|The Mechanism of Hippocampal Replay||MT|
During sleep or awake rest, the hippocampus may replay past experiences (Buhry et al., 2011). At the cellular level, this means that the synapses activated by previous experiences are reactivated, usually in the form of sharp wave ripples (SWRs) (Buzsáki, 1986). The hippocampal replay is a major means for memory consolidation (Karlsson and Frank, 2009; Ego-Stengel and Wilson, 2010; Nokia et al., 2012). This article will focus on the molecular mechanism underlying SWR and its counterpart in the epileptic hippocampus: fast ripples.
Sharp Wave Ripples
SWRs are high-frequency (140 - 250 Hz) oscillations which originate in the CA3 of the hippocampus, and then propagate to CA1 and other brain regions. They could be induced by tubulins departing from postsynaptic density (PSD). As mentioned in another article, tubulins are highly negatively charged. During long-term potentiation (LTP), CaMKII recruits various plasticity-related proteins into PSD (more info), including tubulins which make the postsynaptic potential more hyperpolarized, thereby inhibiting activation of synapses even with high level of AMPA receptors (AMPARs).
SWRs depend on activities mediated by AMPAR (Wu et al., 2005). It has been demonstrated that SWR-like oscillations can be generated in hippocampal slices perfused with magnesium-free medium (Papatheodoropoulos, 2007), suggesting that, with adequate AMPAR level, the relief of Mg2+ block and spontaneous glutamate release (Espinosa and Kavalali, 2009; Sara et al., 2011) are sufficient to activate the neuronal population responsible for the high-frequency oscillations. After LTP, the AMPAR level is high. If somehow the tubulin level at PSD is reduced, the membrane should become more depolarized, allowing the spontaneous glutamate release to activate both AMPAR and NMDAR, resulting in SWRs.
Now, the question is: what causes tubulins to exit PSD?
The Role of CaMKII Inhibitor
Many plasticity-related proteins, including tubulin, are kept in PSD by interacting with CaMKII which in turn is anchored by NMDAR. The CaMKII Inhibitor (CaMKIIN) can interrupt the binding between CaMKII and NMDAR, thereby reducing the amount of CaMKII in PSD (Gouet et al., 2012). CaMKIIN is regulated in an experience-dependent manner. Within 30 minutes after learning, CaMKIIN is up-regulated by the expression of its mRNA (Lepicard et al., 2006). However, the inhibition of CaMKII by CaMKIIN is independent of synaptic activity. Once CaMKIIN is produced, it will exert its function: removing CaMKII and other plasticity-related proteins from PSD. Hence, the major role of CaMKIIN seems to erase memory traces in spines.
The "synaptic homeostasis hypothesis" (SHY) postulates that the role of sleep is to downscale synaptic strength to a baseline level that is energetically sustainable (Tononi and Cirelli, 2006). While SHY is supported by substantial evidence, a question was raised (Frank, 2012): what is the underlying mechanism? CaMKIIN could be the answer.
At present, it is unclear whether CaMKIIN serves as a general homeostatic regulator or erases only selective memory traces. If the former is true, the long-lasting memory traces cannot be stored in spines. Most likely, they are encoded in microtubule tracks (more info).
Fast ripples (250 - 600 Hz) are observed in the epileptic hippocampus, but none of hippocampal neurons can oscillate at such high frequencies. Evidence suggests that the recroded high frequency arises from more than one population of synchronized neurons oscillating out of phase (Ibarz et al., 2010). Thus, two groups of neurons oscillating at 200 Hz out of phase by about half of a cycle will appear as oscillating at 400 Hz.
The temporal lobe epilepsy has been found to associate with significant reduction in α and β tubulin (Yang et al., 2006). More specifically, tubulin expression at PSD is down-regulated (Conference Abstract, 2009). The group of neurons with reduced tubulin level may start the oscillation sooner than the normal group.
Author: Frank Lee