[1] 10w Brain renin-angiotensin system in the pathophysiology of cardiovascular diseases *

[1] 162w Cardiovascular diseases (CVD) are among the main causes of death globally and in this context hypertension represents one of the key risk factors for developing a CVD. It is well established that the peripheral renin-angiotensin system (RAS) plays an important role in regulating blood pressure (BP). All components of the classic RAS can also be found in the brain but, in contrast to the peripheral RAS, how the endogenous RAS is involved in modulating cardiovascular effects in the brain is not fully understood yet. It is a complex system that may work differently in diverse areas of the brain and is linked to the peripheral system by the circumventricular organs (CVO), which do not have a blood brain barrier (BBB). In this review, we focus on the brain angiotensin peptides, their interactions with each other, and the consequences in the central nervous system (CNS) concerning cardiovascular control. Additionally, we present potential drug targets in the brain RAS for the treatment of hypertension.

[1] 94w The autonomic nervous system with its efferent sympathetic and parasympathetic arms plays a crucial role in regulating BP and has an impact on peripheral neurons in order to control muscles and other visceral targets. Cardiovascular target control is primarily noradrenergic and mainly determined by sympathetic neurons and RAS interference. It has also been suggested that central AngII is additionally involved in regulating hypertension based on the sympathetic hyperactivity observed in early stages of CVD pathology [64][65][66]. This means that an excessive sympathetic activity supports the pathogenesis of hypertension and progression of organ damage [67,68].

[2] 316w To identify the specific effects of AngII on sympathetic nervous system (SNS) activity, several studies have been performed. Besides injecting AngII into the NTS, microinjection into the SFO, OVLT, and the PVN can also elevate BP [69]. Additionally, the activation of AT1 receptors by AngII in the anterior ventrolateral medulla elevates the BP by facilitating the SNS, tachycardia, and catecholamine release from the adrenal medulla [48,70,71]. In this context, AngII is able to modulate both the presynaptic sympathetic system and the adrenal medulla, resulting in an enhanced release of norepinephrine (NE) and epinephrine by facilitating catecholamine release from peripheral sympathetic neurons and by enhancing depolarization-dependent exocytosis and via direct ganglionic excitation [72,73] (see also Fig. 2). Consequently, inhibition of the converting enzyme by ACE inhibitors or blockade of the specific AT1 receptors by AT1 receptor blocking agents should lead to a decrease in both NE and epinephrine release. The net catecholamine overflow during stimulation of the sympathetic nerves was not reduced after ACE inhibition [72], which may be attributed to the ACE-dependent inhibition of bradykinin degradation since bradykinin also increases NE release via bradykinin B2 receptors [74]. However, neuronal uptake of catecholamines was increased when rats were treated with ACE inhibitors, which occurred independently of the hypotensive efficacy of ACE inhibition and also independently of central sympathetic activity and which was suggested to contribute to the antihypertensive and cardioprotective effects of ACE [75]. In contrast to ACE inhibitors, AT1 antagonists lowered AngII-induced NE release in a dose-dependent manner [72]. Thus, the antihypertensive efficacy of angiotensin receptor blockers (ARB) depends on their ability to lower NE release from sympathetic terminals and by decreasing vascular sensitivity to NE (Fig 2). The latter effect was found to be related to an intact neuronal NE uptake via NE transporter and presynaptic alpha2-mediated autoinhibition. Furthermore, it was suggested that both mechanisms contribute to the antihypertensive and cardioprotective actions of ARBs [76,77].

[3] 164w In contrast to the previously described AngII effects, the impact of AngII under hypoxic conditions is different. Under hypoxia a bilateral blockade of the AT1 receptors in the RVLM leads to an increase in BP and renal sympathetic nerve activity (RSNA) [62,78] whereas an additional GABA receptor block in the RVLM results in a decrease in RSNA and BP [79]. Under normoxia, the blockade does not have any effects on the BP. One can conclude that the lack of oxygen leads to a tonic sympathoinhibitory influence of AngII in the RVLM, which is dependent on GABA receptors [42,79]. Hence, these results suggest that endogenous AngII balances the tonic excitatory and inhibitory effects on sympathetic premotor neurons via AT1 receptors, thus influencing the sympathetic vasomotor activity depending on the variation in physiological or pathophysiological conditions. Under hypoxic conditions, for example, the balance is shifted to prefer sympathoinhibition. This mechanism is known as the 'push-pull' hypothesis put forward by Dampney et al. (see Fig. 3) [80].

[4] 88w The previously described activation of the SNS by brain AngII leads to the release of vasopressin and inhibits the baroreceptor reflex when AngII is injected into the brain [37] (see also Fig. 5). This SNS activation and the reliable pressor response ensue via activation of AT1 receptors located in the CVOs and via PVN and SON projection to sympathetic nuclei in the brainstem [48,49]. The further projection goes through the spinal cord to the intermediate lateral nucleus and then via the sympathetic chain to terminate in blood vessels.

[5] 38w In particular, the release of vasopressin, a powerful vasoconstrictor, from the PVN and SON is stimulated via AT1 receptor activation by AngII [3]. The vasopressin levels in mice were 2fold increased when AngII was injected intracerebroventricularly (icv) [122].

[6] 349w As mentioned above, AngIV is metabolized from AngIII by APN, APB, or both, which can be inhibited by bestatin and PC18, respectively (Fig. 1) [137] [146]. Comparable to AngIII, no AngIV levels could be detected so far in the brain. However, endogenous levels are nevertheless expected to be present in brain as metabolic enzymes (e.g., APA and APN) that are involved in the conversion from AngII have been detected (Table 1). AngIV binds at the AT1 and AT4 receptors. AT4 receptors do not bind AngII, AngIII, Ang(1-7), or their related receptor antagonists (Fig. 1). Mapping studies revealed that the highest concentrations of AT4 receptors in the brain are distinct from AT1 and AT2 receptors. The AT4 receptor is thought to be linked more closely to the regulation of cognitive, sensory, and motor functions and less to physiologic functions, such as water-electrolyte balance, cardiovascular regulation, and control of thirst behaviors, where AT1 or AT2 receptors play a predominant role (reviewed by Karnik et al. [173]). In 2001, Albiston et al. clearly identified the AT4 receptor as the insulinregulated aminopeptidase (IRAP) cleaving N-terminal amino acids from several peptides (including metenkephalin, dynorphin, oxytocin, arginine-vasopressin, lysine bradykinin, neurokinin A, somatostatin, neuromedin B, and cholecystokinin). They claimed that the AngIV binding-site ligands produce their effects by binding to the active site of IRAP and thus inhibiting the catalytic activity of the enzyme [174]. The AT4 receptor subtype was also claimed to be the HGF/c-Met receptor system that plays a functional role in memory and Parkinson's disease (reviewed in [175]). Consequently, it is not fully known whether the AT4 receptor subtype plays a crucial role in cardiovascular control and regulation. It is likely that the cardiovascular effects of AngIV are only mediated via the AT1 receptor [173]. By using in vitro preparations of endothelium-denuded human saphenous vein, AngIV was demonstrated to act as a vasoconstrictor that was 2.7-fold less potent than AngII. The contractile response of AngIV was mediated via AT1 receptors but not AT2 receptors [176]. As AngII more potently activates the AT1 receptor, AngIV may play a supporting role for the AngII/AT1 signaling.

[7] 366w Various in vivo studies were performed to determine how brain AngIV influences BP regulation. In response to icv administration (10-min infusion of 100pmol/µl aCSF/min), AngIV increased BP in alert rats. However, this pressure response was only 2-fold less than the same amount of AngII or AngIII. An aminopeptidase-resistant AngIV analog produced a significantly greater pressor response than did native AngIV and the duration of the pressor effect was extended. The AngIV-induced pressor response was blocked by the AT1 antagonist losartan, but not by the AT4 antagonist divalinal, indicating the AT1 receptor-dependent mechanism [165,156]. Surprisingly, however, divalinal by itself also increased BP, raising the question as to whether divalinal is specific at AT4 receptors since both stimulation and blocking of AT4 receptors increase BP [156]. Confirming these findings, icv-administered AngIV in rats was also demonstrated to enhance BP, which was accompanied by an increase in renal blood flow and renal vascular resistance. These effects were abolished by pretreatment with centrally administered candesartan but not by the selective AT4 receptor ligands LVV-H7 and AT4-16, proving the hypothesis that AT1 rather than AT4 receptors are involved in the pressure response or renal response to icv administration of AngIV [177]. The BP-enhancing efficacy of brain AngIV was impressively supported in a transgenic mouse model that chronically releases the AngIV peptide specifically in the brain [178]. The brain levels of AngIV in these transgenic mice were only elevated 4-fold compared to the normal endogenous levels. Despite this moderate increase in brain AngIV, transgenic mice were found to be hypertensive compared to the control animals. However, treatment with the ACE inhibitor captopril did not affect BP in transgenic mice and candesartan again normalized the BP [178]. Calcium mobilization assays performed on cultured CHO cells chronically transfected with the AT1 receptor confirm that low-dose AngIV can only mobilize calcium via the AT1 receptor in the presence of AngII, consistent with an allosteric mechanism [178]. In summary, the results of Wright et al., Yang et al.,and Lochard et al. [175,177,178] demonstrate that chronic elevation of AngIV in the brain can induce hypertension via an AT1 receptor-dependent mechanism. Despite these cardiovascular effects, stable, small-molecule AngIV analogs were developed to treat Alzheimer's and Parkinson's diseases [175,179].

[8] 254w Ang(1-7) is another peptide that has come into the focus of research in the last few years [170]. Ang(1-7) can be generated from AngI or II. Neprilysin (also known as neutral endopeptidase), thimet oligopeptidase, or prolylendopeptidase cleaves the last three amino acids from AngI and ACE2 or prolylcarboxypeptidase removes the C-terminal phenylalanine from AngII, all liberating the heptapeptide Ang(1-7). ACE2 can also generate Ang(1-9) from AngI at first, followed by the action of ACE, which releases the last two amino acids. It should be mentioned that none of these enzymes is specific for the angiotensin peptides (reviewed by Bader [170]; Fig. 1). All components of the ACE2/Ang(1-7)/Mas axis are expressed in the brain [6,181,182]. In contrast to AngIII and AngIV, there is a clear evidence for the presence of significant amounts of Ang(1-7) in brain tissue (Table 1). In the hypothalamus of rats, the level of Ang(1-7) is comparable to that of AngI and AngII [180]. Many effects of Ang(1-7) are mediated by the G protein-coupled Mas receptor, which consists of seven transmembrane domains [6,170]. Indeed, the brain is actually the organ with the highest expression of Mas, in particular in the hippocampus and the piriform cortex [181], including the RVLM in which the receptor is abundantly expressed [183,184], the NTS, parvocellular, and magnocellular portions of the PVN, and other areas [184] that are related to cardiovascular control in the CNS [185][186][187]. The Ang(1-7) antagonist D-Alanine-Ang(1-7) (A-779) is able to block the Mas receptor and in consequence selectively abolish the effects of Ang(1-7) [188].

[9] 95w In the periphery the Ang(1-7)/Mas receptor signaling constitutes the counterregulatory axis of the AngII/AT1 receptor path [189]. The ACE2/Ang-(1-7)/Mas axis is established to exert protective actions in hypertension and other cardiovascular disorders via diverse peripheral mechanism, also including the stimulation of prostaglandin and NO release from endothelial cells, the augmentation of bradykinin effects via decreasing ACE activity, and the reduced growth of smooth muscle cells [170]. There is evidence that central effects are also involved in regulating cardiovascular function; however, these brain-related Ang(1-7) actions seem to be more complex and partially discrepant (reviewed by [6,190]).

[10] 277w Discrepant data on BP effects in response to Ang(1-7) have been published, depending on duration (acute or chronic) and areas (CVLM, PVN, RVLM, NTS) of Ang(1-7) administration (Table 2). In particular, Ang(1-7) was repeatedly injected into NTS, RVLM, and CVLM as these areas contribute to BP regulation by modulating baroreflex sensitivity. Thus, a decrease in BP was observed when Ang(1-7) was injected into the NTS (serving as the first brain relay for baroreceptors) and into the dorsal medulla (controlling the parasympathetic tone) of normotensive and SHR rats [192,193] (Table 2). Ang(1-7) has moreover been established to be a powerful facilitator of the bradycardic component of baroreflex control of HR in normotensive [194,195] or hypertensive animals [196][197][198]. Infusion of Ang(1-7) into the lateral ventricle facilitates the baroreceptor reflex only for the bradycardic component, in contrast to the attenuation of the baroreflex that is induced by AngII [194] and the microinjection of Ang(1-7) into the NTS that decreases the MAP in normotensive and SHR rats. Additionally, the infusion into the NTS produces a significant increase in the sensitivity of baroreceptor reflex control of HR [192]. In the brain the NTS seems to be involved in the baroreflex facilitatory effect of Ang(1-7) [192,199]. It was further suggested that bradykinin and Ang(1-7) interact to modulate baroreflex control of HR because simultaneous infusion of these peptides at subeffective rates increases baroreflex sensitivity (Fig. 2). The functional importance of Ang(1-7) for modulating baroreflex sensitivity was impressively demonstrated because the counterbalancing effect of endogenous Ang(1-7) at NTS age-dependently decreases in parallel to decreased production of the peptide by neprilysin [200]. As Ang (1)(2)(3)(4)(5)(6)(7) inhibited AngI-and AngII-mediated NE release [201][202][203] via a NO-mediated pathway [318],

[11] 81w Ang(1-7) also decreases tyrosine hydroxylase (TH) expression, the rate-limiting enzyme in catecholamine biosynthesis [319] without affecting NE uptake and catabolism [204]. It is likely that Ang(1-7) influences baroreflex sensitivity due to its ability to reduce sympathetic tone and modulate the local effects of NE in the brain (Fig. 2). However, this concept is not completely convincing since Ang(1-7) reduced NE release via NO and TH expression and was blocked by an AT2 receptor, but not by a Mas receptor antagonist [205,206].

[12] 175w The microinjection of Ang(1-7) into the CVLM also causes depressor effects similar to those caused by AngII in WKY and SHR rats [191] (see also Table 2). Those similarities can also be observed for Ang (1)(2)(3)(4)(5)(6)(7) in the PVN. In contrast to AngII, the Ang(1-7)-mediated pressure response is attenuated by L-nitro-arginine methyl ester (L-NAME) and neuronal NO synthase blockers [207]. The effects of Ang (1)(2)(3)(4)(5)(6)(7) in the PVN are as effective as those of AngII in enhancing the cardiac sympathetic afferent reflex (CSAR) and increasing sympathetic outflow, renal sympathetic nerve activity (RSNA), and MAP. Both endogenous Ang(1-7) and AngII in the PVN contribute to the enhanced CSAR and sympathetic outflow in renovascular hypertension. Ang(1-7) in the PVN even seems to potentiate the effects of AngII in renovascular hypertension as the simultaneous injection of Ang(1-7) and AngII causes greater enhancing effects on RSNA, MAP, and CSAR than one of them alone in 2K1C rats [208]. Additionally, the chronic infusion of the antagonist A-779 into the PVN prevents hypertension in a rat model of sleep apnea [209].

[13] 288w If Ang(1-7) is injected into the RVLM (see also Tab 2), it enhances the CSAR and increases the RSNA and the MAP in normotensive rats [210]; moreover, it enhances the sympathetic activation in renovascular hypertension [184] as AngII in the RVLM does [211]. In normotensive rats, Ang(1-7) in the RVLM increases femoral blood flow while AngII increases renal blood flow, without changing vascular resistance. In hypertensive rats, both Ang (1)(2)(3)(4)(5)(6)(7) and AngII increase vascular resistance of the renal and mesenteric beds and reduce blood flow in the renal artery [31]. On the other hand, the microinjection of Ang(1-7) close to the RVLM has no significant effects on MAP and RSNA, which conclusively clarifies the specificity of Ang (1)(2)(3)(4)(5)(6)(7) in the RVLM and confirms that it is not a general phenomenon in the brain and also excludes the possibility that the effects of Ang(1-7) were caused by leaking into the surrounding tissue of the RVLM [211]. Furthermore, the microinjection of A-779 into the RVLM reduced the RSNA and MAP and inhibited the CSAR in normal rats [186]. The Ang(1-7)mediated effects in the RVLM are in part modulated by superoxide anions [212]. The binding of Ang(1-7) to the Mas receptor in the RVLM significantly increases NAD(P)H oxidase activity and consequently elevates the levels of superoxide anions, which in turn are responsible for the increase in sympathetic outflow. In support of these findings, the antagonist A-779 decreases the NAD(P)H oxidase activity and the superoxide anion level in the RVLM. The inhibition of NAD(P)H oxidase abolished the increasing effect of Ang(1-7) at the superoxide anion level, which leads to the hypothesis that NAD(P)H oxidase is the major source of the superoxide anions that modulate the effects of Ang(1-7) in the RVLM [212].

[14] 144w The counterregulatory role of Ang(1-7) to AngII (as outlined above) becomes clear when Ang(1-7) is infused icv for 4 weeks. After chronic infusion, the expression of AngII and the AT1 receptor in the brain of SHR and DOCA-salt rats is significantly reduced [213,214] and the MAP is lowered; furthermore, the baroreflex control of arterial pressure and the cardiac autonomic tone are normalized in DOCA-salt rats [213] (see also Table 2). In support of this, an elevated AngII/Ang(1-7) ratio in the medulla of transgenic (mRen2)27 hypertensive rats [215] was shown to be connected by an impaired baroreflex function [216], and the icv infusion of Ang(1-7) in these hypertensive rats reduced BP, normalized baroreceptor control of HR, restored cardiac autonomic balance, reduced cardiac hypertrophy, and decreased imbalance of the AngII/Ang(1-7) ratio in the heart [217]. Interestingly, in Mas receptor knockout mice baroreceptor sensitivity is reduced [218].

[15] 119w In response to chronic administration of 100-200 ng/h Ang(1-7) into the lateral ventricle by mini-pumps, an anti-hypertensive efficacy was observed [213,217,219] (Table 2). In contrast, others convincingly demonstrated that transgenic mice overexpressing brain ACE2 indeed had higher Ang(1-7) brain levels but the BP remained normal [220][221][222]. Moreover, viral gene transfer of ACE2 into the NTS of normotensive WKY rats did not affect BP while viraldriven ACE2 overexpression in SHRs induced a decrease in BP [223][224][225] (see also Table 2). This clearly indicates that the efficacy of brain Ang(1-7) to lower BP depends on whether animals are hypertensive or normotensive. It might also be argued that icv injections were not performed regularly, thus explaining almost negative results upon Ang(1-7) treatment.

[16] 128w Additional effects of brain Ang(1-7) thus beneficially affecting cardiovascular and cerebrovascular diseases may include modulation of the circadian rhythms of MAP and HR [226] and increases in cerebral blood flow [227], bradykinin levels [228], NO release, and eNOS expression [229]. Despite some controversial effects of brain Ang(1-7) on cardiovascular function, the majority of the experimental studies based on acute Ang(1-7) microinjections into specific brain areas and the experimental approaches based on chronic brain Ang(1-7) administration and overexpressing brain ACE2 indicate that high brain Ang(1-7) may have antihypertensive efficacy, particularly when the BP is increased (Table 2). Thus, it was suggested that pharmacological strategies leading to activation of the Ang(1-7)/Mas receptor axis, also in the brain, should be developed in order to improve the treatment of hypertension-triggered CVD [217].

[17] 47w Within in the past few years, further angiotensin peptides have been identified but their overall role in the brain is not fully understood. Indeed, endogenous brain levels have not yet been determined, particularly for most of the non-classical RAS components (see also Fig. 1 and Table 2).

[18] 234w The most recently described novel AOGEN-derived peptide, playing a central role in cardiovascular control, is angiotensin (1-12) (Ang(1-12)) Using the RIA technique, Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) was shown to be abundant in a wide range of organs and tissues, including brain [234] . In This peptide was also termed proangiotensin-12 based upon its peripheral vasoconstrictor properties, suggesting a role as a precursor of AngII [234]. Both ACE and chymase are involved in this metabolism [235] and Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) is processed to Ang(1-7) via neutral endopeptidase (NEP) (Fig. 1). In the periphery, Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) was reported to elicit pressor responses after iv administration and these actions may be related to a rapid conversion to AngII and an AT1 receptor-dependent mechanism [234]. In addition to its presence in various peripheral tissues, Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) was also shown to be in rat brain within the NTS. The concentration of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) is about five times greater than that of AngII [234,236] (see Table 2), perhaps indicating a more important role than that of AngII in the CNS. When central Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) was immunoneutralized, BP was decreased and baroreflex sensitivity and HR variability were improved in hypertensive (mRen2)27 rats, indicating a putative role of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) in regulating BP and autonomic function in hypertension [237]. To further identify the sites of actions, Sapru and his group [238][239][240][241] have intensively studied the cardiovascular actions of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) in response to brain-specific microinjection.

[19] 255w Microinjections of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) into the NTS decreased MAP, HR and SNA [236,242]. Similar observations were made when Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) was injected into the RVLM pressor area, hypothalamic PVN and arcuate nuclei [238][239][240]. These findings suggest that a) the CVLM depressor area has an important role as a relay nucleus in medullary baroreflex pathways; b) second-order NTS neurons project to CVLM neurons by using an excitatory amino acid (probably glutamate) as a neurotransmitter; c) AT1 receptors are present in the CVLM, and microinjections of AngII into the CVLM decrease BP; d) microinjections of Ang(1-7) into the CVLM elicit depressor responses; and e) microinjections of AngII and Ang(1-7) into the CVLM modulate baroreflex sensitivity, the Sapara group recently demonstrated [241] that microinjections of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) into the CVLM of WKY rats elicited decreases in MAP, HR, and greater splanchnic nerve activity, and that initial and delayed phases of these responses are mediated via AngII and Ang(1-7) and mediated via AT1 receptors and Mas, respectively. Since blockade of GABA receptors in the RVLM attenuated cardiovascular responses provoked by microinjections of Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) into the ipsilateral CVLM, it was also concluded that GABA receptors in the RVLM partly mediate the cardiovascular responses to Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12). Although cardiovascular effects have been demonstrated in response to icv injections, it seems doubtful that the described effects can be attributed to Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) itself as no specific receptor has been suggested to mediate Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12) effects. Instead, it seems more likely that effects are related to metabolism into AngII or other peptides.

[20] 242w The second nonclassical peptide that likely plays a role in consideration of brain-related cardiovascular effects is angiotensin (3-7) (Ang(3-7)). Ang(3-7) can be derived from AngII via AngIV. AngIV is converted to Ang(3-7) by carboxypeptidase P (CPP) and prolyl oligopeptidase (POP) cleavage of the Pro-Phe bond. It further can be derived from Ang(1-7) (Fig. 1). Its concentration in the brain has not been measured so far (Table 1). Thus, information regarding its biological brain-related function is only based on experiments after organ-specific application. While iv injections of Ang(3-7) alone did not alter BP and HR in rats, it potentiated bradykinin-induced hypotensive effects [243]. Considering brain-related cardiovascular effects, Ang (3)(4)(5)(6)(7) was demonstrated to be an additional peptide of the RAS by acting as a neuromodulator, at least in the RVLM, since Ang(3-7) microinjections into the RVLM increased BP, similar to Ang(1-7) and AngII [271]. However, the Ang(3-7)-induced pressor effects were not modified by losartan or A779, whereas the Ang(3-7) analog d-Ala(7)-Ang-(3-7) completely abolished them, indicating that the effects are not mediated via Mas or AT1 receptors [271]. Since Ang(3-7) also has a high affinity at AT4 receptors [244] and AngIV also has pressor effects (see above), the AT4 receptor may be involved. However, as outlined above the AngIV-induced hypertension was mainly related to an AT1 receptordependent mechanism. Thus, the exact mechanism including the receptor interaction of Ang(3-7) effects in the RVLM and its influence in other parts of the CNS remain unclear for now.

[21] 43w In addition to being present in many different tissues such as arteries, thymus, heart, lung, skeletal muscle, white and brown adipose tissue, and several others, another new peptide of the RAS, called alamandine, was found in the brain, particularly in the cerebellum [245,246].

[22] 151w Alamandine is formed by the hydrolysis of AngA by ACE2 [247,248] or from the Ang (1)(2)(3)(4)(5)(6)(7) peptide by decarboxylation of its aspartate residue. Alamandine is a heptapeptide whose amino acid sequence only differs from that of Ang(1-7) in having a N-terminal alanine instead of an aspartate residue (Fig. 1). So far, alamandine has been detected in rats, mice, and humans and acts via the Mas-related G protein-coupled receptor (MrgD) [195,248]. As alamandine stimulated NO release in MrgD-transfected cells, but not in Mas-transfected cells, it was concluded that alamandine is a natural ligand for MrgD receptor [249,250]. In the brain, alamandine induces a vasopressor effect after microinjection into the RVLM of rats, whereas a vasodepressor effect is triggered in the CVLM [249]. However, these brain-related effects of alamandine are not defined in the presence of MrgD antagonists. In addition, an icv infusion of alamandine potentiates the bradycardic component of the baroreflex [250].

[23] 95w Additional observations demonstrated increased alamandine plasma levels in patients with a renal disease [249]. The bioavailability of acetylcholine is increased in the presence of alamandine and it promotes acetylcholine-mediated vasodilatation in the thoracic aorta and iliac artery. Consequently, alamandine seems to antagonize the AngA-mediated effects, leading to a negative feedback loop [251] that may be conceivable in the brain, too, but for now further details about alamandine and its receptor or AngA are not known. However, these factors might play an important role in central cardiovascular regulation as the other components of the RAS do.

[24] 201w For the other nonclassical RAS peptides that are known to be present in the periphery, no data are available yet to show central effects of angiotensin A (AngA), angiotensin (2-10) (Ang(2-10)), or angiotensin (1-5) (Ang(1-5)). AngA, an AngII-derived octapeptide, is formed by the decarboxylation of aspartic acid to alanine, which is the only difference in the amino acid sequence in comparison to AngII (Fig. 1) [251,252] but its physiological concentration in human plasma is lower than 20% that of AngII [251]. Due to the fact that AngA interacts with AT1 und AT2 in a comparable way and that its peripheral effects are similar to those of AngII, it is very well possible that AngA has effects in the CNS similar to those of AngII considering vasoconstriction, but so far, it remains unclear whether AngA has direct brain-related effects, particularly concerning cardiovascular function. The same applies to Ang(2-10); a group showed that Ang(2-10) can be converted into AngIII [253] and AngIII is the most active form of brain angiotensin with tonic stimulatory effects on BP (see above), but its brain-related relevance for cardiovascular function seems unclear and has not been investigated further up to now as it has been for Ang (1)(2)(3)(4)(5).

[1] 430w In the past few years, knowledge about the brain RAS and particularly the different angiotensins and related receptors for regulating cardiovascular function has expanded dramatically. Considering that the conversion of brain AngII to AngIII can be blocked by APA inhibitors followed by a decrease in BP, the development of APA inhibitors seems to be a promising new strategy to develop new, pharmacologically available alternatives for treating cardiovascular dysfunction, in particular hypertension and post-MI. As mentioned above, the initial results of a phase 2 clinical trial on QGC001 indicate that systolic BP is lowered in hypertensive patients without revealing any great safety concerns. At ESH conference in Milan (June 2017), first data were presented showing that QGC001 led to a decrease in blood pressure. The higher was the baseline blood pressure, the greater was this antihypertensive effect [171]. In particular, hypertensive patients with the specific LRHV hormonal profile were claimed to benefit from brain APA blockade to prevent the conversion from AngII to AngIII, provided that it is true that that brain AngIII is more effective in controlling BP than AngII (for detailed discussion, see section above). Considering these positive effects, Quantum Genomics intends to continue QGC001 development by conducting the NEW HOPE trial in 25 centers in the United States on 250 high cardiovascular risk patients with results being expected in the first half of 2019 [171]. It nevertheless remains largely uncertain as to whether, on the one hand, such a screening program really takes place in clinical routine before initiating QGC001 treatment and whether, on the other hand, the resulting increase in brain AngII itself --particularly in patients with normal/high renin levels --induces detrimental effects such as cognitive dysfunction ( [254][255][256] and also the review of JM Saavedra in this issue) or other brain disorders (e.g., anxiety and depression) [257]. However, it was recently found that AngII and AngIII levels alike were significantly higher in postmortem brain tissue of patients with Alzheimer's disease (AD) than in age-matched controls and that AngIII, rather than AngII, was strongly associated with amyloid-β load and tau load and that levels of APA were significantly reduced in AD. This indicates that the APA/AngIII pathway is dysregulated and that elevated AngIII may contribute to the pathogenesis of AD [258]. Hence, even if the full results of QGC001 Phase IIa trial (to be presented during the next congress of the European Society for Hypertension, Milan, in June 2017) disclose the best data regarding efficacy and safety, it remains to be seen in long-term clinical trails whether cognitive dysfunction is altered in patients treated with APA inhibitors.

[2] 64w It also seems feasible that an increase in brain AngII could also impair glucose metabolism, particularly in individuals suffering from metabolic syndrome as plasma insulin and glucose have been demonstrated in rats to be markedly decreased when AngII was chronically injected into the brain [129] whereas others have found that insulin was indeed decreased while glucose was increased after long-term AngII icv injections [131].

[3] 462w A second strategy for new antihypertensive drugs may be to develop stable Ang(1-7) analogs acting at Mas and also penetrating the BBB that can improve the treatment of hypertensiontriggered CVD. The development of AVE0991, a nonpeptide Ang(1-7) receptor Mas agonist that is orally active, indeed represents an important step for exploring the effects of Ang-(1-7) and testing its potential as a cardiovascular drug [259,260]. However, it must be investigated in further studies whether AVE0991 might cross the BBB, thus contributing to antihypertensive effects. The question of whether activation of reactive oxygen species, which may increase the sympathetic outflow (as seen after single Ang(1-7) microinjections specifically into the RVLM), harbors clinical risk has not been answered yet and needs to be addressed in further studies. Two further Mas agonists, CGEN-856 and CGEN-857, were identified by using a computational discovery platform and were claimed to have a therapeutic value because they induce vasorelaxing, antihypertensive, and cardioprotective effects [260]; however both CGEN-856 and CGEN-857 are peptides and it seems rather unlikely that such structures can be administered orally and cross the intestinal barrier to enter the brain. Figure 2: Opposing influence of brain AngII and Ang(1-7) on blood pressure via regulating neuronal firing, norepinephrine (NE) release, and sympathetic outflow, thereby affecting baroreflex sensitivity. AngI is cleaved by ACE into AngII and further by ACE2 into Ang(1-7). Ang(1-7) is also produced by NEP and PEP from AngI. In the central nervous system, AngII increases neuronal firing, NE release, and sympathetic outflow via AT1 receptors. Baroreflex sensitivity is diminished upon AngII stimulation by increasing eNOS activity and enhancing NO release, which further promotes the release of the inhibitory transmitter GABA. AngII effects can be blocked by sartans. When ACE is blocked by ACE inhibitors neuronal firing, NE release, and sympathetic outflow are diminished via a bradykinin B2-dependent pathway. Ang(1-7) decreases neuronal firing, NE release, and sympathetic outflow upon Mas activation followed by increased NO activity and NO release. Ang(1-7) also decreases ACE activity, thus increasing bradykinin and leading to B2 receptor stimulation.  stimulation; inhibition,  increase;  decrease. hypoxia leads to sympathoinhibition. AngII, angiotensin II; BP, blood pressure; RSNA, renal sympathetic nerve activity; RVLM, rostral ventrolateral medulla Figure 4: Interaction between AngII and the hypothalamo-pituitary-adrenal (HPA) axis. AT1 and AT2 receptors are expressed in all organs of the HPA axis. In response to AngII and/or stress the synthesis and secretion of corticotropin-releasing hormone (CRH), adrenocorticotropic hormone (ACTH), and glucocorticoids are increased. These reactions may further be enhanced by hypertension and obesity. HPA axis activation was suggested to contribute to an impaired glucose control as well as to an increased palatable food intake to compensate for stress;  increase. dose of saline, AngII, or AngIII in the presence or absence of EC33 or EC27. meanSEM Fig. 5

[4] 39w Table 2: The effect of brain Ang(1-7) on blood pressure (BP) in dependency of duration of administration (acute or chronic), area of application (CVLM, RVLM, PVN, icv), and experimental animal model (normotensive or hypertensive rats, transgenic rats overexpressing ACE2)

[1] 205w Normotensive Wistar rats  acute Ang(1-7) in NTS [192] Hypertensive SHR rats  acute Ang(1-7) in NTS [192] Normotensive SD rats  acute Ang(1-7) in NTS [193] Normotensive SD rats  acute Ang(1-7) in dorsal medulla [193] Normotensive Wistar rats  acute Ang(1-7) in CVLM [207] Normotensive SD rats  acute Ang(1-7) in CVLM [208] Hypertensive 2K1C SD rats  acute Ang(1-7) in CVLM [208] Normotensive SD rats  acute Ang(1-7) in RVLM [210] Normotensive Wistar rats  acute icv A779 [198] Hypertensive SHR rats  acute icv A779 [198] DOCA-salt hypertension rats  chronic icv Ang(1-7) [213] Hypertensive (mRen2)27 rats  chronic icv Ang(1-7) [217] Hypertensive (mRen2)27 rats  chronic icv Ang(1-7) [219] Normotensive SD rats  chronic icv Ang(1-7) [131] Normotensive TgR/AsAOGEN) rats  chronic icv Ang(1-7) [131] Hypertensive sleep apnea SD rats  chronic A779 in PVN [209] Normotensive hACE2 mice  brain ACE2 [222] Hypertensive DOCA hACE2 mice  brain ACE2 [222] Normotensive Syn-hACE2 mice  brain ACE2 [220] Normotensive Syn-hACE2 mice  brain ACE2 [221] SHR-ACE2 gene transfer  brain ACE2 [223] WKY-ACE2 gene transfer  brain ACE2 [223] SHR-ACE2 gene transfer  brain ACE2 [224] WKY-ACE2 gene transfer  brain ACE2 [224]  reduction;  no change;  increase

[1] 152w Identifying pathophysiological pathways that underlie the development of cardiovascular diseases (CVD) has been an important research topic for many years due to the fact that CVD increasingly has become a major problem. Indeed, 26% of the adult population suffers from hypertension -the chronic elevation of the 24h average blood pressure (BP) -which constitutes the main risk factor for developing a CVD [1,2]. According to the World Health Organization (WHO), CVD constituted the number one cause of mortality at the global level and accounted for 30% of all global deaths. This is expected to increase further from 17.3 million to 23.6 million by 2030 (reviewed in [3]). As BP is controlled by vascular resistance and cardiac output, both of which are regulated by the autonomic nervous system and by the renin-angiotensin system (RAS), it seems reasonable to define these mechanisms in detail so as to develop potential drug targets for treating hypertension [4].

[2] 47w For several decades, it has been well established that, in addition to the peripheral RAS, a local system exists in the brain, synthesizing all components of the classic RAS [5][6][7], although it has been difficult to detect appreciable levels of renin in the central nervous system (CNS).

[3] 365w Based on the findings of Ganten [8], the groups of Lindpaintner [9] and Sigmund [10] confirmed that secreted prorenin and nonsecreted renin are existent in human and rodent brains and that overexpression of these two compounds leads to hypertension (reviewed in [6]). However, van Thiel et al. recently concluded from their results that the renin observed in the brain is plasma-derived renin and that, consequently, angiotensin II (AngII) in the brain is taken up from the blood rather than synthesized in the brain itself, as renin is the key enzyme for generating angiotensins (see Fig. 1). The levels of renin in the brain correlated with the plasma levels and the brain/plasma ratio decreased dramatically after treatment with the angiotensin type 1 (AT1) receptor antagonist olmesartan [11]. On the other hand, they suggested that if angiotensins were formed in the brain, a nonrenin proenzyme would be responsible for cleavage of angiotensin I (AngI) out of angiotensinogen (AOGEN), whereas the absence of AngI in brain tissue outside the compartment of blood led them to presume that AngI is very likely not locally synthesized in the brain. Other groups have proposed for years that intracellular renin derived from an alternative transcript is responsible for the generation of angiotensin in the brain (reviewed in [5]) and for the capability to increase BP [10]. Finally, multiple authors argued that prorenin rather than renin generates angiotensins in the brain [12][13][14]. However, van Thiel et al. could not detect prorenin levels in the brain that would have supported local generation in the brain [11]. Therefore, a local brain RAS becomes uncertain because it would mean that most of the components occurring in the brain derive from the periphery although angiotensin peptides are not able to cross the blood brain barrier (BBB) [6]. Nonetheless, it could be shown that the two systems are not completely isolated from each other. The circumventricular organs (CVO), which do not have a BBB but instead possess fenestrated capillaries, provide a complex interface between the circulating RAS components and the endogenous brain RAS [7,15,16] and form a link between kidney and brain [17,18]. Additionally, several groups showed that BBB permeability changes under pathophysiological conditions such as hypertension [19][20][21],

[4] 235w with the consequence that angiotensins can pass the BBB. On the one hand, earlier studies have shown that AOGEN, the precursor of all angiotensin peptides, is mainly synthesized by glial cells [22] and some research groups have shown that neurons are able to produce the peptide, too [23,24]. On the other hand, however, van Thiel did not detect AOGEN in the astrocytes of rat and mouse brain [11]. The discussion as to whether all RAS components of the brain are synthesized de novo (by the activity of renin or not) is ongoing and to resolve this issue, more investigations need to be conducted. In the following, we will review all that is known about the particular RAS components in the brain so far and their central effects considering cardiovascular control because many groups have shown that components of RAS, being present in the brain, have central effects. This raises the question, therefore, whether the components are synthesized in the brain or delivered from the periphery. BP, heart rate (HR), and left ventricular pressure in transgenic rats with a brain-specific AOGEN deficiency [TGR (AsAOGEN)] were reported to be lower than in Sprague Dawley (SD) rats [25][26][27][28]. Moreover, the development of sympathetic hyperactivity and cardiac dysfunction was dampened [29] and the sympathetic tonus, at least to the renal nerve, and vagal tonus to the heart were changed [28] in rats with a marked deficiency in brain AOGEN.

[5] 77w Indeed, the increase in BP and cardiac hypertrophy in response to AngII administered iv was also diminished in TGR(ASrAOGEN) compared to SD controls [30]. All these findings suggest that brain RAS has a substantial impact on cardiovascular (dys)function without exactly attributing them to a specific angiotensin (e.g., AngII, AngIII, AngIV, Ang (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12), Ang (2)(3)(4)(5)(6)(7)(8)(9)(10) Ang (1)(2)(3)(4)(5) or Ang (3)(4)(5)(6)(7), Ang (1)(2)(3)(4)(5)(6)(7), AngA or Alamandine), which will be discussed in detail in the following (see also Fig. 1).

[6] 78w Central RAS modulation influences the major vasomotor center in the brainstem, the rostral ventrolateral medulla (RVLM), containing sympathetic premotor neurons that are responsible for generating and maintaining vasomotor tone and resting levels of arterial BP [31,32]. A few studies have shown that changes in the firing rate of RVLM neurons are involved in the development and maintenance of hypertension [33][34][35]. In spontaneously hypertensive rats (SHR), the RVLM shows increased neuronal excitability [33] and overexpression of AT1 receptors [36,37].

[7] 220w The first component of the brain RAS that should be mentioned is AngII. It is generated from the AOGEN peptide by renin and angiotensin-converting enzyme (ACE) activity [38] (see also Fig, 1). Substantial amounts of ACE, AngI, and AngII have been detected in the brain tissue (Table 1). AngII is responsible for regulating fluid balance, sodium intake, and thirst, and, consequently, for the maintenance of BP [39]. It increases water and sodium intake and BP, alters the modulation of baroreceptor reflex, increases vasopressin secretion, and alters the secretion of reproductive hormones as a result of behavioral and physiological changes [7,40]. Due to the fact that ACE is an exoenzyme, AngII is taken in via neurons after extracellular formation from AOGEN. The other option of this process would be internal formation of AngII by a non-ACE mechanism of cathepsin G or tonin [41] (Fig 1). The relative expression of AngII is primarily high in the nuclei of the hypothalamus, such as in the paraventricular (PVN) and the supraoptic (SON) nuclei, the nucleus of the solitary tract (NTS) in the brainstem, the amygdala, and the preoptic area [7]. The NTS as the site of afferent baroreceptor termination in the medulla and the PVN have a significant effect on sympathetic output indirectly through modulation of the neuronal activity of the RVLM area [42].

[8] 157w With the help of highly selective nonpeptide AngII receptor antagonists such as losartan (AT1) and PD123177 (AT2), it has been possible to distinguish between two different AngII receptor subtypes, termed AT1 and AT2 receptors, respectively [43,44]. All AngII effects are mediated by these two receptors [45,46]. High concentrations of angiotensin receptors are located in regions regulating thirst and body fluids such as the subfornical organs (SFO) and the area postrema [39,47]. There is a high density of AT1 receptors in the PVN and in the SON [48,49] and the NTS. They are possibly synthesized in perikaryons located elsewhere and are axonally transported to the sites where AngII actions are mediated in the brain [1]. In contrast, AT2 receptors are mainly located in the thalamus and the cerebellum [50,51], where they are highly abundant. The role of AT2 receptors in cardiovascular and renal diseases and non-CV functions is extensively covered in the review of T. Unger, this issue.

[9] 222w Several studies in SHRs indicate that the brain RAS is more active in these animals than in the normotensive controls. In addition, the brain RAS is different in that the number of AngII receptors in the brain of SHRs is higher than in Wistar Kyoto (WKY) rats [52,53], especially in the SFO and the vascular organ of lamina terminalis (OVLT) [53]. Additionally, SHRs are more sensitive to acute AngII administration into the RVLM [54][55][56], and the group around Ganten found a BP reduction after the direct infusion of saralasin, a partial AngII antagonist, into the brain ventricles [57]. A central lowering of BP could be elicited by the antagonist but only in SHRs, not in the periphery and not in normotensive controls. Different studies in which converting enzyme inhibitors, including captopril [58,59], enalapril [60], and ramiprilat, were injected in SHRs have also shown a reduction in BP, which is usually elevated by sodium intake in these animals [61]. Moreover, a selective blockade of the AT1 receptor in SHRs decreases BP and sympathetic nerve activity (SNA) [62,63]. Therefore, blocking the brain RAS may be advantageous as it simultaneously decreases sympathetic tone and, consequently, vascular resistance. Moreover, blocking the brain RAS blocks angiotensin-induced baroreflex inhibition and decreases vasopressin release, which in turn reduces blood volume [1]. These mechanisms will be outlined in detail below.

[10] 40w The AngII-dependent activation of the hypothalamopituitary-adrenal (HPA) axis may also contribute to the pathogenesis of hypertension (Fig. 4). Both AT1 and AT2 receptors are expressed in all organs of the HPA axis [81][82][83][84][85] and are regulated by restraint stress [86][87][88][89].

[11] 484w Moreover, AngII influences the synthesis and secretion of corticotropin-releasing hormone (CRH), adrenocorticotropic hormone (ACTH), and glucocorticoids [90][91][92]. Immobilization activates hypothalamic CRH gene expression via AT1 receptors [91], and hypothalamic AT1A receptor mRNA was reduced after chronic AT1 receptor blockade [93], thus indicating that at least stress-dependent hypertension may partially be attributed to a HPA axis-dependent mechanism. Hence, the AngII-induced release of ACTH and corticosterone was higher in SHR than in normotensive WKY rats, which was further attributed to higher expression levels of AT1A receptors in the pituitary gland and lower levels of AT1B receptors in pituitary and adrenal glands [94]. A similar correlation between hypertension and enhanced stress sensitivity of the HPA axis has also been demonstrated in patients suffering from hypertension [95][96][97][98] or CVD [99]. Ten years ago, we demonstrated in SHRs that the CRHinduced ACTH and corticosterone release is reduced after treatment with an ARB or ACE inhibitor [100]. The ineffectiveness of a calcium blocker on the HPA axis reactivity indicates, on the one hand, that enhanced stress sensitivity does not develop as a phenomenon secondary to hypertension and, on the other, that stress sensitivity is not normalized by general antihypertensive strategies. Accordingly, attenuation of HPA axis reactivity by RAS inhibitors is based on mechanisms independent of BP regulation [100]. Thus, the suppression of HPA axis reactivity must be considered for its possible contribution to the antihypertensive actions of RAS inhibitors. Nevertheless, baseline levels of ACTH and corticosterone were admittedly not affected by therapy with ARB or ACE inhibitors, suggesting that an attenuation of HPA axis reactivity is not an essential mechanism for antihypertensive activity. Despite these findings, however, it should be conceded that there are conflicting data which do not strengthen the notion of a brain-related mechanism as ACTH responses to either shaking or immobilization stress in rats are not changed by central AT1 receptor antagonists or ACE inhibitors [91,101,102] and the stress sensitivity increased in transgenic rats with low brain AOGEN [26]. In this context, it should briefly be mentioned that the HPA axis activation was also suggested to have deteriorating effects on two additional key symptoms of the metabolic syndrome, thus provoking CVD. Firstly, AngII-dependent activation of HPA axis also contributes to an impaired glucose control and this effect was enlarged in obesity [103,104]. Vice versa, the alleviation of stress reactions after chronic ARB treatment contributes to the hypoglycemic actions as well [105]. Secondly, considering the findings that glucocorticoids have an important effect on food intake and that palatable food intake is increased to compensate for stress [105][106][107], we recently observed that stress-induced increase of tasty diet was lowered by ARB treatment, which may indicate that a decreased HPA reactivity induces anti-obese effects [105]. Although Kintscher et al. [108] also observed weight loss in patients after irbesartan treatment, it is still unclear whether food intake was also reduced in humans and whether the HPA axis-related mechanism is indeed involved.

[12] 375w Another parameter that must be considered in the context of AngII effects is baroreceptor sensitivity (Fig. 2). Barosensitive efferent activity is presumed to be the most crucial parameter for long-term BP regulation and, under physiological conditions, the reflex prevents a rise in BP by modulating the dynamics of the blood vessel wall. The reflex is regulated by central regions such as RVLM, NTS, and the hypothalamus [4]. In particular, the NTS plays an important role in regulating the reflex because the first synapse of the baroreceptor afferent of the CNS is located in the NTS [109]. Several studies showed a reduction in the cardiac component of the baroreceptor response after intra-parenchymal AngII injection into the NTS of rats [45,[110][111][112]. Inhibiting the reflex by direct AngII injection into the NTS of the brainstem and the action of AngII on the first synapse of the vagus in the NTS causes an increase in BP without a compensatory decrease in HR [113]. This mechanism might occur naturally in SHRs because these rats have dampened baroreceptor reflexes and higher levels of AngII and its corresponding receptors in the NTS. In support of these findings, the cardiac baroreflex was also increased by injecting the AngII antagonist [Sar1,Thr8]-AngII into the NTS in anesthetized rats [114], suggesting the presence of an angiotensinergic 'tone' in this nucleus and, hence, a physiological homeostatic role. Another study could show that the AngII in NTS-associated depression of baroreceptor reflex gain is sensitive to nitric oxide synthase (NOS) blockade by the microinjection of an AngII antagonist into the NTS [115]. The depressant action of AngII in the NTS on the baroreceptor reflex is mediated by activation of endothelial NOS (eNOS), releasing nitric oxide (NO) that promotes release of the inhibitory transmitter GABA [116]. A later study supports the hypothesis that AngII can trigger the release of NO from the blood vessel wall in the NTS via a vascularneuronal signaling mechanism [117] and when the eNOS activity was genetically blocked in NTS, the effect of locally administered AngII in NTS was missing [118]. The level of eNOS mRNA in the NTS of SHRs is relatively higher than in WKY rats [119] and a chronic inhibition of the eNOS activity in NTS results in an enhanced baroreceptor reflex [118].

[13] 108w The consequences of an overproduction of brain RAS components or the corresponding receptors by a genetic defect have been illustrated in transgenic mice in which some additional copies of the AOGEN gene were inserted. The resulting excessive amount of angiotensin peptides induced a chronic activation of the mechanism described above and caused hypertension in these animals, whereas AOGEN levels were not elevated in the plasma [120]. On the other hand, mice with brain AOGEN gene knockout have lower BP [7]. Furthermore, in transgenic rats possessing reduced AOGEN levels in the brain, it could be shown that a reduction in brain AOGEN enhanced the level of AT1 receptors [121].

[14] 102w Considering that obesity is a pandemic phenomenon, at least in western civilization, obesity is a risk for cardiovascular dysfunction, and considering that AngII and AOGEN positively correlate with body mass [123][124][125], the question arises as to whether the RAS plays an indirect role in aggravating cardiovascular function and, if so, whether it is mediated via peripheral and/or brain related pathways. Since "RAS and weight regulation" is a separate topic and only indirectly contributes to CVD, we only briefly outline some specific aspects of brain AngII in this review without going into detail concerning weight-regulating effects after treatment with ARBs and ACE inhibitors.

[15] 132w Chronic iv administration of AngII induces weight loss [126] by decreasing food intake [127] and stimulating SNA, lipolysis, and thermogenesis [128], consequently suggesting that brainrelated effects are important. Indeed, in response to icv AngII injections, food intake and weight gain increased whereas whole-body heat production and oxygen consumption were decreased [129]. Accordingly, food intake and body mass increased and energy expenditure decreased when animals with a specific deletion in the AT1A receptor in the PVN were fed with a high-fat diet [130], likely indicating that the elevated levels of AngII accompanying obesity may serve as a negative feedback signal that activates PVN neurons to alleviate weight gain. In confirmation of this hypothesis, TGR(ASrAOGEN) rats were reported to be protected against developing obesity even if they were fed with a high-calorie diet [131].

[16] 96w Admittedly, however, these effects are not clearly attributable to brain AngII since other angiotensin metabolites such as angiotensin (1-7) (Ang(1-7)) also reveal anti-adipose potency [132]. Our group recently demonstrated that chronic icv administration of Ang (1)(2)(3)(4)(5)(6)(7) was ineffective on growth, weight regulation, and energy balance in SD and TGR(ASrAOGEN) rats while AngII following chronic icv injections clearly revealed anti-obese and growthdiminishing efficacy in SD and TGR(ASrAOGEN) rats, respectively. These results show that brain Ang(1-7) has minor effects on weight regulation while brain AngII is a key contributor for regulating energy homeostasis and weight in obesity [131].

[17] 333w The heptapeptide angiotensin III (AngIII) was previously shown to be one of the main effector peptides of the brain RAS [133]. Although AngII is well accepted to be the main active form of central angiotensin peptides of the brain RAS, several studies suggested that AngIII is the more active effector in terms of BP control, AT1-mediated vasopressin release, and changes in drinking behavior [134][135][136][137][138][139][140][141]. This hypothesis is supported by several studies in which the selective blockade of brain AngIII formation results in a decrease in BP [134] [ 136,138,142]. Additionally, the iontophoretic administration into the hypothalamic PVN or SFO neurons has shown that AngIII is more potent than AngII [143,144]. Furthermore, the icv injection of anti-serum developed by Song et al., which has an anti-catalytic activity against glutamyl aminopeptidase A (APA), significantly reduced the pressor and drinking responses induced by icv infusion of AngII but failed to influence AngIII-induced drinking and pressor responses [141]. Another hint that AngIII is the more active peptide in the brain compared to AngII is that AngIII microiontophoresis induces a higher firing rate than AngII in two brain structures involved in BP regulation: the SFO [144] and the PVN [143,145]. The mean arterial pressure (MAP) is similarly increased in response to icv-infused AngII, AngIII, and the metabolically resistant analogs d-Asp(1)AngII and d-Arg(1)AngIII in alert, freely moving rats [138], indicating that pressure responses of AngIII are comparable to those of AngII. It has also been suggested that the brain AT1 receptor may favorably bind AngIII instead of AngII in mediating BP, but it was also shown that AngII and AngIII have approximately the same affinity to AT1 [146,147] and AT2 receptors [148,149], although the AT2 receptor appears to be maximally sensitive to AngIII [150]. The discussion of whether AngII and/or AngIII is the better endogenous ligand at the receptors continues. It was also claimed that brain-related effects of AngIII on cardiovascular function may be related to central SNS stimulation since AngIII greatly decreased neuronal and non-neuronal NE uptake [151].

[18] 87w In contrast, however, other studies do not support the AngIII hypothesis claiming that conversion of AngII to AngIII is necessary for brain-related AngII effects as a) similar pressor and dipsogenic effects as well as an impact on salt appetite have also been found when AngII and AngIII are injected into the cerebral ventricles, the NTS, or the area postrema (AP) of hypertensive or normotensive rats [146,[152][153][154]; and b) pressor and dipsogenic effects of aminopeptidase-resistant AngII analogs administered icv were similar or even greater than AngII itself [155,156].

[19] 360w AngIII is generated from AngII when brain APA cleaves the N-terminal aspartic residue from AngII [150]. ACE can also convert AngI directly to AngIII. AngIII is metabolized to angiotensin IV (AngIV) by the aminopeptidase N (APN) as well as by aminopeptidase B (APB) (Fig. 1) [137,146]. Although we cannot find any published reports on the detectable concentration of AngIII in the brain, it is very likely to be present as APA was detected in a cerebral tissue (see also Tab. 1). APA is a homodimeric, membrane-bound glycoprotein; activity is highest in the hypothalamic nuclei and in the medulla [157,158] but it is also active in other parts of the brain such as the CVO, PVN, SON, and NTS [159]. His 389 [160] and Glu 386 [161] from the zincbinding motif H 385 EXXH 389 as well as Tyr 471 [162] are involved in the catalytic process of APA by which the transition state complex which is formed during catalysis is stabilized by creating a hydrogen bond between its hydroxy group and the oxyanion of the tetrahedral intermediate [162]. Another important residue is Glu 352 , which contributes to APA exopeptidase activity via interaction with the N-terminal part of the substrate [163]. Ca 2+ elevates APA activity and is suggested to be involved in the binding of the side chain carboxylate of the N-terminal acidic amino acid of the substrate [164]. This conversion of brain AngII to AngIII can be blocked by APA inhibitors. A brain-specific blockade of APA lowered the levels of endogenous AngIII in the brain, thus inducing a decrease in BP. In this regard, it was demonstrated more than 20 years ago that the Ang II-induced pressor response was reduced after icv administration of the APA inhibitor amastatin [165]. Amastatin is a peptidic, naturally occurring, competitive, and reversible APA inhibitor that was isolated from Streptomyces sp. In contrast, AngII-induced pressor response was increased after pretreatment with the aminopeptidase B inhibitor bestatin [165] and the brain levels of AngIII and BP also increased when brain APN was blocked by PC18 [136]. Hence, APA was thought to be a potential target for the treatment of different forms of hypertension [159].

[20] 194w EC33 is a nonpeptidic APA-blocking substance, inhibiting brain APA activity in a dosedependent manner with an IC50 value of 12 µg when injected icv [134]. Confirming the amastatin findings, EC33 administered icv not only increased the half-life of AngII and completely blocked the formation of AngIII [159], it also abolished the AngII-induced pressor response [166]. However, EC33 did not change the pressor effect induced by AngIII itself [136]. These findings indicate that AngII must be converted to AngIII in order to increase the arterial BP [167]. As mentioned above, AngII is established to release vasopressin via a brainrelated mechanism, thus contributing to vasoconstriction. However, the central, AngII- half-lives increase in the hypothalamus when APN is blocked by EC27 (Fig. 5A); and c) the increase in plasma vasopressin levels upon icv AngII administration was completely abolished when conversion into AngIII via APA was blocked by EC33 while (AngIIIstimulated) vasopressin levels were markedly enhanced when AngIII degradation was blocked by the APN inhibitor EC27 [145] (see also Fig. 5C). Moreover, the AngIII-stimulated vasopressin release is attributed to an AT1 receptor-dependent mechanism because the EC27-induced increase in plasma vasopressin was hampered in the presence of saralasin [145].

[21] 329w Considering, on the one hand, that vasopressin contributes to high BP and, on the other, that vasopressin release is mainly related to AngIII, hypertensive patients with low renin and high vasopressin (LRHV profile) concentrations were claimed to benefit from the concept of blocking brain APA to lower BP. Thus, a clinically available APA inhibitor must be ingestible and able to cross the BBB. RB150 (renamed as QGC001 by Quantum Genomics) fulfills these requirements [134]. RB150 is a prodrug and consists of two EC33 molecules linked by a disulfide bridge that is cleaved by reductases only after entering the brain [159]. In vitro studies on recombinant purified mouse APA showed that the inhibitory potency of the reduced form of RB150 obtained in the presence of dithiothreitol (DTT), which facilitates the generation of EC33 from RB150, was similar to that of EC33, while in the absence of DTT, RB150 with an intact disulfide bridge was inactive on APA [134]. The brain-dependent metabolism of RB150 into EC33 minimizes the risk that APA inhibitors will induce Ang II accumulation in the periphery. In vivo, RB150 is cleaved by reductases after entering the brain. RB150 was demonstrated in different models of high BP to be highly effective when orally administered and crossed the intestinal, hepatic, and hemophilic barriers to enter the brain [159]. EC33 and RB150 selectively inhibit APA as these compounds have no affinity toward other zinc metalloproteases such as APN, ACE, ACE2, ECE1, or NEP and do not block AT1 and AT2 or endothelin A and B receptors, which are all known to be involved in BP regulation [142]. In hypertensive DOCA-salt rats [134,142] and SHRs [168], the formation of brain AngIII decreased and BP was lowered for several hours. The decrease in BP in SHR was partly attributed to a decrease in sympathetic tone, reducing vascular resistance. This treatment did not modify systemic RAS activity. Concomitant oral administration of RB150 with the ACE inhibitor enalapril potentiated the RB150-induced BP decrease [168].

[22] 181w Considering the findings, too, that enzymatic characteristics of APA do not differ between humans and rodents and that the presence of APA in several human brain nuclei is sensitive to angiotensins and involved in BP regulation [158], RB150 was consequently evaluated in a phase I clinical trial in healthy and normotensive volunteers. The results showed an appropriate gastrointestinal absorption and metabolism into EC33 after 1.25 mg QGC001, no influence on the systemic RAS or on systolic or diastolic BP or HR, as well as drug tolerance [169]. A randomized, double-blind, placebo-controlled phase 2 clinical trial was performed and completed in April 2016 to evaluate the efficacy and safety of QGC001 oral doses in patients with grade I or II essential hypertension [170]. Quantum Genomics reported that "data showed positive signals over several parameters of the study, particularly on the primary endpoint, the drop in daytime systolic BP measured as ambulatory pressure in hypertensive patients" without giving details. Quantum Genomics also announced that "the full results will be disclosed during the congress of the European Society for Hypertension in June 2017 [171].

[23] 170w Considering the findings that the brain RAS contributes to sympathetic hyperactivity and left ventricle (LV) remodeling and dysfunction after myocardial infarction (MI), Huang et al. investigated whether AngIII is also involved in these effects [172]. Four weeks after inducing MI in rats by coronary artery ligation, they demonstrated that brain APA activity was increased, sympathoexcitatory and pressor responses to air stress enhanced, and arterial baroreflex function impaired; furthermore, left ventricular end-diastolic pressure (LVEDP) was increased and ejection fraction (EF) and dP/dt(max) were decreased. When animals were treated with RB150 for 4 weeks post-MI, brain APA activity and responses to stress and baroreflex function were normalized and LVEDP, EF, and dP/dt(max) were improved while losartan was less effective without altering brain APA activity. These results indicate, on the one hand, that brain APA and AngIII appear to play a crucial role in the sympathetic hyperactivity and LV dysfunction in rats post-MI and, on the other, that blocking APA by RB150, for example, may represent a potential candidate for CNS-targeted therapy post-MI.

[24] 64w Here, we would like to mention that Ang(1-7) was also demonstrated to have antiobese efficacy as transgenic rats overexpressing Ang(1-7) showed the following characteristics: 1. reduced body weight following chow feeding; 2. no development of obesity following cafeteria diet (CD) feeding; 3. lower energy intake; 4. intact leptin sensitivity despite CD feeding; 5. enlarged energy expenditure; and 6. no insulin resistance despite CD feeding.

[25] 133w Considering the findings, Ang(1-7) increases in response to chronic treatment with ARB [230,231]. Additionally, we further found that the antiobese efficacy of ARB can at least be partially attributed to an Ang(1-7)/Mas-dependent mechanism because weight loss after chronic telmisartan administration in SD rats was antagonized after a high, but not a low dose of the Mas receptor antagonist A779 [132]. Since transgenic rats with an AOGEN deficiency did not develop CD-induced obesity either [232], we further investigated whether the antiobese effects of Ang(1-7) may be related to a brain-dependent mechanism. However, after icv administration of Ang(1-7), weight-reducing effects could only be observed when high doses were administered whereas AngII elicited convincing and well-known effects already at lower doses, suggesting that the brain-related mechanisms of Ang(1-7) for antiobese effects may rather be minor [233].