Introduction

Angiotensin-(1–7) [Ang-(1–7)] was discovered 1988 as a product of angiotensin I (Ang I) degradation by enzymes from the brainstem [69]. Angiotensin-converting enzyme (ACE), which classically generates Ang II from Ang I, was shown to degrade Ang-(1–7) into inactive peptides, in particular Ang-(1–5) [86] (Fig. 1). Ang-(1–7) binds and activates the Ang II AT1 receptor only at supraphysiological concentrations, and an interaction with the AT2 receptor has been shown but is still controversial [46, 90]. Thus, this heptapeptide represented the first molecule of a novel axis of the renin–angiotensin system (RAS). However, before the ACE homologue ACE2 and Mas as receptor were described, it was not clear how the peptide was generated and by which pathways it signals, not to speak about its physiological functions. This review tries to give a timely appraisal of the large body of evidence suggesting that the new RAS axis, ACE2/Ang-(1–7)/Mas, is important for cardiovascular physiology and beyond.

Fig. 1
figure 1

Metabolic pathways of angiotensin peptides. Angiotensin peptides are metabolized by several subsequent enzymatic steps: First, renin cleaves angiotensinogen, into angiotensin I (Ang I). Ang I can be metabolized by angiotensin-converting enzyme (ACE) resulting in the production of the bioactive octapeptide angiotensin II (Ang II), which interacts with AT1 and AT2 receptors. Alternatively, it can be processed first by ACE2 to the inactive peptide Ang-(1–9) and then by ACE to Ang-(1–7) or by neutral endopeptidase 24.11 (NEP) or prolylendopeptidase (PEP) directly to Ang-(1–7). Ang-(1–7) can also be generated by ACE2 from Ang II and interacts with its receptor Mas

Full size image

Ang-(1–7)-generating enzymes

Ang-(1–7) can be generated from Ang I or II. Neprilysin (also known as neutral endopeptidase 24.11), thimet oligopeptidase, or prolylendopeptidase release the last three amino acids from Ang I [93] and ACE2 [16, 84] or prolylcarboxypeptidase (PRCP) [78] remove the C-terminal phenylalanine from Ang II, all liberating the heptapeptide (Fig. 1). ACE2 can also first generate Ang-(1–9) from Ang I followed by the action of ACE which releases the last two amino acids. It has to be noted that all these enzymes are not specific for the angiotensin peptides. For example, ACE2 also metabolizes kinins, apelins, and neurotensin [88], and PRCP degrades kinins, alpha melanocyte-stimulating hormone [89] and activates plasma prekallikrein [102]. Furthermore, ACE2 and its homologue collectrin are involved in amino acid transport in the kidney and gut and have, thus, additional functions beyond proteolysis [43]. Importantly, ACE2 and PRCP (also called angiotensinase C) not only generate Ang-(1–7) but at the same time degrade the potent effector peptide of the RAS, Ang II. Therefore, any alteration in expression or activity of these enzymes switches the net effect of the RAS between the two sides of the coin, leave alone their other functions. The relative contribution of the different Ang-(1–7)-generating enzymes may vary from tissue to tissue [18, 86]. Nevertheless, in the brain and kidney, Ang-(1–7) may even be the major product of angiotensin metabolism [18, 86].

In the following chapters, I will reduce the enzymes generating Ang-(1–7) to ACE2, since most studies have been done with this protein, but it is by no mean excluded that other enzymes, in particular PRCP, may be even more important in certain tissues or situations. A major problem to distinguish the enzymes and their functional relevance is the lack of specific inhibitors.

Mas

In 2003, Santos et al. finally solved two problems, by describing the long-sought receptor for Ang-(1–7) and deorphanizing Mas [74]. Before this discovery of Mas as Ang-(1–7) receptor, it had been shown to be a receptor for Ang II [41], but this was soon unveiled to be an artifact of the frog oocytes used. Probably the known interactions between Mas and the AT1 receptor [9, 42, 73] had changed the response of the oocytes to Ang II when Mas was expressed [2]. Originally, Mas had been discovered as proto-oncogene, but also the transforming activity could later not be confirmed and may have been due to a unique genetic rearrangement in the transformed cells which may have affected the imprinting of the neighboring genes [1, 53]. Alternatively, these cells may have dramatically overexpressed Mas and then its described ligand-independent activity may have transformed them [99]. However, this issue has never been totally solved. Interestingly, Mas agonism is now even discussed as antitumor strategy [29].

Mas is predominantly expressed in brain and testis, but was also detected in the kidney, heart, and vessels [56]. Mas belongs to the family of G protein-coupled receptors with seven transmembrane domains; however, its G protein coupling is still debated. At high concentrations in cells using synthetic agonists and antagonists, a coupling to Gq proteins has been described [99]. However, Ang-(1–7) does not elicit such a response. At more physiological receptor concentrations and with Ang-(1–7) as ligand, Mas induces arachidonic acid release from cells and intracellular Akt phosphorylation [67, 74]. A recent phosphoproteomic study identified several signaling pathways induced by Ang-(1–7) in human endothelial cells again including Akt [87]. However, the complete signaling pathways employed by Mas still await clarification. It is even not clear whether Ang-(1–7) has more receptors or whether Mas binds other natural ligands.

ACE2/Ang-(1–7)/Mas in vessels

Ang-(1–7) has been reported to be vasodilatory [47, 63], antithrombotic [27, 44], and antiproliferative [45, 79]. Most of these actions are mediated by changes in the redox balance in the vascular wall initiated by Ang-(1–7) via Mas [96]. Ang-(1–7) triggers NO release by Akt phosphorylation inducing the activation of endothelial NO synthase and inhibits Ang II-induced reactive oxygen species (ROS) production in endothelial cells [66, 67]. Accordingly, vessels of Mas-deficient mice produce more ROS and less NO leading to an impaired in vivo endothelial function and increased blood pressure [62, 96]. In the opposite, an improved endothelial function was observed in stroke-prone spontaneously hypertensive rats (SHRSP) expressing a human ACE2 transgene in vascular smooth muscle cells [64]. Increased Ang-(1–7) generation in the vascular wall appears to be the main mediator of this effect since also an improved endothelial function in the renal artery could be elicited in diabetic SHR by chronic treatment with Ang-(1–7) again by the reduction of oxidative stress [5].

Furthermore, activation of bradykinin signaling [49, 60] and attenuation of Ang II actions have been implicated in the vasculoprotective actions of Ang-(1–7). The known interactions between Mas and the AT1 receptor may have contributed to the latter effects [9, 42, 73].

In atherosclerosis, the ACE2/Ang-(1–7)/Mas axis was also shown to be protective. The genetic ablation of ACE2 significantly increases [81, 83], and transgenic vascular ACE2 overexpression decreases [51, 100] plaque formation in atherosclerotic apolipoprotein E or LDL receptor-deficient mice. In one study, the transfer of ACE2-deficient bone marrow into LDL receptor-deficient mice was already sufficient to aggravate plaque formation indicating that the enzyme on leukocytes is particularly beneficial in the atherosclerotic process [82]. Moreover, long-term Ang-(1–7) treatment induces protective effects in such animals [80]. In these cases again, an improvement of the redox balance by Ang-(1–7) has been reported to be pivotal for the anti-atherogenic effect.

ACE2/Ang-(1–7)/Mas in kidney

ACE2 and Mas are expressed in the kidney. ACE2 was found in endothelial cells of vessels, but the mesangium and glomerular endothelium were negative for ACE2. However, ACE2 is most highly expressed in the brush border of proximal tubular cells, while epithelial cells from other parts of the nephron showed weak cytoplasmic staining [36]. Mas is localized to the proximal and distal tubules, but also found in the glomerulus [11]. The majority of studies describe ACE2, Ang-(1–7), and Mas as protective factors in different kidney diseases [22]: Ang II-induced kidney damage and diabetic nephropathy are aggravated in ACE2-deficient mice, and the Ang II effects are ameliorated by recombinant ACE2 in wild-type animals [94, 101]. Ang-(1–7) infusion reverts diabetic renal damage in mice and rats [5, 33, 57], and Mas agonists protect the kidney from ischemia/reperfusion damage [3]. The mechanism involved in most of these cases seems to be a reduction in oxidative stress and reduced fibrosis by the components of the ACE2/Ang-(1–7)/Mas axis. Accordingly, Mas-deficient mice develop a spontaneous nephropathy with microalbuminuria [61]. Nevertheless, there have also been reports about aggravation of renal damage by Ang-(1–7) and Mas [7, 19]. The reasons for this discrepancy are not yet clarified.

Ang-(1–7) has also been shown to be involved in the normal function of the kidney by influencing sodium reabsorption. The effects of the peptide seem to be biphasic with an antidiuretic action at low concentration and diuretic effects at high levels [22, 30].

ACE2/Ang-(1–7)/Mas in heart

In the heart, ACE2 is mainly localized to the vascular endothelium and smooth muscle but was also detected in cardiomyocytes [8], and Mas was mainly described on cardiomyocytes [70]. Mice deficient for Mas and ACE2 show a reduced cardiac contractile function aggravating with age [10, 70]. Infusion of Ang-(1–7) rescues this phenotype in mice lacking ACE2 indicating that the loss of this peptide is an important component of the pathophysiology [59]. Diabetic cardiomyopathy is also exacerbated in the absence of ACE2, again based on increased oxidative stress [58]. In the opposite, local overexpression of ACE2 in the heart by lentiviral gene transfer elicited cardioprotective actions in several disease models [13, 15, 37]. A cardioprotective role of Ang-(1–7) was also observed in cardiac damage models, such as isoproterenol or Ang II-induced hypertrophy or ischemia/reperfusion injury, when the peptide was either infused or overexpressed from transgenes [4, 24, 35, 55, 72]. Besides NO-releasing, antioxidative, NO-increasing, and direct antihypertrophic effects on cardiomyocytes, the main actions of Ang-(1–7) in the heart seem to be the regulation of genes involved in fibrosis in cardiac fibroblasts via Mas [14, 31, 34, 40, 58, 59]. Accordingly, Mas agonists attenuate heart failure after myocardial infarction [50, 54].

ACE2/Ang-(1–7)/Mas in lung

ACE2 was found in type I and type II alveolar epithelial cells of normal lungs [36]. However, the cellular localization of Mas in the lung is not yet reported [85]. ACE2-deficient mice are more prone to lung injury in several disease models [38, 39], and recombinant ACE2 ameliorates the symptoms in the bleomycin-induced lung injury model [65]. Since also a lentivirally delivered Ang-(1–7) release construct has the same effect, the generation of this peptide by ACE2 seems to be of major importance [77] and not only the degradation of Ang II.

ACE2/Ang-(1–7)/Mas in brain

All components of the ACE2/Ang-(1–7)/Mas axis are expressed in the brain [56, 95, 97]. For Mas, the brain is even the organ with the highest expression, in particular in the hippocampus and the piriform cortex [56]. Therefore, it came to no surprise that Mas affects behavior and electrophysiology of the hippocampus [48, 91, 92]. Concerning cardiovascular actions, Ang-(1–7) and Mas have been shown to enhance baroreflex sensitivity and influence blood pressure in different directions depending on the brain area studied [12, 25, 95]. Local overexpression of ACE2 by viral transfection in the medulla of SHRSP resulted in a decrease in blood pressure [98]. When the same technology was applied in the subfornical organ of mice, a significant reduction in the pressor effect of infused Ang II was observed [21]. Local administration of Ang-(1–7) did not have the same effect, indicating that in this case, the degradation of Ang II may be a major action of ACE2. Accordingly, also transgenic mice in which ACE2 was targeted to the brain were protected from Ang II-induced neurogenic hypertension [20].

ACE2/Ang-(1–7)/Mas in metabolism

The role of Ang-(1–7) and Mas in metabolic regulation has become increasingly clear in recent years. We have demonstrated that Mas deficiency in mice induces a metabolic syndrome-like state, with dyslipidemia, lower glucose tolerance and insulin sensitivity, hyperinsulinemia, decreased glucose uptake in white adipose cells, and an increase in adipose tissue mass [71]. In accordance, chronically increased Ang-(1–7) levels in transgenic rats reduce the amount of fat tissue and plasma lipid levels and enhance glucose tolerance and insulin sensitivity [68]. Insulin sensitivity is increased by an enhancing effect on its intracellular signaling by Ang-(1–7) leading to an increased Akt phosphorylation and GLUT4 translocation to the plasma membrane in different tissues [32, 33]. ACE2 is also involved in the regulation of insulin secretion in the pancreas [6]. Taken together, these observations provide strong evidence that the components of the ACE2/Ang-(1–7)/Mas axis have an important role in metabolic regulation.

Therapeutic perspectives of the ACE2/Ang-(1–7)/Mas axis

Based on the mainly protective actions of the ACE2/Ang-(1–7)/Mas axis described in this review, first attempts are under way to exploit the novel branch of the RAS for therapeutic purposes [23]. At present, recombinant human ACE2 is clinically tested for the treatment of lung and heart diseases. Moreover, ACE2 activating substances have been discovered and may also be used for pharmacological applications in the near future [26]. The third group of substances interfering with the ACE2/Ang-(1–7)/Mas axis are Mas agonists. These include just the peptide Ang-(1–7) itself in oral formulations [52], chemically slightly changed versions including cyclic peptides [17], or peptides with different sequences [76]. Most of these substances have already shown beneficial effects in animal models of lung diseases, hypertension, and diabetes [26, 28, 54, 75]. However, the first clinical trials are still under way and will finally clarify whether the other side of the RAS coin can be a successful therapeutic target.