II. LTP Induction Component 1—Mechanisms Upstream of the NMDA Receptor That Directly Regulate NMDA Receptor Function
A. The Structure of the NMDA Receptor
B. Kinase Regulation of the NMDA Receptor
C. Redox Regulation of the NMDA Receptor
D. Polyamine Regulation of the NMDA Receptor
III. LTP Induction Component 2—Mechanisms Upstream of the NMDA Receptor That Control Membrane Depolarization
A. Dendritic Potassium Channels
B. Voltage-Dependent Sodium Channels (and Calcium Channels?)
C. AMPA Receptor Function
IV. LTP Induction Component 3—The Components of the Synaptic Infrastructure That Are Necessary for the NMDA Receptor and the Synaptic Signal Transduction Machinery to Function Normally
A. Cell Adhesion Molecules and the Actin Matrix
B. Presynaptic Processes
C. Anchoring and Interacting Proteins of the Postsynaptic Compartment
V. LTP Induction Component 4—Feed-Forward and Feedback Mechanisms That Regulate the Level of Calcium Attained
B. The Spine Apparatus
C. Mitochondrial Calcium-Handling
VI. LTP Induction Component 5—Extrinsic Signals That Regulate the Response to the Calcium Influx
A. The cAMP Gate for LTP Induction
B. The PLC/PKC/Neurogranin System
VII. LTP Induction Component 6—The Mechanisms for the Generation of the Actual Persisting Biochemical Signals
VIII. Summary—Models for Biochemical Information Processing in LTP Induction A. Four-Way Coincidence Detection
As we discussed in the last chapter, one problem with LTP is that it looks simple. You give a hippocampal slice a brief period of high-frequency stimulation, and for all the world it looks like you flipped a light switch and put the synapses in a potentiated state. In Chapter 4, we talked about the consensus that elevation of postsynaptic calcium is what triggers LTP. In this chapter, we will talk about the complex biochemical mechanisms involved in this process.
The typical discussion of LTP induction starts out with something like: "NMDA receptors open, a triggering burst of calcium comes in, and then many complicated LTP induction processes are started." To my embarrassment, I have promulgated that idea many times, but now I realize that it is exactly the wrong way to look at the LTP induction process. Most people, including myself, have historically thought of the LTP induction process as starting with calcium elevation postsynaptically. It is more appropriate to think of the LTP induction process as ending with postsynaptic calcium being elevated. This is because many important biochemical components of the LTP induction machinery are upstream of the NMDA receptor. There probably also are postsy-naptic biochemical events triggered by the initial stages of calcium influx that feed back and regulate later stages of calcium influx.
One point here is that once postsynaptic calcium gets to a sufficient concentration, it will trigger LTP. However, many important mechanisms upstream of reaching the triggering level of calcium determine whether that threshold level of calcium gets reached. These upstream mechanisms are extremely important as well—they are the biochemical computation that the synapse performs in deciding if LTP should be triggered.
A second point is that these processes are mechanisms for sophisticated signal integration at the molecular level, and many of these mechanisms are themselves associative in nature. In this chapter, we will see several examples where the simultaneous presence of two signals leads to a unique event—biochemical coincidence detection. The neuron capitalizes on these biochemical simultaneity detectors, where they are necessary for LTP induction to be triggered, in order to be able to achieve the sort of sophisticated logical operations necessary for the hippocampus to serve as a multimodal signal integrator as we discussed in Chapter 3.
Given the hundreds of individual molecular events that have been reported as being involved in LTP induction in the literature, how can one begin to organize this immense molecular system into a coherent picture? In order to help myself think about this complicated, interactive molecular system, I have begun to think about the biochemistry of LTP induction as comprising several basic systems. Specifically, I have found it useful to think about LTP induction as involving the following components:
1. Mechanisms upstream of the NMDA receptor that directly regulate NMDA receptor function.
2. Mechanisms upstream of the NMDA receptor that control membrane depolarization.
3. The components of the synaptic infrastructure that are necessary for the NMDA receptor and the synaptic signal transduction machinery to function normally.
4. Feed-forward and feedback mechanisms that regulate the level of calcium attained.
5. Extrinsic signals that regulate the response to the calcium influx.
6. The mechanisms for the generation of the actual persisting biochemical signals.
These various stages and components of the LTP induction machinery are schematized in Figure 1. Essentially in this chapter we will be filling in the molecular details necessary to flesh out this general model.
As an editorial aside, I believe that the general lack of appreciation of the role of all these mechanisms, which are upstream of calcium reaching a level sufficient to trigger LTP, is a big part of why there has been so much confusion about LTP induction mechanisms. Perturbing the system at any point can lead to a block of LTP. The typical block-mimic-measure hypothesis testing approach is not sufficiently robust to parse out such a complicated mechanism when the read-out (LTP measured physiologically) is (1) so far downstream and (2) triggered largely in an all-or-none fashion.
Note that we will be discussing mechanisms for the induction of E-LTP by and large, although, of course, the same sophisticated control mechanisms could also participate in regulating the triggering of L-LTP. However, it's not clear to what extent the induction mechanisms for E-LTP and L-LTP are sequential versus parallel in the context of the intact cell. For example, the same signal transduction enzymes that contribute to E-LTP induction locally at a postsynaptic spine compartment likely also contribute to altered gene expression in the nuclear compartment. However, these two processes may be completely sequestered from each other in the cell (or not). The answer to this specific question will, of course, require a great deal of detailed information at the cytoarchitectural level that we simply don't have at this point.
Nevertheless, in some instances it is known, by definition, that there are unique mechanisms contributing to L-LTP induction. We will talk about mechanisms that uniquely contribute to L-LTP induction and expression in Chapter 8, and focus on the complexities of the very earliest steps in the induction of E-LTP and L-LTP in this chapter. In the next chapter (Chapter 7), we will discuss the mechanisms for generating persisting signals that contribute to E-LTP maintenance and expression. Lest you become complacent with how clearly delineated this is starting to sound, keep in mind that some of the persisting signals in E-LTP maintenance (Chapter 7) may well contribute to L-LTP induction as well (Chapter 8), although this is quite speculative at this point.
To re-cap: in this chapter, we will talk about the complex biochemical mechanisms that compute whether the right signals have arrived at a synapse so that LTP should be induced. This is numbers 1-5 in the preceding list. In the next chapter, we will discuss how that decision, manifest as a triggering level of calcium, is translated into a persisting signal subserving E-LTP. This is point number 6 in the preceding list specifically applied to E-LTP. In Chapter 8, we will talk about the additional mechanisms involved in computing whether L-LTP should also be triggered, and the known biochemical mechanisms uniquely contributing to inducing and maintaining that phase of LTP.
Finally, I note once again that there are many different types of LTP in the mammalian CNS—hippocampal, cortical, cerebellar, NMDA receptor-dependent, and -independent, just to name a few prominent categories. I therefore need to specify to the best extent possible exactly which LTP I am discussing. In the next three chapters, I will be discussing NMDA receptor-dependent LTP at Schaffer-collateral/commissural synapses in area CA1 of rat or mouse hippocampus. In most cases, I will be discussing LTP induced using multiple, spaced trains of 1 second, 100-Hz stimuli. I chose this subtype of LTP because it has available the widest variety of direct biochemical data, and this type of protocol induces both E-LTP and L-LTP.
Several sections later in this chapter, I also will discuss LTP induced with theta-frequency stimulation or theta-burst stimulation. This type of LTP induction protocol is most interesting in the context of the signal integration aspects of LTP induction (1-3). LTP induction using these types of protocols is subject to modulation by a wide variety of external signals and regulation of membrane properties. Theta-type LTP provides the richest examples of how the biochemical machinery of LTP induction might be involved in the complex multimodal information processing that the hippocampus performs. Specifically, for this form of LTP, we can see examples of how the synapse has the capacity at the molecular level to ascertain if three, four, five, or more signals are simultaneously present. These are the types of mechanisms that of necessity must underlie the complex associations that the hippocampus is involved in processing at the cognitive level.
II. LTP INDUCTION COMPONENT 1—MECHANISMS UPSTREAM OF THE NMDA RECEPTOR THAT DIRECTLY REGULATE NMDA RECEPTOR FUNCTION
The NMDA receptor is a biochemical signal integrator. It's capacity for signal integration and coincidence detection is not limited to ascertaining the simultaneous presence of glutamate and depolarization. It also senses biochemical signals used in computing the degree of calcium influx that it will allow. In this section, we will discuss the biochemical mechanisms known to regulate the NMDA receptor directly, which have been implicated in LTP induction, memory formation, or both. These biochemical processes, as far as I know at present, are not all-or-none like the glutamate/depolarization mechanism but rather serve to modulate the magnitude of postsynaptic calcium influx. (In the section after this one, we will talk about biochemical processes that are used to control the NMDA receptor in an all-or-none fashion, indirectly through controlling the membrane potential.)
The NMDA receptor is also a temporal integrator, and these mechanisms are limited to biochemical processes, as opposed to biophysical processes. The time frame in which the NMDA receptor can detect the simultaneous presence of depolarization and glutamate is, of course, quite limited because the membrane depolarization is so brief. In contrast, a biochemical signal such as elevation of a second messenger or increased protein kinase activity has a much longer half-life. Thus if the synapse wants to set up a temporal integration mechanism on the seconds (or longer) time-scale, it must use these types of processes. The capacity of the NMDA receptor to be modulated by protein kinases and other messenger molecules allows for this sort of temporal integration as well.
The NMDA receptor is a glutamate-gated cation channel and as such is a multisubunit transmembrane protein. Current models hypothesize that it is a tetrameric hetero-oligomeric protein with more than one glutamate binding site. It is, of course, voltage-dependent, and this arises from the voltage-dependent Mg block of the pore that we discussed in Chapter 4. The protein has binding sites for zinc, polyamines, and glycine (a co-agonist necessary for activity).
Abundantly expressed individual sub-units of the receptor are named NR1, NR2A, NR2B, NR2C, and NR2D—the somewhat complex nomenclature arose for historical reasons related to different groups that cloned the first NMDA receptor subunits. One functional NMDA receptor is comprised of one or more NR1 subunits plus one or more NR2-type subunits. The NR2 subunits determine the calcium permeability of the channel and can influence the voltage-dependence of its activation, kinetics of opening, and other biophysical properties. NR1 and NR2A and 2B are phosphorylated at a number of different sites—NR1 by PKC and the cyclic AMP dependent protein kinase (PKA), NR2A by cyclin-dependent kinase 5, and NR2A and NR2B by various tyrosine kinases such as src and fyn.
The NMDA receptor is subject to a wide variety of direct modulatory influences, some of which are listed in Table 1. Please keep in mind that not every single biochemical factor known to man to influence the NMDA receptor is listed in Table 1. The table is limited to the intersection of those things that (1) I know about and (2) have been experimentally implicated as being involved in LTP per se. The same general rule applies to all the tables included in this chapter, and I welcome feedback on any missing items.
One of the oncogene products produced by the Rous sarcoma virus, which also has a homologue in the mammalian genome, is the tyrosine kinase src. Src family tyrosine kinases like src and fyn directly
Src family tyrosine kinases (src, fyn)
Was this article helpful?
A course in techniques and skills for mentalists, magicians and students. For students, improve your grades with less effort! But this book is also.... The ideal for any stage mentalist or magician by establishing credibility of amazing skills with an easy to follow instructional book on using the amazing power of your memory.