Acute Axonal Injury Depolarization and Ionic Deregulation

Central mammalian axons are critically dependent on a continuous supply of oxygen and glucose for survival. Interruption of the supply of these substrates rapidly leads to failure of excitability (although some central tracts such as mouse optic nerve can maintain excitability for longer periods in vitro [Tekkok et al., 2003]), depolarization and ionic deregulation (Utzschneider et al., 1991; Ransom et al., 1992; LoPachin and Stys, 1995; Leppanen and Stys, 1997; Fern et al., 1998; Stys, 1998; Stys et al., 1998; Brown et al., 2001a). Figure 1 shows an example of the acute effects of anoxia on rat optic nerve in vitro. Within minutes, rat optic nerves lose the capacity to conduct action potentials (Fig. 1A) in parallel with a rapid and robust depolarization (Fig. 1B). Glycolysis is able to sustain a small residual resting membrane potential during anoxia as shown by the modest additional depolarization after application of ouabain, an Na pump inhibitor. This experiment shows that Na-K-ATPase activity was only minimally supported during anoxia. This does not imply that glycolysis plays only a minor role in the overall energy supply of central axons. During anoxic conditions, mitochondria are known to consume ATP by reverse operation of the ATP-synthase (Nicholls and Budd, 2000). Therefore it is likely that the inability of glycolysis to support the required energy demands is also due to paradoxical consumption of ATP by dysfunctional mitochondria. Aglycemia (without anoxia) shows a longer delay before depolarization and action potential failure become apparent (Stys, 1998; Stys et al., 1998; Brown et al., 2001a), likely caused by additional supply by glycogen stores from supporting astrocytes, with shuttling of lactate to axons by specific monocarboxylate transporters (Wender et al., 2000; Brown et al., 2003).

A more detailed study of the ionic deregulation that occurs in the acute phases of anoxia/ischemia showed that extracellular [K] rises to =15 mM, along with an acidification of the extracellular space by 0.3 pH units (Ransom et al., 1992). The rise in [K]o closely parallels the rapid anoxic depolarization and loss of excitability. Studies using electron probe microanalysis have revealed that it is largely axons that source the observed increase in [K]o, with axo-plasmic [K] falling to 10% of normal, together with a parallel increase in [Na] from approximately 20 to approximately

Figure 1 Representative traces of compound action potential (A) and resting membrane potential (B) from rat optic nerve recorded in vitro and exposed to anoxia (beginning at time 0). In the rat, optic nerve action potential conduction is abolished within minutes of anoxia onset, paralleling the rapid phase of depolarization. Blocking the Na-K-ATPase with ouabain (B) after the resting potential has plateaued in anoxia reveals that a small component is supported by glycolysis. (C) Time course of changes in axoplasmic [Na], [K], and [Ca] in anoxic optic nerve axons measured by electron probe microanalysis (LoPachin and Stys, 1995). [Na] and [K] are shown as calculated free concentrations in mM, whereas Ca, which exists largely in the bound state in cells, is shown as mmol/kg dry weight. Axoplasmic Na increases from a resting level of approximately 20 mM, with a parallel severe loss of axoplasmic K. Total axonal Ca content increases gradually during the hour of anoxic exposure to about five times normal. (A, B: Reproduced from Stys, P. K. [1998]. J. Cereb. Blood Flow Metab., 18:2-25, with permission from Lippincott, Williams and Wilkins. C: Reproduced from Stys, P. K., and Waxman, S. G. [2003]. Ischemic white matter damage. In: "Myelin Biology and Disorders," R. Lazzarini, J. Griffin, H. Lassmann, K.-A. Nave, R. J. Miller, and B. D. Trapp, eds. Copyright 2003, with permission from Elsevier.)

Figure 1 Representative traces of compound action potential (A) and resting membrane potential (B) from rat optic nerve recorded in vitro and exposed to anoxia (beginning at time 0). In the rat, optic nerve action potential conduction is abolished within minutes of anoxia onset, paralleling the rapid phase of depolarization. Blocking the Na-K-ATPase with ouabain (B) after the resting potential has plateaued in anoxia reveals that a small component is supported by glycolysis. (C) Time course of changes in axoplasmic [Na], [K], and [Ca] in anoxic optic nerve axons measured by electron probe microanalysis (LoPachin and Stys, 1995). [Na] and [K] are shown as calculated free concentrations in mM, whereas Ca, which exists largely in the bound state in cells, is shown as mmol/kg dry weight. Axoplasmic Na increases from a resting level of approximately 20 mM, with a parallel severe loss of axoplasmic K. Total axonal Ca content increases gradually during the hour of anoxic exposure to about five times normal. (A, B: Reproduced from Stys, P. K. [1998]. J. Cereb. Blood Flow Metab., 18:2-25, with permission from Lippincott, Williams and Wilkins. C: Reproduced from Stys, P. K., and Waxman, S. G. [2003]. Ischemic white matter damage. In: "Myelin Biology and Disorders," R. Lazzarini, J. Griffin, H. Lassmann, K.-A. Nave, R. J. Miller, and B. D. Trapp, eds. Copyright 2003, with permission from Elsevier.)

100 mM (Fig. 1C). A gradual but substantial increase in total axonal [Ca] was also observed during the 60-minute anoxic exposure, which worsens in large axons after reoxy-genation (LoPachin and Stys, 1995; Stys and LoPachin, 1996). As expected, the observed rise in total axonal Ca is paralleled by a reduction in extracellular [Ca] as measured by ion-sensitive microelectrodes (Brown et al., 1998), reflecting the translocation of this ion across the axolemma. Free [Ca] (that sensed by Ca-sensitive dyes) also increases substantially during anoxia/ischemia in central axons, with estimates exceeding 10 mM after 20 to 30 minutes of in vitro ischemia (Ren et al., 2000; Nikolaeva and Stys, 2003). Of interest, although as expected total axoplasmic [Ca] (as measured by electron probe microanalysis) does not increase during anoxia in Ca-free perfusate, both optic nerve and dorsal column axons exhibit substantial free [Ca] increases, even in Ca-free bath. In dorsal column axons, this internally sourced Ca rise is sufficient to severely damage the axons (Ouardouz et al., 2003). Reoxygenation appears particularly harmful to mitochondria, which suffer a greater than 100-fold increase in total matrix Ca, occurring largely after oxygen is restored (LoPachin and Stys, 1995; Stys and LoPachin, 1996). This is not unexpected as these organelles accumulate Ca electrophoretically via a Ca uniporter, which relies on the negative matrix potential (Crompton, 1985; Nicholls, 1985). This potential is rapidly dissipated when electron transport fails during anoxia/ischemia, limiting the amount of Ca entry into the matrix. During reoxygenation, however, with a high axoplasmic free [Ca] (see previously), re-energized mitochondria paradoxically accumulate large quantities of this ion, which will in turn short-circuit the newly re-established flux of protons pumped by the respiratory chain complexes, and overload the matrix with Ca, causing further damage to these organelles. Similar Ca changes are seen in CNS axons injured by mechanical trauma (LoPachin et al., 1999).

Axonal Ca overload is not merely an epiphenomenon but appears central in the pathogenesis of structural and functional injury. Removal of Ca from the perfusate during in vitro anoxia or ischemia prevents net accumulation of Ca and reduces free [Ca] increases and is also highly neuroprotective under many conditions (Stys et al., 1990; Imaizumi et al., 1997; Tekkok and Goldberg, 2001). One notable exception is ischemic injury of spinal dorsal columns, where in addition to control of extracellular Ca influx, release from intracellular stores must also be reduced for functional improvement (Ouardouz et al., 2003). Regardless of the source, however, in central white matter an increased concentration of Ca appears to be the "final common pathway" of cell death, as proposed for other cell types by Schanne and colleagues 25 years ago (Schanne et al., 1979; Orrenius and Nicotera, 1996). For this reason, deciphering the various modes of cellular Ca overload, whether sourced from outside or inside the cell, is of great importance.

An initial series of in vitro experiments on mature rat optic nerve in the early 1990s showed that acute anoxic injury to this structure is almost entirely dependent on influx of extracellular Ca. Removal of this cation from the perfusate allowed virtually complete functional (Stys et al., 1990) and ultrastructural (Waxman et al., 1993) recovery. Follow-up studies uncovered a Na-dependence of the Ca entry. Removal of bath Ca or Na strongly protects optic nerves against in vitro anoxia (Fig. 2A). Similarly, preventing Na influx through voltage-gated Na channels by applying TTX is highly protective, indicating that Na loading (later directly shown to be axonal [Stys and LoPachin, 1998]), not merely depolarization, secondarily leads to

O2 + glucose

2K 3Na Na 3Na 1Ca

O2 + glucose

2K 3Na Na 3Na 1Ca

2a 3

axoplasm

K 2b

2a 3

axoplasm

Figure 2 (A) Bar graph summarizing the degree of compound action potential recovery recorded electrophysiologically after exposure of adult rat optic nerves to in vitro anoxia. In normal artificial CSF containing 2 mM Ca, CAP magnitude recovers to approximately 20% of control after 1 hour of anoxia. Removing Ca from the perfusate (with the addition of the Ca chela-tor EGTA) allows virtually complete recovery, as does removal of Na ions from the bath. Blocking voltage-gated Na channels with TTX is also highly protective, whereas increasing Na permeability with the Na channel inactiva-tion blocker veratridine worsens anoxic injury. The Na-Ca exchange inhibitor bepridil is also very protective. (B) Taken together, these results suggest a Ca-and Na-dependent mechanism of anoxic optic nerve injury summarized in the diagram. Energy failure (1) causes Na influx through non-inactivating Na channels (2a) and K loss through K channels (2b). The Na accumulation and depolarization together drive the Na-Ca exchanger to admit damaging amounts of axoplasmic Ca from the extracellular space. (A: data from Stys et al., 1992a; B: Modified from Stys et al., 1992a.)

axoplasmic Ca accumulation. Reverse operation of the NaCa exchanger and resultant Ca influx would be recruited by these steps, which was directly shown by the protective effects of Na-Ca exchange inhibitors (Fig. 2A) (Stys et al., 1992a; Imaizumi et al., 1997; Li et al., 2000; Tekkok et al., 2000; Brown et al., 2001a). This preliminary model of axonal Ca overload is illustrated in Fig. 2B. According to this model, energy depletion leads to Na pump failure, influx of Na through noninactivating Na channels (Stys et al., 1993; Taylor, 1993), and stimulation of the Na-efflux, Ca-import mode of the Na-Ca exchanger, shown to be present on central axons (Steffensen et al., 1997; Craner et al., 2004a) (see also Chapter 7). (More recently, a K-coupled Na-K-Ca exchanger has been found in the CNS

outside of its traditional territory in the retina [Dong et al., 2001; Kiedrowski et al., 2002]. [For a review, see Lytton, 2002]). Its contribution to Ca loading cannot be ruled out, if it were proven to be present in axons. Indeed, given the K dependence of this transporter and the known collapse of the K gradient and K accumulation in the extracellular space, the Na-K-Ca exchanger would be even more strongly driven than the K-independent Na-Ca exchanger to import Ca into cells).

The proximal role played by voltage-gated Na channels appears to extend to other modes of white matter injury, such as inflammatory demyelination. Two recent studies indicate that use-dependent Na channel blockers, such as phenytoin (Lo et al., 2003) and flecainide (Bechtold et al., 2004), significantly improve the survival of CNS axons, their ability to conduct impulses, and clinical scores in rats with EAE, an animal model of MS (see Chapter 29). Taken together, it is becoming apparent that the fundamental steps illustrated in Fig. 2B may represent a ubiquitous cascade of events unleashed in injured white matter, regardless of initial insult. This raises important therapeutic opportunities for a variety of white matter disorders.

0 0

Post a comment