In secondary systemtic amyloidosis, amyloid fibrils ... - Science Direct

4 downloads 0 Views 595KB Size Report
Jun 9, 1981 - The inhibitory effects of the plasma protease inhibitors antithrombin III, a2-macroglobulin and a l-antitrypsin on the activity of human factor Xa ...
Biochimica et Biophysica Acta, 701 (1982) 19-23

19

Elsevier Biomedical Press BBA31038

THE HETEROGENEITY OF PROTEIN AA IN SECONDARY (REACTIVE) SYSTEMIC AMYLOIDOSIS PER W E S T E R M A R K

Department of Pathology, University of Uppsala, P.O. Box 553, S-751 22 Uppsala (Sweden) (Received June 9th, 1981)

Key words: Protein AA; Amyloidosis; Amyloid fibril

In secondary systemtic amyloidosis, amyloid fibrils have protein AA as a main subunit protein. As judged from gel chromatography and electrophoresis, this protein is rather homogeneous. In the present paper it is shown, however, that protein AA is very heterogeneous and composed of many peptides with different isoelectric points. However, their antigenic properties and amino acid compositions vary only little. It is concluded that protein AA is as heterogeneous as its postulated precursor, the acute phase reactant serum AA and that a theory that only one or a few serum protein AA's can give rise to amyloid fibrils, might be wrong.

Introduction The fibrils in secondary (reactive) systemic amyloidosis are mainly composed of protein AA [1]. This protein has a relatively constant size and consists of 76 amino acid residues, although a few variations have been described [2-4]. The amino acid sequence of protein AA has been determined and some variations in the primary structure of the protein have been reported [5,6]. These variations occur in the C-terminal part of the molecule while the N-terminal part is constant besides a variation in the presence or absence of arginine as the aminoterminal residue [5,7,8]. Protein AA is believed to originate from a normal plasma acute phase reactant [9-11]. This protein has recently been shown to be an apolipoprotein of the high-density lipoprotein fraction [ 12,13]. The apolipoprotein, named serum AA, has a molecular weight of about 11000-14000 [ 14-17] and contains around 109 amino acid residues [14]. Serum protein AA has the same N-terminal amino acid sequence as protein AA [14,15,17] but contains about 33 amino acid residues more in the C-terminal part of the molecule [14]. Although not 0167-4838/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

proved, protein AA is probably formed by an enzymatic splitting of serum protein AA [18,19]. Serum protein AA is a heterogeneous protein and it has recently been shown that at least six different species of the protein exist [14]. In spite of similar size and similar amino acid composition, these species have quite different isoelectric points. The reason for this is not clear. Although serum protein AA is an acute phase reactant always occurring as an inflammatory response, e.g., in rheumatoid arthritis, only some persons develop amyloidosis. The reason for this is quite unclear, but it has been suggested that only some serum protein AA-forms might be amyloidogenous [14,20]. In that case protein AA should be a more homogeneous protein than serum protein AA. The present study shows, however, that protein AA is also a highly heterogeneous protein.

Material and Methods Amyloid fibrils were extracted as described [21] from the spleen of two patients (582 and Es59 *) * Thanks are due to Dr. Lars Vejlens for this material.

20 and from the kidney of one person (98) with rheumatoid arthritis and systemic amyloidosis. Protein AA was purified by gel filtration mainly as described previously [8,22]. Electrophoretic methods. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate (SDS) was performed according to Swank and Munkres [23]. For analytic isoelectric focusing, thin layer plates of 5% polyacrylamide gel containing 6 M urea and 2% Ampholine (LKBProdukter AB, Bromma, Sweden) pH range 3.5-9.5 were used [24]. Protein AA was dissolved (8 m g / m l ) in 6 M urea and applied to the cathodic side of the gel and the focusing was run at a constant power of 40 W. After fixation, the gels were stained with Coomassie blue and photographed. Immunological methods Antiserum to protein AA was induced in rabbits by immunization with alkali-degraded amyloid fibrils as described [25,26]. The antiserum was absorbed with normal human serum (1 : 5) and double immunodiffusion was performed in 1% agarose. For immunoelectric focusing, stripes from the isoelectric focusing plates were cut out without fixation and brough to a glass plate and an agarose gel was cast close to them. Anti-protein AA antiserum was applied in a furrow made in the agarose gel. Preparative isoelectric focusing. Preparative isoelectric focusing was performed in an LKB 8100-110 column, using a sucrose density gradient containing 6 M freshly prepared urea and 2% Ampholine, pH 3.5-9.5 [24]. About 25 mg protein AA were dissolved in 6 M urea and applied to the heavy sucrose solution before the formation of the gradient. Electrofocusing was performed at 5 W for 48-72h, with a final voltage of 1200V. The column was then eluted with a continuous registration of the absorbance at 280 nm. Fractions of 2ml were collected. After that the pH o f the fractions was measured then suitable fractions were pooled and dialyzed exhaustively against distilled water and lyophilized. Amino acid analysis. " Amino acid analyses were performed by the Central Amino Acid Analysis Laboratory, Department of Biochemistry, University of Uppsala, after hydrolysis in 6 M HC1 .at 110°C for 24 h.

Results Gel filtration of dissolved amyloid fibrils on a Sepharose 6B column revealed a typical pattern with a late major retarded peak comprising protein AA. SDS-polyacrylamide gel electrophoresis of this protein showed one rather broad band corresponding to a molecular weight of about 9000.

Analytic isoelectric focusing Analytic isoelectric focusing resulted in a rather complex pattern. Protein AA from all three patients turned out to be highly heterogeneous with several peptides within the anode half of the gel plate (Fig. 1). Four rather distinct groups of bands ( A - D , Fig. 1) could be discerned. Of these, group B and C dominated, preeminently in the case of 582 and Es59. Group D especially consisted of several closely located bands. The pattern of the

A II

g, I]

,Z.

Fig. 1. Thin layer polyacrylamidegel isoelectric focusing, pH 3.5-9.5, of protein APt of three different patients (582, Es 59, 98, from left to right). Protein AA of all three preparations is highly heterogeneousand consists of severalpeptideswithin the pH intervalof about 4.7-6.9. The patterns resembleeach other but distinct differencesare seen.

21

.6

-4== -2

'~'t

20



30 FRACTION NO.



40

it

2o

3'o

FRACTION NO.

,b

il

Fig. 2. Elution profiles of protein AA 582 (A) and 98 (B) after' column isoelectric focusing, pH 3.5-10. Six distinct peaks are seen, 1 + 2 corresponding to band group D in Fig 1, 3 to group C, 4 to group B and 5 + 6 to group A.

three different AA proteins resembled each other closely, but the proportions between the different band differed. The isoelectric focusing pattern was unaffected by whether a newly purified protein or a protein that had been stored for a long time was used. The pattern was also the same in proteins purified from different lots of fibrils which had been isolated from organs at different times.

Preparative isoelectricfocusing The elution patterns of protein AA from 582 and 98 are shown in Fig. 2. As is seen, four major and two minor peaks occurred. Peak 1 and 2 correspond to band group D in the analytical isoelectric focusing, peak 3 to band group C, peak4 to band group B and peak 5 and 6 band group A. As is seen, the p I of the peaks varies between about 4.7 and 6.9 but the major part of protein AA has an isoelectric point of 5.1-5.5. Peak 1 in 582 is rather high although the amount of protein was low. This discrepancy is due to a precipitation of protein in that region during the focusing. Protein AA 98 showed the same main peaks but also an additional one at pI 4.5. This corresponds to a heavily stained band in band group D, occurring in this protein. In protein AA 98 the proportions of the main peaks are also little different with peak 2 being comparably high and broad. The amino acid compositions of the different fractions are given in Table I, together with that of a previously reported protein AA [5]. Due to lack of material not all peak materials could be studied. As is seen, the amino acid compositions are typical of protein AA with high values of aspartic

acid/asparagine, glutamic acid/glutamine, alanine and arginine and low values of threonine, proline, valine and leucine. Peak 1 in 582 is slightly different since the values of threonine, valine and leucine are comparably high. The lack of methionine and proline is also peculiar. There are also a few other variations among the peptides, the most notable being the lack of proline in peak 2 of 582 and the low values of methionine in most peak materials of this case.

Immunological findings Protein AA from both patients reacted with anti-protein AA antiserum giving one line of identity. All different peak proteins also reacted with this antiserum, giving one line of identity (Fig. 3). Immunoisoeletric focusing revealed one line of precipitation along the whole area where protein bands occurred.

Fig. 3. Double immunodiffusion of anti protein AA antiserum (central well) tested against different peak proteins of protein AA 582, obtained by preparative isoelectric focusing. Material from peak 1 in well 1, peak 3 in well 2, peak 5 in well 3, peak 4 in well 4, peak 3 in well 5 and peak 6 in well 6. There are no certain antigenic differences.

22 TABLE I The amino acid composition of subspecies of protein AA0 purified by isoelectric focusing from two different individuals expressed as residues per 100 residues found. The peak numbers correspond to Fig. 2. n.d., not determined. Peak

582

98

Protein AA

(5) Peak

1

2

3

4

5

1

2

3

Asp Thr Ser Glu Pro Gly Ala ~Cys Val Met lle Leu Tyr Phe His Lys A~ T~

13.7 2.2 9.1 10.5 Trace 12.2 13.6 n.d. 3.0 Trace 4.7 5.0 6.3 7.4 1.6 3.2 7.6 n.d.

14.4 Trace 8.4 8.8 0.0 12.6 17.3 n.d. 1.4 0.8 4.4 2.5 6.2 9.1 2.2 2.8 9.1 n.d.

14.2 Trace 8.2 8.4 1.2 12.5 17.2 n.d. 1.3 0.6 4.3 2.0 6.3 9.2 2.3 2.9 9.5 n.d.

13.7 0.5 8.0 8.1 1.4 12.3 16.7 n.d. 1.5 1.4 4.2 2.2 6.1 8.9 2.2 3.1 9.9 n.d.

13.0 0.9 8.0 8,4 1.5 12.7 15.6 n.d. 1.7 1.6 4.3 2.9 6.0 8.6 2.3 3.1 9.4 n.d.

12.9 0.4 8.0 8.8 1.5 11.9 18.4 0.3 a 1.3 2.8 a 3.6 2.9 5.5 8.5 2.3 2.6 8.6 n.d.

12.2 0.4 8.8 9.5 1.5 12.1 17.3 0.2 a 1.4 2.6 a 3.6 2.7 5.6 8.6 2.4 2.7 8.9 n.d.

12.7 0.4 8.7 8.5 1.8 12.3 16.9 0.1 a 1.5 2.7 a 3.7 2.2 5.7 8.7 2.4 2.7 9.3 n.d.

13.2 0.0 7.9 7.9 1,3 11.8 15.8 0.0 1.3 2.6 3.9 1.3 5.3 9.2 2.6 2.6 10.5 2.6

* Determined in oxidized sample,

Discussion Protein AA usually consists of 76 amino acids [5,6] and only a few shorter and one larger AA protein have been sequenced [2-4]. By SDSpolyacrylamide gel electrophoresis and by gel filtration protein AA has been found to be homogeneous although variants with two bands in electrophoresis [2,27] or two peaks in gel filtration [8] have been described. On the other hand, there is evidence that protein AA in the mouse is rather heterogeneous [28,29]. The results in the present study show that human protein AA, although homogeneous in size, consists of several polypeptides with rather big differences in isoelectric points. The peptides with a p I of about 5.1-5.5 dominated which is in accordance with the findings of Linke et al. [30] who reported that a major part of protein AA had a p l of 5.1. Antiserum to protein AA reacted with peptides in the whole pH range where any peptides occurred. It cannot, of course, be ruled out that some of the minor bands seen in the analytic

electrofocusing were contaminants and not protein AA. The reason for the heterogeneity in charge is not clear. The protein did not contain any hexosamine. Furthermore, the amino acid composition varied only rather slightly. The findings are thus comparable to those of Bausserman et al. [14] concerning serum protein AA, which also is heterogeneous in spite of only small variations in amino acid composition. It has been proposed that one or a few polypeptides of all serum protein AA variants could be especially prone to form fibrils, i.e., be amyloidogeneous [14,20,29]. This idea has received some support from studies on mouse amyloid, where more heterogeneity exists in the N-terminus of serum protein AA than that of protein AA [29]. One would expect protein AA to be more or less homogeneous and not as heterogeneous as serum protein AA. However, a heterogeneous protein AA could also arise through cleavage at different positions in a homogeneous serum protein AA. In that case greater differences in amino acid composition

23

should be seen and variations in the size of the different polypeptides should exist. Therefore, other factors than the existence of one specific amyloidogeneous serum protein AA protein might be of importance in determining whether amyloidosis develops or not. An unusual finding was the lack of proline in peak 1 and 2 of 582. Proline has been found constantly in position 49 in all AA proteins sequenced until now [2,4-6] except for one very short protein, which contained only 45 residues [3]. Vahne occurs only beyond position 52 [2,4-6] and since this amino acid was found in peak 1 and 2 a lack of proline is not explained by a restricted size of this protein AA. Acknowledgements

Supported by the Swedish Medical Research Council (Project No. 5941) and the Research Fund of King GustafV. Thanks are due to Anni Bedy and Christer Tengvar for skilled technical help. References 1 Benditt, E.P., Eriksen, N., Hermodson, M.A. and Ericsson, L.H. (1971) FEBS Lett. 19, 169-173 2 Sletten, K., Husby, G. and Natvig, J.B. (1976) Biochem. Biophys. Res. Commun. 69, 19-25 3 Ein, D., Kimura, S., Terry, W.D., Magnotta, J. and Glenner, G.G. (1972) J. Biol. Chem. 247, 5653-5655 4 Mo'yner, K., Sletten, K., Husby, G. and Natvig, J.B. (1980) Scand. J. Immunol. I I, 549-554 5 Levin, M., Franklin, E.C., Frangione, B. and Pras, M. (1972) J. Clin. Invest. 51, 2773-2776 6 Sletten, K. and Husby, G. (1974) Eur. J. Biochem. 41, 117-125 7 Pras, M., Zaretzky, J., Frangione, B. and Franklin, E.C. (1980) Am. J. Med. 68, 291-294 8 Westermark, P., Sletten, K. and Eriksson, M. (1979) Lab. Invest. 41,427-431 9 Husby, G. and Natvig, J.B. (1974) J. Clin. Invest. 53, 1054-1061

10 Levin, M., Pras, M. and Franklin, E.C. (1973) J. Exp. Med. 138, 373-380 I1 Gorevic, P.D., Rosenthal, C.J. and Franklin, E.C. (1976) Clin. Immunol. Immunopathol. 6, 83-93 12 Benditt, E.P. and Eriksen, N. (1977) Proc. Natl. Acad. Sci. USA 74, 4025-40128 13 Skogen, B., Barresen, A.L., Natvig, J.B., Berg, K. and Michaelsen, T. (1979) Scand. J. Immunol. 10, 39-45 14 Bausserman, L.L., Herbert, P.N. and McAdam, K.P.W.J. (1980) J. Exp. Med. 152, 164-656 15 Anders, R.F., Natvig, J.B., Michaelsen, T.E. and Husby, G. (1975) Scand. J. Immunol. 4, 397-401 16 Linke R.P., Sipe, J.D., Pollock, P.S., Ignaczak, T.F. and Glenner, G.G. (1975) Proc. Natl. Acad. Sci. USA 72, 14731476 17 Rosenthal, C.J., Franklin, E.C., Frangione, B. and Greenspan, J. (1976) J. Immunol. 116, 1415-1418 18 Lavie, G., Zucker-Franklin, D. and Franklin, E.C. (1978) J. Exp. Med. 148, 1020-1031 19 Skogen, B., Thorsteinsson, L. and Natvig, J.B. (1980) Scand. J. Immunol. 1I, 533-540 20 Glenner, G.G. (1980) New Engl. J. Med. 302, 1283-!292, 1333-1343 21 Pras, M., Schubert, M., Zucker-Franklin, D., Rimon, A. and Franklin, E.C. (1968) J. Clin, Invest. 47, 924-933 22 Glenner, G.G., Harada, M. and Isersky, C. (1972) Prep. Biochem. 2, 39-51 23 Swank, R.T. and Munkres, K.D. (1971) Anal. Biochem. 39, 462-477 24 Haglund, H. (1970) Methods of Biochemical Analysis, Vol. 19, LKB-Produkter AB, Stockholm 25 Pras, M., Zucker-Franklin, D., Rimon, A. and Franklin, E.C. (1969) J. Exp. med. 130, 777-795 26 Husby, G. and Natvig, J.B. (1972) Clin. EXp. Immunol. 10, 635-647 27 Benditt, E.P. and Eriksen, N. (1972) Lab. Invest. 26, 615625 28 Loewenstein, J., Rimon, A. and Frensdorff, A. (1980) in Amyloid and Amyloidosis (Glenner, G.G., Costa, P.P. and Freitas, A.F., eds.), pp. 471-474, Excerpta Medica, Amsterdam 29 Gorevic, P.D., Levo, Y., Frangione, B. and Franklin, E.C. (1978) J. Immunol. 121, 138-140 30 Linke, R.P., Sipe, J.D., Pollock, P.S., Ignaczak, T.F. and Glenner, G.G. (1975) Proc. Natl. Acad. Sci. USA 72, 14731476

24

Biochimica et Biophvsica A eta, 701 (1982) 24- 31 Elsevier Biomedical Press

BBA 31036

INHIBITION OF HUMAN FACTOR Xa BY VARIOUS PLASMA PROTEASE INHIBITORS V I N C E N T ELLIS, M I C H A E L SCULLY *, IAN M A C G R E G O R ** and VIJAY K A K K A R

Thrombosis Research Unit, King's College Hospital Medical School, Denmark Hill, 'London SE5 8 R X (U.K.) (Received October 13th, 1981)

Kev words: Factor Xa inhibition; Protease inhibitor," (Human plasma)

The inhibitory effects of the plasma protease inhibitors antithrombin III, a2-macroglobulin and a l-antitrypsin on the activity of human factor Xa have been studied using purified proteins. The rate of inhibition was determined by measuring the residual factor Xa activity at timed intervals utilizing the synthetic peptide susbtrate Bz-lle-Glu(piperidyl)-Giy.Arg-pNA. Kinetic analysis with varying molar concentrations of inhibitors demonstrated that the inhibition of factor Xa by antithrombin III, az-macroglobulin and a l-antitrypsin followed second-order kinetics. Calculated values of the rate constants for the inhibition of factor Xa by antithrombin III, az-macroglobulin and al-antitrypsin were 5.8-104, 4.00.104 and 1.36.104 M -Lmin -1, respectively. The plasma concentrations of the inhibitors can he used to assess their potential relative effectiveness against factor Xa. In plasma this was found as al-antitrypsin>antithrombin I I I > a 2macroglobulin in the ratio 4.64: 2.08:1.0. Cephalin was shown to inhibit the rate of reaction between factor Xa and antithrombin III.

Introduction Factor X is a plasma glycoprotein involved in the blood coagulation cascade and can be converted to its active form, factor Xa a serine protease, by both the intrinsic pathway (Factor IXa) and the extrinsic pathway (factor VII and tissue factor), (for recent review see Ref. 1). Factor Xa is the activator of factor II (prothrombin) and occupies a central position linking the two blood coagulation pathways. Control of factor Xa levels by plasma protease inhibitors may therefore be pivotal in the regulation of the coagulation process. A number of plasma serine protease inhibitors * To whom correspondence should be addressed. ** Present address: Scottish National Blood-Transfusion Service, Headquarters Unit Laboratories, 2 Forest Road, Edinburgh EHI 2QN, U.K. 016%4838/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

are thought to be involved in the regulation of activated coagulation factors, of which the most important are found in high molar concentrations in plasma ( > 3 /~M). These are a~-antitrypsin, a2-macroglobulin and antithrombin III. a 1 Antitrypsin is the most abundant of the plasma protease inhibitors and is responsible for over 90% of the trypsin-inhibitory activity of human plasma [2]. It is an effective inhibitor of factor XIa [3] and thrombin [4] as well as many other proteasts yet despite its trypsin-like nature there have been very few reports of the inhibition of factor Xa by antitrypsin [5]. The inhibitory range of a 2 macroglobulin is less broad than that of a 2antitrypsin. The mechanism of its action is unique in that the inhibitor-enzyme reaction involves a proteolytic cleavage of the inhibitor molecule probably followed by a physical entrapment of the enzyme [6]. a2-Macroglobulin has been shown to inhibit plasma kallikrein [7] plasmin [8] and

25 thrombin [9]. Antithrombin III is found at similar molar concentrations to a 2-macroglobulin and has been shown to inhibit the action of factors IXa [10], XIIa [11], Xa [12] and thrombin [13] by the formation of a complex involving a 1 : 1 stoichiometry. In the case of factor XIa a stoichiometry of 2 mol inhibitor to 1 mol enzyme was found [14]. The reaction between all these enzymes and antithrombin III is greatly accelerated by the presence of catalytic amounts of heparin. Downing et al., [15] have shown antithrombin III to be the most important thrombin inhibitor in plasma. We have studied the kinetics of the reactions between factor Xa and eq-antitrypsin, ot2 macroglobulin and antithrombin III in order to be able to directly compare the activity of the three inhibitors. From the data it has been possible to make an assessment of their relative potential in plasma in the inhibition of factor Xa. The effects of heparin, a widely used anticoagulant and phospholipid (cephalin), a cofactor in the activation of prothrombin by factor Xa, on the reactions between the enzyme and their inhibitors have also been studied. -

Materials and Methods

Human factor X was prepared from 1400 ml of outdated citrated plasma by a modification of the method of Mertens and Bertina [16]. The factor X containing fractions from the DEAE-Sephadex A50 column step were dialysed against 5 mM KC1/10 mM-triethanolamine-HC1, pH 6.35/3 mM CaCI 2 and applied to a column of heparinSepharose equilibrated in the same buffer. Heparin-Sepharose was prepared by coupling 1 mg crude porcine mucosal heparin (Riker Laboratories) to 1 ml Sepharose 4B (Pharmacia) using the method of March et al. [17]. Factor X was eluted from the heparin-Sepharose column using a linear gradient of KC1 upto 500 mM. Factor X-containing fractions were pooled and dialysed against 0.2 M potassium phosphate, pH 6.8, and applied to a column of hydroxyapatite (Bio-Gel-HT) equilibrated in the same buffer. After washing the column the factor X was eluted with a linear gradient of 0.2-0.4 M KH2PO 4, pH 6.8. The fractions were dialysed against 0.05 M Tris-HCl/ 0.1 M NaCI, pH 7.5. Factor X was activated by

incubation with Russell's Viper Venom (Sigma Chem. Co. Poole, U.K.) covalently bound to Sepharose 4B. The activation peptide was removed from the factor Xa by ion exchange chromatography on DEAE-Sephadex A-50 equilibrated with 0.05 M Tris-HCl/0.1 M NaC1, pH 7.5, the column being eluted with a linear gradient from 0.1-1.0 M NaC1 in the same buffer. Factor Xa was dialysed against 0.05 M Tris-HCl/0.1 M NaC1, pH 7.4 and stored in aliquots at -20°C. at-Antitrypsin was prepared from out-dated human plasma by the method of Kurecki et al. [18]. Human antithrombin III purified by the method of Miller-Anderson et al. [19] was obtained from AB Kabi Stockholm and human a 2macroglobulin was a gift from Dr. A.J. Barrett, Strangeways Laboratories, Cambridge. Russell's Viper Venom (Sigma, Poole, U.K.) was purified by the method of Jesty et al. [20] and linked to Sepharose 4B by the method of March et al. [17]. Protein concentrations were measured by absorption at 280 nm assuming an E21s~0of 11.6 for factor X [21], 4.84 for al-antitrypsin [22], 8.7 for a 2-macroglobulin [23] and 6.1 for antithrombin III [201. Factor X was assayed by the clotting method of Fujikawa et al. [24] using factors VII and X deficient plasma, (Sigma, Poole, U.K.). Factor X was also assayed by hydrolysis of Bz-Ile-Glu(-OR)Gly-Arg-pNA, (S-2222 Kabi, Stockholm, Sweden), after activation by Russell's Viper Venom [25]. A unit of factor X in both assays is defined as equivalent to that present in 1 ml normal pooled plasma after complete activation. SDS-polyacrylamide gel electrophoresis [26] was used to determine the homogeneity of all enzymes used. Crossed immunoelectrophoresis was used to determine the purity of preparations using antisera to human al-antitrypsin and az-macroglobulin, (Dako-Paks, Copenhagen, Denmark) and to antithrombin III, (Behringwerke AG, Marburg, F.R.G.). The inhibition of factor Xa by al-antitrypsin and antithrombin III was determined by measurement of the residual factor Xa activity against tile chromogenic peptide substrate, Bz- Ile- Glu(piperidyl)-Gly-Arg-pNA, (S-2337, Kabi, Stockholm, Sweden) [27]. Inhibition reactions were car-