Cyclooxygenase-2–Derived Prostaglandin F2α Mediates Endothelium-Dependent Contractions in the Aortae of Hamsters With Increased Impact During Aging
Hypertension and vascular dysfunction result in the increased release of endothelium-derived contracting factors (EDCFs), whose identity is poorly defined. We tested the hypothesis that endothelial cyclooxygenase (COX)-2 can generate EDCFs and identified the possible EDCF candidate. Changes in isometric tension of aortae of young and aged hamsters were recorded on myograph. Real-time changes in intracellular calcium concentrations ([Ca2+]i) in native aortic endothelial cells were measured by imaging. Endothelium-dependent contractions were triggered by acetylcholine (ACh) after inhibition of nitric oxide production and they were abolished by COX-2 but not COX-1 inhibitors or by thromboxane–prostanoid receptor antagonists. 2-Aminoethoxydiphenyl borate (cation channel blocker) eliminated endothelium-dependent contractions and ACh-stimulated rises in endothelial cell [Ca2+]i. RT-PCR and Western blotting showed COX-2 expression mainly in the endothelium. Enzyme immunoassay and high-performance liquid chromatography-coupled mass spectrometry showed release of prostaglandin (PG)F2α and prostacyclin (PGI2) increased by ACh; only PGF2α caused contraction at relevant concentrations. COX-2 expression, ACh-stimulated contractions, and vascular sensitivity to PGF2α were augmented in aortae from aged hamsters. Human renal arteries also showed thromboxane–prostanoid receptor–mediated ACh- or PGF2α-induced contractions and COX-2–dependent release of PGF2α. The present study demonstrates that PGF2α, derived from COX-2, which is localized primarily in the endothelium, is the most likely EDCF underlying endothelium-dependent, thromboxane–prostanoid receptor–mediated contractions to ACh in hamster aortae. These contractions involved increases in endothelial cell [Ca2+]i. The results support a critical role of COX-2 in endothelium-dependent contractions in this species with an increased importance during aging and, possibly, a similar relevance in humans.
Besides neuronal and hormonal regulation, vascular tone is modulated locally by a delicate balance between endothelium-derived relaxing (EDRFs) and contracting (EDCFs) factors,1,2 with the latter being less well-defined but emerging in hypertension, obesity, hyperlipidemia, diabetes, and aging.3
A number of molecules have been proposed as possible EDCF candidates under pathophysiological conditions. These include prostaglandin (PG)H2, thromboxane (TX)A2, leukotrienes, endothelin 1, and superoxide anions. The release of these tentative EDCFs can be triggered by acetylcholine (ACh), angiotensins I/II, ADP, and ATP.4 The contribution of additional cyclooxygenase (COX)-derived metabolites, ie, PGE2, PGD2, and PGF2α, has been postulated. The precise nature of these EDCFs varies among species and vascular beds.1,3
Two isoforms of COX have been identified in blood vessels. COX-1 is constitutively expressed and believed to participate in physiological responses, whereas COX-2 is a highly inducible enzyme.5 At least in the rat aorta, EDCFs appear to be COX-1–derived prostanoids generated in the endothelium, which diffuse to contract the underlying vascular smooth muscle by activating thromboxane–prostanoid (TP) receptors.3 In arteries of spontaneously hypertensive or diabetic rats, the expression of COX-1 is upregulated, and the augmented endothelium-dependent contractions are inhibited by COX-1 inhibitors.6–8 COX-1–derived prostacyclin, TXA2, or endoperoxides all contribute to endothelium-dependent contractions.7,9
However, this generally accepted distinction between “constitutive” and “inducible” isoform of COX appears to be an overgeneralization. Indeed, COX-2 can be expressed constitutively in the endothelium of the rat pulmonary and human renal blood vessels and in cultured endothelial cells.10,11 A COX-2–specific inhibitor attenuates arachidonic acid–induced vasodilatation of canine coronary arteries,12 supporting a physiological role for COX-2 in vascular function. COX-2 is upregulated under pathological conditions including renovascular hypertension,13 reflux nephropathy,14 and diabetes.15 For instance, the elevated arteriolar tone and blood pressure in type 2 diabetic mice is associated with the augmented production of COX-2–derived vasoconstrictor prostanoids,16 even though the source of this production is unclear. In deoxycorticosterone acetate salt–induced hypertension, the expression of COX-2 is enhanced and this is related to the increased contraction of the aorta to ACh, probably because of the exaggerated oxidative stress in the vascular wall.17
COX-2 can be upregulated by physiological shear stress from pulsatile flow.18,19 However, its actual role in the endothelial regulation of the normal vascular tone is uncertain. Identification of COX-2–mediated generation of EDCFs can help to elucidate the cellular mechanisms of endothelial dysfunction and potentially uncover novel therapeutic targets. Because the lipid profile and arachidonic acid metabolism of hamsters resemble that of humans20–22 and COX-2 may be important for both the physiological and pathological regulation of vascular reactivity,5,18,19 we hypothesized that COX-2 rather than COX-1 is mediating the generation of EDCFs in the aorta of young and healthy hamsters and that COX-2 expression and COX-2–mediated vascular responses increase with aging. We also studied whether this pathway is relevant to humans by studying human renal arteries. The present findings revealed PGF2α as a physiological EDCF, which can be generated by COX-2 in the endothelium and is of increasing importance during aging.
Materials and Methods
Most experiments were performed on aortae from young and aged (≈3- or >18-month-old) hamsters. This part of the study was approved by the Animal Ethics Committee, Chinese University of Hong Kong. Human renal arteries were obtained during surgery after informed consent from 4 patients aged 59 to 75 years. An expanded Materials and Methods section can be found in the online data supplement at http://circres.ahajournals.org.
Isometric Force Measurement
Blood vessels were prepared as described in the online data supplement. Briefly, to visualize the endothelium-dependent contractions, aortic rings with endothelium were exposed for 30 minutes to 100 μmol/L NG-nitro-l-arginine methyl ester (L-NAME) to eliminate the relaxant effect of endothelium-derived nitric oxide (NO) before the application of ACh (0.1 to 10 μmol/L). This contraction was absent in rings without endothelium. The effects of various inhibitors and antagonists (eg, COX-1 and COX-2 inhibitors, TP receptor antagonists) were tested on ACh-induced endothelium-dependent contractions following a 30-minute incubation with each drug. Specificity of these inhibitors and antagonists was tested against contraction induced by 60 mmol/L KCl or 50 nmol/L U46619.
In Situ Endothelial Cell [Ca2+]i Imaging
A calcium imaging technique was used to visualize real-time changes in intracellular calcium levels ([Ca2+]i) in native endothelial cells of the intact hamster aorta.23
Reverse Transcription–Polymerase Chain Reaction
Expression levels of COX-2 mRNA in rings with and without endothelium were detected by RT-PCR (see the online data supplement).
Primers for PCR were COX-2 (216-bp) sense (5′-TGA TCC CCA AGG CAC GAA-3′) and antisense (5′-ACC TCT CCA CCA ATG ACC TGA-3′)24 and GAPDH (171-bp) sense (5′-ACC CAG AAG ACT GTG GAT GG-3′) and antisense (5′-CAC ATT GGG GGT AGG AAC AC-3′). Melting temperature for COX-2 and GAPDH primers were 60°C and 57°C, respectively. PCR products were run on 1.5% agarose gel in 1×Tris-acetate-EDTA buffer at 80 V. Ethidium bromide–stained bands were visualized under UV illumination using FluorChem (version 2.00, Alpha Innotech Corp, San Leandro, Calif).
Expression of COX-1, COX-2, platelet endothelial cell adhesion molecule (PECAM)-1, F-series–prostanoid (FP) receptor, and TP receptor protein in aortic rings was determined by Western blot analysis (see the online data supplement).
Enzyme Immunoassay and High-Performance Liquid Chromatography–Coupled Mass Spectrometry Measurement of Prostaglandins
PGF2α, PGE2, PGD2, 6-keto PGF1α (for PGI2), TXB2 (for TXA2), and 8-isoprostanes were assayed. Details can be found in the online data supplement.
Detection of Reactive Oxygen Species Formation
Reactive oxygen species (ROS) were detected by electron paramagnetic resonance (EPR). The detailed method can be found in the online data supplement.
Chemicals and drugs can be found in the online data supplement.
Endothelium-dependent contractions were expressed as active tension [force recorded/(2×ring length)]. Results are means±SEM of n rings from different animals. For statistical analysis, Student’s t test or 2-way ANOVA, followed by Bonferroni post tests were used when more than 2 treatments were compared (GraphPad Software, San Diego, Calif). P<0.05 was considered significantly different.
Essential Role of COX-2 in Endothelium-Dependent Contractions in Aortae From Young Hamsters
In the presence, but not in the absence of NG-nitro-l-arginine methyl ester (L-NAME), ACh elicited pronounced contraction of aortic rings with endothelium, with a maximal response of 5.93±0.16 mN/mm (Figure 1A and 1B), corresponding to ≈70% of the contractile response (8.55±0.32 mN/mm) induced by 60 mmol/L KCl. Removal of the endothelium abolished the contractions to ACh (Figure 1C). The endothelium-dependent contractions were attenuated or eliminated by the nonselective COX inhibitor indomethacin (Figure 2A). Likewise, 3 structurally different selective COX-2 inhibitors (NS-398, DuP-697, and celecoxib) reduced or abolished the endothelium-dependent contractions (Figure 2B through 2D). The specificity of COX-2 inhibition was confirmed by the lack of inhibitory effects of the 3 inhibitors on contractions induced by 60 mmol/L KCl and U46619 (supplemental Figure II, A and B). By contrast, neither the COX-1 selective inhibitors (valeryl salicylate [VAS] and sc-560) nor the inhibitor of 5- and 12-lipoxygenase (baicalein) inhibited the response (supplemental Figure I, A through C). The endothelium-dependent contractions were unaffected by treatment with actinomycin-D (10 μmol/L, RNA synthesis inhibitor) or cycloheximide (10 μmol/L, protein synthesis inhibitor) (supplemental Figure I, D).
Endothelium-Dependent Contractions Mediated Through TP Receptors
The endothelium-dependent contractions were attenuated or abolished in aortic rings treated with 3 structurally distinct selective TP receptor antagonists, terutroban (S18886, 3 to 100 nmol/L), L-655,240 (0.1 to 1 μmol/L), or GR 32191 (100 nmol/L) (Figure 3A through 3C). On the contrary, the thromboxane synthase inhibitor ozagrel hydrochloride (10 μmol/L) did not affect the contraction (Figure 3D). The specificity of the TP receptor antagonists was tested against contractions induced by 60 mmol/L KCl and U46619. Treatment with these antagonists inhibited or prevented the U46619-induced contraction without affecting that to 60 mmol/L KCl (supplemental Figure II, C and D).
Dependency on Extracellular Ca2+
Endothelium-dependent contractions were absent following the removal of extracellular calcium ions. Reintroduction of 2.5 mmol/L CaCl2 to the bathing solution restored contraction to 10 μmol/L ACh (Figure 4A). Exposure of rings to 2-aminoethoxydiphenyl borate (2-APB) (3 to 50 μmol/L, a nonselective cation channel blocker) diminished or abolished the endothelium-dependent contractions (Figure 4B) without affecting the response to 60 mmol/L KCl or U46619 (n=4).
En face fluorescence images from viable individual native endothelial cells of cut-open aortic segments were examined (Figure 4C, a). The fluorescence signal indicative of the [Ca2+]i was absent after mechanical removal of the endothelium. It increased following the addition of ACh in the presence of L-NAME only in arterial tissues with endothelium (Figure 4C, a*). Treatment with 2-APB (50 μmol/L, the concentration that abolished endothelium-dependent contractions), prevented the increases in [Ca2+]i (Figure 4C, b*). By contrast, S18886 (0.1 μmol/L) had no effect on the Ca2+ fluorescence signal (Figure 4C, c*). The ACh-stimulated real-time increase in endothelial cell [Ca2+]i was eliminated by 2-APB but not by S18886 (Figure 4D).
Localization of COX-2
The expression of COX-2 mRNA was significantly higher in aortae with endothelium than those without (Figure 5A). The COX-2 protein expression was reduced following the mechanical removal of the endothelium (Figure 5B), which we had confirmed by the reduced protein levels of the endothelium-specific marker PECAM-1 (Figure 5C). The protein expression of COX-1 was slightly but insignificantly greater in the aortae with endothelium than those without (supplemental Figure I, E).
Role of ROS
Endothelium-dependent contractions to ACh were unaffected by tiron (1 mmol/L) plus diethyldithiocarbamate acid (100 μmol/L) (membrane-permeable free radical scavengers), tempol (100 μmol/L, superoxide dismutase mimetic), or apocynin (100 μmol/L, NADPH oxidase inhibitor) (supplemental Figure III, A). L-NAME–treated aortic rings showed no EPR signal for superoxide anions or peroxynitrite in response to ACh (supplemental Figure III, B and C), whereas the addition of hypoxanthine plus xanthine oxidase (HXXO, a mixture to release superoxide anions) gave rise to 3 distinct EPR signals (supplemental Figure III, D).
PGF2α As the EDCF
Six possible EDCF candidates, ie, PGF2α, PGE2, TXA2, PGD2, PGI2, and 8-isoprostanes were assayed chemically. ACh at 3 μmol/L stimulated a significant rise in the release of PGF2α and PGI2 (detected as 6-keto PGF1α) but not PGE2, TXA2 (detected as TXB2), and PGD2 from aortic rings with endothelium (Figure 6A and 6B). The release of both PGF2α and 6-keto PGF1α was largely inhibited by removal of the endothelium (Figure 6C and 6D). The level of 8-isoprostanes was very low (supplemental Figure III, E). Among the 5 assayed prostanoids, only the release of PGF2α (≈0.8 ng/mL) and 6-keto PGF1α (≈7 ng/mL) evoked by 3 μmol/L ACh was inhibited or abolished by treatment with celecoxib or 2-APB but not by VAS (Figure 6C and 6D). These results are consistent with those obtained by high-performance liquid chromatography-coupled mass spectrometry (HPLC-MS), which showed that the amount of PGF2α (≈1.0 ng/mL) was comparable to that (≈0.8 ng/mL) assayed by enzyme immunoassay (EIA) (supplemental Figure IV).
To further investigate the role of PGF2α and PGI2 in endothelium-dependent contractions, the effects of PGF2α (1 to 30 ng/mL), PGI2 (3.7 to 370 ng/mL), and its stable analog cicaprost (10 to 100 ng/mL) were tested in the presence of 100 μmol/L L-NAME. PGF2α induced contraction of the aortic rings at the relatively low concentration of 1 ng/mL (Figure 6E), comparable with the level of ACh-induced release measured by EIA (≈0.8 ng/mL). The contraction was reduced by S18886 (0.1 μmol/L). By contrast, neither PGI2 nor cicaprost produced a contraction (Figure 6F and 6G), even at a concentration 50-fold higher than that detected in the solution bathing aortae exposed to ACh (≈7 ng/mL for 6-keto PGF1α). Exogenous PGI2 did not relax phenylephrine-contracted aortae (supplemental Figure V).
PGE2, whose release was independent of ACh stimulation (Figure 6A), produced very small contractions at the assayed concentration (≈0.8 ng/mL), and such contractions were insensitive to S18886 (supplemental Figure V).
Augmented Endothelium-Dependent Contractions in Aortae From Aged Hamsters
ACh-induced endothelium-dependent contractions were significantly higher in aortae from aged (>18-month-old) than young (≈3-month-old) hamsters (Figure 7A), and ACh was able to trigger contractions even in the absence of L-NAME in aortae from the aged animals (supplemental Figure VI, H), which were inhibited or abolished by celecoxib, DuP-697, and S18886 (Figure 7B and 7C) and by 2-APB but not by VAS or sc-560 (supplemental Figure VI). Aged aortae expressed a significantly higher level of COX-2, which was again mainly localized to the endothelium (Figure 7D), whereas the expression of COX-1 was comparable in the 2 age groups (supplemental Figure VI, J). Finally, ACh tended to stimulate more release of PGF2α and 6-keto PGF1α in aortae from aged than young hamsters, and the release of PGF2α was inhibited significantly by celecoxib (Figure 7E and supplemental Figure VI, K).
PGF2α produced larger contractions in L-NAME–treated aortae from aged than young hamsters (Figure 7F), whereas contractions to KCl (30 to 50 mmol/L) or phenylephrine (0.1 to 1 μmol/L) were comparable (data not shown). The protein expression of TP receptors was similar in aortae from both age groups (supplemental Figure VI, L). By contrast, in the aged hamster aortae, PGI2 produced neither a relaxation nor a contraction, and PGE2-induced contractions were again insensitive to S18886 (supplemental Figure V).
Human Renal Arteries
In the presence of 100 μmol/L L-NAME, ACh induced contractions in arteries from patients with hypertension and diabetes mellitus, and these contractions were reversed by S18886 (supplemental Figure VII, A). PGF2α-induced contractions were antagonized by S18886 but not by AL-8810, the FP receptor antagonist (supplemental Figure VII, B and C). Pretreatment with S18886, but not AL-8810, prevented exogenous PGF2α-induced contractions (supplemental Figure VII, D). Western blot analysis revealed little or no expression of FP receptor in human arteries in contrast to well-expressed TP receptor (supplemental Figure VII, E). HPLC-MS measurement showed that those arteries released both PGF2α and 6-keto PGF1α in response to 100 μmol/L ACh, but only the release of PGF2α was inhibited by celecoxib (10 μmol/L), whereas VAS (30 μmol/L) was without effect (supplemental Figure VII, F).
Endothelium-dependent contractions are observed generally in arteries of aged or diseased animals, including high fat diet–induced obese mice, spontaneously hypertensive rats, and diabetic rats,3,8,25 in which endothelial function is already impaired. The present study demonstrates that in the aorta of young and healthy hamsters endothelium-dependent contractions can be evoked via COX-2–mediated production of PGF2α, which acts on the TP receptor in vascular smooth muscle cells. Our studies on human renal arteries revealed that this pathway could be of relevance also in humans.
Similar to previous observations in other blood vessels,3,6,8,25 the occurrence of endothelium-dependent contractions in the aorta of the young hamster is unmasked by the presence of L-NAME, which eliminates the production of endothelium-derived NO. Because the contraction is not observed in the absence of the inhibitor of NO synthase, endothelium-dependent relaxations would have predominated over contractions in the aorta of young healthy hamsters.
Arachidonic acid, released from cell membranes by phospholipases, can be metabolized via different pathways to generate vasoactive substances. Lipoxygenases convert arachidonic acid to HPETEs (hydroperoxyeicosatetraenoic acids) and then to either HETEs (hydroxyeicosatetraenoic acids) or leukotrienes. Cyclooxygenases oxygenate arachidonic acid to form PGG2 and PGH2, which are further converted to various prostanoids including PGD2, PGE2, PGF2α, TXA2, and PGI2 via their respective synthases.26 In the present study, baicalein was used to inhibit the lipoxygenase pathway, yet this caused no suppression of the ACh-induced endothelium-dependent contractions. By contrast, incubation with a relatively low concentration of indomethacin abolished the response. These observations permit the conclusion that arachidonic acid metabolites formed under the catalytic action of cyclooxygenases are the most likely EDCF candidate(s) mediating the endothelium-dependent contractions in the aorta of healthy hamsters.
COX-1 is known to be expressed constitutively in most tissues, whereas COX-2 is highly inducible by proinflammatory cytokines, tumor promoters, and mitogens.27 Recent studies suggest that COX-2 is also constitutively expressed in the kidney, brain, and arteries.10,11,28 In the cardiovascular system, endothelial cells express COX-2 in response to shear stress under normal physiological condition.18 When COX-2 is present, it contributes to PGI2 synthesis29,30 and can activate silent reservoirs of PGI2 synthase in most tissues.31 In the present study, the endothelium-dependent contractions of the aorta from healthy hamsters were mediated by COX-2, whereas the constitutively expressed COX-1 did not play a major role, as evidenced by the pronounced attenuation of the response by NS-398, DuP-697, and celecoxib but not by sc-560 and VAS. Inhibition of RNA synthesis and protein synthesis by actinomycin-D and cycloheximide, respectively, did not alter the endothelium-dependent contractions, indicating that COX-2 was expressed constitutively in the aorta and that its presence was not induced acutely by ACh. The molecular biological comparison of aortae with and without endothelium permitted the conclusion that COX-2 mRNA and protein expressions, demonstrated by using RT-PCR and Western blot analysis, respectively, are localized mainly in the endothelium. Thus, endothelial COX-2 appears to represent the major enzyme responsible for the generation of EDCF(s) in the aorta of healthy hamsters. After the production of EDCF(s) by COX-2 in endothelial cells, it diffuses to the vascular smooth muscle cells, where it acts on the TP receptor to cause contraction. The involvement of TP receptors was demonstrated in the present study by the use of specific antagonists (S18886, GR 32191, and L-655,240), which markedly decreased or abolished the endothelium-dependent contractions evoked by ACh. In contrast to the well-expressed TP receptor, the FP receptor is minimally expressed in hamster aortae. Besides, ACh- or PGF2α-induced contractions were not reduced by the FP receptor antagonist AL-8810 (supplemental Figure VIII), thus discounting a significant role of the FP receptor in the endothelium-dependent contractions, although its natural agonist PGF2α is proposed to be the EDCF in hamster aortae.
The present data show that Ca2+ influx into endothelial cells is crucial for the occurrence of the endothelium-dependent contractions. This conclusion is based on the observation that preparations incubated in Ca2+-free solution showed no contraction until Ca2+ was reintroduced into the bathing solution. Ca2+ ions possibly enter endothelial cells via nonselective cation channels, as evidenced by the effect of 2-APB in attenuating endothelium-dependent contractions and abolishing the ACh-stimulated elevation of [Ca2+]i in in situ imaging of endothelial cells. In addition, 2-APB inhibited the ACh-induced release of COX-2–derived prostanoids in the aorta with endothelium, whereas it did not affect U46619- or KCl-induced contraction of aortic rings, illustrating that it acts on the endothelial cells.
By comparing the results from EIA and HPLC-MS and the subsequent functional studies performed using the myograph with exogenously added prostanoids, PGF2α appears to be the most likely EDCF candidate. Indeed, PGF2α was released endogenously from the aortic endothelium in physiological amounts that correspond to its potent effect in eliciting contraction of the smooth muscle. Although PGI2 was also released in considerable amounts, it failed to evoke any contraction or relaxation per se, even at a concentration 50 times higher than the one detected, suggesting that it may not contribute to endothelium-dependent contractions as it does in the aorta of spontaneously hypertensive rats.7,9 The present study can discount the possible involvement of PGE2 because its release was not stimulated by ACh and the small contractions induced by exogenous PGE2 were insensitive to the TP receptor antagonism.
Because TP receptors are involved in the response, it is logical to speculate that TXA2 may contribute to endothelium-dependent contractions in the hamster aorta. However, this possibility is made unlikely by 2 observations. First, in the presence of ozagrel, a thromboxane synthase inhibitor,32 endothelium-dependent contractions to ACh remained unaltered. Second, ACh did not increase the release of TXA2. Thus TXA2 is not a major EDCF candidate in the hamster aorta. By contrast, TXA2 contributes to the endothelium-dependent contractions of the canine basilar artery and the SHR aorta.7,33
COX is involved in the generation of ROS in vascular tissues,34 which are normally neutralized by NO. It appears necessary to test whether ROS play a role in the production and the action of EDCF, because L-NAME inhibits NO production and unmasks the endothelium-dependent contractions. The present evidence from functional, EPR and EIA studies points against such a possibility. In the functional studies, neither free radical scavengers nor an NADPH oxidase inhibitor attenuated the endothelium-dependent contractions to ACh. The EPR study showed that basal levels of superoxide anions and peroxynitrite formed by the combination of superoxide and NO35 were undetectable in the hamster aorta and remained so even in the presence of L-NAME and ACh. The EIA study demonstrated that 8-isoprostanes, which are generated in vivo by the free radical-catalyzed, nonenzymatic peroxidation of arachidonic acid,36 were released in very small amounts compared with other prostaglandins. Thus, in the healthy blood vessels examined in the present studies, ROS play no role in catalyzing the production and release of EDCF, unlike in the aorta of the SHR.6
The present study also provided evidence for positive impact of aging on endothelium-dependent contractions and showed a marked enhancement of endothelium-dependent contractions in aortae from aged hamsters (>18 months old) as compared with those from young hamsters (3 months old). One distinctive feature in aged hamsters, besides greater contractions in L-NAME–treated aortae, is that ACh was able to trigger endothelium-dependent contractions in the absence of L-NAME, albeit to a lesser degree. The present results suggest that aging is likely to reduce the NO bioavailability, ultimately leading to the occurrence of endothelium-dependent contractions without inhibition of NO synthesis.
The enhanced endothelium-dependent contraction is most likely related to an increased expression and activity of COX-2, a slightly (although not significantly) augmented ACh-stimulated release of PGF2α and, most importantly, the increased vascular sensitivity to PGF2α with unaltered expression of the TP receptor in aortae from aged hamsters.
The preliminary results obtained in human renal arteries imply that a COX-2 metabolite, possibly PGF2α, produces TP receptor–dependent vasoconstrictions consistent with the observations on hamster aortae. In addition, we provide preliminary evidence that the release of PGF2α but not that of PGI2, in response to ACh was inhibited by celecoxib, thus indirectly indicating a crucial role of COX-2 in ACh-induced contractions of human renal arteries.
To conclude, the present study demonstrates a role of endothelial COX-2 in the regulation of vascular tone in the aorta of the healthy hamster. The present data show that endothelial COX-2 catalyzes the formation of PGF2α, which represents a physiological EDCF at least in this preparation. Through binding to TP receptors, PGF2α produces endothelium-dependent contractions to ACh (Figure 8). The present preliminary data in human arteries could have clinical relevance in humans because the same or a similar pathway also exists in the human renal arteries being tested. Our data also demonstrate that this pathway is of increasing importance during aging in hamsters.
Sources of Funding
This study was supported by the Hong Kong General Research Fund (Chinese University of Hong Kong) (4362/04M and 465308), the Chinese University of Hong Kong Li Ka Shing Institute of Health Sciences, the Chinese University of Hong Kong Focused Investment Scheme, and the Germany–Hong Kong Joint Research Scheme/Deutsche Forschungsgemeinschaft. S.L.W. is a postgraduate student and F.P.L. is a postdoctoral fellow.
Original received May 20, 2008; revision received December 2, 2008; accepted December 4, 2008.
Lüscher TF, Vanhoutte PM. Endothelium-derived contractions to acetylcholine in the aorta of the spontaneously hypertensive rats. Hypertension. 1986; 8: 344–348.
Furchgott RF, Vanhoutte PM. Endothelium-derived relaxing and contracting factors. FASEB J. 1989; 3: 2007–2018.
Gluais P, Paysant J, Badier-Commander C, Verbeuren T, Vanhoutte PM, Feletou M. In SHR aorta, calcium ionophore A-23187 releases prostacyclin and thromboxane A2 as endothelium-derived contracting factors. Am J Physiol Heart Circ Physiol. 2006; 291: H2255–H2264.
Baber SR, Champion HC, Bivalacqua TJ, Hyman AL, Kadowitz PJ. Role of cyclooxygenase-2 in the generation of vasoactive prostanoids in the rat pulmonary and systemic vascular beds. Circulation. 2003; 108: 896–901.
Therland KL, Stubbe J, Thiesson HC, Ottosen PD, Walter S, Sørensen GL, Skøtt O, Jensen BL. Cycloxygenase-2 is expressed in vasculature of normal and ischemic adult human kidney and is colocalized with vascular prostaglandin E2 EP4 receptors. J Am Soc Nephrol. 2004; 15: 1189–1198.
Hennan JK, Huang J, Barrett TD, Driscoll EM, Willens DE, Park AM, Crofford LJ, Lucchesi BR. Effects of selective cyclooxygenase-2 inhibition on vascular responses and thrombosis in canine coronary arteries. Circulation. 2001; 104: 820–825.
Hartner A, Cordasic N, Goppelt-Struebe M, Veelken R, Hilgers KF. Role of macula densa cyclooxygenase-2 in renovascular hypertension. Am J Physiol Renal Physiol. 2003; 284: F498–F502.
Bagi Z, Erdei N, Toth A, Li W, Hintze TH, Koller A, Kaley G. Type 2 diabetic mice have increased arteriolar tone and blood pressure: enhanced release of COX-2-derived constrictor prostaglandins. Arterioscler Thromb Vasc Biol. 2005; 25: 1610–1616.
Topper JN, Cai J, Falb D, Gimbrone MA Jr. Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese superoxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc Natl Acad Sci U S A. 1996; 93: 10417–10422.
Nicolosi RJ. Dietary fat saturation effects on low-density-lipoprotein concentrations and metabolism in various animal models. Am J Clin Nutr. 1997; 65: 1617S–1627S.
Arbeeny CM, Meyers DS, Bergquist KE, Gregg RE. Inhibition of fatty acid synthesis decreases very low density lipoprotein secretion in the hamster. J Lipid Res. 1992; 33: 843–851.
Valeille K, Férézou J, Amsler G, Quignard-Boulangé A, Parquet M, Gripois D, Dorovska-Taran V, Martin JC. A cis-9, trans-11-conjugated linoleic acid-rich oil reduces the outcome of atherogenic process in hyperlipidemic hamster. Am J Physiol Heart Circ Physiol. 2005; 289: H652–H659.
Simmons DL, Botting RM, Hla T. Cyclooxygenase isozymes: the biology of prostaglandin synthesis and inhibition. Pharmacol Rev. 2004; 56: 387–437.
Fries S, Grosser T. The cardiovascular pharmacology of COX-2 inhibition. Hematology Am Soc Hematol Educ Program. 2005: 445–451.
McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA, FitzGerald GA. Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci U S A. 1999; 96: 272–277.
Katusic ZS, Shepherd JT, Vanhoutte PM. Endothelium-dependent contractions to calcium ionophore A23187, arachidonic acid, and acetylcholine in canine basilar arteries. Stroke. 1988; 19: 476–479.