Coagulation assay with improved specificity to factor V mutants insensitive to activated protein C

Thrombosis

Pergamon

Research, Vol. 80, No. 3, pp. 255-264 1995 Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved 0049.3848/95 $9.50 + .oO

0049-3848(95)00174-3

COAGULATION ASSAY WITH IMPROVED SPECIFICITY TO FACTOR V MUTANTS INSENSITIVE TO ACTIVATED PROTEIN C Michael Kraus, Norbert Zander and Karl Fickenscher Research Laboratories of Behringwerke AG, D-35001 Marburg, Germany

(Received

Abstract

5 March

1995 by Editor J. Stiirzebecher;

revised/accepted

18 August

1995)

The prevalence of a new hereditary defect in the protein C anticoagulant pathway, the factor V-Leiden, has been reported to range between 20% to 60% in familial thrombophilia. In addition to differences in patient groups, these very divergent numbers might also be due to the detection method applied. In most studies a modified APTT was used, where activated protein C (APC) is added simultaneously with the start of the clotting reaction. However, this method is also influenced by other factors like protein S, factor VIII or lupus anticoagulants. Furthermore, heparin or oral anticoagulant therapy might interfere. We tried to develop a coagulation assay dependent only on those mutant forms of factor V stable against proteolytic attack by APC. For this purpose, samples were first diluted with a factor V deficient plasma (f.V-dp). Then, coagulation was initiated either on the intrinsic pathway (APTT), or on the extrinsic pathway (PT) or, by directly activating factor X (RVVT). Additionally, APC was added, which prolongation of the clotting time. Deficiencies in protein S or the presence of factor V-Leiden resulted in a less pronounced clotting time prolongation. Titration of protein S-deficient plasma samples with f.V-dp diminished this effect. In contrast, in samples with factor VLeiden the difference to the clotting time obtained with normal plasma even increased in the order APTT>>RVVT>PT. In the APTT-based method high concentrations of factor VIII shortened the clotting times, thus mimicking a factor V-Leiden defect. This could be compensated for up to 4 U/ml factor VIII by using a f.V-dp containing factor VIII at physiological concentration. Neither unfractionated nor LMW-heparin (up to 2 U/ml) interfered with the determination. In a brief investigation on 16 plasma samples from patients under oral anticoagulation 5 (30%) showed a similar behaviour as observed with normal plasma from factor V-Leiden carriers. These results let us suggest that by simply mixing the patient sample with a factor V-deficient plasma factor V-Leiden might be detected also in patients under oral anticoagulant therapy,

Key words: activated protein C, factor V-Leiden, APC-resistance, oral anticoagulation Corresponding author: Dr Michael Kraus, Research Laboratories, Behringwerke AG, D-35001 Marburg, Germany

256

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

Inherited disorders of protein C or protein S are well known as thrombotic risk factors (1). The recent investigations by Dahlblck et al. (2) led to the discovery of a new hereditary defect in the protein C anticoagulant pathway: the factor V-Leiden (3). This mutation renders activated factor V stable against proteolytic attack by activated protein C (APC). The reports on the prevalence of this mutation in thrombophilic patients show a considerable variation between 21% (4) and 64% (5). One reason for these very different findings might be the selection of patient groups studied. However, the specificity of the detection method applied is also of major importance. The most specific method for identification of factor V-Leiden is gene analysis. Since gene analysis is a very expensive and sophisticated technology, in most studies a modified APTT has been used. In this assay APC is added simultaneously with starting the clotting reaction, thus causing a prolongation of the clotting time. In presence of factor V-Leiden, this prolongation is less pronounced. However, this effect might not only be caused by factor V-Leiden. The same method has already been used previously for investigation of the influence of protein S, factor VIII:C, factor VIII:vWF, factor IX or lupus anticoagulants on the functionality of the protein Csystem (6,7). Therefore, the involvement of factor V could be found only after exclusion of these interferences (8). Furthermore, heparin or coumarin therapy might mask a defect of the protein C system, since they also prolong the clotting time. Therefore, we investigated the possibility whether and how the mixture of patient sample with factor V-deficient plasma (f.V-dp) would allow a more specific detection of the presence of mutant forms of factor V, such as factor V-Leiden, in various coagulation assays.

MATERIAL

AND METHODS

Liquemin@ (H o ff mann-LaRoche, Basel) was used as unfractionated heparin, Fragmin@ (KabiVitrum, Munchen) as low molecular weight (‘LMW’)-heparin. For the RVVT, LA-Search@ from Gradipore, Sydney, was used. All other reagents and plasmas were supplied by Behringwerke AG, Marburg. As a source of purified factor VIII, Beriate@ was used. The heterozygous factor V-Leiden defect was detected in plasmas from otherwise healthy blood donors using the APC-Sensitivity Reagents from Behringwerke AG (9). Only plasmas with clotting times below 70 set in the APCT assay (see below) were used, which is much shorter than observed for total protein S-deficiency or abnormal factor VIII-concentrations. Genetic analysis, which was performed by Professor Witt (Freiburg), confirmed the factor V-Leiden defect in all plasmas of which whole blood was available for investigation (see legend to figure 3). Plasmas from patients under oral anticoagulant therapy were generously provided by Professor Seitz, Marburg. Factor V-deficient plasmas were prepared by heat inactivation and obtained from Behringwerke AG, Marburg. For determination of the APTT, 100 ul sample were mixed with 100 ul Pathromtin@, a kaolin based APTT-reagent containing phospholipids. After 3 min at +37 “C, the clotting reaction was started by addition of 100 ul of a 25 mM calcium chloride solution. For determination of the APC dependent clotting time (‘APCT’), the clotting reaction was released by addition of a APC/calcium chloride solution from the APC-Sensitivity Reagents from Behringwerke AG (9). For a PT-based variant, Thromborel Se, a tissue factor/phospholipid preparation from human placenta was diluted 1: 1000 in 50 mM Tris-(hydroxymethyl)-aminomethane, pH 7.4, containing 0.4 U/ml APC, 10 mM calcium chloride and 0.01% (w/v) phospholipids from soy beans. 100 ul of sample were mixed with 200 ul of this PT/APC-mixture and the clotting time was recorded.

257

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

For the RVVT-based assay 0.4 U/ml activated protein C were added to the LA-Search@ reagent containing phospholipids and Russell’s viper venom (‘RVV’), which activates factor X and factor V in the sample. 200 ul of sample were mixed with 200 ul of the RW/APC mixture and the clotting time was recorded. All reagents containing APC were heated to +37 “C prior to use. In these APC-dependent assays, an impairment of the protein C system results in a less pronounced prolongation of the clotting time relative to a normal plasma pool. Therefore, samples having shorter clotting times than normal plasma are considered positive in the resp. assay. The assays were performed either on a Schnitger & Gross coagulometer (Amelung AG, Lemgo) or on a Behring Fibrintimer A@ (Behringwerke AG, Marburg). (For an overview on the methods see Table I).

RENJLTS Neutralisation of factors of the PC-system besides factor V-Leiden The behaviour of a normal plasma pool (NPP), a protein C- or a protein S-immunodepleted plasma and a plasma of a blood donor with confirmed heterozygous factor V-Leiden defect were investigated in the APC-dependent coagulation assays. Additionally, protein S-depleted plasma was mixed I+1 with factor V-Leiden plasma to simulate a coexistence of both defects. In all untreated samples, the APTT was within the normal range (25 to 40 set), thus any defect in procoagulatoty factors could be excluded. The concentration of APC in the reagents as described in ‘Material and Methods’ had been optimized previously to meet the following requirements. First, the prolongation of the clotting time in NPP should be at least twofold to the clotting time in absence of APC. Secondly, in order to obtain a good discrimination between NPP and heterozygous factor V-Leiden plasma, the clotting times between these two plasmas should also at least differ by a factor of 2 in presence of APC (Table I). In the presence of exogenously added APC, a protein S-deficient plasma was found positive as indicated by a shorter clotting time (see Fig. 1 A for the APTT-, 1B for the RVVT- and 1C for the PT-based method), while a protein C-deficiency had no effect. As can be deduced from Fig. 1, in the APTT-based method, a heterozygous factor V-Leiden plasma revealed even shorter clotting times than the protein S deficient plasma. However, this discrimination was less

TABLE I. Overview on the Clotting Methods used.

NPP f.V-L PW

APTT PT + ratio + ratio 34.0 120.6 3.5 45.6 126.4 2.8 32.5 61.8 1.9 46.0 64.1 1.4 intrinsic extrinsic

32.4 30.4

RVVT + ratio 81.6 2.5 38.6 1.3 factor X

Shown are and the clotting times (set) obtained with a normal plasma pool (NPP) or a heterozygous factor V-Leiden plasma (f.V-L) in absence (-) or in the presence (+) of activated protein C. PW = underlying coagulation pathway.

258

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

pronounced in the RVVT-based method and even no difference in clotting times between these plasmas could be found in the PT-based method. A mixture of protein S deficient plasma and factor V-Leiden could not be discriminated from a protein S deficiency alone in any of these assays.

50 0

L

.. ...

..T--

‘...

elIu31...

-50

-100 I 100

PC-def I

portion f. V-deficient

of plasma

PS-def

f V-L

1113 0% •I]50% q

f. V-UPS-def 60%

l

70%

n

I

75%

FIG. 1. Influence of factor V-deficient plasma content (O-75% of sample) on APC sensitivity applying either (A) APTT-, (E3)RVVT- or (C) PT-based methods. The bars represent the clotting times (set) relative to that obtained with a normal plasma pool using the same mixture with fV-deficient plasma. The first bar in each serie (0% f V-dp) indicates the currently used methodology. A negative value indicates a pathological sample. PC-def = protein C-, PS-def = protein S-immunodepleted plasma; f V-L = heterozygous factor V-Leiden plasma; fV-LiPS-def = heterozygous factor V-Leiden mixed l+l with protein Simmunodepleted plasma.

Vol. 80, No. 3

FACTOR V APC RESISTANT ASSAY

In order to compensate for these interferences, the samples were diluted with increasing portions of a factor V-depleted plasma and then applied as before to the respective assay. The factor V deficiency thus introduced led to an additional prolongation of the clotting times: for NPP up to 180 set in the APTT-, up to 120 see in the RVVT- and up to 160 set in the PT-based method. For comparability, the clotting time obtained with NPP at each mixture with f.V-dp was used as reference point in the respective method. Thus, the differences of the clotting times of the samples to that obtained with NPP are indicated in Fig. 1. Negative values indicate an impairement of the protein C-system. With increasing concentration of EV-dp, the difference between normal plasma and the factor V-Leiden plasma even increased in the APTT and to a minor degree also in the RVVT. In the PT-based method, however, the difference to NPP remained the same. In the APTT-based method, any influence of protein S was neutralized at a EV-dp portion of about 70%, while more than 75% would be required in the RVVT-based method. In the mixture of protein S-deficient plasma with the heterozygous factor V-Leiden plasma simulating a coexistence of both defects, the clotting times relativ to NPP were unchanged. Thus, mixing the patient sample with at least 3 portions of the f.V-dp (75% f.V-dp), only the type of factor V present in the sample determines the test result in the APTT-based method. In the PT-based method only a minor neutralization of protein S-deficiency could be achieved. Therefore, this attempt is not suited for an improved method for detection of factor V-Leiden and, thus, was omitted from further investigations, Injluence and neuiralisation of factor VIII After mixing of the patient sample with f.V-dp, any interference by clotting factors of the procoagulant pathway can be neglected, since all missing proteins are supplied by the fV-dp. However, this is not necessarily to be expected for factor VIII. As a cofactor, variations in factor VIII concentration have a higher impact on clotting time than any of the other factors (besides factor V), especially in an APTT-based assay. Since it is well known that factor VIII concentrations might increase quite significantly under various clinical conditions as e.g. inflammation, acute phase or pregnancy, the influence of factor VIII has been examined. A normal plasma pool and a heterozygous factor V-Leiden plasma were spiked with factor VIII or mixed with F.VIII-depleted plasma, and then used in the APTT-based method (Fig. 2) at a 1+3 mixture with two different qualities of the EV-substrate plasma. One f.V-substrate plasma contained no factor VIII (Fig. 2A), while the other f.V-substrate plasma contained 1 U/ml (Fig. 2B). In case of factor VIII-deficiency, the ciotting times were prolonged so that a disturbance of the protein C-system might be missed. However, while a factor VIII deficiency can already be recognized by an abnormal APTT, increased factor VIII-concentrations would be missed. Factor VIII accelerates the clotting reaction, thus mimicking an impaired APC-response in an otherwise normal plasma. This was confirmed especially in the case that the f.V-substrate plasma contained no factor VIII (Fig. 2A). Using a fV-substrate plasma containing 1 U/ml factor VIII, the influence of increased factor VIII in the sample was less pronounced (Fig. 2B). Even at factor VIII concentrations of up to 4 U/ml in the NPP, there was no overlap with the heterozygous factor V-Leiden plasma at normal factor VIII concentrations (dashed line in Fig. 2). The RVVTbased method was not dependent on factor VIII at all. Influence of heparin The reagents used for the APTT- and RVVT-based method contain heparin neutralizing substances. Therefore, in the mixtures with f.V-substrate plasma no heparin interference could be found even up to 2 U/ml of either unfractionated or LMW-heparin.

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

100 50 -(A) :\

0

1

2

3

4

factor WI concentration [u/ml] FIG. 2 Influence of factor VIII on the APTT-based method at a 1+3 mixture of samples with f.V-deficient plasma containing (A) no or (B) 1 U/ml factor VIII. Indicated are the clotting times relative to a normal plasma pool (dotted line). The clotting time difference obtained for a heterozygous factor V-Leiden plasma at 1 U/ml factor VIII is indicated by the dashed line. (-•-) = normal plasma pool; ( --U- - ) = heterozygous factor V-Leiden plasma.

Determinations in plasmas from patients under oral anticoagulant treatment. Up to now, the commonly used APCT-assay cannot be applied to plasma From patients under oral anticoagulation. However, based on the results described above, we wondered whether the mixture of patient sample with f.V-dp might also serve to identify factor V-Leiden in patients even under oral anticoagulant therapy. For this purpose, we investigated 2 single samples and 5 pools from patients under oral anticoagulation (OA). Furthermore, 25 normal plasmas and 13 samples of normal blood donors positive in the APCT were used. From 7 of these positive samples, material for gene analysis was available, which in all cases confirmed the presence of factor V-Leiden. However, no material for gene analysis was available for the OA plasmas. The results are summarized in Fig. 3, wherein the differences of the clotting times between sample and the NPP, which was used as a reference, are indicated for each assay. In the commonly applied APCT (Fig. 3A), the normal plasmas had clotting times close to that of the

Vol. 80, No. 3

FACTOR V APC RESISTANT ASSAY

261

NPP, while the factor V-Leiden plasmas had clotting times, which were significantly about 50 seconds shorter than that of the NPP (p < 0.001). The samples from OA patients (pools are denoted by boxes in Fig. 3) had longer clotting times than the NPP, although factor VIIIconcentrations were within normal range. This confirms that plasma from patients under OA cannot be used in the normal APCT. After mixing the samples 1+3 with f.V-dp (see Fig. 3B) the clotting time relative to that obtained with NPP even became shorter with heterozygous factor V-Leiden carriers (p < 0.001). Furthermore, one of the single OA plasmas (denoted by + in Fig. 3) now had clotting times similar to normal factor V-Leiden carriers, while the other (denoted by a * in Fig. 3) behaved similar to normals, The OA plasma pools were distributed between both. For a further characterization of the two single OA samples, the samples were mixed with increasing portions of f.V-dp as described before and the clotting time was determined in the APTT-based method. For the ‘normal’ OA plasma (Fig. 4A), the impairement of the APCdependent clotting time was neutralized with increasing portions of EV-dp, as has been observed for protein S-deficient plasma. For the ‘factor V-Leiden’-like OA plasma, the difference of clotting times to that of a NPP even increased as has been observed for the normal factor VLeiden carriers (Fig. 4B). However, gene analysis was not available. If taken for true, this would made up an incidence of 3 1% (5/16) of factor V-Leiden carriers among patients suffering from thrombosis. This is quite consistent with other numbers reported so far. This finding might help to explain the behaviour of the OA plasma pools (Fig. 3). Since about l/3 of the factor V in this pools will be factor V-Leiden, the response to APC will be impaired to some extent and thus, the clotting times obtained will be somewhere between normals or true heterozygous factor VLeiden plasma.

FIG. 3. Determination of APC sensitivity in the APTT-based method (A) without or (B) after mixing the samples l-1-3 with factor V-deficient plasma. Indicated are the clotting times relative to a normal plasma pool (horizontal line). N = normal plasmas, EV-L = heterozygous factor V-Leiden plasmas, OA = plasmas from patients under oral anticoagulant therapy (M = Iyophilized plasma pools). (a) and (+) represent the samples shown in Fig. 4 A and B, respectively.

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

q 0% q 50% q 60% 70%

A

6

FIG. 4. Titration of two plasmas from patients under oral anticoagulant therapy with various concentrations of N-deficient plasma applying the APTT-based method. The bars represent the clotting times (set) relative to that obtained with a normal plasma pool. A, B = plasmas represented by (*) resp. (+) in Fig. 3.

DISCUSSION During the last year the discovery of the factor V-Leiden mutation (38) and its strong association with familial thrombophilia (4) and thrombotic risk (5) led to a tremendously increased interest in this field. However, the reports published so far differ quite significantly in the actual prevalence of this mutation. Besides differences in type and size of patient groups, a major reason for this might be the interpretation of the assays applied. The golden standard for identification of factor V-Leiden seems to be the PCR technique. However, this is a sophisticated, expensive technique with its own pitfalls. A simple coagulation assay which only depends on factor V resistant to degradation by activated protein C would be much more convenient for routine use. Even more, a functional assay offers the chance to identity mutations in factor V additional to the Arg 3 Gln exchange described for factor VLeiden (3) having a similar effect on the resistance towards APC (16). It was known from previous investigations (1,6,7,10) and is also shown here, that the commonly applied, modified APTT, in which activated protein C is added with the initiation of the clotting reaction, is influenced by other factors involved in the protein C reaction cascade. Furthermore, increased factor VIII concentrations might lead to misinterpretation of the results. In an attempt to neutralize these effects the samples were mixed with different portions of a factor V-deficient plasma. The clotting time of the resulting mixtures in presence of activated protein C was determined in modification to Amer et al. (6) applying standard coagulation assay methods: APTT, PT or RVVT.

Vol. 80. No. 3

FACTOR V APC RESISTANT ASSAY

263

With increasing portions of the factor V-deficient plasma, the interference by protein S could be neutralized in the APTT- and RVVT-based method only. In the RVVT-based method, at least 80% of the sample had to be factor V-deficient plasma, while only 70% were needed in the APTT-based method. This was astonishing, as it had been expected that direct activation of the factor V-dependent coagulation step using RVVT would be less influenced by other factors in comparison to the APTT. Perhaps, these findings support the hypothesis of a synergistic function of factor Va (12) or factor V (13) in degradation of factor VIIIa, which is still in discussion. As expected, factor VIII did not influence the RVVT- and PT-based methods, but the APTTbased methods. However, the factor VIII interference could be deminished by using a factor Vdeficient plasma containing factor VIII itself. Heparin did not interfere with any of these assays. One reason for the different behaviour of the various types of clotting methods used might be the composition of the phospholipids of the reagents. It is well known that protein S (14) and APC (15) require a specific amount and composition of phospholipids to achieve optimal functionality. In the PT-based method all the necessary phospholipids were supplied. Therefore, in order to minimize protein S effects, it might be advisable to use a phospholipid composition outside of optimal range. However, it was beyond the scope of this investigation to get into further details. The influence of lupus anticoagulants specifically interfering with protein C- or protein S-phospholipid complexes was not investigated as the abundance of such antibodies is not yet clear (6,lO). The APCT assay commonly used could not be applied to samples of patients under oral anticoagulant therapy. The present investigations suggest that by mixing the plasma at least 1+3 with f.V-deficient plasma, a factor V-Leiden defect might even be detected in patients under oral anticoagulant therapy. However, it seems advisable to use only fresh samples, so that factor VIII concentration is at least within normal range. Furthermore, the use of factor V-deficient plasma with normal factor VIII-concentration is certainly more suitable for mixing. In this study, no gene analysis was available for confirmation of factor V-Leiden in the plasma from oral anticoagulated patients investigated. Therefore, these findings will have to be verified by further clinical investigations, In summary, mixing the sample at least 1+3 with a factor V-deficient plasma before applying the commonly modified APTT seems to give the best specificity for the presence of a factor VLeiden like mutation. Since the reagents necessary are already available, this method should soon be feasible for routine use. Note: Some of these data had been presented at the 39th Annual Meeting of the German Society for Thrombosis and Haemostasis in Berlin, february 1995. At the XVth Congress of the International Society on Thrombosis and Haemostasis in Jerusalem, june 1995, other groups reported similar results. In consequence, the Scientific and Standardization Committee recommended to use a mixture of patient plasma and factor V deficient plasma for a more reliable detection of factor V-Leiden in coagulation assays.

Acknowledgements We thank Professor I. Witt, Freiburg, for performing the genetic study in the APC-resistant blood samples and Professor R. Seitz for the OA plasmas. We also thank Ms. A. Ochs and Ms. A. Pilgrim for skillfU1 technical assistance.

264

FACTOR V APC RESISTANT ASSAY

Vol. 80, No. 3

REFERENCES 1. ESMON, C.T. Protein S and protein C - Biochemistry, physiology, and clinical manifestation of deficiencies. Trends Cardiovasc Med 2,2 14-2 19, 1992. 2. DAHLBACK, B., CARLSSON, M. and SVENSSON, P.J. Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Nat1 Acad Sci USA 90,10041008, 1993. 3. BERTINA, R.M., KOELEMAN, B.P.C., KOSTER, T., ROSENDAAL, F-R., DIRVEN, R.J., DE RONDE, H., VAN DER VELDEN, P.A. and REITSMA, P.H. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 369, 64-67, 1994. 4. SVENSSON, P.J. and DAHLBACK, B. Resistance to activated protein C as a basis for venous thrombosis. New Engl J Med 330, 5 17-522, 1994. 5. GRIFFIN, J.H., EVATT, B., WIDEMAN, C. and FERNANDEZ, J.A. Anticoagulant protein C pathway defective in majority of thrombophilic patients. Blood a, 1989-1993, 1993. 6. AMER, L., KISIEL,W., SEARLES, R.P. and WILLIAMS, R.C. Impairment of the protein C anticoagulant pathway in a patient with systemic lupus erythematosus, anticardiolipin antibodies and thrombosis. Thromb Res 57,247-258, 1990. 7. RICK, M.E., ESMON, N.L. and KRIZEK, D.M. Factor IXa and von Willebrand factor mod+ the inactivation of factor VIII by activated protein C. J Lab Clin Med l& 415-421, 1990. 8. DAI-ILBACK, B. and HILDEBRAND, B. Inherited resistance to activated protein C is corrected by anticoagulant cofactor activity found to be a property of factor V. Proc Nat1 Acad Sci USA 9l, 1396-1400, 1994. 9. KRAUS, M. and WAGNER, C. Evaluation of APC-sensitivity in normal blood donors using different reagents and instruments. Thromb Res 76,23 l-236, 1994. 10. MALIA R.G., KITCHEN, S., GREAVES, M. and PRESTON, F.E. Inhibition of activated protein C and its cofactor protein S by antiphospholipid antibodies. Brit J Haematol 76, IOI107,199o. 11. COOPER P.C., HAMPTON, K.K., MAKRIS, M., ABUZENADAH, A., PAUL, B. and PRESTON, F.E. Further evidence that activated protein C resistance can be misdiagnosed as inherited functional protein S deficiency. Brit J Haematol 88, 201-203, 1994. 12. SALEM, H.H., BROZE, G.J., MILETICH, J.P. and MAJERUS, P.W. Human coagulation factor Va is a cofactor for the activation of protein C. Proc Nat1 Acad Sci USA 80, 1584-1588, 1983. 13. SHEN, L. and DAHLBACK, B. Factor V and protein S as synergistic cofactors in activated protein C in degradation offactor VIIIa. J Biol Chem 269, 18735-18738, 1994. 14. BAKKER, H.M., TANS, G., JANSSEN-CLAESSEN, T., THOMASSEN, M.C.L.G.D., HEMKER, H.C., GRIFFIN, J.H. and ROSING, J. The effect of phospholipids, calcium ions and protein S on rate constants of human factor Va inactivation by activated human protein C. Eur J Biochem 20& 171-178, 1992. 15. HORIE, S., ISHII, H., HARA, H. and KAZAMA, M. Enhancement of thrombinthrombomodulin-catalysed protein C activation by phosphatidylethanolamine containing unsaturated fatty acids: possible physiological significance of phosphatidylethanolamine in anticoagulant activity of thrombomodulin. Biochem J 3&l, 683-691, 1994. 16. ZGLLER, B., SVENSSON, P.J., HE, X. and DAHLBACK, B. Identification of the same factor V gene mutation in 47 out of 50 thrombosis-prone families with inherited resistance to activated protein C. J Clin Invest 94,2521-2524, 1994.