Targets Of The Calcium Trigger

CaMKII and CaMKIV. CaMKIV is most likely involved in regulating neuronal gene expression, and we will return to this molecule in the next chapter. CaMKII has achieved especial notoriety for its importance in the induction, maintenance, and expression of LTP, as we will discuss in the next section. Before proceeding, I point out that a particularly outstanding review of the role of CaMKII in LTP has been published recently by John Lisman, Howard Schulman, and Holly Cline (6). I refer you to this review for additional details and insightful analysis.

The Ca2+/calmodulin-dependent protein kinase II (CaMKII) is enriched in the brain and exhibits multifunctional roles in calcium-mediated signal transduction processes. CaMKII is composed of homologous alpha and beta subunits with a size of 52 and 60 kDa, respectively. The CaMKII holoenzyme (i.e., the functional structure)

is a dodecamer (12 subunits) and is a mixture of both alpha and beta subunits. The individual subunit structure and the structure of the holoenzyme are given in Figure 3. Each subunit comprises three domains: an interaction domain that allows formation of the holoenzyme complex, a regulatory domain that binds calmodulin and regulates the enzyme's activity, and a catalytic domain that executes the phospho-transfer reaction from Mg2+/ATP to the substrate protein.

The activity of CaMKII is highly sensitive to calcium influx, and the calcium-dependency of activation has an absolute requirement for calmodulin. In the simplest mode of regulation of CaMKII activity, calcium binds to calmodulin, the complex activates CaMKII, and enzymatic activity returns to baseline after calcium levels diminish (see Figure 4). This obviously can be relevant to LTP induction but is not a

Transient CaMKII Activation

Receptor r

NMDAR Association (persistently active)

FIGURE 4 Three different effects of Ca/CaM on CaMKII. Calcium and calmodulin can produce CaMKII activation by various mechanisms, with the duration of the activation differing among the mechanisms. CaMKII can be transiently activated by direct binding of Ca/CaM—the activation terminates when calcium levels return to baseline. CaMKII can also be translocated to the NMDA receptor, independent of autophosphorylation, which leads to a calcium-independent activation lasting seconds to minutes. CaMKII which undergoes autophos-phorylation at Thr286 can be rendered active independent of Ca/CaM, which is a mechanism for persistent activation. This persistently activated, autophosphosphorylated CaMKII can also associate with the NMDA receptor as well. See text for additional discussion.

mechanism for generating a persistently activated enzyme that can serve as an LTP maintenance molecule.

However, the activity of the enzyme and its sensitivity to successive increases in calcium concentrations are altered following CaMKII autophosphorylation. Autophospho-rylation, the act of a kinase phosphory-lating itself, occurs in CaMKII in response to calcium/calmodulin stimulation. Changes in the phosphorylation state of the alpha or beta subunits of CaMKII alter the kinetic properties of the holoenzyme—the enzyme becomes active independent of any need for calcium/calmodulin. Thus, autophos-phorylation of CaMKII can generate a persistently active (aka autonomously active) enzyme—a persisting biochemical signal in response to a transient intitiating event!

The principal autophosphorylation site, and the site that renders the enzyme autonomously active, is Thr286 in CaMKII alpha (and the homologous Thr 287 site in the beta subunit). Autophosphorylation at this site occurs via an intraholoenzyme but intersubunit reaction. In other words, individual CaMKII subunits phosphory-late their neighbors, but not themselves. Thus, CaMKII autophosphorylation is self-delimited to a single 12-subunit holoenzyme. Activation of CaMKII by Ca2+/calmodulin causes the alpha and beta subunits to undergo autophosphorylation at Thr286/287, rendering the enzyme autonomously active and partially insensitive to further increases in Ca2+/CaM concentrations. An additional consequence of autophosphory-lation is that activated CaMKII associates with NMDA receptors (7, 8), a process that also can lead to persistent activation of the enzyme (see Figure 4). Thus, autophospho-rylation at the Thr286 site leads to a persistently activated CaMKII enzyme, localized postsynaptically at its site of activation—an ideal molecular memory trace.

How is it that the autophosphorylation of CaMKII is maintained in the face of protein phosphatases that can reverse the autophosphorylation? Current models2 (9) capitalize on data indicating that protein phosphatase activity is low in the PSD, where the autophosphorylated CaMKII is located. Thus, all that needs occur is that the rate of intersubunit autophosphory-lation be greater than the net rate of CaMKII dephosphorylation in order for the phosphorylation to persist for the life of the enzyme. Thus, the capacity for intersubunit trans-phosphorylation synergizes with a low level of phosphatase activity to give a persisting signal in the cell.

The necessity of CaMKII autophos-phorylation in LTP is reasonably well established. The induction of NMDA receptor-dependent LTP requires CaMKII activation in the postsynaptic neuron (10, 11), and mice deficient for alpha CaMKII show deficits in hippocampal LTP (12, 13). The sites of CaMKII autophosphorylation are also important in LTP induction. Mutations of the Thr286 site to prevent autophosphorylation or, conversely to produce a calcium-independent form of CaMKII, result in LTP deficits for some types of LTP-inducing stimulation (14). Persistently activated CaMKII, and indeed increased CaMKII autophosphorylation at Thr286, have been demonstrated in LTP.

Moreover, activation of NMDA receptors results in translocation of CaMKII from the cytosol to the postsynaptic density regions, and LTP-inducing stimulation triggers a transient translocation of CaMKII from the cytosol to the PSD that is largely dependent on the autophosphorylation state of the CaMKII at Thr286. Finally, injection or

2In the mid-1980s there was much excitement about the idea that autophosphorylated CaMKII might serve as a self-perpetuating signal that could subserve permanent memory storage. We will return to this idea in more detail in Chapter 12. The bottom line, however, is that a variety of experimental results suggest that perpetual activation of CaMKII does not occur with LTP-inducing stimulation. Direct assays for autonomously activated CaMKII indicate that the autophosphorylated enzyme only persists for 1 to 2 hours in the cell.

transfection of autonomously active CaMKII into neurons likewise leads to enhancement of synaptic strength and an occlusion of LTP induced by tetanic stimulation, further data consistent with a role of autonomously active CaMKII in E-LTP maintenance.

There is, however, one fly in the ointment concerning the present model for a role for CaMKII as an LTP maintenance molecule—CaMKII inhibitors applied after LTP-inducing stimulation may not reverse LTP (15). How can one rationalize that a compound that blocks CaMKII phospho-transferase activity does not lead to a reversal of LTP? One possibility is that CaMKII is playing a structural role in LTP that does not necessitate phosphorylation of substrates per se. Another possibility is that simultaneous CaMKII and PKC activity are triggered in LTP, and either is sufficient for E-LTP expression. Finally, it is possible that phosphatase inhibition synergizes with CaMKII autophosphory-lation in LTP maintenance and expression, so that even if CaMKII phosphotransferase activity is blocked, the synapse can still stay potentiated for some period of time before substrate dephosphorylation occurs. (None of these possibilities excludes any of the others.) Thus, some mechanistic details still need to be worked out regarding the role of CaMKII as a persisting signal in E-LTP. Regardless, autonomously active, autophos-phorylated CaMKII still serves as the prototype molecular information storage device at present.

Inhibitory Autophosphorylation of CaMKII

Additional in vitro experiments indicate that autophosphorylation of CaMKII at additional sites can also occur. Specifically, autophosphorylation of Thr 305/306 sites on alpha and beta CaMKII occurs following Thr286 autophosphorylation under circumstances of prolonged or robust stimulation. This can inhibit both the calcium/CaM-independent activity and the calcium/CaM-dependent activity. Thr305/

306 autophosphorylation also leads to dissociation of the enzyme from the PSD. We will return to a possible role for this event in a human mental retardation syndrome, Angelman Syndrome, in Chapter 10.

CaMKII as a Temporal Integrator

Finally, it is important to note that while CaMKII can serve as an information storage molecule in E-LTP maintenance, a role for this mechanism is not limited to longer-term information storage. CaMKII can by similar mechanisms serve to allow temporal integration between spaced periods of NMDA receptor activation. That CaMKII can by itself serve as a temporal integrator was elegantly demonstrated in a series of studies by DeKoninck and Schulman (16). When CaMKII sees repeated, spaced pulses of calcium and calmodulin, it integrates these signals and gives a readout of calcium spike frequency in terms of autonomous CaMKII activity (see Figure 5). Thus, increasing frequencies and levels of calcium give increased autonomous

FIGURE 5 Temporal integration of a Ca signal by CaMKII, as a function of availability of calmodulin. Graphical representation of autonomous CaMKII activity illustrating dependence on the frequency of calcium concentration spikes and the concentration of calmodulin. Adapted from data in De Koninck and Schulman (16). Reproduced with permission from Dineley et al. (126).

FIGURE 5 Temporal integration of a Ca signal by CaMKII, as a function of availability of calmodulin. Graphical representation of autonomous CaMKII activity illustrating dependence on the frequency of calcium concentration spikes and the concentration of calmodulin. Adapted from data in De Koninck and Schulman (16). Reproduced with permission from Dineley et al. (126).

CaMKII activity. In this amazing example, all that is needed for temporal integration is a single molecular complex sensing the ambient level of free calcium. This is likely a means by which multiple, spaced tetanic stimuli are able to selectively produce unique long-lasting effects on CaMKII activity postsynaptically. The generation of autonomous activity is also dependent on the level of free calmodulin, recalling the possible role of the PKC/neurogranin gate in regulating CaMKII, which we discussed at the end of the last chapter.

Two Additional Targets of CaM: Adenylyl Cyclase and NOS

Adenylyl cyclase (AC), the enzyme that converts ATP to the second messenger cAMP, is also a target of CaM in neurons. The CaM-sensitive AC isoforms are AC1 and AC8, and work from Dan Storm and his colleagues has demonstrated that simultaneous knockout of these two genes gives a pronounced LTP phenotype (see Figure 6, reviewed in 17). Both E-LTP and L-LTP are affected in the double knockouts. Calcium/calmodulin stimulation of AC likely contributes to regulating the cAMP gate in a temporal integration fashion (see the last chapter), and this may be the mechanism of attenuation of E-LTP in the AC knockout mice. However, there is no evidence that there is any persisting activation of AC that contributes to E-LTP maintenance (18) so AC likely serves only in the induction phase of E-LTP. L-LTP is completely lost in AC knockout mice, and we will return to the important role of AC in L-LTP induction in the next chapter.

Nitric Oxide Synthase also is CaMsensitive. When NOS is activated by CaM, it converts arginine to citrulline plus the free radical species NO (nitric oxide). It is clear that generation of NO through NOS activation modulates LTP induction, but the precise mechanisms by which this happens are still being worked out (19). As we discussed in the last chapter, NMDA receptors are modulated by NO, and

NO-sensitive guanylyl cyclase has also been implicated as a target of NO in LTP induction, although there has been substantial argument over the role of this mechanism in LTP (see references 20, 21, and 22). The ras/ERK cascade is also a potential target of NO. These are all relevant mechanisms for NO in modulating LTP induction, but as with AC it seems unlikely that there is any role for ongoing NOS activation in E-LTP maintenance.

One of the most interesting potential roles for NO derives from its capacity to cross cell membranes directly. Thus, NO like other membrane-permeant species could be a "retrograde messenger" that carries a signal from the postsynaptic compartment to the presynaptic compartment (23). One specific retrograde messenger role that has been proposed for NO is activation of presynaptic guanylyl cyclase and the cyclic GMP-dependent protein kinase (PKG; 24).This proposed role also is limited to LTP induction as well and has not been proposed to serve an LTP maintenance function. However, presy-naptic NO and NO-derived reactive species might also be particularly important in generating persisting signals presynap-tically. We will return to this idea later when we talk about the possible role of oxidatively modified PKC in E-LTP maintenance.

In summary, then, both AC and NOS are additional targets of CaM in E-LTP induction. They serve as important triggering mechanisms, involved in generating persisting signals. They are not, however, persistently activated and do not directly participate in E-LTP maintenance.

B. A Second Target of Calcium: PKC

The calcium/phospholipid-dependent protein kinases (PKCs) are pluripotent regulators of synaptic transmission and neuronal function. PKC has not been as extensively studied as its cousin CaMKII in the context of LTP; nevertheless, there is a fairly broad literature implicating PKC as a

FIGURE 6 LTP in adenylyl cyclase-deficient mice. DKO mice lacking both the AC1 and AC8 calmodulin-sensitive isoforms of adenylyl cyclase have a defect in LTP. (A) LTP was induced in AC1 mutant (AC1-M, shaded triangles; n = 9 mice, 17 slices) and AC8 mutant mice (AC8-M, open diamonds; n = 13 mice, 25 slices) slices by four tetanic trains of 100-Hz (200 ms, 6 seconds apart) stimulus. There was no statistically significant difference between the mean fEPSP of AC1-M and AC8-M mutant mice 180 minutes after tetanization (p = .28). Representative fEPSP traces taken from area CA1 in AC1-M and AC8-M slices at 1 minute and 180 minutes after tetani are shown in the insets. Each trace is superimposed over a baseline trace taken 2 minutes before tetani for ease of comparison. Scale bar, 1 mV, 10 ms. (B) LTP was diminished in DKO mice. Wild-type mice (closed circles; n = 13 mice, 26 slices) preparations gave an L- LTP lasting up to 180 minutes, whereas potentiation in the DKO (open circles; n = 8 mice, 19 slices) preparations declined to near baseline values within 80 minutes. Representative fEPSP traces taken from area CA1 in wild-type and DKO brain slices at 1 minute and 180 minutes after tetani are shown in insets. The dashed lines represent an arbitrary cutoff (120%), below which potentiation was considered to be near baseline. Scale bar, 1 mV, 10 ms. Figure and legend reproduced from Wong et al. (127).

molecule contributing to the maintenance of E-LTP. PKC inhibitors can block the expression of E-LTP (25-28), PKC is persistently activated in E-LTP (29-36), and activation of PKC or injection of the enzyme into the postsynaptic cell elicits synaptic potentiation (37-40; reviewed in Weeber et al. (42); see Box 1). Thus, the molecule meets the three principal criteria establishing it as a candidate E-LTP maintenance molecule.

In mammals the PKC enzyme family is quite heterogeneous and comprises 11

Mind and Memory Mastery

Mind and Memory Mastery

Tap into the memory secrets of geniuses… Discover The Untold Mind And Memory Secrets Used By The World’s Smartest People To Help You Unleash Your Full Potential! Finally You Can Fully Equip Yourself With These “Must Have” Mind And Memory Tools For Creating Your Ideal Lifestyle! Let’s face it, We all know the fact that we aren't using more than 10% of our brains. If we could, we'd be achieving much more and giving our fullest gifts to the world.

Get My Free Ebook

Post a comment