**3.4 Molecular mechanisms of LTP**

The cellular and molecular mechanisms of LTP induction are comprised of many events such as covalent modification of pre-existing proteins, the activation of cellular programs for gene expression and increased protein synthesis. The regulatory events move from the synapse to the nucleus and then back to the synapse in the course of LTP induction.

LTP induction experiments have mostly been done in hippocampal excitatory synapses. The hippocampus is divided into three distinctive regions composed of three distinctive kinds of cells. The dentate gyrus (DG), which is composed of granule cells and the CA3 and CA1 regions, which are composed of pyramidal cells having different properties. These regions are connected by well defined pathways through which signals traverse the hippocampus. The perforant fibre pathway (pp) from the entorhinal cortex forms excitatory connections with the granule cells of the DG. The granule cells give rise to axons that form the mossy fibre pathway (mf), which connects with the pyramidal cells in area CA3 of the hippocampus. The pyramidal cells of the CA3 region project to the pyramidal cells in CA1 by means of the Schaffer collateral pathway (Fig. 1).

LTP is widely studied in the CA1 region of the hippocampus (Bliss and Collingridge, 1993; Reymann and Frey, 2007). The establishment of LTP in the CA1 region requires both presynaptic activity and large postsynaptic depolarization. The original stimulus protocol used by Bliss and Lomo in the anesthetized rabbit ranged from 10 to 100 Hz (Bliss & Lomo, 1973). Since that time, a variety of LTP induction protocols from different research groups have emerged in the literature. Most involve trains of high-frequency stimulation (tetanization) that are delivered to presynaptic axons. The tetanization typically lasts several seconds and is delivered at frequencies of 25 to 400 Hz.

The induction of LTP requires an influx of calcium into the postsynaptic neuron that can be either through NMDAR dependent or NMDAR-independent mechanisms.

Molecular Mechanisms in Synaptic Plasticity 301

LTP. This calcium influx activates several important signaling pathways involving different protein kinases and phosphatases. One of the kinases activated by the influx of calcium through NMDARs is Ca2+/calmodulin dependent protein kinase II (CaMKII), which is known as the memory molecule. CaMKII is a Ser/Thr protein kinase, abundant in glutamatergic postsynaptic terminal. The activation of CaMKII by Ca2+/CaM complex leads to the formation of autophosphorylated enzyme at Thr286 position, which will make it calcium independent. Thus the Thr286 autophosphorylated form of the enzyme will maintain its activity even though Ca2+/CaM complex is removed from its regulatory domain. The autophosphorylation can enhance binding affinity of the enzyme for Ca2+/CaM by a 1000 fold. Studies have shown that Thr286 autophosphorylated enzyme is required for the induction of LTP. Upon activation, CaMKII can rapidly translocate to the postsynaptic density (PSD), where postsynaptic receptors such as AMPAR and NMDAR are concentrated. The translocated CaMKII can bind to different subunits of NMDAR such as GluN1, GluN2A and GluN2B, which are the ideal postsynaptic adapters. Of these, GluN2B-CaMKII interaction is well characterized and is essential for the induction and maintenance of LTP (Barria & Malinow, 2005; Lisman et al., 2011). The AMPAR is one of the substrates for CaMKII (as well as for PKC) in the PSD where CaMKII can phosphorylate GluA1 subunit of AMPAR at Ser831. This phosphorylation of GluA1 by CaMKII (Barria et al., 1997b) leads to an increased conductance of homomeric GluA1 channels (Derkach et al., 1999) and is believed to be one of the major contributors to the enhanced efficacy of glutamatergic synapses in CA1 area of hippocampus during LTP

LTP can occur either in AMPAR containing synapses or in synapses lacking AMPAR. When a glutamatergic synapse is formed, only NMDAR will be present in the postsynaptic membrane. Such synapses lacking AMPA receptors are called *silent synapses*, where AMPAR gets inserted in the postsynaptic membrane during the activation of nearby synapses. As a consequence of NMDAR activation and the resulting Ca2+ influx into the post synaptic dendrite, new AMPARs get inserted into the post synaptic membrane. This '*AMPAfication'* of the synapse makes the transmission stronger (Bear, 2001). Thus enhanced AMPAR activity either by increase in AMPAR abundance in the synapse or by increase in the conductivity of AMPARs is the key postsynaptic mechanism leading to increase in EPSP response seen in LTP. Studies have shown that activated forms of α-CaMKII can enhance the synaptic trafficking of AMPARs. PKA can also participate in *AMPAfication* by phosphorylating GluA1 at Ser845 which enhances AMPAR exocytosis (Oh, 2005). AMPAR recruitment mediated by PKA is shown in Fig. 3. Activation of PKA also boosts the activity of CaMKII indirectly by decreasing the competing protein phosphatase activity especially protein phosphatase 1(PP1). PKA inhibits PP1 by activating the inhibitor of PP1 called

Several other protein kinases, including protein kinase C (PKC), PKA, the tyrosine kinase Src, and mitogen-activated protein kinase (MAPK), have also been suggested to contribute to LTP (Teyler et al., 1987). The evidence in support of critical roles for these kinases is, however, considerably weaker than that for CaMKII. PKC has been suggested to play a role analogous to that of CaMKII, because PKC inhibitors have been reported to block LTP and increasing postsynaptic PKC activity can enhance synaptic transmission (Hu et al., 1987).

(Fig. 2).

inhibitor-1(Bryne, 2009).

Fig. 1. Hippocampus

#### **3.4.1 NMDAR dependent mechanism (NMDAR-LTP)**

The best understood form of LTP is induced by the activation of the NMDAR complex. This subtype of glutamate receptor allows electrical events at the postsynaptic membrane to be transduced into chemical signals which, in turn, are thought to activate both pre and postsynaptic mechanisms to generate a persistent increase in synaptic strength.

 Glutamate is a major excitatory neurotransmitter in the brain. During nerve impulse transmission, glutamate will be released into the synapse from the presynaptic terminal. Glutamate receptors present on the postsynaptic membrane are the initial triggers for the ensuing postsynaptic calcium signaling mechanism responsible for the induction of LTP. NMDA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and kainate receptors are the ionotropic-glutamate receptors present on the postsynaptic membrane. Among these NMDA and AMPA receptors play an important role in the induction of LTP. NMDARs are formed from hetero-tetrameric assemblies of GluN1 (previously NR1) subunits with GluN2A-D (NR2A-D) and Glu3A/B (NR3A/B). NMDARs require the binding of L-glutamate and the co-agonist glycine, as well as depolarization, to become activated and conduct Na+, K+ and Ca2+ ions. AMPARs are composed of four subunits, GluA1-4 (previously GluR1-4). The Q/R edited GluA2 subunit is critical for the biophysical properties of AMPARs producing low conductance, non-rectifying, Ca2+-impermeable AMPARs. Postnatally the great majority of AMPARs contain edited GluA2 in excitatory synapses.

Glutamate binding to the AMPA receptor leads to a sodium influx into the postsynaptic compartment. This leads to depolarization causing release of Mg2+ block present on the NMDA receptor. The binding of glutamate and the removal of Mg2+ block causes NMDA receptor to open and conduct Ca2+ and Na+ into the cell. The influx of Ca2+ is essential for LTP induction. With repeated activation of the neuron, sufficient calcium will enter into the postsynaptic compartment and triggers the molecular events needed for the induction of

The best understood form of LTP is induced by the activation of the NMDAR complex. This subtype of glutamate receptor allows electrical events at the postsynaptic membrane to be transduced into chemical signals which, in turn, are thought to activate both pre and

 Glutamate is a major excitatory neurotransmitter in the brain. During nerve impulse transmission, glutamate will be released into the synapse from the presynaptic terminal. Glutamate receptors present on the postsynaptic membrane are the initial triggers for the ensuing postsynaptic calcium signaling mechanism responsible for the induction of LTP. NMDA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and kainate receptors are the ionotropic-glutamate receptors present on the postsynaptic membrane. Among these NMDA and AMPA receptors play an important role in the induction of LTP. NMDARs are formed from hetero-tetrameric assemblies of GluN1 (previously NR1) subunits with GluN2A-D (NR2A-D) and Glu3A/B (NR3A/B). NMDARs require the binding of L-glutamate and the co-agonist glycine, as well as depolarization, to become activated and conduct Na+, K+ and Ca2+ ions. AMPARs are composed of four subunits, GluA1-4 (previously GluR1-4). The Q/R edited GluA2 subunit is critical for the biophysical properties of AMPARs producing low conductance, non-rectifying, Ca2+-impermeable AMPARs. Postnatally the great majority of AMPARs contain edited GluA2 in excitatory

Glutamate binding to the AMPA receptor leads to a sodium influx into the postsynaptic compartment. This leads to depolarization causing release of Mg2+ block present on the NMDA receptor. The binding of glutamate and the removal of Mg2+ block causes NMDA receptor to open and conduct Ca2+ and Na+ into the cell. The influx of Ca2+ is essential for LTP induction. With repeated activation of the neuron, sufficient calcium will enter into the postsynaptic compartment and triggers the molecular events needed for the induction of

postsynaptic mechanisms to generate a persistent increase in synaptic strength.

Fig. 1. Hippocampus

synapses.

**3.4.1 NMDAR dependent mechanism (NMDAR-LTP)** 

LTP. This calcium influx activates several important signaling pathways involving different protein kinases and phosphatases. One of the kinases activated by the influx of calcium through NMDARs is Ca2+/calmodulin dependent protein kinase II (CaMKII), which is known as the memory molecule. CaMKII is a Ser/Thr protein kinase, abundant in glutamatergic postsynaptic terminal. The activation of CaMKII by Ca2+/CaM complex leads to the formation of autophosphorylated enzyme at Thr286 position, which will make it calcium independent. Thus the Thr286 autophosphorylated form of the enzyme will maintain its activity even though Ca2+/CaM complex is removed from its regulatory domain. The autophosphorylation can enhance binding affinity of the enzyme for Ca2+/CaM by a 1000 fold. Studies have shown that Thr286 autophosphorylated enzyme is required for the induction of LTP. Upon activation, CaMKII can rapidly translocate to the postsynaptic density (PSD), where postsynaptic receptors such as AMPAR and NMDAR are concentrated. The translocated CaMKII can bind to different subunits of NMDAR such as GluN1, GluN2A and GluN2B, which are the ideal postsynaptic adapters. Of these, GluN2B-CaMKII interaction is well characterized and is essential for the induction and maintenance of LTP (Barria & Malinow, 2005; Lisman et al., 2011). The AMPAR is one of the substrates for CaMKII (as well as for PKC) in the PSD where CaMKII can phosphorylate GluA1 subunit of AMPAR at Ser831. This phosphorylation of GluA1 by CaMKII (Barria et al., 1997b) leads to an increased conductance of homomeric GluA1 channels (Derkach et al., 1999) and is believed to be one of the major contributors to the enhanced efficacy of glutamatergic synapses in CA1 area of hippocampus during LTP (Fig. 2).

LTP can occur either in AMPAR containing synapses or in synapses lacking AMPAR. When a glutamatergic synapse is formed, only NMDAR will be present in the postsynaptic membrane. Such synapses lacking AMPA receptors are called *silent synapses*, where AMPAR gets inserted in the postsynaptic membrane during the activation of nearby synapses. As a consequence of NMDAR activation and the resulting Ca2+ influx into the post synaptic dendrite, new AMPARs get inserted into the post synaptic membrane. This '*AMPAfication'* of the synapse makes the transmission stronger (Bear, 2001). Thus enhanced AMPAR activity either by increase in AMPAR abundance in the synapse or by increase in the conductivity of AMPARs is the key postsynaptic mechanism leading to increase in EPSP response seen in LTP. Studies have shown that activated forms of α-CaMKII can enhance the synaptic trafficking of AMPARs. PKA can also participate in *AMPAfication* by phosphorylating GluA1 at Ser845 which enhances AMPAR exocytosis (Oh, 2005). AMPAR recruitment mediated by PKA is shown in Fig. 3. Activation of PKA also boosts the activity of CaMKII indirectly by decreasing the competing protein phosphatase activity especially protein phosphatase 1(PP1). PKA inhibits PP1 by activating the inhibitor of PP1 called inhibitor-1(Bryne, 2009).

Several other protein kinases, including protein kinase C (PKC), PKA, the tyrosine kinase Src, and mitogen-activated protein kinase (MAPK), have also been suggested to contribute to LTP (Teyler et al., 1987). The evidence in support of critical roles for these kinases is, however, considerably weaker than that for CaMKII. PKC has been suggested to play a role analogous to that of CaMKII, because PKC inhibitors have been reported to block LTP and increasing postsynaptic PKC activity can enhance synaptic transmission (Hu et al., 1987).

Molecular Mechanisms in Synaptic Plasticity 303

Fig. 3. AMPAR exocytosis regulation by PKA. Ca2+ signaling can activate PKA via adenyl cyclase-cAMP pathway. PKA can phosphorylate GluA1 subunit of AMPAR at Ser845 and this leads to the recruitment of AMPARs into extrasynaptic site. This extrasynaptic pool of

This phosphorylation can initiate transcription of CRE-associated genes. One protein that is regulated by the CREB family of transcription factors is brain-derived neurotrophic factor (BDNF), a key regulator in the conversion of E-LTP to L-LTP. BDNF can bind to a specific receptor tyrosine kinase, TrkB. This binding results in dimerization and autophosphorylation of the Trk receptors, leading to activation of the tyrosine kinases. Activated receptors in general are capable of triggering a number of signal transduction cascades including the MAPK pathway, the phosphatidylinositol 3-kinase (PI3K) pathway, and the phospholipase Cγ (PLC-γ) pathway. The signals thus generated also can pass on to the nucleus to cause further

PKA can also recruit MAPK to the nucleus where it can phosphorylate other kinases and transcription factors (eg: CREB) to activate gene transcription. Extra cellular signal regulated protein kinase (ERK), is a member of the mitogen-activated family of protein kinases, which play a crucial role in L-LTP. ERK activity is required to initiate the local translation of messenger RNAs (mRNAs) that are present at spines into functional proteins. Another function of ERK is its rapid translocation into the nucleus of the neuron where it phosphorylates several regulatory transcription factors. This leads to the transcription of several mRNAs that are transported along dendrites toward the spines and their synapses.

activation of transcription factors and alterations in gene expression (Lu, 2003).

AMPARs can then diffuse to PSD during NMDAR activation.

Fig. 2. Molecular mechanisms of NMDAR Dependent LTP.

The calcium influx through NMDAR also activates adenyl cyclase, which generates cAMP in the postsynaptic compartment. This second messenger generated thus triggers a series of downstream signalling mechanisms, which function more in LTP maintenance. The local increase in cAMP levels leads to the activation of PKA by causing the catalytic subunits of this enzyme to dissociate from the regulatory subunits.

The activated PKA can regulate gene expression. PKA can modify transcription by phosphorylating several different transcription factors, one of which is the cAMP response element binding protein (CREB). CREB is a nuclear protein that modulates transcription of genes containing cAMP response elements (CRE) in their promoters (Kandel, 2001). The catalytic subunits of PKA can translocate to nucleus and phosphorylate serine-133 on CREB.

Fig. 2. Molecular mechanisms of NMDAR Dependent LTP.

this enzyme to dissociate from the regulatory subunits.

CREB.

The calcium influx through NMDAR also activates adenyl cyclase, which generates cAMP in the postsynaptic compartment. This second messenger generated thus triggers a series of downstream signalling mechanisms, which function more in LTP maintenance. The local increase in cAMP levels leads to the activation of PKA by causing the catalytic subunits of

The activated PKA can regulate gene expression. PKA can modify transcription by phosphorylating several different transcription factors, one of which is the cAMP response element binding protein (CREB). CREB is a nuclear protein that modulates transcription of genes containing cAMP response elements (CRE) in their promoters (Kandel, 2001). The catalytic subunits of PKA can translocate to nucleus and phosphorylate serine-133 on

Fig. 3. AMPAR exocytosis regulation by PKA. Ca2+ signaling can activate PKA via adenyl cyclase-cAMP pathway. PKA can phosphorylate GluA1 subunit of AMPAR at Ser845 and this leads to the recruitment of AMPARs into extrasynaptic site. This extrasynaptic pool of AMPARs can then diffuse to PSD during NMDAR activation.

This phosphorylation can initiate transcription of CRE-associated genes. One protein that is regulated by the CREB family of transcription factors is brain-derived neurotrophic factor (BDNF), a key regulator in the conversion of E-LTP to L-LTP. BDNF can bind to a specific receptor tyrosine kinase, TrkB. This binding results in dimerization and autophosphorylation of the Trk receptors, leading to activation of the tyrosine kinases. Activated receptors in general are capable of triggering a number of signal transduction cascades including the MAPK pathway, the phosphatidylinositol 3-kinase (PI3K) pathway, and the phospholipase Cγ (PLC-γ) pathway. The signals thus generated also can pass on to the nucleus to cause further activation of transcription factors and alterations in gene expression (Lu, 2003).

PKA can also recruit MAPK to the nucleus where it can phosphorylate other kinases and transcription factors (eg: CREB) to activate gene transcription. Extra cellular signal regulated protein kinase (ERK), is a member of the mitogen-activated family of protein kinases, which play a crucial role in L-LTP. ERK activity is required to initiate the local translation of messenger RNAs (mRNAs) that are present at spines into functional proteins. Another function of ERK is its rapid translocation into the nucleus of the neuron where it phosphorylates several regulatory transcription factors. This leads to the transcription of several mRNAs that are transported along dendrites toward the spines and their synapses.

Molecular Mechanisms in Synaptic Plasticity 305

phosphatase switch in an energy efficient manner (Cheriyan et al., 2011). Activated CaMKII can function in enhancing AMPAR currents and its recruitment. This will contribute to the

Fig. 4. CaMKII-phosphatase bistable switch model. Continuous interconversion between

Activated PKA can also contribute to the maintenance of LTP by involving in a selfsustaining mechanism, in addition to its role in promoting AMPAR exocytosis. As described earlier, activated PKA can alter gene expression via cAMP-PKA-CREB pathway. One gene activated by CREB encodes a ubiquitin hydrolase, a component of a specific ubiquitin protease that leads to the regulated proteolysis of the regulatory subunit of PKA. This results in persistent activity of PKA, leading to persistent phosphorylation of PKA substrates such as CREB, MAPK, etc., thereby completing a self-sustaining cycle that can be

L-LTP requires *de novo* protein synthesis. The long lasting activity changes require nuclear transcription followed by delivery of newly synthesized proteins to the synapse to yield synaptic remodeling. Newly synthesized proteins delivered by non-directed transport from the cell body must be captured locally at the activated synapse in order to function in an input-specific manner (Doyle, 2011). For this, the activated synapse requires a local signal that allows it to capture proteins or mRNAs for protein-synthesis-dependent LTP or LTD. This process has been termed *synaptic tagging.* Based on this proposal, synaptic activity generates a tag, which "captures" the plasticity-related proteins (PRPs) derived outside of synapses (Lu et al., 2011). These findings indicate a tight and extensive dialogue between the

mRNA localization to the synapses depends on synaptic activity and the mechanism behind this transport is largely a mystery. This transport mechanism is highly complex and involves multiple mRNA binding proteins. This process can be divided into different stages, (1) the presence of cis-acting localization elements (LEs) or zipcodes generally located in the 3'-untranslated region (3'-UTR) of localized transcripts, (2) the recognition of these signals by trans-acting RNA-binding proteins (RBPs), (3) the assembly of RBPs and their cargo

CaMKII and phospho-CaMKII is catalysed by the kinase activity of Thr286 autophosphorylated CaMKII and the phosphatase activity of PP1.

stably maintained.

**3.4.1.2 Synaptic tagging hypothesis** 

synapse and the nucleus in both directions. **3.4.1.3 mRNA transport into the dendrites** 

maintenance of enhanced AMPAR mediated postsynaptic response.

The tyrosine kinases Src and Fyn indirectly affect LTP by modulating NMDAR function. These Src family of tyrosine kinases can alter NMDAR function by phosphorylating GluN2A and 2B subunits, thereby relieving a basal zinc inhibition of the NMDAR. Phosphorylation of GluN2A or 2B thus potentiates the current through NMDAR complex. The increase in calcium concentration thus produced can contribute to the process of LTP.

Recent studies indicate that another subclass of glutamate receptors, the metabotropic glutamate receptors (mGluRs) are also involved in LTP induction (Bashir, 1993). In addition to activating ion channel-linked receptors, glutamate activates G protein-coupled metabotropic receptors which exist in eight different types labeled mGluR1 to mGluR8 which are classified into groups I, II, and III. Receptor types are grouped based on receptor structure and physiological activity. mGluR subtypes 1 and 5 (group I mGluRs) are positively coupled to phospholipase C (PLC), and trigger elevations in intracellular inositol triphosphate (IP3) and diacylglycerol (DAG), followed by mobilization of Ca2+ and activation of PKC (Benquet, 2002). Group I mGluRs are known to modulate the function of NMDAR by binding to PDZ proteins near to NMDAR (Yu, 1997). The activation of mGluRs, especially mGluR5 is involved in the induction of large amplitude or long-lasting late phase LTP of AMPAR-mediated transmission induced by strong or repeated stimulation protocols (Anwyl, 2009).

#### **3.4.1.1 Maintenance of LTP**

While LTP induction involves enhancement of synaptic efficacy largely by the biochemical events of E-LTP, the long term maintenance of the potentiated state demands for stable and self-sustaining biochemical mechanisms. In the dynamic milieu of the cell where most changes are reversible, stable alterations can be brought about by changes in the size of molecular pools that are dynamically maintained or by establishment of cyclic pathways which can maintain themselves. Increased exocytosis of AMPARs to the synaptic membrane could increase the size of the AMPAR pool in the synapse thereby increasing the response of the synapse. Phosphorylation of AMPARs leads to an increase in the pool of AMPARs with increased conductivity. However sustained maintenance of the larger pools requires adjustments in the kinetics of the pathways that influence these pools. One of the molecules that had been viewed as a candidate for maintenance of the stable state is CaMKII. Theoretical analysis indicates that the pool of CaMKII molecules in the special chemical environment of the PSD acts as a bistable switch. According to this model, the activity level of kinases and phosphatases determine which kind of synaptic plasticity, LTP or LTD is induced. A switch of this kind turns on, when a threshold number of Thr286 sites on the kinase are phosphorylated. Thr286-autophosphorylation converts CaMKII to an autonomously active 'on' state. The 'on' state of the switch can last for very long periods, because the kinase acts faster than the PSD phosphatase on Thr286 sites (Lisman, 2002). In the early phase of LTP, phospho-CaMKII generated will be more due to the fast activity of the kinase. Activated form of CaMKII can bind to the GluN2B subunit of the NMDAR as described earlier. This binding leads to saturation of CaMKII at very low concentration of ATP and thereby stabilizes the activity of the kinase against variations in the concentrations of ATP at synapses (Pradeep et al., 2009). This binding also leads to reduction in the rates of the phosphorylation and dephosphorylation reactions, resulting in a reduction in the amount of ATP consumed while running the simultaneous kinase and phosphatase reactions. Thereby this biochemical mechanism permits the functioning of the kinase-

The tyrosine kinases Src and Fyn indirectly affect LTP by modulating NMDAR function. These Src family of tyrosine kinases can alter NMDAR function by phosphorylating GluN2A and 2B subunits, thereby relieving a basal zinc inhibition of the NMDAR. Phosphorylation of GluN2A or 2B thus potentiates the current through NMDAR complex. The increase in calcium concentration thus produced can contribute to the process of LTP. Recent studies indicate that another subclass of glutamate receptors, the metabotropic glutamate receptors (mGluRs) are also involved in LTP induction (Bashir, 1993). In addition to activating ion channel-linked receptors, glutamate activates G protein-coupled metabotropic receptors which exist in eight different types labeled mGluR1 to mGluR8 which are classified into groups I, II, and III. Receptor types are grouped based on receptor structure and physiological activity. mGluR subtypes 1 and 5 (group I mGluRs) are positively coupled to phospholipase C (PLC), and trigger elevations in intracellular inositol triphosphate (IP3) and diacylglycerol (DAG), followed by mobilization of Ca2+ and activation of PKC (Benquet, 2002). Group I mGluRs are known to modulate the function of NMDAR by binding to PDZ proteins near to NMDAR (Yu, 1997). The activation of mGluRs, especially mGluR5 is involved in the induction of large amplitude or long-lasting late phase LTP of AMPAR-mediated transmission

While LTP induction involves enhancement of synaptic efficacy largely by the biochemical events of E-LTP, the long term maintenance of the potentiated state demands for stable and self-sustaining biochemical mechanisms. In the dynamic milieu of the cell where most changes are reversible, stable alterations can be brought about by changes in the size of molecular pools that are dynamically maintained or by establishment of cyclic pathways which can maintain themselves. Increased exocytosis of AMPARs to the synaptic membrane could increase the size of the AMPAR pool in the synapse thereby increasing the response of the synapse. Phosphorylation of AMPARs leads to an increase in the pool of AMPARs with increased conductivity. However sustained maintenance of the larger pools requires adjustments in the kinetics of the pathways that influence these pools. One of the molecules that had been viewed as a candidate for maintenance of the stable state is CaMKII. Theoretical analysis indicates that the pool of CaMKII molecules in the special chemical environment of the PSD acts as a bistable switch. According to this model, the activity level of kinases and phosphatases determine which kind of synaptic plasticity, LTP or LTD is induced. A switch of this kind turns on, when a threshold number of Thr286 sites on the kinase are phosphorylated. Thr286-autophosphorylation converts CaMKII to an autonomously active 'on' state. The 'on' state of the switch can last for very long periods, because the kinase acts faster than the PSD phosphatase on Thr286 sites (Lisman, 2002). In the early phase of LTP, phospho-CaMKII generated will be more due to the fast activity of the kinase. Activated form of CaMKII can bind to the GluN2B subunit of the NMDAR as described earlier. This binding leads to saturation of CaMKII at very low concentration of ATP and thereby stabilizes the activity of the kinase against variations in the concentrations of ATP at synapses (Pradeep et al., 2009). This binding also leads to reduction in the rates of the phosphorylation and dephosphorylation reactions, resulting in a reduction in the amount of ATP consumed while running the simultaneous kinase and phosphatase reactions. Thereby this biochemical mechanism permits the functioning of the kinase-

induced by strong or repeated stimulation protocols (Anwyl, 2009).

**3.4.1.1 Maintenance of LTP** 

phosphatase switch in an energy efficient manner (Cheriyan et al., 2011). Activated CaMKII can function in enhancing AMPAR currents and its recruitment. This will contribute to the maintenance of enhanced AMPAR mediated postsynaptic response.

Fig. 4. CaMKII-phosphatase bistable switch model. Continuous interconversion between CaMKII and phospho-CaMKII is catalysed by the kinase activity of Thr286 autophosphorylated CaMKII and the phosphatase activity of PP1.

Activated PKA can also contribute to the maintenance of LTP by involving in a selfsustaining mechanism, in addition to its role in promoting AMPAR exocytosis. As described earlier, activated PKA can alter gene expression via cAMP-PKA-CREB pathway. One gene activated by CREB encodes a ubiquitin hydrolase, a component of a specific ubiquitin protease that leads to the regulated proteolysis of the regulatory subunit of PKA. This results in persistent activity of PKA, leading to persistent phosphorylation of PKA substrates such as CREB, MAPK, etc., thereby completing a self-sustaining cycle that can be stably maintained.

#### **3.4.1.2 Synaptic tagging hypothesis**

L-LTP requires *de novo* protein synthesis. The long lasting activity changes require nuclear transcription followed by delivery of newly synthesized proteins to the synapse to yield synaptic remodeling. Newly synthesized proteins delivered by non-directed transport from the cell body must be captured locally at the activated synapse in order to function in an input-specific manner (Doyle, 2011). For this, the activated synapse requires a local signal that allows it to capture proteins or mRNAs for protein-synthesis-dependent LTP or LTD. This process has been termed *synaptic tagging.* Based on this proposal, synaptic activity generates a tag, which "captures" the plasticity-related proteins (PRPs) derived outside of synapses (Lu et al., 2011). These findings indicate a tight and extensive dialogue between the synapse and the nucleus in both directions.

#### **3.4.1.3 mRNA transport into the dendrites**

mRNA localization to the synapses depends on synaptic activity and the mechanism behind this transport is largely a mystery. This transport mechanism is highly complex and involves multiple mRNA binding proteins. This process can be divided into different stages, (1) the presence of cis-acting localization elements (LEs) or zipcodes generally located in the 3'-untranslated region (3'-UTR) of localized transcripts, (2) the recognition of these signals by trans-acting RNA-binding proteins (RBPs), (3) the assembly of RBPs and their cargo

Molecular Mechanisms in Synaptic Plasticity 307

Although a vast majority of studies of NMDAR dependent LTP have been conducted, there are also a few mechanisms that are independent of NMDAR that have been studied. Following section will briefly describe molecular mechanisms of NMDAR independent

NMDAR-independent forms of LTP also can be induced at the Schaffer collateral pathway in CA1. This allows for a comparison of two different types of LTP at the same synapse. NMDAR-independent LTP in CA1 can be elicited by use of four and a half seconds, 200 Hz stimuli separated by five seconds. LTP induced by this stimulation protocol is insensitive to NMDAR selective antagonist such as APV. 200 Hz LTP was shown to be blocked by nifedipine (Grover and Teyler, 1990), a voltage gated calcium channel (VGCC) blocker. This observation led to the conclusion that 200 Hz-LTP stimulation elicits sufficiently large and prolonged membrane depolarization, resulting in the opening of voltage dependent calcium channels, to trigger elevation of postsynaptic calcium sufficient to trigger LTP. It is also reported that L-type Ca2+ channel-dependent synaptic plasticity significantly contributes to

NMDAR-independent LTP at the Schaffer collateral pathway in CA1 can also be induced by the bath application of the K+ channel blocker tetraethylammonium (TEA) (TEA-LTP) (Aniksztejn and Ben-Ari, 1991) and is referred to as LTPk. This nonspecific potassium channel blocker can cause membrane excitability. Like 200 Hz-LTP, TEA-LTP is insensitive to NMDAR antagonists, and is blocked by blockade of voltage sensitive calcium channels. The induction of LTPk is dependent on synaptic activity, as its induction is blocked by AMPAR antagonists. Similar to 200 Hz LTP, the current model for TEA-LTP is that synaptic depolarization via glutamate receptor activation, augmented by the hyperexcitable membrane due to K+ channel blockade, leads to a relatively large and prolonged membrane depolarization. This leads to the triggering of LTP through postsynaptic calcium influx via

A good model system for studying NMDAR-independent LTP is the mossy fiber inputs into CA3 pyramidal neurons. The mossy fiber synapses are unique, large synapses with unusual

Early protocols for the induction of LTP in cultures of dissociated hippocampal neurons comprised repetitive high frequency presynaptic stimulation (HFS-LTP) as mentioned above. In some cases this was coupled with postsynaptic depolarization and in others cultures were preincubated with blockers of different channels. High frequency stimulation activates only a small fraction of synapses, making it difficult to detect molecular and cellular changes associated with LTP. Most biochemical analysis and imaging studies require a high proportion of synapses to be potentiated. Therefore, a range of strategies

presynaptic specializations. The mechanism will be described in later section (3.5.).

spatial learning in the behaving mouse (Moosmang et al., 2005).

**3.4.2 NMDAR-independent mechanisms** 

**3.4.2.2 Tetra-Ethyl-Ammonium LTP** 

**3.4.2.3 Mossy fiber LTP in CA3** 

**3.4.3 Chemical LTP (Chem-LTP)** 

forms of LTP.

the VGCCs.

**3.4.2.1 200 Hz LTP** 

RNAs into transport ribonucleo-protein particles (RNPs) as a functional complex, (4) the translocation of transport RNPs along the microtubule (MT) cytoskeleton to their final destination at synapses in a translationally repressed state, (5) the anchoring of these particles at or underneath activated synapses in a translationally repressed state and finally (6) the activation of translation of the localized mRNAs (Doyle, 2011).

One of the specific immediately expressed candidate gene is activity-regulated cytoskeletonassociated protein (Arc). Newly synthesized Arc mRNA is targeted rapidly to synapses that have recently undergone specific forms of synaptic activity where it is locally translated. Targeting of *Arc* mRNA depends on NMDAR activity. An increase in Arc expression promotes stable expansion of the F-actin network in dendritic spines, which is believed to underlie morphological enlargement of the synapse and stable LTP (Bramham, 2010).

#### **3.4.1.4 Spine enlargement**

Most excitatory synapses in the brain terminate on dendritic spines. Spines are specialized perturbations on dendrites that contain PSD. The PSD includes receptors, channels and signaling molecules that couple synaptic activity with postsynaptic biochemistry. Spines provide a closed compartment that allows rapid changes in the concentrations of signaling molecules, such as calcium, and hereby make efficient responses to inputs possible. Longterm changes in spine morphology could contribute to the modulation of synaptic transmission that occurs in LTP. Shortening or widening the neck of a spine affects calcium influx into the dendrite. Spine enlargement depends on the structure of cytoskeletal filaments. Actin filaments of microfilaments are in close association with PSD. Reorganization of actin filament contributes to the spine enlargement process in LTP. The AMPA class of glutamate receptors has been found to have a stabilizing effect on spine morphology. Rho GTPases and their downstream effectors have an important role in regulating the cytoskeleton, and consequently in regulating spine and dendritic morphology, in response to extracellular stimulation. AMPAR activation by spontaneous glutamate release at synapses is sufficient to maintain dendritic spines (Lamprecht & LeDoux, 2004*)*.

#### **3.4.1.5 Presynaptic mechanisms**

Activation of both pre and postsynaptic sites are necessary for the generation of LTP on the basis of Hebbian theory. Neurotransmitter release is one of the presynaptic mechanisms eliciting the induction of LTP. An increase in neuro-transmitter release can be observed together with the postsynaptic mechanisms. This is due to the activation of presynaptic terminals by some factors released by the postsynaptic compartment or cell (Williams et al., 1989). A prominent candidate for such a messenger is arachidonic acid or one of its metabolites, because these compounds can readily cross cell membranes. This can be generated by the degradation of phospholipids by the enzyme phospholipase A2, a calcium dependent enzyme (Bliss, 1990). Nitric oxide (NO) is another retrograde messenger produced by Ca2+/CaM activated nitric oxide synthase (NOS), which can activate the synthesis of cyclic GMP presynaptic terminal by activating two NO-sensitive guanylyl cyclases (NO-GCs) (NO-GC1 and NO-GC2) leading to increased neurotransmitter release. The physiological consequences of increase in NO/cGMP and the associated cellular mechanisms involved are not well understood.

#### **3.4.2 NMDAR-independent mechanisms**

Although a vast majority of studies of NMDAR dependent LTP have been conducted, there are also a few mechanisms that are independent of NMDAR that have been studied. Following section will briefly describe molecular mechanisms of NMDAR independent forms of LTP.

#### **3.4.2.1 200 Hz LTP**

306 Neuroscience – Dealing with Frontiers

RNAs into transport ribonucleo-protein particles (RNPs) as a functional complex, (4) the translocation of transport RNPs along the microtubule (MT) cytoskeleton to their final destination at synapses in a translationally repressed state, (5) the anchoring of these particles at or underneath activated synapses in a translationally repressed state and finally

One of the specific immediately expressed candidate gene is activity-regulated cytoskeletonassociated protein (Arc). Newly synthesized Arc mRNA is targeted rapidly to synapses that have recently undergone specific forms of synaptic activity where it is locally translated. Targeting of *Arc* mRNA depends on NMDAR activity. An increase in Arc expression promotes stable expansion of the F-actin network in dendritic spines, which is believed to

Most excitatory synapses in the brain terminate on dendritic spines. Spines are specialized perturbations on dendrites that contain PSD. The PSD includes receptors, channels and signaling molecules that couple synaptic activity with postsynaptic biochemistry. Spines provide a closed compartment that allows rapid changes in the concentrations of signaling molecules, such as calcium, and hereby make efficient responses to inputs possible. Longterm changes in spine morphology could contribute to the modulation of synaptic transmission that occurs in LTP. Shortening or widening the neck of a spine affects calcium influx into the dendrite. Spine enlargement depends on the structure of cytoskeletal filaments. Actin filaments of microfilaments are in close association with PSD. Reorganization of actin filament contributes to the spine enlargement process in LTP. The AMPA class of glutamate receptors has been found to have a stabilizing effect on spine morphology. Rho GTPases and their downstream effectors have an important role in regulating the cytoskeleton, and consequently in regulating spine and dendritic morphology, in response to extracellular stimulation. AMPAR activation by spontaneous glutamate release at synapses is sufficient to maintain dendritic spines (Lamprecht &

Activation of both pre and postsynaptic sites are necessary for the generation of LTP on the basis of Hebbian theory. Neurotransmitter release is one of the presynaptic mechanisms eliciting the induction of LTP. An increase in neuro-transmitter release can be observed together with the postsynaptic mechanisms. This is due to the activation of presynaptic terminals by some factors released by the postsynaptic compartment or cell (Williams et al., 1989). A prominent candidate for such a messenger is arachidonic acid or one of its metabolites, because these compounds can readily cross cell membranes. This can be generated by the degradation of phospholipids by the enzyme phospholipase A2, a calcium dependent enzyme (Bliss, 1990). Nitric oxide (NO) is another retrograde messenger produced by Ca2+/CaM activated nitric oxide synthase (NOS), which can activate the synthesis of cyclic GMP presynaptic terminal by activating two NO-sensitive guanylyl cyclases (NO-GCs) (NO-GC1 and NO-GC2) leading to increased neurotransmitter release. The physiological consequences of increase in NO/cGMP and the associated cellular

underlie morphological enlargement of the synapse and stable LTP (Bramham, 2010).

(6) the activation of translation of the localized mRNAs (Doyle, 2011).

**3.4.1.4 Spine enlargement** 

LeDoux, 2004*)*.

**3.4.1.5 Presynaptic mechanisms** 

mechanisms involved are not well understood.

NMDAR-independent forms of LTP also can be induced at the Schaffer collateral pathway in CA1. This allows for a comparison of two different types of LTP at the same synapse. NMDAR-independent LTP in CA1 can be elicited by use of four and a half seconds, 200 Hz stimuli separated by five seconds. LTP induced by this stimulation protocol is insensitive to NMDAR selective antagonist such as APV. 200 Hz LTP was shown to be blocked by nifedipine (Grover and Teyler, 1990), a voltage gated calcium channel (VGCC) blocker. This observation led to the conclusion that 200 Hz-LTP stimulation elicits sufficiently large and prolonged membrane depolarization, resulting in the opening of voltage dependent calcium channels, to trigger elevation of postsynaptic calcium sufficient to trigger LTP. It is also reported that L-type Ca2+ channel-dependent synaptic plasticity significantly contributes to spatial learning in the behaving mouse (Moosmang et al., 2005).

#### **3.4.2.2 Tetra-Ethyl-Ammonium LTP**

NMDAR-independent LTP at the Schaffer collateral pathway in CA1 can also be induced by the bath application of the K+ channel blocker tetraethylammonium (TEA) (TEA-LTP) (Aniksztejn and Ben-Ari, 1991) and is referred to as LTPk. This nonspecific potassium channel blocker can cause membrane excitability. Like 200 Hz-LTP, TEA-LTP is insensitive to NMDAR antagonists, and is blocked by blockade of voltage sensitive calcium channels. The induction of LTPk is dependent on synaptic activity, as its induction is blocked by AMPAR antagonists. Similar to 200 Hz LTP, the current model for TEA-LTP is that synaptic depolarization via glutamate receptor activation, augmented by the hyperexcitable membrane due to K+ channel blockade, leads to a relatively large and prolonged membrane depolarization. This leads to the triggering of LTP through postsynaptic calcium influx via the VGCCs.

#### **3.4.2.3 Mossy fiber LTP in CA3**

A good model system for studying NMDAR-independent LTP is the mossy fiber inputs into CA3 pyramidal neurons. The mossy fiber synapses are unique, large synapses with unusual presynaptic specializations. The mechanism will be described in later section (3.5.).

#### **3.4.3 Chemical LTP (Chem-LTP)**

Early protocols for the induction of LTP in cultures of dissociated hippocampal neurons comprised repetitive high frequency presynaptic stimulation (HFS-LTP) as mentioned above. In some cases this was coupled with postsynaptic depolarization and in others cultures were preincubated with blockers of different channels. High frequency stimulation activates only a small fraction of synapses, making it difficult to detect molecular and cellular changes associated with LTP. Most biochemical analysis and imaging studies require a high proportion of synapses to be potentiated. Therefore, a range of strategies

Molecular Mechanisms in Synaptic Plasticity 309

GluR2–GluR3 heteromeric complexes. An activity dependent synaptic delivery of GluR2 has been shown during the induction of LTP in PF-PC synapses. This activity driven process involves NO-mediated binding of *N*-ethylmaleimide sensitive factor (NSF) to GluR2. In PF LTP, GluR2 synaptic delivery is also facilitated by dephosphorylation of GluR2 at Ser880.

Fig. 5. Cellular anatomy of the cerebellum. Adapted from Ramnani, 2006

Spinal LTP has been demonstrated in different areas of the spinal cord. The ventral and the superficial dorsal horn, Wide Dynamic Range (WDR) neurons and superficial neurons in the spinal cord that project to the parabrachial area in the brain stem are some of the sites where LTP has been demonstrated. It has been suggested that the generation of LTP in spinal cord may be one mechanism, whereby acute pain may be transformed into a chronic pain state. LTP in superficial spinal dorsal horn involves simultaneous activation of multiple receptors like the NMDAR, the Neurokinin 1 (NK-1) receptor for substance P and mGluRs. This LTP is likely to occur in both the sensory and the affective pain pathways. LTP in deep spinal WDR neurons have a pivotal role in transmission of painful inputs. As with LTP in the superficial spinal cord, activation of the ionotrophic glutamate receptors (AMPA and NMDA subtypes) and the NK1 receptor seems crucial for the induction of LTP in deep

**3.5.2 LTP in spinal cord** 

WDR neurons (Rygh et al, 2005).

were applied to chemically induce LTP (Chem-LTP). Chem-LTP is an alternative to high frequency stimulation and has the advantage that it can activate all the cells in the culture. One example of Chem-LTP is mentioned below.

### **3.4.3.1 Forskolin/rolipram-induced LTP**

Forskolin/rolipram-induced LTP was predominantly used in slice cultures; it can also be applied for dissociated hippocampal neuronal cultures. This form of chemically induced, highly sensitive plasticity state is based on the increase of intracellular cAMP levels by the application of the adenylyl cyclase activator forskolin (50 µM) and the phosphodiesterase inhibitor rolipram (0.1 µM) in Mg2+ and 2-Cl-adenosine free artificial cerebrospinal fluid for 16 min (Otmakhov, 2004). This induction procedure is bypassing the need for synaptic activation, and by raising cAMP concentration directly activates PKA and signaling pathways that underlie synaptic plasticity. However, froskolin/rolipram-LTP still require NMDAR activation and involve the recruitment of CaMKII to dendritic spines (Molnar, 2011).
