Endocytic pathway for surfactant protein A in human macrophages: binding, clathrin-mediated uptake, and trafficking through the endolysosomal pathway

Joy E. Crowther, Larry S. Schlesinger

Abstract

In the noninflamed lung, surfactant protein A (SP-A) acts as an anti-inflammatory molecule through its effects on macrophage (MΦ) function, modulating cytokine and reactive oxygen and nitrogen intermediate production. The receptors responsible for these effects of SP-A on human MΦ are not clear, although SP-A binding to several proteins has been described. In this study, we demonstrate high-affinity specific binding of SP-A to primary human MΦ. SP-A binding was inhibited by EGTA, indicating calcium dependence. However, mannan did not inhibit SP-A binding, suggesting that binding is mediated by a direct protein-protein interaction that does not involve carbohydrate recognition. Our laboratory has previously shown that SP-A is rapidly endocytosed by human MΦ into discrete vesicles. Although previous work indicates that SP-A is ultimately degraded by murine MΦ over time, the trafficking pathway of SP-A through MΦ after uptake has not been reported and is of potential biological importance. We examined trafficking of SP-A in human MΦ by electron and confocal microscopy and show for the first time that SP-A is endocytosed by primary human MΦ through clathrin-coated pits and colocalizes sequentially over time with the early endosome marker EEA1, late endosome marker lamp-1, and lysosome marker cathepsin D. We conclude that SP-A binds to receptor(s) on human MΦ, is endocytosed by a receptor-mediated, clathrin-dependent process, and trafficks through the endolysosomal pathway. These studies provide further insight into the interactions of SP-A with the MΦ cell surface and intracellular compartments that play important roles in SP-A modulation of lung MΦ biology.

  • surfactant proteins
  • alveolar macrophage
  • endocytosis

alveolar macrophages (alveolar MΦ) are the first line of cellular defense against inhaled environmental particles and infectious microorganisms that reach the lungs. These cells express a broad range of immune receptors, including Fcγ-receptors, complement receptors, Toll-like receptors (TLR), and scavenger receptors (17), as well as the MΦ mannose receptor (36), dendritic cell-specific ICAM-3-grabbing nonintegrin (34), and dectin-1 (37). These receptors contribute to the high phagocytic capacity of these cells, especially with regard to nonopsonic phagocytosis. Despite constant stimulation by inhaled particles and pathogens, alveolar MΦ display an anti-inflammatory phenotype that includes altered cytokine responses (8) and reduced oxidant production in response to stimuli (26), a phenomenon described elsewhere as “alternative activation” (11). There has been considerable interest in the discovery of those factors that induce alternative activation states in MΦ. Data are emerging that indicate components of pulmonary surfactant, especially surfactant proteins (SP) A and D, play important roles in host innate immunity, both as innate defense proteins and as modulators of cellular immune activities (6, 41).

Pulmonary surfactant is a complex mixture of proteins and lipids best known for its role in reducing surface tension at the air-liquid interface in the lung. Resident alveolar MΦ are bathed in surfactant and have been shown to ingest abundant amounts of this material (24). Therefore, the role of surfactant components in the induction and/or maintenance of an anti-inflammatory alveolar MΦ phenotype has been of great interest (41). The most abundant surfactant-associated protein (by weight) is SP-A, a member of the collectin protein family, which also includes immune molecules such as mannose-binding protein and complement component C1q. In addition to its ability to specifically bind carbohydrates and lipid with high affinity in the presence of calcium, SP-A has been shown to have multiple effects on MΦ biology in vitro, including increased pattern recognition receptor activity, increased phagocytosis, altered production of proinflammatory cytokines, and decreased production of nitric oxide in response to stimuli (33, 42). Recent work in our laboratory has shown that SP-A also inhibits production of reactive oxygen intermediates by decreasing the activity of the NADPH oxidase through a reduction in p47phox association with the phagosome (7). These effects were observed both with preincubation of the cells with SP-A and when SP-A was added simultaneously with a phagocytic or chemical stimulus, indicating the importance of early interactions of SP-A with the MΦ.

SP-A effects on MΦ are believed to be mediated by interactions of this protein with its receptor(s) on the MΦ surface (42). Several receptor candidates have been identified; however, the role of these proteins in SP-A binding to primary human MΦ remains unclear. After binding, SP-A enters discrete vesicles in the MΦ cytoplasm (10). These vesicles have not been characterized, nor has the nature of the endocytic pathway been explored. In this study, we characterized SP-A binding to primary human monocyte-derived MΦ using fluorescently labeled SP-A. We then examined the kinetics of SP-A uptake using a novel MΦ ELISA and determined the role of clathrin in SP-A endocytosis by electron microscopy using a biotin-avidin bridging technique to label MΦ-bound SP-A with gold in situ. Finally, we characterized the trafficking of internalized SP-A in MΦ using indirect immunofluorescence and confocal microscopy to colocalize SP-A with markers of the endolysosomal pathway. Our data provide evidence that SP-A binds with high affinity to human MΦ in a calcium-dependent manner, is endocytosed by a receptor-mediated process through clathrin-coated pits, and trafficks through the endolysosomal pathway in primary human MΦ.

MATERIALS AND METHODS

Proteins and chemical reagents.

Bovine serum albumin was obtained from Sigma (St. Louis, MO). Cell culture medium was obtained from GIBCO BRL/Invitrogen (Carlsbad, CA). Human serum albumin (25% solution) and heparin sodium were obtained through Ohio State University Pharmacy Services. Rabbit polyclonal anti-SP-A was the generous gift of Dr. Francis X. McCormack (Univ. of Cincinnati, Cincinnati, OH). Biotinylated human low-density lipoprotein (LDL) was purchased from Intracel (Frederick, MD). Monoclonal mouse anti-LDL was acquired from Research Diagnostics (Flanders, NJ). Polyclonal goat anti-EEA1 and normal goat and rabbit IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse monoclonal anti-lamp-1 was obtained from the Developmental Studies Hybridoma Bank at the University of Iowa (Iowa City, IA). Mouse monoclonal anti-cathepsin D and mouse IgG isotype controls were purchased from BD Biosciences (San Diego, CA). Goat anti-rabbit IgG horseradish peroxidase (HRP) and goat anti-mouse IgG HRP were purchased from Bio-Rad (Hercules, CA). Alexa 647 succinimidyl ester, biotin-xx succinimidyl ester, Alexa 488 chicken anti-rabbit IgG, Alexa 546 donkey anti-goat IgG, and Alexa 647 chicken anti-mouse IgG were purchased from Molecular Probes (Eugene, OR).

SP-A purification.

SP-A was purified as reported previously (7). Briefly, human alveolar proteinosis bronchial lavage fluid was centrifuged at 20,000 g, washed repeatedly, and eluted using 2 mM EDTA at 4°C. SP-A was purified from the supernatant by separation over a mannose-sepharose affinity chromatography column, and EDTA was removed by dialysis against 10 mM HEPES plus 25 mM NaCl. All steps were performed under sterile conditions at 4°C whenever possible. Endotoxin-free water (Baxter, Round Lake, IL) was used for all steps to reduce LPS contamination. Purified protein was examined by SDS-PAGE and Western blot using rabbit anti-SP-A antiserum, and endotoxin contamination was determined using a Limulus Amoebocyte Lysate assay kit with Escherichia coli LPS standards (Bio-Whittaker/Cambrex, East Rutherford, NJ). Endotoxin concentration was ≤0.25 pg/μg protein.

SP-A labeling.

Purified SP-A was incubated with Alexa 647 succinimidyl ester or biotin-xx succinimidyl ester for 1 h at room temperature according to the manufacturer's instructions, and free label was removed by extensive dialysis against 10 mM HEPES. Labeled proteins were examined by SDS-PAGE under reducing and denaturing conditions and stained for total protein using colloidal Coomassie or transferred to nitrocellulose for Western blot using rabbit anti-SP-A antiserum. Endotoxin concentration was ≤0.25 pg/μg protein.

Liposome aggregation.

Unilamellar vesicles were generated by probe tip sonication of lipid [dipalmitoylphosphatidylcholine-phosphatidylethanolamine-phosphatidylglycerol (9:3:2)] at 1 mg/ml in buffer (5 mM Tris·HCl + 0.1 mM EDTA + 150 mM NaCl, pH 7.4) for 30 min on a setting of 5. SP-A-mediated aggregation of liposomes was measured by monitoring the change in absorbance at 400 nm using a plate spectrophotometer as described (16). SP-A (20 μg/ml) or buffer control was added to the wells containing lipid (400 μg/ml), and the change in absorbance at 400 nm was monitored in the presence or absence of CaCl2 (2.5 mM) for 25 min.

Cell preparation.

Monocyte-derived MΦ were obtained as described previously (7). Briefly, blood was obtained from healthy adult volunteers using an approved institutional review board protocol for human subjects at the Ohio State University. Human peripheral blood mononuclear cells (PBMC) were isolated from heparinized peripheral blood using Ficoll-Paque (Amersham, Piscataway, NJ) and cultured in Teflon wells (Savillex, Minnetonka, MN) in RPMI 1640 plus 20% autologous serum for 5–6 days at 37°C at 5% CO2. MΦ were isolated from cultured PBMC by being adhered to tissue culture plates (BD Falcon, Bedford, MA) for 2 h at 37°C in 10% autologous serum and were washed and repleted with Dulbecco's PBS plus 10 mM HEPES plus 1 mg/ml human serum albumin plus 0.1% glucose (DPBS-HHG). Cell viability was determined by trypan blue exclusion and was ≥90%.

Homologous competitive binding assay.

Cells were adhered to four-well plates, washed, and rested overnight in RPMI 1640 plus 10% serum. MΦ were lifted using rubber policemen and resuspended in DPBS-HHG. MΦ in suspension at 1E5 cells/tube were incubated with 0.6 μg of fluorescently labeled SP-A plus or minus increasing amounts of unlabeled SP-A competitor at 4°C for 2 h. Cells were washed and fixed in 2% paraformaldehyde, and fluorescence indicative of SP-A binding was measured using a Beckman FacsCalibur flow cytometer. Best-fit curves with nonlinear regression as determined using SigmaPlot software (SPSS, Chicago, IL) were used to calculate binding affinity using the equation y = min + [(max − min)/(1 + 10x −logIC50)], where x is the total concentration of unlabeled ligand (nM), y is the amount of cell-associated labeled ligand [relative fluorescence units (RFU)], min is the minimum value of y, which represents nonspecific binding, and max is the maximum value of y, which represents total binding. The dissociation constant (Kd) was determined using the equation Ki = IC50/{1 + ([r]/Kd)}, where [r] is the concentration of labeled protein and Ki is the dissociation constant of the unlabeled protein (4); the unlabeled and labeled proteins being the same, this equation resolved to Kd = IC50 − [r].

Saturation binding assay.

Cells were adhered to four-well plates and were washed and rested overnight in RPMI 1640 plus 10% serum. MΦ were lifted using rubber policemen and resuspended in DPBS-HHG. MΦ in suspension at 1E5 cells/tube were incubated with increasing amounts of fluorescently labeled SP-A plus or minus 10 mM EDTA at 4°C for 2 h. Cells were washed and fixed in 2% paraformaldehyde, and fluorescence indicative of SP-A binding was measured in triplicate tubes using a Beckman FacsCalibur flow cytometer. Best-fit curves with nonlinear regression as determined using SigmaPlot software were used to calculate binding affinity using the equation y = (Bmax x)/(Kd + x), where x is the total concentration of ligand (nM), Bmax is the maximum number of binding sites, and y is the specific binding (RFU).

Competitive binding assay with inhibitors.

Cells were adhered to four-well plates and were washed and rested overnight in RPMI 1640 plus 10% serum. MΦ were lifted using rubber policemen and resuspended in DPBS-HHG. For test conditions containing no calcium, MΦ were incubated in DPBS-HHG without calcium for 1 h at 37°C, washed, and then cooled to 4°C before addition of SP-A. MΦ in suspension at 1E5 cells/tube were incubated with 1 μg of fluorescently labeled SP-A plus or minus 10 mM EDTA, 10 mM EGTA plus 7 mM MgCl2, 1% mannose (wt/vol), and 2.5 mg/ml mannan or in DPBS-HHG without calcium at 4°C for 2 h. The addition of EDTA (pH 7.6) or EGTA (pH 7.6) to buffered assay media in a volume ≤2% of the total volume did not alter the pH of the assay media. Cells were washed and fixed in 2% paraformaldehyde, and fluorescence indicative of SP-A binding was measured in triplicate tubes using a Beckman FacsCalibur flow cytometer. Background fluorescence (MΦ with no SP-A) was subtracted from all test conditions, and values were set relative to the positive control (MΦ + SP-A, no inhibitors).

Kinetics of SP-A association with human MΦ.

Cells were adhered to six-well plates and washed and repleted in DPBS-HHG. MΦ were incubated with 10 μg/ml biotin-SP-A for 0–30 min at 37°C and then washed. Cell-associated ligand was recovered by resuspending cells in hypotonic buffer [10 mM HEPES + 10 mM EDTA + Complete Protease Inhibitor Cocktail (Roche Applied Science, Indianapolis, IN)] and lysing cells by passage through a 30-gauge needle. Lysis was monitored by hemocytometer and halted when nuclear breakage became visible. Lysates were applied to avidin-coated 96-well plates (Sigma) at 4°C overnight. Plates were washed in DPBS plus 0.05% Tween 20, and captured biotin-SP-A was stained using rabbit anti-SP-A (1:5,000) for 1 h at room temperature followed by goat anti-rabbit IgG HRP (1:3,000). Plates were developed using a Horseradish Peroxidase Substrate kit (Bio-Rad) and read at 415 nm using a plate spectrophotometer (Molecular Devices, Sunnyvale, CA). Cell-associated SP-A was determined using a standard curve after correction for background.

Clathrin-dependent endocytosis of SP-A.

Cells were adhered to four-well plates, washed, and rested overnight in RPMI 1640 plus 10% serum. MΦ were lifted using rubber policemen and washed. Cells were cooled to 4°C and incubated in DPBS-HHG with 20 μg/ml biotinylated SP-A, unlabeled SP-A control, or 10 μg/ml biotinylated LDL control for 1 h at 4°C. MΦ were washed and bound biotin-SP-A or biotin-LDL were labeled in situ by incubation with avidin-gold (1:10 dilution) for 30 min at 4°C. Cells were washed, repleted with DPBS-HHG, and incubated at 37°C for 5 min. MΦ were then fixed using 2.5% glutaraldehyde in 0.1 M phosphate buffer containing 0.1 M sucrose for 30 min and processed for transmission electron microscopy. Approximately 40 MΦ cross sections from 2 separate resin blocks were examined in each experimental group. Percent colocalization of gold particles with electron-dense pits and/or vesicles was determined by counting a minimum of 100 consecutive gold particles in each experimental group.

Electron microscopy.

Fixed cells were washed in 0.1 M phosphate buffer containing 0.1 M sucrose and resuspended in 2% agarose. Cell pellets in agarose were postfixed in 1% osmium tetroxide for 1 h, washed, and then dehydrated in graded ethanols for 10 min at each concentration. Dehydrated pellets were then infiltrated using propylene oxide and Spurr resin before being embedded overnight at 60°C. The resin blocks were trimmed to expose the underlying agar blocks, and sections were cut using a Reichert Ultracut E microtome and transferred to uncoated 300-mesh copper grids. The sections on copper grids were stained for 15 min with 2% aqueous uranyl acetate and for 5 min with Reynold's lead citrate and then washed thoroughly. Sections on grids were examined using a Philips CM-12 electron microscope.

Colocalization of SP-A with markers of the endolysosomal pathway.

Cells were adhered to acid-cleaned glass coverslips in 24-well plates at ∼4E5 MΦ/ml and repleted in DPBS-HHG. MΦ were incubated with 20 μg/ml SP-A at 37°C for 0–90 min, washed, fixed in 2% paraformaldehyde, and permeabilized in 100% methanol. Coverslips were blocked in 10% heat-inactivated fetal calf serum plus 5 mg/ml bovine serum albumin overnight at 4°C and then stained with goat anti-EEA1 antibody (1:100), mouse anti-lamp-1 antibody (1:250), mouse anti-cathepsin D antibody (1:500), rabbit SP-A antibody (1:2,000), and/or normal IgG controls for 1 h at room temperature, followed by Alexa 488 chicken anti-rabbit IgG, Alexa 546 donkey anti-goat IgG, and/or Alexa 647 chicken anti-mouse IgG (1:250). Coverslips were washed and mounted using Vectashield mounting medium (Vector Laboratories, Burlingame, CA) and examined using a ×63 oil-immersion objective on a Zeiss 510 META laser scanning confocal microscope.

Confocal microscopy.

Slides were examined using a ×63 oil-immersion objective on a Zeiss 510 META laser scanning confocal microscope. Cell-associated SP-A was identified using fluorescence and differential interference contrast, and the percent positive colocalization of SP-A with the indicated markers was determined for each time point by counting an average of 113 discrete SP-A vesicles on triplicate coverslips.

Statistics.

The numbers of experiments cited are independent experiments performed on separate occasions with a minimum of two different donors. A two-tailed Student's t-test was used to analyze differences between test groups, except where a test group result was expressed as a relative value and then analyzed with respect to the control, in which case a one-sample t-test was used to determine statistical significance. Differences between groups were considered significant at P < 0.05.

RESULTS

Labeled SP-A proteins are functional.

SP-A was purified from human bronchoalveolar lavage and labeled with biotin or Alexa 647 as indicated in materials and methods. The labeled proteins were analyzed by electrophoretic separation under reducing and denaturing conditions and by Western blotting and were compared with unlabeled (mock-treated) SP-A protein control. Coomassie staining of unlabeled SP-A, biotinylated SP-A, and fluorescently labeled SP-A revealed a major protein band with an apparent molecular mass of ∼31 kDa and a minor protein band of ∼57 kDa (Fig. 1A). The identity of these bands as SP-A was confirmed by Western blot (Fig. 1B); using equal protein loading, labeled and unlabeled SP-A exhibited similar sensitivity in detection by rabbit polyclonal anti-SP-A antibody.

Fig. 1.

Analysis of labeled surfactant protein A (SP-A) proteins. SP-A was purified from human bronchoalveolar lavage and labeled with biotin or Alexa 647 as indicated in materials and methods. The labeled proteins were analyzed by electrophoretic separation under reducing and denaturing conditions with equal protein loading, and gel was stained with colloidal Coomassie for total protein (A) or transferred to nitrocellulose for Western blot (B) using rabbit anti-SP-A antiserum. Shown are molecular mass markers (MW) in kDa, unlabeled SP-A (lane 1), biotinylated SP-A (SP-AB; lane 2), and fluorescently labeled SP-A (SP-A647; lane 3). Representative gel and blot are from 2 independent experiments. C: functional activity of labeled and unlabeled SP-A was assessed by measuring their ability to aggregate surfactant-like liposomes in the presence of calcium, as indicated by an increase in liposome A400 in the presence of SP-A and calcium. Data are expressed as absorbance at 400 nm ± SD for triplicate wells and are representative of 2 independent experiments. *P < 0.0003 compared with liposomes with no SP-A.

We next assessed the functional activity of the labeled and unlabeled SP-A using a standard assay for determining SP-A activity, i.e., measurement of its ability to aggregate surfactant-like liposomes in the presence of calcium (16). As seen in Fig. 1C, unlabeled SP-A preparations significantly increased liposome A400 in the presence of calcium, indicating that SP-A aggregated the liposomes. This effect was SP-A and calcium dependent, as no significant increase in A400 was seen in the absence of protein (Fig. 1C), and addition of EDTA to the reaction reversed SP-A's effects (data not shown). Labeling of SP-A with biotin or Alexa 647 did not inhibit the ability of SP-A to aggregate lipid (Fig. 1C), indicating these proteins are functionally active.

SP-A binding to human MΦ is receptor mediated.

To examine binding of SP-A to human MΦ, we used a homologous competitive binding assay in which MΦ were incubated with a fixed amount of fluorescently labeled SP-A (0.6 μg) plus increasing amounts of unlabeled SP-A competitor for 2 h at 4°C. Cell fluorescence indicative of labeled SP-A binding to the MΦ was then determined using flow cytometry. The resulting sigmoidal curve is shown in Fig. 2, where the upper plateau represents total binding of SP-A to the cells, the lower plateau represents nonspecific binding, and the difference between the two plateaus represents the amount of specific binding in the presence of unlabeled competitor. These data indicate that binding of SP-A to human MΦ is specific.

Fig. 2.

SP-A binding to primary human macrophages is specific. Cells were adhered to 4-well tissue culture plates, washed, and rested overnight in RPMI 1640 + 10% serum. Macrophages (MΦ) were lifted using rubber policemen and resuspended in Dulbecco's PBS + 10 mM HEPES + 1 mg/ml human serum albumin + 0.1% glucose (DPBS-HHG). MΦ in suspension at 1E5 cells/tube were incubated with fluorescently labeled SP-A (0.6 μg) ± increasing amounts of unlabeled SP-A competitor at 4°C for 2 h, washed, and then fluorescence indicative of SP-A binding was measured using flow cytometry. Data are expressed as percent of total binding (binding in the absence of inhibitor) and are representative of 3 independent experiments.

SP-A binding to alveolar type II cells (5, 15) and rat MΦ (5, 19) has been shown to be dependent on the presence of divalent cations. To further examine the interactions of SP-A with human MΦ, we incubated primary human MΦ in suspension with increasing amounts of fluorescently labeled SP-A, using 10 mM EDTA to chelate divalent cations [i.e., nonspecific binding, as defined previously (5)]. EDTA inhibited 87.4 ± 1.1% of SP-A binding to human MΦ (means ± SE at 1 μg of SP-A, P < 0.0002, n = 3). As shown in Fig. 3, SP-A binding to MΦ was saturable and divalent cation dependent.

Fig. 3.

SP-A binding to human MΦ is saturable and dependent on divalent cations. MΦ in suspension at 1E5 cells/tube were incubated in triplicate tubes with increasing amounts of fluorescently labeled SP-A ± 10 mM EDTA at 4°C for 2 h and washed, and then fluorescence indicative of SP-A binding was measured using flow cytometry. Data are expressed as mean fluorescence intensity (MFI) ± SD and are representative of 3 independent experiments.

We used the two assays described above to calculate the affinity of SP-A binding to human MΦ. Our results demonstrate that SP-A binding to human MΦ is of high affinity that is specific, saturable, and dependent on divalent cations; the binding affinities obtained by nonlinear regression in both assays were comparable, with a mean Kd of 9.95 ± 3.68 nM (means ± SE, n = 6).

SP-A binding to human MΦ is calcium dependent and not inhibitable by mannan.

SP-A binding to MΦ could be mediated through specific binding of SP-A to carbohydrates on the MΦ cell surface via its carbohydrate recognition domain (CRD). The binding of the SP-A CRD to carbohydrates is dependent on calcium (6). EDTA chelates both Ca2+ and Mg2+; therefore, we next investigated whether the binding of SP-A to human MΦ was calcium dependent, using 1 μg of SP-A in the presence or absence of 10 mM EGTA plus an excess of MgCl2. As shown in Fig. 4A, EGTA inhibited 84.6 ± 0.2% of total SP-A binding to MΦ (n = 2, P < 0.001), indicating that this binding is dependent on calcium. This was supported by experiments in which incubation of MΦ with SP-A in calcium-free media led to an 82.8 ± 9.2% decrease in SP-A binding compared with controls (n = 2, P < 0.05).

Fig. 4.

SP-A binding to human MΦ is calcium dependent and not mannan inhibitable. MΦ in suspension at 1E5 cells/tube were incubated in triplicate tubes with 1 μg of fluorescently labeled SP-A with or without 10 mM EDTA, 10 mM EGTA + 7 mM MgCl2, or media without calcium (A), 1% mannose, or 2.5 mg/ml mannan (B) at 4°C for 2 h and washed, and then fluorescence indicative of SP-A binding was measured using flow cytometry. Data are expressed as percent of total binding (where “total” is the amount of SP-A bound in the presence of calcium and absence of inhibitors) ± SE for 2–3 independent experiments. *P < 0.001, **P < 0.05 compared with total binding.

The SP-A CRD binds with high affinity to mannose moieties (6). Therefore, to further examine the involvement of carbohydrate recognition by the SP-A CRD in SP-A binding to MΦ, we incubated human MΦ with 1 μg of SP-A in the presence or absence of 1% soluble mannose monosaccharide or 2.5 mg/ml mannan, a mannose polymer from Saccharomyces cerevisiae. As shown in Fig. 4B, mannose or mannan did not inhibit SP-A binding to MΦ, which suggests that binding of SP-A to MΦ does not involve carbohydrate recognition. The requirement for calcium in SP-A binding is therefore likely due to the involvement of a calcium-dependent MΦ receptor, a requirement for calcium to maintain structural conformation of a recognized SP-A epitope, or both.

Kinetics of SP-A association with human MΦ.

We next examined the kinetics of SP-A association with MΦ using a new avidin-capture technique to measure cell-associated biotin-labeled SP-A. MΦ were incubated with biotin-SP-A for 0–30 min at 37°C and then lysed. Cell-associated biotin-SP-A was captured from lysates using avidin-coated microtiter plates and then detected by specific antibody ELISA. As shown in Fig. 5, SP-A association with MΦ was very rapid and reached a steady state maximum by 10 min. Approximately 90% of cell-associated SP-A at early time points (2–10 min) was resistant to removal from MΦ with EGTA before cell lysis (data not shown), suggesting that cell-associated SP-A is rapidly internalized, consistent with receptor-mediated uptake.

Fig. 5.

SP-A rapidly associates with human MΦ at 37°C. MΦ were adhered to 6-well plates and were incubated with 10 μg/ml biotin-SP-A for 0–30 min at 37°C and then washed and lysed. Cell-associated ligand was captured by binding to avidin-coated 96-well plates. Plates were washed and stained using rabbit anti-SP-A followed by goat anti-rabbit IgG horseradish peroxidase. Plates were developed and read at 415 nm. Cell-associated SP-A was determined using a standard curve after correction for background. Data are expressed as average cell-associated SP-A (μg) ± SD over time for triplicate wells and are representative of 3 independent experiments.

Endocytosis of SP-A is clathrin mediated.

Endocytosis of receptor-bound ligands may occur through several mechanisms, one of which is clathrin-coated vesicle assembly and internalization (14). We examined endocytosis of SP-A using biotin-labeled SP-A conjugated to avidin-gold in situ. MΦ were incubated with biotin-SP-A, unlabeled SP-A negative control, or biotin-LDL positive control for 5 min at 37°C, fixed in glutaraldehyde, and then embedded, sectioned, stained, and examined using transmission electron microscopy. Gold labeling of cells incubated with unlabeled SP-A controls was low and nonspecific (Fig. 6A). As shown in Fig. 6B, in MΦ incubated with biotin-SP-A, the vast majority of gold particles was localized to electron-dense-coated pits and vesicles, similar to the localization of gold seen in cells incubated with biotin-LDL controls (1) (78.6% and 72.3% positive, respectively, Fig. 6C), indicating involvement of clathrin in the receptor-mediated uptake of SP-A by MΦ.

Fig. 6.

SP-A is localized to clathrin-coated pits and vesicles in human macrophages. MΦ in suspension at ∼1E7 cells/tube were incubated with 20 μg/ml biotinylated SP-A or unlabeled SP-A control or 10 μg/ml biotinylated LDL control at 4°C for 1 h and washed, and bound biotin-SP-A or biotin-LDL was labeled in situ by incubation with avidin-gold for 30 min at 4°C. Cells were then incubated at 37°C for 5 min, fixed, and processed for transmission electron microscopy. A: unlabeled SP-A control shows nonspecific gold staining only. B: gold indicative of biotin-SP-A is localized to clathrin-coated pits (arrow) and vesicles (arrowheads). C: gold indicative of biotin-LDL is localized to clathrin-coated pits (arrow) and vesicles (arrowhead). Images are representative of 2 independent experiments. Each bar = 120 nm.

SP-A colocalizes with markers of the endolysosomal pathway.

The biological effects of SP-A on MΦ function could result from signals generated through cell surface receptor(s) bound by extracellular SP-A and/or signal cascades initiated by interactions of intracellular SP-A with proteins encountered as SP-A traffics through the cell after endocytosis. We have previously shown that after binding, SP-A is endocytosed by human MΦ into discrete vesicles (10). We therefore wanted to examine the fate of SP-A in MΦ by determining whether SP-A enters the MΦ endolysosomal pathway, since previous studies have shown that SP-A is degraded by alveolar MΦ (43). To examine the localization of SP-A in the MΦ, we treated MΦ with 20 μg/ml SP-A for 5–90 min at 37°C, washed, fixed, and permeabilized the cells, and then stained for SP-A and the early endosome marker EEA1, the endosome/lysosome marker lamp-1, or the lysosome marker cathepsin D. Stained cells were examined by confocal microscopy. SP-A showed maximum colocalization with EEA1 (Fig. 7A) at 5–15 min, followed by a steady decline to baseline over the remainder of the assay. SP-A colocalization with lamp-1 (Fig. 7B) and cathepsin D (Fig. 7C) was initially low (baseline) and then increased over time. Quantification of SP-A colocalization with each of these markers was determined for each time point by counting and is shown in Fig. 7D. Together, these data provide evidence that SP-A traffics through the endolysosomal pathway in human MΦ and that the SP-A-containing vesicles mature into lysosomes.

Fig. 7.

SP-A colocalizes with markers of the endolysosomal pathway. Cells were adhered to glass coverslips in 24-well plates at 4E6 cells/ml (∼2E5 MΦ/well), washed, and repleted in DPBS-HHG. Cells were incubated with SP-A (20 μg/ml) for 5–90 min at 37°C, washed, and then fixed in 2% paraformaldehyde and permeabilized in methanol. Coverslips were blocked overnight at 4°C and then stained with goat anti-EEA1 antibody, mouse anti-lamp-1 antibody, mouse anti-cathepsin D antibody, rabbit anti-SP-A antibody, and/or normal IgG controls, followed by Alexa 488 chicken anti-rabbit IgG and Alexa 647 chicken anti-mouse IgG or Alexa 546 donkey anti-goat IgG. Cells were visualized using confocal microscopy. Colocalization of SP-A (red) with EEA1 (A, green), lamp-1 (B, green), or cathepsin D (C, green) is indicated by yellow staining. Images are representative of 3–4 independent experiments. Each bar = 10 μm. D: cell-associated SP-A was identified using confocal fluorescence and differential interference contrast microscopy, and the percent positive colocalization of SP-A with EEA1, lamp-1, and cathepsin D was determined for each time point by counting an average of 113 discrete SP-A vesicles each on triplicate coverslips. Results are expressed as percent positive colocalization ± SE for 3–4 independent experiments.

DISCUSSION

Alveolar MΦ are believed to originate from two sources: local replication and monocyte/MΦ migration. Limited data in vivo suggest that alveolar MΦ replication may occur in the absence of inflammatory stimuli, whereas significant proliferation of alveolar MΦ in vivo has been observed in several inflammatory models of lung disease (8). Under normal, noninflammatory conditions, however, the primary source of MΦ in the alveoli appears to be transmigration of blood monocytes and/or interstitial MΦ into the alveolar space, where they encounter lung-specific molecules capable of maturing them into prototypical alternatively activated MΦ (11, 17). In this report, we have used the human monocyte-derived MΦ, an established in vitro model system, to model the initial interactions of monocytes/MΦ, which are migrating into the alveolar space, with SP-A. SP-A is the most abundant surfactant-associated protein by weight (33). Human SP-A contains a hydroxyproline-rich collagen-like domain, a CRD, and one N-linked oligosaccharide attachment site in the CRD (20). SP-A monomers assemble into an octadecamer, which provides high valency binding sites for many molecules, including carbohydrates, surfactant lipids, and cell surface receptors. The interactions of SP-A with its receptor(s) on MΦ have been reported to affect multiple phenotypic and functional aspects of MΦ biology (33, 42). Here we characterized the nature of the receptor-mediated binding and uptake of SP-A by human MΦ and identify for the first time the trafficking route of SP-A in human MΦ after uptake. In performing these studies, we used newly produced labeled SP-A proteins, a novel MΦ ELISA for SP-A uptake, and an electron microscopy bridging technique for SP-A localization in MΦ.

The receptors responsible for the effects of SP-A on MΦ are still under investigation, although several SP-A binding proteins have been identified. Malhotra et al. (18) showed that SP-A bound purified C1q receptor and could inhibit binding of C1q to U-937 cells, and Chroneos et al. (5) identified an SP-A binding protein called SP-R210 on U-937 cells and rat bone marrow-derived MΦ, type II epithelial cells, and alveolar MΦ. Studies using transfected cell lines have shown SP-A binds to signal inhibitory regulatory protein (SIRP)-α (9) and the calreticulin/CD91 complex (9). SP-A also binds to soluble recombinant CD14 (32), glycoprotein-340 (39), and TLR2 (23), although its ability to bind the membrane forms of these proteins has not been shown.

The contribution of these or other proteins to SP-A binding to primary human MΦ is not clear. Our approach in this study was to characterize the nature of SP-A binding to these cells. We found that SP-A binding is specific and saturable, indicative of receptor-mediated binding. Data from competition and saturation experiments provide evidence for a single high-affinity receptor class, although we cannot completely rule out that SP-A binds to more than one receptor type with similar high affinity.

We found that SP-A binding to primary human MΦ is dependent on Ca2+. This is consistent with Ca2+-dependent interactions of the SP-A CRD with carbohydrates on the MΦ surface (6). However, the addition of mannose monosaccharide or mannan, a mannose polymer, had no effect on SP-A binding. Thus our data suggest that the dependence of SP-A binding on Ca2+ is due to the participation of a Ca2+-dependent receptor and/or a requirement for Ca2+ to maintain structural conformation of a recognized epitope on SP-A. The latter possibility is supported by work showing that Ca2+ alters the conformation of the collagenase-resistant fragment of SP-A (28, 30, 35). Chroneos et al. (5) observed mannan-resistant binding of SP-A to rat alveolar MΦ and U-937 cells. In contrast, SP-A binding to rat and bovine MΦ is mannan or mannose inhibitable (19, 29). Therefore, the repertoire of SP-A receptors expressed by human, rat, and bovine MΦ appears to differ, which may account for differences in the biological effects of SP-A reported in these cells.

In addition to the characteristics of SP-A binding noted above, recent work in our laboratory indicates that the SP-A receptor(s) on human MΦ undergoes homologous desensitization after initial ligation by SP-A (3). Together with published literature, our data would be consistent with involvement of SPR-210, TLR2, and/or SIRP-α in the binding and anti-inflammatory effects of SP-A observed in our MΦ system (7). Given the plurality of proteins that bind SP-A, it is likely that no single receptor is responsible for the effects of SP-A on MΦ. Instead, the effects of SP-A on a MΦ are dependent on the outcome of pro- and anti-inflammatory signals generated through receptors at the cell surface, the ratios (and subsequent results) of which may be controlled by multiple factors including SP-A presentation (i.e., “heads” vs. “tails”), cell activation state, local concentration of cofactors such as Ca2+, and the presence of other immunomodulatory molecules. The MΦ SP-A receptor repertoire may change during maturation and/or differentiation of these cells (by SP-A and/or other local tissue factors). Such changes would be expected to alter SP-A signaling, and therefore its effects on MΦ function, in migrating vs. resident MΦ. Identification of these receptors and their signal pathways in human MΦ is a focus of ongoing investigation in our laboratory.

One issue that has been raised with regard to SP-A binding and signaling of MΦ in the lung is its ubiquitous nature. Alveolar MΦ are bathed in and ingest abundant amounts of surfactant material (24). SP-A concentrations in rat lung hypophase have been estimated based on lavage to be 300 μg/ml to 1.8 mg/ml (41). It would be expected that a cell residing in such an environment might quickly become desensitized to SP-A signaling. Our collective data support the notion that SP-A serves as a local tissue factor contributing to the induction and/or maintenance of the alveolar MΦ phenotype: resident alveolar MΦ are in a continuous state of desensitization to SP-A, whereas migrating monocytes or interstitial MΦ are sensitive to the effects of SP-A upon first encounter.

In the alveoli, SP-A is produced by type II epithelial cells and secreted as free protein and then endocytosed by these same cells and resecreted in association with surfactant lipids (27). Therefore, it is likely that there are two pools of SP-A in the alveoli: lipid-associated SP-A and free protein. The majority of SP-A is lipid associated and concentrated in tubular myelin (25). SP-A from this pool could interact with MΦ through receptor-independent pathways due to its sequestration by lipid and have little or no independent biological effect on the MΦ. This hypothesis is supported by work by Golioto and Wright (12), who showed that Survanta or surfactant-like liposomes inhibited the SP-A-mediated increase in phagocytosis of Streptococcus pneumoniae and group B Streptococcus, as well as work by Huang et al. (13), who found that Infasurf inhibited SP-A effects on TNF-α and IL-1β in THP-1 cells. The remaining small pool of SP-A, i.e., free protein, would demonstrate specific binding and signaling in MΦ and be responsible for the effects of SP-A on the biology of these cells. Studies on SP-A binding and MΦ effects to date have revealed many consequences of the binding of purified SP-A to MΦ in the absence of lipid. In the lung, changes in overall SP-A concentration as well as shifts in the ratio of free vs. lipid-bound SP-A may alter the MΦ response to SP-A due to increased or decreased availability of free SP-A and the resulting changes in sensitivity to signaling by this protein.

We found rapid association and internalization of SP-A with MΦ at 37°C, consistent with receptor-mediated endocytosis of SP-A by human MΦ. The endocytic pathway for SP-A in MΦ after receptor-mediated binding has received little attention but is likely to contribute to the biological activities of this protein. Receptor-mediated endocytosis of proteins is classically defined by the involvement of clathrin in this process. Data are emerging regarding signal complexes associated with clathrin-dependent processes that mediate the unique biological effects of protein-receptor interactions in this context. Ryan et al. (31) have previously shown that SP-A is endocytosed by rat type II epithelial cells through clathrin-coated pits. We found that SP-A colocalized with clathrin-coated pits and vesicles in primary human MΦ, indicating that SP-A is endocytosed via a clathrin-dependent process in these cells. These data are supported by a report indicating that chlorpromazine, which has been shown to inhibit clathrin-dependent LDL uptake in fibroblasts (40), inhibits SP-A degradation in rabbit alveolar MΦ (2).

Receptor-bound proteins are commonly internalized and degraded in lysosomes. This method allows for specific protein clearance as well as signal downregulation. However, in addition to interactions with its receptors on the MΦ surface, SP-A may interact with signaling proteins after internalization. Recent data suggest that the endosome serves as a unique compartment with distinct signal transduction pathways compared with the plasma membrane and that protein trafficking may therefore regulate both the magnitude and specificity of the cell response (22, 38). Prior work in our laboratory by Gaynor et al. (10) has shown that internalized SP-A is localized to discrete vesicles within MΦ. In this study, we have shown for the first time that SP-A colocalizes over time with the early endosomal marker EEA1, the late endosomal and lysosomal marker lamp-1, and the lysosomal marker cathepsin D, indicating that SP-A vesicles mature through the endolysosomal pathway into lysosomes. The steady-state kinetics of SP-A association seen after 10 min (Fig. 5) therefore likely reflect a balance between binding of free SP-A to the MΦ and degradation of internalized SP-A in the lysosomal compartment. This is supported by work by Wright and Youmans (43), who have shown that SP-A is degraded over time by alveolar MΦ. However, in addition to acting as a means of controlling signals generated by SP-A, the trafficking of SP-A through this pathway may serve a second purpose in the MΦ. Recent work has revealed that SP-A has direct antimicrobial effects through binding to microbial membranes and increasing membrane permeability (21, 44). This ability of SP-A may be uniquely utilized by the MΦ as an intracellular weapon in its antimicrobial arsenal through targeted fusion of SP-A-containing vesicles with pathogen-containing vesicles in the phagolysosomal pathway. This possibility is currently being explored in our laboratory.

In summary, we have demonstrated that SP-A binds to primary human MΦ with high affinity in a receptor-mediated manner. This binding is calcium dependent and not inhibitable by mannan. Our data indicate that after binding to its receptor(s), SP-A is endocytosed by human MΦ via a clathrin-dependent pathway and trafficks through the endolysosomal pathway in these cells, where SP-A-containing vesicles mature into lysosomes. Further characterization of the receptor(s) for SP-A in primary human MΦ is necessary to elucidate the mechanism(s) underlying its biological effects on these cells, especially with regard to possible regulation of MΦ function in the lung through control of SP-A states (i.e., free vs. lipid bound). The current study provides further evidence that pulmonary surfactant constituent SP-A contributes to the alternate activation phenotype of human alveolar MΦ through specific interactions of this protein with its receptor(s) on the MΦ surface.

GRANTS

This work was supported by National Institute of Allergy and Infectious Diseases Grant AI-059639.

Acknowledgments

The authors thank K. Wolken at Ohio State University Campus Microscopy and Imaging Facility for technical assistance in the electron microscopy work.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract