F2-isoprostanes are a family of compounds produced in vivo by free radical peroxidation of arachidonic acid (1). Their quantification in biologic fluids represents a reliable marker of lipid peroxidation (2). Among the F2-isoprostanes, the 15-F2t-isoprostane (15-F2t-IsoP (eicosanoid nomenclature committee) (3), also named isoprostaglandin F2α type-III (4), formerly named 8-iso-PGF2α) has been shown to be a potent vasoconstrictor in animal and human vascular beds (see (5) for full review). These effects are mediated predominantly by TP-receptor stimulation (6) and may be modulated by the endothelium through a pathway independent of TP-receptor stimulation (7,8).
In humans, two major endogenous β-oxidized metabolites of 15-F2t-IsoP have recently been described in plasma and urine: 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP (9,10), which we have recently synthesized from L-glucose (11) (seeFig. 1 for the stereochemistry). In addition, 15-F2t-IsoP has been shown in the rabbit to undergo alcohol group oxidation leading to 15-keto-15-F2t-IsoP, which in turn undergoes β-oxidation (12). Quantification of 15-F2t-IsoP metabolites gives an integrated index of isoprostane production over time. These compounds are produced in vivo and may sometimes be more abundant than 15-F2t-IsoP in plasma (13). No data are currently available on the vascular properties of these compounds. In this study, we investigated the vascular actions (contraction and relaxation) of 15-keto-15-F2t-IsoP, 2,3-dinor-15-F2t-IsoP, and 2,3-dinor-5,6-dihydro-15-F2t-IsoP in comparison with 15-F2t-IsoP on the rat thoracic aorta and the involvement of TP-receptor activation and endothelial modulation.
The methods used for the measurement of isometric tension in rings of rat thoracic aorta were as previously reported (14). In accordance with the French law and the local ethical committee guidelines for animal research, male Wistar rats (260–400 g, Janvier, Le Genest St Isle, France) were housed in climate-controlled conditions and provided with standard rat chow. Animals were terminally anesthetized with an intraperitoneal injection of 60 mg/kg−1 of sodium pentobarbital (Sanofi, Libourne, France). Heparin (150 IU, Sanofi Winthrop, Gentilly, France) was injected intravenously. Then, the thoracic aorta was quickly excised, cleaned of connective tissue, and cut into 4-mm lengths. Six rings were taken from each thoracic aorta. The endothelium was removed from some aortic rings by gently rolling the tip of a plastic forceps inside the vessel. Rings were mounted between two stainless steel wires and immersed in organ chambers filled with 6 ml of Krebs solution maintained at 37°C and gassed with 95% oxygen and 5% carbon dioxide. The Krebs solution had the following composition (mM): NaCl 118, KCl 4.7, CaCl2 2.5, MgSO4 1, KH2PO4 1, NaHCO3 25, and glucose 11. Responses were measured with an isometric force displacement transducer (UF-1 Pioden Controls limited, Canterbury, U.K.) and continuously recorded (Linseis L200E, Bioblock, Illkirch, France). A resting force of 1.5 g was applied to each tissue ring. The aortic rings were left to stabilize for 60 min and washed every 15 min with Krebs solution. The rings were then challenged twice with KCl (90 mM) at a 10-min interval to ensure that responses were reproducible. Aortic rings were tested 30 min later for a functional endothelium by their ability to relax in response to acetylcholine (10−6 to 10−4M) when precontracted with phenylephrine (10−7M). The experiments were performed after a further 60-min equilibration period. All the concentration-response curves were made by adding increasing concentrations of the agonist to the organ bath in 0.5-log unit steps, waiting until a plateau was obtained for each step. One curve was obtained from each ring.
Protocol design for contractile responses
Concentration-response curves were expressed as a percentage of KCl 90 mM-induced contraction. Five series of contractile experiments were performed:
- The first series compared the contractile properties of F2-isoprostane metabolites (15-keto-15-F2t-IsoP, 2,3-dinor-15-F2t-IsoP, and 2,3-dinor-5,6-dihydro-15-F2t-IsoP) with their parent compound 15-F2t-IsoP and its prostaglandin isomer prostaglandin F2α with U46619, a TP-receptor agonist (15), and with fluprostenol, an FP-receptor agonist (16,17). Isoprostanes derived from eicosapentaenoic (15-F3t-Isoprostane, 15-F3t-IsoP) and docosahexaenoic acid (4[RS]-F4t-Isoprostane, 4[RS]-F4t-IsoP) have also been tested (18,19) (seeFig. 1 for stereochemistry).
- The role of the endothelium in the contractile responses was assessed by comparing the responses to 15-keto-15-F2t-IsoP, 15-F2t-IsoP, and U46619 in rings with an intact endothelium and the endothelium denuded.
- The activation of TP-receptors by 15-keto-15-F2t-IsoP, 15-F2t-IsoP, and U46619 was assessed on rings pretreated for 30 min with increasing concentrations of GR32191 (10−9 to 10−7M), a TP-receptor antagonist (20). Alone, GR32191 (up to 10−7M) had no effect on the basal tone.
- To test the antagonist activity of the F2-isoprostane metabolites, 15-keto-15-F2t-IsoP, 2,3-dinor-15-F2t-IsoP, and 2,3-dinor-5,6-dihydro-15-F2t-IsoP on the contractile response to 15-F2t-IsoP, concentration-responses curves to 15-F2t-IsoP were obtained 30 min after pretreatment with 15-keto-15-F2t-IsoP (10−7, 3.10−7, 10−6, and 3.10−6M), 2,3-dinor-15-F2t-IsoP (10−5M), and 2,3-dinor-5,6-dihydro-15-F2t-IsoP (10−5M).
Protocol design for relaxation
To observe any relaxation, the rings were contracted by the addition of U46619 (3.10−8 M) or phenylephrine (10−7M), which produced a stable contraction for the duration of the experiment. The choice of U46619 and phenylephrine concentrations was based on their pD2 values (obtained from preliminary experiments using phenylephrine, data not shown). Concentration-relaxation curves were expressed as a percentage of the agonist-induced contraction. Two series of experiments were performed:
- After a stable contraction to U46619 was obtained, 15-F2t-IsoP or the F2-isoprostane metabolites (15-keto-15-F2t-IsoP, 2,3-dinor-15-F2t-IsoP or 2,3-dinor-5,6-dihydro-15-F2t-IsoP) were added in a cumulative fashion, in intact rings or in endothelium denuded rings. Paired time-control curves were obtained.
- To characterize the involvement of TP receptors in the relaxation of either 15-F2t-IsoP or 15-keto-15-F2t-IsoP, phenylephrine was used to contract the aorta rings. 15-F2t-IsoP or 15-keto-15-F2t-IsoP were added in a cumulative fashion in intact rings, following 30 min pretreatment with vehicle or GR32191 (10−7M) in endothelium intact rings. GR32191 had no effect on the contraction plateau. Paired time-control curves were obtained.
15-F2t-IsoP (8-iso-prostaglandin F2α), 15-keto-15-F2t-IsoP (8-iso-15-keto prostaglandin F2α), prostaglandin F2α, fluprostenol, and 15-F3t-IsoP (8-iso-prostaglandin F3α) were purchased from Cayman (Ann Arbor, MI, U.S.A.). U46619 (9,11-dideoxy-9α, 11 α-methanoepoxy-prostaglandin F2α), L-phenylephrine, and acetylcholine were purchased from Sigma (Saint Quentin Fallavier, France). GR32191 ([1R-[1 α(Z), 2β, 3β, 5 α (+)-7-[[1,1´-biphenyl)-4-yl]methoxy]-3-hydroxy-2-(1-piperidinyl) cclopentyl] -4-4heptanoic acid], hydrochloride) was kindly provided by Glaxo Wellcome (Steventage, UK).
2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP were totally synthesized from commercially available L-glucose, according to our procedure (11). The synthesis of 4(RS)-F4t-IsoP was performed in our laboratory, as described (21).
All isoprostanes were dissolved in methanol at 10−2M except 15-F3t-IsoP, which was dissolved in methyl acetate at 3.10−3M. Stock solutions were then diluted in distilled water before being added to the organ baths. The highest concentration of methanol was 0.3%, which had no direct effect on the vascular tone in preliminary experiments.
The effective concentration of agent that caused 50% of maximal contraction (EC50) was determined from each curve by a logistic, curve-fitting equation. EC50 values were expressed as pD2 (-log EC50). The contractile responses were compared in terms of potency (pD2) and maximal contraction (Emax). For the relaxation, the maximal effect observed was compared.
To determine 15-keto-15-F2t-IsoP antagonist affinity, apparent pA2 values were determined according to the Schild analysis (22). The pA2 value and the slope of the linear regression line were calculated. Slope of the Schild plot that was not different from the unity was indicative that the interaction was competitive. pKB values were calculated for GR32191 10−7M, because the lack of significant rightward shift of the curves observed at GR32191 10−9 and 10−8M did not allow a pA2 determination.
Unpaired t tests were used to test the statistical significance between two means. More than two means were compared with the use of analysis of variance (and Bonferroni's test as post-hoc test). Values of p < 0.05, corrected by the number of comparisons made, were considered significant. All values are expressed as mean ± SEM.
Vasoconstrictor effects of F2-isoprostanes and their metabolites in rat aorta
The F2-isoprostane metabolite, 15-keto-15-F2t-IsoP, induced vasoconstriction in a concentration-dependent manner, whereas 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP had no effect (Table 1). The following rank order of potency was determined by comparison of pD2 values: U46619 > 15-F2t-IsoP > PGF2α = 15-keto-15-F2t-IsoP > fluprostenol (p < 0.05 for 15-keto-15-F2t-IsoP vs. 15-F2t-IsoP and fluprostenol, and p < 0.0001 vs. U46619). The maximal contraction to 15-keto-15-F2t-IsoP was significantly lower than U46619 and PGF2α (p < 0.01 vs. PGF2α and p < 0.05 vs. U46619). The isoprostanes derived from eicosapentaenoic (15-F3t-IsoP) and docosahexaenoic acid (4(RS)-F4t-IsoP) induced no contraction.
Endothelial modulation of 15-keto-15-F2t-IsoP, 15-F2t-IsoP, and U46619 induced vasoconstriction in rat aorta
Endothelium removal had no influence on the contraction to 15-keto-15-F2t-IsoP (Fig. 2A). In contrast, the maximal contraction to 15-F2t-IsoP was significantly increased in endothelium-denuded rings (Emax: 186 ± 14% vs. 148 ± 7% in endothelium-intact rings, p < 0.05, Fig. 2B). Similarly, the maximal contraction to U46619 was significantly increased in endothelium-denuded rings (Emax: 184 ± 9% vs. 162 ± 6% in endothelium-intact rings, p < 0.05, Fig. 2C).
Effect of the TP-receptor antagonist GR32191 on the contractile responses induced by 15-keto-15-F2t-IsoP, 15F2t-IsoP, and U46619 in rat aorta
GR32191 caused an inhibition of the vasoconstriction induced by 15-keto-15F2t-IsoP, with a significant decrease in the Emax values for GR32191 10−7M (Fig. 3A). The pD2 values in presence of the vehicles, GR32191 10−9, 10−8M, and 10−7M were 5.81 ± 0.05, 5.75 ± 0.04, 6.02 ± 0.15, and 5.22 ± 0.05, respectively (p < 0.0001 for GR32191 10−7M vs. other groups).
GR32191 10−7M produced a rightward shift of 15-F2t-IsoP concentration-response curves, associated with a nonsignificant decrease in the Emax values (Fig. 3B). The pD2 values in presence of the vehicles, GR32191 10−9, 10−8M, and 10−7M were 6.59 ± 0.03, 6.53 ± 0.07, 6.46 ± 0.09, and 5.80 ± 0.09, respectively (p < 0.0001 for GR32191 10−7M vs. other groups). The pKB value was 7.72 ± 0.09 for GR32191 10−7M. Similarly, GR32191 10−7M produced a rightward shift of U46619 concentration-response curves. The pD2 values in presence of the vehicles, GR32191 10−9, 10−8 M, and 10−7M were 7,71 ± 0.06, 7.64 ± 0.18, 7.72 ± 0.12, and 6.97 ± 0.1, respectively (p < 0.001 for GR32191 10−7M vs. other groups), with a pKB value of 7.64 ± 0.12 for GR32191 10−7M (Fig. 3C).
Antagonist effect of F2-isoprostane metabolites on 15F2t-IsoP-induced contraction
Pretreatment with 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP did not affect 15-F2t-IsoP concentration-response curves (n = 4, data not shown).
In contrast, 15-keto-15-F2t-IsoP pretreatment induced a concentration-dependent contraction and also competitively inhibited the response to 15-F2t-IsoP (Table 2). When concentration ratios of EC50 values were used, a Schild regression of this data was linear with a slope of 0.974, not significantly different from the unity, and a pA2 value of 6.13.
Characterization of the potential relaxation of 15-F2t-IsoP and its metabolites on rat aorta
2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP induced no variation of the contraction plateau induced by U46619 (n = 4, data not shown).
15-keto-15-F2t-IsoP induced a weak relaxation of rat aortic rings contracted with U46619 that was observed at high concentrations, which was not altered by endothelium denudation (Fig. 4A). Such a relaxation was significantly different from the control curves only at high concentrations (10−5M) (−22 ± 5% for endothelium intact rings vs. −23 ± 3% for endothelium denuded rings vs. −8 ± 4 for the time-control curve, p < 0.01 vs. time-control curves). In comparison with the time-control curves, no significant relaxation was observed in rat aortic rings contracted with phenylephrine (Fig. 4B). Pretreatment with GR32191 (10−7M) induced a rightward shift of 15-keto-15-F2t-IsoP contraction curves.
In comparison with the time-control curves, no significant relaxation to 15-F2t-IsoP was observed in rat aortic rings contracted with phenylephrine or U46619 (Fig. 5, A and B). However, a contractile response was observed. The maximal contraction was enhanced in the absence of endothelium (Fig. 5A). Pretreatment with GR32191 (10−7M) induced a rightward shift of 15-F2t-IsoP contraction curves (Fig. 5B).
In this study, we found that the F2-isoprostane metabolites, 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP, did not affect the basal tone of the rat aorta. In addition, these compounds had neither antagonist effects on 15-F2t-IsoP-induced contractions, nor relaxing effects on aortic rings precontracted with U46619. As a consequence, these data clearly show that 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP, the major 15-F2t-isoprostanes metabolites (9,10), have no vascular effects on rat aorta. In addition, the F3-isoprostane, 15-F3t-IsoP, produced in response to eicosapentaenoic acid peroxidation (18), had no contractile effect. In comparison with 15-F2t-IsoP, two carbon atoms shorten the carboxyl moiety of 2,3-dinor-15-F2t-IsoP, whereas 15-F3t-IsoP has an additional double bond on the lower side chain. Altogether, these data strongly support the hypothesis that the structure of the upper and the lower side chains of 15-F2t-IsoP are key determinants in the receptor-agonist interaction inducing 15-F2t-IsoP vasoconstriction. As well as 15-F3t-IsoP, 4(RS)-F4t-IsoP, produced by docosahexaenoic acid peroxidation, had no vascular effects suggesting that unlike arachidonic acid-dependent isoprostanes, eicosapentaenoic and docosahexaenoic acid-dependent isoprostanes have no vascular effects.
Unlike the 2,3-dinor metabolites, 15-keto-15-F2t-IsoP induced a contraction of rat aortic rings with a potency equivalent to PGF2α, lower than the parent compound 15-F2t-IsoP and the TP-receptor agonist U46619 but higher than the FP-receptor agonist fluprostenol. Responses to 15-keto-15-F2t-IsoP were inhibited by GR32191, a highly potent and selective TP receptor antagonist (20,23), suggesting that 15-keto-15-F2t-IsoP-induced contraction was mediated through TP-receptor activation, as already shown for 15-F2t-IsoP (6,24). However, the concentration-dependent inhibition by GR32191 of the contractile response to 15-keto-15-F2t-IsoP was associated with a decreased maximal response, unlike that of U46619 and 15-F2t-IsoP. In addition, the maximal contractile response to 15-keto-15-F2t-IsoP was lower than for U46619. If 15-keto-15-F2t-IsoP is acting at the TP-receptor, this suggests a partial agonist. The action of a partial agonist requires a higher level of receptor occupancy than that of a full agonist. At high concentrations of a TP-receptor antagonist, the receptor reserve is reduced and 15-keto-15-F2t-IsoP then is unable to occupy a sufficient number of receptors to induce a maximal contraction, therefore explaining the decrease in the maximal response to 15-keto-15-F2t-IsoP observed in presence of GR32191 10−7M. The action of 15-keto-15-F2t-IsoP as a partial agonist at the TP-receptor is also supported by the antagonist effect of 15-keto-15-F2t-IsoP observed on 15-F2t-IsoP-induced contraction in rat aorta.
15-keto-15-F2t-IsoP exerted a weak relaxation on the U46619-precontracted rat aorta. This relaxant effect was not altered by endothelium removal, and therefore, the contractile response was not modified in endothelium-denuded preparations. The partial agonism of 15-keto-15-F2t-IsoP at the TP-receptor explains the relaxation observed in response to 15-keto-15-F2t-IsoP on U46619-precontracted rat aorta, because the level of precontraction induced by U46619 is close to the maximal contractile response to 15-keto-15-F2t-IsoP, leaving no room for an additional contractile effect and therefore unmasking the relaxant effect.
An important unresolved question is whether the vascular properties of 15-keto-15-F2t-IsoP may have any biologic relevance. The limited data available concerning 15-keto-15-F2t-IsoP formation, which has only been described in the rabbit (12), needs to be further implemented in humans to enable a clearer answer.
In conclusion, among the F2-isoprostane metabolites, 2,3-dinor-15-F2t-IsoP and 2,3-dinor-5,6-dihydro-15-F2t-IsoP did not cause vasorelaxation or vasoconstriction on rat thoracic aorta. In contrast, 15-keto-15-F2t-IsoP is probably acting as a partial agonist at the TP-receptor, mediating contraction, and induces a weak endothelium-independent relaxation at high concentrations.
1. Morrow JD, Hill KE, Burk RF, et al. A series of prostaglandin F2
-like compounds are produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci U S A 1990; 87:9383–7.
2. Roberts LJ, Morrow JD. Measurement of F2
as an index of oxidative stress in vivo. Free Radic Biol Med 2000; 28:505–13.
3. Taber DF, Morrow JD, Robert II. LJ A nomenclature system for the isoprostanes
. Prostaglandins 1997; 53:63–7.
4. Rokach J, Khanapure SP, Hwang SW, et al. Nomenclature of isoprostanes
: a proposal. Prostaglandins 1997; 54:853–73.
5. Cracowski JL, Devillier P, Durand T, et al. Vascular biology of the isoprostanes
. J Vasc Res 2001; 38:93–103.
6. Audoly LP, Rocca B, Fabre JE, et al. Cardiovascular responses to the isoprostanes
-III and iPE2
-III are mediated via the thromboxane A2
receptor in vivo. Circulation 2000; 101:2833–40.
7. Jourdan KB, Evans TW, Curzen NP, et al. Evidence for a dilator function of 8-iso prostaglandin F2α
in rat pulmonary artery. Br J Pharmacol 1997; 120:1280–5.
8. Wilson SH, Best PJ, Lerman LO, et al. Enhanced coronary vasoconstriction to oxidative stress product, 8-epi-prostaglandin F2α
, in experimental hypercholesterolemia. Cardiovasc Res 1999; 44:601–7.
9. Chiabrando C, Valagussa A, Rivalta C, et al. Identification and measurement of endogenous beta-oxidation metabolites of 8-epi-Prostaglandin F2α
. J Biol Chem 1999; 274:1313–9.
10. Roberts II, LJ Moore KP, Zachert WE, et al. Identification of the major urinary metabolite of the F2
-isoprostane 8-iso-prostaglandin F2α
in humans. J Biol Chem 1996; 271:20617–20.
11. Durand T, Guy A, Henry O, et al. Total syntheses of four metabolites of 15-F2t
. Eur J Org Chem 2001; 4:809–19.
12. Basu S. Metabolism of 8-iso-prostaglandin F2α
. FEBS Lett 1998; 428:32–6.
13. Burke A, Lawson JA, Meagher EA, et al. Specific analysis in plasma and urine of 2,3-dinor-5,6-dihydro-isoprostane F2α
-III, a metabolite of isoprostane F2α
-III and an oxidation product of gamma-linolenic acid. J Biol Chem 2000; 275:2499–504.
14. Stanke-Labesque F, Devillier P, Veitl S, et al. Cysteinyl leukotrienes are involved in angiotensin II-induced contraction of aorta from spontaneously hypertensive rats. Cardiovasc Res 2001; 49:152–60.
15. Coleman RA, Humphrey PP, Kennedy L, et al. Comparison of the action of U-46619, a prostaglandin H2
-analogue, with those of prostaglandin H2
and thromboxane A2
on some isolated smooth muscle preparations. Br J Pharmacol 1981; 73:773–8.
16. Lake S, Gullberg H, Wahlqvist J, et al. Cloning of the rat and human prostaglandin F2α
receptors and the expression of the rat prostaglandin F2α
receptor. FEBS Lett 1994; 355:317–25.
17. Abramovitz M, Boie Y, Nguyen T, et al. Cloning and expression of a cDNA for the human prostanoid FP receptor. J Biol Chem 1994; 269:2632–6.
18. Nourooz-Zadeh J, Halliwell B, Anggard EE. Evidence for the formation of F3
during peroxidation of eicosapentaenoic acid. Biochem Biophys Res Commun 1997; 236:467–72.
19. Roberts II, LJ Montine TJ, Markesbery WR, et al. Formation of isoprostane-like compounds (neuroprostanes) in vivo from docosahexaenoic acid. J Biol Chem 1998; 273:13605–12.
20. Humphrey PP, Hallet P, Hornby EJ, et al. Pathophysiological actions of thromboxane A2
and their pharmacological antagonisms by thromboxane receptor blockade with GR32191. Circulation 1990; 81(suppl I):42–52.
21. Durand T, Guy A, Vidal JP, et al. Total synthesis of 4(RS)-F4t
-isoprostane methyl ester. Tetrahedron Lett 2000; 41:3859–62.
22. Arunlakshana O, Schild HO. Some quantitative uses of drug antagonists. Br J Pharmacol Chemother 1959; 14:48–58.
23. Lumley P, White BP, Humphrey PP. GR32191, a highly potent and specific thromboxane A2
receptor blocking drug on platelets and vascular and airways smooth muscle in vitro. Br J Pharmacol 1989; 97:783–94.
24. Kromer B, Tippins JR. Coronary artery constriction by the isoprostane 8-epi-prostaglandin F2α
. Br J Pharmacol 1996; 119:1276–80.