Protein Synthesis And Memory Research Paper

The nature of protein synthesis in the brain became of particular interest for Behavioral scientists as a result of demonstrations in a number of vertebrate and invertebrate species that chemical agents that inhibit protein synthesis in the brain block the formation of long-term memory (LTM) of a variety of training tasks, yet do not block acquisition of the task (Agranoff et al. 1999). More recently, analogous effects of protein synthesis inhibition have been reported in a reductionist model of learning and memory, termed long-term potentiation (LTP).

For the purposes of this research paper, learning and short-term memory (STM) formation are defined by experimental measures of improved performance of an adaptive response observed during or shortly after a training session (acquisition), while LTM formation is evidenced by retention of the learned response in a subsequent training session, typically administered some days or weeks later. An experimental advantage of using an interventive agent such as a protein synthesis blocker in studying learning and memory is at the same time its great disadvantage: while it can establish the timing of an inferred molecular mechanism taking place in an otherwise intact animal, it does not specify cells or even brain regions critical for LTM formation, nor does it identify specific proteins on which memory formation is presumed to be dependent. Such proteins could hypothetically be identified by complementary correlative studies, for example, by examining the nature of newly synthesized proteins on the basis of their incorporation of radiolabeled precursor amino acids during the period of LTM formation. Together, the interventive and correlative approaches have proven useful in establishing the molecular bases of other complex cellular processes.

1. Proteins Defined

Proteins are biological macromolecules consisting of a linear array of 20 component -amino acids. The amino acids are joined in chemically stable peptide linkages that are not easily disrupted by thermal, chemical or mechanical stresses encountered by living cells. A protein typically consists of several hundred amino acids in a precise sequence (termed its primary structure), much like a very long word written in a 20letter alphabet. The secondary and tertiary structures of a protein refer to how the polypeptide chain is folded into helices, ribbons and sheets. This overall conformation of proteins is largely dictated by the chemical properties of the individual amino acids in its primary structure.

Proteins play both structural and functional roles in cells. Membranes, consisting of lipid bilayers in which proteins are embedded, define a neuron and its intracellular organelles, such as the nucleus, mitochondria, synaptic vesicles, and endoplasmic reticulum. Proteins may be primarily structural, adding integrity to the membrane, or serving as a bridge to other membranes or to one of a group of fibrillar proteins including tubulin, actin and neurofilament proteins, known collectively as the cytoskeleton. Functional roles of proteins are exemplified by the enzymes, which are generally soluble proteins, present in a globular conformation. Hundreds of enzymes have been characterized and shown each to be a chemical catalyst that accelerates the rate of a specific biochemical reaction, without itself being consumed. Structural and functional proteins together catalyze and otherwise mediate the business of the cell (Voet and Voet 1995), expressing its genetic phenotype, including its appearance, its metabolism and its repertoire of responses. Knowledge of the primary structure of a protein is useful in determining its relation to other known proteins, in some instances yielding information about its evolutionary lineage as well as providing insight into its function. Short amino acid sequences within proteins (domains) identify probable catalytic active sites, the number of times a protein chain spans a membrane and the probable site of a protein’s binding to other known proteins. There are also short amino acid sequences that identify the nature and likely sites of ‘decoration’ of proteins with small chemical groups following their synthesis. In addition to acting as catalysts, proteins can mediate recognition of a variety of small molecules that are intracellular chemical messengers. Membrane proteins on the cell surface may also act as receptors that recognize a variety of small molecules, including extracellular neuromodulators, hormones, neurotransmitters and environmental chemical signals (odorants) (Siegel et al. 1999). Proteins can recognize other proteins, even xenobiotic (novel) ones, as in the immune response. While proteins are clearly the arbiters of all cellular activities, not all cellular actions depend on ongoing protein synthesis, as will be seen.

1.1 Proteins Are Synthesized On Ribosomes From Messenger RNA

The primary structure of a protein ultimately is determined genetically by linearly coded triplet deoxynucleotides in DNA in the cell nucleus, which are transcribed to a complementary sequence of nucleotide triplets in the messenger RNA (mRNA) for the protein. The newly synthesized mRNA is translated (‘decoded’) into the new protein on polyribosomes in a cellular region near the nucleus, the RNA triplets establishing the primary structure of the new protein. The entire process, beginning with the availability of DNA template for transcription and ending with protein synthesis, generally takes minutes to hours (Voet and Voet 1995). Antibiotic inhibitors of protein synthesis such as puromycin and acetoxycycloheximide are structurally unrelated, and block translation by different molecular mechanisms, yet each has been shown to block LTM formation, confirming that their Behavioral effects are indeed the result of their inhibition of protein synthesis rather than of some unknown side effects of the agents. Inhibition of RNA synthesis has also been found to block LTM formation.

Upon release from the polysomes, newly synthesized proteins may be subjected to post-translational modification, typically the addition of stable, small chemical substitutents, such as fatty acid, isoprenoid, carbohydrate, and phosphate or methyl groups. These modifications are catalyzed by specific enzymes, and may also require cofactors such as calcium ions or cosubstrates such as ATP. Phosphorylation of a protein by ATP requires a protein kinase, although some proteins can autophosphorylate themselves. As a consequence of such modifications, a protein may be tagged with a chemical ‘zip code’ that designates its intracellular destination. For example, a lipid substituent will target a protein to a membrane site, while phosphorylation could direct a protein to a nuclear site. The overall conformation of proteins can be affected by these post-translational changes. A newly formed protein may be inactive until it is cleaved by a specific protein-splitting enzyme that removes a builtin inhibitory sequence.

Many proteins are part of intracellular messenger systems that act sequentially to produce multiplicative cascades, resulting ultimately in a cellular response, such as hyperpolarization, depolarization, contraction, or secretion. Unlike the ribosomal synthesis of proteins, in which the translation of proteins is measured in tens of minutes, post-translational modification can occur rapidly, even in milliseconds, and unlike typical protein synthesis, is not restricted to the perinuclear region of the cell. Post-translational modification of proteins can occur at any time in the life of a protein, and can even target proteins for degradation.

1.2 Brain Proteins Are In A Dynamic State Of Formation And Degradation

The thousands of protein species present in a brain cell are in a metabolically dynamic state, i.e., they are continuously being synthesized by ribosomes and are then broken down into their component amino acids by the action of enzymes called proteases. The released amino acids may be metabolized or can be reincorporated into new proteins. Turnover studies employing radiolabeled amino acid precursors show that a given cellular protein in a specific cell type is turning over with a characteristic half-life of minutes to weeks. Each protein is present at a characteristic steady state level, i.e., the rate of synthesis of the protein is equal to its rate of degradation. Even though most neurons are formed early in life and remain in a postmitotic state through adulthood, their component structures are nevertheless in a dynamic metabolic equilibrium at a molecular level. A neuron may thus maintain an unchanging microscopic appearance throughout life, yet, with the exception of DNA, every component molecule, including all of its neuronal proteins, is constantly being replaced. There is considerable evidence that the cell’s morphology and other cellular modifications of long duration are maintained by continual feedback to nuclear DNA, signaling initiation of transcription of mRNAs for constitutive proteins as they are depleted, or the induction of novel proteins, as occurs in cellular differentiation during development and in nerve regeneration. Thus, changes in synaptic connections might be similarly maintained by such transcriptional feedback loops in LTM storage. Translational and post-translational feedback loops that mediate memory formation have been proposed (Lisman and Fallon 1999).

1.3 Temporal Considerations In Establishing The Role Of Protein Synthesis In Memory Formation

Since acquisition of a new Behavior can occur within seconds, it is immediately apparent that it cannot be either directly or indirectly mediated by the relatively slow process of de no o protein synthesis. Acquisition could, however, involve a sustained post-translational modification at the synapse, which can occur within milliseconds and at any cellular locus. As noted, the macromolecular process of protein synthesis requires initiation of transcription in the nucleus followed by ribosomal translation, and takes minutes to hours, a time frame compatible with LTM, but not with STM formation. Additional time may be required for a putative chemical message from the presumed site of STM formation in the synaptic region to be transported to the nucleus in the neuronal cell body and to initiate transcription and translation. Still more time may be required for newly synthesized proteins to be transported from the cell body to a synaptic destination, where they could mediate long-term neuroplastic changes.

Rates of protein flow away from the cell body (anterograde axonal transport) and from the dendrites and axons toward the cell body (retrograde axonal transport) were established in neurons with long processes. The amount of time required for trafficking of a chemical signal from a synapse to the nucleus and of a newly synthesized protein back to the synapse hypothetically could be extrapolated from known rates of axonal transport, together with knowledge of the morphology of the neuroplastic neurons hypothesized to be mediating the process. Ribosomal protein synthesis at a distance from the nucleus (Steward 1997) could shorten the calculated transport times. As will be addressed in the following section, the advent of antibiotic agents that selectively block protein synthesis offered the opportunity to study these complex issues experimentally.

2. Agents That Inhibit Protein Synthesis Block Formation Of Long-Term Memory

In the early 1960’s, the excitement over the identification of the genetic code spurred renewed speculation and experimentation directed at the possible molecular nature of the engram. In 1962 Flexner and co-workers reported that intracerebral injection of the antibiotic puromycin, an aminonucleoside that prematurely cleaves the growing peptide chain from ribosomes, blocked brain protein synthesis in the mouse brain and could also block formation of memory of a Y-maze shock avoidance task (Flexner et al. 1962). This was followed by studies in the goldfish in 1964 (reviewed in Agranoff et al. 1999), and subsequently in a variety of other vertebrates and invertebrates.

2.1 Effects Of Protein Synthesis Blockers On LTM Formation In Experimental Animals

Injection of puromycin into the cranial fluid (intracranially; IC) surrounding the goldfish brain produced a profound block in brain protein synthesis for several hours. Studies on acquisition (STM) and on LTM formation of a light-coupled-to-shock shuttlebox task revealed that a block in brain protein synthesis had no effect on acquisition of avoidance Behavior during an initial multitrial training session, but eliminated retention of the improved performance that would otherwise have been seen in a second session administered a week later. The depth and duration of the inhibition of brain protein synthesis and the effectiveness of the block of LTM formation correlated well. Injection of the blocker into the fish body cavity rather than IC was without effect on LTM formation. IC injection of acetoxycycloheximide, another inhibitor of protein synthesis, was equally effective in blocking LTM formation.

The Behavioral effects of inhibition of protein synthesis differ from that of a number of neurodisruptive agents and treatments, such as electroconvulsive shock, in that animals exhibit normal Behavior during the action of the interventive agent. This permitted an exploration of STM decay. It was known that fish injected IC with puromycin before the first training session would show increasing avoidance scores, comparable to those seen in uninjected control fish; yet in a second training session a week later, such fish performed at a naıve level, i.e., had formed no LTM of the first session. Groups of fish trained under conditions of blocked protein synthesis (and that nevertheless demonstrated normal acquisition, as noted above) were then retrained at various times following the initial session. It was found that performance gradually dropped to the level of naıve fish over a 2–3 day period, providing evidence for the decay of STM for this task. Block of protein synthesis immediately after training also blocked LTM formation, whereas if animals were returned to their home tanks and injected with puromycin a few hours later, there was no detectable effect on LTM. This period of LTM susceptibility was comparable in its duration to that seen with electroconvulsive shock and other disruptive agents administered after training. When fish were detained in the training apparatus following the first training session, the consolidation time was prolonged, suggesting that the onset of LTM formation required an ‘environmental trigger.’ This phenomenon, in which continued exposure of an experimental subject to its training environment delays the onset of LTM formation, has also been termed ‘detention.’

Behavioral studies with numerous other species, including young chicks and rats (for review, see Davis and Squire 1984), support the general hypothesis that LTM formation requires ongoing protein synthesis, while STM formation does not. Among vertebrates, studies in Hermessenda and in Aplysia (Bailey et al. 1992) support this hypothesis (see Agranoff et al. 1999). Drosophila have proven useful in Behavioral studies, with genetic mutants in which selected proteins are mutated, deleted or overexpressed, implicating various neurotransmitter-related and second messenger system proteins in memory formation (see Agranoff et al. 1999). Effects of protein blockers on human memory have not been reported, but the possibility did not escape the imagination of a novelist (MacDonald 1968).

2.2 Protein Synthesis Inhibition In Long-Term Potentiation, A Reductionist Model Of Learning And Memory

Attempts to identify specific proteins, the synthesis of which are crucial to LTM formation in the brain, have been beset with many experimental problems, not the least of which is knowledge of the precise anatomical sites of the presumed altered molecular events underlying an observed altered Behavioral performance interpreted as LTM formation. LTP can be induced and studied ex vivo in localized tissue preparations such as brain hippocampal slices, and has been proposed as a cellular model of LTM formation. Blockers of protein synthesis have been shown to prevent LTP formation, a result taken by some as its validation as a model of LTM formation. Numerous pharmacological, biochemical, molecular, and genetic studies have been performed to implicate one or another structural protein, enzyme, receptor, or intracellular messenger pathway in the development of LTP, as a result of which over 100 known molecules have been implicated in LTP formation, including intracellular messengers, vesicular trafficking proteins, proteases, kinases, etc. (Sanes and Lichtman 1999). Eventually it may prove to be the case that the ongoing synthesis of one or only a few of these proteins is indeed necessary and sufficient for LTP formation, and can account for the observed block of LTM formation by protein synthesis inhibitors. It would seem more likely that the synthesis of many such proteins is necessary for the complex process of LTM formation to proceed. This embarrassment of riches in candidate compounds continues to grow as additional previously undiscovered cellular proteins become implicated in neuroplasticity, in that their selective deletion or inactivation prevents LTP or LTM formation. Further information regarding the temporal sequence of the observed molecular events accompanying LTP may prove useful in establishing whether they are indeed causal or epiphenomenal in its formation.

2.3 Protein Nutrition And Mental Function

Half of the 20 amino acids of which proteins are composed can be synthesized in the human body, while the remaining 10, termed ‘essential amino acids’ (EAAs), must come from the diet, principally in the form of ingested proteins that are degraded to their component amino acids in the gut and absorbed into the bloodstream. Eventually some will cross the blood brain barrier, which has efficient EAA uptake systems. Also, as the brain’s own proteins are broken down, the released amino acids are conserved and become a rich source of EAAs for new brain protein. Thus, under conditions of dietary EAA deficiency in adults, as occurs in starvation, brain protein metabolism is relatively protected. This is not true for the developing brain in infancy and early childhood, periods during which the brain must grow. For this, the brain relies on an exogenous source of EAAs, without which permanent cognitive deficits can ensue, as in kwashiorkor, a protein deficiency disease of children encountered in impoverished tropical countries. A blood amino acid imbalance is seen in the genetic defect phenylketonuria in which there is a high blood level of the amino acid phenylalanine, while blood levels of the other EAAs are normal. This imbalance overloads and blocks the amino acid transport system at the blood/brain barrier and effectively prevents entry of the other EAAs, eventually resulting in mental retardation. If infants with this genetic defect are raised on a diet deficient in phenylalanine, the brain develops normally, and as they mature, the toxic effect of dietary phenylalanine is much reduced.

3. Summary

It is apparent from the foregoing that in a number of animal species, disruption of protein synthesis in the brain for a period of hours has no effect on acquisition of a Behavioral task, but blocks the formation of LTM. Since all brain proteins turn over metabolically, it is presumed that positive feedback loops are induced during LTM consolidation that involve transcription and/or translation of new protein, but how and where these loops function and are distributed in the complex structure of the brain is not yet known. Selective inhibition or elimination of various proteins by genetic means has been shown to block memory formation, but the salience of each new finding is dampened by the increasing number of identified genetic defects or deletions that can be shown to lead to a defect in memory and learning. The synthesis of many proteins each may prove necessary (but not sufficient) for LTM formation.

STM formation, as defined in human and primate learning and memory studies, may be measured in fractions of seconds, with lifetimes that are compatible with the durations of electrophysiological measurements of neuronal activity. Other forms of STM formation, primarily in lower vertebrates and invertebrates, may be long-lived (seconds to hours or even days) as indicated in Sect. 2.1. Intermediate stages in STM formation have been proposed in studies on taste aversion in the chick (Gibbs and Ng 1979). It is possible that the longer lasting forms of STM reside in synaptic regions and are mediated by posttranslational feedback loops between proteins, and thus would not be affected by the protein synthesis blockers, as has been observed. It must be remembered that the various molecular, pharmacological, Behavioral, and electrophysiological interpretations of memory formation are not mutually exclusive and in many instances may reflect observation of the same neurobiological phenomenon from dissimilar vantage points.

Bibliography:

1. Agranoff B W, Cotman C W, Uhler M D 1999 Learning and memory. In: Siegel G J, Agranoff B W, Albers R W, Fisher S K, Uhler M D (eds.) Basic Neurochemistry: Molecular, Cellular and Medical Aspects, 6th edn. Lippincott-Williams & Wilkins, Philadelphia New York, Chap. 50
2. Bailey C H, Montarolo P, Chen M, Kandel E R, Schacher S 1992 Inhibitors of protein and RNA synthesis block structural changes that accompany long-term heterosynaptic plasticity in Aplysia. Neuron 9: 749–58
3. Davis H P, Squire L R 1984 Protein synthesis and memory: A review. Psychological Bulletin 96: 518–59
4. Flexner J B, Flexner L B, Stellar E, de la Haba G, Roberts R B 1962 Inhibition of protein synthesis in brain and learning and memory following puromycin. Journal of Neurochemistry 9: 595–605
5. Gerard R W 1953 What is memory? Scientific American 189: 118–25
6. Gibbs M E, Ng K T 1979 Neuronal depolarization and the inhibition of short-term memory formation. Physiological Behavior 23: 369–75
7. Lisman J E, Fallon J R 1999 What maintains memories? Science 283: 339–40
8. MacDonald J D 1968 The Girl in the Plain Brown Wrapper. Fawcett, Greenwich, CT
9. Sanes J R, Lichtman J W 1999 Can molecules explain long-term potentiation? Nature Neuroscience 2: 597–604
10. Siegel G J, Agranoff B W, Albers R W, Fisher S K, Uhler M D (eds.) 1999 Basic Neurochemistry: Molecular, Cellular and Medical Aspects, 6th edn. Lippincott-Williams & Wilkins, Philadelphia New York
11. Steward O 1997 mRNA localization in neurons: A multipurpose mechanism? Neuron 18: 9–12
12. Voet D, Voet J G 1995 Biochemistry, 2nd edn. Wiley, New York