The role of voltage-gated chloride channels in type II pyrethroid insecticide poisoning

Pyrethroids act on mammalian sodium channels, but we have previously shown that low concentrations of the type II pyrethroid deltamethrin also decreas e the open channel probability (P-o) of voltage-gated chloride channels. Th is effect would be expected to amplify the sodium channel-mediated signs of poisoning produced by pyrethroids. In the present study we evaluated poten tial chloride channel agonists in vitro, and then tested the most effective of these on pyrethroid-poisoned rats to determine the practical significan ce of chloride channel effects in vivo. Patch clamp experiments showed that , for voltage-gated maxi chloride channels in excised, inside-out patches f rom mouse N1E 115 neuroblastoma cells, ivermectin (10(-7) M) and pentobarbi tone (10(-6) M) significantly increased open channel probability (p less th an or equal to 0.01 and p less than or equal to 0.02, respectively), wherea s phenobarbitone, hexobarbitone, mephobarbitone, thiopentone, and barbituri c acid did not. This suggested that, if chloride channels were important in vivo, ivermectin and pentobarbitone should antagonize type II pyrethroid p oisoning and phenobarbitone should not. Male F344 rats were then pretreated with ivermectin (4 mg/kg iv), equisedative doses of either pentobarbitone (15 mg/kg ip) or phenobarbitone (45 mg/kg ip), or solvent controls. This wa s followed by deltamethrin (1.5 or 2 mg/kg iv) or the type I pyrethroid cis methrin (4 mg/kg iv). Ivermectin produced a marked fall in deltamethrin-ind uced salivation (p less than or equal to 0.05) and also (in anesthetized ra ts) in repetitive electromyogram discharge and muscle twitch (p less than o r equal to 0.01 and p less than or equal to 0.05, respectively). Pentobarbi tone significantly reduced the motor signs score due to deltamethrin (p les s than or equal to 0.01). Ivermectin therefore protected against the periph eral signs of deltamethrin poisoning and pentobarbitone protected against t he central signs. As expected phenobarbitone had no protective effects. The motor signs produced by the type I pyrethroid cismethrin (which does not a ct on chloride channels) were not diminished by either barbiturate. The per ipheral benzodiazepine receptor blocker PK11195 did not diminish the protec tive action of ivermectin on the muscle twitch (p less than or equal to 0.0 5), although it partially reversed the block of salivation (p less than or equal to 0.05). These results support the hypothesis that the voltage-depen dent chloride channel is a toxicologically significant additional site of a ction for deltamethrin and that the use of chloride channel agonists can pr ovide a rationale for a novel and effective therapy against type II pyrethr oid poisoning. (C) 2000 Academic Press.