Snake venom toxins

The medical threat of mamba envenoming in sub-Saharan Africa revealed by genus-wide analysis of venom composition, toxicity and antivenomics profiling of available antivenoms

Mambas (genus Dendroaspis) are among the most feared of the medically important elapid snakes found in sub-Saharan Africa, but many facets of their biology, including the diversity of venom composition, remain relatively understudied. Here, we present a reconstruction of mamba phylogeny, alongside genus-wide venom gland transcriptomic and high-resolution top-down venomic analyses. Whereas the green mambas, D. viridis, D. angusticeps, D. j. jamesoni and D. j. kaimosae, express 3FTx-predominant venoms, black mamba (D. polylepis) venom is dominated by dendrotoxins I and K. The divergent terrestrial ecology of D. polylepis compared to the arboreal niche occupied by all other mambas makes it plausible that this major difference in venom composition is due to dietary variation. The pattern of intrageneric venom variability across Dendroaspis represented a valuable opportunity to investigate, in a genus-wide context, the variant toxicity of the venom, and the degree of paraspecific cross-reactivity between antivenoms and mamba venoms. To this end, the immunological profiles of the five mamba venoms were assessed against a panel of commercial antivenoms generated for the sub-Saharan Africa market. This study provides a genus-wide overview of which available antivenoms may be more efficacious in neutralising human envenomings caused by mambas, irrespective of the species responsible. The information gathered in this study lays the foundations for rationalising the notably different potency and pharmacological profiles of Dendroaspis venoms at locus resolution. This understanding will allow selection and design of toxin immunogens with a view to generating a safer and more efficacious pan-specific antivenom against any mamba envenomation.

BIOLOGICAL SIGNIFICANCE

The mambas (genus Dendroaspis) comprise five especially notorious medically important venomous snakes endemic to sub-Saharan Africa. Their highly potent venoms comprise a high diversity of pharmacologically active peptides, including extremely rapid-acting neurotoxins. Previous studies on mamba venoms have focused on the biochemical and pharmacological characterisation of their most relevant toxins to rationalize the common neurological and neuromuscular symptoms of envenomings caused by these species, but there has been little work on overall venom composition or comparisons between them. Only very recently an overview of the composition of the venom of two Dendroaspis species, D. angusticeps and D. polylepis, has been unveiled through venomics approaches. Here we present the first genus-wide transcriptomic-proteomic analysis of mamba venom composition. The transcriptomic analyses described in this paper have contributed 29 (D. polylepis), 23 (D. angusticeps), 40 (D. viridis), 25 (D. j. jamesoni) and 21 (D. j. kaimosae), novel full-length toxin sequences to the non-redundant Dendroaspis sequence database. The mamba genus-wide venomic analysis demonstrated that major D. polylepis venom components are Kunitz-fold family toxins. This feature is unique in relation to the relatively conserved three-finger toxin (3FTx)-dominated venom compositions of the green mambas. Venom variation was interpreted in the context of dietary variation due to the divergent terrestrial ecology of D. polylepis compared to the arboreal niche occupied by all other mambas. Additionally, the degree of cross-reactivity conservation of mamba venoms was assessed by antivenomics against a panel of commercial antivenoms generated for the sub-Saharan Africa market. This study provides a genus-wide overview to infer which available antivenoms may be capable of neutralising human envenomings caused by mambas, irrespective of the species responsible. The information gathered in this study lays the foundations for rationalising the pharmacological profiles of mamba venoms at locus resolution. This understanding will contribute to the generation of a safer and more efficacious pan-Dendroaspis therapeutic antivenom against any mamba envenomation.

Predicted secondary structure of snake venom toxins from their primary structures

We have predicted the secondary structure of 38 snake venom toxins using the method of Chou and Fasman. Our predictions indicate that beta-chain and random coil structures predominate in these proteins. The conformations of long neurotoxins, short neurotoxins and cytotoxins are less similar than previously believed. Cytotoxins contain 40--50% of beta-structure and they form a notably homogeneous group. Short neurotoxins contain less beta-structure (13--30%) and more random coil than cytotoxins, and they also form a more heterogeneous group in terms of secondary structure. The characteristics of long neurotoxins are intermediate to the above mentioned groups. Experimental evidence supporting these propositions is discussed.

Proline brackets and identification of potential functional sites in proteins: toxins to therapeutics

Protein toxins induce their specific pharmacological effects through protein protein interaction. Identification of these protein-protein interaction sites could lead to prototypes of highly specific therapeutic agents. However, deciphering the structure function relationships of protein toxins and locating the functional sites is a difficult, tedious and cumbersome task. We recently developed a novel predictive method to identify potential protein protein interaction sites directly from the amino acid sequence of a protein (R. M. Kini and H. J. Evans (1996) FEBS Lett. 385, 81-86) based on the presence of proline residues, a common residue found predominantly in the flanking segments of protein-protein interaction sites (R. M. Kini and H. J. Evans (1996) Biochem. Biophys. Res. Commun. 212, 1115-1124). It is a simple and straight-forward method. This review describes the new method and its application to solve structure function relationships of protein toxins. The method is useful in identifying functional sites in toxins, despite the subtle and complex nature of their structure function relationships and saves significant amounts of time and money.

Black widow spider alpha-latrotoxin: a presynaptic neurotoxin that shares structural homology with the glucagon-like peptide-1 family of insulin secretagogic hormones

alpha-Latrotoxin is a presynaptic neurotoxin isolated from the venom of the black widow spider Latrodectus tredecimguttatus. It exerts toxic effects in the vertebrate central nervous system by depolarizing neurons, by increasing [Ca2+]i and by stimulating uncontrolled exocytosis of neurotransmitters from nerve terminals. The actions of alpha-latrotoxin are mediated, in part, by a GTP-binding protein-coupled receptor referred to as CIRL or latrophilin. Exendin-4 is also a venom toxin, and it is derived from the salivary gland of the Gila monster Heloderma suspectum. It acts as an agonist at the receptor for glucagon-like peptide-1(7-36)-amide (GLP-1), thereby stimulating secretion of insulin from pancreatic beta-cells of the islets of Langerhans. Here is reported a surprising structural homology between alpha-latrotoxin and exendin-4 that is also apparent amongst all members of the GLP-1-like family of secretagogic hormones (GLP-1, glucagon, vasoactive intestinal polypeptide, secretin, pituitary adenylyl cyclase activating polypeptide). On the basis of this homology, we report the synthesis and initial characterization of a chimeric peptide (Black Widow GLP-1) that stimulates Ca2+ signaling and insulin secretion in human beta-cells and MIN6 insulinoma cells. It is also reported here that the GTP-binding protein-coupled receptors for alpha-latrotoxin and exendin-4 share highly significant structural similarity in their extracellularly-oriented amino-termini. We propose that molecular mimicry has generated conserved structural motifs in secretagogic toxins and their receptors, thereby explaining the evolution of defense or predatory strategies that are shared in common amongst distantly related species including spiders, lizards, and snakes. Evidently, the toxic effects of alpha-latrotoxin and exendin-4 are explained by their ability to interact with GTP-binding protein-coupled receptors that normally mediate the actions of endogenous hormones or neuropeptides.

Flanking proline residues identify the L-type Ca2+ channel binding site of calciseptine and FS2

Calciseptine and FS2 are 60-amino acid polypeptides, isolated from venom of the black mamba (Dendroaspis polylepis polylepis), that block voltage-dependent L-type Ca2+ channels. We predicted that these polypeptides contain an identical functional site between residues 43 and 46 by searching for proline residues that mark the flanks of protein-protein interaction sites [Kini, R. M., and Evans, H. J. (1966) FEBS Lett. 385, 81-86]. The predicted Ca2+ channel binding site also occurs in closely related toxins, C10S2C2 and S4C8. Therefore, it is likely that these toxins also will block L-type Ca2+ channels. To test the proposed binding site on calciseptine and FS2, an eight-residue peptide, named L-calchin (L-type calcium channel inhibitor), was synthesized and examined for biological activity. As expected for an L-type Ca2+ channel blocker, L-calchin reduced peak systolic and developed pressure in isolated rat heart Langendorff preparations without affecting diastolic pressure or heart rate. Furthermore, L-calchin caused a voltage-independent block of L-type Ca2+ channel currents in whole-cell patch-clamped rabbit ventricular myocytes. Thus the synthetic peptide exhibits the L-type Ca2+ channel blocking properties of the parent molecules, calciseptine and FS2, but with a lower potency. These results strongly support the identification of a site in calciseptine and FS2 that is important for binding to L-type Ca2+ channels and reinforce the importance of proline brackets flanking protein-protein interaction sites.

Characterization of the lower-molecular-mass fraction of venoms from Dendroaspis jamesoni kaimosae and Micrurus fulvius using capillary-electrophoresis electrospray mass spectrometry

Capillary electrophoresis (CE) with electrospray ionization (ESI) and selected ion-monitoring mass-spectrometric (SIM-MS) detection has been used to provide as much information as possible about the lower molecular-mass fraction (peptides of molecular masses up to 8500 Da) of the venoms of Dendroaspis jamesoni kaimosae (Jameson's Mamba) and Micrurus fulvius (Eastern Coral Snake). Method development was based on the venom of D. jamesoni kaimosae, which contains some previously described peptides, with subsequent application to the completely unknown venom of M. fulvius. CE-ESI-SIM-MS provides a rapid and extremely sensitive method for the detection and molecular-mass determination of peptides present in venoms. It has been utilized to provide molecular masses and thus, by inference, confirmation of the peptide compositions for those toxins which have been previously described in the literature. Our methodology indicates the presence of 83 peptides in the venom of D. jamesoni kaimosae and 49 peptides in the venom of M. fulvius in the molecular-mass range 6000-8500 Da.

Application of electrospray mass spectrometry and matrix-assisted laser desorption ionization time-of-flight mass spectrometry for molecular weight assignment of peptides in complex mixtures

Electrospray mass spectrometry (ES/MS) and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI/TOF/MS) were used to provide mass spectra from seven elapid snake venoms. Spectral interpretation was much simpler for MALDI/TOF/MS. ES/MS proved more useful for the provision of molecular weight data for very closely related peptides, but suppression of higher molecular weight compounds was seen to occur during flow injection analysis. MALDI/TOF/MS proved useful for providing a complete picture of the venom, but the low resolution led to obscuring of major ions, and the mass accuracy was poorer for known peptides. Suppression also occurred during MALDI/TOF/MS but could be overcome using alternative matrices because the spectra were very dependent on the choice of matrix. ES/MS and MALDI/TOF/MS provide complementary and confirmatory information such that for the anal sis of complex peptide mixtures (snake venoms), the use of both techniques is desirable.

Cloning of a novel TGF-beta related cytokine, the vgr, from rat brain: cloning of and comparison to homologous human cytokines

Here we describe cloning of a TGF-beta related cytokine from a rat brain cDNA library. This novel cytokine, the vgr (vegetal related), is homologous to the vegetal (Vg1) gene of Xenopus (DL Weeks and DA Melton, Cell, 51:861-867, 1987). In rat brain mRNA a single 3.5 kb RNA could be detected by Northern blot analysis. Thus, this new cytokine is constitutively expressed in the central nervous system. A monoclonal antibody reactive with a synthetic peptide of vgr revealed a faint vgr-like immunoreactivity throughout the CNS, with more pronounced staining of hippocampal neurons, ependymal cells, cells of the choroid plexus, and hypophysis. Using the rat cDNA, two homologous human cytokine cDNAs encoding the human vgr and op-1 were cloned.

Snake venom toxins--I. The primary structure of a long neurotoxin S4C6 from Aspidelaps scutatus (shield or shield-nose snake) venom

1. Long neurotoxin S4C6 from Aspidelaps scutatus venom was purified by gel filtration and ion exchange chromatography (Joubert, 1987). 2. It contains 68 amino acids including 10 half-cystines. The toxicity of toxin S4C6 was determined and a LD50 of 0.13 +/- 0.04 micrograms/g mouse was found. 3. The complete primary structure of long neurotoxin S4C6 has been elucidated. In the toxin the 10 structurally invariant amino acids of the neurotoxins and cytoxins and the five functionally invariant amino acids of the neurotoxins are conserved.

The complete primary structures of two reduced and S-carboxymethylated Angusticeps-type toxins from Dendroaspis angusticeps (green mamba) venom

Toxins C10S2C2 and C13S1C1 from Dendroaspis angusticeps venom were purified by gel filtration and ion-exchange chromatography. Whereas toxin C10S2C2 comprises 60 amino acids, toxin C13S1C1 contains only 58 but they each include eight half-cystine residues. The complete primary structures of the toxins have been elucidated. The sequences and the invariant amino acids of toxins C10S2C2 and C13S1C1 from D. angusticeps venom resemble those of the angusticeps-type toxins. In the two toxins the ten structurally invariant amino acids of the neurotoxins and cytotoxins are conserved, but the toxins contain none of the three functionally-invariant amino acids of the neurotoxins. Further, the eight cystine residues of the angusticeps-type toxins are in similar locations to those in short neurotoxins of known structure so they are presumed to link similarly. The only structural characteristic of the angusticeps-type toxins which binds them together as a group, is the serine residue in position 5. The toxicities of the angusticeps-type toxins differ among themselves but appear to be of considerably lower toxicity relative to that of the neurotoxin group.

Some properties and the complete primary structures of two reduced and S-carboxymethylated polypeptides (S5C1 and S5C10) from Dendroaspis jamesoni kaimosae (Jameson's mamba) venom

Two polypeptides (protein S5C1 and toxin S5C10) were purified from Dendroaspis jamesoni kaimosae venom. Whereas protein S5C1 comprises 61 amino acid residues, toxin S5C10 contains 58 and they each comprise four disulphide bridges. The complete primary structures of the two polypeptides have been elucidated. The sequences of protein S5C1 and toxin S5C10 are structurally homologous to the short neurotoxins Type I, but they are much less toxic. In toxin S5C10 one of the functionally invariant amino acid residues, lysine 26, of the Type I neurotoxin has been replaced by a serine. In contrast protein S5C1 has the feature that it contains ten or eleven structurally invariant amino acids and apparently only one of the five functionally invariant residues.

Conformation of snake toxic polypeptides studied by a method of prediction and circular dichroism

Short and long neurotoxins as well as cardiotoxins belong to three distinct families of homologous toxic polypeptides extracted from cobra venoms. A study of their conformation was undertaken by using the method of Chou and Fasman for prediction of secondary structures of proteins. To improve the reliability of this method, an averaging scheme was developed. The data obtained showed that all toxins have a predominant trend for beta-sheet nucleation. Moreover, predicted beta-sheet strands fitted well those actually observed from X-ray data. Thus, it seems that all toxins share similarities in their secondary structure. This proposition was supported by a comparative study of the CD spectra of a set of toxins. Nevertheless, the present data suggest also that each type of toxins possesses localized structural individualities which might be responsible for the biological and/or immunological specificities.

Molecular evolution of snake venom toxins

Phylogenetic trees were constructed for 62 venom toxins of snakes of Proteroglyphae suborder using matrix method. The resulting tree from Minimum Spanning Tree-Cluster Analysis technique had the lowest "percent deviation" (8.55). The taxonomic relationship of these toxins agrees very well with zoological opinions. However, the appearance of the tree did not directly provide a plausible evolutionary model for the toxins. A model was derived from nodal ancestral sequence calculations, comparisons between intra- and intergenerical rates of amino acid change, and generally held ideas about protein evolution. According to the model, short neurotoxin is the ancient form of snake venom toxins. The courses of evolution leading to the present intraspecific homologous toxins are explained by gene duplication and allelomorphism.

Amino acid sequences of neurotoxins I and III of the elapidae snake Naja mossambica massambica

The amino acid sequences of two neurotoxins of the African cobra Naja mossambica mossambica have been determined using almost uniquely phenylisothiocyanate degradation in a liquid protein sequencer programmed alternatively with 'protein' and 'peptide' programs. When compared to known sequences of so-called 'short' neurotoxins belonging to other Elapidae snakes, neurotoxins I and III of Naja mossambica mossambica are very similar to the cobrotoxin, a neurotoxin isolated from the formosan cobra Naja atra atra.

Snake venom toxins. The amino-acid sequence of a short-neurotoxin homologue from Dendroaspis polylepis polylepis (black mamba) venom

The third most abundant component of black mamba venom, named FS2, was sequenced with the aid of sequenator studies and peptides derived by tryptic and chymotryptic digestion. Cyanogen bromide digests provided extra information to support the proposed structure. This protein is a homologue of the short neurotoxins of snake venom, but is much less toxic. Its structure is quite different from both neurotoxins and the other mamba proteins, called angusticeps types (neurotoxin homologues). Comparison of the known angusticeps-type toxins from mamba venom with mamba neurotoxins and each other leads to proposals that these proteins of low toxicity are inventions of the group of mambas and that three different, as yet unknown, functions will be associated with the three subgroups that are discernable.

Primary structures of cardiotoxin analogues II and IV from the venom of Naja jaja atra

Cardiotoxin analogues II and IV were isolated from the venom of Naja naja atra by gel filtration on Sephadex G-50 followed by CM-cellulose chromatography. The venom contains at least four cardiotoxin analogues that account for about 54% of the weight of the lyophilized crude venom. These four cardiotoxin analogues, named cardiotoxin analogues I, II, III, and IV, show strong cytotoxicity to Yoshida sarcoma cells but the lethal toxicity is one-order less. These toxins contain 60 amino acid residues in a single peptide chain. Cardiotoxin analogue IV differs from cardiotoxin analogue II only by the presence of arginine in place of a leucine residue at position 1. A comparison of the amino acid sequences of these toxins with that of cobrotoxin, a neurotoxin containing 62 amino acid residues obtained from the same snake venom, shows that about 20 amino acid residues, including 8 half cystine residues, are identical, assuming 3 residues deletion and 2 residues insertion in the cardiotoxin molecule.

Snake venom toxins. Purification and properties of low-molecular-weight polypeptides of Dendroaspis polylepis polylepis (black mamba) venom

Twelve low-molecular-weight proteins, of which eleven have subcutaneous LD50 values of less than 40 mug/g mouse, were purified from Dendroaspis polylepis polylepis venom. Ion-exchange chromatography on Amberlite CG-50 and ion-exchange chromatography on carboxymethyl-cellulose and/or phosphocellulose was used for the purification. The amino-terminal sequences of these proteins were determined and used to indicate that five groups of low-molecular-weight polypeptides are to be found in black mamba venom. Proteins from two of these groups which have low toxicity individually, when used together show synergism, in that their toxicity in combination is greater than the sum of their individual toxicities.

Purification of six neurotoxins from the venom of Dendroaspis viridis. Primary structure of two long toxins

Six neurotoxins were purified from Dendroaspis viridis venom using gel filtration and equilibrium chromatography. The amino acid sequences of two of these neurotoxins (72 and 73 residues, five disulphide bridges) have been determined using almost exclusively automated Edman degradation. These two sequences are very similar: the only differences lies in the presence of one extra glycine at the C-terminal end of one of them. There is a good homology with the sequences of toxins now isolated from other Elapidae and Hydrophidae venoms.

Amino-acid sequence of toxin III of Naja haje

The complete amino acid sequence of toxin III of Naja haje (72 residues) has been established mainly by use of a protein sequenator (identification of 70 residues). The two C-terminal residues have been determined by digestion with carboxypeptidases A and B. Addition of succinylated protein or peptide greatly improved the performance of the sequenator for the Edman degradation of peptides: on one peptide (39 residues) degradation went to step 34 with a protein program and on two peptides (10 and 13 residues) degradation reached the last amino acid with a peptide program (use of dimethylbenzylamine). Amino acid analysis of tryptic peptides obtained by digestion of the C-terminal cyanogen bromide peptide are in full agreement with the sequence established by automatic degradation. The sequence of toxin III of Naja haje is unique and is very similar to that of Naja nivea alpha (although there are 9 differences), of Naja melanoleuca b (11 differences) and also to that of Naja naja A (18 differences).