Spinal cord injury (SCI) is a devastating condition afflicting over 13,000 people annually in North America and is an important cause of mortality and neurological morbidity1.
Although early pharmacological intervention after SCI with methylprednisolone2,3 or GM-1 ganglioside4 results in modest neurological improvement, the overall impact of these treatments remains minimal. Therefore, novel therapeutic approaches are required to improve the neurological outcome of these patients.
Mechanisms of cellular injury after neurotrauma
Posttraumatic degradation of white matter is an important aspect of the pathophysiology of SCI. Axonal degeneration after traumatic SCI evolves over minutes to days after injury5,6,7,8.
The pathophysiology of neurotrauma involves an initial or “primary” mechanical insult followed by a complex series of molecular and cellular events termed “secondary injury”, which include ischemia, rises in intracellular Na+ and Ca2+, glutamate toxicity, free-radical mediated cell damage and apoptosis1,6,7,9-17.
Relevance of axonal dysfunction to spinal cord injury
The spinal cord is rarely totally transected even after severe SCI associated with complete paralysis18. The injury site is characterized by central cavitation and a subpial rim of surviving axons with varying degrees of demyelination8,19-23.
We examined the relationships among the severity of SCI, the number of surviving axons at the injury site and the extent of neurological recovery of rats24.
A logarithmic relationship was found between the number of myelinated axons surviving in the subpial rim of the spinal cord after SCI and recovery of neurological function. Persistence of approximately 45,000 axons after SCI (~12% of normal number) was associated with recovery of significant hind limb locomotor function. This finding suggests that relatively small changes in neuroanatomical integrity or restoration of function of a small number of axons in the CNS can impact substantially in clinical neurological recovery.
Our laboratory and others have shown that axons that persist in the subpial rim after SCI display dysfunctional conduction properties including prolonged refractory periods, high frequency conduction failure, conduction block at subphysiological temperatures, increased threshold of activation and reduced conduction velocity25-30.
Role of K+ channels in axonal dysfunction after SCI
Experiments by Chiu and Ritchie31 demonstrated that K+ channels are localized in the myelin covered paranodal or internodal regions of peripheral mammalian myelinated axons. They showed that there was little K+ conductance in intact myelinated nerves and that there was a significant increase in outward K+ current upon acute myelin disruption. Several functions ascribed to the “fast” K+ channels include action potential repolarization32, stabilizing the node to prevent re-excitation of the node after a single stimulus33-36, limiting excessive axonal depolarization and inactivation of nodal Na+ channels37, and increasing the security of axonal conduction by contributing to the nodal potential33.
There is evidence supporting the concept that “fast” or rapidly activating K+ channels are concealed by myelin in the paranodal or internodal regions. Following disruption of the myelin there is an increased activity of these channels38 which “clamp” the membrane potential close to the equilibrium potential of K+ and result in axonal conduction blockade. Our laboratory and others have shown that altered activity of 4-aminopyridine (4-AP) sensitive K+ channels is associated with axonal dysfunction after SCI21-22 ,39-43. The results from our laboratory are congruent with those of Blight26 who has shown that 4-AP can relieve temperature dependent conduction block in chronically injured spinal cord axons and Kocsis44 who reported that traumatic disruption of paranodal myelin loops of spinal cord white matter was associated with enhanced sensitivity to 4-AP. Blight and Gruner observed that intravenous administration of 4-AP to cats with chronic thoracic SCI augmented the vestibulospinal free fall responses45. Furthermore, preliminary clinical studies with 4-AP have shown enhanced central axonal conduction and modest degrees of functional improvement in patients with chronic SCI45-50. It is noteworthy that 4-AP blocks a variety of K+ channels51 at various PNS and CNS sites so the specific mechanism of this drug is still unknown57. While the effect of 4-AP after SCI has been assumed to be on the basis of blockade of rapidly activating, voltage-gated K+ channels in juxtaparanodal regions of axons26, K+ channels are also located on astrocytes52, oligodendrocytes53 and oligodendrocyte precursor cells53-55.
Molecular structure, electrophysiological properties and pharmacological sensitivity of Kv1.1 and Kv1.2 K+ channels
Both Kv1.1 and Kv1.2 encode K+ channel proteins having 6 transmembrane (S1-S6) domains and one intervening pore (P or H5) sequence. Kv1.1 in both rat (RCK1)56 and mouse (MBK1 or MK1)57,58 consists of 495 amino acids. The length of the cDNA sequence varies from 3822 bp for the rat RCK1 channel56 to 2222 bp for the mouse MK1 channel. Kv1.2 encodes a 498 amino acid length sequence for the rat (RCK5)59 and a 499 amino acid protein for the mouse (MK2)60. The lengths of cDNA encoding for RCK5 are 2409 bp and 3383 bp for MK2. The K+ channel is tetrameric, requiring the co-assembly of four α subunit proteins to form one functional channel61,62. Kv1.1 and Kv1.2 can co-assemble to form heteromultimeric K+ channels63,64.
In general, Kv1.1 and Kv1.2 encode voltage-gated K+ channels that express outward currents with depolarizing pulses59. Both Kv1.1 and Kv1.2 activate rapidly and inactivate slowly. Detailed biophysical parameters are listed for rat and mouse from the following references59, 63. A variety of pharmacological blockers inhibit Kv1.1 and Kv1.2 K+ channel currents with different rank and order of affinities59,65. Kv1.1 is blocked by 4-AP at affinities ranging from IC50=89 mM to IC50=1 mM59,65-67. Kv1.2 is also sensitive to 4-AP blockade mainly in the submillimolar concentration ranging from IC50=74 mM to IC50=0.8 mM 1 mM59,66,68. Kv1.1 has a high affinity for TEA blockade (IC50=0.6 mM; 59 Kd=0.3 mM)65, while Kv1.2 is insensitive to TEA ( IC50=129 mM59, Kd=560 mM)65, both Kv1.1 ( IC50=12 nM59, Kd=20 nM)65, and Kv1.2 (IC50=4 nM59, Kd=17 nM)65, are highly sensitive to nM blockade by Dendrotoxin (or α-Dendrotoxin).
Subcellular distribution of Kv1.1 and Kv1.2 K+ channels in myelinated axons and changes following challenge
In CNS white matter, Kv1.1 and Kv1.2 colocalize in the juxtaparanodal regions of myelinated axons, for example the corpus callosum, brain stem and spinal cord64,69,70. Co-immunoprecipitation reactions also showed that Kv1.1 forms heteromultimeric K+ channels with Kv1.271. Kv1.1 and Kv1.2 also colocalize in juxtaparanodal regions of peripheral myelinated axons72-74.
Defects in the myelination of axons may play an important role in determining the expression patterns of K+ channels on axons. Wang et al70 examined the expression patterns of Kv1.1 and Kv1.2 in the Shiverer mutant mice, which have hypomyelinated CNS and PNS axons. In the brain stem, subcortical and cerebellar white matter and corpus callosum axons of wild-type mice had punctate staining of both Kv1.1 and Kv1.2. However, axons in Shiverer mice had diffuse staining and elevated expression of both Kv1.1 and Kv1.2 protein. mRNA levels of both Kv1.1 and Kv1.2 were elevated in forebrain, hindbrain and cerebellar regions.
In demyelination lesions of the rat sciatic nerve, Rasband et al73. observed a redistribution of Kv1.1 and Kv1.2. In the intact nerve, Kv1.1 and Kv1.2 had paired juxtaparanodal localization on axons . At six to nine days following demyelination, there was a heterogeneous effect on K+ channel distribution. Some axons had diffuse staining at nodes, some axons were devoid of K+ channel staining, but none had paranodal localization of K+ channels. During remyelination (>12 days following demyelination), Kv1.1 expression was found at the node and over time redistributed to the paranodal/juxtaparanodal sites but were never completely contained in their original juxtaparanodal sites.
Studies on injuries to peripheral nerves have also shown changes in K+ channel expression on axons. Nerves from patients with neuromas showed dispersed labelling in the internodes of myelin ensheathed axons75. Ishikawa et al76. reported a decrease in Kv1.1 and Kv1.2 K+ channel proteins in 16 to 24 hr cultured dorsal root ganglion cells 14 days after peripheral axotomy.
Evidence for involvement of Kv1.1 and Kv1 .2 potassium channels in posttrauma ticaxonal dysfunction
With chronic SCI, surviving axons showed a variety of electrophysiological abnormalities including reduced conduction velocity, increased threshold of activation, high frequency conduction failure and an increased refractory period30. These changes were associated with increased sensitivity to 4-AP, which blocks “fast” K+ channels, α-DTX which selectively targets K+ channels of the Shaker gene family, but not TEA or CsCl which block “slow” and inward rectifier K+ channels respectively77.
Based on these observations, we examined the hypothesis that upregulation and/or altered distribution of Kv1.1 and 1.2, which form 4-AP and α-DTX sensitive K+ channels, and have been reported to be on CNS axons, is associated with the electrophysiological abnormalities seen after chronic SCI. Quantitative confocal immunofluorescence microscopy demonstrated increased axonal expression of Kv1.1 and 1.2 (confirmed by quantitative immunoblotting normalized to the axonal protein NF200)77. In addition, there was a markedly altered distribution of these K+ channel subunits after SCI with dispersed labeling along the internodes (in contrast to uninjured axons where labeling was concentrated in the juxtaparanodal regions). Triple-labelled immunofluorescence microscopy using fluorescently tagged (Texas Red) α-DTX showed that α-DTX colocalized with Kv1.1 and Kv1.2 on neurofilament positive axons. This result suggests that it is the increased activity of the α-DTX sensitive “fast” axonal potassium channels, Kv1.1 and Kv1.2, that contributes to axonal dysfunction after chronic SCI. These results strongly support the hypothesis that increased expression and altered distribution of Kv1.1 and 1.2 K+ channel subunits on axons contribute to posttraumatic axonal dysfunction. Therefore, the K+ channel subunits Kv1.1 and Kv1.2 may represent clinically important targets for gene therapy or molecular targeted pharmacological approaches to treat spinal cord injured patients.
- Tator, C. H. & Fehlings, M. G. (1991) J Neurosurg 75, 15.
- Bracken, M. B. et al. (1990) N Engl J Med 322, 1405.
- Bracken, M. B. et al. (1998) J Neurosurg 89, 699.
- Geisler, F., Dorsey, F. & Coleman, W. (1991) New England Journal of Medicine 324, 1829.
- Schumacher, A. , Eubanks, J. & Fehlings, M. G. (1999) Neuroscience 91, 733.
- Schumacher, P. A. , Siman, R. G. & Fehlings, M. G. (2000) J Neurochem 74, 1646.
- Li, S. , Mealing, G. A., Morley, P. & Stys, P. K. (1999) J Neurosci 19, RC16.
- Balentine, J. D. (1978) Laboratory Investigation 39, 254.
- Povlishock, J. T. & Christman, C. W. (1995) J Neurotrauma 12, 555.
- Povlishock, J. T. & Pettus, E. H. (1996) Acta Neurochir Suppl (Wien) 66, 81.
- Agrawal, S. K. & Fehlings, M. G. (1996) J Neurosci 16, 545.
- Agrawal, S. K. & Fehlings, M. G. (1997) J Neurosci 17, 1055.
- Agrawal, S. K. & Fehlings, M. G. (1998) J Neurotrauma 14, 81.
- Agrawal, S. K. , Theriault , E. & Fehlings, M. G. (1998) J Neurotrauma 15, 929.
- Casha, S., Yu, W. & Fehlings, M. (2000) Neuroscience, in press.
- Schumacher, P. A. , Eubanks, J. & Fehlings, M. (1999) Neuroscience 91, 733.
- Li, S. & Stys, P. K. (2000) J Neurosci 20, 1190.
- Kakulas, B. A. (1984) Central Nervous System Trauma 1, 117.
- Blight, A. R. (1983) Neuroscience 10, 521.
- Blight, A. R. & Decrescito, V. (1986) Neuroscience 19, 321.
- Fehlings, M. G. & Nashmi, R. (1995) Brain Res 677, 291.
- Fehlings, M. G. & Nashmi, R. (1997) J Neurosci Methods 71, 215.
- Fehlings, M. G. & Tator, C. H. (1992) Brain Res 579.
- Fehlings, M. G. & Tator, C. H. (1995) Exp Neurol 132, 220.
- Fehlings, M. G. & Nashmi, R. (1995) Brain Research 677, 291.
- Blight, A. R. (1989) Brain Research Bulletin 22, 47.
- Blight, A. R. (1983) Neuroscience 10, 1471.
- Fehlings, M. G. & Nashmi, R. (1996) Brain Research 736, 135.
- Fehlings, M. G. & Nashmi, R. (1997) Journal of Neuroscience Methods 71, 215.
- Nashmi, R. & Fehlings, M. G. (2000) Neuroscience (submitted).
- Chiu, S. Y. & Ritchie, J. M. (1980) Nature 284, 170.
- Kocsis, J. D., Gordon, T. R. & Waxman, S. G. (1986) Brain Res 383, 357.
- Chiu, S. Y. & Ritchie, J. M. (1984) Proc R Soc Lond B Biol Sci 220, 415.
- Poulter, M. O. Hashiguchi, T. & Padjen, A. L. (1989) J Neurophysiol 62, 174.
- Poulter, M. O. & Padjen, A. L. (1995) Neuroscience 68, 497.
- David, G., Barrett , J. N. & Barrett, E. F. (1993) J Physiol (Lond) 472, 177.
- David, G., Barrett , J. N. & Barrett, E. F. (1992) Journal of Physiology 445, 277.
- Chiu, S. Y. & Ritchie, J. M. (1980) Nature 284, 170.
- Blight, A. R. (1989) Brain Res Bull 22, 47.
- Brau, M. E. et al. (1990) J Physiol (Lond) 420, 365.
- Corrette, B. J. et al. (1991) Pf lugers Arch 418, 408.
- Fehlings, M. G. & Nashmi, R. (1996) Brain Res 736, 135.
- Waxman, S. G. (1989) J Neurol Sci 91, 1.
- Kocsis, J. D. (1985) Exp Brain Res 57, 620.
- Blight, A. R. & Gruner, J. A. (1987) J Neurol Sci 82, 145.
- Blight, A. R., et al. (1991) J Neurotrauma 8, 103.
- Hansebout, R. R. et al (1993) J Neurotrauma 10, 1.
- Hayes, K. C. et al. (1993) Paraplegia 31, 216.
- Hayes, K. C. et al. (1994) J Neurotrauma 11, 433.
- Qiao, J. et al (1997) J Neurotrauma 14, 135.
- Albert, J. L. & Nerbonne, J. M. (1995) J Neurophysiol 73, 2163.
- MacFarlane, S. N. & Sontheimer, H. (1997) J Neurosci 17, 7316.
- At tali, B. et al. (1997) J Neurosci 17, 8234.
- Soliven, B. et al. (1988) J Neurosci 8, 2131.
- Sontheimer, H. et al. (1989) Neuron 2, 1135.
- Baumann, A. et al. (1988) Embo J 7, 2457.
- Chandy, K. G. et al. (1990) Science 247, 973.
- Tempel, B. L., Jan, Y. N. & Jan, L. Y. (1988) Nature 332, 837.
- Stuhmer, W. et al. (1989) Embo J 8, 3235.
- Chandy, K. G. et al. (1990) Science 247, 973.
- MacKinnon, R. (1991) Nature 350, 232.
- Liman, E. R., Tytgat, J. & Hess, P. (1992) Neuron 9, 861.
- Hopkins, W. F. et al. (1994). Pflugers Arch 428, 382.
- Wang, H. et al (1993) Nature 365, 75.
- Grissmer, S. et al. (1994) Mol Pharmacol 45, 1227.
- Castle, N. A. et al. (1994) Mol Pharmacol 45, 1242.
- Stephens, G. J. et al (1994) J Physiol (Lond) 477, 187.
- Hart, P. J. et al. (1993) Proc Natl Acad Sci USA 90, 9659.
- Wang, H. et al. (1994) J Neurosci 14, 4588.
- Wang, H. et al. (1995) Neuron 15, 1337.
- Wang, H. et al. (1993) Nature 365, 75.
- Mi, H. et al. (1995) J Neurosci 15, 3761.
- Rasband, M. N. et al. (1998) J Neurosci 18, 36.
- Vabnick, I. et al. (1999) J Neurosci 19, 747.
- England, J. D. et al. (1998) Neurosci Lett 255, 37.
- Ishikawa, K. et al. (1999) Muscle Nerve 22, 502.
- Nashmi, R. , Jones, O. T. & Fehlings, M. G. (2000) Eur J Neurosci 12, 491.