A Phosphodiesterase-5 Inhibitor Vardenafil Enhances Angiogenesis Through a Protein Kinase G–Dependent Hypoxia-Inducible Factor-1/Vascular Endothelial Growth Factor Pathway
Objective— We examined whether phosphodiesterase-5 (PDE5) inhibition can promote ischemia-induced angiogenesis.
Methods and Results— Unilateral hindlimb ischemia was generated by resecting right femoral artery in wild-type C3H/He mice, treated with either vehicle or a PDE5 inhibitor vardenafil (10 mg/kg per day). Four weeks after surgery, vardenafil significantly enhanced blood flow recovery and augmented capillary collateral formation in ischemic muscle (blood flow ratios of ischemic/nonischemic leg: 0.52±0.17 [vehicle] versus 0.92±0.09 [vardenafil], P<0.01). Vardenafil upregulated protein expression of vascular endothelial growth factor and hypoxia-inducible factor (HIF)-1α in ischemic muscle and enhanced mobilization of Sca-1/Flk-1-positive endothelial progenitor cells (EPCs) in peripheral blood and bone marrow, contributing to neovascularization. Vardenafil also promoted capillary-like tube formation of human umbilical vein endothelial cells and increased the number of human blood mononuclear cell-derived EPCs in vitro. Furthermore, reporter assays showed that vardenafil and cGMP activated the transactivation activity of HIF-1 under hypoxia. These effects of vardenafil were markedly inhibited by genetic ablation of endothelial nitric oxide synthase, a soluble guanylate cyclase inhibitor, and a protein kinase G inhibitor, respectively.
Conclusion— Our results suggest that PDE5 inhibition enhances ischemia-induced angiogenesis with mobilization of EPCs through a protein kinase G–dependent HIF-1/vascular endothelial growth factor pathway. PDE5 inhibition may have a therapeutic potential to treat ischemic cardiovascular diseases.
Targeting cyclic guanosine monophosphate (cGMP)-specific phosphodiesterase-5 (PDE5) has recently attracted much interest not only in erectile dysfunction but also in several cardiovascular diseases.1,2 PDE5 hydrolyzes and degrades cGMP that is formed from guanosine triphosphate by action of guanylate cyclases (GCs), which are activated by nitric oxide (NO) and natriuretic peptides.3 PDE5 is highly expressed in smooth muscle cells of the corpora cavernosa4 and pulmonary arteries,5 and it has potent effects on vascular tone in these vasculatures.2,5 Orally active PDE5 inhibitors, such as sildenafil, tadalafil, and vardenafil, are widely used to clinically enhance erectile function, and the clinical use of sildenafil has been approved for the treatment of pulmonary arterial hypertension. PDE5 is also expressed in various other tissues, such as skeletal muscle, platelets, and myocardium,4 and therefore, the effects of PDE5 inhibitors are complex. In contrast to the relatively well-known effects of PDE5 inhibitors on vascular smooth muscle cells, little is known about the effects of PDE5 inhibitors on vascular endothelial cell functions and properties, in particular those associated with angiogenesis.
Angiogenesis (postnatal neovascularization) is the process that includes dissolution of matrix, endothelial cell proliferation and migration, and organization to form a tube, which is tightly regulated by a balance of many angiogenic and antiangiogenic factors.6 Stimulation of neovascularization is considered a promising therapeutic approach to treat conditions that include ischemic heart disease and peripheral artery disease, although excessive and deregulated angiogenesis can contribute to the pathogenesis of cancer, diabetic microvascular disease, and rheumatoid arthritis.7 Cumulative evidence suggests that circulating endothelial progenitor cells (EPCs) mobilized from bone marrow (BM) significantly contribute to and are also required for angiogenesis.8 EPCs are known to be mobilized from BM into the circulation via several angiogenic factors, such as vascular endothelial growth factor (VEGF).9 Furthermore, endothelial NO synthase (eNOS) not only enhances growth, migration, and tube formation of endothelial cells, leading to angiogenesis but also plays an essential role in the mobilization of EPCs from BM at tissue ischemia.10 Here, to better understand the role of the NO/cGMP pathway in the angiogenic process, including the mobilization of EPCs, and to evaluate the additional therapeutic potential of PDE5 inhibitors for tissue ischemia, we investigated the effects of a PDE5 inhibitor, vardenafil, on angiogenesis in vivo and in vitro. The results suggest that vardenafil enhances ischemia-induced angiogenesis with mobilization of EPCs through a protein kinase G (PKG)–dependent hypoxia-inducible factor (HIF)-1/VEGF pathway, leading to the notion that PDE5 inhibition may have a therapeutic potential to treat patients with ischemic cardiovascular diseases, especially peripheral arterial disease.
Wild-type C57BL/6 and C3H/He mice were purchased from SLC (Hamamatsu, Japan). eNOS-deficient (eNOS−/−) mice were purchased from Jackson Laboratory.11 Transgenic mice (C57BL/6 background) that ubiquitously express enhanced green fluorescent protein (GFP) were donated by Dr. M. Okabe (Osaka University). All experimental procedures and protocols were approved by the institutional committee for animal research at the University of Tokyo and complied with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication no. 86-23; revised 1985).
Mouse Hindlimb Ischemia Model
Unilateral hindlimb ischemia was induced in 12- to 16-week-old male mice by resecting the right femoral and saphenous arteries.11 Mice were classified into the 7 groups (n=8 to 10 per group) and treated with either vehicle (phosphate buffered saline); 2.5 or 10 mg/kg per day of a PDE5 inhibitor vardenafil (Bayer); IP injection of 5 mg/kg of a soluble GC inhibitor 1H-[1,2,4]oxadiazolo[4,3-a] quinoxaline-1-1 (ODQ; Sigma) every 3 days with or without 10 mg/kg per day vardenafil; or IP injection of 1 mg/kg of a specific PKG inhibitor, DT-3 (Sigma), every 3 days with or without 10 mg/kg per day vardenafil. Vehicle or vardenafil was administered continuously by an implanted subcutaneous osmotic pump (Alzet; Durect Corp), and the treatment was initiated at 2 days before the hindlimb surgery. Hindlimb blood perfusion was measured using a laser Doppler perfusion imaging system (Moor Instruments). The results were expressed as the blood flow ratio of the right (ischemic) to left (nonischemic) limb.10,11 Overall, wild-type C3H/He mice were used to determine the angiogenic effects of vardenafil, because C3H/He mice reveal more severe impairment of blood flow recovery after hindlimb ischemic surgery compared with wild-type C57BL/6 mice.11 In contrast, wild-type C57BL/6 mice were used as control to evaluate the angiogenic effects of vardenafil in eNOS−/− mice.
Induction of BM-Chimeric Mice
BM transplantation (BMT) was performed as described previously.12 One day after the lethal X irradiation, 8-week-old male wild-type C57BL/6 mice received unfractionated BM cells (3×106) derived from GFP-transgenic mice. Eight weeks after BMT, the unilateral hindlimb ischemia surgery was induced in the recipient mice.
Four weeks after the ischemia surgery, the mice were euthanized, and the thigh and calf muscles were embedded in paraffin after fixation in methanol. Then, sections (5 μm) were deparaffinized and incubated with monoclonal antibodies against murine CD31 (BD Biosciences Pharmingen) and murine E-selectin (R&D Systems), followed by incubation with biotinylated secondary antibody and subsequent use of the avidin-biotin complex technique (Vector Laboratories). Capillaries were identified based on positive staining for CD31 and their morphological appearance.11 Six different fields from each muscle were randomly selected, and capillaries were counted. Capillary density was expressed as the number of capillaries/mm2.10,11
Immunofluorescence double staining was performed on the muscle harvested from the BMTGFP→wild mice after surgery. The frozen muscle sections were doubly stained with an anti-GFP antibody (Molecular Probes) and either an anti-CD31, anti-CD45 (BD Biosciences Pharmingen), or anti-E-selectin antibody, followed by incubation with Alexa 488–conjugated (Molecular Probes) and Cy3-conjugated (Jackson Immunoresearch) secondary antibodies. The nuclei were counterstained with Hoechst 33258 (Sigma). The sections were observed under a confocal microscope (Fluoview FV300; Olympus). Six different fields from each muscle were randomly selected, and the double-positive cells for GFP and E-selectin were counted.12
Proteins were extracted from the thigh and calf muscles homogenized in a lysis buffer containing 150 mmol/L NaCl, 10 mmol/L Tris-HCl (pH 7.4), 1% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate, and a protease inhibitor cocktail (Sigma). Protein samples (10 μg) were separated by SDS-PAGE and transferred to a polyvinylidene fluoride membrane (Hybond-P; GE Healthcare Biosciences). The membrane was incubated with a polyclonal antibody against VEGF (dilution, 1:1000; Santa Cruz Biotechnology) and a monoclonal antibody against HIF-1α (1:1000; Santa Cruz Biotechnology), followed by incubation with horseradish peroxidase–conjugated secondary antibody. Then, the enhanced chemiluminescence system (ECL Plus; GE Healthcare) was used for detection, and bands were visualized and quantified using a luminoanalyzer (LAS-1000; Fujifilm). Signal intensity was normalized to β-actin expression (Sigma).
Measurement of Plasma and Tissue cGMP Levels
To investigate the systemic and local effects of vardenafil treatment, plasma and hindlimb muscle tissue cGMP levels in mice with or without hindlimb ischemia were determined by radioimmunoassay using a Yamasa cGMP assay kit according to the manufacturer’s instructions.
Flow Cytometry Analysis
In a separate experiment, peripheral blood and BM of vehicle- or vardenafil-treated mice with or without hindlimb ischemia were harvested at 1, 7, 14, and 28 days after the ischemic surgery (n=4 per group). After the lysis of erythrocytes using a mouse erythrocyte lysing kit (R&D Systems), the viable mononuclear cells (MNCs) were analyzed for the expression of Sca-1-fluorescein isothiocyanate (FITC) (eBioscience) and Flk-1(VEGFR2)-PE (eBioscience). Isotype-identical antibodies were used as negative control. Quantitative fluorescence analysis was performed by using an Epics Altra instrument (Beckman Coulter). Data were evaluated with Expo32 software.
In Vitro Capillary-Like Tube Formation Assay
An in vitro angiogenesis assay was performed using an angiogenesis kit (KZ-1000; Kurabo), which consists of a 2-dimensional coculture system of human umbilical vein endothelial cells (HUVECs) and normal human dermal fibroblasts, according to the manufacturer’s instructions. Either vardenafil (10 to 1000 nmol/L), 10 μmol/L ODQ with or without 100 nmol/L vardenafil, 1 μmol/L DT-3 with or without 100 nmol/L vardenafil, or 50 ng/mL of a neutralizing anti-human VEGF goat polyclonal antibody (R&D Systems) with or without 100 nmol/L vardenafil was added to the medium. Human recombinant VEGF-A and an antitrypanosomal agent, suramin (Kurabo), were used as positive and negative control, respectively.13 At 3 days, the supernatants on culture medium in the respective wells were harvested to measure the concentration of VEGF by enzyme-linked immunosorbent assay, using a commercially available kit (R&D Systems). At 11 days, cells were fixed with 10% formalin and stained with an anti-CD31 antibody. The length and area of CD31-positive endothelial cells forming capillary-like tubes were quantified by an angiogenesis image analyzing system (Kurabo) in 5 randomly selected high-power fields (×20). All assays were performed in triplicate.
EPC Culture Assay
Human peripheral blood MNCs were isolated from healthy volunteers by density-gradient centrifugation with Histopaque-1077 (Sigma) and were resuspended in endothelial cell culture medium (EBM-2; Cambrex BioScience) supplemented with 1 μg/mL hydrocortisone, 3 μg/mL bovine pituitary extract, 10 ng/mL human recombinant VEGF (R&D Systems), and 10% FBS (complete EGM-2 medium). MNCs were cultured at a density of 1×107 cells per well of a 6-well plate precoated with type 1 rat tail collagen (BD Biosciences) at 37°C, 5% CO2, in a humidified incubator. After 24 hours, nonadherent cells and debris were aspirated, adherent cells were washed once, and complete EGM-2 medium was added to each well. Medium was changed daily for 7 days and then every other day.14,15 Throughout incubation, MNCs were concurrently stimulated with either vardenafil (10 to 1000 nmol/L), 10 μmol/L ODQ with or without 100 nmol/L vardenafil, 1 μmol/L DT-3 with or without 100 nmol/L vardenafil, 50 ng/mL of neutralizing anti-human VEGF antibody with or without 100 nmol/L vardenafil, or 100 ng/mL human recombinant endothelial growth factor. At days 14 to 21, colonies of endothelial cells that originated from adherent cells appeared, and the number of endothelial cell colonies was counted by visual inspection using an inverted microscope. At day 21, adherent cells forming endothelial cell colonies were incubated with 10 μg/mL 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindo-carbocyanine perchlorate (DiI-Ac-LDL; Molecular Probes) for 2 hours. Then, cells were fixed with 2% paraformaldehyde and stained with FITC-labeled Ulex europaeus agglutinin I (UEA I; EY laboratories). The nuclei were counterstained with Hoechst 33258 (Sigma). Adherent cells staining positively for both DiI-Ac-LDL and FITC-UEA I were judged to be EPCs.10,14–17 The number of EPCs was counted per well by 2 independent investigators.
Genomic DNA was prepared from human peripheral blood MNCs. The promoter region of VEGF gene, comprising the VEGF sequence from −1096 to +5 relative to the transcription initiation site, was obtained by polymerase chain reaction (PCR) with 2 primers: the forward primer was 5′-CGGCTAGCAGCCCCCTGGCCTCAGTT-3′, and the reverse primer was 5′-GGAAGCTTAAGCCTCCGCGATCCTCC-3′. The 5′-flanking region of the VEGF gene contains the consensus HIF-1 binding-site, hypoxia-responsive element (HRE) sequence 5′-TACGTGGG-3′, located between −975 and −968.18,19 The PCR product was digested with NheI and HindIII and subcloned into pGL3 basic plasmid containing the firefly luciferase gene (Promega). This constructed plasmid (pHRE-Luc) and a control vector pRL-SV40 (Promega), which contains the Renilla luciferase gene, were used for the transfection into human embryonic kidney (HEK) 293 cells.
Transfection of the Constructed Plasmid into HEK 293 Cells and the Dual-Luciferase Assays
HEK 293 cells were transfected by a lipofection reagent (Superfect; Qiagen) with 2 μg of pHRE-Luc and 0.2 μg of pRL-SV40. After 27 hours of incubation, the transfected HEK 293 cells were stimulated with either vardenafil (10 to 1000 nmol/L), 1 mmol/L 8-bromo (Br)-cGMP (a stable analog of cGMP; Biomol International), 10 μmol/L ODQ with or without either 100 nmol/L vardenafil or 1 mmol/L 8-Br-cGMP, or 1 μmol/L DT-3 with or without either 100 nmol/L vardenafil or 1 mmol/L 8-Br-cGMP, under either normoxic conditions or hypoxic conditions of less than 1% O2 (5% CO2), which was generated by a CO2-generating oxygen absorbent product (Anaeropack A-07; Mitsubishi Gas Chemical), for the next 4 hours. Then, the cell lysates were obtained, and the dual-luciferase activities were measured using a reporter assay system (Promega) with a luminometer. The relative fold induction of firefly luciferase activity normalized to Renilla luciferase activity was determined. All transfections were done in duplicate, and the experiments were repeated 3 times.
Transfection of Small Interfering RNA Against HIF-1α
A duplex 21-nucleotide small interfering RNA (siRNA) against HIF-1α was designed by Ambion (Applied Biosystems no. 144735) according to the manufacturer’s instructions. The double-stranded HIF-1α-targeting siRNA was 5′-CCUCAGUGUGGGUAUAAGATT-3′ (sense) and 5′-UCUUAUACCCACACUGAGGTT-3′ (antisense). A negative control siRNA (Ambion), which does not target any gene product, was used in the same experiment. Approximately 2×105 HEK 293 cells per well were grown at 60% to 80% confluence in a 6-well plate. The HEK 293 cells were incubated with siRNA (30 nmol/L) using a transfection agent (Lipofectamine 2000; Invitrogen) for 4 hours, and then complete growth media was added. After 48 hours, gene silencing was monitored by quantitative RT-PCR using the extracted total RNA. Thereafter, as mentioned above, the siRNA-transfected HEK 293 cells were cotransfected with pHRE-Luc and pRL-SV40, and after 27 hours of incubation, those were stimulated with or without either vardenafil or 8-Br-cGMP, under either normoxic or hypoxic (<1% O2) conditions for 4 hours. In the same manner, the cell lysates were obtained, and the dual-luciferase activities were measured with a luminometer.
Data are presented as mean±SD. Comparisons of means were performed by 1-way ANOVA followed by the Scheffé post hoc test, and the blood flow ratios were compared by repeated-measures ANOVA followed by the Student t test. Statistical significance was defined as P<0.05.
Vardenafil Treatment Enhances Blood Flow Recovery and Augments Capillary Collateral Formation After Hindlimb Ischemia
Unilateral hindlimb ischemia was induced in wild-type C3H/He mice. Blood flow of the ischemic and nonischemic legs was monitored weekly by laser Doppler imaging (Figure 1A). In control mice treated with vehicle, blood flow of the ischemic leg recovered gradually and reached approximately half the blood flow of the nonischemic leg at 4 weeks (Figure 1B). Vardenafil treatment significantly enhanced blood flow recovery in the ischemic leg in a dose-dependent manner (blood flow ratio at 4 weeks: control, 0.52±0.17; 2.5 mg/kg per day vardenafil, 0.74±0.14, P<0.05 versus control; 10 mg/kg per day vardenafil, 0.92±0.09, P<0.01 versus control). In contrast, coadministration of ODQ and DT-3 diminished the enhanced blood flow recovery by vardenafil, respectively.
Collateral formation was evaluated based on the capillary density of the ischemic hindlimb muscle. The anti-CD31 and anti-E-selectin immunostaining revealed that vardenafil treatment augmented capillary collateral formation and increased the capillary density in the ischemic leg in a dose-dependent manner (control, 276±62/mm2; 2.5 mg/kg per day vardenafil, 455±70/mm2, P<0.05 versus control; 10 mg/kg per day vardenafil, 636±129/mm2, P<0.01 versus control) (Figure 1C and 1D). In contrast, coadministration of ODQ and DT-3 attenuated the increase of capillary density by vardenafil, respectively. Vardenafil treatment did not increase the capillary density in the nonischemic leg (control, 167±45/mm2, versus 10 mg/kg per day vardenafil, 187±37/mm2, P=not significant), suggesting that vardenafil does not enhance angiogenesis in nonischemic tissue.
Next, we examined the effects of vardenafil therapy on plasma and hindlimb muscle tissue cGMP levels in mice with or without hindlimb ischemia. Vardenafil treatment (10 mg/kg per day) tends to mildly increase plasma cGMP levels in both normal and hindlimb ischemia mice, although it is not statistically significant (Supplemental Figure IA, available online at http://atvb.ahajournals.org). In fact, vardenafil treatment did not alter the systemic blood pressure and heart rate of mice with or without hindlimb ischemia (data not shown). In contrast, muscle tissue cGMP levels in vardenafil-treated mice with hindlimb ischemia were significantly increased, particularly in the ischemic hindlimb, compared with vehicle-treated mice, although muscle tissue cGMP levels in normal mice were not different irrespective of vardenafil treatment (Supplemental Figure IB).
Vardenafil Treatment Upregulates Protein Expressions of VEGF and HIF-1α in the Ischemic Hindlimb Muscle
VEGF plays a pivotal role in angiogenesis.6,7 Thus, we examined the protein expressions of VEGF and HIF-1α, which is a transcriptional factor that upregulates transcription of the VEGF gene,20 in hindlimb muscles. In the ischemic muscle of mice treated with vehicle, VEGF expression was mildly upregulated at 2 weeks, although it returned to the baseline level at 4 weeks. Vardenafil treatment markedly upregulated protein expressions of both VEGF and HIF-1α in the ischemic muscle, with a maximum at 2 weeks, in a dose-dependent manner, and the upregulation was detected even at 4 weeks in mice treated with 10 mg/kg per day vardenafil (Figure 2A and 2B). As shown, vardenafil treatment did not upregulate expression of these proteins in the nonischemic muscle. Coadministration of ODQ and DT-3 significantly attenuated the increase of VEGF and HIF-1α protein expression by vardenafil, respectively.
PDE5 Inhibitors Fail to Improve Blood Flow Recovery and Capillary Collateral Formation in eNOS−/− Mice With Hindlimb Ischemia
To examine the role of eNOS/NO on the angiogenic effects by PDE5 inhibitors including vardenafil and another PDE5 inhibitor, sildenafil (Pfizer), unilateral hindlimb ischemia was induced in both wild-type C57BL/6 and eNOS−/− mice. After surgery, blood flow recovery and capillary collateral formation in eNOS−/− mice were severely impaired compared with wild-type mice (Supplemental Figure IIA and IIB), as reported previously.10,11 Intriguingly, vardenafil and sildenafil treatment improved blood flow recovery and capillary collateral formation in wild-type mice in a similar fashion but could not improve them in eNOS−/− mice. In addition, vardenafil and sildenafil treatment did not upregulate VEGF or HIF-1α expressions in the ischemic leg in eNOS−/− mice, unlike the case in wild-type mice (Supplemental Figure IIC). These results suggest that eNOS/NO are essential for the PDE5 inhibitors-induced angiogenesis and upregulation of HIF-1α/VEGF at tissue ischemia.
Vardenafil Treatment Enhances Mobilization of EPCs From BM, Contributing to Neovascularization
The time-dependent effects of vardenafil treatment on the numbers of Sca-1/Flk-1-positive cells, generally considered EPCs,21,22 in peripheral blood and BM of mice with or without hindlimb ischemia are shown in Figure 3A and 3B. In normal mice, vardenafil treatment did not cause a significant increase in EPCs. On the contrary, in hindlimb ischemia mice, vardenafil treatment significantly increased numbers of EPCs in the peripheral blood and BM in a dose-dependent manner from 7 days after the surgery, with a maximum at 14 days. Coadministration of ODQ and DT-3 attenuated the increase of EPCs by vardenafil, respectively.
Next, we generated unilateral hindlimb ischemia in the BMTGFP→wild mice, which were treated with vehicle or vardenafil. In the BMT mice, BM-derived cells were positive for GFP. Immunofluorescence double staining of ischemic muscle sections in vehicle-treated mice (Figure 3C) revealed that the majority of BM-derived cells in muscle tissue were negative for CD31 but positive for CD45, suggesting that those cells were considered hematopoietic lineage cells. In contrast, BM-derived endothelial-like cells were detected as the double-positive cells for GFP and E-selectin (Figure 3D). Notably, vardenafil treatment significantly increased the number of BM-derived endothelial-like cells participating in the newly formed capillary network in the ischemic muscle in a dose- and time-dependent manner (Figure 3D and 3E). At 4 weeks after surgery, the percentages of GFP/E-selectin double-positive cells among E-selectin-positive endothelial cells in the ischemic muscle were as follows: control, 0.5±0.5%; 2.5 mg/kg per day vardenafil, 3.8±1.3%; and 10 mg/kg per day vardenafil, 7.5±2.8% (P<0.01 versus control). GFP/E-selectin-double-positive cells were not detected in the nonischemic hindlimb muscle irrespective of vardenafil treatment.
Induction of In Vitro Capillary-Like Tube Formation by Vardenafil
The angiogenic effect of vardenafil was also investigated in an in vitro angiogenesis assay using a commercially available kit. As shown in Figure 4A, vardenafil treatment significantly enhanced capillary-like tube formation of HUVECs in a concentration-dependent manner. Those observational findings were confirmed by the quantitative results of the luminal length and area (Figure 4B and 4C). In addition, vardenafil treatment significantly increased the concentration of VEGF in the supernatants on culture medium (control, 127.9±64.3 pg/mL, versus 100 nmol/L vardenafil, 613.2±187.3 pg/mL; P<0.01). In contrast, the angiogenic effect of vardenafil was markedly inhibited by exposure of ODQ, DT-3, and a neutralizing anti-human VEGF antibody (Figure 4A through 4C).
Vardenafil Increases the Cell and Colony Numbers of Human Blood MNC-Derived EPCs
Human blood MNCs were cultured in endothelial-cell culture medium in the absence or presence of vardenafil. As seen previously,14,15 at days 14 to 21, colonies of endothelial cells that originated from adherent cells appeared and were identified as well-circumscribed monolayers of cobblestone-appearing cells (Figure 5A). Notably, vardenafil treatment significantly increased the number of endothelial cell colonies generated from 107 MNCs (control, 0.25±0.45, versus 100 nmol/L vardenafil, 1.17±0.72 per well; P<0.01) (Figure 5B). Almost all of the adherent cells forming the endothelial cell colony were double-positive for DiI-Ac-LDL and FITC-UEA I (Figure 5C), and those cells are thought of and defined as EPCs.10,14–17 Vardenafil also significantly increased the number of EPCs in a concentration-dependent manner, with a 2- to 4-fold increase compared with control (Figure 5D). The effects of vardenafil on the cell and colony numbers of human MNC-derived EPCs were relatively comparable to those of endothelial growth factor (positive control), and in contrast, those were abrogated by exposure of ODQ, DT-3, and a neutralizing anti-human VEGF antibody, respectively (Figure 5B and 5D).
Effects of Vardenafil and cGMP on the Transactivation Activity of HIF-1
We carried out the dual-luciferase reporter assay to investigate whether vardenafil regulates the transactivation activity of HIF-1, which plays a crucial role in VEGF transcription. We cotransfected pHRE-Luc, which contains the HIF-1 binding site HRE sequence (5′-TACGTGGG-3′) in the promoter region of VEGF gene, with a control vector (pRL-SV40) into HEK 293 cells. Then, the transfected HEK 293 cells were incubated under either normoxic or hypoxic (<1% O2) conditions in the absence or presence of vardenafil or 8-Br-cGMP. HEK 293 cells stimulated with vardenafil (100 nmol/L) or 8-Br-cGMP (1 mmol/L) under hypoxic conditions resulted in 6.3- or 7.2-fold increase of the luciferase activity compared with control cells under normoxic conditions (P<0.01), although they did not increase the activity under normoxic conditions (Figure 6A), suggesting that vardenafil and 8-Br-cGMP activate the HIF-1 transactivation activity primarily at hypoxia, respectively. ODQ prevented the increase of the luciferase activity induced by vardenafil but could not attenuate that by 8-Br-cGMP, whereas DT-3 inhibited both the vardenafil-induced and 8-Br-cGMP-induced increase of the luciferase activity (Figure 6A).
Next, to confirm that the vardenafil- and 8-Br-cGMP-induced augmentation in VEGF transcription is dependent on HIF-1, the expression of HIF-1α was suppressed using the double-stranded HIF-1α-targeting siRNA in another experiment series using HEK 293 cells. Quantitative RT-PCR analyses revealed that the HIF-1α expression was significantly downregulated by the HIF-1α-targeting siRNA (−82.4±7.9%), although a negative control siRNA did not affect the HIF-1α expression. When the negative control siRNA-transfected HEK 293 cells were cotransfected with pHRE-Luc and pRL-SV40, the subsequent stimulation with vardenafil or 8-Br-cGMP under hypoxic conditions similarly increased the luciferase activity compared with control (Figure 6B). In contrast, when the HIF-1α-targeting siRNA-transfected HEK 293 cells were cotransfected with pHRE-Luc and pRL-SV40, the subsequent stimulation with vardenafil or 8-Br-cGMP did not increase the luciferase activity compared with control, even under hypoxic conditions (Figure 6B).
The Vardenafil-Induced Angiogenic Effects Are Dependent on the Activity of HIF-1
Finally, to examine whether the vardenafil-induced angiogenic effects in vivo are indeed dependent on the activity of HIF-1, in a separate experiment, we generated unilateral hindlimb ischemia in wild-type C3H/He mice, which were treated with either vehicle, vardenafil (10 mg/kg per day), or vardenafil (10 mg/kg per day) plus IP injection of 100 mg/kg of an HIF-1 inhibitor, YC-1 (Cayman Chemical), every 3 days. As shown in Supplemental Figure III, YC-1 significantly attenuated the vardenafil-induced blood flow recovery, capillary collateral formation, VEGF upregulation in ischemic muscle, and increase of the number of Sca-1/Flk-1-positive cells in the peripheral blood, respectively.
We next examined the role of the activity of HIF-1 on the vardenafil-induced angiogenic effects in vitro, using the same angiogenesis kit (KZ-1000; Kurabo) as in Figure 4. As shown in Supplemental Figure IV, the enhanced capillary-like tube formation of HUVECs by vardenafil was markedly inhibited by exposure of an HIF-1 inhibitor YC-1 (60 mg/mL).
We here show that the PDE5 inhibitor vardenafil significantly enhanced blood flow recovery and augmented capillary collateral formation in ischemic hindlimb muscle, along with the upregulation of VEGF and HIF-1α protein expressions. Vardenafil also enhanced the mobilization of EPCs in peripheral blood and BM, contributing to neovascularization in the ischemic tissue. In an in vitro study, vardenafil promoted capillary-like tube formation of HUVECs and increased the numbers of human blood MNC-derived EPCs. In addition, the luciferase reporter assay data showed that vardenafil and cGMP activated the HIF-1 transactivation activity at hypoxia, suggesting that HIF-1 might be a target of vardenafil and cGMP in the context of ischemia-induced angiogenesis. These angiogenic effects of vardenafil were significantly attenuated by genetic ablation of eNOS, the soluble GC inhibitor ODQ, and the PKG inhibitor DT-3, respectively, suggesting that vardenafil exerts the angiogenic effects through an eNOS/soluble GC/PKG pathway. Our findings are consistent with the report by Zhang et al,23 who revealed that another PDE5 inhibitor, sildenafil, promoted angiogenesis in the ischemic boundary regions in a rat stroke model.
The angiogenic effects of vardenafil were associated with the upregulation of VEGF in vivo and were blocked by a neutralizing anti-human VEGF antibody in vitro, suggesting that vardenafil also exerts those effects through VEGF. VEGF is a potent proangiogenic cytokine and induces EPC mobilization from BM in postnatal neovascularization.9 Evidence from the literature suggests that NO appears to be a downstream mediator of VEGF-induced angiogenesis24; however, conversely, it has also been reported that NO and sildenafil enhanced angiogenesis through the increase of VEGF synthesis in tissue ischemia.23 Although the exact mechanisms remain unclear, it has been suggested that a putative paracrine loop between endothelial cells producing NO and vascular smooth muscle cells producing VEGF might be attributable to the NO/cGMP-induced VEGF synthesis, at least in part.
Transcriptional activation of the VEGF gene is mediated by a variety of cellular transcription factors represented by HIF-1.20 HIF-1 is a heterodimer composed of a constitutively expressed β subunit (HIF-1β) and oxygen-regulated HIF-1α, which degrades readily under normoxic conditions, leading to transcriptional deactivation of HIF-1.25 Under hypoxic conditions, HIF-1α protein is stabilized because of limited substrate (O2), resulting in transcriptional activation of HIF-1. HIF-1 plays a crucial role in VEGF synthesis by binding to the HRE in the VEGF promoter region and by subsequently upregulating VEGF transcription.20 Our study revealed that vardenafil enhanced synthesis of HIF-1α in ischemic tissue and that vardenafil and cGMP activated the HIF-1 transactivation activity at hypoxia, leading to the increase of VEGF secretion. Thus, HIF-1 might be a target of vardenafil and cGMP under ischemic conditions, and VEGF might be a downstream effector in the vardenafil-mediated angiogenesis. This notion is supported by the fact that an HIF-1 inhibitor YC-1 significantly attenuated the angiogenic effects of vardenafil in vivo and in vitro. Interestingly, as seen in the present study, it has been reported that NO also enhances synthesis of HIF-1α through the soluble GC/cGMP pathway, which induces phosphorylation of translational regulatory proteins, such as p70 S6 kinase.26
Our study revealed that vardenafil-induced angiogenic effects, including the activation of a HIF-1/VEGF pathway, were blocked by deletion of eNOS gene, ODQ, and DT-3, respectively. These findings are in part consistent with a previous study27 that has reported that another PDE5 inhibitor, sildenafil, stimulates angiogenesis, which is blocked by ODQ or DT-3. In contrast, our findings are inconsistent with another previous study,28 in which Senthilkumar et al reported that sildenafil promotes ischemia-induced angiogenesis even in eNOS−/− mice in the same model as used in the present study, although both vardenafil and sildenafil could not enhance angiogenesis in eNOS−/− mice in our study. The reason for this discrepancy might be possibly attributable to the differences in the administration route of a PDE5 inhibitor (drinking water or subcutaneous osmotic pumps) or in the extent of injury derived from the ischemic surgery, but remains enigmatic. However, in the present study, DT-3 inhibited the angiogenic effects of vardenafil more strongly compared with ODQ, suggesting that the vardenafil-induced angiogenic effects are mediated by PKG more directly. The difference in the inhibitory responses between ODQ and DT-3, although it was modest, may be derived from the existence and function of a natriuretic peptides/receptor-linked GC pathway, which is another cGMP/PKG-activating pathway.3,29
Numerous studies have revealed that circulating EPCs mobilized from BM significantly contribute to angiogenesis in ischemic tissue.8 Although the mobilization of EPCs from BM is a complex mechanism, it has been suggested that both eNOS expression and local secretion of metalloproteinase-9, a downstream activator of NO, in BM are essential for the mobilization of EPCs.10 Recent clinical studies demonstrated that the number of circulating EPCs positively correlates with endothelial function in humans and that patients with erectile dysfunction show reduced EPC numbers.30 Interestingly, it has also been reported that vardenafil increases the number of circulating EPCs in humans.31 These findings in the clinical study are similar to our findings that vardenafil promoted the mobilization of EPCs in peripheral blood and BM, contributing to neovascularization at tissue ischemia, and also increased the numbers of human blood MNC-derived EPCs. Although the exact mechanisms by which vardenafil induces the mobilization and proliferation of EPCs have not been fully determined, it is likely that they are closely associated with the upregulation of VEGF synthesis by vardenafil.9,22
In conclusion, our findings suggest that the PDE5 inhibitor vardenafil enhances ischemia-induced angiogenesis with mobilization of EPCs through a PKG-dependent HIF-1/VEGF pathway. Thus, PDE5 inhibition may have a therapeutic potential to treat ischemic conditions such as peripheral arterial disease.
Sources of Funding
This study was supported in part by grants from the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labor and Welfare of Japan (M. Sata) and by a research fellowship from the Japan Society for the Promotion of Science (M. Sahara).
Received on: December 2, 2009; final version accepted on: April 5, 2010.
Reffelmann T, Kloner RA. Therapeutic potential of phosphodiesterase 5 inhibition for cardiovascular disease. Circulation. 2003; 108: 239–244.
Rybalkin SD, Yan C, Bornfeldt KE, Beavo JA. Cyclic GMP phosphodiesterases and regulation of smooth muscle function. Circ Res. 2003; 93: 280–291.
Sebkhi A, Strange JW, Phillips SC, Wharton J, Wilkins MR. Phosphodiesterase type 5 as a target for the treatment of hypoxia-induced pulmonary hypertension. Circulation. 2003; 107: 3230–3235.
Sata M. Role of circulating vascular progenitors in angiogenesis, vascular healing, and pulmonary hypertension: lessons from animal models. Arterioscler Thromb Vasc Biol. 2006; 26: 1008–1014.
Asahara T, Takahashi T, Masuda H, Kalka C, Chen D, Iwaguro H, Inai Y, Silver M, Isner JM. VEGF contributes to postnatal neovascularization by mobilizing bone marrow-derived endothelial progenitor cells. EMBO J. 1999; 18: 3964–3972.
Sata M, Nishimatsu H, Suzuki E, Sugiura S, Yoshizumi M, Ouchi Y, Hirata Y, Nagai R. Endothelial nitric oxide synthase is essential for the HMG-CoA reductase inhibitor cerivastatin to promote collateral growth in response to ischemia. FASEB J. 2001; 15: 2530–2532.
Sahara M, Sata M, Morita T, Nakamura K, Hirata Y, Nagai R. Diverse contribution of bone marrow-derived cells to vascular remodeling associated with pulmonary arterial hypertension and arterial neointimal formation. Circulation. 2007; 115: 509–517.
Ingram DA, Mead LE, Tanaka H, Meade V, Fenoglio A, Mortell K, Pollok K, Ferkowicz MJ, Gilley D, Yoder MC. Identification of a novel hierarchy of endothelial progenitor cells using human peripheral and umbilical cord blood. Blood. 2004; 104: 2752–2760.
Yoder MC, Mead LE, Prater D, Krier TR, Mroueh KN, Li F, Krasich R, Temm CJ, Prchal JT, Ingram DA. Redefining endothelial progenitor cells via clonal analysis and hematopoietic stem/progenitor cell principals. Blood. 2007; 109: 1801–1809.
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Rehman J, Li J, Orschell CM, March KL. Peripheral blood “endothelial progenitor cells” are derived from monocyte/macrophages and secrete angiogenic growth factors. Circulation. 2003; 107: 1164–1169.
Forsythe JA, Jiang BH, Iyer NV, Agani F, Leung SW, Koos RD, Semenza GL. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol. 1996; 16: 4604–4613.
Laufs U, Werner N, Link A, Endres M, Wassmann S, Jürgens K, Miche E, Böhm M, Nickenig G. Physical training increases endothelial progenitor cells, inhibits neointima formation, and enhances angiogenesis. Circulation. 2004; 109: 220–226.
Urbich C, Dimmeler S. Endothelial progenitor cells: characterization and role in vascular biology. Circ Res. 2004; 95: 343–353.
Zhang R, Wang L, Zhang L, Chen J, Zhu Z, Zhang Z, Chopp M. Nitric oxide enhances angiogenesis via the synthesis of vascular endothelial growth factor and cGMP after stroke in the rat. Circ Res. 2003; 92: 308–313.
Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O'Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI, Dhanda A, Tian YM, Masson N, Hamilton DL, Jaakkola P, Barstead R, Hodgkin J, Maxwell PH, Pugh CW, Schofield CJ, Ratcliffe PJ. C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell. 2001; 107: 43–54.
Kasuno K, Takabuchi S, Fukuda K, Kizaka-Kondoh S, Yodoi J, Adachi T, Semenza GL, Hirota K. Nitric oxide induces hypoxia-inducible factor 1 activation that is dependent on MAPK and phosphatidylinositol 3-kinase signaling. J Biol Chem. 2004; 279: 2550–2558.
Senthilkumar A, Smith RD, Khitha J, Arora N, Veerareddy S, Langston W, Chidlow JH Jr, Barlow SC, Teng X, Patel RP, Lefer DJ, Kevil CG. Sildenafil promotes ischemia-induced angiogenesis through a PKG-dependent pathway. Arterioscler Thromb Vasc Biol. 2007; 27: 1947–1954.
Kuhn M, Völker K, Schwarz K, Carbajo-Lozoya J, Flögel U, Jacoby C, Stypmann J, van Eickels M, Gambaryan S, Hartmann M, Werner M, Wieland T, Schrader J, Baba HA. The natriuretic peptide/guanylyl cyclase–a system functions as a stress-responsive regulator of angiogenesis in mice. J Clin Invest. 2009; 119: 2019–2030.