Determination of human plasma kininogen by a single radial immunodiffusion method and the bradykinin equivalent

Clinica Chimica Acta, 142 (1984) 211-220

211

Elsevier CCA 02956

Determination of human plasma kininogen by a single radial immunodiffusion method and the bradykinin equivalent Ulla Hamberg * and Tytti K~rkk~inen Department of Biochemistry, University of Helsinki, Unioninkatu 35, SF-O0170 Helsinki 17 (Finland)

(ReceivedJanuary 25th 1984; revisionJune 1st, 1984) Key words: Single radial immunodiffusion of kininogen," Bradykinin equivalent," Kininogen RIA," Monospecific high avidity anti-kininogen (heavy chain) serum," Conformationalantigenic determinants

Summary An SRI method is presented for the quantitation of native human plasma kininogen using a monospecific high avidity antiserum prepared against the conformational determinants of kininogen heavy chain (H c antigen). The SRI value of total human plasma kininogen averaged 0.260 +__0.052 g/1 ( +_2 SD) using a standard curve of normal human plasma in each assay (measuring range 0.016-0.260 g/l). The SRI data, confirmed by RIA, were compared with determination of the bradykinin equivalent by bioassay 4.30 _+ 0.052 mg/1 ( _ 2 SD), generally accepted as a measure of total plasma kininogen. The SRI method does not discriminate between LMr and H M r kininogen but detects an increase of total kininogen independently of bradykinin release as a consequence of activated proteolysis. It is suggested that the SRI method can be applied to study levels of kininogen in clinical research.

Introduction Kininogens are single-chain heterogeneous glycoproteins in mammalian blood plasma existing in two principal forms, a high ( H M r 110000) and a low ( L M r about 60000) molecular mass kininogen [1]. As precursor of the covalently bound vasoactive kinin (bradykinin) peptide released by limited proteolysis the kininogen-kinin

* To whom correspondenceshould be addressed. 0009-8981/84/$03.00 © 1984 ElsevierSciencePublishers B.V.

212 system has been extensively studied with regard to bradykinin as a local nonapeptide hormone shown to participate as a mediator in inflammatory reactions [2-4]. In human plasma kallikrein (EC 3.4.21.8) releases bradykinin more readily from H M r, while L M r kininogen is the largest reservoir of the vasoactive peptide. H M r kininogen has more recently attracted attention as a cofactor in the contact activation of blood coagulation and fibrinolysis [4]. This diversity of functions and also the interrelations with some prostaglandins [3,5] suggests that the kininogen-kinin system plays a role in many pathophysiological conditions emphasizing the need of convenient assay methods. The present report deals with the presentation of an immunoassay for the determination of human plasma kininogens utilizing a monospecific anti-kininogen (heavy chain) serum developed in this laboratory [6-9]. This antiserum recognizes all heterogeneous forms of human kininogen by their common heavy chain provided that it contains the conformation dependent antigenic determinants of native plasma kininogen. The single radial immunodiffusion method (SRI) is compared with the determination of the bradykinin equivalent by bioassay [1], the method most frequently applied for estimation of plasma kininogens. Materials and methods Plasma samples

Normal blood bank plasma (anticoagulants CPD or ACD) from several healthy donors was obtained from the Finnish Red Cross Blood Transfusion Center (FRC, JE-1 Blood-Pack, Fenwal, Finland) and processed by standard FRC procedures. Fractions of 0.2 ml were stored at - 20 o C and applied as SRI standards throughout this investigation. Single donor blood was taken from healthy volunteers using plastic tubes containing ACD anticoagulant solution (citric acid, trisodium citrate, dextrose) 1 m l / 9 ml blood or heparin 30 NIH U / m l . Single donor plasma samples containing CPD anticoagulant (citrate, phosphate, dextrose) were obtained from FRC. T h e bradykinin equioalent

The content of biologically active kininogen (containing the bradykinin peptide) in plasma was determined by the standardized method used in this laboratory [10]. Plasma samples (1-2 ml) diluted at pH 2 with the same volume of ice cold acid saline were heated to about 95 o C, pH brought to 7.4-7.8 and incubated for 20 rain at 37°C with 200/~g trypsin/ml plasma (EC 3.4.21.4, Worthington T R L / T P C K ) . Released bradykinin was extracted with two volumes of boiling ethanol and the supernatant collected by centrifugation was dried under reduced pressure (Rotavapor). The sample was redissolved in ice cold saline to the original plasma volume. Bradykinin was determined by bioassay on the isolated guinea pig ileum suspended in aerated Tyrode solution in a thermostatted (7-10 ml) bath at 37 o C. Contractions were registered with a Harward 386 smooth muscle transducer using the Grass Model 79 C Polygraph. A 2-3 point bioassay was performed with 3-min intervals between doses in the linear dose response range using synthetic bradykinin

213 ( M r 1060) as standard (Sigma, St Louis, MO, USA) or synthesized as described in [11]. The sensitivity of the ileum (threshold dose) varied between 6.25-15.6 n g / m l bath volume. The amount of bradykinin/ml plasma, designated the bradykinin equivalent, is generally accepted as a direct measure of the kininogen content [1]. The specific activity (#g bradykinin/mg kininogen protein) gives the amount of active peptide bound to the protein. Preparation of antigens and antisera Antigenically pure kininogens were prepared from Cohn's fraction IV (Hc-antigen) [7], L M r and H M r kininogens were isolated from fresh blood bank plasma [12-14]. Final purification was performed with antibody specific immunoadsorbents as described earlier [7,9]. The purity was ascertained by SDS-gel electrophoresis and Ouchterlony double diffusion analysis. The respective antisera were raised in rabbits (2.4-3.2 kg) by injecting (i.d./s.c.) about 200 #g antigen in Freunds complete adjuvant (Difco Laboratories, Detroit, MI, USA) every fortnight. Commercial antisera and standard human serum (NHS) were from Behringwerke AG, Marburg/Lahn, FRG, and anti-normal human serum (a-NHS) was from Organon Teknika, Oss, The Netherlands. Single radial immunodiffusion (SR1) The quantitative determination of kininogen was performed [15] using glass plates (5 x 7.5 cm), 1.5% agar gel (Special Agar Noble, Difco), 0.04 mol/1 diaethylbarbiturate buffer pI-I, 8.6 and 2.5% antiserum. A standard of human pooled blood bank plasma (FRC, 63 ml anticoagulant solution/500 ml blood) was applied on each plate in dilutions 1 : 1, 1 : 2, 1 : 4, 1 : 8 and 1 : 16. The volume of all samples was 2 ~1 in 2-mm diameter wells in the agar plate. Plasma and protein samples were diluted/dissolved in saline. The diffusion was complete after 18-24 h at room temperature (the change of d 2 was negligible when measured after 48 h). The squared diameters (d 2) of the precipitation areas were plotted against the standard plasma concentration. The slope and intercept of the d2-axis was calculated using the least square root method. Radioimmunoassay (RIA ) H c kininogen (5 /~g) was radioiodinated to a specific activity of 69.2 # C i / # g . Details of the RIA procedure have been given before [6,9]. In the RIA assays H c and H M r kininogen standards were immunologically pure proteins applied in a concentration of 1-100 ng/200 #1. Determination of kallikrein actioity Kallikrein activity was determined as described in detail before [6] using 0.09 m m o l / l synthetic substrate H-D-Pro-Phe-Arg-p-nitroanilide (S-2302, Kabi, Stockholm, Sweden). Glass contact activation of plasma was performed by thawing a normal plasma sample kept frozen in a glass tube. Kallikrein activity was determined immediately and in a diluted plasma sample (500 #1 of plasma added to 250 #1 of buffer) after 30 re_in standing in an ice bath.

214

L M r kininogen was incubated with pig pancreas kallikrein (EC 3.4.21.8) (KZC 1/75, 1180 K U / m g , Bayer AG, Wuppertal, FRG) at pH 8.0 with 0.2 mol/1 Tris-HCl or 0.05 mol/1 NH4HCO 3 buffer. Incubation was performed at 37°C for various times. The reaction was stopped by addition of 10000 KIU Trasylol @ (Bayer AG, Wuppertal, FRG) in 0.1 ml of buffer. The reaction mixture was applied to a Sephadex G-25 column (0.8 × 35 cm) and immunoreactive kininogen was measured by SRI against the anti-H c serum in the protein pool (V0).

Results The bradykinin equivalent

The result of 35 determinations of the bradykinin equivalent in normal single donor FRC plasma is shown in Table I. The range of these determinations was 3.78 to 4.82 mg/1 with no significant difference ( p < 0.4) between the ACD and CPD

TABLE I The bradykinin equivalent (mg/I) in normal human plasma determined by bioassay on the isolated guinea pig ileum Plasma sample

No. of determined (mg/l)

Bradykinin (mean + 2 sv) =

Normal blood bank A C D / C P D Single donor A C D / C P D Single donor heparin

24 11 10

4.32 + 0.44 b 4.24 -I-0.68 b 3.19 + 0.54

a Uncorrected for the anticoagulant dilution. b Difference statistically not significant ( p < 0.4) mean 4.30+0.52 (rag/I).

o'(mm=) 40

d=(mm~') a

4O

30

30

2o

20

10 1:16 1:4 1:2 ! 0.0'16 0.0'65 0.130 0.2~00 BKG protein mg/ml

n 1:1 ! 0.260

b

10

n

1!16 li4 li2 , 0.016 0.065 0.130 0.200 BKG protein mg/ml

lil 0.260

Fig. 1. Immunoreactive kininogen (BKG) in serially diluted normal blood bank plasma ( O ) compared with serial dilutions of immunologically pure H c kininogen protein (Al~0 .m = 7.0) (e) determined by single radial immunodiffusion against (a) anti-H c kininogen serum (R 16 BI IV, titre 0.83 g/l), (b) anti-LM r kininogen serum (R3 B1 V, titre 0.19 g/l) 2.5 and 10% in the respective agar plates. Titres were determined by Becker's method [19].

215

plasma. This yields the spec act 16.5 # g / m g (see Fig. 1, 0.260 g/i) in agreement with the earlier spec act obtained with purified LMf kininogen [12]. The first value of 20 # g / m g was reported by Habermann [16] for bovine kininogen. With heparinized plasma the significantly lower (p < 0.001) 3.19 m g / l was obtained within the range 2.65 to 3.7.3 m'g/l. The difference (1.11 m g / l accounts for bradykinin release with heparin activated kallikrein [17] from HMf kininogen. This is in agreement with results by Uchida and Katori [18] obtained with a different technique estimating separately the release of 1 /~g/ml from HM~ kininogen, 3.08 /~g/ml from L M r kininogen and 4.11 # g / m l for the total bradykinin equivalent with trypsin. The SRI method does not discriminate between H M r and L M r, but measures the total kininogen by the heavy chain regardless of the content of kinin segment [6-9]. Kininogen in normal plasma by" S R I

Serial dilutions of immunologically pure H c kininogen protein (0.016-0.260 g/l) coincided with the serially diluted pooled blood bank plasma (Fig. la, b). Negligible variation occurred between single donor and blood bank plasma containing an unknown ratio of male and female plasma (Table II). The difference between sexes, however, as investigated so far is statistically significant (p < 0.01). Pooled blood bank and single donor sera also were assayed (Table II). The normal serum kininogen was on average 13% higher (0.313 g/l) compared with ACD plasma values which can be accounted for by the anticoagulant dilution. The higher value (0.349 g/l) with the male serum seems significant but requires confirmatory studies. As shown in Fig. 1, there was no difference in antigenic response with anti-H c and anti-LMr kininogen. Fig. lb shows the corresponding SRI analysis employing

TABLE II Kininogen (g/l) in normal human plasma and serum determined by single radial immunodiffusion Sample

Sex

No. of

Mean

so

( X 4- 2 so)

Range

Samples Pooled blood bank plasma Single donor

women a

12 56

0.265 0.260

0.021 0.026

0.265 4- 0.042 0.260 4- 0.052

0.223-0.307 0.208-0.312

men

85

0.270

0.022

0.2704-0.044

0.226-0.314

women

7 4

0.313 0.311

0.022 0.042

0.313 4- 0.044 0.311 4- 0.084

0.269-0.357 0.227-0.395

men

5

0.349

0.006

0.349 _ 0.012

0.337-0.361

plasma

Pooled blood bank serum b Single donor serum

a Submitted to routine questions only; women donors taking contraceptives excluded from normal group. b Pooled normal human serum was from the FRC collected by standard procedures. The plasma sample values were uncorrected for the anticoagulant dilution. The titre of the antiserum used was 1.1 g/i.

216

antiserum against L M r kininogen using 10% in the agar plate due to the lower titre [19]. Radioimmunoassay

The total kininogen content in normal blood bank and one single do.nor plasma estimated by RIA is given in Table III. The results (range 0.250-0.272 g/l) tested with the same antiserum are in good agreement with data obtained by SRI (Table II). The precision was determined using normal plasma. Ten dilutions of the same sample in duplicate were assayed in two different experiments (Nos. 1-2) and five dilutions of a single donor plasma (No. 3). The difference between the first two assays is statistically not significant (p > 0.5). Analyses of antigenic determinants

The recognition of the conformational determinants in HMr, L M r and H c kininogens by the antiserum directed again these determinants is demonstrated by RIA in Fig. 2. After reduction, alkylation and separation of the respective RCMheavy chains the antiserum prepared against the conformational determinants in the H c antigen no longer responds. As shown before [9] by the higher association

TABLE III Kininogen in normal human plasma (g/l) by RIA a Plasma

Assay

No.

Mean

SD

CV

Pooled blood bank

1 2 3

10 10 5

0.256 0.250 0.272

0.021 0.021 0.022

8.2 8.4 8.1

Single donor

(~)

a Anti.H c kininogen serum (R 16 BI XII diluted 1:20000).

B/B, 0.51"0 ~-'"

I

: a : l e dilution

2

:3

1:16

1:8 1:4 1'0 ng BKG

1:2

1:1

16o

Fig. 2. RIA inhibition curves of different kinlnogen (BKG) antigens ( O O) L M r, (@ @) H c and (11 II) H~4"r kininogen; heavy chairl containing only sequential determinants (- - -) from R C M - H M r (11) and R C M - H C (@) antigens were prepared by reduction and carboxymethylation [13] of 1.4 mg ldninogen. The assay was performed with the antiserum containing antibodies against the conformational determinants of the heavy chain (R 16 BI XII).

217 constant of the H c kininogen antiserum the unreduced antigen had the highest avidity for the conformation-dependent combining sites. Kallikrein and the immunoreactive kininogen

The activatidn of plasma kallikrein by glass contact and dilution may occur when handling plasma samples. Glass contact together with the dilution effect yielded an activation of about 35% of the totally activated plasma kallikrein with Cephotest comparing favourably with our earlier estimates [6,7]. With glass contact alone a comparatively small kallikrein effect was obtained (5%) in plasma. Table IV shows the result of the incubation of immunologically pure L M r kininogen with pancreas kallikrein. Regardless of the length of the incubation time the recovery of the SRI determined antigen was on average 95%. Discussion

There are several advantages in applying antiserum against the conformational determinants of the kininogen heavy chain [9] for the immunochemical determination of plasma kininogen. The first is obviously the resistance of the H c antigen to limited proteolysis by kallikrein. Secondly, all native kininogen forms containing the intact heavy chain can be measured [6-9]. The most obvious advantage is avoiding the error produced by relying only on determination of the bradykinin equivalent, since, particularly in pathological material due to activation of kallikrein and other plasma proteolytic enzymes, bradykinin can be lacking in the protein to an unknown extent. The unusual kinin-free H c kininogen from plasma fraction IV is a derivative of L M r kininogen with a heterogeneous light chain and antigenically intact heavy chain still connected by the disulfide bond [9]. The comparatively high plasma kallikrein activity in the starting material [7] explains the heterogeneity, but does not affect the antigenicity of the isolated antigen, nor does pancreatic kallikrein affect the SRI determination of L M r kininogen (Table IV). The higher avidity of the antibodies against this antigen show that the conformation-dependent determinants dominate the antigenic structure. TABLE IV Resistance of antigenic determinants to degradation by kallikrein assayed by single radial immunodiffusion with anti-He kininogen serum L M r kininogen

Kallikrein

Incubation

KU a

time

(h) 1 2 3

91 182 91

1 2 16

Kininogen (mg) After incubation b

Before

3.59 3.69 1.89

a The amount of enzymehydrolyzing1/~mol of Bz-Arg-OEt/min [26]. b Assayed after Sephadex(3-25 gel filtration of incubation mixture.

3.55 3.21 1.87

218 As shown by Uchida and Katori [18] incubation of plasma with glass powder causes the release of bradykinin from H M r kininogen. Since plasma kallikrein can be activated already by dilution and contact with a glass surface as shown by the present findings, special care must be taken when handling plasma samples for the determination of the bradykinin equivalent. This is particularly relevant when the bradykinin equivalent is used for estimation of plasma kininogen. Furthermore, two forms of L M r kininogen differing by the location of the bradykinin sequence are known. In kininogen I, bradykinin is situated at the C-terminal end of the protein and more easily inactivated.by plasma carboxypeptidase removing the C-terminal arginine [20]. Thus the peptide released with kallikrein or trypsin may be inactivated. The error, possibly deriving from this inactivation of bradykinin, is avoided with the present SRI method. As shown earlier [6] degraded forms of H M r kininogen (M r 105000, 90000, 86 000) were found after isolating the protein from normal plasma, which respond to the anti-H c serum and can be detected regardless of the degree of kallikrein cleavage of the light chain [9] and the part of the H M , . molecule responsive to the procoagulant activity. These cleavage mechanisms have not yet been conclusively shown in detail for the human kininogens [21]. The 27% content of H A I r kininogen, derived at by a preliminary estimation of the bradykinin equivalent in heparinized plasma (Table I), is within the range (16-30%) obtained by gel filtration in our earlier experiments. Quantiative immunochemical methods for the determination of L M r kininogen are rare. H M r kininogen is mostly determined by its coagulation activity [1,4]. Bouma et al [22] found 70 /lg/ml in plasma by Laurell's rocket immunoelectrophoresis method, and Proud et al [23] reported 90.8 _+ 2.5/Lg/ml by RIA. The difficulty to avoiding molecular changes during preparative procedures or with plasma unprotected against the activation of F XII clearly shows the advantage of an immunoassay utilizing the resistance of the determinants of the heavy chain to limited proteolysis. Sakamoto and Nishikaze [24] determined L M r kininogen in human serum using a calibration curve with the purified protein and obtained 111.3 /~g/ml. The reason for this low estimate compared with our results may depend on the use of a calibration curve with L M r kininogen of too low a specific activity [12]. In clinical material, when activation of plasma kallikrein is indicated, the SRI method apparently seems well-suited to give information about the concentration of plasma kininogen regardless of an eventual escape of bradykinin. Our studies with acute phase patient sera in inflammation showed increases of the SRI kininogen between 0.410-0.640 g/l, whereas the bradykinin equivalent varied between 1.9-4.0 m g / l (unpubl. results). As an acute phase reactant, kininogen, follows the increases of haptoglobin, al-antitrypsin and al-acid glycoprotein [25]. TheSe results demonstrate that an activation of proteolytic enzymes occurred during inflammation followed by the release of bradykinin. More recently we have found that the increase of kininogen as a pregnancy-associated protein followed the increase of the pregnancy-specific beta-l-glycoprotein and normalized after delivery (to be published). It is suggested that the SRI method presented can be applied in clinical research to study levels of plasma kininogen.

219

Acknowledgements This investigation was supported by grants from the Signe and Ane Gyllenberg Foundation and the Societas Scientiarum Fennica, Helsinki. Valuable assistance was given by Mrs. Mirja Kurki. For research materials we wish to thank Prof. G.L. Haberland, Bayer AG. Wuppertal, FRG, Drs. E. Berggren and R. Lund6n, AB Kabi, Stockholm, Sweden and Dr. G. Myllyl~, the Finnish Red Cross Blood Transfusion Centre, Helsinki, Finland.

References, 1 Kato H, Nagasawa S, Ivanaga S. HMW and LMW kininogens. Meth Enzymol 1981; 80: 172-198. 2 Garcia Leme J. Bradykinin-System. In: Vane JR, Ferreira SH, eds. Inflammation. Handbook of experimental pharmacology Vol 50/I, New York: Springer-Verlag, 1978: 464-522. 3 Ryan GB, Majno G. Acute inflammation. Am J Pathol 1977; 86: 184-276. 4 Cochrane CG, Griffin JH. The biochemistry and pathophysiology of the contact system of plasma. Adv Immunol 1981; 33: 221-306. 5 Goldstein RH, Polgar P. The effect and interaction of bradykinin and prostaglandins on protein and collagen production by lung fibroblasts. J Biol Chem 1982; 257: 8630-8633. 6 Syv~nen A-C, Turpeinen U, Siimesmaa S, Hamberg U. A radioimmunoassay for the detection of molecular forms of human plasma kininogen. FEBS Lett 1981; 129: 241-341. 7 K~kk~inen T, Syv~men A-C, Turpeinen U, Hamberg U. Isolation and immunologic properties of a heterogeneous antigen with the characteristics of the heavy chain of human plasma kininogen. Mol Immunol 1982; 19: 179-189. 8 Hamberg U, Syv~'~en A-C, SHmesmaa S. Identification in human plasma of low M r protein fragments with antigenic determinants of kininogen. FEBS Lett 1982; 138: 128-132. 9 Syv~LnenA-C, K~rkk~dnen T, Hamberg U. Conformation and sequence dependent antigenic determinants in human low molecular weight kJninogen. Mol Immunol 1983; 20: 669-678. 10 Hamberg U. Kininogen in human plasma after fibrinolytic activation. Scand J Clin Lab Invest 1969; 24 Suppl 107: 37-47. 11 Ragnarsson U, Karlsson SM, Hamberg U. Synthesis of peptides by fragment condensation on a solid support. 1. Application in preparation of bradykinin. Int J Peptide Prot Res 1975; 7: 307-312. 12 Hamberg U, Elg P, Nissinen E, Stelwagen P. Purification and heterogeneity of human kirunogen. Int J Peptide Protein Res 1975; 7: 261-280. 13 Kerbiriou DM, Griffin JH. Human high molecular weight kJninogen. Studies of structure-function relationship and of proteolysis of the molecule occurring during contact activation of plasma. J Biol Chem 1979; 254: 12020-12027. 14 Hamberg U, Tallberg Th. Preparation of monospecific antiserum to human LMW lcimnogen using immunosorbents. J Immunol Methods 1972; 2: 17-24. 15 Mancini (3, Carbonara AO, Heremans .IF. Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 1965; 2: 235-254. 16 Habermann E. Uber pH-bedingte Modificationen des kirtln-liefernden a-globulins (IGninogen) aus Rinderserum und das Molekulargewicht yon Kininogen I. Biochem Z 1963; 337: 440-448. 17 Armstrong DAJ, Stewart JW. Anti-heparin agents as inh/bitors of plasma kinin formation. Nature 1962; 194: 689. 18 Uchida Y, Katori M. Differential assay method for high molecular weight and low molecular weight klninogens. Thromb Res 1979; 15: 127-134. 19 Becker W. Determination of antisera titres using the single radial immunodiffusion method. Immunochemistry 1969; 6: 539-546. 20 Pierce .IV, Webster ME. The purification and some properties of two different kallidinogens from

220

21

22

23 24 25 26

human plasma. In: Erd6s EG, Back N, Sicuteri F, eds. Hypotensive peptides. New York: SpringerVerlag, 1966: 130-138. Mori K, Sakamoto W, Nagasawa S. Studies on human high molecular weight (HMW) kininogen. II1. Cleavage of HMW kininogen by the action of human salivary kallikrein. J Biochem 1981; 90: 503-509. Bouma BN, Kerbiriou DM, Vlooswijk RAA, Griffin JH. Immunological studies of prekallikrein, kallikrein and high-molecular weight kininogen in normal and deficient plasmas and in normal plasma after cold-dependent activation. J Lab Clin Med 1980; 96: 693-709. Proud D, Pierce JV, Pisano .J.J. Radioimmunoassay of human high molecular weight kininogen in normal and deficient plasmas. J Lab Clin Med 1980; 95: 563-574. Sakamoto W, Nishikaze O. Purification and immunochemical properties of human low molecular weight kininogen. J Biochem 1979; 86: 1549-1557. Hamberg U, Tallberg Th. Kininogen as an acute phase plasma protein. Abstract Xlth Int. Congress Biochemistry Toronto 1979; Molecular mechanisms of disease: 670. Kutzbach C, Schmidt-Kastner G. Kallikrein from pig pancreas. Hoppe-Seyler's Z Physiol Chem 1972; 353: 1099-1106.