THROMBOSIS RESEARCH 58; 481-468,199O 0049-3848190 $3.00 + .OO Printed in the USA. Copyright (c) 1990 Pergamon Press pk. All rights reserved.
SCREENING OF PROTEIN S DEFICIENCY USING A FUNCI-JONAL ASSAY IN PATIENTS WJTH VENOUS AND ARTERJAL THROMBOSIS
Marie-Louise
Wiesel, Jean-Luc
Charmantier, Jean-Marie Freyssinet, Lelia Grunebaum, Simone Schuhler, Jean-Pierre Cazenave. Service d’Hemostase et de Thrombose, INSERM U. 311, Centre Regional de Transfusion Sanguine, 10 rue Spielmann, 67085 Strasbourg CCdex, France.
(Received
ABSTRACT
30.10.1989;
accepted
in revised form 1.3.1990
by Editor M.C. Boffa)
Protein S is the vitamin K-dependent cofactor of activated protein C which functions as a potent anticoagulant by degrading activated factors V and VJJJ in a Ca2+ and phospholipid-dependent reaction. Protein S circulates under two forms, free (- 40%) or bound to C4b-binding protein (C4b-bp); only the free form supports the cofactor activity for activated protein C. Total protein S antigen is usually measured by rocket immunoelectrophoresis. Free protein S antigen is measured by the same technique but after precipitation of the protein S-C4b-bp complex by PEG 8000. However, these immunological assays do not detect functional alterations of protein S which can be responsible for thrombosis. This paper describes a functional assay for free protein S based on its ability to promote the prolongation of clotting time following factor Va inactivation by activated protein C when coagulation is triggered by factor Xa. Using this assay a prolongation of about 100 s between 0 and 1 U/ml protein S is measured, allowing a reliable and rapid determination of functional protein S. The correlation coefficient between functional protein S and free antigenic protein S is 0.921. This functional protein S assay has allowed the detection of 34 cases of protein S deficiency, confirmed by immunological assays, and their classitication. The striking observation is the high frequency (- 25%) of arterial thrombosis in these patients. The rapid determination of functional protein S in patients with venous or arterial thrombosis is of diagnostic interest and should allow the detection of mutant protein S in combination with an JmmunologicaJ assay.
JNTRODUCTJON Protein S (PS) is the vitamin K-dependent cofactor of activated protein C (APC) reaction which inhibits activated factors V and VJJJ in a Ca 2+ and phospholipid-dependent APC could also stimulate fibrinolysis by inhibiting plasminogen activator inhibitors 1 and The cofactor activity of PS is expressed through the formation of a ternary complex with
Keywords:
Protein S, functional
assay, arterial thrombosis, 461
venous thrombosis.
(1,2) (3,4). 3 (5). APC
462
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
and anionic phospholipids (6). This complex would modulate protection of factor Va by factor Xa at the phospholipid interface (7). PS circulates under two forms, free (- 40 %) or bound to C4b-binding protein (C4b-bp), a regulatory protein of the complement system (- 60 %) (8,9). Only the free form of PS supports the cofactor activity for APC (10). The relation between hereditary PS deficiency and venous thrombosis was established in 1984 (11,12). Two types of anomalies have been described : i) Type I deficiency, where most PS is complexed to C4b-bp; hence there is no or only low levels of circulating active free PS in plasma with normal or subnormal levels of bound PS. ii) Type II deficiency, where both bound and free PS are reduced (13). The overall prevalence is identical to that of protein C (PC) deficiency, each representing - 10 % of biological anomalies associated with recurrent familial thrombosis. Clinical manifestations resemble those of PC deficiency for venous thrombosis. However, contrary to PC deficiency, significant tendency to arterial thrombosis has been reported by several groups in the case of PS reduction (14-18). The common technique used for PS evaluation is rocket immunoelectrophoresis for the detection of either total PS antigen or free PS after precipitation of the C4b-bp complex with PEG 8000 (12,13). However these immunological assays do not enable detection of functional PS molecular defects potentially responsible for thrombosis. Furthermore such assays usually require at least 24 hours for results. A more rapid technique would be of benefit especially in the case of arterial thrombosis. So far, two types of functional assays have been designed, based on the ability of PS to promote the prolongation of clotting time by APC. They both measure free PS, i.e, functional PS. The triggers of coagulation can be either factor Xa or contact phase activation. In the first type the prolongation of clotting time is due to factor Va inactivation, while in the second one it reflects the inactivation of both factors VIIIa and Va (10,11,19,20). Such assays give a prolongation of the clotting time of only - 25 s which could make interpretation difficult. More recently a PS assay using PC activated by venom activator giving a prolongation of - 50 s between 0 and 1 U/ml of PS was published (21). In the present paper we describe a functional assay of PS in which the clotting time is prolonged by about 100 s between 0 and 1 U/ml PS. Using this assay, we detected 34 PS deficiencies among a group of patients with vascular thrombosis and we compared the functional and immunological methods of screening. PATIENTS ---___._--I--_ MATERIALS
AND METHODS ____
Vitamin K-dependent proteins C, S and factor X were purified as described (22) by a combination of adsorption on precipitating barium salts, fractionation by ammonium sulphate precipitation, ion exchange chromatography at different pH and afftity chromatography on insolubilized dextran sulphate. APC was obtained as described (23) by activation of purified PC 2 hours at 37°C with human a thrombin (15 NIH units/ml, final concentration for 0.75 mg/rnl PC) in the presence of 2mM EDTA. The activation reaction was terminated by the addition of an excess of hi&in (Diagnostica Stago, Asnibres, France). A final concentration of 20 nM APC was obtained by dilution in 50 mM Tris buffer pH 7.5, containing 0.1 M NaCl and 0.3 % (w/v) human albumin (CRTS ctrasbourg, France). APC was stored at -80°C and upon thawing it was used within a limit of one hour while kept at 0°C. Factor Xa was obtained as described (24) by activation of purified factor X with Russell’s viper venom . Factor Xa was diluted to a final concentration of 3 nM in the above buffer containing also 0.3 % (w/v) human albumin and stored and used as APC. Anti-PS antibodies were raised in rabbits which had been immunized by 4 weekly injections and a booster injection one month later, using 50 pg of purified human PS/injection. IgG fraction was obtained as the non-binding protein when the rabbit serum was adsorbed on DEAE Trisacryl-M (IBF, Villeneuve-la-Garenne, France) previously equilibrated in 25 mM Tris buffer pH 8.8 containing 35 mM NaCl. The IgG fraction was coupled to CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden), 10 mg of IgG/ml of swollen gel. Thirty ml of a pool of normal titrated human plasma, screened for the absence of viral contamination, was subjected to anti-PS adsorption on a 1 x 10 cm column containing the affinity gel, at a flow rate
Vol. 58, No. 5
SCREENING OF PROTEIN S DEFICIENCY...
463
of 12 ml/hour at room temperature. The depleted plasma contained no detectable PS and had normal levels (U/ml, mean + SD, n = 10) of factors V (0.97 + 0.07), VIII (1.04 f 0.1 l), II (0.92 f 0.08), fibrinogen (2.04 f 0.22 g/l) and - 0.25 U/ml C4b-bp (1 U of C4b-bp corresponding to the amount of C4b-bp present in one ml of control plasma). Phospholipids were purchased from Stago (Cephaline Stago, Asnieres, France) and were used diluted 15 in 50 mM Tris buffer pH 7.5 , containing 0.1 M NaCl, after reconstitution as recommended by the manufacturer. Blood samples (9 vol.) were collected into 0.13 M trisodium citrate (1 vol.); plasma was obtained after 2 min centrifugation at 10 000 g and stored in several aliquots at -80°C in order to avoid repeated freezing and thawing. Normal reference plasma was obtained by mixing titrated plasmas obtained from 20 healthy donnors, freezing at -80°C being carried out within one hour of blood collection.
During the last seven years, about 50 patients were analyzed per year, all suffering from recurrent venous or arterial thrombosis or thrombosis without apparent clinical etiology (25). All these patients were systematically screened for PC, PS, antithrombin III, heparin cofactor II and factor XII deficiencies, dysfibrinogenemia, impaired fibrinolysis and lupus like anticoagulants. Up to now routine assay for PS was rocket immunoelectrophoresis after precipitation of the PSC4b-bp complex with PEG 8000. For each patient, plasma aliquots were stored frozen at -80-C for further investigations which allowed us to detect 34 PS deficiencies during these last seven years. When identified during acute thrombotic episode, patients were controlled at distance from the accident without significant modification of their PS status. Immunological
assay of protein S
Antigenic PS was measured by rocket immunoelectrophoresis in agarose gels, Total PS was measured under conditions that favour dissociation of the PS-C4b-bp complex, i.e, low voltage and long migration time, at 2O’C. Free PS antigen was measured under the same electrophoretic conditions, but after precipitation of the PS-C4b-bp complex with polyethylene glycol 8000 at a final concentration of 3.75 % for half an hour in melting ice (13). F_unctional ass_ay of proteirS A calibration curve was constructed by serially diluting reference plasma in PS deficient plasma, 100 % PS corresponding to 1:8 dilution. Such a dilution procedure enables minimization of variations due to different levels of factor V in the samples to be tested. To 100 cl_lof plasma mixture (either reference plasma + PS-deficient plasma, or plasma to be assayed + PSdeficient plasma), 50 pl of phospholipid suspension was added together with 50 pl of APC. Preincubation for 1 min at 37°C was allowed. Then 50 p.l of 40 mM CaC12 was added. After 1 min at 37°C the clotting reaction was triggered by addition of 50 pl of factor Xa solution. Clotting times were recorded in duplicate on a semi automatic coagulometer KC10 (Dade, Amehmg, FBG). Plasmas with unknown PS levels were tested at two dilutions , 1:8 and 1: 16. In order to avoid any degradation of factor Xa or APC, these reagents were used within one hour after thawing. RESULTS The specificity of the assay was demonstrated by the following points. In the absence of PS (PS-deficient plasma), the clotting time was 47.7 f 3.7 s (n = 10 ). The addition of 10 pg of purified PS/ml of PS-deficient plasma, corresponding to physiological concentrations of free PS (26), resulted in a prolongation of the clotting time from 47.7 to 132 s, when used at a dilution of 1:8 as described in Materials and Methods . The reproducibility of the assay was evaluated with the intra and interassay variation coefficients which were 2.2 % (n = 10) and 4.02 % (n = 10) respectively. Normal values for functional PS in plasma from 30 healthy blood donnors were 1.08 f 0.39 units/ml. Free PS antigen determination gave 0.91 f 0.36 units/ml. Fig. 1 represents
464
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
c
8 0.25
0.50
0.75
1.0
PROTEIN S (U/ml) Figure. 1 Calibration curve fur the asay of functional protein S. Reference pool plasma was diluted in protein S deficient plasma, 1 U/ml corresponding to 1:s dilution, 0.75 U/ml to 1:10.7, 0.5 U/ml to 1:16 and 0.25 U/ml to 1:32. Plasma to be assayed was diluted at a ratio of 1:8 and 1:16 in protein S deficient plasma. To 100 pl of plasma mixture (either reference plasma or plasma to be tested + PS deficient plasma), 50 l.tl of phospholipid and 50 pl of 20 nM APC were added. After preincubation for 1 min at 37’C. 50 pl of 40 mM CaC12 was added followed by another 1 min incubation at 37°C and the clotting reaction was triggered by addition of 50 pl of 3 nM factor Xa. Clotting time was recorded in duplicate on a semi automatic coagulometer.
2
1.5
5 i= g
1.0
0
z
0
0
u
-J
2
0.5
W
5 3
IA 0.0
0.0
0.5 1.0 FREE ANTIGEN Wml)
1.5
Pigum2 Correlation between functional protein S and free protein S antigen. Twenty seven patients with immunologicalIy confiied protein S deficiency, belonging to type I (0) and to type II deficiency (A) according to Comp et al (13) and 22 relatives (0) were tested for protein S functional activity. The correlation coefficient was r = 0.921. Other experimental details were as described in Materials and Methods.
Vol. 58, No. 5
SCREENING
OF PROTEIN
S DEFICIENCY...
465
a typical calibration
curve showing the dependence of clotting times on PS levels. Out of 34 patients with immunologicaIly confiied PS deficiency, either type I or type II, 27 were tested for PS functional activity, as well as 22 relatives. A good linear correlation between free PS antigen and functional PS could be observed (r = 0.921),( Fig.2 ). The combination of functional and antigenic assays allowed the classification of the 34 reported deficiencies. i) Total + free antigenic protein $ and functional protein S deficiencies. This group, type II deficiency according to Comp et al. (13), represents - 35 % of PS deficiencies. The mean level of total PSwas 0.56 f0.07 units/ml, that of free antigenic PS was 0.27 f 0.09 units/ml and that of functional PS was 0.29 + 0.13 units/ml. When measured, the mean level of C4b-bp was within the normal range, 0.82 + 0.06 units/ml. All patients from this group had severe thrombosis : multiple phlebitis, pulmonary embolism, mesenteric infarction and one case of myocardial infarction at the age of 20. Except one patient who fiit suffered from thrombosis at the age of 50, the mean age of occurrence of thrombosis was 22. ii) Fun_ctional protein S and-free antigenic protein S deficiencies with normal total antigenic protein S. This group, type I deficiency according to Comp ef al. (13), represents 65 % of PS deficiencies. The mean level of total antigenic PS was 0.9 f 0.23 units/ml, whereas the mean level of functional PS was reduced to 0.37 f 0.2 units/ml and that of free antigenic PS was lowered in parallel to 0.34 + 0.13 units/ml. Mean level of C4b-bp was within normal range : 1.15 + 0.21 U/ml. All patients also suffered from severe thrombosis except children under 15. In both groups - 25 % of the patients had arterial thrombosis. In order to assess the sensitivity of functional PS to oral anticoagulant treatment, 10 anticoagulated patients were tested (INK = 3.45 + 0.68 ). Mean total PS antigen was 0.65 f 0.12 units/ml, mean free PS antigen 0.39 + 0.11 units/ml and functional PS 0.34 + 0.079 units/ml. DISCUSSI,oN Since PS functions as a cofactor for APC, its reduction or absence should result in a shortening of the clotting time of plasma exposed to this potent natural anticoagulant. The design of a functional PS assay based on the use of a PSdeficient substrate plasma implies the adjustment of several parameters in order to correlate the prolongation of the clotting time to the amount of free PS in the plasma to be tested. Furthermore the variations of factors V and VIII, the precursors of the substrates of activated PC, and/or of other vitamin K- dependent factors should have a minimal influence. The present assay takes into account the sole inactivation of factor Va by APC since coagulation is initiated by factor Xa. It appears to be specific as PSdeficient plasma behaved as normal reference plasma when resupplemented with physiological levels of free PS (- 10 pg/ml, fmal concentration) at comparable dilutions. A prolongation of the clotting time of - 100 s in the presence of 100 % PS allows a reliable determination of functional PS in patients’plasma. Linearity between prolongation of the clotting time and PS amount (Fig. 1) probably allows a more precise measurement of low levels of PS (i.e. < 0.2 U/ml) than the immunological assay. This could be responsible for the discrepancies between functional and antigenic PS in this area of low PS levels (Fig. 2). In addition, as the lowest dilution of reference plasma in PS deficient plasma corresponding to 100 % PS is 1:8, this should minimize the effects of possible abnormal levels of factor V or vitamin K-dependent factors in sample plasmas. Dilutions were made just before assay as advance preparation could have resulted in an apparent diminution of free PS due to possible complexation with the 25% remaining C4b-bp in PS deficient plasma. Another advantage of this functional assay is its rapidity of execution since it is almost as fast as other routine coagulation tests. In our experience PS deficiency seems to be more threatening for thrombosis than PC deficiency which appears to be largely asymptomatic (27). Therefore a rapid diagnosis should be of great help in the decision to prescribe anticoagulant treatment in order to diminish the risk of (re)thrombosis. The only limitation of the assay is the stability of some labile preparations, in particular APC and factor Xa. For these compounds storage at -8O’C is crucial as welI as their use within one hour after thawing. To our knowledge there is no available commercial preparation of human factor Xa or APC of established stability. Preparations from
SCREENING OF PROTEIN S DEFICIENCY...
466
Vol. 58, No. 5
bovine origin might appear to be suitable, however our first aim was to measure PS activity in a human homogeneous system. The striking observation is the high frequency, _ 25 %, of arterial thrombosis in patients with PS deficiency, all of them being under predisposing conditions. It is clear that, in our study, type I deficiency (with normal total antigenic PS and decreased functional and free antigenic PS) is the most frequent anomaly. This emphasizes the need for rapid measurement of free PS in plasma of patients with thrombosis. The present functional assay fulfils this requirement. A further immunological determination or confirmation should allow the classification of the patients with functional PS defect, taking into account those with mutant PS since the functional assay should detect them. This information is of prime importance in the understanding of the complex transmission of these defects as two genes have been identified for PS, one being a pseudogene (28). Acknowledgments We thank Dr. J.N. Mulvihill for reviewing the English of this paper and Mrs. E. Schoettel and C. Helbourg for typing the manuscript. This work was partly supported by the Fonds Regional d’Alsace de Recherche-Developpement.
1.
DI SCIPIO R.G., HERMODSON M.A., YATES S.G., DAVIE E.W. A comparison of human prothrombin, Factor IX (Christmas Factor), Factor X (Stuart Factor) and protein S. Biochemistry, 16,698-706,1977.
2.
WALKER F.J. Regulation bovine protein S. J,Biol.
3.
WALKER F.J., SEXTON P.W., ESMON C.T. The inhibition of blood coagulation by activated protein C through the selective inactivation of activated factor V. B_ioch&n. Biophys._Acta, 571,333-342, 1979.
4.
WALKER F.J., CHAVIN S.I., FAY P.J. Inactivation of factor VIII by activated and protein S. Arch.Biochem. Biophys., 252,322-328, 1987.
5.
D’ANGELO A., LOCKHART M.S., D’ANGELO S.V., TAYLOR F.B. Protein S is a cofactor for activated protein C neutralization of an inhibitor of plasminogen activation released from platelets. Blood, 69,231-237,1987.
6.
WALKER F.J. The regulation of activated protein C by protein S . The phospholipid in factor Va inactivation. J. ~Biol?_Chern_ 256,11128- 1113 1, 1980.
7.
SOLYMOSS S., TUCKER M.M., TRACY P.B. Kinetics of inactivation of membranebound factor Va by activated protein C. Protein S modulates factor Xa protection. J. Biol, Chem., 263,14884-14890,1988.
8.
DAHLBACK B., STENFLO J. High molecular weight complex in human plasma between vitamin K-dependent protein S and complement component C4b-binding protein. Prqc. NatI. Acad Sci. US-A, 78,2512-2516,198l.
9.
DAHLBACK B. Purification of human C4b-binding protein and formation with vitamin Kdependent protein S. B&hem. J., 209,847-856,1983.
10.
DAHLBACK B. Inhibition of protein Ca cofactor function by C4b-binding protein. J, Biol.Chem,, 261,12022-12027,
of activated protein C by a new protein. Chem., 2555521-5524, 1980.
A possible
function
for
protein C
role
of
of its complex
of human and bovine protein 1986.
S
Vol. 58, No. 5
SCREENING OF PROTEIN S DEFICIENCY...
467
11.
COMP P.C., NIXON R.R., COOPER M.R., ESMON C.T. Familial protein S deficiency is associated with recurrent thrombosis. J. Clin. Invest., 74,2082-2088, 1984.
12.
SCHWARZ H.P., FISCHER M., HOPMEIER P., BATARD M.A., GRIFFIN J.H. Plasma protein S deficiency in familial thrombotic disease. Blood, 64,1297-1300,1984.
13.
COMP P.C., DORAY D., PATTON D., ESMON CT. Abnormal plasma distribution of protein S occurs in functional protein S deficiency. Blood, 67,504-508, 1986.
14.
MANNUCCI P.M., TRIPOD1 A., BERTINA R.M. Protein S deficiency associated with “juvenile” arterial and venous thrombosis. Thromb. Haemostas,, 55,440 (abstr.) , 1986.
15.
COLLER B.S., OWEN J., JESTY J., HOROWITZ D., REITMAN M.J., SPEAR J., YEH T., COMP P.C. Deficiency of plasma protein S, protein C or antithrombin HI and arterial thrombosis. Arteriosclerosis, 7,456462, 1987.
16.
ISRAELS S.J., SESHIA S.S. Childhood stroke associated with protein C or S deficiency. J. Pediat., 111,562-564, 1987.
17.
GIROLAMI A., SIMIONI P., LAZZARO A.R., CORDIANO I. Severe arterial cerebral thrombosis in a patient with protein S deficiency (moderatly reduced total and markedly reduced free protein S) a family study. Thromb. Haemostas., 61,144147,1989.
18.
SIE P., BONEU B., BIERME R., WIESEL M.-L., GRUNEBAUM L., CAZENAVE Arterial thrombosis and protein S deficiency. Thromb. Haemostas., 62, 1040, 1989.
19.
SUZUKI K., NISHIOKA J. Protein S activity measured using Protac, a snake venom derived activator of protein C. Thm_mb,Res., 49,241-251, 1988.
20.
VAN DE WAART P., PREISSNER K.T., BECHTOLD J.R., MULLER-BERGHAUS functional test for protein S activity in plasma. Thromb. Res., 48,427-437, 1987.
21.
KOBAYASHI I., AMEMIYA N., END0 T., OKUYAMA K., TAMURA K. KUME S. Functional activity of protein S determined with use of protein C activated by venom activator. Clin. Chem., 35, 1644-1648, 1989.
22.
FREYSSINET J.-M., SCHUHLER S., SCHUHLER E., GAUCHY J., RAVANAT C., WIESEL M.L., GRUNEBAUM L., STIERLE A., FOLLEA G., CAZENAVE J.-P. Vitamin K-dependent proteins of plasma : structure-function relationships, purification and therapeutic applications. In: Biotechnology of Plasma Proteins. J.F. Stoltz, C. Rivat (Eds.). Paris : Colloque Inserm (175), 1989, pp. 305-314.
23.
FREYSSINET J.-M., GAUCHY J., CAZENAVE J.-P. The effect of phospholipids on the activation of protein C by the human thrombin-thrombomodulin complex. B&hem. J., 238,151-157,1986.
24.
FREYSSINET J.-M., WIESEL M.-L., GRUNEBAUM L., PEREILLO J.-M., GAUCHY J., SCHULER S., FREUND G., CAZENAVE J.-P. Activation of human protein C by blood coagulation factor Xa in the presence of anionic phospholipids : enhancement by sulphated polysaccharides. Biochem. J., 261,341-348,1989.
25.
WIESEL M.-L., GRUNEBAUM L., FREYSSINET J.-M., CAZENAVE causes of familial thrombosis. Fibrinolysis, 2, supp.2, 14-15, 1988.
26.
HEEB M.J., GRIFFIN J.H. The biochemistry of protein S. In: Protein C and Related Proteins. R.M. Bertina (Ed.) Edinburgh : Churchill Livingstone, 1988, pp. 55-70.
J.-P.
G. A
J.-P. Biological
468
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
27.
MILETICH J., SHERMAN L., BROZE G. Jr. Absence of thrombosis in subjects with heterozygous protein C deficiency. N. Engl. J. Med., 317,991~996, 1987.
28.
PLOOS VAN AMSTEL J.K., REITSMA P.H., BERTINA R.M. Identification protein S p gene as a pseudogene. Thromb. Haemostas., 62,272 (abstr.), 1989.
of the
SCREENING OF PROTEIN S DEFICIENCY USING A FUNCI-JONAL ASSAY IN PATIENTS WJTH VENOUS AND ARTERJAL THROMBOSIS
Marie-Louise
Wiesel, Jean-Luc
Charmantier, Jean-Marie Freyssinet, Lelia Grunebaum, Simone Schuhler, Jean-Pierre Cazenave. Service d’Hemostase et de Thrombose, INSERM U. 311, Centre Regional de Transfusion Sanguine, 10 rue Spielmann, 67085 Strasbourg CCdex, France.
(Received
ABSTRACT
30.10.1989;
accepted
in revised form 1.3.1990
by Editor M.C. Boffa)
Protein S is the vitamin K-dependent cofactor of activated protein C which functions as a potent anticoagulant by degrading activated factors V and VJJJ in a Ca2+ and phospholipid-dependent reaction. Protein S circulates under two forms, free (- 40%) or bound to C4b-binding protein (C4b-bp); only the free form supports the cofactor activity for activated protein C. Total protein S antigen is usually measured by rocket immunoelectrophoresis. Free protein S antigen is measured by the same technique but after precipitation of the protein S-C4b-bp complex by PEG 8000. However, these immunological assays do not detect functional alterations of protein S which can be responsible for thrombosis. This paper describes a functional assay for free protein S based on its ability to promote the prolongation of clotting time following factor Va inactivation by activated protein C when coagulation is triggered by factor Xa. Using this assay a prolongation of about 100 s between 0 and 1 U/ml protein S is measured, allowing a reliable and rapid determination of functional protein S. The correlation coefficient between functional protein S and free antigenic protein S is 0.921. This functional protein S assay has allowed the detection of 34 cases of protein S deficiency, confirmed by immunological assays, and their classitication. The striking observation is the high frequency (- 25%) of arterial thrombosis in these patients. The rapid determination of functional protein S in patients with venous or arterial thrombosis is of diagnostic interest and should allow the detection of mutant protein S in combination with an JmmunologicaJ assay.
JNTRODUCTJON Protein S (PS) is the vitamin K-dependent cofactor of activated protein C (APC) reaction which inhibits activated factors V and VJJJ in a Ca 2+ and phospholipid-dependent APC could also stimulate fibrinolysis by inhibiting plasminogen activator inhibitors 1 and The cofactor activity of PS is expressed through the formation of a ternary complex with
Keywords:
Protein S, functional
assay, arterial thrombosis, 461
venous thrombosis.
(1,2) (3,4). 3 (5). APC
462
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
and anionic phospholipids (6). This complex would modulate protection of factor Va by factor Xa at the phospholipid interface (7). PS circulates under two forms, free (- 40 %) or bound to C4b-binding protein (C4b-bp), a regulatory protein of the complement system (- 60 %) (8,9). Only the free form of PS supports the cofactor activity for APC (10). The relation between hereditary PS deficiency and venous thrombosis was established in 1984 (11,12). Two types of anomalies have been described : i) Type I deficiency, where most PS is complexed to C4b-bp; hence there is no or only low levels of circulating active free PS in plasma with normal or subnormal levels of bound PS. ii) Type II deficiency, where both bound and free PS are reduced (13). The overall prevalence is identical to that of protein C (PC) deficiency, each representing - 10 % of biological anomalies associated with recurrent familial thrombosis. Clinical manifestations resemble those of PC deficiency for venous thrombosis. However, contrary to PC deficiency, significant tendency to arterial thrombosis has been reported by several groups in the case of PS reduction (14-18). The common technique used for PS evaluation is rocket immunoelectrophoresis for the detection of either total PS antigen or free PS after precipitation of the C4b-bp complex with PEG 8000 (12,13). However these immunological assays do not enable detection of functional PS molecular defects potentially responsible for thrombosis. Furthermore such assays usually require at least 24 hours for results. A more rapid technique would be of benefit especially in the case of arterial thrombosis. So far, two types of functional assays have been designed, based on the ability of PS to promote the prolongation of clotting time by APC. They both measure free PS, i.e, functional PS. The triggers of coagulation can be either factor Xa or contact phase activation. In the first type the prolongation of clotting time is due to factor Va inactivation, while in the second one it reflects the inactivation of both factors VIIIa and Va (10,11,19,20). Such assays give a prolongation of the clotting time of only - 25 s which could make interpretation difficult. More recently a PS assay using PC activated by venom activator giving a prolongation of - 50 s between 0 and 1 U/ml of PS was published (21). In the present paper we describe a functional assay of PS in which the clotting time is prolonged by about 100 s between 0 and 1 U/ml PS. Using this assay, we detected 34 PS deficiencies among a group of patients with vascular thrombosis and we compared the functional and immunological methods of screening. PATIENTS ---___._--I--_ MATERIALS
AND METHODS ____
Vitamin K-dependent proteins C, S and factor X were purified as described (22) by a combination of adsorption on precipitating barium salts, fractionation by ammonium sulphate precipitation, ion exchange chromatography at different pH and afftity chromatography on insolubilized dextran sulphate. APC was obtained as described (23) by activation of purified PC 2 hours at 37°C with human a thrombin (15 NIH units/ml, final concentration for 0.75 mg/rnl PC) in the presence of 2mM EDTA. The activation reaction was terminated by the addition of an excess of hi&in (Diagnostica Stago, Asnibres, France). A final concentration of 20 nM APC was obtained by dilution in 50 mM Tris buffer pH 7.5, containing 0.1 M NaCl and 0.3 % (w/v) human albumin (CRTS ctrasbourg, France). APC was stored at -80°C and upon thawing it was used within a limit of one hour while kept at 0°C. Factor Xa was obtained as described (24) by activation of purified factor X with Russell’s viper venom . Factor Xa was diluted to a final concentration of 3 nM in the above buffer containing also 0.3 % (w/v) human albumin and stored and used as APC. Anti-PS antibodies were raised in rabbits which had been immunized by 4 weekly injections and a booster injection one month later, using 50 pg of purified human PS/injection. IgG fraction was obtained as the non-binding protein when the rabbit serum was adsorbed on DEAE Trisacryl-M (IBF, Villeneuve-la-Garenne, France) previously equilibrated in 25 mM Tris buffer pH 8.8 containing 35 mM NaCl. The IgG fraction was coupled to CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden), 10 mg of IgG/ml of swollen gel. Thirty ml of a pool of normal titrated human plasma, screened for the absence of viral contamination, was subjected to anti-PS adsorption on a 1 x 10 cm column containing the affinity gel, at a flow rate
Vol. 58, No. 5
SCREENING OF PROTEIN S DEFICIENCY...
463
of 12 ml/hour at room temperature. The depleted plasma contained no detectable PS and had normal levels (U/ml, mean + SD, n = 10) of factors V (0.97 + 0.07), VIII (1.04 f 0.1 l), II (0.92 f 0.08), fibrinogen (2.04 f 0.22 g/l) and - 0.25 U/ml C4b-bp (1 U of C4b-bp corresponding to the amount of C4b-bp present in one ml of control plasma). Phospholipids were purchased from Stago (Cephaline Stago, Asnieres, France) and were used diluted 15 in 50 mM Tris buffer pH 7.5 , containing 0.1 M NaCl, after reconstitution as recommended by the manufacturer. Blood samples (9 vol.) were collected into 0.13 M trisodium citrate (1 vol.); plasma was obtained after 2 min centrifugation at 10 000 g and stored in several aliquots at -80°C in order to avoid repeated freezing and thawing. Normal reference plasma was obtained by mixing titrated plasmas obtained from 20 healthy donnors, freezing at -80°C being carried out within one hour of blood collection.
During the last seven years, about 50 patients were analyzed per year, all suffering from recurrent venous or arterial thrombosis or thrombosis without apparent clinical etiology (25). All these patients were systematically screened for PC, PS, antithrombin III, heparin cofactor II and factor XII deficiencies, dysfibrinogenemia, impaired fibrinolysis and lupus like anticoagulants. Up to now routine assay for PS was rocket immunoelectrophoresis after precipitation of the PSC4b-bp complex with PEG 8000. For each patient, plasma aliquots were stored frozen at -80-C for further investigations which allowed us to detect 34 PS deficiencies during these last seven years. When identified during acute thrombotic episode, patients were controlled at distance from the accident without significant modification of their PS status. Immunological
assay of protein S
Antigenic PS was measured by rocket immunoelectrophoresis in agarose gels, Total PS was measured under conditions that favour dissociation of the PS-C4b-bp complex, i.e, low voltage and long migration time, at 2O’C. Free PS antigen was measured under the same electrophoretic conditions, but after precipitation of the PS-C4b-bp complex with polyethylene glycol 8000 at a final concentration of 3.75 % for half an hour in melting ice (13). F_unctional ass_ay of proteirS A calibration curve was constructed by serially diluting reference plasma in PS deficient plasma, 100 % PS corresponding to 1:8 dilution. Such a dilution procedure enables minimization of variations due to different levels of factor V in the samples to be tested. To 100 cl_lof plasma mixture (either reference plasma + PS-deficient plasma, or plasma to be assayed + PSdeficient plasma), 50 pl of phospholipid suspension was added together with 50 pl of APC. Preincubation for 1 min at 37°C was allowed. Then 50 p.l of 40 mM CaC12 was added. After 1 min at 37°C the clotting reaction was triggered by addition of 50 pl of factor Xa solution. Clotting times were recorded in duplicate on a semi automatic coagulometer KC10 (Dade, Amehmg, FBG). Plasmas with unknown PS levels were tested at two dilutions , 1:8 and 1: 16. In order to avoid any degradation of factor Xa or APC, these reagents were used within one hour after thawing. RESULTS The specificity of the assay was demonstrated by the following points. In the absence of PS (PS-deficient plasma), the clotting time was 47.7 f 3.7 s (n = 10 ). The addition of 10 pg of purified PS/ml of PS-deficient plasma, corresponding to physiological concentrations of free PS (26), resulted in a prolongation of the clotting time from 47.7 to 132 s, when used at a dilution of 1:8 as described in Materials and Methods . The reproducibility of the assay was evaluated with the intra and interassay variation coefficients which were 2.2 % (n = 10) and 4.02 % (n = 10) respectively. Normal values for functional PS in plasma from 30 healthy blood donnors were 1.08 f 0.39 units/ml. Free PS antigen determination gave 0.91 f 0.36 units/ml. Fig. 1 represents
464
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
c
8 0.25
0.50
0.75
1.0
PROTEIN S (U/ml) Figure. 1 Calibration curve fur the asay of functional protein S. Reference pool plasma was diluted in protein S deficient plasma, 1 U/ml corresponding to 1:s dilution, 0.75 U/ml to 1:10.7, 0.5 U/ml to 1:16 and 0.25 U/ml to 1:32. Plasma to be assayed was diluted at a ratio of 1:8 and 1:16 in protein S deficient plasma. To 100 pl of plasma mixture (either reference plasma or plasma to be tested + PS deficient plasma), 50 l.tl of phospholipid and 50 pl of 20 nM APC were added. After preincubation for 1 min at 37’C. 50 pl of 40 mM CaC12 was added followed by another 1 min incubation at 37°C and the clotting reaction was triggered by addition of 50 pl of 3 nM factor Xa. Clotting time was recorded in duplicate on a semi automatic coagulometer.
2
1.5
5 i= g
1.0
0
z
0
0
u
-J
2
0.5
W
5 3
IA 0.0
0.0
0.5 1.0 FREE ANTIGEN Wml)
1.5
Pigum2 Correlation between functional protein S and free protein S antigen. Twenty seven patients with immunologicalIy confiied protein S deficiency, belonging to type I (0) and to type II deficiency (A) according to Comp et al (13) and 22 relatives (0) were tested for protein S functional activity. The correlation coefficient was r = 0.921. Other experimental details were as described in Materials and Methods.
Vol. 58, No. 5
SCREENING
OF PROTEIN
S DEFICIENCY...
465
a typical calibration
curve showing the dependence of clotting times on PS levels. Out of 34 patients with immunologicaIly confiied PS deficiency, either type I or type II, 27 were tested for PS functional activity, as well as 22 relatives. A good linear correlation between free PS antigen and functional PS could be observed (r = 0.921),( Fig.2 ). The combination of functional and antigenic assays allowed the classification of the 34 reported deficiencies. i) Total + free antigenic protein $ and functional protein S deficiencies. This group, type II deficiency according to Comp et al. (13), represents - 35 % of PS deficiencies. The mean level of total PSwas 0.56 f0.07 units/ml, that of free antigenic PS was 0.27 f 0.09 units/ml and that of functional PS was 0.29 + 0.13 units/ml. When measured, the mean level of C4b-bp was within the normal range, 0.82 + 0.06 units/ml. All patients from this group had severe thrombosis : multiple phlebitis, pulmonary embolism, mesenteric infarction and one case of myocardial infarction at the age of 20. Except one patient who fiit suffered from thrombosis at the age of 50, the mean age of occurrence of thrombosis was 22. ii) Fun_ctional protein S and-free antigenic protein S deficiencies with normal total antigenic protein S. This group, type I deficiency according to Comp ef al. (13), represents 65 % of PS deficiencies. The mean level of total antigenic PS was 0.9 f 0.23 units/ml, whereas the mean level of functional PS was reduced to 0.37 f 0.2 units/ml and that of free antigenic PS was lowered in parallel to 0.34 + 0.13 units/ml. Mean level of C4b-bp was within normal range : 1.15 + 0.21 U/ml. All patients also suffered from severe thrombosis except children under 15. In both groups - 25 % of the patients had arterial thrombosis. In order to assess the sensitivity of functional PS to oral anticoagulant treatment, 10 anticoagulated patients were tested (INK = 3.45 + 0.68 ). Mean total PS antigen was 0.65 f 0.12 units/ml, mean free PS antigen 0.39 + 0.11 units/ml and functional PS 0.34 + 0.079 units/ml. DISCUSSI,oN Since PS functions as a cofactor for APC, its reduction or absence should result in a shortening of the clotting time of plasma exposed to this potent natural anticoagulant. The design of a functional PS assay based on the use of a PSdeficient substrate plasma implies the adjustment of several parameters in order to correlate the prolongation of the clotting time to the amount of free PS in the plasma to be tested. Furthermore the variations of factors V and VIII, the precursors of the substrates of activated PC, and/or of other vitamin K- dependent factors should have a minimal influence. The present assay takes into account the sole inactivation of factor Va by APC since coagulation is initiated by factor Xa. It appears to be specific as PSdeficient plasma behaved as normal reference plasma when resupplemented with physiological levels of free PS (- 10 pg/ml, fmal concentration) at comparable dilutions. A prolongation of the clotting time of - 100 s in the presence of 100 % PS allows a reliable determination of functional PS in patients’plasma. Linearity between prolongation of the clotting time and PS amount (Fig. 1) probably allows a more precise measurement of low levels of PS (i.e. < 0.2 U/ml) than the immunological assay. This could be responsible for the discrepancies between functional and antigenic PS in this area of low PS levels (Fig. 2). In addition, as the lowest dilution of reference plasma in PS deficient plasma corresponding to 100 % PS is 1:8, this should minimize the effects of possible abnormal levels of factor V or vitamin K-dependent factors in sample plasmas. Dilutions were made just before assay as advance preparation could have resulted in an apparent diminution of free PS due to possible complexation with the 25% remaining C4b-bp in PS deficient plasma. Another advantage of this functional assay is its rapidity of execution since it is almost as fast as other routine coagulation tests. In our experience PS deficiency seems to be more threatening for thrombosis than PC deficiency which appears to be largely asymptomatic (27). Therefore a rapid diagnosis should be of great help in the decision to prescribe anticoagulant treatment in order to diminish the risk of (re)thrombosis. The only limitation of the assay is the stability of some labile preparations, in particular APC and factor Xa. For these compounds storage at -8O’C is crucial as welI as their use within one hour after thawing. To our knowledge there is no available commercial preparation of human factor Xa or APC of established stability. Preparations from
SCREENING OF PROTEIN S DEFICIENCY...
466
Vol. 58, No. 5
bovine origin might appear to be suitable, however our first aim was to measure PS activity in a human homogeneous system. The striking observation is the high frequency, _ 25 %, of arterial thrombosis in patients with PS deficiency, all of them being under predisposing conditions. It is clear that, in our study, type I deficiency (with normal total antigenic PS and decreased functional and free antigenic PS) is the most frequent anomaly. This emphasizes the need for rapid measurement of free PS in plasma of patients with thrombosis. The present functional assay fulfils this requirement. A further immunological determination or confirmation should allow the classification of the patients with functional PS defect, taking into account those with mutant PS since the functional assay should detect them. This information is of prime importance in the understanding of the complex transmission of these defects as two genes have been identified for PS, one being a pseudogene (28). Acknowledgments We thank Dr. J.N. Mulvihill for reviewing the English of this paper and Mrs. E. Schoettel and C. Helbourg for typing the manuscript. This work was partly supported by the Fonds Regional d’Alsace de Recherche-Developpement.
1.
DI SCIPIO R.G., HERMODSON M.A., YATES S.G., DAVIE E.W. A comparison of human prothrombin, Factor IX (Christmas Factor), Factor X (Stuart Factor) and protein S. Biochemistry, 16,698-706,1977.
2.
WALKER F.J. Regulation bovine protein S. J,Biol.
3.
WALKER F.J., SEXTON P.W., ESMON C.T. The inhibition of blood coagulation by activated protein C through the selective inactivation of activated factor V. B_ioch&n. Biophys._Acta, 571,333-342, 1979.
4.
WALKER F.J., CHAVIN S.I., FAY P.J. Inactivation of factor VIII by activated and protein S. Arch.Biochem. Biophys., 252,322-328, 1987.
5.
D’ANGELO A., LOCKHART M.S., D’ANGELO S.V., TAYLOR F.B. Protein S is a cofactor for activated protein C neutralization of an inhibitor of plasminogen activation released from platelets. Blood, 69,231-237,1987.
6.
WALKER F.J. The regulation of activated protein C by protein S . The phospholipid in factor Va inactivation. J. ~Biol?_Chern_ 256,11128- 1113 1, 1980.
7.
SOLYMOSS S., TUCKER M.M., TRACY P.B. Kinetics of inactivation of membranebound factor Va by activated protein C. Protein S modulates factor Xa protection. J. Biol, Chem., 263,14884-14890,1988.
8.
DAHLBACK B., STENFLO J. High molecular weight complex in human plasma between vitamin K-dependent protein S and complement component C4b-binding protein. Prqc. NatI. Acad Sci. US-A, 78,2512-2516,198l.
9.
DAHLBACK B. Purification of human C4b-binding protein and formation with vitamin Kdependent protein S. B&hem. J., 209,847-856,1983.
10.
DAHLBACK B. Inhibition of protein Ca cofactor function by C4b-binding protein. J, Biol.Chem,, 261,12022-12027,
of activated protein C by a new protein. Chem., 2555521-5524, 1980.
A possible
function
for
protein C
role
of
of its complex
of human and bovine protein 1986.
S
Vol. 58, No. 5
SCREENING OF PROTEIN S DEFICIENCY...
467
11.
COMP P.C., NIXON R.R., COOPER M.R., ESMON C.T. Familial protein S deficiency is associated with recurrent thrombosis. J. Clin. Invest., 74,2082-2088, 1984.
12.
SCHWARZ H.P., FISCHER M., HOPMEIER P., BATARD M.A., GRIFFIN J.H. Plasma protein S deficiency in familial thrombotic disease. Blood, 64,1297-1300,1984.
13.
COMP P.C., DORAY D., PATTON D., ESMON CT. Abnormal plasma distribution of protein S occurs in functional protein S deficiency. Blood, 67,504-508, 1986.
14.
MANNUCCI P.M., TRIPOD1 A., BERTINA R.M. Protein S deficiency associated with “juvenile” arterial and venous thrombosis. Thromb. Haemostas,, 55,440 (abstr.) , 1986.
15.
COLLER B.S., OWEN J., JESTY J., HOROWITZ D., REITMAN M.J., SPEAR J., YEH T., COMP P.C. Deficiency of plasma protein S, protein C or antithrombin HI and arterial thrombosis. Arteriosclerosis, 7,456462, 1987.
16.
ISRAELS S.J., SESHIA S.S. Childhood stroke associated with protein C or S deficiency. J. Pediat., 111,562-564, 1987.
17.
GIROLAMI A., SIMIONI P., LAZZARO A.R., CORDIANO I. Severe arterial cerebral thrombosis in a patient with protein S deficiency (moderatly reduced total and markedly reduced free protein S) a family study. Thromb. Haemostas., 61,144147,1989.
18.
SIE P., BONEU B., BIERME R., WIESEL M.-L., GRUNEBAUM L., CAZENAVE Arterial thrombosis and protein S deficiency. Thromb. Haemostas., 62, 1040, 1989.
19.
SUZUKI K., NISHIOKA J. Protein S activity measured using Protac, a snake venom derived activator of protein C. Thm_mb,Res., 49,241-251, 1988.
20.
VAN DE WAART P., PREISSNER K.T., BECHTOLD J.R., MULLER-BERGHAUS functional test for protein S activity in plasma. Thromb. Res., 48,427-437, 1987.
21.
KOBAYASHI I., AMEMIYA N., END0 T., OKUYAMA K., TAMURA K. KUME S. Functional activity of protein S determined with use of protein C activated by venom activator. Clin. Chem., 35, 1644-1648, 1989.
22.
FREYSSINET J.-M., SCHUHLER S., SCHUHLER E., GAUCHY J., RAVANAT C., WIESEL M.L., GRUNEBAUM L., STIERLE A., FOLLEA G., CAZENAVE J.-P. Vitamin K-dependent proteins of plasma : structure-function relationships, purification and therapeutic applications. In: Biotechnology of Plasma Proteins. J.F. Stoltz, C. Rivat (Eds.). Paris : Colloque Inserm (175), 1989, pp. 305-314.
23.
FREYSSINET J.-M., GAUCHY J., CAZENAVE J.-P. The effect of phospholipids on the activation of protein C by the human thrombin-thrombomodulin complex. B&hem. J., 238,151-157,1986.
24.
FREYSSINET J.-M., WIESEL M.-L., GRUNEBAUM L., PEREILLO J.-M., GAUCHY J., SCHULER S., FREUND G., CAZENAVE J.-P. Activation of human protein C by blood coagulation factor Xa in the presence of anionic phospholipids : enhancement by sulphated polysaccharides. Biochem. J., 261,341-348,1989.
25.
WIESEL M.-L., GRUNEBAUM L., FREYSSINET J.-M., CAZENAVE causes of familial thrombosis. Fibrinolysis, 2, supp.2, 14-15, 1988.
26.
HEEB M.J., GRIFFIN J.H. The biochemistry of protein S. In: Protein C and Related Proteins. R.M. Bertina (Ed.) Edinburgh : Churchill Livingstone, 1988, pp. 55-70.
J.-P.
G. A
J.-P. Biological
468
SCREENING OF PROTEIN S DEFICIENCY...
Vol. 58, No. 5
27.
MILETICH J., SHERMAN L., BROZE G. Jr. Absence of thrombosis in subjects with heterozygous protein C deficiency. N. Engl. J. Med., 317,991~996, 1987.
28.
PLOOS VAN AMSTEL J.K., REITSMA P.H., BERTINA R.M. Identification protein S p gene as a pseudogene. Thromb. Haemostas., 62,272 (abstr.), 1989.
of the
