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

Emerging Treatments and Technologies ORIGINAL ARTICLE

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

1 2 GUNVOR AHLBORG, MD, PHD ADRIAN GONON, MD, PHD The vascular responses to ET-1 are 2 2 ALEXEY SHEMYAKIN, MD JOHN PERNOW, MD, PHD mediated via two receptor subtypes: ET 2 A FELIX BÖHM, MD, PHD and ETB receptors (10,11). Both types of receptors are located on vascular smooth muscle cells and mediate vasoconstric- OBJECTIVE — Endothelin (ET)-1 is a vasoconstrictor and proinflammatory peptide that may tion. The ETB receptor is also located on inhibit glucose uptake. The objective of the study was to investigate if ET (selective ETA and dual endothelial cells and mediates vasodilata- ϩ ETA ETB) receptor blockade improves insulin sensitivity in patients with insulin resistance and tion by stimulating release of NO and coronary artery disease. prostacyclin. Early reports show that Ϯ ET-1 interferes with glucose metabolism RESEARCH DESIGN AND METHODS — Seven patients (aged 58 2 years) with as indicated by a drop in splanchnic glu- insulin resistance and coronary artery disease completed three hyperinsulinemic-euglycemic clamp protocols: a control clamp (saline infusion), during ET receptor blockade (BQ123), and cose production and peripheral glucose A utilization during ET-1 infusion in during combined ETA (BQ123) and ETB receptor blockade (BQ788). Splanchnic blood flow (SBF) and renal blood flow (RBF) were determined by infusions of cardiogreen and p- healthy subjects (12). Ferri et al. (4) dem- aminohippurate. onstrated a negative correlation between total glucose uptake and circulating ET-1 RESULTS — Total-body glucose uptake (M) differed between the clamp protocols with the levels in non–insulin-dependent diabe- highest value in the BQ123ϩBQ788 clamp (P Ͻ 0.05). The M value corrected by insulin was tes. The notion that ET-1 modulates insu- higher in the BQ123ϩBQ788 than in the control clamp (P Ͻ 0.01) or the BQ123 clamp (P Ͻ lin sensitivity was supported by the 0.05). There was no difference between the control clamp and the BQ123 clamp. Mean arterial demonstration that ET-1 reduces insulin pressure did not change during the control clamp, whereas it decreased during both the BQ123 sensitivity in healthy volunteers (13). (P Ͻ 0.01) and BQ123ϩBQ788 (P Ͻ 0.05) clamps. RBF increased and renal vascular resistance decreased in the BQ123ϩBQ788 clamp (P Ͻ 0.05) but not in the BQ123 clamp. There was no Furthermore, the ET-1 precursor, big change in SBF in either clamp. ET-1, reduces insulin sensitivity via an action mediated by the ETA receptor in ϩ CONCLUSIONS — Dual ETA ETB receptor blockade acutely enhances insulin sensitivity healthy subjects (14). However, no study in patients with insulin resistance and coronary artery disease, indicating an important role for has investigated whether endogenous endogenous ET-1. ET-1 contributes to reduced insulin sensitivity in patients with insulin resistance. Diabetes Care 30:591–596, 2007 The present study was therefore designed to investigate the impact of ET-1 receptor nsulin resistance is a key component of resistant states such as type 2 diabetes (4), blockade on insulin sensitivity in patients the metabolic syndrome and is associ- obesity (5), essential hypertension (6), with insulin resistance and coronary ar- I ated with increased cardiovascular risk and coronary artery disease (7) are asso- tery disease. Since ETB receptors on vas- (1). Impairment of endothelial function, ciated with elevated plasma levels of the cular smooth muscle have been shown to characterized by reduced bioactivity of endothelium-derived vasoconstrictor and be upregulated in atherosclerosis (15) ϩ nitric oxide (NO), is an early finding in proinflammatory peptide endothelin and dual ETA ETB receptor blockade patients with insulin resistance (2) and (ET)-1 (8,9). Furthermore, insulin may (but not selective ETA receptor blockade) type 2 diabetes (3). Besides the negative directly stimulate the secretion of ET-1 improves endothelial function in subjects effects on NO availability, insulin- from endothelial cells (8). with insulin resistance (16), we compared ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● the effects of selective ETA and dual ET ϩET receptor antagonism on insulin 1 A B From the Department of Laboratory Medicine, Division of Clinical Physiology, Karolinska Institutet, Stock- sensitivity. holm, Sweden; and the 2Department of Medicine, Division of Cardiology, Karolinska Institutet, Stockholm, Sweden. Address correspondence and reprint requests to John Pernow, Department of Cardiology, Karolinska RESEARCH DESIGN AND University Hospital, Solna, S-171 76 Stockholm, Sweden. E-mail: [email protected]. METHODS — Seven patients (aged Received for publication 22 September 2006 and accepted in revised form 9 December 2006. Ϯ Ϯ 2 Abbreviations: ET, endothelin; MAP, mean arterial blood pressure; PAH, p-aminohippurate; RBF, renal 58 2 years, BMI 31.7 2.6 kg/m ) with blood flow; SBF, splanchnic blood flow. a previous history of impaired glucose tol- A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion erance and coronary artery disease were factors for many substances. recruited. Impaired glucose tolerance was DOI: 10.2337/dc06-1978 defined as fasting blood glucose Ͻ6.1 © 2007 by the American Diabetes Association. Ն The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby mmol/l and a blood glucose 7.8 mmol/l marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 2 h after an oral glucose loading (75 g).

DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 591 Endothelin blockade improves insulin sensitivity

Patients were classified as having diabetes tional uptake, F (equal to the arterio- Calculations and statistics if fasting blood glucose was Ն6.1 mmol/l venous difference divided by the arterial The total-body glucose uptake (M;mg Ϫ Ϫ (on at least on two occasions) or blood concentration), of cardiogreen and PAH kg 1 min 1) was calculated during three glucose concentration was Ն11.1 mmol/l were not influenced by the infusion of the 20-min sampling periods from 60 to 120 2 h after an oral glucose loading. Based on ET-1 blockers (17) or the clamp proce- min during the clamp (i.e., 60–80, 80– these criteria, five patients were classified dure (14). As there was no change in the F 100, and 100–120 min). This means that as having diabetes and two as having im- values, a constant F value of 0.8 for car- the first sampling was made during ET paired glucose tolerance. Coronary artery diogreen and 0.9 for PAH was used in the receptor blockade on the occasions when disease was defined as a history of previ- present study. antagonists were administered. The M ous myocardial infarction or significant Insulin, dissolved in 0.9% saline and value was then corrected for the mean of coronary stenosis verified by coronary an- blood, was infused at a rate correspond- the two plasma insulin values obtained giography. The patients were taking aspi- ing to 804 mU/m2 body surface area dur- during each period to calculate the M/I rin (n ϭ 7), statins (n ϭ 6), fibrates (n ϭ ing the first 8 min, followed by 40 mU/m2 value, which represents a measure of in- 1), ACE inhibitors (n ϭ 5), ␤-blockers per min for 112 min. Fasting blood glu- sulin sensitivity. Splanchnic and renal (n ϭ 7), and oral antidiabetic agents (n ϭ cose level was maintained by adjusting vascular resistance were calculated as 5). Average total, LDL, and HDL choles- the infusion rate of a 20% glucose solu- MAP divided by SBF or RBF, respectively, terol levels were 4.0 Ϯ 0.3, 2.3 Ϯ 0.3, and tion. Arterial blood samples were taken and measured 60 min into the clamp, i.e., 0.9 Ϯ 0.1 mmol/l, respectively. Average every 5 min for determination of blood before administration of the antagonists serum creatinine was 87 Ϯ 5 ␮mol/l. The glucose. Arterial samples for plasma insu- (baseline), and thereafter every 20 min. A subjects were informed of the nature, pur- lin, cardiogreen, PAH, and hematocrit two-way ANOVA with repeated measures pose, and possible risk involved in the were collected at 60, 80, 100, and 120 on two factors was used to analyze the study before giving informed consent. min of the clamp procedure. data. The factors were clamp (three levels) The investigation was carried out in ac- In the control clamp, an infusion of and time (three to four levels). The inter- cordance with the Declaration of Helsinki saline was started at 60 min into the action in the ANOVA refers to the statis- and was approved by the ethics commit- clamp and was maintained for 15 min. In tical test of whether the effect of one tee of the Karolinska Institute. the BQ123 clamp, the ETA receptor an- factor, as measured by differences in the The study consisted of three different tagonist BQ123 was infused at a rate of 5 response averages, is different for differ- Ϫ Ϫ hyperinsulinemic-euglycemic clamp pro- nmol kg 1 min 1. In the combined ent levels of the other factor. In the case of ϩ tocols: 1) a control clamp with saline in- BQ123 BQ788 clamp, the ETB receptor significant interaction, post hoc interac- fusion, 2) a clamp with infusion of the antagonist BQ788 was infused at a rate of tion tests were performed between each Ϫ Ϫ ET receptor antagonist BQ123, and 3)a 4 nmol kg 1 min 1 together with pair of the clamp conditions across the A Ϫ Ϫ clamp with a combined infusion of BQ123 (5 nmol kg 1 min 1). The in- time intervals. Due to positively skewed BQ123 and the ETB receptor antagonist fusions of the antagonists started 60 min distribution, the data for M/I were log BQ788. There was at least 1 week be- into the clamp and were maintained for transformed. Simple effects tests were tween the clamp studies. There were no 15 min. The doses of BQ123 and BQ788 also performed, i.e., effects of one factor dropouts. The investigations were per- were based on previous studies demon- holding the other factor fixed. For these formed in random order, and the patients strating effective hemodynamic responses tests, Fischer’s protected least significant were unaware of the order of the clamps and antagonism of vascular effects evoked difference was performed. Data are pre- and thus blinded to the treatments. On by ET-1 (17–20). sented as means Ϯ SE. the study day, the patient arrived in the Glucose was analyzed in whole blood laboratory after an overnight fast. All according to the glucose dehydrogenase RESULTS medication was withheld on the study method by using a HemoCue B glucose day. Two thin catheters were percutane- photometer (HemoCue, A¨ ngelholm, Swe- Hemodynamic effects ously inserted into one antecubital vein of den) with a precision corresponding to an There was no change in heart rate in either each arm for infusions. The receptor an- SD Ϯ0.3 mmol/l. Plasma insulin was an- group during the clamps, and there were tagonists, saline, cardiogreen, and p- alyzed by radioimmunoassay (12). The no differences in heart rate between the aminohippurate (PAH) were given in one hematocrit was analyzed with a micro- clamps. There was no difference in MAP arm, while glucose and insulin were in- capillary hematocrit centrifuge and cor- between the groups at 60 min, i.e., before fused in the contra-lateral arm. Another rected for trapped plasma. administration of antagonists. However, catheter was introduced into the brachial MAP differed between the three clamps artery for sampling of blood and measure- Drugs (interaction, clamp ϫ time P Ͻ 0.001). ment of systemic arterial blood pressure. BQ123 (Clinalfa, Laüfelfingen, Switzer- When compared with the control clamp, Heart rate and mean arterial blood pres- land) and BQ788 (Neosystem, Stras- MAP was reduced following administra- sure (MAP) were continuously monitored bourg, France) were dissolved in sterile tion of BQ123 (clamp ϫ time P Ͻ 0.01) and recorded every 20 min throughout 0.9% NaCl, which was sterile filtered and following administration of BQ123ϩ the studies. Splanchnic blood flow (SBF) through a Millipore filter and stored fro- BQ788 (clamp ϫ time P Ͻ 0.05) (Table and renal blood flow (RBF) were deter- zen at Ϫ80°C thereafter. On the day of the 1). mined by the constant infusions of car- experiments, all substances were diluted RBF at 60 min did not differ between diogreen, PAH, and the hematocrit, as to the proper concentrations in sterile the clamps, and it did not change in the previously described (12). In a previous 0.9% NaCl. Insulin (Actrapid, 100 IE/ml; control or BQ123 clamps. In contrast, study, hepatic and renal vein catheters Novo Nordisk, Bagsværd, Denmark) was RBF increased by 24% (P Ͻ 0.01) follow- were introduced to ascertain that frac- dissolved in 0.9% saline and blood. ing administration of BQ123ϩBQ788

592 DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 Ahlborg and Associates

Table 1—Hemodynamic data during the study

60 min 80 min 100 min 120 min ANOVA P Heart rate (bpm) Control 65 Ϯ 465Ϯ 468Ϯ 569Ϯ 6NS BQ123 65 Ϯ 463Ϯ 467Ϯ 667Ϯ 5NS Dual BQ 63 Ϯ 463Ϯ 564Ϯ 366Ϯ 4NS MAP (mmHg) Control 94 Ϯ 496Ϯ 498Ϯ 596Ϯ 4NS BQ123 103 Ϯ 397Ϯ 291Ϯ 3* 96 Ϯ 3† Ͻ0.01 Dual BQ 97 Ϯ 491Ϯ 5† 93 Ϯ 4† 93 Ϯ 4† Ͻ0.05 Arterial glucose (mmol/l) Control 8.2 Ϯ 1.2 8.0 Ϯ 1.1 8.0 Ϯ 1.2 7.9 Ϯ 1.3 NS BQ123 8.2 Ϯ 1.3 8.0 Ϯ 1.3 8.1 Ϯ 1.3 8.0 Ϯ 1.2 NS Dual BQ 7.9 Ϯ 1.3 8.0 Ϯ 1.3 8.4 Ϯ 1.2 8.5 Ϯ 1.2 NS Plasma insulin (pmol/l) Control 730 Ϯ 51 731 Ϯ 30 577 Ϯ 58† 714 Ϯ 52 NS BQ123 622 Ϯ 46 644 Ϯ 54 663 Ϯ 28 629 Ϯ 44 NS Dual BQ 644 Ϯ 40 624 Ϯ 66 633 Ϯ 49 704 Ϯ 42 NS Data are means Ϯ SE, n ϭ 7. Times 60–120 min denote minutes into the clamp. ANOVA calculated for all time points within each group. *P Ͻ 0.01, †P Ͻ 0.05 vs. 60 min, i.e., before the administration of antagonists. BQ123, clamp with administration of BQ123; Dual BQ, clamp with administration of BQ123ϩBQ788.

(Fig. 1A). RBF was signiﬁcantly higher in BQ123 clamp (P Ͻ 0.001). RBF did not There were signiﬁcant differences in the BQ123ϩBQ788 clamp compared differ between the control and the BQ123 renal vascular resistance between the with the control clamp (P Ͻ 0.01, inter- clamps. There was no difference in SBF clamp protocols (P Ͻ 0.05, clamp ϫ time action clamp ϫ time P Ͻ 0.05) and the between the three clamps (Fig. 1B). P Ͻ 0.01) (Fig. 1C). Renal vascular resis-

Figure 1—Changes in renal and splanchnic hemodynamics during the control clamp (Control), selective ETA receptor blockade (BQ123), and dual ϩ ϩ o ETA ETB receptor blockade (BQ123 788). , period of saline/antagonist infusion. Signiﬁcant differences between groups are indicated. Data are presented as means and SE; n ϭ 7.

DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 593 Endothelin blockade improves insulin sensitivity

Figure 2—Total-body glucose uptake (M) and insulin sensitivity (M/I) values at the three sampling periods during the control clamp (Control), ϩ ϩ selective ETA receptor blockade (BQ123), and dual ETA ETB receptor blockade (BQ123 788). Values were calculated during three 20-min periods following administration of saline/antagonists and then corrected for the mean of the two plasma insulin values obtained during each period (M/I). Signiﬁcant differences between groups are indicated. *P Ͻ 0.05. Data are presented as means and SE; n ϭ 7.

tance was significantly lower in the CONCLUSIONS — The main find- ministration of the ETB receptor antago- BQ123ϩBQ788 clamp than in the con- ing of the present study is that total-body nist alone was not tested, since selective Ͻ ϫ Ͻ trol clamp (P 0.05, clamp time P glucose uptake (M) and insulin sensitivity ETB receptor blockade in the absence of 0.05). Furthermore, renal vascular resis- (M/I) were acutely increased during dual ETA receptor blockade may exert adverse ϩ ϩ tance was lower in the BQ123 BQ788 ETA ETB receptor blockade in obese pa- hemodynamic effects in patients with clamp than in the BQ123 clamp (P Ͻ tients with insulin resistance and coro- heart failure (18). It is often assumed that 0.05). There were no significant differ- nary artery disease. On the other hand, no most of the pathophysiological effects of ences in splanchnic vascular resistance difference in M or M/I values were ob- ET-1 are mediated via stimulation of the between the clamp protocols (Fig. 1D). served following selective ETA receptor ETA receptor. In vitro studies have shown blockade. These observations indicate that ET-1 inhibits insulin-stimulated glu- Effect on glucose and insulin that ET-1 contributes to insulin resistance cose uptake in isolated rat adipocytes (22) Arterial glucose values remained un- and that the ETB receptor plays an impor- and isolated skeletal muscle strips (21) changed and did not differ between the tant role. mainly through the ETA receptor. Fur- clamps. There were no differences in ar- Previous studies (2) have demon- thermore, chronic selective ETA receptor terial insulin levels between the three strated that plasma levels of ET-1 are ele- antagonism improves insulin sensitivity clamp protocols (Table 1). Glucose up- vated in patients with insulin resistance and reduces hyperinsulinemia in animal take and insulin sensitivity were deter- and/or type 2 diabetes. Furthermore, a models (23). We have shown that ETA mined during saline infusion in the negative correlation between total glucose receptor blockade improves, while selec- control clamp and following antagonist uptake and circulating ET-1 levels was tive ETB receptor blockade impairs, insu- administration in the BQ clamps. There found in diabetic patients (4). In that lin sensitivity during infusion of the ET-1 were significant (P Ͻ 0.05) differences in study, no correction was done for the in- precursor big ET-1 in healthy humans total-body glucose uptake, M values, be- sulin levels, and the correlation between (14). In the present study, insulin sensi- ϩ tween the three clamp protocols (Fig. 2A). ET-1 and insulin sensitivity could not be tivity improved only after dual ETA ETB The M value was significantly higher in elucidated. Other studies show that ad- receptor blockade but not following se- ϩ the in the BQ123 BQ788 clamp than in ministration of exogenous ET-1 reduces lective ETA receptor blockade, indicating the control clamp and in comparison with insulin sensitivity in healthy humans (13) that the ETB receptor plays a different role the BQ123 clamp (P Ͻ 0.05). There was and in experimental animals (21). How- in health and disease. It has previously no difference in M value between the con- ever, the role of endogenous ET-1 in the been demonstrated that the expression of trol and the BQ123 clamps. regulation of insulin sensitivity in cardio- smooth muscle cell ETB receptors that There were differences in insulin sen- vascular disease is still incompletely un- mediate vasoconstriction are increased in sitivity, expressed as M/I values, between derstood. Based on these previous atherosclerotic human arteries (15). The Ͻ the three clamps (P 0.02). The M/I observations, we hypothesized that en- expression of ETB receptors is also in- value was significantly higher in the dogenous ET-1 is involved in the regula- creased in mice overexpressing ET-1 (24), BQ123ϩBQ788 clamp than in the con- tion of glucose uptake by insulin in suggesting that there may be a change in Ͻ trol clamp (P 0.01) and the BQ123 patients with insulin resistance and coro- balance between ETA and ETB receptors clamp (P Ͻ 0.05). The M/I value tended nary artery disease. To our knowledge, in states of enhanced expression of ET-1. to be higher already at the first measure- the present study is the first to demon- The increased expression of ETB receptors ϩ ment during dual receptor blockade with strate that dual ETA ETB receptor block- has functional consequences, as dual ϩ ϩ BQ123 BQ788, and it became signifi- ade acutely increased the M and the M/I ETA ETB receptor blockade results in cantly higher in comparison with the values in this group of patients, clearly more pronounced vasodilatation than se- other groups at measurements two and supporting our hypothesis. lective ETA receptor blockade in the fore- three (Fig. 2B). There was no difference The effect on insulin sensitivity was arm of patients with atherosclerosis (25), between the control clamp and the investigated using both selective ETA and as well as in the renal vascular bed of the ϩ BQ123 clamp. dual ETA ETB receptor blockade. Ad- present study. Furthermore, dual

594 DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 Ahlborg and Associates

ϩ ETA ETB receptor blockade but not se- cular smooth muscle cells (21,26). Fur- its negative effect is antagonized by dual ϩ lective ETA improves endothelium- thermore, a recent study (27) suggests ETA ETB receptor blockade. dependent vasodilatation in clinically that ET-1 impairs glucose transporter healthy subjects with insulin resistance GLUT4 trafficking via interference with (16). Collectively, these observations in- phosphatidylinositol 4,5-bisphosphate– Acknowledgments— This study was sup- ϩ dicate that dual ETA ETB receptor block- regulated cytoskeletal events. These ob- ported by grants from the Swedish Research ade may result in more favorable effects servations suggest that ET receptor Council Medicine (10374 and 10857), the Swedish Heart and Lung Foundation, the than selective ETA in certain pathophysi- blockade may result in increased insulin ological situations. The present data are in Actelion Research Award, the King Gustav and sensitivity via effects on insulin signaling. Queen Victoria Foundation, the Stockholm accordance with this notion and suggest However, effects related to increased re- ϩ County Council, and the Novo Nordisk that dual ETA ETB receptor blockade gional blood flow and bioavailability of Foundation. may be preferred over selective ETA re- insulin by ET receptor blockade cannot ceptor blockade in order to improve insu- be ruled out. lin sensitivity in patients with insulin A limitation of the present study is resistance and coronary artery disease. References that only the acute effect of ET receptor 1. Dekker JM, Girman C, Rhodes T, Nijpels The changes in hemodynamic param- blockade on insulin resistance was inves- eters support a role for ET-1 in the regu- G, Stehouwer CD, Bouter LM, Heine RJ: tigated in a limited study group. No infor- Metabolic syndrome and 10-year cardio- lation of vascular tone. Renal vascular mation is available regarding the long- vascular disease risk in the Hoorn Study. resistance but not splanchnic vascular re- ϩ term effects of dual ETA ETB receptor Circulation 112:666–673, 2005 sistance was reduced by dual receptor blockade on insulin sensitivity. Further- 2. Wheatcroft SB, Williams IL, Shah AM, blockade, suggesting regional differences more, it cannot be ruled out that a small Kearney MT: Pathophysiological implica- regarding the influence of ET-1 on vascu- significant effect in insulin sensitivity by tions of insulin resistance on vascular en- lar tone in this patient group. The more selective ET receptor blockade may be dothelial function. Diabet Med 20:255– ϩ A 268, 2003 pronounced effect of dual ETA ETB re- detected in larger study groups. Based on ceptor blockade, compared with selective 3. De Vriese AS, Verbeuren TJ, Van de the present findings, studies using orally Voorde J, Lameire NH, Vanhoutte PM: ETA receptor blockade, on renal hemody- available ET receptor antagonists in larger namics is in agreement with the effect on Endothelial dysfunction in diabetes. Br J patient groups are warranted to further Pharmacol 130:963–974, 2000 M and M/I values. The finding is also in establish the effect on insulin sensitivity. 4. Ferri C, Carlomagno A, Coassin S, Baldon- agreement with observations in the fore- Nevertheless, the present study was ade- cini R, Cassone Faldetta MR, Laurenti O, arm of patients with atherosclerosis (25) quately sized to detect a significant effect Properzi G, Santucci A, De Mattia G: Cir- and indicates importance of ET receptors B of dual ET ϩET receptor blockade in culating endothelin-1 levels increase dur- contributing to vasoconstrictor tone in A B ing euglycemic hyperinsulinemic clamp comparison with the control group and cardiovascular disease. However, the re- in lean NIDDM men. Diabetes Care 18: selective ET receptor blockade. Another sults contrast those obtained in patients A 226–233, 1995 with chronic renal failure and hyperten- limitation is that only one dose of each 5. Ferri C, Bellini C, Desideri G, Di Fran- antagonist was tested. Both protocols cesco L, Baldoncini R, Santucci A, De Mattia sion (19). Selective ETA receptor blockade, but not dual ET ϩET receptor using antagonists reduced MAP, demon- G: Plasma endothelin-1 levels in obese A B hypertensive and normotensive men. Dia- blockade, increased RBF and reduced re- strating pharmacological effects. Further- more, comparable doses have previously betes 44:431–436, 1995 nal vascular resistance. Important differ- 6. Kohno M, Yasunari K, Murakawa K, ences between these studies are that been demonstrated to exert hemodynamic effects in patients with heart failure Yokokawa K, Horio T, Fukui T, Takeda T: patients in our study had normal renal Plasma immunoreactive endothelin in es- function, based on serum creatinine lev- (18) and to antagonize the vasoconstric- sential hypertension. Am J Med 88:614– els, and normal blood pressure. Further tor effect of exogenous ET-1 in healthy 618, 1990 studies are needed to evaluate the hemo- humans (14,17,20), suggesting efficient 7. Zeiher AM, Goebel H, Schachinger V, Ih- doses of both antagonists. It is difficult to ling C: Tissue endothelin-1 immunoreac- dynamic effects of selective ETA receptor ϩ speculate whether baseline medication tivity in the active coronary atherosclerotic blockade versus dual ETA ETB receptor blockade in patients with cardiovascular may have influenced insulin sensitivity. plaque: a clue to the mechanism of increased vasoreactivity of the culprit lesion in disease. However, baseline medication was un- changed during the study, suggesting that unstable angina. Circulation 91:941–947, The mechanism behind the effect of 1995 ET-1 on insulin sensitivity is not yet fully any influence on insulin sensitivity was 8. Ferri C, Pittoni V, Piccoli A, Laurenti O, understood. Previous studies on healthy similar on all occasions. In addition, ad- Cassone MR, Bellini C, Properzi G, humans and on isolated skeletal muscle ministration of an ACE inhibitor does not Valesini G, De Mattia G, Santucci A: Insu- strips suggest that this effect is indepen- attenuate the positive effect of ET receptor lin stimulates endothelin-1 secretion from dent of a reduction in skeletal muscle blockade on endothelial function in pa- human endothelial cells and modulates its blood flow (21). Instead, the results imply tients with atherosclerosis (28). circulating levels in vivo. J Clin Endocrinol a direct influence of ET-1 on insulin- In conclusion, the present study dem- Metab 80:829–835, 1995 stimulated glucose transport. Accord- onstrates that dual ET ϩET receptor 9. Yanagisawa M, Kurihara H, Kimura S, To- A B mobe Y, Kobayashi M, Mitsui Y, Yazaki Y, ingly, ET-1 decreases the expression of blockade acutely enhances insulin sensi- Goto K, Masaki T: A novel potent vaso- insulin receptor substrate-1 and its acti- tivity in patients with insulin resistance constrictor peptide produced by vascular vation of phosphatidylinositol 3-kinase and coronary artery disease. The data endothelial cells. Nature 332:411–415, pathway and insulin-stimulated Akt support that ET-1 is involved in the de- 1988 phosphorylation in skeletal muscle vas- velopment of insulin resistance and that 10. Arai H, Hori S, Aramori I, Ohkubo H, Na-

DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 595 Endothelin blockade improves insulin sensitivity

kanishi S: Cloning and expression of a G: ETA receptors mediate vasoconstric- mainly through ETA receptors. Metabo- cDNA encoding an endothelin receptor. tion, whereas ETB receptors clear endo- lism 47:1468–1471, 1998 Nature 348:730–732, 1990 thelin-1 in the splanchnic and renal 23. Berthiaume N, Wessale JL, Opgenorth TJ, 11. Sakurai T, Yanagisawa M, Takuwa Y, circulation of healthy men. Clin Sci (Lond) Zinker BA: Metabolic responses with en- Miyazaki H, Kimura S, Goto K, Masaki T: 104:143–151, 2003 dothelin antagonism in a model of insulin Cloning of a cDNA encoding a non- 18. Cowburn PJ, Cleland JG, McDonagh TA, resistance. Metabolism 54:735–740, 2005 isopeptide-selective subtype of the endo- McArthur JD, Dargie HJ, Morton JJ: Com- 24. Amiri F, Virdis A, Neves MF, Iglarz M, thelin receptor. Nature 348:732–735, parison of selective ET(A) and ET(B) re- Seidah NG, Touyz RM, Reudelhuber 1990 ceptor antagonists in patients with TL, Schiffrin EL: Endothelium-restricted 12. Ahlborg G, Weitzberg E, Lundberg JM: chronic heart failure. Eur J Heart Fail overexpression of human endothelin-1 Endothelin-1 infusion reduces splanch- 7:37–42, 2005 causes vascular remodeling and endothe- nic glucose production in humans. J Appl 19. Goddard J, Johnston NR, Hand MF, Cum- lial dysfunction. Circulation 110:2233– Physiol 77:121–126, 1994 ming AD, Rabelink TJ, Rankin AJ, Webb 2240, 2004 13. Ottosson-Seeberger A, Lundberg JM, DJ: Endothelin-A receptor antagonism re- 25. Bohm F, Ahlborg G, Johansson BL, Hans- Alvestrand A, Ahlborg G: Exogenous en- duces blood pressure and increases renal son LO, Pernow J: Combined endothelin dothelin-1 causes peripheral insulin resis- blood ﬂow in hypertensive patients with receptor blockade evokes enhanced vaso- tance in healthy humans. Acta Physiol chronic renal failure: a comparison of se- dilatation in patients with atherosclerosis. Scand 161:211–220, 1997 lective and combined endothelin receptor Arterioscler Thromb Vasc Biol 22:674– 14. Ahlborg G, Lindstrom J: Insulin sensitiv- blockade. Circulation 109:1186–1193, ity and big ET-1 conversion to ET-1 after 2004 679, 2002 ETA- or ETB-receptor blockade in hu- 20. Spratt JC, Goddard J, Patel N, Strachan 26. Jiang ZY, Zhou QL, Chatterjee A, Feener mans. J Appl Physiol 93:2112–2121, 2002 FE, Rankin AJ, Webb DJ: Systemic ETA EP, Myers MG Jr, White MF, King GL: 15. Iwasa S, Fan J, Shimokama T, Nagata M, receptor antagonism with BQ-123 blocks Endothelin-1 modulates insulin signaling Watanabe T: Increased immunoreactivity ET-1 induced forearm vasoconstriction through phosphatidylinositol 3-kinase of endothelin-1 and endothelin B receptor and decreases peripheral vascular resis- pathway in vascular smooth muscle cells. in human atherosclerotic lesions: a possi- tance in healthy men. Br J Pharmacol 134: Diabetes 48:1120–1130, 1999 ble role in atherogenesis. Atherosclerosis 648–654, 2001 27. Strawbridge AB, Elmendorf JS: Endothe- 146:93–100, 1999 21. Wilkes JJ, Hevener A, Olefsky J: Chronic lin-1 impairs glucose transporter trafﬁck- 16. Shemyakin A, Bohm F, Wagner H, Efen- endothelin-1 treatment leads to insulin ing via a membrane-based mechanism. dic S, Bavenholm P, Pernow J: Enhanced resistance in vivo. Diabetes 52:1904– J Cell Biochem 97:849–856, 2006 endothelium-dependent vasodilatation 1909, 2003 28. Bohm F, Beltran E, Pernow J: Endothelin by dual endothelin receptor blockade in 22. Lee YC, Juan CC, Fang VS, Hsu YP, Lin receptor blockade improves endothelial individuals with insulin resistance. J Car- SH, Kwok CF, Ho LT: Evidence that en- function in atherosclerotic patients on an- diovasc Pharmacol 47:385–390, 2006 dothelin-1 (ET-1) inhibits insulin-stimu- giotensin converting enzyme inhibition. 17. Bohm F, Pernow J, Lindstrom J, Ahlborg lated glucose uptake in rat adipocytes J Intern Med 257:263–271, 2005

596 DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007