Radmila Mileusnic Abstract
The role of protein synthesis in long-term memory formation is still an area of intense scientific interest, which encompasses the study of mechanisms involved in gene expression and molecular mechanisms underlying synaptic plasticity. A number of low molecular weight compounds have been used to inhibit or enhance this fundamental cellular process. The pivotal role of protein synthesis in long-term memory formation suggest yet again that an understanding of how protein synthesis can be activated and regulated by events that ultimately lead to memory consolidation can lead to better understanding of the processes that keep our memories alive.
Protein synthesis allows the remarkable capacity of nervous system to modify its neuronal architecture as a consequence of learning. The variety of protein molecules involved in this process encompasses very different type of proteins, from transcription factors and enzymes involved in neurotransmitter metabolism, to cell adhesion molecules.
Today, one takes a great risk when questioning the role of protein synthesis in memory formation. However, some readers may find it unusual that there is often a sign of scepticism regarding this dogma.19,21 The main reason for the doubt is a paradoxical feature of proteins. Namely, proteins simultaneously encompass two rather different features: they are not stable molecules, yet they ensure continuity of our memories throughout our lifetime. Protein turnover replaces different types of proteins, including synaptic proteins, probably hundred times during our lifetime with no evident consequences for long-lasting synaptic modifications. Therefore, the question of how organisms store information for very long periods of time in spite of constant molecular turnover is one of the most captivating questions for many neuroscientists.
Bearing in mind the physical limits of a chapter, I have been 'forced' to omit many important papers that paved the way to a better understanding of the role of protein synthesis in memory formation. Thus, I would like to apologise to all of the zealous researchers whose name and work is not mentioned in this chapter.
The search for chemical substrates of learning and memory has a long history. Halstead46 was perhaps the first to suggest that 'engrams' might be stored in "template" protein molecules in nerve cells. Correlational studies of Hyden focused on RNA and protein changes induced by training procedures, and implicated RNA changes in memory formation.47-49 Although no replication of this type of experiment was ever reported from outside Hyden's laboratory, an unfortunate cul-de-sac stemming from this work was the hypothesis on memory transfer.52,71 As a consequence, the conceptual model of "specific memories stored in specific molecules" was abandoned rather quickly and research shifted from "memory molecules" to biochemical and molecular processes that might result in modified synaptic interactions, either through (a)
structural changes, which inevitably involve proteins as building blocks of synaptic membranes or (b) process changes, which would, again, involve synthesis of protein molecules such as enzymes, neurotransmitters, receptors, etc.
A reasonable test for the hypothesis that new proteins are necessary for long term memory was to substantially prevent the expression of new proteins using protein synthesis inhibitors. A vast literature has developed using different drugs and antibiotics that inhibit protein synthesis. Early studies10,11,17 showed that actinomycin D, an antibiotic that in low concentrations inhibits transcription without appreciably affecting DNA replication could produce impairments in learning and memory. Results were criticised because of the severe toxicity of the drug. However, this objection has been overcome with the introduction of less toxic drugs such as anisomycin, which inhibit peptide bond synthesis. The most convincing data on the importance of protein synthesis for long-term memory formation came from work of Flood et strated that the stronger the training and consequently the greater the amount of learning, the greater the duration of inhibition of protein synthesis had to be in order to produce amnesia.
However, there was some scepticism regarding the conclusion that treatment with protein synthesis inhibitors demonstrates that synthesis of new proteins is required for long-term memory. Importantly, it was not the concept that was in doubt, only whether antibiotics used as protein synthesis inhibitors demonstrate this notion (Table 1). Another reason for scepticism resided in the fact that, for example, puromycin produces effects on learning and memory that are qualitatively different from drugs such as acetocyloheximide, cycloheximide and anisomycin.24 Namely, Flexner and Flexner29 published data on the effect of NaCl which, when injected into the brains of mice some days after training, reversed the amnesia produced by puromycin given at the time of training, suggesting that puromycin did not prevented the long-term memory formation, but only its expression. On the other hand, Rosenbaum and colleagues,97 while replicating this unusual finding of the Flexner's found that NaCl did not have an effect on amnesia produced by acetocycloheximide. Moreover, many other drugs 'attenuated' the amnestic
Table 1. Inhibitors of protein synthesis in eucaryotes most often used in research into the role of proteins in long-term memory formation
Inhibitors of Protein Synthesis
Site of Action
Prokaryotes and Eucaryotes: binds tightly to duplex DNA by intercalating phenoxazine ring between neighbouring base pairs in DNA
Structural analogue of the 3'-terminal end of aminoacyl-tRNA. Enters aminoacyl site on the ribosome and is incorporated into the growing polypeptide chain that causes release of shortened polypeptide chains from the ribosome with puromycin attached.
Blocks peptide bond formation by binding to 60S ribosomal subunit. Concentration used for protein inhibition causes internucleosomal fragmentation and activates the stress-responsive pathways.
effect of protein synthesis inhibitors while not affecting the rate of protein synthesis.24 For example, Flood et al.32 found that amnesia induced by anisomycin could be attenuated by either amphetamine, strychnine or picrotoxin if these drugs were given at the time of training with no significant effect on the inhibition of protein synthesis induced by anisomycin. It seems that all these drugs act in this manner when their effect is opposed against the effect of amnestic agents such as protein synthesis inhibitors. The scepticism was nourished even further with the set of data showing that specific conditions such as stress, could bring back memories that were supposedly never formed.67 Thus, one inescapable question arose: what are the contributors, of central or peripheral nature, to the process of memory formation?
De Wied116 was among the first to suggest that hormones might influence the processes that underlie memory formation. It become obvious that the learning experience and the hormonal response have to occur within certain temporal limits for the hormonal response to influence the learning process and consequent memory formation. The list of hormones influencing learning processes is impressively long (see review by de Wied and Kovacs in this book). Many of them, such as vassopressine, affect learning through peripheral autonomic sequalae, while others, such as steroid hormones, modulate the process of gene expression.
The concept that memory storage is time-dependent process74 brought new light on the idea that some form of neural activity must underlie the setting down of memory traces and that while the experience of events and the memories formed may be continuous, the underlying molecular processes are clearly discontinuous with defined time courses. This conceptual framework of McGaugh introduced post-training treatments as a new experimental approach in studies of different phases of memory consolidation. Among the earliest published data supporting the idea that the underlying molecular processes in memory consolidation are discontinuous with defined time courses are studies on goldfish by Agranoff et al.3,5 They clearly demonstrate that animals injected intracranially with puromycin just before or within 30 min of a learning experience showed no problems with initial learning but had markedly impaired long-term memory. However, injections given more than 30 min after learning, did not prevent the formation of long-term memory. Thus, it was concluded that the most crucial difference between short- and long-term memory is that short-term memory is resistant to protein synthesis inhibitors while long-term memory is not (Fig. 1). From that time until today countless experiments confirmed that short-term memory is protein synthesis-independent, the formation of long-term memory is a time- and protein synthesis-dependent process, while the formation of long-lasting memory (consolidation of memory) is protein synthesis independent (for review see ref. 75).
The large body of evidence on the effect of antibiotics on brain protein synthesis can be summarised, according to Squire and Davis,106 into few basic findings: first, the effect of antibiotics on memory formation is a phenomenon found across different animal species and in wide variety of training paradigms (see also Strock and Welzl in this book). Second, in order to induce amnesia protein synthesis has to be inhibited up to a level of 90 - 95%, and finally, if inhibition is established 30 or more minutes after training, no amnesia develops. Despite the concise summary offered by Squire,105 these studies left memory researchers with the already mentioned paradox: if protein synthesis is necessary for long-term memory then how could any other molecule that does not affect protein synthesis reverse the amnesia that was induced by drugs such as anisomycin? This paradox lead to the development of research into the role of drugs, such as amphetamine, that could substantially alter neural processing67,98 and hormones, that could modify gene expression (for review see refs. 63, 72 and 73) thus altering the formation of long term-memories.
Changes in the rate of protein synthesis associated with learning have been extensively studied as potentially crucial factor in consolidating memories. Despite the fact that many experiments suggested that there may be a general increase in the rate of cerebral protein synthesis, incorporation of radiolabelled amino acids into brain proteins in trained animals relative to control ones in the period immediately following training4,23,24,95 appeared to be affected by stress responses and as such, may be due to the release of the hormones known to respond during stress25 rather than to consolidation. Moreover, the increased amino acid incorporation was often neither region-specific nor tissue specific - further evidence of a nonspecific response to activation.
The posttranslational modification of proteins attracted the attention of many laboratories because any process by which chemical groups can be added to or subtracted from the proteins very rapidly, resulting in major changes in the properties of the molecule, is likely to be utilised by neurons during the process of activity-dependent modifications.4,23 Two types of covalent modification, very different in nature, attracted the most attention: phosphorylation and glycosylation. The process of cAMP-induced protein phosphorylation59 offered fast, reliable, target-specific modification, an ideal candidate for signal transduction that could ultimately lead to modification of gene expression. Glycosylation, in contrast to phosphorylation, is a much slower process, but is capable of producing enormously heterogeneous and complex glycoprotein structures, making glycoproteins ideal candidates for intercellular interactions.
The first indication that protein phosphorylation was sensitive to behavioural manipulation came from the work of Machlus, Wilson and Glassman65 on nuclear proteins, but subsequently emphasis has been shifted toward synaptic and membrane proteins.14,38,55,62,100 It took neuroscience about 20 years after Wilson and Glassman's publication to 'rediscover' the role of activity-driven phosphorylation of transcription factors (TFs) as the crucial event in regulation of gene expression in the process of memory formation (for review see refs. 1, 18, 22, 69, 113 and 115).
The first indication that glycosylation of proteins is sensitive to behavioural manipulations came from work of Entingh et al. 7 who discovered changes in uridine metabolism following learning and suggested that uridine metabolites might be used as building blocks for glycopro-tein synthesis. Two research groups, one in Magdeburg68,88 and the other at The Open University91,92,94,110 took another approach to investigate the role of glycoproteins in memory formation. Namely, they introduced a new 'tool' into the field: 2-D-galactose as competitor of fucose and consequent protein glycolsylation.
The 'Local' Protein Synthesis
When Fisher and Litvak28 and Guiditta et al.42 published their findings showing that axo-plasmic proteins were labelled when isolated squid giant axon was incubated in vitro with radioactive amino acids, they set the stage for a number of studies indicating that translation might occur in cell compartments other than cell body. But, studies of glial-axonal protein transfer60'61'114 provided strong suspicions of local protein synthesis and led to the belief that proteins were synthesised in periaxonal glial cells and secondarily transferred to the sub-adjacent axon by intercellular transfer. This view was generalised to all axons and became a commonly accepted view, despite evidence that the giant axon contained all of the prerequisites for such protein synthesis. 1,43,44,50 The idea of local protein synthesis was brought to light for the second time with the isolation of mitochondrial fractions from brain which contained a substantial portion of sheared off nerve terminals resealed into osmotically sensitive particles named synaptosomes.20'45'87 This method allowed determination of whether presynaptic terminals contribute to protein synthesising activity8 of neuronal cell bodies. Using two different inhibitors of protein synthesis, cycloheximide (inhibitor of ribosomal protein synthesis) and chloramphenicol (inhibitor of mitochondrial protein synthesis), Yellin et al.117 confirmed that both cytoplasmic and mitochondrial systems of protein synthesis are present in synaptosomes.
The history of presynaptic protein synthesis is as turbulent as the history of protein synthesis. However, the convincing experimental evidence of Steward and Levy108 which showed the existence of synapse-associated polyribosomal complexes (SPRCs) selectively localised in distal processes beneath postsynaptic sites on the dendrites and recent data indicating that presynap-tic translational activity exists across different species cleared the clouds of suspicion and more importantly, lead to the belief that presynaptic protein synthesis is required for long-term plasticity changes.66
Although the present times are characterised by breath-taking technological developments, time-dependent involvement of cellular processes enabling formation of lasting memories (consolidation theory proposed by Müller and Pilzecker in 1900 and revisited by McGaugh in ref.
74) is still shaping research into protein synthesis. Unfortunately, the contribution of Glassman and his two stage hypothesis of molecular cascade in memory consolidation, which postulated two distinct waves of protein synthesis during which an activator protein synthesised during phase 1 will act as an activator of the genes coding for phase 2 proteins,39 is often forgotten.
One question that suddenly reemerged during the last four years and has lead to rather intensive arguments, that is the discussion of how stable long-lasting memories are after reactivation or retrieval. The experiments LeDoux and colleagues83,84, aimed to test the stability of memory trace in terms of activation of molecular events. They showed that the same manipulations that could cause amnesia after initial learning, such as inhibition of protein synthesis around the time of training, can also lead to memory loss right after reactivation or retrieval. In other words, LeDoux and colleagues argue that any attempt to access memory when it is consolidated will bring the memory trace into a labile state making it vulnerable to the effects of inhibition of the same cellular and molecular processes that were critical for the original consolidation.83,84 Hence the term reconsolidation entered the field yet again. The term reconsolidation as such is contradiction in adjecto. The problem is not of semantic nature but in the fact that anisomycin interfered with both new memories and reactivated memories, in other words, anisomycin rendered animals amnestic for the original learning task. On the other hand, the experiments of Morris, Anokhin, Sara and Taubenfeld contradicted the work of LeDoux's goup and showed that at the cellular level, NMDA receptors104 or transcription factor C/EBPß11 are not involved in retrieval of previously established memories. The stability of the memory trace during retrieval studied by the group of Anokhin,64 using protein synthesis inhibitors in association with a reminder procedure, showed that administration of anisomycin and cycloheximide induced the development of temporary amnesia whose duration gradually
Figure 2. The time-windows of sensitivity of protein synthesis to antibiotics. The progress in imaging techniques, which can identify brain regions active during acquisition and retrieval, combined with visualisation of neural activity and gene expression could help us to elucidate the problem of'reconsolidation'.
Figure 2. The time-windows of sensitivity of protein synthesis to antibiotics. The progress in imaging techniques, which can identify brain regions active during acquisition and retrieval, combined with visualisation of neural activity and gene expression could help us to elucidate the problem of'reconsolidation'.
declined as the interval between training and reminder increased. The same conclusion was drawn from the experiments of Sara.101,1 2 At a system level, hippocampus-dependent learning tasks have always been the 'darling' of research into memory consolidation. To identify whether hippocampal activity contributes to these processes independently, Riedel and colleagues used a novel method of inactivating synaptic transmission using a water-soluble antagonist ofAMPA/ kainate glutamate receptors and addressed the reconsolidation question by temporarily inactivating hippocampus prior to spatial memory test.90 Their findings indicated that hippocampal neural activity is necessary for both encoding and retrieval of spatial memory and for either trace consolidation or long-term storage. On the other hand, experiments of McGaugh showed that other brain areas, such as amygdala, may have an important function in memory consolidation after a learning task has taken place. 6
As for the novelty of the problem, it was in 1989 when Matthies68 actually 'broke' the concept of protein synthesis independent phase of memory consolidation and showed the existence of the second wave of protein synthesis, which occurs 4-6 hr after learning experience, using a brightness discrimination task in rats (Fig. 2). The bimodal feature of protein synthesis was found in another extensively studied task, the one-trial passive avoidance task in chicks.33 In spite of anatomical differences between rats and birds, the similarities in cellular processes between these two species is remarkable, from the very early events, such as involvement of glutamate receptors,1 ,51 to the bimodal protein synthesis. Thus, the concept of 'stability' of long term memory, historically speaking, has already been challenged. Nevertheless, the reappearance of the consolidation problem emphasises, yet again, the complex nature of learning, consolidation, retrieval and extinction as well as our need to carefully study all these processes on both molecular and behavioural levels.
De Novo Protein Synthesis (with Paraphernaliae)
During the last two decades most of the research efforts have been directed toward discovering the sequence of events that will ultimately lead to activity-driven transcription and consequently protein synthesis. The major break-through was made by Kandel and colleagues40,82 who unravelled the mechanism by which extracellular signals capable of inducing covalent modification of constitutive transcription regulators, mediated through second messengers, regulate the induction of gene expression (for review see refs. 9 and 54). Although inducible transcription factors, often called immediate early genes (IEGs), are not the focus of this chap ter, one could not discuss protein synthesis without referring to IEGs (see also Greenwood et al. in this book). Many of the IEGs products are regulated by kinases, such as fos and CREB6,7,13 or NF-kB78 and are set in motion by a different state of organism almost on a minute-to-minute basis. Although it is obvious that the regulation of gene expression does not necessarily depend on de novo synthesis, but could be achieved by posttranslational modification of existing IEGs, it is important to emphasise that: (a) IEGs work only because of their combinatorial properties,107 (b) CREB alone or any other transcription factor for that matter cannot be sufficient for initiation of DNA transcription or for any physiological process such as memory storage.
When a role for a particular protein is proposed on a basis of evidence from single-gene knockout approach, one should be aware of many interpretative difficulties precluding firm conclusions. 5,37,70 The reason for doubt is not solely due to the problems associated with knockout animals. When the idea that kinases can modify proteins already present within the synapse was proposed it was believed that these were synaptic transmission processes set in motion by the event itself.99 Experiments in which kinase inhibitors were applied to the region of synapses showed that inhibition of phosphorylation ceases to be effective at about 1 hr after long-term potentiation (LTP) is initiated. However, one hour after the LTP tetanus or low frequency stimulation, the synthesis of Nf-kB and TFIIIa as well as p50 and p65 mRNA are increased.78,79 The most probable reason for this rapid synthesis of transcription factors following phosphorylation events is that these proteins must be replenished.77 Increased activation and/or synthesis of transcription factors could lead to more synthesis of target proteins as well as repressors and terminators of transcription1 that would ultimately cause cessation of gene transcription. A good example of this complexity comes from work of Kinney et al.58 describing the effect of decreased binding of hippocampal TFs to the E-box that is correlated with increased GAP-43 mRNA. All these events, rather different in nature, such as: transmembrane signalling, phosphorylation cascade, protein kinase activation, recruitment of transcription factors somehow merge their effect at the same time-point, 1-hr after the initial signal transduction event. Thus, one can assume that the transition from short-term memory to events that will trigger the consolidation is likely to occur at this point in time.
There is a time point at about 6 hr after the initial signal transduction event that may herald another transition point. Here, we find studies with widely disparate methods and animal models pointing to this time point.53,58,7896 Across different species, model-systems and training tasks, memory consolidation occurs only after the initiation of the late-transcription phase that leads to the synthesis of a variety of proteins among which the cell adhesion proteins implicated in morphological changes at specific synapses (Fig. 3).
One could ask the question 'why search for glycoproteins'? Long-term remodelling must involve changes to the structure or geometry of synaptic or dendritic membranes, hence attention was drawn to glycoproteins, especially cell adhesion molecules (idea formulated as a general theory in ref. 26). And indeed, over the past 15 years there has been a remarkable convergence of evidence pointing to (a) activity-driven changes in the synthesis of cell adhesion molecules and (b) a key role for the cell adhesion molecules in the process of memory formation.
Early learning in the chick has proved a particularly fertile system in which to study the cellular and molecular processes of memory formation and consolidation. The protocol which has been most widely used for those studies is a one-trial passive avoidance learning task, based on the disposition of young chicks to peck spontaneously at small objects and to remember their characteristics for long periods. This model has all the experimental advantages of one-trial learning paradigms and since its first description by Cherkin in 196916 it has been widely used for the biochemical and pharmacological study of the molecular events involved in memory formation.
Following the pioneering work of Gibbs and Ng,35,36 which utilised a pharmacological dissection procedure to identify biochemically sensitive periods in the minutes following training on this task, a combination of interventive and correlative studies has revealed a cascade of molecular processes occurring in defined brain regions, notably the left intermediate medial hyperstriatum ventrale and lobusparolfactorius. Briefly, within minutes of pecking at the bitter
Figure 3. Pre and post-training intervals sensitive to 2-deoxy-Galactose.
bead, there is: (i) enhanced glutamate release, (ii) up-regulation of NMDA-sensitive glutamate receptors, and (iii) the opening of N-type conotoxin-sensitive calcium channels. These synaptic transients result in the activation of protein kinases and expression of IEGs such as c-fos and c-jun and consequently, the family of late genes coding for glycoproteins which, inserted into the pre and post-synaptic membranes, alter synaptic structure and connectivity (for review see ref. 93). Several aspects of this cascade and its time-dependencies are reminiscent of those occurring during and following LTP However a one-trial task may not be typical of learning in general, because many instances of animal and human learning are based on the acquision of experience in a number of repeated trials, involving processes such as generalisation, categorisation and discrimination.
We have been able to identify a number of pre and post-synaptic membrane glycoproteins, which show enhanced fucose incorporation between 1 and 24 hr after training (presynaptic 50 kD; post-synaptic 33, 100-120 150-180 kD). This enhanced fucosylation was accompanied by increased activity of fucokinase. Moreover, the anti-metabolite 2-deoxygalactose, if injected around the time of training, and 4-6 hr after training , produced amnesia in animals tested 24hr later (for review see ref. 93).
What might be the significance of those 2 time-windows of sensitivity to the fucosylation inhibitor - presumably representing 2 waves of glycoprotein synthesis? A clue comes from manipulating the nature of the training experience: in a normal training protocol chicks peck at the bead which was made aversive by immersion in 100% methylanthranilate (MeA), and will avoid that bead for at least 48 hr subsequently. However, if the aversant is made weaker (10%), the avoidance response is initially as strong as for 100% MeA, but retention does not persist. Much more importantly, in the case of weak training fucosylation does not occur. These data led us to believe that, for the memory of passive avoidance training to endure, a functionally discrete second wave of neural activity, including glycoprotein, synthesis is required.
Two questions inevitably arise: (i) as even a single cell type uses multiple molecular mechanisms in adhering to the other cells (and to the extracellular matrix), the specificity of cell-cell adhesion must result from the integration of a number of different adhesion systems, some of which are associated with specialised cell junctions while others are not, and (ii) whether this cascade is unique to the specific case of one-trial learning in the chick or whether it is generalisable to other forms of avian and mammalian learning?
Our work on different synaptic transmembrane glycoproteins showed that the glycoproteins are recruited at different points in time and so are susceptible to blockade only at the time
at which they are recruited. This conclusion was based on two different but complementary approaches: (i) an antibody approach, by which we were able to interfere with the expression of protein function by blocking the extracellular domain of protein in question (Fig. 4), and (ii) antisense approach, by which we were able to downregulate gene expression, hence to test the significance of de novo protein synthesis (Fig. 5).
L1, NCAM and APP antibodies were injected into the IMHV at various time-points before and after training. These groups received antibodies 30-min pretraining and 5.5 and 8-hr post-training. These times were similar to the time-windows of amnestic action observed with 2-d-Gal. Animals were tested 24 hr after training. Memory retention was impaired, compared with saline controls, (a) in the case of anti-L1 during both time-windows of protein synthesis; (b) in the case of anti-NCAM only during the second time-window, and (c) in the case of anti-APP only during the first time-window of protein synthesis. Since blocking the protein function by use of specific antibodies outside of specific time window were without effect, we addressed the question of the importance of de novo protein synthesis by use of synthetic oligodeoxynucleotides.
The two distinct time-windows of behavioural response to APP and NCAM downregulation confirmed that: (a) induction of de novo synthesis of different synaptic transmembrane glyco-proteins occurs at different points in time after training, and (b) the specificity of cell-cell adhesion indeed results from the integration of a number of different adhesion systems.80'81
"... be always prepared to rewrite the encyclopaedia"
Umberto Eco, Serendipities, Chapter: The Force of Falsity, 1998
Although it is safe to say that protein synthesis leads to structural changes that are the physical substrate for long-term memory, there are time points in which the protein synthesis seems to be constitutive and independent of input. The experimental data that supports this notion are obtained from studies of transcriptional activation of GAP-43, which occurs 12, 24, 48 and 72 hr after LTP.85,86 There is obviously a 2-3 day delay in promoter activation and mRNA synthesis. At this point we should start to doubt the prevailing view. What maintains memories for lifetime if proteins are synthesised, utilised, and finally replenished reflecting the recruitment of gene regulatory events that are input independent?
And, there are more questions to be asked. Namely, does dependence on the protein synthesis indicate that the proteins required for enduring synaptic modification are made on mRNAs
that are present constitutively, or on mRNAs that are synthesised as a consequence of learning-induced transcriptional activation? Modification of existing proteins already positioned at synapses to be modified could potentially solve the problem of synaptic specificity,34 because rapid changes in protein function can occur at the site of change. But, if the required proteins are made on constitutively expressed mRNAs, the precise mechanism of their increased trans-lational activation should be synaptic activity. And if that is the case, what could be the 'pick-and chose' mechanism by which cells precisely regulate the translation of different mRNAs from the mix of mRNAS that are in place. If the required proteins are synthesised as a consequence of learning-induced transcriptional activation and the new gene expression is required, what is the nature of the process that allows signalling from the synapse to control the transcriptional activation of the neuron? What are the mechanisms by which the newly synthesised proteins essential for activity-dependent synaptic modification are selectively delivered to the synapses that are to be modified? How is this coordinated and how all of these molecules actually 'fit' into the process of memory consolidation?
The cellular mechanism for targeting newly synthesised mRNAs to synaptic sites on dendrites have been partially revealed by Steward and Worley109 through studies on the intracellular transport and synaptic targeting of Arc (activity-regulated cytoskeleton associated protein). The synthesis of Arc mRNA is induced by patterns of synaptic activity that also includes LTP. Arc mRNA is rapidly transported into dendrites and the de novo synthesised arc protein is assembled into the synaptic junctional complex of the recently activated synapses. Moreover, the experiments with Arc antisense showed that preventing of Arc induction impairs consolidation of long-term memory thus suggesting that Arc protein induction is fundamental not only in marking neurons which have been sufficiently depolarised to activate NMDA receptors, but also in the stabilisation of activity-dependent processes.
The recent success of 'proteomics', the large-scale analysis of proteins, provided the basis for a better understanding of the interactive network of proteins within multiprotein complexes associated with neurotransmitter receptors and cell adhesion proteins involved in signalling. However, understanding the regulation of the flow of genetic information between the genome and 'proteome' is still uncharted territory. The first step in this flow is dependent on successful gene expression. The expression of gene product seems to depend on a complex network of ribonucleoproteins (mRNAPs) regulated by ELAV/Hu proteins.112 This is particularly important for the expression of IEGs, which products could mediate the second wave of gene expression, in other words, the second wave of protein synthesis. Thus, the understanding of the properties of the mRNP expressed in neurons, their infrastucture and organised network, the approach called "ribonomics", could potentially provide research into memory formation with invaluable knowledge of the temporal and spatial regulation of activity-driven regulation of gene expression. Many of the neuronal mRNAs, to which ELAV/Hu proteins bind, encode transcription factors including CREB, ERG-1 and fos and appear to be packaged at distances far from the nucleus. Their localised expression in response to external stimuli may, when the transcription factors are transported back to nucleus, influence cellular events following stimulation of neurons.5 Moreover, it seems to be that ELAV/Hu proteins are capable of regulating expression of clustered mRNP subset encoding proteins with related functional properties. The advantage of such a guarded regulatory pathway is that it could represent a controlled synapse-specific mode of direct activation of specific genes, and subsequent de novo synthesis of specific proteins, without involvement of multiple signal transduction cascade.56,57 This mode of cellular response has the potential to be a selective mechanism by which dendritic branches, which received the appropriate input, activate de novo protein synthesis and modulate with great precision long-term modifications, the strength of the connections, underlying long-term memory formation.
1. Abel T, Lattal M. Molecular mechanisms of memory acquisition, consolidation and retrieval. Curr Opin Neurobiol 2001; 11:180-187.
2. Agranoff BW. Agents that block memory. In: Quarton GC et al. eds. The Neuroscience: A study programme. NewYork: Rockefeller Univ Press, 1967:756-764.
3. Agranoff BW. Learning and memory: Bochemical Approaches. In: Siegel GJ et al. eds. Basic Neu-rochemistry. Boston: Little Brown, 1981:801-820.
4. Agranoff BW, Davis RE, Brink JJ. Chemical study of memory fixation in goldfish. Brain Res 1966; 1, 303-309.
5. Albright TD et al. Neural Science: A century of progress and the mysteries that remain. Cell 2000; 100:S1-S55.
6. Anokhin KV, Rose SPR. Learning-induced increase of immediate early gene messenger RNA in the chick forebrain. Eur J Neurosci 1991; 3:162-167.
7. Anokhin KV et al. Effect of early experience on c-fos gene expression in the chick forebrain. Brain Res 1991; 544:101-107.
8. Austin L, Morgan IG. Incorporation of 14C-labelled leucine into synaptosomes from rat cerebral cortex in vitro. J Neurochem 1967; 14:377-387.
9. Bailey CH, Bartsch D, Kandel ER. Toward a molecular definition of long-term memory storage. Proc Natl Acad Sci USA 1996; 93:13455-13502.
10. Barondes SH, Jarvik ME. The influence of actinomycin D on brain RNA synthesis and memory. J Neurochem 1964; 11:187-189.
11. Barraco RA, Stettner LJ. Antibiotics and memory. Psychol Bull 1976; 83:242-302.
12. Bourtchouladze R, Rose SPR. Memory formation in day old chicks requires NMDA but not nonNMDA glutamate receptors. Eur J Neurosci 1992; 4:533-538.
13. Bourtchouladze R et al. Deficient long-term memory in mice with a targeted mutation of the cAMP-responsive element-binding protein. Cell 1994; 79:59-68.
14. Browning M et al. Synaptic phosphoproteins: Specific changes after repetitive stimulation of the hippocampal slice. Science 1979; 203:60-62.
15. Brusa R. Genetically modified mice in neuropharmacology. Pharmacol Res 1999; 39:405-419.
16. Cherkin A. Kinetics of memory consolidation: Role of amnestic treatment parameters. Proc Natl Acad Sci USA 1969; 63:1094-1101.
17. Cohen HD, Barondes SH. Further studies on learning and memory after intracerebral injection of actinomycin D. J Neurochem. 1966; 13:207-211.
18. Curran T. Morgan JI. Memories of fos. BioEssays 1987; 7:255-258.
19. DeOrtiz SP, Arshavsky YI. DNA Recombination as a possible mechanism in declarative memory: A hypothesis. J Neurosci Res 2001; 63:72-81.
20. De Robertis E et al. On the isolation of nerve ending and synaptic vesicles. J Biophys Biochem Cytol 1961; 9:229-235.
21. Dietrich A, Been W. Memory and DNA. J Theor Biol 2001; 208:145-169.
22. Dragunow M. A role for immediate-early transcription factors in learning and memory. Behav Genet 1996; 26:293-299.
23. Dunn AJ. The Chemistry of learning and the formation of memory. In: Gispen WH, ed. Molecular and functional neurobiology. Amsterdam, Elsevier, 1976; 347-387.
24. Dunn AJ. Neurochemistry of learning and memory: An evaluation of recent data. Ann Rev Psychol 1980; 31:343-390.
25. Dunn AJ. Kramarcy NR. Effects of ACTH and related peptides on cerebral RNA and protein synthesis. Pharmacol Therap 1984; 12:353-372.
26. Edelman GM. Cell adhesion and the molecular processes of morphogenesis. Ann Rev Biochem
27. Entingh D et al. Brain uridine monophosphate: Reduced incorporation of uridine during avoidance learning. Brain Res 1974; 70:131-138.
28. Fisher S, Litvak S. The incorporation of microinjected 12C-aminoacid into TCA insoluble fraction of the giant axon in the squid. J Cell Physiol 1967; 70:69-74.
29. Flexner LB, Flexner JB. Intracerebral saline: Effect on memory of trained mice treated with puro-mycin. Science 1978; 159:330 -331.
30. Flood JF et al. The influence of duration of protein synthesis inhibition on memory. Physiol Behav 1973; 10:555 - 562.
31. Flood JF et al. Relation of memory formation to controlled amounts of brain protein synthesis. Physiol Behav 1975; 15:97-102.
32. Flood JF et al. Modification of anisomycin-induced amnesia by stimulants and depressants. Science
33. Freeman FM, Rose SPR. Two time windows of anisomycin-induced amnesia for passive avoidance training in the day-old chick. Neurobiol Learn Mem 1995; 63:291-295.
34. Frey U, Morris RGM. Synaptic tagging and long-term potentiation. Nature 1997; 385:533-536.
35. Gibbs ME, Ng K. Counteractive effects of norepinephrine and amphetamine on quabain-induced amnesia. Pharmacol Biochem Behav 1977; 6:533-537.
36. Gibbs ME, Ng K. Behavioural stages in memory formation. Neurosci Lett 1979; 13:279-283.
37. Gingrich JA, Hen R. The broken mouse: The role of development, plasticity and environment in the interpretation of phenotypic changes in knockout mice. Curr Opinion Neurobiol 2000; 10:146-152.
38. Gispen WH et al. Phosphorylation of proteins in synaptosome-enriched fraction of brain after a short-term training experience: The effects of various treatments. Behav Biol 1977; 21:358-363.
39. Glassman E. The biochemistry of learning. An evaluation of the role of RNA and proteins. Ann Rev Biochem 1969; 38:605-646.
40. Goelet P et al. The long and the short of long-term memory - a molecular framework. Nature 1986; 322:419-422.
41. Guiditta A, Cupello A, Lazzarini G. Ribosomal RNA in the axoplasm of the squid giant axon. J Neurochem 1980; 34:1757-1760.
42. Guiditta A, Dettbarn WD, Brzin M. Protein synthesis in the isolated giant axon of the squid. Proc Natl Acad Sci USA 1968; 59:1284-1287.
43. Guiditta A, Hunt T, Santella L. Messenger RNA in squid axoplasm. Biol Bull 1983; 165:526-528.
44. Guiditta A et al. Factors for protein synthesis in the axoplasm of the squid giant axon. J Neurochem 1977; 28:1393-1395.
45. Gray EG, Whittaker VP. The isolation of nerve endings from brain: An electron-microscopic study of cell fragments derived by homogenisation and centrifugation. J Anat 1962; 96:79-88.
46. Halstead WC. Cerebral mechanisms for behaviour. New York: Wiley, 1951.
47. Hydén H, Egyhazi E. Nuclear RNA changes of nerve cells during a learning experiment in rats. Proc Natl Acad Sci USA 1962; 48:1366-1373.
48. Hydén H, Egyhazi E. Changes in RNA content and base composition in cortical neurons in rats in a learning experiment involving a transfer of handedness. Proc Natl Acad Sci USA 1964; 52:1030-1035.
49. Hydén H, Lange PW. S-100 protein: Correlation with behaviour. Proc Natl Acad Sci USA 1970; 67:1959-1966.
50. Ingoglia NA et al. Incorporation of [3H]amino acids into proteins in a partially purified fraction of axoplasm: Evidence for transfer RNA mediated, post-translational protein modification in squid giant axon. J Neurosci 1983; 3:2463-2473.
51. Izquierdo I, Medina JH. Memory formation: The sequence of biochemical events in the hippocampus and its connection to activity in other brain structures. Neurobiol Learn Mem 1997; 68:285-316.
52. Jacobson AL et al. Differential approach tendencies produced by injection of ribonucleic acid from trained rats. Science 1965; 150:636-637.
53. Jork R, Grecksch G, Matthies H. Impairment of glycoprotein fucosylation in rat hippocampus and the consequences on memory formation. Pharmacol Biochem Res 1986; 25:1137-1144.
54. Kandel ER, Pittenger C. The past, the future and the biology of memory storage. Philos Trans R Soc Lond B Biol Sci 1999; 354:2027-2052.
55. Kandel ER, Schwartz JH. Molecular biology of learning: Modulation of transmitter release. Science 1982; 218:433-443.
56. Keene J. Why is Hu where? Shuttling of early-response-gene messenger RNA subsets. Proc Natl Acad Sci USA 1999; 96:5-7.
57. Keene J. Ribonucleoprotein infrastructure regulating the flow of genetic information between genome and the proteome. Proc Nat Acad Sci USA 2001; 98:7108-7024.
58. Kinney WR et al. Prolong activation in E-box binding after a single systemic kainate injection: Potential relation to F1/GAP-43 gene expression. Mol Brain Res 1996; 38:25-36.
59. Kao JF, Greengard P. An adenosine3',5'-monophosphate-dependent protein kinase from Escherichia coli. J Biol Chem 1969; 244:3417- 3419.
60. Lasek RJ et al. Analysis of axoplasmic RNA from invertebrate giant axon. Nature New Biol 1973; 244:162-165.
61. Lasek JR et al. Transfer of newly synthesised proteins from Schwann cells to the squid giant axon. Proc Natl Acad Sci USA 1974; 71:1188-1192.
62. Levitan IB, Lemos JR, Novak-Hopper I. Protein phosphorylation and regulation of ion channels. Trends Neurosci 1983; 6:496-499.
63. Liang KC, Bennett C, McGaugh JL. Peripheral epinephrine modulates the effect of posttraining amygdala stimulation on memory. Behav Brain Res 1985; 15:93-100.
64. Litvin OO, Anokhin KV. Mechanisms of memory reorganisation during retrieval of acquired behavioural experience in chicks: The effects of protein synthesis inhibition in the brain. Neurosci Behav Physiol 2000; 30:671-678.
65. Machlus B, Wilson JE, Glassman E. Brain phosphoproteins: The effect of short experiences on phosphorylation of nuclear proteins of rat brain. Behav Biol 1974; 10:43-62.
66. Martin R. Ribosomes in peripheral and presynaptic domains of axon. In: Telkeen AW, Korf J, eds. Neurochemistry. New York: Plenum Press, 1997:661-665.
67. Martinez Jr JL. Endogenous modulators of learning and memory. In: Martinez JL, Kesner RP, eds. Learning and Memory. New York: Academic Press, 1981:127-155.
68. Matthies H. Plasticity in the nervous system - an approach to memory research. In: Amone Marsane C, Matties H, eds. Neural Plasticity and Memory formation. New York: Raven Press, 1982:1-15.
69. Mayford M, Kandel ER. Genetic approaches to memory storage. Trends Gen 1999; 15:463-470.
70. Mayford M, Abel T, Kandel ER. Transgenic approaches to cognition. Curr Opin Neurobiol 1995; 5:141-148.
71. McConnell JV. Memory transfer through cannibalism in planarians. J Neuropsychiatry Suppl 1962; 1:42-48.
72. McEwen B. Neuronal and cognitive effects of oestrogens. Introduction. Novartis Found Symp 2000; 230:1-6.
73. McEwen B. Effects of adverse experience for brain structure and function. Biol Psychiatry, 2000; 48:721-731.
74. McGaugh, JL. Time-dependent processes in memory storage. Science 1966; 153:1351-1358.
75. McGaugh JL. Memory - a century of consolidation. Science 2000; 287:248-251.
76. McGaugh JL, Cahill L, Roozendaal B. Involvement of the amygdala in memory storage: Interaction with other brain systems. Proc Natl Acad Sci USA 1996; 93:13508-13514.
77. Meberg PJ et al. Protein kinase C and F1/GAP-43 gene expression in hippocampus inversely related to synaptic enhancement lasting 3 days. Proc Natl Acad Sci USA 1993; 90:12050-12054.
78. Meberg PJ et al. Gene expression of the transcription factor NF-kB in hippocampus: Regulation by synaptic activity. Mol Brai Res 1996; 38:179-190.
79. Meberg PJ et al. MARCKS and protein H1/GAP-43 RNA in chick brain: Effects of imprinting. Mol Brain Res 1996: 35:149-156.
80. Mileusnic R, Lancashire C, Rose SPR. Sequence-specific impairment of memory formation by NCAM antisense oligodeoxynucleotides. Learning Memory 1999; 6:120-127.
81. Mileusnic R et al. APP is required during an early phase of memory formation. Eur J Neurosci 2000; 12:4487-4495.
82. Montarolo PG et al. A critical period for macromolecular synthesis in long-term heterosynaptic facilitation in Aplysia. Science 1986; 234:1249-1254.
83. Nader K, Schafe GE, LeDoux JE. Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature 2000; 406:722-726.
84. Nader K, Schafe GE, LeDoux JE. The labile nature of consolidation theory. Nature Rev Neurosci 2000; 1:216-219.
85. Namgung U, Routtenberg A. Transcriptional and post-transcriptional replication of brain growth protein: Regional differentiation and induction of GAP-43. Europ J Neurosci 2000; 12:3214-3136.
86. Namgung U, Matsuyama S, Routtenberg A. Long-term potentiation activates GAP-43 promoter: Selective participation of hippocampal mossy cells. Proc Natl Acad Sci USA 1997; 94:11675-11680.
87. Petrushka E, Guiditta A. Electron microscopy of two subcellular fractions isolated from cerebral cortex homogenate. J Biophys Biochem Cytol 1959; 6:129-132.
88. Popov N et al. PAGE-autoradiography of fucose incorporation into rat hippocampal glycoproteins after acquisition of a brightness discrimination. BBA 1983; 42:763-776.
89. Rees HD et al. Effect of sensory stimulation on the uptake and incorporation of radioactive lysine into protein of mouse brain and liver. Brain Res 1974; 68:143-156.
90. Riedel G et al. Reversible neural inactivation reveals hippocampal participation in several memory processes. Nature Neurosci 1999; 2:898-905.
91. Rose SPR. Strategies in studying the cell biology of learning and memory. In: Squire LR, Butters N, eds. The Neuropsychology of memory. New York: Guilford Press, 1984:547-554.
92. Rose SPR. Cell adhesion molecules, glucocorticoids and memory. Trends Neurosci 1995; 18:502-506.
93. Rose SPR. God's organism: The chick as a model system for memory studies. Learning Memory 2000; 7:1- 17.
94. Rose SPR, Harding S. Training increase [3H]-fucose incorporation in chick brain only if followed by memory storage. Neurosci 1984; 12:663-667.
95. Rose SPR, Haywood J. Experience, learning and brain metabolism. In: Davidson AN, ed. Biochemical correlates of brain structure and function. New York: Academic Press, 1977:249 — 292.
96. Rose SPR, Jork R. Long-term memory formation in chicks is blocked by 2-deoxygalacose, a fucose analog. Behav Neural Biol 1987; 48:246-258.
97. Rosenbaum M, Cohen HD, Barondes SH. Effect of intracerebral saline on amnesia produced by inhibitors of cerebral protein synthesis. Comm Behav Biol Part A 1968; 2:47 -50.
98. Rosenzweig MR, Bennett EL, Flood JL. Pharmacological modulation of formation of long-term memory. In: Adam G et al. eds. Brain and Behaviour. Adv Physiol Sci Vol 17 1981:101 - 111, London: Pergamon Press,
99. Routtenberg A. Anatomical localisation of phosphoprotein and glycoprotein substrate of memory. Prog Neurobiol 1979; 12:85-113.
100. Routtenberg A, Benson GE. In vitro phosphorylation of a 41.000-MW protein band selectively increased 24 hr after footshock or learning. Behav Neurol Biol 1980; 29:168-175.
101. Sara SJ. Retrieval and reconsolidation: Toward a neurobiology of remembering. Learning Memory 2000; 7:73-84.
102. Sara SJ. Strengthening the shaky trace through retrieval. Nature Rev Neurosci 2000; 1:212-213.
103. Schafe EG et al. Memory consolidation of Pavlovian fear conditioning: A cellular and molecular perspective. Trends Neurosci 2001; 24:540-546.
104. Steele RJ, Morris RGM. Delayed-dependent impairment of matching-to-place task with chronic and intrahippocampal infusion of the NMDA-antagonist D-AP5. Hippocampus 1999; 9:118-136.
105. Squire LR. Short-term memory as a biological entity. In: Deutsch D, Deutsch JA, eds. Short-term memory. New York: Academic Press, 1975:1-40.
106. Squire LR, Davis HP. The pharmacology of memory: A neurobiological perspective. Ann Rev Pharmacol. Toxicol 1981; 21:323 -356.
107. Struhl K. Mechanism of diversity in gene expression patterns. Neuron 1991; 7:177-181.
108. Steward O, Levy WB. Preferential localisation of polyribosomes under the base of dendritic spines of granule cells of the dentate gyrus. J Neurosci 1982; 2:84-91.
109. Steward O, Worley P. A cellular mechanism for targeting newly synthesised mRNAs to synaptic sites on dendrites. Proc Natl Acad Sci USA 2001; 98:70622-7068.
110. Sukumar R, Rose SPR, Bourgoyne RD. Incresed incorpotaion of [3H]fucose into chick brain gly-coproteins following training on a passive avoidance task. J Neurochem 1980; 34:1000-1006.
111. Taubenfeld SM et al. The consolidation of new but not reactivated memory requires hippocampal C/EBP. Nature Neurosci 2001; 4:813-818.
112. Tenenbaum SM et al. Identifying mRNA subsets in messenger ribonucleoprotein complexes by using cDNA. Proc Natl Acad Sci USA 2000; 97:14085-14090.
113. Tully T. Discovery of genes involved in learning and memory: An experimental synthesis of Hirschian and Benzerian perspective. Proc Natl Acad Sci USA 1996; 93:13460-13467.
114. Tytell M, Lasek JR. Glial polypeptides transferred into the squid giant axon. Brain Res 1984; 363:161-164.
115. Walton M et al. Immediate early gene transcription and synaptic modulation. J Neurosci Res 1999; 58:96-106.
116. Wied D de. Pituitary-adrenal system hormones and behaviour. In: Schmitt FO, Worden FG, eds. The neuroscience third study programme. Cambridge: MIT Press, 1974:653-666.
117. Yellin TO, Butler BJ, Stein HH. Inhibition of protein synthesis in mitochondria by cyclohexemide.
Fed Proc 1967; 26:833-835.
Was this article helpful?