Sodium Channel Blockers

1. Tetrodotoxin and Saxitoxin

From the molecular cascade depicted in Fig. 1, block of Na+ channels with tetrodotoxin (TTX) or saxitoxin (STX) would be predicted to confer protection of axons in various models of axon degeneration. Experimental studies using the in vitro anoxic optic nerve injury and spinal cord injury models support this prediction (Stys et al., 1992a; Agrawal and Fehlings, 1996; Imaizumi et al., 1997; Teng and Wrathall, 1997). In the optic nerve anoxic-injury model, for example, 1|M TTX provides greater than 80% of preanoxic recovery of CAP after 60 minutes of anoxic insult; 1|M of STX also preserves CAP function after anoxic injury, although not to as great an extent (58%) (Stys et al., 1992a).

TTX also prevents against the morphological axonopathy caused by exposure of isolated optic nerve to NO (Garthwaite et al., 2002).

Further support for neuroprotection of white matter axons by TTX has been reported in an in vivo spinal cord injury model. TTX has a protective effect on spinal cord axons after contusive dorsal column injury (Teng and Wrathall, 1997; Rosenberg et al., 1999), suggesting that sodium channels may be generally involved in a final common injury cascade that produces degeneration of spinal cord axons after a variety of insults. Phenytoin also has a neuroprotec-tive effect in experimental spinal cord contusive injury (Hains et al., 2004).

2. Tertiary Anesthetics

Tertiary anesthetics are known to reversibly block sodium channels (Narahashi et al., 1967; Strichartz, 1973; Cahalan, 1978). Administration of lidocaine or procaine (1|M) in vitro significantly enhances CAP recovery in the anoxic optic nerve (88% to 94%, compared to 34% recovery of nontreated, optic nerve [Stys et al., 1992c]). Lower concentrations (0.1 |M) of lidocaine and procaine that were neuroprotective in the anoxic optic nerve (with CAP recovery of 72% to 74%) did not markedly block normoxic conduction, suggesting that it is possible to use some sodium channel blocking agents to protect axons at concentrations that do not compromise axonal function. Procaine treatment has also been reported to provide some improvement of recovery of spinal cord dorsal column white matter CAP (83% vs. 72% for untreated control) after in vitro dorsal column compression injury (Agrawal and Fehlings, 1996).

Lidocaine also protects isolated ventral and dorsal roots from NO-induced injury. Using an in vivo spinal cord preparation to examine NO-induced axonal pathology, Kapoor et al. (2003) have shown that in spinal cord perfused in medium containing lidocaine, 83% of spinal roots exposed to NO regained "good" impulse conduction (defined as >50% CAP amplitude) compared to only 43% of spinal roots that regained conduction after NO exposure without treatment (Kapoor et al., 2003). Lidocaine treatment was also reported to prevent structural injury of axons in this study.

3. Quaternary Anesthetics

Some quaternary anesthetic analogs, such as QX-314 and QX-222, offer state-specific Na+ channel block, where the agents preferentially bind open Na+ channels (Wang et al., 1987; Khodorov, 1991). This characteristic is therapeutically advantageous because doses that achieve neuroprotection potentially may have little or no effect in terms of impeding normal action potential (Stys et al., 1992c). In the anoxic optic nerve model of white matter injury, QX-314 and QX-222 used at neuroprotective concentrations caused only minimal reduction in the CAP under normoxic conditions (5% of CAP for QX-314 and no reduction for QX-222). Nonetheless, the neuroprotective effects of these quaternary agents were substantial, with dramatic treatment recovery of the CAP after 60 minutes of anoxic injury (CAP recovery = 99.6% for QX-314 and 81.4% for QX-222) (Stys et al., 1992c).

Investigations of neuroprotection by QX-314 have been extended to an in vivo model of spinal cord compression injury (Agrawal and Fehlings, 1997). In these studies, after T1 spinal cord injury, animals were randomized to placebo or to receive 2 to -10 nM QX-314, delivered by microinjection to the injury site 15 minutes postinjury, as a one-time dose. Six weeks later, QX-314-treated animals demonstrated significantly increased functional integrity of spinal cord long tract axons, recognized through retrograde fluorogold labeling of supraspinal nuclei, with ~50% more fluorogold-positive neurons in the QX-314 animals compared to control animals. However, there was no clear difference in injury-site morphometrics or in final hind-limb behavioral testing between treated and untreated animals (Agrawal and Fehlings 1997).

4. Anticonvulsant Drugs

Phenytoin and carbamazepine are two clinically available and widely used anticonvulsant agents known to inhibit persistent neuronal sodium channels (Schwarz and Grigat, 1989; Ragsdale et al., 1996; Ragsdale and Avoli, 1998). These agents were demonstrated to be effective as neuroprotective agents in the optic nerve anoxia model of white matter injury (Fern et al., 1993). Both agents promoted significant recovery of postanoxic CAP in a dose-dependent manner, where maximal neuroprotection plateaued below the clinical reference range, and at concentrations that had no suppressing effect on conduction under normoxic conditions (Fig. 3). At these concentrations, phenytoin produced 69% recovery and carbamazepine produced 53% recovery of the CAP, compared to 34% recovery for untreated anoxic axons (Fern et al., 1993).

Lamotrigine, a newer anticonvulsant agent, is also known to block sodium channels (Kuo, 1998) and has been shown to be structurally neuroprotective of optic nerve axons subjected to oxygen-glucose deprivation injury in vitro (Garthwaite et al., 1999). Lamotrigine-treated optic nerves showed greater than 50% preservation of axons after oxygen-glucose deprivation in this study, compared to 100% axonopathy of oxygen-deprived axons. Lamotrigine at the concentration used for neuroprotection had only a small inhibitory effect on the normal optic nerve CAP. A clinically relevant note is that even the lowest neuroprotective concentration of lamotrigine used in this study (100 ||M) was still significantly in excess of levels seen in the cerebrospinal fluid of epileptic patients taking lamotrigine (Garthwaite etal., 1999).

Lo et al. (2002, 2003) carried out a series of studies to test whether oral administration of sodium channel-blocking anticonvulsant drugs in an animal model of MS would have neuroprotective properties similar to those seen within the isolated optic nerve model. Because phenytoin was more efficacious at neuroprotection than carbamazepine in the in vitro optic nerve anoxia experiments (Fern et al., 1993), phenytoin was tested as an neuroprotective agent in an in vivo animal model of MS, oligodendrocyte glycoprotein (MOG)-induced EAE.

The first study (Lo et al., 2002) extrapolated from the earlier in vitro demonstration (Fern et al., 1993) of a neuroprotection effect in the optic nerve, and examined optic nerve axon loss in EAE to assess the possible protective effects of phenytoin. In this study, mice with EAE were treated with oral phenytoin, beginning 10 days after immunological induction with MOG peptide, at doses that were titrated to achieve serum phenytoin levels within the human therapeutic range. At 28 days post-MOG injection, the number of axons in treated and untreated optic nerves

Figure 3 In vitro administration of phenytoin prevents irreversible optic nerve dysfunction. (A) One hour of anoxia substantially reduces the compound action potential (CAP) of an isolated rodent optic nerve, whereas treatment with 1 |M of phenytoin significantly restores postanoxic CAP recovery. (B) The presence of 1 |M phenytoin increases CAP recovery from 35% to 58% and a wash-out after the initial 60 minutes of reoxygenation has no additional effect. (C) Significant effects on CAP recovery occurs at concentrations below the clinically defined therapeutic range for phenytoin. (From Fern et al., 1993.)

Figure 3 In vitro administration of phenytoin prevents irreversible optic nerve dysfunction. (A) One hour of anoxia substantially reduces the compound action potential (CAP) of an isolated rodent optic nerve, whereas treatment with 1 |M of phenytoin significantly restores postanoxic CAP recovery. (B) The presence of 1 |M phenytoin increases CAP recovery from 35% to 58% and a wash-out after the initial 60 minutes of reoxygenation has no additional effect. (C) Significant effects on CAP recovery occurs at concentrations below the clinically defined therapeutic range for phenytoin. (From Fern et al., 1993.)

were compared after immunolabeling with antineurofila-ment antibodies. To identify all axons within these fields, nerve fibers were labeled using a combination of antibodies directed against both phosphorylated (SMI-31) and non-phosphorylated (SMI-32) neurofilaments (Lovas et al., 2000; Lo et al., 2002; Wujek et al., 2002). Axon counts of untreated EAE revealed 49% loss of optic nerve axons. EAE treated with oral phenytoin showed robust neuroprotection, with only 12% loss of optic nerve axons (Fig. 4). Clinical scores were significantly better, indicating less functional impairment, in phenytoin-treated mice.

Given the neuroprotective effects seen of phenytoin treatment in optic nerve, and in view of the substantial contribution of spinal cord atrophy to disability in MS, the effect of phenytoin was examined in a second study (Lo et al., 2003) on spinal cord white matter axons in EAE. In this study cer-

Figure 4 Oral administration of phenytoin in vivo to EAE mice results in significant axonal neuroprotection in the optic nerve. In EAE, there is a considerable loss of optic nerve axons as is quantified by neurofilament counts. Phenytoin administration in EAE results in a significant increase in optic nerve axon numbers. (Modified from Lo et al., 2002.)

Figure 4 Oral administration of phenytoin in vivo to EAE mice results in significant axonal neuroprotection in the optic nerve. In EAE, there is a considerable loss of optic nerve axons as is quantified by neurofilament counts. Phenytoin administration in EAE results in a significant increase in optic nerve axon numbers. (Modified from Lo et al., 2002.)

vical spinal cord axons were counted within standardized 500 ^m2 fields located at predetermined sites in the corticospinal tract, dorsal, lateral, and ventral columns. (This lesion-independent method of axon counting permitted the same white matter tract regions to be compared in different animals, and avoided selection bias due to lesion variability.) Mice with untreated EAE displayed a significant loss of axons within the corticospinal tracts and dorsal columns compared to control animals (non-EAE) (Fig. 5A,B). The density of axons labeled with neurofilament antibodies after 27 to 28 days of untreated EAE decreased substantially, with a 63% dropout of axons in the corticospinal tract and a 43% loss of axons in the dorsal columns (cuneate fasciculus). Treatment with phenytoin beginning at day 10 reduced the loss of corticospinal axons from 63% to 28% and reduced the loss of cuneate fasciculus axons from 43% to 17% (Lo et al., 2003).

Protection of axons by phenytoin was reflected functionally both in the electrophysiological recordings of the spinal cord CAP, which measures function in dorsal column axons, and in clinical behavioral testing (Figs. 6 and 7). The CAP in the untreated EAE group was significantly smaller or even unrecordable compared to the phenytoin-treated control group (Fig. 6B). In contrast, robust CAPs with a normal positive-negative configuration were observed in phenytoin-treated EAE, with a threshold similar to that in controls, and the average CAP amplitude in the phenytoin-treated EAE group was not significantly different than in the phenytoin-treated control group. (Fig. 6C). Thus in addition to protecting the structural integrity of spinal cord axons in EAE, phenytoin maintains their ability to conduct action potentials.

Clinical scores were assessed in this study using a standard EAE rating scale. Untreated EAE mice manifested progressive clinical impairment, with an average clinical score of 3.8 ±0.18 (representing complete hind-limb paralysis) at days 27 and 28. Phenytoin-treated EAE mice at day 27 and 28 exhibited a less severe clinical course, with an average clinical score of 1.5 ± 0.26, which represents a limp tail and minor righting reflex abnormalities (Fig. 7) (Lo et al., 2003).

More recently, the effects of carbamazepine in EAE have been assessed in EAE (Lo and Waxman, unpublished), and protective effects similar to those of phenytoin have been observed.

4. Sodium Channel Blocking Antiarrhythmic Drugs

Of the four established classes of antiarrhythmic drugs, class I agents function by acting on sodium channels, a mechanism of action that predicts this class of agent to be reason-

Figure 5 (A) Cervical spinal cord axons in the corticospinal tract (CST) and dorsal column are lost in untreated EAE. (B) Quantification of neurofilaments in the CST and dorsal columns of untreated EAE compared to phenytoin-treated EAE demonstrates reduction of both CST and dorsal column axon counts in untreated EAE (#P < 0.05) and a significant increase in axons in phenytoin-treated EAE (*P < 0.05). (Modified from Lo et al., 2003.)

Figure 5 (A) Cervical spinal cord axons in the corticospinal tract (CST) and dorsal column are lost in untreated EAE. (B) Quantification of neurofilaments in the CST and dorsal columns of untreated EAE compared to phenytoin-treated EAE demonstrates reduction of both CST and dorsal column axon counts in untreated EAE (#P < 0.05) and a significant increase in axons in phenytoin-treated EAE (*P < 0.05). (Modified from Lo et al., 2003.)

Figure 6 Neurophysiological spinal cord compound action potential (CAP) recordings of untreated and phenytoin-treated EAE. (A) Stimulating "S" and recording "R" electrodes are placed 8 mm apart at L4-5 and T11-T12, respectively. (B) The biphasic CAP wave seen in controls is highly attenuated in untreated EAE, and is restored in phenytoin-treated EAE (arrow indicates stimulus artifact). (C) Average CAP amplitudes (± SEM, mV) in phenytoin-treated controls, untreated EAE, and phenytoin-treated EAE at different stimulating current intensities (*P < 0.01, phenytoin-treated EAE compared to untreated EAE). (D) Average CAP area (± SEM, mV x ms) at different stimulus intensities in pheny-toin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05, phenytoin-treated EAE compared to untreated EAE). (E) Average supramaximal CAP area (± SEM, mV x ms) in phenytoin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05 phenytoin-treated EAE compared to untreated EAE). (F) Average conduction velocity (±SEM, m/s) in phenytoin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05, phenytoin-treated EAE compared to untreated EAE). (From Lo et al., 2003.)

Figure 6 Neurophysiological spinal cord compound action potential (CAP) recordings of untreated and phenytoin-treated EAE. (A) Stimulating "S" and recording "R" electrodes are placed 8 mm apart at L4-5 and T11-T12, respectively. (B) The biphasic CAP wave seen in controls is highly attenuated in untreated EAE, and is restored in phenytoin-treated EAE (arrow indicates stimulus artifact). (C) Average CAP amplitudes (± SEM, mV) in phenytoin-treated controls, untreated EAE, and phenytoin-treated EAE at different stimulating current intensities (*P < 0.01, phenytoin-treated EAE compared to untreated EAE). (D) Average CAP area (± SEM, mV x ms) at different stimulus intensities in pheny-toin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05, phenytoin-treated EAE compared to untreated EAE). (E) Average supramaximal CAP area (± SEM, mV x ms) in phenytoin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05 phenytoin-treated EAE compared to untreated EAE). (F) Average conduction velocity (±SEM, m/s) in phenytoin-treated control, untreated EAE, and phenytoin-treated EAE (*P < 0.05, phenytoin-treated EAE compared to untreated EAE). (From Lo et al., 2003.)

Figure 7 Phenytoin treatment ameliorates EAE clinical dysfunction. Clinical scores (mean ± SEM) are shown for untreated-EAE (filled circles) and phenytoin-treated EAE (filled boxes). Oral administration of phenytoin was started on day 10, as indicated by the horizontal bar. (From Lo et al., 2003.)

able candidates with neuroprotective properties in disorders of white matter. Various antiarrhythmic agents, some of which are used in clinical practice, have been tested in the rat optic nerve anoxia model (Stys, 1995) as well as in NO-mediated axonal degeneration models (Kapoor et al., 2003) and EAE (Bechtold et al., 2004). In the in vitro anoxia optic nerve model, the most effective agent tested (as assessed by postanoxic CAP restoration) was prajmaline, with 82% CAP recovery. Tocainide provided 79% recovery and ajmaline provided 79% recovery (untreated controls had 32% recovery in postanoxia electrophysiological testing). Disopyramide and bupivacaine were slightly less effective and resulted in CAP recovery in the range of 72% to 75%, and procainamide was ineffective. However, an important consideration for any of these agents for potential clinical use is their "efficacy index," which depends on the dose required for neuroprotection and any effect of that dose on normal conduction, which is undesirable. Prajmaline and tocainide at neuroprotective doses suppressed the preanoxic CAP by only 10%, whereas ajmaline (which provided almost the same magnitude of CAP protection as prajmaline) significantly depressed normal conduction by about 21%. Disopyramide, bupivacaine, and verapamil produced a 12% to 15% reduction in preanoxia CAP.

Stys and Lesiuk (1996) demonstrated that mexiletine provides neuroprotection of both in the in vitro anoxic optic nerve model and in an in situ ischemia-induced optic nerve injury model. Within the concentration range tested (10 to 1 mM) for mexiletine, the peak postanoxic recovery was seen at 100 (53% recovery of postanoxic CAP vs. 21% for untreated anoxic optic nerves). The magnitude of recovery above and below 100 was reduced. At this maximum neuroprotective concentration, the preanoxic CAP was suppressed slightly by 15%. In another set of in vivo experiments using ischemic conditions, the optic nerve electrophysiological responses to ischemia with and without mexiletine were examined in situ. The postischemic (60 minutes) CAP of untreated optic nerves in these experiments was dramatically reduced to 21% of control levels. Treatment with mexiletine (80 mg/kg) resulted in modestly improved recovery of the CAP, to 31% of preischemic control (Stys and Lesiuk, 1996).

Another Na+ channel-blocking antiarrhythmic agent, fle-cainide, was studied using a model of the NO-mediated injury of spinal roots (Kapoor et al., 2003; see also Chapter 18). Flecainide, at concentrations that allow normal impulse propagation, was perfused over spinal dorsal roots, resulting in a significant increase in the number of spinal roots which regained conduction following NO exposure. Whereas only 27% of spinal roots (3 of 11 roots assessed) recovered "good" conduction (defined as >50% CAP amplitude) after exposure to NO, 88% of flecainide-treated spinal roots (22 of 25 roots) showed good CAP recovery after NO exposure (Kapoor et al., 2003). Furthermore morphological analysis demonstrated that treatment with flecainide maintained the axonal structural integrity of axons, whereas exposure to NO alone caused dispersion of the axolemma and dissolution of organelles within the axoplasma, consistent with wallerian degeneration (Kapoor et al., 2003).

Bechtold et al. (2004) (see also Chapter 18) recently reported that flecainide (delivered subcutaneously) has a protective effect in EAE induced by inoculation of dark agouti rats with syngeneic spinal cord homogenate, reducing axon loss in the spinal cord from 48% to 3% to 17% depending on the dose schedule. As with phenytoin, treatment with flecainide protected axonal conduction and improved clinical outcome. Thus two sodium-channel blocking drugs, phenytoin and flecainide, have been demonstrated to protect against loss of spinal cord axons in EAE.

0 0

Post a comment