Clinical Chemistry 45:6 807– 813 (1999)
Enzymes and Protein Markers
Assay of Procarboxypeptidase U, a Novel Determinant of the Fibrinolytic Cascade, in Human Plasma Katinka A. Schatteman,1 Filip J. Goossens,1 Simon S. Scharpe´,2 Hugo M. Neels,2 and Dirk F. Hendriks1*
Background: Procarboxypeptidase U (proCPU) is a novel proenzyme found in human plasma. The active form, carboxypeptidase U (CPU; EC 126.96.36.199), retards the rate of fibrinolysis through its ability to cleave C-terminal lysine residues on fibrin partially degraded by plasmin. This reduces the number of high-affinity plasminogen-binding sites on fibrin. Methods: We developed an assay to determine the proCPU concentration in human plasma. The assay involved quantitative conversion of proCPU to active CPU by thrombin-thrombomodulin, a very efficient activator of proCPU, followed by determination of the enzymatic activity of CPU with the substrate hippurylL-arginine, using an HPLC-assisted determination of the released hippuric acid. Using this method, we established a reference interval based on 490 healthy individuals. Results: The mean proCPU concentration, determined after activation of the zymogen in diluted plasma and expressed as CPU activity, was 964 U/L, with a SD of 155 U/L. The population showed a gaussian distribution. However, we noticed important differences related to age and the use of hormone preparations. Conclusions: The sensitivity and precision of the method make it suitable for routine clinical determinations and as a reference procedure.
plasma compartment, this family of enzymes is represented by carboxypeptidase N (CPN,3 EC 188.8.131.52; also known as lysine carboxypeptidase, kininase I, anaphylatoxin inactivator, or plasma carboxypeptidase B) and the more recently identified carboxypeptidase U (CPU; EC 184.108.40.206; also known as carboxypeptidase R, plasma carboxypeptidase B, or TAFIa). We first reported in 1988 on the presence of a labile carboxypeptidase activity in fresh human serum that interfered with the assay of CPN (1–3 ). This novel carboxypeptidase activity was not detectable in human plasma, but appeared after the coagulation of blood. Because of the marked instability of this enzyme, we named it CPU (where “U” indicates unstable) (3, 4 ). Campbell and co-workers (5, 6 ) confirmed these findings and called the enzyme carboxypeptidase R, where “R” represents arginine. Eaton et al. (7 ) discovered it as a contaminant during the purification of a2-antiplasmin, utilizing plasminogen-Sepharose affinity chromatography. They cloned the CPU cDNA from human liver and deduced the amino acid sequence of the proenzyme, designating it plasma carboxypeptidase B because the active enzyme very much behaves like pancreatic carboxypeptidase B in terms of enzyme activity (7, 8 ). This name, however, is rather confusing because the term “plasma carboxypeptidase B” had long been used as a synonym for plasma carboxypeptidase N (9 ). In addition, pancreatic carboxypeptidase B, a pancreatic digestive carboxypeptidase, can be detected in the plasma compartment of patients with acute pancreatitis (10, 11 ). Wang et al. (12 ) confirmed the identity between proCPU and the
© 1999 American Association for Clinical Chemistry
Basic carboxypeptidases are a group of enzymes that cleave a single basic amino acid, lysine or arginine, from the COOH terminus of peptides and proteins. In the
1 Department of Clinical Biochemistry, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium. 2 Laboratory of Clinical Biology, OCMW Hospitals, Antwerp, Belgium. *Author for correspondence. Fax 32-3-820.27.45; e-mail [email protected]
uia.ua.ac.be. Received November 9, 1998; accepted March 24, 1999.
3 Nonstandard abbreviations: CPN, carboxypeptidase N; CPU, carboxypeptidase U; Hip-Arg, hippuryl-l-arginine; PPACK, d-phenylalanyl-lprolylarginyl chloromethyl ketone; and SDS-PAGE, sodium dodecyl sulfatepolyacrylamide gel electrophoresis.
Schatteman et al.: Assay of Plasma proCPU
protein isolated by Eaton et al. (7 ). We localized the subregional mapping of the human proCPU gene as locus q14.11 to chromosome 13 (13 ). Bajzar et al. (14 ) independently found this enzyme and showed that it can be activated by thrombin and that upon activation, it can inhibit fibrinolysis. Consequently, they named it thrombin-activatable fibrinolysis inhibitor. Recently, much attention has been focused on the role of CPU in the fibrinolytic system (Fig. 1). It had been postulated that CPU, generated during coagulation, dampens the fibrinolytic system by acting on fibrin that has been partially degraded by plasmin (4, 15–17 ). This hypothesis was substantiated when it was demonstrated that proCPU exhibits a strong affinity for plasminogen and that it can be activated by plasmin and thrombin (7 ). The thrombin-thrombomodulin complex is the most likely physiological activator because thrombomodulin increases the catalytic efficiency of CPU activation by thrombin by a factor of 1250 (18 ). Plasmin degradation of fibrin exposes C-terminal lysine residues that are essential for the high-affinity binding of plasminogen to fibrin and the subsequent activation of plasminogen to plasmin (19 –21 ). Thus CPU, activated locally through the action of thrombin-thrombomodulin, is able to control the rate of fibrinolysis by cleaving some of the generated C-terminal lysine residues (15 ). This effect was demonstrated directly in experiments involving confocal microscopy (22 ). Recent data indicate that CPU indeed plays an important role in plasminogen activation because it is able to delay tissue plasminogen activator-induced clot lysis in vitro and since inhibition or depletion of CPU enhances the rate of clot lysis (14 –16, 18, 22–24 ). In vivo evidence for a role of CPU in the fibrinolytic system is provided by two studies. In a canine model of intracoronary thrombosis, it was shown that CPU is formed in vivo and that increased
Fig. 1. CPU as a link between coagulation and fibrinolysis and its role in hemostasis. proCPU can be activated by plasmin and thrombin to its active form CPU but is activated most efficiently by the thrombin-thrombomodulin complex. It has been postulated that CPU generated during coagulation dampens fibrinolysis. This is accomplished by the cleaving of some of the C-terminal lysine residues that are exposed on fibrin after plasmin degradation. These C-terminal lysine residues are essential for the high-affinity binding of plasminogen to fibrin and thus are necessary for rapid plasmin formation and fibrin dissolution.
CPU activity is associated with prolonged time to reperfusion (17 ). In a study using a thrombosis model in rabbits, inhibition of CPU enhanced clot lysis (25 ). The aim of our present study was to design an assay to determine proCPU concentrations in human plasma by quantitatively converting proCPU to CPU by means of a standardized activation with the thrombin-thrombomodulin complex. Subsequently, the activity of the generated CPU would be measured by an HPLCassisted method. Because a recent report shows that the proCPU concentration in human plasma is a major determinant of the clot lysis time (26 ) and because no reference values for proCPU concentrations are known, we determined proCPU concentrations in a healthy population to establish a reference interval for this potential new marker of fibrinolytic capacity.
Materials and Methods plasma samples Plasma was collected from 490 healthy adult volunteers in accordance with NCCLS guidelines. The subjects were men, ages 20 – 61 years; women receiving hormonal therapy (contraception or climacteric substitution therapy), ages 20 – 60 years; and women not receiving hormonal therapy, ages 20 – 60 years. Blood was collected into buffered 32 g/L sodium citrate anticoagulant (0.109 mol/L sodium citrate), using an evacuated tube system with a final ratio of nine parts blood to one part buffered citrate, by volume. Plasma was prepared by centrifugation at 3000g for 15 min at room temperature and was stored at 270 °C until analysis. Control plasma was obtained by pooling citrated plasma from 10 healthy volunteers. Aliquots were stored at 270 °C.
Clinical Chemistry 45, No. 6, 1999
Inactivated serum—serum in which neither activatable proCPU nor basic carboxypeptidase activity could be detected—was prepared by incubating human serum for 12 h at 56 °C.
chemicals Hippuryl-l-arginine (Hip-Arg) was obtained from Bachem Feinchemicalien. Hippuric acid was obtained from Fluka. o-Methylhippuric acid was synthesized from glycine and o-methylbenzoylchloride (UCB) by a procedure analogous to that used for the synthesis of hippuric acid (27 ). Human thrombin was obtained from Sigma. Rabbit lung thrombomodulin was obtained from American Diagnostica. The thrombin inhibitor d-phenylalanyll-prolylarginyl chloromethyl ketone (PPACK) was obtained from Alexis Biochemicals. Aprotinin was obtained from Bayer. HEPES was purchased from Calbiochem. All other reagents used were of high purity grade and were from Merck.
instruments The HPLC system consisted of a model 302 pump, a model 303 solvent delivery system, a model 802 C manometric module, a model 401 dilutor, and a model 231 autosampling injector (all from Gilson), a model 450 ultraviolet detector (Waters Associates), and a 150 3 4.6 mm (i.d.) Spherisorb ODS-2, 5 mm, column (Alltech). All chromatography steps were performed on a Waters 650 Advanced Protein Purification System (Millipore). Columns and chromatography media were from Pharmacia. The materials used for electrophoresis were purchased from Bio-Rad Laboratories. An Eppendorf centrifuge 5417, with a fixed-angle F 45-24-11 rotor, or an Heraeus Megafuge 1.0 R centrifuge, with a bucket 2252 rotor, was used. For the evaporation of samples, a Savant Speedvac SC 100 (Savant Instruments) was used.
methods Purification of human proCPU. Human citrated plasma was diluted 1:1 with 50 mmol/L phosphate buffer, pH 7.5. Aprotinin was added to a final concentration of 14 mg/L. The plasma was then depleted of plasminogen by chromatography on lysine-Sepharose as described previously (28 ). All buffers used for the lysine-Sepharose column contained aprotinin in a final concentration of 1.4 mg/L. The eluate of the lysine-Sepharose column was used for further purification of proCPU, based on the method of Bajzar et al. (14 ). The purification yielded a single protein band at a molecular mass of 60 kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The enzyme was purified an average 24 000-fold with a specific activity of 300 kU/g. Protein concentrations were measured with the bicinchoninic acid detection reagent (Pierce Chemicals) or with Coomassie Brilliant Blue G-250, according to the method by Bradford (29 ). SDS-PAGE was performed at 200 V in 10% gels accord-
ing to the method by Laemmli (30 ). Gels (10%) were prepared on the day of electrophoresis, following the Bio-Rad protocol. Proteins were visualized by silver staining. After SDS-PAGE, the purified proenzyme was electroblotted (1 h at 150 mA) to a PVDF membrane. The membrane was subjected to Edman degradation in an automated sequencer (Applied Biosystems Model 470 A). NH2-terminal sequencing reveal the sequence FQSGQVLAALPRTSRQVQVL, which confirmed its identity as plasma carboxypeptidase B (7 ). Determination of carboxypeptidase activity. The carboxypeptidase activities were measured with an HPLC-assisted assay using the substrate Hip-Arg as described elsewhere (12, 31 ). The procedure was as follows. For the chromatographic assay, 10 mL of enzyme solution was added to 40 mL of substrate solution (30 mmol/L Hip-Arg, 50 mmol/L HEPES, pH 8.0). This mixture was incubated for 30 min at 37 °C; the reaction was then stopped by the addition of 50 mL of 1 mol/L HCl. After the addition of 10 mL of internal standard (o-methylhippuric acid), hippuric acid and omethylhippuric acid were extracted from the acidified solution into 300 mL of ethyl acetate by vortex-mixing for 30 s. The mixture was centrifuged for 2 min at 20 000g in an Eppendorf centrifuge. A 200-mL aliquot of the supernatant ethyl acetate layer was removed and evaporated to dryness. The dry residue was redissolved in 150 mL of the chromatographic mobile phase, and 15 mL was injected onto the column (the described volumes were used for determination of CPU in diluted plasma after activation). The mobile phase used was a mixture of 850 mL of 10 mmol/L potassium phosphate buffer, pH 3.5, and 150 mL of acetonitrile. The separation of hippuric acid from the substrate was obtained within 3 min by reversed-phase HPLC. Each batch of substrate was checked for the presence of free hippuric acid by assaying a sample blank (10 mL of distilled water added to 40 mL of substrate). A stock solution of the internal standard was prepared by dissolving 291 mg of o-methylhippuric acid in 25 mL of ethanol and diluting to 100 mL with distilled water. For the determination of CPU in diluted plasma (after activation), a 1:6 dilution of the internal standard stock solution in 250 mL/L ethanol was used. Duplicate determinations were performed for all enzyme assays. For the calibration procedure, four 1000-mL aliquots of the hippuric acid calibrator solution (22.5 mmol/L in ethyl acetate) were evaporated to dryness. Substrate (40 mL) was added to each tube, and the assay procedure was performed as described above, beginning with the addition of 10 mL of distilled water. For the assay, one unit of enzyme activity was defined as the amount of enzyme required to hydrolyze 1 mmol of substrate per minute at 37 °C under the conditions described. The carboxypeptidase activity (U/L) was calculated using the mean ratio of the peak heights of hippuric
Schatteman et al.: Assay of Plasma proCPU
acid and the peak heights of the internal standard of the sample, blank, and calibrator. Determination of optimal conditions for activation of proCPU. The activation of purified proCPU with thrombin in the presence of thrombomodulin was performed at room temperature and at 37 °C in the presence and absence of 5 mmol/L CaCl2. Different concentrations of thrombinthrombomodulin were tested. The following protocol was used. To 5 mL of purified proCPU (from different batches with a corresponding CPU activity of 900-2650 U/L), 10 mL of thrombin-thrombomodulin in 20 mmol/L HEPES, 5 mmol/L CaCl2, 0.1 mL/L Tween 80, pH 7.4, was added. After the mixture was incubated for time intervals of 1–120 min, the activation was stopped by the addition of 5 mL of 20 mmol/L PPACK. Subsequently, 5 mL of inactivated serum was added. In citrated plasma samples, CPU was determined after activation of the zymogen, proCPU, using a similar protocol. Optimal concentrations of thrombin-thrombomodulin were determined. The following protocol was used. Citrated plasma samples were diluted 20-fold in 20 mmol/L HEPES, pH 7.4. The thrombin-thrombomodulin complex was prepared in the same buffer as described above and mixed with an equal amount of 80 mmol/L CaCl2 in 20 mmol/L HEPES, pH 7.4. A 20-mL aliquot of this mixture was added to 20 mL of plasma diluted 1:20 and incubated at room temperature. The activation of proCPU in plasma was stopped by the addition of 10 mL of 20 mmol/L PPACK, and the activity was measured with Hip-Arg and analyzed by HPLC. To determine CPN activity, similar measurements were performed without the addition of thrombin-thrombomodulin. The CPU activity was the difference in the value obtained after activation minus the value obtained without activation; this activity was used as a measure for the proCPU concentration in plasma. The effect of sodium ions on the activation of purified proCPU and proCPU in diluted plasma was evaluated in the presence or absence of 150 mmol/L NaCl. Reference values of proCPU, determined after activation with thrombin-thrombomodulin. Citrated plasma from 490 healthy adults was analyzed in duplicate as follows. Plasma samples, stored at 270 °C, were thawed quickly at 37 °C until they had just thawed. To 20 mL of a plasma sample diluted 1:20 in 20 mmol/L HEPES, pH 7.4, was added 20 mL of a 1:1 mixture of thrombin-thrombomodulin and 80 mmol/L CaCl2 in 20 mmol/L HEPES, pH 7.4. The final concentrations of thrombin and thrombomodulin during activation were 0.4 kU/L and 16 nmol/L, respectively. After incubation for 10 min at room temperature, the activation was stopped with 10 mL of 20 mmol/L PPACK. Activity was measured with the substrate Hip-Arg, and the released hippuric acid was determined by HPLC. Statistical analysis was performed with SigmaStat, Ver. 1.
Table 1. Determination of optimal conditions for the activation of purified proCPU and proCPU in diluted plasma.a Ca21, mmol/L
CPU activity at 22 °C, U/L
CPU activity at 37 °C, U/L
381 6 58 922 6 23
457 6 3 332 6 31
a All measurements were performed in duplicate. Purified proCPU was activated with thrombin (0.2 kU/L) and thrombomodulin (16 nmol/L) for 10 min at room temperature or at 37 °C, in the presence or absence of 5 mmol/L Ca21 ions.
Results optimal conditions for activation of proCPU The results obtained with the different activation conditions are shown in Tables 1 and 2 and Fig. 2. Thrombin and thrombomodulin concentrations of 0.2 kU/L and 16 nmol/L, respectively, produced optimal activation for purified proCPU. No higher activities were obtained when the concentration of either compound was increased. On the other hand, when the thrombin and thrombomodulin concentrations were ,0.1 kU/L and ,8 nmol/L respectively, the activation potential was reduced substantially. Optimum activation was at room temperature in the presence of Ca21 ions (Table 1). Fig. 2 shows an activation profile of purified proCPU as a function of time. Maximal activities were obtained within 5–10 min at room temperature. SDS-PAGE revealed a complete conversion of proCPU to its active form. Upon longer activation (20 –100 min), a gradual loss of activity was detected, which was caused by conformational changes of CPU (18, 32 ) followed by further cleavage of CPU (7, 14, 32 ). Similar results were obtained for proCPU in plasma. Maximum activity was obtained with a concentration of 0.4 kU/L thrombin and 16 nmol/L thrombomodulin (Table 2), with activation times not exceeding 15 min. For purified proCPU and for proCPU in plasma, lower activities were obtained when activation was performed in the presence of sodium ions. This result is similar to the one observed for protein C and is explained by the fact that the activity and specificity of thrombin are controlled in an allosteric fashion by the binding of Na1 to a single Table 2. Activation of proCPU in diluted plasma, using different concentrations of thrombin-thrombomodulin, as assessed with an HPLC-assisted assay. Thrombin,a kU/L
CPU activity, U/L
0.025 0.05 0.1 0.2 0.4 0.4 0.8 1.6
2 4 8 16 32 16 16 16
322 6 7 582 6 15 720 6 72 828 6 32 900 6 16 887 6 25 884 6 43 898 6 41
a Thrombin and thrombomodulin concentrations are the final concentrations during activation. All measurements were performed in duplicate.
Clinical Chemistry 45, No. 6, 1999
without activation from the total carboxypeptidase activity obtained after activation.
Fig. 2. Activation of proCPU by the thrombin-thrombomodulin complex. Purified proCPU was activated with thrombin (0.2 kU/L) and thrombomodulin (16 nmol/L) at room temperature for several different time intervals. The reaction was then stopped by the addition of PPACK, and activity was measured using an HPLC-assisted assay. The indicated concentrations of thrombin and thrombomodulin are the final concentrations during activation.
site. The Na1-bound state is the fast form, whereas the Na1-free state is the slow form (33 ). This may imply that for the activation of proCPU, the slow form of thrombin, i.e., the Na1-free form, is important. In all additional experiments, purified proCPU was activated as follows: 10 mL of 0.3 kU/L thrombin-24 nmol/L thrombomodulin (the final concentration of thrombin was 0.2 kU/L, and the final concentration of thrombomodulin was 16 nmol/L) in 20 mmol/L HEPES, 5 mmol/L CaCl2, 0.1 mL/L Tween 80, pH 7.4, was added to 5 mL of purified proCPU. The activation was stopped after a 10-min incubation at room temperature by the addition of 5 mL of 20 mmol/L PPACK. Subsequently, 5 mL of inactivated serum was added and the carboxypeptidase activity was measured. The protocol for determining proCPU in plasma is described in Materials and Methods. The proCPU concentration, expressed as CPU activity in U/L after activation with thrombin-thrombomodulin, was obtained by subtracting the carboxypeptidase activity present in a sample
To study the linearity as a function of CPU activity, purified proCPU was added to control plasma. Dilutions were made in inactivated serum. Activation was as described above, and activity was measured. The linearity of the assay could be demonstrated by assaying serially diluted plasma with added proCPU (total CPU activity, 7885 U/L). Linear regression analysis gave r 5 0.999 for dilutions of 1:128, 1:64, 1:32, 1:16, 1:8, 1:4, and 1:1 and undiluted proCPU (activity, 62–7885 U/L). The linearity as a function of incubation time was checked by measuring the activity at regular time intervals ranging from 10 to 120 min. Linear regression analysis gave r 5 0.999 (time, 0 –50 min). Linearity diminished after incubation time exceeded 70 min. Although the half-life of CPU at 37 °C is short (t1/2 5 10 min) (2, 32 ), a stabilizing effect of the substrate during assay was observed. This is in accordance with our previous results (3, 4 ).
reproducibility of the proCPU assay We evaluated the within- and between-day precision of a control plasma. The mean value of a control plasma, measured on different days (n 5 11) was 1040 6 48 U/L (between-day CV, 4.6%). The mean value of a control plasma, measured several times on the same day (n 5 10) was 1034 6 31 U/L (within-day CV, 3.0%)
sensitivity The HPLC method described is able to detect enzyme activities as low as 1 U/L. For the measurement of very low carboxypeptidase activities, a slightly modified procedure (longer incubation time and different volume ratio of substrate to sample ) was used. If a dilution value of 20 (dilution of the plasma sample) 3 2.5 (dilution upon activation) is used for the measurement of CPU in diluted plasma, an activity of 50 U/L plasma can be detected.
Table 3. Reference values of proCPU in human plasma. Activity,a U/L Men Age, years
20–29 52 30–39 49 40–49 57 50–61 58 Subtotal for women
Mean (SD) [range]
904 930 995 1004
(150) [421–1264] (124) [645–1245] (158) [440–1373] (156) [661–1405]
Women not receiving hormone therapy n
32 40 38 34 144
961 (153) [421–1405]
Mean (SD) [range]
920 886 875 1052 930
(123) [638–1109] (140) [475–1234] (136) [546–1102] (142) [790–1268] (152) [475–1268]
Women receiving hormone therapy n
29 32 31 38 130
Mean (SD) [range]
987 1036 1003 997 1006
(126) [768–1307] (186) [698–1516] (164) [679–1367] (126) [747–1245] (152) [679–1516]
966 (156) [475–1516]
Carboxypeptidase activity was measured after activation of proCPU in diluted plasma with thrombin-thrombomodulin and without activation with thrombinthrombomodulin. CPU activity was calculated by subtracting the carboxypeptidase activity without activation from the carboxypeptidase activity obtained after activation with thrombin-thrombomodulin.
Schatteman et al.: Assay of Plasma proCPU
reference values Reference values of proCPU (expressed as CPU activity in U/L after activation with thrombin-thrombomodulin) from a population of 216 men and 274 women were determined. The mean values, SDs, and minimums and maximums of the different groups, expressed in U/L, are presented in Table 3. All groups displayed a gaussian distribution except for women in the 30 –39 age group receiving hormone therapy. Statistical analysis was performed using the Student t-test or the Mann–Whitney rank-sum test. The overall population of 490 healthy volunteers had a mean proCPU concentration corresponding to a CPU activity of 964 U/L (SD, 155 U/L). No statistical difference was observed between men (961 U/L; SD, 153 U/L) and women (966 U/L; SD, 156 U/L), although there was a significant difference between women receiving hormonal therapy (1006 U/L; SD, 152 U/L) and women not receiving hormonal therapy (930 U/L; SD, 152 U/L; P ,0.0001). In men, there was a positive correlation between the proCPU concentration and age (r 5 0.2865; P ,0.001), with a significant increase in activity starting with the 40 – 49 age group (P 5 0.0024). Women not receiving hormonal therapy had a considerable increase in proCPU concentration in the 50 – 60 age group (P ,0.0001). Under the age of 50 years, similar values were observed in women not receiving hormone therapy as in men under 40 years of age. In women receiving hormonal therapy, no age-related differences were observed. Increased proCPU concentrations were obtained in women receiving hormonal therapy in comparison with women not receiving hormonal therapy [P 5 0.0002 (age group, 30 –39 years); P 5 0.0007 (age group, 40 – 49 years)]. In the 50 – 61 age groups, no difference could be observed between the groups.
Discussion In this study, we have described a method to determine proCPU concentrations in human plasma. The zymogen is converted quantitatively to its active form when thrombin and thrombomodulin are used as the activating complex. Subsequently, the enzymatic activity is determined by an HPLC-assisted assay, using hippuryl-l-arginine as the substrate (12, 31 ). Detection of the enzyme-produced product is simple, fast (separation of hippuric acid and the substrate within 3 min), and can be partly automated. Using a slightly modified procedure (longer incubation time, different volume ratio of substrate to sample), we are able to detect carboxypeptidase activity as low as 1 U/L. The problem of interference of CPN is circumvented by the measurement of the carboxypeptidase activity in each plasma sample without previous activation. The CPN activity, which is limited to 5–10% of the total activity, is subtracted from the carboxypeptidase activity obtained after activation, and this value is a measure for the proCPU concentration in plasma. The proposed method is very reproducible both for the activation step
and the enzymatic determination. Our preliminary data are indicative of a good correlation between proCPU measurement using either the activity assay or antigen determination by ELISA ( n 5 120, r 5 0.803; manuscript in preparation). Recently, much attention has been focused on the role of CPU in fibrinolysis (Fig. 1). proCPU exhibits a strong affinity for plasminogen and can be activated by plasmin and thrombin (7 ), but is activated most efficiently by the thrombin-thrombomodulin complex (18 ). The potential role of the proCPU/CPU system in controlling fibrinolysis is further demonstrated by a study of Mosnier et al. (26 ), which revealed a positive correlation between the proCPU antigen concentration and the time to clot lysis. This indicates that an increased proCPU concentration in plasma signifies a retardation of clot lysis and thus could parallel increased thrombo-embolic risk. In this respect, the proCPU concentrations we measured in the different age and sex groups and in relation to the use of hormonal therapy seem very interesting. Indeed, we noticed increases in proCPU concentrations in relation to age for men, increases for women in the 50 – 61 age groups and increases in women taking oral contraceptives. Further studies on patient groups with increased thrombo-embolic risk need to be undertaken to determine whether a high proCPU concentration is a risk factor for thrombo-embolic disease.
We gratefully acknowledge the expert technical assistance of Yani Sim. K. Schatteman is a research assistant and F. Goossens is a senior research assistant of the Fund for Scientific Research Flanders (FWO-Vlaanderen).
References 1. Hendriks D, Scharpe´ S, van Sande M. Partial purification and characterization of a new arginine carboxypeptidase from human serum [Abstract]. J Clin Chem Clin Biochem 1988;26:305. 2. Hendriks D, Scharpe´ S, van Sande M, Lommaert MP. A labile enzyme in fresh human serum interferes with the assay of carboxypeptidase N. Clin Chem 1989;35:177. 3. Hendriks D, Scharpe´ S, van Sande M, Lommaert MP. Characterization of a carboxypeptidase in human serum distinct from carboxypeptidase N. J Clin Chem Clin Biochem 1989;27:277– 85. 4. Hendriks D, Wang W, Scharpe´ S, Lommaert M, van Sande M. Purification and characterization of a new arginine carboxypeptidase in human serum. Biochim Biophys Acta 1990;1034:86 –92. 5. Campbell W, Okada H. An arginine specific carboxypeptidase generated in blood during coagulation or inflammation which is unrelated to carboxypeptidase N or its subunits. Biochem Biophys Res Commun 1989;162:933–9. 6. Campbell W, Yonezo K, Shinohara T, Okada H. An arginine carboxypeptidase generated during coagulation is diminished or absent in patients with rheumatoid arthritis. J Lab Clin Med 1990;115:610 –2. 7. Eaton D, Malloy B, Tsai S, Henzel W, Drayna D. Isolation, molecular cloning, and partial characterization of a novel carboxypeptidase B from human plasma. J Biol Chem 1991;266: 21833– 8. 8. Tsai S, Drayna D. The gene encoding human plasma carboxypep-
Clinical Chemistry 45, No. 6, 1999
tidase B (CPB2) resides on chromosome 13. Genomics 1992;14: 549 –50. Skidgel R. Basic carboxypeptidases: regulators of peptide hormone activity. Trends Pharm Sci 1988;9:299 –304. Geokas M, Wollesen F, Rinderknecht H. Radioimmunoassay for pancreatic carboxypeptidase B in human serum. J Lab Clin Chem 1974;84:574 – 83. Delk A, Durie P, Fletcher T, Largman C. Radioimmunoassay of active pancreatic enzymes in sera from patients with acute pancreatitis. I. Active carboxypeptidase B. Clin Chem 1985;31: 1294 –300. Wang W, Hendriks D, Scharpe´ S. Carboxypeptidase U: a plasma carboxypeptidase with high affinity for plasminogen. J Biol Chem 1994;269:15937– 44. Vanhoof G, Wauters J, Schatteman K, Hendriks D, Goossens F, Bossuyt P, et al. The gene for human carboxypeptidase U (CPU)—a proposed novel regulator of plasminogen activation— maps to 13q14.11. Genomics 1996;38:454 –5. Bajzar L, Manuel R, Nesheim M. Purification and characterization of TAFI, a thrombin-activatable fibrinolysis inhibitor. J Biol Chem 1995;270:14477– 84. Redlitz A, Tan A, Eaton D, Plow E. Plasma carboxypeptidases as regulators of the plasminogen system. J Clin Investig 1995;96: 2534 – 8. Bajzar L, Nesheim M, Tracy P. The profibrinolytic effect of protein C in clots formed from plasma is TAFI-dependent. Blood 1996; 88:2093–100. Redlitz A, Nicolini F, Malycky J, Topol E, Plow E. Inducible carboxypeptidase activity: a role in clot lysis in vivo. Circulation 1996;93:1328 –30. Bajzar L, Morser J, Nesheim M. TAFI, or plasma procarboxypeptidase B, couples the coagulation and fibrinolytic cascades through the thrombin-thrombomodulin complex. J Biol Chem 1996;271: 16603– 8. Suenson E, Lu¨tzen O, Thorsen S. Initial plasmin-degradation of fibrin as the basis of a positive feed-back mechanism in fibrinolysis. Eur J Biochem 1984;140:513–22. Fleury V, Angle´s-Cano E. Characterization of the binding of plasminogen to fibrin surfaces: the role of carboxy-terminal lysines. Biochemistry 1991;30:7630 – 8.
21. Pannell R, Black J, Gurewich V. Complementary modes of action of tissue-type plasminogen activator and pro-urokinase by which their synergistic effect on clot lysis may be explained. J Clin Investig 1988;81:853–9. 22. Sakharov D, Plow E, Rijken D. On the mechanism of the antifibrinolytic activity of plasma carboxypeptidase B. J Biol Chem 1997;272:14477– 82. 23. Broze G, Higuchi D. Coagulation-dependent inhibition of fibrinolysis: role of carboxypeptidase-U and the premature lysis of clots from hemophilic plasma. Blood 1996;88:3815–25. 24. Broze G. Thrombin-dependent inhibition of fibrinolysis. Curr Opin Hematol 1996;3:390 – 4. 25. Minnema M, Friederich P, Levi M, von dem Borne P, Mosnier L, Meijers J, et al. Enhancement of rabbit jugular vein thrombolysis by neutralization of factor XI: in vivo evidence for a role of factor XI as an anti-fibrinolytic factor. J Clin Investig 1998;101:10 – 4. 26. Mosnier L, von dem Borne P, Meijers J, Bouma B. Plasma TAFI levels explain the variable clot lysis time in healthy individuals [Abstract]. Blood 1997;90:10. 27. Vogel AI. Practical organic chemistry. London: Longman, 1970: 584. 28. Deutsch D, Mertz E. Plasminogen: purification from human plasma by affinity chromatography. Science 1970;170:1095– 6. 29. Bradford M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 1976;72:248 –54. 30. Laemmli U. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680 –5. 31. Hendriks D, Scharpe´ S, van Sande M. Assay of carboxypeptidase N activity in serum by liquid-chromatographic determination of hippuric acid. Clin Chem 1985;31:1936 –9. 32. Boffa M, Wang W, Bajzar L, Nesheim M. Plasma and recombinant thrombin-activatable fibrinolysis inhibitor (TAFI) and activated TAFI compared with respect to glycosylation, thrombin/thrombomodulin-dependent activation, thermal stability, and enzymatic properties. J Biol Chem 1988;273:2127–35. 33. Dang Q, Vindigni A, Di Cera E. An allosteric switch controls the procoagulant and anticoagulant activities of thrombin. Proc Natl Acad Sci U S A 1995;92:5977– 81.