP-A and binding of antibody was determined by ELISA as outlined in MATERIALS AND METHODS. (a) Representative experiment showing concentration-dependent binding of P15G5-16 monoclonal antibody to type II cells (open squares). There was no significant binding of irrelevant control IgG antibody (closed diamonds). Data for binding of anti-receptor antibody were analyzed after correction by subtracting control data. (b) Representative experiment showing saturable binding of P15G5-16 antibody to type II cells. Most of the binding was inhibited by SP-A. Specific binding was determined by subtracting nonspecific binding of antibody in the presence of SP-A from the total binding in the absence of SP-A. Similar results were obtained in three other experiments. (c) SP-A almost completely inhibits binding of monoclonal antibody to type II cells (n = 4). Reduced inhibition at 85 µg of antibody indicates competitive interaction. There was no significant inhibition of binding of irrelevant antibody (CON) by SP-A. Percent inhibition was calculated as: [1  (OD with SP-A/OD without SP-A)] × 100. Data are expressed as means ± SEM. *P < 0.001 versus no antibody; P < 0.025 versus 0.017, 17 µg antibody.

Since SP-A enhances phospholipid uptake by type II cells in a process that appears to be receptor-mediated, we tested whether these purified glycoproteins were components of that receptor using the monoclonal antibody to the > 200-kD protein. This antibody was used in a series of experiments that studied the effects of the antibody on the SP-A-mediated enhancement of phospholipid uptake by type II cells. We found that the antibody (55 µg/ml) completely blocked the effect of SP-A on uptake of phosphatidylcholine (Figure 9). The antibody itself did not affect uptake, as shown in Figure 9.

View larger version (36K): [in this window] [in a new window]   Figure 9.   Anti-SP-A receptor antibody blocks the effect of SP-A on phospholipid uptake by type II cells. Primary cultures of type II cells were incubated with [3H]DPPC-labeled liposomes in media alone (control; open bars), media containing monoclonal antibody to the SP-A receptor (light crosshatched bars), media containing 5 µg/ml SP-A (dark crosshatched bars), or media containing monoclonal antibody and 5 µg/ml SP-A (stippled bars). As expected, SP-A enhanced uptake of phospholipid compared with controls. Monoclonal antibody to the SP-A receptor completely blocked the effect of SP-A on phospholipid uptake. The antibody by itself had no effect on phospholipid uptake by type II cells. Data are expressed as means ± SEM of four experiments. *P < 0.025 versus control; P < 0.01 versus monoclonal antibody; P < 0.005 versus antibody + SP-A.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have purified and partially characterized a novel receptor for SP-A expressed on type II cells. There are two major peptides that bind to SP-A in a calcium-dependent manner that does not involve the mannose-binding region of SP-A. The ligand affinity chromatography that we have developed separates the mannose-containing SP-A binding proteins from the calcium-dependent SP-A binding proteins by elution with buffer containing calcium and 100 mM mannose. The specific, calcium-dependent SP-A-binding proteins are then eluted using calcium-free buffer containing EDTA. Under nonreducing conditions, there are two calcium-dependent SP-A-binding proteins with apparent M_r of 86 and > 200 kD that were isolated from type II cell membranes and cell membranes prepared from whole lung. These are not proteins that simply bind to the matrix in the affinity column, because ligand affinity columns lose their function and cannot isolate these SP-A binding proteins after 1 yr, and we did not detect proteins using columns containing matrix that has not been covalently linked to SP-A.

Two lines of evidence indicate that the > 200-kD protein is composed of multiple ~ 50-kD proteins which assemble through the formation of disulfide bonds. First, comparison of silver-stained gels of samples run under reducing and nonreducing conditions reveals that the larger protein is not present under reducing conditions, whereas an ~ 50-kD protein is present (Figure 1b). Second, immunoblots of proteins separated under reducing and nonreducing conditions confirm this finding (Figure 3a).

Several lines of evidence show that the ~ 50- and 86-kD proteins are glycoproteins that contain about 30% of their molecular mass as N-linked carbohydrate. Deglycosylation with endoglycosidase shows that the ~ 50-kD protein has a relative mobility equivalent to 35 kD by silver stain (data not shown), by ligand blot (Figure 2a), and by immunoblot (Figure 3b). The 86-kD protein has a relative mobility of ~ 60 kD after digestion with endoglycosidase by silver stain (data not shown) and by immunoblot (Figure 3b).

These glycoproteins are specifically expressed on type II cells as shown by the ELISA binding isotherms, immunofluorescence experiments, and immunohistochemistry. Inhibition of binding of monoclonal antibody by SP-A, and blockade of the effects of SP-A on phospholipid uptake by type II cells using monoclonal antibodies to the > 200-kD complex, provide conclusive evidence that the 50-kD peptides are components of the receptor for SP-A that is expressed by the type II alveolar epithelial cell.

Together, our data indicate that a receptor for SP-A specifically expressed by type II cells contains a 50-kD glycoprotein that assembles by formation of disulfide bonds to form a > 200-kD protein that binds to SP-A in a calcium-dependent manner that does not involve the mannose-binding region of SP-A. The 50-kD glycoprotein retains the ability specifically to bind SP-A, and the binding site for SP-A is located on the protein core of the 50-kD glycoprotein.

SP-A binds to itself, but several lines of evidence indicate that the 50-kD glycoprotein we have isolated is not SP-A. First, the apparent M_r of reduced, glycosylated SP-A is 32-36 kD (depending on the degree of post-translational modification). Second, the apparent M_r of the primary translation product of rat SP-A is 28 kD, and the apparent M_r of this deglycosylated SP-A receptor is 35 kD. Third, none of the polyclonal or monoclonal antibodies to this receptor recognize SP-A in lung homogenates or type II cell proteins, and anti-SP-A antibody does not recognize the 50-kD glycoprotein. Therefore, the receptor we have isolated is not SP-A itself.

Recently, Chroneos and colleagues purified a 200-kD cell surface receptor from membranes derived from the U937 macrophage cell line and membranes from whole rat lung (20). They found that polyclonal antibodies to this protein inhibited binding of SP-A to alveolar macrophages and type II cells. This antibody also blocked the inhibitory effect of SP-A on phospholipid secretion by type II cells. Binding affinity to alveolar macrophages was 4-fold higher than binding affinity to type II cells. In addition, the receptor isolated from U937 cells and rat bone marrow-derived macrophages was soluble in detergent, whereas this 200-kD protein isolated from rat alveolar macrophages and type II cells was insoluble in detergent. These differences raise the possibility that there are two different isoforms or molecules with a M_r of 200 kD. Electrophoresis under reducing conditions did not alter the relative mobility of this protein (20). The high-M_r protein (> 200 kD) that we have isolated on type II cells is composed of smaller peptides based on the change in the relative mobility when subjected to electrophoresis under reducing conditions. Furthermore, the antibodies to the SP-A receptor expressed on type II cells that we have isolated do not block the inhibitory effects of SP-A on phospholipid secretion by type II cells (data not shown), but these antibodies do block the effect of SP-A on phosphatidylcholine uptake (Figure 9). Therefore, it is not unreasonable to suggest that type II cells express at least two different receptors for SP-A that mediate secretion and uptake separately.

Other investigators have indirectly studied a receptor for SP-A using anti-idiotype antibodies against SP-A. Strayer and associates found a 32-kD protein using this approach (21), but they were unable to isolate this protein and study it directly. They found that these antibodies also bound to alveolar cell membrane proteins from three different species, raising the question of cross-reactivity and thus nonspecificity of binding of the antibody. They also isolated complementary DNA (cDNA) clones from both human and porcine lung cDNA libraries using these antibodies. Using this cDNA they found nonspecific expression of the messenger RNA in various tissues (heart, kidney, and small intestine), which also raises the question of specificity of this protein. It is hard to interpret this data and compare our results. It is possible that this 32-kD protein isolated using the anti-idiotype antibody approach might be the mannose-containing SP-A-binding protein that we found, but which we believe is unlikely to be the specific receptor for SP-A expressed on type II cells that regulates surfactant uptake and clearance by the type II cell.

Stevens and coworkers used a similar approach by generating auto-anti-idiotypic antibodies in mice and found that one anti-idiotype antibody (2H5) bound two proteins present in type II cell membranes: a 170-200-kD protein under nonreducing conditions and a 55-kD protein under reducing conditions (22). Immunoprecipitates of type II cell membrane proteins using the 2H5 antibody were able to bind to SP-A in a calcium-dependent manner (22). Furthermore, they showed that binding of biotinylated SP-A to type II cells was inhibited by the 2H5 antibody (22), and that this antibody was able to block phospholipid uptake (23); but they have not yet isolated or characterized this protein. It is possible that the ~ 50-kD glycoprotein subunit of the receptor that we have isolated is the protein that is recognized by the anti-idiotypic antibody that Stevens and associates have described. In addition, we have also shown that these smaller glycoproteins form disulfide bonds to assemble into a protein with apparent M_r of > 200 kD.

In summary, the receptor that we have isolated binds to SP-A in a calcium-dependent manner that does not involve the mannose-binding region of SP-A. The purified 50-kD glycoprotein retains the ability to bind to the ligand, SP-A. Furthermore, we have determined that SP-A binds to a binding site on the protein backbone of this glycoprotein. This receptor appears to be fundamental in the regulation of surfactant metabolism in the lung. We speculate that abnormalities in the expression of this receptor may underlie pathologic conditions in which surfactant accumulates in the airspaces, such as alveolar proteinosis.