- Email: [email protected]
Molecular cloning and expression of structural domains of bothropasin, a P-III metalloproteinase from the venom of Bothrops jararaca
Toxicon 41 (2003) 217–227 www.elsevier.com/locate/toxicon
Molecular cloning and expression of structural domains of bothropasin, a P-III metalloproteinase from the venom of Bothrops jararaca Marina T. Assakuraa, Carlos A. Silvaa, Reinhard Menteleb, Antonio C.M. Camargoa, Solange M.T. Serranoa,* a
Laborato´rio de Bioquı´mica e Biofı´sica, Instituto Butantan, Av. Vital Brasil 1500, CEP 05503-900 Sa˜o Paulo, SP, Brazil b Department of Clinical Chemistry and Clinical Biochemistry, Klinikum Innenstadt, Universita¨t Mu¨nchen, D-80336 Mu¨nchen, Germany Received 31 July 2002; accepted 3 September 2002
Abstract Mature P-III snake metalloproteinases are soluble venom components which belong to the Reprolysin sub family and are structurally related to the mammalian membrane-bound A Disintegrin And Metalloproteinase (ADAMs). Here we present the molecular cloning of bothropasin, a metalloproteinase with hemorrhagic and myonecrotic activities isolated from the venom of Bothrops jararaca. The full-length cDNA encoding the bothropasin precursor was cloned by immunoscreening and its authenticity was confirmed by the amino acid sequence of internal fragments obtained from an autolyzed sample of native bothropasin. The predicted bothropasin precursor is comprised of the elements of a P-III venom metalloproteinase: signal sequence, pro-, metalloproteinase, disintegrin-like and cysteine-rich domains. In the autolysis process of native bothropasin, the disintegrin-like and cysteine-rich domains remained intact while the metalloproteinase domain was cleaved at different sites. The attempts made to obtain the recombinant precursor form of bothropasin using bacterial, yeast and mammalian cell expression systems failed to produce it in an amount sufficient to analyze the activation of the zymogen. Nevertheless, the study of the expression of the individual domains of bothropasin using a bacterial system resulted in the production of recombinant pro-and disintegrin-like þ cysteine-rich domains but not the metalloproteinase domain. These results along with the autolysis pattern of the native protein suggest a role for the metalloproteinase domain in the structural stability of bothropasin. q 2002 Elsevier Science Ltd. All rights reserved. Keywords: Metalloproteinase; Disintegrin; Autolysis; Recombinant protein expression
1. Introduction Envenomation with snake venoms of the families Crotalidae and Viperidae produces pronounced local hemorrhage and muscle damage. These venoms are a rich source of active zinc metalloproteinases, which have been extensively investigated (Hite et al., 1992; Paine et al., 1992; Takeya et al., 1993; Bjarnason and Fox, 1995). The metalloproteinases isolated from snake venoms are mem* Corresponding author. Fax: þ 55-11-3726-1024. E-mail address: [email protected] (S.M.T. Serrano).
bers of the Reprolysin subfamily of enzymes, which also includes the A Disintegrin And Metalloproteinase (ADAMs). They are secreted as preproenzymes and contain additional regulatory modules, which are responsible for interactions with the extracellular matrix and integrins. The mature P-I class proteins have only a proteinase domain, whereas the P-II, P-III, and P-IV classes have disintegrin or disintegrin-like, cysteine-rich, and lectin-like domains found carboxy to the proteinase domain, respectively. Similar domain structures are found in the ADAMs, which additionally contain an epidermal growth factor-like domain, a transmembrane region and a cytoplasmic tail (Black and White, 1998).
0041-0101/03/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 0 2 ) 0 0 2 7 9 - 9
218
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
The venom of Bothrops jararaca contains hemorrhagic toxins that cause disruption of the basement membrane of the vascular endothelium resulting in bleeding (Kamiguti et al., 1991). At present, five metalloproteinases with hemorrhagic activity have been isolated from this venom. HF1, HF2 and HF3 are highly active hemorrhagic toxins (Mandelbaum and Assakura, 1988). Jararhagin is another hemorrhagic proteinase of 52 kDa, which has been extensively analyzed and shown to cause inhibition of collageninduced platelet aggregation (Paine et al., 1992; Kamiguti et al., 1996; Moura-da-Silva et al., 1999). Bothropasin is a single chain metalloproteinase of 48 kDa isolated from the same venom, whose caseinolytic activity corresponds to , 15% of the total activity of the crude venom (Mandelbaum et al., 1982). It causes hemorrhage on rabbit skin with a MHD of 1 mg. The same dose causes myonecrosis in mouse tibialis anterior muscle. On the B-chain of oxidized insulin, bothropasin shows specificity for leucine and phenylalanine bonds. Polyclonal antibodies raised in rabbits against bothropasin neither give a precipitin line with hemorrhagic factors (HF1 and HF2) nor neutralize their hemorrhagic activity (Mandelbaum and Assakura, 1988). This specific antiserum was used to screen a B. jararaca venom gland cDNA expression library in order to identify putative clones encoding bothropasin. Metalloproteinases of the Reprolysin subfamily are synthesized as zymogens, which are proteolytically activated by a mechanism where the N-terminal propeptide is cleaved off (Stocker et al., 1995). The in vitro activation/processing of the recombinant zymogen form of the Crotalus atrox P-II class metalloproteinase atrolysin E by the venom enzymes atrolysin A and atrolysin E itself was reported (Shimokawa et al., 1996). However, the pro-domain of venom proteinases also contains a conserved pair of basic residues K180 – K181, which might be cleaved immediately downstream by a proprotein convertase-dependent pathway in the gland tissue leading to the activation of the zymogen as it is the case of the matrix metalloproteinases (Nagase and Woessner, 1999) and the ADAMs (Primakoff and Myles, 2000). To investigate the ability of a proconvertase to process the precursor of bothropasin we attempted to produce the recombinant proprotein using different expression systems. However, due to the very low expression level obtained from the pro-protein we tested the expression of its individual domains in order to gain some knowledge about the structure features that might be important for the successful expression of snake venom PIII class metalloproteinases.
2. Materials and methods 2.1. Molecular cloning of bothropasin A B. jararaca cDNA expression library was constructed in lZAPII from the venom gland poly(A)þ RNA using a cDNA
synthesis kit (Stratagene) according to the manufacturer’s instructions. Total RNA was extracted from venom glands of five specimens of B. jararaca with RNAzol B (WAKChemie). Double-stranded cDNA was synthesized using 5 mg of poly(A)þ RNA template, ligated to Eco RI – Xho Idigested lZAPII DNA, and packaged. The library contained over 106 independent phages. Anti-bothropasin antibodies were prepared as described elsewhere (Mandelbaum and Assakura, 1988). Absorption of anti-Escherichia coli and anti-l phage antibodies was performed as described in Sambrook et al. (1989). The library was plated on E. coli XL1-Blue MRF’ and grown at 37 8C for 8 h. Recombinant protein expression was induced with 10 mM isopropyl-1-thio-b-D -galactopyranoside saturated nitrocellulose (Schleischer and Schuell) and the filters were processed for immunostaining with antibothropasin antibodies, as described in Sambrook et al. (1989). Plasmids were excised from lZAPII vector and recircularized in the presence of ExAssiste (Stratagene) helper phage to form phagemid pBluescript. Plasmids containing longest inserts were selected for sequencing of both strands in an ABI Prism 310 Genetic Analyzer (Applied Biosystems) using an ABI PRISM Dye Terminator Cycle Sequencing kit (Perkin Elmer) and oligonucleotide primers to internal sequence to obtain overlapping sequence information. The sequence obtained was deposited to the GenBank database (accession AF056025). 2.2. Expression of pro-bothropasin in E. coli The expression plasmid pGEX-4T2-BPMDC, that contains cDNA encoding amino acid residues 13 – 610 and includes pro, metalloproteinase, disintegrin-like and cysteine-rich domains was constructed by amplification of the cDNA fragment by PCR and subcloning into pGEMe-T vector (Promega). The plasmid was sequenced on both strands to ensure that the coding sequence was correct. To generate a fusion protein with glutathione S-transferase (GST), the insert from the pGEMe-T vector was isolated as a Bam H I– Xho I fragment and cloned into pGEX-4T (Amersham Pharmacia). The constructed plasmid was transformed into E. coli strain DH5a, and grown in 50 ml Luria –Bertani medium containing 100 mg ml21 ampicillin at 37 8C to a cell density of A600 ¼ 0.6 – 0.8. The expression of recombinant protein was induced by adding concentrations ranging from 0.3 to 1.5 mM isopropyl thio-b-D galactopyranoside (IPTG) and incubation was continued for 3 h at 30 8C. Cells were processed as described below for the expression of individual domains of bothropasin. The expression plasmid pET17b-BPMDC containing cDNA encoding residues 13 – 610 corresponding to the sequence of pro-bothropasin was constructed by subcloning the insert from pGEMe-T vector (see above) into pET17b (Novagen) previously digested with Bam H I – Xho I. The plasmid was transformed into both E. coli XL1BlueMRF’ and BL21(DE3)pLysS strains, and grown for induction of
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
protein expression as described earlier. After cell lysing by sonication in lysis buffer (50 mM Tris – HCl, pH 8.0, 2 mM EDTA, 1% Triton X-100, 10 mg ml21 lyzozyme) inclusion bodies were recovered by centrifugation, washed four times with 50 mM Tris –HCl, pH 8.0 containing 2 mM EDTA and submitted to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS –PAGE) and Western blot analysis. 2.3. Expression of pro-bothropasin in P. pastoris A cDNA fragment encoding amino acid residues 13 – 610 corresponding to the sequence of pro-bothropasin was amplified by PCR incorporating nucleotide sequences of Sna B I and Bam H I (sense) and Not I and Bam H I (antisense). The purified PCR product was ligated into the pGEM-11Zf vector. The insert from the pGEM-11Zf vector was isolated and cloned into pPIC9K (Invitrogen). The construct, named pPIC9K-BPMDC, was transformed into the E. coli strain XL1Blue MRF’ and sequenced prior to transformation into yeast. Following the identification of Hisþ transformants of P. pastoris, the screening for clones expressing pro-bothropasin was performed by growing 20 ml in buffered glycerol medium (BMGY) with vigorous shaking at 30 8C in loosely covered flasks for 2 days, and resuspending the cell pellets in 2 ml buffered methanol medium (BMMC) or non-buffered methanol medium (MMC), and growing under the same conditions for 5 days. Media were clarified by centrifugation and subjected to SDS – PAGE and Western blot analysis to assess the level of expression. 2.4. Expression of pro-bothropasin in COS-7 cells A cDNA fragment corresponding to amino acid residues 1 – 610 corresponding to the complete coding sequence of the bothropasin precursor was amplified by PCR incorporating nucleotide sequences of Bam H I (sense) and Xho I (antisense). The purified PCR product was ligated into the pGEMe-T vector. The insert from the pGEMe-T vector was isolated and cloned into the expression vector pcDNA3 (Invitrogen). The construct, named pcDNA3-BPMDC, was transformed into the E. coli strain XL1Blue MRF’ and sequenced prior to transfection into COS-7 cells. For transient expression of the precursor of bothropasin, monolayers of 106 COS-7 cells grown in petri dishes of 10 cm were transfected with 10 mg of pcDNA3-BPMDC using Exgen 500 (Euromedex) as a gene-delivering agent. Eighteen hours after transfection, 4 ml of 2xDulbecco’s modified Eagle’s medium (DMEM) containing 20% Nuserume (Becton Dickinson) were added to each plate containing the same volume of DMEM and incubation was prolonged for 24 h at 37 8C. The medium from each plate was substituted by 1 £ DMEM containing 10% Nu-serume and incubation was prolonged for another 48 h period. Cells were harvested with phosphate buffered saline containing 0.5 mM EDTA and cell lysates were obtained by treating cells with
219
40 mM Tris – HCl, pH 7.5, 40 mM MgCl2, 1% Triton X-100. The content of cells and the media were analyzed by SDS – PAGE and Western blot for protein expression. 2.5. Expression of bothropasin individual domains in E. coli Three separate expression plasmids were constructed: pGEX-4T1-BPM, that includes cDNA encoding residues 13 – 394 and encodes the pro and metalloproteinase domains; pGEX-4T1-BP that contains cDNA encoding residues 13 – 189 and has the pro domain only; and pGEX4T1-BDC that contains cDNA encoding residues 395– 610 and includes the disintegrin-like and Cys-rich domains. The bothropasin cDNA fragments were amplified by PCR and subcloned into pGEMe-T vector (Promega). The plasmids were sequenced on both strands to ensure that their coding sequences were correct. To generate fusion proteins with GST, the inserts from the pGEMe-T vector were isolated as Bam H I – Xho I fragments and cloned into pGEX-4T (Amersham Pharmacia). Each of the constructed plasmids was transformed into E. coli strain DH5a, and grown as described earlier for induction of protein expression in E. coli. Cells were collected by centrifugation at 4000 g for 5 min and suspended in 1 ml lysis buffer (50 mM Tris – HCl pH 7.5, 150 mM NaCl, 1% Triton X-100). After lysis by sonication cell lysate was centrifuged and 80 ml of glutathione-Sepharose suspension were added to supernatant and incubated for 16 h at 4 8C. The resin was washed successively with lysis buffer to eliminate unbound proteins and incubated with thrombin cleavage buffer (50 mM Tris – HCl pH 8.0, 150 mM NaCl, 2.5 mM CaCl2) and 4U thrombin (Sigma) for 2 h at 37 8C. The recombinant protein obtained in the supernatant was analyzed by SDS – PAGE, Western blot analysis and N-terminal amino acid sequencing. 2.6. Analytical procedures Protein concentrations were determined by the Bio-Rad protein assay kit using bovine serum albumin as a standard. SDS –PAGE was carried out according to Laemmli (1970). For Western blot, the protein samples were transferred to nitrocellulose membrane and developed as described earlier for immunoscreening. Fragments originated from autolysed bothropasin were isolated by RP-HPLC on a RP18 column (Reprosil, 1 £ 150 mm, Maisch). N-terminal amino acid sequencing of peptides originated from the autolysis of native bothropasin, of the internal peptide obtained by digestion of the 68 kDa fusion protein GST-BPM with endoproteinase Lys-C, and of protein bands transferred to PVDF membrane (Bio-Rad) was performed by Edman degradation using a 473A sequencer (Applied Biosystems) following manufacturer’s operating protocols. Platelet suspensions were prepared from fresh citrated human blood obtained from healthy donors, according to Mustard et al. (1972). The native and recombinant proteins
220
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
were incubated with platelet suspensions containing 3 £ 108 ml21 for 5 min at 37 8C before the addition of 0.5 mg ml21 collagen (Sigma). The extent of aggregation was estimated as the percentage increase of light transmission taking the value obtained for platelets incubated with control buffer and stimulated with collagen as 100%.
3. Results 3.1. Cloning of a cDNA encoding bothropasin and sequence analysis Several immunoreactive clones were isolated from the cDNA bank and those with largest inserts were selected for further studies. Fig. 1 shows the full-length nucleotide and the deduced amino acid sequences of the bothropasin precursor encoded by clone 8 £ 3.3. The cDNA consists of 2334 bases including a 73 bases 50 -end non-coding region, an open reading frame of 1830 bases, and a 30 -end non-coding region of 431 bases with termination codon, AATAAA sequence and polyadenylation sites. The cDNA sequence of bothropasin predicts a pro-protein with striking similarities to precursors of the Reprolysin subfamily, belonging to class P-III (Bjarnason and Fox, 1995). Native bothropasin undergoes autolysis under storage conditions at 4 8C in 50 mM Tris – HCl pH 7.5, containing 0.5 mM CaCl2, originating main fragments of 40 –30 kDa which are detected after 7 days by SDS – PAGE and Western blot analysis (Fig. 3(A) and (B), lane 3). Autolyzed bothropasin was submitted to RP-HPLC on a C18 column to isolate fragments and the N-terminal sequence of five internal peptides confirmed the authenticity of the isolated cDNA (Fig. 1). The N-terminal sequence of the fragment that starts at position 395 (L –G – T – D– I…) and its molecular mass (30 kDa) indicate that it corresponds to the intact disintegrin-like and Cys-rich domains of bothropasin. The other four peptides sequenced derived from the metalloproteinase domain. The molecular mass of the predicted mature protein of 46.6 kDa is lower than the value of 48 kDa estimated by SDS – PAGE (Mandelbaum et al., 1982) probably due to carbohydrate moieties. One putative N-glycosylation site was detected at Asn372 located in the proteinase domain (Fig. 1) of the bothropasin precursor. Fig. 2 shows the alignment of the predicted amino acid sequence of bothropasin with other venom metalloproteinase precursors of the Reprolysin subfamily. The putative prodomain contains the conserved sequence P– K –M – C– G – V– T present in the correspondent positions of other venom metalloproteinase precursors and which resembles a region in the propeptide of matrixins called Cys-switch which is involved in the activation of these enzymes (Nagase and Woessner, 1999). Bothropasin exhibits the characteristic elongated consensus motif HEXXHXXGXXH, whose three histidine residues are involved in binding of the catalytic
essential zinc ion (Murphy et al., 1991). Moreover, the same domain contains a conserved methionine residue, which in the metzincins is located beneath the active site metal as part of a super imposable ‘Met-turn’ (Stocker et al., 1995). In the disintegrin-like domain of bothropasin, the sequence SECD replaces the RGD sequence that is frequently found in the disintegrin domains of the P-II class precursors (Fig. 2). The 28 Cys residues found in the disintegrin-like/cysteine-rich domains of bothropasin are conserved among P-III metalloproteinases. 3.2. Expression of pro-bothropasin The cDNA-coding region for pro-bothropasin (BPMDC) was produced by PCR and subcloned into different expression vectors. Initially, we tried the expression of the precursor as a soluble cytoplasmic protein in fusion with GST by transforming E. coli DH5a cells with the expression vector pGEX-4T2BPMDC. However, the cells induced with IPTG concentrations ranging from 0.3 to 1.5 mM IPTG failed to express the recombinant protein (not shown). In another attempt the expression vector pET17b-BPMDC was used to transform the E. coli strains XL1BlueMRF’ and BL21(DE3)pLysS. Bacterial cells of both strains induced with 0.5 mM IPTG failed to produce the expected 70 kDa recombinant precursor (not shown). However, the analysis of proteins produced as inclusion bodies by BL21(DE3)pLysS showed three protein bands of , 68, 40 and 30 kDa which were detected by Western blot using the anti-bothropasin antibodies (Fig. 3(A) and (B), lane 5). This suggests that the recombinant protein either underwent autolysis or was degraded by bacterial proteinases. Various attempts made to get information on the amino acid sequence of these protein bands failed because the recombinant protein was present in very low amount and mixed with bacterial proteins. Furthermore, the use of a cocktail of proteinase inhibitors (1 mM PMSF, 5 mM EDTA, 20 mg/ml pepstatin A, 20 mg/ml leupeptin) during the steps of sample processing for protein electrophoresis did not prevent the hydrolysis of the recombinant protein (not shown). Alternatively to the expression using procaryotic cells, we tested the expression of the precursor by two eucaryotic systems. In case of mammalian COS-7 cells, the analysis of media samples and cell lysates collected daily for up to 4 days and submitted to SDS – PAGE and Western blot did not indicate the expression of bothropasin precursor (not shown). On the other hand, the methylotrophic yeast P. pastoris secreted a very low amount of a protein of , 70 kDa detected only by Western blot using anti-bothropasin antibodies (not shown). 3.3. Expression of bothropasin individual domains The cDNA sequences coding for the pro-domain, or the pro-domain plus the metalloproteinase domain, or the disintegrin-like plus the Cys-rich domains were subcloned
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
221
Fig. 1. cDNA sequence of clone 8 £ 3.3 encoding bothropasin, with the translated open reading frame from the start codon ATG to the termination codon TAG (both in italic). A polyadenylation signal is shown in lowercase letters. The deduced amino acid sequence is shown in one-letter symbols. Underlined amino acid residues denote sequences confirmed by protein sequencing. The shaded Asn residue indicates a putative N-glycosylation site.
222
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 2. Comparison of deduced amino acid sequence of bothropasin with other members of the Reprolysin family. a, bothropasin; b, jararhagin (Paine et al., 1992); c, catrocollastatin (Zhou et al., 1995); d, trigramin (Neeper and Jacobson, 1990); e, atrolisin-A (Hite et al., 1994); f, HR1B (Takeya et al., 1990). The numbering of residues corresponds to that of bothropasin precursor. Putative bothropasin domains: signal peptide: 1– 20; pro-domain: 21–189; proteinase domain: 190–394; disintegrin-like domain: 395–487; carboxyl cysteine-rich domain: 488–610. Residues conserved in all P-III class mature proteins (a, b, c, e, and f) are shaded.
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 3. Expression of pro-bothropasin (BPMDC). (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions obtained by the expression of pro-bothropasin (BPMDC) in E. coli BL21(DE3)pLysS. (1) Mol. mass markers. (2) Native bothropasin. (3) Autolysed bothropasin. (4) and (5) Protein fractions of inclusion bodies of bacteria transformed with plasmids pET17b and pET-17bBPMDC, and induced with 0.5 mM IPTG, respectively. Arrows indicate protein bands recognized by anti-bothropasin antibodies.
in the expression vector pGEX-4T1 which allows the expression of soluble recombinant proteins in fusion with GST. E. coli DH5a cells were transformed with the resultant expression plasmids: pGEX-4T1-BP, pGEX-4T1-BPM and pGEX-4T1-BDC, respectively, and protein expression was induced with 0.5 mM IPTG for different incubation times. For the expression of the protein containing the prodomain plus the metalloproteinase domain of bothropasin (BPM) the cells were cultured for up to 6 h. Samples of the culture media were taken every hour and submitted to chromatography on Glutathione Sepharose 4B (GS4B) to follow the production of the fusion protein. As shown in Fig. 4, the eluted fractions contained a major protein band of about 68 kDa that was recognized by antibodies antibothropasin (Fig. 4, panel B). The Western blot also revealed a minor band of about 48 kDa that could represent the pro þ metalloproteinase domains released from the GST by a bacterial proteinase. The identity of the 68 kDa fusion protein was confirmed by N-terminal amino acid sequencing, which showed the sequence of GST and by the sequencing of an internal peptide, -L– Q – R– E – T –Y – F– I –E – P– L – K, present in the pro-domain of bothropasin. The yield of the fusion protein GST-BPM was, however, very low, estimated as 25– 50 mg/l culture. To verify the possibility that the instability of the metalloproteinase domain observed with native bothropasin was impairing the expression of the pro þ metalloproteinase domains (BPM) we tested the expression of the pro-domain (BP) by the same system. The recombinant pro-domain (BP) was detected by SDS – PAGE as a major band of about 46 kDa, which matches the expected mol mass calculated for the pro-domain fused to GST (Fig. 4(D), lane 3). The recombinant protein was isolated by chromatography on
223
GS4B and after cleavage of the fusion protein by thrombin a major protein band of about 20 kDa was obtained. The N-terminal amino acid sequence of this band was identified as G –S –Y – V– A –F –P –Y – Q– G – where the first four residues are contributed from the expression vector. The remaining sequence starting at A– F– P corresponds to the amino-terminal sequence of the pro-domain of bothropasin. The BP protein was expressed at ,3 mg/l culture medium, which is 100 times higher than the yield obtained from the construct containing the metalloproteinase domain. We also tested the expression of the remaining domains of bothropasin precursor: disintegrin-like and Cys-rich, which should play a role in the interaction of snake venom metalloproteinases with integrins. The recombinant fusion protein GST-BDC was detected by SDS –PAGE and by Western blot using anti-bothropasin antibodies as a ,52 kDa protein, which shifted to , 30 kDa after release of GST by thrombin (Fig. 5(A) and (B), lanes 4, 5). The Nterminal amino acid sequence of the BDC protein was determined to be G– S– Y –V – L – G– T – D– I, where the residues L – G – T – D– I… correspond to the amino-terminal sequence of the disintegrin-like domain (Fig. 5(C)). The yield of the BDC protein was estimated as 1 mg/l culture medium. The recombinant protein BDC was then tested for its ability to inhibit collagen-induced platelet aggregation. As shown in Fig. 5 (panel D), BDC inhibited aggregation of platelet suspensions in a dose-dependent manner with an estimated IC50 value of 1.4 mM, which is ,5 times higher than the value of 290 nm, obtained for native bothropasin tested under the same conditions.
4. Discussion In the present study we cloned and analyzed the expression of the zymogen form and of the individual domains of bothropasin, a metalloproteinase with hemorrhagic and myonecrotic activities, isolated from the venom of B. jararaca. The deduced amino acid sequence of the bothropasin precursor consisted of 610 amino acids forming a multi-domain structure comprised of a signal peptide followed by the pro, metalloproteinase, disintegrin-like and cysteine-rich domains found in the P-III class venom metalloproteinases. Bothropasin is a highly active proteolytic enzyme inhibited by metal-chelating compounds and the sequence of its metalloproteinase domain included the consensus zinc-binding site, H – E – X– X– H– X – X – G – X– X– H. The nucleotide sequence of the cDNA encoding bothropasin showed 96% identity with the one coding for jararhagin, a metalloproteinase isolated from the venom of B. jararaca that shows similar biochemical and biophysical properties to bothropasin (Paine et al., 1992). The differences found were: 11 bases in the pro-domain, 21 bases in the metalloproteinase domain, three bases in the disintegrin-like and cysteine-rich domains and 14 bases in
224
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 4. Expression of pro þ metalloproteinase domains (BPM) and pro-domain (BP) of bothropasin by E. coli DH5a. (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions of BPM obtained by affinity to GS4B. (1) Mol. mass markers. (2)– (6) Protein fractions eluted with reduced glutathione, induction with 0.5 mM IPTG for 1, 2, 3, 4 and 6 h, respectively. (C) N-terminal amino acid sequence of a peptide obtained from the protein band GST-BPM (indicated by an arrow, lane 6). (D) Fractions of BP obtained by affinity to GS4B. (1) Mol. mass markers. (2), (3) Protein fractions eluted with reduced glutathione from pGEX-4T1 and pGEX4T1-BP, respectively. (4) Protein bound to GS4B after thrombin cleavage. (5) Protein eluted from GS4B after thrombin-cleavage. (E) N-terminal amino acid sequence of the protein band BP (indicated by an arrow in D, lane 5).
the 3’UTR including a 8-base insertion in the bothropasin cDNA starting at position 2058 (Fig. 1). Seven nucleotide substitutions are synonymous, found mainly in the prodomain while 26 non-synonymous substitutions led to 18 amino acid changes in the metalloproteinase domain and one in the disintegrin-like domain. The cDNA coding for jararhagin was obtained from a library constructed with the mRNA isolated from the venom glands of one B. jararaca specimen (Paine et al., 1992) while the nucleotide sequence of bothropasin was cloned from a pool of venom gland tissue from five snakes. Since the differences between the cDNAs coding for bothropasin and jararhagin were found along all the extension of their sequences one can exclude the possibility that an alternative splicing of an identical precursor RNA generated the substitutions. On the other hand, the differences in amino acid sequence between these two similar metalloproteinases as well as the ones found among other venom metalloproteinases could be the result
of a process of accelerated evolution already described for the phospholipases and serineproteinases, which led to the accumulation of non-synonymous nucleotide substitutions in the regions encoding the mature proteins (Nakashima et al., 1995; Deshimaru et al., 1996; Ohno et al., 1998). Since the primary function of viperid venom is to immobilize and digest prey and prey animals vary in their susceptibility to venom components, the synthesis of various mature metalloproteinases by the venom gland provided viperid snakes with a large repertoire of tools to interact with animal target proteins. Since to our best knowledge there is no report on the processing mechanism of venom P-III metalloproteinase precursors, one of the goals of this work was the expression of pro-bothropasin in order to test the removal of its prodomain by venom proteinases, as it was reported for atrolysin E (Shimokawa et al., 1996), and by a pro-protein convertase, as is the case of the activation of some matrix
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
225
Fig. 5. Expression of disintegrin-like þ Cys-rich domains of bothropasin (BDC) by E. coli DH5a. (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions obtained by affinity to GS4B. (1) Mol. mass markers. (2) Whole cell extract. (3) Protein fraction not bound to GS4B. (4) Protein fraction bound to GS4B. (5) Protein eluted from GS4B after thrombin-cleavage. (C) N-terminal amino acid sequence of the protein band BDC (indicated by an arrow in A, lane 5). (D) Inhibition of collagen-induced aggregation of platelet suspensions by native bothropasin, estimated IC50 ¼ 290 nM and recombinant BDC, estimated IC50 ¼ 1.4 mM. Data were confirmed by two other experiments.
proteinases (Nagase and Woessner, 1999) and the ADAMs (Primakoff and Myles, 2000). For this purpose, we tested three different expression systems: bacterial, yeast and mammalian cells. Although eucaryotic cells have been used to express other recombinant venom toxins, both mammalian COS-7 and yeast P. pastoris cells failed to produce the bothropasin precursor. Moreover, from the three strains of E. coli used in the attempt to obtain the recombinant precursor only the strain BL21(DE3)pLysS, which is recommended for use in expressing toxic genes, was able to produce an apparently degraded recombinant protein recognized by anti-bothropasin antibodies. These results suggest that either the precursor of bothropasin was toxic and was degraded to different extents by the proteinases
from the cells used as host for protein expression or it was unstable under the conditions used for expression and analysis by SDS – PAGE. Since our attempts to obtain the zymogen form of bothropasin failed, we compared the expression of its individual domains to identify the protein region that was impairing protein expression. We chose E. coli as a host and the expression vector pGEX-4T because of the easy handling and the possibility of isolating the fusion protein by affinity chromatography on glutathione-Sepharose. Our pGEX-4T constructs for the expression of the pro-domain (BP), pro þ metalloproteinase domains (BPM) and disintegrin-like þ cysteine-rich domains (BDC) of bothropasin in fusion with GST enabled us to obtain the recombinant
226
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
proteins and their identities were confirmed by amino acid sequencing. Besides, the protein BDC was showed to be active by inhibiting collagen-induced platelet aggregation in a dose-dependent fashion. However, a comparison of their expression levels clearly showed that the metalloproteinase domain of bothropasin was the main reason for the difficult expression of the recombinant proteins in this work. While the recombinant pro-domain yielded 3 mg/l culture and from the disintegrin-like þ cysteine-rich domains 1 mg/l culture was obtained, the expression of the pro þ metalloproteinase domains was estimated as no higher than 50 mg/l culture. An inspection of the primary structure of the metalloproteinase domain of the Reprolysins shows that they are highly variable. The most conserved features of the metalloproteinase domains are the sequence around their extended Zn-binding sites and the positions of the cysteine residues. P-III metalloproteinase domains usually contain six conserved cysteines plus a seventh residue found in variable positions. Bothropasin has seven Cys residues in the metalloproteinase domain, however, the titration of free cysteines in bothropasin resulted negative (not shown). Although venom P-III metalloproteinases are found in the venom predominantly in the non-autolysed status, polypeptides of about , 30 kDa corresponding to the disintegrinlike þ cysteine-rich domains of these metalloproteinases have been isolated revealing a single chain toxin capable of inhibiting collagen-induced platelet aggregation (Usami et al., 1994). The presence of 28 cysteine residues in these domains of the P-III class proteins suggests that they fold in a stable, compact fashion as it was shown by the determination of the disulfide pattern of catrocollastatin C, a disintegrin-like/cysteine-rich protein isolated from C. atrox venom (Calvete et al., 2000). The sequence of the peptides obtained from an autolysed bothropasin sample indicated that the metalloproteinase domain was cleaved at various sites while the disintegrin-like and cysteine-rich domains remained intact. Taken together, these results indicate that the metalloproteinase domain of bothropasin is unstable and may have decreased the yield of expression of the recombinant proteins containing this domain by the systems used in this work.
Acknowledgements This work was supported by a grant from FAPESP, Brazil (CAT/CEPID 98/14307-9). We thank Marcelo L. Santoro (Instituto Butantan, Brazil) for help with the platelet aggregation assays, and Anne Wisner (Institut Pasteur, France) for advising on the protein expression in COS-7 cells. We also thank Prof. Jay W. Fox (University of Virginia, USA) for the helpful discussions.
References Bjarnason, J.B., Fox, J.W., 1995. Snake venom metalloendopeptidases: reprolysins. Meth. Enzymol. 248, 345–368. Black, R.A., White, J.M., 1998. ADAMs: focus on the protease domain. Curr. Opin. Cell Biol. 10, 654–659. Calvete, J.J., Moreno-Murciano, M.P., Sanz, L., Jurgens, M., Schrader, M., Raida, M., Benjamin, D.C., Fox, J.W., 2000. The disulfide bond pattern of catrocollastatin C, a disintegrin-like/ cysteine-rich protein isolated from Crotalus atrox venom. Protein Sci. 9, 1365–1373. Deshimaru, M., Ogawa, T., Nakashima, K., Nobuhisa, I., Chijiwa, T., Shimohigashi, Y., Fukumaki, Y., Niwa, M., Yamashina, I., Hattori, S., Ohno, M., 1996. Accelerated evolution of crotalinae snake venom gland serine proteases. FEBS Lett. 397, 83–88. Hite, L.A., Shannon, J.D., Bjarnason, J.B., Fox, J.W., 1992. Sequence of a clone encoding the zinc metalloproteinase hemorrhagic toxin e from Crotalus atrox: evidence for signal, zymogen, and disintegrin-like structures. Biochemistry 31, 6203–6211. Hite, L.A., Jia, L.G., Bjarnason, J.B., Fox, J.W., 1994. cDNA sequences for four snake venom metalloproteinases: structure, classification, and their relationship to mammalian reproductive proteins. Arch. Biochem. Biophys. 308, 182 –191. Kamiguti, A.S., Cardoso, J.L., Theakston, R.D., Sano-Martins, I.S., Hutton, R.A., Rugman, F.P., Warrell, D.A., Hay, C.R., 1991. Coagulopathy and hemorrhage in human victims of Bothrops jararaca envenoming in Brazil. Toxicon 29, 961 –972. Kamiguti, A.S., Hay, C.R.M., Zuzel, M., 1996. Inhibition of collagen-induced platelet aggregation as the result of cleavage of a2b1-integrin by the snake venom metalloproteinase jararhagin. Biochem. J. 320, 635–641. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 –685. Mandelbaum, F.R., Assakura, M.T., 1988. Antigenic relationship of hemorrhagic factors and proteases isolated from the venoms of three species of Bothrops snakes. Toxicon 26, 379–385. Mandelbaum, F.R., Assakura, M.T., Reichl, A.P., 1982. Isolation and characterization of a proteolytic enzyme from the venom of the snake Bothrops jararaca (jararaca). Toxicon 20, 955–972. Moura-da-Silva, A.M., Linica, A., Della-Casa, M.S., Kamiguti, A.S., Ho, P.L., Crampton, J.M., Theakston, R.D., 1999. Jararhagin ECD-containing disintegrin domain: expression in Escherichia coli and inhibition of the platelet-collagen interaction. Arch. Biochem. Biophys. 369, 295– 301. Murphy, G.J.P., Murphy, G., Reynolds, J.J., 1991. The origin of matrix metalloproteinases and their familial relationships. FEBS Lett. 289, 4–7. Mustard, J.F., Perry, D.W., Ardlie, N.G., Packham, M.A., 1972. Preparations of suspensions of washed platelet from humans. Br. J. Haematol. 22, 193 –204. Nagase, H., Woessner, J.F. Jr., 1999. Matrix metalloproteinases. J. Biol. Chem. 274, 21491–21494. Nakashima, K., Nobuhisa, I., Deshimaru, M., Nakai, M., Ogawa, T., Shimohigashi, Y., Fukumaki, Y., Hattori, M., Sakaki, Y., Hattori, S., Ohno, M., 1995. Accelerated evolution in the protein-coding regions is universal in crotalinae snake venom gland phospholipase A2 isozyme genes. Proc. Natl. Acad. Sci. USA 92, 5605–5609.
M.T. Assakura et al. / Toxicon 41 (2003) 217–227 Neeper, M.P., Jacobson, M.A., 1990. Sequence of a cDNA encoding the platelet aggregation inhibitor trigramin. Nucl. Acids Res. 18, 4255–4261. Ohno, M., Menez, R., Ogawa, T., Danse, J.M., Shimohigashi, Y., Fromen, C., Ducancel, F., Zinn-Justin, S., Le Du, M.H., Boulain, J.C., Tamiya, T., Menez, A., 1998. Molecular evolution of snake toxins: is the functional diversity os snake toxins associated with a mechanism of accelerated evolution? Prog. Nucl. Acid Res. Mol. Biol. 59, 307 –364. Paine, M.J.I., Desmond, H.P., Theakston, R.D.G., Crampton, J.M., 1992. Purification, cloning, and molecular characterization of a high molecular weight hemorrhagic metalloproteinase, jararhagin, from Bothrops jararaca venom. J. Biol. Chem. 267, 22869– 22876. Primakoff, P., Myles, D.G., 2000. The ADAM gene family: surface proteins with adhesion and protease activity. Trends Genet. 16, 83–87. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, second ed., Cold Spring Harbour Laboratory, Spring Harbor, NY. Shimokawa, K., Jia, L.G., Wang, X.M., Fox, J.W., 1996. Expression, activation, and processing of the recombinant snake venom metalloproteinase, pro-atrolysin E. Arch. Biochem. Biophys. 335, 283 –294.
227
Stocker, W., Grams, F., Baumann, U., Reinemer, P., Gomis-Ru¨th, F.-X., Mckay, D.B., Bode, W., 1995. The metzincins— topological and sequential relations between the astacins, adamalysins, serralysins, and matrixins (collagenases) define a super family of zinc-peptidases. Protein Sci. 4, 823 –840. Takeya, H., Oda, K., Miyata, T., Omori-Satoh, T., Iwanaga, S., 1990. The complete amino acid sequence of the high molecular mass hemorrhagic protein HR1B isolated from the venom of Trimeresurus flavoviridis. J. Biol. Chem. 265, 16068–16073. Takeya, H., Miyata, T., Nishino, N., Omori-Satoh, T., Iwanaga, S., 1993. Snake venom hemorrhagic and non-hemorrhagic metalloendopeptidases. Meth. Enzymol. 223, 365 –378. Usami, Y., Fujimura, Y., Miura, S., Shima, H., Yoshida, E., Yoshioka, A., Hirano, K., Suzuki, M., Titani, K., 1994. A 28 kDa-protein with disintegrin-like structure (jararhagin-C) purified from Bothrops jararaca venom inhibits collagen- and ADP-induced platelet aggregation. Biochem. Biophys. Res. Commun. 201, 331–339. Zhou, Q., Smith, J.B., Grossman, M.H., 1995. Molecular cloning and expression of catrocollastatin, a snake-venom protein from Crotalus atrox (western diamondback rattlesnake) which inhibits platelet adhesion to collagen. Biochem. J. 307, 411 –417.
Molecular cloning and expression of structural domains of bothropasin, a P-III metalloproteinase from the venom of Bothrops jararaca Marina T. Assakuraa, Carlos A. Silvaa, Reinhard Menteleb, Antonio C.M. Camargoa, Solange M.T. Serranoa,* a
Laborato´rio de Bioquı´mica e Biofı´sica, Instituto Butantan, Av. Vital Brasil 1500, CEP 05503-900 Sa˜o Paulo, SP, Brazil b Department of Clinical Chemistry and Clinical Biochemistry, Klinikum Innenstadt, Universita¨t Mu¨nchen, D-80336 Mu¨nchen, Germany Received 31 July 2002; accepted 3 September 2002
Abstract Mature P-III snake metalloproteinases are soluble venom components which belong to the Reprolysin sub family and are structurally related to the mammalian membrane-bound A Disintegrin And Metalloproteinase (ADAMs). Here we present the molecular cloning of bothropasin, a metalloproteinase with hemorrhagic and myonecrotic activities isolated from the venom of Bothrops jararaca. The full-length cDNA encoding the bothropasin precursor was cloned by immunoscreening and its authenticity was confirmed by the amino acid sequence of internal fragments obtained from an autolyzed sample of native bothropasin. The predicted bothropasin precursor is comprised of the elements of a P-III venom metalloproteinase: signal sequence, pro-, metalloproteinase, disintegrin-like and cysteine-rich domains. In the autolysis process of native bothropasin, the disintegrin-like and cysteine-rich domains remained intact while the metalloproteinase domain was cleaved at different sites. The attempts made to obtain the recombinant precursor form of bothropasin using bacterial, yeast and mammalian cell expression systems failed to produce it in an amount sufficient to analyze the activation of the zymogen. Nevertheless, the study of the expression of the individual domains of bothropasin using a bacterial system resulted in the production of recombinant pro-and disintegrin-like þ cysteine-rich domains but not the metalloproteinase domain. These results along with the autolysis pattern of the native protein suggest a role for the metalloproteinase domain in the structural stability of bothropasin. q 2002 Elsevier Science Ltd. All rights reserved. Keywords: Metalloproteinase; Disintegrin; Autolysis; Recombinant protein expression
1. Introduction Envenomation with snake venoms of the families Crotalidae and Viperidae produces pronounced local hemorrhage and muscle damage. These venoms are a rich source of active zinc metalloproteinases, which have been extensively investigated (Hite et al., 1992; Paine et al., 1992; Takeya et al., 1993; Bjarnason and Fox, 1995). The metalloproteinases isolated from snake venoms are mem* Corresponding author. Fax: þ 55-11-3726-1024. E-mail address: [email protected] (S.M.T. Serrano).
bers of the Reprolysin subfamily of enzymes, which also includes the A Disintegrin And Metalloproteinase (ADAMs). They are secreted as preproenzymes and contain additional regulatory modules, which are responsible for interactions with the extracellular matrix and integrins. The mature P-I class proteins have only a proteinase domain, whereas the P-II, P-III, and P-IV classes have disintegrin or disintegrin-like, cysteine-rich, and lectin-like domains found carboxy to the proteinase domain, respectively. Similar domain structures are found in the ADAMs, which additionally contain an epidermal growth factor-like domain, a transmembrane region and a cytoplasmic tail (Black and White, 1998).
0041-0101/03/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 0 2 ) 0 0 2 7 9 - 9
218
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
The venom of Bothrops jararaca contains hemorrhagic toxins that cause disruption of the basement membrane of the vascular endothelium resulting in bleeding (Kamiguti et al., 1991). At present, five metalloproteinases with hemorrhagic activity have been isolated from this venom. HF1, HF2 and HF3 are highly active hemorrhagic toxins (Mandelbaum and Assakura, 1988). Jararhagin is another hemorrhagic proteinase of 52 kDa, which has been extensively analyzed and shown to cause inhibition of collageninduced platelet aggregation (Paine et al., 1992; Kamiguti et al., 1996; Moura-da-Silva et al., 1999). Bothropasin is a single chain metalloproteinase of 48 kDa isolated from the same venom, whose caseinolytic activity corresponds to , 15% of the total activity of the crude venom (Mandelbaum et al., 1982). It causes hemorrhage on rabbit skin with a MHD of 1 mg. The same dose causes myonecrosis in mouse tibialis anterior muscle. On the B-chain of oxidized insulin, bothropasin shows specificity for leucine and phenylalanine bonds. Polyclonal antibodies raised in rabbits against bothropasin neither give a precipitin line with hemorrhagic factors (HF1 and HF2) nor neutralize their hemorrhagic activity (Mandelbaum and Assakura, 1988). This specific antiserum was used to screen a B. jararaca venom gland cDNA expression library in order to identify putative clones encoding bothropasin. Metalloproteinases of the Reprolysin subfamily are synthesized as zymogens, which are proteolytically activated by a mechanism where the N-terminal propeptide is cleaved off (Stocker et al., 1995). The in vitro activation/processing of the recombinant zymogen form of the Crotalus atrox P-II class metalloproteinase atrolysin E by the venom enzymes atrolysin A and atrolysin E itself was reported (Shimokawa et al., 1996). However, the pro-domain of venom proteinases also contains a conserved pair of basic residues K180 – K181, which might be cleaved immediately downstream by a proprotein convertase-dependent pathway in the gland tissue leading to the activation of the zymogen as it is the case of the matrix metalloproteinases (Nagase and Woessner, 1999) and the ADAMs (Primakoff and Myles, 2000). To investigate the ability of a proconvertase to process the precursor of bothropasin we attempted to produce the recombinant proprotein using different expression systems. However, due to the very low expression level obtained from the pro-protein we tested the expression of its individual domains in order to gain some knowledge about the structure features that might be important for the successful expression of snake venom PIII class metalloproteinases.
2. Materials and methods 2.1. Molecular cloning of bothropasin A B. jararaca cDNA expression library was constructed in lZAPII from the venom gland poly(A)þ RNA using a cDNA
synthesis kit (Stratagene) according to the manufacturer’s instructions. Total RNA was extracted from venom glands of five specimens of B. jararaca with RNAzol B (WAKChemie). Double-stranded cDNA was synthesized using 5 mg of poly(A)þ RNA template, ligated to Eco RI – Xho Idigested lZAPII DNA, and packaged. The library contained over 106 independent phages. Anti-bothropasin antibodies were prepared as described elsewhere (Mandelbaum and Assakura, 1988). Absorption of anti-Escherichia coli and anti-l phage antibodies was performed as described in Sambrook et al. (1989). The library was plated on E. coli XL1-Blue MRF’ and grown at 37 8C for 8 h. Recombinant protein expression was induced with 10 mM isopropyl-1-thio-b-D -galactopyranoside saturated nitrocellulose (Schleischer and Schuell) and the filters were processed for immunostaining with antibothropasin antibodies, as described in Sambrook et al. (1989). Plasmids were excised from lZAPII vector and recircularized in the presence of ExAssiste (Stratagene) helper phage to form phagemid pBluescript. Plasmids containing longest inserts were selected for sequencing of both strands in an ABI Prism 310 Genetic Analyzer (Applied Biosystems) using an ABI PRISM Dye Terminator Cycle Sequencing kit (Perkin Elmer) and oligonucleotide primers to internal sequence to obtain overlapping sequence information. The sequence obtained was deposited to the GenBank database (accession AF056025). 2.2. Expression of pro-bothropasin in E. coli The expression plasmid pGEX-4T2-BPMDC, that contains cDNA encoding amino acid residues 13 – 610 and includes pro, metalloproteinase, disintegrin-like and cysteine-rich domains was constructed by amplification of the cDNA fragment by PCR and subcloning into pGEMe-T vector (Promega). The plasmid was sequenced on both strands to ensure that the coding sequence was correct. To generate a fusion protein with glutathione S-transferase (GST), the insert from the pGEMe-T vector was isolated as a Bam H I– Xho I fragment and cloned into pGEX-4T (Amersham Pharmacia). The constructed plasmid was transformed into E. coli strain DH5a, and grown in 50 ml Luria –Bertani medium containing 100 mg ml21 ampicillin at 37 8C to a cell density of A600 ¼ 0.6 – 0.8. The expression of recombinant protein was induced by adding concentrations ranging from 0.3 to 1.5 mM isopropyl thio-b-D galactopyranoside (IPTG) and incubation was continued for 3 h at 30 8C. Cells were processed as described below for the expression of individual domains of bothropasin. The expression plasmid pET17b-BPMDC containing cDNA encoding residues 13 – 610 corresponding to the sequence of pro-bothropasin was constructed by subcloning the insert from pGEMe-T vector (see above) into pET17b (Novagen) previously digested with Bam H I – Xho I. The plasmid was transformed into both E. coli XL1BlueMRF’ and BL21(DE3)pLysS strains, and grown for induction of
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
protein expression as described earlier. After cell lysing by sonication in lysis buffer (50 mM Tris – HCl, pH 8.0, 2 mM EDTA, 1% Triton X-100, 10 mg ml21 lyzozyme) inclusion bodies were recovered by centrifugation, washed four times with 50 mM Tris –HCl, pH 8.0 containing 2 mM EDTA and submitted to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS –PAGE) and Western blot analysis. 2.3. Expression of pro-bothropasin in P. pastoris A cDNA fragment encoding amino acid residues 13 – 610 corresponding to the sequence of pro-bothropasin was amplified by PCR incorporating nucleotide sequences of Sna B I and Bam H I (sense) and Not I and Bam H I (antisense). The purified PCR product was ligated into the pGEM-11Zf vector. The insert from the pGEM-11Zf vector was isolated and cloned into pPIC9K (Invitrogen). The construct, named pPIC9K-BPMDC, was transformed into the E. coli strain XL1Blue MRF’ and sequenced prior to transformation into yeast. Following the identification of Hisþ transformants of P. pastoris, the screening for clones expressing pro-bothropasin was performed by growing 20 ml in buffered glycerol medium (BMGY) with vigorous shaking at 30 8C in loosely covered flasks for 2 days, and resuspending the cell pellets in 2 ml buffered methanol medium (BMMC) or non-buffered methanol medium (MMC), and growing under the same conditions for 5 days. Media were clarified by centrifugation and subjected to SDS – PAGE and Western blot analysis to assess the level of expression. 2.4. Expression of pro-bothropasin in COS-7 cells A cDNA fragment corresponding to amino acid residues 1 – 610 corresponding to the complete coding sequence of the bothropasin precursor was amplified by PCR incorporating nucleotide sequences of Bam H I (sense) and Xho I (antisense). The purified PCR product was ligated into the pGEMe-T vector. The insert from the pGEMe-T vector was isolated and cloned into the expression vector pcDNA3 (Invitrogen). The construct, named pcDNA3-BPMDC, was transformed into the E. coli strain XL1Blue MRF’ and sequenced prior to transfection into COS-7 cells. For transient expression of the precursor of bothropasin, monolayers of 106 COS-7 cells grown in petri dishes of 10 cm were transfected with 10 mg of pcDNA3-BPMDC using Exgen 500 (Euromedex) as a gene-delivering agent. Eighteen hours after transfection, 4 ml of 2xDulbecco’s modified Eagle’s medium (DMEM) containing 20% Nuserume (Becton Dickinson) were added to each plate containing the same volume of DMEM and incubation was prolonged for 24 h at 37 8C. The medium from each plate was substituted by 1 £ DMEM containing 10% Nu-serume and incubation was prolonged for another 48 h period. Cells were harvested with phosphate buffered saline containing 0.5 mM EDTA and cell lysates were obtained by treating cells with
219
40 mM Tris – HCl, pH 7.5, 40 mM MgCl2, 1% Triton X-100. The content of cells and the media were analyzed by SDS – PAGE and Western blot for protein expression. 2.5. Expression of bothropasin individual domains in E. coli Three separate expression plasmids were constructed: pGEX-4T1-BPM, that includes cDNA encoding residues 13 – 394 and encodes the pro and metalloproteinase domains; pGEX-4T1-BP that contains cDNA encoding residues 13 – 189 and has the pro domain only; and pGEX4T1-BDC that contains cDNA encoding residues 395– 610 and includes the disintegrin-like and Cys-rich domains. The bothropasin cDNA fragments were amplified by PCR and subcloned into pGEMe-T vector (Promega). The plasmids were sequenced on both strands to ensure that their coding sequences were correct. To generate fusion proteins with GST, the inserts from the pGEMe-T vector were isolated as Bam H I – Xho I fragments and cloned into pGEX-4T (Amersham Pharmacia). Each of the constructed plasmids was transformed into E. coli strain DH5a, and grown as described earlier for induction of protein expression in E. coli. Cells were collected by centrifugation at 4000 g for 5 min and suspended in 1 ml lysis buffer (50 mM Tris – HCl pH 7.5, 150 mM NaCl, 1% Triton X-100). After lysis by sonication cell lysate was centrifuged and 80 ml of glutathione-Sepharose suspension were added to supernatant and incubated for 16 h at 4 8C. The resin was washed successively with lysis buffer to eliminate unbound proteins and incubated with thrombin cleavage buffer (50 mM Tris – HCl pH 8.0, 150 mM NaCl, 2.5 mM CaCl2) and 4U thrombin (Sigma) for 2 h at 37 8C. The recombinant protein obtained in the supernatant was analyzed by SDS – PAGE, Western blot analysis and N-terminal amino acid sequencing. 2.6. Analytical procedures Protein concentrations were determined by the Bio-Rad protein assay kit using bovine serum albumin as a standard. SDS –PAGE was carried out according to Laemmli (1970). For Western blot, the protein samples were transferred to nitrocellulose membrane and developed as described earlier for immunoscreening. Fragments originated from autolysed bothropasin were isolated by RP-HPLC on a RP18 column (Reprosil, 1 £ 150 mm, Maisch). N-terminal amino acid sequencing of peptides originated from the autolysis of native bothropasin, of the internal peptide obtained by digestion of the 68 kDa fusion protein GST-BPM with endoproteinase Lys-C, and of protein bands transferred to PVDF membrane (Bio-Rad) was performed by Edman degradation using a 473A sequencer (Applied Biosystems) following manufacturer’s operating protocols. Platelet suspensions were prepared from fresh citrated human blood obtained from healthy donors, according to Mustard et al. (1972). The native and recombinant proteins
220
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
were incubated with platelet suspensions containing 3 £ 108 ml21 for 5 min at 37 8C before the addition of 0.5 mg ml21 collagen (Sigma). The extent of aggregation was estimated as the percentage increase of light transmission taking the value obtained for platelets incubated with control buffer and stimulated with collagen as 100%.
3. Results 3.1. Cloning of a cDNA encoding bothropasin and sequence analysis Several immunoreactive clones were isolated from the cDNA bank and those with largest inserts were selected for further studies. Fig. 1 shows the full-length nucleotide and the deduced amino acid sequences of the bothropasin precursor encoded by clone 8 £ 3.3. The cDNA consists of 2334 bases including a 73 bases 50 -end non-coding region, an open reading frame of 1830 bases, and a 30 -end non-coding region of 431 bases with termination codon, AATAAA sequence and polyadenylation sites. The cDNA sequence of bothropasin predicts a pro-protein with striking similarities to precursors of the Reprolysin subfamily, belonging to class P-III (Bjarnason and Fox, 1995). Native bothropasin undergoes autolysis under storage conditions at 4 8C in 50 mM Tris – HCl pH 7.5, containing 0.5 mM CaCl2, originating main fragments of 40 –30 kDa which are detected after 7 days by SDS – PAGE and Western blot analysis (Fig. 3(A) and (B), lane 3). Autolyzed bothropasin was submitted to RP-HPLC on a C18 column to isolate fragments and the N-terminal sequence of five internal peptides confirmed the authenticity of the isolated cDNA (Fig. 1). The N-terminal sequence of the fragment that starts at position 395 (L –G – T – D– I…) and its molecular mass (30 kDa) indicate that it corresponds to the intact disintegrin-like and Cys-rich domains of bothropasin. The other four peptides sequenced derived from the metalloproteinase domain. The molecular mass of the predicted mature protein of 46.6 kDa is lower than the value of 48 kDa estimated by SDS – PAGE (Mandelbaum et al., 1982) probably due to carbohydrate moieties. One putative N-glycosylation site was detected at Asn372 located in the proteinase domain (Fig. 1) of the bothropasin precursor. Fig. 2 shows the alignment of the predicted amino acid sequence of bothropasin with other venom metalloproteinase precursors of the Reprolysin subfamily. The putative prodomain contains the conserved sequence P– K –M – C– G – V– T present in the correspondent positions of other venom metalloproteinase precursors and which resembles a region in the propeptide of matrixins called Cys-switch which is involved in the activation of these enzymes (Nagase and Woessner, 1999). Bothropasin exhibits the characteristic elongated consensus motif HEXXHXXGXXH, whose three histidine residues are involved in binding of the catalytic
essential zinc ion (Murphy et al., 1991). Moreover, the same domain contains a conserved methionine residue, which in the metzincins is located beneath the active site metal as part of a super imposable ‘Met-turn’ (Stocker et al., 1995). In the disintegrin-like domain of bothropasin, the sequence SECD replaces the RGD sequence that is frequently found in the disintegrin domains of the P-II class precursors (Fig. 2). The 28 Cys residues found in the disintegrin-like/cysteine-rich domains of bothropasin are conserved among P-III metalloproteinases. 3.2. Expression of pro-bothropasin The cDNA-coding region for pro-bothropasin (BPMDC) was produced by PCR and subcloned into different expression vectors. Initially, we tried the expression of the precursor as a soluble cytoplasmic protein in fusion with GST by transforming E. coli DH5a cells with the expression vector pGEX-4T2BPMDC. However, the cells induced with IPTG concentrations ranging from 0.3 to 1.5 mM IPTG failed to express the recombinant protein (not shown). In another attempt the expression vector pET17b-BPMDC was used to transform the E. coli strains XL1BlueMRF’ and BL21(DE3)pLysS. Bacterial cells of both strains induced with 0.5 mM IPTG failed to produce the expected 70 kDa recombinant precursor (not shown). However, the analysis of proteins produced as inclusion bodies by BL21(DE3)pLysS showed three protein bands of , 68, 40 and 30 kDa which were detected by Western blot using the anti-bothropasin antibodies (Fig. 3(A) and (B), lane 5). This suggests that the recombinant protein either underwent autolysis or was degraded by bacterial proteinases. Various attempts made to get information on the amino acid sequence of these protein bands failed because the recombinant protein was present in very low amount and mixed with bacterial proteins. Furthermore, the use of a cocktail of proteinase inhibitors (1 mM PMSF, 5 mM EDTA, 20 mg/ml pepstatin A, 20 mg/ml leupeptin) during the steps of sample processing for protein electrophoresis did not prevent the hydrolysis of the recombinant protein (not shown). Alternatively to the expression using procaryotic cells, we tested the expression of the precursor by two eucaryotic systems. In case of mammalian COS-7 cells, the analysis of media samples and cell lysates collected daily for up to 4 days and submitted to SDS – PAGE and Western blot did not indicate the expression of bothropasin precursor (not shown). On the other hand, the methylotrophic yeast P. pastoris secreted a very low amount of a protein of , 70 kDa detected only by Western blot using anti-bothropasin antibodies (not shown). 3.3. Expression of bothropasin individual domains The cDNA sequences coding for the pro-domain, or the pro-domain plus the metalloproteinase domain, or the disintegrin-like plus the Cys-rich domains were subcloned
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
221
Fig. 1. cDNA sequence of clone 8 £ 3.3 encoding bothropasin, with the translated open reading frame from the start codon ATG to the termination codon TAG (both in italic). A polyadenylation signal is shown in lowercase letters. The deduced amino acid sequence is shown in one-letter symbols. Underlined amino acid residues denote sequences confirmed by protein sequencing. The shaded Asn residue indicates a putative N-glycosylation site.
222
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 2. Comparison of deduced amino acid sequence of bothropasin with other members of the Reprolysin family. a, bothropasin; b, jararhagin (Paine et al., 1992); c, catrocollastatin (Zhou et al., 1995); d, trigramin (Neeper and Jacobson, 1990); e, atrolisin-A (Hite et al., 1994); f, HR1B (Takeya et al., 1990). The numbering of residues corresponds to that of bothropasin precursor. Putative bothropasin domains: signal peptide: 1– 20; pro-domain: 21–189; proteinase domain: 190–394; disintegrin-like domain: 395–487; carboxyl cysteine-rich domain: 488–610. Residues conserved in all P-III class mature proteins (a, b, c, e, and f) are shaded.
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 3. Expression of pro-bothropasin (BPMDC). (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions obtained by the expression of pro-bothropasin (BPMDC) in E. coli BL21(DE3)pLysS. (1) Mol. mass markers. (2) Native bothropasin. (3) Autolysed bothropasin. (4) and (5) Protein fractions of inclusion bodies of bacteria transformed with plasmids pET17b and pET-17bBPMDC, and induced with 0.5 mM IPTG, respectively. Arrows indicate protein bands recognized by anti-bothropasin antibodies.
in the expression vector pGEX-4T1 which allows the expression of soluble recombinant proteins in fusion with GST. E. coli DH5a cells were transformed with the resultant expression plasmids: pGEX-4T1-BP, pGEX-4T1-BPM and pGEX-4T1-BDC, respectively, and protein expression was induced with 0.5 mM IPTG for different incubation times. For the expression of the protein containing the prodomain plus the metalloproteinase domain of bothropasin (BPM) the cells were cultured for up to 6 h. Samples of the culture media were taken every hour and submitted to chromatography on Glutathione Sepharose 4B (GS4B) to follow the production of the fusion protein. As shown in Fig. 4, the eluted fractions contained a major protein band of about 68 kDa that was recognized by antibodies antibothropasin (Fig. 4, panel B). The Western blot also revealed a minor band of about 48 kDa that could represent the pro þ metalloproteinase domains released from the GST by a bacterial proteinase. The identity of the 68 kDa fusion protein was confirmed by N-terminal amino acid sequencing, which showed the sequence of GST and by the sequencing of an internal peptide, -L– Q – R– E – T –Y – F– I –E – P– L – K, present in the pro-domain of bothropasin. The yield of the fusion protein GST-BPM was, however, very low, estimated as 25– 50 mg/l culture. To verify the possibility that the instability of the metalloproteinase domain observed with native bothropasin was impairing the expression of the pro þ metalloproteinase domains (BPM) we tested the expression of the pro-domain (BP) by the same system. The recombinant pro-domain (BP) was detected by SDS – PAGE as a major band of about 46 kDa, which matches the expected mol mass calculated for the pro-domain fused to GST (Fig. 4(D), lane 3). The recombinant protein was isolated by chromatography on
223
GS4B and after cleavage of the fusion protein by thrombin a major protein band of about 20 kDa was obtained. The N-terminal amino acid sequence of this band was identified as G –S –Y – V– A –F –P –Y – Q– G – where the first four residues are contributed from the expression vector. The remaining sequence starting at A– F– P corresponds to the amino-terminal sequence of the pro-domain of bothropasin. The BP protein was expressed at ,3 mg/l culture medium, which is 100 times higher than the yield obtained from the construct containing the metalloproteinase domain. We also tested the expression of the remaining domains of bothropasin precursor: disintegrin-like and Cys-rich, which should play a role in the interaction of snake venom metalloproteinases with integrins. The recombinant fusion protein GST-BDC was detected by SDS –PAGE and by Western blot using anti-bothropasin antibodies as a ,52 kDa protein, which shifted to , 30 kDa after release of GST by thrombin (Fig. 5(A) and (B), lanes 4, 5). The Nterminal amino acid sequence of the BDC protein was determined to be G– S– Y –V – L – G– T – D– I, where the residues L – G – T – D– I… correspond to the amino-terminal sequence of the disintegrin-like domain (Fig. 5(C)). The yield of the BDC protein was estimated as 1 mg/l culture medium. The recombinant protein BDC was then tested for its ability to inhibit collagen-induced platelet aggregation. As shown in Fig. 5 (panel D), BDC inhibited aggregation of platelet suspensions in a dose-dependent manner with an estimated IC50 value of 1.4 mM, which is ,5 times higher than the value of 290 nm, obtained for native bothropasin tested under the same conditions.
4. Discussion In the present study we cloned and analyzed the expression of the zymogen form and of the individual domains of bothropasin, a metalloproteinase with hemorrhagic and myonecrotic activities, isolated from the venom of B. jararaca. The deduced amino acid sequence of the bothropasin precursor consisted of 610 amino acids forming a multi-domain structure comprised of a signal peptide followed by the pro, metalloproteinase, disintegrin-like and cysteine-rich domains found in the P-III class venom metalloproteinases. Bothropasin is a highly active proteolytic enzyme inhibited by metal-chelating compounds and the sequence of its metalloproteinase domain included the consensus zinc-binding site, H – E – X– X– H– X – X – G – X– X– H. The nucleotide sequence of the cDNA encoding bothropasin showed 96% identity with the one coding for jararhagin, a metalloproteinase isolated from the venom of B. jararaca that shows similar biochemical and biophysical properties to bothropasin (Paine et al., 1992). The differences found were: 11 bases in the pro-domain, 21 bases in the metalloproteinase domain, three bases in the disintegrin-like and cysteine-rich domains and 14 bases in
224
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
Fig. 4. Expression of pro þ metalloproteinase domains (BPM) and pro-domain (BP) of bothropasin by E. coli DH5a. (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions of BPM obtained by affinity to GS4B. (1) Mol. mass markers. (2)– (6) Protein fractions eluted with reduced glutathione, induction with 0.5 mM IPTG for 1, 2, 3, 4 and 6 h, respectively. (C) N-terminal amino acid sequence of a peptide obtained from the protein band GST-BPM (indicated by an arrow, lane 6). (D) Fractions of BP obtained by affinity to GS4B. (1) Mol. mass markers. (2), (3) Protein fractions eluted with reduced glutathione from pGEX-4T1 and pGEX4T1-BP, respectively. (4) Protein bound to GS4B after thrombin cleavage. (5) Protein eluted from GS4B after thrombin-cleavage. (E) N-terminal amino acid sequence of the protein band BP (indicated by an arrow in D, lane 5).
the 3’UTR including a 8-base insertion in the bothropasin cDNA starting at position 2058 (Fig. 1). Seven nucleotide substitutions are synonymous, found mainly in the prodomain while 26 non-synonymous substitutions led to 18 amino acid changes in the metalloproteinase domain and one in the disintegrin-like domain. The cDNA coding for jararhagin was obtained from a library constructed with the mRNA isolated from the venom glands of one B. jararaca specimen (Paine et al., 1992) while the nucleotide sequence of bothropasin was cloned from a pool of venom gland tissue from five snakes. Since the differences between the cDNAs coding for bothropasin and jararhagin were found along all the extension of their sequences one can exclude the possibility that an alternative splicing of an identical precursor RNA generated the substitutions. On the other hand, the differences in amino acid sequence between these two similar metalloproteinases as well as the ones found among other venom metalloproteinases could be the result
of a process of accelerated evolution already described for the phospholipases and serineproteinases, which led to the accumulation of non-synonymous nucleotide substitutions in the regions encoding the mature proteins (Nakashima et al., 1995; Deshimaru et al., 1996; Ohno et al., 1998). Since the primary function of viperid venom is to immobilize and digest prey and prey animals vary in their susceptibility to venom components, the synthesis of various mature metalloproteinases by the venom gland provided viperid snakes with a large repertoire of tools to interact with animal target proteins. Since to our best knowledge there is no report on the processing mechanism of venom P-III metalloproteinase precursors, one of the goals of this work was the expression of pro-bothropasin in order to test the removal of its prodomain by venom proteinases, as it was reported for atrolysin E (Shimokawa et al., 1996), and by a pro-protein convertase, as is the case of the activation of some matrix
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
225
Fig. 5. Expression of disintegrin-like þ Cys-rich domains of bothropasin (BDC) by E. coli DH5a. (A) SDS–PAGE and (B) Western-blot with anti-bothropasin antibodies of fractions obtained by affinity to GS4B. (1) Mol. mass markers. (2) Whole cell extract. (3) Protein fraction not bound to GS4B. (4) Protein fraction bound to GS4B. (5) Protein eluted from GS4B after thrombin-cleavage. (C) N-terminal amino acid sequence of the protein band BDC (indicated by an arrow in A, lane 5). (D) Inhibition of collagen-induced aggregation of platelet suspensions by native bothropasin, estimated IC50 ¼ 290 nM and recombinant BDC, estimated IC50 ¼ 1.4 mM. Data were confirmed by two other experiments.
proteinases (Nagase and Woessner, 1999) and the ADAMs (Primakoff and Myles, 2000). For this purpose, we tested three different expression systems: bacterial, yeast and mammalian cells. Although eucaryotic cells have been used to express other recombinant venom toxins, both mammalian COS-7 and yeast P. pastoris cells failed to produce the bothropasin precursor. Moreover, from the three strains of E. coli used in the attempt to obtain the recombinant precursor only the strain BL21(DE3)pLysS, which is recommended for use in expressing toxic genes, was able to produce an apparently degraded recombinant protein recognized by anti-bothropasin antibodies. These results suggest that either the precursor of bothropasin was toxic and was degraded to different extents by the proteinases
from the cells used as host for protein expression or it was unstable under the conditions used for expression and analysis by SDS – PAGE. Since our attempts to obtain the zymogen form of bothropasin failed, we compared the expression of its individual domains to identify the protein region that was impairing protein expression. We chose E. coli as a host and the expression vector pGEX-4T because of the easy handling and the possibility of isolating the fusion protein by affinity chromatography on glutathione-Sepharose. Our pGEX-4T constructs for the expression of the pro-domain (BP), pro þ metalloproteinase domains (BPM) and disintegrin-like þ cysteine-rich domains (BDC) of bothropasin in fusion with GST enabled us to obtain the recombinant
226
M.T. Assakura et al. / Toxicon 41 (2003) 217–227
proteins and their identities were confirmed by amino acid sequencing. Besides, the protein BDC was showed to be active by inhibiting collagen-induced platelet aggregation in a dose-dependent fashion. However, a comparison of their expression levels clearly showed that the metalloproteinase domain of bothropasin was the main reason for the difficult expression of the recombinant proteins in this work. While the recombinant pro-domain yielded 3 mg/l culture and from the disintegrin-like þ cysteine-rich domains 1 mg/l culture was obtained, the expression of the pro þ metalloproteinase domains was estimated as no higher than 50 mg/l culture. An inspection of the primary structure of the metalloproteinase domain of the Reprolysins shows that they are highly variable. The most conserved features of the metalloproteinase domains are the sequence around their extended Zn-binding sites and the positions of the cysteine residues. P-III metalloproteinase domains usually contain six conserved cysteines plus a seventh residue found in variable positions. Bothropasin has seven Cys residues in the metalloproteinase domain, however, the titration of free cysteines in bothropasin resulted negative (not shown). Although venom P-III metalloproteinases are found in the venom predominantly in the non-autolysed status, polypeptides of about , 30 kDa corresponding to the disintegrinlike þ cysteine-rich domains of these metalloproteinases have been isolated revealing a single chain toxin capable of inhibiting collagen-induced platelet aggregation (Usami et al., 1994). The presence of 28 cysteine residues in these domains of the P-III class proteins suggests that they fold in a stable, compact fashion as it was shown by the determination of the disulfide pattern of catrocollastatin C, a disintegrin-like/cysteine-rich protein isolated from C. atrox venom (Calvete et al., 2000). The sequence of the peptides obtained from an autolysed bothropasin sample indicated that the metalloproteinase domain was cleaved at various sites while the disintegrin-like and cysteine-rich domains remained intact. Taken together, these results indicate that the metalloproteinase domain of bothropasin is unstable and may have decreased the yield of expression of the recombinant proteins containing this domain by the systems used in this work.
Acknowledgements This work was supported by a grant from FAPESP, Brazil (CAT/CEPID 98/14307-9). We thank Marcelo L. Santoro (Instituto Butantan, Brazil) for help with the platelet aggregation assays, and Anne Wisner (Institut Pasteur, France) for advising on the protein expression in COS-7 cells. We also thank Prof. Jay W. Fox (University of Virginia, USA) for the helpful discussions.
References Bjarnason, J.B., Fox, J.W., 1995. Snake venom metalloendopeptidases: reprolysins. Meth. Enzymol. 248, 345–368. Black, R.A., White, J.M., 1998. ADAMs: focus on the protease domain. Curr. Opin. Cell Biol. 10, 654–659. Calvete, J.J., Moreno-Murciano, M.P., Sanz, L., Jurgens, M., Schrader, M., Raida, M., Benjamin, D.C., Fox, J.W., 2000. The disulfide bond pattern of catrocollastatin C, a disintegrin-like/ cysteine-rich protein isolated from Crotalus atrox venom. Protein Sci. 9, 1365–1373. Deshimaru, M., Ogawa, T., Nakashima, K., Nobuhisa, I., Chijiwa, T., Shimohigashi, Y., Fukumaki, Y., Niwa, M., Yamashina, I., Hattori, S., Ohno, M., 1996. Accelerated evolution of crotalinae snake venom gland serine proteases. FEBS Lett. 397, 83–88. Hite, L.A., Shannon, J.D., Bjarnason, J.B., Fox, J.W., 1992. Sequence of a clone encoding the zinc metalloproteinase hemorrhagic toxin e from Crotalus atrox: evidence for signal, zymogen, and disintegrin-like structures. Biochemistry 31, 6203–6211. Hite, L.A., Jia, L.G., Bjarnason, J.B., Fox, J.W., 1994. cDNA sequences for four snake venom metalloproteinases: structure, classification, and their relationship to mammalian reproductive proteins. Arch. Biochem. Biophys. 308, 182 –191. Kamiguti, A.S., Cardoso, J.L., Theakston, R.D., Sano-Martins, I.S., Hutton, R.A., Rugman, F.P., Warrell, D.A., Hay, C.R., 1991. Coagulopathy and hemorrhage in human victims of Bothrops jararaca envenoming in Brazil. Toxicon 29, 961 –972. Kamiguti, A.S., Hay, C.R.M., Zuzel, M., 1996. Inhibition of collagen-induced platelet aggregation as the result of cleavage of a2b1-integrin by the snake venom metalloproteinase jararhagin. Biochem. J. 320, 635–641. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 –685. Mandelbaum, F.R., Assakura, M.T., 1988. Antigenic relationship of hemorrhagic factors and proteases isolated from the venoms of three species of Bothrops snakes. Toxicon 26, 379–385. Mandelbaum, F.R., Assakura, M.T., Reichl, A.P., 1982. Isolation and characterization of a proteolytic enzyme from the venom of the snake Bothrops jararaca (jararaca). Toxicon 20, 955–972. Moura-da-Silva, A.M., Linica, A., Della-Casa, M.S., Kamiguti, A.S., Ho, P.L., Crampton, J.M., Theakston, R.D., 1999. Jararhagin ECD-containing disintegrin domain: expression in Escherichia coli and inhibition of the platelet-collagen interaction. Arch. Biochem. Biophys. 369, 295– 301. Murphy, G.J.P., Murphy, G., Reynolds, J.J., 1991. The origin of matrix metalloproteinases and their familial relationships. FEBS Lett. 289, 4–7. Mustard, J.F., Perry, D.W., Ardlie, N.G., Packham, M.A., 1972. Preparations of suspensions of washed platelet from humans. Br. J. Haematol. 22, 193 –204. Nagase, H., Woessner, J.F. Jr., 1999. Matrix metalloproteinases. J. Biol. Chem. 274, 21491–21494. Nakashima, K., Nobuhisa, I., Deshimaru, M., Nakai, M., Ogawa, T., Shimohigashi, Y., Fukumaki, Y., Hattori, M., Sakaki, Y., Hattori, S., Ohno, M., 1995. Accelerated evolution in the protein-coding regions is universal in crotalinae snake venom gland phospholipase A2 isozyme genes. Proc. Natl. Acad. Sci. USA 92, 5605–5609.
M.T. Assakura et al. / Toxicon 41 (2003) 217–227 Neeper, M.P., Jacobson, M.A., 1990. Sequence of a cDNA encoding the platelet aggregation inhibitor trigramin. Nucl. Acids Res. 18, 4255–4261. Ohno, M., Menez, R., Ogawa, T., Danse, J.M., Shimohigashi, Y., Fromen, C., Ducancel, F., Zinn-Justin, S., Le Du, M.H., Boulain, J.C., Tamiya, T., Menez, A., 1998. Molecular evolution of snake toxins: is the functional diversity os snake toxins associated with a mechanism of accelerated evolution? Prog. Nucl. Acid Res. Mol. Biol. 59, 307 –364. Paine, M.J.I., Desmond, H.P., Theakston, R.D.G., Crampton, J.M., 1992. Purification, cloning, and molecular characterization of a high molecular weight hemorrhagic metalloproteinase, jararhagin, from Bothrops jararaca venom. J. Biol. Chem. 267, 22869– 22876. Primakoff, P., Myles, D.G., 2000. The ADAM gene family: surface proteins with adhesion and protease activity. Trends Genet. 16, 83–87. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, second ed., Cold Spring Harbour Laboratory, Spring Harbor, NY. Shimokawa, K., Jia, L.G., Wang, X.M., Fox, J.W., 1996. Expression, activation, and processing of the recombinant snake venom metalloproteinase, pro-atrolysin E. Arch. Biochem. Biophys. 335, 283 –294.
227
Stocker, W., Grams, F., Baumann, U., Reinemer, P., Gomis-Ru¨th, F.-X., Mckay, D.B., Bode, W., 1995. The metzincins— topological and sequential relations between the astacins, adamalysins, serralysins, and matrixins (collagenases) define a super family of zinc-peptidases. Protein Sci. 4, 823 –840. Takeya, H., Oda, K., Miyata, T., Omori-Satoh, T., Iwanaga, S., 1990. The complete amino acid sequence of the high molecular mass hemorrhagic protein HR1B isolated from the venom of Trimeresurus flavoviridis. J. Biol. Chem. 265, 16068–16073. Takeya, H., Miyata, T., Nishino, N., Omori-Satoh, T., Iwanaga, S., 1993. Snake venom hemorrhagic and non-hemorrhagic metalloendopeptidases. Meth. Enzymol. 223, 365 –378. Usami, Y., Fujimura, Y., Miura, S., Shima, H., Yoshida, E., Yoshioka, A., Hirano, K., Suzuki, M., Titani, K., 1994. A 28 kDa-protein with disintegrin-like structure (jararhagin-C) purified from Bothrops jararaca venom inhibits collagen- and ADP-induced platelet aggregation. Biochem. Biophys. Res. Commun. 201, 331–339. Zhou, Q., Smith, J.B., Grossman, M.H., 1995. Molecular cloning and expression of catrocollastatin, a snake-venom protein from Crotalus atrox (western diamondback rattlesnake) which inhibits platelet adhesion to collagen. Biochem. J. 307, 411 –417.