Eur. J. Biochem.

229, 83-90 (1995)

0 FEBS 1995

Structural domains of streptokinase involved in the interaction

with plasminogen
Pedro RODRfGUEZ'.2, Pablo FUEh?TES2,Mario BARR02, Julio G. ALVAREZZ, Emilio MWOZZ, DCsir6 COLLEN' and H. Roger LUNEN'
' Center for Molecular and Vascular Biology, University of Leuven, Belgium
Center for Genetic Engineering and Biotechnology, Thrombolytics Division, Havana, Cuba

(Received 8 December 1994/27 January 1995) - EJB 94 1897/4

Two fragments of recombinant streptokinase, comprising amino acids Val143 -Lys293 (17-kDa rSK)
or Vd143-Lys386 (26-kDa rSK), were cloned and expressed in Escherichia coli, purified to homogeneity
and their interactions with plasmin(ogen) were evaluated. Both 17-kDa rSK and 26-kDa rSK bound
to plasminogen with a 1:1 stoichiometry and with affinity constants of 3.0X108M-' and 12XlO'M-',
respectively, as compared to 6.3 X 10' M-' for the binding of intact recombinant streptokinase to plasmin-
ogen. Binding of 17-kDa rSK to plasminogen-Sepharose was displaced by addition of increasing concen-
trations of recombinant streptokinase, whereas bound recombinant streptokinase was not displayed by
17-kDa rSK. In equimolar mixtures of plasminogen and 26-kDa rSK, the appearance of amidolytic activ-
ity as monitored with a chromogenic substrate, was significantly delayed compared to the equimolar
mixture with recombinant streptokinase (60% of the maximal activity after 30 min, compared to maxi-
mum activity within s 2 min). In contrast, no amidolytic activity was generated in equimolar mixtures of
plasminogen and 17-kDa rSK. Plasminogen was rapidly activated by catalytic amounts (1:100 molar
ratio) of recombinant streptokinase (60-70% within 10-15 min), whereas only 4% of the plasminogen
was activated within 60 min with 26-kDa rSK, and no plasmin was generated with 17-kDa rSK. Com-
plexes of plasmin with 17-kDa rSK or with 26-kDa rSK were very rapidly inhibited by a,-antiplasmin
(apparent second-order inhibition rate constant of approximately 2X lo7M-' s-I), whereas the complex
with recombinant streptokinase was resistant to inhibition. With 26-kDa rSK, inhibition by a,-antiplasmin
resulted in dissociation of the complexes and recycling of functionally active 26-kDa rSK to other plas-
minogen molecules ; 17-kDa rSK, in contrast, remained associated with the plasmin -a,-antiplasmin com-
plex.
These findings suggest that different regions of the streptokinase molecule are involved in binding to
plasminogen, in active-site exposure, and in impairment of the inhibition of plasmin by a,-antiplasmin.
Thus, the 17-kDa region spanning Val143-Lys293 in streptokinase mediates its binding to plasminogen
but does not induce activation. Furthermore, this region does not interfere with the inhibition of the
complex with plasmin by a,-antiplasmin.
Keywords. Streptokinase ; plasminogen ; a,-antiplasmin.

Streptokinase converts plasminogen from human and several
other mammalian species into plasmin. This activation involves
formation of a 1:1 stoichiometric complex between plasminogen
and streptokinase, followed by exposure of an active-site in the
plasminogen moiety [l -31. Both the plasminogen and streptoki-
nase moieties of the complex are subsequently modified; plas-
minogen is converted to plasmin, whereas different degraded
forms of streptokinase lacking N-terminal and/or C-terminal
Correspondence to H. R. Lijnen, Center for Molecular and Vascular
Biology, University of Leuven, 0 & N, Herestraat 49, B-3000 Leuven,
Belgium
Fax: +32 16 345990.
Abbreviations. 17-kDa rSK, fragment of recombinant streptokinase
consisting of residues Va1143-Lys293 ;26-kDa rSK, fragment of recom-
binant streptokinase consisting of residues Va1143-Lys386; -CH,Cl,
chloromethane; D-ValPheLysCH,-plasmin, human plasmin inactivated
with D-ValPheLysCH2C1; D-GlpPheLys-NH-Np, D-pyroglutamyl-phe-
nylalanyl-lysine-p-nitroanilide;NpGdnPhCOOH, p-nitrophenyl-p'-gua-
nidinobenzoate; k.,,, association Tate constant ;k,,., dissociation rate con-
stant; K,,association affinity constant.
Enzyme. Plasmin (EC 3.4.21.7).
portions have been isolated from complexes with human, rabbit
and dog plasmin [4-61. Furthermore, modified forms have been
obtained following chemical or enzymic cleavage of native
streptokinase [7] or upon expression of streptokinase in Strepto-
coccus sanguis [8] and in Escherichia coli [9].
Studies with such streptokinase fragments suggested that the
N-terminus of streptokinase (up to Ser60) may be an indepen-
dent domain with functions which are unrelated to binding or
activation of plasminogen. This is substantiated by the findings
that recombinant fusion proteins in which the 13 N-terminal
amino acids were replaced with fragments of unrelated proteins
had intact plasminogen activator activity [lo], that chemical
cleavage of the 15 N-terminal residues of mature streptokinase
does not result in loss of activity [ll], and that the streptokinase
region comprising residues Va120-Pro58 shares a common epi-
tope with fibronectin [12]. Furthermore, a 36-kDa streptokinase
fragment with N-terminal Ser60 (and the C-terminus presumably
as Lys386) and a 26-kDa fragment with N-terminus Ser60 (and
C-terminus presumably Lys293) were found to form a complex
with plasminogen, with normal plasminogen activator activity
for the 36-kDa fragment [4], but with impaired activity for the
84 Rodriguez et al. ( E m J. Biochem. 229)

26-kDa fragment [5]. Monoclonal antibodies directed against the characterized as described elsewhere [20-221. Plasmin was in-
region 1-253 of streptokinase were shown to inhibit the binding activated by alkylation with D-valyl-phenylalanyl-lysine-
of streptokinase to plasminogen but not to affect the catalytic chloromethane (D-ValPheLysCH2C1; 2 100-fold molar excess)
efficiency of a plasmin(ogen)-streptokinase complex, whereas and complete inhibition was confirmed by measurement of
non-overlapping monoclonal antibodies, directed against the re- residual activity with D-pyroglutamyl-phenylalanyl-lysine-p-ni-
gion 120-352, inhibited both the binding of streptokinase to troanilide (D-GlpPheLys-NH-Np). Plasminogen was coupled to
plasminogen and the activity of the complex [13]. These results CNBr-activated Sepharose 4B (Pharmacia) by incubation in
are in agreement with previous findings that the streptokinase 0.1 M sodium phosphate, pH 6.0, containing 1 M NaCl for 2 h
region Val143-Lys293 contains a structure involved in binding at room temperature; this yielded coupling of 5 mg plasmino-
to plasminogen [14]. The C-terminal fragment of streptokinase gedml swollen gel. Recombinant Glu-plasminogen with the
(Leu294-Lys386) was previously shown to play an important active-site Ser741 replaced by Ala ( [S741A]-recombinant plas-
role in plasminogen activation [5, 8, 15, 161. minogen) was obtained and characterized as described elsewhere
In the present study, two recombinant streptokinase frag- [23, 241. The chromogenic substrate for plasmin, D-GlpPheLys-
ments, comprising residues Val143 -Lys293 (17-kDa rSK) and NH-Np, was purchased from Chromogenix, p-nitrophenyl-p'-
Va1143-Lys386 (26-kDa rSK) were used to analyze structure/ guanidinobenzoate (NpGdnPhCOOH) from E. Merck and the
function relationships of streptokinase with respect to the plasmin inhibitor D-ValPheLysCHZC1was custom synthesized at
domains involved in binding to and activation of human plas- UCB .
minogen. 17-kDa rSK was labeled with '''I using the lactoperoxidase
method [25], to a specific radioactivity of O.7X1O6c p d p g .
Determination of affinity constants and the binding stoi-
MATERIALS AND METHODS chiometry of streptokinase moieties to plasmin(ogen). Asso-
ciation rate constants (kass)and dissociation rate constants (kdlSc)
Cloning and expression of 17-kDa rSK and 26-kDa rSK. were determined by real-time biospecific interaction analysis
Two recombinant streptokinase fragments, comprising amino using the BIAcoreTMinstrument (Pharmacia, Biosensor AB),
acids Va1143-Lys293 (17-kDa rSK) and Va1143-Lys386 (26- essentially as described elsewhere [26]. The plasmin(ogen)
kDa rSK) were obtained from the cytosol fraction of E. coli cells moieties (the ligand) were immobilized onto the sensor chip
transformed with an expression vector pTrp [17], containing the using protein solutions at 10 pg/ml in 10 mM sodium acetate,
corresponding streptokinase gene fragments. pH 5.0, at a flow rate of 5 pVmin for 6 min. The streptokinase
These fragments, comprising nucleotides 427-879 or 427- moieties (the analyte) were injected at different concentration
1158 of the previously cloned streptokinase gene (rSK C-2) [I71 (100-SO0 nM for streptokinase, recombinant streptokinase and
were obtained using PCR. The 5' amplification primers for both 26-kDa rSK and 100-1000nM for 17-kDa rSK) in 10mM
fragments introduced an EcoRI restriction site (italics) and an Hepes, 3.4 mM EDTA, 0.15 M NaCl and 0.005% surfactant
ATG codon (in bold-faced print) for initiation of translation (5'- P20, pH 7.2, at a flow rate of 5 pvmin for 6 min (association
GAGAATTCATGGTTAGACCATATAAAG-3'). The 3' amplifi- phase), followed by buffer also at a flow rate of 5 pl/min, for
cation primers introduced a BamHI restriction site (italics) in 6-30 rnin (dissociation phase). Values for k,, and kdlsJwere de-
both sequences (5'-ATGGATCCTTATTTCAAGTGACTGC-3' rived from the sensorgrams, apparent affinity constants (K,)
for 17-kDa rSK and 5'-GTGGATCCTTATTTATCATAGGCTA- were calculated as kaSJ/kdlSS, and the binding stoichiometry was
3' for 26-kDa rSK), allowing oriented cloning of the amplified determined as described elsewhere [26].
fragments in pTrp. PCR was carried out using 1 pg recombinant Competition between different streptokinase moieties for
streptokinase C-2 plasmid and 100 pM of each oligoprimer in binding to plasminfogen). Competition between different strep-
100 pl 0.1 M Tris/HCl, pH 8.8, containing 10 mM NaC1, 10 mM tokinase moieties for binding sites on plasminogen was eval-
MgC1, and 100mM dithiothreitol at 100°C for 5 min. 2 U uated by consecutive binding experiments. Therefore, 2.5-pl
Thermus aquaticus DNA polymerase (Perkin Elmer-Cetus) and samples of plasminogen-Sepharose (5 mg plasminogedml gel)
dNTPs to a final concentration of 200 pM each were added and in 20mM Tris/HCl, pH7.4, containing 150 mM NaCl and
the samples were subjected to 30 PCR cycles. For 17-kDa rSK, 0.2 mM NpGdnPhCOOH, were incubated with excess recombi-
the first cycle consisted of 1 min denaturation at 95"C, 45 s at nant streptokinase (10 pg/2.5 p1 gel, corresponding to four times
52°C and 1 min at 75°C; the other cycles were performed in the maximal binding capacity of the gel) for 20min at 4°C.
the same way but with annealing at 70°C. For 26-kDa rSK, the After extensive washing with the same buffer to remove un-
first cycle was carried out for 1 min at 95"C, 45 s at 50°C and bound recombinant streptokinase, the gel was incubated with
1 rnin at 75°C; the other cycles were performed for 1 min at increasing concentrations of 17-kDa rSK (0-4 pg in a total vol-
95"C, 45 s at 68°C and 1 min at 75°C. After amplification, the ume of 5 pl) for 35-40 min at 4°C. After extensive washing,
fragments (474 bp and 753 bp, respectively) were concentrated the gel was eluted in 5 pl 1% SDS sample buffer by heating at
by ethanol precipitation, digested with EcoRI and BarnHI, ex- 100°C for 40 s. 1-p1 samples of the eluate were subjected to
tracted from low-melting-point agarose gel (Sigma) and cloned SDS/PAGE (Phast SystemTM,Pharmacia) using 8 -25 % gradient
into the expression vector pTrp, to generate the plasmid pEKG3 gels under non-reducing conditions. Alternatively, the plasmino-
which was used for transformation of E. coli K-12 cells (strain gen-Sepharose was first saturated with 17-kDa rSK (4 pg for
W3110) [17]. The sequences of the 17-kDa rSK and 26-kDa 2.5 pl gel, corresponding to five times the maximal binding
rSK fragments were confirmed by DNA sequencing [18]. capacity of the gel) and subsequent binding of recombinant
Proteins and reagents. Natural streptokinase was StreptaseR streptokinase (0-10 pg) was evaluated as described above.
obtained from Hoechst. Streptokinase devoid of albumin was The different molecular forms of streptokinase in the eluates
purchased from Boehringer Mannheim. Recombinant streptoki- were quantified by densitometric scanning of SDSPAGE.
nase, expressed in E. coli, was obtained as described elsewhere Complex formation of streptokinase moieties with plas-
[141. Protein concentrations were determined according to Brad- minogen. The generation of an active-site in complexes of Glu-
ford [19]. plasminogen with streptokinase moieties was monitored as fol-
Human Glu-plasminogen (92-kDa native plasminogen with lows. Plasminogen (final concentration 1.O pM) was incubated
N-terminal Glu), plasmin and a,-antiplasmin were prepared and with streptokinase, recombinant streptokinase, 17-kDa rSK or
 Rodriguez et al. (Eur J. Biochem. 229)
26-kDa rSK (final concentration 1.0 pM) at 37°C in 0.05 M
sodium phosphate, pH 7.4, containing 0.01 % Tween 80. At dif-
ferent time intervals (0-30 min), generation of an active-site in
the plasmin(ogen)-streptokinase complexes was measured with
D-GlpPheLys-NH-Np (final concentration 0.3 mM) after 50-fold
dilution of samples in 0.05 M Tris/HCl buffer, pH 7.4, contain-
ing 38 mM NaCl and 0.01 % Tween 80. The molecular form of
the plasminogen and streptokinase moieties in the complexes
was monitored by SDSPAGE using 8-25% gradient gels
without reduction or after reduction of samples by heating at
100°C for 3 rnin in the presence of 1% SDS and 1% dithioer-
ythritol.
Activation of plasminogen by catalytic amounts of strep-
tokinase moieties. Activation of Glu-plasminogen (final con-
centration 10 pM) with streptokinase, recombinant streptoki-
nase, 17-kDa rSK or 26-kDa rSK (final concentration 100 nM)
was monitored at 37°C in 0.05 M sodium phosphate, pH 7.4,
containing 0.01% Tween 80. Therefore, at different time
intervals (0-95 rnin), generated plasmin was measured with D-
GlpPheLys-NH-Np (final concentration 0.3 mM) after 50-500-
fold dilution of samples. The molecular forms of the plasmino-
gen and streptokinase moieties were monitored by SDSPAGE,
as described above.
Inhibition of plasmin-streptokinasecomplexes by a,-anti-
plasmin. Complexes between plasmin and recombinant strepto-
kinase or 17-kDa rSK were formed by incubation of plasmin
(final concentration 100 nM) with the streptokinase moiety
(equimolar to 50-fold molar excess) for 1 rnin at 0°C in 0.05 M
sodium phosphate, pH 7.4, containing 0.01 % Tween 80. Inhibi-
tion of the plasmin -recombinant-streptokinase or plasmin-
17 kDa-rSK complex (final concentration 8 nM) by a,-antiplas-
min (final concentration 10 nM) was monitored continuously
in the presence of D-GlpPheLys-NH-Np (final concentration
0.3 mM) as described previously [27]. Residual complex was
determined at different time intervals (0-3 min) and the
apparent second-order rate constant was calculated as described
~71.
In view of the previously observed instability of 26-kDa rSK
in a complex with plasmin, alternative experiments were per-
formed to evaluate inhibition by a,-antiplasmin. a,-antiplasmin
(final concentration 120 nM) was mixed with increasing con-
centrations of recombinant streptokinase or 26-kDa rSK (0-
800 nM) in 0.05 M sodium phosphate, pH 7.4, containing 0.01 %
Tween 80 at O"C, and plasmin was added to a final concentra-
tion of 100 nM. After incubation for exactly 2 min, residual
plasmin activity was measured with D-GlpPheLys-NH-Np (final
concentration 0.3 mM) after 50-fold dilution, as described
above.
In additional experiments, equimolar mixtures of plasmin
with 26-kDa rSK (final concentration 2 pM) were treated with
a,-antiplasmin (20 % molar excess), and immediately adsorbed
to Lys-Sepharose (10 pl swollen geVlO p1 sample) in NaCl/P,
(0.05 M sodium phosphate, pH 7.4, with 0.15 M NaC1) contain-
ing 0.01 % Tween 80 at 4°C for 35 min. Bound protein was
eluted with 0.1 M TrisLHCl, pH 8.4, containing 2% SDS at
100°C for 2 min. The distribution of streptokinase moieties
between the breakthrough and the eluate was monitored by SDS/
PAGE, as described above. The relative amounts of streptoki-
nase moieties in the starting materials, breakthroughs and eluates
were determined by densitometric scanning.
To further investigate these molecular interactions, plasmin
(final concentration 0.60 pM) was incubated for 2.5 min at 0°C
in 0.05 M sodium phosphate, pH 7.4, containing 0.01 % Tween
80 with 26-kDa rSK (final concentration 0.50 pM) before addi-
tion of a,-antiplasmin (final concentration 0.70 pM). Complete
inhibition within 2 rnin was confirmed with D-GlpPheLys-NH-
85
Fig. 1. SDS/polyacrylamide gel electrophoresis (8-25 % gradient
gels) under reducing conditions. Lane 2, streptokinase; lane 3, recom-
binant streptokinase; lane 4, 17-kDa rSK; lane 5, 26-kDa rSK; lanes 1
and 6, protein calibration mixture containing phosphorylase b (97 ma),
albumin (67 ma), ovalbumin (45 ma), carbonic anhydrase (30 ma),
trypsin inhibitor (20.1 kDa) and a-lactalbumin (14.4 ma).
Np, as described above. The mixture was then diluted fivefold in
a plasminogen solution (final concentration 10 pM) and plasmin
activity was monitored as a function of time (0-95 min) after
50-fold dilution of samples, as described above. Control experi-
ments under the same conditions were performed with uncom-
plexed 26-kDa rSK, and with a mixture of D-ValPheLysCH,-
plasmin, 26-kDa rSK and a,-antiplasmin at the same final con-
centrations. At the end of the experiments, the molecular form
of the plasmin(ogen) and streptokinase moieties was analyzed
by SDSPAGE, as described above.
The molecular interactions between plasmin, 17-kDa rSK
and a,-antiplasmin were evaluated by gel-filtration chromatogra-
phy. Therefore, 200 pl samples of 17-kDa rSK at a protein con-
centration of 20 pM, containing approximately 50000 cpm T -
17-kDa rSK, were applied to a 30 cmX1.0 cm Superdexm 200
HR column equilibrated in 20 mM TrisMCl, pH 8.0, containing
0.15 M NaCl and 0.01% Tween 80 at 4"C, at a flow rate of
45 ml/h. Equimolar mixtures of 17-kDa rSK and D-ValPhe-
LysCH,-plasmin, and of 17-kDa rSK and plasmin with the addi-
tion of a,-antiplasmin (20% excess over plasmin) were also gel
filtered under the same conditions. The absorbance at 280 nm
was monitored continuously and fractions of 350 pl were
collected for measurement of radioactivity. The molecular
composition of the peak fractions was analyzed by SDSPAGE
under non-reducing conditions.
RESULTS
Purification and physicochemical characterization of 17-kDa
rSK and 26-kDa rSK. After sonification of E. coli cells trans-
formed with the expression vector containing the 474-bp or
753-bp DNA fragments, 17-kDa rSK and 26-kDa rSK were de-
tected in the soluble cytosol fraction. Densitometric scanning of
SDSPAGE (not shown) revealed that 17-kDa rSK and 26-kDa
rSK were expressed to levels of approximately 7% and 15% of
the total protein. The proteins were purified from 1-g batches of
total protein, by chromatography on DEAE-Sepharose fast-flow
(Pharmacia AB; 4cmX4 cm column) in 0.02 M Tris/HCl,
pH 8.0, at a flow rate of 1300 ml/h, at 4"C, and eluted with a
linear gradient of 0-0.4M NaCi. The recovery was approxi-
mately 40 % for both recombinant streptokinase moieties. In a
second step, the proteins were adsorbed onto octyl-Sepharose
(4 cmX4 cm column) in 0.02 M Tris/HCl, pH 8.0, containing
1.0 M (NH4),S04 at a flow rate of 630 mVh at 4°C. Under these
conditions, 17-kDa rSK did not bind to the column and was
recovered to approximately 50% in the flow-through. 26-kDa
rSK bound and was eluted with a linear gradient of 1.0-0.5 M
(NH,),SO, in the same buffer, with a recovery of 15-20%.
Purified 17-kDa rSK and 26-kDa rSK were obtained in final
yields of 14 mg/g cytosol and 10 mg/g cytosol, respectively.
86 Rodriguez et al. (Eul: J. Biochem. 229)

Table 1. Apparent affinity constants of different streptokinase moieties for binding to [S741A]recombinant plasminogen and D-ValPhe-
LysCH,-plasmin. Values given are mean 2 SD of three determinations.

Analyte [S741A]Recombinant plasminogen D-ValPheLysCH,-plasmin

M-1 . s-1 s-' M-' M-1. s-1 S-' M-'

Streptokinase 120 t 9.8 0.36 2 0.03 3.3 2 0.5 220 t 8.4 0.72 2 0.05 3.2 t 0.30
Recombinant streptokinase 280 t 13 0.46 2 0.02 6.3 -t 0.5 320 2 45 0.29 2 0.02 11 2 1.6
17-kDa rSK 58 2 3.1 0.20 2 0.02 3.0 ? 0.2 33 2 3.0 0.34 ? 0.08 1.0 2 0.2
26-kDa rSK 870 2 28 0.77 ? 0.10 12 ? 1.6 440 2 56 1.6 2 0.04 2.7 2 0.3

rSK I 17-kDa rSK (mol I mol) 17-kDa rSK I rSK (mol I mol)
Fig. 2. Consecutive binding of recombinant streptokinase and 17-kDa rSK to plasminogen-Sepharose in the presence of excess
NpGdnPhCOOH. 17-kDa rSK (A) or recombinant streptokinase (B) were insolubilized onto plasminogen-Sepharose before addition of increasing
concentrations of free recombinant streptokinase or 17-kDa rSK, respectively. The data were obtained by densitometric scanning of non-reduced
SDSPAGE using samples eluted from the gel. Binding of 17-kDa rSK ( 0 )or of recombinant streptokinase (V)(in % of the maximal binding
capacity of the gel) is plotted versus the molar excess of added over insolubilized streptokinase moiety. The inset in (A) shows the binding of
17-kDa rSK without competing ligand (lane 1). and after addition of up to a fourfold molar excess of recombinant streptokinase (lanes 3-7); the
inset in (B) shows the binding of recombinant streptokinase without competing ligand (lane 1) and after addition of a fivefold molar excess of
17-kDa rSK (lane 2). Lane 2 in (A) and lane 3 in (B) represent a protein calibration mixture, as described in Fig. 1.

Both proteins appeared homogeneous on SDS/PAGE under non- The affinities of recombinant streptokinase and of the iso-
reducing conditions and migrated, under reducing conditions lated 26-kDa rSK and 17-kDa rSK fragments for both [S741A]-
with an apparent M, of 17 kDa and 26 kDa, respectively, com- recombinant plasminogen and D-ValPheLysCH,-plasmin were
pared to 47 kDa for intact recombinant streptokinase (Fig. 1). similar or somewhat higher than that of natural streptokinase.
Immunoblotting onto nitrocellulose sheets confirmed the reac- Values for kdlSsranged over 0.20-0.77X SK', corresponding
tivity of 17-kDa rSK and 26-kDa rSK with a polyclonal rabbit to half-lifes of the complexes of 15-58 min, with the exception
antiserum raised against intact recombinant streptokinase (data of the complex between 26-kDa rSK and active-site-blocked
not shown). plasmin, which had a half-life of only 7 min. The K, values of
N-terminal amino acid sequence analysis (Applied Bio- 1.O-12X1O8 M-' correspond to dissociation constants of 1.0-
systems 477A protein sequencer with identification of amino 10 nM.
acids by HPLC) yielded the following homogenous se- Analysis of the stoichometry of these interactions indicated
quences (295%), with yields in pmol in parenthesis: 17-kDa that both recombinant streptokinase fragments as well as intact
rSK, Va1(466)-Arg(278)-Pro(143)-Tyr(226)-Lys(272)-Glu(203)- natural and recombinant streptokinase bind to plasmin(ogen)
Lys(337); 26-kDa rSK, Val(l7)-Arg(l7)-Pro(22)-Tyr(33)- in a 1:l stoichiometry. The values obtained for binding to
Lys(24)-Glu(23)-Lys(31).For recombinant streptokinase two [S741A]recombinant plasminogen and to D-ValPheLysCH,-plas-
sequences were obtained, approximately 70 % corresponding to min were (mean+SD; n = 6) 1.250.22, 0.96?0.13, 1.150.26
Met(1298)-Ile(l399)-Ala(1497)-Gly(585)-Pro(467) and 30 % in and 0.94 2 0.10 moVmol for streptokinase, recombinant strepto-
which the N-terminal Met residue was removed during bacterial kinase, 17-kDa rSK and 26-kDa rSK, respectively.
processing [Ile(652)-Ala(686)-Gly(283)-Pro(183)-Glu(l99)].
Competition between different streptokinase moieties for
Determination of affinity constants and the binding stoichi- binding to plasmin(ogen). Consecutive binding experiments for
ometry of streptokinase moieties to plasmin(ogen). Table 1 recombinant streptokinase followed by 17-kDa rSK, or of
shows k,,,, kdlsSand K, for binding of different streptokinase 17-kDa rSK followed by recombinant streptokinase, to plasmin-
species to plasmin(ogen). ogen-Sepharose in the presence of excess NpGdnPhCOOH re-
 Rodriguez et al. (Eul: J. Biochem. 229)
600 I
0 10 20 30 40
TIME (min)
Fig. 3. Generation of active sites in equimolar mixtures (1 pM each)
of plasminogen with Streptokinase (A), recombinant streptokinase
(V),17-kDa rSK (0)and 26-kDa rSK (W).The amidolytic activity
measured with D-GlpPheLys-NH-Np after 50-fold dilution of samples is
plotted as a function of time. The data represent mean 2 SD of three
experiments. The insets show SDSPAGE under reducing conditions of
samples taken from the incubation mixtures. In (A), samples were taken
from the incubation mixtures at 4 m i n with streptokinase (lane 2) or
recombinant streptokinase (lane 3) and at 30 min from the incubation
mixtures with 26-kDa rSK (lane 4) or 17-kDa rSK (lane 1);(B) of the
inset shows the time course of degradation of 26-kDa rSK. Samples
were taken before addition of 26-kDa rSK (lane 5 ) , at 30 s (lane 4),
1 min (lane 3) and 2 rnin (lane 2). Lane 5 in (A) and lane 1 in (B)
represent a protein calibration mixture, as described in Fig. 1. Plg, plas-
minogen ; plasmin A, plasmin A-chain; plasmin B, plasmin-B-chain.
vealed that 17-kDa rSK was displaced from plasminogen by ad-
dition of increasing concentrations of recombinant streptokinase,
whereas bound recombinant streptokinase was not displaced by
addition of increasing concentrations of 17-kDa rSK. No detect-
able amounts of 17-kDa were bound to plasminogen-Sepharose
which was previously saturated with recombinant streptokinase
(Fig. 2).
Complex formation of streptokinase moieties with plasmino-
gen. Fig. 3 shows that, in equimolar mixtures of plasminogen
and streptokinase or recombinant streptokinase, the active site,
as monitored with a chromogenic substrate, was rapidly ex-
posed. In contrast, no active-site could be detected in equimolar
mixtures of plasminogen and 17-kDa rSK. With 26-kDa rSK,
the initial rate of complex formation appeared to be significantly
lower with approximately 60% of the maximal amidolytic
activity generated after 30 min.
SDSPAGE under reducing conditions (Fig. 3A) revealed
quantitative conversion of plasminogen to plasmin after 4 min
in the incubation mixtures with streptokinase (Fig. 3A lane 2),
and recombinant streptokinase (Fig. 3A lane 3). In the incuba-
tion mixtures with 17-kDa rSK, no conversion of plasminogen
to plasmin or degradation of the recombinant streptokinase frag-
ment was observed up to 30 rnin (Fig, 3A lane 1). In the incuba-
tion mixture with 26-kDa rSK, the recombinant streptokinase
fragment was completely degraded after 30 min (Fig. 3A lane
4) and partial conversion of plasminogen to plasmin occurred.
The observed amidolytic activity probably corresponds partly to
87
*Oo0 I
9- 6ooo
'R
z
0
i+
f 4000
Y
8
z
5
5n 2000
l!LGEzL
0
0 20 40 60
TIME (min)
Fig. 4. Activation of plasminogen (final concentration 10 pM), as a
function of time, by streptokinase (A), recombinant streptokinase
(V),17-kDa rSK (0)and 26-kDa rSK (W) (final concentration
100 nM each). The data represent meanfSD of three experiments.
The inset shows SDSRAGE under reducing conditions of samples taken
from the incubation mixtures at 10 rnin with streptokinase (lane 2) or
recombinant streptokinase (lane 3) and at 60 min from the incubation
mixtures with 26-kDa rSK (lane 4) or 17-kDa rSK (lane 5). Lane 1
represents plasminogen and lane 6 shows a protein calibration mixture,
as described in Fig. 1. Plg, plasminogen; plasmin A, plasmin A chain;
plasmin B, plasmin B chain.
plasmin generated through activation of plasminogen by traces
of active plasmin- 26-kDa rSK complex. Densitometric scan-
ning revealed a residual plasminogen concentration of only
20 %, suggesting that some degradation of generated plasmin
had occurred. The time course for degradation of 26-kDa rSK,
(Fig. 3B) shows that 26-kDa rSK remained intact at 30 s
(Fig. 3 B lane 4), but was rapidly degraded at 1 min (lane 3) and
2 min (lane 2).
Activation of plasminogen by catalytic amounts of streptoki-
nase moieties. Fig. 4 shows rapid time-dependent activation of
plasminogen by streptokinase and by recombinant streptokinase,
as monitored by quantification of generated plasmin with D-
GlpPheLys-NH-Np. Under the conditions used, approximately
60-70% of the plasminogen was activated within 10-15 min.
The absence of quantitative recovery of plasmin activity is due
to the relative instability of plasmin at 37°C. With 17-kDa rSK,
no plasmin activity or conversion of plasminogen was detectable
after 60min. Catalytic amounts of a complex of plasmin with
17-kDa rSK were also unable to activate plasminogen (data not
shown). With 26-kDa rSK, very slow plasminogen activation
was observed, corresponding to approximately 4% of the maxi-
mal activity at 60-90 min, whereas 230% of the plasminogen
was converted, as determined by densitometric scanning, indi-
88 Rodriguez et . (Eur. J. Biochem. 229)
I I
J
0 phase, but after approximately 80 min the amount of plasmin
200
8 -
-P'g .
PlasminA
-Plasmin B
0
0 20 40 60 80 100
TIME (min)
Fig.5. Activation of plasminogen by 26-kDa rSK or by equimolar
mixtures of plasmin, 26-kDa rSK and a,-antiplasmin. Plasminogen
(final concentration 10 pM) was incubated with (a) 26-kDa rSK (B),(b)
a mixture of plasmin, 26-kDa rSK and a,-antiplasmin (0)or (c) a mix-
ture of D-ValPheLysCH,-plasmin, 26-kDa rSK and a2-antiplasmin (+)
at a final concentration of approximately 100 nM each. Generation of
plasmin activity as a function of time was monitored with D-GlpPheLys-
NH-Np. The data represent mean 2 SD of three experiments. The inset
shows reduced SDSPAGE of samples taken after incubation for 95 min
with 26-kDa rSK (lane 2), with the plasmid26-kDa rSWa,-antiplasmin
mixture (lane 3) and with the ~-ValPheLysCH,-plasmid26-kDa rSW
a,-antiplasmin mixture (lane 4). Lane 1 represents a protein calibration
mixture, as described in Fig. I. Plg, plasminogen; plasmin A, plasmin
A chain; plasmin B, plasmin B chain.
cating significant plasmin degradation under the experimental
conditions.
SDSPAGE under reducing conditions (Fig. 4), revealed
nearly quantitative conversion, within 10 min, of plasminogen
to plasmin in the experiments carried out with streptokinase
(lane 2) and recombinant streptokinase (lane 3). In the experi-
ment with 26-kDa rSK, some (degraded) plasmin was detected
after 60 min (lane 4), whereas in the experiment with 17-kDa
rSK, no plasmin was detected (lane 5).
Inhibition of plasmin-streptokinase complexes by a,-anti-
plasmin. The formed plasmin- 17-kDa rSK complex was inhib-
ited by a,-antiplasmin with an apparent second-order rate con-
stant which was indistinguishable from that of the inhibition of
free plasmin by a,-antiplasmin (approximately 2 X lo7 M-' S K I ) .
In contrast, complexes of plasmin with recombinant streptoki-
nase were not inhibited to any measurable extent by a,-antiplas-
min.
Addition of an up to eightfold molar excess of 26-kDa rSK
did not affect the inhibition of plasmin by a,-antiplasmin (100%
plasmin inhibited; mean of two experiments) indicating normal
inhibition of the plasmin-26-kDa-rSK complex. In contrast, ad-
dition of recombinant streptokinase resulted in a concentration-
dependent reduction of the inhibition of plasmin by a,-antiplas-
min, apparently because the formed plasmin-recombinant-
streptokinase complex is resistant to inhibition by a,-antiplasmin
(50%, 41%, 7% or 4% of plasmin inhibited at 1:1, 2:1, 4 : l
or 8 : 1 molar excess of recombinant streptokinase; data not
shown).
Additional experiments designed to investigate the mecha-
nism of the observed rapid inhibition of the plasmin-26-kDa-
rSK complex by a,-antiplasmin, suggested that 26-kDa rSK dis-
sociates from the complex into a functionally active form upon
inhibition by a,-antiplasmin. This is suggested by our observa-
tion that time-dependent generation of plasmin activity is ob-
tained following addition to excess plasminogen of an approxi-
mately equimolar mixture of plasmin and 26-kDa rSK, which
was previously neutralized with a,-antiplasmin (Fig. 5). The
time course of plasmin generation showed a pronounced lag
generated was comparable to that found in a control experiment
with the same concentrations of 26-kDa rSK and plasminogen.
An additional experiment revealed that no plasmin activity was
generated when the mixture of D-ValPheLysCH,-plasmin,
26-kDa rSK and a,-antiplasmin was used under the same condi-
tions (Fig. 5). This indicates that no functional 26-kDa rSK is
generated from this inactive complex upon addition of a,-anti-
plasmin.
SDS/PAGE under reducing conditions (Fig. 5, inset) of sam-
ples taken at the end of the experiments confirmed conversion
of plasminogen to plasmin in the experiments with 26-kDa rSK
(lane 2) and with the mixture of plasmin, 26-kDa rSK and
a,-antiplasmin (lane 3) but not in the mixture with D-ValPhe-
LysCH,-plasmin (lane 4). Densitometric scanning of the gels
revealed conversion of approximately 30 % of the plasminogen
(lanes 2 and 3).
Dissociation of 26-kDa rSK from the complex was con-
firmed by adsorption of equimolar mixtures of plasmid26-kDa
rSWa,-antiplasmin to Lys-Sepharose, followed by elution with
0.1 M Tris/HCl, pH 8.4, containing 2% SDS. Densitometric
scanning of SDS/PAGE under non-reducing conditions (data not
shown) revealed that 26-kDa rSK was recovered only in the
breakthrough, whereas only the plasmin- a,-antiplasmin com-
plex was detected in the eluate. A control experiment with
adsorption of an equimolar mixture of D-ValPheLysCH,-plasmin
and 26-kDa rSK to Lys-Sepharose showed quantitative recovery
of both active-site-blocked plasmin and 26-kDa rSK in the
eluate, confirming that the plasmin- 26-kDa-rSK complex re-
mains intact in the absence of a,-antiplasmin (data not
shown).
Gel-filtration experiments confirmed that 17-kDa rSK re-
mains associated with plasmin upon interaction of the complex
with a,-antiplasmin. The elution pattern from SuperdexTM200
HR indeed revealed co-elution of 17-kDa rSK with the plas-
min-a,-antiplasmin complex (Fig. 6). This was confirmed by
radioactivity determination (data not shown), as well as by SDS/
PAGE of the peak fraction (Fig. 6 C, inset). The main bands ob-
served, from top to bottom of the gel, correspond to the plas-
min-a,-antiplasmin complex, residual a,-antiplasmin, 17-kDa
rSK and the 14-kDa peptide released from a,-antiplasmin upon
interaction with plasmin.
DISCUSSION
Streptokinase, a bacterial protein that is routinely used for
thrombolytic therapy, is not an enzyme; it binds to plasminogen,
resulting in the exposure of an active-site in the plasminogen
moiety of the equimolar complex [1-31. The plasmin(ogen)-
streptokinase complex is not neutralized by a,-antiplasmin.
Whereas it is clearly established that streptokinase interacts with
the B chain of plasmin [28], the structural domains of streptoki-
nase involved in the interactions with plasmin(ogen) and a,-anti-
plasmin are less well characterized. Since the three-dimensional
structures of streptokinase and of its complex with plasmin-
(ogen) have not been determined, studies on structure/function
relationships of streptokinase have mainly been performed using
isolated fragments of the molecule [lo-161.
 Rodriguez et al. (EUKJ. Biochem. 229)
0.03
0.02
0.01
E rSK, in contrast to 26-kDa rSK, binds to plasmin at a site remote
C 0.40
0 from the active site, which is not involved in the interaction
00
N
W would bind to plasmin(ogen) in the vicinity of the active-site
0 0.20
z recombinant plasminogen for streptokinase was substantially re-
2PL
0 0.00
v)
m nal region (Ilel -Lys59) which is not involved in plasminogen
a activation, two regions (Ser60-Arg142 and Val294-Lys386)
0.10
0.05
0.00
8 10 12 14 16 18
ELUTION VOLUME (mil
Fig. 6. Elution pattern following gel filtration on a SuperdexTM200
HR column of 17-kDa rSK before and after addition of plasmin and
a,-antiplasmin. Samples applied are 17-kDa rSK (A), equimolar mix-
ture of D-ValPheLysCH,-plasmin and 17-kDa rSK (B), and equimolar
mixture of plasmin and 17-kDa rSK with 20% excess of a,-antiplasmin
(C). The inset shows SDSPAGE under non-reducing conditions of the
peak fractions (lanes 2) and of a protein calibration mixture (lanes l),
as described in Fig. 1.
In the present study, we have evaluated the molecular in-
teractions between plasmin(ogen), a,-antiplasmin and streptoki-
nase using two recombinant streptokinase fragments ; 17-kDa
rSK and 26-kDa rSK, consisting of residues Val143-Lys293
and Val143 -Lys386, respectively.
Both 17-kDa rSK and 26-kDa rSK bound with high affinity
to plasmin(ogen) and did not compete with binding of recombi-
nant streptokinase to plasmin(ogen), whereas recombinant strep-
tokinase could displace 17-kDa rSK from plasmin(ogen). Sta-
bility of the complexes was confirmed by gel-filtration chroma-
tography and by adsorption onto Lys-Sepharose. Functional
analysis of the interaction revealed that no active-site was ex-
posed upon binding of 17-kDa rSK to plasminogen and that no
plasmin activity was generated in mixtures of plasminogen with
catalytic amounts of 17-kDa rSK or plasmin- 17-kDa-rSK com-
plex. In contrast, slow active-site exposure was observed in
equimolar mixtures of 26-kDa rSK and plasminogen, accounting
for the very slow and limited activation of plasminogen by cata-
lytic amounts of 26-kDa rSK (approximately 4% of the activa-
tion obtained with recombinant streptokinase). Together with
previously published results, these data are consistent with the
hypothesis that at least two regions in the streptokinase molecule
are involved in the interaction with plasminogen [29]; one site
located in the region comprising residues Ser60 -Arg142 and
the other in the region comprising Val143 -Lys293.
89
Interestingly, we have observed that complexes of plasmin
with 17-kDa rSK and 26-kDa rSK are very rapidly inhibited by
a,-antiplasmin, in contrast to the complex with intact streptoki-
nase which is resistant to inhibition. Structural analysis of the
reaction mixtures revealed that, upon interaction with a,-anti-
plasmin, 26-kDa rSK dissociates in a functionally active form
from the complex and can be recycled to other plasminogen
molecules, whereas the inactive 17-kDa rSK remains associated
with the plasmin- a,-antiplasmin complex. Apparently, 17-kDa
with a,-antiplasmin. Earlier studies suggested that streptokinase
[30], and in a recent study it was reported that the affinity of
duced by substitution of Arg719 by Glu [31].
Together, the available data confirm that streptokinase is a
protein with at least four different structural regions; an N-termi-
which are crucial for plasminogen activation, and one region
(Val143-Lys293) which binds to plasminogen but does not acti-
vate it. Interestingly, this region does not impair the inhibition
by a,-antiplasmin of the complex with plasmin.
We are grateful to Dr J. Van Damme and to Mr P. Proost (Rega
Institute, KU Leuven, Belgium) for performing N-terminal amino acid
sequence analyses, and to Dr R. Silva (Center for Genetic Engineering
and Biotechnology, Havana, Cuba) for assistance in the design of oligo-
nucteotides. Skilful technical assistance of F. De Cock, B. Van Hoef and
E. Demarsin is gratefully acknowledged.
REFERENCES
1. McClintock, D. K. & Bell, P. H. (1971) The mechanism of activation
of human plasminogen by streptokinase, Biochem. Biophys. Res.
Commun. 43, 694-702.
2. Reddy, K. N. N. & Markus, G. (1972) Mechanism of activation of
human plasminogen by streptokinase. Presence of active center
in streptokinase-plasminogen complex, J. Biol. Chem. 247,
1683- 1691.
3. Schick, L. A. & Castellino, F. J. (1974) Direct evidence for the
generation of an active-site in the plasminogen moiety of the
streptokinase-human plasminogen activator complex, Biochem.
Biophys. Res. Commun. 57, 47-54.
4. Brockway, W. J. & Castellino, F. J. (1974) A characterization of
native streptokinase and altered streptokinase isolated from hu-
man plasminogen activator complex, Biochemistry 13, 2063 -
2070.
5. Siefring, G. E. Jr & Castellino, F. J. (1976) Interaction of streptoki-
nase with plasminogen: isolation and characterization of a strep-
tokinase degradation product, J. Biol. Chem. 251, 3913 - 3920.
6. Reddy, K. N. N. & Kline, D. L. (1976) Isolation of an active deriva-
tive produced from native streptokinase by dog plasmin, Thromb.
Res. 9, 407-411.
7. Misselwitz, R., Kraft, R., Kostka, S., Fabian, H., Welfle, K., Pfeil,
W., Welfle, H. & Gerlach, D. (1992) Limited proteolysis of strep-
tokinase and properties of some fragments, Int. J. Biol. Mucromol.
14, 107-116.
8. Jackson, K. W., Malke, H., Gerlach, D., Ferretti, J. J. & Tang, J.
(1986) Active streptokinase from the cloned gene in Streptococ-
cus sunguis is without the carboxyl-terminal 32 residues, Bio-
chemistry 25, 108-114.
9. Muller, J., Reinert, H. & Malke, H. (1989) Streptokinase mutations
relieving Escherichia coli K-12 (prIA4) of detriments caused by
the wild type skc gene, J. Bucteriol. 171, 2202-2208.
10. Kiessen, C., Schmidt, K. H., Ferretti, J. J. & Malke, H. (1988) Tri-
partite streptokinase gene fusion vectors for gram-positive and
gram-negative procaryotes, Mol. Gen. Genet. 212, 295-300.
90 Rodriguez et al. (Eul: J. Biochem. 229)
11. Malke, H. (1993) Polymorphism of the streptokinase gene: Implica-
tions for the pathogenesis of post streptococcal glomernlonephr-
tis, Zentralbl. Bakteriol. lnt. J. Med. Microbiol. Virol. Parasitol.
Infect. Dis. 278, 246-257.
12. Gonzalez-Gronow,M., Enghild, J. J. & Pizza, S . V. (1993) Streptoki-
nase and human fibronectin share a common epitope: implica-
tions for regulation of fibrinolysis and rheumatoid arthritis, Bio-
chim. Biophys. Acta 1180, 283-288.
13. Reed, G. L., Kussie, P. & Parham-Seren, B. (1993) A functional
analysis of the antigenicity of streptohnase using monoclonal
antibody mapping and recombinant streptokinase fragments, J.
lmmunol. 150, 4407-4415.
14. Rodn’guez, P., Fuentes, D., Muiioz, E., Rivero, D., Orta, D., Albur-
querque, S . , Perez, S . , Besada, V. & Herrera, L. (1994) The strep-
tohnase domain responsible for plasminogen binding, Fibrinoly-
sis 8, 276-285.
15. Malke, H., Roe, B. & Ferretti, J. J. (1987) Streptokinase: expression
of altered forms, in Streptococcal genetics (Ferretti, J. J. & Cur-
tiss, R. 111, eds) pp. 143-149, Amer. SOC.Microbiol., Washington
DC.
16. Wong, S . L., Ye, R. & Nathoo, S . S . (1994) Engineering and produc-
tion of streptokinase in a Bacillus subtilis expression-secretion
system, Appl. Environ. Microbiol. 60, 517-523.
17. Estrada, M. P., HLmandez, L., PLrez, A., Rodriguez, P., Serrano, R.,
Rubiera, R., Pedraza, A., Padron, G., Antuch, W., de la Fuente,
J. & Herrera, L. (1992) High level expression of streptokinase in
Escherichia coli, Biotechnology 10, 1138-1142.
18. Sanger, G., Nicklen, S . & Coulson, A. R. (1977) DNA sequencing
with chain-terminating inhibitors, Proc. Nut1 Acad. Sci. USA 74,
5463 -5467.
19. Bradford, M. M. (1976) A rapid and sensitive method for the quanti-
tation of microgram quantities of protein utilizing the principle of
protein-dye binding, Anal. Biochem. 72, 248 -254.
20. Deutsch, D. G. & Mertz, E. T. (1970) Plasminogen: purification
from human plasma by affinity chromatography, Science 170,
1095- 1096.
21. Nelles, L., Lijnen, H. R., Collen, D. & Holmes, W. E. (1987) Char-
acterization of a fusion protein consisting of amino acids 1 to 263
of tissue-type plasminogen activator and amino acids 144 to 41 1
of urokinase-type plasminogen activator, J. Biol. Chem. 262,
10855-10862.
22. Wiman, B. (1980) Affinity-chromatographic purification of human
alpha-2-antiplasmin, Biochem. J. 191, 229-232.
23. Busby, S . J., Mulvihill, E., Rao, D., Kumar, A. A., Lioubin, P.,
Heipel, M., Sprecher, C., Halfpap, L., Prunkard, D., Gambee, J. &
Foster, D. C. (1991) Expression of recombinant human plasmino-
gen in mammalian cells is augmented by suppression of plasmin
activity, J. Biol. Chem. 266, 15286-15292.
24. Lijnen, H. R., Van Hoef, B., Nelles, L. & Collen, D. (1990) Plasmin-
ogen activation with single-chain urokinase-type plasminogen ac-
tivator (scu-PA). Studies with active site mutagenized plasmino-
gen (Ser74WAla) and plasmin resistant scu-PA (Lysl58-+Glu),
J. Biol. Chem. 265, 5232-5236.
25. Marchalonis, J. J. (1969) An enzymic method for the trace iodin-
ation of immunoglobulins and other proteins, Biochem. J. 113,
299 -305.
26. Lijnen, H. R., De Cock, F. & Collen, D. (1994) Characterization
of the binding of urokinase-type plasminogen activator (u-PA) to
plasminogen, to plasminogen activator inhibitor-1 and to the u-PA
receptor, Eus J. Biochem. 224, 567-574.
27. Lijnen, H. R., Van Hoef, B., De Cock, F., Okada, K., Ueshima, S . ,
Matsuo, 0. & Collen, D. (1991) On the mechanism of fibrin-
specific plasminogen activation by staphylokinase, J. Biol. Chem.
266, 11826- 11832.
28. Summaria, L. & Robbins, K. C. (1976). Isolation of a human plas-
min-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.
29. Young, K. C., Shi, G. Y., Huang, H. C., Chang, Y. F., Hsiao, W.
C. & Wu, H. L. (1993) At least two interaction sites on the strep-
tokinase molecule for plasminogen activation, Thromb. Huemost.
69, 843.
30. Wohl, R. C. (1984) Interference of active-site specific reagents in
plasminogen-streptokinaseactive-site formation, Biochemistry 23,
3799- 3804.
31. Dawson, K. M., Marshall, J. M., Raper, R. H., Gilbert R. J. &
Ponting, C. P. (1994) Substitution of arginine 719 for glutamic
acid in human plasminogen substantially reduces its affinity for
streptokinase, Biochemistry 33, 12042-12047.