Duration of NMDA-dependent synaptic potentiation in piriform cortex in vivo is increased after epileptiform bursting.

Stimulation of afferent fibers with current pulse trains has been reported to induce long-term potentiation (LTP) in piriform cortex in vitro but not in vivo. LTP has been observed in vivo only when trains are paired with behavioral reinforcement and as a consequence of kindled epileptogenesis. This study was undertaken in the urethan-anesthetized rat to determine if the reported failures to observe pulse-train evoked LTP in vivo may be related to a lesser persistence rather than lack of occurrence, if disinhibition might facilitate induction, and to examine the nature of the relationship between seizure activity and LTP. Stimulation of afferent fibers in the lateral olfactory tract with theta-burst trains under control conditions potentiated the monosynaptic field excitatory postsynaptic potential (EPSP) by approximately the same extent (20.3 +/- 2%; n = 12) as reported for the slice. However, in contrast to the slice, potentiation in vivo decayed to a low level within 1-2 h after induction (70% loss in 1.5 h, on average). The N-methyl--aspartate (NMDA)-receptor antagonists -APV and MK-801 blocked the induction of this decremental potentiation. Pharmacological reduction of gamma-aminobutyric acid-mediated inhibition at the recording site did not increase the duration of potentiation. In contrast, theta-burst stimulation applied after recovery from a period of epileptiform bursting induced stable NMDA-dependent potentiation. Mean increase in the population EPSP was approximately the same as under control conditions (21 +/- 2%; n = 6), but in five of six experiments there was little or no decay in potentiation for the duration of the monitoring period (</=6 h). It is concluded that seizure activity has an enabling action on the induction of persistent synaptic potentiation by stimulus trains that bypasses the need for behavioral reinforcement.