Biochemical and Hemostatic Mechanism of A Novel Thrombin-Like Enzyme

Thrombosis Research 124 (2009) 631–639

Contents lists available at ScienceDirect

Thrombosis Research j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t h r o m r e s

Regular Article

Biochemical and Hemostatic Mechanism of A Novel Thrombin-Like Enzyme Song-Shan Tang ⁎, Juan-Hui Zhang 1, Bai-Shan Tang 2, Zhi-Hua Tang 3, Hong-Zhi Li 4, Hao-Jia Yuan 5, Shu-Lian Chui 6, Er-Yao Zhao 7 Department of Biochemistry & Molecular Biology, School of Basic Courses, Guangdong Pharmaceutical University, Guangzhou 510006, China

a r t i c l e

i n f o

Article history: Received 1 March 2009 Received in revised form 2 June 2009 Accepted 3 June 2009 Available online 15 August 2009 Keywords: Agkistrodon acutus snake venom Thrombin-like enzyme Coagulant Agacutin Hemostasis Blood coagulation

a b s t r a c t Thrombin-like enzyme (TLE) plays a significant role in vessel injury hemostasis. A novel snake venom TLE (Agacutin) was purified from Agkistrodon Acutus snake venom. Structural analysis indicated that Agacutin is a heterodimer that has a MW of 29,402 Da, a pI value of 5.39, and optimum activity at 35 °C and pH 7.5. The N-terminal 15 amino acid sequences of Agacutin are DSSGWSSYEGHEYYV (small subunit) and DCSSGWSSYEEHQYY (large subunit). In vitro studies indicated that the coagulation activity of Agacutin was activated by Ca+2 or inhibited by phenylmethanesulfonyl fluoride, but not influenced by heparin or hirudin. The arginine esterase activity and fibrinogen hydrolysis result showed that Agacutin only cleaves α-subunit and releases fibrinopeptide A. In vivo studies indicated that Agacutin iv (0.01-0.05 U/kg) shortened 30.2-49% of the rabbit blood clotting time, or ip (0.5-2.0 U/kg) shortened 29.7-73.1% of the mouse tail bleeding time. Agacutin does not influence APTT, platelet or euglobulin clotting time, and activate Factor II or XIII. It converts fibrinogen into the soluble fibrin that accelerates hemostasis at wound. © 2009 Elsevier Ltd. All rights reserved.

Introduction Snake venom contains various components that affect the mammalian hemostatic system as either procoagulants or anticoagulants [1,2]. These components interact with diverse proteins of the blood coagulation cascade and the fibrinolytic pathway. Generally, according to the results of in vivo and / or vitro analysis, these proteins are classified into the following seven groups: (i) fibrinogen clotting enzymes; (ii) fibrino (geno)lytic enzymes; (iii) plasminogen activators; (iv) prothrombin activators; (v) factor V, X activators; (vi) hemorrhagins; and (vii) platelet aggregation inhibitors [3]. Thrombin-like enzyme (TLE) is protease that belongs to the first group (i). TLE can accelerate the clotting of plasma or

⁎ Corresponding author. Tel.: +86 20 39352192; fax: +86 20 39352187. E-mail address: [email protected] (S.-S. Tang). 1 Now in Department of Obstetrics & Gynecology, Guangdong Armed Police Hospital, Guangzhou 510507, China. 2 Now graduate student in the Second Medical College, Lanzhou University, Lanzhou 730000, China. 3 Now in Pharmacological Lab, Huadong Medical Biotechnological Institute, Nanjing 210009, China. 4 Now in Department of Biology, Guangdong Pharmaceutical University, Guangzhou 510006, China. 5 Now in the Third Lab of Lanzhou Bio-product Institute of Healthy Ministry, Lanzhou 730030, China. 6 Now in Medical Nutrition Department, Guangdong Lingnan Vocational and Technical College, Guangzhou 510503, China. 7 Now in Technical Department, Kunming Longjin Pharmaceutical Co., Ltd, Kunming 650032, China. 0049-3848/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2009.06.022

fibrinogen solution in vitro, but in vivo they have different effects, for example hemostasis and anticoagulation. By exploiting some of the effects, these proteins have been developed into clinical drugs. Reptilase from Bothrops atrox snake venom serves as an example of the successful clinical drugs. Reptilase shortens the blood clotting and bleeding time in vivo and has been used widely in the treatment of wound hemorrhages and in the surgery to prevent bleeding. In 2004, Reptilase was successfully recombined into Pichia pastoris [4]. In 2005, a 35.5 kDa TLE with hemostatic function was purified from Agkistrodon blomhoffii ussurensis snake venom [5]. Interestingly, its N-terminal 15 amino acid residue sequence was the same as that of Reptilase and its hemostatic effect was similar to that of Reptilase. In 2006, Wang YN et al [6] purified a novel Factor X activator from Viper russellii snake venom. This activator showed better hemostatic activity in mice and rats at a dose of 0.000313 U/kg. Since 1995, a number of enzymes with structural and functional diversity have been purified from Agkistrodon acutus snake venom. In 1995, Chen YL et al [7] purified Agkicetin (a heterodimer with 14 and 15 kDa subunits), a GP1b antagonist and platelet inhibitor. In 1998, Chia-Hsin Yeh et al [8] purified Accutin (5.241 kDa), a new member of disintegrin family, which potently inhibits human platelet aggregation. In 1999, Huang QQ et al [9] reported two TLEs, Acuthrombin-A (28 kDa) & Acuthrombin-C (69 kDa). Cheng X et al [10] purified the fibrinogenolytic enzyme Agkisacutacin (a heterodimer with 14 and 15 kDa subunits). Pan H et al [11] successfully cloned Acutin (38 kDa) which caused defibrination. In 2000, Xu XL et al [12] reported two TLEs, Anticlotting Factor-I (ACF-I) and Anticlotting Factor-II (ACF-II) (a heterodimer with 14.6 and 14.7 kDa subunits). In 2001, Yeh CH

632

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

et al [13] reported a platelet glycoprotein Ib antagonist Agkistin (a heterodimer with 16.5 and 15.5 kDa subunits) and Liang XX et al [14] reported a fibrinolytic enzyme F II (a) (26 kDa). In 2004, Li T et al [15] purified Acoagulatin (a heterodimer with 14.4 and 17 kDa subunits), an anticoagulant. The publications on all these enzymes enrich our understanding about the components of the venom. Since 1993, we have tried to purify the coagulant from Agkistrodon acutus snake venom. Fortunately, five novel enzymes which use fibrinogen as substrate have been obtained. Agacutin is the most comprehensively studied one. The biochemical properties and the in vitro and in vivo hemostatic mechanisms of Agacutin are reported in this paper. Materials and methods

separate the two subunits of the heterodimer. The Edman Degradation method (ProciseR cLC Protein Sequencer, Applied Biosystems, USA) was used to determine the sequence of each subunit. Determination of the iso-electric point (pI) According to the method of Righetti P et al [17], iso-electric focusing (IEF) electrophoresis was performed by using PAG gel matrix (T:C = 7:3) and Ampholine (pH 3.5-9.5, Pharmacia Co. Sweden) to form an appropriate pH gradient. A Sanhen electrophoresis apparatus (ECP3000, Beijing Medical Apparatus Co.) was used in the experiment. The pH values which were measured at each 0.5 cm region of the gel were plotted against the corresponding gel lengths and the resulting pI of Agacutin was calculated.

Preparation of Agacutin Study of the biochemical properties 1.0 g of the crude snake venom powder (purchased from the Snake Venom Institute, Guangxi Medical University, China) was dissolved in 8 ml of 0.01 M phosphate buffer (pH 6.8) and centrifuged at 6000 rpm for 20 min. The supernatant was applied to a gel filtration column of Sephadex G-75 and eluted with 0.01 M phosphate buffer (pH 6.8). After the relevant fraction had been collected, it was applied to a DEAE-Sepharose Fast Flow (FF) ion exchange column and eluted with a linear gradient of 0-0.065 M NaCI in 0.01 M phosphate buffer (pH 6.8). After the effective fraction had been dialyzed against 0.01 M phosphate buffer (pH 6.8) for 24 h, the sample was re-applied to the regenerated DEAE-Sepharose FF column to repeat the linear gradient elution of 0-0.065 M NaCI. The effective fraction was subsequently applied to a Sephadex G-25 column for desalting. The preparation procedure was finished after the purified Agacutin was freeze-dried. One vial for the animal experiments contained 1 unit of Agacutin and 1% micro-molecular dextran (purchased from the Sin Food Drug Administration, SFDA). The Econo-system preparative liquid chromatographer (Bio-Rad Co, USA) was used in the preparation.

The optimum temperature or pH value of Agacutin-induced clotting time was obtained from the plot of Agacutin-induced clotting time. A 0.2 ml Agacutin water solution (0.2 U) was mixed with 0.2 ml of standard human plasma (Dade Behring Co., Germany) containing 20-500 U sodium heparin, 2.5-25 mM phenylmethanesulfonyl fluoride (PMSF), or 5-100 AT-U hirudin to measure plasma clotting time at 37 °C. To determine the influence of Ca2+ on ox fibrinogen clotting time, 0.45-9.0 mM CaCI2 was used. Porcine thrombin (Sigma Co.) was required as a control in some experiments. Urea dissolving assay In order to observe whether the plasma clot was re-dissolved, 1 ml of 3 M urea was added to a human plasma clot which was previously prepared by mixing 0.2 ml of human plasma with 0.2 U of Agacutin at 37 °C. Determination of enzyme activity

HPLC analysis of Agacutin High performance liquid chromatography (HPLC) (Waters 6000E, USA) was performed to analyze the purity of Agacutin. For gelfiltration HPLC column (BIOSEP SEC-2000, Pharmacia-LKB Co), an elution of 0.2 M phosphate buffer (pH 6.8) and a flow rate of 1 ml/min were used. For reverse-phase HPLC C4-column (HypersilC4, 300 Å, ODS-BP, 2.6 × 40 mm), Solution A (0.1% trifluoroacetic acid water solution containing 0.1% n-butylamine), Solution B (acetonitrile solution containing 0.1% trifluoroacetic acid and 0.1% n-butylamine) and a flow rate of 1 ml/min were used. The elution gradients of 0 min/ 40% to 10 min/ 64% to 10.1 min/ 64% to 15 min/ 40% (retention time / Solution B ratio) were utilized during the analysis. SDS-PAGE analysis A polyacrylamide gel (T: C = 12: 3.9) was prepared for SDS-PAGE. MALDI-TOF Mass Spectrographic analysis 0.75 µl of Agacutin (200 ng/µl) was mixed with an equal volume of 10 mg/ml mustard acid. The mixture was applied to a mass spectrometer (ABI Co, USA). N-terminal amino acid sequencing of Agacutin Agacutin is a heterodimer consisting of 15 and 16 kDa subunits. In order to sequence the N-terminal 15 amino acid residues, SDS-PAGE (T: C = 15:3.9) and the PVDF membrane electrical transfer technique [16] (Mini Trans-Blot & Power Pac HC, Bio-Rad Co, USA) were used to

Coagulation activity assay The enzyme activity of Agacutin was estimated according to the Reptilase Unit (KU) method [18]. Here, 0.2 ml of Agacutin-water solution (1 KU/ml) should coagulate both 0.2 ml of standard human plasma (Dade Behring Co., Germany) at 37 °C in 60 ± 20 seconds and 1 ml of ox fibrinogen solution (0.4%) (purchased from SFDA) in 232 ± 56 seconds. The plasma, ox fibrinogen and Agacutin water solution were pre-incubated individually for 3 min at 37C in a glass incubator. When the 0.2 ml of plasma or the 1 ml of ox fibrinogen solution was mixed with 0.2 ml of Agacutin at 37 °C, the cloudiness in the mixture occurred. The plasma or ox fibrinogen clotting time was obtained by counting the time from the mixture to the initial cloudiness. KU was used in vivo and vitro experiments below. In the coagulation activity assay, because the activation method has a larger eye observation error, each sample was assayed activity three times. The first two activities were preliminary results. The 3rd result which was recorded in the paper was the most accurate and reliable. Arginine esterase activity assay Arginine esterase activity was determined according to the BAEE method [19]. Hydrolysis analysis of Agacutin to ox fibrinogen subunit 1 µg / µl of ox fibrinogen and 10 U / 100 µl of Agacutin were prepared with 50 mM Tris-HCI (pH 7.5), respectively. 100 µl of the ox fibrinogen and 100 µl of the Agacutin solution were mixed and incubated at 37 °C for 1, 2, 3, 4 h, respectively. The incubation samples were loaded on SDS-PAGE (T: C = 8:3.9). In the experiment porcine thrombin (1 U / 100 µl) was used as control.

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

633

Effect of Agacutin on the blood coagulation system Whole blood samples were collected from New Zealand white rabbit hearts before and after iv with various time points. The LeeWhite method [20] was used in the whole blood clotting time assay. The blood fibrinogen concentration was measured by using the STA Analyzers and STA Fibrinogen Kit (France); APTT was measured with Thrombosil (Ortho Diagnostic System); WBC, RBC and platelet counting was performed with a CELL-DYN3599R apparatus (Abbot Laboratories, USA). All animals (50% male and 50% female) were purchased from the Southern Medical University. Reptilase vials were purchased from Basle Co. (1 KU/vial, Rat No770604). Euglobulin clotting time (ECT) and euglobulin lysis time (ELT) were determined according to the methods in Xu SY et al [20].

Mouse tail bleeding time Mouse tails (non-anesthetized mice) were cut in 30 min after ip, and the resulting blood from the tail was continually absorbed with filter paper until the tail stopped bleeding.

Fig. 2. A: The 2nd DEAE-Sepharose FF column chromatography. The linear elution gradient in the 1ST DEAE-Sepharose FF column chromatography was repeated. B: SDS-PAGE (T: C = 12:3.9) analysis of Agacutin. Lane 1 and 2: 20 and 10 µg sample; Lane 3: Standard molecular weight markers (14.4-97.4 kDa).

Maximum dosage tolerance for mice The maximum tolerant dosage of Agacutin for mice was obtained by using a 0.5 ml volume and 40 U/kg iv once and observing over a 14-day period. Results Purification of Agacutin Purification procedure The crude snake venom solution was applied to a Sephadex G-75 column (Fig. 1A). The second fraction from the Sephadex G-75 chromatography was applied to a DEAE-Sepharose FF column (Fig. 1B). After having been dialyzed in the phosphate buffer, the

Table 1 Summary of the purification procedure. Separation steps

Fig. 1. A: Sephadex G-75 gel filtration column chromatography of the crude venom. A column of 4.5 × 100 cm and flow rate 6 ml/min were used for the first step of the purification. B: The 1st DEAE-Sepharose FF column chromatography. A linear gradient elution was performed with 0-0.065 M NaCI in a phosphate buffer. A column of 3 × 20 cm and a flow rate 15 ml/min were used.

Total protein Total activity Specific activity Weight yield weight (mg) units (U) (U/mg)⁎⁎ (%)

Crude venom 1000 Pre-treatment of venom 900 G-75 471 1st DEAE-Sepharose F.F 56 2st DEAE-Sepharose F.F. 10.3 G-25 8.8

–⁎ – – – – 176 ± 20

–⁎ – – – – 19.7 ± 3.6

100 90 47.1 5.6 1.03 0.88

⁎The actual activity of Agacutin during intermediate steps of the purification process could not be measured due to the impact of salt and other TLE impurities on the plasma clotting time. ⁎⁎Protein concentration was determined with the Lowry method.

634

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

solution was applied to the 2nd DEAE Sepharose FF column. The second fraction from the 2nd DEAE-Sepharose FF column (Fig. 2A) contained pure Agacutin. The purification procedure was finished after the pure Agacutin was desalted with Sephadex G-25 column using water as eluent.

Summary of the purification procedure The results (Table 1) indicated that the final weight yield and specific activity of Agacutin were 0.88% (g/dl) and 19.7 ± 3.6 U / mg protein, respectively. The maximal purification fold was about 100 times.

Fig. 3. A: Gel-filtration HPLC analysis. B: The reverse-phase HPLC analysis. C: MALDI-TOF MS analysis of Agacutin.

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

635

The iso-electric point (pI) The IEF electrophoresis of Agacutin was performed by using PAG (T: C = 7.6:3) and 150 V (Fig. 4A). A pI value of 5.39 was obtained according to the relationship between the gel pH gradient and the gel length (Fig. 4B).

Biochemical properties of Agacutin Influence of temperature and pH on the human plasma clotting activity When the reaction temperature and pH values were plotted against the corresponding clotting time, activity-temperature (Fig. 5A) and activity-pH (Fig. 5B) curves were obtained. The optimum temperature and optimum pH value were determined to be 35 °C and 7.5, respectively.

Influence of PMSF, hirudin, and heparin on the human plasma coagulation activity The results indicated that concentrations of 2.5-25 mM PMSF (Fig. 6A) obviously inhibited the coagulation activity of Agacutin, but 5-100 AT-U hirudin (data not shown) did not influence the activity. The human plasma coagulation activity of Agacutin was not inhibited by heparin, whereas the plasma coagulation activity of thrombin was clearly inhibited by the addition of more than 90 U heparin / 0.4 ml (Fig. 6B).

Fig. 4. IEF electrophoresis analysis of Agacutin. A: IEF image. Lane 1: IEF markers; Lane 2: Agacutin. B: Relationship curve between pH value and gel length.

Purity analysis and subunit composition The results of the gel-filtration HPLC (Fig. 3A) and the reversephase HPLC (Fig. 3B) indicated that Agacutin was purified to high homogeneity. SDS-PAGE analysis showed that the molecular weights of its two subunits were 16,224 Da and 14,736 Da (Fig. 2B). The apparent molecular weight of the heterodimer was 30,960 Da.

MALDI-TOF MS analysis The MALDI-TOF MS result indicated that the accurate molecular weight of Agacutin was 29,402 Da (Fig. 3C).

Isolation of the subunits and the determination of the amino acid sequence The 15 and 16 kDa subunits of Agacutin were isolated by using SDS-PAGE and blotted on PVDF membrane by using an electrical transfer technique (150 mA for 1.5 h). After the protein bands were stained with Coomassie brilliant blue G-250, these bands were excised and prepared for sequence analysis via the Edman Degradation method. The N-terminal 15 amino acid residues of the subunits were determined to be DSSGWSSYEGHEYYV (15-kDa subunit) and DCSSGWSSYEEHQYY (16-kDa subunit), respectively.

Fig. 5. A: Activity-temperature relationship curve. B: Activity-pH relationship curve.

636

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

Influence of Agacutin on the rabbit blood system Analysis of whole blood clotting time The results (Table 2) indicated that, compared with the results before iv, Agacutin iv shortened 30.2 to 49% of the rabbit blood clotting time. The peak effect appeared in 30 min and the effect prolonged for 24 h after iv. Compared with the results before iv, Agacutin or Reptilase had significant differences in the rabbit blood clotting time. In these experiments the effect of Reptilase was delayed for only 6 hours.

Fig. 6. Influence of PMSF (A) and heparin (B) on human plasma clotting activity.

Influence of Ca+2 on ox fibrinogen coagulation activity The results indicated that Agacutin-induced ox fibrinogen clotting time shortened with the increase of Ca+2 concentration (Fig. 7A).

Urea dissolving assay When 1 ml of 3 M Urea solution was added to a human plasma clot previously formed by Agacutin and softly shook, the clot was quickly re-dissolved.

Enzyme activity Coagulation activity Agacutin displayed robust human plasma coagulation activity. Its specific activity for the pure enzyme was enriched to 19.7 ± 3.6 KU per one mg protein, or 48.5 to 51.7 µg of protein per one KU.

Arginine esterase activity The result indicated that 1 mg of the enzyme hydrolyzed 1.4277 µmol of BAEE per minute or 1 U of Agacutin hydrolyzed 0.1856 µmol of BAEE per minute, which indicated that Agacutin is an arginine esterase with weak BAEE activity. Hydrolysis of Agacutin to ox fibrinogen The hydrolysis picture (Fig. 7B) showed that Agacutin only hydrolyzes α-subunit of ox fibrinogen and releases Fibrinopeptide A, whereas porcine thrombin exerts on α and β subunits of fibrinogen (Fig. 7C).

Fig. 7. A: Influence of Ca+2 on the ox fibrinogen clotting activity. B: Hydrolysis analysis of Agacutin to ox fibrinogen. Lane 1: fibrinogen; Lane 2-5: fibrinogen hydrolyzed by Agacutin for 1, 2, 3, 4 h; Lane M: Standard molecular protein markers (14.4-97.6 kDa). α, β, and γ stand for the three subunit bands of fibrinogen. C: Hydrolysis analysis of porcine thrombin to ox fibrinogen. Lane 1: fibrinogen; Lane 2-5: fibrinogen hydrolyzed by thrombin for 1, 2, 3, 4 h; Lane M: Standard molecular protein markers (14.4-97.6 kDa). α, β, and γ stand for the three subunit bands of fibrinogen.

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

637

Table 2 Influence of Agacutin on the rabbit whole blood clotting time (x̄ ± s, n = 7). Group

Before iv

Whole blood clotting time (Sec) before and after iv with various time points 0.17 h

0.5 h

1h

3h

6h

12 h

24 h

Agacutin 0.05U/kg Agacutin0.025U/kg Agacutin 0.01U/kg Reptilase0.025U/kg

590 ± 258 600 ± 200 593 ± 189 620 ± 149

382 ± 83⁎ 346 ± 143⁎ 377 ± 113⁎ 423 ± 230⁎

294 ± 203⁎ 342 ± 103⁎ 412 ± 244⁎ 365 ± 174⁎

368 ± 120⁎ 287 ± 134⁎ 425 ± 150⁎ 386 ± 135⁎

369 ± 133⁎ 411 ± 104⁎ 293 ± 249⁎ 546 ± 162⁎

286 ± 163⁎ 329 ± 91⁎ 345 ± 214⁎ 555 ± 162

479 ± 122 351 ± 121⁎ 455 ± 165 573 ± 134

540 ± 145 413 ± 100⁎ 580 ± 179 640 ± 120

⁎P b 0.05 by t test vs. Result before administered.

Analysis of mouse tail bleeding time Agacutin (0.5, 1.0, 2.0 U/kg) and Reptilase (0.5 U/kg) were administered by ip. The result of Agacutin or Reptilase was less than that of saline group and there was significant difference between Agacutin 0.5 U/kg and saline group (P b 0.01) (Table 3). Rabbit blood fibrinogen concentration Blood fibrinogen concentrations for each experimental group were decreased after iv, and returned to normal levels within 12 hours (Table 4). The results of Reptilase or Agacutin (0.05 U/kg) showed significant difference compared with the results before in vivo. Activated partial thromboplastin time (APTT) assay APTT results indicated that there was not significant difference between the results before and after iv within the same experimental group (Table 5). ELT and ECT After iv, Agacutin decreased the ELT (Table 6), but did not influence the ECT (data not shown). Maximum tolerant toxicity The maximum tolerance level of Agacutin in mice was obtained via a continuous observation of 14- day period after iv (Table 7). During the observation period, no abnormal observations were discovered regarding the behavior, or diet and fur color of the mice. Discussion Since the 1970s, numerous snake venom enzymes, such as Ancrod, Batroxobin, Crotalase, and Reptilase, have been developed into clinical drugs. These drugs play definite roles in blood anticoagulation and hemorrhage diseases. Due to the great efficacy and lower toxicity of Reptilase in clinic, this enzyme is considered one of the elemental clinical drugs in China. SDS-PAGE analysis of the purification fraction and change of rabbit blood clotting time were used to screen TLE. In the purification process, the presence of many snake venom proteins with similar MW, pI, or coagulation activity made the isolation and identification of Agacutin extremely challenging. Salt and other TLE impurities in the purification process had an impact on the precise determination of Agacutin activity. For this reason, in Table 1 there did not show the

Table 3 Results of mouse bleeding time. Group

Dose (U/kg)

Administration

n

Bleeding time (min) (x̄ ± s)

Saline Agacutin Agacutin Agacutin Reptilase

— 0.5 1.0 2.0 0.5

Ip Ip Ip Ip Ip

10 10 11 12 10

34.1 ± 11.1 19.7 ± 7.5⁎⁎ 26.9 ± 12 26 ± 18.0 28.7 ± 17.1

⁎⁎P b 0.01 vs. Saline group.

activity or specific activity data of these intermediates. However the plasma coagulation activities of these intermediates were determined to evaluate whether the samples had activity. In the final step of the purification, the Sephadex G-25 column was used to desalt the enzyme solution so that the actual activity of the pure enzyme could be determined. Biochemical research indicated that Agacutin is a TLE composed of two subunits of 15 and 16 kDa. By using NCBI Blast, the sequence analysis indicated that the N-terminal 15 amino acid residues of the 15 or 16 kDa subunits have maximal 66.7% or 73.3% sequence identity, respectively with the A chain of Agkisacutacin (Aa-X-Bp-I) from Agkistrodon Acutus snake venom [21]. In the research we discovered that the coagulation activity of Agacutin was not influenced by heparin and hirudin, but stimulated by the addition of Ca+2 (0.45-9.0 mM) or inhibited by PMSF (2.5-25 mM), which suggested that Agacutin is Ser-dependent poly-subunit protease. The result of the heparin (20-500 U) assay indicated that the plasma coagulation mechanism of Agacutin is different from that of thrombin, but similar to that of Reptilase. Hirudin (5-100 AT-U) did not influence the clotting time induced by Agactin (0.2 KU) indicating that Agacutin-induced clotting time is independent of thrombin activity. The results of the in vitro enzyme activity indicated that Agacutin has less ox fibrinogen coagulation activity or arginine esterase activity compared with thrombin or Batroxobin, which suggested that Agacutin has a stronger hemostatic function rather than an anticoagulant function caused by the defibrination. A lower affinity of Agacutin to fibrinogen induces a low level of soluble fibrin in the blood vessel, whereas a higher affinity of Batroxobin to fibrinogen causes a lot of fibrinogen to be transformed rapidly into soluble fibrin, which subsequently causes endogenous t-PA release to lead to activation of the fibrinolytic system. The activity of Agacutin was determined by using Reptilase unit because the in vitro and in vivo properties of Agacutin were very similar to those of Reptilase. In the fibrinogen coagulation activity assay, the ox fibrinogen activity standard for Agacutin (I KU: 232 ± 56 seconds) was far less than that of Reptilase (1 KU: 100 ± 20 seconds), which suggested that the affinity of Agacutin to fibrinogen was far less than that of Reptilase to fibrinogen. The arginine esterase activity and the α-subunit hydrolysis of Agacutin to fibrinogen showed that Agacutin specifically cleaves Arg16-Gly17 bond of fibrinogen α-chain and releases Fibrinopeptide A, whereas thrombin hydrolyzes α and β-subunits. The structure of Agacutin is obviously different from that of Reptilase or Batroxobin. Reptilase is a 32 kDa-single-chain glycoprotein. One lyophilized vial of Reptilase contains 1 KU (1 KU = 0.04 NIH Unit), which shortens 1/3-1/2 of the adult blood clotting time. Batroxobin is a 36 kDa-single-chain protein. One vial of Batroxobin contains 10 Batroxobin Units (1 BU = 0.17 NIH Unit), which delays blood clotting time at least 30 min (2 BU in 0.1 ml of enzyme makes 0.3 ml of 4% ox fibrinogen clot in 19 ± 2 seconds). The iv or im of Batroxobin can lead to blood anticoagulation by causing a rapid and prolonged defibrination. The difference of specific activity between the Acutin (Batroxobin-like TLE) 2000 BU/mg (or 8000 KU/mg) and Agacutin 20 KU/mg also explain the difference in effect. These properties help us to understand their inherent differences.

638

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639

Table 4 Influence of Agacutin on the rabbit blood fibrinogen (x̄ ± s, n = 7). Group (U/kg) Agacutin 0.05 Agacutin0.025 Agacutin 0.01 Reptilase0.025

Blood fibrinogen concentration (mg/dl) before and after iv with various time points

Before iv 256.0 ± 32.1 222.1 ± 47.4 254.0 ± 56.1 265.3 ± 34.9

0.17 h

0.5 h

1h

3h

6h

12 h

24 h

188.0 ± 21.6⁎ 210.7 ± 44.4 247.3 ± 55.7 219.8 ± 31.3⁎

186.2 ± 20.9⁎ 204.3 ± 60.2 218.7 ± 107.0 216.5 ± 25.9⁎

174.3 ± 27.2⁎ 206.6 ± 53.9 249.0 ± 57.7 220.7 ± 24.4⁎

178.1 ± 23.7⁎ 209.0 ± 53.1 218.2 ± 42.9⁎ 221.5 ± 38.5⁎

209.9 ± 19.9 219.0 ± 58.6 221.8 ± 83.7 212.5 ± 24.7⁎

254.2 ± 20.2 299.8 ± 37.5 316.5 ± 33.4 245.2 ± 21.5

412.5 ± 23.0 318.0 ± 56.6 370.7 ± 43.3 332.2 ± 34.0

⁎P b 0.05 by t test vs. Result before administered.

Table 5 Influence of Agacutin on the rabbit blood APTT (x̄ ± s, n = 7). Group (U/kg)

Before iv

APTT (Sec) before and after iv with various time points

Agacutin 0.05 Agacutin0.025 Agacutin 0.01 Reptilase0.025

66.10 ± 17.9 84.52 ± 14.7 75.35 ± 15.2 69.38 ± 5.60

0.17 h

0.5 h

1h

3h

6h

12 h

24 h

77.40 ± 21.7 77.48 ± 14.9 76.20 ± 16.4 79.82 ± 15.6

76.57 ± 25.9 72.28 ± 16.8 78.70 ± 19.1 73.53 ± 10.9

67.58 ± 17.9 72.33 ± 15.6 72.68 ± 13.4 79.13 ± 16.0

81.98 ± 15.7 73.45 ± 11.9 56.03 ± 17.0 66.10 ± 15.6

60.43 ± 8.6 83.92 ± 12.4 56.95 ± 14.0⁎ 63.75 ± 12.3

62.03 ± 15.6 66.66 ± 13.4 62.35 ± 11.5 71.02 ± 12.7

68.35 ± 15.7 79.08 ± 13.8 60.87 ± 14.4 70.34 ± 14.5

⁎P b 0.05 by t test vs. Result before administered.

Table 6 Influence of Agacutin on the New Zealand rabbit ELT (x̄ ± s, n = 7). Group

Dose (U/kg)

ELT (min) after iv with various time points 0.17 h

0.5 h

1h

3h

6h

12 h

24 h

Control Agacutin Agacutin Agacutin Reptilase

—0.05 0.025 0.01 0.025

246 ± 8 169 ± 97 134 ± 7⁎⁎ 85 ± 28⁎⁎ 211 ± 86

235 ± 6 155 ± 7⁎ 135 ± 11⁎ 75 ± 6⁎⁎ 202 ± 77

225 ± 76 156 ± 4 142 ± 35⁎ 75 ± 28⁎⁎ 139 ± 86

191 ± 4 151 ± 4⁎⁎ 235 ± 8 136 ± 7⁎⁎ 286 ± 90⁎

202 ± 5 160 ± 6⁎ 168 ± 12⁎ 121 ± 6⁎⁎ 233 ± 46

198 ± 14 163 ± 24⁎ 189 ± 7 86 ± 10⁎⁎ 201 ± 88

257 ± 55 178 ± 6⁎ 246 ± 35 129 ± 12⁎⁎ 268 ± 90

⁎P b 0.05, ⁎⁎P b 0.01 by t test vs. Control.

Table 7 Maximal tolerant dosage result of Agacutin for mice. Administered approach

Dose (U/kg)

n

Death rate (%)

Iv

40

12

0

The in vivo effect of Agacutin indicated that 0.01 to 0.05 U/kg iv shortened 30.2 to 49% of the rabbit blood clotting time. We discovered that the dose 0.025 U/kg suggested the best hemostatic effect, which may be that the dosage did not cause in vivo fibrinolytic effect, whereas a larger dose (0.05 U/kg) caused a pronounced fibrinolysis which leads to a weak hemostatic effect (see ELTs in Table 6). However, in this experiment, we did not understand why the lowest dose (0.01 U/ kg) caused a fibrinolytic effect. In the future studies we will pay particularly attention to this event. The results of the urea dissolving, hirudin (data not shown), and APTT assays suggested that Agacutin did not activate Factor XIII or Factor II. The hemostatic effect of Agacutin was delayed for 24 hours whereas that of Reptilase was only done for 6 hours, which results from the longer half-life of Agacutin (we have obtained the slow clearance half-life of Agacutin: t 1/2 β 939.68-1267 min, the experimental information will be submitted soon). The results of the blood clotting time, blood viscosity (data not shown), and blood fibrinogen concentration indicated that Agacutin generated a continuously soluble fibrin level in the blood vessel. When a wound occurs, the soluble fibrin can accelerate hemostasis at it. In normal blood vessels the soluble fibrin has a several minutes` half-life. The results of the mice tail bleeding times indicated that 0.5 to 2.0 U/kg ip shortened 29.7 to 73.1% of the bleeding time. Because none of the mice were subjected to anesthesia, this might cause an increase of the experimental error values due to

the movement of the mice. The results of these in vivo experiments indicated that Agacutin did not have a significant impact on APTT, WBC, RBC, platelet count and function (release and aggregation activities), or ECT etc. (some results were not showed due to the text space limitations). In the study on the acute toxicity of Agacutin, the LD50 (half death rate) was not obtained due to the minimal toxicity in mice (iv). Analysis of the maximal tolerant dosage indicated that a dose of 40 U/kg, equivalent to 2400 folds of the clinical dosage of Reptilase, did not cause death in any of the twelve mice. Agacutin did not show hemorragin or neurotoxin activity (data not shown). All the results indicated that Agacutin may be a potential coagulant in clinic. Additional research will help us profoundly to understand Agacutin.

Acknowledgements This work was supported by Shenzhen Yihekang Biotech Co., Ltd, Guangdong Greatsun Biochemical Pharmaceutical Co., Ltd, and Kunming Longjin Pharmaceutical Co., Ltd. We thank them for their financial support and their equipment. The research work was also supported by Guangdong Pharmaceutical University Grant 2005SMK22 and KeyTeacher Training Grant.

References [1] Braud S, Bon C, Wisner A. Snake venom proteins acting on haemostasis. Biochimie 2000;82:851–9. [2] Matsui T, Fujimura Y, Titani K. Snake venom proteases affecting haemostasis and thrombosis. Biochim Biophys Acta 2000;1477(1-2):146–56. [3] Markland FS. Snake venoms and the hemostatic system. Toxicon 1998;36 (12):1749–800.

S.-S. Tang et al. / Thrombosis Research 124 (2009) 631–639 [4] Weon-Kyoo Y, Won-Seok C, You-Seok K, Hang-Cheol S, Yangsoo J, Kwang-Hoe C. Functional characterization of recombinant batroxobin, a snake venom thrombinlike enzyme, expressed from pichia pastoris. FEBS Letters 2004;57:167–72. [5] Mu P, Yin JR, Hou RP, Yu HR. Purification of thrombin-like enzyme from the snake venom of Agkistrodon halys (blomhoffii ussurensis) of Changbai mountain. J Jinzhou Med College, 2005;26(5):39–42. [6] Wang YN, Zuo CL, Chen MX, Hu JC. The haemostasis effect on experimental mice of RVV-X from venom of viper russellii in Guangxi. Journal of Snake 2006;18(1):6–10. [7] Chen YL, Tsai JH. Functional and sequence characterization of agkicetin, a new glycoprotein Ib antagonist isolated from Agkistrodon acutus venom. Biochem Biophys Res Commun 1995;210(2):472–7. [8] Yeh CH, Peng HC, Yih JB, Huang TF. A new short chain RGD-containing disintegrin, accutin, inhibits the common pathway of human platelet aggregation. Biochim Biophys Acta 1998;1425(3):493–504. [9] Huang QQ, Teng MK, Niu LW. Purification and characterization of two fibrinogenclotting enzymes from five-pace snake (Agkistrodon acutus) venom. Toxicon 1999;37(7):999–1013. [10] Cheng X, Xu ZY, Liu QD, Li XM, Li XY, Liu J. Purification and Characterization of a Platelet Agglutinating Inhibiting Protein (Agkisacutacin) from Agkistrodon acutus Venom. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao 2000;32(6):653–6. [11] Pan H, Du XY, Yang GZ, Zhou YC, Wu XF. CDNA Cloning and Expression of Acutin, a Thrombin-Like Enzyme from Agkistrodon acutus. Biochem Biophys Res Commun 1999;255:412–5. [12] Xu XL, Liu QL. Purification and Characterization of a Novel Anticlotting Factor (ACF II) from the Venom of Agkistrodon acutus. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao 2000;32(1):84–6.

639

[13] Yeh CH, Chang MC, Peng HC, Huang TF. Pharmacological characterization and antithrombotic effect of Agkistin, a platelet glycoprotein Ib antagonist. Br J Pharmacol 2001;132(4):843–50. [14] Liang XX, Chen LS, Zhou YN, Qiu PX, Yan GM. Purification and Biochemical Characterization of FIIa, a fibrinolytic enzyme from Agkistrodon acutus venom. Toxicon 2001;39(8):1133–9. [15] Li T, Zhu ZG, Yu CL, Wu SG. Purification, identification of acoagulatin, an anticlotting factor from the venom of Chinese Agkistrodon, and observation of its anticlotting effect. Di Yi Jun Yi Da Xue Xue Bao 2004;24(7):836–8. [16] Zhu Z, Gong P, Teng M, Niu L. Purification, N-terminal sequencing, partial characterization, crystallization and preliminary crystallographic analysis of two glycosylated serine proteinases from Agkistrodon acutus venom. Acta Crystallogr D Biol Crystallogr 2003;59:547–50. [17] Righetti P, Tudor G, Gianazza. Iso-electic Focusing: Theory, Methodology and Applications. Amsterdam: Elsevier Press; 1982. [18] Kornalik F. The influence of snake venom enzymes on blood clotting. Pharmac Ther 1985;29:353–405. [19] Wei C, Chen F. Long-term toxicity test of arginine esterase from Agkistrodon halys ussuriensis venom. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 2004;21(3):420–3. [20] Xu SY, Bian RL, Cheng X. Hemostatic drugs. In: Ma CG, Wang YS, Wang ZG, Bian RL, editors. Experimental Methodology in Pharmacology. 2nd ed. Beijing: People`s Healthy Press; 1994. p. 179–207. [21] GenBank accession no.1WT9 A [gi / 93278427], Chain A, Crystal Structure of Aa-XBp-I, A snake venom protein from Deinagkistrodon acutus snake venom.