Role of the Low Density Lipoprotein-related Protein Receptor in Mediation of Factor VIII Catabolism*

In the present study, we found that catabolism of coagulation factor VIII (fVIII) is mediated by the low density lipoprotein receptor-related protein (LPR), a liver multiligand endocytic receptor. In a solid phase assay, fVIII was shown to bind to LRP (K d 116 nm). The specificity was confirmed by a complete inhibition of fVIII/LRP binding by 39-kDa receptor-associated protein (RAP), an antagonist of all LRP ligands. The region of fVIII involved in its binding to LRP was localized within the A2 domain residues 484–509, based on the ability of the isolated A2 domain and the synthetic A2 domain peptide 484–509 to prevent fVIII interaction with LRP. Since vWf did not inhibit fVIII binding to LRP, we proposed that LRP receptor may internalize fVIII from its complex with vWf. Consistent with this hypothesis, mouse embryonic fibroblasts that express LRP, but not fibroblasts genetically deficient in LRP, were able to catabolize 125I-fVIII complexed with vWf, which was not internalized by the cells. These processes could be inhibited by RAP and A2 subunit of fVIII, indicating that cellular internalization and degradation were mediated by interaction of the A2 domain of fVIII with LRP. In vivo studies of125I-fVIII·vWf complex clearance in mice demonstrated that RAP completely inhibited the fast phase of the biphasic125I-fVIII clearance that is responsible for removal of 60% of fVIII from circulation. Inhibition of the RAP-sensitive phase prolonged the half-life of 125I-fVIII in circulation by 3.3-fold, indicating that LRP receptor plays an important role in fVIII clearance.

The plasma glycoprotein factor VIII (fVIII) 1 functions as a cofactor for factor IXa in the factor X activation enzyme complex in the intrinsic pathway of blood coagulation, and its level is decreased or the protein is nonfunctional in patients with hemophilia A. The fVIII protein consists of a homologous A and C domains and a unique B domain which are arranged in the order A1-A2-B-A3-C1-C2 (1). It is processed to a series of Me 2ϩ -linked heterodimers produced by cleavage at the B-A3 junction (2), generating a light chain (LCh) which consists of an acidic region and A3, C1, and C2 domains and a heavy chain (HCh) which consists of the A1, A2, and B domains (Fig. 1).
Transplantational studies both in animals and humans demonstrated that liver hepatocytes are the major fVIII-producing cells (3,4). Immediately after release into circulation, fVIII binds with a high affinity (K d Ͻ 0.5 nM (5,6)) to its carrier protein vWf to form a tight, noncovalent complex. The binding to vWf is required for maintenance of a normal fVIII level in circulation, since vWf stabilizes association of the LCh and HCh (7). This prevents fVIII from binding to phospholipid membranes (8), activation by factor Xa (9), and protein Ccatalyzed inactivation (10). vWf comprises a series of high molecular mass, disulfide-bonded multimers with molecular mass values as high as 2 ϫ 10 7 Da (11) and circulates in plasma at 10 g/ml or 50 nM assuming a molecular mass of 270 kDa for vWf monomers (12). Since the concentration of fVIII in plasma is approximately 1 nM (13), one fVIII molecule is bound per 50 vWf monomers in plasma (14).
Activation of fVIII by thrombin leads to dissociation of activated fVIII (fVIIIa) from vWf and to at least 100-fold increase of the cofactor activity. The fVIIIa is a A1/A2/A3-C1-C2 heterotrimer (15) in which domains A1 and A3 retain the metal ion linkage (Fig. 1). The stable dimer A1/A3-C1-C2 is weakly associated with the A2 subunit through electrostatic forces (15). Spontaneous dissociation of the A2 subunit from the heterotrimer results in non-proteolytic inactivation of fVIIIa (15).
Infusion of fVIII⅐vWf complex, purified plasma, or recombinant fVIII into patients with severe hemophilia A who do not have fVIII (16,17) or in normal individuals (18) results in a similar fVIII disappearance with a half-life of 12-14 h. Although the complex formation between fVIII and vWf is crucial for the normal half-life and level of fVIII in circulation, the mechanisms associated with turnover of fVIII⅐vWf complex are not well defined. We proposed that the fVIII⅐vWf complex is eliminated from plasma via a clearance receptor and tested the possibility whether this receptor is a low density lipoproteinrelated protein receptor (LRP). LRP-mediated cellular endocytosis was shown to be a mechanism of removal of a number of structurally unrelated ligands including several proteins involved in coagulation or fibrinolysis. These ligands are complexes of thrombin with antithrombin III, heparin cofactor II (19), protease nexin I (20), urokinase-type and tissue-type plasminogen activators, respectively, complexes with plasminogen activator inhibitor (21,22), thrombospondin (23), tissue factor pathway inhibitor (24), and factor Xa (25,26).
LRP, a large cell-surface glycoprotein identical to ␣ 2 -macroglobulin receptor (27), is a member of the low density lipoprotein receptor family which also includes the LDL receptor, very low density lipoprotein receptor, vitellogenin receptor, and gly-coprotein 330 receptor. LRP receptor consists of the noncovalently linked 515-kDa ␣-chain (28) containing binding sites for LRP ligands, and the 85-kDa transmembrane ␤-chain. The cluster of 31 cysteine-rich class A repeats participates in binding of different ligands (29). The presence of multiple repeats may be responsible for wide ligand diversity of LRP and its ability to serve as a multiligand clearance receptor. In contrast to the acidic ligand-binding region in LRP, its ligands expose regions rich in positively charged amino acid residues (30).
LRP is a major endocytic receptor in the liver (31) but it also expressed in many cell types and tissues including placenta, lung, and brain (32). A 39-kDa receptor-associated protein (RAP) binds to LRP with high affinity (K d ϭ 4 nM (27)) and inhibits binding and LRP-mediated internalization and degradation of all ligands (30,33), therefore serving as a useful tool for testing whether LRP is involved in endocytosis of a given ligand.
In the present study we demonstrated that fVIII specifically binds to LRP, and that LRP mediates the internalization and subsequent degradation of fVIII in cultured fibroblasts and plays a significant role in the clearance of fVIII in vivo. We also found that interaction of the A2 domain of fVIII with LRP is responsible for mediating catabolism of fVIII and localized the A2 domain region that is directly involved in the binding.
Proteins-LRP was isolated from human placenta as described (37). Human recombinant RAP was expressed in bacteria and purified as described (33). FVIII was purified from therapeutic concentrates of Method M, American Red Cross (38). HCh and LCh were prepared from fVIII as described previously (6). Purification of the A1/A3-C1-C2 dimer and A2 subunit was performed using ion exchange chromatography of thrombin-activated fVIII on a Resource S column (Amersham Pharmacia Biotech) (39). Residual A2 present in the A1/A3-C1-C2 preparation was removed by its passage over an immobilized mAb 8860 column equilibrated in 20 mM Tris, pH 7.4, 0.15 M NaCl, 5 mM CaCl 2 .
Radiolabeling of fVIII and Its A2 Subunit-Prior to iodination, vWf, fVIII, and A2 were dialyzed into 0.2 M sodium acetate, 5 mM calcium nitrate, pH 6.8 (iodination buffer). Five g of fVIII in 30 l of iodination buffer were added to lactoperoxidase beads (Worthington Biochemical Corp.), containing 5 l of Na 125 I (100 mCi/ml, Amersham Pharmacia Biotech) and 5 l of 0.03% H 2 O 2 (Mallincrodt) and incubated for 4 min. Free Na 125 I was removed by chromatography on a PD10 column (Amersham Pharmacia Biotech). The specific radioactivities of 125 I-labeled fVIII and A2 were 3.5 and 5 Ci/g protein, respectively. The activity of 125 I-fVIII determined in the one-stage clotting assay (40) (3740 units/ mg) was similar to that of unlabeled fVIII (3960 units/mg).
Solid-phase Binding Assays-Homologous and heterologous ligand displacement assays were performed as described previously (33). Microtiter wells were coated with purified LRP or BSA (3 g/ml) in 50 mM Tris, 0.15 M NaCl, pH 8.0, for 16 h at 4°C and then blocked with 3% BSA in TBS. Coated wells were incubated with 125 I-fVIII in 20 mM Tris-buffered saline, pH 7.4, containing 5 mM CaCl 2 , 0.05% Tween 20 in the presence or absence of unlabeled competitors (vWf, fVIII, or RAP) for 1 h at 37°C. In the experiment using vWf, 125 I-fVIII was preincubated in the above buffer in the presence of varying concentrations of vWf for 30 min at 37°C. This was followed by determination of radioactivity bound to the wells. Affinity constants were derived from homologous and heterologous displacement data using the computer program LIGAND (41).
Cell-mediated Ligand Internalization and Degradation Assays-A normal mouse embryonic fibroblast line (MEF) and a mouse embryonic fibroblast cell line that is genetically deficient in LRP biosynthesis (PEA 13) were obtained from Dr. Joachim Herz (University of Texas Southwestern Medical Center, Dallas, TX) and maintained as described (42). Cells were seeded at 1 ϫ 10 5 cells/well and allowed to grow Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (Life Technologies, Inc.) for 24 h at 37°C, 5% CO 2 . Cellular internalization and degradation assays were conducted as described previously (43). Internalization and degradation of the 125 I-labeled fVIII and A2 was measured after incubation for varying time intervals at 37°C in 0.5 ml of Dulbecco's modified Eagle's medium containing 2% BSA. Surface binding of radiolabeled ligand was defined as the amount of radioactivity released by treatment of the cells by trypsin (50 g/ml) and proteinase K (50 g/ml) (Sigma) in a buffer containing 5 mM EDTA (44). This treatment was previously shown to release radioligands bound to cell surface (43), therefore a ligand which remained associated with the cells after this treatment was considered as internalized. Degradation was defined as radioactivity in the medium that is soluble in 10% trichloroacetic acid. The value of degradation was corrected for noncellular mediated degradation by subtracting the amount of degradation products generated in parallel wells lacking cells.
Clearance of 125 I-fVIII⅐vWf Complex from Mouse Plasma-Prior to the experiment, 125 I-fVIII, vWf, and RAP were dialyzed into 20 mM Hepes, 0.15 M NaCl, pH 7. weighed, followed by measuring the radioactivity in these tissues.
The kinetics of 125 I-fVIII clearance from circulation was described using a double-exponential model (45), where C is the percent of fVIII remaining in plasma at a given time, k 1 and k 2 are the kinetic rate constants corresponding to fast and slow phases of fVIII clearance, and C 1 and C 2 are percentages of administered radioactivity removed during the fast and slow phases of clearance, respectively. Clearance data for the saturating concentration of RAP (250 M) was fitted using a single exponential equation derived from Equation 1 by deleting the first exponential component. The values of k 1 , k 2 , C 1 , and C 2 constants were derived for each clearance curve by fitting C versus t to Equation 1 using Sigmaplot 3.0 computer program (Jandel Scientific).

Factor VIII Binds to LRP and Its Binding Is Prevented by RAP-
The ability of fVIII to bind to LRP in vitro was examined in the homologous displacement binding assay. In the assay, binding of 125 I-fVIII (1 nM) to purified LRP, but not to BSAcoated wells, was inhibited (Ͼ90%) by an excess of unlabeled fVIII ( Fig. 2A). The quantitative data regarding fVIII interaction with LRP were derived from the homologous displacement of 125 I-fVIII by unlabeled fVIII, which was adequately described by a model containing a single class of fVIII-binding sites with K d of 116 nM. To elucidate whether fVIII in a complex with vWf is also able to bind to LRP, we tested the effect of vWf on 125 I-fVIII binding to immobilized LRP. In this experiment, 125 I-fVIII was preincubated with vWf as described under "Experimental Procedures" to allow complex formation prior to its addition to LRP-coated wells. As shown in Fig. 2A, the binding of fVIII to LRP was not inhibited by addition of up to 1000 nM vWf, which is 20-fold higher than its concentration in plasma (50 nM) (14). This indicates that the complex formation with vWf does not affect fVIII ability to bind to LRP.
RAP, an antagonist of LRP-ligand binding, completely inhibited the binding of 125 I-fVIII to LRP-coated microtiter with K i of 2.5 nM (Fig. 2B), the value is similar to the previously determined affinity (4 nM) of RAP for LRP (27). Together, these results demonstrate specific fVIII binding to LRP.
The Amino Acid Residues 484 -509 within the fVIII A2 Domain Are Responsible for fVIII Binding to Purified LRP-In order to localize the fVIII region(s) involved in interaction with LRP, the binding between 125 I-fVIII and immobilized LRP was competed by unlabeled fVIII fragments. As shown in Fig. 3, HCh and A2 domains of fVIII, but not LCh (AR-A3-C1-C2) or A1/A3-C1-C2 dimer, displaced 125 I-fVIII from LRP in the het-erologous ligand displacement assay. The K i values determined for HCh and A2 were similar, 120 and 132 nM, respectively. The similarity of the above K d value for fVIII binding to LRP and the K i value for inhibition of this binding by isolated A2 subunit indicate that the A2 domain of HCh contains a major site for fVIII binding to LRP.
To localize the region of the A2 domain responsible for interaction with LRP, we tested the effect of anti-A2 monoclonal antibodies with known epitopes on fVIII/LRP binding. Fig. 4A shows that mAb 413 (epitope within the A2 domain residues 484 -509 (35)) but not mAb T5 (epitope within the A2 domain residues 701-740 (34)) is able to block the fVIII/LRP interaction. The concentration of mAb 413 required for 50% inhibition of 125 I-fVIII/LRP binding was 2.5 nM. The low molar excess (2.5-fold) of mAb 413 over fVIII required for 50% inhibition of fVIII/LRP binding is consistent with a previously reported high affinity of mAb 413 for fVIII (46).
Since it was previously demonstrated that mAb 413 recognizes a synthetic peptide consisting of A2 domain residues 484 -509 (35), we tested whether these residues are involved in binding to LRP. As seen from Fig. 4B, the synthetic peptide 484 -509, but not control A2 peptide 432-456, inhibited fVIII binding to LRP in a dose-dependent fashion, indicating that the region 484 -509 of the A2 domain does contain residues critical for fVIII binding to LRP. In the control experiment, no binding of 125 I-fVIII to BSA-coated wells was observed in the presence of peptide 484 -509 (Fig. 4B).
Internalization and Degradation of 125 I-fVIII Complex with vWf by Cultured Fibroblasts Is Mediated by LRP-Since the above data demonstrated a specific interaction between fVIII and LRP, and vWf did not interfere with this interaction, we hypothesized that LRP may also be capable of mediating the cellular internalization of 125 I-fVIII from its complex with vWf. To examine this hypothesis, the cellular uptake and degradation experiments were conducted on MEF which express LRP and on PEA 13 fibroblasts that are genetically deficient in LRP (42). The fVIII⅐vWf complex was prepared by mixing fVIII and vWf at their plasma concentrations of 1 and 50 nM, respectively. As shown in Fig. 5, A and B, MEF cells, but not PEA 13 cells lacking LRP, were capable of internalizing and degrading 125 I-fVIII in the presence of vWf. In addition, internalization and degradation of 125 I-fVIII by MEF but not by PEA 13 fibroblasts was inhibited by RAP, an antagonist of ligand binding to LRP. The ability of RAP to block the uptake and degradation of fVIII/vWf in MEF cells and inability of PEA13 cells to efficiently mediate the uptake and degradation indicates that LRP is a mediator of fVIII/vWf catabolism. To further characterize the degradation pathway of fVIII in MEF cells, we tested the effect of chloroquine (an agent that blocks lysosomal degradation) on fVIII degradation. As seen from Fig. 5B, the degradation of 125 I-fVIII is completely inhibited by chloroquine.
To elucidate whether cellular mediated endocytosis of fVIII in the absence of vWf is also mediated by LRP, we compared internalization and degradation of 125 I-fVIII⅐vWf complex and isolated 125 I-fVIII (Fig. 6). As seen from Fig. 6, A and B, both internalization and degradation of isolated 125 I-fVIII by MEF fibroblasts is approximately 2-fold higher than that in the presence of vWf. RAP inhibited internalization and degradation of 125 I-fVIII to a lesser degree than those of 125 I-fVIII⅐vWf complex. In addition, LRP-deficient PEA 13 fibroblasts were able to internalize and degrade isolated 125 I-fVIII. This indicates that the LRP-mediated pathway is not the sole mechanism of internalization and degradation of fVIII not complexed with vWf.
To determine whether vWf bound to fVIII is also internalized and degraded by MEF cells, we measured internalization and degradation of 125 I-labeled vWf complexed with fVIII. As shown in Fig. 6, A and B, the amounts of internalized and degraded 125 I-vWf by both MEF and PEA13 cells were less than 5% of the corresponding amounts of 125 I-fVIII catabolized from its complex with vWf under the same experimental conditions. This indicates that vWf does not follow fVIII in the LRPmediated pathway and possibly dissociates from fVIII at the early stage of endocytosis, prior to entry of the complex into endosomal compartments.
The A2 Subunit of fVIII Inhibit Endocytosis and Degradation of 125 I-fVIII/vWf by LRP-expressing Cells-Since we demonstrated that the A2 subunit of fVIII is responsible for interaction of fVIII with purified LRP in vitro, we next examined whether A2 is also involved in LRP-mediated internalization and degradation of fVIII⅐vWf complex by LRP-expressing cells. Fig. 7, A and B, demonstrate that 1000-fold excess of the A2 subunit over 125 I-fVIII⅐vWf complex effectively inhibited internalization (by Ͼ70% after 4 h) and degradation (by Ͼ60% after 4 h) of this complex. In contrast, the A1/A3-C1-C2 heterodimer, which did not inhibit fVIII interaction with purified LRP in the above experiments, did not have any effect on 125 I-fVIII endocytosis and degradation by MEF cells (Fig. 7).
Effect of RAP on the Plasma Clearance of 125 I-fVIII-To determine whether LRP is capable of catabolizing fVIII from its complex with vWf in vivo, the effect of RAP on the clearance rate of 125 I-fVIII⅐vWf complex was tested in mouse model. In the experiment, each mouse was injected with approximately 1 g of 125 I-fVIII (45) complexed with vWf in the presence or absence of RAP. The amount of administered fVIII was similar to that previously used to study fVIII clearance in mice (48,49). As shown in Fig. 8, RAP increased the half-life of 125 I-fVIII in mouse plasma. The time courses of fVIII clearance in the presence of 150 or 250 M RAP in the injected sample, suggested that RAP concentration of 150 M is saturating and its further increase does not appreciably affect the fVIII clearance. In the presence of the saturating concentration of RAP the half-life of 125 I-fVIII was prolonged by approximately by 3.3-fold. In addition, in the absence of RAP, most radioactivity was found in the liver but not in kidney, results that are consistent with LRP presence in a high abundance in hepatic tissues (31).
To determine whether 125 I-fVIII remains bound to vWf after injection into mice, some aliquotes of mice plasma taken at various time intervals (5-480 min) were also subjected to fast protein liquid chromatography as above. It was found that for the time intervals of up to 480 min, 125 I-fVIII (Ն90% of the total eluted radiactivity) was eluted as a single peak in the void volume of the column as would be expected for the fVIII⅐vWf complex, demonstrating that 125 I-fVIII remained bound to vWf in circulation in the course of clearance studies.
In the absence of RAP, clearance of fVIII was biphasic, requiring a double exponential model (Equation 1, "Experimental Procedures") to adequately fit the data. The results are consistent with the existence of fast (k 1 ϭ 0.0196 min Ϫ1 ) and slow phases (k 2 ϭ 0.00329 min Ϫ1 ) of fVIII removal from the circulation. By contrast, in the presence of a saturating concentration of RAP (250 M), the clearance could be well described by a single exponential equation with k ϭ 0.00334 min Ϫ1 , a value close to that of the slow phase in the absence of RAP. Since the slow phase of clearance was not RAP-sensitive, we then fitted all four curves to Equation 1 using three fitting parameters: C 1 , C 2 , and k 1 , and a constant value for k 2 ϭ 0.00329 min Ϫ1 . As seen from the plot of residuals (Fig. 8), the errors are randomly distributed along the time range, indicating that this model describes adequately the fVIII clearance data. The parameters derived by fitting the fVIII clearance curves are presented in Table I. The increasing concentrations of RAP progessively reduced both the value of k 1 , the kinetic rate constant corresponding to the fast phase of fVIII clearance, and C 1 , the percentage of fVIII removed by that path. In the presence of a saturating amount of RAP (250 M), k 1 decreased by 25-fold and C 1 decreased to about 6%. Thus, RAP almost completely inhibited the fast phase of fVIII clearance. DISCUSSION In the present study we demonstrated that LRP mediates the internalization and degradation of fVIII by LRP-expressing cells and contributes to fVIII clearance in vivo. This conclusion is based on several independent observations. First, we found that fVIII directly binds to purified LRP in vitro, and that this binding is competed by RAP, an antagonist of ligands binding to LRP. Second, fVIII is internalized and degraded in mouse fibroblasts expressing LRP (MEF) but not in mouse fibroblasts genetically deficient in LRP. Third, RAP effectively inhibits the cellular uptake and degradation of 125 I-fVIII from its complex with vWf by MEF cells, and fourth, clearance of 125 I-fVIII from the circulation in mice can be significantly retarded by the presence of RAP.
Based on the observation that vWf did not inhibit fVIII binding to purified LRP, we propose that LRP-mediated internalization of fVIII from its complex with vWf may be a physiological mechanism of fVIII catabolism. Indeed, fibroblast cells that express LRP, but not fibroblasts genetically deficient in LRP, were able to internalize and degrade 125 I-fVIII from its complex with vWf. These processes were also inhibited by RAP, indicating that cellular internalization and degradation were mediated by interaction of fVIII with LRP.
The physiological relevance of observations utilizing the LRP-expressing cell model system was supported by in vivo clearance studies of 125 I-fVIII⅐vWf complex in mice which demonstrated that RAP prolonged the half-life of 125 I-fVIII in the circulation by 3.3-fold, indicating that a RAP-sensitive receptor, most likely LRP, contributes to the plasma clearance of fVIII. We found that fVIII clearance is a biphasic process consisting of fast and slow components. It is essential to recognize that only the fast phase of the clearance is LRP-mediated since RAP inhibited only this phase but not the slow one. Thus, the slow phase of fVIII clearance is not LRP-mediated and the mechanism responsible for this pathway remains to be identified.
Biphasic fVIII clearance in mice, observed in the present study, is in agreement with results of a previous study in which fVIII clearance in mice was described by the sum of two exponentials (45). The biphasic nature of fVIII clearance was also demonstrated for dogs (50) and humans (17). Based on these parallels, we predict that the fast phase of fVIII clearance in humans is also LRP-mediated. One might then anticipate that inhibition of this pathway would increase the half-life of infused fVIII by severalfold. This would be especially important for the prophylactic treatment of hemophilia A patients who require frequent fVIII infusions to maintain its minimal level in circulation. This might significantly reduce the cost of the care of hemophilia A patients by reducing the frequency of infusions.
One approach to suppress LRP-mediated clearance of fVIII would be to generate a fVIII mutant in which the LRP-binding site is disrupted. We localized the 25-amino acid region 484 -509 within the A2 domain of fVIII, which is responsible for fVIII binding to purified LRP. At the initial stage, different fragments derived from fVIII and activated heterotrimeric fVIII (A1/A2/A3-C1-C2) were tested for inhibition of fVIII interaction with LRP. Since only the A2 domain and HCh, which contains A2, were able to inhibit this interaction with similar K i values, we concluded that A2 may be entirely responsible for fVIII binding to LRP. Further localization of the LRP-binding region within the A2 domain was inferred from finding that a monoclonal antibody with an epitope within residues 484 -509 completely inhibited fVIII interaction with LRP. At the same time, inhibition of fVIII/LRP binding by the synthetic peptide corresponding to residues 484 -509 also indicated that this  with vWf from plasma of mice The values of the kinetic rate constants k 1 and k 2 , corresponding to the fast and slow phases of fVIII clearance and the total percents of the radioactivity administered (C 1 and C 2 , respectively) removed during these phases were determined by fitting of the clearance data shown in Fig. 8 (47), urokinase-type/plasminogen activator inhibitor 1 complex (50), and ␣ 2 -macroglobulin (52) were previously shown to be critical for electrostatic interaction with LRP. Alanine substitution of the basic amino acid residues involved in binding to LRP in the above ligands leads to a substantial reduction of affinity for ligand binding to LRP and partial (51) or complete (52) inhibition of internalization and degradation of the mutants. Therefore, mutation of charged residues within the 484 -509 region of fVIII may be a feasible approach for generation of a recombinant fVIII with a lower rate of LRP-mediated endocytosis and a longer lifetime in circulation.
We found that internalization and degradation of isolated fVIII by LRP-presenting MEF cells was more effective than the corresponding processes for fVIII bound to vWf. Faster catabolism of fVIII in the absence of vWf is consistent with the demonstrated shorter half-life of fVIII in patients with severe von Willebrand disease lacking plasma vWf in comparison with hemophilia A patients who have normal levels of vWf (54,55). Moreover, the half-life of fVIII in von Willebrand disease patients was prolonged by the presence of vWf in the infused fVIII preparation (55). The above observations were previously explained as stabilization of fVIII via complex formation with vWf and secondary vWf-mediated release of endogenous fVIII (7,56). Our data suggest that in addition to the above effects, vWf may reduce the rate of fVIII clearance by preventing the LRP-independent pathway and limiting fVIII clearance to LRP-mediated pathway.
In the recent study related to fVIII interaction with LRP, Lenting et al. (57) made several observations consistent with our study. (i) fVIII binds to purified LRP with the affinity similar (within 2-fold) to that determined in our study. (ii) RAP completely inhibits fVIII binding to purified LRP. (iii) LRPexpressing cells degrade fVIII and this process is only partially inhibited by RAP. In addition, it was shown using the biosensor technique, that (iv) immobilized HCh interacts with the recombinant cluster II of LRP, 2 containing 8 out of 31 cysteine-rich class A repeats which are involved in LRP binding to different ligands (29). However, based on other biosensor experiments using immoblized LRP, Lenting et al. (57) concluded that the major fVIII region involved in binding to LRP is located within LCh (57), whereas in our experiments LCh failed to significantly compete with fVIII for binding to LRP. One possible way to reconcile these findings would be if conformation of the LRP-binding site within the isolated LCh is different from that within the intact fVIII. This was also suggested by Lenting et al. (57), based on their finding that the association and dissociation rate constants for fVIII interaction with LRP differ from those for LCh by 25-fold (57), which raises a possibility that the properties of the binding site localized within the isolated LCh are not identical to those within fVIII molecule. We believe, therefore, that the site localized within the A2 domain in the present study is likely to be the major site responsible for fVIII interaction with LRP. In summary, the current study demonstrates that LRP-mediated catabolism of fVIII constitutes an important pathway of fVIII clearance from circulation.