The involvement of presynaptic mechanisms in the expression of long-term potentiation (LTP), an enhancement of synaptic transmission suggested to take part in learning and memory in the mammalian brain, has not been fully clarified. Although evidence for enhanced vesicle cycling has been reported, it is unknown whether presynaptic terminal excitability could change as has been observed in invertebrate synapses. To address this question, we performed extracellular focal recordings in cerebellar slices. The extracellular current consisted of a pre- (P(1)/N(1)) and postsynaptic (N(2)/SN) component. In ~50% of cases, N(1) could be subdivided into N(1a) and N(1b). Whereas N(1a) was part of the fiber volley (P(1)/N(1a)), N(1b) corresponded to a Ca(2+)-dependent component accounting for 40-50% of N(1), which could be isolated from individual mossy fiber terminals visualized with fast tetramethylindocarbocyanine perchlorate (DiI). The postsynaptic response, given its timing and sensitivity to glutamate receptor antagonists [N(2) was blocked by 10 microM [1,2,3,4-tetrahydro-6-nitro-2,3-dioxo-benzo[f]quinoxaline-7-sulfonamide disodium (NBQX) and SN by 100 microM APV +50 microM 7-Cl-kyn], corresponded to granule cell excitation. N(2) and SN could be reduced by 1) Ca(2+) channel blockers, 2) decreasing the Ca(2+) to Mg(2+) ratio, 3) paired-pulse stimulation, and 4) adenosine receptor activation. However, only the first two manipulations, which modify Ca(2+) influx, were associated with N(1) (or N(1b)) reduction. LTP was induced by theta-burst mossy fiber stimulation (8 trains of 10 impulses at 100 Hz separated by 150-ms pauses). Interestingly, during LTP, both N(1) (or N(1b)) and N(2)/SN persistently increased, whereas P(1) (or P(1)/N(1a)) did not change. Average changes were N(1) = 38.1 +/- 31.9, N(2) = 49.6 +/- 48.8, and SN = 59.1 +/- 35.5%. The presynaptic changes were not observed when LTP was prevented by synaptic inhibition, by N-methyl-D-aspartate and metabotropic glutamate receptor blockage, or by protein kinase C blockage. Moreover, the presynaptic changes were sensitive to Ca(2+) channel blockers (1 mM Ni(2+) and 5 microM omega-CTx-MVIIC) and occluded by K(+) channel blockers (1 mM tetraethylammmonium). Thus regulation of presynaptic terminal excitability may take part in LTP expression at a central mammalian synapse.