Figure 1: The 3 pathways in the hippocampus where LTP is observed which are the scaffer collateral pathway, the mossy fiber pathway and the perforant pathway (Rongo et al, 2002).
NMDA receptor subunits contain 3 membrane spanning regions and a C-terminal tail of various lengths, which produces the intracellular portion of the receptor (Strack et al, 1998). NMDA receptors are heteromeric complexes consisting of NR1, NR2 subunits and less commonly NR3 subunits. NR2B can combine with NR1 to form a heteromeric channel or with NR2A and NR1 and this NR2B is shown to be vital for the sustained increase in T286 phosphorylation of αCaMKII in the postsynaptic density and therefore plays an important role in LTP induction and learning (Zhou et al, 2007).
A CaMKII inhibitor applied postsynaptically after induction of LTP did not prevent the mechanism of LTP (Otmakhov et al, 1997). This demonstrates that the memory storage could be transferred to a downstream event. The inability to prevent LTP maintenance could be due to low PP1 activity that it may take an hour for CaMKII to be in a dephosphorylated state. When this inhibitor was applied before initiation of LTP, LTP was prevented.
NMDA receptors are expressed in CA1 neurons and in response to glutamate release from the active presynaptic cell, NMDA receptors open and depolarisation of the postsynaptic cell occur. The presynaptic and postsynaptic cells fire at the same time to cause glutamate entry into the synaptic cleft and the release of glutamate by the presynaptic cells happens at the exact time when the postsynaptic cell is depolarised. The opening of NMDA-R results in transient Ca2+ entry to the postsynaptic cell to elicit LTP.This stimulates CaMKII which autophosphorylates and in turn modulates synaptic plasticity.
Figure 2: Here this figure shows that the fluorescence of calcium green is mixed with a red colour which suggests the greatest calcium levels. But with AP5, a well known NMDA antagonist, the green and the red colour is not to be seen. This suggests that activation of NMDA receptors induces calcium entry into the dendritic spine (Lisman et al, 2002).
Studies show that in interneurons in striatum radiatum, LTP is induced even in α-CaMKII T286A-mutant mice (Lamsa et al, 2007). This demonstrated that the α-subunit does not need to be autophosphorylated, which is normally observed in CA1 pyramidal cells.
Studies demonstrate that the removal of intermediate and medial hyperstriatum ventrale/mesopallium in the forebrain which is major memory storage in untrained and trained chicks showed that the total amount of T286-autophosphorylated α-CaMKII was more in good learners than in poor learners (Solemania et al, 2005). The rise in autophosphorylated α-CaMKII is observed primarily as the chicks learn by visual imprinting and not so much when they are just subjected to only the stimulus. Visual imprinting is a rapid learning process whereby a certain individual carefully analyses the features of the object and recognises it after a given length of time. This form of imprinting can induce cellular and synaptic changes in the forebrain. As the strength of learning increases, α-CaMKII activity rises and this shows that α-CaMKII is vital for information storage. Furthermore the autophosphorylated CaMKII increases an hour after the end of training and not at 24 hours. Hence this proves that memory is formed at 1 hour and not at 24 hours after training.
Figure 3: In this figure after one hour of training, there was significantly more phosphorylated CaMKII in good learners than in intermediate learners, poor learners and untrained chicks in the intermediate and medial hyperstriatum ventrale/mesopallium which is a forebrain memory storage site. Lanes 1 -4 correspond to good learners,5-8 from intermediate learners,9-12 from poor learners and 13-16 from untrained chicks(Solomonia et al,2005).
The autophosphorylation of αCaMKII at T286 in spines in the lateral amygdala is vital for fear conditioning and impairment of CaMKII activation disrupts synaptic plasticity and fear conditioning in the lateral amygdala (Rodrigues et al, 2004).
Pharmacological agents that suppress CaMKII or calmodulin inhibit LTP when added to recordings from hippocampal slices (Otmakhov et al, 1997). Mice that lacked the α-isoform of CaMKII demonstrated reduced LTP in the hippocampus (Silva et al, 1992). Mice that had no α-CaMKII exhibited dysfunction in spatial memory tasks such as the Morris water test but did well on tasks testing on non-spatial associative memory. These experiments showed that LTP is a process for spatial learning and memory.CaMKII autophosphorylation causes its translocation to the synapse, which could sustain its response to Ca2+ entry. The induction of LTP appears to cause increased presynaptic release of glutamate and the postsynaptic response of AMPA-type glutamate receptors which leads to enhanced synaptic responses (Soderling et al, 1993). Stimulation or autophosphorylation of CaMKII results in binding of the enzyme with the cytoplasmic C-terminal domain of the NMDA receptor NR2B subunit located in the postsynaptic density. Furthermore the kinase also combines with the NR1 subunit via autophosphorylation and as autophosphorylation occurs, the enzyme interacts with many regions on the NMDA receptor and binds with it powerfully (Lisman et al, 2002). The importance of CaMKII autophosphorylation inducing tight interactions of the kinase to the NMDA receptor is that it places it in an ideal position to control synaptic function. Plasticity in the dorsal horn of the spinal cord increases the interaction of CaMKII with NMDA receptors. The interaction of CaMKII to the NMDA receptor controls enzyme activity. After a subunit interacts with NR2B, it continues to be active in the absence of autophosphorylation even after dissociation of the Ca2+ /calmodulin complex. This is due to the sequence on NR2B that the kinase binds to have a resemblance to the autoinhibitory domain that interacts with the T site. In addition NR2B can bind to the T site, which will keep the autoinhibitory gate open even after Ca2+ /calmodulin dissociates. This active state lasts for seconds to minutes and cannot be reversed by phosphatase activity. The interaction of an unphosphorylated subunit to the NMDA receptor results in increased affinity of calmodulin for the kinase and induces ‘trapping’. The trapping effect retains the kinase at the synapse because it stops a secondary autophosphorylation of the domain that interacts with calmodulin, which would result in kinase dissociation from synaptic sites (Shen et al, 2000).
Stimulation of the subunit bound to the NMDA receptor would result in additional autophosphorylation around the ring because interaction of the Ca2+ /calmodulin to a surrounding subunit would be enough to induce autophosphorylation.
NR2B binding is initiated by Ca2+ /calmodulin binding to CaMKII or by T286 autophosphorylation (Bayer et al, 2001). The cytoplasmic domain of NR2B binds about six times more [32 P-T286] CaMKIIα than NR2A does (Strack et al, 1998).This high affinity binding of CaMKII with a 50 amino acid domain in the cytoplasm of NR2B than with NR2A shows that this predominant subunit affects developmental plasticity and have a major impact on learning and memory. Synaptic activity, which activates dendritic calcium influx, induces T286 autophosphorylation in CaMKII and induces binding specifically to residues 1260-1309 of the NR2B subunit (Strack et al, 1998).
The increased and stable binding of autophosphorylated CaMKII to the NR1 and NR2 subunits caused production of new anchoring assemblies for additional AMPA receptors and activates the delivery of AMPA receptors to the membrane (Wu et al, 1996). The binding of CaMKII and NR2B affects the rate of attainment of spatial learning in the hippocampus (Zhou et al, 2007). When CaMKII translocates to the postsynaptic density, the AMPA receptor GluR1 is phosphorylated on serine 831 by this enzyme and this potentiates channel function and increases channel conductance. This mechanism is vital for learning and memory and this phosphorylation occurs during LTP and after half an hour of induction of LTP there is enhanced receptor phosphorylation and this phosphorylation is increased for a further one hour. Therefore CaMKII activity persists for an hour and its activity is needed for controlling GluR1 phosphorylation.
Figure 4: A transgenic mice expressing LBD G521R- cNR2B derived from the fusion of ligand-activated carboxy terminal NR2B with a tamoxifen-dependent mutant of the estrogen receptor ligand binding domain LBD G521R .Activation of LBD G521R- cNR2B protein induces its binding to CaMKII/NR2B binding. This in turn reduces T286 phosphorylation which in this figure shows that in the tamoxifen treated transgenic mice(tg/tam),ser 831 phosphorylation of GluR1 is significantly reduced in the postsynaptic density(Zhou et al,2007).
The mechanism of CaMKII autophosphorylation allows the activity of CaMKII to last from seconds to minutes. This length of time could be adequate to trigger possible downstream events that are the actual molecular memory. CaMKII is also needed for the delivery of AMPAR to ‘silent synapses’ which are synapses that do not have AMPARs (Fink et al, 2002), therefore this seems to be a vital role in LTP and there seems also to be a deduction in failures. Failure rate is the probability that a presynaptic action potential will fail to create a postsynaptic action potential. Furthermore the application of CaMKII decreases synaptic failures, which demonstrates that CaMKII will create functional contacts from silent synapses (Lledo et al, 1995). The number of AMPA receptors at synapses is regulated by two mechanisms, which is the (1) trafficking process that regulates the delivery of and removal of AMPA receptors from the plasma membrane and the (2) anchoring process that retains these AMPA receptors at the synapse. These two processes are mediated by CaMKII and therefore enhances synaptic transmission. The stable interactions of autophosphorylated CaMKII to the NMDA receptor creates a structural process that induces the insertion of AMPA receptor proteins into the postsynaptic density and the consequent anchoring of further AMPA receptors.
Memory storage relies on the phosphorylation state of the CaMKII switch but not on the number of AMPA receptors. The phosphorylation of existing AMPA receptors, insertion of additional AMPA receptors into unfilled anchoring sites and production and filling of new anchoring sites contribute to synaptic strength and strengthening of LTP. If AMPA receptor trafficking was impaired, LTP expression could still be initiated due to direct phosphorylation of remaining receptors. Alpha-CaMKII can interact with densin-180, which is a synaptic adhesion molecule. This molecule consists of a PDZ domain, which permits the CaMKII complex to interact with other postsynaptic density proteins. This scaffold structure allows CaMKII to be located near the AMPARs, which therefore allows LTP induction at synapses.
Actin filaments are involved in promoting a link between the CaMKII/actinin complex and the protein 4.1/SAP97 complex (GluR1-binding proteins). Furthermore actin plays a vital role in placing AMPA receptors at synapses and actin depolymerisation induces a rapid decrease in AMPA-receptor mediated transmission and synaptic AMPA receptors which would then lead to a decrease in LTP.These actin filaments are also one of the reasons why CaMKII concentration is high in the postsynaptic density (Lisden et al, 2002). Synaptic strength needs to be bidirectionally altered for synapses to store memory. Because CaMKII phosphorylation increases the interactions between CaMKII and NMDA receptor, the actin-mediated strengthening could be reversed by CaMKII dephosphorylation. Studies show that mice having a knock in mutation of Ser 831 and Ser 845 in GluR1 to alanine, stops phosphorylation of GluR1 subunits at Ser 831 and hence hippocampal synaptic plasticity and spatial memory is impaired (Lee et al, 2003). GluR1 dephosphorylation by depotentiation techniques causes a decrease in CaMKII phosphorylation at T286.Depotentiation is a reversal of LTP induced by low frequency synaptic activation (Huang et al, 2001).
Neurons have many synapses, which are independently altered by LTP. This synaptic modification permits neurons to store huge amounts of information. It has been suggested that kinase functions as a switch. The stimulus-induced phosphorylation of kinase can potentiate a stable ‘on’ state even in the absence of a stimulus. The enzyme kinase in its ‘on’ state can phosphorylate itself and when the kinase is increasingly phosphorylated, the phosphatase becomes saturated which normally dephosphorylates the kinase. This permits the kinase to rephophorylate sites quicker than the phosphatase can dephosphorylate them. The function of this kinase switch relies on the interactions between kinase and phosphatase molecules, the informational state of the switch can still be stable even though there is protein turnover (Lisman et el, 2002). So the kinase can be used to store vast amounts of memory at synapses. CaMKII is dephosphorylated only by PP2A in the cytosol and PP1 in the postsynaptic density.
When CaMKII becomes hyperphosphorylated, the postsynaptic density PP1 (protein phosphatase 1) is saturated because the T286 sites are in greater concentrations than the Km of PP1.Within a ring of subunits, the six-threonine 286 sites are phosphorylated. When PP1 dephosphorylates a subunit, the dephosphorylated subunit interacts with a single Ca2+ /calmodulin molecule and this subunit gets phosphorylated by the surrounding phosphorylated subunit. This keeps the ring structure in an ‘on’ state (Lisman et al, 2002). The CaMKII acts as a bistable switch. The activity of PP1 in the postsynaptic density needs to be low for bistability to occur otherwise vast amounts of energy would be needed to rephosphorylate T286 sites that are also being dephosphorylated.
In the presence of protein turnover, the ability of the protein switch to store information is vital. Protein turnover occurs in about one month in the case of CaMKII.PP1 saturation by the hyperphosphorylated kinase molecules provides a role for communication between CaMKII and holoenzymes which prevents turnover. Holoenzymes do not phosphorylate each other and if an unphosphorylated holoenzyme replaced an increasingly phosphorylated holoenzyme, the unphosphorylated holoenzyme would be exposed to lower phosphatase activity. This low activity is due to surrounding phosphorylated holoenzymes saturating the phosphatase. This unphosphorylated holoenzyme would be phosphorylated rapidly and therefore this restores the initial state of the protein switch .The properties of the kinase enzyme and PP1 in the postsynaptic density generate a functional protein switch that has the ability to store vast amounts of information.
I finally conclude that CaMKII autophosphorylation is vital for memory attainment but not so important for its maintenance. CaMKII autophosphorylation induces NMDA receptors to bind to CaMKII and initiates LTP induction. CaMKII phosphorylates AMPA receptors, which contributes to mechanisms of synaptic plasticity and memory. The molecular properties of CaMKII allow it to sustain its activity from the initial stimulus and hence functions as a memory molecule.
Figure 5: This figure shows the proposed mechanism that underlies LTP which promotes learning and memory (Hell et al, 2005).
- Bayer KU De Koninck P Leonard AS Hell JW Schulman H (2001) Interaction with the NMDA receptor locks CaMKII in an active conformation. Nature. 411:801-805.
- Colbran RJ Brown AM (2004) Calcium/calmodulin-dependent protein kinase II and synaptic plasticity. Current Opinion in Neurobiology.14: 318-327.
- Esteban JA Shi SH Wilson C Nuriya M Huganir RL Malinow R (2003). PKA phosphorylation of AMPA receptor subunits controls synaptic trafficking underlying plasticity. Nature Neuroscience.6: 136-143.
- Fink CC Meyer T (2002) Molecular mechanisms of CaMKII activation in neuronal plasticity. Current Opinion in Neurobiology.12: 293-299.
- Hell JW (2005) Is CaMKII forgetful. Cell Science.1.
- Huang CC Liang YC Hsu KS (2001) Characterization of the Mechanism Underlying the Reversal of Long Term Potentiation by Low Frequency Stimulation at Hippocampal CA1 Synapses. Journal of Biological Chemistry. 276:48108-48117.
- Miyamoto E Fukunaga K (1996). A role of Ca2+/calmodulin-dependent protein kinase II in the induction of long-term potentiation in hippocampal CA1 area. Neuroscience Research. 24:117-122.
- Lamsa K Irvine EE Giese KP Kullmann DM (2007). NMDA receptor-dependent long-term potentiation in mouse hippocampal interneurons shows a unique dependence on Ca2+/ calmodulin-dependent kinases. Journal of Physiology. 584:885-894.
- Lee HK Takamiya K Han JS Man H., Kim CH., Rumbaugh G Yu S Ding L He C Petralia RS Wenthold RJ Gallagher M Huganir RL (2003) Phosphorylation of the AMPA receptor GluR1 subunit is required for synaptic plasticity and retention of spatial memory. Cell.112: 631-643.
- Lisman J Schulman H Cline H (2002) The molecular basis of CaMKII function in synaptic and behavioural memory. Nature Reviews Neuroscience.3: 175-190.
- Lledo PM Hjelmstad GO Mukherji S Soderling TR Malenka RC Nicoll RA (1995) Calcium/calmodulin-dependent kinase II and long-term potentiation enhance synaptic transmission by the same mechanism. Proceedings of the National Academy of Sciences of the United States of America.92: 11175-11179.
- Otmakhov N Griffith LC Lisman JE (1997) Postsynaptic inhibitors of calcium/calmodulin-dependent protein kinase type II block induction but not maintenance of pairing-induced long-term potentiation. Journal of Neuroscience.17: 5357-5365.
- Rodrigues SM Farb CR Bauer EP LeDoux JE Schafe GE (2004) Pavlovian Fear Conditioning Regulates Thr286 Autophosphorylation of Ca2+ /Calmodulin-Dependent Protein Kinase II at Lateral Amygdala Synapses. Journal of Neuroscience. 24:3281-3288.
- Rongo C (2002) A fresh look at the role of CaMKII in hippocampal synaptic plasticity and memory.BioEssays.24: 223-233.
- Shen K Teruel MN Connor JH Shenolikar S Meyer T (2000) Molecular memory by reversible translocation of calcium/calmodulin-dependent protein kinase II.Nature Neuroscience.3: 881-886
- Silva AJ Stevens CF Tonegawa S Wang Y(1992). Deficient hippocampal long-term potentiation in alpha-calcium-calmodulin kinase II mutant mice. Science.257: 201-206.
- Soderling TR (1993) Calcium/calmodulin-dependent protein kinase II: Role in learning and memory. Molecular and Cellular Biochemistry.127-128: 93-101.
- Solomonia RO Kotorashvili A Kiguradze T McCabe BJ Horn G (2005). Ca2+/calmodulin protein kinase II and memory: Learning-related changes in a localized region of the domestic chick brain. Journal of Physiology. 569:643-653.
- Strack S Colbran RJ (1998). Autophosphorylation-dependent targeting of calcium/calmodulin-dependent protein kinase II by the NR2B subunit of the N-methyl-D-aspartate receptor. Journal of Biological Chemistry. 273: 20689-20692.
- Wu GY Malinow R Cline HT (1996). Maturation of a central glutamatergic synapse (1996) Science. 274:972-976.
- Yang E Schulman H (1999) Structural examination of autoregulation of multifunctional calcium/calmodulin-dependent protein kinase II. Journal of Biological Chemistry. 274: 26199-26208.
- Zhou Y Takahashi E Li W Halt A Wiltgen B Ehninger D Li G.-D Hell JW Kennedy MB Silva AJ (2007) Interactions between the NR2B receptor and CaMKII modulate synaptic plasticity and spatial learning. Journal of Neuroscience.27: 13843-13853.