Báo cáo khoa học: The leech product saratin is a potent inhibitor of platelet integrin a2b1 and von Willebrand factor binding to collagen pdf

The leech product saratin is a potent inhibitor of platelet
integrin a
2
b
1
and von Willebrand factor binding to collagen
Tara C. White
1
, Michelle A. Berny
1
, David K. Robinson
1
, Hang Yin
2
, William F. DeGrado
2,3
,
Stephen R. Hanson
1
and Owen J. T. McCarty
1,4
1 Department of Biomedical Engineering, Oregon Health & Science University, Portland, OR, USA
2 Department of Biochemistry and Biophysics, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
3 Department of Chemistry, University of Pennsylvania, Philadelphia, PA, USA
4 Cell and Developmental Biology, Oregon Health & Science University, Portland, OR, USA
Collagen plays a critical role in mediating the platelet
response to vessel injury in the dynamic environment
of the vasculature. Exposed collagen at sites of vascu-
lar injury initiates two platelet functions fundamental
to the process of primary hemostasis: initial recruit-
ment of circulating platelets, and triggering of the

platelet activation cascade required to stimulate throm-
bus growth [1,2]. The first step in platelet recruitment
to collagen occurs indirectly, via binding of platelet
glycoprotein (GP)Ib to collagen-bound von Willebrand
factor (VWF) [3]. VWF plays a critical role in the teth-
ering of platelets at high shear levels, due to the rapid
on-rate of binding between GPIb and VWF. The rapid
off-rate of GPIb–VWF interactions results in platelet
Keywords
a
2
b
1
; collagen; platelet; saratin;
von Willebrand factor
Correspondence
O. J. T. McCarty, Department of Biomedical
Engineering, Oregon Health & Science
University, 13B-CHH, 3303 SW Bond Ave,
Portland, OR 97239, USA
Fax: +1 503 418 9311
Tel: +1 503 418 9307
E-mail:
(Received 22 October 2006, revised 20
December 2006, accepted 11 January 2007)
doi:10.1111/j.1742-4658.2007.05689.x
Subendothelial collagen plays an important role, via both direct and indir-
ect mechanisms, in the initiation of thrombus formation at sites of vascular
injury. Collagen binds plasma von Willebrand factor, which mediates plate-
let recruitment to collagen under high shear. Subsequently, the direct bind-

ing of the platelet receptors glycoprotein VI and a
2
b
1
to collagen is critical
for platelet activation and stable adhesion. Leeches, have evolved a number
of inhibitors directed towards platelet–collagen interactions so as to prevent
hemostasis in the host during hematophagy. In this article, we describe the
molecular mechanisms underlying the ability of the leech product saratin
to inhibit platelet binding to collagen. In the presence of inhibitors of ADP
and thromboxane A
2
, both saratin and 6F1, a blocking a
2
b
1
mAb, abro-
gated platelet adhesion to fibrillar and soluble collagen. Additionally, sara-
tin eliminated a
2
b
1
-dependent platelet adhesion to soluble collagen in the
presence of an Src kinase inhibitor. Moreover, saratin prevented platelet-
rich plasma adhesion to fibrillar collagen, a process dependent upon both
a
2
b
1
and von Willebrand factor binding to collagen. Furthermore, saratin

specifically inhibited the binding of the a
2
integrin subunit I domain to col-
lagen, and prevented platelet adhesion to collagen under flow to the same
extent as observed in the presence of a combination of mAbs to glycopro-
tein Ib and a
2
b
1
. These results demonstrate that saratin interferes with inte-
grin a
2
b
1
binding to collagen in addition to inhibiting von Willebrand
factor–collagen binding, presumably by binding to an overlapping epitope
on collagen. This has significant implications for the use of saratin as a
tool to inhibit platelet–collagen interactions.
Abbreviations
a
2
I-bio, biotinylated a
2
integrin subunit I domain; DIC, differential interference contrast; FITC, fluorescein isothiocyanate; GP, glycoprotein;
PRP, platelet-rich plasma; TxA
2
, thromboxane A
2
; VWF, von Willebrand factor.
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1481

translocation at the site of injury, allowing adhesive
interactions with slower binding kinetics (such as the
platelet collagen receptors GPVI and a
2
b
1
and a
IIb
b
3
integrins) to mediate platelet adhesion and activation
[4]. Two routes have been proposed for this second
step of platelet adhesion, namely GPVI-mediated
platelet activation either preceding or following a
2
b
1
integrin-mediated platelet adhesion [5,6]. It is notewor-
thy that under static or low-shear conditions, the roles
of VWF and GPIb are dispensable, as the collagen
receptors GPVI and a
2
b
1
can mediate platelet adhesion
independently of VWF.
The evolution of a panoply of molecules to inter-
fere with the process of hemostasis has allowed the
leech to continue its alimentary habit of hemato-
phagy. The presence of anticoagulants in the salivary

glands of the leech, Hirudo medicinalis, was originally
discovered by Haycraft in 1884 and led to the isola-
tion of hirudin, a potent antithrombin anticoagulant
[7]. In addition to molecules that target the coagula-
tion cascade, a number of leech-derived substances
have been discovered that inhibit platelet adhesion
and activation. Three such molecules, LAPP (an
approximately 13 kDa leech antiplatelet protein isola-
ted from Haementeria officinalis) and calin and sara-
tin (approximately 65 kDa and 12 kDa proteins,
respectively, both isolated from H. medicinalis), have
been shown to specifically block platelet–collagen
interactions by inhibiting VWF binding to collagen
[8–11]. Depraetere et al. [12] then went on to demon-
strate that both LAPP and calin block the binding
site on collagen for the platelet integrin a
2
b
1
. The
saratin-binding site on collagen responsible for the
inhibition of VWF binding is presently unknown.
Saratin, which consists of 103 amino acids and con-
tains three disulfide bridges [13], has been cloned and
produced in recombinant form in Hansenula polymor-
pha. Barnes et al. were the first to demonstrate that
saratin specifically blocks purified VWF binding to col-
lagen, as well as potently inhibiting platelet aggregate
formation on immobilized collagen under shear flow
[8], therefore leading to the extensive use in the litera-

ture of saratin as a VWF–collagen inhibitor [14–18].
Furthermore, saratin has been shown to inhibit lumen
stenosis in carotid endarterectomized rats [19] and to
reduce platelet adhesion and intimal hyperplasia in
both a nondiseased environment [20] and in the state
of hyperhomocystinemia [21]. Moreover, Vilahur et al.
demonstrated that local administration of saratin
inhibited atherosclerotic plaque thrombogenicity under
shear conditions [22].
The main collagen-binding site on VWF resides
within the A3 domain (residues 923–1109) of VWF
[23–25]. Structural studies on the VWF A3 domain
showed that it assumes the same fold as the binding
site for collagen on the a
2
b
1
integrin, namely the
homologous integrin a
2
I domain [26]. The present
study demonstrates that saratin interferes with integrin
a
2
b
1
binding to collagen, in addition to inhibiting
VWF–collagen binding, presumably by binding to an
overlapping epitope on collagen. This has significant
implications for the use of saratin as a tool to inhibit

platelet–collagen interactions, and may provide the
basis for the therapeutic use of saratin as a potent
antithrombotic agent.
Results
Delayed collagen-induced aggregation
of platelets in the presence of saratin
We initially investigated the effects of the leech prod-
uct saratin on the ability of platelets to aggregate in
response to fibrillar collagen. Consistent with previous
findings [8], dose–response and maximal aggregation
of platelets did not differ in the presence of saratin
(data not shown). However, onset of aggregation
was significantly delayed in the presence of saratin
(Fig. 1A), and this lag time was particularly evident
at low fibrillar collagen concentrations (Fig. 1B).
Moreover, a similar delay in collagen-induced aggre-
gation was observed in the presence of the a
2
b
1
-
blocking antibody 6F1 (data not shown), consistent
with previous reports demonstrating an a
2
b
1
-depend-
ent lag phase for collagen-induced aggregation [29].
Together, these findings led us to question whether
saratin blocks platelet a

2
b
1
binding to collagen in
addition to functioning as an inhibitor of VWF–
collagen binding, as had been previously described by
Barnes et al. [8].
Dissection of the molecular actions of saratin
on fibrillar collagen
Experiments were designed to evaluate the ability of
saratin to inhibit platelet adhesion to collagen. We
gently pipetted purified human platelets onto surface-
immobilized fibrillar collagen, and recorded the degree
of adhesion and spreading using Normarski differential
interference contrast (DIC) microscopy. In agreement
with previous reports, human platelets undergo com-
plete spreading on fibrillar collagen in the absence of
external stimulation (Fig. 2A). The degree of platelet
adhesion to fibrillar collagen was only slightly reduced
by the presence of either an a
2
b
1
-blocking antibody or
saratin; however, these effects were statistically insigni-
Saratin blocks a
2
b
1
-collagen binding T. C. White et al.

1482 FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS
ficant (Fig. 2B). In comparison, a 40% reduction in
the degree of platelet adhesion to fibrillar collagen was
observed in the presence of apyrase and indomethacin,
inhibitors of the secondary mediators ADP and throm-
boxane A
2
(TxA
2
), respectively (Fig. 2B). Furthermore,
the integrin a
2
b
1
mediates this ADP ⁄ TxA
2
-independent
platelet adhesion to fibrillar collagen, as evidenced by
the abrogation of platelet binding in the presence of the
a
2
b
1
mAb 6F1 (Fig. 2A). A long the se lines, ADP ⁄ TxA
2
-
independent platelet adhesion to fibrillar collagen was
eliminated in the presence of saratin (Fig. 2A). Import-
antly, saratin was not required to be present in suspen-
sion to have an inhibitory effect, as pretreatment of

the collagen surface with saratin was sufficient to
achieve blockade of platelet adhesion in the presence
of apyrase ⁄ indomethacin (Table 1). In contrast, 6F1
needed to be present in the suspension to achieve
blockade (Table 1). However, it is noteworthy that the
inhibition of platelet adhesion observed in the presence
of 6F1 or saratin in the absence of secondary media-
tors could be overcome by treatment of platelet sus-
pensions with the G protein-coupled receptor agonist
thrombin (Fig. 2C,D).
Dissection of the molecular actions of saratin
on soluble collagen
We next aimed to examine platelet attachment to sol-
uble collagen, a process that has been reported to be
predominately mediated via a
2
b
1
integrins [31,32].
Indeed, the presence of the a
2
b
1
mAb 6F1 reduced
platelet adhesion on soluble collagen by over 60%
(Fig. 3A,B). Along these lines, a similar degree of inhi-
bition was observed in the presence of saratin (Fig. 3B).
We extended our studies to examine the effects of
secondary mediators on platelet adhesion to soluble
collagen. In parallel with our observations on fibrillar

collagen, a 50% reduction in platelet adhesion on
soluble collagen was observed in the presence of the
ADP scavenger apyrase and the cyclooxygenase inhib-
itor indomethacin (Fig. 3B). Moreover, ADP ⁄ TxA
2
-
independent platelet adhesion to soluble collagen was
eliminated through the blockade of a
2
b
1
with 6F1 or
treatment of collagen with saratin (Fig. 3A). As was
observed with fibrillar collagen, saratin did not need to
be present in suspension to have an inhibitory effect
on platelet adhesion (Table 1). However, in distinct
contrast to what was observed with fibrillar collagen,
both the a
2
b
1
mAb 6F1 and saratin blocked thrombin-
stimulated platelet adhesion to soluble collagen in the
presence, but not the absence, of inhibitors of secon-
dary mediators (Fig. 3C,D).
It is noteworthy that the presence of saratin did not
have any effect on platelet adhesion to immobilized
fibrinogen (Table 1), indicating that saratin does not
inhibit platelet integrin a
IIb

b
3
binding to fibrinogen.
Saratin blocks Src kinase-independent platelet
adhesion to soluble collagen
A set of experiments was designed to investigate the
role of Src family kinases in supporting platelet adhe-
sion and spreading on soluble collagen. As shown in
Fig. 4B, a 40% reduction in the degree of adhesion
was observed in the presence of the Src kinase inhib-
Fig. 1. The effect of saratin on the time of onset (lag phase) of
shape change in response to fibrillar collagen. (A) Heparinized
human PRP was stimulated with different concentrations of fibrillar
collagen in the absence or presence of saratin (10 lgÆmL
)1
). Chan-
ges in attenuance indicative of shape change and aggregation were
recorded. (B) The delay in onset of platelet shape change (lag time)
in the absence (black bars) and presence (white bars) of saratin is
expressed as time (seconds) between addition of collagen and ini-
tial increase in the attenuance of the platelet suspension. Values
are reported as follows: mean ± SEM from at least three experi-
ments. *P < 0.05,
d
P < 0.01, with respect to vehicle-treated
sample.
T. C. White et al. Saratin blocks a
2
b
1

-collagen binding
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1483
itor PP2, whereas platelets that bound to soluble colla-
gen in an Src kinase-independent manner were unable
to form lamellipodia. Furthermore, the presence of the
a
2
b
1
mAb 6F1 in combination with the Src kinase
inhibitor PP2 eliminated platelet adhesion to soluble
collagen altogether (Fig. 4A). Importantly, saratin was
capable of blocking this Src kinase-independent adhe-
sion to soluble collagen (Fig. 4A), consistent with the
ability of saratin to block a
2
b
1
-mediated platelet bind-
ing. It is noteworthy that this series of experiments
was performed in the absence of inhibitors of ADP
and TxA
2
.
Saratin blocks platelet-rich plasma adhesion
to collagen
Thus far, this study has utilized washed platelets in
order to examine the molecular mechanisms of saratin.
Physiologically, however, platelets are exposed to col-
lagen in the presence of plasma proteins. In order to

investigate the ability of saratin to inhibit receptor-
mediated interactions under physiologic conditions, we
layered platelet-rich plasma (PRP) over immobilized
collagen. Our studies demonstrated that individual
platelets in citrated PRP bound to immobilized soluble
collagen; however, interestingly, these platelets were
unable to form lamellipodia (Fig. 5A). Moreover, the
presence of either the a
2
b
1
mAb 6F1 or saratin abro-
gated this adhesion (Fig. 5A), further demonstrating
the ability of saratin to block a
2
b
1
-mediated platelet
binding. Equivalent results were observed in
PPACK ⁄ heparin-anticoagulated PRP, which preserves
the physiologic levels of divalent cations (36.2 ± 3.2
versus 0.74 ± 0.25 · 10
)2
platelets ⁄ mm
2
on soluble
collagen in the presence or absence of 10 lgÆmL
)1
sar-
atin, respectively; mean ± SEM; n ¼ 3).

In contrast to studies using washed platelets, where
we found individual platelets to be adherent to fibrillar
collagen, platelets in citrated PRP were incorporated
into a fibrous mesh along the collagen fibres (Fig. 5B).
The degree of platelet ⁄ fibrin deposition onto collagen
fibres was unaffected by the presence of 6F1 (Fig. 5B).
However, the presence of saratin eliminated the ability
of PRP to form a fibrous mesh, and significantly
reduced the degree of platelet adhesion to fibrillar colla-
gen. Importantly, we observed a similar level of reduc-
tion in platelet ⁄ fibrin deposition and platelet adhesion
when the a
2
b
1
mAb 6F1 was used in combination with
A
C
B
D
Fig. 2. The effect of saratin on platelet
adhesion on immobilized fibrillar collagen.
Human washed platelets (2 · 10
7
mL
)1
)
were placed on fibrillar collagen-coated cov-
erslips for 45 min at 37 °C, and imaged
using DIC microscopy. In selected experi-

ments, the function-blocking a
2
b
1
mAb 6F1
(10 lgÆmL
)1
) or saratin (10 lgÆmL
)1
) was
added to the platelet suspension, either in
the absence (A, B) or the presence (C, D) of
thrombin (1 UÆmL
)1
). Experiments were per-
formed in the absence (black bars) or pres-
ence (white bars) of the ADP-removing
enzyme apyrase (apy) and the cyclooxyge-
nase inhibitor indomethacin (indo) as indica-
ted. The numbers of adherent platelets
were recorded for five fields of view
(0.013 mm
2
) and expressed as mean ±
SEM from at least three experiments.
*P < 0.01 with respect to platelet adhesion
in the absence of apy ⁄ indo for each respec-
tive treatment;
d
P < 0.01 with respect to

platelet adhesion in the presence of
apy ⁄ indo and absence of 6D1 or saratin.
Saratin blocks a
2
b
1
-collagen binding T. C. White et al.
1484 FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS
antagonists to VWF receptors on platelets, namely the
mAb to GPIb, 6D1, and the mAb to a
IIb
b
3
, LJ-CP8.
Taken together, these data are reflective of the ability of
saratin to both block VWF–collagen binding and to
inhibit a
2
b
1
–collagen interactions.
Inhibition of a
2
integrin subunit I domain binding
to collagen by saratin
In an attempt to determine whether the binding site on
collagen for the platelet receptor a
2
b
1

is blocked by
the leech product saratin, we utilized a biotinylated
recombinant a
2
integrin subunit I domain construct.
Previous studies have shown that the a
2
integrin sub-
unit I domain binds to collagen type I in a dose-
dependent and saturable manner [12]. To investigate
the ability of saratin to inhibit a
2
b
1
binding, coverslips
were coated with fibrillar collagen type I and preincu-
bated with or without saratin. Subsequently, a con-
stant amount of biotinylated a
2
integrin subunit I
domain (a
2
I-bio) was added, and the amount of a
2
I-
bio was detected by adding streptavidin–fluorescein
isothiocyanate (FITC) and visualized using fluores-
cence microscopy. Our results demonstrated that sara-
tin was able to abrogate a
2

I-bio binding to collagen
(Fig. 6). In addition, saratin was able to completely
block VWF binding to immobilized collagen (data not
shown), consistent with previous reports [8,14]. Taken
together, our results definitively demonstrate that sara-
tin potently inhibits a
2
b
1
binding to collagen in addi-
tion to blocking VWF–collagen interactions.
Saratin reduces platelet adhesion to collagen
under flow conditions
We next aimed to examine the effects of saratin on
platelet adhesion in a more physiologically relevant
setting. We therefore investigated platelet recruitment
and aggregation as a result of the perfusion of whole
blood at 1000 s
)1
over immobilized fibrillar collagen.
As shown in Fig. 7, substantial platelet aggregates
form on collagen under flow, producing 39.6 ± 1.9
thrombi per field of view, resulting in 34.7 ± 6.2%
surface coverage (Table 2). Platelet adhesion was seve-
rely reduced in the presence of the GPIb mAb 6D1, as
evidenced by a dramatic reduction in surface coverage
(Fig. 7, Table 2). It is noteworthy that a number of
thrombi consisting of one to three platelets were
observed in the presence of 6D1 (Fig. 7), whereas the
number of these small thrombi was significantly

reduced in the presence of the a
2
b
1
mAb 6F1 in com-
bination with the GPIb mAb 6D1 (Table 2). Import-
antly, the presence of saratin reduced both the
percentage of surface coverage and the amount of
thrombi formed to a similar level as observed in the
presence of the GPIb and a
2
b
1
antagonists. Similar
results were observed in reconstituted blood (data not
shown). Altogether, our results demonstrate that sara-
tin, through blockade of both VWF and a
2
b
1
binding
to collagen, acts as a potent inhibitor of platelet aggre-
gation on collagen under shear flow conditions.
Discussion
Previous studies have demonstrated that the leech
product saratin functions as a potent inhibitor of
VWF binding to collagen [8,14]. In this study, we
extend these findings to demonstrate that saratin addi-
tionally functions as an inhibitor of platelet integrin
a

2
b
1
binding to collagen. This has important implica-
tions for the interpretation of results obtained when
saratin is used as an inhibitor of platelet–collagen
interactions, both in vitro [14–18] and in vivo [19–22].
The current study, in accordance with others
[5,31,32], demonstrates that a
2
b
1
integrins are not
Table 1. Effects of an a
2
b
1
blocker and saratin on platelet adhesion
on collagen. Purified human platelets (2 · 10
7
mL
)1
), in the pres-
ence of apyrase (2 U mL
)1
) and indomethacin (10 lM), were placed
on BSA, fibrillar or soluble collagen or fibrinogen-coated coverslips
for 45 min at 37 °C. In designated experiments, immobilized colla-
gen or fibrinogen was treated with the a
2

b
1
-blocking mAb 6F1
(10 lgÆmL
)1
) or saratin (10 lgÆmL
)1
) for 10 min, followed by wash-
ing with NaCl ⁄ P
i
, prior to exposure to platelets (surface treatment).
In selected experiments, 6F1 (10 lgÆmL
)1
) or saratin (10 lgÆmL
)1
)
was added to and maintained in the suspension with the platelets
throughout the adhesion assay (suspension treatment). Values are
reported as follows: adherent platelets, mean ± SEM of three to
six experiments; platelet surface area, mean ± SEM of 50–300
cells.
Surface
Surface
treatment
Suspension
treatment
Platelet adhesion
(cells ⁄ mm
2
· 10

)2
)
BSA – – 2.2 ± 0.62*
,
**
Fibrillar collagen – – 43.7 ± 2.57
Fibrillar collagen 6F1 – 44.2 ± 1.98
Fibrillar collagen – 6F1 8.1 ± 0.25*
Fibrillar collagen Saratin – 7.9 ± 1.45*
Fibrillar collagen – Saratin 8.2 ± 0.80*
Soluble collagen – – 30.1 ± 1.50
Soluble collagen 6F1 – 31.6 ± 1.21
Soluble collagen – 6F1 2.9 ± 0.74**
Soluble collagen Saratin – 2.7 ± 0.59**
Soluble collagen – Saratin 2.8 ± 0.80**
Fibrinogen – – 63.2 ± 1.03
Fibrinogen Saratin – 62.8 ± 1.88
Fibrinogen – Saratin 64.4 ± 1.60
*
,
**P < 0.01 with respect to platelet adhesion ⁄ surface area on
untreated fibrillar or soluble collagen, respectively.
T. C. White et al. Saratin blocks a
2
b
1
-collagen binding
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1485
essential for platelet, whether purified or in plasma,
adhesion to fibrillar collagen, as GPVI is capable of

triggering platelet activation and release of secondary
mediators (ADP and TxA
2
), which leads to platelet
adhesion independently of a
2
b
1
. However, in the
absence of the actions of secondary mediators,
GPVI-mediated activation alone is insufficient to
induce platelet adhesion to fibrillar collagen in the
absence of a
2
b
1
, consistent with the current paradigm
[5,33,34].
A different picture emerges for platelet adhesion to
soluble collagen. This form of collagen results from
the cleavage of collagen in the nontriple helical region,
where covalent cross-links are found that are required
for the assembly of collagen molecules into the typical
banded structure found in fibrillar collagen. Soluble
collagen therefore lacks the highly repetitive GPVI
recognition sites characteristic of fibrillar collagen,
therefore providing a means of reducing but not ablat-
ing GPVI signaling [1,35]. Consistent with previous
reports [32–34,36,37], our data demonstrate that a
2

b
1
integrins play an important role in mediating platelet
adhesion to soluble collagen in the absence of inhibi-
tors of secondary mediators, whereas a
2
b
1
is essential
for ADP ⁄ TxA
2
-independent platelet adhesion. Addi-
tionally, we demonstrate that PRP binding to immobi-
lized soluble collagen is a
2
b
1
-dependent. Moreover, the
a
2
b
1
dependency of Src kinase-independent adhesion
on soluble collagen further indicates the essential role
of a
2
b
1
in the absence of platelet activation. Interest-
ingly, we found that thrombin stimulation, which pre-

dominantly acts via the G
q
family of proteins [38],
potentiated a
2
b
1
-independent platelet adhesion to sol-
uble collagen only in the presence of the actions of the
G
i
protein-coupled agonist ADP. This supports the
notion that GPVI-mediated platelet activation and
adhesion on the low-GPVI-affinity soluble collagen is
dependent upon a cosignal from G
i
-coupled receptors
in the absence of a
2
b
1
[37,39,40].
The ability of saratin to precisely mirror the effects of
6F1 in the aforementioned experiments, in combination
with the fact that saratin abrogates the binding of the a
2
integrin subunit I domain to collagen, provides unequi-
vocal evidence that this leech product is a potent a
2
b

1
blocker. As the binding sites for VWF and a
2
b
1
on colla-
gen are within close spatial proximity [26], saratin pre-
sumably binds to an overlapping epitope on collagen to
achieve dual blockade of these interactions. Therefore,
A
C
DB
Fig. 3. The effect of saratin on platelet
adhesion on immobilized soluble collagen.
Human washed platelets (2 · 10
7
mL
)1
)
were placed on soluble collagen-coated
coverslips for 45 min at 37 °C, and imaged
using DIC microscopy. In selected experi-
ments, the function-blocking the a
2
b
1
mAb
6F1 (10 lgÆmL
)1
) or saratin (10 lgÆmL

)1
)
was added to the platelet suspension either
in the absence (A, B) or in the presence (C,
D) of thrombin (1 UÆmL
)1
). Experiments
were performed in the absence (black bars)
or the presence (white bars) of the ADP-
removing enzyme apyrase (apy) and the cy-
clooxygenase inhibitor indomethacin (indo)
as indicated. The numbers of adherent
platelets were recorded for five fields of
view (0.013 mm
2
) and expressed as
mean ± SEM from at least three experi-
ments. *P < 0.01 with respect to platelet
adhesion in the absence of apy ⁄ indo for
each respective treatment; **,dP < 0.01
with respect to platelet adhesion in the
absence of receptor inhibitors and absence
or presence of apy ⁄ indo, respectively.
Saratin blocks a
2
b
1
-collagen binding T. C. White et al.
1486 FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS
in light of the role that a

2
b
1
plays in stabilizing collagen-
bound platelets under shear [5,6], the combined ability
of saratin to block both VWF-dependent and VWF-
independent (via a
2
b
1
) pathways of platelet deposition
on collagen makes this leech product a powerful anti-
thrombotic agent.
Experimental procedures
Reagents
Fibrillar type I collagen (Horm) from equine tendon was
purchased from Nycomed (Munich, Germany). Soluble,
nonfibrillar type I collagen from rat tail was purchased
from Sigma (St Loius, MO, USA). LJ-CP8 was generously
provided by Z. M. Ruggeri (Scripps Research Institute,
La Jolla, CA, USA). 6F1 and 6D1 were a kind gift from
A
B
Fig. 4. Saratin blocks Src kinase-independent platelet adhesion
on immobilized soluble collagen. Human washed platelets
(2 · 10
7
mL
)1
) were placed on soluble collagen-coated coverslips for

45 min at 37 °C and (A) imaged using DIC microscopy. In selected
experiments, the function-blocking a
2
b
1
mAb 6F1 (10 lgÆmL
)1
)or
saratin (10 lgÆmL
)1
) was added to the platelet suspension either in
the absence or presence of the Src kinase inhibitor PP2 (20 l
M).
(B) The numbers of adherent platelets were recorded for five fields
of view (0.013 mm
2
) and expressed as mean ± SEM from at least
three experiments. *, dP < 0.01 with respect to platelet adhesion in
the absence or presence of PP2 for each respective inhibitor.
AB
Fig. 5. Saratin inhibits PRP adhesion on collagen. Human platelets
in PRP were layered onto either a soluble (A) or fibrillar (B) collagen-
coated slide for 45 min at 37 °C and imaged using DIC microscopy. In
selected experiments, function-blocking a
2
b
1
mAb 6F1 (10 lgÆmL
)1
),

GPIb mAb 6D1 (10 lgÆmL
)1
), a
IIb
b
3
mAb LJ-CP8 (CP8; 100 lgÆmL
)1
)
or saratin (10 lgÆmL
)1
) was added to PRP. Images are representative
of at least three experiments.
T. C. White et al. Saratin blocks a
2
b
1
-collagen binding
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1487
B. Coller (Rockefeller University, New York, NY, USA).
The Src kinase inhibitor PP2 was purchased from Calbio-
chem (San Diego, CA, USA). Recombinant saratin, pro-
duced in the yeast Han. polymorpha as previously described
[8], was supplied by BioVascular, Inc. (La Jolla, CA, USA).
Other reagents were obtained from Sigma or previously
named sources [27,28].
Preparation of washed platelets
Human venous blood was drawn by venipuncture from
healthy volunteers into sodium citrate and acid ⁄ citrate ⁄
dextrose as previously described [3]. PRP was prepared by

centrifugation of whole blood at 200 g for 20 min (5702 R
centrifuge, Eppendorf, Hamburg, Germany, rotor F-35-30-
17). The platelets were then isolated from PRP by centrifu-
gation at 1000 g for 10 min (5702 R centrifuge, Eppendorf,
Hamburg, Germany, rotor F-35-30-17) in the presence of
prostacyclin (0.1 lgÆmL
)1
). The pellet was resuspended in
modified Hepes ⁄ Tyrodes buffer (129 mm NaCl, 0.34 mm
Na
2
HPO
4
,2.9mm KCl, 12 mm NaHCO
3
,20mm Hepes,
5mm glucose, 1 mm MgCl
2
, pH 7.3) containing
0.1 lgÆmL
)1
prostacyclin, washed, and resuspended
(2 · 10
7
mL
)1
) in Hepes ⁄ Tyrode buffer.
In selected experiments, platelet suspensions were treated
with 10 lgÆmL
)1

6F1, 100 lgÆmL
)1
LJ-CP8, 10 lgÆmL
)1
6D1, 10 lgÆmL
)1
saratin, 1 UÆmL
)1
thrombin, 20 lm PP2,
and ⁄ or 2 UÆmL
)1
apyrase and 10 lm indomethacin for
10 min before use in the assays. It is noteworthy that this
dose of saratin is well above the IC50 reported for platelet–
collagen binding [8]. All experiments were performed in the
absence of exogenously added Ca
2+
.
Platelet adhesion assays
Glass coverslips were incubated with a suspension of
fibrillar collagen (100 lgÆmL
)1
) or soluble collagen
(50 lgÆmL
)1
) overnight at 4 °C. Surfaces were then
blocked with denatured BSA (5 mgÆmL
)1
) for 1 h at
room temperature, and this was followed by subsequent

washing with NaCl ⁄ P
i
before use in spreading assays. In
selected experiments, collagen-coated surfaces were treated
for 10 min with saratin (10 lgÆmL
)1
), and this was fol-
lowed by washing with NaCl ⁄ P
i
. Quiescent platelets failed
to bind or spread on surfaces coated with denatured
BSA (Table 1).
For spreading experiments, washed platelets (2 ·
10
7
mL
)1
) were incubated on collagen-coated coverslips at
37 °C for 45 min. Subsequently, coverslips were gently
washed with Hepes ⁄ Tyrode buffer to remove unbound cells.
Fig. 6. Inhibition of a
2
I-bio binding on collagen by saratin. Coverslips
coated with fibrillar soluble collagen were preincubated with either
vehicle or saratin (10 lgÆmL
)1
) for 10 min. A constant amount of
0.3 l
M a
2

I-bio was added, and bound a
2
I-bio was detected by add-
ing streptavidin–FITC and visualized using DIC and fluorescence
microscopy. Images are representative of three experiments.
Fig. 7. Saratin inhibits platelet adhesion on immobilized collagen
under flow. Anticoagulated human whole blood was perfused over
a fibrillar collagen coverslip at a shear rate of 1000 s
)1
for 4 min. In
selected experiments, blood was pretreated for 10 min with 6D1
(10 lgÆmL
)1
) with or without 6F1 (10 lgÆmL
)1
). In separate experi-
ments, collagen-coated coverslips were pretreated with saratin
(10 lgÆmL
)1
) for 10 min, whereas saratin (30 lgÆmL
)1
) was main-
tained in whole blood during flow. Images are representative of at
least three experiments.
Saratin blocks a
2
b
1
-collagen binding T. C. White et al.
1488 FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS

Platelet spreading was imaged using Ko
¨
hler illuminated
Nomarski DIC optics with a Zeiss 63· oil immersion
1.40 NA plan-apochromat lens on a Zeiss Axiovert 200M
microscope (Carl Zeiss, Thornwood, NY, USA), and recor-
ded using stallion 4.0 (Intelligent Imaging Innovations,
Inc., Denver, CO, USA). To compute the degree of adhe-
sion and surface area of spreading platelets, images were
manually outlined and quantified by determining the num-
ber of pixels within each outline using a Java plug-in for
image j software, as previously described [28]. Imaging a
graticule under the same conditions allowed the conversion
of pixel size to micrometers.
Flow adhesion studies
Glass coverslips were coated with fibrillar collagen as des-
cribed above. Coverslips were assembled onto a flow cham-
ber (Glyotech, Gaithersburg, MD, USA) and mounted on
the stage of an inverted microscope (Zeiss Axiovert 200M).
In selected experiments, coverslips were treated with
10 lgÆmL
)1
saratin for 10 min prior to the flow assay.
PPACK (40 lm) anticoagulated whole blood was perfused
through the chamber for 3 min at a wall shear rate of
1000 s
)1
, and this was followed by washing for 4 min at the
same shear rate with modified Tyrodes buffer and imaged
using DIC microscopy.

Measurement of platelet aggregation
To prepare heparinized PRP, blood was collected from
healthy human donors into syringes containing heparin
sodium (10 UÆmL
)1
final concentration). PRP was obtained
by centrifugation of heparinized blood at 200 g for 15 min
(5702 R centrifuge, Eppendorf, rotor F-35-30-17). Optical
aggregation studies were carried out using a Born aggreg-
ometer (Chronolog, Havertown, PA, USA) with high-speed
stirring (1200 r.p.m.) at 37 °C. Platelet shape change and
aggregation were monitored by measuring changes in light
transmission as previously described [29].
Binding competition assays
The recombinant a
2
I domain-encoding region was gener-
ated, purified and biotinylated as previously described
[30]. Purified material was characterized by SDS ⁄ PAGE,
and the concentration of a
2
I-bio was quantified using a
detergent compatible-protein assay (Biorad, Hercules, CA,
USA).
Coverslips were coated overnight at 4 °C with 1 mgÆmL
)1
fibrillar collagen. Wells were then blocked with denatured
BSA (5 mgÆmL
)1
) for 1 h at room temperature, and this

was followed by subsequent washing with NaCl ⁄ P
i
before
incubation with vehicle or saratin (10 lgÆmL
)1
) for 10 min.
A constant amount of a
2
I-bio (0.3 l m) or VWF
(10 lgÆmL
)1
) was then added and allowed to bind for
90 min. Following copious washing, bound a
2
I-bio or
VWF was detected by adding streptavidin–FITC or anti-
VWF–FITC, respectively, for 1 h at RT, and visualized
using fluorescence microscopy.
Analysis of data
Experiments were carried out at least three times, and ima-
ges shown are representative data from one experiment.
Where applicable, results are shown as mean ± SEM. The
statistical significance of differences between means was
determined by ANOVA. If means were shown to be signifi-
cantly different, multiple comparisons were preformed by
the Tukey test. Probability values of P < 0.01 were consid-
ered to be statistically significant.
Acknowledgements
We would like to thank Steve P. Watson and Andras
Gruber for stimulating discussions, and Dr Barry Col-

ler for the generous gifts of 6F1 and 6D1. Tara C.
White and Michelle A. Berny are ARCS scholars,
and David K. Robinson is the recipient of a Johnson
scholarship. Owen J. T. McCarty is supported by an
American Heart Association Beginning Grant-in-Aid
(0665512Z).
References
1 Nieswandt B & Watson SP (2003) Platelet–collagen
interaction: is GPVI the central receptor? Blood 102,
449–461.
2 Ruggeri ZM (2006) Platelet interactions with vessel wall
components during thrombogenesis. Blood Cells Mol
Dis 36, 145–147.
Table 2. Effects of saratin on platelet adhesion ⁄ aggregation on col-
lagen under flow. Human whole blood was perfused over immobi-
lized collagen at 1000 s
)1
. Blood was pretreated for 10 min with
the GPIb mAb 6D1 (10 lgÆmL
)1
) with or without the a
2
b
1
mAb 6F1
(10 lgÆmL
)1
). In separate experiments, collagen-coated coverslips
were pretreated with saratin (10 lgÆmL
)1

) for 10 min, whereas sar-
atin (30 lgÆmL
)1
) was maintained in whole blood during flow. Val-
ues are reported as mean ± SEM of three experiments.
Treatment
Surface coverage
(%)
Number of thrombi ⁄
field of view
– 34.7 ± 6.18 39.6 ± 1.95
6D1 6.9 ± 2.77* 36.4 ± 5.85
6D1 + 6F1 3.8 ± 0.96* 13.0 ± 4.19**
Saratin 5.7 ± 2.43* 15.4 ± 5.27**
*
,
**P < 0.01 with respect to untreated and 6D1-treated blood,
respectively.
T. C. White et al. Saratin blocks a
2
b
1
-collagen binding
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1489
3 McCarty OJ, Calaminus SDJ, Berndt MC, Machesky
LM & Watson SP (2006) von Willebrand factor med-
iates platelet spreading through glycoprotein Ib and
a
IIb
b

3
in the presence of botrocetin and ristocetin,
respectively. J Thromb Haemost 4, 1367–1378.
4 Watson SP, Auger JM, McCarty OJ & Pearce AC
(2005) GPVI and integrin alphaIIb beta3 signaling in
platelets. J Thromb Haemost 3, 1752–1762.
5 Auger JM, Kuijpers MJ, Senis YA, Watson SP &
Heemskerk JW (2005) Adhesion of human and mouse
platelets to collagen under shear: a unifying model.
FASEB J 19, 825–827.
6 Sarratt KL, Chen H, Zutter MM, Santoro SA, Hammer
DA & Kahn ML (2005) GPVI and alpha2beta1 play
independent critical roles during platelet adhesion and
aggregate formation to collagen under flow. Blood 106,
1268–1277.
7 Brewer DB (2006) Max Schultze (1865), G. Bizzozero
(1882) and the discovery of the platelet. Br J Haematol
133, 251–258.
8 Barnes CS, Krafft B, Frech M, Hofmann UR, Pap-
endieck A, Dahlems U, Gellissen G & Hoylaerts MF
(2001) Production and characterization of saratin, an
inhibitor of von Willebrand factor-dependent platelet
adhesion to collagen. Semin Thromb Hemost 27, 337–
348.
9 Connolly TM, Jacobs JW & Condra C (1992) An inhi-
bitor of collagen-stimulated platelet activation from the
salivary glands of the Haementeria officinalis leech. I.
Identification, isolation, and characterization. J Biol
Chem 267, 6893–6898.
10 Keller PM, Schultz LD, Condra C, Karczewski J &

Connolly TM (1992) An inhibitor of collagen-stimulated
platelet activation from the salivary glands of the Hae-
menteria officinalis leech. II. Cloning of the cDNA and
expression. J Biol Chem 267, 6899–6904.
11 Munro R, Jones CP & Sawyer RT (1991) Calin ) a
platelet adhesion inhibitor from the saliva of the medici-
nal leech. Blood Coagul Fibrinolysis 2, 179–184.
12 Depraetere H, Kerekes A & Deckmyn H (1999) The
collagen-binding leech products rLAPP and calin pre-
vent both von Willebrand factor and alpha2beta1
(GPIa ⁄ IIa)-I-domain binding to collagen in a different
manner. Thromb Haemost 82, 1160–1163.
13 Maurer T, Bomke J, Frech M, Rysiok T & Kalbitzer
HB (2001) Sequential assignment and secondary struc-
ture of saratin, an inhibitor of von Willebrand factor-
dependent platelet adhesion to collagen. J Biomol NMR
21, 77–78.
14 Kuijpers MJ, Schulte V, Oury C, Lindhout T, Broers J,
Hoylaerts MF, Nieswandt B & Heemskerk JW (2004)
Facilitating roles of murine platelet glycoprotein Ib and
alphaIIbbeta3 in phosphatidylserine exposure during
vWF-collagen-induced thrombus formation. J Physiol
558, 403–415.
15 Melis E, Bonnefoy A, Daenens K, Yamamoto H, Ver-
mylen J & Hoylaerts MF (2004) alphaIIbbeta3 antagon-
ism versus antiadhesive treatment to prevent platelet
interactions with vascular subendothelium. J Thromb
Haemost 2, 993–1002.
16 Oury C, Kuijpers MJ, Toth-Zsamboki E, Bonnefoy A,
Danloy S, Vreys I, Feijge MA, De Vos R, Vermylen J,

Heemskerk JW et al. (2003) Overexpression of the plate-
let P2X1 ion channel in transgenic mice generates a
novel prothrombotic phenotype. Blood 101, 3969–3976.
17 Siljander PR, Munnix IC, Smethurst PA, Deckmyn H,
Lindhout T, Ouwehand WH, Farndale RW & Heems-
kerk JW (2004) Platelet receptor interplay regulates col-
lagen-induced thrombus formation in flowing human
blood. Blood 103, 1333–1341.
18 Bonnefoy A, Romijn RA, Vandervoort PA, Van
Rompaey I, Vermylen J & Hoylaerts MF (2006) von
Willebrand factor A1 domain can adequately substitute
for A3 domain in recruitment of flowing platelets to col-
lagen. J Thromb Haemost 4, 2151–2161.
19 Cruz CP, Eidt J, Drouilhet J, Brown AT, Wang Y, Bar-
nes CS & Moursi MM (2001) Saratin, an inhibitor of
von Willebrand factor-dependent platelet adhesion,
decreases platelet aggregation and intimal hyperplasia in
a rat carotid endarterectomy model. J Vasc Surg 34,
724–729.
20 Smith TP, Alshafie TA, Cruz CP, Fan CY, Brown AT,
Wang Y, Eidt JF & Moursi MM (2003) Saratin, an
inhibitor of collagen–platelet interaction, decreases
venous anastomotic intimal hyperplasia in a canine
dialysis access model. Vasc Endovascular Surg 37, 259–
269.
21 Davis JA, Brown AT, Alshafie T, Poirier LA, Cruz CP,
Wang Y, Eidt JF & Moursi MM (2004) Saratin (an
inhibitor of platelet–collagen interaction) decreases
platelet aggregation and homocysteine-mediated post-
carotid endarterectomy intimal hyperplasia in a dose-

dependent manner. Am J Surg 188, 778–785.
22 Vilahur G, Duran X, Juan-Babot O, Casani L & Badi-
mon L (2004) Antithrombotic effects of saratin on
human atherosclerotic plaques. Thromb Haemost 92 ,
191–200.
23 Cruz MA, Yuan H, Lee JR, Wise RJ & Handin RI
(1995) Interaction of the von Willebrand factor (vWF)
with collagen. Localization of the primary collagen-
binding site by analysis of recombinant vWF a domain
polypeptides. J Biol Chem 270, 10822–10827.
24 Lankhof H, van Hoeij M, Schiphorst ME, Bracke M,
Wu YP, Ijsseldijk MJ, Vink T, de Groot PG & Sixma
JJ (1996) A3 domain is essential for interaction of von
Willebrand factor with collagen type III. Thromb Hae-
most 75, 950–958.
25 Staelens S, Hadders MA, Vauterin S, Platteau C,
De Maeyer M, Vanhoorelbeke K, Huizinga EG &
Deckmyn H (2006) Paratope determination of the
Saratin blocks a
2
b
1
-collagen binding T. C. White et al.
1490 FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS
antithrombotic antibody 82D6A3 based on the crystal
structure of its complex with the von Willebrand factor
A3-domain. J Biol Chem 281, 2225–2231.
26 Huizinga EG, Martijn van der Plas R, Kroon J, Sixma
JJ, Gros P (1997). Crystal structure of the A3 domain
of human von Willebrand factor: implications for colla-

gen binding. Structure 5, 1147–1156.
27 McCarty OJ, Larson MK, Auger JM, Kalia N, Atkin-
son BT, Pearce AC, Ruf S, Henderson RH, Tybulewicz
VL, Machesky LM et al. (2005) Rac1 is essential for
platelet lamellipodia formation and aggregate stability
under flow. J Biol Chem 280, 39474–39484.
28 McCarty OJ, Zhao Y, Andrew N, Machesky LM,
Staunton D, Frampton J & Watson SP (2004) Evalua-
tion of the role of platelet integrins in fibronectin-depen-
dent spreading and adhesion. J Thromb Haemost 2,
1823–1833.
29 Jarvis GE, Atkinson BT, Snell DC & Watson SP (2002)
Distinct roles of GPVI and integrin [alpha]2[beta]1 in
platelet shape change and aggregation induced by differ-
ent collagens. Br J Pharmacol 137, 107–117.
30 Yin H, Gerlach LO, Miller MW, Moore DT, Liu D,
Vilaire G, Bennett JS & Degrado WF (2006) Arylamide
derivatives as allosteric inhibitors of the integrin
alpha(2)beta(1) ⁄ type I collagen interaction. Bioorg Med
Chem Lett 16 , 3380–3382.
31 Holtkotter O, Nieswandt B, Smyth N, Muller W, Haf-
ner M, Schulte V, Krieg T & Eckes B (2002) Integrin
alpha 2-deficient mice develop normally, are fertile, but
display partially defective platelet interaction with col-
lagen. J Biol Chem 277, 10789–10794.
32 Nieswandt B, Brakebusch C, Bergmeier W, Schulte V,
Bouvard D, Mokhtari-Nejad R, Lindhout T, Heems-
kerk JW, Zirngibl H & Fassler R (2001) Glycoprotein
VI but not alpha2beta1 integrin is essential for platelet
interaction with collagen. EMBO J 20, 2120–2130.

33 Inoue O, Suzuki-Inoue K, Dean WL, Frampton J &
Watson SP (2003) Integrin alpha2beta1 mediates out-
side-in regulation of platelet spreading on collagen
through activation of Src kinases and PLCgamma2.
J Cell Biol 160, 769–780.
34 Inoue O, Suzuki-Inoue K, McCarty OJ, Moroi M, Rug-
geri ZM, Kunicki TJ, Ozaki Y & Watson SP (2006)
Laminin stimulates spreading of platelets through integ-
rin alpha6beta1-dependent activation of GPVI. Blood
107, 1405–1412.
35 Farndale RW (2006) Collagen-induced platelet activa-
tion. Blood Cells Mol Dis 36, 162–165.
36 Atkinson BT, Jarvis GE & Watson SP (2003) Activa-
tion of GPVI by collagen is regulated by alpha2beta1
and secondary mediators. J Thromb Haemost 1,
1278–1287.
37 Nieswandt B, Bergmeier W, Eckly A, Schulte V, Ohl-
mann P, Cazenave JP, Zirngibl H, Offermanns S &
Gachet C (2001) Evidence for cross-talk between glyco-
protein VI and Gi-coupled receptors during collagen-
induced platelet aggregation. Blood 97, 3829–3835.
38 Brass LF (2003) Thrombin and platelet activation.
Chest 124 (3 Suppl.), 18S–25S.
39 Chen H & Kahn ML (2003) Reciprocal signaling by
integrin and nonintegrin receptors during collagen acti-
vation of platelets. Mol Cell Biol 23, 4764–4777.
40 Thornber K, McCarty OJ, Watson SP & Pears CJ
(2006) Distinct but critical roles for integrin alphabeta
in platelet lamellipodia formation on fibrinogen, col-
lagen-related peptide and thrombin. FEBS J 273,

5032–5043.
T. C. White et al. Saratin blocks a
2
b
1
-collagen binding
FEBS Journal 274 (2007) 1481–1491 ª 2007 The Authors Journal compilation ª 2007 FEBS 1491