Blut (1986) 52:169-177

Blur

© Springer-Verlag 1986

A Simplified Functional Assay for Protein C in Plasma Samples Wilfried Thiel, Klaus T. Preissner, Ulrich Delvos, and Gert Miiller-Berghaus Clinical Research Unit for Blood Coagulation and Thrombosis of the Max-PlanckGesellschaft and Department of Medicine, Justus-Liebig-Universit~it, Giessen, Federal Republic of Germany

Summary. The important role of protein C (PC) in the regulation of hemostasis has been appreciated since the description of patients who were deficient in PC and presented with severe thromboembolic events. The potentially fatal complications associated with PC-deficiency require an early and reliable identification of those patients affected with this inherited disorder. The present study introduces a test procedure for the functional assessment of PC in plasma samples. The test utilizes the thrombin/thrombomodulin complex to achieve complete and rapid formation of activated PC whose proteolytic capacity is subsequently determined with a chromogenic substrate. Homogenate obtained from rabbit lung effectively substituted the purified component thrombomodulin in the assay system. This new approach simplifies the test procedure without losing specificity and accuracy. Proteases, such as plasmin, streptokinase and urokinase did not influence the assay and the inhibitory effect of heparin on the PC-activation could easily be overcome by the addition of protamine sulphate. The PC-activity in a group of unselected patients (n = 50), who did not reveal any abnormalities in global coagulation tests, amounted to 100 + 12% (mean + SD) with a range from 54 to 143°70 when analyzed in comparison to a plasma pool constituted from healthy volunteers. Since the synthesis of PC depends on the availability of vitamin K, patients receiving phenprocoumon have also been analyzed. These patients (n = 103) presented 40 + 11% residual PC-activity accompanied by a concomitant decrease in PC-antigen levels to 43 + 10% (mean + SD). The test described is specific, sensitive, less time-consuming and can be performed on a routine basis. Key words: Protein C - Functional test - Tendency to thrombosis For years clinicians have been searching for blood tests that might lead to the identification of patients who are at high risk of developing thrombosis. Several hereditary deficiencies and abnormalities of coagulation proteins have consistently been shown Offprint requests to: Dr. G. M~ller-Berghaus, Max-Planck-Gesellschaft, GaffkystraBe 11,

D-6300 Giessen, FRG

170

W. Thiel et al.

to predispose to venous thrombosis [14]. The recent observation that patients who were suffering from recurrent thrombosis and thromboembolic complications were congenitally deficient in Protein C (PC), has provided compelling evidence for an important regulatory role of PC in hemostasis [2, 4, 12]. While this autosomal inherited disorder in its heterozygous form is frequently associated with thrombotic events, the homozygous state presents as a purpura fulminans in the affected newborn [3, 16, 20]. P C is a circulating zymogen of a serine protease whose synthesis in the liver depends on the presence o f vitamin K. Thrombin is the only enzyme so far known in vivo that generates activated Protein C (APC). While its activation capacity is rather ineffective in the fluid phase, activation occurs very efficiently when thrombin is bound to the cofactor thrombomodulin at the endothelial cell surface [8]. A P C exerts its anticoagulant effect by inactivating coagulation factors V a and V i l l a [11, 15,22,26] and presumably by decreasing the level of the plasminogen activator-inhibitor [13]. The enzymatic activity of A P C is markedly enhanced by a protein cofactor, Protein S [25], whose synthesis also depends on the availability of vitamin K. Inhibition of A P C in plasma is achieved by a rather slowly occurring complex formation between A P C and its specific inhibitor, APC-inhibitor [21]. The frequently fatal outcome of thromboembolic complications requires reliable tests to identify those patients who are deficient in PC. Since the exclusive determination of antigen levels [1] is not sufficient, additional tests are necessary to obtain adequate information about the function of the molecule. At least four different functional assays for PC have been described [2, 6, 10, 18], which are unfortunately restricted by some severe limitations. Either the assay procedure is rather imprecise, because activation of PC is performed in the absence of thrombomodulin and, thus, only a small part of the PC present in the plasma sample is activated in an acceptable time period [2], or it is rather unspecific, because the prolongation of the activated partial thromboplastin time is employed as a test system [10]. Other approaches are confined to few specialized centers, because the use of purified thrombomodulin [18] is required, or an additional immunoadsorption with antibodies against PC is included in the procedure [6]. The present study describes an assay procedure for the functional activity of PC that appears to be specific, sensitive and reproducible and that, in addition, can routinely be applied in the clinical laboratory.

Materials and Methods Materials

Bovine thrombin and antithrombin III (AT III) were obtained from Behringwerke, Marburg, and hirudin from Pentapharm, Basel (Switzerland). The chromogenic substrate S 2266 was purchased from Kabi, Stockholm (Sweden). All other reagents were obtained from Sigma, Mt~nchen. Human PC was isolated from human plasma using a modification of the purification procedure of Bertina et al. [1]. On polyacrylamide gels in the presence of SDS, the purified protein migrated as a single band (Mr = 60,000) under non-reducing conditions and showed two bands (Mr = 40,000, 18,000) after reduction. Factor X, Protein S and prothrombin were absent from the purified preparation [17]. As source for the cofactor thrombomodulin, rabbit lungs were homogenized in 20 mM Tris buffered saline (pH7.5) and centrifuged at 15,000xg for 25 min at 4°C. The pellet resulting from one lung was extracted twice with 50ml of the same buffer that contained in addition 0.5%

Functional Assay for Protein C

171

Triton X,100 (v/v). The clear supernatant obtained after centrifugation of the second extract was frozen in aliquots and stored at - 7 5 ° C . The ability of the endothelial cell cofactor to inhibit thrombin clotting activity was employed for the standardization of the lung homogenate [7]. 50 gl of the lung homogenate increased the clotting time of a fibrinogen solution (2 mg/ml) from 15 s to more than 100 s when incubated with 0.12 NIH U bovine thrombin. PC-deficient plasma was prepared by two consecutive incubations of normal human plasma with 25mg/ml AI(OH)3 for 30 rain. The final supernatant exhibited less than 3O7o of PCantigen levels of normal plasma. Patients

Two groups of patients and in addition, healthy volunteers (n = 35), who were recruited from the laboratory staff, have been analyzed in this study. The first group of patients (n = 50) was routinely screened before undergoing surgery and showed no abnormalities in the global coagulation tests. The second group (n = 103) was under treatment with phenprocoumon for various thromboembolic events in the past and was routinely monitored by the coagulation laboratory of the Department of Medicine. Blood obtained by venepuncture was anticoagulated with 11 mM trisodium citrate (final concentration) and subsequently centrifuged twice for 15 min at 3,000 × g to obtain platelet-poor plasma. Aliquots were stored frozen at - 7 5 ° C until assayed. Heparinized plasma was mixed with 2 U/ml protamine sulphate (Hoffmann-La Roche, Basel, Switzerland) prior to analysis. For the preparation of serum non-anticoagulated blood was allowed to clot for 2 h at 37 °C and subsequently centrifuged twice to obtain serum. Methods

Prothrombin times were measured using the Hepato-quick reagent containing brain thromboplastin (Boehringer, Mannheim). PC-antigen levels in purified systems and in plasma were determined by a commercially available ELISA kit (Boehringer, Mannheim). Functional assay for PC in plasma: Plasma samples were thawed at 37°C and subsequently cooled to 4°C. The assay procedure of Sala et al. [18] was adopted and modified in the following way: i ml plasma was mixed with 100gl 1 M BaCI/and kept for 10 rain at 4°C. After centrifugation, the resulting pellet was washed with 250 gl Tris-buffer (20mM, pH 8.0) containing 0.15 M BaC12 and 0.15 M NaC1, and thereafter the solution was again centrifuged. Subsequently, proteins were eluted from the barium citrate pellet with 250 gl 0.15 M EDTA (pH 6.0) containing 0.15 M NaC1, 0.5% (w/v) bovine serum albumin (BSA), 0.02% NaN 3. The eluted proteins were precipitated by the addition of 250 ~tl 40% polyethyleneglycol 8,000 for 10 min at 4°C and recovered by centrifugation at 40,000xg for 30 min at 4°C. The pellet was dissolved in 420~tl Tris-buffer (20raM, pH8.0) containing 0.15M NaC1, 10mM CaCI2, 0.l% (w/v) BSA, 0.1% (v/v) Triton X-100. Aliquots of 200 gl were immediately assayed or kept frozen at - 7 5 ° C . A mixture of 5 gl bovine thrombin (25 NIH U/ml) and 50 Ixl rabbit lung homogenate was added to 200 ~tl of the solution to initiate activation of PC. The final concentrations present in the reaction mixture were as follows: 8raM CaCI2, 0.33% (w/v) BSA, 0.1% (v/v) Triton X-100. After 30 min incubation at 37°C the activation of PC was terminated by the addition of 0.25 U ATIII and 1 ATU hirudin in 45 gl 0.15 M NaC1. After addition of 655 gl assay buffer (25 mM Tris, 0.15 M NaC1, pH 8.0) and preincubation for 3 min at 37°C the reaction solution was transferred quantitatively into a plastic cuvette and 50 gl of the chromogenic substrate S 2266 (4 mM) was added. The release of para-nitroaniline was determined at 37°C by continuously recording the absorbance at 405nm for at least 3 min. The values of AA405xmin -1 were converted to PCactivity by using a calibration curve that had been constructed with mixtures of normal and PCdepleted plasma. Control samples which did not contain thrombin and lung homogenate were read as blanks and gave negligible absorbance.

172

W.Thiel et al.

Results

Specificity and Reliability of the Assay Procedure More than 90°7o of PC-antigen present in the samples was recovered by adsorption with barium citrate and 8 0 - 8 5 % could be converted to APC as determined by immunoassay at each step of the assay procedure. The time course of the activation of C obtained after barium citrate adsorption of normal citrated plasma is shown in Figure 1. As maximal activation was achieved after 30 min, this time period was chosen for routine analysis. The presence of heparin in the plasma sample abolished the activation of PC. Addition of protamine sulphate could prevent this effect and an activation profile was generated comparable to that obtained with citrated plasma (Fig. 1). The presence of urokinase (100U/ml), streptokinase (100U/ml) or plasmin (1 U/ml) did not effect the activation of PC within the assay procedure (Table 1). Furthermore, when plasma was depleted of PC by adsorption with AI(OH)3 no amidolytic activity could be generated in the residual supernatant. A linear correlation between the amidolytic activity and the concentration of purified PC or between the different mixtures of normal plasma and PC-depleted plasma was documented (Fig. 2). PC-activity measured correlated well with the amount of antigen present in these samples. The reproducibility of the PC assay was assessed by performing duplicate measurements of a normal plasma pool on 12 consecutive days as well as by performing 12 determinations on one day. The variation coefficient within the assay amounted to 6% and between the assays to 7.5%.

Protein C-Activity in Healthy Individuals and in Phenprocoumontreated Patients The plasma of 35 healthy volunteers (12 female, 23 male, age between 19-37 years) was analyzed for PC-activity and PC-antigen levels. PC-activity was 97 + 8% (mean +_ SD) covering a range from 72 to 109%. PC-antigen levels ranged from 75 to 132%, the mean value being 98 _+ 14%. The latter value corresponded to 4.35 +_ 0.61 ~tg/ml PC-antigen as determined from a calibration curve obtained with purified human PC. The residual PC-activities and the PC-antigen levels have also been determined in the serum of the same donors. A slight decrease of only 13°70 and 10%, respectively, could be documented in these samples (Table 2). During the study, plasma pools, each composed from 10 of those healthy individuals, served as controls (100°70). The PC-activities and PC-antigen levels of patients (n = 50) who revealed no abnormalities in the global clotting tests, showed a considerable larger variation of the values. While the PC-activity amounted to 100 + 12% (mean _+ SD), range: 54 to 143°70, the PC-antigen determined was 88 + 18% with a range from 68 to 125%. In 103 patients receiving phenprocoumon PC-activity was analyzed to be 40 + 11% and PC-antigen to be 43 + 10% (mean + SD). The correlation between levels of functional active PC and PC-antigen levels was found to be linear with a correlation coefficient of r = 0.85 (Fig. 3). A histogram of the levels of functionally active PC in norreals and patients under therapy with phenprocoumon is depicted in Figure 4 (panel A) and is compared with the ratio of PC-activity to PC-antigen obtained from each individual group (panel B). While the differences in PC-activities are apparent, the values of the ratios approximate to 1.0 in both groups. Since the intensity of the anticoagulant therapy was highly variable, patients were divided into subgroups. Pro-

Functional Assay for Protein C

173 0.2n=8 y= 0 . 0 0 1 7 x - 0.0038

0.15-

=

0.15'E E ×E •

E 0.1o

; 0.1"~

xl

0.05.

.~ o.o5<

0 10

20

1

t 40

30

Activation

i

50

0

I 60

12.5

2

t i m e (min)

25

50

75

100

Normal human plasma (%)

Figs. 1, 2. 1 Time course of the activation of Protein C (PC) by thrombin/lung homogenate in the presence of CaC12 (8 raM) at 37°C. The source of PC was either citrated plasma (closed circles), citrated plasma mixed with 1 U/ml heparin (open circles), or citrated plasma mixed with 1 U / m l heparin and 2 U / m l protamine sulphate (open squares). Complete activation was achieved after 30 rain. The inhibitory effect of heparin on the activation of PC was completely reversed by the addition of protamine sulphate (mean, n = 3); 2 Calibration curve for the PCactivity in plasma samples. Mixtures of normal and PC-depleted plasma were analyzed for PCactivity. The absorbance measured was plotted against percent of normal human plasma and showed a linear relationship (mean _+ SD, n = 8)

P h e n p r o c o u m o n - T r e a t ed Patients 1.0-~

140 -

n=103 y=0.98x-0.016 r=0.85

120 -

0.8-

0 ~

,

~ 2.0-

i"

~

1.6-

~.

0

1.6-

100 ""

~ 3"

A •o

"~{t

,,

S 1.4-

t;.

80-

0.6o =

~.

o

,,,

--> t . o -

:" 60~| ?.

0,4-

0 •

";~L~I*" 0,2-

20..

0

0

i 0,2

i 0.4

i

0.6

i 0.8

Protein C antigen (unitslml)

Z:, ooQ

1.2-

Norma/s

i 1.0

4

(n=50)

Phenproeoumon treated (n=I03)

0.8 -

"o

::~.:-.... Q"

0.6-

~

0.4-

.2 ~ 0,2 0

Normals

(n=50)

Phenprocoumontreated (n=103)

Figs. 3, 4. 3 Correlation between the PC-activity and the PC-antigen levels in the plasma of patients (n = 103) anticoagulated with phenprocoumon; 4 Histogram of the levels of functionally active ~C in normals and patients under therapy with phenprocoumon (panel A), the ratio of PC-activity to PC-antigen obtained from each individual is depicted in panel B. While the PCactivities of the two groups markedly differ, a difference for the ratios of PC-activity to PCantigen between the groups is not apparent

174

W. Thiet et al.

Table 1. Effect of proteases on the activity of Protein C (PC) in plasma samples Fibrinolytic agent Plasmin Streptokinase Urokinase Control a

(1 U/ml) (100 U/ml) (100 U/ml)

n

PC-activity (%)a

5 7 6 10

95 97 98 96

+ 9 + 7 + 6 +_ 9

Mean + SD

Table 2. Activities of Protein C (PC) in plasma and serum Sample

PC-activity (Range) (%)a

PC-antigen (Range) (%)a

Plasma Serum

97 + 8 (72-109) 84 + 6 (77-95)

98 +_ 14 (75-132) 88 + 13 (66-120)

a Mean + SD, n = 3 5

Table 3. Analysis of Protein C (PC)-antigen and PC-activity according to the intensity of the oral anticoagulant therapy expressed as International Normalized Ratio (INR) INR

n

PC-activity (O/o)a

PC-antigen (%)a

PC-activity PC-antigen

Total group 1.7-8.1

103

40.5 + 11.2

43.5 + 10.1

0.93

Subgroups below 2.7 2.7-4.7 above 4.7

34 46 23

48.1 + 12.8 37.8 + 11.2 35.9 + 7.5

48.7 + 9.4 41.2 + 10.9 39.6 + 9.6

0.98 0.91 0.90

a Mean +_ SD

thrombin times were expressed as International Normalized Ratio (INR) and three subgroups were arranged according to the following INR: INR ~ 2.7 (above), INR > 4.7 (below) or 2.7 ___ INR > 4.7 (within the therapeutic range). The results obtained are summarized in Table 3. A significant difference between the ratios of PC-activity to PC-antigen of different subgroups was,not observed, however.

Discussion The test described here allows reproducible and reliable measurements of the activity of P C in plasma samples. In the first step of the procedure, the carboxylated mole-

Functional Assay for Protein C

175

cules are separated from the noncarboxylated PC-molecules. This separation is essential since the noncarboxylated proteins have also been shown to be activated by the thrombin/thrombomodulin complex [9]. Furthermore, the APC-inhibitor present in plasma is removed by this precipitation procedure [21]. The use of the thrombin/ thrombomodulin complex in the presence of calcium ions accelerates the activation of PC nearly 20,000-fold, when compared to the activation of PC in the presence of thrombin only [8]. Homogenate obtained from rabbit lungs has been used as a source of the endothelial cofactor thrombomodulin and replaced effectively the purified component employed by other investigators [6, 18]. About 1,000 assays can be performed with the homogenate obtained from a single rabbit lung. Since the specificity of the assay to activate PC was not dependent on Protein S, as was ascertained in adsorbed plasma to which purified PC had been added, construction of calibration curves with mixtures of normal and Al(OH)3-adsorbed plasma reflect the appropriate levels of PC. Although heparin interfered with the PC-assay, probably due to its co-adsorption to the barium citrate precipitate, PC-activities can also be determined in heparinized plasma samples, as the effect of heparin can be overcome by protamine sulphate. In addition, plasma samples of patients who are undergoing a fibrinolytic treatment with streptokinase or urokinase can be analyzed without any interference to the assay procedure. Thus, the assay is applicable in many clinical settings to analyze PC-activity in patients with hemostatic disorders. Since the functional activity of PC is nearly completely retained in serum (84%), even a determination of PC-activity in the patient's serum sample may abready indicate if a PC-deficient state is present. A decreased PC-activity was found in the plasma of patients receiving therapy with phenprocoumon. This finding is not surprising and can easily be explained by the inability of the liver cell to carboxylate the protein in the presence of the vitamin K inhibitor phenprocoumon. Concomitant reduction of synthesis was also observed as documented by a decrease in PC-antigen levels [4]. The splitting of patients into three subgroups according to the intensity of the oral anticoagulant therapy did not reveal any significant changes in PC-activities or the ratio PC-activity to PC-antigen. At least two possibilities have to be discussed to explain these findings. Firstly, the relatively short half-life of PC of about 8 h [23], which is comparable to that of factor VII, may enable the protein to recover more rapidly than the other coagulation factors which influence the duration of the prothrombin time. Secondly, it appears possible that partly carboxylated PC-molecules (PIVKA-PC) may become adsorbed to the barium citrate precipitate and thus, could be activated by the thrombin/thrombomodulin complex, thereby leading to an increase in PC-activity. A comparison of the reported values with data obtained by a clotting assay for APC [24] may help to clarify this phenomenon in the future. The more important conclusion for the clinician is, of course, that an identification of a patient, congenitally deficient in PC, can not be made under treatment with phenprocoumon, since the drug mimics exactly the PC-levels which can be expected in this situation. One has also to keep in mind that normal PC-activities in patients do not exclude the possibility that the action of APC is minimized due to a deficiency of the cofactor Protein S [5, 19].

Acknowledgement. We thank Iris Stehl and Susanne Will for their skillful technical assistance. This study was supported by the Stiftung Volkswagenwerk, Hannover.

176

W. Thiel et al.

References 1. Bertina RM, Broekmans AW, van der Linden IK, Mertens K (1982) Protein C deficiency in a dutch family with thrombotic disease. Thromb Haemostas 4 8 : 1 - 5 2. Bertina RM, Broekmans AW, Krommenhoek-van Es C, van Wijngaarden A (1984) The use of a functional and immunologic assay for plasma protein C in the study of the heterogeneity of congenital protein C deficiency. Thromb Haemostas 5 1 : 1 - 5 3. Branson H, Katz J, Marble R, Griffin JH (1983) Inherited protein C deficiency and coumarin-responsive chronic relapsing purpura fulminans in a newborn infant. Lancet II: 1165-1168 4. Broekmans AW, Veltkamp J J, Bertina RM (1983) Congenital protein C deficiency and venous thromboembolism. N Engl J Med 309:340-344 5. Comp PC, Nixon RR, Cooper MR, Esmon CT (1984) Familial protein S deficiency is associated with recurrent thrombosis. J Clin Invest 74:2082-2088 6. Comp PC, Nixon RR, Esmon CT (1984) Determination of functional levels of protein C, an antithrombotic protein, using thrombin-thrombomodulin complex. Blood 63:15-21 7. Esmon CT, Esmon NL, Harris KW (1982) Complex formation between thrombin and thrombomodulin inhibits both thrombin-catalyzed fibrin formation and factor V activation. J Biot Chem 257:7944-7947 8. Esmon CT, Owen WG (1981) Identification of an endothelial cell cofactor for thrombincatalyzed activation of protein C. Proc Natl Acad Sci USA 78:2249-2252 9. Esmon NL, De Bault LE, Esmon CT (1983) Proteolytic formation and properties of y-carboxyglutamic acid-domainless protein C. J Biol Chem 258:5548-5553 10. Francis RB, Patch MJ (1983) A functional assay for protein C in human plasma. Thromb Res 32:605-613 11. Fulcher CA, Gardiner JE, Griffin JH, Zimmerman TS (1984) Proteolytic inactivation of human factor VIII procoagulant protein by activated human protein C and its analogy with factor V. Blood 63:486-489 12. Griffin JH, Evatt B, Zimmerman TS, Kleiss AJ (1981) Deficiency of protein C in congenital thrombotic disease. J Clin Invest 68:1370-1373 13. van Hinsbergh VWM, Bertina RM, Wijngaarden A, van Tilburg NH, Emeis JJ, Haverkate F (1985) Activated protein C decreases plasminogen activator-inhibitor activity in endothelial cell-conditioned medium. Blood 65:444-451 14. Hirsh J (1981) Blood tests for the diagnosis of venous and arterial thrombosis. Blood 57: 1-8 15. Kisiel W, Ericsson LH, Davie EW (1976) Proteolytic activation of protein C from bovine plasma. Biochemistry 15:4893-4900 16. Marcianiak E, Wilson HD, Marlar RA (1985) Neonatal purpura fulminans - a genetic disorder related to the absence of protein C in blood. Blood 65" 15-20 17. Preissner KT, Mtiller-Berghaus G (1985) The effect of heparin and antithrombin III on the protein C activation by thrombin/thrombomodulin (in prep) 18. Sala N, Owen WG, Collen D (1984) A functional assay of protein C in human plasma. Blood 63:671-675 19. Schwarz HP, Fischer M, Hopmeier P, Betard MA, Griffin JH (1984) Plasma protein S deficiency in familial thrombotic disease. Blood 64: 1297-1300 20. Seligsohn U, Berger A, Abend M, Rubin L, Attias D, Zivelin A, Rapaport SI (1984) Homozygous protein C deficiency manifested by massive venous thrombosis in the newborn. N Engl J Med 310:559-562 21. Suzuki K, Nishioka J, Hashimot0 S (1983) Protein C inhibitor, purification from human plasma and characterization. J Biol Chem 258:163-168 22. Suzuki K, Stenflo J, Dahlb/ick B, Teodorsson B (1983) Inactivation of human coagulation factor V by activated protein C. J Biol Chem 258:1914-1920 23. Vigano S, Mannucci PM, Solinas S, Bottasso B, Mariano G (1984) Decrease in protein C antigen and formation of an abnormal protein soon after starting oral anticoagulant therapy. Br J Haematol 57:213-220 24. Vigano S, D'Angelo A, Comp PC, Esmon CT (1985) A new functional assay for protein C: Assessment of protein C activity during oral anticoagulation. Thromb Haemostas 54: 107 (Abstract)

Functional Assay for Protein C

177

25. Walker FJ (1984) Protein S and the regulation of activated protein C. Sem Thromb Hemostas 10:131-138 26. Walker FJ, Sexton PW, Esmon CT (1979) The inhibition of blood coagulation by activated protein C through the selective inactivation of activated factor V. Biochim Biophys Acta 571:333-342 Received July 9, 1985/Accepted October 28, 1985