Cellular internalization and degradation of antithrombin III-thrombin, heparin cofactor II-thrombin, and alpha 1-antitrypsin-trypsin complexes is mediated by the low density lipoprotein receptor-related protein.

The inhibition of proteinase activity by members of the serine proteinase inhibitor (serpin) family is a critical regulatory mechanism for a variety of biological processes. Once formed, the serpin enzyme complexes (SECs) are removed from the circulation by a hepatic receptor. The present study suggests that this receptor is very likely the low density lipoprotein receptor-related protein (LRP), a prominent liver receptor. In vitro binding studies revealed that antithrombin III (ATIII)-thrombin, heparin cofactor II (HCII)-thrombin, and alpha1-antitrypsin (alpha1AT)-trypsin bound to purified LRP, and their binding was inhibited by the 39-kDa receptor-associated protein (RAP), an antagonist of LRP-ligand binding activity. In contrast, native or modified forms of the inhibitors were unable to bind to LRP. Mouse embryonic fibroblasts, which express LRP, mediate the cellular internalization leading to degradation of these SECs, while mouse fibroblasts genetically deficient in LRP showed no capacity to internalize and degrade these complexes. SECs were also degraded by HepG2 cells, and this process was inhibited by LRP antibodies, RAP, and chloroquine. The cellular-mediated uptake and degradation was specific for SECs; native or modified forms of the inhibitors were not internalized and degraded. Finally, in vivo clearance studies in rats demonstrated that RAP inhibited the clearance of ATIII-125I-thrombin complexes from the circulation. Together, these results indicate that LRP functions as a liver receptor responsible for the plasma clearance of SECs.

; however, some members such as angiotensinogen and ovalbumin are unable to inhibit proteinases (Doolittle, 1983;Hunt and Dayhoff, 1980). The inhibitory serpins are found in three distinct structural conformations; native (active inhibitor), cleaved (an inactive inhibitor generated after cleavage of its reactive center loop), or complexed with a proteinase. The inhibitory action of serpins affect a broad range of biological processes including coagulation, fibrinolysis, inflammation, connective tissue turnover, complement activation, and phagocytosis (for reviews, see Travis and Salvesen (1983), Bjork and Danielsson (1986), and Carrell and Travis (1985)).
␣ 1 -Antitrypsin (␣ 1 AT, also referred to as ␣ 1 -proteinase inhibitor) is the prototypic member of the serpin family and primarily acts to inhibit neutrophil elastase and proteinase 3 (PR3) (Travis and Salvesen, 1982;Carrell and Travis, 1985;Rao et al., 1991). Neutrophil elastase and PR3 degrade connective tissue matrix proteins, and both enzymes can induce emphysema in animal models (Janoff et al., 1977;Senior et al., 1977;Kao et al., 1988). ␣ 1 AT is present in high concentration in the plasma (Ͼ25 M) and protects the connective tissue-rich lung against excessive proteolysis through its inhibitory action on neutrophil elastase and PR3. The protection of lung tissue by ␣ 1 AT is evidenced by studies that have genetically linked decreased functional levels of ␣ 1 AT with the early onset of pulmonary emphysema (Laurell and Eriksson, 1963;Travis and Salvesen, 1983;Carrell, 1986).
Antithrombin III (ATIII), a serpin with extensive sequence similarity to ␣ 1 AT, plays a major role in hemostasis by acting as the principle physiological inhibitor of thrombin and coagulation Factor X a (for reviews see Bjork and Danielsson (1986), Bjork et al. (1989), and Olson and Bjork (1992)). This inhibitor is present in high concentration in the plasma (5 M), and its importance is underscored by the observation that reduced functional levels of ATIII are associated with increased risks of thromboembolic disorders (Edeburg, 1965;Lane et al., 1991).
Heparin cofactor II (HCII) is a 65,000-dalton inhibitor of thrombin, cathepsin G, and chymotrypsin (Tollefsen et al., 1982;Church and Hoffman, 1994;Church et al., 1995) and is present in the human plasma at a concentration of 1.4 M. Like ATIII, its ability to react with target proteinases is greatly accelerated in the presence of heparin. Individuals exhibiting low levels of HCII but having normal levels of ATIII display thrombotic tendencies (Tran et al., 1985;Sie et al., 1985), suggesting that HCII, like ATIII, is an important inhibitor of thrombin in vivo.
An important feature of serpin metabolism was originally identified when Lollar et al. (1980) showed that 125 I-thrombin in complex with ATIII was rapidly cleared from the circulation of rabbits by the liver. Additional studies by several groups (Bauer et al., 1982;Fuchs et al., 1982;Pratt et al., 1988) have demonstrated that the liver mediates rapid clearance of complexes of ATIII⅐thrombin, HCII⅐thrombin, ␣ 1 AT⅐trypsin, and ␣ 1 AT⅐elastase from the circulation. Further, these complexes could compete for one another's clearance, indicating that the plasma elimination of these SECs was occurring via a common receptor pathway. Mast et al. (1991) showed that the kinetics of plasma elimination of the native and cleaved inhibitors were 10 -50 times slower than that of the complexed inhibitors, suggesting that the liver receptor might preferentially clear the SEC as opposed to the free serpin.
The low density lipoprotein receptor (LDLR) and low density lipoprotein receptor-related protein (LRP) are two prominent hepatic receptors that are members of the LDLR superfamily. This receptor family also includes the very low density lipoprotein receptor and glycoprotein 330 (gp330). LRP is expressed in many tissues and cell types  and is present in high abundance in the liver. The requirement for LRP expression during development was demonstrated when targeted disruption of the LRP gene resulted in termination of the null embryos at day 13.5 during development . LRP interacts with numerous ligands and is thought to play key roles in both lipoprotein metabolism and proteinase regulation (for reviews, see Krieger and Herz (1994), Strickland et al. (1994Strickland et al. ( , 1995, and Moestrup (1994)).
In addition to mediating the catabolism of ␣ 2 -macroglobulin (␣ 2 M)-proteinase complexes , LRP also binds to several SECs, including elastase⅐␣ 1 AT (Poller et al., 1995), urinary-type plasminogen activator (uPA)⅐PAI-1 Nykjaer et al., 1992), and tissue plasminogen activator ⅐PAI-1 (Orth et al., 1992). However, the receptor responsible for the hepatic catabolism of a number of other predominant SECs remained obscure. Studies were therefore initiated to determine if LRP could interact with those SECs. In this report we demonstrate that LRP binds directly to complexes of ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin, can mediate their cellular uptake and degradation, and appears responsible for their in vivo clearance from the circulation.

EXPERIMENTAL PROCEDURES
Proteins-LRP was isolated from human placenta as described by Ashcom et al. (1990). Human RAP, expressed in bacteria as a fusion protein with glutathione S-transferase, was prepared and purified as described previously . ␣ 2 M was purified from human plasma by zinc-chelate chromatography as described previously . ␣ 2 M⅐methylamine, designated (␣ 2 M*), was prepared by incubation of native ␣ 2 M with 200 mM methylamine (Sigma) for 30 min at 25°C followed by dialysis against Tris-buffered saline (TBS). HCII was purified from human plasma as described previously (Griffith et al., 1985). Antithrombin III was isolated from human plasma as described (Miller-Andersson et al., 1974). ␣ 1 AT was purchased from Athens Research and Technology (Athens, GA). Inhibitors inactivated by cleavage of the reactive center loop, without complex formation, are referred to as modified or cleaved inhibitors. Modified inhibitors were prepared by incubation of inhibitor in TBS (1 mg/ml) with 100 l of activated immobilized papain-Sepharose (Pierce) for 30 min at 25°C. To monitor the efficiency of papain to cleave inhibitors, cleaved preparations were tested for their ability to inhibit target proteinases. In addition, SDS-polyacrylamide gel electrophoresis analysis of the inhibitors revealed a shift in mobility on 8 -16% polyacrylamide gradient gels (Novex, San Diego, CA) as compared with native inhibitors and also showed that papain treatment did not result in excess proteolysis of the cleaved inhibitors. Thrombin (␣-form) was purchased from Enzyme Research Laboratories (South Bend, IN). Trypsin was purchased from Sigma. To prepare complexes of ATIII⅐thrombin and HCII⅐thrombin, the enzyme and inhibitor were combined in a 1:2 molar ratio and were incubated for 30 min at 25°C. Complexes of ␣ 1 AT⅐trypsin were prepared by incubation of trypsin and ␣ 1 AT at a 1:2 molar ratio for 5 min at 25°C, followed by addition of soybean trypsin inhibitor (Sigma) (in a molar amount equal to that of trypsin added) to inhibit any residual trypsin activity. uPA was provided by Jack Henkin (Abbott Park, IL). Active PAI-1 was purchased from Molecular Innovations (Royal Oak, MI). Complexes of uPA⅐PAI-1 were prepared by incubation of a 1:1 molar ratio of active PAI-1 with uPA for 30 min at 25°C. Proteins were labeled with [ 125 I]iodine to a specific activity ranging from 2 to 10 Ci/g of protein using IODOGEN (Pierce). Bovine serum albumin, fraction V (BSA) was purchased from Sigma.
Antibodies-A rabbit polyclonal antibody to the 515-kDa heavy chain of LRP (R777) and another one to a synthetic peptide from the cytoplasmic domain of the 85-kDa light chain of LRP (R704) have been described elsewhere (Kounnas et al., 1992). R777 antibodies were affinity-purified on a column of LRP-Sepharose. R704 antibodies were purified on protein G-Sepharose (Pharmacia Biotech Inc.). Both of the rabbit antibody preparations were dialyzed against isotonic PBS, heatinactivated for 30 min at 56°C, and filtered with a 0.45 M Acrodisc (Corning, Corning, NY) prior to use in cell assays. The mouse monoclonal antibody to LRP, designated 8G1, has been described .
Solid-phase Binding Assays-Enzyme-linked immunosorbent assays (ELISA) were performed as detailed elsewhere (Kounnas et al., 1993). Briefly, microtiter wells were coated with the native, cleaved, or complexed forms of ATIII, HCII, or ␣ 1 AT (prepared as described under "Proteins") at 10 g/ml in TBS, pH 8.0, for 4 h at 37°C, blocked with 3% BSA, TBS, pH 8.0, then incubated with various concentrations of LRP (0.6 -150 nM) in 3% BSA, TBS, 5 mM CaCl 2 , 0.05% Tween-20 for 18 h at 4°C. For ELISAs measuring the effects of RAP, coated wells were incubated with 20 nM LRP in the presence of increasing concentrations of RAP (0.2-450 nM). LRP binding was detected with the mouse monoclonal antibody 8G1.
Cells-Human hepatocellular carcinoma (HepG2, ATCC HB 8065) cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and were grown in Eagle's minimal essential medium containing 10% bovine calf serum (HyClone Laboratories, Logan, UT), 100 units/ml penicillin (Life Technologies, Inc.), 100 g/ml streptomycin (Life Technologies, Inc.), and 1% L-glutamine (Life Technologies, Inc.). A normal mouse embryonal fibroblast line (MEF) and a mouse embryonal fibroblast cell line that is deficient in LRP biosynthesis (PEA13) were obtained from Dr. Joachim Herz (University of Texas Southwestern Medical Center, Dallas, TX) and maintained as described by .
Cell-mediated Ligand Internalization and Degradation Assays-Cells were seeded into 12-well dishes (Corning) at 2-3 ϫ 10 5 cells/well and allowed to grow for 24 h at 37°C, 5% CO 2 . Cellular internalization and degradation assays were conducted according to Kounnas et al. (1995a). Prior to performing the assays, cultured cells were washed with medium and incubated in medium containing 1% Nutridoma (Boehringer Manneheim), 20 mM Hepes, penicillin/streptomycin, and 1.5% BSA (assay medium). For assays with HepG2 cells, radiolabeled serpin-enzyme complexes (10 nM) in assay medium were incubated with cells for 18 h at 37°C in the presence of RAP (1 M), LRP antibodies (100 g/ml), or chloroquine (0.1 mM, Sigma). For assays using antibodies, cultured cells were preincubated for 1 h at 37°C with assay medium containing antibodies. Antibodies (100 g/ml) were also included during the cellular assay. For assays utilizing cultured fibroblasts, 125 I-proteins (5-10 nM) were added in assay medium either alone or in the presence of RAP (1 M) for the indicated times. Total radioactivity secreted in the cell culture medium that was soluble in 10% trichloroacetic acid was corrected for non-cellular mediated degradation by subtracting the amount of degradation that occurred in parallel wells lacking cells. The amount of radiolabeled ligand that was internalized by cells was defined as the amount of radioactivity that remained associated with the cell pellet following trypsin-EDTA, proteinase K (Sigma) treatment (Chappell et al., 1992).
Clearance of ATIII⅐ 125 I-Thrombin from the Plasma of Rats-Sprague-Dawley rats (200 g) were anesthetized with ketamine (100 mg/ml)/ xylazine (20 mg/ml) at a dose of 90 mg/kg ketamine, 8 mg/kg xylazine. A bolus of 500 l of ATIII⅐ 125 I-thrombin (100 nM) in the presence or absence of RAP (110 M) was injected into the tail vein over a period of approximately 15 s. At selected time intervals following injection (1, 5, 10, and 20 min), blood (200 l) was collected from the vena cava into 10 l of 0.5 M EDTA, and an aliquot (50 l) was counted for its 125 I content.
In those experiments examining the clearance of 125 I-␣ 2 M*, a bolus of 500 l of 125 I-␣ 2 M (10 nM) in the presence or absence of RAP (32 M) was injected. The initial time point, taken 1 min after injection, was considered to represent 100% radioactivity in the circulation. The clearance of each preparation was examined in two rats and the results averaged.

RESULTS
ATIII⅐Thrombin, HCII⅐Thrombin, and ␣ 1 AT⅐Trypsin Bind to LRP-The ability of LRP to bind to native, cleaved, or complexed forms of ATIII, HCII, and ␣ 1 AT was measured using an ELISA. As shown in Fig. 1, LRP binds to ATIII⅐thrombin (panel A), HCII⅐thrombin (panel B), and ␣ 1 AT⅐trypsin (panel C), while no detectable binding to the native or cleaved forms of ATIII (panel A), HCII (panel B), and ␣ 1 AT (panel C) was observed. The lack of saturation in these assays indicates that the affinity of LRP for these SECs is weaker than that measured for other LRP ligands. Apparent half-saturation values measured for the binding interaction of these complexes with LRP are estimated to range from 80 to 120 nM.
Since RAP is known to inhibit the binding of ligands to LRP (Herz et al., 1991), binding assays were performed to test the effect of RAP on the binding of LRP to ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin. Fig. 2 depicts an ELISA in which LRP was incubated with serpin-proteinase-coated wells in the presence of increasing concentrations of RAP. These data demonstrate that RAP is capable of competing for the binding of these SECs to LRP.
Efficient Uptake and Degradation of Complexes of 125 I-Thrombin⅐ATIII, 125 I-Thrombin⅐HCII, and 125 I-Trypsin⅐␣ 1 AT Is Abolished in Fibroblasts Lacking LRP-Cellular uptake and degradation experiments in mouse embryonal fibroblasts (MEF) which express LRP and in a mutant MEF cell line that lacks LRP (PEA13) were conducted in order to determine the role of LRP in the catabolism of 125 I-labeled complexes of ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin. For these experiments, trypsin or ␣-thrombin were labeled with 125 Iiodine and then complexed to an inhibitor. As shown in Fig. 3, little or no detectable free enzyme remained after addition of inhibitors. Fig. 4 shows that LRP-expressing MEF cells, but not LRP-deficient PEA13 cells, were capable of internalizing and degrading complexes of ATIII⅐ 125 I-thrombin (panels A and B), HCII⅐ 125 I-thrombin (panels C and D), and ␣ 1 AT⅐ 125 I-trypsin (panels E and F). Further, in MEF cells, the uptake and degradation of labeled complexes was inhibited by RAP, an antagonist of the binding of LRP to ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin (see Fig. 2). The ability of RAP to block uptake and degradation of 125 I-SECs in LRP-expressing cells and the inability of LRP-deficient cells to mediate SEC uptake and degradation point to LRP as a mediator of the endocytosis of SECs.
LRP Antibodies Inhibit the Degradation of ATIII⅐ 125 I-Thrombin, HCII⅐ 125 I-Thrombin, and ␣ 1 AT⅐ 125 I-Trypsin in HepG2 Cells-To further implicate LRP in mediating the uptake and degradation of ATIII⅐ 125 I-thrombin, HCII⅐ 125 I-thrombin, and ␣ 1 AT⅐ 125 I-trypsin complexes, the effect of LRP antibodies on the degradation of 125 I-labeled SECs by HepG2 cells was determined. Fig. 5 shows that RAP, LRP antibodies, and chloro-quine (a drug that blocks lysosomal degradation) inhibit the degradation of ATIII⅐ 125 I-thrombin (panel A), HCII⅐ 125 I-thrombin (panel B), and ␣ 1 AT⅐ 125 I-trypsin (panel C). In each case, control antibodies had little effect. These data demonstrate that LRP is capable of mediating the endocytosis leading to lysosomal degradation of labeled ATIII⅐thrombin, HCII⅐ thrombin, and ␣ 1 AT⅐trypsin in HepG2 cells, raising the possibility that LRP can function in vivo to mediate the liver clearance of these complexes.
The Uptake and Degradation of 125 I-ATIII, 125 I-HCII, and 125 I-␣ 1 AT Requires Complex Formation with an Enzyme-Previous studies examining the clearance of 125 I-labeled SECs from plasma in a mouse model have shown that while the complexed form of ATIII, HCII, and ␣ 1 AT are cleared rapidly by hepatocytes, the clearance of the native and cleaved inhibitors is slow (Mast et al., 1991). We, therefore performed assays to investigate the LRP-mediated cellular endocytosis of 125 Ilabeled inhibitors in the native, cleaved, or complexed forms. Fig. 6 shows the uptake of 125 I-ATIII (native, cleaved, and thrombin complexed) in MEF (LRP-expressing cells) (panel A) and PEA13 (LRP-deficient cells) (panel B), 125 I-HCII (native, cleaved, and thrombin complexed) in MEF and PEA13 cells (panels C and D, respectively), and 125 I-␣ 1 AT (native, cleaved, and complexed to trypsin) in MEF and PEA13 cells (panels E and F, respectively). These data demonstrate that the labeled inhibitors alone (native or cleaved) are not endocytosed by mouse fibroblasts. However, complexes of 125 I-ATIII⅐thrombin, 125 I-HCII⅐thrombin, and 125 I-␣ 1 AT⅐trypsin are internalized efficiently by LRP-expressing MEF cells but not the mutant LRP-deficient PEA13 cells and this process is inhibited by RAP. Further, these complexes are efficiently degraded in LRPexpressing cells (but not LRP-deficient cells) in a RAP-dependent manner with similar kinetics as shown in Fig. 4 (data not shown). These data reveal that proteinase complex formation of ATIII, HCII, and ␣ 1 AT is required for LRP-mediated endocytosis.
Comparative Degradation of Different LRP Ligands by Mouse Fibroblasts-Complexes of uPA⅐PAI-1 bind to LRP with an affinity constant of 1 nM (Nykjaer et al., 1994), while complexes of ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin bind to LRP with estimated affinities of 80 -120 nM. This observation prompted us to compare the amounts of 125 I-labeled SECs that were degraded by cells in an LRP-dependent manner. The first experiments were designed to determine the relative levels of 125 I-labeled SEC degraded by mouse fibro- blasts expressing LRP (MEF) using a ligand concentration of 1 nM. These data are shown in Fig. 7 (panel A) and reveal that the amount of uPA⅐PAI-1 degraded in an LRP-dependent manner (as determined by the ability of RAP to compete) is approximately 10-fold greater than that of the other SECs examined and approximately 3-fold greater than ␣ 2 M*, another LRP ligand. Further experiments were conducted using a dose of 125 I-labeled ligand (0.6 -150 for ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin (panel B) and 0.6 -16.6 nM for uPA⅐PAI-1 (panel C)). From these data it is evident that at a given concentration of 125 I-labeled SEC, approximately 10-fold more uPA⅐PAI-1 is degraded when compared to the other SECs. For example, at a ligand concentration of 15 nM, approximately 2000 fmol of uPA⅐PAI-1 is degraded in 18 h, an amount that is achieved when a 10-fold higher concentration (150 nM) of 125 Ilabeled ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin is used. These data are consistent with the higher affinity of LRP for uPA⅐PAI-1 (1 nM) and a lower affinity of LRP for ATIII⅐ thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin (approximately 80 -120 nM).

Effect of RAP on the Plasma Clearance of 125 I-␣ 2 M* and ATIII-125 I-Thrombin
Complexes-To evaluate the role of LRP in mediating the plasma clearance of SECs, the effect of RAP on the clearance rate of a representative SEC, the ATIII-thrombin complex, was measured. RAP is known to have a pronounced effect on the removal of 125 I-␣ 2 M* from the mouse circulation , and thus 125 I-␣ 2 M* was used as a control for these experiments. The results, shown in Fig. 8, demonstrate that when co-injected with the ligand, RAP significantly delays the clearance of 125 I-␣ 2 M* (Fig. 8A) and ATIII-125 I-thrombin complexes (Fig. 8B) from the rat circulation. These studies indicate that a RAP-sensitive hepatic receptor, most likely LRP, plays a major role in the removal of ATIII-125 I-thrombin complexes from the circulation.

DISCUSSION
The existence of a hepatic receptor that is responsible for binding and removing SECs from the plasma was demonstrated in early studies (Pizzo and Gonias, 1984;Shifman and Pizzo, 1982;Imber and Pizzo, 1981;Fuchs et al., 1982;Pratt et al., 1988) in which labeled SECs (ATIII⅐thrombin, HCII⅐thrombin, ␣ 1 AT⅐trypsin, and ␣ 1 AT⅐elastase) were shown to be rapidly cleared by the liver (half-lives ϭ 3, 10, 15, and 20 min, respectively). These complexes apparently all bound to the same receptor, since they were able to cross-compete with one another (Pratt et al., 1988). However, clearance of these molecules was not competed with high concentrations of ␣ 2 M⅐proteinase complexes (Fuchs et al., 1982). The clearance mechanism appears specific for the enzyme-complexed form of these serpins, since studies by Mast et al. (1991) demonstrated that the clearance of native and cleaved forms of ATIII, HCII, and ␣ 1 AT occurred at a considerably slower rate (half-lives Ͼ 1 h) than the proteinase-complexed serpins.
The data presented in the current study suggest that LRP is the hepatic receptor responsible for the clearance of these complexes. This conclusion is supported by several independent lines of evidence. First of all, we demonstrated that purified LRP can directly bind to several SECs (ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin) coated on microtiter wells, and that the binding is competed by RAP. Second, SECs are internalized and degraded in mouse fibroblasts that express LRP, but not in mouse fibroblasts genetically deficient in LRP. Third, we demonstrated that LRP antagonists, RAP and LRP antibodies, blocked the cellular uptake and degradation of 125 Ilabeled ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin. Fourth, our studies have shown that the interaction of SECs with LRP has properties expected of the hepatic clearance mechanism; LRP does not recognize the native and cleaved forms of the inhibitors but readily internalizes the proteinase complexed form of the inhibitors. Taken together, these data suggest that the liver receptor responsible for the plasma clearance of ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin is LRP.
To demonstrate this, in vivo clearance studies were performed in rats. The ability of RAP to inhibit hepatic clearance of ATIII-125 I-thrombin indicates that a RAP-sensitive hepatic receptor is responsible for this process. While RAP is known to bind to all members of the LDLR superfamily, only two of these receptors, LDLR and LRP, are known to be expressed in high levels in the liver. Our studies indicate that the LDLR does not play a significant role in the catabolism of SECs, since PEA13 fibroblasts, which express LDLR and are normal in their ability to catabolize LDL Kounnas et al., 1995c), are unable to mediate the cellular catabolism of SECs. Thus, the in vivo clearance studies suggest that LRP is playing a prominent role in the hepatic clearance of SECs. Perlmutter et al. (1990aPerlmutter et al. ( , 1990b) described a SEC receptor on monocytes and HepG2 cells which internalized complexes of ␣ 1 AT⅐trypsin and ␣ 1 AT⅐elastase, and it is highly likely that this receptor is LRP. In the present investigation, we found that the cellular uptake and degradation of ATIII⅐thrombin, HCII⅐ thrombin, and ␣ 1 AT⅐trypsin by Hep G2 cells is inhibited by anti-LRP IgG. In addition to a receptor that mediates the endocytosis of SECs, studies by several investigators suggest that binding of SECs to the cell surface may result in signal transduction events. Perlmutter et al. (1988) demonstrated that binding of elastase⅐␣ 1 AT complexes to monocytes or macrophages resulted in increased gene expression of ␣ 1 AT, while Banda et al. (1988) showed that cleaved ␣ 1 AT stimulated neutrophil chemotaxis. Hoffman et al. (1989) reported that HCII could be cleaved by proteinases to generate products that stimulated neutrophil chemotaxis, suggesting that cleaved serpins release biologically active molecules. While the mechanism of signal transduction by these complexes is unclear, it seems likely that a distinct signaling receptor, different from LRP, may exist and would account for these effects.
The existence of a clearance mechanism for SECs is likely important since such complexes are known to dissociate with time, releasing active enzyme along with the cleaved inhibitor (Travis and Salvesen, 1983;Carrell and Boswell, 1986). ATIII is a primary inhibitor of thrombin and FX a and plays a major role in maintaining normal hemostasis (Bjork et al., 1989;Olson and Bjork, 1992). Disruption of the normal balance of thrombin and inhibitor activity results in either thrombosis (Egeberg, 1965;Lane et al., 1991) or bleeding disorders (Owen et al., 1983); therefore, it is clear that maintaining an appropriate amount of thrombin activity is critical for normal hemostasis. The principle action of ATIII may be to localize clotting by inhibiting thrombin that has escaped into the circulation, and the LRP-mediated clearance of ATIII⅐thrombin complexes may serve to help maintain normal hemostasis.
The imbalance of ␣ 1 AT/proteinase activity has been associated with various pathological conditions such as pulmonary emphysema (Laurell and Eriksson, 1963;Travis and Salvesen, 1983;Carrell, 1986). ␣ 1 AT has been implicated as a primary modulator of elastase and proteinase 3 activity during inflammatory processes. In acute inflammation, infiltrating neutrophils and macrophages, as well as damaged cells, release excessive amounts of proteinases resulting in tissue damage. This damage can be minimized by rapid inhibition of proteinase activity by inhibitors present in body fluids. Elastase and PR3 are released by neutrophils during inflammation and participate in the degradation of extracellular matrix components. At the indicated times, blood (200 l) was collected into 10 l of 0.5 M EDTA and an aliquot (50 l) was counted for its 125 I content. B, a bolus of 500 l of ATIII⅐ 125 Ithrombin (100 nM) in the absence (q) or presence (Ⅺ) of RAP (110 M) was injected into the tail vein of rats. At the indicated times, blood (200 l) was collected into 10 l of 0.5 M EDTA and an aliquot (50 l) was counted for its 125 I content. The clearance of each preparation was examined in two rats, and the data plotted represent the average value Ϯ S.E.
Once these enzymes are inhibited by ␣ 1 AT, the complexes formed are known to be cleared by macrophages and fibroblasts in a receptor-mediated process. LRP is found in high concentration on fibroblasts and macrophages  and has been shown to bind and mediate the uptake of at least two ␣ 1 AT⅐proteinase complexes (Poller et al., 1995;this study) and therefore is very likely to represent the receptor responsible for the endocytosis of ␣ 1 AT⅐proteinase complexes formed during the inflammatory response.
In summary, the current investigation demonstrates that LRP can bind and endocytose complexes of ATIII⅐thrombin, HCII⅐thrombin, and ␣ 1 AT⅐trypsin, but not the native or the cleaved inhibitors. In vivo clearance studies confirm that LRP functions to remove SECs from the plasma. These findings expand the role of LRP as a SEC receptor.