Characterization of Placental Bikunin, a Novel Human Serine Protease Inhibitor*

(Received for publication, December 11, 1996)

Katherine A. Delaria Dagger , Daniel K. Muller §, Christopher W. Marlor §, James E. Brown , Rathindra C. Das Dagger , Steven O. Roczniak Dagger and Paul P. Tamburini par **

From Dagger  Preclinical Research, Biotechnology Unit, § Institute for Bone and Joint Disease and Cancer, and par  Institute for Research Technologies, Pharmaceuticals Division, Bayer Corporation, West Haven, Connecticut 06516 and  Process Science, Biotechnology Unit, Berkeley, California 94710

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

We reported previously the cloning of a novel human serine protease inhibitor containing two Kunitz-like domains, designated as placental bikunin, and the subsequent purification of a natural counterpart from human placental tissue (Marlor, C. W., Delaria, K. A., Davis, G., Muller, D. K., Greve, J. M., and Tamburini, P. P. (1997) J. Biol. Chem. 272, 12202-12208). In this report, the 170 residue extracellular domain of placental bikunin (placental bikunin(1-170)) was expressed in baculovirus-infected Sf9 cells using its putative signal peptide. The resulting 21.3-kDa protein accumulated in the medium with the signal peptide removed and could be highly purified by sequential kallikrein-Sepharose and C18 reverse-phase chromatography. To provide insights as to the potential in vivo functions of this protein, we performed an extensive investigation of the inhibitory properties of recombinant placental bikunin(1-170) and both of its synthetically prepared Kunitz domains. All three proteins inhibited a number of serine proteases involved in the intrinsic pathway of blood coagulation and fibrinolysis. Placental bikunin(1-170) formed inhibitor-protease complexes with a 1:2 stoichiometry and strongly inhibited human plasmin (Ki = 0.1 nM), human tissue kallikrein (Ki = 0.1 nM), human plasma kallikrein (Ki = 0.3 nM) and human factor XIa (Ki = 6 nM). Conversely, this protein was a weaker inhibitor of factor VIIa-tissue factor (Ki = 1.6 µM), factor IXa (Ki = 206 nM), factor Xa (Ki = 364 nM), and factor XIIa (Ki = 430 nM). This specificity profile was to a large extent mimicked, albeit with reduced potency, by the individual Kunitz domains. As predicted from this in vitro specificity profile, recombinant placental bikunin(1-170) prolonged the clotting time in an activated partial thromboplastin time assay.


INTRODUCTION

Blood clotting, resulting either from the extrinsic pathway following tissue injury or the intrinsic pathway following contact activation, involves tightly regulated proteolytic cascades (1). The intrinsic pathway is initiated by activation of factor XII either through proteolysis or contact with negatively charged surfaces. Activated factor XIIa, in turn, converts plasma prekallikrein to kallikrein, which can then activate additional factor XII. Factor XIIa activates factor XI, which, in turn, activates factor IX. Activated factor IX forms a complex with factor VIIIa, phospholipid, and calcium, which converts factor X to factor Xa. Factor X is also activated by the factor VIIa-tissue factor complex operating within the extrinsic pathway. Thrombin generation by factor Xa in complex with factor Va leads ultimately to the formation of the fibrin clot. Thrombus formation is also regulated by the fibrinolytic system whereby plasmin, formed from plasminogen by the action of kallikrein, tissue plasminogen activator (tPA),1 or urokinase, breaks down both fibrinogen and fibrin (2).

Protease inhibitors play critical roles in the regulation of the coagulation and fibrinolytic systems. Tissue factor pathway inhibitor (TFPI), a multivalent Kunitz-type inhibitor, is a major regulator of the extrinsic pathway through factor Xa-dependent inhibition of factor VIIa-tissue factor (3). C1 inhibitor is thought to be the major physiological inhibitor of the intrinsic pathway enzymes plasma kallikrein and factor XII (4). More recently, TFPI-2, a serine protease inhibitor structurally related to TFPI, has been proposed to be an important physiological inhibitor of enzymes involved in both coagulation and fibrinolysis (5).

In a previous study, we identified and cloned a human cDNA encoding a novel protein containing two Kunitz-like domains which we termed placental bikunin (6). In this study a soluble recombinant fragment of this protein, placental bikunin(1-170),2 was expressed and found to be a potent serine protease inhibitor. Using the recombinant protein and both of its synthetic NH2- and COOH-terminal Kunitz domains, we dissect the specificity of the protein more comprehensively. In doing so, we demonstrate that placental bikunin is a potent inhibitor of plasma and tissue kallikreins, plasmin and factor XIa.


EXPERIMENTAL PROCEDURES

Materials

Bovine chymotrypsin, bovine trypsin, and tPA (single chain form from human melanoma cell culture), urokinase, Suc-A-A-P-F-AMC, Boc-L-G-R-AMC, Suc-A-A-A-AMC, Boc-Q-G-R-AMC, Bz-P-F-R-pNa, and P-F-R-AMC were from Sigma. Neutrophil elastase was from Athens Research and Technology, Inc. (Athens, GA). Human plasmin, factor VIIa, tissue factor (lipidated), and (CH3SO2-D-cyclohexylpyrosyl)-G-R-pNa (Spectrozyme tPA substrate) were from American Diagnostica, Inc. (Greenwich, CT). Human plasma kallikrein and human factors IXa, X, Xa, XIa, and XIIa were from Enzyme Research Laboratories (South Lafayette, IN). Tos-G-P-K-AMC, Suc-A-A-P-R-pNa, and Boc-E(OBzl)-A-R-AMC, were from Bachem Bioscience (King of Prussia, PA). H-D-I-P-R-pNa was from Chromogenix AB. Bovine pancreatic kallikrein, human tissue kallikrein, and recombinant aprotinin were kindly provided by Bayer AG (Wuppertal, Germany).

Expression and Purification of Placental Bikunin(1-170)

A cDNA fragment encoding an NH2-terminal 170-amino acid fragment of placental bikunin was expressed in Sf9 cells (7) as follows. Placental bikunin cDNA obtained by polymerase chain reaction and contained within the pCRII vector as described previously (6) was liberated by digestion with HindIII and XbaI. This fragment was gel purified and then cloned into the M13mp19 vector (New England Biolabs, Beverly, MA). In vitro mutagenesis (8) was used to generate a PstI site between the upstream XbaI site and the sequence encoding the ATG start site at the 5' end of the cDNA insert. The oligonucleotide used for the mutagenesis had the sequence CGC GTC TCG GCT GAC CTG GCC CTG CAG ATG GCG CAC GTG TGC GGG. A stop codon (TAG) and BglII/XmaI site was similarly engineered at the 3' end of the cDNA using the oligonucleotide CTG CCC CTT GGC TCA AAG TAG GAA GAT CTT CCC CCC GGG GGG GTG GTT CTG GCG GGG CTG. The stop codon was in-frame with the sequence encoding placental bikunin and caused termination immediately following the lysine at amino acid residue 170 of the mature protein sequence. The mutated cDNA thus encoded a truncated placental bikunin protein containing both Kunitz domains but which was devoid of the putative transmembrane domain. The product from digestion with PstI and BglII was isolated and cloned into the BacPac8 vector (CLONTECH) for expression of placental bikunin(1-170). The expression of placental bikunin(1-170) by Sf9 insect cells was optimal at a multiplicity of infection of 1 when the medium was harvested at 72 h post-infection. Sf9 cell culture supernatant (2-3 liters) was harvested by centrifugation (1,500 × g for 30 min), adjusted to pH 8.0 by the addition of 1 M Tris-HCl (pH 8.0) to a final concentration of 50 mM, then applied at 2.0-ml min-1 to a 5-ml column comprised of bovine pancreatic kallikrein (70 mg) which had been immobilized onto CNBr-activated Sepharose 4B (Pharmacia Biotech Inc.) according to the manufacturer's instructions. After loading, the column was washed with 0.1 M Tris-HCl (pH 8.0) containing 0.1 M NaCl until the A280 nm of the wash could no longer be detected. The column was washed further with 0.1 M Tris-HCl (pH 8.0) containing 0.5 M NaCl and then eluted with 0.2 M acetic acid (pH 2.0). Fractions (2.0 ml) were collected in tubes containing 0.25 ml of 1 M Tris-HCl (pH 7.5) and assayed for the ability to inhibit bovine trypsin as described (6). Active fractions were pooled, adjusted to pH 2.5 with trifluoroacetic acid and subjected to chromatography on a C18 reverse-phase column (1.0 × 25 cm) equilibrated with 0.1% trifluoroacetic acid containing 22.5% acetonitrile at a flow rate of 1 ml/min. The placental bikunin(1-170) was eluted with a linear gradient of 22.5-50% acetonitrile in 0.1% trifluoroacetic acid over 50 min. Fractions containing trypsin inhibitory activity were pooled, lyophilized, redissolved in 5 mM sodium acetate (pH 5.0), 0.1 M NaCl, and stored at -20 °C until needed. The highly purified protein exhibited an epsilon 280 nm of 4.52 × 104 liters M-1 cm-1 based upon composition analysis.

Synthesis and Refolding of the Individual Kunitz Domains of Placental Bikunin

Peptides corresponding to amino acids 7-64 and 102-159 of mature placental bikunin (6) were synthesized on an Applied Biosystems model 433A peptide synthesizer using NMP-HBTU Fmoc chemistry. Synthesis was on a preloaded Gln resin (Nova Biochem, La Jolla, CA) with an 8-fold excess of Fmoc-amino acid/coupling. The side chain protecting groups used were as follows: for cysteine, histidine, tryptophan, and asparagine, trityl; for glutamic acid, serine, tyrosine, aspartic acid, and threonine, t-butyl; for lysine, t-Boc; and for arginine, 2,2,5,7,8-pentamethylchroman-6-sulfonyl. In addition, glycine was protected with N-alpha -Fmoc-N-alpha -(2-Fmoc-oxy-4-methoxybenzyl) when it occurred after aspartic acid. Cleavage and deprotection were performed in 84.5% trifluoroacetic acid, 4.4% thioanisole, 2.2% ethanedithiol, 4.4% liquefied phenol, and 4.4% H2O for 2 h at room temperature. The crude peptide was precipitated with t-butyl methyl ether, centrifuged at 3,000 rpm in a Sorval RT 6000D model table top centrifuge for 5 min, and washed twice in t-butyl methyl ether. Reduced peptides were purified by HPLC (10 ml/min) on a Dynamax 60A C18 column (2.1 × 30 cm) using a 30-min linear gradient of either 23-34% acetonitrile in 0.1% trifluoroacetic acid (placental bikunin(7-64)) or a 30-min linear gradient of 20-30% acetonitrile in 0.1% trifluoroacetic acid (placental bikunin(102-159)).

The purified synthetic peptides corresponding to placental bikunin(7-64) and placental bikunin(102-159) were refolded to yield the respective functional NH2-terminal(7-64) and COOH-terminal(102-159) Kunitz domains using an adaptation of the method of Tam et al. (9) as follows. A solution containing 23% (v/v) dimethyl sulfoxide in 0.1 M Tris-HCl (pH 6.0) was added dropwise to a solution of reduced peptide (2.1 and 2.9 mg/ml for the NH2-terminal(7-64) and COOH-terminal(102-159) domain, respectively) in 0.1 M Tris-HCl (pH 6.0), containing 8 M urea to obtain a final concentration of 0.3 mg/ml peptide in 20% dimethyl sulfoxide, 0.1 M Tris-HCl (pH 6.0), and 1 M urea. The solutions containing either domain were each stirred at 25 °C for 24 h and then diluted 1:10 with 50 mM Tris-HCl (pH 8.0) containing 0.1 M NaCl. Refolded material was isolated by affinity chromatography over immobilized bovine pancreatic kallikrein under conditions described above for the purification of recombinant placental bikunin(1-170). Fractions containing trypsin inhibitory activity were pooled, adjusted to pH 2.5 with trifluoroacetic acid, then applied directly to a Vydac C18 reverse-phase column (5 µm, 0.46 × 25 cm). The NH2-terminal(7-64) Kunitz domain was purified using a 40-min linear gradient of 22.5-50% acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 1.0 ml/min. The COOH-terminal(102-159) Kunitz domain was purified using a 40-min linear gradient of 20-40% acetonitrile in 0.1% trifluoroacetic acid. Active fractions were pooled, lyophilized, dissolved in 0.1% trifluoroacetic acid, and stored at -20 °C until needed.

Determination of the Stoichiometry of the Protease-Inhibitor Complex

To determine the stoichiometry with trypsin, the protease (2.2 nM) was incubated with recombinant placental bikunin (0-2 nM) or aprotinin (0-4 nM) in 1.0 ml of 50 mM Hepes (pH 7.5), 0.1 M NaCl, 2.0 mM CaCl2, 0.01% Triton X-100 (buffer A) for 30 min at room temperature, followed by the addition of Suc-A-A-P-R-pNa (90 µM, final concentration), after which residual enzymatic formation of pNa was monitored at 410 nm. To determine the stoichiometry with plasma kallikrein, 40 nM protease was incubated with recombinant placental bikunin (0-40 nM) in 1.0 ml of 50 mM Tris-HCl (pH 8.0) containing 0.1 M NaCl and 0.01% Triton X-100 (buffer B). After 30 min at room temperature, Bz-P-F-R-pNa was added (300 µM, final concentration) and residual enzymatic activity monitored at 410 nm. Fractional protease activity was calculated as V(+I)/V(-I) and stoichiometries obtained from plots of V(+I)/V(-I) versus ratio of inhibitor to enzyme concentrations.

Determination of Equilibrium Dissociation Constants

Apparent equilibrium dissociation constants (Ki*), were determined as described previously (6) using methods for tight binding inhibitors (10) assuming enzyme:inhibitor stoichiometries of 1:1 and 2:1 for binding to single Kunitz domains and placental bikunin(1-170), (1) respectively. Active site concentrations of trypsin, plasma and tissue kallikreins, and plasmin were determined by titration with p-nitrophenyl-p'-guanidinobenzoate as described (11). The concentration of bovine chymotrypsin was determined by titration with N-trans-cinnamoylimidazole as described (12). The amount of the NH2-terminal(7-64) and COOH-terminal (102-159) Kunitz domain, and aprotinin were determined by titration with active site-titrated trypsin and by amino acid analysis. The concentration of placental bikunin(1-170) was quantified by amino acid analysis. The concentrations of factors VIIa, IXa, X, Xa, XIa, and XIIa, tissue factor, elastase, urokinase, and tPA were based on the manufacturer's specifications. Following preincubation of protease with inhibitor for 5 min at 37 °C in 990 µl of the appropriate buffer (see below), reactions were initiated with substrate to achieve the following initial component concentrations: bovine pancreatic kallikrein (buffer B) with [E0] = 92 pM and 100 µM P-F-R-AMC (Km = 82 ± 8.8 µM); human tissue kallikrein, 50 mM Tris-HCl (pH 9.0), 50 mM NaCl, and 0.01% Triton X-100 with [E0] = 0.35 nM and 10 µM P-F-R-AMC (Km = 5.7 ± 0.4 µM); human neutrophil elastase, 0.1 M Tris-HCl (pH 8.0), and 0.05% Triton X-100 with [E0] = 19 nM and 0.6 mM Suc-A-A-A-AMC (Km = 1,300 µM); human factor Xa, 20 mM Tris-HCl (pH 8.0), 0.1 M NaCl, 0.1% bovine serum albumin with [E0] = 0.87 µM and 0.6 mM Boc-L-G-R-AMC (Km = 750 ± 98 µM); and human factor XIa (buffer A with 1 mg/ml bovine serum albumin) with [E0] = 0.1 nM and 400 µM Boc-E(OBzl)-A-R-AMC (Km = 450 ± 50 µM). The following assays were performed following preincubation of protease with inhibitor for 30 min at 37 °C: trypsin (buffer A) with [E0] = 50 pM and 30 µM Tos-G-P-K-AMC (Km = 22 ± 2 µM); chymotrypsin (buffer A) with [E0] = 50 pM and 80 µM Suc-A-A-P-F-AMC (Km = 70 ± 12 µM); human plasmin, 50 mM Tris-HCl (pH 7.5), 0.1 M NaCl, and 0.02% Triton X-100 with [E0] = 50 pM and 500 µM Tos-G-P-K-AMC (Km = 726 ± 70 µM); human plasma kallikrein (buffer B) with [E0] = 0.2 nM and 100 µM P-F-R-AMC (Km = 457 ± 28 µM); and factor XIIa (buffer B) with [E0] = 4.0 nM, and 200 µM Boc-Q-G-R-AMC (Km = 200 ± 37 µM). Factor VIIa-tissue factor inhibition was measured by incubating enzyme (2.8 nM factor VIIa, 2.8 nM tissue factor) with inhibitor in 100 µl of buffer A for 30 min at 25 °C. Reactions were initiated with 1 mM final (CH3SO2-D-cyclohexylpyrosyl)-G-R-pNa, and the A405 nm was monitored. Factor IXa inhibition was measured as described (5). Briefly, factor IXa (10 nM) was preincubated for 15 min at 37 °C with inhibitor in 50 mM Tris-HCl (pH 7.5) containing 0.1 M NaCl, 250 µM CaCl2, 60 nM poly-D-lysine (Mr = 209,200) and 0.1% polyethylene glycol 8000. Factor X (200 nM) was then added and the incubation continued for 30 min at 37 °C. Substrate (600 µM Boc-L-G-R-AMC) was then added and factor Xa activity measured. To determine the inhibition of tPA, tPA was preincubated with inhibitor for 2 h at room temperature in 20 mM Tris-HCl (pH 7.2) containing 150 mM NaCl and 0.02% sodium azide. Reactions were initiated with substrate to achieve initial component concentrations of: tPA (16.7 nM), inhibitor (0-6.6 µM), I-P-R-pNa (1 mM) in 28 mM Tris-HCl (pH 8.5) containing 0.004% (v/v) Triton X-100 and 0.005% (v/v) sodium azide. The amount of pNa formed after incubation for 2 h at 37 °C was determined from the increase in the A405 nm. Hydrolysis of AMC-conjugated peptides was monitored on a Perkin-Elmer model LS50B fluorometer (excitation = 370 nm, emission = 432 nm) over the first 2 min of the reaction, while hydrolysis of pNa conjugates was monitored at 405 nm on a Hewlett-Packard model HP8452 spectrophotometer. Ki* values were determined from plots of fractional rate versus inhibitor concentration, which were fit by nonlinear regression analysis (Enzfitter by Biosoft, Cambridge, U. K.) using the following equation
V<SUB>i</SUB>/V<SUB>0</SUB>=1−([E]<SUB>0</SUB>+[<UP>I</UP>]<SUB>0</SUB>+K<SUB>i</SUB>*)−[([E]<SUB>0</SUB>+[<UP>I</UP>]<SUB>0</SUB> (Eq. 1)
+K<SUB>i</SUB>*)<SUP>2</SUP>−4[E]<SUB>0</SUB>[<UP>I</UP>]<SUB>0</SUB>]<SUP>1/2</SUP>/2[E]<SUB>0</SUB>
where Vi and V0 are the enzyme activities in the presence or absence of a total inhibitor concentration of [I]0, and [E]0 is the total concentration of enzyme. Ki values were obtained by correction for the effect of substrate using the following equation: Ki = Ki*/(1 + [S]0/Km). Ki values for inhibition by placental bikunin(1-170) were determined assuming an inhibitor:protease stoichiometry of 1:2 and that the two Kunitz inhibitory domains are equivalent and act independently.

Measurement of the Activated Partial Thromboplastin Time (APTT)

The effect of recombinant placental bikunin(1-170) and aprotinin on the APTT was determined as follows. Inhibitor in 20 mM Tris-HCl (pH 7.2) containing 150 mM NaCl and 0.02% sodium azide was added (0.1 ml) to a cuvette within a MLA ElectraR 800 Automatic Coagulation Timer coagulometer (Medical Laboratory Automation, Inc., Pleasantville, NY). The instrument was set to APTT mode with a 300-s activation time, and samples were run in duplicate. Following the addition of 0.1 ml of plasma (Specialty Assayed Reference Plasma 1-6-5185, Helena Laboratories, Beaumont, TX), the APTT reagent (Automated APTT-lot 102345, Organon Teknika Corp., Durham, NC) and CaCl2 (final concentration of 25 mM) were automatically added to initiate clotting, which was monitored automatically.

Analytical Procedures

CNBr digestion was performed as follows. Purified placental bikunin(1-170) in water (165 pmol) was taken to dryness on a Speed-Vac concentrator (Savant Instruments, Farmingdale, NY), resuspended in 0.05 ml of 250 mM Tris-HCl containing 6 M guanidine HCl, 0.5% (v/v) 2-mercaptoethanol, and 1 mM EDTA, and incubated for 2 h in the dark at room temperature before the addition of 4% (v/v) final 4-vinylpyridine. Following an additional 2-h incubation, the sample was applied to a biphasic reaction cartridge containing C18 (Hewlett-Packard), eluted in methanol, then taken to dryness by Speed-Vac concentration. The sample was incubated with 0.05 ml of 70% (v/v) formic acid saturated with CNBr for 12 h, taken to dryness, reconstituted in 0.1 ml of 0.1% (v/v) trifluoroacetic acid, then subjected to NH2-terminal sequencing. NH2-terminal sequence analysis was performed essentially as described (13), but with minor modification (6). Amino acid analysis was performed as described previously (14). Reducing SDS-polyacrylamide gel electrophoresis was performed with 10-20% Tricine-buffered polyacrylamide gels (Novex, San Diego, CA) according to the manufacturer's instructions, and protein was visualized with Coomassie Brilliant Blue R-250 using standard protocols. Electrospray ionization mass spectrometry was performed on a Finnigan TSQ7000 mass spectrophotometer capable of unit resolution, with the following operational parameters: scan rate, 200-2,000 over 3 s; spray voltage, 5 kV; capillary temperature, 220 °C; nitrogen pressure, 50 p.s.i.; auxillary nitrogen, 30 ml/min. Samples were loaded on a PLRP cartridge (Microm Bioresources, Auburn, CA), washed with water, and eluted at 0.2 ml/min with 90% (v/v) acetonitrile, 2% (v/v) acetic acid.


RESULTS

To explore the protease inhibitory capacity of placental bikunin, a cDNA encoding the entire extracellular domain of 170 amino acids of placental bikunin plus its natural signal peptide (6) was prepared by site-directed mutagenesis and expressed in the baculovirus/Sf9 system. The protein was purified from the cell culture supernatant by sequential kallikrein-Sepharose affinity chromatography and C18 reverse-phase HPLC (Table I). An overall enrichment of 240-fold was achieved based on the increase in specific activity; however, the actual fold enrichment of the protein was likely to be much higher as only a fraction of the inhibitory activity of the culture supernatant was due to placental bikunin(1-170). This latter conclusion is based on the fact that the majority (>99%) of the trypsin inhibitory activity present in the starting supernatant did not bind to the kallikrein-Sepharose column, whereas all the placental bikunin(1-170) bound the column based on immunoblot analysis of the column starting material and flow-through (not shown). Fractionation of 2 liters of medium typically yielded between 0.7 and 1 mg of placental bikunin(1-170) based on amino acid analysis.

Table I. Purification of recombinant placental bikunin(1-170) from Sf9 cells


Purification step Volume A280 total Unitsa Specific activity

ml units/A
Supernatant 2,300.0 20,700 6,150,000 297
Kallikrein affinity 23.0 2.76 40,700 14,746
C18 0.4 1.54 11,111 72,150

a A unit of activity is the amount of inhibitor needed to inhibit trypsin activity by 50% with the conditions specified under "Experimental Procedures."

The final preparation was highly pure as judged by SDS-polyacrylamide gel electrophoresis (Fig. 1) and exhibited a molecular mass of 21.3 kDa, consistent with the expected size of placental bikunin(1-170). NH2-terminal sequence analysis of the protein (26 cycles) yielded the same NH2 terminus for as obtained for the natural protein (6). The sequence started at residue +1 and continued with the sequence ADRER-. Furthermore, the amino acid composition was > 95% accurate relative to the theoretical composition of placental bikunin(1-170). Purified placental bikunin(1-170) (100 pmol) was then pyridylethyl alkylated, CNBr digested, and then sequenced without purification of the resulting peptide fragments. Sequencing for 20 cycles yielded the sequence for each of the NH2 termini expected from the digestion of placental bikunin(1-170) (Table II).


Fig. 1. SDS-polyacrylamide gel electrophoresis analysis of placental bikunin(1-170) highly purified from a baculovirus/Sf9 expression system. Three µg of highly purified placental bikunin (lane 2) and molecular size markers of the indicated sizes in kDa (lane 1) were resolved on a 10-20% Tricine-SDS-polyacrylamide gel and developed with Coomassie Blue.
[View Larger Version of this Image (50K GIF file)]

Table II. NH2-terminal sequences of fragments resulting from the digestion of purified placental bikunin(1-170) with CNBr

Digestion was performed according to "Experimental Procedures." Digestion was performed according to "Experimental Procedures."
Sequencea Amount Placental bikunin residue no.

pmol
LRCFrQQENPP-PLG 21 154 -168
ADRERSIHDFCLVSKVVGRC 20 1 -20
FNYeEYCTANAVTGPCRASF 16 100 -119
Pr--Y-V-dGS-Q-F-Y-G 6 25 -43

a Lowercase letters denote tentative amino acid assignments.

Titration of fixed concentrations of either bovine trypsin or human plasma kallikrein with purified placental bikunin(1-170) showed that each molecule of inhibitor can interact with two molecules of protease simultaneously (Fig. 2). On the other hand, titration of trypsin with aprotinin yielded a stoichiometry of 1:1 in the enzyme inhibitor complex (Fig. 2).


Fig. 2. Stoichiometry of complex formation between placental bikunin(1-170) and serine proteases. Increasing concentrations of placental bikunin(1-170) were incubated with a fixed concentration of either trypsin (bullet ) or kallikrein (square ) under stoichiometric conditions followed by determination of residual protease activity. Titration of trypsin with aprotinin (open circle ) is shown for comparison. Stoichiometries are obtained from the x axis intercept of the tangent to curves passing through the y axis intercept.
[View Larger Version of this Image (16K GIF file)]

To attempt to dissect the specificities of the individual domains within this bikunin, synthetic peptides were made corresponding to the NH2-terminal(7-64) and COOH-terminal(102-159 Kunitz domains of placental bikunin. Purified, reduced NH2-terminal(7-64) ([M-H]+ = 6,567.5), and COOH-terminal(102-159) domain ([M-H]+ = 6,835.6) were prepared, each of which had the expected amino acid composition. Refolding of these peptides using dimethyl sulfoxide as the oxidizing agent, followed by purification on a C18 column, yielded purified refolded NH2-terminal(7-64) (2% of the starting reduced peptide) and COOH-terminal(102-150) domains (1.4% of the starting reduced peptide) which exhibited [M-H]+ values of 6,561.2 and 6,829.3, respectively. The mass reduction upon refolding of each peptide (6 ± 1 mass units) suggests that refolding had in each case resulted in the formation of three intrachain disulfide bonds from the six cysteines present within each of the reduced peptides.

A panel of serine proteases was used to determine the relative specificity and potency of recombinant placental bikunin(1-170) and its individual Kunitz domains (Table III). Aprotinin, a Kunitz inhibitor of defined specificity, was included in these studies for comparison. Recombinant placental bikunin(1-170) was a potent inhibitor of several trypsin-like proteases and also of chymotrypsin. Of potential relevance to its putative physiological role, this protein was a potent inhibitor of human plasma and tissue kallikreins, human plasmin and factor XIa. The potency of recombinant placental bikunin(1-170) against plasmin, plasma kallikrein, and factor XIa were 2-, 45-, and 47-fold greater, respectively, than observed with aprotinin. Recombinant placental bikunin(1-170) was also a more potent inhibitor than aprotinin against several other enzymes of the intrinsic pathway of coagulation, including factors IXa, Xa, and XIIa, although the Ki values were in the hundred nanomolar range.

Table III. Comparison of equilibrium dissociation constants for the interaction of placental bikunin(1-170) its individual Kunitz domains, and aprotinin with various serine proteases

Equilibrium dissociation constants (Ki values) are listed in nM. Values are the mean ± S.D. determined from (n) titration points. Equilibrium dissociation constants (Ki values) are listed in nM. Values are the mean ± S.D. determined from (n) titration points.
Protease NH2-terminal(7-64) Kunitz domain COOH-terminal(102-159) Kunitz domain Placental bikunin(1-170)a Aprotinin

Trypsinb 0.03  ± 0.01 (13) 0.05  ± 0.01 (13) 0.01  ± 0.002 (12) 0.02  ± 0.003 (13)
Chymotrypsinb 1.7  ± 0.54 (11) 2.1  ± 0.2 (12) 0.48  ± 0.15 (11) 1.3  ± 0.43 (8)
Plasma kallikreinc 0.67  ± 0.14 (8) 0.73  ± 0.09 (11) 0.31  ± 0.04 (9) 13.4  ± 2.4 (11)
Pancreatic kallikreinb 0.42  ± 0.09 (11) 0.49  ± 0.13 (11) 0.26  ± 0.06 (10) 0.023  ± 0.006 (6)
Tissue kallikreinc 2.3  ± 0.5 (10) 0.13  ± 0.02 (7) 0.13  ± 0.03 (10) 0.004  ± 0.003 (7)
Plasminc 1.04  ± 0.25 (11) 0.54  ± 0.11 (9) 0.10  ± 0.03 (9) 0.18  ± 0.06 (11)
Factor IXa/Xc,d N.I.e (2 µM) 507  ± 13 (5) 206  ± 14 (6) N.I.e (5 µM)
Factor Xac N.D.f 274  ± 60 (7) 364  ± 97 (8) N.I.e (30 µM)
Factor XIac 107  ± 25 (7) 15  ± 3.8 (8) 5.9  ± 1.4 (11) 288  ± 71 (6)
Factor XIIac 477  ± 140 (6) 800  ± 245 (6) 429  ± 134 (6) 6800  ± 2,840 (7)
FVIIa-tissue factorc,d N.D.f N.D.f 1606  ± 264 N.I.e (1 µM)

a Ki values for placental bikunin(1-170) were determined by nonlinear regression analysis assuming [I]0 equal to twice the molar protein concentration.
b Enzymes from bovine sources.
c Enzymes from human sources.
d Apparent equilibrium dissociation constant.
e N.I., no inhibition at the highest inhibitor concentration tested (value in parentheses).
f N.D., not determined.

Collectively, these data suggest that recombinant placental bikunin(1-170) should be a more effective inhibitor of the intrinsic coagulation pathway than aprotinin. To evaluate this hypothesis further, placental bikunin(1-170) and aprotinin were compared for their capacity to prolong the APTT, a measure of the intrinsic pathway of coagulation. As predicted from the in vitro specificity data (Table III), recombinant placental bikunin(1-170) was significantly more potent than aprotinin in prolonging the APTT (Fig. 3). To achieve a doubling of the clotting time, at least 28 µM aprotinin but only 0.7 µM recombinant placental bikunin(1-170) was required.


Fig. 3. Prolongation of the activated partial thromboplastin time of human plasma by placental bikunin(1-170) and aprotinin. The fold prolongation in clotting time initiated by CaCl2 is plotted versus the concentration of recombinant placental bikunin(1-170) (bullet ) and aprotinin (square ). The uninhibited clotting time was 29.6 s.
[View Larger Version of this Image (11K GIF file)]

Recombinant placental bikunin(1-170) was a poor inhibitor of the factor VIIa-tissue factor complex and factor Xa, suggesting that it may be a poor inhibitor of the extrinsic pathway of coagulation. This prediction was confirmed by the finding that placental bikunin(1-170) had no significant effect on the prothrombin time when tested at concentrations up to 2 µM (data not shown). Recombinant placental bikunin(1-170) did not inhibit urokinase, tPA, or elastase at concentrations up to 1 µM (data not shown).

The individual Kunitz domains within placental bikunin were each functional protease inhibitors and exhibited roughly similar specificities, although the NH2-terminal(7-64) Kunitz domain was significantly less potent against tissue kallikrein and factor XIa. Only versus factor XIIa was the NH2-terminal domain more potent than the COOH-terminal domain. Overall, the specificity of the COOH-terminal Kunitz(102-159) domain more closely resembled that of recombinant placental bikunin(1-170) except that on average the potency of the COOH-terminal Kunitz(102-159) for each protease was approximately 5-fold lower.


DISCUSSION

In this report we describe the expression and functional characterization of placental bikunin. Because the natural protein was available only in low amounts (6) the functional studies were performed instead with a recombinant form of the protein. A cDNA sequence encoding a 170-residue placental bikunin fragment containing both Kunitz domains but truncated immediately prior to the putative transmembrane segment (6) was expressed in Sf9 cells. This was done to alleviate concerns that expression of the full-length protein would complicate its recovery as a soluble protein. The putative signal peptide encoded within the cloned cDNA (6) was recognized as such by the Sf9 system as judged by its absence from the NH2-terminal sequence of the purified protein. For comparison with the recombinant protein, the individual Kunitz domains were prepared synthetically and effectively folded as judged by reductions in mass consistent with formation of the expected number of intrachain disulfide bonds. Recombinant placental bikunin(1-170) and the synthetic NH2-terminal(7-64) and COOH-terminal(102-159) Kunitz domains were all potent serine protease inhibitors. Each protein was a potent inhibitor of trypsin, chymotrypsin, plasmin, factor XIa, and tissue and plasma kallikrein, yet a relatively weak inhibitor of factors IXa, Xa, and XIIa. Stoichiometry studies with placental bikunin(1-170) verified that both Kunitz domains were functional and could direct the formation of ternary complexes with selected target proteases. The fact that placental bikunin(1-170) was functional lends credence to the notion that the hydrophobic segment encoded by the full-length cDNA acts as a membrane anchor and does not participate in the globular structure necessary for protease inhibition.

Although placental bikunin(1-170) potently inhibited two key enzymes, kallikrein and factor XIa, of the contact activation/intrinsic pathway, it was an extremely weak inhibitor of factor VIIa-tissue factor complex. This suggests that it is unlikely involved in the regulation of the extrinsic coagulation cascade. Consistent with these observations, placental bikunin (1-170) was extremely potent in extending the APTT time, a measure of the rate of coagulation driven by the intrinsic pathway, but did not extend the prothrombin time, a measure of the rate of coagulation driven by the extrinsic pathway. Based on these findings, placental bikunin exhibits an in vitro activity profile that might be expected of a physiological regulator of both contact activation and the intrinsic pathway of coagulation. On the other hand, the kallikreins play a key upstream role in initiating fibrinolysis through activation of plasmin, which in turn participates directly in fibrinolysis. The high potency of placental bikunin(1-170) against the kallikreins and plasmin suggests that this inhibitor could function to stabilize fibrin clots. The possibility that the native full-length protein might associate with biological membrane is consistent with the notion that it may serve to regulate the processes of contact activation and fibrinolysis since elements of these processes take place at damaged tissue surfaces.

In general, placental bikunin(1-170) exhibited Ki values that were significantly lower than those observed with the individual domains. Similar results have been obtained with TFPI wherein the recombinant protein containing all three of the Kunitz domain had anywhere from a 2-fold to 10-fold greater affinity for the factor VII-tissue factor complex than any of the individual domains (15). Furthermore, cooperatively has been demonstrated for the interaction of TFPI with its target proteases. Specifically, Girard and Broze demonstrated that TFPI inhibits the tissue factor-factor VIIa complex 50-fold more potently when first allowed to complex with factor Xa (16). From a biological perspective this is consistent with a feedback inhibition mechanism allowing some active factor Xa to form before the entire extrinsic pathway terminates. The possibility that placental bikunin may regulate the intrinsic pathway through a similar cooperative mechanism warrants further investigation. We have demonstrated that although the NH2- and COOH-terminal Kunitz domains have similar potencies versus a number of proteases, the COOH-terminal Kunitz domain is 10-fold more potent against factor XIa. Perhaps the COOH-terminal domain of placental bikunin inhibits factor XIa in vivo, which in turn triggers the subsequent formation of a ternary complex through interaction of the NH2-terminal domain with factor XIIa or kallikrein. This would serve as a feedback inhibition mechanism for the extrinsic pathway where some factor XIa is allowed to form before the entire pathway is shut down. On the other hand, the differences in potency against factor XIa between the synthetic NH2- and COOH-terminal Kunitz domains could reflect the fact that the putative N-linked carbohydrate residue within the NH2-terminal domain of native placental bikunin (6) plays a role in determining the specificity of this domain.

Blockage of fibrinolysis and extracorporeal coagulation have been observed in open heart surgery patients receiving aprotinin (17, 18), and these effects are believed to arise from the inhibition of kallikrein, plasmin, and the intrinsic pathway of coagulation (19). Because placental bikunin(1-170) is significantly more potent than aprotinin as an inhibitor of plasma kallikrein, and the intrinsic pathway of coagulation, further exploration of the use of this novel human bikunin in this clinical setting is warranted. In addition, placental bikunin should also be considered for other indications for which aprotinin has been shown to be beneficial preclinically. One potential application is the blockage of the plasmin-dependent invasion of cancer cells into surrounding tissue (20). In addition, placental bikunin may be useful for the reduction of severity of brain edema secondary to brain or spinal chord injury (21, 22), a process that may result from the local release of the vasodilator bradykinin (21-23) through the actions of serine proteases such as kallikrein (24).


FOOTNOTES

*   The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
**   To whom correspondence should be addressed: Institute for Research Technologies, Bayer Research Center, 400 Morgan La., West Haven, CT 06516-4175. Tel.: 203-812-2920; Fax: 203-812-2526; E-mail: pault{at}wh.bayer.com.
1   The abbreviations used are: tPA, tissue plasminogen activator; TFPI, tissue factor pathway inhibitor; Suc, succinyl; AMC, 7-amido-4-methylcoumarin; Boc, t-butoxycarbonyl; Bz, benzoyl; pNa, p-nitroanilide; Tricine, N-[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]glycine; Fmoc, N-(9-fluorenyl)methoxycarbonyl; HPLC, high performance liquid chromatography; APTT, activated partial thromboplastin time.
2   Residues of the placental bikunin sequence and fragments thereof are numbered as defined in Ref. 6 according to their position within the open reading frame for full-length mature placental bikunin. In this scheme, placental bikunin(1-170) refers to a protein containing the NH2-terminal 170 amino acids of the mature protein produced by removal of the signal peptide. The NH2-terminal(7-64) domain and COOH-terminal(102-159) domain refer to individual functional Kunitz domains normally present within residues 7-64 and 102-159, respectively, of mature placental bikunin.

ACKNOWLEDGEMENTS

We acknowledge the following members of the research staff at the Bayer Research Center for important contributions: Carla Pellegrino (NH2-terminal sequencing), Thomas Buckholz (peptide synthesis), John Kupcho (amino acid analysis), Susan Barkowski-Clark and Bradley Rose (Sf9 expression), and Tony Paiva (mass spectrometry).


REFERENCES

  1. Mann, K. G., and Lorand, L. (1993) Methods Enzymol. 222, 1-10 [CrossRef][Medline] [Order article via Infotrieve]
  2. Marder, V. J., Frances, C. W., and Doolittle, R. F. (1982) in Hemostasis and Thrombosis: Basic Principles and Clinical Practice (Colman, R. W., Hirsh, J., Marder, V. J., and Salzman, E. W., eds), pp. 145-163, Lippincott, Philadelphia
  3. Davie, E. W., Fujikawa, K., and Kisiel, W. (1991) Biochemistry 30, 10363-10370 [Medline] [Order article via Infotrieve]
  4. Travis, J., and Salvesen, G. S. (1983) Annu. Rev. Biochem. 52, 655-709 [CrossRef][Medline] [Order article via Infotrieve]
  5. Petersen, L. C., Sprecher, C. A., Foster, D. C., Blumberg, H., Hamamoto, T., and Kisiel, W. (1996) Biochemistry 35, 266-272 [CrossRef][Medline] [Order article via Infotrieve]
  6. Marlor, C. W., Delaria, K. A., Davis, G., Muller, D. K., Greve, J. M., and Tamburini, P. P. (1997) J. Biol. Chem. 272, 12202-12208 [Abstract/Free Full Text]
  7. Luckow, V. A., and Summers, M. D. (1989) Virology 170, 31-39 [Medline] [Order article via Infotrieve]
  8. Kunkel, T. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 488-492 [Abstract]
  9. Tam, J. P., Wu, C.-R., Liu, W., and Zhang, J.-W. (1991) J. Am. Chem. Soc. 113, 6657-6662
  10. Boudier, C., and Bieth, J. G. (1989) Biochim. Biophys. Acta 995, 36-41 [Medline] [Order article via Infotrieve]
  11. Chase, T., and Shaw, E. (1970) Methods Enzymol. 19, 20-27
  12. Schonbaum, G. R., Zerner, B., and Bender, M. L. (1961) J. Biol. Chem. 236, 2930-2935 [Medline] [Order article via Infotrieve]
  13. Miller, C. G. (1994) Methods: Companion Methods Enzymol. 6, 315-333 [CrossRef]
  14. Tamburini, P. P., Dreyer, R. N., Hansen, J., Letsinger, J., Elting, J., Gore-Willse, A., Dally, R., Hanko, R., Osterman, D., Kamarck, M. E., and Yoo-Warren, H. (1990) Anal. Biochem. 186, 363-368 [Medline] [Order article via Infotrieve]
  15. Petersen, L. C., Bjorn, S. E., Olsen, O. H., Nordfang, O., Norris, F., and Norris, K. (1996) Eur. J. Biochem. 234, 310-316
  16. Girard, T. J., and Broze, G. J., Jr. (1993) Methods Enzymol. 222, 195-209 [Medline] [Order article via Infotrieve]
  17. Davis, R., and Whittington, R. (1995) Drugs 49, 954-983 [Medline] [Order article via Infotrieve]
  18. Dietrich, W., Barankay, A., Dilthey, G., Henze, R., Nikeau, E., Sebening, F., and Richter, J. A. (1989) Thorac. Cardiovasc. Surg. 37, 92-98 [Medline] [Order article via Infotrieve]
  19. Fritz, H., and Wunderer, G. (1983) Arzneimittel-Forshung. Drug Res. 33, 479-494
  20. Liu, G., Schuman, M. A., and Cohen, R. L. (1995) Int. J. Cancer 60, 501-506 [Medline] [Order article via Infotrieve]
  21. Kamiya, T., Katayama, Y., Kashigawi, F., and Terashi, A. (1993) Stroke 24, 571-575 [Abstract]
  22. Unterberg, A., Dautermann, C., Baethmann, A., and Muller-Esterl, W. (1986) J. Neurosurg. 64, 269-276 [Medline] [Order article via Infotrieve]
  23. Xu, J., Hsu, C. Y., Junker, H., Chao, S., Hogan, E. L., and Chao, J. (1984) J. Neurochem. 57, 975-980 [Medline] [Order article via Infotrieve]
  24. Coleman, R. W. (1984) J. Clin. Invest. 73, 1249-1253 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.