Citation
Isolation and characterization of a group A streptococcal receptor for human plasmin

Material Information

Title:
Isolation and characterization of a group A streptococcal receptor for human plasmin
Creator:
Broder, Christopher C., 1961-
Publication Date:
Language:
English
Physical Description:
xiii, 174 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Antibodies ( jstor )
Bacteria ( jstor )
Cellulose nitrate ( jstor )
Enzymes ( jstor )
Gels ( jstor )
Molecular chains ( jstor )
Molecules ( jstor )
Plasminogen activators ( jstor )
Receptors ( jstor )
Streptococcus ( jstor )
Dissertations, Academic ( mesh )
Dissertations, Academic -- Immunology and Medical Microbiology -- UF ( mesh )
Immunology and Medical Microbiology thesis Ph.D ( mesh )
Plasmin ( mesh )
Streptococcus pyogenes ( mesh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 162-173).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Christopher C. Broder.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Christopher C. Broder. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
25072858 ( OCLC )
024136819 ( ALEPH )

Downloads

This item has the following downloads:


Full Text













ISOLATION AND CHARACTERIZATION OF A
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN



















By

CHRISTOPHER C. BRODER



















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1989






























This dissertation is dedicated
to the memory of my father
Thomas J. Broder
















ACKNOWLEDGMENTS



I wish to express my sincere thanks to Dr. Michael D.P. Boyle for

giving me the opportunity to work in his laboratory, for his support and guidance, and especially for his patience. It has been a pleasure to work with Mike.

I also wish to give a special thanks to Dr. Richard Lottenberg for all of his help, guidance, and friendship.

I would like to thank the other members of my committee, Drs. R.W. Moyer and J.W. Shands, for their helpful suggestions throughout this study.

I would also like to offer a special thanks to Dr. Kenneth H.

Johnston, my outside examiner, for sending me the monoclonal antibodies, and the solid-phase plasminogen activator assay mentioned in this study and for taking the time to review and discuss my work.

I would also like to express my appreciation to all the people with whom I have worked for the past four years, especially Jeannine Brady, Greg VonMerring, Tim Broeseker, and Lucy DesJardin.

I also offer most special thanks to my parents, Jeanne C. and

Thomas J. Broder, for their never-ending love and support throughout all my endeavors. I also thank all my family, especially my brother Michael.

Finally, I offer my sincere thanks to my wife, Colleen, for all of her unselfish support, which has been essential for me in pursuing my goals.

iii

















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS.................................................... iii

LIST OF TABLES..................................................... vi

LIST OF FIGURES.................................................... vii

KEY TO ABBREVIATIONS............................................... x

ABSTRACT ........................................................... xii

CHAPTER

I. INTRODUCTION............................................. 1

Introduction...................................... 1
Plasmin(ogen) Structure ........................... 1
Plasminogen Activation ............................ 5
Plasmin Regulation................................ 6
Summary and Specific Aims ......................... 7

II. IDENTIFICATION AND CHARACTERIZATION OF A GROUP A
STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN .................. 9

Introduction...................................... 9
Materials and Methods .............................. 9
Results ........................................... 18
Discussion........................................ 43

III. LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN
BINDING TO ITS SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR... 49

Introduction...................................... 49
Materials and Methods .............................. 50
Results ........................................... 58
Discussion ........................................ 79

IV. ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN......... 85

Introduction...................................... 85
Materials and Methods ............................. 86
Results........................................... 99
Discussion........................................ 120




iv










Page

V. COMPARISON OF THE GROUP A STREPTOCOCCAL RECEPTOR FOR
HUMAN PLASMIN WITH STREPTOKINASE ......................... 122

Introduction...................................... 122
Materials and Methods ............................. 123
Results........................................... 130
Discussion........................................ 146

VI. SUMMARY AND CONCLUSIONS .................................. 153

REFERENCES ......................................................... 162

BIOGRAPHICAL SKETCH................................................ 174


















































V
















LIST OF TABLES

Table Page


2-1. Binding of radiolabeled proteins to various nephritogenic and non-nephritogenic group A
streptococci..................................... 19

2-2. Ability of bacterial bound plasmin to solubilize a
fibrin clot....................................... 29

3-1. Summary of inhibition experiments ................... 73

3-2. Measurement of plasmin(ogen) associated with
bacterial pellets................................. 74

5-1. Fluid-phase plasminogen activator activity assay.... 132




































vi
















LIST OF FIGURES

Figure Page

1-1. Schematic representation of the human Glu-plasminogen
molecule ............................................ 4

2-1. Binding of plasmin to bacteria: comparison of the
kinetics of generation of plasmin and its ability
to bind to the group A streptococcus 64/14.......... 22

2-2. Effect of inhibiting the active site of plasmin on
its ability to bind to the group A streptococcal
strain 64/14........................................ 24

2-3. Regulation of bacterial bound enzyme activity by a
variety of different serine protease inhibitors..... 27

2-4. Binding of 1251-plasmin or 1251-plasminogen to the
group A streptococcal strain 64/14 as a function of
pH ......... .. .. .. .......... ...... .. .. ... ... ... ... .. 31

2-5. Binding of 1251-plasmin or 1251-plasminogen to the
group A streptococcal strain 64/14 as a function of
ionic strength...................................... 34

2-6. Specific binding of 1251-plasmin to 107 group A
streptococci, strain 64/14, following a 15 minute
incubation at 37*C in VBS-gel at pH 7.4 ............. 37

2-7. Inhibition of binding of 1251-plasmin to the group A
streptococcal strain 64/14 in VBS-gel containing
various concentrations of epsilon-aminocaproic acid,
lysine, and arginine ................................ 40

2-8. Elution of 1251-plasmin from group A streptococcal
strain 64/14 in VBS-gel containing various concentrations of epsilon-aminocaproic acid, lysine,
and arginine........................................ 42

3-1. SDS-UREA-PAGE analysis of isolated plasmin(ogen)
fragments........................................... 60

3-2. Inhibition of, PPACK reacted, 1251-Lys-plasmin
binding to group A streptococcal plasmin receptor... 63

3-3. Binding of 1251 labeled Glu- and Lys-plasmin(ogens).... 67



vii









Figure Page

3-4. Inhibition of, PPACK reacted, 1251-Lys-plasmin
binding to group A plasmin receptor ................. 69

3-5. Binding of Lys-plasmin(s), derived from Gluplasminogen and Lys-plasminogen, to the group A
streptococcal strain 64/14 as measured by residual
activity in the bacterial free supernatant........... 72

3-6. Binding of Glu- and Lys-plasminogen to the group A
streptococcal strain 64/14 as measured by residual
activatable zymogen in the bacterial free
supernatant......................................... 76

3-7. Characterization of 1251-plasmin(ogen) species eluted
from bacteria....................................... 78

4-1. Dot-blot analysis of solubilized plasmin binding
activities.......................................... 102

4-2. SDS-PAGE and Western blot analysis of mutanolysin
extracted 64/14 bacterial plasmin binding activity.. 104

4-3. Solid-phase plasminogen activation assay................ 109

4-4. Representative profile of an affinity purification
of strain 64/14 mutanolysin extracted plasmin
binding activity.................................... 112

4-5. Analysis of affinity purified plasmin binding material
from the strain 64/14 mutanolysin extract ........... 114

4-6. SDS-PAGE and Western blot analysis of mutanolysin
extracted, affinity purified plasmin binding
activity ............................................ 117

4-7. SDS-PAGE and Western blot analysis of plasmin
receptor protein with a polyclonal rabbit antibody.. 119

5-1. Functional identification and distinction of
streptokinase proteins and plasmin binding receptor
protein ............................................. 134

5-2. Comparison of binding reactivities of streptokinase
proteins and plasmin binding receptor protein with
251-plasmin heavy chain and 125I-plasmin light
chain ............................................... 138

5-3. Analysis of the antigenic relationship of the 64/14
plasmin receptor and streptokinase proteins ......... 141





viii









Figure Page

5-4. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and
group C streptokinase with mouse polyclonal
anti-plasmin receptor antibodies..................... 145

5-5. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and group C streptokinase with mouse anti-group C
streptokinase monoclonal antibodies ................. 148


















































ix
















KEY TO ABBREVIATIONS



a2-AP/a2-PI alpha-2-antiplasmin ATCC American typed culture collection BSA bovine serum albumin cpm counts per minute DNase deoxyribonuclease EACA epsilon aminocaproic acid EDTA ethylenediaminetetraacetic acid ELISA enzyme-linked-immunosorbant-assay FPLC fast performance liquid chromatography g gravity Glu-plasminogen native human plasminogen with NH2-terminal glutamic acid

HC heavy chain of plasmin HRGP histidine-rich glycoprotein IgG immunoglobulin class G KD kilodalton KD dissociation constant KIU kallikrein inhibitor unit LBS lysine-binding site LC light chain of plasmin Lys-plasminogen proteolytically modified form of Glu-plasminogen with NH2-terminal lysine

M molar



x









Mini-PLA low molecular weight plasmin Mini-PLG low molecular weight plasminogen mM millimolar JM micromolar Am micrometer MOPS (3-[N-Morpholino]propanesulfonic acid) Mr relative molecular weight NIH u National Institute of Health unit nm nanometer PAGE polyacrylamide gel electrophoresis PBS phosphate buffered saline PEG polyethylene glycol PLA plasmin PLG plasminogen p pico PMSF phenylmethylsulfonylfloride pNpGB p-nitrophenyl p-guanidinobenzoate HC1 PPACK Phe-Pro-Arg-chloromethylketone RNase ribonuclease S-2251 H-D-Val-Leu-Lys-paranitroanilide SDS sodium dodecylsulfate SK streptokinase tPA tissue-type plasminogen activator Tris (Tris[hydroxymethyl]aminomethane UK urokinase VBS-gel Veronal-buffered saline plus gelatine VPLCK D-Val-Phe-Lys-chloromethyl ketone



xi
















Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


ISOLATION AND CHARACTERIZATION OF A GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN By

CHRISTOPHER C. BRODER

May 1989

Chairman: Michael D.P. Boyle
Major Department: Immunology and Medical Microbiology

The expression of a specific receptor for the key human fibrinolytic enzyme plasmin on the surface of the group A streptococcal strain 64/14 is reported. The receptor was specific for plasmin, and demonstrated no significant reactivity with the zymogen form of the molecule (Gluplasminogen). Bacterial bound plasmin retained its enzymatic activity, and could not be inhibited by the physiological regulator (a2antiplasmin). The receptor demonstrated a high affinity for plasmin (KD=1.0x10-10 M), and binding was maximal at physiologic pH and ionic strength. Furthermore, the receptor-ligand complex was reversibly inhibitable by E-aminocaproic acid and L-lysine. The binding of plasmin to this group A streptococcus was found to occur primarily through interactions with the heavy chain of the plasmin molecule, and was dependent on a specific conformation of the ligand. A functionally active plasmin receptor was obtained from strain 64/14 bacteria by an enzymatic extraction with mutanolysin. Plasmin binding activity was expressed predominantly in a protein having an Mr of approximately 41,000


xii










daltons. The plasmin receptor demonstrated no plasminogen activator activity. A functionally active plasmin receptor protein was purified by affinity chromatography using immobilized plasmin and specific elution with L-lysine or EACA. The strain 64/14 plasmin receptor was compared with secreted streptokinase proteins from five streptococcal isolates including strain 64/14. Only the plasmin receptor-plasmin complex was found to be sensitive to L-lysine or EACA. Polyclonal rabbit and mouse anti-plasmin receptor antibodies were prepared, as well as polyclonal anti-group C streptokinase antibodies. Using these antibodies as well as a bank of mouse monoclonal anti-group C streptokinase antibodies, the 41,000 dalton plasmin receptor protein from strain 64/14 was shown to be antigenically unrelated to either group A or C streptokinase. Thus the strain 64/14 streptococcal receptor for human plasmin is physicochemically, functionally, and antigenically distinct from streptokinase. The importance of a specific receptor for human plasmin on pathogenic streptococci is unclear; however, it may provide a mechanism for the capture of the potent enzyme plasmin which may confer additional invasive properties to the bacteria.























xiii
















CHAPTER ONE
INTRODUCTION



Introduction


Plasmin, a serine protease, is the key component of the mammalian fibrinolytic enzyme system. The main physiological role of the fibrinolytic system is the dissolution of fibrin clots formed in blood vessels. Milstone, in 1941, determined that the lysis of fibrin, by the streptococcal substance described by Tillett and Garner in 1933, was dependent on a 'lytic factor' in human serum. This was later followed by the discovery that the lytic factor was an enzyme precursor, in human plasma, that was converted to an active enzyme by a component in the streptococcal fluid (Christensen, 1945; Kaplan, 1944). This precursor was called plasminogen, the enzyme plasmin, and the streptococcal factor streptokinase (Christensen and Macleod, 1945). The zymogen precursor plasminogen, molecular weight approximately 92,000 daltons, is a singlechain glycosylated protein containing 790 amino acids in known sequence and containing 24 disulfide bridges (Brogden et al., 1973; Sottrup-Jensen et al., 1978; Wiman, 1973, 1977).


Plasmin(ogen) Structure


Native plasminogen (Glu-plasminogen) has glutamic acid as the NH2terminal residue, but is readily converted by the action of plasmin to modified forms of plasminogen which are commonly called Lys-plasminogen.



1







2

These modified forms of plasminogen have lysine, valine, or methionine as their N-terminal amino acid (Wall6n and Wiman, 1970, 1972). These modifications occur by the hydrolysis of the Arg67-Met68, Lys76-Lys77, or Lys77-Val78 peptide bonds. The generation of plasmin from plasminogen occurs through the specific cleavage of a single Arg-Val bond corresponding to the Arg560-Val561 positions (Robbins et al., 1967). This cleavage generates a two chain active plasmin molecule consisting of a heavy chain and light chain held together by disulfide linkages (Groskopf et al., 1969; Wiman, 1973) (see Figure 1-1). The light chain of plasmin has a molecular weight of approximately 25,000 daltons (Robbins and Summaria, 1970; Wiman, 1977) and contains the serine active site. The heavy chain of plasmin has a molecular weight of approximately 63,000 daltons (Robbins and Summaria, 1970), and amino acid sequencing revealed a structure containing 5 homologous triple loop structures known as kringles (Sottrup-Jensen et al., 1978).

Several specific compounds such as lysine, E-aminocaproic acid

(EACA), trans-4-aminomethycyclohexanecarboxylic acid (transexamic acid), and C-terminal lysine peptides bind to certain sites on the plasmin(ogen) molecule. These specific sites are the characteristic 'lysine-binding sites' distinct from the catalytic site (Thorsen, 1975). These compounds affect the properties of plasminogen and plasmin, and play an important role in determining this zymogen-enzyme system's physiological specificity. Chapter Three will go into more detail on the nature of plasmin(ogen)'s lysine binding sites. Affinity chromatography of defined fragments of plasminogen on lysine Sepharose has demonstrated that these 'lysine-binding sites' are located in the portion of the plasmin molecule which becomes the heavy chain upon activation. The

























































Figure 1-1. Schematic representation of the human Glu-plasminogen molecule. H-Glu: N-terminal glutamic acid residue; HO-Asn: C-terminal asparagine residue; Ser: serine residue of the enzyme active site; t-PA and UK: plasminogen activation cleavage site on Glu-plasminogen; LBS-I and LBS-II: lysine-binding sites; 1-5: kringle domains numbers 1 through
5. E: elastase cleavage sites. Adapted from Collen (1980).






















LBS II


3 4 5
E E
t-PA H- snm UK


PLA

H-Glu LS Ser







5

native Glu-plasminogen molecule contains one high affinity lysinebinding site (Kd = 9 pM) and five weaker lysine-binding sites (Kd 5 mM) (Markus et al., 1978a, 1978b). Lys-plasminogen contains one high affinity, one intermediate affinity, and four lower affinity lysine binding sites. The exact number of sites on the plasmin molecule has not been reported. Two of these lysine-binding sites have been mapped to specific regions in the plasminogen molecule. Studies involving equilibrium dialysis experiments on the binding of EACA to isolated fragments of the plasmin(ogen) molecule (see Chapter Three for a complete discussion of these fragments) revealed that the high affinity lysine-binding site was located in the kringle 1 structure, and kringle 4 contained one of the lower affinity sites (Lerch et al., 1980).


Plasminogen Activation


The generation of plasmin from plasminogen is accomplished by plasminogen activators. Three plasminogen activators have been extensively studied. Urokinase (UK) and tissue plasminogen activator (tPA) are proteolytic enzymes (for review see Astrup, 1978); the third, streptokinase (SK), possesses no inherent proteolytic activity. Tissue plasminogen activator, by virtue of its serine protease activity can directly activate plasminogen. This enzyme is present in various tissues and can also bind to fibrin. Urokinase, also a serine protease, is a glycoprotein which has no fibrin binding capacity. Urokinase can also activate plasminogen directly cleaving the Arg560-Val561 peptide bond.

Streptokinase is a unique plasminogen activator which is produced by certain streptococci. The only apparent function of streptokinase, since its initial description by Tillett and Garner (1933), is its ability to







6

activate plasminogen. Unlike the other plasminogen activators, streptokinase has no enzymatic activity. The activation mechanism lies in its ability to form a specific 1:1 stoichiometric complex with plasminogen, as well as with plasmin, which leads to the generation of an active enzyme moiety, presumably through conformational changes in the plasminogen molecule without the cleavage event at the Arg560-Val561 peptide bond (Markus et al., 1976), that can in turn act as a plasminogen activator for plasminogen molecules. This is a function neither of the two proteins possesses alone.


Plasmin Regulation


Once generated, plasmin's activity is also carefully regulated under normal physiological conditions. This regulation is accomplished by a specific inhibitor of plasmin known as a2-antiplasmin (a2-AP) (Aoki et al., 1977; Collen, 1976; Collen et al., 1975). Alpha2-antiplasmin is a single chain glycoprotein with a molecular weight of approximately 70,000 daltons. a2-antiplasmin forms a 1:1 stoichiometric complex very rapidly (estimated rate constant of kl=3x107M-1S-1) (Wiman and Collen, 1978) and neutralizes plasmin's activity through a covalent association with the serine residue in the active site of plasmin. A physiological role of a2-AP as an inhibitor of other proteases other than plasmin appears negligible (Edy and Collen, 1979; Ohlsson and Collen, 1977).

Workers pioneering the techniques of tissue culturing noted that explants of cancer tissue consistently caused proteolytic degradation, liquefying the plasma clots on which they were grown (Carrel and Burrows, 1911; Lambert and Hanes, 1911). Since those early studies the plasminogen-plasmin system, in addition to its role in fibrinolysis, has







7

been implicated in a variety of normal and abnormal processes which involve the destruction or alteration of the extracellular environment, such as tumor cell growth and invasiveness (for review of this extensive literature see Dano et al., 1985), tissue remodeling, embryogenesis (Beers et al., 1975), ovulation (Strickland and Beers, 1976), and trophoblast implantation (Strickland et al., 1976). In fact, plasmin exhibits broad substrate specificity and in addition to fibrin can hydrolyze components of connective tissue and basement membranes such as laminin, proteoglycans, fibronectin, thrombospondin, and type-V collagen, as well as proteolytically activating other proteases (for review see Knudsen et al., 1986) and several plasma proteins (Marder et al., 1982).


Summary and Specific Aims


In the process of examining human serum from patients for antibody reactivity directed against the streptococcal plasminogen activator streptokinase, from patients who received thrombolytic therapy by streptokinase administration, an interesting observation was made by Dr. M.D.P. Boyle and Dr. R. Lottenberg. In experiments which involved immune precipitations using heat killed, Fc-receptor expressing, group C streptococci it was found that control tests involving incubations of the radiolabeled tracer (125I-plasmin) and bacteria revealed an association of plasmin to bacteria in the absence of any added antibody. Testing other groups of streptococci showed that the group A streptococci displayed the highest level of plasmin binding activity, while demonstrating little binding activity for preparations of 1251 plasminogen.







8

The ability of certain group A streptococci to produce a plasminogen activator (e.g. streptokinase) and also to express a receptor for the activation product plasmin, may contribute to the invasive properties of these bacteria. This study has been designed to characterize this plasmin receptor phenomenon more completely in order to increase understanding of any potential role in bacterial pathogenesis.

The specific aims of the study are to

1. Identify and characterize a group A streptococcal receptor for human plasmin (Chapter Two).

2. Map the binding site on human plasmin recognized by the

bacterial plasmin receptor (Chapter Three).

3. Isolate and purify a functionally active group A

streptococcal plasmin receptor (Chapter Four).

4. Compare the group A streptococcal receptor for human

plasmin with streptokinase, (Chapter Five), with

respect to

a. Plasminogen activator activity.

b. Plasmin(ogen) binding domain specificity.

c. Antigenic relatedness.
















CHAPTER TWO
IDENTIFICATION AND CHARACTERIZATION OF A GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN


Introduction


Many group A streptococcal infections are characterized by tissue invasion. A variety of characteristics of these microorganisms contribute to their ability to break down natural tissue barriers and to avoid elimination by the host immune response. Certain surface proteins or secreted products associated with streptococci have been identified that enable these organisms to elude the immune system (Sparling, 1983), and proteins and toxins produced by these bacteria are known to contribute to tissue damage (Ginsburg, 1972; Johnston, 1984; Sparling, 1983). In addition a variety of receptors for host proteins have been described on streptococci. These include receptors for components of the immune system such as Clq (Yarnall et al., 1986), IgG (Kronvall, 1973), and IgA (Russell-Jones et al., 1984), the serum protein fibrinogen (Kronvall et al., 1979), and the stromal structural proteins laminin (Switalski et al., 1984) and fibronectin (Myhre and Krusela, 1983). In this chapter I describe the presence of a specific receptor for human plasmin on certain group A streptococci.


Materials and Methods


Human Plasminogen

Human plasminogen was prepared from human plasma by chromatography on lysine Sepharose and molecular sieving chromatography on Sephadex G9







10

100 (Lottenberg et al., 1985). Plasminogen was quantified by measuring absorbance using a AJ8nm value of 17.0 (Nilsson et al., 1982). Enzymes, Inhibitors and other Reagents

The enzymes urokinase and trypsin were obtained from the Sigma Chemical Company, St. Louis, Mo.; Aprotinin was obtained as Trasylol from Mobay Pharmaceuticals, New York, New York. Phe-pro-arg chloromethylketone (PPACK) was obtained from Cal-Biochem (San Diego, Ca.) P-nitrophenyl, p-guanidobenzoate HC1 (pNpGB) was obtained from Sigma Chemical Co, St. Louis, Mo.; human a2-antiplasmin (a2-AP) was obtained from American Diagnostica Inc., Greenwich, Connecticut. H-D-Val-leu-lysparanitroanilide (S-2251) was obtained from Helena Chemical Co., Beaumont, Texas.

Radioiodination of Proteins

Human plasminogen, urokinase, and trypsin were iodinated by a mild lactoperoxidase method using Enzymobeads (Bio-rad Laboratories Richmond, Calif.) as described by Reis et al., (1983). The labeled proteins were separated from free iodine by passage over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1% gelatin (VBS-gel). The labeled proteins were stored in aliquots containing 0.02% sodium azide at -20*C. Labeled aliquots were used once and discarded. Generation of Plasmin

Plasmin was generated from either radiolabeled or unlabeled

plasminogen by reaction with urokinase. Three .l of urokinase (Sigma 20 u/ml) was added to a 400 p1 solution of 1.0 pM plasminogen containing

0.04 M lysine. The mixture was incubated at 37*C for 45 minutes unless stated otherwise. The efficiency of plasmin generation was followed by









measuring the conversion of the single chain plasminogen molecule (Mr=90,000) into heavy chains (Mr-60,000) and light chains (Mr-25,000) as determined by the migration of radiolabeled proteins, following denaturation and reduction, on 10% SDS-polyacrylamide gels. The migration of labeled proteins was determined by autoradiographic exposure of dried gels to Kodak XAR 5 film with intensifying screens at 70*C for 20 hours.

Bacteria

The group A f-hemolytic streptococcal strain 64 had been previously subjected to mouse passage as described by Reis et al., (1984). The parent strain (64/P), as well as strains isolated after three (64/3) and fourteen (64/14) mouse passages, were grown in either Todd-Hewitt broth (DIFCO, Detroit, Mich.) or chemically defined media (Van De Rijn and Kessler, 1980) overnight at 37C as stationary cultures (Yarnall and Boyle, 1986b). The bacteria were harvested by centrifugation and resuspended in phosphate-buffered saline (PBS), pH 7.4, containing 0.05% Tween-20 and 0.02% sodium azide. The bacteria were heat killed at 80* C for 10 minutes, a treatment that did not alter their plasmin binding potential, but eliminated the production of soluble plasminogen activators which would interfere with these studies. The suspension was centrifuged, the pellet washed twice with PBS and then resuspended at 10% wet weight/volume in PBS containing 0.05% Tween-20 and 0.02% sodium azide. Samples were stored at -20*C. The concentration of a bacterial suspension was determined by counting bacterial chains in a Neubauer hemacytometer (Fisher Scientific, Orlando, FL). Determination of Binding of Radiolabeled Proteins to Bacteria

The light scatter at 550 nm was determined to standardize the

concentration of organisms used in subsequent tests. A light scatter







12

reading of 0.3 corresponded to approximately 2 x 109 organisms/ml (Yarnall and Boyle, 1986b). A standardized number of bacteria (approximately 109 organisms) were incubated with labeled proteins (approximately 30,000 cpm/tube) in a total volume of 400 pl of VBS-gel for 1 hour at 370C. The bacteria were pelleted by centrifugation at 1000 x g for 10 minutes and washed twice with 2.0 ml VBS-gel. The radioactivity associated with the bacteria was determined in a Beckman 5500 autogamma counter. All estimates were carried out in duplicate. Fibrin Plate Assay

Fibrin plates were prepared using 5 cm diameter disposable petri dishes. Ten ml of 0.1% fibrinogen in PBS were clotted with 0.2 ml of bovine thrombin (10 NIH u/ml) in 0.5 M CaC12. Twenty pmoles of plasmin were bound to 100 Ml of a 10% w/v solution of the group A streptococcus, 64/14, in a total volume of 400 pl of VBS-gel. The mixture was incubated at 37*C for 45 minutes. A parallel series of samples containing bacteria with no added plasmin served as the negative control. Fifty pl of a suspension of bacteria or bacteria plus plasmin were placed either directly onto a fibrin plate or onto a 0.22 pm Millipore filter placed between the bacteria and the fibrin plate. The plates were incubated for 20 hours at 370C and the degree of hydrolysis was scored by measuring the area of the zone of clearing from the underside of the plate. In each experiment a control of free plasmin was included and each estimate was carried out in duplicate.

Plasmin-Inhibitor Complex Generation

Plasmin was generated from plasminogen as described above. Three 130 p1 aliquots of the labeled enzyme were placed into separate microtubes (approximately 2 x 106 cpm/tube), and incubated with a 200-







13

fold molar excess of either PPACK, aprotinin, or pNpGB for 10 minutes at room temperature. The volume of each sample was increased by the addition of 200 pl of VBS-gel, and each was applied to a separate G-25 column (PD 10 Pharmacia Fine Chemical) to remove free inhibitor. Five hundred p1 fractions were collected and counted in an autogamma counter to localize the modified 1251 plasmin. Fractions containing the labeled protein-inhibitor complexes were pooled. Aliquots of the three labeled complexes were mixed with a 10-fold molar excess of a2-AP in a final reaction volume of 400 p1 for 10 minutes at room temperature. Plasmin and plasminogen were included as controls. The volumes of each solution were adjusted to 1.1 ml with VBS-gel and then used in a direct binding assay to group A streptococci. Each of the plasmin-inhibitor samples that has been incubated with excess a2-AP was monitored on non-reducing SDS-polyacrylamide gels as described by Weber and Osborn (1969) for the formation of a high molecular weight complex. Determination of Plasmin Activity While Bound to Bacteria

To five microtubes, each containing 100 p1 of a 10% w/v solution of the group A streptococcus, 64/14, in a total volume of 400 p1 VBS-gel, 10 nM plasmin was added and allowed to bind for 40 minutes at 37*C. Five other tubes containing plasmin but no bacteria and one tube containing bacteria alone were prepared as controls. The bacteria were pelleted and washed twice with 1.0 ml VBS-gel and resuspended in 400 p1 VBS-gel containing a 10-fold molar excess of either pNpGB, PPACK, aprotinin, a2AP or buffer alone. All samples were then incubated for 15 minutes at room temperature. The samples containing bacteria were pelleted by centrifugation, washed with 1.0 ml of VBS-gel and resuspended with vigorous vortexing in 400 p1 VBS-gel.







14

To each tube, 20 p1 of an 8.0 mM solution of the chromogenic

substrate, H-D-val-leu-lys-paranitroanilide, was added,to yield a final concentration of substrate in the reaction mixture of 400 pM. The tubes were mixed by vortexing and incubated at 37*C for 25 minutes. At that time the enzyme reaction was quenched by the addition of 400 Cl of 10% (v/v) acetic acid, the samples were then centrifuged for 5 minutes at 10,000 x g and the optical densities of the solutions at 405 nm were determined. The release of paranitroaniline from the synthetic substrate monitored at this wavelength was directly proportional to the enzymatic activity of plasmin. Control samples of substrate alone and substrate plus bacteria were included and each estimate was carried out in duplicate.

Effect of pH on Plasmin Binding to Bacteria

To assess the effect of pH on the bacterium:plasmin(ogen)

interaction, 50 pl of labeled plasminogen or plasmin (approximately 2 x 105 cpm) were added to 1.0 ml of VBS containing 0.05% Tween-20 adjusted to the appropriate pH. After 15 minutes at room temperature, 50 pl of VBS containing approximately 107 bacteria (strain 64/14) were added and the mixture was incubated at 37* C for 15 minutes. The bacterial suspensions were centrifuged at 1,000 x g for 7 minutes to separate bacteria from unbound labeled proteins and the pellets were washed twice with 2.0 ml of VBS at the appropriate pH. The radioactivity associated with the bacterial pellet in duplicate experiments was measured using a Beckman 5500 autogamma counter.

To assess the effect of ionic strength on the bacteriumplasmin(ogen) interaction, similar studies were carried out in solutions containing different concentrations of NaCl with 0.05% Tween-20. The







15

bacterial pellets were washed in the appropriate NaCl concentration to remove unbound labeled proteins.

Effect of Ca++ and Mg++ on Plasmin Binding

Binding of radiolabeled plasmin to group A streptococci strain 64/14 was studied in the following buffers: 1) VBS-gel containing 0.00015 M Ca++ and 0.001 M Mg++, or 2) metal free VBS-gel containing 0.15 M EDTA. In each case 400 Al of buffer were added to 100 Al of VBS-gel containing approximately 107 bacteria and 100 Al of VBS-gel containing 3 x 105 cpm of radiolabeled plasmin. After incubation at 37* C for 15 minutes, the mixtures were centrifuged at 1,000 x g for 7 minutes to separate bacteria from unbound radiolabel, the pellets were washed twice with 2.0 ml of the appropriate buffer, and radioactivity associated with the bacterial pellet in duplicate tubes was measured. Inhibition of Binding of Plasmin by Amino Acids

Labeled plasmin (100 p1 containing approximately 2 x 105 cpm) was added to 200 p1 VBS-gel containing varying concentrations of epsilon-aminocaproic acid (EACA), lysine, or arginine, and incubated at 37* C for 15 minutes. The pH of each solution was 7.0. One hundred p1 of VBS-gel containing 107 bacteria (strain 64/14) were then added and the mixture was incubated at 37*C for 15 minutes. The bacterial suspensions were centrifuged at 1,000 x g for 7 minutes and washed twice with 2.0 ml of VBS-gel containing the same concentration of amino acid present during the incubation period. The percent inhibition of binding was calculated for duplicate samples by comparison with binding in VBS-gel alone.

The ability of EACA, lysine, or arginine to dissociate bound plasmin from the bacteria was examined in the following manner. Labeled plasmin was incubated with 107 bacteria in VBS-gel at 37* C for 15 minutes. The







16

bacteria were pelleted by centrifugation and washed twice with 2.0 ml of VBS-gel. After determining the radioactivity associated with the bacteria, the pellets were resuspended in solutions of VBS-gel containing varying concentrations of amino acid or amino acid analogs (pH 7.0) as described above. The mixtures were incubated at 37* C for 15 minutes and washed twice with VBS-gel containing the appropriate amino acid concentration. The radioactivity associated with the bacteria in duplicate samples was again measured and the percentage dissociated was calculated.

Determination of KD and Receptor Density

Labeled plasmin (25,000 to 250,000 cpm) in 100 pl of VBS-gel was

added to 3 x 106 bacteria in 300 pl of VBS-gel, pH 7.4, and incubated at 37*C for 15 minutes. The bacterial suspensions were centrifuged at 1,000 x g for 10 minutes and washed twice with 2.0 ml of VBS-gel. All determinations were performed in triplicate. Total binding was determined by measuring the radioactivity associated with the bacterial pellet when only labeled plasmin was offered. Non-specific binding was determined by pre-incubating bacteria at 37* C for 15 minutes in VBS-gel, pH 7.4, containing unlabeled plasmin at a 100-fold molar excess of the labeled plasmin. Specific binding was calculated by subtracting non-specific binding from total binding for each amount of labeled plasmin offered. The amount of free labeled plasmin was calculated by subtracting the amount of specifically bound labeled plasmin from the total amount of labeled plasmin offered.

The apparent dissociation constant (KD) was determined by two

methods. A non-linear least squares analysis of the total counts offered vs. the counts bound fit to the simple Michaelis-Menten equation was







17

performed as described by Lottenberg et al., (1985). The concentration of plasmin was determined by converting counts per minute to moles using the known specific activity for the labeled plasminogen. Scatchard analysis (Scatchard, 1949) of these data was also performed as described by Lottenberg et al., (1987). Counts bound vs. counts bound/counts free was plotted and the slope (representing -1/KD) was determined by linear regression. The X-intercept (counts bound) was converted to moles of plasmin. To determine the receptor densities the number of moles of plasmin bound was determined by extrapolating the Scatchard plot and determining the intercept. This represented the maximal binding of plasmin to a known number of bacteria (derived by hemacytometer chamber counts).

Plasmin which had been bound to and eluted from strain 64/14 by treatment with lysine was also examined in similar binding studies. Eluted plasmin was obtained by incubating 2.0 ml of stock 10% wet weight/volume bacterial suspension (strain 64/14) with approximately 20 Ag of labeled plasmin at room temperature for 45 minutes. This suspension was centrifuged at 1,000 x g for 10 minutes and washed once with 10 ml of VBS-gel, and the radioactivity associated with the bacterial pellet was measured. The pellet was then resuspended in VBS-gel containing 20 mM lysine and incubated at room temperature for 30 minutes. The suspension was centrifuged and the supernatant recovered. Approximately 90% of the radioactivity originally associated with the bacterial pellet was dissociated by the lysine treatment. The dissociated plasmin in the supernatant was then subjected to gel filtration on a G-25 column to separate lysine from plasmin. Fractions containing plasmin were collected and stored at -200C.







18

Results


Twenty p hemolytic streptococcal isolates were grown overnight at 37C and tested for their ability to bind radiolabeled plasminogen, plasmin, urokinase, or trypsin as described in the Methods Section. The results (see Table 2-1) showed that all twenty group A isolates bound plasmin but failed to bind significant quantities of plasminogen or any of the other labeled proteins, i.e. less than 10% of the offered label. Furthermore, the expression of plasmin binding ability was shown to be present on bacteria grown in either Todd-Hewitt broth or chemically defined media (data not shown). Plasmin binding was found to be dependent on the concentration of bacteria and was maximal within two minutes at 37*C. Pre-incubation with excess unlabeled plasmin prevented binding of the labeled plasmin. In the absence of unlabeled plasmin, strain 64/14 consistently bound approximately 60% of the radioactive plasmin offered and was used to analyze further the selective plasmin binding activity. In my initial attempts to characterize the differential binding of plasminogen and plasmin to a group A streptococcus I compared the kinetics of generation of plasmin from plasminogen with the ability of labeled protein to bind to the bacteria. Conversion of plasminogen to plasmin occurs when a single arginine-valine bond is split in the zymogen by action of the enzyme urokinase (Groskopf et al., 1969). The zymogen activation reaction can be monitored on SDSpolyacrylamide gels following reduction of disulfide bonds.

The zymogen molecule migrates as a single polypeptide chain with a Mr of approximately 90,000 daltons. The active enzyme plasmin migrates under these conditions as two distinct polypeptide chains (a heavy chain with an Mr of approximately 60,000 daltons and a light chain with an Mr







19

Table 2-1.

Binding of radiolabeled proteins to various nephritogenic
and non-nephritogenic group A streptococci.


STRAIN M-TYPE PLASMINOGEN PLASMIN UROKINASE TRYPSIN


A992* 18 + B923 12 + D897* 12 + B512 4 + B438 18 + B512 NT + A928 55 + 64/14 NT ++ B905 2 + B281 12 + B920 49 ++ B915 49 + A374 12 + B931* 2 + A207 2 + F2030 1 + A547 NT + 64/P NT ++ 648 1 + A995 57 +


* Non-nephritis causing strains
- = Less than 10% bound of total counts offered + = 10% to 30% bound of total counts offered ++ Greater than 30% bound of total counts offered Approximately 3 X 108 bacteria/tube heat killed at 80*C for 10 min. NT = Not typable







20

of approximately 25,000 daltons). The activation reaction can be stopped at any time by addition of a 10-fold molar excess of p-nitrophenyl pguanidinobenzoate (pNpGB). Consequently, it is possible to obtain plasminogen in various stages of activation and compare the ability of the labeled proteins to bind to a group A streptococcus. The results presented in Figure 2-1, panel A demonstrate that the activation of plasminogen to plasmin could be readily monitored. As the conversion of plasminogen to plasmin proceeded, an increase in the binding of labeled protein occurred which correlated with the concentration of plasmin in the reaction mixture (Figure 2-1, panel B).

The conversion of plasminogen to plasmin yields a serine active site that is not expressed in the zymogen. In the next series of experiments the role of the active site in binding of the enzyme to the bacteria was assessed. Plasmin was treated with the active site titrant pNpGB, the small naturally occurring inhibitor aprotinin (Fritz and Wunderer, 1983), the selective histidine modifying agent, phe-pro-arg chloromethyl ketone [PPACK] (Kettner et al., 1980), and the physiological regulator a2 antiplasmin (a2-AP) (Mori and Aoki, 1976). The ability of the various inhibited forms of plasmin to bind to a group A streptococcus was measured. The results presented in Figure 2-2 demonstrate that plasmin treated with pNpGB, aprotinin, or PPACK were all capable of binding to the bacteria in the presence of a2-AP. By contrast unmodified plasmin incubated with the physiological inhibitor, a2-AP, failed to bind. Each of the plasmin-inhibitor samples that had been incubated with excess a2AP was monitored on non-reducing SDS-polyacrylamide gels for the formation of a high molecular weight complex. The high molecular weight band observed in the third lane indicates the formation of a stable


















































Figure 2-1. Binding of plasmin to bacteria: comparison of the kinetics of generation of plasmin and its ability to bind to the group A streptococcus 64/14: Labeled plasminogen was converted to plasmin by treatment with urokinase. The kinetics of generation of plasmin was monitored on SDS-polyacrylamide gels under reducing conditions. The conversion of single chain, high molecular weight, plasminogen (Mr approximately 90,000) into heavy (Mr approximately 60,000) and light chains (Mr approximately 25,000) of plasmin was monitored. At each time point the ability of labeled proteins to bind to the group A streptococcus 64/14 was measured as described in the Methods. The data are presented as the mean the standard deviation of duplicate experiments.










c'lJ









(9

I I J




_I
(10
w w















co o0 0
0 0 0 0 0
I Im I



aNn8 d3
















































Figure 2-2. Effect of inhibiting the active site of plasmin on its ability to bind to the group A streptococcal strain 64/14: The lower panel demonstrates the binding of the group A streptococcus to labeled plasminogen, plasmin, plasmin pretreated with excess a2PI, plasmin treated with excess pNpGB, plasmin treated with excess aprotinin or plasmin treated with excess phe-pro-arg chloromethyl ketone,[PPACK], as described in the Methods. The data are presented as the mean the standard deviation of duplicate experiments. The upper panel demonstrates the analysis of each of the plasmin-inhibitor samples that had been incubated with excess a2PI. Samples were monitored on nonreducing SDS-polyacrylamide gels for the formation of a high molecular weight complex.















% OF TOTAL RADIOACTIVITY BOUND o 0 0 0 0



PLASMINOGEN

PLASMIN

PLASMIN +a2PI

.PLASMIN pNpGB + a2 PI PLASMIN APROTININ + a2 P L I :ZZ I:I:I PLASMIN PPACK + a2 PI







25

complex of plasmin with its physiological inhibitor a2-AP (Figure 2-2, upper panel). Pretreatment of plasmin with aprotinin, pNpGB, or PPACK inhibited the ability of the enzyme to form the covalent linkage inhibition reaction with a2-AP demonstrating that under the experimental conditions used the plasmin active site was modified.

The next series of experiments were designed to determine whether bacterial bound plasmin was capable of retaining its enzymatic activity. Radiolabeled plasmin was generated and incubated with a suspension of group A streptococci for 40 minutes at 37*C. The bacteria with the associated plasmin were recovered by centrifugation, washed twice with buffer, and then tested for their ability to cleave the chromogenic synthetic substrate H-D-val-leu-lys-paranitroanilide (as described in the Methods). In these experiments a control of bacteria alone failed to hydrolyze the chromogenic substrate, while bacteria pre-incubated with plasmin were found to cleave the substrate efficiently. The ability of bacterial bound plasmin to be affected by a variety of different inhibitors was tested. The results in Figure 2-3 demonstrate that addition of pNpGB, PPACK, or aprotinin to the bacterial bound enzyme was capable of inhibiting its enzyme activity for the synthetic substrate. By contrast, addition of a2-AP failed to reduce the enzyme activity (Figure 2-3). All inhibitors were used in excess of that required to totally inhibit an equivalent concentration of plasmin in the fluid phase. Since a2-AP failed to regulate the bacterial bound enzyme, one might predict that the large molecule fibrin, the natural substrate of plasmin, would also be prevented from occupying the substrate pocket in the active site. To test this prediction, bacteria with plasmin bound to their surface were placed on a fibrin plate and their ability to



















































Figure 2-3. Regulation of bacterial bound enzyme activity by a variety of different serine protease inhibitors: Bacterial pellets were preincubated with plasmin, washed and resuspended in buffer containing excess pNpGB, PPACK, aprotinin, a2-AP or buffer alone for 15 minutes at room temperature. Following incubation with the inhibitor the bacteria were pelleted and washed. Enzyme activity was then measured by the ability of the samples to hydrolyze the chromogenic substrate HD-Valleu-gly-paranitroanilide as described in the Methods. The data are presented as the mean the standard deviation of duplicate experiments. The hydrolysis by bacterial bound plasmin in the absence of any inhibitor represents 100% activity.











CN






-J

J Z



-J
z

zZ p



0 0











< Z
wz









FOL
Ilist
ooooo...................
.. .. . Luz








28

mediate dissolution of the fibrin clot was measured. The results presented in Table 2-2 demonstrate that the bacterial bound plasmin still retained its ability to cleave fibrin. These effects could not be accounted for by dissociation of plasmin from the bacteria, since clot lysis did not occur when the microbe-plasmin complex was separated from the clot by a 0.22 pm Millipore filter (Table 2-2). Under these experimental conditions, unbound plasmin was capable of passing through the filter and causing fibrin degradation.

The following series of experiments, which characterize further the interaction of human plasmin and this group A streptococci bacteria were performed by Dr. Tim A. Broeseker, a Fellow in the department of medicine, division of hematology at the University of Florida.

In his initial experiments the binding of labeled plasmin or

plasminogen to the group A streptococcal strain 64/14 as a function of pH was tested. Labeled proteins were pre-equilibrated in VBS-gel buffers of differing pH's before the addition of bacteria. After an incubation period of 15 minutes at 37*C, the radioactivity associated with the bacteria was measured by pelleting the micro-organisms and washing free the unbound label with buffer of the appropriate pH as described in the Methods. Maximal binding of plasmin to the bacteria was observed between pH 5 and 8 with approximately 60% of counts offered being bound by the group A streptococcus 64/14, (Figure 2-4). In contrast, addition of labeled plasminogen to the bacteria over the entire pH range tested (pH 5-9) resulted in direct binding of less than 10% of offered counts. (Figure 2-4). Similar studies were carried out to determine the effect of ionic strength on binding of radiolabeled plasmin and plasminogen to the group A streptococcal strain 64/14. Labeled proteins were pre-







29





Table 2-2.

Ability of bacterial bound plasmin to solubilize a fibrin clot.




Hydrolysis of Fibrina Sample
Directb Indirectc Bacteria Alone

Bacterial Bound Plasmin ++

Plasmin Alone +++ +++













aUnder the experimental conditions described in the Methods section, a (+++) reaction represented a zone of clearing with a diameter of 1.0-1.5 cm, a (++) reaction represented a zone of clearing of 0.5-1.0 cm, a (+) reaction represented a zone of clearing from 0-0.5 cm, and (-) represents no clearing. bSample placed directly onto a fibrin plate. CSample placed onto a filter placed between the bacteria and the fibrin plate.























































Figure 2-4. Binding of 1251-plasmin or 1251-plasminogen to the group A streptococcal strain 64/14 as a function of pH: The data are presented as the mean the standard deviation. Measurements of duplicate experiments were performed and are expressed as the percent of total counts offered (20,000 cpm) which were associated with the bacterial pellet. (O----O) 1251-plasmin; (o-----O) 1251-plasminogen.





31







70 60 ( 50
z
0 40 m 30

20 10

5 6 7 8 9
pH







32

equilibrated in NaC1 solutions of varying ionic strength before the addition of bacteria. Following an incubation period of 15 minutes at 37*C the bacteria were washed with solutions containing the appropriate concentration of NaC1 and the number of counts associated with the bacteria determined. The results in Figure 2-5 demonstrate that plasmin binding was dependent on ionic strength and that optimal binding occurred between 0.1 and 0.4 M NaC1. In this range of salt concentrations, less than 10% of plasminogen bound to bacteria. As the ionic strength was lowered below 0.075 M NaCl, significant binding of plasminogen to the bacteria was observed.

Binding of labeled plasmin to the group A streptococcal strain 64/14 was examined in the presence and absence of divalent cations to determine if these metal ions were important for plasmin binding. Binding studies were carried out in VBS-gel at pH 7.4 containing 0.00015 M Ca++ and 0.001 M Mg++ or in metal free VBS-EDTA-gel at pH 7.4. After incubation at 37*C for 15 minutes, the bacteria were washed twice with the appropriate buffer and radioactivity associated with the bacterial pellets was measured. The amount of plasmin bound by the bacteria was the same in the presence or absence of divalent cations. (data not shown).

After identification of the optimal binding conditions for the

plasmin:bacterium interaction, the affinity of the plasmin receptor for its ligand was determined in 0.15 M VBS-gel buffer at pH 7.4. In the initial studies the group A f-hemolytic strain 64/14 was used. Preliminary kinetic studies were conducted to establish first that equilibrium between bacterial bound and free plasmin had been achieved, and second the conditions under which saturation of bacterial plasmin receptors could be demonstrated. Binding equilibrium was found to be























































Figure 2-5. Binding of 1251-plasmin or 125-plasminogen to the group A streptococcal strain 64/14 as a function of ionic strength: The data are presented as the mean the standard deviation. Measurements of duplicate experiments were performed and are expressed as the percent of total counts offered (20,000 cpm) which were associated with the bacterial pellet. (0- O) 1 B-plasmin; (0-----0) 1251-plasminogen






34










60

50

40 I 30' 20

10


.01 .1 1.0
[NaCI M








35

established within 15 minutes at 37*C. Under these conditions using 106 bacteria and increasing concentrations of labeled plasmin, it was possible to demonstrate a plateau in plasmin binding capacity consistent with saturation of bacterial receptors, as shown in Figure 2-6. To correct for non-specific binding of radiolabel, similar studies were carried out in which binding of radiolabeled plasmin to the bacteria was measured in the presence of 100-fold molar excess of unlabeled plasmin. Non-specific binding demonstrated a linear relationship to counts offered and was less than 5% in all tubes (data not shown). Analysis of this data by least squares and Scatchard analysis demonstrated a KD of approximately 5 x 10-11 M for the association of plasmin with its receptor on the mouse passaged group A streptococcus strain 64/14. Scatchard analysis of the binding data indicates that there is a single population of plasmin receptors on streptococci (see inset, Figure 2-6), and that strain 64/14 possesses approximately 800 receptors per bacterium. Similar studies were carried out with the group A strain 64/P (the original isolate that was used in the previous mouse passage studies (Reis et al., 1984)) and with strain 64/3 isolated following three passages of strain 64 in mice. The 64/3 strain was less virulent in mice than 64/P which in turn was much less virulent than the strain recovered after 14 mouse passages, 64/14 (Reis et al., 1984). The 64/P and 64/3 strains were studied using the same protocol described for the generation of the data using strain 64/14. The results indicated that the plasmin receptor on 64/P had a KD of approximately 1 x 10-10 M and each bacterium expressed approximately 200 receptors. The 64/3 strain displayed a plasmin receptor with a KD of approximately 6 x 10-10 M and approximately 3,500 receptor sites per bacterium. Analysis of the
























































Figure 2-6. Specific binding of 1251-plasmin to 10' group A streptococci, strain 64/14, following a 15 minute incubation at 37*C in VBS-gel at pH 7.4: Measurements of triplicate experiments were performed. Specific binding was determined as described in Methods. The inset represents the Scatchard analysis of the specific binding data.






37










10,000O

8000
z
0 m 6000
-.16
S4000O ~z
.08
2000 a
2 4 6 8 10 12 COUNTS BOUND x 1000 40 80 120 160 200
COUNTS ADDED x 1000








38

Scatchard plots of 64/P and 64/3, like that shown for 64/14 in Figure 26, demonstrated only a single class of plasmin receptors expressed on these bacteria (Broeseker et al., 1988). Plasmin(ogen) contains lysine binding sites which also bind analogous amino acids (Winn et al., 1980). Epsilon-aminocaproic acid (EACA) approximates the side chain structure of lysyl residues incorporated in intact proteins and has higher affinity than lysine for these sites on plasmin(ogen), whereas arginine binds with lower affinity (Winn et al., 1980). In order to assess the possible role of the lysine binding sites of plasmin in its interaction with the bacterial receptor, the binding of plasmin to the group A streptococcus strain 64/14 in the presence of EACA, lysine, or arginine was determined. Binding was measured in 0.15 M VBS-gel, pH 7.4, containing amino acid in increasing concentrations. The percentage inhibition of binding was determined by comparison with the binding in VBS-gel pH, 7.4, buffer alone. The results of these studies are presented in Figure 2-7 and demonstrate that binding of plasmin to the group A streptococcus 64/14 was inhibited by each amino acid in a concentration dependent fashion. Fifty percent inhibition of binding of plasmin to the bacteria was observed at an EACA concentration of 0.15 mM, a lysine concentration of

2.0 mM, and an arginine concentration of 25 mM. In similar studies, plasmin was pre-bound to the group A streptococcus and a concentration dependent elution of bound radiolabel was observed on incubation with EACA, lysine, or arginine (Figure 2-8). The concentration of amino acid required to elute 50% of the bound plasmin was approximately equivalent to that required to inhibit plasmin binding by 50% (compare Figures 2-7 and 2-8).























































Figure 2-7. Inhibition of binding of 1251-plasmin to the group A streptococcal strain 64/14 in VBS-gel containing various concentrations of epsilon-aminocaproic acid, lysine, and arginine: Measurements of duplicate experiments were performed and the data are presented as the mean the standard deviation. The percent inhibition of binding was calculated by comparing with binding in VBS-gel alone. (0----O) epsilon- aminocaproic acid; (0-----0) lysine; (o-.---o) arginine.






40














100- --- -90
80
z 70
0 60 / I-/
z 40 a 30
20
10 -

.01 .1 1 10 100 1000 [mM]





















































Figure 2-8. Elution of 125I-plasmin from group A streptococcal strain 64/14 in VBS-gel containing various concentrations of epsilon-aminocaproic acid, lysine, and arginine : Measurements of duplicate experiments were performed and the data are presented as the mean the standard deviation. Percent eluted was calculated by comparing the radioactivity associated with the bacterial pellet before and after incubation in the given amino acid solution. (O------O) epsilon-aminocaproic acid; (0-----0) lysine; (0---.-0) arginine.






42













100 p-I 90 --7 80 / 70
60 50 /


30




.01 .1 1 10 100 1000 [mM]







43

Discussion


Plasminogen, an inactive zymogen can be converted to the protease plasmin by a variety of plasminogen activators (Collen, 1980). This enzyme demonstrates broad substrate specificity. In addition to fibrin cleavage, plasmin can activate the first component of the classical complement pathway, hydrolyze coagulation factors, degrade components of basement membrane, and break down connective tissue (Atichartakarn et al., 1978; Jones and DeClerck, 1980; Liotta et al., 1981). Furthermore a variety of potent split products are generated as a consequence of plasmin activity, e.g. chemotactic fibrinopeptides (Kay et al., 1974). Effective regulation of plasmin activity is therefore important in order to prevent tissue damage and inflammation. Normally the selective protease inhibitor a2-antiplasmin regulates plasmin activity in man (Aoki et al., 1977).

Interaction of streptococci and streptococcal products with the fibrinolytic system has been recognized for many years (Tillett and Sherry, 1949). The observation that certain streptococci could lyse a fibrin clot lead to the identification and isolation of streptokinase. This secreted protein is known to bind to human plasminogen and plasmin with a similar affinity (Reddey and Markus, 1972). In this study I have identified a surface receptor on certain group A streptococci, grown in either Todd-Hewitt broth or chemically defined media, that specifically binds to plasmin while demonstrating no significant affinity for the zymogen form of the molecule, plasminogen. Thus the surface receptor we have identified is distinct from streptokinase. Furthermore, this binding phenomenon did not appear to be simply a function of the ligand being a serine protease, since no binding activity was demonstrated with the two other serine proteases examined, trypsin and urokinase.







44

Binding of plasmin to its bacterial receptor does not inhibit the ability of the enzyme to cleave either small synthetic substrates or its natural substrate fibrin. Aprotinin, a naturally occurring tight-binding inhibitor of plasmin and phe-pro-arg chloromethylketone which chemically modifies the histidine residue of the active site can react with the bound plasmin and neutralize its enzymatic activity. These findings suggest that the catalytic portion of the plasmin molecule is not interfered with by the association with the bacteria. Of interest was the observation that the enzymatic activity of bacterial bound plasmin could not be regulated by addition of its specific inhibitor, a2antiplasmin. Alpha2-antiplasmin is a potent inhibitor of plasmin in the fluid phase forming a 1:1 stoichiometric complex between the enzyme and inhibitor.

The failure of a2-antiplasmin to regulate bacterial bound plasmin provides the bacteria with a potential mechanism for tissue invasion by virtue of the ability of plasmin to hydrolyze components of connective tissue and basement membranes. Recent studies of the invasive characteristics and metastatic potential of tumor cells has suggested a key role for plasminogen activators in this process (Dano et al., 1985). The ability of certain group A streptococci to produce a plasminogen activator (e.g., streptokinase) and also to express a receptor for the activation product plasmin may account for certain of its invasive properties. Furthermore, since plasmin bound to a group A streptococcus is incapable of inhibition by a2-antiplasmin the bacteria has associated with it a non-regulatable proteolytic activity that may help to contribute to its tissue invasive properties.

A variety of receptors for human proteins have been described on streptococci. These include receptors for key components of the immune







45

system such as Clq (Yarnall et al., 1986), IgG (Kronvall, 1973), and IgA (Russell-Jones et al., 1984), the serum protein fibrinogen (Kronvall et al., 1979), and the stromal structural proteins laminin (Switalski et al., 1984), and fibronectin (Myhre and Kuusela, 1983). The importance of any of these receptors in the pathogenic process remains controversial. Our recent observation that streptococci also have a receptor that is specific for human plasmin adds to this list (Lottenberg et al., 1987). Although the primary substrate for plasmin is fibrin, plasmin is a non-specific protease capable of also hydrolyzing such extracellular matrix proteins as thrombospondin, fibronectin, and laminin, while also exposing matrix components for degradation by other enzymes (Knudsen et al., 1986). The ability to bind plasmin in an active form which can not be regulated by its efficient physiological regulator a2-antiplasmin could provide for surface mediated protease activity and a mechanism for tissue invasiveness by plasmin receptor-positive bacteria (Lottenberg et al., 1987).

In this study it was shown that the group A streptococcus strain

64/14 demonstrates optimal binding of its ligand at physiological pH and ionic strength. The interaction had a high affinity (KD =5 x 10-11 M) and demonstrated a linear Scatchard plot indicating that a single population of plasmin receptors was present on the bacteria (Broeseker et al., 1988). There was no evidence for either an additional low affinity receptor or for any cooperativity, positive or negative, in the binding of ligand with the specific receptor. In agreement with our previous observations (Lottenberg et al., 1987), there was no evidence for specific interaction between bacteria and the native zymogen form of the protein, plasminogen.







46

Plasmin(ogen) has several lysine binding sites located on its heavy chain. The low affinity sites are primarily important for binding to fibrin and the high affinity site is important for the interaction with a2-antiplasmin (Wiman et al., 1979). In order to assess the possible role of these lysine binding sites in the interaction between plasmin and its bacterial receptor, the bacterial binding of plasmin in the presence of increasing concentrations of EACA, lysine, or arginine. The results in Figure 2-7 and 2-8 demonstrate that binding could be inhibited, and bound plasmin could be eluted, by these amino acids in a concentration-dependent manner. Eluted plasmin could be re-bound to bacteria simply by removing lysine from the eluted plasmin solution, indicating a possible importance of the lysine binding sites for the receptor:plasmin interaction. There are 4 or 5 low affinity sites (KD = 5mM) and one high affinity site (KD = 9pM) for EACA (Markus et al., 1978). Lysine and arginine bind to the high affinity site less tightly than does the lysine analog EACA (Wiman and Collen, 1978). Analysis of the inhibition curves for EACA, lysine, and arginine reveal that occupancy of the high affinity lysine binding site on plasmin interferes with binding to the bacteria. It is recognized that occupancy of the high affinity lysine binding site causes gross conformational changes in the plasmin(ogen) molecule, and therefore the possibility for allosteric as well as direct effects needs to be considered. The very high affinity of the receptor for plasmin, approximating the affinity of a2-antiplasmin for plasmin (KD 2 x 10-10 M, Wiman and Collen, 1978), suggests that streptococci may be able to compete effectively with a2-antiplasmin for plasmin. and could explain why bacterial bound plasmin cannot be regulated by a2-antiplasmin. The enzymatic inhibition of plasmin by







47

a2-antiplasmin occurs in a two-step process (Christensen and Clemmenson, 1977; Wiman and Collen, 1978). The first step is a non-covalent association of the a2-antiplasmin molecule with the lysine binding site located in the kringle 1 region of the plasmin molecule, followed by a rapid covalent linkage to the plasmin active site serine residue. The observation that lysine or EACA effects, on the high affinity site of plasmin, disturbs the plasmin bacterial receptor interaction, together with the observation that bound plasmin is not inhibited by a2antiplasmin, suggests that the kringle 1 domain may be important in the binding of plasmin to bacteria, or may be inaccessible.

Previous studies have shown that passage of streptococci in mice heightened virulence with concomitant enhanced expression of certain surface proteins (Burova et al., 1980; Burova et al., 1981; Reis et al., 1984). Group A streptococcal strain 64 exhibits decreased expression of Fc receptors after 3 or 4 mouse passages (strain 64/3 and 64/4) as compared to the parent strain (64/P) (Reis et al., 1984). Following 8 mouse passages this group A streptococcus demonstrates markedly enhanced Fc receptor expression which appears to be a stable characteristic of the selected strain (Reis et al., 1984). In this study the average affinity of the plasmin receptor expressed on strains 64/P, 64/3, and 64/14 was not significantly different, indicating that mouse passage did not have a major selective pressure on the affinity of the plasmin receptor.

There was some variation in the number of plasmin receptors calculated for each bacteria. The 64/P, 64/3, and 64/14 strains displayed 200, 3500, and 800 receptor sites, respectively, per bacterium. Clearly these are average numbers of receptors per bacterium (Broeseker et al., 1988), and given the potential errors in estimating bacterial







48

number, the effect of phase variations in the expression of different proteins by bacteria (Cleary et al., 1987), and the heterogeneity in receptor expression among colonies (Yarnall et al., 1984), we believe that such differences need to be cautiously interpreted. It does not appear from the results of these studies that the degree of plasmin receptor expression correlates with the virulence of these group A isolates in mice. Nonetheless we calculate that binding of active plasmin in the picomolar range with any of the group A isolates studied is achievable. The high affinity for and slow off rate of bound plasmin may make these interactions with streptococci of importance in the infectious process.

The next series of studies, described in the following chapter, were designed to analyze the way in which the bacterial receptor associated with its ligand, the human plasmin molecule.
















CHAPTER THREE
LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN BINDING TO ITS
SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR


Introduction


The studies documented in Chapter Two demonstrate that certain pathogenic group A streptococci, grown in either Todd-Hewitt broth or chemically defined media, express a receptor that binds to human plasmin while demonstrating no significant reactivity with the native zymogen form of the protein, Glu-plasminogen or with other serine class proteases. Bacterial-bound plasmin retains its enzymatic activity and can no longer be regulated by its physiological inhibitor, a2antiplasmin. Optimal binding of plasmin to its bacterial receptor was shown to occur under physiological conditions of ionic strength and pH. This interaction of plasmin with a group A streptococcus had a high affinity with an estimated dissociation constant of approximately 1.0 x 10-10 M. Plasmin binding was inhibited reversibly by lysine or epsilon amino caproic acid, (EACA). These data suggest that the lysine binding kringle structures of the plasmin molecule might be involved in the association of plasmin with the bacterial receptor. In this chapter I describe the experiments performed to localize the region of the plasmin molecule which interacts with the bacterial plasmin receptor. Binding of plasmin to a group A streptococcus is dependent on the conformation of the plasmin molecule, and involves interactions that are distinct from those occurring between other known plasmin(ogen) binding molecules like streptokinase, fibrin, fibrinogen, thrombospondin, or a2-antiplasmin.

49







50

Materials and Methods


Enzymes, Inhibitors and other Reagents

Urokinase and porcine elastase (type IV) were obtained from Sigma Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from Mobay Pharmaceuticals, New York, NY. Phe-Pro-Arg-chloromethylketone (PPACK) was obtained from Calbiochem-Behring, San Diego, CA. Human Lysplasminogen was obtained from American Diagnostica Inc., Greenwich, CT. H-D-Val-Leu-Lys-paranitroanilide (S-2251) was obtained from Helena Laboratories, Beaumont, TX.

Human Plasminogen

Native human plasminogen (Glu-plasminogen) was prepared from human plasma by chromatography on lysine-Sepharose and molecular sieving chromatography on Superose 6 (Pharmacia-FPLC, Piscataway, NJ). The purified protein appeared as a single band on a silver stain of an SDSpolyacrylamide gel. Plasminogen was quantified by measuring absorbance using a A8Onm value of 17.0 (Nilsson et al., 1982). The protein was also quantified antigenically by Laurell Rocket electrophoresis (Laurell, 1966). The purity of the isolated human plasminogen was confirmed by activation of a known quantity of plasminogen with streptokinase and measuring amidolytic activity. The observed and theoretical predicted enzymatic activity were equivalent, within experimental error. Human Lys-plasminogen, a modified form of Glu-plasminogen in which 76 of the NH2-terminal amino acid residues are removed (Glu-1 to Lys-76) was obtained from American Diagnostica Inc., Greenwich, CT. The homogeneity of this Lys-plasminogen preparation was analyzed using both an urea gel electrophoresis procedure and an acetic acid urea gel electrophoresis procedure. This Lys-plasminogen preparation demonstrated the appropriate







51

migratory property (a shift to a lower Mr of approximately 85,000 daltons) in comparison to native Glu-plasminogen (Mr of approximately 92,000 daltons).

Iodination of Proteins

Glu- and Lys-plasminogen were iodinated by the chloramine T method using lodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell (1982). The labeled proteins were separated from free iodine by passage over a G-25 column (PD-10, Pharmacia) and collected in 0.15 M Veronal buffered saline, pH 7.4, containing 0.001 M Mg++, 0.00015 M Ca+, and 0.1% gelatin (VBS-gel). The labeled proteins were stored in aliquots containing 0.02% sodium azide at -20*C. The concentration of 1251 plasminogen was determined antigenically using a sandwich enzyme-linked immunosorbent assay (ELISA) technique utilizing goat anti-human plasminogen IgG fraction from Atlantic Antibodies, Scarborough, ME. This assay could measure plasminogen reliably in the nanogram range. Generation of Plasmin

Lys-plasmin was generated from radiolabeled or unlabeled Glu- or Lys-plasminogen by incubation with urokinase (20 units/ml) in VBS-gel (unless stated otherwise) that contained 0.04 M lysine. The conversion of the single chain zymogen molecule to the two chain plasmin enzyme was monitored on SDS-PAGE under reducing conditions as described previously by Lottenberg et al., (1987). Conversion of the zymogen to the active enzyme was maximal after 30 min incubation at 37*C. Glu-plasmin was generated by a similar procedure with the exception that a 10-fold molar concentration of aprotinin relative to the Glu-plasminogen concentration was added prior to addition of urokinase (Swenson and Thorsen, 1981). Mini-plasmin was generated from mini-plasminogen using the same







52

activation procedure described to generate Lys-plasmin. Phe-pro-argchloromethylketone (PPACK) reacted radiolabeled or unlabeled plasmin was obtained by mixing a 5-fold molar excess of the inhibitor with plasmin and incubating at 37*C for 30 min.

Bacteria

The group A, p-hemolytic, streptococcal strain 64/14 was grown in Todd-Hewitt broth (Difco, Detroit, MI) overnight at 37*C as stationary cultures (Yarnall and Boyle, 1986a). The bacteria were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4, containing 0.05% Tween 20 and 0.02% sodium azide. The bacteria were heat killed at 80*C for 15 min. The suspension was centrifuged and the pellet washed twice with VBS-gel containing 0.02% sodium azide. Aliquots were stored at -20*C. Stocks of 10% wet weight/vol suspensions were prepared in VBS-gelatin containing 0.02% sodium azide. The concentration of a bacterial suspension was determined by counting bacterial chains in a Neubauer hemacytometer (Fisher Scientific, Orlando, FL). Polyacrylamide Gel Electrophoresis

Electrophoresis was carried out as described by Weber and Osborn (1969) with the addition of 6.0 M urea to the polyacrylamide gel. The polyacrylamide gels consisted of a 4% stacking gel layered onto a 10% or 12% polyacrylamide gel containing 0.1% sodium dodecylsulfate, 0.05 M sodium phosphate pH 7.1, 6.0 M urea. Slab gels were used in the Bio-Rad Protean II system (BioRad, Richmond, CA). Protein samples were prepared by mixing an equal volume of sample buffer containing 0.1 M sodium phosphate pH 7.1, 8.0 M Urea, and 4.0% SDS with the protein solution, and heating at 80*C for 2 minutes. Sample buffer containing 0.72 M 'mercaptoethanol was used to prepare protein samples in the reduced state.







53

Preparation of Elastase Digestion Fragments of Plasminogen

Elastase digestion of human plasminogen yields three defined

fragments of the plasminogen molecule (Sottrup-Jensen et al., 1978). These are 1) the lysine-binding domain I (LBS-I), Mr of approx. 38,000 daltons containing kringle domains 1 through 3, 2) Lysine binding domain II (LBS-II), Mr of approx. 10-12,000 daltons consisting of the kringle domain 4, and 3) the non-lysine-binding domain known as mini-plasminogen, Mr of approx. 36,000 daltons containing the remainder of the heavy chain (kringle 5) and intact light chain. Elastase digestion was performed using established conditions (Sottrup-Jensen et al., 1978). Purified Glu-plasminogen (3.0 mg/ml) in 0.05 M Tris, 0.1 M NaCI, pH 8.0, was digested with a 40:1 molar ratio of Glu-Plasminogen to porcine elastase in the presence of 250 KIU/ml aprotinin for 6.5 hours at room temperature with gentle stirring in a total volume of 20 mls. At this time an aliquot containing 50 Mg of protein was removed for analysis by SDS-PAGE and silver staining, to determine the extent of plasminogen digestion. The remainder of the reaction mixture was flash frozen and stored at 70*C. The fragments were subsequently purified by a combination of affinity chromatography on lysine-Sepharose and gel filtration on Superose 6 (Pharmacia FPLC). The concentrations of the purified proteins (see Figure 3-1, panel A) were determined spectrophotometrically, using previously reported A 80nm values of 17.0 for both Glu- and Lysplasminogen (Holvoet et al., 1985), 14.0 for mini-plasminogen (Holvoet et al., 1985), 22.5 for LBS I (Nilsson et al., 1982), 25.0 for LBS II (Nilsson et al., 1982), and 16.0 for plasmin heavy chain and plasmin light chain (Summaria and Robbins, 1976). All proteins were aliquoted and stored at -70*C.







54

Preparation of Plasmin Heavy and Light Chains

Plasmin heavy and light chains were prepared essentially as

described by Summaria and Robbins (1976). Twenty mg of Lys-plasmin, enzymatically inhibited with a 5 fold molar excess of aprotinin in 5 mis of 0.05 M Tris, 0.1 M NaCl, pH 8.0, was reduced by treatment with 0.1 M P-mercaptoethanol for 20 min. at 20*C. The reduced solution was then cooled in an ice slurry and carboxymethylated with 0.1 M sodiumiodoacetate on ice for 10 min. The plasmin heavy and light chains were then separated and purified by a combination of affinity chromatography on lysine-Sepharose, concentration by ammonium sulfate precipitation (4.0 g/10 ml), resuspension in 0.05 M Tris, 0.1 M NaC1, pH 8.0, and subjected to gel filtration on Superose 6 (Pharmacia FPLC). The isolated plasmin and plasminogen fragments were analyzed for purity on a reduced SDS-6 Murea-polyacrylamide gel. As shown in Figure 3-1, panel B, the various fragments demonstrated appropriate molecular sizes and were homogeneous. Concentrations were determined as described above. All proteins were aliquoted and stored at -70*C.

Direct Binding Assay of Radiolabeled Proteins

The ability of radiolabeled plasminogen fragments to bind to the group A streptococcus 64/14 was measured as described previously by Lottenberg et al., (1987). A fixed number of bacteria were incubated with labeled proteins (approximately 30,000 cpm per tube) in a total volume of 400 pl of VBS-gel for 30 min at 37*C. The bacteria were pelleted by centrifugation at 1000 x g for 10 min and the pellets washed twice with 2.0 ml of VBS-gel. The radioactivity associated with the bacteria was determined in a Beckman 5500 Auto gamma counter (Beckman Instruments, Inc., Fullerton, CA). Non-specific background binding was







55

determined in replicate tubes which contained no bacteria. All estimates were performed in duplicate.

Inhibition of Plasmin Binding to Bacteria by Purified Plasmin(ogen)
Fragments

The ability of different concentrations of one or more of the isolated plasminogen fragments to inhibit binding of Phe-Pro-ArgChloromethylketone (PPACK) reacted 1251-Lys-plasmin to the group A streptococcus 64/14 was tested using a modification of the direct binding assay described above. Different concentrations of plasmin(ogen) or plasmin(ogen) fragments were mixed with a fixed dilution of a 10% w/v suspension of streptococcal strain 64/14 and PPACK reacted 125I-Lysplasmin (approx. 30,000 cpm per tube) followed by incubation for 30 min at 37*C. Bacterial associated radioactivity was determined after washing away unbound label as described above. The inhibition of binding of labeled plasmin was calculated by comparing the number of counts bound in the absence of competitor with the number of counts bound when the competitor was present. All samples were corrected for background binding of counts. Counts bound in the tubes from which bacteria were omitted or in tubes in which a 100-fold molar-excess of unlabeled ligand was added. In no case was the background level of radioactivity greater than 5% of the counts offered. Furthermore, background levels in the presence of excess cold competitor, or in the absence of bacteria were not significantly different.

Elution and Analysis of 125I-Lys-plasmin(ogen) from Bacteria

1251-Lys-plasminogen (approx. 100,000 cpm) was added to a 100 'l

aliquot of a 10% w/v solution of strain 64/14 bacteria in a total volume of 400 4l VBS-gel and allowed to incubate at 37*C for 30 min. The bacteria were then pelleted by centrifugation (3000 x g, 10 min) and







56

washed three times with 2.0 ml VBS-gel. The bacterial pellets were resuspended in 300 pl of either VBS-gel containing 0.5% SDS; VBS-gel containing 0.1M EACA; or VBS-gel containing 0.5% SDS and 2.0% jmercaptoethanol, to elute the 1251-Lys-plasminogen from the bacteria. Following a 10 minute incubation at 37*C the bacteria were removed by centrifugation and the supernatant recovered. The eluted material was analyzed by electrophoresis on a 10%-SDS-PAGE-6M-Urea gel under reducing conditions. The gel was dried and the migration of labeled protein determined by autoradiography. Similar studies were also carried out in which the bacterial bound 1251-Lys-plasminogen was treated with a 20 unit/ml concentration of urokinase for 20 minutes at 37*C in a total volume of 300 pl of VBS-gel prior to eluting the bound proteins. Following this plasminogen activation reaction the bacteria were centrifuged and washed twice with 2.0 mls VBS-gel. The residual bound 1251-Lys-plasmin(ogen) was eluted and analyzed as described above. Measurement of Functional Activity of Plasmin(ogen) in Bacterial-Free
Supernatants

The following assay was used to measure binding of the various plasmin(ogen) as an alternative method to using radiolabeled tracers. In these studies, 2.0 pg of Glu-plasminogen, Lys-plasminogen, or Lysplasmin was incubated with 100 Al of a heat killed 10% w/v suspension of the group A streptococcal strain 64/14 for 20 minutes at 37*C in a total reaction volume of 400 pl of VBS-gel. Following incubation, the bacteria were removed by centrifugation at 12,000 x g for 4 minutes in an Ependorf Microfuge and bacterial-free supernatants were obtained. Control tubes for each plasmin(ogen) species containing no bacteria were treated identically and all samples were run in duplicate. The bacterial free supernatants were recovered and enzymatic activity was measured as







57

follows. The bacterial free supernatants or the corresponding control samples were added to plastic cuvettes containing 106 IU streptokinase in a total volume of 900 pl of enzyme assay buffer (0.05 M Tris, 0.05 M NaC1, 0.1% PEG-8000, pH 7.4). For the Glu-and Lys-plasminogen preparations the reaction mixture was incubated at 37*C for 10 minutes to allow plasminogen-streptokinase complexes to form. For Lys-plasmin, a similar incubation with streptokinase was performed to allow for equivalent substrate turnover to that of zymogen-streptokinase complexes. Following incubation, H-D-Val-Leu-Lys-pNA (S-2251) was added to yield a final concentration of 300 pM. Tubes were allowed to incubate for precisely five minutes and then quenched with 100 pl of glacial acetic acid. The amount of substrate hydrolysis, which is directly proportional to the amount of plasmin enzyme present was then quantified by measuring the absorbance of the reaction mixture at 405 nm.

The enzymatic activity of the bacterial free supernatant was

determined by comparison with the enzymatic activity of known standards. The percent of residual enzymatic plasmin(ogen) activity in the bacterial free supernatant was calculated by determining the fraction of total enzymatic activity in a control sample remaining in the supernatant following incubation with bacteria. Control tubes containing bacteria and substrate, and substrate in buffer were included. All assays were performed in duplicate.

Measurement of Functional Activity of Plasmin Associated with Bacteria

The plasmin activity associated with bacterial pellets was examined using the chromogenic substrate as described above. Following binding and centrifugation the pellets were washed 3 times with 1.0 ml of VBS-gel and resuspended in 400 p1 of the enzyme assay buffer. S-2251 was added







58

to yield a final concentration of 300 pjM. The resuspended bacterial pellets were then incubated at 37*C for 20 minutes and quenched with 50 Ml of glacial acetic acid. The bacteria were removed by centrifugation (12,000 x g for 4 minutes) and the optical density of the bacterial free supernatant was measured at 405 nm. Control tubes containing bacteria and substrate, and substrate in buffer were included. All assays were performed in duplicate.


Results


The experiments described in this chapter were designed to map the domains on the human plasmin molecule involved in the high affinity interaction with the group A streptococcal strain 64/14. For these studies a variety of defined plasminogen fragments as well as the heavy and light chains of plasmin were prepared as described in the Methods. The plasminogen fragments obtained were characterized on urea gels, see Figure 3-1. The homogeneous plasminogen fragments were used to compete with intact PPACK reacted 1251-plasmin for receptor sites on the group A streptococcal strain 64/14 (Table 3-1). I have previously demonstrated that plasmin treated with PPACK, p-Nitrophenyl-p-guanidinobenzoate (pNpGB), or aprotinin, does not effect plasmin's binding reactivity to the group A streptococcal strain 64/14 as documented in Chapter Two. Enzymatic inhibition of the 1251-Lys-plasmin and urokinase in the labeled tracer preparation was necessary to prevent the proteolytic conversion of Glu-plasminogen to Lys-plasminogen, or Lys-plasmin (Markus et al., 1978) by the labeled tracer mixture. In all the competitive inhibition experiments described in this study, a constant concentration of PPACK125I plasmin (1.0 x 10-10 M) and a range of concentrations of unlabeled























Figure 3-1. SDS-UREA-PAGE analysis of isolated plasmin(ogen) fragments. Panel A: Elastase digestion fragments of plasmin(ogen): Glu-plasminogen was digested with elastase and fragments purified as described in Materials and Methods. Panel A: Lane 1: Mini-PLG (4.0 pg); 2: Glu-PLG (4.0 pg); 3: Mini-PLA (4.0 pg); 4: LBS-I (4.0 pg); 5: LBS-II (4.0 pg); M: molecular weight standards. Panel B: Plasmin heavy (HC) and light (LC) chain preparations: Lys-plasmin was reduced and carboxymethylated as described in Materials and Methods. Lane 1: Glu-PLG (5.0 Ag); 2: lys-PLA (5.0 Mg); 3: HC (5.0 Ag); 4: LC (5.0 pg). Proteins were electrophoresed under reducing conditions on a SDS-6 M urea-10%-polyacrylamide gel.









A B



12345

1 2 3 4
-66 KD
plasminogen-45 ( o -heavy chain .g -36


-24
U U -light chain



a-14
C.







61

competitor molecules (10-6 M to 10-10 M) were mixed with a fixed concentration of bacteria. Following incubation and washing, the amount of radiolabeled plasmin bound to the bacterial pellet was determined. The quantity of radioactivity bound in the presence or absence of unlabeled competitor was compared and the degree of inhibition calculated (Figure 3-2). The results summarized in Table 3-1 show that unlabeled plasmin inhibits the binding of labeled plasmin efficiently, with 50% inhibition being observed in the presence of 1.2 x 10-8 M Lys-plasmin. Significant inhibition of radiolabeled plasmin binding was also observed when purified heavy chain was used as the competitor. Addition of any of the other plasminogen fragments including isolated lysine binding domains of the heavy chain (LBS I or LBS II) demonstrated no significant inhibitory effect (Figure 3-2). Similarly, mini-plasminogen, miniplasmin, and isolated light chains demonstrated no significant inhibition of binding of radiolabeled lys-plasmin over the concentration range tested (10-6 M to 10-10 M) (Figure 3-2). Identical results were obtained in the inhibition assays involving mini-plasmin, mini-plasminogen, LBS I, LBS II, Lys-plasmin heavy chain, and plasmin light chain in the absence of protease inhibitors in the reaction mixture (data not shown).

Combining equimolar quantities of the elastase digested fragments of plasminogen or plasmin failed to restore any inhibitory potential. Furthermore, combination of isolated light chain and heavy chain demonstrated no synergistic effect in inhibitory capacity compared to the sum of the isolated fragments alone (data not shown). The inhibition curves for isolated heavy chain and intact plasmin (Figure 3-2) demonstrated that both preparations could inhibit binding of labeled Lysplasmin by 100%. However, these curves differed in shape, indicating

















































Figure 3-2. Inhibition of, PPACK reacted, 1251-Lys-plasmin binding to group A streptococcal plasmin receptor: A constant concentration of (1.0 x 10"-0 M) PPACK reacted 1251-Lys-plasmin and an increasing concentration of unlabeled competitor molecules (10-10 M to 10-6 M) were mixed with a fixed concentration of streptococcal strain 64/14. Following incubation and washing (see Materials and Methods), the amount of radiolabeled Lys-plasmin bound to the bacterial pellet was determined. The quantity of radioactivity bound in the presence of unlabeled competitor was compared to the radioactivity bound in the absence of inhibitor and the percent inhibition calculated. (0 -Lysplasmin; o -HC; w -LBS-I; 0 -LBS-II; A -LC; A -Mini-PLG; V -Mini-PLA).







63
















100

90 80 Z 70
z
o 60
LL
0 50
z
0
S40 :F 30
z

S20

10


-10 -9 -8 -7 -6 -5
10 10 10 10 10 10
CONCENTRATION OF COMPETITOR (M)







64

differences in the efficiency of inhibition. The isolated heavy chain was found to be less efficient an inhibitor than the intact plasmin molecule. These findings suggest that there is some component involved in the interaction of plasmin with the bacteria that is either not present on the heavy chain or is altered during the isolation procedure.

Two possibilities to account for these observations were considered. The first was that there are some sites on the heavy chain of the plasmin molecule that are modified when the molecule is purified, thereby changing its efficiency of interaction with the bacterial receptor. The second was that the plasmin light chain, while associated with the heavy chain, confers a different tertiary structure to the molecule than exists on either (or both) of the isolated chains. Such a change in conformation of the molecule might affect its interaction with the bacteria.

It has been established previously that a conformational change occurs when Glu-plasminogen is activated to Lys-plasmin, or when Gluplasminogen is converted to Lys-plasminogen (Swenson and Thorsen, 1981; Markus et al., 1978; Thorsen, 1975). Lys-plasminogen is the zymogen form of plasminogen, lacking the 76 amino acid NH2-terminus of the native protein (Markus et al., 1978). This modification results from the proteolytic activity of plasmin on Glu-plasminogen, which removes the 76 amino acid NH2-terminus, resulting in a new NH2-terminus lysine (for review, see Thorsen et al., 1981). This modification occurs without generation of protease activity. The conversion of Glu-plasminogen to Lys-plasmin or to Lys-plasminogen not only results in a marked conformational change of the protein but also causes an increase in the binding affinity of these molecules to fibrin (Thorsen, 1975), as well as







65

lowering the dissociation constant between these molecules and a2antiplasmin (Swenson and Thorsen, 1981; Wiman et al., 1979).

To examine the possible importance of the conformation of the plasmin(ogen) molecule for binding to bacteria, the ability of the conformationally altered form of plasminogen, Lys-plasminogen, to bind to the group A streptococcus, 64/14, was measured. The isolated protein was radiolabeled and examined by urea gel analysis for homogeneity. The labeled material demonstrated a single band on an autoradiograph (Figure 3-3, panel A, lane 2) at a position corresponding to that reported for the migration of Lys-plasminogen in this gel system (Swensen and Thorsen, 1981). This labeled form of plasminogen was found to bind to the bacteria (Figure 3-3, panel B, lane 2). Similarly, Glu-plasmin generated from Glu-plasminogen in the presence of aprotinin, was also capable of binding to the bacteria (Figure 3-3, panel B, lane 3). The relative efficiency of unlabeled Glu-plasminogen, Lys-plasminogen and Lys-plasmin to compete with labeled plasmin for binding sites on the group A streptococcus 64/14 was tested. Different concentrations of each of these molecules were mixed with a fixed concentration (1.0 x 10-10 M), of PPACK reacted 1251-Lys-plasmin and the extent of inhibition of binding of radiolabel was measured, as described previously. The results of this experiment shown in Figure 3-4 indicate that the inhibition achieved with Lys-plasminogen and Lys-plasmin were identical. These results indicate that the receptor for these ligands are the same, and that the affinity for each protein is equivalent.

The possibility that the results presented in Figure 3-4 could be

accounted for by the conversion of Lys-plasminogen to Lys-plasmin during the reaction was considered. The next series of experiments were




















































Figure 3-3. Binding of 1251 labeled Glu- and Lys-plasmin(ogens): Gluand Lys-plasmin(ogen) were generated as described in Materials and Methods. The labeled tracers were then used in direct binding assays with a fixed concentration of the streptococcal strain 64/14. Panel A is an autoradiograph demonstrating the analysis of each reduced 1251 labeled sample on a SDS-6 M-Urea 12%-polyacrylamide gel by autoradiograph to verify their molecular form. (Glu-H: Glu-heavy chain; Lys-H: Lys-heavy chain; L: light chain). Panel B illustrates the percent of offered cpm bound to bacterial pellets. (Lane 1: Glu-PLG; 2: Lys-PLG; 3: Glu-PLA; 4: Lys-PLA).







67












1. 2. 3. 4.A.






< GIu-H < Lys-H








ft< L
50
B.

40
z

0 30


L 20


10


















































Figure 3-4. Inhibition of, PPACK reacted, 1251-lys-plasmin binding to group A plasmin receptor: A constant concentration of (1 x 10-10 M) PPACK reacted 125-Lys-plasmin and an increasing concentration range of Lys-PLG, Glu-PLG or Lys-PLA (10-10 M to 10-6 M) were mixed with a fixed concentration of the streptococcal strain 64/14. Following incubation and washing (see Materials and Methods), the amount of radiolabeled Lysplasmin bound to the bacterial pellet was determined. The quantity of radioactivity bound in the presence of unlabeled competitor was compared to the radioactivity bound in the absence of inhibitor and the percent inhibition calculated. ( o-Lys-PLA; -Lys-PLG; A Glu-PLG).







69















100 90 80
z
o 70
z
60
0
z 50040
I
Z 30

2010


10-10 10-9 10-8 10-7 10-6 10-5
CONCENTRATION OF COMPETITOR (M)







70

designed to determine whether Lys-plasminogen binds to the bacteria without first being activated. These experiments were carried out by monitoring the distribution of Lys-plasminogen, Glu-plasminogen, or Lysplasmin, in the fluid phase and associated with the bacteria, following incubation of the protein with the bacteria. Unlabeled Glu-plasminogen, Lys-plasminogen, or Lys-plasmin was added to a fixed concentration of the group A streptococci 64/14 and incubated for 30 minutes at 37*C. Following this incubation period, the bacteria were pelleted by centrifugation and the supernatants were recovered and monitored for enzymatic activity either directly for Lys-plasmin, or following activation with excess streptokinase for the sample containing Gluplasminogen or Lys-plasminogen, as described in the Methods. Following incubation with bacteria, and removal of the bacteria by centrifugation, there was no significant Lys-plasmin activity detectable in the bacterial free supernatant (Figure 3-5). By contrast over 98% of the enzymatic potential of Glu-plasminogen was detected in the supernatant, while in similar experiments using Lys-plasminogen less than 10% of the enzymatic potential was measured following activation with streptokinase (Figure 3-6). Because of differences in the efficiency of detection of plasmin activity in the fluid phase compared to its activity when bound to bacteria it is not possible to quantitate accurately the exact percentage of plasmin activity that is bound to bacteria. However, I have demonstrated previously that once associated with bacteria, the plasmin retains its ability to cleave synthetic chromogenic substrates like H-Dval-leu-lys-pNA (S-2251), as documented in Chapter Two. Consequently, the washed pellets from the absorption reaction were incubated with this synthetic substrate. The results presented in Table 3-2 demonstrate























































Figure 3-5. Binding of Lys-plasmin(s), derived from Glu-plasminogen and Lys-plasminogen, to the group A streptococcal strain 64/14 as measured by residual activity in the bacterial free supernatant: (0o -urokinase activated Glu-plasminogen alone; 0 -urokinase activated Glu-plasminogen + bacteria; o -urokinase activated Lys-plasminogen alone; a -urokinase activated Lys-plasminogen + bacteria). For precise experimental details see Materials and Methods.







72


















0.2






E

" 0.1










1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 I 1.0

PLASMIN (nM)





Table 3-1.

Summary of inhibition experiments of PPACK reacted 1251-Lys-plasmin binding to the group A streptococcal strain 64/14 are shown with a
schematic depiction of the portion of the native molecule they represent.**


COMPETITOR Iso%(L M)

Glu-Plasminogen Glul- 0 0 __ 790 > 2.0


Lys-Plasminogen Lys77 ) ) 790 0.010



Lys-Plasmin Lys77 7) 790 0.012 s s
Mini-Plasminogen Val 790 > 1.0 442 790


s-s

LBS-I(K1 -K3) Tyr79 > 1.0 LBS-T(K4) Val354~ > 1.0

Heavy (A) Chain Lys77 - (---- 0.046 Light (B) Chain Valse561 X 790 > 1.0 Activation cleavage site (arginine560-valine561).
-(* Plasmin active site residues (histidine602; aspartic acid645; serine740) from left to right.


**Structure and NH2-amino-terminal residue data were obtained from the work of Sottrup-Jensen et al., (1978). Inhibition is expressed as 50% inhibitory values in (uM) with Lys-plasmin as the standard (see Materials and Methods).







74

Table 3-2.

Measurement of plasmin(ogen) associated with bacterial pellets.





Absorbance at 405 nm
Following a 20 Minute
Incubation at 37*C with HBacteria Pre-Incubated With: D-Val-Leu-Lys-Paranitroanilide

Buffer 0.023 0.002 Glu-plasminogen 0.026 0.002 Lys-plasminogen 0.095 0.004 Lys-plasminI 0.630 0.001 Lys-plasmin2 0.637 0.002







Two pg of the indicated enzyme or zymogen was incubated with a
fixed dilution of the streptococcal strain 64/14. Following incubation the bacteria were pelleted by centrifugation, washed, resuspended in 400 pl VBS-gelatin, and assayed for enzymatic activity by hydrolysis of the chromogenic substrate H-D-Val-Leu-Lys-paranitroanilide (see Materials and Methods). The spontaneous cleavage of the substrate under the experimental conditions in the presence of bacteria alone was an absorbance (405 nm) of 0.024 0.002.

1. Urokinase activated Glu-plasminogen 2. Urokinase activated Lys-plasminogen
























































Figure 3-6. Binding of Glu- and Lys- plasminogen to the group A streptococcal strain 64/14 as measured by residual activatable zymogen in the bacterial free supernatant: ( e Glu-plasminogen + bacteria; o-Glu-plasminogen alone bacteria; *-Lys-plasminogen + bacteria; D-Lysplasminogen alone). For precise experimental details see Materials and Methods.







76



















0.4




0.3



E
'o 0.2




0.1





1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0

PLASMINOGEN (nM)
















































Figure 3-7. Characterization of 1251-Lys-plasmin(ogen) species eluted from bacteria: Eluted labeled proteins were analyzed by electrophoresis on a 10%-SDS-PAGE-6M urea gel under reduced conditions. Lanes 1, 2, and
3 contain labeled proteins from bacteria pre-incubated with 1251-Lysplasminogen and eluted by 0.5% SDS; 0.1 M EACA; or 0.5% SDS containing
2.0% 6-mercaptoethanol respectively. Lanes 4, 5, and 6 are identical to Lanes 1, 2, and 3 with the exception that the bound labeled proteins were pre-incubated with urokinase prior to elution. Lane 7 contains 1251-Lys-plasminogen incubated at 37*C without bacteria for the period of the experiments and Lanes 8 and 9 contain 1251-Lys-plasminogen and 125-Lys-plasmin respectively. For precise experimental details see Methods.






78












123 456 7 8 9






L-PLG




8M-LC







79

that the bacteria incubated with Lys-plasminogen exhibited only a low level of enzymatic activity (approximately 15% of that observed in the samples pre-incubated with Lys-plasmin). The bacterial free supernatant of the sample incubated with Lys-plasminogen demonstrated <10% of the zymogen remained in the supernatant as detected by enzymatic activity following activation with streptokinase (Figure 3-6). Taken together these results indicate that the Lys-plasminogen was removed from the fluid phase without prior or concomitant activation to Lys-plasmin. Furthermore, when the bacterial bound 1251-Lys-plasminogen was eluted from the bacteria and examined by polyacrylamide gel electrophoresis under reducing conditions, a single protein band was observed on the autoradiograph corresponding to the enzymatically inactive modified zymogen form of the protein, Lys plasminogen (Figure 3-7). All of these studies demonstrate that Lys-plasminogen can bind to bacteria without first being converted to Lys-plasmin. This would indicate that the intact native plasminogen molecule (Glu-plasminogen) does not express structures that are recognized by the bacterial plasmin receptor. However, following a conformational change achieved by either conversion to the Lys-plasminogen form of the zymogen or by activation to plasmin, structures are formed or exposed on the molecule that facilitate interaction with the bacteria.


Discussion


Plasmin is the key component of the mammalian fibrinolytic enzyme system which is responsible for fibrin degradation and intravascular blood clot lysis. Active plasmin, which cleaves fibrin, is derived from the circulating zymogen precursor Glu-plasminogen. Glu-plasminogen is a







80

single chain glycosylated protein containing 790 amino acids in known sequence with a molecular weight of approximately 92,000 daltons (Thorsen et al., 1981; Wiman, 1973, 1977). The generation of plasmin from plasminogen is accomplished by proteins known as plasminogen activators. This conversion is brought about by cleavage of a single arginine (560)valine (561) peptide bond which creates, through conformation changes, a two chain active plasmin molecule held together by disulfide linkages (Astrup, 1978). The light chain of plasmin has a molecular weight of approximately 25,000 daltons and contains the serine protease active site (Robbins and Summaria, 1970; Wiman, 1977). The heavy chain of plasmin has a molecular weight of approximately 63,000 daltons (Robbins and Summaria, 1970) and contains 5 homologous triple loop structures known as kringles (Sottrup-Jensen et al., 1978). An additional conformationally distinct form of plasminogen can be generated when Glu-plasminogen is exposed to plasmin. This removes a 76 amino acid peptide from the NH2terminus, thereby generating Lys-plasminogen (Swenson and Thorsen, 1981).

The plasmin(ogen) molecule contains several characteristic 'lysinebinding sites', one with high affinity for the lysine analogue EACA (dissociation constant of 9.0 pM), and four or five with low affinity (dissociation constant of 5 mM) (Markus et al., 1978a, 1978b). The high affinity site has been mapped to the kringle 1 region, and one of the lower affinity sites has been mapped to the kringle 4 region of the plasmin(ogen) molecule (Lerch et al., 1980). These structures are known to participate in the binding of plasmin(ogen) to a2-antiplasmin (a2-AP) (Wiman, 1981) and to fibrin (Swenson and Thorsen, 1981; Wiman et al., 1979) respectively. It is known that binding of lysine and lysine analogs to plasmin(ogen)'s lysine binding sites induces conformational







81

changes in the molecule (Violand et al., 1975). I have shown previously that lysine or a2-AP inhibit the binding of plasmin to the group A streptococcal receptor, as documented in Chapter Two, indicating the possible involvement of the high affinity lysine-binding site in the plasmin-bacterial receptor interaction. A comparison of my findings with studies of the interaction of plasminogen with other naturally occurring plasminogen binding proteins reveals a number of interesting similarities and contrasts. Specific binding to the group A streptococcus, 64/14, was demonstrated with plasmin's heavy chain. However, the isolated heavy chain alone was not as efficient a competitor as intact Lys-plasmin, as evidenced by the non-superimposible nature of the heavy chain and Lys-plasmin inhibition curves (Figure 3-2). It should be noted that 100% inhibition of binding of Lys-plasmin could be achieved by addition of high concentrations of heavy chain, but none of the kringle containing fragments (Lysine-binding domains) alone or in combination had any significant inhibitory effects at similar molar concentrations. This finding stresses the importance of the conformation of the entire heavy chain for binding to bacteria. The bacterial binding of plasmin therefore differs from the kind of interaction seen with a2-AP, as well as with fibrin and fibrinogen, to which plasmin as well as plasminogen, LBS-I, LBS-II, and miniplasmin(ogen) are known to interact (Swenson and Thorsen, 1981; Thorsen et al., 1981; Wiman et al., 1979).

Consistent with my initial observations, documented in Chapter Two, there is no significant binding of the native zymogen, Glu-plasminogen, while the conformationally altered form of the zymogen, Lys-plasminogen, was found to bind specifically to bacteria (Figures 3-3,3-4, and 3-5).







82

This form of the zymogen molecule is known to be conformationally distinct from Glu-plasminogen and is similar in conformation to Lysplasmin (Violand et al., 1975). The binding of Lys-plasminogen to the group A streptococcal receptor is therefore dependent on a specific conformation, most probably of the heavy chain.

Interaction of both plasmin and plasminogen with thrombospondin (an adhesive glycoprotein) has been demonstrated to occur via interaction with the heavy chain of the plasmin(ogen) molecule (Silverstein et al., 1984; Walz et al., 1987). However, efficient binding of thrombospondin with any elastase digestion fragment of plasminogen has not been observed (Walz et al., 1987). I have observed a similar pattern for the interaction of plasmin(ogen) with its bacterial receptor. However, unlike the interaction of thrombospondin with plasmin, the bacterial binding properties were reversible by addition of lysine or lysine analogs (Broeseker et al., 1988).

Histidine-rich glycoprotein (HRGP), an a2-glycoprotein in human plasma, has been reported to compete with a2-AP for the high-affinity lysine-binding site in plasmin (Haupt and Heinburger, 1972; Lijnen et al., 1980). In addition, HRGP also reduces the binding of plasminogen to fibrin by complex formation with the low-affinity lysine binding sites (Lijnen et al., 1980). Furthermore, the characteristic interaction of Glu-plasminogen, Lys-plasminogen, or plasmin and their fragments with fibrin or fibrinogen involves the heavy chain lysine-binding sites (Cenderholm-Williams, 1977). This is distinct from the profile of reactivity for the interaction of these proteins with the group A streptococcus (Table 3-1). It can be seen that Glu-plasminogen shows no reactivity with these bacteria, nor is there any significant reactivity with the isolated lysine-binding fragments LBS I or LBS II.







83

Of particular relevance to this study is the interaction of

plasmin(ogen) with the well characterized streptococcal plasminogen activator streptokinase isolated from group C streptococci (Christensen, 1945; Tillet and Garner, 1933). This secreted streptococcal protein is known to bind rapidly to Glu-plasminogen, Lys-plasminogen, and Lysplasmin (rate constant 5.4 x 107 M-1S-1) forming a 1:1 stoichiometric complex with an estimated dissociation constant of 5 x 10-11 M (Cenderholm-Williams et al., 1979). This interaction occurs via an interaction with the light chain (Summaria and Robbins, 1976). The interaction with the group A streptococcal plasmin receptor is distinct from group C streptokinase in that it does not recognize the Gluplasminogen molecule, and demonstrates no significant reactivity with the isolated light chain of plasmin. Furthermore, the plasmin(ogen)streptokinase complex cannot be dissociated by lysine or lysine analogs (Von-Mering et al., 1988), while the interaction of plasmin with a group A streptococcus is completely reversible by lysine or lysine analogs (Broeseker et al., 1988).

Taken together, these results indicate that the group A

streptococcal plasmin receptor binds in a unique manner to both plasmin and Lys-plasminogen. The predominant interaction is via determinants present on the intact heavy chain. These structures are present in their optimal binding configuration on the intact plasmin molecule and on the modified zymogen, Lys-plasminogen. The studies presented here suggest that the lysine binding sites themselves are not involved in direct interaction of plasmin with the bacteria (Figure 3-2). The observations that plasmin bound to bacteria retains its enzymatic activity for both small synthetic substrates and for fibrin, Chapter Two, are consistent with the observations that the light chain is not involved in binding.







84

The failure of a2-AP to regulate the bound enzyme suggests that the required interaction between a2-AP and plasmin is directly or indirectly inhibited. This may occur because one of the recognition sites for a2-AP in the kringle 1 region of plasmin's heavy chain may not be accessible when plasmin is bound to a streptococcus.

The characteristics of the interaction of human plasmin with the group A streptococcus, 64/14, described in this study indicate that the bacteria can capture a potent protease activity that cannot be regulated by the primary physiological inhibitor of plasmin, a2-AP. This group A streptococci also secretes a plasminogen activator and consequently, in the presence of plasminogen, the bacteria has the potential to both generate plasmin and bind the active enzyme to its surface (DesJardin et al., 1988). The importance of this selective receptor to the infectious disease process of receptor positive bacteria remains to be established.

The purpose of the series of studies described in the next chapter was to isolate and characterize the plasmin binding receptor from the strain 64/14 streptococcus.
















CHAPTER FOUR
ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN


Introduction


The studies presented thus far have documented the existence of a cell surface receptor for human plasmin on group A streptococcal strain 64/14. In addition to this plasmin binding activity, certain group A streptococci have long been known to secrete the plasmin(ogen) binding protein streptokinase, (Mr approx. 48,000 daltons), a non-enzymatic plasminogen activator. This protein, described by Tillet and Garner (1933), non-covalently associates with both plasminogen and plasmin, and was originally identified by virtue of its ability to generate fibrinolytic activity. Streptokinase binds rapidly to the native zymogen Glu-plasminogen (rate content 5.4 x 107 M-1S-1) forming, a 1:1 stoichiometric complex with an estimated dissociation constant of 5 x 10-11 M (Cederholm-Williams et al., 1979). The formation of a complex between streptokinase and plasminogen generates an enzymatic moiety capable of plasminogen activator activity, a property neither protein possesses alone.

The properties of the bacterial plasmin receptor reported thus far are markedly different from streptokinase. While the bacterial plasmin receptor binds preferentially to domains in the heavy chain of the plasmin molecule (see Chapter Three), streptokinase binds to plasmin's light chain (Summaria and Robbins, 1976). Furthermore, streptokinase



85







86

binds to both plasmin and the native zymogen Glu-plasminogen, while the surface associated plasmin receptor shows no significant reactivity for the native zymogen. Despite these clear functional differences, the expression of two proteins by the same bacteria that bind to the key human fibrinolytic protein plasmin with such selectivity raises the possibility that they may be in some way related or derived from a common precursor. Furthermore, the majority of information on the properties of streptokinase have been derived from studies of the plasminogen activator molecule isolated from group C streptococcal strains and evidence for differences in antigenicity and hence possibly function have been reported between streptokinase proteins isolated from group A and group C streptococcal isolates (Dillon and Wannamaker, 1965; Weinstein, 1953). The purpose of the studies presented in this chapter were to isolate a functionally active receptor for human plasmin from strain 64/14 and to compare it with the streptokinase protein that is produced by the same organism.


Materials and Methods


Enzymes. Inhibitors and other Reagents

Urokinase and porcine elastase (type IV) were obtained from Sigma Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from Mobay Pharmaceuticals, New York, NY. D-Val-Phe-Lys-chloromethyl ketone (VPLCK), and Phe-Pro-Arg-chloromethylketone (PPACK) were obtained from Calbiochem-Behring, San Diego, CA. Human Lys-plasminogen was obtained from American Diagnostica Inc., Greenwich, CT. H-D-Val-Leu-Lysparanitroanilide (S-2251) was obtained from Helena Laboratories, Beaumont, TX. Purified group C streptokinase was a gift from Kabivitrum, A.B., Stockholm, Sweden.







87

Bacteria

The Lancefield P hemolytic streptococcal strain 64/14 was grown as a stationary culture at 37*C, in one to two liter batches of a chemically defined media for streptococci described by Van de Rijn and Kessler (1980), containing 0.1% phenol red. The pH of the cultures were maintained at a pH greater than 7.0, as monitored by the indicator dye. For certain experiments, where noted, bacteria were grown in Todd-Hewitt broth (Difco Laboratories, Detroit, MI) Approximately 2.0 to 4.0 g (wet weight) of bacteria could be recovered per liter of media following a 24 to 36 hour incubation at 37*C. Bacteria were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4, containing 0.02% sodium azide. The bacteria were heat killed at 80*C for 15 min. The suspension was centrifuged and the pellet washed twice with PBS containing 0.02% sodium azide. Aliquots could be stored at 20*C, or used immediately for extraction purposes. Radioiodination of Proteins

Human plasminogen was iodinated by the chloramine T method using lodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell (1982). The labeled proteins were separated from free iodine by passage over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1% gelatin (VBS-gel). The labeled proteins were stored in aliquots containing 0.02% sodium azide at -20*C. Labeled aliquots were used once and discarded.

Generation of Plasmin

Plasmin was generated from either radiolabeled or unlabeled

plasminogen by reaction with urokinase. Three p1 of urokinase (Sigma 20




Full Text
40


51
migratory property (a shift to a lower Mr of approximately 85,000
daltons) in comparison to native Glu-plasminogen (Mr of approximately
92,000 daltons).
Iodination of Proteins
Glu- and Lys-plasminogen were iodinated by the chloramine T method
using Iodobeads (Pierce Chem. Co., Rockford, IL) as described by
Markwell (1982). The labeled proteins were separated from free iodine by
passage over a G-25 column (PD-10, Pharmacia) and collected in 0.15 M
Veronal buffered saline, pH 7.4, containing 0.001 M Mg++, 0.00015 M Ca++,
and 0.1% gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. The concentration of 1^1-
plasminogen was determined antigenically using a sandwich enzyme-linked
immunosorbent assay (ELISA) technique utilizing goat anti-human
plasminogen IgG fraction from Atlantic Antibodies, Scarborough, ME. This
assay could measure plasminogen reliably in the nanogram range.
Generation of Plasmin
Lys-plasmin was generated from radiolabeled or unlabeled Glu- or
Lys-plasminogen by incubation with urokinase (20 units/ml) in VBS-gel
(unless stated otherwise) that contained 0.04 M lysine. The conversion
of the single chain zymogen molecule to the two chain plasmin enzyme was
monitored on SDS-PAGE under reducing conditions as described previously
by Lottenberg et al., (1987). Conversion of the zymogen to the active
enzyme was maximal after 30 min incubation at 37C. Glu-plasmin was
generated by a similar procedure with the exception that a 10-fold molar
concentration of aprotinin relative to the Glu-plasminogen concentration
was added prior to addition of urokinase (Swenson and Thorsen, 1981).
Mini-plasmin was generated from mini-plasminogen using the same


Figure 2-7. Inhibition of binding of 125I-plasmin to the group A
streptococcal strain 64/14 in VBS-gel containing various concentrations
of epsilon-aminocaproic acid, lysine, and arginine: Measurements of
duplicate experiments were performed and the data are presented as the
mean the standard deviation. The percent inhibition of binding was
calculated by comparing with binding in VBS-gel alone. (O O)
epsilon- aminocaproic acid; (O O) lysine; (0---0) arginine.


145
12 3 12 3 12 3
KD
116-
84_ f
58- *
ABC


Figure 2-4. Binding of 125I-plasmin or 125I-plasminogen to the group A
streptococcal strain 64/14 as a function of pH: The data are presented
as the mean the standard deviation. Measurements of duplicate
experiments were performed and are expressed as the percent of total
counts offered (20,000 cpm) which were associated with the bacterial
pellet. (O O) 125I-plasmin; (O O) 125I-plasminogen.


89
aliquots of PBS-azide and vacuum drained. Extraction samples,
chromatography fractions or standards were loaded into wells in 50-200 pi
aliquots. Commercially available group C streptokinase (Kabikinase) was
used as a positive control in each assay. All wells were washed twice
with 200 pi aliquots of PBS-azide and vacuum drained, all samples were
assayed in duplicate.
Blots were removed from the apparatus and remaining sites on the
nitrocellulose were blocked by washing a total of four times in 200-250
ml of 5.0 mM sodium diethylbarbiturate, 0.14 M NaCl, 0.5% gelatin, 0.15%
Tween 20, 0.004% NaN3 pH 7.35 (blotting wash buffer I) for 15 minutes
per wash. At this point, blots could be probed as described below or
stored in the fourth wash overnight at 4-8C. If the latter procedure
was followed, blots were washed a fifth time after cold storage in 200-
250 ml blotting wash buffer I for 30 minutes. Results from the two
variations did not differ.
The individual blots were then probed for 3-4 hours at room
temperature while rotating in 10 ml aliquots of the following probing
solution: blotting wash buffer I containing 2.0 mM PMSF and l^I-labeled
plasmin at 2 x 10^-3 x 10^ cpm/ml. The probed blots were then washed
four times in 200-250 ml of 0.01 M EDTA, 0.5 M NaCl 0.25% gelatin, 0.15%
Tween 20, 0.004% NaN3 for 15 minutes per wash. All washing and probing
steps were carried out at ambient temperature. The probed, washed blots
were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.


120
protein in the concentrated culture supernatant of strain 64/14. Neither
the 48,000 dalton plasminogen activator protein present in the
concentrated culture supernatant of strain 64/14 (Figure 4-3), nor group
C streptokinase was recognized by this antibody (Figure 4-7).
Discussion
Group A streptococci have been recognized for many years to secrete
a protein, streptokinase, with a high affinity for both plasminogen and
plasmin (Tillett and Garner, 1933). The experiments documented in
Chapters Two and Three, have described a surface receptor on certain
group A streptococci that displays selective binding activity towards
plasmin, while having minimal reactivity with the zymogen form of the
molecule, Glu-plasminogen. The purpose of the experiments described in
this chapter were to isolate the plasmin receptor and compare it on a
functional basis to the secreted streptokinase protein produced by the
same group A streptococcal strain, 64/14.
A variety of different extraction techniques were compared and
treatment with mutanolysin yielded the highest quantity of soluble
plasmin binding activity. This activity was associated with a 41,000
dalton molecule by Western blot analysis, under both reducing and non
reducing conditions, and no evidence for subunit structure by
intramolecular disulfide bonds was observed. This plasmin binding
molecule was protein in nature, and was totally devoid of plasminogen
activator activity. The 41,000 dalton plasmin receptor protein was
purified from the mutanolysin extract of strain 64/14 by affinity
chromatography using enzymatically inactivated immobilized human plasmin.
Bound receptor activity on the column was specifically eluted with


Table 3-1.
Summary of inhibition experiments of PPACK reacted -Lvs-plasmin
binding to the group A streptococcal strain 64/14 are shown with a
schematic depiction of the portion of the native molecule they represent.**
COMPETIT OR Isq%( a*- m)
Glu-Plasminogen
X. X ,
Li,.I
> 2.0
Lys-Plasminogen
* -?9,
0.010
Li
Lys-Plasmin
x, ?90
L*- s s I
0.012
Mini-Plasminogen
> 1.0
442 -^r-g* 790
Mini-Plasmin
Val442 -X- -X- i 790
Is- S 1
> 1.0
LBS-I(K1 -Kg)
> 1.0
lbs-h(k4)
v.,
Va 354
> 1.0
Heavy (A) Chain
0.046
Light (B) Chain
Valgg j 1jfr -X- 1 790
> 1.0
| Activation cleavage site (arginine5gQ-valinegg .j).
* Plasmin active site residues (histidineg02; aspartic acidg45; serine74Q) from left to right.
Structure and NH2-amino-terminal residue data were obtained from the work
of Sottrup-Jensen et al.. (1978). Inhibition is expressed as 50% inhibitory
values in (/M) with Lys-plasmin as the standard (see Materials and Methods).
U)


7
been implicated in a variety of normal and abnormal processes which
involve the destruction or alteration of the extracellular environment,
such as tumor cell growth and invasiveness (for review of this extensive
literature see Dano et al., 1985), tissue remodeling, embryogenesis
(Beers et al., 1975), ovulation (Strickland and Beers, 1976), and
trophoblast implantation (Strickland et al., 1976). In fact, plasmin
exhibits broad substrate specificity and in addition to fibrin can
hydrolyze components of connective tissue and basement membranes such as
laminin, proteoglycans, fibronectin, thrombospondin, and type-V collagen,
as well as proteolytically activating other proteases (for review see
Knudsen et al.. 1986) and several plasma proteins (Marder et al., 1982).
Summary and Specific Aims
In the process of examining human serum from patients for antibody
reactivity directed against the streptococcal plasminogen activator
streptokinase, from patients who received thrombolytic therapy by
streptokinase administration, an interesting observation was made by Dr.
M.D.P. Boyle and Dr. R. Lottenberg. In experiments which involved immune
precipitations using heat killed, Fc-receptor expressing, group C
streptococci it was found that control tests involving incubations of the
"IOC
radiolabeled tracer (-plasmin) and bacteria revealed an association
of plasmin to bacteria in the absence of any added antibody. Testing
other groups of streptococci showed that the group A streptococci
displayed the highest level of plasmin binding activity, while
demonstrating little binding activity for preparations of 12f>I-
plasminogen.


10
100 (Lottenberg et al.. 1985). Plasminogen was quantified by measuring
absorbance using a value of 17.0 (Nilsson e£ al. 1982).
Enzymes, Inhibitors and other Reagents
The enzymes urokinase and trypsin were obtained from the Sigma
Chemical Company, St. Louis, Mo.; Aprotinin was obtained as Trasylol
from Mobay Pharmaceuticals, New York, New York. Phe-pro-arg
chloromethylketone (PPACK) was obtained from Cal-Biochem (San Diego, Ca.)
P-nitrophenyl, p-guanidobenzoate HC1 (pNpGB) was obtained from Sigma
Chemical Co, St. Louis, Mo.; human a2"anti.p]-asini.n (c*2-AP) was obtained
from American Diagnostica Inc., Greenwich, Connecticut. H-D-Val-leu-lys-
paranitroanilide (S-2251) was obtained from Helena Chemical Co.,
Beaumont, Texas.
Radioiodination of Proteins
Human plasminogen, urokinase, and trypsin were iodinated by a mild
lactoperoxidase method using Enzymobeads (Bio-rad Laboratories Richmond,
Calif.) as described by Reis et al., (1983). The labeled proteins were
separated from free iodine by passage over a G25 column (PD-10 Pharmacia)
and collected in 0.15 M Veronal buffered saline pH 7.35 containing 0.001
M Mg++, 0.00015 M Ca++ and 0.1% gelatin (VBS-gel). The labeled proteins
were stored in aliquots containing 0.02% sodium azide at -20C. Labeled
aliquots were used once and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three /xl of urokinase (Sigma 20
u/ml) was added to a 400 /I solution of 1.0 /xM plasminogen containing
0.04 M lysine. The mixture was incubated at 37C for 45 minutes unless
stated otherwise. The efficiency of plasmin generation was followed by


125
heavy and light chains preparations were examined by SDS-PAGE and
autoradiography in a similar manner.
Polyacrylamide Gel Electrophoresis and Protein Blotting
Electrophoresis was carried out as described by Laemmli (1970).
Polyacrylamide separating gels were 10% and contained 0.1% sodium
dodecylsulfate (SDS), 0.375 M Tris at pH 8.8. Stacking gels were 4% and
contained 0.1% SDS and 0.125 M Tris at pH 6.8. Electrode buffer was
0.024 M Tris, 0.192 M glycine, 0.1% SDS at pH 8.3 Samples were diluted
1:2 with sample buffer containing 0.125 M Tris pH 6.8, 4% SDS, 20%
glycerol, 10% ^3-mercaptoethanol and 0.05% bromophenol blue and heated at
80-90C for 3 minutes. Gels were run at 45 volts constant voltage for
approximately 15-18 hours. Slab gels were used in the Bio-Rad Protean II
system (BioRad, Richmond, CA). Molecular weight markers were run on all
gels. Gels intended for Western blot transfer contained pre-stained
markers (Sigma) applied as a mixture which included: triosephosphate
isomerase (26,600), lactic dehydrogenase (36,500), fumarase (48,500),
pyruvate kinase (58,000), fructase-6-phosphate kinase (84,000), /9-
galactosidase (116,000), and alpha 2-macroglobulin (180,000). After
electrophoresis, gels intended for Western blotting were equilibrated in
25 mM Tris, 0.2 M glycine pH 8.0 containing 20% v/v methanol (electroblot
buffer) for 25 minutes. Protein blotting, from SDS-PAGE gels, was
performed using the 'Trans-Blot SD Semi-Dry' electrophoretic transfer
cell (Bio Rad, Richmond, CA). Nitrocellulose transfer medium, also
equilibrated in electroblot buffer, was sandwiched between the gel and
two sheets of Whatman 3 mm paper. The gel was also backed with two
sheets of 3 mm paper. Blots were blocked by washing a total of four
times in 200-250 ml of 5.0 mM sodium diethylbarbiturate, 0.14 M NaCl,


80
single chain glycosylated protein containing 790 amino acids in known
sequence with a molecular weight of approximately 92,000 daltons (Thorsen
et al.. 1981; Wiman, 1973, 1977). The generation of plasmin from
plasminogen is accomplished by proteins known as plasminogen activators.
This conversion is brought about by cleavage of a single arginine (560)-
valine (561) peptide bond which creates, through conformation changes, a
two chain active plasmin molecule held together by disulfide linkages
(Astrup, 1978). The light chain of plasmin has a molecular weight of
approximately 25,000 daltons and contains the serine protease active site
(Robbins and Summaria, 1970; Wiman, 1977). The heavy chain of plasmin
has a molecular weight of approximately 63,000 daltons (Robbins and
Summaria, 1970) and contains 5 homologous triple loop structures known as
kringles (Sottrup-Jensen et al., 1978). An additional conformationally
distinct form of plasminogen can be generated when Glu-plasminogen is
exposed to plasmin. This removes a 76 amino acid peptide from the NH2-
terminus, thereby generating Lys-plasminogen (Swenson and Thorsen, 1981).
The plasmin(ogen) molecule contains several characteristic 'lysine
binding sites', one with high affinity for the lysine analogue EACA
(dissociation constant of 9.0 fiH), and four or five with low affinity
(dissociation constant of 5 mM) (Markus et al., 1978a, 1978b). The high
affinity site has been mapped to the kringle 1 region, and one of the
lower affinity sites has been mapped to the kringle 4 region of the
plasmin(ogen) molecule (Lerch et al., 1980). These structures are known
to participate in the binding of plasmin(ogen) to a^-antiplasmin (c^'AP)
(Wiman, 1981) and to fibrin (Swenson and Thorsen, 1981; Wiman et al.,
1979) respectively. It is known that binding of lysine and lysine
analogs to plasmin(ogen)'s lysine binding sites induces conformational


172
Weber, K., and M. Osborn. 1969. The reliability of molecular weight
determinations by Dodecyl Sulfate-polyacrylamide gel
electrophoresis. J. Biol. Chem. 244:4406-4412.
Weinstein, L. 1953. Antigenic dissimilarity of streptokinases. Proc.
Soc. Exp. Biol. Med. 83:689-691.
Werb, Z., C.L. Mainardi, C.A. Vater, and E.D. Harris. 1977. Endogenous
activation of latent collagenase by rheumatoid synovial cells:
Evidence for a role of plasminogen activator. N. Engl. J. Med.
296:1017-1023.
Wiman, B. 1973. Primary structure of peptides released during
activation of human plasminogen by urokinase. Eur. J. Biochem.
39:1-9.
Wiman, B. 1977. Primary structure of the B-chain of human plasmin.
Eur. J. Biochem. 76:129-137.
Wiman, B. 1981. Human a^-Antiplasmin. Methods Enzymol. 80:395-408.
Wiman, B., L. Bowman, and D. Collen. 1978. On the kinetics of the
reaction between human antiplasmin and a low- molecular-weight form
of plasmin. Eur. J. Biochem. 87:143-146.
Wiman, B., and D. Collen. 1977. Purification and characterization of
human antiplasmin, the fast-acting plasmin inhibitor in plasma.
Eur. J. Biochem. 78:19-26.
Wiman, B., and D. Collen. 1978. On the kinetics of the reaction between
human antiplasmin and plasmin. Eur. J. Biochem. 84:573-578.
Wiman, B., H.R. Lijnen, and D. Collen. 1979. On the specific
interaction between the lysine-binding sites in plasmin and
complementary sites in a^-antiplasmin and in fibrinogen. Biochem.
Biophys. Acta. 579:142-154.
Wiman, B., and P. Walln. 1973. Activation of human plasminogen by an
insoluble derivative of urokinase. Eur. J. Biochem. 36:25-31.
Wiman, B., and P. Walln. 1975. Structural relationship between
"glutamic acid-*- and lysine forms of human plasminogen and their
interaction with the NH2-terminal alteration peptide as studied by
affinity chromatography. Eur. J. Biochem. 50:489-494.
Winn, E.S., S.P. Hu, S.M. Hochschwender, and R.A. Laursen. 1980.
Studies on the lysine-binding sites of human plasminogen: The
effect of ligand structure on the binding of lysine analogs to
plasminogen. Eur. J. Biochem. 104:579-586.
Woodard, L.F., and R.L. Jasman. 1985. Stable oil-in-water emulsions:
preparation and uses as vaccine vehicles for lipophilic adjuvants.
Vaccine 3:137-144.


146
Figure 5-4 shows the reactivity of the rabbit polyclonal anti-plasmin
receptor protein for comparison. These data clearly demonstrate that
streptokinase and the bacterial plasmin receptor do not contain any
common immunodominant epitopes.
The two bacterial plasmin binding proteins were compared for any
evidence of antigenic relatedness using sixteen mouse monoclonal
antibodies, prepared against group C streptokinase. These experiments
were carried out by Western blot analysis. The results of these
experiments are summarized in Figure 5-5. All the monoclonal antibodies
tested reacted efficiently with group C streptokinase as expected (see
lane 3 of each Western blot). Thirteen of the sixteen monoclonal
antibodies reacted equally well with the streptokinase protein from
strain 64/14 present in the concentrated culture supernatant see lane 2
of each Western blot. However none of the monoclonal antibodies examined
reacted significantly with the extracted 41,000 dalton plasmin receptor,
see lane 1 of each Western blot.
These studies demonstrated certain unique epitopes present on group
C streptokinase molecules that are not present on the group A plasminogen
activator molecule. Overall the group A and group C streptokinase
proteins were found to be antigenically closely related, while the group
A plasmin receptor was totally devoid of any of the antigenic
determinants found on streptokinase.
Discussion
The purpose of this investigation was to compare the streptococcal
strain 64/14 receptor for human plasmin with the well characterized
secreted plasmin(ogen) binding streptococcal protein streptokinase, to


Page
Figure
3-4. Inhibition of, PPACK reacted, '^^I-Lys-plasmin
binding to group A plasmin receptor 69
3-5. Binding of Lys-plasmin(s), derived from Glu-
plasminogen and Lys-plasminogen, to the group A
streptococcal strain 64/14 as measured by residual
activity in the bacterial free supernatant 72
3-6. Binding of Glu- and Lys-plasminogen to the group A
streptococcal strain 64/14 as measured by residual
activatable zymogen in the bacterial free
supernatant 76
3-7. Characterization of -plasmin(ogen) species eluted
from bacteria 78
4-1. Dot-blot analysis of solubilized plasmin binding
activities 102
4-2. SDS-PAGE and Western blot analysis of mutanolysin
extracted 64/14 bacterial plasmin binding activity.. 104
4-3. Solid-phase plasminogen activation assay 109
4-4. Representative profile of an affinity purification
of strain 64/14 mutanolysin extracted plasmin
binding activity 112
4-5. Analysis of affinity purified plasmin binding material
from the strain 64/14 mutanolysin extract 114
4-6. SDS-PAGE and Western blot analysis of mutanolysin
extracted, affinity purified plasmin binding
activity 117
4-7. SDS-PAGE and Western blot analysis of plasmin
receptor protein with a polyclonal rabbit antibody.. 119
5-1. Functional identification and distinction of
streptokinase proteins and plasmin binding receptor
protein 134
5-2. Comparison of binding reactivities of streptokinase
proteins and plasmin binding receptor protein with
^*1 -plasmin heavy chain and ^-^^1-plasmin light
chain 138
5-3. Analysis of the antigenic relationship of the 64/14
plasmin receptor and streptokinase proteins 141
viii


ACKNOWLEDGMENTS
I wish to express my sincere thanks to Dr. Michael D.P. Boyle for
giving me the opportunity to work in his laboratory, for his support and
guidance, and especially for his patience. It has been a pleasure to
work with Mike.
I also wish to give a special thanks to Dr. Richard Lottenberg for
all of his help, guidance, and friendship.
I would like to thank the other members of my committee, Drs. R.W.
Moyer and J.W. Shands, for their helpful suggestions throughout this
study.
I would also like to offer a special thanks to Dr. Kenneth H.
Johnston, my outside examiner, for sending me the monoclonal antibodies,
and the solid-phase plasminogen activator assay mentioned in this study
and for taking the time to review and discuss my work.
I would also like to express my appreciation to all the people with
whom I have worked for the past four years, especially Jeannine Brady,
Greg VonMerring, Tim Broeseker, and Lucy DesJardin.
I also offer most special thanks to my parents, Jeanne C. and
Thomas J. Broder, for their never-ending love and support throughout all
my endeavors. I also thank all my family, especially my brother
Michael.
Finally, I offer my sincere thanks to my wife, Colleen, for all of
her unselfish support, which has been essential for me in pursuing my
goals.
iii


Figure 3-4. Inhibition of. PPACK reacted, 125I-lvs-plasmin binding to
group A plasmin receptor:A constant concentration of (1 x 10'10 M)
PPACK reacted ^-Lys-plasmin and an increasing concentration range of
Lys-PLG, Glu-PLG or Lys-PLA (1010 M to 10"6 M) were mixed with a fixed
concentration of the streptococcal strain 64/14. Following incubation
and washing (see Materials and Methods), the amount of radiolabeled Lys-
plasmin bound to the bacterial pellet was determined. The quantity of
radioactivity bound in the presence of unlabeled competitor was compared
to the radioactivity bound in the absence of inhibitor and the percent
inhibition calculated. ( O-Lys-PLA; -Lys-PLG; Glu-PLG).


99
in the gel, and by Western blotting a small strip of the gel to
nitrocellulose, followed by blocking and probing with ^^1-plasmin, and
autoradiography as described in the polyacrylamide gel electrophoreses
and protein blotting section of the Methods. The stained 41,000 dalton
band was cut from the gel, and a portion containing approximately 300 ig
was emulsified with an equal volume of Freund's complete adjuvant. The
emulsion was injected subcutaneously at 6 sites on a rabbit. The rabbit
was boosted eight times with the 41,000 dalton protein-polyacrylamide gel
emulsified in Freund's incomplete adjuvant (approximately 200 /g per
boost) during a 14 month period. Pre-immune and immune IgG fractions
were prepared from rabbit sera by Protein A-Sepharose (Sigma) affinity
chromatography.
Results
A variety of extraction procedures were compared and were to
determine the optimal method for solubilizing functional plasmin receptor
activity. The extractions were carried out on heat killed 64/14 as
described in detail in the Methods section. The bacterial samples were
washed thoroughly prior to treatment to minimize carry over of culture
media and secreted products to the extraction samples. This would
therefore reduce the likelihood of significant streptokinase
contamination (with the possible exception of intracellular forms).
The extraction techniques included: (1) Lancefield acid and
alkaline extractions; (2) a time course trypsin digestion under
suboptimal conditions for enzyme activity (conditions previously shown to
maximize homogeneity of type III Fc receptor extraction, Reis et al.,
1985); (3) Triton X-100/osmotic shock/lysozyme treatment; (4)


48
number, the effect of phase variations in the expression of different
proteins by bacteria (Cleary e£ al., 1987), and the heterogeneity in
receptor expression among colonies (Yarnall et al., 1984), we believe
that such differences need to be cautiously interpreted. It does not
appear from the results of these studies that the degree of plasmin
receptor expression correlates with the virulence of these group A
isolates in mice. Nonetheless we calculate that binding of active
plasmin in the picomolar range with any of the group A isolates studied
is achievable. The high affinity for and slow off rate of bound plasmin
may make these interactions with streptococci of importance in the
infectious process.
The next series of studies, described in the following chapter, were
designed to analyze the way in which the bacterial receptor associated
with its ligand, the human plasmin molecule.


LIST OF TABLES
Table Page
2-1. Binding of radiolabeled proteins to various nephri-
togenic and non-nephritogenic group A
streptococci 19
2-2. Ability of bacterial bound plasmin to solubilize a
fibrin clot 29
3-1. Summary of inhibition experiments 73
3-2. Measurement of plasmin(ogen) associated with
bacterial pellets 74
5-1. Fluid-phase plasminogen activator activity assay.... 132


83
Of particular relevance to this study is the interaction of
plasmin(ogen) with the well characterized streptococcal plasminogen
activator streptokinase isolated from group C streptococci (Christensen,
1945; Tillet and Garner, 1933). This secreted streptococcal protein is
known to bind rapidly to Glu-plasminogen, Lys-plasminogen, and Lys-
plasmin (rate constant 5.4 x 10^ forming a 1:1 stoichiometric
complex with an estimated dissociation constant of 5 x 10"^ M
(Cenderholm-Williams et al., 1979). This interaction occurs via an
interaction with the light chain (Summaria and Robbins, 1976). The
interaction with the group A streptococcal plasmin receptor is distinct
from group C streptokinase in that it does not recognize the Glu-
plasminogen molecule, and demonstrates no significant reactivity with
the isolated light chain of plasmin. Furthermore, the plasmin(ogen)-
streptokinase complex cannot be dissociated by lysine or lysine analogs
(Von-Mering et al., 1988), while the interaction of plasmin with a group
A streptococcus is completely reversible by lysine or lysine analogs
(Broeseker et al., 1988).
Taken together, these results indicate that the group A
streptococcal plasmin receptor binds in a unique manner to both plasmin
and Lys-plasminogen. The predominant interaction is via determinants
present on the intact heavy chain. These structures are present in their
optimal binding configuration on the intact plasmin molecule and on the
modified zymogen, Lys-plasminogen. The studies presented here suggest
that the lysine binding sites themselves are not involved in direct
interaction of plasmin with the bacteria (Figure 3-2). The observations
that plasmin bound to bacteria retains its enzymatic activity for both
small synthetic substrates and for fibrin, Chapter Two, are consistent
with the observations that the light chain is not involved in binding.


32
equilibrated in NaCl solutions of varying ionic strength before the
addition of bacteria. Following an incubation period of 15 minutes at
37C the bacteria were washed with solutions containing the appropriate
concentration of NaCl and the number of counts associated with the
bacteria determined. The results in Figure 2-5 demonstrate that plasmin
binding was dependent on ionic strength and that optimal binding occurred
between 0.1 and 0.4 M NaCl. In this range of salt concentrations, less
than 10% of plasminogen bound to bacteria. As the ionic strength was
lowered below 0.075 M NaCl, significant binding of plasminogen to the
bacteria was observed.
Binding of labeled plasmin to the group A streptococcal strain 64/14
was examined in the presence and absence of divalent cations to determine
if these metal ions were important for plasmin binding. Binding studies
were carried out in VBS-gel at pH 7.4 containing 0.00015 M Ca++ and 0.001
M Mg++ or in metal free VBS-EDTA-gel at pH 7.4. After incubation at
37C for 15 minutes, the bacteria were washed twice with the appropriate
buffer and radioactivity associated with the bacterial pellets was
measured. The amount of plasmin bound by the bacteria was the same in
the presence or absence of divalent cations, (data not shown).
After identification of the optimal binding conditions for the
plasmin:bacterium interaction, the affinity of the plasmin receptor for
its ligand was determined in 0.15 M VBS-gel buffer at pH 7.4. In the
initial studies the group A /S-hemolytic strain 64/14 was used.
Preliminary kinetic studies were conducted to establish first that
equilibrium between bacterial bound and free plasmin had been achieved,
and second the conditions under which saturation of bacterial plasmin
receptors could be demonstrated. Binding equilibrium was found to be


136
surface of the strain 64/14 streptococci possesses a plasmin binding
activity which appears to be associated with determinants present in the
intact heavy chain of plasmin in a conformationally dependent manner. I
therefore examined the binding specificities for ^^1-labeled plasmin
heavy chain and ^^I-labeled plasmin Light chain, by Western blot
analysis, of the extracted plasmin receptor and streptokinase from strain
64/14, as well as the other streptokinases. The results presented in
Figure 5-2 depict the binding reactivities of extracted plasmin receptor
and the streptokinases from the strains of streptococci described above
1-2^1-plasmin heavy chain Panel A, and -plasmin light chain Panel B.
The data clearly indicates that the group C streptokinase (Kabikinase),
in agreement with earlier findings (Summaria and Robbins, 1976) has a
much stronger reactivity with plasmin light chain than with plasmin heavy
chain. The streptokinase from the group C strain ATCC 12449 also shows
preferential reactivity for plasmin light chain. However, streptokinase
proteins associated with 26RP66 and all three of the group A strains
demonstrated equivalent reactivity with plasmin heavy chain and light
chain. The reactivity of the plasmin receptor for either heavy chain or
light chain was relatively weak. This binding is consistent with my
earlier observations (Chapter Three) which indicated that for optimal
binding reactivity of the bacterial plasmin receptor there appeared to be
a requirement for a specific conformation that was best represented in
the intact plasmin molecule or the intact, conformationally modified,
zymogen, Lys-plasminogen.
These results and the previous findings (Chapters Two, Three, and
Four) provide evidence that the plasmin receptor and streptokinase are
physicochemically and functionally distinct molecules. The


19
Table 2-1.
Binding of radiolabeled proteins to various nephrltogenic
and non-nephritogenic group A streptococci.
STRAIN
M-TYPE
PLASMINOGEN
PLASMIN
UROKINASE
TRYPSIN
A992*
18
-
+
-
-
B923
12
-
+
-
-
D897*
12
-
+
-
-
B512
4
-
+
-
-
B438
18
-
+
-
-
B512
NT
-
+
-
-
A928
55
-
+
-
-
64/14
NT
-
++
-
-
B905
2
-
+
-
-
B281
12
-
+
-
-
B920
49
-
++
-
-
B915
49
-
+
-
-
A374
12
-
+
-
-
B931*
2
-
+
-
-
A207
2
-
+
-
-
F2030
1
-
+
-
-
A547
NT
-
+
-
-
64/P
NT
-
++
-
-
648
1
-
+
-
-
A995
57
-
+
-
-
* Non-nephritis causing strains
- = Less than 10% bound of total counts offered
+ 10% to 30% bound of total counts offered
++ = Greater than 30% bound of total counts offered
Approximately 3 X 10 bacteria/tube heat killed at 80C for 10 min.
NT = Not typable


154
Streptococci and streptococcal products have been known to interact
with the fibrinolytic system for many years (Tillett and Sherry, 1949).
The secreted streptococcal plasminogen activator, streptokinase, was
identified by Tillett and Garner in 1933. This protein is known to bind
to both human plasminogen and plasmin with high affinity (Reddy and
Markus, 1972), through interactions with the light chain of the plasmin
molecule (Summaria and Robbins, 1976). The formation of this
streptokinase-plasmin(ogen) complex generates an enzymatic moiety capable
of plasminogen activator activity, a property neither protein alone
possesses.
In the studies documented in Chapter Two, I have identified and
characterized a group A streptococcal surface receptor that binds human
plasmin while demonstrating no significant affinity for the zymogen form
of the molecule Glu-plasminogen. The expression of this binding activity
was seen with bacteria grown in either Todd-Hewitt broth or chemically
defined media. The binding of plasmin to its bacterial receptor did not
inhibit its enzymatic activity. In fact, bacterial bound plasmin was
shown to be capable of cleaving both synthetic substrates, as well as its
natural substrate, fibrin. The bacterial bound plasmin was also shown to
be fully accessible to small protease inhibitors; specifically,
aprotinin, PPACK, and pNpGB. Together, these data suggested that the
bound plasmin molecule interacts with the surface of the bacteria in a
fashion which leaves the active site accessible to substrates. Of great
interest was the observation that the bacterial bound plasmin enzyme was
not capable of being inhibited by plasmin's main physiological inhibitor
c^-antiplasmin which may have important implications in this bacteria's
pathogenic mechanisms.


This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
May, 1989
Dean, College of Medicine
Dean, Graduate School


Figure 2-2. Effect of inhibiting the active site of plasmin on its
ability to bind to the group A streptococcal strain 64/14: The lower
panel demonstrates the binding of the group A streptococcus to labeled
plasminogen, plasmin, plasmin pretreated with excess a2PI, plasmin
treated with excess pNpGB, plasmin treated with excess aprotinin or
plasmin treated with excess phe-pro-arg chloromethyl ketone,[PPACK], as
described in the Methods. The data are presented as the mean the
standard deviation of duplicate experiments. The upper panel
demonstrates the analysis of each of the plasmin-inhibitor samples that
had been incubated with excess a2PI. Samples were monitored on non
reducing SDS-polyacrylamide gels for the formation of a high molecular
weight complex.


43
Discussion
Plasminogen, an inactive zymogen can be converted to the protease
plasmin by a variety of plasminogen activators (Collen, 1980). This
enzyme demonstrates broad substrate specificity. In addition to fibrin
cleavage, plasmin can activate the first component of the classical
complement pathway, hydrolyze coagulation factors, degrade components of
basement membrane, and break down connective tissue (Atichartakarn et
al. 1978; Jones and DeClerck, 1980; Liotta et al.. 1981). Furthermore a
variety of potent split products are generated as a consequence of
plasmin activity, e.g. chemotactic fibrinopeptides (Kay et al., 1974).
Effective regulation of plasmin activity is therefore important in order
to prevent tissue damage and inflammation. Normally the selective
protease inhibitor a^-antiplasmin regulates plasmin activity in man (Aoki
et al., 1977).
Interaction of streptococci and streptococcal products with the
fibrinolytic system has been recognized for many years (Tillett and
Sherry, 1949). The observation that certain streptococci could lyse a
fibrin clot lead to the identification and isolation of streptokinase.
This secreted protein is known to bind to human plasminogen and plasmin
with a similar affinity (Reddey and Markus, 1972). In this study I have
identified a surface receptor on certain group A streptococci, grown in
either Todd-Hewitt broth or chemically defined media, that specifically
binds to plasmin while demonstrating no significant affinity for the
zymogen form of the molecule, plasminogen. Thus the surface receptor we
have identified is distinct from streptokinase. Furthermore, this
binding phenomenon did not appear to be simply a function of the ligand
being a serine protease, since no binding activity was demonstrated with
the two other serine proteases examined, trypsin and urokinase.


8
The ability of certain group A streptococci to produce a plasminogen
activator (e.g. streptokinase) and also to express a receptor for the
activation product plasmin, may contribute to the invasive properties of
these bacteria. This study has been designed to characterize this
plasmin receptor phenomenon more completely in order to increase
understanding of any potential role in bacterial pathogenesis.
The specific aims of the study are to
1. Identify and characterize a group A streptococcal
receptor for human plasmin (Chapter Two).
2. Map the binding site on human plasmin recognized by the
bacterial plasmin receptor (Chapter Three).
3. Isolate and purify a functionally active group A
streptococcal plasmin receptor (Chapter Four).
4. Compare the group A streptococcal receptor for human
plasmin with streptokinase, (Chapter Five), with
respect to
a. Plasminogen activator activity.
b. Plasmin(ogen) binding domain specificity.
c. Antigenic relatedness.


34


2
These modified forms of plasminogen have lysine, valine, or methionine as
their N-terminal amino acid (Walln and Wiman, 1970, 1972). These
modifications occur by the hydrolysis of the ArggyMetgg, Lysyg-Lysyy, or
Lys77-Val7g peptide bonds. The generation of plasmin from plasminogen
occurs through the specific cleavage of a single Arg-Val bond
corresponding to the Arg550'Val5g^ positions (Robbins et al., 1967).
This cleavage generates a two chain active plasmin molecule consisting of
a heavy chain and light chain held together by disulfide linkages
(Groskopf et al., 1969; Wiman, 1973) (see Figure 1-1). The light chain
of plasmin has a molecular weight of approximately 25,000 daltons
(Robbins and Summaria, 1970; Wiman, 1977) and contains the serine active
site. The heavy chain of plasmin has a molecular weight of approximately
63,000 daltons (Robbins and Summaria, 1970), and amino acid sequencing
revealed a structure containing 5 homologous triple loop structures known
as kringles (Sottrup-Jensen et al., 1978) .
Several specific compounds such as lysine, e-aminocaproic acid
(EACA), trans-4-aminomethycyclohexanecarboxylic acid (transexamic acid),
and C-terminal lysine peptides bind to certain sites on the plasmin(ogen)
molecule. These specific sites are the characteristic 'lysine-binding
sites' distinct from the catalytic site (Thorsen, 1975). These compounds
affect the properties of plasminogen and plasmin, and play an important
role in determining this zymogen-enzyme system's physiological
specificity. Chapter Three will go into more detail on the nature of
plasmin(ogen)'s lysine binding sites. Affinity chromatography of
defined fragments of plasminogen on lysine Sepharose has demonstrated
that these 'lysine-binding sites' are located in the portion of the
plasmin molecule which becomes the heavy chain upon activation. The


115
The activity eluted in three 1.0 ml fractions (Lanes 3-5 of Figure 4-6
Panel A). The greatest activity was found in fraction number two,
corresponding to lane four of Figure 4-6, Panel A, and 50 /I of this
fraction was analyzed by Western blotting. Figure 4-6, Panel B, is an
autoradiograph of Western blotted extracted plasmin receptor preparation
(Lane 1) and plasmin affinity purified receptor (Lane 3), which
demonstrates that the 41,000 dalton molecule has retained functional
plasmin binding activity following affinity purification. Treatment of
the affinity purified material with trypsin destroys the ability of the
41,000 dalton molecule to bind plasmin and results in the disappearance
of the 41,000 dalton stained band on SDS-polyacrylamide gel. These
results clearly indicate the purification of a plasmin binding activity
from a mutanolysin extract of the streptococcal strain 64/14 bacteria by
means of affinity chromatography. Taken together these results indicate
that the extracted surface receptor for human plasmin, and streptokinase
produced by the strain 64/14 streptococcus, are physicochemically
(molecular weight) and functionally (plasminogen activator activity)
distinct molecules.
The isolated 41,000 dalton plasmin receptor protein was used to
immunize a rabbit as described in the Methods. The resulting antibody
was used to probe both the mutanolysin extract of strain 64/14,
concentrated culture supernatant of strain 64/14, and group C
streptokinase (Kabikinase) using a sandwich Western blot, and the results
are shown in Figure 4-7.
Specific antigen-antibody complexes on the nitrocellulose were
detected by probing with *^1-Protein G. This antibody recognized only
the 41,000 dalton band in the mutanolysin extract and a corresponding


97
generating a standard curve), or the samples to be tested were placed
into the microtiter wells in duplicate. To the 20 /il aliquots, 40 /il of
50 mM Tris, pH 7.5, was added. 30 /I of a freshly prepared solution of
human Glu-plasminogen, 20 /tg/ml in 0.01 mM Triton X-100, was then added
to the well and allowed to incubate at 37C for 15 min. 30 /I of
substrate was then added. Substrate is prepared as follows: To 1 volume
substrate (5 mg/ml S-2251 in water) add 3 volumes 1.77 M NaCl in 0.32 M
Tris, pH 7.5, and one volume of water. The plate was then incubated at
37C to allow substrate hydrolysis, and product production is measured at
405 nm. Control wells from which either SK and/or plasminogen were
omitted were included. These controls will indicate whether there is any
plasmin contamination in the plasminogen preparation or if the sample
being tested has any proteolytic activity for the substrate that is not
dependent on plasminogen activation.
Solid Phase Assay for Plasminogen Activators
Samples to be tested for plasminogen activator activity by this
assay (Dr. K. Johnston, personal communication) were first resolved by
SDS-PAGE and transferred to nitrocellulose. The nitrocellulose membranes
were then immersed in blocking buffer (10 mM Tris, pH 8.0 containing 0.5%
Tween-20, 0.5 M NaCl and 1.0% bovine serum albumin) for at least one hour
at room temperature. The substrate overlay is prepared as follows: To a
2.0% agarose solution (Bio Rad Richmond, CA) in 0.15 M phosphate buffered
saline, pH 7.5 was equilibrated at 50C, with the chromogenic substrate
S-2251 at a concentration of 100 /ig/ml. Human plasminogen free of
plasmin activity, was then added to a final concentration of 20 /ig/ml.
The agarose-substrate-plasminogen solution was then applied to an ethanol
washed glass slide slightly larger than the nitrocellulose membrane


58
to yield a final concentration of 300 /M. The resuspended bacterial
pellets were then incubated at 37C for 20 minutes and quenched with 50
pi of glacial acetic acid. The bacteria were removed by centrifugation
(12,000 x g for 4 minutes) and the optical density of the bacterial free
supernatant was measured at 405 nm. Control tubes containing bacteria
and substrate, and substrate in buffer were included. All assays were
performed in duplicate.
Results
The experiments described in this chapter were designed to map the
domains on the human plasmin molecule involved in the high affinity
interaction with the group A streptococcal strain 64/14. For these
studies a variety of defined plasminogen fragments as well as the heavy
and light chains of plasmin were prepared as described in the Methods.
The plasminogen fragments obtained were characterized on urea gels, see
Figure 3-1. The homogeneous plasminogen fragments were used to compete
with intact PPACK reacted ^^^1-plasmin for receptor sites on the group A
streptococcal strain 64/14 (Table 3-1). I have previously demonstrated
that plasmin treated with PPACK, p-Nitrophenyl-p-guanidinobenzoate
(pNpGB), or aprotinin, does not effect plasmin's binding reactivity to
the group A streptococcal strain 64/14 as documented in Chapter Two.
Enzymatic inhibition of the ^i-Lys-plasmin and urokinase in the labeled
tracer preparation was necessary to prevent the proteolytic conversion of
Glu-plasminogen to Lys-plasminogen, or Lys-plasmin (Markus et al., 1978)
by the labeled tracer mixture. In all the competitive inhibition
experiments described in this study, a constant concentration of PPACK-
plasmin (1.0 x 10'^- M) and a range of concentrations of unlabeled


123 4 567123 4567
12 3 4 5 6
KD
I I 6-
84-
5 8-
4 8.5-
36.5-
26.6-i

I I 6-
8 4-
5 8-
3 6.5-
26.6-1
B C
134


121
L-lysine or EACA at the concentrations which reversibly inhibit plasmin
binding to this streptococci (Chapter Two). The affinity purified 41,000
dalton protein was demonstrated to retain functional activity by Western
1 9 S
blot analysis and probing with plasmin.
The secreted protein, streptokinase, identified in the concentrated
culture supernatant of strain 64/14 (Mr approximately 48,000) has the
ability to bind plasminogen and once complexed to the zymogen, can act as
a plasminogen activator converting plasminogen to plasmin. By contrast
the cell bound plasmin receptor lacks plasminogen activator activity and
demonstrates binding specificity towards plasmin rather than the native
zymogen, Glu-plasminogen. The secretion of streptokinase into an
environment containing plasminogen would result in plasminogen activation
and the generation of plasmin. Plasmin generated by this reaction could
then bind specifically to the receptor on the surface of the bacteria.
In experiments documented in Chapter Two, it was recognized that once
plasmin bound to the bacterial surface receptor it retained enzymatic
activity. Furthermore, this cell bound plasmin activity could not be
regulated by plasmin's normal physiological inhibitor a^-antiplasmin.
The ability of bacteria, not only to produce plasminogen activators,
but to associate the enzymatically active product on their cell surface
in an physiologically nonregulatable form may prove to be an important
factor in the ability of these pathogens to invade human tissue. The
expression of two functionally distinct streptococcal proteins with
affinity for proteins of the human fibrinolytic system is intriguing. In
the next chapter, the relationship between the plasmin receptor and
streptokinase protein produced by strain 64/14 is compared
physicochemically, functionally, and antigenically.


102
I 2 3 4 5 6 7


124
concentrated using an Amicon concentrator (Amicon Corp., Danvers, MA)
fitted with a YM-10 membrane at 4C. Supernatants were concentrated
approximately 100 fold, and aliquoted in 0.5 ml fractions flash frozen
and stored at -70C.
Radioiodination of Proteins
Human plasminogen, isolated plasmin heavy chain, and isolated
plasmin light chain were iodinated by the chloramine T method using
Iodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell
(1982). The labeled proteins were separated from free iodine by passage
over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal
buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1%
gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. Labeled aliquots were used once
and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three fil of urokinase (Sigma 20
u/ml) was added to a 400 /xl solution of 1 //M plasminogen containing 0.04
M lysine. The mixture was incubated at 37C for 45 minutes unless stated
otherwise. The efficiency of plasmin generation was followed by
measuring the conversion of the single chain plasminogen molecule (Mr=
90,000 daltons) into heavy chains (Mr=60,000 daltons) and light chains
(Mr=25,000 daltons) as determined by the migration of radiolabeled
proteins, following denaturation and reduction, on 10% SDS-polyacrylamide
gels. The migration of labeled proteins was determined by auto
radiographic exposure of dried gels to Kodak XAR-5 film with intensifying
screens at -70C for 15-20 hours. The integrity of the iodinated plasmin


86
binds to both plasmin and the native zymogen Glu-plasminogen, while the
surface associated plasmin receptor shows no significant reactivity for
the native zymogen. Despite these clear functional differences, the
expression of two proteins by the same bacteria that bind to the key
human fibrinolytic protein plasmin with such selectivity raises the
possibility that they may be in some way related or derived from a common
precursor. Furthermore, the majority of information on the properties of
streptokinase have been derived from studies of the plasminogen activator
molecule isolated from group C streptococcal strains and evidence for
differences in antigenicity and hence possibly function have been
reported between streptokinase proteins isolated from group A and group C
streptococcal isolates (Dillon and Wannamaker, 1965; Weinstein, 1953).
The purpose of the studies presented in this chapter were to isolate a
functionally active receptor for human plasmin from strain 64/14 and to
compare it with the streptokinase protein that is produced by the same
organism.
Materials and Methods
Enzymes. Inhibitors and other Reagents
Urokinase and porcine elastase (type IV) were obtained from Sigma
Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from
Mobay Pharmaceuticals, New York, NY. D-Val-Phe-Lys-chloromethyl ketone
(VPLCK), and Phe-Pro-Arg-chloromethylketone (PPACK) were obtained from
Calbiochem-Behring, San Diego, CA. Human Lys-plasminogen was obtained
from American Diagnostica Inc., Greenwich, CT. H-D-Val-Leu-Lys-
paranitroanilide (S-2251) was obtained from Helena Laboratories,
Beaumont, TX. Purified group C streptokinase was a gift from Kabivitrum,
A.B., Stockholm, Sweden.


135
plasmin receptor and concentrated culture supernatants, which contain
streptokinase, from the strains of streptococcal bacteria tested. Panel
A in Figure 5-1 identifies the molecular species of streptokinase by
means of plasminogen activator potential using a solid phase assay as
described in the Methods. The major species of streptokinase produced by
all but group A B923 had a molecular weight of approximately 48,000
daltons. The major streptokinase molecular species from strain B923 was
slightly smaller, approximately 46,000 daltons. There was no
plasminogen activator activity in the extracted plasmin receptor
preparation. Smaller molecular species of a given streptokinase, as seen
in strain B923, ATCC 12449, and purified Kabikinase, result from
degradation of the larger secreted protein (Johnston and Zabriskie,
1986). The Western blot depicted in Panel B of Figure 5-1 show the ^->I-
plasmin binding activities and demonstrate the major binding activity in
the various culture supernatants is associated with streptokinase,
compare Panel A to B. Panel C of Figure 5-1 shows the ^->I-plasmin
binding results in the presence of EACA, which is known to disrupt the
binding of human plasmin to the 41,000 dalton plasmin receptor. These
results indicate that all of the streptokinases appear to have a binding
activity with human plasmin that is not disrupted in the presence of
EACA. However, the 41,000 dalton extracted plasmin receptor is clearly
shown to be sensitive to the presence of EACA. Compare lane 1 in Panels
B and C of Figure 5-1.
The studies presented in Chapter Three demonstrated that unlike the
group C streptokinase (Kabikinase), which has been shown to associate
with plasmin through interactions with determinants located in the light
chain of the plasmin molecule, the plasmin receptor associated with the


Figure 4-7. SDS-PAGE and Western blot analysis of plasmin receptor
protein with a polyclonal rabbit antibody. The isolated plasmin binding
protein, streptokinase from the same group A strain, and group C
streptokinase were compared for reactivity with a polyclonal rabbit
antibody to the purified plasmin receptor molecule. Lane 1: 2.0 /jg of
group C streptokinase (Kabikinase); lane 2: 50 tl of extracted plasmin
receptor preparation (approx. 5 fig of the 41,000 dalton molecule); lane
3: 60 il of strain 64/14 concentrated supernatant (approx. 2.0 ng of
64/14 streptokinase). The proteins on the nitrocellulose blot were
probed in a sandwich assay first with the polyclonal anti-plasmin
receptor antibody followed by ^^I-Protein q as described in the Methods.
The resulting blot was autoradiographed at -70C for 10 hours with
intensifying screens.


109
KD
I I 6
84
58
48.5
36.
26.6


87
Bacteria
The Lancefield /? hemolytic streptococcal strain 64/14 was grown as a
stationary culture at 37C, in one to two liter batches of a chemically
defined media for streptococci described by Van de Rijn and Kessler
(1980), containing 0.1% phenol red. The pH of the cultures were
maintained at a pH greater than 7.0, as monitored by the indicator dye.
For certain experiments, where noted, bacteria were grown in Todd-Hewitt
broth (Difco Laboratories, Detroit, MI) Approximately 2.0 to 4.0 g (wet
weight) of bacteria could be recovered per liter of media following a 24
to 36 hour incubation at 37C. Bacteria were harvested by
centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4,
containing 0.02% sodium azide. The bacteria were heat killed at 80C
for 15 min. The suspension was centrifuged and the pellet washed twice
with PBS containing 0.02% sodium azide. Aliquots could be stored at -
20C, or used immediately for extraction purposes.
Radioiodination of Proteins
Human plasminogen was iodinated by the chloramine T method using
Iodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell
(1982). The labeled proteins were separated from free iodine by passage
over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal
buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1%
gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. Labeled aliquots were used once
and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three n1 of urokinase (Sigma 20


65
lowering the dissociation constant between these molecules and <*2*
antiplasmin (Swenson and Thorsen, 1981; Wiman et al. 1979).
To examine the possible importance of the conformation of the
plasmin(ogen) molecule for binding to bacteria, the ability of the
conformationally altered form of plasminogen, Lys-plasminogen, to bind to
the group A streptococcus, 64/14, was measured. The isolated protein was
radiolabeled and examined by urea gel analysis for homogeneity. The
labeled material demonstrated a single band on an autoradiograph (Figure
3-3, panel A, lane 2) at a position corresponding to that reported for
the migration of Lys-plasminogen in this gel system (Swensen and Thorsen,
1981). This labeled form of plasminogen was found to bind to the
bacteria (Figure 3-3, panel B, lane 2). Similarly, Glu-plasmin
generated from Glu-plasminogen in the presence of aprotinin, was also
capable of binding to the bacteria (Figure 3-3, panel B, lane 3). The
relative efficiency of unlabeled Glu-plasminogen, Lys-plasminogen and
Lys-plasmin to compete with labeled plasmin for binding sites on the
group A streptococcus 64/14 was tested. Different concentrations of each
of these molecules were mixed with a fixed concentration (1.0 x 10'-*- M) ,
of PPACK reacted *^^1-Lys-plasmin and the extent of inhibition of binding
of radiolabel was measured, as described previously. The results of this
experiment shown in Figure 3-4 indicate that the inhibition achieved
with Lys-plasminogen and Lys-plasmin were identical. These results
indicate that the receptor for these ligands are the same, and that the
affinity for each protein is equivalent.
The possibility that the results presented in Figure 3-4 could be
accounted for by the conversion of Lys-plasminogen to Lys-plasmin during
the reaction was considered. The next series of experiments were


A
B
plasminogen- I
heavy chain
-I ight chain
as
O


160
been demonstrated to activate complement components, hydrolyze components
of connective tissue and basement membranes such as laminin (Liotta et
al. 1981a; Ott et al.. 1981), fibronectin (Liotta et al., 1981),
proteoglycans (Emonds-Alt et al., 1980), thrombospondin (Lawler and
Slayter, 1981) and type-V collagen (Liotta et al., 1981a,b), as well as
proteolytically activating other latent proteases like collageneases
(O'Grady et al., 1981; Stricklin et al., 1977; and Werb et al., 1977).
These collagenases can then catalyze the initial cleavage of the collagen
molecules, which are then susceptible to further proteolytic action by
plasmin (Dan^, 1985). It is now recognized that many normal cell types
also have the ability to produce plasminogen activators, as well as the
ability to specifically bind the plasminogen activators and
plasmin(ogen).
In conclusion, the ability of certain streptococcal bacteria, like
strain 64/14, to both produce a plasminogen activator (i.e.,
streptokinase) and also to express a receptor having an extremely high
affinity for the activation product plasmin may be important for
certain of the invasive properties of these organism. Furthermore, since
plasmin bound to a group A streptococcus is incapable of inhibition by
plasmin's normal physiological regulator a^-antiplasmin, the bacterium
has acquired a non-regulatable proteolytic activity that may contribute
to its tissue-invasive properties. All of these findings would be
consistent with a linked role for streptokinase and the surface bacterial
plasmin receptor. Together these properties of a group A streptococcus
would provide the bacteria with a mechanism to capture an unregulated
enzyme activity that might alter their ability to interact with host
barriers. In fact, the streptococcal strain 64/14 when grown in human


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ISOLATION AND CHARACTERIZATION OF A
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
By
CHRISTOPHER C. BRODER
May 1989
Chairman: Michael D.P. Boyle
Major Department: Immunology and Medical Microbiology
The expression of a specific receptor for the key human fibrinolytic
enzyme plasmin on the surface of the group A streptococcal strain 64/14
is reported. The receptor was specific for plasmin, and demonstrated no
significant reactivity with the zymogen form of the molecule (Glu-
plasminogen). Bacterial bound plasmin retained its enzymatic activity,
and could not be inhibited by the physiological regulator (<*2-
antiplasmin). The receptor demonstrated a high affinity for plasmin
(Kp=l.0x10'I m), and binding was maximal at physiologic pH and ionic
strength. Furthermore, the receptor-ligand complex was reversibly
inhibitable by e-aminocaproic acid and L-lysine. The binding of plasmin
to this group A streptococcus was found to occur primarily through
interactions with the heavy chain of the plasmin molecule, and was
dependent on a specific conformation of the ligand. A functionally
active plasmin receptor was obtained from strain 64/14 bacteria by an
enzymatic extraction with mutanolysin. Plasmin binding activity was
expressed predominantly in a protein having an Mr of approximately 41,000
xii


Figure 2-3. Regulation of bacterial bound enzyme activity by a variety
of different serine protease inhibitors: Bacterial pellets were pre
incubated with plasmin, washed and resuspended in buffer containing
excess pNpGB, PPACK, aprotinin, a2-AP or buffer alone for 15 minutes at
room temperature. Following incubation with the inhibitor the bacteria
were pelleted and washed. Enzyme activity was then measured by the
ability of the samples to hydrolyze the chromogenic substrate HD-Val-
leu-gly-paranitroanilide as described in the Methods. The data are
presented as the mean the standard deviation of duplicate experiments.
The hydrolysis by bacterial bound plasmin in the absence of any
inhibitor represents 100% activity.


119
I 2 3
KD
I I 6-
84 *
5 8- Hj
48.5
36.5-
26.6-


162
REFERENCES
Abiko, Y., M. Iwamoto, and M. Tomikawa. 1969. Pasminogen-plasmin system
V: A stoichiometric equilibrium complex of plasminogen and a
synthetic inhibitor. Biochim. Biophys. Acta. 185:424-431.
Alkjaersig, N. 1964. The purification and properties of human
plasminogen. Biochem. J. 93:171-182.
Aoki, N., M. Moroi, M. Matsuda, and K. Tachiya. 1977. The behavior of
o^-plasmin inhibitor in fibrinolytic states. J. Clin. Invest.
60:361-369.
Astrup, T. 1978. Fibrinolysis: An overview. In: Progress in Chemical
Fibrinolysis and Thrombolysis, Vol. 3, pp. 1-57, Davidson, J.F.,
Towan, R.M., Samana, M.M. and Desnoyers, P.C. (eds.), Raven Press,
New York.
Atichartakarn, V., V.J. Marder, E.P. Kirby, and A.Z. Budzynski. 1978.
Effects of enzymatic degradation on the subunit composition and
biologic properties of human factor VIII. Blood 51:281-297.
Bajpai, A., and J.B. Baker. 1985. Cryptic urokinase binding sites on
human foreskin fibroblasts. Biochem. Biophys. Res. Commun.
133:475-482.
Beebe, D.P. 1987. Binding of tissue plasminogen activator to human
umbilical vein endothelial cells. Thromb. Res. 46:241-254.
Beers, W.H., S. Strickland, and E. Reich. 1975. Ovarian plasminogen
activator: Relationship to ovarian and hormonal regulation. Cell
6:387394.
Bhaduri, S., and Demchick, P.H. 1983. Simple and rapid method for
disruption of bacteria for protein studies. Applied and
Environ. Microbiol. 46:941-943.
Boyle, M.D.P., and K.J. Reis. 1987. Bacterial Fc receptors.
Biotechnology 5:697-703.
Brockway, W., and F. Castellino. 1972. Measurement of the binding of
antifibrinolytic amino acids to various plasminogens. Arch.
Biochem. Biophys. 151:194-199.


53
Preparation of Elastase Digestion Fragments of Plasminogen
Elastase digestion of human plasminogen yields three defined
fragments of the plasminogen molecule (Sottrup-Jensen et al., 1978) .
These are 1) the lysine-binding domain I (LBS-I), Mr of approx. 38,000
daltons containing kringle domains 1 through 3, 2) Lysine binding domain
II (LBS-II), Mr of approx. 10-12,000 daltons consisting of the kringle
domain 4, and 3) the non-lysine-binding domain known as mini-plasminogen,
Mr of approx. 36,000 daltons containing the remainder of the heavy chain
(kringle 5) and intact light chain. Elastase digestion was performed
using established conditions (Sottrup-Jensen et al., 1978). Purified
Glu-plasminogen (3.0 mg/ml) in 0.05 M Tris, 0.1 M NaCl, pH 8.0, was
digested with a 40:1 molar ratio of Glu-Plasminogen to porcine elastase
in the presence of 250 KIU/ml aprotinin for 6.5 hours at room temperature
with gentle stirring in a total volume of 20 mis. At this time an
aliquot containing 50 ng of protein was removed for analysis by SDS-PAGE
and silver staining, to determine the extent of plasminogen digestion.
The remainder of the reaction mixture was flash frozen and stored at -
70C. The fragments were subsequently purified by a combination of
affinity chromatography on lysine-Sepharose and gel filtration on
Superse 6 (Pharmacia FPLC). The concentrations of the purified proteins
(see Figure 3-1, panel A) were determined spectrophotometrically, using
previously reported A^^11111 values of 17.0 for both Glu- and Lys-
plasminogen (Holvoet et al., 1985), 14.0 for mini-plasminogen (Holvoet et
al., 1985), 22.5 for LBS I (Nilsson et al., 1982), 25.0 for LBS II
(Nilsson et al., 1982), and 16.0 for plasmin heavy chain and plasmin
light chain (Summaria and Robbins, 1976). All proteins were aliquoted
and stored at -70C.


31


131
and the commercially available purified Kabikinase (KabiVitrum, Sockbolm,
Sweden) were compared physicochemically, functionally and antigenically.
The well characterized highly purified commercially available group C
streptokinase (Kabikinase) was used as a reference. Streptokinase
proteins were obtained from the bacterial strains by growing them in
chemically defined media under pH controlled conditions in order to
optimize the yield of streptokinase in the culture supernatants (Johnston
and Zabriskie, 1986). The quantity of functionally active streptokinase
was measured using the fluid phase plasminogen activator assay described
in the Methods. This assay used the highly purified streptokinase
(Kabikinase) as a standard. The quantities of streptokinase produced by
the various strains of either group A or C streptococcal bacteria are
summarized in Table 5-1. The concentrations of each of the plasminogen
activators (SK)s were expressed as units relative to Kabikinase.
We have previously demonstrated that the binding of human plasmin to
the group A streptococcal strain 64/14 was both inhibitable and
reversible with L-Lysine or EACA. Furthermore, the binding of human
plasmin to the extracted receptor from this bacteria was also shown to be
inhibitable and reversible with these molecules in Western and dot-blot
IOC
assays using -plasmin as the probe. However, it was demonstrated
that the binding of human plasmin to group C streptokinase (Kabikinase)
was not sensitive to L-Lysine or EACA in similar assays. The following
series of experiments were designed to compare the binding specificities
of the extracted plasmin receptor from strain 64/14 to the streptokinase
produced by strain 64/14 from the same culture, as well as those from two
other group A streptococci, two group C streptococci and Kabikinase.
Shown in Figure 5-1 are parallel 10% SDS-PAGE Western blots of extracted


92
supernatants were then dialyzed at 4C into 20 mM Tris-HCl, 0.15 M NaCl
pH 7.4 containing 1.0 mM iodoacetic acid, 1.0 mM benzamidine HCl.and were
stored at -70C.
Time Course Trypsin Digestion
The bacterial pellet from approximately 11 ml of 10% (w/v) 64/14 in
PBS-azide was collected by centrifugation at 10,000 x g for 10 minutes.
The pellet was washed with 10 ml of 0.05 M KH2PO4, 0.005 M EDTA, 0.02%
NaN3 pH 6.1, centrifuged as before and resuspended to 10% (w/v) in that
buffer. These salt and buffer conditions are not optimal for trypsin
activity and facilitate the extraction of surface proteins without
concomitant proteolysis of the solubilized material. Pancreatic DNAse I
(Sigma) was added to approximately 6.0 ml of this suspension to a final
concentration of 4 g/ml. The sample was vortexed and warmed to 37C.
Bovine pancreatic trypsin (Type I, Sigma) was then added to a final
concentration of 20 fMg/ml and the sample was mixed. A 1.0 ml aliquot was
immediately removed and mixed with a concentrated solution of benzamidine
HC1. The final concentration of benzamidine HC1 in the reaction mixture
was 100 mM, well in excess of what was required to completely inhibit
the activity of the trypsin present in the reaction mixture. This
sample, was mixed and placed on ice, and was designated the zero time of
the experiment. At 5, 10, 30 and 60 minutes 1.0 ml aliquots were removed
from the reaction mixture, and were treated in an identical manner. A
control digestion was prepared by incubating a 1.0 ml aliquot of 10%
bacterial suspension containing 4 /g/ml DNAse I at 37C for 60 minutes
followed by the addition of benzamidine HCl to 100 mM final
concentration. All samples were centrifuged at approximately 10,000 x g
for 10 minutes. Supernatants were collected and stored at -70C. Prior


Figure 3-2. Inhibition of. PPACK reacted. 125I-Lvs-plasmin binding to
group A streptococcal plasmin receptor: A constant concentration of
(1.0 x 10 M) PPACK reacted I-Lys-plasmin and an increasing
concentration of unlabeled competitor molecules (10"1 M to 10"6 M) were
mixed with a fixed concentration of streptococcal strain 64/14.
Following incubation and washing (see Materials and Methods), the amount
of radiolabeled Lys-plasmin bound to the bacterial pellet was
determined. The quantity of radioactivity bound in the presence of
unlabeled competitor was compared to the radioactivity bound in the
absence of inhibitor and the percent inhibition calculated. ( -Lys-
plasmin; o-HC; -LBS-I; D-LBS-II; a-LC; a -Mini-PLG; -Mini-PLA).


O.D. 405nm
72
PLASMIN (nM)


5
native Glu-plasminogen molecule contains one high affinity lysine
binding site (K = 9 fiM.) and five weaker lysine-binding sites (K = 5 mM)
(Markus et al.. 1978a, 1978b). Lys-plasminogen contains one high
affinity, one intermediate affinity, and four lower affinity lysine
binding sites. The exact number of sites on the plasmin molecule has not
been reported. Two of these lysine-binding sites have been mapped to
specific regions in the plasminogen molecule. Studies involving
equilibrium dialysis experiments on the binding of EACA to isolated
fragments of the plasmin(ogen) molecule (see Chapter Three for a
complete discussion of these fragments) revealed that the high affinity
lysine-binding site was located in the kringle 1 structure, and kringle 4
contained one of the lower affinity sites (Lerch et al., 1980).
Plasminogen Activation
The generation of plasmin from plasminogen is accomplished by
plasminogen activators. Three plasminogen activators have been
extensively studied. Urokinase (UK) and tissue plasminogen activator
(tPA) are proteolytic enzymes (for review see Astrup, 1978); the third,
streptokinase (SK), possesses no inherent proteolytic activity. Tissue
plasminogen activator, by virtue of its serine protease activity can
directly activate plasminogen. This enzyme is present in various tissues
and can also bind to fibrin. Urokinase, also a serine protease, is a
glycoprotein which has no fibrin binding capacity. Urokinase can also
activate plasminogen directly cleaving the Arg56o_Val561 peptide bond.
Streptokinase is a unique plasminogen activator which is produced by
certain streptococci. The only apparent function of streptokinase, since
its initial description by Tillett and Garner (1933), is its ability to


46
Plasmin(ogen) has several lysine binding sites located on its heavy
chain. The low affinity sites are primarily important for binding to
fibrin and the high affinity site is important for the interaction with
2-antiplasmin (Wiman et al.. 1979). In order to assess the possible
role of these lysine binding sites in the interaction between plasmin and
its bacterial receptor, the bacterial binding of plasmin in the presence
of increasing concentrations of EACA, lysine, or arginine. The results
in Figure 2-7 and 2-8 demonstrate that binding could be inhibited, and
bound plasmin could be eluted, by these amino acids in a
concentration-dependent manner. Eluted plasmin could be re-bound to
bacteria simply by removing lysine from the eluted plasmin solution,
indicating a possible importance of the lysine binding sites for the
receptor:plasmin interaction. There are 4 or 5 low affinity sites (Kq =
5mM) and one high affinity site (K¡) = 9/zM) for EACA (Markus et al. ,
1978). Lysine and arginine bind to the high affinity site less tightly
than does the lysine analog EACA (Wiman and Collen, 1978). Analysis of
the inhibition curves for EACA, lysine, and arginine reveal that
occupancy of the high affinity lysine binding site on plasmin interferes
with binding to the bacteria. It is recognized that occupancy of the
high affinity lysine binding site causes gross conformational changes in
the plasmin(ogen) molecule, and therefore the possibility for allosteric
as well as direct effects needs to be considered. The very high affinity
of the receptor for plasmin, approximating the affinity of <*2-antiplasmin
for plasmin (Kp = 2 x 10'^ M, Wiman and Collen, 1978), suggests that
streptococci may be able to compete effectively with a2-antiplasmin for
plasmin. and could explain why bacterial bound plasmin cannot be
regulated by Q2*antiplasmin. The enzymatic inhibition of plasmin by


106
as well as the samples of strain 64/14 concentrated supernatant. The
mutanolysin extracts of strain 64/14 were totally devoid of plasminogen
activator activity (data not shown). The strain 64/14 concentrated
supernatant was found to contain approximately 3,555 72 units of SK
activity per ml, or approximately 37 /g/ml of streptokinase, based on the
kabikinase standard. There was a total of 15 mis of concentrated
supernatant from 1 liter of this strain 64/14 bacterial culture. The
possibility that treatment of streptokinase with mutanolysin may destroy
plasminogen activator activity was considered, and under the conditions
of mutanolysin treatment used for extraction no loss of plasminogen
activator activity of mutanolysin treated streptokinase was observed
(data not shown).
A second assay to investigate plasminogen activator activity was
also used. This is a semi-quantitative, solid-phase assay designed to
correlate plasminogen activator activity to molecular weight of the
activator present in a sample (Dr. K. Johnston, personal communication).
In this assay the samples to be analyzed are first electrophoresed by
SDS-PAGE, separating individual proteins, and are then electroblotted
onto nitrocellulose membrane (see Methods). The nitrocellulose membrane
is then blocked with BSA, and applied to an agarose film containing
plasminogen and the chromogenic substrate S-2251. The plasminogen
activator present on the membrane will activate plasminogen in the
agarose which will in turn hydrolyze the S-2251. After allowing a period
of activation (time determined by the amount of plasminogen activator
present in the sample) the paranitroaniline product that passively
adheres to the membrane is chemically fixed to membrane. This fixation
procedure results in the production of a pink band(s) indicative of the


70
designed to determine whether Lys-plasminogen binds to the bacteria
without first being activated. These experiments were carried out by
monitoring the distribution of Lys-plasminogen, Glu-plasminogen, or Lys-
plasmin, in the fluid phase and associated with the bacteria, following
incubation of the protein with the bacteria. Unlabeled Glu-plasminogen,
Lys-plasminogen, or Lys-plasmin was added to a fixed concentration of the
group A streptococci 64/14 and incubated for 30 minutes at 37C.
Following this incubation period, the bacteria were pelleted by
centrifugation and the supernatants were recovered and monitored for
enzymatic activity either directly for Lys-plasmin, or following
activation with excess streptokinase for the sample containing Glu-
plasminogen or Lys-plasminogen, as described in the Methods. Following
incubation with bacteria, and removal of the bacteria by centrifugation,
there was no significant Lys-plasmin activity detectable in the bacterial
free supernatant (Figure 3-5). By contrast over 98% of the enzymatic
potential of Glu-plasminogen was detected in the supernatant, while in
similar experiments using Lys-plasminogen less than 10% of the enzymatic
potential was measured following activation with streptokinase (Figure
3-6). Because of differences in the efficiency of detection of plasmin
activity in the fluid phase compared to its activity when bound to
bacteria it is not possible to quantitate accurately the exact percentage
of plasmin activity that is bound to bacteria. However, I have
demonstrated previously that once associated with bacteria, the plasmin
retains its ability to cleave synthetic chromogenic substrates like H-D-
val-leu-lys-pNA (S-2251), as documented in Chapter Two. Consequently,
the washed pellets from the absorption reaction were incubated with this
synthetic substrate. The results presented in Table 3-2 demonstrate


132
Table 5-1.
Fluid-phase plasminogen activator activity assay.
Plasminogen Activator
Activity*
approx.
Strain Units/ml /ig/ml
Group A
Strep.
64/14
3,55070
40
B923
3,970160
40
A995
3,250160
30
Group C
Strep.
26RP66
2,330180
20
ATCC 12449
44,0001170
460
64/14 mutanolysin
extract
5014
0.5
Plasminogen activator activity (e.g. streptokinase) present in
concentrated culture supernatants, and mutanolysin extracted plasmin
binding activity. Estimates were determined using purified Kabikinase
the standard. See Methods for precise experimental details.
as


82
This form of the zymogen molecule is known to be conformationally
distinct from Glu-plasminogen and is similar in conformation to Lys-
plasmin (Violand et al.. 1975). The binding of Lys-plasminogen to the
group A streptococcal receptor is therefore dependent on a specific
conformation, most probably of the heavy chain.
Interaction of both plasmin and plasminogen with thrombospondin (an
adhesive glycoprotein) has been demonstrated to occur via interaction
with the heavy chain of the plasmin(ogen) molecule (Silverstein et al.,
1984; Walz et al., 1987). However, efficient binding of thrombospondin
with any elastase digestion fragment of plasminogen has not been observed
(Walz et: al., 1987). I have observed a similar pattern for the
interaction of plasmin(ogen) with its bacterial receptor. However,
unlike the interaction of thrombospondin with plasmin, the bacterial
binding properties were reversible by addition of lysine or lysine
analogs (Broeseker et al., 1988).
Histidine-rich glycoprotein (HRGP) an a2glycoProtein *-n human
plasma, has been reported to compete with a^-AP for the high-affinity
lysine-binding site in plasmin (Haupt and Heinburger, 1972; Lijnen et
al., 1980). In addition, HRGP also reduces the binding of plasminogen to
fibrin by complex formation with the low-affinity lysine binding sites
(Lijnen et al., 1980). Furthermore, the characteristic interaction of
Glu-plasminogen, Lys-plasminogen, or plasmin and their fragments with
fibrin or fibrinogen involves the heavy chain lysine-binding sites
(Cenderholm-Williams, 1977). This is distinct from the profile of
reactivity for the interaction of these proteins with the group A
streptococcus (Table 3-1). It can be seen that Glu-plasminogen shows no
reactivity with these bacteria, nor is there any significant reactivity
with the isolated lysine-binding fragments LBS I or LBS II.


79
that the bacteria incubated with Lys-plasminogen exhibited only a low
level of enzymatic activity (approximately 15% of that observed in the
samples pre-incubated with Lys-plasmin). The bacterial free supernatant
of the sample incubated with Lys-plasminogen demonstrated <10% of the
zymogen remained in the supernatant as detected by enzymatic activity
following activation with streptokinase (Figure 3-6). Taken together
these results indicate that the Lys-plasminogen was removed from the
fluid phase without prior or concomitant activation to Lys-plasmin.
Furthermore, when the bacterial bound ^^^1-Lys-plasminogen was eluted
from the bacteria and examined by polyacrylamide gel electrophoresis
under reducing conditions, a single protein band was observed on the
autoradiograph corresponding to the enzymatically inactive modified
zymogen form of the protein, Lys plasminogen (Figure 3-7). All of these
studies demonstrate that Lys-plasminogen can bind to bacteria without
first being converted to Lys-plasmin. This would indicate that the
intact native plasminogen molecule (Glu-plasminogen) does not express
structures that are recognized by the bacterial plasmin receptor.
However, following a conformational change achieved by either conversion
to the Lys-plasminogen form of the zymogen or by activation to plasmin,
structures are formed or exposed on the molecule that facilitate
interaction with the bacteria.
Discussion
Plasmin is the key component of the mammalian fibrinolytic enzyme
system which is responsible for fibrin degradation and intravascular
blood clot lysis. Active plasmin, which cleaves fibrin, is derived from
the circulating zymogen precursor Glu-plasminogen. Glu-plasminogen is a


daltons. The plasmin receptor demonstrated no plasminogen activator
activity. A functionally active plasmin receptor protein was purified by
affinity chromatography using immobilized plasmin and specific elution
with L-lysine or EACA. The strain 64/14 plasmin receptor was compared
with secreted streptokinase proteins from five streptococcal isolates
including strain 64/14. Only the plasmin receptor-plasmin complex was
found to be sensitive to L-lysine or EACA. Polyclonal rabbit and mouse
anti-plasmin receptor antibodies were prepared, as well as polyclonal
anti-group C streptokinase antibodies. Using these antibodies as well as
a bank of mouse monoclonal anti-group C streptokinase antibodies, the
41,000 dalton plasmin receptor protein from strain 64/14 was shown to be
antigenically unrelated to either group A or C streptokinase. Thus the
strain 64/14 streptococcal receptor for human plasmin is
physicochemically, functionally, and antigenically distinct from
streptokinase. The importance of a specific receptor for human plasmin
on pathogenic streptococci is unclear; however, it may provide a
mechanism for the capture of the potent enzyme plasmin which may confer
additional invasive properties to the bacteria.
xiii


28
mediate dissolution of the fibrin clot was measured. The results
presented in Table 2-2 demonstrate that the bacterial bound plasmin still
retained its ability to cleave fibrin. These effects could not be
accounted for by dissociation of plasmin from the bacteria, since clot
lysis did not occur when the microbe-plasmin complex was separated from
the clot by a 0.22 /m Millipore filter (Table 2-2). Under these
experimental conditions, unbound plasmin was capable of passing through
the filter and causing fibrin degradation.
The following series of experiments, which characterize further the
interaction of human plasmin and this group A streptococci bacteria were
performed by Dr. Tim A. Broeseker, a Fellow in the department of
medicine, division of hematology at the University of Florida.
In his initial experiments the binding of labeled plasmin or
plasminogen to the group A streptococcal strain 64/14 as a function of pH
was tested. Labeled proteins were pre-equilibrated in VBS-gel buffers of
differing pH's before the addition of bacteria. After an incubation
period of 15 minutes at 37C, the radioactivity associated with the
bacteria was measured by pelleting the micro-organisms and washing free
the unbound label with buffer of the appropriate pH as described in the
Methods. Maximal binding of plasmin to the bacteria was observed between
pH 5 and 8 with approximately 60% of counts offered being bound by the
group A streptococcus 64/14, (Figure 2-4). In contrast, addition of
labeled plasminogen to the bacteria over the entire pH range tested
(pH 5-9) resulted in direct binding of less than 10% of offered counts.
(Figure 2-4). Similar studies were carried out to determine the effect of
ionic strength on binding of radiolabeled plasmin and plasminogen to the
group A streptococcal strain 64/14. Labeled proteins were pre-


Figure Page
5-4. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and
group C streptokinase with mouse polyclonal
anti-plasmin receptor antibodies 145
5-5. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and
group C streptokinase with mouse anti-group C
streptokinase monoclonal antibodies 148
IX


Figure 2-6. Specific binding of 125I-plasmin to 107 group A
streptococci, strain 64/14. following a 15 minute incubation at 37C in
VBS-gel at pH 7.4: Measurements of triplicate experiments were
performed. Specific binding was determined as described in Methods.
The inset represents the Scatchard analysis of the specific binding
data.


KD
I 6-
84-
I 2 3
48.5-
36.5-
26.6-
5 A3
8E2
2D3
I 2G7
I I 6-
58
48.5
36.5
26.6
IBM | 2H7 9F | 2
Lane I. Plasmin Receptor Protei n
Lane 2. Group A SK
Lane 3. Group C SK
I I H4
5G I |
3H4/C7
6H7
I I H5
6D3
I 2H8 4D8 8F4
148


104


Figure 4-2. SDS-PAGE and Western blot analysis of mutanolvsin extracted
64/14 bacterial plasmin binding activity. Parallel 10% SDS-poly-
acrylamide gels were electrophoresed. One gel was silver stained to
detect protein molecules (Panel A) and the second was Western blotted and
probed with -plasmin, as described in the Methods, and auto-
radiographed for 10 hours at -70C with intensifying screens (Panel B).
Panel A, lane 1: 1.0 /g of group C streptokinase (Kabikinase); lane 2: 50
Hi of strain 64/14 mutanolysin extract. Panel B, lane 1: 50 tl of strain
64/14 mutanolysin extract; lane 2: 1.0 /jg of group C streptokinase
(Kabikinase).


107
presence of plasminogen activator. Samples of the mutanolysin extracted
plasmin binding activity preparation, concentrated supernatant from the
strain 64/14 pH controlled CDM cultures, and purified streptokinase
(Kabikinase) as a positive control were all analyzed with this assay
system. The results are shown in Figure 4-3. There were no bands of
activator activity present in the mutanolysin extracts from the 64/14
bacteria. However, there was a streptokinase activity in the culture
supernatant of this strain. This secreted plasminogen activator activity
co-migrated at the same molecular weight (approx. 48,000 daltons) as the
purified sample of group C streptokinase obtained from Kabivitrum.
Because of the high level of plasmin receptor activity in the
extractions prepared by mutanolysin digestion, this preparation was
chosen as material for further analysis and purification of the strain
64/14 plasmin receptor from bacteria grown in chemically defined media.
Due to the simplicity and specificity offered by affinity chromatography,
the plasmin receptor activity was purified from the mutanolysin extracts
using a plasmin affinity matrix prepared as described in the Methods.
Briefly, approximately 50 mg of purified human plasminogen was activated
to plasmin by incubation in the presence of the plasminogen activator
urokinase. The plasmin was then enzymatically inactivated and coupled to
6.0 mis of the affinity chromatography support Affi-Prep 10 (Bio Rad).
Following ligand coupling, the remaining active sites on the matrix were
blocked with 1.0 M ethanolamine HC1 pH 8.0. The matrix was then washed
with 2.0 M NaCl. The Affi-Prep 10 inactivated plasmin was loaded
intoan HR 10/10 FPLC compatible column (Pharmacia). The affinity matrix
was again washed with 2.0 M NaCl, and then equilibrated in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaNj pH 7.4.


56
washed three times with 2.0 ml VBS-gel. The bacterial pellets were
resuspended in 300 pi of either VBS-gel containing 0.5% SDS; VBS-gel
containing 0.1M EACA; or VBS-gel containing 0.5% SDS and 2.0% /?-
mercaptoethanol, to elute the l^I-Lys-plasminogen from the bacteria.
Following a 10 minute incubation at 37 C the bacteria were removed by
centrifugation and the supernatant recovered. The eluted material was
analyzed by electrophoresis on a 10%-SDS-PAGE-6M-Urea gel under reducing
conditions. The gel was dried and the migration of labeled protein
determined by autoradiography. Similar studies were also carried out in
which the bacterial bound -*-^^I-Lys-plasminogen was treated with a 20
unit/ml concentration of urokinase for 20 minutes at 37C in a total
volume of 300 pi of VBS-gel prior to eluting the bound proteins.
Following this plasminogen activation reaction the bacteria were
centrifuged and washed twice with 2.0 mis VBS-gel. The residual bound
-IOC
I-Lys-plasmin(ogen) was eluted and analyzed as described above.
Measurement of Functional Activity of Plasmin(ogen') in Bacterial-Free
Supernatants
The following assay was used to measure binding of the various
plasmin(ogen) as an alternative method to using radiolabeled tracers.
In these studies, 2.0 pg of Glu-plasminogen, Lys-plasminogen, or Lys-
plasmin was incubated with 100 pi of a heat killed 10% w/v suspension of
the group A streptococcal strain 64/14 for 20 minutes at 37C in a total
reaction volume of 400 pi of VBS-gel. Following incubation, the bacteria
were removed by centrifugation at 12,000 x g for 4 minutes in an Ependorf
Microfuge and bacterial-free supernatants were obtained. Control tubes
for each plasmin(ogen) species containing no bacteria were treated
identically and all samples were run in duplicate. The bacterial free
supernatants were recovered and enzymatic activity was measured as


Figure 5-2. Comparison of binding reactivities of streptokinase proteins
and plasmin binding receptor protein with -plasmin heavy chain and
^-'i-plasmin light chain. Parallel 10% SDS-polyacrylamide gels were
electrophoresed. The proteins on each gel were transferred to separate
nitrocellulose sheets and the membranes blocked and probed with 3I-
plasmin heavy chain (Panel A) or -plasmin light chain (Panel B) as
described in the Methods. The resulting blots were autoradiographed at -
70C for 24 hours with intensifying screens. Lane(s) 1: exracted plasmin
receptor preparation (approx. 5.0 ng); lane(s) 2: approx. 2.0 /ig of 64/14
SK; lane(s) 3: approx. 2.0 g of B923 SK; lane(s) 4: approx. 1.0 ng of
A995 SK; lane(s) 5: approx. 2.0 ng of Kabikinase; lane(s) 6: approx. 2.0
pg of ATCC 12449 SK; lane(s) 7: approx. 1.0 ng of 26RP66 SK.
Streptokinase proteins from the strains 64/14, B923, A995, ATCC 12449,
and 26RP66 were contained in concentrated culture supernatants, see
Methods for precise experimental details.


ISOLATION AND CHARACTERIZATION OF A
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
By
CHRISTOPHER C. BRODER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1989


171
Thorsen, S. 1975. Differences in the binding to fibrin of native
plasminogen and plasminogen modified by proteolytic degradation:
Influence of W-aminocarboxylic acids. Biochem. Biopbys. Acta.
393:55-65.
Tborsen, S., I. Clemmensen, L. Sottrup-Jensen, and S. Magnusson.
1981. Absorption to fibrin of native fragments of known primary
structure from human plasminogen. Biochem. Biophys. Acta. 668:377-
387.
Tillett, W.S., and R.L.L. Garner. 1933. The fibrinolytic activity of
hemolytic streptococci. J. Exp. Med. 58:485-502.
Tillett, W.S., and S. Sherry. 1949. The effect in patients of
streptococcal fibrinolysin (streptokinase) and streptococcal
deoxyribonuclease on fibrinous, purulent and sanguinous pleural
exudations. J. Clin. Invest. 28:173-190.
Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
Procedure and some applications. Proc. Natl. Acad. Sci. USA.
76:4350-4354.
Van De Rijn, I., and R.E. Kessler. 1980. Growth characteristics of
group A streptococci in a new chemically defined medium. Infect,
and Immun. 27:444-448.
Vassallii, J-D., D. Baccino, and D. Belin. 1985. A cellular binding
site for the Mr 55,000 form of the human plasminogen activator,
urokinase. J. Cell. Biol. 100:86-92.
Violand, B.N., J.M. Sodetz, and F.J. Castellino. 1975. The effect of e-
Amino Caproic Acid on the gross conformation of plasminogen and
plasmin. Arch. Biochem. and Biophys. 170:300-305.
Von Mering, G.O., M.D.P. Boyle, C.C. Broder, and R. Lottenberg. 1988.
Isolation of a surface receptor for human plasmin from a pathogenic
group A streptococcus. Clin. Res. 36:464a.
Walln, P., and B. Wiman. 1970. Characterization of human plasminogen:
I. On the relationship between different molecular forms of
plasminogen demonstrated in plasma and found in purified
preparations. Biochim. Biophys. Acta. 221:20-30.
Walln, P., and B. Wiman. 1972. Characterization of human plasminogen:
II. Separation and partial characterization of different molecular
forms of human plasminogen. Biochim. Biophys. Acta. 257:122-134.
Walz, D.A., T. Bacon-Saguley, S. Kendra-Franczak, and P. DePoli. 1987.
Binding of thrombospodin to immobilized ligands specific
interaction with fibrinogen, plasminogen, histidine-rich
glycoprotein, and fibronectin. Sem. Thromb. and Hemo. 13:317-325.


% OF TOTAL RADIOACTIVITY BOUND
PLASMINOGEN
PLASMIN
PLASMIN +a2PI
PLASMIN pNpGB + a2 PI
PLASMIN APROTININ + a2PI
PLASMIN -PPACK + a2PI


CHAPTER ONE
INTRODUCTION
Introduction
Plasmin, a serine protease, is the key component of the mammalian
fibrinolytic enzyme system. The main physiological role of the
fibrinolytic system is the dissolution of fibrin clots formed in blood
vessels. Milstone, in 1941, determined that the lysis of fibrin, by the
streptococcal substance described by Tillett and Garner in 1933, was
dependent on a 'lytic factor' in human serum. This was later followed by
the discovery that the lytic factor was an enzyme precursor, in human
plasma, that was converted to an active enzyme by a component in the
streptococcal fluid (Christensen, 1945; Kaplan, 1944). This precursor
was called plasminogen, the enzyme plasmin, and the streptococcal factor
streptokinase (Christensen and Macleod, 1945). The zymogen precursor
plasminogen, molecular weight approximately 92,000 daltons, is a single
chain glycosylated protein containing 790 amino acids in known sequence
and containing 24 disulfide bridges (Brogden et al., 1973; Sottrup-Jensen
et al. 1978; Wiman, 1973, 1977).
Plasmin(ogen) Structure
Native plasminogen (Glu-plasminogen) has glutamic acid as the NH2-
terminal residue, but is readily converted by the action of plasmin to
modified forms of plasminogen which are commonly called Lys-plasminogen.
1


Figure 2-5. Binding of 125I-plasmin or 125I-plasminogen to the group A
streptococcal strain 64/14 as a function of ionic strength: The data
are presented as the mean the standard deviation. Measurements of
duplicate experiments were performed and are expressed as the percent of
total counts offered (20,000 cpm) which were associated with the
bacterial pellet. (O O) 125I-plasmin; (O O) 125I-plasminogen


128
one hour at 37C; the time of incubation is dependent upon the
concentration of plasminogen activator present. At the termination of
incubation, the nitrocellulose membrane is removed from the agarose-
plasminogen- substrate gel and immersed for 5 min in freshly prepared
0.1% sodium nitrite dissolved in 1.0 N HCl followed immediately by
immersion for 5 min in 0.5% ammonium sulfamate dissolved in 1.0 N HCl.
The membrane is then transferred to a solution containing 0.05% N-l-
napthylethylenediamine in 47.5% ethanol and observed for the appearance
of red bands indicative of plasmin activity. Plasminogen activators
present on the nitrocellulose membrane will activate the plasminogen
present in the agarose which will in turn cleave the S-2251 substrate
incorporated in the agarose. The chromogenic cleavage product
(paranitroaniline) appears yellow and deposits on the nitrocellulose
membrane. The chemical treatment of the membranes as described above
will convert the deposited yellow paranitroaniline to a red color and fix
it to the membrane. Membranes can be stored under water at 4C.
Production of Rabbit Anti-streptokinase and Rabbit Anti-plasmin Receptor
Polyclonal Antibodies
Highly purified streptokinase from a Group C streptococcus was
obtained from KabiVitrum A.B., Stockholm, Sweden. Approximately 1.38 mg
of the antigen was electrophoresed on a 7.5% SDS-PAGE gel and stained
with Coomassie brilliant blue R-250. The single stained band was cut
from the gel and equilibrated in PBS-Azide. A portion of the gel slice
containing approximately 345 /g of streptokinase was emulsified with an
equal volume of Freund's complete adjuvant. The emulsion was injected
subcutaneously at 6 sites on a rabbit. The rabbit was boosted three
times at two week intervals with the streptokinase polyacrylamide gel
emulsified in Freund's incomplete adjuvant (approximately 300 /g per


61
competitor molecules (10" M to 10'^ m) were mixed with a fixed
concentration of bacteria. Following incubation and washing, the amount
of radiolabeled plasmin bound to the bacterial pellet was determined.
The quantity of radioactivity bound in the presence or absence of
unlabeled competitor was compared and the degree of inhibition calculated
(Figure 3-2) The results summarized in Table 3-1 show that unlabeled
plasmin inhibits the binding of labeled plasmin efficiently, with 50%
inhibition being observed in the presence of 1.2 x 10' M Lys-plasmin.
Significant inhibition of radiolabeled plasmin binding was also observed
when purified heavy chain was used as the competitor. Addition of any of
the other plasminogen fragments including isolated lysine binding domains
of the heavy chain (LBS I or LBS II) demonstrated no significant
inhibitory effect (Figure 3-2). Similarly, mini-plasminogen, mini-
plasmin, and isolated light chains demonstrated no significant inhibition
of binding of radiolabeled lys-plasmin over the concentration range
tested (10' M to 10"^ M) (Figure 3-2). Identical results were obtained
in the inhibition assays involving mini-plasmin, mini-plasminogen, LBS I,
LBS II, Lys-plasmin heavy chain, and plasmin light chain in the absence
of protease inhibitors in the reaction mixture (data not shown).
Combining equimolar quantities of the elastase digested fragments of
plasminogen or plasmin failed to restore any inhibitory potential.
Furthermore, combination of isolated light chain and heavy chain
demonstrated no synergistic effect in inhibitory capacity compared to the
sum of the isolated fragments alone (data not shown). The inhibition
curves for isolated heavy chain and intact plasmin (Figure 3-2)
demonstrated that both preparations could inhibit binding of labeled Lys-
plasmin by 100%. However, these curves differed in shape, indicating


% INHIBITION OF BINDING
69
CONCENTRATION OF COMPETITOR (M)


Figure 4-5. Analysis of affinity purified plasmin binding material from
the strain 64/14 mutanolvsin extract. 50 il aliqouts of affinity
purified fractions containing plasmin binding activity were electro-
phoresed on a 10% SDS-PAGE gel and stained with silver. The functional
activity of each sample was monitored by dot-blot analysis and probing
with 125i-plasmin, and is shown below the coresponding lane on the SDS-
polyacrylamide gel.


157
with l^I-plasmin. The 41,000 dalton molecule was specifically purified
from the mutanolysin extract of strain 64/14 by affinity chromatography
with immobilized plasmin and elution with L-lysine or EACA. Analysis of
the affinity purified plasmin binding activity by SDS-PAGE and Western
blotting, demonstrated that 41,000 dalton molecule was the predominant
molecule recovered, as detected by silver staining, and this molecule was
also responsible for the -plasmin binding activity of the affinity
purified sample. This plasmin binding molecule could be destroyed by
trypsin digestion and was consequently protein in nature. The 41,000
dalton plasmin binding protein was demonstrated to be totally devoid of
plasminogen activator activity, distinguishing it from the secreted
plasmin(ogen) binding protein streptokinase.
A detailed comparison between the plasmin receptor activity and
streptokinase produced by the same group A strain are the focus of the
studies in Chapter Five. The purpose of this series of experiments was
to make a comparison of the streptococcal strain 64/14 plasmin receptor
with the known and well characterized secreted streptococcal
plasmin(ogen) binding protein streptokinase. These studies focused on
obtaining plasmin receptor molecules as well as secreted streptokinase
from a single culture of streptococcal strain 64/14. These two molecules
were then compared for: (1) functional activity, (2) plasmin binding
specificity, and (3) antigenic relatedness These studies were performed
using Western blot assays to enable specific molecular species to be
identified and compared directly to function or antigenic reactivity.
In addition to streptokinase from strain 64/14, five other sources of
streptokinase were included in these studies for a more complete
comparison.


Figure 4-3. Solid-phase plasminogen activation assay. The plasminogen
activator activities of mutanolysin extracted and secreted proteins were
monitored using the solid phase plasminogen activator assay (Johnston,
personal communication). Lane 1: 50 /xl of mutanolysin extracted 64/14
plasmin binding material (approx. 5-10 fig of 41,000 dalton band); Lane2:
60 fil of strain 64/14 concentrated supernatant (approx. 2.0 fig of
streptokinase); Lane 3: 2.0 fig of group C streptokinase (Kabikinase).
For precise experimental details see Methods.


167
Leung, L.L.K., R.L. Nachman, and P.C. Harpel. 1984. Complex formation
of platelet thrombospondin with histidine-rich glycoprotein. J.
Clin. Invest. 73:5-12.
Levy, H.B., and Sober, H.A. 1960. A simple chromatographic method for
preparation of gamma globulin. Proc. Soc. Exp. Bio. and Med.
103:250-252.
Lijnen, H.R., Hoylaerts, M., and Collen, D. 1980. Isolation and
characterization of a human plasma protein with affinity for the
lysine binding sites in plasminogen. J. Biol. Chem. 255:10244-
10222.
Liotta, L.A., R.H. Goldfarb, R. Brundage, G.P. Siegal, V. Terranova, and
S. Garbisa. 1981a. Effect of plasminogen activator (urokinase),
plasmin, and thrombin on glycoprotein and collagenous of basement
membrane. Cancer Res. 41:4629-4636.
Liotta, L.A., R.H. Goldfarb, and V.P. Terranova. 1981b. Cleavage of
laminin by thrombin and plasmin: Alpha thrombin selectively cleaves
the beta chain of laminin. Thromb. Res. 21:663-673.
Liu, T-Y., and S.D. Elliott. 1965. Streptococcal proteinase: The
zymogen to enzyme transformation. J. Biol. Chem. 240:1138-1144.
Lottenberg, R., C.C. Broder, and M.D.P. Boyle. 1987. Identification of
a specific receptor for plasmin on a group A streptococcus. Infect.
Immun. 55:1914-1918.
Lottenberg, R., F.R. Dolly, and C.S. Kitchens. 1985. Recurring
thromboembolic disease and pulmonary hypertension associated with
severe hypoplasminogenemia. Am. J. Hematol. 19:181-193.
Malinowski, D.P., J.E. Sadler, and E.W. Davie. 1984. Characterization
of a complementary deoxyribonucleic acid coding for human and
bovine plasminogen. Biochemistry 23:4243-4250.
Marder, V.J., S.E. Martin, and R.W. Caiman. 1982. Clinical aspects of
consumptive thrombohemorrhagic disorders. In: Hemostasis and
Thrombosis: Basic Principles and Clinical Practice, p. 672, R.W.
Colman, J. Hirsh, V.J. Mardar and E.W. Solzman (eds.), J.B.
Lippincott Co., Philadelphia.
Markus, G., J.L. DePosquale, and F.C. Wissler. 1978a. Quantitative
determination of the binding of e-aminocaproic acid to native
plasminogen. J. Biol. Chem. 253:727-732.
Markus, G., J.L. Evers, and G.H. Hubika. 1976. Activator activities of
the transient forms of the human plasminogen streptokinase complex
during its proteolytic conversion to the stable activation complex.
J. Biol. Chem. 251:6495-6504.
Markus, G., J.L. Evers, and G.H. Hubika. 1978b. Comparison of some
properties of native (Glu) and modified (Lys) human plasminogen. J.
Biol. Chem. 253:733-739.


161
plasma has been documented to both be capable of generating plasmin from
plasminogen, as well as binding the active enzyme to its surface
(DesJardin et al. 1989). The importance of this selective receptor to
the infectious disease process of receptor positive bacteria remains to
be established.


117
M | 2 3 4
KD
I 2 3
A


84
The failure of a^-AP to regulate the bound enzyme suggests that the
required interaction between 02-AP and plasmin is directly or indirectly
inhibited. This may occur because one of the recognition sites for <*2-AP
in the kringle 1 region of plasmin's heavy chain may not be accessible
when plasmin is bound to a streptococcus.
The characteristics of the interaction of human plasmin with the
group A streptococcus, 64/14, described in this study indicate that the
bacteria can capture a potent protease activity that cannot be regulated
by the primary physiological inhibitor of plasmin, a^-AP. This group A
streptococci also secretes a plasminogen activator and consequently, in
the presence of plasminogen, the bacteria has the potential to both
generate plasmin and bind the active enzyme to its surface (DesJardin et
al. 1988) The importance of this selective receptor to the infectious
disease process of receptor positive bacteria remains to be established.
The purpose of the series of studies described in the next chapter
was to isolate and characterize the plasmin binding receptor from the
strain 64/14 streptococcus.


98
template. The nitrocellulose membrane containing the sample was then
drained of excess blocking buffer and overlayed on the agarose-
plasminogen- substrate gel. It is important to ensure uniform contact
between the nitrocellulose membrane and the gel. The nitrocellulose
membrane was allowed to remain in contact with the agarose for at least
one hour at 37C; the time of incubation is dependent upon the
concentration of plasminogen activator present. At the termination of
incubation, the nitrocellulose membrane was removed from the agarose-
plasminogen- substrate gel and immersed for 5 min in freshly prepared
0.1% sodium nitrite dissolved in 1.0 N HC1 followed immediately by
immersion for 5 min in 0.5% ammonium sulfamate dissolved in 1.0 N HC1.
The membrane was then transferred to a solution containing 0.05% N-l-
napthylethylenediamine in 47.5% ethanol and observed for the appearance
of red bands indicative of plasmin activity. Plasminogen activators
present on the nitrocellulose membrane will activate the plasminogen
present in the agarose which will in turn cleave the S-2251 substrate
incorporated in the agarose. The chromogenic cleavage product
(paranitroaniline) appears yellow and deposits on the nitrocellulose
membrane. The chemical treatment of the membranes as described above
will convert the deposited yellow paranitroaniline to a red color and fix
it to the membrane. Membranes can be stored under water at 4C.
Preparation of Polyclonal Rabbit Anti-plasmin Receptor Protein Antibody
Mutanolysin extracted 41,000 dalton plasmin receptor protein was
purified by gel electrophoreses on a 10% SDS-PAGE gel and stained with
Coomassie brilliant blue R-250. The single stained band was cut from the
gel and equilibrated in PBS-azide. The location of the 41,000 dalton
plasmin binding band was determined by the position of the stained band


LIST OF FIGURES
Figure Page
1-1. Schematic representation of the human Glu-plasminogen
molecule 4
2-1. Binding of plasmin to bacteria: comparison of the
kinetics of generation of plasmin and its ability
to bind to the group A streptococcus 64/14 22
2-2. Effect of inhibiting the active site of plasmin on
its ability to bind to the group A streptococcal
strain 64/14 24
2-3. Regulation of bacterial bound enzyme activity by a
variety of different serine protease inhibitors 27
2-4. Binding of ^^I-plasmin or 125j-piasmin0gen to the
group A streptococcal strain 64/14 as a function of
pH 31
2-5. Binding of ^^I-plasmin or 125j-piasminogen to the
group A streptococcal strain 64/14 as a function of
ionic strength 34
2-6. Specific binding of ^^I-plasmin to 10^ group A
streptococci, strain 64/14, following a 15 minute
incubation at 37C in VBS-gel at pH 7.4 37
2-7. Inhibition of binding of -plasmin to the group A
streptococcal strain 64/14 in VBS-gel containing
various concentrations of epsilon-aminocaproic acid,
lysine, and arginine 40
1 9 S
2-8. Elution of -plasmin from group A streptococcal
strain 64/14 in VBS-gel containing various con
centrations of epsilon-aminocaproic acid, lysine,
and arginine 42
3-1. SDS-UREA-PAGE analysis of isolated plasmin(ogen)
fragments 60
3-2. Inhibition of, PPACK reacted, ^^I-Lys-plasmin
binding to group A streptococcal plasmin receptor... 63
3-3. Binding of ^-^1 labeled Glu- and Lys-plasmin(ogens) . 67
vii


138
12 3 4 5 6 7 1
KD
84-
5 8-
48.5-
a
36.5-
26.6-*'
2 3 4 5 6 7
A
B


105
possible confusing factor in the isolation and characterization of the
surface plasmin receptor. The production of streptokinase by
streptococci is optimal when the pH of the growth medium is maintained
at 7.0 to 8.0 (Johnston and Zabriskie, 1986). The maintenance of a pH
above 6.8 prevents the activation of an extracellular zymogen to an
active mercaptoproteinase (Elliott and Dole, 1947; Liu and Elliott, 1965)
produced by the bacteria, which would significantly contribute to the
proteolytic hydrolysis of secreted streptokinase. Therefore, solubilized
plasmin receptor activity was prepared by mutanolysin extraction as
before, but from bacteria harvested from chemically defined media
cultures in which the pH was not allowed to become acidic. In addition,
the supernatants from these cultures were collected, filtered and
concentrated as a source of streptokinase from the strain 64/14 bacteria
for comparative analyses.
The production of plasmin receptor from chemically defined media, pH
controlled, cultures was first investigated. Receptor activity was
expressed, mutanolysin extracted, and also affinity purified.
Furthermore, there was no change in the molecular weight of the plasmin
receptor.
The production of streptokinase from the strain 64/14 bacteria was
measured functionally, by use of a quantitative plasminogen activation
assay. The commercially available highly purified group C streptokinase
(Kabikinase) (Kabivitrum, A.B., Stockholm, Sweden.) was used as a
standard in these studies.
The fluid phase assay for plasminogen activator activity (see
Methods for precise experimental details) was also used to measure
plasminogen activator activity in the mutanolysin extraction preparations


CHAPTER FOUR
ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
Introduction
The studies presented thus far have documented the existence of a
cell surface receptor for human plasmin on group A streptococcal strain
64/14. In addition to this plasmin binding activity, certain group A
streptococci have long been known to secrete the plasmin(ogen) binding
protein streptokinase, (Mr approx. 48,000 daltons), a non-enzymatic
plasminogen activator. This protein, described by Tillet and Garner
(1933), non-covalently associates with both plasminogen and plasmin, and
was originally identified by virtue of its ability to generate
fibrinolytic activity. Streptokinase binds rapidly to the native
zymogen Glu-plasminogen (rate content 5.4 x 10^ M'^S"^-) forming, a 1:1
stoichiometric complex with an estimated dissociation constant of 5 x
10"H M (Cederholm-Williams et al.. 1979). The formation of a complex
between streptokinase and plasminogen generates an enzymatic moiety
capable of plasminogen activator activity, a property neither protein
possesses alone.
The properties of the bacterial plasmin receptor reported thus far
are markedly different from streptokinase. While the bacterial plasmin
receptor binds preferentially to domains in the heavy chain of the
plasmin molecule (see Chapter Three), streptokinase binds to plasmin's
light chain (Summaria and Robbins, 1976). Furthermore, streptokinase
85


% INHIBITION OF BINDING
63
CONCENTRATION OF COMPETITOR (M)


BIOGRAPHICAL SKETCH
Christopher Charles Broder was born on June 12, 1961 in White
Plains, New York, the heart of New York suburbia, where he lived for
seventeen years. He graduated from White Plains High School in 1979,
and having been a member of the 'Jacques Cousteau television generation',
went south to attend the Florida Institute of Technology, Melbourne,
Florida. He studied marine science for most of his undergraduate career,
obtained a Dive Master certification and proceeded to explore every
available body of water in the state. He traveled back home in the
summers, and worked for Peckham Industries in Connecticut, where he
helped finance his academic life by working blacktop. He graduated with
a B.S. degree in biology, in 1983, along with Colleen M. Guay whom he
would later marry. He stayed on at F.I.T. entering graduate studies, and
received his M.S. degree in molecular biology in 1985, and hoped he would
never have to return to the wheelbarrow and shovel. In the fall of that
year he moved to Gainesville, Florida, and entered graduate school in the
department of Immunology and Medical Microbiology at the University of
Florida, and began his studies in the laboratory of Dr. Michael D.P.
Boyle in 1986. He has completed his research, and expects to receive
his Ph.D. in May of 1989 from the University of Florida.
174


159
group A streptokinase while three antibodies recognized epitopes on
group C streptokinase which were not present on the group A plasminogen
activator molecule. These findings further confirm the antigenic
difference noted previously between streptokinases of group A and group C
streptococci (Dillon and Wannamaker, 1965; Huang et al.. 1989). These
studies demonstrate that not only does the group A strain 64/14 produce
two distinct proteins with high affinity for human plasmin but there is
no evidence that these proteins share any common structural features.
This dissertation describes the first report of a prokaryotic
cellular receptor for human plasmin. However, eukaryotic cellular
receptors for plasmin(ogen) have been reported recently (Hajjar et al.,
1986; Miles et al., 1986; Miles and Plow 1985; Plow et al., 1986).
Eukaryotic cellular receptors for the plasminogen activators urokinase-
type plasminogen activator (Bajpai and Baker, 1985; Del Rosso et al..
1985; and Vassalli et al., 1985) and tissue-type plasminogen activator
(Beebe, 1987; and Hajjar et al., 1987) have also been described. It has
been concluded by others that the expression of plasminogen receptors
and their colocalization with urokinase- and tissue-type plasminogen
activator receptors provide cells with a basic mechanism for obtaining
proteolytic activity for the purposes of certain cellular functions (Knox
et al. 1987) and for modification of their surroundings (Knudsen et
al., 1986; Sheela and Barret, 1982). It had long been recognized (Carrel
and Burrows, 1911; Lambert and Hanes, 1911) that cancer tissue
consistently caused proteolytic degradation upon culturing. It has since
been shown that this proteolytic activity stems from the ability of many
types of cancer cells to produce and secrete plasminogen activators. In
fact plasmin, in addition to acting within the fibrinolytic system, has


94
allowed to stand on ice for 5 minutes and then collected by
centrifugation at 10,000 x g for 10 minutes. Residual acetone was
evaporated under a stream of air. The pellet was resuspended with
vortexing in 25 ml of 1.0% (v/v) Triton X-100 in PBS-azide and incubated
at room temperature for 5 minutes. An additional 2.5 ml of PBS-azide was
added. The supernatants were collected and treated as described for the
Lancefield extracts.
Mutanolvsin Extraction
This procedure is a modification of the method described by Yarnall
et al.. (1986). approximately 0.9 g wet weight of 64/14 was suspended in
5.0 ml of 20 mM KH2PO4, 1.0 mM EDTA, .02% NaN3 pH 7.0 containing 2.0 mM
PMSF, 10 /tg/ml DNAse I and 50 /ig/ml mutanolysin. The mutanolysin was
purified from a commercial product (Sigma) according to the method
described by Siegal et al., (1981). The suspension was vortexed and
placed at 37C for 4 hours with periodic vortexing. Supernatants were
collected and treated as described above for Lancefield extractions. An
enzyme control for use in the plasmin receptor assay contained 10 /ig/ml
DNase I, 2.0 mM PMSF, and 50 //g/ml mutanolysin in 20 mM Tris-HCl, 0.15 M
NaCl, 1.0 mM iodoacetic acid, 1.0 mM benzamidine HC1, 0.02% NaN3 pH 7.4.
Preparation of Immobilized Human Plasmin Affinity Column
Human plasminogen at a concentration of approximately 5.2 x 10'^ M
was activated to plasmin by incubating the sample in the presence of an
approximately 62 fold lower molar concentration of urokinase (Abbott).
The reaction volume was 10 ml and the primary buffer was 0.05 M Tris,
0.15 M NaCl pH 7.4 containing 40 mM lysine. Conversion was carried out
with constant agitation for one hour at 37C. A 50 /I aliquot was
removed and the remainder flash frozen and stored at -70C. The sample


Figure 5-3. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and streptokinase proteins. Parallel 10% SDS-polyacrylamide
gels were electrophoresed. The proteins in each gel were transferred to
separate nitrocellulose sheets, and one membrane was used in the solid-
phase plasminogen activator assay (Panel A), to identify the molecular
species of plasminogen activator molecules, as described in the Methods.
The second and third membranes were probed in a sandwich assay first with
polyclonal rabbit anti-group C streptokinase IgG (Panel B) or polyclonal
rabbit anti-plasmin receptor IgG (Panel C) followed by ^-1-Protein G,
as described in the Methods. The resulting blots (Panels B and C) were
autoradiographed for 6 hours at -70C with intensifying screens. Lane(s)
1: exracted plasmin receptor preparation (approx. 5.0 Mg); lane(s) 2:
approx. 2.0 Mg of 64/14 SK; lane(s) 3: approx. 2.0 Mg f B923 SK; lane(s)
4: approx. 1.0 Mg of A995 SK; lane(s) 5: approx. 2.0 Mg of Kabikinase;
lane(s) 6: approx. 2.0 Mg of ATCC 12449 SK; lane(s) 7: approx. 1.0 Mg of
26RP66 SK. Streptokinase proteins from the strains 64/14, B923, A995,
ATCC 12449, and 26RP66 were contained in concentrated culture
supernatants, see Methods for precise experimental details.


44
Binding of plasmin to its bacterial receptor does not inhibit the
ability of the enzyme to cleave either small synthetic substrates or its
natural substrate fibrin. Aprotinin, a naturally occurring tight-binding
inhibitor of plasmin and phe-pro-arg chloromethyIketone which chemically
modifies the histidine residue of the active site can react with the
bound plasmin and neutralize its enzymatic activity. These findings
suggest that the catalytic portion of the plasmin molecule is not
interfered with by the association with the bacteria. Of interest was
the observation that the enzymatic activity of bacterial bound plasmin
could not be regulated by addition of its specific inhibitor, o^-
antiplasmin. Alpha2-antiplasmin is a potent inhibitor of plasmin in the
fluid phase forming a 1:1 stoichiometric complex between the enzyme and
inhibitor.
The failure of a^-antiplasmin to regulate bacterial bound plasmin
provides the bacteria with a potential mechanism for tissue invasion by
virtue of the ability of plasmin to hydrolyze components of connective
tissue and basement membranes. Recent studies of the invasive
characteristics and metastatic potential of tumor cells has suggested a
key role for plasminogen activators in this process (Dano et al., 1985).
The ability of certain group A streptococci to produce a plasminogen
activator (e.g., streptokinase) and also to express a receptor for the
activation product plasmin may account for certain of its invasive
properties. Furthermore, since plasmin bound to a group A streptococcus
is incapable of inhibition by a^-antiplasmin the bacteria has associated
with it a non-regulatable proteolytic activity that may help to
contribute to its tissue invasive properties.
A variety of receptors for human proteins have been described on
streptococci. These include receptors for key components of the immune


127
Molecular weight determinations on Western blots were made possible
by the transfer of the prestained molecular markers.
For staining, gels were fixed in a solution of 40% ethanol and 10%
acetic acid, stained with Coomassie brilliant blue R-250 (0.25% w/v in
40% ethanol and 10% acetic acid) for 1 hour, and destained by soaking in
several changes of 10% ethanol and 10% acetic acid containing a small
quantity of DE 52 (Whatman, England) as a dye adsorbent. All other gels
prepared for staining were silver stained according to the procedure of
Merril et al., (1981).
Solid Phase Assay for Plasminogen Activator Activity
Samples to be tested for plasminogen activator activity by this
assay (Dr. K. Johnston, personal communication) are first resolved by
SDS-PAGE and transferred to nitrocellulose. The nitrocellulose membranes
are then immersed in blocking buffer (10 mM Tris, pH 8.0 containing 0.5%
Tween-20, 0.5 M NaCl and 1.0% bovine serum albumin) for at least one hour
at room temperature. The substrate overlay is prepared as follows: To a
2.0% agarose solution (Bio Rad Richmond, CA) in 0.15 M phosphate buffered
saline, pH 7.5 was equilibrated at 50C, with the chromogenic substrate
S-2251 at a concentration of 100 tg/ml. Human plasminogen free of
plasmin activity, is then added to a final concentration of 20 /ig/ml.
The agarose-substrate-plasminogen solution is then applied to an ethanol
washed glass slide slightly larger than the nitrocellulose membrane
template. The nitrocellulose membrane containing the sample is then
drained of excess blocking buffer and overlayed on the agarose-
plasminogen- substrate gel. It is important to ensure uniform contact
between the nitrocellulose membrane and the gel. The nitrocellulose
membrane is allowed to remain in contact with the agarose for at least


11
measuring the conversion of the single chain plasminogen molecule
(Mr=90,000) into heavy chains (Mr=60,000) and light chains
(Mr=25,000) as determined by the migration of radiolabeled proteins,
following denaturation and reduction, on 10% SDS-polyacrylamide gels.
The migration of labeled proteins was determined by autoradiographic
exposure of dried gels to Kodak XAR 5 film with intensifying screens at -
70C for 20 hours.
Bacteria
The group A /3-hemolytic streptococcal strain 64 had been previously
subjected to mouse passage as described by Reis et al., (1984). The
parent strain (64/P), as well as strains isolated after three (64/3) and
fourteen (64/14) mouse passages, were grown in either Todd-Hewitt broth
(DIFCO, Detroit, Mich.) or chemically defined media (Van De Rijn and
Kessler, 1980) overnight at 37C as stationary cultures (Yarnall and
Boyle, 1986b). The bacteria were harvested by centrifugation and
resuspended in phosphate-buffered saline (PBS), pH 7.4, containing 0.05%
Tween-20 and 0.02% sodium azide. The bacteria were heat killed at 80 C
for 10 minutes, a treatment that did not alter their plasmin binding
potential, but eliminated the production of soluble plasminogen
activators which would interfere with these studies. The suspension was
centrifuged, the pellet washed twice with PBS and then resuspended at 10%
wet weight/volume in PBS containing 0.05% Tween-20 and 0.02% sodium
azide. Samples were stored at -20C. The concentration of a bacterial
suspension was determined by counting bacterial chains in a Neubauer
hemacytometer (Fisher Scientific, Orlando, FL).
Determination of Binding of Radiolabeled Proteins to Bacteria
The light scatter at 550 nm was determined to standardize the
concentration of organisms used in subsequent tests. A light scatter


Mini-PLA
low molecular weight plasmin
Mini-PLG
low molecular weight plasminogen
mM
millimolar
fjM
micromolar
/im
micrometer
MOPS
(3-[N-Morpholino]propanesulfonic acid)
Mr
relative molecular weight
NIH u
National Institute of Health unit
nm
nanometer
PAGE
polyacrylamide gel electrophoresis
PBS
phosphate buffered saline
PEG
polyethylene glycol
PLA
plasmin
PLG
plasminogen
P
pico
PMSF
phenylmethylsulfonylfloride
pNpGB
p-nitrophenyl p-guanidinobenzoate HC1
PPACK
Phe- Pro-Arg-chloromethyIketone
RNase
ribonuclease
S-2251
H-D-Val-Leu-Lys-paranitroanilide
SDS
sodium dodecylsulfate
SK
streptokinase
tPA
tissue-type plasminogen activator
Tris
(Tris[hydroxymethyl]aminomethane
UK
urokinase
VBS-gel
Veronal-buffered saline plus gelatine
VPLCK
D-Val-Phe-Lys-chloromethyl ketone
xi


Figure 5-4. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and 64/14 streptokinase and group C streptokinase with mouse
polyclonal anti-plasmin receptor antibodies. Parallel protein samples
were electrophoresed in a 10% SDS-polyacrylamide gel. The proteins were
transferred to nitrocellulose membranes, and the membranes probed in
sandwich assays first with group II polyclonal mouse, anti-plasmin
receptor acities fluid (Panel A) or with group IV polyclonal mouse, anti-
plasmin receptor acities fluid (Panel B) or with polyclonal rabbit, anti-
plasmin receptor IgG followed by goat, anti-mouse IgG antibody (Panels A
and B) followed by Protein G. The resulting blots were auto-
radiographed at -70C for 6 hours with intensifying screens. Lane(s) 1:
approx. 5.0 /g of plasmin receptor protein; lane(s) 2: concentrated 64/14
culture supernatant containing approx. 2.0 ng of 64/14 streptokinase;
lane(s) 3: 4.0 ng of group C streptokinase. See Methods for precise
experimental details.


Figure 5-1. Functional identification and distinction of streptokinase
proteins and plasmin binding receptor protein. Parallel 10% SDS-
polyacrylamide gels were electrophoresed. The proteins on one gel were
transferred to nitrocellulose, and the membrane blocked and used in the
solid-phase plasminogen activation assay (Panel A) as described in the
Methods. The proteins in the second and third gel were transferred to
separate nitrocellulose sheets, and the membranes blocked and probed with
l-2^I-plasmin (Panel B) and '^I-plasmin in the presence of 1.0 mM EACA
(Panel C) as described in the Methods. The resulting blots (Panels B and
C) were autoradiographed at -70C for 24 hours with intensifying screens.
Lane(s) 1: exracted plasmin receptor preparation (approx. 5.0 fig)',
lane(s) 2: approx. 2.0 fig of 64/14 SK; lane(s) 3: approx. 2.0 ng of B923
SK; lane(s) 4: approx. 1.0 fig of A995 SK; lane(s) 5: approx. 2.0 fig of
Kabikinase; lane(s) 6: approx. 2.0 fig of ATCC 12449 SK; lane(s) 7:
approx. 1.0 fig of 26RP66 SK. Streptokinase proteins from the strains
64/14, B923, A995, ATCC 12449, and 26RP66 were contained in concentrated
culture supernatants, see Methods for precise experimental details.


This dissertation is dedicated
to the memory of my father
Thomas J. Broder


96
In order to determine the extent of coupling, the plasmin content of the
dialyzed sample (determined by means of absorbance at 280 nm using an
A?cmnm value of 17.0) was compared with the known starting plasmin
concentration. The efficiency of coupling was estimated to be 90%.
The Affi-Prep 10 Plasmin was loaded into an HR 10/10 FPLC
compatible column (Pharmacia). The affinity matrix was equilibrated in
0.05 M Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, 0.02% NaNj pH 7.2.
When not in use, the column was stored at 4C.
Affinity Purification of Plasmin Receptor
The Affi-Prep 10-Plasmin HR 10/10 column was attached to a Pharmacia
FPLC chromatography system and equilibrated at room temperature in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, 0.02% sodium azide pH 7.2.
1.0-2.0 ml of crude supernatant from the mutanolysin extraction of
bacterial strain of 64/14 (prepared as described above) was loaded onto
the column at a flow rate of 0.02 ml/min. The flow rate was increased to
a 1.0 ml/min rate during the washing step using equilibration buffer
(approx. 200 mis). The column was either eluted at 0.2 ml/min with a 50
ml linear gradient of 0.0 M 0.1 M L-Lysine in equilibration buffer, or
eluted in a single step using equilibration buffer containing 0.1 M L-
Lysine. The absorbance at 280 nm was continuously monitored and 1.0 ml
fractions were collected. After each affinity purification procedure the
column was washed with 20 mis of 2.0 M NaCl, followed by 200 mis of
equilibration buffer and stored at 4C.
Plasminogen Activation Assay for Streptokinase
The following assay (Zolton and Mertz, 1972; Teger-Nilsson et al..
1977) to measure SK activity was carried out in microtiter plates. 20 pi
aliquots of streptokinase standards (a dilution series for the purpose of


54
Preparation of Plasmin Heavy and Light Chains
Plasmin heavy and light chains were prepared essentially as
described by Summaria and Robbins (1976) Twenty mg of Lys-plasmin,
enzymatically inhibited with a 5 fold molar excess of aprotinin in 5 mis
of 0.05 M Tris, 0.1 M NaCl, pH 8.0, was reduced by treatment with 0.1 M
/3-mercaptoethanol for 20 min. at 20C. The reduced solution was then
cooled in an ice slurry and carboxymethylated with 0.1 M sodium-
iodoacetate on ice for 10 min. The plasmin heavy and light chains were
then separated and purified by a combination of affinity chromatography
on lysine-Sepharose, concentration by ammonium sulfate precipitation (4.0
g/10 ml), resuspension in 0.05 M Tris, 0.1 M NaCl, pH 8.0, and subjected
to gel filtration on Superse 6 (Pharmacia FPLC). The isolated plasmin
and plasminogen fragments were analyzed for purity on a reduced SDS-6 M-
urea-polyacrylamide gel. As shown in Figure 3-1, panel B, the various
fragments demonstrated appropriate molecular sizes and were homogeneous.
Concentrations were determined as described above. All proteins were
aliquoted and stored at -70C.
Direct Binding Assay of Radiolabeled Proteins
The ability of radiolabeled plasminogen fragments to bind to the
group A streptococcus 64/14 was measured as described previously by
Lottenberg et al.. (1987). A fixed number of bacteria were incubated
with labeled proteins (approximately 30,000 cpm per tube) in a total
volume of 400 nl of VBS-gel for 30 min at 37C. The bacteria were
pelleted by centrifugation at 1000 x g for 10 min and the pellets washed
twice with 2.0 ml of VBS-gel. The radioactivity associated with the
bacteria was determined in a Beckman 5500 Auto gamma counter (Beckman
Instruments, Inc., Fullerton, CA). Non-specific background binding was


169
Plow, E.F., D.E. Freaney, J. Plescia, and L.A. Miles. 1986. The
plasminogen system and cell surfaces: Evidence for plasminogen and
urokinase receptors on the same cell type. J. Cell. Biol.
103:2411-2420.
Rakoczi, I., B. Wiman, and D. Collen. 1978. On the biological
significance of the specific interaction between fibrin, plasminogen
and antiplasmin. Biochem. Biophys. Acta. 540:295-300.
Reddy, K.N.N., and G. Markus. 1972. Mechanism of activation of human
plasminogen by streptokinase. J. Biol. Chem. 247:1683-1691.
Reis, K.J., E.M. Ayoub, and M.D.P. Boyle. 1983. Detection of receptors
for the Fc region of IgG on streptococci. J. Immunol Methods.
59:83-94.
Reis, K.J., E.M. Ayoub, and M.D.P. Boyle. 1984. Streptococcal Fc
receptors: I. Isolation and partial characterization of the
receptor from a group C streptococcus. J. Immunol. 132:3091-3098.
Reis, K.J., G.O. Von Mering, M.A. Karis, E.L. Faulmann, R. Lottenberg,
and M.D.P. Boyle. 1988. Enzyme-labeled type III bacterial Fc
receptors: A versatile tracer for immunoassay. J. Immunol. Meth.
107:273-280.
Reis, K.J., M. Yarnall, E.M. Ayoub, and M.D.P. Boyle. 1984. Effect of
mouse passage on Fc receptor expression by a group A streptococci.
Scand. J. Immunol. 20:433-439.
Robbins, K.C. 1981. The regulation and control of the blood
fibrinolytic system. In: Progress in Chemical Fibrinolysis and
Thrombolysis, Vol. 5, pp. 3-13, Davidson, J.F., Nilsson, I. Mare
Astedt, B. (eds.), Churchill Livingstone, Edinburgh.
Robbins, K.C., and L. Summaria. 1970. Human plasminogen and plasmin.
Methods Enzymol. 19:184-199.
Robbins, K.C., L. Summaria, B. Hsieh, and R.J. Shah. 1967. The peptide
chains of human plasmin: Mechanism of activation of human
plasminogen to plasmin. J. Biol. Chem. 242:2333-2342.
Russell-Jones, G.J., E.C. Gotschlich, and M.S. Blake. 1984. A surface
receptor specific for human IgA on group B streptococci possessing
the Ibc protein antigen. J. Exp. Med. 160:1467-1475.
Ryan, T.J. 1987. Photoaffinity labeling of human plasmin and
plasminogen. Pharmac. Ther. 34:335-348.
Scatchard, G. 1949. The attractions of proteins for small molecules
and ions. Ann. N.Y. Acad. Sci. 51:660-672.
Scopes, R.K. 1982. Protein Purification: Principles and Practice,
Springer-Verlag, New York.


Figure 2-1. Binding of plasmin to bacteria: comparison of the kinetics
of generation of plasmin and its ability to bind to the group A
streptococcus 64/14: Labeled plasminogen was converted to plasmin by
treatment with urokinase. The kinetics of generation of plasmin was
monitored on SDS-polyacrylamide gels under reducing conditions. The
conversion of single chain, high molecular weight, plasminogen (Mr
approximately 90,000) into heavy (Mr approximately 60,000) and light
chains (Mr approximately 25,000) of plasmin was monitored. At each time
point the ability of labeled proteins to bind to the group A
streptococcus 64/14 was measured as described in the Methods. The data
are presented as the mean the standard deviation of duplicate
experiments.


16
bacteria were pelleted by centrifugation and washed twice with 2.0 ml of
VBS-gel. After determining the radioactivity associated with the
bacteria, the pellets were resuspended in solutions of VBS-gel containing
varying concentrations of amino acid or amino acid analogs (pH 7.0) as
described above. The mixtures were incubated at 37 C for 15 minutes and
washed twice with VBS-gel containing the appropriate amino acid
concentration. The radioactivity associated with the bacteria in
duplicate samples was again measured and the percentage dissociated was
calculated.
Determination of Kp and Receptor Density
Labeled plasmin (25,000 to 250,000 cpm) in 100 /I of VBS-gel was
added to 3 x 10^ bacteria in 300 pi of VBS-gel, pH 7.4, and incubated at
37C for 15 minutes. The bacterial suspensions were centrifuged at 1,000
x g for 10 minutes and washed twice with 2.0 ml of VBS-gel. All
determinations were performed in triplicate. Total binding was
determined by measuring the radioactivity associated with the bacterial
pellet when only labeled plasmin was offered. Non-specific binding was
determined by pre-incubating bacteria at 37 C for 15 minutes in VBS-gel,
pH 7.4, containing unlabeled plasmin at a 100-fold molar excess of the
labeled plasmin. Specific binding was calculated by subtracting
non-specific binding from total binding for each amount of labeled
plasmin offered. The amount of free labeled plasmin was calculated by
subtracting the amount of specifically bound labeled plasmin from the
total amount of labeled plasmin offered.
The apparent dissociation constant (Kp) was determined by two
methods. A non-linear least squares analysis of the total counts offered
vs. the counts bound fit to the simple Michaelis-Menten equation was


130
nitrocellulose-Ag in a non-Freund's adjuvant (T1501). The vehicle of
this adjuvant has been described by Woodard and Jasman (1985), and the
adjuvant (T1501) has been described by Hunter and Bennett (1984). The
use of this combination was developed by Woodard and Jasman. Group 4:
This group was injected with (100/il/50/ig) of sonicated nitrocellulose-Ag
in Ribi Adjuvant intraperitoneally. Part II first boost (2 weeks after
initial inoculation): Mice in Groups 1+3 were boosted with (100/il/10pg)
of sonicated nitrocellulose-Ag in incomplete Freund's Adjuvant in two
sites subcutaneously. Mice in group 2 were boosted with (100tl/10/g) of
sonicated nitrocellulose-Ag in the non-Freund's adjuvant (Woodard and
Jasman, 1985) subcutaneously. Mice in group 4 were boosted with
(lOO/l/10/ig) of sonicated nitrocellulose-Ag in Ribi Adjuvant
intraperitoneally. Part III second boost (4 weeks after initial
inoculation): Mice in group 1 were boosted with (lOO/il/10/xg) of
solubilized nitrocellulose-Ag in dimethylsulfoxide intraperitoneally.
Mice in group 2 were boosted with (100/il/10^g) of sonicated
nitrocellulose-Ag in dd^O plus glycogen intraperitoneally. Mice in
group 3 were boosted with (100tl/10tg) of sonicated nitrocellulose-Ag
plus glycogen intraperitoneally. Mice in group 4 were boosted with
(100/il/10/ig) of sonicated nitrocellulose-Ag in dd^O intraperitoneally.
Four days after the final boost all mice were sacrificed and ascites
harvested. The peritoneal cavities were washed with PBS and also
harvested, and the resulting fluids from each group pooled and stored as
aliquots at -70C, with 0.02% sodium azide, until used.
Results
Streptokinase from three group A Streptococcal strains (64/14, B923
and A995), and two group C streptococcal strains (ATCC 12449 and 26RP66),


13
fold molar excess of either PPACK, aprotinin, or pNpGB for 10 minutes at
room temperature. The volume of each sample was increased by the
addition of 200 /I of VBS-gel, and each was applied to a separate G-25
column (PD 10 Pharmacia Fine Chemical) to remove free inhibitor. Five
hundred /I fractions were collected and counted in an autogamma counter
to localize the modified ^^1 plasmin. Fractions containing the labeled
protein-inhibitor complexes were pooled. Aliquots of the three labeled
complexes were mixed with a 10-fold molar excess of c*2-AP in a final
reaction volume of 400 /il for 10 minutes at room temperature. Plasmin
and plasminogen were included as controls. The volumes of each solution
were adjusted to 1.1 ml with VBS-gel and then used in a direct binding
assay to group A streptococci. Each of the plasmin-inhibitor samples
that has been incubated with excess a^'AP was monitored on non-reducing
SDS-polyacrylamide gels as described by Weber and Osborn (1969) for the
formation of a high molecular weight complex.
Determination of Plasmin Activity While Bound to Bacteria
To five microtubes, each containing 100 /il of a 10% w/v solution of
the group A streptococcus, 64/14, in a total volume of 400 /I VBS-gel, 10
nM plasmin was added and allowed to bind for 40 minutes at 37C. Five
other tubes containing plasmin but no bacteria and one tube containing
bacteria alone were prepared as controls. The bacteria were pelleted and
washed twice with 1.0 ml VBS-gel and resuspended in 400 /il VBS-gel
containing a 10-fold molar excess of either pNpGB, PPACK, aprotinin, a^-
AP or buffer alone. All samples were then incubated for 15 minutes at
room temperature. The samples containing bacteria were pelleted by
centrifugation, washed with 1.0 ml of VBS-gel and resuspended with
vigorous vortexing in 400 /il VBS-gel.


Page
V. COMPARISON OF THE GROUP A STREPTOCOCCAL RECEPTOR FOR
HUMAN PLASMIN WITH STREPTOKINASE 122
Introduction 122
Materials and Methods 123
Results 130
Discussion 146
VI. SUMMARY AND CONCLUSIONS 153
REFERENCES 162
BIOGRAPHICAL SKETCH 174
v


45
system such as Clq (Yarnall et al.. 1986), IgG (Kronvall, 1973), and IgA
(Russell-Jones et al., 1984), the serum protein fibrinogen (Kronvall et
al., 1979), and the stromal structural proteins laminin (Switalski et
al., 1984), and fibronectin (Myhre and Kuusela, 1983). The importance of
any of these receptors in the pathogenic process remains controversial.
Our recent observation that streptococci also have a receptor that is
specific for human plasmin adds to this list (Lottenberg et al., 1987).
Although the primary substrate for plasmin is fibrin, plasmin is a
non-specific protease capable of also hydrolyzing such extracellular
matrix proteins as thrombospondin, fibronectin, and laminin, while also
exposing matrix components for degradation by other enzymes (Knudsen et
al., 1986). The ability to bind plasmin in an active form which can not
be regulated by its efficient physiological regulator o^-antiplasmin
could provide for surface mediated protease activity and a mechanism for
tissue invasiveness by plasmin receptor-positive bacteria (Lottenberg et
al., 1987).
In this study it was shown that the group A streptococcus strain
64/14 demonstrates optimal binding of its ligand at physiological pH and
ionic strength. The interaction had a high affinity (Kp =5 x 10'^ M)
and demonstrated a linear Scatchard plot indicating that a single
population of plasmin receptors was present on the bacteria (Broeseker et
al., 1988). There was no evidence for either an additional low affinity
receptor or for any cooperativity, positive or negative, in the binding
of ligand with the specific receptor. In agreement with our previous
observations (Lottenberg et al., 1987), there was no evidence for
specific interaction between bacteria and the native zymogen form of the
protein, plasminogen.


A280
112
FRACTION NUMBER
JJlL; -i-
' -#'..!*
; bi:j i'i'I' ?
60
50
40
30
20
10
i
L-LYSINE (mM)


64
differences in the efficiency of inhibition. The isolated heavy chain
was found to be less efficient an inhibitor than the intact plasmin
molecule. These findings suggest that there is some component involved
in the interaction of plasmin with the bacteria that is either not
present on the heavy chain or is altered during the isolation procedure.
Two possibilities to account for these observations were considered.
The first was that there are some sites on the heavy chain of the
plasmin molecule that are modified when the molecule is purified, thereby
changing its efficiency of interaction with the bacterial receptor. The
second was that the plasmin light chain, while associated with the heavy
chain, confers a different tertiary structure to the molecule than exists
on either (or both) of the isolated chains. Such a change in
conformation of the molecule might affect its interaction with the
bacteria.
It has been established previously that a conformational change
occurs when Glu-plasminogen is activated to Lys-plasmin, or when Glu-
plasminogen is converted to Lys-plasminogen (Swenson and Thorsen, 1981;
Markus et al.. 1978; Thorsen, 1975). Lys-plasminogen is the zymogen form
of plasminogen, lacking the 76 amino acid NH2-terminus of the native
protein (Markus et al., 1978). This modification results from the
proteolytic activity of plasmin on Glu-plasminogen, which removes the 76
amino acid NH2-terminus, resulting in a new NH2-terminus lysine (for
review, see Thorsen et al., 1981). This modification occurs without
generation of protease activity. The conversion of Glu-plasminogen to
Lys-plasmin or to Lys-plasminogen not only results in a marked
conformational change of the protein but also causes an increase in the
binding affinity of these molecules to fibrin (Thorsen, 1975), as well as


149
explore any possible similarities or differences. Streptokinase is a
unique plasminogen activator. Unlike eukaryotic plasminogen activators,
streptokinase has no enzymatic activity. The activation mechanism lies
in its ability to form a specific 1:1 stoichiometric complex with
plasminogen, as well as with plasmin, which leads to the formation of an
active enzyme moiety presumably through conformational changes in the
plasminogen molecule without the cleavage event at the Arg560'Va]_561
peptide bond (Markus et al., 1976), that can in turn act as a plasminogen
activator for other plasminogen molecules. This is a function neither of
the two proteins alone possesses.
The data presented in Chapters Two, Three, and Four provides good
evidence that the surface associated plasmin binding receptor activity is
distinct from the streptococcal plasminogen activator (streptokinase).
While both the plasmin receptor and streptokinase display a high affinity
interaction towards plasmin; the plasmin receptor specifically recognizes
plasmin while demonstrating no significant reactivity with the native
zymogen precursor Glu-plasminogen (see Chapters Two and Three). By
contrast streptokinase displays similar affinity for plasminogen and
plasmin.
This investigation focused on obtaining plasmin receptor molecules
and secreted streptokinase from a single streptococcal strain 64/14
culture, and comparing the two molecules by three criteria: (1)
functionally, by measuring plasminogen activator activity; (2) plasmin
binding specificity using (a) -plasmin with and without EACA,
(b) plasmin heavy chain and (c) plasmin light chain as probes; and (3)
antigenically, by the production of a polyclonal anti-plasmin receptor
antibodies as well as testing a series of mouse anti-streptokinase


Figure 4-1. Dot-blot analysis of solubilized plasmin binding activities.
Samples to be tested for plasmin binding activity were immobilized on
nitrocellulose in a Bio-dot blot apparatus. The nitrocellulose membrane
was blocked as described in the Methods and probed with -plasmin.
The blot was washed and autoradiographed at -70C for 10 hours using
Kodak X-AR-5 film and an intensifying screen. Row A: columns 2 through
5: Kabikinase 400, 300, 200, and 100 ng respectivly; column 6: trypsin
extraction (60 min); column 7: trypsin enzyme control. Row B: column 1:
lysozyme/detergent/shock extract; column 2: lysozyme/detergent/shock
control; column 3: mutanolysin extract; column 4: mutanolysin enzyme
control; column 5: hot-alkali extract; column 6: hot-acid extract;
column 7: acetone/detergent extract.


Figure 3-1. SDS-UREA-PAGE analysis of isolated olasmin(ogen) fragments.
Panel A: Elastase digestion fragments of plasmin(ogen): Glu-plasminogen
was digested with elastase and fragments purified as described in Materials
and Methods. Panel A: Lane 1: Mini-PLG (4.0 Mg); 2: Glu-PLG (4.0 Mg); 3:
Mini-PLA (4.0 Mg); 4: LBS-I (4.0 Mg); 5: LBS-II (4.0 Mg); M: molecular weight
standards. Panel B: Plasmin heavy (HC) and light (LC) chain preparations:
Lys-plasmin was reduced and carboxymethylated as described in Materials and
Methods. Lane 1: Glu-PLG (5.0 Mg) I 2: lys-PLA (5.0 Mg); 3: HC (5.0 Mg) I t*'-
LC (5.0 Mg)- Proteins were electrophoresed under reducing conditions on a
SDS-6 M urea-10%-polyacrylamide gel.


123
Materials and Methods
Materials
Nitrocellulose was purchased from BioRad, Richmond CA. Affinity
purified goat anti-mouse IgG, heavy and light chains specific, was
purchased from Cappel, Organon Teknika Corp. West Chester, PA. Ribi
Adjuvant System (RAS) was purchased from Ribi Immunochem Research Inc.
Hamilton, MT. Mouse anti-group C streptokinase monoclonal antibodies
were a gift from Dr. K. Johnston. All other chemicals and reagents were
purchased from Sigma Chemical Co. St. Louis, MO.
Bacterial Strains
The Lancefield group A /3 hemolytic streptococcal strains 64/14,
B923, and A995; and the Lancefield group C /3 hemolytic streptococcal
strain 26RP66 and the ATCC strain 12449 were grown as stationary cultures
at 37C for 24 to 36 hrs, in one to two liter batches of a chemically
defined media described by Van De Rijn and Kessler (1980), containing
0.1% phenol red. The pH of the culture was maintained above 7.0, as
monitored by the indicator dye. Approximately 3.0 to 5.0 g (wet weight)
of bacteria could be recovered per liter of media. Bacteria were
harvested by centrifugation, resuspended in phosphate-buffered saline
(PBS), pH 7.4, containing 0.02% sodium azide. The bacteria were heat
killed at 80C for 15 min. The suspension was centrifuged and the pellet
washed three times with PBS containing 0.02% sodium azide. Aliquots
could be stored at -20C, or used immediately for extraction purposes.
Bacterial Culture Supernatants
Culture supernatants from the various strains of bacteria were
recovered following removal of the bacteria by centrifugation. The
supernatants were filtered through a 0.22 /m filter and immediately


158
In agreement with the earlier data, the extracted plasmin receptor
preparation did not possess plasminogen activator activity. The
streptokinase proteins from all strains examined, including strain 64/14,
were identified using a solid-phase plasminogen activator assay. This
assay also allowed the molecular weight of the plasminogen activator to
be defined. The 41,000 dalton plasmin receptor lacked the ability to
activate plasminogen. A further distinction between the streptokinase
proteins and the plasmin receptor was observed when the proteins were
probed with ^i-plasmin in the presence of EACA. All of the
streptokinase proteins examined bound ^^^1-plasmin equally well in the
presence of EACA while the bacterial plasmin receptor failed to bind
plasmin in the presence of EACA.
The antigenic relatedness of the plasmin receptor and streptokinase
was compared using polyclonal antibodies prepared in rabbits and mice to
the purified 41,000 dalton plasmin receptor. Mouse and rabbit polyclonal
antibodies were prepared that could recognize the 41,000 dalton protein
while showing no reactivity with the 48,000 dalton plasminogen activator
molecule (streptokinase) produced by the same group A isolate from which
the plasmin receptor activity had been isolated. The 48,000 dalton,
strain 64/14, streptokinase was found to be reactive with a rabbit
polyclonal antibody to group C streptokinase. This antibody recognized
the plasminogen activator molecule (streptokinase) produced by three
group A and two group C isolates that were studied. This antibody failed
to recognize any epitopes on the 41,000 dalton plasmin receptor molecule.
Furthermore none of 16 monoclonal antibodies prepared to different
epitopes of a group C streptokinase reacted with the plasmin receptor.
Thirteen of these antibodies recognized epitopes on the strain 64/14


38
Scatchard plots of 64/P and 64/3, like that shown for 64/14 in Figure 2-
6, demonstrated only a single class of plasmin receptors expressed on
these bacteria (Broeseker et al., 1988). Plasmin(ogen) contains lysine
binding sites which also bind analogous amino acids (Winn et al., 1980).
Epsilon-aminocaproic acid (EACA) approximates the side chain structure of
lysyl residues incorporated in intact proteins and has higher affinity
than lysine for these sites on plasmin(ogen), whereas arginine binds with
lower affinity (Winn et al., 1980). In order to assess the possible role
of the lysine binding sites of plasmin in its interaction with the
bacterial receptor, the binding of plasmin to the group A streptococcus
strain 64/14 in the presence of EACA, lysine, or arginine was determined.
Binding was measured in 0.15 M VBS-gel, pH 7.4, containing amino acid
in increasing concentrations. The percentage inhibition of binding was
determined by comparison with the binding in VBS-gel pH, 7.4, buffer
alone. The results of these studies are presented in Figure 2-7 and
demonstrate that binding of plasmin to the group A streptococcus 64/14
was inhibited by each amino acid in a concentration dependent fashion.
Fifty percent inhibition of binding of plasmin to the bacteria was
observed at an EACA concentration of 0.15 mM, a lysine concentration of
2.0 mM, and an arginine concentration of 25 mM. In similar studies,
plasmin was pre-bound to the group A streptococcus and a concentration
dependent elution of bound radiolabel was observed on incubation with
EACA, lysine, or arginine (Figure 2-8). The concentration of amino acid
required to elute 50% of the bound plasmin was approximately equivalent
to that required to inhibit plasmin binding by 50% (compare Figures 2-7
and 2-8).


Figure 2-8. Elution of 125I-plasmin from group A streptococcal strain
64/14 in VBS-gel containing various concentrations of
epsilon-aminocaproic acid, lysine, and arginine : Measurements of
duplicate experiments were performed and the data are presented as the
mean the standard deviation. Percent eluted was calculated by
comparing the radioactivity associated with the bacterial pellet before
and after incubation in the given amino acid solution. (O O)
epsilon-aminocaproic acid; (O O) lysine; (0-*-*-0) arginine.


O.D. 405nm
76
PLASMINOGEN (nM)


110
The Affi-Prep 10-plasmin HR 10/10 column was then either attached to
a Pharmacia FPLC chromatography system and used at room temperature, or
used at 4C with a peristaltic pump and fraction collector (Pharmacia).
The mutanolysin extract was dialyzed into 0.05 M Na2HP04, 0.15 M NaCl,
1.0 mM benzamidine HC1, and 0.02% NaN3 pH 7.4. prior to chromatography.
1.0-2.0 mis of cell free mutanolysin extract of strain 64/14 was
routinely applied to the blocked plasmin affinity coliman matrix in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaN3 pH 7.4.
After loading the extract onto the column, the matrix was washed with
this buffer until the OD 280 nm returned to base line absorbance
(approximately 20 mis). Bound plasmin receptor activity was eluted with
either a 50 ml linear gradient of 0.0 0.1 M L-Lysine in 0.05 M Na2HP04,
0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaNj pH 7.4, or in a
single step using buffer containing 0.1 M L-lysine. The absorbance at 280
nm was continuously monitored and 1.0 ml fractions were collected.
Fractions eluted from the affinity column were assayed by the dot
blotting and ^^1-plasmin probing method. The plasmin binding functional
activity eluted from the column correlated to absorbance at 280 nm (see
Figure 4-4). This material was then analyzed by SDS-PAGE followed by
either silver or Coomassie brilliant blue R-250 staining, and Western
blotting and probing with iodinated plasmin. The results shown in Figure
4-5 indicate silver stained band at 41,000 daltons eluted from the
affinity column with lysine. This band corresponds to the plasmin
binding activity seen in the crude extract. The plasmin affinity
purified material was then analyzed by Western blotting and probing with
125
I-plasmin. Figure 4-6, Panel A, shows another sample of plasmin
affinity purified receptor, eluted using 0.1 M lysine in a single step.


Figure 3-5. Binding of Lvs-plasmin(s) derived from Glu-plasminogen and
Lvs-plasminogen. to the group A streptococcal strain 64/14 as measured
by residual activity in the bacterial free supernatant: (O -urokinase
activated Glu-plasminogen alone; -urokinase activated Glu-plasminogen
+ bacteria; -urokinase activated Lys-plasminogen alone; -urokinase
activated Lys-plasminogen + bacteria). For precise experimental details
see Materials and Methods.


47
a2_antiplasmin occurs in a two-step process (Christensen and Clenunenson,
1977; Wiman and Collen, 1978). The first step is a non-covalent
association of the a^-antiplasmin molecule with the lysine binding site
located in the kringle 1 region of the plasmin molecule, followed by a
rapid covalent linkage to the plasmin active site serine residue. The
observation that lysine or EACA effects, on the high affinity site of
plasmin, disturbs the plasmin bacterial receptor interaction, together
with the observation that bound plasmin is not inhibited by a^-
antiplasmin, suggests that the kringle 1 domain may be important in the
binding of plasmin to bacteria, or may be inaccessible.
Previous studies have shown that passage of streptococci in mice
heightened virulence with concomitant enhanced expression of certain
surface proteins (Burova et al., 1980; Burova et al., 1981; Reis et al.,
1984). Group A streptococcal strain 64 exhibits decreased expression of
Fc receptors after 3 or 4 mouse passages (strain 64/3 and 64/4) as
compared to the parent strain (64/P) (Reis et al., 1984). Following 8
mouse passages this group A streptococcus demonstrates markedly enhanced
Fc receptor expression which appears to be a stable characteristic of the
selected strain (Reis et al., 1984). In this study the average affinity
of the plasmin receptor expressed on strains 64/P, 64/3, and 64/14 was
not significantly different, indicating that mouse passage did not have a
major selective pressure on the affinity of the plasmin receptor.
There was some variation in the number of plasmin receptors
calculated for each bacteria. The 64/P, 64/3, and 64/14 strains
displayed 200, 3500, and 800 receptor sites, respectively, per bacterium.
Clearly these are average numbers of receptors per bacterium (Broeseker
et al. 1988), and given the potential errors in estimating bacterial


88
u/ml) was added to a 400 /tl solution of 1 tM plasminogen containing 0.04
M lysine. The mixture was incubated at 37C for 45 minutes unless stated
otherwise. The efficiency of plasmin generation was followed by
measuring the conversion of the single chain plasminogen molecule (Mr=
90,000 daltons) into heavy chains (Mr=60,000 daltons) and light chains
(Mr=25,000 daltons) as determined by the migration of radiolabeled
proteins, following reduction, on 10% SDS-polyacrylamide gels. The
migration of labeled proteins was determined by autoradiographic
exposure of dried gels to Kodak XAR 5 film with intensifying screens at
-70C for 15-20 hours.
Direct Binding Assay of Radiolabeled Proteins
The ability of radiolabeled plasmin(ogen) to bind to the group A
streptococcus 64/14 was measured as described previously by Lottenberg et
al. (1987). A fixed number of bacteria were incubated with labeled
proteins (approximately 30,000 cpm per tube) in a total volume of 400 /tl
of VBS-gel for 30 min at 37C. The bacteria were pelleted by
centrifugation at 1000 x g for 10 min and the pellets washed twice with
2.0 ml of VBS-gel. The radioactivity associated with the bacteria was
determined in a Beckman 5500 Auto gamma counter (Beckman Instruments,
Inc., Fullerton, CA). Non-specific background binding was determined in
replicate tubes which contained no bacteria. All estimates were
performed in duplicate.
Dot-blotting Procedure for the Identification of Plasmin Receptor
Activity
This assay was carried out with a Bio-Rad bio-dot microfiltration
apparatus using a modification of the Bio-Rad procedure. A piece of
nitrocellulose pre-equilibrated in PBS-azide for a minimum of 10 minutes
was fitted into the apparatus. The wells were loaded with 100 tl


42
#
.01


27
ALONE BOUND BOUND BOUND BOUND BOUND
PLASMIN PLASMIN PLASMIN PLASMIN PLASMIN
+
+
PNPGB
PPACK APROTININ
ajpi


Figure 3-6. Binding of Glu- and Lvs- plasminogen to the group A
streptococcal strain 64/14 as measured by residual activatable zymogen
in the bacterial free supernatant: ( Glu-plasminogen + bacteria;
o-Glu-plasminogen alone bacteria; -Lys-plasminogen + bacteria; o-Lys-
plasminogen alone). For precise experimental details see Materials and
Methods.


17
performed as described by Lottenberg et al.. (1985). The concentration
of plasmin was determined by converting counts per minute to moles using
the known specific activity for the labeled plasminogen. Scatchard
analysis (Scatchard, 1949) of these data was also performed as described
by Lottenberg et al.. (1987). Counts bound vs. counts bound/counts free
was plotted and the slope (representing -1/K¡)) was determined by linear
regression. The X-intercept (counts bound) was converted to moles of
plasmin. To determine the receptor densities the number of moles of
plasmin bound was determined by extrapolating the Scatchard plot and
determining the intercept. This represented the maximal binding of
plasmin to a known number of bacteria (derived by hemacytometer chamber
counts).
Plasmin which had been bound to and eluted from strain 64/14 by
treatment with lysine was also examined in similar binding studies.
Eluted plasmin was obtained by incubating 2.0 ml of stock 10% wet
weight/volume bacterial suspension (strain 64/14) with approximately 20
Hg of labeled plasmin at room temperature for 45 minutes. This
suspension was centrifuged at 1,000 x g for 10 minutes and washed once
with 10 ml of VBS-gel, and the radioactivity associated with the
bacterial pellet was measured. The pellet was then resuspended in
VBS-gel containing 20 mM lysine and incubated at room temperature for 30
minutes. The suspension was centrifuged and the supernatant recovered.
Approximately 90% of the radioactivity originally associated with the
bacterial pellet was dissociated by the lysine treatment. The dissociated
plasmin in the supernatant was then subjected to gel filtration on a G-25
column to separate lysine from plasmin. Fractions containing plasmin
were collected and stored at -20C.


165
Ginsburg, I. 1972. Mechanisms of cell and tissue injury induced by
group A streptococci: Relation to post-streptococcal sequelae. J.
Infect. Dis. 126:419-456.
Groskopf, W.R., L. Summaria, and K.C. Robbins. 1969. Studies on the
active center of human plasmin: Partial amino acid sequence of a
peptide containing the active center serine residue. J. Biol.
Chem. 244:3590-3597.
Hajjar, K.A., N.M. Hamel, P.C. Harpel, and R.L. Nachman. 1987. Binding
of tissue plasminogen activator to cultured human endothelial cells.
J. Clin. Invest. 80:1712-1719.
Hajjar, K.A., P.C. Harpel, E.A. Jaffe, and R.L. Nachman. 1987. Binding
of plasminogen to cultured human endothelial cells. J. Biol. Chem.
261:11656-11662.
Haupt, H., and Heimburger, N. 1972. Humanserumproteine mit hoher
affinitat zu carboxymethylcellulose I. Hoppe-Seyler's Z. Physiol.
Chem. 353:1125-1132.
Hayes, M.L., and F.J. Castellino. 1979a. Carbohydrate of human
plasminogen variants-I. Carbohydrate composition, glycopeptide
isolation and characterization. J. Biol. Chem. 254:8768-8771.
Hayes, M.L., and F.J. Castellino. 1979b. Carbohydrate of human
plasminogen variants-II. Structure of the asparagine-linked
oligosaccharide unit. J. Biol. Chem. 254:8772-8776.
Hayes, M.L., and F.J. Castellino. 1979c. Carbohydrate of the human
plasminogen variants -III. Structure of the O-glucosidically linked
oligosaccharide unit. J. Biol. Chem. 254:8777-8780.
Holvoet, P., H.R. Lijnen, and D. Collen. 1985. A monoclonal antibody
specific for lys-plasminogen. J. Biol. Chem. 260:12106-12111.
Huang, T-T., H. Malke, and J.J. Ferretti. 1989. Heterogeneity of the
streptokinase gene in group A streptococci. Infec. Immun.
57:502-506.
Hunter, R.L., and B. Bennett. 1984. The adjuvant activity of nonionic
block polymer surfactants. II. Antibody formation and inflammation
related to the structure of triblock and octablock copolymers. J.
Immunol. 133:3167-3175.
Johnston, K.H., and J.B. Zabriskie. 1986. Purification and partial
characterization of the nephritis strain-associated protein from
Streptococcus pyogenes group A. J. Exp. Med. 163:697-711.
Johnston, R.B. 1984. Current concepts: recurrent bacterial infection
in children. N. Engl. J. Med. 310:1237-1243.
Jones, P.A., and Y.A. DeClerck. 1980. Destruction of extracellular
matrices containing glycoproteins, elastin, and collagen by
metastatic human tumor cells. Cancer Res. 40:3222-3227.


UNIVERSITY OF FLORIDA


CHAPTER SIX
SUMMARY AND CONCLUSIONS
This dissertation describes the first report of a prokaryotic
cellular receptor for the human fibrinolytic enzyme plasmin. This study
was designed to characterize the plasmin receptor more completely in
order to explore what role, if any, it might play in streptococcal
infections. The specific goals of the study were to: (1) identify and
characterize a group A streptococcal receptor for human plasmin; (2) map
the binding site on human plasmin recognized by the bacterial plasmin
receptor; (3) isolate and purify a functionally active group A
streptococcal plasmin receptor; and (4) compare the group A streptococcal
plasmin receptor to the known streptococcal plasmin(ogen) binding protein
streptokinase.
Plasmin is the key component of the mammalian fibrinolytic enzyme
system which is responsible for intravascular blood clot lysis. The two
chain (heavy and light) plasmin serine protease, which cleaves fibrin, is
derived from the circulating single chain zymogen precursor Glu-
plasminogen. This derivation is brought about by plasminogen activators,
which cleave a single peptide bond in the plasminogen molecule (see
Chapter One for a detailed discussion of plasminogen activation). The
light chain of the plasmin molecule contains the enzyme active site
(Robbins and Summaria, 1970; Wiman, 1977). The plasmin heavy chain
contains five homologous triple loop structures known as kringles, which
are responsible for fibrin binding (Sottrup-Jensen et al., 1978).
153


139
possibility that the two proteins arose from a common precursor and
contained at least some common antigenic determinants was considered.
This possibility was examined using a series of polyclonal and monoclonal
antibodies to streptokinase, or rabbit and mouse polyclonal antibodies to
the isolated strain 64/14 plasmin receptor.
A polyclonal rabbit antibody to group C streptokinase (Kabikinase)
was prepared. This antibody could successfully be used in an ELISA
assay system to measure streptokinase in culture supernatants of both
group C and group A streptococcal bacteria (Reis et al.. 1988). This
polyclonal rabbit antibody to the extracted plasmin receptor was
prepared using the same immunization procedure, see Methods for precise
details. Using these polyclonal antibodies the antigenic relationship of
the strain 64/14 plasmin receptor to the streptokinase produced by strain
64/14 as well as the other streptokinases examined was assessed by SDS-
PAGE and Western blotting; probing with the two different polyclonal
antibodies followed by ^¡>1 _labeled Protein G. The results of these
experiments are presented in Figure 5-3. Panel A of Figure 5-3 shows
that the polyclonal rabbit anti-group C streptokinase antibody could
detect efficiently all the streptokinases examined in concentrated
culture supernatants from all the streptococcal isolates tested including
the group A strain 64/14. The amount of streptokinase in each lane was
approximately 1-2 /g, as measured by the fluid phase plasminogen
activator assay. No reactivity was seen in the lane containing the
41,000 dalton extracted plasmin receptor protein. The receptor lanes
contained approximately 5 ig of the 41,000 dalton protein, and was
readily detected when a parallel gel was probed with the polyclonal
rabbit anti-plasmin receptor antibody (see Panel B of Figure 5-3). This


12
reading of 0.3 corresponded to approximately 2 x 10^ organisms/ml
(Yarnall and Boyle, 1986b). A standardized number of bacteria
(approximately 10^ organisms) were incubated with labeled proteins
(approximately 30,000 cpm/tube) in a total volume of 400 /xl of VBS-gel
for 1 hour at 37C. The bacteria were pelleted by centrifugation at 1000
x g for 10 minutes and washed twice with 2.0 ml VBS-gel. The
radioactivity associated with the bacteria was determined in a Beckman
5500 autogamma counter. All estimates were carried out in duplicate.
Fibrin Plate Assay
Fibrin plates were prepared using 5 cm diameter disposable petri
dishes. Ten ml of 0.1% fibrinogen in PBS were clotted with 0.2 ml of
bovine thrombin (10 NIH u/ml) in 0.5 M CaCl2. Twenty pmoles of plasmin
were bound to 100 1 of a 10% w/v solution of the group A streptococcus,
64/14, in a total volume of 400 tl of VBS-gel. The mixture was incubated
at 37C for 45 minutes. A parallel series of samples containing bacteria
with no added plasmin served as the negative control. Fifty ti of a
suspension of bacteria or bacteria plus plasmin were placed either
directly onto a fibrin plate or onto a 0.22 /tm Millipore filter placed
between the bacteria and the fibrin plate. The plates were incubated for
20 hours at 37C and the degree of hydrolysis was scored by measuring the
area of the zone of clearing from the underside of the plate. In each
experiment a control of free plasmin was included and each estimate was
carried out in duplicate.
Plasmin-Inhibitor Complex Generation
Plasmin was generated from plasminogen as described above. Three
130 1 aliquots of the labeled enzyme were placed into separate
microtubes (approximately 2 x 10^ cpm/tube), and incubated with a 200-


150
monoclonal antibodies. All of these studies were carried out using a
Western blotting approach to enable specific molecular species to be
identified and compared in each of these assays. In addition to using
the group A strain 64/14 as a source of streptokinase, five other sources
(three group A and 2 group C) were included for comparison. In all
studies, the commercially available, highly purified group C
streptokinase (Kabikinase) was included as a reference.
In agreement with my earlier work, the extracted plasmin receptor
preparation did not possess plasminogen activator activity (Table 5-1).
This lack of plasminogen-activator potential was further demonstrated
using a solid-phase plasminogen-activator assay (see Panel A of Figure 5-
1) in which plasminogen activator molecules could be identified and
correlated to a molecular size. The assay identified the streptokinase
produced by strain 64/14, having an Mr of 48,000 daltons which comigrated
with Kabikinase and 3 of the other streptokinases examined. The group A
strain B923 secreted a slightly smaller streptokinase molecule (see Panel
A of Figure 5-1). When these preparations were Western blotted and
probed with '-^^I.plasmin the 41,000 dalton plasmin binding activity
could be clearly identified (see Panel B of Figure 5-1) and shown to
exhibit the sensitivity to EACA previously reported (see Panel C of
Figure 5-1). However, the streptokinase proteins including Kabikinase,
and the strain 64/14 streptokinase were not inhibited from binding ^^I-
plasmin in the presence of EACA (Compare Panels B and C of Figure 5-1).
The results of probing these Western-blotted proteins with either
^^^1-plasmin heavy chain or ^-*1 -plasmin light chain were not as
conclusive. Previous data (see Chapter Three) suggested that the 41,000
dalton protein would be detected preferentially by the heavy chain


91
IgG per ml of probing solution (approximately a 1:3000 dilution of
antisera) for three hours, washed twice for 20 min with 300 mis of
blotting wash buffer I, and probed with ^^^I-Protein G at 2 x 10^-3 x 10^
cpm/ml, the probed blots were then washed four times in 200-300 mis of
0.01 M EDTA, 1.0 M NaCl 0.25% gelatin, 0.15% Tween 20 for 15 minutes per
wash. All washing and probing steps were carried out at ambient
temperature. The probed, washed blots were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.
Molecular weight determinations on Western blots were made possible
by the transfer of the prestained molecular weight markers.
Gels to be used for protein identification were either stained with
silver according to the procedure described by Merril et al., (1981), or
with Coomassie brilliant blue R-250 as follows: (0.25% w/v in 40%
ethanol and 10% acetic acid) for 1 hour, and destained by soaking in
several changes of 10% ethanol and 10% acetic acid containing a small
quantity of DE 52 (Whatman, England) as a dye adsorbent.
Lancefield Hot Acid/Hot Alkaline Extractions
These extractions were carried out as described by Lancefield
(1928). Approximately 10 ml aliquots of 10% (wet weight/volume) 64/14 in
PBS-azide were adjusted to pH 2 or 10 using 0.5% M HC1 or 0.5 M NaOH
respectively. The suspensions were boiled for 10 minutes and the pH was
neutralized in each sample using either 0.5 M NaOH or 0.5 M HC1. PMSF
was added to a concentration of 2.0 mM. The samples were centrifuged at
approximately 10,000 x g for 10 minutes. The supernatants were passed
through a 0.22 /m filter to remove any residual cells. The cell-free


173
Yarnall, M., E.M. Ayoub, and M.D.P. Boyle. 1986. Analysis of surface
receptor expression on bacteria isolated from patients with
endocarditis. J. Gen. Microbiol. 132:2049-2052.
Yarnall, M., and M.D.P. Boyle. 1986a. Isolation and characterization of
typella and type lib Fc receptors from a group A streptococcus. J.
Immunol. 24:549-557.
Yarnall, M., and M.D.P. Boyle. 1986b. Isolation and partial
characterization of a type II Fc receptor from a group A
streptococcus. Mol. Cell. Biochem. 70:57-66.
Yarnall, M., K.J. Reis, E.M. Ayoub, and M.D.P. Boyle. 1984. An
immunoblotting technique for the detection of bound and secreted
bacterial Fc receptors. J. Microb. Meth. 3:83-93.
Zolton, R.P., and E.T. Mertz. 1972. Studies on plasminogen. X.
Isolation of plasminogen by affinity chromatography using Sepharose
butesin. Can. J. Biochem. 50:529-537.


Figure 1-1. Schematic representation of the human Glu-plasminogen
molecule. H-Glu: N-terminal glutamic acid residue; HO-Asn: C-terminal
asparagine residue; Ser: serine residue of the enzyme active site; t-PA
and UK: plasminogen activation cleavage site on Glu-plasminogen; LBS-I
and LBS-II: lysine-binding sites; 1-5: kringle domains numbers 1 through
5. E: elastase cleavage sites. Adapted from Collen (1980).


152
group C streptokinase (Kabikinase), however, none could detect the
41,000 dalton plasmin binding protein in the extract of strain 64/14
(Figure 5-5).
Taken together these findings provide evidence that the single
group A streptococcal strain 64/14 produces two proteins with affinity
for human plasmin(ogen), that can be shown to be physicochemically,
functionally, and antigenically distinct.


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
KEY TO ABBREVIATIONS x
ABSTRACT xii
CHAPTER
I. INTRODUCTION 1
Introduction 1
Plasmin(ogen) Structure 1
Plasminogen Activation 5
Plasmin Regulation 6
Summary and Specific Aims 7
II. IDENTIFICATION AND CHARACTERIZATION OF A GROUP A
STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN 9
Introduction 9
Materials and Methods 9
Results 18
Discussion 43
III. LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN
BINDING TO ITS SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR... 49
Introduction 49
Materials and Methods 50
Results 58
Discussion 79
IV. ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN 85
Introduction 85
Materials and Methods 86
Results 99
Discussion 120
iv


170
Sheela, S., and J.C. Barrett. 1982. In vitro degradation of
radiolabeled, intact basement membrane mediated by cellular
plasminogen activator. Carcinogenesis 3:363-369.
Siegal, J.L., S.F. Hurst, E.S. Liberman, S.E. Coleman, and A.S. Bleiweis.
1981. Mutanolysin-induced spheroplasts of Streptococcus mutans are
true protoplasts. Infect. Immun. 31:303-815.
Silverstein, R.L., and R.L. Nachman. 1987. Thrombospondin-plasminogen
interactions: Modulation of plasmin generation. Sem. Thromb. and
Hemo. 13:335-342.
Sottrup-Jensen, L., N. Claeys, M. Zecler, T.E. Peterson, and S.
Magnusson. 1978. Isolation of two lysine-binding fragments and one
'mini'-plasminogen (mw 38,000) by elastase catalyzed-specific
limited proteolysis. In: Progress in Chemical Fibrinolysis and
Thrombolysis, Vol. 3, pp. 191-209, Davidson, J.R., Rowan, R.M.,
Samana, M.M. and Desnoyers, P.C. (eds.), Raven Press, New York.
Sparling, P.F. 1983. Bacterial virulence and pathogenesis: An
overview. Rev. Infect. Dis. 5:637-646.
Stief, T.W., P. Lenz, and V. Becker. 1987. A simple method for
producing degradation products of fibrinogen by an insoluble
derivative of plasmin. Thromb. Res. 48:603-609.
Strickland, S., and W.H. Beers. 1976. Studies on the role of
plasminogen activator in ovulation. J. Biol. Chem. 251:5694-5702.
Strickland, S., E. Reich, and M.I. Sherman. 1976. Plasminogen activator
in early embryogenesis: Enzyme production by trophoblast and
parietal endoderm. Cell 9:231-240.
Stricklin, G.P., E.A. Baner, J.J. Jeffrey, and A.Z. Eisen. 1977. Human
skin collagenase: Isolation of precursor and active forms from both
fibroblast and organ cultures. Biochem. 16:1607-1615.
Summaria, L., and K.C. Robbins. 1976. Isolation of a human plasmin
derived functionally active light (B) chain capable of forming with
streptokinase an equimolar light (B) chain-streptokinase complex
with plasminogen activator activity. J. Biol. Chem. 251:5810-
5813.
Swenson, E., and Thorsen, S. 1981. Secondary-site binding of Glu-
plasmin, Lys-plasmin and mini-plasmin to fibrin. Biochem. J.
197:619- 628.
Switalski, L.M., P. Speziale, M. Hook, T. Waldstrom, and R. Timpl. 1984.
Binding of Streptococcus pyogenes to laminin. J. Biol. Chem.
259:3734-3738.
Teger-Nilsson, A.C., P. Friberger, and E. Gyzander. 1977. Determination
of a new rapid plasmin inhibitor in human blood by means of a
plasmin specific tripeptide substrate. Scand. J. Clin. Lab. Invest.
37:403-409.


126
0.5% gelatin, 0.15% Tween 20, 0.004% NaNj pH 7.35 (blotting wash buffer
I) for 15 minutes per wash. Blots were then probed for 3-4 hours at room
temperature while rotating in one of the following probing solutions:
For probing with either plasmin, plasmin-EACA, plasmin heavy chain, or
plasmin light chain, blots were probed with blotting wash buffer I
containing 2.0 mM PMSF and -*-^I-labeled plasmin at 2 x 10^-3 x 10^ cpm/ml
with or without 1.0 mM EACA, or blotting wash buffer I containing 2.0 mM
PMSF and l^I-plasmin heavy or light chain at 2 x 10^-3 x 10* cpm/ml.
For probing with rabbit, anti-plasmin receptor antibody or anti-group C
streptokinase antibody, blots were probed with blotting wash buffer I
containing 4.3 tg IgG per ml of probing solution (approximately a 1:3000
dilution of antisera) for three hours, washed twice for 20 min with 300
mis of blotting wash buffer I, and probed with *^1-Protein G at 2 x 10^-
3 x 10-* cpm/ml. For probing with mouse, anti-group C streptokinase
monoclonal antibodies, blots were probed with blotting wash buffer I
containing a 1:100 dilution of the stock solutions for three hours,
washed twice for 20 min with 300 mis of blotting wash buffer I, followed
by probing for three hours with blotting wash buffer I containing goat,
anti-mouse IgG, antibody (Cappel) at 1.0 //g/ml, washed twice for 20 min
with 300 mis of blotting wash buffer I, followed by probing with blotting
wash buffer I containing ^^^I-Protein G at 2 x 10^-3 x 10^ cpm/ml.
Following the last probing step all blots were washed four times in 200-
300 mis of 0.01 M EDTA, 1.0 M NaCl 0.25% gelatin, 0.15% Tween 20 for 20
minutes per wash. All washing and probing steps were carried out at
ambient temperature. The probed, washed blots were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.


129
boost). Generation of polyclonal rabbit anti-plasmin receptor 41,000
dalton protein was prepared in a similar manner using the strain 64/14
mutanolysin extracted plasmin receptor preparation, except that the
41,000 dalton protein was separated on 10% SDS-PAGE gels and used in 150
Hg to 200 ng aliquots for immunization. Pre-immune and immune IgG
fractions were prepared from rabbit sera by Protein A-Sepharose (Sigma)
affinity chromatography.
Production of Mouse Anti-plasmin Receptor Polyclonal Antibody
Plasmin receptor protein (41,000 dalton molecule) was gel purified
on 10% SDS-polyacrylamide gels under reducing conditions as described
above. Following electrophoresis each slab gel was then Western-blotted
to a sheet of nitrocellulose membrane. A small vertical strip from each
nitrocellulose blot was cut and probed with -plasmin as described
above and autoradiographed to confirm the position of the 41,000 dalton
plasmin binding protein. The remainder of each nitrocellulose sheet was
then stained with 1.0% Fast Green according to the procedure described by
Chiles et al., (1987) and the position of the 41,000 dalton band located
and aligned with the autoradiographed strip. The marked band on each
nitrocellulose sheet was then carefully cut out to avoid any
contamination, and divided into four equal fractions containing
approximately 500 pg of protein. The strips were then equilibrated in
PBS and sonicated to a fine powder. The immunization schedule was as
follows: Four groups of 6-8 week old out-bred female mice were used with
ten mice per group. Part I (initial inoculation) groups 1+3: Each mouse
in these groups was injected in two sites with (lOO/il/25/ig) of sonicated
nitrocellulose-Ag in complete Freund's Adjuvant subcutaneously. Group 2:
This group was injected in two sites with (100^il/25/ig) of sonicated


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EF0A2VPW7_V7SE67 INGEST_TIME 2015-03-25T19:05:53Z PACKAGE AA00029746_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


Figure 5-5. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and 64/14 streptokinase and group C streptokinase with mouse
anti-group C streptokinase monoclonal antibodies. Parallel protein
samples were electrophoresed on 10% SDS-polyacrylamide gels and
transferred to separate nitrocellulose membranes. Each individual blot
was probed individually in a sandwich assay first with an anti-group C
streptokinase monoclonal antibody, followed by goat, anti-mouse IgG
antibody, followed by ^^^I-Protein G. The resulting blots were auto-
radiographed at -70C for 6 to 8 hours with intensifying screens. The
identity of each monoclonal antibody is depicted below each blot. Lane 1
of each blot contains extracted plasmin receptor protein (approx. 5.0
Hg). Lane 2 of each blot contains 64/14 concentrated culture supernatant
containing approx. 2.0 ng of streptokinase. Lane 3 of each blot contains
2.0 ng of group C streptokinase (Kabikinase). See Methods for precise
experimental details.


100
acetone/Triton X-100 extractions; and (5) mutanolysin digestion. The
IOC
supernatants from these solubilizations were screened for iZ-JI-plasmin
binding by the dot-botting procedure in Methods. Aliquots of cell free
supernatants prepared from the extractions were applied to nitro
cellulose membrane in a dot-blot apparatus. Blocking and washing of the
nitrocellulose membranes was carried out according to Methods. The
nitrocellulose membranes were then probed with ^^^1-plasmin, washed and
autoradiographed.
The results of the screening of the various extracts and
preparations are shown in Figure 4-1. Extraction with mutanolysin
demonstrated the highest yield of soluble plasmin binding activity (see
Figure 4-1, row B, column 3).
The size heterogeneity of the soluble plasmin receptor activity in
the mutanolysin extract of strain 64/14 was assessed by electrophoresis
of a 50 /il aliquot of the extract on both reducing and non-reducing SDS-
polyacrylamide gels which were then stained with silver or electroblotted
onto nitrocellulose and probed with -*-^I-plasmin. The protein staining
pattern of the mutanolysin extract is shown in lane 2 of Figure 4-2.
Plasmin binding activity was concentrated predominately in a band with
an Mr of approximately 41,000 daltons (see Figure 4-2, Panel B). One /g
of purified group C streptokinase (Mr approx. 48,000 daltons) was
electrophoresed as a positive control, (Figure 4.2, lane 1, Panel A;
lane 2, Panel B). An aliquot of the control mutanolysin digestion
mixture and containing all the reactants except the bacteria was analyzed
by SDS-PAGE reveled no plasmin binding activity by Western blotting nor
any significant stainable bands (data not shown).
The possible release of the secreted plasminogen activator
(streptokinase), from the strain 64/14, during the extraction, would be a


35
established within 15 minutes at 37C. Under these conditions using 10^
bacteria and increasing concentrations of labeled plasmin, it was
possible to demonstrate a plateau in plasmin binding capacity consistent
with saturation of bacterial receptors, as shown in Figure 2-6. To
correct for non-specific binding of radiolabel, similar studies were
carried out in which binding of radiolabeled plasmin to the bacteria was
measured in the presence of 100-fold molar excess of unlabeled plasmin.
Non-specific binding demonstrated a linear relationship to counts
offered and was less than 5% in all tubes (data not shown). Analysis of
this data by least squares and Scatchard analysis demonstrated a Kp of
approximately 5 x 10"H M for the association of plasmin with its
receptor on the mouse passaged group A streptococcus strain 64/14.
Scatchard analysis of the binding data indicates that there is a single
population of plasmin receptors on streptococci (see inset, Figure 2-6),
and that strain 64/14 possesses approximately 800 receptors per
bacterium. Similar studies were carried out with the group A strain 64/P
(the original isolate that was used in the previous mouse passage studies
(Reis et al.. 1984)) and with strain 64/3 isolated following three
passages of strain 64 in mice. The 64/3 strain was less virulent in mice
than 64/P which in turn was much less virulent than the strain recovered
after 14 mouse passages, 64/14 (Reis et al., 1984). The 64/P and 64/3
strains were studied using the same protocol described for the
generation of the data using strain 64/14. The results indicated that
the plasmin receptor on 64/P had a Kp of approximately 1 x 10M and
each bacterium expressed approximately 200 receptors. The 64/3 strain
displayed a plasmin receptor with a Kp of approximately 6 x 10"^ M and
approximately 3,500 receptor sites per bacterium. Analysis of the


29
Table 2-2.
Ability of bacterial bound plasmin to solubilize a fibrin clot.
Sample
Hydrolysis of Fibrin3
Direct^
Indirect0
Bacteria Alone
Bacterial Bound Plasmin
++
Plasmin Alone
+++
+++
aUnder the experimental conditions described in the Methods section, a
(+++) reaction represented a zone of clearing with a diameter of 1.0-1.5
cm, a (++) reaction represented a zone of clearing of 0.5-1.0 cm, a (+)
reaction represented a zone of clearing from 0-0.5 cm, and (-)
represents no clearing.
^Sample placed directly onto a fibrin plate.
cSample placed onto a filter placed between the bacteria and the fibrin
plate.


50
Materials and Methods
Enzymes. Inhibitors and other Reagents
Urokinase and porcine elastase (type IV) were obtained from Sigma
Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from
Mobay Pharmaceuticals, New York, NY. Phe-Pro-Arg-chloromethylketone
(PPACK) was obtained from Calbiochem-Behring, San Diego, CA. Human Lys-
plasminogen was obtained from American Diagnostica Inc., Greenwich, CT.
H-D-Val-Leu-Lys-paranitroanilide (S-2251) was obtained from Helena
Laboratories, Beaumont, XX.
Human Plasminogen
Native human plasminogen (Glu-plasminogen) was prepared from human
plasma by chromatography on lysine-Sepharose and molecular sieving
chromatography on Superse 6 (Pharmacia-FPLC, Piscataway, NJ). The
purified protein appeared as a single band on a silver stain of an SDS-
polyacrylamide gel. Plasminogen was quantified by measuring absorbance
using a value of 17.0 (Nilsson et al. 1982). The protein was
also quantified antigenically by Laurell Rocket electrophoresis (Laurell,
1966). The purity of the isolated human plasminogen was confirmed by
activation of a known quantity of plasminogen with streptokinase and
measuring amidolytic activity. The observed and theoretical predicted
enzymatic activity were equivalent, within experimental error. Human
Lys-plasminogen, a modified form of Glu-plasminogen in which 76 of the
NH2'terminal amino acid residues are removed (Glu-1 to Lys-76) was
obtained from American Diagnostica Inc., Greenwich, CT. The homogeneity
of this Lys-plasminogen preparation was analyzed using both an urea gel
electrophoresis procedure and an acetic acid urea gel electrophoresis
procedure. This Lys-plasminogen preparation demonstrated the appropriate


CHAPTER TWO
IDENTIFICATION AND CHARACTERIZATION OF A GROUP A STREPTOCOCCAL
RECEPTOR FOR HUMAN PLASMIN
Introduction
Many group A streptococcal infections are characterized by tissue
invasion. A variety of characteristics of these microorganisms
contribute to their ability to break down natural tissue barriers and to
avoid elimination by the host immune response. Certain surface proteins
or secreted products associated with streptococci have been identified
that enable these organisms to elude the immune system (Sparling, 1983),
and proteins and toxins produced by these bacteria are known to
contribute to tissue damage (Ginsburg, 1972; Johnston, 1984; Sparling,
1983). In addition a variety of receptors for host proteins have been
described on streptococci. These include receptors for components of the
immune system such as Clq (Yarnall et al.. 1986), IgG (Kronvall, 1973),
and IgA (Russell-Jones et al., 1984), the serum protein fibrinogen
(Kronvall et al., 1979), and the stromal structural proteins laminin
(Switalski et al., 1984) and fibronectin (Myhre and Krusela, 1983). In
this chapter I describe the presence of a specific receptor for human
plasmin on certain group A streptococci.
Materials and Methods
Human Plasminogen
Human plasminogen was prepared from human plasma by chromatography
on lysine Sepharose and molecular sieving chromatography on Sephadex G-
9


90
Polyacrylamide Gel Electrophoresis and Protein Blotting
Electrophoresis was carried out as described by Laemmli (1970).
Polyacrylamide separating gels were 10% and contained 0.1% sodium
dodecylsulfate (SDS), 0.375 M Tris at pH 8.8. Stacking gels were 4% and
contained 0.1% SDS and 0.125 M Tris at pH 6.8. Electrode buffer was
0.024 M Tris, 0.192 M glycine, 0.1% SDS at pH 8.3 Samples were diluted
1:2 with sample buffer containing 0.125 M Tris pH 6.8, 4% SDS, 20%
glycerol, 10% /3-mercaptoethanol and 0.05% bromophenol blue and heated at
80-90C for 3 minutes. Gels were run at 45 volts constant voltage for
approximately 15-18 hours. Slab gels were used in the Bio-Rad Protean II
system (BioRad, Richmond, CA). Molecular weight markers were run on all
gels. Gels intended for Western blot transfer contained pre-stained
markers (Sigma) applied as a mixture which included: triosephosphate
isomerase (26,600), lactic dehydrogenase (36,500), fumarase (48,500),
pyruvate kinase (58,000), fructase-6-phosphate kinase (84,000), fi-
galactosidase (116,000), and c*2-macroglobulin (180,000). After
electrophoresis, gels intended for Western blotting were equilibrated in
25 mM Tris, 0.2 M glycine pH 8.0 containing 20% v/v methanol (electroblot
buffer) for 25 minutes. Protein blotting, from SDS-PAGE gels, was
performed using the 'Trans-Blot SD Semi-Dry' electrophoretic transfer
cell (Bio Rad, Richmond, CA). Nitrocellulose transfer medium, also
equilibrated in electroblot buffer, was sandwiched between the gel and
two sheets of Whatman 3 mm paper. The gel was also backed with two
sheets of 3 mm paper. For probing with plasmin, the Western blots were
washed, probed and autoradiographed according to the procedure described
above for dot-blotting. For probing with rabbit, anti-plasmin receptor
antibody, blots were probed with blotting wash buffer I containing 4.3 fig


93
to testing each control and experimental supernatant was centrifuged for
5 minutes at 10,000 x g to remove any particulate material. A control
sample treated in an identical fashion with the exception that no trypsin
was added was included at the 60 minute incubation time to determine the
degree of non-specific release of proteins from the bacteria. In
addition a control from which bacteria were omitted was included in each
assay.
Non-ionic Detergent/Osmotic Shock/Lvsozvme Extraction
This extraction was a modification of the procedure described by
Scopes (1982). Approximately 1.0 g wet weight of 64/14 was combined with
2.5 ml of glycerol and 0.1 ml of 10% (v/v) Triton X-100 in 20 mM KH2PO4,
1.0 mM EDTA 0.02% NaN3 pH 7.6. The cells were dispersed by vortexing and
the sample was placed at 37C for 30 minutes. The suspension was
vortexed a number of times while incubating. Following incubation, the
mixture was adjusted to 20 ml with lysozyme buffer. Lysozyme (Sigma) was
added to 200 ig/ml and DNAse I to 10 /g/ml. The sample was vortexed and
returned to 37C for 30 minutes with frequent vortexing. After
incubation, PMSF was added to 2.0 mM final concentration. The
supernatant was collected and treated as described for the Lancefield
extractions. As enzyme control to be tested with the extract in the
screening assay for plasmin receptor activity contained 200 pg/ml
lysozyme and 10 /g/ml DNAse I in 20 mM Tris-HCl pH 7.4, 0.15 M NaCl, 1.0
mM iodoacetic acid, 1.0 mM benzamidine HC1 and .02% NaN3.
Acetone/Detergent Extraction
This procedure was a modification of the extraction described by
Bhaduri et al.. (1983). approximately 1.0 g wet weight of 64/14 was
suspended in 10 ml of ice cold acetone (Fisher Certified A.C.S. grade),


151
probe. However, it appeared that either heavy or light chain would
detect the 41,000 dalton protein by Western blot assay. In agreement with
work by Summaria and Robbins (1976) the group C streptokinase
(Kabikinase), as well as the streptokinase produced by strain ATCC 12449
were preferentially detected by probing Western blots with light
chain (see Panel B of Figure 5-2). Surprisingly, not all the
streptokinases were seen to react preferentially with ^^I-light chain.
The comparison of streptokinase and plasmin receptor indicated no
antigenic relatedness. Polyclonal rabbit anti-streptokinase antibody and
both polyclonal rabbit and polyclonal mouse anti-plasmin receptor
antibody were prepared. These antibodies were tested for reactivity to
the extracted plasmin receptor and several streptokinases. The rabbit
anti-plasmin receptor antibodies showed no cross-reactivity with
streptokinase from the strain 64/14, or with two other group A
streptokinases, or any of the three group C streptokinase sources studied
(Figure 5-3). Two polyclonal mouse anti-plasmin receptor antibody
preparations also showed no cross-reactivity with 64/14 streptokinase
or with the prototype group C streptokinase (Kabikinase)(Figure 5-4).
The polyclonal rabbit anti-group C streptokinase antibody detected all
forms of streptokinases examined, while showing no cross-reactivity
with the 41,000 dalton plasmin binding protein in the extract of strain
64/14.
Furthermore, 16 mouse monoclonal antibodies were tested for
reactivity towards the strain 64/14 streptokinase, the extracted plasmin
receptor, and group C streptokinase (Kabikinase). Thirteen of the
sixteen monoclonal antibodies recognized the 48,000 dalton 64/14
streptokinase. As expected, all the monoclonal antibodies reacted with


57
follows. The bacterial free supernatants or the corresponding control
samples were added to plastic cuvettes containing 10^ IU streptokinase in
a total volume of 900 il of enzyme assay buffer (0.05 M Tris, 0.05 M
NaCl, 0.1% PEG-8000, pH 7.4). For the Glu-and Lys-plasminogen
preparations the reaction mixture was incubated at 37C for 10 minutes to
allow plasminogen-streptokinase complexes to form. For Lys-plasmin, a
similar incubation with streptokinase was performed to allow for
equivalent substrate turnover to that of zymogen-streptokinase
complexes. Following incubation, H-D-Val-Leu-Lys-pNA (S-2251) was added
to yield a final concentration of 300 /M. Tubes were allowed to incubate
for precisely five minutes and then quenched with 100 /il of glacial
acetic acid. The amount of substrate hydrolysis, which is directly
proportional to the amount of plasmin enzyme present was then quantified
by measuring the absorbance of the reaction mixture at 405 nm.
The enzymatic activity of the bacterial free supernatant was
determined by comparison with the enzymatic activity of known standards.
The percent of residual enzymatic plasmin(ogen) activity in the bacterial
free supernatant was calculated by determining the fraction of total
enzymatic activity in a control sample remaining in the supernatant
following incubation with bacteria. Control tubes containing bacteria
and substrate, and substrate in buffer were included. All assays were
performed in duplicate.
Measurement of Functional Activity of Plasmin Associated with Bacteria
The plasmin activity associated with bacterial pellets was examined
using the chromogenic substrate as described above. Following binding
and centrifugation the pellets were washed 3 times with 1.0 ml of VBS-gel
and resuspended in 400 /I of the enzyme assay buffer. S-2251 was added


ISOLATION AND CHARACTERIZATION OF A
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
By
CHRISTOPHER C. BRODER
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1989

This dissertation is dedicated
to the memory of my father
Thomas J. Broder

ACKNOWLEDGMENTS
I wish to express my sincere thanks to Dr. Michael D.P. Boyle for
giving me the opportunity to work in his laboratory, for his support and
guidance, and especially for his patience. It has been a pleasure to
work with Mike.
I also wish to give a special thanks to Dr. Richard Lottenberg for
all of his help, guidance, and friendship.
I would like to thank the other members of my committee, Drs. R.W.
Moyer and J.W. Shands, for their helpful suggestions throughout this
study.
I would also like to offer a special thanks to Dr. Kenneth H.
Johnston, my outside examiner, for sending me the monoclonal antibodies,
and the solid-phase plasminogen activator assay mentioned in this study
and for taking the time to review and discuss my work.
I would also like to express my appreciation to all the people with
whom I have worked for the past four years, especially Jeannine Brady,
Greg VonMerring, Tim Broeseker, and Lucy DesJardin.
I also offer most special thanks to my parents, Jeanne C. and
Thomas J. Broder, for their never-ending love and support throughout all
my endeavors. I also thank all my family, especially my brother
Michael.
Finally, I offer my sincere thanks to my wife, Colleen, for all of
her unselfish support, which has been essential for me in pursuing my
goals.
iii

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
KEY TO ABBREVIATIONS x
ABSTRACT xii
CHAPTER
I. INTRODUCTION 1
Introduction 1
Plasmin(ogen) Structure 1
Plasminogen Activation 5
Plasmin Regulation 6
Summary and Specific Aims 7
II. IDENTIFICATION AND CHARACTERIZATION OF A GROUP A
STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN 9
Introduction 9
Materials and Methods 9
Results 18
Discussion 43
III. LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN
BINDING TO ITS SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR... 49
Introduction 49
Materials and Methods 50
Results 58
Discussion 79
IV. ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN 85
Introduction 85
Materials and Methods 86
Results 99
Discussion 120
iv

Page
V. COMPARISON OF THE GROUP A STREPTOCOCCAL RECEPTOR FOR
HUMAN PLASMIN WITH STREPTOKINASE 122
Introduction 122
Materials and Methods 123
Results 130
Discussion 146
VI. SUMMARY AND CONCLUSIONS 153
REFERENCES 162
BIOGRAPHICAL SKETCH 174
v

LIST OF TABLES
Table Page
2-1. Binding of radiolabeled proteins to various nephri-
togenic and non-nephritogenic group A
streptococci 19
2-2. Ability of bacterial bound plasmin to solubilize a
fibrin clot 29
3-1. Summary of inhibition experiments 73
3-2. Measurement of plasmin(ogen) associated with
bacterial pellets 74
5-1. Fluid-phase plasminogen activator activity assay.... 132

LIST OF FIGURES
Figure Page
1-1. Schematic representation of the human Glu-plasminogen
molecule 4
2-1. Binding of plasmin to bacteria: comparison of the
kinetics of generation of plasmin and its ability
to bind to the group A streptococcus 64/14 22
2-2. Effect of inhibiting the active site of plasmin on
its ability to bind to the group A streptococcal
strain 64/14 24
2-3. Regulation of bacterial bound enzyme activity by a
variety of different serine protease inhibitors 27
2-4. Binding of ^^I-plasmin or 125j-piasmin0gen to the
group A streptococcal strain 64/14 as a function of
pH 31
2-5. Binding of ^^I-plasmin or 125j-piasminogen to the
group A streptococcal strain 64/14 as a function of
ionic strength 34
2-6. Specific binding of ^^I-plasmin to 10^ group A
streptococci, strain 64/14, following a 15 minute
incubation at 37C in VBS-gel at pH 7.4 37
2-7. Inhibition of binding of -plasmin to the group A
streptococcal strain 64/14 in VBS-gel containing
various concentrations of epsilon-aminocaproic acid,
lysine, and arginine 40
1 9 S
2-8. Elution of -plasmin from group A streptococcal
strain 64/14 in VBS-gel containing various con
centrations of epsilon-aminocaproic acid, lysine,
and arginine 42
3-1. SDS-UREA-PAGE analysis of isolated plasmin(ogen)
fragments 60
3-2. Inhibition of, PPACK reacted, ^^I-Lys-plasmin
binding to group A streptococcal plasmin receptor... 63
3-3. Binding of ^-^1 labeled Glu- and Lys-plasmin(ogens) . 67
vii

Page
Figure
3-4. Inhibition of, PPACK reacted, '^^I-Lys-plasmin
binding to group A plasmin receptor 69
3-5. Binding of Lys-plasmin(s), derived from Glu-
plasminogen and Lys-plasminogen, to the group A
streptococcal strain 64/14 as measured by residual
activity in the bacterial free supernatant 72
3-6. Binding of Glu- and Lys-plasminogen to the group A
streptococcal strain 64/14 as measured by residual
activatable zymogen in the bacterial free
supernatant 76
3-7. Characterization of -plasmin(ogen) species eluted
from bacteria 78
4-1. Dot-blot analysis of solubilized plasmin binding
activities 102
4-2. SDS-PAGE and Western blot analysis of mutanolysin
extracted 64/14 bacterial plasmin binding activity.. 104
4-3. Solid-phase plasminogen activation assay 109
4-4. Representative profile of an affinity purification
of strain 64/14 mutanolysin extracted plasmin
binding activity 112
4-5. Analysis of affinity purified plasmin binding material
from the strain 64/14 mutanolysin extract 114
4-6. SDS-PAGE and Western blot analysis of mutanolysin
extracted, affinity purified plasmin binding
activity 117
4-7. SDS-PAGE and Western blot analysis of plasmin
receptor protein with a polyclonal rabbit antibody.. 119
5-1. Functional identification and distinction of
streptokinase proteins and plasmin binding receptor
protein 134
5-2. Comparison of binding reactivities of streptokinase
proteins and plasmin binding receptor protein with
^*1 -plasmin heavy chain and ^-^^1-plasmin light
chain 138
5-3. Analysis of the antigenic relationship of the 64/14
plasmin receptor and streptokinase proteins 141
viii

Figure Page
5-4. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and
group C streptokinase with mouse polyclonal
anti-plasmin receptor antibodies 145
5-5. Analysis of the antigenic relationship of the 64/14
plasmin receptor and 64/14 streptokinase and
group C streptokinase with mouse anti-group C
streptokinase monoclonal antibodies 148
IX

KEY TO ABBREVIATIONS
c*2 -AP/a2 "PI
alpha-2-antiplasmin
ATCC
American typed culture collection
BSA
bovine serum albumin
cpm
counts per minute
DNase
deoxyribonuclease
EACA
epsilon aminocaproic acid
EDTA
ethylenediaminetetraacetic acid
ELISA
enzyme -linked-immunosorbant-assay
FPLC
fast performance liquid chromatography
g
gravity
Glu-plasminogen
native human plasminogen with NH2-terminal
glutamic acid
HC
heavy chain of plasmin
HRGP
histidine-rich glycoprotein
IgG
immunoglobulin class G
KD
kilodalton
kd
dissociation constant
KIU
kallikrein inhibitor unit
LBS
lysine-binding site
LC
light chain of plasmin
Lys-plasminogen
proteolytically modified form of Glu-plasminogen with
NH2-terminal lysine
M
molar
x

Mini-PLA
low molecular weight plasmin
Mini-PLG
low molecular weight plasminogen
mM
millimolar
fjM
micromolar
/im
micrometer
MOPS
(3-[N-Morpholino]propanesulfonic acid)
Mr
relative molecular weight
NIH u
National Institute of Health unit
nm
nanometer
PAGE
polyacrylamide gel electrophoresis
PBS
phosphate buffered saline
PEG
polyethylene glycol
PLA
plasmin
PLG
plasminogen
P
pico
PMSF
phenylmethylsulfonylfloride
pNpGB
p-nitrophenyl p-guanidinobenzoate HC1
PPACK
Phe- Pro-Arg-chloromethyIketone
RNase
ribonuclease
S-2251
H-D-Val-Leu-Lys-paranitroanilide
SDS
sodium dodecylsulfate
SK
streptokinase
tPA
tissue-type plasminogen activator
Tris
(Tris[hydroxymethyl]aminomethane
UK
urokinase
VBS-gel
Veronal-buffered saline plus gelatine
VPLCK
D-Val-Phe-Lys-chloromethyl ketone
xi

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ISOLATION AND CHARACTERIZATION OF A
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
By
CHRISTOPHER C. BRODER
May 1989
Chairman: Michael D.P. Boyle
Major Department: Immunology and Medical Microbiology
The expression of a specific receptor for the key human fibrinolytic
enzyme plasmin on the surface of the group A streptococcal strain 64/14
is reported. The receptor was specific for plasmin, and demonstrated no
significant reactivity with the zymogen form of the molecule (Glu-
plasminogen). Bacterial bound plasmin retained its enzymatic activity,
and could not be inhibited by the physiological regulator (<*2-
antiplasmin). The receptor demonstrated a high affinity for plasmin
(Kp=l.0x10'I m), and binding was maximal at physiologic pH and ionic
strength. Furthermore, the receptor-ligand complex was reversibly
inhibitable by e-aminocaproic acid and L-lysine. The binding of plasmin
to this group A streptococcus was found to occur primarily through
interactions with the heavy chain of the plasmin molecule, and was
dependent on a specific conformation of the ligand. A functionally
active plasmin receptor was obtained from strain 64/14 bacteria by an
enzymatic extraction with mutanolysin. Plasmin binding activity was
expressed predominantly in a protein having an Mr of approximately 41,000
xii

daltons. The plasmin receptor demonstrated no plasminogen activator
activity. A functionally active plasmin receptor protein was purified by
affinity chromatography using immobilized plasmin and specific elution
with L-lysine or EACA. The strain 64/14 plasmin receptor was compared
with secreted streptokinase proteins from five streptococcal isolates
including strain 64/14. Only the plasmin receptor-plasmin complex was
found to be sensitive to L-lysine or EACA. Polyclonal rabbit and mouse
anti-plasmin receptor antibodies were prepared, as well as polyclonal
anti-group C streptokinase antibodies. Using these antibodies as well as
a bank of mouse monoclonal anti-group C streptokinase antibodies, the
41,000 dalton plasmin receptor protein from strain 64/14 was shown to be
antigenically unrelated to either group A or C streptokinase. Thus the
strain 64/14 streptococcal receptor for human plasmin is
physicochemically, functionally, and antigenically distinct from
streptokinase. The importance of a specific receptor for human plasmin
on pathogenic streptococci is unclear; however, it may provide a
mechanism for the capture of the potent enzyme plasmin which may confer
additional invasive properties to the bacteria.
xiii

CHAPTER ONE
INTRODUCTION
Introduction
Plasmin, a serine protease, is the key component of the mammalian
fibrinolytic enzyme system. The main physiological role of the
fibrinolytic system is the dissolution of fibrin clots formed in blood
vessels. Milstone, in 1941, determined that the lysis of fibrin, by the
streptococcal substance described by Tillett and Garner in 1933, was
dependent on a 'lytic factor' in human serum. This was later followed by
the discovery that the lytic factor was an enzyme precursor, in human
plasma, that was converted to an active enzyme by a component in the
streptococcal fluid (Christensen, 1945; Kaplan, 1944). This precursor
was called plasminogen, the enzyme plasmin, and the streptococcal factor
streptokinase (Christensen and Macleod, 1945). The zymogen precursor
plasminogen, molecular weight approximately 92,000 daltons, is a single
chain glycosylated protein containing 790 amino acids in known sequence
and containing 24 disulfide bridges (Brogden et al., 1973; Sottrup-Jensen
et al. 1978; Wiman, 1973, 1977).
Plasmin(ogen) Structure
Native plasminogen (Glu-plasminogen) has glutamic acid as the NH2-
terminal residue, but is readily converted by the action of plasmin to
modified forms of plasminogen which are commonly called Lys-plasminogen.
1

2
These modified forms of plasminogen have lysine, valine, or methionine as
their N-terminal amino acid (Walln and Wiman, 1970, 1972). These
modifications occur by the hydrolysis of the ArggyMetgg, Lysyg-Lysyy, or
Lys77-Val7g peptide bonds. The generation of plasmin from plasminogen
occurs through the specific cleavage of a single Arg-Val bond
corresponding to the Arg550'Val5g^ positions (Robbins et al., 1967).
This cleavage generates a two chain active plasmin molecule consisting of
a heavy chain and light chain held together by disulfide linkages
(Groskopf et al., 1969; Wiman, 1973) (see Figure 1-1). The light chain
of plasmin has a molecular weight of approximately 25,000 daltons
(Robbins and Summaria, 1970; Wiman, 1977) and contains the serine active
site. The heavy chain of plasmin has a molecular weight of approximately
63,000 daltons (Robbins and Summaria, 1970), and amino acid sequencing
revealed a structure containing 5 homologous triple loop structures known
as kringles (Sottrup-Jensen et al., 1978) .
Several specific compounds such as lysine, e-aminocaproic acid
(EACA), trans-4-aminomethycyclohexanecarboxylic acid (transexamic acid),
and C-terminal lysine peptides bind to certain sites on the plasmin(ogen)
molecule. These specific sites are the characteristic 'lysine-binding
sites' distinct from the catalytic site (Thorsen, 1975). These compounds
affect the properties of plasminogen and plasmin, and play an important
role in determining this zymogen-enzyme system's physiological
specificity. Chapter Three will go into more detail on the nature of
plasmin(ogen)'s lysine binding sites. Affinity chromatography of
defined fragments of plasminogen on lysine Sepharose has demonstrated
that these 'lysine-binding sites' are located in the portion of the
plasmin molecule which becomes the heavy chain upon activation. The

Figure 1-1. Schematic representation of the human Glu-plasminogen
molecule. H-Glu: N-terminal glutamic acid residue; HO-Asn: C-terminal
asparagine residue; Ser: serine residue of the enzyme active site; t-PA
and UK: plasminogen activation cleavage site on Glu-plasminogen; LBS-I
and LBS-II: lysine-binding sites; 1-5: kringle domains numbers 1 through
5. E: elastase cleavage sites. Adapted from Collen (1980).

4
LBS II

5
native Glu-plasminogen molecule contains one high affinity lysine
binding site (K = 9 fiM.) and five weaker lysine-binding sites (K = 5 mM)
(Markus et al.. 1978a, 1978b). Lys-plasminogen contains one high
affinity, one intermediate affinity, and four lower affinity lysine
binding sites. The exact number of sites on the plasmin molecule has not
been reported. Two of these lysine-binding sites have been mapped to
specific regions in the plasminogen molecule. Studies involving
equilibrium dialysis experiments on the binding of EACA to isolated
fragments of the plasmin(ogen) molecule (see Chapter Three for a
complete discussion of these fragments) revealed that the high affinity
lysine-binding site was located in the kringle 1 structure, and kringle 4
contained one of the lower affinity sites (Lerch et al., 1980).
Plasminogen Activation
The generation of plasmin from plasminogen is accomplished by
plasminogen activators. Three plasminogen activators have been
extensively studied. Urokinase (UK) and tissue plasminogen activator
(tPA) are proteolytic enzymes (for review see Astrup, 1978); the third,
streptokinase (SK), possesses no inherent proteolytic activity. Tissue
plasminogen activator, by virtue of its serine protease activity can
directly activate plasminogen. This enzyme is present in various tissues
and can also bind to fibrin. Urokinase, also a serine protease, is a
glycoprotein which has no fibrin binding capacity. Urokinase can also
activate plasminogen directly cleaving the Arg56o_Val561 peptide bond.
Streptokinase is a unique plasminogen activator which is produced by
certain streptococci. The only apparent function of streptokinase, since
its initial description by Tillett and Garner (1933), is its ability to

6
activate plasminogen. Unlike the other plasminogen activators,
streptokinase has no enzymatic activity. The activation mechanism lies
in its ability to form a specific 1:1 stoichiometric complex with
plasminogen, as well as with plasmin, which leads to the generation of an
active enzyme moiety, presumably through conformational changes in the
plasminogen molecule without the cleavage event at the Arg5g0^al56^
peptide bond (Markus et al., 1976), that can in turn act as a plasminogen
activator for plasminogen molecules. This is a function neither of the
two proteins possesses alone.
Plasmin Regulation
Once generated, plasmin's activity is also carefully regulated under
normal physiological conditions. This regulation is accomplished by a
specific inhibitor of plasmin known as a2"anti-Plasn,i-n (a^-AP) (Aoki et
al.. 1977; Collen, 1976; Collen et al., 1975). Alpha2*antiplasmin is a
single chain glycoprotein with a molecular weight of approximately 70,000
daltons. 2-antiplasmin forms a 1:1 stoichiometric complex very rapidly
(estimated rate constant of k^=3xl0^M"'^) (Wiman and Collen, 1978) and
neutralizes plasmin's activity through a covalent association with the
serine residue in the active site of plasmin. A physiological role of
2*AP as an inhibitor of other proteases other than plasmin appears
negligible (Edy and Collen, 1979; Ohlsson and Collen, 1977).
Workers pioneering the techniques of tissue culturing noted that
explants of cancer tissue consistently caused proteolytic degradation,
liquefying the plasma clots on which they were grown (Carrel and Burrows,
1911; Lambert and Hanes, 1911). Since those early studies the
plasminogen-plasmin system, in addition to its role in fibrinolysis, has

7
been implicated in a variety of normal and abnormal processes which
involve the destruction or alteration of the extracellular environment,
such as tumor cell growth and invasiveness (for review of this extensive
literature see Dano et al., 1985), tissue remodeling, embryogenesis
(Beers et al., 1975), ovulation (Strickland and Beers, 1976), and
trophoblast implantation (Strickland et al., 1976). In fact, plasmin
exhibits broad substrate specificity and in addition to fibrin can
hydrolyze components of connective tissue and basement membranes such as
laminin, proteoglycans, fibronectin, thrombospondin, and type-V collagen,
as well as proteolytically activating other proteases (for review see
Knudsen et al.. 1986) and several plasma proteins (Marder et al., 1982).
Summary and Specific Aims
In the process of examining human serum from patients for antibody
reactivity directed against the streptococcal plasminogen activator
streptokinase, from patients who received thrombolytic therapy by
streptokinase administration, an interesting observation was made by Dr.
M.D.P. Boyle and Dr. R. Lottenberg. In experiments which involved immune
precipitations using heat killed, Fc-receptor expressing, group C
streptococci it was found that control tests involving incubations of the
"IOC
radiolabeled tracer (-plasmin) and bacteria revealed an association
of plasmin to bacteria in the absence of any added antibody. Testing
other groups of streptococci showed that the group A streptococci
displayed the highest level of plasmin binding activity, while
demonstrating little binding activity for preparations of 12f>I-
plasminogen.

8
The ability of certain group A streptococci to produce a plasminogen
activator (e.g. streptokinase) and also to express a receptor for the
activation product plasmin, may contribute to the invasive properties of
these bacteria. This study has been designed to characterize this
plasmin receptor phenomenon more completely in order to increase
understanding of any potential role in bacterial pathogenesis.
The specific aims of the study are to
1. Identify and characterize a group A streptococcal
receptor for human plasmin (Chapter Two).
2. Map the binding site on human plasmin recognized by the
bacterial plasmin receptor (Chapter Three).
3. Isolate and purify a functionally active group A
streptococcal plasmin receptor (Chapter Four).
4. Compare the group A streptococcal receptor for human
plasmin with streptokinase, (Chapter Five), with
respect to
a. Plasminogen activator activity.
b. Plasmin(ogen) binding domain specificity.
c. Antigenic relatedness.

CHAPTER TWO
IDENTIFICATION AND CHARACTERIZATION OF A GROUP A STREPTOCOCCAL
RECEPTOR FOR HUMAN PLASMIN
Introduction
Many group A streptococcal infections are characterized by tissue
invasion. A variety of characteristics of these microorganisms
contribute to their ability to break down natural tissue barriers and to
avoid elimination by the host immune response. Certain surface proteins
or secreted products associated with streptococci have been identified
that enable these organisms to elude the immune system (Sparling, 1983),
and proteins and toxins produced by these bacteria are known to
contribute to tissue damage (Ginsburg, 1972; Johnston, 1984; Sparling,
1983). In addition a variety of receptors for host proteins have been
described on streptococci. These include receptors for components of the
immune system such as Clq (Yarnall et al.. 1986), IgG (Kronvall, 1973),
and IgA (Russell-Jones et al., 1984), the serum protein fibrinogen
(Kronvall et al., 1979), and the stromal structural proteins laminin
(Switalski et al., 1984) and fibronectin (Myhre and Krusela, 1983). In
this chapter I describe the presence of a specific receptor for human
plasmin on certain group A streptococci.
Materials and Methods
Human Plasminogen
Human plasminogen was prepared from human plasma by chromatography
on lysine Sepharose and molecular sieving chromatography on Sephadex G-
9

10
100 (Lottenberg et al.. 1985). Plasminogen was quantified by measuring
absorbance using a value of 17.0 (Nilsson e£ al. 1982).
Enzymes, Inhibitors and other Reagents
The enzymes urokinase and trypsin were obtained from the Sigma
Chemical Company, St. Louis, Mo.; Aprotinin was obtained as Trasylol
from Mobay Pharmaceuticals, New York, New York. Phe-pro-arg
chloromethylketone (PPACK) was obtained from Cal-Biochem (San Diego, Ca.)
P-nitrophenyl, p-guanidobenzoate HC1 (pNpGB) was obtained from Sigma
Chemical Co, St. Louis, Mo.; human a2"anti.p]-asini.n (c*2-AP) was obtained
from American Diagnostica Inc., Greenwich, Connecticut. H-D-Val-leu-lys-
paranitroanilide (S-2251) was obtained from Helena Chemical Co.,
Beaumont, Texas.
Radioiodination of Proteins
Human plasminogen, urokinase, and trypsin were iodinated by a mild
lactoperoxidase method using Enzymobeads (Bio-rad Laboratories Richmond,
Calif.) as described by Reis et al., (1983). The labeled proteins were
separated from free iodine by passage over a G25 column (PD-10 Pharmacia)
and collected in 0.15 M Veronal buffered saline pH 7.35 containing 0.001
M Mg++, 0.00015 M Ca++ and 0.1% gelatin (VBS-gel). The labeled proteins
were stored in aliquots containing 0.02% sodium azide at -20C. Labeled
aliquots were used once and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three /xl of urokinase (Sigma 20
u/ml) was added to a 400 /I solution of 1.0 /xM plasminogen containing
0.04 M lysine. The mixture was incubated at 37C for 45 minutes unless
stated otherwise. The efficiency of plasmin generation was followed by

11
measuring the conversion of the single chain plasminogen molecule
(Mr=90,000) into heavy chains (Mr=60,000) and light chains
(Mr=25,000) as determined by the migration of radiolabeled proteins,
following denaturation and reduction, on 10% SDS-polyacrylamide gels.
The migration of labeled proteins was determined by autoradiographic
exposure of dried gels to Kodak XAR 5 film with intensifying screens at -
70C for 20 hours.
Bacteria
The group A /3-hemolytic streptococcal strain 64 had been previously
subjected to mouse passage as described by Reis et al., (1984). The
parent strain (64/P), as well as strains isolated after three (64/3) and
fourteen (64/14) mouse passages, were grown in either Todd-Hewitt broth
(DIFCO, Detroit, Mich.) or chemically defined media (Van De Rijn and
Kessler, 1980) overnight at 37C as stationary cultures (Yarnall and
Boyle, 1986b). The bacteria were harvested by centrifugation and
resuspended in phosphate-buffered saline (PBS), pH 7.4, containing 0.05%
Tween-20 and 0.02% sodium azide. The bacteria were heat killed at 80 C
for 10 minutes, a treatment that did not alter their plasmin binding
potential, but eliminated the production of soluble plasminogen
activators which would interfere with these studies. The suspension was
centrifuged, the pellet washed twice with PBS and then resuspended at 10%
wet weight/volume in PBS containing 0.05% Tween-20 and 0.02% sodium
azide. Samples were stored at -20C. The concentration of a bacterial
suspension was determined by counting bacterial chains in a Neubauer
hemacytometer (Fisher Scientific, Orlando, FL).
Determination of Binding of Radiolabeled Proteins to Bacteria
The light scatter at 550 nm was determined to standardize the
concentration of organisms used in subsequent tests. A light scatter

12
reading of 0.3 corresponded to approximately 2 x 10^ organisms/ml
(Yarnall and Boyle, 1986b). A standardized number of bacteria
(approximately 10^ organisms) were incubated with labeled proteins
(approximately 30,000 cpm/tube) in a total volume of 400 /xl of VBS-gel
for 1 hour at 37C. The bacteria were pelleted by centrifugation at 1000
x g for 10 minutes and washed twice with 2.0 ml VBS-gel. The
radioactivity associated with the bacteria was determined in a Beckman
5500 autogamma counter. All estimates were carried out in duplicate.
Fibrin Plate Assay
Fibrin plates were prepared using 5 cm diameter disposable petri
dishes. Ten ml of 0.1% fibrinogen in PBS were clotted with 0.2 ml of
bovine thrombin (10 NIH u/ml) in 0.5 M CaCl2. Twenty pmoles of plasmin
were bound to 100 1 of a 10% w/v solution of the group A streptococcus,
64/14, in a total volume of 400 tl of VBS-gel. The mixture was incubated
at 37C for 45 minutes. A parallel series of samples containing bacteria
with no added plasmin served as the negative control. Fifty ti of a
suspension of bacteria or bacteria plus plasmin were placed either
directly onto a fibrin plate or onto a 0.22 /tm Millipore filter placed
between the bacteria and the fibrin plate. The plates were incubated for
20 hours at 37C and the degree of hydrolysis was scored by measuring the
area of the zone of clearing from the underside of the plate. In each
experiment a control of free plasmin was included and each estimate was
carried out in duplicate.
Plasmin-Inhibitor Complex Generation
Plasmin was generated from plasminogen as described above. Three
130 1 aliquots of the labeled enzyme were placed into separate
microtubes (approximately 2 x 10^ cpm/tube), and incubated with a 200-

13
fold molar excess of either PPACK, aprotinin, or pNpGB for 10 minutes at
room temperature. The volume of each sample was increased by the
addition of 200 /I of VBS-gel, and each was applied to a separate G-25
column (PD 10 Pharmacia Fine Chemical) to remove free inhibitor. Five
hundred /I fractions were collected and counted in an autogamma counter
to localize the modified ^^1 plasmin. Fractions containing the labeled
protein-inhibitor complexes were pooled. Aliquots of the three labeled
complexes were mixed with a 10-fold molar excess of c*2-AP in a final
reaction volume of 400 /il for 10 minutes at room temperature. Plasmin
and plasminogen were included as controls. The volumes of each solution
were adjusted to 1.1 ml with VBS-gel and then used in a direct binding
assay to group A streptococci. Each of the plasmin-inhibitor samples
that has been incubated with excess a^'AP was monitored on non-reducing
SDS-polyacrylamide gels as described by Weber and Osborn (1969) for the
formation of a high molecular weight complex.
Determination of Plasmin Activity While Bound to Bacteria
To five microtubes, each containing 100 /il of a 10% w/v solution of
the group A streptococcus, 64/14, in a total volume of 400 /I VBS-gel, 10
nM plasmin was added and allowed to bind for 40 minutes at 37C. Five
other tubes containing plasmin but no bacteria and one tube containing
bacteria alone were prepared as controls. The bacteria were pelleted and
washed twice with 1.0 ml VBS-gel and resuspended in 400 /il VBS-gel
containing a 10-fold molar excess of either pNpGB, PPACK, aprotinin, a^-
AP or buffer alone. All samples were then incubated for 15 minutes at
room temperature. The samples containing bacteria were pelleted by
centrifugation, washed with 1.0 ml of VBS-gel and resuspended with
vigorous vortexing in 400 /il VBS-gel.

14
To each tube, 20 1 of an 8.0 mM solution of the chromogenic
substrate, H-D-val-leu-lys-paranitroanilide, was added,to yield a final
concentration of substrate in the reaction mixture of 400 M. The tubes
were mixed by vortexing and incubated at 37C for 25 minutes. At that
time the enzyme reaction was quenched by the addition of 400 ti of 10%
(v/v) acetic acid, the samples were then centrifuged for 5 minutes at
10,000 x g and the optical densities of the solutions at 405 nm were
determined. The release of paranitroaniline from the synthetic substrate
monitored at this wavelength was directly proportional to the enzymatic
activity of plasmin. Control samples of substrate alone and substrate
plus bacteria were included and each estimate was carried out in
duplicate.
Effect of pH on Plasmin Binding to Bacteria
To assess the effect of pH on the bacteriumrplasmin(ogen)
interaction, 50 ti of labeled plasminogen or plasmin (approximately 2 x
10^ cpm) were added to 1.0 ml of VBS containing 0.05% Tween-20 adjusted
to the appropriate pH. After 15 minutes at room temperature, 50 tl of
VBS containing approximately 10^ bacteria (strain 64/14) were added and
the mixture was incubated at 37 C for 15 minutes. The bacterial
suspensions were centrifuged at 1,000 x g for 7 minutes to separate
bacteria from unbound labeled proteins and the pellets were washed twice
with 2.0 ml of VBS at the appropriate pH. The radioactivity associated
with the bacterial pellet in duplicate experiments was measured using a
Beckman 5500 autogamma counter.
To assess the effect of ionic strength on the bacterium-
plasmin(ogen) interaction, similar studies were carried out in solutions
containing different concentrations of NaCl with 0.05% Tween-20. The

15
bacterial pellets were washed in the appropriate NaCl concentration to
remove unbound labeled proteins.
Effect of Ca++ and Mg-1"1- on Plasmin Binding
Binding of radiolabeled plasmin to group A streptococci strain 64/14
was studied in the following buffers: 1) VBS-gel containing 0.00015 M
Ca++ and 0.001 M Mg++, or 2) metal free VBS-gel containing 0.15 M EDTA.
In each case 400 pi of buffer were added to 100 pi of VBS-gel containing
approximately 107 bacteria and 100 pi of VBS-gel containing 3 x 10~* cpm
of radiolabeled plasmin. After incubation at 37 C for 15 minutes, the
mixtures were centrifuged at 1,000 x g for 7 minutes to separate bacteria
from unbound radiolabel, the pellets were washed twice with 2.0 ml of the
appropriate buffer, and radioactivity associated with the bacterial
pellet in duplicate tubes was measured.
Inhibition of Binding of Plasmin by Amino Acids
Labeled plasmin (100 pi containing approximately 2 x 10^ cpm) was
added to 200 pi VBS-gel containing varying concentrations of
epsilon-aminocaproic acid (EACA), lysine, or arginine, and incubated at
37 C for 15 minutes. The pH of each solution was 7.0. One hundred pi
of VBS-gel containing 107 bacteria (strain 64/14) were then added and the
mixture was incubated at 37C for 15 minutes. The bacterial suspensions
were centrifuged at 1,000 x g for 7 minutes and washed twice with 2.0 ml
of VBS-gel containing the same concentration of amino acid present during
the incubation period. The percent inhibition of binding was calculated
for duplicate samples by comparison with binding in VBS-gel alone.
The ability of EACA, lysine, or arginine to dissociate bound plasmin
from the bacteria was examined in the following manner. Labeled plasmin
was incubated with 107 bacteria in VBS-gel at 37 C for 15 minutes. The

16
bacteria were pelleted by centrifugation and washed twice with 2.0 ml of
VBS-gel. After determining the radioactivity associated with the
bacteria, the pellets were resuspended in solutions of VBS-gel containing
varying concentrations of amino acid or amino acid analogs (pH 7.0) as
described above. The mixtures were incubated at 37 C for 15 minutes and
washed twice with VBS-gel containing the appropriate amino acid
concentration. The radioactivity associated with the bacteria in
duplicate samples was again measured and the percentage dissociated was
calculated.
Determination of Kp and Receptor Density
Labeled plasmin (25,000 to 250,000 cpm) in 100 /I of VBS-gel was
added to 3 x 10^ bacteria in 300 pi of VBS-gel, pH 7.4, and incubated at
37C for 15 minutes. The bacterial suspensions were centrifuged at 1,000
x g for 10 minutes and washed twice with 2.0 ml of VBS-gel. All
determinations were performed in triplicate. Total binding was
determined by measuring the radioactivity associated with the bacterial
pellet when only labeled plasmin was offered. Non-specific binding was
determined by pre-incubating bacteria at 37 C for 15 minutes in VBS-gel,
pH 7.4, containing unlabeled plasmin at a 100-fold molar excess of the
labeled plasmin. Specific binding was calculated by subtracting
non-specific binding from total binding for each amount of labeled
plasmin offered. The amount of free labeled plasmin was calculated by
subtracting the amount of specifically bound labeled plasmin from the
total amount of labeled plasmin offered.
The apparent dissociation constant (Kp) was determined by two
methods. A non-linear least squares analysis of the total counts offered
vs. the counts bound fit to the simple Michaelis-Menten equation was

17
performed as described by Lottenberg et al.. (1985). The concentration
of plasmin was determined by converting counts per minute to moles using
the known specific activity for the labeled plasminogen. Scatchard
analysis (Scatchard, 1949) of these data was also performed as described
by Lottenberg et al.. (1987). Counts bound vs. counts bound/counts free
was plotted and the slope (representing -1/K¡)) was determined by linear
regression. The X-intercept (counts bound) was converted to moles of
plasmin. To determine the receptor densities the number of moles of
plasmin bound was determined by extrapolating the Scatchard plot and
determining the intercept. This represented the maximal binding of
plasmin to a known number of bacteria (derived by hemacytometer chamber
counts).
Plasmin which had been bound to and eluted from strain 64/14 by
treatment with lysine was also examined in similar binding studies.
Eluted plasmin was obtained by incubating 2.0 ml of stock 10% wet
weight/volume bacterial suspension (strain 64/14) with approximately 20
Hg of labeled plasmin at room temperature for 45 minutes. This
suspension was centrifuged at 1,000 x g for 10 minutes and washed once
with 10 ml of VBS-gel, and the radioactivity associated with the
bacterial pellet was measured. The pellet was then resuspended in
VBS-gel containing 20 mM lysine and incubated at room temperature for 30
minutes. The suspension was centrifuged and the supernatant recovered.
Approximately 90% of the radioactivity originally associated with the
bacterial pellet was dissociated by the lysine treatment. The dissociated
plasmin in the supernatant was then subjected to gel filtration on a G-25
column to separate lysine from plasmin. Fractions containing plasmin
were collected and stored at -20C.

Results
Twenty hemolytic streptococcal isolates were grown overnight at
37C and tested for their ability to bind radiolabeled plasminogen,
plasmin, urokinase, or trypsin as described in the Methods Section. The
results (see Table 2-1) showed that all twenty group A isolates bound
plasmin but failed to bind significant quantities of plasminogen or any
of the other labeled proteins, i.e. less than 10% of the offered label.
Furthermore, the expression of plasmin binding ability was shown to be
present on bacteria grown in either Todd-Hewitt broth or chemically
defined media (data not shown). Plasmin binding was found to be
dependent on the concentration of bacteria and was maximal within two
minutes at 37C. Pre-incubation with excess unlabeled plasmin prevented
binding of the labeled plasmin. In the absence of unlabeled plasmin,
strain 64/14 consistently bound approximately 60% of the radioactive
plasmin offered and was used to analyze further the selective plasmin
binding activity. In my initial attempts to characterize the
differential binding of plasminogen and plasmin to a group A
streptococcus I compared the kinetics of generation of plasmin from
plasminogen with the ability of labeled protein to bind to the bacteria.
Conversion of plasminogen to plasmin occurs when a single arginine-valine
bond is split in the zymogen by action of the enzyme urokinase (Groskopf
et al. 1969). The zymogen activation reaction can be monitored on SDS-
polyacrylamide gels following reduction of disulfide bonds.
The zymogen molecule migrates as a single polypeptide chain with a
Mr of approximately 90,000 daltons. The active enzyme plasmin migrates
under these conditions as two distinct polypeptide chains (a heavy chain
with an Mr of approximately 60,000 daltons and a light chain with an Mr

19
Table 2-1.
Binding of radiolabeled proteins to various nephrltogenic
and non-nephritogenic group A streptococci.
STRAIN
M-TYPE
PLASMINOGEN
PLASMIN
UROKINASE
TRYPSIN
A992*
18
-
+
-
-
B923
12
-
+
-
-
D897*
12
-
+
-
-
B512
4
-
+
-
-
B438
18
-
+
-
-
B512
NT
-
+
-
-
A928
55
-
+
-
-
64/14
NT
-
++
-
-
B905
2
-
+
-
-
B281
12
-
+
-
-
B920
49
-
++
-
-
B915
49
-
+
-
-
A374
12
-
+
-
-
B931*
2
-
+
-
-
A207
2
-
+
-
-
F2030
1
-
+
-
-
A547
NT
-
+
-
-
64/P
NT
-
++
-
-
648
1
-
+
-
-
A995
57
-
+
-
-
* Non-nephritis causing strains
- = Less than 10% bound of total counts offered
+ 10% to 30% bound of total counts offered
++ = Greater than 30% bound of total counts offered
Approximately 3 X 10 bacteria/tube heat killed at 80C for 10 min.
NT = Not typable

20
of approximately 25,000 daltons). The activation reaction can be stopped
at any time by addition of a 10-fold molar excess of p-nitropbenyl p-
guanidinobenzoate (pNpGB). Consequently, it is possible to obtain
plasminogen in various stages of activation and compare the ability of
the labeled proteins to bind to a group A streptococcus. The results
presented in Figure 2-1, panel A demonstrate that the activation of
plasminogen to plasmin could be readily monitored. As the conversion of
plasminogen to plasmin proceeded, an increase in the binding of labeled
protein occurred which correlated with the concentration of plasmin in
the reaction mixture (Figure 2-1, panel B).
The conversion of plasminogen to plasmin yields a serine active site
that is not expressed in the zymogen. In the next series of experiments
the role of the active site in binding of the enzyme to the bacteria was
assessed. Plasmin was treated with the active site titrant pNpGB, the
small naturally occurring inhibitor aprotinin (Fritz and Wunderer, 1983),
the selective histidine modifying agent, phe-pro-arg chloromethyl ketone
[PPACK] (Kettner et al. 1980) and the physiological regulator 02
antiplasmin (a^-AP) (Mori and Aoki, 1976). The ability of the various
inhibited forms of plasmin to bind to a group A streptococcus was
measured. The results presented in Figure 2-2 demonstrate that plasmin
treated with pNpGB, aprotinin, or PPACK were all capable of binding to
the bacteria in the presence of -AP. By contrast unmodified plasmin
incubated with the physiological inhibitor, a2Ai>> failed to bind. Each
of the plasmin-inhibitor samples that had been incubated with excess 2"
AP was monitored on non-reducing SDS-polyacrylamide gels for the
formation of a high molecular weight complex. The high molecular weight
band observed in the third lane indicates the formation of a stable

Figure 2-1. Binding of plasmin to bacteria: comparison of the kinetics
of generation of plasmin and its ability to bind to the group A
streptococcus 64/14: Labeled plasminogen was converted to plasmin by
treatment with urokinase. The kinetics of generation of plasmin was
monitored on SDS-polyacrylamide gels under reducing conditions. The
conversion of single chain, high molecular weight, plasminogen (Mr
approximately 90,000) into heavy (Mr approximately 60,000) and light
chains (Mr approximately 25,000) of plasmin was monitored. At each time
point the ability of labeled proteins to bind to the group A
streptococcus 64/14 was measured as described in the Methods. The data
are presented as the mean the standard deviation of duplicate
experiments.

22
3 6 9 12 15 24 36
TIME (MINUTES)

Figure 2-2. Effect of inhibiting the active site of plasmin on its
ability to bind to the group A streptococcal strain 64/14: The lower
panel demonstrates the binding of the group A streptococcus to labeled
plasminogen, plasmin, plasmin pretreated with excess a2PI, plasmin
treated with excess pNpGB, plasmin treated with excess aprotinin or
plasmin treated with excess phe-pro-arg chloromethyl ketone,[PPACK], as
described in the Methods. The data are presented as the mean the
standard deviation of duplicate experiments. The upper panel
demonstrates the analysis of each of the plasmin-inhibitor samples that
had been incubated with excess a2PI. Samples were monitored on non
reducing SDS-polyacrylamide gels for the formation of a high molecular
weight complex.

% OF TOTAL RADIOACTIVITY BOUND
PLASMINOGEN
PLASMIN
PLASMIN +a2PI
PLASMIN pNpGB + a2 PI
PLASMIN APROTININ + a2PI
PLASMIN -PPACK + a2PI

25
complex of plasmin with its physiological inhibitor a^'AP (Figure 2-2,
upper panel). Pretreatment of plasmin with aprotinin, pNpGB, or PPACK
inhibited the ability of the enzyme to form the covalent linkage
inhibition reaction with c*2-AP demonstrating that under the experimental
conditions used the plasmin active site was modified.
The next series of experiments were designed to determine whether
bacterial bound plasmin was capable of retaining its enzymatic activity.
Radiolabeled plasmin was generated and incubated with a suspension of
group A streptococci for 40 minutes at 37C. The bacteria with the
associated plasmin were recovered by centrifugation, washed twice with
buffer, and then tested for their ability to cleave the chromogenic
synthetic substrate H-D-val-leu-lys-paranitroanilide (as described in
the Methods). In these experiments a control of bacteria alone failed to
hydrolyze the chromogenic substrate, while bacteria pre-incubated with
plasmin were found to cleave the substrate efficiently. The ability of
bacterial bound plasmin to be affected by a variety of different
inhibitors was tested. The results in Figure 2-3 demonstrate that
addition of pNpGB, PPACK, or aprotinin to the bacterial bound enzyme was
capable of inhibiting its enzyme activity for the synthetic substrate.
By contrast, addition of -AP failed to reduce the enzyme activity
(Figure 2-3). All inhibitors were used in excess of that required to
totally inhibit an equivalent concentration of plasmin in the fluid
phase. Since c*2-AP failed to regulate the bacterial bound enzyme, one
might predict that the large molecule fibrin, the natural substrate of
plasmin, would also be prevented from occupying the substrate pocket in
the active site. To test this prediction, bacteria with plasmin bound
to their surface were placed on a fibrin plate and their ability to

Figure 2-3. Regulation of bacterial bound enzyme activity by a variety
of different serine protease inhibitors: Bacterial pellets were pre
incubated with plasmin, washed and resuspended in buffer containing
excess pNpGB, PPACK, aprotinin, a2-AP or buffer alone for 15 minutes at
room temperature. Following incubation with the inhibitor the bacteria
were pelleted and washed. Enzyme activity was then measured by the
ability of the samples to hydrolyze the chromogenic substrate HD-Val-
leu-gly-paranitroanilide as described in the Methods. The data are
presented as the mean the standard deviation of duplicate experiments.
The hydrolysis by bacterial bound plasmin in the absence of any
inhibitor represents 100% activity.

27
ALONE BOUND BOUND BOUND BOUND BOUND
PLASMIN PLASMIN PLASMIN PLASMIN PLASMIN
+
+
PNPGB
PPACK APROTININ
ajpi

28
mediate dissolution of the fibrin clot was measured. The results
presented in Table 2-2 demonstrate that the bacterial bound plasmin still
retained its ability to cleave fibrin. These effects could not be
accounted for by dissociation of plasmin from the bacteria, since clot
lysis did not occur when the microbe-plasmin complex was separated from
the clot by a 0.22 /m Millipore filter (Table 2-2). Under these
experimental conditions, unbound plasmin was capable of passing through
the filter and causing fibrin degradation.
The following series of experiments, which characterize further the
interaction of human plasmin and this group A streptococci bacteria were
performed by Dr. Tim A. Broeseker, a Fellow in the department of
medicine, division of hematology at the University of Florida.
In his initial experiments the binding of labeled plasmin or
plasminogen to the group A streptococcal strain 64/14 as a function of pH
was tested. Labeled proteins were pre-equilibrated in VBS-gel buffers of
differing pH's before the addition of bacteria. After an incubation
period of 15 minutes at 37C, the radioactivity associated with the
bacteria was measured by pelleting the micro-organisms and washing free
the unbound label with buffer of the appropriate pH as described in the
Methods. Maximal binding of plasmin to the bacteria was observed between
pH 5 and 8 with approximately 60% of counts offered being bound by the
group A streptococcus 64/14, (Figure 2-4). In contrast, addition of
labeled plasminogen to the bacteria over the entire pH range tested
(pH 5-9) resulted in direct binding of less than 10% of offered counts.
(Figure 2-4). Similar studies were carried out to determine the effect of
ionic strength on binding of radiolabeled plasmin and plasminogen to the
group A streptococcal strain 64/14. Labeled proteins were pre-

29
Table 2-2.
Ability of bacterial bound plasmin to solubilize a fibrin clot.
Sample
Hydrolysis of Fibrin3
Direct^
Indirect0
Bacteria Alone
Bacterial Bound Plasmin
++
Plasmin Alone
+++
+++
aUnder the experimental conditions described in the Methods section, a
(+++) reaction represented a zone of clearing with a diameter of 1.0-1.5
cm, a (++) reaction represented a zone of clearing of 0.5-1.0 cm, a (+)
reaction represented a zone of clearing from 0-0.5 cm, and (-)
represents no clearing.
^Sample placed directly onto a fibrin plate.
cSample placed onto a filter placed between the bacteria and the fibrin
plate.

Figure 2-4. Binding of 125I-plasmin or 125I-plasminogen to the group A
streptococcal strain 64/14 as a function of pH: The data are presented
as the mean the standard deviation. Measurements of duplicate
experiments were performed and are expressed as the percent of total
counts offered (20,000 cpm) which were associated with the bacterial
pellet. (O O) 125I-plasmin; (O O) 125I-plasminogen.

31

32
equilibrated in NaCl solutions of varying ionic strength before the
addition of bacteria. Following an incubation period of 15 minutes at
37C the bacteria were washed with solutions containing the appropriate
concentration of NaCl and the number of counts associated with the
bacteria determined. The results in Figure 2-5 demonstrate that plasmin
binding was dependent on ionic strength and that optimal binding occurred
between 0.1 and 0.4 M NaCl. In this range of salt concentrations, less
than 10% of plasminogen bound to bacteria. As the ionic strength was
lowered below 0.075 M NaCl, significant binding of plasminogen to the
bacteria was observed.
Binding of labeled plasmin to the group A streptococcal strain 64/14
was examined in the presence and absence of divalent cations to determine
if these metal ions were important for plasmin binding. Binding studies
were carried out in VBS-gel at pH 7.4 containing 0.00015 M Ca++ and 0.001
M Mg++ or in metal free VBS-EDTA-gel at pH 7.4. After incubation at
37C for 15 minutes, the bacteria were washed twice with the appropriate
buffer and radioactivity associated with the bacterial pellets was
measured. The amount of plasmin bound by the bacteria was the same in
the presence or absence of divalent cations, (data not shown).
After identification of the optimal binding conditions for the
plasmin:bacterium interaction, the affinity of the plasmin receptor for
its ligand was determined in 0.15 M VBS-gel buffer at pH 7.4. In the
initial studies the group A /S-hemolytic strain 64/14 was used.
Preliminary kinetic studies were conducted to establish first that
equilibrium between bacterial bound and free plasmin had been achieved,
and second the conditions under which saturation of bacterial plasmin
receptors could be demonstrated. Binding equilibrium was found to be

Figure 2-5. Binding of 125I-plasmin or 125I-plasminogen to the group A
streptococcal strain 64/14 as a function of ionic strength: The data
are presented as the mean the standard deviation. Measurements of
duplicate experiments were performed and are expressed as the percent of
total counts offered (20,000 cpm) which were associated with the
bacterial pellet. (O O) 125I-plasmin; (O O) 125I-plasminogen

34

35
established within 15 minutes at 37C. Under these conditions using 10^
bacteria and increasing concentrations of labeled plasmin, it was
possible to demonstrate a plateau in plasmin binding capacity consistent
with saturation of bacterial receptors, as shown in Figure 2-6. To
correct for non-specific binding of radiolabel, similar studies were
carried out in which binding of radiolabeled plasmin to the bacteria was
measured in the presence of 100-fold molar excess of unlabeled plasmin.
Non-specific binding demonstrated a linear relationship to counts
offered and was less than 5% in all tubes (data not shown). Analysis of
this data by least squares and Scatchard analysis demonstrated a Kp of
approximately 5 x 10"H M for the association of plasmin with its
receptor on the mouse passaged group A streptococcus strain 64/14.
Scatchard analysis of the binding data indicates that there is a single
population of plasmin receptors on streptococci (see inset, Figure 2-6),
and that strain 64/14 possesses approximately 800 receptors per
bacterium. Similar studies were carried out with the group A strain 64/P
(the original isolate that was used in the previous mouse passage studies
(Reis et al.. 1984)) and with strain 64/3 isolated following three
passages of strain 64 in mice. The 64/3 strain was less virulent in mice
than 64/P which in turn was much less virulent than the strain recovered
after 14 mouse passages, 64/14 (Reis et al., 1984). The 64/P and 64/3
strains were studied using the same protocol described for the
generation of the data using strain 64/14. The results indicated that
the plasmin receptor on 64/P had a Kp of approximately 1 x 10M and
each bacterium expressed approximately 200 receptors. The 64/3 strain
displayed a plasmin receptor with a Kp of approximately 6 x 10"^ M and
approximately 3,500 receptor sites per bacterium. Analysis of the

Figure 2-6. Specific binding of 125I-plasmin to 107 group A
streptococci, strain 64/14. following a 15 minute incubation at 37C in
VBS-gel at pH 7.4: Measurements of triplicate experiments were
performed. Specific binding was determined as described in Methods.
The inset represents the Scatchard analysis of the specific binding
data.

COUNTS BOUND
37

38
Scatchard plots of 64/P and 64/3, like that shown for 64/14 in Figure 2-
6, demonstrated only a single class of plasmin receptors expressed on
these bacteria (Broeseker et al., 1988). Plasmin(ogen) contains lysine
binding sites which also bind analogous amino acids (Winn et al., 1980).
Epsilon-aminocaproic acid (EACA) approximates the side chain structure of
lysyl residues incorporated in intact proteins and has higher affinity
than lysine for these sites on plasmin(ogen), whereas arginine binds with
lower affinity (Winn et al., 1980). In order to assess the possible role
of the lysine binding sites of plasmin in its interaction with the
bacterial receptor, the binding of plasmin to the group A streptococcus
strain 64/14 in the presence of EACA, lysine, or arginine was determined.
Binding was measured in 0.15 M VBS-gel, pH 7.4, containing amino acid
in increasing concentrations. The percentage inhibition of binding was
determined by comparison with the binding in VBS-gel pH, 7.4, buffer
alone. The results of these studies are presented in Figure 2-7 and
demonstrate that binding of plasmin to the group A streptococcus 64/14
was inhibited by each amino acid in a concentration dependent fashion.
Fifty percent inhibition of binding of plasmin to the bacteria was
observed at an EACA concentration of 0.15 mM, a lysine concentration of
2.0 mM, and an arginine concentration of 25 mM. In similar studies,
plasmin was pre-bound to the group A streptococcus and a concentration
dependent elution of bound radiolabel was observed on incubation with
EACA, lysine, or arginine (Figure 2-8). The concentration of amino acid
required to elute 50% of the bound plasmin was approximately equivalent
to that required to inhibit plasmin binding by 50% (compare Figures 2-7
and 2-8).

Figure 2-7. Inhibition of binding of 125I-plasmin to the group A
streptococcal strain 64/14 in VBS-gel containing various concentrations
of epsilon-aminocaproic acid, lysine, and arginine: Measurements of
duplicate experiments were performed and the data are presented as the
mean the standard deviation. The percent inhibition of binding was
calculated by comparing with binding in VBS-gel alone. (O O)
epsilon- aminocaproic acid; (O O) lysine; (0---0) arginine.

40

Figure 2-8. Elution of 125I-plasmin from group A streptococcal strain
64/14 in VBS-gel containing various concentrations of
epsilon-aminocaproic acid, lysine, and arginine : Measurements of
duplicate experiments were performed and the data are presented as the
mean the standard deviation. Percent eluted was calculated by
comparing the radioactivity associated with the bacterial pellet before
and after incubation in the given amino acid solution. (O O)
epsilon-aminocaproic acid; (O O) lysine; (0-*-*-0) arginine.

42
#
.01

43
Discussion
Plasminogen, an inactive zymogen can be converted to the protease
plasmin by a variety of plasminogen activators (Collen, 1980). This
enzyme demonstrates broad substrate specificity. In addition to fibrin
cleavage, plasmin can activate the first component of the classical
complement pathway, hydrolyze coagulation factors, degrade components of
basement membrane, and break down connective tissue (Atichartakarn et
al. 1978; Jones and DeClerck, 1980; Liotta et al.. 1981). Furthermore a
variety of potent split products are generated as a consequence of
plasmin activity, e.g. chemotactic fibrinopeptides (Kay et al., 1974).
Effective regulation of plasmin activity is therefore important in order
to prevent tissue damage and inflammation. Normally the selective
protease inhibitor a^-antiplasmin regulates plasmin activity in man (Aoki
et al., 1977).
Interaction of streptococci and streptococcal products with the
fibrinolytic system has been recognized for many years (Tillett and
Sherry, 1949). The observation that certain streptococci could lyse a
fibrin clot lead to the identification and isolation of streptokinase.
This secreted protein is known to bind to human plasminogen and plasmin
with a similar affinity (Reddey and Markus, 1972). In this study I have
identified a surface receptor on certain group A streptococci, grown in
either Todd-Hewitt broth or chemically defined media, that specifically
binds to plasmin while demonstrating no significant affinity for the
zymogen form of the molecule, plasminogen. Thus the surface receptor we
have identified is distinct from streptokinase. Furthermore, this
binding phenomenon did not appear to be simply a function of the ligand
being a serine protease, since no binding activity was demonstrated with
the two other serine proteases examined, trypsin and urokinase.

44
Binding of plasmin to its bacterial receptor does not inhibit the
ability of the enzyme to cleave either small synthetic substrates or its
natural substrate fibrin. Aprotinin, a naturally occurring tight-binding
inhibitor of plasmin and phe-pro-arg chloromethyIketone which chemically
modifies the histidine residue of the active site can react with the
bound plasmin and neutralize its enzymatic activity. These findings
suggest that the catalytic portion of the plasmin molecule is not
interfered with by the association with the bacteria. Of interest was
the observation that the enzymatic activity of bacterial bound plasmin
could not be regulated by addition of its specific inhibitor, o^-
antiplasmin. Alpha2-antiplasmin is a potent inhibitor of plasmin in the
fluid phase forming a 1:1 stoichiometric complex between the enzyme and
inhibitor.
The failure of a^-antiplasmin to regulate bacterial bound plasmin
provides the bacteria with a potential mechanism for tissue invasion by
virtue of the ability of plasmin to hydrolyze components of connective
tissue and basement membranes. Recent studies of the invasive
characteristics and metastatic potential of tumor cells has suggested a
key role for plasminogen activators in this process (Dano et al., 1985).
The ability of certain group A streptococci to produce a plasminogen
activator (e.g., streptokinase) and also to express a receptor for the
activation product plasmin may account for certain of its invasive
properties. Furthermore, since plasmin bound to a group A streptococcus
is incapable of inhibition by a^-antiplasmin the bacteria has associated
with it a non-regulatable proteolytic activity that may help to
contribute to its tissue invasive properties.
A variety of receptors for human proteins have been described on
streptococci. These include receptors for key components of the immune

45
system such as Clq (Yarnall et al.. 1986), IgG (Kronvall, 1973), and IgA
(Russell-Jones et al., 1984), the serum protein fibrinogen (Kronvall et
al., 1979), and the stromal structural proteins laminin (Switalski et
al., 1984), and fibronectin (Myhre and Kuusela, 1983). The importance of
any of these receptors in the pathogenic process remains controversial.
Our recent observation that streptococci also have a receptor that is
specific for human plasmin adds to this list (Lottenberg et al., 1987).
Although the primary substrate for plasmin is fibrin, plasmin is a
non-specific protease capable of also hydrolyzing such extracellular
matrix proteins as thrombospondin, fibronectin, and laminin, while also
exposing matrix components for degradation by other enzymes (Knudsen et
al., 1986). The ability to bind plasmin in an active form which can not
be regulated by its efficient physiological regulator o^-antiplasmin
could provide for surface mediated protease activity and a mechanism for
tissue invasiveness by plasmin receptor-positive bacteria (Lottenberg et
al., 1987).
In this study it was shown that the group A streptococcus strain
64/14 demonstrates optimal binding of its ligand at physiological pH and
ionic strength. The interaction had a high affinity (Kp =5 x 10'^ M)
and demonstrated a linear Scatchard plot indicating that a single
population of plasmin receptors was present on the bacteria (Broeseker et
al., 1988). There was no evidence for either an additional low affinity
receptor or for any cooperativity, positive or negative, in the binding
of ligand with the specific receptor. In agreement with our previous
observations (Lottenberg et al., 1987), there was no evidence for
specific interaction between bacteria and the native zymogen form of the
protein, plasminogen.

46
Plasmin(ogen) has several lysine binding sites located on its heavy
chain. The low affinity sites are primarily important for binding to
fibrin and the high affinity site is important for the interaction with
2-antiplasmin (Wiman et al.. 1979). In order to assess the possible
role of these lysine binding sites in the interaction between plasmin and
its bacterial receptor, the bacterial binding of plasmin in the presence
of increasing concentrations of EACA, lysine, or arginine. The results
in Figure 2-7 and 2-8 demonstrate that binding could be inhibited, and
bound plasmin could be eluted, by these amino acids in a
concentration-dependent manner. Eluted plasmin could be re-bound to
bacteria simply by removing lysine from the eluted plasmin solution,
indicating a possible importance of the lysine binding sites for the
receptor:plasmin interaction. There are 4 or 5 low affinity sites (Kq =
5mM) and one high affinity site (K¡) = 9/zM) for EACA (Markus et al. ,
1978). Lysine and arginine bind to the high affinity site less tightly
than does the lysine analog EACA (Wiman and Collen, 1978). Analysis of
the inhibition curves for EACA, lysine, and arginine reveal that
occupancy of the high affinity lysine binding site on plasmin interferes
with binding to the bacteria. It is recognized that occupancy of the
high affinity lysine binding site causes gross conformational changes in
the plasmin(ogen) molecule, and therefore the possibility for allosteric
as well as direct effects needs to be considered. The very high affinity
of the receptor for plasmin, approximating the affinity of <*2-antiplasmin
for plasmin (Kp = 2 x 10'^ M, Wiman and Collen, 1978), suggests that
streptococci may be able to compete effectively with a2-antiplasmin for
plasmin. and could explain why bacterial bound plasmin cannot be
regulated by Q2*antiplasmin. The enzymatic inhibition of plasmin by

47
a2_antiplasmin occurs in a two-step process (Christensen and Clenunenson,
1977; Wiman and Collen, 1978). The first step is a non-covalent
association of the a^-antiplasmin molecule with the lysine binding site
located in the kringle 1 region of the plasmin molecule, followed by a
rapid covalent linkage to the plasmin active site serine residue. The
observation that lysine or EACA effects, on the high affinity site of
plasmin, disturbs the plasmin bacterial receptor interaction, together
with the observation that bound plasmin is not inhibited by a^-
antiplasmin, suggests that the kringle 1 domain may be important in the
binding of plasmin to bacteria, or may be inaccessible.
Previous studies have shown that passage of streptococci in mice
heightened virulence with concomitant enhanced expression of certain
surface proteins (Burova et al., 1980; Burova et al., 1981; Reis et al.,
1984). Group A streptococcal strain 64 exhibits decreased expression of
Fc receptors after 3 or 4 mouse passages (strain 64/3 and 64/4) as
compared to the parent strain (64/P) (Reis et al., 1984). Following 8
mouse passages this group A streptococcus demonstrates markedly enhanced
Fc receptor expression which appears to be a stable characteristic of the
selected strain (Reis et al., 1984). In this study the average affinity
of the plasmin receptor expressed on strains 64/P, 64/3, and 64/14 was
not significantly different, indicating that mouse passage did not have a
major selective pressure on the affinity of the plasmin receptor.
There was some variation in the number of plasmin receptors
calculated for each bacteria. The 64/P, 64/3, and 64/14 strains
displayed 200, 3500, and 800 receptor sites, respectively, per bacterium.
Clearly these are average numbers of receptors per bacterium (Broeseker
et al. 1988), and given the potential errors in estimating bacterial

48
number, the effect of phase variations in the expression of different
proteins by bacteria (Cleary e£ al., 1987), and the heterogeneity in
receptor expression among colonies (Yarnall et al., 1984), we believe
that such differences need to be cautiously interpreted. It does not
appear from the results of these studies that the degree of plasmin
receptor expression correlates with the virulence of these group A
isolates in mice. Nonetheless we calculate that binding of active
plasmin in the picomolar range with any of the group A isolates studied
is achievable. The high affinity for and slow off rate of bound plasmin
may make these interactions with streptococci of importance in the
infectious process.
The next series of studies, described in the following chapter, were
designed to analyze the way in which the bacterial receptor associated
with its ligand, the human plasmin molecule.

CHAPTER THREE
LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN BINDING TO ITS
SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR
Introduction
The studies documented in Chapter Two demonstrate that certain
pathogenic group A streptococci, grown in either Todd-Hewitt broth or
chemically defined media, express a receptor that binds to human plasmin
while demonstrating no significant reactivity with the native zymogen
form of the protein, Glu-plasminogen or with other serine class
proteases. Bacterial-bound plasmin retains its enzymatic activity and
can no longer be regulated by its physiological inhibitor, a^-
antiplasmin. Optimal binding of plasmin to its bacterial receptor was
shown to occur under physiological conditions of ionic strength and pH.
This interaction of plasmin with a group A streptococcus had a high
affinity with an estimated dissociation constant of approximately 1.0 x
10"-*- M. Plasmin binding was inhibited reversibly by lysine or epsilon
amino caproic acid, (EACA). These data suggest that the lysine binding
kringle structures of the plasmin molecule might be involved in the
association of plasmin with the bacterial receptor. In this chapter I
describe the experiments performed to localize the region of the plasmin
molecule which interacts with the bacterial plasmin receptor. Binding of
plasmin to a group A streptococcus is dependent on the conformation of
the plasmin molecule, and involves interactions that are distinct from
those occurring between other known plasmin(ogen) binding molecules like
streptokinase, fibrin, fibrinogen, thrombospondin, or a2-antiplasmin.
49

50
Materials and Methods
Enzymes. Inhibitors and other Reagents
Urokinase and porcine elastase (type IV) were obtained from Sigma
Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from
Mobay Pharmaceuticals, New York, NY. Phe-Pro-Arg-chloromethylketone
(PPACK) was obtained from Calbiochem-Behring, San Diego, CA. Human Lys-
plasminogen was obtained from American Diagnostica Inc., Greenwich, CT.
H-D-Val-Leu-Lys-paranitroanilide (S-2251) was obtained from Helena
Laboratories, Beaumont, XX.
Human Plasminogen
Native human plasminogen (Glu-plasminogen) was prepared from human
plasma by chromatography on lysine-Sepharose and molecular sieving
chromatography on Superse 6 (Pharmacia-FPLC, Piscataway, NJ). The
purified protein appeared as a single band on a silver stain of an SDS-
polyacrylamide gel. Plasminogen was quantified by measuring absorbance
using a value of 17.0 (Nilsson et al. 1982). The protein was
also quantified antigenically by Laurell Rocket electrophoresis (Laurell,
1966). The purity of the isolated human plasminogen was confirmed by
activation of a known quantity of plasminogen with streptokinase and
measuring amidolytic activity. The observed and theoretical predicted
enzymatic activity were equivalent, within experimental error. Human
Lys-plasminogen, a modified form of Glu-plasminogen in which 76 of the
NH2'terminal amino acid residues are removed (Glu-1 to Lys-76) was
obtained from American Diagnostica Inc., Greenwich, CT. The homogeneity
of this Lys-plasminogen preparation was analyzed using both an urea gel
electrophoresis procedure and an acetic acid urea gel electrophoresis
procedure. This Lys-plasminogen preparation demonstrated the appropriate

51
migratory property (a shift to a lower Mr of approximately 85,000
daltons) in comparison to native Glu-plasminogen (Mr of approximately
92,000 daltons).
Iodination of Proteins
Glu- and Lys-plasminogen were iodinated by the chloramine T method
using Iodobeads (Pierce Chem. Co., Rockford, IL) as described by
Markwell (1982). The labeled proteins were separated from free iodine by
passage over a G-25 column (PD-10, Pharmacia) and collected in 0.15 M
Veronal buffered saline, pH 7.4, containing 0.001 M Mg++, 0.00015 M Ca++,
and 0.1% gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. The concentration of 1^1-
plasminogen was determined antigenically using a sandwich enzyme-linked
immunosorbent assay (ELISA) technique utilizing goat anti-human
plasminogen IgG fraction from Atlantic Antibodies, Scarborough, ME. This
assay could measure plasminogen reliably in the nanogram range.
Generation of Plasmin
Lys-plasmin was generated from radiolabeled or unlabeled Glu- or
Lys-plasminogen by incubation with urokinase (20 units/ml) in VBS-gel
(unless stated otherwise) that contained 0.04 M lysine. The conversion
of the single chain zymogen molecule to the two chain plasmin enzyme was
monitored on SDS-PAGE under reducing conditions as described previously
by Lottenberg et al., (1987). Conversion of the zymogen to the active
enzyme was maximal after 30 min incubation at 37C. Glu-plasmin was
generated by a similar procedure with the exception that a 10-fold molar
concentration of aprotinin relative to the Glu-plasminogen concentration
was added prior to addition of urokinase (Swenson and Thorsen, 1981).
Mini-plasmin was generated from mini-plasminogen using the same

52
activation procedure described to generate Lys-plasmin. Phe-pro-arg-
chloromethyIketone (PPACK) reacted radiolabeled or unlabeled plasmin was
obtained by mixing a 5-fold molar excess of the inhibitor with plasmin
and incubating at 37C for 30 min.
Bacteria
The group A, /3-hemolytic, streptococcal strain 64/14 was grown in
Todd-Hewitt broth (Difco, Detroit, MI) overnight at 37C as stationary
cultures (Yarnall and Boyle, 1986a). The bacteria were harvested by
centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4,
containing 0.05% Tween 20 and 0.02% sodium azide. The bacteria were heat
killed at 80C for 15 min. The suspension was centrifuged and the pellet
washed twice with VBS-gel containing 0.02% sodium azide. Aliquots were
stored at -20C. Stocks of 10% wet weight/vol suspensions were prepared
in VBS-gelatin containing 0.02% sodium azide. The concentration of a
bacterial suspension was determined by counting bacterial chains in a
Neubauer hemacytometer (Fisher Scientific, Orlando, FL).
Polyacrylamide Gel Electrophoresis
Electrophoresis was carried out as described by Weber and Osborn
(1969) with the addition of 6.0 M urea to the polyacrylamide gel. The
polyacrylamide gels consisted of a 4% stacking gel layered onto a 10% or
12% polyacrylamide gel containing 0.1% sodium dodecylsulfate, 0.05 M
sodium phosphate pH 7.1, 6.0 M urea. Slab gels were used in the Bio-Rad
Protean II system (BioRad, Richmond, CA). Protein samples were prepared
by mixing an equal volume of sample buffer containing 0.1 M sodium
phosphate pH 7.1, 8.0 M Urea, and 4.0% SDS with the protein solution, and
heating at 80C for 2 minutes. Sample buffer containing 0.72 M /3-
mercaptoethanol was used to prepare protein samples in the reduced state.

53
Preparation of Elastase Digestion Fragments of Plasminogen
Elastase digestion of human plasminogen yields three defined
fragments of the plasminogen molecule (Sottrup-Jensen et al., 1978) .
These are 1) the lysine-binding domain I (LBS-I), Mr of approx. 38,000
daltons containing kringle domains 1 through 3, 2) Lysine binding domain
II (LBS-II), Mr of approx. 10-12,000 daltons consisting of the kringle
domain 4, and 3) the non-lysine-binding domain known as mini-plasminogen,
Mr of approx. 36,000 daltons containing the remainder of the heavy chain
(kringle 5) and intact light chain. Elastase digestion was performed
using established conditions (Sottrup-Jensen et al., 1978). Purified
Glu-plasminogen (3.0 mg/ml) in 0.05 M Tris, 0.1 M NaCl, pH 8.0, was
digested with a 40:1 molar ratio of Glu-Plasminogen to porcine elastase
in the presence of 250 KIU/ml aprotinin for 6.5 hours at room temperature
with gentle stirring in a total volume of 20 mis. At this time an
aliquot containing 50 ng of protein was removed for analysis by SDS-PAGE
and silver staining, to determine the extent of plasminogen digestion.
The remainder of the reaction mixture was flash frozen and stored at -
70C. The fragments were subsequently purified by a combination of
affinity chromatography on lysine-Sepharose and gel filtration on
Superse 6 (Pharmacia FPLC). The concentrations of the purified proteins
(see Figure 3-1, panel A) were determined spectrophotometrically, using
previously reported A^^11111 values of 17.0 for both Glu- and Lys-
plasminogen (Holvoet et al., 1985), 14.0 for mini-plasminogen (Holvoet et
al., 1985), 22.5 for LBS I (Nilsson et al., 1982), 25.0 for LBS II
(Nilsson et al., 1982), and 16.0 for plasmin heavy chain and plasmin
light chain (Summaria and Robbins, 1976). All proteins were aliquoted
and stored at -70C.

54
Preparation of Plasmin Heavy and Light Chains
Plasmin heavy and light chains were prepared essentially as
described by Summaria and Robbins (1976) Twenty mg of Lys-plasmin,
enzymatically inhibited with a 5 fold molar excess of aprotinin in 5 mis
of 0.05 M Tris, 0.1 M NaCl, pH 8.0, was reduced by treatment with 0.1 M
/3-mercaptoethanol for 20 min. at 20C. The reduced solution was then
cooled in an ice slurry and carboxymethylated with 0.1 M sodium-
iodoacetate on ice for 10 min. The plasmin heavy and light chains were
then separated and purified by a combination of affinity chromatography
on lysine-Sepharose, concentration by ammonium sulfate precipitation (4.0
g/10 ml), resuspension in 0.05 M Tris, 0.1 M NaCl, pH 8.0, and subjected
to gel filtration on Superse 6 (Pharmacia FPLC). The isolated plasmin
and plasminogen fragments were analyzed for purity on a reduced SDS-6 M-
urea-polyacrylamide gel. As shown in Figure 3-1, panel B, the various
fragments demonstrated appropriate molecular sizes and were homogeneous.
Concentrations were determined as described above. All proteins were
aliquoted and stored at -70C.
Direct Binding Assay of Radiolabeled Proteins
The ability of radiolabeled plasminogen fragments to bind to the
group A streptococcus 64/14 was measured as described previously by
Lottenberg et al.. (1987). A fixed number of bacteria were incubated
with labeled proteins (approximately 30,000 cpm per tube) in a total
volume of 400 nl of VBS-gel for 30 min at 37C. The bacteria were
pelleted by centrifugation at 1000 x g for 10 min and the pellets washed
twice with 2.0 ml of VBS-gel. The radioactivity associated with the
bacteria was determined in a Beckman 5500 Auto gamma counter (Beckman
Instruments, Inc., Fullerton, CA). Non-specific background binding was

55
determined in replicate tubes which contained no bacteria. All estimates
were performed in duplicate.
Inhibition of Plasmin Binding to Bacteria by Purified Plasmin(ogen')
Fragments
The ability of different concentrations of one or more of the
isolated plasminogen fragments to inhibit binding of Phe-Pro-Arg-
Chloromethylketone (PPACK) reacted -*-^^I-Lys-plasmin to the group A
streptococcus 64/14 was tested using a modification of the direct binding
assay described above. Different concentrations of plasmin(ogen) or
plasmin(ogen) fragments were mixed with a fixed dilution of a 10% w/v
suspension of streptococcal strain 64/14 and PPACK reacted -Lys-
plasmin (approx. 30,000 cpm per tube) followed by incubation for 30 min
at 37C. Bacterial associated radioactivity was determined after washing
away unbound label as described above. The inhibition of binding of
labeled plasmin was calculated by comparing the number of counts bound in
the absence of competitor with the number of counts bound when the
competitor was present. All samples were corrected for background
binding of counts. Counts bound in the tubes from which bacteria were
omitted or in tubes in which a 100-fold molar-excess of unlabeled ligand
was added. In no case was the background level of radioactivity greater
than 5% of the counts offered. Furthermore, background levels in the
presence of excess cold competitor, or in the absence of bacteria were
not significantly different.
Elution and Analysis of -Lvs-plasminCogen') from Bacteria
IOC
JI-Lys-plasminogen (approx. 100,000 cpm) was added to a 100 p 1
aliquot of a 10% w/v solution of strain 64/14 bacteria in a total volume
of 400 pi VBS-gel and allowed to incubate at 37C for 30 min. The
bacteria were then pelleted by centrifugation (3000 x g, 10 min) and

56
washed three times with 2.0 ml VBS-gel. The bacterial pellets were
resuspended in 300 pi of either VBS-gel containing 0.5% SDS; VBS-gel
containing 0.1M EACA; or VBS-gel containing 0.5% SDS and 2.0% /?-
mercaptoethanol, to elute the l^I-Lys-plasminogen from the bacteria.
Following a 10 minute incubation at 37 C the bacteria were removed by
centrifugation and the supernatant recovered. The eluted material was
analyzed by electrophoresis on a 10%-SDS-PAGE-6M-Urea gel under reducing
conditions. The gel was dried and the migration of labeled protein
determined by autoradiography. Similar studies were also carried out in
which the bacterial bound -*-^^I-Lys-plasminogen was treated with a 20
unit/ml concentration of urokinase for 20 minutes at 37C in a total
volume of 300 pi of VBS-gel prior to eluting the bound proteins.
Following this plasminogen activation reaction the bacteria were
centrifuged and washed twice with 2.0 mis VBS-gel. The residual bound
-IOC
I-Lys-plasmin(ogen) was eluted and analyzed as described above.
Measurement of Functional Activity of Plasmin(ogen') in Bacterial-Free
Supernatants
The following assay was used to measure binding of the various
plasmin(ogen) as an alternative method to using radiolabeled tracers.
In these studies, 2.0 pg of Glu-plasminogen, Lys-plasminogen, or Lys-
plasmin was incubated with 100 pi of a heat killed 10% w/v suspension of
the group A streptococcal strain 64/14 for 20 minutes at 37C in a total
reaction volume of 400 pi of VBS-gel. Following incubation, the bacteria
were removed by centrifugation at 12,000 x g for 4 minutes in an Ependorf
Microfuge and bacterial-free supernatants were obtained. Control tubes
for each plasmin(ogen) species containing no bacteria were treated
identically and all samples were run in duplicate. The bacterial free
supernatants were recovered and enzymatic activity was measured as

57
follows. The bacterial free supernatants or the corresponding control
samples were added to plastic cuvettes containing 10^ IU streptokinase in
a total volume of 900 il of enzyme assay buffer (0.05 M Tris, 0.05 M
NaCl, 0.1% PEG-8000, pH 7.4). For the Glu-and Lys-plasminogen
preparations the reaction mixture was incubated at 37C for 10 minutes to
allow plasminogen-streptokinase complexes to form. For Lys-plasmin, a
similar incubation with streptokinase was performed to allow for
equivalent substrate turnover to that of zymogen-streptokinase
complexes. Following incubation, H-D-Val-Leu-Lys-pNA (S-2251) was added
to yield a final concentration of 300 /M. Tubes were allowed to incubate
for precisely five minutes and then quenched with 100 /il of glacial
acetic acid. The amount of substrate hydrolysis, which is directly
proportional to the amount of plasmin enzyme present was then quantified
by measuring the absorbance of the reaction mixture at 405 nm.
The enzymatic activity of the bacterial free supernatant was
determined by comparison with the enzymatic activity of known standards.
The percent of residual enzymatic plasmin(ogen) activity in the bacterial
free supernatant was calculated by determining the fraction of total
enzymatic activity in a control sample remaining in the supernatant
following incubation with bacteria. Control tubes containing bacteria
and substrate, and substrate in buffer were included. All assays were
performed in duplicate.
Measurement of Functional Activity of Plasmin Associated with Bacteria
The plasmin activity associated with bacterial pellets was examined
using the chromogenic substrate as described above. Following binding
and centrifugation the pellets were washed 3 times with 1.0 ml of VBS-gel
and resuspended in 400 /I of the enzyme assay buffer. S-2251 was added

58
to yield a final concentration of 300 /M. The resuspended bacterial
pellets were then incubated at 37C for 20 minutes and quenched with 50
pi of glacial acetic acid. The bacteria were removed by centrifugation
(12,000 x g for 4 minutes) and the optical density of the bacterial free
supernatant was measured at 405 nm. Control tubes containing bacteria
and substrate, and substrate in buffer were included. All assays were
performed in duplicate.
Results
The experiments described in this chapter were designed to map the
domains on the human plasmin molecule involved in the high affinity
interaction with the group A streptococcal strain 64/14. For these
studies a variety of defined plasminogen fragments as well as the heavy
and light chains of plasmin were prepared as described in the Methods.
The plasminogen fragments obtained were characterized on urea gels, see
Figure 3-1. The homogeneous plasminogen fragments were used to compete
with intact PPACK reacted ^^^1-plasmin for receptor sites on the group A
streptococcal strain 64/14 (Table 3-1). I have previously demonstrated
that plasmin treated with PPACK, p-Nitrophenyl-p-guanidinobenzoate
(pNpGB), or aprotinin, does not effect plasmin's binding reactivity to
the group A streptococcal strain 64/14 as documented in Chapter Two.
Enzymatic inhibition of the ^i-Lys-plasmin and urokinase in the labeled
tracer preparation was necessary to prevent the proteolytic conversion of
Glu-plasminogen to Lys-plasminogen, or Lys-plasmin (Markus et al., 1978)
by the labeled tracer mixture. In all the competitive inhibition
experiments described in this study, a constant concentration of PPACK-
plasmin (1.0 x 10'^- M) and a range of concentrations of unlabeled

Figure 3-1. SDS-UREA-PAGE analysis of isolated olasmin(ogen) fragments.
Panel A: Elastase digestion fragments of plasmin(ogen): Glu-plasminogen
was digested with elastase and fragments purified as described in Materials
and Methods. Panel A: Lane 1: Mini-PLG (4.0 Mg); 2: Glu-PLG (4.0 Mg); 3:
Mini-PLA (4.0 Mg); 4: LBS-I (4.0 Mg); 5: LBS-II (4.0 Mg); M: molecular weight
standards. Panel B: Plasmin heavy (HC) and light (LC) chain preparations:
Lys-plasmin was reduced and carboxymethylated as described in Materials and
Methods. Lane 1: Glu-PLG (5.0 Mg) I 2: lys-PLA (5.0 Mg); 3: HC (5.0 Mg) I t*'-
LC (5.0 Mg)- Proteins were electrophoresed under reducing conditions on a
SDS-6 M urea-10%-polyacrylamide gel.

A
B
plasminogen- I
heavy chain
-I ight chain
as
O

61
competitor molecules (10" M to 10'^ m) were mixed with a fixed
concentration of bacteria. Following incubation and washing, the amount
of radiolabeled plasmin bound to the bacterial pellet was determined.
The quantity of radioactivity bound in the presence or absence of
unlabeled competitor was compared and the degree of inhibition calculated
(Figure 3-2) The results summarized in Table 3-1 show that unlabeled
plasmin inhibits the binding of labeled plasmin efficiently, with 50%
inhibition being observed in the presence of 1.2 x 10' M Lys-plasmin.
Significant inhibition of radiolabeled plasmin binding was also observed
when purified heavy chain was used as the competitor. Addition of any of
the other plasminogen fragments including isolated lysine binding domains
of the heavy chain (LBS I or LBS II) demonstrated no significant
inhibitory effect (Figure 3-2). Similarly, mini-plasminogen, mini-
plasmin, and isolated light chains demonstrated no significant inhibition
of binding of radiolabeled lys-plasmin over the concentration range
tested (10' M to 10"^ M) (Figure 3-2). Identical results were obtained
in the inhibition assays involving mini-plasmin, mini-plasminogen, LBS I,
LBS II, Lys-plasmin heavy chain, and plasmin light chain in the absence
of protease inhibitors in the reaction mixture (data not shown).
Combining equimolar quantities of the elastase digested fragments of
plasminogen or plasmin failed to restore any inhibitory potential.
Furthermore, combination of isolated light chain and heavy chain
demonstrated no synergistic effect in inhibitory capacity compared to the
sum of the isolated fragments alone (data not shown). The inhibition
curves for isolated heavy chain and intact plasmin (Figure 3-2)
demonstrated that both preparations could inhibit binding of labeled Lys-
plasmin by 100%. However, these curves differed in shape, indicating

Figure 3-2. Inhibition of. PPACK reacted. 125I-Lvs-plasmin binding to
group A streptococcal plasmin receptor: A constant concentration of
(1.0 x 10 M) PPACK reacted I-Lys-plasmin and an increasing
concentration of unlabeled competitor molecules (10"1 M to 10"6 M) were
mixed with a fixed concentration of streptococcal strain 64/14.
Following incubation and washing (see Materials and Methods), the amount
of radiolabeled Lys-plasmin bound to the bacterial pellet was
determined. The quantity of radioactivity bound in the presence of
unlabeled competitor was compared to the radioactivity bound in the
absence of inhibitor and the percent inhibition calculated. ( -Lys-
plasmin; o-HC; -LBS-I; D-LBS-II; a-LC; a -Mini-PLG; -Mini-PLA).

% INHIBITION OF BINDING
63
CONCENTRATION OF COMPETITOR (M)

64
differences in the efficiency of inhibition. The isolated heavy chain
was found to be less efficient an inhibitor than the intact plasmin
molecule. These findings suggest that there is some component involved
in the interaction of plasmin with the bacteria that is either not
present on the heavy chain or is altered during the isolation procedure.
Two possibilities to account for these observations were considered.
The first was that there are some sites on the heavy chain of the
plasmin molecule that are modified when the molecule is purified, thereby
changing its efficiency of interaction with the bacterial receptor. The
second was that the plasmin light chain, while associated with the heavy
chain, confers a different tertiary structure to the molecule than exists
on either (or both) of the isolated chains. Such a change in
conformation of the molecule might affect its interaction with the
bacteria.
It has been established previously that a conformational change
occurs when Glu-plasminogen is activated to Lys-plasmin, or when Glu-
plasminogen is converted to Lys-plasminogen (Swenson and Thorsen, 1981;
Markus et al.. 1978; Thorsen, 1975). Lys-plasminogen is the zymogen form
of plasminogen, lacking the 76 amino acid NH2-terminus of the native
protein (Markus et al., 1978). This modification results from the
proteolytic activity of plasmin on Glu-plasminogen, which removes the 76
amino acid NH2-terminus, resulting in a new NH2-terminus lysine (for
review, see Thorsen et al., 1981). This modification occurs without
generation of protease activity. The conversion of Glu-plasminogen to
Lys-plasmin or to Lys-plasminogen not only results in a marked
conformational change of the protein but also causes an increase in the
binding affinity of these molecules to fibrin (Thorsen, 1975), as well as

65
lowering the dissociation constant between these molecules and <*2*
antiplasmin (Swenson and Thorsen, 1981; Wiman et al. 1979).
To examine the possible importance of the conformation of the
plasmin(ogen) molecule for binding to bacteria, the ability of the
conformationally altered form of plasminogen, Lys-plasminogen, to bind to
the group A streptococcus, 64/14, was measured. The isolated protein was
radiolabeled and examined by urea gel analysis for homogeneity. The
labeled material demonstrated a single band on an autoradiograph (Figure
3-3, panel A, lane 2) at a position corresponding to that reported for
the migration of Lys-plasminogen in this gel system (Swensen and Thorsen,
1981). This labeled form of plasminogen was found to bind to the
bacteria (Figure 3-3, panel B, lane 2). Similarly, Glu-plasmin
generated from Glu-plasminogen in the presence of aprotinin, was also
capable of binding to the bacteria (Figure 3-3, panel B, lane 3). The
relative efficiency of unlabeled Glu-plasminogen, Lys-plasminogen and
Lys-plasmin to compete with labeled plasmin for binding sites on the
group A streptococcus 64/14 was tested. Different concentrations of each
of these molecules were mixed with a fixed concentration (1.0 x 10'-*- M) ,
of PPACK reacted *^^1-Lys-plasmin and the extent of inhibition of binding
of radiolabel was measured, as described previously. The results of this
experiment shown in Figure 3-4 indicate that the inhibition achieved
with Lys-plasminogen and Lys-plasmin were identical. These results
indicate that the receptor for these ligands are the same, and that the
affinity for each protein is equivalent.
The possibility that the results presented in Figure 3-4 could be
accounted for by the conversion of Lys-plasminogen to Lys-plasmin during
the reaction was considered. The next series of experiments were

Glu-
Figure 3-3. Binding of 125I labeled Glu- and Lvs-plasmin(ogens):
and Lys-plasmin(ogen) were generated as described in Materials and
Methods. The labeled tracers were then used in direct binding assays
with a fixed concentration of the streptococcal strain 64/14. Panel A
is an autoradiograph demonstrating the analysis of each reduced 125I-
labeled sample on a SDS-6 M-Urea 12%-polyacrylamide gel by auto
radiograph to verify their molecular form. (Glu-H: Glu-heavy chain;
Lys-H: Lys-heavy chain; L: light chain). Panel B illustrates the
percent of offered cpm bound to bacterial pellets. (Lane 1: Glu-PLG; 2
Lys-PLG; 3: Glu-PLA; 4: Lys-PLA).

67
A
1. 2. 3. 4.

Figure 3-4. Inhibition of. PPACK reacted, 125I-lvs-plasmin binding to
group A plasmin receptor:A constant concentration of (1 x 10'10 M)
PPACK reacted ^-Lys-plasmin and an increasing concentration range of
Lys-PLG, Glu-PLG or Lys-PLA (1010 M to 10"6 M) were mixed with a fixed
concentration of the streptococcal strain 64/14. Following incubation
and washing (see Materials and Methods), the amount of radiolabeled Lys-
plasmin bound to the bacterial pellet was determined. The quantity of
radioactivity bound in the presence of unlabeled competitor was compared
to the radioactivity bound in the absence of inhibitor and the percent
inhibition calculated. ( O-Lys-PLA; -Lys-PLG; Glu-PLG).

% INHIBITION OF BINDING
69
CONCENTRATION OF COMPETITOR (M)

70
designed to determine whether Lys-plasminogen binds to the bacteria
without first being activated. These experiments were carried out by
monitoring the distribution of Lys-plasminogen, Glu-plasminogen, or Lys-
plasmin, in the fluid phase and associated with the bacteria, following
incubation of the protein with the bacteria. Unlabeled Glu-plasminogen,
Lys-plasminogen, or Lys-plasmin was added to a fixed concentration of the
group A streptococci 64/14 and incubated for 30 minutes at 37C.
Following this incubation period, the bacteria were pelleted by
centrifugation and the supernatants were recovered and monitored for
enzymatic activity either directly for Lys-plasmin, or following
activation with excess streptokinase for the sample containing Glu-
plasminogen or Lys-plasminogen, as described in the Methods. Following
incubation with bacteria, and removal of the bacteria by centrifugation,
there was no significant Lys-plasmin activity detectable in the bacterial
free supernatant (Figure 3-5). By contrast over 98% of the enzymatic
potential of Glu-plasminogen was detected in the supernatant, while in
similar experiments using Lys-plasminogen less than 10% of the enzymatic
potential was measured following activation with streptokinase (Figure
3-6). Because of differences in the efficiency of detection of plasmin
activity in the fluid phase compared to its activity when bound to
bacteria it is not possible to quantitate accurately the exact percentage
of plasmin activity that is bound to bacteria. However, I have
demonstrated previously that once associated with bacteria, the plasmin
retains its ability to cleave synthetic chromogenic substrates like H-D-
val-leu-lys-pNA (S-2251), as documented in Chapter Two. Consequently,
the washed pellets from the absorption reaction were incubated with this
synthetic substrate. The results presented in Table 3-2 demonstrate

Figure 3-5. Binding of Lvs-plasmin(s) derived from Glu-plasminogen and
Lvs-plasminogen. to the group A streptococcal strain 64/14 as measured
by residual activity in the bacterial free supernatant: (O -urokinase
activated Glu-plasminogen alone; -urokinase activated Glu-plasminogen
+ bacteria; -urokinase activated Lys-plasminogen alone; -urokinase
activated Lys-plasminogen + bacteria). For precise experimental details
see Materials and Methods.

O.D. 405nm
72
PLASMIN (nM)

Table 3-1.
Summary of inhibition experiments of PPACK reacted -Lvs-plasmin
binding to the group A streptococcal strain 64/14 are shown with a
schematic depiction of the portion of the native molecule they represent.**
COMPETIT OR Isq%( a*- m)
Glu-Plasminogen
X. X ,
Li,.I
> 2.0
Lys-Plasminogen
* -?9,
0.010
Li
Lys-Plasmin
x, ?90
L*- s s I
0.012
Mini-Plasminogen
> 1.0
442 -^r-g* 790
Mini-Plasmin
Val442 -X- -X- i 790
Is- S 1
> 1.0
LBS-I(K1 -Kg)
> 1.0
lbs-h(k4)
v.,
Va 354
> 1.0
Heavy (A) Chain
0.046
Light (B) Chain
Valgg j 1jfr -X- 1 790
> 1.0
| Activation cleavage site (arginine5gQ-valinegg .j).
* Plasmin active site residues (histidineg02; aspartic acidg45; serine74Q) from left to right.
Structure and NH2-amino-terminal residue data were obtained from the work
of Sottrup-Jensen et al.. (1978). Inhibition is expressed as 50% inhibitory
values in (/M) with Lys-plasmin as the standard (see Materials and Methods).
U)

74
Table 3-2.
Measurement of nlasmin(ogen') associated with bacterial pellets.
Bacteria Pre-Incubated With:
Absorbance at 405 nm
Following a 20 Minute
Incubation at 37C with H-
D-Val-Leu-Lys-Paranitroanilide
Buffer
0.023 0.002
Glu-plasminogen
0.026 0.002
Lys-plasminogen
0.095 0.004
Lys-plasmin^
0.630 0.001
O
Lys-plasmin^
0.637 0.002
Two ng of the indicated enzyme or zymogen was incubated with a
fixed dilution of the streptococcal strain 64/14. Following incubation
the bacteria were pelleted by centrifugation, washed, resuspended in 400
/I VBS-gelatin, and assayed for enzymatic activity by hydrolysis of the
chromogenic substrate H-D-Val-Leu-Lys-paranitroanilide (see Materials and
Methods). The spontaneous cleavage of the substrate under the
experimental conditions in the presence of bacteria alone was an
absorbance (405 nm) of 0.024 0.002.
1. Urokinase activated Glu-plasminogen
2. Urokinase activated Lys-plasminogen

Figure 3-6. Binding of Glu- and Lvs- plasminogen to the group A
streptococcal strain 64/14 as measured by residual activatable zymogen
in the bacterial free supernatant: ( Glu-plasminogen + bacteria;
o-Glu-plasminogen alone bacteria; -Lys-plasminogen + bacteria; o-Lys-
plasminogen alone). For precise experimental details see Materials and
Methods.

O.D. 405nm
76
PLASMINOGEN (nM)

1 PR
Figure 3-7. Characterization of I-Lvs-plasminCogen') species eluted
from bacteria: Eluted labeled proteins were analyzed by electrophoresis
on a 10%-SDS-PAGE-6M urea gel under reduced conditions. Lanes 1, 2, and
3 contain labeled proteins from bacteria pre-incubated with 125I-Lys-
plasminogen and eluted by 0.5% SDS; 0.1 M EACA; or 0.5% SDS containing
2.0% £-mercaptoethanol respectively. Lanes 4, 5, and 6 are identical to
Lanes 1, 2, and 3 with the exception that the bound labeled proteins
were pre-incubated with urokinase prior to elution. Lane 7 contains
125I-Lys-plasminogen incubated at 37C without bacteria for the period
of the experiments and Lanes 8 and 9 contain 125I-Lys-plasminogen and
125-Lys-plasmin respectively. For precise experimental details see
Methods.

78

79
that the bacteria incubated with Lys-plasminogen exhibited only a low
level of enzymatic activity (approximately 15% of that observed in the
samples pre-incubated with Lys-plasmin). The bacterial free supernatant
of the sample incubated with Lys-plasminogen demonstrated <10% of the
zymogen remained in the supernatant as detected by enzymatic activity
following activation with streptokinase (Figure 3-6). Taken together
these results indicate that the Lys-plasminogen was removed from the
fluid phase without prior or concomitant activation to Lys-plasmin.
Furthermore, when the bacterial bound ^^^1-Lys-plasminogen was eluted
from the bacteria and examined by polyacrylamide gel electrophoresis
under reducing conditions, a single protein band was observed on the
autoradiograph corresponding to the enzymatically inactive modified
zymogen form of the protein, Lys plasminogen (Figure 3-7). All of these
studies demonstrate that Lys-plasminogen can bind to bacteria without
first being converted to Lys-plasmin. This would indicate that the
intact native plasminogen molecule (Glu-plasminogen) does not express
structures that are recognized by the bacterial plasmin receptor.
However, following a conformational change achieved by either conversion
to the Lys-plasminogen form of the zymogen or by activation to plasmin,
structures are formed or exposed on the molecule that facilitate
interaction with the bacteria.
Discussion
Plasmin is the key component of the mammalian fibrinolytic enzyme
system which is responsible for fibrin degradation and intravascular
blood clot lysis. Active plasmin, which cleaves fibrin, is derived from
the circulating zymogen precursor Glu-plasminogen. Glu-plasminogen is a

80
single chain glycosylated protein containing 790 amino acids in known
sequence with a molecular weight of approximately 92,000 daltons (Thorsen
et al.. 1981; Wiman, 1973, 1977). The generation of plasmin from
plasminogen is accomplished by proteins known as plasminogen activators.
This conversion is brought about by cleavage of a single arginine (560)-
valine (561) peptide bond which creates, through conformation changes, a
two chain active plasmin molecule held together by disulfide linkages
(Astrup, 1978). The light chain of plasmin has a molecular weight of
approximately 25,000 daltons and contains the serine protease active site
(Robbins and Summaria, 1970; Wiman, 1977). The heavy chain of plasmin
has a molecular weight of approximately 63,000 daltons (Robbins and
Summaria, 1970) and contains 5 homologous triple loop structures known as
kringles (Sottrup-Jensen et al., 1978). An additional conformationally
distinct form of plasminogen can be generated when Glu-plasminogen is
exposed to plasmin. This removes a 76 amino acid peptide from the NH2-
terminus, thereby generating Lys-plasminogen (Swenson and Thorsen, 1981).
The plasmin(ogen) molecule contains several characteristic 'lysine
binding sites', one with high affinity for the lysine analogue EACA
(dissociation constant of 9.0 fiH), and four or five with low affinity
(dissociation constant of 5 mM) (Markus et al., 1978a, 1978b). The high
affinity site has been mapped to the kringle 1 region, and one of the
lower affinity sites has been mapped to the kringle 4 region of the
plasmin(ogen) molecule (Lerch et al., 1980). These structures are known
to participate in the binding of plasmin(ogen) to a^-antiplasmin (c^'AP)
(Wiman, 1981) and to fibrin (Swenson and Thorsen, 1981; Wiman et al.,
1979) respectively. It is known that binding of lysine and lysine
analogs to plasmin(ogen)'s lysine binding sites induces conformational

81
changes in the molecule (Violand et al., 1975). I have shown previously
that lysine or c*2-AP inhibit the binding of plasmin to the group A
streptococcal receptor, as documented in Chapter Two, indicating the
possible involvement of the high affinity lysine-binding site in the
plasmin-bacterial receptor interaction. A comparison of my findings
with studies of the interaction of plasminogen with other naturally
occurring plasminogen binding proteins reveals a number of interesting
similarities and contrasts. Specific binding to the group A
streptococcus, 64/14, was demonstrated with plasmin's heavy chain.
However, the isolated heavy chain alone was not as efficient a competitor
as intact Lys-plasmin, as evidenced by the non-superimposible nature of
the heavy chain and Lys-plasmin inhibition curves (Figure 3-2). It
should be noted that 100% inhibition of binding of Lys-plasmin could be
achieved by addition of high concentrations of heavy chain, but none of
the kringle containing fragments (Lysine-binding domains) alone or in
combination had any significant inhibitory effects at similar molar
concentrations. This finding stresses the importance of the
conformation of the entire heavy chain for binding to bacteria. The
bacterial binding of plasmin therefore differs from the kind of
interaction seen with a^-AP, as well as with fibrin and fibrinogen, to
which plasmin as well as plasminogen, LBS-I, LBS-II, and mini-
plasmin(ogen) are known to interact (Swenson and Thorsen, 1981; Thorsen
et al., 1981; Wiman et al., 1979).
Consistent with my initial observations, documented in Chapter Two,
there is no significant binding of the native zymogen, Glu-plasminogen,
while the conformationally altered form of the zymogen, Lys-plasminogen,
was found to bind specifically to bacteria (Figures 3-3,3-4, and 3-5).

82
This form of the zymogen molecule is known to be conformationally
distinct from Glu-plasminogen and is similar in conformation to Lys-
plasmin (Violand et al.. 1975). The binding of Lys-plasminogen to the
group A streptococcal receptor is therefore dependent on a specific
conformation, most probably of the heavy chain.
Interaction of both plasmin and plasminogen with thrombospondin (an
adhesive glycoprotein) has been demonstrated to occur via interaction
with the heavy chain of the plasmin(ogen) molecule (Silverstein et al.,
1984; Walz et al., 1987). However, efficient binding of thrombospondin
with any elastase digestion fragment of plasminogen has not been observed
(Walz et: al., 1987). I have observed a similar pattern for the
interaction of plasmin(ogen) with its bacterial receptor. However,
unlike the interaction of thrombospondin with plasmin, the bacterial
binding properties were reversible by addition of lysine or lysine
analogs (Broeseker et al., 1988).
Histidine-rich glycoprotein (HRGP) an a2glycoProtein *-n human
plasma, has been reported to compete with a^-AP for the high-affinity
lysine-binding site in plasmin (Haupt and Heinburger, 1972; Lijnen et
al., 1980). In addition, HRGP also reduces the binding of plasminogen to
fibrin by complex formation with the low-affinity lysine binding sites
(Lijnen et al., 1980). Furthermore, the characteristic interaction of
Glu-plasminogen, Lys-plasminogen, or plasmin and their fragments with
fibrin or fibrinogen involves the heavy chain lysine-binding sites
(Cenderholm-Williams, 1977). This is distinct from the profile of
reactivity for the interaction of these proteins with the group A
streptococcus (Table 3-1). It can be seen that Glu-plasminogen shows no
reactivity with these bacteria, nor is there any significant reactivity
with the isolated lysine-binding fragments LBS I or LBS II.

83
Of particular relevance to this study is the interaction of
plasmin(ogen) with the well characterized streptococcal plasminogen
activator streptokinase isolated from group C streptococci (Christensen,
1945; Tillet and Garner, 1933). This secreted streptococcal protein is
known to bind rapidly to Glu-plasminogen, Lys-plasminogen, and Lys-
plasmin (rate constant 5.4 x 10^ forming a 1:1 stoichiometric
complex with an estimated dissociation constant of 5 x 10"^ M
(Cenderholm-Williams et al., 1979). This interaction occurs via an
interaction with the light chain (Summaria and Robbins, 1976). The
interaction with the group A streptococcal plasmin receptor is distinct
from group C streptokinase in that it does not recognize the Glu-
plasminogen molecule, and demonstrates no significant reactivity with
the isolated light chain of plasmin. Furthermore, the plasmin(ogen)-
streptokinase complex cannot be dissociated by lysine or lysine analogs
(Von-Mering et al., 1988), while the interaction of plasmin with a group
A streptococcus is completely reversible by lysine or lysine analogs
(Broeseker et al., 1988).
Taken together, these results indicate that the group A
streptococcal plasmin receptor binds in a unique manner to both plasmin
and Lys-plasminogen. The predominant interaction is via determinants
present on the intact heavy chain. These structures are present in their
optimal binding configuration on the intact plasmin molecule and on the
modified zymogen, Lys-plasminogen. The studies presented here suggest
that the lysine binding sites themselves are not involved in direct
interaction of plasmin with the bacteria (Figure 3-2). The observations
that plasmin bound to bacteria retains its enzymatic activity for both
small synthetic substrates and for fibrin, Chapter Two, are consistent
with the observations that the light chain is not involved in binding.

84
The failure of a^-AP to regulate the bound enzyme suggests that the
required interaction between 02-AP and plasmin is directly or indirectly
inhibited. This may occur because one of the recognition sites for <*2-AP
in the kringle 1 region of plasmin's heavy chain may not be accessible
when plasmin is bound to a streptococcus.
The characteristics of the interaction of human plasmin with the
group A streptococcus, 64/14, described in this study indicate that the
bacteria can capture a potent protease activity that cannot be regulated
by the primary physiological inhibitor of plasmin, a^-AP. This group A
streptococci also secretes a plasminogen activator and consequently, in
the presence of plasminogen, the bacteria has the potential to both
generate plasmin and bind the active enzyme to its surface (DesJardin et
al. 1988) The importance of this selective receptor to the infectious
disease process of receptor positive bacteria remains to be established.
The purpose of the series of studies described in the next chapter
was to isolate and characterize the plasmin binding receptor from the
strain 64/14 streptococcus.

CHAPTER FOUR
ISOLATION AND PURIFICATION OF A FUNCTIONALLY ACTIVE
GROUP A STREPTOCOCCAL RECEPTOR FOR HUMAN PLASMIN
Introduction
The studies presented thus far have documented the existence of a
cell surface receptor for human plasmin on group A streptococcal strain
64/14. In addition to this plasmin binding activity, certain group A
streptococci have long been known to secrete the plasmin(ogen) binding
protein streptokinase, (Mr approx. 48,000 daltons), a non-enzymatic
plasminogen activator. This protein, described by Tillet and Garner
(1933), non-covalently associates with both plasminogen and plasmin, and
was originally identified by virtue of its ability to generate
fibrinolytic activity. Streptokinase binds rapidly to the native
zymogen Glu-plasminogen (rate content 5.4 x 10^ M'^S"^-) forming, a 1:1
stoichiometric complex with an estimated dissociation constant of 5 x
10"H M (Cederholm-Williams et al.. 1979). The formation of a complex
between streptokinase and plasminogen generates an enzymatic moiety
capable of plasminogen activator activity, a property neither protein
possesses alone.
The properties of the bacterial plasmin receptor reported thus far
are markedly different from streptokinase. While the bacterial plasmin
receptor binds preferentially to domains in the heavy chain of the
plasmin molecule (see Chapter Three), streptokinase binds to plasmin's
light chain (Summaria and Robbins, 1976). Furthermore, streptokinase
85

86
binds to both plasmin and the native zymogen Glu-plasminogen, while the
surface associated plasmin receptor shows no significant reactivity for
the native zymogen. Despite these clear functional differences, the
expression of two proteins by the same bacteria that bind to the key
human fibrinolytic protein plasmin with such selectivity raises the
possibility that they may be in some way related or derived from a common
precursor. Furthermore, the majority of information on the properties of
streptokinase have been derived from studies of the plasminogen activator
molecule isolated from group C streptococcal strains and evidence for
differences in antigenicity and hence possibly function have been
reported between streptokinase proteins isolated from group A and group C
streptococcal isolates (Dillon and Wannamaker, 1965; Weinstein, 1953).
The purpose of the studies presented in this chapter were to isolate a
functionally active receptor for human plasmin from strain 64/14 and to
compare it with the streptokinase protein that is produced by the same
organism.
Materials and Methods
Enzymes. Inhibitors and other Reagents
Urokinase and porcine elastase (type IV) were obtained from Sigma
Chemical Co., St. Louis., MO. Aprotinin was obtained as Trasylol from
Mobay Pharmaceuticals, New York, NY. D-Val-Phe-Lys-chloromethyl ketone
(VPLCK), and Phe-Pro-Arg-chloromethylketone (PPACK) were obtained from
Calbiochem-Behring, San Diego, CA. Human Lys-plasminogen was obtained
from American Diagnostica Inc., Greenwich, CT. H-D-Val-Leu-Lys-
paranitroanilide (S-2251) was obtained from Helena Laboratories,
Beaumont, TX. Purified group C streptokinase was a gift from Kabivitrum,
A.B., Stockholm, Sweden.

87
Bacteria
The Lancefield /? hemolytic streptococcal strain 64/14 was grown as a
stationary culture at 37C, in one to two liter batches of a chemically
defined media for streptococci described by Van de Rijn and Kessler
(1980), containing 0.1% phenol red. The pH of the cultures were
maintained at a pH greater than 7.0, as monitored by the indicator dye.
For certain experiments, where noted, bacteria were grown in Todd-Hewitt
broth (Difco Laboratories, Detroit, MI) Approximately 2.0 to 4.0 g (wet
weight) of bacteria could be recovered per liter of media following a 24
to 36 hour incubation at 37C. Bacteria were harvested by
centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4,
containing 0.02% sodium azide. The bacteria were heat killed at 80C
for 15 min. The suspension was centrifuged and the pellet washed twice
with PBS containing 0.02% sodium azide. Aliquots could be stored at -
20C, or used immediately for extraction purposes.
Radioiodination of Proteins
Human plasminogen was iodinated by the chloramine T method using
Iodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell
(1982). The labeled proteins were separated from free iodine by passage
over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal
buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1%
gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. Labeled aliquots were used once
and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three n1 of urokinase (Sigma 20

88
u/ml) was added to a 400 /tl solution of 1 tM plasminogen containing 0.04
M lysine. The mixture was incubated at 37C for 45 minutes unless stated
otherwise. The efficiency of plasmin generation was followed by
measuring the conversion of the single chain plasminogen molecule (Mr=
90,000 daltons) into heavy chains (Mr=60,000 daltons) and light chains
(Mr=25,000 daltons) as determined by the migration of radiolabeled
proteins, following reduction, on 10% SDS-polyacrylamide gels. The
migration of labeled proteins was determined by autoradiographic
exposure of dried gels to Kodak XAR 5 film with intensifying screens at
-70C for 15-20 hours.
Direct Binding Assay of Radiolabeled Proteins
The ability of radiolabeled plasmin(ogen) to bind to the group A
streptococcus 64/14 was measured as described previously by Lottenberg et
al. (1987). A fixed number of bacteria were incubated with labeled
proteins (approximately 30,000 cpm per tube) in a total volume of 400 /tl
of VBS-gel for 30 min at 37C. The bacteria were pelleted by
centrifugation at 1000 x g for 10 min and the pellets washed twice with
2.0 ml of VBS-gel. The radioactivity associated with the bacteria was
determined in a Beckman 5500 Auto gamma counter (Beckman Instruments,
Inc., Fullerton, CA). Non-specific background binding was determined in
replicate tubes which contained no bacteria. All estimates were
performed in duplicate.
Dot-blotting Procedure for the Identification of Plasmin Receptor
Activity
This assay was carried out with a Bio-Rad bio-dot microfiltration
apparatus using a modification of the Bio-Rad procedure. A piece of
nitrocellulose pre-equilibrated in PBS-azide for a minimum of 10 minutes
was fitted into the apparatus. The wells were loaded with 100 tl

89
aliquots of PBS-azide and vacuum drained. Extraction samples,
chromatography fractions or standards were loaded into wells in 50-200 pi
aliquots. Commercially available group C streptokinase (Kabikinase) was
used as a positive control in each assay. All wells were washed twice
with 200 pi aliquots of PBS-azide and vacuum drained, all samples were
assayed in duplicate.
Blots were removed from the apparatus and remaining sites on the
nitrocellulose were blocked by washing a total of four times in 200-250
ml of 5.0 mM sodium diethylbarbiturate, 0.14 M NaCl, 0.5% gelatin, 0.15%
Tween 20, 0.004% NaN3 pH 7.35 (blotting wash buffer I) for 15 minutes
per wash. At this point, blots could be probed as described below or
stored in the fourth wash overnight at 4-8C. If the latter procedure
was followed, blots were washed a fifth time after cold storage in 200-
250 ml blotting wash buffer I for 30 minutes. Results from the two
variations did not differ.
The individual blots were then probed for 3-4 hours at room
temperature while rotating in 10 ml aliquots of the following probing
solution: blotting wash buffer I containing 2.0 mM PMSF and l^I-labeled
plasmin at 2 x 10^-3 x 10^ cpm/ml. The probed blots were then washed
four times in 200-250 ml of 0.01 M EDTA, 0.5 M NaCl 0.25% gelatin, 0.15%
Tween 20, 0.004% NaN3 for 15 minutes per wash. All washing and probing
steps were carried out at ambient temperature. The probed, washed blots
were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.

90
Polyacrylamide Gel Electrophoresis and Protein Blotting
Electrophoresis was carried out as described by Laemmli (1970).
Polyacrylamide separating gels were 10% and contained 0.1% sodium
dodecylsulfate (SDS), 0.375 M Tris at pH 8.8. Stacking gels were 4% and
contained 0.1% SDS and 0.125 M Tris at pH 6.8. Electrode buffer was
0.024 M Tris, 0.192 M glycine, 0.1% SDS at pH 8.3 Samples were diluted
1:2 with sample buffer containing 0.125 M Tris pH 6.8, 4% SDS, 20%
glycerol, 10% /3-mercaptoethanol and 0.05% bromophenol blue and heated at
80-90C for 3 minutes. Gels were run at 45 volts constant voltage for
approximately 15-18 hours. Slab gels were used in the Bio-Rad Protean II
system (BioRad, Richmond, CA). Molecular weight markers were run on all
gels. Gels intended for Western blot transfer contained pre-stained
markers (Sigma) applied as a mixture which included: triosephosphate
isomerase (26,600), lactic dehydrogenase (36,500), fumarase (48,500),
pyruvate kinase (58,000), fructase-6-phosphate kinase (84,000), fi-
galactosidase (116,000), and c*2-macroglobulin (180,000). After
electrophoresis, gels intended for Western blotting were equilibrated in
25 mM Tris, 0.2 M glycine pH 8.0 containing 20% v/v methanol (electroblot
buffer) for 25 minutes. Protein blotting, from SDS-PAGE gels, was
performed using the 'Trans-Blot SD Semi-Dry' electrophoretic transfer
cell (Bio Rad, Richmond, CA). Nitrocellulose transfer medium, also
equilibrated in electroblot buffer, was sandwiched between the gel and
two sheets of Whatman 3 mm paper. The gel was also backed with two
sheets of 3 mm paper. For probing with plasmin, the Western blots were
washed, probed and autoradiographed according to the procedure described
above for dot-blotting. For probing with rabbit, anti-plasmin receptor
antibody, blots were probed with blotting wash buffer I containing 4.3 fig

91
IgG per ml of probing solution (approximately a 1:3000 dilution of
antisera) for three hours, washed twice for 20 min with 300 mis of
blotting wash buffer I, and probed with ^^^I-Protein G at 2 x 10^-3 x 10^
cpm/ml, the probed blots were then washed four times in 200-300 mis of
0.01 M EDTA, 1.0 M NaCl 0.25% gelatin, 0.15% Tween 20 for 15 minutes per
wash. All washing and probing steps were carried out at ambient
temperature. The probed, washed blots were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.
Molecular weight determinations on Western blots were made possible
by the transfer of the prestained molecular weight markers.
Gels to be used for protein identification were either stained with
silver according to the procedure described by Merril et al., (1981), or
with Coomassie brilliant blue R-250 as follows: (0.25% w/v in 40%
ethanol and 10% acetic acid) for 1 hour, and destained by soaking in
several changes of 10% ethanol and 10% acetic acid containing a small
quantity of DE 52 (Whatman, England) as a dye adsorbent.
Lancefield Hot Acid/Hot Alkaline Extractions
These extractions were carried out as described by Lancefield
(1928). Approximately 10 ml aliquots of 10% (wet weight/volume) 64/14 in
PBS-azide were adjusted to pH 2 or 10 using 0.5% M HC1 or 0.5 M NaOH
respectively. The suspensions were boiled for 10 minutes and the pH was
neutralized in each sample using either 0.5 M NaOH or 0.5 M HC1. PMSF
was added to a concentration of 2.0 mM. The samples were centrifuged at
approximately 10,000 x g for 10 minutes. The supernatants were passed
through a 0.22 /m filter to remove any residual cells. The cell-free

92
supernatants were then dialyzed at 4C into 20 mM Tris-HCl, 0.15 M NaCl
pH 7.4 containing 1.0 mM iodoacetic acid, 1.0 mM benzamidine HCl.and were
stored at -70C.
Time Course Trypsin Digestion
The bacterial pellet from approximately 11 ml of 10% (w/v) 64/14 in
PBS-azide was collected by centrifugation at 10,000 x g for 10 minutes.
The pellet was washed with 10 ml of 0.05 M KH2PO4, 0.005 M EDTA, 0.02%
NaN3 pH 6.1, centrifuged as before and resuspended to 10% (w/v) in that
buffer. These salt and buffer conditions are not optimal for trypsin
activity and facilitate the extraction of surface proteins without
concomitant proteolysis of the solubilized material. Pancreatic DNAse I
(Sigma) was added to approximately 6.0 ml of this suspension to a final
concentration of 4 g/ml. The sample was vortexed and warmed to 37C.
Bovine pancreatic trypsin (Type I, Sigma) was then added to a final
concentration of 20 fMg/ml and the sample was mixed. A 1.0 ml aliquot was
immediately removed and mixed with a concentrated solution of benzamidine
HC1. The final concentration of benzamidine HC1 in the reaction mixture
was 100 mM, well in excess of what was required to completely inhibit
the activity of the trypsin present in the reaction mixture. This
sample, was mixed and placed on ice, and was designated the zero time of
the experiment. At 5, 10, 30 and 60 minutes 1.0 ml aliquots were removed
from the reaction mixture, and were treated in an identical manner. A
control digestion was prepared by incubating a 1.0 ml aliquot of 10%
bacterial suspension containing 4 /g/ml DNAse I at 37C for 60 minutes
followed by the addition of benzamidine HCl to 100 mM final
concentration. All samples were centrifuged at approximately 10,000 x g
for 10 minutes. Supernatants were collected and stored at -70C. Prior

93
to testing each control and experimental supernatant was centrifuged for
5 minutes at 10,000 x g to remove any particulate material. A control
sample treated in an identical fashion with the exception that no trypsin
was added was included at the 60 minute incubation time to determine the
degree of non-specific release of proteins from the bacteria. In
addition a control from which bacteria were omitted was included in each
assay.
Non-ionic Detergent/Osmotic Shock/Lvsozvme Extraction
This extraction was a modification of the procedure described by
Scopes (1982). Approximately 1.0 g wet weight of 64/14 was combined with
2.5 ml of glycerol and 0.1 ml of 10% (v/v) Triton X-100 in 20 mM KH2PO4,
1.0 mM EDTA 0.02% NaN3 pH 7.6. The cells were dispersed by vortexing and
the sample was placed at 37C for 30 minutes. The suspension was
vortexed a number of times while incubating. Following incubation, the
mixture was adjusted to 20 ml with lysozyme buffer. Lysozyme (Sigma) was
added to 200 ig/ml and DNAse I to 10 /g/ml. The sample was vortexed and
returned to 37C for 30 minutes with frequent vortexing. After
incubation, PMSF was added to 2.0 mM final concentration. The
supernatant was collected and treated as described for the Lancefield
extractions. As enzyme control to be tested with the extract in the
screening assay for plasmin receptor activity contained 200 pg/ml
lysozyme and 10 /g/ml DNAse I in 20 mM Tris-HCl pH 7.4, 0.15 M NaCl, 1.0
mM iodoacetic acid, 1.0 mM benzamidine HC1 and .02% NaN3.
Acetone/Detergent Extraction
This procedure was a modification of the extraction described by
Bhaduri et al.. (1983). approximately 1.0 g wet weight of 64/14 was
suspended in 10 ml of ice cold acetone (Fisher Certified A.C.S. grade),

94
allowed to stand on ice for 5 minutes and then collected by
centrifugation at 10,000 x g for 10 minutes. Residual acetone was
evaporated under a stream of air. The pellet was resuspended with
vortexing in 25 ml of 1.0% (v/v) Triton X-100 in PBS-azide and incubated
at room temperature for 5 minutes. An additional 2.5 ml of PBS-azide was
added. The supernatants were collected and treated as described for the
Lancefield extracts.
Mutanolvsin Extraction
This procedure is a modification of the method described by Yarnall
et al.. (1986). approximately 0.9 g wet weight of 64/14 was suspended in
5.0 ml of 20 mM KH2PO4, 1.0 mM EDTA, .02% NaN3 pH 7.0 containing 2.0 mM
PMSF, 10 /tg/ml DNAse I and 50 /ig/ml mutanolysin. The mutanolysin was
purified from a commercial product (Sigma) according to the method
described by Siegal et al., (1981). The suspension was vortexed and
placed at 37C for 4 hours with periodic vortexing. Supernatants were
collected and treated as described above for Lancefield extractions. An
enzyme control for use in the plasmin receptor assay contained 10 /ig/ml
DNase I, 2.0 mM PMSF, and 50 //g/ml mutanolysin in 20 mM Tris-HCl, 0.15 M
NaCl, 1.0 mM iodoacetic acid, 1.0 mM benzamidine HC1, 0.02% NaN3 pH 7.4.
Preparation of Immobilized Human Plasmin Affinity Column
Human plasminogen at a concentration of approximately 5.2 x 10'^ M
was activated to plasmin by incubating the sample in the presence of an
approximately 62 fold lower molar concentration of urokinase (Abbott).
The reaction volume was 10 ml and the primary buffer was 0.05 M Tris,
0.15 M NaCl pH 7.4 containing 40 mM lysine. Conversion was carried out
with constant agitation for one hour at 37C. A 50 /I aliquot was
removed and the remainder flash frozen and stored at -70C. The sample

95
taken was analyzed by SDS-PAGE under reduced conditions for the
conversion of the single chain plasminogen molecule to the two chain
plasmin form. Once it was established that the plasminogen was fully
activated, the bulk preparation was reacted with a 5 fold molar excess of
D-valyl-L-phenylalanyl-L-lysine chloromethyl ketone (VPLCK) (Calbiochem),
an irreversible inhibitor of the enzyme activity of plasmin. This enzyme
inactivation was carried out at ambient temperature with constant
rotation. The enzymatically inactive plasmin was then concentrated by
ammonium sulfate precipitation (4.0 g / 10 ml), and dialyzed at 4C
against 0.1 M MOPS buffer, pH 7.3, containing 0.02% sodium azide. The
dialyzed inactive plasmin was then chromatographed on Superse 6
(Pharmacia) in 0.1 M MOPS buffer, pH 7.3.
The activated affinity chromatography support Affi-Prep 10 (Bio Rad)
was selected as the matrix for immobilizing the chlormethyl ketone
blocked plasmin. This matrix couples in aqueous buffers by means of an
N-hydroxysuccinimide ester on the end of a 10 carbon space arm to primary
amino groups in the ligand. The ligand is linked by amide bonds to the
terminal carboxyl groups of the Affi-Prep 10 spacer arm. The buffer used
in the coupling reaction was 0.1 M MOPS buffer, pH 7.3. Approximately 50
mg of inactivated plasmin in 18 ml of coupling buffer was incubated with
6.0 ml of washed Affi-Prep 10. The reaction was carried out at 4C for
15 hours with rotation. Following ligand coupling, 100 /xl of 1.0 M
ethanolamine HC1 pH 8 was added to the reaction mixture to block
remaining active sites. This blocking reaction was completed in 1 hour
at 4C with sample rotation. The matrix was washed with two 1.0 ml
aliquots of 1.5 M NaCl. The supernatant from the coupling reaction and
the two 1.5 M NaCl wash volumes combined and dialyzed against PBS at 4C.

96
In order to determine the extent of coupling, the plasmin content of the
dialyzed sample (determined by means of absorbance at 280 nm using an
A?cmnm value of 17.0) was compared with the known starting plasmin
concentration. The efficiency of coupling was estimated to be 90%.
The Affi-Prep 10 Plasmin was loaded into an HR 10/10 FPLC
compatible column (Pharmacia). The affinity matrix was equilibrated in
0.05 M Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, 0.02% NaNj pH 7.2.
When not in use, the column was stored at 4C.
Affinity Purification of Plasmin Receptor
The Affi-Prep 10-Plasmin HR 10/10 column was attached to a Pharmacia
FPLC chromatography system and equilibrated at room temperature in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, 0.02% sodium azide pH 7.2.
1.0-2.0 ml of crude supernatant from the mutanolysin extraction of
bacterial strain of 64/14 (prepared as described above) was loaded onto
the column at a flow rate of 0.02 ml/min. The flow rate was increased to
a 1.0 ml/min rate during the washing step using equilibration buffer
(approx. 200 mis). The column was either eluted at 0.2 ml/min with a 50
ml linear gradient of 0.0 M 0.1 M L-Lysine in equilibration buffer, or
eluted in a single step using equilibration buffer containing 0.1 M L-
Lysine. The absorbance at 280 nm was continuously monitored and 1.0 ml
fractions were collected. After each affinity purification procedure the
column was washed with 20 mis of 2.0 M NaCl, followed by 200 mis of
equilibration buffer and stored at 4C.
Plasminogen Activation Assay for Streptokinase
The following assay (Zolton and Mertz, 1972; Teger-Nilsson et al..
1977) to measure SK activity was carried out in microtiter plates. 20 pi
aliquots of streptokinase standards (a dilution series for the purpose of

97
generating a standard curve), or the samples to be tested were placed
into the microtiter wells in duplicate. To the 20 /il aliquots, 40 /il of
50 mM Tris, pH 7.5, was added. 30 /I of a freshly prepared solution of
human Glu-plasminogen, 20 /tg/ml in 0.01 mM Triton X-100, was then added
to the well and allowed to incubate at 37C for 15 min. 30 /I of
substrate was then added. Substrate is prepared as follows: To 1 volume
substrate (5 mg/ml S-2251 in water) add 3 volumes 1.77 M NaCl in 0.32 M
Tris, pH 7.5, and one volume of water. The plate was then incubated at
37C to allow substrate hydrolysis, and product production is measured at
405 nm. Control wells from which either SK and/or plasminogen were
omitted were included. These controls will indicate whether there is any
plasmin contamination in the plasminogen preparation or if the sample
being tested has any proteolytic activity for the substrate that is not
dependent on plasminogen activation.
Solid Phase Assay for Plasminogen Activators
Samples to be tested for plasminogen activator activity by this
assay (Dr. K. Johnston, personal communication) were first resolved by
SDS-PAGE and transferred to nitrocellulose. The nitrocellulose membranes
were then immersed in blocking buffer (10 mM Tris, pH 8.0 containing 0.5%
Tween-20, 0.5 M NaCl and 1.0% bovine serum albumin) for at least one hour
at room temperature. The substrate overlay is prepared as follows: To a
2.0% agarose solution (Bio Rad Richmond, CA) in 0.15 M phosphate buffered
saline, pH 7.5 was equilibrated at 50C, with the chromogenic substrate
S-2251 at a concentration of 100 /ig/ml. Human plasminogen free of
plasmin activity, was then added to a final concentration of 20 /ig/ml.
The agarose-substrate-plasminogen solution was then applied to an ethanol
washed glass slide slightly larger than the nitrocellulose membrane

98
template. The nitrocellulose membrane containing the sample was then
drained of excess blocking buffer and overlayed on the agarose-
plasminogen- substrate gel. It is important to ensure uniform contact
between the nitrocellulose membrane and the gel. The nitrocellulose
membrane was allowed to remain in contact with the agarose for at least
one hour at 37C; the time of incubation is dependent upon the
concentration of plasminogen activator present. At the termination of
incubation, the nitrocellulose membrane was removed from the agarose-
plasminogen- substrate gel and immersed for 5 min in freshly prepared
0.1% sodium nitrite dissolved in 1.0 N HC1 followed immediately by
immersion for 5 min in 0.5% ammonium sulfamate dissolved in 1.0 N HC1.
The membrane was then transferred to a solution containing 0.05% N-l-
napthylethylenediamine in 47.5% ethanol and observed for the appearance
of red bands indicative of plasmin activity. Plasminogen activators
present on the nitrocellulose membrane will activate the plasminogen
present in the agarose which will in turn cleave the S-2251 substrate
incorporated in the agarose. The chromogenic cleavage product
(paranitroaniline) appears yellow and deposits on the nitrocellulose
membrane. The chemical treatment of the membranes as described above
will convert the deposited yellow paranitroaniline to a red color and fix
it to the membrane. Membranes can be stored under water at 4C.
Preparation of Polyclonal Rabbit Anti-plasmin Receptor Protein Antibody
Mutanolysin extracted 41,000 dalton plasmin receptor protein was
purified by gel electrophoreses on a 10% SDS-PAGE gel and stained with
Coomassie brilliant blue R-250. The single stained band was cut from the
gel and equilibrated in PBS-azide. The location of the 41,000 dalton
plasmin binding band was determined by the position of the stained band

99
in the gel, and by Western blotting a small strip of the gel to
nitrocellulose, followed by blocking and probing with ^^1-plasmin, and
autoradiography as described in the polyacrylamide gel electrophoreses
and protein blotting section of the Methods. The stained 41,000 dalton
band was cut from the gel, and a portion containing approximately 300 ig
was emulsified with an equal volume of Freund's complete adjuvant. The
emulsion was injected subcutaneously at 6 sites on a rabbit. The rabbit
was boosted eight times with the 41,000 dalton protein-polyacrylamide gel
emulsified in Freund's incomplete adjuvant (approximately 200 /g per
boost) during a 14 month period. Pre-immune and immune IgG fractions
were prepared from rabbit sera by Protein A-Sepharose (Sigma) affinity
chromatography.
Results
A variety of extraction procedures were compared and were to
determine the optimal method for solubilizing functional plasmin receptor
activity. The extractions were carried out on heat killed 64/14 as
described in detail in the Methods section. The bacterial samples were
washed thoroughly prior to treatment to minimize carry over of culture
media and secreted products to the extraction samples. This would
therefore reduce the likelihood of significant streptokinase
contamination (with the possible exception of intracellular forms).
The extraction techniques included: (1) Lancefield acid and
alkaline extractions; (2) a time course trypsin digestion under
suboptimal conditions for enzyme activity (conditions previously shown to
maximize homogeneity of type III Fc receptor extraction, Reis et al.,
1985); (3) Triton X-100/osmotic shock/lysozyme treatment; (4)

100
acetone/Triton X-100 extractions; and (5) mutanolysin digestion. The
IOC
supernatants from these solubilizations were screened for iZ-JI-plasmin
binding by the dot-botting procedure in Methods. Aliquots of cell free
supernatants prepared from the extractions were applied to nitro
cellulose membrane in a dot-blot apparatus. Blocking and washing of the
nitrocellulose membranes was carried out according to Methods. The
nitrocellulose membranes were then probed with ^^^1-plasmin, washed and
autoradiographed.
The results of the screening of the various extracts and
preparations are shown in Figure 4-1. Extraction with mutanolysin
demonstrated the highest yield of soluble plasmin binding activity (see
Figure 4-1, row B, column 3).
The size heterogeneity of the soluble plasmin receptor activity in
the mutanolysin extract of strain 64/14 was assessed by electrophoresis
of a 50 /il aliquot of the extract on both reducing and non-reducing SDS-
polyacrylamide gels which were then stained with silver or electroblotted
onto nitrocellulose and probed with -*-^I-plasmin. The protein staining
pattern of the mutanolysin extract is shown in lane 2 of Figure 4-2.
Plasmin binding activity was concentrated predominately in a band with
an Mr of approximately 41,000 daltons (see Figure 4-2, Panel B). One /g
of purified group C streptokinase (Mr approx. 48,000 daltons) was
electrophoresed as a positive control, (Figure 4.2, lane 1, Panel A;
lane 2, Panel B). An aliquot of the control mutanolysin digestion
mixture and containing all the reactants except the bacteria was analyzed
by SDS-PAGE reveled no plasmin binding activity by Western blotting nor
any significant stainable bands (data not shown).
The possible release of the secreted plasminogen activator
(streptokinase), from the strain 64/14, during the extraction, would be a

Figure 4-1. Dot-blot analysis of solubilized plasmin binding activities.
Samples to be tested for plasmin binding activity were immobilized on
nitrocellulose in a Bio-dot blot apparatus. The nitrocellulose membrane
was blocked as described in the Methods and probed with -plasmin.
The blot was washed and autoradiographed at -70C for 10 hours using
Kodak X-AR-5 film and an intensifying screen. Row A: columns 2 through
5: Kabikinase 400, 300, 200, and 100 ng respectivly; column 6: trypsin
extraction (60 min); column 7: trypsin enzyme control. Row B: column 1:
lysozyme/detergent/shock extract; column 2: lysozyme/detergent/shock
control; column 3: mutanolysin extract; column 4: mutanolysin enzyme
control; column 5: hot-alkali extract; column 6: hot-acid extract;
column 7: acetone/detergent extract.

102
I 2 3 4 5 6 7

Figure 4-2. SDS-PAGE and Western blot analysis of mutanolvsin extracted
64/14 bacterial plasmin binding activity. Parallel 10% SDS-poly-
acrylamide gels were electrophoresed. One gel was silver stained to
detect protein molecules (Panel A) and the second was Western blotted and
probed with -plasmin, as described in the Methods, and auto-
radiographed for 10 hours at -70C with intensifying screens (Panel B).
Panel A, lane 1: 1.0 /g of group C streptokinase (Kabikinase); lane 2: 50
Hi of strain 64/14 mutanolysin extract. Panel B, lane 1: 50 tl of strain
64/14 mutanolysin extract; lane 2: 1.0 /jg of group C streptokinase
(Kabikinase).

104

105
possible confusing factor in the isolation and characterization of the
surface plasmin receptor. The production of streptokinase by
streptococci is optimal when the pH of the growth medium is maintained
at 7.0 to 8.0 (Johnston and Zabriskie, 1986). The maintenance of a pH
above 6.8 prevents the activation of an extracellular zymogen to an
active mercaptoproteinase (Elliott and Dole, 1947; Liu and Elliott, 1965)
produced by the bacteria, which would significantly contribute to the
proteolytic hydrolysis of secreted streptokinase. Therefore, solubilized
plasmin receptor activity was prepared by mutanolysin extraction as
before, but from bacteria harvested from chemically defined media
cultures in which the pH was not allowed to become acidic. In addition,
the supernatants from these cultures were collected, filtered and
concentrated as a source of streptokinase from the strain 64/14 bacteria
for comparative analyses.
The production of plasmin receptor from chemically defined media, pH
controlled, cultures was first investigated. Receptor activity was
expressed, mutanolysin extracted, and also affinity purified.
Furthermore, there was no change in the molecular weight of the plasmin
receptor.
The production of streptokinase from the strain 64/14 bacteria was
measured functionally, by use of a quantitative plasminogen activation
assay. The commercially available highly purified group C streptokinase
(Kabikinase) (Kabivitrum, A.B., Stockholm, Sweden.) was used as a
standard in these studies.
The fluid phase assay for plasminogen activator activity (see
Methods for precise experimental details) was also used to measure
plasminogen activator activity in the mutanolysin extraction preparations

106
as well as the samples of strain 64/14 concentrated supernatant. The
mutanolysin extracts of strain 64/14 were totally devoid of plasminogen
activator activity (data not shown). The strain 64/14 concentrated
supernatant was found to contain approximately 3,555 72 units of SK
activity per ml, or approximately 37 /g/ml of streptokinase, based on the
kabikinase standard. There was a total of 15 mis of concentrated
supernatant from 1 liter of this strain 64/14 bacterial culture. The
possibility that treatment of streptokinase with mutanolysin may destroy
plasminogen activator activity was considered, and under the conditions
of mutanolysin treatment used for extraction no loss of plasminogen
activator activity of mutanolysin treated streptokinase was observed
(data not shown).
A second assay to investigate plasminogen activator activity was
also used. This is a semi-quantitative, solid-phase assay designed to
correlate plasminogen activator activity to molecular weight of the
activator present in a sample (Dr. K. Johnston, personal communication).
In this assay the samples to be analyzed are first electrophoresed by
SDS-PAGE, separating individual proteins, and are then electroblotted
onto nitrocellulose membrane (see Methods). The nitrocellulose membrane
is then blocked with BSA, and applied to an agarose film containing
plasminogen and the chromogenic substrate S-2251. The plasminogen
activator present on the membrane will activate plasminogen in the
agarose which will in turn hydrolyze the S-2251. After allowing a period
of activation (time determined by the amount of plasminogen activator
present in the sample) the paranitroaniline product that passively
adheres to the membrane is chemically fixed to membrane. This fixation
procedure results in the production of a pink band(s) indicative of the

107
presence of plasminogen activator. Samples of the mutanolysin extracted
plasmin binding activity preparation, concentrated supernatant from the
strain 64/14 pH controlled CDM cultures, and purified streptokinase
(Kabikinase) as a positive control were all analyzed with this assay
system. The results are shown in Figure 4-3. There were no bands of
activator activity present in the mutanolysin extracts from the 64/14
bacteria. However, there was a streptokinase activity in the culture
supernatant of this strain. This secreted plasminogen activator activity
co-migrated at the same molecular weight (approx. 48,000 daltons) as the
purified sample of group C streptokinase obtained from Kabivitrum.
Because of the high level of plasmin receptor activity in the
extractions prepared by mutanolysin digestion, this preparation was
chosen as material for further analysis and purification of the strain
64/14 plasmin receptor from bacteria grown in chemically defined media.
Due to the simplicity and specificity offered by affinity chromatography,
the plasmin receptor activity was purified from the mutanolysin extracts
using a plasmin affinity matrix prepared as described in the Methods.
Briefly, approximately 50 mg of purified human plasminogen was activated
to plasmin by incubation in the presence of the plasminogen activator
urokinase. The plasmin was then enzymatically inactivated and coupled to
6.0 mis of the affinity chromatography support Affi-Prep 10 (Bio Rad).
Following ligand coupling, the remaining active sites on the matrix were
blocked with 1.0 M ethanolamine HC1 pH 8.0. The matrix was then washed
with 2.0 M NaCl. The Affi-Prep 10 inactivated plasmin was loaded
intoan HR 10/10 FPLC compatible column (Pharmacia). The affinity matrix
was again washed with 2.0 M NaCl, and then equilibrated in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaNj pH 7.4.

Figure 4-3. Solid-phase plasminogen activation assay. The plasminogen
activator activities of mutanolysin extracted and secreted proteins were
monitored using the solid phase plasminogen activator assay (Johnston,
personal communication). Lane 1: 50 /xl of mutanolysin extracted 64/14
plasmin binding material (approx. 5-10 fig of 41,000 dalton band); Lane2:
60 fil of strain 64/14 concentrated supernatant (approx. 2.0 fig of
streptokinase); Lane 3: 2.0 fig of group C streptokinase (Kabikinase).
For precise experimental details see Methods.

109
KD
I I 6
84
58
48.5
36.
26.6

110
The Affi-Prep 10-plasmin HR 10/10 column was then either attached to
a Pharmacia FPLC chromatography system and used at room temperature, or
used at 4C with a peristaltic pump and fraction collector (Pharmacia).
The mutanolysin extract was dialyzed into 0.05 M Na2HP04, 0.15 M NaCl,
1.0 mM benzamidine HC1, and 0.02% NaN3 pH 7.4. prior to chromatography.
1.0-2.0 mis of cell free mutanolysin extract of strain 64/14 was
routinely applied to the blocked plasmin affinity coliman matrix in 0.05 M
Na2HP04, 0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaN3 pH 7.4.
After loading the extract onto the column, the matrix was washed with
this buffer until the OD 280 nm returned to base line absorbance
(approximately 20 mis). Bound plasmin receptor activity was eluted with
either a 50 ml linear gradient of 0.0 0.1 M L-Lysine in 0.05 M Na2HP04,
0.15 M NaCl, 1.0 mM benzamidine HC1, and 0.02% NaNj pH 7.4, or in a
single step using buffer containing 0.1 M L-lysine. The absorbance at 280
nm was continuously monitored and 1.0 ml fractions were collected.
Fractions eluted from the affinity column were assayed by the dot
blotting and ^^1-plasmin probing method. The plasmin binding functional
activity eluted from the column correlated to absorbance at 280 nm (see
Figure 4-4). This material was then analyzed by SDS-PAGE followed by
either silver or Coomassie brilliant blue R-250 staining, and Western
blotting and probing with iodinated plasmin. The results shown in Figure
4-5 indicate silver stained band at 41,000 daltons eluted from the
affinity column with lysine. This band corresponds to the plasmin
binding activity seen in the crude extract. The plasmin affinity
purified material was then analyzed by Western blotting and probing with
125
I-plasmin. Figure 4-6, Panel A, shows another sample of plasmin
affinity purified receptor, eluted using 0.1 M lysine in a single step.

Figure 4-4. Representative profile of an affinity purification of strain
64/14 mutanolvsin extracted nlasmin binding activity. One ml of a
mutanolysin extract of strain 64/14 was applied to a column of
immobilized enzymatically inactive plasmin. Bound material was eluted
with a gradient of L-lysine. All fractions were screened for plasmin
binding activity by dot-blot analysis. The resulting autoradiograph of
dot-blotted fractions (50 n 1) probed with ^-^^1-plasmin is shown for the
coresponding fractions below the X-axis.

A280
112
FRACTION NUMBER
JJlL; -i-
' -#'..!*
; bi:j i'i'I' ?
60
50
40
30
20
10
i
L-LYSINE (mM)

Figure 4-5. Analysis of affinity purified plasmin binding material from
the strain 64/14 mutanolvsin extract. 50 il aliqouts of affinity
purified fractions containing plasmin binding activity were electro-
phoresed on a 10% SDS-PAGE gel and stained with silver. The functional
activity of each sample was monitored by dot-blot analysis and probing
with 125i-plasmin, and is shown below the coresponding lane on the SDS-
polyacrylamide gel.

114
1 2 3456789

115
The activity eluted in three 1.0 ml fractions (Lanes 3-5 of Figure 4-6
Panel A). The greatest activity was found in fraction number two,
corresponding to lane four of Figure 4-6, Panel A, and 50 /I of this
fraction was analyzed by Western blotting. Figure 4-6, Panel B, is an
autoradiograph of Western blotted extracted plasmin receptor preparation
(Lane 1) and plasmin affinity purified receptor (Lane 3), which
demonstrates that the 41,000 dalton molecule has retained functional
plasmin binding activity following affinity purification. Treatment of
the affinity purified material with trypsin destroys the ability of the
41,000 dalton molecule to bind plasmin and results in the disappearance
of the 41,000 dalton stained band on SDS-polyacrylamide gel. These
results clearly indicate the purification of a plasmin binding activity
from a mutanolysin extract of the streptococcal strain 64/14 bacteria by
means of affinity chromatography. Taken together these results indicate
that the extracted surface receptor for human plasmin, and streptokinase
produced by the strain 64/14 streptococcus, are physicochemically
(molecular weight) and functionally (plasminogen activator activity)
distinct molecules.
The isolated 41,000 dalton plasmin receptor protein was used to
immunize a rabbit as described in the Methods. The resulting antibody
was used to probe both the mutanolysin extract of strain 64/14,
concentrated culture supernatant of strain 64/14, and group C
streptokinase (Kabikinase) using a sandwich Western blot, and the results
are shown in Figure 4-7.
Specific antigen-antibody complexes on the nitrocellulose were
detected by probing with *^1-Protein G. This antibody recognized only
the 41,000 dalton band in the mutanolysin extract and a corresponding

Figure 4-6. SDS-PAGE and Western blot analysis of mutanolvsin extracted,
affinity purified plasmin binding activity. Parallel 10% SDS-poly-
acrylamide gels were electrophoresed. One gel was silver stained to
detect the distribution of proteins present (Panel A). The proteins on
the second gel were transfered to nitrocellulose by Western blotting and
probed with -plasmin according to Methods. Panel B shows the results
of the autoradiograph demonstrating functional activity. Panel A, lane
M: molecular weight markers; lane 1: 50 /I of strain 64/14 mutanolysin
extract; lane 2-4: three fractions containing lysine eluted plasmin
binding activity. Panel B, lane 1: 50 pi of strain 64/14 mutanolysin
extract; lane 3: 50 ^1 of the lysine eluted plasmin binding activity from
the fraction shown in lane 3 of Panel A.

117
M | 2 3 4
KD
I 2 3
A

Figure 4-7. SDS-PAGE and Western blot analysis of plasmin receptor
protein with a polyclonal rabbit antibody. The isolated plasmin binding
protein, streptokinase from the same group A strain, and group C
streptokinase were compared for reactivity with a polyclonal rabbit
antibody to the purified plasmin receptor molecule. Lane 1: 2.0 /jg of
group C streptokinase (Kabikinase); lane 2: 50 tl of extracted plasmin
receptor preparation (approx. 5 fig of the 41,000 dalton molecule); lane
3: 60 il of strain 64/14 concentrated supernatant (approx. 2.0 ng of
64/14 streptokinase). The proteins on the nitrocellulose blot were
probed in a sandwich assay first with the polyclonal anti-plasmin
receptor antibody followed by ^^I-Protein q as described in the Methods.
The resulting blot was autoradiographed at -70C for 10 hours with
intensifying screens.

119
I 2 3
KD
I I 6-
84 *
5 8- Hj
48.5
36.5-
26.6-

120
protein in the concentrated culture supernatant of strain 64/14. Neither
the 48,000 dalton plasminogen activator protein present in the
concentrated culture supernatant of strain 64/14 (Figure 4-3), nor group
C streptokinase was recognized by this antibody (Figure 4-7).
Discussion
Group A streptococci have been recognized for many years to secrete
a protein, streptokinase, with a high affinity for both plasminogen and
plasmin (Tillett and Garner, 1933). The experiments documented in
Chapters Two and Three, have described a surface receptor on certain
group A streptococci that displays selective binding activity towards
plasmin, while having minimal reactivity with the zymogen form of the
molecule, Glu-plasminogen. The purpose of the experiments described in
this chapter were to isolate the plasmin receptor and compare it on a
functional basis to the secreted streptokinase protein produced by the
same group A streptococcal strain, 64/14.
A variety of different extraction techniques were compared and
treatment with mutanolysin yielded the highest quantity of soluble
plasmin binding activity. This activity was associated with a 41,000
dalton molecule by Western blot analysis, under both reducing and non
reducing conditions, and no evidence for subunit structure by
intramolecular disulfide bonds was observed. This plasmin binding
molecule was protein in nature, and was totally devoid of plasminogen
activator activity. The 41,000 dalton plasmin receptor protein was
purified from the mutanolysin extract of strain 64/14 by affinity
chromatography using enzymatically inactivated immobilized human plasmin.
Bound receptor activity on the column was specifically eluted with

121
L-lysine or EACA at the concentrations which reversibly inhibit plasmin
binding to this streptococci (Chapter Two). The affinity purified 41,000
dalton protein was demonstrated to retain functional activity by Western
1 9 S
blot analysis and probing with plasmin.
The secreted protein, streptokinase, identified in the concentrated
culture supernatant of strain 64/14 (Mr approximately 48,000) has the
ability to bind plasminogen and once complexed to the zymogen, can act as
a plasminogen activator converting plasminogen to plasmin. By contrast
the cell bound plasmin receptor lacks plasminogen activator activity and
demonstrates binding specificity towards plasmin rather than the native
zymogen, Glu-plasminogen. The secretion of streptokinase into an
environment containing plasminogen would result in plasminogen activation
and the generation of plasmin. Plasmin generated by this reaction could
then bind specifically to the receptor on the surface of the bacteria.
In experiments documented in Chapter Two, it was recognized that once
plasmin bound to the bacterial surface receptor it retained enzymatic
activity. Furthermore, this cell bound plasmin activity could not be
regulated by plasmin's normal physiological inhibitor a^-antiplasmin.
The ability of bacteria, not only to produce plasminogen activators,
but to associate the enzymatically active product on their cell surface
in an physiologically nonregulatable form may prove to be an important
factor in the ability of these pathogens to invade human tissue. The
expression of two functionally distinct streptococcal proteins with
affinity for proteins of the human fibrinolytic system is intriguing. In
the next chapter, the relationship between the plasmin receptor and
streptokinase protein produced by strain 64/14 is compared
physicochemically, functionally, and antigenically.

CHAPTER FIVE
COMPARISON OF THE GROUP A STREPTOCOCCAL RECEPTOR FOR
HUMAN PLASMIN WITH STREPTOKINASE
Introduction
In the previous chapter I have described the isolation of a specific
receptor for human plasmin from a group A streptococcus. This was
achieved by first solubilizing the receptor by treatment with the enzyme
mutanolysin followed by affinity purification on a column of immobilized
plasmin. The purified functionally active receptor material had a Mr of
approximately 41,000 daltons as measured by SDS-polyacrylamide gel
electrophoresis and lacked plasminogen activator activity. These findings
indicate that the plasmin receptor protein was not an intact
streptokinase molecule. The purpose of the studies presented in this
chapter were to perform a more complete comparison of the group A
streptococcal receptor for human plasmin and streptokinase with respect
to plasminogen activator activity; their binding specificities for
domains of the plasmin molecule; and examine possible antigenic
relatedness.
The results presented in this chapter demonstrate that the plasmin
receptor and streptokinase, while both produced by the same strain of
group A streptococci bacteria and having high affinity for plasmin, are
physicochemically, functionally, and antigenically distinct molecules.
122

123
Materials and Methods
Materials
Nitrocellulose was purchased from BioRad, Richmond CA. Affinity
purified goat anti-mouse IgG, heavy and light chains specific, was
purchased from Cappel, Organon Teknika Corp. West Chester, PA. Ribi
Adjuvant System (RAS) was purchased from Ribi Immunochem Research Inc.
Hamilton, MT. Mouse anti-group C streptokinase monoclonal antibodies
were a gift from Dr. K. Johnston. All other chemicals and reagents were
purchased from Sigma Chemical Co. St. Louis, MO.
Bacterial Strains
The Lancefield group A /3 hemolytic streptococcal strains 64/14,
B923, and A995; and the Lancefield group C /3 hemolytic streptococcal
strain 26RP66 and the ATCC strain 12449 were grown as stationary cultures
at 37C for 24 to 36 hrs, in one to two liter batches of a chemically
defined media described by Van De Rijn and Kessler (1980), containing
0.1% phenol red. The pH of the culture was maintained above 7.0, as
monitored by the indicator dye. Approximately 3.0 to 5.0 g (wet weight)
of bacteria could be recovered per liter of media. Bacteria were
harvested by centrifugation, resuspended in phosphate-buffered saline
(PBS), pH 7.4, containing 0.02% sodium azide. The bacteria were heat
killed at 80C for 15 min. The suspension was centrifuged and the pellet
washed three times with PBS containing 0.02% sodium azide. Aliquots
could be stored at -20C, or used immediately for extraction purposes.
Bacterial Culture Supernatants
Culture supernatants from the various strains of bacteria were
recovered following removal of the bacteria by centrifugation. The
supernatants were filtered through a 0.22 /m filter and immediately

124
concentrated using an Amicon concentrator (Amicon Corp., Danvers, MA)
fitted with a YM-10 membrane at 4C. Supernatants were concentrated
approximately 100 fold, and aliquoted in 0.5 ml fractions flash frozen
and stored at -70C.
Radioiodination of Proteins
Human plasminogen, isolated plasmin heavy chain, and isolated
plasmin light chain were iodinated by the chloramine T method using
Iodobeads (Pierce Chem. Co., Rockford, IL) as described by Markwell
(1982). The labeled proteins were separated from free iodine by passage
over a G25 column (PD-10 Pharmacia) and collected in 0.15 M Veronal
buffered saline pH 7.35 containing 0.001 M Mg++, 0.00015 M Ca++ and 0.1%
gelatin (VBS-gel). The labeled proteins were stored in aliquots
containing 0.02% sodium azide at -20C. Labeled aliquots were used once
and discarded.
Generation of Plasmin
Plasmin was generated from either radiolabeled or unlabeled
plasminogen by reaction with urokinase. Three fil of urokinase (Sigma 20
u/ml) was added to a 400 /xl solution of 1 //M plasminogen containing 0.04
M lysine. The mixture was incubated at 37C for 45 minutes unless stated
otherwise. The efficiency of plasmin generation was followed by
measuring the conversion of the single chain plasminogen molecule (Mr=
90,000 daltons) into heavy chains (Mr=60,000 daltons) and light chains
(Mr=25,000 daltons) as determined by the migration of radiolabeled
proteins, following denaturation and reduction, on 10% SDS-polyacrylamide
gels. The migration of labeled proteins was determined by auto
radiographic exposure of dried gels to Kodak XAR-5 film with intensifying
screens at -70C for 15-20 hours. The integrity of the iodinated plasmin

125
heavy and light chains preparations were examined by SDS-PAGE and
autoradiography in a similar manner.
Polyacrylamide Gel Electrophoresis and Protein Blotting
Electrophoresis was carried out as described by Laemmli (1970).
Polyacrylamide separating gels were 10% and contained 0.1% sodium
dodecylsulfate (SDS), 0.375 M Tris at pH 8.8. Stacking gels were 4% and
contained 0.1% SDS and 0.125 M Tris at pH 6.8. Electrode buffer was
0.024 M Tris, 0.192 M glycine, 0.1% SDS at pH 8.3 Samples were diluted
1:2 with sample buffer containing 0.125 M Tris pH 6.8, 4% SDS, 20%
glycerol, 10% ^3-mercaptoethanol and 0.05% bromophenol blue and heated at
80-90C for 3 minutes. Gels were run at 45 volts constant voltage for
approximately 15-18 hours. Slab gels were used in the Bio-Rad Protean II
system (BioRad, Richmond, CA). Molecular weight markers were run on all
gels. Gels intended for Western blot transfer contained pre-stained
markers (Sigma) applied as a mixture which included: triosephosphate
isomerase (26,600), lactic dehydrogenase (36,500), fumarase (48,500),
pyruvate kinase (58,000), fructase-6-phosphate kinase (84,000), /9-
galactosidase (116,000), and alpha 2-macroglobulin (180,000). After
electrophoresis, gels intended for Western blotting were equilibrated in
25 mM Tris, 0.2 M glycine pH 8.0 containing 20% v/v methanol (electroblot
buffer) for 25 minutes. Protein blotting, from SDS-PAGE gels, was
performed using the 'Trans-Blot SD Semi-Dry' electrophoretic transfer
cell (Bio Rad, Richmond, CA). Nitrocellulose transfer medium, also
equilibrated in electroblot buffer, was sandwiched between the gel and
two sheets of Whatman 3 mm paper. The gel was also backed with two
sheets of 3 mm paper. Blots were blocked by washing a total of four
times in 200-250 ml of 5.0 mM sodium diethylbarbiturate, 0.14 M NaCl,

126
0.5% gelatin, 0.15% Tween 20, 0.004% NaNj pH 7.35 (blotting wash buffer
I) for 15 minutes per wash. Blots were then probed for 3-4 hours at room
temperature while rotating in one of the following probing solutions:
For probing with either plasmin, plasmin-EACA, plasmin heavy chain, or
plasmin light chain, blots were probed with blotting wash buffer I
containing 2.0 mM PMSF and -*-^I-labeled plasmin at 2 x 10^-3 x 10^ cpm/ml
with or without 1.0 mM EACA, or blotting wash buffer I containing 2.0 mM
PMSF and l^I-plasmin heavy or light chain at 2 x 10^-3 x 10* cpm/ml.
For probing with rabbit, anti-plasmin receptor antibody or anti-group C
streptokinase antibody, blots were probed with blotting wash buffer I
containing 4.3 tg IgG per ml of probing solution (approximately a 1:3000
dilution of antisera) for three hours, washed twice for 20 min with 300
mis of blotting wash buffer I, and probed with *^1-Protein G at 2 x 10^-
3 x 10-* cpm/ml. For probing with mouse, anti-group C streptokinase
monoclonal antibodies, blots were probed with blotting wash buffer I
containing a 1:100 dilution of the stock solutions for three hours,
washed twice for 20 min with 300 mis of blotting wash buffer I, followed
by probing for three hours with blotting wash buffer I containing goat,
anti-mouse IgG, antibody (Cappel) at 1.0 //g/ml, washed twice for 20 min
with 300 mis of blotting wash buffer I, followed by probing with blotting
wash buffer I containing ^^^I-Protein G at 2 x 10^-3 x 10^ cpm/ml.
Following the last probing step all blots were washed four times in 200-
300 mis of 0.01 M EDTA, 1.0 M NaCl 0.25% gelatin, 0.15% Tween 20 for 20
minutes per wash. All washing and probing steps were carried out at
ambient temperature. The probed, washed blots were air dried.
Autoradiographs were prepared by exposing the nitrocellulose blots
to Kodak XAR-5 film with an intensifying screen for 15-24 hours at -70C
followed by automated film developing.

127
Molecular weight determinations on Western blots were made possible
by the transfer of the prestained molecular markers.
For staining, gels were fixed in a solution of 40% ethanol and 10%
acetic acid, stained with Coomassie brilliant blue R-250 (0.25% w/v in
40% ethanol and 10% acetic acid) for 1 hour, and destained by soaking in
several changes of 10% ethanol and 10% acetic acid containing a small
quantity of DE 52 (Whatman, England) as a dye adsorbent. All other gels
prepared for staining were silver stained according to the procedure of
Merril et al., (1981).
Solid Phase Assay for Plasminogen Activator Activity
Samples to be tested for plasminogen activator activity by this
assay (Dr. K. Johnston, personal communication) are first resolved by
SDS-PAGE and transferred to nitrocellulose. The nitrocellulose membranes
are then immersed in blocking buffer (10 mM Tris, pH 8.0 containing 0.5%
Tween-20, 0.5 M NaCl and 1.0% bovine serum albumin) for at least one hour
at room temperature. The substrate overlay is prepared as follows: To a
2.0% agarose solution (Bio Rad Richmond, CA) in 0.15 M phosphate buffered
saline, pH 7.5 was equilibrated at 50C, with the chromogenic substrate
S-2251 at a concentration of 100 tg/ml. Human plasminogen free of
plasmin activity, is then added to a final concentration of 20 /ig/ml.
The agarose-substrate-plasminogen solution is then applied to an ethanol
washed glass slide slightly larger than the nitrocellulose membrane
template. The nitrocellulose membrane containing the sample is then
drained of excess blocking buffer and overlayed on the agarose-
plasminogen- substrate gel. It is important to ensure uniform contact
between the nitrocellulose membrane and the gel. The nitrocellulose
membrane is allowed to remain in contact with the agarose for at least

128
one hour at 37C; the time of incubation is dependent upon the
concentration of plasminogen activator present. At the termination of
incubation, the nitrocellulose membrane is removed from the agarose-
plasminogen- substrate gel and immersed for 5 min in freshly prepared
0.1% sodium nitrite dissolved in 1.0 N HCl followed immediately by
immersion for 5 min in 0.5% ammonium sulfamate dissolved in 1.0 N HCl.
The membrane is then transferred to a solution containing 0.05% N-l-
napthylethylenediamine in 47.5% ethanol and observed for the appearance
of red bands indicative of plasmin activity. Plasminogen activators
present on the nitrocellulose membrane will activate the plasminogen
present in the agarose which will in turn cleave the S-2251 substrate
incorporated in the agarose. The chromogenic cleavage product
(paranitroaniline) appears yellow and deposits on the nitrocellulose
membrane. The chemical treatment of the membranes as described above
will convert the deposited yellow paranitroaniline to a red color and fix
it to the membrane. Membranes can be stored under water at 4C.
Production of Rabbit Anti-streptokinase and Rabbit Anti-plasmin Receptor
Polyclonal Antibodies
Highly purified streptokinase from a Group C streptococcus was
obtained from KabiVitrum A.B., Stockholm, Sweden. Approximately 1.38 mg
of the antigen was electrophoresed on a 7.5% SDS-PAGE gel and stained
with Coomassie brilliant blue R-250. The single stained band was cut
from the gel and equilibrated in PBS-Azide. A portion of the gel slice
containing approximately 345 /g of streptokinase was emulsified with an
equal volume of Freund's complete adjuvant. The emulsion was injected
subcutaneously at 6 sites on a rabbit. The rabbit was boosted three
times at two week intervals with the streptokinase polyacrylamide gel
emulsified in Freund's incomplete adjuvant (approximately 300 /g per

129
boost). Generation of polyclonal rabbit anti-plasmin receptor 41,000
dalton protein was prepared in a similar manner using the strain 64/14
mutanolysin extracted plasmin receptor preparation, except that the
41,000 dalton protein was separated on 10% SDS-PAGE gels and used in 150
Hg to 200 ng aliquots for immunization. Pre-immune and immune IgG
fractions were prepared from rabbit sera by Protein A-Sepharose (Sigma)
affinity chromatography.
Production of Mouse Anti-plasmin Receptor Polyclonal Antibody
Plasmin receptor protein (41,000 dalton molecule) was gel purified
on 10% SDS-polyacrylamide gels under reducing conditions as described
above. Following electrophoresis each slab gel was then Western-blotted
to a sheet of nitrocellulose membrane. A small vertical strip from each
nitrocellulose blot was cut and probed with -plasmin as described
above and autoradiographed to confirm the position of the 41,000 dalton
plasmin binding protein. The remainder of each nitrocellulose sheet was
then stained with 1.0% Fast Green according to the procedure described by
Chiles et al., (1987) and the position of the 41,000 dalton band located
and aligned with the autoradiographed strip. The marked band on each
nitrocellulose sheet was then carefully cut out to avoid any
contamination, and divided into four equal fractions containing
approximately 500 pg of protein. The strips were then equilibrated in
PBS and sonicated to a fine powder. The immunization schedule was as
follows: Four groups of 6-8 week old out-bred female mice were used with
ten mice per group. Part I (initial inoculation) groups 1+3: Each mouse
in these groups was injected in two sites with (lOO/il/25/ig) of sonicated
nitrocellulose-Ag in complete Freund's Adjuvant subcutaneously. Group 2:
This group was injected in two sites with (100^il/25/ig) of sonicated

130
nitrocellulose-Ag in a non-Freund's adjuvant (T1501). The vehicle of
this adjuvant has been described by Woodard and Jasman (1985), and the
adjuvant (T1501) has been described by Hunter and Bennett (1984). The
use of this combination was developed by Woodard and Jasman. Group 4:
This group was injected with (100/il/50/ig) of sonicated nitrocellulose-Ag
in Ribi Adjuvant intraperitoneally. Part II first boost (2 weeks after
initial inoculation): Mice in Groups 1+3 were boosted with (100/il/10pg)
of sonicated nitrocellulose-Ag in incomplete Freund's Adjuvant in two
sites subcutaneously. Mice in group 2 were boosted with (100tl/10/g) of
sonicated nitrocellulose-Ag in the non-Freund's adjuvant (Woodard and
Jasman, 1985) subcutaneously. Mice in group 4 were boosted with
(lOO/l/10/ig) of sonicated nitrocellulose-Ag in Ribi Adjuvant
intraperitoneally. Part III second boost (4 weeks after initial
inoculation): Mice in group 1 were boosted with (lOO/il/10/xg) of
solubilized nitrocellulose-Ag in dimethylsulfoxide intraperitoneally.
Mice in group 2 were boosted with (100/il/10^g) of sonicated
nitrocellulose-Ag in dd^O plus glycogen intraperitoneally. Mice in
group 3 were boosted with (100tl/10tg) of sonicated nitrocellulose-Ag
plus glycogen intraperitoneally. Mice in group 4 were boosted with
(100/il/10/ig) of sonicated nitrocellulose-Ag in dd^O intraperitoneally.
Four days after the final boost all mice were sacrificed and ascites
harvested. The peritoneal cavities were washed with PBS and also
harvested, and the resulting fluids from each group pooled and stored as
aliquots at -70C, with 0.02% sodium azide, until used.
Results
Streptokinase from three group A Streptococcal strains (64/14, B923
and A995), and two group C streptococcal strains (ATCC 12449 and 26RP66),

131
and the commercially available purified Kabikinase (KabiVitrum, Sockbolm,
Sweden) were compared physicochemically, functionally and antigenically.
The well characterized highly purified commercially available group C
streptokinase (Kabikinase) was used as a reference. Streptokinase
proteins were obtained from the bacterial strains by growing them in
chemically defined media under pH controlled conditions in order to
optimize the yield of streptokinase in the culture supernatants (Johnston
and Zabriskie, 1986). The quantity of functionally active streptokinase
was measured using the fluid phase plasminogen activator assay described
in the Methods. This assay used the highly purified streptokinase
(Kabikinase) as a standard. The quantities of streptokinase produced by
the various strains of either group A or C streptococcal bacteria are
summarized in Table 5-1. The concentrations of each of the plasminogen
activators (SK)s were expressed as units relative to Kabikinase.
We have previously demonstrated that the binding of human plasmin to
the group A streptococcal strain 64/14 was both inhibitable and
reversible with L-Lysine or EACA. Furthermore, the binding of human
plasmin to the extracted receptor from this bacteria was also shown to be
inhibitable and reversible with these molecules in Western and dot-blot
IOC
assays using -plasmin as the probe. However, it was demonstrated
that the binding of human plasmin to group C streptokinase (Kabikinase)
was not sensitive to L-Lysine or EACA in similar assays. The following
series of experiments were designed to compare the binding specificities
of the extracted plasmin receptor from strain 64/14 to the streptokinase
produced by strain 64/14 from the same culture, as well as those from two
other group A streptococci, two group C streptococci and Kabikinase.
Shown in Figure 5-1 are parallel 10% SDS-PAGE Western blots of extracted

132
Table 5-1.
Fluid-phase plasminogen activator activity assay.
Plasminogen Activator
Activity*
approx.
Strain Units/ml /ig/ml
Group A
Strep.
64/14
3,55070
40
B923
3,970160
40
A995
3,250160
30
Group C
Strep.
26RP66
2,330180
20
ATCC 12449
44,0001170
460
64/14 mutanolysin
extract
5014
0.5
Plasminogen activator activity (e.g. streptokinase) present in
concentrated culture supernatants, and mutanolysin extracted plasmin
binding activity. Estimates were determined using purified Kabikinase
the standard. See Methods for precise experimental details.
as

Figure 5-1. Functional identification and distinction of streptokinase
proteins and plasmin binding receptor protein. Parallel 10% SDS-
polyacrylamide gels were electrophoresed. The proteins on one gel were
transferred to nitrocellulose, and the membrane blocked and used in the
solid-phase plasminogen activation assay (Panel A) as described in the
Methods. The proteins in the second and third gel were transferred to
separate nitrocellulose sheets, and the membranes blocked and probed with
l-2^I-plasmin (Panel B) and '^I-plasmin in the presence of 1.0 mM EACA
(Panel C) as described in the Methods. The resulting blots (Panels B and
C) were autoradiographed at -70C for 24 hours with intensifying screens.
Lane(s) 1: exracted plasmin receptor preparation (approx. 5.0 fig)',
lane(s) 2: approx. 2.0 fig of 64/14 SK; lane(s) 3: approx. 2.0 ng of B923
SK; lane(s) 4: approx. 1.0 fig of A995 SK; lane(s) 5: approx. 2.0 fig of
Kabikinase; lane(s) 6: approx. 2.0 fig of ATCC 12449 SK; lane(s) 7:
approx. 1.0 fig of 26RP66 SK. Streptokinase proteins from the strains
64/14, B923, A995, ATCC 12449, and 26RP66 were contained in concentrated
culture supernatants, see Methods for precise experimental details.

123 4 567123 4567
12 3 4 5 6
KD
I I 6-
84-
5 8-
4 8.5-
36.5-
26.6-i

I I 6-
8 4-
5 8-
3 6.5-
26.6-1
B C
134

135
plasmin receptor and concentrated culture supernatants, which contain
streptokinase, from the strains of streptococcal bacteria tested. Panel
A in Figure 5-1 identifies the molecular species of streptokinase by
means of plasminogen activator potential using a solid phase assay as
described in the Methods. The major species of streptokinase produced by
all but group A B923 had a molecular weight of approximately 48,000
daltons. The major streptokinase molecular species from strain B923 was
slightly smaller, approximately 46,000 daltons. There was no
plasminogen activator activity in the extracted plasmin receptor
preparation. Smaller molecular species of a given streptokinase, as seen
in strain B923, ATCC 12449, and purified Kabikinase, result from
degradation of the larger secreted protein (Johnston and Zabriskie,
1986). The Western blot depicted in Panel B of Figure 5-1 show the ^->I-
plasmin binding activities and demonstrate the major binding activity in
the various culture supernatants is associated with streptokinase,
compare Panel A to B. Panel C of Figure 5-1 shows the ^->I-plasmin
binding results in the presence of EACA, which is known to disrupt the
binding of human plasmin to the 41,000 dalton plasmin receptor. These
results indicate that all of the streptokinases appear to have a binding
activity with human plasmin that is not disrupted in the presence of
EACA. However, the 41,000 dalton extracted plasmin receptor is clearly
shown to be sensitive to the presence of EACA. Compare lane 1 in Panels
B and C of Figure 5-1.
The studies presented in Chapter Three demonstrated that unlike the
group C streptokinase (Kabikinase), which has been shown to associate
with plasmin through interactions with determinants located in the light
chain of the plasmin molecule, the plasmin receptor associated with the

136
surface of the strain 64/14 streptococci possesses a plasmin binding
activity which appears to be associated with determinants present in the
intact heavy chain of plasmin in a conformationally dependent manner. I
therefore examined the binding specificities for ^^1-labeled plasmin
heavy chain and ^^I-labeled plasmin Light chain, by Western blot
analysis, of the extracted plasmin receptor and streptokinase from strain
64/14, as well as the other streptokinases. The results presented in
Figure 5-2 depict the binding reactivities of extracted plasmin receptor
and the streptokinases from the strains of streptococci described above
1-2^1-plasmin heavy chain Panel A, and -plasmin light chain Panel B.
The data clearly indicates that the group C streptokinase (Kabikinase),
in agreement with earlier findings (Summaria and Robbins, 1976) has a
much stronger reactivity with plasmin light chain than with plasmin heavy
chain. The streptokinase from the group C strain ATCC 12449 also shows
preferential reactivity for plasmin light chain. However, streptokinase
proteins associated with 26RP66 and all three of the group A strains
demonstrated equivalent reactivity with plasmin heavy chain and light
chain. The reactivity of the plasmin receptor for either heavy chain or
light chain was relatively weak. This binding is consistent with my
earlier observations (Chapter Three) which indicated that for optimal
binding reactivity of the bacterial plasmin receptor there appeared to be
a requirement for a specific conformation that was best represented in
the intact plasmin molecule or the intact, conformationally modified,
zymogen, Lys-plasminogen.
These results and the previous findings (Chapters Two, Three, and
Four) provide evidence that the plasmin receptor and streptokinase are
physicochemically and functionally distinct molecules. The

Figure 5-2. Comparison of binding reactivities of streptokinase proteins
and plasmin binding receptor protein with -plasmin heavy chain and
^-'i-plasmin light chain. Parallel 10% SDS-polyacrylamide gels were
electrophoresed. The proteins on each gel were transferred to separate
nitrocellulose sheets and the membranes blocked and probed with 3I-
plasmin heavy chain (Panel A) or -plasmin light chain (Panel B) as
described in the Methods. The resulting blots were autoradiographed at -
70C for 24 hours with intensifying screens. Lane(s) 1: exracted plasmin
receptor preparation (approx. 5.0 ng); lane(s) 2: approx. 2.0 /ig of 64/14
SK; lane(s) 3: approx. 2.0 g of B923 SK; lane(s) 4: approx. 1.0 ng of
A995 SK; lane(s) 5: approx. 2.0 ng of Kabikinase; lane(s) 6: approx. 2.0
pg of ATCC 12449 SK; lane(s) 7: approx. 1.0 ng of 26RP66 SK.
Streptokinase proteins from the strains 64/14, B923, A995, ATCC 12449,
and 26RP66 were contained in concentrated culture supernatants, see
Methods for precise experimental details.

138
12 3 4 5 6 7 1
KD
84-
5 8-
48.5-
a
36.5-
26.6-*'
2 3 4 5 6 7
A
B

139
possibility that the two proteins arose from a common precursor and
contained at least some common antigenic determinants was considered.
This possibility was examined using a series of polyclonal and monoclonal
antibodies to streptokinase, or rabbit and mouse polyclonal antibodies to
the isolated strain 64/14 plasmin receptor.
A polyclonal rabbit antibody to group C streptokinase (Kabikinase)
was prepared. This antibody could successfully be used in an ELISA
assay system to measure streptokinase in culture supernatants of both
group C and group A streptococcal bacteria (Reis et al.. 1988). This
polyclonal rabbit antibody to the extracted plasmin receptor was
prepared using the same immunization procedure, see Methods for precise
details. Using these polyclonal antibodies the antigenic relationship of
the strain 64/14 plasmin receptor to the streptokinase produced by strain
64/14 as well as the other streptokinases examined was assessed by SDS-
PAGE and Western blotting; probing with the two different polyclonal
antibodies followed by ^¡>1 _labeled Protein G. The results of these
experiments are presented in Figure 5-3. Panel A of Figure 5-3 shows
that the polyclonal rabbit anti-group C streptokinase antibody could
detect efficiently all the streptokinases examined in concentrated
culture supernatants from all the streptococcal isolates tested including
the group A strain 64/14. The amount of streptokinase in each lane was
approximately 1-2 /g, as measured by the fluid phase plasminogen
activator assay. No reactivity was seen in the lane containing the
41,000 dalton extracted plasmin receptor protein. The receptor lanes
contained approximately 5 ig of the 41,000 dalton protein, and was
readily detected when a parallel gel was probed with the polyclonal
rabbit anti-plasmin receptor antibody (see Panel B of Figure 5-3). This

Figure 5-3. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and streptokinase proteins. Parallel 10% SDS-polyacrylamide
gels were electrophoresed. The proteins in each gel were transferred to
separate nitrocellulose sheets, and one membrane was used in the solid-
phase plasminogen activator assay (Panel A), to identify the molecular
species of plasminogen activator molecules, as described in the Methods.
The second and third membranes were probed in a sandwich assay first with
polyclonal rabbit anti-group C streptokinase IgG (Panel B) or polyclonal
rabbit anti-plasmin receptor IgG (Panel C) followed by ^-1-Protein G,
as described in the Methods. The resulting blots (Panels B and C) were
autoradiographed for 6 hours at -70C with intensifying screens. Lane(s)
1: exracted plasmin receptor preparation (approx. 5.0 Mg); lane(s) 2:
approx. 2.0 Mg of 64/14 SK; lane(s) 3: approx. 2.0 Mg f B923 SK; lane(s)
4: approx. 1.0 Mg of A995 SK; lane(s) 5: approx. 2.0 Mg of Kabikinase;
lane(s) 6: approx. 2.0 Mg of ATCC 12449 SK; lane(s) 7: approx. 1.0 Mg of
26RP66 SK. Streptokinase proteins from the strains 64/14, B923, A995,
ATCC 12449, and 26RP66 were contained in concentrated culture
supernatants, see Methods for precise experimental details.

141
12 3 4 5 6 7
I I 6-1
84
58
48.5
36.5-
26.6
I 2 3 4 5 6 7
B
12 3 4 5 6 7
58-
48.5
36.5-
26.6-
c

142
polyclonal anti-plasmin receptor antibody could detect the 41,000 dalton
protein in the extracted plasmin receptor preparation, as well as in the
culture supernatants of all three group A streptococcal strains tested.
However, this antibody did not detect the 48,000 dalton streptokinase
protein in any of the culture supernatants including strain 64/14.
On prolonged exposure of the Western blots probed with the anti-
receptor antibody, minor bands of reactivity could be seen at in the
range of Mr 48,000 to 55,000 daltons in the culture supernatants from
these bacteria. There was concern that the rabbit immunized to prepare
the anti-plasmin receptor antibody may have previously had a
streptococcal infection and therefore pre-existing antibody to
streptococcal proteins. These reactive bands might be attributable to
Fc-binding proteins produced by the group A (Type II a+b) (Yarnall and
Boyle, 1986a) and group C (Type III) Fc-receptor proteins (Boyle and
Reis, 1987). In similar experiments using normal rabbit IgG a similar
pattern of reactive bands were observed when the autoradiographs were
over-exposed, suggesting that these reactivities were not related to
antigen specific interactions with rabbit immune antibody (data not
shown).
In an attempt to eliminate these background reactivities, I
prepared a polyclonal mouse anti-plasmin receptor antibody since the
frequency of mouse anti-streptococcal antibodies in non-immune
populations is less than observed in non-immune rabbits. Furthermore,
mouse IgG reacts very poorly with Type II Fc-receptor (Yarnall and Boyle,
1986b). The polyclonal mouse anti-plasmin receptor antibodies were
prepared as described in the Methods. Four immunization protocols were
performed on separate groups of outbred mice, all involved the use of the

143
41,000 dalton protein-nitrocellulose in a powder form with an adjuvant
(see Methods for precise details). Following the immunization schedule,
ascites fluid and a PBS wash of the pleural cavities were obtained from
each mouse and the material pooled by group.
The amount of mouse IgG in the ascites and washes of the four groups
were first assayed by serial dilution and dot-blotting, followed by
probing with goat anti-mouse IgG, followed by -*-^^1-Protein G. This was
performed to equalize the amount of ascites or wash among the four groups
of mice to be used later. The ascites and pleural washes of the four
groups of mice were all tested against Western blotted extracted plasmin
receptor from strain 64/14, 64/14 concentrated supernatant containing
streptokinase, and the highly purified commercial group C streptokinase
(Kabikinase). Blots were probed in a sandwich assay with mouse
antibody, followed by goat anti-mouse IgG, followed by ^^1-Protein G.
All four groups of mice showed reactivity with the 41,000 dalton
protein. Two of the four groups had a much higher titer based on the
signal intensity seen on the autoradiograph of the Western blot. Results
using these two, higher titer, polyclonal mouse anti-plasmin receptor
antibodies on Western blots are depicted in Figure 5-4. These data
confirm the previous observations with the polyclonal rabbit anti-
plasmin receptor antibody. The mouse IgG in the ascites fluid clearly
shows strong reactivity with the 41,000 dalton protein in the extracted
plasmin receptor preparation, lane 1 of panels A and B, as well as in the
64/14 concentrated supernatant, lane 2 of panels A and B. However, no
significant reactivity is seen at 48,000 dalton position in the 64/14
concentrated supernatant or in the lane containing group C streptokinase
(Kabikinase), lanes 2 and 3 of panels A and B, respectively. Panel C of

Figure 5-4. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and 64/14 streptokinase and group C streptokinase with mouse
polyclonal anti-plasmin receptor antibodies. Parallel protein samples
were electrophoresed in a 10% SDS-polyacrylamide gel. The proteins were
transferred to nitrocellulose membranes, and the membranes probed in
sandwich assays first with group II polyclonal mouse, anti-plasmin
receptor acities fluid (Panel A) or with group IV polyclonal mouse, anti-
plasmin receptor acities fluid (Panel B) or with polyclonal rabbit, anti-
plasmin receptor IgG followed by goat, anti-mouse IgG antibody (Panels A
and B) followed by Protein G. The resulting blots were auto-
radiographed at -70C for 6 hours with intensifying screens. Lane(s) 1:
approx. 5.0 /g of plasmin receptor protein; lane(s) 2: concentrated 64/14
culture supernatant containing approx. 2.0 ng of 64/14 streptokinase;
lane(s) 3: 4.0 ng of group C streptokinase. See Methods for precise
experimental details.

145
12 3 12 3 12 3
KD
116-
84_ f
58- *
ABC

146
Figure 5-4 shows the reactivity of the rabbit polyclonal anti-plasmin
receptor protein for comparison. These data clearly demonstrate that
streptokinase and the bacterial plasmin receptor do not contain any
common immunodominant epitopes.
The two bacterial plasmin binding proteins were compared for any
evidence of antigenic relatedness using sixteen mouse monoclonal
antibodies, prepared against group C streptokinase. These experiments
were carried out by Western blot analysis. The results of these
experiments are summarized in Figure 5-5. All the monoclonal antibodies
tested reacted efficiently with group C streptokinase as expected (see
lane 3 of each Western blot). Thirteen of the sixteen monoclonal
antibodies reacted equally well with the streptokinase protein from
strain 64/14 present in the concentrated culture supernatant see lane 2
of each Western blot. However none of the monoclonal antibodies examined
reacted significantly with the extracted 41,000 dalton plasmin receptor,
see lane 1 of each Western blot.
These studies demonstrated certain unique epitopes present on group
C streptokinase molecules that are not present on the group A plasminogen
activator molecule. Overall the group A and group C streptokinase
proteins were found to be antigenically closely related, while the group
A plasmin receptor was totally devoid of any of the antigenic
determinants found on streptokinase.
Discussion
The purpose of this investigation was to compare the streptococcal
strain 64/14 receptor for human plasmin with the well characterized
secreted plasmin(ogen) binding streptococcal protein streptokinase, to

Figure 5-5. Analysis of the antigenic relationship of the 64/14 plasmin
receptor and 64/14 streptokinase and group C streptokinase with mouse
anti-group C streptokinase monoclonal antibodies. Parallel protein
samples were electrophoresed on 10% SDS-polyacrylamide gels and
transferred to separate nitrocellulose membranes. Each individual blot
was probed individually in a sandwich assay first with an anti-group C
streptokinase monoclonal antibody, followed by goat, anti-mouse IgG
antibody, followed by ^^^I-Protein G. The resulting blots were auto-
radiographed at -70C for 6 to 8 hours with intensifying screens. The
identity of each monoclonal antibody is depicted below each blot. Lane 1
of each blot contains extracted plasmin receptor protein (approx. 5.0
Hg). Lane 2 of each blot contains 64/14 concentrated culture supernatant
containing approx. 2.0 ng of streptokinase. Lane 3 of each blot contains
2.0 ng of group C streptokinase (Kabikinase). See Methods for precise
experimental details.

KD
I 6-
84-
I 2 3
48.5-
36.5-
26.6-
5 A3
8E2
2D3
I 2G7
I I 6-
58
48.5
36.5
26.6
IBM | 2H7 9F | 2
Lane I. Plasmin Receptor Protei n
Lane 2. Group A SK
Lane 3. Group C SK
I I H4
5G I |
3H4/C7
6H7
I I H5
6D3
I 2H8 4D8 8F4
148

149
explore any possible similarities or differences. Streptokinase is a
unique plasminogen activator. Unlike eukaryotic plasminogen activators,
streptokinase has no enzymatic activity. The activation mechanism lies
in its ability to form a specific 1:1 stoichiometric complex with
plasminogen, as well as with plasmin, which leads to the formation of an
active enzyme moiety presumably through conformational changes in the
plasminogen molecule without the cleavage event at the Arg560'Va]_561
peptide bond (Markus et al., 1976), that can in turn act as a plasminogen
activator for other plasminogen molecules. This is a function neither of
the two proteins alone possesses.
The data presented in Chapters Two, Three, and Four provides good
evidence that the surface associated plasmin binding receptor activity is
distinct from the streptococcal plasminogen activator (streptokinase).
While both the plasmin receptor and streptokinase display a high affinity
interaction towards plasmin; the plasmin receptor specifically recognizes
plasmin while demonstrating no significant reactivity with the native
zymogen precursor Glu-plasminogen (see Chapters Two and Three). By
contrast streptokinase displays similar affinity for plasminogen and
plasmin.
This investigation focused on obtaining plasmin receptor molecules
and secreted streptokinase from a single streptococcal strain 64/14
culture, and comparing the two molecules by three criteria: (1)
functionally, by measuring plasminogen activator activity; (2) plasmin
binding specificity using (a) -plasmin with and without EACA,
(b) plasmin heavy chain and (c) plasmin light chain as probes; and (3)
antigenically, by the production of a polyclonal anti-plasmin receptor
antibodies as well as testing a series of mouse anti-streptokinase

150
monoclonal antibodies. All of these studies were carried out using a
Western blotting approach to enable specific molecular species to be
identified and compared in each of these assays. In addition to using
the group A strain 64/14 as a source of streptokinase, five other sources
(three group A and 2 group C) were included for comparison. In all
studies, the commercially available, highly purified group C
streptokinase (Kabikinase) was included as a reference.
In agreement with my earlier work, the extracted plasmin receptor
preparation did not possess plasminogen activator activity (Table 5-1).
This lack of plasminogen-activator potential was further demonstrated
using a solid-phase plasminogen-activator assay (see Panel A of Figure 5-
1) in which plasminogen activator molecules could be identified and
correlated to a molecular size. The assay identified the streptokinase
produced by strain 64/14, having an Mr of 48,000 daltons which comigrated
with Kabikinase and 3 of the other streptokinases examined. The group A
strain B923 secreted a slightly smaller streptokinase molecule (see Panel
A of Figure 5-1). When these preparations were Western blotted and
probed with '-^^I.plasmin the 41,000 dalton plasmin binding activity
could be clearly identified (see Panel B of Figure 5-1) and shown to
exhibit the sensitivity to EACA previously reported (see Panel C of
Figure 5-1). However, the streptokinase proteins including Kabikinase,
and the strain 64/14 streptokinase were not inhibited from binding ^^I-
plasmin in the presence of EACA (Compare Panels B and C of Figure 5-1).
The results of probing these Western-blotted proteins with either
^^^1-plasmin heavy chain or ^-*1 -plasmin light chain were not as
conclusive. Previous data (see Chapter Three) suggested that the 41,000
dalton protein would be detected preferentially by the heavy chain

151
probe. However, it appeared that either heavy or light chain would
detect the 41,000 dalton protein by Western blot assay. In agreement with
work by Summaria and Robbins (1976) the group C streptokinase
(Kabikinase), as well as the streptokinase produced by strain ATCC 12449
were preferentially detected by probing Western blots with light
chain (see Panel B of Figure 5-2). Surprisingly, not all the
streptokinases were seen to react preferentially with ^^I-light chain.
The comparison of streptokinase and plasmin receptor indicated no
antigenic relatedness. Polyclonal rabbit anti-streptokinase antibody and
both polyclonal rabbit and polyclonal mouse anti-plasmin receptor
antibody were prepared. These antibodies were tested for reactivity to
the extracted plasmin receptor and several streptokinases. The rabbit
anti-plasmin receptor antibodies showed no cross-reactivity with
streptokinase from the strain 64/14, or with two other group A
streptokinases, or any of the three group C streptokinase sources studied
(Figure 5-3). Two polyclonal mouse anti-plasmin receptor antibody
preparations also showed no cross-reactivity with 64/14 streptokinase
or with the prototype group C streptokinase (Kabikinase)(Figure 5-4).
The polyclonal rabbit anti-group C streptokinase antibody detected all
forms of streptokinases examined, while showing no cross-reactivity
with the 41,000 dalton plasmin binding protein in the extract of strain
64/14.
Furthermore, 16 mouse monoclonal antibodies were tested for
reactivity towards the strain 64/14 streptokinase, the extracted plasmin
receptor, and group C streptokinase (Kabikinase). Thirteen of the
sixteen monoclonal antibodies recognized the 48,000 dalton 64/14
streptokinase. As expected, all the monoclonal antibodies reacted with

152
group C streptokinase (Kabikinase), however, none could detect the
41,000 dalton plasmin binding protein in the extract of strain 64/14
(Figure 5-5).
Taken together these findings provide evidence that the single
group A streptococcal strain 64/14 produces two proteins with affinity
for human plasmin(ogen), that can be shown to be physicochemically,
functionally, and antigenically distinct.

CHAPTER SIX
SUMMARY AND CONCLUSIONS
This dissertation describes the first report of a prokaryotic
cellular receptor for the human fibrinolytic enzyme plasmin. This study
was designed to characterize the plasmin receptor more completely in
order to explore what role, if any, it might play in streptococcal
infections. The specific goals of the study were to: (1) identify and
characterize a group A streptococcal receptor for human plasmin; (2) map
the binding site on human plasmin recognized by the bacterial plasmin
receptor; (3) isolate and purify a functionally active group A
streptococcal plasmin receptor; and (4) compare the group A streptococcal
plasmin receptor to the known streptococcal plasmin(ogen) binding protein
streptokinase.
Plasmin is the key component of the mammalian fibrinolytic enzyme
system which is responsible for intravascular blood clot lysis. The two
chain (heavy and light) plasmin serine protease, which cleaves fibrin, is
derived from the circulating single chain zymogen precursor Glu-
plasminogen. This derivation is brought about by plasminogen activators,
which cleave a single peptide bond in the plasminogen molecule (see
Chapter One for a detailed discussion of plasminogen activation). The
light chain of the plasmin molecule contains the enzyme active site
(Robbins and Summaria, 1970; Wiman, 1977). The plasmin heavy chain
contains five homologous triple loop structures known as kringles, which
are responsible for fibrin binding (Sottrup-Jensen et al., 1978).
153

154
Streptococci and streptococcal products have been known to interact
with the fibrinolytic system for many years (Tillett and Sherry, 1949).
The secreted streptococcal plasminogen activator, streptokinase, was
identified by Tillett and Garner in 1933. This protein is known to bind
to both human plasminogen and plasmin with high affinity (Reddy and
Markus, 1972), through interactions with the light chain of the plasmin
molecule (Summaria and Robbins, 1976). The formation of this
streptokinase-plasmin(ogen) complex generates an enzymatic moiety capable
of plasminogen activator activity, a property neither protein alone
possesses.
In the studies documented in Chapter Two, I have identified and
characterized a group A streptococcal surface receptor that binds human
plasmin while demonstrating no significant affinity for the zymogen form
of the molecule Glu-plasminogen. The expression of this binding activity
was seen with bacteria grown in either Todd-Hewitt broth or chemically
defined media. The binding of plasmin to its bacterial receptor did not
inhibit its enzymatic activity. In fact, bacterial bound plasmin was
shown to be capable of cleaving both synthetic substrates, as well as its
natural substrate, fibrin. The bacterial bound plasmin was also shown to
be fully accessible to small protease inhibitors; specifically,
aprotinin, PPACK, and pNpGB. Together, these data suggested that the
bound plasmin molecule interacts with the surface of the bacteria in a
fashion which leaves the active site accessible to substrates. Of great
interest was the observation that the bacterial bound plasmin enzyme was
not capable of being inhibited by plasmin's main physiological inhibitor
c^-antiplasmin which may have important implications in this bacteria's
pathogenic mechanisms.

155
The plasmin(ogen) molecule contains several characteristic lysine
binding sites (Alkjaersig, 1964; Abiko et al.. 1969; Brockway and
Castellino, 1972), one high affinity site and four to five sites of lower
affinity (Markus et al., 1978a; Markus et al., 1978b). The high affinity
site is located in the kringle 1 structure (see Figure 1-1), and one of
the lower affinity sites located in the kringle 4 region (Learch et al.,
1980). These lysine-binding sites participate in plasmin(ogen)
interactions with fibrin (Wiman et al., 1979; Swenson and Thorsen, 1981)
and to a^-antiplasmin (Wiman, 1981). Experiments documented in Chapter
Two show that the plasmin-bacterial receptor interaction is one of very
high affinity. Also, the plasmin-bacterial receptor interaction, like
the first step in the plasmin-a2*antiplasmin interaction and the
plasmin(ogen)- fibrin interaction, was reversibly inhibitable by L-lysine
and EACA. Furthermore, plasmin binding to bacterial plasmin receptor was
inhibited by a^-antiplasmin. Together these data suggest the possible
involvement of the high affinity lysine binding site in the plasmin-
bacterial receptor interaction.
The experiments presented in Chapter Three were designed to examine
the nature of the interaction of plasmin with the bacterial plasmin
receptor which would account for the particular observations described
above. Specific binding to the group A streptococcus, 64/14, was
demonstrated with plasmin's heavy chain. However, the isolated heavy
chain alone was not as efficient a competitor as intact Lys-plasmin. It
was noted that although heavy chain alone could completely inhibit Lys-
plasmin binding to bacterial plasmin receptor, in competition assays,
none of the isolated lysine-binding kringle domains alone or in
combination had any significant inhibitory effect. These data stressed

156
the importance of the conformation of the entire heavy chain for optimal
plasmin binding to bacterial receptor. Therefore the bacterial binding
of plasmin differs from the kind of interactions seen with 2"
antiplasmin, fibrin and fibrinogen. Supporting the hypothesis that a
particular conformation was a requirement for optimal binding were the
data, also presented in Chapter Three, that demonstrated the
conformationally distinct form of the native Glu-plasminogen zymogen
(Lys-plasminogen) could specifically interact with the bacterial plasmin
receptor without converting to Lys-plasmin.
Taken together these observations indicate that the binding of
plasmin to the bacterial plasmin receptor is dependent on the
conformation of the plasmin molecule, and involves interactions that are
distinct from those occurring between other known plasmin(ogen) binding
molecules like fibrin, fibrinogen, -antiplasmin, thrombospondin,
histidine-rich glycoprotein, and streptokinase.
The plasmin receptor activity expressed on the surface of the group
A streptococcal strain 64/14 bacteria was isolated and purified to
functional homogeneity to further characterize and distinguish it as a
unique molecule. The isolation and purification results are the subject
of Chapter Four.
Of a variety of different extraction techniques employed, treatment
of the bacteria with the enzyme mutanolysin yielded the highest quantity
of soluble plasmin binding activity. This soluble plasmin binding
activity was predominantly associated with a 41,000 dalton molecule
identified by Western blot analysis using ^^^1-plasmin as a probe. In
agreement with the earlier data, the soluble plasmin binding activity was
inhibited by L-lysine and EACA, in both dot-blot and Western blot assays

157
with l^I-plasmin. The 41,000 dalton molecule was specifically purified
from the mutanolysin extract of strain 64/14 by affinity chromatography
with immobilized plasmin and elution with L-lysine or EACA. Analysis of
the affinity purified plasmin binding activity by SDS-PAGE and Western
blotting, demonstrated that 41,000 dalton molecule was the predominant
molecule recovered, as detected by silver staining, and this molecule was
also responsible for the -plasmin binding activity of the affinity
purified sample. This plasmin binding molecule could be destroyed by
trypsin digestion and was consequently protein in nature. The 41,000
dalton plasmin binding protein was demonstrated to be totally devoid of
plasminogen activator activity, distinguishing it from the secreted
plasmin(ogen) binding protein streptokinase.
A detailed comparison between the plasmin receptor activity and
streptokinase produced by the same group A strain are the focus of the
studies in Chapter Five. The purpose of this series of experiments was
to make a comparison of the streptococcal strain 64/14 plasmin receptor
with the known and well characterized secreted streptococcal
plasmin(ogen) binding protein streptokinase. These studies focused on
obtaining plasmin receptor molecules as well as secreted streptokinase
from a single culture of streptococcal strain 64/14. These two molecules
were then compared for: (1) functional activity, (2) plasmin binding
specificity, and (3) antigenic relatedness These studies were performed
using Western blot assays to enable specific molecular species to be
identified and compared directly to function or antigenic reactivity.
In addition to streptokinase from strain 64/14, five other sources of
streptokinase were included in these studies for a more complete
comparison.

158
In agreement with the earlier data, the extracted plasmin receptor
preparation did not possess plasminogen activator activity. The
streptokinase proteins from all strains examined, including strain 64/14,
were identified using a solid-phase plasminogen activator assay. This
assay also allowed the molecular weight of the plasminogen activator to
be defined. The 41,000 dalton plasmin receptor lacked the ability to
activate plasminogen. A further distinction between the streptokinase
proteins and the plasmin receptor was observed when the proteins were
probed with ^i-plasmin in the presence of EACA. All of the
streptokinase proteins examined bound ^^^1-plasmin equally well in the
presence of EACA while the bacterial plasmin receptor failed to bind
plasmin in the presence of EACA.
The antigenic relatedness of the plasmin receptor and streptokinase
was compared using polyclonal antibodies prepared in rabbits and mice to
the purified 41,000 dalton plasmin receptor. Mouse and rabbit polyclonal
antibodies were prepared that could recognize the 41,000 dalton protein
while showing no reactivity with the 48,000 dalton plasminogen activator
molecule (streptokinase) produced by the same group A isolate from which
the plasmin receptor activity had been isolated. The 48,000 dalton,
strain 64/14, streptokinase was found to be reactive with a rabbit
polyclonal antibody to group C streptokinase. This antibody recognized
the plasminogen activator molecule (streptokinase) produced by three
group A and two group C isolates that were studied. This antibody failed
to recognize any epitopes on the 41,000 dalton plasmin receptor molecule.
Furthermore none of 16 monoclonal antibodies prepared to different
epitopes of a group C streptokinase reacted with the plasmin receptor.
Thirteen of these antibodies recognized epitopes on the strain 64/14

159
group A streptokinase while three antibodies recognized epitopes on
group C streptokinase which were not present on the group A plasminogen
activator molecule. These findings further confirm the antigenic
difference noted previously between streptokinases of group A and group C
streptococci (Dillon and Wannamaker, 1965; Huang et al.. 1989). These
studies demonstrate that not only does the group A strain 64/14 produce
two distinct proteins with high affinity for human plasmin but there is
no evidence that these proteins share any common structural features.
This dissertation describes the first report of a prokaryotic
cellular receptor for human plasmin. However, eukaryotic cellular
receptors for plasmin(ogen) have been reported recently (Hajjar et al.,
1986; Miles et al., 1986; Miles and Plow 1985; Plow et al., 1986).
Eukaryotic cellular receptors for the plasminogen activators urokinase-
type plasminogen activator (Bajpai and Baker, 1985; Del Rosso et al..
1985; and Vassalli et al., 1985) and tissue-type plasminogen activator
(Beebe, 1987; and Hajjar et al., 1987) have also been described. It has
been concluded by others that the expression of plasminogen receptors
and their colocalization with urokinase- and tissue-type plasminogen
activator receptors provide cells with a basic mechanism for obtaining
proteolytic activity for the purposes of certain cellular functions (Knox
et al. 1987) and for modification of their surroundings (Knudsen et
al., 1986; Sheela and Barret, 1982). It had long been recognized (Carrel
and Burrows, 1911; Lambert and Hanes, 1911) that cancer tissue
consistently caused proteolytic degradation upon culturing. It has since
been shown that this proteolytic activity stems from the ability of many
types of cancer cells to produce and secrete plasminogen activators. In
fact plasmin, in addition to acting within the fibrinolytic system, has

160
been demonstrated to activate complement components, hydrolyze components
of connective tissue and basement membranes such as laminin (Liotta et
al. 1981a; Ott et al.. 1981), fibronectin (Liotta et al., 1981),
proteoglycans (Emonds-Alt et al., 1980), thrombospondin (Lawler and
Slayter, 1981) and type-V collagen (Liotta et al., 1981a,b), as well as
proteolytically activating other latent proteases like collageneases
(O'Grady et al., 1981; Stricklin et al., 1977; and Werb et al., 1977).
These collagenases can then catalyze the initial cleavage of the collagen
molecules, which are then susceptible to further proteolytic action by
plasmin (Dan^, 1985). It is now recognized that many normal cell types
also have the ability to produce plasminogen activators, as well as the
ability to specifically bind the plasminogen activators and
plasmin(ogen).
In conclusion, the ability of certain streptococcal bacteria, like
strain 64/14, to both produce a plasminogen activator (i.e.,
streptokinase) and also to express a receptor having an extremely high
affinity for the activation product plasmin may be important for
certain of the invasive properties of these organism. Furthermore, since
plasmin bound to a group A streptococcus is incapable of inhibition by
plasmin's normal physiological regulator a^-antiplasmin, the bacterium
has acquired a non-regulatable proteolytic activity that may contribute
to its tissue-invasive properties. All of these findings would be
consistent with a linked role for streptokinase and the surface bacterial
plasmin receptor. Together these properties of a group A streptococcus
would provide the bacteria with a mechanism to capture an unregulated
enzyme activity that might alter their ability to interact with host
barriers. In fact, the streptococcal strain 64/14 when grown in human

161
plasma has been documented to both be capable of generating plasmin from
plasminogen, as well as binding the active enzyme to its surface
(DesJardin et al. 1989). The importance of this selective receptor to
the infectious disease process of receptor positive bacteria remains to
be established.

162
REFERENCES
Abiko, Y., M. Iwamoto, and M. Tomikawa. 1969. Pasminogen-plasmin system
V: A stoichiometric equilibrium complex of plasminogen and a
synthetic inhibitor. Biochim. Biophys. Acta. 185:424-431.
Alkjaersig, N. 1964. The purification and properties of human
plasminogen. Biochem. J. 93:171-182.
Aoki, N., M. Moroi, M. Matsuda, and K. Tachiya. 1977. The behavior of
o^-plasmin inhibitor in fibrinolytic states. J. Clin. Invest.
60:361-369.
Astrup, T. 1978. Fibrinolysis: An overview. In: Progress in Chemical
Fibrinolysis and Thrombolysis, Vol. 3, pp. 1-57, Davidson, J.F.,
Towan, R.M., Samana, M.M. and Desnoyers, P.C. (eds.), Raven Press,
New York.
Atichartakarn, V., V.J. Marder, E.P. Kirby, and A.Z. Budzynski. 1978.
Effects of enzymatic degradation on the subunit composition and
biologic properties of human factor VIII. Blood 51:281-297.
Bajpai, A., and J.B. Baker. 1985. Cryptic urokinase binding sites on
human foreskin fibroblasts. Biochem. Biophys. Res. Commun.
133:475-482.
Beebe, D.P. 1987. Binding of tissue plasminogen activator to human
umbilical vein endothelial cells. Thromb. Res. 46:241-254.
Beers, W.H., S. Strickland, and E. Reich. 1975. Ovarian plasminogen
activator: Relationship to ovarian and hormonal regulation. Cell
6:387394.
Bhaduri, S., and Demchick, P.H. 1983. Simple and rapid method for
disruption of bacteria for protein studies. Applied and
Environ. Microbiol. 46:941-943.
Boyle, M.D.P., and K.J. Reis. 1987. Bacterial Fc receptors.
Biotechnology 5:697-703.
Brockway, W., and F. Castellino. 1972. Measurement of the binding of
antifibrinolytic amino acids to various plasminogens. Arch.
Biochem. Biophys. 151:194-199.

163
Broeseker, T.A., M.D.P. Boyle, and R. Lottenberg. 1988.
Characterization of the interaction of human plasmin with its
specific receptor on a group A streptococcus. Microbial
Patho. 5:19-27.
Brogden, R.N., T.M. Speight, and G.S. Avery. 1973. Streptokinase: A
review of its clinical pharmacology, mechanism of action and
therapeutic uses. Drugs 5:357-445.
Burova, L., P. Christensen, R. Grubb, A. Johnson, G. Samuelsson, C.
Schalen, and M.L. Svenson. 1980. Changes in virulence, M protein
and IgG Fc receptor activity in a type 12 group A streptococcal
strain during mouse passage. Acta Path. Microbiol. Scand.
88:199-205.
Burova, L., P. Christensen, R. Grubb, I.S. Krasilnikov, G. Samuelsson, C.
Schalen, M.L. Svenson, and U. Zatterstrom. 1981. IgG-Fc receptors
in T-type 12 group A streptococci from clinical specimens: Absence
from M-type 12 and presence in M-type 22. Acta Path. Microbiol.
Scand. 89:433-435.
Carrel, A., and M.T. Burrows. 1911. Cultivation in vitro of malignant
tumors. J. Exp. Med. 13:571-575.
Castellino, F.J., Sodetz, J.M., Brockway, W.J., and Sefring, G.E., Jr.
1976. Streptokinase. Methods Enzymol. 45:244-257.
Cederholm-Williams, S.A. 1977. The binding of plasminogen (mol. wt.
84,000) and plasmin to fibrin. Thromb. Res. 11:421-423.
Cederholm-Williams, S.A., F. DeCock, H.R. Lijnen, and D. Collen. 1979.
Kinetics of the reactions between streptokinase, plasmin and 02-
antiplasmin. Eur. J. Biochem. 100:125-132.
Chiles, T.C., T.W. O'Brien, and M.S. Kilberg. 1987. Production of
monospecific antibodies to a low-abundance hepatic membrane protein
using nitrocellulose immobilized protein as antigen. Anal. Biochem.
163:136-142.
Christensen, L.R. 1945. Streptococcal fibrinolysis: Proteolytic
reaction due to serum enzyme activated by streptoccus fibrinolysin.
J. Gen. Physiol. 28:363-383.
Christensen, L.R., and C.M. Macleod. 1945. Proteolytic enzyme of serum:
Characterization, activation and reaction with inhibitors. J. Gen.
Physiol. 28:559-583.
Christensen, U., and I. Clemmenson. 1977. Kinetic properties of the
primary inhibitor of plasmin from human plasma. Biochem. J.
163:389-391.
Cleary, P., S. Jones, J. Robbins, and W. Simpson. 1987. Phase variation
and M-protein expression in group A streptococci. In: Streptococcal
Genetics, p.102, Ferretti, J.J., and R. Curtiss III (eds.), American
Society for Microbiology, Washington, D.C..

164
Cleland, W.W. 1967. The statistical analysis of enzyme kinetic data.
Adv. Enzymol. 29:1-32.
Collen, D. 1976. Identification and some properties of a new fast
reacting plasmin inhibitor in human plasma. Eur. J. Biochem.
69:209-216.
Collen, D. 1980. On the regulation and control of fibrinolysis.
Thromb. and Hemostat. 43:77-89.
Collen, D., F. DeCock, and M. Verstralte. 1975. Immunochemical
distinction between antiplasmin and a^-antitrypsin. Thromb. Res.
7:245-249.
Dan<£, K. P.A. Andreasen, J. Grondahl-Hansen, P. Kristensen, L.S.
Neilsen, and L. Skriver. 1985. Plasminogen activators, tissue
degradation, and cancer. Adv. Cancer Res. 44:139-267.
Davies, M.C., M.E. Englert, and E.C. DeRenzo. 1964. Interaction of
streptokinase and human plasminogen. J. Biol. Chem. 239:2651-2656.
Del Rosso, M., G. Dini, and G. Fibbi. 1985. Receptors for plasminogen
activator, urokinase, in normal and Rous Sarcoma virus-transformed
mouse fibroblasts. Cancer Res. 45:630-636.
DesJardin, L., M.D.P. Boyle, and R. Lottenberg. 1988. Group A
streptococci grown in human plasma generate and bind plasmin to a
specific surface receptor. In: Proceedings of the 88th Annual
Meeting of the American Society for Microbiology, p. 57, B-169a,
Hartman, P.A., and Morello, J.A. (eds), American Society for
Microbiology, Washington, D.C..
Diano, M., A. Le Bivic, and M. Hirn. 1987. A method for the production
of highly specific polyclonal antibodies. Anal. Biochem.
166:224229.
Dillon, H.C., and L.W. Wannamaker. 1965. Physical and immunological
differences among streptokinases. J. Exp. Med. 121:351-371.
Edy, J., and D. Collen. 1979. The Interaction in human plasma of
antiplasmin, the fast-reacting plasmin inhibitor with plasmin,
thrombin, trypsin, and chymotrypsin. Biochim. Biophys. Acta
484:423-432.
Elliott, S.D., and V.P. Dole. 1947. An inactive precursor of
streptococcal proteinase. J. Exp. Med. 85:305-310.
Emonds-Alt, X., E. Quisquater, and G. Vaes. 1980. Proteoglycan- and
fibrin-degrading neutral proteinase activities of Lewis lung
carcinoma cells. Eur. J. Cancer 16:1257-1261.
Fritz, H., and G. Wunderer. 1983. Biochemistry and applications of
aprotinin, the kallikrein inhibitor from bovine organs. Arzneim.-
Forsch./Drug Res. 33:479-494.

165
Ginsburg, I. 1972. Mechanisms of cell and tissue injury induced by
group A streptococci: Relation to post-streptococcal sequelae. J.
Infect. Dis. 126:419-456.
Groskopf, W.R., L. Summaria, and K.C. Robbins. 1969. Studies on the
active center of human plasmin: Partial amino acid sequence of a
peptide containing the active center serine residue. J. Biol.
Chem. 244:3590-3597.
Hajjar, K.A., N.M. Hamel, P.C. Harpel, and R.L. Nachman. 1987. Binding
of tissue plasminogen activator to cultured human endothelial cells.
J. Clin. Invest. 80:1712-1719.
Hajjar, K.A., P.C. Harpel, E.A. Jaffe, and R.L. Nachman. 1987. Binding
of plasminogen to cultured human endothelial cells. J. Biol. Chem.
261:11656-11662.
Haupt, H., and Heimburger, N. 1972. Humanserumproteine mit hoher
affinitat zu carboxymethylcellulose I. Hoppe-Seyler's Z. Physiol.
Chem. 353:1125-1132.
Hayes, M.L., and F.J. Castellino. 1979a. Carbohydrate of human
plasminogen variants-I. Carbohydrate composition, glycopeptide
isolation and characterization. J. Biol. Chem. 254:8768-8771.
Hayes, M.L., and F.J. Castellino. 1979b. Carbohydrate of human
plasminogen variants-II. Structure of the asparagine-linked
oligosaccharide unit. J. Biol. Chem. 254:8772-8776.
Hayes, M.L., and F.J. Castellino. 1979c. Carbohydrate of the human
plasminogen variants -III. Structure of the O-glucosidically linked
oligosaccharide unit. J. Biol. Chem. 254:8777-8780.
Holvoet, P., H.R. Lijnen, and D. Collen. 1985. A monoclonal antibody
specific for lys-plasminogen. J. Biol. Chem. 260:12106-12111.
Huang, T-T., H. Malke, and J.J. Ferretti. 1989. Heterogeneity of the
streptokinase gene in group A streptococci. Infec. Immun.
57:502-506.
Hunter, R.L., and B. Bennett. 1984. The adjuvant activity of nonionic
block polymer surfactants. II. Antibody formation and inflammation
related to the structure of triblock and octablock copolymers. J.
Immunol. 133:3167-3175.
Johnston, K.H., and J.B. Zabriskie. 1986. Purification and partial
characterization of the nephritis strain-associated protein from
Streptococcus pyogenes group A. J. Exp. Med. 163:697-711.
Johnston, R.B. 1984. Current concepts: recurrent bacterial infection
in children. N. Engl. J. Med. 310:1237-1243.
Jones, P.A., and Y.A. DeClerck. 1980. Destruction of extracellular
matrices containing glycoproteins, elastin, and collagen by
metastatic human tumor cells. Cancer Res. 40:3222-3227.

166
Kaplan, M.H. 1944. Nature and role of the lytic factor in hemolytic
streptococcal fibrinolysis. Proc. Soc. Exp. Biol, and Med.
57:40-43.
Kay, A.B., D.S. Petter, and R. McKenzie. 1974. The identification of
fibrinopeptide B as a chemotactic agent derived from human
fibrinogen. Brit. J. Haematol. 27:669-677.
Kettner, C., C. Mirabelli, J.V. Pierce, and E. Shaw. 1980. Active site
mapping of human and rat urinary kallikreins by peptidyl
chloromethyl ketones. Arch. Biochem. Biophys. 202:420-430.
Knox, P., S. Crooks, M.C. Scaife, and S. Patel. 1987. Role of
plasminogen, plasmin, and plasminogen activators in the migration of
fibroblasts into plasma clots. J. Cell. Physiol. 132:501-508.
Knudsen, B.S., R.L. Silverstein, L.L.K. Leung, P.C. Harpel, and R.L.
Nachman. 1986. Binding of plasminogen to extracellular Matrix. J.
Biol. Chem. 261:10765-10771.
Kronvall, G. 1973. A surface component in group A, C, and G
streptococci with non-immune reactivity for immunoglobulin G. J.
Immunol. 111:1401-1406.
Kronvall, G., C. Schonbeck, and E. Myhre. 1979. Fibrinogen binding
structures in B-hemolytic streptococci group A,C, and G. Acta.
Path. Microbiol. Scand., Sect. B, 87:303-310.
Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly
of the head of bacteriophage T4. Nature 227:680-685.
Lambert, R.A., and F.M. Hanes. 1911. Characteristics of growth of
sarcoma and carcinoma cultivated .in vitro. J. Exp. Med. 13:495-504.
Lancefield, R.C. 1928. The antigenic complex of Streptococcus
haemolvticus. I. Demonstration of a type specific substance in
extracts of Streptococcus haemolvticus. J. Exp. Med. 47:91-103.
Laurell, C.B. 1966. Quantitative estimation of proteins by
electrophoresis in agarose gel containing antibodies. Anal.
Biochem. 15:45-52.
Lawler, J., and H. Slayter. 1981. The release of heparin binding
peptides from platelet thrombospondin by proteolytic action of
thrombin, plasmin and trypsin. Thromb. Res. 22:267-279.
Lerch, P.G., and E.G. Rickli. 1980. Studies on the chemical nature of
lysine-binding sites and their localization in human plasminogen.
Biochim. Biophys. Acta 625:374-378.
Lerch, P.G., E.G. Rickli, W. Lergier, and D. Gillessen. 1980.
Localization of individual lysine-binding regions in human
plasminogen and investigations on their complex-forming properties.
Eur. J. Biochem. 107:7-13.

167
Leung, L.L.K., R.L. Nachman, and P.C. Harpel. 1984. Complex formation
of platelet thrombospondin with histidine-rich glycoprotein. J.
Clin. Invest. 73:5-12.
Levy, H.B., and Sober, H.A. 1960. A simple chromatographic method for
preparation of gamma globulin. Proc. Soc. Exp. Bio. and Med.
103:250-252.
Lijnen, H.R., Hoylaerts, M., and Collen, D. 1980. Isolation and
characterization of a human plasma protein with affinity for the
lysine binding sites in plasminogen. J. Biol. Chem. 255:10244-
10222.
Liotta, L.A., R.H. Goldfarb, R. Brundage, G.P. Siegal, V. Terranova, and
S. Garbisa. 1981a. Effect of plasminogen activator (urokinase),
plasmin, and thrombin on glycoprotein and collagenous of basement
membrane. Cancer Res. 41:4629-4636.
Liotta, L.A., R.H. Goldfarb, and V.P. Terranova. 1981b. Cleavage of
laminin by thrombin and plasmin: Alpha thrombin selectively cleaves
the beta chain of laminin. Thromb. Res. 21:663-673.
Liu, T-Y., and S.D. Elliott. 1965. Streptococcal proteinase: The
zymogen to enzyme transformation. J. Biol. Chem. 240:1138-1144.
Lottenberg, R., C.C. Broder, and M.D.P. Boyle. 1987. Identification of
a specific receptor for plasmin on a group A streptococcus. Infect.
Immun. 55:1914-1918.
Lottenberg, R., F.R. Dolly, and C.S. Kitchens. 1985. Recurring
thromboembolic disease and pulmonary hypertension associated with
severe hypoplasminogenemia. Am. J. Hematol. 19:181-193.
Malinowski, D.P., J.E. Sadler, and E.W. Davie. 1984. Characterization
of a complementary deoxyribonucleic acid coding for human and
bovine plasminogen. Biochemistry 23:4243-4250.
Marder, V.J., S.E. Martin, and R.W. Caiman. 1982. Clinical aspects of
consumptive thrombohemorrhagic disorders. In: Hemostasis and
Thrombosis: Basic Principles and Clinical Practice, p. 672, R.W.
Colman, J. Hirsh, V.J. Mardar and E.W. Solzman (eds.), J.B.
Lippincott Co., Philadelphia.
Markus, G., J.L. DePosquale, and F.C. Wissler. 1978a. Quantitative
determination of the binding of e-aminocaproic acid to native
plasminogen. J. Biol. Chem. 253:727-732.
Markus, G., J.L. Evers, and G.H. Hubika. 1976. Activator activities of
the transient forms of the human plasminogen streptokinase complex
during its proteolytic conversion to the stable activation complex.
J. Biol. Chem. 251:6495-6504.
Markus, G., J.L. Evers, and G.H. Hubika. 1978b. Comparison of some
properties of native (Glu) and modified (Lys) human plasminogen. J.
Biol. Chem. 253:733-739.

168
Markwell, M.A.K. 1982. A new solid-state reagent to iodinate
proteins. Anal. Biochem. 125:427-432.
McClintock, D.K., and P.H. Bell. 1971. The mechanism of activation of
human plasminogen by streptokinase. Biochem. Biophys. Res. Commun.
3:694-702.
Merril, C.R., D. Goldman, S.A. Sedman, and M.H. Ebert. 1981.
Ultrasensitive stain for proteins in polyacrylamide gels shows
regional variation in cerebrospinal fluid proteins. Science
211:1437-1438.
Miles, L.A., E.G. Levin, and E.F. Plow. 1986. Localization of
fibrinolytic proteins on peripheral blood cells and endothelial
cells. Circulation 74: Suppl. II:246a.
Miles, L.A., and E.F. Plow. 1985. Binding and activation of plasminogen
on the platelet surface. J. Biol. Chem. 260:4303-4311.
Miles, L.A., and E.F. Plow. 1987. Receptor mediated binding of the
fibrinolytic components, plasminogen and urokinase to peripheral
blood cells. Thromb. Haemost. 58:936-942.
Milstone, H. 1941. A factor in normal human blood which participates in
streptococcal fibrinolysis. J. Immunol. 42:109-116.
Mori, M., and N. Aoki. 1976. Isolation and characterization of c*2-
plasmin inhibitor from human plasma. J. Biol. Chem. 251:5956-5965.
Myhre, E.B., and P. Krusela. 1983. Binding of human fibronectin to
group A,C, and G streptococci. Infect. Immun. 40:29-34.
Nilsson, T., I. Sjoholm, and B. Wiman. 1982. Circular dichroism
studies on c*2-antiplasmin and its interaction with plasmin and
plasminogen. Biochem. Biophys. Acta. 705:264-270.
O'Grady, R.L., L.I. Upfold, and R.W. Stephens. 1981. Rat mammary
carcinoma cells secrete active collagenase and active latent enzyme
in the stroma via plasminogen activator. Int. J. Cancer 28:509-515.
Ohkuni, H., J. Friedman, I. Van De Rijn, V.A. Fischetti, T. Poon-King,
and J.B. Zabriskie. 1983. Immunological studies of post
streptococcal sequelae: Serological studies with an extracellular
protein associated with nephritogenic streptococci. Clin. Exp.
Immunol. 54:185-193.
Ohlsson, K., and D., Collen. 1977. Comparison of the reactions of
neutral granulocytic proteases with the major plasma protease
inhibitors and with antiplasmin. Scad. J. Clin. Lab. Invest.
37:345-350.
Ott, U., E. Odermatt, J. Engel, H. Furthmayr, and R. Timpl. 1982.
Protease resistance and conformation of laminin. Eur. J. Biochem.
123:63-72.

169
Plow, E.F., D.E. Freaney, J. Plescia, and L.A. Miles. 1986. The
plasminogen system and cell surfaces: Evidence for plasminogen and
urokinase receptors on the same cell type. J. Cell. Biol.
103:2411-2420.
Rakoczi, I., B. Wiman, and D. Collen. 1978. On the biological
significance of the specific interaction between fibrin, plasminogen
and antiplasmin. Biochem. Biophys. Acta. 540:295-300.
Reddy, K.N.N., and G. Markus. 1972. Mechanism of activation of human
plasminogen by streptokinase. J. Biol. Chem. 247:1683-1691.
Reis, K.J., E.M. Ayoub, and M.D.P. Boyle. 1983. Detection of receptors
for the Fc region of IgG on streptococci. J. Immunol Methods.
59:83-94.
Reis, K.J., E.M. Ayoub, and M.D.P. Boyle. 1984. Streptococcal Fc
receptors: I. Isolation and partial characterization of the
receptor from a group C streptococcus. J. Immunol. 132:3091-3098.
Reis, K.J., G.O. Von Mering, M.A. Karis, E.L. Faulmann, R. Lottenberg,
and M.D.P. Boyle. 1988. Enzyme-labeled type III bacterial Fc
receptors: A versatile tracer for immunoassay. J. Immunol. Meth.
107:273-280.
Reis, K.J., M. Yarnall, E.M. Ayoub, and M.D.P. Boyle. 1984. Effect of
mouse passage on Fc receptor expression by a group A streptococci.
Scand. J. Immunol. 20:433-439.
Robbins, K.C. 1981. The regulation and control of the blood
fibrinolytic system. In: Progress in Chemical Fibrinolysis and
Thrombolysis, Vol. 5, pp. 3-13, Davidson, J.F., Nilsson, I. Mare
Astedt, B. (eds.), Churchill Livingstone, Edinburgh.
Robbins, K.C., and L. Summaria. 1970. Human plasminogen and plasmin.
Methods Enzymol. 19:184-199.
Robbins, K.C., L. Summaria, B. Hsieh, and R.J. Shah. 1967. The peptide
chains of human plasmin: Mechanism of activation of human
plasminogen to plasmin. J. Biol. Chem. 242:2333-2342.
Russell-Jones, G.J., E.C. Gotschlich, and M.S. Blake. 1984. A surface
receptor specific for human IgA on group B streptococci possessing
the Ibc protein antigen. J. Exp. Med. 160:1467-1475.
Ryan, T.J. 1987. Photoaffinity labeling of human plasmin and
plasminogen. Pharmac. Ther. 34:335-348.
Scatchard, G. 1949. The attractions of proteins for small molecules
and ions. Ann. N.Y. Acad. Sci. 51:660-672.
Scopes, R.K. 1982. Protein Purification: Principles and Practice,
Springer-Verlag, New York.

170
Sheela, S., and J.C. Barrett. 1982. In vitro degradation of
radiolabeled, intact basement membrane mediated by cellular
plasminogen activator. Carcinogenesis 3:363-369.
Siegal, J.L., S.F. Hurst, E.S. Liberman, S.E. Coleman, and A.S. Bleiweis.
1981. Mutanolysin-induced spheroplasts of Streptococcus mutans are
true protoplasts. Infect. Immun. 31:303-815.
Silverstein, R.L., and R.L. Nachman. 1987. Thrombospondin-plasminogen
interactions: Modulation of plasmin generation. Sem. Thromb. and
Hemo. 13:335-342.
Sottrup-Jensen, L., N. Claeys, M. Zecler, T.E. Peterson, and S.
Magnusson. 1978. Isolation of two lysine-binding fragments and one
'mini'-plasminogen (mw 38,000) by elastase catalyzed-specific
limited proteolysis. In: Progress in Chemical Fibrinolysis and
Thrombolysis, Vol. 3, pp. 191-209, Davidson, J.R., Rowan, R.M.,
Samana, M.M. and Desnoyers, P.C. (eds.), Raven Press, New York.
Sparling, P.F. 1983. Bacterial virulence and pathogenesis: An
overview. Rev. Infect. Dis. 5:637-646.
Stief, T.W., P. Lenz, and V. Becker. 1987. A simple method for
producing degradation products of fibrinogen by an insoluble
derivative of plasmin. Thromb. Res. 48:603-609.
Strickland, S., and W.H. Beers. 1976. Studies on the role of
plasminogen activator in ovulation. J. Biol. Chem. 251:5694-5702.
Strickland, S., E. Reich, and M.I. Sherman. 1976. Plasminogen activator
in early embryogenesis: Enzyme production by trophoblast and
parietal endoderm. Cell 9:231-240.
Stricklin, G.P., E.A. Baner, J.J. Jeffrey, and A.Z. Eisen. 1977. Human
skin collagenase: Isolation of precursor and active forms from both
fibroblast and organ cultures. Biochem. 16:1607-1615.
Summaria, L., and K.C. Robbins. 1976. Isolation of a human plasmin
derived functionally active light (B) chain capable of forming with
streptokinase an equimolar light (B) chain-streptokinase complex
with plasminogen activator activity. J. Biol. Chem. 251:5810-
5813.
Swenson, E., and Thorsen, S. 1981. Secondary-site binding of Glu-
plasmin, Lys-plasmin and mini-plasmin to fibrin. Biochem. J.
197:619- 628.
Switalski, L.M., P. Speziale, M. Hook, T. Waldstrom, and R. Timpl. 1984.
Binding of Streptococcus pyogenes to laminin. J. Biol. Chem.
259:3734-3738.
Teger-Nilsson, A.C., P. Friberger, and E. Gyzander. 1977. Determination
of a new rapid plasmin inhibitor in human blood by means of a
plasmin specific tripeptide substrate. Scand. J. Clin. Lab. Invest.
37:403-409.

171
Thorsen, S. 1975. Differences in the binding to fibrin of native
plasminogen and plasminogen modified by proteolytic degradation:
Influence of W-aminocarboxylic acids. Biochem. Biopbys. Acta.
393:55-65.
Tborsen, S., I. Clemmensen, L. Sottrup-Jensen, and S. Magnusson.
1981. Absorption to fibrin of native fragments of known primary
structure from human plasminogen. Biochem. Biophys. Acta. 668:377-
387.
Tillett, W.S., and R.L.L. Garner. 1933. The fibrinolytic activity of
hemolytic streptococci. J. Exp. Med. 58:485-502.
Tillett, W.S., and S. Sherry. 1949. The effect in patients of
streptococcal fibrinolysin (streptokinase) and streptococcal
deoxyribonuclease on fibrinous, purulent and sanguinous pleural
exudations. J. Clin. Invest. 28:173-190.
Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
Procedure and some applications. Proc. Natl. Acad. Sci. USA.
76:4350-4354.
Van De Rijn, I., and R.E. Kessler. 1980. Growth characteristics of
group A streptococci in a new chemically defined medium. Infect,
and Immun. 27:444-448.
Vassallii, J-D., D. Baccino, and D. Belin. 1985. A cellular binding
site for the Mr 55,000 form of the human plasminogen activator,
urokinase. J. Cell. Biol. 100:86-92.
Violand, B.N., J.M. Sodetz, and F.J. Castellino. 1975. The effect of e-
Amino Caproic Acid on the gross conformation of plasminogen and
plasmin. Arch. Biochem. and Biophys. 170:300-305.
Von Mering, G.O., M.D.P. Boyle, C.C. Broder, and R. Lottenberg. 1988.
Isolation of a surface receptor for human plasmin from a pathogenic
group A streptococcus. Clin. Res. 36:464a.
Walln, P., and B. Wiman. 1970. Characterization of human plasminogen:
I. On the relationship between different molecular forms of
plasminogen demonstrated in plasma and found in purified
preparations. Biochim. Biophys. Acta. 221:20-30.
Walln, P., and B. Wiman. 1972. Characterization of human plasminogen:
II. Separation and partial characterization of different molecular
forms of human plasminogen. Biochim. Biophys. Acta. 257:122-134.
Walz, D.A., T. Bacon-Saguley, S. Kendra-Franczak, and P. DePoli. 1987.
Binding of thrombospodin to immobilized ligands specific
interaction with fibrinogen, plasminogen, histidine-rich
glycoprotein, and fibronectin. Sem. Thromb. and Hemo. 13:317-325.

172
Weber, K., and M. Osborn. 1969. The reliability of molecular weight
determinations by Dodecyl Sulfate-polyacrylamide gel
electrophoresis. J. Biol. Chem. 244:4406-4412.
Weinstein, L. 1953. Antigenic dissimilarity of streptokinases. Proc.
Soc. Exp. Biol. Med. 83:689-691.
Werb, Z., C.L. Mainardi, C.A. Vater, and E.D. Harris. 1977. Endogenous
activation of latent collagenase by rheumatoid synovial cells:
Evidence for a role of plasminogen activator. N. Engl. J. Med.
296:1017-1023.
Wiman, B. 1973. Primary structure of peptides released during
activation of human plasminogen by urokinase. Eur. J. Biochem.
39:1-9.
Wiman, B. 1977. Primary structure of the B-chain of human plasmin.
Eur. J. Biochem. 76:129-137.
Wiman, B. 1981. Human a^-Antiplasmin. Methods Enzymol. 80:395-408.
Wiman, B., L. Bowman, and D. Collen. 1978. On the kinetics of the
reaction between human antiplasmin and a low- molecular-weight form
of plasmin. Eur. J. Biochem. 87:143-146.
Wiman, B., and D. Collen. 1977. Purification and characterization of
human antiplasmin, the fast-acting plasmin inhibitor in plasma.
Eur. J. Biochem. 78:19-26.
Wiman, B., and D. Collen. 1978. On the kinetics of the reaction between
human antiplasmin and plasmin. Eur. J. Biochem. 84:573-578.
Wiman, B., H.R. Lijnen, and D. Collen. 1979. On the specific
interaction between the lysine-binding sites in plasmin and
complementary sites in a^-antiplasmin and in fibrinogen. Biochem.
Biophys. Acta. 579:142-154.
Wiman, B., and P. Walln. 1973. Activation of human plasminogen by an
insoluble derivative of urokinase. Eur. J. Biochem. 36:25-31.
Wiman, B., and P. Walln. 1975. Structural relationship between
"glutamic acid-*- and lysine forms of human plasminogen and their
interaction with the NH2-terminal alteration peptide as studied by
affinity chromatography. Eur. J. Biochem. 50:489-494.
Winn, E.S., S.P. Hu, S.M. Hochschwender, and R.A. Laursen. 1980.
Studies on the lysine-binding sites of human plasminogen: The
effect of ligand structure on the binding of lysine analogs to
plasminogen. Eur. J. Biochem. 104:579-586.
Woodard, L.F., and R.L. Jasman. 1985. Stable oil-in-water emulsions:
preparation and uses as vaccine vehicles for lipophilic adjuvants.
Vaccine 3:137-144.

173
Yarnall, M., E.M. Ayoub, and M.D.P. Boyle. 1986. Analysis of surface
receptor expression on bacteria isolated from patients with
endocarditis. J. Gen. Microbiol. 132:2049-2052.
Yarnall, M., and M.D.P. Boyle. 1986a. Isolation and characterization of
typella and type lib Fc receptors from a group A streptococcus. J.
Immunol. 24:549-557.
Yarnall, M., and M.D.P. Boyle. 1986b. Isolation and partial
characterization of a type II Fc receptor from a group A
streptococcus. Mol. Cell. Biochem. 70:57-66.
Yarnall, M., K.J. Reis, E.M. Ayoub, and M.D.P. Boyle. 1984. An
immunoblotting technique for the detection of bound and secreted
bacterial Fc receptors. J. Microb. Meth. 3:83-93.
Zolton, R.P., and E.T. Mertz. 1972. Studies on plasminogen. X.
Isolation of plasminogen by affinity chromatography using Sepharose
butesin. Can. J. Biochem. 50:529-537.

BIOGRAPHICAL SKETCH
Christopher Charles Broder was born on June 12, 1961 in White
Plains, New York, the heart of New York suburbia, where he lived for
seventeen years. He graduated from White Plains High School in 1979,
and having been a member of the 'Jacques Cousteau television generation',
went south to attend the Florida Institute of Technology, Melbourne,
Florida. He studied marine science for most of his undergraduate career,
obtained a Dive Master certification and proceeded to explore every
available body of water in the state. He traveled back home in the
summers, and worked for Peckham Industries in Connecticut, where he
helped finance his academic life by working blacktop. He graduated with
a B.S. degree in biology, in 1983, along with Colleen M. Guay whom he
would later marry. He stayed on at F.I.T. entering graduate studies, and
received his M.S. degree in molecular biology in 1985, and hoped he would
never have to return to the wheelbarrow and shovel. In the fall of that
year he moved to Gainesville, Florida, and entered graduate school in the
department of Immunology and Medical Microbiology at the University of
Florida, and began his studies in the laboratory of Dr. Michael D.P.
Boyle in 1986. He has completed his research, and expects to receive
his Ph.D. in May of 1989 from the University of Florida.
174

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Michael D.P. Boyle,/Ph.D., Chairman
Professor of Immunology and
Medical Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
C/
Richard Lottenberg, M.D.,
Associate Professor of Biochemistry
and Molecular Biology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Richard W. Moyer, Pf
Professor of ImmunoLgy and
Medical Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
LO \
/sepn W. Shands, Jr., M.D.
rofessor of Immunology and
Medical Microbiology

This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor of
Philosophy.
May, 1989
Dean, College of Medicine
Dean, Graduate School

UNIVERSITY OF FLORIDA



142
polyclonal anti-plasmin receptor antibody could detect the 41,000 dalton
protein in the extracted plasmin receptor preparation, as well as in the
culture supernatants of all three group A streptococcal strains tested.
However, this antibody did not detect the 48,000 dalton streptokinase
protein in any of the culture supernatants including strain 64/14.
On prolonged exposure of the Western blots probed with the anti-
receptor antibody, minor bands of reactivity could be seen at in the
range of Mr 48,000 to 55,000 daltons in the culture supernatants from
these bacteria. There was concern that the rabbit immunized to prepare
the anti-plasmin receptor antibody may have previously had a
streptococcal infection and therefore pre-existing antibody to
streptococcal proteins. These reactive bands might be attributable to
Fc-binding proteins produced by the group A (Type II a+b) (Yarnall and
Boyle, 1986a) and group C (Type III) Fc-receptor proteins (Boyle and
Reis, 1987). In similar experiments using normal rabbit IgG a similar
pattern of reactive bands were observed when the autoradiographs were
over-exposed, suggesting that these reactivities were not related to
antigen specific interactions with rabbit immune antibody (data not
shown).
In an attempt to eliminate these background reactivities, I
prepared a polyclonal mouse anti-plasmin receptor antibody since the
frequency of mouse anti-streptococcal antibodies in non-immune
populations is less than observed in non-immune rabbits. Furthermore,
mouse IgG reacts very poorly with Type II Fc-receptor (Yarnall and Boyle,
1986b). The polyclonal mouse anti-plasmin receptor antibodies were
prepared as described in the Methods. Four immunization protocols were
performed on separate groups of outbred mice, all involved the use of the


78


143
41,000 dalton protein-nitrocellulose in a powder form with an adjuvant
(see Methods for precise details). Following the immunization schedule,
ascites fluid and a PBS wash of the pleural cavities were obtained from
each mouse and the material pooled by group.
The amount of mouse IgG in the ascites and washes of the four groups
were first assayed by serial dilution and dot-blotting, followed by
probing with goat anti-mouse IgG, followed by -*-^^1-Protein G. This was
performed to equalize the amount of ascites or wash among the four groups
of mice to be used later. The ascites and pleural washes of the four
groups of mice were all tested against Western blotted extracted plasmin
receptor from strain 64/14, 64/14 concentrated supernatant containing
streptokinase, and the highly purified commercial group C streptokinase
(Kabikinase). Blots were probed in a sandwich assay with mouse
antibody, followed by goat anti-mouse IgG, followed by ^^1-Protein G.
All four groups of mice showed reactivity with the 41,000 dalton
protein. Two of the four groups had a much higher titer based on the
signal intensity seen on the autoradiograph of the Western blot. Results
using these two, higher titer, polyclonal mouse anti-plasmin receptor
antibodies on Western blots are depicted in Figure 5-4. These data
confirm the previous observations with the polyclonal rabbit anti-
plasmin receptor antibody. The mouse IgG in the ascites fluid clearly
shows strong reactivity with the 41,000 dalton protein in the extracted
plasmin receptor preparation, lane 1 of panels A and B, as well as in the
64/14 concentrated supernatant, lane 2 of panels A and B. However, no
significant reactivity is seen at 48,000 dalton position in the 64/14
concentrated supernatant or in the lane containing group C streptokinase
(Kabikinase), lanes 2 and 3 of panels A and B, respectively. Panel C of


COUNTS BOUND
37


52
activation procedure described to generate Lys-plasmin. Phe-pro-arg-
chloromethyIketone (PPACK) reacted radiolabeled or unlabeled plasmin was
obtained by mixing a 5-fold molar excess of the inhibitor with plasmin
and incubating at 37C for 30 min.
Bacteria
The group A, /3-hemolytic, streptococcal strain 64/14 was grown in
Todd-Hewitt broth (Difco, Detroit, MI) overnight at 37C as stationary
cultures (Yarnall and Boyle, 1986a). The bacteria were harvested by
centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4,
containing 0.05% Tween 20 and 0.02% sodium azide. The bacteria were heat
killed at 80C for 15 min. The suspension was centrifuged and the pellet
washed twice with VBS-gel containing 0.02% sodium azide. Aliquots were
stored at -20C. Stocks of 10% wet weight/vol suspensions were prepared
in VBS-gelatin containing 0.02% sodium azide. The concentration of a
bacterial suspension was determined by counting bacterial chains in a
Neubauer hemacytometer (Fisher Scientific, Orlando, FL).
Polyacrylamide Gel Electrophoresis
Electrophoresis was carried out as described by Weber and Osborn
(1969) with the addition of 6.0 M urea to the polyacrylamide gel. The
polyacrylamide gels consisted of a 4% stacking gel layered onto a 10% or
12% polyacrylamide gel containing 0.1% sodium dodecylsulfate, 0.05 M
sodium phosphate pH 7.1, 6.0 M urea. Slab gels were used in the Bio-Rad
Protean II system (BioRad, Richmond, CA). Protein samples were prepared
by mixing an equal volume of sample buffer containing 0.1 M sodium
phosphate pH 7.1, 8.0 M Urea, and 4.0% SDS with the protein solution, and
heating at 80C for 2 minutes. Sample buffer containing 0.72 M /3-
mercaptoethanol was used to prepare protein samples in the reduced state.


156
the importance of the conformation of the entire heavy chain for optimal
plasmin binding to bacterial receptor. Therefore the bacterial binding
of plasmin differs from the kind of interactions seen with 2"
antiplasmin, fibrin and fibrinogen. Supporting the hypothesis that a
particular conformation was a requirement for optimal binding were the
data, also presented in Chapter Three, that demonstrated the
conformationally distinct form of the native Glu-plasminogen zymogen
(Lys-plasminogen) could specifically interact with the bacterial plasmin
receptor without converting to Lys-plasmin.
Taken together these observations indicate that the binding of
plasmin to the bacterial plasmin receptor is dependent on the
conformation of the plasmin molecule, and involves interactions that are
distinct from those occurring between other known plasmin(ogen) binding
molecules like fibrin, fibrinogen, -antiplasmin, thrombospondin,
histidine-rich glycoprotein, and streptokinase.
The plasmin receptor activity expressed on the surface of the group
A streptococcal strain 64/14 bacteria was isolated and purified to
functional homogeneity to further characterize and distinguish it as a
unique molecule. The isolation and purification results are the subject
of Chapter Four.
Of a variety of different extraction techniques employed, treatment
of the bacteria with the enzyme mutanolysin yielded the highest quantity
of soluble plasmin binding activity. This soluble plasmin binding
activity was predominantly associated with a 41,000 dalton molecule
identified by Western blot analysis using ^^^1-plasmin as a probe. In
agreement with the earlier data, the soluble plasmin binding activity was
inhibited by L-lysine and EACA, in both dot-blot and Western blot assays


14
To each tube, 20 1 of an 8.0 mM solution of the chromogenic
substrate, H-D-val-leu-lys-paranitroanilide, was added,to yield a final
concentration of substrate in the reaction mixture of 400 M. The tubes
were mixed by vortexing and incubated at 37C for 25 minutes. At that
time the enzyme reaction was quenched by the addition of 400 ti of 10%
(v/v) acetic acid, the samples were then centrifuged for 5 minutes at
10,000 x g and the optical densities of the solutions at 405 nm were
determined. The release of paranitroaniline from the synthetic substrate
monitored at this wavelength was directly proportional to the enzymatic
activity of plasmin. Control samples of substrate alone and substrate
plus bacteria were included and each estimate was carried out in
duplicate.
Effect of pH on Plasmin Binding to Bacteria
To assess the effect of pH on the bacteriumrplasmin(ogen)
interaction, 50 ti of labeled plasminogen or plasmin (approximately 2 x
10^ cpm) were added to 1.0 ml of VBS containing 0.05% Tween-20 adjusted
to the appropriate pH. After 15 minutes at room temperature, 50 tl of
VBS containing approximately 10^ bacteria (strain 64/14) were added and
the mixture was incubated at 37 C for 15 minutes. The bacterial
suspensions were centrifuged at 1,000 x g for 7 minutes to separate
bacteria from unbound labeled proteins and the pellets were washed twice
with 2.0 ml of VBS at the appropriate pH. The radioactivity associated
with the bacterial pellet in duplicate experiments was measured using a
Beckman 5500 autogamma counter.
To assess the effect of ionic strength on the bacterium-
plasmin(ogen) interaction, similar studies were carried out in solutions
containing different concentrations of NaCl with 0.05% Tween-20. The


CHAPTER THREE
LOCALIZATION OF THE DOMAIN OF PLASMIN INVOLVED IN BINDING TO ITS
SPECIFIC GROUP A STREPTOCOCCAL RECEPTOR
Introduction
The studies documented in Chapter Two demonstrate that certain
pathogenic group A streptococci, grown in either Todd-Hewitt broth or
chemically defined media, express a receptor that binds to human plasmin
while demonstrating no significant reactivity with the native zymogen
form of the protein, Glu-plasminogen or with other serine class
proteases. Bacterial-bound plasmin retains its enzymatic activity and
can no longer be regulated by its physiological inhibitor, a^-
antiplasmin. Optimal binding of plasmin to its bacterial receptor was
shown to occur under physiological conditions of ionic strength and pH.
This interaction of plasmin with a group A streptococcus had a high
affinity with an estimated dissociation constant of approximately 1.0 x
10"-*- M. Plasmin binding was inhibited reversibly by lysine or epsilon
amino caproic acid, (EACA). These data suggest that the lysine binding
kringle structures of the plasmin molecule might be involved in the
association of plasmin with the bacterial receptor. In this chapter I
describe the experiments performed to localize the region of the plasmin
molecule which interacts with the bacterial plasmin receptor. Binding of
plasmin to a group A streptococcus is dependent on the conformation of
the plasmin molecule, and involves interactions that are distinct from
those occurring between other known plasmin(ogen) binding molecules like
streptokinase, fibrin, fibrinogen, thrombospondin, or a2-antiplasmin.
49


168
Markwell, M.A.K. 1982. A new solid-state reagent to iodinate
proteins. Anal. Biochem. 125:427-432.
McClintock, D.K., and P.H. Bell. 1971. The mechanism of activation of
human plasminogen by streptokinase. Biochem. Biophys. Res. Commun.
3:694-702.
Merril, C.R., D. Goldman, S.A. Sedman, and M.H. Ebert. 1981.
Ultrasensitive stain for proteins in polyacrylamide gels shows
regional variation in cerebrospinal fluid proteins. Science
211:1437-1438.
Miles, L.A., E.G. Levin, and E.F. Plow. 1986. Localization of
fibrinolytic proteins on peripheral blood cells and endothelial
cells. Circulation 74: Suppl. II:246a.
Miles, L.A., and E.F. Plow. 1985. Binding and activation of plasminogen
on the platelet surface. J. Biol. Chem. 260:4303-4311.
Miles, L.A., and E.F. Plow. 1987. Receptor mediated binding of the
fibrinolytic components, plasminogen and urokinase to peripheral
blood cells. Thromb. Haemost. 58:936-942.
Milstone, H. 1941. A factor in normal human blood which participates in
streptococcal fibrinolysis. J. Immunol. 42:109-116.
Mori, M., and N. Aoki. 1976. Isolation and characterization of c*2-
plasmin inhibitor from human plasma. J. Biol. Chem. 251:5956-5965.
Myhre, E.B., and P. Krusela. 1983. Binding of human fibronectin to
group A,C, and G streptococci. Infect. Immun. 40:29-34.
Nilsson, T., I. Sjoholm, and B. Wiman. 1982. Circular dichroism
studies on c*2-antiplasmin and its interaction with plasmin and
plasminogen. Biochem. Biophys. Acta. 705:264-270.
O'Grady, R.L., L.I. Upfold, and R.W. Stephens. 1981. Rat mammary
carcinoma cells secrete active collagenase and active latent enzyme
in the stroma via plasminogen activator. Int. J. Cancer 28:509-515.
Ohkuni, H., J. Friedman, I. Van De Rijn, V.A. Fischetti, T. Poon-King,
and J.B. Zabriskie. 1983. Immunological studies of post
streptococcal sequelae: Serological studies with an extracellular
protein associated with nephritogenic streptococci. Clin. Exp.
Immunol. 54:185-193.
Ohlsson, K., and D., Collen. 1977. Comparison of the reactions of
neutral granulocytic proteases with the major plasma protease
inhibitors and with antiplasmin. Scad. J. Clin. Lab. Invest.
37:345-350.
Ott, U., E. Odermatt, J. Engel, H. Furthmayr, and R. Timpl. 1982.
Protease resistance and conformation of laminin. Eur. J. Biochem.
123:63-72.


74
Table 3-2.
Measurement of nlasmin(ogen') associated with bacterial pellets.
Bacteria Pre-Incubated With:
Absorbance at 405 nm
Following a 20 Minute
Incubation at 37C with H-
D-Val-Leu-Lys-Paranitroanilide
Buffer
0.023 0.002
Glu-plasminogen
0.026 0.002
Lys-plasminogen
0.095 0.004
Lys-plasmin^
0.630 0.001
O
Lys-plasmin^
0.637 0.002
Two ng of the indicated enzyme or zymogen was incubated with a
fixed dilution of the streptococcal strain 64/14. Following incubation
the bacteria were pelleted by centrifugation, washed, resuspended in 400
/I VBS-gelatin, and assayed for enzymatic activity by hydrolysis of the
chromogenic substrate H-D-Val-Leu-Lys-paranitroanilide (see Materials and
Methods). The spontaneous cleavage of the substrate under the
experimental conditions in the presence of bacteria alone was an
absorbance (405 nm) of 0.024 0.002.
1. Urokinase activated Glu-plasminogen
2. Urokinase activated Lys-plasminogen


95
taken was analyzed by SDS-PAGE under reduced conditions for the
conversion of the single chain plasminogen molecule to the two chain
plasmin form. Once it was established that the plasminogen was fully
activated, the bulk preparation was reacted with a 5 fold molar excess of
D-valyl-L-phenylalanyl-L-lysine chloromethyl ketone (VPLCK) (Calbiochem),
an irreversible inhibitor of the enzyme activity of plasmin. This enzyme
inactivation was carried out at ambient temperature with constant
rotation. The enzymatically inactive plasmin was then concentrated by
ammonium sulfate precipitation (4.0 g / 10 ml), and dialyzed at 4C
against 0.1 M MOPS buffer, pH 7.3, containing 0.02% sodium azide. The
dialyzed inactive plasmin was then chromatographed on Superse 6
(Pharmacia) in 0.1 M MOPS buffer, pH 7.3.
The activated affinity chromatography support Affi-Prep 10 (Bio Rad)
was selected as the matrix for immobilizing the chlormethyl ketone
blocked plasmin. This matrix couples in aqueous buffers by means of an
N-hydroxysuccinimide ester on the end of a 10 carbon space arm to primary
amino groups in the ligand. The ligand is linked by amide bonds to the
terminal carboxyl groups of the Affi-Prep 10 spacer arm. The buffer used
in the coupling reaction was 0.1 M MOPS buffer, pH 7.3. Approximately 50
mg of inactivated plasmin in 18 ml of coupling buffer was incubated with
6.0 ml of washed Affi-Prep 10. The reaction was carried out at 4C for
15 hours with rotation. Following ligand coupling, 100 /xl of 1.0 M
ethanolamine HC1 pH 8 was added to the reaction mixture to block
remaining active sites. This blocking reaction was completed in 1 hour
at 4C with sample rotation. The matrix was washed with two 1.0 ml
aliquots of 1.5 M NaCl. The supernatant from the coupling reaction and
the two 1.5 M NaCl wash volumes combined and dialyzed against PBS at 4C.


67
A
1. 2. 3. 4.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Michael D.P. Boyle,/Ph.D., Chairman
Professor of Immunology and
Medical Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
C/
Richard Lottenberg, M.D.,
Associate Professor of Biochemistry
and Molecular Biology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Richard W. Moyer, Pf
Professor of ImmunoLgy and
Medical Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
LO \
/sepn W. Shands, Jr., M.D.
rofessor of Immunology and
Medical Microbiology


164
Cleland, W.W. 1967. The statistical analysis of enzyme kinetic data.
Adv. Enzymol. 29:1-32.
Collen, D. 1976. Identification and some properties of a new fast
reacting plasmin inhibitor in human plasma. Eur. J. Biochem.
69:209-216.
Collen, D. 1980. On the regulation and control of fibrinolysis.
Thromb. and Hemostat. 43:77-89.
Collen, D., F. DeCock, and M. Verstralte. 1975. Immunochemical
distinction between antiplasmin and a^-antitrypsin. Thromb. Res.
7:245-249.
Dan<£, K. P.A. Andreasen, J. Grondahl-Hansen, P. Kristensen, L.S.
Neilsen, and L. Skriver. 1985. Plasminogen activators, tissue
degradation, and cancer. Adv. Cancer Res. 44:139-267.
Davies, M.C., M.E. Englert, and E.C. DeRenzo. 1964. Interaction of
streptokinase and human plasminogen. J. Biol. Chem. 239:2651-2656.
Del Rosso, M., G. Dini, and G. Fibbi. 1985. Receptors for plasminogen
activator, urokinase, in normal and Rous Sarcoma virus-transformed
mouse fibroblasts. Cancer Res. 45:630-636.
DesJardin, L., M.D.P. Boyle, and R. Lottenberg. 1988. Group A
streptococci grown in human plasma generate and bind plasmin to a
specific surface receptor. In: Proceedings of the 88th Annual
Meeting of the American Society for Microbiology, p. 57, B-169a,
Hartman, P.A., and Morello, J.A. (eds), American Society for
Microbiology, Washington, D.C..
Diano, M., A. Le Bivic, and M. Hirn. 1987. A method for the production
of highly specific polyclonal antibodies. Anal. Biochem.
166:224229.
Dillon, H.C., and L.W. Wannamaker. 1965. Physical and immunological
differences among streptokinases. J. Exp. Med. 121:351-371.
Edy, J., and D. Collen. 1979. The Interaction in human plasma of
antiplasmin, the fast-reacting plasmin inhibitor with plasmin,
thrombin, trypsin, and chymotrypsin. Biochim. Biophys. Acta
484:423-432.
Elliott, S.D., and V.P. Dole. 1947. An inactive precursor of
streptococcal proteinase. J. Exp. Med. 85:305-310.
Emonds-Alt, X., E. Quisquater, and G. Vaes. 1980. Proteoglycan- and
fibrin-degrading neutral proteinase activities of Lewis lung
carcinoma cells. Eur. J. Cancer 16:1257-1261.
Fritz, H., and G. Wunderer. 1983. Biochemistry and applications of
aprotinin, the kallikrein inhibitor from bovine organs. Arzneim.-
Forsch./Drug Res. 33:479-494.


155
The plasmin(ogen) molecule contains several characteristic lysine
binding sites (Alkjaersig, 1964; Abiko et al.. 1969; Brockway and
Castellino, 1972), one high affinity site and four to five sites of lower
affinity (Markus et al., 1978a; Markus et al., 1978b). The high affinity
site is located in the kringle 1 structure (see Figure 1-1), and one of
the lower affinity sites located in the kringle 4 region (Learch et al.,
1980). These lysine-binding sites participate in plasmin(ogen)
interactions with fibrin (Wiman et al., 1979; Swenson and Thorsen, 1981)
and to a^-antiplasmin (Wiman, 1981). Experiments documented in Chapter
Two show that the plasmin-bacterial receptor interaction is one of very
high affinity. Also, the plasmin-bacterial receptor interaction, like
the first step in the plasmin-a2*antiplasmin interaction and the
plasmin(ogen)- fibrin interaction, was reversibly inhibitable by L-lysine
and EACA. Furthermore, plasmin binding to bacterial plasmin receptor was
inhibited by a^-antiplasmin. Together these data suggest the possible
involvement of the high affinity lysine binding site in the plasmin-
bacterial receptor interaction.
The experiments presented in Chapter Three were designed to examine
the nature of the interaction of plasmin with the bacterial plasmin
receptor which would account for the particular observations described
above. Specific binding to the group A streptococcus, 64/14, was
demonstrated with plasmin's heavy chain. However, the isolated heavy
chain alone was not as efficient a competitor as intact Lys-plasmin. It
was noted that although heavy chain alone could completely inhibit Lys-
plasmin binding to bacterial plasmin receptor, in competition assays,
none of the isolated lysine-binding kringle domains alone or in
combination had any significant inhibitory effect. These data stressed


Results
Twenty hemolytic streptococcal isolates were grown overnight at
37C and tested for their ability to bind radiolabeled plasminogen,
plasmin, urokinase, or trypsin as described in the Methods Section. The
results (see Table 2-1) showed that all twenty group A isolates bound
plasmin but failed to bind significant quantities of plasminogen or any
of the other labeled proteins, i.e. less than 10% of the offered label.
Furthermore, the expression of plasmin binding ability was shown to be
present on bacteria grown in either Todd-Hewitt broth or chemically
defined media (data not shown). Plasmin binding was found to be
dependent on the concentration of bacteria and was maximal within two
minutes at 37C. Pre-incubation with excess unlabeled plasmin prevented
binding of the labeled plasmin. In the absence of unlabeled plasmin,
strain 64/14 consistently bound approximately 60% of the radioactive
plasmin offered and was used to analyze further the selective plasmin
binding activity. In my initial attempts to characterize the
differential binding of plasminogen and plasmin to a group A
streptococcus I compared the kinetics of generation of plasmin from
plasminogen with the ability of labeled protein to bind to the bacteria.
Conversion of plasminogen to plasmin occurs when a single arginine-valine
bond is split in the zymogen by action of the enzyme urokinase (Groskopf
et al. 1969). The zymogen activation reaction can be monitored on SDS-
polyacrylamide gels following reduction of disulfide bonds.
The zymogen molecule migrates as a single polypeptide chain with a
Mr of approximately 90,000 daltons. The active enzyme plasmin migrates
under these conditions as two distinct polypeptide chains (a heavy chain
with an Mr of approximately 60,000 daltons and a light chain with an Mr


20
of approximately 25,000 daltons). The activation reaction can be stopped
at any time by addition of a 10-fold molar excess of p-nitropbenyl p-
guanidinobenzoate (pNpGB). Consequently, it is possible to obtain
plasminogen in various stages of activation and compare the ability of
the labeled proteins to bind to a group A streptococcus. The results
presented in Figure 2-1, panel A demonstrate that the activation of
plasminogen to plasmin could be readily monitored. As the conversion of
plasminogen to plasmin proceeded, an increase in the binding of labeled
protein occurred which correlated with the concentration of plasmin in
the reaction mixture (Figure 2-1, panel B).
The conversion of plasminogen to plasmin yields a serine active site
that is not expressed in the zymogen. In the next series of experiments
the role of the active site in binding of the enzyme to the bacteria was
assessed. Plasmin was treated with the active site titrant pNpGB, the
small naturally occurring inhibitor aprotinin (Fritz and Wunderer, 1983),
the selective histidine modifying agent, phe-pro-arg chloromethyl ketone
[PPACK] (Kettner et al. 1980) and the physiological regulator 02
antiplasmin (a^-AP) (Mori and Aoki, 1976). The ability of the various
inhibited forms of plasmin to bind to a group A streptococcus was
measured. The results presented in Figure 2-2 demonstrate that plasmin
treated with pNpGB, aprotinin, or PPACK were all capable of binding to
the bacteria in the presence of -AP. By contrast unmodified plasmin
incubated with the physiological inhibitor, a2Ai>> failed to bind. Each
of the plasmin-inhibitor samples that had been incubated with excess 2"
AP was monitored on non-reducing SDS-polyacrylamide gels for the
formation of a high molecular weight complex. The high molecular weight
band observed in the third lane indicates the formation of a stable


Figure 4-4. Representative profile of an affinity purification of strain
64/14 mutanolvsin extracted nlasmin binding activity. One ml of a
mutanolysin extract of strain 64/14 was applied to a column of
immobilized enzymatically inactive plasmin. Bound material was eluted
with a gradient of L-lysine. All fractions were screened for plasmin
binding activity by dot-blot analysis. The resulting autoradiograph of
dot-blotted fractions (50 n 1) probed with ^-^^1-plasmin is shown for the
coresponding fractions below the X-axis.


55
determined in replicate tubes which contained no bacteria. All estimates
were performed in duplicate.
Inhibition of Plasmin Binding to Bacteria by Purified Plasmin(ogen')
Fragments
The ability of different concentrations of one or more of the
isolated plasminogen fragments to inhibit binding of Phe-Pro-Arg-
Chloromethylketone (PPACK) reacted -*-^^I-Lys-plasmin to the group A
streptococcus 64/14 was tested using a modification of the direct binding
assay described above. Different concentrations of plasmin(ogen) or
plasmin(ogen) fragments were mixed with a fixed dilution of a 10% w/v
suspension of streptococcal strain 64/14 and PPACK reacted -Lys-
plasmin (approx. 30,000 cpm per tube) followed by incubation for 30 min
at 37C. Bacterial associated radioactivity was determined after washing
away unbound label as described above. The inhibition of binding of
labeled plasmin was calculated by comparing the number of counts bound in
the absence of competitor with the number of counts bound when the
competitor was present. All samples were corrected for background
binding of counts. Counts bound in the tubes from which bacteria were
omitted or in tubes in which a 100-fold molar-excess of unlabeled ligand
was added. In no case was the background level of radioactivity greater
than 5% of the counts offered. Furthermore, background levels in the
presence of excess cold competitor, or in the absence of bacteria were
not significantly different.
Elution and Analysis of -Lvs-plasminCogen') from Bacteria
IOC
JI-Lys-plasminogen (approx. 100,000 cpm) was added to a 100 p 1
aliquot of a 10% w/v solution of strain 64/14 bacteria in a total volume
of 400 pi VBS-gel and allowed to incubate at 37C for 30 min. The
bacteria were then pelleted by centrifugation (3000 x g, 10 min) and


4
LBS II


Figure 4-6. SDS-PAGE and Western blot analysis of mutanolvsin extracted,
affinity purified plasmin binding activity. Parallel 10% SDS-poly-
acrylamide gels were electrophoresed. One gel was silver stained to
detect the distribution of proteins present (Panel A). The proteins on
the second gel were transfered to nitrocellulose by Western blotting and
probed with -plasmin according to Methods. Panel B shows the results
of the autoradiograph demonstrating functional activity. Panel A, lane
M: molecular weight markers; lane 1: 50 /I of strain 64/14 mutanolysin
extract; lane 2-4: three fractions containing lysine eluted plasmin
binding activity. Panel B, lane 1: 50 pi of strain 64/14 mutanolysin
extract; lane 3: 50 ^1 of the lysine eluted plasmin binding activity from
the fraction shown in lane 3 of Panel A.


166
Kaplan, M.H. 1944. Nature and role of the lytic factor in hemolytic
streptococcal fibrinolysis. Proc. Soc. Exp. Biol, and Med.
57:40-43.
Kay, A.B., D.S. Petter, and R. McKenzie. 1974. The identification of
fibrinopeptide B as a chemotactic agent derived from human
fibrinogen. Brit. J. Haematol. 27:669-677.
Kettner, C., C. Mirabelli, J.V. Pierce, and E. Shaw. 1980. Active site
mapping of human and rat urinary kallikreins by peptidyl
chloromethyl ketones. Arch. Biochem. Biophys. 202:420-430.
Knox, P., S. Crooks, M.C. Scaife, and S. Patel. 1987. Role of
plasminogen, plasmin, and plasminogen activators in the migration of
fibroblasts into plasma clots. J. Cell. Physiol. 132:501-508.
Knudsen, B.S., R.L. Silverstein, L.L.K. Leung, P.C. Harpel, and R.L.
Nachman. 1986. Binding of plasminogen to extracellular Matrix. J.
Biol. Chem. 261:10765-10771.
Kronvall, G. 1973. A surface component in group A, C, and G
streptococci with non-immune reactivity for immunoglobulin G. J.
Immunol. 111:1401-1406.
Kronvall, G., C. Schonbeck, and E. Myhre. 1979. Fibrinogen binding
structures in B-hemolytic streptococci group A,C, and G. Acta.
Path. Microbiol. Scand., Sect. B, 87:303-310.
Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly
of the head of bacteriophage T4. Nature 227:680-685.
Lambert, R.A., and F.M. Hanes. 1911. Characteristics of growth of
sarcoma and carcinoma cultivated .in vitro. J. Exp. Med. 13:495-504.
Lancefield, R.C. 1928. The antigenic complex of Streptococcus
haemolvticus. I. Demonstration of a type specific substance in
extracts of Streptococcus haemolvticus. J. Exp. Med. 47:91-103.
Laurell, C.B. 1966. Quantitative estimation of proteins by
electrophoresis in agarose gel containing antibodies. Anal.
Biochem. 15:45-52.
Lawler, J., and H. Slayter. 1981. The release of heparin binding
peptides from platelet thrombospondin by proteolytic action of
thrombin, plasmin and trypsin. Thromb. Res. 22:267-279.
Lerch, P.G., and E.G. Rickli. 1980. Studies on the chemical nature of
lysine-binding sites and their localization in human plasminogen.
Biochim. Biophys. Acta 625:374-378.
Lerch, P.G., E.G. Rickli, W. Lergier, and D. Gillessen. 1980.
Localization of individual lysine-binding regions in human
plasminogen and investigations on their complex-forming properties.
Eur. J. Biochem. 107:7-13.


Glu-
Figure 3-3. Binding of 125I labeled Glu- and Lvs-plasmin(ogens):
and Lys-plasmin(ogen) were generated as described in Materials and
Methods. The labeled tracers were then used in direct binding assays
with a fixed concentration of the streptococcal strain 64/14. Panel A
is an autoradiograph demonstrating the analysis of each reduced 125I-
labeled sample on a SDS-6 M-Urea 12%-polyacrylamide gel by auto
radiograph to verify their molecular form. (Glu-H: Glu-heavy chain;
Lys-H: Lys-heavy chain; L: light chain). Panel B illustrates the
percent of offered cpm bound to bacterial pellets. (Lane 1: Glu-PLG; 2
Lys-PLG; 3: Glu-PLA; 4: Lys-PLA).


6
activate plasminogen. Unlike the other plasminogen activators,
streptokinase has no enzymatic activity. The activation mechanism lies
in its ability to form a specific 1:1 stoichiometric complex with
plasminogen, as well as with plasmin, which leads to the generation of an
active enzyme moiety, presumably through conformational changes in the
plasminogen molecule without the cleavage event at the Arg5g0^al56^
peptide bond (Markus et al., 1976), that can in turn act as a plasminogen
activator for plasminogen molecules. This is a function neither of the
two proteins possesses alone.
Plasmin Regulation
Once generated, plasmin's activity is also carefully regulated under
normal physiological conditions. This regulation is accomplished by a
specific inhibitor of plasmin known as a2"anti-Plasn,i-n (a^-AP) (Aoki et
al.. 1977; Collen, 1976; Collen et al., 1975). Alpha2*antiplasmin is a
single chain glycoprotein with a molecular weight of approximately 70,000
daltons. 2-antiplasmin forms a 1:1 stoichiometric complex very rapidly
(estimated rate constant of k^=3xl0^M"'^) (Wiman and Collen, 1978) and
neutralizes plasmin's activity through a covalent association with the
serine residue in the active site of plasmin. A physiological role of
2*AP as an inhibitor of other proteases other than plasmin appears
negligible (Edy and Collen, 1979; Ohlsson and Collen, 1977).
Workers pioneering the techniques of tissue culturing noted that
explants of cancer tissue consistently caused proteolytic degradation,
liquefying the plasma clots on which they were grown (Carrel and Burrows,
1911; Lambert and Hanes, 1911). Since those early studies the
plasminogen-plasmin system, in addition to its role in fibrinolysis, has


163
Broeseker, T.A., M.D.P. Boyle, and R. Lottenberg. 1988.
Characterization of the interaction of human plasmin with its
specific receptor on a group A streptococcus. Microbial
Patho. 5:19-27.
Brogden, R.N., T.M. Speight, and G.S. Avery. 1973. Streptokinase: A
review of its clinical pharmacology, mechanism of action and
therapeutic uses. Drugs 5:357-445.
Burova, L., P. Christensen, R. Grubb, A. Johnson, G. Samuelsson, C.
Schalen, and M.L. Svenson. 1980. Changes in virulence, M protein
and IgG Fc receptor activity in a type 12 group A streptococcal
strain during mouse passage. Acta Path. Microbiol. Scand.
88:199-205.
Burova, L., P. Christensen, R. Grubb, I.S. Krasilnikov, G. Samuelsson, C.
Schalen, M.L. Svenson, and U. Zatterstrom. 1981. IgG-Fc receptors
in T-type 12 group A streptococci from clinical specimens: Absence
from M-type 12 and presence in M-type 22. Acta Path. Microbiol.
Scand. 89:433-435.
Carrel, A., and M.T. Burrows. 1911. Cultivation in vitro of malignant
tumors. J. Exp. Med. 13:571-575.
Castellino, F.J., Sodetz, J.M., Brockway, W.J., and Sefring, G.E., Jr.
1976. Streptokinase. Methods Enzymol. 45:244-257.
Cederholm-Williams, S.A. 1977. The binding of plasminogen (mol. wt.
84,000) and plasmin to fibrin. Thromb. Res. 11:421-423.
Cederholm-Williams, S.A., F. DeCock, H.R. Lijnen, and D. Collen. 1979.
Kinetics of the reactions between streptokinase, plasmin and 02-
antiplasmin. Eur. J. Biochem. 100:125-132.
Chiles, T.C., T.W. O'Brien, and M.S. Kilberg. 1987. Production of
monospecific antibodies to a low-abundance hepatic membrane protein
using nitrocellulose immobilized protein as antigen. Anal. Biochem.
163:136-142.
Christensen, L.R. 1945. Streptococcal fibrinolysis: Proteolytic
reaction due to serum enzyme activated by streptoccus fibrinolysin.
J. Gen. Physiol. 28:363-383.
Christensen, L.R., and C.M. Macleod. 1945. Proteolytic enzyme of serum:
Characterization, activation and reaction with inhibitors. J. Gen.
Physiol. 28:559-583.
Christensen, U., and I. Clemmenson. 1977. Kinetic properties of the
primary inhibitor of plasmin from human plasma. Biochem. J.
163:389-391.
Cleary, P., S. Jones, J. Robbins, and W. Simpson. 1987. Phase variation
and M-protein expression in group A streptococci. In: Streptococcal
Genetics, p.102, Ferretti, J.J., and R. Curtiss III (eds.), American
Society for Microbiology, Washington, D.C..


15
bacterial pellets were washed in the appropriate NaCl concentration to
remove unbound labeled proteins.
Effect of Ca++ and Mg-1"1- on Plasmin Binding
Binding of radiolabeled plasmin to group A streptococci strain 64/14
was studied in the following buffers: 1) VBS-gel containing 0.00015 M
Ca++ and 0.001 M Mg++, or 2) metal free VBS-gel containing 0.15 M EDTA.
In each case 400 pi of buffer were added to 100 pi of VBS-gel containing
approximately 107 bacteria and 100 pi of VBS-gel containing 3 x 10~* cpm
of radiolabeled plasmin. After incubation at 37 C for 15 minutes, the
mixtures were centrifuged at 1,000 x g for 7 minutes to separate bacteria
from unbound radiolabel, the pellets were washed twice with 2.0 ml of the
appropriate buffer, and radioactivity associated with the bacterial
pellet in duplicate tubes was measured.
Inhibition of Binding of Plasmin by Amino Acids
Labeled plasmin (100 pi containing approximately 2 x 10^ cpm) was
added to 200 pi VBS-gel containing varying concentrations of
epsilon-aminocaproic acid (EACA), lysine, or arginine, and incubated at
37 C for 15 minutes. The pH of each solution was 7.0. One hundred pi
of VBS-gel containing 107 bacteria (strain 64/14) were then added and the
mixture was incubated at 37C for 15 minutes. The bacterial suspensions
were centrifuged at 1,000 x g for 7 minutes and washed twice with 2.0 ml
of VBS-gel containing the same concentration of amino acid present during
the incubation period. The percent inhibition of binding was calculated
for duplicate samples by comparison with binding in VBS-gel alone.
The ability of EACA, lysine, or arginine to dissociate bound plasmin
from the bacteria was examined in the following manner. Labeled plasmin
was incubated with 107 bacteria in VBS-gel at 37 C for 15 minutes. The


141
12 3 4 5 6 7
I I 6-1
84
58
48.5
36.5-
26.6
I 2 3 4 5 6 7
B
12 3 4 5 6 7
58-
48.5
36.5-
26.6-
c


114
1 2 3456789


CHAPTER FIVE
COMPARISON OF THE GROUP A STREPTOCOCCAL RECEPTOR FOR
HUMAN PLASMIN WITH STREPTOKINASE
Introduction
In the previous chapter I have described the isolation of a specific
receptor for human plasmin from a group A streptococcus. This was
achieved by first solubilizing the receptor by treatment with the enzyme
mutanolysin followed by affinity purification on a column of immobilized
plasmin. The purified functionally active receptor material had a Mr of
approximately 41,000 daltons as measured by SDS-polyacrylamide gel
electrophoresis and lacked plasminogen activator activity. These findings
indicate that the plasmin receptor protein was not an intact
streptokinase molecule. The purpose of the studies presented in this
chapter were to perform a more complete comparison of the group A
streptococcal receptor for human plasmin and streptokinase with respect
to plasminogen activator activity; their binding specificities for
domains of the plasmin molecule; and examine possible antigenic
relatedness.
The results presented in this chapter demonstrate that the plasmin
receptor and streptokinase, while both produced by the same strain of
group A streptococci bacteria and having high affinity for plasmin, are
physicochemically, functionally, and antigenically distinct molecules.
122


22
3 6 9 12 15 24 36
TIME (MINUTES)


25
complex of plasmin with its physiological inhibitor a^'AP (Figure 2-2,
upper panel). Pretreatment of plasmin with aprotinin, pNpGB, or PPACK
inhibited the ability of the enzyme to form the covalent linkage
inhibition reaction with c*2-AP demonstrating that under the experimental
conditions used the plasmin active site was modified.
The next series of experiments were designed to determine whether
bacterial bound plasmin was capable of retaining its enzymatic activity.
Radiolabeled plasmin was generated and incubated with a suspension of
group A streptococci for 40 minutes at 37C. The bacteria with the
associated plasmin were recovered by centrifugation, washed twice with
buffer, and then tested for their ability to cleave the chromogenic
synthetic substrate H-D-val-leu-lys-paranitroanilide (as described in
the Methods). In these experiments a control of bacteria alone failed to
hydrolyze the chromogenic substrate, while bacteria pre-incubated with
plasmin were found to cleave the substrate efficiently. The ability of
bacterial bound plasmin to be affected by a variety of different
inhibitors was tested. The results in Figure 2-3 demonstrate that
addition of pNpGB, PPACK, or aprotinin to the bacterial bound enzyme was
capable of inhibiting its enzyme activity for the synthetic substrate.
By contrast, addition of -AP failed to reduce the enzyme activity
(Figure 2-3). All inhibitors were used in excess of that required to
totally inhibit an equivalent concentration of plasmin in the fluid
phase. Since c*2-AP failed to regulate the bacterial bound enzyme, one
might predict that the large molecule fibrin, the natural substrate of
plasmin, would also be prevented from occupying the substrate pocket in
the active site. To test this prediction, bacteria with plasmin bound
to their surface were placed on a fibrin plate and their ability to


KEY TO ABBREVIATIONS
c*2 -AP/a2 "PI
alpha-2-antiplasmin
ATCC
American typed culture collection
BSA
bovine serum albumin
cpm
counts per minute
DNase
deoxyribonuclease
EACA
epsilon aminocaproic acid
EDTA
ethylenediaminetetraacetic acid
ELISA
enzyme -linked-immunosorbant-assay
FPLC
fast performance liquid chromatography
g
gravity
Glu-plasminogen
native human plasminogen with NH2-terminal
glutamic acid
HC
heavy chain of plasmin
HRGP
histidine-rich glycoprotein
IgG
immunoglobulin class G
KD
kilodalton
kd
dissociation constant
KIU
kallikrein inhibitor unit
LBS
lysine-binding site
LC
light chain of plasmin
Lys-plasminogen
proteolytically modified form of Glu-plasminogen with
NH2-terminal lysine
M
molar
x


81
changes in the molecule (Violand et al., 1975). I have shown previously
that lysine or c*2-AP inhibit the binding of plasmin to the group A
streptococcal receptor, as documented in Chapter Two, indicating the
possible involvement of the high affinity lysine-binding site in the
plasmin-bacterial receptor interaction. A comparison of my findings
with studies of the interaction of plasminogen with other naturally
occurring plasminogen binding proteins reveals a number of interesting
similarities and contrasts. Specific binding to the group A
streptococcus, 64/14, was demonstrated with plasmin's heavy chain.
However, the isolated heavy chain alone was not as efficient a competitor
as intact Lys-plasmin, as evidenced by the non-superimposible nature of
the heavy chain and Lys-plasmin inhibition curves (Figure 3-2). It
should be noted that 100% inhibition of binding of Lys-plasmin could be
achieved by addition of high concentrations of heavy chain, but none of
the kringle containing fragments (Lysine-binding domains) alone or in
combination had any significant inhibitory effects at similar molar
concentrations. This finding stresses the importance of the
conformation of the entire heavy chain for binding to bacteria. The
bacterial binding of plasmin therefore differs from the kind of
interaction seen with a^-AP, as well as with fibrin and fibrinogen, to
which plasmin as well as plasminogen, LBS-I, LBS-II, and mini-
plasmin(ogen) are known to interact (Swenson and Thorsen, 1981; Thorsen
et al., 1981; Wiman et al., 1979).
Consistent with my initial observations, documented in Chapter Two,
there is no significant binding of the native zymogen, Glu-plasminogen,
while the conformationally altered form of the zymogen, Lys-plasminogen,
was found to bind specifically to bacteria (Figures 3-3,3-4, and 3-5).


1 PR
Figure 3-7. Characterization of I-Lvs-plasminCogen') species eluted
from bacteria: Eluted labeled proteins were analyzed by electrophoresis
on a 10%-SDS-PAGE-6M urea gel under reduced conditions. Lanes 1, 2, and
3 contain labeled proteins from bacteria pre-incubated with 125I-Lys-
plasminogen and eluted by 0.5% SDS; 0.1 M EACA; or 0.5% SDS containing
2.0% £-mercaptoethanol respectively. Lanes 4, 5, and 6 are identical to
Lanes 1, 2, and 3 with the exception that the bound labeled proteins
were pre-incubated with urokinase prior to elution. Lane 7 contains
125I-Lys-plasminogen incubated at 37C without bacteria for the period
of the experiments and Lanes 8 and 9 contain 125I-Lys-plasminogen and
125-Lys-plasmin respectively. For precise experimental details see
Methods.