(Received for publication, June 25, 1996, and in revised form, April 11, 1997)
From the Unité des Venins, Institut Pasteur, 25 rue du Docteur Roux, 75724 Paris Cedex 15, France and
§ Kunming Institute of Zoology, Chinese Academy of Sciences,
Kunming 650223, Yunnan, China
The specific plasminogen activator from Trimeresurus stejnegeri venom (TSV-PA) is a serine proteinase presenting 23% sequence identity with the proteinase domain of tissue type plasminogen activator, and 63% with batroxobin, a fibrinogen clotting enzyme from Bothrops atrox venom that does not activate plasminogen. TSV-PA contains six disulfide bonds and has been successfully overexpressed in Escherichia coli (Zhang, Y., Wisner, A., Xiong, Y. L., and Bon, C. (1995) J. Biol. Chem. 270, 10246-10255).
To identify the functional domains of TSV-PA, we focused on three short peptide fragments of TSV-PA showing important sequence differences with batroxobin and other venom serine proteinases. Molecular modeling shows that these sequences are located in surface loop regions, one of which is next to the catalytic site. When these sequences were replaced in TSV-PA by the equivalent batroxobin residues none generated either fibrinogen-clotting or direct fibrinogenolytic activity. Two of the replacements had little effect in general and are not critical to the specificity of TSV-PA for plasminogen. Nevertheless, the third replacement, produced by the conversion of the sequence DDE 96a-98 to NVI, significantly increased the Km for some tripeptide chromogenic substrates and resulted in undetectable plasminogen activation, indicating the key role that the sequence plays in substrate recognition by the enzyme.
The plasminogen activator from Trimeresurus stejnegeri venom (TSV-PA)1 is a 234-residue single chain glycoprotein with an apparent molecular weight of 33 kDa (1). Like physiological tissue type plasminogen activator (t-PA), TSV-PA specifically cleaves the Arg561-Val562 plasminogen bond to generate two-chain plasmin, a key enzyme in fibrinolysis (2, 3). Sequence homology with trypsin and other venom serine proteinases (4) indicates that TSV-PA belongs to the family of serine proteinases (5). Trypsin-like serine proteinases, which cleave the peptide bond following arginine or lysine, display very different substrate and inhibitor specificities. They were among the first enzymes to be studied extensively (6). In particular, the role of the "specificity pocket" in determining the "primary", or P1,2 specificity of the enzyme has long been recognized (7). The existence and critical importance of other additional structural determinants of proteinase specificity have also been established (8), but both their location and their role remain unknown.
The sequence of TSV-PA exhibits a high degree of identity (60-66%) with other snake venom proteinases, such as Vipera russelli venom factor V activator (9), batroxobin from Bothrops atrox venom (10), and Agkistrodon contortrix venom protein C activator (11), which present considerable differences in their substrate specificities. For example, TSV-PA, which efficiently activates plasminogen, does not clot or degrade fibrinogen and does not activate or degrade factor X, prothrombin, and protein C (1). On the other hand, batroxobin, which shows a thrombin-like activity (12), does not act on plasminogen. TSV-PA shares a weak sequence identity with the serine proteinase domain of human t-PA (23%) and urokinase type plasminogen activator (u-PA) (21%) yet is functionally analogous to them. Moreover, in common with the highly nonspecific trypsin, TSV-PA is a one-chain enzyme, and it possesses six disulfide bonds, of which five are topologically equivalent; the sequence identity with trypsin is 40%.
For the expression and refolding of recombinant TSV-PA (rTSV-PA) in bacteria we used a strategy similar to the one reported by Maeda et al. (13) for batroxobin. This led us to undertake structure-function studies on the substrate specificity of TSV-PA by site-directed mutagenesis to examine the role of particular residue(s) or region(s). We thus looked at three peptide fragments (residues SNNFQ 60-64, DDE 96a-98, and SWRQV 173-176) that show quite different amino acid sequences between TSV-PA and batroxobin. These fragments are located in solvent-exposed loops in a sequence homology trypsin-based three-dimensional model of TSV-PA. Consequently, we exchanged the TSV-PA sequence by the equivalent batroxobin residues by site-directed mutagenesis. A TSV-PA variant whose modifications are located close to the catalytic site lost its plasminogen activation properties.
Human thrombin, human Lys-plasminogen, and
urokinase (two-chain form, Mr 33,000) were
obtained from Sigma. Bovine factor Xa was from Pierce. Human fibrinogen
(grade L) from Kabi Vitrum was pretreated with diisopropyl
fluorophosphate according to the instructions of the manufacturer.
Natural TSV-PA was purified from T. stejnegeri venom as
described before (1). The concentration of plasminogen was determined
by measuring the absorption at 280 nm, using an absorption coefficient
1% of 16.8 (14).
Methods
Construction of a TSV-PA Expression PlasmidThe pET
expression vector (Novagen) was used for expression of TSV-PA in
Escherichia coli. To subclone the TSV-PA open reading frame
from the cDNA clone D16 into the pET17b vector, a polymerase chain
reaction was conducted with a forward primer, En, and a reverse primer,
EcR, in the presence of an adapter primer, E. Primer E contains the
following elements: 1) a BamHI site to facilitate subcloning
(italicized); 2) the coding sequence for a tetrapeptide recognition
site of factor Xa (boldface); and 3) 5 N-terminal residues of the
TSV-PA coding sequence (underlined):
5-CCGGATCCAATCGAAGGTCGTGTCTTTGGAGGTGAT-3
. Primer En corresponds to the first 24 nucleotides of primer E. Primer
EcR, 5
-AGAATTCTCACGGAGGGCAGGTCGC-3
is an antisense oligonucleotide that contains the following: 1)
an EcoRI site to facilitate subcloning (italicized); 2) a
stop codon (boldface); and 3) the coding sequence for the 5 C-terminal
residues of TSV-PA (underlined). The product, amplified with primers En
and EcR, was first subcloned into a pGEM vector (Promega). After
digestion with BamHI and EcoRI, the recovered
TSV-PA open reading frame was ligated into the pET17b vector at
BamHI and EcoRI sites, to generate the expression
plasmid, designated pET(tsvpa). The constructed plasmid was sequenced
on both strands to ensure that the coding sequence of TSV-PA was
correct.
The construction of expression plasmids for TSV-PA variants was followed by overlap extension using polymerase chain reaction techniques, mainly as described by Ho et al. (15). Using pET(tsvpa)15 as the template, one fragment was amplified between primer Vn, in the sense direction, containing specific alterations in the nucleotide sequence and an antisense T7 terminal primer (GCTAGTTATTGCTCAGCGG), located downstream of the TSV-PA coding sequence. Another fragment was amplified between the antisense primer VnR (reverse complementary to Vn) and a sense primer located in the T7 promoter sequence (TAATACGACTCACTATAGGG). The resulting fragments were purified by gel electrophoresis on polyacrylamide gels and combined in a subsequent "fusion" reaction in which the overlapping ends annealed. Polymerase chain reaction amplification was then carried out between the T7 promoter and terminal primers. The resulting 1-kilobase pair fragment was also purified by gel electrophoresis, digested by BamHI and EcoRI, and ligated into the pET17b vector. The entire sequence encoding mutated TSV-PA was then verified by sequencing both DNA strands. The synthetic oligonucleotides used to generate the TSV-PA variants are listed in Table I.
|
The E. coli strain BL21(DE3) was transformed
with plasmid of wild type, pET(tsvpa), or of TSV-PA variants. Cells
were grown at 37 °C up to an absorbance of about 0.6 at 600 nm and
then induced with 0.4 mM
isopropyl-1-thio--D-galactopyranoside and cultivated for
an additional period of 3 h at 37 °C. After cell lysing,
inclusion bodies containing TSV-PA or mutated TSV-PA were recovered by
centrifugation, washed three times with 50 mM Tris-HCl, pH
8.0, containing 2.5 mM EDTA, and then dissolved in 8 M urea in distilled water at room temperature for 30 min.
The sample was dialyzed against 50 mM Tris-HCl, pH 8.0, 0.25 M NaCl, 2 mM EDTA at room temperature for
150 min. Bovine factor Xa was added at an enzyme:substrate ratio of
1:100, and digestion was conducted overnight at 37 °C. The protein
was then fully denatured by dilution into 10 volumes of 8 M
urea in 50 mM Tris-HCl, pH 8.0, 2 mM EDTA,
containing 50 mM
-mercaptoethanol and incubated in the
buffer for 120 min at room temperature. Refolding was started by
diluting the sample 50 times in 50 mM Tris-HCl, pH 8.0, 2 mM EDTA. The refolding process was allowed to proceed at
room temperature and was monitored by assaying the amidolytic activity
of TSV-PA on chromogenic substrate S-2238. When the enzyme activity
reached a plateau, after about 30-48 h, the refolding mixture was
stored at 4 °C for subsequent purification. N-terminal sequence
determinations of the recombinant protein were performed by Edman
degradation with a 470A gas phase sequencer.
The purification of refolded wild type and mutated proteins was performed according to the last purification step used for natural TSV-PA from T. stejnegeri venom (1). The refolded protein was concentrated in an Amicon stirred cell concentrator under nitrogen pressure.
Measurement of Enzyme ConcentrationEnzyme concentrations were determined by the Bio-Rad protein assay. To determine the molar concentration of the active site enzyme, wild type and mutated rTSV-PA were subjected to active site titration with 4-methylumbelliferyl p-guanidinobenzoate (16), using a Kontron spectrofluorimeter. Titrations were performed in 50 mM Tris-HCl (pH 8.0) containing 100 mM NaCl.
SDS-Polyacrylamide Gel ElectrophoresisSDS-PAGE was performed on a Phastsystem apparatus (Pharmacia) following the manufacturer's directions.
Immunochemical Analysis of rTSV-PAPurified rTSV-PA was tested by enzyme-linked immunosorbent assay using rabbit polyclonal antibodies raised against natural TSV-PA. Competition experiments were performed as described by Choumet et al. (17). Briefly, various concentrations of natural or rTSV-PA were incubated for 12 h at 4 °C with a fixed dilution of anti-TSV-PA antibodies as deduced from preliminary enzyme-linked immunosorbent assay calibration. One hundred µl of each mixture were then transferred onto a microtitration plate previously coated with native TSV-PA and incubated for 1 h at 37 °C. Subsequent steps were performed as described above. The concentration of natural or rTSV-PA that gave half of the maximal response (IC50) was determined.
Chromogenic Assays and Kinetic AnalysesThe amidolytic activity of the enzymes was measured as described before (1). The kinetic parameters Km and kcat were determined by analysis of double-reciprocal plots of the initial velocity as a function of substrate concentration. The concentrations of the chromogenic substrates S-2238 and S-2251 varied from 0.3 mM to 0.01 mM in assays of all enzymes. The enzyme concentrations used were between 8 and 17 nM, except in the case of the TSV-PA variant D96aN/D97V/E98I, which was used at 55 nM. S-2302 and S-2444 substrate concentrations varied from 0.03 mM to 0.4 mM, and the final enzyme concentrations varied from 26 to 55 nM.
Plasminogen Activation Assays and Kinetic AnalysisThe initial rate of plasminogen activation was measured with Lys-plasminogen as described previously (1). For kinetic analyses, Lys-plasminogen concentrations varied from 10 to 100 nM in the case of natural and recombinant wild type TSV-PA, the final concentration of TSV-PA being 10 nM. In the case of TSV-PA variants, Lys-plasminogen concentrations varied from 50 to 1000 nM, and the final concentration of the enzyme varied from 10 to 28 nM.
Topological Alignment and Modeling of TSV-PAModeling of the three-dimensional structure of TSV-PA (computations, visualization and analysis) was performed using the BIOPOLYMER and HOMOLOGY modules of the BIOSYM Technologies Inc. (San Diego, CA) software package.
The set of structurally conserved regions (SCRs) thus determined was
based on the crystal structure of trigonal bovine pancreas trypsin
(Protein Data Bank entry code 3PTN; Ref. 18). The choice for trypsin as
the reference structure, rather than the functionally analogous u-PA
(19) or t-PA (20) was justified by the fact that the B-chain of the
C-terminal catalytic proteinase domain of u-PA contains several
insertion fragments of considerable size, resulting in a total length
of 248 residues compared with the 234 residues of TSV-PA. The length of
trypsin (223 residues) is then closer to that of TSV-PA. Furthermore,
five of the six disulfide bridges of TSV-PA are found in the same
positions along the sequence in trypsin, as compared with four in the
case of u-PA, one of the u-PA bridges linking the catalytic and kringle
domains. The SCRs determined through structural alignment are in line
with those of Greer (21), although they include minor modifications that take into consideration changes in the definition of certain segments due to an updated set of serine proteinases. As a first approximation, the loops were modeled as follows. When the loop length
was the same, the backbone coordinates of TSV-PA were transferred from
those of trypsin; in the case of one-residue insertions, the
corresponding residue was inserted in the sequence, and the conformation of the new loop was energy-minimized. The conformations of
the side chains of mutated residues were energy-minimized without any
further refinement such as conformation search or rotamer assignment.
Loop 142-151 of TSV-PA was obtained from the corresponding loop in
u-PA, since the sequence homology was significantly higher than with
trypsin. TSV-PA contains a C-terminal amino acid extension absent in
trypsin and involved in the 91-245e cysteine bridge unique to TSV-PA.
This fragment was model-built and optimized such as to ensure the
formation of the corresponding disulfide bond. The stereochemical
quality of the model was checked with the program PROCHECK (22). The
numbering of the TSV-PA sequence is based on that of
-chymotrypsin.
After overexpression of
TSV-PA in E. coli, a major protein of 30 kDa could be
detected in whole cell extracts by SDS-PAGE. After denaturation, the
renaturation process yielded 20% of active enzyme. Active rTSV-PA was
purified from misfolded protein and from contaminant bacterial proteins
by chromatography in a Mono-Q anion exchange column (Fig.
1). The protein corresponding to peak 2 was able to hydrolyze chromogenic substrates and possessed plasminogen activation activity. This fraction corresponded indeed to rTSV-PA. The
inset in Fig. 1 shows an SDS-PAGE analysis of the different fractions obtained by expression of TSV-PA in E. coli.
Purified rTSV-PA reacted with polyclonal antibodies prepared against
the natural enzyme. The IC50 determined through competition
experiments was similar for recombinant and natural TSV-PA
(5·107 M; data not shown).
Activity Assays and Kinetic Analyses of rTSV-PA
Titration of the active site in each molecule of natural and rTSV-PA gives a ratio of active site concentration to protein concentration of 0.97 ± 0.02 for both, indicating that rTSV-PA was fully active.
The activity of rTSV-PA on chromogenic substrates was characterized by
the Km and kcat parameters
obtained from Lineweaver-Burk plots for the synthetic substrates
S-2238, S-2251, S-2302, and S-2444. Table II shows that the kinetic
parameters of rTSV-PA with these substrates, when determined under the
same conditions in parallel experiments, were not significantly
different from those of natural TSV-PA. We also analyzed the kinetics
of plasminogen activation by TSV-PA. TSV-PA activates Lys-plasminogen
with a second order rate of 3.8 µM1·min
1 and with small
Michaelis and catalytic constants (Km = 53 nM, kcat = 0.2 min
1).
We obtained the same result for rTSV-PA (Table II).
Km of TSV-PA on Lys-plasminogen activation is 35 times smaller than that of u-PA (Km = 1.9 µM; Ref. 23). On the other hand, TSV-PA cleaves
plasminogen with a kcat 100 times and 50 times lower than that of u-PA and t-PA, respectively (24, 25).
|
The topology-based sequence alignment
between TSV-PA and trypsin is shown in Fig.
2. Fig. 3
shows a MOLSCRIPT representation (26) of the trypsin-like
three-dimensional model of TSV-PA. As in all members of the family of
serine proteinases (21), the topology of TSV-PA consists of two
subdomains of antiparallel -barrel structure (three strands in each
subdomain) and a C-terminal helical segment. The spatial organization
of the catalytic triad residues His57, Asp102,
and Ser195 (7) at the active site cleft between the two
subdomains is preserved in TSV-PA. The overall dimensions of TSV-PA are
slightly larger than those of trypsin but smaller than those of u-PA.
His217a appears to reduce solvent accessibility to the
gorge of the S1 specificity pocket. In the S2 pocket, a valine replaces
the imidazole ring of His99 in u-PA. As a consequence,
small hydrophobic residues usually found at the P2 position of known
TSV-PA substrates can be accommodated. The C-terminal peptide of
TSV-PA, contiguous to the large 234-243 helix, comprises a
-turn
for residues 243-245a with Gly244 in the i + 1 position and an extended conformation for the
Asp245b-Pro245g segment, which lies
approximately perpendicular to the axis of the helix. This spatial
arrangement will ensure the formation of the
Cys91/Cys245e disulfide bridge characteristic
of TSV-PA. All amide bonds are in trans.
Sequence comparison between TSV-PA and batroxobin identifies several
regions where the two enzymes present a significant difference in amino
acid sequence. Among those, the three segments, A, B, C (Table
I), are exposed to the solvent in our
three-dimensional model (Fig. 3). The loop ending at Val99
is smaller than that of u-PA and corresponds to region B of TSV-PA, made up of the charged peptide KKDDE 95-98, while the more hydrophobic SADTLA 95-98 segment lines the S2 pocket of u-PA. Region B is in close
spatial proximity to the catalytic triad of the molecule, and the
active site is thus slightly more accessible in TSV-PA than in u-PA,
but less so than in trypsin. The S3 pocket is essentially composed of
Gly216, and the 95-99 loop does not extend to it as in
u-PA. Upon imposing the canonical conformation of positions P5 to P1
of the inhibitors complexed to trypsin, human
-thrombin, and u-PA
(Protein Data Bank codes 1MCT, 1ABJ, and 1LMW, respectively) to the
corresponding sequence of plasminogen, two features become apparent.
First, the cyclic side chain of position P3 (Pro559) points
away from region B, and only the main chain-main chain hydrogen bond
with Gly216 remains. Second, positions P6
(Lys556) and P5 (Lys557) of plasminogen,
located next to the small disulfide loop containing the
Arg561-Val562 scissile bond (27) lie proximal
to Asp97 of TSV-PA. Finally, the
Cys128-Cys232 disulfide bond in trypsin is
replaced by a stacking interaction between Pro128 and
Phe232 in TSV-PA, thus helping stabilize the long
Pro128-Val135 unstructured segment.
Other features of the three-dimensional model of TSV-PA common to the family of serine proteinases include the glycine at position 69, the characteristic CGG 42-44 sequence, and an internal salt bridge between the amino terminus and the side chain of the catalytic site residue Asp194. Interestingly enough, in contrast to all trypsin-like serine proteinases that possess a glycine 193 at the S1 subsite, TSV-PA has a phenylalanine.
Mutagenesis of TSV-PATo examine whether the three regions mentioned above might be implicated in the differential substrate specificities of TSV-PA, the variants constructed by replacing the corresponding amino acid residues by those of batroxobin (Table I), were expressed and purified as for the wild type enzyme. Under the same conditions, the renaturation yield of the variants varied from 7 to 20% and their elution from the Mono-Q column was slightly shifted according to their electric charge (results not shown).
For each TSV-PA variant the active site to protein concentration ratio per molecule, as determined by the Bio-Rad protein assay, was close to 1, as was the case for wild type rTSV-PA. Immunochemical analysis of TSV-PA variants by enzyme-linked immunosorbent assay showed that the IC50 values measured with each variant were similar to that of natural and rTSV-PA, suggesting that the mutations performed did not affect the epitope recognized by anti-TSV-PA antibodies and that the structural changes are due only to segment replacement and not to folding differences.
Activity of TSV-PA Variants on Chromogenic SubstratesAll TSV-PA variants hydrolyzed S-2238, S-2444, and S-2251 but with significant quantitative differences (Table III). Detailed kinetic analyses revealed that replacement of residues 60-64, SNNFQ by RRFMR (mutant A), did not significantly modify the Km and kcat values of TSV-PA for these three substrates. Replacement of the distinctive electrically charged TSV-PA DDE 96a-98 region by the neutral sequence of batroxobin (NVI), mutant B, significantly increased the Km value for S-2238 and S-2251 and abolished the activity on S-2444. We thus prepared the single point mutants B1 (D96aN), B2 (D97V), and B3 (E98I). As shown in Table III, the Km values increased in some cases and decreased in others. The changes in kcat displayed the same behavior too. For example, the kcat value of D97V for S-2444 was about 8-fold lower than that of the wild type enzyme, whereas it was more than 3-fold higher for S-2238. In mutant C, replacement of the 173-176 peptide (SWRQV to NGLPA) modified slightly the kcat value of the enzyme for these substrates, increasing it for S-2238 and S-2251 and decreasing it for S-2444 by about a factor of 2. The double mutant displayed the combined effects of individual replacements for S-2238, with slight increases in the Km values of all three substrates, and 2-3-fold increases in the kcat values of S-2238 and S-2251.
|
Replacement of the 60-64 and 173-176 peptide (mutants A and C, and double mutant AC) did not abolish plasminogen activation by TSV-PA (Table III). However, the Km for plasminogen was increased 4-fold over that of the wild type TSV-PA for mutant C and 22-fold for mutant AC, resulting in a significant change of the enzymatic activity of plasminogen. Interestingly enough, mutant B showed no detectable plasminogen activation activity, even when the sensitivity of the assay was increased 1000 times. Mutants B1, B2, and B3 retained some plasminogen activation activity but with significantly decreased affinity and rate constants. The most important effect was observed for B2, for which Km increased by a factor of 25. For mutants B1, B2, and B3, the observed plasminogen activation changes widely (from 10- to 60-fold) with respect to the tripeptides (Table III).
Like TSV-PA, the variants did not show any direct fibrinogenolytic activity and did not clot a solution of purified fibrinogen, indicating the absence of thrombin-like activity (data not shown).
Studies of structure-function relationships and of potential thrombolytic activities in vivo require in general large amounts of protein difficult to obtain by purifying TSV-PA from its natural source; T. stejnegeri snakes are rather rare, and their venom is quite expensive. To overcome this obstacle, we prepared rTSV-PA by production in the efficient E. coli expression system. With 12 cysteines and six disulfide bridges, the successful expression of rTSV-PA relied on the isolation of the protein from inclusion bodies, followed by a denaturation-renaturation process in appropriate redox conditions to allow for the correct formation of disulfide bridges. The kinetic parameters of rTSV-PA with synthetic substrates and plasminogen were like those of natural TSV-PA and were not influenced by the absence of glycosylation. Anti-TSV-PA antibodies recognized the natural and the recombinant TSV-PA with comparable sensitivity.
Trypsin and the serine proteinases of snake venoms share a common structure and are believed to have evolved from a common ancestor (5). While the central region surrounding the catalytic triad is highly conserved structurally (28), the surfaces of snake venom serine proteinases may show considerable shape differences given the wide variations in loop sequence and length. The arrangement of disulfide bonds is identical to that of trypsin, except that the Cys128-Cys232 bond of trypsin is replaced by one involving the C-terminal cysteine (4), i.e. the Cys91-Cys245e bond in TSV-PA. The functional significance of this modification, as well as that of the 6-7-residue insertion at the C terminus, is unknown.
As opposed to trypsin, which displays a broad substrate specificity, venom serine proteinases, whose primary structure is highly conserved (60-75% identity among members of the family), show very specific recognition for their particular substrates. This is especially so for TSV-PA, factor V activator, and batroxobin, which act, respectively on plasminogen, factor V, and fibrinogen but do not cleave other macromolecular substrates. Sequence alignment reveals that sequence differences among venom serine proteinases are clustered in several regions that can naturally be thought to contribute to their substrate specificity. The three-dimensional model of TSV-PA (Fig. 3) shows that peptides A, B, and C (Table I), whose sequences differ considerably between TSV-PA and batroxobin, are located on surface regions of TSV-PA. The 6-residue loop containing peptide B is made up almost entirely of charged side chains (KKDDEV) and, of the three peptides, it is the closest to the catalytic site. The data in this work show that the conversion of DDE 96a-98 of TSV-PA to NVI of batroxobin (peptide B) resulted in undetectable activity for plasminogen and in no fibrinogen clotting activity. Replacement of each of the negatively charged residues in the site indicated that D97V is the most critical one. This point mutation resulted in a 125-fold decrease of TSV-PA activity for plasminogen activation. A possible explanation of this effect is the spatial proximity of the DDE loop to the catalytic site. Once the substrate is appropriately placed on the site, this closeness leads to direct electrostatic interactions between Asp97 of TSV-PA and the residues vicinal to the scissile peptide bond of plasminogen (Lys556- Lys557). This is in agreement with the crystallographic structure determination of u-PA and t-PA, which shows that the 99-loop of t-PA contains three carbonyl groups directed toward the active site serving as anchoring parts for hydrogen bonds (20) and that a two-residue hydrophobic extension extends the same loop to form a lip to the binding cleft for u-PA (19). On the other hand, mutant B increased the Km for the chromogenic substrates S-2238 and S-2251 but showed no activity for S-2444. These findings support the idea that replacement of residues in segment B affects tripeptide substrate cleavage and plasminogen differentially, the effect on plasminogen being much larger. Substitution of region A of TSV-PA, SNNFQ 60-64, by that of batroxobin (RRFMR) does not significantly modify the catalytic parameters of TSV-PA. This suggests, in fact, that this loop has little effect in general and is not critical to the plasminogen specificity of TSV-PA. In a similar vein, the experimental data indicate that the region corresponding to loop C, SWRQV 173-176 of TSV-PA, has a moderate effect on all substrates.
In general, the specificities of TSV-PA, t-PA, and u-PA for small chromogenic substrates are quite different (Chromogenix Catalogue, Sweden, and our own data). Yet, just like t-PA and u-PA, TSV-PA acts specifically on plasminogen and exerts no activity on other blood coagulation factors.
TSV-PA shows a high specificity for plasminogen and promises to be an interesting biochemical and pharmacological tool for investigating fibrinolysis in vitro and in vivo. Altogether, our results suggest that, in analogy to t-PA (8), the interaction of TSV-PA with macromolecular substrates such as plasminogen is of a complex nature and depends on secondary site binding. Finally, plasminogen activators seem to show no apparent correlation between their activity toward small substrates on one hand and plasminogen on the other. Comparative structural studies of t-PA, u-PA, and TSV-PA may help to further understand the nature of this discrimination for plasminogen.
We thank Dr. X. Cousin (Unité des Venins, Institut Pasteur, Paris) for helpful advice in conducting the molecular biology experiments, Dr J. d'Alayer (Laboratoire de Microséquence, Institut Pasteur, Paris) for performing the amino acid sequence analyses, Dr. U. K. Nowak (Oxford Center for Molecular Sciences, University of Oxford, UK) for making available the coordinates of u-PA prior to appearance in the Protein Data Bank, and Dr. S. Anderson (Unité des Venins, Institut Pasteur, Paris) for stylistic revision of the manuscript.