GENETIC SUBSTRATES OF MEMORY

[The understanding of biological substrates of learning and memory has been greatly facilitated by the application of the techniques of molecular biology and genetics. Studies done in the 1960s argued that longer-term memory processes required protein synthesis (i.e., required certain genes to increase or alter their expression of proteins). Technology developed in the 1990s has made it possible to "knock out" or "knock in" specific genes experimentally, as well as to make use of naturally occurring mutants. This work has involved both invertebrates (e.g., the fruit fly and C. elegans) and vertebrates, mostly mice. Learning is a particularly interesting aspect of behavior to manipulate genetically since memory involves changes in behavior as a result of experiences in the environment. As with language in humans, the ability to learn and the mechanisms of memory storage are largely determined genetically, but what is learned depends almost entirely on experience.

The first article in this series focuses on theA mygdalaand fear conditioning (see NEURAL SUBSTRATES OF CLASSICAL CONDITIONING: FEAR CONDITIONING, FREEZING). Induction of fear conditioning alters the expression of immediate early genes in the amygdala, and blocking gene expression in the amygdala prevents long-term fear memory. Using viral vector gene-transfer technology, increasing the expression of the transcriptional activator CREB (Cyclic AMP Response Element Binding) in the amygdala enhances fear memory, whereas decreasing CREB expression in the amygdala impairs fear memory.

Genetic approaches have provided much new information concerning theC erebellarsubstrates of classical eye-blink condition (see NEURAL SUBSTRATES OF CLASSICAL CONDITIONING: DISCRETE BEHAVIORAL RESPONSES). The Purkinje cell degeneration (pcd) mutant mouse loses all Purkinje neurons in the cerebellar cortex within four weeks after birth. These animals are able to learn eyeblink conditioning, but the learning is slower and to a lower degree than normal, and it slows extinction. Lesions of the interpositus nucleus in the pcd mouse completely prevent learning. So both cerebellar cortex and interpositus are involved but the basic association seems to be formed in the interpositus. Mutant and knock-out mice that are deficient in cerebellar long-term depression (LTD) are also much impaired in learning and in adaptive timing of the CR, indicating an important modulatory role of the cerebellar cortex.

The third article, on theH ippocampus, is written by a pioneer in this new field, Alcino Silva. The first studies to use genetics to manipulate plasticity and learning in mammals (mice) targeted kinases (calcium-calmodulin dependent kinase II) and the tyrosine kinase Fyn. These studies showed that disruption of hippocampal-synaptic plasticity (i.e., long-term potentiation, LTP) resulted in hippocampal-dependentlearning deficits. Recent approaches involve targeted, selective, temporally controlled expression of normal or altered genes, an extraordinarily promising technology.]