Hyperinsulinemia may contribute to hypertension by increasing sympathetic activity and vascular resistance. We sought to determine if insulin increases central sympathetic neural outflow and vascular resistance in humans. We recorded muscle sympathetic nerve activity (MSNA; microneurography, peroneal nerve), forearm blood flow (plethysmography), heart rate, and blood pressure in 14 normotensive males during 1-h infusions of low (38 mU/m2/min) and high (76 mU/m2/min) doses of insulin while holding blood glucose constant. Plasma insulin rose from 8±1 uU/ml during control, to 72±8 and 144±13 MtU/ml during the low and high insulin doses, respectively, and fell to 15±6 gU/ml 1 h after insulin infusion was stopped.MSNA, which averaged 21.5±1.5 bursts/min in control, increased significantly (P < 0.001) during both the low and high doses of insulin (±5.4 and ±9.3 bursts/min, respectively) and further increased during 1-h recovery (+15.2 bursts/min). Plasma norepinephrine levels (119±19 pg/ml during control) rose during both low (258±25; P < 0.02) and high (285±95; P < 0.01) doses of insulin and recovery (316±23; P < 0.01). Plasma epinephrine levels did not change during insulin infusion. Despite the increased MSNA and plasma norepinephrine, there were significant (P < 0.001) increases in forearm blood flow and decreases in forearm vascular resistance during both doses of insulin. Systolic pressure did not change significantly during infusion of insulin and diastolic pressure fell -4-5 mmHg (P < 0.01). This study suggests that acute increases in plasma insulin within the physiological range elevate sympathetic neural outflow but produce forearm vasodilation and do not elevate arterial pressure in normal humans. (J. Clin. Invest.

To determine whether nitric oxide (NO) synthase activity exists in rat skeletal muscle, media from incubated rat extensor digitorum longus muscle preparations were assayed for NO with a chemiluminescent detection system. Although small amounts of NO were detected in media alone, the addition of muscle increased NO concentration in the media by 30-fold. The release of NO into the media diminished over time. Either arginine (10(-6) M), sodium nitroprusside (10(-6) M), or prior electrical stimulation in vivo caused 50-200% increases (P < 0.05) in NO concentration. NG-monomethyl-L-arginine monoacetate (10(-6) M), an NO synthase inhibitor, decreased both basal 2-deoxyglucose transport and NO efflux, indicating that NO may play a role in modulating skeletal muscle carbohydrate metabolism. These data indicate that NO is released from an incubated skeletal muscle preparation and presents the possibility that muscle-derived NO may play an important metabolic role.

Nitric oxide synthase (NOS) is expressed in skeletal muscle. However, the role of nitric oxide (NO) in glucose transport in this tissue remains unclear. To determine the role of NO in modulating glucose transport, 2-deoxyglucose (2-DG) transport was measured in rat extensor digitorum longus (EDL) muscles that were exposed to either a maximally stimulating concentration of insulin or to an electrical stimulation protocol, in the presence of NG-monomethyl-L-arginine, a NOS inhibitor. In addition, EDL preparations were exposed to sodium nitroprusside (SNP), an NO donor, in the presence of submaximal and maximally stimulating concentrations of insulin. NOS inhibition reduced both basal and exercise-enhanced 2-DG transport but had no effect on insulin-stimulated 2-DG transport. Furthermore, SNP increased 2-DG transport in a dose-responsive manner. The effects of SNP and insulin on 2-DG transport were additive when insulin was present in physiological but not in pharmacological concentrations. Chronic treadmill training increased protein expression of both type I and type III NOS in soleus muscle homogenates. Our results suggest that NO may be a potential mediator of exercise-induced glucose transport.

D-Glucose protectable cytochalasin B (CB) binding to subcellular membrane fractions was used to determine glucose transporter number in red (quadriceps-gastrocnemius-soleus) and white (quadriceps-gastrocnemius) rat muscle. CB binding was twofold higher in isolated plasma membranes of red than of white muscle. In contrast, the number of transporters in an isolated insulin-sensitive intracellular membrane organelle was similar in the two muscle groups. Immunoblotting and immunofluorescence microscopy with anti-GLUT4 and anti-GLUT1 antibodies indicated that both GLUT1 and GLUT4 transporter isoforms account for the higher abundance of CB binding sites in plasma membranes of red than of white muscle. Immunofluorescence localized GLUT4 to both the surface and the interior of the muscle cell and demonstrated that type I (slow twitch oxidative) and type IIa (fast twitch oxidative-glycolytic) fibers are enriched in GLUT4 protein compared with type IIb (fast twitch glycolytic) fibers. In contrast, GLUT1 reactivity was restricted to the surface of the muscle cell and was also highly enriched in the perineurial sheaths of peripheral nerves and the capsules of muscle spindles present in both red and white muscles. Insulin caused a twofold increase in CB binding in isolated plasma membranes of red or white muscles with a corresponding 40-50% decrease in CB binding in isolated intracellular membranes. These changes in CB binding were paralleled by similar changes in the membrane distribution of the GLUT4 glucose transporter isoform and in glucose transport activity measured after insulin perfusion of hindquarter muscles. In contrast, insulin did not change the distribution of either GLUT1 glucose transporters or Na(+)-K(+)-ATPase alpha 1-subunits. The molar ratio of GLUT4 to GLUT1 in red and white muscle plasma membranes was found to be 4:1 in the basal state and 7:1 in the insulin-stimulated state. These results indicate that red muscle contains a higher amount of GLUT1 and GLUT4 transporters at the plasma membrane than white muscle in the basal and insulin-stimulated states but that GLUT4 translocation does not differ between muscle types. In addition, GLUT4 expression correlates with the metabolic nature (oxidative vs. glycolytic) of skeletal muscle fibers, rather than with their contractile properties (slow twitch vs. fast twitch).

It has been suggested that there are separate insulin-stimulated and contraction-stimulated glucose transport pathways in skeletal muscle. This study examined the effects of nitric oxide on glucose transport in rat skeletal muscle by use of an isolated sarcolemmal membrane preparation and the nitric oxide synthase inhibitor N omega-nitro-L-arginine methyl ester (L-NAME), administered in the drinking water (1 mg/ml). Female Sprague-Dawley rats were divided into five groups: control, acute exercise, acute exercise+L-NAME, insulin stimulated, and insulin stimulated+L-NAME. Exercise (45 min of exhaustive treadmill running) increased glucose transport (37 +/- 2 to 76 +/- 5 pmol.mg-1.15 s-1) and this increase was completely inhibited by L-NAME (40 +/- 4 pmol.mg-1.15 s-1). A maximum dose of insulin increased glucose transport (87 +/- 10 pmol.mg-1.15 s-1), and adding L-NAME had no effect (87 +/- 11 pmol.mg-1.15 s-1). In addition, exercise, but not exercise+L-NAME, increased sarcolemma GLUT-4 content. This study confirms that there are separate pathways for contraction- and insulin-stimulated glucose transport. More importantly, although exercise and insulin both significantly increased glucose transport, L-NAME had no effect on insulin-stimulated glucose transport but blocked the exercise-stimulated transport. We conclude that nitric oxide is involved in the signal transduction mechanism to increase glucose transport during exercise.

This study examined the effects of acute exercise on skeletal muscle nitric oxide synthase (NOS) activity. Female Sprague-Dawley rats were divided into three groups: control, exercise, and exercise + N G-nitro-l-arginine methyl ester (l-NAME). In the exercise + l-NAME group, l-NAME was administered in the drinking water (1 mg/ml) for 2 days and subsequently the exercise and exercise + l-NAME groups underwent a 45-min bout of exhaustive treadmill running after which NOS activity and muscle glycogen were measured. In the control and exercise groups, 1-amino- S-methylisothiourea (AMITU), a selective neuronal NOS inhibitor, with and without additional nonselective NOS blockade [with N G-monomethyl-l-arginine (l-NMMA)], was used in vitro to assess the contribution of nNOS to total NOS activity. The exercise bout increased NOS activity by 37% in exercise compared with control groups, and both groups had significantly greater NOS activity compared with exercise + l-NAME. AMITU decreased total NOS activity in the control and exercise groups by 31.8 and 30.2%, respectively, and these activities were significantly greater than AMITU +l-NMMA in both control and exercise groups. We conclude that 1) there is basal neuronal NOS and endothelial NOS activity in skeletal muscle, 2) an acute exercise bout increases NOS activity in skeletal muscle, and 3) glycogen depletion during exercise occurs irrespective of NOS activity.

Insulin stimulates signaling reactions that include insulin receptor autophosphorylation and tyrosine kinase activation, insulin receptor substrate-1 (IRS-1) tyrosine phosphorylation, and phosphatidylinositol 3-kinase (PI 3-kinase) activation. Muscle contraction has metabolic effects similar to insulin, and contraction can increase insulin sensitivity, but little is known about the molecular signals that mediate the effects of contraction. To investigate the effects of muscle contraction on insulin signaling, rats were studied after contraction of hindlimb muscles by electrical stimulation, maximal insulin injection in the absence of contraction, or contraction followed by insulin injection. Insulin increased tyrosine phosphorylation of the insulin receptor and IRS-1, whereas contraction alone had no effect. Contraction before insulin injection decreased the insulin effect on receptor and IRS-1 phosphorylation by 20-25%. Increased tyrosine phosphorylation of other proteins by insulin and/or contraction was not observed. Contraction alone had little effect on PI 3-kinase activity, but contraction markedly blunted the insulin-stimulated activation of IRS-1 and insulin receptor-immunoprecipitable PI 3-kinase. In conclusion, skeletal muscle contractile activity does not result in tyrosine phosphorylation of molecules involved in the initial steps of insulin signaling. Although contractile activity increases insulin sensitivity and responsiveness in skeletal muscle, contraction causes a paradoxical decrease in insulin-stimulated tyrosine phosphorylation and PI 3-kinase activity.

We have previously demonstrated that physiological hyperinsulinemia in normotensive humans increases sympathetic nerve activity but not arterial pressure since it also causes skeletal muscle vasodilation. However, in the presence of insulin resistance and/or hypertension, insulin may cause exaggerated sympathetic activation or impaired vasodilation and thus elevate arterial pressure. This study sought to determine if insulin causes a pressor response in borderline hypertensive humans by producing exaggerated increases in sympathetic neural outflow or impaired vasodilation. We recorded muscle sympathetic nerve activity (microneurography, peroneal nerve), forearm blood flow, heart rate, and blood pressure in 13 borderline hypertensive subjects during a 1-hour insulin infusion (38 microunits/m2/min) while holding blood glucose constant. Plasma insulin rose from 12 +/- 3 microunits/ml (mean +/- SEM) during control to 73 +/- 7 microunits/ml during insulin infusion and fell to 9 +/- 2 microunits/ml 2 hours after insulin infusion was stopped. Muscle sympathetic nerve activity, which averaged 25 +/- 2 bursts per minute in control, increased significantly during insulin infusion (+9 bursts per minute) and remained elevated 1.5 hours into recovery (+7 bursts per minute, p less than 0.001). Despite increased muscle sympathetic nerve activity, there were significant (p less than 0.001) increases in forearm blood flow and decreases in forearm vascular resistance during insulin infusion. Further, systolic and diastolic pressures fell approximately 3 and 6 mm Hg, respectively, during insulin infusion (p less than 0.01). This study suggests that acute physiological increases in plasma insulin elevate sympathetic neural outflow in borderline hypertensive humans but produce vasodilation and do not elevate arterial pressure.