Platelet-activating factor (PAF), a biologically active lipid (1-O-alkyl-2-acetyl-sn-glycero-3-phosphoholine), is identified in different regions of brain, including hippocampus. Specific PAF-activated receptors (PAFRs) are expressed in corresponding brain areas. PAF has been proposed to be a retrograde messenger of long-term potentiation (LTP): the antagonist of PAFRs, ginkgolide B (or BN52021) prevents induction of LTP. Recently it has been found that ginkgolide B is also an efficient blocker of the glycine receptor (GlyR) operated chloride channels (IC(50)=270+/-10 nM in hippocampal pyramidal neurons). The question is as follows: is the alteration of LTP by BN52021 due to the PAF antagonism or to the inhibition of glycine-gated chloride channels? We have studied the effects of ginkgolides B and J on LTP induced in the CA1 area of rat hippocampus. Ginkgolide J which is the weakest blocker of PAFR (IC(50)=54 microM, as compared to IC(50)=2.5 microM for ginkgolide B) inhibits GlyR-operated channels with IC(50)=2.0 microM. This assures a convenient concentration window which allows to inhibit GlyR-operated channels without affecting PAFRs. An amount of 5 microM of ginkgolide J did not prevent the induction of LTP, while ginkgolide B (5 microM) completely inhibited this phenomenon. The effect of ginkgolide B on LTP did not alter considerably if GlyRs were blocked by strychnine (2 microM). Strychnine itself had no significant effect on the induction of LTP. Both ginkgolides and strychnine significantly facilitated short-term potentiation (STP). Our data support a hypothesis according to which ginkgolides affect LTP by inhibiting PAFRs.