Demyelination in spinal cord injury

Demyelination in spinal cord injury

Journal of the Neurological Sciences, 1989, 91:1-14 1 Elsevier JNS 03162 Review article Demyelination in spinal cord injury Stephen G. W a x m a ...

1MB Sizes 0 Downloads 19 Views

Journal of the Neurological Sciences, 1989, 91:1-14



JNS 03162

Review article

Demyelination in spinal cord injury Stephen G. W a x m a n Department of Neurology, Yale University School of Medicine, New Haven, CT (U.S.A.), and PVA/EPVA Center for Neuroscience and Regeneration Research, VA Medical Center, West Haven, CT (U.S.A.)

(Received 5 January, 1989) (Accepted 5 January, 1989)

SUMMARY Morphological and physiological studies demonstrate that demyelination constitutes a significant component of the pathology in compressive spinal cord injury. In many cases of spinal cord injury, a rim of demyelinated axons surrounds a central core of hemorrhagic necrosis. This provides a pathophysiological basis for "discomplete" spinal cord injuries, characterized by apparently complete transection as judged by clinical criteria, but with neurophysiological evidence of conduction through the level of damage. Recovery of conduction in demyelinated axons may permit recovery of function, and can be mediated by several mechanisms, including remyelination by oligodendrocytes or Schwann cells. Alternatively, conduction of action potentials can occur in the absence of remyelination, but this requires plasticity of the demyelinated axon. The biophysics of conduction favors recovery of electrogenesis after demyelination of small diameter axons. This may account, in part, for the observation that functional recovery is more common after demyelination of visual, compared to spinal, axons. Restoration or modification of conduction in demyelinated fibers represents an important strategy for promoting functional recovery in spinal cord injury.

Key words: Spinal cord injury; Demyelination; Remyelination; Myelin

Correspondence to: Stephen G. Waxman,M.D., Ph.D., Department of Neurology,LCI 707, Yale Medical School, 333 Cedar Street, New Haven, CT 06510, U.S,A. Tel.: (203) 785-4086.

0022-510X/89/$03.50 © 1989Elsevier Science Publishers B.V. (BiomedicalDivision)

The development of new strategies for treating spinal cord injury will depend on an understanding of its pathology and pathophysiology. In this regard, it is becoming clear that there is a spectrum of types of spinal cord injury. Transection is seen primarily in the context of penetrating injuries (e.g., knife and gunshot wounds). In contrast. contusion, characterized by central hemorrhagic necrosis, is often seen in closed injury (as occurs in most motor vehicle accidents). This article will discuss the emerging evidence that demyelination constitutes a significant component of the pathology in closed spinal cord injury, and will review some recent aspects of the pathophysiology of demyelination which are relevant to recovery of function in spinal cord Injury. Even early pathological studies suggested that demyelination occurs in compressive spinal cord injury. For example, Holmes (1906), in a description of chronic spinal cord compression due to tubercular disease of the vertebral column, observed preservation of axons with loss of myelin, and suggested that demyelination interfered with conduction in ascending and descending axons within the spinal white matter. More recently, morphological observations at both the light and electron microscopic levels have provided evidence for demyelination in white matter tracts in compressive spinal cord injury (Gledhill et al 1973; McDonald 1974; Harrison and McDonald 1977; Blight 1983a; Griffiths and McCulloch 1983). There is evidence which suggests that inflammatory mechanisms contribute to demyelination after spinal cord injury; for example, vesicular degeneration of myelin, together with stripping of myelin by macrophages, has been reported 7 days following experimental spinal cord injury (Blight 1985). Electrophysiological studies (Blight 1983b) which demonstrate decreased conduction velocity, prolonged refractory period, and decreased safety factor in axons studied at the site of spinal cord injury, are consistent with the presence of demyelination. These observations provide a pathological correlate for the observation (Dimitrijevic 1988) that, m some patients in whom spinal cord injury appears to be clinically complete, there is neurophysiological evidence for residual descending influences on spinal reflex activity. These latter observations indicate that, in addition to complete and incomplete spinal cord lesions, there is a third type of lesion, characterized by apparently complete spinal cord transection as judged by clinical criteria, but with neurophysiological evidence of axonal conduction through the site of injury. Dimitrijevic (1988) has termed these lesions "discomplete". Within these lesions, conduction may occur in ascending and descending fibers which maintain anatomic continuity. It is now well established that, under some conditions, conduction can be restored in previously demyelinated axons (Waxman 1988a). A number of studies demonstrate that demyelination is followed by remyelination (mediated either by oligodendrocytes or by Schwann cells) in experimental spinal cord injury (Gledhill and McDonald t977: Harrison and McDonald 1977; Griffiths and McCulloch 1983). The elegant observations of Blight and Young (1989) demonstrate that remyelination by Schwann cells. which presumably migrate into the spinal cord from the peripheral nervous system, is capable of restoring action potential conduction in sensory tracts after experimental spinal cord injury. This is an important observation, because it provides a possible basis for recovery of conduction after trauma to the spinal cord.

Recovery of conduction, in previously demyelinated fibers, provides a pathophysiological basis for clinical remissions in disorders such as multiple sclerosis (McDonald 1974; Waxman 1981, 1988b). Yet clinical remission occurs much more frequently in multiple sclerosis than in spinal cord injury. One purpose of the present paper is to briefly discuss the pathophysiology of demyelination in the context of spinal cord injury, and to consider mechanisms that might contribute to recovery of function following injury to the spinal cord.


Figure 1 shows, in schematic form, the ion channel organization of the mammalian myelinated fiber. The complex structure of the myelinated fiber, which includes a segregation of various types of ion channels, has important implications for the pathophysiology of demyelination. As indicated in Fig. 1, voltage-sensitive sodium channels (gNa) are clustered at a high density within the axon membrane at the node of Ranvier. It is likely that the density of sodium channels in the axon membrane at the node is at least 1000/#m z, in contrast to a density of < 25/#m 2 in the internodal axon membrane under the myelin sheath (Ritchie and Rogart 1977; Waxman 1977; Chiu and Ritchie 1981). This non-uniform distribution of sodium channels has important implications with respect to excitability of the nodal/internodal domains of the axon membrane. Electrophysiological studies have also demonstrated a complex organization of potassium channels along mammalian myelinated fibers. These studies have utilized pharmacological agents, which specifically block various types of potassium channels, in order to examine their distribution in myelinated fibers. Two pharmacological probes have proven particularly useful: 4-aminopyridine (4-AP) blocks the rapidly activating (or "fast") potassium channel (gKf)" Tetraethylammonium (TEA), in contrast, blocks the delayed rectifying channel or "slow" potassium channel (gxs)' Physiological studies show that fast (4-AP-sensitive) potassium channels are present in the paranodal and internodal parts of the axon membrane, under the myelin

CJNa ,-__gK_~_^__gl_~

CJKf _.

Fig. 1. Schematic diagram showing molecular organization of the myelinated fiber, gr~a = sodium channels, gKf = fast potassium channels, gKs = slow potassium channels, g m = inward rectifier channels. Note that sodium channels are clustered in high density at the node of Ranvier and are present in much lower densities in the internode, grr, responsible for repolarization of the action potential, are present in the internodal axon membrane; these channels are "unmasked" by demyelination.

sheath (Fig. 2) (Waxrnan and Foster 1980; Ritchie and Chiu 1981). Chiu and Ritchie (i 980, 1981) demonstrated the appearance of a voltage-sensitive potassium conductance in acutely demyelinated axons, concomitant with the increase in capacitance that occurs during acute demyelination. Voltage clamp (Brismar 1981) and longitudinal current analyses (Bostock et al. 1981) of chronically demyelinated fibers have also demonstrated that potassium channels are present in the bared axon membrane. It is likely that these channels, which are exposed by demyelination, correspond to the fast potassium




3 ImV



\ Fig. 2. Fast potassium channels mediate rapid repolarization of the action potential but are masked in myelinated fibers. Compound action potentials show the effect of blockade of potassium channelswith externally appfied 4-aminopyridin¢(4-AP), in premyelinated (A). mature myelinated (B), and demyelinated (C) axons in rat sciatic nerve. In premyelinated axons (A), after blockade of gin, repolarizat~ is delayed. In mydinated a x o n s (B),gKf are masked by the overlying myelin,and:externallyapp~d4-APdoes not block repolarization. Followingdemyeliaationwith lysophosphatidyi choline (C),grjare umnaaske.d;and b!oekade of these channels with 4-AP increases amplitude and duration of the action potential (arrow).

channels located in the internodal axon membrane. Since these channels can mediate rapid repolarization of the action potential (Kocsis et al. 1982, 1986), when unmasked by damage to the myelin, these channels will tend to hold the membrane close to E K, the potassium equilibrium potential (close to resting potential) and thereby interfere with action potential electrogenesis. More recent studies (Kocsis et al. 1986, 1987; Baker et al. 1987; Gordon et al. 1988) show that "slow" potassium channels (gKs) are also present in the axon membrane in some mammalian myelinated fibers. These channels, which are blocked by TEA, are activated by prolonged depolarization of the axon. Whereas the fast potassium channels mediate repolarization of the action potential, the slow potassium channels are activated during high frequency discharge, and modulate repetitive firing patterns (Fig. 3). These slow potassium channels thus participate in accommodative properties of axons. The available evidence suggests that the slow potassium channels are present in the axon membrane at the node of Ranvier, and possibly also in the internodal axon membrane. There is also evidence for a fourth type of ion channel, the "inward rectifier" (gIR), in the axon membrane of myelinated fibers (Baker et al. 1987; Eng et al. 1988). These channels, which are blocked by exposure to cesium, activate in response to hyperpolarization of the axon membrane. Although their properties are not yet fully understood, these channels appear to be permeable to both sodium and potassium ions. Since they limit the degree ofhyperpolarization that the membrane can sustain, these channels may play a role in modulating the excitability of the axon, or may control hyperpolarization of the axon membrane by other channels or pumps. The differentiation of nodal and internodal regions of the axon membrane in mammalian myelinated fibers, as shown in Fig. 1, has significant implications for the pathophysiology of demyelinated fibers: (i) Following acute loss of myelin, the low density of sodium channels in demyelinated (formerly internodal) axon regions will result in decreased electrical excitability. As a result of this there is a requirement, if conduction is to occur in demyelinated fibers, for reorganization which can support action potential electrogenesis through regions characterized by relatively low densities of sodium channels. (ii) In addition, unmasking of fast potassium channels, after damage to the myelin, will tend to hold the demyelinated axon membrane at a value close to the potassium equilibrium potential, impeding conduction. This observation suggests that pharmacological blockade of 4-AP-sensitive potassium channels should result in increased safety factor for conduction in demyelinated fibers (Davis and Schauf 1981). A number of studies have demonstrated restoration of conduction following exposure of demyelinated fibers to 4-AP (Sherrett et al. 1980; Targ and Kocsis 1985). Moreover, transient symptomatic improvement has been observed in multiple sclerosis patients treated with 4-AP (Stefoski et al. 1987). In this regard, it is notable that Blight and Gruver (1987) have demonstrated augmentation of reflexes mediated by bulbospinal pathways, suggesting restoration of conduction in axons traversing the lesion, in spinal cord-injured cats following treatment with 4-AP. (iii) It is possible that changes in the organization of slow potassium channels, or the inward rectifier, may alter the accommodative properties or bursting patterns in demyelinated fibers. It is not yet clear what effect this will have on neural coding.







'mY I


g'e B 60

msec 4-AP+TEA



/ I Lh:IL










6o msec

Fig. 3. (A,B) Recordings from rat sensory axon, showing modulation of repetitive firing after blockade of slow potassium channels (gr~) by TEA. An after-hyperpolarizing potential (A, open arrow) terminates action potential burst activity. This potential is produced by activation of gKs channels, and is abolished (B) by TEA. Note the increased burst activity (B, arrow) aider blockade ofgt,. (C,D) Recordings from motor axon in sciatic nerve. An atter-hYl~polarization (open arrow) is present afrer repetitive activity (C) and is attenuated by blockade ofgK, with TEA (D). Modified from Kocsis et al. (1987).

REMYELINATION IN SPINAL CORD INJURY T h e available evidence strongly suggests that some fibers are r e m y d i n a t e d in experimental spinal c o r d injury (Gledhill et al. 1973; Gtedhill a n d M c D o n a l d 1977; H a r r i s o n and M c D o n a l d 1977; Griffiths a n d M e C u l l o e h 1983; Blight and_Young 1989).

It is expected, on theoretical grounds, that remyelination should promote conduction through previously demyelinated fibers, assuming that remyelinated nodes of Ranvier develop membrane properties similar to those in normal fibers (Koles and Rasminsky 1972). Morphological (Weiner et al. 1981) and pharmacological (Ritchie et al. 1981) studies provide evidence, in fact, for the development of relatively normal sodium channel densities at remyelinated nodes of Ranvier. Biophysical studies (Koles and Rasminsky 1972) suggest that even thin remyelinated sheaths should support the conduction of action potentials. Since internode distances are reduced along remyelinated fibers, conduction velocity is expected to be reduced compared to normal (Brill et al. 1977), but it should be recalled that reduced conduction velocity does not necessarily lead to conduction deficit (Halliday and McDonald 1977; Waxman 1988b). Thus, remyelination should still be expected to lead to clinical improvement on the basis of reversal of conduction block. There is electrophysiological evidence for restoration of conduction following remyelination in dorsal columns of the cat after injection of lysophosphatidyl choline (Smith et al. 1981). Similarly, the recent study by Blight and Young (1989) suggests that conduction is restored in some spinal cord sensory fibers, following experimental spinal cord injury, as a result of remyelination. Notably, this latter demonstration suggests restoration of conduction following remyelination by Schwann cells. These findings are consistent with earlier work (Black et al. 1986), which showed that, following delayed remyelination by either oligodendrocytes or Schwann cells, spinal cord axons develop membrane characteristics similar to those of normal myelinated fibers. The recent physiological findings are particularly important in providing electrophysiological correlates that suggest a favorable effect on conduction that is mediated by Schwann cell myelination within the spinal cord. In this regard, attempts to promote CNS remyelination via Schwann cells (Harrison 1980; Sims and Gilmore 1983; Blakemore and Crang 1985; Duncan et al. 1988a,b) are of considerable interest.


Some demyelinated fibers demonstrate the capability to conduct action potentials through regions of demyelination. For example, continuous conduction of action potentials, manifested by continuously distributed inward membrane current, has been demonstrated along rat ventral root fibers demyelinated with diphtheria toxin (Bostock and Sears 1976, 1978). Continuous conduction was observed beginning 4-6 days following demyelination, and propagated at 1/20 to 1/40 of the velocity expected for saltatory conduction. These observations indicate that the demyelinated internodal axon membrane in some fibers develops electrical excitability, i.e., a sodium channel density sufficient to support action potential conduction in areas of demyelination. Morphological studies have demonstrated, in some demyelinated axon regions, the development of increased densities of sodium channels (Foster et al. 1980; Coria et al. 1985; Meiri et al. 1985). This raises the important question of how sodium channels are synthesized and transported to appropriate parts of the axon. Neuronal

synthesis of new channels may underlie the development of increased sodium channel densities in demyelinated axons. However, it is also possible that nearby glial or Schwann cells may function as extraneuronal sites of synthesis of ion channels that are subsequently transferred to the axon (Gray and Ritchie 1985; Black and Waxman 1988; Black et al. 1989). It should be noted, however, that conduction through demyelinated regions does not necessarily require the acquisition of sodium channel densities as high as those at nodes of Ranvier. Waxman and Brill (1978) demonstrated that, if the prerequisites for invasion of demyelinated axon regions are met, a sodium channel density that is much lower than at normal nodes, can support continuous conduction of action potentials in demyelinated regions. It has been suggested (Waxman and Brill 1978) that development of continuous conduction may be mediated by redistributton of pre-existing channels, which were formally localized at nodes of Ranvier. following demyelination. It is also possible that, in some fibers, there is a sufficient density of sodium channels in the internode to support conduction following demyelination. For example. conduction of action potentials in premyelinated fibers (which do not yet have myelin sheaths) in rat optic nerve is mediated by a sodium channel density of approximately 2//~m2 (Waxman et al. 1988). As a result of the high input resistance (which is proportional to [diameter] - 3/2) of small diameter axons, activation of a small number of sodium channels can depolarize these fibers to threshold (Hille 1970). Waxman et al. (1988) suggested that, in some demyelinated fibers, a similar mechanism may contribute to restoration of conduction. While the sodium channel density of the internodal axon membrane is much lower than that at the node, the internodal sodium channel density may exceed 2/#m 2 (Ritchie and Rogart 1977; Waxman and Quick 1977). Prineas and Connell (1978, 1979) have observed that axon diameter is reduced in demyelinated regions. An example is shown in Fig. 4. It is possible that, in some demyelinated axons, the reduction in diameter is sufficient to provide an increase in input resistance that permits conduction to occur on the basis of pre-existing internodal sodium channels. It is interesting that patients with multiple sclerosis (including the "spinal" form of multiple sclerosis) often display asymptomatic increases in latency of the visual evoked potential; whereas patients with optic neuritis (including patients who go on to develop clinically definite MS) much less frequently display asymptomatic delays in conduction of the somatosensory evoked potential (Hume and Waxman 1988). This is especially notable when interpreted in the context of the longer conduction distance along the somatosensory pathway. This observation suggests that functional recovery occurs tess frequently, following demyelination of the dorsal columns, than after demyetination within the optic nerves and chiasm. The observation, that asymptomatic slowing of conduction is more common along the visual pathways than along spinal somatosensory pathways, may have implications with respect to mechanisms underlying recovery of function in demyelinated fibers. Myelinated axons within the optic nerve exhibit relatively small diameters, with a mean axonal diameter of 0.77 #m in the rat (Foster et al. 1982). In contrast, myelinated axons within the dorsal columns exhibit larger diameters (see e.g., Hildebrand 1971). As noted above, input impedance for cylindrical axons is proportional to (diameter) -3/2. Thus, the increase of input

Fig. 4. Axonal diameter is reduced in demyelinatedfibers. This electron micrograph shows a demyelinated fiber after injection of lysophosphatidylcholineinto rat sciatic nerve. The myelin sheath is vesiculated(V). Note the reduction in diameter of the axon (A) in region of total demyetinationto the right of the arrow. × 6600. From Smith et al. (1983). impedance, following demyelination of a small diameter fiber, might be expected to be greater than that occurring after demyelination of a larger fiber. This would provide an explanation for the observation in multiple sclerosis patients, that demyelination is less likely to produce clinical deficits (i.e., is more likely to be asymptomatic) in fiber tracts characterized by small diameter such as the optic nerve, than in spinal tracts that are characterized by larger diameter fibers such as the dorsal columns. If this explanation is correct, it has important implications for spinal cord injury.

IMPEDANCE MATCHING AND CONDUCTION IN INJURED AXONS Development of an adequate sodium channel density in the demyelinated region does not necessarily insure reliable impulse conduction after axonal injury. A critical factor in determining whether conduction will occur into a demyelinated zone is impedance mismatch. Impedance mismatch can occur at the transition between myelinated and demyelinated areas, on the basis of the abrupt increase in surface area of the axon membrane in the region of demyelination, which produces decreased current density due to capacitative loading (Waxman 1977). Observations of conduction through chronically demyelinated axons (see e.g., Bostock and Sears 1978) suggest that impedance mismatch is overcome in some instances. Mechanisms for impedance



oE o_




Im~,e 2














I)4 456


E 0


I rnle¢










Fig. 5. (A) Impedance mismatch interferes with conduction in demyelinated axons, The computer s'maulations in this figure show conduction through a focally demyelinated axon. Evenlin the presence of an adequate density of sodium ehannets in the demyelinated zone (D~-D4), conduction can fall,as a result of impedance mismatch. Modified from Waxman and Brill (1978). (B) Interposition of short myelinated segments (A-B; B-D1), proximal to the demyetinated zone, can overcome impedance mismatch and facilitate invasion of action potentials which propagate along the demyelinated axon. Modified from Waxman and Brill (1978)~

11 matching include the development of relatively short internodes proximal to the demyelinated region (Waxman and Brill 1978), the development of larger-than-normal current densities at the node proximal to the demyelinated area, reduction of diameter of the demyelinated axon region (Bostock and Sears 1978), or the development of specialized transition zones (characterized by relatively high sodium channel densities or low potassium channel densities) at the edge of the demyelinated zone (Waxman and Wood 1984). Figure 5A illustrates impedance mismatch in a focally demyelinated fiber (between nodes D 1 and D4), under the assumption that the demyelinated axon membrane has reorganized so as to exhibit a high density of sodium channels (Waxman and Brilt 1978). Conduction of action potentials fails at the junction between normal and demyelinated axon zones (D1), despite the assumption of excitability of the axon in the demyelinated domain. In contrast, Fig. 5B shows unimpeded action potential conduction in a similar fiber, with two interposed short internodes proximal to the demyelinated zone (A-B; B - D 0. Similarly, interposition of a small "transition zone" of altered axon membrane properties (high sodium channel density, or low potassium channel density) at the border between myelinated and demyelinated regions, can facilitate invasion of action potentials into the demyelinated region (Waxman and Wood 1984). In comparing multiple sclerosis (where conduction frequently occurs through demyelinated axons) and spinal cord injury (where conduction through demyelinated axons appears to be less common), attention must thus be directed toward the structure of the injured axons at the junction between myelinated and demyelinated areas. Short myelinated internodes (Suzuki et al. 1969; Gledhill et al. 1973; Prineas and Connell 1978, 1979), have been described in central demyelinated fibers in a variety of pathologies. Yet, it is quite possible that the fine structure, of myelinating cells as well as axons, is different in various pathologies. Thus, it is possible that the degree of impedance mismatch may be different in spinal cord injury, as compared to multiple sclerosis. In this regard, therapeutic modification of remyelination, at the edge of the area of injury, deserves careful attention. CONCLUDINGREMARKS There is a spectrum of strategies for promoting functional recovery after spinal cord injury, including regeneration of axons (if necessary via transplantation of appropriate bridging tissue or substrates; Aguayo et al. 1985; Smith and Silver 1988), neurogenesis (Anderson and Waxman 1985), recruitment of central pattern generators below the site of transection (Grillner and Dubuc 1988) and utilization of alternative conduction pathways (Nathan and Smith 1969). Clearly, modification of demyelinated fibers, with the acquisition of the capability for impulse conduction, may provide an additional mechanism for conduction of ascending and descending neural information in spinal cord injury. In this regard, it is significant that preservation of conduction in less than 10~o of the axons in the spinal cord can suffice to support locomotor capability after cord transection (see Windle et al. 1958; Young 1988). Thus, restoration of conduction in all spinal cord axons (or even a majority) may not be required for recovery

12 o f s o m e function. T h i s r e d u n d a n c y , t o g e t h e r w i t h b i o m e c h a n i c a l o r g a n i z a t i o n o f the spinal c o r d w h i c h f a v o r s p r e s e r v a t i o n o f s o m e o f the p e r i p h e r a l l y l o c a t e d w h i t e m a t t e r a x o n s ( Y o u n g 1988), suggests t h a t further a t t e n t i o n s h o u l d b e d i r e c t e d t o w a r d u n d e r s t a n d i n g d e m y e l i n a t i o n , a n d r e c o v e r y o f c o n d u c t i o n f o l l o w i n g d e m y e l i n a t i o n , in spinal c o r d injury.

ACKNOWLEDGEMENTS W o r k in the a u t h o r ' s l a b o r a t o r y h a s b e e n s u p p o r t e d in p a r t by g r a n t s f r o m t h e National Institute of Neurological Disorders and Stroke and the National Multiple Sclerosis Society, a n d by the M e d i c a l R e s e a r c h Service, V e t e r a n s A d m i n i s t r a t i o n . I also t h a n k the D a n i e l H e u m a n n

F u n d for s u p p o r t .

REFERENCES Aguayo, A.J. (1985)Axonal regeneration from injured neurons in mammalian central nervous system. In: C. Cotman (Ed.), Synaptic Plasticity, Guilford Press, New York, pp. 457-483. Anderson. M.J. and S.G. Waxman (1985) Neurogenesis in adult vertebrate spinal cord in situ and in vitro: a new model system. Ann. N.Y. Acad. Sci., 457: 213-233. Baker, M., H. Bostock, P. Grafe and P. Martins (1987) Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J. PhysioL, 383: 45-68. Black, J.A. and S.G. Waxman (1988)The perinodal astroeyte. Glia. 1: 169-183. Black, J.A., S.G. Waxman. T.J. Sims and S.A. Gilmore (1986) Effects of delayed myelination by oligodendrocytes and Schwann cells on the macromolecular structure of axonal membrane in rat spinal cord. J. Neurocytol.. 15:745-762 Black, J.A., B. Friedman, S.G. Waxman. L.W. Elmer and K.J. Angelides (1989) Immuno-ultrastructurat localization of sodium channels at nodes of Ranvier and perinodal astrocytes in rat optic nerve. Proc. R. Sac. Land. Ser. B, m press. Blakemore. W.F. (1977) Remyetination of CNS axons by Schwann cells transplanted from sciatic nerve. Nature. 266: 68-69. Blakemore. W. F. and A. J. Crang (1985) The use of cultured autologous Schwann cells to remyelinate areas of demyelination in the central nervous system. J. NeuroL Sci., 70: 207-223. Blight, A.R. (1983a) Cellular morphology of chronic spinal cord injury in the cat: analysis of myelinated axons by line-sampling. Neuroscience. 10: 521-543. Blight, A.R. (1983b) Axonal physiology of chronic spinal cord injury in the cat: intraceltular recording in vitro. Neuroscience, 10: 1471-1486. Blight, A. R. (t985) Delayed demyetination and macrophage invasion: a candidate for secondary celt damage in spinal cord injury. Central Nervous System Trauma. 2: 299-314. Blight, A. R. and J.A. Gruner (1987) Augmentation by 4-aminopyridine of vestibulospinal free fall responses in chronic spinal-injured cats. J. Neurol. Sci.. 82: 145-t59. Blight, A.R. and W. Young (1989) Central axons in injured cat spinal cord recover etectrophyslological function following remyelination by Sehwann cells. J. Neurol. Sci., 91: 15-34. Bostock. H, and T. A. Sears (1976) Continuous conduction in demyelinated mammalian nerve fibres. Nature. 263: 786-787. Bostock. H. and T.A. Sears (1978) The internodal axon membrane: electrical excitability and continuous conduction m segmental demyelination. J. PhysioL, 280: 273-301. Bostock. H.. T. A. Sears and R.M. Sherratt (1981) The effects of 4-aminopyridine and tetraethylammonium ions on normal and demyelinated mammalian nerve fibers. J. Physiol.. 313: 301-315. Brill, M.H.. S.G. Waxman, J.W. Moore and R.W. Joyner (1977) Conduction velocity and spike configuration in myelinated fibers: computed dependence on intemode distance. J. Neurol. Neurosurg, Psychiat.. 40: 769-774. Brismar. T. (1981) Specific permeability properties of demyelinated rat nerve fibers. Aeta Physiol. Scand.. 113: 167-176.

13 Chiu, S.Y. and J.M. Ritchie (1980) Potassium channels in nodal and internodal axonal membrane in mammalian myelinated fibers. Nature, 284: 170-171. Chiu, S. Y. and J. M. Ritchie (1981) Evidence for the presence of potassium channels in the internodal region of acutely demyelinated mammalian single nerve fibres. J. Physiol., 313:415-437. Coria, F., F. Silos, R. Fernandez, F. Mouton and M. La Farga (1985) Demyelination-induced plasticity in the axon membrane: an ultrastructural cytochemical study of lead neuropathy in the rat. Neurosci. Lett., 58: 359-364. Davis, F.A. and C.L. Schauf (1981) Approaches to the development of pharmacological interventions in multiple sclerosis. In: Waxman, S. G. and J. M. Ritchie (Eds.), Demyelinating Disease: Basic and Clinical Electrophysiology, Raven Press, New York. Dimitrijevic, M. R. (1988) Residual motor functions in spinal cord injury. In: Waxman, S. G. (Ed.), Functional Recovery in Neurological Disease, Raven Press, New York, pp. 139-155. Dimitrijevic, M.R., M.M. Dimitrijevic, J. Faganel and A.M. Sherwood (1984) Suprasegmentally induced motor unit activity in paralyzed muscles of patients with established spinal cord injury. Ann. Neurol., 16: 216-221. Duncan, 1. D., J. P. Hammang and S.A. Gilmore (1988) Schwann cell myelination of the myelin deficient rat spinal cord following X-irradiation. Glia, 1: 233-239. Duncan, I.D., J.P. Hammang, K.F. Jackson, P.M. Wood, R.P. Bunge and L. Langford (1988) Transplantation of oligodendrocytes and Schwann cells into the spinal cord of the myelin-deficient rat. J. Neurocytol., 17: 351-360. Eng, D. L., T. R Gordon, J.D. Kocsis and S.G. Waxman (1988) Inward rectification in mammalian CNS axons. Abstr. Soc. Neurosci., 14: 946. Foster, R.E., C.C. Whalen and S.G. Waxman (1980) Reorganization of the axonal membrane of demyelinated nerve fibers: morphological evidence. Science, 210: 661-663. Foster, R. E., B. Connors and S.G. Waxman (1982) Rat optic nerve: electrophysiological, pharmacological and anatomical studies during development. Dev. Brain Res., 3: 361-376. GledhiU, R.F. and W.I. McDonald (1977) Morphological characteristics of central demyelination and remyelination: a single-fiber study. Ann. Neurol., 1: 552-560. Gledhill, R.F., B.M. Harrison and W.I. McDonald (1973) Demyelination and remyelination after acute spinal cord compression. Exp. Neurol., 38: 472-487. Gordon, T. R., J. D. Kocsis and S.G. Waxman (1988) Evidence for the presence of two types of potassium channels in the rat optic nerve. Brain Res., 447: 1-10. Gray, P.T. and J.M. Ritchie (1985) Ion channels in Schwann and glial cells. Trends Neurosci., 8: 411-415. Griffiths, I.R. and M.C. McCuUoch (1983) Nerve fibers in spinal cord impact injuries. 1. Changes in the myelin sheath during the initial five weeks. 3. NeuroL Sci., 58: 335-345. Grillner, S. and R. Dubuc (1988) Control of locomotion in vertebrates: spinal and supraspinal mechanisms. In: Waxman, S.G. (Ed.), Functional Recovery in Neurological Disease, Raven Press, New York, pp. 425-454. Halliday, A. M. and W.I. McDonald (1977) Pathophysiology of demyelinating disease. Br. Med. Bull., 33: 21-27. Harrison, B.M. (1980) Remyelination by cells introduced into a stable demyelinating lesion in the central nervous system. J. Neurol. Sci., 46: 63-81. Harrison, B.M. and W. I. McDonald (1977) Remyelination after transient experimental compression of the spinal cord. Ann. Neurol., 1:542-551. Hildebrand, C. ( 1971) Caliber spectra of some fiber tracts in feline central nervous system during postnatal development. Acta PhysioL Scand., Suppl., 364: 5-42. Hille, B. (1970) Ion channels in nerve membranes. Progr. Biophys. Mol. Biol., 21: 1-32. Holmes, G. (1906) On the relation between loss of function and structural change in focal lesions of the central nervous system, with special reference to secondary degeneration. Brain, 29: 514-523. Hume, A.L. and S.G. Waxman (1988) Evoked potentials in suspected multiple sclerosis: diagnostic value and prediction of clinical course. J. Neurol. Sci., 83: 191-210. Kocsis, J. D., S.G. Waxman, C. Hildebrand and J.A. Ruiz (1982) Regenerating mammalian nerve fibres: changes in action potential waveform and firing characteristics following blockage of potassium conductance. Proc. R. Soc. Lond. Ser. B, 217: 277-287. Kocsis, J. D., T. R. Gordon and S. G. Waxman (1986) Mammalian optic nerve fibers display two pharmacologically distinct potassium channels. Brain Res., 393: 357-361. Kocsis, J.D., D.L. Eng, T.R. Gordon and S.G. Waxman (1987) Functional differences between 4-aminopyridine and tetraethylammonium-sensitive potassium channels in myelinated axons. Neurosci. Lett., 75: 193-198. Koles, Z.J. and M. Rasminsky (1972) A computer simulation of conduction in demyelinated nerve fibres. J. Physiol., 227: 351-364.

14 McDonald. W. 1. (1974) Remyelination in relation to clinical lesions of the central nervous system. Br. Med. Bull.. 30: 186-189. Meiri, H., S. Pre-Chen and A.D. Koczya (1985) Sodium channel localization in rat sciatic nerve following lead-induced demyelination, Brain Res., 359: 326-331. Nathan, P.W. and M.C. Smith C1969) Effects of two unilateral cordotomies on the motility of the lower limbs. Brain. 96: 471-494. Prineas, J. and F. Connell (1978) The fine structure of chronically active multiple sclerosis plaques. Neurology, 28: 68-75. Prineas, J. and F. Connell (1979) Remyelination in multiple sclerosis. Ann. Neurol.. 5: 22-31. Ritchie. J. M. and R. B. Rogart (1977) The density of sodium channels in mammalian myelinatednerve fibers and nature of the axonal membrane under the myelin sheath. Proc. Natl. Acad. Sci. USA. 74:211-215. Ritchie, J.M. and S.Y. Chin (1981) Distribution of sodium and potassium channels in mammalian myelinated nerve. In: Waxman, S. G. and J. M. Ritehie (Eds.),DemyelinatingDiseases: Basic and Clinical Electrophysiology, Raven Press, New York. Ritchie, J.M.. H.P. Rang and R. Pellegrino (1981) Sodium and potassium channels in demyelinated and remyelinated mammalian nerve. Nature, 294: 257-259. Sherratt, R.M., H. Bostock and T.A. Sears (1980) Effects of 4-aminopyridine on normal and demyelinated mammalian nerve fibres. Nature, 283: 570-572. Sims, T.J. and S.A. Gilmore (1983) Interactions between intraspinal Schwann cells and the cellular constituents normally occurring in rat spinal cord: an ultrastructural study in the irradiated rat. Brain Res., 276: 17-30. Smith. G. M and J. Silver (1988) Transplantation of immature and mature astrocytes and their effect on scar formation in lesioned CNS. In: Gash, D.M. and J. R. Sladek, Jr. (Eds.), Progress in Brain Research. Vol. 78, Elsevier Science Publishers, Amsterdam-New York-Oxford, chapter 45, pp. 353-361. Smith, K.J., W.F. Blakemore and W.I. McDonald (1981) The restoration of conduction by central remyelination. Brain, 104: 383-404. Smith. M.E., J.D. Kocsis and S.G. Waxman (1983) Myelin protein metabolism in demyelination and remyetination in the sciatic nerve. Brain Res., 270: 37-44. Stefoski. D., F.A. Davis, M. Faut and C.L. Sehauf (1987) 4-Aminopyridine improves "clinical signs in multiple sclerosis. Ann. Neurol., 21: 71-77. Suzuki. K.. J.M. Andrews, J.J. Waltz and R. D. Terry (1969) Ultrastructural studies of multiple sclerosis. Lab. Invest.. 20: 444-451. Targ, E.F. and J. D. Kocsis (1985) 4-Aminopyridine leads to restoration of conduction in demyelinated rat sciatic nerve. Brain Res., 328: 358-361. Waxman, S.G. (1977) Conduction in myelinated, unmyelinated, and demyelinated fibers. Arch. Neurol.. 34: 585-590. Waxman, S.G. (1981) Clinicopathological correlations in multiple sclerosis and related diseases. In: Waxman. S.G. and J.M. Ritchie (Eds.), Demyelinating Diseases: Basic and Clinical Electrophysiology, Raven Press, New York, pp. 299-312. Waxman, S. G. (1988a) Biophysical mechanisms of impulse conduction in demyelinated axons. In: Waxman, S.G. (Ed.), Functional Recovery in Neurological Disease, Raven Press, New York. pp. 185-214. Waxman, S. G. (1988b) Clinical course and electrophysiology of multiple sclerosis. In: Waxman, S. G. (Ed.), Functional Recovery in Neurological Disease, Raven Press, New York, pp. 157-184. Waxman, S.G. and M.H. Brill (1978) Conduction through demyelinated plaques in multiple sclerosis: computer simulations of facilitation by short internodes. J. Neurol. Neurosurg, Psychiat,, 41: 408-4t7. Waxman. S.G. and R. E. Foster (1980) Ionic channel distribution and heterogeneity of the axon membrane in myelinated fibers. Brain Res. Rev., 2: 205-234. Waxman. S.G. and D.C. Quick (1977) Cytochemical differentiation of the axon membrane in A- and C-fibers. J. Neurol. Neurosurg. Psychiat,, 40: 769-774. Waxman, S.G. and S. L. Wood (1984) Impulse conduction in inhomogeneous axons; effects of variation in voltage-sensitive ionic conductances on invasion of demyelinated axon segments and preterminal fibers. Brain Res., 294: 111-122. Waxman, S. G., J. A. Black, J. D. Kocsis and J.M. Ritchie (1989) Low density of sodium channels supports action potential conduction in axons of neonatal optic nerve. Proc. Natl. Acad. Sci. USA. in press. Weiner. L.P., S.G. Waxman, S.A. Stohlman and A. Kwan (1980) Remyelination following viral-induced demyelination: ferric ion-ferrocyanide staining of nodes of Ranvier within the CNS. Ann. NeuroL. 8: 580-583. Windle. W. F , J.O. Smart and J.J. Beers (1958) Residual function after subtotal spinal cord resection in adult cats. Neurology, 8:518-522. Young, W. (I 988) Recovery mechanisms in spinal cord injury: implications for regenerative therapy. In: Seit. F.J. (Ed.), Neural Regeneration and Transplantation, A.R. Liss Publ.. New York, pp. 157-169.