Whereas the presynaptic action of Crotalus durissus terrificus venom is well-established, Bothrops venoms have historically been considered to have only postsynaptic and muscular effects. However, some studies have also suggested a presynaptic action for these venoms. In this work, we used chick biventer cervicis preparations to compare the presynaptic actions of two Bothrops venoms (B. insularis and B. neuwiedi) with that of C. d. terrificus venom. At 10 mug/ml, all venoms produced irreversible blockade of the twitch tension responses, with no reduction in acetylcholine (ACh)-induced contractures and only a slight decrease in potassium induced-contractures. The times (in min) required to produce 50% neuromuscular blockade (C. d. terrificus: 16.3 +/- 0.7, n = 8; B. insularis: 30.0 +/- 1.9, n = 5; B. neuwiedi: 42.0 +/- 2.0, n = 8 mean +/- SEM) were significantly different among the venoms (p < 0.01). Lowering the temperature at which the experiments were done (from 37 to 24 degreesC) prevented neuromuscular blockade by the three venoms, indicating that enzyme activity may be involved in this response. At concentrations capable of causing complete neuromuscular blockade, creatine kinase release remained close to levels seen ill control preparations incubated with Krebs solution alone (500-1200 IU/l). Commercial crotalic antivenom, but not bothropic antivenom, protected against the neuromuscular blockade caused by B. insularis and B. neuwiedi venoms. These observations indicate that bothropic venoms may contain components which act presynaptically in a manner similar to C. d. terrificus venom, and that at low venom concentrations a direct action oil skeletal muscle does not contribute to this presynaptic neurotoxicity. (C) 2003 Elsevier Ltd. All rights reserved.436633638