Identification of novel serum proteins in a Japanese viper: Homologs of mammalian PSP94

Identification of novel serum proteins in a Japanese viper: Homologs of mammalian PSP94

Biochemical and Biophysical Research Communications 359 (2007) 330–334 www.elsevier.com/locate/ybbrc Identification of novel serum proteins in a Japan...

415KB Sizes 2 Downloads 56 Views

Biochemical and Biophysical Research Communications 359 (2007) 330–334 www.elsevier.com/locate/ybbrc

Identification of novel serum proteins in a Japanese viper: Homologs of mammalian PSP94 Narumi Aoki, Akie Sakiyama, Masanobu Deshimaru *, Shigeyuki Terada Department of Chemistry, Faculty of Science, Fukuoka University, 8-19-1 Nanakuma, Jonan-ku, Fukuoka 814-0180, Japan Received 9 May 2007 Available online 22 May 2007

Abstract Three small serum proteins (SSP-1, -2, and -3), with molecular masses of 6.5–10 kDa, were isolated from Habu (Trimeresurus flavoviridis) serum, and the amino acid sequences were determined by protein and cDNA analysis. Despite only limited sequence identity to any mammalian prostatic secretory protein of 94 amino acids (PSP94), all of the Cys residues in these SSPs were well conserved. SSPs are the first PSP94 family proteins to be identified in reptiles. SSP-1 and -3 weakly inhibited the proteolytic activity of a snake venom metalloproteinase. On the other hand, SSP-2 formed a tight complex with triflin, a snake venom-derived Ca2+ channel blocker that suppresses the smooth muscle contraction. This suggests a role for SSP-2 in the self defense system of venomous snakes.  2007 Elsevier Inc. All rights reserved. Keywords: CRISP; Metalloproteinase inhibitor; PSP94; Snake serum; Trimeresurus flavoviridis; Triflin

PSP94 (prostatic secretory protein of 94 amino acids), also called b-microseminoprotein or prostatic inhibin peptide, is a 10.7-kDa, nonglycosylated, and cysteine-rich protein [1]. Even though PSP94 was first isolated as a major protein from human seminal plasma [2], it was later found to be present at the same level in women as in men [3]. Along with roles in fertility [4,5], PSP94 has postulated systemic functions including growth regulation and induction of apoptosis in prostate cancer cells [6], as well as regulation of calcium levels in hypercalcemia of malignancy [7]. Expression of PSP94 decreases progressively during the development of prostate cancer [8,9]. Recently, it was reported that PSP94 binds to a protein in human blood (PSP94-binding protein) [10] and to human cysteine-rich secretory protein 3 (CRISP-3) [11]. CRISP family proteins are widely distributed in mammals, reptiles, and amphibians, and are involved in a variety of biological reactions. Crystallographic study of CRISPs has suggested that they are composed of two domains

*

Corresponding author. Fax: +81 92 865 6030. E-mail address: [email protected] (M. Deshimaru).

0006-291X/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.05.091

[12]. The N-terminal part of PSP94-binding protein is homologos to the N-terminal domain of CRISPs. PSP94 proteins have also been identified in other mammals [13–15] and in ostrich [16]. However, the amino acid sequences of these proteins are very different from that of the human protein. In addition to the 10 cysteines that form five disulfide bonds, there are only 16 identical amino acids in the mammalian proteins, suggesting that they evolved with few constraints and at a relatively rapid rate [17]. The proton NMR spectra of pig and human PSP94s have been assigned, and it was shown that these are b-sheet proteins composed of two domains [18,19]. Proteins that inhibit snake venom hemorrhagic metalloproteinases (MPs) have been isolated from the sera of certain snakes: for example, Habu serum factor (HSF) from Habu (Trimeresurus flavoviridis) [20] and BJ46a from Bothrops jararaca [21]. Recently, while purifying the similar anti-hemorrhagic factor from the serum of Chinese Mamushi viper (Gloydius blomhoffi brevicaudus), we obtained a novel 13-kDa protein that inhibits brevilysin H6 [22], a snake venom MP with weak hemorrhagic activity. We termed this protein small serum protein (SSP). Analysis of the N-terminal 50 residues indicated that SSP

N. Aoki et al. / Biochemical and Biophysical Research Communications 359 (2007) 330–334

belongs to the PSP94 protein family with respect to the topological similarity of cysteine residues, despite the low sequence homology. We also found four similar proteins (SSP-1, -2, and -3) in the serum of Habu. SSPs are the first PSP94 family proteins to be found in reptiles. Here, we report the properties and amino acid sequences of these SSPs. SSP-1 showed an inhibitory effect toward brevilysin H6; a slight inhibition of brevilysin H6 by SSP3 was also observed. With respect to the physiological role of SSPs, we show evidence that SSP-2 has a strong binding affinity for triflin, which belongs to the CRISP family and was isolated from T. flavoviridis venom as the smooth muscle contraction blocker [23]. This suggests that SSP-2 exists in snake blood as a self-defense material against the toxic effects of their own venom by an accidental envenomation. As far as we know, SSP-2 is the first protein that specifically binds to the ion channel modulator protein. Materials and methods Materials. Blood from T. flavoviridis from the Amami Oshima Islands was collected by decapitation. The serum was separated by centrifugation and stored at 20 C. Brevilysin H6 was prepared as described previously [22]. The hemorrhagic factors HR1A and HR1B were prepared as described [24]. Triflin was purified from T. flavoviridis venom as described [23]. All other reagents were purchased from Wako Pure Chemicals (Osaka). Separation of snake serum fraction. Serum was fractionated by cold ethanol into E0.75–E2.0 as described previously [25]. Fraction E2.0 was dissolved in a small amount of water and put on a lBondasphere 5l-C8˚ column (1.9 · 15 cm, Waters). Elution was carried out with a linear 300 A gradient of acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 5.0 ml/ min, and was monitored at 230 nm using a flow cell of 1 mm light path. Measurement of proteolytic activity. Proteolytic activity was measured in 5 mM CaCl2 50 mM Tris–HCl (pH 8.5) using fluorescein isothiocyanate-labeled casein as the substrate [26]. The increase in fluorescence was analyzed on an FP-550A spectrofluorometer (Jasco) at 520 nm with excitation at 490 nm. Electrophoresis. SDS–PAGE was carried out on 14% gels by Laemmli’s method [27], or on 16.5% polyacrylamide gels under reducing conditions [28]. Bovine serum albumin (67,000), ovalbumin (46,000), carbonic anhydrase (30,000), chymotrypsinogen (25,000), soybean trypsin inhibitor (21,500), lysozyme (14,300), aprotinin (6500), and insulin Bchain (3400) were used for molecular weight markers. After running the gels under a constant current, they were stained with 0.1% Coomassie brilliant blue R-250 and destained with 10% acetic acid. Native PAGE was performed in a manner analogous to that for SDS–PAGE, but using gels and a developing solvent without SDS [29]. Sequence analysis. Protein was S-pyridylethylated [30] and the amino acid sequence of the modified protein was determined by an automatic protein sequencer PPSQ 21 (Shimadzu). Affinity chromatography. Affinity adsorbent was prepared by reacting an N-hydroxysuccimide-activated HiTrap column (1 ml, GE Healthcare Bio-Science) with an equimolar mixture of the three SSPs (4.0 mg total) following the manufacturer’s instructions. Sample dissolved in 0.15 M NaCl, 20 mM phosphate buffer (pH 7.4) was loaded onto the resulting SSP-HiTrap column, and washed with the same buffer. Absorbed materials on the column were then eluted using 0.5 M NaCl 0.1 M glycine–HCl buffer (pH 3.0) and 1 ml fractions were collected. Western blotting assay with polyclonal anti-sera. Japanese White rabbits were immunized with an equimolar mixture of the three SSPs (100 lg total) or the purified triflin (250 lg) emulsified in an equal volume of complete Freund’s adjuvant. After 2 weeks, boosters of each antigen emulsified in incomplete Freund’s adjuvant were given by intradermal

331

injection; this was followed by another injection of peptide after a further 2 weeks. Anti-sera were collected 12 days after the last boost. For immunodetection of SSPs and triflin, proteins separated by native PAGE were electroblotted onto PVDF membranes, which were then blocked using skimmed milk and incubated for 16 h at 4 C with anti-SSPs or anti-triflin rabbit anti-serum diluted 1:400. Bound IgG antibodies were detected by use of horseradish peroxidase-labeled goat anti-rabbit IgG antibody diluted 1:5000 (Rockland Inc.). The target proteins were revealed by reacting blots with 0.02% 3,3 0 -diaminobenzidine containing 0.06% H2O2 in water. Synthesis of partial fragment of SSP-1 cDNA. Total RNA was extracted from 0.5 g of T. flavoviridis liver by acid guanidinium-phenol-chloroform (AGPC) method, and reverse transcribed to synthesize cDNA first strands using adaptor-linked oligo(dT) primer (5 0 -GGCCACGC GTCGACTAG TAC-(dT)17-3 0 ). cDNAs obtained were used as template for 3 0 -RACE (rapid amplification of cDNA ends) reaction. Synthetic oligonucleotides, SSP-1N (5 0 -CANGAYGCNATGGCNCCNAARAARTGYGT-3 0 ) and 3 0 -Adp (5 0 -GGCCACGCGTCGACTAGTAC-3 0 ), were used for PCR amplification. SSP-1N primer was designed upon N-terminal amino acid sequence of SSP-1 and 3 0 -Adp corresponded to the adaptor sequence within adaptor-linked oligo(dT) primer. Amplification product was once subcloned into plasmid vector, and its nucleotide sequence was determined. As the result, it was confirmed to be the cDNA partial fragment for SSP-1. This cDNA fragment was radiolabeled with [a-32P]dCTP (3000 Ci/mmol) using random primer DNA labeling kit (Takara Bio, Japan) and used for hybridization screening of cDNA library. Construction of T. flavoviridis liver cDNA Iibrary. T. flavoviridis liver cDNA library was constructed using Creator SMART cDNA Library Construction kit (BD Biosciences) according to manufacturer’s instruction. Briefly, cDNA first strand was synthesized using 1 lg of total RNA, followed by five cycles of PCR for non-specific enrichment of full-length cDNAs. The cDNA fragments were then ligated to pDNR-LIB vector. When the plasmid clones were used to transform Escherichia coli JM109, resulting library contained 2.0 · 106 independent clones. Cloning and sequence determination of cDNAs encoding SSPs. Clones (2 · 105) from unamplified cDNA library were plated on LB agar plates and bacterial colonies were transferred onto Hybond-NX membranes (GE Healthcare Bio-Science, USA) and fixed by UV irradiation. The resulted replica membranes were prehybridized in Church’s hybridization solution at 65 C for 30 min and then hybridized with radiolabeled SSP-1 cDNA overnight at 50 C in Church’s hybridization solution. Membranes were finally washed twice for 15 min at 50 C with 1· SSC, 0.1% SDS and hybridization signals were visualized using BioImage Analyzer (Fuji Film, Japan). Accordingly, 20 bacterial colonies were isolated, cultured and their plasmids were purified by standard Alkali-SDS method. Nucleotide sequences of cDNA inserts were determined using ABI PRISM 377 DNA Sequencing System (Applied Biosystems, CA, USA).

Results Purification and sequence determination of SSPs Serum from T. flavoviridis was ethanol precipitated [25], and proteins precipitated by a two-fold excess of ethanol (fraction E2.0) were fractionated by reverse phase HPLC. Although this fraction contained no anti-hemorrhagic factor, significant MP-inhibitory activity was detected in peak 1 (Fig. 1). The apparent molecular mass of peak 1 was 10 kDa as estimated by SDS–PAGE (Fig. 1). Since peaks 2 and 3 also had a low molecular mass, the protein species which represent peaks 1, 2, and 3 was designated SSP-1, -2, and -3, respectively. The molecular masses of SSP-1, -2, and -3 were determined to be 10,059.3, 9916.7, and 6552.7 kDa, respectively, by time-of-flight mass spectrometry (data not

332

N. Aoki et al. / Biochemical and Biophysical Research Communications 359 (2007) 330–334

composed of 88 and 90 amino acids, respectively, while SSP-3 comprises only 60 residues (Fig. 2). When the primary structures of SSP-1 and -2 were compared with those of mammalian PSP94 family proteins, a similar topology of 10 Cys residues was found between them, irrespective of the limited sequence similarities. SSP-3 is unique among PSP94 family proteins, as it lacks the C-terminal 30 residues of this family. It has only six cysteines, and the third Cys residue, found in all the other proteins at position 38, is substituted by Ser. Inhibitory effect of SSPs on snake venom MP

Fig. 1. Separation of three SSPs by HPLC. Acetonitrile gradient of 20– 50% is shown in a dashed line. SDS–PAGE analysis of HPLC-purified SSPs is shown in the inset. Lanes 1–3 are SSP-1, -2, and -3, respectively. M, molecular weight markers.

shown). The sequence of N-terminal residues was determined for each pyridylethylated protein, as indicated by the underlined sequences in Fig. 2. To elucidate the entire amino acid sequences of three SSPs, the respective cDNA clones were isolated from T. flavoviridis liver cDNA library. Based on the N-terminal amino acid sequence of SSP-1, a 5 0 -truncated cDNA fragment for SSP-1 was amplified by 3 0 -RACE method. This cDNA fragment was then 32P-labeled and employed as the probe for screening of the cDNA library. Probe-hybridized replica membranes were washed under a low stringency condition (at 50 C with 1· SSC, 0.1% SDS) and plasmid clones with various signal strengths were isolated in order to cover the clones for all three SSPs. Finally, 12 cDNA clones encoding SSPs were obtained; two SSP-1 clones, nine SSP-2 clones, and a single SSP-3 clone. According to their nucleotide sequences, SSP-1 and -2 are

When the MP inhibitory activity of these proteins was measured using brevilysin H6, an enzyme isolated from G. blomhoffi brevicaudus venom, SSP-1 and -3 both showed weak but significant inhibition (Fig. 3). However, neither protein was inhibitory towards other MPs, such as thermolysin, HR1A, and HR1B [24] (data not shown). The proteolytic activities of trypsin, chymotrypsin, and papain were also not influenced. On the other hand, SSP-2 showed no inhibitory activity towards brevilysin H6, even at high concentrations.

Fig. 3. Metalloproteinase inhibitory activities of SSPs. Brevilysin H6 (0.3 lM) was used as the enzyme. Data are the average values of three independent experiments.

Fig. 2. Sequence alignment of SSPs with human, pig, and mouse PSP94 proteins (GenBank Accession Nos. AJ13356, S41663, and J89840, respectively). The residue numbering is based on the sequence of human PSP94. Amino acid sequences determined by direct analysis of the pyridylethylated-proteins are underlined. The conserved cysteine residues are marked by asterisks.

N. Aoki et al. / Biochemical and Biophysical Research Communications 359 (2007) 330–334

Identification of the physiological target of SSPs Since SSPs and brevilysin H6 are present in different animals, H6 can not be a physiological target of SSP. To determine the true target of the SSPs, affinity chromatography with an SSP-HiTrap column was used. To find targets for all SSPs, a mixture of the three SSPs was used as the ligand. A 25-kDa protein remained specifically bound on the column when crude T. flavoviridis venom was applied to the column (data not shown). The exact mass, 24,713.8, and the sequence of the N-terminal 30 residues (NVDFDSESPRKPEIQNEIIDLHNSLRRSVN), have shown this protein to be triflin, a toxic protein in Habu venom that blocks smooth muscle contraction [23]. In order to determine which SSP or SSPs bind to triflin, a small excess of SSP was added to triflin, and complex formation was analyzed by native PAGE. Triflin alone migrated as two bands. Neither SSP-1 nor SSP-3 interacted with triflin (Fig. 4A). On the other hand, SSP-2 bound to triflin resulting in the appearance of new bands. When increasing amounts of SSP-2 were mixed with a constant amount of triflin, bands corresponding to unbound triflin disappeared and new bands appeared, as shown in Fig. 4B. The stoichiometry of SSP-2 and triflin was close to 1:1. A Western blotting assay of the mixture of SSP-2

Fig. 4. Native PAGE analysis of the binding of SSPs with triflin. (A) A small excess of SSP-1 or -3 was added to triflin and analyzed. (B) Complex formation of SSP-2 with triflin. To a constant amount of triflin were added several amount of SSP-2: 0, 0.4, 0.8, 1.2, 1.6, 2.0, and 2.4 molar equivalents were used in lanes 1–7, respectively. Arrows indicate bands of the complex. (C) Immunochemical staining of SSP-2–triflin complex. Equimolar mixture of SSP-2 and triflin was analyzed. After blotted on PVDF membrane, the bands were stained with anti-SSP and anti-triflin antibodies.

333

and triflin showed that both proteins bind tightly to form a complex (Fig. 4C). Two new bands in Fig. 4C could be stained by both anti-SSP and anti-triflin antibodies. Furthermore, the major band of the complex was transferred to PVDF membrane and subjected to a protein sequencer, giving two N-terminal sequences, AXGIG and NVDFD, which correspond to those for SSP-2 and triflin, respectively. Discussion Sequence comparisons show that SSP-1, -2, and -3 are approximately 23%, 27%, and 30% identical to human PSP94, respectively (Fig. 2). However, all of the cysteine residues are conserved in these proteins, suggesting that the three-dimensional conformations of the snake and the human proteins may be similar. Genes encoding several PSP94 proteins are known to have evolved rapidly [18]. The limited sequence identity of these three SSPs with mammalian PSP94s indicates that SSPs may also have evolved at a relatively fast rate. Cloning experiments for these three SSPs are now in progress to prove this idea. Human PSP94 binds to CRISP-3, an abundant constituent of human seminal plasma, and forms a stable, noncovalent complex [11]. In addition, human PSP94 seems to exist in serum in a complex with a PSP94-binding protein [10]. The structural similarity of the N-terminal domains of CRISP-3 and PSP94-binding protein suggests that they both bind PSP94 through their N-terminal domains. Complex formation by SSP-2 and triflin may be reasonable since the latter is also a CRISP family protein. We have not yet determined whether or not SSP-2 suppresses the biological activity of triflin. If so, SSP-2 is an anti-toxic protein that inhibits the neurotoxin-like activity of snake venom CRISP. All SSPs seem to be present in blood as high molecular weight forms, like human PSP94. No SSP was recovered in the ultrafiltrate of Habu serum using a membrane with a molecular weight cut-off of 30,000, implicating complex formation of SSPs among themselves or with other proteins (data not shown). The native states of SSPs in blood are now being elucidated. The triflin we purified from T. flavoviridis venom seemed to be homogenous because it migrated as a single band on SDS–PAGE. However, two bands were observed on native PAGE (Fig. 4). The minor band may represent a novel isoform of triflin, since it reacted with anti-triflin antibody as well as bound to SSP-2 (Fig. 4B and C). In cloning experiments targeting the cDNA encoding triflin, we obtained a cDNA clone that codes Asn110-Ala-Val-Val114 instead of the reported sequence Asp110-Ala-Val-Ile114 [23] (unpublished). At present, the physiological functions of Habu SSP-1 and -3 remain uncertain. Inhibition of brevilysin H6 by these SSPs was weak and apparently non-stoichiometric. Furthermore, SSP-1 did not inhibit any hemorrhagic MPs isolated from Habu venom. SSP-1 is the most abundant of the three SSPs and is present in the sera of both

334

N. Aoki et al. / Biochemical and Biophysical Research Communications 359 (2007) 330–334

sexes (data not shown). It may play a role in fertility like PSP94 [4,5]. References [1] J.Y. Dube, G. Frenette, R. Paquin, P. Chapdelaine, J. Tremblay, R.R. Tremblay, C. Lazure, N. Seidah, M.J. Chretien, Isolation from human seminal plasma of an abundant 16-kDa protein originating from the prostate, its identification with a 94-residue peptide originally described as b-inhibin, Androl. 8 (1987) 182–189. [2] H. Lilja, P.A. Abrahamsson, Three predominant proteins secreted by the human prostate gland, Prostate 12 (1988) 29–38. [3] H. von der Kammer, E. Krauhs, G. Aumuller, K.H. Scheit, Characterization of a monoclonal antibody specific for prostatic secretory protein of 94 amino acids (PSP94) and development of a two-site binding enzyme immunoassay for PSP94, Clin. Chim. Acta. 187 (1990) 207–219. [4] A.R. Sheth, N. Arabatti, M. Carlquist, H. Jornvall, Characterization of a polypeptide from human seminal plasma with inhibin (inhibition of FSH secretion)-like activity, FEBS Lett. 165 (1984) 11–15. [5] C.F. Chao, S.T. Chiou, H. Jeng, W.C. Chang, The porcine sperm motility inhibitor is identical to b-microseminoprotein and is a competitive inhibitor of Na+, K+ -ATPase, Biochem. Biophys. Res. Commun. 218 (1996) 623–628. [6] S.V. Garde, V.S. Basrur, L. Li, M.A. Finkelman, A. Krishan, L. Wellham, E. Ben-Josef, M. Haddad, J.D. Taylor, A.T. Porter, D.G. Tang, Prostate secretory protein (PSP94) suppresses the growth of androgen-independent prostate cell line (PC3) and xenografts by inducing apoptosis, Prostate 38 (1999) 118–125. [7] N. Shukeir, A. Arakelian, S. Kadhim, S. Garde, S.A. Rabbani, Prostate secretory protein (PSP-94) decreases tumor growth and hypercalcemia of malignancy in a syngenic in vivo model of prostate cancer, Cancer Res. 63 (2003) 2072–2078. [8] P.S. Chan, L.W. Chan, J.W. Xuan, J.L. Chin, H.L. Choi, F.L. Chan, In situ hybridization study of PSP94 (prostatic secretory protein of 94 amino acids) expression in human prostates, Prostate 41 (1999) 99–109. [9] M. Stanbrough, G.J. Bubley, K. Ross, T.R. Golub, M.A. Rubin, T.M. Penning, Increased expression of genes converting adrenal androgens to testosterone in androgen- independent prostate cancer, Cancer Res. 66 (2006) 2815–2825. [10] J.R. Reeves, J.W. Xuan, K. Arfanis, C. Morin, S.V. Garde, M.T. Ruiz, J. Wisniewski, C. Panchal, J.E. Tanner, Identification, purification and characterization of a novel human blood protein with binding affinity for prostate secretory protein of 94 amino acids, Biochem. J. 385 (2005) 105–114. [11] L. Udby, A. Lundwall, A.H. Johnsen, P. Fernlund, C. ValtonenAndre, A.M. Blom, H. Lilja, N. Borregaard, L. Kjeldsen, A. Bjartell, (-Microseminoprotein binds CRISP-3 in human seminal plasma, Biochem. Biophys. Res. Commun. 333 (2005) 555–561. [12] Y. Shikamoto, K. Suto, Y. Yamazaki, T. Morita, H. Mizuno, Crystal structure of a CRISP family Ca2+-channel blocker derived from snake venom, J. Mol. Biol. 350 (2005) 735–743. [13] P. Fernlund, L.B. Granberg, P. Roepstor, Amino acid sequence of bmicroseminoprotein from porcine seminal plasma, Arch. Biochem. Biophys. 309 (1994) 70–76.

[14] J.W. Xuan, J. Kwong, F.L. Chan, M. Ricci, Y. Imasato, H. Sakai, G.H. Fong, C. Panchal, J.L. Chin, cDNA, genomic cloning, and gene expression analysis of mouse PSP94 (prostate secretory protein of 94 amino acids), DNA Cell Biol. 18 (1999) 26. [15] M. Makinen, C. Valtonen-Andre, A. Lundwall, New world, but not Old World, monkeys carry several genes encoding b-microseminoprotein, Eur. J. Biochem. 264 (1999) 407–414. [16] C. Lazure, M. Villemure, D. Gauthier, R.J. Naude, M. Mbikay, Characterization of ostrich (Struthio camelus) b-microseminoprotein (MSP): identification of homologous sequences in EST databases and analysis of their evolution during speciation, Protein Sci. 10 (2001) 2207–2218. [17] S. Nolet, D. St-Louis, M. Mbikay, M. Chretien, Rapid evolution of prostatic protein PSP94 suggested by sequence divergence between rhesus monkey and human cDNAs, Genomics 9 (1991) 775–777. [18] I. Wang, Y.C. Lou, K.P. Wu, S.H. Wu, W.C. Chang, C. Chen, Novel solution structure of porcine b-microseminoprotein, J. Mol. Biol. 346 (2005) 1071–1082. [19] H. Ghasriani, K. Teilum, Y. Johnsson, P. Fernlund, T. Drakenberg, Solution structures of human and porcine b-microseminoprotein, J. Mol. Biol. 362 (2006) 502–515. [20] Y. Yamakawa, T. Omori-Satoh, Primary structure of the antihemorrhagic factor in serum of the Japanese Habu: a snake venom metalloproteinase inhibitor with a double headed cystatin domain, J. Biochem. 112 (1992) 583–589. [21] R.H. Valente, B. Dragulev, J. Perales, J.W. Fox, G.B. Domont, BJ46a, a snake venom metalloproteinase inhibitor. Isolation, characterization, cloning and insights into its mechanism of action, Eur. J. Biochem. 268 (2001) 3042–3052. [22] S. Fujimura, K. Oshikawa, S. Terada, E. Kimoto, Primary structure and autoproteolysis of brevilysin H6 from the venom of Gloydius halys brevicaudus, J. Biochem. 128 (2000) 167–173. [23] Y. Yamazaki, H. Koike, Y. Sugiyama, K. Motoyoshi, T. Wada, S. Hishinuma, M. Mita, T. Morita, Cloning and characterization of novel snake venom proteins that block smooth muscle contraction, Eur. J. Biochem. 269 (2002) 2708–2715. [24] T. Omori-Satoh, S. Sadahiro, Resolution of the major hemorrhagic component of Trimeresurus flavoviridis venom into two parts, Biochim. Biophys. Acta 580 (1979) 392–404. [25] M. Deshimaru, C. Tanaka, A. Tokunaga, M. Goto, S. Terada, Efficient method for purification of antihemorrhagic factor (HSF) in the serum of Japanese Habu (Trimeresurus flavoviridis), Fukuoka Univ. Sci. Rep. 33 (2003) 45–53. [26] S.S. Twining, Fluorescein isothiocyanate-labeled casein assay for proteolytic enzymes, Anal. Biochem. 143 (1984) 30–34. [27] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680–685. [28] H. Schagger, G. von Jagow, Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa, Anal. Biochem. 166 (1987) 368–379. [29] B.J. Davis, Disc Electrophoresis II. Method and Application to Human Serum Proteins, Ann. N. Y. Acad. Sci. 121 (1964) 404–427. [30] M. Friedman, L.H. Krull, J.F. Cavins, The chromatographic determination of cystine and cysteine residues in proteins as S-b-(4pyridylethyl)cysteine, J. Biol. Chem. 245 (1970) 3868–3871.