Synaptic plasticity and learning and memory: LTP and beyond.

Long-term potentiation (LTP) of synaptic activity is by far the most popular and widely researched model of synaptic plastic changes that might occur during learning. Numerous recent reports, however, have not found a correlation between the inducibility of LTP in the hippocampus and the ability of animals to learn hippocampus-dependent tasks. For example, some experiments with gene deletion (knockout) mice strains have shown that in some strains LTP is not inducible in the dentate gyrus, in area CA3, or CA1, but the animals are still able to learn spatial tasks. This apparent mismatch has rejuvenated the discussion concerning whether LTP is a good model for mechanisms that underlie memory formation in the nervous system. This review analyzes the conditions under which LTP is induced or learning takes place and suggests reasons for the mismatches that can occur and what we can learn from them. High-frequency stimulation protocols and in vitro assays cannot be seen to resemble natural firing patterns or conditions found in the brain. More physiological experimental conditions, especially in vivo recording in awake animals, could lead the way to the development of improved models of learning mechanisms that better correlate with learning abilities of animals.