Characterization of the binding of serum amyloid P to type IV collagen.

Serum amyloid P (SAP), a member of the evolutionarily conserved pentraxin family, is a normal component of a number of basement membranes, including glomerular and alveolar. In vitro SAP binds to a variety of proteins including fibronectin, proteoglycans, and the collagen-like region of the complement component C1q. In these studies, binding of SAP to type IV collagen, a major component of basement membrane, was examined. Purified SAP binds to human and mouse type IV collagen but not type I, II, or III collagens. Binding of SAP to type IV collagen is dependent on the presence of Ca2+. This binding is saturable with a Kd approximately 1.2 x 10(-7) M based on solid phase binding and 4 x 10(-8) M based on the IC50 value from fluid phase binding data. Binding of SAP to type IV collagen was inhibited by both SAP and C-reactive protein (CRP). However, a 5-fold molar excess of CRP as compared with SAP was required to inhibit the SAP binding by 50%. Binding of SAP to type IV collagen was inhibited by both collagen IV and C1q but not by phosphatidylethanolamine or bovine serum albumin. The inhibition data indicate that SAP may bind to the triple helical region of type IV collagen via a site distinct from its galactan binding site. The most likely site of SAP involved in its interaction with type IV collagen may be the region spanning amino acid residues 108-120, which shows a great deal of sequence homology (60% strict identity) with the CRP region implicated in its binding to the collagen-like region of the C1q molecule.

Serum amyloid P (SAP) 1 is a 10-subunit (two flat pentameric discs stacked face to face) M r Ϸ 230,000 glycoprotein coded by a single gene on human and mouse chromosome 1 (1,2). It is a member of the evolutionarily conserved pentraxin family (1). These are proteins that are made up of five identical noncovalently bound subunits arranged in a flat pentameric disc (1,3). In vitro SAP binds to proteoglycans and fibronectin in a specific Ca 2ϩ -dependent manner (4 -6). Furthermore, SAP binds to the collagen-like region of complement component C1q (7). SAP is a normal component of a group of basement membranes including glomerular and alveolar basement membranes (8,9). It comprises approximately 10% of the protein released from glomerular basement membrane after collagenase treatment (8). Association of SAP with glomerular basement membranes is completely disrupted or disturbed in a number of nephritides such as Alport's syndrome (10), membranous glomerulonephritis, and membranoproliferative glomerulonephritis (11).
Basement membranes are multicomponent structures that perform a variety of functions. They are involved in maintenance of the differentiated state and basal and apical polarity of the cells as well as maintenance of the organ structure and filtration functions (12,13). A number of factors affect the structure and function of basement membranes (e.g. the relative concentration of each component and the affinity of the interaction between components as well as their structure) (14 -16). It is therefore possible that SAP, via its interaction with various components of the extracellular matrix, modifies the structure and function of the basement membranes with which it is associated. In the present study, the binding of SAP to type IV collagen was examined. The data indicate that SAP binding to type IV collagen is Ca 2ϩ -dependent, specific and saturable with a K d Ϸ 1.2 ϫ 10 Ϫ7 M for immobilized and a K d Ϸ 4 ϫ 10 Ϫ8 M (based on the IC 50 value of the soluble phase binding assay) for soluble type IV collagen. Furthermore, SAP probably binds via its C1q binding region to the triple helical region of type IV collagen. The interaction of SAP with type IV collagen and other components of the basement membrane may affect the structure of the basement membranes, thereby affecting their function.

EXPERIMENTAL PROCEDURES
Materials-Human SAP and CRP were purchased from Calbiochem. Each protein gave a single band of M r Ϸ 23,000 and 25,000 when size fractionated on SDS-polyacrylamide gel (12% polyacrylamide) under reducing conditions. Mouse SAP was purified from acute phase mouse serum (provided by R. F. Mortensen, Ohio State University) by Ca 2ϩdependent affinity chromatography on a column of phosphatidylethanolamine conjugated to agarose beads (Sigma) as described previously (17) followed by anion exchange chromatography on a Mono Q column (Pharmacia Biotech Inc.). The purified protein gave a single band of M r Ϸ 23,000 upon size fractionation on SDS-polyacrylamide gel electrophoresis (12% polyacrylamide) under reducing conditions. Human type I, II, and III collagens were obtained from Life Technologies, Inc., and Sigma. Mouse type IV collagen was from Life Technologies, Inc. and was isolated from Engelbreth-Holm-Swarm tumor cells grown subcutaneously in lathrytic mice. Size fractionation on SDS-polyacrylamide gel electrophoresis (7% polyacrylamide) gave two bands of M r Ϸ 170,000 and 190,000 and minor high molecular weight bands (M r of intact type IV collagen was estimated at Ϸ550,000). Type IV collagen extracted from human placenta after pepsin digestion was purchased from Sigma, and anti-human SAP and CRP antibodies were from Dako Corp. Monoclonal mouse anti-human SAP and phosphatidylethanolamine were from Sigma, and anti-mouse SAP antibody and human C1q were from Calbiochem.
125 I Labeling of SAP-Purified human SAP was iodinated by mixing 0.250 mCi of 125 I-labeled sodium iodide (Amersham Corp.) with 500 g of SAP in Tris-buffered saline (TBS, 20 mM Tris, 150 mM NaCl, 10 mM EDTA) in a glass tube coated with IODO-GEN reagent (Pierce). The * This study was supported by a Children's Hospital Research Foundation trustee grant. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
reaction was allowed to proceed for 7 min at 25°C. Unincorporated radioactivity was separated by desalting on a Microcon 50 microconcentrator (Amicon, Inc.). The remaining protein was diluted in TBS containing 1 mM EDTA. Radioactivity of the final protein preparation was 95-98% precipitable by trichloroacetic acid. The specific activity of 125 I-SAP was 0.1-0.3 Ci/pmol. Iodinated SAP gave a single band of M r Ϸ 23,000 on SDS-polyacrylamide gel electrophoresis and retained its Ca 2ϩ -dependent binding to phosphatidylethanolamine.
ELISA Binding Assay-Type IV collagen or other proteins (1 g/well) were coated overnight at 4°C onto microtiter plates (Corning Inc.) using carbonate buffer (45.3 mM NaHCO 3 and 18.2 mM Na 2 CO 3 , pH 9.6). The plates were washed with TBS washing buffer (20 mM Tris, 150 mM NaCl, 2 mM CaCl 2 , 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, pH 7.45) containing 10 mg/ml blocking reagent. Dilutions of SAP (100 l) in TBS dilution buffer (20 mM Tris, 150 mM NaCl, 2 mM CaCl 2 , 0.05% Tween 20, and 5 mg/ml blocking reagent, pH 7.45) were added to triplicate wells, and binding was allowed to proceed for 3 h at 37°C. Wells were washed 3 times with TBS washing buffer, and plates were incubated with the appropriate dilutions of primary antibody in TBS dilution buffer for 24 h at 4°C. Wells were washed 3 times in TBS washing buffer, and horseradish peroxidase-conjugated antibodies (Calbiochem) were diluted (1:2000) in TBS dilution buffer, added to each well, and allowed to bind for 90 min at 25°C. Plates were washed 3 times with TBS washing buffer, the substrate solution (10 g/ml O-phenylenediamine dihydrochloride in 50 mM citric acid, 100 mM Na 2 HPO 4 , 0.0003% H 2 O 2 , pH 5.0) was added to each well and the color reaction was allowed to develop at 25°C and then stopped by the addition of 9.6% H 2 SO 4 . Absorption by each well at 492 nm was determined using an ELISA plate reader (TiterTek). Inhibition of SAP binding to type IV collagen was determined by allowing 25 g/ml SAP to bind to immobilized type IV collagen in the presence of soluble type IV collagen and soluble C1q for 4 h at 37°C. Experiments determining the Ca 2ϩ dependence of the SAP interaction with type IV collagen were performed using the same conditions; however, the Ca 2ϩ levels for dilution and washing buffers were adjusted before the SAP binding step.
Solid Phase Binding Assay Using 125 I-SAP-Immulon I Removawells (Dynatech Laboratories, Inc.) were coated with 60 l of 50 g/ml mouse type IV collagen in phosphate-buffered saline. The amount of type IV collagen bound to the wells for each experiment was determined by directly measuring the protein bound in each well of a 12-well strip coated with 60 l of 50 g/ml type IV collagen in phosphate-buffered saline against known type IV collagen standard using the BCA protein quantitation assay (Pierce). Total bound collagen was determined to be approximately 780 Ϯ 80 ng/well. These results agree with preliminary experiments that used trace amounts of 125 I-collagen (20,000 cpm) mixed with 3 g of type IV collagen to determine the fraction of type IV collagen that was bound to each well. Results from these studies indicated that approximately 25-30% of 125 I-type IV collagen (830 Ϯ 110 ng/well) was bound to each well. Wells were washed 3 times with TBS washing buffer and blocked with TBS blocking buffer. Plates were then washed 3 times. Various dilutions of 125 I-SAP in TBS dilution buffer (100 l/well) were added to duplicate wells and incubated for 16 h at 4°C. To calculate the total amount of SAP added to each well, 2 l of each sample were counted using an Isoflex automatic ␥-counter. The results were converted to picomolar concentrations using the calculated specific activity of 125 I-SAP and an M r Ϸ 230,000 for the SAP decamer. After washing the wells 3 times with TBS washing buffer, the bound SAP was measured by counting the entire well. Specific binding was calculated by subtracting nonspecific binding (bound counts in the presence of 100-fold excess of unlabeled SAP in low concentration samples or in the presence of 10 mM EDTA in all samples) from total binding (binding in TBS dilution buffer with 2 mM CaCl 2 ). The results were converted to picomolar concentration using the calculated specific activity of 125 I-SAP. Reversible fluid phase binding of SAP to type IV collagen was examined as follows. Type IV collagen (0.5-450 g/ml) was incubated for 24 h at 4°C with 17 g/ml 125 I-SAP. The fluid phase binding samples were then added to microtiter wells coated with type IV collagen or bovine serum albumin (BSA) (100 l of sample/well). Plates were then incubated for 16 h at 4°C. Plates were washed 3 times with TBS containing 2 mM CaC1 2 and 0.5% Tween 20. Bound counts/ min were determined by counting individual wells in a ␥-counter. Inhibition of SAP binding by SAP and CRP were performed by determining the binding of 125 I-SAP to type IV collagen in the presence of increasing concentrations of SAP and CRP.

RESULTS
Binding of SAP to Immobilized Type IV Collagen-Binding of SAP to immobilized type IV collagen was examined. Human SAP binds to mouse type IV collagen. In a non-quantitative ELISA, maximal SAP binding was observed at approximately 20 g/ml SAP (data not shown). Binding of SAP (25 g/ml) to immobilized type IV collagen was measured as a function of time (5 min to 15 h) at 37 and 4°C. SAP binding to immobilized type IV collagen proceeded rapidly, reaching near maximum levels by approximately 3-4 h for samples incubated at 37°C and by 8 -9 h for samples incubated at 4°C (Fig. 1).
To characterize the binding of SAP to type IV collagen, binding of 125 I-SAP to type IV collagen was examined and subjected to Scatchard analysis. Serial dilutions of 125 I-SAP (0.2-132.25 g/ml or 0.1-57.5 pmol in 100 l) were added to immobilized type IV collagen. Binding of SAP to type IV collagen approached saturation when 15-20 pmol in 100 l (34.5-46 g/ml) of SAP were added to plates coated with approximately 1.3-1.5 pmol (720 -830 ng/well) of type IV collagen ( Fig. 2A). Scatchard analysis of the binding data indicated a K d Ϸ 1.2 ϫ 10 Ϫ7 M (Fig. 2B).
Binding of SAP to type IV collagen was further examined using a reversible solution phase binding assay. Binding of half-maximal saturation concentrations of SAP (17 g/ml as determined by quantitative binding assays) to increasing levels of soluble type IV collagen (0.5-450 g/ml) was carried out for 24 h at 4°C. Samples (100 l) were then added to microtiter wells coated with type IV collagen. Specific binding of SAP to immobilized type IV collagen was determined, and the percent inhibition of SAP binding was calculated (Fig. 3). The IC 50 of type IV collagen was shown to be approximately 20 g/ml (4 ϫ 10 Ϫ8 M). Scatchard transformation of the data from these experiments was not linear. This may be due to the polymerization of soluble type IV collagen at neutral pH or the binding of soluble SAP-type IV collagen complexes to immobilized collagen IV since the SAP decamer possesses multiple binding sites for type IV collagen.
Since the binding experiments above used heterologous sources of protein (i.e. human SAP and mouse type IV collagen), binding of both mouse and human SAP with mouse type IV collagen was compared using ELISA binding assays. Both human and mouse SAP bind to type IV collagen (Fig. 4), and similar saturation concentrations were observed for both proteins (Ϸ20 g/ml). Examination of the binding of human SAP to human and mouse type IV collagen (Fig. 5) as well as type I, II, and III collagens, mouse type IV collagen, and C1q (Fig. 6) indicated that SAP binds to human and mouse type IV colla- gens and C1q, but not type I, II, and III collagens. Furthermore, SAP does not bind to other proteins such as C1 inhibitor, BSA, and apoferritin (data not shown). The binding of SAP to human type IV collagen is lower than that of its binding to mouse type IV collagen. This may be due to proteolytic cleavage of human type IV collagen by pepsin during its extraction. The results indicate that binding of SAP to type IV collagen is saturable and of a relatively high affinity. Furthermore, these data indicate that this binding is specific for type IV collagen since SAP did not bind to other collagen molecules.
Effect of Ca 2ϩ on the Binding of SAP to Type IV Collagen-Binding of SAP to a variety of its ligands is dependent on the presence of Ca 2ϩ (4,18,19). In order to determine the effect of Ca 2ϩ concentration on the binding of SAP to immobilized type IV collagen, its binding in the presence of Ca 2ϩ (0.5-7 mM) and EDTA (1-10 mM) was examined in an ELISA binding assay (Fig. 7). Binding of SAP to type IV collagen was enhanced by 4 -40-fold in the presence of Ca 2ϩ as compared with its binding to type IV collagen in the absence of Ca 2ϩ . Enhanced binding was observed in the presence of 0.5 mM CaC1 2 , whereas binding levels in the presence of higher calcium concentration (1-7 mM) remained at similar levels throughout the experiments. Bind-ing levels of SAP at 5 mM Ca 2ϩ were more consistently lower in all experiments. However, this binding was not significantly different from other values obtained in the presence of Ca 2ϩ concentrations between 1 and 7 mM. As little as 1 mM EDTA diminished the binding of SAP to background levels. Binding of SAP to BSA was minimal. These data indicate that SAP binding to type IV collagen is dependent on the presence of Ca 2ϩ levels in the physiological range.
Inhibition of Binding of SAP to Immobilized Type IV Collagen by SAP and CRP-Human CRP is an acute phase reactant in which hepatic synthesis and plasma levels increase by up to 1000-fold in response to inflammation (20). It has a high degree of structural and amino acid sequence homology with SAP (1, 2, 21). Previous studies indicate that both SAP and CRP bind to C1q via its collagen-like region (7,22). Binding of CRP to C1q is mediated via amino acid residues 108 -120 (23). Comparison

FIG. 2. Binding of 125 I-SAP to immobilized type IV collagen (Co(IV)).
A, plates coated with type IV collagen were washed 3 times in TBS washing buffer. Dilutions of 125 I-SAP (0.1-57.5 pmol in 100 l of TBS dilution buffer) were added to each well, and binding was allowed to proceed for 16 h at 4°C. Samples were then removed, and wells were washed 3 times with TBS washing buffer, dried, and counted in a ␥-counter. The specific binding of SAP was calculated based on the specific activity of 125 I-SAP. B, Scatchard analysis of the binding data resulted in a K d Ϸ 1.2 ϫ 10 Ϫ7 M and an SAP to type IV collagen ratio of 1:8.8. Data represent the combined results from two independent experiments. Each data point is the mean value of duplicate samples.

FIG. 3. Binding of 125 I-SAP to type IV collagen (Co IV) in solution.
Microtiter plates were coated with type IV collagen or BSA. 125 I-SAP at 17 g/ml was incubated with increasing amounts of type IV collagen (0 -450 g/ml) for 24 h at 4°C. The samples were then transferred to the microtiter wells (100 l/well), and binding of SAP to coated wells was quantitated. The percent inhibition was determined by using the following formula: (100 Ϫ (binding in the presence of an inhibitor/ binding in the absence of any inhibitor) ϫ 100). The results represent the mean values of four independent experiments.

FIG. 4. Binding of human and mouse SAP to type IV collagen.
Various amounts of human (Ⅺ) and mouse (q) SAP (100 l/well) were added to wells coated with mouse type IV collagen. The extent of binding was determined using an ELISA binding assay. Data represent the mean of three independent experiments. of CRP, mouse, and human SAP primary structures indicates that there is a high degree of amino acid sequence identity in this area (23); therefore, SAP may bind to C1q via a region similar to the CRP C1q binding region. The ability of SAP and CRP to inhibit the binding of SAP to type IV collagen was examined (Fig. 8). BSA, at concentrations of up to 750 g/ml, could not significantly inhibit the binding of SAP to type IV collagen. Both SAP and CRP inhibited the binding of SAP to type IV collagen; however, an approximately 5-fold molar excess of CRP as compared with SAP was required for 50% inhibition of the binding of SAP to type IV collagen. The data indicate that SAP and CRP bind to the same region or to closely located regions of the type IV collagen molecule.
SAP Binding to Type IV Collagen Is Inhibited by Soluble Type IV Collagen and C1q but Not by Phosphatidylethanolamine-SAP binds to a number of substrates including C4 binding protein and ␤-amyloid fibrils via its galactan binding site (24,25). Binding of the SAP analog CRP to C1q is mediated via its amino acid residues 108 -120, which is distinct from the phosphorylcholine binding site of CRP (23,26). It is possible that similar residues in SAP (108 -120 region) are involved in its interaction with C1q and type IV collagen. The primary sequence of CRP in this region exhibits a high degree of amino acid identity (54 -67%) with human and mouse SAP (23), and the substitutions in both SAP molecules either do not affect the binding to C1q or enhance this binding in the context of the CRP molecule (23). In order to examine the role of the galactan binding site in type IV collagen binding, binding of SAP to type IV collagen in the presence of phosphatidylethanolamine, C1q, and type IV collagen was examined. Binding of SAP to type IV FIG. 5. Binding of human SAP to human and mouse type IV collagen. Various amounts of SAP (100 l/well) were allowed to bind to immobilized human (Ⅺ) and mouse (f) type IV collagen, and the extent of binding was determined. The data represent the mean of four independent experiments.
FIG. 6. Comparison of SAP binding to C1q and various isoforms of collagen. Plates were coated with type I (E), type II (ࡗ), type III (å), mouse type IV (f) collagens, and C1q (Ⅺ). Binding of increasing levels of SAP to immobilized ligands was examined by ELISA. The data represent the mean of two independent experiments.

FIG. 7. Effect of Ca 2؉ on the binding of SAP to type IV collagen.
Binding of SAP (100 l of 25 g/ml) to immobilized type IV collagen (solid bars) and BSA (open bars) was examined in the presence or absence of Ca 2ϩ . The binding assay was conducted as described under "Experimental Procedures" except that the dilution of SAP, the three washing steps before the addition of SAP, and two of the washing steps after the removal of SAP were performed using TBS buffers containing the appropriate levels of Ca 2ϩ or EDTA. The amounts of bound SAP were calculated based on a standard curve. The data are mean values from four independent experiments.
FIG. 8. Inhibition of SAP binding to type IV collagen by SAP and CRP. Inhibition of 125 I-SAP (100 l of 25 g/ml) to immobilized type IV collagen by CRP (Ⅺ) and SAP (f) was examined. Labeled SAP and appropriate levels of the inhibitor (100 l/well) were added to type IV collagen-coated wells, and binding was allowed to proceed for 16 h at 4°C. Wells were washed 3 times, allowed to dry and counted in a ␥-counter. The percent inhibition was determined using the formula in the legend for Fig. 3. The results represent the mean values from four independent experiments. collagen is inhibited by C1q and type IV collagen but not by phosphatidylethanolamine or BSA (Fig. 9). Comparison of the IC 50 of C1q (Ϸ75 g/ml) with type IV collagen (Ϸ40 g/ml) indicates that an ϳ2-fold molar excess of C1q as compared with type IV collagen is required to inhibit the binding of SAP to type IV collagen by 50%. The data indicate that the galactan binding site of SAP is not involved in its binding to type IV collagen. Binding of the SAP analog CRP to the collagen-like region of C1q is mediated via amino acid residues 108 -120. This region is conserved in a number of pentraxins (23). It is therefore possible that this region is important in the interaction of SAP with type IV collagen. Studies are under way to further characterize the collagen binding sites of the SAP molecule. Furthermore, SAP most probably binds to the triple helical region of type IV collagen; however, the triple helical conformation by itself is not sufficient for binding of SAP to type IV collagen. DISCUSSION SAP is a component of a number of basement membranes including alveolar, glomerular, and sweat gland basement membranes (8,9). The disposition of SAP in the glomerular basement membrane is disrupted or altered in a number of nephritides such as Alport's syndrome (10), membranous glomerulonephritis, and membranoproliferative glomerulonephritis (11). Previous studies have shown that SAP binds to fibronectin and proteoglycans such as heparan and dermatan sulfates in vitro (4 -6). SAP also binds to C1q via its collagenlike region in vitro (7). Based on the in vivo association of SAP with specific basement membranes, its in vitro binding to proteoglycans, and its interaction with the collagen-like region of the C1q molecule, experiments were conducted to characterize the interaction of SAP with the basement membrane-derived type IV collagen. The data indicate that SAP binding to type IV collagen is Ca 2ϩ -dependent and saturable with a K d Ϸ 1.2 ϫ 10 Ϫ7 M in the solid phase and 4 ϫ 10 Ϫ8 M (based on the IC 50 value obtained in reversible soluble phase binding assay) in the soluble phase. These differences in binding may be due to different conformations assumed by immobilized versus fluid phase collagen. They may also reflect complex molecular interactions in liquid and solid phase as indicated by non-linear Scatchard data. These interactions may include binding of SAP to immobilized collagen or to soluble type IV collagen molecules that may be followed by either binding of the SAP component of the SAP-type IV collagen complex to immobilized type IV collagen or binding of the type IV collagen component of the complex to immobilized collagen. Inhibition studies indicate that SAP binds to the triple helical region of type IV collagen, probably via the region involved in its binding to C1q (amino acid residues 108 -120 based on studies examining the CRP region involved in its binding to C1q). Furthermore, the triple helical region motif (Gly-X-TYr) is not the only requirement for this binding since SAP does not bind to collagens I, II, and III.
The significance of the binding of SAP to type IV collagen and other components of the basement membrane is not yet understood. The ability of SAP to bind to type IV collagen as well as other ECM components suggests that it may facilitate the formation of the matrix or stabilize this matrix by bridging its components. Type IV collagen binds to a variety of other molecules via its triple helical region (30). Previous studies indicate that type IV collagen polymerizes to form a lattice via several interactions. These include head to head interactions of non-collagenous domains with one another (dimerization) as well as interaction of the amino-terminal 7 S domains (tetramerization) (27). Self-assembly of type IV collagen can also proceed laterally via its triple helical regions or via binding of the non-collagenous domain to the triple helical region of another molecule to form multimers (27,29). Studies examining the interaction of laminin and heparin with type IV collagen demonstrated that both molecules show significant binding to the triple helical region of type IV collagen (28,30). Since SAP binding to type IV collagen is most likely via type IV collagen's triple helical region, it may modify the assembly of type IV collagen or its interaction with proteoglycans and laminin and alter the structure of the basement membrane and its function.
Crystallographic studies of the SAP molecule indicate that it is a very compact and highly structured protein (31,32). Probably due to its highly ordered and compact structure, the SAP molecule is resistant to proteolysis (31,32). Previous studies have shown that SAP binds to a number of proteins and protects them from proteolytic digestion (25). For example, SAP binds to Alzheimer's ␤-amyloid peptide and amyloid A fibrils and protects them from proteolysis by enzymes such as trypsin, chymotrypsin, and Pronase (25). These observations suggest that SAP may act by binding to components of the ECM and protect them from digestion by extracellular matrix proteases and by altering the turnover rate and the structure of the basement membrane, thereby affecting its function.
Although the role of SAP in the structure and function of the basement membranes with which it is associated is not clear, it has been shown that in a variety of nephritides its association with the glomerular basement membrane is altered. It is possible that SAP via its interaction with type IV collagen and other ECM components may alter the assembly and function of the basement membrane. The absence of SAP or its altered disposition in the basement membrane, probably as a secondary defect to the initial injury, could lead to further changes in basement membrane assembly or turnover. This may partially account for basement membrane abnormalities observed in certain nephritides. The role of SAP in basement membrane structure and function is not clear; however, previous studies and the data presented here indicate that further studies of the effect of SAP on the assembly of ECM components and of the FIG. 9. Inhibition of SAP binding to immobilized type IV collagen by C1q, type IV collagen, and phosphatidylethanolamine. Inhibition of SAP binding to immobilized type IV collagen by BSA (E), C1q (ࡗ), type IV collagen (f), and phosphatidylethanolamine (q) was examined. Various concentrations of each inhibitor were mixed with 2.5 g of SAP (100 l final volume) and incubated for 30 min at 37°C. The mixture was added to immobilized type IV collagen. The percent inhibition was determined using the formula outlined in the legend for Fig.  3. Data represent mean values from three independent experiments.