Heparin Inhibits the Binding of (cid:1) 2-glycoprotein I to Phospholipids and Promotes the Plasmin-mediated Inactivation of This Blood Protein ELUCIDATION OF THE CONSEQUENCES OF THE TWO BIOLOGICAL EVENTS IN PATIENTS WITH THE ANTI-PHOSPHOLIPID SYNDROME*

, The phospholipid-binding plasma protein (cid:1) 2-glyco-protein I ( (cid:1) 2-GPI) is the primary antigen recognized by the circulating autoantibodies in patients with the “an-ti-phospholipid syndrome” (APS). Although heparin is routinely used in the treatment and prophylaxis of APS patients, the primary heparin-binding site within (cid:1) 2-GPI has not been identified. More importantly, how heparin exerts its beneficial effects in vivo in APS patients has not been deduced at the molecular level. Using an expression/site-directed mutagenesis approach, we now show that the positively charged site that resides in the first domain of (cid:1) 2-GPI is not the primary heparin-bind-ing site. Rather it is the second positively charged site located within the fifth domain of the protein that also binds to phospholipids. Lys 284 , Lys 286 , and Lys 287 in this domain are essential for the interaction of (cid:1) 2-GPI with heparin. vitrogen). recombinant purified from the conditioned medium by affinity chromatography using a polyclonal anti- (cid:1) 2-GPI antibody. The obtained recombinant proteins were subjected to SDS-PAGE/immunoblot analysis (25, 26). Wild type and domain deletion mutants of (cid:1) 2-GPI were also subjected to automated Edman sequencing using an Applied Biosystems sequencer (Foster City, CA). A (cid:1) 2-GPI mutant (F307*/C288A) the C-terminal tail cDNA termination codon Phe Cys dimerization of recombinant oligonucleotide-directed vitro mutagenesis

The phospholipid-binding plasma protein ␤2-glycoprotein I (␤2-GPI) is the primary antigen recognized by the circulating autoantibodies in patients with the "anti-phospholipid syndrome" (APS).Although heparin is routinely used in the treatment and prophylaxis of APS patients, the primary heparin-binding site within ␤2-GPI has not been identified.More importantly, how heparin exerts its beneficial effects in vivo in APS patients has not been deduced at the molecular level.Using an expression/site-directed mutagenesis approach, we now show that the positively charged site that resides in the first domain of ␤2-GPI is not the primary heparin-binding site.Rather it is the second positively charged site located within the fifth domain of the protein that also binds to phospholipids.Lys 284 , Lys 286 , and Lys 287 in this domain are essential for the interaction of ␤2-GPI with heparin.These data indicate that ␤2-GPI binds to heparin in a relatively specific manner even though the affinity for the interaction is rather low.Lys 317 resides in the center of the high affinity phospholipid-binding site.Surprisingly, heparin at concentrations that can be achieved in vivo during anticoagulation therapy greatly enhances the plasmin-mediated cleavage of the Lys 317 -Thr 318 site in ␤2-GPI.Because the cleaved form cannot bind to phospholipids effectively, the combined actions of heparin and plasmin result in a diminished ability of ␤2-GPI to recognize phospholipids.This, in turn, decreases the prothrombotic activity of the endogenous circulating anti-␤2-GPI antibodies in the patients.Thus, heparin exerts its beneficial effects in APS patients by at least two distinct mechanisms.
␤2-glycoprotein I (␤2-GPI) 1 (also known as apolipoprotein H) is a highly glycosylated 50-kDa protein that resides in human plasma at ϳ200 g/ml (1).␤2-GPI is a member of the complement control superfamily of proteins (2).It is composed of five contiguous domains, four of which are highly homologous to one another.Each of the first four domains contains ϳ60 amino acids and two disulfide bonds.The fifth domain is aberrant in that it contains ϳ80 amino acids, a long C-terminal tail, and an extra disulfide bond (3,4).␤2-GPI has high affinity for negatively charged phospholipids.It was this very characteristic that led to its identification as the primary target antigen recognized by the autoantibodies in patients with the "antiphospholipid syndrome" (APS) (5,6).
APS is characterized clinically by recurrent venous and arterial thrombosis, miscarriages, and thrombocytopenia (7,8).A number of studies have shown that there is a significant correlation between thrombotic manifestations and the presence of anti-␤2-GPI antibodies in these patients (9 -11).The physiological function of ␤2-GPI in normal individuals remains to be determined.However, the protein has both negative and positive effects on coagulation in vitro.␤2-GPI can inhibit the contact-activation pathway of coagulation, the prothrombinase activity on activated platelets, and the ADP-mediated aggregation of platelets (12)(13)(14).␤2-GPI also can inhibit the anticoagulant effect of activated protein C (15).Much of the pro-and anticoagulant properties of ␤2-GPI are mediated via its ability to bind anionic phospholipids.Thus, ␤2-GPI competes effectively with varied coagulation factors for catalytic surfaces (16).
Using synthetic peptides, the major phospholipid-binding site within ␤2-GPI has been mapped to the Cys 281 -Cys 288 loop that resides in the fifth domain of the protein.Lys 284 , Lys 286 , and Lys 287 have been implicated in the ionic interaction of ␤2-GPI with phospholipids (17,18).However, Mehdi and coworkers (19,20) have shown that optimal interaction additionally requires the integrity of the C-terminal tail and nearby residues Leu 313 -Trp 316 .The latter residues are important because they are directly inserted into the phospholipid bilayer as an anchor.
␤2-GPI will bind to many negatively charged macromolecules, including heparin glycosaminoglycan.The primary function of heparin in normal individuals is to bind to mast cell proteases, and it probably does not interact with ␤2-GPI under physiological conditions.However, heparin is regularly used in the treatment of APS patients both in acute therapy and prophylaxis (21,22).The manner by which ␤2-GPI binds to heparin has not been deduced.A positively charged, low affinity phospholipid-binding site has been identified in the first domain of ␤2-GPI (23), thereby raising the possibility that this domain is the primary heparin-binding site.It also has not been ascertained whether or not heparin alters the biological activity of ␤2-GPI.We now demonstrate that the primary he-parin-binding site does not reside in the first domain.Rather, it resides in the fifth domain of ␤2-GPI, which is the same site that interacts with phospholipids.We also demonstrate that heparin promotes the plasmin-mediated inactivation of ␤2-GPI.Thus, heparin exerts its beneficial effects in APS patients by two distinct mechanisms.

EXPERIMENTAL PROCEDURES
␤2-GPI Preparations-Native ␤2-GPI was purified from normal human plasma using cardiolipin affinity chromatography as described previously (24).Eleven ␤2-GPI mutants were used in this study to define the heparin-binding site in ␤2-GPI.Eight of the mutants were generated previously to define the primary phospholipid-binding site and the primary antigenic site recognized by the autoantibodies in APS patients (25,26).Three additional ␤2-GPI mutants were generated to evaluate the relative importance of the first domain in ␤2-GPI/heparin interactions.Recombinant wild type and domain deletion mutants of ␤2-GPI were generated as described previously (25,26).Briefly, the cDNAs encoding native human ␤2-GPI (as well as domains I-IV, domains II-V, domains III-V, and just domain V of ␤2-GPI) were inserted into the baculovirus vector pBacPAK6 (CLONTECH Laboratories).Spodoptera frugiperda (Sf9) cells were grown in a monolayer, and then the cells were infected with the resulting constructs.The resulting cells were cultured 3-5 days at 27 °C in serum-free medium (Sf-900II; In- vitrogen).The resulting recombinant proteins were purified from the conditioned medium by affinity chromatography using a rabbit polyclonal anti-␤2-GPI antibody.The obtained recombinant proteins were subjected to SDS-PAGE/immunoblot analysis (25,26).Wild type and domain deletion mutants of ␤2-GPI were also subjected to automated Edman sequencing using an Applied Biosystems sequencer (Foster City, CA).
A ␤2-GPI mutant (F307*/C288A) lacking the C-terminal tail was prepared as described previously (27).Briefly, two mutations were introduced simultaneously into the human cDNA that ultimately resulted in a translation termination codon at Phe 307 and the conversion of Cys 288 to Ala.The latter mutation had to be performed to prevent the spontaneous dimerization of the resulting recombinant protein.An oligonucleotide-directed in vitro mutagenesis kit (CLONTECH Laboratories) was used to convert Lys 286 to Glu; Lys 286 and Lys 287 to Glu; and Lys 284 , Lys 286 , and Lys 287 to Glu in the fifth domain of the molecule (18).In the first domain the positively charged residue Arg 43 was mutated to Gly, Gly 40 was mutated to Glu, a negatively charged residue, while Thr 50 was mutated to Ala, both uncharged residues.The nucleotide sequence of each mutated cDNA was confirmed by standard sequencing technology.Each cDNA was subcloned into the pBacPAK-6 vector, and the recombinant protein was expressed and purified as outlined above.
Binding of Native and Mutated Forms of ␤2-GPI to Immobilized Heparin and Cardiolipin-A heparin-BSA complex was prepared by chemically coupling the heparin chains via their reducing ends to BSA according to the recommendations of the manufacturer (Sigma).Binding of ␤2-GPI to immobilized heparin was performed using a modification of the method of Najjam and co-workers (28).Nunc Maxisorp plates were incubated overnight at 4 °C with heparin-BSA complex or BSA (Sigma) alone at a concentration of 5.0 g/ml in 13 mM EDTA and 50 mM Tris-HCl, pH 7.4.Each plate was washed four times in phosphatebuffered saline, pH 7.2, containing 0.05% Tween (v/v).Wild type, domain deletion mutants, and point mutants of ␤2-GPI were added at equimolar concentrations ranging from 0.2-1.0M, and each plate was incubated for an additional 2 h at room temperature.Each plate was washed, 100 l of rabbit polyclonal antibody to ␤2-GPI was added at a concentration of 1 g/ml, and the plate was incubated for another 2 h at room temperature.The plate was then washed, and 100 l of alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma) was added at a dilution of 1:1000.After a 1-h incubation at room temperature, the plate was washed again, and 100 l of para-nitrophenyl phosphate (1 mg/ml) in 1 M diethanolamine buffer, 0.5 mM MgCl 2 , pH 9.8 (Sigma) was added to each well.The optical density was read at 405 nm using a Power-Wave X Select (Bio-Tek Instruments Inc.).The absorbance values of the BSA control wells were similar to background levels.Thus, these values were subtracted from that of the heparin-BSA wells in all cases as a measure of specific binding to heparin.Binding of native ␤2-GPI and its mutants to cardiolipin was determined using an anti-cardiolipin antibody enzyme-linked immunosorbent assay that has been described previously (17).
Competitive Inhibition of ␤2-GPI Binding to Heparin Using a Synthetic Peptide and Determination of the Binding Affinity of Wild Type ␤2-GPI for Immobilized Heparin-Wild type ␤2-GPI was incubated in the presence of increasing concentrations (25-200 M) of a synthetic peptide, 281 CKNKEKKC 288 , spanning the phospholipid-binding region on the fifth domain of the molecule or the control peptide, a scrambled peptide containing this region of the fifth domain.Binding of wild type ␤2-GPI to immobilized heparin in the presence of these peptides was then determined using the protocol described above.The level of inhibition of wild type ␤2-GPI binding to immobilized heparin was determined in the presence of increasing concentrations (1-20 g/ml) of heparin in fluid phase or haptoglobin (as an irrelevant control) using the protocol described above.Haptoglobin was chosen as a control because it is a member of the complement control family of proteins similar to ␤2-GPI.The binding affinity of fluid phase ␤2-GPI for heparin was calculated according to the method of Friguet and co-workers (29).
Evaluation of the Influence of Heparin on the Plasmin-mediated Cleavage of ␤2-GPI-Preliminary experiments were performed to determine the concentration of plasmin required to generate approximately one-third cleaved and two-thirds intact ␤2-GPI using a variation of the method of Okhura and co-workers (30).This was achieved when ␤2-GPI and plasmin were incubated at a molar ratio of ϳ90:1 for 30 min at 37 °C; therefore these conditions were used in all subsequent experiments.The reaction was terminated by snap freezing the digests in dry ice.The samples were diluted 10-fold in 10 mM Tris-HCl buffer, pH 8.0, and applied to a HiTrap heparin column (Sigma) equilibrated with 0.1 M NaCl and 10 mM Tris-HCl, pH 8.0.Both intact and cleaved ␤2-GPI was eluted from the heparin column using a linear gradient from 0.1 to 1.0 M NaCl in Tris-HCl buffer, pH 8.0.Cleaved ␤2-GPI bound the heparin column with a lower affinity than the intact protein and therefore eluted earlier from the column at a lower salt concentration.
To examine the effect of heparin on the plasmin-mediated cleavage of ␤2-GPI, heparin (0.1-10 units/ml) was added to ␤2-GPI and plasmin at a constant molar ratio, and the resulting sample was incubated for 30 min at 37 °C.The reaction was terminated by snap freezing in dry ice.The samples were diluted 10-fold in 10 mM Tris-HCl buffer, pH 8.0, and applied to the heparin column as described above.Aliquots from the reaction mixture were removed at 0-and 30-min time intervals and subjected to SDS-PAGE under reducing conditions on an 8 -16% gradient gel (Novex ® , Invitrogen BV).Proteins were visualized by Coomassie Blue staining of the gels.

Binding of Recombinant and Domain Deletion Mutants of ␤2-GPI to Heparin-A ␤2-GPI mutant lacking domain I bound
to heparin in a manner similar to that obtained with native ␤2-GPI.Although the mutant that lacked both domains I and II bound at a reduced efficiency, the mutant that lacked just domain V failed to bind to heparin at any examined concentration (Fig. 1A).When a mutant that contained just the fifth domain of ␤2-GPI was examined, it was found that this small portion of the protein bound to heparin at a rate of 22-48% of that obtained with wild type ␤2-GPI depending on the concentration of the mutant used in the assay.These findings indicate that ␤2-GPI binds to heparin primarily via its fifth domain.␤2-GPI cleaved at its Lys 317 -Thr 318 site did not bind to heparin at either 10 or 25 g/ml.However, a ␤2-GPI mutant lacking the internal loop in the fifth domain of the molecule (F307*/C288A) bound heparin in a manner similar to that obtained with wild type ␤2-GPI.The latter finding suggests that the internal loop that is normally inserted into the lipid bilayer is not involved in the recognition of heparin (Fig. 1B).Neither the ␤2-GPI mutant lacking the C-terminal tail nor the proteolytically processed ␤2-GPI is able to recognize cardiolipin (Fig. 1C).
Binding of Recombinant and ␤2-GPI Mutants with Point Mutations in the First and Fifth Domains to Heparin-␤2-GPI mutants possessing point mutations in the first domain of this plasma protein bound to heparin effectively (Fig. 2A).Thus, in agreement with the above domain deletion mutants, the first domain of ␤2-GPI does not appear to be essential for the interaction of this plasma protein with heparin.Compared with wild type ␤2-GPI, a single point mutation at Lys 286 in the fifth domain of the molecule resulted in a 69% reduction in the ability of the recombinant protein to bind to heparin.Binding efficiency was reduced to 80% when both Lys 286 and Lys 287 were mutated, and no binding to heparin occurred when Lys 284 , Lys 286 , and Lys 287 were all mutated.Thus, all three positively charged residues in domain V appear to be involved at varying degrees in the interaction of ␤2-GPI with heparin (Fig. 2B).
Competitive Inhibition of the Binding of ␤2-GPI to Heparin and Determination of the Affinity of the Interaction-Binding of wild type ␤2-GPI to heparin was competitively inhibited using the 281 CKNKEKKC 288 synthetic peptide that corresponds to a portion of the fifth domain that binds to phospholipids.A 50% inhibition was obtained at a peptide concentration of 150 M (Fig. 3).The binding of ␤2-GPI to heparin also was competitively inhibited with fluid phase heparin, but not haptoglobin, at equimolar concentrations (Fig. 4).An IC 50 was obtained at 20 g of heparin/ml.Based on these data, the dissociation constant K d for the fluid phase interaction of ␤2-GPI and heparin was estimated to be 4.3 ϫ 10 Ϫ5 M.
Effect of Heparin on the Plasmin-mediated Cleavage of ␤2-GPI-Incubation of ␤2-GPI with heparin (10 units) in the absence of plasmin did not induce cleavage of ␤2-GPI above the control nor did it promote the dissociation of ␤2-GPI from the heparin column (Fig. 5, A and B).Interestingly heparin, at concentrations from 2.5 to 10 units/ml, greatly enhanced plasmin cleavage of ␤2-GPI as demonstrated by an increase in the peak area in the chromatograph representing cleaved ␤2-GPI compared with intact ␤2-GPI above the control (Fig. 5, C and  E).The peak area representing cleaved and intact ␤2-GPI was 10.4 and 16.5 cm 2 , respectively, for ␤2-GPI incubated with plasmin alone.However in the presence of heparin (10 units) there was a complete reversal in the chromatographic profile as demonstrated by an increase in peak area for cleaved ␤2-GPI to 15 cm 2 with a concomitant decrease for that representing intact ␤2-GPI to 12.1 cm 2 .At lower concentrations of heparin there was no increase in the level of cleavage by plasmin detected above the control without heparin.
SDS-PAGE analysis of ␤2-GPI incubated with plasmin alone showed a slight increase in mobility for ␤2-GPI cleaved at the Lys 317 -Thr 318 site compared with intact ␤2-GPI (Fig. 6), while a similar banding pattern was observed when ␤2-GPI was incubated with plasmin in the presence of heparin suggesting that no further cleavage product was generated.FIG. 5. HiTrap heparin chromatography of ␤2-GPI incubated with plasmin in the presence or absence of heparin.␤2-GPI was incubated for 30 min at 37 °C in the absence of plasmin and heparin (A), in the presence of heparin (10 units) alone (B), in the presence of plasmin alone at a molar ratio of substrate to plasmin of 90:1 (C), or with 5 (D) and 10 (E) units of heparin.The peaks that correspond to cleaved (C) and intact (I) ␤2-GPI are indicated.These results are representative of three independent experiments.The peak areas for cleaved and intact ␤2-GPI are 10.4 and 16.5 cm 2 in panel C, 12.7 and 11.7 cm 2 in panel D, and 15.0 and 12.1 cm 2 in panel E, respectively.ABS, absorbance.

DISCUSSION
␤2-GPI will bind to a number of negatively charged surfaces, including those that contain phospholipids, DNA, or heparin (16,31,32).␤2-GPI is abundant in plasma, and this protein is highly conserved throughout mammalian evolution (33).Nev-ertheless, its physiological function still has not been deduced.Based on numerous in vitro studies, ␤2-GPI has both positive and negative effects on coagulation.It enhances the uptake of apoptotic bodies, and it binds oxidized low density lipoproteins (12-15, 34, 35).The heparin binding property of ␤2-GPI was first noted in 1979.Although this property is routinely exploited to purify ␤2-GPI from human plasma (32), the heparinbinding site in ␤2-GPI has not been defined.Whether or not heparin alters ␤2-GPI in some way also has not been deduced.
The first domain within ␤2-GPI contains a positively charged region that includes Arg 39 , Arg 43 , and Lys 44 .Because it has been reported that this domain binds negatively charged phospholipids, an expression/site-directed mutagenesis approach was used to evaluate the possible importance of this positively charged domain in heparin interaction.Conversion of Arg 43 to Gly did not dramatically alter the ability of ␤2-GPI to bind to heparin.Furthermore a synthetic peptide spanning residues Val 37 -Pro 48 in domain I failed to inhibit the binding of wild type ␤2-GPI to heparin at any concentration examined despite its overall positive charge (data not shown).Based on these data, we conclude that the first domain in ␤2-GPI is not the primary heparin-binding site.Using various mutants of ␤2-GPI, we instead found that ␤2-GPI binds to heparin primarily via its fifth domain.We also show that residues at Lys 284 , Lys 286 , and Lys 287 in this domain are of critical importance in the interaction.␤2-GPI uses the same three Lys residues in its interaction with phospholipids (18).However, high affinity interaction only occurs when the hydrophobic segment corresponding to residues Leu 313 -Trp 316 are inserted into the phospholipid bilayer as shown in the model depicted in Fig. 7. ␤2-GPI that had been cleaved at the Lys 317 -Thr 318 site was unable to bind to immobilized heparin.While this finding initially suggested that the hydrophobicity of the C-terminal loop is also important in heparin recognition, a separate mutant that lacked the hydrophobic loop bound to heparin just as efficiently as wild type ␤2-GPI.The latter finding now suggests FIG. 6. SDS-PAGE analysis of ␤2-GPI digests.Aliquots (5 g) of the reaction mixture were taken at time 0 (lanes 3 and 5) and 30 min (lanes 4 and 6) in the absence (lanes 3 and 4) and presence (lanes 5 and 6) of heparin (10 units), respectively.The resulting digests were subject to SDS-PAGE analysis using an 8 -16% gradient gel under reducing conditions.Shown is the resulting Coomassie Blue-stained gel.Shown in lane 1 is ␤2-GPI that had been incubated in the absence of plasmin and heparin.Shown in lanes 2 and 7 are plasmin (5 g) incubated in the absence and presence of heparin, respectively.The plasmin-derived bands seen in lanes 2 and 7 are not seen in lanes 3-6 because considerably less plasmin is present in lanes 3-6.Shown on the right in lane 8 are molecular mass markers (Sigma).FIG. 7. Schematic representation of the interaction of ␤2-GPI with phospholipids and heparin in the presence and absence of plasmin.␤2-GPI normally binds to phospholipids via domain V (A).However, during heparin therapy, ␤2-GPI also binds to heparin via domain V (B).Thus, heparin functions in the first mechanism as a competitive inhibitor for the phospholipid-binding site.If plasmin is also present, the peptide bond linking residues 317 and 318 is cleaved resulting in inactivation of domain V (C).Heparin promotes the plasmin-mediated proteolytic event.Thus, heparin functions in the second mechanism as a catalytic cofactor in the processing event.
that cleavage of ␤2-GPI at the Lys 317 -Thr 318 site alters the spatial array of the three critical Lys residues.In support of this conclusion, a recent report proposed that cleavage of ␤2-GPI at the Lys 317 -Thr 318 site disturbs the nearby electrostatic environment (36).It is believed that the integrity of the 317/ 318 peptide bond is important in tethering the cluster of positively charged and hydrophobic residues in this region (36).While McNally and co-workers (37) proposed that heparin might enhance the binding of ␤2-GPI to phospholipids, we conclude that the opposite most likely occurs in vivo because the heparin and phospholipid recognition sites are the same.
Horbach and colleagues (38) reported that ␤2-GPI is often cleaved at the Lys 317 -Thr 318 site in the plasma of patients with disseminated intravascular coagulation and patients treated with streptokinase.The cleaved form of ␤2-GPI was correlated with varied in vitro markers of fibrinolytic activity.Thus, it has been proposed that the cleaved form of ␤2-GPI is probably generated during normal fibrinolysis.More recently, the presence of the processed form of ␤2-GPI was demonstrated in patients with leukemia and with lupus anticoagulant activity (39).It was proposed that plasmin is the likely protease that generates the proteolytically processed form of ␤2-GPI.Unexplained was the mechanism that restricts the plasmin-mediated attack to a single site in the protein.Thus, we decided to investigate whether or not heparin alters and/or restricts the plasmin-mediated cleavage of ␤2-GPI.We initially suspected that heparin would sterically inhibit the plasmin-mediated cleavage of ␤2-GPI because the heparin-binding site is close to the primary plasmin-susceptible site.However, in contrast to what we expected, heparin at concentrations that are reached therapeutically in vivo greatly enhanced the plasmin-mediated cleavage of ␤2-GPI.In addition, the proteolytically processed ␤2-GPI was no longer able to bind to heparin.There are many potential plasmin cleavage sites in ␤2-GPI.Thus, the ability of heparin to enhance attack at the Lys 317 -Thr 318 site is quite remarkable.Considering that the cleaved form of ␤2-GPI cannot bind to phospholipids, interaction with heparin should greatly reduce the prothrombotic effects of anti-␤2-GPI antibodies.Finally, it has been proposed that the proteolytically processed form of ␤2-GPI may be cleared more rapidly from the circulation than native ␤2-GPI (38).
In summary, our data indicate at least two distinct mechanisms by which heparin achieves its therapeutic effect in APS patients.In the first mechanism, heparin sterically prevents the binding of ␤2-GPI to negatively charged phospholipids, which in turn prevents the deposition of the anti-␤2-GPI antibodies in tissues.In the second mechanism, heparin potentiates the generation of an inactive form of ␤2-GPI by plasmin.

FIG. 1 .FIG. 2 .FIG. 3 .
FIG. 1. Binding of wild type ␤2-GPI, varied domain deletion mutants of ␤2-GPI, a proteolytically processed form of ␤2-GPI, and a mutant of ␤2-GPI lacking the C-terminal loop to immobilized heparin and cardiolipin.Wild type (WT) ␤2-GPI (A, B, and C), domain deletion mutants of ␤2-GPI (A), and a proteolytically processed form of ␤2-GPI and a mutant of ␤2-GPI lacking the C-terminal loop (B and C) were evaluated for their ability to bind to immobilized heparin (A and B) and cardiolipin (C) at the indicated concentrations.The absorbance readings for wells coated with BSA alone were similar to background and subtracted from wells coated with heparin-BSA complex in the experiments depicted in panels A and B. Similar findings were obtained in a second experiment.ABS, absorbance.

FIG. 4 .
FIG. 4. Competitive inhibition of wild type ␤2-GPI binding to immobilized heparin with fluid phase heparin and haptoglobin.Inhibition of wild type ␤2-GPI binding to immobilized heparin with fluid phase heparin (Ⅺ) and haptoglobin (ࡗ) was examined at concentrations ranging from 0.6 to 20 g/ml.ABS, absorbance.