Maintenance

One of the most striking properties of memory is its extreme duration; people have memories dating as far back as twenty years or more. Similarly, one of the features of long-term potentiation or LTP (at least in certain brain structures) is its duration; it has been a major challenge for neurobiology to provide a cellular mechanism for long-lasting changes in synaptic transmission that outlast protein turnover. The discovery of the critical role of the NMDA receptors in triggering LTP indicated the existence of distinct stages in the establishment of LTP, which are generally defined as induction and maintenance (Bliss and Lynch, 1988). The former refers to events activated during the high-frequency stimulation, whereas the latter designates the processes that are responsible for the changes in synaptic transmission that underlie LTP and their long-term maintenance. In general, three types of mechanisms have been proposed as maintenance mechanisms: those producing increased transmitter release; those resulting in changes in spine electrical properties; and those implicated in regulating receptor properties (Landfield and Deadwyler, 1988). Since the 1980s evidence against and for each of these mechanisms has been obtained, and the most salient features of the arguments are presented in this entry.

Presynaptic Mechanisms

An increase in transmitter release could account for an increase in synaptic transmission; it is generally admitted, for instance, that short-term potentiation is due to increased transmitter release (Zucker, 1989). Biochemical and electrophysiological evidence has been obtained in support of the hypothesis that increased transmitter release accounts for LTP. Glutamate is likely to be the neurotransmitter used by synapses exhibiting LTP in various hippocampal pathways, and increased glutamate levels have been found after LTP induction in perfusates obtained with push-pull cannulas implanted in the dentate gyrus (Dolphin, Errington, and Bliss, 1982). Increased glutamate release was also found after LTP induction in hippocampal slices and in synaptosomes prepared from hippocampus of rats in which LTP had been induced. Since the induction of LTP involves the postsynaptic activation of NMDA receptors, a retrograde signal has been postulated that relays the postsynaptic activation to the presynaptic terminal. Researchers initially proposed arachidonic acid to be such a retrograde signal, since activation of NMDA receptors stimulates phospholipase A2 (PLA2), and PLA2 inhibitors prevent LTP formation (Bliss and Lynch, 1988). Arachidonic acid could also provide a link with the presynaptic machinery involved in the regulation of transmitter release because it activates protein kinases that have been shown to participate in phosphorylation reactions important for transmitter release. Researchers later proposed that nitric oxide as well as carbon monoxide were retrograde signals, although the evidence for this conclusion remains highly controversial (Hawkins, Son, and Arancio, 1998; Malenka and Nicoll, 1999). While this mechanism could conceivably provide a satisfactory explanation for a short-lasting enhancement of transmitter release, in its present version it does not account for a long-lasting increase.

Quantal analysis of synaptic transmission (Martin, 1966) is an approach that, in principle, could provide an unambiguous answer to the question of the locus of the changes underlying LTP. It uses the intrinsic variability of transmitter release and statistical methods to determine the parameters generally thought to govern synaptic transmission; in effect, the probability of release, the number of release sites, and the elementary size of a postsynaptic response elicited by a quantum of neurotransmitter. Three groups have reported results obtained with quantal analysis in field CA1 before and after LTP induction in hippocampal slices. Two groups observed an increase in quantal content (Malinow and Tsien, 1990; Bekkers and Stevens, 1990), whereas one group observed an increase in quantal size (Foster and McNaughton, 1991). In addition to this ambiguity in the results, quantal analysis requires a number of assumptions that might not be satisfied at hippocampal synapses. In particular, several studies have indicated the existence of synapses with functional NMDA receptors but few, if any, functional AMPA receptors (the socalled silent synapses). Transformation of silent synapses into active synapses by LTP-inducing stimulation could account for the inconsistencies in results from quantal analysis (Malenka and Nicoll, 1999).

Finally, increased transmitter release could be accounted for by an increase in the number of synapses. Anatomical studies using quantitative electron microscopy have provided evidence that the number of certain categories of synapses, in particular sessile synapses, is increased in field CA1 following LTP induction (Lee, Schottler, Oliver, and Lynch, 1980; Chang and Greenough, 1984). Although an increased number of synapses could explain the duration of LTP and account for its maintenance, it is not yet clear whether the increase in sessile synapses represents the formation of new synapses or the transformation of existing synapses by shape modifications (Toni et al., 1999).

Changes in Spine Electrical Properties

Since their discovery, there has been much speculation concerning the roles of dendritic spines in synaptic transmission and the possibilities that synaptic plasticity could be due to alterations in their electrical properties (Rall, 1978; Coss and Perkel, 1985). Because of their shape—large heads relative to long, narrow necks—they are considered to be high-resistance elements coupling voltage changes at the synapses with voltage changes in dendritic shafts. Based upon calculation and computer simulation, researchers have proposed that decreased neck resistance could account for increased synaptic responses (Wilson, 1988), which could affect the AMPA receptors more specifically than the NMDA receptors (Lynch and Baudry, 1991). Such a decrease in spine resistance could be due to an increase in neck dimension, and anatomical evidence indicates that LTP is accompanied by an increase in spines with wider and shorter necks. The recent discoveries of active elements (channels or pumps) in dendritic spines provide alternative means of modifying their electrical properties (Johnston, Magee, Colbert, and Cristie, 1996), although there is no evidence that these elements are modified following LTP induction.

Changes in Receptor Properties

One of the most important characteristics of LTP is that it is expressed by an increase in the component of the synaptic response resulting from the activation of the AMPA receptors with little changes in the component generated by the activation of NMDA receptors. Thus an alteration of the properties of the AMPA receptors has been proposed as a potential maintenance mechanism. Several studies have shown that both the ligand binding and the ionic conductance properties of the AMPA receptors are affected by a variety of manipulations. In particular, changes in the lipid environment of the AMPA receptors modify its affinity for agonists, an observation that could account for the involvement of PLA2 in LTP. Studies conducted in the 1990s on the molecular biology of the AMPA receptors have provided exciting results concerning the nature and properties of these receptors. AMPA receptors belong to a family of receptors encoded by at least four related genes, each existing in two closely similar versions generated by alternative splicing of the genes, and designated flip and flop. Researchers have proposed that AMPA receptors exist as oligomers composed of a number of possible arrangements of flip and flop elements, with the flip elements providing larger currents than the flop (Sommer et al., 1990). Twenty-first-century research has suggested that LTP could result from a change in the composition of receptor subunits or in changes in the rates of receptor insertion or internalization in the synaptic membrane from a subsynaptic pool of receptors (Malinow, Mainen, and Hayashi, 2000). Another dimension related to the stabilization of the changes underlying LTP maintenance has recently been included to the search for the mechanisms of LTP. It is now apparent that LTP stabilization involves adhesion molecules and that there is a time window following LTP induction during which processes leading to LTP consolidation can be reversed, thus producing depotentiation (Lynch, 1998). There has been considerable discussion to determine whether depotentiation represents the same process as long-term depression of synaptic transmission. Although both processes share a number of features, there is not yet a general consensus as to the mechanisms and the significance of these forms of synaptic plasticity.

Conclusion

Although several mechanisms discussed in this entry could account for some characteristics of LTP, most of them run into difficulties when they have to explain its stability. This feature is almost certain to eliminate most mechanisms, however attractive they might appear, that are based solely on conformational or posttranslational changes in proteins. In particular, several speculative mechanisms of LTP maintenance have been based on the idea of biochemical switches, constituted of biochemical reactions involving positive feedback. The most popular of these biochemical switches is derived from the properties of autophosphorylation of calcium/calmodulin kinase type II, which is a very prominent protein in postsynaptic densities and especially in hippocampus (Crick, 1984; Lisman, 1994). The mechanism involved in LTP maintenance therefore will probably require structural modifications that can confer lasting stability to biochemical modifications responsible for changes in synaptic transmission. The coupling of these processes with those responsible for the maintenance of the adhesive properties of synaptic contacts represents a key feature in the evolution of adaptive properties of brain neuronal networks. Finally, the various mechanisms described above underline the richness of mechanisms that have evolved to produce synaptic plasticity and that are likely to contribute to the multiple forms of potentiation that are being uncovered at a variety of central synapses.

See also:GLUTAMATE RECEPTORS AND THEIR CHARACTERIZATION; GUIDE TO THE ANATOMY OF THE BRAIN: NEURON; MORPHOLOGICAL BASIS OF LEARNING AND MEMORY: VERTEBRATES

Bibliography

Bekkers, J. M., and Stevens, C. F. (1990). Presynaptic mechanism for long-term potentiation in the hippocampus. Nature 346, 724-729.

Bliss, T. V. P., and Lynch, M. A. (1988). Long-term potentiation: Mechanisms and properties. In P. W. Landfield and S. A. Deadwyler, eds., Long-term potentiation: From biophysics to behavior. New York: Liss.

Chang, F., and Greenough, W. T. (1984). Transient and enduring morphological correlates of synaptic activity and efficacy change in the rat hippocampal slice. Brain Research 309, 35-46.

Coss, R. G., and Perkel, D. H. (1985). The function of dendritic spines: A review of theoretical issues. Behavior, Neurology, and Biology 44, 151-185.

Crick, F. (1984). Memory and molecular turn-over. Nature 312, 101.

Dolphin, A. C., Errington, M. L., and Bliss, T. V. P. (1982). Long-term potentiation of the perforant path in vivo is associated with increased glutamate release. Nature 297, 496-498.

Foster, T. C., and McNaughton, B. L. (1991). Long-term synaptic enhancement in CA1 is due to increased quantal size, not quantal content. Hippocampus 1, 79-91.

Hawkins, R. D., Son, H., and Arancio, O. (1998). Nitric oxide as a retrograde messenger during long-term potentiation in hippocampus. Progress in Brain Research 118, 155-172.

Johnston, D., Magee, J. C., Colbert, C. M., and Cristie, B. R. (1996). Active properties of neuronal dendrites. Annual Review of Neuroscience 19, 165-186.

Landfield, P. W., and Deadwyler, S. A., eds. (1988). Long-term potentiation: From biophysics to behavior. New York: Liss.

Lee, K., Schottler, F., Oliver, M., and Lynch, G. (1980). Brief bursts of high frequency stimulation produce two types of structural changes in rat hippocampus. Journal of Neurophysiology 44, 247-258.

Lisman, J. (1994). The CaM kinase II hypothesis for the storage of synaptic memory. Trends in Neurosciences 17, 406-412.

Lynch, G. (1998). Memory and the brain: Unexpected chemistries and a new pharmacology. Neurobiology of Learning and Memory 70, 80-100.

Lynch, G., and Baudry, M. (1991). Re-evaluating the constraints on hypotheses regarding LTP expression. Hippocampus 1, 9-14.

Malenka, R. C., and Nicoll, R. A. (1999). Long-term potentiation—a decade of progress? Science 285, 1,870-1,874.

Malinow, R., Mainen, Z. F., and Hayashi, Y. (2000). LTP mechanisms: From silence to four-lane traffic. Current Opinion in Neurobiology 3, 352-357.

Malinow, R., and Tsien, R. W. (1990). Presynaptic enhancement shown by whole-cell recordings of long-term potentiation in hippocampal slices. Nature 346, 177-180.

Martin, A. R. (1966). Quantal nature of synaptic transmission. Physiological Review 46, 51-66.

Rall, W. (1978). Dendritic spines and synaptic potency. In R. Porter, ed., Studies in Neurophysiology. Cambridge, UK: Cambridge University Press.

Sommer, B., Keinanen, K., Verdoorn, T. A., Wisden, W., Burnashev, N., Herb, A., Kohler, M., Takagi, T., Sakmann, B., and Seeburg, P. H. (1990). Flip and flop: A cell-specific functional switch in glutamate-operated channels of the CNS. Science 249, 1,580-1,585.

Toni, N., Buchs, P.-A., Nikonenko, I., Bron, C. R., and Muller, D. (1999). LTP promotes formation of multiple spine synapses between an axon terminal and a dendrite. Nature 401, 421-425.

Wilson, C. J. (1988). Cellular mechanisms controlling the strength of synapses. Journal of Electron Microscope Technology 10, 293-313.

Zucker, R. S. (1989). Short-term synaptic plasticity. Annual Review of Neuroscience 12, 13-22.

MichelBaudry

GaryLynch