| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular & Cellular Proteomics 6:973-986, 2007.
© 2007 by The American Society for Biochemistry and Molecular Biology, Inc.
,
,¶,,¶,||,**,||,**,,¶¶
From The Queensland Institute of Medical Research, P. O. Royal Brisbane Hospital, Brisbane 4029, Australia and the Central Clinical Division and Southern Clinical Division, Faculty of Health Sciences, || Institute for Molecular Biosciences, and School of Molecular and Microbial Sciences, University of Queensland, Brisbane 4072, Australia
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-neurotoxins, both represented by different sequence variants. The higher molecular weight range contained proteases, nucleotidases, oxidases, and homologs of mammalian coagulation factors. This information together with the identification of several novel proteins (metalloproteinases, vespryns, phospholipase A2 inhibitors, protein-disulfide isomerase, 5'-nucleotidases, cysteine-rich secreted proteins, C-type lectins, and acetylcholinesterases) aids in understanding the lethal mechanisms of elapid snake venoms and represents a valuable resource for future development of novel human therapeutics.
Although there is much known about the protein compositions of venoms from Asian and American snakes, comparatively little is known of Australian snakes. This is despite the top 10 most toxic snakes (determined by LD50 (lethal dose in mice to kill 50%)) being Australian elapid snakes (8). Australian snakes are most closely related to Asian snakes, and although they have significantly higher toxicity, they cause far fewer deaths by envenomation than their Asian counterparts, most likely due to lower population density and more widespread availability of suitable health care in Australia (8). Proteins and peptides comprise the majority of the dry weight of elapid snake venoms, whereas other components include metallic cations, carbohydrates, nucleosides, biogenic amines, and low levels of free amino acids and lipids (9, 10). Venoms from Australian elapids can be loosely divided into procoagulant and anticoagulant types. The procoagulant venoms contain serine proteases (prothrombin activators) that cleave prothrombin to produce thrombin in the coagulation cascade resulting in coagulation. Snake venom prothrombin activators are classified into four groups based on functional characteristics, structural properties, and cofactor requirements (11). Group A and B prothrombin activators are metalloproteinases, whereas Group C and D prothrombin activators are serine proteinases. Some Australian snake venoms are known to contain Group C or Group D prothrombin activators (12). Group C prothrombin activators resemble the mammalian Factor Xa (FXa)1-Factor Va (FVa) complex, whereas Group D activators are structurally and functionally similar to FXa alone (12, 13). The Group C prothrombin activator from the Australian brown snake, P. textilis, has been extensively studied and named Pseutarin C (14). The 1461-amino acid FVa-like non-enzymatic subunit of Pseutarin C has been shown to share 50% sequence identity and significant structural homology with human FVa (15). Another hemostasis-related family of proteins identified in Australian snake venom is the textilinin family of plasmin inhibitors (16). These 7-kDa proteins share 
45% identity with aprotinin, a bovine Kunitz-type serine protease inhibitor that acts on plasmin and kallikrein to reduce blood loss during surgery. Six isoforms of textilinin have been identified in P. textilis venom gland-derived cDNA (7).
Australian snake venoms are also known to contain PLA2s and peptidic neurotoxins (for a review, see Ref. 17). Australian elapid PLA2s have seven conserved disulfide bonds and molecular masses of 13–15 kDa. As well as phospholipase activity, individual PLA2s are known to have myotoxic, neurotoxic, and/or anticoagulant activities. The -neurotoxins found in Australian elapids are postsynaptic blocking short or long chain neurotoxins (18, 19). Short and long chain neurotoxins have similar effects and bind with high affinity to skeletal nicotinic acetylcholine receptors. The short chain neurotoxins have four disulfide bonds and are 60 amino acids in length, whereas the long chain neurotoxins have five disulfide bonds and are 73 amino acids in length. More recently, neurotoxic cysteine-rich secreted proteins (CRISPs) have been identified and characterized from the Australian elapids Pseudechis australis and Pseudechis porphyriacus (20, 21). These 211-amino acid proteins show the ability to block cyclic nucleotide-gated ion channels involved in signaling from the visual and olfactory systems.
Other proteins known to exist in Australian elapid venoms include L-amino-acid oxidase, nerve growth factor (NGF), and natriuretic peptides. L-Amino-acid oxidase was identified in P. australis venom and in purified form was found to be antibacterial when tested against the important pathogen Aeromonas hydrophila (22). This protein has also been identified in Viperidae snake venoms. NGF has been identified in multiple Asian and African snake venoms but has been reported only recently as a component of Australian elapid venoms (23). NGF from Oxyuranus scutellatus venom is gly co sylated and shows the ability to induce neurite outgrowth of PC12 cells, a standard assay for NGF activity (23). Natriuretic peptides from Oxyuranus microlepidotus venom show potent arterial smooth muscle relaxant effects (24). Natriuretic peptides have been identified and characterized from several Australian elapid venoms and represent the smallest known proteins from these venoms at only 35–39 amino acids (25).
The present study describes a thorough screening and identification of the venom proteins present in 18 Australian elapid snake species representing nine genera of the most venomous snakes on earth. In addition, post-translational modifications such as gly co sy la tion, phos pho ryl a tion, and 
-carboxylation were examined in the venom proteins using specific stains and antibodies. Glycoforms of multiple proteins were identified and characterized using lectin binding specificity. Also antivenom reactivity of venom proteins was examined using tiger snake (Notechis) antivenom raised in horses for clinical use in envenomated people. Many proteins previously unknown in Australian snake venoms were identified based on sequence matches to venom proteins from other snakes. This information sheds light on evolutionary relationships between the different snake species and the clinical manifestations of envenomation. This comprehensive proteomics analysis brings Australian elapid snakes in line with their well studied American and Asian counterparts and represents a valuable resource for the future development of novel human therapeutics.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-carboxyglutamate (Gla) residues was from American Diagnostica Inc. (Greenwich, CT). Tiger snake (Notechis scutatus) antivenom was from Commonwealth Serum Laboratories Ltd. (Victoria, Australia). All other chemicals were from Sigma unless stated otherwise.
|
Identification of Venom Components and de Novo Sequencing Using MALDI-TOF Mass Spectrometry—
Protein spots from silver-stained 2D PAGE were excised, washed in water, and destained as described previously (28). Trypsin was added, and proteins were allowed to digest overnight at 37 °C prior to extraction. Extracted peptides were dried; resuspended in 50% ACN, 0.1% TFA; mixed 1:1 with matrix (10 mg/ml -cyano-4-hydroxycinamic acid in 60% acetonitrile, 25 mM ammonium bicarbonate); and spotted on a MALDI plate. Peptides were analyzed by MALDI-TOF/TOF MS/MS using a 4700 Proteomics Analyzer (Applied Biosystems) operated in positive ion reflectron mode. MS data were acquired using 2000 shots of a neodymium-doped yttrium aluminium garnet (Nd:YAG) laser at 355 nm with a 200-Hz repetition rate and fixed intensity. MS data were calibrated via a plate wide external calibration using the 4700 Mass Standards kit (Applied Biosystems) containing des-Arg-bradykinin (MH+ 904.458), angiotensin I (MH+ 1296.685), Glu-fibrinopeptide B (MH+ 1570.677), ACTH (1–17 clip, MH+ 2093.087), ACTH (18–39 clip, MH+ 2465.199), and ACTH (7–38 clip, MH+ 3657.929). The top 50 most intense peptides detected for each spot in the MS mode were automatically selected for MS/MS analysis using 3000 laser shots at a fixed intensity 20% greater than that used for MS. MS/MS data were calibrated against the MS/MS fragments of the m/z = 1296.685 angiotensin I peptide in the standards. With regard to the MS/MS analysis, no CID was used. Metastable decay was used exclusively to obtain fragment ion information without the introduction of CID gas into source 2. Interference from metastable decay of the precursor ions postsource 2 was precluded using metastable ion suppression
Alternatively peptides were analyzed using a Microflex MALDI-TOF PSD instrument (Bruker Daltonics, Bremen, Germany) operated in positive ion reflectron mode. MS data were acquired using 350 shots of a nitrogen laser at 355 nm with a 20-Hz repetition rate and varying intensity. MS data were calibrated via close external calibration using peptide standards (New England Biolabs) containing angiotensin I (MH+ 1296.69), neurotensin (MH+ 1672.92), ACTH (1–17 clip, MH+ 2093.09), ACTH (18–39 clip, MH+ 2465.20), and ACTH (7–38 clip, MH+ 3657.93). The top three most intense peptides detected for each spot in the MS mode were automatically selected for PSD analysis using 15 segments with 200 shots per segment. Initial laser power was 25% greater than for MS. The segments were stitched together by Flex Analysis software (Version 2.4, Bruker Daltonics) for sequence interpretation. PSD data were internally calibrated.
MALDI-TOF/TOF MS/MS data from the 4700 Proteomics Analyzer were automatically analyzed using the GPS Explorer suite of software (Version 3.5 Build 321, Applied Biosystems). For each spot a combined MS and MS/MS analysis was performed in-house using a Mascot search engine (Version 1.9) and the Celera Discovery System database (CDS Combined KBMS2.1.20030813 containing 1,335,729 sequences, dated May 5, 2006). MS peptide tolerance was 100 ppm, and MS/MS tolerance was 0.3 Da. MALDI-TOF PSD data from the Microflex were automatically analyzed using Biotools 3.0 software (Bruker Daltonics). For each spot MS and PSD data were separately searched in-house using the Mascot search engine (Version 1.9) and the Chordata taxonomic subset of the National Center for Biotechnology Information non-redundant database (NCBInr 20060111; Chordata subset; 627,293 sequences; dated January 11, 2006). In both cases, mass tolerance was set at 150 ppm. All searches took into account carbamidomethylated cysteine and oxidized methionine. For the purposes of protein identification, no other post-translational modifications were considered, and the top matched protein from the database searches was included in Supplemental Table 1. The scores used to evaluate the quality of matches for MS/MS data appear in Supplemental Table 1 in columns titled "Ion Score" and "Ion Score confidence interval" calculated by the Mascot search engine. Protein and ion scores increase where the experimental fragmentation data matches theoretical fragmentation data. With regard to acceptance criteria for protein identification, those candidates with Mascot scores greater than the 95% confidence threshold (generally a protein score of 65) and whose protein mass and pI correlated with the 2D PAGE spot were automatically accepted. Those candidates with Mascot scores below the 95% confidence threshold but whose identity matched to a known snake venom protein were also included. Candidates still unknown were selected for de novo peptide sequencing.
For de novo peptide sequencing, MALDI-TOF/TOF MS/MS data were opened in Data Explorer (Applied Biosystems, Version 4.2) and deisotoped, and raw text peak lists were exported. The peak lists were analyzed using the automatic de novo function of PEAKS Studio software Version 2.4 (29) (Bioinformatics Solutions Inc., Ontario, Canada). Peptide sequences of 6 or more amino acids with a 100% confidence call using parent ion and fragment mass error tolerances of 0.1 Da were collected and matched to the NCBI non-redundant protein database using the protein Basic Local Alignment Search Tool (BLAST) algorithm (Version 2.2.14, www.ncbi.nlm.nih.gov/BLAST/).
Enzymatic Deglycosylation—
Venom samples (50 µg) were precipitated in 8:1 acetone:methanol for 16 h at –20 °C. Samples were centrifuged for 30 min at 16000 x g and 4 °C to pellet protein. The supernatants were removed and discarded, and the protein pellets were allowed to air dry for 10 min under ambient conditions. PNGase F (New England Biolabs) was subsequently used according to the manufacturer's instructions. Reactions were allowed to proceed for 16 h at 37 °C.
Lectin Binding Specificity—
Venom samples (100 µg) were separated by 1D PAGE as described above and transferred onto nitrocellulose membrane in Towbin buffer at 100 V for 1 h. Glycosylation profiles of the venom proteins were examined using a panel of eight lectins labeled with FITC (Fluorescein Lectin Kit 1, Vector Laboratories) as described previously (30).
Phosphoprotein and Glycoprotein Staining—
Detection of phosphoproteins and glycoproteins in 1D PAGE was performed using the ProQ Diamond and ProQ Emerald fluorescent reagents, respectively (Molecular Probes, Invitrogen) according to the manufacturer's instructions. ProQ Emerald is a fluorescent form of the periodic acid-Schiff stain for carbohydrates that has a broad range of specificity for glycol containing molecules.
Immunoblotting and Antivenom Reactivity—
Immunoblotting for Gla residues and the heavy chain of the FXa-like protease was performed as described previously (31). Tiger snake antivenom reactivity was performed similarly to immunoblotting as follows. Proteins were separated by 1D SDS-PAGE, transferred to nitrocellulose membrane, blocked in 5% skim milk, and incubated overnight in a 1:2000 dilution of antivenom in PBS with 0.05% Tween 20. After extensive washing, the membrane was incubated in a 1:2000 dilution of goat anti-horse Ig conjugated to horseradish peroxidase, and signal was detected using ECL reagent.
|
RESULTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
PLA2s and PLA2 Inhibitors—
The most frequently encountered proteins identified across the species were PLA2s that were found in all 18 venoms. In all cases there were multiple PLA2 isoforms representing sequence variants. Because of the large number of isoforms and associated activities for these proteins, information on the specific PLA2s identified within the different species can be accessed through the NCBI accession numbers provided in Supplemental Table 1. In the majority of cases the PLA2 isoforms were in the molecular size range of 13–17 kDa as predicted from published data and in some cases their cDNA sequences (34). However, in other cases (for example Acanthophis, Fig. 2F) there was considerable size variation with PLA2s ranging from 13 to 30 kDa. Surprisingly we also detected a protein homologous to the isoform of PLA2 inhibitor (35) (NCBI accession number CAB56615) in both Pseudonaja nuchalis and O. microlepidotus. This is the first description of a PLA2 inhibitor in the venom of any snake. Previous reports have demonstrated that PLA2 inhibitors are present in the blood of Elapidae, Hydrophidae, Boidae, and Colubridae genera, and it was assumed that they were present to provide protection against any PLA2s appearing inadvertently in the blood (for a review, see Ref. 36).
Neurotoxins and Protease Inhibitors—
Australian snake venoms are known to contain a multitude of neurotoxins including short and long chain -neurotoxins and neurotoxic PLA2s (8). Long chain neurotoxins were found in the majority of venoms. These were readily identifiable in all species of Notechis and Pseudonaja but only in one (P. australis) of the Pseudechis species. On the other hand, short chain neurotoxins were only detected in three species (Table II). The Kunitz-type serine protease inhibitor textilinin was identified in all species of the Pseudonaja genus as expected. We have previously identified multiple forms of textilinin both at the protein and cDNA level in P. textilis (7, 16). Proteins with homology to textilinin were also found in the Notechis, Acanthophis, Pseudechis, Tropidechis, and Oxyuranus genera. This molecule is an antifibrinolytic agent that inhibits plasmin (7).
Prothrombin Activators—
Another snake protein known to affect hemostasis is the FXa-like prothrombin activator previously identified in Pseudonaja and other genera (37). The heavy chain of this protein was identified in all the Notechis, Oxyuranus, and Pseudonaja species along with Tropidechis carinatus, Hoplocephalus stephensii, Rhinoplocephalus nigrescens, and P. porphyriacus. This is consistent with the presence of procoagulant activity in all of these snakes (8). The heavy chain was identified as a train of spots at 35 kDa and with pI generally ranging from 6 to 9. This peptide was not found in the Austrelaps or other three Pseudechis species or in Acanthophis antarcticus. This is in accordance with previous data showing that these snakes possess largely anticoagulant venoms (8). In addition, the light chain of the FXa-like molecule is known to contain Gla residues, and immunoblotting with Gla-specific antibody (Fig. 3A) showed reactivity only to species where the heavy chain was also identified by 2D PAGE and MS. It is also notable that the light chain was only identified by mass spectrometry in three species. This is in agreement with previous data where it has been difficult to detect the presence of this peptide by both Coomassie Blue and silver staining (31). The non-enzymatic cofactor of the prothrombin activator, FVa-like protein, was identified only in Pseudonaja and Oxyuranus, the only two genera known to contain a complete prothrombin activator (15, 37).
|
Metalloproteinases—
In the higher molecular size region, seven of the 18 2D PAGE maps (39%), proteins were matched via MS to proteins homologous to snake venom metalloproteinases (SVMPs). These proteins have been observed in abundance in venoms of viper snakes and also in some elapids (39, 40). The peptides observed identified the Australian elapid metalloproteinases as PIII type SVMPs that contain a reprolysin zinc metalloprotease domain, a disintegrin domain, and a C-terminal cysteine-rich domain.
C-type Lectins—
Protein spots of 15 kDa from six of the 2D PAGE maps were matched via MS to a C-type lectin from Bungarus sp. venom. C-type lectins are non-enzymatic, calcium-de pend ent, sugar-binding proteins and have been identified previously in several Viperidae and a limited number of Elapidae venoms.
Other Protein Families—
Acetylcholinesterase, NGF, 5'-nucleotidase, CRISP, L-amino-acid oxidase, protein-disulfide isomerase, and glucose-regulated protein 78 (GRP78) were also isolated in 2D PAGE maps of some of the venoms. Acetylcholinesterase has been characterized previously from Bungarus fasciatus venom (41) and in this study matched to a 70-kDa series of spots spanning 2 pI units in five venoms (A. antarcticus, Notechis ater niger, N. scutatus, Austrelaps superbus, and P. australis). NGF has been identified recently as a component of six Australian elapid venoms (23) and in this study was identified in a further six snakes. 5'-Nucleotidase activity has been demonstrated in a variety of snake venoms, including the Australian elapids (42). In support of these data, protein spots matching to mammalian 5'-nucleotidase were observed in seven different venoms including the Oxyuranus species. 2D gel spots corresponding to CRISPs were identified in four venoms. L-Amino-acid oxidase has been identified previously in P. australis venom and in this study was identified in 12 Australian elapid venoms, including P. australis. Protein-disulfide isomerase was identified in P. nuchalis (Western brown snake). Two discrete spots of 50 kDa in the PAGE separation of the venom of Notechis ater serventyi (Supplemental Fig. 2C, spots 19 and 38) were matched via MS to GRP78. This protein has also been identified recently in the venom of another Australian elapid, P. textilis (31).
Post-translational Modification of Venom Proteins
We have shown previously that 2D PAGE separation of P. textilis venom proteins gives rise to multiple horizontal trains of spots for several proteins, representing differing numbers of charged groups attached post-translationally (31). This phenomenon was evident in the 2D maps for all 18 species of snake venoms in the current study. A clear example of this was observed for the heavy chain of FXa-like protease (35 kDa) present in the three Notechis species (Supplemental Fig. 1, A–C), the three Pseudonaja species (Supplemental Fig. 1, D–F), P. porphyriacus (Supplemental Fig. 1K), and the two Oxyuranus species (Supplemental Fig. 1, N and O). In N. scutatus, this heavy chain was observed as spots ranging in pI from 8 to 10 (Supplemental Fig. 2A, spots 26–28), whereas the predicted pI from the cDNA sequence (NCBI accession number AAX37262) is 8.7. This suggests modifications are added to the heavy chain isoforms resulting in both loss and gain of charge with only minor influence on molecular size. Virtually all the proteins above 30 kDa in all 18 species were represented by multiple horizontal trains of spots.
Although a number of different post-translational modifications have been reported for snake venom proteins (33), no evidence has been presented for changes in phos pho ryl a tion status. Mammalian FVa has been shown to be phos pho ryl a ted on Ser-692, which may be important for its inactivation by activated protein C (43). This site is not present in the FVa-like molecule from P. textilis, but Rao et al. (15) have described several predicted phos pho ryl a tion sites in this molecule. In an effort to establish the phos pho ryl a tion status of venom proteins in all 18 species we used the ProQ Diamond stain, which has been reported to be specific for phosphoproteins (44). The results in Fig. 3B reveal the presence of strongly staining phos pho ryl a ted proteins over a wide range of molecular sizes in seven of the nine genera studied. Acanthophis and Oxyuranus displayed only low intensity staining. Curiously there was a wide range of staining intensity of protein bands within genera as clearly demonstrated by the variation among the three Notechis species (Fig. 3B, lanes 1–3). The lack of signal in the case of N. ater niger (Fig. 3B, lane 2) is not due to lower protein loading as the same amount of protein was used as in Fig. 1 where the silver-stained gel shows equal loading. No phos pho ryl a tion was detected for FXa-like protease heavy and light chains, which are present in abundance in Notechis, Pseudonaja, Oxyuranus, and other species. The identity of these phos pho ryl a ted proteins remains to be determined.
Glycoproteins were also detected using the specific stain ProQ Emerald (Fig. 3C). Glycosylation of a high molecular weight region corresponding in size to the FVa-like protein was particularly prominent for all three Pseudonaja species. Some labeling was also seen for the Oxyuranus species, which also contain Factor Va-like protein. As expected, neither the Notechis, P. porphyriacus, T. carinatus, R. nigrescens, nor H. stephensii had evidence of gly co sy la tion in the high molecular weight region. Although all of these have prothrombinase activity and possess an FXa-like protease they do not have the FVa molecule.
The rather extensive amount of phos pho ryl a tion observed in the Notechis and other species such as T. carinatus could mean direct phos pho ryl a tion of serine, threonine, and/or tyrosine residues on the protein or the presence of phos pho ryl a ted glycan groups on glycoproteins. Accordingly we treated crude venom from these snakes with either degly co sylase (PNGase F) or phosphatase. A Coomassie-stained gel showed that degly co sy la tion significantly altered the pattern of protein migration (Fig. 4A, lanes 2, 5, 8, and 11), whereas phosphatase treatment (lanes 3, 6, 9, and 12) showed no discernible effect. Staining for phos pho ryl a ted proteins with ProQ Diamond revealed the same degree of loss of phosphate groups after treatment with either degly co sylase or phosphatase (Fig. 4B), suggesting that the phosphate groups were attached to the glycans rather than directly to the protein. The prominent band at 35 kDa lost its ProQ Diamond reactivity for both N. scutatus and N. ater serventyi crude venoms after treatment with degly co sylase and phosphatase (Fig. 4B, arrow). We demonstrated that this band corresponds to the heavy chain of the FXa-like molecule by immunoblotting and showed that glycosidase treatment generated two discrete bands, one corresponding to the native protein (30 kDa) and a second band (27 kDa) corresponding to the calculated mass of the native protein (Fig. 4C). The existence of the upper 30-kDa band suggests incomplete degly co sy la tion. Although only two species of the FXa-like protease heavy chain were detected on the 1D gels, up to seven discrete spots corresponding to this protein were resolved by 2D PAGE. Immunoblotting with an antivenom against N. scutatus venom also detected these two bands in a region corresponding to the FXa-like heavy chain (Fig. 4D). There was also evidence of cleavage of additional proteins of higher molecular size. It is of interest that this antivenom also detected a similar distribution and cleavage pattern of proteins in the other two Notechis species as well as T. carinatus (Fig. 4D). The strong reactivity of the antivenom in the low molecular size region corresponding to neurotoxins is consistent with the reported ability of this antivenom to neutralize neurotoxin activity (45).
|
22 kDa was detected by SNA and UEA (Fig. 5, C and E) that detect sialic acid or one of its constituents. Although this band is in a region corresponding to the light chain of the FXa-like protease, it is unlikely that this protein has been modified because it was detected in several species that do not have a prothrombin activator. Lectin binding in the high molecular weight region was prominent for WGA and ConA for the vast majority of the 18 species, whereas very little reactivity was seen in this region for the other six lectins. However, there was some reactivity seen with RCA120 (N-acetylgalactosamine and galactose) in a few species.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Using mass spectrometry, we identified multiple isoforms of PLA2s in all of the Australian elapid venoms. Snake venom PLA2s have been implicated in a number of physiological processes including hemorrhage, myotoxicity, hemolytic and hypotensive activities as well as edema, platelet aggregation, cardiotoxicity, and pre- and postsynaptic neurotoxicity (48). These proteins have been identified in almost all snake venoms, regardless of genera, and it is the PLA2 enzymes of the Australian elapids that constitute the single most widely studied family of toxins from these snakes (49). Consequently protein sequence data for these enzymes is readily accessible for PMM matching. Multiple protein spots corresponding to these enzymes were detected across the entire pI range of 3–10 and generally in the molecular size region of 13–15 kDa. Previous evidence from Australian elapids, including P. australis, N. scutatus, and A. superbus, indicates that there are multiple PLA2 isoforms within the venom of these snakes (50, 51). A more recent study using cDNA cloning in O. scutellatus identified eight clones coding for PLA2 sequence variants with predicted mature protein sizes ranging from 13.3 to 14.4 kDa and ranging in pI values from 4.6 to 8.4 (34), whereas in the present study, PLA2s ranged in size from 13 to 50 kDa and in pI values from 3 to 10.
The PMM and MS/MS data obtained here (Supplemental Table 1) represents a large number of individual peptides that correspond to a variety of individual PLA2s, consistent with previous data that provide evidence for multiple variants of these enzymes (46). A number of larger molecular size spots from 2D PAGE also contained PLA2 peptides. However this may be due to incomplete dissociation of molecular complexes during the 2D PAGE procedure, the existence of novel higher molecular size forms, and/or post-translational modifications such as gly co sy la tion as suggested by the lectin binding experiments (Fig. 5). An example of this is the Taipoxin complex, a potently active toxin complex within the venom of O. scutellatus composed of three PLA2 chains: , ß, and a gly co sylated -chain (46). The -chain has a predicted molecular size of 14.6 kDa (Swiss-Prot accession number P00616) yet it migrates to 25 kDa on 2D PAGE (Supplemental Fig. 2, spots 1 and 2) likely due to the existence of N-linked carbohydrate structure (46). The -chain has been observed previously to migrate at 26.9 kDa and has a predicted pI of 4.1, which is consistent with that observed on the 2D PAGE (Supplemental Fig. 2N, spots 1 and 2). Natriuretic peptides were the only known protein family not found in the present study possibly due to their small molecular size (35–39 amino acids) (25).
Novel Snake Venom Proteins—
In this study we identified two novel proteins, protein-disulfide isomerase (PDI) and a PLA2 inhibitor not previously reported in the venom of any snake. PDIs are multidomain, multifunctional members of the thioredoxin superfamily that catalyze thiol oxidation, disulfide reduction, and isomerization (52). Isomerization occurs directly by intramolecular disulfide rearrangement or through cycles of reduction and oxidation (53). Twelve peptides corresponding to a PDI sequence in the NCBI protein database were identified in a protein spot from P. nuchalis whose molecular size corresponds with the predicted size of the mature protein (58 kDa). The database sequence represents a conceptual translation of a cDNA sequence isolated from the venom gland of O. scutellatus (NCBI accession number AAY33972) and has not yet been published.3 Because several of the toxic components of the venom are known to contain essential disulfide bridges (e.g. PLA2s, textilinins, and prothrombin activator components), it is possible that PDI plays a role in the synthesis and maintenance of these proteins in their active structures in the venom gland. Interestingly human protein-disulfide isomerase has been shown recently to suppress tissue factor coagulant activity by targeting a disulfide bond in tissue factor that is critical for its coagulation activity (54). This suggests that protein-disulfide isomerase could also play a role as an anticoagulant in the venom.
The second protein, PLA2 inhibitor, has not been described in snake venom, but there are several reports of the existence of such an inhibitor in the blood of American, Asian, and Australian snakes (35, 55). Hains and Broady (35) purified and characterized PLA2 inhibitors from the sera of N. scutatus, N. ater niger, N. ater serventyi, O. scutellatus, O. microlepidotus, and P. textilis. They proposed that PLA2 inhibitors exist in snake blood as a defense mechanism against the toxic PLA2 enzymes produced in the venom glands. We were therefore surprised to find the protein in the venom of P. nuchalis. Three spots were identified corresponding to two forms of the -chain and one of the ß-chain suggesting that it is closely related, if not identical, to the serum inhibitor. It is not immediately clear why a PLA2 inhibitor would be present in the venom particularly in such low abundance compared with PLA2s themselves. However, because not all snake venom PLA2 enzymes have PLA2 activity, it is possible that the inhibitor is specific for only a subclass that could damage the epithelial lining of the venom gland where they are produced. It would seem unlikely that the PLA2 inhibitor that we identified was the result of serum contamination of the venom sample as no other serum proteins were detected in this or any other venom sample.
Novel Proteins in Australian Snake Venoms—
Four families of proteins (vespryn, metalloproteinase, C-type lectin, and acetylcholinesterase) that have been detected previously in snake venoms from other regions are reported here for the first time in venoms of Australian elapids. Pung et al. (38) first described Ohanin as a novel protein from O. hannah (king cobra) venom and subsequently named this protein subfamily "vespryns" (56). They cloned a full-length cDNA for this protein of 1558 bp. The predicted protein contained a propeptide of 22 amino acids at the N terminus and an additional "propeptide" of 63 amino acids at the C terminus. The mature protein was 107 amino acids in length. It was not clear from the report of Pung et al. (56) how the C-terminal region was identified as a propeptide. Ohanins are related in sequence to PRY-SPRY (B30.2-like) domain-containing proteins that include mammalian ryanodine receptor subtypes and the dual specificity kinase, splA, found in Dictyostelium discoideum. There are only three protein sequences in the public domain with >50% shared identity to Ohanin. These related proteins are Thaicobrin from Naja kaouthia (UniProt accession number P82885), an Ohanin-like protein from Lachesis muta (GenBankTM accession number DQ396476), and vespryn-POGU1 from the Australian bearded dragon, Pogona barbata (GenBank accession number AAZ75637) (57). Indeed Fry et al. (57) used cDNA libraries and phylogenetic analysis of transcripts such as vespryn-POGU1 to show that nine toxin types are shared between lizards and snakes, supporting their theory for a single, early origin of the venom system in lizards and snakes.
SVMPs have been identified in some viperid and elapid venoms and are reported to play key roles in the development of such symptoms as hemorrhage, edema, hypotension, hypovolemia, inflammation, and necrosis (39). SVMPs were found in seven of the 18 species examined here. The molecular size varied from 60 to 80 kDa with pI in the range of 4–6. The mass spectrometry data matched to snake venom metalloproteinases from the M12B reprolysin subfamily. This subfamily possesses three different domains: the reprolysin catalytic domain, a disintegrin domain that inhibits the binding of ligands to integrin receptors, and a cysteine-rich domain. The peptide sequences identified here were found to match these regions. It is likely that these enzymes are present in most Australian elapid snakes because it was possible to successfully amplify metalloproteinase cDNAs using PCR primers designed from the cobra cDNA sequence (GenBank accession number AY101383)4 from seven different genera.2 The presence of these enzymes in a wide variety of snakes suggests that they play important roles in toxicity. SVMPs possess multidomain functions, and it has been shown that the catalytic site is not the only crucial domain enabling this class of toxin to cause hemorrhage and inflammation (58). This is evidenced by a form of SVMP, jararhagin C, that lacks the catalytic domain yet is capable of activating an acute inflammatory response in a mouse model (58). Platelets are the main target of these enzymes in the disruption of hemostasis that occurs by inhibition of interaction with collagen and/or von Willebrand factor using various mechanisms that result in systemic bleeding.
C-type lectins were also identified in Australian elapid snake venoms for the first time. These calcium-de pend ent sugar-binding proteins have been characterized extensively from Viperidae venoms. However, only a limited number of C-lectins have been identified in Elapidae venoms, and these remain to be fully characterized. Viper venom C-lectins have been shown to cause edema and increased vascular permeability and to agglutinate erythrocytes (59, 60). The C-lectins identified from the Australian elapids match most closely to a galactose-binding C-lectin sequence from the krait B. fasciatus, suggesting that they also bind this sugar. Although this protein was only identified in six of the 18 species, C-lectin cDNAs have been successfully amplified from seven different genera of Australian elapids5 suggesting that this protein is a common venom component in these snakes.
Acetylcholinesterase activity has been reported in several snake venoms previously; however, only one snake venom acetylcholinesterase has been isolated and characterized, that from B. fasciatus (41). Acetylcholinesterase was identified via MS in five species as an 80-kDa train of spots. This molecular size is consistent with that of B. fasciatus acetylcholinesterase, which is known to contain N-linked gly co sy la tion. Glycosylation of the Australian forms may account for the train of spots observed in this study.
Post-translational Modifications—
We have provided evidence here for at least three types of post-translational modification in the venom proteins from the different genera of the Australian elapids: phos pho ryl a tion, gly co sy la tion, and -carboxylation. Only one previous report, from our laboratory, has provided evidence for the presence of phos pho ryl a ted snake venom proteins (31). It seems unlikely that this was artifactual binding because the phos pho ryl a ted spots correlated well with protein staining, and the ProQ Diamond technique has been shown to be very specific for protein phos pho ryl a tion (44). This method of staining failed to detect phosphate in DNA or RNA and did not detect sulfated glycans. Birrell et al. (31) demonstrated that ProQ Diamond detected proteins in P. textilis that corresponded in size to the major toxins in that venom including FVa-like and FXa-like proteins as well as PLA2s. In this study, phos pho ryl a ted proteins were detected in most snake species, and they varied in molecular size from 15 kDa up to greater than 100 kDa. Proteins are normally phos pho ryl a ted as parts of signaling mechanisms, leading to alterations in their conformation and/or their capacity to interact with other proteins. In this case it is more likely that the phos pho ryl a ted species detected are glycoproteins containing phos pho ryl a ted carbohydrate. Phosphorylated glycoproteins are widely distributed in nature and play key roles in a variety of biological processes. For example, soluble lysosomal proteins are synthesized in the endoplasmic reticulum and are cotranslationally gly co sylated on specific asparagine residues. As these proteins move through the secretory pathway, they are selectively recognized by a phosphotransferase that initiates a two-step reaction that results in the generation of the mannose 6-phosphate modification on specific N-linked oligosaccharides. This modification causes the lysosomal protein to bind mannose 6-phosphate receptors as part of the process of endocytosis and lysosomal targeting (61). Another example of a glycoprotein in which the carbohydrate moiety is phos pho ryl a ted is cathepsin D (62). Cathepsin D is an aspartic peptidase that is found at elevated levels in some tumor tissues; however, the role of the phos pho ryl a ted carbohydrate is unknown. To understand the role of putative phos pho ryl a ted glycoproteins in snake venom it will be necessary to identify these proteins, determine whether phos pho ryl a tion status influences their interaction with other proteins, and determine whether phos pho ryl a tion alters the function of these proteins in specific functional assays, e.g. proteolytic activity, coagulability, or neurotoxicity.
Glycosylation of proteins was detected using lectin binding and ProQ Emerald staining. A high molecular weight signal was detected in the venom from several species that corresponded to FVa-like protein in 1D SDS-PAGE (Fig. 3C). This is in good agreement with previous results using 2D PAGE with P. textilis that detected spots of the same molecular size as FVa-like and FXa-like proteins. This was confirmed with ConA (glucose- and mannose-specific) and WGA (GlcNAc- and sialic acid-specific) lectins. It is not evident what role this modification may have on components of the prothrombin activator, but it is possible that the abundance of gly co sy la tion of these proteins influences their stability and/or function. Increased stability would result in greater diffusion and transport of gly co sylated toxins upon envenomation.
Clinical Implications—
In general, the results of this study support previous clinical data from patients envenomated by Australian elapid species. The major clinical effects of Australian elapid venoms include paralysis, coagulopathy, and muscle destruction, and these symptoms can be attributed largely to three groups of proteins, neurotoxins, prothrombin activators, and PLA2s (8). Inspection of the 2D maps showed that these three groups of proteins account for a large proportion of the total venom proteins, confirming their high abundance. We identified several venom proteins not known previously in Australian snake venoms, but reported in other snake venoms, that may have clinical manifestations. For example, SVMPs from several Viperidae venoms can cause hemorrhage, edema, hypotension, hypovolemia, inflammation, and necrosis (39). Similarly C-type lectins that have also been characterized from Viperidae venoms can cause edema and increased vascular permeability and can agglutinate erythrocytes (59, 60). Thus it may be possible that some clinical effects attributed previously to PLA2s, such as necrosis and inflammation, are in fact caused by the SVMPs. The clinical effects of other proteins identified in Australian elapid snakes for the first time here also remain to be determined. Moreover some of these proteins may have unique activities applicable to the development of new human therapeutics.
Conclusions—
We have described here a comprehensive analysis of the venom proteins of 18 Australian elapid snakes that are among the most venomous snakes worldwide. This represents a significant advance in the proteomics of snake venom proteins and an important resource for comparison of protein patterns both between species of the same genera and across genera. The vast majority of these proteins were identified using mass spectrometry of spots isolated from 2D PAGE.
Because a significant number of the peptides isolated failed to match with any peptides in the NCBI non-redundant protein database it was necessary to use de novo peptide sequencing by tandem mass spectrometry, which significantly increased the total number of proteins identified. In all, we identified six previously undescribed families of proteins in Australian elapid snakes. Some of these, including vespryns, metalloproteinases, C-type lectins, and 5'-nucleotidase were described previously in non-Australian snake venoms (38, 63–66). Protein-disulfide isomerase and PLA2 inhibitor have not been reported previously in any snake venom, although a very recent report has described a cDNA coding for PDI (67), and a PLA2 inhibitor has been reported in snake serum (36). In addition to identifying all the major venom proteins in the Australian snakes, we partially characterized them using specific antibodies, phosphoprotein- and glycoprotein-specific stains, enzymatic digestion, lectin binding, and antivenom reactivity. Further characterization of these proteins will assist in delineating their role in the venoms, their clinical effects, and their potential as therapeutic agents.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Published, MCP Papers in Press, February 21, 2007, DOI 10.1074/mcp.M600419-MCP200
1 The abbreviations used are: FXa, Factor Xa; FVa, Factor Va; 2D, two-dimensional; 1D, one-dimensional; Gla, -carboxyglutamate; PNGase F, peptidyl-N-glycosidase F; ConA, concanavalin A; WGA, wheat germ agglutinin; RCA120, Ricinus communis agglutinin 120; GalNAc, N-acetylgalactosamine; GlcNAc, N-acetylglu co sa mine; PLA2, phospholipase A2; NGF, nerve growth factor; CRISP, cysteine-rich secreted protein; PDI, protein-disulfide isomerase; SVMP, snake venom metalloproteinase; GRP78, glucose-regulated protein 78; ACTH, adrenocorticotropic hormone; clip, corticotropin-like intermediate lobe peptide; PMM, peptide mass mapping; SNA, Sambucus nigra lectin; UEA, Ulex europaeus agglutinin. 
2 G. W. Birrell, S. T. H. Earl, P. P. Masci, J. de Jersey, and M. F. Lavin, manuscript in preparation.
3 R. E. Welton and J. N. Burnell, unpublished data.
4 D. Sako and G. D. Shaw, unpublished data.
5 S. T. H. Earl, G. W. Birrell, P. P. Masci, J. de Jersey, and M. F. Lavin, manuscript in preparation.
* This work was supported by QRxPharma and the Australian Research Council Linkage Program. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
S The on-line version of this article (available at http://www.mcponline.org) contains supplemental material.
¶ Both authors contributed equally to this work.
** Present address: The Queensland Inst. of Medical Research, P. O. Royal Brisbane Hospital, Brisbane 4029, Australia.
¶¶ To whom correspondence should be addressed: The Queensland Cancer Fund Research Unit, The Queensland Inst. of Medical Research, P. O. Box Royal Brisbane Hospital, Herston, Brisbane 4029, Australia. Tel.: 617-3362-0341; Fax: 617-3362-0106; E-mail: martin.lavin{at}qimr.edu.au
|
REFERENCES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
