INTRODUCTION

Serum amyloid P (SAP)¹ is a M_r ~230,000 glycoprotein encoded by a single gene on human chromosome 1 (1). It is a member of the highly conserved pentraxin family of proteins (2). These proteins are made up of five identical noncovalently bound subunits arranged in a flat pentameric disc (2). The biological role of pentraxins is not yet clear, however, they have been shown to mediate a variety of biological activities. SAP binds to the collagen-like region of C1q and activates the classical complement pathway (3). It binds to C4b-binding protein (C4bp) and prevents the factor I-mediated inactivation of C4b (4-6). It also binds to extracellular matrix (ECM) components such as type IV collagen, fibronectin, and proteoglycans (7-10). The binding of SAP to sulfated glycosaminoglycans such as heparin and dextran sulfate proteoglycans is thought to mediate the extravascular procoagulant activity of SAP (11). In vivo SAP is found in association with amyloid deposits of Alzheimer's disease and secondary amyloidosis (12). It is a normal component of a number of basement membranes including glomerular basement membrane (GBM), alveolar basement membrane, and sweat gland basement membrane (13, 14). The association of SAP with GBM is disrupted in a number of nephritides such as Alport's Syndrome, type II membranoproliferative glomerulonephritis, and membranous glomerulonephritis (15, 16).

Basement membranes are complex structures made up of a number of glycoprotein components (17). They form sheet-like structures in close association with tissues and organs. They are involved in maintaining the morphology of specific organs, filtration functions, and maintenance of the differentiated state and basal apical polarity of the cells associated with them (17, 18). The structure, and consequently the function, of a basement membrane is determined by factors such as the shape of its components, their concentration, their affinity, and their interaction (18, 19). Laminin isotypes are major components of all basement membranes (20). They are cruciform-shaped proteins that are composed of three polypeptide chains (21). They interact with other ECM components such as nidogen, type IV collagen, and proteoglycans (22-24). They also bind to cell surface receptors such as integrins and help in anchoring the cells to the ECM (24). Via the above interactions, various laminin isotypes play an important role in maintaining the structure and function of different basement membranes and their adjacent tissues (25-27).

Previous studies indicate that SAP constitutes approximately 10% of the protein released from the GBM after collagenase treatment (13). The pattern of association of SAP with specific basement membranes, its altered disposition in the GBM in a number of nephritides, and its interactions with components of the ECM suggest that it may play a role in determining the structure and function of the basement membrane. Therefore, characterizing the binding of SAP to ECM components and its effect on their interactions will provide a better understanding of the biological function of this molecule and the mechanism by which it affects the structure of the specific basement membranes with which it is associated. In the present study the binding of SAP to laminin was examined. SAP binding to laminin was specific, saturable, and dependent on the presence of calcium. The K_d of the interaction was 2.74 × 10⁷ M, and the molar ratio of SAP/laminin was 1:7.1. The inhibition studies indicated that the binding of SAP to laminin is most likely mediated via its galactan binding region or a site near this region. SAP was also shown to enhance the polymerization of laminin molecules in solution in a concentration-dependent manner. The binding of SAP to laminin did not alter the ability of laminin to serve as a cell adhesion substrate.

EXPERIMENTAL PROCEDURES

Human SAP and C-reactive protein (CRP) were purchased from Calbiochem. Each protein gave a single band of M_r ~23,0000 and 25,0000, respectively, when size-fractionated by SDS-polyacrylamide gel (12% acrylamide) under reducing conditions. Mouse SAP was purified from acute-phase mouse serum by Ca²⁺-dependent affinity chromatography on a column of phosphatidylethanolamine (PE) conjugated to agarose beads (Sigma) as described previously (28), followed by anion-exchange chromatography on a Mono Q column (Pharmacia Biotech Inc.). The purified protein gave a single band after size fractionation by SDS-polyacrylamide gel electrophoresis under reducing conditions. Mouse, rat, and human laminin and merosin were purchased from Life Technologies, Inc. Size fractionation of these proteins by SDS-polyacrylamide gel electrophoresis gave bands of M_r ~400,000 and M_r ~200,000. Crystalline PE, phosphorylcholine chloride (PC), and monoclonal anti-human SAP antibody were purchased from Sigma. Human SAP and CRP antibodies were purchased from DAKO.

Purified human SAP was iodinated by mixing 0.250 mCi of [¹²⁵I]sodium iodide (Amersham Corp.) with 500 µg of SAP in Tris-buffered saline (TBS; 20 mM Tris, 150 mM NaCl, and 10 mM EDTA) in a glass vial precoated with Iodogen reagent (Pierce). The reaction was allowed to proceed for 5 min at 25 °C. Unincorporated radioactivity was removed by desalting on a Microcon 50 microconcentrator (Amicon). The remaining protein was diluted in TBS containing 1 mM EDTA. Radioactivity of the final protein preparation was 95-98% precipitable with trichloroacetic acid. The iodinated SAP molecule ran as a single band on SDS-polyacrylamide gel and retained its Ca²⁺-dependent binding to phosphatidylethanolamine.

Laminin (1 µg/well) was coated overnight at 4 °C onto 96-well microtiter plates (Corning) using carbonate buffer (45.3 mM NaHCO₃ and 18.2 mM Na₂CO₃, pH 9.6). The plates were washed with TBS buffer (20 mM Tris, 150 mM NaCl, 2 mM CaCl₂, and 0.05% Tween 20, pH 7.45) containing 10 mg/ml blocking reagent (Boehringer Mannheim) and blocked with TBS blocking buffer (20 mM Tris, 150 mM NaCl, and 10 mg/ml blocking reagent, pH 7.45). Dilutions of SAP in TBS buffer were added to triplicate wells, and binding was allowed to proceed for 3 h at 37 °C. Wells were washed and incubated for 24 h at 4 °C with rabbit anti-human SAP antibody (DAKO). After washing, goat anti-rabbit IgG antibody conjugated to horseradish peroxidase (Calbiochem) was added to each well and allowed to bind for 90 min at 25 °C. Plates were washed, substrate solution (10 µg/ml O-phenylenediamine dihydrochloride in 50 mM acetic acid, 100 mM Na₂HPO₄, and 0.0003% H₂O₂, pH 5.0) were added to each well, and the color was allowed to develop for 15 min at 25 °C. The reaction was then stopped by the addition of 9.6% H₂SO₄. The absorption of each well at 492 nm was determined using an ELISA plate reader (Titer Tek, Co.).

Immulon I Removawells (Dynatech, Inc.) were coated with 60 µl of 50 µg/ml mouse laminin in phosphate-buffered saline (PBS). Wells were washed with TBS buffer and blocked with TBS blocking buffer. Wells were washed and incubated with dilutions of ¹²⁵I-SAP in TBS buffer (100 µg/well) for 20 h at 4 °C. To determine the amount of SAP added to each well, 2 µl of each sample were counted using a gamma counter (Isoflex, Co.). The results were converted to picomole concentrations using the calculated specific activity of ¹²⁵I-SAP. After washing the wells with TBS buffer, the bound radioactivity was measured by counting the entire well. Specific binding was calculated by subtracting nonspecific binding (bound radioactivity in the presence of 100-fold SAP in low SAP concentration samples or in the presence of 10 mM EDTA all samples) from total binding (binding in TBS buffer). The results were converted to picomole concentrations using the specific activity of ¹²⁵I-SAP. The amount of laminin bound to the wells was determined directly by measuring the bound protein levels using the BCA protein quantitation assay (Pierce). Briefly, wells coated with 60 µl of 50 µg/ml laminin were incubated with 100 µl of protein quantification reagent. Wells containing known levels of laminin were used to generate a standard curve. The bound laminin levels were determined by measuring the average A₅₆₂ of 12 wells and calculating the bound laminin levels based on the standard curve. Total bound laminin was determined to be approximately 1.7 ± 0.17 µg/well.

Samples containing 2.5 µg of ¹²⁵I-SAP and various concentrations of SAP and CRP (0-750 µg/ml) were added to each of the triplicate wells coated with 100 µl of 10 µg/ml laminin and incubated at 4 °C for 20 h. Wells were then washed with TBS buffer, and the bound counts were determined. The percentage of inhibition was determined by assuming that the binding was 100% in the absence of the inhibitor. The binding of ¹²⁵I-SAP in the presence of the inhibitor was then calculated as a percentage of the above. The inhibition of SAP binding to laminin by PE and PC was examined by ELISA. SAP (25 µg/ml) was incubated with increasing concentrations of PE and PC (0-500 mM) for 1 h at 37 °C. Samples (100 µl) were then added to laminin-coated wells and stored at 4 °C for 20 h. The rest of the assay was performed as described for the ELISA binding assay.

The effect of SAP on the polymerization of laminin was examined in a turbidometric assay. The development of turbidity in a solution of laminin (350 µg/ml) was monitored in the absence or presence of increasing concentrations of SAP (30-150 µg/ml). Laminin was thawed and dialyzed against 1 liter each of 100 mM Tris-HCl, pH 7.4; 0.5 M CaCl₂ and 100 mM Tris-HCl, pH 7.4; and 100 mM Tris-HCl, pH 7.4; and PBS. SAP and bovine serum albumin (BSA) were also dialyzed against PBS. All dialysis steps were performed at 4 °C for 24 h. All solutions contained 10 µl/liter of 100 mM phenylmethylsulfonyl fluoride. Both laminin and SAP preparations were cleared of aggregates by centrifugation. Aliquots of laminin in the presence or absence of SAP in a final volume of 1000 µl were incubated at 37 °C, and the change in their absorbance at 360 nm was monitored for 80 min.

The effect of SAP on the adhesion of human umbilical vein endothelial cells (American Type Culture Collection) to laminin substrate was examined using the methodology described by Sriramarao et al. (29). Briefly, 96-well microtiter plates (Linbro/Titertek, ICN Biomedicals Inc.) were coated with 1 pmol/well of laminin or BSA in PBS for 24 h at 4 °C. The remaining binding sites were blocked with PBS containing 10 mg/ml blocking reagent. Increasing amounts of SAP (0.01-5 µg/well in 50 µl) were added to each well, and the plates were incubated for 5 h at 37 °C. Cells were harvested by treatment with 0.5 mM EDTA and washed twice in Hanks' balanced salt solution. Cells were resuspended in Ultraculture serum-free medium (BioWhittaker) to a concentration of 2 × 10⁶ cells/ml, and 50 µl of the suspension were added to each well. The plates were incubated for 2 h at 37 °C. Nonadherent cells were removed by washing the plates with PBS containing 1 mM Mg²⁺ and Ca²⁺. Adherent cells were fixed and stained with PBS containing 3.75% paraformaldehyde and 0.5% crystal violet. Wells were washed twice with PBS, and adherent cells were quantitated by measuring the absorbance at 595 nm on a microtiter plate reader.

RESULTS

Nonquantitative ELISAs indicate that SAP binds to immobilized laminin and that the binding reaches equilibrium within 3-4 h at 37 °C (data not shown). To characterize the interaction of SAP with immobilized laminin quantitatively, the direct binding of ¹²⁵I-SAP to immobilized laminin was examined. Dilutions (100 µl/well) of ¹²⁵I-SAP (0.1-58 pmol/100 µl) were added to immobilized laminin. The binding of SAP to laminin approached saturation when 28-35 pmol of SAP were added to plates coated with 1.8-2.1 pmol of laminin (Fig. 1A). Scatchard analysis of the saturation binding data indicates that SAP binding to laminin has a K_d ~2.74 × 10⁷ M, a molar ratio of SAP (native decameric form)/laminin of 1:7.1 at saturation (Fig. 1B). Because the binding experiments described here use heterologous sources of protein (human SAP and mouse laminin), the binding of mouse SAP and human SAP to mouse laminin was compared by ELISA. SAP from both species binds to immobilized mouse laminin (Fig. 2). Saturation concentrations for both human and mouse SAP were approximately 20 µg/ml. The binding of human SAP to immobilized human laminin, mouse laminin-1 (composed of 1, 1, and 1 chains), rat laminin, and human merosin (composed of 2, 1, and 1 chains) was measured by ELISA. SAP binds to all four molecules. A comparison of SAP binding to human and mouse laminin indicates that SAP binding to human laminin is lower than SAP binding to mouse laminin (Fig. 3A). Similarly, the binding of SAP to merosin was lower than SAP binding to either mouse or rat laminin (Fig. 3B). This lower binding may be due to proteolytic cleavage of human laminin by pepsin during its extraction or to the differential binding of SAP by the  chains of laminin and merosin. These results indicate that SAP binding to laminin is saturable and has a relatively high affinity. Furthermore, this interaction is consistent between proteins from the same species or proteins isolated from different species.

7

The binding of pentraxins to their ligands and the polymerization of laminin are both Ca²⁺-dependent reactions (30, 31). SAP binding to a number of proteins including C1q (3), proteoglycans (10), type IV collagen (7), and C4bp (5) is dependent on the presence of Ca²⁺. The role of Ca²⁺ in the binding of SAP to laminin was examined (Fig. 4) by determining the extent of the binding of SAP (2.5 µg in 100 µl) to laminin in the presence (0.5-7 mM CaCl₂) or absence of Ca²⁺ (0-10 mM EDTA). The binding of SAP to laminin was enhanced up to 8-fold in the presence of Ca²⁺. Increased binding was observed in the presence of 0.5-1 mM Ca²⁺, whereas the binding of SAP to laminin in the presence of higher Ca²⁺ concentrations (2-7 mM) remained constant, about 8-fold greater than the binding observed in the absence of Ca²⁺. The binding of SAP to laminin diminished significantly upon the addition of EDTA. Furthermore, other divalent cations (Mg²⁺, Mn²⁺, and Zn²⁺) could not replace Ca²⁺ (data not shown). Under all conditions the binding of SAP to immobilized BSA was minimal. These data indicate that SAP binding to laminin is dependent on the presence of Ca²⁺ and that enhanced binding is observed within the physiological range of Ca²⁺.

CRP and SAP exhibit extensive structural and amino acid sequence homology. Previous studies indicate that CRP binds to laminin in a Ca²⁺-dependent manner via its PC binding site (32). Based on the structural and sequence homology of SAP and CRP and their binding to laminin, the ability of CRP to interfere with the binding of SAP to laminin was examined. Both SAP and CRP inhibited the binding of ¹²⁵I-SAP to immobilized laminin (Fig. 5), but BSA, even at very high concentrations (1000 µg/ml), did not inhibit the binding of SAP to laminin (data not shown). The data in Fig. 5 indicate that an approximately 6-fold molar excess of CRP as compared to SAP is required to inhibit the binding of ¹²⁵I-SAP to immobilized laminin by 50%. The results indicate that the binding of SAP to laminin is specific and occurs at a site that is similar or closely located to the CRP binding site on the laminin molecule.

To determine the role of the SAP galactan binding site in its interaction with laminin, the ability of PE, which binds to the galactan binding site of SAP (33), to inhibit the binding of SAP to laminin was examined. The binding of SAP to laminin was inhibited by PE but not by PC (Fig. 6). The maximum inhibition achieved in these assays was approximately 65%, which was observed when SAP was preincubated with 500 mM PE. These results indicate that SAP binding to laminin is mediated via its galactan binding site.

Inhibition of SAP binding to laminin by PE () and PC ().

The interaction of SAP with laminin raises the possibility that it may affect the polymerization of laminin, which in turn may affect the overall structure and function of the basement membrane. To determine the effect of SAP on the polymerization of laminin, turbidity assays were performed in the absence or presence of SAP (30-150 µg/ml) (Fig. 7). A comparison of the laminin only to laminin/SAP samples indicates that SAP enhanced the polymerization rate of the laminin in solution in a concentration-dependent manner. Control samples containing 150 µg/ml SAP were also examined. The A₃₆₀ of SAP remained constant throughout the experiment, indicating that self-polymerization of SAP is not responsible for the increased turbidity of the laminin/SAP samples. Furthermore, the presence of BSA (100 µg/ml) did not affect the polymerization of laminin. These data suggest that SAP can bind to laminin or polymerized laminin in solution and enhance its polymerization reaction or lattice formation.

The molecular composition and structure of basement membranes are major determinants of their interaction with adjacent cells and the phenotypes of these cells. The binding of SAP to laminin may affect its interaction with other ECM components and lead to changes in the structure of the basement membrane. This may in turn alter the cell matrix interactions and modify the phenotype of these cells. The effect of SAP on the ability of laminin to serve as a cell binding substrate was examined. Immobilized SAP did not support cell attachment. In addition, neither the binding of SAP to immobilized laminin (Fig. 8) nor its incorporation into a laminin matrix before immobilization (data not shown) had any effects on the binding of human umbilical vein endothelial cells to laminin.

DISCUSSION

SAP is a member of the pentraxin family of proteins (2). These are proteins with a high degree of sequence and structural homology that have been evolutionarily conserved (2, 34, 35). SAP is a constitutive component of human plasma (36). In vivo observations indicate that it is an integral part of a specific group of basement membranes including the GBM (13, 14). SAP binds to ECM components such as proteoglycans (e.g. heparan and dermatan sulfate proteoglycans) (10), fibronectin (8, 9), and type IV collagen (7). In the present study, the binding of SAP to laminin was examined. SAP binds to laminin molecules derived from a variety of sources. SAP binding to laminin was calcium-dependent, saturable, and specific, with a calculated K_d of 2.74 × 10⁷ M. Inhibition studies indicate that SAP binding to laminin is inhibited by both soluble SAP and its analog, CRP, suggesting that these proteins may bind to similar or closely located binding sites on the laminin molecule. The binding of SAP to laminin was inhibited by PE but not by PC, suggesting that the binding of SAP to laminin is mediated via its galactan binding site. The involvement of the lectin binding sites of pentraxins in their interactions with other molecules has been extensively documented. For example, the binding of CRP to laminin is mediated via its PC binding site (32). Previous studies indicate that the binding of SAP to C4bp and heparan and dermatan sulfate proteoglycans as well as amyloid fibrils is mediated through its galactan binding sites (6, 37, 38). Although the exact nature of this site is not known, studies by Loveless et al. (39) indicate that it includes amino acid residues 27-38 of the SAP molecule.

The structure and function of a basement membrane is determined by the interaction of its constituent components, their assembly, and their turnover (17, 18). Laminin is a major component of all basement membranes (20). It interacts with the cells that are in contact with the ECM (25-27), other ECM components (22, 23, 40, 41), and itself (31). Through these interactions laminin influences the structure and function of the basement membrane. Polymerization of laminin is an important step in the formation and maintenance of the basement membrane structure. Turbidity assays indicate that SAP enhances the polymerization of soluble laminin or laminin polymers. This observation as well as the Scatchard analysis data showing that SAP can bind multiple laminin molecules suggests that SAP may act as a nucleating or scaffolding agent by simultaneously binding to a number of molecules. This potential function becomes even more interesting when the ability of SAP to bind to different components of the ECM such as type IV collagen and proteoglycans is considered. The incorporation of SAP into the ECM may affect the structure and function of this matrix through a number of mechanisms. The binding of SAP to ECM components may modify their interaction and the kinetics of their assembly, thereby modifying the structure and, consequently, the function of the basement membrane. SAP may contribute to the maintenance of the net negative charge and the structure of the glomerular capillary wall basement membrane and the integrity of its filtration functions (13). Another potential mechanism by which SAP may affect the structure of the basement membrane is the alteration of basement membrane turnover. The SAP molecule is very compact and highly structured (42). It is also very resistant to proteolysis (43). The binding of SAP to amyloid fibrils derived from secondary amyloidosis and Alzheimer's disease plaques protects these molecules from digestion by a variety of proteases (38). It is therefore possible that the binding of SAP to components of the ECM and its incorporation into the basement membrane may protect this matrix from digestion by ECM proteases, modify its turnover and, consequently, affect its structure and function. Basement membranes and some of their components are cell adhesion substrates that influence the phenotypes of their adjacent tissue. Data presented here as well as those examining the effect of SAP on cell adhesion to type IV collagen and fibronectin² indicate that SAP probably does not play a role in tissue matrix interaction. The potential influence of SAP on the cell matrix interaction, however, cannot be completely ruled out because the mixture of ECM components and their interaction may create an environment that is quite different than the conditions used in the in vitro cell adhesion assays.

The majority of the present studies were performed using Engelbreth-Holm-Swarm tumor matrix-derived laminin. Differential localization of laminin isoforms has been well documented (44-46), and it is possible that SAP may have a more or less dramatic effect on the interaction and polymerization of other laminin isoforms specifically expressed in those basement membranes with which SAP is associated. This possibility is to a certain extent supported by the variable degrees of SAP binding to different laminin isoforms and laminin molecules from different species.

The association of SAP with a specific group of basement membranes has been documented in a number of studies (13, 14). This association cannot simply be due to the exposure of these basement membranes to circulating SAP because it is absent from a number of basement membranes that are exposed to high levels of SAP (basement membranes of the liver sinuses and venous sinuses of the spleen) (14). Furthermore, the association of SAP with the GBM is completely disrupted or altered in a variety of nephritides (15, 16). It is possible that the absence of SAP or its altered disposition as a result of an initial injury could lead to further alterations in the basement membrane. This may partially account for some of the pathological changes associated with diseases such as the nephritides mentioned here. The biological role of SAP in the basement membrane is not known. However, previous in vivo observations and in vitro experimental data describing the interaction of SAP with the ECM components indicate that it may play an important role in the structure and function of those basement membranes with which it is associated. Immunohistochemical studies indicate that SAP is associated with a specific group of basement membranes including the GBM, alveolar basement membrane, and sweat gland basement membrane as well as the basement membranes in the posterior chamber of the eye. It is possible that SAP binding to the ECM components can modify their interactions and lead to the development of specific capacities in those basement membranes with which it is associated. Further examination of the role of SAP in the basement membrane and the mechanism(s) by which it modifies the structure of the basement membrane is required to determine its biological significance and its role as a structural protein.