Background and purpose:
Diabetes-associated vascular dysfunction contributes to increased cardiovascular risk. We investigated whether the phosphodiesterase-5 inhibitor sildenafil would improve vascular function in diabetic rats.
Diabetes-associated vascular dysfunction contributes to increased cardiovascular risk. We investigated whether the phosphodiesterase-5 inhibitor sildenafil would improve vascular function in diabetic rats.
Male Wistar rats were injected with streptozotocin (50 mg kg-1, i.v.) to induce insulin-deficient diabetes. Direct effects of sildenafil as well as modification of endothelium-dependent and -independent vasorelaxation were investigated in vitro. The effects of acute and chronic (2 week) treatment in vivo of sildenafil on vascular function were also characterized in isolated aortic segments in organ bath chambers 4 weeks after diabetes induction.
Sildenafil induced a concentration-dependent vasorelaxation, which was attenuated by the nitric oxide (NO) synthase inhibitor, NG-nitro-L-arginine. Acetylcholine-induced endothelium-dependent as well as endothelium-independent relaxation induced by the NO donor, DEA-NONOate, was significantly reduced in aortae from diabetic rats. Incubation with sildenafil in vitro normalized both endothelium-dependent and -independent relaxation in aortae from diabetic rats. Acute as well as chronic in vivo treatment with sildenafil resulted in enhanced endothelium-dependent and -independent vasorelaxation. Superoxide formation was increased in diabetes, associated with enhanced membrane expression of the NAD(P)H oxidase subunit gp91phox and Rac, which were both reduced by chronic treatment with sildenafil.
We demonstrate that sildenafil treatment rapidly and chronically improves vascular relaxation in diabetic rats. Treatment with sildenafil might provide a similarly beneficial effect in diabetic patients.
British Journal of Pharmacology (2008) 153, 886–893; doi:10.1038/sj.bjp.0707459; published online 24 September 2007
reactive oxygen species
Diabetes is strongly associated with various cardiovascular diseases, which are the primary cause of morbidity and mortality among patients with diabetes, accounting for more than 80% of deaths (Tschoepe and Menart-Houtermans, 2002). Diabetes alone confers a long-term cardiovascular risk similar to that observed among non-diabetic patients with prior myocardial infarction (Haffner et al., 1998). Patients with diabetes have early development of abnormal endothelial function, aggressive atherosclerosis and adverse arterial remodelling (Tschoepe and Menart-Houtermans, 2002). The risk for cardiovascular disease is already substantially elevated before diagnosis of diabetes (Hu et al., 2002) and the progression of atherosclerosis is accelerated in diabetics (Wagenknecht et al., 2003). Reduced nitric oxide (NO) bioavailability and abundant formation of reactive oxygen species (ROS) within the vascular wall are the key determinants in endothelial dysfunction resulting in an imbalance between NO and ROS. Impaired endothelial function has been described in very early stages of diabetes mellitus and hyperglycaemia, and decreased insulin-sensitivity as well as increased oxidative stress have been proposed as possible contributors (reviewed by Guerci et al., 2001). The increased amount of ROS and the lack of NO are deleterious to smooth muscle cell function.
Endothelium-derived NO activates soluble guanylyl cyclase in smooth muscle cells resulting in enhanced cGMP concentrations and vasorelaxation. Sildenafil is a potent and selective inhibitor of phosphodiesterase (PDE)-5, and thus decreases hydrolysis of cGMP. Sildenafil enhances NO/cGMP-mediated signalling in the corpus cavernosum and is approved for treatment of erectile dysfunction. As already noted for endothelial dysfunction, erectile dysfunction is common in diabetic patients and both disorders are clearly linked. Inhibiting cGMP degradation by sildenafil might be a rational approach to treat patients with diabetes, coronary artery disease or heart failure (Gross, 2005). Sildenafil dilates epicardial coronary arteries, improves endothelial dysfunction and inhibits platelet activation in patients with coronary artery disease (Halcox et al., 2002), and acutely enhances flow-mediated vasodilation in patients with heart failure (Hryniewicz et al., 2005). Similarly, PDE-5 inhibition with sildenafil prevents the smoking-induced decrease in flow-mediated vasodilation and increases impaired NO bioactivity (Kimura et al., 2003; Vlachopoulos et al., 2004).
In the present study, we demonstrate that sildenafil improves NO-mediated vasorelaxation in vitro as well as following acute and chronic treatment in vivo, in an experimental model of diabetes, whereby chronic treatment significantly reduces oxidative stress.
These investigations conform to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996).
Male Wistar rats (250–300 g, obtained from Harlan-Winkelmann, Borchen, Germany) were housed in temperature-controlled cages (20–22 °C) with a 12-h light–dark cycle, and given free access to water and formulated diets.
A single dose streptozotocin (STZ) regimen was used to induce pancreatic islet cell destruction and persistent hyperglycaemia. STZ (10 mg ml−1, Sigma, Deisenhofen, Germany) was freshly dissolved in sterile sodium citrate buffer (25 mM, pH 4.5) and used within 10 min. Rats received a single 50 mg kg−1 intravenous injection of STZ or citrate buffer (control). Blood glucose was monitored using a one-touch blood glucose meter (Ascensia Elite, Bayer-Vital GmbH, Leverkusen, Germany). Hyperglycaemia was defined as a random blood glucose level >20 mM at 2 and 4 weeks after injection. Diabetic rats were randomized to placebo or sildenafil (5 mg kg−1 day−1 by gavage; Pfizer) at day 14. Two weeks later, vasomotor function was measured to assess chronic effects of sildenafil. The last dose of sildenafil was given on the day before the experiments were performed. Acute in vivo effects of sildenafil were investigated in diabetic animals 4 weeks after STZ injection 2 h following one single application of the study drug. In vitro effects of sildenafil were evaluated in aortae from 4-week diabetic as well as healthy control rats in separate experiments.
The descending thoracic aorta was dissected under deep anaesthesia induced by isoflurane following removal of the heart and cleaned of connective tissue. One section was used for measurement of superoxide production, while the other was cut into 3 mm rings, which were mounted in an organ bath (FMI, Seeheim, Germany) for isometric force measurements. The rings were equilibrated for 30 min under a resting tension of 2 g in oxygenated (95% O2; 5% CO2) Krebs–Henseleit solution (NaCl 118 mM, KCl 4.7 mM, MgSO4 1.2 mM, CaCl2 1.6 mM, KH2PO4 1.2 mM, NaHCO3 25 mM, glucose 12 mM; pH 7.4, 37 °C) containing diclofenac (1 μM). All vascular reactivity studies were performed in vessels with endothelium. Rings were repeatedly contracted by KCl (with a maximum of 100 mM) until reproducible responses were obtained. The relaxant response to cumulative concentrations of acetylcholine was assessed after preconstriction with phenylephrine (0.3–3.0 μM) to comparable levels (in g: control: 1.93±0.04, STZ-placebo: 1.89±0.06, STZ-sildenafil: 1.85±0.06). Furthermore, relaxant responses to the endothelium-independent vasodilator 2-(N,N-diethylamino)-diazenolate-2-oxide (DEA-NONOate; Alexis Biochemicals, San Diego, CA, USA) were determined after preconstriction with phenylephrine (in g: control: 2.05±0.06, STZ-placebo: 2.13±0.05, STZ-sildenafil: 2.11±0.06) in the presence of NG-nitro-L-arginine (L-NNA, 100 nM).
In separate in vitro experiments, the effect of cumulative concentrations of sildenafil as well as preincubation with sildenafil (100 nM, 5 min) prior to endothelium-dependent and -independent vasorelaxation in preconstricted aortic rings was tested.
Vascular superoxide formation was measured using lucigenin-enhanced chemiluminescence (Schäfer et al., 2003). The light reaction between superoxide and lucigenin (5 μM) was detected in a luminometer (Wallac, Freiburg, Germany) during incubation of rings in a HEPES-modified Krebs buffer (pH 7.40).
The oxidative fluorescent dye, hydroethidine, was used to evaluate in situ production of superoxide as described previously (Schäfer et al., 2003). Unfixed frozen ring segments were cut into 10-μm-thick sections and placed on a glass slide. Hydroethidine (2 μM) was topically applied to each tissue section and coverslipped. Slides were incubated in a light-protected humidified chamber at 37 °C for 30 min.
Stained sections were investigated using a Nikon Eclipse E600 microscope equipped with a C1 confocal scanning head and a 20-fold oil immersion objective. Pictures were acquired and prepared for presentation using the EZ-C1 3.00 software from Nikon. Image analysis was performed using Image J (Rasband WS, ImageJ, US National Institutes of Health, Bethesda, MD, USA, http://rsb.info.nih.gov/ij/, 1997–2006.). Aortic rings from diabetic animals and control tissues were processed and imaged in parallel. Microscope settings and laser intensities were identical for acquisition of images from diabetic and control specimens.
Aorta samples with endothelium were homogenized in ice-cold Tris-buffer (30 mM Tris-HCl, pH 7.5, 10 mM EGTA, 5 mM EDTA, 1 mM dithiothreitol and 250 mM sucrose). The homogenates were centrifuged at 8000 g for 10 min at 4 °C. To obtain membranous fractions, the supernatants were then centrifuged at 120 000 g for 45 min at 4 °C. The resulting pellets were resuspended in Tris-buffer and solubilized with 1% Triton X-100 and 1% sodium cholate. Membranous extracts were mixed with sample loading buffer (B7703, BioLabs, Frankfurt, Germany) and separated on 10% sodium dodecyl sulphate (SDS)-polyacrylamide gel under reducing conditions. Proteins were electrotransferred onto polyvinylidine difluoride membrane (Immun-Blot 0.2 μm, Bio-Rad, Munich, Germany). The bands were detected using chemiluminescence assay (ECL+Plus, Amersham, Munich, Germany). Primary antibodies used recognize: gp91phox (611414, BD Biosciences Pharmingen, Heidelberg, Germany), p47phox (610355, BD Bioscience Pharmingen), Rac (05–389, Upstate antibodies) and β-actin (4967, Cell Signaling Technology, Frankfurt, Germany).
Aorta extracts were mixed with 3 × SDS sample buffer (187.5 mM Tris-HCl (pH 6.8), 6% (w/v) SDS, 30% glycerol and 0.03% (w/v) bromophenol blue and 15% v/v 2-mercaptoethanol) at 0 °C. Samples were loaded on 7.5% polyacrylamide gels and subjected to electrophoresis. Gels and buffers were cooled to 4 °C prior to electrophoresis and the buffer tank placed in an ice-bath during electrophoresis. Endothelial NO synthase (eNOS) dimer/monomer protein was detected by western blot analysis using an anti-eNOS antibody (N-30020, Transduction Laboratories, Lexington, KY, USA).
For immunohistochemical analysis, frozen 5 μm sections of aorta were stained using primary antibody against CD68 (MCA341R, Serotec, Disseldorf, Germany). Briefly, sections were fixed in cold acetone for 5 min followed by pretreatment with 0.3% hydrogen peroxide for 20 min to inhibit endogenous peroxidase activity. Subsequently, sections were blocked with 2% goat serum for 30 min and incubated with the primary antibody for 1 h at room temperature. After rinsing with PBS, the sections were incubated for 30 min with a biotinylated secondary antibody (550337, BD Biosciences Pharmingen), followed by incubation with Streptavidin-horseradish peroxidase (550946, BD Biosciences Pharmingen) and diaminobenzidine (SK4100, Vector Linaris, Wertheim-Bettingen, Germany). Sections were counterstained with eosin.
Values shown are means±s.e.mean for curves and bar graphs. Relaxant responses are given as percentage relaxation relative to the preconstriction level. Statistical analysis was performed using Prism 4 (GraphPad Software, San Diego, CA, USA) by repeated measures analysis of variance followed by Tukey–Kramer multiple comparisons test. Superoxide formation was analysed by analysis of variance followed by a Tukey post hoc test where appropriate; P<0.05 was considered statistically significant.
Unless otherwise stated, all chemicals were obtained from Sigma (Deisenhofen, Germany) in the highest purity available.
Diabetes-induced increases in blood glucose (control: 10.5±1.9 mM, STZ-placebo: 32.0±0.6 mM, n=15 per group, P<0.01) were unaffected by sildenafil treatment of diabetic animals (32.2±0.7 mM), while the reduction in body weight (control: 336±16 g, STZ-placebo: 252±9 g, n=15 per group, P<0.01) was slightly further lowered following treatment with sildenafil (204±6 g, n=15, P<0.05).
A concentration response to sildenafil in vitro was obtained in isolated preconstricted aortic rings from diabetic and non-diabetic rats. Sildenafil concentration-dependently induced vasorelaxation of preconstricted aortic rings. The concentration response to sildenafil was significantly attenuated in aortic rings following preincubation with the NO synthase inhibitor L-NNA, indicating that enhanced endogenous NO signalling accounts for the majority of sildenafil-induced vasorelaxation at lower and medium concentrations in both groups (Figure 1a).
Acute in vitro incubation with sildenafil at a threshold concentration (100 mM) enhanced acetylcholine-induced, NO-mediated, endothelium-dependent vasorelaxation of preconstricted aortic rings from diabetic rats to levels comparable to vessels from healthy rats (Figure 1b). This effect seemed to be predominantly attributable to enhanced smooth muscle sensitivity to NO, because sildenafil also markedly enhanced endothelium-independent vasorelaxation to the NO donor DEA-NONOate, which was significantly impaired in aortic rings from diabetic rats (Figure 1c).
For assessment of acute effects of sildenafil in vivo, aortic rings were harvested from diabetic rats 2 h after a single application of sildenafil (5 mg kg−1 body weight). Endothelium-dependent as well as -independent vasorelaxation was significantly improved (Figure 2).
In addition to the above-mentioned direct (Figure 1) and acute (Figure 2) effects of sildenafil on vascular function, we assessed the impact of chronic sildenafil treatment (2 weeks) with the last dose given >24 h before the experiments. Acetylcholine-induced endothelium-dependent vasorelaxation, which was impaired in diabetic rats, was significantly improved after chronic treatment with sildenafil (Figures 3a and c).
The concentration–response curve for the NO-donor DEA-NONOate, which was used to assess endothelium-independent vasorelaxation, was shifted to the right in aortae from diabetic rats and was normalized after chronic treatment with sildenafil as indicated by the respective EC50 values (Figures 3b and d).
One very important contributor to reduced sensitivity to NO in diabetes is oxidative stress. Aortic superoxide production was therefore assessed by lucigenin-enhanced chemiluminescence. In aortic tissues, luminescence signals were significantly increased in rats with diabetes and significantly reduced by chronic treatment with sildenafil (Figure 4a). In contrast, direct in vitro incubation with sildenafil (100 nM) did not inhibit superoxide generation (superoxide generation: STZ+buffer 0.91±0.11 c.p.m. mg−1 dry weight; STZ+sildenafil 0.90±0.18 c.p.m. mg−1 dry weight). Expression of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase subunits gp91phox and p47phox was significantly increased in aortae from diabetic rats. gp91phox expression was significantly reduced by chronic treatment with sildenafil (Figure 4b), whereas the increased p47phox expression was not changed (data not shown). Translocation of the small G protein Rac-1 to the plasma membrane activates NADPH oxidase. There was a trend towards increased Rac-1 expression in the aortic membrane in diabetes vs control, and sildenafil significantly reduced Rac-1 expression changes (Figure 4c). Confocal scanning microscopy images, which utilized hydroethidium to visualize superoxide formation, indicated increased ROS production throughout the vascular wall in aortic rings from diabetic vs control animals, and markedly reduced signals in rats treated with sildenafil as compared to untreated diabetic animals (Figure 4d).
In addition to these mechanisms demonstrating ROS production throughout the vascular wall, endothelium-specific increased superoxide formation by uncoupled eNOS with reduced dimer/monomer ratio has been observed in diabetes (Alp et al., 2003; Cai et al., 2005). We found that eNOS dimer/monomer ratio was significantly reduced in the aortae of diabetic animals, which was reversed by chronic treatment with sildenafil (Figure 5a).
To exclude superoxide formation by infiltrating macrophages, we performed immunohistochemistry to detect CD68+ cells. We did not find significant infiltration of aortae from diabetic animals by CD68+ cells (Figure 5b).
In this study, we demonstrate that acute and chronic treatment with the PDE-5 inhibitor sildenafil in experimental diabetes improves vascular function and reduces superoxide formation, suggesting a potential general mechanism of improved signalling through the NO/cGMP-signalling cascade.
Patients with erectile dysfunction in the absence of clinical cardiovascular disease have a deficient endothelium-dependent and -independent vasodilation occurring before the development of other overt functional or structural systemic vascular disease (Kaiser et al., 2004). Erectile dysfunction in diabetic men correlates with endothelial dysfunction and reduced NO bioavailability/activity might provide a unifying explanation (De Angelis et al., 2001). Based on these studies, we determined whether sildenafil, which is an approved therapy for erectile dysfunction, beneficially modulates endothelial function in experimental diabetes. We firstly demonstrated that sildenafil exerts a similar concentration-dependent vasorelaxation in isolated aortic rings from diabetic and healthy rats. Prior inhibition of NO synthase by L-NNA showed that this direct effect of sildenafil was truly NO dependent, at least at low and medium concentrations of sildenafil. Based on that concentration response, we performed further experiments, in which a threshold concentration of sildenafil was added before stimulating endogenous NO release or supplementing exogenous NO. Sildenafil enhanced acetylcholine-induced NO-mediated vasorelaxation in aortic rings from diabetic rats. Similarly, the impaired responsiveness to exogenous NO in aortic rings from diabetic rats was significantly improved by pre-incubation with sildenafil. These effects were also observed when sildenafil was given in vivo 2 h before the in vitro experiments. Thus, acute application of sildenafil is sufficient to enhance the impaired vascular NO/cGMP signalling in diabetes.
Chronic treatment of diabetic rats with sildenafil resulted in a significant improvement of endothelium-dependent as well as -independent vasorelaxation indicating improved signalling through the NO/cGMP-signalling cascade, even more than 24 h after the last dosing. A recent study in hypertensive patients demonstrated that sildenafil had the potential for chronic treatment in addition to its specific local indication as acute supportive treatment in erectile dysfunction (Oliver et al., 2006). Endothelial dysfunction is a common feature in cardiovascular diseases characterized by an imbalance between NO and ROS. Oxidant stress is a major cause of reduced endothelial NO bioavailability in diabetes and is involved in the pathogenesis and progression of diabetic tissue damage (Guzik et al., 2002; Landmesser et al., 2006). Increased expression of NAD(P)H oxidase subunits, enhanced NAD(P)H oxidase and protein kinase C activity as well as increased levels of the endogenous eNOS inhibitor asymmetric dimethylarginine result in enhanced oxidative stress and reduced NO bioavailability in diabetes (Hink et al., 2001). Treatment with sildenafil reduced lipid peroxidation and increased total antioxidant capacity in plasma of diabetic rats (Milani et al., 2005). Sildenafil inhibited NAD(P)H oxidase-dependent superoxide formation in corpus cavernosum from hypercholesterinaemic rabbits thereby improving smooth muscle relaxation even in the absence of endogenous NO release (Shukla et al., 2005). Sildenafil also reduced superoxide formation and NAD(P)H oxidase expression in cultured porcine pulmonary artery endothelial cells as well as corpus cavernosum smooth muscle cells (Koupparis et al., 2005; Muzaffar et al., 2005a, 2005b). We demonstrate that chronic treatment with sildenafil reduces the enhanced superoxide formation and expression of the NAD(P)H oxidase subunit gp91phox in aortae from diabetic rats. eNOS becomes uncoupled under certain circumstances such as high glucose challenge to produce superoxide instead of NO (Bauersachs and Schäfer, 2005; Forstermann and Munzel, 2006). However, a major contribution of uncoupled eNOS to superoxide formation in the present experimental condition is unlikely. Enhanced formation of superoxide in all layers of the aortic wall together with the increased expression of certain NAD(P)H oxidase subunits and expression changes of Rac-1 favours NAD(P)H oxidase as the major source of ROS. This is further supported by the fact that L-NNA did not significantly reduce superoxide formation in aortae from diabetic rats (data not shown). Nevertheless, there was a significant shift in eNOS dimer/monomer ratio towards monomer in aortae from diabetic rats as previously reported in mouse aortae (Alp et al., 2003; Cai et al., 2005), which was reversed by chronic treatment with sildenafil.
The reaction product of superoxide and NO, peroxynitrite, may compensate in part for the loss of vascular NO bioavailability in diabetes (Zobali et al., 2001). It has recently been demonstrated that peroxynitrite is capable of inducing cGMP-dependent vasorelaxation, which can be enhanced by the PDE-5 inhibitor sildenafil (Li et al., 2005). Thereby, PDE-5 inhibition would not only result in the enhancement of the residual truly NO-mediated signal but also amplify the peroxynitrite-induced vasorelaxation.
In summary, chronic treatment with sildenafil reduced superoxide formation and enhanced smooth muscle reactivity to exogenous NO in our study, thereby reversing the deleterious events leading to vascular dysfunction, and improving vascular homoeostasis. Chronic regular treatment with sildenafil has recently been shown to constitute an effective antihypertensive therapy and may also be a valuable tool to improve diabetic vascular dysfunction. However, its use for chronic treatment in clinical practice is limited by its short duration of action requiring multiple doses per day (Oliver et al., 2006) and this pharmacological approach is currently only applied for diabetic patients with a very well-defined subset of endothelial dysfunction, namely erectile dysfunction (Basu and Ryder, 2004; Pegge et al., 2006).
We thank Meike Leutke for expert technical assistance. The study was partially supported by a research grant from Pfizer and by the Deutsche Forschungsgemeinschaft (SFB 355 and 688 to T.R.).
The study was partially supported by a research grant from Pfizer.