Surfactant protein A down-regulates epidermal growth factor receptor by mechanisms different from those of surfactant protein D

We recently reported that the lectin surfactant protein D (SP-D) suppresses epidermal growth factor receptor (EGFR) signaling by interfering with ligand binding to EGFR through an interaction between the carbohydrate-recognition domain (CRD) of SP-D and N-glycans of EGFR. Here, we report that surfactant protein A (SP-A) also suppresses EGF signaling in A549 human lung adenocarcinoma cells and in CHOK1 cells stably expressing human EGFR and that SP-A inhibits the proliferation and motility of the A549 cells. Results with 125I-EGF indicated that SP-A interferes with EGF binding to EGFR, and a ligand blot analysis suggested that SP-A binds EGFR in A549 cells. We also found that SP-A directly binds the recombinant extracellular domain of EGFR (soluble EGFR or sEGFR), and this binding, unlike that of SP-D, was not blocked by EDTA, excess mannose, or peptide:N-glycosidase F treatment. We prepared a collagenase-resistant fragment (CRF) of SP-A, consisting of CRD plus the neck domain of SP-A, and observed that CRF directly binds sEGFR but does not suppress EGF-induced phosphorylation of EGFR in or proliferation of A549 cells. These results indicated that SP-A binds EGFR and down-regulates EGF signaling by inhibiting ligand binding to EGFR as well as SP-D. However, unlike for SP-D, SP-A lectin activity and EGFR N-glycans were not involved in the interaction between SP-A and EGFR. Furthermore, our results suggested that oligomerization of SP-A is necessary to suppress the effects of SP-A on EGF signaling.

Pulmonary surfactant consists of lipids and proteins and covers alveolar surfaces (1)(2)(3). Four specific proteins have been designated as surfactant protein A (SP-A), 2 SP-B, SP-C, and SP-D. They are produced by alveolar type II cells and Clara cells. SP-B and SP-C are hydrophobic proteins, which are essential for reducing surface tension. SP-A and SP-D are hydrophilic proteins and belong to the collectin subgroup of the C-type lectin superfamily. They are mosaic proteins, consisting of collagen-like domains, neck domains, and carbohydrate-recognition domains (CRD) (ϭ lectin domain), known to preferentially bind to glucose and mannose (4). The monomeric proteins of SP-A and SP-D, which are 28 -36 and 43 kDa, respectively, are assembled to form a trimeric helix at the collagen-like domain. These trimers are then oligomerized, resulting in the formation of a bouquet-like octadecamer for SP-A and a cruciform dodecamer for SP-D. SP-A and SP-D have been shown to be implicated in host defense and regulation of inflammatory responses in the lung (4,5).
Epidermal growth factor receptor (EGFR) is a member of the ErbB family. It is a type I transmembrane glycoprotein of 170 kDa, comprising a ligand-binding extracellular domain, a transmembrane domain, an intracellular tyrosine kinase domain, and a C-terminal regulatory region. By binding to a ligand, conformational rearrangement occurs, which gives rise to a "extended form," in which the dimerization arm projecting from domain II mediates homo-and heterodimers, followed by the activation of downstream signaling such as the Ras/ERK or PI3K/Akt pathways (6 -9). The signaling is involved in a wide variety of cellular events such as proliferation, differentiation, migration, and adhesion. Aberrant expression or dysregulation of EGFR has been implicated in cell transformation and cancer (10). EGFR-tyrosine kinase inhibitors show anti-tumor activity for patients with advanced non-small cell lung cancer (9).
We recently reported that SP-D suppresses the progression of lung cancer cells by down-regulation of epidermal growth factor (EGF) signaling (11). We found that SP-D binds to oligomannose type N-glycans of EGFR via lectin activity and interferes with the ligand-binding of EGFR. Moreover, previous reports have suggested that SP-A expression levels are inversely correlated with lung cancer progression. For example, genetic mutations in SP-A are associated with lung cancer progression (12,13). A high MUC1/SP-A expression ratio is related to a poor outcome in patients with lung adenocarcinoma (14). Another recent study has demonstrated that SP-A gene transfection suppresses the progression of tumors in subcutaneous xenograft mouse models by regulating the polarization of tumor-associated macrophages (15). However, the molecular mechanisms associated with the property of SP-A have not been fully elucidated.
The aims of this study were to assess whether SP-A modulates EGF signaling as well as SP-D. We found that SP-A also blocks the ligand binding of EGFR and down-regulated EGF signaling in lung cancer cells. We further demonstrated that SP-A directly binds to the extracellular domain of EGFR via the CRD and/or the neck domain, but through a means other than lectin activity, and that oligomerization of SP-A might be involved in the signaling-suppressing effects. The results indicated that SP-A exerts tumor-suppressive effects by mechanisms differing from those of SP-D.

SP-A suppresses EGF signaling in human lung cancer cells and CHOK1 cells stably expressing human EGFR
First, we confirmed that EGFR-tyrosine kinase inhibitor gefitinib almost completely blocked the EGF-induced phosphorylation of EGFR, ERK, and Akt in A549 human lung adenocarcinoma cells (Fig. 1A). The results indicated that the phosphorylation of ERK and Akt was predominantly mediated through EGFR and not a parallel pathway. We then assessed the effects of SP-A on EGF signaling in A549 cells. As shown in Fig. 1B, SP-A suppressed the EGF-induced phosphorylation of EGFR, ERK, and Akt in a dose-dependent manner. The same results were obtained with H441 human lung adenocarcinoma cells (data not shown). Next, we established CHOK1 cells stably expressing human EGFR, which express undetectable levels of endogenous EGFR. SP-A also suppressed the EGF-induced phosphorylation of EGFR, ERK, and Akt in a dose-dependent manner in CHOK1 cells stably expressing EGFR (Fig. 1C).

SP-A suppresses the proliferation, migration, and invasion of A549 cells
Next, we examined the effects of SP-A on the proliferation of lung cancer cells. A549 cells were incubated with 10 g/ml SP-A, and the cell proliferation was assayed after 24, 48, and 72 h. As shown in Fig. 2A, left panel, SP-A suppressed the proliferation of A549 cells. Dose dependence was also confirmed ( Fig. 2A, right panel). We also examined the effects of combination treatment with SP-A and gefitinib on A549 cells. SP-A augmented the viability-suppressing effects of gefitinib at low concentrations. However, at high concentrations of gefitinib, SP-A effects were not observed (Fig. 2B). These results suggested that the viability-suppressing effects of SP-A are all or partly mediated through EGFR.
We then evaluated the effects of SP-A on the migration and invasion of A549 cells. When SP-A was added, the number of EGF-induced migration and invasion cells was significantly decreased (Fig. 2C). We also assessed the effects of combination treatments with SP-A and gefitinib in migration and invasion assays. When gefitinib was at the concentration strongly suppressing the motility of A549 cells, any potential additive effects of SP-A could not be evaluated (Fig. 2D). Furthermore, we conducted wound-healing assay, indicating that SP-A decreased the wound-healing areas in A549 cells (Fig. 2E). These data suggested that SP-A inhibited the cell growth and motility of A549 cells.

SP-A reduces the binding of EGF to EGFR in A549 cells
To determine the mechanisms by which SP-A suppressed EGF signaling in A549 cells, a binding analysis of 125 I-EGF to EGFR in the presence and absence of SP-A was performed. SP-A significantly reduced the saturation level of bound EGF in A549 cells (Fig. 3A). When suppression patterns were examined, SP-A suppressed EGF binding in a dose-dependent manner (Fig. 3B). Additionally, we examined whether SP-A altered cell-surface expression levels of EGFR. Cell-surface proteins of A549 cells, which were cultured in the absence or presence of SP-A, were biotinylated. Probing of immunoprecipitated EGFR from cellular lysates with streptavidin revealed that the cellsurface expression levels of EGFR were not altered by SP-A (Fig.  4). These data suggested that SP-A suppressed EGF signaling in part by blocking the binding of EGF to EGFR.

SP-A binds to the extracellular domain of EGFR in a different manner from SP-D
In an earlier study (11), we presented evidence that SP-D directly binds to EGFR through the interaction between the CRD of SP-D and the oligomannose type of N-glycans of EGFR. We examined whether SP-A directly bound to EGFR in A549 cells by using a ligand blot. The whole-cell lysate of A549 cells was immunoprecipitated with anti-EGFR monoclonal antibody Ab-11 and subjected to SDS-PAGE, and ligand blot analysis was subsequently conducted. As shown in Fig. 5A, the anti-SP-A polyclonal antibody and monoclonal antibodies PE-10 and PC-6 detected the bands, suggesting that SP-A bound to the EGFR of A549 cells on the membrane. The same results were obtained with H441 cells and CHOK1 cells stably expressing EGFR (Fig. 5, B and C).
To examine the particular mechanisms by which SP-A directly interacts with EGFR, we prepared the recombinant extracellular domain of EGFR (soluble EGFR ϭ sEGFR). Recombinant sEGFR with or without PNGase F treatment was subjected to SDS-PAGE. Immunoblotting by using anti-EGFR monoclonal antibody Ab-5, the epitope of which is the extracellular domain of EGFR and anti-His tag polyclonal antibody, was performed. This was followed by lectin blot analysis. As shown in Fig. 6A, it was found that the molecular mass decreased, and the reactivity against concanavalin A (ConA) or datura stramonium agglutinin (DSA) was diminished by the PNGase F treatment, suggesting that both an oligomannose type and a complex type of N-glycans were successfully removed from sEGFR. We also examined whether SP-A bound to sEGFR coated onto microtiter wells. SP-A exhibited a concentration-dependent binding to coated sEGFR in the presence of Ca 2ϩ (Fig. 6B). Inclusion of 2 mM EDTA instead of Ca 2ϩ or excess mannose did not inhibit the binding of SP-A to sEGFR (Fig. 6, C and D). Cleavage of N-glycans by PNGase F did not affect the SP-A binding to EGFR, indicating that unlike SP-D, SP-A binds to sEGFR but not via N-glycans (Fig. 6E). Additionally, a binding buffer containing high concentrations of NaCl suppressed the binding of SP-A to sEGFR, suggesting that SP-A electrostatically interacts with sEGFR (Fig. 6F).

SP-A down-regulates EGF signaling
In the previous study, we indicated that SP-D binds to EGFR with the dissociation constant K D ϭ 3.2 ϫ 10 Ϫ8 M (11). We examined the binding parameters of SP-A with sEGFR using surface plasmon resonance analysis. The passage of SP-A at various concentrations over immobilized sEGFR on sensor chip C1 yielded an association rate constant of k a ϭ 5.5 ϫ 10 5 M Ϫ1 s Ϫ1 and dissociation rate constant of k d ϭ 1.1 ϫ 10 Ϫ1 S Ϫ1 , for a consequent K D (k d /k a ) ϭ 2.4 ϫ 10 Ϫ7 M (Fig. 5G). To examine whether SP-A binds to N-glycans of EGFR via lectin activity in the same way as SP-D, surface plasmon resonance analysis using a buffer with EDTA, buffer with mannose, or PNGase  (Fig. 6I) interfered with the binding of SP-A to sEGFR. When sEGFR treated with PNGase F was used, the interaction was not blocked either (K D (k d /k a ) ϭ 2.5 ϫ 10 Ϫ7 M) (Fig. 6J). In addition, the binding of SP-A to sEGFR was reduced in a buffer containing high concentrations of NaCl (Fig. 6K). These results suggested that SP-A electrostatically interacts with the extracellular domain of EGFR but not via lectin activity.

Figure 1. SP-A suppresses EGF signaling in A549 human lung cancer cells and CHOK1 cells stably expressing human EGFR.
A, A549 human lung adenocarcinoma cells were serum-starved overnight and incubated with 1 M gefitinib for 2 h at 37°C. After incubation, the cells were stimulated with 10 ng/ml EGF for 10 min at 37°C. The cell lysate was prepared, and 15 g of protein/lane were subjected to Western blotting (WB) using indicated antibodies. The data are representative of three independent experiments. B, A549 cells were serum-starved overnight and incubated with various concentrations of SP-A for 2 h at 37°C. After incubation, the cells were washed in a medium without serum and stimulated with 10 ng/ml EGF for 10 min at 37°C. The cell lysate was prepared, and 15 g of protein/lane were subjected to Western blotting using the indicated antibodies. The lower panels display the densitometric evaluation, and data are presented as mean Ϯ S.D. (error bars) from three independent experiments. Data were expressed as percentage of phosphorylation relative to that of control cells treated with EGF and without SP-A. C, same experiment as B was performed using CHOK1 cells stably expressing human EGFR. Lower panels display densitometric analyses, and data are presented as mean Ϯ S.D. (error bars) from three independent experiments. Student's t test or Welch's t test was used for statistical comparisons. *, p Ͻ 0.05; **, p Ͻ 0.01 (compared with control).

SP-A down-regulates EGF signaling Oligomerization of SP-A is necessary for suppressing effects on EGF signaling
To identify the binding mechanisms of SP-A to EGFR, we prepared a collagenase-resistant fragment (CRF), in which N-terminal collagenase domains of SP-A were removed by collagenase treatment, consisting of the neck domain and CRD of SP-A, and CRF of SP-A was purified by using gel filtration. The N-terminal sequence of the purified CRF of SP-A determined by the protein sequencing was Gly-Pro-Pro and Gly-Leu-Pro-Ala, indicating that the N-terminal region and the collagenous domain of SP-A were removed, and the CRF starts at amino acid residue Gly-75 or Gly-78 of SP-A (16,17). In addition, we previously performed gel-filtration chromatography and confirmed that SP-A was eluted with an apparent molecular mass greater than 1.5 Mda, and CRF was recovered at an apparent molecular mass of 60 kDa (16,18). Electrophoretic analysis revealed that CRF migrated as a monomer of ϳ20 kDa in the presence of SDS (Fig. 7A) (16). These results suggested that CRF assembled as trimers under a non-denaturing condition.

SP-A down-regulates EGF signaling
We examined whether CRF bound to sEGFR by using surface plasmon resonance analysis. The passage of CRF at various concentrations over immobilized sEGFR on sensor chip C1 yielded k a ϭ 4.7 ϫ 10 3 M Ϫ1 s Ϫ1 and k d ϭ 8.7 ϫ 10 Ϫ4 s Ϫ1 , for a consequent K D (k d /k a ) ϭ 1.8 ϫ 10 Ϫ7 M (Fig. 7B). These results suggested that the binding sites of SP-A to EGFR are in CRD and/or the neck domain of SP-A.
Next, we assessed whether CRF of SP-A alters the EGF-induced phosphorylation of EGFR and proliferation in A549 cells. As shown in Fig. 7, C and D, CRF of SP-A suppressed neither the phosphorylation of EGFR nor cell proliferation. These data suggested that the oligomerization of SP-A is necessary for the suppressing effects of SP-A on EGF signaling.

Discussion
In this study, we examined the effects of SP-A on EGFR in lung cancer cells. SP-A suppressed EGF signaling in human lung adenocarcinoma cells and CHOK1 cells stably expressing human EGFR. We demonstrated that SP-A inhibited the proliferation and motility of A549 cells. A binding study using 125 I-EGF suggested that SP-A blocked the binding of EGF to EGFR. Ligand blotting suggested that SP-A bound to the EGFR of A549 cells. We purified recombinant sEGFR and showed that SP-A directly bound to sEGFR. Surface plasmon resonance analysis revealed that the dissociation constant of SP-A and

SP-A down-regulates EGF signaling
sEGFR was K D ϭ 2.4 ϫ 10 Ϫ7 M. The binding of SP-A to sEGFR was not blocked by EDTA, excess mannose, or PNGase F treatment, suggesting that the binding is not via lectin activity of SP-A. Additionally, we demonstrated that CRF of SP-A bound to sEGFR, but CRF did not suppress the EGF-induced phosphorylation of EGFR or the proliferation of A549 cells. These findings suggest that SP-A down-regulates EGFR activity by different mechanisms from SP-D and that oligomerization of SP-A is important for the suppressing effects of SP-A on EGF signaling.
SP-A has been implicated in the regulation of innate immune responses in the lung (3,4,19). For example, SP-A prevents dissemination of infectious microbes by their biological activities, including agglutination and growth inhibition. SP-A also promotes clearance of microbes by enhancing phagocytosis in macrophages. In addition, SP-A interacts with the other pattern-recognition molecules, including toll-like receptors (TLRs) and TLR-associated molecules CD14 and MD-2, and regulates inflammatory responses. In addition to the antibacterial activity, SP-A has various anti-tumor activities. It was previously reported that SP-A expression in human lung adenocarcinoma cells suppressed the progression of tumors in subcutaneous xenografts and in lung metastasis mouse models by regulating and activating NK cells by controlling the polarization of tumor-associated macrophages (15). Furthermore, CD245 as myosin 18A, a receptor for SP-A, induced the enhancing NK cell cytotoxicity toward tumor cells (20). The suppressing effects of SP-A on EGF signaling might be involved in one of the various mechanisms of SP-A anti-tumor activity. It is possible that SP-A interacts with other receptors and modulates their functions. Those mechanisms might be involved in the anti-tumor effects of SP-A observed in the previous studies.

Figure 5. SP-A binds to EGFR in A549 cells, H441 cells, and CHOK1 cells stably expressing EGFR.
A, whole-cell lysate of A549 cells was immunoprecipitated (IP) with anti-EGFR monoclonal antibody Ab-11 or control IgG (0. 8 g) at 4°C for 16 h. The samples and BSA (200 ng/lane) were subjected to SDS-PAGE and transferred onto PVDF membranes. The membranes were incubated with or without SP-A (1 g/ml) for 16 h. The membranes not incubated with SP-A were subjected to Western blotting using an anti-human EGFR monoclonal antibody (WB: EGFR). The membranes incubated with SP-A were then treated with an anti-human SP-A polyclonal antibody (SP-A pAb) or monoclonal antibodies PE10 (SP-A mAb (PE10)) or PC6 (SP-A mAb (PC6)), which was followed by incubation with HRP-labeled anti-rabbit IgG or anti-mouse IgG (upper panel). As a negative control, the membranes in the absence of incubation with SP-A were also incubated with an anti-human SP-A polyclonal antibody (WB: SP-A pAb) or monoclonal antibodies PE10 (WB: SP-A mAb (PE10)) or PC6 (WB: SP-A mAb (PC6)), which was followed by incubation with HRP-labeled anti-rabbit IgG or anti-mouse IgG (lower panel). The data are representative of three independent experiments. B and C, experimental paradigm described in A was performed in H441 cells (B) and CHOK1 cells stably expressing EGFR (C) by using an anti-human SP-A polyclonal antibody, which was followed by incubation with HRP-labeled anti-rabbit IgG. The data are representative of three independent experiments.

SP-A down-regulates EGF signaling
We showed that the structural arrangement of the dodecamer of SP-A is crucial for the regulation of EGFR (Fig. 7). We previously reported the direct interaction between SP-A and the complex of TLR 4 with MD-2 and the importance of oligomerization in the immunomodulatory function of SP-A (16). It has been reported that the structural organization of SP-A influences its function and can be linked to disease severity in patients with cystic fibrosis (21). The single nucleotide polymorphisms (SNPs) that cause an amino acid variation in SP-D influence oligomerization and function (22) and are associated with risk of lung cancer (23). Additional studies have reported that the status of oligomerization is highly involved in the func-tions of another collectin, mannose-binding lectin (MBL), which forms a bouquet-like octadecamer as SP-A (24,25). It is important to research the mechanisms by which the oligomerization affects the anti-tumor effects of SP-A.
This study showed that SP-A down-regulates EGFR by mechanisms differing from those of SP-D. Our study indicated that through the interaction between CRD of SP-D and N-glycans of EGFR, SP-D suppresses EGF signaling by interfering with ligand binding to EGFR, EGFR dimerization, or EGFR autophosphorylation (11,26). In contrast, lectin activity of SP-A and N-glycans of EGFR were not involved in the interaction between SP-A and EGFR. Previous studies also indicate the Figure 6. SP-A binds to the extracellular domain of EGFR. A, sEGFR was produced in Flp-In CHOK1 cells and purified as described under "Experimental procedures." 0.5 g of proteins with or without PNGase F treatment were subjected to SDS-PAGE, which was followed by Coomassie Brilliant Blue R-250 staining (CBB). sEGFR with or without PNGase F treatment was electrophoresed, transferred onto PVDF membranes, and subjected to immunoblotting using an anti-EGFR monoclonal antibody Ab-5 (WB: EGFR (Ab5)) and an anti-His tag polyclonal antibody (WB: His-tag) and then HRP-labeled anti-mouse or anti-rabbit IgG. The same membranes were incubated with biotinylated ConA (Con A) and DSA (DSA) and then HRP-labeled streptavidin. B, indicated concentrations of SP-A were incubated with sEGFR (100 ng/well) or BSA (100 ng/well) coated onto microtiter wells at room temperature for 2 h in the presence of 2 mM CaCl 2 . ELISA was performed as described under "Experimental procedures." The data shown are the means Ϯ S.D. (error bars) from three independent experiments. A Student's t test or Welch correction was used for statistical comparisons. *, p Ͻ 0.05; **, p Ͻ 0.01. C, same experiment as that in B was performed in the presence of 2 mM EDTA instead of CaCl 2 . D, same experiment as that in B was performed in the presence of 2 mM CaCl 2 with 0.2 M mannose. E, same experiment as that in B was performed with sEGFR with or without PNGase F treatment. F, same experiment as that in B was performed in the presence of 2 mM CaCl 2 with 0.5 M NaCl. G, parameters of bindings of sEGFR to SP-A were determined by surface plasmon resonance analysis as described under "Experimental procedures." Sensorgrams for the bindings of various concentrations of SP-A to sEGFR immobilized on a sensor chip are shown. For the running buffer, 5 mM Tris-HCl (pH 7.4), containing 0.15 M NaCl and 2 mM CaCl 2 , was used. H, same experiment as that in G was performed in a running buffer containing 2 mM EDTA instead of CaCl 2 . I, same experiment as that in G was performed in a running buffer containing 0.2 M mannose. J, sensorgrams for the bindings of various concentrations of SP-A to sEGFR with PNGase F treatment immobilized on a sensor chip are shown. The same running buffer as that in G was used. K, sensorgrams for binding of SP-A (40 nM) to sEGFR immobilized on a sensor chip in a running buffer containing 0.15 or 0.5 M NaCl are superimposed. RU, response units.

SP-A down-regulates EGF signaling
functional differences between SP-A and SP-D. Both SP-A and SP-D modulate LPS-induced cellular responses by direct interaction with CD14 in the innate immune system of the lung (3). Although SP-D binds to CD14 via its lectin property, SP-A does not bind to CD14 via lectin activity. Rather, the neck domain of SP-A is involved in the binding to CD14 (27,28). The antiinfluenza virus functions of SP-A are lectin activity-independent, whereas the effects of SP-D are lectin activity-dependent (4). We previously reported that SP-A interacts with human ␤-defensin 3 (hBD3), independently of its lectin activity, although the functional region of SP-A lies within CRD (29). To further investigate the various functions of SP-A, it might be ideal to focus on the lectin-independent activity of SP-A.
We cannot quantify the actual concentrations of SP-A in alveolar space. However, it has been reported that its concentrations can be presumed based on the recovery of the proteins in the bronchoalveolar lavage fluids and the extrapolated hypophase volume. The calculated concentration of SP-A ranges bindings of sEGFR to CRF of SP-A were determined by surface plasmon resonance analysis as described under "Experimental procedures." Sensorgrams for the bindings of various concentrations of CRF of SP-A to sEGFR immobilized on a sensor chip are shown. C, A549 cells were serum-starved overnight and incubated with various concentrations of CRF of SP-A for 2 h at 37°C. After incubation, the cells were washed in a medium without serum and stimulated with 10 ng/ml EGF for 10 min at 37°C. The cell lysate was prepared, and 15 g of protein/lane were subjected to Western blotting (WB) using indicated antibodies. Lower panels display densitometric analyses, and data are presented as mean Ϯ S.D. (error bars) from three independent experiments. Data were expressed as percentage of phosphorylation relative to that of control cells treated with EGF and without CRF. Student's t test or Welch's t test was used for statistical comparisons. D, A549 cells were incubated with various concentrations of SP-A or CRF of SP-A, and the cell proliferation was assayed after 72 h. The data shown are the mean Ϯ S.D. (error bar) from three independent experiments. Data were expressed as percentage of proliferation relative to that of untreated control cells. Student's t test or Welch correction was used for statistical comparisons. **, p Ͻ 0.01 (compared with CRF with SP-A at the same concentration).

SP-A down-regulates EGF signaling
from 180 to 1800 g/ml (30,31). Additionally, it has been reported that about half of all lung adenocarcinomas express SP-A (32)(33)(34)(35). Therefore, the concentrations of SP-A in the cancer tissue of lung cancer patients are thought to be higher than those in healthy persons. The SP-A concentrations used in this study are within the best estimates of the physiological ranges of the lung tissue. EGFR is expressed in the basolateral surface of the epithelial cells (26), whereas SP-A is secreted into the alveolar space. Therefore, we infer that a small amount of SP-A in serum or interstitial tissues might weakly regulate EGF signaling in normal alveolar epithelial cells. Meanwhile, when the normal lung structures are destroyed by lung tumors, it is possible that a large amount of SP-A in the alveolar space interacts with EGFR in lung cancer cells and down-regulates strongly.
According to a recent study, phosphorylation of EGFR was increased in the lungs by mycoplasma membrane fraction stimulation in SP-A knock-out mice (36). These results suggested that SP-A down-regulates EGFR activity in vivo. We showed that SP-A directly interacted with the extracellular domain of EGFR and inhibited EGF-induced phosphorylation of EGFR in vitro. Further in vivo studies are needed to examine whether SP-A suppresses EGF signaling and as well as whether recombinant SP-A inhibits lung cancer progression. SP-A exists in the blood of healthy persons, so it is expected that recombinant SP-A has the potential to be a new therapeutic reagent candidate with few side effects.

Reagents and antibodies
Human recombinant EGF was purchased from Sigma, and 125 I-EGF was purchased from PerkinElmer Life Sciences. Gefitinib was purchased from SYNkinase (Melbourne, Australia). The monoclonal antibody to EGFR D38B1 (catalog no. 4267), the polyclonal antibodies to Akt (catalog no. 9272) and ERK (catalog no. 9102), and phospho-specific polyclonal antibodies to EGFR (Tyr-1068, catalog no. 2234), Akt (catalog no. 9271), and ERK (catalog no. 9101) were purchased from Cell Signaling Technology (Danvers, MA). The monoclonal antibodies to EGFR Ab-5 (catalog no. MS-316) and Ab-11 (catalog no. MS-396) were from NeoMarkers (Fremont, CA). The polyclonal antibody to His tag was from MBL (D291-3) (Nagoya, Japan). Anti-rabbit HRP IgG (catalog no. W4011) and antimouse HRP IgG (catalog no. W4021) were from Promega (Madison, WI). To produce the polyclonal antibody against human SP-A, purified recombinant human SP-A was injected into New Zealand White rabbits intramuscularly, and the antiserum was purified (16). Hybridomas producing anti-human SP-A monoclonal antibodies PE10 and PC6 were established, and antibodies were prepared as described previously (37).

Cell culture
The A549 human lung adenocarcinoma cell line was obtained from ATCC (Manassas, VA) and maintained in DMEM (Sigma) with 10% (v/v) FCS. The H441 human lung adenocarcinoma cell line was obtained from ATCC and maintained in RPMI 1640 medium (Sigma) with 10% (v/v) FCS. The Flp-In CHO cell line was obtained from Invitrogen, and main-tained in Ham's F-12 (Sigma) with 10% (v/v) FCS. CHOK1 cells expressing SP-A were grown in glutamate-free Glasgow Minimum Essential Medium (Sigma) containing 10% (v/v) dialyzed FCS.

Purification of recombinant human SP-A
Recombinant human SP-A was expressed in CHOK1 cells using the glutamine synthetase gene amplification system and purified using a mannose-Sepharose 6B column (16,38). The endotoxin content in the recombinant SP-A was Ͻ1.0 pg/g protein as determined by the Limulus amebocyte assay (16). The physical form of the recombinant SP-A used in this study was confirmed by rotary shadow electron microscopy (18).

Establishment of human EGFR and the extracellular domain of EGFR (sEGFR) stable expression cells
To establish human EGFR and sEGFR stable expression cells, the Flp-In system (Invitrogen) was used (11). cDNA for human EGFR and the Myc-His-tagged sEGFR (residues 1-618 of the mature protein) were subcloned into a pcDNA5/FRT expression vector and transfected into Flp-In CHOK1 cells with pOG44 plasmids.

Protein sample preparation and Western blotting
Cells were lysed as described previously (11,39). The concentrations of protein samples were measured using a protein assay kit from Bio-Rad. The samples were subjected to SDS-PAGE and transferred to PVDF membranes (Millipore, Billerica, MA). The membranes were blocked with 5% (w/v) BSA for 1 h, and the blots were probed overnight at 4°C with the indicated antibody. Immunoreactivity was detected by using anti-mouse or anti-rabbit IgG-conjugated HRP and then developed by Super-Signal West Pico (Pierce). A luminous analyzer was used for densitometric analysis.

Cell proliferation assay
A549 cells were plated onto a 96-well. After a 24-h incubation with DMEM with 10% (v/v) FCS, cells were treated with different concentrations of SP-A as indicated in the figures. After 24, 48, and 72 h, the WST-1 reagent (Takara, Japan) was added to each well, and the absorbance at 440 nm was measured on the plate reader.

Cell migration and invasion assays
Cell migration and invasion assays were performed using the transwell double chamber (BioCoat Matrigel Invasion Chamber, BD Biosciences). A549 cells were seeded into the upper insert in DMEM with 0.1% (v/v) bovine serum albumin (BSA) with or without SP-A, and DMEM with 10% (v/v) FCS as a chemoattractant was then added to the bottom wells. After fixation with 4% paraformaldehyde/PBS(Ϫ), the cells were permeabilized and stained with DAPI (200 ng/ml) in PBS(Ϫ). The cells were counted under a fluorescence microscope (Keyence, Osaka, Japan).

Wound-healing assay
Cell wound-healing assays were conducted using ibidi chambers (ibidi, Martinsried, Germany). A549 cells were applied into SP-A down-regulates EGF signaling each well. After incubation for 24 h, the culture inserts were removed, and the dishes were filled with a serum-free medium. EGF (100 ng/ml) and SP-A (20 g/ml) were added to the medium, and the cells were incubated for 24 h. After the incubation, the cells were stained with Diff-Quik (Sysmex, Kobe, Japan).

Binding analysis of 125 I-EGF to EGFR
A549 cells were seeded in a 24-well plate and allowed to adhere overnight. The cells were washed with DMEM containing 0.1% (w/v) BSA and incubated with the indicated concentration of SP-A for 2 h at 37°C in the same medium. The medium was removed; the cells were washed with the same medium, and 125 I-EGF was added in the presence of unlabeled EGF to reach the indicated concentration. Non-specific binding was determined by adding 1,000 nM cold EGF. After incubation for 2 h at 4°C, the cells were washed three times with ice-cold PBS(Ϫ) containing 0.1% (w/v) BSA and hydrolyzed in 0.5 ml of 1 N NaOH at room temperature for 30 min. The radioactivity of the cell lysate was quantified using a ␥-counter.

Immunoprecipitation
The whole-cell lysates of A549 cells were incubated with anti-EGFR monoclonal antibody Ab-11 (0.8 g) and protein A-Sepharose beads (Sigma) (15 l) at 4°C for 16 h. After being washed with a lysis buffer, the beads were boiled in the SDS sample buffer. The immunoprecipitates were separated by SDS-PAGE and transferred onto PVDF membranes.

Cell-surface biotinylation assay
A549 cells were serum-starved overnight. After a 2-h incubation with 20 g/ml SP-A, the cells were washed twice with PBS(ϩ) and treated with 0.5 mg/ml sulfo-N-hydroxysulfosuccinimide long-chain biotin (Sulfo-NHS-LC-biotin; Pierce) for 30 min at 4°C. The biotinylation reaction was quenched by washing the cells with 50 mM Tris-HCl (pH 7.4) containing 0.15 M NaCl, and the cells were rinsed with PBS(ϩ) and harvested. Whole-cell lysates were immunoprecipitated with the monoclonal anti-EGFR antibody (clone Ab-11), separated by SDS-PAGE, and transferred to PVDF membranes. Membranes were blocked with 3% (w/v) BSA and visualized by incubation with HRP-conjugated streptavidin (Vector Laboratories, Burlingame, CA).

Ligand blot
EGFR in A549 cells was immunoprecipitated, separated by SDS-PAGE, and transferred onto PVDF membranes. After being blocked with 3% (w/v) BSA for 1 h, the membranes were incubated with SP-A (1 g/ml) overnight at 4°C. After being washed with 20 mM Tris-HCl (pH 7.4) containing 0.15 M NaCl, 0.05% (v/v) Tween 20, bound SP-A was probed overnight at 4°C with an anti-human SP-A polyclonal antibody, monoclonal antibody PE10, or monoclonal antibody PC6.

Purification of recombinant sEGFR
sEGFR stable expression cells were cultured, and the medium was collected. The expressed Myc-His-tagged sEGFR was purified by a series of column chromatographies on His-Trap HP5 (GE Healthcare), Mono Q (GE Healthcare), and HiLoad Superdex 200 pg (GE Healthcare) using the AKTA purifier system (GE Healthcare) (39). Purity of sEGFR was determined by SDS-PAGE, and the protein identification was confirmed by the MALDI-TOF/TOF mass spectrometer (AB Sciex, Framingham, MA).

Treatment of sEGFR with PNGase F
sEGFR was treated with PNGase F (Takara, Japan) (10 milliunits/25 g protein) for 24 h at 37°C under non-reducing conditions. After incubation, the state of cleavage was confirmed by SDS-PAGE, immunoblotting, and a lectin blot. The cleavage was dialyzed against PBS(Ϫ) before being used for ELISA and surface plasmon resonance analysis.

Lectin blot
sEGFR treated with or without PNGase F was also electrophoresed and transferred onto PVDF membranes. After being blocked with 3% (w/v) BSA for 1 h, the membranes were incubated with ConA or DSA (J-Oil Mills, Tokyo, Japan) (4 g/ml) overnight at 4°C and visualized by incubation with HRP-labeled streptavidin (Vector Laboratories, Burlingame, CA).

Binding assay of SP-A to sEGFR with ELISA
sEGFR (100 ng/well) was coated onto microtiter wells, and nonspecific binding was blocked with 5 mM Tris-HCl buffer (pH 7.4) containing 0.15 M NaCl, 2 mM CaCl 2 , and 5% (w/v) BSA (binding buffer). The indicated concentrations of SP-A in binding buffer were added and incubated for 2 h. After incubation, the wells were washed with 5 mM Tris-HCl buffer (pH 7.4) containing 0.15 M NaCl, 2 mM CaCl 2 (washing buffer) and incubated with an anti-SP-A polyclonal antibody (2 g/ml) for 1 h. After being washed with washing buffer, HRP-conjugated goat anti-rabbit IgG was added and further incubated for 1 h. Finally, a peroxidase reaction was performed by using o-phenylenediamine as a substrate. To eliminate the effect of Ca 2ϩ on the binding, in some experiments we included 2 mM EDTA instead of CaCl 2 in binding buffer. To examine the effect of mannose, we included 200 mM mannose in binding buffer in some experiments. To examine the electrostatic interaction, we included 0.5 M NaCl in binding buffer.

Preparation of CRF of SP-A
The collagenase digestion of SP-A was performed as described previously (16,29). SP-A was treated with collagenase III from Clostridium histolyticum (Advance Biofactures Corp.) at 37°C for 22 h. After incubation, CRF was isolated by gel filtration using Superose 610/300 GL (GE Healthcare).

Binding assay of SP-A or CRF of SP-A to sEGFR with surface plasmon resonance analysis
sEGFR (20 g/ml) in 10 mM sodium acetate (pH 5.0) was immobilized on a sensor chip C1 of the Biacore 3000 system (Biacore, Uppsala, Sweden), according to the manufacturer's specifications. For the running buffer, 5 mM Tris-HCl (pH 7.4), containing 0.15 M NaCl and 2 mM CaCl 2 was used. The association rate constant (k a ) and the dissociation rate constant (k d )

SP-A down-regulates EGF signaling
were calculated according to the BIAevaluation software (Version 3.1, Biacore AB).

Statistics
Data are presented as mean Ϯ S.D. Differences in cell proliferation, migration, invasion, and absorbances at 492 nm in ELISA were evaluated by Student's t test or Welch's t test. Differences were considered statistically significant at a two-tailed test in which p Ͻ 0.05.