The hallmark of the secondary phase of traumatic brain injury (TBI) is a triad of excitotoxicity, free radical-induced lipid peroxidation (LP), and Ca2+ dysregulation (Bullock and Fujisawa 1992; Tymianski and Tator 1996; Hall et al. 2010). Glutamate release after injury causes an influx of Ca2+ into neuronal cells via activation of NMDA receptors (Arundine and Tymianski 2004). Activation of NMDA receptors has been shown to contribute to post-traumatic LP (Ozsuer et al. 2005) probably by causing an early increase of cytosolic Ca2+ which ignites the production of free radicals by several mechanisms including the Ca2+ induced activation of phospholipases and arachidonic acid cascade, conversion of xanthine dehydrogenase to xanthine oxidase, induction of nitric oxide synthases and mitochondrial leak (Hall and Springer 2004). The reactive species will attack cellular and mitochondrial membranes causing LP and protein oxidative damage (Beckman and Koppenol 1996; Violi et al. 1999; Singh et al. 2007). The inflicted oxidative damage causes further deterioration of Ca2+ homeostasis (Hall et al. 1998) probably by targeting mitochondria and stimulating the formation of the mitochondrial permeability transition pore (mPTP) (Castilho et al. 1995) which contributes to delayed Ca2+ dysregulation (Jacquard et al. 2006). These events collectively culminate in a buildup of cytosolic Ca2+ that will ultimately lead to neuronal degeneration though massive activation of cellular proteases like calpain (Kampfl et al. 1997).
Calpains are non-lysosomal Ca2+-dependent cysteine proteases that function at neutral pH. However, Under physiological conditions, calpains exist as inactive proenzymes in the cytosol (Wang and Yuen 1994; Kawasaki and Kawashima 1996). Once activated by increased cytosolic Ca2+ load after TBI, calpains degrade a large number of cellular proteins including cytoskeletal proteins such as α-spectrin (Roberts-Lewis and Siman 1993; Posmantur et al. 1996; Saatman et al. 1996a; Newcomb et al. 1997; Pike et al. 1998; Buki et al. 1999; Kupina et al. 2001, 2002, 2003; Deng et al. 2007) leading ultimately to post-traumatic neurodegeneration and neurological dysfunction (Saatman et al. 2000). α-Spectrin is an integral component of the cytoskeleton, especially in axons, dendrites and pre-synaptic terminals (Goodman et al. 1995). Calpain-mediated degradation of α-spectrin leads to the formation of breakdown products of two distinctive molecular weights; 150 kDa and 145 kDa which are considered footprints of calpain activation (Roberts-Lewis and Siman 1993; Bartus et al. 1995; Pike et al. 2001) and are a reliable predictors of outcome after TBI (Pike et al. 2001). It should be pointed out that the 150 kDa α-spectrin breakdown product is also generated by another cysteine protease caspase 3 (Wang 2000). However, our past work has indicated that calpain is the primary mediator of α-spectrin degradation in the mouse controlled cortical impact (CCI)-TBI model (Thompson et al. 2006; Deng et al. 2007). Similarly, studies by others in animal models of TBI (Liu et al. 1989; Pike et al. 1998, 2001; Aikman et al. 2006) and human TBI (Cardali and Maugeri 2006; Pineda et al. 2007; Brophy et al. 2009) have also shown that the contribution of calpain to post-TBI cytoskeletal degradation far exceeds that of caspase 3.
Studies done in a rat TBI model by (Buki et al. 1999) suggest that calpain-mediated α-spectrin degradation ultimately culminates in overt damage to the cytoskeleton, leading to irreversible damage of the axon and probably contributes to axotomy. Moreover, Inhibition of calpain–mediated proteolysis has proved to be a neuroprotective strategy since calpain inhibitors have been shown to salvage α-spectrin, attenuate axonal injury, and/or to improve motor and/or cognitive functions (Saatman et al. 1996b; Kampfl et al. 1997; Posmantur et al. 1997; Kupina et al. 2001; Ai et al. 2007). Recently published clinical studies have largely validated the use of α-spectrin degradation and immunoblot assessment of its proteolytic fragments in cerebrospinal fluid as biomarkers, the levels of which seem to correlate with TBI diagnosis, severity in terms of the Glasgow Coma Scale score and most importantly outcome (Mondello et al. 2010).
In the present report, we tested the hypothesis that LP plays a significant role in triggering post-traumatic calpain-mediated α-spectrin proteolysis, by determining whether pharmacologically scavenging lipid peroxyl radicals (LOO•) would attenuate calpain-mediated α-spectrin proteolysis after severe CCI-TBI. This was carried out with the use of U-83836E (Hall et al. 1991), which is a potent, selective and dual mechanism LOO• scavenger that we recently demonstrated to attenuate cortical mitochondrial oxidative damage and preserve mitochondrial functions following TBI including aerobic respiration and Ca2+-buffering capacity (Mustafa et al. 2010).
- Top of page
- Materials and methods
In our previous investigation, we demonstrated that U-83836E treatment can attenuate post-traumatic LP in cerebral cortical tissue or mitochondria together with a preservation of aerobic respiratory function and Ca2+-buffering capacity (Mustafa et al. 2010). Consistent with that overall effect, and the preservation of Ca2+-buffering capacity in particular, the current results have demonstrated that the LOO• scavenger U-83836E produces a dose-related attenuation of α-spectrin degradation after TBI. As discussed earlier, post-traumatic α-spectrin degradation is produced by both calpain and another cysteine protease caspase 3. The 150 kDa α-spectrin fragment is produced by both proteases whereas the 145 kDa fragment is calpain-specific (Wang 2000). However, since we detected very little of the caspase 3-specific 120 kDa fragment (Wang 2000) in these or other TBI experiments (Kupina et al. 2001, 2002, 2003; Thompson et al. 2006; Deng et al. 2007; Deng-Bryant et al. 2008), we believe that the leading source of post-traumatic α-spectrin degradation is calpain. From this, we conclude that even though U-83836E does not directly interact with calpain, it nevertheless attenuates the activation of this Ca2+-activated protease by preserving neuronal intracellular Ca2+ homeostatic mechanisms which are known to be compromised by post-traumatic LP (Hall et al. 2010) and thereby decreases post-traumatic Ca2+ overload and it’s activation of neuronal calpain. Calpain-mediated α-spectrin proteolysis peaks at 24 h after CCI-TBI in mice (Deng et al. 2007), and it has been repeatedly implicated in post-traumatic neurodegeneration (Posmantur et al. 1994; Kampfl et al. 1996; Posmantur et al. 1996; Kampfl et al. 1997; Buki and Povlishock 2006; Thompson et al. 2006; Deng et al. 2007; Saatman et al. 2010). Post-traumatic activation of calpains is mediated via the evolution of intracellular Ca2+ dysregulation after injury (Kampfl et al. 1997). The early phase of Ca2+ dysregulation is an acute manifestation of glutamate-mediated excitotoxicity which triggers an influx of Ca2+ via activation of NMDA receptors (Hayes et al. 1988; Faden et al. 1989; Randall and Thayer 1992). This initial increase in intracellular Ca2+ triggers a ‘burst’ of highly reactive free radicals (Hall et al. 1993; Lewen et al. 2000) contributed to by multiple mechanisms including membrane phospholipase and arachidonic acid cascade activation, increased biogenic amine release and turnover and mitochondrial ‘leak’ (Hall and Braughler 1993). The radical burst initiates LP oxidative damage which can increase exponentially as it propagates through cell membranes (Hall and Braughler 1993).
The ensuing LP damage exacerbates the loss of intracellular Ca2+ regulation by several mechanisms as illustrated in Fig. 4. First of all, it has been shown that the accumulation of LP products such as 4-hydroxynonenal (4-HNE) impairs glutamate transport mechanisms (Keller et al. 1997; Pedersen et al. 1999) which prolongs the extracellular duration of synaptically-released glutamate and thus the duration of its activation of NMDA receptors and inward Ca2+ influxes. Second, LP damages membrane phospholipid architecture (Hall et al. 2010) which can accentuate cell membrane ionic permeabilities, and most importantly that of Ca2+. This is evident by the occurrence of Ca2+ influxes in cultured cells following oxidant treatment which has been shown to be ameliorated by U-83836E treatment (Kimura et al. 1992; Munns and Leach 1995). Third, LP damage targets the Ca2+ ATPase within the cell membrane impairing its ability to pump Ca2+ either into the endoplasmic reticulum or through the plasma membrane (Rohn et al. 1993, 1996; Durmaz et al. 2003). Fourth, LP can potentially disrupt Ca2+ homeostasis by mobilizing Ca2+ from endoplasmic reticular stores (Racay et al. 1997). Fifth, LP in mitochondrial membranes aggravates mitochondrial dysfunction (Gadelha et al. 1997; Kowaltowski and Vercesi 1999; Sullivan et al. 1999; Singh et al. 2006b; Mbye et al. 2008) including a compromise of the ability of mitochondria to buffer cytosolic Ca2+ (Singh et al. 2006b; Mustafa et al. 2010). If oxidative damage to the mitochondrion is severe enough, it can trigger formation of the mPTP leading to mitochondrial permeability transition, collapse of the mitochondrial membrane potential and a dramatic release of mitochondrial matrix Ca2+ into the cytoplasm (Gadelha et al. 1997; Nicholls and Budd 2000). These events culminate in a dramatic increase in cytosolic Ca2+ causing massive activation of calpain. Such immense protease activation following TBI is revealed by the evolution of high levels of cytoskeletal SBDPs after TBI (Kampfl et al. 1997; Posmantur et al. 1997; Pike et al. 1998, 2001). Thus, it is our hypothesis that LP is a major contributor to the delayed Ca2+ dysregulation following TBI. Consequently, inhibiting LP by scavenging LOO• will protect cellular and mitochondrial membranes and hence maintain Ca2+ homeostasis sufficiently to prevent the delayed post-traumatic phase of Ca2+ dysregulation. Consistent with this theory, a single i.v. dose of U-83836E administered at 15 min post-injury was able to decrease α-spectrin degradation at its peak at 24 h after injury.
Figure 4. Mechanisms by which free radical-induced lipid peroxidation contributes to intracellular Ca2+ overload and calpain activation after TBI (see DISCUSSION for details).
Download figure to PowerPoint
The fact that delaying the single dose treatment beyond 15 min did not have a significant effect is attributed to the nature of the LP process. Lipid peroxidation can be initiated by a variety of free radicals generated by the reactive nitrogen species peroxynitrite or by iron-dependent mechanisms (Hall et al. 2010). However, once initiated, branching and propagation of LP reactions damage is dependent on LOO• which is generated within the process of LP (Spiteller 2006; Hall et al. 2010). As more time passes after injury before the start of antioxidant treatment, then more and more LOO• radicals are generated and the propagation of LP becomes amplified and therefore more difficult to stop. As a result, when we delayed the treatment from 15 min to 1 h, the single dose was not sufficient to stop the LP which was already well along. In contrast, when we employed the multiple dosing paradigm to increase the total amount of U-83836E administered and prolong its antioxidant action, it significantly attenuated the levels of SBDPs even though the treatment was delayed up to 12 h after injury. The ability of the drug to attenuate calpain-mediated α-spectrin proteolysis even when the treatment is delayed for 12 h is indicative that both branching and propagation reactions of LP and their contribution to Ca2+ dysregulation are still active at least out to 12 h following TBI. This is supported by our previously reported data which shows that in addition to the early increase in calpain-mediated α-spectrin proteolysis which begins during the first hour and plateaus at 6 h, there is a significant secondary increase in α-spectrin degradation between 12 and 24 h following CCI-TBI followed by a progressive decline thereafter (Deng et al. 2007).
As explained above, and illustrated in Fig. 4, there are several mechanisms by which LP contributes to post-traumatic intracellular Ca2+ dysregulation, and a cell-permeable LP inhibitor like U-83836E could potentially counteract each of these. Having said that, we demonstrated in our previous experiments that administration of a 3 mg/kg i.v. dose to mice at 15 min after CCI-TBI was able to significantly preserve the bioenergetics and Ca2+ buffering capacity of mitochondria isolated from the injured cortex (Mustafa et al. 2010). In the same study, the mitochondrial functional protection was accompanied by a significant reduction in the levels of the mitochondrial protein-bound LP product 4-HNE. In other work in our laboratory, it has been shown that direct application of 4-HNE to isolated cortical mitochondria potently impairs their bioenergetic function (Vaishnav et al. 2010). Thus, it seems certain that U-83836E is attenuating calpain-mediated cytoskeletal damage at least in part via mitochondrial functional protection. Further support for this notion comes from the fact that the peak of mitochondrial functional collapse after CCI-TBI does not occur until 12 h post-injury (Singh et al. 2006a) which coincides with our present observation of a U-83836E repeated dose therapeutic window of 12 h. In other words, the 12 h window is demonstrable because the absolute window for mitochondrial functional preservation is also 12 h. Secondary to the 12 h bottoming out of cortical mitochondrial function, we have documented a delayed secondary spike in α-spectrin degradation between 12 and 24 h post-injury in the CCI-TBI model (Deng et al. 2007). The coincidence of these events provides an explanation for how a 12 h therapeutic window in regards to stopping cytoskeletal proteolysis could be feasible. Nevertheless, the requirement for U-83836E to be effective when delayed until 12 h is that intensive (i.e. repeated) dosing be applied to stop the probably intense LP that is well developed by 12 h after injury.
Finally, it should be pointed out that this is not our first demonstration that a 12 h therapeutic window in regards to pharmacological inhibition of post-TBI calpain-mediated cytoskeletal damage can occur or that such a prolonged window is associated with a compound that acts to protect mitochondria. In recent work, we examined the cytoskeletal and neuroprotective effects of the novel compound NIM811 (Mbye et al. 2008, 2009). NIM811 is a non-immunosuppressive cyclosporine A analog, and like cyclosporine A, it appears to protect mitochondria by binding to cyclophilin D and preventing the formation of the mPTP (Waldmeier et al. 2002). Identical to our present therapeutic window data with U-83836E, we observed that NIM811 can attenuate post-TBI calpain-mediated α-spectrin degradation if administered as late as 12 h after injury (Mbye et al. 2009) and that this is paralleled by a significant reduction in mitochondrial LP-related oxidative damage (Mbye et al. 2008) and ultimately a decrease in cortical neurodegeneration and motor dysfunction (Mbye et al. 2009). This further emphasizes the contribution of mitochondrial dysfunction to overall post-traumatic Ca2+ dysregulation as well as the conclusion that it is amenable to pharmacological intervention as late as 12 h post-TBI.