Expression and characterization of recombinant protein S with the Ser 460 Pro mutation

To characterize the putative biochemical modifications induced by the Ser 460 to Pro (Heerlen) mutation in protein S (PS), we expressed both wild-type (wt) and mutated recombinant PS in HEK cells. In SDS-polyacrylamide gels, r-PS Heerlen migrated at 71 kDa whereas r-wt PS migrated at 73 kDa, a difference abolished after deglycosylation by N-glycosidase, suggesting that the Ser 460 Pro mutation abolishes N-glycosylation of Asn 458. The affinity of r-wt PS and r-PS Heerlen for C4b-binding protein (C4b-BP) and for phospholipid vesicles was similar. Neither the enhancement of APC-dependent prolongation of the APTT, nor the specific enhancement of FVa and FVIIIa proteolysis by APC in purified systems was affected by the mutation. However, the Ser 460 Pro mutation induced a slight conformational change in the SHBG domain of the PS molecule, as shown by reduced binding affinity for monoclonal antibodies. The type III phenotype associated with the Heerlen mutation might thus result from a slightly modified rate of synthesis or catabolism. The resulting moderate decrease in the circulating PS concentration may modify the equilibrium between free PS and C4b-BP/PS complexes.     

Monoclonal antibodies directed to the calcium-free conformation of human protein S

Four mouse hybridomas secreting monoclonal antibodies specific for human protein S (PS) have been generated. The antibodies, all of the IgG1 subclass, were designated S2, S3, S8, and S10. In a fluid phase radioimmunoassay, the binding of monoclonal antibodies to PS was about 30% greater in the presence of EDTA and totally inhibited in presence of Ca2+. Using the same technique, we performed displacement curves of 125I-labeled PS by purified PS, thrombin-cleaved PS, normal plasma, plasma from a patient on warfarin therapy, and plasma from a patient with no free PS and only PS bound to C4b-binding protein. The slopes of the curves show that the monoclonal antibodies reacted equally with all the tested forms of PS indicating that the antigenic site(s) to which the monoclonal antibodies are directed are present and exposed in free and bound PS, in thrombin-cleaved PS, and in the coumarin form of the protein. Each EDTA-dependent antibody, immobilized on Sepharose 4B-CNBr was used to purify PS from the barium citrate-absorbed, ammonium sulphate-soluble fraction of plasma. The fraction eluted from the immunoabsorbent with a buffer containing 4 mmol/l CaCl2 and analysed by SDS-PAGE, contained two bands, one migrating with conventionally purified PS and the other with purified C4b-binding protein. Homogeneous PS was obtained by chromatography of the barium citrate adsorbate on a DEAE-Sephadex column. The protein peak containing the bulk of PS was subsequently applied to the immunoadsorbent and eluted with 4 mmol/l CaCl2.(ABSTRACT TRUNCATED AT 250 WORDS)     

Determination of total, free and complexed protein s in plasma by ELISA, and comparison with a standard electroimmunoassay

An enzyme-linked immunosorbent assay (ELISA) for measuring total, free and complexed protein S in plasma was developed. To assay free protein S, C4b-binding protein-bound protein S (C4b-BP-PS) was extracted by addition of polyethyleneglycol (PEG) 6000 (5%, final concentration) to plasma samples. Microtiter plates were coated with rabbit anti-human protein S, and bound protein S was detected with labelled anti-protein S antibody. Diluted plasma samples were incubated in the plates overnight at 22 °C to permit C4b-BP-PS complexes to dissociate. Mean variation coefficients of 2.1 and 3.2% (intra-assay) and 4.3 and 7.9% (inter-assay) were found for total and free protein S assays, respectively. The ELISA measures free and complexed protein S with equal efficiency as is demonstrated by the fact that the sum of free protein S and C4b-BP-PS complex levels in normal individuals, women in their third trimester of gestation and patients with acute deep vein thrombosis (DVT), equaled the level of total protein S present in the corresponding plasma. Total protein S values obtained with the ELISA, in all groups studied, correlated well with those obtained with a standard electroimmunoassay (EIA) (r=0.93; N=40). However, total protein S levels measured by EIA were lower than those assayed by ELISA in pregnant women and in DVT patients. Furthermore, addition of several amounts of purified C4b-BP to NHP, which reduced the recovery of free protein S, did not influence the total protein S values measured by ELISA but slightly decreased the recovery of total protein S measured by EIA. These results indicate the necessity of using assays which accurately and reliably measure the total amount of protein S antigen. After addition of C4b-BP to NHP, the residual functional protein S level was lower than the residual level of free protein S antigen; this lends support to the idea that C4b-BP-PS complex inhibits the activated protein C cofactor activity of protein S.     

Multicentre evaluation of IL Test Free PS: a fully automated assay to quantify free protein S

Deficiency of the anticoagulant vitamin K-dependent protein S (PS) is associated with increased risk of venous thrombosis. In human plasma, PS circulates in two forms: as free protein (free PS) and PS bound to C4b-binding protein (C4BP), a regulator of the complement system. Assays for free PS have higher sensitivity and specificity for protein S deficiency than assays for total protein S. We have extensively evaluated the analytical performance of a novel assay for free PS, the IL Test Free Protein S, which takes advantage of the affinity of C4BP for free PS, and compared its performance to existing methods. IL Test Free Protein S is a rapid, fully automated turbidimetric assay consisting of two reagents: a C4BP coated latex and an anti-PS monoclonal antibody coated latex. The test range, precision and linearity were adequate and the assay tolerated high concentrations of interfering substances of clinical significance. The reference range agreed with previously published studies. The analysis of 903 patient samples belonging to 20 different clinical categories with the new assay yielded free PS results that agreed well with those obtained using the assays established in the participating laboratories. The study demonstrated the IL Test Free Protein S to be rapid, reliable and easy to perform.     

Four missense mutations identified in the protein S gene of thrombosis patients with protein S deficiency: effects on secretion and anticoagulant activity of protein S

Four missense mutations, G54R, T589I, K155E, and Y595C, were identified in the protein S (PS) gene of the patients with PS deficiency and venous thrombosis. Three patients were heterozygous for the novel mutations, G54R, T589I, and Y595C, while a remaining one patient was homozygous for the K155E mutation, which is known to be a polymorphism in the Japanese population. A family study revealed that the Y595C mutation was associated with a Type I PS deficiency and the K155E mutation with a Type II PS deficiency, while no family study was performed for the patients with the G54R and T589I mutations. To determine whether these four mutations play a causative role in PS deficiency, the four PS mutants and wild-type PS were stably expressed in human embryo kidney (HEK) 293 cells. Pulse-chase experiments showed intracellular degradation and decreased secretion of the Y595C mutant. In the activated protein C (APC) cofactor assays, the specific activity of the K155E mutant decreased to 58% of that of wild-type PS. The APC cofactor activity of the three mutants, G54R, K155E, and T589I, were inhibited by C4b-binding protein (C4BP) with a dose dependency similar to that of wild-type PS. These results indicate that the Y595C and the K155E mutations are responsible for a secretion defect and a decreased anticoagulant activity of PS, respectively. The remaining two mutations, G54R and T589I, however, did not produce any definite abnormality leading to a low plasma PS activity.     

Multicenter evaluation of three commercial methods for measuring protein S antigen.

This collaborative study was designed to assess the performance of commercial methods for protein S (PS) antigen measurement. Twenty-five different samples were distributed deep-frozen (24 plasmas) or lyophilized (one plasma) to five laboratories. They were analyzed blind in each laboratory by the method used locally and by three commercial methods which included two electroimmunoassays (EIA), Asseraplate-PS (Diagnostica Stago), Rellplate-S (American Diagnostica) and an ELISA system, Asserachrom-PS (Diagnostica Stago). 1. Reproducibility. Average between-laboratory coefficients of variation were 15.4%, 17.6% and 25.3% for Asserachrom-PS, Asseraplate-PS and Rellplate-S. 2. Specificity. Results of all methods showed that PS is underestimated when C4b binding protein is high. This influence was particularly evident for the ELISA Asserachrom-PS and disappeared when the antibody-antigen incubation period was prolonged to overnight. 3. Sensitivity. In all laboratories ELISA detected even the lowest PS concentration (4 U/dl), whereas the two EIAs were less sensitive (lower detection limit 14 U/dl). All methods and laboratories correctly diagnosed a plasma sample from a PS congenitally deficient patient. Conclusions. This study shows that better standardization of PS immunoassays is necessary to improve accuracy and reduce interlaboratory variability before a candidate plasma standard can be successfully calibrated in an international collaborative study.     

Combined inherited protein S and heparin co-factor II deficiency in a patient with upper limb thrombosis: a family study.

A 42-year-old Italian woman presenting with spontaneous deep vein thrombosis of the right arm, was found to have inherited a deficiency of both protein S (PS) and heparin co-factor II (HC II). The two defects seemed to segregate independently, since her son exhibited only a HC II deficiency while one of her sisters manifested only the PS defect. All affected patients appeared heterozygous for one or other or both deficiency states. The proposita and her sister exhibited a congenital PS deficiency consisting of normal or near normal levels of total PS antigen and C4b-binding protein (C4b-BP) but a moderate reduction both of free PS antigen and of PS functional activity. In addition, the proposita and her son had half normal levels of HC II antigen and activity. Except for the proposita, all were asymptomatic. Inherited deficiencies either of PS or of HC II have been associated with thrombotic manifestations. Since the proposita had an inherited combined defect of the two proteins, severe thrombotic events might be expected. However, this was not found to be the case. The role of HC II deficiency in the pathogenesis of thrombosis whether alone or combined remains to be fully investigated.     

The anticoagulant properties of a modified form of protein S

Protein S (PS) is a vitamin K-dependent anticoagulant that acts as a cofactor to activated protein C (APC). To date PS has not been shown to possess anticoagulant activity in the absence of APC. In this study, we have developed monoclonal antibody to protein S and used to purify the protein to homogeneity from plasma. Affinity purified protein S (PSM), although identical to the conventionally purified protein as judged by SDS-PAGE, had significant anticoagulant activity in the absence of APC when measured in a factor Xa recalcification time. Using SDS-PAGE we have demonstrated that prothrombin cleavage by factor Xa was inhibited in the presence of PSM. Kinetic analysis of the reaction revealed that PSM competitively inhibited factor Xa mediated cleavage of prothrombin. PS preincubated with the monoclonal antibody, acquired similar anticoagulant properties. These results suggest that the interaction of the monoclonal antibody with PS results in an alteration in the protein exposing sites that mediate the observed anticoagulant effect. Support that the protein was altered was derived from the observation that PSM was eight fold more sensitive to cleavage by thrombin and human neutrophil elastase than conventionally purified protein S. These observations suggest that PS can be modified in vitro to a protein with APC-independent anticoagulant activity and raise the possibility that a similar alteration could occur in vivo through the binding protein S to a cellular or plasma protein.     

Membrane binding and anticoagulant properties of protein S natural variants

Introduction

Protein S (PS) is a vitamin K-dependent plasma glycoprotein with a key role in the control of coagulation pathway on phospholipid membranes. We compared anticoagulant and membrane binding properties of PS altered by natural mutations (N217S, DelI203D204) affecting the epidermal growth factor like-domain 4 (EGF4) and causing PS deficiency.

Materials and methods

Binding of recombinant, immunopurified PS (rPS) to several conformation-specific antibodies, to C4BP and to phospholipid liposomes was investigated by ELISA. PS binding to cells was analysed by flow cytometry. PS inhibitory activities were studied in plasma and purified systems.

Results and conclusions

Conformational changes produced by mutations were revealed by mapping with calcium-dependent antibodies. The immunopurified recombinant mutants (rPS) showed at 200-800nM concentration reduced inhibition of coagulation (rPS217S, 10.2-17.3%; rPSDelI203D204, 5.8-8.9% of rPSwt) in FXa 1-stage clotting assay with APC. In thrombin generation assays the inhibition of ETP was reduced to 51.6% (rPS217S) and 24.1% (rPSDelI203D204) of rPSwt. A slightly shortened lag time (minutes) was also observed (rPS217S, 2.58; rPSDelI203D204, 2.33; rPSwt, 3.17; PS deficient plasma, 2.17).In flow cytometry analysis both mutants efficiently bound apoptotic cells in adhesion or in suspension. The affinity for phosphatidylserine-rich vesicles (apparent Kd: rPSwt 27.7 ± 1.6 nM, rPS217S 146.0 ± 16.1 nM and rPSDelI203D204 234.1 ± 28.1 nM) was substantially increased by membrane oxidation (10.9 ± 0.6, 38.2 ± 3.5 and 81.4 ± 6.0 nM), which resulted in a virtually normal binding capacity of mutants at physiological PS concentration.These properties help to define the molecular bases of PS deficiency, and provide further elements for PS-mediated bridging of coagulation and inflammation.     

Behavior of protein S during long-term oral anticoagulant therapy

It has recently been reported that a natural anticoagulant, protein S (PS), is depressed during oral anticoagulation. Since more detailed information is required from the clinical standpoint, we measured plasma levels of PS [both total and free (not complexed) PS antigen], C4b-binding protein (C4bp) and other vitamin K-dependent proteins (factors II, VII, IX, X and protein C) in 60 plasma samples from patients on long-term oral anticoagulant therapy with warfarin. Together with the reduction of other vitamin K-dependent plasma proteins, PS decreased during warfarin treatment, being dependent on the intensity of the therapy. A considerable variation in plasma PS levels was also observed among individuals with a similar intensity of anticoagulation. Plasma concentration of C4bp was closely correlated with total PS level, and free PS/total PS ratio was independent of thrombotest values. These findings indicate that long-term oral anticoagulant therapy results in the suppression of the synthesis of PS, and that its reduction is on the whole balanced with C4bp and vitamin K-dependent coagulation factors. It was suggested that the metabolism of C4bp might be regulated by the plasma PS level, although this hypothesis needs further exploration.     

Analytical considerations for free protein S assays in protein S deficiency

Protein S is an anticoagulant protein that circulates in plasma in complex with C4b-binding protein (C4BP) or in free form. Deficiency of protein S increases the risk of venous thrombosis. Measurement of free protein S, as compared to total levels, has been shown to be superior for prediction of protein S deficiency. We studied the effects of different handling protocols for an immuno- and a ligand (C4BP)-based assay for free protein S. When the assay was performed at 37 degrees C, the levels of free protein S in plasma from protein S deficient patients were approximately twice those obtained at room temperature. The reason for this phenomenon was that plasmas from protein S deficient patients exhibited a time-, temperature-, and dilution-dependent increase in free protein S, which was more pronounced than corresponding dilution of the normal plasma that was used to create the standard curve. These findings demonstrate the importance of assay procedure and sample handling in assays for free protein S.     

A case report of deficiency in an inhibitor of calcium-dependent association of protein S with C4B-binding protein suggested by a modified crossed immunoelectrophoresis

We have experienced a coagulation factor VIII-deficient patient whose plasma has normal protein S (PS) activity and masses of free PS and its bound form in complex with C4b-binding protein (C4BP). Although the patient's plasma showed a normal ratio of free PS to PS-C4BP complex in the presence of 5 mM EDTA, the plasma gave an abnormally retarding major C4BP peak together with a major PS peak in the crossed immunoelectrophoresis (CIE) in the presence of 2 mM CaCl₂. It was revealed that the major peak was formed by a mixture of PS-C4BP complex and free form. The addition of normal human plasma (NHP) to the patient's plasma inhibited the retardation of the major PS-C4BP complex. These suggest that the patient's plasma lacks some component(s) to inhibit Ca²⁺-dependent association of PS with C4BP.     

Immunoradiometric assay for the calcium-stabilized conformation of human protein S

Rabbit polyclonal anti-protein S serum was fractionated with immobilized human protein S to establish solid-phase immunoradiometric assays recognizing Ca(II)-dependent and NonCa(II)-dependent epitopes of human protein S. The two assays were specific for PS:Ca(II)Ag and PS:NonCa(II)Ag and highly sensitive with a lower limit of detection of about 2.5 ng/ml. PS:Ca(II)Ag and PS:NonCa(II)Ag levels were measured in immunopurified protein S, thrombin-modified protein S and chymotrypsin-cleaved protein S. Only in chymotrypsin-cleaved protein S an important discrepancy between the two antigen levels was observed. Ranges for the concentration of PS:Ca(II)Ag and PS:Non-Ca(II)Ag and their ratio were established in plasma of healthy individuals (0.92 +/- 0.13 U/ml, 0.98 +/- 0.21 U/ml, 0.96 +/- 0.17, respectively). In a group of patients using oral anticoagulant therapy the ratio PS:Ca(II)Ag/PS:NonCa(II)Ag decreased at increasing intensity of anticoagulation suggesting the presence of sub- and noncarboxylated protein S molecules. In plasma of patients with a hereditary type I protein S deficiency PS:Ca(II)Ag and PS:NonCa(II)Ag were reduced to the same extent: mean ratio 1.02 +/- 0.12 in the group not on oral anticoagulant treatment and 0.94 +/- 0.10 in the group on oral anticoagulant therapy. Analysis of patients with a history of unexplained thrombo-embolic disease did not reveal individual patients with a PS:Ca(II)Ag/PS:NonCa(II)Ag ration below the lower limit of the normal range (mean ratio 1.05 +/- 0.17), suggesting that the frequency of genetic protein S variants with defects in the Ca(II)-stabilized conformation is very low.     

Acquired activated protein C resistance is associated with IgG antibodies to protein S in patients with systemic lupus erythematosus

The objective of this study was to clarify the roles of anti-phospholipid antibodies (aPLs) in the pathogenesis of acquired activated protein C resistance (APC-R) in patients with systemic lupus erythematosus (SLE). We examined several aPLs levels (lupus anticoagulant, anti-cardiolipin antibodies, anti-β2-glycoprotein I antibodies, anti-protein C antibodies, and anti-protein S antibodies), the APC-R test, and the factor V Leiden test in 85 SLE patients. Acquired APC-R, which was not found in any patient with the factor V Leiden mutation, was present in 26 (30.6%) of 85 patients, and confirmed that acquired APC-R was a significant risk factor for thromboembolic complications [odd ratio (OR), 3.36; 95% confidence interval (CI), 1.24-9.11]. Multivariate logistic analysis revealed that both LA and anti-PS strongly associated with the presence of APC-R, and that the correlation between anti-PS and APC-R was much stronger (OR, 46.7; 95%CI, 6.99-311) than that between LA and APC-R (OR, 11.3; 95%CI, 2.26-57.0). Furthermore, the mean value of APC sensitivity ratios was significantly lower in SLE patients with anti-PS (mean ± SD, 1.68 ± 0.37, p < 0.0001) than in those without anti-PS (2.23 ± 0.40). These results suggest that acquired APC-R is most strongly attributable to functional interference of the APC pathway by anti-PS, which contribute to risk of thromboembolic complications.     

Identification of different forms of human C4b-binding protein lacking beta-chain and protein S binding ability

Human C4b-binding protein (C4BP) is a multimeric regulatory component of the complement system that circulates in plasma either as a free protein or in a noncovalent complex with the vitamin K-dependent protein S. The major form of C4BP is composed of seven identical alpha-chains (70 kDa) and one beta-chain (45 kDa). C4BP was purified from human plasma after barium citrate adsorption using anti-C4BP monoclonal antibody affinity chromatography. C4BP-high and low Mr forms were both obtained from the barium citrate precipitate and the supernatant. C4BP-high and low forms from the barium citrate precipitate were separated by sodium dodecylsulfate polyacrylamide slab gel electrophoresis and extracted with Triton X-100. Both forms contained the beta-chain as was demonstrated on sodium dodecylsulfate polyacrylamide slab gel electrophoresis under reduced conditions after silver-staining and with Western-blotting using monoclonal antibodies specific for the beta-chain. The C4BP-high and low forms demonstrated similar protein S binding affinity (KA: 3.18 x 10(8) and 3.21 x 10(8) M-1, respectively) in a C4BP-protein S binding assay and a protein S ligand blot using a peroxidase-conjugated monoclonal anti-protein S antibody. The barium citrate supernatant contained two forms of C4BP-high and one form of C4BP-low. One form of C4BP-high did contain the beta-chain and was capable of protein S binding (KA: 4.35 x 10(8) M-1). The two other forms of C4BP lacked the beta-chain and were unable to bind protein S.     

Determination of plasma protein S--the protein cofactor of activated protein C

Protein S, an important cofactor of activated protein C, and C4b-binding protein were purified from human plasma. Specific antibodies against the purified proteins were raised in rabbits and used for the development of immunologic assays for these proteins in plasma: an immunoradiometric assay for protein S (which measures both free protein S and protein S complexed with C4b-binding protein) and an electroimmunoassay for C4b-binding protein. Ranges for the concentrations of these proteins were established in healthy volunteers and patients using oral anticoagulant therapy. A slight decrease in protein S antigen was observed in patients with liver disease (0.78 +/- 0.25 U/ml); no significant decrease in protein S was observed in patients with DIC (0.95 +/- 0.25 U/ml). Criteria were developed for the laboratory diagnosis of an isolated protein S deficiency.     

A mutation in the protein S pseudogene is linked to protein S deficiency in a thrombophilic family

Probands from 15 unrelated families with hereditary protein S deficiency type I, that is having a plasma total protein S concentration fifty percent of normal, were screened for abnormalities in their protein S genes by Southern analysis. Two probands were found to have a deviating DNA pattern with the restriction enzyme MspI. In the two patients the alteration concerned the disappearance of a MspI restriction site, CCGG, giving rise to an additional hybridizing MspI fragment. Analysis of relatives of both probands showed that in one family the mutation does not co-segregate with the phenotype of reduced plasma protein S. In the family of the other proband, however, complete linkage between the mutated gene pattern and the reduced total protein S concentration was found: 12 heterozygous relatives showed the additional MspI fragment but none of the investigated 26 normal members of the family. The mutation is shown to reside in the PS beta gene, the inactive protein S gene. The cause of type I protein S deficiency, a defect PS alpha gene has escaped detection by Southern analysis. No recombination has occurred between the PS alpha gene and the PS beta gene in 23 informative meioses. This suggests that the two protein S genes, located near the centromere of chromosome 3, are within 4 centiMorgan of each other.     

Plasma protein S activity measured using Protac, a snake venom derived activator of protein C

Functional activity of protein S, a cofactor of activated protein C-dependent inhibition of blood coagulation, in human plasma was measured by using Protac, a snake venom derived activator of protein C. This assay appeared to be specific for protein S, because 1) the activated partial thromboplastin time of protein S-depleted plasma depended on the purified protein S added in the presence of Protac; and 2) the level of protein C in plasma sample (0 to 10 micrograms/ml) had no influence on the clotting time. The cofactor activity of protein S in the plasma of normal men (n = 16) and women (n = 14) was 99.4 +/- 23.8% and 98.6 +/- 24.5% respectively. The protein S activity in the plasma of pregnant women at pre- and post-partum (n = 14), and that in the plasma of patients under warfarin therapy (n = 20) were 46.2 +/- 18.9%, 45.8 +/- 19.6% and 24.0 +/- 15.7%, respectively. In these plasmas, the levels of protein S activity were lower than those of total protein S antigen, but were similar to those of free protein S antigen. In 16 patients out of two families with congenital protein S deficiency, the protein S activity, the free antigen and the total antigen were 9.4 +/- 6.9%, 13.3 +/- 4.6% and 57.4 +/- 20.7%, respectively. There was no significant relationship between the level of protein S activity and that of a complemental C4b-binding protein antigen in any of these patients.     

A collaborative study on the measurement of protein S antigen in plasma.

A collaborative study on the measurement of protein S (PS) antigen (total and free) in a freeze-dried ampouled test plasma by assay against local house standard plasmas was carried out in eleven laboratories. Potency estimates of total PS showed good agreement between laboratories with a geometric coefficient of variation (gcv) of 5.9% and an overall combined potency of 0.84 units per ml. Potency estimates of free PS antigen in the test sample were associated with increased variability between laboratories resulting from the polyethylene glycol (PEG) precipitation step which is used to separate free PS from PS bound to the C4b binding protein. Free PS in the test could be expressed relative to either total PS in the house standards (e.g. 0.28 units per ml) or relative to free PS in the house standards following PEG precipitation (e.g. 0.71 units free PS per ml).     

C4b-binding protein inhibits the factor V-dependent but not the factor V-independent cofactor activity of protein S in the activated protein C-mediated inactivation of factor VIIIa

Activated protein C (APC) is an important inactivator of coagulation factors Va and VIIIa. In the inactivation of factors Va and VIIIa, protein S serves as a cofactor to APC. Protein S can bind to C4b-binding protein (C4BP), and thereby loses its cofactor activity to APC. By modulating free protein S levels, C4BP is an important regulator of protein S cofactor activity. In the factor VIIIa inactivation, protein S and factor V act as synergistic cofactors to APC. We investigated the effect of C4BP on both the factor V-independent and factor V-dependent cofactor activity of protein S in the factor VIIIa inactivation using a purified system. Protein S increased the APC-mediated inactivation of factor VIIIa to 60% and in synergy with protein S, factor V at equimolar concentrations increased this effect further to 90%. The protein S/factor V synergistic effect was inhibited by preincubation of protein S and factor V with a four-fold molar excess of C4BP. However, C4BP did not inhibit the factor V-independent protein S cofactor activity in the purified system whereas it inhibited the cofactor activity in plasma. We conclude that C4BP-bound protein S retains its cofactor activity to APC in the factor VIIIa inactivation.