Tissue factor pathway inhibitor is highly susceptible
to chymase-mediated proteolysis
Tsutomu Hamuro
1
, Hiroshi Kido
2
, Yujiro Asada
3
, Kinta Hatakeyama
3
, Yuushi Okumura
2
,
Youichi Kunori
4
, Takashi Kamimura
4
, Sadaaki Iwanaga
1
and Shintaro Kamei
1
1 Therapeutic Protein Products Research Department, The Chemo-Sero-Therapeutic Research Institute, Kaketsuken, Japan
2 Division of Enzyme Chemistry, Institute for Enzyme Research, University of Tokushima, Japan
3 Department of Pathology, Faculty of Medicine, University of Miyazaki, Japan
4 Institute for Biomedical Research, Teijin Pharma Limited, Japan
Tissue factor pathway inhibitor (TFPI) is the main
inhibitor of tissue factor-induced blood coagulation.
Human TFPI contains 276 amino acids that comprise
an acidic N-terminal domain followed by three tandem
Kunitz-type trypsin inhibitor domains and a C-ter-
minal basic amino-acid cluster region [1]. The first

ase inhibitor that primarily inhibits the extrinsic pathway of blood coagula-
tion. It is synthesized by various cells and its expression level increases in
inflammatory environments. Mast cells and neutrophils accumulate at sites
of inflammation and vascular disease where they release proteinases as well
as chemical mediators of these conditions. In this study, the interactions
between TFPI and serine proteinases secreted from human mast cells and
neutrophils were examined. TFPI inactivated human lung tryptase, and its
inhibitory activity was stronger than that of antithrombin. In contrast,
mast cell chymase rapidly cleaved TFPI even at an enzyme to substrate
molar ratio of 1 : 500, resulting in markedly decreased TFPI anticoagulant
and anti-(factor Xa) activities. N-Terminal amino-acid sequencing and MS
analyses of the proteolytic fragments revealed that chymase preferentially
cleaved TFPI at Tyr159-Gly160, Phe181-Glu182, Leu89-Gln90, and
Tyr268-Glu269, in that order, resulting in the separation of the three indi-
vidual Kunitz domains. Neutrophil-derived proteinase 3 also cleaved TFPI,
but the reaction was much slower than the chymase reaction. In contrast,
a-chymotrypsin, which shows similar substrate specificities to those of chy-
mase, resulted in a markedly lower level of TFPI degradation. These data
indicate that TFPI is a novel and highly susceptible substrate of chymase.
We propose that chymase-mediated proteolysis of TFPI may induce a
thrombosis-prone state at inflammatory sites.
Abbreviations
pNA, p-nitroanilide; TFPI, tissue factor pathway inhibitor; TFPI-C, TFPI with truncated C-terminal basic-amino-acid region.
FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS 3065
various stimuli produced in inflammatory states [6–9].
It is well known that neutrophils and mast cells are
important effector cells at sites of vascular perturba-
tion. These cells release secretory granules that contain
a variety of biologically active substances. In parti-
cular, neutrophil-derived proteolytic enzymes partici-

of TFPI may induce a thrombosis-prone state at
inflammatory sites.
Results
Inhibitory properties of TFPI on proteases
To investigate the inhibitory properties of TFPI against
human mast cell-derived and neutrophil-derived pro-
teinases, inhibition assays were performed using appro-
priate synthetic substrates. For the mast cell-derived
proteinases, TFPI inhibited the activity of tryptase with
a 50% inhibitory concentration (IC
50
)of 10 lm,
whereas it did not inhibit the activity of chymase
(Fig. 1A,B). Next, we tested the effects of TFPI on the
amidolytic activities of elastase, cathepsin G, and neu-
trophil-derived proteinase 3. As shown in Fig. 1C,D,
TFPI inhibited the amidolytic activities of elastase
(IC
50
¼ 1.4 lm) and cathepsin G (IC
50
¼ 0.13 lm),
which agrees with a previous report [12]. In contrast,
TFPI produced only weak inhibition of the amidolytic
activity of proteinase 3 (Fig. 1E), even though this pro-
tein is structurally similar to elastase and cathepsin G.
Inhibitory properties of TFPI derivatives
on tryptase
It is well known that the conversion of tryptase into
an active tetrameric form requires sulfated polysaccha-

previous report that used heparin antagonists [17].
Proteolysis of TFPI by chymase and other
proteinases
To examine whether TFPI is degraded by mast cell-
derived and neutrophil-derived proteinases, each protei-
nase was incubated with TFPI at a molar ratio of
1 : 500. Surprisingly, chymase rapidly cleaved TFPI,
even at a low enzyme to substrate ratio. As shown in
Fig. 3A, TFPI was cleaved by chymase within 15 min,
as evidenced by the two smeared bands (bands 1 and
2) with molecular masses of 20–30 kDa observed on
SDS ⁄ PAGE. Subsequently, the level of the approxi-
mately 15-kDa band increased, and TFPI (43 kDa) as
well as bands 1 and 2 disappeared entirely after incu-
bation for 20 h. These results indicate that TFPI was
completely converted into fragments with molecular
masses of  15 kDa by chymase. Chymostatin, an
inhibitor of chymotrypsin-type serine proteinases, com-
pletely blocked the chymase-mediated proteolysis of
Chymase-mediated proteolysis of TFPI T. Hamuro et al.
3066 FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS
Chymase activity
(% of control)
A
TFPI (µM)
B
Tryptase activity
(% of control)
TFPI (µM)
TFPI (µ

20
40
60
80
100
120
0
20
40
60
80
100
120
0
20
40
60
80
0
20
40
60
80
100
120
Fig. 1. Inhibition of mast cell-derived and neutrophil-derived proteinases by TFPI. Each proteinase was incubated with various concentrations
of TFPI at 37 °C, and the residual proteinase activity was measured. (A) Chymase (7 n
M) was assayed using Suc-Ala-Ala-Pro-Phe-pNA
(5 m
M). (B) Tryptase (14 nM) was assayed using H-D-Ile-Pro-Arg-pNA (0.15 mM). (C) Neutrophil elastase (40 nM) was assayed using MeO-

10
1
0
0
20
40
60
80
100
120
Tryptase activity
(% of control)
Tryptase activity
(% of control)
Tryptase activity
(% of control)
Fig. 2. Inhibition of tryptase by TFPI deriva-
tives and antithrombin. TFPI (A), TFPI-C (B),
and antithrombin (C) were incubated with
tryptase in the presence of 0.5 lgÆmL
)1
(d)
or 500 lgÆmL
)1
heparin (s) for 60 min. A
chromogenic substrate was then added, and
the amidolytic activity of tryptase was
measured. Data are presented as the
mean ± SD from three independent experi-
ments.

To characterize the proteolysis of TFPI by chymase,
we digested TFPI with chymase for 20 h and separated
the resulting peptides using RP-HPLC. As shown in
Fig. 4A, six major peptide peaks were separated using
97
66
45
30
20.1
14.4
kDa
Band 1
Band 2
Mr 0 15 30 60 120 180 1200
min
97
66
45
30
20.1
14.4
kDa
Mr 0 15 30 60 120 180
min
AB
97
66
45
30
20.1

kDa
Mr 0 15 30 60 120 180
min
Fig. 3. Cleavage of TFPI by mast cell-derived and neutrophil-derived serine proteinases. Chymase (A), tryptase (B), a-chymotrypsin (C), neu-
trophil elastase (D), cathepsin G (E), or proteinase 3 (F) at a concentration of 7 n
M was incubated with TFPI (3.5 lM) for the indicated times
at 37 °C. Proteins were separated on 15–25% polyacrylamide gels under reducing conditions and the gels were stained with Coomassie Bril-
liant Blue. Intermediate degradation products were designated as bands 1 and 2 (A). Mr, Low molecular mass marker.
Chymase-mediated proteolysis of TFPI T. Hamuro et al.
3068 FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS
this type of chromatography. Four of them, designated
peaks III, IV, V, and VI, all migrated at  15 kDa
during SDS ⁄ PAGE (Fig. 4B). The broad peak IV was
also visualized as a smeared band on the gel, even
though only a single N-terminal residue was detected
for this fragment. Because recombinant TFPI contains
a variety of carbohydrate chains, this fragment may
have included N-linked carbohydrate chains [18]. To
identify the chymase cleavage sites in TFPI, the frag-
ments that resulted in these six peaks were analyzed to
determine their amino-acid compositions, N-terminal
sequences, and mass spectra. The results of these ana-
lyses are summarized in Table 1. Peaks III and IV cor-
responded to the third and second Kunitz domains of
TFPI, respectively. Multiple MS signals were observed
for the peak IV fragment, supporting the idea that this
fragment was glycosylated. Peaks V and VI both cor-
responded to the first Kunitz domain of TFPI; only
peak VI, however, was consistent with the calculated
MS value of this domain. We assumed that the peak V

Absorbance at 214nm
Peak I
Peak II
Peak III
Peak IV
Peak VI
Peak V
66
45
30
20.1
14.4
kDa
Mr
Peak III
BA
Peak IV
Peak V
Peak VI
97
Fig. 4. Separation of TFPI degradation products produced by treatment with chymase. (A) TFPI digested with chymase for 20 h was applied
to a Vydac C8 column that had been equilibrated with 0.1% trifluoroacetic acid. The products were eluted with 0.1% trifluoroacetic acid con-
taining a linear concentration gradient of acetonitrile from 0% to 50%. (B) SDS ⁄ PAGE of the peak III, IV, V, and VI fractions after RP-HPLC.
Mr, Low molecular mass marker.
Table 1. N-Terminal amino-acid sequences and mass spectra of degradation products. The individual fragments designated in Figs 2 and 3
were separated by RP-HPLC, and the amino-acid sequences and mass spectra were determined.
Fragment Sequence Cleavage site Observed mass (m ⁄ z) Calculated mass (m ⁄ z) Deduced structure
Band 1
a
DSEEDEE –– –

sequences starting with Gln90, Gly160, Glu182, and
Glu269 were observed. These sequences were present
at an approximate molar ratio of 2 : 5 : 4 : 1, respect-
ively. These data revealed that chymase first cleaved
TFPI at Tyr159-Gly160 to generate two molecules
(bands 1 and 2), which was followed by cleavage at
Phe181-Glu182 and Leu89-Gln90 to generate the frag-
ments corresponding to peaks IV, V, and VI. Finally,
chymase-mediated cleavage at Tyr268-Glu269 gener-
ated the peaks I and III. Taken together, these data
indicate that chymase selectively cleaved TFPI into five
fragments that were not disulfide linked, three of
which contained individual Kunitz domains.
Anticoagulant activity of the TFPI degradation
products produced by chymase treatment
The effects of the chymase-mediated degradation on
TFPI function were evaluated by testing the residual
anticoagulant activities of the fragments and also by
determining the amount of residual TFPI antigen
using an ELISA. Figure 5 shows the time course of
the decrease in anticoagulant and anti-(factor Xa)
activities determined under conditions in which 70 nm
TFPI was incubated at 37 °C with 7 nm chymase. As
shown in Fig. 5A, TFPI antigen promptly disap-
peared; the two fragments generated by cleavage at
Tyr159-Gly160 were not detected with this ELISA sys-
tem. The anticoagulant and anti-(factor Xa) activities
also decreased in a time-dependent manner; after 5 min,
however, these activities decreased at relatively slow
rates (Fig. 5B,C). In particular, 20% of the anti-

0
20
40
60
80
100
Incubation time (min)
Anticoagulant activity
(% of control)
TFPI antigen
(% of control)
Anti-factor Xa activity
(% of control)
A
B
C
Fig. 5. Effects of chymase on the anticoagulant and anti-(factor Xa)
activities of TFPI. TFPI (70 n
M) was incubated with chymase (7 nM)
at 37 °C. After various incubation times, the reaction was termin-
ated with chymostatin, and aliquots of the sample were subjected
to ELISA, a dilute tissue factor clotting assay, and anti-(factor Xa)
assay (see Experimental procedures). (A) Residual TFPI antigen. (B)
Residual anticoagulant activity of TFPI. (C) Residual anti-(factor Xa)
activity of TFPI. Data are presented as the mean ± SD from three
independent experiments.
Chymase-mediated proteolysis of TFPI T. Hamuro et al.
3070 FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS
because the efficiency of enzymatic proteolysis was
estimated on the basis of the amount of remaining

matrices, lung, heart, and epithelial surfaces, are
effector cells that participate in innate and acquired
immunity [23–26]. In pathological conditions, such as
inflammation, fibrosis, and malignancy, mast cells as
well as neutrophils and macrophages accumulate at
the affected sites. Recent studies indicate that mast
cells also accumulate at sites of atrial appendages
[27], deep venous thrombosis [28], periprostate vein
thrombosis [29], and atherosclerotic plaques [30–33].
These findings imply that mast cells are involved in
thrombosis and fibrinolysis. In fact, mast cells express
tissue-type plasminogen activator and urokinase-type
plasminogen activator receptor [34]. Little, however,
is known about the functional roles of proteinases
released from mast cell granules during thrombosis
and other pathological states. In this report, we
focused on the reactivity of TFPI with serine protein-
ases released from mast cells.
Table 2. Apparent kinetic constants for the proteolytic cleavage of
TFPI by chymase, neutrophil elastase, and proteinase 3. The velo-
city of TFPI degradation was measured using an ELISA as des-
cribed in Experimental procedures. Kinetic constants were
calculated from a Lineweaver–Burk plot. Values are expressed as
the mean ± SD from three independent experiments.
Enzyme Substrate
K
m
(lM)
k
cat

30
20.1
14.4
B
TFPI antigen
(% of control)
Incubation time (min)
180
12060
0
0
20
40
60
80
100
Mr
Fig. 6. Effects of various polysaccharides on chymase-mediated proteolysis of TFPI. (A) TFPI (3.5 lM) was incubated with chymase (7 nM)
for 60 min at 37 °C in the presence of each polysaccharide (100 lgÆmL
)1
). Proteins were separated on 15–25% polyacrylamide gels under
reducing conditions, and the gels were stained with Coomassie Brilliant Blue. The arrow indicates TFPI. (B) TFPI (70 n
M) was incubated with
chymase (7 n
M) in the presence or absence of heparin (100 lgÆmL
)1
) for up to 180 min. After various incubation times, the reaction was ter-
minated with chymostatin, and the level of TFPI that remained was measured using an ELISA. (d) Control incubation; (j) incubation with
low molecular weight heparin (LMWH); (h) incubation with unfractionated heparin (UFH). Mr, Low molecular mass marker.
T. Hamuro et al. Chymase-mediated proteolysis of TFPI

which TFPI inactivates tryptase requires further
investigation.
Secondly, we found that chymase efficiently cleaved
TFPI, even at a very low enzyme to substrate molar
ratio (1 : 500). As shown in Fig. 7, TFPI is known to
be degraded by several proteinases, including throm-
bin, plasmin, factor Xa, matrix metalloproteinases,
and neutrophil elastase [11–14,38–40]. Those results,
however, were obtained from reactions performed at
high enzyme to substrate molar ratios, or after long
incubations. The present study revealed that chymase
selectively cleaves TFPI at four peptide bonds
(Tyr159-Gly160, Phe181-Glu182, Leu89-Gln90, and
Tyr268-Glu269 in that order), which separated the
three individual Kunitz inhibitor domains and abol-
ished the anticoagulant activity of TFPI. The previ-
ously reported natural substrates of human chymase
include angiotensin I [41], bradykinin [42], C1-inhibitor
[43], interleukin-1b [44], neurotensin [45], interstitial
procollagenase (proMMP-1) [46], kit ligand [47], big
endothelins [48], type-I procollagen [49], a2-macroglo-
bulin [50], profilin [51], albumin [52], and connective
tissue-activating peptide III [53]. The cleaving sites of
these natural substrates are summarized in Table 3.
Using a combinatorial peptide screening method, Ray-
mond et al. [52] demonstrated that chymase preferen-
tially acts at sites with Tyr or Phe as the P1 residue,
which is supported by the presence of these residues at
the P1 positions in the natural substrates (Table 3). In
agreement with those results, we found that three of

14
Fig. 7. Schematic structure of TFPI and cleavage sites by chymase.
The cleavage sites in TFPI are summarized in this figure. Data
obtained in this study including the four chymase cleavage sites
are shown below the TFPI structure, whereas previously deter-
mined data are shown above the TFPI structure. The thick arrows
indicate the locations of the sites cleaved by chymase, which
include the amino acids and residue numbers. The open circles and
branches indicate O-linked glycosylation sites and N-linked glycosy-
lation sites, respectively. Our findings suggest that the threonine
residue at amino-acid position 14 carried an O-linked carbohydrate
in half of the TFPI molecules used here. The solid bar indicates a
‘hot region’, which contains cleavage sites for thrombin (IIa), plas-
min (Pm), factor Xa (Xa), neutrophil elastase, proteinase 3, and
chymase. The thin arrows indicate the cleavage sites for each pro-
teolytic enzyme. MMP, Matrix metalloproteinase.
Table 3. Sites of hydrolysis of natural substrates of human chy-
mase.
P1 site P4–P1_P1¢ Substrate Reference
Tyr NEAY_V Interleukin 1b [44]
RRPY_I Neurotensin [45]
VVPY_G Endothelin-1 [48]
VGFY_E a2-Macroglobulin [50]
RETY_G Albumin [52]
VDNY_G TFPI This work
KIAY_E TFPI This work
Phe IHPF_H Angiotensin I [41]
FSPF_R Bradykinin [42]
KMLF_V C1 inhibitor [43]
TKPF_M Kit ligand [47]

chymase, because both of these proteins bind to hep-
arin [55,56]. Heparin, which is produced and secreted
by mast cells, did not inhibit the cleavage of TFPI by
chymase. Moreover, heparan sulfate did not influence
the proteolysis of TFPI. It was reported that heparin
has no affect on the amidolytic activity of chymase
for a chromogenic substrate, whereas it inhibited the
chymase-mediated proteolysis of casein and angio-
tensin I [56,57], suggesting that the regulation of chy-
mase activity by heparin is dependent on the
substrate. Therefore, cell-surface TFPI, which is bound
to proteoglycans, could be cleaved by chymase.
Although Valentin & Schousboe [58] reported that
TFPI interacts with acidic phospholipids such as phos-
phatidylserine in vitro, we found that phosphatidyl-
serine did not affect the cleavage of TFPI (data not
shown).
The human gastrointestinal tract contains numerous
mast cells, which are located primarily in the lamina
propria mucosa. We confirmed that a large number
of chymase-positive mast cells are located around
microvessels in the lamina propria mucosa, and TFPI
was detected on the intraluminal surface of these
microvessels (K. Hatakeyama and Y. Asada, unpub-
lished data). It was previously reported that human
intestinal mast cells produce and release tumor necro-
sis factor-a in response to Gram-negative bacteria
such as Escherichia coli [59]. Furthermore, tumor nec-
rosis factor-a induces the expression of tissue factor
on vascular endothelial cells [60]. Clot formation

epsin G were obtained from Calbiochem (La Jolla, CA,
USA). Human neutrophil proteinase 3 was purchased from
Athens Research and Technology (Athens, GA, USA).
Bovine a-chymotrypsin was obtained from Worthington
Biochemical Corp. (Lakewood, NJ, USA). Recombinant
human chymase was expressed in Trichoplusia ni insect cells
using a baculovirus expression system and purified from the
culture medium as described previously [61]. Activated
human factor X (factor Xa) was prepared by incubating
purified factor X with Russell’s viper venom factor X activa-
tor (Haematologic Technologies, Essex Junction, VT, USA)
and then separating factor Xa by gel filtration on a column
of Sephacryl S-200 (Amersham Biosciences, Piscataway,
NJ, USA) as described in a previous paper [62]. Human
antithrombin was purified from human plasma using a
procedure based on heparin affinity chromatography [63].
Recombinant human TFPI was expressed in Chinese ham-
ster ovary (CHO) cells and purified from the culture medium
as described previously [4]. TFPI-C, which lacked the C-ter-
minal basic region and ended at Lys249, was separated from
T. Hamuro et al. Chymase-mediated proteolysis of TFPI
FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS 3073
full-length TFPI [4]. TFPI expressed in CHO cells had
N-linked carbohydrate chains at Asn117 and Asn167, and
O-linked carbohydrate chains at Ser174 and Thr175 [18].
Inhibition assay of TFPI
All inhibition experiments were performed in Tris ⁄ NaCl
buffer (50 mm Tris ⁄ HCl containing 150 mm NaCl, pH 7.5)
at 37 °C in 96-well microtiter plates. TFPI was incubated
with each proteinase and synthetic substrate, and the

using chymase, a-chymotrypsin, elastase, cathepsin G, and
proteinase 3 were carried out in Tris ⁄ NaCl. Reactions using
tryptase were carried out in Tris ⁄ NaCl containing
500 lgÆmL
)1
unfractionated heparin. At the indicated time
points, samples were taken from the reaction mixture and
subjected to SDS ⁄ PAGE under reducing conditions. Pro-
tein bands were visualized by staining with Coomassie Bril-
liant Blue R-250.
RP-HPLC
RP-HPLC was carried out on a Vydac 208TP54 C8-300
column (Cypress International Ltd, Tokyo, Japan). After
the sample was injected, the column was washed with
a solution of 0.1% trifluoroacetic acid for 10 min. TFPI
fragments were eluted with a linear gradient of this
solution containing 24–37% acetonitrile at a flow rate of
1.0 mLÆmin
)1
. Each peak fraction was pooled, lyophilized,
and dissolved in water for further analyses.
Amino-acid composition analysis, N-terminal
sequencing, and MS analysis
The amino-acid compositions of the fragments were ana-
lyzed using an AccQTag
TM
system (Waters, Milford, MA,
USA) according to the manufacturer’s protocol. Automa-
ted Edman degradation was carried out using an Applied
Biosystems 492 protein sequencer and standard methods.

30-min incubation, development was terminated by the
addition of 100 lL 0.5 m H
2
SO
4
, and the absorbance at
405 nm was measured using a THERMOmax microplate
spectrometer. The concentration of TFPI was calculated
from a standard curve prepared with known amounts of
TFPI.
Dilute tissue factor clotting assay
TFPI (70 nm) was incubated with chymase (7 nm)at37°C
in Tris ⁄ NaCl. After various incubation times, the reaction
was terminated with 100 lm chymostatin, and a 15-lL ali-
quot of the sample was added to 135 lL human control
plasma and mixed for 2 min at 37 °C. Then, 150 lL throm-
boplastin [diluted 1 : 40 in 50 mm Tris ⁄ HCl (pH 7.0)
Chymase-mediated proteolysis of TFPI T. Hamuro et al.
3074 FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS
containing 100 mm NaCl, 30 mm CaCl
2
, and 0.2% BSA]
was added to the sample ⁄ plasma mixture, and the clotting
time was measured using a Fibrintimer coagulometer (Dade
Behring Ltd, Tokyo, Japan). The concentration of the dilu-
ted thromboplastin was adjusted to yield clotting times
of about 30 s in the absence of TFPI. The clotting time
was related to the anticoagulant activity (expressed as the
percentage of the control sample) using a reference curve
constructed with known amounts of TFPI.

Burk plot.
Acknowledgements
This work was supported through Special Coordina-
tion Funds of the Ministry of Education, Culture,
Sports, Science and Technology, the Japanese Govern-
ment (to TH and SK). We thank Dr Yu-ichi Kami-
kubo for valuable advice and helpful discussion. We
also thank Kumiko Arita, Rumiko Onitsuka and
Michiko Kihara for technical assistance.
References
1 Wun TC, Kretzmer KK, Girard TJ, Miletich JP &
Broze GJ Jr (1988) Cloning and characterization of a
cDNA coding for the lipoprotein-associated coagulation
inhibitor shows that it consists of three tandem Kunitz-
type inhibitory domains. J Biol Chem 263, 6001–6004.
2 Petersen LC, Bjorn SE, Olsen OH, Nordfang O, Norris
F & Norris K (1996) Inhibitory properties of separate
recombinant Kunitz-type-protease-inhibitor domains
from tissue-factor-pathway inhibitor. Eur J Biochem
235, 310–316.
3 Wesselschmidt R, Likert K, Girard T, Wun TC & Broze
GJ Jr (1992) Tissue factor pathway inhibitor: the car-
boxy-terminus is required for optimal inhibition of fac-
tor Xa. Blood 79, 2004–2010.
4 Enjyoji K, Miyata T, Kamikubo Y & Kato H (1995)
Effect of heparin on the inhibition of factor Xa by
tissue factor pathway inhibitor: a segment, Gly212-
Phe243, of the third Kunitz domain is a heparin-binding
site. Biochemistry 34, 5725–5735.
5 Mine S, Yamazaki T, Miyata T, Hara S & Kato H

13 Belaaouaj AA, Li A, Wun TC, Welgus HG & Shapiro
SD (2000) Matrix metalloproteinases cleave tissue factor
pathway inhibitor. Effects on coagulation. J Biol Chem
275, 27123–27128.
14 Cunningham AC, Hasty KA, Enghild JJ & Mast AE
(2002) Structural and functional characterization of
tissue factor pathway inhibitor following degradation
by matrix metalloproteinase-8. Biochem J 367, 451–458.
15 Schwartz LB & Bradford TR (1986) Regulation of tryp-
tase from human lung mast cells by heparin. Stabiliza-
tion of the active tetramer. J Biol Chem 261, 7372–7379.
T. Hamuro et al. Chymase-mediated proteolysis of TFPI
FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS 3075
16 Pereira PJ, Bergner A, Macedo-Ribeiro S, Huber R,
Matschiner G, Fritz H, Sommerhoff CP & Bode W
(1998) Human beta-tryptase is a ring-like tetramer with
active sites facing a central pore. Nature 392, 306–311.
17 Hallgren J, Estrada S, Karlson U, Alving K & Pejler G
(2001) Heparin antagonists are potent inhibitors of mast
cell tryptase. Biochemistry 40, 7342–7349.
18 Nakahara Y, Miyata T, Hamuro T, Funatsu A,
Miyagi M, Tsunasawa S & Kato H (1996) Amino
acid sequence and carbohydrate structure of a recom-
binant human tissue factor pathway inhibitor
expressed in Chinese hamster ovary cells: one N- and
two O-linked carbohydrate chains are located between
Kunitz domains 2 and 3 and one N-linked carbo-
hydrate chain is in Kunitz domain 2. Biochemistry 35,
6450–6459.
19 Sandset PM, Abildgaard U & Larsen ML (1988)

D, Sperr WR, Grossschmidt K, Bankl H, Lechner K &
Valent P (1995) Increase and redistribution of cardiac
mast cells in auricular thrombosis. Possible role of kit
ligand. Circulation 91, 275–283.
28 Bankl HC, Grossschmidt K, Pikula B, Bankl H,
Lechner K & Valent P (1999) Mast cells are augmented
in deep vein thrombosis and express a profibrinolytic
phenotype. Hum Pathol 30, 188–194.
29 Bankl HC, Samorapoompichit P, Pikula B, Latinovic L,
Bankl H, Lechner K & Valent P (2001) Characterization
of human prostate mast cells and their increase in peri-
prostatic vein thrombosis. Am J Clin Pathol 116,
97–106.
30 Kaartinen M, Penttila A & Kovanen PT (1994) Accu-
mulation of activated mast cells in the shoulder region
of human coronary atheroma, the predilection site of
atheromatous rupture. Circulation 90, 1669–1678.
31 Kaartinen M, Penttila A & Kovanen PT (1994) Mast
cells of two types differing in neutral protease composi-
tion in the human aortic intima. Demonstration of tryp-
tase- and tryptase ⁄ chymase-containing mast cells in
normal intimas, fatty streaks, and the shoulder region
of atheromas. Arterioscler Thromb 14, 966–972.
32 Kovanen PT, Kaartinen M & Paavonen T (1995) Infil-
trates of activated mast cells at the site of coronary
atheromatous erosion or rupture in myocardial infarc-
tion. Circulation 92, 1084–1088.
33 Kovanen PT (1997) Chymase-containing mast cells in
human arterial intima: implications for atherosclerotic
disease. Heart Vessels Suppl. 125–127.

chymase. Biochem Biophys Res Commun 127, 443–449.
43 Schoenberger O, Sprows J, Schechter N, Cooperman B
& Rubin H (1989) Limited proteolysis of C1-inhibitor
Chymase-mediated proteolysis of TFPI T. Hamuro et al.
3076 FEBS Journal 274 (2007) 3065–3077 ª 2007 The Authors Journal compilation ª 2007 FEBS
by chymotrypsin-like proteinases. FEBS Lett 259, 165–
167.
44 Mizutani H, Schechter N, Lazarus G, Black RA &
Kupper TS (1991) Rapid and specific conversion of pre-
cursor interleukin 1 beta (IL-1 beta) to an active IL-1
species by human mast cell chymase. J Exp Med 174,
821–825.
45 Kinoshita A, Urata H, Bumpus F & Husain A (1991)
Multiple determinants for the high substrate specificity
of an angiotensin II-forming chymase from the human
heart. J Biol Chem 266, 19192–19197.
46 Saarinen J, Kalkkinen N, Welgus HG & Kovanen PT
(1994) Activation of human interstitial procollagenase
through direct cleavage of the Leu83-Thr84 bond by
mast cell chymase. J Biol Chem 269, 18134–18140.
47 Longley B, Tyrrell L, Ma Y, Williams D, Halaban R,
Langley K, Lu H & Schechter N (1997) Chymase clea-
vage of stem cell factor yields a bioactive, soluble pro-
duct. Proc Natl Acad Sci USA 94, 9017–9021.
48 Nakano A, Kishi F, Minami K, Wakabayashi H,
Nakaya Y & Kido H (1997) Selective conversion of big
endothelins to tracheal smooth muscle-constricting 31-
amino acid-length endothelins by chymase from human
mast cells. J Immunol 159, 1987–1992.
49 Kofford MW, Schwartz LB, Schechter NM, Yager DR,

tor-initiated coagulation reactions. J Biol Chem 265,
16786–16793.
56 Sayama S, Iozzo RV, Lazarus GS & Schechter NM
(1987) Human skin chymotrypsin-like proteinase chy-
mase. Subcellular localization to mast cell granules and
interaction with heparin and other glycosaminoglycans.
J Biol Chem 262, 6808–6815.
57 Murakami M, Karnik SS & Husain A (1995) Human
prochymase activation. A novel role for heparin in
zymogen processing. J Biol Chem 270, 2218–2223.
58 Valentin S & Schousboe I (1996) Factor Xa enhances
the binding of tissue factor pathway inhibitor to acidic
phospholipids. Thromb Haemost 75, 796–800.
59 Berrettini M, Parise P, Ricotta S, Iorio A, Peirone C
& Nenci GG (1996) Increased plasma levels of tissue
factor pathway inhibitor (TFPI) after n-3 polyunsatu-
rated fatty acids supplementation in patients with
chronic atherosclerotic disease. Thromb Haemost 75,
395–400.
60 Tijburg PN, Ryan J, Stern DM, Wollitzky B, Rimon S,
Rimon A, Handley D, Nawroth P, Sixma JJ & de
Groot PG (1991) Activation of the coagulation mechan-
ism on tumor necrosis factor-stimulated cultured
endothelial cells and their extracellular matrix. The role
of flow and factor IX ⁄ IXa.
J Biol Chem 266, 12067–
12074.
61 Kunori Y, Kawamura T & Takagi K (1997) Purification
of recombinant human chymase expressed in insect cells
and identification of unique cleavage site of self-diges-