Long-term potentiation: persisting problems and recent results.

In this paper we discuss recent experimental results pertinent to three unresolved issues regarding the long-term potentiation (LTP) effect: the nature of its enduring substrates, the biochemical mechanisms that produce it, and its potential role in memory. LTP appears to be triggered by a postsynaptic influx of calcium and is associated with alterations in the shape of dendritic spines and probably the formation of new synapses. We discuss the possibility that morphological reorganization also modifies membrane surface chemistry of synaptic elements. Evidence is presented that LTP is not associated with changes in presynaptic calcium currents. Activation of protein kinase C is shown to be insufficient for the induction of LTP, although it may play a modulatory role. The hypothesis that activation of a calcium-sensitive protease (calpain) is pivotal to the establishment of LTP is supported by experiments showing that a calpain inhibitor, leupeptin, blocks LTP. Furthermore, activation of NMDA receptors, an event implicated in LTP induction, is accompanied by calcium-sensitive proteolysis of spectrin, a major dendritic cytoskeletal protein. The finding that stimulation patterns designed to mimic naturally-occurring cell discharge patterns are highly effective for LTP induction greatly strengthens the hypothesis that LTP actually occurs during the encoding of information in cortical systems. Potential contributions of LTP to learning are explored using computer simulations of a simple cortical network.