Dual Endothelin Receptor Blockade Acutely Improves Insulin Sensitivity in Obese Patients With Insulin Resistance and Coronary Artery Disease

Seven patients (aged 58 ± 2 years, BMI 31.7 ± 2.6 kg/m²) with a previous history of impaired glucose tolerance and coronary artery disease were recruited. Impaired glucose tolerance was defined as fasting blood glucose <6.1 mmol/l and a blood glucose ≥7.8 mmol/l 2 h after an oral glucose loading (75 g). Patients were classified as having diabetes if fasting blood glucose was ≥6.1 mmol/l (on at least on two occasions) or blood glucose concentration was ≥11.1 mmol/l 2 h after an oral glucose loading. Based on these criteria, five patients were classified as having diabetes and two as having impaired glucose tolerance. Coronary artery disease was defined as a history of previous myocardial infarction or significant coronary stenosis verified by coronary angiography. The patients were taking aspirin (n = 7), statins (n = 6), fibrates (n = 1), ACE inhibitors (n = 5), β-blockers (n = 7), and oral antidiabetic agents (n = 5). Average total, LDL, and HDL cholesterol levels were 4.0 ± 0.3, 2.3 ± 0.3, and 0.9 ± 0.1 mmol/l, respectively. Average serum creatinine was 87 ± 5 μmol/l. The subjects were informed of the nature, purpose, and possible risk involved in the study before giving informed consent. The investigation was carried out in accordance with the Declaration of Helsinki and was approved by the ethics committee of the Karolinska Institute.

The study consisted of three different hyperinsulinemic-euglycemic clamp protocols: 1) a control clamp with saline infusion, 2) a clamp with infusion of the ET_A receptor antagonist BQ123, and 3) a clamp with a combined infusion of BQ123 and the ET_B receptor antagonist BQ788. There was at least 1 week between the clamp studies. There were no dropouts. The investigations were performed in random order, and the patients were unaware of the order of the clamps and thus blinded to the treatments. On the study day, the patient arrived in the laboratory after an overnight fast. All medication was withheld on the study day. Two thin catheters were percutaneously inserted into one antecubital vein of each arm for infusions. The receptor antagonists, saline, cardiogreen, and p-aminohippurate (PAH) were given in one arm, while glucose and insulin were infused in the contra-lateral arm. Another catheter was introduced into the brachial artery for sampling of blood and measurement of systemic arterial blood pressure. Heart rate and mean arterial blood pressure (MAP) were continuously monitored and recorded every 20 min throughout the studies. Splanchnic blood flow (SBF) and renal blood flow (RBF) were determined by the constant infusions of cardiogreen, PAH, and the hematocrit, as previously described (12). In a previous study, hepatic and renal vein catheters were introduced to ascertain that fractional uptake, F (equal to the arteriovenous difference divided by the arterial concentration), of cardiogreen and PAH were not influenced by the infusion of the ET-1 blockers (17) or the clamp procedure (14). As there was no change in the F values, a constant F value of 0.8 for cardiogreen and 0.9 for PAH was used in the present study.

Insulin, dissolved in 0.9% saline and blood, was infused at a rate corresponding to 804 mU/m² body surface area during the first 8 min, followed by 40 mU/m² per min for 112 min. Fasting blood glucose level was maintained by adjusting the infusion rate of a 20% glucose solution. Arterial blood samples were taken every 5 min for determination of blood glucose. Arterial samples for plasma insulin, cardiogreen, PAH, and hematocrit were collected at 60, 80, 100, and 120 min of the clamp procedure.

In the control clamp, an infusion of saline was started at 60 min into the clamp and was maintained for 15 min. In the BQ123 clamp, the ET_A receptor antagonist BQ123 was infused at a rate of 5 nmol · kg⁻¹ · min⁻¹. In the combined BQ123+BQ788 clamp, the ET_B receptor antagonist BQ788 was infused at a rate of 4 nmol · kg⁻¹ · min⁻¹ together with BQ123 (5 nmol · kg⁻¹ · min⁻¹). The infusions of the antagonists started 60 min into the clamp and were maintained for 15 min. The doses of BQ123 and BQ788 were based on previous studies demonstrating effective hemodynamic responses and antagonism of vascular effects evoked by ET-1 (17–20).

Glucose was analyzed in whole blood according to the glucose dehydrogenase method by using a HemoCue B glucose photometer (HemoCue, Ängelholm, Sweden) with a precision corresponding to an SD ±0.3 mmol/l. Plasma insulin was analyzed by radioimmunoassay (12). The hematocrit was analyzed with a microcapillary hematocrit centrifuge and corrected for trapped plasma.

BQ123 (Clinalfa, Läufelfingen, Switzerland) and BQ788 (Neosystem, Strasbourg, France) were dissolved in sterile 0.9% NaCl, which was sterile filtered through a Millipore filter and stored frozen at −80°C thereafter. On the day of the experiments, all substances were diluted to the proper concentrations in sterile 0.9% NaCl. Insulin (Actrapid, 100 IE/ml; Novo Nordisk, Bagsværd, Denmark) was dissolved in 0.9% saline and blood.

The total-body glucose uptake (M; mg · kg⁻¹ · min⁻¹) was calculated during three 20-min sampling periods from 60 to 120 min during the clamp (i.e., 60–80, 80–100, and 100–120 min). This means that the first sampling was made during ET receptor blockade on the occasions when antagonists were administered. The M value was then corrected for the mean of the two plasma insulin values obtained during each period to calculate the M/I value, which represents a measure of insulin sensitivity. Splanchnic and renal vascular resistance were calculated as MAP divided by SBF or RBF, respectively, and measured 60 min into the clamp, i.e., before administration of the antagonists (baseline), and thereafter every 20 min. A two-way ANOVA with repeated measures on two factors was used to analyze the data. The factors were clamp (three levels) and time (three to four levels). The interaction in the ANOVA refers to the statistical test of whether the effect of one factor, as measured by differences in the response averages, is different for different levels of the other factor. In the case of significant interaction, post hoc interaction tests were performed between each pair of the clamp conditions across the time intervals. Due to positively skewed distribution, the data for M/I were log transformed. Simple effects tests were also performed, i.e., effects of one factor holding the other factor fixed. For these tests, Fischer’s protected least significant difference was performed. Data are presented as means ± SE.

There was no change in heart rate in either group during the clamps, and there were no differences in heart rate between the clamps. There was no difference in MAP between the groups at 60 min, i.e., before administration of antagonists. However, MAP differed between the three clamps (interaction, clamp × time P < 0.001). When compared with the control clamp, MAP was reduced following administration of BQ123 (clamp × time P < 0.01) and following administration of BQ123+ BQ788 (clamp × time P < 0.05) (Table 1).

RBF at 60 min did not differ between the clamps, and it did not change in the control or BQ123 clamps. In contrast, RBF increased by 24% (P < 0.01) following administration of BQ123+BQ788 (Fig. 1A). RBF was significantly higher in the BQ123+BQ788 clamp compared with the control clamp (P < 0.01, interaction clamp × time P < 0.05) and the BQ123 clamp (P < 0.001). RBF did not differ between the control and the BQ123 clamps. There was no difference in SBF between the three clamps (Fig. 1B).

There were significant differences in renal vascular resistance between the clamp protocols (P < 0.05, clamp × time P < 0.01) (Fig. 1C). Renal vascular resistance was significantly lower in the BQ123+BQ788 clamp than in the control clamp (P < 0.05, clamp × time P < 0.05). Furthermore, renal vascular resistance was lower in the BQ123+BQ788 clamp than in the BQ123 clamp (P < 0.05). There were no significant differences in splanchnic vascular resistance between the clamp protocols (Fig. 1D).

Arterial glucose values remained unchanged and did not differ between the clamps. There were no differences in arterial insulin levels between the three clamp protocols (Table 1). Glucose uptake and insulin sensitivity were determined during saline infusion in the control clamp and following antagonist administration in the BQ clamps. There were significant (P < 0.05) differences in total-body glucose uptake, M values, between the three clamp protocols (Fig. 2A). The M value was significantly higher in the in the BQ123+BQ788 clamp than in the control clamp and in comparison with the BQ123 clamp (P < 0.05). There was no difference in M value between the control and the BQ123 clamps.

There were differences in insulin sensitivity, expressed as M/I values, between the three clamps (P < 0.02). The M/I value was significantly higher in the BQ123+BQ788 clamp than in the control clamp (P < 0.01) and the BQ123 clamp (P < 0.05). The M/I value tended to be higher already at the first measurement during dual receptor blockade with BQ123+BQ788, and it became significantly higher in comparison with the other groups at measurements two and three (Fig. 2B). There was no difference between the control clamp and the BQ123 clamp.

The main finding of the present study is that total-body glucose uptake (M) and insulin sensitivity (M/I) were acutely increased during dual ET_A+ET_B receptor blockade in obese patients with insulin resistance and coronary artery disease. On the other hand, no difference in M or M/I values were observed following selective ET_A receptor blockade. These observations indicate that ET-1 contributes to insulin resistance and that the ET_B receptor plays an important role.

Previous studies (2) have demonstrated that plasma levels of ET-1 are elevated in patients with insulin resistance and/or type 2 diabetes. Furthermore, a negative correlation between total glucose uptake and circulating ET-1 levels was found in diabetic patients (4). In that study, no correction was done for the insulin levels, and the correlation between ET-1 and insulin sensitivity could not be elucidated. Other studies show that administration of exogenous ET-1 reduces insulin sensitivity in healthy humans (13) and in experimental animals (21). However, the role of endogenous ET-1 in the regulation of insulin sensitivity in cardiovascular disease is still incompletely understood. Based on these previous observations, we hypothesized that endogenous ET-1 is involved in the regulation of glucose uptake by insulin in patients with insulin resistance and coronary artery disease. To our knowledge, the present study is the first to demonstrate that dual ET_A+ET_B receptor blockade acutely increased the M and the M/I values in this group of patients, clearly supporting our hypothesis.

The effect on insulin sensitivity was investigated using both selective ET_A and dual ET_A+ET_B receptor blockade. Administration of the ET_B receptor antagonist alone was not tested, since selective ET_B receptor blockade in the absence of ET_A receptor blockade may exert adverse hemodynamic effects in patients with heart failure (18). It is often assumed that most of the pathophysiological effects of ET-1 are mediated via stimulation of the ET_A receptor. In vitro studies have shown that ET-1 inhibits insulin-stimulated glucose uptake in isolated rat adipocytes (22) and isolated skeletal muscle strips (21) mainly through the ET_A receptor. Furthermore, chronic selective ET_A receptor antagonism improves insulin sensitivity and reduces hyperinsulinemia in animal models (23). We have shown that ET_A receptor blockade improves, while selective ET_B receptor blockade impairs, insulin sensitivity during infusion of the ET-1 precursor big ET-1 in healthy humans (14). In the present study, insulin sensitivity improved only after dual ET_A+ET_B receptor blockade but not following selective ET_A receptor blockade, indicating that the ET_B receptor plays a different role in health and disease. It has previously been demonstrated that the expression of smooth muscle cell ET_B receptors that mediate vasoconstriction are increased in atherosclerotic human arteries (15). The expression of ET_B receptors is also increased in mice overexpressing ET-1 (24), suggesting that there may be a change in balance between ET_A and ET_B receptors in states of enhanced expression of ET-1. The increased expression of ET_B receptors has functional consequences, as dual ET_A+ET_B receptor blockade results in more pronounced vasodilatation than selective ET_A receptor blockade in the forearm of patients with atherosclerosis (25), as well as in the renal vascular bed of the present study. Furthermore, dual ET_A+ET_B receptor blockade but not selective ET_A improves endothelium-dependent vasodilatation in clinically healthy subjects with insulin resistance (16). Collectively, these observations indicate that dual ET_A+ET_B receptor blockade may result in more favorable effects than selective ET_A in certain pathophysiological situations. The present data are in accordance with this notion and suggest that dual ET_A+ET_B receptor blockade may be preferred over selective ET_A receptor blockade in order to improve insulin sensitivity in patients with insulin resistance and coronary artery disease.

The changes in hemodynamic parameters support a role for ET-1 in the regulation of vascular tone. Renal vascular resistance but not splanchnic vascular resistance was reduced by dual receptor blockade, suggesting regional differences regarding the influence of ET-1 on vascular tone in this patient group. The more pronounced effect of dual ET_A+ET_B receptor blockade, compared with selective ET_A receptor blockade, on renal hemodynamics is in agreement with the effect on M and M/I values. The finding is also in agreement with observations in the forearm of patients with atherosclerosis (25) and indicates importance of ET_B receptors contributing to vasoconstrictor tone in cardiovascular disease. However, the results contrast those obtained in patients with chronic renal failure and hypertension (19). Selective ET_A receptor blockade, but not dual ET_A+ET_B receptor blockade, increased RBF and reduced renal vascular resistance. Important differences between these studies are that patients in our study had normal renal function, based on serum creatinine levels, and normal blood pressure. Further studies are needed to evaluate the hemodynamic effects of selective ET_A receptor blockade versus dual ET_A+ET_B receptor blockade in patients with cardiovascular disease.

The mechanism behind the effect of ET-1 on insulin sensitivity is not yet fully understood. Previous studies on healthy humans and on isolated skeletal muscle strips suggest that this effect is independent of a reduction in skeletal muscle blood flow (21). Instead, the results imply a direct influence of ET-1 on insulin-stimulated glucose transport. Accordingly, ET-1 decreases the expression of insulin receptor substrate-1 and its activation of phosphatidylinositol 3-kinase pathway and insulin-stimulated Akt phosphorylation in skeletal muscle vascular smooth muscle cells (21,26). Furthermore, a recent study (27) suggests that ET-1 impairs glucose transporter GLUT4 trafficking via interference with phosphatidylinositol 4,5-bisphosphate–regulated cytoskeletal events. These observations suggest that ET receptor blockade may result in increased insulin sensitivity via effects on insulin signaling. However, effects related to increased regional blood flow and bioavailability of insulin by ET receptor blockade cannot be ruled out.

A limitation of the present study is that only the acute effect of ET receptor blockade on insulin resistance was investigated in a limited study group. No information is available regarding the long-term effects of dual ET_A+ET_B receptor blockade on insulin sensitivity. Furthermore, it cannot be ruled out that a small significant effect in insulin sensitivity by selective ET_A receptor blockade may be detected in larger study groups. Based on the present findings, studies using orally available ET receptor antagonists in larger patient groups are warranted to further establish the effect on insulin sensitivity. Nevertheless, the present study was adequately sized to detect a significant effect of dual ET_A+ET_B receptor blockade in comparison with the control group and selective ET_A receptor blockade. Another limitation is that only one dose of each antagonist was tested. Both protocols using antagonists reduced MAP, demonstrating pharmacological effects. Furthermore, comparable doses have previously been demonstrated to exert hemodynamic effects in patients with heart failure (18) and to antagonize the vasoconstrictor effect of exogenous ET-1 in healthy humans (14,17,20), suggesting efficient doses of both antagonists. It is difficult to speculate whether baseline medication may have influenced insulin sensitivity. However, baseline medication was unchanged during the study, suggesting that any influence on insulin sensitivity was similar on all occasions. In addition, administration of an ACE inhibitor does not attenuate the positive effect of ET receptor blockade on endothelial function in patients with atherosclerosis (28).

In conclusion, the present study demonstrates that dual ET_A+ET_B receptor blockade acutely enhances insulin sensitivity in patients with insulin resistance and coronary artery disease. The data support that ET-1 is involved in the development of insulin resistance and that its negative effect is antagonized by dual ET_A+ET_B receptor blockade.

This study was supported by grants from the Swedish Research Council Medicine (10374 and 10857), the Swedish Heart and Lung Foundation, the Actelion Research Award, the King Gustav and Queen Victoria Foundation, the Stockholm County Council, and the Novo Nordisk Foundation.