Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The vasoactive responses to angiotensin IV (Ang IV), the NH₂-terminal deleted hexapeptide obtained from angiotensin II (Ang II) by the successive effects of aminopeptidase A and aminopeptidase N, have been studied extensively in various vascular beds. However, conflicting results were observed. Several reports concluded that Ang IV is a constrictor agent in the mesenteric and hindlimb vascular beds of the cat (Garrison et al., 1995; Garrison and Kadowitz, 1996; Champion et al., 1996; Champion and Kadowitz, 1997) and rat (Gardiner et al., 1993; Champion et al., 1998). These vasoconstrictor responses were mediated by type 1 angiotensin receptor (AT₁) stimulation and also showed that Ang IV had a much lower affinity toward AT₁ than its precursor, Ang II (Champion and Kadowitz, 1997; Champion et al., 1998; Lambert et al., 1998). Other reports demonstrated a vasodilator effect of Ang IV in the renal cortical (Coleman et al., 1998a), cochlear (Coleman et al., 1998b), and brain (Kramar et al., 1997, 1998) circulation, resulting in an increase of blood flow in these organs. This effect was related to a specific type 4 angiotensin receptor (AT₄) subtype because it was not inhibited by the AT₁ antagonists, whereas Divalinal, an AT₄ antagonist, abolished the Ang IV vasodilator response (Kramar et al., 1997; Coleman et al., 1998a). Further complexity originates from the necessity to distinguish the direct and indirect effects of Ang IV. Indeed, several studies reported that Ang IV-induced vasodilation was dependent of nitric oxide release. For example, pretreatment of rats with N^G-nitro-L-arginine methyl ester (L-NAME) blocked the vasodilator effect of Ang IV and of its analog, norleucine 1-Ang IV on the cerebral circulation (Kramar et al., 1998). Similar results were obtained for the renal cortical blood flow (Coleman et al., 1998a). It has been also shown that both L-NAME and meclofenamate, a cyclooxygenase inhibitor, shifted to the left the vasoconstrictor response curve of Ang IV in the pulmonary circulation of the rat, suggesting that Ang IV stimulates the release of vasodilator prostaglandins and nitric oxide (Nossaman et al., 1995). In accordance with these results, Yoshida et al. (1996) reported that the vasoconstrictor response to Ang IV in the rat was enhanced by L-NAME initially and by the association of L-NAME and indomethacin, another cyclooxygenase inhibitor, at a later phase. Opposite results were reported by Li et al. (1997), who concluded that indomethacin did not influence the concentration-response curve for Ang IV in human isolated saphenous veins. Finally, it is also possible that in addition to AT₁ and AT₄, Ang IV also activates the type 2 angiotensin receptor (AT₂) because specific AT₂ antagonists enhanced the pressor response to Ang IV (Nossaman et al., 1995).

The purpose of the present study was to try to distinguish between these multiple effects of Ang IV, with the use of infused rat mesenteric arteries, which represent a well known model for the study of resistance arteries (Halpern et al., 1984). Taken together, our results demonstrate the presence of specific and functionally active AT₄s in supplement of the classical AT₁s and AT₂s in rat resistance arteries.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Twelve-week-old male Wistar-Kyoto rats (Iffa-Credo, Lyon, France), weighing 300 g were anesthetized with pentobarbital (50 mg/kg i.p.). A median laparotomy was performed, and a segment of mesenteric artery, ~150 µm in internal diameter, was isolated, cannulated at both ends, and mounted in a video-monitored perfusion system (Halpern et al., 1984) as described previously (Henrion et al., 1997a,b). Briefly, the cannulated artery was bathed in a 5-ml organ bath containing physiological salt solution of the following composition: 135mM NaCl, 15 mM NaHCO₃, 4.6 mM KCl, 1.5 mM CaCl₂, 1.2 mM MgSO₄, 11 mM glucose, 5 mM HEPES, pH, 7.4. pO₂ was maintained at 160 mm Hg and pCO₂ at 37 mm Hg. The bath solution was changed continuously at a rate of 4 ml/min. The artery was perfused with a similar physiological salt solution at a rate of 100 µl/min under a pressure of 50 mm Hg. Pressures in both ends of the artery segment were monitored by pressure transducers. Pressure in the proximal end of the vessel was controlled by a servo perfusion system. Arterial diameter was recorded continuously by a video-monitored system (Living System Instrumentation Inc., Burlington, VT). Pressure and diameter measurements were collected by a Biopac data acquisition system ( MP 100; Biopac, La Jolla, CA) and recorded and analyzed on a Macintosh Quadra computer (Apple, Cupertino, CA) by Acqknowledge software (Biopac, La Jolla, CA). The viability of each vessel was tested before each experiment by testing the responsiveness of the smooth muscle to KCl (80 mM) and phenylephrine (0.1 µM). The presence of intact functional endothelium was assessed by testing the vasodilator effect of acetylcholine (1 µM) after preconstriction of the mesenteric arteries with phenylephrine (0.1 µM). Vessels that did not fully contract with 80 mM KCl and did not fully relax when 1 µM acetylcholine was applied were excluded.

The procedures followed for the care and euthanasia of rats were conducted in accordance with the European Community standards on the care and use of laboratory animals (Ministère de l'Agriculture, France, authorization no. 00577). In a first series of experiments, arteries were exposed to Ang IV under controlled conditions (n = 7) or after addition of the following drugs: 1) the nitric oxide synthase blocker, L-NAME, (1 µM; n = 4); 2) the cyclooxygenase blocker, indomethacin (10 µM; n = 4); 3) L-NAME (1 µM) plus indomethacin (10 µM; n = 4); 4) the Ang II type 1 receptor blockers, losartan (1 µM; n = 6) or candesartan cilexetil (10 nM; n = 5); 5) the Ang II type 2 receptor blocker, PD 123319 (1 µM; n = 5); and 6) candesartan cilexetil (10 nM) plus PD 123319 (1 µM; n = 5). The agonist was added 30 min later. In a second series of experiments, arteries were preconstricted with phenylephrine (at a concentration sufficient to obtain 50% of maximal contraction: 0.1 to 1 µM) and then exposed to Ang IV under controlled conditions (n = 9) or after addition of either L-NAME (1 µM; n = 5), indomethacin (10 µM; n = 5), L-NAME (1 µM) plus indomethacin (10 µM; n = 5), candesartan cilexetil (10 nM; n = 5), PD 123319 (1 µM; n = 5), or candesartan cilexetil (10 nM) plus PD 123319 (1 µM; n = 6). Arteries were incubated with the different drugs for 30 min. In a third series of experiments, arteries were preconstricted with serotonin (concentration sufficient to obtain 50% of maximal contraction: 0.1 to 1 µM) instead of phenylephrine, and the protocol described above was repeated. In all of these studies, drugs were added in the organ bath and the perfusion fluid as well. Similarly, the agonist was added 30 min later in both media.

Statistical Analysis. Results are expressed as means ± S.E. Significances of the differences between the different groups were determined by analysis of variance (one-factor ANOVA). Means were compared by Dunnett's test when appropriate. p values <.05 were considered to be significant.

Drugs. HEPES, L-NAME, indomethacin, serotonin, phenylephrine, acetylcholine, Ang II, and Ang IV were purchased from Sigma Chemical Co.(St. Louis, MO). PD 123319 was provided by Parke-Davis (Paris, France). Candesartan cilexetil and losartan were donated by Astra Hässle AB (Mölndal, Sweden) and Merck, Sharp and Dohme Research Laboratories (West Point, PA), respectively. Other reagents were purchased from Prolabo (Paris, France).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In isolated mesenteric resistance arteries, addition of Ang IV (1 µM) induced a significant contraction from 173 ± 3 µm to 133 ± 7 µm (p < .05; n = 7; Fig. 1). Concentrations of Ang IV below 1 µM had no significant effect. Pretreatment of the arteries with losartan (1 µM; n = 6; Fig. 1) or candesartan cilexetil (10 nM; n = 7; Fig. 1) suppressed the vasoconstrictor effect of 1 µM Ang IV. However, after pretreatment of the arteries with PD 123319 (1 µM; n = 5) or with candesartan cilexetil (10 nM) plus PD 123319 (1 µM; n = 5; Fig. 1), Ang IV (1 µM) still induced a significant contraction (p < .05). Ang IV (1 µM)-dependent contraction was not significantly affected by L-NAME (1 µM; n = 4; Fig. 2), whereas it was significantly enhanced by indomethacin (10 µM; n = 4) or indomethacin (10 µM) plus L-NAME (1 µM; n = 4; p < .05). Typical recordings indicated that L-NAME, indomethacin, losartan, PD 123319 (data not shown), and candesartan cilexetil (Fig. 1) had no significant effect on the basal arterial diameter when studied individually.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Typical recordings obtained with isolated mesenteric resistance arteries perfused at a rate of 100 µl/min under a pressure of 50 mm Hg. A, contractile effect of Ang IV (1 µM). B, no contractile effect of Ang IV in the presence of the Ang II AT1 blocker candesartan cilexetil (10 nM for 30 min). C, contractile effect of Ang IV in an artery pretreated for 30 min with candesartan cilexetil and with the Ang II AT2 blocker PD 123319 (1 µM). The averaged data obtained from several arteries are given in the bar graph showing the diameter values before (control; n = 7) and after addition of Ang IV (Ang IV; n = 7). The effects of Ang IV in the presence of the Ang II AT1 blockers candesartan cilexetil (CV; 10 nM; n = 7) or losartan (LOS; 1 µM; n = 6), the Ang II AT2 blocker PD 123319 (PD; 1 µM; n = 5), and the association of 10 nM candesartan cilexetil plus 1 µM PD 123319 (+CV +PD; n = 5) are also shown. *p < .05, one-way ANOVA, compared with control.

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Effect of Ang IV (Ang IV; 1 µM; n = 7) in isolated rat mesenteric resistance arteries. The effects of Ang IV (1 µM) in the presence of the nitric oxide synthase blocker, L-NAME (L-N; 1 µM; n = 4) and the cyclooxygenase blocker, indomethacin (indo; 10 µM; n = 4) separately or in combination were also studied. Control refers to the arterial diameter before addition of Ang IV. *p < .05, one-way ANOVA, compared with control. #p < .05, one-way ANOVA, compared with Ang IV + L-N.

Whereas Ang II (1 µM) induced immediate tachyphylaxis, repeated Ang IV (1 µM)-dependent contractions could be observed in the same vessel (Fig. 3).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Typical recordings obtained with isolated mesenteric resistance arteries exposed to either Ang IV (1 µM; lower panel) or Ang II (1 µM; upper panel). Thirty minutes after washout, the arteries were re-exposed to Ang IV or Ang II.

After precontraction of mesenteric resistance arteries with phenylephrine (0.1 to 1 µM), Ang IV (1 µM) induced another and significant contraction from 196 ± 4 µm to 141 ± 6 µm (n = 9; p < .05; Fig. 4). In these conditions, pretreatment of the arteries with indomethacin (10 µM) significantly attenuated the amplitude of the decrease in diameter induced by Ang IV (p < .05; n = 5). Addition of L-NAME (1 µM; n = 5) or of L-NAME (1 µM) plus indomethacin (10 µM) restored the Ang IV-induced contraction to its initial level (n = 5, Fig. 5).

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Contractile effect of Ang IV (Ang IV; 1 µM) in isolated mesenteric resistance arteries preconstricted with phenylephrine (Phe; 0.1 to 1 µM, a concentration sufficient to reach 50% of maximal contraction); n = 9 per group. *p < .05, one-way ANOVA, compared with control. #p < .05, one-way ANOVA, compared with phenylephrine.

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Typical recording obtained from an isolated mesenteric resistance artery preconstricted with phenylephrine (0.1 to 1 µM). The artery was first treated with indomethacin (10 µM) for 30 min and exposed to Ang IV (1 µM). The artery was then treated with L-NAME (1 µM) in addition to indomethacin and again exposed to Ang IV (1 µM). The averaged data obtained from several arteries are given in the bar graph showing the arterial diameter of the arteries preconstricted with phenylephrine (Phe; n = 9) and then subjected to Ang IV before (Phe + Ang IV; n = 9) and after treatment of the arteries with indomethacin (indo; n = 5), L-NAME (L-N; n = 5), or indomethacin plus L-NAME (n = 5). *p < .05, one-way ANOVA, compared with Phe. #p < .05, one-way ANOVA, compared with Phe + Ang IV.

After precontraction with phenylephrine and pretreatment of the arteries with candesartan cilexetil (10 nM; n = 7), Ang IV (1 µM) induced a significant dilation (p < .05; Fig. 6). On the contrary, after precontraction with phenylephrine and pretreatment of the arteries with PD 123319 (1 µM; n = 6) or with candesartan cilexetil (10 nM) plus PD 123319 (1 µM; n = 6), Ang IV (1 µM) induced a significant contraction (p < .05; Fig. 6). A similar protocol was performed after precontraction with serotonin and produced the same results (data not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 6.   Typical recordings obtained from isolated mesenteric resistance arteries preconstricted with phenylephrine (0.1 to 1 µM). A, artery pretreated with the Ang II AT1 blocker, candesartan cilexetil (10 nM) for 30 min and then exposed to Ang IV (1 µM). B, artery pretreated with candesartan cilexetil (10 nM) plus the Ang II AT2 blocker, PD 123319 (1 µM) for 30 min before addition of Ang IV (1 µM). The averaged data obtained from several arteries are given in the bar graph showing the arterial diameter of the arteries preconstricted with phenylephrine (Phe; n = 9) and then exposed to Ang IV before (Phe + Ang IV; n = 9) and after pretreatment of the arteries with candesartan cilexetil (CV; 10 nM; n = 7), PD 123319 (PD; 1 µM; n = 6), or 10 nM candesartan cilexetil plus 1 µM PD 123319 (+CV +PD; n = 6). *p < .05, one-way ANOVA, compared with Phe. #p < .05, one-way ANOVA, compared with Phe + Ang IV.

To make sure that AT₁ blockade was complete under the conditions selected, Ang II (1 µM) was added to vessels pretreated for 30 min with candesartan cilexetil (10 nM). No contraction was observed as demonstrated by the similarity of the vessel diameter values in the presence of candesartan cilexetil alone (333 ± 31 µm) and after addition of Ang II (332 ± 32 µm; n = 3). These results are in accordance with those of Shibouta et al. (1993) and Li et al. (1997). Regarding losartan, the concentration of 1 µM used in the present study largely exceeds that of its K_B (4 nM) as measured by Corriu et al. (1995) on mesenteric arteries exposed to Ang II.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present results show first that Ang IV is a constrictor agent in rat resistance arteries via stimulation of AT₁. This effect is observed with both normal and preconstricted vessels, preconstriction in the latter case being obtained with either phenylephrine or serotonin. The degree of constriction appreciated by the decrease in the vessel diameter was of 23 and 28% under these two conditions, respectively. This would correspond in vivo and, according to Poiseuille's law, to a 2.8- to 3.7-fold increase in resistance. Losartan and candesartan cilexetil, two specific AT₁ antagonists, abolished this effect when studied in normal vessels. Candesartan cilexetil was active at a 100-fold lower concentration than losartan in accordance with its higher affinity toward AT₁ (Nishikawa et al., 1997). Interestingly, PD 123319, an AT₂ antagonist, did not modify the Ang IV-dependent vasoconstriction, but it suppressed the inhibitory effect of candesartan cilexetil. Closely related results were obtained on preconstricted vessels: PD 123319 was also inactive on Ang IV-induced vasoconstriction, and it abolished the moderate vasodilator effect of Ang IV on vessels that had been pretreated with cilexetil candesartan. As reported by others (Gardiner et al., 1993; Garrison et al., 1995; Champion et al., 1998; Lambert et al., 1998), a high concentration of Ang IV (1 µM) was required in both normal and precontracted vessels to obtain the vasoconstrictor effect.

Ang IV-induced vasoconstriction was not influenced by L-NAME, a nitric oxide synthase inhibitor, but it was potentiated by indomethacin or L-NAME and indomethacin in combination, suggesting that Ang IV could stimulate the release of vasodilator prostaglandins that attenuate the vasoconstrictor effect. Opposed results were found in preconstricted vessels. In this case, pretreatment with indomethacin diminished the magnitude of the decrease in the diameter of Ang IV-treated mesenteric arteries, suggesting a role of Ang IV on the release of vasoconstrictor prostaglandins or thromboxanes. There has been, to our knowledge, no report of an effect of Ang IV on prostaglandin synthesis. Such an effect cannot be excluded because Ang IV has been shown to stimulate cytosolic calcium in several preparations, including smooth muscle cells (Dostal et al., 1990) and mesangial cells (Chansel et al., 1998), which might result in phospholipase A₂ activation, arachidonic acid release, and prostaglandin synthesis. Moreover, indomethacin potentiated the vasoconstrictor effect of Ang IV in the rat kidney (Coleman et al., 1998a) and pulmonary circulation (Nossaman et al., 1995). These studies, like ours, do not demonstrate unequivocally that Ang IV acts in part on artery vasoreactivity via prostaglandin production. They show only that there is an equilibrium between Ang IV and various cyclooxygenase products that can be displaced when cyclooxygenase activity is blocked. Consistent with this view is the presence of prostaglandins and thromboxanes in the perfusate of mesenteric arteries. There are both vasodilator (prostaglandin I₂) and vasoconstrictor (prostaglandin F₂ and thromboxane A₂) products. Prostaglandin I₂ and thromboxane A₂ are detected in the form of their metabolites, 6-keto-prostaglandin F₁ and thromboxane B₂, respectively, and a flow dependence of prostanoid production is observed (Matrougui et al., 1997). However, the fact that in our study, indomethacin alone did not modify mesenteric artery vasoreactivity is rather in favor of an Ang IV-dependent prostaglandin release. This conclusion at present is indirect and should be confirmed by measurement of prostaglandin and thromboxane production by the mesenteric arteries after Ang IV treatment. Interestingly, the vasoconstrictor effect of Ang IV could be obtained repeatedly with the same magnitude, contrary to that of Ang II, suggesting an absence of desensitization by Ang IV of AT₁. The AT₁-mediated vasoconstrictor effect of Ang IV is thus predominant in rat resistance arteries as observed by others in various vascular beds (Kramar et al., 1997, 1998; Coleman et al., 1998a,b), but its physiological significance can be questioned considering the high levels of Ang IV required.

Our data also show that Ang IV does not interact with nitric oxide in normal or preconstricted vessels. Under the latter conditions, L-NAME restored Ang IV-induced contraction that had been attenuated with indomethacin, thus indicating that nitric oxide counterbalanced in part the vasoconstrictor effect of unidentified cyclooxygenase products. Noveri et al. (1994) also demonstrated the noninvolvement of nitric oxide in the response to Ang IV in the rat cerebral circulation, but in this case, contrary to the data shown on Fig. 5, it was a vasodilator response. Their opinion was based on the fact that L-NAME did not antagonize the increase in cerebral blood flow in rats with subarachnoid hemorrhage treated by Ang IV. Moreover, Ang IV did not alter nitric oxide synthase activity in cerebral arteries in vitro. Opposite results were reported recently in the cerebral as well as in the renal circulation (Coleman et al., 1998a; Kramar et al., 1998). Furthermore, Ang IV significantly increased endothelial cell constitutive nitric oxide synthase activity and cellular cGMP content in porcine pulmonary endothelial cells. This effect depended on specific AT₄s and caused pulmonary arterial vasodilation (Patel et al., 1998). In our preparation of perfused rat mesenteric artery, the role of nitric oxide was less apparent; it could be demonstrated only when preconstricted vessels were exposed to Ang IV and indomethacin in combination.

Ang IV exerted a vasodilator effect on precontracted mesenteric arteries when these vessels were pretreated with cilexetil candesartan. This vasodilator effect was suppressed by PD 123319, an AT₂ antagonist. Moreover, Ang IV was vasoconstrictor when both AT₁ and AT₂ were blocked. Two conclusions can be drawn from these results: 1) There exists a vasodilator component in the whole effect of Ang IV that is mediated by activation of AT_2. 2) Ang IV activates its own receptors, which are distinct from AT₁ and AT₂, and this activation results in vasoconstriction. Ang IV displays a low affinity toward AT₂ (Bouley et al., 1998), but our finding that inhibition of AT₂ unravels a vasodilatory effect of Ang IV is in accordance with the previous conclusions of Nossaman et al. (1995) in the pulmonary circulation of the rat and of Haberl (1994) in rabbit brain arterioles. In contrast, other studies did not report any effect of AT₂ antagonists on Ang IV-induced vasoconstriction (Garrison et al., 1995; Kramar et al., 1997; Champion et al., 1998). Our results show that Ang IV is still a vasoconstrictor agent when AT₁ and AT₂ are blocked. Mesenteric artery vasoconstriction is observed under these conditions with both normal and preconstricted vessels. This result was unexpected because previously published studies concluded that the specific Ang IV receptor designated as AT₄ mediates vasodilator effects (Kramar et al., 1997; Coleman et al., 1998a,b). However, vasoconstriction is more consistent than vasodilation with the Ang IV-dependent increase in cytosolic calcium that has been observed in many preparations (Dostal et al., 1990; Dulin et al., 1995; Chansel et al., 1998). Location of AT₄ in the vessel wall of the mesenteric artery is still to be determined. Abundant specific AT₄s that are distinct from AT₁ and AT₂ have been characterized in porcine (Riva and Galzin, 1996) and bovine aortic endothelial cells as well as in coronary venular endothelial cells (Hall et al., 1995). Cultured vascular smooth muscle cells (Hall et al., 1993) and mesangial cells (Chansel et al., 1998) also possess AT₄s. Therefore, the AT₄s responsible for the changes in vasoreactivity described in the present study are likely to be distributed both in the endothelial cells and the smooth muscle cells of the mesenteric artery wall.

Our study thus confirms the complexity of Ang IV effects on arterial smooth muscle cell contraction. Ang IV is essentially a vasoconstrictor agent via stimulation of AT₁. This vasoconstrictor effect is modified when the production by the vessel wall of prostaglandins and thromboxanes in the presence of Ang IV is suppressed. Ang IV can also produce a vasodilator effect via activation of AT₂ when AT₁ is blocked. Finally, when both AT₁ and AT₂ are blocked, there exists a residual and direct vasoconstrictor effect of Ang IV that is likely to be mediated by specific AT₄s.