In this final section we will begin to address issues related to the E-LTP to L-LTP transition. I include them in this chapter because they involve the generation of persisting signals, specifically the increased synthesis of proteins; thus, these mechanisms fit well within the general theme of this chapter. Also, there are data to suggest that regulation of protein synthesis may be a downstream target of some of the specific persisting signals involved in E-LTP, such as persistently activated PKC (see Figure 15), and these data fit in this chapter for that reason. However, these mechanisms also can be looked upon as part of the induction process for protein synthesis-dependent L-LTP, and thus serve as a transition to the next chapter where we talk about L-LTP as well. I also emphasize that these mechanisms are a gray area right now and are very much an area of ongoing discovery. Many aspects of the specific models and diagrams I will present in this section are speculative.
There was a resurgence of interest in this area when Kelsey Martin in Eric Kandel's laboratory published a seminal finding
demonstrating that localized dendritic protein synthesis is involved in synapse-specific potentiation of synaptic transmission in Aplysia sensory neurons (84). This discovery along with a variety of earlier findings led to the formulation of the model that local dendritic protein synthesis could provide a solution to the vexing problem of synapse-specificity for protein synthesis-dependent L-LTP. The conundrum was that L-LTP is dependent on protein synthesis, but dogma had it that protein synthesis happened exclusively in the rough ER in the cell body. The synapse specificity problem can be solved simply if there is activity-regulated protein synthesis limited to specific dendritic or synaptic regions. Formulation of the "local protein synthesis" model for L-LTP capitalized on earlier groundbreaking work from Ozzie Steward's group and Bill Greenough's group that indicated the existence of dendritically localized protein synthesis machinery (polyribosomes).
Work from a wide variety of labs has supported the relevance of local protein synthesis to explaining synapse specificity of protein-synthesis-dependent LTP, and indeed to activity-dependent synaptic plasticity in the CNS in general. It is clear that there is activity-dependent regulation of protein synthesis for a variety of proteins, notably among them CaMKII (85), PKMZ (as described earlier), and Arc (see Box 1 in Chapter 3). It is also now quite clear that local, regulated protein synthesis occurs in dendrites (reviewed in references 86 and 87). Of course, the oft-replicated finding that protein synthesis inhibitors can block L-LTP is what precipitated the local protein synthesis model in the first place (1, 2).
How is local protein synthesis regulated in LTP? The mechanisms for regulating protein synthesis are themselves horren-dously complicated, even without the added complexity of trying to understand how neuronal activity-dependent mechanisms might impinge upon them. Figure 15 and the following discussion summarizes a plethora of papers and reviews from the laboratories of Bill Greenough, Steve Warren, Erin Schuman, Ozzie Steward, Paul Worley, Cliff Abraham, and Eric Klann. It is a brief summary of some of the signal trans-duction mechanisms that are hypothesized to operate in the LTP-associated regulation of local protein synthesis in CA1 pyramidal neurons (86-88).
Protein synthesis must, of course, begin with the recognition of an mRNA by the ribosomal complex, which allows the intitiation of peptide chain elongation starting from the 5' end of the message. One mechanism for translational initiation involves the eukaryotic translation initiation factor 4e (eIF4e). Activated eIF4e associates with a number of co-activating proteins and this complex recruits the ribosome to the mRNA, which initiates the process of scanning for the AUG start codon. Activation of eIF4e is regulated by phosphorylation at one major site, Ser 209. The kinase likely to mediate this phos-phorylation, at least based on the data available at present, is MNK1. MNK1 is mitogen-activated protein kinase-interacting-kinase 1 (MNK1). MNK1 is regulated by ERK MAPK, which as we have already discussed is involved in the induction of L-LTP.
Another target of ERK that may transduce a signal to the protein synthesis machinery is ribosomal S6 kinase 2 (RSK2). ERK directly phosphorylates and activates RSK2, which can then act upon the ribosome complex. While the role of RSK2 in regulating protein synthesis is not clear, both it and its target glycogen synthase kinase 3P (GSK3P) have been proposed to regulate protein synthesis through phos-phorylation of ribosome-associated initiation factors (see Figure 15)—one specific candidate in this context is eIF2B (89). Thus, the ERK pathway has been proposed to regulate protein synthesis by a variety of mechanisms still being defined, but overall it is appealing to hypothesize a role for this cascade in regulating neuronal protein synthesis.
Another player implicated in regulating dendritic protein synthesis is PKC (90). Metabotropic glutamate receptors linked to PLC, of course, lead to PKC activation directly and ERK activation indirectly. Bill Greenough's group has found that metabotropic receptors via this pathway affect phosphorylation of ribosome-associated proteins (89). A parsimonious model integrating these findings is given in Figure 15, which shows one possible means of coupling glutamate receptors to protein synthesis in dendrites.
What are the messages regulated by this mechanism? One of the most interesting possibilities is FMRP. FMRP is the protein encoded by the fragile X mental retardation type 1 gene (91). FMRP is an mRNA binding protein that has both a nuclear localization signal and a nuclear export signal—it is hypothesized to be involved in trafficking of mRNAs. Moreover, FMRP colocalizes with polyribosomes in neuronal cell bodies and dendrites. FMRP knockout mice have altered dendritic spine morphology (in cortical neurons at least), which is consistent with a role for FMRP in regulating localized protein synthesis in dendrites. In addition, FMRP knockout animals exhibit alterations in mGluR-induced LTD at CA1 synapses (92). mGluR agonists also can regulate the synthesis of FMRP itself (90). Thus, one clear candidate as a target of local protein synthesis is FMRP, a protein involved in a human mental retardation syndrome. We will return to FMRP in Chapter 10 as well.
While FMRP is a target of the local synthesis machinery, as an mRNA binding protein localized to dendrites, it also likely contributes to regulating local protein synthesis as well. Two known targets of FMRP
Was this article helpful?