OUP user menu

Cyclooxygenase-2 preserves flow-mediated remodelling in old obese Zucker rat mesenteric arteries

Eric J. Belin de Chantemèle , Emilie Vessières , Anne-Laure Guihot , Bertrand Toutain , Laurent Loufrani , Daniel Henrion
DOI: http://dx.doi.org/10.1093/cvr/cvp411 516-525 First published online: 23 December 2009

Abstract

Aims Resistance arteries have a key role in the control of local blood flow and pressure, and chronic increases in blood flow induce endothelium-dependent outward hypertrophic remodelling. The incidence of metabolic syndrome increases with age, and the combination of these two risk factors impairs endothelium integrity, in part through an inflammatory process. We hypothesized that cyclooxygenase-2 (COX2) would affect remodelling in 12-month-old obese rats compared with young rats.

Methods and results Mesenteric arteries of obese and lean Zucker rats were alternatively ligated to generate high flow (HF) in the median artery. After 21 days, arteries were isolated for in vitro analysis. After 21 days, outward hypertrophic remodelling occurred in HF arteries in obese (498 ± 20 vs. 443 ± 18 μm intraluminal diameter in normal flow (NF) arteries, P < 0.01), but not in lean rats (454 ± 17 vs. 432 ± 14, NS; n = 12 per group). Endothelium-dependent (acetylcholine)-mediated relaxation (AMR) was lower in obese than in lean rats. AMR was reduced by NO-synthase blockade in all groups, and eNOS expression was higher in HF than in NF arteries without difference between lean and obese rats. Indomethacin further reduced AMR in HF arteries from obese rats only. Obesity increased COX2 immunostaining in mesenteric arteries. Acute COX2 inhibition (NS398) significantly reduced AMR in HF arteries from obese rats only, suggesting production of vasodilator prostanoid(s). In obese rats chronically treated with the COX2 inhibitor celecoxib, outward remodelling did not occur in HF arteries and AMR was improved without reaching the level found in lean rats.

Conclusion COX2 preserved in part flow-mediated arterial remodelling in old obese rats. Nevertheless, this effect was not sufficient to keep endothelium-dependent relaxation to the level obtained in lean rats.

  • Resistance arteries
  • Shear stress
  • Ageing
  • Cyclooxygenase-2
  • Metabolic syndrome

1. Introduction

The metabolic syndrome is a rising health problem affecting a large proportion of the population worldwide. It is defined as an association of three or more of the following risk factors: impaired glucose tolerance or insulin resistance, increased blood pressure, increased plasma triglycerides, and/or low HDL cholesterol, obesity, and microalbuminuria.1 The main vascular consequences of the metabolic syndrome are endothelial dysfunction2 and vascular remodelling leading to arterial wall hypertrophy.3

Resistance arteries (RAs) play a major role in the control of local blood flow.4 They are affected by the metabolic syndrome because the capacity of the endothelium to produce vasodilator agents is reduced early in this disease in large vessels2 and RAs.5 In addition, the capacity of the RAs to adapt to chronic changes in blood flow, essential in growth, ischaemic events, and exercise is impaired.6 In response to a chronic increase in blood flow, RAs undergo an outward hypertrophic remodelling in association with improved capacity of the endothelium to induce vasorelaxation.7 We have previously shown, in a model of obesity, that RAs submitted to a chronic increase in blood flow in vivo were able to increase their diameter and wall thickness, but without the expected improvement in endothelium-mediated dilation.6 Indeed, the production of reactive oxygen species (ROS), important for vascular remodelling,8 is excessive in obese Zucker rats, a model of metabolic syndrome.

The frequency of the metabolic syndrome increases dramatically with age, and the capacity of RAs to remodel in older obese Zucker rats is still unknown. Ageing per se induces important changes in vascular structure and reactivity.9 Indeed, with age, flow-mediated dilation is altered in RAs.10 Furthermore, RAs lose the capacity to increase their diameter in response to chronic rises in blood flow.11 Nevertheless, RAs from old rats remain able to improve endothelium-mediated dilation in response to a chronic rise in blood flow.12 Thus, the response of RAs to chronic increases in blood flow in old obese Zucker rats cannot be deduced from these previous studies. The first goal of this study will therefore consist in analysing the combine effects of obesity and ageing on the ability of RAs to remodel in response to chronic changes in blood flow.

Both ageing and obesity lead to vascular inflammation. Cyclooxygenase-2 (COX2) derivatives modulate vascular function and more specifically endothelial function. Flow-induced outward remodelling requires a concomitant release of NO and ROS to form peroxynitrite and activate metalloproteinases.8,13 Indeed, inflammation precedes the change in diameter,14 and this initial phase with macrophage infiltration, cytokine production, and COX2 induction might be excessive in obese rats. Studies in obese Zucker rats demonstrate endothelial dysfunction in several vascular beds, associated with a reduction in NO production, an alteration of endothelium-mediated dilation, and an increased vascular oxidative stress,15 as well as an induction of COX2 in older rats.16 The second goal of the current study was therefore to demonstrate that obesity and aged-induced COX2 over-expression would affect vascular remodelling in response to chronic rise in blood flow.

2. Methods

2.1 Animals

Three- and 12-month-old male obese and lean Zucker rats (Charles River) were anaesthetized (Isoflurane, 2.5%) and submitted to surgery in order to modify blood flow in the mesenteric circulation as described previously.8,11 Before anaesthesia, animals were treated with buprenorphine (TEMGESIC®; 0.1 mg/kg, s.c.). In brief, three consecutive first-order arteries were used. Ligatures (7-0 silk surgical thread) were applied to second-order branches of the first and third arteries. The artery located between two ligated arteries was designed as a high-flow (HF) artery. Equivalent arteries located at distance of the ligated arteries were used as control arteries (normal flow, NF). At the end of the surgery, another dose of buprenorphine was administered. As determined by ultrasonic flowmetry (Transonic), arterial ligation resulted in a similar increase in blood flow in lean and obese Zucker rats during the surgical procedure (from 426 ± 24 to 841 ± 37 μL/min in lean rats and from 504 ± 38 to 920 ± 42 μL/min in obese Zucker rats, n = 4 per group, NS). After 21 days, animals were anaesthetized, blood pressure measured17, and the arteries submitted to HF or NF were isolated. In another series of experiments, rats were treated with the COX2 inhibitor celecoxib (25 mg/kg, daily gavage). Twelve rats were used per group, and each experiment described below was performed on a different segment of the artery obtained from each animal. The investigation conforms with 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) and had authorization no. 6422 (protocol approved by the regional ethical committee: CEEA-PdL: Comité d'Ethique en Expérimentation Animale des Pays-de-la-Loire).

2.2 Blood parameters

Before sacrifice, glycaemia was quantified on a sample of arterial blood using a glucometer as described previously.6 Plasma 6keto-PGF and PGE2 were then measured as described previously.18

2.3 Pressure–diameter relationship in mesenteric arteries in vitro

Arterial segments were cannulated at both ends and mounted in a video-monitored perfusion system19 as described previously.20,21 Briefly, cannulated arterial segments were bathed in a 5 mL organ bath containing a Ca2+-free physiological salt solution containing ethylenbis-(oxyethylenenitrolo) tetra-acetic acid (2 mmol/L) and sodium nitroprusside (SNP, 10 μmol/L). Pressure steps (10–150 mmHg) were then performed in order to determine passive arterial diameter. Pressure and diameter measurements were collected using a Biopac data acquisition system (Biopac MP100 and Acqknowledge® software; La Jolla, CA, USA). The arterial segment was then fixed with formaldehyde under a pressure of 75 mmHg in order to measure media cross-section area and wall thickness as described previously.22

2.4 Pharmacological profile of isolated mesenteric arteries

Other arterial segments were mounted on a wire-myograph (DMT, Aarhus, Denmark) as described previously.23,24 Arteries were bathed in a physiological salt solution. Wall tension was applied as described previously.25 Artery viability was tested using a potassium-rich solution (80 mmol/L). Endothelium integrity was evaluated using acetylcholine (1 μmol/L) induced relaxation after phenylephrine (1 μmol/L) induced pre-constriction. Cumulative concentration–response curve to acetylcholine (0.001–10 μmol/L) was performed after phenylephrine-induced pre-constriction (3 μmol/L). The same concentration–response curve was repeated after incubation of the arteries with the NO-synthase inhibitor L-NAME (100 μmol/L, 20 min) and then (after washout) in the presence of L-NAME plus the cyclooxygenase inhibitor indomethacin (10 μmol/L, 20 min).26 In another series of experiments, a concentration–response curve to acetylcholine (0.001–10 μmol/L) was performed before and after incubating arterial rings with the selective COX2 inhibitor NS398 (10 μmol/L, 20 min)26 or with tetrahydrobiopterin (10 μmol/L) plus l-arginine (100 μmol/L). Endothelium-independent relaxation to SNP was obtained at the end of the protocol.27

The remaining arterial segments were pooled and then homogenized. Proteins (25 μg total protein from each sample) were separated by SDS–PAGE using a 4% stacking gel followed by a 10% running gel. Proteins were detected with specific antibodies (Transduction Laboratories: COX2 1 : 200, eNOS 1 : 1000, phospho-eNOS (S1177) 1 : 1000, PGE2 synthase 1 : 100, PGI2 synthase and actin 1 : 1000 in TBST). Protein expression was visualized using the ECL-Plus Chemiluminescence kit (Amersham).6

2.5 Histological analysis of COX2

Arterial segments were mounted in embedding medium (Tissu-Tek, Miles, Inc.), frozen in isopentane pre-cooled in liquid nitrogen, and stored at −80°C. COX2 was detected on transverse cross-sections (7 μm thick) using primary goat anti-COX2 polyclonal antibodies (1/200, Santa Cruz Biotechnology), followed by the fluorescent secondary antibody (1/200, Fluoroprobes) as shown previously.11,26 In negative control experiments, the primary antibody was omitted. A positive control was obtained with the mesenteric arteries of a rat treated with lipopolysaccharide. Positive staining was visualized using confocal microscopy and QED-Image software (Solamere Technology). Image analysis was performed using Histolab (Microvision). The auto-fluorescent internal and external laminas, excluded for the fluorescence quantification, were used to delimit the media. Care was taken to take all the pictures in the same condition of laser power, gain, and exposure time.

2.6 Statistical analysis

Results are expressed as means ± SEM. Significance of the difference between arteries was determined by ANOVA (ANOVA for repeated measurements for the comparison of concentration–response curves). Means were compared by paired t-test or by the Bonferroni test for multigroup comparisons (STATVIEW). Values of P < 0.05 were considered to be significant.

3. Results

3.1 Physiological parameters

As reported in Table 1, in 3-month-old animals, obese Zucker rats present a higher body weight and plasma cholesterol level compared with lean rats. In 12-month-old animals, body weight, mean arterial pressure, blood glucose, triglycerides, and cholesterol were significantly higher in obese than in lean rats. Chronic celecoxib had no significant effect on these parameters. Plasma level 6keto-PGF was significantly enhanced in old obese Zucker rats but reduced by chronic celecoxib (Table 1). No changes in plasma PGE2 levels were observed with obesity, ageing, or the combination of both (Table 1). These results are supported by an increased PGI2 level in old obese rats, together with a rise in monocyte chemotactic protein-1 and intercellular adhesion molecule-1 (Supplementary material online, Figure S1).

View this table:
Table 1

Body weight, mean arterial pressure (MAP), and blood parameters measured in 3- (3 m) and 12-month-old (12 m) lean and obese Zucker rats

VehicleCelecoxib
Lean (3 m)Obese (3 m)Lean (12 m)Obese (12 m)Lean (12 m)Obese (12 m)
Body weight (g)335 ± 11527 ± 28*502.2 ± 12.6628.3 ± 34*474.2 ± 14.3603.8 ± 15.3*
MAP (mmHg)97 ± 3101 ± 5105.8 ± 6.4120.6 ± 4.5*101 ± 2.7118 ± 5.4*
Glucose (mmol/L)8.6 ± 0.6889.58 ± 0.8812.6 ± 1.417.2 ± 1.9*11.5 ± 1.918.2 ± 1.1*
Triglycerides (mmol/L)7.2 ± 1.810 ± 1.97.5 ± 1.112.3 ± 1.5*7.0 ± 1.812.2 ± 2.0*
Cholesterol (mmol/L)4.1 ± 0.28.4 ± 0.2*4.4 ± 0.211.3 ± 1.5*4.8 ± 1.314.7 ± 2.5*
6keto-PGF (mg/dL)146 ± 20168 ± 28155 ± 23421 ± 40*132 ± 28202 ± 31*#
PGE2 (mg/dL)56 ± 1165 ± 1360 ± 1071 ± 1452 ± 965 ± 10
  • Rats were treated with celecoxib or vehicle. Values are mean ± SEM (n = 12 per group).

  • *P < 0.05, obese vs. lean rats.

  • #P < 0.05, celecoxib vs. 12-month-old rats treated with vehicle.

3.2 Pressure–diameter relationship in mesenteric arteries

Increasing intraluminal pressure induced a significant increase in arterial diameters of isolated vessels. In NF arteries, obesity significantly decreased arterial diameter in young animals while it did not further change intraluminal diameter, already enhanced by ageing (Figure 1A). In response to the chronic increase in blood flow, young lean and obese animals enhanced similarly their arterial diameter (Figure 1B). However, in 1-year-old animals, obese rats only presented an increase in arterial diameter in response to HF. Indeed, arterial diameter was higher in HF arteries compared with NF in obese rats only (Figure 1A and B). As reflected by a higher media cross-sectional area in the HF arteries, both young and old lean and obese Zucker rats presented hypertrophic remodelling in response to HF. Nevertheless, the media cross-sectional area was significantly higher in obese than in lean rats in both NF and HF arteries (Figure 1C).

Figure 1

(A) Changes in arterial diameter in response to stepwise increases in pressure in HF and NF mesenteric RAs isolated from 3- and 12-month-old obese and lean Zucker rats. (B) Changes in diameter in HF arteries compared with NF vessels in 3- and 12-month-old lean and obese Zucker rats. (C) Media cross-sectional area. Values are mean ± SEM (n = 12 per group). *P < 0.01, HF vs. NF in each group. #P < 0.01, obese vs. lean.

3.3 Endothelium-dependent dilation and eNOS expression

As the endothelium and eNOS have a key role in flow-mediated remodelling, we assessed eNOS activity indirectly via the measurement of eNOS and phospho-eNOS (S1177) expression levels as well as through the evaluation of the endothelial (acetylcholine)-mediated relaxation (AMR). Neither obesity nor ageing affected eNOS expression level in NF arteries, but eNOS expression level was higher in all HF arteries compared with NF arteries but without difference between groups (Figure 2A). Similarly, while looking at eNOS phophorylation (S1177), we observed a similar increase between groups, suggesting that obesity, ageing, or the combination of both did not affect blood flow-induced eNOS activation (Figure 2B).

Figure 2

(A) Endothelial NO-synthase (eNOS, A) and phosphorylated-eNOS (B) expression levels were measured using western blot analysis in HF and NF mesenteric RAs isolated from 3- and 12-month-old obese and lean Zucker rats. Data are shown as ratio eNOS/beta-actin (values for the intensity of the beta-actin bands are given on each bar). Concentration–response curves to acetylcholine performed in the presence of L-NAME or L-NAME plus indomethacin in NF (C) and HF (D) arteries taken from 3- and 12-month-old lean and obese Zucker rats. Values are mean ± SEM (n = 12 per group). *P < 0.01, HF vs. NF in each group. #P < 0.01, obese vs. lean. &P < 0.01, L-NAME plus indomethacin vs. L-NAME.

Acetylcholine induced a concentration-dependent relaxation in isolated mesenteric arteries. No endothelial dysfunction was observed in NF and HF arteries isolated from 1-year-old lean rats (Figure 2C and D; Supplementary material online, Figure S2). However, obesity reduced AMR in both NF and HF arteries isolated from young and old obese rats (Figure 2C and D; Supplementary material online, Figure S2). In both young and old lean rats, relaxation to acetylcholine tended to be higher in HF compared with NF arteries without reaching significance (Figure 2C and D). In young and old obese Zucker rats, the chronic increase in blood flow impeded endothelium-dependent relaxation instead of improving it. Indeed, HF arteries taken from young and old obese rats presented a lower relaxation to acetylcholine compared with NF arteries (Figure 2C and D).

As reported in Figure 2C and D, inhibition of NOS with L-NAME reduced acetylcholine-dependent relaxation in all groups. However, the inhibitory effect of L-NAME was increased in the HF arteries of young lean rats, whereas it was decreased in those of the young obese rats, suggesting a reduced involvement of NO in the endothelium-dependent relaxation of these arteries (Supplementary material online, Figure S3). Furthermore, ageing abolished HF-induced increase in NO participation (inhibitory effects of L-NAME) in lean Zucker rats. In obese rats, ageing did not further affect the inhibitory effect of L-NAME (Supplementary material online, Figure S3). Indeed, no difference was observed between arteries of the young and 1-year-old obese Zucker rats. To determine the involvement of COX derivatives in the endothelium-dependent relaxation, concentration–response curves to acetylcholine have been repeated in the presence of L-NAME and the unspecific COX inhibitor, indomethacin. In the presence of L-NAME, indomethacin further reduced the relaxation of NF and HF arteries with the exception of the arteries from old obese Zucker rats (Figure 2C and D; Supplementary material online, Figure S3).

To determine whether eNOS uncoupling could explain the endothelial dysfunction, concentration–response curves to acetylcholine were repeated in the presence of the eNOS cofactor tetrathydrobiopterin and its substrate, l- arginine (Supplementary material online, Figure S4) or in the presence of the antioxidant tempol (Supplementary material online, Figure S5). In agreement with our previous study performed with 3–month-old obese Zucker rats,6 data from this study confirmed that supplementation in tetrahydrobiopterin and l-arginine or tempol restored endothelial function in both NF and HF arteries taken from 3- and 12-month-old obese Zucker rats.

3.4 COX2 expression

COX2 expression was barely detectable in NF and HF arteries taken from young lean and obese rats. However, in 12-month-old rats, obesity increased COX2 expression independent of changes in blood flow (Figure 3A). These results have been confirmed by measuring COX2 expression level by western blot analysis (Figure 3B).

Figure 3

Representative pictures (A) and quantification of COX2 immunostaining performed in HF and NF arteries isolated from 3- and 12-month-old lean and obese Zucker rats. (B) Representative blots and quantification of COX2 expression in arteries isolated from 12-month-old lean and obese Zucker rats. Values are mean ± SEM (n = 12 per group). #P < 0.01, obese vs. lean.

3.5 Role of COX2 in vascular reactivity

Acute treatment of the isolated arteries with NS398 did not affect the endothelial function of young animals and had no effect on the AMR in NF and HF arteries isolated from old lean rats (Figure 4). However, in 12-month-old obese Zucker rats, NS398 significantly increased relaxation in NF arteries, whereas it decreased relaxation in HF arteries (Figure 4). These latter results, respectively, suggest that the association of obesity and ageing promoted the production of constrictor COX2 derivatives, whereas in aged obese rats, the chronic rise in blood flow increased the participation of dilator prostaglandins.

Figure 4

Concentration–response curves to acetylcholine performed in the absence and in the presence of NS398 (10 μmol/L, 20 min) in NF (A and C) and HF (B and D) arteries isolated from 3- (A and B) and 12-month-old (C and D) obese and lean Zucker rats. Values are mean ± SEM (n = 12 per group). #P < 0.01, obese vs. lean. &P < 0.01, NS398 vs. vehicle.

Neither changes in blood flow nor acute treatment with NS398 affected phenylephrine-induced constriction in young animals (Supplementary material online, Figure S6). Similarly, in 12-month-old animals, neither obesity nor the chronic inhibition of COX2 affected phenylephrine-induced constriction of NF arteries. In aged animals, obesity significantly reduced the phenylephrine-induced contraction of HF arteries. In these latter animals, acute treatment with NS398 restored the phenylephrine-induced constriction of the HF arteries to the level of the lean animals, suggesting that phenylephrine-induced constriction was counterbalanced by dilator prostaglandins in old obese animals. SNP-induced relaxation was not affected by the different experimental conditions investigated (Supplementary material online, Figure S7).

3.6 COX2 and vascular remodelling

To determine the role of COX2 in flow-mediated remodelling in old obese Zucker rats, 12-month-old obese Zucker rats were chronically treated with celecoxib. The treatment completely abolished the remodelling induced by the chronic increased in blood flow and restored the intraluminal diameter to the level of the NF arteries of old lean and obese Zucker rats (Figure 5A). These results suggest that HF-induced remodelling is COX2-dependent in old obese Zucker rats. Moreover, in celecoxib-treated rats, media cross-sectional area was higher in HF than in NF arteries and higher in obese than in lean rats (Figure 5B).

Figure 5

Internal diameter (A) and cross-sectional area (B) in HF and NF mesenteric arteries isolated from 12-month-old obese and lean Zucker rats treated chronically with the COX2 inhibitor celecoxib. Acetylcholine-dependent relaxation was measured in NF (C) and HF arteries (D) in the presence of L-NAME and in the presence of L-NAME plus indomethacin. Maximal relaxation (E) obtained in response to acetylcholine (calculated from Figure 4A and 5CD). Values are mean ± SEM (n = 12 per group). *P < 0.01, HF vs. NF in each group. #P < 0.01, obese vs. lean. &P < 0.01, celecoxib vs. vehicle.

In obese rats, chronic treatment with celecoxib reduced acetylcholine-dependent relaxation in HF to a level lower to that observed in NF arteries. In lean rats, acetylcholine-dependent relaxation was significantly higher in HF than in NF arteries, without reaching significance (Figure 5CE).

4. Discussion

We found that the presence of COX2 in old obese rats allowed remodelling to occur despite age and prevented, in part, the reduction in endothelium-mediated relaxation associated with obesity and ageing. In physiological conditions, vascular adaptation to a chronic increase in blood flow includes an increased diameter and wall thickness and an improved endothelial function.8,28 This adaptation of the microvasculature in response to a chronic increase in flow has a major role in pathological processes such as hypertension, ischaemic diseases, and diabetes. This remodelling is also highly involved in growth, pregnancy, and exercise, the last of which is recommended for obese patients.

As reported previously,11 ageing blunts diameter enlargement in response to chronic increases in blood flow. Nevertheless, a beneficial effect of the chronic rise in flow remains with an increased endothelium-mediated dilation.11,12 Surprisingly, the combination of ageing and obesity, in old obese Zucker rats, leads to a significant increase in diameter in response to a chronic rise in blood flow. However, the increase in flow further decreases endothelium-mediated relaxation in obese rats, whereas it improved it in old lean rats, as described previously.11

One-year-old obese Zucker rats are resistant to insulin, slightly hypertensive, obese, and dyslipidaemic, with some degree of vascular dysfunction.29 In metabolic syndrome, obesity,30 hyperinsulinaemia,31 or both are associated with an alteration of endothelium-dependent relaxation in large arteries.32 Although the obese Zucker rat is considered a good animal model to study the metabolic syndrome, the effects of ageing on this model remain rarely studied.

As the capacity of the endothelium to produce NO is important for flow-induced diameter enlargement,11 we measured eNOS expression levels and eNOS phosphorylation (S1177) in HF arteries. Expression levels of total or phophorylated eNOS were found to be similarly increased, with flow, in aged lean and obese rats, and the inhibitory effect of L-NAME on AMR was preserved in HF arteries from aged obese rats (Figure 2). In fact, mesenteric arteries from obese rats did not present a significant change in eNOS expression, in agreement with previous observations in rat aorta.33 However, arteries from obese rats have a lower capacity to produce NO34 that is certainly worsened by the high ROS level found in obese rats, leading to a reduced NO bioavailability as previously shown.35,36 This is also supported by the uncoupled eNOS observed in young6 and old (present study) obese Zucker rats. Nevertheless, in the present study, this assumption is only supported by indirect observation based on the capacity of l-arginine associated with tetrahydrobiopterin to improve endothelium-mediated relaxation. All together, these results may explain, at least in part, the difference in relaxation observed between obese and lean rats despite equal eNOS expression levels.

In this study, although we observed that indomethacin did not affect endothelium-dependent relaxation in young animals, we reported different effects of this cyclooxygenase inhibitor between old lean and old obese Zucker rats. Indeed, indomethacin further reduced the relaxation in HF arteries from obese rats without affecting NF vessels. This may suggest the involvement of dilator prostanoids in the relaxation. Indeed, we found that blood 6-keto-PGF, the stable metabolite of prostacyclin, was higher in obese than in lean rats. The positive COX2 immunostaining as well as the increased expression level of this latter cyclooxygenase, reported in the arteries isolated from obese Zucker rats, support the participation of COX2 derivatives in the vascular relaxation. Our data are in agreement with previous studies reporting higher levels of COX2 in renal arteries from young obese Zucker rats.37 Even if COX2 expression is mainly associated with vascular hypercontractility,38 it may also exert a protective effect against endothelial dysfunction. Indeed, an increased COX2 expression and prostaglandin-mediated relaxation have been reported in coronary arteries from old patients with type 2 diabetes.39 Furthermore, COX2 over-expression is beneficial for prevention of vascular dysfunction induced by streptozotocin-induced type 1 diabetes and pulmonary hypertension.40,41

As the presence of COX2 was associated with a vasodilatory effect in the HF arteries of our 1-year-old obese Zucker rats, we hypothesized that COX2 may be involved in flow-induced remodelling and notably improve it. We then performed the ligation of the mesenteric arteries in old obese Zucker rats chronically treated with the COX2 inhibitor, celecoxib. After 3 weeks of ligation, no diameter enlargement was found in old obese Zucker rats, suggesting a key role of COX2 in this process. This is in agreement with a previous study reporting that COX2 inhibition or PGI2 receptor deletion enhances blood flow reduction after carotid artery ligation.42 The current study is also in agreement with previous data showing that inflammation is necessary for the dissociation of the extracellular matrix and the diameter enlargement in response to an increased flow, although macrophages and cytokines are not detected in the arterial wall 4 days after ligation.43 In the present study, acute COX2 inhibition with NS398 reduced the endothelium-dependent relaxation of HF arteries. These results clearly demonstrated the participation of COX2-derived vasorelaxing agents in acetylcholine-induced relaxation. The release of dilator agents in HF arteries of aged obese Zucker rats was further confirmed by the blunted phenylephrine-mediated constriction of these latter arteries, which was restored with acute NS398 infusion. All together, these data report that chronic rise in blood flow stimulates the release of dilator prostanoids in the arteries of old obese rats.

In both young and aged obese Zucker rats, chronic rise in blood flow further reduced AMR. These observations, previously reported in one of our studies,6 are in agreement with the decrease in relaxation observed in HF arteries isolated from young Zucker diabetic fatty rats.8 In these two latter studies, we found that an excessive superoxide production in HF arteries reduces endothelium-dependent relaxation. In the present study, we showed that the presence of COX2 prevented, in part, the reduction in relaxation in the HF arteries. Thus, COX2-derived vasorelaxing agents reduced the fall in endothelium-mediated dilation. Nevertheless, AMR remained much lower than in lean rats. An excessive superoxide production most likely induces this effect, as shown in younger obese Zucker rats.6

While acute infusion of NS398 reduced the endothelial function of HF arteries taken from aged obese Zucker rats, this specific COX2 inhibitor improved AMR in NF arteries. A similar improvement in endothelial function with COX2 inhibition has been reported in aorta isolated from old hamsters.16 These results suggest that chronic increase in blood flow, in aged obese Zucker rats, induced a switch from the secretion of constrictor COX2 derivatives to dilator prostanoids. The difference between NF and HF arteries is probably the result of flow-induced activation of vasodilator processes28,44,45 as well as the result of the flow-mediated paradoxical production of ROS.13 Nevertheless, the resulting response in ageing is improved relaxation in response to both flow and vasodilators (present study and Dumont et al.11).

In contrast to acetylcholine, SNP-induced relaxation was neither affected by changes in flow nor by obesity, ageing, or the combination of both. These results indicate that the responsiveness of vascular smooth muscle cells to NO was not affected by obesity, ageing, or by alterations in flow. Phenylephrine-induced contraction was slightly higher in HF than in NF arteries in lean rats. Indeed, the increased wall mass may induce this higher contractility. On the other hand, the contraction was lower in HF arteries from obese rats and was restored to control (lean HF) level by COX2 blockade used either acutely or chronically. This further supports the existence of vasodilator prostanoids derived from COX2 in HF arteries from obese rats.

4.1 Perspective

The capacity of RAs to remodel in response to a chronic increase in blood flow is essential in both physiological and pathological conditions, especially in ageing when ischaemic diseases become more frequent. We found that the association of obesity with ageing further deteriorates endothelium-dependent relaxation in arteries submitted to a chronic increase in blood flow. Although arteries from old obese rats remained able to increase their diameter, this excessive reduction in endothelium-mediated relaxation is worrying despite the apparent protective effect of COX2. Indeed, the hypothesis that stimulation by shear stress improves endothelial function via increasing NO bioavailability in both young and old subjects is not verified in obesity, especially when associated with ageing. Most cardiovascular diseases, such as hypertension, diabetes, and atherosclerosis, share common features with ageing-associated endothelial dysfunction. Thus, our results suggest that blood flow may not act as an efficient stimulus to improve endothelial function in ageing as well as in other cardiovascular diseases if associated with obesity or metabolic syndrome. The role of COX2 in endothelium-dependent relaxation is in agreement with previous studies in hypertension26 or obesity46 in young animals. In addition, the apparent vascular protective effects of COX2 might be one more explanation for the development of side effects with the chronic use of COX2 inhibitors and might notably be involved in the increased incidence of cardiovascular events.47,48

5. Conclusion

We found that COX2 preserved in part the capacity of the mesenteric RAs to respond to a chronic increase in blood flow in old obese rats. This observation might be taken as a positive effect of COX2 as the responsiveness of RAs to a chronic increase in blood flow is essential in ischaemic diseases, more frequent in ageing. Nevertheless, whether this observation may be extended to other vascular territories remains to be determined.

Funding

This work was supported in part by the Foundation for Medical Research (FRM), Paris, France. E.B.C. was a post-doctoral fellow of the Centre National d'Etudes Spatiales (CNES), France.

Acknowledgements

The authors gratefully acknowledge the assistance of Mrs Jessica Osmond in the correction of the English language in this manuscript.

We thank the local Animal Care Unit of the University of Angers and Jérôme Roux, Pierre Legras and Dominique Gilbert for their kind help in treating the rats.

Conflict of interest: none declared.

References

View Abstract