Search

Reduced peripheral nerve perfusion participates in the aetiology of diabetic neuropathy. 5-Hydroxtryptamine causes vasa nervorum vasoconstriction and platelet aggregation, which are enhanced by diabetes. To assess whether these mechanisms could contribute to neuropathy, the effects of 5-hydroxytryptamine 5-HT2 receptor antagonist treatment were examined in streptozotocin-induced diabetic rats. One study determined the dose-response relationship for AT1015 (N-[2-[4-(5H-dibenzo[a,d]cyclohepten-5-ylidene)piperidino]ethyl]-1-formyl-4-piperidinecarboxamide monohydrochloride monohydrate). Two weeks AT1015 treatment after 6 weeks of diabetes dose-dependently corrected 19.7%, 54.1%, and 15.7% deficits in sciatic nerve motor conduction velocity and blood flow, and saphenous nerve sensory conduction: ED50 values were 0.52, 0.74 and 0.15 mg/kg(-1)/day(-1), respectively. In a second study, high-dose AT1015 (3 mg/kg(-1)/day(-1)) actions were compared with those of the 5HT2 receptor antagonists, ritanserin (10 mg/kg(-1)/day(-1)) and sarpogrelate (100 mg/kg(-1)/day(-1)), and the anti-platelet phosphodiesterase III inhibitor, cilostazol (100 mg/kg(-1)/day(-1)). Two weeks treatment with these drugs produced a marked correction (82.6-99.7%) of a 19.8% sciatic motor conduction deficit in diabetic rats. Similarly, 44.7% and 14.9% reductions in sciatic endoneurial blood flow and saphenous sensory conduction velocity were completely reversed. Thus, 5-HT2 receptor antagonists had marked beneficial effects in experimental diabetic neuropathy, and AT1015 appears suitable for further evaluation in clinical trials.

Diabetes mellitus is a major cause of peripheral neuropathy, commonly manifested as distal symmetrical polyneuropathy. This review examines evidence for the importance of vascular factors and their metabolic substrate from human and animal studies. Diabetic neuropathy is associated with risk factors for macrovascular disease and with other microvascular complications such as poor metabolic control, dyslipidaemia, body mass index, smoking, microalbuminuria and retinopathy. Studies in human and animal models have shown reduced nerve perfusion and endoneurial hypoxia. Investigations on biopsy material from patients with mild to severe neuropathy show graded structural changes in nerve microvasculature including basement membrane thickening, pericyte degeneration and endothelial cell hyperplasia. Arterio-venous shunting also contributes to reduced endoneurial perfusion. These vascular changes strongly correlate with clinical defects and nerve pathology. Vasodilator treatment in patients and animals improves nerve function. Early vasa nervorum functional changes are caused by the metabolic insults of diabetes, the balance between vasodilation and vasoconstriction is altered. Vascular endothelium is particularly vulnerable, with deficits in the major endothelial vasodilators, nitric oxide, endothelium-derived hyperpolarising factor and prostacyclin. Hyperglycaemia and dyslipidaemia driven oxidative stress is a major contributor, enhanced by advanced glycation end product formation and polyol pathway activation. These are coupled to protein kinase C activation and omega-6 essential fatty acid dysmetabolism. Together, this complex of interacting metabolic factors accounts for endothelial dysfunction, reduced nerve perfusion and function. Thus, the evidence emphasises the importance of vascular dysfunction, driven by metabolic change, as a cause of diabetic neuropathy, and highlights potential therapeutic approaches.

Aims/hypothesis: Increased oxidative stress has been linked to diabetic neurovascular complications, which are reduced by antioxidants. Our aim was to assess the contribution of hydroxyl radicals to early neuropathic changes by examining the effects of treatment with the specific scavenger, dimethylthiourea, on nerve function and neural tissue blood flow in diabetic rats., Methods: Diabetes was induced by streptozotocin. Measurements comprised sciatic nerve motor and saphenous nerve sensory conduction velocity. Responses to noxious mechanical and thermal stimuli were estimated by Randall-Sellito and Hargreaves tests respectively. Sciatic nerve and superior cervical ganglion blood flow were measured by hydrogen clearance microelectrode polarography., Results: Eight weeks of diabetes reduced motor and sensory conduction velocity by 19.9% and 15.7% respectively, and these were completely corrected by 2 weeks of dimethylthiourea treatment. The ED50 for motor conduction was 9 mg kg(-1) x day(-1). Mechanical and thermal nociceptive sensitivities were 18.9% and 25.0% increased by diabetes, respectively, indicating hyperalgesia which was 70% reduced by dimethylthiourea. Sciatic endoneurial and superior cervical ganglion blood flows were 51.2% and 52.4% reduced by diabetes and there was an approximately 80% improvement with treatment., Conclusion/interpretation: Hydroxyl radicals seem to make a major contribution to neuropathy and vasculopathy in diabetic rats. Treatment with the hydroxyl scavenger, dimethylthiourea, was highly effective. The data suggest that the development of potent hydroxyl radical scavengers suitable for use in man could markedly enhance the potential therapeutic value of an antioxidant approach to the treatment of diabetic neuropathy and vascular disease.

Aims/hypothesis: Increased oxidative stress has been causally linked to diabetic neurovascular complications, which are attenuated by antioxidants. There are several possible sources of reactive oxygen species in diabetes. Our aim was to assess the contribution of free radicals, produced by transition metal catalysed reactions, to early neuropathic changes. To this end, we examined, firstly, the effects of an extracellular high molecular weight chelator, hydroxyethyl starch-deferoxamine, which is expected to be confined to vascular space, on nerve perfusion and conduction deficits in diabetic rats and, secondly, the action of a single chelator dose., Methods: Diabetes was induced by streptozotocin. In vivo measurements comprised sciatic nerve motor conduction velocity and endoneurial perfusion, monitored by hydrogen clearance microelectrode polarography., Results: We found that 8 weeks of diabetes reduced sciatic blood flow and conduction velocity by 48.3 % and 19.9% respectively. Two weeks of intravenous treatment corrected these deficits. Starch vehicle was ineffective. The time-course of action of a single hydroxyethyl starch-deferoxamine injection was examined in diabetic rats. There was a rapid increase in nerve blood flow on day 1, which remained within the non-diabetic range for 9 days before declining to the diabetic level at day 27. In contrast, conduction velocity changes were slower, reaching the non-diabetic range at day 6 and declining to the diabetic level at day 27., Conclusion/interpretation: Extracellular transition metal catalysed reactions play a major role in the neurovascular deficits of experimental diabetes. Given the long-lasting effect of a single treatment, extracellular metal chelator therapy could be suitable for further assessment in clinical trials.

Aims/hypothesis: Increased protein kinase C activity has been linked to diabetic vascular complications in the retina and kidney, which were attenuated by protein kinase C antagonist treatment. Neuropathy has a vascular component, therefore, the aim was to assess whether treatment with WAY151 003 or chelerythrine, inhibitors of protein kinase C regulatory and catalytic domains respectively, could correct nerve blood flow, conduction velocity, Na(+),K(+)-ATPase, and glutathione deficits in diabetic rats., Methods: Diabetes was induced by streptozotocin. Sciatic nerve conduction velocity was measured in vivo and sciatic endoneurial perfusion was monitored by microelectrode polarography and hydrogen clearance. Glutathione content and Na(+),K(+)-ATPase activity were measured in extracts from homogenised sciatic nerves., Results: After 8 weeks of diabetes, sciatic blood flow was 50 % reduced. Two weeks of WAY151 003 (3 or 100 mg/kg) treatment completely corrected this deficit and chelerythrine dose-dependently improved nerve perfusion. The inhibitors dose-dependently corrected a 20 % diabetic motor conduction deficit, however, at high doses ( > 3.0 mg/kg WAY151003; > 0.1 mg/kg chelerythrine) conduction velocity was reduced towards the diabetic level. Sciatic Na(+),K(+)-ATPase activity, 42 % reduced by diabetes, was partially corrected by low but not high dose WAY151 003. In contrast, only a very high dose of chelerythrine partially restored Na(+),K(+)-ATPase activity. A 30 % diabetic deficit in sciatic glutathione content was unchanged by protein kinase C inhibition. The benefits of WAY151 003 on blood flow and conduction velocity were blocked by nitric oxide synthase inhibitor co-treatment., Conclusion/interpretation: Protein kinase C contributes to experimental diabetic neuropathy by a neurovascular mechanism rather than through Na(+),K(+)-ATPase defects.

Diabetes mellitus is associated with impotence in animal models and patients. Raised reactive oxygen species contribute to diabetic neurovascular deficits, which are amenable to antioxidant treatment. Our aim was to examine the effects of streptozotocin-induced diabetes in rats and long-term treatment with the antioxidant, alpha-lipoic acid, on responses of an in vitro corpus cavernosum preparation. Diabetes duration was 8 weeks and preventive and reversal (4 weeks untreated diabetes, 4 weeks of treatment) studies were done. Four and 8 weeks of diabetes caused an about 41% reduction in endothelium-dependent nitric oxide mediated relaxation to acetylcholine in phenylephrine-precontracted cavernosum. This deficit was prevented (93.9+/-7.1%) by treatment with alpha-lipoic acid; reversal studies showed 64.9+/-19.5% correction. Neither diabetes nor treatment with alpha-lipoic acid altered endothelium-independent relaxation to the nitric oxide donor, sodium nitroprusside. Stimulation of corpus cavernosum autonomic innervation caused noradrenergic-mediated contractions that were unaffected by diabetes or alpha-lipoic acid. Non-adrenergic, non-cholinergic nerve responses, largely dependent on nitric oxide, were seen after phenylephrine precontraction in the presence of atropine and guanethidine. Non-adrenergic, non-cholinergic stimulation caused frequency dependent relaxation to a maximum of about 40%. Diabetes reduced this to about 25%, however with preventive alpha-lipoic acid treatment, non-adrenergic, noncholinergic relaxation was in the nondiabetic range. In the reversal alpha-lipoic acid treated diabetic group, its deficit was corrected by 52.1+/-14.6%. Thus, diabetes reduces endothelium and non-adrenergic, noncholinergic nerve nitric oxide-mediated relaxation of corpus cavernosum smooth muscle, which is likely to be the organic base for impotence. Prevention and partial correction by alpha-lipoic acid emphasises the importance of reactive oxygen species and suggests a potential therapeutic approach.

Elevated oxidative stress and impaired n-6 essential fatty acid metabolism contribute to defective nerve conduction velocity (NCV) and perfusion in diabetic rats, which may be corrected by free radical scavenger and gamma-linolenic acid (GLA) treatments. Alpha-lipoic acid (LPA) has antioxidant actions and both LPA racemate (racLPA) and GLA treatments produced benefits in clinical neuropathy trials. The aims were to study LPA action on neurovascular function in diabetic rats and to investigate potential interactions for co-treatment with GLA and other essential fatty acids. After 6 weeks of diabetes, 2 weeks of racLPA treatment corrected 20% sciatic motor and 14% saphenous sensory NCV deficits. The ED50 for motor NCV restoration was approximately 38 mg kg(-1) day(-1). racLPA also corrected a 49% diabetic deficit in sciatic endoneurial blood flow. R and S-LPA enantiomers were equipotent in correcting NCV and blood flow deficits. Treatment of diabetic rats with low doses (20 mg kg(-1) day(-1)) of racLPA and GLA, while having modest effects on their own, showed evidence of marked synergistic action in joint treatment, completely correcting motor NCV and blood flow deficits. This was also noted for the novel compound, SOC0150, which contains equimolar proportions of LPA and GLA (ED50 9.3 mg kg(-1) day(-1), containing 3.5 mg LPA). NCV effects also showed marked synergism when racLPA:GLA ratios were varied over a 1:3-3:1 range. In contrast, a compound containing LPA and the n-3 component, docosahexaenoic acid, showed similar activity to LPA alone. Thus, LPA-GLA interactions yield drug combinations and compounds with an order of magnitude increase in efficacy against experimental diabetic neuropathy and are worthy of consideration for clinical trials.

Aldose reductase inhibitors (ARIs) attenuate diabetic complications in several tissues, including lens, retina, kidney, blood vessels, striated muscle and peripheral nerve. However, it is unclear whether their action in diabetes mellitus depends directly on inhibiting the conversion of glucose to sorbitol by aldose reductase or indirectly by reducing the sorbitol available for subsequent metabolism to fructose by sorbitol dehydrogenase. To identify the polyol pathway step most relevant to complications, particularly neuropathy, we compared the biochemical effects of a sorbitol dehydrogenase inhibitor, WAY-135706, (250 mg.kg-1.day-1) and an ARI, WAY-121509, (10 mg.kg-1.day-1) on a variety of tissues, and their effects on nerve perfusion and conduction velocity. After 6 weeks of untreated streptozotocin diabetes, rats were treated for 2 weeks. Sorbitol was elevated 2.1-32.6-fold by diabetes in lens, retina, kidney, aorta, diaphragm, erythrocytes and sciatic nerve; this was further increased (1.6-8.2-fold) by WAY-135706 whereas WAY-121509 caused a marked reduction. Fructose 1.6-8.0-fold elevated by diabetes in tissues other than diaphragm, was reduced by WAY-135706 and WAY-121509, except in the kidney. Motor and sensory nerve conduction velocities were decreased by 20.2 and 13.9%, respectively with diabetes. These deficits were corrected by WAY-121509, but WAY-135706 was completely ineffective. A 48.6% diabetes-induced deficit in sciatic nutritive endoneurial blood flow was corrected by WAY-121509, but was unaltered by WAY-135706. Thus, despite profound sorbitol dehydrogenase inhibition, WAY-135706 had no beneficial effect on nerve function. The data demonstrate that aldose reductase activity, the first step in the polyol pathway, makes a markedly greater contribution to the aetiology of diabetic neurovascular and neurological dysfunction than does the second step involving sorbitol dehydrogenase.

We investigated whether anti-oxidant treatment with alpha-tocopherol (1 g kg-1 day-1) could prevent the blunting of the normal maturational increase in motor and sensory nerve conduction velocity when diabetes is induced by streptozotocin in young rats. A further study in the same rats examined effects on myelinated fibre regeneration distance 14 days after a punctate sciatic nerve lesion by a liquid nitrogen cooled probe. In non-diabetic rats between 8 and 14 weeks of age, sciatic motor and saphenous sensory conduction velocity increased by approximately 28% (P < 0.001) and 21% (P < 0.001) respectively. Diabetes induced at 8 weeks blunted this maturational change by 65% for sciatic motor and almost completely for saphenous sensory fibres (P < 0.001). Treatment with alpha-tocopherol from diabetes induction totally prevented motor conduction velocity deficits (P < 0.001). For sensory saphenous nerve, conduction abnormalities were markedly attenuated (72%, P < 0.001) although a significant deficit remained compared to age-matched non-diabetic rats (P < 0.01). Sciatic nerve myelinated fibre regeneration distance, 14 days post lesion, was 15% reduced (P < 0.001) by untreated diabetes. However, in diabetic rats treated with alpha-tocopherol, regeneration distance was significantly improved (P < 0.001), being within the non-diabetic range. Thus, the data highlight the importance of reactive oxygen species in the aetiology of impaired nerve maturation and regeneration in experimental diabetes and indirectly support the view that anti-oxidant treatment could have a therapeutic role in patients.

Essential fatty acid metabolism is impaired by diabetes mellitus and gamma-linolenic acid rich treatments such as evening primrose oil correct deficits in nerve conduction and endoneurial blood flow in diabetic rats. Other mechanistically unrelated treatments, such as antioxidants and aldose reductase inhibitors have a similar effect and there may be positive interactions with multiple treatments. Our aim was to compare the efficacy of a novel essential fatty acid derivative, ascorbyl gamma-linolenic acid, with that of gamma-linolenic acid in correcting diabetic neurovascular deficits. Eight weeks of diabetes caused 20.4 and 48.2% reductions in sciatic motor conduction velocity and nutritive endoneurial blood flow, respectively. Treatment was given for the last 2 weeks with gamma-linolenic acid (100 mg.kg-1.day-1) either in pure form or as ascorbyl gamma-linolenic acid, an equivalent dose of ascorbate (21 mg.kg-1.day-1) or jointly with ascorbate and gamma-linolenic acid. Conduction velocity was corrected by 39.8, 87.4, 13.2 and 66.8% with gamma-linolenic acid, ascorbyl gamma-linolenic acid, ascorbate and gamma-linolenic acid plus ascorbate, respectively. Corresponding ameliorations of the nutritive blood flow deficit were 44.0, 87.4, 87.4, 13.2 and 65.7%. For the gamma-linolenic acid plus ascorbate combinatin, and especially for ascorbyl gamma-linolenic acid, the magnitude of correction for conduction velocity and blood flow was greater than expected for simple addition of ascorbate and gamma-linolenic acid, indicating a synergistic interaction. Thus, with an efficacy 40 times that of evening primrose oil in rats, ascorbyl gamma-linolenic acid may be a suitable candidate for clinical trials of diabetic neuropathy.

We examined whether chronic treatment with the free radical scavengers butylated hydroxytoluene (1 g kg-1 day-1) and N-acetyl-L-cysteine (250 mg kg-1 day-1), or the inhibitor of advanced glycosylation reactions, aminoguanidine (1 g kg-1 day-1), could prevent the development of relaxation and contraction abnormalities in aorta from 2 month streptozotocin-diabetic rats. Diabetes caused a 24% deficit in maximal endothelium-dependent relaxation to acetylcholine for phenylephrine precontracted aortas (P < 0.01). This was unaffected by tissue-bath glucose concentration (5.5 or 40 mM), or by addition of 1 mM L-arginine. Butylated hydroxytoluene, N-acetyl-L-cysteine and aminoguanidine treatments gave substantial protection, maximum relaxation remaining in the non-diabetic range. Neither diabetes nor treatment affected endothelium-independent relaxation to glyceryl trinitrate. To test the suggestion that aminoguanidine could act as an inhibitor of constitutive nitric oxide synthase, acute aminoguanidine effects on endothelium-dependent relaxation to acetylcholine were also examined. No inhibition was noted. A modest increase in phenylephrine sensitivity with diabetes (P < 0.05) was unaffected by butylated hydroxytoluene or N-acetyl-L-cysteine, but partially prevented by aminoguanidine (P < 0.05). The data, therefore, provide evidence for the involvement of reactive oxygen species and the advanced glycosylation process particularly for impaired endothelium-dependent relaxation in experimental diabetes.

Impaired omega-6 essential fatty acid metabolism and exaggerated polyol pathway flux contribute to the neurovascular abnormalities in streptozotocin-diabetic rats. The potential interactions between these mechanisms were examined by comparing the effects of threshold doses of aldose reductase inhibitors and evening primrose oil, alone and in combination, on neurovascular deficits. In addition, high-dose aldose reductase inhibitor and evening primrose oil treatment effects were challenged by co-treatment with the cyclo-oxygenase inhibitor, flurbiprofen, or the nitric oxide synthase inhibitor, NG-nitro-L-arginine. Eight weeks of diabetes caused an 18.9% reduction in sciatic motor conduction velocity (p < 0.001). This was only modestly ameliorated by a 0.1% dietary supplement of evening primrose oil or the aldose reductase inhibitors ZD5522 (0.25 mg.kg-1.day-1 and WAY121 509 (0.2 mg.kg-1.day-1 for the final 2 weeks. However, joint treatment with primrose oil and ZD5522 or WAY121 509 caused marked 71.5 and 82.4% corrections, respectively, of the conduction deficit. Sciatic nutritive blood flow was 43.1% reduced by diabetes (p < 0.001) and this was corrected by 67.8% with joint ZD5522 and primrose oil treatment (p < 0.001). High-dose WAY121 509 (10 mg. kg-1.day-1 and primrose oil (10% dietary supplement) prevented sciatic conduction velocity and nutritive blood flow deficits in 1-month diabetic rats (p < 0.001). However, these effects were abolished by flurbiprofen (5 mg.kg(-1).day-1 and NG-nitro-L-arginine (10 mg.kg-1.day-1) co-treatment (p < 0.001). Thus, the data provide evidence for synergistic interactions between polyol pathway/nitric oxide and essential fatty acid/cyclo-oxygenase systems in the control of neurovascular function in diabetic rats, from which a potential therapeutic advantage could be derived.

We examined the ability of the oxygen free radical scavenger, BM 15.0639 (400 mg kg-1 day-1), to correct existing sciatic motor nerve conduction velocity and endoneurial blood flow deficits in streptozotocin-diabetic rats. Rats were treated for 1 month following 1 month of untreated diabetes. Effects of treatment in non-diabetic rats were also examined. A further experiment determined the dose-response relationship for correction of conduction velocity abnormalities by BM15.0639. Diabetes caused 20.9% and 22.7% deficits in motor conduction velocity after 1 and 2 months respectively (both P < 0.001). Rats treated with BM15.0639 after the first month of untreated diabetes had conduction velocity values that were not significantly different from those for non-diabetic controls, but were significantly elevated compared to 1 or 2 months untreated diabetes (both P < 0.001). The ED50 for correction of nerve conduction velocity was approximately 36 mg kg-1 day-1. Sciatic nutritive endoneurial blood flow was 46.5% and 50.5% decreased by 1 and 2 months diabetes respectively (both P < 0.001). This was more than corrected by BM15.0639 treatment of diabetic rats, flow being approximately 33% greater than normal (P < 0.05). In contrast, 1 month BM15.0639 treatment had no effect on blood flow or conduction velocity in non-diabetic rats. Co-treatment of BM15.0639-treated diabetic rats with the nitric oxide (NO) synthase inhibitor, NG-nitro-L-arginine (10 mg kg-1 day-1) largely abolished the anti-oxidant effect on conduction velocity and blood flow. Thus, the data highlight the importance of oxygen free radical activity for the neurovascular deficits in experimental diabetes which are at least in part caused by impaired NO production or release.

We examined the effect in rats of 2 months of streptozotocin-induced diabetes mellitus on relaxation and contraction of aortas in vitro. A further diabetic group was treated from time of diabetes induction with a 1% dietary supplement of vitamin E. Diabetes caused a 26.5% deficit (p < 0.001) in maximum endothelium-dependent relaxation to acetylcholine in phenylephrine-precontracted aortas. This was 64.3% attenuated (p < 0.01) by vitamin E treatment; maximum relaxation was not significantly altered compared to non-diabetic rats. Vitamin E treatment of non-diabetic rats did not significantly affect acetylcholine-induced relaxation. Diabetes or treatment did not significantly alter acetylcholine sensitivity. Endothelium-independent relaxation response to glyceryl trinitrate was not affected by diabetes or vitamin E treatment, indicating that vascular smooth muscle responses to nitric oxide remained unaltered. There was a 35.4% reduction in the maximum contractile response to phenylephrine with diabetes (p < 0.05) which was unaffected by vitamin E treatment. The data suggest that the chronic deficit in nitric oxide-mediated endothelium-dependent relaxation in diabetes depends largely upon excess activity of reactive oxygen species. Treatment with vitamin E to increase free radical scavenging specifically protected vascular endothelium although it had no effect on deficits in vascular smooth muscle contractile responses.

Increased generation of reactive oxygen species, coupled with impaired endogenous scavenging mechanisms, plays a prominent role in the aetiology of neurovascular abnormalities in experimental diabetes mellitus. We examined the efficacy of the natural anti-oxidants vitamins C, E and beta-carotene in preventing nerve conduction and nutritive blood flow deficits in streptozotocin-diabetic rats. One month of diabetes caused a 19.1% reduction in sciatic motor conduction velocity (p < 0.001). This was approximately prevented 80-90% by high-dose (1000 mg.kg-1.day-1) vitamin E and beta-carotene treatments (p < 0.001). Vitamin C had lesser effects; the maximum protection found for motor conduction velocity was 36% using a dose of 150 mg.kg-1.day-1 (p < 0.001). High dose (500 mg.kg-1.day-1 (p < 0.001). High dose (500 mg.kg-1.day-1) vitamin C had a lesser effect on conduction than intermediate doses. Joint vitamin C and lower dose (500 mg.kg-1.day-1) vitamin E treatment had a predominantly additive preventive effect against nerve dysfunction. Resistance to hypoxic conduction failure for sciatic nerve in vitro was markedly increased by diabetes and this remained relatively unaffected by treatment. Sciatic nutritive endoneurial blood flow, measured using microelectrode polarography and hydrogen clearance, was reduced 46.1% by 1 month of diabetes (p < 0.001). This was prevented to the extent of 87%, 36% and 98% by vitamins E, C and beta-carotene, respectively (p < 0.01). These data emphasize the role of oxidative stress in the development of early neurovascular changes in experimental diabetes and show that naturally available scavengers have a neuroprotective action.

The effects of carvedilol (10 mg kg-1 day-1), a peripheral vasodilator and oxygen free radical scavenger, on nerve conduction and blood flow were examined in streptozotocin-diabetic rats. Rats were treated from induction of diabetes for 1 month in a preventive experiment; effects of treatment in non-diabetic rats were also examined. Diabetes caused a 20.0% (P < 0.001) reduction in sciatic motor conduction and a 14.7% (P < 0.001) deficit in saphenous sensory conduction. Diabetic rats given carvedilol treatment had motor and sensory conduction velocities which did not differ significantly from those of non-diabetic controls, but were greater than for untreated diabetes (P < 0.001). Carvedilol treatment did not significantly alter nerve conduction in non-diabetic rats. The nutritive (capillary) component of sciatic endoneurial blood flow, measured by microelectrode polarography and hydrogen clearance, was 45.8% (P < 0.001) reduced by diabetes. This was completely corrected (P < 0.001) by carvedilol treatment in diabetic rats, flow being in the upper half of the non-diabetic range. In non-diabetic rats, carvedilol caused a 30.6% (P < 0.05) elevation in blood flow. Hydrogen clearance data also revealed a shift away from non-nutritive (arterio-venous anastomotic) towards nutritive flow in treated diabetic rats (P < 0.05). We conclude that the data support a neurovascular hypothesis of diabetic neuropathy. Carvedilol, via vasodilator and anti-oxidant actions, prevents the nutritive blood flow changes that lead to nerve dysfunction.(ABSTRACT TRUNCATED AT 250 WORDS)

Evening primrose oil (EPO) is rich in the omega-6 essential fatty acid component, gamma-linolenic acid. The aim of the investigation was to determine whether EPO treatment prevented a reduction in sciatic nerve perfusion and oxygenation in streptozotocin-diabetic rats. Rats were treated from diabetes induction with 10 g EPO kg-1 day-1. Sciatic blood flow was measured by microelectrode polarography and hydrogen clearance. Diabetes caused 47.7% +/- 3.4% (P < 0.001) and 58.8% +/- 4.8% (P < 0.001) reduction in the nutritive (capillary) and the non-nutritive (large vessel) components of endoneurial blood flow, respectively, which were prevented by EPO. Treatment had no significant effect on nutritive flow in non-diabetic rats; however, the rate of non-nutritive flow increased by 97.7% +/- 38.9% (P < 0.01). Sciatic endoneurial oxygen tension was measured by microelectrode polarography. Diabetes resulted in a 44.7% +/- 3.4% reduction in mean oxygen tension (P < 0.001), which was largely (82.3% +/- 10.2%) prevented by EPO treatment (P < 0.001). Thus, EPO prevents impairment of blood flow and endoneurial oxygenation in experimental diabetes. It is likely that this neurovascular action accounts for the beneficial effects of treatment on nerve function in diabetic rats and patients.

Abnormal vascular endothelium function may contribute to the reduced nerve perfusion implicated in the aetiology of neuropathy in diabetes mellitus. The aim was to test the hypothesis that a powerful vasoconstrictor, endothelin-1, could be involved in nerve dysfunction in streptozotocin-diabetic rats. After 6 weeks of untreated diabetes, rats were implanted with osmotic minipumps which continuously delivered the endothelin-1 antagonist, BQ-123, to the circulation via a jugular vein cannula. Sciatic motor conduction velocity, monitored serially, was increased after 4 days, treatment (p = 0.028), and reached asymptote by 9-11 days (p = 0.0001), when the degree of amelioration was approximately 60% of the initial diabetic deficit. Treatment of non-diabetic rats for 13 days with BQ-123 had no significant effect on motor conduction velocity. Sensory saphenous nerve conduction velocity was measured acutely after 20 days, BQ-123 treatment. The amelioration of a sensory deficit was approximately 80% (p < 0.001); the resultant conduction velocity value was not significantly different from that of a non-diabetic control group. After 20 days, treatment, sciatic nutritive endoneurial blood flow was measured by microelectrode polarography and hydrogen clearance. A 48% deficit with untreated diabetes (p < 0.001) was 64% ameliorated by BQ-123 treatment (p < 0.001). In non-diabetic rats, BQ-123 treatment had no effect on blood flow. We conclude that endothelin-1 does not seem to be involved in the control of nerve blood flow in non-diabetic rats; however, it makes a major contribution to the perfusion deficit in experimental diabetes. This has deleterious consequences for nerve conduction, and it is possible that endothelin-1 receptor blockade may have therapeutic potential in diabetic patients.

We examined the effects of aldose reductase inhibition on nerve biochemistry and function, blood flow and endoneurial oxygenation in experimental diabetes mellitus. After 1 month untreated diabetes in rats, treatment with the novel sulphonylnitromethane aldose reductase inhibitor, ZENECA ZD5522, prevented a progressive increase in sciatic nerve resistance to hypoxic conduction failure (p < 0.05). Motor conduction velocity deficits after 4 months untreated diabetes were rapidly returned to normal within 12 days (p < 0.0001) by ZD5522 treatment. Following 2-months untreated diabetes, examination of 1 month ZD5522 treatment dose-response relationships for correction of nerve sorbitol and fructose accumulations and reduction in myo-inositol concentration, sciatic motor and saphenous sensory conduction velocity and sciatic blood flow by laser-Doppler flowmetry revealed poor agreement between nerve function and biochemical indices. In addition, polyol accumulation differed between sciatic and saphenous nerves, the latter showing ten-fold lower sorbitol concentrations. Laser-Doppler blood flow was 60% decreased by untreated diabetes (p < 0.001) and there was a strong correlation between ZD5522-mediated increases in blood flow and conduction velocity (p < 0.0001). Measurement of nutritive endoneurial blood flow by microelectrode polarography and hydrogen clearance showed 44% and 45% deficits for 1 and 2 months untreated diabetes (p < 0.001) that were prevented by ponalrestat and ZD5522 treatments, respectively. In contrast, 2 months myo-inositol treatment from diabetes induction did not prevent reduction in blood flow or sciatic motor conduction velocity. A 37% reduction in endoneurial oxygen tension after 2 months diabetes (p < 0.001) was completely prevented by ZD5522 treatment (p < 0.001). The data show that a very high degree of polyol pathway blockade is necessary to correct nerve functional deficits and that aldose reductase inhibitors have a neurovascular action that does not depend on restoration of nerve myo-inositol.

Increased oxygen free radical activity, coupled with reduced protection against oxidative stress, could play a role in the aetiology of neurovascular abnormalities in experimental diabetes mellitus. To test this hypothesis, non-diabetic and streptozotocin-diabetic rats were treated with the anti-oxidant probucol or the pro-oxidant primaquine. One-month diabetes caused 21.4% and 13.6% reduction in sciatic motor and saphenous sensory conduction velocity (p < 0.001). These deficits were prevented by probucol treatment (p < 0.001). After 1-month untreated diabetes, conduction velocity deficits were reversed by a further month of probucol treatment (p < 0.001). For non-diabetic rats, primaquine treatment caused a 12.9% reduction in motor conduction velocity (p < 0.001), which was prevented by probucol treatment (p < 0.001). Primaquine treatment did not affect diabetic rats. Sciatic nerve nutritive endoneurial blood flow, measured using microelectrode polarography and hydrogen clearance, was 48.0% reduced by 2-month diabetes (p < 0.001). This was completely prevented by probucol treatment (p < 0.001). Primaquine treatment did not affect blood flow in diabetic rats. However, in non-diabetic rats it caused a 30.0% reduction (p < 0.01) which was prevented by probucol treatment (p < 0.05). Sciatic endoneurial oxygen tensions were also measured by microelectrode polarography. Mean tension was 38.8% reduced by diabetes (p < 0.001). This was prevented by probucol treatment. Non-diabetic rats given primaquine treatment showed a 21.7% reduction in endoneurial oxygen tension (p < 0.01). The data suggest that vascular-mediated nerve dysfunction in diabetes depends on oxidative stress, and that similar effects in non-diabetic rats may be produced by pro-oxidant treatment. This provides evidence for the potentially important role of oxygen free radical activity in diabetic neuropathy.

We examined the effect of the angiotensin II receptor blocker, ZD 8731, on nerve function, capillary density, and blood flow in streptozotocin-diabetic rats. Deficits in sciatic motor and saphenous sensory nerve conduction velocity of 21% and 15%, respectively, were observed after 1 month of diabetes mellitus (p < 0.001). These were completely ameliorated by a further month of ZD 8731 treatment (p < 0.001). Treatment of non-diabetic rats for 1 month with ZD 8731 had no effect on motor or sensory conduction velocity. Sciatic nerve capillary density was not significantly affected by 1- or 2-month untreated diabetes, however, there was a 15% increase in density with ZD 8731 treatment (p < 0.001). Treatment of non-diabetic rats for 1 month had no effect on capillary density. Diabetes prolonged the time taken for 80% conduction failure by 19% (p < 0.05) and 49% (p < 0.001) for 1 and 2 months of diabetes, respectively, when sciatic nerve was exposed to hypoxia in vitro. ZD 8731 treatment during the second month of diabetes limited the prolongation to 22%, not significantly different from 1 month of untreated diabetes but less than for the 2-month diabetic group (p < 0.001). Concentrations of sciatic nerve polyol pathway metabolites were elevated six-fold and myo-inositol was reduced 40% by diabetes; ZD 8731 treatment was without effect. Acute experiments examined the effect of ZD 8731 on sciatic nerve blood flow using laser-Doppler flowmetry. In non-diabetic rats, blood flow changes followed the dose-dependent reductions in systemic arterial pressure and there were no significant variations in sciatic vascular resistance.(ABSTRACT TRUNCATED AT 250 WORDS)

Essential fatty acid metabolism is impaired by diabetes mellitus and this may be important in the aetiology of peripheral nerve dysfunction. The effects of gamma-linolenic acid (omega-6) and fish oil (omega-3) alone, and in combination, on nerve function and capillarization were examined in 2-month streptozotocin-diabetic rats. Diabetes resulted in approximately 15% and 23% decreases in saphenous sensory and sciatic motor nerve conduction velocities, respectively (p < 0.001). Motor and sensory conduction velocities were in the non-diabetic range after both preventive and reversal omega-6 treatment of diabetic rats (p < 0.001). No significant changes occurred in omega-6 treated non-diabetic rats. Preventive omega-3 treatment was largely ineffective. Reversal treatment with a combination of omega-6 and omega-3 fatty acids was marginally effective and improved motor (p < 0.05), but not sensory conduction velocity. In vitro measurement of sciatic nerve resistance to hypoxic conduction failure in diabetic rats revealed a 56% increase in the time taken for the compound action potential amplitude to be reduced by 80% (p < 0.01) compared to non-diabetic rats. This was partially prevented by omega-6 treatment (29% increase, p < 0.01). Reversal omega-6 treatment had a lesser effect (37% increase, p < 0.05 compared to untreated diabetic rats). omega-3 treatment had no significant effect on conduction failure time. Sciatic endoneurial capillary density increased by 11% with preventive omega-6 treatment (p < 0.05), but was unaffected by reversal omega-6 and by omega-3 treatments.(ABSTRACT TRUNCATED AT 250 WORDS)

We examined the potential for some of the abnormalities of vascular endothelium found in diabetes mellitus to cause neuropathic changes. Non-diabetic rats were treated for 2 months with the cyclo-oxygenase inhibitor flurbiprofen (5 mg.kg-1.day-1) to reduce prostacyclin production, the nitric oxide synthase inhibitor NG-nitro-L-arginine (5 or 25 mg.kg-1.day-1), or combined treatment. There were dose-dependent reductions in sciatic motor and saphenous sensory conduction velocity. The two inhibitors acted synergistically, thus, the 5-6% motor conduction deficits (p < 0.01) produced by either flurbiprofen or NG-nitro-L-arginine (5 mg.kg-1.day-1) increased to 17% (p < 0.001) for combined treatment. With NG-nitro-L-arginine (25 mg.kg-1.day-1) and flurbiprofen, motor and sensory conduction velocity were reduced by 23% (p < 0.001) and 12% (p < 0.001), respectively, matching the deficits following 2-month streptozotocin diabetes. NG-nitro-L-arginine (25 mg.kg-1.day-1) and flurbiprofen together produced a 13% prolongation of the time taken for 80% hypoxic conduction failure in vitro (p < 0.05) and a 10% reduction in sciatic capillary density. A second investigation tested an alternative hypothesis that overproduction of nitric oxide was responsible for vascular-related complications in diabetes, the prediction being that NG-nitro-L-arginine (5 mg.kg-1.day-1) would prevent nerve dysfunction. However, rather than prophylaxis during 2-month streptozotocin diabetes, treatment exacerbated nerve abnormalities.(ABSTRACT TRUNCATED AT 250 WORDS)

The effects of the prostacyclin analogue iloprost on nerve function were examined in streptozotocin-diabetic rats. Rats were treated either with iloprost from induction of diabetes over 2 months in a preventive experiment, or for 1 month following a 1 month untreated period of diabetes in a reversal experiment. One and 2 months untreated diabetic control, non-diabetic control, and iloprost-treated non-diabetic groups were also used. Diabetes of 1 month duration caused a 21% (P < 0.001) reduction in sciatic motor conduction velocity and a 14% (P < 0.001) deficit in saphenous sensory conduction. This was not significantly changed over a subsequent month without treatment. Diabetic rats given iloprost treatment in both preventive and reversal studies had motor and sensory conduction velocities not significantly different from those of non-diabetic controls, but greater than for untreated diabetes (P < 0.01). Iloprost treatment did not have a significant effect on nerve conduction in non-diabetic rats. The time taken for sciatic nerve compound action potential amplitude to be reduced by 80% under hypoxic conditions in vitro was progressively elevated by 19% and 57% after 1 and 2 months diabetes respectively. Iloprost treatment significantly attenuated this for both preventive (47%, P < 0.001) and reversal (50%, P < 0.001) studies. There was no effect on hypoxic resistance for non-diabetic rats. In the preventive group there was a 28% increase in sciatic nerve endoneurial capillary density (P < 0.001), a lesser effect (16%, P < 0.05) in the reversal group, and no effect in non-diabetic rats.(ABSTRACT TRUNCATED AT 250 WORDS)

We tested the notion that oxidative stress makes an important contribution to the aetiology of diabetic neuropathy. The effect of treatment with a 1% dietary supplement of the anti-oxidant butylated hydroxytoluene was studied during 2 months of streptozotocin-induced diabetes mellitus. In final experiments, sciatic motor and saphenous sensory conduction velocities were measured in vivo, and resistance to hypoxic conduction failure for sciatic trunk was examined in vitro. There were 20% and 12% decreases in motor and sensory conduction velocity, respectively after 2 months of diabetes (p < 0.001). There were completely prevented by butylated hydroxytoluene treatment (p < 0.001). Resistance to hypoxic conduction failure, shown by the time taken for sciatic compound action potential amplitude to decline by 80%, was 55% increased by diabetes, and this was limited to 31% (p < 0.01) by treatment. There were no significant effects of treatment on the 9-10 fold elevation of sciatic nerve sorbitol and fructose levels with diabetes, or on the non-significant 22% reduction in myoinositol content. Butylated hydroxytoluene treatment also did not affect sciatic nerve capillary density. We conclude that oxidative stress makes an important contribution to the aetiology of early experimental diabetic neuropathy. Amelioration of oxidative stress could potentially be a final common mechanism whereby a number of diverse treatments exert a beneficial effect on diabetic nerve function.

The effect of treatment of rats with the angiotensin converting enzyme inhibitor lisinopril after 5 weeks of untreated streptozotocin-diabetes was examined by daily monitoring of sciatic motor conduction velocity to tibialis anterior muscle. Diabetes produced a 31.5% decrease in conduction velocity (P < 0.001). Lisinopril treatment caused a progressive improvement which was significant after 3 days (P = 0.002), full normalization being achieved by 6 days (P < 0.0001). After 7 days of treatment there followed a 7-day washout period in which no lisinopril was given. During this time conduction velocity declined to untreated diabetic levels over 3 days. A subsequent treatment period resulted in complete normalization of conduction velocity within 2 days (P < 0.0001). Thus, the marked functional effects seen for vasodilator treatment with lisinopril suggest that angiotension converting enzyme inhibitors may have potential therapeutic value in the treatment of diabetic neuropathy.

Recent data suggests that reduced nerve blood flow is implicated in the aetiology of experimental diabetic neuropathy, which may be prevented by manipulations that reduce receptor-mediated vasoconstrictor activity. This investigation examines the effects of nifedipine, a voltage-sensitive calcium channel antagonist which has a direct vasodilatory effect on vessels, on nerve conduction, hypoxic resistance and capillary density in streptozotocin-induced diabetic rats. Treated and non-treated non-diabetic and diabetic groups were employed. Diabetes duration was 2 months. Treatment was preventive, groups received a nifedipine dietary supplement (40 mg.kg-1.day-1) for 2 months from the start of the study. Conduction was measured in sciatic motor branches supplying tibialis anterior and gastrocnemius muscles, and sensory saphenous nerve. Diabetes resulted in a 23-28% reduction in motor conduction velocity (p < 0.001), and a 15% deficit for sensory saphenous nerve (p < 0.001). In the nifedipine-treated diabetic group, motor and sensory conduction deficits were minimal compared with non-treated diabetes (p < 0.001). Nifedipine treatment had no significant effect on conduction velocity in non-diabetic rats. In vitro measurement of sciatic nerve hypoxic resistance revealed a 60% increase in the time taken for compound action potential amplitude to reach half its initial value with diabetes (p < 0.001). This was not significantly affected by nifedipine treatment. Experimental diabetes or nifedipine treatment did not significantly alter sciatic nerve endoneurial capillary density. We conclude that nifedipine, a vasodilator which acts directly on vascular smooth muscle, prevents nerve conduction deficits in experimental diabetes.

The effects of 3 months streptozotocin-induced diabetes mellitus on contraction and relaxation of aorta were examined in vitro. A further diabetic group was treated with a novel sulphonylnitromethane-based aldose reductase inhibitor for 3 months following diabetes induction. Diabetes resulted in reduced maximal tension production, particularly for responses to phenylephrine (p < 0.001) and serotonin (p < 0.001). However, with aldose reductase inhibitor treatment, responses were in the non-diabetic range. The ratio of maximum contractions to noradrenaline and phenylephrine were 28% elevated by diabetes (p < 0.01), which may suggest increased alpha 2-adrenoreceptor-mediated responses. Endothelium-independent relaxation to glyceryl trinitrate was unaffected by diabetes or treatment. By contrast, there were 38% deficits in endothelium-dependent relaxation to acetylcholine (p < 0.001) and Ca2+ ionophore A23187 (p < 0.001) with diabetes which were prevented by aldose reductase inhibitor treatment (p < 0.001). A 121% shift in the concentration giving a 50% maximum effect for acetylcholine towards lower sensitivity with diabetes (p < 0.001) was also largely corrected by treatment (p < 0.001). A non-diabetic group treated with aldose reductase inhibitor showed a 30% decrease in the 50% effective concentration for acetylcholine (p < 0.05). A 15% deficit in maximum relaxation to the ATP-sensitive K+ channel opener cromakalim for the diabetic group (p < 0.001) was prevented by aldose reductase inhibitor treatment (p < 0.01). We conclude that there are polyol pathway related abnormalities for contraction, some aspects of endothelium-independent relaxation, but particularly for endothelium-dependent relaxation in aorta from chronic streptozotocin-diabetic rats.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of 2 months aminoguanidine treatment on nerve conduction abnormalities was studied in streptozotocin-diabetic rats. Treatment with aminoguanidine from the induction of diabetes mellitus prevented a 22% decrease in sciatic motor nerve conduction velocity (p less than 0.001), and a 10% deficit in sensory saphenous conduction velocity (p less than 0.01). There was a 49% increase in resistance of sciatic nerve to hypoxic conduction failure in vitro. This was not significantly affected by aminoguanidine treatment. Sciatic nerve polyol pathway metabolites, sorbitol and fructose, were elevated 10-fold by diabetes (p less than 0.001). Myo-inositol levels were 18% decreased by diabetes. Aminoguanidine treatment had no significant effect on sorbitol, fructose or myo-inositol levels. Aminoguanidine has been identified as both an inhibitor of the formation of advanced glycation end products, and an aldose reductase inhibitor. The data suggest that beneficial actions on nerve function do not depend on the latter property. They support the notion that advanced glycation end products contribute to the aetiology of early diabetic neuropathy, possibly acting via a vascular mechanism, and that aminoguanidine treatment may have therapeutic applications.

The effects of the angiotensin converting enzyme inhibitor lisinopril on slow and fast twitch muscle contractile properties, nerve conduction and hypoxic resistance, and muscle and nerve capillary density were examined in streptozotocin-diabetic rats. Prolongation of soleus contraction and relaxation were partially prevented by treatment (p less than 0.01). A 22% deficit in fast twitch extensor digitorum longus tetanic tension production was also ameliorated (p less than 0.01). Sciatic motor and sensory conduction velocity, 25% and 12% reduced by diabetes respectively, were 75% normalized by lisinopril (p less than 0.01). There was a 47% increase in resistance to hypoxic conduction block with diabetes (p less than 0.01). Lisinopril treatment resulted in normal hypoxic resistance. Capillarization of nerve and muscle was little affected by diabetes; however, there was a 17% increase in capillary density in sciatic nerve, and a 40% increase in extensor digitorum longus muscle with lisinopril (p less than 0.01). For soleus, a smaller treatment-induced increase in capillary density led to an elevated capillary/muscle fibre ratio (p less than 0.01). These results suggest that lisinopril promoted angiogenesis. It was concluded that the beneficial effect of preventive lisinopril treatment is likely to depend upon a reduction of peripheral vascular resistance and improvement of tissue blood flow, which implicates relative hypoxia as an important factor in the development of myopathy and neuropathy in experimental diabetes.

This study examined the effect of chronic electrical activation on conduction velocity deficits after three months of streptozotocin-induced diabetes. There were 30% and 20% reductions in conduction velocity in diabetic animals for tibialis anterior and saphenous nerves, respectively (p less than 0.01). Unilateral electrical stimulation of the common peroneal nerve, which contains axons supplying tibialis anterior but not saphenous nerve, was carried out in a group of diabetic and a group of normal control rats. Stimulation was given over seven days, at 10 Hz for 8 h/day. Final experiments were carried out at least 17 h after the last stimulation session. In normal rats stimulation had no effect on conduction velocity in either nerve. In diabetic animals, however, tibialis anterior conduction was within the normal control range for the stimulated nerve. In contrast, the contralateral unstimulated nerve had reduced conduction velocity (p less than 0.001), which was within the unoperated diabetic control range. There were no effects on saphenous nerve conduction, comparing stimulated and unstimulated legs. We conclude that chronic increases in nerve electrical activation promote mechanisms that reverse conduction deficits in diabetic rats.

This study examined the preventative effect of an aldose reductase inhibitor, ponalrestat, on contractile properties of heart left ventricular papillary muscles in rats having streptozotocin induced diabetes for 13 weeks. Both contraction and relaxation were slowed by diabetes. The time to reach peak twitch tension was increased by 21%, and the time to relax to half peak tension was increased by 29% (p less than 0.01, respectively compared to normal control animals). With ponalrestat treatment, the increase in contraction time was only 11%, and relaxation was only slowed by 4% (p less than 0.05 and p less than 0.01, respectively compared to diabetic controls). Diabetes also reduced maximum rates of contraction (13%) and relaxation (19%) and prolonged the time taken to reach peak relaxation rate (36%, p less than 0.01). Ponalrestat had no effect on maximum contraction rates but was particularly effective in normalising relaxation rates (p less than 0.01). Deficiencies with diabetes were noted over a range of stimulation frequencies (0.1-4.0 Hz), and ponalrestat treatment was beneficial except at the highest rates. Diabetes and ponalrestat effects were observed over a temperature range of 25-37 degrees C. Ventricular sorbitol levels showed a 17-fold increase with diabetes (p less than 0.01) and this was reduced by 67% with ponalrestat (p less than 0.01). There were no changes in ventricular myo-inositol. It is possible that ponalrestat treatment prevented a defect in sarcoplasmic reticulum calcium handling which could be responsible in part for the deficits in contraction ability and mainly for the deficits in relaxation ability in diabetic cardiomyopathy, although this remains to be tested directly.

This study examined the effects of an aldose reductase inhibitor, Sorbinil, on neuropathy over a 6-month period in streptozotocin-diabetic rats. Sorbinil treatment prevented the 10-fold increase in nerve sorbitol found with diabetes. It produced a 60% improvement in tibial nerve motor conduction velocity after 6 months. Morphometric profiles of nerves were also normalized. Axon area was reduced by 14% in untreated diabetic rats compared to age-matched controls, whereas Sorbinil-treated animals showed normal age-related axon growth. Myelin area was increased by 28% in untreated diabetic animals, but was the same as age-matched controls with Sorbinil treatment. Nerve myo-inositol levels were reduced by 45% after three months of untreated diabetes, but were normal after six months. Sorbinil treatment tended to restore myo-inositol levels toward normal over the shorter time period. It was concluded that axon growth retardation is the most likely cause of the conduction deficit seen in long-term experimental diabetes.