Longterm Changes in Synaptic Strength

Short-term changes in synaptic strength affect neurotransmitter release on a time-scale of seconds to minutes after a burst of activity. In addition, neuronal activity can lead to longer-term aftereffects that alter neurotransmitter release on a time-scale of hours or days. Such long-lasting changes require different cellular mechanisms from those that underlie short-term synaptic enhancement and depression. In this section, we will examine a particularly well studied example of these long-term changes: long-term potentiation (abbreviated LTP). As the name implies, LTP involves enhancement of synaptic strength lasting a week or more. Although LTP occurs at a variety of sites in the nervous system, we will concentrate on LTP in synaptic connections in a brain region called the hippocampus, which is involved in the formation of new memories.

In LTP, a burst of high-frequency activity in a presynaptic input enhances subsequent postsynaptic excitatory responses. Activity in one synapse can affect subsequent responses evoked by another synapse on the same postsyn-aptic cell (i.e., the potentiation is heterosynaptic), provided the synapses are active at approximately the same time (i.e., the potentiation is associative). Thus, a weakly stimulated synapse that is active contemporaneously with strong stimulation of the postsynaptic cell becomes potentiated. LTP is initiated in active synapses whenever the synaptic activity is paired with depolarization of the postsynaptic cell. If the postsynaptic neuron is depolarized by injecting positive current into the cell through a microelectrode, LTP is triggered in synaptic responses to the presynaptic cells that were active (even at a low rate) during the artificial depolarization. Synapses that were silent during the postsynaptic depolarization are not potentiated.

How does depolarization of the postsynaptic cell affect subsequent syn-aptic responses, and why does the potentiation affect only those synapses that are active during the depolarization? To answer these questions, we must first examine the anatomical arrangement of the excitatory synapses in the hippocampus and the properties of the postsynaptic receptor molecules that detect the neurotransmitter, glutamate, released by the presynaptic terminals. As with many other excitatory synapses in the central nervous system, the synaptic terminals contact the dendrites of hippocampal neurons at short, hairlike protuberances called dendritic spines. At high magnification, each spine is seen to consist of a knob-like swelling connected via a thin neck of cytoplasm to the main branch of the dendrite, as shown schematically in Figure 9-17. The thin connecting neck allows each spine to behave as a separate intracellular compartment, within which biochemical events can occur in isolation from the rest of the cell. Thus, internal signals can be generated in one spine without spreading to and affecting other spines on the dendrite. Each spine receives input from a single excitatory synaptic terminal. The combination of one terminal and one spine forms a functional synaptic unit that can be modulated separately from the other units on the dendrite of a single neuron. This structural organization may play a central role in the ability of LTP to selectively enhance transmission at active synapses, leaving inactive synapses unaffected.

Two types of glutamate receptors, called NMDA receptors and AMPA receptors, are located in the postsynaptic membrane of the dendritic spine. Both receptor types are glutamate-gated cation channels that have about equal permeability to sodium and potassium, but in addition, NMDA receptors permit influx of calcium ions while AMPA receptors do not. Another important difference between the two receptor types is that AMPA receptors open when glutamate binds, regardless of the membrane potential, whereas NMDA receptors require both glutamate and depolarization to open. NMDA receptors are blocked by external magnesium ions at negative membrane potentials. Block of the channel by magnesium, a divalent cation like calcium, is relieved during depolarization, allowing sodium, potassium, and calcium to move through the open channel. Influx of calcium through the open NMDA channel is the actual trigger for LTP, which explains why both activity and postsynaptic depolarization is required to initiate LTP. Activity provides glutamate and depolarization relieves block by magnesium, both of which are necessary to open NMDA channels and permit calcium to enter the dendritic spine.

The increase in internal calcium has multiple targets in the dendritic spine, summarized in Figure 9-18. LTP reflects, at least in part, an increase in neurotransmitter release from the presynaptic terminal, which raises the question of how an increase in postsynaptic calcium can influence presynaptic events. A retrograde messenger is required to transmit information from the dendritic spine to the synaptic terminal. Among the cellular targets for elevated calcium in the dendritic spine is nitric oxide synthase, which is an enzyme that produces nitric oxide (NO). NO is membrane permeant and can diffuse from the dendritic spine to the presynaptic terminal, where it might potentiate transmitter

Figure 9-17 Excitatory synapses onto hippocampal pyramidal cells are made onto spike-like protrusions of the dendrites, called dendritic spines.

Presynaptic targets (guanylyl cyclase ?)

Increased transmitter release

Increased transmitter release

synthase +

Other targets

NMDA receptor

synthase +

Ca • Calmodulin

Calmodulin

Ca2+

Other targets

NMDA receptor

Figure 9-18 Elevated intracellular calcium activates several cellular signals in dendritic spines. Calcium influx through NMDA receptors increases intracellular calcium, which binds to calmodulin. Calcium/calmodulin then activates two enzymes: nitric oxide synthase (NO synthase) and calcium/ calmodulin-dependent protein kinase II (CaM kinase II). Calcium also activates protein kinase C. NO synthase produces nitric oxide (NO) from arginine, and the membrane permeant messenger is thought to diffuse to the presynaptic terminal. NO then interacts with cellular signaling pathways, possibly including guanylyl cyclase, to potentiate transmitter release.

release by activating guanylyl cyclase, the synthetic enzyme for the second messenger cyclic GMP.

How might elevated calcium in a spine trigger postsynaptic factors that might also contribute to LTP? Several possible mechanisms for enhanced postsynaptic sensitivity to glutamate have been suggested. Figure 9-18 illustrates some other cellular targets for calcium in the dendritic spine, including two different kinases: protein kinase C (PKC) and calcium/calmodulin-dependent kinase II (CaM kinase II). When activated by elevated calcium, these enzymes phosphorylate specific target proteins in the postsynaptic cell. Phosphorylation is a cellular signal often used to activate or inactivate various kinds of proteins. In the case of LTP, the targets for phosphorylation by calcium-dependent kinases have not been established. Phosphorylation may increase the number of functional postsynaptic glutamate receptors, either because phosphoryla-tion allows the channels to open in response to glutamate or because phos-phorylation allows channels to attach to the cytoskeleton, anchoring them at the appropriate position in the postsynaptic membrane. Increased glutamate sensitivity might also arise from insertion of additional AMPA receptors into the postsynaptic membrane. All of these factors can potentiate e.p.s.p.'s by making more glutamate receptors available in the postsynaptic membrane to respond to glutamate released by the presynaptic terminal.

The excitatory synapses in the hippocampus demonstrate long-term depression (LTD), as well as LTP. If a synaptic input is activated at a low rate for a few minutes without strong activity in other synapses, the size of the e.p.s.p. elicited by that synaptic input diminishes and remains at this lower level for many hours. In this regard, LTD can be considered the opposite of LTP. In LTP, the effectiveness of a weakly stimulated synaptic input is enhanced when paired with strong activation of other pathways. In LTD, the effectiveness of a weakly stimulated synapse becomes reduced if its activation occurs in the absence of strong stimulation in other synaptic inputs. If LTP is induced at a particular synapse, it can subsequently be reversed by LTD. This fact suggests that LTD represents an erasure mechanism for LTP in the hippocampus: unless activation of a synaptic input is consistently strongly activated or paired with strong activation of other inputs, potentiation of that input is reversed by LTD.

0 0

Post a comment