Crotalid Snake Venom Subproteomes Unraveled by the Antiophidic

Mar 6, 2009 - Neves-Ferreira , A. G. C.; Cardinale , N.; Rocha , S. L. G.; Perales , J.; Domont , G. B. Isolation and characterization of DM40 and DM4...
0 downloads 0 Views 3MB Size
Download PDF
Crotalid Snake Venom Subproteomes Unraveled by the Antiophidic Protein DM43 Surza L. G. Rocha,†,‡,§,# Ana G. C. Neves-Ferreira,†,§,# Monique R. O. Trugilho,†,§ Alex Chapeaurouge,†,§ Ileana R. Leo ´ n,†,§ Richard H. Valente,†,§ Gilberto B. Domont,‡,§ and Jonas Perales*,†,§ Laborato´rio de Toxinologia, Pavilha˜o Ozo´rio de Almeida, Instituto Oswaldo Cruz, Fiocruz, Av. Brasil, 4365, 21040-900 Rio de Janeiro, RJ, Brazil, Laborato´rio de Quı´mica de Proteı´nas, Departamento de Bioquı´mica, Instituto de Quı´mica, Universidade Federal do Rio de Janeiro, 21949-900 Rio de Janeiro, RJ, Brazil, Rede Proteoˆmica do Rio de Janeiro Received November 10, 2008

Snake venoms are mixtures of proteins and peptides with different biological activities, many of which are very toxic. Several animals, including the opossum Didelphis aurita, are resistant to snake venoms due to the presence of neutralizing factors in their blood. An antihemorrhagic protein named DM43 was isolated from opossum serum. It inhibits snake venom metalloproteinases through noncovalent complex formation with these enzymes. In this study, we have used DM43 and proteomic techniques to explore snake venom subproteomes. Four crotalid venoms were chromatographed through an affinity column containing immobilized DM43. Bound fractions were analyzed by one- and two-dimensional gel electrophoresis, followed by identification by MALDI-TOF/TOF mass spectrometry. With this approach, we could easily visualize and compare the metalloproteinase compositions of Bothrops atrox, Bothrops jararaca, Bothrops insularis, and Crotalus atrox snake venoms. The important contribution of proteolytic processing to the complexity of this particular subproteome was demonstrated. Fractions not bound to DM43 column were similarly analyzed and were composed mainly of serine proteinases, C-type lectins, C-type lectin-like proteins, L-amino acid oxidases, nerve growth factor, cysteine-rich secretory protein, a few metalloproteinases (and their fragments), and some unidentified spots. Although very few toxin families were represented in the crotalid venoms analyzed, the number of protein spots detected was in the hundreds, indicating an important protein variability in these natural secretions. DM43 affinity chromatography and associated proteomic techniques proved to be useful tools to separate and identify proteins from snake venoms, contributing to a better comprehension of venom heterogeneity. Keywords: Snake • venom • toxin • metalloproteinase • inhibitor • proteomics • mass spectrometry

Introduction Systematically describing snake venom components is just the very first, though critical, step toward better understanding and treating human envenomation, which is still an important public health hazard in tropical developing countries.1 Worldwide, the true estimate of snake-bites per year may exceed 5 million, including more than 100 000 deaths.2 Particularly in Brazil, the National Ministry of Health reported a total number of 28 321 cases of envenomation in 2005, which is certainly an underestimation.3 As reviewed by Koh and colleagues,4 the exploration of snake venom complexity may additionally provide natural biomolecules “that can be used as molecular probes in human health * To whom correspondence should be addressed. Phone: +55-21-25984393ext 215. Fax: +55-21-2590-9490. E-mail: [email protected]. † Instituto Oswaldo Cruz, Fiocruz. ‡ Universidade Federal do Rio de Janeiro. § Rede Proteoˆmica do Rio de Janeiro. # These authors have contributed equally to this work. 10.1021/pr800977s CCC: $40.75

 2009 American Chemical Society

and disease”. Currently, five toxin-based drugs are approved by the U.S. Food and Drug Administration for treating hypertension, heart disease/coagulopathies, severe chronic pain, and diabetes; five others directed at heart diseases and cancer therapies are currently undergoing clinical trials.5 Historically, most studies leading to the discovery of the toxins referred to above have made use of classical protein chemistry analytical techniques, which involve the single isolation of proteins/peptides, following a slower reductionist methodology.6 Instead, the proteomic approach may increase the speed with which new toxins can be identified, allowing for the study of several molecules simultaneously.7 Over the past few years, the Toxinology field has witnessed an extraordinary growth in the number of articles employing proteomic techniques. The current state of snake venom proteomics was discussed recently in an excellent review published by Fox and Serrano.8 It is now increasingly clear that, although they apparently do not belong to very many different classes, venom components can be highly diversified in terms of postJournal of Proteome Research 2009, 8, 2351–2360 2351 Published on Web 03/06/2009

research articles translational modifications and dynamic range. The analysis of such complex protein mixtures requires technologies with very good resolving power and sensitivity. Presently, no single analytical method can fully meet this demand and a combination of different techniques for separating and identifying the polypeptides must be used. As a general rule, venom decomplexation before mass spectrometry seems to be the best strategy for maximum proteome coverage.9 Very few papers have chosen to use prefractionation procedures for focusing on a special group of proteins, such as a Sepharose-Con A affinity column to select for glycosylated toxins10 or size exclusion chromatography for purifying low molecular mass peptides from venom.11 Although, when using crude venoms, some studies have used different stains, lectins, and antibodies to specifically visualize post-translational modifications and toxin-specific subproteomes.12,13 In the current investigation, a natural antiophidic protein named DM43, isolated from the serum of the opossum Didelphis aurita14 (formerly Didelphis marsupialis),15 was used for the first time as an affinity tag to reduce the complexity of snake venoms. This target-driven approach allowed for the specific depletion of the very abundant family of metalloproteinases (SVMPs) from crotalid venoms, facilitating the analysis of other toxins in these complex secretions, as well as of the SVMPs themselves.

Experimental Section Snake Venoms. Pooled lyophilized venoms from Bothrops atrox, Bothrops jararaca, and Bothrops insularis were kindly provided by Instituto Butantan (Sa˜o Paulo, Brazil). Pooled lyophilized Crotalus atrox venom was purchased from the Miami Serpentarium (Miami, FL). Affinity Chromatography. Following the manufacturer’s instructions, a 1 mL Hitrap NHS-activated affinity column (GE Healthcare, Chalfont St Giles, U.K.) was coupled with 7 mg of DM43.16 The preparation used was determined to be homogeneous based on the criteria of electrophoresis, Edman N-terminal sequencing, and mass spectrometry. With the use ¨ KTA Purifier FPLC system (GE Healthcare), nonsaturatof an A ing amounts of venom (3 mg) were separately chromatographed through the affinity column, with protein detection set at 280 nm. The column was initially washed with 20 mM Tris-HCl, pH 7.5, and 20 mM CaCl2. Bound fractions were eluted with 100 mM Glycine-HCl, pH 2.7, and 20 mM CaCl2 and collected over 1 M Tris to neutralize the pH. Bound and nonbound fractions (0.5 mL/tube) were collected on ice at a flow rate of 1 mL/min and their protein content was determined by the Folin-Lowry method using bovine serum albumin (BSA) as a standard.17 These fractions were precipitated in 10% ice-cold TCA (trichloroacetic acid) for 1 h,18 and after two washes with cold acetone, the pellets were air-dried and stored at -20 °C for further analysis. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE). TCA-precipitated samples were first analyzed by 12% SDS-PAGE in the mini-Protean II system (BioRad Laboratories, Hercules, CA) under reducing conditions and using 4% stacking gels.19 Low molecular weight markers were from GE Healthcare and the gels were stained with 0.2% Coomassie blue R250. Two-Dimensional Gel Electrophoresis (2-DE). Samples were analyzed using 18 cm Immobiline DryStrips (IPG) pH 4-7 (GE Healthcare) in the first dimension and homogeneous SDSPAGE (15%) in the second dimension.20 Briefly, TCA-precipi2352

Journal of Proteome Research • Vol. 8, No. 5, 2009

Rocha et al. tated fractions (1 mg) were solubilized overnight in 350 µL of rehydration buffer [8 M urea, 2% CHAPS (3-[(3-Cholamidopropyl)dimethylammonio]-1-propanesulfonate), 1% (v/v) IPG buffer 4-7, 0.002% BPB (bromophenol blue), and 100 mM DTT (dithiothreitol)] and used for rehydration of the IPG strips for 12 h at 30 V. Isoelectric focusing (IEF) was performed on the IPGPhor II system (GE Healthcare) for 68 000 Vh at 20 °C. After IEF, each strip was initially incubated for 15 min in 10 mL of 50 mM Tris-HCl, pH 8.8, 6 M urea, 30% (v/v) glycerol, 2% SDS, 0.002% BPB, and 100 mg DTT, followed by a second 15 min incubation step in the same buffer solution, replacing DTT with 400 mg of iodoacetamide. Strips were then rinsed in Trisglycine electrode buffer, transferred to homogeneous 15% SDSPAGE, and overlaid with 0.5% agarose in running buffer containing BPB. Molecular weight markers were mixed (v/v) with 1% agarose and applied to filter papers positioned above the gels, close to the anodic end of the strips. Gels were run in a DALTSix system (GE Healthcare) at 2.5 W/gel for 30 min, then at 100 W until the dye front reached the gel bottom. At this time, the run was extended for an additional 30% of the completed run time. Additionally, DM43 bound-fractions were also analyzed by 2-DE under nonreducing conditions. All gels were fixed overnight in 40% ethanol and 10% acetic acid, washed for 30 min with fresh fixing solution, stained with 0.2% Coomassie blue R-250 in 40% ethanol and 10% acetic acid for 3 h, and partially destained with fresh fixing solution. Complete destaining was obtained through incubation in 1% acetic acid for several days. Gels were digitalized using a conventional light scanner and image analysis was performed using either the software ImageMaster 2D Elite or ImageMaster Platinum (GE Healthcare). For visualization of glycosylated proteins, 7 cm IPG strips (pH 4-7) were loaded with 300 µg of affinity-derived samples from B. jararaca venom, with IEF conditions scaled down to 17 000 Vh. The second dimension was run on 15% SDS-PAGE gels using the MiniProtean III system (BioRad). Staining was performed with the GelCode Glycoprotein Staining Kit (Pierce Biotechnology, Rockford, IL) following the manufacturer’s instructions and gels were scanned as previously described. Western-Blotting and Edman Degradation. The bound fraction from B. insularis venom was submitted to SDS-PAGE as previously described and transferred to Immobilon-P PVDF membrane (Millipore, Bedford, MA) at 100 V for 60 min using 25 mM Tris, 192 mM glycine, and 20% methanol as blotting buffer (pH 8.3). After staining for 2 min with 0.1% Coomassie blue R250 in 50% methanol, the membrane was briefly destained in 50% methanol, protein bands were excised, and the Nterminal amino acid sequences were determined by Edman degradation carried out in a Shimadzu PSQ-23A protein sequencer (Shimadzu, Kyoto, Japan). Searches for sequence homology were performed with the BLAST program.21 In-Gel Trypsin Digestion of Proteins. Protein spots were excised from gels, destained with 25 mM ammonium bicarbonate and 50% acetonitrile, dehydrated with acetonitrile, and digested with trypsin (166 ng/15 µL of 25 mM ammonium bicarbonate) for 16-24 h at 37 °C, using the 96-well ZipPlate micro-SPE and the MultiScreen vacuum manifold system (Millipore). After digestion, peptides were extracted from the gel by several washes with 1% formic acid under vacuum and further eluted from the C18 resin with 1% formic acid in 60% methanol. The hydrolysates were concentrated by vacuum centrifugation to approximately 10 µL and stored at -20 °C until use. Gel pieces from a “blank” region and from the BSA

Crotalid Venom Subproteomes Unraveled by DM43 molecular mass marker were used as negative and positive controls, respectively. MALDI-TOF/TOF MS (Matrix-Assisted Laser Desorption Ionization Time-of-Flight/Time-of-Flight Mass Spectrometry). Roughly 0.3 µL of the sample solution was mixed with an equal volume of a saturated matrix solution [10 mg/mL R-cyano-4-hydroxycinnamic acid (Aldrich, Milwaukee, WI) in 50% acetonitrile/0.1% trifluoroacetic acid] on the target plate and allowed to dry at room temperature. Raw data for protein identification were obtained on the 4700 Proteomics Analyzer (Applied Biosystems, Foster City, CA). Both MS and MS/MS data were acquired with a neodymium-doped yttrium aluminum garnet (Nd:YAG) laser with a 200-Hz repetition rate. Typically, 1600 shots were accumulated for spectra in the MS mode, while 2400 shots were accumulated for spectra in the MS/MS mode. Up to six of the most intense ion signals with a signal-to-noise ratio above 30 were selected as precursors for MS/MS acquisition, excluding common trypsin autolysis peaks and matrix ion signals. External calibration in MS mode was performed using a mixture of four peptides: des-Arg1-Bradykinin (m/z ) 904.468); angiotensin I (m/z ) 1296.685); Glu1fibrinopeptide B (m/z ) 1570.677); and ACTH (18-39) (m/z ) 2465.199). MS/MS spectra were externally calibrated using known fragment ion masses observed in the MS/MS spectrum of angiotensin I. Following data acquisition, a peak list was obtained from the MS/MS raw data using the “Peaks to Mascot” function in the 4000 Series Explorer software (Applied Biosystems). Database Search. Uninterpreted tandem mass spectra were searched against the nonredundant protein sequence database from the National Center for Biotechnology Information (NCBI) using the Mascot (version 2.1) MS/MS ion search tool (http:// www.matrixscience.com). In addition, MS/MS data that resulted from spots of the B. jararaca venom were searched against a curated in-house B. jararaca database (907 sequence entries).22 The search parameters were as follows: no restrictions on protein molecular weight, one tryptic missed cleavage allowed, nonfixed modifications of methionine (oxidation) and cysteine (carbamidomethylation); pyroglutamate formation at the N-terminal glutamine of peptides with no other posttranslational modifications taken into account. Peptide mass tolerance in the searches was 0.8 Da for MS spectra and 0.6 Da for MS/MS spectra. Peptides were considered to be identified when the scoring value exceeded the identity or extensive homology threshold value calculated by Mascot. In cases of protein identification based on a single peptide, the minimum threshold of the probability based Mascot score was 40. Otherwise, mass spectra with lower scores, but presenting a reasonable tandem mass spectrum, were manually verified.23 To define the SVMP domain coverage, the metalloproteinase sequences obtained were aligned against full-length jararhagin (accession number CAA48323) using the program Multalin available at http://bioinfo.genotoul.fr/multalin/multalin.html.24

Results and Discussion More than 90% of all venomous snake bites occurring in Brazil involve species of the genus Bothrops.25 In the current proteomic study, venoms from the following Brazilian snakes were analyzed: B. atrox, a species that accounts for the majority of envenomations in the Amazon region; B. jararaca, which is responsible for most accidents in the southeast region of the country; and B. insularis, a phylogenetically close sister of B. jararaca and endemic species from a small coastal island whose

research articles toxinology is still poorly understood. Additionally, we also studied C. atrox, the most medically important species from North America whose venom induces a pathology similar to that induced by B. jararaca venom.26 Immobilized DM43 affinity chromatography was used as a depletion technique to selectively remove the abundant metalloproteinases from the venoms referred to above, enriching them for proteins expressed at lower levels. Figures 1-4B show the typical chromatographic profiles. If it is assumed that the areas under the curves at 280 nm of the bound and nonbound chromatographic fractions represent the total amount of protein in each venom (100%), then the following percentages of proteins interacting with the column are obtained: B. atrox (41.0%), B. insularis (30.5%), C. atrox (33.5%), and B. jararaca (19.2%). These values are suggestive of the relative metalloproteinase content of the venoms and are in accordance with literature describing them as highly proteolytic.27,28 Particularly in the case of the B. insularis proteome, a recent publication has shown that 30% of all identified 2-DE protein spots were assigned to the metalloproteinase family.29 On SDS-PAGE, the DM43-bound fraction patterns obtained using the various venoms were very similar, showing two main bands migrating in the 43-67 kDa region and a third fainter band between 20.1 and 30 kDa (Figures 1-4B, Insets, lane 2). The nonbound fractions were more heterogeneous, including several bands with different staining intensities, which ranged from 94 to 14.4 kDa (Figures 1-4B, Insets, lane 1). In accordance with previously published work using crude venoms,12,30 proteins from both the bound and nonbound venom fractions presented isoelectric points (pI) focused along the pH range from 3 to 10 (data not shown). However, since a majority of the spots were concentrated between pH 4 and 7 (Figures 1-4C,D), this higher resolution 18 cm IPG strip was chosen for further studies. Metalloproteinases and their fragments constituted the major components of venom fractions that interacted with the DM43 column (Figures 1-4D), as identified by MS/MS (Tables 1-4 in Supporting Information). The preferential binding of SVMPs to the column was expected, given that DM43 is a well-characterized inhibitor of snake venom metalloproteinases.14 SVMPs are key toxins in the pathogenesis induced by Viperidae venoms and represent one of their most abundant components, which ultimately lead to microvessel disruption and hemostatic disturbances.31 Transcriptomic studies have shown that SVMPs are by far the most expressed molecules in B. insularis32 and B. jararaca22 venoms and correspond to 41.7% and 53.1% of all toxin transcripts, respectively. These results do not seem to correlate well with our proteomic data. Transcriptomic and proteomic data sets from Bitis gabonica gabonica snake venom have been compared and poor correlation was also observed.33 The modest correlation between mRNA expression and actual protein levels is not surprising, and has been mainly attributed to substantial variations in post-translational processing and/or several analytical limitations that can often bias quantitative results.34 When working with snake venoms, one must additionally consider the possibility of composition variation due to geographical distribution, seasonal changes, diet, habitat, age, and sexual dimorphism.35 Systematic efforts to catalog full-length transcripts and secreted proteins are needed in order to ascertain the relationship between the gene products of snake venom glands. It is well-known that SVMPs are zinc-dependent enzymes synthesized as zymogens in the venom glands. They are Journal of Proteome Research • Vol. 8, No. 5, 2009 2353

research articles

Rocha et al.

Figure 1. Proteomic analysis of B. atrox venom. (A) 2-DE of crude venom on 18 cm IPG strips pH 4-7, 15% SDS-PAGE, under reducing conditions. (B) Affinity chromatography profile of crude venom submitted to a Hitrap NHS-activated affinity column coupled with DM43. Inset: 12% SDS-PAGE under reducing conditions of molecular mass markers (MM), crude venom (cv), nonbound (1), and bound (2) fractions from the affinity chromatography. (C) 2-DE of the venom fraction not bound to the DM43 affinity column. (D) 2-DE of the venom fraction bound to the DM43 affinity column. All gels were stained with Coomassie blue. Circled spots were excised, trypsinized, and analyzed by MS/MS. Protein identification details are shown in Table 1 of Supporting Information. Red, metalloproteinase; blue, serine proteinase; black, PLA2; green, C-type lectin; purple, L-amino acid oxidase.

proteolytically processed to generate mature SVMPs, which present different domain compositions.36 When only one domain is present (the catalytic one), they are classified as PIa SVMPs (20-30 kDa). PII SVMPs (35-50 kDa) present a canonical disintegrin domain located C-terminal to the catalytic domain. In addition to the catalytic domain, PIII SVMPs (50-80 kDa) show disintegrin-like and cysteine-rich domains. Some PIII SVMPs can generate multimeric structures as they form disulfide bonds with two lectin-like domains. This class of enzyme, formerly classified as PIV SVMP, now constitutes the PIIId subclass. Other SVMP subclasses (PII a-e and PIII a-d) have been categorized based on two main criteria, susceptibility to proteolytic processing and to dimerization. With the use of the affinity of DM43 to capture SVMPs from crotalid venoms, we have shown that B. atrox, B. insularis, and B. jararaca venoms show quite comparable metalloproteinase subproteome profiles. On the other hand, the venoms of B. jararaca and C. atrox, which were expected to be similar considering they induce similar associated pathologies,7 showed strikingly different metalloproteinase patterns. In the size region of the gel where the PIII class members are expected, two main lines of spots of 60 and 45 kDa, with pIs ranging from 4.5 to 2354

Journal of Proteome Research • Vol. 8, No. 5, 2009

5.5, were observed in bothropic venoms (Figures 1-3D). In C. atrox venom (Figure 4D), more diffused spots were detected mainly in the 45 kDa region. Some less abundant spots of ∼23 kDa and a more basic pI (∼6.2) were detected only in the bothropic venoms. They were identified as metalloproteinases and the sequencing of several peptides belonging to the catalytic domain suggests they correspond to class PI SVMPs. An apparent molecular mass of 60 kDa is characteristic of full-length PIIIs and was further confirmed by sequencing peptides from more than one domain (catalytic, disintegrinlike, and/or cysteine-rich) in several of these spots. In the case of the B. insularis bound fraction, Edman degradation sequencing analysis of the 60 kDa band from SDS-PAGE (Figure 2B, Inset, lane 2) was unsuccessful, probably due to N-terminal blockage by a pyroglutamyl residue, a known characteristic of several snake venom metalloproteinase.37 However, sequencing of the 45 kDa band from this same venom has given rise to the following two parallel N-terminal sequences, KKHDNAQL and LTAIDFNGPT, which correspond to positions 93-100 and 101-110 of the catalytic domain of mature PIII from B. insularis (accession number Q8QG88). The 45 kDa band is barely visible on the SDS-PAGE profile of crude venoms (Figures 1-4B,

Crotalid Venom Subproteomes Unraveled by DM43

research articles

Figure 2. Proteomic analysis of B. insularis venom. (A) 2-DE of crude venom on 18 cm IPG strips pH 4-7, 15% SDS-PAGE, under reducing conditions. (B) Affinity chromatography profile of crude venom submitted to a Hitrap NHS-activated affinity column coupled with DM43. Inset: 12% SDS-PAGE under reducing conditions of molecular mass markers (MM), crude venom (cv), nonbound (1), and bound (2) fractions from the affinity chromatography. (C) 2-DE of the venom fraction not bound to the DM43 affinity column. (D) 2-DE of the venom fraction bound to the DM43 affinity column. All gels were stained with Coomassie blue. Circled spots were excised, trypsinized, and analyzed by MS/MS. Protein identification details are shown in Table 2 of Supporting Information. Red, metalloproteinase; blue, serine proteinase; black, PLA2; green, C-type lectin; purple, L-amino acid oxidase; yellow, CRISP; light blue, NGF.

Insets, lane cv), and was probably generated during sample manipulation, either by autoproteolysis and/or acid hydrolysis due to the elution conditions of the affinity column. Working with the PIII SVMP brevilysin H6 from Gloydius halys brevicaudus, Fujimura and colleagues38 first described the generation of a 45 kDa autocatalytic fragment in 50 mM Tris-HCl, pH 8.5. The susceptible cleavage site was located between Leu98 and Leu99, but no hydrolysis was observed using the same buffer in the presence of 10 mM Ca2+ ions. On the basis of the homology assumption with adamalysin II and H2proteinase, the authors have speculated that calcium would bind to a long loop near the 98-99 scissile bond, stabilizing the enzyme’s tertiary structure. In another paper on SVMP autolysis, Moura-da-Silva and colleagues39 suggested the existence of alternative forms of proteolytically susceptible PIII jararhagin contributes to B. jararaca venom complexity. A limited amount of a 43 kDa fragment was observed by SDSPAGE under reducing conditions only after incubating jararhagin for 4 days at pH 3.0. Since most SVMPs have little or no proteolytic activity at this pH, its generation was attributed to acid hydrolysis. We have shown here that, when working under very acidic conditions (pH 2.7), even the presence of 20 mM calcium ions was not able to suppress the rapid generation of this preferential 45 kDa fragment.

Several spots identified as SVMPs in the venom bound fractions showed molecular masses compatible with neither PI, nor with known PIII SVMPs (Figures 1-4D).40 All SVMP fragments were aligned with the archetypical full-length PIII metalloproteinase jararhagin using the Multalin program. In Tables 1-4 (Supporting Information, column “Protein family”), the SVMP domain where each peptide was aligned is indicated. In several very abundant spots of about 27 kDa and pI 4.9, only peptides from disintegrin-like and/or cysteine-rich domains (DC domains) were identified. Most of these representatives of DC domains were clearly generated proteolytically from PIII SVMPs during sample manipulation, as already described for the 45 kDa fragments, since they do not appear at such an intensity on the crude venoms gels (Figures 1-4A). Interestingly, we have previously shown that DM43 does not bind to jararhagin-C,41 the DC domain split from the class PIII SVMP jararhagin, indicating the essential role of the metalloproteinase domain for the interaction.14 If this is also true for other DC domains, then the proteolytic processing of class PIII enzymes most likely occurred after their binding to immobilized DM43 through their metalloproteinase domains. The parts of the metalloproteinase domains that were probably split from these PIII SVMPs were observed in the very low molecular mass region of the gel (67 kDa) have been observed in all crude venoms (Figures 1-4A) and mainly in the DM43 nonbound fractions (Figures 1-4C). They were more abundant in B. insularis and B. atrox venoms and virtually absent in B. jararaca venom. MS/MS analysis of the former sample identified several peptides matching disintegrin and/or metalloproteinase domains, pointing to the existence of metalloproteinases with higher molecular masses than usually observed. Interestingly, transcriptomic studies have shown the predominant expression of class PIII SVMPs in B. insularis venom when compared to B. jararaca venom.22,32 SVMPs with rather atypical high molecular masses have been identified by LC-MS/MS in SDS-PAGE slices from C. atrox venom run under reducing conditions. Their presence has been most likely attributed to a smearing of the proteinase through the gel lane.9 It appears that this artifact did not occur in our 2-DE analysis where very well focused spots have been observed. The rather low number of protein families represented in snake venoms has been reported by several authors working with proteomic strategies applied to Viperidae venoms. In addition to the higher abundance protein families identified herein, the following classes have been described in the proteomic analysis of some viperid venoms: glutaminyl cyclase, Kunitz- and cystatin-type protease inhibitors, vascular apoptosis-inducing protein, vascular endothelial growth factor (VEGF), disintegrin, bradykinin-potentiating peptide, and natriuretic peptide.8 Several protein spots, even though they presented good quality mass spectra, could not be successfully identified. They are being carefully sequenced manually and may constitute new proteins revealing some yet unknown structural features. This result is corroborated by several transcriptomic studies, which have found that

Crotalid Venom Subproteomes Unraveled by DM43 10-47% of snake venom clones cannot be assigned to any database entry.22

Concluding Remarks For the first time, a large amount of consistent MS/MS identification data was produced for B. atrox, B. jararaca, B. insularis, and C. atrox snake venoms analyzed by 2-DE, after affinity chromatography with DM43. Together with estimates of pI, molecular mass and relative abundance, and visualization of post-translational modifications, the MS/MS results allowed for a more complete description of their subproteomes. We have also shown that DM43 interacts with metalloproteinases from several crotalid venoms, including two of the most medically important in Brazil. This is a rather relevant result considering that it has been recently suggested that antivenom therapy should definitely include rapid in situ administration of natural inhibitors of venom toxins after the snakebite.51 Such natural inhibitors most likely target SVMP catalytic centers, which are highly conserved among different taxa, despite their antigenic diversity.

Acknowledgment. This study was supported by grants from Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico (CNPq), Rede Proteoˆmica do Rio de Janeiro/ Fundac¸a˜o de Amparo a` Pesquisa do Estado do Rio de Janeiro (FAPERJ), and Fundac¸a˜o Oswaldo Cruz/PDTIS, Brazil. We are grateful to Rodolpho M. Albano, from Departamento de Bioquı´mica, Universidade do Estado do Rio de Janeiro, for B. jararaca database access. The authors thank Rodrigo da Cunha Mexas and Bruno Eschenazi, from Servic¸o de Produc¸a˜o e Tratamento de Imagens-IOC, for preparing high resolution figures for publication, and for creating the cartoon, respectively. Supporting Information Available: Figure 1, 2-DE of venom fractions bound to the DM43 affinity column on 18 cm IPG strips pH 4-7, 15% SDS-PAGE, under nonreducing conditions. All gels were stained with Coomassie blue. Venoms are from (A) B. atrox, (B) B. insularis, (C) B. jararaca and (D) C. atrox. Tables 1-4, protein identification details for B. atrox, B. insularis, B. jararaca, and C. atrox. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Gutie´rrez, J. M.; Theakston, R. D.; Warrell, D. A. Confronting the neglected problem of snake bite envenoming: the need for a global partnership. PLoS Med. 2006, 3, e150. (2) Chippaux, J. P. Snake-bites: appraisal of the global situation. Bull. WHO 1998, 76, 515–524. (3) Gutie´rrez, J. M.; Gondo Higashi, H.; Hui Wen, F.; Burnouf, T. Strengthening antivenom production in Central and South American public laboratories: Report of a workshop. Toxicon 2007, 49, 30–35. (4) Koh, D. C.; Armugam, A.; Jeyaseelan, K. Snake venom components and their applications in biomedicine. Cell. Mol. Life Sci. 2006, 63, 3030–3041. (5) Fox, J. W.; Serrano, S. M. Approaching the golden age of natural product pharmaceuticals from venom libraries: an overview of toxins and toxin-derivatives currently involved in therapeutic or diagnostic applications. Curr. Pharm. Des. 2007, 13, 2927–2934. (6) Tu, A. T. Overview of snake venom chemistry. In Natural Toxins II, Singh, B. R. ; Tu, A. T. , Eds.; Plenum Press: New York, 1996. (7) Fox, J. W.; Shannon, J. D.; Stefansson, B.; Kamiguti, A. S.; Theakston, D. G.; Serrano, S. M.; Camargo, A. C. M.; Sherman, N. Role of discovery science in Toxinology: examples in venom proteomics. In Perspectives in Molecular Toxinology; Me´nez, A. , Ed.; John Wiley & Sons, Ltd.: Chichester, U.K., 2002; pp 97-105. (8) Fox, J. W.; Serrano, S. M. Exploring snake venom proteomes: multifaceted analyses for complex toxin mixtures. Proteomics 2008, 8, 909–920.

research articles (9) Fox, J. W.; Ma, L.; Nelson, K.; Sherman, N. E.; Serrano, S. M. Comparison of indirect and direct approaches using ion-trap and Fourier transform ion cyclotron resonance mass spectrometry for exploring viperid venom proteomes. Toxicon 2006, 47, 700–714. (10) Nawarak, J.; Phutrakul, S.; Chen, S. T. Analysis of lectin-bound glycoproteins in snake venom from the Elapidae and Viperidae families. J. Proteome Res. 2004, 3, 383–392. (11) Wermelinger, L. S.; Dutra, D. L.; Oliveira-Carvalho, A. L.; Soares, M. R.; Bloch, C., Jr.; Zingali, R. B. Fast analysis of low molecular mass compounds present in snake venom: identification of ten new pyroglutamate-containing peptides. Rapid Commun. Mass Spectrom. 2005, 19, 1703–1708. (12) Serrano, S. M.; Shannon, J. D.; Wang, D.; Camargo, A. C.; Fox, J. W. A multifaceted analysis of viperid snake venoms by two-dimensional gel electrophoresis: an approach to understanding venom proteomics. Proteomics 2005, 5, 501–510. (13) Birrell, G. W.; Earl, S. T.; Wallis, T. P.; Masci, P. P.; de Jersey, J.; Gorman, J. J.; Lavin, M. F. The diversity of bioactive proteins in Australian snake venoms. Mol. Cell. Proteomics 2007, 6, 973–986. (14) Neves-Ferreira, A. G. C.; Perales, J.; Fox, J. W.; Shannon, J. D.; Makino, D. L.; Garratt, R. C.; Domont, G. B. Structural and functional analyses of DM43, a snake venom metalloproteinase inhibitor from Didelphis marsupialis serum. J. Biol. Chem. 2002, 277, 13129–13137. (15) Cerqueira, R.; Lemos, B. Morphometric differentiation between Neotropical black-eared opossums Didelphis marsupialis and D. aurita (Didelphimorphia, Didelphidae). Mammalia 2000, 64, 319– 327. (16) Neves-Ferreira, A. G. C.; Cardinale, N.; Rocha, S. L. G.; Perales, J.; Domont, G. B. Isolation and characterization of DM40 and DM43, two snake venom metalloproteinase inhibitors from Didelphis marsupialis serum. Biochim. Biophys. Acta 2000, 1474, 309–320. (17) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem. 1951, 193, 265–275. (18) Stone, K. L.; Williams, K. R. Enzymatic digestion of proteins and HPLC peptide isolation. In A Practical Guide to Protein and Peptide Purification for Microsequencing; Matsudaira, P. , Ed.; Academic Press: San Diego, CA, 1993; pp 43-69. (19) Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 1970, 227, 680– 685. (20) Go¨rg, A.; Weiss, W.; Dunn, M. J. Current two-dimensional electrophoresis technology for proteomics. Proteomics 2004, 4, 3665– 3685. (21) Altschul, S. F.; Madden, T. L.; Schaffer, A. A.; Zhang, J.; Zhang, Z.; Miller, W.; Lipman, D. J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997, 25, 3389–3402. (22) Cidade, D. A.; Sima˜o, T. A.; Davila, A. M.; Wagner, G.; L., J.-d.-A. I.; Ho, P. L.; Bon, C.; Zingali, R. B.; Albano, R. M. Bothrops jararaca venom gland transcriptome: analysis of the gene expression pattern. Toxicon 2006, 48, 437–461. (23) Medzihradszky, K. F. Peptide sequence analysis. Methods Enzymol. 2005, 402, 209–244. (24) Corpet, F. Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res. 1988, 16, 10881–10890. (25) Arau ´ jo, F. A. A. ; Santalu ´ cia, M.; Cabral, R. F. Epidemiologia dos acidentes por animais pec¸onhentos. In Animais pec¸onhentos no Brasil: biologia, clı´nica e terapeˆutica dos acidentes; Costa-Cardoso, J. L., Ed.; Sarvier: Sa˜o Paulo, Brazil, 2003; pp 6-12. (26) Campbell, J. A.; Lamar, W. W. The Venomous Reptiles of Latin America; Cornell University Press: New York, 1989. (27) Sanchez, E. F.; Freitas, T. V.; Ferreira-Alves, D. L.; Velarde, D. T.; Diniz, M. R.; Cordeiro, M. N.; Agostini-Cotta, G.; Diniz, C. R. Biological activities of venoms from South American snakes. Toxicon 1992, 30, 95–103. (28) Fox, J. W.; Bjarnason, J. B. Atrolysins: metalloproteinases from Crotalus atrox venom. Methods Enzymol. 1995, 248, 368–387. (29) Valente, R. H.; Guimara˜es, P. R.; Junqueira, M.; Neves-Ferreira, A. G. C.; Soares, M. R.; Chapeaurouge, A.; Trugilho, M. R.; Leo´n, I. R.; Rocha, S. L.; Oliveira-Carvalho, A. L.; Wermelinger, L. S.; Dutra, D. L.; Lea˜o, L. I.; Junqueira-de-Azevedo, I. L.; Ho, P. L.; Zingali, R. B.; Perales, J.; Domont, G. B. Bothrops insularis venomics: A proteomic analysis supported by transcriptomicgenerated sequence data. J. Proteomics 2009, 72, 241–255. (30) Guercio, R. A.; Shevchenko, A.; Lopez-Lozano, J. L.; Paba, J.; Sousa, M. V.; Ricart, C. A. Ontogenetic variations in the venom proteome of the Amazonian snake Bothrops atrox. Proteome Sci. 2006, 4, 11. (31) Gutie´rrez, J. M.; Rucavado, A.; Escalante, T.; Diaz, C. Hemorrhage induced by snake venom metalloproteinases: biochemical and

Journal of Proteome Research • Vol. 8, No. 5, 2009 2359

research articles (32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

2360

biophysical mechanisms involved in microvessel damage. Toxicon 2005, 45, 997–1011. Junqueira-de-Azevedo, I. L.; Ho, P. L. A survey of gene expression and diversity in the venom glands of the pitviper snake Bothrops insularis through the generation of expressed sequence tags (ESTs). Gene 2002, 299, 279–291. Calvete, J. J.; Marcinkiewicz, C.; Sanz, L. Snake venomics of Bitis gabonica gabonica. Protein family composition, subunit organization of venom toxins, and characterization of dimeric disintegrins bitisgabonin-1 and bitisgabonin-2. J. Proteome Res. 2007, 6, 326– 236. Cox, B.; Kislinger, T.; Emili, A. Integrating gene and protein expression data: pattern analysis and profile mining. Methods 2005, 35, 303–314. Chippaux, J. P.; Williams, V.; White, J. Snake venom variability: methods of study, results and interpretation. Toxicon 1991, 29, 1279–1303. Fox, J. W.; Serrano, S. M. Insights into and speculations about snake venom metalloproteinase (SVMP) synthesis, folding and disulfide bond formation and their contribution to venom complexity. FEBS J. 2008, 275, 3016–3030. Gomis-Ru ¨ th, F. X.; Kress, L. F.; Kellermann, J.; Mayr, I.; Lee, X.; Huber, R.; Bode, W. Refined 2.0 A X-ray crystal structure of the snake venom zinc-endopeptidase adamalysin II. Primary and tertiary structure determination, refinement, molecular structure and comparison with astacin, collagenase and thermolysin. J. Mol. Biol. 1994, 239, 513–544. Fujimura, S.; Oshikawa, K.; Terada, S.; Kimoto, E. Primary structure and autoproteolysis of brevilysin H6 from the venom of Gloydius halys brevicaudus. J. Biochem. (Tokyo) 2000, 128, 167–173. Moura-da-Silva, A. M.; Della-Casa, M. S.; David, A. S.; Assakura, M. T.; Butera, D.; Lebrun, I.; Shannon, J. D.; Serrano, S. M.; Fox, J. W. Evidence for heterogeneous forms of the snake venom metalloproteinase jararhagin: a factor contributing to snake venom variability. Arch. Biochem. Biophys. 2003, 409, 395–401. Fox, J. W.; Serrano, S. M. Structural considerations of the snake venom metalloproteinases, key members of the M12 reprolysin family of metalloproteinases. Toxicon 2005, 45, 969–985. Usami, Y.; Fujimura, Y.; Miura, S.; Shima, H.; Yoshida, E.; Yoshioka, A.; Hirano, K.; Suzuki, M.; Titani, K. A 28 kDa-protein with disintegrin-like structure (jararhagin-C) purified from Bothrops

Journal of Proteome Research • Vol. 8, No. 5, 2009

Rocha et al.

(42)

(43)

(44)

(45) (46)

(47) (48)

(49)

(50) (51)

jararaca venom inhibits collagen- and ADP-induced platelet aggregation. Biochem. Biophys. Res. Commun. 1994, 201, 331–339. Nakashima, K.; Ogawa, T.; Oda, N.; Hattori, M.; Sakaki, Y.; Kihara, H.; Ohno, M. Accelerated evolution of Trimeresurus flavoviridis venom gland phospholipase A2 isozymes. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5964–5968. Moura-da-Silva, A. M.; Theakston, R. D.; Crampton, J. M. Evolution of disintegrin cysteine-rich and mammalian matrix-degrading metalloproteinases: gene duplication and divergence of a common ancestor rather than convergent evolution. J. Mol. Evol. 1996, 43, 263–269. Deshimaru, M.; Ogawa, T.; Nakashima, K.; Nobuhisa, I.; Chijiwa, T.; Shimohigashi, Y.; Fukumaki, Y.; Niwa, M.; Yamashina, I.; Hattori, S.; Ohno, M. Accelerated evolution of crotalinae snake venom gland serine proteases. FEBS Lett. 1996, 397, 83–88. Ogawa, T.; Chijiwa, T.; Oda-Ueda, N.; Ohno, M. Molecular diversity and accelerated evolution of C-type lectin-like proteins from snake venom. Toxicon 2005, 45, 1–14. Rocha, S. L.; Lomonte, B.; Neves-Ferreira, A. G. C.; Trugilho, M. R. O.; Junqueira-de-Azevedo, I. L.; Ho, P. L.; Domont, G. B.; Gutie´rrez, J. M.; Perales, J. Functional analysis of DM64, an antimyotoxic protein with immunoglobulin-like structure from Didelphis marsupialis serum. Eur. J. Biochem. 2002, 269, 6052– 6062. Wijeyewickrema, L. C.; Berndt, M. C.; Andrews, R. K. Snake venom probes of platelet adhesion receptors and their ligands. Toxicon 2005, 45, 1051–1061. Neves-Ferreira, A. G. C.; Perales, J.; Ovadia, M.; Moussatche´, H.; Domont, G. B. Inhibitory properties of the antibothropic complex from the South American opossum (Didelphis marsupialis) serum. Toxicon 1997, 35, 849–863. Hirabayashi, J.; Kusunoki, T.; Kasai, K. Complete primary structure of a galactose-specific lectin from the venom of the rattlesnake Crotalus atrox. Homologies with Ca2(+)-dependent-type lectins. J. Biol. Chem. 1991, 266, 2320–2326. Yamazaki, Y.; Morita, T. Structure and function of snake venom cysteine-rich secretory proteins. Toxicon 2004, 44, 227–231. Gutie´rrez, J. M.; Lomonte, B.; Leo´n, G.; Rucavado, A.; Chaves, F.; Angulo, Y. Trends in snakebite envenomation therapy: scientific, technological and public health considerations. Curr. Pharm. Des. 2007, 13, 2935–2950.

PR800977S