|The Roles of Microtubules
and Tau Proteins in Neuronal Excitability
Hyperexcitability is an early sign of Alzheimer's disease (Dickerson et al., 2005; Putcha et al., 2011), Parkinson's disease (Blandini et al., 1996), amyotrophic lateral sclerosis (van Zundert et al., 2012) and Huntington's Disease (Klapstein et al., 2001). The microtubule-associated protein Tau is implicated in all of these neurodegenerative disorders (Braak et al., 1993; Wills et al., 2010; Mimuro et al., 2007; Fernández-Nogales et al., 2014). In Paper 1, it was described briefly how the Tau protein may modulate neuronal excitability by interacting with the microtubules at the axon initial segment (AIS). Further details are discussed in this paper.
The Tau protein has six isoforms produced from a single gene through alternative RNA splicing (Figure 1). They differ in the number of inserts at the N-terminal half and the number of repeats at the C-terminal half . The number of inserts may be 0, 1 or 2, depending on whether the exon 2 and/or 3 are included during RNA splicing. The number of repeats may be either 3 or 4. The 4-repeat (4R) Tau includes the second repeat encoded by exon 10. In a healthy adult brain, the levels of 4R and 3R Tau proteins are approximately equal. Distortion of the balance toward 4R Tau may lead to neurodegeneration, as observed in Huntington's Disease (Fernández-Nogales et al., 2014), Alzheimer's disease (Yasojima et al., 1999; Ginsberg et al., 2006; Glatz et al., 2006), Parkinson's disease (Caffrey et al., 2006), and frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17) (Ghetti et al., 2015). Their pathogenic mechanisms will be discussed in later papers. This article focuses on the role of Tau proteins in excitability. It will be shown that the 4R Tau increases excitability.
The Microtubule Model for Excitability (MTME)
MTME was originally proposed for wireless communication in the brain, where microtubules at the AIS may serve as the receiving antennas for converting electromagnetic (EM) waves into neuronal excitability. A microtubule is made up of tubulin which has two isoforms, α and β, that usually form a heterodimer. In a tubulin heterodimer, the number of negatively charged amino acids exceeds that of positively charged amino acids by about 50 (Minoura and Muto, 2006). As a result, a microtubule is highly negatively charged along the entire length (Baker et al., 2001). Its association with the AIS membrane should have the same effects as hyperpolarization, i.e., inhibitory. Dissociation from the AIS membrane is equivalent to depolarization.
Ankyrin-G, together with the microtubule end binding protein EB1 or EB3 (denoted by EB1/3), play a crucial role in anchoring microtubules to the AIS membrane (Leterrier et al., 2011). Higher Ankyrin-G level will provide more "anchor points", which restrict the microtubule to bend away from the membrane, thereby reducing excitability (Figure 2). During long range synchronization, the EM waves can induce microtubule vibration to disrupt the anchoring, resulting in microtubule bending away from the membrane. If all anchor points are disrupted, the entire microtubule will detach from the membrane.
This model is supported by two studies that have demonstrated the effects of Ankyrin-G on excitability. The first study employed TsA201 cells to express Nav1.6 channels and/or Ankyrin-G. If only Nav1.6 channels are expressed, a significant persistent sodium current INaP (conducted by Nav1.6 channels) was observed. However, co-expression with Ankyrin-G reduced INaP (Shirahata et al., 2006). Nav1.6 is abundant at the AIS, with a critical role for the regulation of excitability (O'Brien and Meisler, 2013). The second study investigated the effects of amyloid precursor protein (APP) over-expression in a transgenic mouse model. It was found that the APP over-expression up-regulates a microRNA, miR-342-5p, which in turn down-regulates the expression of Ankyrin-G (Sun et al., 2014). In agreement with MTME, the APP transgenic mice exhibited hyperexcitability (Wesson et al., 2011; Bezzina et al., 2015).
Modulation of Excitability by Tau Proteins
Tau is a microtubule-associated protein. It can also interact with EB1/3 (Sayas et al., 2015). Thus, Tau may increase excitability by interfering with the association between microtubules and the membrane. This explains why in animal models Tau reduction or knockout attenuates hyperexcitability (Holth et al., 2013; DeVos et al., 2013; Li et al., 2014).
Hyperphosphorylated Tau has been shown to reduce excitability (Hatch et al., 2017). Phosphorylation is a process that adds a negatively charged phosphate group to a protein. As a result, hyperphosphorylation of the Tau protein should reduce its binding with the negatively charged microtubules. In the absence of the Tau interference, MTME predicts that the microtubule should be more likely to associate with the membrane, thereby attenuating excitability.
Tau hyperphosphorylation is a major characteristic of Alzheimer disease (AD). During the development of the disease, the hippocampus and entorhinal cortex first exhibited hyperexcitability, but then followed by hypoactivation as the disease progresses (Dickerson et al., 2005). The later hypoactive phase is likely due to Tau hyperphosphorylation. The early hyperactive phase may arise from elevated 4R Tau.
In the Tau protein, the repeat region is the microtubule binding domain. Thus, the 4R Tau should bind to the microtubule more tightly (Bunker et al., 2004), and prevent microtubule association with the membrane more effectively than the 3R Tau. Therefore, the 4R Tau should have greater impact on enhancing excitability than the 3R Tau. Elevated 4R Tau may lead to hyperexcitability, which could be the origin of neurodegeneration.
The Effect of Microtubule Depolymerization
The idea that microtubules might play a role in neuronal excitability has been proposed several decades ago, based on the observation that the excitability of squid giant axons correlates with microtubule assembly (Matsumoto and Sakai, 1979). It was further found that the internal perfusions which cause microtubules to depolymerize reduce the excitability and those supporting microtubule assembly increase the peak sodium current (Sakai et al., 1985). Recently, similar results were obtained from epilepsy models in the rat: the microtubule-depolymerizing agent nocodazole reduced burst activity and seizure severity (Carletti et al., 2016). Nocodazole is likely to act on the microtubules at AIS, as the chemical agent has been shown to modulate AIS location in a microtubule-dependent manner (Hatch et al., 2017).
How can microtubule depolymerization reduce excitability? It is important to note that, beneath the plasma membrane of a neuron, there is a submembrane cytoskeleton composed of actin and spectrin (Figure 4). The actin/spectrin layer exists not only in axons (Xu et al., 2013), but also in the somatodendritic compartments of various neuronal types, across different animal species (He et al., 2016; Han et al., 2017). Although both actin and spectrin are also highly negatively charged (Elzinga et al., 1973; Speicher et al., 1983), their distance with the membrane surface is fixed. These charges do not contribute to the dynamic change of the membrane potential field that governs the opening of voltage-gated ion channels. By contrast, the AIS contains sparse microtubules (Palay et al., 1968), allowing a microtubule to bend or translocate significantly within the AIS. As calculated in Paper 1, translocation of a microtubule within the AIS is sufficient to influence channel gating.
Due to the presence of the submembrane cytoskeleton, a microtubule cannot contact the intracellular membrane surface. The free tubulin dimer is a small molecular complex capable of penetrating the actin/spectrin layer to inhibit action potentials with its negative electric field. However, the actin/spectrin layer is also negatively charged, which would repel tubulin. Therefore, the entry of tubulin into the submembrane cytoskeleton may require multivalent counterions (Ha and Liu, 1999), such as Ca2+ ions. High concentration of Ca2+ ions is known to cause microtubule depolymerization to produce free tubulin dimers (O'Brien et al., 1997). During seizures (Paper 14) or traumatic brain injury (Paper 13), the intensive neural activities often lead to Ca2+ overload. Not only can the elevated Ca2+ ions induce microtubule depolymerization to produce free tubulin dimers, they also serve as the counterions mediating the attraction between the actin/spectrin layer and tubulin. This explains how microtubule depolymerization can attenuate neuronal excitability (Figure 5). We see that, through tubulin inhibition, the microtubule depolymerization resulting from Ca2+ overload provides a negative feedback to mitigate Ca2+ toxicity.
For more than three decades, the evidence (Matsumoto and Sakai, 1979) that microtubules might play a role in neuronal excitability has been largely ignored. This situation was changed a few years ago, when several groups provided direct evidence for the involvement of microtubule-associated protein Tau in excitability (Holth et al., 2013; DeVos et al., 2013; Li et al., 2014). Since then, further evidence continues to accumulate. The microtubule-depolymerizing agent nocodazole has been demonstrated to reduce burst activity and seizure severity (Carletti et al., 2016). Very recently, Hatch et al. (2017) showed that the hyperphosphorylated Tau could reduce excitability by modulating AIS in a microtubule-dependent manner. This finding supports MTME which centers on the microtubules at AIS. However, Hatch et al. attributed the reduced excitability to the relocation of AIS away from the soma, whereas MTME suggests that the reduced excitability is due to the modulation of Tau on microtubule dynamics.
The shift of AIS away from the soma does not necessarily attenuate excitability. There are two competing factors. On one hand, a more distal AIS will increase the voltage attenuation from synapses to AIS which is the initiation site of action potentials. As a result, the possibility for AIS potential to reach the threshold is decreased. On the other hand, the large somatodendritic membrane area acts as a current sink for sodium current generated in the AIS. A more distal AIS would reduce the current flow, thereby increasing the local depolarization at the AIS (Hamada et al., 2016). For pyramidal neurons (used in the study of Hatch et al.), computer modeling indicates that a more distally located AIS should enhance, rather than reduce, excitability. However, due to the two opposite factors, the change in excitability by AIS relocation is only modest (Gulledge and Bravo, 2016).
The presence of the submembrane cytoskeleton allows the free tubulin dimers to exert a stronger inhibitory effect than microtubules. This may account for the reduced excitability by microtubule depolymerization. Paper 15 and subsequent several papers will present evidence that the free tubulin dimers also play crucial roles in transmitter release at axon terminals, and in memory retrieval at dendritic spines.
Author: Frank Lee