Báo cáo khoa học: Identification and characterization of novel salivary thrombin inhibitors from the ixodidae tick, Haemaphysalis longicornis doc

Identification and characterization of novel salivary thrombin
inhibitors from the ixodidae tick,
Haemaphysalis longicornis
Shiroh Iwanaga
1
, Masakazu Okada
1
, Haruhiko Isawa
3
, Akihiro Morita
2
, Masao Yuda
2
and Yasuo Chinzei
2
1
Laboratory of Chemistry and Utilization of Animal Resources, Faculty of Agriculture, Kobe University, Japan;
2
Department of Medical Zoology, School of Medicine, Mie University, Tsu, Japan
3
Laboratory of Physiology and Biochemistry, Department of Medical Entomology,NationalInstituteofInfectious Diseases, Tokyo, Japan
Novel antithrombin molecules were identified from the
ixodidae tick, Haemaphysalis longicornis. These molecules,
named madanin 1 and 2, are 7-kDa proteins and show no
significant similarities to any previously identified proteins.
Assays using human plasma showed that madanin 1 and 2
dose-dependently prolonged both activated partial throm-
boplastin time and prothrombin time, indicating that they
inhibit both the intrinsic and extrinsic pathways. Direct
binding assay by surface plasmon resonance measurement
demonstrated that madanin 1 and 2 specifically interacted

with thrombin. Furthermore, it was clearly shown that
madanin 1 and 2 inhibited conversion of fibrinogen into
fibrin by thrombin, thrombin-catalyzed activation of
factor V and factor VIII, and thrombin-induced aggregation
of platelets without affecting thrombin amidolytic activity.
These results suggest that madanin 1 and 2 bind to the anion-
binding exosite 1 on the thrombin molecule, but not to the
active cleft, and interfere with the association of fibrinogen,
factor V, factor VIII and thrombin receptor on platelets with
an anion-binding exosite 1. They appear to be exosite 1-
directed competitive inhibitors.
Keywords: anticoagulant; Haemaphysalis longicornis; sali-
vary gland; thrombin inhibitor; tick.
Thrombin has various physiological functions and plays
important roles in hemostasis. For example, in the final step
of blood clot formation, thrombin converts soluble fibri-
nogen into fibrin and subsequently triggers cross-linking
between fibrin monomers by activating factor XIII [1]. It
also amplifies its own generation by activating nonenzy-
matic cofactors V and VIII as well as factor XI [2,3].
Conversely, it suppresses its own generation by activating
protein C [4], which inactivates factor Va and factor VIIIa
together with protein S [5], when bound to the endothelial
membrane receptor thrombomodulin. In addition, throm-
bin induces platelet aggregation via proteolytic activation of
G-protein-coupled protease-activated receptors (PARs)
[6,7]. Specific interactions of thrombin with these substrates,
cofactors, and receptors involve not only the catalytic site
and the primary binding pocket, but also secondary
recognition sites, termed anion-binding exosite 1 and 2.

Anion-binding exosite 1 interacts with negatively charged
domains on fibrinogen [8], PARs [6,7,9], and thrombomo-
dulin [10,11]. Anion-binding exosite 2 interacts with heparin
[12], promoting inhibition of thrombin by antithrombin III
[13] and heparin cofactor II [14]. Furthermore, both exosites
areinvolvedintherecognitionoffactorVandfactorVIII
by thrombin [15].
The salivary glands of blood-sucking animals, such as
leeches, insects, and ticks, contain various anticoagulants
[16]. These substances inhibit the host hemostatic response so
that the blood-sucking organism can feed smoothly on host
blood. The best known anticoagulant identified from blood-
sucking organisms is hirudin, a highly specific thrombin
inhibitor, isolated from the medical leech, Hirudo medicinalis
[17]. It interacts with two distinct sites on the thrombin
molecule: its N-terminal and C-terminal domains bind to the
active site and anion-binding exosite 1, respectively [18,19].
This binary binding mechanism appears to contribute to its
potent inhibitory activity. It has also been demonstrated that
the peptide alone derived from the C-terminal domain of
hirudin is able to inhibit various thrombin functions [20–23].
This indicates that anion-binding exosite 1 is essential in
interactions between thrombin and its substrates, and that
competitive binding to anion-binding exosite 1 is one strategy
of thrombin inhibition.
In this paper, we describe two novel anticoagulants
identified from the ixodidae tick, H. longicornis.These
molecules exhibit no sequence similarities to any previously
known proteins. We show that the recombinant anticoagu-
lant molecules clearly prolong both activated partial

thromboplastin time (APTT) and prothrombin time (PT),
and specifically bind to thrombin. We further demonstrate
that these molecules inhibit the conversion of fibrinogen
into fibrin, activation of factor V and factor VIII, and
aggregation of platelets by thrombin without inhibiting
thrombin amidolytic activity toward a small synthetic
substrate. These results suggest that these factors are novel
exosite 1 competitive inhibitors like the C-terminal peptide
of hirudin.
Correspondence to S. Iwanaga, Laboratory of Chemistry and
Utilization of Animal Resources, Faculty of Agriculture,
Kobe University, Kobe 657-8501, Japan.
E-mail:
Abbreviations: APTT, activated partial thromboplastin time; PT,
prothrombin time; PAR, protease activated receptor; SPR, surface
plasmon resonance; RU, resonance unit.
Enzyme: Thrombin (EC 3.4.21.5).
(Received 13 November 2002, revised 1 March 2003,
accepted 7 March 2003)
Eur. J. Biochem. 270, 1926–1934 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03560.x
Experimental procedures
Materials
Human thrombin, bovine thrombin, human factor X/Xa,
human factor IXa, and human factor V were purchased from
Enzyme Research Laboratories. The following absorption
coefficients (e
0.1%,280
) and molecular masses were used to
determine protein concentrations: human thrombin, 18.3,
37 kDa; bovine thrombin, 19.5, 37 kDa; human factor X,

11.6, 58.8 kDa; human factor Xa, 11.6, 46 kDa; human
factor IXa, 14.9, 56 kDa; human factor V, 9.6, 330 kDa.
Human factor VIII was obtained from American Diagnos-
tica Inc., Greenwich, CT, USA and the concentration
adjusted to 0.25 UÆlL
)1
. Human fibrinogen was from
Sigma-Aldrich. Chromogenic substrates, S-2238 and
S-2222, were obtained from AB Kabi. Restriction enzyme
was purchased from Invitrogen. All other reagents were
analytical grade and obtained from either Nacalai Tesque,
Kyoto, Japan or Wako Pure Chemical Industry, Osaka,
Japan.
Mass sequence analysis of cDNA clones
H. longicornis salivary gland mRNA was isolated from 10
pairs of glands of ticks at three distinct feeding stages
(unfed, slow feeding, and rapid feeding) by using a
QuickPrep Micro mRNA Purification Kit (Amersham
Bioscience). Three cDNA libraries were constructed from
each isolated mRNA using SuperScript Plasmid System for
cDNA Synthesis and Plasmid Cloning Kit (Invitrogen). In
total, approximately 2000 cDNA clones were picked
randomly from three libraries, and their partial DNA
sequences were determined using T7 primer and an ABI
PRISM Big Dye Terminator Cycle Sequencing Kit (Applied
Biosystems). Sequence similarity searches of clones were
carried out using the
BLAST
program at the National Center
for Biotechnology Information (NCBI). In addition, signal

peptides of deduced amino-acid sequences were predicted
using the
SIGNAL P
Program at the Center for Biological
Sequence Analysis (CBS).
Expression and purification of recombinant proteins
DNA fragments encoding predicted mature regions of
recombinant proteins were amplified by PCR using each
specific primer set. Amplified PCR products were subcloned
into the NdeI–HindIII site of pET22b expression vector
(Novergen). After verification of nucleotide sequences of
constructed plasmids, recombinant proteins were expressed
according to the manufacturer’s instructions. Cells expres-
sing recombinant proteins were resuspended in 20 mL
50 m
M
Tris/HCl, pH 7.5, and frozen at )20 °C. Cells were
thawed in ice-cold water and sonicated. Cell lysates were
centrifuged, and supernatants subjected to gel-filtration
chromatography on a Sephadex G75 column (1.8 · 90 cm).
Fractions containing purified recombinant proteins were
pooled and stored at )20 °C until use.
The purity of the recombinant proteins was confirmed by
RP-HPLC using a Wakosil 5C4 column (4.6 mm · 20 cm;
Wako) pre-equlibrated in 0.1% trifluoroacetic acid. Bound
proteins were eluted with a linear gradient of 0–100%
acetonitrile/0.1% trifluoroacetic acid. The molecular masses
of purified recombinant madanin 1 and 2 were determined
on a Voyager MALDI-TOF mass spectrometer (PerSeptive
Biosystems) with a 337-nm N

2
laser and ion reflector.
Assay of effects of madanin 1 and 2 on plasma
coagulation
Citrated human normal plasma (20 lL) and recombinant
proteins (30 lL) were preincubated for 5 min at 37 °C.
Mixtures were activated for 2 min at 37 °Cwith30lL25%
actin (Dade Behring) in the APTT assay and with 30 lL
rabbit brain thromboplastin (Ortho Diagnostic System) in
the PT assay. Clotting reactions were started by the addition
of 25 lL50m
M
CaCl
2
, and the clotting time was measured
using a coagulometer.
Binding analysis using surface plasmon resonance (SPR)
SPR measurement was performed using a BIAcore 3000
instrument (BIAcore). Thrombin (bovine) was immobilized
on the surface of a sensor chip CM5 by the amine coupling
procedure according to the manufacturer’s instructions.
Binding analyses were carried out using Hepes-buffered
saline (10 m
M
Hepes,pH7.4,150m
M
NaCl, 5 m
M
CaCl
2

,
and 0.005% Tween 20) as running buffer at 25 °C. A 40 lL
volume of various concentrations of the samples was
injected on to the sensor chip at a flow rate of 20 lLÆmin
)1
.
Association was monitored during a 2-min injection of
analyte. Dissociation was monitored for 2 min after return
to the running buffer. Regeneration of the sensor chip
surface was achieved with a pulse injection of 1
M
NaCl.
The binding data were analyzed using the evaluation
software (BIAevaluation) to determine the dissociation
constants (K
d
).
Assay of effects of madanin 1 and 2 on fibrinogenolytic
activity of thrombin
Inhibition of fibrin clot formation by the recombinant
proteins was measured using fibrinogen as a substrate.
Substrate solution was prepared by the addition of 1 part
arabic gum (15%, w/v) to 7 parts fibrinogen (100 mg per
7 mL), and preincubated at 37 °C for 15 min. Thrombin
(3.9 n
M
, final concentration) was mixed with various
concentrations of recombinant proteins, and the mixtures
added to prewarmed substrate solution. The prolongation of
fibrin clot formation was measured using the coagulometer.

Assay of effects of madanin 1 and 2 on activation of
factor V and factor VIII by thrombin
The effect of the madanins on the activation of factor V by
thrombin was determined as follows. Factor V (240 p
M
,
final concentration) in buffer A (50 m
M
Tris/HCl, pH 7.5,
150 m
M
NaCl, 5 m
M
CaCl
2
, and 0.1% BSA) was preincu-
bated for 2 min at 37 °C with thrombin (20 p
M
,final
concentration) in the presence of various concentrations of
the recombinant proteins and added to buffer A containing
400 n
M
prothrombin, 20 p
M
factor Xa, and 40 l
M
phos-
pholipid. After the addition of thrombin and the recom-
binant proteins, the reaction mixtures were incubated for

Ó FEBS 2003 Identification of tick thrombin inhibitors, madanin 1 and 2 (Eur. J. Biochem. 270) 1927
5 min at room temperature. Prothrombin activation by
prothrombinase was stopped by the addition of EDTA
(5 m
M
, final concentration). The activity of generated
thrombin, which reflects the amount of activated factor V
in the sample, was measured using chromogenic substrate
S-2238.
The effect of the madanins on the activation of factor
VIIIbythrombinwasdeterminedasfollows.FactorVIII
(0.15 UÆmL
)1
, final concentration) in buffer A was pre-
incubated for 5 min at 37 °C with thrombin (2.5 p
M
, final
concentration) and recombinant proteins, then added to
buffer A containing 400 n
M
factor X, 1 n
M
factor IXa, and
40 l
M
phospholipid. The reaction mixtures were incubated
for 2 min at room temperature. After the incubation, the
activity of factor Xa, which reflects the amount of activated
factor VIII, was measured using chromogenic substrate
S-2222.

Assay of effects of madanin 1 and 2 on thrombin-
induced platelet aggregation
Washed platelets were prepared as follows. Blood was
mixed with acid citrate dextrose, incubated for 30 min at
room temperature, and centrifuged at 300 g for 10 min.
Prostaglandin E
1
(PGE
1
) was added to the supernatant
to a final concentration of 20 ngÆmL
)1
. The mixture was
incubated for 15 min at room temperature and centrifuged
at 1300 g for 20 min. The precipitated platelets were washed
three times with a modified Tyrode’s buffer (134 m
M
NaCl,
3m
M
KCl, 0.3 m
M
NaH
2
PO
4
,2m
M
MgCl
2

,12m
M
NaHCO
3
,5m
M
glucose, 5 m
M
Hepes, 3.5 mgÆmL
)1
BSA,
1m
M
EGTA) containing 20 ngÆmL
)1
prostaglandin E
1
and
20 ngÆmL
)1
apyrase. The resulting platelets were suspended
with Tyrode’s buffer containing 2 m
M
CaCl
2
.
Inhibition of thrombin-induced platelet aggregation was
measured using washed platelets. Briefly, 550 lLwashed
platelets (3 · 10
5

platelets per lL) in Tyrode’s buffer
containing 2 m
M
CaCl
2
,0.2mgÆmL
)1
fibrinogen, and
1.0 m
M
Gly-Pro-Arg-Pro peptide was preincubated at
37 °C for 3 min. Then 50 lL of a mixture containing
thrombin (0.1 n
M
, final concentration) and recombinant
proteins was added to prewarmed washed platelets. Platelet
aggregation was monitored using an aggregometer.
Results
cDNA cloning and expression of madanin 1 and 2
Three distinct cDNA libraries were constructed from the
salivary glands of H. longicornis at different feeding stages:
1-GCTTTGACCGCAATGAAGCACTTCGCAATTTTGATTCTTGCTGTTGTGGCCAGTGCCGTG
- M K H F A I L I L A V V A S A V
61-GTGATGGCATACCCGGAGAGAGATTCAGCGAAGGAGGGCAACCAAGAGCAAGAGAGAGCT
-V M A Y P E R D S A K E G N Q E Q E R A
121-CTGCATGTAAAGGTACAAAAACGTACTGATGGTGATGCTGACTACGATGAATATGAGGAA
-L H V K V Q K R T D G D A D Y D E Y E E
181-GATGGGACGACTCCTACTCCGGATCCAACTGCACCAACTGCTAAACCACGGCTTCGAGGA
-D G T T P T P D P T A P T A K P R L R G
241-AATAAGCCTTGAATCAATGATGTTCTATTTTTTATAGCGTCCCGATGGCGGTGATGTTGT

-N K P *
301-AGGCTGGAAGCAAATAAAAATACGAAGAGTGACTTCAAAAAAAAAAAAAAAAAAAAAAAA
A
1-GCTTTGACGGCAATGAAGCACTTCGTAATTTTGATTCTTGCTGTTGTGGCCAGTGCCGTG
M K H F V I L I L A V V A S A V
61-GTGATGGCATACCCGGAGAGAGATTCAGCAAAGGACGGCAACCAAGAGAAAGAGAGAGCT
V M A Y P E R D S A K D G N Q E K E R A
121-CTGCTAGTTAAAGTACAAGAACGCTATCAAGGTAATCAAGGTGACTACGATGAATATGAC
L L V K V Q E R Y Q G N Q G D Y D E Y D
181-CAAGATGAGACCACTCCTCCTCCGGATCCAACTGCACAAACTGCAAGACCACGGCTTCGA
Q D E T T P P P D P T A Q T A R P R L R
241-CAAAATCAGGATTGAATCAATGGTGTTCTAGATTTCTATAACCTACCGACGGCGGCAATT
Q N Q D *
301-TTGTGGGGTCCAAACAAATAAAACTACAAAGTGGGACCTCAAAAAAAAAAAAAAAAAAAA
B
Madanin-1 MKHFAILILAVVASAVVMAYPERDSAKEGNQEQERALHVKVQKRTDG-DADYDEYEEDGT
Madanin-2 MKHFVILILAVVASAVVMAYPERDSAKDGNQEKERALLVKVQERYQGNQGDYDEYDQDET
**** ********************** **** **** **** * * ***** * *
Madanin-1 TPTPDPTAPTAKPRLRGNKP
Madanin-2 TPPPDPTAQTARPRLRQNQD
** ***** ** **** *
C
Fig. 1. Nucleotide sequences and deduced
amino-acid sequences of madanin 1 and 2. The
first 19 amino acids are predicted to be the
signal peptide sequences for both madanin 1
(A) and madanin 2 (B). (C) Comparison of
amino-acid sequences of madanin 1 and 2.
Sequence alignment was performed using the
ClustalW program at the Bioinformatics

Center Institute for Chemical Research. The
same amino-acid residues are indicated by
stars.
1928 S. Iwanaga et al.(Eur. J. Biochem. 270) Ó FEBS 2003
unfed, slow feeding, and rapid feeding. A total of 1889
cDNA clones were picked from three cDNA libraries, and
their partial nucleotide sequences determined. Sequence
similarity searches of all cDNA clones were performed using
the
BLAST
program. A signal peptide prediction was also
carried out using the SignalP Program, because many
physiologically active molecules identified from the salivary
gland are secreted proteins. Predicted secreted proteins were
classified into several protein families based on amino-acid
sequence similarities (data not shown).
The major protein family found in the rapid feeding stage
cDNA library was named the Ômadanin familyÕ after the
JapanesenameforH. longicornis, Hutatoge-chimadani.
The madanin family consists of two proteins, madanin 1
and 2 (Fig. 1A,B). They exhibited no sequence similarities
to any other previously identified proteins and shared 79%
sequence identity (Fig. 1C). The cDNA of madanin 1 and 2
contained 240 and 243 bp ORFs encoding 79 and 80
amino-acids residues, respectively. The first 19 amino acids
in both were predicted to be the signal peptide. The
calculated molecular mass of the mature regions of madanin
1 and 2 were 6770.9 Da and 7122.42 Da, respectively.
To investigate the biological activities of madanin 1 and 2,
the recombinant molecules were produced in Escherichia

coli BL21(DE3) cells using expression vector pET22b. Their
expression was confirmed by SDS/PAGE. The recombinant
proteins were purified by gel-filtration chromatography
using Sephadex G75, and purity was evaluated by
RP-HPLC (data not shown). The MALDI spectrum of
madanin 1 and 2 exhibited main [M + H]
+
ions of m/z
6899.39 and 7244.46, respectively (data not shown). As a
methionine residue was added to the N-terminus of the
recombinant proteins when produced using pET22b, the
experimental and calculated masses were almost identical.
Madanin 1 and 2 are novel anticoagulants
of
H. longicornis
It was previously reported that anticoagulants are present in
the salivary glands of H. longicornis [24]. According to this
report, an extract of the salivary glands prolonged both
APTT and PT, suggesting the presence of an inhibitor of
either factor Xa or thrombin. However, the anticoagulant
molecule has not yet been identified. To investigate the
physiological function of madanin 1 and 2, we first
examined their antihemostatic activities using normal
human plasma. As shown in Fig. 2, madanin 1 and 2
prolonged both APTT and PT in a dose-dependent manner,
demonstrating that they are novel anticoagulants in H. lon-
gicornis and inhibit both the intrinsic and extrinsic coagu-
lation pathways.
Thrombin is a target molecule of madanin 1 and 2
The results obtained from the APTT and PT assays suggest

that madanins act as inhibitors of factor Xa and/or
thrombin. Thus, we investigated, by SPR using BIAcore,
their ability to bind to factor Xa and thrombin. All assays
were performed twice with immobilized 821.8 resonance
units (RU) factor Xa or 1037.0 RU thrombin on sensor
chips, as described in Experimental Procedures. This assay
clearly showed that madanin 1 and 2 specifically interacted
with thrombin, but not with factor Xa. Both bound
thrombin in a dose-dependent manner, as shown in
Fig. 3. From these results, it is clear that thrombin is a
target molecule of madanin 1 and 2. However, the
sensorgrams showed abnormal patterns, indicating poor
interaction, in which the dissociation of the complex was
very fast and the reaction between ligand and analyte
equilibrated very rapidly.
As the association and dissociation phases in the inter-
actions of madanin 1 and 2 with thrombin were very short,
it was difficult to analyze the kinetic constants for the inter-
action. Therefore, we evaluated the equilibrium constants
Fig. 2. Effect of madanin 1 and 2 on prolongation of APTT and PT. The
effects of madanin 1 and 2 on intrinsic and extrinsic pathways were
investigated by APTT (A) and PT (B) assays, respectively. Various
concentrations of madanin 1 and 2 were incubated with citrated
human plasma, and the mixture was activated with diluted APTT and
PT reagents. After activation, CaCl
2
was added to the mixture, and
clotting times were measured. (s) Madanin 1; (d) madanin 2.
Ó FEBS 2003 Identification of tick thrombin inhibitors, madanin 1 and 2 (Eur. J. Biochem. 270) 1929
using

BIAEVALUATION
software according to the following
equation:
R
eq
¼ K
a
C=ð1 þ K
a
CÞ
where R
eq
is the value of resonance units in an equilibrated
state, K
a
is the association constant, and C is the concen-
tration of the analyte used in the assay. K
d
was derived from
the relationship, K
d
¼ 1/K
a
. Under the conditions of this
experiment, K
d
values for binding of madanin 1 and 2 to the
immobilized thrombin were 4.18 and 2.96 l
M
, respectively.

The two madanins had similar potencies for interaction with
thrombin.
Madanin 1 and 2 inhibit the conversion of fibrinogen
into fibrin by thrombin
Next, we examined whether madanin 1 and 2 inhibit the
amidolytic activity of thrombin. Assays were performed
using a synthetic substrate for thrombin (S-2238). Thrombin
activity remained fully intact despite a 1000-fold molar
excess of madanin 1 and 2 added to thrombin, indicating
that they do not inhibit amidolytic activity of thrombin with
a small chromogenic substrate (data not shown).
Although madanin 1 and 2 were found by SPR meas-
urements to bind to thrombin, they did not inhibit
hydrolysis of small synthetic substrates by thrombin.
Therefore, we examined whether they inhibit cleavage of
the physiological substrate, fibrinogen. As shown in Fig. 4,
they prolonged fibrin clot formation by thrombin in a dose-
dependent manner, showing that they prevented thrombin
from cleaving fibrinogen and are able to inhibit the function
of thrombin without affecting its amidolytic activity. The
results of these two experiments suggest that thrombin
inhibition by madanin 1 and 2 is caused by competitive
binding to the fibrinogen-binding site (anion-binding exo-
site 1) on the thrombin molecule, and not from binding to
the active site.
Fig. 3. SPR analysis of the interaction between
madanin 1 and 2 and thrombin. Interactions
between thrombin and madanin 1 (A) and
madanin 2 (B) were investigated by SPR
measurement. Thrombin was immobilized on

a sensor chip at levels of 1037.0 RU. Sensor-
grams were obtained by injection of madanin
1 and 2 at different concentrations ranging
from 0.25 l
M
to 5 l
M
at a flow rate of
20 lLÆmin
)1
and are indicated as solid lines.
The sensor chip surface was regenerated with
1
M
NaCl before each injection. The inter-
actions between factor Xa and madanin 1 and 2
were also investigated. Factor Xa was coupled
to a sensor chip at levels of 821.8 RU. The
interactions were measured by injection of
5 l
M
madanin 1 and 2. The sensorgrams are
shown as dotted lines in each case.
1930 S. Iwanaga et al.(Eur. J. Biochem. 270) Ó FEBS 2003
Madanin 1 and 2 inhibit activation of factors V
and VIII and aggregation of platelets by thrombin
It has been reported that the anion-binding exosite 1 is
involved in molecular recognition of thrombin for factor V,
factor VIII, and PAR on platelets. Therefore, we investi-
gated the effects of madanin 1 and 2 on thrombin-catalyzed

activation of factors V and VIII and thrombin-induced
aggregation of platelets. As shown in Figs 5 and 6,
activation of prothrombin and factor X by prothrombinase
and tenase were inhibited, respectively, by the presence of
madanin 1 and 2. It was further confirmed that they do not
inhibit the activity of prothrombinase itself. The activities of
thrombin and factor Xa in each assay reflect the amount of
factors Va and VIIIa, respectively, present. Thus, it is
indicated that madanin 1 and 2 inhibit activation of factors
V and VIII by thrombin and then prevent these cofactors
from forming prothrombinase and tenase complexes. On
the other hand, as shown in Fig. 7, madanin 1 and 2
suppressed thrombin-induced aggregation of platelets in a
dose-dependent manner. Platelet aggregation begins with
proteolytic cleavage of PAR by thrombin. Therefore, these
results also support the conclusion that madanin 1 and 2 are
direct competitive inhibitors of the anion-binding exosite 1
on thrombin.
Discussion
In this study, we have identified novel anticoagulants,
named madanin 1 and 2, from the salivary gland of
H. longicornis. This is the first description of antihemostatic
factors in H. longicornis. Madanin 1 and 2 prolonged both
APTT and PT, indicating that they are anticoagulants for
the common pathway of coagulation. SPR analysis showed
that they specifically interacted with thrombin, not with
factor Xa. An inhibition assay with fibrinogen as substrate
showed that they inhibited the fibrinogenolytic activity of
thrombin. Therefore, we conclude that they inhibit blood
coagulation by inhibiting the function of thrombin.

Madanin 1 and 2 also inhibited the activation of
thrombin-catalyzed cofactors V and VIII as well as the
conversion of fibrinogen into fibrin. Furthermore, they
inhibited thrombin-induced platelet aggregation via proteo-
lytic activation of PAR. However, they did not inhibit the
amidolytic activity of thrombin in an assay using a synthetic
Fig. 4. Effect of madanin 1 and 2 on formation of fibrin clot by thrombin.
Substrate solution containing fibrinogen and arabic gum was prepared
as described in Experimental Procedures and prewarmed at 37 °C.
Thrombin and various concentrations of madanin 1 and 2 were mixed
and added to the substrate solution. The time to fibrin clot formation
was measured using a coagulometer. (s) Madanin 1; (d) madanin 2.
Fig. 5. Effect of madanin 1 and 2 on the activation of factor V by
thrombin. Factor V was preincubated with thrombin in the presence of
madanin 1 or 2 and added to buffer A containing prothrombin
(400 n
M
), factor Xa (20 p
M
), phospholipid (40 l
M
), and CaCl
2
(5 m
M
). Thrombin activity was measured using a chromogenic sub-
strate (S-2238).
Ó FEBS 2003 Identification of tick thrombin inhibitors, madanin 1 and 2 (Eur. J. Biochem. 270) 1931
substrate. It has been reported that fibrinogen and PAR
bind to the anion-binding exosite 1 on thrombin and that

anion-binding exosite 1 and 2 are involved in the inter-
actions between thrombin and its cofactors. Structural data
on thrombin show that anion-binding exosite 1 and 2 are
located opposite each other on the thrombin molecule [25–
27]. Taking the inhibitory profiles and molecular sizes of
madanin 1 and 2 into consideration, it is most likely that
they are competitive inhibitors directed to the anion-binding
exosite 1 of thrombin.
Madanin 1 and 2 show no similarities in their amino-acid
sequences to thrombin inhibitors from any blood-sucking
organisms. However, the clusters of acidic residues found in
the central regions of madanin are similar to those found in
hirudin [17], tsetse thrombin inhibitor [28,29], anophelin
[30,31], and thrombostatin [32]. These acidic regions show
electrostatic interactions with positively charged anion-
binding exosite 1. Furthermore, our recent studies indicate
that N-terminally truncated madanin 1 maintains the ability
to bind to thrombin (unpublished data). Thus, madanin 1
and 2 may bind to anion-binding exosite 1 through the
acidic residue clusters.
The K
d
values of madanin 1 and 2 determined by SPR
analysis (4.18 and 2.96 l
M
, respectively) are significantly
higher than those of other anion-binding exosite 1 inhibitors
[17,28,31]. In the SPR analysis, thrombin was immobilized
Fig. 7. Effect of madanin 1 and 2 on platelet aggregation by thrombin.
Washed platelets were prepared as described in Experimental Proce-

dures. Thrombin and the madanins were mixed and added to the
washed platelets (3 · 10
5
platelets per lL) in the presence of 2 m
M
CaCl
2
,0.2 mgÆmL
)1
fibrinogen, and 1.0 m
M
Gly-Pro-Arg-Pro peptide.
Platelet aggregation was monitored with an aggregometer. Complete
aggregation was obtained in the absence of madanin 1 and 2.
Fig.6.Effectofmadanin1and2onactivationoffactorVIIIby
thrombin. Factor VIII was preincubated with thrombin in the presence
of madanin 1 or 2 and added to buffer A containing factor X (400 n
M
),
factor IXa (1 n
M
), phospholipid (40 l
M
), and CaCl
2
(5 m
M
). Factor
Xa activity was measured using a chromogenic substrate (S-2222).
1932 S. Iwanaga et al.(Eur. J. Biochem. 270) Ó FEBS 2003

by amine coupling. In this immobilizing reaction, the
e-amino group of the lysine residue on the anion-binding
exosite 1 of thrombin may be coupled to the carboxy group
on the sensor chip. This may result in the weak interactions
observed. In fact, the K
d(app)
values calculated from the
results in Figs 5 and 6 by the methods of Henderson [33] are
85–170-fold lower than those obtained from SPR analysis.
The K
d(app)
values of madanin 1 and 2 are 25 and 34.5 n
M
,
respectively. Therefore, it is possible that the K
d
values of
interactions between madanins and thrombin are lower
than those determined by SPR analysis.
Physiological coagulation is initiated by formation of a
factor VIIa–tissue factor complex at the injury site. This
complex activates factor X, which is followed by generation
of small amounts of thrombin [34]. The generated thrombin
further activates factor V and factor VIII, which leads to the
generation of a large amount of thrombin. This amplifica-
tion step is thought to be crucial for physiological coagu-
lation, as deficiencies of factor V and factor VIII cause
serious bleeding. Therefore, it is possible that madanins
inhibit blood coagulation at this initial step by inhibiting the
activation of factor V and factor VIII by thrombin and

contribute considerably to tick blood feeding.
In conclusion, novel thrombin inhibitors(madanin 1and2)
from H. longicornis have been identified. They inhibit
various physiological functions of thrombin without inter-
fering with its catalytic activity and may play an important
role in tick blood feeding. Recent studies report that anion-
binding exosite 1 inhibitors such as the C-terminal peptide of
hirudin are attractive therapeutic drugs for arterial throm-
bosis [35]. Thus, further studies on the inhibitory mechanisms
of madanin 1 and 2 may provide useful information for the
development of therapeutic agents for thrombosis.
Acknowledgements
This work was supported by a grant from the Japan Society for
Promoting Science: Future Developmental Research (to Y. C.) and by
a grant from Mitsubishi Pharma Research Foundation (to S. I.).
References
1. Davie, E.W., Fujikawa, K. & Kisiel, W. (1991) The coagulation
cascade: initiation, maintenance, and regulation. Biochemistry 30,
10363–10370.
2. Kane, W.H. & Davie, E.W. (1988) Blood coagulation factors V
and VIII: structural and functional similarities and their
relationship to hemorrhagic and thrombotic disorders. Blood 71,
539–555.
3. Gailani, D. & Broze, G.J. Jr (1991) Factor XI activation in a
revised model of blood coagulation. Science 253, 909–912.
4. Walker, F.J., Sexton, P.W. & Esmon, C.T. (1979) The inhibition
of blood coagulation by activated Protein C through the selective
inactivation of activated Factor V. Biochim. Biophys. Acta 571,
333–342.
5. Esmon, C.T. (1987) The regulation of natural anticoagulant

pathways. Science 235, 1348–1352.
6. Vu, T.K., Hung, D.T., Wheaton, V.I. & Coughlin, S.R. (1991)
Molecular cloning of a functional thrombin receptor reveals
a novel proteolytic mechanism of receptor activation. Cell 64,
1057–1068.
7. Coughlin, S.R. (1999) How the protease thrombin talks to cells.
Proc.NatlAcad.Sci.USA96, 11023–11027.
8. Fenton,J.W.II,Olson,T.A.,Zanbinski,M.P.&Wilner,G.D.
(1988) Anion-binding exosite of human alpha-thrombin and fibrin
(ogen) recognition. Biochemistry 27, 7106–7112.
9. Coughlin, S.R. (2000) Thrombin signaling and protease-activated
receptors. Nature (London) 407, 258–264.
10. Wu, Q.Y., Sheehan, J.P., Tsiang, M., Lentz, S.R., Birktoft, J.J. &
Sadler, J.E. (1991) Single amino acid substitutions dissociate
fibrinogen-clotting and thrombomodulin-binding activities of
human thrombin. Proc. Natl Acad. Sci. USA 88, 6775–6779.
11. Tsiang, M., Jain, A.K., Dunn, K.E., Rojas, M.E., Leung, L.K. &
Gibbs, C.S. (1995) Functional mapping of the surface residues of
human thrombin. J. Biol. Chem. 270, 16854–16863.
12. Fenton, J.W. II, Witting, J.I., Pouliott, C. & Fareed, J. (1989)
Thrombin anion-binding exosite interactions with heparin and
various polyanions. Ann. NY Acad. Sci. 556, 158–165.
13. Rosenberg, R.D. & Damus, P.S. (1973) The purification and
mechanism of action of human antithrombin-heparin cofactor.
J. Biol. Chem. 248, 6490–6505.
14. Tollefsen, D.M., Pestka, C.A. & Monafo, W.J. (1982) Heparin
cofactor. II. Purification and properties of a heparin-dependent
inhibitor of thrombin in human plasma. J. Biol. Chem. 257, 2162–
2169.
15. Esmon, C.T. & Lollar, P. (1996) Involvement of thrombin anion-

binding exosites 1 and 2 in the activation of factor V and factor
VIII. J. Biol. Chem. 271, 13882–13887.
16. Arocha-Pinango, C.L., Marchi, R., Carvajal, Z. & Guerrero, B.
(1999) Invertebrate compounds acting on the hemostatic
mechanism. Blood Coag. Fibrinol. 10, 43–68.
17. Markwardt, F. & Walsmann, P. (1967) Purification and analysis
of the thrombin inhibitor hirudin. Hoppe Seylers Z. Physiol.
Chem. 348, 1381–1386.
18. Bode, W. & Huber, R. (1991) Refined structure of the hirudin–
thrombin complex. J. Mol. Biol. 221, 583–601.
19. Rydel, T.J., Ravichandran, K.G., Tulinsky, A., Bode, W., Huber,
R., Roitsch, C. & Fenton, J.W. II. (1990) The structure of a
complex of recombinant hirudin and human alpha-thrombin.
Science 249, 277–280.
20. Naski, M.C., Fenton, J.W. II, Maranagore, J.M., Olson, S.T. &
Shafer, J.A. (1990) The COOH-terminal domain of hirudin. An
exosite-directed competitive inhibitor of the action of alpha-
thrombin on fibrinogen. J. Biol. Chem. 265, 13484–13489.
21. Tsiang, M., Lentz, S.R., Dittman, W.A., Wen, D., Scarpati, E. &
Sadler, J.E. (1990) Equilibrium binding of thrombin to recom-
binant human thrombomodulin: effect of hirudin, fibrinogen,
factor Va, and peptide analogues. Biochemistry 29, 10602–10612.
22. Hortin, G.L. & Trimpe, B.L. (1991) Allosteric changes in
thrombin’s activity produced by peptides corresponding to seg-
ments of natural inhibitors and substrates. J. Biol. Chem. 266,
6866–6871.
23. Vu,T.K.H.,Wheaton,V.,Hung,D.T.,Charo,I.&Coughlin,
S.R. (1991) Domains specifying thrombin–receptor interaction.
Nature (London) 353, 674–677.
24. Anastopoulos, P., Thurn, M.J. & Broady, K.W. (1991) Anti-

coagulant in the tick Ixodes holocyclus. Aust.Vet.J.68, 366–367.
25. Stubb, M.T. & Bode, W. (1993) A player of many parts: the
spotlight falls on thrombin’s structure. Thromb. Res. 69, 1–58.
26. Bode, W., Turk, D. & Karshikov, A. (1992) The refined 1.9-A
˚
X-ray crystal structure of D-Phe-Pro-Arg chloromethylketone-
inhibited human alpha-thrombin: structure analysis, overall
structure, electrostatic properties, detailed active-site geometry,
and structure-function relationships. Protein Sci. 1, 426–471.
27. Di Cera, E., Dang, Q.D. & Ayala, Y. (1997) Molecular mechan-
isms of thrombin function. Cell Mol. Life Sci. 53, 701–730.
28. Cappello, M., Bergum, P.W., Vlasuk, G.P., Furmidge, B.A.,
Pritchard, D.I. & Aksoy, S. (1996) Isolation and characterization
of the tsetse thrombin inhibitor: a potent antithrombotic peptide
Ó FEBS 2003 Identification of tick thrombin inhibitors, madanin 1 and 2 (Eur. J. Biochem. 270) 1933
from the saliva of Glossina morsitans morsitans. Am.J.Trop.Med.
Hyg. 54, 475–480.
29. Cappello, M., Li, S., Ghen, X., Li, C.B., Harrison, L., Nara-
shimhan, S., Beard, C.B. & Aksoy, S. (1998) Tsetse thrombin
inhibitor: blood meal-induced expression of an anticoagulant in
salivary glands and gut tissue of Glossina morsitans morsitans.
Proc.NatlAcad.Sci.USA95, 14290–14295.
30. Valenzuela, J.G., Francischetti, I.M.B. & Ribeiro, J.M. (1999)
Purification, cloning, and synthesis of a novel salivary anti-
thrombin from the mosquito Anopheles albimanus. Biochemistry
38, 11209–11215.
31. Francischetti, I.M.B., Valenzuela, J.G. & Ribeiro, J.M. (1999)
Anophelin: kinetics and mechanism of thrombin inhibition. Bio-
chemistry 38, 16678–16685.
32. Zhang, D., Cupp, M.S. & Cupp, E.W. (2002) Thrombostasin:

purification, molecular cloning and expression of a novel anti-
thrombin protein from horn fly saliva. Insect Biochem. Mol. Biol.
32, 321–330.
33. Henderson, P.J.F. (1972) A linear equation that describes
the steady-state kinetics of enzymes and subcellular particles
interacting with tightly bound inhibitors. Biochem. J. 127, 321–
333.
34. Haffman, M. & Monroe, D.M. (2001) A cell-based model of
hemostasis. Thromb. Haemost. 85, 958–965.
35. Bates, S.M. & Weitz, J.I. (1998) Direct thrombin inhibitors for
treatment of arterial thrombosis: potential differences between
bivalirudin and hirudin. Am.J.Cardiol.82, 12–18.
1934 S. Iwanaga et al.(Eur. J. Biochem. 270) Ó FEBS 2003