Human protease nexin-I. Further characterization using a highly specific polyclonal antibody

We have used purified protease nexin-I (PN-I) from human fibroblasts to develop a polyclonal antibody that specifically blocks the PN-I-mediated cellular binding of thrombin and urokinase. Anti-PN-I IgG did not inhibit the binding of 125I-epidermal growth factor-binding protein to fibroblasts, which is mediated by protease nexin-II, another cell-secreted, serine protease inhibitor that is distinct from PN-I. This furthers the belief that the protease nexins are distinct from one another. In addition, while anti-PN-I IgG immunoprecipitated PN-I X thrombin complexes, it did not do so with antithrombin-III X thrombin. Metabolically labeled PN-I was also immunoprecipitated by IgG, indicating that the protein can be labeled in vivo. The antibody also recognized primarily one band on Western transfers of conditioned medium from fibroblast cultures. These results suggest that anti-PN-I will be useful in probing the physiological role of PN-I as well as its biosynthesis.     

Characterization of the receptor for protease nexin-I:protease complexes on human fibroblasts

Fibroblasts as well as several other cell types, secrete a number of protease inhibitors into their culture media. Among these inhibitors are the protease nexins, a class of proteins which covalently bind serine proteases, thereby inactivating their specific targets. Protease nexin-I, first discovered in human foreskin fibroblasts, binds thrombin, plasmin, and urokinase with high affinity, forming covalently linked complexes. Human fibroblasts bind complexes of protease nexin-I and its target protease via a cell-surface, high-affinity receptor. We have analyzed a number of characteristics of this receptor, and found them to be typical of class II receptors in general. At 4 degrees C binding of PN-I:protease complexes was competed by heparin. In addition, binding was independent of the particular protease bound to the PN-I; purified complexes of PN-I with thrombin or urokinase competed equipotently for [125]I-thrombin:PN-I binding. As the pH of the binding buffer was lowered, binding to cells increased. A twofold increase in binding was attained by lowering the pH from 7.5 to 4.5. This phenomenon was not due to irreversible, pH-induced changes to either the cell surface or the labeled complexes. At 37 degrees C, the removal of labeled complexes from culture medium was rapid; approximately 80% was removed by 4 hours under given conditions. The internalization of complexes was also very rapid, with an estimated ke (endocytic rate constant) of 1.0 min-1. At neutral pH, fibroblasts bind complexes in a saturable manner. Scatchard analysis yields a receptor number of 250,000 per cell and a Kd of 1 nM.     

The glioma cell-derived neurite promoting activity protein is functionally and immunologically related to human protease nexin-I

Protease nexin-I (PN-I, Mr approximately 43,000) is representative of a newly described class of cell-secreted protease inhibitors. PN-I has been purified to apparent homogeneity, partially sequenced, and monospecific antibodies have been raised against it. PN-I is a potent inhibitor of urokinase, thrombin, plasmin, and trypsin. In addition, cells have specific receptors that mediate the uptake of covalently linked complexes formed between PN-I and its protease substrates. In the present studies, we have investigated the relationship between human PN-I and a protease inhibitor derived from C6 glioma cells in culture that has neurite-promoting activity. On the basis of co-purification on heparin-Sepharose, identical molecular weight, antibody cross-reactivity, and receptor cross-reactivity, we conclude that PN-I and the glioma-cell-derived inhibitor are equivalent molecules.     

Biosynthesis of the adenosine deaminase-binding protein in human fibroblasts and hepatoma cells

The adenosine deaminase-binding protein has previously been localized to the cell surface of human fibroblasts (Andy, R. J., and Kornfeld, R. (1982) J. Biol. Chem. 257, 7922-7925). In this study we examine the biosynthesis of binding protein in human fibroblasts, human hepatoma HepG2 cells, and a human kidney tumor cell line. Binding protein immunoprecipitated from radioiodinated detergent-extracted fibroblast membranes has a molecular weight of 120,000 when analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. An additional band of Mr 100,000 is also present which we believe is a result of proteolysis of the 120,000 band. Purified soluble kidney binding protein has an Mr of 112,000. Binding protein from fibroblasts pulse-labeled with [35S]methionine for 15 min migrates as a 110-kDa band on sodium dodecyl sulfate-polyacrylamide gels. Within 30-60 min of chase, the intensity of the 110-kDa band is diminished, and a 120-kDa band has appeared. Binding protein reaches the cell surface of fibroblasts within 30-60 min of chase. The same results are obtained with the other cell lines studied. Thus, binding protein is initially synthesized as a precursor of 110 kDa which chases into a 120-kDa mature form. The shift of 10 kDa is probably due to processing of its oligosaccharide chains since soluble kidney-binding protein contains 7-9 complex N-linked chains. Upon endoglycosidase H treatment, the 110,000 precursor shifts to a Mr of 89,000 while the 120,000 mature band shifts to 115,000, consistent with the presence of 7-9 high mannose chains on the precursor and 1-2 high mannose chains on the mature form. These results and the presence of complex N-linked chains on binding protein were confirmed by lectin affinity chromatography of glycopeptides derived from [2-3H]mannose-labeled binding protein. Analysis of [6-3H]glucosamine-labeled binding protein indicates the presence of 1 sialic acid residue per chain.     

Interactions of serine proteases with cultured fibroblasts

This review summarizes the mechanisms by which several serine proteases, particularly urokinase, thrombin, and elastase, interact with cultured fibroblasts. Many of these studies were prompted by findings that interactions of these proteases with cells and the extracellular matrix are important in a number of physiologic and pathologic processes. Two main pathways have been identified for specific interactions of these proteases with fibroblasts. One involves surface binding sites for the free protease that appear to bind only one particular protease. An unusual feature collectively shared by the binding sites for urokinase, thrombin, and elastase is that the bound protease is not detectably internalized by the fibroblasts. The other pathway by which serine proteases interact with fibroblasts involves proteins named protease nexins (PNs). Three PNs have been identified. They are secreted by fibroblasts and inhibit certain serine proteases by forming a covalent complex with the protease catalytic site serine. The complexes then bind back to the fibroblasts via the PN portion of the complex and are internalized and degraded. Recent studies showing that the fibroblast surface and extracellular matrix accelerate the inactivation of thrombin by PN-1 support the hypothesis that the PNs control protease activity at and near the cell surface. The PNs differ from plasma protease inhibitors in their molecular properties, absence in plasma, site of synthesis, and site of clearance of the inhibitor:protease complexes.     

Kinetics of the formation of thrombin-thrombospondin complexes: involvement of a 77-kDa intermediate

Thrombin forms sodium dodecyl sulfate stable complexes of 77 and greater than 450 kDa with proteins secreted by activated platelets. The kinetics of formation of these complexes were investigated by addition of 125I-thrombin to the supernatant solution of A23187-activated platelets. Complexes were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis either with or without reduction of disulfide bonds. When analyzed on nonreduced gels, the 77-kDa complex reached a maximum at about 3 min and then declined as the greater than 450-kDa complex increased. On reduced gels (on which there was no greater than 450-kDa complex) the 77-kDa complex approached the level of the greater than 450-kDa complex on nonreduced gels. The half-time of formation was less than 1 min for the 77-kDa complex and about 15 min for the greater than 450-kDa complex. These time courses suggested that the 77-kDa complex was incorporated into the greater than 450-kDa complex as an essential precursor. Formation of complexes was inhibited by a competitive inhibitor or a noncompetitive inhibitor of thrombin, and the pH dependence of formation of both complexes was similar to the pH dependence for catalytic activity of thrombin. Ca2+ inhibited formation of the greater than 450-kDa complex but not of the 77-kDa complex. A model is presented in which thrombin and a secreted protein form a 77-kDa complex by a process that involves the active site of thrombin. The 77-kDa complex is then incorporated into a greater than 450-kDa complex by thiol-disulfide exchange with thrombospondin, a process that is inhibited by Ca2+. Thrombin in the greater than 450-kDa complex had no catalytic activity.     

Covalent binding of human thrombin to a human endothelial cell-associated protein

Binding of 125I-thrombin to endothelial cells derived from human umbilical vein was studied in tissue culture. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography revealed covalent binding of thrombin in a 72-kDa complex. This binding is specific and requires the catalytically active site of the enzyme. Formation of the complex could be detected as early as 3 min after addition of thrombin or with a thrombin concentration as low as 0.5 nM. This irreversible binding exhibits thrombin dose-dependence and reaches maximum levels at a concentration of 50 nM (10 fmol/10(5) cells). Some characteristics of the 72-kDa complex were compared to those of the complexes formed between thrombin and protease nexin originating from fibroblasts or platelets: (i) its electrophoretic mobility on SDS-PAGE is identical to that of the thrombin-platelet protease nexin complex, (ii) heparin prevents the appearance of the complex on the cell surface, (iii) plasmin in a 100-fold molar excess prevents the covalent linkage of thrombin, suggesting that the protease specificity of the endothelial component involved in the complex might not be restricted to thrombin. Yet no release, nor any secretion of the endothelial protein, could be detected. These results indicate that active thrombin binds covalently to a specific endothelial protein that is in several respects similar to fibroblast or platelet protease nexin and provides a thrombin binding site distinct from thrombomodulin and glycosaminoglycans.     

Binding of protease nexin-1 to the fibroblast surface alters its target proteinase specificity

Protease nexin-1 is a protein proteinase inhibitor that is secreted by a variety of cultured cells and rapidly forms complexes with thrombin, urokinase, and plasmin; the complexes then bind back to the cells and are internalized and degraded. In fibroblast cultures, protease nexin-1 is localized to the extracellular matrix. Here we report that protease nexin-1, which is bound to the surface of fibroblasts, forms complexes with thrombin, but not urokinase or plasmin. Experiments were conducted to determine directly if protease nexin-1 binding to the fibroblast surface alters its proteinase specificity. To do this, cell surface protease nexin-1 was inhibited using anti-protease nexin-1 monoclonal antibodies that stoichiometrically block its ability to form complexes with target proteinases. Then, purified protease nexin-1 was added to these cells; the cell-bound molecule formed complexes with thrombin, but not urokinase or plasmin. Similar experiments showed that protease nexin-1 bound to preparations of fibroblast extracellular matrix also formed complexes with thrombin, but not urokinase or plasmin. Components of the extracellular matrix other than heparin-like glycosaminoglycans are required for this regulation since heparin did not block the formation of complexes between protease nexin-1 and urokinase or plasmin. These results suggest that protease nexin-1 is primarily a thrombin inhibitor in interstitial fluids where much of it would be bound to cell surfaces.     

Physiological properties and differential glycosylation of phosphorylated and nonphosphorylated forms of osteopontin secreted by normal rat kidney cells

In a previous study we have shown that normal rat kidney (NRK) cells in vitro secrete a 69-kDa osteopontin in both phosphorylated (pp69) and nonphosphorylated (np69) forms. Only pp69 interacts with the cell surface and np69 forms a heat-dissociable complex with plasma fibronectin, suggesting functional modulation of osteopontin by phosphorylation. Using tunicamycin, an inhibitor of N-linked glycosylation, and peptide:N-glycosidase F, which removes N-linked oligosaccharide chains from glycoproteins, we show here that np69, but not pp69, contains N-linked carbohydrates. Our results also demonstrate that tunicamycin treatment does not inhibit the cell surface binding of pp69; however, np69 secreted by the treated cells fails to complex with plasma fibronectin, suggesting importantly, our data show that pp69 forms a heat-stable complex with cell surface fibronectin, suggesting that it is an integral component of the extracellular matrix of NRK cells. Finally, sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of deglycosylated and in vitro translated osteopontin suggests that the acidic nature of osteopontin as well as its post-translational modifications play a role in the anomalous behavior of osteopontin in sodium dodecyl sulfate gels, observed in several laboratories. The data presented here provide evidence for possible functional roles of 69-kDa osteopontin and suggest that its physiological properties are regulated by post-translational modifications.     

Protease nexin. Properties and a modified purification procedure

The present paper describes chemical and functional properties of protease nexin, a serine protease inhibitor released from cultured human fibroblasts. It is shown that protease nexin is actually synthesized by fibroblasts and represents about 1% of their secreted protein. Analysis of the amino acid composition of purified protease nexin indicates that it is evolutionarily related to antithrombin III and heparin cofactor II. Protease nexin contains approximately 6% carbohydrate, with 2.3% amino sugar, 1.1% neutral sugar, and 3.0% sialic acid. The Mr calculated from equilibrium sedimentation analysis is 43,000. Protease nexin is a broad specificity inhibitor of trypsin-like serine proteases. It reacts rapidly with trypsin (kassoc = 4.2 +/- 0.4 X 10(6) M-1 s-1), thrombin (kassoc = 6.0 +/- 1.3 X 10(5) M-1 s-1), urokinase (kassoc = 1.5 +/- 0.1 X 10(5) M-1 s-1), and plasmin (kassoc = 1.3 +/- 0.1 X 10(5) M-1 s-1), and slowly inhibits Factor Xa and the gamma subunit of nerve growth factor but does not inhibit chymotrypsin-like proteases or leukocyte elastase. In the presence of heparin, protease nexin inhibits thrombin at a nearly diffusion-controlled rate. Two heparin affinity classes of protease nexin can be detected. The present characterization pertains to the fraction of protease nexin having the higher affinity for heparin. The low affinity material, which is the minor fraction, is lost during purification.     

Vasoactive intestinal peptide receptor on liver plasma membranes: characterization as a glycoprotein

The receptor for vasoactive intestinal peptide (VIP) was identified in rat liver plasma membranes after covalent cross-linking to 125I-VIP by three different agents [disuccinimido dithiobis(propionate), disuccinimido suberate, and succinimido 4-azidobenzoate] and examined by sodium dodecyl sulfate-acrylamide electrophoresis. Regardless of the presence of reducing conditions, two molecular species of the putative VIP binding unit were identified as broad autoradiographic bands of 80,000 and 56,000 daltons (Da). Both the large and small species showed the same high affinity for 125I-VIP binding and subsequent cross-linking (half-maximal inhibition at 3 nM unlabeled VIP). The 80-kDa species was partially converted to the 56-kDa form by denaturing conditions and was extensively degraded when incubated at 20 degrees C for 30 min with 1 microgram/mL chymotrypsin, trypsin, or elastase to fragments that that migrated similarly to the 56-kDa unit. In contrast, the 56-kDa moiety was resistant to attack by serine proteases. Both the 80- and 56-kDa species were microheterogeneous due at least in part to the presence of carbohydrate chains, each species binding fractionally to wheat germ agglutinin (WGA)-agarose (approximately 50%). The WGA-bound fraction (eluted with N-acetylglucosamine) was relatively retarded on acrylamide gels as compared to the WGA-unbound fraction. Exposure of the 80- and 56-kDa species to endo-beta-acetylglucosaminidase F reduced the apparent molecular mass of each by 19 kDa, indicating the presence of complex N-linked carbohydrate chains. The receptor species do not appear to have high-mannose N-linked chains since they did not interact with concanavalin A and were not cleaved by endo-beta-acetylglucosaminidase H.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization of an endoserine protease secreted by Arthrobacter aureus.

A 22-kDa endoserine protease secreted by an Arthrobacter aureus strain was identified. The optimal temperature for this protease was 70 degrees C. Protease production was maximal at the end of the exponential growth phase, and the protease was induced by amino acids and peptides and repressed by carbon sources.     

A 58-kDa resident protein of the cis Golgi cisterna is not terminally glycosylated

A 58-kDa Golgi protein (gp58) was previously identified and found to be concentrated in cis Golgi cisternae in several cell types (Saraste, J., Palade, G.E., and Farquhar, M.G. (1987) J. Cell Biol. 105, 2021-2029). In this study the protein was partially purified from rat pancreas and mouse myeloma cells in order to characterize its oligosaccharides. It migrated on sodium dodecyl sulfate-polyacrylamide gels as a 57-58-kDa doublet under reducing conditions or as a single approximately 116-kDa band under nonreducing conditions. Pancreatic gp58 was susceptible to alpha-N-acetylgalactosaminidase digestion and it bound concanavalin A, Helix pomatia, Dolichos biflorus, soybean agglutinin, and Bauhinia purpurea lectins, but not Ricinus communis agglutinin or lectins from Griffonia simplicifolia-1, Arachis hypogaea, and Limulus polyphemus. It bound Ricinus communis agglutinin after galactosylation with GlcNAc galactosyltransferase. These data demonstrate that pancreatic p58 contains immature N-linked moieties with nonreducing terminal GlcNAc residues as well as the initiating GalNAc of O-linked glycoproteins. Myeloma gp58 was sensitive to endo-beta-N-acetylglucosaminidase H, and oligosaccharide analysis of its [3H]glucosamine-labeled glycopeptides indicated that it also contained immature N-linked glycans. Some of the latter consist of high mannose chains (high affinity for concanavalin A, endo-beta-N-acetylglucosaminidase H-sensitive), but the predominant (95%) species are neutral tri- or tetraantennary N-linked chains containing GlcNAc (no binding to concanavalin A). Glycopeptides from biosynthetically labeled myeloma cells did not contain detectable base labile oligosaccharides, indicating that unlike pancreatic p58, myeloma gp58 may not be an O-linked glycoprotein. Neither pancreatic nor myeloma gp58 contained terminally processed oligosaccharides, indicating that gp58 has not been modified by trans-Golgi glycosyltransferases. Thus, the oligosaccharide content of gp58 is consistent with the assumption that this protein is retained in the cis Golgi cisternae during biosynthesis instead of being transported across the Golgi stacks and targeted back to the cis Golgi from the trans side.     

Protease secretion by Erwinia chrysanthemi. Proteases B and C are synthesized and secreted as zymogens without a signal peptide

The gene encoding the secreted 53-kDa metalloprotease (protease B) and the 5' end of the gene encoding the secreted 55-kDa metalloprotease (protease C) of the Gram-negative bacterium Erwinia chrysanthemi have been sequenced. The predicted sequences of the two proteases do not have typical signal sequences at their NH2 termini. Both proteases are synthesized as inactive higher molecular weight precursors (zymogens proB and proC) which are secreted into the external medium where divalent cation-mediated activation occurs. The activation of proB occurs with a t1/2 of less than 5 min at 37 degrees C in Luria broth medium, whereas that of proC occurs with a t1/2 of about 150 min. The NH2 termini of purified proteases B, proB, and C were sequenced. ProB starts at the initiator methionine whereas B and C start, respectively, at residues +16 and +18 of the sequence deduced from the nucleotide sequence. A short NH2-terminal extension is therefore removed during the activation process, most likely by an autocatalytic mechanism. Protease B shows a high degree of sequence homology with the secreted 50-kDa metalloprotease of Serratia marcescens, which also lacks a signal peptide and for which an inactive higher molecular weight form has been reported.     

A recombinant group 1 house dust mite allergen, rDer f 1, with biological activities similar to those of the native allergen

Serum IgE directed against Der f 1, a protease found in the feces of Dermatophagoides farinae, correlates well with allergic sensitization to house dust mite in humans and is a risk factor for developing asthma. Native Der f 1 (nDer f 1) is produced as a pre-pro form and processed to an approximately 25-kDa mature form. We have expressed recombinant forms of Der f 1 (rDer f 1) in Pichia pastoris using AOX1-promoter expression vectors. Fusion of either the pro-enzyme form or the mature form to the Saccharomyces cerevisiae alpha factor pre-pro sequence resulted in secretion of the mature form of the protein from P. pastoris. The secreted protein was heterogeneously glycosylated at a single N-glycosylation site and had an apparent molecular mass of 35-50 kDa. Both the alpha factor signal peptide and the pro-enzyme region were efficiently processed during secretion. A version of the pro-enzyme with a mutated consensus N-linked glycosylation site was secreted from P. pastoris as a mature, unglycosylated, approximately 25-kDa protein. The IgE binding activity of this unglycosylated rDer f 1 was similar to that of glycosylated forms produced by P. pastoris and to nDer f 1 obtained from mites. Thus, oligosaccharides are not required for secretion from P. pastoris or for IgE binding in vitro. Recombinant and native versions of Der f 1 displayed protease activity on casein zymogram gels. The availability of a highly purified recombinant Der f 1 will facilitate experimental and clinical studies of mite allergy.     

The purification of protease IV of E.coli and the demonstration that it is an endoproteolytic enzyme

A strong proteolytic activity is unmasked and solubilized when $\text{E. coli}$ outer membrane fragments are preincubated with 0.083% sodium dodecyl sulfate. This proteolytic activity cleaves αS1 casein into the same degradation products as protease IV, a recently described protease of $\text{E.coli}$ located in the outer membrane (Ph. Régnier, preceding paper), it is concluded that sodium dodecyl sulfate solubilizes the same protease. Protease IV has been purified 11,200 fold, probably to homogenetiy, by sodium dodecyl sulfate polyacrylamide gel electrophoresis followed by elution of the protein from gel slices. The purified enzyme is fully active, its molecular weight, determined from its migration in denaturating gels is 23,500. αS1 casein is cleaved by protease IV into two large polypeptides which are not further degraded and some small peptides of about 5,000 daltons. The production of discrete polypeptide species suggests that protease IV is an endoproteolytic enzyme.     

Protease-nexin I as an androgen-dependent secretory product of the murine seminal vesicle.

A search for inhibitors of urokinase-type plasminogen activator (uPA) in the male and female murine genital tracts revealed high levels of a uPA ligand in the seminal vesicle. This ligand is functionally, biochemically and immunologically indistinguishable from protease-nexin I (PN-I), a serpin ligand of thrombin and uPA previously detected only in mesenchymal cells and astrocytes. A survey of murine tissues indicates that PN-I mRNA is most abundant in seminal vesicles, where it represents 0.2-0.4% of the mRNAs. PN-I is synthesized in the epithelium of the seminal vesicle, as determined by in situ hybridization, and is secreted in the lumen of the gland. PN-I levels are much lower in immature animals, and strongly decreased upon castration. Testosterone treatment of castrated males rapidly restores PN-I mRNA levels, indicating that PN-I gene expression is under androgen control.     

Alternative model for the internal structure of laminin

A monoclonal antibody to laminin, LMN-1, was generated by immunizing rats with laminin from the EHS tumor and fusing the rat spleen cells with mouse NS-1 myeloma cells. Laminin fragments were generated by proteolytic digestion with thrombin, thermolysin, and chymotrypsin. Monoclonal antibody binding fragments were identified by immunoblotting. Fragments which bound monoclonal antibody LMN-1 included a 440-kilodalton (kDa) chymotrypsin fragment and thermolysin fragments of 440 and 110 kDa. These fragments could also be generated from within a 600-kDa thrombin fragment. Digestion of the 440-kDa chymotrypsin fragment with thermolysin generated the 110-kDa antibody binding fragment and a 330-kDa nonbinding fragment. Immunoblotting was performed on extracts of PYS-2 cells and EHS cells using polyclonal and monoclonal antibodies to laminin. Polyclonal antibodies stained the intact 850-kDa complex and the 200- and 400-kDa subunits, while monoclonal LMN-1 stained only the 400-kDa subunit and the complete molecule. Rotary shadowing of monoclonal LMN-1 bound to laminin molecules indicated that the binding site was within the long arm of laminin. Changes in the model of the internal organization of the laminin molecule are proposed, based on the binding of LMN-1 to the 400-kDa subunit and specific proteolytic fragments. The locations of the major thrombin and chymotrypsin fragments in the model are rotated 180 degrees relative to the previously described model [Ott, U., Odermatt, E., Engel, J., Furthmayr, H., & Timpl, R. (1982) Eur. J. Biochem. 123, 63-72] to include part of the 400-kDa subunit of laminin.     

Functional and structural similarities between protease nexin I and C1 inhibitor

Protease nexin I is a proteinase inhibitor that is secreted by human fibroblasts and forms stable complexes with certain serine proteinases; the complexes then bind to the fibroblasts and are rapidly internalized and degraded. In this report, we show that this inhibitor, which is present in very low concentrations in plasma, has functional and structural similarities to C1 inhibitor, an abundant proteinase inhibitor in plasma. Both inhibitors complex and inactivate certain proteinases that previously were known to rapidly react only with C1 inhibitor. Kinetic inhibition studies show that protease nexin I inhibits Factor XIIa and plasma kallikrein with second-order rate constants of 2.3 x 10(3) and 2.5 x 10(5) M-1 s-1, respectively, which are similar to the rate constants for inhibition of these proteinases by C1 inhibitor. Protease nexin I inhibits C1s about one-tenth as rapidly as does C1 inhibitor. Alignment of the amino acid sequences of protease nexin I and C1 inhibitor shows that these proteins have similarity at their reactive centers (from sites P7 to P1). The remaining regions of the two proteins share much less similarity. In contrast to protease nexin I, C1 inhibitor is not secreted by human fibroblasts. Although 125I-C1s-protease nexin I complexes readily bind to human fibroblasts, binding of 125I-C1s-C1 inhibitor complexes or other 125I-proteinase-C1-inhibitor complexes to these cells is not detectable. Thus, protease nexin I and C1 inhibitor may control some common regulatory proteinases in the extravascular and vascular compartments, respectively.