内皮衍生舒张因子合成前体降压机制的研究

本研究试图探讨内皮衍生舒张因子（ＥＤＲＦ）合成前体——左旋精氨酸（Ｌ－Ａｒｇ）的降压机制。从实验动物、临床及离体与整体功能上观察其血液动力学效应。结果显示：给予自发性高血压大鼠（ＳＨＲ）和ＥＤＲＦ合成抑制剂——左旋硝基精氨酸（Ｌ－ＮＮＡ）诱导的高血压大鼠，以及高血压病（ＥＨ）患者静脉滴注Ｌ－Ａｒｇ，均能产生明显的降压作用，其大鼠主动脉环对去甲肾上腺素（ＮＥ）及内皮素（ＥＴ）收缩反应明显减低，对乙酰胆碱（Ａｃｈ）的舒张反应明显增强，ＥＨ患者的心排血量明显增加，外周血管阻力明显减低，并伴有体内主要的收缩因子ＥＴ水平明显下降，ＥＤＲＦ效应产物ｃＧＭＰ明显升高，且尿亚硝酸盐试验阳性。结论：本研究提示Ｌ－Ａｒｇ通过增加体内ＥＤＲＦ的合成释放，产生有效的降压作用。     

红细胞抗高血压因子和川芎嗪对大鼠血管平滑肌cGMP含量的影响

观察红细胞抗高血压因子（ＡＨＦ）和中药川芎嗪（ＴＭＰ）对血管平滑肌细胞环一磷酸鸟苷（ｃＧＭＰ）产生的影响。方法实验用８～１０周的Ｗｉｓｔａｒ大鼠（ｎ＝６）。分离主动脉（Ａ）及肠系膜动脉（ＭＡ），其中Ａ一半去内皮，另一半保留内皮完整，将肌条分别制备成匀浆，用放射免疫分析法测定ｃＧＭＰ含量。结果ＡＨＦ（１０－５ｇ／ｍｌ）和ＴＭＰ（１０－４ｍｏｌ／Ｌ）对血管平滑肌细胞（ＶＳＭＣ）ｃＧＭＰ生成有显著刺激作用。在ＡＨＦ作用下，Ａ组有、无内皮组和ＭＡ组ｃＧＭＰ含量分别是对照组的１．２５倍，１．２６倍和１．７２倍；在ＴＭＰ作用下的实验组ｃＧＭＰ含量分别为对照组的１．６０倍，１．５０倍和１．５２倍，与对照组比较差异均有显著性（Ｐ＜０．０５或Ｐ＜０．００１）。结论ＡＨＦ和ＴＭＰ均能升高Ａ和ＭＡｃＧＭＰ水平，且ＡＨＦ对阻力血管的影响明显高于容量血管，而ＴＭＰ对两种血管的ｃＧＭＰ水平影响无明显差异，ＡＨＦ与ＴＭＰ对血管ｃＧＭＰ的增高作用似乎与内皮无关。     

慢性肺心病患者血浆内皮素-1、环磷酸鸟苷与肺血流动力学的关系及硝普钠对其的影响

对１８例慢阻肺缓解期和２０例肺心病急性发作期患者行血浆ＥＴ－１（内皮素－１）、ｃＧＭＰ（环磷酸鸟苷）放免测定并同时行血流动力学、ＰａＯ２检测及硝普钠治疗观察。结果发现，肺心病患者血浆ＥＴ－１水平显著高于慢阻肺患者，并与ＰＡＰＭ、ＰＶＲ呈显著正相关，与ＰａＯ２显著负相关。血浆ＥＴ－１水平的增加同时伴随血浆ｃＧＭＰ水平继发性的增加，两者呈显著正相关。硝普钠治疗后血浆ｃＧＭＰ水平显著增加，而血浆ＥＴ－１水平无明显变化。提示：缺氧性肺动脉高压不仅使血浆ＥＴ－１水平增加，同时伴随血浆ｃＧＭＰ继发性增加。ＥＴ－１／ｃＧＭＰ比值可能对肺循环压力的调节起着重要作用。     

淋巴因子激活的杀伤细胞与自发性高血压大鼠的一氧化氮样内…

目的 探讨淋巴因子激活的杀伤细胞（ＬＡＫ细胞）与一氧化氮样内皮依赖性舒血管因子（ＥＤＲＦ）作用的相关性。方法动态观察自发性高血压大鼠（ＳＨＲ，ｎ＝５）ＬＡＫ细胞地增殖、活性及其所表达的ＳＯＤ样物质活性当时、自由基清除效应。同时观察胸主动脉环对胆碱（Ａｃｈ）的舒张反应以及上述ＬＡＫ细胞与ＳＯＤ标准品对主动脉环（ｎ＝３０）的Ａｃｈ舒张反应的影响，与等量ＷＫＹ大鼠比较。结果 与ＷＫＹ比，ＳＨＲ的ＬＡＫ细     

淋巴因子激活的杀伤细胞与自发性高血压大鼠的一氧化氮样内皮依赖性舒血管因子

目的探讨淋巴因子激活的杀伤细胞（ＬＡＫ细胞）与一氧化氮样内皮依赖性舒血管因子（ＥＤＲＦ）作用的相关性。方法动态观察自发性高血压大鼠（ＳＨＲ，ｎ＝５）ＬＡＫ细胞的增殖、活性及其所表达的ＳＯＤ样物质活性当量、自由基清除效应。同时观察胸主动脉环对乙酰胆碱（Ａｃｈ）的舒张反应，以及上述ＬＡＫ细胞与ＳＯＤ标准品对主动脉环（ｎ＝３０）的Ａｃｈ舒张反应的影响，与等量ＷＫＹ大鼠比较。结果与ＷＫＹ比，ＳＨＲ的ＬＡＫ细胞增殖、活性及其所表达的ＳＯＤ样物质和其自由基清除效应均明显降低（Ｐ＜０．０５），经ＬＡＫ细胞（无论来自ＷＫＹ还是ＳＨＲ）孵育后的主动脉环对Ａｃｈ的舒张反应均有不同程度增加。标准ＳＯＤ与ＬＡＫ细胞有相同效应。结论ＬＡＫ细胞可能通过表达ＳＯＤ样物质来减少ＮＯ氧化分解从而增强主动脉环的ＮＯ样舒血管作用。     

内皮源舒张因子合成抑制剂致大鼠心血管变化及意义

目的：探讨内皮源舒张因子（ＥＤＲＦ）合成受抑制时的心血管病理生理的变化及意义。方法：实验组大鼠（ｎ＝６）长期应用ＥＤＲＦ抑制剂──Ｌ－硝基精氨酸１５ｍｇ，ｋｇ~（－１）／ｄ，腹腔注射，共２８天），对照组（ｎ＝６）给予注射用水。结果：与对照组比较，Ｌ－硝基精氨酸使大鼠血压明显增高、心率减慢、心功能增强，主动脉环对去甲肾上腺素（１０~（１１）～１０~（－５）ｍｏｌ／Ｌ）和内皮素（１０~（－１０）～１０~（－６）ｍｏｌ／Ｌ）的收缩反应增强（Ｐ＜０．０５），对乙酰胆碱（１０~（－９）～１０~（－５）ｍｏｌ／Ｌ）和降钙素基因相关肽（１０~（－１０）～１０~（－８）ｍｏｌ／Ｌ）舒张反应减弱（Ｐ均＜０．０５），血浆和组织内皮素增高，主动脉组织环磷酸鸟苷活性减低，血浆血管紧张素Ⅱ和肾素活性减低，血浆丙二醛增高。肾入球小动脉内膜增厚。结论：ＥＤＲＦ对心血管有重要调控功能，ＥＤＲＦ减少可导致心血管一系列病理生理变化。     

L-arg对高胆固醇血症兔血管反应性的影响

本研究观察ＮＯ前体物Ｌ－ａｒｇ的抗动脉粥样硬化作用机理。将１６只雄性新西兰白兔按体重随机分为３组：正常对照组（Ｄ组,ｎ＝５）,单纯胆固醇组（Ｃ组,ｎ＝６）和复合组（Ｌ组,ｎ＝５）。实验开始前及实验第５周和第１０周抽血检测血浆总胆固醇,第１０周处死动物,检测血浆和胸主动脉组织中内皮素－１（ＥＴ－１）、环－磷酸鸟苷（ｃＧＭＰ）含量,并测试了胸主动脉环对去甲肾上腺素（ＮＥ）的反应性。结果发现Ｌ－精氨酸对血浆总胆固醇无明显影响,但可使血浆及组织中ＥＴ－１含量显著下降,ｃＧＭＰ含量上升,补充ＮＯ前体物可改善胸主动脉环对缩血管物质去甲肾上腺素的反应性。由此可见,补充Ｌ－ａｒｇ可纠正高胆固醇引起的ＥＴ－１／ｃＧＭＰ失衡,具有一定的抗动脉粥样硬化作用。     

甲状腺激素对左室心肌弛张功能的影响及其可能机制

本文用心脏超声及放射免疫法分别检测２０例甲状腺功能亢进症（甲亢）患者予抗甲状腺素治疗前后之心脏弛张功能（ＩＩａ－ＭＶＯ、ｍａｘＲＦＶ、ＰｅａｋＥ、ＰｅａｋＡ、Ｅ／Ａ）、心脏前与后负荷（ＬＶＥＤＶ、ＥＳＳ）、左室收缩力（ＥＳＳ／ＬＶＥＳＶ）及血清Ｔ３和血浆环磷酸腺苷（ｃＡＭＰ）。结果显示：治疗前甲亢组与对照组比较，ＩＩａ－ＭＶＯ明显缩短、ｍａｘＲＦＶ、ＰｅａｋＥ、ＰｅａｋＡ明显增高。甲亢组血浆ｃＡＭＰ、Ｉｌａ－ＭＶＯ、ｍａｘＲＦＶ、ＰｅａｋＥ和Ｅ／Ａ与血清Ｔ３水平有显著的简单相关。血浆ｃＡＭＰ、ＩＩａ－ＭＶＯ、ｍａｘＲＦＶ和ＰｅａｋＥ与Ｔ３有显著偏相关，但ＰｅａｋＥ与Ｔ３之偏相关系数小于年龄及前负荷与Ｔ３之偏相关系数。甲亢组经平均１８０天抗甲状腺素治疗后，其左室舒张功能、血浆ｃＡＭＰ与治疗前比较有显著差异，而与对照组无差异。     

胰岛素增强内皮素—1的缩血管作用

为观察胰岛素对内皮素－１（ＥＴ－１）缩血管作用的影响，将高血压大鼠（ＳＨＲ）和正常血压大鼠（ＷＫＹ）的离体主动脉环置于高出生理浓度的胰岛素环境中。结果发现：ＥＴ－１对ＷＫＹ／ＳＨＲ离体血管环的作用特点是低浓度下明显收缩，随着浓度的增加收缩作用显著增强，血管去内皮后增强程度减弱；与ＷＫＹ血管相比，ＳＨＲ血管达到最大收缩的时间显著延长。加入胰岛素后，ＥＴ－１对ＷＫＹ和ＳＨＲ主动脉环收缩或舒张的反应特点相同，但作用强度明显加强，ＷＫＹ增加的最大收缩为２４％而ＳＨＲ为１６％（Ｐ＜０．０５或Ｐ＜０．０１）。认为：（１）高血压动脉血管对ＥＴ－１具有耐受性；（２）胰岛素对ＥＴ－１的缩血管作用具有增敏性，增敏程度不受血管内皮完整性的影响；（３）高胰岛素血症可通过对血管活性物质的增敏作用参与高血压病的发生和发展。     

内皮损伤后大鼠胸主动脉血管反应性的动态观察

本研究在大鼠胸主动脉内皮损伤内膜增生的模型上观察了内皮损伤后血管反应性的动态变化。实验大鼠随机分为对照组、术后３天组、术后１０天组、术后２１天组４组。组织学检查提示内皮损伤后的大鼠胸主动脉术后２１天内皮尚不能完全修复，且有新生内膜形成。血管条张力测定试验表明，３天组大鼠受损血管对去甲肾上腺素（ＮＥ）的最大收缩反应下降（与对照组比，Ｐ＜０．０００１），对阶梯浓度ＮＥ的反应曲线下移，而１０天组已恢复正常；３天组大鼠受损血管对乙酰胆碱（Ａｃｈ）的最大舒张反应亦显著受损（与对照组比，Ｐ＜０．０００１），对阶梯浓度Ａｃｈ的反应曲线下移，且术后２１天仍不能恢复正常；受损血管对硝普钠介导的非内皮依赖性舒张反应不受影响。提示血管内皮受损可致血管反应性异常，且长时间不能恢复。     

一氧化氮在高脂饲养兔胸主动脉的变化和意义

观察高脂饲养兔胸主动脉血管环对去甲肾上腺素（ＮＥ）的反应性及环磷酸鸟苷（ｃＧＭＰ）含量的变化。结果表明其对ＮＥ的反应性降低，组织ｃＧＭＰ的含量减少。左旋精氨酸（Ｌ－Ａｒｇ）和硝普钠（ＳＮＰ）可恢复此反应性，而一氧化氮合成酶（ＮＯＳ）阻滞剂Ｌ－硝基精氨酸（Ｌ－ＮＮＡ）却加重其降低的反应性。提示高脂饲养兔血管环中的一氧化氮（ＮＯ）减少，Ｌ－Ａｒｇ及ＳＮＰ可以逆转之，暗示Ｌ－Ａｒｇ／ＮＯ通路障碍可能是动脉粥样硬化的又一机理。     

Mechanical deformation of vessel wall and shear stress determine the basal release of endothelium-derived relaxing factor in the intact rabbit coronary vascular bed.

We investigated the mechanisms that are responsible for the basal release of endothelium-derived relaxing factor (EDRF), which is likely to be identical with nitric oxide, in the intact coronary circulation. The increase in cGMP content of platelets passing through the coronary bed of the isolated rabbit heart was used as an index of EDRF release. Platelet cGMP content after passage through the heart under control conditions (flow rate of 20 ml/min) amounted to 0.50 +/- 0.10 pmol/mg protein. Inhibition of endothelial nitric oxide synthesis by 30 microM NG-nitro-L-arginine (L-NNA) reduced this amount by more than 60%. Increasing flow rate from 20 ml/min to 40 and 60 ml/min led to flow-dependent dilation as reflected by the subsequent drop in perfusion pressure after an initial rise. The flow-dependent dilation was associated with a significant increase in the normalized platelet cGMP content. L-NNA abolished completely both the flow-dependent dilation and the increase in platelet cGMP content. Increasing shear stress by a strong vasoconstriction (1 nM endothelin-1) at constant flow was also accompanied by a 2.5-fold increase in platelet cGMP content. To investigate whether mechanical forces applied to the vascular wall by the myocardial contraction cycle were also a stimulus for EDRF release, cardiac arrest was induced by a continuous infusion of mepivacaine (final concentration, 0.02%). Under these conditions, a decrease in platelet cGMP content comparable to that after nitric oxide synthesis inhibition was observed in the arrested heart.(ABSTRACT TRUNCATED AT 250 WORDS)     

Reduced production of cGMP underlies the loss of endothelium-dependent relaxations in the canine basilar artery after subarachnoid hemorrhage.

Endothelium-dependent relaxations are inhibited during chronic vasospasm after subarachnoid hemorrhage in the canine basilar artery, although the luminal release of endothelium-derived relaxing factor (EDRF) is maintained. The present study investigated the mechanisms underlying the impaired vascular reactivity and in particular whether the loss of responsiveness of the smooth muscle to EDRF is due to an impaired production of cGMP. Bradykinin and nitric oxide evoked concentration-dependent relaxations in isolated canine basilar arteries with and without endothelium, respectively, which were reduced in the subarachnoid hemorrhage group. Relaxations evoked by M&B22,948 (an inhibitor of cGMP phosphodiesterases) were smaller, but those evoked by the lipophilic cGMP analogue 8-bromo-cGMP were potentiated slightly in the subarachnoid hemorrhage group. The resting levels of cGMP in rings with endothelium (reflecting the effect of spontaneous release of EDRF) and those evoked by bradykinin in rings with endothelium and by nitric oxide in rings without endothelium were diminished in the subarachnoid hemorrhage group. These data indicate that the altered endothelium-mediated relaxations of the smooth muscle after subarachnoid hemorrhage is due, at least in part, to an impaired activation of soluble guanylate cyclase leading to a reduced production of cGMP in the smooth muscle.     

Influence of endothelin on human platelet aggregation and prostacyclin generation from human vascular endothelial cells in culture.

The effects of endothelin (ET) on the function of cultured human umbilical vein endothelial cells (HUVEC) and that of human platelets were investigated with reference to endothelium-derived relaxing factor (EDRF) and PGI2. Considering the platelets, ET had no effect on platelet-rich plasma (PRP) aggregation, the generation of thromboxane A2 ([TXA2]) from platelets, and cytosolic free calcium ion concentration ([Ca++]i), cAMP content ([cAMP]i) or cGMP content ([cGMP]i) in platelets. In contrast, the addition of the solution in which HUVEC had been incubated with ET to PRP produced a decrease in PRP aggregation, [TXA2], and [Ca++]i, and an increase not only in [cAMP]i but also in [cGMP]i in platelets. In the HUVEC pretreated with acetylsalicylic acid (aspirin), this increase of [cGMP]i was not affected, but the HUVEC-mediated decrease in PRP aggregation, [TXA2], and [Ca++]i induced by ET were not completely abolished. However, the pretreatment of HUVEC with a combination of aspirin and L-NG-monomethyl arginine (LNMMA) as an inhibitor of EDRF completely abolished the HUVEC-mediated decrease in PRP aggregation, [TXA2] and [Ca++]i induced by ET, and also abolished the enhancement of [cGMP]i and [cAMP]i in platelets. The PGI2 of HUVEC was enhanced by ET with no changes in [Ca++]i, [cAMP]i and [cGMP]i. The ET-induced enhancement was remarkably attenuated by pretreating the HUVEC with aspirin, but not with LNMMA. We conclude that ET attenuates the aggregation of platelets through a decrease in [TXA2] by an increase in [cAMP]i via the increase in PGI2 of HUVEC, and by an increase in [cGMP]i via EDRF.     

Evidences for the existence of endothelium-derived relaxing factor in the renal medulla

The basal levels of cGMP in renal medulla slices were enhanced when the slices were stimulated with both endothelium-dependent (acetylcholine) and endothelium-independent (molsidomine) vasodilators. When preincubated with NG-mono-methyl-L-arginine, a specific inhibitor of endothelium-derived relaxing factor, only the acetylcholine-stimulated increase was completely abolished. Furthermore, a preincubation with L-arginine, a selective precursor of endothelium-derived relaxing factor, enhanced the cGMP levels. The results indicate that the renal medulla, presumably the endothelial cells of the vasa recta, is able to produce endothelium-derived relaxing factor.     

Endothelium-dependent responses of cerebral arterioles to adenosine 5'-diphosphate.

The goal of this study was to determine the role of endothelium-derived relaxing factor (EDRF) in dilatation of cerebral arterioles in response to adenosine 5'-diphosphate (ADP). Using intravital microscopy, we measured the diameters of pial arterioles in vivo in rats during superfusion with ADP, acetylcholine and nitroglycerin before and during topical application of inhibitors of nitric oxide synthesis (NG-monomethyl-L-arginine; L-NMMA and NW-nitro-L-arginine; L-NNA). Prior to suffusion with L-NMMA and L-NNA, ADP (10 and 100 microM), acetylcholine (0.1 and 1.0 microM), and nitroglycerin (1.0 and 10 microM) produced dose-related increases in the diameter of pial arterioles. Following application of L-NMMA (1.0 and 10 microM) and L-NNA (0.1 and 1.0 microM), dilatation of cerebral arterioles in response to ADP was significantly inhibited. In addition, L-NMMA and L-NNA significantly inhibited dilatation of cerebral arterioles in response to acetylcholine, but did not alter vasodilatation in response to nitroglycerin. Thus, our findings suggest that ADP dilates cerebral arterioles via the production of nitric oxide, or a nitric-oxide-containing compound.     

Development and mechanism of a specific supersensitivity to nitrovasodilators after inhibition of vascular nitric oxide synthesis in vivo.

The mechanism of the increased sensitivity to nitrovasodilators after removal of endothelial nitric oxide (NO) was investigated in vitro and in vivo. The vasoconstrictor potency of phenylephrine and the force of contraction of rat isolated aortic rings were significantly enhanced after endothelium removal or treatment with inhibitors of endothelial NO synthase. Furthermore, these procedures led to a significant decrease in the basal levels of cGMP in the vascular rings. Moreover, the potency of glyceryl trinitrate (n3Gro) and sodium nitroprusside (SNP) as relaxing agents and the ability of SNP to induce increases in cGMP in aortic rings were significantly enhanced in those rings denuded of endothelium or treated with the inhibitors. These procedures did not affect the vasodilator actions of isoprenaline or 8-bromo-cGMP. In the anesthetized rat, treatment with the inhibitors enhanced significantly the hypotensive responses to n3Gro without affecting those to isoprenaline. These data indicate that the removal of the basal NO-mediated vasodilator tone in the cardiovascular system leads, at the level of the soluble guanylate cyclase, to a specific supersensitivity to nitrovasodilators in vivo. The existence of such a phenomenon has important implications for understanding the local physiological control of blood flow, its pathological disturbances, and the mechanism of action of nitrovasodilators.     

Endothelium-derived Relaxing Factor Inhibits Shear Stress-induced Platelet Aggregation

The effect of endothelium-derived relaxing factor (EDRF) and related compounds on platelet aggregation in response to physiological and pathological levels of arterial wall shear stress (30-120 dyne/cm(2)) was investigated. Platelets in plasma, or washed platelets, aggregated markedly in response to shear stresses generated by a cone-plate viscometer. Pre-treatment of platelets with the S-nitrosothiol compounds S-nitroso-N-acetylcysteine or S-nitrosocysteine, or with nitric oxide (NO) or SIN-1 (which is non-enzymatically metabolized to NO), resulted in decreased platelet aggregation in response to shear stress. Non-hydrolyzable analogues of cyclic guanosine 3',5'-monophosphate (cGMP) also inhibited shear stress-induced platelet aggregation, and specific pharmacological manipulations of NO and cGMP (with methylene blue or the cGMP phosphodiesterase inhibitor M&B 22 984) resulted in alterations of intraplatelet levels of cGMP that correlated with the degree of inhibition of shear stress-induced platelet aggregation. These results demonstrate that EDRF and related compounds inhibit platelet aggregation that is initiated by shear stress, and suggest that this physiologically relevant mechanism of platelet aggregation may be regulated by intraplatelet cGMP.     

内皮素变化及心钠素对哮喘影响的实验研究

目的 研究哮喘豚鼠血浆及支气管肺泡灌洗液中（BALF）内皮素（ET－1）含量的变化及外源性心钠素（ANF）对ET－1的影响。方法 采用放射免疫分析法测定哮喘豚鼠血浆及BALF中ET－1和cGMP含量及给哮喘豚鼠输注不同剂量ANF后血浆及BALF中ET－1和cGMP含量。结果 哮喘豚鼠组血浆及BALF中ET－1显高于正常对照组（P＜0．01）。输注ANF后血浆及BALF中ET－1水平明显下降（P＜0．01），并呈剂量依赖性；cGMP水平明显升高（P＜0．01），亦呈剂量依赖性。结论 ET－1在哮喘发病中起有重要作用；ANF具有抑制ET－1合成及分泌的作用，拮抗其生物效应，可望成为治疗哮喘的一种新方法。     

Stimulation of cyclic GMP production via AT2 and B2 receptors in the pressure-overloaded aorta after banding

Abdominal aortic banding induces upregulation of the angiotensin II (Ang II) type-2 (AT2) receptor, thereby decreasing the contractile response to Ang II in the thoracic aorta of the rat. The aim of this study was to use a mouse model to clarify the mechanisms by which the banding elicits upregulation of the aortic AT2 receptor and the subsequent attenuation of Ang II responsiveness. Concomitantly with the elevation in blood pressure and plasma renin concentration after banding, AT2-receptor mRNA levels in the thoracic aorta rapidly increased in mice within 4 days. Upregulation of the AT2 receptor, as well as blood pressure elevation after banding, was abolished by losartan administration. The contractile response to Ang II was depressed in aortic rings of banding mice but not of sham mice, and was restored by either the AT2-receptor antagonist PD123319 or the bradykinin B2-receptor antagonist icatibant. cGMP content in the thoracic aorta of banding mice was 9-fold greater than that of sham mice, and the elevation was reduced to sham levels 1 hour after intravenous injection of PD123319 or icatibant. When aortic rings were incubated with Ang II, cGMP content increased in banding rings but not in sham rings; the pretreatment with PD123319 or icatibant inhibited Ang II-induced cGMP production. These results suggest that aortic banding induces upregulation of the AT2 receptor through increased circulating Ang II via the AT1 receptor, thereby activating a vasodilatory pathway in vessels through the AT2 receptor via the kinin/cGMP system.