Role of superoxide and thromboxane receptors in acute angiotensin II-induced vasoconstriction of rabbit vessels

Sandra L. Pfister, Kasem Nithipatikom, William B. Campbell


This study explored the hypothesis that a portion of angiotensin II-induced contractions is dependent on superoxide generation and release of a previously unidentified arachidonic acid metabolite that activates vascular smooth muscle thromboxane receptors. Treatment of rabbit aorta or mesentery artery with the thromboxane receptor antagonist SQ29548 (10 μM) reduced angiotensin II-induced contractions (maximal contraction in aorta; control vs. SQ29548: 134 ± 16 vs. 93 ± 10%). A subset of rabbits deficient in vascular thromboxane receptors also displayed decreased contractions to angiotensin II. The superoxide dismutase mimetic Tiron (30 mM) attenuated angiotensin II-induced contractions only in rabbits with functional vascular thromboxane receptors (maximal contraction in aorta; control vs. Tiron: 105 ± 5 vs. 69 ± 11%). Removal of the endothelium or treatment with a nitric oxide synthase inhibitor, nitro-l-arginine (30 μM) did not alter angiotensin II-induced contractions. Tiron and SQ29548 decreased angiotensin II-induced contractions in the denuded aortas by a similar percentage as that observed in intact vessels. The cyclooxygenase inhibitor indomethacin (10 μM) or thromboxane synthase inhibitor dazoxiben (10 μM) had no effect on angiotensin II-induced contractions indicating that the vasoconstrictor was not thromboxane. Angiotensin II increased the formation of a 15-series isoprostane. Isoprostanes are free radical-derived products of arachidonic acid. The unidentified isoprostane increased when vessels were incubated with the superoxide-generating system xanthine/xanthine oxidase. Pretreatment of rabbit aorta with the isoprostane isolated from aortic incubations enhanced angiotensin II-induced contractions. Results suggest the factor activating thromboxane receptors and contributing to angiotensin II vasoconstriction involves the superoxide-mediated generation of a 15-series isoprostane.

  • oxidative stress
  • arachidonic acid
  • isoprostane

alterations in the balance of endothelium-dependent relaxing and contracting factors promote endothelial dysfunction and contribute to the severity of a number of cardiovascular diseases. Previous studies (4) from our laboratory identified thromboxane (TX) A2 as an endothelium-dependent contracting factor (EDCF) that activates a G-protein-coupled receptor referred to as the TP receptor. Endothelial cells produce other constrictors including endothelin and angiotensin II. Angiotensin II stimulates AT1 receptors in the vascular wall leading to vasoconstriction, activation of NAD(P)H oxidase (8), and enhanced superoxide production. The resultant increase in oxidative stress by angiotensin II can activate additional vasopressor mechanisms such as TXA2 or other arachidonic acid metabolites. Mistry et al. (17) first reported an interaction of angiotensin II with TP receptors in studies that showed that a TP receptor antagonist but not a TX synthase inhibitor attenuated the development of angiotensin II-induced hypertension. Despite numerous studies confirming interaction between angiotensin II and the TP receptor, the identity of the compound that acts at the TP receptor to mediate the response to angiotensin II has never been clearly shown. A potential candidate is an isoprostane. Isoprostanes are formed by the peroxidation of arachidonic acid by oxygen free radicals (3, 6, 30). The first class of isoprostanes identified contained F-type prostane rings, analogous to PGF, and were called F2-isoprostanes. Certain F2-isoprostanes elicit contractions via an interaction with TP receptors (11, 14, 16, 32). Acute or chronic angiotensin II infusion increases plasma 8-iso-PGF concentrations in humans, pigs, and rats (9, 18, 29).

The present study characterized the role of TP receptors in the acute vasoconstrictor effects of angiotensin II in isolated rabbit blood vessels. Studies explored the hypothesis that angiotensin II increases superoxide, resulting in the formation of an isoprostane that acts at the vascular TP receptor and thereby contributes to angiotensin II-induced contraction.



Animal protocol was approved by the Institutional Animal Care and Use Committee of the Medical College of Wisconsin, and procedures were performed in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals (1996). Two- to three-month-old male New Zealand White (NZW) rabbits were obtained from New Franken Rabbitry (New Franken, WI) and Kuiper Rabbit Ranch (Gary, IN). Animals were housed in the Medical College of Wisconsin Animal Care Facilities and maintained on a standard rabbit chow diet and given tap water ad libitum. Rabbits were anesthetized with sodium pentobarbital (120 mg/kg iv), and thoracic aorta removed and placed in Krebs-bicarbonate buffer of the following composition (in mM): 118 NaCl, 4 KCl, 3.3 CaCl2, 24 NaHCO3, 1.4 KH2PO4, 1.2 MgSO4, and 11 glucose pH 7.4. Second- or third-order branches from the superior mesentery arteries (200–300 μm) were isolated and placed in HEPES solution consisting of the following (in mM): 150 NaCl, 5.0 KCl, 1.8 CaCl2, 1.0 MgCl2, 10 HEPES, and 5.5 glucose pH 7.4. Arteries were cleaned of adhering fat and connective tissue and used for either vascular reactivity or eicosanoid metabolism studies. We previously identified a subgroup of NZW rabbits that have a significant decrease in vascular smooth muscle cell TP receptors and are referred to as vTP− (5). Rabbits with vascular TP receptors are called vTP+. Thromboxane mimetics like U46619 do not contract blood vessels from vTP− rabbits. However, as previously reported, contractions to other vasoconstrictors, like KCl, norepinephrine and endothelin are identical in vTP+ and vTP− rabbits (5). For all described studies, the presence or absence of functional vascular TP receptors was confirmed by testing for U46619-induced contractions.

Vascular reactivity.

Rings of aorta (3–4 mm) were suspended in 6-ml organ baths containing Krebs bicarbonate buffer that was warmed to 37°C and continuously aerated with a 95% O2-5% CO2 mixture. Isometric tension was measured with force-displacement transducers (Grass Instruments) and AD Instruments ETH-400 amplifiers and recorded on a Macintosh computer using MacLab 8e software as previously described (23). Resting tension was adjusted to 2 g, and the vessels equilibrated for 1 h. The KCl concentration of the baths was increased to 40 mM until stable, reproducible contractions were produced. Responses to the TXA2 mimetic U46619 (10−10–10−7 M) were obtained. Aortic rings that contracted to KCl but not to U46619 were identified as vTP− (Fig. 1A). Aortic rings that contracted to both KCl and U46619 were vTP+. Maximal KCl contractions in vTP+ and vTP− were similar (vTP+ vs. vTP−: 2.4 ± 0.1 vs. 2.7 ± 0.1 g). Mesenteric arterial segments (1.5-mm long) were threaded on two stainless steel wires (40-μm diameter) and mounted on a four-chamber wire myograph (model 610M; Danish Myo Technology) as previously described (1). Arteries were equilibrated at 37°C for 30 min in physiological saline solution containing the following (in mM): 119 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgSO4, 24 NaHCO3, 1.2 KH2PO4, 0.026 EDTA, and 5.5 glucose, bubbled with 95% O2-5% CO2. The resting tension was set at 1 mN. Arteries were stimulated two times with KCl (60 mM) plus phenylephrine (10 μM) for 3–5 min at 10-min intervals before the initiation of experimental protocols. Cumulative concentration response curves to angiotensin II (10−11–10−7 M) or 8-iso-PGF (10−9–10−6 M) were obtained. Other vessels were pretreated with the TP receptor antagonist SQ29548 (10 μM), the TX synthase inhibitor dazoxiben (10 μM), the cyclooxygenase (COX) inhibitor indomethacin (10 μM), the nitric oxide (NO) synthase inhibitor NG-nitro-l-arginine (l-NNA; 30 μM), the SOD mimetic Tiron (30 μM), or vehicle before the administration of angiotensin II. Pretreatment time was 10 min for all the inhibitors except Tiron, which was added 30 min before angiotensin II. The pretreatment time was based on results from previous studies that showed that both the time period and inhibitor concentration blocked the compounds of interest (4, 40). The aortic endothelium was purposefully removed in some vessels by gently rubbing the intimal surface with a cotton-tipped swab. Relaxation responses to acetylcholine were tested and found to be absent in the denuded vessels. All results were expressed as the percent contraction of the maximal KCl response. Removal of endothelium did not change KCl responses.

Fig. 1.

Contractile effects of U46619 (A) and the isoprostane 8-iso-PGF (B) in aortas from rabbits with an increase in vascular smooth muscle cell TP receptor (vTP+) and those with a decrease (vTP−). Concentration-response curves were obtained by the cumulative addition of drugs. Values are means ± SE expressed as %response to KCl (40 mM).

14C-arachidonic acid metabolism.

Rabbit aorta was obtained from vTP+ rabbits as described above. To determine the effect of angiotensin II on arachidonic acid metabolism, strips of aorta were placed in HEPES buffer and incubated at 37°C for 15 min with 14C-arachidonic acid (0.05 μCi, 50 μM) and angiotensin II (10 μM) as previously described (26). After incubation, the HEPES buffer was removed, acidified to pH 3.0 with glacial acetic acid, and extracted over Bond Elut octadecylsilica (ODS) columns. The ODS columns were washed sequentially with 5 ml water and ethanol. The acidified sample (made 15% vol/vol with ethanol) was then added to the column and washed with 5 ml each of 15% ethanol and water. The arachidonic acid metabolites were eluted with 6 ml ethyl acetate, evaporated to dryness under N2, and stored at −40° C until analysis by reverse-phase HPLC. The prostaglandin-like metabolites of arachidonic acid were separated using a reverse-phase HPLC system utilizing a Nucleosil-C18 column. Solvent A was water containing 0.025 M phosphoric acid, and solvent B was acetonitrile. The program consisted of a 40-min isocratic phase with 31% solvent B in solvent A, followed by a 20-min linear gradient to 100% solvent B and a 10-min isocratic phase with 100% solvent B. The flow rate was 1 ml/min. Elution times of radioactive peaks were compared with retention times of known prostaglandin standards.

Isoprostane analysis.

Arteries were incubated in HEPES buffer containing vehicle, angiotensin II (10−7 M), or xanthine (100 μM)/xanthine oxidase (0.03 U/ml) for 15 min at 37°C. After incubation the HEPES buffer was removed, the internal standard ([2H4]8-iso-PGF) was added to the buffer and the buffer was acidified to pH 3.0 with glacial acetic acid and extracted over ODS extraction columns as described above (26). The isoprostanes were eluted with 6 ml ethyl acetate and then were evaporated to dryness under N2 and stored at −80°C until analysis. Samples were analyzed by using liquid chromatography-electrospray ionization mass spectrometry (LC-ESI-MS, Agilent 1100 LC/MSD, SL model) using a modification of a method previously described by Nithipatikom et al. (20). The samples were separated on a reverse-phase C18 column (Kromasil, 250 × 2 mm) using water/acetonitrile with 0.1% acetic acid as a mobile phase at the flow rate of 0.3 ml/min. The gradient started at 35% acetonitrile, linearly increased to 100% acetonitrile in 35 min, and held for 10 min. The detection was made in the negative mode. Selective ion monitoring is used for quantitation. The mass-to-charge ratio (m/z) of 353 and 357 is used for 8-iso-PGF, and [2H4]8-iso-PGF, respectively. A standard curve was constructed over the range of 5 to 1,000 pg per injection. The concentration of 8-iso-PGF in the samples was calculated by comparing its ratio of peak area to the standard curve. The results were normalized to the wet weight of the vessel. Studies were repeated three times. To further characterize products, MS/MS analysis was performed by electrospray ionization triple quadruple mass spectrometer (Waters). The parent ion m/z 353 was fragmented by collision-induced dissociation using argon gas. Only the precursor ion is allowed to pass through the first quadrupole, and the ion is activated with argon in the second quadrupole. Product ion spectra were recorded for the m/z range of 50 to 380. Data were acquired in the profile mode. Results were processed using Masslynx software (Micromass).

Biological activity of 8.5-min isoprostane.

Aortas from four to eight rabbits were incubated as before with angiotensin II. Identical control (cell free) incubations without tissue were carried out in parallel. Following incubation and extraction, the samples were chromatographed on the LC as described above. Fractions eluting with the 8.5-min peak were collected, extracted with cyclohexane/ethyl acetate, dried down under a stream of N2 and stored under nitrogen at −40° C until vascular reactivity studies were performed. The biological sample (or cell-free control) was suspended in 100 μl of ethanol (10 μl per 6-ml bath was the maximal concentration administered) and tested for vasoconstrictor activity under basal tone and before the administration of increasing concentrations of angiotensin II (10−11–10−7 M) or phenylephrine (10−9–10−5 M). Results were compared with vessels treated with a submaximal concentration of 8-iso-PGF (5 nM) added before the vasoconstrictor.


U46619, SQ29548, 8-iso-PGE2, 8-iso-PGF, [2H4]-8-iso-PGF, and TXB2 were from Cayman Chemical (Ann Arbor, MI). 14C-arachidonic acid was obtained from New England Nuclear. Angiotensin II was from Peninsula Labs (San Carlos, CA). Indomethacin, phenylephrine, Tiron, and l-NNA were from Sigma (St. Louis, MO); dazoxiben was from Pfizer.

Statistical analysis.

The vascular reactivity data were expressed as the means ± SE. Statistical evaluation of the vascular reactivity data was performed by using a repeated-measures two-way ANOVA followed Bonferronni posttest when significant differences were present. A value of P < 0.05 was considered statistically significant.


Pretreatment of vTP+ aortas with the TP receptor antagonist SQ29548 resulted in a rightward shift of the concentration-response curve to angiotensin II (Fig. 2A). The maximal contraction to angiotensin II was reduced from 134 ± 16 to 93 ± 10% with SQ29548 (P < 0.05). The contractile response of aortic rings from vTP+ and vTP− rabbits to angiotensin II is shown in Fig. 2B. Angiotensin II produced a concentration-dependent contractile response in both vTP+ and vTP− rabbits but the contraction was greater in vTP+ rabbits compared with vTP− rabbits (maximal response; vTP− vs. vTP+: 69 ± 6 vs. 132 ± 11%; P < 0.01). The log EC50 values for angiotensin II in the vTP+ and vTP− aortas were not different (1.36 and 1.63 nM, respectively). SQ29548 had no effect on angiotensin II-induced contractions of vTP− aortas (data not shown). Studies were also performed using a resistance-sized vessel, the mesenteric artery, and results shown in Fig. 3. Similar to what was observed in the aorta, blockade of TP receptors or absence of TP receptors in vTP− rabbits attenuated angiotensin II-induced contractions in mesenteric arteries.

Fig. 2.

Contractile responses to angiotensin II (ANG II). Concentration-response curves were obtained by the cumulative addition of angiotensin II in the presence or absence of SQ29548 (10 μM; A) in vTP+ or vTP− aortas (B). Values are means ± SE expressed as %response to KCl (40 mM). Each value represents means ± SE for n = 12. *P ≤ 0.05, SQ29548-treated compared with control (A) or vTP+ vs. vTP− (B).

Fig. 3.

Contractile responses to angiotensin II (ANG II). Concentration-response curves were obtained by the cumulative addition of angiotensin II in the presence or absence of SQ29548 (10 μM) in vTP+ or vTP− mesenteric arteries. Values are means ± SE expressed as %response to KCl (60 mM). Each value represents means ± SE for n = 12. *P ≤ 0.05, SQ29548-treated compared with control (A) or vTP+ vs. vTP− (B).

Next, the studies determined if TXA2 was the mediator of the enhanced angiotensin II-induced contractions in the vTP+ rabbits. Blockade of TXA2 synthesis with the COX inhibitor indomethacin or the specific TX synthase inhibitor dazoxiben had no effect on angiotensin II-induced contractions in vTP+ aortas (data not shown). Segments of vessels from vTP+ aortas were incubated with 14C-arachidonic acid in the presence and absence of angiotensin II (10 μM), and the 14C-metabolites were resolved by reverse-phase HPLC. Radioactive products comigrating with 6-keto-PGF, PGE2, and PGF were detected. There was no evidence that rabbit aortas produced TXB2, the stable metabolite of TXA2, either under basal conditions or when stimulated with angiotensin II (Fig. 4).

Fig. 4.

Metabolism of 14C-arachidonic acid (AA) in vTP+ aortas. Vessels with (B) and without (A) angiotensin II (0.1 μM) were incubated with HEPES buffer at 37°C for 15 min with 14C-AA. Metabolites of arachidonic acid were separated by reverse-phase HPLC as explained in the materials and methods. Migration time of known standard eicosanoids are shown above the chromatograms. TXB2, thromboxane B2; cpm, counts per minute.

Removal of the endothelium (data not shown) or prior treatment with the NO synthase inhibitor l-NNA (data not shown) had no effect on angiotensin II-induced contractions. However, if denuded vessels were pretreated with the TP receptor antagonist, angiotensin II contractions were reduced (Fig. 5A). Because reactive oxygen species are increased by angiotensin II in vascular smooth muscle cells, an additional study evaluated angiotensin II-induced contractions in vTP+ rabbits in the presence of the SOD mimetic Tiron. Inhibition of superoxide in both intact and denuded vessels attenuated angiotensin II-induced contractions (Fig. 5, B and C). To compare responses in denuded vessels with intact vessels and to control for variations in responses to angiotensin II that occur between rabbits, the Tiron and SQ29548 results were expressed as percent inhibition from control responses in both the intact and denuded vessels. Results are shown in Table 1 and indicated that a similar inhibitory effect is seen when intact and denuded vessel responses are compared.

Fig. 5.

Effect of Tiron (30 mM; B and C) and SQ29548 (10 μM; A) on contractile responses to angiotensin II in vTP+ aortas with (B) and without endothelium (A and C). Values are means ± SE expressed as %response to KCl (40 mM). Each value represents means ± SE for n = 8. *P ≤ 0.05, Tiron-treated compared with control.

View this table:
Table 1.

Percent inhibition of angiotensin II-induced contractions in the presence of SQ29548 and Tiron

The hypothesis was explored that the isoprostane 8-iso-PGF may contribute to angiotensin II-induced contractions in the rabbit arteries. The 8-iso-PGF elicited concentration-dependent contractions in vTP+ aortas. The effect was dependent on an interaction with the TP receptor since no contractions were observed in the vTP− aortas or when vTP+ aortas were treated with SQ29548 (Fig. 1B). The production of 8-iso-PGF is evident in isolated aortas of vTP+ and vTP− rabbits, but there was no change in 8-iso-PGF production following angiotensin II treatment (vTP+ basal vs. angiotensin II: 0.58 ± 0.15 vs. 0.47 ± 0.06 pg/mg; vTP− basal vs. angiotensin II: 0.53 ± 0.02 vs. 0.41 ± 0.14 pg/mg; n = 4). Similar results were observed in vTP+ mesenteric arteries (basal vs. angiotensin II: 0.59 ± 1.7 vs. 1.2 ± 1.7; n = 3). However, the LC-ESI-MS method used to quantify 8-iso-PGF indicated that there were a number of other peaks in which the most abundant molecular ion was m/z 353, the mass that is indicative of a F2-isoprostane structure derived from arachidonic acid. The ion chromatograms of the mass spectra data from the control and angiotensin II-treated samples were overlayed, and a single product that migrated at ∼8.5 min was increased in the angiotensin II-treated aortas (Fig. 6A). A similar product was observed if mesenteric vessels (Fig. 6B) were used or if the aortic vessels were pretreated with the superoxide-generating system xanthine/xanthine oxidase (Fig. 6C). When denuded vessels were incubated with angiotensin II, the production of the 8.5-min peak was similar to that observed in intact vessels (data not shown). Pretreatment of vessels with Tiron (30 mM) diminished the production of the 8.5-min peak (data not shown). The mass spectrum of the 8.5-min product is shown in Fig. 7. The dominant ion in the spectrum is m/z 353, which corresponds to the [M-H]-ion and indicates a molecular weight of 354. This is indicative of the 15-series F2-isoprostanes. Collision-induced dissociation of the m/z 353 ion produced a series of daughter ions. The sequential losses of water (m/z 353 to m/z 335 [M-H-H2O], m/z 317 [M-H-2H2O], and m/z 299 [M-H-3H2O]) indicated the presence of three hydroxyl groups. Further confirmation that this product is an isomer of the 15-series F2-isoprostanes comes from the m/z 309 and m/z 193 ions. The most characteristic ion for the F series occurs with the loss of 44 [m/z 309, M-H-C2H4O] that is a charge-site remote loss of C2H4O from the cyclopentane ring. The loss of 44 occurs most predominantly in isoprostanes that have the 1,3-cyclic-diol structure and has been previously described for prostaglandins and TX, which share this common structural modality (19). The loss of 44 can also be due to carboxyl group and loss of CO2. The m/z 193 ion is another characteristic ion of the F series.

Fig. 6.

Representative of liquid chromatography-electrospray ionization mass spectrometry (LC-ESI-MS) selected ion chromatogram (m/z = 353) of vTP+ vessels incubated in HEPES buffer with and without angiotensin II [0.1 μM; aorta (A) or mesentery artery (B)] or xanthine (100 μM)/xanthine oxidase (0.03 U/ml; aorta; C). Samples were extracted and analyzed by LC-ESI-MS as described in the materials and methods. Migration times of known isoprostane standards are shown. Arrow points to the 8.5-min peak increased by angiotensin II or xanthine/xanthine oxidase. These studies were repeated 3 separate times.

Fig. 7.

MS-MS analysis of 8.5 peak increased by angiotensin II in rabbit aorta. Inset: chemical structure of the 8-iso-PGF isomer for reference only. m/z, Mass-to-charge ratio.

The 8.5-min peak was isolated from aortas incubated with angiotensin II and tested for activity on isolated rings of rabbit aorta. The 8.5-min factor had no effect on basal tone. However, if the factor was added before the vessel was contracted with angiotensin II, there was an enhanced contractile response to angiotensin II [contraction to angiotensin II (10−8 M); control vs. 8.5-min peak: 39 ± 5 vs. 64 ± 8%; P < 0.05; Fig. 8A]. The vehicle control from a cell free incubation had no effect on angiotensin-II -induced contractions (data not shown). There is evidence that subthreshold concentrations of 8-iso-PGE2 and 8-iso-PGF potentiate vasoconstrictions induced by angiotensin II in the isolated perfused rabbit ear (31). This effect was also seen with 8-iso-PGF (5 nM) using isolated rings of rabbit aorta [contraction to angiotensin II (5 × 10−8 M); control vs. 8-iso-PGF: 27 ± 9 vs. 56 ± 11%; P < 0.05; Fig. 8B]. This effect of the 8.5-min peak or 8-iso-PGF also occurred if phenylephrine was used to contract the blood vessels [data not shown; contraction to phenylephrine (10−7 M); control vs. 8.5-min peak: 50 ± 4 vs. 60 ± 4%; control vs. 8-iso-PGF: 56 ± 3 vs. 70 ± 5%; P < 0.05]. In vTP− rabbits, there was no effect of 8-iso-PGF or the 8.5-min peak to enhance angiotensin II- or phenylephrine-induced contractions (data not shown).

Fig. 8.

Effect of subthreshold concentration of 8-iso-PGF (5 nM; B) and the 8.5-min isoprostane peak (A) on angiotensin II-induced contractions. Fractions corresponding to the 8.5-min peak were collected from the LC system shown in Fig. 6. Following extraction, the product was added to vTP+ rabbit aortic rings suspended in tissue baths before the administration of increasing concentration of angiotensin II (10−10–10−6 M). Data are expressed as percentage of the KCl contraction and are shown as means ± SE for n = 4. *P < 0.05, for 8-iso-PGF or 8.5-min peak vs. vehicle control.


This study explored the hypothesis that a portion of angiotensin II-induced contractions is dependent on superoxide generation and the release of a mediator that activates vascular smooth muscle cell TP receptors. The major findings confirmed that a TP receptor antagonist attenuates angiotensin II-induced contractions in rabbit aorta and mesentery artery. If rabbits lack vascular TP receptors, there are also reduced contractile responses to angiotensin II. Neither TXA2 nor PGH2 contribute to angiotensin II-induced contraction, but an inhibitor of superoxide reduces angiotensin II-induced contractions. Furthermore, angiotensin II increases the production of a F2-isoprostane that is not 8-isoPGF, and this F2-isoprostane can potentiate angiotensin II-induced contractions through a mechanism that involves the vascular TP receptor.

The first series of experiments showed that angiotensin II caused a concentration-related contractile response that was partially inhibited by pretreatment with SQ29548. Previous studies (17, 37) by others indicated that in certain blood vessels angiotensin II increased TXA2 synthesis. There is no evidence that the rabbit aorta produces TXA2 (24), and if vessels were pretreated with dazoxiben, a specific inhibitor of TX synthase, there was no difference in angiotensin II-induced contractions compared with the controls. Some studies (2, 27) reported that the endoperoxide metabolite PGH2 mediates endothelium-dependent contractions via an interaction with the TP receptor. However, in the present study, pretreatment of aortas with the COX inhibitor indomethacin also had no effect on the response to angiotensin II eliminating the COX metabolites PGH2 or TXA2 as mediators. Finally, to support that a portion of angiotensin II-induced contractions required an interaction with the vascular TP receptor, we used the vascular TP-receptor-deficient rabbits. These rabbits have been well characterized in previous studies (5, 25) and have been shown to contract similarly to agonists like KCl, endothelin, and norepinephrine. This is the first study to evaluate angiotensin II responses, and results showed that contractions were decreased in vTP− compared with the vTP+ rabbits.

To explain the mechanistic interaction between angiotensin II and TP receptors, it is hypothesized that angiotensin II increases superoxide production and that superoxide contributes to contractions via a mechanism that involves the TP receptor. A scavenger of superoxide, Tiron, inhibited a portion of the contractions to angiotensin II in the vTP+ rabbit aorta. In vTP− aortas, Tiron had no effect on angiotensin II-induced contractions supporting a role for the TP receptor in the response. In the isolated rabbit aorta, vascular cells other than the endothelial cell are the major source of superoxide. Pagano and coworkers (21, 22) showed that mechanical removal of the endothelium had no effect on the production of superoxide when vessels were incubated with the SOD inhibitor diethyldithiocarbamate. The present studies showed that angiotensin II-induced contractions were similar in both the intact and denuded vessels. Furthermore, a portion of the angiotensin II-induced contractions in the denuded vessels was still dependent on superoxide and TP receptor activation, as both Tiron and SQ29548 attenuated the response. The magnitude of inhibition to Tiron or SQ29548 was similar in intact and denuded vessels. Other studies (28, 34) have shown that the increase in superoxide by angiotensin II inactivates NO. If this pathway was present in the rabbit aorta, then treatment of the vessels with the NO inhibitor l-NNA should make more superoxide available and enhance angiotensin II contractions. However, in rabbit aorta, l-NNA did not alter angiotensin II-induced contractions.

Numerous reports support an interaction between TXA2 and angiotensin II (7, 13, 17, 38, 39). In mice with a targeted disruption of the TP receptor gene and subjected to chronic angiotensin II infusion, the absence of TP receptors blocked the development of hypertension (7, 12). Wilcox and coworkers (33, 36) studied rabbits following acute low dose (60 ng·kg−1·min−1), or high dose (200 ng·kg−1·min−1) angiotensin II infusion. High-dose but not low-dose angiotensin II increased blood pressure and markers of oxidative stress in the renal afferent arterioles. Isolated renal afferent arterioles from both groups had enhanced angiotensin II-mediated contractions compared with the sham infusion. Blockade of TP receptors decreased the contractions. With high-dose angiotensin II infusion, there was an even greater increase in angiotensin II contractions that was inhibited by both a TP receptor antagonist and SOD. The present study did not examine a model of angiotensin II-induced hypertension but instead investigated whether the acute vasoconstrictor response to angiotensin II was also dependent on the activation of the TP receptor. In both a conduit vessel (aorta) and resistance-sized vessel (mesenteric artery), a portion of the response to angiotensin II was inhibited by a TP receptor antagonist.

Isoprostanes represent a unique series of PG-like compounds that were identified as products of the peroxidation of arachidonic acid catalyzed by oxygen free radicals (3, 6, 30). At least two isomers of the isoprostanes, 8-iso-PGF and 8-iso-PGE2, possess significant biological activity (35) acting as vasoconstrictors through a TP receptor mechanism (10, 14, 16, 32). Since the arachidonic acid inhibitor studies indicated that neither TXA2 nor PGH2 contributed to angiotensin II-induced contractions in the rabbit aorta, we explored the possibility that angiotensin II released an isoprostane, like 8-iso-PGF, that may then contribute to angiotensin II-induced contractions. While we were able to measure 8 iso-PGF in the rabbit vessels, angiotensin II did not increase its production. Based on the mechanism of isoprostane synthesis, four endoperoxide regioisomers are formed and then further reduced to the isoprostanes. The four regioisomers are denoted as either 5-(type VI), 12-(type V), 8-(type IV), or 15-(type III) series depending on the carbon atom to which the side chain hydroxyl is attached. For example, 8-iso-PGF is also called 15-F2T-Iso P or iPF-III. As an indication of the large number of potential isoprostanes, each of the four regioisomers can theoretically be comprised of eight racemic diastereomers. Therefore, the possibility exists for the formation of 64 different F2-isoprostanes. There is direct evidence for the formation of each of the four classes of regioisomers from both in vitro and in vivo studies (15). Very few of these regioisomers have been studied for effects on vascular tone. In the present study, the LC-ESI-MS results showed that angiotensin II increased a unique isoprostane from blood vessels and that this production was not dependent on the endothelium. LC/MS/MS analysis of this metabolite revealed a molecular weight of 354 that is indicative of the F2-isoprostanes. The ions observed at m/z 309 and m/z 193 ions gave further evidence that the isoprostane increased by angiotensin II is of the F series (19). Further characterization studies are ongoing to determine the stereochemistry of the hydroxyl and alkyl groups and should provide more definitive structural identification of the isoprostane isomer.

Finally, it was important to show that the F2-isoprostane had biological activity if it is a mediator of enhanced angiotensin II-induced vasoconstriction. The F2-isoprostane had no vasoconstrictor activity on basal tone. A possible explanation may relate to the amount of the unknown metabolite that is isolated from the rabbit aorta and subsequently tested for biological activity. Because we are isolating an unknown factor, it is difficult to quantify its production, and therefore, it may be that we are not adding a large enough concentration to see a direct constrictor effect. Several extraction and HPLC steps are required for isolation, which would also contribute to a loss of mass of the unknown product. However, the F2-isoprostane did enhance angiotensin II-mediated contractions. A similar enhancement was observed if a subthreshold concentration of 8-iso-PGF was tested. The effect is dependent on the TP receptor since there was no enhanced effect in vTP− rabbits. The ability of isoprostanes to amplify responses is not restricted to angiotensin, as both 8-iso-PGF and the newly identified F2-isoprostane also increased phenylephrine-induced contractions through a TP receptor-mediated mechanism. The results suggest that isoprostanes increase sensitivity to vasoconstrictor agonists, a potentially detrimental effect in conditions like atherosclerosis, diabetes, and hypertension. This mechanism may be especially important in diseases in which oxidative stress, leads to increased generation of isoprostanes. Isoprostanes are well recognized as biomarkers of oxidative stress but these compounds also display important biological effects. Much of the available data on isoprostane biological effects are related to known synthetic derivatives, like 8-isoPGF and 8-iso-PGE2. This study describes novel findings concerning previously unidentified, biologically derived isoprostanes that are potentially linked to the TP receptor. Further characterization of this pathway has the potential to advance our knowledge in the possible causes of vascular diseases, ultimately leading to better therapeutic treatments.


This study was supported by research National Heart, Lung, and Blood Institute Grants HL-093181 (to S. L. Pfister) and HL-37981 (to W. B. Campbell).


No conflicts of interest, financial or otherwise, are declared by the author(s).


We thank Marilyn Isbell and Thivashnee Pillay for technical assistance.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
View Abstract