Studies on sea snake venom.

Erabutoxins a and b are neurotoxins isolated from venom of a sea snake Laticauda semifasciata (erabu-umihebi). Amino acid sequences of the toxins indicated that the toxins are members of a superfamily consisting of short and long neurotoxins and cytotoxins found in sea snakes and terrestrial snakes. The short neurotoxins to which erabutoxins belong act by blocking the nicotinic acetylcholine receptor on the post synaptic membrane in a manner similar to that of curare. X-ray crystallography and NMR analyses showed that the toxins have a three-finger structure, in which three fingers made of three loops emerging from a dense core make a gently concave surface of the protein. The sequence comparison and the location of essential residues on the protein suggested the mechanism of binding of the toxin to the acetylcholine receptor. Classification of snakes by means of sequence comparison and that based on different morphological features were inconsistent, which led the authors to propose a hypothesis "Evolution without divergence."

Modeling the third loop of short-chain snake venom neurotoxins: roles of the short-range and long-range interactions.

The influence of long-range interactions on local structures is an important issue in understanding protein folding process and protein structure stability. Using short-chain snake venom neurotoxin as a model system, we have studied the conformational properties of eight different loop III sequences either in the environment of one of the short-chain neurotoxin, erabutoxin b (PDB ID 1nxb), or in free state by Monte Carlo simulated annealing method. The surrounding protein structure was found to be crucial in stabilizing the loop conformation. Although all the eight peptides prefer type V beta turn in solution, three of them (KPGI, KPGV, KSGI) turn to type II beta turn and the other five (KKGI, KKGV, KNGI, KQGI, and KRGV) are confined to more rigid type V beta turn conformation in the protein structure. Using flexible tetra-glycine-peptide to screen the backbone conformational space in the protein environment also validates the results. This study shows that long-range interactions do contribute to the stability and the types of conformation for a surface loop in protein, while short-range interactions may only provide candidate conformations, which then have to be filtered by the long-range interactions further.

High resolution x-ray analysis of two mutants of a curaremimetic snake toxin.

A previous mutational analysis of erabutoxin a (Ea), a curaremimetic toxin from sea snake venom, showed that the substitutions S8G and S8T caused, respectively, 176-fold and 780-fold affinity decreases for the nicotinic acetylcholine receptor (AchR). In view of the fact that the side-chain of Ser8 is buried in the wild-type toxin, we wondered whether these affinity changes reflect a direct binding contribution of S8 to the receptor and/or conformational changes that could have occurred in Ea as a result of the introduced mutations. To approach this question, we solved X-ray structures of the two mutants S8G and S8T at high resolution (0.18 nm and 0.17 nm, with R factors of 18.0% and 17.9%, respectively). The data show that none of the mutations significantly modified the toxin structure. Even within the site where the toxin binds to the receptor the backbone conformation remained unchanged. Therefore, the low affinities of the mutants S8T and S8G cannot be explained by a large conformational change of the toxin structure. Although we cannot exclude the possibility that undetectable structural changes have occurred in the toxin mutants, our data support the view that, although buried between loop I and II, S8 is part of the functional epitope of the toxin.

The length of a single turn controls the overall folding rate of "three-fingered" snake toxins.

Snake curaremimetic toxins are short all-beta proteins, containing several disulfide bonds which largely contribute to their stability. The four disulfides present in snake toxins make a "disulfide beta-cross"-fold that was suggested to be a good protein folding template. Previous studies on the refolding of snake toxins (Ménez, A. et al. (1980) Biochemistry 19, 4166-4172) showed that this set of natural homologous proteins displays different rates of refolding. These studies suggested that the observed different rates could be correlated to the length of turn 2, one out of five turns present in the toxins structure and close to the "disulfide beta-cross". To demonstrate this hypothesis, we studied the refolding pathways and kinetics of two natural isotoxins, toxin alpha (Naja nigricollis) and erabutoxin b (Laticauda semifasciata), and two synthetic homologues, the alpha mutants, alpha60 and alpha62. These mutants were designed to probe the peculiar role of the turn 2 on the refolding process by deletion or insertion of one residue in the turn length that reproduced the natural heterogeneity at that locus. The refolding was studied by electrospray mass spectrometry (ESMS) time-course analysis. This analysis permitted both the identification and quantitation of the population of intermediates present during the process. All toxins were shown to share the same sequential scheme for disulfide bond formation despite large differences in their refolding rates. The results presented here demonstrate definitely that no residues except those forming turn 2 accounted for the observed differences in the refolding rate of toxins.

Functional determinants by which snake and cone snail toxins block the alpha 7 neuronal nicotinic acetylcholine receptors.

Snakes and cone snails produce toxins which block muscular and/or neuronal nicotinic acetylcholine receptors (AChRs). This paper mostly focuses on the determinants by which a snake long chain curaremimetic toxin and the cone snail toxin ImI bind specifically to the alpha 7 neuronal receptor. In both cases, the site involves a small turn-like structure constrained by two half-cystines.

Structure of dimeric and monomeric erabutoxin a refined at 1.5 A resolution.

Erabutoxin a has been crystallized in its monomeric and dimeric forms. The structures were refined at 1.50 and 1.49 A resolution, respectively, using synchrotron radiation data. The crystals belong to space group P212121, with cell dimensions a = 49.84, b = 46.62, c = 21.22 A for the monomer and a = 55.32, b = 53.54, c = 40.76 A for the dimer. Using starting models from earlier structure determinations, the monomeric structure refined to an R value of 16.7% (8004 unique reflections, 17.0-1.50 A resolution range), while the dimeric structure has been solved by the molecular-replacement method with a final R value of 16.9% (19 444 unique reflections, 17.4-1.49 A resolution range). The high-resolution electron-density maps clearly revealed significant discrete disorder in the proteins and allowed an accurate determination of the solvent structure. For the monomer, the side chains of six residues were modelled with alternate conformers and 106 sites for water molecules and one site for a sulfate ion were included in the final model, whereas for the dimer, 206 sites for water molecules were included and both C-terminal residues together with the side chains of 11 residues adopted alternative conformations. A comparison was made with earlier structure determinations. The features of the solvent structure of the erabutoxin molecules are discussed in detail.

Genomic structures of cardiotoxin 4 and cobrotoxin from Naja naja atra (Taiwan cobra).

Two genomic DNAs with the size of 2.3 kb and 2.4 kb, which were isolated from the liver of Naja naja atra (Taiwan cobra), encoded the precursors of cardiotoxin 4 and cobrotoxin, respectively. Both genes shared virtually identical overall organization with three exons separated by two introns, which were inserted in the similar positions of the gene's coding regions. Moreover, their nucleotide sequences shared approximately 84.2% identity. This result reveals the evolutionary relationship between cardiotoxin and cobrotoxin. The exon/intron structures of cardiotoxin 4 and cobrotoxin genes were similar to that reported for erabutoxin c gene, a neurotoxin genomic DNA from a sea snake (Laticauda semifasciata). However, in contrast to the finding that the intron 2 of these genes had a similar size, a notable variation with the size of intron 1 was observed (1233 bp, 1269 bp and 197 bp for cardiotoxin 4, cobrotoxin and erabutoxin c genes, respectively). The different size with intron 1 is due to the middle region at the first intron of cardiotoxin 4 and cobrotoxin genes, which encoded small nucleolar RNA (snoRNA), being absent in that of erabutoxin c gene. These results, together with the finding of the potential mobility of snoRNA genes during evolution, suggest that intron insertions or deletions of snoRNA genes occur with the evolutionary divergence of snake neurotoxins and cardiotoxins.

High-level production and isotope labeling of snake neurotoxins, disulfide-rich proteins.

The aim of this work was to produce and to label snake neurotoxins, disulfide-rich proteins. A mutant of a snake toxin, erabutoxin a, was used as a model. Its N-terminal part was fused to ZZ, a synthetic IgG-binding domain of protein A (B. Nilsson et al., 1987, Protein Eng. 1, 107-113), thus preventing degradation in the bacterial cytoplasm and providing a simple affinity-purification method on IgG Sepharose. A soluble fusion protein was obtained with a yield of 60 mg/L, corresponding to 20 mg/L toxin. The toxin moiety was folded on the column while the hybrid was still bound. The oxidoreducing conditions for the refolding were optimized and were found to be oxidative but with a need for reducing molecules. The concentration of the hybrid bound to the column could be increased up to 3.3 mg/ml without significantly altering the folding process. CNBr cleavage of the fusion protein followed by a purification step yielded about 2 mg of biologically active toxin mutant per gram of dry cell weight. This procedure was applied to produce 55 mg of a toxin uniformly labeled with 15N.

Transformation of a non-enzymatic toxin into a toxoid by genetic engineering.

Curaremimetic toxins are typical non-enzymatic toxins that bind to their target [the nicotinic acetylcholine receptor (AChR)] through multiple residues. Nevertheless, we show that the concomitant substitutions of only three of the ten functionally important residues of such a toxin sufficed to cause an affinity decrease of the toxin for AChR that is higher than four orders of magnitude. Despite these triple mutations, the overall conformation of the mutated protein remains similar to that of a related recombinant toxin, as judged from both circular dichroism analysis and investigation of antigenicity, using monoclonal and polyclonal antibodies. Furthermore, we show that the detoxified toxin is capable of eliciting antibodies that neutralize the binding of a wild-type toxin to AChR. Therefore, transformation of a non-enzymatic toxin into a toxoid can be achieved, like in the case of enzymatic toxins, by introducing a small number of mutations at positions identified to be critical for expression of toxicity.

Probing local secondary structure by fluorescence: time-resolved and circular dichroism studies of highly purified neurotoxins.

The relationship between beta-sheet secondary structure and intrinsic tryptophan fluorescence parameters of erabutoxin b, alpha-cobratoxin, and alpha-bungarotoxin were examined. Nuclear magnetic resonance and x-ray crystallography have shown that these neurotoxins have comparable beta-sheet, beta-turn, and random coil secondary structures. Each toxin contains a single tryptophan (Trp) residue within its beta-sheet. The time-resolved fluorescence properties of native erabutoxin b and alpha-cobratoxin are best described by triple exponential decay kinetics, whereas native alpha-bungarotoxin exhibits more than four lifetimes. The disulphide bonds of each toxin were reduced to facilitate carboxymethylation and amidocarboxymethylation. The two different toxin derivatives of all three neurotoxins displayed triple exponential decay kinetics and were completely denatured as evidenced by circular dichroism (random coil). The concentration (c) values of the three fluorescence decay times (time-resolved fluorescence spectroscopy (TRFS)) were dramatically different from those of the native toxins. Each neurotoxin, treated with different concentrations of guanidinium hydrochloride (GuHCl), was studied both by circular dichroism and TRFS. Disappearance of the beta-sheet secondary structural features with increasing concentrations of GuHCl was accompanied by a shift in the relative contribution (c value) of each fluorescence decay time (TRFS). It was found that certain disulphide residues confer added stability to the beta-sheet secondary structure of these neurotoxins and that the center of the beta-sheet is last to unfold. These titrations show that Trp can be used as a very localized probe of secondary structure.

Theoretical analysis of the structure of the peptide fasciculin and its docking to acetylcholinesterase.

The fasciculins are a family of closely related peptides that are isolated from the venom of mambas and exert their toxic action by inhibiting acetylcholinesterase (AChE). Fasciculins belong to the structural family of three-fingered toxins from Elapidae snake venoms, which include the alpha-neurotoxins that block the nicotinic acetylcholine receptor and the cardiotoxins that interact with cell membranes. The features unique to the known primary and tertiary structures of the fasciculin molecule were analyzed. Loop I contains an arginine at position 11, which is found only in the fasciculins and could form a pivotal anchoring point to AChE. Loop II contains five cationic residues near its tip, which are partly charge-compensated by anionic side chains in loop III. By contrast, the other three-fingered toxins show full charge compensation within loop II. The interaction of fasciculin with the recognition site on acetylcholinesterase was investigated by estimating a precollision orientation followed by determination of the buried surface area of the most probable complexes formed, the electrostatic field contours, and the detailed topography of the interaction surface. This approach has led to testable models for the orientation and site of bound fasciculin.

Genetic engineering of snake toxins. The functional site of Erabutoxin a, as delineated by site-directed mutagenesis, includes variant residues.

Using site-directed mutagenesis, we previously identified some residues that probably belong to the site by which Erabutoxin a (Ea), a sea snake toxin, recognizes the nicotinic acetylcholine receptor (AcChoR) (Pillet, L., Trémeau, O., Ducancel, F. Drevet, P., Zinn-Justin, S., Pinkasfeld, S., Boulain, J.-C., and Ménez, A. (1993) J. Biol. Chem. 268, 909-916). We have now studied the effect of mutating 26 new positions on the affinity of Ea for AcChoR. The mutations are F4A, N5V, H6A, Q7L, S9G, Q10A, P11N, Q12A, T13V, T14A, K15A, T16A, delta S18, E21A, Y25F, Q28A, S30A, T35A, I36R, P44V, T45A, V46A, K47A, P48Q, I50Q, and S53A. Binding affinity decreases upon mutation at Gln-7, Gln-10 and to a lesser extent at His-6, Ser-9 and Tyr-25 whereas it increases upon mutation at Ile-36. Other mutations have no effect on Ea affinity. In addition, new mutations of the previously explored Ser-8, Asp-31, Arg-33, and Glu-38 better explain the functional role of these residues in Ea. The previous and present mutational analysis suggest that the "functional" site of Ea covers a homogeneous surface of at least 680 A2, encompassing the three toxin loops, and includes both conserved and variant residues. The variable residues might contribute to the selectivity of Ea for some AcChoRs, including those from fish, the prey of sea snakes.

Tertiary structure of erabutoxin b in aqueous solution as elucidated by two-dimensional nuclear magnetic resonance.

The three-dimensional structure of erabutoxin b, a short-chain neurotoxic peptide purified from the venom of the sea snake Laticauda semifasciata, was determined in aqueous solution by two-dimensional proton nuclear magnetic resonance and simulated annealing-based calculations. On the basis of 883 assigned nuclear Overhauser effect (NOE) connectivities, 676 final distance constraints were derived and used together with 38 torsion angle (phi, chi 1) constraints, four distance constraints derived from disulfide bridges and 30 distance constraints derived from hydrogen bonds. A total of 14 converged structures were obtained from 50 runs of calculations. The atomic root-mean-square difference about the mean coordinate positions (excluding the residues 18 to 22) is 0.60 A for backbone atoms (N, C alpha and C'). The protein consists of a core region from which three finger-like loops emerge outwards. It includes a short, two-stranded antiparallel beta-sheet of residues 2 to 5 and 13 to 16, a three-stranded antiparallel beta-sheet involving residues 23 to 30, 35 to 41 and 50 to 56, and four disulfide bridges in the core region. Comparison with two crystal structures of erabutoxin b at 1.4 A and 1.7 A resolution indicated that the solution and the crystal structures were very similar, but less defined regions were observed at the localized region of the tip of the central loop and the outside of the third loop in solution. Other short-chain alpha-neurotoxins showed structural characteristics similar to those of erabutoxin b.

Engineering of protein epitopes: a single deletion in a snake toxin generates full binding capacity to a previously unrecognized antibody.

Structural features associated with the ability of a monoclonal antibody (mAb) to discriminate between protein variants are identified and engineered. The variants are the curaremimetic toxin alpha from Naja nigricollis and erabutoxin a or b from Laticauda semifasciata, which differ from each other by 16 substitutions and one insertion. The neutralizing mAb M alpha 1 recognizes with high affinity a topographical epitope on the surface of toxin alpha, but fails to recognize the erabutoxins although they possess most of the residues forming the presumed epitope. Examinations of the toxin alpha and erabutoxin 3-D structures and molecular dynamics simulations reveal several differences between the variants. In particular, the region involving the beta-turn 17-24 is organized differently. Analysis of the differences found in this region suggest that the insertion (or deletion) at position 18 of the variant amino acid sequences is particularly important in determining the differential cross-reactivity. To test this proposal, residue 18 was deleted in one erabutoxin using site-directed mutagenesis, and the biological properties of the resulting mutant were examined. We found that full antigenicity was restored in the previously unrecognized variant. The implications of this finding are discussed.

Dynamics of the active loop of snake toxins as probed by time-resolved polarized tryptophan fluorescence.

The local environment and dynamics of the single tryptophan residue in the respective active loops of cardiotoxin and alpha-neurotoxin from Naja nigricollis and of erabutoxin b from Laticauda semifasciata have been studied by steady-state and time-resolved polarized fluorescence and analyzed with distributions of decay times. Trp11 in loop I of cardiotoxin exhibits a very broad and complex distribution of fluorescence lifetimes at 20 degrees C. Despite its relatively external location in the toxin, the residue appears to be partly shielded from water and shows restricted but significant conformational fluctuations on the picosecond and nanosecond time scales. The thermal stability of cardiotoxin allowed a study of its static and dynamic fluorescence properties over a large range of temperatures. Interconversions in the intermediate nanosecond range lead to a thorough reorganization of the cardiotoxin fluorescence lifetime distribution with temperature. On the contrary, the fluorescence kinetics of Trp29 in loop II of the two neurotoxins is dominated by about 80% of a major decay time, which suggests that a nearly unique local conformation of the toxin is maintained over all time scales above the sub-nanosecond range. The fluorescence anisotropy decays show that the residue also has extremely limited rotational freedom down to the picosecond time scale. These findings are in good agreement with structural and dynamic information previously reported on the different toxins from NMR and X-ray crystallographic studies. The different dynamic properties around the tryptophan residue of the cardiotoxin and neurotoxin active loops can be analyzed within the frame of their different respective mechanisms of toxicity.

Three-dimensional crystal structure of recombinant erabutoxin a at 2.0 A resolution.

Recombinant erabutoxin a (Ea(r)) has been crystallized by vapour diffusion in hanging drops. The crystals belong to space group P2(1)2(1)2(1) with cell dimensions a = 55.8 A, b = 53.4 A, c = 40.8 A. Diffraction data have been recorded on a FAST detector up to 2.0 A. The atomic crystal structure of Ea(r) has been determined by initial refinement of the structure of the isotoxin erabutoxin b (Eb) the crystals of which were grown under identical conditions. The R-factor was 23% at 2.0 A resolution. The secondary and tertiary structures of Ea(r) are shown to be identical with that of wild-type Eb, within the experimental error.

Comparison of refolding patterns of erabutoxin b and cardiotoxin 3.10.2 from snake venom.

The refolding patterns of erabutoxin b (a neurotoxin) and cardiotoxin 3.10.2 (from Naja naja siamensis venom) have been studied by reducing both the proteins by treatment with reduced dithiothreitol followed by renaturation by treatment with oxidised dithiothreitol. Isoelectric focusing of the samples trapped at varying time intervals during renaturation of the proteins reveals formation of intermediates in the folding pathway with cardiotoxin 3.10.2. having fewer intermediates than erabutoxin b and faster rate of refolding (1 hr and 3 hr respectively).

Solution conformation of cobrotoxin: a nuclear magnetic resonance and hybrid distance geometry-dynamical simulated annealing study.

The solution conformation of cobrotoxin has been determined by using proton nuclear magnetic resonance spectroscopy. With the combination of various two-dimensional NMR techniques, the 1H-NMR spectrum of cobrotoxin was completely assigned (Yu et al., 1990). A set of 435 approximate interproton distance restraints was derived from nuclear Overhauser enhancement (NOE) measurements. These NOE constraints, in addition to the 29 dihedral angle constraints (from coupling constant measurements) and 26 hydrogen bonding restraints (from the pattern of short-range NOEs), form the basis of 3-D structure determination by the hybrid distance geometry-dynamical simulated annealing method. The 23 structures that were obtained satisfy the experimental restraints, display small deviation from idealized covalent geometry, and possess good nonbonded contacts. Analysis of converged structures indicated that there are two antiparallel beta sheets (double and triple stranded), duly confirming our earlier observations. These are well defined in terms of both atomic root mean square (RMS) differences and backbone torsional angles. The average backbone RMS deviation between the calculated structures and the mean structure, for the beta-sheet regions, is 0.92 A. The mean solution structure was compared with the X-ray crystal structure of erabutoxin b, the homologous protein. This yielded information that both structures resemble each other except at the exposed loop/surface regions, where the solution structure seems to possess more flexibility.

Genetic engineering of snake toxins. Role of invariant residues in the structural and functional properties of a curaremimetic toxin, as probed by site-directed mutagenesis.

To study the site by which erabutoxin a (Ea) from Laticauda semifasciata binds to the nicotinic acetylcholine receptor, we mutated most residues that are shared with other curaremimetic toxins and studied the structural and biological consequences of introduced mutations. By site-directed mutagenesis, we changed Ser-8 into Gly (EaS8G), Lys-27 into Glu (EaK27E), Trp-29 into Phe (EaW29F) and His (EaW29H), Asp-31 into His (EaD31H), Phe-32 into Leu (EaF32L), Arg-33 into Lys (EaR33K) and Glu (EaR33E), Gly-34 into Ser (EaG34S), Glu-38 into Gln (EaE38Q) and Lys (EaE38K), Gly-49 into Val (EaG49V), and Leu-52 into Ala (EaL52A). All mutants were homogeneous as judged by various analytical procedures. EaE38Q, EaG49V, and EaL52A bound the nicotinic acetylcholine receptor with apparent Kd values close to 10(-10) M, virtually identical to wild Ea. Therefore, Glu-38, Gly-49, and Leu-52 are not important elements in the expression of curaremimetic function in Ea. Mutations of Phe-32 and Gly-34 provoked a 7-fold affinity decrease, suggesting that these residues moderately contribute to function. The 176-fold affinity decrease due to mutation of Ser-8 may reflect some structural change that operates in the polypeptide chain of the mutant, as detected by circular dichroism. Decreases in affinity by a factor of 175, 67, 46, and 318 were seen upon mutations of Lys-27 into Glu, Trp-29 into Phe, Asp-31 into His, and Arg-33 into Glu, with no concomitant change in secondary structure. These residues appear to be important elements of the curaremimetic function of Ea. Thus, a picture of the contribution of conserved residues to the function of a curaremimetic toxin is proposed on the basis of experimental evidence.

1.9-A resolution structure of fasciculin 1, an anti-acetylcholinesterase toxin from green mamba snake venom.

The crystal structure of fasciculin 1, a potent acetylcholinesterase inhibitor from green mamba snake venom, has been solved by the multiple isomorphous replacement method complemented with anomalous scattering and subsequently refined at 1.9-A resolution. The overall structure of fasciculin is similar to those of the short alpha-neurotoxins and cardiotoxins, with a dense core rich in disulfide bridges and three long loops disposed as the central fingers of a hand. A comparison of these three prototypic toxin types shows that fasciculin 1 has structural features that are intermediate between those of the other two molecules. Its core region, which can be defined as a continuous stretch of conserved residues, is very similar to that of erabutoxin b, whereas the orientation of its long loops resembles that of cardiotoxin VII4. This result introduces a new element in the study of phylogenetic relationships of snake toxins and suggests that, after divergency from an ancestral gene, convergent evolution may have played an important factor in the evolution of these proteins. In fasciculin 1, several arginine and lysine residues are well ordered and relatively exposed to the solvent medium and may play a role in the binding to the peripheral site of acetylcholinesterases.