See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/38033037
A new Factor Xa inhibitor from Amblyomma cajennense with a unique domain composition Article in Archives of Biochemistry and Biophysics · October 2009 DOI: 10.1016/j.abb.2009.10.009 · Source: PubMed
CITATIONS
READS
29
33
6 authors, including: Oscar Henrique Pereira Ramos
Janaina Souza Ventura
Atomic Energy and Alternative Energies Com…
Instituto Butantan
59 PUBLICATIONS 669 CITATIONS
10 PUBLICATIONS 270 CITATIONS
SEE PROFILE
SEE PROFILE
Paulo Ho
Ana Marisa Chudzinski-Tavassi
Instituto Butantan
Instituto Butantan
204 PUBLICATIONS 5,421 CITATIONS
88 PUBLICATIONS 1,073 CITATIONS
SEE PROFILE
SEE PROFILE
Some of the authors of this publication are also working on these related projects: Uses of cytoprotective molecules (hemolin, lipocalin and peptides) as tools for searching novel molecular targets in osteoarthitis View project
All content following this page was uploaded by Paulo Ho on 07 March 2015. The user has requested enhancement of the downloaded file.
Archives of Biochemistry and Biophysics 493 (2010) 151–156
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
A new Factor Xa inhibitor from Amblyomma cajennense with a unique domain composition I.F.C. Batista a, O.H.P. Ramos a, J.S. Ventura a, I.L.M. Junqueira-de-Azevedo b, P.L. Ho b, A.M. Chudzinski-Tavassi a,* a b
Laboratório de Bioquímica e Biofísica, Instituto Butantan, São Paulo, SP, Brazil Centro de Biotecnologia, Instituto Butantan, São Paulo, SP, Brazil
a r t i c l e
i n f o
Article history: Received 21 September 2009 and in revised form 14 October 2009 Available online 22 October 2009 Keywords: Kunitz FXa inhibitor Anticoagulant Molecular modeling Molecular dynamics
a b s t r a c t Bioactive compounds of great interest are found in the saliva of hematophagous organisms. While exploring a cDNA library derived from the salivary glands of the tick Amblyomma cajennense, a transcript that codes for a protein with unique structure (containing an N-terminal Kunitz-type domain and a C-terminus with no homology to any annotated sequences) was found. The recombinant mature form of this protein (13.5 kDa) was produced in Escherichia coli BL21 (DE3), and it was able to inhibit Factor Xa (FXa) and extend global blood clotting times in vitro and ex vivo. Static and dynamic predictions of its tertiary structure indicate regions that may be related to its FXa inhibitor function. Ó 2009 Elsevier Inc. All rights reserved.
Introduction Hemostasis control involves a complex system responsible for the maintenance of blood properties, and Factor X (FX) has a central role on this process merging the extrinsic with the intrinsic pathways leading to a common pathway. The active Factor X (FXa) specifically catalyses the conversion of prothrombin into thrombin, leading to the thrombin and fibrin generation [1,2]. This activation is physiologically performed by Factor VII/Tissue Factor (TF) and Factor IXa/Factor VIII and occurs more efficiently in the presence of phospholipids, which could be considered as a cofactor to accelerate the catalysis, and are found on activated platelets, damaged extravascular cells and procoagulant microparticles [3,4]. Tissue Factor is considered the initiator of blood coagulation, and when it is exposed to the blood, it binds circulating FVIIa. The TF:VIIa enzymatic complex catalyzes the formation of the proteases FXa and Factor IXa (FIXa) from their zymogens. FIXa converts FX into FXa, and in the presence of FVa, FXa catalyses the conversion of prothrombin into thrombin, which then converts fibrinogen into fibrin [5]. The blood coagulation process is mainly regulated by antithrombin (AT), tissue factor pathway inhibitor (TFPI) and protein
* Corresponding author. Address: Laboratório de Bioquímica e Biofísica, Instituto Butantan, Av. Vital Brasil, CEP 05503-900, 1500 São Paulo, SP, Brazil. Fax: +55 11 3726 7222x2018. E-mail address:
[email protected] (A.M. Chudzinski-Tavassi). 0003-9861/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2009.10.009
C. The main endogenous FXa inhibitor is AT, although the presence of heparin is required to improve the efficiency of this reaction [6]. TFPI is a protein with an acidic N-terminal region followed by three Kunitz-type serine protease inhibitor domains and a basic C-terminal region. TFPI affects blood coagulation by directly inhibiting FXa (using its second domain) and FVIIa/TF (via its first domain). This unique mechanism for inhibition of both FXa and FVIIa/TF makes TFPI the only physiologically functional inhibitor of the first step of blood coagulation [7–8]. Therapeutic approaches based on selective inhibitors of specific factors, especially FXa inhibitors, are potential options for clinical treatments of thrombosis [9]. FXa inhibitors are particularly attractive because the levels of serine proteases are amplified at each step. Thus, anticoagulants which target coagulation factors located at the beginning of the cascade, such as FXa, might be more effective than those directly targeting thrombin [10]. The current therapy for blood clotting disturbances includes administration of conventional unfractionated heparin (UFH) and some new agents, such as low molecular weight heparin (LMWH), FXa inhibitors and direct thrombin inhibitors [11]. Various FXa inhibitors have been isolated from hematophagous organisms such as leeches (Antistasin from Haementeria officinalis, Lefaxin from Haementeria depressa and Therostasin from Theromyzon tessulatum) [12–14] and ticks. The tick anticoagulant peptide (TAP) is a 60 amino acid protein, related to inhibitors of Kunitz family, isolated from Ornithidoros moubata [15]. Another such example is Ixolaris, a two Kunitz domain inhibitor that binds to the FXa heparin-binding exosite [16].
152
I.F.C. Batista et al. / Archives of Biochemistry and Biophysics 493 (2010) 151–156
In a previous report, the A. cajennense salivary gland transcriptome was characterized by expressed sequence tags (ESTs) [17]. The present paper describes the cloning and heterologous expression of a cDNA that encodes an FXa inhibitor from the tick Amblyomma cajennense. The gene that codes for the recombinant mature inhibitor, which was named Amblyomin-X (Amblyomma Factor Xa inhibitor), was expressed with a 6His-Tag fused at its N-terminus. Amblyomin-X is a 13,491 Da single chain protein that contains seven Cys residues. The Amblyomin-X inhibits the FXa blood clotting enzyme and prolongs global blood clotting assays.
Materials and methods Cloning, expression and purification The EST profile of A. cajennense [17] and the cDNA library construction [18] were previously described. Among the 1754 deposited sequences, a clone (EC780109) that codes for a TFPI-like protein was selected for expression. A DNA fragment encoding the mature inhibitor was amplified by PCR. The forward primer contained an XhoI restriction site and the nucleotide sequence coding for the first seven amino acids of the putative mature protein; the reverse primer was SP6, which anneals to the promoter located in the pGEM11Zf+ vector. The PCR product was sub-cloned into the pGEM-T plasmid (Promega) and, after digestion with XhoI and EcoRI, the excised insert was subcloned into the pAE expression vector [19]. The sequence of this construct was verified by automatic DNA sequencing (ABI 3100 sequencer, Applied Biosystems), and the plasmid was used to transform Escherichia coli BL21 (DE3) cells. Transformed cells were inoculated into 100 mL of 2YT/Amp and grown overnight at 37 °C/220 rpm. The culture was then diluted with 2YT/Amp medium to 1 L and grown under the same conditions until the absorbance at 600 nm reached 0.6 U. Next, IPTG was added (1 mM final concentration), and the induced culture was incubated for three additional hours. The cells were collected by centrifugation (5000 rpm/20 min/4 °C), resuspended in 50 mL of Buffer A (100 mM Tris–HCl, pH 8.0, 300 mM NaCl, 5 mM Imidazole) and disrupted in a French Press. The inclusion bodies were harvested by centrifugation and washed with 15 mL of Buffer B (Buffer A containing 2 M Urea and 10 mM b-mercaptoethanol). Amblyomin-X was solubilized overnight at room temperature with gentle shaking in 50 mL of Buffer C (Buffer A containing 8 M urea). Then, the resuspended material was diluted drop-by-drop into 2 L of Buffer A, and this solution was loaded onto a Ni2+–Sepharose column (10-mL column, Amersham Pharmacia Biotech), followed by 10 volumes of Buffer A, 10 volumes of washing buffer (Buffer A containing 60 mM Imidazole) and 5 volumes of elution buffer (Buffer A containing 1 M imidazole). Fractions (1 mL/each) were collected, pooled and dialyzed against saline and analyzed by SDS–PAGE. The protein concentration was evaluated by colorimetric assay as described by Bradford [20] and purity was estimated by SDS-gel scanning using a scanner (DNR Bio-Imaging Sistem – MIniBisPro).
Factor Xa Inhibition assays FXa inhibition was assayed in the presence of phospholipids (phosphatidylserine/phosphatidylcholine) prepared as previously described [21]. Different concentrations of Amblyomin-X (3.8–38 lM) were incubated with 0.5 lM FXa (20 min/37 °C in 50 mM Tris–HCl, pH 8.0), and inhibition was determined by FXa residual activity on the chromogenic substrate Z-D-Arg-Gly-ArgpNA (S-2765 Chromogenix), which liberates the chromophoric
group pNA (p-nitroaniline) [22]. The formed colour was photometrically monitored at 405 nm (SpectraMax 190, Molecular devices). Coagulation tests Activated partial thromboplastin time (APTT [23]), prothrombin time (PT [24]) and procoagulant activity (PCA [25])1 assays were performed using human normal plasma and commercially available kits (Diacelin for APTT and Diaplastin for PT, both from Diamed) in presence and absence of Amblyomin-X. For the in vitro APTT assay, human normal plasma (100 ll), cefalin (100 lL) and Amblyomin-X (2.5–12.5 lM previously diluted in 100 ll 0.1 M Tris–HCl pH 8.0) were mixed and incubated (3 min/ 37 °C). To initiate the coagulation process, 100 lL of 0.025 M CaCl2 was added, and the time for clot formation was recorded. To measure the in vitro PT, human normal plasma (100 ll) was incubated with different concentrations of Amblyomin-X (3.8– 18 lM previously diluted in 100 lL 0.1 M Tris–HCl pH 8.0 buffer) for 2 min at 37 °C. Then, 100 lL of the kit reagent was added, and the clotting time was measured. The PCA was assayed in the presence of Amblyomin-X (diluted in 50 mM HEPES buffer), and alternatively, apoptotic bodies (Apobodies) from cultures of Chinese hamster ovary (CHO) cells were used as an alternative source of phospholipids in this assay [26]. For ex vivo coagulation assays, Amblyomin-X was dissolved in saline and intraperitoneally administered (0, 1 or 2 mg/kg) in C57BL/6 J mice (five animals/group). Five and 24 h after drug administration, blood was collected (cardiac punction), immediately centrifuged (1000g, 10 min), and the plasma was used to measure the APTT and PT as previously described. Molecular modeling and dynamics To understand Amblyomin-X structure–function relationships, a molecular model was built using a mixed protocol based on ab initio and homology methods. Due to the unavailability of a known solved structure with sufficient similarity to the last 50 C-terminal residues of Amblyomin-X, an ab initio method, implemented in Rosetta [27], was employed for its tertiary structure prediction. Along with other suitable templates corresponding to the N-terminal His-Tag (Roadblock/LC7 dynein, PDB: 1Y4O) and Kunitz-type domains (Human Bikunin, PDB: 1BIK; Chymotrypsin inhibitor from Bungarus fasciatus, PDB: 1JC6; and human Kunitz domain 5 of collagen type VI alpha 3 chain, PDB: 1KTH), the predicted C-terminus was used for a reconstruction of the whole structure of the recombinant protein by homology modeling based on spatial restraints implemented in Modeler 9v4 [28]. The model with the best variable target function was selected from 100 starting models and submitted to energy minimization (1 step of steepest descent each 100 steps of conjugate gradient), hydration (19.969 water molecules), ion additions and system charge neutralization (37 Na+ and 28 Cl ions). The system was submitted to molecular dynamics simulation (MD) during 5 ns (box dimensions: 92.0472 70.8347 96.5105 Å). The OPLSAA force field was employed for force calculations. Energy minimization, system generation and molecular dynamics were performed using the GROMACS package [29]. The resulting trajectory from MD was analyzed concerning the root mean square deviation of backbone atom coordinates, radius of gyration, secondary structure, solvent accessible surface area and salt bridges. The quality of the
1 Abbreviations used: PCA, procoagulant activity test; APTT, activated partial thromboplastin time; TP, prothrombin time; PS, phosphatidylserine; PC, phosphatidylcholine; CHO, Chinese hamster ovarian cells; AB, apoptotic bodies; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; MD, molecular dynamics simulation.
153
I.F.C. Batista et al. / Archives of Biochemistry and Biophysics 493 (2010) 151–156
model was checked with Procheck [30], Verify 3D [31] and WhatIf [32] software before and after MD.
Results Amblyomin-X production and in vitro assays Amblyomin-X was overexpressed in E. coli after IPTG induction and solubilized under denaturing conditions (8 M urea and b-mercaptoethanol). The refolded protein was purified using a nickel-chelating resin (Sepharose Fast Flow, GE), and the fractions (Fig. 1) were dialyzed against saline solution (150 mM NaCl), pooled and stored at 20 °C until use. The table 1 summarizes the purification process. The purified protein was able to inhibit FXa activity in the presence of a phospholipid mixture, in a dose dependent manner (Fig. 2). Recombinant Amblyomin-X (12.5 lM) was able to extend the APTT of normal human plasma (3.2-fold increase), and the PT (4.5-fold increase). In the PCA assay, the effect of the protein (2.5 lM) was studied in presence of a PS/PC mixture or CHO apoptotic bodies. Amblyomin-X increased PCA 1.6-fold in the presence of PS/PC and 2-fold using CHO apoptotic bodies (Table 2). During the ex vivo experiments, when the animals were treated with 1.0 mg/kg Amblyomin-X, no significant alterations were
Fig. 2. Factor Xa inhibition curve in the presence of phospholipids and with increasing concentrations of Amblyomin-X.
Table 2 Global blood clotting assays.
Ctrl 2.5 lM 3.75 lM 6.25 lM 12.5 lM
Ctrl 2.5 lM
Ctrl 1 mg/kg 2 mg/kg
Ctrl 1 mg/kg 2 mg/kg * **
APTT in vitro
PT
43.5 ± 0.9 60.8 ± 2.0* 74.9 ± 1.8* 92.8 ± 1.7* 138.7 ± 2.7*
14.3 ± 0.5 18.4 ± 1.1 31.6 ± 0.9* 65.2 ± 2.0* 89.3 ± 0.9*
PCA PS:PC
Apobodies
128.6 ± 1.7 209.9 ± 1.16*
127.2 ± 1.6 266.7 ± 1.7*
APTT (ex vivo) 5h
24 h
35.2 ± 1.2 31.3 ± 2.2 47.4 ± 2.1**
35.6 ± 2.4 32.8 ± 0.07 43.1 ± 3.1**
PT (ex vivo) 5h
24 h
22.7 ± 2.4 21.5 ± 1.4 22.0 ± 1.7
19.1 ± 1.7 22.1 ± 1.7 19.8 ± 1.8
p < 0.001. p < 0.01.
observed on the APTT. However, when the animals received 2.0 mg/kg, an increase in the APTT was observed. Conversely, Amblyomin-X treatment did not cause PT alterations (Table 2). Fig. 1. Fifteen percentage SDS–PAGE electrophoresis. (1) Molecular mass marker, (2) non-induced BL21-DE3 culture, (3) IPTG induced BL21-DE3 culture, (4) solubilized inclusion bodies – loading material onto Ni–Sepharose, (5) pool of the fractions eluted with 1 M imidazole, (6) pool of the fractions eluted with 1 M imidazole after dialysis against saline, under non-reduce conditions, and (7) pool of the fractions eluted with 1 M imidazole after dialysis against saline, under reduce conditions by b-mercaptoethanol.
Table 1 Amblyomin-X purification from BL21(DE3) inclusion bodies’. Purification step
Total protein (mg)
Purity (%)
Yield (%)
Protein recovered in inclusion bodies Affinity chromatography After dialysis
4.4
1
100
1.0 0.92
2.5 2.3
44 42
Molecular modeling and dynamics For the initial model construction, a mixed procedure involving ab initio prediction of the C-terminal 50 residues (with no solved homologous structure found) of Amblyomin-X, followed by homology modeling of the whole protein, was proposed. The quality evaluation of models before and after MD indicates that the latter have better overall structure, abolishing bad contacts and bad chemical atomic environments. For the initial model, 87.7% of the residues were found in the most favorable region of the phi and psi dihedral angles of a Ramachandran plot, and the remaining 12.3% were found in additional allowed regions. In the final model, 89.6% and 10.4% were found in the same regions, respectively. For both cases, no residues were found in generously allowed or disallowed regions.
154
I.F.C. Batista et al. / Archives of Biochemistry and Biophysics 493 (2010) 151–156
During the MD simulation, a tendency of increased b-strand content with a concomitant decrease in a-helices (until close to 1.75 ns) was observed when b-strand content started to decrease and a-helix content started to increase, tending to stabilize around 19 and 12 residues, respectively. The final model shows 22.2% of the residues forming b-strands, 9.4% a-helices and 68.4% random coil structures. Compared to the starting structure, the root mean square deviation of the backbone atom positions found its stable peak after 1 ns, while there was a tendency for the structure to assume a more compact conformation throughout the simulation as indicated by the decrease in its radius of gyration and solvent accessible surface area.
The final model (available under request) exhibits close similarity with Kunitz-type domains (residues 10–67). The N-terminal His-Tag (residues 1–9) folds towards the interface between the Kunitz-type domain and the C-terminal domain (residues 68–117) of the molecule, establishing one hydrogen bond and one salt bridge with the former. The predicted model suggests a C-terminal domain comprising a short helix and a short beta sheet with two strands. Concerning the entire structure, salt bridges, the hydrogen bond network and hydrophobic contacts all seem to have relevance for structure stabilization. Four protein bands were observed after SDS–PAGE of Amblyomin-X. All of them were sequenced, and only fragments of the
Fig. 3. (a) Structural alignment of selected Kunitz domains using Cys as references: Bikunin domain 2 (Bikunin, PDB: 1BIK), TFPI – second Kunitz-type domain 2 (TFPI, PDB: 1ADZ), Tick anticoagulant protein (TAP, PDB: 1TAP) and Amblyomin-X. Key inhibitory residues (grey background) and Cys (boxes). (b) Sequence alignment involving Amblyomin-X (Kunitz-type domain), Antistasin (AAA29193.1) and Ghilanten (AAB21233.1). The Antistasin reactive site is indicated by the box. (c) Sequence alignment involving Amblyomin-X (Kunitz-type domain) and Ixolaris (AAK83022.1), Cys represented with grey background.
I.F.C. Batista et al. / Archives of Biochemistry and Biophysics 493 (2010) 151–156
Amblyomin-X were found in them, implicating different quaternary forms such as monomers, dimers, trimers and tetramers. When the material eluted from the nickel column was loaded onto a C4 column, a single 13.5 kDa protein was eluted, suggesting that the conditions of this chromatography were favorable for complex dissociation. The inhibitory activity of Amblyomin-X was also tested on trypsin, thrombin, plasmin, tissue and plasma kallikreins, but none of their activities were affected (results not shown). Discussion In recent years, there has been an increase in the number of described tick anticoagulant proteins, including TAP (Tick Anticoagulant Peptide from O. moubata [33]), Boophilin (from Boophilus microplus; [34]), Ixolaris and Penthalaris (from Ixodes scapularis [16,35]), and reports describing the cloning and expression of active molecules [34,36,16]. TAP and Boophilin are examples of Kunitz domain proteins with structures solved in complex with target proteases (FXa and thrombin, respectively). A primary structure analysis was performed with Kunitz domains using cystine bonds as references (Fig. 3a: bikunin domain 2, TFPI second Kunitz-type domain 2, TAP and Amblyomin-X). The TFPI extorts a very crucial hole in the coagulation system [6], it is a multivalent Kunitz-type inhibitor, inhibits FXa via its second domain and factor VIIa/TF complex via its first domain. TFPI is able to inhibit the extrinsic pathway and affects both PT and APTT tests [37], which are a useful model to verify the influence of the inhibitors in the blood clotting cascade [38]. Amblyomin-X is able to extend the global coagulation tests in vitro, however when normal mice were injected with the Amblyomin-X (2 mg/kg) just the APTT was altered. To explain the differences higher doses Amblyomin-X should be injected in the animals to check if the same effects could be observed. The first domain of Amblyomin-X is 40% similar to the Kunitz domains present in TFPI and Ixolaris. During the in vitro experiments, an increase in the coagulation parameters of PT and APTT was observed when plasma was incubated with Amblyomin-X,
155
suggesting that the FVIIa/TF complex and prothrombin conversion might be inhibited. However, the APTT performed using the plasma of animals treated with 2 mg/kg of Amblyomin-X was weakly modified when compared to untreated mice. Perhaps the differences observed are related to the structural characteristics of the Amblyomin-X Kunitz domain as it is not identical to the related TFPI sequence. Regardless, to confirm if Amblyomin-X exerts the same activity as TFPI, higher doses of Amblyomin-X should be administered to the animals in future experiments. Antistasin (a potent FXa inhibitor isolated from H. officinalis) and Ghilantens (a potent anticoagulant and antimetastatic protein from Haementeria ghilianii) likely display different inhibitory efficiencies because of their C-termini [39]. To compare AmblyominX, the primary structures of Antistasin and Ghilantens were aligned with Amblyomin-X. Although the Cys residues appear conserved, Amblyomin-X did not display other similar regions in the alignment (Fig. 3b). Among the tick anticoagulant proteins, TAP is probably the best studied inhibitor and the one with the most potent affinity (0.18 nM). The unexpected binding mode of its single Kunitz domain to FXa [34] reveals the positioning of an N-terminal residue side chain (Tyr) in the S1 pocket of FXa. Boophilin is a two Kunitz-type domain inhibitor of thrombin, trypsin and plasmin. Its complex with thrombin [33] reveals that the first Kunitz domain (K-1) interacts with the active site of the serine protease. In this case, the Arg17 side chain (second residue from the N-terminus) occupies the S1 pocket. The second Kunitz domain (K-2) has several negatively charged residues located in the sequence ESEEECE (residues 130–136, located mainly in K-2 C-terminal helix) that seem to have relevant roles concerning the binding of this domain to exosite 1 of the protease. Comparing their binding modes, these tick inhibitors present some obvious similarities such as the docking of their N-termini into the serine protease active sites of their target proteases. Conversely, the binding mode of bovine pancreatic trypsin inhibitor (BPTI), which binds to trypsin by a positively charged residue present at the tip of the salient loop that is characteristic of Kunitz domains, differs considerably from that observed for TAP and
Fig. 4. Schematic representation of Amblyomin-X structure. (a) Main chain (ribbons) and side chains (sticks). Relevant structural regions: Kunitz domain, C-terminal domain, negative residues cluster, His-Tag, hydrophobic core and salt bridges (dashed lines). (b) Putative residues relevant to FXa inhibition (Lys residues represented by sticks). Nglycosylation sites predicted by prosite are indicated by a star symbol.
156
I.F.C. Batista et al. / Archives of Biochemistry and Biophysics 493 (2010) 151–156
Boophilin. This suggests a possible case of divergent evolution of Kunitz domains that led to different mechanisms of serine proteases inhibition. However, analyzing complexes of Kunitz-type inhibitors demonstrates that the presence of a positive electrostatic potential in the corresponding interface region, which is usually surrounded by a negatively charged electrostatic potential, is a conserved feature. Amblyomin-X constitutes a unique example among the anticoagulant proteins described to date. It shares similarity with Kunitz domains of tick anticoagulant proteins but simultaneously harbors a C-terminal domain (last 50 residues) that scarcely resembles other known proteins as indicated by BLAST searches against non-redundant sequences. Similarly, a search using the predicted structure for this segment against the DALI database yields no hits. At the same time that this fact brings exciting possibilities for the expansion of knowledge, it makes potential predictions concerning structure–function relationships more difficult. Despite this, the sequence of Amblyomin-X was aligned to the Ixolaris, another tick FXa inhibitor (Fig. 3c). According to this information, four possible models were proposed for the Amblyomin-X/FX complex. The first hypothesis proposes that Amblyomin-X binding resembles that observed for the TAP/FXa and Boophilin/Thrombin complexes. Taking into account our experimental results and molecular dynamics simulation, it is likely that the N-terminal His-Tag on recombinant Amblyomin-X interferes with its activity, and this is in agreement with the high concentrations required for FXa inhibition (Fig. 4a). A second hypothesis is that the C-terminal domain of Amblyomin-X, which has a strong negatively charged electrostatic surface, could bind FXa exosites (positively charged) to produce the experimentally observed inhibition. Despite the fact that this domain is not predicted to adopt a Kunitz-type conformation, the cluster of negatively charged residues in the Amblyomin-X C-terminus may function similarly to the second Kunitz domains of Boophilin and Ixolaris which bind to the positively charged exosites of their target serine proteases. The third hypothesis is that the binding of the Amblyomin-X N-terminus to the active site of FXa is accompanied by additional binding of its negative C-terminal domain to positive cluster of FXa’s exosites. The last hypothesis is that Lys30, present on the characteristic loop of Kunitz domains of serine protease inhibitors, could bind to the active site of FXa. However, in this case, it is possible that a predicted N-glycosylation site found close to this residue could impair its function in the native protein (Fig. 4b). The expression of domains separately may bring light to this issue. A new construct excluding the His-Tag has been developed by our group, and the expression of truncated and mutated forms is the next natural step. Finally, the described homology modeling/ ab initio mixed protocol suggests that this could be a useful method for predicting the structure of proteins that contain both novel regions and regions homologous to proteins with known structures. Acknowledgment This work was supported by FAPESP, CNPq, CAT/CEPID, INCTTOX.
References [1] B.A. Lwaleed, A.J. Cooper, D. Voegeli, K. Getliffe, Biol. Res. Nurs. 9 (2007) 97–107. [2] K. Borensztajn, M.P. Peppelenbosch, C.A. Spek, Trends Mol. Med. 14 (2008) 429–440. [3] R.G. Di Scipio, M.A. Hermodson, E.W. Davie, Biochemistry 16 (1977) 5253– 5260. [4] O. Morel, N. Morel, J.M. Freyssinet, F. Toti, Platelets 19 (2008) 9–23. [5] B. Engelmann, T. Luther, I. Müller, Thromb. Haemost. 89 (2003) 3–8. [6] J.C. Rau, L.M. Beaulieu, J.A. Huntington, F.C. Church, J. Thromb. Haemost. 5 (2007) 102–115. [7] S.A. Maroney, A.E. Mast, Transfus. Apher. Sci. 38 (2008) 9–14. [8] G.J. Broze Jr., Blood Coagul. Fibrinolysis 9 (1998) S89–S92. [9] K.M. Bromfield, N.S. Quinsey, P.J. Duggan, R.N. Pike, Chem. Biol. Drug Des. 68 (2006) 11–19. [10] J. Ansell, J.J. Caro, M. Salas, R.J. Dolor, W. Corbett, A. Hudnut, S. Seyal, N.D. Lordan, I. Proskorovsky, G. Wygant, Am. J. Med. Qual. 22 (2007) 327–333. [11] C.J. McCann, I.B. Menown, Vasc. Health Risk Manag. 4 (2008) 305–313. [12] E. Nutt, T. Gasic, J. Rodkey, G.J. Gasic, J.W. Jacobs, P.A. Friedman, E. Simpson, J. Biol. Chem. 263 (1988) 10162–10167. [13] F. Faria, E.M. Kelen, C.A. Sampaio, C. Bon, N. Duval, A.M. Chudzinski-Tavassi, Thromb. Haemost. 82 (1999) 1469–1473. [14] V. Chopin, M. Salzet, J. Baert, F. Vandenbulcke, P.E. Sautiére, J.P. Kerckaert, J. Malecha, J. Biol. Chem. 275 (2000) 32701–32707. [15] M.P. Neeper, L. Waxman, D.E. Smith, C.A. Schulman, M. Sardana, R.W. Ellis, L.W. Schaffer, P.K. Siegl, G.P. Vlasuk, J. Biol. Chem. 265 (1990) 17746– 17752. [16] I.M. Francischetti, J.G. Valenzuela, J.F. Andersen, T.N. Mather, J.M. Ribeiro, Blood 99 (2002) 3602–3612. [17] I.F. Batista, A.M. Chudzinski-Tavassi, F. Faria, S.M. Simons, D.M. Barros-Batestti, M.B. Labruna, L.I. Leão, P.L. Ho, I.L. Junqueira-de-Azevedo, Toxicon 51 (2008) 823–834. [18] I.L. Junqueira-de-Azevedo, S.H. Farsky, M.L. Oliveira, P.L. Ho, J. Biol. Chem. 276 (2001) 39836–39842. [19] C.R. Ramos, P.A. Abreu, A.L. Nascimento, P.L. Ho, Braz. J. Med. Biol. Res. 37 (2004) 1103–1109. [20] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254. [21] Y. Barenholz, D. Gibbes, B.J. Litman, J. Goll, T.E. Thompson, R.D. Carlson, Biochemistry 16 (1977) 2806–2810. [22] E.M. van Wijk, J.L. Smit, Thromb. Res. 34 (3) (1984) 263–268. [23] J.J. Hoffmann, P.N. Meulendijk, Thromb. Haemost. 39 (1978) 640–645. [24] A.J. Quick, Am. J. Clin. Pathol. 45 (1966) 105–109. [25] N.S. Key, A. Slungaard, L. Dandelet, S.C. Nelson, C. Moertel, L.A. Styles, F.A. Kuypers, R.R. Bach, Blood 91 (1998) 4216–4223. [26] B.B. Sorensen, L.V. Rao, D. Tornehave, S. Gammeltoft, L.C. Petersen, Blood 102 (2003) 1708–1715. [27] C.A. Rohl, C.E. Strauss, K.M. Misura, D. Baker, Methods Enzymol. 383 (2004) 66–93. [28] A. Sali, Mol. Med. Today 1 (1995) 270–277. [29] S.D. Van Der, E. Lindahl, B. Hess, G. Groenhof, A.E. Mark, H.J. Berendsen, J. Comput. Chem. 26 (2005) 1701–1718. [30] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Cryst. 26 (1993) 283–291. [31] J.U. Bowie, R. Luthy, D. Eisenberg, Science 253 (5016) (1991) 164–170. [32] G. Vriend, J. Mol. Graph. 8 (1) (1990) 52–56. [33] L. Waxman, D.E. Smith, K.E. Arcuri, G.P. Vlasuk, Science 248 (4955) (1990) 593–596. Erratum in: Science 1990; 248:1473.. [34] S. Macedo-Ribeiro, C. Almeida, B.M. Calisto, T. Friedrich, R. Mentele, J. Stürzebecher, P. Fuentes-Prior, P.J. Pereira, PLoS ONE 3 (2008) e1624. [35] I.M. Francischetti, T.N. Mather, J.M. Ribeiro, Thromb. Haemost. 91 (2004) 886– 898. [36] A. Wei, R.S. Alexander, J. Duke, H. Ross, S.A. Rosenfeld, C.H. Chang, J. Mol. Biol. 283 (1998) 147–154. [37] Han Xin, Thomas J. Girard, Pamela Baum, Dana R. Abendschein, George J. Broze Jr., Arterioscler. Thromb. Vasc. Biol. 19 (1999) 2563–2567. [38] N. Kriz, C.S. Rinder, H.M. Rinder, Clin. Lab. Med. 29 (2) (2009) 159–174. [39] D.L. Danalev, L.T. Vezenkov, B. Grigorova, Bioorg. Med. Chem. Lett. 15 (19) (2005) 4217–4220.
