The Molecules Of

Recent studies have given a number of important insights into the molecular mechanisms underlying cerebellar LTD. Cerebellar LTD induction requires mGluR and AMPAR activation, and postsynaptic Ca2+ influx through voltage-gated Ca2+ channels (reviewed in reference 10). The mGluR activation produces DAG, which subsequently along with calcium activates PKC. This PKC activation is necessary for the induction of cerebellar LTD (60), as has been shown in a variety of inhibitor studies, and PKC activators mimic LTD (i.e., they cause synaptic depression).

In an interesting mirror image of hippocampal LTP, AMPAR internalization appears to underly cerebellar LTD. Interference with clathrin endocytosis blocks the induction of cerebellar LTD, and induction of AMPAR internalization produces an LTD that occludes stimulus-induced LTD (61, 62). PKC phos-phorylates the GluR2/3 AMPA receptor at serine 880 (serine 885 in GluR3), and this phosphorylation decreases the binding of the GluR to glutamate receptor-interacting protein/AMPAR binding protein (GRIP/ABP) (63). As we have already discussed, GRIP is a PDZ domain containing protein, which serves as an adaptor to cross-link AMPAR

to other neuronal proteins including cytoskeletal elements. Thus, LTD requires PKC phosphorylation of the GluR2/3 receptors, which regulates intereactions with several PDZ domain-containing proteins and apparently controls their rate of internalization.

Although the mechanisms are unclear at this point, there also are different molecular stages of cerebellar LTD. Later stages of LTD are blocked by protein synthesis inhibitors, just like hippocampal LTP. Also, late LTD is dependent upon CREB activation, a process that is dependent on CaMKIV activation (64, 65).

Again, these molecular studies allow us to make generalizations about the molecular basis of learning and memory. The same signal transduction processes (PKC in this case) can lead to either synaptic strengthening or weakening depending on the cellular context in which that process is imbedded. Signal transduction mechanisms are used for information processing at synapses, detecting cell surface signals and translating them into the appropriate change in cellular properties. Finally, that multiple molecular mechanisms trigger biochemical processes of differing durations is the rule rather than the exception.

In Figure 9 is a simplified diagram of the hippocampus—only area CA1 is specified to any extent. The hippocampus receives "sensory input" and neuromodulatory "arousal/attention/emotion" signals. These are simply a lumping together of the various inputs to the hippocampus and area CA1 that we have discussed extensively in previous chapters. The

"consolidation signal" designates activity that is triggered externally or internally as part of the consolidation process—it basically represents the AMPA/kaninate receptor-dependent activity that is known to be necessary for memory consolidation. The consolidation signal could also involve several neuromodulatory systems, as Jim McGaugh's work has highlighted.

FIGURE 9 A simple model for how LTP might participate in memory consolidation. See text for discussion. (A) Activity in a sensory input plus activation of an ACh input into area CA1 (142). PP/DG/MF = inputs via the perforant path, dentate gyrase, and mossy fiber Pathway. EC/PC = entorhinal cortex / perirhinal cortex. Note also that a recall signal routed through the cerebral cortex elicits a given pre-training behavioral output. (B) LTP at a set of synapses in area CA1. (C) A consolidation signal played through the potentiated pathway results in synaptic potentiation in downstream synapses, including in the cerebral cortex. (D) L-LTP is now established at specific synapses in the cortex. (E) Sending a recall signal through the modified cortical synapses elicits a modified behavioral output. (F) Even when synaptic potentiation is lost in the hippocampus and its immediate targets, the modified behavior persists.

FIGURE 9 A simple model for how LTP might participate in memory consolidation. See text for discussion. (A) Activity in a sensory input plus activation of an ACh input into area CA1 (142). PP/DG/MF = inputs via the perforant path, dentate gyrase, and mossy fiber Pathway. EC/PC = entorhinal cortex / perirhinal cortex. Note also that a recall signal routed through the cerebral cortex elicits a given pre-training behavioral output. (B) LTP at a set of synapses in area CA1. (C) A consolidation signal played through the potentiated pathway results in synaptic potentiation in downstream synapses, including in the cerebral cortex. (D) L-LTP is now established at specific synapses in the cortex. (E) Sending a recall signal through the modified cortical synapses elicits a modified behavioral output. (F) Even when synaptic potentiation is lost in the hippocampus and its immediate targets, the modified behavior persists.

The output from area CA1 goes to the entorhinal cortex and perirhinal cortex and these are treated in block fashion as a way-station to the cerebral cortex. The entirety of the cerebral cortex is reduced to two neurons in the model, which is about how many functioning cortical neurons I have left after a long day of writing.

When memory is to be assessed, the cortex receives a "recall signal"—you might imagine this as an environmental signal such as being asked the question, "What is the capital of Alabama?" Before training, your "behavioral output" is the answer "Birmingham." After training and the ministrations of your hippocampus, your behavioral output is altered, and your answer becomes "Montgomery."

How might the hippocampus bring about this change, capitalizing (so to speak) on its capacity for LTP? You receive sensory input in the form of new information, such as the preceding sentence, manifest as the firing of hippocampal neurons. The importance of the information is clear, so your hippocampal synapses receive a blast of neuromodulatory neurotransmitter (i.e., your professor just told you that the question will be on the exam). The simultaneous activity of neuronal projections into area CA1 and the neuromodulatory signal leads to the formation of LTP at one (for our purposes) of your Schaffer/collateral synapses. These events are diagrammed in Figure 9A, B.

If you are going to store this new information long enough to make it to the test next week, your hippocampus has to receive the "consolidation signal" (Figure 9C). This relays neuronal activity through the hippocampus, including your newly potentiated CA1 synapse. This heightened firing of one of your CA1 pyramidal neurons is manifest as a potentiation of synapses downstream in the entorhinal cortex/perirhinal cortex, and ultimately potentiation of a synapse downstream of there in one of your two functioning cortical neurons (Figure 9D).

The potentiated synapse in your cerebral cortex is the result of a very long-lasting variety of LTP.5 The potentiated cortical synapse participates in a network of neurons storing information. Its potentiated state leads to an altered behavioral output when its circuit receives a recall signal (Figure 9E). Your cortical circuit stays altered long past the duration of the potentiation in the hippocampus (Figure 9F). Hippocampal synapses are free to relax back to their original state—the hippocampus after all is not storing the memory long-term, just participating in consolidating the memory in the cortex. Thus, your hippocampal synapses return to an unpotentiated state, but their potentiation was an absolute requirement for storing the memory in the cortex.

Now exam day arrives, and your behavior has been modified appropriately as a result of sensory input—in this case the correct information concerning the capital of Alabama. You receive the recall signal— "What is the capital of Alabama?" Your two functioning cortical neurons fire and give the modified behavioral output— "Montgomery." You make a 100 on the test and go on to get your Ph.D. in Neuroscience—all because of a little hippocampal LTP.

Please keep in mind that the model is deliberately oversimplified. I made it up using the absolute minimum number of components that I could, simply to illustrate the basic idea of a role for the hippocampus in cortical memory consolidation. Nevertheless, it is an interesting exercise to expand upon the model yourself. What happens if you potentiate two synapses in the hippocampus? What

5This aspect of the model is probably accurate. Long-lasting memory storage in the cortex may well involve LTP or LTP-like processes as well (53, 54). Ultra-long-term LTP of the sort we discussed at the end of the last chapter may be the mechanism for cortical information storage, triggered by the potentiated outputs of the hippocampus and entorhinal cortex.

constraints do you place on the model if you only allow yourself to use homosy-naptic LTP? How can you increase the complexity of the information processing capacity if you add more details to the dentate gyrus/CA3 region? Exploring these various options, while still keeping the model quite simple, is an edifying exercise in the power of synaptic plasticity in information processing.

Was this article helpful?

0 0
Eliminating Stress and Anxiety From Your Life

Eliminating Stress and Anxiety From Your Life

It seems like you hear it all the time from nearly every one you know I'm SO stressed out!? Pressures abound in this world today. Those pressures cause stress and anxiety, and often we are ill-equipped to deal with those stressors that trigger anxiety and other feelings that can make us sick. Literally, sick.

Get My Free Ebook


Post a comment