Demyelinated Axons

Although demyelinated axons can sometimes conduct impulses successfully, they never conduct as well as normal axons. Conduction remains much slower and less secure than normal, and it is also prone to conduction failure, especially if conducting trains of impulses at higher frequencies. These conduction deficits result in a range of functional deficits in patients.

I. Conduction Slowing

Techniques that allow the progress of the action potential to be monitored as it conducts along the demyelinated axolemma (Bostock and Sears, 1978) have clearly revealed that while conduction remains fast in the normally myeli-nated portions of the axolemma, it is markedly slowed along the demyelinated region (McDonald and Sears, 1970). In fact, conduction is slowed to approximately 1 m/sec (0.5 to 2.5 m/sec), and so each 1 mm of demyelination contributes approximately 1 mssec of delay. Except in particular pathways, this slowing usually has little consequence for patients (Halliday et al., 1973), but it has proven to be particularly valuable clinically, assisting in the diagnosis of demyelinat-ing disease because of delays in the latency of the sensory or motor compound action potentials in patients with peripheral neuropathy, or in delays in the latency of the visual (Halliday et al., 1973), somatosensory, or auditory-evoked potentials in patients with MS.

2. Conduction of Pairs of Impulses

The slow conduction along demyelinated axolemma inevitably limits the ability of the axons to transmit pairs (and trains) of impulses because the second action potential of a pair, traveling close behind the first, will run into the relative and absolute refractory periods of the first impulse as this is slowed at the site of demyelination. McDonald and Sears (1970) coined the phrase the refractory period of transmission (RPT) to describe this deficit, defining the RPT as the maximum interval between two conducted impulses such that the second impulse fails to be conducted successfully through the lesion. In a subsequent study of central axons proven to be segmentally demyelinated, the refractory period of the normal portion of the axon was always in the range 0.5 to 1.4 m/sec, but this was prolonged to 1.0 to 6.0 m/sec if the site of demyelination was included in the conduction pathway (Fig. 4). In one axon the RPT was prolonged to 27 m/sec when the lesion was included in the pathway (Felts et al., 1997).

3. Conduction of Trains of Impulses

The problems associated with conducting pairs of impulses are magnified when considering the conduction of trains of impulses. One consideration is that the RPT of the second and later impulses is longer than that of the first (McDonald and Sears, 1970), slightly increasing with increasing numbers of impulses until a plateau is reached. This effect is due, in part, to the fact that the second impulse conducts across the demyelinated region in the relative refractory period of the first, and so conducts a little more slowly than it, and so on. The maximum frequency of transmission therefore slowly declines over time. In a classic study (Fig. 5) (McDonald and Sears, 1970) conduction through the lesion was restricted to only 410 Hz, although the normal portion of the same axon was conducted at 1,000 Hz. Even the low frequency was maintained for only three impulses before alternate impulses failed to be transmitted.

Another problem arising with repeated activation is the appearance of intermittent periods of complete conduction

Figure 4 Recordings from a central demyelinated axon either excluding (A) or including (B) the lesioned portion in the conduction pathway. The normal portion of the axon can conduct two closely spaced impulses resulting from electrical stimuli presented 0.77 ms apart (A, third record), but not less than 0.77 ms apart (A, second record). However, the demyelinated portion of the same axon was only able to conduct impulses spaced by more than 1.32 ms (B). (Reproduced from Felts et al., 1995.)

rT"h

Figure 4 Recordings from a central demyelinated axon either excluding (A) or including (B) the lesioned portion in the conduction pathway. The normal portion of the axon can conduct two closely spaced impulses resulting from electrical stimuli presented 0.77 ms apart (A, third record), but not less than 0.77 ms apart (A, second record). However, the demyelinated portion of the same axon was only able to conduct impulses spaced by more than 1.32 ms (B). (Reproduced from Felts et al., 1995.)

block (Fig. 6) (Felts et al., 1995); a high-frequency impulse train is abruptly "chopped" into periods where impulses are transmitted faithfully at high frequency, separated by periods of conduction failure. These periods are due to membrane hyperpolarization in response to potentiated activity rT"h

Figure 5 Single unit activity in a dorsal root filament teased from an intercostal nerve caudal to an experimental demyelinating lesion in the dorsal columns induced by the prior injection of diphtheria toxin (hatched region). The stimulus artifacts appear as dotted lines and the action potentials as solid lines. Whereas stimulation at S2 excludes the lesion from the conduction pathway and the axon can conduct faithfully at 1,000 Hz, stimulation at S1 includes the lesion and conduction sometimes fails even at 410 Hz. (Reproduced from McDonald and Sears, 1970.)

0.5 msec tfV 1000 Hz

0.5 msec

Figure 5 Single unit activity in a dorsal root filament teased from an intercostal nerve caudal to an experimental demyelinating lesion in the dorsal columns induced by the prior injection of diphtheria toxin (hatched region). The stimulus artifacts appear as dotted lines and the action potentials as solid lines. Whereas stimulation at S2 excludes the lesion from the conduction pathway and the axon can conduct faithfully at 1,000 Hz, stimulation at S1 includes the lesion and conduction sometimes fails even at 410 Hz. (Reproduced from McDonald and Sears, 1970.)

of the electrogenic Na+/K+ ATPase (sodium pump) caused by the raised intra-axonal sodium concentration resulting from the impulse activity (Bostock and Grafe, 1985; Vagg et al., 1998). The phenomenon can appear after just 1 second of stimulation at 500 Hz (McDonald and Sears, 1970) or within 10 to 30 seconds of stimulation at 100 to 200 Hz (R. Kapoor, P. A. Felts, and K. J. Smith, unpublished observations). These latter frequencies are within the physiological range, and so the phenomenon is likely to impair normal sensation and motor function, contributing to the reduced flicker fusion frequency in some patients (Titcombe and Willison, 1961), to the "fading" or blurring of vision upon prolonged fixation of gaze (McDonald, 1998; Waxman,

Figure 6 Intermittent complete conduction block exhibited by a central demyelinated axon in an intra-axonal record (resting potential -60mV) obtained at or near a site of demyelination induced by the injection of ethidium bromide into the dorsal column 14 days previously. The axon was initially able faithfully to follow a stimulus applied at 200 Hz (not shown), but after about 10 seconds of such stimulation, the axon entered intermittent periods of complete conduction block. The action potentials are the tall spikes of regular amplitude. The irregular spikes during the periods of conduction block are due to stimulus artefacts occasionally captured by the analog-to-digital converter. The axon conducted trains of 20 or so action potentials separated by periods of conduction block. The periods of conduction block coincide exactly with periods of membrane hyperpolarization, indicating that the block is mediated by activity of the electrogenic sodium/potassium ATPase. Calibration bar = 0.5 sec. (Reproduced from Felts et al., 1995.)

Figure 6 Intermittent complete conduction block exhibited by a central demyelinated axon in an intra-axonal record (resting potential -60mV) obtained at or near a site of demyelination induced by the injection of ethidium bromide into the dorsal column 14 days previously. The axon was initially able faithfully to follow a stimulus applied at 200 Hz (not shown), but after about 10 seconds of such stimulation, the axon entered intermittent periods of complete conduction block. The action potentials are the tall spikes of regular amplitude. The irregular spikes during the periods of conduction block are due to stimulus artefacts occasionally captured by the analog-to-digital converter. The axon conducted trains of 20 or so action potentials separated by periods of conduction block. The periods of conduction block coincide exactly with periods of membrane hyperpolarization, indicating that the block is mediated by activity of the electrogenic sodium/potassium ATPase. Calibration bar = 0.5 sec. (Reproduced from Felts et al., 1995.)

1981), and to the progressive weakness experienced by some patients upon walking only a short distance (McDonald, 1975). Drugs that inhibit the Na+/K+ ATPase have been reported to improve conduction in central demyelinated axons (Kaji and Sumner, 1989) and some patients with MS (Kaji et al., 1990). Intermittent conduction block has now been demonstrated in a range of human neuropathies (e.g., Cappelen-Smith et al., 2000).

Although preceding impulse activity typically impairs function in demyelinated axons, it can happen, at least in demyelinated Xenopus axons, that successful impulse transmission is enhanced by prior activity. The enhancement occurs if the axon is conditioned by the timing of the preceding activity so that a subsequent impulse arrives at the lesion while it exhibits supernormality (Shrager, 1993).

4. Temporary Exacerbation Caused by Warming: Uhthoff's Phenomenon

The security of conduction in demyelinated axons is markedly affected by temperature, and this effect can readily be demonstrated in the laboratory in both central (Fig. 7) (Smith et al., 2000) and peripheral experimentally demyelinated axons (Bostock et al., 1981; Davis et al., 1975; Rasminsky, 1973). The effects can be sufficiently strong that even subtle changes in body temperature can have a profound effect on the expression of symptoms by patients with

Figure 7 Two families of superimposed records obtained from excised dorsal columns examined in vitro, showing the effects of temperature on axonal conduction. The records on the left are from normal tissue, while those on the right include an experimental demyelinating lesion in the conduction pathway, induced 21 days previously. The records were obtained as the temperature in the central recording lane (containing the lesion, where present) was raised from 25°C to 37°C in 1°C intervals. (A) The temperature changes had little effect on conduction along normal axons. (B) In contrast, temperature changes had prominent effects on conduction in demyelinated axons, distinguished by their longer latency. At cooler temperatures many more demyelinated axons were able to conduct than at normal body temperature, when nearly all the demyelinated axons failed to conduct. (Reproduced from Smith et al., 2000.)

Figure 7 Two families of superimposed records obtained from excised dorsal columns examined in vitro, showing the effects of temperature on axonal conduction. The records on the left are from normal tissue, while those on the right include an experimental demyelinating lesion in the conduction pathway, induced 21 days previously. The records were obtained as the temperature in the central recording lane (containing the lesion, where present) was raised from 25°C to 37°C in 1°C intervals. (A) The temperature changes had little effect on conduction along normal axons. (B) In contrast, temperature changes had prominent effects on conduction in demyelinated axons, distinguished by their longer latency. At cooler temperatures many more demyelinated axons were able to conduct than at normal body temperature, when nearly all the demyelinated axons failed to conduct. (Reproduced from Smith et al., 2000.)

MS. Indeed, many patients with MS exhibit a worsening of some symptoms upon body warming, and the worsening is sufficiently robust and reproducible that at one time the phenomenon was used in the diagnosis of MS in the form of the "hot bath test" (Guthrie, 1951; Malhotra and Goren, 1981). The effect was first described in 1890 by Uhthoff, and the phenomenon now bears his name. Uhthoff-like phenomena can be provoked by a hot shower (Waxman and Geschwind, 1983), sunbathing (Berger and Sheremata, 1985; Harbison et al., 1989; Avis and Pryse-Phillips, 1995), exercise (van Diemen et al., 1992; Edmund and Fog, 1955), or even by the normal circadian change in body temperature (Fig. 8) (Scherokman et al., 1985). Just as warming can be deleterious to function, so cooling can be beneficial and can sometimes be realized after a cool bath or simply after drinking cold water. Indeed, improvement in vision has been documented after drinking iced water, and these functional improvements are accompanied by an increase in the amplitude of the visual-evoked potential (McDonald, 1986; Hopper et al., 1972).

It is generally accepted that temperature affects the success of conduction in demyelinated axons mainly by affecting the duration of the action potential (Paintal, 1966) at the driving node, the node just before the demyelinated region that is responsible for driving the current to depolarize the demyelinated axolemma to its firing threshold. The change in duration arises from the fact that the temperature coefficient for sodium inactivation is larger than that for sodium activation (Davis and Schauf, 1981). In axons balanced on the "knife edge" between conduction block and successful conduction (see Safety Factor) this change can easily be sufficient to decide whether conduction succeeds or fails. Thus although warming will increase the conduction velocity (Fig. 9) (Swadlow et al., 1981), thereby offsetting the reduction in conduction velocity produced by demyelination, the net effect of warming is typically deleterious to the patient, at least with regard to the expression of symptoms. It is worth mentioning that although the mechanism described (changes in action potential duration) may be paramount in most lesions in patients, the expression of Uhthoff's phenomenon may involve multiple mechanisms, perhaps including the modulation of nitric oxide production (see Chapter 18).

5. Therapeutic Strategies Based on Uhthoff's Phenomenon

That cooling can improve conduction in demyelinated axons has encouraged the search for pharmacological agents that might be able to achieve the same benefit, but at normal body temperature. At a time when it was believed that potassium currents curtailed action potential duration in mammalian axons, it was imaginatively surmised that blocking these currents might mimic the effects of body cooling and so restore conduction to demyelinated axons (Bostock et al., 1978; Sherratt et al., 1980; Davis and Schauf, 1981;

Figure 8 Visual-evoked potentials from a pilot with MS who experienced blurred vision each afternoon (visual acuity indicated). Visual acuity improved substantially in the afternoon within 3 minutes of drinking iced water, and this improvement was accompanied by an increase in the visual-evoked response. On a different day a similar improvement in acuity was accompanied by a reduction in the temperature of the tympanic membrane of 0.25°C. (Reproduced from Scherokman et al., 1985.)

Figure 8 Visual-evoked potentials from a pilot with MS who experienced blurred vision each afternoon (visual acuity indicated). Visual acuity improved substantially in the afternoon within 3 minutes of drinking iced water, and this improvement was accompanied by an increase in the visual-evoked response. On a different day a similar improvement in acuity was accompanied by a reduction in the temperature of the tympanic membrane of 0.25°C. (Reproduced from Scherokman et al., 1985.)

McDonald and Sears, 1970). Indeed, the potassium channel blocking agent 4-aminopyridine can be effective in restoring conduction to demyelinated axons (Sears and Bostock, 1981; Targ and Kocsis, 1985; Bowe et al., 1987), and prolonging action potential duration with scorpion toxin is also effective (Bostock et al., 1978). Based on findings such as these, 4-AP is being evaluated as a possible symptomatic therapy in MS (see Chapter 10), although whether the undoubtedly beneficial effects of the drug in patients are due to effects at the lesion or on synapses (Smith et al., 2000) is currently uncertain. Laboratory studies (Smith et al., 2000) have noted that at the very low concentrations that can safely be achieved in patients (the drug is strongly proconvulsant) it has no apparent effects on demyelinated axons, but it strongly potentiates synaptic transmission.

III. Hyperexcitability

Demyelinated axons not only can develop excitability along the demyelinated portion over time but also can become hyperexcitable, such that they generate trains of ectopic impulses. Such impulses are generated at the site of demyelination, conducting away from it in both directions (Smith and McDonald, 1982; Baker and Bostock, 1992). The impulse trains are composed of either continuous (at 10 to 50 Hz) or bursting discharges (e.g., 0.1 to 5 sec bursts separated by gaps of 0.1 to 100 sec) (Fig. 10), with the discharge pattern typically being consistent for any particular axon if left undisturbed. However, different patterns of impulse activity can continue simultaneously in different axons within a single lesion. Continuous discharges can arise because of the appearance of a slow, persistent inward sodium current along the demyelinated axolemma (Kapoor et al., 1997; Rizzo et al., 1996), whereas bursting discharges can arise from an inward potassium current that can arise if potassium ions accumulate in a compartment surrounding axons (Kapoor et al., 1993; Felts et al., 1995), although a definitive link between these mechanisms and these particular patterns of hyperexcitability has not been established (see reviews by Smith et al., 1997; Mogyoros et al., 2000; Baker, 2000; and Chapter 9).

Figure 9 Axonal conduction velocity increases, but safety factor is reduced, as temperature increases. (A) Changes in conduction latency for callosal axons, measured in rabbits, as body temperature is altered. Note the decrease in latency, reflecting increased conduction velocity, as temperature rises. (B, C) Extracellular recordings of action potentials elicited in a cal-losal efferent neuron by stimulation of its axon in the corpus callosum, at 36.5°C (B) and at 38.7°C (C), display more rapid conduction (resulting in earlier onset of the action potential) at higher temperature. For the second action potential evoked at high frequency, the safety factor is reduced, resulting in an inflection (arrow in B) between the initial segment spike and somatodendritic spike at 36.5°C. At 38.7C° the action potential fails to invade the soma (arrow in C) because of a superimposed temperature-induced decrease in safety factor. (Reproduced from Swadlow et al., 1981.)

Figure 9 Axonal conduction velocity increases, but safety factor is reduced, as temperature increases. (A) Changes in conduction latency for callosal axons, measured in rabbits, as body temperature is altered. Note the decrease in latency, reflecting increased conduction velocity, as temperature rises. (B, C) Extracellular recordings of action potentials elicited in a cal-losal efferent neuron by stimulation of its axon in the corpus callosum, at 36.5°C (B) and at 38.7°C (C), display more rapid conduction (resulting in earlier onset of the action potential) at higher temperature. For the second action potential evoked at high frequency, the safety factor is reduced, resulting in an inflection (arrow in B) between the initial segment spike and somatodendritic spike at 36.5°C. At 38.7C° the action potential fails to invade the soma (arrow in C) because of a superimposed temperature-induced decrease in safety factor. (Reproduced from Swadlow et al., 1981.)

Kapoor et al., 1992). It is intuitively likely that the trains of ectopic activity arising in sensory axons will contribute to, if not underlie, the positive sensory symptoms commonly associated with MS, such as tingling paraesthesia and perhaps pain. Evidence to support this view has been provided by recordings made during neck movement in patients exhibiting Lhermitte's sign (Nordin et al., 1984) (see Mechanosensitivity).

Observations in patients that sodium channel blocking agents (e.g., carbamazepine, lamotrigine, lidocaine, mexile-tine) are effective in controlling positive symptoms (Cianchetti et al., 1999; Sakurai and Kanazawa, 1999) also support a strong role for sodium channels in the generation of ectopic impulses.

0 0

Post a comment