Adrenomedullin (ADM) is a novel hypotensive peptide that exists in abundant amounts in normal adrenal medulla as well as in pheochromocytoma tissue arising from adrenal medulla⁽¹⁾. The peptide, consisting of 52 amino acids, circulates in plasma in considerable concentration (3.3 ± 0.4 pM)⁽²⁾ and has a potent and long-lasting hypotensive effect⁽³⁾. Synthetic rat ADM has been reported to have a potent and long lasting hypotensive effect in anesthetized rats⁽⁴⁾. Studies by Lippton and co-workers demonstrated that intravenous bolus injections of human ADM and a fragment, ADM_13-52 decreased systemic arterial pressure in a dose-dependent manner without altering cardiac output or heart rate and the systemic vasodilator activity of ADM was conserved across species⁽⁵⁾. ADM was found to have greater systemic vasodilator activity in its native form⁽⁶⁾; ADM and ADM_13-52 had greater vasodilator effect in thromboxane A₂ (TXA₂)-activated pulmonary artery than in the systemic artery, and ADM_1-12 had no effect on pulmonary and systemic artery in vivo⁽⁷⁾. Other studies suggest that the mechanism of ADM-induced vasorelaxation is stimulation of cyclic AMP in vascular smooth muscle cells (VSMC)⁽⁸⁾ and is largely independent of nitric oxide (NO)⁽⁹⁾.

Acute hypoxia-induced vasoconstriction of pulmonary artery has been demonstrated in different animal species^(10-14). Chronic hypoxic lung disease in humans also leads to pulmonary arterial hypertension⁽¹⁵⁾. Dinh-Xuan and colleagues⁽¹⁶⁾ showed that endothelium-dependent vasorelaxation is impaired in pulmonary arteries from patients with end-stage chronic hypoxic lung diseases and suggested that the impairment of endothelium-dependent pulmonary artery relaxation contributes to the development of pulmonary hypertension in chronic lung disease. The hypoxic pulmonary vasoconstriction does not respond to ordinary vasodilators that modulate cyclic nucleotides^(16,17). Because of the potent pulmonary vasodilator properties of ADM, we examined the effect of ADM and its fragment ADM_13-52 on pulmonary artery rings subjected to hypoxic constriction.

MATERIALS AND METHODS

Preparation of pulmonary arterial rings

Rats were anesthetized with halothane. The chest was opened, and heparin sulfate (100 IU) was injected into the right ventricle. The heart and lungs were then removed en bloc and placed in ice-cold Krebs-Ringer buffer (K-R: CaCl₂ 2.5 mM, MgCl₂ 1.2 mM, KCl 4.7 mM, NaCl 118 mM, KH₂PO₄ 1.2 mM, D-glucose 11.1 mM, EDTA-Na 0.01 and 12.5 mM NaHCO₃) gassed with 95% O₂/5% CO₂ for a final pH of 7.35-7.45⁽¹⁸⁾. The extrapulmonary left and right branches of the pulmonary artery and the main pulmonary artery were cleaned of all visible connective tissue and fat and cut into 3- 4-mm rings. In some arterial rings, the endothelium was removed by threading the vessel onto a lightly sanded surgical steel rod and rotating the vessel. Each of the rings was mounted on wire stirrup, suspended in a tissue bath filled with 20% O₂/5% CO₂/75% N₂ continuously bubbled K-R buffer at 37°C⁽¹⁸⁾. The wire stirrups were connected to force transducers (Grass Instruments, force-displacement FT03, Quincy, MA, U.S.A.) to record changes in isometric force on a four-channel recorder (Grass Instruments, model 79E)⁽¹⁸⁾. Each pulmonary ring was equilibrated at an optimal resting tone of 0.7 g for 1 h while the buffer was changed every 30 min. After equilibration, the vascular rings were tested for viability by addition of 40 mM KCl, which resulted in 60-80% of maximal contraction. Endothelial integrity of rings was established by the presence of characteristic 60-90% decrease in preexisting tone in response to acetylcholine (Ach, 10^-6M). The deendothelialized rings failed to relax in response to Ach.

Preparation of aortic rings

Aorta was removed from the anesthetized rat, placed in oxygen-saturated buffer, cleaned of connective tissue and fat, and cut into 3- 4-mm rings. Each ring was mounted on wire stirrup, suspended in a tissue both filled with 95% O₂/5% CO₂ continuously bubbled K-R buffer at 37°C. The changes in isometric force were recorded as already described. Each aortic ring was equilibrated at an optimal resting tone of 2 g for 2 h, with the buffer changed every 30 min. The VSM and endothelial integrity was established as already described.

Experimental protocol

After equilibration, rings with and without intact endothelium were contracted with appropriate concentration of norepinephrine (NE 10^-8-10^-6M) to obtained 0.6-0.7 g/mg tissue of tension and then exposed to cumulative concentrations of ADM (10^-10-10^-6M) or ADM_13-52 (2 × 10^-11-2 × 10^-7M). Parallel endothelium-intact pulmonary arterial rings were treated with the NO synthesis inhibitorN^ω-nitro-L-arginine methyl ester (L-NAME) 10^-4M or the cyclooxygenase inhibitor indomethacin 5 × 10^-5M 10 min before being contracted with NE.

Some endothelium-intact pulmonary arterial rings were exposed to sustained hypoxia, which was induced by switching 20% O₂/5% CO₂/75% N₂ to 95% N₂/5% CO₂. The hypoxia lasted ≈1 h. Under hypoxia, those rings were contracted with NE and exposed to cumulative concentrations of ADM or ADM_13-52. Parallel rings with intact endothelium were pretreated with indomethacin 5 × 10^-5M 10 min before hypoxia. Baseline tension during hypoxia was kept the same as that during normoxia.

Reagents

NE, L-NAME, and Ach were purchased from Sigma Chemical (St. Louis, MO, U.S.A.). Indomethacin was obtained from Merck, Sharp and Dohme (West Point, PA, U.S.A.). ADM and ADM_13-52 were obtained from Phoenix Pharmaceuticals (Belmont, CA, U.S.A.).

Statistical procedures

Vasorelaxation was calculated as percent change from the preexisting tone⁽¹⁸⁾. Data are mean ± SEM. Student'st test for paired or unpaired comparisons was used to establish difference between two means. When more than two means were compared, an analysis of variance were used to determine whether differences existed among the means, and subsequent comparison was made (if applicable) by appropriate tests (e.g., Student-Newman-Keuls test).

RESULTS

Relaxation of aortic and pulmonary arterial rings under normoxia

ADM resulted in a concentration-dependent relaxation of NE-precontracted pulmonary arterial rings with intact endothelium. The ADM-mediated vasorelaxation was abolished by pretreatment of pulmonary arterial rings with L-NAME or by deendothelialization(Fig. 1). ADM induced only modest relaxation of endothelium-intact aortic rings precontracted with NE (<20% of relaxation in pulmonary arterial rings)(Fig. 1).

Fragment ADM_13-52 caused a concentration-dependent relaxation of precontracted pulmonary rings with intact endothelium. The magnitudes of pulmonary arterial ring relaxation by ADM or ADM_13-52 were similar. ADM_13-52 induced relaxation was abolished by pretreatment with L-NAME or by deendothelialization but not by pretreatment with indomethacin(Fig. 1). ADM_13-52 failed to relax precontracted aortic rings with intact endothelium(Fig. 1).

Relaxation of endothelium-intact pulmonary arterial rings under sustained hypoxia

Under hypoxic conditions, ADM resulted in a modest relaxation of precontracted pulmonary arterial rings and the degree of relaxation was concentration dependent, but the extent of vasorelaxation was only one third of that under normoxic condition (p < 0.01). This relaxation was abolished by pretreatment of rings with indomethacin(Fig. 2). In contrast, ADM_13-52 failed to show any relaxation in pulmonary arterial rings during hypoxia(Fig. 2). Studies in pulmonary arterial rings subjected to treatment with L-NAME or deendothelialization under hypoxic condition were not performed, since treatment with L-NAME or deendothelialization abolished the ADM- and ADM_13-52-mediated vasorelaxation during normoxia.

DISCUSSION

Our results show that ADM is a more potent relaxant of pulmonary artery than of aortic rings and a more potent relaxant than ADM_13-52 in aortic rings. Furthermore, ADM, but not ADM_13-52, relaxes hypoxic pulmonary arterial rings through an indomethacin-sensitive pathway, and intact endothelium and NO synthesis is necessary for ADM-induced relaxation.

Lippton and colleagues⁽⁷⁾, studying the vasorelaxant effect of ADM in the pulmonary and systemic vascular beds of the intact cats, reported that ADM and its truncated form ADM_13-52 had little effect on baseline lobar arteries. In contrast, when pulmonary vasomotor tone was actively increased by intralobar arterial infusion of TXA₂ analogue U46,619, intralobar arterial bolus injections of ADM or ADM_13-52 dilated lobar arteries in a dose-dependent manner, whereas only a very high concentration of ADM and ADM_13-52 dilated systemic arteries⁽⁷⁾. The present study confirmed the more potent vasorelaxant effect of ADM, as well as that of ADM_13-52, in pulmonary arterial rings than in aortic rings. In addition, the present study showed absence of relaxation of aortic rings in response to fragment ADM_13-52, indicating that the entire ADM peptide sequence is necessary for its systemic vasorelaxant effect to manifest.

In the study of Lippton and colleagues⁽⁷⁾, fragment ADM_1-12 had no vasodilator effect. Other reports indicate retention of systemic and microvasculature dilatory effect of fragment ADM_13-52^(5,6,19). Therefore, it has been suggested that fragment ADM_13-52 is the active fragment of ADM and that fragment ADM_1-12 is irrelevant to the vasodilator activity of ADM. However, all these studies were conducted under normoxic conditions. We extended the observations on the role of ADM in vasorelaxation during hypoxia and showed that ADM relaxed hypoxic pulmonary arterial rings, whereas ADM_13-52 did not. These observations imply that ADM_1-12 must be involved in the vasodilator activity of ADM under hypoxic conditions.

Several studies have documented that acute hypoxia causes vasoconstriction of pulmonary artery^(10-14,18) and that chronic hypoxic lung disease leads to pulmonary arterial hypertension⁽¹⁵⁾. Pulmonary arteries from patients with end-stage chronic hypoxic lung disease show loss of endothelium-dependent relaxation, which suggests that impairment of endothelium-dependent pulmonary artery relaxation may contribute to the development of pulmonary artery hypertension in chronic lung disease. Effects of some vasorelaxants in restoring pulmonary artery relaxation have been examined in experimental and clinical conditions^(16,17). However, no therapy has yet been successful. In recent studies, in vivo administration of NO precursor L-arginine (L-Arg) to chronically hypoxic rats did not appear to improve the chronic hypoxia-induced loss of endothelium-dependent relaxation, although the tissue L-Arg level was increased⁽¹⁷⁾. Therefore, agents other than NO precursor, such as ADM, should be evaluated for therapeutic use in hypoxic pulmonary hypertension.

ADM was discovered in human pheochromocytoma extract by monitoring of the activity that increased rat platelet cyclic AMP⁽¹⁾. Ishizaka and co-workers⁽⁸⁾ reported that ADM stimulates cyclic AMP formation in rat VSMC, which may be the mediator for ADM-induced vasodilation. Gardiner and associates⁽²⁰⁾ showed that ADM induced hypotension in Long Evans rats and that NO synthesis inhibitor L-NAME slightly but significantly inhibited the vasodilator effect of ADM. In the present study, ADM- and ADM_13-52-induced relaxation of pulmonary arterial rings was abolished by the NO synthesis inhibitor L-NAME as well as by deendothelialization. These results indicate that the presence of endothelium as well as synthesis of NO is necessary for ADM-mediated vasorelaxation, at least in the pulmonary artery.

Endothelium releases NO and prostacyclin (PGI₂) and both are potent vasodilators⁽²¹⁾. PGI₂ stimulated NO release and potentiated its action in pig coronary arteries⁽²²⁾. NO has also been reported to activate cyclooxygenase enzymes⁽²³⁾. Therefore, NO and PGI₂ potentiate the vasorelaxant effect of each other, and one of these mediators may have a dominant role under certain conditions. We noted that the pulmonary vasorelaxant effect of ADM and ADM_13-52 during normoxia requires the presence of intact endothelium and synthesis of NO and that the vasorelaxant effect of ADM under normoxia is not affected by cyclooxygenase inhibitor indomethacin. However, under sustained hypoxia, ADM-mediated relaxation of pulmonary arterial rings is completely blocked by indomethacin. These data indicate that NO may be the dominant mediator for ADM-induced vasorelaxation under normoxia, whereas PGI₂ release becomes the dominant mediator of ADM-induced vasorelaxation during hypoxia. This is especially important because hypoxia impairs NO synthesis, release, and activity^(16-18).

Our results show that the vasorelaxant effect of ADM is more prominent in pulmonary artery than in the aorta; the vasorelaxant effect of ADM is more potent than that of ADM_13-52 in aortic rings; endothelium/NO is necessary for the ADM-induced vasorelaxation during normoxia; and the cyclooxygenase pathway may be critical in the pulmonary vasorelaxant effect of ADM during hypoxia. ADM and ADM_13-52 may prove to be of therapeutic value in hypoxic pulmonary hypertension.

REFERENCES