Photoaffinity labeling of functionally different lysine-binding sites in human plasminogen and plasmin

Biochimica et Biophysica Acta 830 (1985) 187-194

187

Elsevier BBA 32260

P h o t o a f f i n i t y labeling of functionally different lysine-binding sites in h u m a n p l a s m i n o g e n and plasmin T h o m a s J. R y a n a n d M a r y C. K e e g a n The Wadsworth Center for Laboratories and Research, New York State Department of Health, Albany. N Y 12201 (U.S.A.)

(Received January 2nd, 1985) (Revised manuscript received May 3rd, 1985)

Key words: Lysine-bindingsite; Plasminogen; Plasmin; Photoaffinity label; (Human)

Photoaffinity labeling of human plasmin using 4-azidobenzoylglycyI-L-lysine inhibits clot lysis activity, wnue the activity toward the active-site titrant, p-nitrophenyl-p'-guanidinobenzoate, or a-casein are maintained. Photoaffinity labeling of native Glu-plasminogen with the same reagent causes incorporation of approximately 1.5 moi label per tool plasminogen. This labeled plasminogen can be activated to plasmin by either urokinase or streptokinase. The resulting plasmin has full clot lysis activity and can be subsequently photoaffinity labeled with a loss of clot lysis activity. The rate of activation of labeled plasminogen by urokinase is increased relative to that of native plasminogen, c-Aminocaproic acid blocks incorporation of photoaffinity label into both plasminogen and plasmin, indicating that the labeling is specific to the lysine~binding sites. The labels are located in the kringle ! + 2 + 3 fragment in either photoaffinity-labeled plasminogen or plasmin. These results indicate that the specific lysine-binding site blocked in plasmin acts in concert with the active-slte in binding and using fibrin as a substrate. This clot lysis regulating site is not available for labeling in plasminogen, but is exposed or changed upon activation to plasmin. The different lysine-binding sites labeled in plasminogen may regulate the conformation of the molecule as evidence by an enhanced rate of activation to plasmin.

Introduction The two major components of the fibrinolytic system, plasminogen and plasmin, contain 'lysinebinding sites' distinct from the active site. These sites bind antifibrinolytic compounds such as caminocaproic acid and C-terminal lysine peptides. Native Glu-plasminogen contains one high-affinity site and five weaker binding sites [1], which are in the N-terminal portion of the plasminogen structure containing five homologous peptide domains termed 'kringles' [2]. The lysine-binding sites mediate both the binding of plasminogen to fibrin for inclusion in a clot [3] and the enhancement of the rate of activation of plasminogen when bound to fibrin [4]. The sites also regulate the action of

plasmin on fibrin [5] and the inactivation of plasmin by the main plasmin inhibitor in blood, a2-plasmin inhibitor [6]. A unified model for the role of the lysine-binding sites in fibrinolysis has been proposed [7]. In a previous study [8], we photoaffinity labeled plasmin using the nitrene precursor, 4-azido[carboxyl-14C]benzoylglycyl-L-lysine at a specific lysine-binding site that acts in concert with the active site in binding fibrin as a substrate. This photolabile compound, when irradiated, generates a nitrene capable of insertion into any available bond. Using the same procedure native Gluplasminogen has now been specifically labeled at the lysine-binding sites and characterized to determine the roles of the blocked sites. In addition,

0167-4838/85/$3.30 ~ 1985 Elsevier Science Publishers B.V. (Biomedical Division)

188 the plasmin resulting from activation of labeled plasminogen has been compared with directly labeled plasmin. Initial degradation studies havc been carried out to locate the labels in the overall structure of the two molecules.

Experimental procedures Materials Reagents. Bovine fibrinogen (95% ciottable, Pentex, Miles Laboratories) and streptokinase (Varidase, Lederle Laboratories) were used as received. Human thrombin was prepared by the method of Fenton et al. [9] except that Taipan venom (Australian Reptile Park, Gosford, N.S.W.) was used to activate the prothrombin. The thrombin had a specific activity of 1 421 US units/mg and was 84% active by p-nitrophenyl-p'-guanidinobenzoate titration [10]. Human urokinase (WinKinase, Winthrop Laboratories) was supplied through the courtesy of Dr. Joseph Fratantoni, National Heart, Lung and Blood Institute, National Institutes of Health. Cohn fraction III pastes were obtained as gifts from Dr. Brian Landis, Armour Pharmaceutical Company. Plasminogen. Native Glu-plasminogen was isolated from outdated human plasma [11] using the modifications of Markus et al. [1]. Preparations gave a single band when analyzed by SDS-polyacrylamide gel electrophoresis on 12.5% polyacrylamide gels [12], had specific activities of 19 to 27 units/mg, and only N-terminal glutamic acid when analyzed by a dansylation procedure [13,141. The activities of plasminogen, plasmin and urokinase are expressed in units of enzyme activity as defined by the Committee on Thrombolytic Agents [15]. Lys-plasminogen was isolated from Cohn fraction III pastes by procedure of Liu and Mertz [16] except that 2% poly(ethylene glycol) 6000 and 0.02 M L-lysine were incorporated in the extraction buffer to precipitate unwanted protein. This plasminogen also gave a single band when analyzed by SDS-polyacrylamide gel electrophoresis and had activities of 23-28 units/mg. Plasmin. Lys-plasminogen was activated to plasmin with urokinase by the method of Robbins and Summaria [17], except that only 4--5 h were required. Preparations were 60 to 70% active by p-nitrophenyl-p'-guanidinobenzoate titration as-

suming Mr = 81000 and L2s ":~0 ,,,~,= 17.0 [18]. Analysis by SDS-polyacrylamide gel electrophoresis indicated that the major contaminant was degraded plasmin, and plasminogen. Photoaffinity labeling reagents. 4-Azido[ carboxyl- 14C]benzoylglycyl-l.-lysine monohydrochloride monohydrate was synthesized as previously described [8]. The final product had 384299 cpm/~mol. 4-Azido[ benzoyl-3,5-SH]benzoylglycyl H,-lysine monohydrochloride monohydrate was prepared as described for the 14C-labeled compound using succinimyl 4-azido[3,5-3H]benzoate (New England Nuclear) to acylate glycyl-~-tbutoxycarbonyl-I,-lysine acetate. The vield was 67%. The peptide had 182 289 cpm/p, mol.

Methods Photoaffinity labeling. The method has been described [8]. A solution of 4-azidobenzoylglycyl-Llysine and plasmin(ogen) in 0.3 M potassium phosphate buffer (pH 7.6) was irradiated with 300 nm light at 2°C for 30 min. For plasmin, 10% glycerol was included in the buffer to stabilize the enzyme. The solution was either diluted 1:3 with water or dialyzed against 0.1 M potassium phosphate buffer (pH 7.6) to reduce the concentration of labeling reagent, photoproducts and, in the case of competition experiments, ~-aminocaproic acid. The sample was applied to a column of L-lysinecoupled Sepharose, equilibrated in the dialysis buffer. Unbound label and photoproducts were washed off and the photolabeled protein eluted with the equilibration buffer containing 15 mM c-aminocaproic acid. After activity measurements, the labeled plasmin samples are inhibited with 10 mM diisopropylfluorophosphate to prevent autolysis during subsequent procedures. Activity measurements. Clot lysis activity was measured by the method of Carlin and Saldeen [19]. An aliquot of the photolysis solution was diluted approx. 1:50 with 0.1 M Tris buffer (pH 7.5)/0.15 M NaCI. A portion of this solution (20 ~tl) was mixed with 0.6 ml of the dilution buffer and 0.2 ml of fibrinogen solution (5 mg/ml). The clot was initiated by adding 0.1 ml of human thrombin solution (73 US units/ml). The lysis time was taken as the time from the addition of thrombin to the complete disappearance of clot fragments. Samples were diluted to give lysis times

189 between 3 and 20 min. The loss of clot lysis activity was determined by comparing the activity of the starting plasmin solution containing 4azidobenzoylglycyl-L-lysinewith the activity of the same solution after photolysis. Plasminogen samples were activated with urokinase (200 units/ml) for 5 min at 37°C before measuring the clot lysis activity. Activities toward a-casein (Worthington Diagnostic Systems) were measured by a modification of the method of Johnson et al. [15] in which 16% trichloroacetic acid was used to precipitate casein and plasmin. A portion of the supernatant (1 ml) was mixed with 1 ml of 4 M sodium acetate buffer (pH 6.2), reacted with ninhydrin (1 ml) at 100°C for 15 min [20], and diluted with ethanol/water (1:1, 5 ml). The absorbance at 570 nm was used to determine the concentration of acid-soluble peptides. The assay was standardized using a plasminogen supplied by Dr. Alan J. Johnson, New York University Medical Center. Protein concentrations in photoaffinity labeled samples were determined by the method of Beardon [21] using native plasminogen as the standard. The kinetic parameters for the activation of native or labeled plasminogen by urokinase were determined by the method of Philo and Gaffney [22] with modifications suggested by Gilboa et al. [23]. Plasminogen (0.74-11.9/xM) in 0.05 M Tris buffer (pH 7.4)/0.10 M NaCI/0.01 M Triton X100 was incubated with urokinase (22 units/ml) for 10 min at 37°C. Photoaffinity-labeled plasminogen (1.1-15.1 #M) was similarly activated with urokinase (2.2 units/ml). The activation mixture (50 #i) was diluted 1:20 in a cuvette with activation buffer containing 0.11 M lysine and the plasmin-specific substrate, H-D-Val-Leu-Lys-pnitroanilide (0.35 raM). The initial increase in absorbance at 405 nm due to the hydrolysis of the substrate by the plasmin produced during the activation step was measured at 37°C. The kinetic constants K m and V,,a, were calculated from double-reciprocal plots of (A~5 ,m/min) -1 versus [plasminogen]-1. The increase in absorbance was converted to tool plasmin by constructing a calibration curve using human plasmin of known molarity. The values for K m and Vma~ were determined using a computer program for a Dig-

ital Equipment Corp. VAX 11/780 computer, which fits experimental data to the LineweaverBurk equation using a weighted linear regression technique based on the algorithm of Paulson and Nicklin [24]. Elastase digestion of labeledproteins. Photoaffinity-labeled plasmin(ogen) was digested with procine pancreatic elastase (Worthington Diagnostic Systems) in the presence of soybean trypsin inhibitor, and the resulting fragments separated by chromatography on L-lysine°coupled Sepharose and Sephadex G-75 essentially as described by Sottrup-Jensen et al. [2]. The peptide fragments were hydrolyzed in 6 M hydrochloric acid in nitrogenflushed tubes at 110°C for 24 h, and amino acid analyses were performed in duplicate with a JEOL automatic amino acid analyzer.

Results

Photoaffinity labeling of plasmin We have compared the activities of plasmin and photoaffinity-labeled plasmin both on the nonspecific protein substrate, a-casein, and on a fibrin clot. The data in Table I for duplicate photolabeling experiments, using a 13-fold molar excess of labeling reagent to plasmin, indicate that labeling of plasmin does not diminish its activity toward casein but does significantly reduce activity towards its specific substrate, fibrin. Photolabeled plasmin samples required approximately twice the activity toward casein to produce a clot lysis time comparable to that of unmodified plasmin. To confirm that these changes in activity are due to blocking of the lysine-binding sites, we have examined the effect of the antifibrinolytic compound, ~-aminocaproic acid, on the incorporation of labeling reagent into plasmin. The amount of label covalently bound decreased with increasing c-aminocaproic acid concentration. The mol label per moi plasmin at selected molarities of ~aminocaproic acid were: 0.65 (0 M), 0.17 (0.05 M), 0.08 (0.50 M) and 0.03 (1.0 M). At the highest concentration of c-aminocaproic acid (1.0 M) 95% of the labeling was blocked, indicating that the lysine-binding sites in plasmin had been specifically modified.

190 TABLE I EFFECT OF DIRECT PHOTOAFFINITY LABELING ON PLASMIN ACTIVITY TOWARD AN ACTIVE-SITE TITRANT. FIBRIN. AND CASEIN The plasmin was photoaffinity labeled with p-azidobenzoylglycyI-L-lysine and the indicated amounts of activity toward the nonspecific substrate, a-casein, were assayed in a clot lysis assay. Sample

Titrant activity a

Activity on casein b

Activity on fibrin, lysis time (min) ~

% Initial lysis activity d

Plasmin e Photolabeled plasmin f Photolabeled plasmin f

1O0 96 90

0.006 0.014 0.013

12.0 14.0 16.5

1O0 36 33

% activity toward the active-site titrant p-nitrophenyl-p'-guanidinobenzoate, setting starting plasmin as 100%. t, Activity on casein of the sample used in clot lysis assay; unit of enzyme activity as defined by the Committee on Thrombolytic Agents [151. c Time required to completely dissolve a fibrin clot. ,t Clot lysis activity of starting plasmin was set equal to 100%. e Specific activity = 19 units/mg. r Specific activity = 25 units/mg.

Photoaffinity labeling of plasminogen The photoaffinity labeling of native Glu-plasminogen by 4-azJdo{benzoyl-3,5-3H]benzoylglycyl L-lysine was also inhibited by c-aminocaproic acid. The mol label per mol plasminogen at varying c-aminocaproic acid concentrations were: 1.55 (0 M), 0.43 (0.05 M), 0.11 (0.50 M), and 0.13 (1.0 M). At a concentration of 1 M c-aminocaproic acid, 92% of label incorporation was blocked. When plasminogen was photoaffinity labeled with a 20fold excess of reagent, approx. 1.5 mol of label per mol of plasminogen were covalently attached. This modified plasminogen could be activated by either urokinase or streptokinase. The kinetic parameters for the urokinase catalyzed activation of native Glu-plasminogen were found to be K m = 8.9 + 0.1 ~M and Vmax = 1.23 + 0.08 pmol. min- ~ • unit- ~. The parameters for activation of photoaffinitylabeled Glu-plasminogen (1.7 mol label/mol protein) were K m = 10.1 -+_0.2 ~aM and Vma~ = 25.7 + 0.1 pmol. min -~. unit-1. The proteolytic activity of plasmins derived from native plasminogen and photolabeled plasminogen were determined on casein and on a fibrin clot. In Table II, the activities are shown to be equal on both substrates. Photoaffinity-labeled Lys-plasminogen (plasminogen lacking the N-terminal peptide Glu 1Lys76 ) could also be activated to plasmin with full clot iysis activitiy (Table II).

Location of the labeled sites within the plasmin(ogen) molecule Photoaffinity-labeled plasminogen and plasmin TABLE 1I EFFECT OF PHOTOAFFIN1TY LABELING ON THE ACTIVITY O F THE PLASMINS R E S U L T I N G F R O M ACTIVATION OF PHOTOAFFINITY-LABELED PLASMINOGENS Plasminogen was activated with streptokinase and the plasmin activity measured on the nonspecific substrate a-casein. The plasminogen was then activated with urokinase, and the resultant plasmin measured in a clot lysis assay. Sample

Caseinolytic activity (units) c

Lysis time (min)

Glu-plasminogen (24 units/nag) Glu-plasminogen irradiated 30 rain without labeling reagent Glu-plasminogen phototabeled " Lys-plasminogen (27 units/mg) Lys-plasminogen irradiated 30 min without labeling reagent Lys-plasminogen photolabeled b

0.019

10.1

0.017 0.020 0.018

10.5 10.6 13.0

0.018 0.017

12.3 10.3

• 1.6 mol label/mol plasminogen; 20 units/mg. 1.3 mol label/tool plasminogen; 30 units/mg. c Activity on casein of sample used in clot lysis assay. Unit of plasminogen activity as defined by the Committee on Thrombolytic Agents [151.

191 08: 500

1

E c

I000

•

.

--.

II

E c

04

o20

E o.

<

,

i

400

u

?

u

START 0 2 m EACA 5OO

%

0

,

~o

,

,,o

I

~o

s'o

o

0

9

~

'

:

:

'.

20

40 FRACTION

'

FRACTION NUMBER

Fig. 1. Chromatography of elastase-digested photoaffinitylabeled plasmin on lysine-coupled Sepharose. The column (2,0 ×28 cm) was equilibrated in 0.1 M ammonium bicarbonate buffer (pH 8.3) at 2°C. At fraction 48, the buffer was changed to 0.1 M ammonium bicarbonate/0.2 M (-aminocaproic acid (EACA) (pH 8.3). A23o,m (0); 14C cpm (zx). Flow rate = 9.5 ml.cm-2.h-1 w e r e d i g e s t e d w i t h e l a s t a s e , a n d t h e p r o d u c t s isol a t e d u s i n g a p r o c e d u r e [2] w h i c h s e p a r a t e s k r i n g l e 1 + 2 + 3, k r i n g l e 4 a n d k r i n g l e 5 a t t a c h e d t o t h e

EACA

8

60

f~,

80

200

100

NUMBER

Fig. 2. Chromatography of lysine-coupled Sepharose-bound fraction (II) on Sephadex G-75. The column (2.5 × 95 cm) was equilibrated with 0.1 M ammonium bicarbonate buffer (pH 8.3) at 2°C. A230,m (O); 14C cpm (,',). Flow rate = 6.1 mlcm - 2. h - i. EACA, ¢-aminocaproic acid.

light chain. When the elastase digest of labeled p l a s m i n w a s a p p l i e d to a c o l u m n o f l y s i n e - c o u p l e d S e p h a r o s e , e s s e n t i a l l y all o f t h e r a d i o a c t i v i t y w a s bound and could be eluted with (-aminocaproic a c i d (Fig. 1). T h e p o o l e l u t e d w i t h c - a m i n o c a p r o i c

TABLE III AMINO ACID COMPOSITION OF LABELED PEPTIDE FRAGMENTS FROM PLASMIN AND PLASMINOGEN The results are expressed as molar ratios of the constituent amino acids based on aspartic acid value of 32 for the kringle 1 + 2 + 3 fragment. The values are for a 24 h hydrolysis and are uncorrected. Amino acid

Theoretical a

Plasmin fragment

Lys-plasminogen fragment

Glu-plasminogen fragment

Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine

17 10 15 32 22-25 b 20-21 25-28 25 -29 15 6- 8 5- 6 3 6 9-11 13 3

16.5 10.2 14.4 32 22.5 19.6 27.2 22.0 17.6 7.3 8.0 2.6 5.8 10.2 14.3 3.4

14.7 9.3 13.9 32 22.4 19.0 26.7 24.8 17.6 8.7 9.0 2.7 5.8 10.8 13.4 4.4

17.0 9.3 14.4 32 22.7 24.0 26.4 28.6 20.8 8.8 6.1 1.2 5.6 10.2 12.4 4.1

a The values are based on the sequence data in Ref. 2. h The values expressed as ranges are the number of amino acid residues expected for mixtures of the two molecular weight forms ot kringle 1 + 2 + 3 (residues Tyr-79 to Val-337, Tyr-79 to Val-353).

192

acid separated into three peaks when gel filtered on Sephadex G-75 (Fig. 2). The first major peak (pool A) contained all of the radioactivity. These pools were analyzed by SDS-polyacrylamide gel electrophoresis (data not shown). Radioactive label from pool A was found in two peptides of M r = 4 0 0 0 0 and 42000 with trace amounts in other fragments with similar molecular weights. Amino acid analysis (Table IIl) was consistent with these fragments being two molecular weight variants (residues Tyr-79 to Val-337. Tyr-79 to Val-353) of the kringle 1 + 2 + 3 portion of the plasmin molecule [2]. Pool B of the Sephadex G-75 chromatography contained a single peptide, which when analyzed by HPLC (TSK G-3,000 SW column in 0.1 M potassium phosphate (pH 7.0)) was homogeneous, with an M r between 7000 and 12000, consistent with that of kringle 4 [2]. The last peak from the Sephadex G-75 chromatography was identified as c-aminocaproic acid based on its molecular weight and reaction with ninhydrin. The fragments which did not bind to lysinecoupled Sepharose (Fig. 1, Pool I) were found to consist of one peptide with a molecular weight corresponding to the light chain of plasmin and trace amounts of other peptides of different molecular weight. Photoaffinity-labeled plasminogens gave a similar pattern of label incorporation when elastase digested and the fragments separated as described above (data not shown). Essentially all of the radioactive label was found in the fragment identified as kringle 1 + 2 + 3 by SDS-polyacrylamide gel electrophoresis and amino acid analysis.

Photoaffinity labeling of the plasmin resulting from activation of labeled plasminogen When photoaffinity-labeled Glu-plasminogen (1.5 mol 3H label/mol plasminogen) was activated by urokinase, a 52% active plasmin was obtained. When this plasmin was photoaffinity labeled with a 50-fold excess of 4-azJdo[carboxyl-14C]benzoylglycyl-L-lysine, 50% of the clot lysis activity was lost and 0.52 mol of 14C-containing label per mol plasmin was incorporated. After the double-labeled diisopropylfluorophosphate-inhibited plasmin was elastase digested and the fragments isolated as described above, both 14C and 3H were found in

the same fragment, which consisted of two bands of molecular weight 40000--42000 when analyzed by SDS-polyacrylamide gel electrophoresis. This was consistent with both labels being located in the peptide fragment corresponding to kringle 1 + 2+3. Discussion

We previously reported that human plasmin incorporated 0.55-0.75 mol label per mol enzyme when photoaffinity-labeled with 4-azido[carboxyl14C]-benzoylglycyl-L-lysine. The modified enzyme maintained activity toward an active-site titrant but lost clot lysis activity and had a reduced affinity for lysine-coupled Sepharose [8]. In this previous study, ~-aminocaproic acid appeared to afford poor protection against photoaffinity labeling. We have subsequently found that it is difficult to remove all traces of noncovalently bound labeling reagent and photoproducts. By being more careful to ensure the complete removal of noncovalently bound label, we have now demonstrated competition by c-aminocaproic acid with the photoaffinity-labeling reagent for alkylating specifically at the lysine-binding sites of plasmin. The high concentrations of c-aminocaproic acid required to block photoaffinity labeling completely may reflect the fact that we are attempting to inhibit a kinetic process that proceeds to 4-5 half-lives rather than just establish a competition between two ligands for binding to a set of sites. The effects of the photoaffinity-labeling of plasmin are confined to a loss of its specific lytic activity on a fibrin clot, since photolabeled plasmin maintains full activity towards both the active site titrant, p-nitrophenyl-p'-guanicYlnobenzoate, and the nonspecific protein substrate, a-casein. The slight increase in caseinolytic activity observed after photolabeling was not examined in detail, but may result from stabilization of the enzyme during the assay. The lysine-binding site(s) being labeled in plasmin appear to function in concert with the active site in binding fibrin as a substrate. Since photolabeling of the lysine-binding sites in either plasminogen or plasmin did not diminish the activity on a-casein of the resulting plasmin, it is expected that if the lysine-binding sites were completely blocked, plasmin would still

193 retain a diminished ability to digest fibrin. Plasmin would lose its unique clot lysis acitivity but retain its trypsin-like proteolytic activity. In an earlier report [8], we showed that photoaffinity-labeled native plasminogen displayed a reduced affinity for l_-Iysine-coupled Sepharose. Relative to that required for native plasminogen, a lower concentration of c-aminocaproic acid eluted the photoaffinity-labeled protein from the affinity matrix. Native plasminogen after being photoaffinity labeled could be activated to plasmin by either urokinase or streptokinase. Determination of the kinetic constants indicated that the acceleration of the activation by urokinase was due to an approximately 20-fold increase in Vmax. A similar increase has been observed when c-aminocaproic acid binds to low-affinity lysine-binding sites and causes a conformational change in the Glu-plasminogen molecule [25]. Since photoaffinity-labeled Lys-plasminogen generates a plasmin with full clot lysis activity, the N-terminal preactivation peptide in Glu-plasminogen does not appear to be involved in blocking a clot lysis regulating lysine binding site. These results can be interpreted in terms of a model of plasmin in which fibrin is b o u n d as a substrate both at the active-site and at a lysine-binding site in the kringle 1 + 2 + 3 portion of the molecule. This lysine-binding site is blocked by direct photoaffinity labeling of plasmin. Different lysine-binding sites in kringle 1 + 2 + 3 are blocked when Glu-plasminogen or Lys-plasminogen are photoaffinity-labeled. Markus and co-workers [26] showed that the kringle 1 + 2 + 3 fragment contains at least two lysine-binding sites, one strong site and one of weaker affinity. Kringle 1, as an isolated peptide fragment, has been shown to have one high-affinity binding site [27]. The photoaffinity label in plasminogen may be distributed between these sites. At least one of the sites appears to control the c o n f o r m a t i o n of the molecule, since the rate of activation is enhanced by photoaffinity labeling. The fact that labeled Glu- or Lys-plasminogen when activated yield plasmins with full activities toward casein and fibrin indicates that the lysine-binding site in plasmin used for binding of fibrin as a substrate is exposed or changed in c o n f o r m a t i o n upon activation, since it cannot be labeled in plasminogen. Plasmin should contain

both sets of sites, but the newly exposed site must have a higher affinity for the labeling reagent. If this model is correct, plasmin resulting from the activation of photoaffinity-labeled plasminogen should be susceptible to further photoaffinity labeling at the exposed lysine-binding site with a loss of clot lysis activity. By use of reagents containing either 14C or 3H, this has been successfully demonstrated. We are currently locating the labeled sites within the kringle 1 + 2 + 3 fragment to expand and to test this model further.

Acknowledgement This work was supported in part by a grant-inaid from the Northeastern New York Chapter of the American Heart Association.

References 1 Markus, G.. DePasquale, J. and Wissler. F.C. (1978) J. Biol. Chem. 253. 727-732 2 Sottrup-Jensen. L.. Claeys, H., Zajdel, M., Peterson. T.E. and Magnusson, S. (1978) in Progress in Chemical Fibrinolysis and Thrombolysis (Davidson, J.F., Rowan, R.M., Samama, M.M. and Desnoyers. P.C., eds.), Vol. 3, pp. 191-209, Raven Press. New York 3 Adams Lucas, M., Fretto L.J. and McKee, P.A. (1983) J. Biol. Chem. 258. 4249-4256 4 Claeys, H. and Vermylen, J. (1974) Biochim. Biophys. Acta 342, 351-359 5 Landmann. H. (1973) Thromb. Diath. Haemorrh. 29, 253-275 6 Wiman, B. and Wallen, P. (1977) Thromb. Res. 10, 213-222 7 Wiman, B. and Collen, D. (1978) Nature (Lond.) 272. 549- 550 8 Ryan, T.J. (1981) Biochem. Biophys. Res. Commun. 98. 1108-1114 9 Fenton, J.W. II, Fasco, M.J., Stackrow, A.B., Aronson, D.L., Young, A.M. and Finlayson, J.S. (1977) J. Biol. Chem. 252, 3587-3598 10 Chase, T., Jr. and Shaw, E. (1969) Biochemistry 8, 2212-2224 11 Deutch, D.G. and Mertz, E.T. (1970) Science 170, 1095-1096 12 Laemmli, U.K. (1970) Nature 227, 680-685 13 Gros, C. and Labouesse, B. (1969) Eur. J. Biochem. 7, 463-470 14 Weiner, A.M., Platt, T. and Weber, K. (1972) J. Biol. Chem. 247, 3242-3251 15 Johnson, A.J., Kline, D.L. and Alkjaersig, N. (1969) Thromb. Diath. Haemorrh. 21,259-272 16 Liu, T.H. and Mertz. E.T. (1971) Can. J. Biochem. 49, 1055-1061

194 17 Robbins, K.C. and Summaria, k. (1976) Methods Enzvmol+ 45, 257--273 18 Robbins, K.C. and Summaria. L. (1970) Methods [:,nzymol. 19, 184-199 19 Carlin, G. and Saldeen, T. (1978) Thromb. Res. 12. 681-686 20 Moore, S. (1968) J. Biol. Chem. 243. 6281-6283 21 Beardon. J.C.. Jr. (1978) Biochim Biophys. Acta 533. 525-529 22 Philo. R.D. and Gaffney, P.J. (1981) Thromb. Res. 21. 81-88 23 Gilboa. N.. Erdman, J. and Urizar, R.E. (1982) Kidney Int. 22, 80-83

24 Paulson. A.S. and Nicklin. E.H. <19~;3) A p p l Star 32 32-50 25 Markus, G.. Evers, J i . and Hobika. G.H. (197g) J. Biol Chem. 253, 733-739 26 Markus. G., Camiolo. S.M.. Sottrup-Jensen. L. and Mag nusson. S. (1981) in Progress in Fibrinolvsis (Davidson J.F., Nelsson, I.M. and Astedt, B.. eds.), Vol. 5. pp. 125-128 Churchill Livingstone. Edinburgh 27 Lerch, P.G., Rickli, E.E.. LergJer. W. and Gillessen. D (1980) Eur. J. Biochem. 107, 7-13