Geon Long-Term Memory: PKC and Microtubule Tracks MT

 

Protein kinase C (PKC) is a promising target for the treatment of Alzheimer's disease, because mounting evidence has suggested that it plays a keyl role in long-term memory (LTM) (Sun and Alkon, 2012). This chapter will present evidence that LTM could be encoded in the microtubule tracks transporting PSD-95, which is regulated by PKC.

PKC Is Critical for Long-Term Memory

PKC has several isoforms. Bryostatin-1 is specific for PKC α and ε. Another PKC isoform, PKMζ, was proposed to be a critical memory molecule (Sacktor, 2011). However, at the beginning of 2013, two papers published in Nature have shown that PKMζ is not required for learning and memory (Volk et al., 2013; Lee et al, 2013).

PKC Regulates Microtubule Dynamics and PSD-95 Expression

PKC activation has been shown to promote microtubule advance in neuronal growth (Kabir et al., 2001). In dendrites, microtubule polymerization requires phosphorylation of the microtubule-associated protein 2 (MAP2), which is controlled by PKC (Ainsztein and Purich, 1994; Quinlan and Halpain, 1996). Therefore, LTM could be created and maintained by PKC partly through its regulation on microtubule dynamics. More importantly, PKC also stimulates PSD-95 expression and induces synaptogenesis.

PSD-95 is a scaffold protein, capable of recruiting other plasticity-related proteins (PRPs) to form a synapse (Keith and El-Husseini, 2008). It plays a critical role in the synaptogenesis mediated by NF-κB (Boersma et al., 2011). The PKC activator, Bryostatin, has also been demonstrated to boost not only PSD-95 expression, but also synaptogenesis (Mogha et al., 2012). Therefore, the existence of a synapse at a dendritic site depends on the availability of PSD-95 at that site. PSD-95 is synthesized in the soma and then transported to dendrites via microtubules. If a microtubule track (MTT) has been constructed for the transport of PSD-95 to a dendritic site which does not have a synapse yet, a new synapse can be created at that site. Conversely, if a synapse cannot get PSD-95, the synapse will become silent (Zhao et al., 2013).

In the spine, PSD-95 is highly labile. It constantly moves into and out of the spine, exchanging with PSD-95 in neighboring spines by diffusion (Gray et al., 2006). Any memory trace stored in a specific spine will be lost within days. Therefore, the spines alone are unlikely to store memory that lasts for years. Since PSD-95 is transported by microtubules, the long-term memory could be encoded in specific MTTs which deliver PSD-95 to target sites (see The Targeted Transport of PSD-95).

The Microtubule Track (MTT) Hypothesis

The memory of an event arises from the activation of a group of synapses, which depend on synaptic strength. According to the MTT hypothesis, the synaptic strength is determined by three major factors: MTTs, PSD-95 expression and CaMKII activity. The long-term memory is encoded in MTTs, while the short-term fluctuation of the synaptic strength is shaped by PSD-95 expression and CaMKII activity.

MTTs. Most PRPs and their mRNAs do not need specific tracks for their delivery to any particular spines. PSD-95 is probably the only one which requires targeted transport. Other PRPs which are diffusing randomly in dendrites can then be recruited by PSD-95 at the target site. Since PSD-95 has the capability to form new synapses, the memory of an event should always exist as long as the MTTs for the transport of PSD-95 to corresponding spines remain intact, even after these spines have been destroyed! The spines affect how easy an event can be recalled, whereas the MTTs for PSD-95 dictate the existence of the memory.

PSD-95 expression. Once a specific MTT for the delivery of PSD-95 to a synapse has been established, the synaptic strength can be modulated by the level of PSD-95 expression. The more PSD-95 are synthesized in the soma, the more they will be delivered to the synapse, where they can retain AMPAR at the synapse, thereby enhancing the synapse. Consistent with this view, it has been demonstrated that overexpression of PSD-95 increases the AMPAR-mediated currents (Béïque and Andrade, 2003).

CaMKII activity. Without stimulation, CaMKII may diffuse randomly in dendrites. Some of them could be recruited into spines by PSD-95. Upon stimulation, those in the dendritic shaft could be carried by microtubules into the spines and capture other PRPs (see CaMKII Entry into Spines and Synaptic Tagging). However, the memory traces maintained by CaMKII cannot last very long, due to the constitutively active CaMKII inhibitor, CaMKIIN (Gouet et al., 2012). It has been found that CaMKIIN is significantly up-regulated during sleep (Chiara Cirelli, personal communication). Hence, a substantial fraction of memory traces in the spines will be erased during sleep. The memories without being encoded in MTTs will be lost forever, but those encoded in MTTs can still be recalled, depending on CaMKII activity and PSD-95 expression.

It is important to note that CaMKII requires Ca2+/calmodulin for activation, which is usually induced by the Ca2+ entry through NMDAR. Thus, the modulation of synaptic strength by CaMKII is stimulation-dependent. By contrast, PSD-95 can modulate the synaptic strength in its native form. This allows PKC to maintain the basic synaptic strength in the absence of any stimulation.

Since PKC plays a central role in LTM, the memory consolidation during sleep should involve PKC activation. Further details are discussed in the next three chapters.

 

Author: Frank Lee
First published: April, 2013
Last updated: May, 2013