Decreased capture and retrograde transport of iodinated NGF in the sciatic nerve was observed in diabetic rats many years ago (34). Reduced retrograde transport of iodinated NGF in ileal mesenteric nerves has also been demonstrated (35). These observations imply that even in the absence of any deficit in production of NGF in diabetes, a deficit in the amount delivered to the cell body might be expected. In diabetic rats, there are reduced levels of NGF in the submandibular gland, superior cervical ganglion, and sciatic nerve (3638). NGF levels have also been shown to be decreased in the serum of diabetic patients with symptomatic peripheral neuropathy (39).
Work in our laboratory has shown that with increasing duration of diabetes, progressive reductions in NGF mRNA appear in leg muscle and sciatic nerve followed by reductions in skin. There is a profound reduction in the retrograde transport of NGF in the sciatic nerve, which can be reversed by intensive insulin treatment, and dose-related increases in sciatic nerve NGF retrograde transport were seen with recombinant human NGF (rhNGF) treatment (33). Additionally, it is clear that there are also deficits in the production of NGF target genes, and deficits in expression of SP and CGRP are easily demonstrable in experimental diabetic neuropathy (32).
It is quite apparent that there are deficits in NGF expiation in experimental diabetes, but this does not explain the earlier observation of reduced capture and transport of exogenous NGF. Recent work in our laboratory has revealed a marked decrease in the retrograde transport of the p75NTR, which closely follows the changes seen with NGF transport in diabetes (40). No changes in transport of trkA were observed, but it is not yet possible to suggest which are the precedent changes because NGF availability is known to affect the expression of the p75NTR (41).
NT-3 mRNA levels are reduced in leg muscle from diabetic rats, but an assessment of NT-3 neurotrophic support is difficult because gene targets of NT-3 have yet to be identified. Work in our laboratory showed that treat ment of diabetic rats with rhNT-3 for the last 4 weeks of a 12-week period of diabetes could completely normalize the reduced sensory nerve conduction velocity, which is characteristic of diabetic rats (16). This implies that NT-3 may be even more instrumental than NGF in the development of important functional deficits in diabetic neuropathy. However, although treatment of animals with NGF or NT-3 prevents some of the deficits characteristic of experimental neuropathy (14,16), these agents do not influence reduced nerve blood flow or motor nerve conduction velocity (42), other classic hallmarks of experimental neuropathy.
Shifts in cellular redox balance due to increased levels of free radicals may cause or result from neuronal injury (43,44). Oxygen free radicals are generated as a consequence of ischemia-reperfusion, inflammation, traumatic, and oxidative injury and are associated with neuronal cell death (45,46).
Peripheral nerves, including the sciatic nerve, have inherent low antioxidant defenses compared with the central nervous system, because total reduced glutathione (GSH) content and activities of GSH utilizing enzymes like glutathione peroxidase (GSH-Px) are about 10-fold lower than they are in brain (47).
NGF stimulates cellular resistance to oxidative stress in PC 12 cells (48). In particular, NGF protects from oxidative injury induced by hydrogen peroxide and 6-hydroxydopamine, (48-50) both of which generate hydroxyl radicals. From these observations, it might be expected that NGF regulates cellular oxidant-antioxidant equilibrium.
NGF has been found to regulate the expression of the antioxidant enzymes, catalase (Cat), and GSH-Px. Application of NGF to PC12 cells in culture results in an increase in the transcription of the mRNA for Cat and GSH-Px and, in addition, appears to stabilize the transcript for Cat (51). This is associated with an increase in the activities of both enzymes (52). Furthermore, newborn rat astrocytes in culture synthesise NGF in a dose-dependent fashion in response to superoxide anion, as generated by xanthine/xanthine oxidase and to hydrogen peroxide (53).
Diabetes is associated with increases in oxidative stress in humans and in experimental animal models. Chronic hyperglycemia per se results in autoxi-
dative glycation/oxidation and lipid peroxidation (54-56), and hyperglycemia alone will cause lipid peroxidation of peripheral nerve in vitro (57).
Diabetic peripheral nerve has increased levels of conjugated dienes (end products from peroxidation of polyunsaturated fatty acids) (57,58), decreased levels of GSH (19), and reduced activity of copper/zinc-superoxide dismutase that is reversible with reinstatement of moderate glycemic control (58).
Further support for the role of oxidative stress in the pathogenesis of diabetic neuropathy come from the effectiveness of antioxidant treatment in reversing some of the functional neurological deficits observed in experimental animal models.
Probucol, a powerful free radical scavenger, normalizes both decreases in endoneurial nerve blood flow and motor nerve conduction velocity (59,60). Dietary treatment with 1 % butylated hydroxytoluene or a-tocopherol also has similar effects (61,62). Intravenous administration of GSH can also partially prevent motor nerve conduction velocity slowing in diabetic rats (63). Beneficial effects on conduction velocity have also been reported using the metal chelator deferoxamine (64).
Was this article helpful?