Human Pulmonary Surfactant Protein (SP-A), a Protein Structurally Homologous to C l q , Can Enhance FcR- and CR1-mediated Phagocytosis*

Clq, a subunit of the first component (Cl) of the classical complement pathway, and the pulmonary surfactant protein SP-A are structurally homologous mol- ecules, each having an extended collagen-like domain contiguous with a non-collagenous domain. It is the collagen-like region of Clq that binds to mononuclear phagocytes and mediates the enhancement of phago- cytosis of opsonized particles by these cells. Because SP-A enhances the endocytosis of phospholipids by alveolar type I1 cells and alveolar macrophages, we examined whether these two molecules were function- ally interchangeable. The phagocytosis of sheep erythrocytes opsonized with IgG or with IgM and comple- ment was enhanced by the adherence of monocytes or macrophages, respectively, to SP-A. The enhanced re- sponse was dependent on the concentration of SP-A used for coating the surfaces, similar to that seen when monocytes were adhered to Clq-coated surfaces. Both the percentage of cells ingesting the opsonized targets and the number of targets ingested per cell increased with increasing concentrations of SP-A. No such en- hancement was seen with cells adhered to albumin, iron-saturated transferrin, or uncoated surfaces. How- ever, SP-A did not substitute for Clq in the formation of hemolytically active and for radioactivity. Hemolytic Assays-Clq hemolytic activity was determined by the method of Kolb et al. (41). The assays of C1 activity reconstituted from purified Clq, Clr, and Cls were performed basically as described by Rapp and Borsos (37) with the modifications described by Tenner and Frank (42). These assays were performed with or without prein- cubation of SP-A with Clq, Clr, and Cls.

Clq, a subunit of the first component (Cl) of the classical complement pathway, and the pulmonary surfactant protein SP-A are structurally homologous molecules, each having an extended collagen-like domain contiguous with a non-collagenous domain.
It is the collagen-like region of Clq that binds to mononuclear phagocytes and mediates the enhancement of phagocytosis of opsonized particles by these cells. Because SP-A enhances the endocytosis of phospholipids by alveolar type I1 cells and alveolar macrophages, we examined whether these two molecules were functionally interchangeable. The phagocytosis of sheep erythrocytes opsonized with IgG or with IgM and complement was enhanced by the adherence of monocytes or macrophages, respectively, to SP-A. The enhanced response was dependent on the concentration of SP-A used for coating the surfaces, similar to that seen when monocytes were adhered to Clq-coated surfaces. Both the percentage of cells ingesting the opsonized targets and the number of targets ingested per cell increased with increasing concentrations of SP-A. No such enhancement was seen with cells adhered to albumin, iron-saturated transferrin, or uncoated surfaces. However, SP-A did not substitute for Clq in the formation of hemolytically active C 1. C l q did not stimulate lipid uptake by alveolar type I1 cells or alveolar macrophages and had only a slight inhibitory effect on the binding of SP-A to alveolar type I1 cells. Thus, these results suggested that a function which requires interactions of both the collagenous and the non-collagenous regions (Le. initiation of the classic complement cascade) could not be mimicked by a protein sharing structural macromolecular similarity but lacking sequence homology in the non-collagen-like region. However, SP-A could substitute for C l q in stimulating a function previously shown to be mediated by the collagen-like domains of the C 1 q molecule. collagenous amino acid sequences. SP-A,' a major pulmonary surfactant protein (I), mannose-binding protein (2), corespecific lectin (3), acetylcholinesterase (4), and conglutinin (5) are the other proteins outside the collagens to have such a structure. Clq is a 460,000-dalton glycoprotein comprised of six subunits, each of which contains three distinct polypeptide chains (6). The NH2-terminal third of all three chains of Clq has an amino acid sequence and structure very similar to collagen (7), with repeating Gly-X-Y triplets and a high amount of proline and hydroxyproline. While C l q in serum is normally in association with Clr2s2 as the C1 complex, upon dissociation of C112S2 by C1 inhibitor after C1 activation by antibody-antigen complexes or other C1 activators (reviewed in Ref. 8), the collagen-like "tail" region of C l q becomes more exposed. We have shown that the region of Clq which is exposed upon dissociation of Clrzs2 from Clq can interact with specific cells of immunological importance: specifically B lymphocytes, monocytes, macrophages, polymorphonuclear leukocytes, and a small population of cells lacking B and T cell markers (9,10). Other laboratories have demonstrated that lymphoblastoid cells, fibroblasts, and a variety of other cell types bind C l q (11)(12)(13)(14)(15). The physiologic significance of these receptors for Clq is currently unclear. However, surface-bound C l q stimulates oxidative metabolism in neutrophils (16), modulates FcR-mediated (17) and CR1-mediated (18,19) phagocytosis in monocytes and macrophages, stimulates immunoglobulin production by B cells (20), and enhances fibroblast adherence (21).
Human (22, 23) and canine (24) SP-A have been cloned and the amino acid sequence derived from the sequenced cDNA. While slightly different sequences have been published for human SP-A (22, 23), it is yet unknown if these represent more than one gene for SP-A or genetic polymorphism. Like Clq, SP-A also has collagen-like sequence in the aminoterminal half of the protein, short NH2-terminal domains containing interchain disulfide bonds, and a break in the Gly-X-Y repeat pattern near the middle of the collagen-like sequence. At the amino acid sequence level these proteins are homologous in the collagen-like region primarily due to the repeating Gly-X-Y triplet and a high percentage of proline/ hydroxyproline. Voss et al. (25) recently showed that the macromolecular structure of SP-A is remarkably similar to the hexameric structure of Clq as visualized in the electron microscope. SP-A is suspected of having a role in the regulation of the level of lung surfactant, as it has been shown to enhance the uptake of lipids by isolated alveolar type I1 cells (26). Given the ability of Clq to modulate cellular phagocytic function and SP-A to enhance lipid uptake, we investigated the possibility that the common structural features of the collagen-like domains may provide a basis for common biological functions.

MATERIALS AND METHODS
Media, Reagents, and Antibodies-RPMI 1640 medium was purchased from Gibco and HL1 medium from Ventrex Laboratories (Portland, ME). Fetal bovine serum was purchased from Hyclone (Logan, Utah). L-Glutamine and gentamicin were obtained from M. A. Bioproducts (Walkersville, MD). The human serum albumin used in the phagocytosis assay and phorbol dibutyrate were purchased from Sigma. The human serum albumin used for the elutriation buffer was obtained from Travenol Laboratories, Inc., Glendale, CA. Medium and fetal bovine serum for the alveolar type I1 cell and macrophage studies were obtained from the University of California Cell Culture Facility.
F(ab')? fragments of a polyclonal goat anti-human Clq were prepared as described previously (10) except that octanoic acid was used to obtain the serum IgG fraction (27). The IgG fraction was affinitypurified by passage over Sepharose-Clq and elution with isothiocyanate. Anti-Clq antibodies were also obtained commercially from Atlantic Antibodies, Scarborough, ME (IgG fraction) and Genzyme (Boston, MA) (antisera). Monoclonal antibodies to C l q were the generous gift of Drs. Verne Schumaker (University of California, Los Angeles) and Linda Curtiss (Scripps Clinic and Research Foundation, La Jolla, CA). Rabbit anti-SP-A antiserum was passed over Sepharose 4B-human serum and the IgG fraction isolated by DEAE-cellulose chromatography. Monoclonal antibodies to recombinant human SP-A were provided by Dr. J a n Marian and Jeanine Roderms (California Biotechnology, Incorporated, Mountain View, CA).
Protein Isolation-Clq was isolated from plasma-derived human serum by the method of Tenner et al. (28) modified as described. 2 The preparations used were fully active as determined by hemolytic titration and homogeneous as assessed by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (Fig. 1). SP-A was isolated from the lavage fluid of patients with alveolar proteinosis as described previously (26). Briefly, the surfactant was purified by density gradient centrifugation. The lipids were extracted with butanol. The butanol-insoluble proteins were extracted with 20 mM octylglucopyranoside to remove contaminating serum proteins. SP-A was solubilized in 5 mM Tris and dialyzed against the same solution. SP-A preparations were greater than 99% pure as assessed by densitometry. The monomeric (34 kDa) and dimeric (68 kDa) forms of SP-A (43) from a representative preparation of SP-A are seen in Fig. 1. It has been previously determined that the NHZ-terminal (residues 1-22) amino acid sequence of SP-A from alveolar proteinosis patients is identical to that of normal human SP-A and that it has the ability to enhance lipid uptake by type I1 cells (22). For the binding assays, rat SP-A was isolated in an identical manner and iodinated with Enzymobeads (Bio-Rad).
Cells-Human peripheral blood monocytes were isolated by counterflow elutriation using a modification of the technique of Lionetti et al. (29) as described (30). Macrophages as defined here were elutriated monocytes that had been cultured in Teflon jars (Savillex Corporation, Minnetonka, MN) at 1 X 10' cells/ml in HL-1 culture medium (a serum-free, defined medium) containing 2 mM L-glutamine and 10 pg/ml gentamicin. On days 2-6 of culture, macrophages were collected and washed three times in phosphate-buffered saline before use. Macrophages were assessed for the presence of myeloperoxidase by the use of a diaminobenzidine-based reaction (leukocyte peroxidase kit no. 391, Sigma). Freshly isolated alveolar type I1 cells were prepared by elastase digestion of rat lung. Cells were purified by differential adherence according to the methods of Dobbs et al. (31). The type I1 cell preparations averaged 85-90% pure. Alveolar macrophages were isolated by lavaging rat lungs eight times with a buffered salt solution containing EGTA.
Phagocytosis Assay-Phagocytosis was assessed as described previously (18,36). Sheep erythrocytes bearing IgG anti-sheep red blood cells (EAlfi) or IgM anti-Forssman antibody (EA,M) and C4b (EAC4b) were prepared as described previously (36). In experiments in which monocytes were cultured in HL-1, 10 ng/ml phorbol dibutyrate was added with the opsonized targets. This pretreatment of  10.6 pg of Clq and 13 pg of SP-A were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions using a 5-15% gradient gel according to Laemmli (40). Molecular weights on right refer to standards (Bethesda Research Laboratories) run in the middle lane.
HL-1-derived macrophages (which are less "activated" than serumcultured macrophages) optimizes phagocytosis of complement-opsinized targets as elaborated previously (18). The number of E-targets ingested per 100 effector cells was defined as the phagocytic index, whereas the percentage of effector cells ingesting a t least one E-target was defined as the percent phagocytosis. At least 200 effector cells were assessed per well, and each experiment used duplicate sample wells per condition. In some experiments, F(ab')? anti-Clq (in phosphate-buffered saline) was added to Clq-or SP-A-coated wells for 60 min a t ambient temperature prior to effector cell addition. In each experiment, unopsonized E and E opsonized with IgM anti-Forssman alone (E-IgM) were used as controls. These control E-targets were not ingested by monocytes or macrophages under any conditions. Phospholipid Uptake Assay with Lung Cells in Suspension-Isolated alveolar type I1 cells or alveolar macrophages (2 X 10') were incubated in 1 ml of a minimal essential medium containing Krebs' improved salts. The cells were incubated a t 37 "C for 60 min in room temperature air. Small unilamellar liposomes were prepared from surfactant-like lipids (Avanti Polar Lipids) consisting of dipalmitoylphosphatidylcholine, egg phosphatidylcholine, phosphatidylglycerol, and cholesterol in a molar ratio of 10:5:2:3. The liposomes were labeled with a trace amount of ['4C]dipalmitoylphosphatidylcholine. Proteins as designated were added to the cells after addition of liposomes. After an hour of incubation, the cells were separated from the media by centrifugation, washed twice by centrifugation, and analyzed for radioactivity and cell number. All of the methods are described in detail by Wright et al. (26).
Phospholipid Uptake with Adherent Lung Cells-Alveolar macrophages and isolated type I1 cells were resuspended in Dulbecco's modified Eagle's Medium (DME) containing 10% fetal bovine serum and 50 pg/ml gentamicin. Cells were plated in 24-well plates, at a density of 1 X 10' cells/0.5 ml/well, that had been precoated with either SP-A or Clq or buffer (as in the phagocytosis experiments (18)). After 16 h, the cells were washed and 0.5 ml of DME containing 0.1% bovine serum albumin and liposomes with and without SP-A or C l q were added. After 60 min of incubation a t 37 "C, cells were washed one time with DME and three times with Dulbecco's phosphate-buffered saline. Samples were solubilized in 1% sodium dodecyl sulfate and analyzed for radioactivity.
SP-A Binding Assays-Iodinated SP-A (1 pg/ml) was incubated with rat alveolar type I1 cells in DME containing 0.1% bovine serum albumin (10' cells adhered per well in 24-well plate) a t 4 "C for 60 min. For some samples an excess of unlabeled SP-A (50 pg/ml) or collagen type IV (Sigma), fibronectin, or Clq (100 pg/ml) was included in the incubation media. Cells were washed one time with SP-A Enhances FcR-and CRl -mediated Phagocytosis 13925 DME and three times with Dulbecco's phosphate-buffered saline. Samples were solubilized in 0.1 N NaOH and analyzed for radioactivity. Hemolytic Assays-Clq hemolytic activity was determined by the method of Kolb et al. (41). The assays of C1 activity reconstituted from purified Clq, Clr, and Cls were performed basically as described by Rapp and Borsos (37) with the modifications described by Tenner and Frank (42). These assays were performed with or without preincubation of SP-A with Clq, Clr, and Cls.

SP-A Enhances the Ingestion of IgG-opsonized Targets by
Human Monocytes-Previous studies have demonstrated that extracellular matrix proteins (32), Clq (17), and collagen type IV (33),3 but not collagen type I (17), enhance the phagocytosis of opsonized particles by mononuclear phagocytes. Therefore, the ability of SP-A to mediate a similar effect on phagocytosis was assessed. Human peripheral blood monocytes purified by counterflow elutriation were allowed to adhere to Lab-Tek chamber wells that had been pretreated with increasing concentrations of Clq, SP-A, or iron-saturated transferrin. Data from one of three similar experiments, presented in Fig. 2, demonstrate that surface-bound SP-A can enhance Fc receptor-mediated phagocytosis in a concentration-dependent manner. This enhancement was seen both as an increase in the percentage of cells ingesting the target particles and as an increase in the number of targets ingested per cell.
Enhancement of Complement (CRl) Receptor-mediated Phagocytosis by SP-A-Fragments of the complement protein C3 have long been known to be important opsonins (34). The C3b receptor, CR1, on monocytes and macrophages is also capable of interacting with the complement protein C4b (35). Monoclonal antibody to CR1 completely blocks the interaction of EAC4b with these cells under conditions identical to the phagocytosis assays used here (36). Recently it has been demonstrated that the interaction of C l q with mononuclear cells enhances CR1-mediated phagocytosis (18) . Fig. 3 demonstrates that adherence of culture-derived macrophages to surfaces coated with different preparations of the SP-A protein also results in the enhancement of CR1-mediated phagocytosis. The enhancement of phagocytosis was dose-dependent and similar to that seen with Clq. No such enhancement was seen when cells were adhered to iron-saturated transferrin, bovine serum albumin, or human serum albumin (data not shown).
In an attempt to investigate the specificity of the cell interaction site on these structurally homologous proteins, we tested whether a polyclonal antibody to Clq, which had been previously shown to inhibit the Clq-induced enhancement of phagocytosis, could also inhibit the SP-A-induced enhancement of phagocytosis. Interestingly, in two separate experiments, while the F(ab')2 anti-Clq inhibited both the percentage of cells ingesting targets and the phagocytic index by 81.5 f 3.5% and 88.5 k 4.9%, respectively, no inhibition of SP-Aenhanced phagocytosis was detected in the presence of this F(ab'), anti-Clq (Fig. 4). In addition, a monoclonal antibody which inhibits Clq-mediated enhanced phagocytosis did not affect the enhanced response generated by surface-bound SP-A.
Clq Does Not Enhance the Uptake of Phospholipids by Isolated Alveolar Type 1 1 Cells or Macrophages-Given the above effect of SP-A on mononuclear cell phagocytic activity and the previously established ability of SP-A to enhance uptake of phospholipids by alveolar type I1 cells and macrophages (26), we investigated the possibility that a common element in the structure of these two proteins may be suffi-

SP-A Enhances
FcR-and CRl -mediated Phagocytosis  cient to trigger both ingestion functions. Thus, we tested whether Clq could mimic SP-A in its ability to enhance phospholipid uptake or whether Clq could interfere with the SP-A stimulated uptake. Rat alveolar type I1 cells and alveolar macrophages were isolated and subsequently incubated with phospholipid-containing liposomes in the presence or absence of SP-A or Clq. Table I demonstrates that while SP-A significantly stimulated phospholipid uptake as reported previously (26), Clq did not. Similar results were obtained with cells that had adhered to SP-Aor Clq-coated surfaces. In addition, the binding of lZ5I-SP-A to type I1 cells was only marginally inhibited by excess unlabeled Clq (Table 11). Thus, it appears that SP-A contains distinct structural properties which mediate these specific functions.
SP-A Is Unable to Substitute for Clq in Complement-mediated Hemolysis-Clq, as a subcomponent of the first component of complement CI, binds through its globular head regions to the Fc portion of immunoglobulins. This interaction triggers the conversion of the associated proenzyme subunits Clr and Cls to active serine esterases and thus activates the classical complement pathway. Since it is the collagenlike region of Clq that associates with both Clr and Cls (8) and interacts with the leukocyte Clq receptor (lo), we questioned whether SP-A could substitute for Clq in associating with Clr2s2 to form active C1. Using Clq-deficient serum and sheep erythrocytes coated with anti-sheep red blood cell IgG, the addition of SP-A (0.05-2.56 pg) did not lead to complement-mediated hemolysis, under conditions in which 5 ng of Clq resulted in maximal lysis (data not shown). In an alternative C1 assay (37) of the inhibition of C1 formation in which Clrzsn was equimolar with Clq (rather than in excess as in the previous assay), SP-A consistently inhibited C1 hemolytic activity, suggesting competition of SP-A and Clq for Clrzsp. However, this inhibition of hemolytic C1 formation was relatively weak in that a 20-fold excess of SP-A over Clq produced only 26-56% inhibition of the control hemolytic activity (i.e. in the absence of SP-A).
Analysis of Reactivity of Anti-Clq and Anti-SP-A with C l q and SP-A-Given the ability of SP-A to enhance the FcRand CR1-mediated phagocytosis, we screened available polyclonal and monoclonal antibodies reactive with Clq or SP-A for their reactivity with SP-A and Clq, respectively. By dot blot analysis no antibody tested was found to cross-react with both proteins (Fig. 5). It is known that the affinity-purified anti-Clq is largely reactive with the collagen-like region of Clq,4 and that two of the monoclonals reactive with intact SP-A are not reactive with the collagenase-resistant fragment.5

DISCUSSION
Clq and pulmonary surfactant protein, SP-A, are structurally homologous proteins. The data presented in this paper indicate that in vitro SP-A can enhance FcR-mediatedphagocytosis by monocytes and macrophages and CR1-mediated phagocytosis by macrophages, activities previously shown to be modulated by Clq as well as some components of extracellular matrix (32). However, SP-A can not substitute for Clq in the formation of hemolytically active C1, the recognition component of the classical complement pathway. Clq does not mimic the ability of SP-A to enhance the uptake of phospholipids by isolated alveolar type I1 cells or macrophages and has only a slight inhibitory effect on the binding of SP-A to alveolar type I1 cells. These observations are consistent with what is known about the homology between Clq and SP-A and the specific domains involved in the mediation of function by these proteins.
The homology between Clq and SP-A exists at the macromolecular level in that electron microscopy (25) demonstrated a nearly identical image of the two proteins as six globular domains connected by short arms to a common stem, and at the amino acid sequence level with repeating Gly-X-Y triplets and a high percentage of proline in the NHp-terminal region of the polypeptide chains. Excluding the common glycine residues in the collagen-like region of the molecules, 15, 21, and 20% of the amino acid residues of the Clq A, B, and C chains, respectively, are identical to the SP-A sequence. In contrast, the non-collagen-like region of Clq is not homologous to the non-collagenous region of SP-A. The isoelectric points of the intact proteins are also quite different, Clq being highly basic (PI > lo), whereas SP-A is acidic (PI = 4.6-5.0) (39). Thus, it is not surprising that SP-A does not cause complement-mediated hemolysis, as the globular (non-collagenous) region of Clq contains the interaction site for the immunoglobulin bound to the red cells. However, SP-A does i..lterfere, albeit weakly, with the efficient assembly of C1, which in-;dves the interaction of Clrzsz with the collagen-like domain of Clq. The observation that Clq was unable to mediate phospholipid uptake by alveolar type 11 cells or alveolar macrophages is consistent with the recent evidence of Wright and colleagues (44) demonstrating that the collagenase-resistant fragment of SP-A (i.e. the portion lacking homology with Clq) binds to alveolar type I1 cells. A third protein which is similarly homologous to Clq is the cation-dependent, mannose-binding protein/core-specific lectin (2,3). Analysis by Drickamer et al. (2) has shown this protein to be homologous with SP-A in the non-collagenous as well as the collagen-like region. Mannose-binding protein binds to mannose-containing structures on pathogenic organisms and has already been shown to play a role in cellular endocytic function (38). It has been speculated that both mannose-binding protein and, by analogy, SP-A, which also has mannose-binding capacity, play a protective role for the host. Similarly, Clq, in addition to its role in the initiation of the classical complement pathway, has been shown to interact with cells of the immune system, triggering enhancement of immune cell function in vitro (16)(17)(18)(19)(20). Since the non-collagenlike region of Clq is not homologous to either SP-A or mannose-binding protein, it can be speculated that the "recognition" of a specific target (Fc, mannose, phospholipid, etc.) is mediated by nonhomologous regions of the proteins, while the collagen-like regions mediate phagocytic functions.

Clq
Interestingly, in this study none of the antibodies to Clq or SP-A, including both polyclonal and monoclonal antibodies, cross-reacted with both of these proteins. Furthermore, neither the affinity-purified polyclonal anti-Clq nor a monoclonal anti-Clq known to interact with the collagen-like domain of Clq and to block the phagocytosis-enhancing effect of Clq inhibited the effect of SP-A on phagocytosis of either IgG or complement-opsonized targets. This would suggest that this monoclonal antibody and the polyclonal anti-Clq contain antibody that binds near, but not at, the cell binding residues of the Clq molecule, thereby sterically hindering ligand/receptor interaction. Alternatively, the cell interaction domains of Clq and SP-A may not be identical, or the proteins may act through different receptors. Nevertheless, while as yet there is no evidence that the cell interaction domains of any of these ligands are identical, the possibility exists that receptors for these molecules in the cell membrane may share common features and, in fact, may constitute a family of proteins.
It should be noted that, because neither Clq nor SP-A was required to be present on the particle being ingested in the phagocytic assays described above (Figs. 2 and 3), these molecules can function as activation ligands rather than merely as opsonic ligands which mediate binding of the targets to the cell, as discussed by Brown (32). Importantly, iron-saturated transferrin, which interacts with a cell membrane receptor, does not mediate this enhancement of function, indicating that this is not due to a non-specific ligand/receptor interaction (Fig. 2). Further definition of the specific receptor interaction domains on each of the proteins which share a macromolecular structural homology and are able to trigger phagocytosis will ultimately determine the structural features required for a ligand to trigger this cellular function. Whether the involved structures are identical among the different proteins and whether the mechanism of signal transduction for each ligand/receptor system is similar should provide insight for future potential manipulations of these systems.