post translational modification

http://www.biotech-online.com/featured-articles/ptm-of-brain-proteins-as-substrate-for-memories-that-last-a-lifetime/index.html


The brain mechanisms underlying the long-term storage of memories remains a fundamental unsolved problem in biology. The current prevailing view is that neural activity generated by learning alters synaptic relationships and then instigates de novo protein synthesis. These newly synthesised molecules then go to those very same synapses, stabilise them and as a consequence, stabilise the memory itself.

At our peril, we have advanced a model whereby post-translational modification (PTM) of proteins already present at synapses is sufficient for long-lasting memory storage. The PTM mechanisms are sufficiently large in number to enable rather subtle multidimensional scaling within the synapse itself. A few of these candidates – phosphorylation, polymerization, translocation, and proteolysis – have already been linked to memory formation, though not necessarily the long-lasting form.

Here we briefly describe our proposal that PTM of synaptic proteins is both necessary and sufficient for information storage acting to instruct and guide the formation of networks that represent the memories of lessons learned. In contrast, translational machinery acts to replenish depleted proteins in a permissive fashion.

The implications of this research for the development of drugs to aid memory are discussed.by Jerome L. Rekart and Aryeh RouttenbergPost-translational modification (PTM) of cellular proteins is universally recognised as a pivotal biochemical mechanism responsible for regulating the functionality of cellular proteins and thus critical to the normal functioning of eukaryotic cells. It is not too surprising, therefore, that it has also been found to be an important mediator of the neuronal changes associated with learning and memory in species ranging from insects [1] to primates [2].

The link between PTM of brain proteins and information storage in the brain has long been viewed as critical for the initial stages of memory formation [3, 4, 5] but not its long-lasting stage. Despite the ubiquity of PTMs and their necessity for cellular functioning, the prevailing view of memory storage holds that though necessary for the storage of information in the short-term, they are not sufficient for the long-term cellular changes underlying long-lasting memory.

Instead, long-term changes are believed to be protein synthesis (PS)-dependent, relying on signalling cascades that in turn regulate translational machinery [6]. De novo PS is thus seen to coordinate the cellular changes associated with long-lasting memory. This two-phase view of cellular learning, sometimes referred to as the dual-trace hypothesis [7], with an initial protein synthesis-independent step followed by a long-term process requiring protein synthesis (PS-dependent), is now current “wisdom” and thus articulated without reservation in major neuroscience textbooks [8].

The position taken here is that this view is essentially untenable because the fundamental cellular events underlying long-lasting memory formation require only PTM of synaptic proteins. Translational machinery serves to replace depleted proteins after molecular degradation brought on by protein PTM in the service of selective and instructive neural plasticity and memory-related processes.

Thus, protein synthesis is a permissive rather than an instructive process in the long-term storage of information. The centrality of protein synthesis to the cellular changes underlying learning and memory is thus far from settled. Indeed, there are a number of non-trivial logical criticisms and methodological concerns related to the study of protein synthesis, which have been enumerated in detail elsewhere [5, 9, 10, 11] including the requirements for the production, transport and targeting of new proteins to activated synapses.

The fact that many proteins have half-lives of a day or so, while memories last for a lifetime, is but one unresolved issue. The situation becomes overly complex when one imagines a system that would need to extract punctate temporal events from the inexorable flow of time’s arrow.

That is, the discontinuity required by our ability to attach chronology to a given memory requires that the cell disrupt the normal, continual dendritic transport of material to synapses in order to render it punctate at specific synaptic locales. This appears to be a formidable barrier.

In contrast, the PTM hypothesis obviates the need for this complex trafficking issue as part of the instructive mechanism. Instructions can thus be satisfactorily implemented by PTM of brain proteins already present at the synapse at the time of learning [5, 9] giving rise to a controlled, rapidly-adaptive, memory system that is resistant to disruption.

Perseverance of memory then results consequent to maintenance of some residual of the post-translational modification of proteins already present at activated sites. It is known that the initial stages of memory and the consequent activity-dependent modification of synaptic strength are regulated by neural plasticity-related enzymes, such as kinases and phosphatases and their protein substrates.

What is perhaps unique about the PTM model is that such modification implementing plasticity-related enzymes is sufficient for memory formation and its perseverance [9], is clearly necessary [12] and, given the wide range of mechanisms included under the PTM rubric, it is also possibly exclusive.The PTM model eschews the long-term commitment of energy and resources suggested by the PS-model.

Counterintuitively, a stable memory does not require a stable synapse. Indeed, this PTM-synaptic malleability is consistent with data and theories of neural homeostatic plasticity, which require reversibility of cellular changes [13] so as to thwart the occurrence of a thermonuclear engram [5].

The removal of translation as a necessary component for information storage increases the responsiveness and adaptability of the cell to input and removes the necessity for a targeting or “tagging” mechanism to direct nascent proteins to their proper destinations as all of the necessary enzymes and substrate molecules will already be present at the appropriate synapse.T

he PTM model is embedded within and builds upon the context of the neural network. Thus, the ongoing post-learning duplication of cortical and subcortical networks underlying a given memory provides a safeguard against the disruption of memory due to loss of circuitry, or indeed new learning, or traumatic experiences or local brain damage.

Moreover, cryptic rehearsal, previously thought of a ‘spontaneous activity’ can act to prolong PTM states, acting as a regulated positive feedback system [5].IMPLICATIONS OF THE PTM MODELInstead of the ‘dual trace’ model currently in vogue, the PTM model advocates a single trace mechanism, obviating the temporal categorisation of memory into distinct ‘stages.

’ The prevailing model of memory divides the phenomenon into distinct processes, in which short-term memory ‘processes’, believed to be protein synthesis-independent, may last after the learning event from seconds up to several hours. Then the trace becomes PS-dependent and thus categorised as either long-term. This may be preceded by the recently-discussed ‘intermediate-term’ memories [14], which last from hours to days.

Then at some undetermined time memory undergoes sufficient consolidation and transitions from ‘long-term’ to ‘long-lasting’, the latter believed to persist as long as the organism. The situation becomes murky when considering the view that multiple ‘waves’ of protein synthesis are necessary for long-term memory [15].

As such evidence is based on the use of protein synthesis inhibitors, all of the requisite caveats [5] apply.Although labels like ‘short-term’ or ‘long-lasting’ memory provide a useful shorthand for describing the age of a memory, the PTM model proposed renders the use of poorly defined temporal categories of memory stages as unnecessary because the mechanisms underlying memory storage are the same 50 years after the event as they are 50 minutes after its initial occurrence: post-translational modification of the molecules already present at
synapses.

RAMIFICATIONS OF THE PTM MODEL: APPROACHES TO ESTABLISHING ITS VALIDITY

In order for the PTM hypothesis to be of value it needs to be tested with strong inference experiments so that the outcome allows one to confirm or deny its suitability. Another value-added approach is to determine whether such a model can suggest pharmaceuticals that will reverse memory impairments in neuroclinical conditions and also enhance memory ability in normal, healthy adults.

As reviewed elsewhere, the PTM hypothesis is supported by
a) the ability to form memories in the presence of near total protein synthetic inhibition,

b) the demonstration of long-lasting PTM of proteins linked to long-lasting memory, and c) the disruption of memory by PTM inhibitors days after the memory is formed.

In a) above, such evidence serves to falsify the assertion that protein synthesis is instructive for memory.

In b) identifying substrate molecules that are phosphorylated, dephosphorylated, etc. in response to stimuli and their associated kinases and phosphatases (in the case of phosphorylation) provides initial clues for new drug targets.

In c) the central tenet of the PTM-model of information storage is evaluated: the meta-stability of the neural networks underlying memories as regulated by PTM.

In an initial test of the PTM theory, Holahan and Routtenberg [16] demonstrated that injection (3 weeks after learning and 1 hour before retention test) of H-7, the broad spectrum serine-threonine kinase inhibitor, into anterior cingulate cortex interferes with memory retention for the original event.

As a negative result would have been taken as evidence against the PTM hypothesis, in fact, the PTM model survives this strong inference test. Moreover, the work of Shema, Sacktor and Dudai (2007) [17], along with our work, suggests that a potentially important drug target opportunity may be synaptic proteins that are protein kinase C substrates.

Ultimately, the PTM hypothesis should provide important leads in the development of drugs to aid memory, for those who are memory-impaired and for those normal, healthy adults who may wish to enhance this ability. We have previously provided models of how phosphorylation regulated by protein kinase C, alters the protein-protein interaction between a PKC substrate and those proteins that regulate transmitter release [18].

Discerning the epitopes where these synaptic phosphoproteomic intereactions occur could provide the substrate for selectively facilitating synaptic function and thereby facilitating the memory formation process.

Clearly, this is a bold suggestion at this juncture. Nonetheless, it serves to illustrate how a new hypothesis of memory formation may allow the formulation of a new direction in regulating our information storage capacity through the selective modification of critical sites on memory-related,


PTM-modified synaptic proteins. REFERENCES1. Fiala, A. et al. Journal of Neuroscience 1999:19;10125-10134. 2. Nelson, R.B et al. Brain Research 1987:416;387-392. 3. Routtenberg, A. Progress in Neurobiology 1979:12(2);85-113. 4. Routtenberg, A. (1982). Memory formation as a post-translational modification of brain proteins. In C.A. Marsden and H. Matthies (Eds.), Mechanisms and Models of Neural Plasticity: IBRO Symposium on Learning and Memory,Vol. 9 (pp. 17-24). New York: Raven Press.5. Routtenberg, A., & Rekart, J.L. Trends in Neurosciences 2005:28(1);12-19.6. Kandel, E.R. Science 2001:294; 1030-1038.7. Hebb, D. O. (1949). The Organization of Behavior. New York: John Wiley.8. Bear, M.F. et al. (2006). Neuroscience. New York: Lippincott. 9. Routtenberg, A. Neurobiology of Learning and Memory 2008:89(3);225-233. 10. Gold, P.E. Neurobiology of Learning and Memory 2008:89(3);201-211.11. Rudy, J.W. Neurobiology of Learning and Memory 2008:89(3);219-224.12. Colley, P.A. et al. Journal of Neuroscience 1990:10(10);3353-3360.13. Turrigiano, G. Current Opinion in Neurobiology 2007:17(3);318-324.14. Stough, S. et al. Current Opinion in Neurobiology 2006:16(6);672-678. 15. Richter, K et al. Learning & Memory 2005:12(4);407-413.16. Holahan, M.R. & Routtenberg, A. Hippocampus 2007:17;93-97.17. Shema, R. et al. Science:2007;317, 951-953.18. Routtenberg, A. et al. Proceedings of the National Academy of Sciences U.S.A 2000:97;7657-7662.AUTHORSJerome L. RekartDepartment of PsychologyRivier CollegeNashuaNH, USAAryeh Routtenberg*Departments of Psychology, Neurobiology & PhysiologyNeuroscience InstituteSwift Hall Room 102Northwestern University2029 Sheridan Rd, Evanston, IL, USATel: +1 (847) 491-3628Fax: +1 (847) 491-3557Email: aryeh@northwestern.eduAuthor for correspondance