Long-Term Potentiation (LTP)

Long-term potentiation is a durable increase in the strength of a synapse produced by a short burst of high-frequency or theta-patterned stimulation of its inputs. The best-characterised form, at Schaffer collateral synapses onto hippocampal CA1 neurons, needs the NMDA receptor to act as a coincidence detector: it opens only when glutamate binds and the postsynaptic cell is already depolarised enough to expel the magnesium block. Incoming calcium activates calcium-calmodulin-dependent kinase II, which drives AMPA receptors into the postsynaptic membrane. A late phase, dependent on new transcription and protein synthesis, extends the effect for many hours. The readout is usually field excitatory postsynaptic potential slope as a percentage of baseline.

The phenomenon was first described in 1973 by Bliss and Lomo in the rabbit hippocampus, and the acute slice preparation has been the workhorse since. In humans there is no direct equivalent; the closest analogues are paired associative transcranial magnetic stimulation and repetitive sensory tetanisation protocols, both indirect and noisy.

Because this is the leading cellular account of encoding, a compound that enhances it has a mechanistically coherent story. That is where the entitlement ends: potentiation measured in a slice is a change in a field potential, not a memory, and the step between them requires behavioural evidence in an intact animal.

Two failures recur. Bath concentrations applied to a slice are frequently far above any unbound brain concentration a systemic dose could produce, so the effect is real and irrelevant. And more potentiation is not automatically better, since saturating plasticity degrades the signal-to-noise of later encoding, and compounds that enlarge potentiation have impaired learning in the same animals.