|Wireless Communication in the Brain
III. Evidence from Traumatic Brain Injury
Traumatic brain injury (TBI) refers to the brain injury caused by an external mechanical force such as head striking or blast exposure (McAllister, 2011; Hill et al., 2016). The mild TBI (mTBI), also known as concussion, accounts for 75% of all head injuries (Girgis et al., 2016). It often leads to post-traumatic epilepsy (Kovacs et al., 2014) and increases the risk for Parkinson's disease (PD) and Alzheimer's disease (AD) (Hutson et al., 2011; Crane et al., 2016; Li et al., 2017). The mechanisms of TBI and subsequent vulnerability to neurodegeneration are not well understood.
Like neurodegeneration, Ca2+ ions also play a critical role in TBI. It appears that the external mechanical force may excite neurons, resulting in Ca2+ overload (McAllister, 2011). Paper 12 has shown that the Microtubule Model for Excitability can explain neuronal excitation by both electromagnetic and mechanical forces. The ultrasound discussed in the previous paper resembles sound blast created by explosives. Therefore, TBI could result from the same mechanism. This paper will provide evidence to support this hypothesis.
The Mechanism of Ca2+ Overload
It has been well documented that TBI is caused by elevated cytosolic Ca2+ level which triggers a cascade of pathological events like calpain activation, oxidative stress, caspase activation and microtubule destruction (McAllister, 2011). The mechanism that links the mechanical force with Ca2+ overload is not known. It was proposed that the mechanical force might produce pores in the cell membrane for Ca2+ influx, a process termed "mechanoporation". However, a study found that the trauma-induced Ca2+ influx could be modulated by tetrodotoxin-sensitive sodium channels and attenuated by the voltage-gated calcium channel blocker. There was no evidence of calcium entry through mechanically produced pores (Wolf et al., 2001).
The axon initial segment (AIS) is enriched with voltage-gated sodium and potassium channels for the generation of action potentials (Figure 1). It contains sodium channels Nav 1.1, Nav1.2 and Nav1.6 which are all sensitive to tetrodotoxin. Several types of voltage-gated calcium channel are also present in AIS (Bender et al., 2012; Yu et al., 2010). According to the Microtubule Model for Excitability, these ion channels could be regulated by microtubules (Paper 2). More specifically, their voltage sensors could be influenced by the negative electric field from the microtubules near the AIS membrane. As illustrated in Paper 12, the mechanically induced cell deformation may cause microtubule bending (buckling), resulting in partial detachment from the AIS membrane, and consequently higher excitability. The mechanically generated action potential further activates the voltage-gated calcium channels not only in AIS, but also in other neuronal compartments.
The Ca2+ entry through voltage-gated calcium channels may trigger the release of Ca2+ ions from the intracellular store, endoplasmic reticulum (ER) (Staal et al., 2010), which is distributed throughout the cell including dendrites and axon (Wu et al., 2017). In resting cells, the cytosolic Ca2+ concentration is maintained around 100 nM. Up to ~ 500 nM, Ca2+ works synergistically with IP3 to activate IP3 receptors (IP3-sensitive calcium channels), releasing more Ca2+ from ER. At higher concentrations, cytosolic Ca2+ inhibits IP3 receptor opening (Hanson et al., 2004).
AIS Shortening and the Initial Decrease in Excitability
In mTBI, the axonal injury occurs primarily within AIS (Greer et al., 2013). It has also been shown that AIS is shortened when measured 2 days (Vascak et al., 2017) or 2 weeks post-trauma (Baalman et al., 2013). These findings suggest that mTBI is caused mainly by Ca2+ influx via AIS, rather than through NMDA receptors at the synapse. The excitability was found to decrease within 1 - 4 hr after trauma, and then gradually recover in a couple of days (Ping and Jin, 2016).
The AIS shortening and attenuated excitability can also be induced by prolonged (3 hr) elevated neural activity (Evans et al., 2015). Their coincidence may lead to the notion that the attenuated excitability after trauma is due to the shortened AIS per se. A computer simulation does find that the neuronal excitability increases with longer AIS (Gulledge and Bravo, 2016). However, this simulation assumes corresponding changes in total sodium conductance, which is not true for mTBI.
The AIS length is typically measured on the basis of Ankyrin-G distribution. Ankyrin-G is a scaffold protein, responsible for anchoring various proteins to AIS. It is thought that the localization of Nav1.6 to AIS also requires Ankyrin-G. Surprisingly, the mTBI-induced AIS shortening, as indicated by Ankyrin-G distribution, was not accompanied by the shortening of Nav1.6 distribution (Vascak et al., 2017). Furthermore, the excitability has started to recover 4 hr post-trauma, but the AIS length remains shorter for at least two days. Therefore, the post-traumatic excitability is unlikely governed by the AIS length alone.
While the shortened AIS per se cannot account for the decrease in post-traumatic excitability, its accompanied microtubule depolymerization could lead to significant decrease in excitability. The formation of AIS is controlled by microtubules as well as casein kinase 2 (CK2) and cyclin-dependent kinase 5 (Cdk5) for the phosphorylation of sodium and potassium channels, respectively (Sanchez-Ponceet al., 2011; Nelson and Jenkins, 2017). Conversely, the AIS shortening may result from microtubule depolymerization and channel dephosphorylation. This notion is supported by two findings. (1) The AIS length decreased from the distal end without changing the start position (Vascak et al., 2017), consistent with microtubule depolymerization from the plus end. (2) Calcineurin activation, which catalyzes dephosphorylation, is required for activity-induced AIS shortening (Evans et al., 2015). Moreover, high Ca2+ concentration can directly cause microtubule depolymerization (O'Brien et al., 1997) and activate calcineurin. Therefore, Ca2+ overload alone is sufficient to cause AIS shortening. Then, how can this process result in hypoexcitability?
Here enters tubulin.
Under the AIS membrane, there is an actin/spectrin layer (Figure 1) preventing microtubules from contacting the membrane surface, for otherwise the negative electric field from microtubules would inhibit neuronal firing. Tubulin, being a small molecule, has the capacity to penetrate the actin/spectrin layer and reach the membrane surface to inhibit neuronal firing with its negative electric field. Microtubule depolymerization produces free tubulin. Hence, microtubule depolymerization can reduce excitability. This was first discovered from the studies of squid giant axons more than three decades ago (Sakai et al., 1985). A recent study also found that the microtubule-depolymerizing agent, nocodazole, reduced burst activity and seizure severity in the rat hippocampus (Carletti et al., 2016). Furthermore, tubulin is implicated in TBI (Wilson et al., 2012; López-García et al., 2016; Tzekov et al., 2016; Lipponen et al., 2016). The next several papers will present evidence that the free tubulin produced by Ca2+-induced microtubule depolymerization plays key roles in seizure termination, memory consolidation and memory retrieval.
Post-Traumatic Epilepsy and the Risk for Neurodegeneration
Except for the initial brief period (~ 4 hr), TBI generally increases excitability which may last for weeks or even years. The long-lasting hyperexcitability may lead to post-traumatic epilepsy (Yang et al., 2010; Webster et al., 2017) and increase the risk for neurodegeneration. Two possible mechanisms are proposed below: APP upregulation and BDNF reduction.
Amyloid precursor protein (APP) is a marker of axonal injury resulting from traumatic or other causes (Hayashi et al., 2015). Its immunohistochemical stain has been used in forensic practice (Reichard et al., 2005). Shortly after trauma, APP is upregulated (Plummer et al., 2016), possibly for repairing axonal injury, as APP plays a critical role in the formation of nodes of Ranvier (Xu et al., 2014). The APP upregulation has been shown to exert beneficial effects (Plummer et al., 2016; Hefter and Draguhn, 2017). However, it is prone to develop epilepsy and neurodegeneration because APP processing may generate its intracellular domain, AICD, which is pro-epileptic (Paper 7). The increased AICD activity has been demonstrated to associate with Parkinson's disease (Chen et al., 2017) and Alzheimer's disease (Ghosal et al., 2009). The Down syndrome is a genetic disorder with three copies of chromosome 21 where the APP gene is located. Down patients are susceptible to seizures and AD (Paper 9).
According to BDNF Cascade Hypothesis (Paper 4), BDNF deficiency may lead to hyperexcitability and neurodegeneration. Inflammation is implicated in epileptogenesis after TBI (Webster et al., 2017; Semple et al., 2017). Mechanistically, this could be due to the reduction of BDNF caused by inflammation (Calabrese et al., 2014), which in turn may result from Ca2+-induced oxidative stress (Biswas, 2016). This mechanism is supported by a recent finding that 30 days post-trauma, neuroinflammation increases while BDNF decreases (Impellizzeri et al., 2016).
Figure 3 summarizes the TBI-associated pathological cascade based on the Microtubule Model for Excitability, BDNF Cascade Hypothesis for neurodegeneration, and currently available experimental evidence.
Author: Frank Lee