|Synaptic Reactivation: The Role of Tubulins||Memory|
The mind is restless. It can still wander from one thought to another, even in the absence of any sensory stimuli. Such phenomenon has several alternative names. If it occurs while awake, it may be called mind wandering, day dreaming, stimulus-independent thought, or spontaneous thought. During light sleep, the mind wandering is simply called a dream. In deep sleep, the mind (or thought) does not exist because we are not conscious during this period (see Chapter 11). However, neurons in a few brain regions still fire spontaneously and intensively to consolidate memory.
Spontaneous Synaptic Reactivation
The mind wandering, dreaming or spontaneous neuronal firing is correlated with the replay of past experiences (Euston et al., 2007; Peyrache et al., 2009; Buhry et al., 2011; Wamsley and Stickgold, 2011). At the synaptic level, this implies that the synapses activated by previous experiences can be reactivated spontaneously. The spontaneous synaptic reactivation may emerge in the neocortex, hippocampus, or amygdala. It can provide information about which synapses should be consolidated. Indeed, the synaptic reactivation has been demonstrated to play a key role in consolidating labile memory traces into long-term memory (Hoffman and McNaughton, 2002; Carr et al., 2011; Born and Wilhelm, 2012).
How can the synapses be spontaneously reactivated? The answer may lie in tubulins.
Regulation of Neuronal Excitability by Tubulins
A microtubule is composed of α and β tubulin heterodimers. The α tubulin contains 65 acidic residues (aspartate and glutamate) and 40 basic residues (lysine and arginine) while in the β tubulin 62 residues are acidic and 37 residues are basic (Minoura and Muto, 2006). Therefore, at the physiological pH value, a tubulin has 25 net negative residues. The negative electrostatic potentials are distributed over the entire microtubule (Baker et al., 2001). In the cytosol, this negative field is strong enough to induce layers of ordered water up to 200 nanometers from the microtubule surface (Pokorný, 2004).
Tubulins are present in the postsynaptic density (PSD), a protein complex located underneath the postsynaptic membrane of the dendritic spine. By using the Nano-Depth-Tagging method, the positions of various proteins in PSD have been determined (Yun-Hong et al., 2011, Figure 9). Tubulin is located just underneath the postsynaptic membrane where ligand-gated ion channels are located. One of these channels is the NMDA receptor (NMDAR). While NMDAR is intrinsically ligand-gated (by glutamate), its open probability also depends on Mg2+ block (Mayer et al., 1984). A depolarizing field may repel the Mg2+ ion out of the pore to facilitate channel opening, whereas a hyperpolarizing field can strengthen the Mg2+ block.
The Mg2+ block plays a critical role in preventing seizures. By removing Mg2+ ions from the medium, the rat hippocampal slice exhibited spontaneous seizure-like events (Anderson et al., 1986). Hence, the highly negatively charged tubulin may stop seizures by enhancing Mg2+ block. In patients with temporal lobe epilepsy, significant reduction of α and β tubulin has been observed (Yang et al., 2006). More specifically, their expression at PSD is down-regulated (Conference Abstract, 2009). Furthermore, mutations in the β-tubulin gene TUBB2A cause infantile-onset epilepsy (Cushion et al., 2014).
It has been shown that, during long-term potentiation (LTP), microtubules may polymerize all the way to PSD and, within 20 seconds to 30 minutes, depolymerize back to the dendritic shaft (more info). Some of the depolymerized tubulins could be captured into PSD by CaMKII, which is known to interact with tubulins (Yoshimura et al., 2000). The translocation of tubulins into PSD may prevent over-excitation. More importantly, these tubulins may trigger synaptic reactivation as they exit PSD.
Reactivation by Tubulin Exit from PSD
Many plasticity-related proteins, including tubulin, are retained in PSD by interacting with CaMKII which, upon LTP induction, is anchored by NMDAR. The CaMKII Inhibitor (CaMKIIN) can disrupt the binding between CaMKII and NMDAR, thereby reducing the amount of CaMKII in PSD (see Chapter 5). Without CaMKII, tubulins will also exit PSD, but AMPARs may not be immediately affected by the departure of CaMKII, because they are embedded in the postsynaptic membrane (Cheng et al., 2009). The interaction between CaMKII and tubulins explains why CaMKII is implicated in neuronal excitability and epilepsy (Liu and Murray, 2012).
Spontaneous glutamate release from the presynaptic axon terminal is quite common (Espinosa and Kavalali, 2009; Sara et al., 2011). The LTP induction results in high level of AMPARs at the postsynaptic membrane. However, in the presence of tubulins at PSD, the spontaneous glutamate release is too small to generate significant excitatory postsynaptic potential (EPSP). When tubulins exit PSD while the AMPAR level remains high, the synapse can be reactivated by the spontaneous glutamate release. The more AMPARs at the postsynaptic membrane, the stronger the synapse will be reactivated.
CaMKIIN increases rapidly in the hippocampus and amygdala after learning (Gouet et al., 2012). The CaMKII inhibitor is also upregulated significantly in the cerebral cortex during sleep (Chiara Cirelli, personal communication). Thus, the memory traces maintained by CaMKII are constantly erased by CaMKIIN. These labile memory traces will be lost forever unless they are consolidated into long-term memory which, according to the MTT Hypothesis, is encoded in the microtubule tracks transporting PSD-95. Trivial memories should not be consolidated. The importance of a memory trace is indicated by the strength of reactivation. Chapter 10 will show how the microtubule tracks can be constructed only to the strongly reactivated synapses.
Author: Frank Lee