The site of expression of NMDA receptor-dependent LTP: New fuel for an old fire

Few phenomena in neuroscience have attracted as much attention as long-term potentiation (LTP) of synaptic transmission in the mammalian cortex (Bliss and Collingridge, 1993). This is explained by the probable involvement of LTP both in the formation and storage of memories and in neuronal injury. Much of the interest in LTP also centers around a long-running debate about the nature of the enhancement in synaptic transmission: does LTP involve an increase in transmitter release or an enhanced postsynaptic response to neurotransmitter, or both? This review describes several recent attempts to resolve the locus of LTP, which also shed light on the basic mechanisms of synaptic transmission in the CNS. At the excitatory synapse between Schaffer collaterals and pyramidal cells in the CA1 region of the hippocampus, LTP is triggered by synchronous high frequency stimulation of many presynaptic fibers, sufficient to depolarize the postsynaptic cells. It can also be induced with low frequency stimulation of individual presynaptic fibers, as long as this is coupled with postsynaptic depolarization, which can be elicited by current injection through a microelectrode or by strong stimulation of an independent input converging on the same postsynaptic cells. There is general agreement that the induction of LTP at this and many other synapses requires Ca 2+ entry into the postsynaptic dendritic spine via the N-methyl-o-aspartate (NMDA) subtype of glutamate receptors. At the resting potential, NMDA receptor-linked channels are blocked by extracellular Mg 2+, but membrane depolarization relieves this block and allows Ca ~+ influx. The NMDA receptors thus function as coincidence detectors, first postulated by Donald Hebb to allow the activity-dependent storage of information in neuronal networks. Events downstream from this, however, are less clear: protein kinases are involved, and possibly retrograde messengers transmitting a signal from the postsynaptic spine to the presynaptic terminal, but the locus of the resulting persistent modification is controversial (Bliss and Collingridge, 1993; Larkman and Jack, 1995; Nicoll and Malenka, 1995).