Chronic Blockade of Endothelin Receptors Improves Ischemia-Induced Angiogenesis in Rat Hindlimbs Through Activation of Vascular Endothelial Growth Factor–NO Pathway
This study investigated in vivo the putative angiogenic role of endothelin (ET)-1 in a model of ischemia-induced angiogenesis. Ischemia was produced by unilateral femoral artery occlusion in Wistar rats submitted to either chronic ET-1 infusion (2 nmol · kg−1 · min−1) or to a dual ETA/ETB receptor antagonist (bosentan, 100 mg · kg−1 · d−1) for 3 and 28 days. Arterial density was evaluated by microangiography and measurement of capillary and arteriolar density in hindlimb muscles. ET-1 infusion had no effect on ischemia-induced angiogenesis and was associated with a slight decrease in vascular endothelial growth factor (VEGF) content measured by Western blot analysis. Conversely, bosentan induced a marked increase in vessel density at 3 and 28 days (1.4-fold and 1.7-fold, respectively, compared with no treatment; P<0.05), which was associated with an increase in VEGF and endothelial NO synthase levels in ischemic legs (by 31±8% and 45±23%, respectively, at 3 days and by 65±13% and 55±15%, respectively, at 28 days; P<0.05 versus nontreated rats). At day 28, the proangiogenic effect of bosentan was abolished when NO synthesis inhibitor NG-nitro-l-arginine methyl ester (10 mg · kg−1 · d−1) or VEGF-neutralizing antibody (2.5 μg/kg twice a week) were coadministered with bosentan. Those results provide the first evidence of an early and sustained proangiogenic effect of endothelin antagonism associated with an upregulation of VEGF and endothelial NO synthase in vivo.
Angiogenesis is defined as the development of new vessels from preexisting blood vessels. In pathological conditions, this process may be either deleterious, such as in tumor growth, or potentially beneficial, such as in ischemic diseases.
In ischemic tissues, the presumed mechanism of hypoxia-induced angiogenesis involves the elevation of hypoxia-induced factor-1α, resulting in an increased expression of growth factors, such as vascular endothelial growth factor (VEGF).1 The angiogenic response to VEGF might involve the production of NO, as previously described in ischemic hindlimbs.2 Indeed, VEGF increases the expression of endothelial NO synthase (eNOS) and inducible NO synthase (iNOS) NO synthase.3
Endothelin (ET)-1 has in vitro mitogenic properties4 and potentiates the action of growth factors, such as platelet-derived growth factor.5 In vitro studies indicate that ET-1 is able to stimulate eNOS through ETB receptors, inducing cell proliferation and migration.6,7 Moreover, ET-1 also activates VEGF synthesis,8 and hypoxia is known to induce ET-1 production.9 Taken together, these data suggest that ET-1 could be a proangiogenic agent in a model of ischemia-induced angiogenesis. However, there is no study to confirm this hypothesis in vivo. Furthermore, recent in vivo studies have shown conflicting results: on one hand, a proangiogenic role for ET-1 has been described in a model of microgel plug assay10; on the other hand, a putative proangiogenic property of endothelin antagonists has been found in tumor progression.11
The aim of the present study was then to investigate the in vivo angiogenic role of ET-1 in a model of surgically induced hindlimb ischemia and to define the cellular mechanism by investigating changes in the VEGF-NO pathway. The angiogenic role of ET-1 was assessed by stimulating ETA receptors with a subpressor dose of exogenous ET-1 or by blocking ETA and ETB receptors with bosentan. We also measured the changes in VEGF and in the different isoforms of NO synthases (eNOS, iNOS, and neuronal NO synthase [nNOS]) levels in ischemic tissues.
The right femoral artery of 12-week-old male normotensive Wistar rats (IFFA-CREDO, L’Arbresle, France) was occluded (3-0 silk suture) under pentobarbital anesthesia (50 mg/kg IP). Animals were then randomly allocated to receive for 3 and/or 28 days (1) a placebo (n=8), (2) ET-1 infused at a subpressor dose (2 pmol · kg−1 · min−1 [Sigma Chemical Co] via subcutaneous osmotic minipump [model 2002, Alzet], n=7), (3) dual ETA/ETB antagonist bosentan (100 mg · kg−1 · d−1 in powder chow, kindly provided by Drs Martine and Jean-Paul Clozel, Actelion, Switzerland; n=9), (4) bosentan (100 mg ·kg−1 · d−1) and a nonselective NO synthesis inhibitor NG-nitro-l-arginine methyl ester (L-NAME, 10 mg · kg−1 · d−1 in drinking water) as previously described12 (n=6), or (5) bosentan (100 mg · kg−1 · d−1) and a VEGF-neutralizing antibody (2.5 μg IP twice a week, R&D Systems; n=5).
For each group, systolic arterial blood pressure was measured before angiography by using the tail-cuff method (BP recorder 8006, W+W Electronic).
Quantification of Angiogenesis
Vessel density was evaluated by high-definition microangiography13,14 at the end of the treatment period, as previously described.15 Rats were anesthetized (pentobarbital, 50 mg/kg IP), and a medial laparotomy was performed to introduce a polyethylene catheter into the abdominal aorta. A contrast medium (barium sulfate, 1 g/mL) was injected into the abdominal aorta. Angiography of the right and left hindlimbs was then assessed, and images were acquired by a digital x-ray transducer (20×30 mm). Images were then assembled to obtain a complete view of the hindlimbs (Figure 1). Image analysis was performed by image framing.15 Briefly, images were thresholded and filtered to identify the arteries. The percentage of pixels per image occupied by arteries was determined, in a blinded manner, in the quantification area, which was limited by the place of the ligature on the femoral artery, the knee, the edge of the femur, and the external limit of the leg. Vessel density was also assessed in the nonischemic leg. In previous experiments (data not shown), we determined that in our model, the angiogenic score for vessel density was stable after 4 weeks of femoral artery ligation.
Capillary and Arteriolar Densities
Microangiographic analysis was completed, as previously described,15 by assessment of capillary and arteriolar densities in sections of hindlimb muscles. Briefly, ischemic and nonischemic muscles were dissected and first incubated in PBS containing 5% BSA at room temperature and then incubated for 1 hour with either mouse monoclonal antibody directed against smooth muscle actin 1 (dilution 1:50) to identify arterioles or with rabbit polyclonal antibody directed against total fibronectin (dilution 1:50) to identify capillaries. Arteriole immunohistochemistry was achieved by treating sections with a biotinylated secondary antibody with a horseradish peroxidase–streptavidin conjugate. Capillaries were revealed with a fluorescent FITC anti-rabbit antibody (Figure 2). Capillary and arteriolar densities were then calculated in randomly chosen fields of a definite area by using Optilab/Pro software.
Determination of Protein Expression
Tissue samples dissected from the same area used for angiography quantification were thawed and homogenized in 300 μL of buffer (200 mmol/L sucrose and 20 mmol/L HEPES, pH7.4) containing protease inhibitors. Protein content was then determined by the method of Bradford (Bio-Rad).
Proteins (5 μg for VEGF preparation and 45 μg for NO synthase preparation) were separated in denaturing SDS/12% polyacrylamide gels and then blotted onto a nitrocellulose sheet (Hybond ECL, Amersham). Polyclonal antibodies against the following proteins were used: VEGF (dilution of 1:2000, Santa Cruz), eNOS (1:10 000), nNOS (1:2500), and iNOS (1:2500; NO synthase antibodies were purchased from Transduction Laboratories). Specific protein was detected by chemiluminescent reaction (ECL+ kit, Amersham) followed by exposition of the membranes to Hyperfilm ECL (Amersham). Proteins were then stained with Ponceau red (Sigma) for 5 minutes. Quantifications were performed by densitometric analysis after scanning by using a Bio-Rad gel Doc 1000. Results are expressed as the ratio of quantification of the specific band on autoradiogram to quantification of the transferred total protein bands stained with Ponceau red.
Results are expressed as mean±SEM.
One-way ANOVA was used to compare vessel densities. A value of P<0.05 was considered significant.
Western Blot Analysis
A paired t test was used to compare the expression of proteins in ischemic versus nonischemic hindlimbs. Then, an unpaired t test was performed to compare the effect of treatments on protein expression between groups. A value of P<0.05 was considered significant.
Systolic Blood Pressure
Systolic blood pressure was not significantly modified by treatment with bosentan alone (151±10 mm Hg), by bosentan and VEGF-neutralizing antibody (158±4 mm Hg), or by ET-1 infusion (163±4 mm Hg) compared with no treatment (148±7 mm Hg, P=NS). Only treatment with L-NAME associated with bosentan induced a significant increase in systolic blood pressure (201±6 mm Hg, P<0.05).
Effect of ET-1 and Endothelin Receptor Blockade on Vessel Density
In the nonischemic hindlimbs, neither bosentan nor ET-1 treatment altered the vascular density compared with no treatment whatever the length of treatment (Figure 3). In the ischemic hindlimbs, vessel density was significantly (1.4-fold and 1.7-fold) increased in rats treated with bosentan compared with the nontreated rats (P<0.05) at 3 and 28 days, respectively. Conversely, ET-1 infusion had no impact on vessel density in ischemic legs.
Capillary and Arteriolar Density
Microangiographic data were confirmed by capillary and arteriolar data analysis. In nonischemic hindlimbs, capillary and arteriolar densities were not modified whatever the treatment (Figure 4). At day 3, bosentan treatment increased capillary density in ischemic hindlimbs (583±38 versus 432±45 vessels/mm2 in nontreated rats, P<0.05; Figure 4A). At day 28, capillary and arteriolar densities were increased in ischemic hindlimbs from bosentan-treated rats compared with nontreated rats (772±71 and 21.6±1.9, respectively, versus 506±14 and 14.2±1.0 vessels/mm2, respectively; P<0.05), whereas ET-1 infusion had no effect on vessel densities (481±75 and 14.6±2 vessels/mm2, P=NS; Figures 2 and 4⇓B).
Effect of Endothelin Receptor Blockade on VEGF and NO Synthase Production
In nontreated animals, there was no difference in VEGF, eNOS, and nNOS protein levels between ischemic and nonischemic legs (data not shown).
Treatment with bosentan increased the expression of VEGF and eNOS in ischemic hindlimbs at 3 days (by 31±8% and by 45±23%, respectively, versus no treatment; P<0.05) and 28 days (by 66±13% and by 55±15%, respectively, versus no treatment; P<0.05) but not that of nNOS (Figure 5).
At 28 days, ET-1 infusion decreased the expression of VEGF in ischemic legs (by 26±11% versus no treatment, P<0.05) but had no effect on the expression of eNOS (Figure 5) and nNOS.
We could not detect iNOS protein in ischemic and nonischemic hindlimbs after 3 and 28 days whatever the treatment (data not shown).
Effect of L-NAME and VEGF Antibody on Endothelin Receptor Blockade–Induced Angiogenesis at Day 28
Neither bosentan+L-NAME nor bosentan+VEGF-neutralizing antibody treatment altered the vascular density in nonischemic hindlimbs compared with nontreated hindlimbs 28 days after ischemia. In the ischemic hindlimbs, VEGF-neutralizing antibody and L-NAME treatment associated with bosentan inhibited the increase in vessel density induced by bosentan alone (Figures 1 and 3⇑).
Capillary and Arteriolar Density
Treatment with VEGF-neutralizing antibody or with L-NAME in addition to bosentan abolished the bosentan–induced elevation in capillary and arteriole density in ischemic hindlimbs (437±23 and 15.4±2.0 vessels/mm2, respectively, and 464±12 and 14.6±1.3 vessels/mm2, respectively; Figures 2 and 4⇑B).
Western blot analysis of eNOS expression showed that treatment with VEGF antibody in addition with bosentan inhibited the increase in eNOS expression observed with bosentan treatment alone (Figure 5).
The present study demonstrates that chronic blockade of endothelin receptors is associated with an early and sustained increase of ischemia-induced angiogenesis in vivo and that this effect is mediated by VEGF and eNOS pathways.
Ever since the discovery of endothelin, no in vivo study has clearly investigated the potential role of endothelin in angiogenesis. ET-1 seems to be a proangiogenic agent in Matrigel, in a nonischemic model.10 Conversely, a recent study has described an increase of angiogenesis markers associated with ET-1 receptor blockade in colon tumors.11 In addition, an in vivo study has shown a beneficial effect of ETA/ETB antagonism on survival after myocardial infarction, suggesting a putative improvement of ischemia-induced angiogenesis.16 Furthermore, in a model of critical hindlimb ischemia, endothelin system was activated, whereas endothelin receptor antagonism has been shown to be protective against muscle injury, although no mechanism was proposed.17 Thus, the angiogenic role of endothelin receptors in vivo remains to be defined.
Our present finding clearly demonstrates that ET-1 blockade had a marked proangiogenic effect in an ischemic leg after femoral occlusion. Those results appear to be contradictory with previous in vitro studies that have suggested a proangiogenic role for ET-1. In fact, previous in vitro studies were performed on a single cellular type over a short time range, <24 hours, and did not clearly demonstrate a direct proangiogenic role for ET-1 but only activities that could account for a proangiogenic role (ie, cell proliferation and cell migration).4–6 Moreover, several stimuli are involved in the angiogenic response after femoral artery occlusion, such as ischemia and inflammation.13,15 Indeed, we have previously shown that modulation of the inflammatory response modifies the angiogenic process in the same model of ischemia-induced angiogenesis.15 Then, the apparent discrepancy between our results and previous in vitro studies might indicate that numerous pathways and cell types are involved in the angiogenic response to ischemia in vivo.
Bosentan modulates ischemia-induced angiogenesis by increasing the speed of revascularization and maintaining sustained activation of the process.
A possible mechanism by which bosentan enhances angiogenesis might involve the early (day 3) and sustained (day 28) increase in VEGF and eNOS levels in treated animals. We confirmed this latter hypothesis, inasmuch as we determined that blockade of VEGF pathway and of NO synthesis totally abolished bosentan-enhanced angiogenesis. Moreover, VEGF antibody treatment associated with bosentan completely inhibited the increase in eNOS production observed in rats treated with bosentan alone. Therefore, eNOS lies downstream from VEGF in the proangiogenic process associated with bosentan treatment. Those results are in agreement with previous in vivo studies showing an involvement of NO in the proangiogenic effect of VEGF.2,18
On the other hand, compared with no treatment, ET-1 infusion induced only a slight decrease in VEGF level in ischemic hindlimbs without a change in vascular density. This lack of effect could be due to a nonrelevant change in VEGF expression or to a compensatory mechanism involving other angiogenic agents, such as platelet-derived growth factor or basic fibroblast growth factor. In this view, this decrease in VEGF content was not associated with changes in eNOS protein levels. In addition, eNOS levels measured in the hindlimbs of ET-1–treated rats might correspond to the baseline of expression of the protein. Hence, eNOS protein content in these rats was not different from that of nontreated rats. Therefore, we can speculate that other stimuli regulate baseline eNOS levels and maintain eNOS protein to a physiological level despite the slight decrease in VEGF content.19 Thus, the present results show a strong correlation between angiogenesis, VEGF, and eNOS, as previously suggested by in vitro studies,3,7,18,20 emphasizing a putative common pathway involving VEGF and NO in ischemia-induced angiogenesis in vivo.
One can discuss the impact of hypertension induced by L-NAME treatment on the angiogenic process. In fact, impairment of angiogenesis by high blood pressure remains unclear,21 and there are no data available regarding this interaction. Moreover, blood pressure has been described as having no impact on coronary capillary angiogenesis in hypertensive rats.22
There are no data concerning the role of nNOS in angiogenesis. Our results indicate that after 28 days, neither chronic blockade, activation of ET-1 receptors, nor sustained variations in VEGF production affected nNOS expression. Thus, nNOS might not be involved in the angiogenic process associated with tissue ischemia.
Although hypoxia upregulates iNOS and VEGF levels,3,23 the lack of iNOS expression in ischemic hindlimb after 3 and 28 days may be related (1) to an earlier activation of iNOS expression before 3 days of ischemia and/or (2) to already sufficient revascularization of the leg, resulting in a loss of iNOS activation.
In conclusion, our present study demonstrates for the first time that chronic blockade of ETA/ETB receptors stimulates ischemia-induced angiogenesis in vivo. This effect was directly dependent on VEGF and eNOS pathways. We also determined that chronically infused ET-1 is not proangiogenic in ischemic and nonischemic tissues.
This study was supported by grants from INSERM and Université Paris VII-Denis Diderot.
Received March 16, 2001; revision accepted June 25, 2001.
Yang Z, Krasnici N, Lüscher TF. Endothelin-1 potentiates human smooth muscle cell growth to PDGF: effects of ETA and ETB receptor blockade. Circulation. 1999; 100: 5–8.
Noiri E, Hu Y, Bahou WF, Keese CR, Giaever I, Goligorsky MS. Permissive role of nitric oxide in endothelin-induced migration of endothelial cells. J Biol Chem. 1997; 272: 1747–1752.
Matsuura A, Yamochi W, Hirata K, Kawashima S, Yokoyama M. Stimulatory interaction between vascular endothelial growth factor and endothelin-1 on each gene expression. Hypertension. 1998; 32: 89–95.
Kourembanas S, Marsden PA, McQuillan LP, Faller DV. Hypoxia induces endothelin gene expression and secretion in cultured human endothelium. J Clin Invest. 1991; 88: 1054–1057.
Arnal JF, Warin L, Michel JB. Determinants of aortic cyclic guanosine monophosphate in hypertension induced by chronic inhibition of nitric oxide synthase. J Clin Invest. 1992; 90: 647–652.
Ito WD, Arras M, Scholz D, Winkler B, Htun P, Schaper W. Angiogenesis but not collateral growth is associated with ischemia after femoral artery occlusion. Am J Physiol. 1997; 273: H1255–1265.
Silvestre JS, Mallat Z, Duriez M, Tamarat R, Bureau MF, Scherman D, Duverger N, Branellec D, Tedgui A, Levy BI. Antiangiogenic effect of interleukin-10 in ischemia-induced angiogenesis in mice hindlimb. Circ Res. 2000; 87: 448–452.
Mulder P, Richard V, Derumeaux G, Hogie M, Henry JP, Lallemand F, Compagnon P, Mace B, Comoy E, Letac B, et al. Role of endogenous endothelin in chronic heart failure: effect of long-term treatment with an endothelin antagonist on survival, hemodynamics, and cardiac remodeling. Circulation. 1997; 96: 1976–1982.
Fleming I, Busse R. Signal transduction of eNOS activation. Cardiovasc Res. 1999; 43: 432–541.
Leibovich SJ, Polverini PJ, Fong TW, Harlow LA, Koch AE. Production of angiogenic activity by human monocytes requires an L-arginine/nitric oxide-synthase-dependent effector mechanism. Proc Natl Acad Sci U S A. 1994; 91: 4190–4194.
Tomanek RJ, Searls JC, Lachenbruch PA. Quantitative changes in the capillary bed during developing peak and stabilized cardiac hypertrophy in the spontaneously hypertensive rat. Circ Res. 1982; 51: 295–304.
Jung F, Palmer LA, Zhou N, Johns A. Hypoxic regulation of inducible nitric oxide synthase via hypoxia inducible factor-1 in cardiac myocytes. Circ Res. 2000; 86: 319–325.