| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Hematology and Oncology, Department of Internal Medicine, University of Michigan, Ann Arbor 48109-0640; 2Thromgen, Incorporated, Ann Arbor 48104; and 3Department of Chemistry, Michigan State University, East Lansing, Michigan 48824
Submitted 11 June 2002 ; accepted in final form 3 February 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-strand with
Ser214-Gly216 and interacts with His57,
Asp189, and Ser195 of the catalytic triad. RPPGF
competitively inhibits 
-thrombin from hydrolyzing
Sar-Pro-Arg-paranitroanilide with a Ki = 1.75 ±
0.03 mM. Other mechanisms were sought to explain why RPPGF inhibits thrombin
activation of platelets at concentrations below that which inhibits its active
site. Soluble RPPGF blocks biotinylated NATLDPRSFLLR of the thrombin cleavage
site on protease-activated receptor (PAR)1 from binding to the peptide RPPGC
(IC50 = 20 µM). The soluble recombinant extracellular domain of
PAR1 (rPAR1EC) blocks biotinylated RPPGF binding to
rPAR1EC (IC50 = 50 µM) bound to microtiter plates,
but rPAR1EC deletion mutants missing the sequence LDPR or PRSF do
not. RPPGF and related forms prevent the thrombin-like enzyme thrombocytin
from proteolyzing rPAR1EC at concentrations that do not block
thrombocytin's active site. These studies indicate that RPPGF is a
bifunctional inhibitor of thrombin: it binds to PAR1 to prevent thrombin
cleavage at Arg41 and interacts with the active site of
-thrombin.
protease-activated receptor 1; bradykinin-(1–5); thrombin inhibitor; thrombin receptor
-thrombin-induced platelet activation
(3). RPPGF and related
compounds inhibit both - and 
-thrombin-induced platelet
aggregation and secretion without inhibiting ADP-, collagen-, or
U-46619-induced platelet activation
(3). RPPGF does not inhibit
thrombin binding to platelets, SFLLRN-induced platelet activation, or
-thrombin cleavage of H-D-Phe-Pip-Arg-paranitroanilide
(3). It directly binds to the
platelet membrane (5). RPPGF
prevents -thrombin from cleaving the peptide NATLDPRSFLLR, which spans
the thrombin cleavage site on protease-activated receptor 1 (PAR1), between
Arg and Ser (3). The importance of RPPGF and related compounds has been shown in animal studies (5, 6). High concentrations of RPPGF are as effective as aspirin in preventing electrolytic injury-induced left circumflex coronary artery thrombosis in dogs (5). Combined RPPGF and aspirin result in a summating antithrombotic effect (5). In the Folts canine model for induced cyclic flow variations, a multiantigenic peptide (MAP) form of RPPGF, MAP4-RPPGF, significantly reduced cyclic flow variations similar to aspirin or clopidogrel (6). In an ex vivo model of balloon angioplasty injury to the vessel wall, RPPGF was as effective as aspirin in preventing platelet adherence to the vessel wall and fibrin formation (20). The present investigation characterizes the mechanism(s) of RPPGF inhibition of thrombin. RPPGF inhibits thrombin activation of platelets in two ways. It directly binds to the active site of thrombin, and it binds to the extracellular domain of PAR1 to prevent thrombin cleavage after Arg41.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-thrombin (3,250 or 3,000 U/mg),
-thrombin, and hirudin were purchased from Hematologic Technologies
(Essex Junction, VT). Bovine serum albumin was purchased from Sigma (St.
Louis, MO). Thrombocytin was a generous gift from the late S. Niewiarowski,
Temple University (Philadelphia, PA).
Peptides. Peptides RPPGF, RPPGC, MAP4-RPPGF, FPRPG, GPRP, and
NATLDPRSFLLR (NAT12) were synthesized by the Protein and Carbohydrate
Structure Facility, University of Michigan (Ann Arbor, MI), and at Multiple
Peptide System (San Diego, CA). MAP4-RPPGF is a multiantigenic form of RPPGF
consisting of a -Ala core with an attached Lys followed by two attached
Lys and four molecules of RPPGF attached to the two latter Lys
(6). Each peptide was made with
the COOH-terminal amino acid covalently attached to a solid phase support, and
succeeding amino acids were coupled sequentially to the NH2
terminus (4,
6). Additional peptides,
biotinylated RPPGF (biotin-RPPGF) and biotinylated NATLDPRSFLLR-amide
(biotin-NAT12), were synthesized at Quality Controlled Biochemical (Hopkinton,
MA) and at Multiple Peptide System. All peptides were purified by preparative
reverse-phase HPLC and characterized by analytic HPLC, amino acid analysis,
and mass spectrometry. Each peptide was colorless, odorless, and soluble in
water. Each peptide is described using the single capital letter designation
for each L-amino acid. Some longer peptides are named by the single
letter code for their first three amino acids, followed by the total number of
amino acids in the peptide.
Antibodies. Polyclonal antibodies to the thrombin cleavage sites on its three cell receptors were raised in goats at Quality Controlled Biochemical, by immunizing with peptides 35NATLDPRSFLLRN47,35KPTLPIKTFRGAP47, or 41SILPAPRGYPGQV53 from the extracellular domain of human PAR1, PAR3, or PAR4, respectively. The numbering listed for these peptides here and elsewhere is that of the amino acid sequence on the extracellular domains of each of these human PARs (8, 27, 28). Each anti-PAR antibody was purified on affinity chromatography by coupling each peptide to AfFigel-15, Bio-Rad (Richmond, CA) in dimethyl sulfoxide. Anti-His6 monoclonal antibody (His-tag monoclonal antibody) was purchased from Novagen (Madison, WI). Monoclonal antibodies SPAN12 and WEDE15 were obtained from Immunotech, a subsidiary of Coulter (Miami, FL). SPAN12 was raised against peptide 35NATLDPRSFLLR46, and WEDE15 was raised against peptide 51KYEPFWEDEEKNES64, the hirudin-like domain on PAR1 (14).
Other materials. Full-length cDNA to human PAR1 was kindly provided by Dr. Lawrence Brass, University of Pennsylvania (Philadelphia, PA). A bacterial expression system with the pET19b vector was purchased from Novagen; fura 2-AM was obtained from Molecular Probes (Eugene, OR). Hirugen [sulfated acyl-hirudin53–65, N-acetyl-NGDFEEIPEEY (SO3) LQ] was obtained from Bachem Bioscience (King of Prussia, PA). Rabbit thrombomodulin (TM) was purchased from American Diagnostica (Greenwich, CT).
Ability of RPPGF to inhibit the coagulant activity of
-thrombin. The ability of RPPGF to inhibit the activated
partial thromboplastin time (APTT), prothrombin time (PT), or thrombin
clotting time (TCT) was investigated. The APTT and PT were performed as
previously reported in the absence or presence of 0.03125–3 mM RPPGF
(5,
6). The TCT experiments were
performed by incubating 100 µl of pooled normal human plasma (George King;
Overland Park, KS) in the absence or presence of 0.03125–3 mM RPPGF at
37°C for 5 min. At the end of incubation, 8 nM human -thrombin (1
U/ml) was added, and the time to clot formation was measured. All the
coagulant assays were performed in a Amelung KC4 coagulation meter (Sigma). In
other experiments, normal human plasma was made to 3 mM with L-Arg,
and the APTT, PT, or TCT was measured.
Crystal structure of the human -thrombin-thrombostatin
complex. Investigations were performed to determine whether RPPGF
directly interacted with -thrombin. A thrombin-hirugen complex was
prepared with human -thrombin (Enzyme Research Laboratories; South
Bend, IN). The thrombin-hirugen complex was made by adding 10-fold molar
excess hirugen to a solution of 1 mg/ml -thrombin in 750 mM NaCl and
0.1 M sodium phosphate buffer (pH 7.3). Independently, peptide NAT12 was added
to RPPGF in a 1:1 molar ratio to form a complex (25 mg/ml in water). Each
complex, thrombin-hirugen and NAT12-RPPGF, was left overnight at 4°C to
complete complexation. The thrombin-hirugen complex then was concentrated to 4
mg/ml. Ten-fold molar excess NAT12-RPPGF complex was then added to the
concentrated thrombin-hirugen complex. The resultant presumed quaternary
complex was left to stand for a day before the crystallization was set up.
Crystal Screen 1 (Hampton Research; Laguna Hills, CA) was used to scan the
initial conditions. Small crystals were obtained from factorials 20
and 31, both of which had PEG 4000 as the precipitant. Crystals
suitable for X-ray diffraction were grown in 23.5% PEG 4000 and 0.1 M sodium
acetate (pH 4.6) containing 0.2 M ammonium sulphate. A crystal of 0.2 x
0.2 x 0.15 mm was used for collecting the diffraction data, which
diffracted X-rays to a resolution of 2.1 Å.
Intensity data were measured with a R-AXIS II imaging plate detector
equipped with Molecular Structure focusing mirrors using CuK radiation from a
Rigaku RU200 rotating anode generator operating at 5 kW power with a fine
focus filament (0.3 x 3.0 mm). The crystal to detector distance was 102
mm, and each frame was collected for 15 min with an oscillation of 2.5°.
The total scan range for the data collection was 125°, and the total
exposure time was <15 h. The unit cell was determined by autoindexing
(7), and the processing of the
raw data was carried out with the Rigaku R-AXIS data-processing software
package. The crystals were orthorhombic: unit cell dimensions of a =
79.15 Å, b = 104.97 Å, and c = 45.18 Å,
belonging to the space group P21212. Of a total of
23,504 independent measurements, 17,158 measurements corresponded to
I/
(I) > 1.5, where I is the intensity of the reflections (73%,
Rmerge = 0.068). This data set is 78% complete for a
resolution of 2.5 Å and contains one-half the reflections for a
resolution between 2.3 and 2.1 Å.
The orientation of the thrombin molecule in the crystal was determined
using Patterson search techniques as implemented in the program AMoRe
(17). The thrombin coordinates
used for the molecular replacement calculations were of the CVS995-thrombin
complex (entry 1DIT
[PDB]
, Brookhaven Protein Data Bank) stripped of solvent and the
inhibitor molecule (10). The
cross-rotation search was conducted at two resolution ranges, 8.0–3.5
and 10.0–3.5 Å, which had a unique solution of 7 
above the
mean. A translational search performed in the space group
P21212 also gave an outstanding solution (correlation of
0.62, R = 37%). The starting model was refined using restrained
least-squares methods with the program PROFFT
(2). The modeling and electron
density fitting was carried out on a Silicon Graphics Indigo workstation using
the program CHAIN (21). After
15 cycles of overall B refinement (temperature factor or B factor) at a
resolution of 2.8 Å, the R factor converged to 29%. Individual
B refinement reduced the R factor further to 25%. Different electron
density maps were analyzed to find hirugen and the bound peptide. The
structure was then refined using standard techniques, including solvent, to a
final R factor of 20.2% with 105 solvent water molecules
(12). The average thermal
factors for the protein and water molecules were 24.6 and 25.7
Å2, respectively.
Ability of RPPGF to directly inhibit the enzymatic activity of
-thrombin. Investigations were performed to determine whether
the enzymatic activity of -thrombin could be directly inhibited by
RPPGF. Initial studies determined the Km of
-thrombin for the chromogenic substrates
H-D-Phe-Pip-Arg-paranitroanilide (S2238, DiaPharma; Franklin, OH)
and Sar-Pro-Arg-paranitroanilide (Sigma) in 0.01 M Tris and 0.15 M NaCl (pH
7.6). Additional studies were performed to determine whether 0.125–2 mM
RPPGF would interfere with the ability of -thrombin (1.25 nM) to
hydrolyze 0.7 mM S2238 or 0.6 mM Sar-Pro-Arg-paranitroanilide in 0.01 M Tris
and 0.15 M NaCl (pH 7.6). Further studies were performed to determine the
nature of the inhibition of -thrombin (1.25 nM) by 0.125–2 mM
RPPGF to hydrolyze Sar-Pro-Arg-paranitroanilide. In these experiments,
increasing concentrations of RPPGF (0–2 mM) were incubated with 1.25 nM
-thrombin in the presence of 0.04–0.5 mM
Sar-Pro-Arg-paranitroanilide at 37°C for 1 h. The rate of substrate
hydrolysis over 1 h was monitored at 37°C. Preliminary experiments
revealed that the hydrolysis of this substrate by -thrombin (1.25 nM)
under these conditions was linear for over 1 h. The data in the absence or
presence of RPPGF were plotted on a Lineweaver-Burk reciprocal plot of
1/V vs. 1/[S], where 1/V is the maximal velocity and [S] is
the concentration of the substrate. The Ki of RPPGF
inhibition of -thrombin was determined by two methods: a plot of the
negative x-intercept reciprocal (apparent Km) or
the slope of the reciprocal plot at each inhibitor concentration versus the
inhibitor concentration (23).
Both methods were used to verify the determination of the
Ki of the interaction.
Additional investigations determined whether the weak inhibitory activity
of RPPGF on the enzymatic activity of -thrombin could be modulated by
occupancy of exosite 1 by TM or sulfated N-acetyl
hirudin53–64
(26). In these experiments,
1.25 nM -thrombin was mixed with either 2 nM TM or 1 µM hirugen in
20 mM Tris acetate, 140 mM NaCl, 5 mM KCl, 1 mM MgCl2, and 1 mM
CaCl2 (pH 7.5) and incubated at room temperature for 6–15 min
before addition to the assay. The incubation mixtures were then added to
Sar-Pro-Arg-paranitroanilide (0.5 mM in 0.01 M Tris and 0.15 M NaCl; pH 7.6)
in the absence (control) or presence of increasing concentrations of RPPGF
(0–3 mM), followed by further incubation at 37°C for 1 h. The
hydrolysis of the substrate over 1 h was determined by monitoring the
absorbance at 405 nm. The change in the rate of substrate hydrolysis over 1 h
in the samples containing various concentrations of RPPGF was compared with
samples that did not contain the peptide.
Cloning and expression of the extracellular fragment of human
PAR1. A portion of the extracellular domain of human PAR1
(Ala26-Ser99) (rPAR1EC) was expressed in
Escherichia coli using a bacterial expression system (Novagen) and
its pET19b vector. Oligonucleotide primers for PCR were designed to place
NdeI and XhoI restriction sites at the 5' and 3'
ends of the coding sequence, respectively. PCR using human PAR1 cDNA as the
template prepared human PAR1 DNA encoding residues
Ala26-Ser99. The NdeI-XhoI PCR
fragment was ligated to the NdeI/XhoI sites of pET19b to
create pET19b/PAR1EC. This plasmid was then used to transform
NovaBlue, an E. coli K12 strain. The insert of the cloned DNA was
sequenced, and it showed 100% fidelity with the DNA sequence of the targeted
NH2-terminal extracellular domain of PAR1. The
His10-DDDDK-PAR1EC fusion construct was produced in
E. coli strain BL21(DE3) by transforming with
pET19b/PAR1EC and inducing it with 1 mM
isopropyl--D-thiogalactopyranoside (IPTG) for 2 h. The
expressed rPAR1EC (Ala26-Ser99) fusion
protein was purified from bacterial cytosolic fractions by nickel-chelate
affinity chromatography (HisTrap Affinity Column, Amersham Pharmacia Biotech;
Piscataway, NJ). The bound recombinant protein was eluted with 20 mM
phosphate, 0.5 M NaCl, and 500 mM imidazole (pH 7.4), dialyzed into 0.01 M
Tris and 0.15 M NaCl (pH 8), and stored in aliquots at -70°C. Recombinant
PAR1EC was characterized by 16.5% Tris-tricine SDS-PAGE,
NH2-terminal sequencing, immunoblotting with anti-PAR1 antibodies,
and cleavage experiments using thrombin or thrombocytin (see
RESULTS).
Site-directed mutagenesis. The pET19b/PAR1EC plasmid
was used to express several deletion mutants (mutants I–V and
VII) of rPAR1EC (Table
1). Oligonucleotide-directed mutagenesis using the GeneEditor
mutagenesis kit (Promega; Madison, WI) was performed to generate the mutants
of rPAR1EC in the expression vector pET19b/PAR1EC.
Briefly, mutants were selected based on the incorporation of a second site
mutation in -lactamase, which alters its substrate specificity, allowing
resistance in transformed bacteria to cefotaxime and ceftriaxone in addition
to ampicillin. Incorporation of the deletion was verified by DNA sequencing.
The mutagenesis primers for the introduction of site-directed deletion were as
follows: mutant I primer,
5'-GCAACAAATGCCACCCCCCGGTCATTTCTTC-3';
mutant II primer,
5'-AATGCCACCTTAGATTCATTTCTTCTCAGG-3';
mutant III primer,
5'-GCCACCTTAGATCCCTTTCTTCTCAGGAAC-3';
mutant IV primer, 5'-GCAACA
AATGCCACCTCATTTCTTCTCAGG-3'; mutant V primer,
5'-AATGCCACCTTAGATCTTCTCAGGA-ACCCC-3'; and
mutant VII primer,
5'-ACCCCCAATGATAAAGAGGATGAGGAGAAAAATG-3'.
Plasmid DNA was prepared using reagents supplied by Qiagen (Valencia, CA), and
recombinant site-directed deletion mutants of the extracellular domain of
human PAR1 were expressed in E. coli strain BL21(DE3) after induction
with 1 mM IPTG. The expressed mutant proteins were purified from bacterial
cytosol by nickel-chelate affinity chromatography as described above.
|
Examining the interaction of RPPGF with PAR1. Peptide RPPGC at 5–10 µg/well in 0.1 M Na2CO3 (pH 9.6) was bound to microtiter plate wells (F96 CERT.MAXISORP, No. 439454, Nunc-Immuno Plate, Fisher Scientific; Chicago, IL) by overnight incubation at 4°C or for 2 h at 37°C. In preliminary experiments, it was found that more biotin-NAT12 bound to cuvette wells linked with the peptide RPPGC than with RPPGF. Presumably, the COOH-terminal Cys on RPPGC allowed for more of it to adsorb to microtiter plates than RPPGF. After incubation, the contents of the wells were discarded, and the wells were washed with 0.01 M Na2PO4 and 0.15 M NaCl (pH 7.4) containing 0.01% Tween 20 (PBS-Tween). After each well was blocked with 1% radioimmuno-assay grade bovine serum albumin, the bound peptides were incubated with 30 µM biotin-NAT12 in the absence or presence of increasing concentrations of RPPGF, NAT12, or scrambled peptides for 1 h at 37°C. Detection of bound biotinylated peptide was performed by using Immuno-Pure streptavidin horseradish peroxidase conjugate from Pierce Chemical (Rockville, IL) and peroxidase-specific fast reacting substrate turbo-3,3',5,5'-tetramethylbenzidine (turbo-TMB, Pierce) as previously described (4). The color reaction was stopped by the addition of 1 M phosphoric acid, and the bound biotinylated peptide had its absorbance measured at 450 nm using a Microplate auto reader EL311 (Bio-Tek; Winooski, VT).
In other binding experiments, purified rPAR1EC (1 µg/well) was coupled to microtiter plates as described above. After each well was blocked, 1 mM RPPGF was incubated with the microtiter plate cuvette wells. Detection of the bound RPPGF was performed using a polyclonal rabbit antibody to RPPGF (Dainippon Pharmaceutical; Osaka, Japan), followed by the addition of a goat anti-rabbit antibody conjugated with horseradish peroxidase (Sigma). The antibody to RPPGF was specific to this peptide and did not recognize bradykinin, RPPGFSPFR (data not shown). After incubation, the enzyme conjugate on the cuvette well was measured by peroxidase-specific fast reacting substrate turbo-TMB (Pierce), as described above (4).
Additional experiments determined whether soluble rPAR1EC or the six deletion mutants blocked biotin-RPPGF binding to rPAR1EC coupled to cuvette wells. In these experiments, rPAR1EC at 1 µg/well was linked to microtiter plate cuvette wells in 0.1 M Na2CO3 (pH 9.6) by overnight incubation at 4°C. After it was linked, 10 µM biotin-RPPGF was incubated with the cuvette wells in the absence or presence of increasing concentrations of rPAR1EC, its deletion mutants, or 5 mM L-Arg (Sigma). The amount of biotin-RPPGF bound was determined as described above (4).
Digestion of rPAR1EC by proteolytic
enzymes in the absence or presence of RPPGF or MAP4-RPPGF.
rPAR1EC (1 µg) in 0.01 mM Tris · HCl and 0.2 M NaCl (pH
8.0) was incubated at 37°C for 15 or 45 min with 1 nM -thrombin or
0.5 µg/ml (16.7 nM) thrombocytin, respectively, in the absence or presence
of 0.125–1 mM RPPGF or 0.05–0.5 mM MAP4-RPPGF. The reaction
mixture was separated by SDS-PAGE and stained with Coomassie Brilliant blue.
The cleaved gel images were scanned using Scion Image software (ScionCorp;
Frederick, MD).
Ability of RPPGF to directly inhibit the enzymatic activity of thrombocytin. Studies determined whether 0.5–2 mM of RPPGF or 0.2–1 mM MAP4-RPPGF interfered with the ability of 0.5 µg/ml thrombocytin (16.7 nM) to hydrolyze 1.0 mM H-D-Phe-Pip-Arg-paranitroanilide in 0.01 M Tris and 0.15 M NaCl (pH 7.6) buffer. The initial rate of hydrolysis of the substrate was determined by monitoring absorbance at 405 nm over 30 min.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-thrombin.
Investigations were performed to determine the mechanism(s) by which RPPGF
inhibited thrombin-induced platelet aggregation. Initial studies determined
whether RPPGF inhibited the coagulant activity of thrombin
(Fig. 1). As the concentration
of RPPGF increased from 0.031 to 3 mM, there was a significant prolongation of
the TCT at ≥0.125 mM (P < 0.01), the APTT at ≥0.25 mM
(P < 0.016), and the PT at ≥0.5 mM (P < 0.003). At
the 0.5 mM concentration of RPPGF, the TCT or APTT was more significantly
prolonged (P < 0.0005 and P < 0.001 levels,
respectively) than the PT. The increased sensitivity of inhibition of the APTT
over the PT may have arisen from the fact that RPPGF inhibits factor XI
activation (6). Because RPPGF
had a direct interaction with thrombin, studies were performed to determine
whether L-Arg itself influenced these assays. L-Arg at a
final concentration of 3 mM did not prolong any of the coagulant assays. These
data indicate that the effect of RPPGF on the coagulant assays was specific
for this peptide and not its NH2-terminal Arg.
|
Investigations to determine whether there is evidence for a physical
interaction between RPPGF and thrombin. Studies were then performed to
determine whether RPPGF interacts with preformed -thrombin crystals. A
preformed NAT12-RPPGF complex was incubated with an -thrombin-hirugen
crystal. Examination of its electron density map revealed the presence of
residues RPPG of RPPGF bound in the active site, in a retro manner, i.e., the
NH2-terminal arginine is in the active site
(Fig. 2), with hirugen bound at
exosite I (Fig. 3)
(21,
22). However, even though a
solution of RPPGF and NAT12 was used for cocrystallization with
hirugen-thrombin, electron density maps only revealed RPPG interacting with
the thrombin active site and did not reveal any bound NAT12 or parts thereof
in the crystal structure (Figs.
2,
3,
4).
|
|
|
The interaction of RPPGF with the active site of -thrombin is shown
in Fig. 4. RPPGF was a
retrobinder to the active site of -thrombin because its NH2
terminal fitted into the specificity (S1) pocket and the main chain ran
parallel (from the NH2 to COOH terminal) with respect to
Ser214-Gly216
(13,
19). In normal substrate
binding, the main chain of the peptide runs from the COOH to NH2
terminal. The location of the Arg of RPPG, being inserted in the S1 pocket,
was clearly evident in fitting the peptide to the electron density map
(Fig. 2). With the
NH2-terminal Arg in the S1 pocket, the Pro residues were
unambiguously assigned positions, one occupying the S2 site (Figs.
2 and
4). This Pro interacted with
His57, Trp60D, Trp215, and Gly216
(Table 1). The two adjacent Pro
limit the conformation of RPPG, preventing it from forming good hydrogen
bonding interactions of a parallel -strand.
The NH2-terminal arginyl is located in the S1 pocket of the active site (Fig. 4). It interacted with His57, Asp189, Ser195, and Ser214 (Table 1). The binding orientation of RPPG is unusual because the main chain direction is opposite to the main chain direction of thrombin substrates. There are generally two hydrogen bonds from substrate/inhibitor to the Gly216 backbone oxygen and nitrogen atoms. In the present structure, however, only one longer hydrogen bond is observed. The presence of Arg in the specificity pocket allowed for all the expected interactions with the surrounding molecules, including the water-mediated interactions of the NE atom with Gly216 (Table 1). The arginyl side chain atoms have lower thermal factors compared with the remainder of RPPGF, suggesting better binding in the specificity S1 pocket. The final coordinates of the ternary complex have been deposited in the Protein Data Bank (code 1NY2 [PDB] ).
Ability of RPPGF to directly inhibit the enzymatic activity of
-thrombin. Because RPPGF physically interacted with the active
site of -thrombin, investigations determined whether RPPGF interfered
with the enzymatic activity of -thrombin. Previous studies using the
chromogenic substrate H-D-Phe-Pip-Arg-paranitroanilide showed that
RPPGF did not interfere with the ability of -thrombin to hydrolyze that
substrate (11). The
Km for -thrombin to hydrolyze
H-D-Phe-Pip-Arg-paranitroanilide was 6 µM in 0.01 M Tris and
0.15 M NaCl at pH 8.3 and 37 µM in the same buffer at pH 7.6
(26). Because RPPGF at only
332–680 µM inhibited -thrombin's ability to activate human
platelets, -thrombin had a higher affinity to the chromogenic substrate
H-D-Phe-Pip-Arg-paranitroanilide substrate than to RPPGF
(3,
5,
6). However, using an
-thrombin substrate, Sar-Pro-Arg-paranitroanilide, which has a lower
Km (138 µM), progressive inhibition of
-thrombin's ability to hydrolyze the substrate was seen when the RPPGF
concentration was >1 mM (data not shown). At 2 mM RPPGF, there was 50%
inhibition of -thrombin-induced hydrolysis of
Sar-Pro-Arg-paranitroanilide. Alternatively, RPPGF was unable to inhibit
-thrombin's ability to hydrolyze
H-D-Phe-Pip-Arg-paranitroanilide at concentrations up to 3 mM, as
previously reported (data not shown)
(3). These data suggest that
RPPGF interacted weakly with the active site of -thrombin to inhibit
thrombin's amidolytic activity.
Investigations were then performed to determine the kinetics of RPPGF
inhibition of -thrombin (Fig.
5). With the use of Sar-Pro-Arg-paranitroanilide concentrations
around the Km of -thrombin hydrolysis, there was a
progressive increase in the apparent Km as the
concentration of RPPGF increased in the reaction mixture
(23). On reciprocal plot
1/V versus 1/[S], the maximal velocity of each reaction,
1/V, was approximately the same in the absence or presence of
increasing concentrations of RPPGF (0.125–2.0 mM). Similarly, on the
reciprocal plot, there was a progressive increase in the
Km (Fig.
5). The results of this experiment indicated competitive
inhibition, consistent with RPPGF interacting with the active site of
-thrombin (23). The
Ki of RPPGF inhibiting -thrombin from hydrolyzing
this substrate was 1.75 ± 0.03 mM. Additional investigations were
performed to determine whether occupancy of thrombin's exosite I altered
RPPGF's ability to inhibit thrombin. Thrombomodulin or sulfated
N-acetyl hirugen53–64 did not alter the ability of
RPPGF to inhibit thrombin (data not shown). These latter data indicated that
the interaction of RPPGF with the active site of the thrombin-hirugen complex
is representative of the interaction of RPPGF with free thrombin and is not
modified by proteins or peptides occupying thrombin's exosite I.
|
Interaction between RPPGF and a peptide that includes the thrombin cleavage site on the extracellular fragment of PAR1. The above studies indicated that RPPGF was a weak, active site binding inhibitor of thrombin. This interpretation of the mechanism by which RPPGF inhibits thrombin activation of platelets, however, did not completely correlate with some previous in vitro and in vivo studies (5). FITC-labeled RPPGF binds to platelets, and thrombin inhibition by RPPGF of canine platelets in vivo was present 150 min after no RPPGF (i.e., <1 µM) could be measured in plasma. These latter data suggested that RPPGF may prevent thrombin activation of platelets by another mechanism than just active site thrombin inhibition as described above. Investigations were performed to determine whether there is a physical interaction between RPPGF and thrombin's major receptor on platelets, PAR1.
Initial studies determined whether there were any interaction between RPPGF and a peptide (NAT12) that included the thrombin cleavage site on the extracellular domain of human PAR1. Preliminary investigations showed that biotin-NAT12 did not bind nonspecifically to microtiter plate wells, albumin, or irrelevant peptides (data not shown). When peptide RPPGC, a peptide that binds plastic more avidly, was coupled to microtiter plates, RPPGF blocked biotin-NAT12 peptide binding with an IC50 of 20 µM (Fig. 6). Unlabeled peptide NAT12 also blocked biotin-NAT12 binding with an IC50 of 500 µM. Alternatively, two scrambled peptides of RPPGF, FPRPG or GPRP, did not block biotin-NAT12 binding (Fig. 6). These experiments suggest that there might be a physical interaction between RPPGF and the thrombin cleavage site on the extracellular domain of PAR1.
|
Preparation and characterization of the extracellular domain of human PAR1. To further examine whether there is an interaction between RPPGF and the extracellular domain of PAR1, a recombinant form of a portion of the extracellular domain of PAR1 (rPAR1EC) was prepared. With the use of pET19b/PAR1EC, a fusion protein was produced at 11 kDa, as seen on Coomassie-stained Tris-tricine SDS-PAGE gels, respectively (see Fig. 10, rPAR1 lane). On amino acid sequencing, the rPAR1EC protein produced by pET19b had complete amino acid identity with native human PAR1 from amino acids 26–99 (data not shown). The recombinant extracellular domain of human PAR1 was recognized by a goat polyclonal antibody to PAR1 or monoclonal antibody SPAN12, both raised to a peptide (NATLDPRSFLLR) from the thrombin cleavage region on human PAR1, and by a monoclonal antibody (WEDE15) to the hirudin-like binding domain on human PAR1 (Fig. 7) (27). The wild-type rPAR1EC gave a doublet on immunoblot with anti-PAR1, SPAN12, and WEDE15 antibodies (Fig. 7, rPAR1EC lane). Also, the rPAR1EC fragment was recognized by a monoclonal antibody to the His tag present in the NH2 terminus of the recombinant fusion protein (data not shown). The doublets of rPAR1EC seen on immunoblot in Fig. 7 represent some COOH-terminal proteolysis because the NH2 terminal His tag was present on all wild-type and mutant forms of the recombinant protein.
|
|
When rPAR1EC was coupled to a microtiter plate, only goat anti-human PAR1 antibody raised to peptide NATLDPRSFLLR recognized the recombinant protein (data not shown). Polyclonal antisera to the analogous regions of human PAR3 or PAR4 did not recognize rPAR1EC linked to the microtiter plate cuvette wells (data not shown). Further specificity studies showed that when rPAR1EC was coupled to a microtiter plate, peptide RPPGF specifically bound to rPAR1EC, as detected by a specific anti-RPPGF antibody followed by a secondary antibody (Fig. 8). These experiments indicate that immunochemically recognizable rPAR1EC was prepared and that it bound RPPGF. This interaction between RPPGF and rPAR1EC was specific for the entire peptide. L-Arg (5 mM), unlike RPPGF, did not block 50 µM biotin-RPPGF binding to rPAR1EC bound to microtiter plates (data not shown).
|
Investigations were then performed to determine the specific region on the rPAR1EC that bound RPPGF. Six recombinant deletion mutant forms of rPAR1EC were prepared by site-directed mutagenesis (Table 2). Five of these deletion mutants (mutants I–V) were missing two or four amino acids around the thrombin cleavage site on the rPAR1EC. A sixth deletion mutant (mutant VII) was missing five amino acids at the hirudin-like binding region on PAR1 (Table 2). On immunoblot, polyclonal antibody to the peptide NATLDPRSFLLR was able to detect each of the mutagenized rPAR1EC, although the wild-type protein and the deletion mutant of the hirudin-like binding region (mutant VII) were recognized better (Fig. 7). The monoclonal antibody SPAN12 only detected wild-type, mutant V, and mutant VII rPAR1EC. Monoclonal antibody WEDE15 recognized all rPAR1EC except mutant VII, which had the deletion of amino acids YEPFW in the hirudin-like binding region (Table 2 and Fig. 7).
|
In competition-inhibition binding experiments, when rPAR1EC was coupled to a microtiter plate, soluble wild-type rPAR1EC inhibited biotin-RPPGF binding to linked rPAR1EC with an IC50 of 20 µM (Fig. 9). Deletion mutant VII of rPAR1EC of the hirudin-like binding site inhibited biotin-RPPGF binding to wild-type rPAR1EC with an IC50 = 65 µM (Fig. 9). Deletion mutants I, II,or III, which have amino acids LD, PR, or RS removed, respectively, also inhibited biotin-RPPGF binding to rPAR1EC (data not shown). However, deletion mutant IV, which had amino acids LDPR eliminated, or deletion mutant V, which had amino acids PRSF eliminated, did not block biotin-RPPGF binding to linked wild-type rPAR1EC (Fig. 9). These data indicate that RPPGF binding to rPAR1EC required the amino acids LDPR or PRSF to be present. Deletion mutants I–III, missing amino acids LD, PR, or RS, had a sufficient amount of structure present to compete with biotin-RPPGF binding to wild-type rPAR1EC like the natural sequence.
|
Ability of RPPGF and MAP4-RPPGF to inhibit thrombin's and
thrombocytin's proteolysis of rPAR1EC. Investigations
were then performed to determine whether RPPGF and MAP4-RPPGF, a MAP form of
RPPGF, blocked thrombin and thrombocytin proteolysis of PAR1. Thrombocytin is
a thrombin-like snake venom enzyme that activates platelets by proteolysis of
PAR1 (19,
22). Initial experiments
determined whether RPPGF or MAP4-RPPGF prevented -thrombin or
thrombocytin proteolysis of wild-type rPAR1EC
(Fig. 10). As the
concentration of RPPGF increased from 0.125 to 1.0 mM, there was reduced
thrombin proteolysis of rPAR1EC
(Fig. 10A,
top). In the presence of 1 mM RPPGF, 24% of rPAR1EC
remained uncleaved by thrombin (Fig.
10A, top). When the thrombin-like enzyme
thrombocytin was used, 0.5 and 1.0 mM RPPGF allowed for only 44% and 40% of
rPAR1EC to be proteolyzed, respectively. More than 55% of
rPAR1EC remained uncleaved
(Fig. 10A,
bottom). Additional studies were performed with a cluster peptide of
RPPGF-MAP4-RPPGF. As the concentration of MAP4-RPPGF increased from 0.05 to
0.5 mM, there was reduced thrombin proteolysis of rPAR1EC
(Fig. 10A,
top). MAP4-RPPGF (0.2 and 0.5 mM) prevented thrombin from
proteolyzing 78% and 82% of rPAR1EC, respectively
(Fig. 10B,
top). Similarly, 0.5 mM MAP4-RPPGF allowed for only 24% cleavage of
rPAR1EC by thrombocytin (Fig.
10B, bottom).
The concentrations of RPPGF and MAP4-RPPGF that prevented thrombin cleavage of rPAR1EC in Fig. 10, A and B, overlapped with those that prolonged the TCT (Fig. 1) (6). The prevention of cleavage data in Fig. 10 did not obviously confirm that these peptides blocked thrombin cleavage of rPAR1EC by binding to it. However, 2.0 mM RPPGF and 1.0 mM MAP4-RPPGF, concentrations twice that needed to prevent >50% cleavage of rPAR1EC, did not directly inhibit 16.7 nM thrombocytin hydrolysis of the substrate H-D-Phe-Pip-Arg-paranitroanilide (Fig. 10C). These latter data indicate that the concentrations of RPPGF and MAP4-RPPGF used in the thrombocytin cleavage experiments (Fig. 10, A and B, bottom) could only have prevented the proteolysis of rPAR1EC by binding to the recombinant protein. These latter data, along with the experiments using deletion mutants of rPAR1EC, indicated that RPPGF and related compounds also inhibit thrombin activation of platelets by binding to and preventing thrombin cleavage of PAR1 as a result of the bound peptide.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-thrombin to
cleave PAR1 by both binding to the extracellular domain of human PAR1 near the
thrombin cleavage site and by binding to the active site of -thrombin
itself. The affinity of RPPGF to bind to the extracellular domain of PAR1 is
higher than its affinity to directly interact with the active site of
thrombin. RPPGF blocks biotin-NATLDPRSFLLR binding to linked RPPGC with an
IC50 of 20 µM; rPAR1EC blocks biotin-RPPGF binding to
linked rPAR1EC with an IC50 of 50 µM. Alternatively,
RPPGF blocks -thrombin's hydrolysis of Sar-Pro-Arg-pNA with a
Ki of 1.75 mM. These findings explain the in vivo
observation that RPPGF-induced platelet inhibition occurs at concentrations
below that which inhibit the active site of thrombin
(5,
6). Furthermore, it explains
the observation that after infusions of RPPGF in dogs, there is a prolonged
antiplatelet effect beyond the prolongation of the thrombin time and
measurable plasma levels of the peptide
(5,
6). Because RPPGF and
MAP4-RPPGF do not inhibit the enzymatic activity of thrombocytin at the
concentrations that inhibit thrombocytin proteolysis of PAR1, their major
antiplatelet effect must be due to their ability to bind the extracellular
domain of PAR1 to prevent proteolysis. These data suggest that inhibition of
thrombin's active site and prevention of thrombin's ability to cleave PAR1,
due to occupancy by RPPGF, could be occurring at the same time. This
assessment is supported by the inability of two deletion mutants of
rPAR1EC to inhibit RPPGF binding to wild-type recombinant protein.
Finally, it is possible that RPPGF has a better ability to bind PAR1 in cell
membranes than in solution to prevent thrombin cleavage. This finding might be
why RPPGF appears to be a better platelet inhibitor in vivo
(5,
6).
The finding that the NH2-terminal portion of RPPGF (i.e.,
retrobinding) interacts directly with the active site of -thrombin was
not expected. Crystal structures of thrombin complexed with retrobinding
inhibitors are not very common. Only four such structures have been
recognized: thrombin-BMS183507
(25), thrombin-nazumamide A
(18), and thrombin-Sel 2711 or
2770 (13). These inhibitors
have modified nonpeptide residues and are therefore different from RPPGF. Only
the BMS183507 inhibitor forms a three hydrogen-bonded parallel -sheet
with Ser214-Gly216, whereas Sel 2711 and 2770 have two
of the hydrogen bonds, and nazumamide A has none. Stabilization of the
thrombin-nazumamide A complex appears to primarily come from interactions in
the S1-S3 binding subsites of thrombin. The parallel -strand
interactions of RPPG are intermediate: two -strands involve a poor
hydrogen bonding angle (Table
1), whereas Pro3 N does not have a hydrogen to make a
hydrogen bond. The side chain interactions of BMS183507 and nazumamide A in
the S1 pocket also are different than RPPGF and the selectide inhibitors. The
former two have only one hydrogen bond in the salt bridge with
Asp189 compared with doubly hydrogen-bonded salt bridges for others
(Fig. 4). Moreover, only RPPGF
has specific interactions with the catalytic triad of thrombin. It is the only
retrobinder to the active site of -thrombin with all naturally
occurring amino acid residues. The interactions between thrombin and the RPPG
peptide are largely through Arg1-Pro2, being
overwhelmingly dominated by Arg1
(Table 1). This fact is
probably the reason for the relatively weak thrombin binding constant
(Ki = 1.75 mM) of RPPGF to the active site of
thrombin.
The experiments to determine whether there is a direct interaction between
RPPGF and -thrombin were begun by the observation that high
concentrations of RPPGF influenced thrombin-mediated coagulant activity. The
APTT, PT, and TCT are prolonged by high concentrations of RPPGF. The
crystallization and kinetic experiments explain this observation. However,
inhibition of the APTT and PT by RPPGF does not necessarily mean that these
peptides are directly interfering only with thrombin. Recent studies indicate
that RPPGF and MAP4-RPPGF interfere with thrombin activation of factor XI
(6). In our animal
investigations, the plasma concentration of RPPGF that inhibited coronary
thrombosis was ≤8.7 µM, a value well below the minimal concentration of
RPPGF that directly interacts with the active site of -thrombin
(5,
6). The present data indicate
that RPPGF binds to the extracellular domain of PAR1 at amino acids LDPR or
PRSF to prevent thrombin cleavage. This activity is probably the main
mechanism by which RPPGF when infused into dogs interferes with thrombin
activation of platelets in vivo.
There is little evidence to date that the interaction of RPPGF with the
active site of thrombin is physiological. However, the native sequence of
RPPGF may provide for a model compound to direct development efforts to
prepare substrate-selective inhibitors of -thrombin. RPPGF can be
described as a substrate-directed thrombin inhibitor because it is partially
directed to the thrombin substrate PAR1. Having compounds that separately
inhibit the ability of -thrombin to activate platelets without
interacting with the active site of thrombin would be novel agents to examine
for their clinical effect on arterial thrombosis. Selective thrombin
inhibitors that interact with PAR1 to prevent thrombin cleavage may also be
useful in regulating -thrombin's mitogenic and inflammatory activities
and the role of PAR1 in cancer cell metastasis.
|
ACKNOWLEDGMENTS |
|---|
|
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.
Present address of A. A. K. Hasan: National Institutes of Health, National Heart, Lung, and Blood Institute, 6701 Rockledge Dr., MSC 7950, Bethesda, MD 20892-7950.
Present address of R. Krishnan: BioCryst Pharmaceuticals Incorporated, Birmingham, AL 35244.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-thrombin.
Protein Sci 1:
426-471, 1992.[Web of Science][Medline]
-thrombin-induced
platelet activation. Circulation
94: 517-528,
1996.-thrombinb complexed with fibrinopeptide A(7–16): evidence for
residual catalytic activity. Acta Crystallogr D Crystallogr
Biol 56: 406-410,
2000.[Medline]
-thrombin. J Mol Biol
221: 1379-1393,
1991.[Web of Science][Medline]
-thrombin. J Mol Biol
246: 14-20,
1996.
This article has been cited by other articles:
|
|
 |
|
  A. M. Marney, J. Ma, J. M. Luther, T. A. Ikizler, and N. J. Brown Endogenous Bradykinin Contributes to Increased Plasminogen Activator Inhibitor 1 Antigen following Hemodialysis J. Am. Soc. Nephrol., October 1, 2009; 20(10): 2246 - 2252. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Shariat-Madar, F. Mahdi, M. Warnock, J. W. Homeister, S. Srikanth, Y. Krijanovski, L. J. Murphey, A. A. Jaffa, and A. H. Schmaier Bradykinin B2 receptor knockout mice are protected from thrombosis by increased nitric oxide and prostacyclin Blood, July 1, 2006; 108(1): 192 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Zamorano, S. Suchindran, and J. V. Gainer 3'-Untranslated region of the type 2 bradykinin receptor is a potent regulator of gene expression Am J Physiol Renal Physiol, February 1, 2006; 290(2): F456 - F464. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Shariat-Madar, F. Mahdi, and A. H. Schmaier Recombinant prolylcarboxypeptidase activates plasma prekallikrein Blood, June 15, 2004; 103(12): 4554 - 4561. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
