Spinal intrathecal injection of the peptide Dynorphin A has been shown to reliably induce ischemia, neuronal injury and persistent flaccid hindlimb paralysis.57-63 Concentrations of the excitatory amino acid neurotransmitters Glu and Asp are significantly increased in lumbar cerebrospinal fluid in the rat, shortly after the onset of dynorphin A-induced hindlimb paralysis.62 In addition, a variety of competitive and noncompetitive inhibitors of the NMDA receptor complex have been shown to significantly improve recovery of hindlimb motor function following this insult.62-66 In light of these findings suggesting excitoxic mechanisms in this model, we chose to use it to assess GCP II inhibition as a means to ameliorate excitoxic injury in the rat spinal cord.
Anesthetized male Sprague-Dawley rats (250-300 g) were injected intrathecally with dynorphin A or vehicle, between the lumbar L4-L5 vertebrae according to a published method.62,63 Injections were delivered in a total volume of 22 ^ [containing the dynorphin A (20 nmoles), 2-PMPA (1-4 umoles) or vehicle]. 2-PMPA was provided by Guilford Pharmaceuticals (Baltimore, MD). Locomotor function was independently evaluated by blinded observers at 2 h and again at 24 h following dynorphin A injection, using a 5 point ordinal scale. Neurological scores were assigned as follows: 4 = normal motor function; 3 = mild paraparesis, with the ability to support weight and walk with impairment; 2 = paraparesis, making walking movements without supporting weight; 1 = severe paraparesis, in which rats could make limited hindlimb movement, but not walking movement; 0 = flaccid paralysis, with complete absence of any hindlimb movement.
For amino acid measurements, CSF (approximately 70 |il) was collected on ice for approximately 10 min beginning 15 min after spinal subarachnoid injections of dynorphin A, acidified and frozen until assayed by method of Robinson (1993).67 After euthanasia and formalin infusion, a section of lumbosacral spinal cord corresponding to L5-L6, was sectioned (6 ^m) and stained with cresyl violet. Neurons with a cell dimension > 25 ^m and a visible nucleus were counted by an observer blinded as to treatment condition.
Dynophin A reliably caused flaccid hindlimb paralysis persisting beyond 24 h and significant neuronal injury throughout the lumbosacral cord. When coadministered with dynorphin A, 2-PMPA caused significant dose-dependent improvements in motor scores by 2 and 24 h postinjection. Higher doses of 2-PMPA alone (> 4 ^moles) also caused neurological deficits. On the basis of their assigned neurological scores (0-4), rats were categorized as ambulatory (3 or 4) or nonambulatory (0,1,or 2), at 24 hr postinjection. 80% of the rats co-treated with 4 ^oles of 2-PMPA could walk when tested 24 h postinjury, in contrast to 17% of the rats co-treated with saline vehicle.68
To rule out the possibility that protection against dynorphin A-induced spinal cord injury might result from direct interactions of 2-PMPA and dynorphin A, we also evaluated the effects of 2-PMPA in rats receiving spinal subarachnoid injections of somatostatin, a structurally unrelated peptide that has also been shown to cause hindlimb paralysis, vasospasm and ischemic spinal cord injury in rats.59,61 2-PMPA also significantly improved recovery of hindlimb motor function in rats acutely paralyzed after spinal subarachnoid injection of 25 nmoles of somatostatin.
The protective effects of 2-PMPA were also clearly evident by histopathological assessment of spinal cords removed from rats 72 h after dynorphin A injection. In contrast to the necrosis, hemorrhage, and cellular infiltration typically seen in the spinal cords of dynorphin A-injured rats co-treated with saline vehicle, significant sparing of neurons in both dorsal and ventral horns were observed in rats co-treated with 4 ^oles of 2-PMPA. The spongy rarefaction of white matter that was characteristically seen in dynorphin A-injured cords was also less evident in the cords of rats co-treated with 2-PMPA. Cell counts in the ventrolateral portions of the L5-L6 anterior horn quantitatively revealed the deleterious and salutary effects of dynorphin A and 2-PMPA, respectively. Rats paralyzed by dynorphin A had a significant loss of large neuronal cell bodies in this region, whereas the dynorphin A-injected rats co-treated with PMPA had a dose-related preservation of these cells that in all likelihood are a-motoneurons that are directly involved in locomotion.68,69
In lumbar CSF samples taken following dynorphin A injection, the concentrations of Glu were significantly increased by approximately 3 fold, to 12.3 ^M. When coinjected with dynorphin A, 2-PMPA (which by itself did not alter CSF concentrations of this or other amino acids) the dynorphin A-induced elevations in Glu were significantly reduced.69
As has been previously shown with ischemic injuries to the brain46,47 and with hypoxic or metabolic injuries to neurons in culture,4651 52 2-PMPA exhibited dose-dependent neuroprotective actions against the spinal cord injuries and persistent neurological deficits caused by spinal subarachnoid injection of dynorphin A. Significant motor recovery was evident about 2 h after dynorphin A injection in rats cotreated with 2-PMPA, suggesting that the protective effects of this compound resulted from the interruption of secondary pathophysiological events rather than from a direct interruption of or interference with the initial pharmacological actions of dynorphin A that trigger an ischemic insult to the lumbosacral cord.
2-PMPA eliminated the dynorphin A-induced increases in Glu in CSF samples taken from the lumbar injection site, as would be expected if the hydrolysis of NAAG serves as a primary source of the elevated extracellular Glu seen under these pathophysiological conditions. 2-PMPA did not elicit a change in CSF Glu concentrations in rats not injected with dynorphin A, indicating that the effects of the drug on extracellular Glu were restricted to ameliorating the disrupted tissue milieu associated with injury. The diminished elevations in Glu concentrations might have also resulted from alternative actions of 2-PMPA that served to lessen the severity of the dynorphin A-induced injury and in turn reduce the associated release and extracellular accumulation of Glu. Although csf levels of NAAG were not measured in the present study, in a reported study using the rat MCAo model, reductions in extracellular Glu along with elevations in extracellular NAAG concentrations followed treatment with 2-PMPA,46 as might be predicted with GCP II as the site of drug action.
Beta-NAAG is a non-hydrolyzable analog of NAAG reported to protect cultured neurons against both hypoxia and NMDA-induced injury.70 We have also found that Beta-NAAG was highly protective in vivo against spinal injury induced by Dynorphin-A.71 Although it has been suggested that NAAG may be neuroprotective via agonist effects at mGluR3 receptors, it has been reported that beta-NAAG is an antagonist at mGluR3 receptors.72 Accordingly, the neuroprotective effects of beta-NAAG might be mediated via reduction of synaptic availability of Glu, or direct or indirect antagonism at the NMDA receptor.
As noted, the concentrations of NAAG in the spinal cord are the highest in the
Was this article helpful?