Marshall_1994_Toxicon_32_1433

The scorpion venom Leiurus quinquestriatus hebreus was fractionated by chromatography in order to isolate toxins that affected binding of radiolabelled dendrotoxin to K+ channel proteins on synaptosomal membranes and that facilitated acetylcholine release in chick biventer cervicis nerve-muscle preparations. In addition to the previously characterized charybdotoxin, three toxins were isolated: 14-2, 15-1 and 18-2. Toxin 14-2 has a blocked N-terminus and because of low quantities, it has not been sequenced; 15-1 is a newly sequenced toxin of 36 residues with some overall homology to charybdotoxin and noxiustoxin; 18-2 is identical to charybdotoxin-2. The apparent Ki against dendrotoxin binding were: charybdotoxin, 3.8 nM; 14-2, 150 nM; 15-1, 50 nM; and 18-2, 0.25 nM. Toxin 14-2 (75 nM-1.5 microM) had a presynaptic facilitatory effect on neuromuscular preparations. Toxin 15-1 augmented responses to direct muscle stimulation, probably because it blocked Ca(2+)-activated K+ currents in muscle fibres. Toxin 18-2 (charybdotoxin-2) had a potent presynaptic facilitatory action, with less effect on direct muscle stimulation. This contrasts with the relatively weak neuromuscular effects of the highly homologous charybdotoxin. On a Ca(2+)-activated K+ current in mouse motor nerve endings, charybdotoxin and toxin 18-2 produced maximal block at around 100 nM, whereas 15-1 was inactive at 300 nM. Charybdotoxin can increase quantal content, but this is more likely to result from block of voltage-dependent K+ channels than Ca(2+)-activated channels: the increase in transmitter release occurred in conditions in which little IKCa would be present; higher concentration of charybdotoxin and longer exposure times were required to increase transmitter release than those needed to block IKCa, and the facilitatory effects of charybdotoxin and toxin 18-2 correlated more with their effects on dendrotoxin binding than on block of IKCa.