Preserved Glucoregulation but Attenuation of the Vascular Actions of Insulin in Mice Heterozygous for Knockout of the Insulin Receptor

IRKO mice (15,16) were obtained from the Medical Research Council Mammalian Genetics Unit, Harwell, Oxfordshire, U.K. Animals were bred on a C57BL/6J background in a conventional animal facility with a 12-h light/dark cycle and received standard laboratory diet. Male IRKO mice aged 8–12 weeks were compared with age- and sex-matched wild-type littermates. Genotyping was performed using PCR on tail genomic DNA, with primers specific for the gene-targeting cassette.

Intraperitoneal glucose and insulin tolerance tests were performed in conscious, fasted animals (18). Blood glucose was measured at 30-min intervals following intraperitoneal glucose (1 mg/g body wt) or insulin injection (0.75 units/kg; Actrapid, Novo Nordisk, Bagsvaerd, Denmark) using a glucometer (Hemocue, Sheffield, U.K.). Plasma insulin was measured by enzyme-linked immunoassay (CrystalChem, Downers Grove, IL) using rat insulin standards. Fasting plasma free fatty acids and triglycerides were measured by colorimetric assays (Roche, Mannheim, Germany and ThermoTrace, Victoria, Australia, respectively).

Systolic blood pressure was measured using tail cuff plethysmography (XBP 1000; Kent Scientific, Torrington, CT) in conscious, restrained mice at an ambient temperature of 24–26°C (19). Animals were habituated to the restraining apparatus and tail cuff inflation on three occasions before measurements were taken. The mean of six recordings on each occasion was taken, and mean data from three separate recording periods were compared between groups.

Vasomotor function was assessed in ex vivo thoracic aortic rings mounted in organ baths containing Krebs Henseleit buffer (composition [in mmol/l]: NaCl 119, KCl 4.7, KH₂PO₄ 1.18, NaHCO₃ 25, MgSO₄ 1.19, CaCl₂ 2.5, and glucose 11.0) gassed with 95%O₂/5%CO₂ (19–21). Rings were equilibrated at a resting tension of 3 g for 45 min before the experiments.

The cumulative dose response to the constrictor phenylephrine (PE) (1 nmol/l to 10 μmol/l) was first assessed. After washing and re-equilibration, relaxation responses to acetylcholine (1 nmol/l to 10 μmol/l) or sodium nitroprusside (0.1 nmol/l to 1 μmol/l) were assessed in separate rings preconstricted to ∼70% of their maximal PE-induced tension. Relaxation was expressed as the percentage of preconstricted tension. Basal NO bioactivity was assessed in rings maximally constricted with PE by measuring the further increase in tension induced by exposure to the NOS inhibitor N_G-monomethyl-l-arginine (l-NMMA) (0.1 mmol/l) for 30 min.

To assess the effect of insulin on NO production, we compared the cumulative dose response to PE (1 nmol/l to 10 μmol/l) before and after a 2-h incubation with insulin (100 mU/ml Actrapid; Novo Nordisk) (22,23). This dose of insulin was found to be optimal in preliminary experiments. To evaluate the contribution of NO to the effect of insulin, this experiment was also performed in the presence of l-NMMA (0.1mmol/l) (24) or in endothelium-denuded rings.

To investigate the effects of blood pressure per se on vascular responsiveness to insulin, additional experiments were undertaken in 8-week-old male wild-type C57BL/6J mice treated with norepinephrine (4.2 mg · kg^–1 · day^–1 s.c.) or 0.9% saline by osmotic minipump (Alzet Model 1002) for 14 days (25). Systolic blood pressure was measured at days −1, 3, 6, 9, and 12. Blood glucose and plasma insulin were measured after 13 days of infusion, and aortic ring studies were performed after 14 days of infusion.

Relative eNOS mRNA expression in aorta was measured by real-time RT-PCR. Total RNA was extracted (Mini Fibrous Kit; Quiagen), and equal amounts were reverse transcribed using Superscript II reverse transcriptase (Invitrogen) and random decamer oligonucleotides. PCR analysis was performed in duplicate using an ABI Prism 7000 sequence detection system, using primers and a FAM-labeled probe specific for murine eNOS (Assays-on-Demand; Applied Biosystems). β-Actin mRNA expression, assayed with a specific FAM-labeled probe, was used to control for mRNA loading. cDNA was amplified for 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. eNOS mRNA expression for each sample was calculated using the standard curve method (User Bulletin no. 2; Applied Biosystems) and adjusted for β-actin expression.

Fasted mice were anesthetized with sodium pentobarbitone before caval injection of insulin (Actrapid 5 units) or vehicle (0.9% NaCl) (26). After 5 min, the thoracic aorta was excised and snap frozen. Immunoblotting was performed on total aortic tissue homogenates extracted on ice with 0.5% Triton X-100, as previously described (27). Soluble protein concentration was determined using a Bio-Rad kit. Equal amounts of protein (25 μg/sample) were separated on 10% polyacrylamide gels and transferred to polyvinylidine fluoride membranes. Membranes were blocked with 5% nonfat milk/PBS/0.2% Tween 20 before probing with monoclonal antibodies against total murine eNOS (BD Biosciences), phospho-eNOS (Ser¹¹⁷⁷; BD Biosciences), and smooth muscle α-actin (Sigma). Specific bands were detected by enhanced chemiluminescense (Amersham) and quantified by densitometry.

All data are expressed as means ± SE. For isometric tension studies, concentration-response relationships were compared between groups by two-way repeated measures ANOVA. Other variables were compared using Student’s t test. P < 0.05 was considered significant.

IRKO mice were morphologically indistinguishable from their wild-type littermates, with no significant differences in total body, organ, or perigonadal fat depot weight (Table 1).

Blood glucose levels were similar in IRKO and wild-type mice both in the fasting state and after a glucose challenge (Fig. 1A). Glucocompetence, assessed by the response to an intraperitoneal glucose load, was also similar in IRKO and wild-type mice (Fig. 1B). Insulin levels were similar in the fasting state (0.60 ± 0.17 vs. 0.34 ± 0.14 ng/ml for IRKO and wild-type mice, respectively) but were higher in IRKO mice after a carbohydrate challenge (1.08 ± 0.11 vs. 0.62 ± 0.13 ng/ml) (Fig. 1C). The hypoglycemic response to an intraperitoneal insulin load was not significantly different in IRKO and wild-type mice (Fig. 1D). Fasting free fatty acid and triglyceride levels were also similar in the two groups (Table 1).

Systolic blood pressure was significantly higher in IRKO (124 ± 4 mmHg) than wild-type (110 ± 3 mmHg; P = 0.01) mice (Table 1).

Basal NO bioactivity, assessed by the increment in tension induced by l-NMMA in PE-preconstricted rings, was significantly lower in IRKO than wild-type mice (Fig. 2). Relaxation in response to the endothelial-dependent vasodilator acetylcholine was similar in IRKO and wild-type groups (Fig. 3A). Likewise, endothelium-independent relaxation to sodium nitroprusside was no different between groups (Fig. 3B).

Preincubation of wild-type aortic rings with insulin significantly attenuated the contractile response to PE (Fig. 3C), with the maximal contractile response (E_max) reduced to 66 ± 9% of control levels (P < 0.01) but with no change in the half-maximal effective concentration (not shown). This effect of preincubation was completely abrogated in aortic rings from IRKO mice (Fig. 3D). Consistent with the insulin-induced decrease in PE response due to insulin-stimulated NO release, the decrement in PE response was abolished by coincubation of wild-type rings with l-NMMA (Fig. 3E). Moreover, endothelial denudation also abolished this response in wild-type rings (Fig. 3F).

There was no significant difference in aortic eNOS mRNA expression between IRKO and wild-type mice. (The relative expression of eNOS mRNA, normalized to β-actin mRNA, in IRKO mice compared with wild-type controls, was 1.07 ± 0.19 [P = 0.74].)

Expression of aortic eNOS protein, normalized for expression of smooth muscle α-actin, was similar in IRKO and wild-type mice (Figs. 4A and 5). In the unstimulated state, eNOS phosphorylation assessed using an antibody specific for phosphorylation of Ser¹¹⁷⁷ was at the limit of detection in both groups (Fig. 5). However, after insulin stimulation, there was a robust increase in phospho-eNOS in wild-type aorta, which was significantly attenuated in the IRKO group (Figs. 4B and 5).

To exclude the possibility that the vascular abnormalities observed in IRKO mice may be related to increased blood pressure per se, we also studied wild-type mice receiving chronic subcutaneous infusion of norepinephrine. Norepinephrine infusion led to a significant rise in systolic blood pressure within 3 days and maintained for 2 weeks (Fig. 6A), which was of a similar magnitude to that observed in IRKO mice in the resting state (Table 1). Blood glucose and plasma insulin levels were similar in norepinephrine- and vehicle-infused mice (Fig. 6B and C). In aortic ring studies, insulin preincubation attenuated the contractile response to PE to a similar degree in both norepinephrine- and vehicle- infused mice (Fig. 7). Basal NO bioactivity was also similar in both groups of mice. (The increment in tension induced by l-NMMA in PE-preconstricted rings was 175 ± 20% in norepinephrine-infused mice compared with 187 ± 20% in vehicle-infused animals [P = 0.36].)

The principal findings of the present study are that mice heterozygous for knockout of the insulin receptor have substantial abnormalities of vascular function despite very mild metabolic dysfunction. Heterozygous IRKO mice are glucocompetent, with hyperinsulinemia in response to a glucose load as the only detectable metabolic abnormality. Nevertheless, the animals have increased blood pressure, evidence of reduced basal NO bioactivity, and marked blunting of insulin-mediated NO release from aorta. At least some of the vascular abnormality may be related to defective insulin-mediated activation of eNOS through phosphorylation of PKB-sensitive sites (10–13). Thus, even during relatively mild metabolic dysfunction (when glucose homeostasis is maintained by hyperinsulinemia), in this model there is a dramatic reduction in insulin-mediated and basal bioactive NO production in conduit vessels that is associated with an increase in arterial blood pressure. These results may have important implications for our understanding of the pathophysiology of vascular disease and accelerated atherosclerosis in insulin-resistant patients and patients presenting with type 2 diabetes.

Compelling evidence supports endothelial cell dysfunction, in particular a reduction in the bioactivity of NO, as a key early event in atherogenesis (7). NO is generated in endothelial cells by the enzyme eNOS. Classical agonists for eNOS activation, such as acetylcholine, act via G-protein–coupled receptors to raise cytosolic Ca²⁺ and activate eNOS. Recently, however, a complementary pathway for eNOS activation involving its phosphorylation has been described (10–13), which may be of particular relevance to the insulin-resistant state. Agonists that act through tyrosine kinase receptors, such as insulin and IGF-I, induce the activation of PI3K and PKB (Akt), leading to eNOS phosphorylation and enhanced production of NO. Interestingly, shear stress, which may be the major determinant of “basal” NO production in vivo, also activates eNOS at least in part through a similar pathway (28). These data suggest at least one mechanism through which abnormalities of insulin signaling or homeostasis may impact endothelial function.

In the present study, we found that acetylcholine-mediated vasodilatation was not different between IRKO and wild-type animals. However, the IRKO animals had a selective abnormality of basal NO bioactivity and insulin-dependent vasodilatation. These results strongly suggest that there may be a selective abnormality of insulin-dependent pathways for eNOS activation. Indeed, we found that while there was no alteration in eNOS mRNA or protein expression, insulin-stimulated eNOS phosphorylation was significantly reduced in the IRKO group. It is unlikely that the defect seen in the IRKO mice is simply due to reduced receptor density per se. Since the insulin receptor itself has intrinsic activity (29), it is highly plausible that it regulates the downstream signaling cascade described above. This hypothesis is consistent with our finding of reduced basal NO bioactivity, the signaling pathway for which shares a number of components of the downstream insulin-signaling pathway. However, the precise nature of the defects leading to blunted eNOS activation in the present report requires further investigation.

In endothelial cells, insulin can also influence the expression levels of eNOS through the PI3K pathway. For example, two structurally different PI3K inhibitors, wortmannin (100 nmol/l) and LY294002 (50 nmol/l), blunt the effect of physiological concentrations of insulin to increase eNOS mRNA and protein expression (rev. in 13). Consistent with these studies, we have previously reported that in mice overexpressing IGF binding protein-1, which have postglucose hyperinsulinemia but normal whole-body insulin sensitivity, eNOS mRNA expression and basal NO bioactivity are increased (19). In the same model, there was also other evidence of increased NO production in response to hyperinsulinemia, manifested as blunted PE-mediated vasoconstriction and a substantial fall in postprandial blood pressure. An elegant study by Vicent et al. (30) recently showed that targeted knockout of the insulin receptor in endothelial cells resulted in reduction of eNOS expression, although the functional effects of insulin-mediated NO release were not assessed. The present report provides the important finding that in the presence of a global reduction of insulin receptors, mice have preserved glucose homeostasis but substantial disruption of basal- and insulin-mediated vascular NO release, secondary to impaired eNOS phosphorylation and activation rather than reduced expression. These results also suggest that insulin-mediated effects on eNOS transcription may not be as sensitive to impaired signaling as eNOS phosphorylation and activation.

The IRKO mice studied in the present report had hyperinsulinemia in response to a glucose load. It has been suggested that hyperinsulinemia per se leads to endothelial dysfunction (31), although studies in subjects with diabetes treated with insulin do not support this hypothesis (32). Previous work from our laboratory in mice overexpressing IGF binding protein-1 in fact suggests that hyperinsulinemia in the presence of normal insulin signaling may have favorable effects on vasodilator function (19), as discussed above. These findings support a favorable effect of hyperinsulinemia when its signaling pathway is intact.

Hypertension has been shown to be associated with endothelial dysfunction (33) and insulin resistance (34). To explore the possibility that the increased blood pressure demonstrated in the IRKO mice resulted in impaired insulin-mediated NO release, we performed studies in normal mice rendered hypertensive by norepinephrine infusion. However, we found no change in insulin-mediated NO release in the norepinephrine-treated mice, suggesting that the vascular abnormalities found in IRKO mice were unlikely to be a result of increased blood pressure in these animals. The mechanisms responsible for the blood pressure rise in these animals remain to be established, although it is possible that the effect of insulin on vasopressor pathways, such as the sympathetic nervous system (35) or the endothelin system (36), unopposed by NO may at least in part contribute. This warrants further study.

Whether the vascular actions of insulin in experimental studies reflect a physiologic or a pharmacologic effect of insulin is a subject of ongoing debate (37,38). This controversy is not readily addressed by the results of this study, in which vascular function was assessed ex vivo, although the demonstration of increased blood pressure in IRKO mice suggests that the vascular actions of insulin in vivo are of physiological importance.

Accelerated atherosclerosis is the principal cause of death in type 2 diabetes, and it is well appreciated that much of the risk may be related to the long period of compensated insulin resistance before the onset of overt hyperglycemia in these patients. Insulin receptor defects alone are a rare cause of insulin resistance in humans (39). However, defects in PI3K-mediated signaling are detectable in insulin-resistant subjects before the onset of diabetes (40); therefore, studying the influence of impaired insulin signaling on vascular function is important. Using IRKO mice allows the relationship between the metabolic and vascular effects of impaired insulin signaling to be investigated without the influence of hyperglycemia, obesity, or abnormal growth, which confound some models of insulin resistance (41). The findings of the present study indicate that despite adequate metabolic compensation there can be substantial disruption of pathways stimulating the release of the antiatherosclerotic molecule NO in states of mild insulin resistance. These findings support an aggressive and early approach to detecting abnormal vascular function in individuals at high risk for developing type 2 diabetes as well as in those who already have diabetes.

This work was funded by the British Heart Foundation (BHF). S.B.W. and E.D. are BHF Clinical PhD Fellows, M.T.K. a BHF Intermediate Fellow, and A.M.S. holds the BHF Chair of Cardiology at King’s College London. P.A.C. is funded by Diabetes U.K.