Chemical Synthesis of a Functional Fluorescent-Tagged α-Bungarotoxin.

α-bungarotoxin is a large, 74 amino acid toxin containing five disulphide bridges, initially identified in the venom of Bungarus multicinctus snake. Like most large toxins, chemical synthesis of α-bungarotoxin is challenging, explaining why all previous reports use purified or recombinant α-bungarotoxin. However, only chemical synthesis allows easy insertion of non-natural amino acids or new chemical functionalities. Herein, we describe a procedure for the chemical synthesis of a fluorescent-tagged α-bungarotoxin. The full-length peptide was designed to include an alkyne function at the amino-terminus through the addition of a pentynoic acid linker. Chemical synthesis of α-bungarotoxin requires hydrazide-based coupling of three peptide fragments in successive steps. After completion of the oxidative folding, an azide-modified Cy5 fluorophore was coupled by click chemistry onto the toxin. Next, we determined the efficacy of the fluorescent-tagged α-bungarotoxin to block acetylcholine (ACh)-mediated currents in response to muscle nicotinic receptor activation in TE671 cells. Using automated patch-clamp recordings, we demonstrate that fluorescent synthetic α-bungarotoxin has the expected nanomolar affinity for the nicotinic receptor. The blocking effect of fluorescent α-bungarotoxin could be displaced by incubation with a 20-mer peptide mimicking the α-bungarotoxin binding site. In addition, TE671 cells could be labelled with fluorescent toxin, as witnessed by confocal microscopy, and this labelling was partially displaced by the 20-mer competitive peptide. We thus demonstrate that synthetic fluorescent-tagged α-bungarotoxin preserves excellent properties for binding onto muscle nicotinic receptors.

NMR structure of alpha-bungarotoxin free and bound to a mimotope of the nicotinic acetylcholine receptor.

A combinatorial library approach was used to produce synthetic peptides mimicking the snake neurotoxin binding site of nicotinic receptors. Among the sequences, which inhibited binding of alpha-bungarotoxin to muscle and neuronal nicotinic receptors, HRYYESSLPWYPD, a 14-amino acid peptide with considerably higher toxin-binding affinity than the other synthesized peptides, was selected, and the structure of its complex with the toxin was analyzed by NMR. Comparison of the solution structure of the free toxin and its complex with this peptide indicated that complex formation induced extensive conformational rearrangements mainly at finger II and the carboxy terminus of the protein. The peptidyl residues P10 and Y4 seemed to be critical for peptide folding and complex stability, respectively. The latter residue of the peptide strongly interacted with the protein by entering a small pocket delimited by D30, C33, S34, R36, and V39 toxin side chains.

Mimotopes of the nicotinic receptor binding site selected by a combinatorial peptide library.

Peptide libraries allow selecting new molecules, defined as mimotopes, which are able to mimic the structural and functional features of a native protein. This technology can be applied for the development of new reagents, which can interfere with the action of specific ligands on their target receptors. In the present study we used a combinatorial library approach to produce synthetic peptides mimicking the snake neurotoxin binding site of nicotinic receptors. On the basis of amino acid sequence comparison of different alpha-bungarotoxin binding receptors, we designed a 14 amino acid combinatorial synthetic peptide library with five invariant, four partially variant, and five totally variant positions. Peptides were synthesized using SPOT synthesis on cellulose membranes, and binding sequences were selected using biotinylated alpha-bungarotoxin. Each variant position was systematically identified, and all possible combinations of the best reacting amino acids in each variant position were tested. The best reactive sequences were identified, produced in soluble form, and tested in BIACORE to compare their kinetic constants. We identified several different peptides that can inhibit the binding of alpha-bungarotoxin to both muscle and neuronal nicotinic receptors. Peptide mimotopes have a toxin-binding affinity that is considerably higher than peptides reproducing native receptor sequences.

Synthetic peptides of neurotoxins and rabies virus glycoprotein behave as antagonists in a functional assay for the acetylcholine receptor.

Peptides of portions of loop 2 (the "toxic" loop) of snake venom curare-mimetic neurotoxins (alpha-bungarotoxin and king cobra toxin b) and of a structurally similar region of the rabies virus glycoprotein were synthesized. The effect of the peptides on carbachol-induced 22Na+ flux into BC3H-1 cells, which contain nicotinic acetylcholine receptors on their surfaces, was measured. Both the neurotoxin and glycoprotein peptides inhibited ion transport with IC50 values of 10(-4) M to 7 x 10(-7) M. The most effective peptides correspond to neurotoxin loop 2 and inhibited 22Na+ flux in the micromolar range comparable to the competitive antagonist d-tubocurarine. These findings show that neurotoxin loop 2 and the corresponding rabies virus glycoprotein segment interact with the agonist binding site of teh acetylcholine receptor and that short synthetic peptides representing portions of larger molecules by themselves can exert a biological effect on a large macromolecular complex like the acetylcholine receptor.

Chemical engineering of a three-fingered toxin with anti-alpha7 neuronal acetylcholine receptor activity.

Though it possesses four disulfide bonds the three-fingered fold is amenable to chemical synthesis, using a Fmoc-based method. Thus, we synthesized a three-fingered curaremimetic toxin from snake with high yield and showed that the synthetic and native toxins have the same structural and biological properties. Both were characterized by the same 2D NMR spectra, identical high binding affinity (K(d) = 22 +/- 5 pM) for the muscular acetylcholine receptor (AChR) and identical low affinity (K(d) = 2.0 +/- 0.4 microM) for alpha7 neuronal AchR. Then, we engineered an additional loop cyclized by a fifth disulfide bond at the tip of the central finger. This loop is normally present in longer snake toxins that bind with high affinity (K(d) = 1-5 nM) to alpha7 neuronal AchR. Not only did the chimera toxin still bind with the same high affinity to the muscular AchR but also it displayed a 20-fold higher affinity (K(d) = 100 nM) for the neuronal alpha7 AchR, as compared with the parental short-chain toxin. This result demonstrates that the engineered loop contributes, at least in part, to the high affinity of long-chain toxins for alpha7 neuronal receptors. That three-fingered proteins with four or five disulfide bonds are amenable to chemical synthesis opens new perspectives for engineering new activities on this fold.