- Email: [email protected]
Endothelial dysfunction induced by triglycerides is not restored by exenatide in rat conduit arteries ex vivo
Regulatory Peptides 157 (2009) 8–13
Contents lists available at ScienceDirect
Regulatory Peptides j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / r e g p e p
Endothelial dysfunction induced by triglycerides is not restored by exenatide in rat conduit arteries ex vivo David Nathanson a,⁎, Özlem Erdogdu a, John Pernow b, Qimin Zhang a, Thomas Nyström a a b
Karolinska Institutet, Department of Internal Medicine, South Hospital, Sweden Karolinska Institutet, Department of Cardiology, Karolinska Hospital, Stockholm, Sweden
a r t i c l e
i n f o
Article history: Received 11 July 2008 Received in revised form 14 May 2009 Accepted 2 July 2009 Available online 10 July 2009 Keywords: Endothelial function eNOS Exenatide GLP-1 Intralipid
a b s t r a c t Exenatide (synthetic exendin-4) is a stable analogue of glucagon-like peptide 1 (GLP-1) and has recently been approved for clinical use against type 2 diabetes. Exenatide is believed to exert its effects via the GLP-1 receptor with almost the same potency as GLP-1 in terms of lowering blood glucose. Short term exenatide treatment normalizes the altered vascular tone in type 2 diabetic rats, probably due to the reduction in glycemia. The aim of this study was to investigate whether exenatide directly protects against triglyceride-induced endothelial dysfunction in rat femoral arterial rings ex vivo. Short term pre-incubation with Intralipid® (0.5 and 2%) was found to dose-dependently induce endothelial dysfunction, in that it elicited a significant reduction in AChinduced vasorelaxation by 29% and 35%, respectively. Paradoxically, this occurred with a concomitant increase in endothelial nitric oxide synthase (eNOS) activity. No such reduction in vasorelaxation by Intralipid® was seen in response to the NO donor sodium nitroprusside (SNP), revealing an endothelium-dependent vascular dysfunction by Intralipid®. However, exenatide did not protect against Intralipid®-induced endothelial dysfunction. More surprisingly, the maximum vasorelaxation induced by exenatide (without Intralipid®) was only 3 ± 2%, compared to the 23 ± 4%, 38 ± 4%, 79 ± 3% and 97 ± 4% relaxations induced by GLP-1, GLP-1 (9-36), ACh and SNP, respectively. This unexpected finding prompted us to ascertain that the exenatide preparation was biologically active, and both exenatide (10− 11 mol/l) and GLP-1 (10− 9 mol/l) significantly increased insulin secretion in pancreatic β-cells from ob/ob mice in vitro. In conclusion, exenatide could neither confer any acute protective effects against triglyceride-induced endothelial dysfunction nor exert any significant vasorelaxant actions in this model of rat conduit arteries ex vivo. © 2009 Elsevier B.V. All rights reserved.
1. Introduction One salient feature in type 2 diabetes is endothelial dysfunction, which closely correlates to coronary heart disease (CHD) [1]. Also, dyslipidemia, involving hypertriglyceridemia and elevation in circulating free fatty acids (FFAs), is common in type 2 diabetes and impacts the endothelium negatively [2,3]. Endothelial dysfunction is a prognostic factor for the outcome of CHD [4,5]. In attempts to improve endothelial function, numerous studies have evaluated lifestyle and pharmacology interventions to decrease the risk for CHD [1,6]. Beside its well known antihyperglycemic effect, glucagon-like peptide-1 (GLP-1) has been shown to ameliorate endothelial dysfunction in type 2 diabetic patients with CHD [7] and dose-dependently induce vasorelaxation in rat conduit arteries ex vivo [8]. Subsequently, it was also recently demonstrated that GLP-1 per se augments endothelium-dependent vasodilatation in non-diabetic humans [9]. ⁎ Corresponding author. Södersjukhuset, Ringvägen 52, SE-118 83 Stockholm, Sweden. Tel.: +46 86163449; fax: +46 86163146. E-mail address: [email protected] (D. Nathanson). 0167-0115/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2009.07.003
GLP-1 acts as an incretin, i.e., released from intestinal cells following ingestion of food, this peptide lowers blood glucose by augmenting the secretion of insulin and inhibiting glucagon secretion, as well as by inhibiting bowel motility and promoting the feeling of satiety [10]. GLP-1 is degraded rapidly by the ubiquitous enzyme dipeptidyl peptidase IV (DPPIV), as a result of which its half-life in the circulation is no more than 1–2 min [10]. For this reason, both longer-acting GLP-1 analogues and inhibitors of DPPIV have been developed. Since these analogues are believed to exert all of their actions through the GLP-1 receptor (GLP-1R), their biological effects are widely assumed be the same as those of native GLP-1. Long-term improvements of glycemic control by exenatide treatment occur along with reductions in body weight [11]. Interestingly, also lipid parameters, blood pressure, and proinflammatory biomarkers have been shown to improve in patients treated with exenatide, supporting a shift toward a more favorable cardiovascular risk profile that might impact positively on the endothelium [11]. Administration of exenatide to diabetic rats nearly normalizes their vascular tone [12]. However, it is not clear whether these effects merely reflect an improved metabolic state or a direct vascular action of exenatide [12].
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
Therefore, in the present study, we wanted to characterize possible direct effects of exenatide on the vasculature, thereby gaining mechanistic insights into the salutary actions of exenatide on endothelial function. The aim of this study was to investigate whether exenatide protects against endothelial dysfunction, induced by a triglyceride-rich fat emulsion, in rat femoral arterial rings ex vivo. 2. Materials and methods This study was approved by the regional ethics committee for animal research and conforms to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institute of Health (NIH publication No. 85-23, revised 1985). 2.1. Preparation of arterial rings Sixty-nine male Sprague–Dawley rats (weight 250–350 g) were anesthetized with a mixture of fluanisonum and fentanylum (Hypnorm®, Janssen, Beerse, Belgium) and midazolam (Dormicum®, Hoffman–LaRoche, Basel, Switzerland) (2.5, 0.08 and 1.25 mg/kg, respectively, i.m.). The rats were then killed by excision of the heart. The femoral arteries were carefully dissected free from surrounding tissue, removed and put in Krebs–Henseleit (KH) solution [in mmol/l: NaCl 118, KCl 4.7, KH2PO4 1.2, MgSO4(H2O) 1.2, NaHCO3 25.2, CaCl2 2.5 and glucose 11.1]. Circular segments (1–2 mm in length) of the artery were mounted on two thin metal holders, one of which was connected to a force displacement transducer (model FT03, Grass Instrument Co, Quincy, MA) and the other to a movable device that allowed the application of a passive tension of 5 mN. The tension was recorded on a polygraph (model 7B, Grass). The mounted vascular segments were kept in 2 ml organ baths containing KH solution at 37 °C and continuously bubbled with 5% CO2 in O2 to maintain a pH of 7.4. After preparation, the vascular segments were allowed to equilibrate for 60 min. 2.2. Functional experiments The contractile function of the vascular segments was tested by administration of phenylephrine (Phe) (Sigma, final concentration 10− 5 mol/l) and with K+-rich (127 mmol/l) KH solution, prepared by replacing NaCl with equimolar amounts of KCl. Endotheliumdependent and endothelium-independent relaxations were determined by administration of acetylcholine (ACh) (Sigma) and the NO donor sodium nitroprusside (SNP) (Alexis Corporation, Läufelfingen, Switzerland), respectively. ACh and SNP were added to the organ baths at cumulatively increasing concentrations (10− 9–10− 5 mol/l and 10− 9– 10− 6 mol/l, respectively) during a stable contractile tone induced by Phe (10− 5 mol/l). The relaxant response following pre-incubation with a studied substance was always compared to the preceding control response in the same vascular segment. All substances used in the protocol were added in 50 µl volumes. Exenatide (Neosystem, Strasbourg, France) was added to the organ baths at cumulative increasing concentrations (10− 13–10− 8 mol/l) during baseline tension to evaluate contractile effects per se. Furthermore, exenatide (10− 9 mol/l) was in separate experiments added 10 min before a dose–response curve for Phe to test for co-activation of Phe-induced contractions. The relaxant effects of exenatide, GLP-1(736)amide (Neosystem, Strasbourg, France) and GLP-1 (9-36) (Bachem, Bubendorf, Switzerland) were evaluated by adding cumulatively increasing concentrations (10− 13–10− 8 mol/l) of these drugs to artery segments precontracted with Phe (10− 5 mol/l). The relaxant effect by GLP-1 was also tested together with the DPPIV inhibitor sitagliptin (5 µM) which was, in separate experiments, added 5 min before Pheinduced contraction. Sitagliptin was also added during baseline tension and during Phe-induced contraction to evaluate any contractile and vasorelaxant effects of the DPPIV inhibitor per se. Finally, to investigate whether exenatide might co-activate the ACh relaxation, exenatide
9
(2.5 nmol/l) was added in separate experiments to the organ baths 10 min before contraction with Phe (10− 5 mol/l). Thereafter the vessels were relaxed by ACh in increasing concentrations (10− 9– 10− 5 mol/l). To induce endothelial dysfunction, the vascular segments were preincubated for 20 min with Intralipid® [100 mg/ml] (PharmaciaUpjohn, Uppsala, Sweden) diluted in KH solution to final concentrations of 0.5 and 1%, corresponding to an approximate triglyceride level in vivo of 5 and 10 mmol/l, respectively. To examine any protective effect of exenatide against the triglyceride-induced endothelial dysfunction, exenatide (final concentration of 2.5 nmol/l) was administered to the organ baths 20 min before the vascular response test of ACh and SNP. After the end of the protocols, artery rings were frozen and stored at − 80 °C for subsequent immunoblotting analysis. For determination of cAMP contents, the thoracic aorta was removed, cleaned and cut into two sections. Each vessel segment was equilibrated for 20 min in KH at 37 °C and bubbled with 5% CO2 in O2 to maintain a pH of 7.4, before incubation. One vessel segment was then incubated for 15 min with GLP-1 (0.1 µM) or exenatide (2.5 nM), respectively, and the other with an equivalent volume of the adenylyl cyclase activator forskolin (10 µM) or DMSO (10 µM), serving as positive control. Vessels were then blotted dry and frozen in liquid nitrogen and stored at −80 °C for further analysis. 2.3. Immunoblotting Frozen artery rings were thawed and homogenized in 100 µl buffer containing 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 1% TritonX-100, 10% glycerol, 20 mM Tris (pH 7.8), 1 mM EDTA with phosphatase inhibitors 0.5 mM Na3VO4, 1 µg/ml leupeptin, 0.2 mM PMSF, 10 mM NaF, 1 µg/ml aprotinin, 5 mM Na-pyrophosphate and 100 mM benzamidine. The tissue was minced and homogenized and incubated on ice for 30 min and centrifuged (5000 g, 5 min). Protein concentrations were determined by the bicinchoninic acid method (Pierce Chemical Co., Rockford, IL). Phosphorylation of the enzyme endothelial nitric oxide synthase (eNOS) was examined by anti-phospho specific antibody as a measure of eNOS activity (Santa Cruz). Expression levels of the GLP-1 receptor were examined with 2 µg/ml rabbit anti-NH2-terminal GLP-1 receptor antibody (a generous gift of Dr. Bernard Thorens). 2.4. Determinations of cAMP For determination of cAMP concentrations, frozen aortae were weighed and homogenized (1500 rpm) using a reusable pellet pestle (Kimble-Kontes, Vineland, New Jersey, USA) fixed into an overhead stirrer (SS10, Stuart Scientific, Surrey, UK). Homogenates were suspended in trichloroacetic acid (5% (w/v); 500 µl/100 mg of tissue), centrifuged (1500 g, 10 min) and the supernatant was removed. Trichloroacetic acid was extracted from samples using water-saturated diethyl ether before total cAMP contents were measured using an enzyme immunoassay kit (Cayman Europe, Tallinn, Estonia). 2.5. Insulin secretion from pancreatic β-cells Pancreatic islets were isolated from 12 month-old ob/ob mice and dispersed into cells in a buffer containing EDTA (2-[2-(Bis(carboxymethyl)amino)ethyl-(carboxymethyl)amino]acetic acid) [13]. The cells were allowed to recover in RPMI-1640 medium containing 10% FCS overnight. Cells were washed in KRBH buffer containing 135 mM NaCl, 3.6 mM KCl, 5 mM NaHCO3, 0.5 mM NaH2PO4, 0.5 mM MgCl2, 1.5 mM CaCl2, 10 mM HEPES pH 7.4 in the presence of 0.1% BSA. Equal amounts of cells were incubated at 37 °C with exenatide or GLP-1 for 20 min. At the end of the incubation, cells were spun down and the insulin concentration in the supernatants was analyzed using mouse insulin ELISA kits (Mercodia, Uppsala, Sweden).
10
2.6. Binding of
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13 125
I-GLP-1
Human coronary artery endothelial cells (HCAECs; Clonetics) or mouse insulinoma cells (MIN6) were collected and washed three times in KRBH buffer. Cells were re-suspended in KRBH buffer containing 1% BSA. Incubation was performed for 30 min at 37 °C (or 4 °C overnight) in the presence of 125I-GLP-1 (100 pmol/tube) and in the presence or absence of increasing concentrations of unlabeled GLP-1 (0.01 nM– 1000 nM) with 2 × 105 cells/tube. Incubation was terminated by aspirating 2 × 100 µl of cell suspensions from each assay tube, followed by centrifugation through 250 µl of a mixture of phthalic acid esters. The supernatants were eliminated and the tips of the tubes containing cell pellets were excised and counted in a gamma counter [13]. The 100 µl remaining in each assay tube was used as reference. 2.7. Statistics The relaxations induced by ACh were expressed as a percentage of the precontracted tone. The precontracted level was set to 0%, and the baseline level, corresponding to maximal relaxation, to 100%. All values were expressed as means ± S.E.M. Friedman's test (ANOVA) was used to evaluate effects between related samples (conditions and concentration). Differences were deemed to be statistically significant at P b 0.05. 3. Results All vessels tested responded well to ACh and the NO donor SNP in a dose-dependent manner, demonstrating a valid working model for studies of endothelium-dependent and -independent relaxation effects (Fig. 1a). Exenatide, applied during basal tension, was without any discernable contractile effect on the femoral artery rings (data not shown). Moreover, pre-incubation with exenatide did not co-activate
Fig. 1. Differential vasorelaxant responses of arterial rings ex vivo. Vasorelaxant function of femoral artery rings from Sprague–Dawley rats after pre-constriction with phenylephrine (10− 5 M) a) Relaxations induced by acetylcholine (ACh) and by the NO donor sodium nitroprusside (SNP) (n = 14). b) Relaxations induced by control (Krebs-solution) (n = 4), exenatide (n = 16), GLP-1 (n = 7), GLP-1 + sitagliptin (n = 4) and GLP-1 (9-36) (n = 6) ⁎denotes P b 0.01 for a chance difference between exenatide vs GLP-1, GLP-1 + sitagliptin and GLP-1 (9-36), respectively. †Denotes P b 0.05 for a chance difference between GLP-1 (9-36) vs GLP-1 + sitagliptin and GLP-1, respectively. ‡ Denotes P b 0.05 for a chance difference between GLP-1 (9-36) vs GLP-1. Means are compared by ANOVA. Results are presented as mean ± S.E.M.
the ACh-induced vasodilatation, or the Phe-induced contractile tone (data not shown). More surprisingly, exenatide was without any discernable relaxant effects during Phe-induced contractile tone compared to GLP-1 and GLP-1(9-36), which demonstrated a significant dose-dependent vascular relaxation (Fig. 1b). However, the maximal relaxation obtained with the highest concentration of GLP-1 and GLP-1 (9-36) was 23 ± 4% and 38 ± 4%, respectively (Fig.1b), compared to 79 ± 3% and 97 ± 4% relaxation induced by the controls, ACh and SNP, respectively (Fig. 1a). Pre-incubation with Intralipid® resulted in a concentrationdependent inhibition of ACh-induced endothelium-dependent relaxation (Fig. 2a–b). A significant reduction in ACh-induced relaxation was obtained already with 0.5% Intralipid® (Fig. 2a). The highest concentration of Intralipid® (2%) caused a rightward shift of the concentration– response curve for ACh by approximately one order of magnitude (Fig. 2b). The relaxation induced by the highest concentration of Ach (10− 7 mol/l) was 71 ± 8% in the presence of 0.5% Intralipid® vs 93 ± 5% in controls (Fig. 2a), and in the presence of 2% Intralipid® 53± 6% vs 87 ± 2% in controls (Fig. 2b). Pre-incubation with Intralipid® (0.5 or 2%) did not affect endothelium-independent relaxation evoked by the NO-donor SNP (Fig. 3a–b), demonstrating an endothelialdependent dysfunction effect by the triglyceride-rich fat emulsion. Short term exenatide exposure did not protect against the triglyceriderich fat emulsion-induced endothelial dysfunction (Fig. 2). Immunoblotting data revealed a single band with a molecular mass of 67 kDa, corresponding to the GLP-1 receptor seen in MIN6 cells, demonstrating expression of the GLP-1 receptor on the femoral artery, albeit with less abundance compared to the insulin-secreting cells. Immunoblotting data also revealed a significant increase in eNOS activation by Intralipid® incubation at both 0.5 and 2% (Fig. 4). This activation was not significantly influenced after pre-incubation with exenatide or by exenatide itself (Fig. 4). However, there was a clear tendency towards a decreased Intralipid®-induced eNOS activation by co-incubation with exenatide (Fig. 4).
Fig. 2. Endothelial dysfunction evoked by Intralipid® is not reversed by exenatide. Relaxations induced by acetylcholine (ACh) under control conditions and after a 20-min pre-incubation with exenatide (2.5 nmol/l) in the presence of a) Intralipid® 0.5% (n = 10) and b) Intralipid® 2% (n = 8). ⁎ and † denote P b 0.05 and P b 0.01 for chance differences vs controls. Results are presented as mean ± S.E.M.
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
11
The somewhat unexpected negative findings with exenatide prompted us to ascertain that the exenatide preparation used was indeed biologically active. To this end, we performed insulin secretion experiments in vitro with pancreatic β-cells isolated from ob/ob mice, and found a similar enhancement of insulin secretion evoked by exenatide as with equipotent concentrations of GLP-1 (Fig. 5). Binding of 125I-GLP-1 to human coronary artery endothelial cells in vitro was studied in the presence of increasing concentrations of unlabeled GLP-1, and compared to the insulin-secreting cell line MIN6. Incubation of MIN6 cells with 125I-GLP-1 resulted in approximately 30% specific binding, consistent with high GLP-1 receptor expression. In contrast, the endothelial cells showed no specific binding of 125 I-GLP-1 (not shown). 4. Discussion
Fig. 3. Intralipid® does not impair endothelium-independent vasorelaxation. Vascular relaxations induced by the nitric oxide donor sodium nitroprusside (SNP) under control conditions and after a 20-min pre-incubation with exenatide (2.5 nmol/l) in the presence of a) Intralipid® 0.5% (n = 10) and b) Intralipid® 2% (n = 8). ⁎ and † denote P b 0.05 and P b 0.01 for chance differences vs controls. Results are presented as mean± S.E.M.
GLP-1, exenatide and forskolin did all increase the cAMP levels of the tissue, however, no differences in the cAMP levels was observed (78.0 ± 16.3, 56.0 ± 11.5 and 51.3 ± 25.3 pmol/l (mg tissue)− 1, N.S.).
Fig. 4. a) Expression of GLP-1 receptor in rat femoral artery and in mouse insulinoma cells (MIN6). b) Intralipid® activates eNOS. Western blot analysis of expression of eNOS phosphorylated at Ser-1177 in rat aortic segments after 20-min incubation of Intralipid® at 0.5% (IL0.5%) or 2% (IL2%), with/out pre-incubation with 2.5 nM exenatide (EX; n = 6 and n = 6, respectively). P-eNOS, phosphorylated (activated) eNOS. ⁎ Denotes P b 0.05 for a chance difference vs control. Bars represent mean± S.E.M with densitometric readings of P-eNOS normalized to those of beta-actin.
This study demonstrates that the triglyceride-rich fat emulsion Intralipid® acutely and dose-dependently induces endothelial-dependent dysfunction, with a concomitant increase in eNOS activity, in rat conduit artery rings ex vivo. However, no protection against such dysfunction was afforded by exenatide. Even more surprisingly, exenatide did not exert any discernable vasorelaxant effects, compared to GLP-1 or GLP-1 (9-36), in this system. Moreover, the fat emulsion did not impact vascular relaxation elicited by the NO donor nitroprusside (reflecting endothelium-independent relaxation), suggesting that the triglyceride-rich fat emulsion only affects the endothelium-dependent pathway of vasorelaxation. Endothelial dysfunction is a major feature of the early stages of the atherosclerotic process and can be used to predict the outcome of CVD in man [4,5]. Although the molecular causes of endothelial dysfunction have not yet been elucidated in detail, numerous studies indicate that impaired signaling by NO and/or attenuated biosynthesis of this messenger play a central role [14]. In the presence of suboptimal concentrations of substrate or cofactors for the synthesis of NO, eNOS might be uncoupled and produce reactive oxygen species (ROS), such as superoxide anion and hydrogen peroxide, thereby giving rise to oxidative stress. Normally, ROS are scavenged via multiple intra- and extracellular defense mechanisms. However, high concentrations of ROS can exceed the capacity of these systems and quench NO, thereby exacerbating the oxidative stress [15]. In the present study, eNOS activity was substantially increased by short-term exposure to a triglyceriderich fat emulsion ex vivo, concomitant with a significant functional impairment in endothelium-dependent vasodilatation. We can only speculate whether this functional impairment might be a consequence of an increase in ROS production, because we did not measure ROS production. However, it has earlier been demonstrated that hyperglycemia-induced endothelial dysfunction is associated with a heightened eNOS activity due to an uncoupling of the enzyme, thereby inducing oxidative stress [16]. Also, it was recently reported that short-term exposure to triglyceride-rich fat emulsion increases eNOS activity in an
Fig. 5. Exenatide and GLP-1 stimulate insulin release. Insulin secretion from ob/ob mice islet cells after exposure to 10 mM glucose (control) with/out exenatide (EX) at 1, 10, or 20 nM; or GLP-1 at 100 nM. ⁎ Denotes P b 0.05 for a chance difference vs control. Bars represent mean percentage of control ± S.E.M for three independent experiments.
12
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
endothelium-derived cell line [17]. However, in that study it was not addressed whether this increased eNOS activity affected the endothelium functionally [17]. In the present study, although the enhanced eNOS activity induced by Intralipid® tended to decrease after co-exposure to exenatide, it did not restore the endothelial dysfunction evoked by the fat emulsion. Additionally, exenatide alone did not influence eNOS activity in this setting. At least in rat artery rings ex vivo, it seems that exenatide acutely does not affect the key enzyme in NO production regulating vascular tone. It is becoming increasingly clear that GLP-1 and its derivatives, apart from their well characterized glycemic actions, also have certain vascular effects [18]. The GLP-1 mimetic exenatide, besides HbA1c reduction, also shows beneficial effects on several cardiovascular risk factors, i.e., lowering of serum triglycerides and blood pressure. Some studies have even suggested direct vascular effects of exenatide per se [12,19]. However, in the present study, we were not able to demonstrate any direct protective effect of exenatide on endothelial dysfunction induced by a triglyceriderich fat emulsion. More surprisingly, exenatide did not produce any vasorelaxant effects on the conduit vessel. It could be questioned why we used just one single dose, and not a dose response curve of exenatide, to prove our hypothesis. However, we believe that the concentration of exenatide we used in this protocol is equipotent, or even higher, than that obtained in the clinical situation. We also changed the Intralipid® concentrations to resemble a clinical situation in which moderate to high triglyceride concentrations prevail. This caused a rightward shift of the concentration–response curve for ACh by approximately one order of magnitude, indicating a dose–response impairment in endothelial function, but without any protective effects afforded by exenatide. Since our aim was to investigate whether the clinically used drug exenatide protects against triglyceride-induced endothelial dysfunction, we did not include native GLP-1 in the protocol. However, after obtaining the somewhat unexpected result that exenatide failed to exert any vasorelaxant effects in the rat conduit artery, we wanted to compare the vasodilatation properties between exenatide vs GLP-1 and GLP-1(9-36). The latter two dose-dependently relaxed rat the arteries, confirming what we and others have demonstrated previously [8,20–22]. To ascertain that the batch of exenatide we used was active, we also investigated its insulin-releasing capacity in pancreatic β-cells, demonstrating a dose-dependent stimulatory effect. There may be several conceivable explanations as to why exenatide did not produce vasorelaxation in this system, while GLP-1 and GLP-1 (9-36) did. First, most studies reporting vasoactive effects by exenatide were conducted in vivo, in which any direct vasoactive influence might be confounded by other factors (e.g. changes in glycemia) [12,19,23]. Very recently it was reported that exenatide does not evoke any vasorelaxant action in a mouse model [20], thus lending support to our findings. Second, even if ligand-binding analyses in other tissues have not revealed any major differences between exenatide and native GLP-1 in terms of GLP-1 receptor affinity [24,25], [26], we cannot rule out that the affinity to the GLP-1 receptor on endothelial cells might differ between exenatide and GLP-1. In an attempt to investigate the GLP-1 receptor binding capacity of human coronary artery endothelial cells in vitro (see Results section), more specifically aiming at comparing receptor affinity between exenatide and GLP-1, we were unable to detect any specific binding of radiolabeled GLP-1 to these cells compared to insulin-secreting MIN6 cells. Third, as recently demonstrated, the vasorelaxant effects of GLP-1 might not be mediated through occupancy of its receptor [20]. This has actually been demonstrated, in as much as its metabolite GLP-1 (9-36) relaxes the artery independent of the GLP-1 receptor [20]. Fourth, there is only a ~50% homology in the amino-acid structure between exenatide and GLP-1, which might explain the differences in vascular effects observed. The effects on glycemia seem to be equal between GLP-1 and exenatide, and most effects of exenatide on glycemia-related endpoints are believed (or assumed) to be conveyed via its signaling through the GLP-1 receptor, which also was demonstrated in present study finding no differences in cAMP productions
between GLP-1 and exenatide. However, we cannot rule out an alternative scenario in terms of vascular responses to these two peptides. In conclusion, the stable GLP-1 analogue exenatide does not acutely ameliorate endothelial dysfunction induced by a triglyceride-rich emulsion in this ex vivo model of rat conduit artery rings. Further in contrast to native GLP-1, and its metabolite GLP-1 (9-36), exenatide surprisingly does not exert any vasorelaxant effects in this system. To date, research on GLP-1 and its derivatives has focused primarily on hyperglycemia endpoints and elucidation of its cardiovascular effects is still in its infancy. As our findings are restricted to acute effects ex vivo on non-diabetic rat conduit arteries, we do not know which effects, if any, chronic exenatide treatment confers to vascular function in vivo in type 2 diabetic patients. Thus, and in the light of the frequency and severity of premature CVD in type 2 diabetes, we still believe that this track should be intensely pursued in future research efforts. Acknowledgments We thank Professor Åke Sjöholm for the critical reading of the manuscript, and Professor Jon Lundberg for allowing us to occupy his laboratory. We thank Margareta Stensdotter and Adrian Gonon for the excellent technical support. Financial support was received from the Swedish Society for Medical Research, the Swedish Society of Medicine, the European Foundation for the Study of Diabetes, Olle Engkvist Byggmästare Foundation, Åke Wibergs Foundation, Loo and Hans Foundation, Tore Nilsson Foundation, Fredrik and Ingrid Thurings Foundation, Merck Sharp & Dohme (Medical School Grant) and the Swedish Research Council (# 10857, 72X-12550, 72X-14507, and 72P-14787). References [1] Caballero AE. Endothelial dysfunction in obesity and insulin resistance: a road to diabetes and heart disease. Obes Res 2003;11:1278–89. [2] Lundman P, Eriksson M, Schenck-Gustafsson K, Karpe F, Tornvall P. Transient triglyceridemia decreases vascular reactivity in young, healthy men without risk factors for coronary heart disease. Circulation 1997;96:3266–8. [3] Steinberg HO, Tarshoby M, Monestel R, Hook G, Cronin J, Johnson A, Bayazeed B, Baron AD. Elevated circulating free fatty acid levels impair endotheliumdependent vasodilation. J Clin Invest 1997;100:1230–9. [4] Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes Jr DR, Lerman A. Longterm follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 2000;101:948–54. [5] Schachinger V, Britten MB, Zeiher AM. Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation 2000;101:1899–906. [6] Calles-Escandon J, Cipolla M. Diabetes and endothelial dysfunction: a clinical perspective. Endocr Rev 2001;22:36–52. [7] Nyström T, Gutniak MK, Zhang Q, Zhang F, Holst JJ, Ahren B, Sjöholm A. Effects of glucagon-like peptide-1 on endothelial function in type 2 diabetes patients with stable coronary artery disease. Am J Physiol Endocrinol Metab 2004;287:E1209–15. [8] Nyström T, Gonon AT, Sjöholm A, Pernow J. Glucagon-like peptide-1 relaxes rat conduit arteries via an endothelium-independent mechanism. Regul Pept 2005;125:173–7. [9] Basu A, Charkoudian N, Schrage W, Rizza RA, Basu R, Joyner MJ. Beneficial effects of GLP-1 on endothelial function in humans: dampening by glyburide but not by glimepiride. Am J Physiol Endocrinol Metab 2007;293:E1289–95. [10] Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696–705. [11] Blonde L, Klein EJ, Han J, Zhang B, Mac SM, Poon TH, Taylor KL, Trautmann ME, Kim DD, Kendall DM. Interim analysis of the effects of exenatide treatment on A1C, weight and cardiovascular risk factors over 82 weeks in 314 overweight patients with type 2 diabetes. Diabetes Obes Metab 2006;8:436–47. [12] Özyazgan S, Kutluata N, Af S, Ozda SB, Akkan AG. Effect of glucagon-like peptide-1 (7-36) and exendin-4 on the vascular reactivity in streptozotocin/nicotinamideinduced diabetic rats. Pharmacology 2005;74:119–26. [13] Zhang Q, Berggren PO, Larsson O, Hall K, Tally M. Insulin-like growth factor II inhibits glucose-induced insulin exocytosis. Biochem Biophys Res Commun 1998;243:117–21. [14] Harrison DG. Cellular and molecular mechanisms of endothelial cell dysfunction. J Clin Invest 1997;100:2153–7. [15] Tesfamariam B, Cohen RA. Free radicals mediate endothelial cell dysfunction caused by elevated glucose. Am J Physiol 1992;263:H321–6. [16] Cosentino F, Hishikawa K, Katusic ZS, Lüscher TF. High glucose increases nitric oxide synthase expression and superoxide anion generation in human aortic endothelial cells. Circulation 1997;96:25–8.
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13 [17] Halle M, Eriksson P, Tornvall P. Effects of free fatty acids and a triglyceride-rich fat emulsion on endothelial nitric oxide synthase. Eur J Clin Investig 2005;35:154–5. [18] Nyström T. The potential beneficial role of glucagon-like peptide-1 in endothelial dysfunction and heart failure associated with insulin resistance. Horm Metab Res 2008;40:593–606. [19] Gardiner SM, March JE, Kemp PA, Bennett T. Mesenteric vasoconstriction and hindquarters vasodilatation accompany the pressor actions of exendin-4 in conscious rats. J Pharmacol Exp Ther 2006;316:852–9. [20] Ban K, Noyan-Ashraf MH, Hoefer J, Bolz SS, Drucker DJ, Husain M. Cardioprotective and vasodilatory actions of glucagon-like peptide 1 receptor are mediated through both glucagon-like peptide 1 receptor-dependent and -independent pathways. Circulation 2008;117:2340–50. [21] Golpon HA, Puechner A, Welte T, Wichert PV, Feddersen CO. Vasorelaxant effect of glucagon-like peptide-(7-36)amide and amylin on the pulmonary circulation of the rat. Regul Pept 2001;102:81–6. [22] Richter G, Feddersen O, Wagner U, Barth P, Göke R, Göke B. GLP-1 stimulates secretion of macromolecules from airways and relaxes pulmonary artery. Am J Physiol 1993;265:L374–81.
13
[23] Yamamoto H, Lee CE, Marcus JN, Williams TD, Overton JM, Lopez ME, Hollenberg AN, Baggio L, Saper CB, Drucker DJ, Elmquist JK. Glucagon-like peptide-1 receptor stimulation increases blood pressure and heart rate and activates autonomic regulatory neurons. J Clin Invest 2002;110:43–52. [24] Thorens B, Porret A, Buhler L, Deng SP, Morel P, Widmann C. Cloning and functional expression of the human islet GLP-1 receptor. Demonstration that exendin-4 is an agonist and exendin-(9-39) an antagonist of the receptor. Diabetes 1993;42:1678–82. [25] Göke R, Fehmann HC, Linn T, Schmidt H, Krause M, Eng J, Göke B. Exendin-4 is a high potency agonist and truncated exendin-(9-39)-amide an antagonist at the glucagon-like peptide 1-(7-36)-amide receptor of insulin-secreting beta-cells. J Biol Chem 1993;268:19650–5. [26] Fehmann HC, Jiang J, Schweinfurth J, Dorsch K, Wheeler MB, Boyd AE, Göke B. Ligand-specificity of the rat GLP-I receptor recombinantly expressed in Chinese hamster ovary (CHO-) cells. Z Gastroenterol 1994;32:203–7.
Contents lists available at ScienceDirect
Regulatory Peptides j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / r e g p e p
Endothelial dysfunction induced by triglycerides is not restored by exenatide in rat conduit arteries ex vivo David Nathanson a,⁎, Özlem Erdogdu a, John Pernow b, Qimin Zhang a, Thomas Nyström a a b
Karolinska Institutet, Department of Internal Medicine, South Hospital, Sweden Karolinska Institutet, Department of Cardiology, Karolinska Hospital, Stockholm, Sweden
a r t i c l e
i n f o
Article history: Received 11 July 2008 Received in revised form 14 May 2009 Accepted 2 July 2009 Available online 10 July 2009 Keywords: Endothelial function eNOS Exenatide GLP-1 Intralipid
a b s t r a c t Exenatide (synthetic exendin-4) is a stable analogue of glucagon-like peptide 1 (GLP-1) and has recently been approved for clinical use against type 2 diabetes. Exenatide is believed to exert its effects via the GLP-1 receptor with almost the same potency as GLP-1 in terms of lowering blood glucose. Short term exenatide treatment normalizes the altered vascular tone in type 2 diabetic rats, probably due to the reduction in glycemia. The aim of this study was to investigate whether exenatide directly protects against triglyceride-induced endothelial dysfunction in rat femoral arterial rings ex vivo. Short term pre-incubation with Intralipid® (0.5 and 2%) was found to dose-dependently induce endothelial dysfunction, in that it elicited a significant reduction in AChinduced vasorelaxation by 29% and 35%, respectively. Paradoxically, this occurred with a concomitant increase in endothelial nitric oxide synthase (eNOS) activity. No such reduction in vasorelaxation by Intralipid® was seen in response to the NO donor sodium nitroprusside (SNP), revealing an endothelium-dependent vascular dysfunction by Intralipid®. However, exenatide did not protect against Intralipid®-induced endothelial dysfunction. More surprisingly, the maximum vasorelaxation induced by exenatide (without Intralipid®) was only 3 ± 2%, compared to the 23 ± 4%, 38 ± 4%, 79 ± 3% and 97 ± 4% relaxations induced by GLP-1, GLP-1 (9-36), ACh and SNP, respectively. This unexpected finding prompted us to ascertain that the exenatide preparation was biologically active, and both exenatide (10− 11 mol/l) and GLP-1 (10− 9 mol/l) significantly increased insulin secretion in pancreatic β-cells from ob/ob mice in vitro. In conclusion, exenatide could neither confer any acute protective effects against triglyceride-induced endothelial dysfunction nor exert any significant vasorelaxant actions in this model of rat conduit arteries ex vivo. © 2009 Elsevier B.V. All rights reserved.
1. Introduction One salient feature in type 2 diabetes is endothelial dysfunction, which closely correlates to coronary heart disease (CHD) [1]. Also, dyslipidemia, involving hypertriglyceridemia and elevation in circulating free fatty acids (FFAs), is common in type 2 diabetes and impacts the endothelium negatively [2,3]. Endothelial dysfunction is a prognostic factor for the outcome of CHD [4,5]. In attempts to improve endothelial function, numerous studies have evaluated lifestyle and pharmacology interventions to decrease the risk for CHD [1,6]. Beside its well known antihyperglycemic effect, glucagon-like peptide-1 (GLP-1) has been shown to ameliorate endothelial dysfunction in type 2 diabetic patients with CHD [7] and dose-dependently induce vasorelaxation in rat conduit arteries ex vivo [8]. Subsequently, it was also recently demonstrated that GLP-1 per se augments endothelium-dependent vasodilatation in non-diabetic humans [9]. ⁎ Corresponding author. Södersjukhuset, Ringvägen 52, SE-118 83 Stockholm, Sweden. Tel.: +46 86163449; fax: +46 86163146. E-mail address: [email protected] (D. Nathanson). 0167-0115/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2009.07.003
GLP-1 acts as an incretin, i.e., released from intestinal cells following ingestion of food, this peptide lowers blood glucose by augmenting the secretion of insulin and inhibiting glucagon secretion, as well as by inhibiting bowel motility and promoting the feeling of satiety [10]. GLP-1 is degraded rapidly by the ubiquitous enzyme dipeptidyl peptidase IV (DPPIV), as a result of which its half-life in the circulation is no more than 1–2 min [10]. For this reason, both longer-acting GLP-1 analogues and inhibitors of DPPIV have been developed. Since these analogues are believed to exert all of their actions through the GLP-1 receptor (GLP-1R), their biological effects are widely assumed be the same as those of native GLP-1. Long-term improvements of glycemic control by exenatide treatment occur along with reductions in body weight [11]. Interestingly, also lipid parameters, blood pressure, and proinflammatory biomarkers have been shown to improve in patients treated with exenatide, supporting a shift toward a more favorable cardiovascular risk profile that might impact positively on the endothelium [11]. Administration of exenatide to diabetic rats nearly normalizes their vascular tone [12]. However, it is not clear whether these effects merely reflect an improved metabolic state or a direct vascular action of exenatide [12].
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
Therefore, in the present study, we wanted to characterize possible direct effects of exenatide on the vasculature, thereby gaining mechanistic insights into the salutary actions of exenatide on endothelial function. The aim of this study was to investigate whether exenatide protects against endothelial dysfunction, induced by a triglyceride-rich fat emulsion, in rat femoral arterial rings ex vivo. 2. Materials and methods This study was approved by the regional ethics committee for animal research and conforms to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institute of Health (NIH publication No. 85-23, revised 1985). 2.1. Preparation of arterial rings Sixty-nine male Sprague–Dawley rats (weight 250–350 g) were anesthetized with a mixture of fluanisonum and fentanylum (Hypnorm®, Janssen, Beerse, Belgium) and midazolam (Dormicum®, Hoffman–LaRoche, Basel, Switzerland) (2.5, 0.08 and 1.25 mg/kg, respectively, i.m.). The rats were then killed by excision of the heart. The femoral arteries were carefully dissected free from surrounding tissue, removed and put in Krebs–Henseleit (KH) solution [in mmol/l: NaCl 118, KCl 4.7, KH2PO4 1.2, MgSO4(H2O) 1.2, NaHCO3 25.2, CaCl2 2.5 and glucose 11.1]. Circular segments (1–2 mm in length) of the artery were mounted on two thin metal holders, one of which was connected to a force displacement transducer (model FT03, Grass Instrument Co, Quincy, MA) and the other to a movable device that allowed the application of a passive tension of 5 mN. The tension was recorded on a polygraph (model 7B, Grass). The mounted vascular segments were kept in 2 ml organ baths containing KH solution at 37 °C and continuously bubbled with 5% CO2 in O2 to maintain a pH of 7.4. After preparation, the vascular segments were allowed to equilibrate for 60 min. 2.2. Functional experiments The contractile function of the vascular segments was tested by administration of phenylephrine (Phe) (Sigma, final concentration 10− 5 mol/l) and with K+-rich (127 mmol/l) KH solution, prepared by replacing NaCl with equimolar amounts of KCl. Endotheliumdependent and endothelium-independent relaxations were determined by administration of acetylcholine (ACh) (Sigma) and the NO donor sodium nitroprusside (SNP) (Alexis Corporation, Läufelfingen, Switzerland), respectively. ACh and SNP were added to the organ baths at cumulatively increasing concentrations (10− 9–10− 5 mol/l and 10− 9– 10− 6 mol/l, respectively) during a stable contractile tone induced by Phe (10− 5 mol/l). The relaxant response following pre-incubation with a studied substance was always compared to the preceding control response in the same vascular segment. All substances used in the protocol were added in 50 µl volumes. Exenatide (Neosystem, Strasbourg, France) was added to the organ baths at cumulative increasing concentrations (10− 13–10− 8 mol/l) during baseline tension to evaluate contractile effects per se. Furthermore, exenatide (10− 9 mol/l) was in separate experiments added 10 min before a dose–response curve for Phe to test for co-activation of Phe-induced contractions. The relaxant effects of exenatide, GLP-1(736)amide (Neosystem, Strasbourg, France) and GLP-1 (9-36) (Bachem, Bubendorf, Switzerland) were evaluated by adding cumulatively increasing concentrations (10− 13–10− 8 mol/l) of these drugs to artery segments precontracted with Phe (10− 5 mol/l). The relaxant effect by GLP-1 was also tested together with the DPPIV inhibitor sitagliptin (5 µM) which was, in separate experiments, added 5 min before Pheinduced contraction. Sitagliptin was also added during baseline tension and during Phe-induced contraction to evaluate any contractile and vasorelaxant effects of the DPPIV inhibitor per se. Finally, to investigate whether exenatide might co-activate the ACh relaxation, exenatide
9
(2.5 nmol/l) was added in separate experiments to the organ baths 10 min before contraction with Phe (10− 5 mol/l). Thereafter the vessels were relaxed by ACh in increasing concentrations (10− 9– 10− 5 mol/l). To induce endothelial dysfunction, the vascular segments were preincubated for 20 min with Intralipid® [100 mg/ml] (PharmaciaUpjohn, Uppsala, Sweden) diluted in KH solution to final concentrations of 0.5 and 1%, corresponding to an approximate triglyceride level in vivo of 5 and 10 mmol/l, respectively. To examine any protective effect of exenatide against the triglyceride-induced endothelial dysfunction, exenatide (final concentration of 2.5 nmol/l) was administered to the organ baths 20 min before the vascular response test of ACh and SNP. After the end of the protocols, artery rings were frozen and stored at − 80 °C for subsequent immunoblotting analysis. For determination of cAMP contents, the thoracic aorta was removed, cleaned and cut into two sections. Each vessel segment was equilibrated for 20 min in KH at 37 °C and bubbled with 5% CO2 in O2 to maintain a pH of 7.4, before incubation. One vessel segment was then incubated for 15 min with GLP-1 (0.1 µM) or exenatide (2.5 nM), respectively, and the other with an equivalent volume of the adenylyl cyclase activator forskolin (10 µM) or DMSO (10 µM), serving as positive control. Vessels were then blotted dry and frozen in liquid nitrogen and stored at −80 °C for further analysis. 2.3. Immunoblotting Frozen artery rings were thawed and homogenized in 100 µl buffer containing 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 1% TritonX-100, 10% glycerol, 20 mM Tris (pH 7.8), 1 mM EDTA with phosphatase inhibitors 0.5 mM Na3VO4, 1 µg/ml leupeptin, 0.2 mM PMSF, 10 mM NaF, 1 µg/ml aprotinin, 5 mM Na-pyrophosphate and 100 mM benzamidine. The tissue was minced and homogenized and incubated on ice for 30 min and centrifuged (5000 g, 5 min). Protein concentrations were determined by the bicinchoninic acid method (Pierce Chemical Co., Rockford, IL). Phosphorylation of the enzyme endothelial nitric oxide synthase (eNOS) was examined by anti-phospho specific antibody as a measure of eNOS activity (Santa Cruz). Expression levels of the GLP-1 receptor were examined with 2 µg/ml rabbit anti-NH2-terminal GLP-1 receptor antibody (a generous gift of Dr. Bernard Thorens). 2.4. Determinations of cAMP For determination of cAMP concentrations, frozen aortae were weighed and homogenized (1500 rpm) using a reusable pellet pestle (Kimble-Kontes, Vineland, New Jersey, USA) fixed into an overhead stirrer (SS10, Stuart Scientific, Surrey, UK). Homogenates were suspended in trichloroacetic acid (5% (w/v); 500 µl/100 mg of tissue), centrifuged (1500 g, 10 min) and the supernatant was removed. Trichloroacetic acid was extracted from samples using water-saturated diethyl ether before total cAMP contents were measured using an enzyme immunoassay kit (Cayman Europe, Tallinn, Estonia). 2.5. Insulin secretion from pancreatic β-cells Pancreatic islets were isolated from 12 month-old ob/ob mice and dispersed into cells in a buffer containing EDTA (2-[2-(Bis(carboxymethyl)amino)ethyl-(carboxymethyl)amino]acetic acid) [13]. The cells were allowed to recover in RPMI-1640 medium containing 10% FCS overnight. Cells were washed in KRBH buffer containing 135 mM NaCl, 3.6 mM KCl, 5 mM NaHCO3, 0.5 mM NaH2PO4, 0.5 mM MgCl2, 1.5 mM CaCl2, 10 mM HEPES pH 7.4 in the presence of 0.1% BSA. Equal amounts of cells were incubated at 37 °C with exenatide or GLP-1 for 20 min. At the end of the incubation, cells were spun down and the insulin concentration in the supernatants was analyzed using mouse insulin ELISA kits (Mercodia, Uppsala, Sweden).
10
2.6. Binding of
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13 125
I-GLP-1
Human coronary artery endothelial cells (HCAECs; Clonetics) or mouse insulinoma cells (MIN6) were collected and washed three times in KRBH buffer. Cells were re-suspended in KRBH buffer containing 1% BSA. Incubation was performed for 30 min at 37 °C (or 4 °C overnight) in the presence of 125I-GLP-1 (100 pmol/tube) and in the presence or absence of increasing concentrations of unlabeled GLP-1 (0.01 nM– 1000 nM) with 2 × 105 cells/tube. Incubation was terminated by aspirating 2 × 100 µl of cell suspensions from each assay tube, followed by centrifugation through 250 µl of a mixture of phthalic acid esters. The supernatants were eliminated and the tips of the tubes containing cell pellets were excised and counted in a gamma counter [13]. The 100 µl remaining in each assay tube was used as reference. 2.7. Statistics The relaxations induced by ACh were expressed as a percentage of the precontracted tone. The precontracted level was set to 0%, and the baseline level, corresponding to maximal relaxation, to 100%. All values were expressed as means ± S.E.M. Friedman's test (ANOVA) was used to evaluate effects between related samples (conditions and concentration). Differences were deemed to be statistically significant at P b 0.05. 3. Results All vessels tested responded well to ACh and the NO donor SNP in a dose-dependent manner, demonstrating a valid working model for studies of endothelium-dependent and -independent relaxation effects (Fig. 1a). Exenatide, applied during basal tension, was without any discernable contractile effect on the femoral artery rings (data not shown). Moreover, pre-incubation with exenatide did not co-activate
Fig. 1. Differential vasorelaxant responses of arterial rings ex vivo. Vasorelaxant function of femoral artery rings from Sprague–Dawley rats after pre-constriction with phenylephrine (10− 5 M) a) Relaxations induced by acetylcholine (ACh) and by the NO donor sodium nitroprusside (SNP) (n = 14). b) Relaxations induced by control (Krebs-solution) (n = 4), exenatide (n = 16), GLP-1 (n = 7), GLP-1 + sitagliptin (n = 4) and GLP-1 (9-36) (n = 6) ⁎denotes P b 0.01 for a chance difference between exenatide vs GLP-1, GLP-1 + sitagliptin and GLP-1 (9-36), respectively. †Denotes P b 0.05 for a chance difference between GLP-1 (9-36) vs GLP-1 + sitagliptin and GLP-1, respectively. ‡ Denotes P b 0.05 for a chance difference between GLP-1 (9-36) vs GLP-1. Means are compared by ANOVA. Results are presented as mean ± S.E.M.
the ACh-induced vasodilatation, or the Phe-induced contractile tone (data not shown). More surprisingly, exenatide was without any discernable relaxant effects during Phe-induced contractile tone compared to GLP-1 and GLP-1(9-36), which demonstrated a significant dose-dependent vascular relaxation (Fig. 1b). However, the maximal relaxation obtained with the highest concentration of GLP-1 and GLP-1 (9-36) was 23 ± 4% and 38 ± 4%, respectively (Fig.1b), compared to 79 ± 3% and 97 ± 4% relaxation induced by the controls, ACh and SNP, respectively (Fig. 1a). Pre-incubation with Intralipid® resulted in a concentrationdependent inhibition of ACh-induced endothelium-dependent relaxation (Fig. 2a–b). A significant reduction in ACh-induced relaxation was obtained already with 0.5% Intralipid® (Fig. 2a). The highest concentration of Intralipid® (2%) caused a rightward shift of the concentration– response curve for ACh by approximately one order of magnitude (Fig. 2b). The relaxation induced by the highest concentration of Ach (10− 7 mol/l) was 71 ± 8% in the presence of 0.5% Intralipid® vs 93 ± 5% in controls (Fig. 2a), and in the presence of 2% Intralipid® 53± 6% vs 87 ± 2% in controls (Fig. 2b). Pre-incubation with Intralipid® (0.5 or 2%) did not affect endothelium-independent relaxation evoked by the NO-donor SNP (Fig. 3a–b), demonstrating an endothelialdependent dysfunction effect by the triglyceride-rich fat emulsion. Short term exenatide exposure did not protect against the triglyceriderich fat emulsion-induced endothelial dysfunction (Fig. 2). Immunoblotting data revealed a single band with a molecular mass of 67 kDa, corresponding to the GLP-1 receptor seen in MIN6 cells, demonstrating expression of the GLP-1 receptor on the femoral artery, albeit with less abundance compared to the insulin-secreting cells. Immunoblotting data also revealed a significant increase in eNOS activation by Intralipid® incubation at both 0.5 and 2% (Fig. 4). This activation was not significantly influenced after pre-incubation with exenatide or by exenatide itself (Fig. 4). However, there was a clear tendency towards a decreased Intralipid®-induced eNOS activation by co-incubation with exenatide (Fig. 4).
Fig. 2. Endothelial dysfunction evoked by Intralipid® is not reversed by exenatide. Relaxations induced by acetylcholine (ACh) under control conditions and after a 20-min pre-incubation with exenatide (2.5 nmol/l) in the presence of a) Intralipid® 0.5% (n = 10) and b) Intralipid® 2% (n = 8). ⁎ and † denote P b 0.05 and P b 0.01 for chance differences vs controls. Results are presented as mean ± S.E.M.
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
11
The somewhat unexpected negative findings with exenatide prompted us to ascertain that the exenatide preparation used was indeed biologically active. To this end, we performed insulin secretion experiments in vitro with pancreatic β-cells isolated from ob/ob mice, and found a similar enhancement of insulin secretion evoked by exenatide as with equipotent concentrations of GLP-1 (Fig. 5). Binding of 125I-GLP-1 to human coronary artery endothelial cells in vitro was studied in the presence of increasing concentrations of unlabeled GLP-1, and compared to the insulin-secreting cell line MIN6. Incubation of MIN6 cells with 125I-GLP-1 resulted in approximately 30% specific binding, consistent with high GLP-1 receptor expression. In contrast, the endothelial cells showed no specific binding of 125 I-GLP-1 (not shown). 4. Discussion
Fig. 3. Intralipid® does not impair endothelium-independent vasorelaxation. Vascular relaxations induced by the nitric oxide donor sodium nitroprusside (SNP) under control conditions and after a 20-min pre-incubation with exenatide (2.5 nmol/l) in the presence of a) Intralipid® 0.5% (n = 10) and b) Intralipid® 2% (n = 8). ⁎ and † denote P b 0.05 and P b 0.01 for chance differences vs controls. Results are presented as mean± S.E.M.
GLP-1, exenatide and forskolin did all increase the cAMP levels of the tissue, however, no differences in the cAMP levels was observed (78.0 ± 16.3, 56.0 ± 11.5 and 51.3 ± 25.3 pmol/l (mg tissue)− 1, N.S.).
Fig. 4. a) Expression of GLP-1 receptor in rat femoral artery and in mouse insulinoma cells (MIN6). b) Intralipid® activates eNOS. Western blot analysis of expression of eNOS phosphorylated at Ser-1177 in rat aortic segments after 20-min incubation of Intralipid® at 0.5% (IL0.5%) or 2% (IL2%), with/out pre-incubation with 2.5 nM exenatide (EX; n = 6 and n = 6, respectively). P-eNOS, phosphorylated (activated) eNOS. ⁎ Denotes P b 0.05 for a chance difference vs control. Bars represent mean± S.E.M with densitometric readings of P-eNOS normalized to those of beta-actin.
This study demonstrates that the triglyceride-rich fat emulsion Intralipid® acutely and dose-dependently induces endothelial-dependent dysfunction, with a concomitant increase in eNOS activity, in rat conduit artery rings ex vivo. However, no protection against such dysfunction was afforded by exenatide. Even more surprisingly, exenatide did not exert any discernable vasorelaxant effects, compared to GLP-1 or GLP-1 (9-36), in this system. Moreover, the fat emulsion did not impact vascular relaxation elicited by the NO donor nitroprusside (reflecting endothelium-independent relaxation), suggesting that the triglyceride-rich fat emulsion only affects the endothelium-dependent pathway of vasorelaxation. Endothelial dysfunction is a major feature of the early stages of the atherosclerotic process and can be used to predict the outcome of CVD in man [4,5]. Although the molecular causes of endothelial dysfunction have not yet been elucidated in detail, numerous studies indicate that impaired signaling by NO and/or attenuated biosynthesis of this messenger play a central role [14]. In the presence of suboptimal concentrations of substrate or cofactors for the synthesis of NO, eNOS might be uncoupled and produce reactive oxygen species (ROS), such as superoxide anion and hydrogen peroxide, thereby giving rise to oxidative stress. Normally, ROS are scavenged via multiple intra- and extracellular defense mechanisms. However, high concentrations of ROS can exceed the capacity of these systems and quench NO, thereby exacerbating the oxidative stress [15]. In the present study, eNOS activity was substantially increased by short-term exposure to a triglyceriderich fat emulsion ex vivo, concomitant with a significant functional impairment in endothelium-dependent vasodilatation. We can only speculate whether this functional impairment might be a consequence of an increase in ROS production, because we did not measure ROS production. However, it has earlier been demonstrated that hyperglycemia-induced endothelial dysfunction is associated with a heightened eNOS activity due to an uncoupling of the enzyme, thereby inducing oxidative stress [16]. Also, it was recently reported that short-term exposure to triglyceride-rich fat emulsion increases eNOS activity in an
Fig. 5. Exenatide and GLP-1 stimulate insulin release. Insulin secretion from ob/ob mice islet cells after exposure to 10 mM glucose (control) with/out exenatide (EX) at 1, 10, or 20 nM; or GLP-1 at 100 nM. ⁎ Denotes P b 0.05 for a chance difference vs control. Bars represent mean percentage of control ± S.E.M for three independent experiments.
12
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13
endothelium-derived cell line [17]. However, in that study it was not addressed whether this increased eNOS activity affected the endothelium functionally [17]. In the present study, although the enhanced eNOS activity induced by Intralipid® tended to decrease after co-exposure to exenatide, it did not restore the endothelial dysfunction evoked by the fat emulsion. Additionally, exenatide alone did not influence eNOS activity in this setting. At least in rat artery rings ex vivo, it seems that exenatide acutely does not affect the key enzyme in NO production regulating vascular tone. It is becoming increasingly clear that GLP-1 and its derivatives, apart from their well characterized glycemic actions, also have certain vascular effects [18]. The GLP-1 mimetic exenatide, besides HbA1c reduction, also shows beneficial effects on several cardiovascular risk factors, i.e., lowering of serum triglycerides and blood pressure. Some studies have even suggested direct vascular effects of exenatide per se [12,19]. However, in the present study, we were not able to demonstrate any direct protective effect of exenatide on endothelial dysfunction induced by a triglyceriderich fat emulsion. More surprisingly, exenatide did not produce any vasorelaxant effects on the conduit vessel. It could be questioned why we used just one single dose, and not a dose response curve of exenatide, to prove our hypothesis. However, we believe that the concentration of exenatide we used in this protocol is equipotent, or even higher, than that obtained in the clinical situation. We also changed the Intralipid® concentrations to resemble a clinical situation in which moderate to high triglyceride concentrations prevail. This caused a rightward shift of the concentration–response curve for ACh by approximately one order of magnitude, indicating a dose–response impairment in endothelial function, but without any protective effects afforded by exenatide. Since our aim was to investigate whether the clinically used drug exenatide protects against triglyceride-induced endothelial dysfunction, we did not include native GLP-1 in the protocol. However, after obtaining the somewhat unexpected result that exenatide failed to exert any vasorelaxant effects in the rat conduit artery, we wanted to compare the vasodilatation properties between exenatide vs GLP-1 and GLP-1(9-36). The latter two dose-dependently relaxed rat the arteries, confirming what we and others have demonstrated previously [8,20–22]. To ascertain that the batch of exenatide we used was active, we also investigated its insulin-releasing capacity in pancreatic β-cells, demonstrating a dose-dependent stimulatory effect. There may be several conceivable explanations as to why exenatide did not produce vasorelaxation in this system, while GLP-1 and GLP-1 (9-36) did. First, most studies reporting vasoactive effects by exenatide were conducted in vivo, in which any direct vasoactive influence might be confounded by other factors (e.g. changes in glycemia) [12,19,23]. Very recently it was reported that exenatide does not evoke any vasorelaxant action in a mouse model [20], thus lending support to our findings. Second, even if ligand-binding analyses in other tissues have not revealed any major differences between exenatide and native GLP-1 in terms of GLP-1 receptor affinity [24,25], [26], we cannot rule out that the affinity to the GLP-1 receptor on endothelial cells might differ between exenatide and GLP-1. In an attempt to investigate the GLP-1 receptor binding capacity of human coronary artery endothelial cells in vitro (see Results section), more specifically aiming at comparing receptor affinity between exenatide and GLP-1, we were unable to detect any specific binding of radiolabeled GLP-1 to these cells compared to insulin-secreting MIN6 cells. Third, as recently demonstrated, the vasorelaxant effects of GLP-1 might not be mediated through occupancy of its receptor [20]. This has actually been demonstrated, in as much as its metabolite GLP-1 (9-36) relaxes the artery independent of the GLP-1 receptor [20]. Fourth, there is only a ~50% homology in the amino-acid structure between exenatide and GLP-1, which might explain the differences in vascular effects observed. The effects on glycemia seem to be equal between GLP-1 and exenatide, and most effects of exenatide on glycemia-related endpoints are believed (or assumed) to be conveyed via its signaling through the GLP-1 receptor, which also was demonstrated in present study finding no differences in cAMP productions
between GLP-1 and exenatide. However, we cannot rule out an alternative scenario in terms of vascular responses to these two peptides. In conclusion, the stable GLP-1 analogue exenatide does not acutely ameliorate endothelial dysfunction induced by a triglyceride-rich emulsion in this ex vivo model of rat conduit artery rings. Further in contrast to native GLP-1, and its metabolite GLP-1 (9-36), exenatide surprisingly does not exert any vasorelaxant effects in this system. To date, research on GLP-1 and its derivatives has focused primarily on hyperglycemia endpoints and elucidation of its cardiovascular effects is still in its infancy. As our findings are restricted to acute effects ex vivo on non-diabetic rat conduit arteries, we do not know which effects, if any, chronic exenatide treatment confers to vascular function in vivo in type 2 diabetic patients. Thus, and in the light of the frequency and severity of premature CVD in type 2 diabetes, we still believe that this track should be intensely pursued in future research efforts. Acknowledgments We thank Professor Åke Sjöholm for the critical reading of the manuscript, and Professor Jon Lundberg for allowing us to occupy his laboratory. We thank Margareta Stensdotter and Adrian Gonon for the excellent technical support. Financial support was received from the Swedish Society for Medical Research, the Swedish Society of Medicine, the European Foundation for the Study of Diabetes, Olle Engkvist Byggmästare Foundation, Åke Wibergs Foundation, Loo and Hans Foundation, Tore Nilsson Foundation, Fredrik and Ingrid Thurings Foundation, Merck Sharp & Dohme (Medical School Grant) and the Swedish Research Council (# 10857, 72X-12550, 72X-14507, and 72P-14787). References [1] Caballero AE. Endothelial dysfunction in obesity and insulin resistance: a road to diabetes and heart disease. Obes Res 2003;11:1278–89. [2] Lundman P, Eriksson M, Schenck-Gustafsson K, Karpe F, Tornvall P. Transient triglyceridemia decreases vascular reactivity in young, healthy men without risk factors for coronary heart disease. Circulation 1997;96:3266–8. [3] Steinberg HO, Tarshoby M, Monestel R, Hook G, Cronin J, Johnson A, Bayazeed B, Baron AD. Elevated circulating free fatty acid levels impair endotheliumdependent vasodilation. J Clin Invest 1997;100:1230–9. [4] Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes Jr DR, Lerman A. Longterm follow-up of patients with mild coronary artery disease and endothelial dysfunction. Circulation 2000;101:948–54. [5] Schachinger V, Britten MB, Zeiher AM. Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation 2000;101:1899–906. [6] Calles-Escandon J, Cipolla M. Diabetes and endothelial dysfunction: a clinical perspective. Endocr Rev 2001;22:36–52. [7] Nyström T, Gutniak MK, Zhang Q, Zhang F, Holst JJ, Ahren B, Sjöholm A. Effects of glucagon-like peptide-1 on endothelial function in type 2 diabetes patients with stable coronary artery disease. Am J Physiol Endocrinol Metab 2004;287:E1209–15. [8] Nyström T, Gonon AT, Sjöholm A, Pernow J. Glucagon-like peptide-1 relaxes rat conduit arteries via an endothelium-independent mechanism. Regul Pept 2005;125:173–7. [9] Basu A, Charkoudian N, Schrage W, Rizza RA, Basu R, Joyner MJ. Beneficial effects of GLP-1 on endothelial function in humans: dampening by glyburide but not by glimepiride. Am J Physiol Endocrinol Metab 2007;293:E1289–95. [10] Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet 2006;368:1696–705. [11] Blonde L, Klein EJ, Han J, Zhang B, Mac SM, Poon TH, Taylor KL, Trautmann ME, Kim DD, Kendall DM. Interim analysis of the effects of exenatide treatment on A1C, weight and cardiovascular risk factors over 82 weeks in 314 overweight patients with type 2 diabetes. Diabetes Obes Metab 2006;8:436–47. [12] Özyazgan S, Kutluata N, Af S, Ozda SB, Akkan AG. Effect of glucagon-like peptide-1 (7-36) and exendin-4 on the vascular reactivity in streptozotocin/nicotinamideinduced diabetic rats. Pharmacology 2005;74:119–26. [13] Zhang Q, Berggren PO, Larsson O, Hall K, Tally M. Insulin-like growth factor II inhibits glucose-induced insulin exocytosis. Biochem Biophys Res Commun 1998;243:117–21. [14] Harrison DG. Cellular and molecular mechanisms of endothelial cell dysfunction. J Clin Invest 1997;100:2153–7. [15] Tesfamariam B, Cohen RA. Free radicals mediate endothelial cell dysfunction caused by elevated glucose. Am J Physiol 1992;263:H321–6. [16] Cosentino F, Hishikawa K, Katusic ZS, Lüscher TF. High glucose increases nitric oxide synthase expression and superoxide anion generation in human aortic endothelial cells. Circulation 1997;96:25–8.
D. Nathanson et al. / Regulatory Peptides 157 (2009) 8–13 [17] Halle M, Eriksson P, Tornvall P. Effects of free fatty acids and a triglyceride-rich fat emulsion on endothelial nitric oxide synthase. Eur J Clin Investig 2005;35:154–5. [18] Nyström T. The potential beneficial role of glucagon-like peptide-1 in endothelial dysfunction and heart failure associated with insulin resistance. Horm Metab Res 2008;40:593–606. [19] Gardiner SM, March JE, Kemp PA, Bennett T. Mesenteric vasoconstriction and hindquarters vasodilatation accompany the pressor actions of exendin-4 in conscious rats. J Pharmacol Exp Ther 2006;316:852–9. [20] Ban K, Noyan-Ashraf MH, Hoefer J, Bolz SS, Drucker DJ, Husain M. Cardioprotective and vasodilatory actions of glucagon-like peptide 1 receptor are mediated through both glucagon-like peptide 1 receptor-dependent and -independent pathways. Circulation 2008;117:2340–50. [21] Golpon HA, Puechner A, Welte T, Wichert PV, Feddersen CO. Vasorelaxant effect of glucagon-like peptide-(7-36)amide and amylin on the pulmonary circulation of the rat. Regul Pept 2001;102:81–6. [22] Richter G, Feddersen O, Wagner U, Barth P, Göke R, Göke B. GLP-1 stimulates secretion of macromolecules from airways and relaxes pulmonary artery. Am J Physiol 1993;265:L374–81.
13
[23] Yamamoto H, Lee CE, Marcus JN, Williams TD, Overton JM, Lopez ME, Hollenberg AN, Baggio L, Saper CB, Drucker DJ, Elmquist JK. Glucagon-like peptide-1 receptor stimulation increases blood pressure and heart rate and activates autonomic regulatory neurons. J Clin Invest 2002;110:43–52. [24] Thorens B, Porret A, Buhler L, Deng SP, Morel P, Widmann C. Cloning and functional expression of the human islet GLP-1 receptor. Demonstration that exendin-4 is an agonist and exendin-(9-39) an antagonist of the receptor. Diabetes 1993;42:1678–82. [25] Göke R, Fehmann HC, Linn T, Schmidt H, Krause M, Eng J, Göke B. Exendin-4 is a high potency agonist and truncated exendin-(9-39)-amide an antagonist at the glucagon-like peptide 1-(7-36)-amide receptor of insulin-secreting beta-cells. J Biol Chem 1993;268:19650–5. [26] Fehmann HC, Jiang J, Schweinfurth J, Dorsch K, Wheeler MB, Boyd AE, Göke B. Ligand-specificity of the rat GLP-I receptor recombinantly expressed in Chinese hamster ovary (CHO-) cells. Z Gastroenterol 1994;32:203–7.