Diabetologia (1993) 36:1230-1237 Diabetologia Springer-Vertag 1993 Originals Angiotensin II receptor blockade improves nerve function, modulates nerve blood flow and stimulates endoneurial angiogenesis in streptozotocin-diabetic ratsand nerve function E. K. Maxfield, N. E. Cameron, M. A. Cotter, K. C. Dines Department of Biomedical Sciences, University of Aberdeen, Scotland, UK Summary. We examined the effect of the angiotensin II re- (p < 0.001). Concentrations of sciatic nerve polyol pathway ceptor blocker, ZD 8731, on nerve function, capillary density, metabolites were elevated six-fold and myo-inositol was re- and blood flow in streptozotocin-diabetic rats. Deficits in sci- duced 40 % by diabetes; ZD 8731 treatment was without ef- atic motor and saphenous sensory nerve conduction velocity fect. Acute experiments examined the effect of ZD 8731 on of 21% and 15 %, respectively, were observed after i month sciatic nerve blood flow using laser-Doppler flowmetry. In of diabetes mellitus (p < 0.001). These were completely ame- non-diabetic rats, blood flow changes followed the dose-de- liorated by a further month of ZD 8731 treatment (p < 0.001). pendent reductions in systemic arterial pressure and there Treatment of non-diabetic rats for 1 month with ZD 8731 had were no significant variations in sciatic vascular resistance. In no effect on motor or sensory conduction velocity. Sciatic marked contrast, nerve blood flow was elevated by 47 % nerve capillary density was not significantly affected by 1- or (p<0.01), and vascular resistance decreased by 32% 2-month untreated diabetes, however, there was a 15 % in- (p < 0.01) in diabetic rats despite similar changes in blood crease in density with ZD 8731 treatment (p < 0.001). Treat- pressure compared with the non-diabetic group. Thus, the in- ment of non-diabetic rats for 1 month had no effect on capil- vestigation has identified abnormalities in vasa nervorum re- lary density. Diabetes prolonged the time taken for 80% activity which are ameliorated by angiotensin II receptor conduction failure by 19 % (p < 0.05) and 49 % (p < 0.001) for blockade and may contribute to experimental diabetic neu- 1 and 2 months of diabetes, respectively, when sciatic nerve ropathy. was exposed to hypoxia in vitro. ZD 8731 treatment during the second month of diabetes limited the prolongation to Key words: Neuropathy, hypoxia, nerve blood flow, nerve 22 %, not significantly different from 1 month of untreated conduction, capillary density, angiotensin II receptor block- diabetes but less than for the 2-month diabetic group ade, diabetic rat. Early changes in nerve function in rats and patients with [7, 17-21]. Such alterations may be relevant to the devel- diabetes mellitus include reduced nerve conduction veloc- opment of early neuropathic changes because treatment ity (NCV) and increased resistance to hypoxic conduc- with the angiotensin-converting enzyme inhibitor, lisino- tion failure (RHCF) [1]. The vascular hypothesis for the pril, prevented the development of motor and sensory aetiology of diabetic neuropathy attributes neurological nerve dysfunction in streptozotocin-diabetic rats [22]. The changes to endoneurial hypoxia caused by a reduction in neurovascular actions of lisinopril are potentially com- nutritive blood flow [1-6]. Vascular changes in diabetes are plex; in addition to blocking angiotensin II synthesis, an- complex and may involve increased reactivity to vasocon- giotensin converting enzyme inhibitors also potentiate strictors [7] reduced endothelial nitric oxide action [8-10], bradykinin-induced vasodilation and reduce the degrada- the formation of advanced glycation end-products [11-13], tion of some other vaso- or neuro-active peptides [23, 24]. deficiencies in essential fatty acid metabolism and prosta- Therefore, to assess whether angiotensin II-mediated cyclin production [14,15], and oxidative stress [16]. vasoconstriction could contribute to neuropathy, the ef- Changes in the renin-angiotensin system could contrib- fects of the angiotensin II receptor antagonist, ZD 8731, ute to the vascular abnormalities of diabetes [7]. While on nerve function was evaluated in streptozotocin- there is some disagreement in the literature, increased diabetic rats. plasma renin activity and elevated levels of angiotensin Vasodilator treatment can cause endoneurial an- converting enzyme and angiotensin II have been reported giogenesis in diabetic rats, and this may be an indirect in- E. K. Max field et al.: Angiotensin II receptor blockade and nerve function [231 Table 1. Body weights and plasma glucose concentrations for non- quinoline which is a selective AT~ angiotensin receptor antago- diabetic and diabetic rats nist [27, 28]. ZD 8731 was dissolved in the drinking water at a con- centration which resulted in rats receiving approximately 50 mg. Body weight (g) Plasma kg -1. day -1. The dose was chosen to give approximately 90 % inhibi- glucose tion of the pressor response to a maximal angiotensin II challenge in (mmolfl) conscious rats [27]. Other groups of rats were given streptozotocin Group n Start End (40-45 mg.kg -1 in 20 mmol.1 ~ sodium citrate buffer, pH 4.5, i.p.). Diabetes was verified 24 h later by estimating hyperglycaemia and Non-diabetic glucosuria (Visidex II and Diastix; Ames, Slough, UK). Samples for Control 23 483 + 10 6.9 + 0.4 plasma glucose measurement were taken from the tail vein on the ZD 8731-treated 8 455+ 5 458-+6 8.6+0.5 day of the final experiments. Diabetic animals were divided into Diabetic three groups, one of which was untreated for 2 months, one was un- 1-month 10 462_+ 8 415 + 12 36.7 + 2.3 treated for 1 month, and the third group was untreated for 1 month 2-month 23 493 _+ 10 371 _+ 8 41.8 + 1.9 followed by i month of treatment with ZD 8731 at the dose used for ZD 8731-treated 13 453+12 388+14 40.0+3.1 non-diabetic rats. In final experiments (1-1.5 g.kg -1 urethane anaesthesia i.p.), Data are group means + SEM NCV was measured in vivo between the sciatic notch and the knee for the motor branches supplying the tibialis anterior (peroneal divi- A B 70 sion) and gastrocnemius (tibial division) muscles. Sensory NCV was T measured in the saphenous nerve between the groin and ankle. Rec- tal and nerve temperatures were monitored and regulated between 6O 36.5 and 37.5 ~ The methods have previously been described in de- E tail [29, 30]. Sciatic nerve hypoxic resistance was measured in vitro as pre- O viously described [14]. The contralateral sciatic trunk was removed >~.~ 50 and mounted on bipolar stimulating (proximal end) and recording (distal end) electrodes in a chamber containing Krebs' solution at 35~ with 5 mmol/1 glucose for nerves from non-diabetic and 40 - 40 mmol/1 glucose for nerves from diabetic rats. Bathing fluid was C CT 1D 2D REV C CT 1D 2D REV gassed with 95 % 02, 5 % CO2. Nerves were equilibrated for 30 min, then the chamber was re-filled with mineral oil pre-gassed for 1 h C 65 - with 100 % N2 and gassing was continued. Nerves were stimulated with just supramaximal pulses (1 Hz, 0.05 ms width, 10 mA) and compound action potential was monitored at 2-rain intervals until it fell below 10 % of its initial value. At the end of the experiment, the rats were killed by exsanguina- tion. Immediately prior to this, approximately 2.5 cm of the sciatic nerve trunk, between the sciatic notch and its bifurcation at the knee, was removed and divided into five pieces which were mounted together, along with skeletal muscle which acted as a support tissue. Samples were frozen in isopentane pre-chilled in liquid nitrogen, 45 - 10-gm sections were cut on a cryostat, and capillary endothelium was C CT 1D 2D REV stained for alkaline phosphatase using the method of Zaida et al. [25]. Three sections were taken, each 90-gm apart, and all capillaries Fig.IA-C. Sciatic motor nerve (A) gastrocnemius and (B) tibialis in the nerve fascicles were counted with the aid of a projection micro- anterior conduction velocity and (C) saphenous sensory nerve con- scope. Fascicle outlines were traced and their areas were measured duction velocity in non-diabetic, untreated diabetic and diabetic rats using a digitizing pad linked to a micro computer. treated with ZD 8731 after 1 month of diabetes. C, non-diabetic Before exsanguination, part of the sciatic nerve proximal and dis- group, n = 23; CT, non-diabetic group treated for 1 month with ZD tal to the sample for capillary measurements, was also taken for 8731 (50 mg. kg -1. day 1), n = 8; 1D, 1-month diabetic group, n = 10; nerve sugar and polyol measurement and frozen in liquid nitrogen. 2 D, 2-month diabetic group, n = 23; REM ZD 8731-treated diabetic Trimethylsilyl derivatives were prepared from aqueous depro- group (50 mg-kg -~. day-t), n = 13. Error bars are SEM teinized extracts and analysed by gas chromatography [31]. In a second investigation, groups of 9 non-diabetic and 13 2- month diabetic rats were used to compare the acute effects of ZD dicator of chronic increases in blood flow [22, 25, 26]. 8731 administration on systemic arterial blood pressure, sciatic Thus, the effect of ZD 8731 on endoneurial capillary den- nerve blood flow and vascular resistance. Rats were anaesthetized sity was also examined. A further study, using laser-Dop- with inactin (50-150 mg-kg -1 i.p.). The trachea was cannulated for artificial ventilation and a carotid cannula was used to monitor mean pler techniques, on the acute effects of ZD 8731 adminis- systemic blood pressure. The jugular vein was cannulated for drug tration on sciatic nerve blood flow was also carried out on administration. The sciatic nerve was exposed between the sciatic separate groups of diabetic and non-diabetic rats.