The alternative renin-angiotensin system

Abstract

The renin-angiotensin system is the most important peptide hormone system in the regulation of cardiovascular homeostasis. Its “classical” arm consists of the enzymes, renin and angiotensin-converting enzyme, generating angiotensin II from angiotensinogen which activates its AT₁ receptor thereby increasing blood pressure, retaining salt and water, and inducing cardiovascular hypertrophy and fibrosis. However, angiotensin II can also activate a second receptor, the AT₂ receptor. Moreover, the removal of the C-terminal phenylalanine from angiotensin II by angiotensin-converting enzyme 2 yields angiotensin-(1-7) and this peptide interacts with its receptor Mas. When the aminoterminal Asp of angiotensin-(1-7) is decarboxylated, alamandine is generated which activates the Mas-related G-protein coupled receptor D, MrgD. Since Mas, MrgD, and the AT₂ receptor have opposing effects to the “classical” AT₁ receptor, they and the enzymes and peptides activating them are called the “alternative” or “protective” arm of the renin-angiotensin system. This review will cover the historical aspects and the current standing of this recent addition to the biology of the renin-angiotensin system.

Introduction

In ancient Rome, the two-faced mythological Janus God symbolized the duality of the principles of presiding over all beginnings and transitions, whether abstract or concrete¹. The bipolar concepts ascribed to this ancient God apply to understanding how a stable equilibrium between interdependent positive and negative physiological processes maintains homeostasis. Irvine Page, in the “Mosaic Theory”,² posited that essential hypertension resulted from the disturbance of the equilibrium among the mechanisms participating in the control of tissue perfusion. This idea has proven helpful in dissecting the mechanisms of hypertension. However, three decades would elapse before it was applied to unraveling the biochemical vocabulary of the alternative arm of the renin-angiotensin system (RAS) counteracting angiotensin II (Ang II) pathological actions on hydromineral balance, tissue perfusion, and cell growth and differentiation.

The RAS is key contributor in regulating of cardiovascular homeostasis. Its “classical” arm consists of the enzyme renin released by the kidney and acting in the blood stream on the liver-derived angiotensinogen to generate angiotensin (Ang) I (Figure 1). This inactive decapeptide is shortened by the last two amino acids through angiotensin-converting enzyme (ACE) mostly expressed on endothelial cells to yield the active peptide Ang II, which by acting on its AT₁ receptor (AT₁R) increases blood pressure and induces cardiovascular hypertrophy and fibrosis. Based on its pivotal actions in blood pressure control and end-organ damage, inhibitors of the “classical” RAS such as renin and ACE inhibitors as well as AT₁R antagonists are corner stones in cardiovascular medicine. From the discovery of renin in 1898³ until about 35 years ago the described enzymes, peptides and receptors were the only known components of the RAS. However, in 1989 a second receptor for Ang II, the AT₂ receptor (AT₂R), was discovered with opposing effects and distinct pharmacology compared to AT₁R^4-6. At about the same time, the heptapeptide Ang-(1-7) lacking the C-terminal phenylalanine of Ang II, was first described to have physiological effects, also opposing the ones of its precursor, Ang II^7,8. For a long time, the main enzyme generating Ang-(1-7) from Ang II and its receptor remained unknown until the discovery of ACE2 and Mas, in 2000^9,10 and 2003¹¹, respectively. However, this was not the final discovery of novel RAS components. In 2013, Lautner et al. described the physiological actions of the peptide alamandine, which is Ang-(1-7) with an alanine instead of aspartic acid at position 1¹². While its receptor, Mas-related G-protein coupled receptor type D (MrgD), was also already described in this publication, the enzyme substituting aspartic acid 1 to alanine 1 to yield alamandine and also Ang-A from Ang II¹³, remains somewhat enigmatic. Since Mas, MrgD and the AT₂R have opposing effects to the receptor for the classical RAS components, AT₁R, they and the enzymes and peptides activating them are called the “alternative” or “protective” RAS. This review will cover the historical aspects and the current standing of this recent addition to the biology of the RAS.

Figure 1.

Schematic outline of the renin-angiotensin system (RAS) with alternative, protective axis boxed in pink. Abbreviations are: AD, aspartate decarboxylase; NEP, neprilysin; PRCP, prolylcarboxypeptidase; PREP, prolylendopeptidase. Other abbreviations as defined in text.

AT₂ receptor

From the mid 1980ies, evidence started to accumulate that more than one receptor exists for Ang II. These first, still preliminary indications were based on differences in the inactivation of ¹²⁵I-Ang II radioligand binding by dithiothreitol, which breaks down disulphide bonds¹⁴. A first breakthrough in the identification of AT₂Rs occurred in 1989, when several pharmaceutical companies in their attempt to develop angiotensin receptor antagonists (ARBs) for the treatment of hypertension “accidentally” synthetized angiotensin receptor ligands which were selective for either the AT₁R (Ex89 or DuP753/losartan) or the AT₂R (CGP42112A, PD123319, PD123177)^4,5. Based on this selectivity, a Nomenclature Committee of the AHA Council for High Blood Pressure Research agreed on a nomenclature for the newly discovered receptor subtypes. AT₁Rs were defined as those receptors where the binding of Ang II could be displaced by DUP753/Ex89, and AT₂Rs when Ang II binding could be displaced by CGP42112A, PD123319, or PD123177¹⁵. The final proof for the existence of angiotensin receptor subtypes was provided in 1991 and 1993 by cloning of the AT₁R and the AT₂R by the Victor Dzau and Tadashi Inagami laboratories^16,17. They found that both receptors belong to the family of class A G-protein–coupled receptors (GPCRs) but share only 34% sequence homology. Both groups were also the first to create AT₂R knockout (KO)mice^18,19, and through the phenotypic description of these mice established the receptor as a biologically active component of the RAS. A main functional consequence of AT₂R knockout in both studies was a stronger increase in blood pressure in response to Ang II infusion compared to wildtype mice, which at that time (1995), was still totally unexpected, since the effect was the opposite of what was known as an Ang II effect – which was a blood pressure increase. The authors of these two seminal papers concluded correctly that the AT₂R apparently mediates effects opposing those of the AT₁R. A few other publications around the same time observed AT₂R effects, opposing the known Ang II/AT₁R effects as well as an increase in phosphotyrosine phosphatase activity²⁰, an increase in delayed-rectifier K1 current (IKv) in neurons²¹, and anti-proliferative effects in vascular endothelial cells in vitro²² and (after AT₂R transfection) in vivo²³. It is nowadays generally accepted that AT₂R actions are in opposition of AT₁R actions – sometimes (but not always) by direct interference with AT₁R signalling. Nevertheless, AT₂R actions are not necessarily dependent on concomitant AT₁R activation (to exert a counteracting effect), but AT₂R can also mediate effects “on its own” as will be discussed below.

Since AT₂R effects are fundamentally different from effects of the AT₁R, it is logical to assume that the receptors themselves are fundamentally different as well regarding their structure and the initiation of cell signalling. This assumption was in fact already made by the authors of both studies reporting the cloning of the AT₂R^17,20 with Mukoyama et al. concluding that this receptor “may belong to a unique class of seven-transmembrane receptors”. These studies also found already that the AT₂R does not couple to G-proteins in a conventional, “AT₁R-like” way (with activation of the Gq/Ca²⁺/PKC or the Gia/oa/adenylyl cyclase inhibition pathways). The recent crystallization of the AT₂R in complex with 6 different ligands confirmed that the AT₂R is a 7-transmembrane domain receptor which shares important motifs with conventional GPCRs (Figure 2A). Nevertheless, although data are available for 7 AT₂R crystals in complex with 6 different ligands²⁴, these studies probably raised as many questions as they answered. For example, it is still not known whether an inactive state of the receptor exists or whether it is always in the conformation that corresponds to an active or active-like state – even if AT₂R antagonists bind to and block the receptor. In all crystal structures published up to now, the receptor was found in the active-like state. This may either mean that the AT₂R is indeed always in an active-like state regardless of the type of ligand bound (which is for example the case for the β₂-adrenergic receptor²⁵), which would fit to functional studies reporting constitutive activity of the receptor²⁶, or all of the ligands that have been included into crystal structures of the receptor so far, have (at least partial) agonistic activity. The latter is a reasonable consideration, because the ligands used for the first published crystal structure, Compounds 1 and 2, are undefined regarding their intrinsic activity at the receptor^6,27, and EMA401, which is part of another crystal structure, was recently identified to be a partial agonist²⁸. Another unsolved question related to the AT₂R is the role of helix 8, which in the crystal with Ang II was found in the canonical conformation, i.e. in a position parallel to the intracellular surface of the cell membrane (Figure 2B)²⁴. However, in crystals with the undefined ligands Compounds 1 and 2 and with EMA401, helix 8 flipped towards helices TM5 and TM6, which is assumed to hinder β-arrestin and G-protein binding to the 3^rd intracellular loop (Figure 2C). While it is an interesting question, whether helix 8 in the non-canonical position prevents canonical G-protein signalling of the AT₂R, it is perhaps not so very critical, since canonical signalling of the AT₂R is already prevented by the lack of the consensus motifs required for the binding of G-proteins and β-arrestin within the 3^rd intracellular loop¹⁷. However, evidence is strong that the AT₂R can bind and signal through Gi-proteins, which may bind to the 2^nd intracellular loop⁶ and the activation of protein phosphatases but not in the canonical inhibition of adenylyl cyclase. Moreover, a specific AT₂R interaction protein (ATIP)²⁹ binds to the C-terminus of the AT₂R and initiates signalling cascades leading to anti-proliferation and apoptosis through PPARγ activation or tyrosine kinase inhibition²⁹.

Figure 2. Crystal structure of the AT₂-receptor.

(A) Crystal structure of the AT₂-receptor (blue) bound to the endogenous agonist angiotensin II (red). (B) Magnification of the intracellular portion of the AT₂-receptor bound to angiotensin II with the receptor in an active-like conformation and helix 8 in the canonical position, i.e. parallel to the intracellular surface of the cell membrane. (C) Magnification of the intracellular portion of the AT₂-receptor bound to the uncharacterised ligand Compound 1 with the receptor in an active-like conformation and helix 8 in a non-canonical position, i.e. shifted “inwards” and covering the binding sites for G-proteins and β-arrestin on the 3rd intracellular loop (3^rd ICL).

Generally, biological effects of the AT₂R are more apparent in disease situations, in which the receptor usually has protective effects. The rather few known physiological actions include natriuresis/diuresis, blood pressure (BP) lowering effects, vasodilation (which in normotension does not lead to a BP lowering effect) and some effects in the intestine such as activation of sodium/glucose cotransporter-1 (SGLT1)⁶. Typically, in diseased or injured tissue, expression of the receptor is increased, which is likely one explanation why AT₂R effects are more apparent in pathologies. Another reason is the often-inhibitory nature of its actions such as anti-inflammatory, anti-fibrotic, anti-proliferative or anti-apoptotic effects, which are inapparent in a health situation. Preclinical disease models, in which therapeutic effects of AT₂R stimulation have been shown to include models of hypertension, myocardial infarction, heart failure, vascular remodelling and fibrosis, atherosclerosis, vascular inflammation, aortic aneurysm, preeclampsia, chronic kidney disease, diabetic and ischemic nephropathy, focal segmental glomerulosclerosis, cyclosporine nephropathy, hypertensive nephropathy, sickle cell nephropathy, Dupuytren’s disease, various cancers including melanoma, pulmonary hypertension, acute lung injury, stroke, spinal cord injury, neuromyelitis optica, cerebral malaria, multiple sclerosis, diabetic neuropathy, obesity, insulin resistance, etc⁶.

The only area, in which the effect mediated by AT₂Rs is heavily discussed, is (neuropathic) pain. Several studies support either an analgesic effect of AT₂R stimulation or an analgesic effect of AT₂R blockade, respectively³⁰. Moreover, EMA401, which was reported to have analgesic effects due to AT₂R blockade, turned out to be a partial AT₂R-agonist, it is now hard to decide whether the therapeutic effect was due to AT₂R stimulation or blockade²⁸.

The preclinical findings with AT₂R agonists and antagonists have initiated a number of drug development programs, of which the most advanced currently is a multi-centre, open-label, single-arm Phase II clinical trial with the AT₂R agonist C21 in patients with idiopathic pulmonary fibrosis (IPF) (ClinicalTrials.gov ID NCT04533022). Interim analysis of this trial indicated that treatment with C21 may be able to halt progression of IPF [American Journal of Respiratory and Critical Care Medicine 2023;207:A2531 (Abstract)], however the final report of the trial, which is expected for 03/2024, will have to be awaited for any final conclusion on the therapeutic efficacy of C21 in IPF. C21 was previously successfully tested in a Phase II clinical trial in patients with moderate to severe courses of COVID-19³¹, but a subsequent Phase III trial did not show statistically significant treatment effects, probably due to the low number of severe, deteriorating cases in the later stages of the pandemic (ClinicalTrials.gov Identifier: NCT04880642). EMA401 (Olodanrigan) was successfully tested in a Phase II clinical trial in patients with post-herpetic, neuropathic pain³². However, its further development was terminated because of drug-specific (not class-specific) preclinical liver toxicity. Moreover, in a follow-up Phase II trial, a statistically significant treatment effect in patients with post-herpetic neuralgia could not be confirmed³² (ClinicalTrials.gov Identifier: NCT 03094195). Nevertheless, several research groups and companies (e.g., Confo Therapeutics and Eli Lilly) continue to work on the design and synthesis of optimised AT₂R antagonists. It will be interesting to see whether these molecules will be superior to EMA401 with regard to their analgesic effect, or whether the loss of agonistic property will also lead to the loss of therapeutic efficacy.

Generally, AT₂R-targeting drugs are quite far advanced in various drug development projects, and the next few years will decide whether any of them will make it to clinical approval.

Angiotensin-(1-7)

Following the synthesis of Ang II, great efforts were directed to unravel its mechanism of action and the importance of the peptide aromatic groups in position 4, 6, and 8 for agonistic activity. The demonstration that phenylalanine (Phe) was essential for Ang II-mediated myotropic and adrenal catecholamine agonistic responses suggested that the free carboxyl group in position 8 was an obligatory requisite for binding to Ang II receptors³³. However, this assumption was wrong. A study performed in Ferrario’s laboratory revealed the generation of the Ang-(1-7) metabolite in canine brainstem homogenates incubated with Ang I³⁴, and the first demonstration of its biological activity stimulating vasopressin release from the rat hypothalami-neurohypophysial system⁸. A decade of persistent and independent investigation of Ang-(1-7) biological actions in Ferrario’s^35,36 and Santos³⁷ laboratories yielded general recognition after identification of an ACE homolog -ACE2- functioning as a monocarboxypeptidase^9,10. ACE2 cleaves Ang-(1-9) from Ang I and Ang-(1-7) from Ang II. The knowledge gained in revealing Ang-(1-7) biological actions and the concomitant identification of Mas as the receptor at which Ang-(1-7) binds¹¹ led to the recognition of a counter-regulatory mechanism in which Ang II and Ang-(1-7) act to oppose each of their known biological actions as the effectors of the opposing ACE/Ang II/AT₁R and ACE2/Ang-(1-7)/Mas axes of the RAS, respectively. Although this terminology has gained general acceptance, this reading of the RAS biochemical physiology ignores the critical role of other peptidases in cleaving Ang-(1-7) from Ang I³⁸.

The collective work on Ang-(1-7) suggests that Ang-(1-7) functions primarily as a tissue rather than a circulating hormone, as illustrated by its presence in the interstitial fluid and tissues at concentrations several orders of magnitude higher than in plasma, as discovered not only by antibody-based assays^36,39, which may suffer from limited specificity⁴⁰, but also by LC-MS/MS⁴¹. Biologically active tissue levels of peptides are notoriously underestimated due to the very short half-life of peptides and the fact that only interstitial fluid levels are relevant for membrane-bound receptor interaction, which are diluted when concentrations are measured in whole tissues. There is a high diversity of Ang-(1-7)-forming endopeptidases present in cardiovascular, gastrointestinal, and reproductive organs. Tissue endopeptidases cleaving Ang-(1-7) from Ang I include neprilysin (NEP), prolylendopeptidase (PREP), and thimet oligopeptidase (Figure 1). Besides ACE2, also PREP and prolylcarboxypeptidase (PRCP), can generate Ang-(1-7) from Ang II^37,38. The differential compartmentalization of Ang II and Ang-(1-7) forming membrane-bound and intracellular enzymes may determine which pathway is activated through the metabolism of Ang I, Ang II, or both. This conclusion is in keeping with the demonstration of Ang-(1-7) production by a metallopeptidase contrasting with Ang II biogenesis by ACE and chymase in atrial tissue homogenates⁴².

The space assigned to this article excludes a comprehensive discussion of Ang-(1-7) biological actions. Excellent reviews of Ang-(1-7) contribution to cardiovascular and renal function^43-47, antithrombotic mechanisms, amelioration of obesity-induced oxidative stress, antitumoral mechanism, hydromineral and carbohydrate metabolism, and reproductive biology are published^36,37,48,49. Most of Ang-(1-7) actions are mediated by the GPCR Mas¹¹.

The relevance of Ang-(1-7) contribution to the pathogenesis and evolution of human diseases remains in its infancy partly because of the absence of well-controlled clinical trial outcomes⁵⁰. The attractiveness of exploring whether a deficit in circulating or tissue-born Ang-(1-7) synthesis or activity contributes to the pathogenesis of primary hypertension and heart disease was first unveiled in Ferrario’s laboratory^51-53. The characterization of Ang-(1-7) in human plasma and urine provided a tool to address whether changes in Ang-(1-7) may serve as a biomarker of reduced activity of the heptapeptide in diseases of the heart and the blood vessels. The demonstration of a negative correlation between diastolic blood pressure and plasma Ang-(1-7) concentrations in primary hypertensive patients medicated with captopril evidenced a vasodepressor role of te hheptapeptide⁵³. Pursuing this hypothesis, it was further shown that 24h Ang-(1-7) urinary excretion was significantly lower in newly diagnosed untreated primary hypertensive patients⁵¹. Treatment with a dual inhibitor of neprilysin and ACE (omapatrilat) controlled the blood pressure of salt-sensitive hypertensive patients by a mechanism involving increased urinary Ang-(1-7) excretion⁵². These studies established a foundation for exploring whether an intrinsic downregulation of the RAS counterregulatory axis is a culprit in the pathogenesis of primary hypertension. A further rigorous examination of this hypothesis remains to be evaluated. A growing literature demonstrates higher plasma Ang-(1-7) in the blood of normal women compared to men⁵⁴ and an association between lower Ang-(1-7)/creatinine ratios and high blood pressure in preterm-born young adults⁵⁵.

As reviewed elsewhere^56,57, interventional studies in human subjects confirm vasodilator actions of Ang-(1-7) in the coronary, splanchnic, and renal circulations. On the other hand, translation of these observational studies to clinical therapy remains elusive even though new orally active Ang-(1-7) formulations⁵⁸ and Mas agonists⁵⁹ are available. Interest in exploring the therapeutic benefit of pharmacological activation of the ACE2/Ang-(1-7)/Mas axis is reflected in the registration of 372 clinical trials when searching clinicaltrials.gov using the term -angiotensin-(1-7)-. As illustrated in Figure 3, 191 clinical trials (52%) are recorded as completed, while 50 others are categorized as recruiting. Of the total number of currently registered clinical trials, 175 are classified as addressing diseases of the heart, cerebral circulation, heart failure, and hypertension (Figure 3, bottom panel), while 50 others are classified as addressing kidney disease and type 2 diabetes. Seventy-three other trials reflect the interest in evaluating the therapeutic effectiveness of augmenting the activity of the ACE2/Ang-(1-7)/Mas axis in the treatment of SARS-CoV-2 infection^60-62. The outcome of one of these trials using a novel Mas agonist in COVID-19 patients was recently reported and showed therapeutic efficiency of the drug⁵⁹.

Figure 3. Clinical trials addressing the alternative RAS.

We interrogated ClinicalTrials.gov (www.clinicaltrials.gov) using the term -angiotensin-(1-7)- to obtain detailed trial information about ongoing and completed trials. As documented in the top panel, the largest number of recorded trials are categorized as completed. A second group of trials are classified as recruiting, while 27 others are marked as terminated. The pie in the bottom portion of the Figure shows that the highest number of clinical trials are classified as exploring Ang-(1-7) mechanisms in hypertension (n=99), followed by COVID-19 (n=73), cardiovascular disease (n=39), kidney disease (n= 30), heart failure (n=26), and diabetes (n=20).

G protein-coupled receptor MAS

In 1986, a new gene was isolated from DNA of a human epidermoid carcinoma cell line, identified as a proto-oncogene, and named MAS⁶³. The name MAS is an abbreviation of the last name of the person who donated the human tumor from which the MAS gene was derived⁶⁴. Initially, the function of the Mas protein was unknown, and it was only in the early 2000s that it was identified as the GPCR through which Ang-(1–7) signals¹¹. The possible relationship between Mas and the RAS was first suggested by Jackson et al.⁶⁵ who expressed MAS in Xenopus oocytes and in a mammalian cell line. Ang I, II and III induced an inward current in oocytes and in transfected cells, Ang II and III led to intracellular Ca²⁺ release and initiation of DNA synthesis⁶⁵. Based on these observations Mas was suggested to be a functional Ang II receptor. Several follow-up studies supported this hypothesis.⁶⁴ However, Ambroz et al.⁶⁶ showed that the Ca²⁺ release after Ang II treatment was only observed in MAS-transfected cells additionally expressing endogenous Ang II receptors. Cloning of the real Ang II receptor, AT₁R, in 1991^67,68 and the later discovery of a direct interaction between Mas and AT₁R⁶⁹ partly explained the original observations of Jackson et al.⁶⁵ and indicated that MAS is not an Ang II receptor per se but modulates AT₁R signaling^37,70.

In 1994, a novel angiotensin antagonist was described, the heptapeptide, D-Ala⁷-Ang-(1-7)⁷¹. This compound named A-779 showed a high degree of selectivity for Ang-(1-7). The results obtained with A-779 were the first strong evidence for the existence of an Ang-(1-7) receptor³⁷. However, only a decade later Mas was identified as the receptor for Ang-(1-7) based on the absence of ¹²⁵I-Ang-(1-7) binding to kidney slices of mice with genetic deletion of Mas¹¹. The Ang-(1-7)-induced vasorelaxation of thoracic aortic rings and the anti-diuretic effect of the peptide in water-loaded mice were also abolished in Mas-Knockout (KO) animals¹¹. Many other effects of Ang-(1-7) were also absent in Mas-KO animals or A-779 treated rats and mice³⁷.

The selectivity of Mas for Ang-(1-7) was hard to prove by the use of radiolabeled Ang-(1-7) which resulted in high unspecific binding even in renal tissue from Mas-KO mice¹¹, probably due to the fact that iodination of Ang-(1-7) alters the binding properties of the heptapeptide⁷². Binding of fluorescent Ang-(1-7) has been demonstrated in several preparations including testis slices and MAS-transfected CHO cells^73,74. Moreover, coupling of Ang-(1-7) to its receptor causes inhibition of MAP kinase or cyclooxygenase-2 (COX-2)-dependent pathways and stimulates nitric oxide (NO)/cGMP-dependent pathways^75,76.

Infusion of Ang-(1-7) or antagonists as well as gain-of-function and loss-of-function models (transgenic animals and genetic deletion of Mas) allowed to establish the relevance of Ang-(1-7) and/or Mas in physiological processes. As reported for Ang II this heptapeptide has pleiotropic effects in the body, including NO-mediated vasorelaxation, baroreflex modulation, cardioprotection, and beneficial metabolic effects³⁷. In most instances the effects of Ang-(1-7) negatively modulates the effects of Ang II³⁷. Most if not all effects of Ang-(1-7) are absent in Mas-KO mice and in many instances, genetic deletion of Mas produces alterations opposed to those produced by Ang-(1-7) administration³⁷. A role for MrgD in partially mediating the Ang-(1-7)-induced production of cAMP in vitro has been suggested, but Ang-(1-7) was only half as potent in MRGD-transfected cells than in MAS-transfected cells⁷⁶. Whether this effect by MrgD contributes to other actions of Ang-(1-7) directly or indirectly (via alamandine, see below) remains to be determined. A direct role of AT₂R in the effects of Ang-(1-7) has been suggested based on the blockade by the putative AT₂R antagonist PD123319⁷³. However, this compound can also block MrgD even on tissues of AT₂R KO mice or Ang-(1-7) effects in MAS-transfected cells^12,76. In addition the affinity of Ang-(1-7) for AT₂R is very low compared to Ang II (~500-fold less)⁷⁷. On the other hand, the possibility that Mas/AT₂R heterodimers can contribute for Ang-(1-7) effects at least in some cell types can not be disregarded⁷⁸. Concerning AT₁R, the fact that Mas can antagonize AT₁R⁷⁹ may explain why Ang-(1-7) can also act as a biased agonist of AT₁R^80,81. A role for the interaction of Mas with the endothelin B or bradykinin B2 receptors in the Ang-(1-7) effects has been suggested in some circumstances^37,64.

Stimulation of MAS-transfected cells increased NO production^73,74, NOS/AKT phosphorylation ^37,82, arachidonic acid products release¹¹, intracellular Ca²⁺ upregulation⁸³, and cyclic AMP^76,84. However, it has been reported that Ang-(1-7) has no effect on MAS-transfected cells but negatively modulates AT₁R signaling⁸⁵. The NOS/NO pathway was not tested in this particular study. An atypical signaling of peptide ligands (NPFF and angiotensin fragments including Ang-(1-7)) through Mas (G-protein-independent signaling) was reported in a study focusing Ca²⁺ influx in MAS-transfected HEK293 cells⁸⁶. Future studies are needed to clarify the results of these in vitro studies which were not supported by other findings using other cell types^87-89 or even similar cells^76,90. In spite of these discrepancies, most studies support a direct activation of Mas by Ang-(1-7), but the distinct response of a specific cell on Ang-(1-7) stimulation may depend on its expression of additional peptide receptors besides Mas⁶⁴.

Alamandine

In 2007, Jankowski and collaborators reported the identification of a novel angiotensin peptide, formed by decarboxylation of the aspartate residue of angiotensin II. They named this peptide, angiotensin A¹³. Ang A acts through AT₁R and AT₂R producing effects similar to those of Ang II. Taking in account the possibility that ACE2 could cleave Ang A, forming a des-Phe⁸ heptapeptide, Santos et al.¹² incubated Ang A with recombinant ACE2. Indeed, the heptapeptide Ala¹-Ang-(1-7) was formed. It was named alamandine. This heptapeptide can also be formed directly from Ang-(1-7), in the rat coronary circulation ¹². Differently from Ang A, the receptor for alamandine was not similar to the one of the parent peptide. Alamandine produces many effects similar to those of Ang-(1-7) but acts mainly by stimulating MrgD¹². This receptor is also stimulated by the amino acid β-alanine with equal potency and the relative importance of the two ligands for its activation may depend on their concentrations in tissues^91,92. In addition, differing from Ang-(1-7)/Mas, alamandine signaling appears to involve mainly AMPK, instead of AKT, at least in cardiomyocytes⁹¹. Accordingly, the effects of alamandine are usually blocked by the angiotensin antagonist D-Pro⁷-Ang-(1-7) but not by A-779¹².

Are all the actions of alamandine and Ang-(1-7) similar? This appears to be not the case, since in the insular cortex microinjection of alamandine but not of Ang-(1-7), altered renal nerve activity and increased blood pressure and heart rate⁹³. Whether similar selectivity is present in other brain regions or in other organs remains to be elucidated.

The pathophysiological relevance of alamandine/MrgD deficiency was suggested by the occurrence of dilated cardiomyopathy in mice with genetic deletion of MrgD⁹⁴. Important alterations in the expression of genes related to sugar and fat metabolism were also reported in MrgD-KO mice⁹⁵. It should be mentioned that further studies are necessary to reveal the aspartate-decarboxylating enzyme(s) involved in the formation of Ang A and alamandine in the body. In this respect, Jha et al.⁹⁶ recently postulated a role of bacterial β-decarboxylases in the formation of alamandine in the gut but these enzymes cannot explain the presence of the peptide in most other parts of the body.

Angiotensin Converting Enzyme 2

In 2020, ACE2 got sad notoriety as uptake receptor for the SARS-CoV-2 virus causing the COVID-19 pandemic⁹⁷. However, its history began in 2000 with its independent discovery by two groups, both reporting that it is the long-sought enzyme degrading Ang II into Ang-(1-7)^9,10.

However, the zinc-dependent monocarboxypeptidase ACE2 also cuts other peptide substrates inside and outside of the RAS with a proline residue at the penultimate position, such as Ang I to Ang-(1-9) and Ang-A to alamandine¹² as well as des-Arg⁹-bradykinin, des-Arg¹⁰-kallidin, neurotensin, opioid peptides and apelins⁹⁸. The ACE2 gene is located on the X-chromosome and encodes an 805 amino acid type-1 transmembrane protein with a molecular weight of ~120 kDa. It consists of two domains with independent functions which probably arose from a gene fusion event in evolution⁹⁹ (Figure 4). ACE2’s amino-terminal domain is homologous to ACE and contains the peptidase function, while its carboxy-terminal domain consisting of the cytoplasmic, transmembrane, and juxtamembrane parts has homology to collectrin, a protein essential for amino acid absorption in the kidney¹⁰⁰. Indeed, ACE2 is also essential for the uptake of certain amino acids, however not in the kidney but in the gut¹⁰¹. ACE2 chaperones the amino acid transporters B(0)AT1 (SLC6A19) and SIT1 (SLC6A20) to the plasma membrane of gut epithelial cells¹⁰¹. When ACE2 is absent, these transporters disappear leading to significant reductions in circulating tryptophan and consequently in its metabolite serotonin in blood and brain¹⁰². Recently, a second isoform of ACE2 has been described being transcribed from an interferon-inducible promoter in intron 9 of the ACE2 gene¹⁰³. However, the resulting protein does not contain the enzymatically active and the virus-binding sites and may only be relevant for the amino acid transport function of ACE2, albeit it is probably not even membrane-bound since it also lacks the signal peptide. Thus, the physiological role of this isoform remains enigmatic.

Figure 4. Structure of human ACE2.

ACE2 is a dimeric type 1 transmembrane protein consisting of a carboxypeptidase and a collectrin-like domain, which can be cleaved by sheddases such as TMPRSS2, MT1-MMP, and ADAM17^{103,110,111,128}. The signal peptide (SP) is released during synthesis and the mature protein starts with residue 18, a glutamine which is cyclized forming pyroglutamate¹²⁹. Transmembrane domain (TM), Dimerization domain (DD), Active site (AS), Sheddase cleavage sites (CS), Binding sites for the spike proteins of coronaviruses (V).

ACE2 is widely expressed with highest levels in the gut, followed depending on the species by testis, kidney, bladder, lung, liver, heart, blood vessels, pancreas, adipose tissue, skin, and brain¹⁰⁴. In blood vessels and in the heart, most single-cell RNA-sequencing data show very low or absent ACE2 expression on endothelial cells and cardiomyocytes, respectively, and instead find it in pericytes^105,106.

Size-exclusion chromatography and cryoelectron microscopy showed that ACE2 forms a dimer^107,108, but there are also reports on it being a monomer in the plasma membrane¹⁰⁹. In its juxtamembrane region there are cutting sites for sheddases, such as ADAM17, MT1-MMP, and TMPRSS2, which release the catalytically active ectodomain, called soluble ACE2 (sACE2)^110,111. The function of sACE2 is still obscure and in particular in COVID-19 it is very controversial, with some groups postulating it to be the main mediator of virus uptake by binding to AT₁R¹¹¹ and others assigning this role solely to the membrane-bound form¹¹². Nevertheless, the blood levels of sACE2 became a biomarker for severe COVID-19¹¹³ and other cardiovascular and inflammatory diseases¹¹⁴ probably because its shedding is mediated by inflammation. The resulting loss of ACE2 activity in tissues may also contribute to the severity of the inflammatory diseases by impairing the local degradation capabilities for proinflammatory factors, such as Ang II and des-Arg⁹-bradykinin. Moreover, autoantibodies against ACE2 appear in the blood of COVID-19 patients and may also contribute to the acute and chronic consequences of the infection¹¹⁵.

Most functions of ACE2 have been discovered using genetically modified animal models (for review see³⁷). Several groups have generated ACE2-deficient mice on different genetic backgrounds leading to controversies concerning the cardiovascular effects of ACE2 deletion. Some groups reported heart dysfunction or high blood pressure while others could not confirm these phenotypes^37,116. Increased blood pressure in the absence of ACE2 may be caused by endothelial dysfunction with reduced NO and increased reactive oxygen species generation¹¹⁶. Accordingly, spontaneously hypertensive stroke-prone rats (SHRSP) express low levels of ACE2 and their hypertensive phenotype could be ameliorated by overexpression of human ACE2 in the vasculature leading to reduced oxidative stress and improved endothelial function¹¹⁷. ACE2 expression in the brain may also be involved in its cardiovascular actions, since ACE2 deletion increased oxidative stress in the brain, and activated the sympathetic nervous system³⁷. Accordingly, transgenic animals overexpressing human ACE2 in neurons were protected from hypertension induced by peripheral and central infusions of Ang II and DOCA-salt treatment³⁷. Additionally, increased central ACE2 overexpression ameliorated cardiac hypertrophy induced by Ang II, heart failure induced by coronary ligation, and stroke induced by middle cerebral artery occlusion³⁷. Mice which ubiquitously overexpress ACE2 in all tissues exhibited reduced infarct size and preserved cardiac function in a myocardial infarction model induced by coronary artery ligation. Overexpression of human ACE2 in podocytes or vascular smooth muscle ameliorated diabetes- or ageing-induced nephropathy^37,118, respectively. Conversely, ACE2 deletion resulted in increased pressure overload-, obesity-, diabetes-, or infarct-induced cardiac injury and in diabetic, obstructive, hypertensive, and shock-induced kidney injury³⁷. Moreover, ACE2 deletion aggravated plaque formation, vascular inflammation, and aneurysm formation in mouse models for atherosclerosis³⁷. With respect to these multiple reports on protective functions of ACE2 it was puzzling that transgenic mice overexpressing human ACE2 in the heart developed ventricular tachycardia and sudden death due to dysregulated connexins¹¹⁹. Most likely, the pleiotropism of ACE2 explains this conundrum and not the alternative RAS is playing a major role in this surprising phenotype, since simply overexpressing Ang-(1-7) in the heart was cardioprotective^120,121.

ACE2 is also involved in the regulation of metabolism and obesity, however, again the findings are puzzling. While ACE2-deficient mice exhibit impaired glucose homeostasis, reduced β-cell mass in the pancreas, disturbed thermogenesis¹²², and aggravated liver fibrosis and steatosis, some groups report a lower body weight under chow diet and decreased weight gain after high-fat diet in these mice³⁷. Again, the pleiotropic actions of ACE2 altering the activity of the RAS and the serotonin system may explain these seemingly paradoxical results since reduced serotonin in the periphery is known to protect from high-fat diet induced obesity¹²³.

Not only since COVID-19, ACE2 has been characterized as a major protective enzyme in inflammatory lung diseases³⁷. ACE2 deficient mice showed aggravated pathologies in the acute respiratory distress syndrome (ARDS), in bleomycin-induced lung injury, pulmonary hypertension, and several viral-induced lung injuries. Zhang et al.¹²⁴ discovered that phosphorylation of ACE2 at position 680 ameliorates pulmonary hypertension by reducing ubiquitination of the protein, which has been well established as destabilizing mechanism for ACE2¹²⁵. Nevertheless, this was a surprising finding since serine at position 680 is thought to be located extracellularly (Figure 4) outside the reach of cytosolic kinases.

Most of the described phenotypes of ACE2-deficient mice are probably based on the peptidase function of the enzyme and involve the shifting of the balance between the classical and the alternative RAS caused by the lack of ACE2, since they are phenocopied in Mas-deficient animals and opposing effects are observed in animals overexpressing Ang-(1-7)³⁷, but other peptide substrates are not excluded. Therefore, therapeutic applications of ACE2 aim on stimulating its enzymatic activity on angiotensin peptides. However, while very efficient inhibitors, such as MLN-4760 and DX600³⁷, have been described, the postulated small-molecule activators, XNT and DIZE, are controversially discussed and have probably additional effects¹²⁶. Therefore, at the moment, the only approved way to increase its activity in a patient is the application of recombinant sACE2, which was originally developed as drug to treat ARDS and COVID-19¹²⁷. However, besides one case study¹²⁷, no reports on positive clinical trials have been published. One reason may be that recombinant sACE2 does not reach the relevant tissue sites for acting on the local RAS and therefore smaller forms of the enzyme are being developed⁶¹.

Conclusions and Future Directions

In summary, the lessons learned from the study of Ang-(1-7), alamandine, and AT₂R as principal effectors counteracting Ang II excesses in regulating tissue perfusion and cellular health have expanded knowledge of how the RAS functions in health and disease. Activating AT₂R, increasing the ACE2-dependent production or the Mas-dependent action of Ang-(1-7) represent novel therapeutic opportunities for a plethora of diseases. As outlined above numerous clinical trials are already in progress testing such novel therapeutics which exploit the beneficial actions of the alternative RAS. We believe that in the near future we will see the first compounds used in clinics activating ACE2, Mas, or activating or blocking the AT₂R. In recounting how knowledge of the ACE2/Ang-(1-7)/Mas axis has evolved in celebrating this centennial, the lessons learned illustrate the fine tuning of the homeostatic mechanisms regulating tissue perfusion and the function of Claude Bernard’s milieu intérieur.