Quantitative analysis of fibrin-binding affinity of fibrinolytic components by frontal, affinity chromatography

THROMBOSIS RESEARCH Supplement VIII; 81-90, 1988 0049-3848/88 $3.00 t .OO Printed in the USA. Copyright (c) 1988 Pergamon Press plc. All rights reserved.

QUANTITATIVE ANALYSIS OF FIBRIN-BINDING AFFINITY OF FIBRINOLYTIC COMPONENTS BY FRONTAL AFFINITY CHROMATOGRAPHY

+Kazama

M., Tahara

C., Abe T. and *Kasai

K.

Department of Medicine, School of Medicine and *Department of Pharmaceutical Sciences, Teikyo University, Tokyo, JAPAN

ABSTRACT Binding affinity of fibrinolytic factors to insolubilized lysine and fibrin was quantitatively measured by frontal affinity chromatography using lysine-Toyopearl and fibrin-Sepharose column. The highest binding affinity was found with recombinant tissue-type plasminogen activator,(t-PA), followed by lysyl-plasminogen and glutamylplasminogen (Glu-PLg) with intermediate affinity, but very low affinity by single chain UK-type plasminogen activator, high molecular weight UK and low molecular weight UK. At the coexistence of EACA, fibrin-binding affinity of Glu-PLg was greatly reduced, but those of UK's were substantially unchanged. It was concluded that high fibrin-binding affinity of t-PA and plasminogens were largely related to the lysine-binding affinity of these enzymes, but that of UK's would be related to the other binding affinity.

INTRODUCTION Fibrin-binding affinity is one of the major properties of fibrinolytic factors. Recent advances in the production of thrombolytic agents have been aiming to create those which possess the higher fibrin-binding affinity. Although the fibrinbinding affinity of these factors has been analyzed with the use of clotting fibrin or insolubilized fibrin, these results did not necessarily draw an clear conclusion on this matter. Because they were devoid of quantitative analytical methods or because they were drawing improper interpretation, conflicting ---------------------------------_-----_------_----------_-----_-

frontal affinity chromatography, words: fibrin-binding, plasminogens, urokinases, + To whom correspondence should be addressed

Key

81

lysine-binding, tPA

82

conclusions

FRONTAL AFFINITY CHROMATOGRAPHY

Suppl.

VIII,

1988

have been described.

Frontal affinity chromatography was first reported by Kasai and Ishii(1). They used glycylglycyl-L-arginine-Sepharose column to calculate Kd between trypsin and arginine, and Ki of NThis method of analysis substituted arginines to trypsin. facilitates very convenient for investigation of the specific In this paper, binding interaction between protein and ligands. affinity of fibrinolytic factors to insolubilized lysine or fibrin was quantitatively measured by the use of frontal affinity chromatography. MATERIALS AND METHODS 1. Principle of Frontal Affinity Chromatography The elution volume of a solution containing a solute is dependent on its binding affinity to a ligand conjugated to a column. When a constant concentration of a solute was passed through a column with a immobilized ligand, the concentration in the eluted solution will increase until it reaches the same level of the starting solution, and then this maximum level will be maintained constant. In this state, a dynamic equilibrium will be established , where the amount of the solute adsorbing to the ligand is equal to that dissociating from (2). The approximate elution volumes of solutes A and B can be substituted with the apparent elution volume, Va and Vb, where the concentration of the each solute in the eluted solution if the elution reaches to l/2 of the starting concentration, profile forms point symmetry. When solute A has no affinity to the column, difference between Vb and Va indicates the binding affinity of the solute B to the column, namely Ve. The conjugated amounts of ligand and constituents as well as concentrations are crucial to obtain good results of frontal affinity chromatography. The suitable condition has to be decided by preliminary experiments. 2. Fibrinolytic Factors Glu-plasminogen (Glu-PLg) was prepared from fresh human plasma by affinity chromatography with lysine-Sepharose column. The final preparation had the activity of g.OCTAu/mg protein. Human lysyl-plasminogen (Lys-PLg) was supplied from Green Cross (Osaka, Japan), of which activity was 21 CU/mg protein and NH2terminal amino acids were lysine and methionine when it was analysed by Edman's method. Low molecular weight urokinase Nippon Abbott, Co. Ltd.( Japan). High (H-UK) and recombinant single chain activator (scu-PA) were supplied from exhibited the activities of 156,000 respectively. Recombinant tissue-type PA) was supplied from Eizai, Co.,

(L-UK) was supplied from molecular weight urokinase urokinase-type plasminogen Green Cross (Japan), which and 150,0001U/mg protein plasminogen activator (tLtd.(Japan), which was a

Suppl. VIII,

1988

FRONTAL AFFINITY CHROMATOGRAPHY

8:

mixture of single and two-chain types containing finger domain and exhibited activity of 10,0001U/mg protein. Human fibrinogen was purified from human Cohn Fraction-I (Fibrinogen-Midori:Green Cross) using 2M beta_alanine)(3), followed by the removal of plasminogen using lysine-Sepharose column. The clottability of the final preparation was more than 90%. 3. Activity Assays of Fibrinolytic Factors Amidolytic activity of plasminogens was measured using PLG KIT (Daiichi Chemicals, Co. Ltd, Japan) and those of L-UK, H-UK and t-PA was measured using%2444 and 2288 (Kabi Vitrum, Sweden) and that of scu-PA was measured according to the method by Kasai The coexistence of EACA or arginine up to 100mM in et a1.(4). specimens did not influence the activity assay of fibrinolytic factors. 4. Preparation of Affinity Columns(l) For preparation of immobilized lysine column, Toyopearl HW65 (Toyo Soda Co. Ltd., Japan) was activated with BrCN. Twenty milliliters of 50 ug/ml of Pro-Phe-Lys (Sigma Chemical Co., U.S.A.) dissolved in O.lM NaHC03 was mixed with 20ml of the activated gel at 5+C overnight. After the gel was blocked with O.lM ethanolamine at room temperature for 2 hrs, the gel was washed and suspended in 50mM Tris buffer (pH=7.2) containing 0.15M NaCl and the gel was packed in a 2ml-sized plastic syringe to prepare a gel column of 2.0ml (8X14mm). Immobilized fibrin monomer column was prepared by a modification of the method by Heene and Mathias(5). The modification was that 2.5mg/ml of purified human fibrinogen was reacted with l.Oml CN -activated Sepharose (Pharmacia Fine Chemicals) at 4+C overnight. The fibrin-Sepharose gel was packed into a 2ml-sized plastic syringe to prepare a gel column of l.Oml (8X7mm). 5. Calculation of Ve's of Fibrinolytic Factors by Frontal Affinity Chromatography After the dilute solution of each fibrinolytic factor was prepared with 50mM Tris buffered saline (pH=7.2), it was continuously passed through a column of lysine-Toyopearl or fibrin-Sepharose at the rate of Gml/hr at 4+C until the activity level in the eluted solution reached to the same level in the The applied specimen to obtain the frontal elution profile. column was washed with 200mM EACA, 5M urea and then the starting The elution buffer successively between chromatographic runs. were calculated as it was described volumes (Ve's) of specimens in the principle.

RESULTS 1.

Lysine Binding Affinity of Fibrinolytic Factors The frontal elution profiles of Glu-PLg, Lys-PLg and albumir were obtained by passing each 35.9nmol (3.2ng/ml) of plasminogens or 135 ug/ml of albumin through lysine-Toyopearl column. Elutior

FRONTAL AFFINITY

84

Suppl.

CHROMATOGRAPHY

VIII,

1988

Lys-PLg was 1.8 and 10.4ml volumes (Ve's) of Glu-PLg and respectively and thus the relative lysine-binding affinity of Lys-PLg was 5.8 folds that of Glu-PLg (Fig. 1).

(“1.) 100.

"FM+-

jO.____ !_,

I

I

G'U~Y-L

Lys-PLg

ELUTION

__

VOLUME

;

,I

20

Elution Volume

(ml)

VE (ml)

RELATIVE VE

2.2

0

0

6.4

4.2

1

12.6

10.4

2.5

33

(ml)

Fig. 1 Frontal Affinity Chromatography of Plasminogens on Pro-Phe-Lys-ToyopearlColumn Frontal elution profiles were obtained by successively passing albumin, Glu-PLg or Lys-PLg through an identical 2.0ml LysToyopearl column. Ve's of plasminogens were calculated by subtracting the elution volume of albumin from those of Glu-PLg or Lys-PLg. The lysine-binding affinity of plasminogen activators were also compared. Each 40.0nmol solution of,L-UK, H-UK, scu-PA or t-PA was passed successively through an identical lysineToyopearl column. Ve's of L-UK, H-UK, scu-PA and t-PA were 0.3m1, 1.8m1, 2.6ml and 62.lml respectively. The relative lysine-binding affinity of scu-PA was 1.5 folds and that of t-PA was 35 folds that of H-UK. As regards to lysine-binding affinity, the highest affinity was found with t-PA, followed by Lys-PLg and then by Glu-PLg, while that of scu-PA was a little greater than that of H-UK (Table 1). Lysine-binding

Table 1 Affinity of Plasminogen Ve (ml)

Albumin L-UK H-UK scu-PA t-PA

0 0.3 1.8 2.6 62.1

Activators

Relative Affinity 0 0.16 1.0 1.44 34.5

Suppl. VIII,

1988

FRONTAL AFFINITY CHROMATOGRAPHY

85

2. Fibrin-binding Affinity of Fibrinolytic Factors The frontal profiles of Glu-PLg, Lys-PLg and albumin were obtained on fibrin-Sepharose column. Each 35.9nmol of plasminogens or 135 ug/ml of albumin was passed through a fibrinSepharose column. Each plasminogen was eluted forming stepwise profile, indicating that the fibrin-binding affinity was heterogenous within one preparation of PLg. Although the higher than that of Glu-PLg, affinity was found with Lys-PLg the difference was not great between both plasminogens (data not shown). The elution profiles on fibrin-Sepharose column were obtained also of L-UK, H-UK, scu-PA and tPA respectively. The Ve of tPA was markedly greater than those of urokinases, more than 100 folds that of H-UK. L-UK had no fibrin-binding affinity, and that of scu-PA was 1.5 folds that of H-UK (Table 2).

Fibrin-bindinq

Table 2 Affinity of Plasminogen Ve(ml)

Albumin L-UK H-UK scu-PA t-PA

0 0 0.6 1.0 108

Activators

Relative

Affinity

0 0 1.0 1.7 180

3. Influence of EACA on Lysine-binding or Fibrin-binding Affinities of Fibrinolytic Factors. After equilibration of 2.0ml of lysine-Toyopearl column with 50mM NaCl and different con50mM Tris buffer containing centrations of EACA, 40nM Lys-PLg dissolved in the same buffer was passed through the column to analyze dose dependent change of The Ve's of Lys-PLg were 3.8ml and the frontal elution profile. 8.8ml at the coexistence of l.OmM and O.OlmM EACA respectively, whilethatof the control run without EACA was 13.6ml. When this not experiment were performed with SCU-PA, Ve of scu-PA was abolished even at the coexistence of 100mM EACA. The similar experiments were performed with Glu-PLg, H-UK, scu-PA and t-PA using fibrin-Sepharose column equilibrated with The affinity of Glu-PLg was different concentrations of EACA. Ve's of UK's almost lost at the coexistence of ImM EACA, while were not changed even at the coexistence of 100mM EACA. Those of t-PA were 9.2mland 35mlatthe coexistence of IOOmM and ImM EACA while that of the control run without EACA was respectively, However, the fibrin affinity of t-PA was not completely 250ml. abolished even at the higher concentration of EACA than IOOmM (Table 3).

FRONTAL AFFINITY CHROMATOGRAPHY

Suppl. VIII,

Table 3 Dose Dependent Decreases -of Ve's of Fibrinolytic at the Coexistence --of EACA on FGrin-Sepharose -Concentration 100 10 Glu-PLg H-UK scu -PA t-PA

3.2

3.8 9.2

5.0

Factors Column

of EACA (mM) 1 .o 1

2.6

1988

.o

-

35

0 28.8mi

2.9ml 3.3ml 250 ml

4. Influence of Lysine and Arginine on Fibrin-binding Affinity of Urokinases of IOOmM EACA, frontal affinity At the coexistence chromatography of H-UK or scu-PA was performed on fibrinAfter frontal elution profiles were obtained, Sepharose column. the column was washed with Tris buffer until no activity was detected in the eluate , then the column was flushed with IOOmM The apparent elution of arginine contained in the same buffer. UK's was found by the 2nd elution (Fig. 2).

PM 100

1OOmM 4 J

Elution

Volume (ml) Fig.

2

Independent elution effect of lysine and arginine -on the fibrinbinding affinity of H-UK andscu-PA After the adsorptionof H-UK or scu-PA on fibrin-Sepharose column, they were eluted with IOOmM arginine forming an. apparent the fibrin-binding affinity of these peak, which indicated enzymes were related to both lysine and arginine.

DISCUSSION

form

Frontal analysis in affinity chromatography is advantageous both theoretical and experimental viewpoints for the

Suppl. VIII,

1988

FRONTAL AFFINITY CHROMATOGRAPHY

analysis of Kinetics between enzymes and ligands, because chromatographic data can be related easily to the amount interacting molecules and the equilibrium constant(2). binding-affinity is regulated at least by two factors, i.e. dissociation constant of enzymes and the available numbers of We utilized this method for quantitative analysis ligands. lysine- and fibrin- binding affinity of fibrinolytic factors.

8;

the of The the the of

By incorporation experiment of clotting fibrin, Whitaker et a1.(6) reported that Lys-PLg was incorporated 4-10 folds that of Glu-PLg into clotting fibrin, according to their concentrations. And the amount of incorporation was reduced when clotting was which suggested interference of the performed in plasma, antiplasmin in plasma, incorporation of these enzymes by lbrin binding affinity of Lucas et al. (7) also reported higher $, Lys-Plg than that of Glu-PLg, utilizing uniform suspensions of human fibrin. They calculated the dissociation constants of Lysor corss-linked, as PLg and Glu-PLg to fibrin, noncross-linked 0.32uM and 38uM respectively. As regards to different results of fibrin binding affinity of plasminogens (8,9), they criticized that these investigators did not consider nonspecific binding and trapping during fibrin clot formation. Utilizing frontal affinity chromatography, nonspecific trapping of fibrinolytic factors was prevented and enabled us the The calculation of relative binding affinity of plasminogens. that of Glu-PLg. lysine-bindyng of Lys-PLg was 5.8 folds However, the affinity to fibrin was not possible, because there was heterogenous fibrin-binding affinities in one preparation of each plasminogen. High fibrin-binding affinity of scu-PA (pro-UK) has been Kasai et a1.(4) reported 40 to 65% fibrin clot binding reported. of pro-UK compared with 0 to 10% binding of H-UK or two-chain UK (tc-UK). The recovery of pro-UK in the supernatant was measured Sumi et al. (10) and Bando et al.(ll) by amidolytic activity. reported that (H-) UK did not bind to fibrin-Sepharose column but proUK was almostadsorbedtothe columnand 70% of its enzymatic activity was recovered by the elution with 0.2M arginine. On the other hand, Stump et a1.(12) observed no binding of pro-UK to fibrin clot in their experiments in which the recovery Gurewich (13) demonstrated of UK was measured by its antigen. that there was full recovery of radioactivity from a labeled UK in the clot supernatant in the similar experiment. The only change observed in the probe was conversion to a two-chain form, inhibited the Thrombin which was devoid of amidolytic activity. activatability of pro-UK(4,14,15) and this degeneration will be associated with a loss of fibrin-binding affinity of pro-UK (16). Gurewich criticized absorption experiments of pro-UK which used clotting fibrin. By these methods of analysis, the presence of the pro-UK could well have been obscured by its of binding degradation to a two-chain nonbinding derivative. Thrombin effect on the method of may introduce an artifact that, depending may masquerade as fibrin binding when activity is analysis,

88

FRONTAL AFFINITY CHROMATOGRAPHY

suppi. VIII,

1988

measured, or conversely may obscure it when antigenic activity or Using fibrin clotted by Ancrod, radioactivity is measured. Curewich (16) found that no significant fibrin clot binding of pro-UK, based on recovery of amidolytic activity or radioactivity and the experimental was demonstrable, in the supernatant, findings with pro-UK and with tc-UK were comparable. In our experiment of frontal affinity chromatography using the degenerative influence of thrombin fibrin-Sepharose column, on scu-PA was negligible, if any. The amidolytic activity in the effluent was not found per se, but was fully developed by the addition of small amount of plasmin, which demonstrated that the applied scu-PA was almost recovered intact in the effluent. Using frontal affinity chromatography, it was revealed that lysine- as well as fibrin-binding affinity of UK family were very low. Among them, the binding affinity of scu-PA onto lysine and fibrin was 1.44 folds and 1.7 folds that of H-UK respectively. However, the weak lysine- or fibrin-binding affinity of UK family was not abolished by the coexistence of EACA in the solution, but they were eluted from the column with arginine. These results suggested that the fibrin-binding affinity of these enzymes would relate to another than lysine-binding. It is well known that NH -terminal residue is related to the binding site of enzymes to & igh molecular weight substrates. Enzymes with shorter NH2-terminal residues, such as trypsin and chymotrypsin, have less substrate specificity but those with longer NH2-terminal residues exhibit high substrate specificity(l7). Finger domain is a component of NH2-terminal residue of tPA and fibronectin. Because this domain is not found in UK molecule, this is supposed to specific site of high fibrin binding site of tPA and fibronectin (18). However, recombinant tPA without finger domain exhibited also high fibrin affinity. This tPA was adsorbed onto insolubilized fibrin and eluted with 0.2M Arginine.(l9). It is strongly suggested that the high lysine binding site of t-PA to fibrin was locatedinthe kringle2 segment(20). We found markedly high lysine- and fibrin-binding affinity of t-PA by frontal affinity chromatography. The lysine-binding affinity of t-PA was 34.5 folds and its fibrin-binding was 180 folds respectively that of H-UK. Its high fibrin-binding affinity was reduced to l/25 by the coexistence of IOOmM EACA in the solution, which suggested that the fibrin-binding affinity of this enzyme would largely depend on the lysine-binding. Frontal affinity chromatography using the column of Pro-PheLys-Toyopearl or fibrin monomer-Sepharose enabled us to quantify lysine-binding and fibrin-binding affinity of fibrinolytic factors, and it was also revealed that high fibrin-binding affinity of plasminogens and t-PA may be related largely to lysine-binding affinity. This study was partly supported Government of Health and Welfare,

by the fund from 1987.

the Japanese

Suppl. VIII,

1988

FRONTAL AFFINITY CHROMATOGRAPHY

89

REFERENCES

analvsis of affinity (1) Kasai, K. and Ishii, S. Quantitative chromatography of trypsin. A new technique for-investigation of protein-ligand interaction. J. Biochem. 77, 261-264, 1975. (2) Kasai, K. Theory for its biomolecules.

and Oda, Y. Frontal affinity chromatography: application to studies on specific interactions of J. Chromatography 376, 33-47, 1986. -

(3) STRAUGHN III, W. preparing fibrinogen. 1966.

and WAGNER, R.H. A simple method for Thromb. Diath. Haemorrh. 16, 198-205,

(4) KASAI, S., ARIMURA, H., NISHIDA, M, and SUYAMA, T. Proteolytic cleavage of single chain pro-urokinase induces conformational change which follow activation of the zymogen and reduction of its high affinity of fibrin. J. Biol. -Chem. 260, 2377-2381, 1985. (5) HEENE, D.L. and MATTHIAS, derivatives on insolubilized Thromb. Res. 2, 137-154, 1973. ---

F.R. Adsorption fibrinogen and

of fibrinogen fibrinmonomer.

(6) WHITAKER, A-N., ROWE, E.A., MASCI, P.P., JOE, F. and GAFFNEY, P.J. The binding of glu- and lys-plasminogen to fibrin and their subsequent effects on fibrinolysis. Thromb. Res. 19, 381-391, --1980.

(7) LUCAS, M.A., FRETTO, L.J. and MCKEE, P.A. The binding of human plasminogen to fibrin and fibrinogen. J. Biol. Chem. 258, --4249-4256, 1983. (8) CERERHOLM-WILLIAMS, S.A. and streptokinase-plasminogen Trans. 5, 1441-1443, 1977.

The binding of plasminogen, plasmin Biochem. Sot. activator to fibrin.

(9) SUENSON, E. and THORSEN, S. Secondary-site plasmin, Lys-plasmin and miniplasmin to fibrin. 619-68, 1981.

binding of GluBiochem. -J. 197,

(10) SUMI, H., MARUYAMA, M., MATSUO, 0, MIHARA, H. and TOKI, N. Higher fibrin-binding and thrombolytic properties of single (letter) molecular weight urokinase. polypeptide chain-high Thromb. Haemost. 47, 297, 1982. (11) BANDO, H., OKADA, K. and MATSUO, pro-urokinase in vitro. Fibrinolysis

effect 0. Thrombolytic 1, 169-176, 1987.

of

(12) Stump, D.C., Lijnen, H.R. and Collen, D. Purification and characterization of single-chain urokinase-type plasminogen Chem. 261, 1274J. Biol. -activator from human cell cultures. 1278, (13)

1986.

GUREWICH,

V. and PANELL,

R.

Inactivation

of single

chain

90

FRONTAL AFFINITY CHROMATOGRAPHY

Suppl. VIII,

1988

urokinase (pro-urokinase) by thrombin and thrombin like enzymes; Release of the findings to the interpretation of fibrin binding experiments. Blood 69, 769-772, 1987. Plasmin and thrombin (14) CONFORTI, G. and LOSKUTOFF, D.J. modulate plasminogen activation by fibro-sarcoma cells. (Abstr.) Thromb. Haemost. 54, 71, 1985. Localization of the (15) ICHINOSE, A., TAKIO K. and FUJIKAWA, K. binding site of tissue-type plasminogen activator to fibrin. J. Clin. Invest. 78, 193-169, 1986. (16) GUREWICH, V. and PANELL, R. Fibrin binding and zymogenic single -chain properties of urokinase (Pro -urokinase). Sem.Thromb. Hemost. 13, 146-151, 1987. (17) FRUTON, J.S. The specificity of proteinases toward protein and Biological Control. Reich, E., substrates. In: Proteases Rifkin, D.B. and Shaw, E (Ed.), Cold Spring Harbor Laboratory, Cold Spring Harbor, N-Y., 1975, ~~30-50. (18) BANYAI, L., VARADI, A. and PATTHY, L. Common evolutionary origin of the fibrinobinding structure of fibrionectin and tissue-type plasminogen activator. FEBS Letters 163, 37-41, 1983.

(19) KAGITANI, H., TAGAWA, M., HATANAKA, K., IKARI, T., SAITO, A ., BANDO, H., OKADA, K. and MATSUO, 0. Expression in E. coli of finger-domain lacking tissue-type plasminogen activator with high fibrin affinity. FEBS Letters 189, 145-149, 1985. (20) ICHINOSE, A., FUJIKAWA, K. and SUYAMA T. The activation prourokinase by kallikrein and its inactivation by thrombin. Biol. Chem. 261, 348-349, 1986. ---

of J 2