Life Sciences 76 (2005) 2607–2619

www.elsevier.com/locate/lifescie

A novel anticoagulant purified from fish protein hydrolysate

inhibits factor XIIa and platelet aggregation

Niranjan Rajapakse, Won-Kyo Jung, Eresha Mendis,

Sung-Hoon Moon, Se-Kwon Kim*

Department of Chemistry, Pukyong National University, Daeyon-3-Dong, Nam-Gu, Busan 608-737, Korea

Received 18 October 2004; accepted 10 December 2004

Abstract

A novel fish protein having anticoagulant and antiplatelet properties was enzymatically extracted from the

marine fish, yellowfin sole (Limanda aspera) and purified to homogeneity producing an overall purification fold of

206.6. MALDI-TOF mass spectroscopic and SDS-PAGE analysis identified the purified protein as 12.01 kDa

single-chain monomeric protein. It inhibited the activated coagulation factor XII (FXIIa) by forming an inactive

complex regardless of Zn2+ mediation, and was named, yellowfin sole anticoagulant protein (YAP). In addition,

YAP act to antagonize platelet membrane glycoprotein integrin, to arrest platelet aggregation. However, YAP was

not able to block the adhesion of platelets to collagen, which mediate via major collagen receptors, GPIa/IIa on

platelet membrane. Furthermore, YAP did not possess plasminogen activator-like activity to activate fibrinolysis. In

fact, our findings indicate that YAP binds with FXIIa and platelet membrane integrins to inhibit thrombosis in vitro.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Anticoagulant; Platelet aggregation; Protein hydrolysate; Activated factor XII (FXIIa); Platelet integrin

Introduction

Autoactivation of factor XII (FXII or Hageman factor) into its active form (FXIIa) in contact with

negatively charged surfaces triggers intrinsic pathway of coagulation along with high molecular weight

0024-3205/$ -

doi:10.1016/j.l

* Correspond

E-mail add

see front matter D 2005 Elsevier Inc. All rights reserved.

fs.2004.12.010

ing author. Tel.: +82 51 620 6375; fax: +82 51 628 8147.

ress: sknkim@mail.pknu.ac.kr (S.-K. Kim).

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192608

kininogen (HMWK) and prekallikrein (Miller et al., 1980). Some evidences suggest that Zn2+ binding to

FXII induces a conformational change that makes the protein more susceptible for the development of

enzymatic activity during autoactivation (Schousboe, 1993). Therefore, the simple fact that a deficiency

of FXII prolongs activated partial thromboplastin time (APTT) in vitro has been clearly recognized to be

important in coagulation. At the same time, platelets also play an important role in coagulation by

aggregating and adhering to the underline collagen via platelet membrane glycoprotein integrins

(Clemetson and Clemetson, 2001).

The importance of anticoagulants and antiplatelets as treatments for the prevention of ischaemic

events in patients with cardiovascular diseases is becoming increasingly recognized. Other than the

natural anticoagulants found in the coagulation cascade, a number of anticoagulant and antiplatelet

protein compounds have also been identified from diverse natural sources (Urata et al., 2003; Tanaka-

Azevedo et al., 2003). Natural anticoagulants found in above biological systems have been evolved

under genetic control with particular structural sequences to regulate coagulation events. Despite the

traditional means of identifying already existing anticoagulant compounds from these sources, only a

few efforts have been laid to develop novel anticoagulants from natural bioresources. In this regard,

proteolysis can be applied as a potential technique to identify such molecules from sequences of protein

sources. Therefore, in this study we attempted to identify anticoagulant compounds from protein

sequences of fish muscle following enzymatic proteolysis. We report that, the new anticoagulant protein

inhibits coagulation in vitro by forming an inactive complex with FXIIa in the contact system.

Furthermore, we discuss the significance of the anticoagulant protein to act as a platelet aggregation

inhibitor.

Materials and methods

Fish frame and reagents

Yellowfin sole (Limanda aspera) frame was obtained from Daerim Co. (Busan, South Korea). DEAE-

Sephadex A-25, Sephadex G-25, bovine thrombin, fibrinogen, plasmin, p-nitrophenyl phosphate, pepsin,

papain and a-chymotrypsin were purchased from Sigma Chemical Co. (St. Louis, MO). AlcalaseR and

NeutraseR were supplied by Novo Nordisk Co. (Denmark). All PAGE reagents were also purchased

from Sigma Chemical Co. (St. Louis, MO). Coagulation and specific factor assay reagents were products

of Instrumentation Laboratory Co. (Lexington, MA). FXIIa was purchased from ICN Pharmaceuticals

Inc. (Aurora, Ohio, USA). The other reagents were of the highest grade commercially available.

Extraction of fish protein

Yellowfin sole frame containing considerable amount of protein after filleting process was subjected

to hydrolyze separately with seven different proteases, AlcalaseR, NeutraseR, pepsin, papain, a-

chymotrypsin, trypsin and previously extracted tuna pyloric caeca crude enzyme (TPCCE) (Kim et al.,

1997). Hydrolysis was performed for different time intervals and the optimum enzymic conditions were

adjusted as required. After the extraction, hydrolysate was analyzed for the degree of hydrolysis and

expressed as percentage protein content measured by the method of Lowry et al. (1951). Finally, the

hydrolysate was centrifuged at 12000 � g for 10 min at 48C and the supernatant was lyophilized.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2609

Blood collection and preparation of plasma

Nine parts of blood was drawn by venipuncture into one part of 3.2% sodium citrate from three

healthy volunteers and pooled. The blood was separately centrifuged at 1600 � g for 10 min and 150 �g for 10 min to obtain platelet poor plasma (PPP) and platelet rich plasma (PRP) respectively. Separated

plasma was utilized for in vitro testing within 24 h.

Plasma-based coagulation assays

All coagulation assays were performed with three individual replicates using an ACLRCoagulation Analyzer (ACLR 7000, Instrumentation Laboratory Co., Lexington, MA) and mean

values were taken. Normal citrated PPP (80 Al) was incubated with the anticoagulant compound (20

Al) for 3 min at 378C. For activated partial thromboplastin time (APTT) assay, 100 Al of APTT

reagent was added to the above mixture and incubated at 378C for 3 min and clotting time was

recorded following the addition of 100 Al of 20 mM CaCl2. In prothrombin time (PT) assay, pre-

warmed 200 Al of PT reagent was added to the incubated mixture of PPP and anticoagulant and

clotting time was recorded. For thrombin time (TT) measurements, 100 Al of anticoagulant was

incubated with human thrombin (100 Al) for 3 min at 378C and clotting time was determined after the

addition of 200 Al of PPP.

Purification of anticoagulant

Lyophilized crude extract was dissolved in 20 mM tris-HCl buffer (pH 7.7) and separated using

a DEAE-Sephadex A-25 ion exchange column (b 35.0 � 280.0 mm). Adsorbed fractions were

collected using a 0–1.5 M NaCl at 215 nm and fractions resulted under a same elution peak were

pooled. Followed by desalting, pooled fractions were lyophilized and tested for APTT assay. An

aqueous solution of potent fraction was loaded onto a Sephadex G-25 gel filtration column (b 20.0 �750.0 mm). Separation was obtained at a flow rate of 0.5 ml/min with distilled water and eluted

fractions (3 ml) were pooled after spectrophotometric measurement. The active fraction selected by

APTT was injected into a capcell pack C18 UG120 (b 20.0 � 250.0 mm) HPLC column (Shiseido

Fine Chemicals, Tokyo, Japan). A linear gradient of acetonitrile (0–50% v/v, in 50 min)

in distilled water was maintained and elution peaks were collected at 215 nm. The peak

responsible for the highest anticoagulant activity was finally purified on a Zorbax SB C18 (b 4.6 �250.0 mm) HPLC column (Agilent Technologies, USA) with a linear gradient of acetonitrile

(0–12.5% v/v, in 75 min). Throughout the purification, all chromatographic procedures were

carried out at 48C.

Determination of purity and molecular weight

The accurate native molecular mass and purity of the anticoagulant were determined by matrix-

assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectroscopy on a Voyager DE-

PRO mass spectrometer (Applied Biosystems, Boston, MA) at a laser power of 3000 kWcm�2. Reduced

molecular weight was studied by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-

PAGE) on 12.5% gel following the method of Laemmli (1970).

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192610

Identification of amino acid composition and N-terminal sequence

Amino acid composition of a-chymotryptic crude extract was analyzed with an automatic amino acid

analyzer, Biochrom 20 (Biochrom Ltd., Cambridge, UK) after hydrolysis in 6 N HCl at 1108C for 24 h

in a vacuum.

The N-terminal amino acid sequence of the purified anticoagulant was obtained by automated Edman

degradation on a Perkin-Elmer (model 470A, Applied Biosystem Division, Branchburg, NJ) protein

sequencer equipped with an HPLC system.

Determination of specific factor activity

The specific activity of activated coagulation factors was determined by modified clotting assays of

APTT using ILR Test Factor Deficient Plasma assay kits (Instrumentation Laboratory Co., Lexington,

MA). Briefly, various concentrations of anticoagulant treated platelet poor human plasma were tested to

normalize prolong clotting time of specific factor deficient standard human plasma. The percentage

activity of specific coagulation factor was calculated following the manufacturer’s instructions.

PAGE analysis of FXIIa inhibition

Native PAGE was used to identify the inhibition of FXIIa. A mixture of FXIIa and the purified

anticoagulant was incubated at 378C in the presence or absence of Zn2+ (5 nM) for 75 min. After the

reaction, 15 Al aliquots were migrated on a 12.5% separating gel at pH 8.8 using Tris-HCl as the

electrolyte buffer. Calibration molecular weight markers used were aprotinin (MW 6.5 kDa), lysozyme

(MW 14.3 kDa), h-lactoglobulin (MW 18.4 kDa), egg albumin (MW 44.0 kDa) and bovine albumin

(66.0 kDa).

Fibrinolytic assay

Fibrinolytic activity of the purified anticoagulant was estimated by a modified fibrin plate assay of

Dametto et al. (2000) using plasmin as the standard. The fibrinogen solution (4.0 ml of 0.5% human

fibrinogen in 0.1 M barbital buffer, pH 7.7, containing 1.6 mM CaCl2, 0.7 mM MgCl2 and 92.0 mM

NaCl) was mixed with agarose solution (4.0 ml of 2.0% agarose in the same barbital buffer), 0.2 ml of

20 NIH U/ml thrombin in 0.9% saline and 0.1 ml of 1.0 M CaCl2. The mixture was poured into a

Petri dish and allowed to stand for 1 h at room temperature to obtain a fibrin clot layer. Various

concentrations of YAP solutions (0–600 AM) in plasminogen (10 Al) were spotted on the fibrin layer

and incubated at 378C for 10 h. Fibrinolytic activity was determined by measuring two perpendicular

diameters of the lysed zone.

Platelet aggregation inhibition assay

Microplate method of Bendar et al. (1995) was performed to detect platelet aggregation. Briefly, PRP

was adjusted with Ca2+ free Tyrode’s buffer (137.0 mM NaCl, 2.7 mM KCl, 0.5 mM NaH2PO4, 12.0

mM NaHCO3, 1.0 mM MgCl2 and 3.5 g/l bovine serum albumin) to obtain an approximate plate count

of 2 � 105/Al. Above solution (100 Al) was preincubated with 10 Al of differentially concentrated

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2611

purified anticoagulant (0–600 AM) at 378C for 12 min and 100 Al aliquots were added to individual

wells. Aggregation of platelets was initiated by adding 20 Al of ADP (50 AM) or thrombin (50 nM).

Change in optical density (OD) at 630 nm against PPP was monitored kinetically to evaluate platelet

aggregation by using an Emaxk microplate reader (Molecular Devices Co., Sunnyvale, CA). The rate

change in OD at 630 nm over 5 min was used to measure the platelet aggregation. Diluted PRP in the

absence of purified anticoagulant was kept as the control.

Platelet adhesion inhibition assay

The ability of platelets to adhere with collagen via specific integrin on cell membrane was measured

using a 96-well microtiter plate assay as described by Bellavite et al. (1994). In brief, various

concentrations (0–400 AM) of anticoagulant (10 Al) were preincubated with 150 Al of platelet solutionadjusted with Ca2+ free Tyrode’s buffer (approximately 2 � 105 platelets/Al) at 308C for 10 min. An

aliquot (150 Al) of above mixtures was added to type I collagen plated microtiter plate and incubated at

the same temperature. After 60 min of incubation, non-adherent platelets were completely washed off

with phosphatebuffered saline and adherent platelets were solubilized with 100 Al of buffer containing1.0% Triton X-100 and 5.0 mM p-nitrophenyl phosphate. After incubating for 60 min, the amount of

acid phosphate release (that represents the amount of platelets lysed) was measured using an Emaxkmicroplate reader at 405 nm. The platelet adhesion percentage was calculated using the OD of control,

where PRP was not preincubated with the anticoagulant.

Statistics

Results are presented as mean F SD of the mean (n = 3). Student’s t-test was used to determine the

level of significance.

Results

Extraction and purification of the anticoagulant

In a preliminary study, seven hydrolytic enzymes (as mentioned in the Materials and methods section)

were separately utilized to hydrolyze yellowfin sole frame protein and screened for potent anticoagulant

activity. The anticoagulant activities of hydrolysates were changed with the time course action of

hydrolysis and the highest inhibition of coagulation exhibited after four hours of reaction time (data

not shown). All hydrolysates could significantly prolong APTT but not PT or TT (P b 0.05). Those

results suggested that anticoagulant compounds present in all hydrolysates act to inhibit coagulation

factor(s) in intrinsic pathway of coagulation. Among the enzymes tested, a-chymotrypsin, papin and

trypsin could result comparatively a higher degree of hydrolysis (45–50%). Moreover, the highest

prolongation of APTT (68.4 F 0.5 s) was observed with a-chymotryptic hydrolysate and the same

enzyme was selected for the extraction of anticoagulant protein. The amino acid composition of a-

chymotryptic hydrolysate was rich in acidic amino acids (glutamic acid and aspartic acid), lysine,

leucine and glycine (Table 1). Together they constituted 53.34% of the total amino acid residues. This

Table 1

Amino acid composition of a-chymotryptic hydrolysate of yellofin sole frame protein

Amino acid Mole %

Asp 9.92

Thr 4.32

Ser 4.74

Glu 17.72

Pro 5.28

Gly 7.97

Ala 6.66

Cys 1.10

Val 4.66

Met 1.63

Ile 3.82

Leu 8.17

Tyr 2.66

Phe 4.15

His 2.95

Lys 9.56

Arg 4.68

Total 100.00

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192612

provides a better indication about negatively charged nature of anticoagulant compounds present in the

hydrolysate.

At the first purification with DEAE-Sephadex A-25 anion exchange chromatography, 79.3% of the

material applied was recovered in the same flow. Adsorbed 20.7% was eluted with a linear gradient

of 0–1.5 M NaCl and identified as three distinct fraction groups by APTT assay (Fig. 1A). The third

fraction pool (F-III, fractions 258–276) eluted with nearly a 1.0 M NaCl exhibited the highest

anticoagulation in vitro. Three fraction groups obtained from gel filtration chromatography displayed

a broader elution profile analyzed by APTT assay (Fig. 1B). The active fraction pool (F-IIIc)

containing a heterogeneous population of less hydrophilic molecules were separated into five distinct

peaks (P1–P5) following semi-preparative C18 HPLC (Fig. 1C). The peak (P2) eluted with 9–10%

acetonitrile had a significantly higher anticoagulant activity (P b 0.01) compared to the other

fractions observed. That was further purified by analytical C18 HPLC column into a clear single peak

(Fig. 1D) producing an overall purification fold of 206.6. Among all the purification steps carried out,

ion exchange chromatography on DEAE-Sephadex A-25 column could result the highest purification

fold and HPLC using capcell pack C18 UG120 column contributed the second highest fold

purification (Table 2).

Purity, molecular weight and amino acid sequence

SDS-PAGE on 12.5% slab gel showed that YAP migrates as a single band closer to the 14.3 kDa

marker. The MALDI-TOF mass spectrum of the purified anticoagulant exhibited a clear peak at 12.01

kDa position after resolution of isotopes (Fig. 2). The molecular mass of the native protein corresponds

well to the predicted molecular weight of 12 kDa under denatured conditions by SDS-PAGE. Therefore,

the purified compound appeared to be highly homogenous and existed as a monomeric protein mass of

0.0

1.5

F-IF-II

F-III

Pool

A

F-IIIa

F-IIIb

F-IIIc

Pool

B

D

0

50

150

200

0

50

P1P2

P3

P4P5

C

F-IF-II

F-III

Pool

A

Fraction number 0 50 100 150 200 250 300

0

1

2

3

4

25

30

35

40

45

50

55

60

F-IF-II

F-III

Pool

A

F-IIIa

F-IIIb

F-IIIc

Pool

B

0 20 40

0 20 40 5010 30

60 800.0

0.4

0.8

1.2

1.6

0

50

100

150

200

250

F-IIIa

F-IIIb

F-IIIc

Pool

D

0 15 30 45 60 750.0

0.3

0.6

0.9

1.2

0

5

10

15

20D

0.0

0.5

1.0

1.5

2.0

P1P2

P3

P4P5

C

Retention time (min)

Fraction number

Retention time (min)

100

P1P2

P3

P4P5

C

Abs

orba

nce

at 2

15 n

mA

bsor

banc

e at

215

nm

Ace

toni

trile

(%

)P

rolo

ngat

ion

of A

PT

T (

s)

Abs

orba

nce

at 2

15 n

mA

bsor

banc

e at

215

nm

Ace

toni

trile

(%

)N

aCl (

M)

Pro

long

atio

n of

AP

TT

(s)

Pro

long

atio

n of

AP

TT

(s)

Fig. 1. Purification of a potent coagulation inhibitor from a-chymotryptic hydrolysate of yellowfin sole frame protein. (A) Ion

exchange chromatography on a DEAE-Sephadex A-25 column. Lyophilized crude extract was dissolved in 20 mM tris-HCl

buffer (pH 7.7) and loaded onto the ion exchange column. Adsorbed fractions were eluted with 0–1.5 M NaCl gradient and

collected monitoring absorbance at 280 nm. Eluted fractions were tested for the inhibition of contact pathway coagulation by

APTT assay. (B) Gel filtration chromatography on Sephadex G-25 column. Potent fraction pool (F III) obtained from ion

exchange chromatography was applied onto the gel filtration column and fractions eluted with distilled water were tested for

APTT assay. (C) The lyophilized potent fraction (F IIIc) obtained from gel filtration chromatography was dissolved in distilled

water and injected into a capcell pack C18 UG120 HPLC column. The column was eluted with 0–50% (v/v) acetonitrile gradient

and elution peaks were assayed for APTT. (D) The highest anticoagulation activie peak (P2) obtained from capcell pack C18

UG120 HPLC was finally purified by Zorbax SB C18 HPLC column with 0–12.5% (v/v) acetonitrile gradient.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2613

12.01 kDa. We designated this protein as yellowfin sole anticoagulant protein (YAP) and N-terminal

amino acid sequence data determined to be TDGSEDYGILEIDSR.

Inhibition of coagulation and specific factor activity

As shown in Fig. 3, YAP prolonged APTT in a dose-dependant manner. However, APTT was not

affected by low concentrations of YAP. A marked increment in APTT prolongation was observed when

the concentration of YAP was higher than 2 AM. In addition, YAP was tested for PT and TT. Neither PT

nor TT was affected with the treatment of YAP, reflecting the results of crude protein extract (data not

shown). These results indicated that YAP is a specific inhibitor of the intrinsic pathway of coagulation.

Table 2

Purification of yellowfin sole anticoagulant protein (YAP)

Purification step Total protein

(mg)

Total anticoagulation

(min)

Specific anticoagulation

activity (s/Ag)Purification

fold

Recovery

(%)

Crude extraction 22463.21 254583.00 0.68 1.00 100.00

DEAE Sephadex A-25 105.54 117642.80 66.88 98.35 46.21

Sephadex G-25 13.59 21563.18 95.20 140.00 8.47

Capcell pack C18 UG120 1.24 2773.67 134.21 197.37 1.09

Zorbax SB C18 0.87 2036.67 140.46 206.56 0.80

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192614

YAP treated plasma was tested for the inhibitory effect on specific clotting factors in intrinsic pathway

of coagulation. A significant activity inhibition (62.4%, P b 0.05) of activated factor XII (FXIIa) was

observed when plasma was treated with final YAP concentration of 1.0 AM. Further, we investigated the

inhibition of FXIIa using serial dilutions of YAP and a dose dependant inhibition of FXIIa activity was

observed (Table 3). When the final concentration of YAP was about 1.5 AM, FXIIa activity was

completely inhibited. Interestingly, pre-incubation of YAP with 50 nM FXIIa could impair the inhibition

of FXIIa in a parallel vein. These results confirmed that YAP is an inhibitor of FXIIa.

50

YAP

66

44

1814.3

6.5

MW (kDa)

12012.14

16791.2611204.27

14340.96

14449.87

9999.0 11999.2 13999.4 15999.6 17999.8

Mass (m/z)

% In

tens

ity

100

90

80

70

60

40

30

20

10

0

3.0E+5

20000.0

Fig. 2. Molecular weight determination of YAP. MALDI-TOF mass spectrometry was performed to identify the native

molecular mass and SDS-PAGE was carried out on a 12.5% polyacrylamide gel to estimate the molecular weight under

reducing conditions. Molecular weight of YAP was estimated using standard molecular weight markers following the

visualization of protein bands by Coomassie Brilliant Blue.

YAP concentration (µM)

0 42 6 8 10 12

Pro

long

atio

n of

AP

TT

(s)

0

50

100

150

200

250

300

350

Fig. 3. Dose dependant inhibition of contact system induced coagulation by YAP. Various concentrations of YAP were

preincubated with platelet poor plasma and tested for the inhibition of coagulation by APTT assay.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2615

YAP mediated inhibitory mechanism of FXIIa

To investigate the inhibitory mode of YAP on FXIIa, native PAGE was performed. As shown in Fig.

4, mixtures of YAP and FXIIa in the presence or absence of Zn2+ (5 nM) were incubated at 378C and

migrated on a 12.5% slab gel indicating their protein bands at lane 3 (L3) and lane 4 (L4). YAP (12.01

kDa) and FXIIa (80 kDa) were also separately migrated on the same gel (L1 and L2 respectively). Both

reaction mixtures (L3 and L4) produced a single new band larger than 88 kDa, corresponding neither to

the band of YAP nor to the band of FXIIa. The new band (estimated molecular weight of 92 kDa)

migrated lesser distance than FXIIa, indicating a formation of a complex between YAP and FXIIa.

Effect of YAP on platelet aggregation and adhesion

When YAP was incubated with ADP or thrombin-induced PRP, a dose-dependent inhibition of

platelet aggregation (antiplatelet aggregation) was observed (Fig. 5). Thrombin-induced platelet

aggregation was inhibited to a higher degree by all YAP concentrations than ADP-induced. YAP

concentrations of 600 AM and 660 AM completely blocked thrombin-induced and ADP-induced platelet

aggregation respectively. In general, aggregation of platelets due to a variety of agonists including

thrombin is mediated via GPIIb/IIIa (Phillips et al., 1988). In fact, our observations suggest that YAP

may contain a potent inhibitor domain for the cell surface glycoprotein receptor GPIIb/IIIa.

Table 3

Inhibitory effect of YAP on FXIIa activity

Final YAP concentration (AM) FXIIa inhibition (%)

No added FXIIa Added FXIIa (50 nM)

0.5 36.67 22.27

1.0 62.43 48.51

1.5 100.00 78.83

M1 M2 L1 L2 L3 L4

88

66

44

18146.5

MW(kDa)

Fig. 4. Native PAGE determination of inhibitory mechanism of FXIIa. YAP and FXIIa were incubated at 378C in the presence

(L3) or absence (L4) of 5 nM Zn2+ and migrated on a 12.5% polyacrylamide gel in the absence of h-mercaptoethanol. M1 and

M2 indicate the standard molecular weight markers. L1 and L2 represent YAP and FXIIa respectively after visualized by

Coomassie Brilliant Blue.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192616

Further, we characterized the ability of YAP to antagonize platelet adhesion to collagen that mediates

via GPIa/IIa, the major collagen binding receptor on platelet surface (Depraetere et al., 1999). For that

we performed a 96-well microtiter plate assay to estimate the acid phosphate (a platelet enzyme) activity

of collagen-adhered platelets, which represents the amount of platelets adhered. However, all the

concentrations of YAP tested did not exhibit inhibitory activity on platelet adhesion to collagen via GPIa/

IIa integrin (data not shown).

YAP concentration (µM)

0 100 200 300 400 500 600

Agg

rega

tion

of p

late

lets

(δ O

D63

0 nm

/min

)

0

4

8

12

16

20

Fig. 5. Inhibitory effect of YAP on platelet aggregation. Different concentrations of YAP (0–600 AM) inhibit platelet

aggregation induced by 20 AM ADP (closed bars in black color) or thrombin (open bars in gray color). Platelet aggregation was

recorded by monitoring the linear change in optical density at 630 nm. yOD630 = change in optical density at 630 nm � 1000.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2617

Discussion

This study was designed aiming to purify a novel anticoagulant from an enzymatic hydrolysate of

marine fish protein source. For this purpose, a-chymotrypsin resulted yellowfin sole protein hydrolysate

was employed consecutive chromatographic separations followed by anticoagulant activity testing.

Finally, a single-chain monomeric protein mass of 12.01 kDa having potent anticoagulant properties was

identified and designated yellowfin sole anticoagulant protein (YAP). As observed in some other reported

anticoagulant proteins (Jung et al., 2002; Hojima et al., 1980), YAP consisted of considerable amount of

negatively charged amino acids in its sequence. The negative charge nature of anticoagulant compounds

expects to facilitate interaction with coagulation factors in the coagulation cascade. Dose-dependant

prolongation of intrinsic pathway of coagulation identified by APTT assay and ineffectiveness to prolong

PT and TT strongly indicated YAP as a specific inhibitor of the intrinsic pathway of coagulation.

We evaluated the inhibitory effect of YAP on specific clotting factors in intrinsic pathway of

coagulation and observed the purified anticoagulant protein is a specific inhibitor of FXIIa.

Autoactivation of factor XII into its active form (FXIIa) in contact with negatively charged surfaces

triggers intrinsic pathway of coagulation and follows sequential activation of number of other

coagulation factors (Miller et al., 1980). Up-to-date, several natural protein sequences have been

identified to inhibit different coagulation factors in intrinsic coagulation cascade (Jung et al., 2002; Urata

et al., 2003). However, only a few reports are available on anticoagulant proteins that inhibit FXIIa. Corn

trypsin inhibitor (12 kDa) with comparatively similar molecular weight acts same as YAP to inhibit

FXIIa (Hojima et al., 1980). Hamadarin, another anticoagulant protein isolated from the saliva of

Anopheles stephensi has shown that it inhibits reciprocal activation of FXII and kallikrein by directly

binding to both FXII and high molecular weight kininogen (Isawa et al., 2002). To investigate the

inhibitory mode of YAP on FXIIa, native PAGE experiments were performed. If YAP cleaves or binds

with FXIIa, the resultant product(s) should migrate on PAGE gel at different rates due to the difference in

molecular weight and relative charge. At 1:1 molar ratio of YAP and FXIIa, a new protein band was

observed and its molecular weight was closer to the estimated combined molecular weight (92 kDa) of

YAP and FXIIa. Therefore, our results confirm that inactivation of FXIIa by YAP is proceeded via

formation of an inactive complex rather than cleavage of FXIIa. It is expected that, Zn2+ induces some

conformational changes in FXII to bind with negatively charged activation surfaces (Schousboe, 1993).

However, incubation of FXIIa with YAP did not reveal the requirement of Zn2+ for the complex

formation between FXIIa and YAP. Further, this suggests that YAP may bind to FXIIa regardless of

metal ion induced conformational change in receptor binding domain. On the other hand, the

physiological concentration of FXIIa in activated normal human plasma is about 375 nM (Tans and

Rosing, 1987) and the complete inhibition of blood coagulation approximately at the same concentration

of YAP, agree with our explanation on the inactive complex formation between FXIIa and YAP.

Activation and aggregation of platelets also play an important role in the regulation of vascular-

platelet hemostasis. These functions are mainly determined by changes of the vascular wall in response

to multiple physiologic agonists such as, adenosine diphosphate (ADP), thrombin, thromboxane A2 and

collagen (Plow and Ginsberg, 1989). Activated platelets interact with each other (aggregation of

platelets) and with some other extracellular components such as collagen and fibrinogen (adhesion of

platelets) mainly via two distinct platelet surface glycoprotein integrins, GPIIb/IIIa (aIIbh3) and GPIa/IIa

(a2h1). Following activation, platelets secrete ADP and other agonists stimulating neighboring platelets

and provoking integrin GPIIb/IIIa-mediated Ca2+-dependent platelet aggregation. Antagonists of platelet

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–26192618

aggregation (platelet inhibitors) work towards blocking the binding of platelet surface integrins with

specific molecular targets during thrombosis (Cohen, 1996). Generally, aggregation of platelets due to a

variety of agonists including thrombin is mediated via GPIIb/IIIa. When we tested YAP for thrombin and

ADP-induced platelet aggregation separately, dose-dependent inhibitions were observed. Therefore, it

can be suggested that YAP may interact with platelet function via GPIIb/IIIa. For clear understanding of

YAP’s binding ability with other main platelet intergrin GPIa/IIa, adhesion of platelets with collagen was

evaluated. However, none of the YAP concentrations could compete with collagen to bind with GPIa/IIa.

Therefore, it can be suggested that YAP inhibit platelet surface integrin GPIIb/IIIa but not GPIa/IIa.

There is a growing interest in the role of FXII and FXIIa as potent initiators of contact activation-

dependent fibrinolysis. This is partly because, FXII shares structural similarity among some other plasma

fibrinolytic proteins namely, plasminogen, tissue-type plasminogen activator (tPA) and urokinase-type

plasminogen activator (uPA) (Tans and Rosing, 1987). Further, DNA sequence data have proven that

FXII exist multiple domains with extensive sequence homology with regions of tPA and fibronectin

(Cool and MacGillivray, 1987). However, the role of FXII as a plasminogen activator in fibrinolysis is

still unclear. But, some evidences strengthen the fact that FXIIa can induce fibrinolysis indirectly, by

activating prekallikrein into kallikrein, which in turn activates uPA and generate tPA in the presence of

bradykinin (Binnema et al., 1991).

Since YAP could bind with FXIIa, it may also have a possibility to bind with plasminogen, an

important zymogen in plasma fibrinolytic system. In this context, we assessed the plasminogen activator-

like property in YAP, utilizing fibrin plate method. But, all of YAP concentrations tested could not lyse

contact fibrin. This indicates that YAP does not involve in activating plasminogen or cleaving fibrin

cross-links by a direct action (data not shown). Therefore, our results speculate that YAP does not act as

a direct fibrinolytic agent by activating plasminogen. On the other hand, it is reasonable to argue YAP as

an indirect inhibitor of fibrinolysis, since it inhibits FXIIa, which suppose to be an initiator of contact

activation-dependant fibrinolysis. Related literature describes that natural fibrinolytic molecule,

hementin isolated from leeches can activate the fibrinolytic system of human plasma and resembles

functional similarity with streptokinase (Swadesh and Budzynski, 1990). However, in this study we did

not assess the ability of YAP to act on other fibrinolytic proteins like urokinase and streptokinase.

In the present work, we identified the new anticoagulant fish protein as an inhibitor of FXIIa and

platelet membrane integrin GPIIb/IIIa. Therefore, we introduce YAP as a multifunctional anticoagulant,

which exerts its action via different mechanisms. To our knowledge, this is the first report on

anticoagulant activity observed in fish protein and it can be suggest that fish protein hydrolysates are

worthwhile targets of investigating potent bioactive molecules.

Acknowledgement

This work was supported by the Brain Korea 21 project.

References

Bellavite, P., Andrioli, G., Guzzo, P., Arigliano, P., Chirumbolo, S., Manzato, F., Santonastaso, C., 1994. A colorimetric method

for the measurement of platelet adhesion in microtiter plates. Analytical Biochemistry 216, 444–450.

N. Rajapakse et al. / Life Sciences 76 (2005) 2607–2619 2619

Bendar, B., Condra, C., Gould, R.J., Connolly, T.M., 1995. Platelet aggregation monitored in a 96-well microplate reader is

useful for evaluation of platelet agonists and antagonists. Thrombosis Research 77, 453–463.

Binnema, D.J., Dooijewaard, G., Turion, P.N.C., 1991. An analysis of the activators of single-chain urokinase-type plasminogen

activator (scu-PA) in the dextran sulphate euglobulin fraction of normal plasma and of plasmas deficient in factor XII and

prekallikrein. Thrombosis and Haemostasis 65, 144–148.

Clemetson, K.J., Clemetson, J.M., 2001. Platelet collagen receptors. Thrombosis and Haemostasis 86, 189–197.

Cohen, M., 1996. Platelet glycoprotein IIb/IIIa receptor inhibitors in coronary artery diseases. Annals of Internal Medicine 124,

843–844.

Cool, D.E., MacGillivray, R.T.A., 1987. Characterization of the human blood coagulation factor XII gene. The Journal of

Biological Chemistry 262, 13662–13673.

Dametto, M., David, A.P., Azzolini, S.S., Campos, I.T.N., Tanaka, A.M., Gomes, A., Andreotti, R., Tanaka, A.S., 2000.

Purification and characterization of a trypsin-like enzyme with fibrinolytic activity present in the abdomen of horn fly.

Haematobia irritans irritans (Diptera: Muscidae). Journal of Protein Chemistry 19, 515–521.

Depraetere, H., Kerekes, A., Deckmyn, H., 1999. The collagen-binding leech products rLAPP and Calin prevent both von

Willebrand factor and a2h1 (GPIa/IIa)-I-domain binding to collagen in a different manner. Thrombosis and Haemostasis 82,

1160–1163.

Hojima, Y., Pierce, J.V., Pisano, J.J., 1980. Hageman factor fragment inhibitor in corn seeds: purification and characterization.

Thrombosis Research 20, 149–162.

Isawa, H., Yuda, M., Orito, Y., Chinzei, Y., 2002. A mosquito salivary protein inhibits activation of the plasma contact system

by binding to factor XII and high molecular weight kininogen. The Journal of Biological Chemistry 277, 27651–27658.

Jung, W.K., Je, J.Y., Kim, H.J., Kim, S.K., 2002. A novel anticoagulant protein from Scapharca broughtonii. Journal of

Biochemistry and Molecular Biology 35, 199–205.

Kim, S.K., Jeon, Y.J., Byeun, H.G., Kim, Y.T., Lee, C.K., 1997. Enzymatic recovery of cod frame proteins with crude protease

from tuna pyloric caeca. Fisheries Science 63, 421–427.

Laemmli, U.K., 1970. Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature 227,

680–685.

Lowry, O.H., Rosebrough, A.L., Farr, A.L., Randall, R.J., 1951. Protein measurement with Folin-phenol reagent. The Journal of

Biological Chemistry 193, 165–275.

Miller, G., Silverberg, M., Kaplan, A.P., 1980. Autoactivation of human Hageman factor. Biochemical and Biophysical

Research Communications 92, 803–810.

Phillips, D.R., Charo, I.F., Parise, L.V., Fitzgerald, L.A., 1988. The platelet membrane glycoprotein IIb-IIIa complex. Blood 71,

831–843.

Plow, E.F., Ginsberg, M.H., 1989. Cellular adhesion: GPIIb-IIIa as a prototypic adhesion receptor. Progress in Hemostasis and

Thrombosis 9, 117–156.

Schousboe, I., 1993. Contact activation in human plasma is triggered by zinc ion modulation of factor XII (Hageman factor).

Blood Coagulation and Fibrinolysis 4, 671–678.

Swadesh, J.K., Budzynski, I.Y.H.Z., 1990. Purification and characterization of hementin, a fibrinogenolytic protease from the

leech Haementeria ghilianii. Journal of Chromatography 502, 359–369.

Tanaka-Azevedo, A.M., Tanaka, A.S., Sano-Martins, I.S., 2003. A new blood coagulation inhibitor from the snake Bothrops

jararaka plasma: isolation and characterization. Biochemical and Biophysical Research Communications 308, 706–712.

Tans, G., Rosing, J., 1987. Structural and functional characterization of FXII. Seminars in Thrombosis and Hemostasis 13,

1–14.

Urata, J., Shojo, H., Kaneko, Y., 2003. Inhibition mechanisms of hematophagous invertebrate compounds acting on the host

blood coagulation and platelet aggregation pathways. Biochimie 85, 493–500.