Synaptic Plasticity

Synaptic Plasticity Overview

Synaptic connections provide the brain's wiring; unlike computer wiring, synaptic strength is dynamic and changes with activity.
Changes in synaptic transmission result from plasticity, varying from milliseconds to years.
Short-term plasticity alters neurotransmitter release from presynaptic terminals in response to action potentials.
- Facilitation, augmentation, and potentiation enhance neurotransmitter release due to calcium ions in the presynaptic terminal.
- Synaptic depression decreases neurotransmitter release, possibly due to depletion of synaptic vesicles.
Long-term plasticity alters synaptic transmission for 30 minutes or longer.
- Examples include long-term potentiation and long-term depression.
- Initial changes involve post-translational modifications of existing proteins (e.g., glutamate receptor trafficking), while later phases involve gene expression and new protein synthesis.
- These changes lead to enduring modifications, including synapse growth, and can permanently alter brain function.

Short-Term Synaptic Plasticity

Chemical synapses can undergo plastic changes that either strengthen or weaken synaptic transmission.
Mechanisms occur on timescales from milliseconds to days or longer.
Short-term forms last a few minutes or less and are observed during repeated activation.
Different forms vary in time courses and mechanisms.

Synaptic Facilitation

Rapid increase in synaptic strength occurs when two or more action potentials invade the presynaptic terminal within a few milliseconds of each other (Figure 8.1A).
The effect of the first action potential lasts for tens of milliseconds (Figure 8.1B).
Result of prolonged elevation of presynaptic calcium levels following synaptic activity.
Recent evidence suggests that synaptotagmin 7, a Ca2+-binding protein found on the plasma membrane and related to synaptotagmins serves as a target of this residual $Ca^{2+}$ signal that serves as the $Ca^{2+}$ sensor for triggering neurotransmitter release. (see Chapter 5).

Synaptic Depression

Synaptic depression opposes facilitation, causing neurotransmitter release to decline during sustained synaptic activity (Figure 8.1C, top).
Depends on the amount of neurotransmitter released.
Lowering external $Ca^{2+}$ concentration slows the rate of depression (see Figure 8.1C).
Total amount of depression is proportional to the amount of transmitter released (Figure 8.1D).
Caused by progressive depletion of a pool of synaptic vesicles available for release.
High release rates deplete vesicles rapidly, causing more depression; reduced release rates slow depletion, yielding less depression.
The strength of transmission declines until the vesicle pool is replenished by mobilization from a reserve pool.
More depression is observed after reducing the reserve pool by impairing synapsin, a protein that maintains vesicles in the reserve pool (see Chapter 5).

Synaptic Potentiation and Augmentation

Elicited by repeated synaptic activity and increase transmitter release from presynaptic terminals.
Enhance the ability of incoming $Ca^{2+}$ to trigger fusion of synaptic vesicles with the plasma membrane, but act over different timescales.
Augmentation rises and falls over a few seconds (see Figure 8.1C, bottom).
Potentiation acts over tens of seconds to minutes (Figure 8.1E). Potentiation can greatly outlast the tetanic stimulus that induces it, and is often called post-tetanic potentiation (PTP).
Thought to arise from prolonged elevation of presynaptic calcium levels during synaptic activity, but mechanisms are poorly understood.
Augmentation may result from $Ca^{2+}$ enhancing the actions of the presynaptic SNARE-regulatory protein munc-13 (see Figure 5.11).
Potentiation may arise when $Ca^{2+}$ activates presynaptic protein kinases that phosphorylate substrates (such as synapsin) that regulate transmitter release.

Interaction of Short-Term Plasticity Forms

During repetitive synaptic activity, the various forms of short-term plasticity can interact to cause synaptic transmission to change in complex ways
At the peripheral neuromuscular synapse, repeated activity first causes an accumulation of $Ca^{2+}$ in the presynaptic terminal that allows facilitation and then augmentation to enhance synaptic transmission (Figure 8.2).
The ensuing depletion of synaptic vesicles then causes depression to dominate and weaken the synapse.
*Presynaptic action potentials that occur within 1 to 2 minutes after the end of the tetanus release more neurotransmitter because of the persistence of post-tetanic potentiation.
Although their relative contributions vary from synapse to synapse, these forms of short-term synaptic plasticity collectively cause transmission at all chemical synapses to change dynamically as a consequence of the recent history of synaptic activity.

Long-Term Synaptic Plasticity in Aplysia

Facilitation, depression, augmentation, and potentiation modify synaptic transmission over timescales of a few minutes or less.
These mechanisms are likely responsible for short-lived changes in brain circuitry, but alone cannot provide the basis for changes in brain function that persist for weeks, months, or years.
*While these mechanisms probably are responsible for many short-lived changes in brain circuitry, they cannot provide the basis for changes in brain function that persist for weeks, months, or years.
Many synapses exhibit long-lasting forms of synaptic plasticity that are plausible substrates for more permanent changes in brain function. Because of their duration, these forms of synaptic plasticity may be cellular correlates of learning and memory.
An obvious obstacle to exploring synaptic plasticity in the brains of humans and other mammals is the enormous number of neurons and the complexity of synaptic connections.
One way to circumvent this dilemma is to examine plasticity in far simpler nervous systems. The assumption in this strategy is that plasticity is so fundamental that its essential cellular and molecular underpinnings are likely to be conserved in the nervous systems of very different organisms. *Eric Kandel and his colleagues studied Aplysia californica (sea slug; Figure 8.3A).
- This sea slug has only a few tens of thousands of neurons, many of which are quite large (up to 1 mm in diameter) and in stereotyped locations within the ganglia that make up the animal’s nervous system (Figure 8.3B).
- Attributes make it practical to monitor the electrical activity of specific, identifiable nerve cells, and thus to define the synaptic circuits involved in mediating the limited behavioral repertoire of Aplysia.

Behavioral Plasticity in Aplysia

Exhibit several elementary forms of behavioral plasticity.
- Habituation: A process that causes the animal to become less responsive to repeated occurrences of a stimulus.
  - For example, a light touch to the siphon of an Aplysia results in withdrawal of the animal’s gill, but habituation causes the gill withdrawal to become weaker during repeated stimulation of the siphon (Figure 8.3C).
- Sensitization: A process that allows an animal to generalize an aversive response—elicited by a noxious stimulus—to a variety of other, non-noxious stimuli.
  - In Aplysia that have habituated to siphon touching, sensitization of gill withdrawal is elicited by pairing a strong electrical stimulus to the animal’s tail with another light touch to the siphon.
  - This pairing causes the siphon stimulus to again elicit a strong withdrawal of the gill (see Figure 8.3C, right) because the noxious stimulus to the tail sensitizes the gill withdrawal reflex to light touch.
  - Even after a single stimulus to the tail, the gill withdrawal reflex remains enhanced for at least an hour (Figure 8.3D). This can be viewed as a simple form of short-term memory.
  - With repeated pairing of tail and siphon stimuli, this behavior can be altered for days or weeks (Figure 8.3E), thus demonstrating a simple form of long-term memory.

Neural Circuits of Gill Withdrawal

The small number of neurons in the Aplysia nervous system makes it possible to define the synaptic circuits involved in gill withdrawal and to monitor the activity of individual neurons in these circuits.
Critical neurons:
- Mechanosensory neurons: Innervate the siphon.
- Motor neurons: Innervate muscles in the gill.
- Interneurons: Receive inputs from a variety of sensory neurons (Figure 8.4A).
Touching the siphon activates the mechanosensory neurons, which form excitatory synapses that release glutamate onto both the interneurons and the motor neurons; thus, touching the siphon increases the probability that both of these postsynaptic targets will produce action potentials.
The interneurons form excitatory synapses on motor neurons, further increasing the likelihood of the motor neurons firing action potentials in response to mechanical stimulation of the siphon.
When the motor neurons are activated by the summed synaptic excitation of the sensory neurons and interneurons, they release acetylcholine that excites the muscle cells of the gill, producing gill withdrawal.

Plasticity Mechanisms

Both habituation and sensitization appear to arise from plastic changes in synaptic transmission in this circuit.
- During habituation, transmission at the glutamatergic synapse between the sensory and motor neurons is depressed (see Figure 8.4B, left).
  - This synaptic depression is thought to be responsible for the decreasing ability of siphon stimuli to evoke gill contractions during habituation.
- In contrast, sensitization modifies the function of this circuit by recruiting additional neurons.
  - The tail shock that evokes sensitization activates sensory neurons that innervate the tail.
  - These sensory neurons in turn excite modulatory interneurons that release serotonin onto the presynaptic terminals of the sensory neurons of the siphon (see Figure 8.4A).
  - Serotonin enhances transmitter release from the siphon sensory neuron terminals, leading to increased synaptic excitation of the motor neurons (Figure 8.4B).
The short-term sensitization is apparently due to recruitment of additional circuit elements, namely the modulatory interneurons that strengthen synaptic transmission in the gill withdrawal circuit.

Short-Term Sensitization Mechanism

Serotonin released by the modulatory interneurons binds to G-protein-coupled receptors on the presynaptic terminals of the siphon sensory neurons (step 1), which stimulates production of the second messenger, cAMP (step 2).
Cyclic AMP binds to the regulatory subunits of protein kinase A (PKA; step 3), liberating catalytic subunits of PKA that are then able to phosphorylate several proteins, probably including $K^+$ channels (step 4).
The net effect of the action of PKA is to reduce the probability that the $K^+$ channels open during a presynaptic action potential.
This effect prolongs the presynaptic action potential, thereby opening more presynaptic $Ca^{2+}$ channels (step 5).
*There is evidence that the opening of presynaptic Ca2+ channels is also directly enhanced by serotonin.
*That modulation of the sensory neuron–motor neuron synapse lasts approximately 1 hour (Figure 8.4C).
Finally, the enhanced influx of $Ca^{2+}$ into the presynaptic terminals increases the amount of transmitter released onto motor neurons during a sensory neuron action potential (step 6).
In summary, a signal transduction cascade that involves neurotransmitters, second messengers, protein kinases, and ion channels mediates short-term sensitization of gill withdrawal. This cascade ultimately causes a short-term enhancement of synaptic transmission between the sensory and motor neurons within the gill withdrawal circuit.

Long-Term Sensitization Mechanism

The same serotonin-induced enhancement of glutamate release that mediates short-term sensitization is also thought to underlie long-term sensitization.
During long-term sensitization, this circuitry is affected for up to several weeks.
The prolonged duration of this form of plasticity is evidently due to changes in gene expression and thus protein synthesis (Figure 8.5B).
*With repeated training (i.e., additional tail shocks), the serotonin-activated PKA involved in short-term sensitization also phosphorylates—and thereby activates—the transcriptional activator CREB.
CREB binding to the cAMP response elements (CREs) in regulatory regions of nuclear DNA increases the rate of transcription of downstream genes.
CREB stimulates the synthesis of an enzyme, ubiquitin hydrolase, that stimulates degradation of the regulatory subunit of PKA. This causes a long-lasting increase in the amount of free catalytic subunit, meaning that some PKA is persistently active and no longer requires serotonin to be activated.
CREB also stimulates another transcriptional activator protein called C/EBP. C/EBP stimulates transcription of other, unknown genes that cause addition of synaptic terminals, yielding a long-term increase in the number of synapses between the sensory and the motor neurons. Such structural increases are not seen following short-term sensitization. They may represent the ultimate cause of the long-lasting change in overall strength of the relevant circuit connections that produce a long-lasting enhancement in the gill withdrawal response.
*Another protein involved in the long-term synaptic facilitation is a cytoplasmic polyadenylation element binding protein, somewhat confusingly called CPEB. CPEB activates mRNAs and may be important for local control of protein synthesis.
*Most intriguing, CPEB has self-sustaining properties like those of prion proteins (see Clinical Applications, Chapter 19), which could allow CPEB to remain active in perpetuity and thereby mediate permanent changes in synaptic structure to generate long-term sensitization.

Fruit Fly Genetics of Learning and Memory

*Genetic mutations such as dunce, rutabaga, and amnesiac, suggest that a central pathway for learning and memory in the fly is signal transduction mediated by the cyclic nucleotide cAMP.

Generalizations about Synaptic Plasticity

Synaptic plasticity can lead to changes in circuit function and, ultimately, to behavioral plasticity.
Plastic changes in synaptic function can be short-term effects that rely on post-translational modification of existing synaptic proteins, or they can be long-term changes that require changes in gene expression, new protein synthesis, and growth of new synapses (as well as enlarging or eliminating existing synapses).
*Thus, it appears that short- and long-term changes in synaptic function have different mechanistic underpinnings.

Long-Term Potentiation (LTP) in the Mammalian Brain

Some patterns of synaptic activity produce a long-lasting increase in synaptic strength known as long-term potentiation (LTP), whereas other patterns of activity produce a long-lasting decrease in synaptic strength, known as long-term depression (LTD).
*LTP and LTD are broad terms that describe only the direction of change in synaptic efficacy; in fact, different cellular and molecular mechanisms can be involved in producing LTP or LTD at different synapses throughout the brain.
LTP and LTD are produced by different histories of activity and are mediated by different complements of intracellular signal transduction pathways in the nerve cells involved.
Long-term synaptic plasticity has been most thoroughly studied at excitatory synapses in the mammalian hippocampus.
*Hippocampus is especially important in the formation and retrieval of some forms of memory (see Chapter 30).
LTP was discovered in the rabbit hippocampus by Terje Lomo and Timothy Bliss.
Progress in understanding the mechanism of LTP has relied heavily on in vitro studies of slices of living hippocampus.
*Pyramidal neurons lie in densely packed layers; the major regions include CA1 and CA3.
*The dendrites of pyramidal cells in the CA1 region form a thick band (the stratum radiatum), where they receive synapses from Schaffer collaterals, the axons of pyramidal cells in the CA3 region.
*The arrangement of neurons allows the hippocampus to be sectioned such that most of the relevant circuitry is left intact.

LTP Induction

*Electrical stimulation of Schaffer collaterals generates EPSPs in the postsynaptic CA1 cells (Figure 8.7A,B).

If the Schaffer collaterals are stimulated only two or three times per minute, the size of the EPSP elicited in the CA1 neurons remains constant.
However, a brief, high-frequency train of stimuli to the same axons causes LTP, which is evident as a long-lasting increase in EPSP amplitude (Figure 8.7B,C).
While the maximum duration of LTP is not known, in some cases LTP can last for more than a year (Figure 8.7D).
LTP occurs at each of the three excitatory synapses of the hippocampus shown in Figure 8.6.
LTP also is found at excitatory synapses in a variety of brain regions—including the cortex, amygdala, and cerebellum—and at some inhibitory synapses as well.

Properties of LTP

LTP requires strong activity in both presynaptic and postsynaptic neurons.
If action potentials in a small number of presynaptic Schaffer collaterals—which evoke transmitter release that produces subthreshold EPSPs that would not normally yield LTP—are paired with strong depolarization of the postsynaptic CA1 cell, the activated Schaffer collateral synapses undergo LTP (Figure 8.8).
*Such a requirement for coincident presynaptic and postsynaptic activity is the central postulate of a theory of learning devised by Donald Hebb; Hebb proposed that coordinated activity of a presynaptic terminal and a postsynaptic neuron would strengthen the synaptic connection between them.
*Requirement indicates the involvement of a coincidence detector that allows LTP to occur only when both presynaptic and postsynaptic neurons are active.
- The synapse specificity of LTP means that each of the tens of thousands of synapses on a hippocampal neuron can store information, thereby making it possible for the millions of neurons in the hippocampus to store a vast amount of information.
- LTP is input specific: When LTP is induced by activation of one synapse, it does not occur in other, inactive synapses that contact the same neuron (see Figure 8.7C).
  - LTP is restricted to activated synapses rather than to all of the synapses on a given cell (Figure 8.9A).
Associativity: If one pathway is weakly activated at the same time that a neighboring pathway onto the same cell is strongly activated, both synaptic pathways undergo LTP (Figure 8.9B).
* Associativity is another consequence of the coincidence detection feature of LTP, specifically pairing of the activity of the weak synapse with the coincident generation of action potentials by the strong synapse.
*Selective enhancement of conjointly activated sets of synaptic inputs is often considered a cellular analog of associative learning, where two stimuli are required for learning to take place; the best-known type of associative learning is classical (Pavlovian) conditioning.

NMDA Receptors

*Antagonists of the NMDA-type glutamate receptor prevent LTP during high-frequency stimulation of the Schaffer collaterals, but have no effect on the synaptic response evoked by low-frequency stimulation.

NMDA receptor channel is permeable to $Ca^{2+}$ but is blocked by $Mg^{2+}$ at the normal resting membrane potential.
NMDA receptor is a molecular coincidence detector: The channel of this receptor opens (to induce LTP) only when glutamate is bound to the receptor, and the postsynaptic cell is depolarized to relieve the $Mg^{2+}$ block of the channel pore.
During low-frequency synaptic transmission, glutamate released by the Schaffer collaterals binds to both NMDA-type and AMPA-type glutamate receptors; however, $Mg^{2+}$ blockade prevents current flow through the NMDA receptors, so that the EPSP is mediated entirely by the AMPA receptors (Figure 8.10, left).
Summation of EPSPs during high-frequency stimulation leads to a prolonged depolarization that expels $Mg^{2+}$ from the NMDA channel pore (see Figure 8.10, right).

Role of Calcium

Induction of LTP is due to accumulation of postsynaptic $Ca^{2+}$ as a result of $Ca^{2+}$ influx through NMDA receptors.
Imaging studies have shown that activation of NMDA receptors increases postsynaptic $Ca^{2+}$ levels and that these $Ca^{2+}$ signals can be restricted to the dendritic spines of individual synapses.
*Furthermore, injection of Ca2+ chelators blocks LTP induction, whereas elevation of $Ca^{2+}$ levels in postsynaptic neurons potentiates synaptic transmission.
A rise in postsynaptic $Ca^{2+}$ concentration serves as a second-messenger signal that induces LTP.

AMPA Receptors

*Excitatory synapses can dynamically regulate their postsynaptic AMPA receptors via the same sorts of membrane trafficking processes that occur in presynaptic neurons during neurotransmitter release.

LTP apparently is due to synaptotagmin-mediated insertion of AMPA receptors into the postsynaptic membrane.
The resulting increase in the number of AMPA receptors increases the response of the postsynaptic cell to released glutamate (Figure 8.11A), yielding a strengthening of synaptic transmission that can last for as long as LTP is maintained (Figure 8.11B).
*LTP does not affect the number of postsynaptic NMDA receptors; thus, while these receptors are crucial for induction of LTP, they do not play a major role in LTP expression.
*Can even add new AMPA receptors to “silent” synapses that did not previously have postsynaptic AMPA receptors
The strengthening of synaptic transmission during LTP arises from an increase in the sensitivity of the postsynaptic cell to glutamate.

CaMKII and PKC

* $Ca^{2+}$ also activates complex postsynaptic signal transduction cascades that include at least two $Ca^{2+}$ -activated protein kinases: $Ca^{2+}$ /calmodulin-dependent protein kinase type II (CaMKII) and protein kinase C (PKC; see Chapter 7).

CaMKII, which is the most abundant postsynaptic protein at Schaffer collateral synapses, seems to play an especially important role. This enzyme is activated by stimuli that induce LTP (Figure 8.12), and pharmacological inhibition or genetic deletion of CaMKII prevents LTP.
Activation of CaMKII in single dendritic spines during long-term potentiation. Nature 458: 299–304.
*It is thought that CaMKII and PKC phosphorylate downstream targets, including both AMPA receptors and other targets, that collectively facilitate delivery of extrasynaptic AMPA receptors to the synapse.

LTP Signaling Pathway

$Ca^{2+}$ entering through postsynaptic NMDA receptors leads to activation of synaptotagmins and protein kinases that regulate trafficking of AMPA receptors, thereby enhancing the postsynaptic response to glutamate released from the presynaptic terminal (Figure 8.13).
*The scheme depicted in Figure 8.13 can account for the changes in synaptic transmission that occur over the first 1 to 2 hours after LTP is induced.
However, there is also a later phase of LTP that depends on changes in gene expression and the synthesis of new proteins.
The contributions of this late phase can be observed by treating synapses with drugs that inhibit protein synthesis: Blocking protein synthesis prevents LTP measured several hours after a stimulus but does not affect LTP measured at earlier times (Figure 8.14).
This late phase of LTP appears to be initiated by protein kinase A, which goes on to activate transcription factors such as CREB, which stimulate the expression of other proteins
*There is evidence that the number and size of synaptic contacts increase during LTP (Figure 8.15B,C).

Overall

In the mammalian hippocampus, LTP has many parallels to the long-term changes in synaptic transmission underlying behavioral sensitization in Aplysia. Both consist of an early, transient phase that relies on protein kinases to produce post-translational changes in membrane ion channels, and both have later, long-lasting phases that require changes in gene expression mediated by CREB.
Both forms of long-term synaptic plasticity are likely to be involved in long-term storage of information, although the role of LTP in memory storage in the hippocampus is not firmly established.

Long-Term Depression (LTD)

LTD occurs at the synapses between the Schaffer collaterals and the CA1 pyramidal cells in the hippocampus.
Whereas LTP at these synapses requires brief, high-frequency stimulation, LTD occurs when the Schaffer collaterals are stimulated at a low rate—about 1 Hz—for long periods (10–15 minutes).
*The major determinant of whether LTP or LTD arises appears to be the nature of the $Ca^{2+}$ signal in the postsynaptic cell: Small and slow rises in $Ca^{2+}$ lead to depression, whereas large and fast increases in $Ca^{2+}$ trigger potentiation.

LTD Mechanisms

Small and slow rises in $Ca^{2+}$ lead to depression, whereas large and fast increases in $Ca^{2+}$ trigger potentiation.
LTD appears to result from activation of phosphatases, specifically PP1 and PP2B (calcineurin), a $Ca^{2+}$ -dependent phosphatase (see Chapter 7). Evidence in support of this idea is that inhibitors of these phosphatases prevent LTD but do not block LTP.
*LTD is often associated with a loss of synaptic AMPA receptors where this loss probably arises from internalization of AMPA receptors into sorting endosomes in the postsynaptic cell (Figure 8.16C), due to the same clathrin-dependent endocytosis mechanisms important for synaptic vesicle recycling in the presynaptic terminal (see Chapter 5).

LTD in the Cerebellum

*Quite a different form of LTD is observed in the cerebellum.
*Masao Ito and Masanobu Kano described LTD of synaptic inputs onto cerebellar Purkinje cells
*Purkinje neurons in the cerebellum receive two distinct types of excitatory input: climbing fibers and parallel fibers.
*Associativity arises from the combined actions of two distinct intracellular signal transduction pathways that are activated in the postsynaptic Purkinje cell in response to activity of the climbing fiber and parallel fiber synapses.
Long-term synaptic depression in the Purkinje cells.

Cerebellar Pathways

Glutamate released from the parallel fiber terminals activates two types of receptors, the AMPA-type and metabotropic glutamate receptors.
* Glutamate binding to the AMPA receptor results in mild membrane depolarization, whereas binding to the metabotropic receptor produces the second messengers inositol triphosphate (IP3) and diacylglycerol (DAG).
Glutamate released by climbing fibers also activates AMPA receptors, which strongly depolarizes the Purkinje cell membrane potential and initiates a second signal transduction pathway: an influx of $Ca^{2+}$ through voltage-gated channels and a subsequent increase in intracellular $Ca^{2+}$ concentration.
*It activates both IP3 receptors and PKC which serve as coincidence detectors.
*
**In contrast to LTD in the hippocampus, cerebellar LTD requires the activity of protein kinases, rather than phosphatases, and does not involve Ca2+ entry through NMDA receptors.
*As is the case for LTP at the hippocampal Schaffer collateral synapse, as well as for long-term synaptic plasticity in Aplysia, CREB appears to be required for a late phase of cerebellar LTD. It is not yet known which proteins are synthesized as a consequence of CRE activation.

Spike Timing-Dependent Plasticity (STDP)

*The precise temporal relationship between activity in the pre- and postsynaptic cells can also be an important determinant of the amount and direction of long-term synaptic plasticity.
*if the presynaptic activity is preceded by a postsynaptic action potential, LTD will occur.

the postsynaptic action potential follows pre-synaptic activity, LTP will occur (Figure 8.18A,B).
*The stronger inputs would be more likely to produce suprathreshold EPSPs
*If a synapse generates a suprathreshold EPSP, the resulting postsynaptic action potential would rapidly follow presynaptic activity, and the resulting LTP would encode the fact that the postsynaptic action potential resulted from the activity of that synapse.
*Has been postulated that endocannabinoids may also be required for LTD induction during STDP.

Clinical Applications: Epilepsy

*Epilepsy is mediated by the rhythmic firing of large groups of neurons.
*Kindling can induce seizures; this is essentially permanent.
Modern thinking about the causes of cures of epilepsy have focused on where seizures originate and the mechanisms that make the affected region hyperexcitable.